734 files changed, 361362 insertions, 4 deletions

diff --git a/COPYRIGHT.txt b/COPYRIGHT.txt
index 7d4b53a7cb..7bb65dced4 100644
--- a/COPYRIGHT.txt
+++ b/COPYRIGHT.txt
@@ -174,6 +174,33 @@ Copyright: 2015-2020 Google, Inc.
   2002, NVIDIA Corporation.
 License: glslang
 
+Files: ./thirdparty/graphite/
+Comment: Graphite engine
+Copyright: 2010, SIL International
+License: MPL-2.0
+
+Files: ./thirdparty/harfbuzz/
+Comment: HarfBuzz text shaping library
+Copyright: 2010,2011,2012,2013,2014,2015,2016,2017,2018,2019,2020 Google, Inc.
+ 2018,2019,2020 Ebrahim Byagowi
+ 2019,2020 Facebook, Inc.
+ 2012 Mozilla Foundation
+ 2011 Codethink Limited
+ 2008,2010 Nokia Corporation and/or its subsidiary(-ies)
+ 2009 Keith Stribley
+ 2009 Martin Hosken and SIL International
+ 2007 Chris Wilson
+ 2006 Behdad Esfahbod
+ 2005 David Turner
+ 2004,2007,2008,2009,2010 Red Hat, Inc.
+ 1998-2004 David Turner and Werner Lemberg
+License: HarfBuzz
+
+Files: ./thirdparty/icu4c/
+Comment: International Components for Unicode
+Copyright: 1991-2020, Unicode
+License: Unicode
+
 Files: ./thirdparty/jpeg_compressor/
 Comment: jpeg-compressor
 Copyright: 2012, Rich Geldreich
@@ -1180,6 +1207,46 @@ License: FTL
    Robert Wilhelm    <robert.wilhelm@freetype.org>
    Werner Lemberg    <werner.lemberg@freetype.org>
 
+License: HarfBuzz
+ HarfBuzz is licensed under the so-called "Old MIT" license.  Details follow.
+ For parts of HarfBuzz that are licensed under different licenses see individual
+ files names COPYING in subdirectories where applicable.
+ .
+ Copyright (C) 2010,2011,2012,2013,2014,2015,2016,2017,2018,2019,2020  Google, Inc.
+ Copyright (C) 2018,2019,2020  Ebrahim Byagowi
+ Copyright (C) 2019,2020  Facebook, Inc.
+ Copyright (C) 2012  Mozilla Foundation
+ Copyright (C) 2011  Codethink Limited
+ Copyright (C) 2008,2010  Nokia Corporation and/or its subsidiary(-ies)
+ Copyright (C) 2009  Keith Stribley
+ Copyright (C) 2009  Martin Hosken and SIL International
+ Copyright (C) 2007  Chris Wilson
+ Copyright (C) 2006  Behdad Esfahbod
+ Copyright (C) 2005  David Turner
+ Copyright (C) 2004,2007,2008,2009,2010  Red Hat, Inc.
+ Copyright (C) 1998-2004  David Turner and Werner Lemberg
+ .
+ For full copyright notices consult the individual files in the package.
+ .
+ .
+ Permission is hereby granted, without written agreement and without
+ license or royalty fees, to use, copy, modify, and distribute this
+ software and its documentation for any purpose, provided that the
+ above copyright notice and the following two paragraphs appear in
+ all copies of this software.
+ .
+ IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ DAMAGE.
+ .
+ THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+
 License: MPL-2.0
  Mozilla Public License Version 2.0
  ==================================
@@ -1650,6 +1717,41 @@ License: Tamsyn
  In no event will the author be held liable for damages arising from the use
  of this font.
 
+License: Unicode
+ COPYRIGHT AND PERMISSION NOTICE (ICU 58 and later)
+ .
+ Copyright (C) 1991-2020 Unicode, Inc. All rights reserved.
+ Distributed under the Terms of Use in https://www.unicode.org/copyright.html.
+ .
+ Permission is hereby granted, free of charge, to any person obtaining
+ a copy of the Unicode data files and any associated documentation
+ (the "Data Files") or Unicode software and any associated documentation
+ (the "Software") to deal in the Data Files or Software
+ without restriction, including without limitation the rights to use,
+ copy, modify, merge, publish, distribute, and/or sell copies of
+ the Data Files or Software, and to permit persons to whom the Data Files
+ or Software are furnished to do so, provided that either
+ (a) this copyright and permission notice appear with all copies
+ of the Data Files or Software, or
+ (b) this copyright and permission notice appear in associated
+ Documentation.
+ .
+ THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF
+ ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
+ WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ NONINFRINGEMENT OF THIRD PARTY RIGHTS.
+ IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS
+ NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL
+ DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+ PERFORMANCE OF THE DATA FILES OR SOFTWARE.
+ .
+ Except as contained in this notice, the name of a copyright holder
+ shall not be used in advertising or otherwise to promote the sale,
+ use or other dealings in these Data Files or Software without prior
+ written authorization of the copyright holder.
+
 License: Zlib
  This software is provided 'as-is', without any express or implied
  warranty.  In no event will the authors be held liable for any damages
diff --git a/SConstruct b/SConstruct
index d82ceb514b..10f961cff8 100644
--- a/SConstruct
+++ b/SConstruct
@@ -144,6 +144,9 @@ opts.Add(BoolVariable("builtin_certs", "Use the built-in SSL certificates bundle
 opts.Add(BoolVariable("builtin_enet", "Use the built-in ENet library", True))
 opts.Add(BoolVariable("builtin_freetype", "Use the built-in FreeType library", True))
 opts.Add(BoolVariable("builtin_glslang", "Use the built-in glslang library", True))
+opts.Add(BoolVariable("builtin_graphite", "Use the built-in Graphite library", True))
+opts.Add(BoolVariable("builtin_harfbuzz", "Use the built-in HarfBuzz library", True))
+opts.Add(BoolVariable("builtin_icu", "Use the built-in ICU library", True))
 opts.Add(BoolVariable("builtin_libogg", "Use the built-in libogg library", True))
 opts.Add(BoolVariable("builtin_libpng", "Use the built-in libpng library", True))
 opts.Add(BoolVariable("builtin_libtheora", "Use the built-in libtheora library", True))
diff --git a/modules/text_server_adv/SCsub b/modules/text_server_adv/SCsub
new file mode 100644
index 0000000000..819ae4f4da
--- /dev/null
+++ b/modules/text_server_adv/SCsub
@@ -0,0 +1,419 @@
+#!/usr/bin/env python
+
+Import("env")
+Import("env_modules")
+
+if env["builtin_harfbuzz"]:
+    env_harfbuzz = env_modules.Clone()
+
+    # Thirdparty source files
+    thirdparty_dir = "#thirdparty/harfbuzz/"
+    thirdparty_sources = [
+        "src/hb-aat-layout.cc",
+        "src/hb-aat-map.cc",
+        "src/hb-blob.cc",
+        "src/hb-buffer-serialize.cc",
+        "src/hb-buffer.cc",
+        "src/hb-common.cc",
+        #'src/hb-coretext.cc',
+        #'src/hb-directwrite.cc',
+        "src/hb-draw.cc",
+        "src/hb-face.cc",
+        "src/hb-fallback-shape.cc",
+        "src/hb-font.cc",
+        "src/hb-ft.cc",
+        #'src/hb-gdi.cc',
+        #'src/hb-glib.cc',
+        #'src/hb-gobject-structs.cc',
+        "src/hb-graphite2.cc",
+        "src/hb-icu.cc",
+        "src/hb-map.cc",
+        "src/hb-number.cc",
+        "src/hb-ot-cff1-table.cc",
+        "src/hb-ot-cff2-table.cc",
+        "src/hb-ot-color.cc",
+        "src/hb-ot-face.cc",
+        "src/hb-ot-font.cc",
+        "src/hb-ot-layout.cc",
+        "src/hb-ot-map.cc",
+        "src/hb-ot-math.cc",
+        "src/hb-ot-meta.cc",
+        "src/hb-ot-metrics.cc",
+        "src/hb-ot-name.cc",
+        "src/hb-ot-shape-complex-arabic.cc",
+        "src/hb-ot-shape-complex-default.cc",
+        "src/hb-ot-shape-complex-hangul.cc",
+        "src/hb-ot-shape-complex-hebrew.cc",
+        "src/hb-ot-shape-complex-indic-table.cc",
+        "src/hb-ot-shape-complex-indic.cc",
+        "src/hb-ot-shape-complex-khmer.cc",
+        "src/hb-ot-shape-complex-myanmar.cc",
+        "src/hb-ot-shape-complex-thai.cc",
+        "src/hb-ot-shape-complex-use-table.cc",
+        "src/hb-ot-shape-complex-use.cc",
+        "src/hb-ot-shape-complex-vowel-constraints.cc",
+        "src/hb-ot-shape-fallback.cc",
+        "src/hb-ot-shape-normalize.cc",
+        "src/hb-ot-shape.cc",
+        "src/hb-ot-tag.cc",
+        "src/hb-ot-var.cc",
+        "src/hb-set.cc",
+        "src/hb-shape-plan.cc",
+        "src/hb-shape.cc",
+        "src/hb-shaper.cc",
+        "src/hb-static.cc",
+        "src/hb-style.cc",
+        "src/hb-subset-cff-common.cc",
+        "src/hb-subset-cff1.cc",
+        "src/hb-subset-cff2.cc",
+        "src/hb-subset-input.cc",
+        "src/hb-subset-plan.cc",
+        "src/hb-subset.cc",
+        "src/hb-ucd.cc",
+        "src/hb-unicode.cc",
+        #'src/hb-uniscribe.cc'
+    ]
+    thirdparty_sources = [thirdparty_dir + file for file in thirdparty_sources]
+
+    if env["platform"] == "android" or env["platform"] == "linuxbsd" or env["platform"] == "server":
+        env_harfbuzz.Append(CCFLAGS=["-DHAVE_PTHREAD"])
+
+    if env["platform"] == "javascript":
+        if env["threads_enabled"]:
+            env_harfbuzz.Append(CCFLAGS=["-DHAVE_PTHREAD"])
+        else:
+            env_harfbuzz.Append(CCFLAGS=["-DHB_NO_MT"])
+
+    env_harfbuzz.Append(
+        CPPPATH=[
+            "#thirdparty/harfbuzz/src",
+            "#thirdparty/freetype/include",
+            "#thirdparty/graphite/include",
+            "#thirdparty/icu4c/common/",
+        ]
+    )
+    env_harfbuzz.Append(
+        CCFLAGS=["-DHAVE_ICU_BUILTIN", "-DHAVE_ICU", "-DHAVE_FREETYPE", "-DHAVE_GRAPHITE2", "-DGRAPHITE2_STATIC",]
+    )
+    env_harfbuzz.disable_warnings()
+    env_thirdparty = env_harfbuzz.Clone()
+    env_thirdparty.disable_warnings()
+    lib = env_thirdparty.add_library("harfbuzz_builtin", thirdparty_sources)
+
+    # Needs to be appended to arrive after libscene in the linker call,
+    # but we don't want it to arrive *after* system libs, so manual hack
+    # LIBS contains first SCons Library objects ("SCons.Node.FS.File object")
+    # and then plain strings for system library. We insert between the two.
+    inserted = False
+    for idx, linklib in enumerate(env["LIBS"]):
+        if isinstance(linklib, (str, bytes)):  # first system lib such as "X11", otherwise SCons lib object
+            env["LIBS"].insert(idx, lib)
+            inserted = True
+            break
+    if not inserted:
+        env.Append(LIBS=[lib])
+
+if env["builtin_graphite"]:
+    env_graphite = env_modules.Clone()
+
+    # Thirdparty source files
+    thirdparty_dir = "#thirdparty/graphite/"
+    thirdparty_sources = [
+        "src/gr_char_info.cpp",
+        "src/gr_face.cpp",
+        "src/gr_features.cpp",
+        "src/gr_font.cpp",
+        "src/gr_logging.cpp",
+        "src/gr_segment.cpp",
+        "src/gr_slot.cpp",
+        "src/CmapCache.cpp",
+        "src/Code.cpp",
+        "src/Collider.cpp",
+        "src/Decompressor.cpp",
+        "src/Face.cpp",
+        #'src/FileFace.cpp',
+        "src/FeatureMap.cpp",
+        "src/Font.cpp",
+        "src/GlyphCache.cpp",
+        "src/GlyphFace.cpp",
+        "src/Intervals.cpp",
+        "src/Justifier.cpp",
+        "src/NameTable.cpp",
+        "src/Pass.cpp",
+        "src/Position.cpp",
+        "src/Segment.cpp",
+        "src/Silf.cpp",
+        "src/Slot.cpp",
+        "src/Sparse.cpp",
+        "src/TtfUtil.cpp",
+        "src/UtfCodec.cpp",
+        "src/FileFace.cpp",
+        "src/json.cpp",
+    ]
+    if not env_graphite.msvc:
+        thirdparty_sources += ["src/direct_machine.cpp"]
+    else:
+        thirdparty_sources += ["src/call_machine.cpp"]
+
+    thirdparty_sources = [thirdparty_dir + file for file in thirdparty_sources]
+
+    env_graphite.Append(CPPPATH=["#thirdparty/graphite/src", "#thirdparty/graphite/include"])
+    env_graphite.Append(CCFLAGS=["-DGRAPHITE2_STATIC", "-DGRAPHITE2_NTRACING", "-DGRAPHITE2_NFILEFACE"])
+    env_graphite.disable_warnings()
+    env_thirdparty = env_graphite.Clone()
+    env_thirdparty.disable_warnings()
+    lib = env_thirdparty.add_library("graphite_builtin", thirdparty_sources)
+
+    # Needs to be appended to arrive after libscene in the linker call,
+    # but we don't want it to arrive *after* system libs, so manual hack
+    # LIBS contains first SCons Library objects ("SCons.Node.FS.File object")
+    # and then plain strings for system library. We insert between the two.
+    inserted = False
+    for idx, linklib in enumerate(env["LIBS"]):
+        if isinstance(linklib, (str, bytes)):  # first system lib such as "X11", otherwise SCons lib object
+            env["LIBS"].insert(idx, lib)
+            inserted = True
+            break
+    if not inserted:
+        env.Append(LIBS=[lib])
+
+if env["builtin_icu"]:
+    env_icu = env_modules.Clone()
+
+    # Thirdparty source files
+    thirdparty_dir = "#thirdparty/icu4c/"
+    # Thirdparty source files
+    thirdparty_sources = [
+        "common/appendable.cpp",
+        "common/bmpset.cpp",
+        "common/brkeng.cpp",
+        "common/brkiter.cpp",
+        "common/bytesinkutil.cpp",
+        "common/bytestream.cpp",
+        "common/bytestrie.cpp",
+        "common/bytestriebuilder.cpp",
+        "common/bytestrieiterator.cpp",
+        "common/caniter.cpp",
+        "common/characterproperties.cpp",
+        "common/chariter.cpp",
+        "common/charstr.cpp",
+        "common/cmemory.cpp",
+        "common/cstr.cpp",
+        "common/cstring.cpp",
+        "common/cwchar.cpp",
+        "common/dictbe.cpp",
+        "common/dictionarydata.cpp",
+        "common/dtintrv.cpp",
+        "common/edits.cpp",
+        "common/errorcode.cpp",
+        "common/filteredbrk.cpp",
+        "common/filterednormalizer2.cpp",
+        "common/icudataver.cpp",
+        "common/icuplug.cpp",
+        "common/loadednormalizer2impl.cpp",
+        "common/localebuilder.cpp",
+        "common/localematcher.cpp",
+        "common/localeprioritylist.cpp",
+        "common/locavailable.cpp",
+        "common/locbased.cpp",
+        "common/locdispnames.cpp",
+        "common/locdistance.cpp",
+        "common/locdspnm.cpp",
+        "common/locid.cpp",
+        "common/loclikely.cpp",
+        "common/loclikelysubtags.cpp",
+        "common/locmap.cpp",
+        "common/locresdata.cpp",
+        "common/locutil.cpp",
+        "common/lsr.cpp",
+        "common/messagepattern.cpp",
+        "common/normalizer2.cpp",
+        "common/normalizer2impl.cpp",
+        "common/normlzr.cpp",
+        "common/parsepos.cpp",
+        "common/patternprops.cpp",
+        "common/pluralmap.cpp",
+        "common/propname.cpp",
+        "common/propsvec.cpp",
+        "common/punycode.cpp",
+        "common/putil.cpp",
+        "common/rbbi.cpp",
+        "common/rbbi_cache.cpp",
+        "common/rbbidata.cpp",
+        "common/rbbinode.cpp",
+        "common/rbbirb.cpp",
+        "common/rbbiscan.cpp",
+        "common/rbbisetb.cpp",
+        "common/rbbistbl.cpp",
+        "common/rbbitblb.cpp",
+        "common/resbund.cpp",
+        "common/resbund_cnv.cpp",
+        "common/resource.cpp",
+        "common/restrace.cpp",
+        "common/ruleiter.cpp",
+        "common/schriter.cpp",
+        "common/serv.cpp",
+        "common/servlk.cpp",
+        "common/servlkf.cpp",
+        "common/servls.cpp",
+        "common/servnotf.cpp",
+        "common/servrbf.cpp",
+        "common/servslkf.cpp",
+        "common/sharedobject.cpp",
+        "common/simpleformatter.cpp",
+        "common/static_unicode_sets.cpp",
+        "common/stringpiece.cpp",
+        "common/stringtriebuilder.cpp",
+        "common/uarrsort.cpp",
+        "common/ubidi.cpp",
+        "common/ubidi_props.cpp",
+        "common/ubidiln.cpp",
+        "common/ubiditransform.cpp",
+        "common/ubidiwrt.cpp",
+        "common/ubrk.cpp",
+        "common/ucase.cpp",
+        "common/ucasemap.cpp",
+        "common/ucasemap_titlecase_brkiter.cpp",
+        "common/ucat.cpp",
+        "common/uchar.cpp",
+        "common/ucharstrie.cpp",
+        "common/ucharstriebuilder.cpp",
+        "common/ucharstrieiterator.cpp",
+        "common/uchriter.cpp",
+        "common/ucln_cmn.cpp",
+        "common/ucmndata.cpp",
+        "common/ucnv.cpp",
+        "common/ucnv2022.cpp",
+        "common/ucnv_bld.cpp",
+        "common/ucnv_cb.cpp",
+        "common/ucnv_cnv.cpp",
+        "common/ucnv_ct.cpp",
+        "common/ucnv_err.cpp",
+        "common/ucnv_ext.cpp",
+        "common/ucnv_io.cpp",
+        "common/ucnv_lmb.cpp",
+        "common/ucnv_set.cpp",
+        "common/ucnv_u16.cpp",
+        "common/ucnv_u32.cpp",
+        "common/ucnv_u7.cpp",
+        "common/ucnv_u8.cpp",
+        "common/ucnvbocu.cpp",
+        "common/ucnvdisp.cpp",
+        "common/ucnvhz.cpp",
+        "common/ucnvisci.cpp",
+        "common/ucnvlat1.cpp",
+        "common/ucnvmbcs.cpp",
+        "common/ucnvscsu.cpp",
+        "common/ucnvsel.cpp",
+        "common/ucol_swp.cpp",
+        "common/ucptrie.cpp",
+        "common/ucurr.cpp",
+        "common/udata.cpp",
+        "common/udatamem.cpp",
+        "common/udataswp.cpp",
+        "common/uenum.cpp",
+        "common/uhash.cpp",
+        "common/uhash_us.cpp",
+        "common/uidna.cpp",
+        "common/uinit.cpp",
+        "common/uinvchar.cpp",
+        "common/uiter.cpp",
+        "common/ulist.cpp",
+        "common/uloc.cpp",
+        "common/uloc_keytype.cpp",
+        "common/uloc_tag.cpp",
+        "common/umapfile.cpp",
+        "common/umath.cpp",
+        "common/umutablecptrie.cpp",
+        "common/umutex.cpp",
+        "common/unames.cpp",
+        "common/unifiedcache.cpp",
+        "common/unifilt.cpp",
+        "common/unifunct.cpp",
+        "common/uniset.cpp",
+        "common/uniset_closure.cpp",
+        "common/uniset_props.cpp",
+        "common/unisetspan.cpp",
+        "common/unistr.cpp",
+        "common/unistr_case.cpp",
+        "common/unistr_case_locale.cpp",
+        "common/unistr_cnv.cpp",
+        "common/unistr_props.cpp",
+        "common/unistr_titlecase_brkiter.cpp",
+        "common/unorm.cpp",
+        "common/unormcmp.cpp",
+        "common/uobject.cpp",
+        "common/uprops.cpp",
+        "common/ures_cnv.cpp",
+        "common/uresbund.cpp",
+        "common/uresdata.cpp",
+        "common/usc_impl.cpp",
+        "common/uscript.cpp",
+        "common/uscript_props.cpp",
+        "common/uset.cpp",
+        "common/uset_props.cpp",
+        "common/usetiter.cpp",
+        "common/ushape.cpp",
+        "common/usprep.cpp",
+        "common/ustack.cpp",
+        "common/ustr_cnv.cpp",
+        "common/ustr_titlecase_brkiter.cpp",
+        "common/ustr_wcs.cpp",
+        "common/ustrcase.cpp",
+        "common/ustrcase_locale.cpp",
+        "common/ustrenum.cpp",
+        "common/ustrfmt.cpp",
+        "common/ustring.cpp",
+        "common/ustrtrns.cpp",
+        "common/utext.cpp",
+        "common/utf_impl.cpp",
+        "common/util.cpp",
+        "common/util_props.cpp",
+        "common/utrace.cpp",
+        "common/utrie.cpp",
+        "common/utrie2.cpp",
+        "common/utrie2_builder.cpp",
+        "common/utrie_swap.cpp",
+        "common/uts46.cpp",
+        "common/utypes.cpp",
+        "common/uvector.cpp",
+        "common/uvectr32.cpp",
+        "common/uvectr64.cpp",
+        "common/wintz.cpp",
+    ]
+    thirdparty_sources = [thirdparty_dir + file for file in thirdparty_sources]
+    thirdparty_sources += ["icu_data/icudata_stub.cpp"]
+
+    env_icu.Append(CPPPATH=["#thirdparty/icu4c/common/"])
+    env_icu.Append(
+        CXXFLAGS=["-DU_STATIC_IMPLEMENTATION", "-DU_COMMON_IMPLEMENTATION", "-DPKGDATA_MODE=static",]
+    )
+
+    env_icu.disable_warnings()
+    env_thirdparty = env_icu.Clone()
+    env_thirdparty.disable_warnings()
+    lib = env_thirdparty.add_library("icu_builtin", thirdparty_sources)
+
+    # Needs to be appended to arrive after libscene in the linker call,
+    # but we don't want it to arrive *after* system libs, so manual hack
+    # LIBS contains first SCons Library objects ("SCons.Node.FS.File object")
+    # and then plain strings for system library. We insert between the two.
+    inserted = False
+    for idx, linklib in enumerate(env["LIBS"]):
+        if isinstance(linklib, (str, bytes)):  # first system lib such as "X11", otherwise SCons lib object
+            env["LIBS"].insert(idx, lib)
+            inserted = True
+            break
+    if not inserted:
+        env.Append(LIBS=[lib])
+
+env_text_server_adv = env_modules.Clone()
+env_text_server_adv.Append(
+    CPPPATH=[
+        "#thirdparty/harfbuzz/src",
+        "#thirdparty/freetype/include",
+        "#thirdparty/graphite/include",
+        "#thirdparty/icu4c/common/",
+    ]
+)
+env_text_server_adv.add_source_files(env.modules_sources, "*.cpp")
diff --git a/modules/text_server_adv/config.py b/modules/text_server_adv/config.py
new file mode 100644
index 0000000000..1c8cd12a2d
--- /dev/null
+++ b/modules/text_server_adv/config.py
@@ -0,0 +1,5 @@
+def can_build(env, platform):
+    return True
+
+def configure(env):
+    pass
diff --git a/modules/text_server_adv/icu_data/icudata_stub.cpp b/modules/text_server_adv/icu_data/icudata_stub.cpp
new file mode 100644
index 0000000000..13f0ac0c50
--- /dev/null
+++ b/modules/text_server_adv/icu_data/icudata_stub.cpp
@@ -0,0 +1,63 @@
+/*************************************************************************/
+/*  icudata_stub.cpp                                                     */
+/*************************************************************************/
+/*                       This file is part of:                           */
+/*                           GODOT ENGINE                                */
+/*                      https://godotengine.org                          */
+/*************************************************************************/
+/* Copyright (c) 2007-2020 Juan Linietsky, Ariel Manzur.                 */
+/* Copyright (c) 2014-2020 Godot Engine contributors (cf. AUTHORS.md).   */
+/*                                                                       */
+/* Permission is hereby granted, free of charge, to any person obtaining */
+/* a copy of this software and associated documentation files (the       */
+/* "Software"), to deal in the Software without restriction, including   */
+/* without limitation the rights to use, copy, modify, merge, publish,   */
+/* distribute, sublicense, and/or sell copies of the Software, and to    */
+/* permit persons to whom the Software is furnished to do so, subject to */
+/* the following conditions:                                             */
+/*                                                                       */
+/* The above copyright notice and this permission notice shall be        */
+/* included in all copies or substantial portions of the Software.       */
+/*                                                                       */
+/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,       */
+/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF    */
+/* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.*/
+/* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY  */
+/* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,  */
+/* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE     */
+/* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.                */
+/*************************************************************************/
+
+#include "unicode/udata.h"
+#include "unicode/utypes.h"
+#include "unicode/uversion.h"
+
+typedef struct {
+	uint16_t header_size;
+	uint8_t magic_1, magic_2;
+	UDataInfo info;
+	char padding[8];
+	uint32_t count, reserved;
+	int fake_name_and_data[4];
+} ICU_data_header;
+
+extern "C" U_EXPORT const ICU_data_header U_ICUDATA_ENTRY_POINT = {
+	32,
+	0xDA, 0x27,
+	{ sizeof(UDataInfo),
+			0,
+#if U_IS_BIG_ENDIAN
+			1,
+#else
+			0,
+#endif
+			U_CHARSET_FAMILY,
+			sizeof(UChar),
+			0,
+			{ 0x54, 0x6F, 0x43, 0x50 },
+			{ 1, 0, 0, 0 },
+			{ 0, 0, 0, 0 } },
+	{ 0, 0, 0, 0, 0, 0, 0, 0 },
+	0, 0,
+	{ 0, 0, 0, 0 }
+};
diff --git a/platform/android/detect.py b/platform/android/detect.py
index 0accacb679..60d4146712 100644
--- a/platform/android/detect.py
+++ b/platform/android/detect.py
@@ -215,7 +215,7 @@ def configure(env):
     env.Append(CPPFLAGS=["-isystem", env["ANDROID_NDK_ROOT"] + "/sources/cxx-stl/llvm-libc++abi/include"])
 
     # Disable exceptions and rtti on non-tools (template) builds
-    if env["tools"]:
+    if env["tools"] or env["builtin_icu"]:
         env.Append(CXXFLAGS=["-frtti"])
     else:
         env.Append(CXXFLAGS=["-fno-rtti", "-fno-exceptions"])
diff --git a/platform/javascript/detect.py b/platform/javascript/detect.py
index 71189cf697..9f584d0899 100644
--- a/platform/javascript/detect.py
+++ b/platform/javascript/detect.py
@@ -71,6 +71,8 @@ def configure(env):
             )
         # Tools need more memory. Initial stack memory in bytes. See `src/settings.js` in emscripten repository (will be renamed to INITIAL_MEMORY).
         env.Append(LINKFLAGS=["-s", "TOTAL_MEMORY=33554432"])
+    elif env["builtin_icu"]:
+        env.Append(CCFLAGS=["-frtti"])
     else:
         # Disable exceptions and rtti on non-tools (template) builds
         # These flags help keep the file size down.
diff --git a/platform/linuxbsd/detect.py b/platform/linuxbsd/detect.py
index a8bc3a09f6..277aafc107 100644
--- a/platform/linuxbsd/detect.py
+++ b/platform/linuxbsd/detect.py
@@ -205,14 +205,31 @@ def configure(env):
 
     # freetype depends on libpng and zlib, so bundling one of them while keeping others
     # as shared libraries leads to weird issues
-    if env["builtin_freetype"] or env["builtin_libpng"] or env["builtin_zlib"]:
+    if (
+        env["builtin_freetype"]
+        or env["builtin_libpng"]
+        or env["builtin_zlib"]
+        or env["builtin_graphite"]
+        or env["builtin_harfbuzz"]
+    ):
         env["builtin_freetype"] = True
         env["builtin_libpng"] = True
         env["builtin_zlib"] = True
+        env["builtin_graphite"] = True
+        env["builtin_harfbuzz"] = True
 
     if not env["builtin_freetype"]:
         env.ParseConfig("pkg-config freetype2 --cflags --libs")
 
+    if not env["builtin_graphite"]:
+        env.ParseConfig("pkg-config graphite2 --cflags --libs")
+
+    if not env["builtin_icu"]:
+        env.ParseConfig("pkg-config icu-uc --cflags --libs")
+
+    if not env["builtin_harfbuzz"]:
+        env.ParseConfig("pkg-config harfbuzz harfbuzz-icu --cflags --libs")
+
     if not env["builtin_libpng"]:
         env.ParseConfig("pkg-config libpng16 --cflags --libs")
 
diff --git a/platform/server/detect.py b/platform/server/detect.py
index d9ac357679..bf4744a234 100644
--- a/platform/server/detect.py
+++ b/platform/server/detect.py
@@ -138,14 +138,31 @@ def configure(env):
 
     # freetype depends on libpng and zlib, so bundling one of them while keeping others
     # as shared libraries leads to weird issues
-    if env["builtin_freetype"] or env["builtin_libpng"] or env["builtin_zlib"]:
+    if (
+        env["builtin_freetype"]
+        or env["builtin_libpng"]
+        or env["builtin_zlib"]
+        or env["builtin_graphite"]
+        or env["builtin_harfbuzz"]
+    ):
         env["builtin_freetype"] = True
         env["builtin_libpng"] = True
         env["builtin_zlib"] = True
+        env["builtin_graphite"] = True
+        env["builtin_harfbuzz"] = True
 
     if not env["builtin_freetype"]:
         env.ParseConfig("pkg-config freetype2 --cflags --libs")
 
+    if not env["builtin_graphite"]:
+        env.ParseConfig("pkg-config graphite2 --cflags --libs")
+
+    if not env["builtin_icu"]:
+        env.ParseConfig("pkg-config icu-uc --cflags --libs")
+
+    if not env["builtin_harfbuzz"]:
+        env.ParseConfig("pkg-config harfbuzz harfbuzz-icu --cflags --libs")
+
     if not env["builtin_libpng"]:
         env.ParseConfig("pkg-config libpng16 --cflags --libs")
 
@@ -233,7 +250,17 @@ def configure(env):
     env.Append(CPPDEFINES=["SERVER_ENABLED", "UNIX_ENABLED"])
 
     if platform.system() == "Darwin":
-        env.Append(LINKFLAGS=["-framework", "Cocoa", "-framework", "Carbon", "-lz", "-framework", "IOKit"])
+        env.Append(
+            LINKFLAGS=[
+                "-framework",
+                "Cocoa",
+                "-framework",
+                "Carbon",
+                "-lz",
+                "-framework",
+                "IOKit",
+            ]
+        )
 
     env.Append(LIBS=["pthread"])
 
diff --git a/thirdparty/README.md b/thirdparty/README.md
index 37518857c7..d011832210 100644
--- a/thirdparty/README.md
+++ b/thirdparty/README.md
@@ -184,6 +184,39 @@ Files extracted from upstream source:
 
 Patches in the `patches` directory should be re-applied after updates.
 
+## Graphite engine
+
+- Upstream: https://github.com/silnrsi/graphite
+- Version: 1.3.14
+- License: MPL-2.0
+
+Files extracted from upstream source:
+- the `include` folder
+- the `src` folder
+- `COPYING`, `ChangeLog`
+
+## HarfBuzz
+
+- Upstream: https://github.com/harfbuzz/harfbuzz
+- Version: 2.7.2
+- License: HarfBuzz
+
+Files extracted from upstream source:
+- the `src` folder
+- `AUTHORS`, `COPYING`, `NEWS`, `THANKS`
+
+## International Components for Unicode
+
+- Upstream: https://github.com/unicode-org/icu
+- Version: 68.1
+- License: Unicode
+
+Files extracted from upstream source:
+- the `common` folder
+- `APIChangeReport.md`, `LICENSE`
+
+Files generated from upstream source:
+- the `icudt68l.dat` built with the provided `godot_data.json` config file (see https://github.com/unicode-org/icu/blob/master/docs/userguide/icu_data/buildtool.md for instructions)
 
 ## jpeg-compressor
 
diff --git a/thirdparty/graphite/COPYING b/thirdparty/graphite/COPYING
new file mode 100644
index 0000000000..f6630af533
--- /dev/null
+++ b/thirdparty/graphite/COPYING
@@ -0,0 +1,26 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+    Alternatively, you may use this library under the terms of the Mozilla
+    Public License (http://mozilla.org/MPL) or under the GNU General Public
+    License, as published by the Free Sofware Foundation; either version
+    2 of the license or (at your option) any later version.
+*/
diff --git a/thirdparty/graphite/ChangeLog b/thirdparty/graphite/ChangeLog
new file mode 100644
index 0000000000..e36110e1c1
--- /dev/null
+++ b/thirdparty/graphite/ChangeLog
@@ -0,0 +1,238 @@
+1.3.14
+    . Bug fixes
+    . Allow features to be hidden (for aliases)
+    . Move to python3
+    . Rename doc files from .txt to .asc
+
+1.3.13
+    . Resolve minor spacing issue in rtl non-overlap kerning
+    . python3 for graphite.py
+    . Better fuzzing
+    . Better building on windows
+
+1.3.12
+    . Graphite no longer does dumb rendering for fonts with no smarts
+    . Segment caching code removed. Anything attempting to use the segment cache gets given a regular face instead
+    . Add libfuzzer support
+    . Builds now require C++11
+    . Improvements to Windows 64 bit builds
+    . Support different versions of python including 32 bit and python 3
+    . Various minor bug fixes
+
+1.3.11
+    . Fixes due to security review
+    . Minor collision avoidance fixes
+    . Fix LZ4 decompressor against high compression
+
+1.3.10
+    . Address floating point build parameters to give consistent positioning results across platforms
+    . Various bug fixes
+
+1.3.9
+    . Add Collision COLL_ISSPACE to allow for visible spaces in collision avoidance
+    . Add segment and pass direction information to tracing output
+    . Bug fix rule length testing in 32-bit
+    . Increase slanted margin distances for collision avoidance
+    . Change kerning algorithm to simple outline expansion. Seems to make no visible difference.
+    . Add trace2svg to test tools
+
+1.3.8
+    . Various bug fixes arising from fuzzing
+    . Fix regression that stopped piglatin from working
+    . Make collision avoidance kerning give more regular results
+    . Minor modification to clustering algorithm to handle variable width chars
+
+1.3.7
+    . Bug fixes
+    . Start to deprecate SegCache. This will be going away in a later release.
+
+1.3.6
+    . Bug fixes
+
+1.3.5
+    . Bug fixes
+        . Security bug fix
+        . Fix ARM misalignment problem
+        . Track latest cmake
+
+1.3.4
+    . Transition from Mercurial to Git
+    . Bug fixes
+        . Fix Collision Kerning ignoring some diacritics
+        . Handle pass bits 16-31 to speed up fonts with > 16 passes
+        . Various minor fuzz bug fixes
+        . Make Coverity happy
+        . Add GR_FALLTHROUGH macro for clang c++11
+
+1.3.3
+    . Slight speed up in Collision Avoidance
+    . Remove dead bidi code
+    . Bug fixes
+        . Between pass bidi reorderings and at the end
+        . Decompressor fuzz bugs
+        . Other fuzz bugs
+
+1.3.2
+    . Remove full bidi. All segments are assumed to be single directioned.
+    . Bug fixes:
+        . Decompressor corner cases
+        . Various fuzz bugs
+
+1.3.1
+    . Deprecation warning: Full bidi support is about to be deprecated. Make contact
+      if this impacts you.
+    . Change compression block format slightly to conform to LZ4
+    . Bug fixes:
+        . Handle mono direction text with diacritics consistently. Fonts
+          now see the direction they expect consistently and bidi now
+          gives expected results.
+        . Fixed lots of fuzz bugs
+        . Coverity cleanups
+        . Build now works for clang and/or asan and/or afl etc.
+
+1.3.0
+    . Add collision avoidance
+        . Shift Collider
+        . Kern Collider
+        . Octabox outlines and subboxes
+    . Add compressed Silf and Glat table support
+    . Bug fixes:
+        . Stop loops forming in the child, sibling tree
+        . Handle bidi mirroring correctly if no bidi occurring
+
+1.2.4
+    . Face failure now has error code reporting via debug logging
+        . can now call gr_start_logging(NULL, fname)
+        . gr2fonttest --alltrace added
+    . Format 14 table support
+        . Not done. To be handled entirely in the compiler
+    . Bidi support for Unicode 6.3 Isolating direction controls
+    . Fonts no longer require a glyf/loca table. In such cases the bounding box is always 0.
+    . Clang ASAN build support added for testing.
+    . Handle out of memory sanely.
+    . Documentation improvements
+    . Bug fixes:
+        . Enforce fonts having to store glyph attributes by monotonically increasing attribute number
+        . zeropadding was not getting called on feature tags
+        . automatic associations for unassociated characters
+        . use direct engine on Mac
+        . various extreme case reading 1 past the end errors fixed
+        . remove tabs from sources so that it becomes readable again
+
+1.2.3
+    . Bug fixes only:
+        . fix byte swapping when testing cmap subtable lengths
+        . work around armel compilation problems with conditional operators
+        . fix pseudoglyph support for advance and bbox
+
+1.2.2
+    . Add support for passKeySlot (makes Charis 2x faster) up to 32 passes
+    . Add telemetry output to json if enabled in build GRAPHITE2_TELEMETRY
+    . Shrink font memory footprint particularly in the fsm
+    . Add -S to comparerenderer
+    . Bug fixes:
+        . Fix shift.x being reversed for rtl text
+        . Fix faulty fallback justification
+        . Fix bad cmap handling
+        . Support compiling on old Solaris where bidi attributes clash with register names
+        . Follow the crowd in using Windows.h
+
+1.2.1
+    . Bug fixes:
+        . Allow glyph reattachment
+        . Allow signed glyph attributes
+        . Various build problems with MacOS, old gcc versions, etc.
+        . Various overrun read errors fixed
+
+1.2.0
+    . API Changes:
+        . Added Windows friendly gr_start_logging and gr_stop_logging, now per face
+        . Added gr_make_face_with_ops, gr_make_face_with_seg_cache_and_ops
+        . Added gr_make_font_with_ops
+        . Added gr_face_is_char_supported
+        . Added gr_face_info to give info to apps about face capabilities
+        . Deprecated gr_make_face, gr_make_face_with_seg_cache, gr_make_font_with_advance_fn
+        . Deprecated graphite_start_logging and graphite_stop_logging
+            . These functions are stubbed now and do nothing, but do compile and link.
+        . Bump API version to 3
+    . Add C# wrapper to contrib
+    . Handle justification information in a font and do something useful with it
+    . Builds clang clean (has done for a while)
+    . Bug fixes
+    . Windows build and bug fixes
+    . Add extra information to json debug output
+    . Added windows build documentation
+    . Added freetype sample code and test
+
+1.1.3
+    . Default build has GRAPHITE2_COMPARE_RENDERER to OFF to reduce dependencies
+    . Builds on Mac with clang
+    . Debug output improvements
+    . Tidy up perl wrappers
+    . Fuzz tester improvements
+    . Various bug fixes for bad font handling
+
+1.1.2
+    . Support feature ids < 4 chars when space padded for inclusion in FF 14.
+    . More fuzztesting and removal of causes of valgrind bad reads and sigabrts
+    . Remove contrib/android into its own repo (http://hg.palaso.org/grandroid)
+    . Update comparerenderer to latest harfbuzzng api
+
+1.1.1
+    . Missing Log.h included
+    . perl wrappers updated
+
+1.1.0
+    . Refactored debug output to use json
+    . Renamed VM_MACHINE_TYPE to GRAPHITE2_VM_TYPE
+    . Renamed DISABLE_SEGCACHE to GRAPHITE2_NSEGCACE
+    . Renamed DISBALE_FILE_FACE to GRAPHITE2_NFILEFACE
+    . Renamed ENABLE_COMPARE_RENDERER to GRAPHTIE2_COMPARE_RENDERER
+    . Renamed DOXYGEN_CONFIG to GRAPHITE2_DOXYGEN_CONFIG
+    . Renamed GR2_CUSTOM_HEADER to GRAPHITE2_CUSTOM_HEADER
+    . Renamed GR2_EXPORTING to GRAPHITE2_EXPORTING
+    . Added GRAPHITE2_STATIC for static only builds
+    . Added GRAPHITE2_NTRACING to compile out tracing code
+    . Documented GRAPHITE2_{EXPORTING,STATIC,NTRACING} in hacking.txt
+    . Bump libtool version to 2.1.0
+    . dumb font rendering works
+    . slot user attributes are now signed rather than unsigned
+    . add support for long class maps
+    . Rename perl library to avoid nameclash on Windows
+    . Various robustness fixes
+    . Moved internal .h files into src/inc
+    . Parallelise fuzztest
+    . General build improvements, particularly on Windows
+
+1.0.3
+    . Fix UTF16 surrogate support
+    . script and lang tags may be space padded or null padded
+    . Remove need for WORDS_BIGENDIAN, do it all automatically
+    . Remove all #include <new>. Use CLASS_NEW_DELETE instead.
+    . Fix comparerenderer to work with current hbng
+    . Add valgrind to fuzztest to ensure good memory use at all times
+    . Fix new fuzztest exposed bugs.
+    . Fix bugs exposed by Mozilla security review
+    . Add continuous integration build on Windows support
+
+1.0.2
+    . Fix Windows build
+    . Comparerenderer uses hbng enforcing ot rendering
+    . Add Bidi .hasChar support and refactor mirroring code
+    . Make cmake default Release rather than debug
+    . Don't compile in a boat load of TtfUtil that isn't used, saving 15% of binary
+    . Chase the FSF around its latest office moves
+    . WORDS_BIGENDIAN is set at the top so tests now pass on ppc, etc.
+    . More words in the manual
+
+1.0.1
+    . Release is the default build in cmake now.
+    . Refactor cmake build to not rebuild things so much.
+    . Include a missing file
+    . Remove -nostdlibs, making gcc happy everywhere
+    . Update comparerenderer to latest hbng interface
+    . Add changelog
+
+1.0.0
+    . First major release of perfect code!
+
diff --git a/thirdparty/graphite/include/graphite2/Font.h b/thirdparty/graphite/include/graphite2/Font.h
new file mode 100644
index 0000000000..fe569295a5
--- /dev/null
+++ b/thirdparty/graphite/include/graphite2/Font.h
@@ -0,0 +1,389 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+    Alternatively, the contents of this file may be used under the terms
+    of the Mozilla Public License (http://mozilla.org/MPL) or the GNU
+    General Public License, as published by the Free Software Foundation,
+    either version 2 of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "graphite2/Types.h"
+
+#define GR2_VERSION_MAJOR   1
+#define GR2_VERSION_MINOR   3
+#define GR2_VERSION_BUGFIX  14
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+typedef struct gr_face          gr_face;
+typedef struct gr_font          gr_font;
+typedef struct gr_feature_ref   gr_feature_ref;
+typedef struct gr_feature_val   gr_feature_val;
+
+/**
+* Returns version information on this engine
+*/
+GR2_API void gr_engine_version(int *nMajor, int *nMinor, int *nBugFix);
+
+/**
+* The Face Options allow the application to require that certain tables are
+* read during face construction. This may be of concern if the appFaceHandle
+* used in the gr_get_table_fn may change.
+* The values can be combined
+*/
+enum gr_face_options {
+    /** No preload, no cmap caching, fail if the graphite tables are invalid */
+    gr_face_default = 0,
+    /** Dumb rendering will be enabled if the graphite tables are invalid. @deprecated Since 1.311 */
+    gr_face_dumbRendering = 1,
+    /** preload glyphs at construction time */
+    gr_face_preloadGlyphs = 2,
+    /** Cache the lookup from code point to glyph ID at construction time */
+    gr_face_cacheCmap = 4,
+    /** Preload everything */
+    gr_face_preloadAll = gr_face_preloadGlyphs | gr_face_cacheCmap
+};
+
+/** Holds information about a particular Graphite silf table that has been loaded */
+struct gr_faceinfo {
+    gr_uint16 extra_ascent;     /**< The extra_ascent in the GDL, in design units */
+    gr_uint16 extra_descent;    /**< The extra_descent in the GDL, in design units */
+    gr_uint16 upem;             /**< The design units for the font */
+    enum gr_space_contextuals {
+        gr_space_unknown = 0,       /**< no information is known. */
+        gr_space_none = 1,          /**< the space character never occurs in any rules. */
+        gr_space_left_only = 2,     /**< the space character only occurs as the first element in a rule. */
+        gr_space_right_only = 3,    /**< the space character only occurs as the last element in a rule. */
+        gr_space_either_only = 4,   /**< the space character only occurs as the only element in a rule. */
+        gr_space_both = 5,          /**< the space character may occur as the first or last element of a rule. */
+        gr_space_cross = 6          /**< the space character occurs in a rule not as a first or last element. */
+    } space_contextuals;
+    unsigned int has_bidi_pass : 1; /**< the table specifies that a bidirectional pass should run */
+    unsigned int line_ends : 1;     /**< there are line end contextuals somewhere */
+    unsigned int justifies : 1;     /**< there are .justify properties set somewhere on some glyphs */
+};
+
+typedef struct gr_faceinfo gr_faceinfo;
+
+/** type describing function to retrieve font table information
+  *
+  * @return a pointer to the table in memory. The pointed to memory must exist as
+  *          long as the gr_face which makes the call.
+  * @param appFaceHandle is the unique information passed to gr_make_face()
+  * @param name is a 32bit tag to the table name.
+  * @param len returned by this function to say how long the table is in memory.
+  */
+typedef const void *(*gr_get_table_fn)(const void* appFaceHandle, unsigned int name, size_t *len);
+
+/** type describing function to release any resources allocated by the above get_table table function
+  *
+  * @param appFaceHandle is the unique information passed to gr_make_face()
+  * @param pointer to table memory returned by get_table.
+  */
+typedef void (*gr_release_table_fn)(const void* appFaceHandle, const void *table_buffer);
+
+/** struct housing function pointers to manage font table buffers for the graphite engine. */
+struct gr_face_ops
+{
+        /** size in bytes of this structure */
+    size_t              size;
+        /** a pointer to a function to request a table from the client. */
+	gr_get_table_fn 	get_table;
+        /** is a pointer to a function to notify the client the a table can be released.
+          * This can be NULL to signify that the client does not wish to do any release handling. */
+	gr_release_table_fn	release_table;
+};
+typedef struct gr_face_ops	gr_face_ops;
+
+/** Create a gr_face object given application information and a table functions.
+  *
+  * @return gr_face or NULL if the font fails to load for some reason.
+  * @param appFaceHandle This is application specific information that is passed
+  *                      to the getTable function. The appFaceHandle must stay
+  *                      alive as long as the gr_face is alive.
+  * @param face_ops      Pointer to face specific callback structure for table
+  *                      management. Must stay alive for the duration of the
+  *                      call only.
+  * @param faceOptions   Bitfield describing various options. See enum gr_face_options for details.
+  */
+GR2_API gr_face* gr_make_face_with_ops(const void* appFaceHandle/*non-NULL*/, const gr_face_ops *face_ops, unsigned int faceOptions);
+
+/** @deprecated Since v1.2.0 in favour of gr_make_face_with_ops.
+  * Create a gr_face object given application information and a getTable function.
+  *
+  * @return gr_face or NULL if the font fails to load for some reason.
+  * @param appFaceHandle This is application specific information that is passed
+  *                      to the getTable function. The appFaceHandle must stay
+  *                      alive as long as the gr_face is alive.
+  * @param getTable      Callback function to get table data.
+  * @param faceOptions   Bitfield describing various options. See enum gr_face_options for details.
+  */
+GR2_DEPRECATED_API gr_face* gr_make_face(const void* appFaceHandle/*non-NULL*/, gr_get_table_fn getTable, unsigned int faceOptions);
+
+/** @deprecated   Since 1.3.7 this function is now an alias for gr_make_face_with_ops().
+  *
+  * Create a gr_face object given application information, with subsegmental caching support
+  *
+  * @return gr_face or NULL if the font fails to load.
+  * @param appFaceHandle is a pointer to application specific information that is passed to getTable.
+  *                      This may not be NULL and must stay alive as long as the gr_face is alive.
+  * @param face_ops      Pointer to face specific callback structure for table management. Must stay
+  *                      alive for the duration of the call only.
+  * @param segCacheMaxSize Unused.
+  * @param faceOptions   Bitfield of values from enum gr_face_options
+  */
+GR2_DEPRECATED_API gr_face* gr_make_face_with_seg_cache_and_ops(const void* appFaceHandle, const gr_face_ops *face_ops, unsigned int segCacheMaxSize, unsigned int faceOptions);
+
+/** @deprecated   Since 1.3.7 this function is now an alias for gr_make_face().
+  *
+  * Create a gr_face object given application information, with subsegmental caching support.
+  * This function is deprecated as of v1.2.0 in favour of gr_make_face_with_seg_cache_and_ops.
+  *
+  * @return gr_face or NULL if the font fails to load.
+  * @param appFaceHandle is a pointer to application specific information that is passed to getTable.
+  *                      This may not be NULL and must stay alive as long as the gr_face is alive.
+  * @param getTable      The function graphite calls to access font table data
+  * @param segCacheMaxSize   How large the segment cache is.
+  * @param faceOptions   Bitfield of values from enum gr_face_options
+  */
+GR2_DEPRECATED_API gr_face* gr_make_face_with_seg_cache(const void* appFaceHandle, gr_get_table_fn getTable, unsigned int segCacheMaxSize, unsigned int faceOptions);
+
+/** Convert a tag in a string into a gr_uint32
+  *
+  * @return gr_uint32 tag, zero padded
+  * @param str a nul terminated string of which at most the first 4 characters are read
+  */
+GR2_API gr_uint32 gr_str_to_tag(const char *str);
+
+/** Convert a gr_uint32 tag into a string
+  *
+  * @param tag contains the tag to convert
+  * @param str is a pointer to a char array of at least size 4 bytes. The first 4 bytes of this array
+  *            will be overwritten by this function. No nul is appended.
+  */
+GR2_API void gr_tag_to_str(gr_uint32 tag, char *str);
+
+/** Get feature values for a given language or default
+  *
+  * @return a copy of the default feature values for a given language. The application must call
+  *          gr_featureval_destroy() to free this object when done.
+  * @param pFace The font face to get feature values from
+  * @param langname The language tag to get feature values for. If there is no such language or
+  *                  langname is 0, the default feature values for the font are returned.
+  *                  langname is right 0 padded and assumes lowercase. Thus the en langauge
+  *                  would be 0x656E0000. Langname may also be space padded, thus 0x656E2020.
+  */
+GR2_API gr_feature_val* gr_face_featureval_for_lang(const gr_face* pFace, gr_uint32 langname);
+
+/** Get feature reference for a given feature id from a face
+  *
+  * @return a feature reference corresponding to the given id. This data is part of the gr_face and
+  *          will be freed when the face is destroyed.
+  * @param pFace Font face to get information on.
+  * @param featId    Feature id tag to get reference to.
+  */
+GR2_API const gr_feature_ref* gr_face_find_fref(const gr_face* pFace, gr_uint32 featId);
+
+/** Returns number of feature references in a face **/
+GR2_API gr_uint16 gr_face_n_fref(const gr_face* pFace);
+
+/** Returns feature reference at given index in face **/
+GR2_API const gr_feature_ref* gr_face_fref(const gr_face* pFace, gr_uint16 i);
+
+/** Return number of languages the face knows about **/
+GR2_API unsigned short gr_face_n_languages(const gr_face* pFace);
+
+/** Returns a language id corresponding to a language of given index in the face **/
+GR2_API gr_uint32 gr_face_lang_by_index(const gr_face* pFace, gr_uint16 i);
+
+/** Destroy the given face and free its memory **/
+GR2_API void gr_face_destroy(gr_face *face);
+
+/** Returns the number of glyphs in the face **/
+GR2_API unsigned short gr_face_n_glyphs(const gr_face* pFace);
+
+/** Returns a faceinfo for the face and script **/
+GR2_API const gr_faceinfo *gr_face_info(const gr_face *pFace, gr_uint32 script);
+
+/** Returns whether the font supports a given Unicode character
+  *
+  * @return true if the character is supported.
+  * @param pFace    face to test within
+  * @param usv      Unicode Scalar Value of character to test
+  * @param script   Tag of script for selecting which set of pseudo glyphs to test. May be NULL.
+  */
+GR2_API int gr_face_is_char_supported(const gr_face *pFace, gr_uint32 usv, gr_uint32 script);
+
+#ifndef GRAPHITE2_NFILEFACE
+/** Create gr_face from a font file
+  *
+  * @return gr_face that accesses a font file directly. Returns NULL on failure.
+  * @param filename Full path and filename to font file
+  * @param faceOptions Bitfile from enum gr_face_options to control face options.
+  */
+GR2_API gr_face* gr_make_file_face(const char *filename, unsigned int faceOptions);
+
+/** @deprecated   Since 1.3.7. This function is now an alias for gr_make_file_face().
+  *
+  * Create gr_face from a font file, with subsegment caching support.
+  *
+  * @return gr_face that accesses a font file directly. Returns NULL on failure.
+  * @param filename Full path and filename to font file
+  * @param segCacheMaxSize Specifies how big to make the cache in segments.
+  * @param faceOptions   Bitfield from enum gr_face_options to control face options.
+  */
+GR2_DEPRECATED_API gr_face* gr_make_file_face_with_seg_cache(const char *filename, unsigned int segCacheMaxSize, unsigned int faceOptions);
+#endif      // !GRAPHITE2_NFILEFACE
+
+/** Create a font from a face
+  *
+  * @return gr_font Call font_destroy to free this font
+  * @param ppm Resolution of the font in pixels per em
+  * @param face Face this font corresponds to. This must stay alive as long as the font is alive.
+  */
+GR2_API gr_font* gr_make_font(float ppm, const gr_face *face);
+
+/** query function to find the hinted advance of a glyph
+  *
+  * @param appFontHandle is the unique information passed to gr_make_font_with_advance()
+  * @param glyphid is the glyph to retireve the hinted advance for.
+ */
+typedef float (*gr_advance_fn)(const void* appFontHandle, gr_uint16 glyphid);
+
+/** struct housing function pointers to manage font hinted metrics for the
+  * graphite engine. */
+struct gr_font_ops
+{
+        /** size of the structure in bytes to allow for future extensibility */
+    size_t              size;
+        /** a pointer to a function to retrieve the hinted
+          * advance width of a glyph which the font cannot
+          * provide without client assistance.  This can be
+          * NULL to signify no horizontal hinted metrics are necessary. */
+    gr_advance_fn       glyph_advance_x;
+        /** a pointer to a function to retrieve the hinted
+          * advance height of a glyph which the font cannot
+          * provide without client assistance.  This can be
+          * NULL to signify no horizontal hinted metrics are necessary. */
+    gr_advance_fn       glyph_advance_y;
+};
+typedef struct gr_font_ops  gr_font_ops;
+
+/** Creates a font with hinted advance width query functions
+  *
+  * @return gr_font to be destroyed via font_destroy
+  * @param ppm size of font in pixels per em
+  * @param appFontHandle font specific information that must stay alive as long
+  *        as the font does
+  * @param font_ops pointer font specific callback structure for hinted metrics.
+  *        Need only stay alive for the duration of the call.
+  * @param face the face this font corresponds to. Must stay alive as long as
+  *        the font does.
+  */
+GR2_API gr_font* gr_make_font_with_ops(float ppm, const void* appFontHandle, const gr_font_ops * font_ops, const gr_face *face);
+
+/** Creates a font with hinted advance width query function.
+  * This function is deprecated. Use gr_make_font_with_ops instead.
+  *
+  * @return gr_font to be destroyed via font_destroy
+  * @param ppm size of font in pixels per em
+  * @param appFontHandle font specific information that must stay alive as long
+  *        as the font does
+  * @param getAdvance callback function reference that returns horizontal advance in pixels for a glyph.
+  * @param face the face this font corresponds to. Must stay alive as long as
+  *        the font does.
+  */
+GR2_API gr_font* gr_make_font_with_advance_fn(float ppm, const void* appFontHandle, gr_advance_fn getAdvance, const gr_face *face);
+
+/** Free a font **/
+GR2_API void gr_font_destroy(gr_font *font);
+
+/** get a feature value
+  *
+  * @return value of specific feature or 0 if any problems.
+  * @param pfeatureref   gr_feature_ref to the feature
+  * @param feats gr_feature_val containing all the values
+  */
+GR2_API gr_uint16 gr_fref_feature_value(const gr_feature_ref* pfeatureref, const gr_feature_val* feats);
+
+/** set a feature value
+  *
+  * @return false if there were any problems (value out of range, etc.)
+  * @param pfeatureref   gr_feature_ref to the feature
+  * @param val   value to set the feature to
+  * @param pDest the gr_feature_val containing all the values for all the features
+  */
+GR2_API int gr_fref_set_feature_value(const gr_feature_ref* pfeatureref, gr_uint16 val, gr_feature_val* pDest);
+
+/** Returns the id tag for a gr_feature_ref **/
+GR2_API gr_uint32 gr_fref_id(const gr_feature_ref* pfeatureref);
+
+/** Returns number of values a feature may take, given a gr_feature_ref **/
+GR2_API gr_uint16 gr_fref_n_values(const gr_feature_ref* pfeatureref);
+
+/** Returns the value associated with a particular value in a feature
+  *
+  * @return value
+  * @param pfeatureref gr_feature_ref of the feature of interest
+  * @param settingno   Index up to the return value of gr_fref_n_values() of the value
+  */
+GR2_API gr_int16 gr_fref_value(const gr_feature_ref* pfeatureref, gr_uint16 settingno);
+
+/** Returns a string of the UI name of a feature
+  *
+  * @return string of the UI name, in the encoding form requested. Call gr_label_destroy() after use.
+  * @param pfeatureref   gr_feature_ref of the feature
+  * @param langId    This is a pointer since the face may not support a string in the requested
+  *                  language. The actual language of the string is returned in langId
+  * @param utf   Encoding form for the string
+  * @param length    Used to return the length of the string returned in bytes.
+  */
+GR2_API void* gr_fref_label(const gr_feature_ref* pfeatureref, gr_uint16 *langId, enum gr_encform utf, gr_uint32 *length);
+
+/** Return a UI string for a possible value of a feature
+  *
+  * @return string of the UI name, in the encoding form requested. nul terminated. Call gr_label_destroy()
+  *          after use.
+  * @param pfeatureref   gr_feature_ref of the feature
+  * @param settingno     Value setting index
+  * @param langId        This is a pointer to the requested language. The requested language id is
+  *                      replaced by the actual language id of the string returned.
+  * @param utf   Encoding form for the string
+  * @param length    Returns the length of the string returned in bytes.
+  */
+GR2_API void* gr_fref_value_label(const gr_feature_ref* pfeatureref, gr_uint16 settingno/*rather than a value*/, gr_uint16 *langId, enum gr_encform utf, gr_uint32 *length);
+
+/** Destroy a previously returned label string **/
+GR2_API void gr_label_destroy(void * label);
+
+/** Copies a gr_feature_val **/
+GR2_API gr_feature_val* gr_featureval_clone(const gr_feature_val* pfeatures);
+
+/** Destroys a gr_feature_val **/
+GR2_API void gr_featureval_destroy(gr_feature_val *pfeatures);
+
+#ifdef __cplusplus
+}
+#endif
diff --git a/thirdparty/graphite/include/graphite2/Log.h b/thirdparty/graphite/include/graphite2/Log.h
new file mode 100644
index 0000000000..a5a6947fab
--- /dev/null
+++ b/thirdparty/graphite/include/graphite2/Log.h
@@ -0,0 +1,85 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+    Alternatively, the contents of this file may be used under the terms
+    of the Mozilla Public License (http://mozilla.org/MPL) or the GNU
+    General Public License, as published by the Free Software Foundation,
+    either version 2 of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <graphite2/Types.h>
+#include <graphite2/Font.h>
+#include <stdio.h>
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+/** deprecated mechanism that doesn't do anything now. */
+typedef enum {
+    GRLOG_NONE = 0x0,
+    GRLOG_FACE = 0x01,
+    GRLOG_SEGMENT = 0x02,
+    GRLOG_PASS = 0x04,
+    GRLOG_CACHE = 0x08,
+
+    GRLOG_OPCODE = 0x80,
+    GRLOG_ALL = 0xFF
+} GrLogMask;
+
+/** Start logging all segment creation and updates on the provided face.  This
+  * is logged to a JSON file, see "Segment JSON Schema.txt" for a precise
+  * definition of the file
+  *
+  * @return true    if the file was successfully created and logging is correctly
+  * 			    initialised.
+  * @param face     the gr_face whose segments you want to log to the given file
+  * @param log_path a utf8 encoded file name and path to log to.
+  */
+GR2_API bool gr_start_logging(gr_face * face, const char *log_path);
+
+
+/** Stop logging on the given face.  This will close the log file created by
+  * gr_start_logging.
+  *
+  * @param face the gr_face whose segments you want to stop logging
+  */
+GR2_API void gr_stop_logging(gr_face * face);
+
+/** Start logging to a FILE object.
+  * This function is deprecated as of 1.2.0, use the _face versions instead.
+  *
+  * @return        True on success
+  * @param logfile FILE reference to output logging to
+  * @param mask    What aspects of logging to report (ignored)
+  */
+GR2_API bool graphite_start_logging(FILE * logFile, GrLogMask mask);    //may not do anthing if disabled in the implementation of the engine.
+
+/** Stop logging to a FILE object.
+  * This function is deprecated as of 1.2.0, use the _face versions instead.
+  */
+GR2_API void graphite_stop_logging();
+
+#ifdef __cplusplus
+}
+#endif
diff --git a/thirdparty/graphite/include/graphite2/Segment.h b/thirdparty/graphite/include/graphite2/Segment.h
new file mode 100644
index 0000000000..0e24f5d795
--- /dev/null
+++ b/thirdparty/graphite/include/graphite2/Segment.h
@@ -0,0 +1,461 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+    Alternatively, the contents of this file may be used under the terms
+    of the Mozilla Public License (http://mozilla.org/MPL) or the GNU
+    General Public License, as published by the Free Software Foundation,
+    either version 2 of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "graphite2/Types.h"
+#include "graphite2/Font.h"
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+enum gr_break_weight {
+    gr_breakNone = 0,
+    /* after break weights */
+    gr_breakWhitespace = 10,
+    gr_breakWord = 15,
+    gr_breakIntra = 20,
+    gr_breakLetter = 30,
+    gr_breakClip = 40,
+    /* before break weights */
+    gr_breakBeforeWhitespace = -10,
+    gr_breakBeforeWord = -15,
+    gr_breakBeforeIntra = -20,
+    gr_breakBeforeLetter = -30,
+    gr_breakBeforeClip = -40
+};
+
+enum gr_justFlags {
+    /// Indicates that this segment is a complete line
+    gr_justCompleteLine = 0,
+    /// Indicates that the start of the slot list is not at the start of a line
+    gr_justStartInline = 1,
+    /// Indicates that the end of the slot list is not at the end of a line
+    gr_justEndInline = 2
+};
+
+/** Used for looking up slot attributes. Most are already available in other functions **/
+enum gr_attrCode {
+    /// adjusted glyph advance in x direction in design units
+    gr_slatAdvX = 0,
+    /// adjusted glyph advance in y direction (usually 0) in design units
+    gr_slatAdvY,
+    /// returns 0. Deprecated.
+    gr_slatAttTo,
+    /// This slot attaches to its parent at the given design units in the x direction
+    gr_slatAttX,
+    /// This slot attaches to its parent at the given design units in the y direction
+    gr_slatAttY,
+    /// This slot attaches to its parent at the given glyph point (not implemented)
+    gr_slatAttGpt,
+    /// x-direction adjustment from the given glyph point (not implemented)
+    gr_slatAttXOff,
+    /// y-direction adjustment from the given glyph point (not implemented)
+    gr_slatAttYOff,
+    /// Where on this glyph should align with the attachment point on the parent glyph in the x-direction.
+    gr_slatAttWithX,
+    /// Where on this glyph should align with the attachment point on the parent glyph in the y-direction
+    gr_slatAttWithY,
+    /// Which glyph point on this glyph should align with the attachment point on the parent glyph (not implemented).
+    gr_slatWithGpt,
+    /// Adjustment to gr_slatWithGpt in x-direction (not implemented)
+    gr_slatAttWithXOff,
+    /// Adjustment to gr_slatWithGpt in y-direction (not implemented)
+    gr_slatAttWithYOff,
+    /// Attach at given nesting level (not implemented)
+    gr_slatAttLevel,
+    /// Line break breakweight for this glyph
+    gr_slatBreak,
+    /// Ligature component reference (not implemented)
+    gr_slatCompRef,
+    /// bidi directionality of this glyph (not implemented)
+    gr_slatDir,
+    /// Whether insertion is allowed before this glyph
+    gr_slatInsert,
+    /// Final positioned position of this glyph relative to its parent in x-direction in pixels
+    gr_slatPosX,
+    /// Final positioned position of this glyph relative to its parent in y-direction in pixels
+    gr_slatPosY,
+    /// Amount to shift glyph by in x-direction design units
+    gr_slatShiftX,
+    /// Amount to shift glyph by in y-direction design units
+    gr_slatShiftY,
+    /// attribute user1
+    gr_slatUserDefnV1,
+    /// not implemented
+    gr_slatMeasureSol,
+    /// not implemented
+    gr_slatMeasureEol,
+    /// Amount this slot can stretch (not implemented)
+    gr_slatJStretch,
+    /// Amount this slot can shrink (not implemented)
+    gr_slatJShrink,
+    /// Granularity by which this slot can stretch or shrink (not implemented)
+    gr_slatJStep,
+    /// Justification weight for this glyph (not implemented)
+    gr_slatJWeight,
+    /// Amount this slot mush shrink or stretch in design units
+    gr_slatJWidth = 29,
+    /// SubSegment split point
+    gr_slatSegSplit = gr_slatJStretch + 29,
+    /// User defined attribute, see subattr for user attr number
+    gr_slatUserDefn,
+    /// Bidi level
+    gr_slatBidiLevel = 56,
+    /// Collision flags
+    gr_slatColFlags,
+    /// Collision constraint rectangle left (bl.x)
+    gr_slatColLimitblx,
+    /// Collision constraint rectangle lower (bl.y)
+    gr_slatColLimitbly,
+    /// Collision constraint rectangle right (tr.x)
+    gr_slatColLimittrx,
+    /// Collision constraint rectangle upper (tr.y)
+    gr_slatColLimittry,
+    /// Collision shift x
+    gr_slatColShiftx,
+    /// Collision shift y
+    gr_slatColShifty,
+    /// Collision margin
+    gr_slatColMargin,
+    /// Margin cost weight
+    gr_slatColMarginWt,
+    // Additional glyph that excludes movement near this one:
+    gr_slatColExclGlyph,
+    gr_slatColExclOffx,
+    gr_slatColExclOffy,
+    // Collision sequence enforcing attributes:
+    gr_slatSeqClass,
+    gr_slatSeqProxClass,
+    gr_slatSeqOrder,
+    gr_slatSeqAboveXoff,
+    gr_slatSeqAboveWt,
+    gr_slatSeqBelowXlim,
+    gr_slatSeqBelowWt,
+    gr_slatSeqValignHt,
+    gr_slatSeqValignWt,
+
+    /// not implemented
+    gr_slatMax,
+    /// not implemented
+    gr_slatNoEffect = gr_slatMax + 1
+};
+
+enum gr_bidirtl {
+    /// Underlying paragraph direction is RTL
+    gr_rtl = 1,
+    /// Set this to not run the bidi pass internally, even if the font asks for it.
+    /// This presumes that the segment is in a single direction. Most of the time
+    /// this bit should be set unless you know you are passing full paragraphs of text.
+    gr_nobidi = 2,
+    /// Disable auto mirroring for rtl text
+    gr_nomirror = 4
+};
+
+typedef struct gr_char_info     gr_char_info;
+typedef struct gr_segment       gr_segment;
+typedef struct gr_slot          gr_slot;
+
+/** Returns Unicode character for a charinfo.
+  *
+  * @param p Pointer to charinfo to return information on.
+  */
+GR2_API unsigned int gr_cinfo_unicode_char(const gr_char_info* p/*not NULL*/);
+
+/** Returns breakweight for a charinfo.
+  *
+  * @return Breakweight is a number between -50 and 50 indicating the cost of a
+  * break before or after this character. If the value < 0, the absolute value
+  * is this character's contribution to the overall breakweight before it. If the value
+  * > 0, then the value is this character's contribution to the overall breakweight after it.
+  * The overall breakweight between two characters is the maximum of the breakweight
+  * contributions from the characters either side of it. If a character makes no
+  * contribution to the breakweight on one side of it, the contribution is considered
+  * to be 0.
+  * @param p Pointer to charinfo to return information on.
+  */
+GR2_API int gr_cinfo_break_weight(const gr_char_info* p/*not NULL*/);
+
+/** Returns the slot index that after this character is after in the slot stream
+  *
+  * In effect each character is associated with a set of slots and this returns
+  * the index of the last slot in the segment this character is associated with.
+  *
+  * @return after slot index between 0 and gr_seg_n_slots()
+  * @param p Pointer to charinfo to return information on.
+  */
+GR2_API int gr_cinfo_after(const gr_char_info* p/*not NULL*/);
+
+/** Returns the slot index that before this character is before in the slot stream
+  *
+  * In effect each character is associated with a set of slots and this returns
+  * the index of the first slot in the segment this character is associated with.
+  *
+  * @return before slot index between 0 and gr_seg_n_slots()
+  * @param p Pointer to charinfo to return information on.
+  */
+GR2_API int gr_cinfo_before(const gr_char_info* p/*not NULL*/);
+
+/** Returns the code unit index of this character in the input string
+  *
+  * @return code unit index between 0 and the end of the string
+  * @param p Pointer to charinfo to return information on.
+  */
+GR2_API size_t gr_cinfo_base(const gr_char_info* p/*not NULL*/);
+
+/** Returns the number of unicode characters in a string.
+  *
+  * @return number of characters in the string
+  * @param enc Specifies the type of data in the string: utf8, utf16, utf32
+  * @param buffer_begin The start of the string
+  * @param buffer_end Measure up to the first nul or when end is reached, whichever is earliest.
+  *            This parameter may be NULL.
+  * @param pError If there is a structural fault in the string, the location is returned
+  *               in this variable. If no error occurs, pError will contain NULL. NULL
+  *               may be passed for pError if no such information is required.
+  */
+GR2_API size_t gr_count_unicode_characters(enum gr_encform enc, const void* buffer_begin, const void* buffer_end, const void** pError);
+
+/** Creates and returns a segment.
+  *
+  * @return a segment that needs seg_destroy called on it. May return NULL if bad problems
+  *     in segment processing.
+  * @param font Gives the size of the font in pixels per em for final positioning. If
+  *             NULL, positions are returned in design units, i.e. at a ppm of the upem
+  *             of the face.
+  * @param face The face containing all the non-size dependent information.
+  * @param script This is a tag containing a script identifier that is used to choose
+  *               which graphite table within the font to use. Maybe 0. Tag may be 4 chars
+  *               NULL padded in LSBs or space padded in LSBs.
+  * @param pFeats Pointer to a feature values to be used for the segment. Only one
+  *               feature values may be used for a segment. If NULL the default features
+  *               for the font will be used.
+  * @param enc Specifies what encoding form the string is in (utf8, utf16, utf32)
+  * @param pStart Start of the string
+  * @param nChars Number of unicode characters to process in the string. The string will
+  *               be processed either up to the first NULL or until nChars have been
+  *               processed. nChars is also used to initialise the internal memory
+  *               allocations of the segment. So it is wise not to make nChars too much
+  *               greater than the actual number of characters being processed.
+  * @param dir Specifies whether the segment is processed right to left (1) or left to
+  *            right (0) and whether to run the internal bidi pass, if a font requests it.
+  *            See enum gr_bidirtl for details.
+  */
+GR2_API gr_segment* gr_make_seg(const gr_font* font, const gr_face* face, gr_uint32 script, const gr_feature_val* pFeats, enum gr_encform enc, const void* pStart, size_t nChars, int dir);
+
+/** Destroys a segment, freeing the memory.
+  *
+  * @param p The segment to destroy
+  */
+GR2_API void gr_seg_destroy(gr_segment* p);
+
+/** Returns the advance for the whole segment.
+  *
+  * Returns the width of the segment up to the next glyph origin after the segment
+  */
+GR2_API float gr_seg_advance_X(const gr_segment* pSeg/*not NULL*/);
+
+/** Returns the height advance for the segment. **/
+GR2_API float gr_seg_advance_Y(const gr_segment* pSeg/*not NULL*/);
+
+/** Returns the number of gr_char_infos in the segment. **/
+GR2_API unsigned int gr_seg_n_cinfo(const gr_segment* pSeg/*not NULL*/);
+
+/** Returns a gr_char_info at a given index in the segment. **/
+GR2_API const gr_char_info* gr_seg_cinfo(const gr_segment* pSeg/*not NULL*/, unsigned int index/*must be <number_of_CharInfo*/);
+
+/** Returns the number of glyph gr_slots in the segment. **/
+GR2_API unsigned int gr_seg_n_slots(const gr_segment* pSeg/*not NULL*/);      //one slot per glyph
+
+/** Returns the first gr_slot in the segment.
+  *
+  * The first slot in a segment has a gr_slot_prev_in_segment() of NULL. Slots are owned
+  * by their segment and are destroyed along with the segment.
+  */
+GR2_API const gr_slot* gr_seg_first_slot(gr_segment* pSeg/*not NULL*/);    //may give a base slot or a slot which is attached to another
+
+/** Returns the last gr_slot in the segment.
+  *
+  * The last slot in a segment has a gr_slot_next_in_segment() of NULL
+  */
+GR2_API const gr_slot* gr_seg_last_slot(gr_segment* pSeg/*not NULL*/);    //may give a base slot or a slot which is attached to another
+
+/** Justifies a linked list of slots for a line to a given width
+  *
+  * Passed a pointer to the start of a linked list of slots corresponding to a line, as
+  * set up by gr_slot_linebreak_before, this function will position the glyphs in the line
+  * to take up the given width. It is possible to specify a subrange within the line to process.
+  * This allows skipping of line initial or final whitespace, for example. While this will ensure
+  * that the subrange fits width, the line will still be positioned with the first glyph of the
+  * line at 0. So the resulting positions may be beyond width.
+  *
+  * @return float   The resulting width of the range of slots justified.
+  * @param pSeg     Pointer to the segment
+  * @param pStart   Pointer to the start of the line linked list (including skipped characters)
+  * @param pFont    Font to use for positioning
+  * @param width    Width in pixels in which to fit the line. If < 0. don't adjust natural width, just run justification passes
+  *                 to handle line end contextuals, if there are any.
+  * @param flags    Indicates line ending types. Default is linked list is a full line
+  * @param pFirst   If not NULL, the first slot in the list to be considered part of the line (so can skip)
+  * @param pLast    If not NULL, the last slot to process in the line (allow say trailing whitespace to be skipped)
+  */
+GR2_API float gr_seg_justify(gr_segment* pSeg/*not NULL*/, const gr_slot* pStart/*not NULL*/, const gr_font *pFont, double width, enum gr_justFlags flags, const gr_slot* pFirst, const gr_slot* pLast);
+
+/** Returns the next slot along in the segment.
+  *
+  * Slots are held in a linked list. This returns the next in the linked list. The slot
+  * may or may not be attached to another slot. Returns NULL at the end of the segment.
+  */
+GR2_API const gr_slot* gr_slot_next_in_segment(const gr_slot* p);
+
+/** Returns the previous slot along in the segment.
+  *
+  * Slots are held in a doubly linked list. This returns the previos slot in the linked
+  * list. This slot may or may not be attached to it. Returns NULL at the start of the
+  * segment.
+  */
+GR2_API const gr_slot* gr_slot_prev_in_segment(const gr_slot* p);
+
+/** Returns the attachment parent slot of this slot.
+  *
+  * Attached slots form a tree. This returns the parent of this slot in that tree. A
+  * base glyph which is not attached to another glyph, always returns NULL.
+  */
+GR2_API const gr_slot* gr_slot_attached_to(const gr_slot* p);
+
+/** Returns the first slot attached to this slot.
+  *
+  * Attached slots form a singly linked list from the parent. This returns the first
+  * slot in that list. Note that this is a reference to another slot that is also in
+  * the main segment doubly linked list.
+  *
+  * if gr_slot_first_attachment(p) != NULL then gr_slot_attached_to(gr_slot_first_attachment(p)) == p.
+  */
+GR2_API const gr_slot* gr_slot_first_attachment(const gr_slot* p);
+
+/** Returns the next slot attached to our attachment parent.
+  *
+  * This returns the next slot in the singly linked list of slots attached to this
+  * slot's parent. If there are no more such slots, NULL is returned. If there is
+  * no parent, i.e. the passed slot is a cluster base, then the next cluster base
+  * in graphical order (ltr, even for rtl text) is returned.
+  *
+  * if gr_slot_next_sibling_attachment(p) != NULL then gr_slot_attached_to(gr_slot_next_sibling_attachment(p)) == gr_slot_attached_to(p).
+  */
+GR2_API const gr_slot* gr_slot_next_sibling_attachment(const gr_slot* p);
+
+
+/** Returns glyph id of the slot
+  *
+  * Each slot has a glyphid which is rendered at the position given by the slot. This
+  * glyphid is the real glyph to be rendered and never a pseudo glyph.
+  */
+GR2_API unsigned short gr_slot_gid(const gr_slot* p);
+
+/** Returns X offset of glyph from start of segment **/
+GR2_API float gr_slot_origin_X(const gr_slot* p);
+
+/** Returns Y offset of glyph from start of segment **/
+GR2_API float gr_slot_origin_Y(const gr_slot* p);
+
+/** Returns the glyph advance for this glyph as adjusted for kerning
+  *
+  * @param p    Slot to give results for
+  * @param face gr_face of the glyphs. May be NULL if unhinted advances used
+  * @param font gr_font to scale for pixel results. If NULL returns design
+  *             units advance. If not NULL then returns pixel advance based
+  *             on hinted or scaled glyph advances in the font. face must be
+  *             passed for hinted advances to be used.
+  */
+GR2_API float gr_slot_advance_X(const gr_slot* p, const gr_face* face, const gr_font *font);
+
+/** Returns the vertical advance for the glyph in the slot adjusted for kerning
+  *
+  * Returns design units unless font is not NULL in which case the pixel value
+  * is returned scaled for the given font
+  */
+GR2_API float gr_slot_advance_Y(const gr_slot* p, const gr_face* face, const gr_font *font);
+
+/** Returns the gr_char_info index before us
+  *
+  * Returns the index of the gr_char_info that a cursor before this slot, would put
+  * an underlying cursor before. This may also be interpretted as each slot holding
+  * a set of char_infos that it is associated with and this function returning the
+  * index of the char_info with lowest index, from this set.
+  */
+GR2_API int gr_slot_before(const gr_slot* p/*not NULL*/);
+
+/** Returns the gr_char_info index after us
+  *
+  * Returns the index of the gr_char_info that a cursor after this slot would put an
+  * underlying cursor after. This may also be interpretted as each slot holding a set
+  * of char_infos that it is associated with and this function returning the index of
+  * the char_info with the highest index, from this set.
+  */
+GR2_API int gr_slot_after(const gr_slot* p/*not NULL*/);
+
+/** Returns the index of this slot in the segment
+  *
+  * Returns the index given to this slot during final positioning. This corresponds
+  * to the value returned br gr_cinfo_before() and gr_cinfo_after()
+  */
+GR2_API unsigned int gr_slot_index(const gr_slot* p/*not NULL*/);
+
+/** Return a slot attribute value
+  *
+  * Given a slot and an attribute along with a possible subattribute, return the
+  * corresponding value in the slot. See enum gr_attrCode for details of each attribute.
+  */
+GR2_API int gr_slot_attr(const gr_slot* p/*not NULL*/, const gr_segment* pSeg/*not NULL*/, enum gr_attrCode index, gr_uint8 subindex); //tbd - do we need to expose this?
+
+/** Returns whether text may be inserted before this glyph.
+  *
+  * This indicates whether a cursor can be put before this slot. It applies to
+  * base glyphs that have no parent as well as attached glyphs that have the
+  * .insert attribute explicitly set to true. This is the primary mechanism
+  * for identifying contiguous sequences of base plus diacritics.
+  */
+GR2_API int gr_slot_can_insert_before(const gr_slot* p);
+
+/** Returns the original gr_char_info index this slot refers to.
+  *
+  * Each Slot has a gr_char_info that it originates from. This is that gr_char_info.
+  * The index is passed to gr_seg_cinfo(). This information is useful for testing.
+  */
+GR2_API int gr_slot_original(const gr_slot* p/*not NULL*/);
+
+/** Breaks a segment into lines.
+  *
+  * Breaks the slot linked list at the given point in the linked list. It is up
+  * to the application to keep track of the first slot on each line.
+  */
+GR2_API void gr_slot_linebreak_before(gr_slot *p/*not NULL*/);
+
+#ifdef __cplusplus
+}
+#endif
diff --git a/thirdparty/graphite/include/graphite2/Types.h b/thirdparty/graphite/include/graphite2/Types.h
new file mode 100644
index 0000000000..916c91191b
--- /dev/null
+++ b/thirdparty/graphite/include/graphite2/Types.h
@@ -0,0 +1,79 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+    Alternatively, the contents of this file may be used under the terms
+    of the Mozilla Public License (http://mozilla.org/MPL) or the GNU
+    General Public License, as published by the Free Software Foundation,
+    either version 2 of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <stddef.h>
+
+typedef unsigned char   gr_uint8;
+typedef gr_uint8        gr_byte;
+typedef signed char     gr_int8;
+typedef unsigned short  gr_uint16;
+typedef short           gr_int16;
+typedef unsigned int    gr_uint32;
+typedef int             gr_int32;
+
+enum gr_encform {
+  gr_utf8 = 1/*sizeof(uint8)*/, gr_utf16 = 2/*sizeof(uint16)*/, gr_utf32 = 4/*sizeof(uint32)*/
+};
+
+
+// Define API function declspec/attributes and how each supported compiler or OS
+// allows us to specify them.
+#if defined __GNUC__
+  #define _gr2_and ,
+  #define _gr2_tag_fn(a)        __attribute__((a))
+  #define _gr2_deprecated_flag  deprecated
+  #define _gr2_export_flag      visibility("default")
+  #define _gr2_import_flag      visibility("default")
+  #define _gr2_static_flag      visibility("hidden")
+#endif
+
+#if defined _WIN32 || defined __CYGWIN__
+  #if defined __GNUC__  // These three will be redefined for Windows
+    #undef _gr2_export_flag
+    #undef _gr2_import_flag
+    #undef _gr2_static_flag
+  #else  // How MSVC sepcifies function level attributes adn deprecation
+    #define _gr2_and
+    #define _gr2_tag_fn(a)       __declspec(a)
+    #define _gr2_deprecated_flag deprecated
+  #endif
+  #define _gr2_export_flag     dllexport
+  #define _gr2_import_flag     dllimport
+  #define _gr2_static_flag
+#endif
+
+#if defined GRAPHITE2_STATIC
+  #define GR2_API             _gr2_tag_fn(_gr2_static_flag)
+  #define GR2_DEPRECATED_API  _gr2_tag_fn(_gr2_deprecated_flag _gr2_and _gr2_static_flag)
+#elif defined GRAPHITE2_EXPORTING
+  #define GR2_API             _gr2_tag_fn(_gr2_export_flag)
+  #define GR2_DEPRECATED_API  _gr2_tag_fn(_gr2_deprecated_flag _gr2_and _gr2_export_flag)
+#else
+  #define GR2_API             _gr2_tag_fn(_gr2_import_flag)
+  #define GR2_DEPRECATED_API  _gr2_tag_fn(_gr2_deprecated_flag _gr2_and _gr2_import_flag)
+#endif
diff --git a/thirdparty/graphite/src/CmapCache.cpp b/thirdparty/graphite/src/CmapCache.cpp
new file mode 100644
index 0000000000..d070019a34
--- /dev/null
+++ b/thirdparty/graphite/src/CmapCache.cpp
@@ -0,0 +1,155 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+#include "inc/Main.h"
+#include "inc/CmapCache.h"
+#include "inc/Face.h"
+#include "inc/TtfTypes.h"
+#include "inc/TtfUtil.h"
+
+
+using namespace graphite2;
+
+const void * bmp_subtable(const Face::Table & cmap)
+{
+    const void * stbl;
+    if (!cmap.size()) return 0;
+    if (TtfUtil::CheckCmapSubtable4(stbl = TtfUtil::FindCmapSubtable(cmap, 3, 1, cmap.size()), cmap + cmap.size())
+     || TtfUtil::CheckCmapSubtable4(stbl = TtfUtil::FindCmapSubtable(cmap, 0, 3, cmap.size()), cmap + cmap.size())
+     || TtfUtil::CheckCmapSubtable4(stbl = TtfUtil::FindCmapSubtable(cmap, 0, 2, cmap.size()), cmap + cmap.size())
+     || TtfUtil::CheckCmapSubtable4(stbl = TtfUtil::FindCmapSubtable(cmap, 0, 1, cmap.size()), cmap + cmap.size())
+     || TtfUtil::CheckCmapSubtable4(stbl = TtfUtil::FindCmapSubtable(cmap, 0, 0, cmap.size()), cmap + cmap.size()))
+        return stbl;
+    return 0;
+}
+
+const void * smp_subtable(const Face::Table & cmap)
+{
+    const void * stbl;
+    if (!cmap.size()) return 0;
+    if (TtfUtil::CheckCmapSubtable12(stbl = TtfUtil::FindCmapSubtable(cmap, 3, 10, cmap.size()), cmap + cmap.size())
+     || TtfUtil::CheckCmapSubtable12(stbl = TtfUtil::FindCmapSubtable(cmap, 0, 4, cmap.size()), cmap + cmap.size()))
+        return stbl;
+    return 0;
+}
+
+template <unsigned int (*NextCodePoint)(const void *, unsigned int, int *),
+          uint16 (*LookupCodePoint)(const void *, unsigned int, int)>
+bool cache_subtable(uint16 * blocks[], const void * cst, const unsigned int limit)
+{
+    int rangeKey = 0;
+    uint32          codePoint = NextCodePoint(cst, 0, &rangeKey),
+                    prevCodePoint = 0;
+    while (codePoint < limit)
+    {
+        unsigned int block = codePoint >> 8;
+        if (!blocks[block])
+        {
+            blocks[block] = grzeroalloc<uint16>(0x100);
+            if (!blocks[block])
+                return false;
+        }
+        blocks[block][codePoint & 0xFF] = LookupCodePoint(cst, codePoint, rangeKey);
+        // prevent infinite loop
+        if (codePoint <= prevCodePoint)
+            codePoint = prevCodePoint + 1;
+        prevCodePoint = codePoint;
+        codePoint =  NextCodePoint(cst, codePoint, &rangeKey);
+    }
+    return true;
+}
+
+
+CachedCmap::CachedCmap(const Face & face)
+: m_isBmpOnly(true),
+  m_blocks(0)
+{
+    const Face::Table cmap(face, Tag::cmap);
+    if (!cmap)  return;
+
+    const void * bmp_cmap = bmp_subtable(cmap);
+    const void * smp_cmap = smp_subtable(cmap);
+    m_isBmpOnly = !smp_cmap;
+
+    m_blocks = grzeroalloc<uint16 *>(m_isBmpOnly ? 0x100 : 0x1100);
+    if (m_blocks && smp_cmap)
+    {
+        if (!cache_subtable<TtfUtil::CmapSubtable12NextCodepoint, TtfUtil::CmapSubtable12Lookup>(m_blocks, smp_cmap, 0x10FFFF))
+            return;
+    }
+
+    if (m_blocks && bmp_cmap)
+    {
+        if (!cache_subtable<TtfUtil::CmapSubtable4NextCodepoint, TtfUtil::CmapSubtable4Lookup>(m_blocks, bmp_cmap, 0xFFFF))
+            return;
+    }
+}
+
+CachedCmap::~CachedCmap() throw()
+{
+    if (!m_blocks) return;
+    unsigned int numBlocks = (m_isBmpOnly)? 0x100 : 0x1100;
+    for (unsigned int i = 0; i < numBlocks; i++)
+        free(m_blocks[i]);
+    free(m_blocks);
+}
+
+uint16 CachedCmap::operator [] (const uint32 usv) const throw()
+{
+    if ((m_isBmpOnly && usv > 0xFFFF) || (usv > 0x10FFFF))
+        return 0;
+    const uint32 block = 0xFFFF & (usv >> 8);
+    if (m_blocks[block])
+        return m_blocks[block][usv & 0xFF];
+    return 0;
+};
+
+CachedCmap::operator bool() const throw()
+{
+    return m_blocks != 0;
+}
+
+
+DirectCmap::DirectCmap(const Face & face)
+: _cmap(face, Tag::cmap),
+  _smp(smp_subtable(_cmap)),
+  _bmp(bmp_subtable(_cmap))
+{
+}
+
+uint16 DirectCmap::operator [] (const uint32 usv) const throw()
+{
+    return usv > 0xFFFF
+            ? (_smp ? TtfUtil::CmapSubtable12Lookup(_smp, usv, 0) : 0)
+            : TtfUtil::CmapSubtable4Lookup(_bmp, usv, 0);
+}
+
+DirectCmap::operator bool () const throw()
+{
+    return _cmap && _bmp;
+}
+
diff --git a/thirdparty/graphite/src/Code.cpp b/thirdparty/graphite/src/Code.cpp
new file mode 100644
index 0000000000..ec5ab298ca
--- /dev/null
+++ b/thirdparty/graphite/src/Code.cpp
@@ -0,0 +1,782 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This class represents loaded graphite stack machine code.  It performs
+// basic sanity checks, on the incoming code to prevent more obvious problems
+// from crashing graphite.
+// Author: Tim Eves
+
+#include <cassert>
+#include <cstddef>
+#include <cstdlib>
+#include <cstring>
+#include "graphite2/Segment.h"
+#include "inc/Code.h"
+#include "inc/Face.h"
+#include "inc/GlyphFace.h"
+#include "inc/GlyphCache.h"
+#include "inc/Machine.h"
+#include "inc/Rule.h"
+#include "inc/Silf.h"
+
+#include <cstdio>
+
+#ifdef NDEBUG
+#ifdef __GNUC__
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#endif
+#endif
+
+
+using namespace graphite2;
+using namespace vm;
+
+namespace {
+
+inline bool is_return(const instr i) {
+    const opcode_t * opmap = Machine::getOpcodeTable();
+    const instr pop_ret  = *opmap[POP_RET].impl,
+                ret_zero = *opmap[RET_ZERO].impl,
+                ret_true = *opmap[RET_TRUE].impl;
+    return i == pop_ret || i == ret_zero || i == ret_true;
+}
+
+struct context
+{
+    context(uint8 ref=0) : codeRef(ref) {flags.changed=false; flags.referenced=false;}
+    struct {
+        uint8   changed:1,
+                referenced:1;
+    } flags;
+    uint8       codeRef;
+};
+
+} // end namespace
+
+
+class Machine::Code::decoder
+{
+public:
+    struct limits;
+    static const int NUMCONTEXTS = 256;
+
+    decoder(limits & lims, Code &code, enum passtype pt) throw();
+
+    bool        load(const byte * bc_begin, const byte * bc_end);
+    void        apply_analysis(instr * const code, instr * code_end);
+    byte        max_ref() { return _max_ref; }
+    int         out_index() const { return _out_index; }
+
+private:
+    void        set_ref(int index) throw();
+    void        set_noref(int index) throw();
+    void        set_changed(int index) throw();
+    opcode      fetch_opcode(const byte * bc);
+    void        analyse_opcode(const opcode, const int8 * const dp) throw();
+    bool        emit_opcode(opcode opc, const byte * & bc);
+    bool        validate_opcode(const byte opc, const byte * const bc);
+    bool        valid_upto(const uint16 limit, const uint16 x) const throw();
+    bool        test_context() const throw();
+    bool        test_ref(int8 index) const throw();
+    bool        test_attr(attrCode attr) const throw();
+    void        failure(const status_t s) const throw() { _code.failure(s); }
+
+    Code              & _code;
+    int                 _out_index;
+    uint16              _out_length;
+    instr             * _instr;
+    byte              * _data;
+    limits            & _max;
+    enum passtype       _passtype;
+    int                 _stack_depth;
+    bool                _in_ctxt_item;
+    int16               _slotref;
+    context             _contexts[NUMCONTEXTS];
+    byte                _max_ref;
+};
+
+
+struct Machine::Code::decoder::limits
+{
+  const byte       * bytecode;
+  const uint8        pre_context;
+  const uint16       rule_length,
+                     classes,
+                     glyf_attrs,
+                     features;
+  const byte         attrid[gr_slatMax];
+};
+
+inline Machine::Code::decoder::decoder(limits & lims, Code &code, enum passtype pt) throw()
+: _code(code),
+  _out_index(code._constraint ? 0 : lims.pre_context),
+  _out_length(code._constraint ? 1 : lims.rule_length),
+  _instr(code._code), _data(code._data), _max(lims), _passtype(pt),
+  _stack_depth(0),
+  _in_ctxt_item(false),
+  _slotref(0),
+  _max_ref(0)
+{ }
+
+
+
+Machine::Code::Code(bool is_constraint, const byte * bytecode_begin, const byte * const bytecode_end,
+           uint8 pre_context, uint16 rule_length, const Silf & silf, const Face & face,
+           enum passtype pt, byte * * const _out)
+ :  _code(0), _data(0), _data_size(0), _instr_count(0), _max_ref(0), _status(loaded),
+    _constraint(is_constraint), _modify(false), _delete(false), _own(_out==0)
+{
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::category _code_cat(face.tele.code);
+#endif
+    assert(bytecode_begin != 0);
+    if (bytecode_begin == bytecode_end)
+    {
+      // ::new (this) Code();
+      return;
+    }
+    assert(bytecode_end > bytecode_begin);
+    const opcode_t *    op_to_fn = Machine::getOpcodeTable();
+
+    // Allocate code and data target buffers, these sizes are a worst case
+    // estimate.  Once we know their real sizes the we'll shrink them.
+    if (_out)   _code = reinterpret_cast<instr *>(*_out);
+    else        _code = static_cast<instr *>(malloc(estimateCodeDataOut(bytecode_end-bytecode_begin, 1, is_constraint ? 0 : rule_length)));
+    _data = reinterpret_cast<byte *>(_code + (bytecode_end - bytecode_begin));
+
+    if (!_code || !_data) {
+        failure(alloc_failed);
+        return;
+    }
+
+    decoder::limits lims = {
+        bytecode_end,
+        pre_context,
+        rule_length,
+        silf.numClasses(),
+        face.glyphs().numAttrs(),
+        face.numFeatures(),
+        {1,1,1,1,1,1,1,1,
+         1,1,1,1,1,1,1,255,
+         1,1,1,1,1,1,1,1,
+         1,1,1,1,1,1,0,0,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0,0,
+         0,0,0,0,0,0,0, silf.numUser()}
+    };
+
+    decoder dec(lims, *this, pt);
+    if(!dec.load(bytecode_begin, bytecode_end))
+       return;
+
+    // Is this an empty program?
+    if (_instr_count == 0)
+    {
+      release_buffers();
+      ::new (this) Code();
+      return;
+    }
+
+    // When we reach the end check we've terminated it correctly
+    if (!is_return(_code[_instr_count-1])) {
+        failure(missing_return);
+        return;
+    }
+
+    assert((_constraint && immutable()) || !_constraint);
+    dec.apply_analysis(_code, _code + _instr_count);
+    _max_ref = dec.max_ref();
+
+    // Now we know exactly how much code and data the program really needs
+    // realloc the buffers to exactly the right size so we don't waste any
+    // memory.
+    assert((bytecode_end - bytecode_begin) >= ptrdiff_t(_instr_count));
+    assert((bytecode_end - bytecode_begin) >= ptrdiff_t(_data_size));
+    memmove(_code + (_instr_count+1), _data, _data_size*sizeof(byte));
+    size_t const total_sz = ((_instr_count+1) + (_data_size + sizeof(instr)-1)/sizeof(instr))*sizeof(instr);
+    if (_out)
+        *_out += total_sz;
+    else
+    {
+      instr * const old_code = _code;
+      _code = static_cast<instr *>(realloc(_code, total_sz));
+      if (!_code) free(old_code);
+    }
+   _data = reinterpret_cast<byte *>(_code + (_instr_count+1));
+
+    if (!_code)
+    {
+        failure(alloc_failed);
+        return;
+    }
+
+    // Make this RET_ZERO, we should never reach this but just in case ...
+    _code[_instr_count] = op_to_fn[RET_ZERO].impl[_constraint];
+
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::count_bytes(_data_size + (_instr_count+1)*sizeof(instr));
+#endif
+}
+
+Machine::Code::~Code() throw ()
+{
+    if (_own)
+        release_buffers();
+}
+
+
+bool Machine::Code::decoder::load(const byte * bc, const byte * bc_end)
+{
+    _max.bytecode = bc_end;
+    while (bc < bc_end)
+    {
+        const opcode opc = fetch_opcode(bc++);
+        if (opc == vm::MAX_OPCODE)
+            return false;
+
+        analyse_opcode(opc, reinterpret_cast<const int8 *>(bc));
+
+        if (!emit_opcode(opc, bc))
+            return false;
+    }
+
+    return bool(_code);
+}
+
+// Validation check and fixups.
+//
+
+opcode Machine::Code::decoder::fetch_opcode(const byte * bc)
+{
+    const byte opc = *bc++;
+
+    // Do some basic sanity checks based on what we know about the opcode
+    if (!validate_opcode(opc, bc))  return MAX_OPCODE;
+
+    // And check its arguments as far as possible
+    switch (opcode(opc))
+    {
+        case NOP :
+            break;
+        case PUSH_BYTE :
+        case PUSH_BYTEU :
+        case PUSH_SHORT :
+        case PUSH_SHORTU :
+        case PUSH_LONG :
+            ++_stack_depth;
+            break;
+        case ADD :
+        case SUB :
+        case MUL :
+        case DIV :
+        case MIN_ :
+        case MAX_ :
+        case AND :
+        case OR :
+        case EQUAL :
+        case NOT_EQ :
+        case LESS :
+        case GTR :
+        case LESS_EQ :
+        case GTR_EQ :
+        case BITOR :
+        case BITAND :
+            if (--_stack_depth <= 0)
+                failure(underfull_stack);
+            break;
+        case NEG :
+        case TRUNC8 :
+        case TRUNC16 :
+        case NOT :
+        case BITNOT :
+        case BITSET :
+            if (_stack_depth <= 0)
+                failure(underfull_stack);
+            break;
+        case COND :
+            _stack_depth -= 2;
+            if (_stack_depth <= 0)
+                failure(underfull_stack);
+            break;
+        case NEXT_N :           // runtime checked
+            break;
+        case NEXT :
+        case COPY_NEXT :
+            ++_out_index;
+            if (_out_index < -1 || _out_index > _out_length || _slotref > _max.rule_length)
+                failure(out_of_range_data);
+            break;
+        case PUT_GLYPH_8BIT_OBS :
+            valid_upto(_max.classes, bc[0]);
+            test_context();
+            break;
+        case PUT_SUBS_8BIT_OBS :
+            test_ref(int8(bc[0]));
+            valid_upto(_max.classes, bc[1]);
+            valid_upto(_max.classes, bc[2]);
+            test_context();
+            break;
+        case PUT_COPY :
+            test_ref(int8(bc[0]));
+            test_context();
+            break;
+        case INSERT :
+            if (_passtype >= PASS_TYPE_POSITIONING)
+                failure(invalid_opcode);
+            ++_out_length;
+            if (_out_index < 0) ++_out_index;
+            if (_out_index < -1 || _out_index >= _out_length)
+                failure(out_of_range_data);
+            break;
+        case DELETE :
+            if (_passtype >= PASS_TYPE_POSITIONING)
+                failure(invalid_opcode);
+            if (_out_index < _max.pre_context)
+                failure(out_of_range_data);
+            --_out_index;
+            --_out_length;
+            if (_out_index < -1 || _out_index > _out_length)
+                failure(out_of_range_data);
+            break;
+        case ASSOC :
+            if (bc[0] == 0)
+                failure(out_of_range_data);
+            for (uint8 num = bc[0]; num; --num)
+                test_ref(int8(bc[num]));
+            test_context();
+            break;
+        case CNTXT_ITEM :
+            valid_upto(_max.rule_length, _max.pre_context + int8(bc[0]));
+            if (bc + 2 + bc[1] >= _max.bytecode)    failure(jump_past_end);
+            if (_in_ctxt_item)                      failure(nested_context_item);
+            break;
+        case ATTR_SET :
+        case ATTR_ADD :
+        case ATTR_SUB :
+        case ATTR_SET_SLOT :
+            if (--_stack_depth < 0)
+                failure(underfull_stack);
+            valid_upto(gr_slatMax, bc[0]);
+            if (attrCode(bc[0]) == gr_slatUserDefn)     // use IATTR for user attributes
+                failure(out_of_range_data);
+            test_attr(attrCode(bc[0]));
+            test_context();
+            break;
+        case IATTR_SET_SLOT :
+            if (--_stack_depth < 0)
+                failure(underfull_stack);
+            if (valid_upto(gr_slatMax, bc[0]))
+                valid_upto(_max.attrid[bc[0]], bc[1]);
+            test_attr(attrCode(bc[0]));
+            test_context();
+            break;
+        case PUSH_SLOT_ATTR :
+            ++_stack_depth;
+            valid_upto(gr_slatMax, bc[0]);
+            test_ref(int8(bc[1]));
+            if (attrCode(bc[0]) == gr_slatUserDefn)     // use IATTR for user attributes
+                failure(out_of_range_data);
+            test_attr(attrCode(bc[0]));
+            break;
+        case PUSH_GLYPH_ATTR_OBS :
+        case PUSH_ATT_TO_GATTR_OBS :
+            ++_stack_depth;
+            valid_upto(_max.glyf_attrs, bc[0]);
+            test_ref(int8(bc[1]));
+            break;
+        case PUSH_ATT_TO_GLYPH_METRIC :
+        case PUSH_GLYPH_METRIC :
+            ++_stack_depth;
+            valid_upto(kgmetDescent, bc[0]);
+            test_ref(int8(bc[1]));
+            // level: dp[2] no check necessary
+            break;
+        case PUSH_FEAT :
+            ++_stack_depth;
+            valid_upto(_max.features, bc[0]);
+            test_ref(int8(bc[1]));
+            break;
+        case PUSH_ISLOT_ATTR :
+            ++_stack_depth;
+            if (valid_upto(gr_slatMax, bc[0]))
+            {
+                test_ref(int8(bc[1]));
+                valid_upto(_max.attrid[bc[0]], bc[2]);
+            }
+            test_attr(attrCode(bc[0]));
+            break;
+        case PUSH_IGLYPH_ATTR :// not implemented
+            ++_stack_depth;
+            break;
+        case POP_RET :
+            if (--_stack_depth < 0)
+                failure(underfull_stack);
+            GR_FALLTHROUGH;
+            // no break
+        case RET_ZERO :
+        case RET_TRUE :
+            break;
+        case IATTR_SET :
+        case IATTR_ADD :
+        case IATTR_SUB :
+            if (--_stack_depth < 0)
+                failure(underfull_stack);
+            if (valid_upto(gr_slatMax, bc[0]))
+                valid_upto(_max.attrid[bc[0]], bc[1]);
+            test_attr(attrCode(bc[0]));
+            test_context();
+            break;
+        case PUSH_PROC_STATE :  // dummy: dp[0] no check necessary
+        case PUSH_VERSION :
+            ++_stack_depth;
+            break;
+        case PUT_SUBS :
+            test_ref(int8(bc[0]));
+            valid_upto(_max.classes, uint16(bc[1]<< 8) | bc[2]);
+            valid_upto(_max.classes, uint16(bc[3]<< 8) | bc[4]);
+            test_context();
+            break;
+        case PUT_SUBS2 :        // not implemented
+        case PUT_SUBS3 :        // not implemented
+            break;
+        case PUT_GLYPH :
+            valid_upto(_max.classes, uint16(bc[0]<< 8) | bc[1]);
+            test_context();
+            break;
+        case PUSH_GLYPH_ATTR :
+        case PUSH_ATT_TO_GLYPH_ATTR :
+            ++_stack_depth;
+            valid_upto(_max.glyf_attrs, uint16(bc[0]<< 8) | bc[1]);
+            test_ref(int8(bc[2]));
+            break;
+        case SET_FEAT :
+            valid_upto(_max.features, bc[0]);
+            test_ref(int8(bc[1]));
+            break;
+        default:
+            failure(invalid_opcode);
+            break;
+    }
+
+    return bool(_code) ? opcode(opc) : MAX_OPCODE;
+}
+
+
+void Machine::Code::decoder::analyse_opcode(const opcode opc, const int8  * arg) throw()
+{
+  switch (opc)
+  {
+    case DELETE :
+      _code._delete = true;
+      break;
+    case ASSOC :
+      set_changed(0);
+//      for (uint8 num = arg[0]; num; --num)
+//        _analysis.set_noref(num);
+      break;
+    case PUT_GLYPH_8BIT_OBS :
+    case PUT_GLYPH :
+      _code._modify = true;
+      set_changed(0);
+      break;
+    case ATTR_SET :
+    case ATTR_ADD :
+    case ATTR_SUB :
+    case ATTR_SET_SLOT :
+    case IATTR_SET_SLOT :
+    case IATTR_SET :
+    case IATTR_ADD :
+    case IATTR_SUB :
+      set_noref(0);
+      break;
+    case NEXT :
+    case COPY_NEXT :
+      ++_slotref;
+      _contexts[_slotref] = context(uint8(_code._instr_count+1));
+      // if (_analysis.slotref > _analysis.max_ref) _analysis.max_ref = _analysis.slotref;
+      break;
+    case INSERT :
+      if (_slotref >= 0) --_slotref;
+      _code._modify = true;
+      break;
+    case PUT_SUBS_8BIT_OBS :    // slotref on 1st parameter
+    case PUT_SUBS :
+      _code._modify = true;
+      set_changed(0);
+      GR_FALLTHROUGH;
+      // no break
+    case PUT_COPY :
+      if (arg[0] != 0) { set_changed(0); _code._modify = true; }
+      set_ref(arg[0]);
+      break;
+    case PUSH_GLYPH_ATTR_OBS :
+    case PUSH_SLOT_ATTR :
+    case PUSH_GLYPH_METRIC :
+    case PUSH_ATT_TO_GATTR_OBS :
+    case PUSH_ATT_TO_GLYPH_METRIC :
+    case PUSH_ISLOT_ATTR :
+    case PUSH_FEAT :
+    case SET_FEAT :
+      set_ref(arg[1]);
+      break;
+    case PUSH_ATT_TO_GLYPH_ATTR :
+    case PUSH_GLYPH_ATTR :
+      set_ref(arg[2]);
+      break;
+    default:
+        break;
+  }
+}
+
+
+bool Machine::Code::decoder::emit_opcode(opcode opc, const byte * & bc)
+{
+    const opcode_t * op_to_fn = Machine::getOpcodeTable();
+    const opcode_t & op       = op_to_fn[opc];
+    if (op.impl[_code._constraint] == 0)
+    {
+        failure(unimplemented_opcode_used);
+        return false;
+    }
+
+    const size_t     param_sz = op.param_sz == VARARGS ? bc[0] + 1 : op.param_sz;
+
+    // Add this instruction
+    *_instr++ = op.impl[_code._constraint];
+    ++_code._instr_count;
+
+    // Grab the parameters
+    if (param_sz) {
+        memcpy(_data, bc, param_sz * sizeof(byte));
+        bc               += param_sz;
+        _data            += param_sz;
+        _code._data_size += param_sz;
+    }
+
+    // recursively decode a context item so we can split the skip into
+    // instruction and data portions.
+    if (opc == CNTXT_ITEM)
+    {
+        assert(_out_index == 0);
+        _in_ctxt_item = true;
+        _out_index = _max.pre_context + int8(_data[-2]);
+        _slotref = int8(_data[-2]);
+        _out_length = _max.rule_length;
+
+        const size_t ctxt_start = _code._instr_count;
+        byte & instr_skip = _data[-1];
+        byte & data_skip  = *_data++;
+        ++_code._data_size;
+        const byte *curr_end = _max.bytecode;
+
+        if (load(bc, bc + instr_skip))
+        {
+            bc += instr_skip;
+            data_skip  = instr_skip - byte(_code._instr_count - ctxt_start);
+            instr_skip =  byte(_code._instr_count - ctxt_start);
+            _max.bytecode = curr_end;
+
+            _out_length = 1;
+            _out_index = 0;
+            _slotref = 0;
+            _in_ctxt_item = false;
+        }
+        else
+        {
+            _out_index = 0;
+            _slotref = 0;
+            return false;
+        }
+    }
+
+    return bool(_code);
+}
+
+
+void Machine::Code::decoder::apply_analysis(instr * const code, instr * code_end)
+{
+    // insert TEMP_COPY commands for slots that need them (that change and are referenced later)
+    int tempcount = 0;
+    if (_code._constraint) return;
+
+    const instr temp_copy = Machine::getOpcodeTable()[TEMP_COPY].impl[0];
+    for (const context * c = _contexts, * const ce = c + _slotref; c < ce; ++c)
+    {
+        if (!c->flags.referenced || !c->flags.changed) continue;
+
+        instr * const tip = code + c->codeRef + tempcount;
+        memmove(tip+1, tip, (code_end - tip) * sizeof(instr));
+        *tip = temp_copy;
+        ++code_end;
+        ++tempcount;
+        _code._delete = true;
+    }
+
+    _code._instr_count = code_end - code;
+}
+
+
+inline
+bool Machine::Code::decoder::validate_opcode(const byte opc, const byte * const bc)
+{
+    if (opc >= MAX_OPCODE)
+    {
+        failure(invalid_opcode);
+        return false;
+    }
+    const opcode_t & op = Machine::getOpcodeTable()[opc];
+    if (op.impl[_code._constraint] == 0)
+    {
+        failure(unimplemented_opcode_used);
+        return false;
+    }
+    if (op.param_sz == VARARGS && bc >= _max.bytecode)
+    {
+        failure(arguments_exhausted);
+        return false;
+    }
+    const size_t param_sz = op.param_sz == VARARGS ? bc[0] + 1 : op.param_sz;
+    if (bc - 1 + param_sz >= _max.bytecode)
+    {
+        failure(arguments_exhausted);
+        return false;
+    }
+    return true;
+}
+
+
+bool Machine::Code::decoder::valid_upto(const uint16 limit, const uint16 x) const throw()
+{
+    const bool t = (limit != 0) && (x < limit);
+    if (!t) failure(out_of_range_data);
+    return t;
+}
+
+inline
+bool Machine::Code::decoder::test_ref(int8 index) const throw()
+{
+    if (_code._constraint && !_in_ctxt_item)
+    {
+        if (index > 0 || -index > _max.pre_context)
+        {
+            failure(out_of_range_data);
+            return false;
+        }
+    }
+    else
+    {
+      if (_max.rule_length == 0
+          || (_slotref + _max.pre_context + index >= _max.rule_length)
+          || (_slotref + _max.pre_context + index < 0))
+      {
+        failure(out_of_range_data);
+        return false;
+      }
+    }
+    return true;
+}
+
+bool Machine::Code::decoder::test_context() const throw()
+{
+    if (_out_index >= _out_length || _out_index < 0 || _slotref >= NUMCONTEXTS - 1)
+    {
+        failure(out_of_range_data);
+        return false;
+    }
+    return true;
+}
+
+bool Machine::Code::decoder::test_attr(attrCode) const throw()
+{
+#if 0   // This code is coming but causes backward compatibility problems.
+    if (_passtype < PASS_TYPE_POSITIONING)
+    {
+        if (attr != gr_slatBreak && attr != gr_slatDir && attr != gr_slatUserDefn
+                                 && attr != gr_slatCompRef)
+        {
+            failure(out_of_range_data);
+            return false;
+        }
+    }
+#endif
+    return true;
+}
+
+inline
+void Machine::Code::failure(const status_t s) throw() {
+    release_buffers();
+    _status = s;
+}
+
+
+inline
+void Machine::Code::decoder::set_ref(int index) throw() {
+    if (index + _slotref < 0 || index + _slotref >= NUMCONTEXTS) return;
+    _contexts[index + _slotref].flags.referenced = true;
+    if (index + _slotref > _max_ref) _max_ref = index + _slotref;
+}
+
+
+inline
+void Machine::Code::decoder::set_noref(int index) throw() {
+    if (index + _slotref < 0 || index + _slotref >= NUMCONTEXTS) return;
+    if (index + _slotref > _max_ref) _max_ref = index + _slotref;
+}
+
+
+inline
+void Machine::Code::decoder::set_changed(int index) throw() {
+    if (index + _slotref < 0 || index + _slotref >= NUMCONTEXTS) return;
+    _contexts[index + _slotref].flags.changed= true;
+    if (index + _slotref > _max_ref) _max_ref = index + _slotref;
+}
+
+
+void Machine::Code::release_buffers() throw()
+{
+    if (_own)
+        free(_code);
+    _code = 0;
+    _data = 0;
+    _own  = false;
+}
+
+
+int32 Machine::Code::run(Machine & m, slotref * & map) const
+{
+//    assert(_own);
+    assert(*this);          // Check we are actually runnable
+
+    if (m.slotMap().size() <= size_t(_max_ref + m.slotMap().context())
+        || m.slotMap()[_max_ref + m.slotMap().context()] == 0)
+    {
+        m._status = Machine::slot_offset_out_bounds;
+        return 1;
+//        return m.run(_code, _data, map);
+    }
+
+    return  m.run(_code, _data, map);
+}
diff --git a/thirdparty/graphite/src/Collider.cpp b/thirdparty/graphite/src/Collider.cpp
new file mode 100644
index 0000000000..1929b39a58
--- /dev/null
+++ b/thirdparty/graphite/src/Collider.cpp
@@ -0,0 +1,1115 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <algorithm>
+#include <limits>
+#include <cmath>
+#include <string>
+#include <functional>
+#include "inc/Collider.h"
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/GlyphCache.h"
+#include "inc/Sparse.h"
+
+#define ISQRT2 0.707106781f
+
+// Possible rounding error for subbox boundaries: 0.016 = 1/64 = 1/256 * 4
+// (values in font range from 0..256)
+// #define SUBBOX_RND_ERR 0.016
+
+using namespace graphite2;
+
+////    SHIFT-COLLIDER    ////
+
+// Initialize the Collider to hold the basic movement limits for the
+// target slot, the one we are focusing on fixing.
+bool ShiftCollider::initSlot(Segment *seg, Slot *aSlot, const Rect &limit, float margin, float marginWeight,
+    const Position &currShift, const Position &currOffset, int dir, GR_MAYBE_UNUSED json * const dbgout)
+{
+    int i;
+    float mx, mn;
+    float a, shift;
+    const GlyphCache &gc = seg->getFace()->glyphs();
+    unsigned short gid = aSlot->gid();
+    if (!gc.check(gid))
+        return false;
+    const BBox &bb = gc.getBoundingBBox(gid);
+    const SlantBox &sb = gc.getBoundingSlantBox(gid);
+    //float sx = aSlot->origin().x + currShift.x;
+    //float sy = aSlot->origin().y + currShift.y;
+    if (currOffset.x != 0.f || currOffset.y != 0.f)
+        _limit = Rect(limit.bl - currOffset, limit.tr - currOffset);
+    else
+        _limit = limit;
+    // For a ShiftCollider, these indices indicate which vector we are moving by:
+    // each _ranges represents absolute space with respect to the origin of the slot. Thus take into account true origins but subtract the vmin for the slot
+    for (i = 0; i < 4; ++i)
+    {
+        switch (i) {
+            case 0 :	// x direction
+                mn = _limit.bl.x + currOffset.x;
+                mx = _limit.tr.x + currOffset.x;
+                _len[i] = bb.xa - bb.xi;
+                a = currOffset.y + currShift.y;
+                _ranges[i].initialise<XY>(mn, mx, margin, marginWeight, a);
+                break;
+            case 1 :	// y direction
+                mn = _limit.bl.y + currOffset.y;
+                mx = _limit.tr.y + currOffset.y;
+                _len[i] = bb.ya - bb.yi;
+                a = currOffset.x + currShift.x;
+                _ranges[i].initialise<XY>(mn, mx, margin, marginWeight, a);
+                break;
+            case 2 :	// sum (negatively sloped diagonal boundaries)
+                // pick closest x,y limit boundaries in s direction
+                shift = currOffset.x + currOffset.y + currShift.x + currShift.y;
+                mn = -2 * min(currShift.x - _limit.bl.x, currShift.y - _limit.bl.y) + shift;
+                mx = 2 * min(_limit.tr.x - currShift.x, _limit.tr.y - currShift.y) + shift;
+                _len[i] = sb.sa - sb.si;
+                a = currOffset.x - currOffset.y + currShift.x - currShift.y;
+                _ranges[i].initialise<SD>(mn, mx, margin / ISQRT2, marginWeight, a);
+                break;
+            case 3 :	// diff (positively sloped diagonal boundaries)
+                // pick closest x,y limit boundaries in d direction
+                shift = currOffset.x - currOffset.y + currShift.x - currShift.y;
+                mn = -2 * min(currShift.x - _limit.bl.x, _limit.tr.y - currShift.y) + shift;
+                mx = 2 * min(_limit.tr.x - currShift.x, currShift.y - _limit.bl.y) + shift;
+                _len[i] = sb.da - sb.di;
+                a = currOffset.x + currOffset.y + currShift.x + currShift.y;
+                _ranges[i].initialise<SD>(mn, mx, margin / ISQRT2, marginWeight, a);
+                break;
+        }
+    }
+
+	_target = aSlot;
+    if ((dir & 1) == 0)
+    {
+        // For LTR, switch and negate x limits.
+        _limit.bl.x = -1 * limit.tr.x;
+        //_limit.tr.x = -1 * limit.bl.x;
+    }
+    _currOffset = currOffset;
+    _currShift = currShift;
+    _origin = aSlot->origin() - currOffset;     // the original anchor position of the glyph
+
+	_margin = margin;
+	_marginWt = marginWeight;
+
+    SlotCollision *c = seg->collisionInfo(aSlot);
+    _seqClass = c->seqClass();
+	_seqProxClass = c->seqProxClass();
+    _seqOrder = c->seqOrder();
+    return true;
+}
+
+template <class O>
+float sdm(float vi, float va, float mx, float my, O op)
+{
+    float res = 2 * mx - vi;
+    if (op(res, vi + 2 * my))
+    {
+        res = va + 2 * my;
+        if (op(res, 2 * mx - va))
+            res = mx + my;
+    }
+    return res;
+}
+
+// Mark an area with a cost that can vary along the x or y axis. The region is expressed in terms of the centre of the target glyph in each axis
+void ShiftCollider::addBox_slope(bool isx, const Rect &box, const BBox &bb, const SlantBox &sb, const Position &org, float weight, float m, bool minright, int axis)
+{
+    float a, c;
+    switch (axis) {
+        case 0 :
+             if (box.bl.y < org.y + bb.ya && box.tr.y > org.y + bb.yi && box.width() > 0)
+            {
+                a = org.y + 0.5f * (bb.yi + bb.ya);
+                c = 0.5f * (bb.xi + bb.xa);
+                if (isx)
+                    _ranges[axis].weighted<XY>(box.bl.x - c, box.tr.x - c, weight, a, m,
+                                                (minright ? box.tr.x : box.bl.x) - c, a, 0, false);
+                else
+                    _ranges[axis].weighted<XY>(box.bl.x - c, box.tr.x - c, weight, a, 0, 0, org.y,
+                                                m * (a * a + sqr((minright ? box.tr.y : box.bl.y) - 0.5f * (bb.yi + bb.ya))), false);
+            }
+            break;
+        case 1 :
+            if (box.bl.x < org.x + bb.xa && box.tr.x > org.x + bb.xi && box.height() > 0)
+            {
+                a = org.x + 0.5f * (bb.xi + bb.xa);
+                c = 0.5f * (bb.yi + bb.ya);
+                if (isx)
+                    _ranges[axis].weighted<XY>(box.bl.y - c, box.tr.y - c, weight, a, 0, 0, org.x,
+                                                m * (a * a + sqr((minright ? box.tr.x : box.bl.x) - 0.5f * (bb.xi + bb.xa))), false);
+                else
+                    _ranges[axis].weighted<XY>(box.bl.y - c, box.tr.y - c, weight, a, m,
+                                                (minright ? box.tr.y : box.bl.y) - c, a, 0, false);
+            }
+            break;
+        case 2 :
+            if (box.bl.x - box.tr.y < org.x - org.y + sb.da && box.tr.x - box.bl.y > org.x - org.y + sb.di)
+            {
+                float d = org.x - org.y + 0.5f * (sb.di + sb.da);
+                c = 0.5f * (sb.si + sb.sa);
+                float smax = min(2 * box.tr.x - d, 2 * box.tr.y + d);
+                float smin = max(2 * box.bl.x - d, 2 * box.bl.y + d);
+                if (smin > smax) return;
+                float si;
+                a = d;
+                if (isx)
+                    si = 2 * (minright ? box.tr.x : box.bl.x) - a;
+                else
+                    si = 2 * (minright ? box.tr.y : box.bl.y) + a;
+                _ranges[axis].weighted<SD>(smin - c, smax - c, weight / 2, a, m / 2, si, 0, 0, isx);
+            }
+            break;
+        case 3 :
+            if (box.bl.x + box.bl.y < org.x + org.y + sb.sa && box.tr.x + box.tr.y > org.x + org.y + sb.si)
+            {
+                float s = org.x + org.y + 0.5f * (sb.si + sb.sa);
+                c = 0.5f * (sb.di + sb.da);
+                float dmax = min(2 * box.tr.x - s, s - 2 * box.bl.y);
+                float dmin = max(2 * box.bl.x - s, s - 2 * box.tr.y);
+                if (dmin > dmax) return;
+                float di;
+                a = s;
+                if (isx)
+                    di = 2 * (minright ? box.tr.x : box.bl.x) - a;
+                else
+                    di = 2 * (minright ? box.tr.y : box.bl.y) + a;
+                _ranges[axis].weighted<SD>(dmin - c, dmax - c, weight / 2, a, m / 2, di, 0, 0, !isx);
+            }
+            break;
+        default :
+            break;
+    }
+    return;
+}
+
+// Mark an area with an absolute cost, making it completely inaccessible.
+inline void ShiftCollider::removeBox(const Rect &box, const BBox &bb, const SlantBox &sb, const Position &org, int axis)
+{
+    float c;
+    switch (axis) {
+        case 0 :
+            if (box.bl.y < org.y + bb.ya && box.tr.y > org.y + bb.yi && box.width() > 0)
+            {
+                c = 0.5f * (bb.xi + bb.xa);
+                _ranges[axis].exclude(box.bl.x - c, box.tr.x - c);
+            }
+            break;
+        case 1 :
+            if (box.bl.x < org.x + bb.xa && box.tr.x > org.x + bb.xi && box.height() > 0)
+            {
+                c = 0.5f * (bb.yi + bb.ya);
+                _ranges[axis].exclude(box.bl.y - c, box.tr.y - c);
+            }
+            break;
+        case 2 :
+            if (box.bl.x - box.tr.y < org.x - org.y + sb.da && box.tr.x - box.bl.y > org.x - org.y + sb.di
+                && box.width() > 0 && box.height() > 0)
+            {
+                float di = org.x - org.y + sb.di;
+                float da = org.x - org.y + sb.da;
+                float smax = sdm(di, da, box.tr.x, box.tr.y, std::greater<float>());
+                float smin = sdm(da, di, box.bl.x, box.bl.y, std::less<float>());
+                c = 0.5f * (sb.si + sb.sa);
+                _ranges[axis].exclude(smin - c, smax - c);
+            }
+            break;
+        case 3 :
+            if (box.bl.x + box.bl.y < org.x + org.y + sb.sa && box.tr.x + box.tr.y > org.x + org.y + sb.si
+                && box.width() > 0 && box.height() > 0)
+            {
+                float si = org.x + org.y + sb.si;
+                float sa = org.x + org.y + sb.sa;
+                float dmax = sdm(si, sa, box.tr.x, -box.bl.y, std::greater<float>());
+                float dmin = sdm(sa, si, box.bl.x, -box.tr.y, std::less<float>());
+                c = 0.5f * (sb.di + sb.da);
+                _ranges[axis].exclude(dmin - c, dmax - c);
+            }
+            break;
+        default :
+            break;
+    }
+    return;
+}
+
+// Adjust the movement limits for the target to avoid having it collide
+// with the given neighbor slot. Also determine if there is in fact a collision
+// between the target and the given slot.
+bool ShiftCollider::mergeSlot(Segment *seg, Slot *slot, const SlotCollision *cslot, const Position &currShift,
+		bool isAfter,  // slot is logically after _target
+		bool sameCluster, bool &hasCol, bool isExclusion,
+        GR_MAYBE_UNUSED json * const dbgout )
+{
+    bool isCol = false;
+    const float sx = slot->origin().x - _origin.x + currShift.x;
+    const float sy = slot->origin().y - _origin.y + currShift.y;
+    const float sd = sx - sy;
+    const float ss = sx + sy;
+    float vmin, vmax;
+    float omin, omax, otmin, otmax;
+    float cmin, cmax;   // target limits
+    float torg;
+    const GlyphCache &gc = seg->getFace()->glyphs();
+    const unsigned short gid = slot->gid();
+    if (!gc.check(gid))
+        return false;
+    const BBox &bb = gc.getBoundingBBox(gid);
+
+    // SlotCollision * cslot = seg->collisionInfo(slot);
+    int orderFlags = 0;
+    bool sameClass = _seqProxClass == 0 && cslot->seqClass() == _seqClass;
+    if (sameCluster && _seqClass
+        && (sameClass || (_seqProxClass != 0 && cslot->seqClass() == _seqProxClass)))
+		// Force the target glyph to be in the specified direction from the slot we're testing.
+        orderFlags = _seqOrder;
+
+    // short circuit if only interested in direct collision and we are out of range
+    if (orderFlags || (sx + bb.xa + _margin >= _limit.bl.x && sx + bb.xi - _margin <= _limit.tr.x)
+                    || (sy + bb.ya + _margin >= _limit.bl.y && sy + bb.yi - _margin <= _limit.tr.y))
+
+    {
+        const float tx = _currOffset.x + _currShift.x;
+        const float ty = _currOffset.y + _currShift.y;
+        const float td = tx - ty;
+        const float ts = tx + ty;
+        const SlantBox &sb = gc.getBoundingSlantBox(gid);
+        const unsigned short tgid = _target->gid();
+        const BBox &tbb = gc.getBoundingBBox(tgid);
+        const SlantBox &tsb = gc.getBoundingSlantBox(tgid);
+        float seq_above_wt = cslot->seqAboveWt();
+        float seq_below_wt = cslot->seqBelowWt();
+        float seq_valign_wt = cslot->seqValignWt();
+        float lmargin;
+        // if isAfter, invert orderFlags for diagonal orders.
+        if (isAfter)
+        {
+            // invert appropriate bits
+            orderFlags ^= (sameClass ? 0x3F : 0x3);
+            // consider 2 bits at a time, non overlapping. If both bits set, clear them
+            orderFlags = orderFlags ^ ((((orderFlags >> 1) & orderFlags) & 0x15) * 3);
+        }
+
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+            dbgout->setenv(0, slot);
+#endif
+
+        // Process main bounding octabox.
+        for (int i = 0; i < 4; ++i)
+        {
+            switch (i) {
+                case 0 :	// x direction
+                    vmin = max(max(bb.xi - tbb.xa + sx, sb.di - tsb.da + ty + sd), sb.si - tsb.sa - ty + ss);
+                    vmax = min(min(bb.xa - tbb.xi + sx, sb.da - tsb.di + ty + sd), sb.sa - tsb.si - ty + ss);
+                    otmin = tbb.yi + ty;
+                    otmax = tbb.ya + ty;
+                    omin = bb.yi + sy;
+                    omax = bb.ya + sy;
+                    torg = _currOffset.x;
+                    cmin = _limit.bl.x + torg;
+                    cmax = _limit.tr.x - tbb.xi + tbb.xa + torg;
+                    lmargin = _margin;
+                    break;
+                case 1 :	// y direction
+                    vmin = max(max(bb.yi - tbb.ya + sy, tsb.di - sb.da + tx - sd), sb.si - tsb.sa - tx + ss);
+                    vmax = min(min(bb.ya - tbb.yi + sy, tsb.da - sb.di + tx - sd), sb.sa - tsb.si - tx + ss);
+                    otmin = tbb.xi + tx;
+                    otmax = tbb.xa + tx;
+                    omin = bb.xi + sx;
+                    omax = bb.xa + sx;
+                    torg = _currOffset.y;
+                    cmin = _limit.bl.y + torg;
+                    cmax = _limit.tr.y - tbb.yi + tbb.ya + torg;
+                    lmargin = _margin;
+                    break;
+                case 2 :    // sum - moving along the positively-sloped vector, so the boundaries are the
+                            // negatively-sloped boundaries.
+                    vmin = max(max(sb.si - tsb.sa + ss, 2 * (bb.yi - tbb.ya + sy) + td), 2 * (bb.xi - tbb.xa + sx) - td);
+                    vmax = min(min(sb.sa - tsb.si + ss, 2 * (bb.ya - tbb.yi + sy) + td), 2 * (bb.xa - tbb.xi + sx) - td);
+                    otmin = tsb.di + td;
+                    otmax = tsb.da + td;
+                    omin = sb.di + sd;
+                    omax = sb.da + sd;
+                    torg = _currOffset.x + _currOffset.y;
+                    cmin = _limit.bl.x + _limit.bl.y + torg;
+                    cmax = _limit.tr.x + _limit.tr.y - tsb.si + tsb.sa + torg;
+                    lmargin = _margin / ISQRT2;
+                    break;
+                case 3 :    // diff - moving along the negatively-sloped vector, so the boundaries are the
+                            // positively-sloped boundaries.
+                    vmin = max(max(sb.di - tsb.da + sd, 2 * (bb.xi - tbb.xa + sx) - ts), -2 * (bb.ya - tbb.yi + sy) + ts);
+                    vmax = min(min(sb.da - tsb.di + sd, 2 * (bb.xa - tbb.xi + sx) - ts), -2 * (bb.yi - tbb.ya + sy) + ts);
+                    otmin = tsb.si + ts;
+                    otmax = tsb.sa + ts;
+                    omin = sb.si + ss;
+                    omax = sb.sa + ss;
+                    torg = _currOffset.x - _currOffset.y;
+                    cmin = _limit.bl.x - _limit.tr.y + torg;
+                    cmax = _limit.tr.x - _limit.bl.y - tsb.di + tsb.da + torg;
+                    lmargin = _margin / ISQRT2;
+                    break;
+                default :
+                    continue;
+            }
+
+#if !defined GRAPHITE2_NTRACING
+            if (dbgout)
+                dbgout->setenv(1, reinterpret_cast<void *>(-1));
+#define DBGTAG(x) if (dbgout) dbgout->setenv(1, reinterpret_cast<void *>(-x));
+#else
+#define DBGTAG(x)
+#endif
+
+            if (orderFlags)
+            {
+                Position org(tx, ty);
+                float xminf = _limit.bl.x + _currOffset.x + tbb.xi;
+                float xpinf = _limit.tr.x + _currOffset.x + tbb.xa;
+                float ypinf = _limit.tr.y + _currOffset.y + tbb.ya;
+                float yminf = _limit.bl.y + _currOffset.y + tbb.yi;
+                switch (orderFlags) {
+                    case SlotCollision::SEQ_ORDER_RIGHTUP :
+                    {
+                        float r1Xedge = cslot->seqAboveXoff() + 0.5f * (bb.xi + bb.xa) + sx;
+                        float r3Xedge = cslot->seqBelowXlim() + bb.xa + sx + 0.5f * (tbb.xa - tbb.xi);
+                        float r2Yedge = 0.5f * (bb.yi + bb.ya) + sy;
+
+                        // DBGTAG(1x) means the regions are up and right
+                        // region 1
+                        DBGTAG(11)
+                        addBox_slope(true, Rect(Position(xminf, r2Yedge), Position(r1Xedge, ypinf)),
+                                        tbb, tsb, org, 0, seq_above_wt, true, i);
+                        // region 2
+                        DBGTAG(12)
+                        removeBox(Rect(Position(xminf, yminf), Position(r3Xedge, r2Yedge)), tbb, tsb, org, i);
+                        // region 3, which end is zero is irrelevant since m weight is 0
+                        DBGTAG(13)
+                        addBox_slope(true, Rect(Position(r3Xedge, yminf), Position(xpinf, r2Yedge - cslot->seqValignHt())),
+                                        tbb, tsb, org, seq_below_wt, 0, true, i);
+                        // region 4
+                        DBGTAG(14)
+                        addBox_slope(false, Rect(Position(sx + bb.xi, r2Yedge), Position(xpinf, r2Yedge + cslot->seqValignHt())),
+                                        tbb, tsb, org, 0, seq_valign_wt, true, i);
+                        // region 5
+                        DBGTAG(15)
+                        addBox_slope(false, Rect(Position(sx + bb.xi, r2Yedge - cslot->seqValignHt()), Position(xpinf, r2Yedge)),
+                                        tbb, tsb, org, seq_below_wt, seq_valign_wt, false, i);
+                        break;
+                    }
+                    case SlotCollision::SEQ_ORDER_LEFTDOWN :
+                    {
+                        float r1Xedge = 0.5f * (bb.xi + bb.xa) + cslot->seqAboveXoff() + sx;
+                        float r3Xedge = bb.xi - cslot->seqBelowXlim() + sx - 0.5f * (tbb.xa - tbb.xi);
+                        float r2Yedge = 0.5f * (bb.yi + bb.ya) + sy;
+                        // DBGTAG(2x) means the regions are up and right
+                        // region 1
+                        DBGTAG(21)
+                        addBox_slope(true, Rect(Position(r1Xedge, yminf), Position(xpinf, r2Yedge)),
+                                        tbb, tsb, org, 0, seq_above_wt, false, i);
+                        // region 2
+                        DBGTAG(22)
+                        removeBox(Rect(Position(r3Xedge, r2Yedge), Position(xpinf, ypinf)), tbb, tsb, org, i);
+                        // region 3
+                        DBGTAG(23)
+                        addBox_slope(true, Rect(Position(xminf, r2Yedge - cslot->seqValignHt()), Position(r3Xedge, ypinf)),
+                                        tbb, tsb, org, seq_below_wt, 0, false, i);
+                        // region 4
+                        DBGTAG(24)
+                        addBox_slope(false, Rect(Position(xminf, r2Yedge), Position(sx + bb.xa, r2Yedge + cslot->seqValignHt())),
+                                        tbb, tsb, org, 0, seq_valign_wt, true, i);
+                        // region 5
+                        DBGTAG(25)
+                        addBox_slope(false, Rect(Position(xminf, r2Yedge - cslot->seqValignHt()),
+                                        Position(sx + bb.xa, r2Yedge)), tbb, tsb, org, seq_below_wt, seq_valign_wt, false, i);
+                        break;
+                    }
+                    case SlotCollision::SEQ_ORDER_NOABOVE : // enforce neighboring glyph being above
+                        DBGTAG(31);
+                        removeBox(Rect(Position(bb.xi - tbb.xa + sx, sy + bb.ya),
+                                        Position(bb.xa - tbb.xi + sx, ypinf)), tbb, tsb, org, i);
+                        break;
+                    case SlotCollision::SEQ_ORDER_NOBELOW :	// enforce neighboring glyph being below
+                        DBGTAG(32);
+                        removeBox(Rect(Position(bb.xi - tbb.xa + sx, yminf),
+                                        Position(bb.xa - tbb.xi + sx, sy + bb.yi)), tbb, tsb, org, i);
+                        break;
+                    case SlotCollision::SEQ_ORDER_NOLEFT :  // enforce neighboring glyph being to the left
+                        DBGTAG(33)
+                        removeBox(Rect(Position(xminf, bb.yi - tbb.ya + sy),
+                                        Position(bb.xi - tbb.xa + sx, bb.ya - tbb.yi + sy)), tbb, tsb, org, i);
+                        break;
+                    case SlotCollision::SEQ_ORDER_NORIGHT : // enforce neighboring glyph being to the right
+                        DBGTAG(34)
+                        removeBox(Rect(Position(bb.xa - tbb.xi + sx, bb.yi - tbb.ya + sy),
+                                        Position(xpinf, bb.ya - tbb.yi + sy)), tbb, tsb, org, i);
+                        break;
+                    default :
+                        break;
+                }
+            }
+
+            if (vmax < cmin - lmargin || vmin > cmax + lmargin || omax < otmin - lmargin || omin > otmax + lmargin)
+                continue;
+
+            // Process sub-boxes that are defined for this glyph.
+            // We only need to do this if there was in fact a collision with the main octabox.
+            uint8 numsub = gc.numSubBounds(gid);
+            if (numsub > 0)
+            {
+                bool anyhits = false;
+                for (int j = 0; j < numsub; ++j)
+                {
+                    const BBox &sbb = gc.getSubBoundingBBox(gid, j);
+                    const SlantBox &ssb = gc.getSubBoundingSlantBox(gid, j);
+                    switch (i) {
+                        case 0 :    // x
+                            vmin = max(max(sbb.xi-tbb.xa+sx, ssb.di-tsb.da+sd+ty), ssb.si-tsb.sa+ss-ty);
+                            vmax = min(min(sbb.xa-tbb.xi+sx, ssb.da-tsb.di+sd+ty), ssb.sa-tsb.si+ss-ty);
+                            omin = sbb.yi + sy;
+                            omax = sbb.ya + sy;
+                            break;
+                        case 1 :    // y
+                            vmin = max(max(sbb.yi-tbb.ya+sy, tsb.di-ssb.da-sd+tx), ssb.si-tsb.sa+ss-tx);
+                            vmax = min(min(sbb.ya-tbb.yi+sy, tsb.da-ssb.di-sd+tx), ssb.sa-tsb.si+ss-tx);
+                            omin = sbb.xi + sx;
+                            omax = sbb.xa + sx;
+                            break;
+                        case 2 :    // sum
+                            vmin = max(max(ssb.si-tsb.sa+ss, 2*(sbb.yi-tbb.ya+sy)+td), 2*(sbb.xi-tbb.xa+sx)-td);
+                            vmax = min(min(ssb.sa-tsb.si+ss, 2*(sbb.ya-tbb.yi+sy)+td), 2*(sbb.xa-tbb.xi+sx)-td);
+                            omin = ssb.di + sd;
+                            omax = ssb.da + sd;
+                            break;
+                        case 3 :    // diff
+                            vmin = max(max(ssb.di-tsb.da+sd, 2*(sbb.xi-tbb.xa+sx)-ts), -2*(sbb.ya-tbb.yi+sy)+ts);
+                            vmax = min(min(ssb.da-tsb.di+sd, 2*(sbb.xa-tbb.xi+sx)-ts), -2*(sbb.yi-tbb.ya+sy)+ts);
+                            omin = ssb.si + ss;
+                            omax = ssb.sa + ss;
+                            break;
+                    }
+                    if (vmax < cmin - lmargin || vmin > cmax + lmargin || omax < otmin - lmargin || omin > otmax + lmargin)
+                        continue;
+
+#if !defined GRAPHITE2_NTRACING
+                    if (dbgout)
+                        dbgout->setenv(1, reinterpret_cast<void *>(j));
+#endif
+                    if (omin > otmax)
+                        _ranges[i].weightedAxis(i, vmin - lmargin, vmax + lmargin, 0, 0, 0, 0, 0,
+                                                sqr(lmargin - omin + otmax) * _marginWt, false);
+                    else if (omax < otmin)
+                        _ranges[i].weightedAxis(i, vmin - lmargin, vmax + lmargin, 0, 0, 0, 0, 0,
+                                                sqr(lmargin - otmin + omax) * _marginWt, false);
+                    else
+                        _ranges[i].exclude_with_margins(vmin, vmax, i);
+                    anyhits = true;
+                }
+                if (anyhits)
+                    isCol = true;
+            }
+            else // no sub-boxes
+            {
+#if !defined GRAPHITE2_NTRACING
+                    if (dbgout)
+                        dbgout->setenv(1, reinterpret_cast<void *>(-1));
+#endif
+                isCol = true;
+                if (omin > otmax)
+                    _ranges[i].weightedAxis(i, vmin - lmargin, vmax + lmargin, 0, 0, 0, 0, 0,
+                                            sqr(lmargin - omin + otmax) * _marginWt, false);
+                else if (omax < otmin)
+                    _ranges[i].weightedAxis(i, vmin - lmargin, vmax + lmargin, 0, 0, 0, 0, 0,
+                                            sqr(lmargin - otmin + omax) * _marginWt, false);
+                else
+                    _ranges[i].exclude_with_margins(vmin, vmax, i);
+
+            }
+        }
+    }
+    bool res = true;
+    if (cslot->exclGlyph() > 0 && gc.check(cslot->exclGlyph()) && !isExclusion)
+    {
+        // Set up the bogus slot representing the exclusion glyph.
+        Slot *exclSlot = seg->newSlot();
+        if (!exclSlot)
+            return res;
+        exclSlot->setGlyph(seg, cslot->exclGlyph());
+        Position exclOrigin(slot->origin() + cslot->exclOffset());
+        exclSlot->origin(exclOrigin);
+        SlotCollision exclInfo(seg, exclSlot);
+        res &= mergeSlot(seg, exclSlot, &exclInfo, currShift, isAfter, sameCluster, isCol, true, dbgout );
+        seg->freeSlot(exclSlot);
+    }
+    hasCol |= isCol;
+    return res;
+
+}   // end of ShiftCollider::mergeSlot
+
+
+// Figure out where to move the target glyph to, and return the amount to shift by.
+Position ShiftCollider::resolve(GR_MAYBE_UNUSED Segment *seg, bool &isCol, GR_MAYBE_UNUSED json * const dbgout)
+{
+    float tbase;
+    float totalCost = (float)(std::numeric_limits<float>::max() / 2);
+    Position resultPos = Position(0, 0);
+#if !defined GRAPHITE2_NTRACING
+	int bestAxis = -1;
+    if (dbgout)
+    {
+		outputJsonDbgStartSlot(dbgout, seg);
+        *dbgout << "vectors" << json::array;
+    }
+#endif
+    isCol = true;
+    for (int i = 0; i < 4; ++i)
+    {
+        float bestCost = -1;
+        float bestPos;
+        // Calculate the margin depending on whether we are moving diagonally or not:
+        switch (i) {
+            case 0 :	// x direction
+                tbase = _currOffset.x;
+                break;
+            case 1 :	// y direction
+                tbase = _currOffset.y;
+                break;
+            case 2 :	// sum (negatively-sloped diagonals)
+                tbase = _currOffset.x + _currOffset.y;
+                break;
+            case 3 :	// diff (positively-sloped diagonals)
+                tbase = _currOffset.x - _currOffset.y;
+                break;
+        }
+        Position testp;
+        bestPos = _ranges[i].closest(0, bestCost) - tbase;     // Get the best relative position
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+            outputJsonDbgOneVector(dbgout, seg, i, tbase, bestCost, bestPos) ;
+#endif
+        if (bestCost >= 0.0f)
+        {
+            isCol = false;
+            switch (i) {
+                case 0 : testp = Position(bestPos, _currShift.y); break;
+                case 1 : testp = Position(_currShift.x, bestPos); break;
+                case 2 : testp = Position(0.5f * (_currShift.x - _currShift.y + bestPos), 0.5f * (_currShift.y - _currShift.x + bestPos)); break;
+                case 3 : testp = Position(0.5f * (_currShift.x + _currShift.y + bestPos), 0.5f * (_currShift.x + _currShift.y - bestPos)); break;
+            }
+            if (bestCost < totalCost - 0.01f)
+            {
+                totalCost = bestCost;
+                resultPos = testp;
+#if !defined GRAPHITE2_NTRACING
+                bestAxis = i;
+#endif
+            }
+        }
+    }  // end of loop over 4 directions
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+        outputJsonDbgEndSlot(dbgout, resultPos, bestAxis, isCol);
+#endif
+
+    return resultPos;
+
+}   // end of ShiftCollider::resolve
+
+
+#if !defined GRAPHITE2_NTRACING
+
+void ShiftCollider::outputJsonDbg(json * const dbgout, Segment *seg, int axis)
+{
+    int axisMax = axis;
+    if (axis < 0) // output all axes
+    {
+        *dbgout << "gid" << _target->gid()
+            << "limit" << _limit
+            << "target" << json::object
+                << "origin" << _target->origin()
+                << "margin" << _margin
+                << "bbox" << seg->theGlyphBBoxTemporary(_target->gid())
+                << "slantbox" << seg->getFace()->glyphs().slant(_target->gid())
+                << json::close; // target object
+        *dbgout << "ranges" << json::array;
+        axis = 0;
+        axisMax = 3;
+    }
+    for (int iAxis = axis; iAxis <= axisMax; ++iAxis)
+    {
+        *dbgout << json::flat << json::array << _ranges[iAxis].position();
+        for (Zones::const_iterator s = _ranges[iAxis].begin(), e = _ranges[iAxis].end(); s != e; ++s)
+            *dbgout << json::flat << json::array
+                        << Position(s->x, s->xm) << s->sm << s->smx << s->c
+                    << json::close;
+        *dbgout << json::close;
+    }
+    if (axis < axisMax) // looped through the _ranges array for all axes
+        *dbgout << json::close; // ranges array
+}
+
+void ShiftCollider::outputJsonDbgStartSlot(json * const dbgout, Segment *seg)
+{
+        *dbgout << json::object // slot - not closed till the end of the caller method
+                << "slot" << objectid(dslot(seg, _target))
+				<< "gid" << _target->gid()
+                << "limit" << _limit
+                << "target" << json::object
+                    << "origin" << _origin
+                    << "currShift" << _currShift
+                    << "currOffset" << seg->collisionInfo(_target)->offset()
+                    << "bbox" << seg->theGlyphBBoxTemporary(_target->gid())
+                    << "slantBox" << seg->getFace()->glyphs().slant(_target->gid())
+                    << "fix" << "shift";
+        *dbgout     << json::close; // target object
+}
+
+void ShiftCollider::outputJsonDbgEndSlot(GR_MAYBE_UNUSED json * const dbgout,
+	 Position resultPos, int bestAxis, bool isCol)
+{
+    *dbgout << json::close // vectors array
+    << "result" << resultPos
+	//<< "scraping" << _scraping[bestAxis]
+	<< "bestAxis" << bestAxis
+    << "stillBad" << isCol
+    << json::close; // slot object
+}
+
+void ShiftCollider::outputJsonDbgOneVector(json * const dbgout, Segment *seg, int axis,
+	float tleft, float bestCost, float bestVal)
+{
+	const char * label;
+	switch (axis)
+	{
+		case 0:	label = "x";			break;
+		case 1:	label = "y";			break;
+		case 2:	label = "sum (NE-SW)";	break;
+		case 3:	label = "diff (NW-SE)";	break;
+		default: label = "???";			break;
+	}
+
+	*dbgout << json::object // vector
+		<< "direction" << label
+		<< "targetMin" << tleft;
+
+	outputJsonDbgRemovals(dbgout, axis, seg);
+
+    *dbgout << "ranges";
+    outputJsonDbg(dbgout, seg, axis);
+
+    *dbgout << "bestCost" << bestCost
+        << "bestVal" << bestVal + tleft
+        << json::close; // vectors object
+}
+
+void ShiftCollider::outputJsonDbgRemovals(json * const dbgout, int axis, Segment *seg)
+{
+    *dbgout << "removals" << json::array;
+    _ranges[axis].jsonDbgOut(seg);
+    *dbgout << json::close; // removals array
+}
+
+#endif // !defined GRAPHITE2_NTRACING
+
+
+////    KERN-COLLIDER    ////
+
+inline
+static float localmax (float al, float au, float bl, float bu, float x)
+{
+    if (al < bl)
+    { if (au < bu) return au < x ? au : x; }
+    else if (au > bu) return bl < x ? bl : x;
+    return x;
+}
+
+inline
+static float localmin(float al, float au, float bl, float bu, float x)
+{
+    if (bl > al)
+    { if (bu > au) return bl > x ? bl : x; }
+    else if (au > bu) return al > x ? al : x;
+    return x;
+}
+
+// Return the given edge of the glyph at height y, taking any slant box into account.
+static float get_edge(Segment *seg, const Slot *s, const Position &shift, float y, float width, float margin, bool isRight)
+{
+    const GlyphCache &gc = seg->getFace()->glyphs();
+    unsigned short gid = s->gid();
+    float sx = s->origin().x + shift.x;
+    float sy = s->origin().y + shift.y;
+    uint8 numsub = gc.numSubBounds(gid);
+    float res = isRight ? (float)-1e38 : (float)1e38;
+
+    if (numsub > 0)
+    {
+        for (int i = 0; i < numsub; ++i)
+        {
+            const BBox &sbb = gc.getSubBoundingBBox(gid, i);
+            const SlantBox &ssb = gc.getSubBoundingSlantBox(gid, i);
+            if (sy + sbb.yi - margin > y + width / 2 || sy + sbb.ya + margin < y - width / 2)
+                continue;
+            if (isRight)
+            {
+                float x = sx + sbb.xa + margin;
+                if (x > res)
+                {
+                    float td = sx - sy + ssb.da + margin + y;
+                    float ts = sx + sy + ssb.sa + margin - y;
+                    x = localmax(td - width / 2, td + width / 2,  ts - width / 2, ts + width / 2, x);
+                    if (x > res)
+                        res = x;
+                }
+            }
+            else
+            {
+                float x = sx + sbb.xi - margin;
+                if (x < res)
+                {
+                    float td = sx - sy + ssb.di - margin + y;
+                    float ts = sx + sy + ssb.si - margin - y;
+                    x = localmin(td - width / 2, td + width / 2, ts - width / 2, ts + width / 2, x);
+                    if (x < res)
+                        res = x;
+                }
+            }
+        }
+    }
+    else
+    {
+        const BBox &bb = gc.getBoundingBBox(gid);
+        const SlantBox &sb = gc.getBoundingSlantBox(gid);
+        if (sy + bb.yi - margin > y + width / 2 || sy + bb.ya + margin < y - width / 2)
+            return res;
+        float td = sx - sy + y;
+        float ts = sx + sy - y;
+        if (isRight)
+            res = localmax(td + sb.da - width / 2, td + sb.da + width / 2, ts + sb.sa - width / 2, ts + sb.sa + width / 2, sx + bb.xa) + margin;
+        else
+            res = localmin(td + sb.di - width / 2, td + sb.di + width / 2, ts + sb.si - width / 2, ts + sb.si + width / 2, sx + bb.xi) - margin;
+    }
+    return res;
+}
+
+
+bool KernCollider::initSlot(Segment *seg, Slot *aSlot, const Rect &limit, float margin,
+    const Position &currShift, const Position &offsetPrev, int dir,
+    float ymin, float ymax, GR_MAYBE_UNUSED json * const dbgout)
+{
+    const GlyphCache &gc = seg->getFace()->glyphs();
+    const Slot *base = aSlot;
+    // const Slot *last = aSlot;
+    const Slot *s;
+    int numSlices;
+    while (base->attachedTo())
+        base = base->attachedTo();
+    if (margin < 10) margin = 10;
+
+    _limit = limit;
+    _offsetPrev = offsetPrev; // kern from a previous pass
+
+    // Calculate the height of the glyph and how many horizontal slices to use.
+    if (_maxy >= 1e37f)
+    {
+        _sliceWidth = margin / 1.5f;
+        _maxy = ymax + margin;
+        _miny = ymin - margin;
+        numSlices = int((_maxy - _miny + 2) / (_sliceWidth / 1.5f) + 1.f);  // +2 helps with rounding errors
+        _edges.clear();
+        _edges.insert(_edges.begin(), numSlices, (dir & 1) ? 1e38f : -1e38f);
+        _xbound = (dir & 1) ? (float)1e38f : (float)-1e38f;
+    }
+    else if (_maxy != ymax || _miny != ymin)
+    {
+        if (_miny != ymin)
+        {
+            numSlices = int((ymin - margin - _miny) / _sliceWidth - 1);
+            _miny += numSlices * _sliceWidth;
+            if (numSlices < 0)
+                _edges.insert(_edges.begin(), -numSlices, (dir & 1) ? 1e38f : -1e38f);
+            else if ((unsigned)numSlices < _edges.size())    // this shouldn't fire since we always grow the range
+            {
+                Vector<float>::iterator e = _edges.begin();
+                while (numSlices--)
+                    ++e;
+                _edges.erase(_edges.begin(), e);
+            }
+        }
+        if (_maxy != ymax)
+        {
+            numSlices = int((ymax + margin - _miny) / _sliceWidth + 1);
+            _maxy = numSlices * _sliceWidth + _miny;
+            if (numSlices > (int)_edges.size())
+                _edges.insert(_edges.end(), numSlices - _edges.size(), (dir & 1) ? 1e38f : -1e38f);
+            else if (numSlices < (int)_edges.size())   // this shouldn't fire since we always grow the range
+            {
+                while ((int)_edges.size() > numSlices)
+                    _edges.pop_back();
+            }
+        }
+        goto done;
+    }
+    numSlices = int(_edges.size());
+
+#if !defined GRAPHITE2_NTRACING
+    // Debugging
+    _seg = seg;
+    _slotNear.clear();
+    _slotNear.insert(_slotNear.begin(), numSlices, NULL);
+    _nearEdges.clear();
+    _nearEdges.insert(_nearEdges.begin(), numSlices, (dir & 1) ? -1e38f : +1e38f);
+#endif
+
+    // Determine the trailing edge of each slice (ie, left edge for a RTL glyph).
+    for (s = base; s; s = s->nextInCluster(s))
+    {
+        SlotCollision *c = seg->collisionInfo(s);
+        if (!gc.check(s->gid()))
+            return false;
+        const BBox &bs = gc.getBoundingBBox(s->gid());
+        float x = s->origin().x + c->shift().x + ((dir & 1) ? bs.xi : bs.xa);
+        // Loop over slices.
+        // Note smin might not be zero if glyph s is not at the bottom of the cluster; similarly for smax.
+        float toffset = c->shift().y - _miny + 1 + s->origin().y;
+        int smin = max(0, int((bs.yi + toffset) / _sliceWidth));
+        int smax = min(numSlices - 1, int((bs.ya + toffset) / _sliceWidth + 1));
+        for (int i = smin; i <= smax; ++i)
+        {
+            float t;
+            float y = _miny - 1 + (i + .5f) * _sliceWidth; // vertical center of slice
+            if ((dir & 1) && x < _edges[i])
+            {
+                t = get_edge(seg, s, c->shift(), y, _sliceWidth, margin, false);
+                if (t < _edges[i])
+                {
+                    _edges[i] = t;
+                    if (t < _xbound)
+                        _xbound = t;
+                }
+            }
+            else if (!(dir & 1) && x > _edges[i])
+            {
+                t = get_edge(seg, s, c->shift(), y, _sliceWidth, margin, true);
+                if (t > _edges[i])
+                {
+                    _edges[i] = t;
+                    if (t > _xbound)
+                        _xbound = t;
+                }
+            }
+        }
+    }
+    done:
+    _mingap = (float)1e37;      // less than 1e38 s.t. 1e38-_mingap is really big
+    _target = aSlot;
+    _margin = margin;
+    _currShift = currShift;
+    return true;
+}   // end of KernCollider::initSlot
+
+
+// Determine how much the target slot needs to kern away from the given slot.
+// In other words, merge information from given slot's position with what the target slot knows
+// about how it can kern.
+// Return false if we know there is no collision, true if we think there might be one.
+bool KernCollider::mergeSlot(Segment *seg, Slot *slot, const Position &currShift, float currSpace, int dir, GR_MAYBE_UNUSED json * const dbgout)
+{
+    int rtl = (dir & 1) * 2 - 1;
+    if (!seg->getFace()->glyphs().check(slot->gid()))
+        return false;
+    const Rect &bb = seg->theGlyphBBoxTemporary(slot->gid());
+    const float sx = slot->origin().x + currShift.x;
+    float x = (sx + (rtl > 0 ? bb.tr.x : bb.bl.x)) * rtl;
+    // this isn't going to reduce _mingap so skip
+    if (_hit && x < rtl * (_xbound - _mingap - currSpace))
+        return false;
+
+    const float sy = slot->origin().y + currShift.y;
+    int smin = max(1, int((bb.bl.y + (1 - _miny + sy)) / _sliceWidth + 1)) - 1;
+    int smax = min((int)_edges.size() - 2, int((bb.tr.y + (1 - _miny + sy)) / _sliceWidth + 1)) + 1;
+    if (smin > smax)
+        return false;
+    bool collides = false;
+    bool nooverlap = true;
+
+    for (int i = smin; i <= smax; ++i)
+    {
+        float here = _edges[i] * rtl;
+        if (here > (float)9e37)
+            continue;
+        if (!_hit || x > here - _mingap - currSpace)
+        {
+            float y = (float)(_miny - 1 + (i + .5f) * _sliceWidth);  // vertical center of slice
+            // 2 * currSpace to account for the space that is already separating them and the space we want to add
+            float m = get_edge(seg, slot, currShift, y, _sliceWidth, 0., rtl > 0) * rtl + 2 * currSpace;
+            if (m < (float)-8e37)       // only true if the glyph has a gap in it
+                continue;
+            nooverlap = false;
+            float t = here - m;
+            // _mingap is positive to shrink
+            if (t < _mingap || (!_hit && !collides))
+            {
+                _mingap = t;
+                collides = true;
+            }
+#if !defined GRAPHITE2_NTRACING
+            // Debugging - remember the closest neighboring edge for this slice.
+            if (m > rtl * _nearEdges[i])
+            {
+                _slotNear[i] = slot;
+                _nearEdges[i] = m * rtl;
+            }
+#endif
+        }
+        else
+            nooverlap = false;
+    }
+    if (nooverlap)
+        _mingap = max(_mingap, _xbound - rtl * (currSpace + _margin + x));
+    if (collides && !nooverlap)
+        _hit = true;
+    return collides | nooverlap;   // note that true is not a necessarily reliable value
+
+}   // end of KernCollider::mergeSlot
+
+
+// Return the amount to kern by.
+Position KernCollider::resolve(GR_MAYBE_UNUSED Segment *seg, GR_MAYBE_UNUSED Slot *slot,
+        int dir, GR_MAYBE_UNUSED json * const dbgout)
+{
+    float resultNeeded = (1 - 2 * (dir & 1)) * _mingap;
+    // float resultNeeded = (1 - 2 * (dir & 1)) * (_mingap - margin);
+    float result = min(_limit.tr.x - _offsetPrev.x, max(resultNeeded, _limit.bl.x - _offsetPrev.x));
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+    {
+        *dbgout << json::object // slot
+                << "slot" << objectid(dslot(seg, _target))
+				<< "gid" << _target->gid()
+                << "limit" << _limit
+                << "miny" << _miny
+                << "maxy" << _maxy
+                << "slicewidth" << _sliceWidth
+                << "target" << json::object
+                    << "origin" << _target->origin()
+                    //<< "currShift" << _currShift
+                    << "offsetPrev" << _offsetPrev
+                    << "bbox" << seg->theGlyphBBoxTemporary(_target->gid())
+                    << "slantBox" << seg->getFace()->glyphs().slant(_target->gid())
+                    << "fix" << "kern"
+                    << json::close; // target object
+
+        *dbgout << "slices" << json::array;
+        for (int is = 0; is < (int)_edges.size(); is++)
+        {
+            *dbgout << json::flat << json::object
+                << "i" << is
+                << "targetEdge" << _edges[is]
+                << "neighbor" << objectid(dslot(seg, _slotNear[is]))
+                << "nearEdge" << _nearEdges[is]
+                << json::close;
+        }
+        *dbgout << json::close; // slices array
+
+        *dbgout
+            << "xbound" << _xbound
+            << "minGap" << _mingap
+            << "needed" << resultNeeded
+            << "result" << result
+            << "stillBad" << (result != resultNeeded)
+            << json::close; // slot object
+    }
+#endif
+
+    return Position(result, 0.);
+
+}   // end of KernCollider::resolve
+
+void KernCollider::shift(const Position &mv, int dir)
+{
+    for (Vector<float>::iterator e = _edges.begin(); e != _edges.end(); ++e)
+        *e += mv.x;
+    _xbound += (1 - 2 * (dir & 1)) * mv.x;
+}
+
+////    SLOT-COLLISION    ////
+
+// Initialize the collision attributes for the given slot.
+SlotCollision::SlotCollision(Segment *seg, Slot *slot)
+{
+    initFromSlot(seg, slot);
+}
+
+void SlotCollision::initFromSlot(Segment *seg, Slot *slot)
+{
+    // Initialize slot attributes from glyph attributes.
+	// The order here must match the order in the grcompiler code,
+	// GrcSymbolTable::AssignInternalGlyphAttrIDs.
+    uint16 gid = slot->gid();
+    uint16 aCol = seg->silf()->aCollision(); // flags attr ID
+    const GlyphFace * glyphFace = seg->getFace()->glyphs().glyphSafe(gid);
+    if (!glyphFace)
+        return;
+    const sparse &p = glyphFace->attrs();
+    _flags = p[aCol];
+    _limit = Rect(Position(int16(p[aCol+1]), int16(p[aCol+2])),
+                  Position(int16(p[aCol+3]), int16(p[aCol+4])));
+    _margin = p[aCol+5];
+    _marginWt = p[aCol+6];
+
+    _seqClass = p[aCol+7];
+	_seqProxClass = p[aCol+8];
+    _seqOrder = p[aCol+9];
+	_seqAboveXoff = p[aCol+10];
+	_seqAboveWt = p[aCol+11];
+	_seqBelowXlim = p[aCol+12];
+	_seqBelowWt = p[aCol+13];
+	_seqValignHt = p[aCol+14];
+	_seqValignWt = p[aCol+15];
+
+    // These attributes do not have corresponding glyph attribute:
+    _exclGlyph = 0;
+    _exclOffset = Position(0, 0);
+}
+
+float SlotCollision::getKern(int dir) const
+{
+    if ((_flags & SlotCollision::COLL_KERN) != 0)
+        return float(_shift.x * ((dir & 1) ? -1 : 1));
+    else
+    	return 0;
+}
+
+bool SlotCollision::ignore() const
+{
+	return ((flags() & SlotCollision::COLL_IGNORE) || (flags() & SlotCollision::COLL_ISSPACE));
+}
diff --git a/thirdparty/graphite/src/Decompressor.cpp b/thirdparty/graphite/src/Decompressor.cpp
new file mode 100644
index 0000000000..42dc9113e5
--- /dev/null
+++ b/thirdparty/graphite/src/Decompressor.cpp
@@ -0,0 +1,125 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2015, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cassert>
+
+#include "inc/Decompressor.h"
+#include "inc/Compression.h"
+
+using namespace lz4;
+
+namespace {
+
+inline
+u32 read_literal(u8 const * &s, u8 const * const e, u32 l) {
+    if (l == 15 && s != e)
+    {
+        u8 b = 0;
+        do { l += b = *s++; } while(b==0xff && s != e);
+    }
+    return l;
+}
+
+bool read_sequence(u8 const * &src, u8 const * const end, u8 const * &literal,
+                    u32 & literal_len, u32 & match_len, u32 & match_dist)
+{
+    u8 const token = *src++;
+
+    literal_len = read_literal(src, end, token >> 4);
+    literal = src;
+    src += literal_len;
+
+    // Normal exit for end of stream, wrap arround check and parital match check.
+    if (src > end - sizeof(u16) || src < literal)
+        return false;
+
+    match_dist  = *src++;
+    match_dist |= *src++ << 8;
+    match_len = read_literal(src, end, token & 0xf) + MINMATCH;
+
+    // Malformed stream check.
+    return src <= end-MINCODA;
+}
+
+}
+
+int lz4::decompress(void const *in, size_t in_size, void *out, size_t out_size)
+{
+    if (out_size <= in_size || in_size < MINSRCSIZE)
+        return -1;
+
+    u8 const *       src     = static_cast<u8 const *>(in),
+             *       literal = 0,
+             * const src_end = src + in_size;
+
+    u8 *       dst     = static_cast<u8*>(out),
+       * const dst_end = dst + out_size;
+
+    // Check the in and out size hasn't wrapped around.
+    if (src >= src_end || dst >= dst_end)
+        return -1;
+
+    u32 literal_len = 0,
+        match_len = 0,
+        match_dist = 0;
+
+    while (read_sequence(src, src_end, literal, literal_len, match_len,
+                         match_dist))
+    {
+        if (literal_len != 0)
+        {
+            // Copy in literal. At this point the a minimal literal + minminal
+            // match plus the coda (1 + 2 + 5) must be 8 bytes or more allowing
+            // us to remain within the src buffer for an overrun_copy on
+            // machines upto 64 bits.
+            if (align(literal_len) > out_size)
+                return -1;
+            dst = overrun_copy(dst, literal, literal_len);
+            out_size -= literal_len;
+        }
+
+        // Copy, possibly repeating, match from earlier in the
+        //  decoded output.
+        u8 const * const pcpy = dst - match_dist;
+        if (pcpy < static_cast<u8*>(out)
+              || match_len > unsigned(out_size - LASTLITERALS)
+              // Wrap around checks:
+              || out_size < LASTLITERALS || pcpy >= dst)
+            return -1;
+        if (dst > pcpy+sizeof(unsigned long)
+            && align(match_len) <= out_size)
+            dst = overrun_copy(dst, pcpy, match_len);
+        else
+            dst = safe_copy(dst, pcpy, match_len);
+        out_size -= match_len;
+    }
+
+    if (literal > src_end - literal_len || literal_len > out_size)
+        return -1;
+    dst = fast_copy(dst, literal, literal_len);
+
+    return int(dst - (u8*)out);
+}
diff --git a/thirdparty/graphite/src/Face.cpp b/thirdparty/graphite/src/Face.cpp
new file mode 100644
index 0000000000..3e106050d7
--- /dev/null
+++ b/thirdparty/graphite/src/Face.cpp
@@ -0,0 +1,366 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cstring>
+#include "graphite2/Segment.h"
+#include "inc/CmapCache.h"
+#include "inc/debug.h"
+#include "inc/Decompressor.h"
+#include "inc/Endian.h"
+#include "inc/Face.h"
+#include "inc/FileFace.h"
+#include "inc/GlyphFace.h"
+#include "inc/json.h"
+#include "inc/Segment.h"
+#include "inc/NameTable.h"
+#include "inc/Error.h"
+
+using namespace graphite2;
+
+namespace
+{
+enum compression
+{
+    NONE,
+    LZ4
+};
+
+}
+
+Face::Face(const void* appFaceHandle/*non-NULL*/, const gr_face_ops & ops)
+: m_appFaceHandle(appFaceHandle),
+  m_pFileFace(NULL),
+  m_pGlyphFaceCache(NULL),
+  m_cmap(NULL),
+  m_pNames(NULL),
+  m_logger(NULL),
+  m_error(0), m_errcntxt(0),
+  m_silfs(NULL),
+  m_numSilf(0),
+  m_ascent(0),
+  m_descent(0)
+{
+    memset(&m_ops, 0, sizeof m_ops);
+    memcpy(&m_ops, &ops, min(sizeof m_ops, ops.size));
+}
+
+
+Face::~Face()
+{
+    setLogger(0);
+    delete m_pGlyphFaceCache;
+    delete m_cmap;
+    delete[] m_silfs;
+#ifndef GRAPHITE2_NFILEFACE
+    delete m_pFileFace;
+#endif
+    delete m_pNames;
+}
+
+float Face::default_glyph_advance(const void* font_ptr, gr_uint16 glyphid)
+{
+    const Font & font = *reinterpret_cast<const Font *>(font_ptr);
+
+    return font.face().glyphs().glyph(glyphid)->theAdvance().x * font.scale();
+}
+
+bool Face::readGlyphs(uint32 faceOptions)
+{
+    Error e;
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::category _glyph_cat(tele.glyph);
+#endif
+    error_context(EC_READGLYPHS);
+    m_pGlyphFaceCache = new GlyphCache(*this, faceOptions);
+
+    if (e.test(!m_pGlyphFaceCache, E_OUTOFMEM)
+        || e.test(m_pGlyphFaceCache->numGlyphs() == 0, E_NOGLYPHS)
+        || e.test(m_pGlyphFaceCache->unitsPerEm() == 0, E_BADUPEM))
+    {
+        return error(e);
+    }
+
+    if (faceOptions & gr_face_cacheCmap)
+        m_cmap = new CachedCmap(*this);
+    else
+        m_cmap = new DirectCmap(*this);
+    if (e.test(!m_cmap, E_OUTOFMEM) || e.test(!*m_cmap, E_BADCMAP))
+        return error(e);
+
+    if (faceOptions & gr_face_preloadGlyphs)
+        nameTable();        // preload the name table along with the glyphs.
+
+    return true;
+}
+
+bool Face::readGraphite(const Table & silf)
+{
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::category _silf_cat(tele.silf);
+#endif
+    Error e;
+    error_context(EC_READSILF);
+    const byte * p = silf;
+    if (e.test(!p, E_NOSILF) || e.test(silf.size() < 20, E_BADSIZE)) return error(e);
+
+    const uint32 version = be::read<uint32>(p);
+    if (e.test(version < 0x00020000, E_TOOOLD)) return error(e);
+    if (version >= 0x00030000)
+        be::skip<uint32>(p);        // compilerVersion
+    m_numSilf = be::read<uint16>(p);
+
+    be::skip<uint16>(p);            // reserved
+
+    bool havePasses = false;
+    m_silfs = new Silf[m_numSilf];
+    if (e.test(!m_silfs, E_OUTOFMEM)) return error(e);
+    for (int i = 0; i < m_numSilf; i++)
+    {
+        error_context(EC_ASILF + (i << 8));
+        const uint32 offset = be::read<uint32>(p),
+                     next   = i == m_numSilf - 1 ? uint32(silf.size()) : be::peek<uint32>(p);
+        if (e.test(next > silf.size() || offset >= next, E_BADSIZE))
+            return error(e);
+
+        if (!m_silfs[i].readGraphite(silf + offset, next - offset, *this, version))
+            return false;
+
+        if (m_silfs[i].numPasses())
+            havePasses = true;
+    }
+
+    return havePasses;
+}
+
+bool Face::readFeatures()
+{
+    return m_Sill.readFace(*this);
+}
+
+bool Face::runGraphite(Segment *seg, const Silf *aSilf) const
+{
+#if !defined GRAPHITE2_NTRACING
+    json * dbgout = logger();
+    if (dbgout)
+    {
+        *dbgout << json::object
+                    << "id"         << objectid(seg)
+                    << "passes"     << json::array;
+    }
+#endif
+
+//    if ((seg->dir() & 1) != aSilf->dir())
+//        seg->reverseSlots();
+    if ((seg->dir() & 3) == 3 && aSilf->bidiPass() == 0xFF)
+        seg->doMirror(aSilf->aMirror());
+    bool res = aSilf->runGraphite(seg, 0, aSilf->positionPass(), true);
+    if (res)
+    {
+        seg->associateChars(0, seg->charInfoCount());
+        if (aSilf->flags() & 0x20)
+            res &= seg->initCollisions();
+        if (res)
+            res &= aSilf->runGraphite(seg, aSilf->positionPass(), aSilf->numPasses(), false);
+    }
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+{
+        seg->positionSlots(0, 0, 0, seg->currdir());
+        *dbgout             << json::item
+                            << json::close // Close up the passes array
+                << "outputdir" << (seg->currdir() ? "rtl" : "ltr")
+                << "output" << json::array;
+        for(Slot * s = seg->first(); s; s = s->next())
+            *dbgout     << dslot(seg, s);
+        *dbgout         << json::close
+                << "advance" << seg->advance()
+                << "chars"   << json::array;
+        for(size_t i = 0, n = seg->charInfoCount(); i != n; ++i)
+            *dbgout     << json::flat << *seg->charinfo(int(i));
+        *dbgout         << json::close  // Close up the chars array
+                    << json::close;     // Close up the segment object
+    }
+#endif
+
+    return res;
+}
+
+void Face::setLogger(FILE * log_file GR_MAYBE_UNUSED)
+{
+#if !defined GRAPHITE2_NTRACING
+    delete m_logger;
+    m_logger = log_file ? new json(log_file) : 0;
+#endif
+}
+
+const Silf *Face::chooseSilf(uint32 script) const
+{
+    if (m_numSilf == 0)
+        return NULL;
+    else if (m_numSilf == 1 || script == 0)
+        return m_silfs;
+    else // do more work here
+        return m_silfs;
+}
+
+uint16 Face::findPseudo(uint32 uid) const
+{
+    return (m_numSilf) ? m_silfs[0].findPseudo(uid) : 0;
+}
+
+int32 Face::getGlyphMetric(uint16 gid, uint8 metric) const
+{
+    switch (metrics(metric))
+    {
+        case kgmetAscent : return m_ascent;
+        case kgmetDescent : return m_descent;
+        default:
+            if (gid >= glyphs().numGlyphs()) return 0;
+            return glyphs().glyph(gid)->getMetric(metric);
+    }
+}
+
+void Face::takeFileFace(FileFace* pFileFace GR_MAYBE_UNUSED/*takes ownership*/)
+{
+#ifndef GRAPHITE2_NFILEFACE
+    if (m_pFileFace==pFileFace)
+      return;
+
+    delete m_pFileFace;
+    m_pFileFace = pFileFace;
+#endif
+}
+
+NameTable * Face::nameTable() const
+{
+    if (m_pNames) return m_pNames;
+    const Table name(*this, Tag::name);
+    if (name)
+        m_pNames = new NameTable(name, name.size());
+    return m_pNames;
+}
+
+uint16 Face::languageForLocale(const char * locale) const
+{
+    nameTable();
+    if (m_pNames)
+        return m_pNames->getLanguageId(locale);
+    return 0;
+}
+
+
+
+Face::Table::Table(const Face & face, const Tag n, uint32 version) throw()
+: _f(&face), _sz(0), _compressed(false)
+{
+    _p = static_cast<const byte *>((*_f->m_ops.get_table)(_f->m_appFaceHandle, n, &_sz));
+
+    if (!TtfUtil::CheckTable(n, _p, _sz))
+    {
+        release();     // Make sure we release the table buffer even if the table failed its checks
+        return;
+    }
+
+    if (be::peek<uint32>(_p) >= version)
+        decompress();
+}
+
+void Face::Table::release()
+{
+    if (_compressed)
+        free(const_cast<byte *>(_p));
+    else if (_p && _f->m_ops.release_table)
+        (*_f->m_ops.release_table)(_f->m_appFaceHandle, _p);
+    _p = 0; _sz = 0;
+}
+
+Face::Table & Face::Table::operator = (const Table && rhs) throw()
+{
+    if (this == &rhs)   return *this;
+    release();
+    new (this) Table(std::move(rhs));
+    return *this;
+}
+
+Error Face::Table::decompress()
+{
+    Error e;
+    if (e.test(_sz < 5 * sizeof(uint32), E_BADSIZE))
+        return e;
+    byte * uncompressed_table = 0;
+    size_t uncompressed_size = 0;
+
+    const byte * p = _p;
+    const uint32 version = be::read<uint32>(p);    // Table version number.
+
+    // The scheme is in the top 5 bits of the 1st uint32.
+    const uint32 hdr = be::read<uint32>(p);
+    switch(compression(hdr >> 27))
+    {
+    case NONE: return e;
+
+    case LZ4:
+    {
+        uncompressed_size  = hdr & 0x07ffffff;
+        uncompressed_table = gralloc<byte>(uncompressed_size);
+        if (!e.test(!uncompressed_table || uncompressed_size < 4, E_OUTOFMEM))
+        {
+            memset(uncompressed_table, 0, 4);   // make sure version number is initialised
+            // coverity[forward_null : FALSE] - uncompressed_table has been checked so can't be null
+            // coverity[checked_return : FALSE] - we test e later
+            e.test(lz4::decompress(p, _sz - 2*sizeof(uint32), uncompressed_table, uncompressed_size) != signed(uncompressed_size), E_SHRINKERFAILED);
+        }
+        break;
+    }
+
+    default:
+        e.error(E_BADSCHEME);
+    };
+
+    // Check the uncompressed version number against the original.
+    if (!e)
+        // coverity[forward_null : FALSE] - uncompressed_table has already been tested so can't be null
+        // coverity[checked_return : FALSE] - we test e later
+        e.test(be::peek<uint32>(uncompressed_table) != version, E_SHRINKERFAILED);
+
+    // Tell the provider to release the compressed form since were replacing
+    //   it anyway.
+    release();
+
+    if (e)
+    {
+        free(uncompressed_table);
+        uncompressed_table = 0;
+        uncompressed_size  = 0;
+    }
+
+    _p = uncompressed_table;
+    _sz = uncompressed_size;
+    _compressed = true;
+
+    return e;
+}
diff --git a/thirdparty/graphite/src/FeatureMap.cpp b/thirdparty/graphite/src/FeatureMap.cpp
new file mode 100644
index 0000000000..014a88fd08
--- /dev/null
+++ b/thirdparty/graphite/src/FeatureMap.cpp
@@ -0,0 +1,293 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cstring>
+
+#include "inc/Main.h"
+#include "inc/bits.h"
+#include "inc/Endian.h"
+#include "inc/FeatureMap.h"
+#include "inc/FeatureVal.h"
+#include "graphite2/Font.h"
+#include "inc/TtfUtil.h"
+#include <cstdlib>
+#include "inc/Face.h"
+
+
+using namespace graphite2;
+
+namespace
+{
+    static int cmpNameAndFeatures(const void *ap, const void *bp)
+    {
+        const NameAndFeatureRef & a = *static_cast<const NameAndFeatureRef *>(ap),
+                                & b = *static_cast<const NameAndFeatureRef *>(bp);
+        return (a < b ? -1 : (b < a ? 1 : 0));
+    }
+
+    const size_t    FEAT_HEADER     = sizeof(uint32) + 2*sizeof(uint16) + sizeof(uint32),
+                    FEATURE_SIZE    = sizeof(uint32)
+                                    + 2*sizeof(uint16)
+                                    + sizeof(uint32)
+                                    + 2*sizeof(uint16),
+                    FEATURE_SETTING_SIZE = sizeof(int16) + sizeof(uint16);
+
+    uint16 readFeatureSettings(const byte * p, FeatureSetting * s, size_t num_settings)
+    {
+        uint16 max_val = 0;
+        for (FeatureSetting * const end = s + num_settings; s != end; ++s)
+        {
+            const int16 value = be::read<int16>(p);
+            ::new (s) FeatureSetting(value, be::read<uint16>(p));
+            if (uint16(value) > max_val)    max_val = value;
+        }
+
+        return max_val;
+    }
+}
+
+FeatureRef::FeatureRef(const Face & face,
+    unsigned short & bits_offset, uint32 max_val,
+    uint32 name, uint16 uiName, flags_t flags,
+    FeatureSetting *settings, uint16 num_set) throw()
+: m_face(&face),
+  m_nameValues(settings),
+  m_mask(mask_over_val(max_val)),
+  m_max(max_val),
+  m_id(name),
+  m_nameid(uiName),
+  m_numSet(num_set),
+  m_flags(flags)
+{
+    const uint8 need_bits = bit_set_count(m_mask);
+    m_index = (bits_offset + need_bits) / SIZEOF_CHUNK;
+    if (m_index > bits_offset / SIZEOF_CHUNK)
+        bits_offset = m_index*SIZEOF_CHUNK;
+    m_bits = bits_offset % SIZEOF_CHUNK;
+    bits_offset += need_bits;
+    m_mask <<= m_bits;
+}
+
+FeatureRef::~FeatureRef() throw()
+{
+    free(m_nameValues);
+}
+
+bool FeatureMap::readFeats(const Face & face)
+{
+    const Face::Table feat(face, TtfUtil::Tag::Feat);
+    const byte * p = feat;
+    if (!p) return true;
+    if (feat.size() < FEAT_HEADER) return false;
+
+    const byte *const feat_start = p,
+               *const feat_end = p + feat.size();
+
+    const uint32 version = be::read<uint32>(p);
+    m_numFeats = be::read<uint16>(p);
+    be::skip<uint16>(p);
+    be::skip<uint32>(p);
+
+    // Sanity checks
+    if (m_numFeats == 0)    return true;
+    if (version < 0x00010000 ||
+        p + m_numFeats*FEATURE_SIZE > feat_end)
+    {   //defensive
+        m_numFeats = 0;
+        return false;
+    }
+
+    m_feats = new FeatureRef [m_numFeats];
+    uint16 * const  defVals = gralloc<uint16>(m_numFeats);
+    if (!defVals || !m_feats) return false;
+    unsigned short bits = 0;     //to cause overflow on first Feature
+
+    for (int i = 0, ie = m_numFeats; i != ie; i++)
+    {
+        const uint32    label   = version < 0x00020000 ? be::read<uint16>(p) : be::read<uint32>(p);
+        const uint16    num_settings = be::read<uint16>(p);
+        if (version >= 0x00020000)
+            be::skip<uint16>(p);
+        const uint32    settings_offset = be::read<uint32>(p);
+        const uint16    flags  = be::read<uint16>(p),
+                        uiName = be::read<uint16>(p);
+
+        if (settings_offset > size_t(feat_end - feat_start)
+            || settings_offset + num_settings * FEATURE_SETTING_SIZE > size_t(feat_end - feat_start))
+        {
+            free(defVals);
+            return false;
+        }
+
+        FeatureSetting *uiSet;
+        uint32 maxVal;
+        if (num_settings != 0)
+        {
+            uiSet = gralloc<FeatureSetting>(num_settings);
+            if (!uiSet)
+            {
+                free(defVals);
+                return false;
+            }
+            maxVal = readFeatureSettings(feat_start + settings_offset, uiSet, num_settings);
+            defVals[i] = uiSet[0].value();
+        }
+        else
+        {
+            uiSet = 0;
+            maxVal = 0xffffffff;
+            defVals[i] = 0;
+        }
+
+        ::new (m_feats + i) FeatureRef (face, bits, maxVal,
+                                       label, uiName, 
+                                       FeatureRef::flags_t(flags),
+                                       uiSet, num_settings);
+    }
+    new (&m_defaultFeatures) Features(bits/(sizeof(uint32)*8) + 1, *this);
+    m_pNamedFeats = new NameAndFeatureRef[m_numFeats];
+    if (!m_pNamedFeats)
+    {
+        free(defVals);
+        return false;
+    }
+    for (int i = 0; i < m_numFeats; ++i)
+    {
+        m_feats[i].applyValToFeature(defVals[i], m_defaultFeatures);
+        m_pNamedFeats[i] = m_feats[i];
+    }
+
+    free(defVals);
+
+    qsort(m_pNamedFeats, m_numFeats, sizeof(NameAndFeatureRef), &cmpNameAndFeatures);
+
+    return true;
+}
+
+bool SillMap::readFace(const Face & face)
+{
+    if (!m_FeatureMap.readFeats(face)) return false;
+    if (!readSill(face)) return false;
+    return true;
+}
+
+
+bool SillMap::readSill(const Face & face)
+{
+    const Face::Table sill(face, TtfUtil::Tag::Sill);
+    const byte *p = sill;
+
+    if (!p) return true;
+    if (sill.size() < 12) return false;
+    if (be::read<uint32>(p) != 0x00010000UL) return false;
+    m_numLanguages = be::read<uint16>(p);
+    m_langFeats = new LangFeaturePair[m_numLanguages];
+    if (!m_langFeats || !m_FeatureMap.m_numFeats) { m_numLanguages = 0; return true; }        //defensive
+
+    p += 6;     // skip the fast search
+    if (sill.size() < m_numLanguages * 8U + 12) return false;
+
+    for (int i = 0; i < m_numLanguages; i++)
+    {
+        uint32 langid = be::read<uint32>(p);
+        uint16 numSettings = be::read<uint16>(p);
+        uint16 offset = be::read<uint16>(p);
+        if (offset + 8U * numSettings > sill.size() && numSettings > 0) return false;
+        Features* feats = new Features(m_FeatureMap.m_defaultFeatures);
+        if (!feats) return false;
+        const byte *pLSet = sill + offset;
+
+        // Apply langauge specific settings
+        for (int j = 0; j < numSettings; j++)
+        {
+            uint32 name = be::read<uint32>(pLSet);
+            uint16 val = be::read<uint16>(pLSet);
+            pLSet += 2;
+            const FeatureRef* pRef = m_FeatureMap.findFeatureRef(name);
+            if (pRef)   pRef->applyValToFeature(val, *feats);
+        }
+        // Add the language id feature which is always feature id 1
+        const FeatureRef* pRef = m_FeatureMap.findFeatureRef(1);
+        if (pRef)   pRef->applyValToFeature(langid, *feats);
+
+        m_langFeats[i].m_lang = langid;
+        m_langFeats[i].m_pFeatures = feats;
+    }
+    return true;
+}
+
+
+Features* SillMap::cloneFeatures(uint32 langname/*0 means default*/) const
+{
+    if (langname)
+    {
+        // the number of languages in a font is usually small e.g. 8 in Doulos
+        // so this loop is not very expensive
+        for (uint16 i = 0; i < m_numLanguages; i++)
+        {
+            if (m_langFeats[i].m_lang == langname)
+                return new Features(*m_langFeats[i].m_pFeatures);
+        }
+    }
+    return new Features (m_FeatureMap.m_defaultFeatures);
+}
+
+
+
+const FeatureRef *FeatureMap::findFeatureRef(uint32 name) const
+{
+    NameAndFeatureRef *it;
+
+    for (it = m_pNamedFeats; it < m_pNamedFeats + m_numFeats; ++it)
+        if (it->m_name == name)
+            return it->m_pFRef;
+    return NULL;
+}
+
+bool FeatureRef::applyValToFeature(uint32 val, Features & pDest) const
+{
+    if (val>maxVal() || !m_face)
+      return false;
+    if (pDest.m_pMap==NULL)
+      pDest.m_pMap = &m_face->theSill().theFeatureMap();
+    else
+      if (pDest.m_pMap!=&m_face->theSill().theFeatureMap())
+        return false;       //incompatible
+    if (m_index >= pDest.size())
+        pDest.resize(m_index+1);
+    pDest[m_index] &= ~m_mask;
+    pDest[m_index] |= (uint32(val) << m_bits);
+    return true;
+}
+
+uint32 FeatureRef::getFeatureVal(const Features& feats) const
+{
+  if (m_index < feats.size() && m_face
+      && &m_face->theSill().theFeatureMap()==feats.m_pMap)
+    return (feats[m_index] & m_mask) >> m_bits;
+  else
+    return 0;
+}
diff --git a/thirdparty/graphite/src/FileFace.cpp b/thirdparty/graphite/src/FileFace.cpp
new file mode 100644
index 0000000000..7e663876a7
--- /dev/null
+++ b/thirdparty/graphite/src/FileFace.cpp
@@ -0,0 +1,115 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cstring>
+#include "inc/FileFace.h"
+
+
+#ifndef GRAPHITE2_NFILEFACE
+
+using namespace graphite2;
+
+FileFace::FileFace(const char *filename)
+: _file(fopen(filename, "rb")),
+  _file_len(0),
+  _header_tbl(NULL),
+  _table_dir(NULL)
+{
+    if (!_file) return;
+
+    if (fseek(_file, 0, SEEK_END)) return;
+    _file_len = ftell(_file);
+    if (fseek(_file, 0, SEEK_SET)) return;
+
+    size_t tbl_offset, tbl_len;
+
+    // Get the header.
+    if (!TtfUtil::GetHeaderInfo(tbl_offset, tbl_len)) return;
+    if (fseek(_file, long(tbl_offset), SEEK_SET)) return;
+    _header_tbl = (TtfUtil::Sfnt::OffsetSubTable*)gralloc<char>(tbl_len);
+    if (_header_tbl)
+    {
+        if (fread(_header_tbl, 1, tbl_len, _file) != tbl_len) return;
+        if (!TtfUtil::CheckHeader(_header_tbl)) return;
+    }
+
+    // Get the table directory
+    if (!TtfUtil::GetTableDirInfo(_header_tbl, tbl_offset, tbl_len)) return;
+    _table_dir = (TtfUtil::Sfnt::OffsetSubTable::Entry*)gralloc<char>(tbl_len);
+    if (fseek(_file, long(tbl_offset), SEEK_SET)) return;
+    if (_table_dir && fread(_table_dir, 1, tbl_len, _file) != tbl_len)
+    {
+        free(_table_dir);
+        _table_dir = NULL;
+    }
+    return;
+}
+
+FileFace::~FileFace()
+{
+    free(_table_dir);
+    free(_header_tbl);
+    if (_file)
+        fclose(_file);
+}
+
+
+const void *FileFace::get_table_fn(const void* appFaceHandle, unsigned int name, size_t *len)
+{
+    if (appFaceHandle == 0)     return 0;
+    const FileFace & file_face = *static_cast<const FileFace *>(appFaceHandle);
+
+    void *tbl;
+    size_t tbl_offset, tbl_len;
+    if (!TtfUtil::GetTableInfo(name, file_face._header_tbl, file_face._table_dir, tbl_offset, tbl_len))
+        return 0;
+
+    if (tbl_offset > file_face._file_len || tbl_len > file_face._file_len - tbl_offset
+            || fseek(file_face._file, long(tbl_offset), SEEK_SET) != 0)
+        return 0;
+
+    tbl = malloc(tbl_len);
+    if (!tbl || fread(tbl, 1, tbl_len, file_face._file) != tbl_len)
+    {
+        free(tbl);
+        return 0;
+    }
+
+    if (len) *len = tbl_len;
+    return tbl;
+}
+
+void FileFace::rel_table_fn(const void* appFaceHandle, const void *table_buffer)
+{
+    if (appFaceHandle == 0)     return;
+
+    free(const_cast<void *>(table_buffer));
+}
+
+const gr_face_ops FileFace::ops = { sizeof FileFace::ops, &FileFace::get_table_fn, &FileFace::rel_table_fn };
+
+
+#endif                  //!GRAPHITE2_NFILEFACE
diff --git a/thirdparty/graphite/src/Font.cpp b/thirdparty/graphite/src/Font.cpp
new file mode 100644
index 0000000000..faf3715f9d
--- /dev/null
+++ b/thirdparty/graphite/src/Font.cpp
@@ -0,0 +1,58 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/Face.h"
+#include "inc/Font.h"
+#include "inc/GlyphCache.h"
+
+using namespace graphite2;
+
+Font::Font(float ppm, const Face & f, const void * appFontHandle, const gr_font_ops * ops)
+: m_appFontHandle(appFontHandle ? appFontHandle : this),
+  m_face(f),
+  m_scale(ppm / f.glyphs().unitsPerEm()),
+  m_hinted(appFontHandle && ops && (ops->glyph_advance_x || ops->glyph_advance_y))
+{
+    memset(&m_ops, 0, sizeof m_ops);
+    if (m_hinted && ops)
+        memcpy(&m_ops, ops, min(sizeof m_ops, ops->size));
+    else
+        m_ops.glyph_advance_x = &Face::default_glyph_advance;
+
+    size_t nGlyphs = f.glyphs().numGlyphs();
+    m_advances = gralloc<float>(nGlyphs);
+    if (m_advances)
+    {
+        for (float *advp = m_advances; nGlyphs; --nGlyphs, ++advp)
+            *advp = INVALID_ADVANCE;
+    }
+}
+
+
+/*virtual*/ Font::~Font()
+{
+    free(m_advances);
+}
diff --git a/thirdparty/graphite/src/GlyphCache.cpp b/thirdparty/graphite/src/GlyphCache.cpp
new file mode 100644
index 0000000000..282bdc18fd
--- /dev/null
+++ b/thirdparty/graphite/src/GlyphCache.cpp
@@ -0,0 +1,492 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Font.h"
+
+#include "inc/Main.h"
+#include "inc/Face.h"     //for the tags
+#include "inc/GlyphCache.h"
+#include "inc/GlyphFace.h"
+#include "inc/Endian.h"
+#include "inc/bits.h"
+
+using namespace graphite2;
+
+namespace
+{
+    // Iterator over version 1 or 2 glat entries which consist of a series of
+    //    +-+-+-+-+-+-+-+-+-+-+                +-+-+-+-+-+-+-+-+-+-+-+-+
+    // v1 |k|n|v1 |v2 |...|vN |     or    v2   | k | n |v1 |v2 |...|vN |
+    //    +-+-+-+-+-+-+-+-+-+-+                +-+-+-+-+-+-+-+-+-+-+-+-+
+    // variable length structures.
+
+    template<typename W>
+    class _glat_iterator : public std::iterator<std::input_iterator_tag, std::pair<sparse::key_type, sparse::mapped_type> >
+    {
+        unsigned short  key() const             { return uint16(be::peek<W>(_e) + _n); }
+        unsigned int    run() const             { return be::peek<W>(_e+sizeof(W)); }
+        void            advance_entry()         { _n = 0; _e = _v; be::skip<W>(_v,2); }
+    public:
+        _glat_iterator(const void * glat=0) : _e(reinterpret_cast<const byte *>(glat)), _v(_e+2*sizeof(W)), _n(0) {}
+
+        _glat_iterator<W> & operator ++ () {
+            ++_n; be::skip<uint16>(_v);
+            if (_n == run()) advance_entry();
+            return *this;
+        }
+        _glat_iterator<W>   operator ++ (int)   { _glat_iterator<W> tmp(*this); operator++(); return tmp; }
+
+        // This is strictly a >= operator. A true == operator could be
+        // implemented that test for overlap but it would be more expensive a
+        // test.
+        bool operator == (const _glat_iterator<W> & rhs) { return _v >= rhs._e - 1; }
+        bool operator != (const _glat_iterator<W> & rhs) { return !operator==(rhs); }
+
+        value_type          operator * () const {
+            return value_type(key(), be::peek<uint16>(_v));
+        }
+
+    protected:
+        const byte     * _e, * _v;
+        size_t        _n;
+    };
+
+    typedef _glat_iterator<uint8>   glat_iterator;
+    typedef _glat_iterator<uint16>  glat2_iterator;
+}
+
+const SlantBox SlantBox::empty = {0,0,0,0};
+
+
+class GlyphCache::Loader
+{
+public:
+    Loader(const Face & face);    //return result indicates success. Do not use if failed.
+
+    operator bool () const throw();
+    unsigned short int units_per_em() const throw();
+    unsigned short int num_glyphs() const throw();
+    unsigned short int num_attrs() const throw();
+    bool has_boxes() const throw();
+
+    const GlyphFace * read_glyph(unsigned short gid, GlyphFace &, int *numsubs) const throw();
+    GlyphBox * read_box(uint16 gid, GlyphBox *curr, const GlyphFace & face) const throw();
+
+    CLASS_NEW_DELETE;
+private:
+    Face::Table _head,
+                _hhea,
+                _hmtx,
+                _glyf,
+                _loca,
+                m_pGlat,
+                m_pGloc;
+
+    bool            _long_fmt;
+    bool            _has_boxes;
+    unsigned short  _num_glyphs_graphics,        //i.e. boundary box and advance
+                    _num_glyphs_attributes,
+                    _num_attrs;                    // number of glyph attributes per glyph
+};
+
+
+
+GlyphCache::GlyphCache(const Face & face, const uint32 face_options)
+: _glyph_loader(new Loader(face)),
+  _glyphs(_glyph_loader && *_glyph_loader && _glyph_loader->num_glyphs()
+        ? grzeroalloc<const GlyphFace *>(_glyph_loader->num_glyphs()) : 0),
+  _boxes(_glyph_loader && _glyph_loader->has_boxes() && _glyph_loader->num_glyphs()
+        ? grzeroalloc<GlyphBox *>(_glyph_loader->num_glyphs()) : 0),
+  _num_glyphs(_glyphs ? _glyph_loader->num_glyphs() : 0),
+  _num_attrs(_glyphs ? _glyph_loader->num_attrs() : 0),
+  _upem(_glyphs ? _glyph_loader->units_per_em() : 0)
+{
+    if ((face_options & gr_face_preloadGlyphs) && _glyph_loader && _glyphs)
+    {
+        int numsubs = 0;
+        GlyphFace * const glyphs = new GlyphFace [_num_glyphs];
+        if (!glyphs)
+            return;
+
+        // The 0 glyph is definately required.
+        _glyphs[0] = _glyph_loader->read_glyph(0, glyphs[0], &numsubs);
+
+        // glyphs[0] has the same address as the glyphs array just allocated,
+        //  thus assigning the &glyphs[0] to _glyphs[0] means _glyphs[0] points
+        //  to the entire array.
+        const GlyphFace * loaded = _glyphs[0];
+        for (uint16 gid = 1; loaded && gid != _num_glyphs; ++gid)
+            _glyphs[gid] = loaded = _glyph_loader->read_glyph(gid, glyphs[gid], &numsubs);
+
+        if (!loaded)
+        {
+            _glyphs[0] = 0;
+            delete [] glyphs;
+        }
+        else if (numsubs > 0 && _boxes)
+        {
+            GlyphBox * boxes = (GlyphBox *)gralloc<char>(_num_glyphs * sizeof(GlyphBox) + numsubs * 8 * sizeof(float));
+            GlyphBox * currbox = boxes;
+
+            for (uint16 gid = 0; currbox && gid != _num_glyphs; ++gid)
+            {
+                _boxes[gid] = currbox;
+                currbox = _glyph_loader->read_box(gid, currbox, *_glyphs[gid]);
+            }
+            if (!currbox)
+            {
+                free(boxes);
+                _boxes[0] = 0;
+            }
+        }
+        delete _glyph_loader;
+        _glyph_loader = 0;
+	// coverity[leaked_storage : FALSE] - calling read_glyph on index 0 saved
+	// glyphs as _glyphs[0]. Setting _glyph_loader to nullptr here flags that
+	// the dtor needs to call delete[] on _glyphs[0] to release what was allocated
+	// as glyphs
+    }
+
+    if (_glyphs && glyph(0) == 0)
+    {
+        free(_glyphs);
+        _glyphs = 0;
+        if (_boxes)
+        {
+            free(_boxes);
+            _boxes = 0;
+        }
+        _num_glyphs = _num_attrs = _upem = 0;
+    }
+}
+
+
+GlyphCache::~GlyphCache()
+{
+    if (_glyphs)
+    {
+        if (_glyph_loader)
+        {
+            const GlyphFace *  * g = _glyphs;
+            for(unsigned short n = _num_glyphs; n; --n, ++g)
+                delete *g;
+        }
+        else
+            delete [] _glyphs[0];
+        free(_glyphs);
+    }
+    if (_boxes)
+    {
+        if (_glyph_loader)
+        {
+            GlyphBox *  * g = _boxes;
+            for (uint16 n = _num_glyphs; n; --n, ++g)
+                free(*g);
+        }
+        else
+            free(_boxes[0]);
+        free(_boxes);
+    }
+    delete _glyph_loader;
+}
+
+const GlyphFace *GlyphCache::glyph(unsigned short glyphid) const      //result may be changed by subsequent call with a different glyphid
+{
+    if (glyphid >= numGlyphs())
+        return _glyphs[0];
+    const GlyphFace * & p = _glyphs[glyphid];
+    if (p == 0 && _glyph_loader)
+    {
+        int numsubs = 0;
+        GlyphFace * g = new GlyphFace();
+        if (g)  p = _glyph_loader->read_glyph(glyphid, *g, &numsubs);
+        if (!p)
+        {
+            delete g;
+            return *_glyphs;
+        }
+        if (_boxes)
+        {
+            _boxes[glyphid] = (GlyphBox *)gralloc<char>(sizeof(GlyphBox) + 8 * numsubs * sizeof(float));
+            if (!_glyph_loader->read_box(glyphid, _boxes[glyphid], *_glyphs[glyphid]))
+            {
+                free(_boxes[glyphid]);
+                _boxes[glyphid] = 0;
+            }
+        }
+    }
+    return p;
+}
+
+
+
+GlyphCache::Loader::Loader(const Face & face)
+: _head(face, Tag::head),
+  _hhea(face, Tag::hhea),
+  _hmtx(face, Tag::hmtx),
+  _glyf(face, Tag::glyf),
+  _loca(face, Tag::loca),
+  _long_fmt(false),
+  _has_boxes(false),
+  _num_glyphs_graphics(0),
+  _num_glyphs_attributes(0),
+  _num_attrs(0)
+{
+    if (!operator bool())
+        return;
+
+    const Face::Table maxp = Face::Table(face, Tag::maxp);
+    if (!maxp) { _head = Face::Table(); return; }
+
+    _num_glyphs_graphics = static_cast<unsigned short>(TtfUtil::GlyphCount(maxp));
+    // This will fail if the number of glyphs is wildly out of range.
+    if (_glyf && TtfUtil::LocaLookup(_num_glyphs_graphics-1, _loca, _loca.size(), _head) == size_t(-2))
+    {
+        _head = Face::Table();
+        return;
+    }
+
+    if ((m_pGlat = Face::Table(face, Tag::Glat, 0x00030000)) == NULL
+        || (m_pGloc = Face::Table(face, Tag::Gloc)) == NULL
+        || m_pGloc.size() < 8)
+    {
+        _head = Face::Table();
+        return;
+    }
+    const byte    * p = m_pGloc;
+    int       version = be::read<uint32>(p);
+    const uint16    flags = be::read<uint16>(p);
+    _num_attrs = be::read<uint16>(p);
+    // We can accurately calculate the number of attributed glyphs by
+    //  subtracting the length of the attribids array (numAttribs long if present)
+    //  and dividing by either 2 or 4 depending on shor or lonf format
+    _long_fmt              = flags & 1;
+    ptrdiff_t tmpnumgattrs       = (m_pGloc.size()
+                               - (p - m_pGloc)
+                               - sizeof(uint16)*(flags & 0x2 ? _num_attrs : 0))
+                                   / (_long_fmt ? sizeof(uint32) : sizeof(uint16)) - 1;
+
+    if (version >= 0x00020000 || tmpnumgattrs < 0 || tmpnumgattrs > 65535
+        || _num_attrs == 0 || _num_attrs > 0x3000  // is this hard limit appropriate?
+        || _num_glyphs_graphics > tmpnumgattrs
+        || m_pGlat.size() < 4)
+    {
+        _head = Face::Table();
+        return;
+    }
+
+    _num_glyphs_attributes = static_cast<unsigned short>(tmpnumgattrs);
+    p = m_pGlat;
+    version = be::read<uint32>(p);
+    if (version >= 0x00040000 || (version >= 0x00030000 && m_pGlat.size() < 8))       // reject Glat tables that are too new
+    {
+        _head = Face::Table();
+        return;
+    }
+    else if (version >= 0x00030000)
+    {
+        unsigned int glatflags = be::read<uint32>(p);
+        _has_boxes = glatflags & 1;
+        // delete this once the compiler is fixed
+        _has_boxes = true;
+    }
+}
+
+inline
+GlyphCache::Loader::operator bool () const throw()
+{
+    return _head && _hhea && _hmtx && !(bool(_glyf) != bool(_loca));
+}
+
+inline
+unsigned short int GlyphCache::Loader::units_per_em() const throw()
+{
+    return _head ? TtfUtil::DesignUnits(_head) : 0;
+}
+
+inline
+unsigned short int GlyphCache::Loader::num_glyphs() const throw()
+{
+    return max(_num_glyphs_graphics, _num_glyphs_attributes);
+}
+
+inline
+unsigned short int GlyphCache::Loader::num_attrs() const throw()
+{
+    return _num_attrs;
+}
+
+inline
+bool GlyphCache::Loader::has_boxes () const throw()
+{
+    return _has_boxes;
+}
+
+const GlyphFace * GlyphCache::Loader::read_glyph(unsigned short glyphid, GlyphFace & glyph, int *numsubs) const throw()
+{
+    Rect        bbox;
+    Position    advance;
+
+    if (glyphid < _num_glyphs_graphics)
+    {
+        int nLsb;
+        unsigned int nAdvWid;
+        if (_glyf)
+        {
+            int xMin, yMin, xMax, yMax;
+            size_t locidx = TtfUtil::LocaLookup(glyphid, _loca, _loca.size(), _head);
+            void *pGlyph = TtfUtil::GlyfLookup(_glyf, locidx, _glyf.size());
+
+            if (pGlyph && TtfUtil::GlyfBox(pGlyph, xMin, yMin, xMax, yMax))
+            {
+                if ((xMin > xMax) || (yMin > yMax))
+                    return 0;
+                bbox = Rect(Position(static_cast<float>(xMin), static_cast<float>(yMin)),
+                    Position(static_cast<float>(xMax), static_cast<float>(yMax)));
+            }
+        }
+        if (TtfUtil::HorMetrics(glyphid, _hmtx, _hmtx.size(), _hhea, nLsb, nAdvWid))
+            advance = Position(static_cast<float>(nAdvWid), 0);
+    }
+
+    if (glyphid < _num_glyphs_attributes)
+    {
+        const byte * gloc = m_pGloc;
+        size_t      glocs = 0, gloce = 0;
+
+        be::skip<uint32>(gloc);
+        be::skip<uint16>(gloc,2);
+        if (_long_fmt)
+        {
+            if (8 + glyphid * sizeof(uint32) > m_pGloc.size())
+                return 0;
+            be::skip<uint32>(gloc, glyphid);
+            glocs = be::read<uint32>(gloc);
+            gloce = be::peek<uint32>(gloc);
+        }
+        else
+        {
+            if (8 + glyphid * sizeof(uint16) > m_pGloc.size())
+                return 0;
+            be::skip<uint16>(gloc, glyphid);
+            glocs = be::read<uint16>(gloc);
+            gloce = be::peek<uint16>(gloc);
+        }
+
+        if (glocs >= m_pGlat.size() - 1 || gloce > m_pGlat.size())
+            return 0;
+
+        const uint32 glat_version = be::peek<uint32>(m_pGlat);
+        if (glat_version >= 0x00030000)
+        {
+            if (glocs >= gloce)
+                return 0;
+            const byte * p = m_pGlat + glocs;
+            uint16 bmap = be::read<uint16>(p);
+            int num = bit_set_count((uint32)bmap);
+            if (numsubs) *numsubs += num;
+            glocs += 6 + 8 * num;
+            if (glocs > gloce)
+                return 0;
+        }
+        if (glat_version < 0x00020000)
+        {
+            if (gloce - glocs < 2*sizeof(byte)+sizeof(uint16)
+                || gloce - glocs > _num_attrs*(2*sizeof(byte)+sizeof(uint16)))
+                    return 0;
+            new (&glyph) GlyphFace(bbox, advance, glat_iterator(m_pGlat + glocs), glat_iterator(m_pGlat + gloce));
+        }
+        else
+        {
+            if (gloce - glocs < 3*sizeof(uint16)        // can a glyph have no attributes? why not?
+                || gloce - glocs > _num_attrs*3*sizeof(uint16)
+                || glocs > m_pGlat.size() - 2*sizeof(uint16))
+                    return 0;
+            new (&glyph) GlyphFace(bbox, advance, glat2_iterator(m_pGlat + glocs), glat2_iterator(m_pGlat + gloce));
+        }
+        if (!glyph.attrs() || glyph.attrs().capacity() > _num_attrs)
+            return 0;
+    }
+    return &glyph;
+}
+
+inline float scale_to(uint8 t, float zmin, float zmax)
+{
+    return (zmin + t * (zmax - zmin) / 255);
+}
+
+Rect readbox(Rect &b, uint8 zxmin, uint8 zymin, uint8 zxmax, uint8 zymax)
+{
+    return Rect(Position(scale_to(zxmin, b.bl.x, b.tr.x), scale_to(zymin, b.bl.y, b.tr.y)),
+                Position(scale_to(zxmax, b.bl.x, b.tr.x), scale_to(zymax, b.bl.y, b.tr.y)));
+}
+
+GlyphBox * GlyphCache::Loader::read_box(uint16 gid, GlyphBox *curr, const GlyphFace & glyph) const throw()
+{
+    if (gid >= _num_glyphs_attributes) return 0;
+
+    const byte * gloc = m_pGloc;
+    size_t      glocs = 0, gloce = 0;
+
+    be::skip<uint32>(gloc);
+    be::skip<uint16>(gloc,2);
+    if (_long_fmt)
+    {
+        be::skip<uint32>(gloc, gid);
+        glocs = be::read<uint32>(gloc);
+        gloce = be::peek<uint32>(gloc);
+    }
+    else
+    {
+        be::skip<uint16>(gloc, gid);
+        glocs = be::read<uint16>(gloc);
+        gloce = be::peek<uint16>(gloc);
+    }
+
+    if (gloce > m_pGlat.size() || glocs + 6 >= gloce)
+        return 0;
+
+    const byte * p = m_pGlat + glocs;
+    uint16 bmap = be::read<uint16>(p);
+    int num = bit_set_count((uint32)bmap);
+
+    Rect bbox = glyph.theBBox();
+    Rect diamax(Position(bbox.bl.x + bbox.bl.y, bbox.bl.x - bbox.tr.y),
+                Position(bbox.tr.x + bbox.tr.y, bbox.tr.x - bbox.bl.y));
+    Rect diabound = readbox(diamax, p[0], p[2], p[1], p[3]);
+    ::new (curr) GlyphBox(num, bmap, &diabound);
+    be::skip<uint8>(p, 4);
+    if (glocs + 6 + num * 8 >= gloce)
+        return 0;
+
+    for (int i = 0; i < num * 2; ++i)
+    {
+        Rect box = readbox((i & 1) ? diamax : bbox, p[0], p[2], p[1], p[3]);
+        curr->addSubBox(i >> 1, i & 1, &box);
+        be::skip<uint8>(p, 4);
+    }
+    return (GlyphBox *)((char *)(curr) + sizeof(GlyphBox) + 2 * num * sizeof(Rect));
+}
diff --git a/thirdparty/graphite/src/GlyphFace.cpp b/thirdparty/graphite/src/GlyphFace.cpp
new file mode 100644
index 0000000000..bc5e63a9f0
--- /dev/null
+++ b/thirdparty/graphite/src/GlyphFace.cpp
@@ -0,0 +1,48 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/GlyphFace.h"
+
+
+using namespace graphite2;
+
+int32 GlyphFace::getMetric(uint8 metric) const
+{
+    switch (metrics(metric))
+    {
+        case kgmetLsb       : return int32(m_bbox.bl.x);
+        case kgmetRsb       : return int32(m_advance.x - m_bbox.tr.x);
+        case kgmetBbTop     : return int32(m_bbox.tr.y);
+        case kgmetBbBottom  : return int32(m_bbox.bl.y);
+        case kgmetBbLeft    : return int32(m_bbox.bl.x);
+        case kgmetBbRight   : return int32(m_bbox.tr.x);
+        case kgmetBbHeight  : return int32(m_bbox.tr.y - m_bbox.bl.y);
+        case kgmetBbWidth   : return int32(m_bbox.tr.x - m_bbox.bl.x);
+        case kgmetAdvWidth  : return int32(m_advance.x);
+        case kgmetAdvHeight : return int32(m_advance.y);
+        default : return 0;
+    }
+}
diff --git a/thirdparty/graphite/src/Intervals.cpp b/thirdparty/graphite/src/Intervals.cpp
new file mode 100644
index 0000000000..0fe99a127a
--- /dev/null
+++ b/thirdparty/graphite/src/Intervals.cpp
@@ -0,0 +1,298 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <algorithm>
+#include <cmath>
+#include <limits>
+
+#include "inc/Intervals.h"
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/debug.h"
+#include "inc/bits.h"
+
+using namespace graphite2;
+
+#include <cmath>
+
+inline
+Zones::Exclusion  Zones::Exclusion::split_at(float p) {
+    Exclusion r(*this);
+    r.xm = x = p;
+    return r;
+}
+
+inline
+void Zones::Exclusion::left_trim(float p) {
+    x = p;
+}
+
+inline
+Zones::Exclusion & Zones::Exclusion::operator += (Exclusion const & rhs) {
+    c += rhs.c; sm += rhs.sm; smx += rhs.smx; open = false;
+    return *this;
+}
+
+inline
+uint8 Zones::Exclusion::outcode(float val) const {
+    float p = val;
+    //float d = std::numeric_limits<float>::epsilon();
+    float d = 0.;
+    return ((p - xm >= d) << 1) | (x - p > d);
+}
+
+void Zones::exclude_with_margins(float xmin, float xmax, int axis) {
+    remove(xmin, xmax);
+    weightedAxis(axis, xmin-_margin_len, xmin, 0, 0, _margin_weight, xmin-_margin_len, 0, 0, false);
+    weightedAxis(axis, xmax, xmax+_margin_len, 0, 0, _margin_weight, xmax+_margin_len, 0, 0, false);
+}
+
+namespace
+{
+
+inline
+bool separated(float a, float b) {
+    return a != b;
+    //int exp;
+    //float res = frexpf(fabs(a - b), &exp);
+    //return (*(unsigned int *)(&res) > 4);
+    //return std::fabs(a-b) > std::numeric_limits<float>::epsilon(); // std::epsilon may not work. but 0.5 fails exising 64 bit tests
+    //return std::fabs(a-b) > 0.5f;
+}
+
+}
+
+void Zones::insert(Exclusion e)
+{
+#if !defined GRAPHITE2_NTRACING
+    addDebug(&e);
+#endif
+    e.x = max(e.x, _pos);
+    e.xm = min(e.xm, _posm);
+    if (e.x >= e.xm) return;
+
+    for (iterator i = _exclusions.begin(), ie = _exclusions.end(); i != ie && e.x < e.xm; ++i)
+    {
+        const uint8 oca = e.outcode(i->x),
+                    ocb = e.outcode(i->xm);
+        if ((oca & ocb) != 0) continue;
+
+        switch (oca ^ ocb)  // What kind of overlap?
+        {
+        case 0:     // e completely covers i
+            // split e at i.x into e1,e2
+            // split e2 at i.mx into e2,e3
+            // drop e1 ,i+e2, e=e3
+            *i += e;
+            e.left_trim(i->xm);
+            break;
+        case 1:     // e overlaps on the rhs of i
+            // split i at e->x into i1,i2
+            // split e at i.mx into e1,e2
+            // trim i1, insert i2+e1, e=e2
+            if (!separated(i->xm, e.x)) break;
+            if (separated(i->x,e.x))   { i = _exclusions.insert(i,i->split_at(e.x)); ++i; }
+            *i += e;
+            e.left_trim(i->xm);
+            break;
+        case 2:     // e overlaps on the lhs of i
+            // split e at i->x into e1,e2
+            // split i at e.mx into i1,i2
+            // drop e1, insert e2+i1, trim i2
+            if (!separated(e.xm, i->x)) return;
+            if (separated(e.xm, i->xm)) i = _exclusions.insert(i,i->split_at(e.xm));
+            *i += e;
+            return;
+        case 3:     // i completely covers e
+            // split i at e.x into i1,i2
+            // split i2 at e.mx into i2,i3
+            // insert i1, insert e+i2
+            if (separated(e.xm, i->xm)) i = _exclusions.insert(i,i->split_at(e.xm));
+            i = _exclusions.insert(i, i->split_at(e.x));
+            *++i += e;
+            return;
+        }
+
+        ie = _exclusions.end();
+    }
+}
+
+
+void Zones::remove(float x, float xm)
+{
+#if !defined GRAPHITE2_NTRACING
+    removeDebug(x, xm);
+#endif
+    x = max(x, _pos);
+    xm = min(xm, _posm);
+    if (x >= xm) return;
+
+    for (iterator i = _exclusions.begin(), ie = _exclusions.end(); i != ie; ++i)
+    {
+        const uint8 oca = i->outcode(x),
+                    ocb = i->outcode(xm);
+        if ((oca & ocb) != 0)   continue;
+
+        switch (oca ^ ocb)  // What kind of overlap?
+        {
+        case 0:     // i completely covers e
+            if (separated(i->x, x))  { i = _exclusions.insert(i,i->split_at(x)); ++i; }
+            GR_FALLTHROUGH;
+            // no break
+        case 1:     // i overlaps on the rhs of e
+            i->left_trim(xm);
+            return;
+        case 2:     // i overlaps on the lhs of e
+            i->xm = x;
+            if (separated(i->x, i->xm)) break;
+            GR_FALLTHROUGH;
+            // no break
+        case 3:     // e completely covers i
+            i = _exclusions.erase(i);
+            --i;
+            break;
+        }
+
+        ie = _exclusions.end();
+    }
+}
+
+
+Zones::const_iterator Zones::find_exclusion_under(float x) const
+{
+    size_t l = 0, h = _exclusions.size();
+
+    while (l < h)
+    {
+        size_t const p = (l+h) >> 1;
+        switch (_exclusions[p].outcode(x))
+        {
+        case 0 : return _exclusions.begin()+p;
+        case 1 : h = p; break;
+        case 2 :
+        case 3 : l = p+1; break;
+        }
+    }
+
+    return _exclusions.begin()+l;
+}
+
+
+float Zones::closest(float origin, float & cost) const
+{
+    float best_c = std::numeric_limits<float>::max(),
+          best_x = 0;
+
+    const const_iterator start = find_exclusion_under(origin);
+
+    // Forward scan looking for lowest cost
+    for (const_iterator i = start, ie = _exclusions.end(); i != ie; ++i)
+        if (i->track_cost(best_c, best_x, origin)) break;
+
+    // Backward scan looking for lowest cost
+    //  We start from the exclusion to the immediate left of start since we've
+    //  already tested start with the right most scan above.
+    for (const_iterator i = start-1, ie = _exclusions.begin()-1; i != ie; --i)
+        if (i->track_cost(best_c, best_x, origin)) break;
+
+    cost = (best_c == std::numeric_limits<float>::max() ? -1 : best_c);
+    return best_x;
+}
+
+
+// Cost and test position functions
+
+bool Zones::Exclusion::track_cost(float & best_cost, float & best_pos, float origin) const {
+    const float p = test_position(origin),
+                localc = cost(p - origin);
+    if (open && localc > best_cost) return true;
+
+    if (localc < best_cost)
+    {
+        best_cost = localc;
+        best_pos = p;
+    }
+    return false;
+}
+
+inline
+float Zones::Exclusion::cost(float p) const {
+    return (sm * p - 2 * smx) * p + c;
+}
+
+
+float Zones::Exclusion::test_position(float origin) const {
+    if (sm < 0)
+    {
+        // sigh, test both ends and perhaps the middle too!
+        float res = x;
+        float cl = cost(x);
+        if (x < origin && xm > origin)
+        {
+            float co = cost(origin);
+            if (co < cl)
+            {
+                cl = co;
+                res = origin;
+            }
+        }
+        float cr = cost(xm);
+        return cl > cr ? xm : res;
+    }
+    else
+    {
+        float zerox = smx / sm + origin;
+        if (zerox < x) return x;
+        else if (zerox > xm) return xm;
+        else return zerox;
+    }
+}
+
+
+#if !defined GRAPHITE2_NTRACING
+
+void Zones::jsonDbgOut(Segment *seg) const {
+
+    if (_dbg)
+    {
+        for (Zones::idebugs s = dbgs_begin(), e = dbgs_end(); s != e; ++s)
+        {
+            *_dbg << json::flat << json::array
+                << objectid(dslot(seg, (Slot *)(s->_env[0])))
+                << reinterpret_cast<ptrdiff_t>(s->_env[1]);
+            if (s->_isdel)
+                *_dbg << "remove" << Position(s->_excl.x, s->_excl.xm);
+            else
+                *_dbg << "exclude" << json::flat << json::array
+                    << s->_excl.x << s->_excl.xm
+                    << s->_excl.sm << s->_excl.smx << s->_excl.c
+                    << json::close;
+            *_dbg << json::close;
+        }
+    }
+}
+
+#endif
diff --git a/thirdparty/graphite/src/Justifier.cpp b/thirdparty/graphite/src/Justifier.cpp
new file mode 100644
index 0000000000..78c11e6a51
--- /dev/null
+++ b/thirdparty/graphite/src/Justifier.cpp
@@ -0,0 +1,282 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+#include "inc/Segment.h"
+#include "graphite2/Font.h"
+#include "inc/debug.h"
+#include "inc/CharInfo.h"
+#include "inc/Slot.h"
+#include "inc/Main.h"
+#include <cmath>
+
+using namespace graphite2;
+
+class JustifyTotal {
+public:
+    JustifyTotal() : m_numGlyphs(0), m_tStretch(0), m_tShrink(0), m_tStep(0), m_tWeight(0) {}
+    void accumulate(Slot *s, Segment *seg, int level);
+    int weight() const { return m_tWeight; }
+
+    CLASS_NEW_DELETE
+
+private:
+    int m_numGlyphs;
+    int m_tStretch;
+    int m_tShrink;
+    int m_tStep;
+    int m_tWeight;
+};
+
+void JustifyTotal::accumulate(Slot *s, Segment *seg, int level)
+{
+    ++m_numGlyphs;
+    m_tStretch += s->getJustify(seg, level, 0);
+    m_tShrink += s->getJustify(seg, level, 1);
+    m_tStep += s->getJustify(seg, level, 2);
+    m_tWeight += s->getJustify(seg, level, 3);
+}
+
+float Segment::justify(Slot *pSlot, const Font *font, float width, GR_MAYBE_UNUSED justFlags jflags, Slot *pFirst, Slot *pLast)
+{
+    Slot *end = last();
+    float currWidth = 0.0;
+    const float scale = font ? font->scale() : 1.0f;
+    Position res;
+
+    if (width < 0 && !(silf()->flags()))
+        return width;
+
+    if ((m_dir & 1) != m_silf->dir() && m_silf->bidiPass() != m_silf->numPasses())
+    {
+        reverseSlots();
+        std::swap(pFirst, pLast);
+    }
+    if (!pFirst) pFirst = pSlot;
+    while (!pFirst->isBase()) pFirst = pFirst->attachedTo();
+    if (!pLast) pLast = last();
+    while (!pLast->isBase()) pLast = pLast->attachedTo();
+    const float base = pFirst->origin().x / scale;
+    width = width / scale;
+    if ((jflags & gr_justEndInline) == 0)
+    {
+        while (pLast != pFirst && pLast)
+        {
+            Rect bbox = theGlyphBBoxTemporary(pLast->glyph());
+            if (bbox.bl.x != 0.f || bbox.bl.y != 0.f || bbox.tr.x != 0.f || bbox.tr.y == 0.f)
+                break;
+            pLast = pLast->prev();
+        }
+    }
+
+    if (pLast)
+        end = pLast->nextSibling();
+    if (pFirst)
+        pFirst = pFirst->nextSibling();
+
+    int icount = 0;
+    int numLevels = silf()->numJustLevels();
+    if (!numLevels)
+    {
+        for (Slot *s = pSlot; s && s != end; s = s->nextSibling())
+        {
+            CharInfo *c = charinfo(s->before());
+            if (isWhitespace(c->unicodeChar()))
+            {
+                s->setJustify(this, 0, 3, 1);
+                s->setJustify(this, 0, 2, 1);
+                s->setJustify(this, 0, 0, -1);
+                ++icount;
+            }
+        }
+        if (!icount)
+        {
+            for (Slot *s = pSlot; s && s != end; s = s->nextSibling())
+            {
+                s->setJustify(this, 0, 3, 1);
+                s->setJustify(this, 0, 2, 1);
+                s->setJustify(this, 0, 0, -1);
+            }
+        }
+        ++numLevels;
+    }
+
+    Vector<JustifyTotal> stats(numLevels);
+    for (Slot *s = pFirst; s && s != end; s = s->nextSibling())
+    {
+        float w = s->origin().x / scale + s->advance() - base;
+        if (w > currWidth) currWidth = w;
+        for (int j = 0; j < numLevels; ++j)
+            stats[j].accumulate(s, this, j);
+        s->just(0);
+    }
+
+    for (int i = (width < 0.0f) ? -1 : numLevels - 1; i >= 0; --i)
+    {
+        float diff;
+        float error = 0.;
+        float diffpw;
+        int tWeight = stats[i].weight();
+        if (tWeight == 0) continue;
+
+        do {
+            error = 0.;
+            diff = width - currWidth;
+            diffpw = diff / tWeight;
+            tWeight = 0;
+            for (Slot *s = pFirst; s && s != end; s = s->nextSibling()) // don't include final glyph
+            {
+                int w = s->getJustify(this, i, 3);
+                float pref = diffpw * w + error;
+                int step = s->getJustify(this, i, 2);
+                if (!step) step = 1;        // handle lazy font developers
+                if (pref > 0)
+                {
+                    float max = uint16(s->getJustify(this, i, 0));
+                    if (i == 0) max -= s->just();
+                    if (pref > max) pref = max;
+                    else tWeight += w;
+                }
+                else
+                {
+                    float max = uint16(s->getJustify(this, i, 1));
+                    if (i == 0) max += s->just();
+                    if (-pref > max) pref = -max;
+                    else tWeight += w;
+                }
+                int actual = int(pref / step) * step;
+
+                if (actual)
+                {
+                    error += diffpw * w - actual;
+                    if (i == 0)
+                        s->just(s->just() + actual);
+                    else
+                        s->setJustify(this, i, 4, actual);
+                }
+            }
+            currWidth += diff - error;
+        } while (i == 0 && int(std::abs(error)) > 0 && tWeight);
+    }
+
+    Slot *oldFirst = m_first;
+    Slot *oldLast = m_last;
+    if (silf()->flags() & 1)
+    {
+        m_first = pSlot = addLineEnd(pSlot);
+        m_last = pLast = addLineEnd(end);
+        if (!m_first || !m_last) return -1.0;
+    }
+    else
+    {
+        m_first = pSlot;
+        m_last = pLast;
+    }
+
+    // run justification passes here
+#if !defined GRAPHITE2_NTRACING
+    json * const dbgout = m_face->logger();
+    if (dbgout)
+        *dbgout << json::object
+                    << "justifies"  << objectid(this)
+                    << "passes"     << json::array;
+#endif
+
+    if (m_silf->justificationPass() != m_silf->positionPass() && (width >= 0.f || (silf()->flags() & 1)))
+        m_silf->runGraphite(this, m_silf->justificationPass(), m_silf->positionPass());
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+    {
+        *dbgout     << json::item << json::close; // Close up the passes array
+        positionSlots(NULL, pSlot, pLast, m_dir);
+        Slot *lEnd = pLast->nextSibling();
+        *dbgout << "output" << json::array;
+        for(Slot * t = pSlot; t != lEnd; t = t->next())
+            *dbgout     << dslot(this, t);
+        *dbgout         << json::close << json::close;
+    }
+#endif
+
+    res = positionSlots(font, pSlot, pLast, m_dir);
+
+    if (silf()->flags() & 1)
+    {
+        if (m_first)
+            delLineEnd(m_first);
+        if (m_last)
+            delLineEnd(m_last);
+    }
+    m_first = oldFirst;
+    m_last = oldLast;
+
+    if ((m_dir & 1) != m_silf->dir() && m_silf->bidiPass() != m_silf->numPasses())
+        reverseSlots();
+    return res.x;
+}
+
+Slot *Segment::addLineEnd(Slot *nSlot)
+{
+    Slot *eSlot = newSlot();
+    if (!eSlot) return NULL;
+    const uint16 gid = silf()->endLineGlyphid();
+    const GlyphFace * theGlyph = m_face->glyphs().glyphSafe(gid);
+    eSlot->setGlyph(this, gid, theGlyph);
+    if (nSlot)
+    {
+        eSlot->next(nSlot);
+        eSlot->prev(nSlot->prev());
+        nSlot->prev(eSlot);
+        eSlot->before(nSlot->before());
+        if (eSlot->prev())
+            eSlot->after(eSlot->prev()->after());
+        else
+            eSlot->after(nSlot->before());
+    }
+    else
+    {
+        nSlot = m_last;
+        eSlot->prev(nSlot);
+        nSlot->next(eSlot);
+        eSlot->after(eSlot->prev()->after());
+        eSlot->before(nSlot->after());
+    }
+    return eSlot;
+}
+
+void Segment::delLineEnd(Slot *s)
+{
+    Slot *nSlot = s->next();
+    if (nSlot)
+    {
+        nSlot->prev(s->prev());
+        if (s->prev())
+            s->prev()->next(nSlot);
+    }
+    else
+        s->prev()->next(NULL);
+    freeSlot(s);
+}
diff --git a/thirdparty/graphite/src/NameTable.cpp b/thirdparty/graphite/src/NameTable.cpp
new file mode 100644
index 0000000000..d42b7f95bd
--- /dev/null
+++ b/thirdparty/graphite/src/NameTable.cpp
@@ -0,0 +1,254 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/Main.h"
+#include "inc/Endian.h"
+
+#include "inc/NameTable.h"
+#include "inc/UtfCodec.h"
+
+using namespace graphite2;
+
+NameTable::NameTable(const void* data, size_t length, uint16 platformId, uint16 encodingID)
+ : m_platformId(0), m_encodingId(0), m_languageCount(0),
+   m_platformOffset(0), m_platformLastRecord(0), m_nameDataLength(0),
+   m_table(0), m_nameData(NULL)
+{
+    void *pdata = gralloc<byte>(length);
+    if (!pdata) return;
+    memcpy(pdata, data, length);
+    m_table = reinterpret_cast<const TtfUtil::Sfnt::FontNames*>(pdata);
+
+    if ((length > sizeof(TtfUtil::Sfnt::FontNames)) &&
+        (length > sizeof(TtfUtil::Sfnt::FontNames) +
+         sizeof(TtfUtil::Sfnt::NameRecord) * ( be::swap<uint16>(m_table->count) - 1)))
+    {
+        uint16 offset = be::swap<uint16>(m_table->string_offset);
+        if (offset < length)
+        {
+            m_nameData = reinterpret_cast<const uint8*>(pdata) + offset;
+            setPlatformEncoding(platformId, encodingID);
+            m_nameDataLength = uint16(length - offset);
+            return;
+        }
+    }
+    free(const_cast<TtfUtil::Sfnt::FontNames*>(m_table));
+    m_table = NULL;
+}
+
+uint16 NameTable::setPlatformEncoding(uint16 platformId, uint16 encodingID)
+{
+    if (!m_nameData) return 0;
+    uint16 i = 0;
+    uint16 count = be::swap<uint16>(m_table->count);
+    for (; i < count; i++)
+    {
+        if (be::swap<uint16>(m_table->name_record[i].platform_id) == platformId &&
+            be::swap<uint16>(m_table->name_record[i].platform_specific_id) == encodingID)
+        {
+            m_platformOffset = i;
+            break;
+        }
+    }
+    while ((++i < count) &&
+           (be::swap<uint16>(m_table->name_record[i].platform_id) == platformId) &&
+           (be::swap<uint16>(m_table->name_record[i].platform_specific_id) == encodingID))
+    {
+        m_platformLastRecord = i;
+    }
+    m_encodingId = encodingID;
+    m_platformId = platformId;
+    return 0;
+}
+
+void* NameTable::getName(uint16& languageId, uint16 nameId, gr_encform enc, uint32& length)
+{
+    uint16 anyLang = 0;
+    uint16 enUSLang = 0;
+    uint16 bestLang = 0;
+    if (!m_table)
+    {
+        languageId = 0;
+        length = 0;
+        return NULL;
+    }
+    for (uint16 i = m_platformOffset; i <= m_platformLastRecord; i++)
+    {
+        if (be::swap<uint16>(m_table->name_record[i].name_id) == nameId)
+        {
+            uint16 langId = be::swap<uint16>(m_table->name_record[i].language_id);
+            if (langId == languageId)
+            {
+                bestLang = i;
+                break;
+            }
+            // MS language tags have the language in the lower byte, region in the higher
+            else if ((langId & 0xFF) == (languageId & 0xFF))
+            {
+                bestLang = i;
+            }
+            else if (langId == 0x409)
+            {
+                enUSLang = i;
+            }
+            else
+            {
+                anyLang = i;
+            }
+        }
+    }
+    if (!bestLang)
+    {
+        if (enUSLang) bestLang = enUSLang;
+        else
+        {
+            bestLang = anyLang;
+            if (!anyLang)
+            {
+                languageId = 0;
+                length = 0;
+                return NULL;
+            }
+        }
+    }
+    const TtfUtil::Sfnt::NameRecord & nameRecord = m_table->name_record[bestLang];
+    languageId = be::swap<uint16>(nameRecord.language_id);
+    uint16 utf16Length = be::swap<uint16>(nameRecord.length);
+    uint16 offset = be::swap<uint16>(nameRecord.offset);
+    if(offset + utf16Length > m_nameDataLength)
+    {
+        languageId = 0;
+        length = 0;
+        return NULL;
+    }
+    utf16Length >>= 1; // in utf16 units
+    utf16::codeunit_t * utf16Name = gralloc<utf16::codeunit_t>(utf16Length + 1);
+    if (!utf16Name)
+    {
+        languageId = 0;
+        length = 0;
+        return NULL;
+    }
+    const uint8* pName = m_nameData + offset;
+    for (size_t i = 0; i < utf16Length; i++)
+    {
+        utf16Name[i] = be::read<uint16>(pName);
+    }
+    utf16Name[utf16Length] = 0;
+    if (!utf16::validate(utf16Name, utf16Name + utf16Length))
+    {
+        free(utf16Name);
+        languageId = 0;
+        length = 0;
+        return NULL;
+    }
+    switch (enc)
+    {
+    case gr_utf8:
+    {
+        utf8::codeunit_t* uniBuffer = gralloc<utf8::codeunit_t>(3 * utf16Length + 1);
+        if (!uniBuffer)
+        {
+            free(utf16Name);
+            languageId = 0;
+            length = 0;
+            return NULL;
+        }
+        utf8::iterator d = uniBuffer;
+        for (utf16::const_iterator s = utf16Name, e = utf16Name + utf16Length; s != e; ++s, ++d)
+            *d = *s;
+        length = uint32(d - uniBuffer);
+        uniBuffer[length] = 0;
+        free(utf16Name);
+        return uniBuffer;
+    }
+    case gr_utf16:
+        length = utf16Length;
+        return utf16Name;
+    case gr_utf32:
+    {
+        utf32::codeunit_t * uniBuffer = gralloc<utf32::codeunit_t>(utf16Length  + 1);
+        if (!uniBuffer)
+        {
+            free(utf16Name);
+            languageId = 0;
+            length = 0;
+            return NULL;
+        }
+        utf32::iterator d = uniBuffer;
+        for (utf16::const_iterator s = utf16Name, e = utf16Name + utf16Length; s != e; ++s, ++d)
+            *d = *s;
+        length = uint32(d - uniBuffer);
+        uniBuffer[length] = 0;
+        free(utf16Name);
+        return uniBuffer;
+    }
+    }
+    free(utf16Name);
+    languageId = 0;
+    length = 0;
+    return NULL;
+}
+
+uint16 NameTable::getLanguageId(const char * bcp47Locale)
+{
+    size_t localeLength = strlen(bcp47Locale);
+    uint16 localeId = m_locale2Lang.getMsId(bcp47Locale);
+    if (m_table && (be::swap<uint16>(m_table->format) == 1))
+    {
+        const uint8 * pLangEntries = reinterpret_cast<const uint8*>(m_table) +
+            sizeof(TtfUtil::Sfnt::FontNames)
+            + sizeof(TtfUtil::Sfnt::NameRecord) * ( be::swap<uint16>(m_table->count) - 1);
+        uint16 numLangEntries = be::read<uint16>(pLangEntries);
+        const TtfUtil::Sfnt::LangTagRecord * langTag =
+            reinterpret_cast<const TtfUtil::Sfnt::LangTagRecord*>(pLangEntries);
+        if (pLangEntries + numLangEntries * sizeof(TtfUtil::Sfnt::LangTagRecord) <= m_nameData)
+        {
+            for (uint16 i = 0; i < numLangEntries; i++)
+            {
+                uint16 offset = be::swap<uint16>(langTag[i].offset);
+                uint16 length = be::swap<uint16>(langTag[i].length);
+                if ((offset + length <= m_nameDataLength) && (length == 2 * localeLength))
+                {
+                    const uint8* pName = m_nameData + offset;
+                    bool match = true;
+                    for (size_t j = 0; j < localeLength; j++)
+                    {
+                        uint16 code = be::read<uint16>(pName);
+                        if ((code > 0x7F) || (code != bcp47Locale[j]))
+                        {
+                            match = false;
+                            break;
+                        }
+                    }
+                    if (match)
+                        return 0x8000 + i;
+                }
+            }
+        }
+    }
+    return localeId;
+}
diff --git a/thirdparty/graphite/src/Pass.cpp b/thirdparty/graphite/src/Pass.cpp
new file mode 100644
index 0000000000..db31c22d46
--- /dev/null
+++ b/thirdparty/graphite/src/Pass.cpp
@@ -0,0 +1,1107 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/Main.h"
+#include "inc/debug.h"
+#include "inc/Endian.h"
+#include "inc/Pass.h"
+#include <cstring>
+#include <cstdlib>
+#include <cassert>
+#include <cmath>
+#include "inc/Segment.h"
+#include "inc/Code.h"
+#include "inc/Rule.h"
+#include "inc/Error.h"
+#include "inc/Collider.h"
+
+using namespace graphite2;
+using vm::Machine;
+typedef Machine::Code  Code;
+
+enum KernCollison
+{
+    None       = 0,
+    CrossSpace = 1,
+    InWord     = 2,
+    reserved   = 3
+};
+
+Pass::Pass()
+: m_silf(0),
+  m_cols(0),
+  m_rules(0),
+  m_ruleMap(0),
+  m_startStates(0),
+  m_transitions(0),
+  m_states(0),
+  m_codes(0),
+  m_progs(0),
+  m_numCollRuns(0),
+  m_kernColls(0),
+  m_iMaxLoop(0),
+  m_numGlyphs(0),
+  m_numRules(0),
+  m_numStates(0),
+  m_numTransition(0),
+  m_numSuccess(0),
+  m_successStart(0),
+  m_numColumns(0),
+  m_minPreCtxt(0),
+  m_maxPreCtxt(0),
+  m_colThreshold(0),
+  m_isReverseDir(false)
+{
+}
+
+Pass::~Pass()
+{
+    free(m_cols);
+    free(m_startStates);
+    free(m_transitions);
+    free(m_states);
+    free(m_ruleMap);
+
+    if (m_rules) delete [] m_rules;
+    if (m_codes) delete [] m_codes;
+    free(m_progs);
+}
+
+bool Pass::readPass(const byte * const pass_start, size_t pass_length, size_t subtable_base,
+        GR_MAYBE_UNUSED Face & face, passtype pt, GR_MAYBE_UNUSED uint32 version, Error &e)
+{
+    const byte * p              = pass_start,
+               * const pass_end = p + pass_length;
+    size_t numRanges;
+
+    if (e.test(pass_length < 40, E_BADPASSLENGTH)) return face.error(e);
+    // Read in basic values
+    const byte flags = be::read<byte>(p);
+    if (e.test((flags & 0x1f) &&
+            (pt < PASS_TYPE_POSITIONING || !m_silf->aCollision() || !face.glyphs().hasBoxes() || !(m_silf->flags() & 0x20)),
+            E_BADCOLLISIONPASS))
+        return face.error(e);
+    m_numCollRuns = flags & 0x7;
+    m_kernColls   = (flags >> 3) & 0x3;
+    m_isReverseDir = (flags >> 5) & 0x1;
+    m_iMaxLoop = be::read<byte>(p);
+    if (m_iMaxLoop < 1) m_iMaxLoop = 1;
+    be::skip<byte>(p,2); // skip maxContext & maxBackup
+    m_numRules = be::read<uint16>(p);
+    if (e.test(!m_numRules && m_numCollRuns == 0, E_BADEMPTYPASS)) return face.error(e);
+    be::skip<uint16>(p);   // fsmOffset - not sure why we would want this
+    const byte * const pcCode = pass_start + be::read<uint32>(p) - subtable_base,
+               * const rcCode = pass_start + be::read<uint32>(p) - subtable_base,
+               * const aCode  = pass_start + be::read<uint32>(p) - subtable_base;
+    be::skip<uint32>(p);
+    m_numStates = be::read<uint16>(p);
+    m_numTransition = be::read<uint16>(p);
+    m_numSuccess = be::read<uint16>(p);
+    m_numColumns = be::read<uint16>(p);
+    numRanges = be::read<uint16>(p);
+    be::skip<uint16>(p, 3); // skip searchRange, entrySelector & rangeShift.
+    assert(p - pass_start == 40);
+    // Perform some sanity checks.
+    if ( e.test(m_numTransition > m_numStates, E_BADNUMTRANS)
+            || e.test(m_numSuccess > m_numStates, E_BADNUMSUCCESS)
+            || e.test(m_numSuccess + m_numTransition < m_numStates, E_BADNUMSTATES)
+            || e.test(m_numRules && numRanges == 0, E_NORANGES)
+            || e.test(m_numColumns > 0x7FFF, E_BADNUMCOLUMNS))
+        return face.error(e);
+
+    m_successStart = m_numStates - m_numSuccess;
+    // test for beyond end - 1 to account for reading uint16
+    if (e.test(p + numRanges * 6 - 2 > pass_end, E_BADPASSLENGTH)) return face.error(e);
+    m_numGlyphs = be::peek<uint16>(p + numRanges * 6 - 4) + 1;
+    // Calculate the start of various arrays.
+    const byte * const ranges = p;
+    be::skip<uint16>(p, numRanges*3);
+    const byte * const o_rule_map = p;
+    be::skip<uint16>(p, m_numSuccess + 1);
+
+    // More sanity checks
+    if (e.test(reinterpret_cast<const byte *>(o_rule_map + m_numSuccess*sizeof(uint16)) > pass_end
+            || p > pass_end, E_BADRULEMAPLEN))
+        return face.error(e);
+    const size_t numEntries = be::peek<uint16>(o_rule_map + m_numSuccess*sizeof(uint16));
+    const byte * const   rule_map = p;
+    be::skip<uint16>(p, numEntries);
+
+    if (e.test(p + 2*sizeof(uint8) > pass_end, E_BADPASSLENGTH)) return face.error(e);
+    m_minPreCtxt = be::read<uint8>(p);
+    m_maxPreCtxt = be::read<uint8>(p);
+    if (e.test(m_minPreCtxt > m_maxPreCtxt, E_BADCTXTLENBOUNDS)) return face.error(e);
+    const byte * const start_states = p;
+    be::skip<int16>(p, m_maxPreCtxt - m_minPreCtxt + 1);
+    const uint16 * const sort_keys = reinterpret_cast<const uint16 *>(p);
+    be::skip<uint16>(p, m_numRules);
+    const byte * const precontext = p;
+    be::skip<byte>(p, m_numRules);
+
+    if (e.test(p + sizeof(uint16) + sizeof(uint8) > pass_end, E_BADCTXTLENS)) return face.error(e);
+    m_colThreshold = be::read<uint8>(p);
+    if (m_colThreshold == 0) m_colThreshold = 10;       // A default
+    const size_t pass_constraint_len = be::read<uint16>(p);
+
+    const uint16 * const o_constraint = reinterpret_cast<const uint16 *>(p);
+    be::skip<uint16>(p, m_numRules + 1);
+    const uint16 * const o_actions = reinterpret_cast<const uint16 *>(p);
+    be::skip<uint16>(p, m_numRules + 1);
+    const byte * const states = p;
+    if (e.test(2u*m_numTransition*m_numColumns >= (unsigned)(pass_end - p), E_BADPASSLENGTH)
+            || e.test(p >= pass_end, E_BADPASSLENGTH))
+        return face.error(e);
+    be::skip<int16>(p, m_numTransition*m_numColumns);
+    be::skip<uint8>(p);
+    if (e.test(p != pcCode, E_BADPASSCCODEPTR)) return face.error(e);
+    be::skip<byte>(p, pass_constraint_len);
+    if (e.test(p != rcCode, E_BADRULECCODEPTR)
+        || e.test(size_t(rcCode - pcCode) != pass_constraint_len, E_BADCCODELEN)) return face.error(e);
+    be::skip<byte>(p, be::peek<uint16>(o_constraint + m_numRules));
+    if (e.test(p != aCode, E_BADACTIONCODEPTR)) return face.error(e);
+    be::skip<byte>(p, be::peek<uint16>(o_actions + m_numRules));
+
+    // We should be at the end or within the pass
+    if (e.test(p > pass_end, E_BADPASSLENGTH)) return face.error(e);
+
+    // Load the pass constraint if there is one.
+    if (pass_constraint_len)
+    {
+        face.error_context(face.error_context() + 1);
+        m_cPConstraint = vm::Machine::Code(true, pcCode, pcCode + pass_constraint_len,
+                                  precontext[0], be::peek<uint16>(sort_keys), *m_silf, face, PASS_TYPE_UNKNOWN);
+        if (e.test(!m_cPConstraint, E_OUTOFMEM)
+                || e.test(m_cPConstraint.status() != Code::loaded, m_cPConstraint.status() + E_CODEFAILURE))
+            return face.error(e);
+        face.error_context(face.error_context() - 1);
+    }
+    if (m_numRules)
+    {
+        if (!readRanges(ranges, numRanges, e)) return face.error(e);
+        if (!readRules(rule_map, numEntries,  precontext, sort_keys,
+                   o_constraint, rcCode, o_actions, aCode, face, pt, e)) return false;
+    }
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::category _states_cat(face.tele.states);
+#endif
+    return m_numRules ? readStates(start_states, states, o_rule_map, face, e) : true;
+}
+
+
+bool Pass::readRules(const byte * rule_map, const size_t num_entries,
+                     const byte *precontext, const uint16 * sort_key,
+                     const uint16 * o_constraint, const byte *rc_data,
+                     const uint16 * o_action,     const byte * ac_data,
+                     Face & face, passtype pt, Error &e)
+{
+    const byte * const ac_data_end = ac_data + be::peek<uint16>(o_action + m_numRules);
+    const byte * const rc_data_end = rc_data + be::peek<uint16>(o_constraint + m_numRules);
+
+    precontext += m_numRules;
+    sort_key   += m_numRules;
+    o_constraint += m_numRules;
+    o_action += m_numRules;
+
+    // Load rules.
+    const byte * ac_begin = 0, * rc_begin = 0,
+               * ac_end = ac_data + be::peek<uint16>(o_action),
+               * rc_end = rc_data + be::peek<uint16>(o_constraint);
+
+    // Allocate pools
+    m_rules = new Rule [m_numRules];
+    m_codes = new Code [m_numRules*2];
+    int totalSlots = 0;
+    const uint16 *tsort = sort_key;
+    for (int i = 0; i < m_numRules; ++i)
+        totalSlots += be::peek<uint16>(--tsort);
+    const size_t prog_pool_sz = vm::Machine::Code::estimateCodeDataOut(ac_end - ac_data + rc_end - rc_data, 2 * m_numRules, totalSlots);
+    m_progs = gralloc<byte>(prog_pool_sz);
+    byte * prog_pool_free = m_progs,
+         * prog_pool_end  = m_progs + prog_pool_sz;
+    if (e.test(!(m_rules && m_codes && m_progs), E_OUTOFMEM)) return face.error(e);
+
+    Rule * r = m_rules + m_numRules - 1;
+    for (size_t n = m_numRules; r >= m_rules; --n, --r, ac_end = ac_begin, rc_end = rc_begin)
+    {
+        face.error_context((face.error_context() & 0xFFFF00) + EC_ARULE + int((n - 1) << 24));
+        r->preContext = *--precontext;
+        r->sort       = be::peek<uint16>(--sort_key);
+#ifndef NDEBUG
+        r->rule_idx   = uint16(n - 1);
+#endif
+        if (r->sort > 63 || r->preContext >= r->sort || r->preContext > m_maxPreCtxt || r->preContext < m_minPreCtxt)
+            return false;
+        ac_begin      = ac_data + be::peek<uint16>(--o_action);
+        --o_constraint;
+        rc_begin      = be::peek<uint16>(o_constraint) ? rc_data + be::peek<uint16>(o_constraint) : rc_end;
+
+        if (ac_begin > ac_end || ac_begin > ac_data_end || ac_end > ac_data_end
+                || rc_begin > rc_end || rc_begin > rc_data_end || rc_end > rc_data_end
+                || vm::Machine::Code::estimateCodeDataOut(ac_end - ac_begin + rc_end - rc_begin, 2, r->sort) > size_t(prog_pool_end - prog_pool_free))
+            return false;
+        r->action     = new (m_codes+n*2-2) vm::Machine::Code(false, ac_begin, ac_end, r->preContext, r->sort, *m_silf, face, pt, &prog_pool_free);
+        r->constraint = new (m_codes+n*2-1) vm::Machine::Code(true,  rc_begin, rc_end, r->preContext, r->sort, *m_silf, face, pt, &prog_pool_free);
+
+        if (e.test(!r->action || !r->constraint, E_OUTOFMEM)
+                || e.test(r->action->status() != Code::loaded, r->action->status() + E_CODEFAILURE)
+                || e.test(r->constraint->status() != Code::loaded, r->constraint->status() + E_CODEFAILURE)
+                || e.test(!r->constraint->immutable(), E_MUTABLECCODE))
+            return face.error(e);
+    }
+
+    byte * const moved_progs = prog_pool_free > m_progs ? static_cast<byte *>(realloc(m_progs, prog_pool_free - m_progs)) : 0;
+    if (e.test(!moved_progs, E_OUTOFMEM))
+    {
+        free(m_progs);
+        m_progs = 0;
+        return face.error(e);
+    }
+
+    if (moved_progs != m_progs)
+    {
+        for (Code * c = m_codes, * const ce = c + m_numRules*2; c != ce; ++c)
+        {
+            c->externalProgramMoved(moved_progs - m_progs);
+        }
+        m_progs = moved_progs;
+    }
+
+    // Load the rule entries map
+    face.error_context((face.error_context() & 0xFFFF00) + EC_APASS);
+    //TODO: Coverity: 1315804: FORWARD_NULL
+    RuleEntry * re = m_ruleMap = gralloc<RuleEntry>(num_entries);
+    if (e.test(!re, E_OUTOFMEM)) return face.error(e);
+    for (size_t n = num_entries; n; --n, ++re)
+    {
+        const ptrdiff_t rn = be::read<uint16>(rule_map);
+        if (e.test(rn >= m_numRules, E_BADRULENUM))  return face.error(e);
+        re->rule = m_rules + rn;
+    }
+
+    return true;
+}
+
+static int cmpRuleEntry(const void *a, const void *b) { return (*(RuleEntry *)a < *(RuleEntry *)b ? -1 :
+                                                                (*(RuleEntry *)b < *(RuleEntry *)a ? 1 : 0)); }
+
+bool Pass::readStates(const byte * starts, const byte *states, const byte * o_rule_map, GR_MAYBE_UNUSED Face & face, Error &e)
+{
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::category _states_cat(face.tele.starts);
+#endif
+    m_startStates = gralloc<uint16>(m_maxPreCtxt - m_minPreCtxt + 1);
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::set_category(face.tele.states);
+#endif
+    m_states      = gralloc<State>(m_numStates);
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::set_category(face.tele.transitions);
+#endif
+    m_transitions      = gralloc<uint16>(m_numTransition * m_numColumns);
+
+    if (e.test(!m_startStates || !m_states || !m_transitions, E_OUTOFMEM)) return face.error(e);
+    // load start states
+    for (uint16 * s = m_startStates,
+                * const s_end = s + m_maxPreCtxt - m_minPreCtxt + 1; s != s_end; ++s)
+    {
+        *s = be::read<uint16>(starts);
+        if (e.test(*s >= m_numStates, E_BADSTATE))
+        {
+            face.error_context((face.error_context() & 0xFFFF00) + EC_ASTARTS + int((s - m_startStates) << 24));
+            return face.error(e); // true;
+        }
+    }
+
+    // load state transition table.
+    for (uint16 * t = m_transitions,
+                * const t_end = t + m_numTransition*m_numColumns; t != t_end; ++t)
+    {
+        *t = be::read<uint16>(states);
+        if (e.test(*t >= m_numStates, E_BADSTATE))
+        {
+            face.error_context((face.error_context() & 0xFFFF00) + EC_ATRANS + int(((t - m_transitions) / m_numColumns) << 8));
+            return face.error(e);
+        }
+    }
+
+    State * s = m_states,
+          * const success_begin = m_states + m_numStates - m_numSuccess;
+    const RuleEntry * rule_map_end = m_ruleMap + be::peek<uint16>(o_rule_map + m_numSuccess*sizeof(uint16));
+    for (size_t n = m_numStates; n; --n, ++s)
+    {
+        RuleEntry * const begin = s < success_begin ? 0 : m_ruleMap + be::read<uint16>(o_rule_map),
+                  * const end   = s < success_begin ? 0 : m_ruleMap + be::peek<uint16>(o_rule_map);
+
+        if (e.test(begin >= rule_map_end || end > rule_map_end || begin > end, E_BADRULEMAPPING))
+        {
+            face.error_context((face.error_context() & 0xFFFF00) + EC_ARULEMAP + int(n << 24));
+            return face.error(e);
+        }
+        s->rules = begin;
+        s->rules_end = (end - begin <= FiniteStateMachine::MAX_RULES)? end :
+            begin + FiniteStateMachine::MAX_RULES;
+        if (begin)      // keep UBSan happy can't call qsort with null begin
+            qsort(begin, end - begin, sizeof(RuleEntry), &cmpRuleEntry);
+    }
+
+    return true;
+}
+
+bool Pass::readRanges(const byte * ranges, size_t num_ranges, Error &e)
+{
+    m_cols = gralloc<uint16>(m_numGlyphs);
+    if (e.test(!m_cols, E_OUTOFMEM)) return false;
+    memset(m_cols, 0xFF, m_numGlyphs * sizeof(uint16));
+    for (size_t n = num_ranges; n; --n)
+    {
+        uint16     * ci     = m_cols + be::read<uint16>(ranges),
+                   * ci_end = m_cols + be::read<uint16>(ranges) + 1,
+                     col    = be::read<uint16>(ranges);
+
+        if (e.test(ci >= ci_end || ci_end > m_cols+m_numGlyphs || col >= m_numColumns, E_BADRANGE))
+            return false;
+
+        // A glyph must only belong to one column at a time
+        while (ci != ci_end && *ci == 0xffff)
+            *ci++ = col;
+
+        if (e.test(ci != ci_end, E_BADRANGE))
+            return false;
+    }
+    return true;
+}
+
+
+bool Pass::runGraphite(vm::Machine & m, FiniteStateMachine & fsm, bool reverse) const
+{
+    Slot *s = m.slotMap().segment.first();
+    if (!s || !testPassConstraint(m)) return true;
+    if (reverse)
+    {
+        m.slotMap().segment.reverseSlots();
+        s = m.slotMap().segment.first();
+    }
+    if (m_numRules)
+    {
+        Slot *currHigh = s->next();
+
+#if !defined GRAPHITE2_NTRACING
+        if (fsm.dbgout)  *fsm.dbgout << "rules" << json::array;
+        json::closer rules_array_closer(fsm.dbgout);
+#endif
+
+        m.slotMap().highwater(currHigh);
+        int lc = m_iMaxLoop;
+        do
+        {
+            findNDoRule(s, m, fsm);
+            if (m.status() != Machine::finished) return false;
+            if (s && (s == m.slotMap().highwater() || m.slotMap().highpassed() || --lc == 0)) {
+                if (!lc)
+                    s = m.slotMap().highwater();
+                lc = m_iMaxLoop;
+                if (s)
+                    m.slotMap().highwater(s->next());
+            }
+        } while (s);
+    }
+    //TODO: Use enums for flags
+    const bool collisions = m_numCollRuns || m_kernColls;
+
+    if (!collisions || !m.slotMap().segment.hasCollisionInfo())
+        return true;
+
+    if (m_numCollRuns)
+    {
+        if (!(m.slotMap().segment.flags() & Segment::SEG_INITCOLLISIONS))
+        {
+            m.slotMap().segment.positionSlots(0, 0, 0, m.slotMap().dir(), true);
+//            m.slotMap().segment.flags(m.slotMap().segment.flags() | Segment::SEG_INITCOLLISIONS);
+        }
+        if (!collisionShift(&m.slotMap().segment, m.slotMap().dir(), fsm.dbgout))
+            return false;
+    }
+    if ((m_kernColls) && !collisionKern(&m.slotMap().segment, m.slotMap().dir(), fsm.dbgout))
+        return false;
+    if (collisions && !collisionFinish(&m.slotMap().segment, fsm.dbgout))
+        return false;
+    return true;
+}
+
+bool Pass::runFSM(FiniteStateMachine& fsm, Slot * slot) const
+{
+    fsm.reset(slot, m_maxPreCtxt);
+    if (fsm.slots.context() < m_minPreCtxt)
+        return false;
+
+    uint16 state = m_startStates[m_maxPreCtxt - fsm.slots.context()];
+    uint8  free_slots = SlotMap::MAX_SLOTS;
+    do
+    {
+        fsm.slots.pushSlot(slot);
+        if (slot->gid() >= m_numGlyphs
+         || m_cols[slot->gid()] == 0xffffU
+         || --free_slots == 0
+         || state >= m_numTransition)
+            return free_slots != 0;
+
+        const uint16 * transitions = m_transitions + state*m_numColumns;
+        state = transitions[m_cols[slot->gid()]];
+        if (state >= m_successStart)
+            fsm.rules.accumulate_rules(m_states[state]);
+
+        slot = slot->next();
+    } while (state != 0 && slot);
+
+    fsm.slots.pushSlot(slot);
+    return true;
+}
+
+#if !defined GRAPHITE2_NTRACING
+
+inline
+Slot * input_slot(const SlotMap &  slots, const int n)
+{
+    Slot * s = slots[slots.context() + n];
+    if (!s->isCopied())     return s;
+
+    return s->prev() ? s->prev()->next() : (s->next() ? s->next()->prev() : slots.segment.last());
+}
+
+inline
+Slot * output_slot(const SlotMap &  slots, const int n)
+{
+    Slot * s = slots[slots.context() + n - 1];
+    return s ? s->next() : slots.segment.first();
+}
+
+#endif //!defined GRAPHITE2_NTRACING
+
+void Pass::findNDoRule(Slot * & slot, Machine &m, FiniteStateMachine & fsm) const
+{
+    assert(slot);
+
+    if (runFSM(fsm, slot))
+    {
+        // Search for the first rule which passes the constraint
+        const RuleEntry *        r = fsm.rules.begin(),
+                        * const re = fsm.rules.end();
+        while (r != re && !testConstraint(*r->rule, m))
+        {
+            ++r;
+            if (m.status() != Machine::finished)
+                return;
+        }
+
+#if !defined GRAPHITE2_NTRACING
+        if (fsm.dbgout)
+        {
+            if (fsm.rules.size() != 0)
+            {
+                *fsm.dbgout << json::item << json::object;
+                dumpRuleEventConsidered(fsm, *r);
+                if (r != re)
+                {
+                    const int adv = doAction(r->rule->action, slot, m);
+                    dumpRuleEventOutput(fsm, *r->rule, slot);
+                    if (r->rule->action->deletes()) fsm.slots.collectGarbage(slot);
+                    adjustSlot(adv, slot, fsm.slots);
+                    *fsm.dbgout << "cursor" << objectid(dslot(&fsm.slots.segment, slot))
+                            << json::close; // Close RuelEvent object
+
+                    return;
+                }
+                else
+                {
+                    *fsm.dbgout << json::close  // close "considered" array
+                            << "output" << json::null
+                            << "cursor" << objectid(dslot(&fsm.slots.segment, slot->next()))
+                            << json::close;
+                }
+            }
+        }
+        else
+#endif
+        {
+            if (r != re)
+            {
+                const int adv = doAction(r->rule->action, slot, m);
+                if (m.status() != Machine::finished) return;
+                if (r->rule->action->deletes()) fsm.slots.collectGarbage(slot);
+                adjustSlot(adv, slot, fsm.slots);
+                return;
+            }
+        }
+    }
+
+    slot = slot->next();
+    return;
+}
+
+#if !defined GRAPHITE2_NTRACING
+
+void Pass::dumpRuleEventConsidered(const FiniteStateMachine & fsm, const RuleEntry & re) const
+{
+    *fsm.dbgout << "considered" << json::array;
+    for (const RuleEntry *r = fsm.rules.begin(); r != &re; ++r)
+    {
+        if (r->rule->preContext > fsm.slots.context())
+            continue;
+        *fsm.dbgout << json::flat << json::object
+                    << "id" << r->rule - m_rules
+                    << "failed" << true
+                    << "input" << json::flat << json::object
+                        << "start" << objectid(dslot(&fsm.slots.segment, input_slot(fsm.slots, -r->rule->preContext)))
+                        << "length" << r->rule->sort
+                        << json::close  // close "input"
+                    << json::close; // close Rule object
+    }
+}
+
+
+void Pass::dumpRuleEventOutput(const FiniteStateMachine & fsm, const Rule & r, Slot * const last_slot) const
+{
+    *fsm.dbgout     << json::item << json::flat << json::object
+                        << "id"     << &r - m_rules
+                        << "failed" << false
+                        << "input" << json::flat << json::object
+                            << "start" << objectid(dslot(&fsm.slots.segment, input_slot(fsm.slots, 0)))
+                            << "length" << r.sort - r.preContext
+                            << json::close // close "input"
+                        << json::close  // close Rule object
+                << json::close // close considered array
+                << "output" << json::object
+                    << "range" << json::flat << json::object
+                        << "start"  << objectid(dslot(&fsm.slots.segment, input_slot(fsm.slots, 0)))
+                        << "end"    << objectid(dslot(&fsm.slots.segment, last_slot))
+                    << json::close // close "input"
+                    << "slots"  << json::array;
+    const Position rsb_prepos = last_slot ? last_slot->origin() : fsm.slots.segment.advance();
+    fsm.slots.segment.positionSlots(0, 0, 0, fsm.slots.segment.currdir());
+
+    for(Slot * slot = output_slot(fsm.slots, 0); slot != last_slot; slot = slot->next())
+        *fsm.dbgout     << dslot(&fsm.slots.segment, slot);
+    *fsm.dbgout         << json::close  // close "slots"
+                    << "postshift"  << (last_slot ? last_slot->origin() : fsm.slots.segment.advance()) - rsb_prepos
+                << json::close;         // close "output" object
+
+}
+
+#endif
+
+
+inline
+bool Pass::testPassConstraint(Machine & m) const
+{
+    if (!m_cPConstraint) return true;
+
+    assert(m_cPConstraint.constraint());
+
+    m.slotMap().reset(*m.slotMap().segment.first(), 0);
+    m.slotMap().pushSlot(m.slotMap().segment.first());
+    vm::slotref * map = m.slotMap().begin();
+    const uint32 ret = m_cPConstraint.run(m, map);
+
+#if !defined GRAPHITE2_NTRACING
+    json * const dbgout = m.slotMap().segment.getFace()->logger();
+    if (dbgout)
+        *dbgout << "constraint" << (ret && m.status() == Machine::finished);
+#endif
+
+    return ret && m.status() == Machine::finished;
+}
+
+
+bool Pass::testConstraint(const Rule & r, Machine & m) const
+{
+    const uint16 curr_context = m.slotMap().context();
+    if (unsigned(r.sort + curr_context - r.preContext) > m.slotMap().size()
+        || curr_context - r.preContext < 0) return false;
+
+    vm::slotref * map = m.slotMap().begin() + curr_context - r.preContext;
+    if (map[r.sort - 1] == 0)
+        return false;
+
+    if (!*r.constraint) return true;
+    assert(r.constraint->constraint());
+    for (int n = r.sort; n && map; --n, ++map)
+    {
+        if (!*map) continue;
+        const int32 ret = r.constraint->run(m, map);
+        if (!ret || m.status() != Machine::finished)
+            return false;
+    }
+
+    return true;
+}
+
+
+void SlotMap::collectGarbage(Slot * &aSlot)
+{
+    for(Slot **s = begin(), *const *const se = end() - 1; s != se; ++s) {
+        Slot *& slot = *s;
+        if(slot && (slot->isDeleted() || slot->isCopied()))
+        {
+            if (slot == aSlot)
+                aSlot = slot->prev() ? slot->prev() : slot->next();
+            segment.freeSlot(slot);
+        }
+    }
+}
+
+
+
+int Pass::doAction(const Code *codeptr, Slot * & slot_out, vm::Machine & m) const
+{
+    assert(codeptr);
+    if (!*codeptr) return 0;
+    SlotMap   & smap = m.slotMap();
+    vm::slotref * map = &smap[smap.context()];
+    smap.highpassed(false);
+
+    int32 ret = codeptr->run(m, map);
+
+    if (m.status() != Machine::finished)
+    {
+        slot_out = NULL;
+        smap.highwater(0);
+        return 0;
+    }
+
+    slot_out = *map;
+    return ret;
+}
+
+
+void Pass::adjustSlot(int delta, Slot * & slot_out, SlotMap & smap) const
+{
+    if (!slot_out)
+    {
+        if (smap.highpassed() || slot_out == smap.highwater())
+        {
+            slot_out = smap.segment.last();
+            ++delta;
+            if (!smap.highwater() || smap.highwater() == slot_out)
+                smap.highpassed(false);
+        }
+        else
+        {
+            slot_out = smap.segment.first();
+            --delta;
+        }
+    }
+    if (delta < 0)
+    {
+        while (++delta <= 0 && slot_out)
+        {
+            slot_out = slot_out->prev();
+            if (smap.highpassed() && smap.highwater() == slot_out)
+                smap.highpassed(false);
+        }
+    }
+    else if (delta > 0)
+    {
+        while (--delta >= 0 && slot_out)
+        {
+            if (slot_out == smap.highwater() && slot_out)
+                smap.highpassed(true);
+            slot_out = slot_out->next();
+        }
+    }
+}
+
+bool Pass::collisionShift(Segment *seg, int dir, json * const dbgout) const
+{
+    ShiftCollider shiftcoll(dbgout);
+    // bool isfirst = true;
+    bool hasCollisions = false;
+    Slot *start = seg->first();      // turn on collision fixing for the first slot
+    Slot *end = NULL;
+    bool moved = false;
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+        *dbgout << "collisions" << json::array
+            << json::flat << json::object << "num-loops" << m_numCollRuns << json::close;
+#endif
+
+    while (start)
+    {
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)  *dbgout << json::object << "phase" << "1" << "moves" << json::array;
+#endif
+        hasCollisions = false;
+        end = NULL;
+        // phase 1 : position shiftable glyphs, ignoring kernable glyphs
+        for (Slot *s = start; s; s = s->next())
+        {
+            const SlotCollision * c = seg->collisionInfo(s);
+            if (start && (c->flags() & (SlotCollision::COLL_FIX | SlotCollision::COLL_KERN)) == SlotCollision::COLL_FIX
+                      && !resolveCollisions(seg, s, start, shiftcoll, false, dir, moved, hasCollisions, dbgout))
+                return false;
+            if (s != start && (c->flags() & SlotCollision::COLL_END))
+            {
+                end = s->next();
+                break;
+            }
+        }
+
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+            *dbgout << json::close << json::close; // phase-1
+#endif
+
+        // phase 2 : loop until happy.
+        for (int i = 0; i < m_numCollRuns - 1; ++i)
+        {
+            if (hasCollisions || moved)
+            {
+
+#if !defined GRAPHITE2_NTRACING
+                if (dbgout)
+                    *dbgout << json::object << "phase" << "2a" << "loop" << i << "moves" << json::array;
+#endif
+                // phase 2a : if any shiftable glyphs are in collision, iterate backwards,
+                // fixing them and ignoring other non-collided glyphs. Note that this handles ONLY
+                // glyphs that are actually in collision from phases 1 or 2b, and working backwards
+                // has the intended effect of breaking logjams.
+                if (hasCollisions)
+                {
+                    hasCollisions = false;
+                    #if 0
+                    moved = true;
+                    for (Slot *s = start; s != end; s = s->next())
+                    {
+                        SlotCollision * c = seg->collisionInfo(s);
+                        c->setShift(Position(0, 0));
+                    }
+                    #endif
+                    Slot *lend = end ? end->prev() : seg->last();
+                    Slot *lstart = start->prev();
+                    for (Slot *s = lend; s != lstart; s = s->prev())
+                    {
+                        SlotCollision * c = seg->collisionInfo(s);
+                        if (start && (c->flags() & (SlotCollision::COLL_FIX | SlotCollision::COLL_KERN | SlotCollision::COLL_ISCOL))
+                                        == (SlotCollision::COLL_FIX | SlotCollision::COLL_ISCOL)) // ONLY if this glyph is still colliding
+                        {
+                            if (!resolveCollisions(seg, s, lend, shiftcoll, true, dir, moved, hasCollisions, dbgout))
+                                return false;
+                            c->setFlags(c->flags() | SlotCollision::COLL_TEMPLOCK);
+                        }
+                    }
+                }
+
+#if !defined GRAPHITE2_NTRACING
+                if (dbgout)
+                    *dbgout << json::close << json::close // phase 2a
+                        << json::object << "phase" << "2b" << "loop" << i << "moves" << json::array;
+#endif
+
+                // phase 2b : redo basic diacritic positioning pass for ALL glyphs. Each successive loop adjusts
+                // glyphs from their current adjusted position, which has the effect of gradually minimizing the
+                // resulting adjustment; ie, the final result will be gradually closer to the original location.
+                // Also it allows more flexibility in the final adjustment, since it is moving along the
+                // possible 8 vectors from successively different starting locations.
+                if (moved)
+                {
+                    moved = false;
+                    for (Slot *s = start; s != end; s = s->next())
+                    {
+                        SlotCollision * c = seg->collisionInfo(s);
+                        if (start && (c->flags() & (SlotCollision::COLL_FIX | SlotCollision::COLL_TEMPLOCK
+                                                        | SlotCollision::COLL_KERN)) == SlotCollision::COLL_FIX
+                                  && !resolveCollisions(seg, s, start, shiftcoll, false, dir, moved, hasCollisions, dbgout))
+                            return false;
+                        else if (c->flags() & SlotCollision::COLL_TEMPLOCK)
+                            c->setFlags(c->flags() & ~SlotCollision::COLL_TEMPLOCK);
+                    }
+                }
+        //      if (!hasCollisions) // no, don't leave yet because phase 2b will continue to improve things
+        //          break;
+#if !defined GRAPHITE2_NTRACING
+                if (dbgout)
+                    *dbgout << json::close << json::close; // phase 2
+#endif
+            }
+        }
+        if (!end)
+            break;
+        start = NULL;
+        for (Slot *s = end->prev(); s; s = s->next())
+        {
+            if (seg->collisionInfo(s)->flags() & SlotCollision::COLL_START)
+            {
+                start = s;
+                break;
+            }
+        }
+    }
+    return true;
+}
+
+bool Pass::collisionKern(Segment *seg, int dir, json * const dbgout) const
+{
+    Slot *start = seg->first();
+    float ymin = 1e38f;
+    float ymax = -1e38f;
+    const GlyphCache &gc = seg->getFace()->glyphs();
+
+    // phase 3 : handle kerning of clusters
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+        *dbgout << json::object << "phase" << "3" << "moves" << json::array;
+#endif
+
+    for (Slot *s = seg->first(); s; s = s->next())
+    {
+        if (!gc.check(s->gid()))
+            return false;
+        const SlotCollision * c = seg->collisionInfo(s);
+        const Rect &bbox = seg->theGlyphBBoxTemporary(s->gid());
+        float y = s->origin().y + c->shift().y;
+        if (!(c->flags() & SlotCollision::COLL_ISSPACE))
+        {
+            ymax = max(y + bbox.tr.y, ymax);
+            ymin = min(y + bbox.bl.y, ymin);
+        }
+        if (start && (c->flags() & (SlotCollision::COLL_KERN | SlotCollision::COLL_FIX))
+                        == (SlotCollision::COLL_KERN | SlotCollision::COLL_FIX))
+            resolveKern(seg, s, start, dir, ymin, ymax, dbgout);
+        if (c->flags() & SlotCollision::COLL_END)
+            start = NULL;
+        if (c->flags() & SlotCollision::COLL_START)
+            start = s;
+    }
+
+#if !defined GRAPHITE2_NTRACING
+    if (dbgout)
+        *dbgout << json::close << json::close; // phase 3
+#endif
+    return true;
+}
+
+bool Pass::collisionFinish(Segment *seg, GR_MAYBE_UNUSED json * const dbgout) const
+{
+    for (Slot *s = seg->first(); s; s = s->next())
+    {
+        SlotCollision *c = seg->collisionInfo(s);
+        if (c->shift().x != 0 || c->shift().y != 0)
+        {
+            const Position newOffset = c->shift();
+            const Position nullPosition(0, 0);
+            c->setOffset(newOffset + c->offset());
+            c->setShift(nullPosition);
+        }
+    }
+//    seg->positionSlots();
+
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+            *dbgout << json::close;
+#endif
+    return true;
+}
+
+// Can slot s be kerned, or is it attached to something that can be kerned?
+static bool inKernCluster(Segment *seg, Slot *s)
+{
+    SlotCollision *c = seg->collisionInfo(s);
+    if (c->flags() & SlotCollision::COLL_KERN /** && c->flags() & SlotCollision::COLL_FIX **/ )
+        return true;
+    while (s->attachedTo())
+    {
+        s = s->attachedTo();
+        c = seg->collisionInfo(s);
+        if (c->flags() & SlotCollision::COLL_KERN /** && c->flags() & SlotCollision::COLL_FIX **/ )
+            return true;
+    }
+    return false;
+}
+
+// Fix collisions for the given slot.
+// Return true if everything was fixed, false if there are still collisions remaining.
+// isRev means be we are processing backwards.
+bool Pass::resolveCollisions(Segment *seg, Slot *slotFix, Slot *start,
+        ShiftCollider &coll, GR_MAYBE_UNUSED bool isRev, int dir, bool &moved, bool &hasCol,
+        json * const dbgout) const
+{
+    Slot * nbor;  // neighboring slot
+    SlotCollision *cFix = seg->collisionInfo(slotFix);
+    if (!coll.initSlot(seg, slotFix, cFix->limit(), cFix->margin(), cFix->marginWt(),
+            cFix->shift(), cFix->offset(), dir, dbgout))
+        return false;
+    bool collides = false;
+    // When we're processing forward, ignore kernable glyphs that preceed the target glyph.
+    // When processing backward, don't ignore these until we pass slotFix.
+    bool ignoreForKern = !isRev;
+    bool rtl = dir & 1;
+    Slot *base = slotFix;
+    while (base->attachedTo())
+        base = base->attachedTo();
+    Position zero(0., 0.);
+
+    // Look for collisions with the neighboring glyphs.
+    for (nbor = start; nbor; nbor = isRev ? nbor->prev() : nbor->next())
+    {
+        SlotCollision *cNbor = seg->collisionInfo(nbor);
+        bool sameCluster = nbor->isChildOf(base);
+        if (nbor != slotFix         						// don't process if this is the slot of interest
+                      && !(cNbor->ignore())    				// don't process if ignoring
+                      && (nbor == base || sameCluster       // process if in the same cluster as slotFix
+                            || !inKernCluster(seg, nbor))   // or this cluster is not to be kerned
+//                            || (rtl ^ ignoreForKern))       // or it comes before(ltr) or after(rtl)
+                      && (!isRev    // if processing forwards then good to merge otherwise only:
+                            || !(cNbor->flags() & SlotCollision::COLL_FIX)     // merge in immovable stuff
+                            || ((cNbor->flags() & SlotCollision::COLL_KERN) && !sameCluster)     // ignore other kernable clusters
+                            || (cNbor->flags() & SlotCollision::COLL_ISCOL))   // test against other collided glyphs
+                      && !coll.mergeSlot(seg, nbor, cNbor, cNbor->shift(), !ignoreForKern, sameCluster, collides, false, dbgout))
+            return false;
+        else if (nbor == slotFix)
+            // Switching sides of this glyph - if we were ignoring kernable stuff before, don't anymore.
+            ignoreForKern = !ignoreForKern;
+
+        if (nbor != start && (cNbor->flags() & (isRev ? SlotCollision::COLL_START : SlotCollision::COLL_END)))
+            break;
+    }
+    bool isCol = false;
+    if (collides || cFix->shift().x != 0.f || cFix->shift().y != 0.f)
+    {
+        Position shift = coll.resolve(seg, isCol, dbgout);
+        // isCol has been set to true if a collision remains.
+        if (std::fabs(shift.x) < 1e38f && std::fabs(shift.y) < 1e38f)
+        {
+            if (sqr(shift.x-cFix->shift().x) + sqr(shift.y-cFix->shift().y) >= m_colThreshold * m_colThreshold)
+                moved = true;
+            cFix->setShift(shift);
+            if (slotFix->firstChild())
+            {
+                Rect bbox;
+                Position here = slotFix->origin() + shift;
+                float clusterMin = here.x;
+                slotFix->firstChild()->finalise(seg, NULL, here, bbox, 0, clusterMin, rtl, false);
+            }
+        }
+    }
+    else
+    {
+        // This glyph is not colliding with anything.
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+        {
+            *dbgout << json::object
+                            << "missed" << objectid(dslot(seg, slotFix));
+            coll.outputJsonDbg(dbgout, seg, -1);
+            *dbgout << json::close;
+        }
+#endif
+    }
+
+    // Set the is-collision flag bit.
+    if (isCol)
+    { cFix->setFlags(cFix->flags() | SlotCollision::COLL_ISCOL | SlotCollision::COLL_KNOWN); }
+    else
+    { cFix->setFlags((cFix->flags() & ~SlotCollision::COLL_ISCOL) | SlotCollision::COLL_KNOWN); }
+    hasCol |= isCol;
+    return true;
+}
+
+float Pass::resolveKern(Segment *seg, Slot *slotFix, GR_MAYBE_UNUSED Slot *start, int dir,
+    float &ymin, float &ymax, json *const dbgout) const
+{
+    Slot *nbor; // neighboring slot
+    float currSpace = 0.;
+    bool collides = false;
+    unsigned int space_count = 0;
+    Slot *base = slotFix;
+    while (base->attachedTo())
+        base = base->attachedTo();
+    SlotCollision *cFix = seg->collisionInfo(base);
+    const GlyphCache &gc = seg->getFace()->glyphs();
+    const Rect &bbb = seg->theGlyphBBoxTemporary(slotFix->gid());
+    const float by = slotFix->origin().y + cFix->shift().y;
+
+    if (base != slotFix)
+    {
+        cFix->setFlags(cFix->flags() | SlotCollision::COLL_KERN | SlotCollision::COLL_FIX);
+        return 0;
+    }
+    bool seenEnd = (cFix->flags() & SlotCollision::COLL_END) != 0;
+    bool isInit = false;
+    KernCollider coll(dbgout);
+
+    ymax = max(by + bbb.tr.y, ymax);
+    ymin = min(by + bbb.bl.y, ymin);
+    for (nbor = slotFix->next(); nbor; nbor = nbor->next())
+    {
+        if (nbor->isChildOf(base))
+            continue;
+        if (!gc.check(nbor->gid()))
+            return 0.;
+        const Rect &bb = seg->theGlyphBBoxTemporary(nbor->gid());
+        SlotCollision *cNbor = seg->collisionInfo(nbor);
+        if ((bb.bl.y == 0.f && bb.tr.y == 0.f) || (cNbor->flags() & SlotCollision::COLL_ISSPACE))
+        {
+            if (m_kernColls == InWord)
+                break;
+            // Add space for a space glyph.
+            currSpace += nbor->advance();
+            ++space_count;
+        }
+        else
+        {
+            space_count = 0;
+            if (nbor != slotFix && !cNbor->ignore())
+            {
+                seenEnd = true;
+                if (!isInit)
+                {
+                    if (!coll.initSlot(seg, slotFix, cFix->limit(), cFix->margin(),
+                                    cFix->shift(), cFix->offset(), dir, ymin, ymax, dbgout))
+                        return 0.;
+                    isInit = true;
+                }
+                collides |= coll.mergeSlot(seg, nbor, cNbor->shift(), currSpace, dir, dbgout);
+            }
+        }
+        if (cNbor->flags() & SlotCollision::COLL_END)
+        {
+            if (seenEnd && space_count < 2)
+                break;
+            else
+                seenEnd = true;
+        }
+    }
+    if (collides)
+    {
+        Position mv = coll.resolve(seg, slotFix, dir, dbgout);
+        coll.shift(mv, dir);
+        Position delta = slotFix->advancePos() + mv - cFix->shift();
+        slotFix->advance(delta);
+        cFix->setShift(mv);
+        return mv.x;
+    }
+    return 0.;
+}
diff --git a/thirdparty/graphite/src/Position.cpp b/thirdparty/graphite/src/Position.cpp
new file mode 100644
index 0000000000..d2fdbd4e7c
--- /dev/null
+++ b/thirdparty/graphite/src/Position.cpp
@@ -0,0 +1,97 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/Position.h"
+#include <cmath>
+
+using namespace graphite2;
+
+bool Rect::hitTest(Rect &other)
+{
+    if (bl.x > other.tr.x) return false;
+    if (tr.x < other.bl.x) return false;
+    if (bl.y > other.tr.y) return false;
+    if (tr.y < other.bl.y) return false;
+    return true;
+}
+
+Position Rect::overlap(Position &offset, Rect &other, Position &othero)
+{
+    float ax = (bl.x + offset.x) - (other.tr.x + othero.x);
+    float ay = (bl.y + offset.y) - (other.tr.y + othero.y);
+    float bx = (other.bl.x + othero.x) - (tr.x + offset.x);
+    float by = (other.bl.y + othero.y) - (tr.y + offset.y);
+    return Position((ax > bx ? ax : bx), (ay > by ? ay : by));
+}
+
+float boundmin(float move, float lim1, float lim2, float &error)
+{
+    // error is always positive for easy comparison
+    if (move < lim1 && move < lim2)
+    { error = 0.; return move; }
+    else if (lim1 < lim2)
+    { error = std::fabs(move - lim1); return lim1; }
+    else
+    { error = std::fabs(move - lim2); return lim2; }
+}
+
+#if 0
+Position Rect::constrainedAvoid(Position &offset, Rect &box, Rect &sdbox, Position &other, Rect &obox, Rect &osdbox)
+{
+    // a = max, i = min, s = sum, d = diff
+    float eax, eay, eix, eiy, eas, eis, ead, eid;
+    float beste = INF;
+    Position res;
+    // calculate the movements in each direction and the error (amount of remaining overlap)
+    // first param is movement, second and third are movement over the constraining box
+    float ax = boundmin(obox.tr.x + other.x - box.bl.x - offset.x + 1, tr.x - offset.x, INF, &eax);
+    float ay = boundmin(obox.tr.y + other.y - box.bl.y - offset.y + 1, tr.y - offset.y, INF, &eay);
+    float ix = boundmin(obox.bl.x + other.x - box.tr.x - offset.x + 1, bl.x - offset.x, INF, &eix);
+    float iy = boundmin(obox.bl.y + other.y - box.tr.y - offset.y + 1, bl.y - offset.y, INF, &eiy);
+    float as = boundmin(ISQRT2 * (osdbox.tr.x + other.x + other.y - sdbox.bl.x - offset.x - offset.y) + 1, tr.x - offset.x, tr.y - offset.y, &eas);
+    float is = boundmin(ISQRT2 * (osdbox.bl.x + other.x + other.y - sdbox.tr.x - offset.x - offset.y) + 1, bl.x - offset.x, bl.y - offset.y, &eis);
+    float ad = boundmin(ISQRT2 * (osdbox.tr.y + other.x - other.y - sdbox.bl.y - offset.x + offset.y) + 1, tr.y - offset.y, tr.x - offset.x, &ead);
+    float id = boundmin(ISQRT2 * (osdbox.bl.y + other.x - other.y - sdbox.tr.y - offset.x + offset.y) + 1, bl.y - offset.y, bl.x - offset.x, &eid);
+
+    if (eax < beste)
+    { res = Position(ax, 0); beste = eax; }
+    if (eay < beste)
+    { res = Position(0, ay); beste = eay; }
+    if (eix < beste)
+    { res = Position(ix, 0); beste = eix; }
+    if (eiy < beste)
+    { res = Position(0, iy); beste = eiy; }
+    if (SQRT2 * (eas) < beste)
+    { res = Position(as, ad); beste = SQRT2 * (eas); }
+    if (SQRT2 * (eis) < beste)
+    { res = Position(is, is); beste = SQRT2 * (eis); }
+    if (SQRT2 * (ead) < beste)
+    { res = Position(ad, ad); beste = SQRT2 * (ead); }
+    if (SQRT2 * (eid) < beste)
+    { res = Position(id, id); beste = SQRT2 * (eid); }
+    return res;
+}
+#endif
diff --git a/thirdparty/graphite/src/Segment.cpp b/thirdparty/graphite/src/Segment.cpp
new file mode 100644
index 0000000000..62edd4250f
--- /dev/null
+++ b/thirdparty/graphite/src/Segment.cpp
@@ -0,0 +1,423 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/UtfCodec.h"
+#include <cstring>
+#include <cstdlib>
+
+#include "inc/bits.h"
+#include "inc/Segment.h"
+#include "graphite2/Font.h"
+#include "inc/CharInfo.h"
+#include "inc/debug.h"
+#include "inc/Slot.h"
+#include "inc/Main.h"
+#include "inc/CmapCache.h"
+#include "inc/Collider.h"
+#include "graphite2/Segment.h"
+
+
+using namespace graphite2;
+
+Segment::Segment(size_t numchars, const Face* face, uint32 script, int textDir)
+: m_freeSlots(NULL),
+  m_freeJustifies(NULL),
+  m_charinfo(new CharInfo[numchars]),
+  m_collisions(NULL),
+  m_face(face),
+  m_silf(face->chooseSilf(script)),
+  m_first(NULL),
+  m_last(NULL),
+  m_bufSize(numchars + 10),
+  m_numGlyphs(numchars),
+  m_numCharinfo(numchars),
+  m_defaultOriginal(0),
+  m_dir(textDir),
+  m_flags(((m_silf->flags() & 0x20) != 0) << 1),
+  m_passBits(m_silf->aPassBits() ? -1 : 0)
+{
+    freeSlot(newSlot());
+    m_bufSize = log_binary(numchars)+1;
+}
+
+Segment::~Segment()
+{
+    for (SlotRope::iterator i = m_slots.begin(); i != m_slots.end(); ++i)
+        free(*i);
+    for (AttributeRope::iterator i = m_userAttrs.begin(); i != m_userAttrs.end(); ++i)
+        free(*i);
+    for (JustifyRope::iterator i = m_justifies.begin(); i != m_justifies.end(); ++i)
+        free(*i);
+    delete[] m_charinfo;
+    free(m_collisions);
+}
+
+void Segment::appendSlot(int id, int cid, int gid, int iFeats, size_t coffset)
+{
+    Slot *aSlot = newSlot();
+
+    if (!aSlot) return;
+    m_charinfo[id].init(cid);
+    m_charinfo[id].feats(iFeats);
+    m_charinfo[id].base(coffset);
+    const GlyphFace * theGlyph = m_face->glyphs().glyphSafe(gid);
+    m_charinfo[id].breakWeight(theGlyph ? theGlyph->attrs()[m_silf->aBreak()] : 0);
+
+    aSlot->child(NULL);
+    aSlot->setGlyph(this, gid, theGlyph);
+    aSlot->originate(id);
+    aSlot->before(id);
+    aSlot->after(id);
+    if (m_last) m_last->next(aSlot);
+    aSlot->prev(m_last);
+    m_last = aSlot;
+    if (!m_first) m_first = aSlot;
+    if (theGlyph && m_silf->aPassBits())
+        m_passBits &= theGlyph->attrs()[m_silf->aPassBits()]
+                    | (m_silf->numPasses() > 16 ? (theGlyph->attrs()[m_silf->aPassBits() + 1] << 16) : 0);
+}
+
+Slot *Segment::newSlot()
+{
+    if (!m_freeSlots)
+    {
+        // check that the segment doesn't grow indefinintely
+        if (m_numGlyphs > m_numCharinfo * MAX_SEG_GROWTH_FACTOR)
+            return NULL;
+        int numUser = m_silf->numUser();
+#if !defined GRAPHITE2_NTRACING
+        if (m_face->logger()) ++numUser;
+#endif
+        Slot *newSlots = grzeroalloc<Slot>(m_bufSize);
+        int16 *newAttrs = grzeroalloc<int16>(m_bufSize * numUser);
+        if (!newSlots || !newAttrs)
+        {
+            free(newSlots);
+            free(newAttrs);
+            return NULL;
+        }
+        for (size_t i = 0; i < m_bufSize; i++)
+        {
+            ::new (newSlots + i) Slot(newAttrs + i * numUser);
+            newSlots[i].next(newSlots + i + 1);
+        }
+        newSlots[m_bufSize - 1].next(NULL);
+        newSlots[0].next(NULL);
+        m_slots.push_back(newSlots);
+        m_userAttrs.push_back(newAttrs);
+        m_freeSlots = (m_bufSize > 1)? newSlots + 1 : NULL;
+        return newSlots;
+    }
+    Slot *res = m_freeSlots;
+    m_freeSlots = m_freeSlots->next();
+    res->next(NULL);
+    return res;
+}
+
+void Segment::freeSlot(Slot *aSlot)
+{
+    if (aSlot == nullptr) return;
+    if (m_last == aSlot) m_last = aSlot->prev();
+    if (m_first == aSlot) m_first = aSlot->next();
+    if (aSlot->attachedTo())
+        aSlot->attachedTo()->removeChild(aSlot);
+    while (aSlot->firstChild())
+    {
+        if (aSlot->firstChild()->attachedTo() == aSlot)
+        {
+            aSlot->firstChild()->attachTo(nullptr);
+            aSlot->removeChild(aSlot->firstChild());
+        }
+        else
+            aSlot->firstChild(nullptr);
+    }
+    // reset the slot incase it is reused
+    ::new (aSlot) Slot(aSlot->userAttrs());
+    memset(aSlot->userAttrs(), 0, m_silf->numUser() * sizeof(int16));
+    // Update generation counter for debug
+#if !defined GRAPHITE2_NTRACING
+    if (m_face->logger())
+        ++aSlot->userAttrs()[m_silf->numUser()];
+#endif
+    // update next pointer
+    if (!m_freeSlots)
+        aSlot->next(nullptr);
+    else
+        aSlot->next(m_freeSlots);
+    m_freeSlots = aSlot;
+}
+
+SlotJustify *Segment::newJustify()
+{
+    if (!m_freeJustifies)
+    {
+        const size_t justSize = SlotJustify::size_of(m_silf->numJustLevels());
+        byte *justs = grzeroalloc<byte>(justSize * m_bufSize);
+        if (!justs) return NULL;
+        for (ptrdiff_t i = m_bufSize - 2; i >= 0; --i)
+        {
+            SlotJustify *p = reinterpret_cast<SlotJustify *>(justs + justSize * i);
+            SlotJustify *next = reinterpret_cast<SlotJustify *>(justs + justSize * (i + 1));
+            p->next = next;
+        }
+        m_freeJustifies = (SlotJustify *)justs;
+        m_justifies.push_back(m_freeJustifies);
+    }
+    SlotJustify *res = m_freeJustifies;
+    m_freeJustifies = m_freeJustifies->next;
+    res->next = NULL;
+    return res;
+}
+
+void Segment::freeJustify(SlotJustify *aJustify)
+{
+    int numJust = m_silf->numJustLevels();
+    if (m_silf->numJustLevels() <= 0) numJust = 1;
+    aJustify->next = m_freeJustifies;
+    memset(aJustify->values, 0, numJust*SlotJustify::NUMJUSTPARAMS*sizeof(int16));
+    m_freeJustifies = aJustify;
+}
+
+// reverse the slots but keep diacritics in their same position after their bases
+void Segment::reverseSlots()
+{
+    m_dir = m_dir ^ 64;                 // invert the reverse flag
+    if (m_first == m_last) return;      // skip 0 or 1 glyph runs
+
+    Slot *t = 0;
+    Slot *curr = m_first;
+    Slot *tlast;
+    Slot *tfirst;
+    Slot *out = 0;
+
+    while (curr && getSlotBidiClass(curr) == 16)
+        curr = curr->next();
+    if (!curr) return;
+    tfirst = curr->prev();
+    tlast = curr;
+
+    while (curr)
+    {
+        if (getSlotBidiClass(curr) == 16)
+        {
+            Slot *d = curr->next();
+            while (d && getSlotBidiClass(d) == 16)
+                d = d->next();
+
+            d = d ? d->prev() : m_last;
+            Slot *p = out->next();    // one after the diacritics. out can't be null
+            if (p)
+                p->prev(d);
+            else
+                tlast = d;
+            t = d->next();
+            d->next(p);
+            curr->prev(out);
+            out->next(curr);
+        }
+        else    // will always fire first time round the loop
+        {
+            if (out)
+                out->prev(curr);
+            t = curr->next();
+            curr->next(out);
+            out = curr;
+        }
+        curr = t;
+    }
+    out->prev(tfirst);
+    if (tfirst)
+        tfirst->next(out);
+    else
+        m_first = out;
+    m_last = tlast;
+}
+
+void Segment::linkClusters(Slot *s, Slot * end)
+{
+    end = end->next();
+
+    for (; s != end && !s->isBase(); s = s->next());
+    Slot * ls = s;
+
+    if (m_dir & 1)
+    {
+        for (; s != end; s = s->next())
+        {
+            if (!s->isBase())   continue;
+
+            s->sibling(ls);
+            ls = s;
+        }
+    }
+    else
+    {
+        for (; s != end; s = s->next())
+        {
+            if (!s->isBase())   continue;
+
+            ls->sibling(s);
+            ls = s;
+        }
+    }
+}
+
+Position Segment::positionSlots(const Font *font, Slot * iStart, Slot * iEnd, bool isRtl, bool isFinal)
+{
+    Position currpos(0., 0.);
+    float clusterMin = 0.;
+    Rect bbox;
+    bool reorder = (currdir() != isRtl);
+
+    if (reorder)
+    {
+        Slot *temp;
+        reverseSlots();
+        temp = iStart;
+        iStart = iEnd;
+        iEnd = temp;
+    }
+    if (!iStart)    iStart = m_first;
+    if (!iEnd)      iEnd   = m_last;
+
+    if (!iStart || !iEnd)   // only true for empty segments
+        return currpos;
+
+    if (isRtl)
+    {
+        for (Slot * s = iEnd, * const end = iStart->prev(); s && s != end; s = s->prev())
+        {
+            if (s->isBase())
+                currpos = s->finalise(this, font, currpos, bbox, 0, clusterMin = currpos.x, isRtl, isFinal);
+        }
+    }
+    else
+    {
+        for (Slot * s = iStart, * const end = iEnd->next(); s && s != end; s = s->next())
+        {
+            if (s->isBase())
+                currpos = s->finalise(this, font, currpos, bbox, 0, clusterMin = currpos.x, isRtl, isFinal);
+        }
+    }
+    if (reorder)
+        reverseSlots();
+    return currpos;
+}
+
+
+void Segment::associateChars(int offset, size_t numChars)
+{
+    int i = 0, j = 0;
+    CharInfo *c, *cend;
+    for (c = m_charinfo + offset, cend = m_charinfo + offset + numChars; c != cend; ++c)
+    {
+        c->before(-1);
+        c->after(-1);
+    }
+    for (Slot * s = m_first; s; s->index(i++), s = s->next())
+    {
+        j = s->before();
+        if (j < 0)  continue;
+
+        for (const int after = s->after(); j <= after; ++j)
+        {
+            c = charinfo(j);
+            if (c->before() == -1 || i < c->before())   c->before(i);
+            if (c->after() < i)                         c->after(i);
+        }
+    }
+    for (Slot *s = m_first; s; s = s->next())
+    {
+        int a;
+        for (a = s->after() + 1; a < offset + int(numChars) && charinfo(a)->after() < 0; ++a)
+        { charinfo(a)->after(s->index()); }
+        --a;
+        s->after(a);
+
+        for (a = s->before() - 1; a >= offset && charinfo(a)->before() < 0; --a)
+        { charinfo(a)->before(s->index()); }
+        ++a;
+        s->before(a);
+    }
+}
+
+
+template <typename utf_iter>
+inline void process_utf_data(Segment & seg, const Face & face, const int fid, utf_iter c, size_t n_chars)
+{
+    const Cmap    & cmap = face.cmap();
+    int slotid = 0;
+
+    const typename utf_iter::codeunit_type * const base = c;
+    for (; n_chars; --n_chars, ++c, ++slotid)
+    {
+        const uint32 usv = *c;
+        uint16 gid = cmap[usv];
+        if (!gid)   gid = face.findPseudo(usv);
+        seg.appendSlot(slotid, usv, gid, fid, c - base);
+    }
+}
+
+
+bool Segment::read_text(const Face *face, const Features* pFeats/*must not be NULL*/, gr_encform enc, const void* pStart, size_t nChars)
+{
+    assert(face);
+    assert(pFeats);
+    if (!m_charinfo) return false;
+
+    // utf iterator is self recovering so we don't care about the error state of the iterator.
+    switch (enc)
+    {
+    case gr_utf8:   process_utf_data(*this, *face, addFeatures(*pFeats), utf8::const_iterator(pStart), nChars); break;
+    case gr_utf16:  process_utf_data(*this, *face, addFeatures(*pFeats), utf16::const_iterator(pStart), nChars); break;
+    case gr_utf32:  process_utf_data(*this, *face, addFeatures(*pFeats), utf32::const_iterator(pStart), nChars); break;
+    }
+    return true;
+}
+
+void Segment::doMirror(uint16 aMirror)
+{
+    Slot * s;
+    for (s = m_first; s; s = s->next())
+    {
+        unsigned short g = glyphAttr(s->gid(), aMirror);
+        if (g && (!(dir() & 4) || !glyphAttr(s->gid(), aMirror + 1)))
+            s->setGlyph(this, g);
+    }
+}
+
+bool Segment::initCollisions()
+{
+    m_collisions = grzeroalloc<SlotCollision>(slotCount());
+    if (!m_collisions) return false;
+
+    for (Slot *p = m_first; p; p = p->next())
+        if (p->index() < slotCount())
+            ::new (collisionInfo(p)) SlotCollision(this, p);
+        else
+            return false;
+    return true;
+}
diff --git a/thirdparty/graphite/src/Silf.cpp b/thirdparty/graphite/src/Silf.cpp
new file mode 100644
index 0000000000..44d3c96171
--- /dev/null
+++ b/thirdparty/graphite/src/Silf.cpp
@@ -0,0 +1,439 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cstdlib>
+#include "graphite2/Segment.h"
+#include "inc/debug.h"
+#include "inc/Endian.h"
+#include "inc/Silf.h"
+#include "inc/Segment.h"
+#include "inc/Rule.h"
+#include "inc/Error.h"
+
+
+using namespace graphite2;
+
+namespace { static const uint32 ERROROFFSET = 0xFFFFFFFF; }
+
+Silf::Silf() throw()
+: m_passes(0),
+  m_pseudos(0),
+  m_classOffsets(0),
+  m_classData(0),
+  m_justs(0),
+  m_numPasses(0),
+  m_numJusts(0),
+  m_sPass(0),
+  m_pPass(0),
+  m_jPass(0),
+  m_bPass(0),
+  m_flags(0),
+  m_dir(0),
+  m_aPseudo(0),
+  m_aBreak(0),
+  m_aUser(0),
+  m_aBidi(0),
+  m_aMirror(0),
+  m_aPassBits(0),
+  m_iMaxComp(0),
+  m_aCollision(0),
+  m_aLig(0),
+  m_numPseudo(0),
+  m_nClass(0),
+  m_nLinear(0),
+  m_gEndLine(0)
+{
+    memset(&m_silfinfo, 0, sizeof m_silfinfo);
+}
+
+Silf::~Silf() throw()
+{
+    releaseBuffers();
+}
+
+void Silf::releaseBuffers() throw()
+{
+    delete [] m_passes;
+    delete [] m_pseudos;
+    free(m_classOffsets);
+    free(m_classData);
+    free(m_justs);
+    m_passes= 0;
+    m_pseudos = 0;
+    m_classOffsets = 0;
+    m_classData = 0;
+    m_justs = 0;
+}
+
+
+bool Silf::readGraphite(const byte * const silf_start, size_t lSilf, Face& face, uint32 version)
+{
+    const byte * p = silf_start,
+               * const silf_end = p + lSilf;
+    Error e;
+
+    if (e.test(version >= 0x00060000, E_BADSILFVERSION))
+    {
+        releaseBuffers(); return face.error(e);
+    }
+    if (version >= 0x00030000)
+    {
+        if (e.test(lSilf < 28, E_BADSIZE)) { releaseBuffers(); return face.error(e); }
+        be::skip<int32>(p);    // ruleVersion
+        be::skip<uint16>(p,2); // passOffset & pseudosOffset
+    }
+    else if (e.test(lSilf < 20, E_BADSIZE)) { releaseBuffers(); return face.error(e); }
+    const uint16 maxGlyph = be::read<uint16>(p);
+    m_silfinfo.extra_ascent = be::read<uint16>(p);
+    m_silfinfo.extra_descent = be::read<uint16>(p);
+    m_numPasses = be::read<uint8>(p);
+    m_sPass     = be::read<uint8>(p);
+    m_pPass     = be::read<uint8>(p);
+    m_jPass     = be::read<uint8>(p);
+    m_bPass     = be::read<uint8>(p);
+    m_flags     = be::read<uint8>(p);
+    be::skip<uint8>(p,2); //  max{Pre,Post}Context.
+    m_aPseudo   = be::read<uint8>(p);
+    m_aBreak    = be::read<uint8>(p);
+    m_aBidi     = be::read<uint8>(p);
+    m_aMirror   = be::read<uint8>(p);
+    m_aPassBits = be::read<uint8>(p);
+
+    // Read Justification levels.
+    m_numJusts  = be::read<uint8>(p);
+    if (e.test(maxGlyph >= face.glyphs().numGlyphs(), E_BADMAXGLYPH)
+        || e.test(p + m_numJusts * 8 >= silf_end, E_BADNUMJUSTS))
+    {
+        releaseBuffers(); return face.error(e);
+    }
+
+    if (m_numJusts)
+    {
+        m_justs = gralloc<Justinfo>(m_numJusts);
+        if (e.test(!m_justs, E_OUTOFMEM)) return face.error(e);
+
+        for (uint8 i = 0; i < m_numJusts; i++)
+        {
+            ::new(m_justs + i) Justinfo(p[0], p[1], p[2], p[3]);
+            be::skip<byte>(p,8);
+        }
+    }
+
+    if (e.test(p + sizeof(uint16) + sizeof(uint8)*8 >= silf_end, E_BADENDJUSTS)) { releaseBuffers(); return face.error(e); }
+    m_aLig       = be::read<uint16>(p);
+    m_aUser      = be::read<uint8>(p);
+    m_iMaxComp   = be::read<uint8>(p);
+    m_dir        = be::read<uint8>(p) - 1;
+    m_aCollision = be::read<uint8>(p);
+    be::skip<byte>(p,3);
+    be::skip<uint16>(p, be::read<uint8>(p));    // don't need critical features yet
+    be::skip<byte>(p);                          // reserved
+    if (e.test(p >= silf_end, E_BADCRITFEATURES))   { releaseBuffers(); return face.error(e); }
+    be::skip<uint32>(p, be::read<uint8>(p));    // don't use scriptTag array.
+    if (e.test(p + sizeof(uint16) + sizeof(uint32) >= silf_end, E_BADSCRIPTTAGS)) { releaseBuffers(); return face.error(e); }
+    m_gEndLine  = be::read<uint16>(p);          // lbGID
+    const byte * o_passes = p;
+    uint32 passes_start = be::read<uint32>(p);
+
+    const size_t num_attrs = face.glyphs().numAttrs();
+    if (e.test(m_aPseudo   >= num_attrs, E_BADAPSEUDO)
+        || e.test(m_aBreak >= num_attrs, E_BADABREAK)
+        || e.test(m_aBidi  >= num_attrs, E_BADABIDI)
+        || e.test(m_aMirror>= num_attrs, E_BADAMIRROR)
+        || e.test(m_aCollision && m_aCollision >= num_attrs - 5, E_BADACOLLISION)
+        || e.test(m_numPasses > 128, E_BADNUMPASSES) || e.test(passes_start >= lSilf, E_BADPASSESSTART)
+        || e.test(m_pPass < m_sPass, E_BADPASSBOUND) || e.test(m_pPass > m_numPasses, E_BADPPASS) || e.test(m_sPass > m_numPasses, E_BADSPASS)
+        || e.test(m_jPass < m_pPass, E_BADJPASSBOUND) || e.test(m_jPass > m_numPasses, E_BADJPASS)
+        || e.test((m_bPass != 0xFF && (m_bPass < m_jPass || m_bPass > m_numPasses)), E_BADBPASS)
+        || e.test(m_aLig > 127, E_BADALIG))
+    {
+        releaseBuffers();
+        return face.error(e);
+    }
+    be::skip<uint32>(p, m_numPasses);
+    if (e.test(unsigned(p - silf_start) + sizeof(uint16) >= passes_start, E_BADPASSESSTART)) { releaseBuffers(); return face.error(e); }
+    m_numPseudo = be::read<uint16>(p);
+    be::skip<uint16>(p, 3); // searchPseudo, pseudoSelector, pseudoShift
+    m_pseudos = new Pseudo[m_numPseudo];
+    if (e.test(unsigned(p - silf_start) + m_numPseudo*(sizeof(uint32) + sizeof(uint16)) >= passes_start, E_BADNUMPSEUDO)
+        || e.test(!m_pseudos, E_OUTOFMEM))
+    {
+        releaseBuffers(); return face.error(e);
+    }
+    for (int i = 0; i < m_numPseudo; i++)
+    {
+        m_pseudos[i].uid = be::read<uint32>(p);
+        m_pseudos[i].gid = be::read<uint16>(p);
+    }
+
+    const size_t clen = readClassMap(p, passes_start + silf_start - p, version, e);
+    m_passes = new Pass[m_numPasses];
+    if (e || e.test(clen > unsigned(passes_start + silf_start - p), E_BADPASSESSTART)
+          || e.test(!m_passes, E_OUTOFMEM))
+    { releaseBuffers(); return face.error(e); }
+
+    for (size_t i = 0; i < m_numPasses; ++i)
+    {
+        uint32 pass_start = be::read<uint32>(o_passes);
+        uint32 pass_end = be::peek<uint32>(o_passes);
+        face.error_context((face.error_context() & 0xFF00) + EC_ASILF + unsigned(i << 16));
+        if (e.test(pass_start > pass_end, E_BADPASSSTART)
+                || e.test(pass_start < passes_start, E_BADPASSSTART)
+                || e.test(pass_end > lSilf, E_BADPASSEND)) {
+            releaseBuffers(); return face.error(e);
+        }
+
+        enum passtype pt = PASS_TYPE_UNKNOWN;
+        if (i >= m_jPass) pt = PASS_TYPE_JUSTIFICATION;
+        else if (i >= m_pPass) pt = PASS_TYPE_POSITIONING;
+        else if (i >= m_sPass) pt = PASS_TYPE_SUBSTITUTE;
+        else pt = PASS_TYPE_LINEBREAK;
+
+        m_passes[i].init(this);
+        if (!m_passes[i].readPass(silf_start + pass_start, pass_end - pass_start, pass_start, face, pt,
+            version, e))
+        {
+            releaseBuffers();
+            return false;
+        }
+    }
+
+    // fill in gr_faceinfo
+    m_silfinfo.upem = face.glyphs().unitsPerEm();
+    m_silfinfo.has_bidi_pass = (m_bPass != 0xFF);
+    m_silfinfo.justifies = (m_numJusts != 0) || (m_jPass < m_pPass);
+    m_silfinfo.line_ends = (m_flags & 1);
+    m_silfinfo.space_contextuals = gr_faceinfo::gr_space_contextuals((m_flags >> 2) & 0x7);
+    return true;
+}
+
+template<typename T> inline uint32 Silf::readClassOffsets(const byte *&p, size_t data_len, Error &e)
+{
+    const T cls_off = 2*sizeof(uint16) + sizeof(T)*(m_nClass+1);
+    const uint32 max_off = (be::peek<T>(p + sizeof(T)*m_nClass) - cls_off)/sizeof(uint16);
+    // Check that the last+1 offset is less than or equal to the class map length.
+    if (e.test(be::peek<T>(p) != cls_off, E_MISALIGNEDCLASSES)
+            || e.test(max_off > (data_len - cls_off)/sizeof(uint16), E_HIGHCLASSOFFSET))
+        return ERROROFFSET;
+
+    // Read in all the offsets.
+    m_classOffsets = gralloc<uint32>(m_nClass+1);
+    if (e.test(!m_classOffsets, E_OUTOFMEM)) return ERROROFFSET;
+    for (uint32 * o = m_classOffsets, * const o_end = o + m_nClass + 1; o != o_end; ++o)
+    {
+        *o = (be::read<T>(p) - cls_off)/sizeof(uint16);
+        if (e.test(*o > max_off, E_HIGHCLASSOFFSET))
+            return ERROROFFSET;
+    }
+    return max_off;
+}
+
+size_t Silf::readClassMap(const byte *p, size_t data_len, uint32 version, Error &e)
+{
+    if (e.test(data_len < sizeof(uint16)*2, E_BADCLASSSIZE)) return ERROROFFSET;
+
+    m_nClass  = be::read<uint16>(p);
+    m_nLinear = be::read<uint16>(p);
+
+    // Check that numLinear < numClass,
+    // that there is at least enough data for numClasses offsets.
+    if (e.test(m_nLinear > m_nClass, E_TOOMANYLINEAR)
+     || e.test((m_nClass + 1) * (version >= 0x00040000 ? sizeof(uint32) : sizeof(uint16)) > (data_len - 4), E_CLASSESTOOBIG))
+        return ERROROFFSET;
+
+    uint32 max_off;
+    if (version >= 0x00040000)
+        max_off = readClassOffsets<uint32>(p, data_len, e);
+    else
+        max_off = readClassOffsets<uint16>(p, data_len, e);
+
+    if (max_off == ERROROFFSET) return ERROROFFSET;
+
+    if (e.test((int)max_off < m_nLinear + (m_nClass - m_nLinear) * 6, E_CLASSESTOOBIG))
+        return ERROROFFSET;
+
+    // Check the linear offsets are sane, these must be monotonically increasing.
+    assert(m_nClass >= m_nLinear);
+    for (const uint32 *o = m_classOffsets, * const o_end = o + m_nLinear; o != o_end; ++o)
+        if (e.test(o[0] > o[1], E_BADCLASSOFFSET))
+            return ERROROFFSET;
+
+    // Fortunately the class data is all uint16s so we can decode these now
+    m_classData = gralloc<uint16>(max_off);
+    if (e.test(!m_classData, E_OUTOFMEM)) return ERROROFFSET;
+    for (uint16 *d = m_classData, * const d_end = d + max_off; d != d_end; ++d)
+        *d = be::read<uint16>(p);
+
+    // Check the lookup class invariants for each non-linear class
+    for (const uint32 *o = m_classOffsets + m_nLinear, * const o_end = m_classOffsets + m_nClass; o != o_end; ++o)
+    {
+        const uint16 * lookup = m_classData + *o;
+        if (e.test(*o + 4 > max_off, E_HIGHCLASSOFFSET)                        // LookupClass doesn't stretch over max_off
+         || e.test(lookup[0] == 0                                                   // A LookupClass with no looks is a suspicious thing ...
+                    || lookup[0] * 2 + *o + 4 > max_off                             // numIDs lookup pairs fits within (start of LookupClass' lookups array, max_off]
+                    || lookup[3] + lookup[1] != lookup[0], E_BADCLASSLOOKUPINFO)    // rangeShift:   numIDs  - searchRange
+         || e.test(((o[1] - *o) & 1) != 0, ERROROFFSET))                         // glyphs are in pairs so difference must be even.
+            return ERROROFFSET;
+    }
+
+    return max_off;
+}
+
+uint16 Silf::findPseudo(uint32 uid) const
+{
+    for (int i = 0; i < m_numPseudo; i++)
+        if (m_pseudos[i].uid == uid) return m_pseudos[i].gid;
+    return 0;
+}
+
+uint16 Silf::findClassIndex(uint16 cid, uint16 gid) const
+{
+    if (cid > m_nClass) return -1;
+
+    const uint16 * cls = m_classData + m_classOffsets[cid];
+    if (cid < m_nLinear)        // output class being used for input, shouldn't happen
+    {
+        for (unsigned int i = 0, n = m_classOffsets[cid + 1] - m_classOffsets[cid]; i < n; ++i, ++cls)
+            if (*cls == gid) return i;
+        return -1;
+    }
+    else
+    {
+        const uint16 *  min = cls + 4,      // lookups array
+                     *  max = min + cls[0]*2; // lookups aray is numIDs (cls[0]) uint16 pairs long
+        do
+        {
+            const uint16 * p = min + (-2 & ((max-min)/2));
+            if  (p[0] > gid)    max = p;
+            else                min = p;
+        }
+        while (max - min > 2);
+        return min[0] == gid ? min[1] : -1;
+    }
+}
+
+uint16 Silf::getClassGlyph(uint16 cid, unsigned int index) const
+{
+    if (cid > m_nClass) return 0;
+
+    uint32 loc = m_classOffsets[cid];
+    if (cid < m_nLinear)
+    {
+        if (index < m_classOffsets[cid + 1] - loc)
+            return m_classData[index + loc];
+    }
+    else        // input class being used for output. Shouldn't happen
+    {
+        for (unsigned int i = loc + 4; i < m_classOffsets[cid + 1]; i += 2)
+            if (m_classData[i + 1] == index) return m_classData[i];
+    }
+    return 0;
+}
+
+
+bool Silf::runGraphite(Segment *seg, uint8 firstPass, uint8 lastPass, int dobidi) const
+{
+    assert(seg != 0);
+    size_t             maxSize = seg->slotCount() * MAX_SEG_GROWTH_FACTOR;
+    SlotMap            map(*seg, m_dir, maxSize);
+    FiniteStateMachine fsm(map, seg->getFace()->logger());
+    vm::Machine        m(map);
+    uint8              lbidi = m_bPass;
+#if !defined GRAPHITE2_NTRACING
+    json * const dbgout = seg->getFace()->logger();
+#endif
+
+    if (lastPass == 0)
+    {
+        if (firstPass == lastPass && lbidi == 0xFF)
+            return true;
+        lastPass = m_numPasses;
+    }
+    if ((firstPass < lbidi || (dobidi && firstPass == lbidi)) && (lastPass >= lbidi || (dobidi && lastPass + 1 == lbidi)))
+        lastPass++;
+    else
+        lbidi = 0xFF;
+
+    for (size_t i = firstPass; i < lastPass; ++i)
+    {
+        // bidi and mirroring
+        if (i == lbidi)
+        {
+#if !defined GRAPHITE2_NTRACING
+            if (dbgout)
+            {
+                *dbgout << json::item << json::object
+//							<< "pindex" << i   // for debugging
+                            << "id"     << -1
+                            << "slotsdir" << (seg->currdir() ? "rtl" : "ltr")
+                            << "passdir" << (m_dir & 1 ? "rtl" : "ltr")
+                            << "slots"  << json::array;
+                seg->positionSlots(0, 0, 0, seg->currdir());
+                for(Slot * s = seg->first(); s; s = s->next())
+                    *dbgout     << dslot(seg, s);
+                *dbgout         << json::close
+                            << "rules"  << json::array << json::close
+                            << json::close;
+            }
+#endif
+            if (seg->currdir() != (m_dir & 1))
+                seg->reverseSlots();
+            if (m_aMirror && (seg->dir() & 3) == 3)
+                seg->doMirror(m_aMirror);
+        --i;
+        lbidi = lastPass;
+        --lastPass;
+        continue;
+        }
+
+#if !defined GRAPHITE2_NTRACING
+        if (dbgout)
+        {
+            *dbgout << json::item << json::object
+//						<< "pindex" << i   // for debugging
+                        << "id"     << i+1
+                        << "slotsdir" << (seg->currdir() ? "rtl" : "ltr")
+                        << "passdir" << ((m_dir & 1) ^ m_passes[i].reverseDir() ? "rtl" : "ltr")
+                        << "slots"  << json::array;
+            seg->positionSlots(0, 0, 0, seg->currdir());
+            for(Slot * s = seg->first(); s; s = s->next())
+                *dbgout     << dslot(seg, s);
+            *dbgout         << json::close;
+        }
+#endif
+
+        // test whether to reorder, prepare for positioning
+        bool reverse = (lbidi == 0xFF) && (seg->currdir() != ((m_dir & 1) ^ m_passes[i].reverseDir()));
+        if ((i >= 32 || (seg->passBits() & (1 << i)) == 0 || m_passes[i].collisionLoops())
+                && !m_passes[i].runGraphite(m, fsm, reverse))
+            return false;
+        // only subsitution passes can change segment length, cached subsegments are short for their text
+        if (m.status() != vm::Machine::finished
+            || (seg->slotCount() && seg->slotCount() > maxSize))
+            return false;
+    }
+    return true;
+}
diff --git a/thirdparty/graphite/src/Slot.cpp b/thirdparty/graphite/src/Slot.cpp
new file mode 100644
index 0000000000..0fdb098952
--- /dev/null
+++ b/thirdparty/graphite/src/Slot.cpp
@@ -0,0 +1,529 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/Silf.h"
+#include "inc/CharInfo.h"
+#include "inc/Rule.h"
+#include "inc/Collider.h"
+
+
+using namespace graphite2;
+
+Slot::Slot(int16 *user_attrs) :
+    m_next(NULL), m_prev(NULL),
+    m_glyphid(0), m_realglyphid(0), m_original(0), m_before(0), m_after(0),
+    m_index(0), m_parent(NULL), m_child(NULL), m_sibling(NULL),
+    m_position(0, 0), m_shift(0, 0), m_advance(0, 0),
+    m_attach(0, 0), m_with(0, 0), m_just(0.),
+    m_flags(0), m_attLevel(0), m_bidiCls(-1), m_bidiLevel(0),
+    m_userAttr(user_attrs), m_justs(NULL)
+{
+}
+
+// take care, this does not copy any of the GrSlot pointer fields
+void Slot::set(const Slot & orig, int charOffset, size_t sizeAttr, size_t justLevels, size_t numChars)
+{
+    // leave m_next and m_prev unchanged
+    m_glyphid = orig.m_glyphid;
+    m_realglyphid = orig.m_realglyphid;
+    m_original = orig.m_original + charOffset;
+    if (charOffset + int(orig.m_before) < 0)
+        m_before = 0;
+    else
+        m_before = orig.m_before + charOffset;
+    if (charOffset <= 0 && orig.m_after + charOffset >= numChars)
+        m_after = int(numChars) - 1;
+    else
+        m_after = orig.m_after + charOffset;
+    m_parent = NULL;
+    m_child = NULL;
+    m_sibling = NULL;
+    m_position = orig.m_position;
+    m_shift = orig.m_shift;
+    m_advance = orig.m_advance;
+    m_attach = orig.m_attach;
+    m_with = orig.m_with;
+    m_flags = orig.m_flags;
+    m_attLevel = orig.m_attLevel;
+    m_bidiCls = orig.m_bidiCls;
+    m_bidiLevel = orig.m_bidiLevel;
+    if (m_userAttr && orig.m_userAttr)
+        memcpy(m_userAttr, orig.m_userAttr, sizeAttr * sizeof(*m_userAttr));
+    if (m_justs && orig.m_justs)
+        memcpy(m_justs, orig.m_justs, SlotJustify::size_of(justLevels));
+}
+
+void Slot::update(int /*numGrSlots*/, int numCharInfo, Position &relpos)
+{
+    m_before += numCharInfo;
+    m_after += numCharInfo;
+    m_position = m_position + relpos;
+}
+
+Position Slot::finalise(const Segment *seg, const Font *font, Position & base, Rect & bbox, uint8 attrLevel, float & clusterMin, bool rtl, bool isFinal, int depth)
+{
+    SlotCollision *coll = NULL;
+    if (depth > 100 || (attrLevel && m_attLevel > attrLevel)) return Position(0, 0);
+    float scale = font ? font->scale() : 1.0f;
+    Position shift(m_shift.x * (rtl * -2 + 1) + m_just, m_shift.y);
+    float tAdvance = m_advance.x + m_just;
+    if (isFinal && (coll = seg->collisionInfo(this)))
+    {
+        const Position &collshift = coll->offset();
+        if (!(coll->flags() & SlotCollision::COLL_KERN) || rtl)
+            shift = shift + collshift;
+    }
+    const GlyphFace * glyphFace = seg->getFace()->glyphs().glyphSafe(glyph());
+    if (font)
+    {
+        scale = font->scale();
+        shift *= scale;
+        if (font->isHinted() && glyphFace)
+            tAdvance = (m_advance.x - glyphFace->theAdvance().x + m_just) * scale + font->advance(glyph());
+        else
+            tAdvance *= scale;
+    }
+    Position res;
+
+    m_position = base + shift;
+    if (!m_parent)
+    {
+        res = base + Position(tAdvance, m_advance.y * scale);
+        clusterMin = m_position.x;
+    }
+    else
+    {
+        float tAdv;
+        m_position += (m_attach - m_with) * scale;
+        tAdv = m_advance.x >= 0.5f ? m_position.x + tAdvance - shift.x : 0.f;
+        res = Position(tAdv, 0);
+        if ((m_advance.x >= 0.5f || m_position.x < 0) && m_position.x < clusterMin) clusterMin = m_position.x;
+    }
+
+    if (glyphFace)
+    {
+        Rect ourBbox = glyphFace->theBBox() * scale + m_position;
+        bbox = bbox.widen(ourBbox);
+    }
+
+    if (m_child && m_child != this && m_child->attachedTo() == this)
+    {
+        Position tRes = m_child->finalise(seg, font, m_position, bbox, attrLevel, clusterMin, rtl, isFinal, depth + 1);
+        if ((!m_parent || m_advance.x >= 0.5f) && tRes.x > res.x) res = tRes;
+    }
+
+    if (m_parent && m_sibling && m_sibling != this && m_sibling->attachedTo() == m_parent)
+    {
+        Position tRes = m_sibling->finalise(seg, font, base, bbox, attrLevel, clusterMin, rtl, isFinal, depth + 1);
+        if (tRes.x > res.x) res = tRes;
+    }
+
+    if (!m_parent && clusterMin < base.x)
+    {
+        Position adj = Position(m_position.x - clusterMin, 0.);
+        res += adj;
+        m_position += adj;
+        if (m_child) m_child->floodShift(adj);
+    }
+    return res;
+}
+
+int32 Slot::clusterMetric(const Segment *seg, uint8 metric, uint8 attrLevel, bool rtl)
+{
+    Position base;
+    if (glyph() >= seg->getFace()->glyphs().numGlyphs())
+        return 0;
+    Rect bbox = seg->theGlyphBBoxTemporary(glyph());
+    float clusterMin = 0.;
+    Position res = finalise(seg, NULL, base, bbox, attrLevel, clusterMin, rtl, false);
+
+    switch (metrics(metric))
+    {
+    case kgmetLsb :
+        return int32(bbox.bl.x);
+    case kgmetRsb :
+        return int32(res.x - bbox.tr.x);
+    case kgmetBbTop :
+        return int32(bbox.tr.y);
+    case kgmetBbBottom :
+        return int32(bbox.bl.y);
+    case kgmetBbLeft :
+        return int32(bbox.bl.x);
+    case kgmetBbRight :
+        return int32(bbox.tr.x);
+    case kgmetBbWidth :
+        return int32(bbox.tr.x - bbox.bl.x);
+    case kgmetBbHeight :
+        return int32(bbox.tr.y - bbox.bl.y);
+    case kgmetAdvWidth :
+        return int32(res.x);
+    case kgmetAdvHeight :
+        return int32(res.y);
+    default :
+        return 0;
+    }
+}
+
+#define SLOTGETCOLATTR(x) { SlotCollision *c = seg->collisionInfo(this); return c ? int(c-> x) : 0; }
+
+int Slot::getAttr(const Segment *seg, attrCode ind, uint8 subindex) const
+{
+    if (ind >= gr_slatJStretch && ind < gr_slatJStretch + 20 && ind != gr_slatJWidth)
+    {
+        int indx = ind - gr_slatJStretch;
+        return getJustify(seg, indx / 5, indx % 5);
+    }
+
+    switch (ind)
+    {
+    case gr_slatAdvX :      return int(m_advance.x);
+    case gr_slatAdvY :      return int(m_advance.y);
+    case gr_slatAttTo :     return m_parent ? 1 : 0;
+    case gr_slatAttX :      return int(m_attach.x);
+    case gr_slatAttY :      return int(m_attach.y);
+    case gr_slatAttXOff :
+    case gr_slatAttYOff :   return 0;
+    case gr_slatAttWithX :  return int(m_with.x);
+    case gr_slatAttWithY :  return int(m_with.y);
+    case gr_slatAttWithXOff:
+    case gr_slatAttWithYOff:return 0;
+    case gr_slatAttLevel :  return m_attLevel;
+    case gr_slatBreak :     return seg->charinfo(m_original)->breakWeight();
+    case gr_slatCompRef :   return 0;
+    case gr_slatDir :       return seg->dir() & 1;
+    case gr_slatInsert :    return isInsertBefore();
+    case gr_slatPosX :      return int(m_position.x); // but need to calculate it
+    case gr_slatPosY :      return int(m_position.y);
+    case gr_slatShiftX :    return int(m_shift.x);
+    case gr_slatShiftY :    return int(m_shift.y);
+    case gr_slatMeasureSol: return -1; // err what's this?
+    case gr_slatMeasureEol: return -1;
+    case gr_slatJWidth:     return int(m_just);
+    case gr_slatUserDefnV1: subindex = 0; GR_FALLTHROUGH;
+      // no break
+    case gr_slatUserDefn :  return subindex < seg->numAttrs() ?  m_userAttr[subindex] : 0;
+    case gr_slatSegSplit :  return seg->charinfo(m_original)->flags() & 3;
+    case gr_slatBidiLevel:  return m_bidiLevel;
+    case gr_slatColFlags :		{ SlotCollision *c = seg->collisionInfo(this); return c ? c->flags() : 0; }
+    case gr_slatColLimitblx:SLOTGETCOLATTR(limit().bl.x)
+    case gr_slatColLimitbly:SLOTGETCOLATTR(limit().bl.y)
+    case gr_slatColLimittrx:SLOTGETCOLATTR(limit().tr.x)
+    case gr_slatColLimittry:SLOTGETCOLATTR(limit().tr.y)
+    case gr_slatColShiftx :	SLOTGETCOLATTR(offset().x)
+    case gr_slatColShifty :	SLOTGETCOLATTR(offset().y)
+    case gr_slatColMargin :	SLOTGETCOLATTR(margin())
+    case gr_slatColMarginWt:SLOTGETCOLATTR(marginWt())
+    case gr_slatColExclGlyph:SLOTGETCOLATTR(exclGlyph())
+    case gr_slatColExclOffx:SLOTGETCOLATTR(exclOffset().x)
+    case gr_slatColExclOffy:SLOTGETCOLATTR(exclOffset().y)
+    case gr_slatSeqClass :	SLOTGETCOLATTR(seqClass())
+    case gr_slatSeqProxClass:SLOTGETCOLATTR(seqProxClass())
+    case gr_slatSeqOrder :	SLOTGETCOLATTR(seqOrder())
+    case gr_slatSeqAboveXoff:SLOTGETCOLATTR(seqAboveXoff())
+    case gr_slatSeqAboveWt: SLOTGETCOLATTR(seqAboveWt())
+    case gr_slatSeqBelowXlim:SLOTGETCOLATTR(seqBelowXlim())
+    case gr_slatSeqBelowWt:	SLOTGETCOLATTR(seqBelowWt())
+    case gr_slatSeqValignHt:SLOTGETCOLATTR(seqValignHt())
+    case gr_slatSeqValignWt:SLOTGETCOLATTR(seqValignWt())
+    default : return 0;
+    }
+}
+
+#define SLOTCOLSETATTR(x) { \
+        SlotCollision *c = seg->collisionInfo(this); \
+        if (c) { c-> x ; c->setFlags(c->flags() & ~SlotCollision::COLL_KNOWN); } \
+        break; }
+#define SLOTCOLSETCOMPLEXATTR(t, y, x) { \
+        SlotCollision *c = seg->collisionInfo(this); \
+        if (c) { \
+        const t &s = c-> y; \
+        c-> x ; c->setFlags(c->flags() & ~SlotCollision::COLL_KNOWN); } \
+        break; }
+
+void Slot::setAttr(Segment *seg, attrCode ind, uint8 subindex, int16 value, const SlotMap & map)
+{
+    if (ind == gr_slatUserDefnV1)
+    {
+        ind = gr_slatUserDefn;
+        subindex = 0;
+        if (seg->numAttrs() == 0)
+            return;
+    }
+    else if (ind >= gr_slatJStretch && ind < gr_slatJStretch + 20 && ind != gr_slatJWidth)
+    {
+        int indx = ind - gr_slatJStretch;
+        return setJustify(seg, indx / 5, indx % 5, value);
+    }
+
+    switch (ind)
+    {
+    case gr_slatAdvX :  m_advance.x = value; break;
+    case gr_slatAdvY :  m_advance.y = value; break;
+    case gr_slatAttTo :
+    {
+        const uint16 idx = uint16(value);
+        if (idx < map.size() && map[idx])
+        {
+            Slot *other = map[idx];
+            if (other == this || other == m_parent || other->isCopied()) break;
+            if (m_parent) { m_parent->removeChild(this); attachTo(NULL); }
+            Slot *pOther = other;
+            int count = 0;
+            bool foundOther = false;
+            while (pOther)
+            {
+                ++count;
+                if (pOther == this) foundOther = true;
+                pOther = pOther->attachedTo();
+            }
+            for (pOther = m_child; pOther; pOther = pOther->m_child)
+                ++count;
+            for (pOther = m_sibling; pOther; pOther = pOther->m_sibling)
+                ++count;
+            if (count < 100 && !foundOther && other->child(this))
+            {
+                attachTo(other);
+                if ((map.dir() != 0) ^ (idx > subindex))
+                    m_with = Position(advance(), 0);
+                else        // normal match to previous root
+                    m_attach = Position(other->advance(), 0);
+            }
+        }
+        break;
+    }
+    case gr_slatAttX :          m_attach.x = value; break;
+    case gr_slatAttY :          m_attach.y = value; break;
+    case gr_slatAttXOff :
+    case gr_slatAttYOff :       break;
+    case gr_slatAttWithX :      m_with.x = value; break;
+    case gr_slatAttWithY :      m_with.y = value; break;
+    case gr_slatAttWithXOff :
+    case gr_slatAttWithYOff :   break;
+    case gr_slatAttLevel :
+        m_attLevel = byte(value);
+        break;
+    case gr_slatBreak :
+        seg->charinfo(m_original)->breakWeight(value);
+        break;
+    case gr_slatCompRef :   break;      // not sure what to do here
+    case gr_slatDir : break;
+    case gr_slatInsert :
+        markInsertBefore(value? true : false);
+        break;
+    case gr_slatPosX :      break; // can't set these here
+    case gr_slatPosY :      break;
+    case gr_slatShiftX :    m_shift.x = value; break;
+    case gr_slatShiftY :    m_shift.y = value; break;
+    case gr_slatMeasureSol :    break;
+    case gr_slatMeasureEol :    break;
+    case gr_slatJWidth :    just(value); break;
+    case gr_slatSegSplit :  seg->charinfo(m_original)->addflags(value & 3); break;
+    case gr_slatUserDefn :  m_userAttr[subindex] = value; break;
+    case gr_slatColFlags :  {
+        SlotCollision *c = seg->collisionInfo(this);
+        if (c)
+            c->setFlags(value);
+        break; }
+    case gr_slatColLimitblx :	SLOTCOLSETCOMPLEXATTR(Rect, limit(), setLimit(Rect(Position(value, s.bl.y), s.tr)))
+    case gr_slatColLimitbly :	SLOTCOLSETCOMPLEXATTR(Rect, limit(), setLimit(Rect(Position(s.bl.x, value), s.tr)))
+    case gr_slatColLimittrx :	SLOTCOLSETCOMPLEXATTR(Rect, limit(), setLimit(Rect(s.bl, Position(value, s.tr.y))))
+    case gr_slatColLimittry :	SLOTCOLSETCOMPLEXATTR(Rect, limit(), setLimit(Rect(s.bl, Position(s.tr.x, value))))
+    case gr_slatColMargin :		SLOTCOLSETATTR(setMargin(value))
+    case gr_slatColMarginWt :	SLOTCOLSETATTR(setMarginWt(value))
+    case gr_slatColExclGlyph :	SLOTCOLSETATTR(setExclGlyph(value))
+    case gr_slatColExclOffx :	SLOTCOLSETCOMPLEXATTR(Position, exclOffset(), setExclOffset(Position(value, s.y)))
+    case gr_slatColExclOffy :	SLOTCOLSETCOMPLEXATTR(Position, exclOffset(), setExclOffset(Position(s.x, value)))
+    case gr_slatSeqClass :		SLOTCOLSETATTR(setSeqClass(value))
+	case gr_slatSeqProxClass :	SLOTCOLSETATTR(setSeqProxClass(value))
+    case gr_slatSeqOrder :		SLOTCOLSETATTR(setSeqOrder(value))
+    case gr_slatSeqAboveXoff :	SLOTCOLSETATTR(setSeqAboveXoff(value))
+    case gr_slatSeqAboveWt :	SLOTCOLSETATTR(setSeqAboveWt(value))
+    case gr_slatSeqBelowXlim :	SLOTCOLSETATTR(setSeqBelowXlim(value))
+    case gr_slatSeqBelowWt :	SLOTCOLSETATTR(setSeqBelowWt(value))
+    case gr_slatSeqValignHt :	SLOTCOLSETATTR(setSeqValignHt(value))
+    case gr_slatSeqValignWt :	SLOTCOLSETATTR(setSeqValignWt(value))
+    default :
+        break;
+    }
+}
+
+int Slot::getJustify(const Segment *seg, uint8 level, uint8 subindex) const
+{
+    if (level && level >= seg->silf()->numJustLevels()) return 0;
+
+    if (m_justs)
+        return m_justs->values[level * SlotJustify::NUMJUSTPARAMS + subindex];
+
+    if (level >= seg->silf()->numJustLevels()) return 0;
+    Justinfo *jAttrs = seg->silf()->justAttrs() + level;
+
+    switch (subindex) {
+        case 0 : return seg->glyphAttr(gid(), jAttrs->attrStretch());
+        case 1 : return seg->glyphAttr(gid(), jAttrs->attrShrink());
+        case 2 : return seg->glyphAttr(gid(), jAttrs->attrStep());
+        case 3 : return seg->glyphAttr(gid(), jAttrs->attrWeight());
+        case 4 : return 0;      // not been set yet, so clearly 0
+        default: return 0;
+    }
+}
+
+void Slot::setJustify(Segment *seg, uint8 level, uint8 subindex, int16 value)
+{
+    if (level && level >= seg->silf()->numJustLevels()) return;
+    if (!m_justs)
+    {
+        SlotJustify *j = seg->newJustify();
+        if (!j) return;
+        j->LoadSlot(this, seg);
+        m_justs = j;
+    }
+    m_justs->values[level * SlotJustify::NUMJUSTPARAMS + subindex] = value;
+}
+
+bool Slot::child(Slot *ap)
+{
+    if (this == ap) return false;
+    else if (ap == m_child) return true;
+    else if (!m_child)
+        m_child = ap;
+    else
+        return m_child->sibling(ap);
+    return true;
+}
+
+bool Slot::sibling(Slot *ap)
+{
+    if (this == ap) return false;
+    else if (ap == m_sibling) return true;
+    else if (!m_sibling || !ap)
+        m_sibling = ap;
+    else
+        return m_sibling->sibling(ap);
+    return true;
+}
+
+bool Slot::removeChild(Slot *ap)
+{
+    if (this == ap || !m_child || !ap) return false;
+    else if (ap == m_child)
+    {
+        Slot *nSibling = m_child->nextSibling();
+        m_child->nextSibling(NULL);
+        m_child = nSibling;
+        return true;
+    }
+    for (Slot *p = m_child; p; p = p->m_sibling)
+    {
+        if (p->m_sibling && p->m_sibling == ap)
+        {
+            p->m_sibling = p->m_sibling->m_sibling;
+            ap->nextSibling(NULL);
+            return true;
+        }
+    }
+    return false;
+}
+
+void Slot::setGlyph(Segment *seg, uint16 glyphid, const GlyphFace * theGlyph)
+{
+    m_glyphid = glyphid;
+    m_bidiCls = -1;
+    if (!theGlyph)
+    {
+        theGlyph = seg->getFace()->glyphs().glyphSafe(glyphid);
+        if (!theGlyph)
+        {
+            m_realglyphid = 0;
+            m_advance = Position(0.,0.);
+            return;
+        }
+    }
+    m_realglyphid = theGlyph->attrs()[seg->silf()->aPseudo()];
+    if (m_realglyphid > seg->getFace()->glyphs().numGlyphs())
+        m_realglyphid = 0;
+    const GlyphFace *aGlyph = theGlyph;
+    if (m_realglyphid)
+    {
+        aGlyph = seg->getFace()->glyphs().glyphSafe(m_realglyphid);
+        if (!aGlyph) aGlyph = theGlyph;
+    }
+    m_advance = Position(aGlyph->theAdvance().x, 0.);
+    if (seg->silf()->aPassBits())
+    {
+        seg->mergePassBits(uint8(theGlyph->attrs()[seg->silf()->aPassBits()]));
+        if (seg->silf()->numPasses() > 16)
+            seg->mergePassBits(theGlyph->attrs()[seg->silf()->aPassBits()+1] << 16);
+    }
+}
+
+void Slot::floodShift(Position adj, int depth)
+{
+    if (depth > 100)
+        return;
+    m_position += adj;
+    if (m_child) m_child->floodShift(adj, depth + 1);
+    if (m_sibling) m_sibling->floodShift(adj, depth + 1);
+}
+
+void SlotJustify::LoadSlot(const Slot *s, const Segment *seg)
+{
+    for (int i = seg->silf()->numJustLevels() - 1; i >= 0; --i)
+    {
+        Justinfo *justs = seg->silf()->justAttrs() + i;
+        int16 *v = values + i * NUMJUSTPARAMS;
+        v[0] = seg->glyphAttr(s->gid(), justs->attrStretch());
+        v[1] = seg->glyphAttr(s->gid(), justs->attrShrink());
+        v[2] = seg->glyphAttr(s->gid(), justs->attrStep());
+        v[3] = seg->glyphAttr(s->gid(), justs->attrWeight());
+    }
+}
+
+Slot * Slot::nextInCluster(const Slot *s) const
+{
+    Slot *base;
+    if (s->firstChild())
+        return s->firstChild();
+    else if (s->nextSibling())
+        return s->nextSibling();
+    while ((base = s->attachedTo()))
+    {
+        // if (base->firstChild() == s && base->nextSibling())
+        if (base->nextSibling())
+            return base->nextSibling();
+        s = base;
+    }
+    return NULL;
+}
+
+bool Slot::isChildOf(const Slot *base) const
+{
+    for (Slot *p = m_parent; p; p = p->m_parent)
+        if (p == base)
+            return true;
+    return false;
+}
diff --git a/thirdparty/graphite/src/Sparse.cpp b/thirdparty/graphite/src/Sparse.cpp
new file mode 100644
index 0000000000..aa43113669
--- /dev/null
+++ b/thirdparty/graphite/src/Sparse.cpp
@@ -0,0 +1,62 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cassert>
+#include "inc/Sparse.h"
+#include "inc/bits.h"
+
+using namespace graphite2;
+
+const sparse::chunk sparse::empty_chunk = {0,0};
+
+sparse::~sparse() throw()
+{
+    if (m_array.map == &empty_chunk) return;
+    free(m_array.values);
+}
+
+
+sparse::mapped_type sparse::operator [] (const key_type k) const throw()
+{
+    mapped_type         g = key_type(k/SIZEOF_CHUNK - m_nchunks) >> (sizeof k*8 - 1);
+    const chunk &       c = m_array.map[g*k/SIZEOF_CHUNK];
+    const mask_t        m = c.mask >> (SIZEOF_CHUNK - 1 - (k%SIZEOF_CHUNK));
+    g *= m & 1;
+
+    return g*m_array.values[g*(c.offset + bit_set_count(m >> 1))];
+}
+
+
+size_t sparse::capacity() const throw()
+{
+    size_t n = m_nchunks,
+           s = 0;
+
+    for (const chunk *ci=m_array.map; n; --n, ++ci)
+        s += bit_set_count(ci->mask);
+
+    return s;
+}
diff --git a/thirdparty/graphite/src/TtfUtil.cpp b/thirdparty/graphite/src/TtfUtil.cpp
new file mode 100644
index 0000000000..2eb46a11fb
--- /dev/null
+++ b/thirdparty/graphite/src/TtfUtil.cpp
@@ -0,0 +1,2053 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+/*--------------------------------------------------------------------*//*:Ignore this sentence.
+
+File: TtfUtil.cpp
+Responsibility: Alan Ward
+Last reviewed: Not yet.
+
+Description
+    Implements the methods for TtfUtil class. This file should remain portable to any C++
+    environment by only using standard C++ and the TTF structurs defined in Tt.h.
+-------------------------------------------------------------------------------*//*:End Ignore*/
+
+
+/***********************************************************************************************
+    Include files
+***********************************************************************************************/
+// Language headers
+//#include <algorithm>
+#include <cassert>
+#include <cstddef>
+#include <cstring>
+#include <climits>
+#include <cwchar>
+//#include <stdexcept>
+// Platform headers
+// Module headers
+#include "inc/TtfUtil.h"
+#include "inc/TtfTypes.h"
+#include "inc/Endian.h"
+
+/***********************************************************************************************
+    Forward declarations
+***********************************************************************************************/
+
+/***********************************************************************************************
+    Local Constants and static variables
+***********************************************************************************************/
+namespace
+{
+#ifdef ALL_TTFUTILS
+    // max number of components allowed in composite glyphs
+    const int kMaxGlyphComponents = 8;
+#endif
+
+    template <int R, typename T>
+    inline float fixed_to_float(const T f) {
+        return float(f)/float(2^R);
+    }
+
+/*----------------------------------------------------------------------------------------------
+    Table of standard Postscript glyph names. From Martin Hosken. Disagress with ttfdump.exe
+---------------------------------------------------------------------------------------------*/
+#ifdef ALL_TTFUTILS
+    const int kcPostNames = 258;
+
+    const char * rgPostName[kcPostNames] = {
+        ".notdef", ".null", "nonmarkingreturn", "space", "exclam", "quotedbl", "numbersign",
+        "dollar", "percent", "ampersand", "quotesingle", "parenleft",
+        "parenright", "asterisk", "plus", "comma", "hyphen", "period", "slash",
+        "zero", "one", "two", "three", "four", "five", "six", "seven", "eight",
+        "nine", "colon", "semicolon", "less", "equal", "greater", "question",
+        "at", "A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M",
+        "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W", "X", "Y", "Z",
+        "bracketleft", "backslash", "bracketright", "asciicircum",
+        "underscore", "grave", "a", "b", "c", "d", "e", "f", "g", "h", "i",
+        "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w",
+        "x", "y", "z", "braceleft", "bar", "braceright", "asciitilde",
+        "Adieresis", "Aring", "Ccedilla", "Eacute", "Ntilde", "Odieresis",
+        "Udieresis", "aacute", "agrave", "acircumflex", "adieresis", "atilde",
+        "aring", "ccedilla", "eacute", "egrave", "ecircumflex", "edieresis",
+        "iacute", "igrave", "icircumflex", "idieresis", "ntilde", "oacute",
+        "ograve", "ocircumflex", "odieresis", "otilde", "uacute", "ugrave",
+        "ucircumflex", "udieresis", "dagger", "degree", "cent", "sterling",
+        "section", "bullet", "paragraph", "germandbls", "registered",
+        "copyright", "trademark", "acute", "dieresis", "notequal", "AE",
+        "Oslash", "infinity", "plusminus", "lessequal", "greaterequal", "yen",
+        "mu", "partialdiff", "summation", "product", "pi", "integral",
+        "ordfeminine", "ordmasculine", "Omega", "ae", "oslash", "questiondown",
+        "exclamdown", "logicalnot", "radical", "florin", "approxequal",
+        "Delta", "guillemotleft", "guillemotright", "ellipsis", "nonbreakingspace",
+        "Agrave", "Atilde", "Otilde", "OE", "oe", "endash", "emdash",
+        "quotedblleft", "quotedblright", "quoteleft", "quoteright", "divide",
+        "lozenge", "ydieresis", "Ydieresis", "fraction", "currency",
+        "guilsinglleft", "guilsinglright", "fi", "fl", "daggerdbl", "periodcentered",
+        "quotesinglbase", "quotedblbase", "perthousand", "Acircumflex",
+        "Ecircumflex", "Aacute", "Edieresis", "Egrave", "Iacute",
+        "Icircumflex", "Idieresis", "Igrave", "Oacute", "Ocircumflex",
+        "apple", "Ograve", "Uacute", "Ucircumflex", "Ugrave", "dotlessi",
+        "circumflex", "tilde", "macron", "breve", "dotaccent", "ring",
+        "cedilla", "hungarumlaut", "ogonek", "caron", "Lslash", "lslash",
+        "Scaron", "scaron", "Zcaron", "zcaron", "brokenbar", "Eth", "eth",
+        "Yacute", "yacute", "Thorn", "thorn", "minus", "multiply",
+        "onesuperior", "twosuperior", "threesuperior", "onehalf", "onequarter",
+        "threequarters", "franc", "Gbreve", "gbreve", "Idotaccent", "Scedilla",
+        "scedilla", "Cacute", "cacute", "Ccaron", "ccaron",
+        "dcroat" };
+#endif
+
+} // end of namespace
+
+/***********************************************************************************************
+    Methods
+***********************************************************************************************/
+
+/* Note on error processing: The code guards against bad glyph ids being used to look up data
+in open ended tables (loca, hmtx). If the glyph id comes from a cmap this shouldn't happen
+but it seems prudent to check for user errors here. The code does assume that data obtained
+from the TTF file is valid otherwise (though the CheckTable method seeks to check for
+obvious problems that might accompany a change in table versions). For example an invalid
+offset in the loca table which could exceed the size of the glyf table is NOT trapped.
+Likewise if numberOf_LongHorMetrics in the hhea table is wrong, this will NOT be trapped,
+which could cause a lookup in the hmtx table to exceed the table length. Of course, TTF tables
+that are completely corrupt will cause unpredictable results. */
+
+/* Note on composite glyphs: Glyphs that have components that are themselves composites
+are not supported. IsDeepComposite can be used to test for this. False is returned from many
+of the methods in this cases. It is unclear how to build composite glyphs in some cases,
+so this code represents my best guess until test cases can be found. See notes on the high-
+level GlyfPoints method. */
+namespace graphite2
+{
+namespace TtfUtil
+{
+
+
+/*----------------------------------------------------------------------------------------------
+    Get offset and size of the offset table needed to find table directory.
+    Return true if success, false otherwise.
+    lSize excludes any table directory entries.
+----------------------------------------------------------------------------------------------*/
+bool GetHeaderInfo(size_t & lOffset, size_t & lSize)
+{
+    lOffset = 0;
+    lSize   = offsetof(Sfnt::OffsetSubTable, table_directory);
+    assert(sizeof(uint32) + 4*sizeof (uint16) == lSize);
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Check the offset table for expected data.
+    Return true if success, false otherwise.
+----------------------------------------------------------------------------------------------*/
+bool CheckHeader(const void * pHdr)
+{
+    const Sfnt::OffsetSubTable * pOffsetTable
+        = reinterpret_cast<const Sfnt::OffsetSubTable *>(pHdr);
+
+    return pHdr && be::swap(pOffsetTable->scaler_type) == Sfnt::OffsetSubTable::TrueTypeWin;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get offset and size of the table directory.
+    Return true if successful, false otherwise.
+----------------------------------------------------------------------------------------------*/
+bool GetTableDirInfo(const void * pHdr, size_t & lOffset, size_t & lSize)
+{
+    const Sfnt::OffsetSubTable * pOffsetTable
+        = reinterpret_cast<const Sfnt::OffsetSubTable *>(pHdr);
+
+    lOffset = offsetof(Sfnt::OffsetSubTable, table_directory);
+    lSize   = be::swap(pOffsetTable->num_tables)
+        * sizeof(Sfnt::OffsetSubTable::Entry);
+
+    return true;
+}
+
+
+/*----------------------------------------------------------------------------------------------
+    Get offset and size of the specified table.
+    Return true if successful, false otherwise. On false, offset and size will be 0.
+----------------------------------------------------------------------------------------------*/
+bool GetTableInfo(const Tag TableTag, const void * pHdr, const void * pTableDir,
+                           size_t & lOffset, size_t & lSize)
+{
+    const Sfnt::OffsetSubTable * pOffsetTable
+        = reinterpret_cast<const Sfnt::OffsetSubTable *>(pHdr);
+    const size_t num_tables = be::swap(pOffsetTable->num_tables);
+    const Sfnt::OffsetSubTable::Entry
+        * entry_itr = reinterpret_cast<const Sfnt::OffsetSubTable::Entry *>(
+            pTableDir),
+        * const  dir_end = entry_itr + num_tables;
+
+    if (num_tables > 40)
+        return false;
+
+    for (;entry_itr != dir_end; ++entry_itr) // 40 - safe guard
+    {
+        if (be::swap(entry_itr->tag) == TableTag)
+        {
+            lOffset = be::swap(entry_itr->offset);
+            lSize   = be::swap(entry_itr->length);
+            return true;
+        }
+    }
+
+    return false;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Check the specified table. Tests depend on the table type.
+    Return true if successful, false otherwise.
+----------------------------------------------------------------------------------------------*/
+bool CheckTable(const Tag TableId, const void * pTable, size_t lTableSize)
+{
+    using namespace Sfnt;
+
+    if (pTable == 0 || lTableSize < 4) return false;
+
+    switch(TableId)
+    {
+    case Tag::cmap: // cmap
+    {
+        const Sfnt::CharacterCodeMap * const pCmap
+            = reinterpret_cast<const Sfnt::CharacterCodeMap *>(pTable);
+        if (lTableSize < sizeof(Sfnt::CharacterCodeMap))
+            return false;
+        return be::swap(pCmap->version) == 0;
+    }
+
+    case Tag::head: // head
+    {
+        const Sfnt::FontHeader * const pHead
+            = reinterpret_cast<const Sfnt::FontHeader *>(pTable);
+        if (lTableSize < sizeof(Sfnt::FontHeader))
+            return false;
+        bool r = be::swap(pHead->version) == OneFix
+            && be::swap(pHead->magic_number) == FontHeader::MagicNumber
+            && be::swap(pHead->glyph_data_format)
+                    == FontHeader::GlypDataFormat
+            && (be::swap(pHead->index_to_loc_format)
+                    == FontHeader::ShortIndexLocFormat
+                || be::swap(pHead->index_to_loc_format)
+                    == FontHeader::LongIndexLocFormat)
+            && sizeof(FontHeader) <= lTableSize;
+        return r;
+    }
+
+    case Tag::post: // post
+    {
+        const Sfnt::PostScriptGlyphName * const pPost
+            = reinterpret_cast<const Sfnt::PostScriptGlyphName *>(pTable);
+        if (lTableSize < sizeof(Sfnt::PostScriptGlyphName))
+            return false;
+        const fixed format = be::swap(pPost->format);
+        bool r = format == PostScriptGlyphName::Format1
+            || format == PostScriptGlyphName::Format2
+            || format == PostScriptGlyphName::Format3
+            || format == PostScriptGlyphName::Format25;
+        return r;
+    }
+
+    case Tag::hhea: // hhea
+    {
+        const Sfnt::HorizontalHeader * pHhea =
+            reinterpret_cast<const Sfnt::HorizontalHeader *>(pTable);
+        if (lTableSize < sizeof(Sfnt::HorizontalHeader))
+            return false;
+        bool r = be::swap(pHhea->version) == OneFix
+            && be::swap(pHhea->metric_data_format) == 0
+            && sizeof (Sfnt::HorizontalHeader) <= lTableSize;
+        return r;
+    }
+
+    case Tag::maxp: // maxp
+    {
+        const Sfnt::MaximumProfile * pMaxp =
+            reinterpret_cast<const Sfnt::MaximumProfile *>(pTable);
+        if (lTableSize < sizeof(Sfnt::MaximumProfile))
+            return false;
+        bool r = be::swap(pMaxp->version) == OneFix
+            && sizeof(Sfnt::MaximumProfile) <= lTableSize;
+        return r;
+    }
+
+    case Tag::OS_2: // OS/2
+    {
+        const Sfnt::Compatibility * pOs2
+            = reinterpret_cast<const Sfnt::Compatibility *>(pTable);
+        if (be::swap(pOs2->version) == 0)
+        { // OS/2 table version 1 size
+//          if (sizeof(Sfnt::Compatibility)
+//                  - sizeof(uint32)*2 - sizeof(int16)*2
+//                  - sizeof(uint16)*3 <= lTableSize)
+            if (sizeof(Sfnt::Compatibility0) <= lTableSize)
+                return true;
+        }
+        else if (be::swap(pOs2->version) == 1)
+        { // OS/2 table version 2 size
+//          if (sizeof(Sfnt::Compatibility)
+//                  - sizeof(int16) *2
+//                  - sizeof(uint16)*3 <= lTableSize)
+            if (sizeof(Sfnt::Compatibility1) <= lTableSize)
+                return true;
+        }
+        else if (be::swap(pOs2->version) == 2)
+        { // OS/2 table version 3 size
+            if (sizeof(Sfnt::Compatibility2) <= lTableSize)
+                return true;
+        }
+        else if (be::swap(pOs2->version) == 3 || be::swap(pOs2->version) == 4)
+        { // OS/2 table version 4 size - version 4 changed the meaning of some fields which we don't use
+            if (sizeof(Sfnt::Compatibility3) <= lTableSize)
+                return true;
+        }
+        else
+            return false;
+        break;
+    }
+
+    case Tag::name:
+    {
+        const Sfnt::FontNames * pName
+            = reinterpret_cast<const Sfnt::FontNames *>(pTable);
+        if (lTableSize < sizeof(Sfnt::FontNames))
+            return false;
+        return be::swap(pName->format) == 0;
+    }
+
+    case Tag::glyf:
+    {
+        return (lTableSize >= sizeof(Sfnt::Glyph));
+    }
+
+    default:
+        break;
+    }
+
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the number of glyphs in the font. Should never be less than zero.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+size_t GlyphCount(const void * pMaxp)
+{
+    const Sfnt::MaximumProfile * pTable =
+            reinterpret_cast<const Sfnt::MaximumProfile *>(pMaxp);
+    return be::swap(pTable->num_glyphs);
+}
+
+#ifdef ALL_TTFUTILS
+/*----------------------------------------------------------------------------------------------
+    Return the maximum number of components for any composite glyph in the font.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+size_t  MaxCompositeComponentCount(const void * pMaxp)
+{
+    const Sfnt::MaximumProfile * pTable =
+            reinterpret_cast<const Sfnt::MaximumProfile *>(pMaxp);
+    return be::swap(pTable->max_component_elements);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Composite glyphs can be composed of glyphs that are themselves composites.
+    This method returns the maximum number of levels like this for any glyph in the font.
+    A non-composite glyph has a level of 1.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+size_t  MaxCompositeLevelCount(const void * pMaxp)
+{
+    const Sfnt::MaximumProfile * pTable =
+            reinterpret_cast<const Sfnt::MaximumProfile *>(pMaxp);
+    return be::swap(pTable->max_component_depth);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the number of glyphs in the font according to a differt source.
+    Should never be less than zero. Return -1 on failure.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+size_t LocaGlyphCount(size_t lLocaSize, const void * pHead) //throw(std::domain_error)
+{
+
+    const Sfnt::FontHeader * pTable
+        = reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    if (be::swap(pTable->index_to_loc_format)
+        == Sfnt::FontHeader::ShortIndexLocFormat)
+    // loca entries are two bytes and have been divided by two
+        return (lLocaSize >> 1) - 1;
+
+    if (be::swap(pTable->index_to_loc_format)
+        == Sfnt::FontHeader::LongIndexLocFormat)
+     // loca entries are four bytes
+        return (lLocaSize >> 2) - 1;
+
+    return -1;
+    //throw std::domain_error("head table in inconsistent state. The font may be corrupted");
+}
+#endif
+
+/*----------------------------------------------------------------------------------------------
+    Return the design units the font is designed with
+----------------------------------------------------------------------------------------------*/
+int DesignUnits(const void * pHead)
+{
+    const Sfnt::FontHeader * pTable =
+            reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    return be::swap(pTable->units_per_em);
+}
+
+#ifdef ALL_TTFUTILS
+/*----------------------------------------------------------------------------------------------
+    Return the checksum from the head table, which serves as a unique identifer for the font.
+----------------------------------------------------------------------------------------------*/
+int HeadTableCheckSum(const void * pHead)
+{
+    const Sfnt::FontHeader * pTable =
+            reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    return be::swap(pTable->check_sum_adjustment);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the create time from the head table. This consists of a 64-bit integer, which
+    we return here as two 32-bit integers.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+void HeadTableCreateTime(const void * pHead,
+    unsigned int * pnDateBC, unsigned int * pnDateAD)
+{
+    const Sfnt::FontHeader * pTable =
+            reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    *pnDateBC = be::swap(pTable->created[0]);
+    *pnDateAD = be::swap(pTable->created[1]);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the modify time from the head table.This consists of a 64-bit integer, which
+    we return here as two 32-bit integers.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+void HeadTableModifyTime(const void * pHead,
+    unsigned int * pnDateBC, unsigned int *pnDateAD)
+{
+    const Sfnt::FontHeader * pTable =
+            reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+   ;
+    *pnDateBC = be::swap(pTable->modified[0]);
+    *pnDateAD = be::swap(pTable->modified[1]);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return true if the font is italic.
+----------------------------------------------------------------------------------------------*/
+bool IsItalic(const void * pHead)
+{
+    const Sfnt::FontHeader * pTable =
+            reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    return ((be::swap(pTable->mac_style) & 0x00000002) != 0);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the ascent for the font
+----------------------------------------------------------------------------------------------*/
+int FontAscent(const void * pOs2)
+{
+    const Sfnt::Compatibility * pTable = reinterpret_cast<const Sfnt::Compatibility *>(pOs2);
+
+    return be::swap(pTable->win_ascent);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the descent for the font
+----------------------------------------------------------------------------------------------*/
+int FontDescent(const void * pOs2)
+{
+    const Sfnt::Compatibility * pTable = reinterpret_cast<const Sfnt::Compatibility *>(pOs2);
+
+    return be::swap(pTable->win_descent);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the bold and italic style bits.
+    Return true if successful. false otherwise.
+    In addition to checking the OS/2 table, one could also check
+        the head table's macStyle field (overridden by the OS/2 table on Win)
+        the sub-family name in the name table (though this can contain oblique, dark, etc too)
+----------------------------------------------------------------------------------------------*/
+bool FontOs2Style(const void *pOs2, bool & fBold, bool & fItalic)
+{
+    const Sfnt::Compatibility * pTable = reinterpret_cast<const Sfnt::Compatibility *>(pOs2);
+
+    fBold = (be::swap(pTable->fs_selection) & Sfnt::Compatibility::Bold) != 0;
+    fItalic = (be::swap(pTable->fs_selection) & Sfnt::Compatibility::Italic) != 0;
+
+    return true;
+}
+#endif
+
+/*----------------------------------------------------------------------------------------------
+    Method for searching name table.
+----------------------------------------------------------------------------------------------*/
+bool GetNameInfo(const void * pName, int nPlatformId, int nEncodingId,
+        int nLangId, int nNameId, size_t & lOffset, size_t & lSize)
+{
+    lOffset = 0;
+    lSize = 0;
+
+    const Sfnt::FontNames * pTable = reinterpret_cast<const Sfnt::FontNames *>(pName);
+    uint16 cRecord = be::swap(pTable->count);
+    uint16 nRecordOffset = be::swap(pTable->string_offset);
+    const Sfnt::NameRecord * pRecord = reinterpret_cast<const Sfnt::NameRecord *>(pTable + 1);
+
+    for (int i = 0; i < cRecord; ++i)
+    {
+        if (be::swap(pRecord->platform_id) == nPlatformId &&
+            be::swap(pRecord->platform_specific_id) == nEncodingId &&
+            be::swap(pRecord->language_id) == nLangId &&
+            be::swap(pRecord->name_id) == nNameId)
+        {
+            lOffset = be::swap(pRecord->offset) + nRecordOffset;
+            lSize = be::swap(pRecord->length);
+            return true;
+        }
+        pRecord++;
+    }
+
+    return false;
+}
+
+#ifdef ALL_TTFUTILS
+/*----------------------------------------------------------------------------------------------
+    Return all the lang-IDs that have data for the given name-IDs. Assume that there is room
+    in the return array (langIdList) for 128 items. The purpose of this method is to return
+    a list of all possible lang-IDs.
+----------------------------------------------------------------------------------------------*/
+int GetLangsForNames(const void * pName, int nPlatformId, int nEncodingId,
+        int * nameIdList, int cNameIds, short * langIdList)
+{
+    const Sfnt::FontNames * pTable = reinterpret_cast<const Sfnt::FontNames *>(pName);
+        int cLangIds = 0;
+    uint16 cRecord = be::swap(pTable->count);
+        if (cRecord > 127) return cLangIds;
+    //uint16 nRecordOffset = swapw(pTable->stringOffset);
+    const Sfnt::NameRecord * pRecord = reinterpret_cast<const Sfnt::NameRecord *>(pTable + 1);
+
+    for (int i = 0; i < cRecord; ++i)
+    {
+        if (be::swap(pRecord->platform_id) == nPlatformId &&
+            be::swap(pRecord->platform_specific_id) == nEncodingId)
+        {
+            bool fNameFound = false;
+            int nLangId = be::swap(pRecord->language_id);
+            int nNameId = be::swap(pRecord->name_id);
+            for (int j = 0; j < cNameIds; j++)
+            {
+                if (nNameId == nameIdList[j])
+                {
+                    fNameFound = true;
+                    break;
+                }
+            }
+            if (fNameFound)
+            {
+                // Add it if it's not there.
+                int ilang;
+                for (ilang = 0; ilang < cLangIds; ilang++)
+                    if (langIdList[ilang] == nLangId)
+                        break;
+                if (ilang >= cLangIds)
+                {
+                    langIdList[cLangIds] = short(nLangId);
+                    cLangIds++;
+                }
+                if (cLangIds == 128)
+                    return cLangIds;
+            }
+        }
+        pRecord++;
+    }
+
+    return cLangIds;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the offset and size of the font family name in English for the MS Platform with Unicode
+    writing system. The offset is within the pName data. The string is double byte with MSB
+    first.
+----------------------------------------------------------------------------------------------*/
+bool Get31EngFamilyInfo(const void * pName, size_t & lOffset, size_t & lSize)
+{
+    return GetNameInfo(pName, Sfnt::NameRecord::Microsoft, 1, 1033,
+        Sfnt::NameRecord::Family, lOffset, lSize);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the offset and size of the full font name in English for the MS Platform with Unicode
+    writing system. The offset is within the pName data. The string is double byte with MSB
+    first.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+bool Get31EngFullFontInfo(const void * pName, size_t & lOffset, size_t & lSize)
+{
+    return GetNameInfo(pName, Sfnt::NameRecord::Microsoft, 1, 1033,
+        Sfnt::NameRecord::Fullname, lOffset, lSize);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the offset and size of the font family name in English for the MS Platform with Symbol
+    writing system. The offset is within the pName data. The string is double byte with MSB
+    first.
+----------------------------------------------------------------------------------------------*/
+bool Get30EngFamilyInfo(const void * pName, size_t & lOffset, size_t & lSize)
+{
+    return GetNameInfo(pName, Sfnt::NameRecord::Microsoft, 0, 1033,
+        Sfnt::NameRecord::Family, lOffset, lSize);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the offset and size of the full font name in English for the MS Platform with Symbol
+    writing system. The offset is within the pName data. The string is double byte with MSB
+    first.
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+bool Get30EngFullFontInfo(const void * pName, size_t & lOffset, size_t & lSize)
+{
+    return GetNameInfo(pName, Sfnt::NameRecord::Microsoft, 0, 1033,
+        Sfnt::NameRecord::Fullname, lOffset, lSize);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the Glyph ID for a given Postscript name. This method finds the first glyph which
+    matches the requested Postscript name. Ideally every glyph should have a unique Postscript
+    name (except for special names such as .notdef), but this is not always true.
+    On failure return value less than zero.
+       -1 - table search failed
+       -2 - format 3 table (no Postscript glyph info)
+       -3 - other failures
+
+    Note: this method is not currently used by the Graphite engine.
+----------------------------------------------------------------------------------------------*/
+int PostLookup(const void * pPost, size_t lPostSize, const void * pMaxp,
+                        const char * pPostName)
+{
+    using namespace Sfnt;
+
+    const Sfnt::PostScriptGlyphName * pTable
+        = reinterpret_cast<const Sfnt::PostScriptGlyphName *>(pPost);
+    fixed format = be::swap(pTable->format);
+
+    if (format == PostScriptGlyphName::Format3)
+    { // format 3 - no Postscript glyph info in font
+        return -2;
+    }
+
+    // search for given Postscript name among the standard names
+    int iPostName = -1; // index in standard names
+    for (int i = 0; i < kcPostNames; i++)
+    {
+        if (!strcmp(pPostName, rgPostName[i]))
+        {
+            iPostName = i;
+            break;
+        }
+    }
+
+    if (format == PostScriptGlyphName::Format1)
+    { // format 1 - use standard Postscript names
+        return iPostName;
+    }
+
+    if (format == PostScriptGlyphName::Format25)
+    {
+        if (iPostName == -1)
+            return -1;
+
+        const PostScriptGlyphName25 * pTable25
+            = static_cast<const PostScriptGlyphName25 *>(pTable);
+        int cnGlyphs = GlyphCount(pMaxp);
+        for (gid16 nGlyphId = 0; nGlyphId < cnGlyphs && nGlyphId < kcPostNames;
+                nGlyphId++)
+        { // glyph_name_index25 contains bytes so no byte swapping needed
+          // search for first glyph id that uses the standard name
+            if (nGlyphId + pTable25->offset[nGlyphId] == iPostName)
+                return nGlyphId;
+        }
+    }
+
+    if (format == PostScriptGlyphName::Format2)
+    { // format 2
+        const PostScriptGlyphName2 * pTable2
+            = static_cast<const PostScriptGlyphName2 *>(pTable);
+
+        int cnGlyphs = be::swap(pTable2->number_of_glyphs);
+
+        if (iPostName != -1)
+        { // did match a standard name, look for first glyph id mapped to that name
+            for (gid16 nGlyphId = 0; nGlyphId < cnGlyphs; nGlyphId++)
+            {
+                if (be::swap(pTable2->glyph_name_index[nGlyphId]) == iPostName)
+                    return nGlyphId;
+            }
+        }
+
+        { // did not match a standard name, search font specific names
+            size_t nStrSizeGoal = strlen(pPostName);
+            const char * pFirstGlyphName = reinterpret_cast<const char *>(
+                &pTable2->glyph_name_index[0] + cnGlyphs);
+            const char * pGlyphName = pFirstGlyphName;
+            int iInNames = 0; // index in font specific names
+            bool fFound = false;
+            const char * const endOfTable
+                = reinterpret_cast<const char *>(pTable2) + lPostSize;
+            while (pGlyphName < endOfTable && !fFound)
+            { // search Pascal strings for first matching name
+                size_t nStringSize = size_t(*pGlyphName);
+                if (nStrSizeGoal != nStringSize ||
+                    strncmp(pGlyphName + 1, pPostName, nStringSize))
+                { // did not match
+                    ++iInNames;
+                    pGlyphName += nStringSize + 1;
+                }
+                else
+                { // did match
+                    fFound = true;
+                }
+            }
+            if (!fFound)
+                return -1; // no font specific name matches request
+
+            iInNames += kcPostNames;
+            for (gid16 nGlyphId = 0; nGlyphId < cnGlyphs; nGlyphId++)
+            { // search for first glyph id that maps to the found string index
+                if (be::swap(pTable2->glyph_name_index[nGlyphId]) == iInNames)
+                    return nGlyphId;
+            }
+            return -1; // no glyph mapped to this index (very strange)
+        }
+    }
+
+    return -3;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Convert a Unicode character string from big endian (MSB first, Motorola) format to little
+    endian (LSB first, Intel) format.
+    nSize is the number of Unicode characters in the string. It should not include any
+    terminating null. If nSize is 0, it is assumed the string is null terminated. nSize
+    defaults to 0.
+    Return true if successful, false otherwise.
+----------------------------------------------------------------------------------------------*/
+void SwapWString(void * pWStr, size_t nSize /* = 0 */) //throw (std::invalid_argument)
+{
+    if (pWStr == 0)
+    {
+//      throw std::invalid_argument("null pointer given");
+        return;
+    }
+
+    uint16 * pStr = reinterpret_cast<uint16 *>(pWStr);
+    uint16 * const pStrEnd = pStr + (nSize == 0 ? wcslen((const wchar_t*)pStr) : nSize);
+
+        for (; pStr != pStrEnd; ++pStr)
+          *pStr = be::swap(*pStr);
+//  std::transform(pStr, pStrEnd, pStr, read<uint16>);
+
+//      for (int i = 0; i < nSize; i++)
+//      { // swap the wide characters in the string
+//          pStr[i] = utf16(be::swap(uint16(pStr[i])));
+//      }
+}
+#endif
+
+/*----------------------------------------------------------------------------------------------
+    Get the left-side bearing and and advance width based on the given tables and Glyph ID
+    Return true if successful, false otherwise. On false, one or both value could be INT_MIN
+----------------------------------------------------------------------------------------------*/
+bool HorMetrics(gid16 nGlyphId, const void * pHmtx, size_t lHmtxSize, const void * pHhea,
+                         int & nLsb, unsigned int & nAdvWid)
+{
+    const Sfnt::HorizontalMetric * phmtx =
+        reinterpret_cast<const Sfnt::HorizontalMetric *>(pHmtx);
+
+    const Sfnt::HorizontalHeader * phhea =
+        reinterpret_cast<const Sfnt::HorizontalHeader *>(pHhea);
+
+    size_t cLongHorMetrics = be::swap(phhea->num_long_hor_metrics);
+    if (nGlyphId < cLongHorMetrics)
+    {   // glyph id is acceptable
+        if ((nGlyphId + 1) * sizeof(Sfnt::HorizontalMetric) > lHmtxSize) return false;
+        nAdvWid = be::swap(phmtx[nGlyphId].advance_width);
+        nLsb = be::swap(phmtx[nGlyphId].left_side_bearing);
+    }
+    else
+    {
+        // guard against bad glyph id
+        size_t lLsbOffset = sizeof(Sfnt::HorizontalMetric) * cLongHorMetrics +
+            sizeof(int16) * (nGlyphId - cLongHorMetrics); // offset in bytes
+        // We test like this as LsbOffset is an offset not a length.
+        if (lLsbOffset >= lHmtxSize - sizeof(int16) || cLongHorMetrics == 0)
+        {
+            nLsb = 0;
+            return false;
+        }
+        nAdvWid = be::swap(phmtx[cLongHorMetrics - 1].advance_width);
+        nLsb = be::peek<int16>(reinterpret_cast<const byte *>(phmtx) + lLsbOffset);
+    }
+
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return a pointer to the requested cmap subtable. By default find the Microsoft Unicode
+    subtable. Pass nEncoding as -1 to find first table that matches only nPlatformId.
+    Return NULL if the subtable cannot be found.
+----------------------------------------------------------------------------------------------*/
+const void * FindCmapSubtable(const void * pCmap, int nPlatformId, /* =3 */ int nEncodingId, /* = 1 */ size_t length)
+{
+    const Sfnt::CharacterCodeMap * pTable = reinterpret_cast<const Sfnt::CharacterCodeMap *>(pCmap);
+    uint16 csuPlatforms = be::swap(pTable->num_subtables);
+    if (length && (sizeof(Sfnt::CharacterCodeMap) + 8 * (csuPlatforms - 1) > length))
+        return NULL;
+    for (int i = 0; i < csuPlatforms; i++)
+    {
+        if (be::swap(pTable->encoding[i].platform_id) == nPlatformId &&
+                (nEncodingId == -1 || be::swap(pTable->encoding[i].platform_specific_id) == nEncodingId))
+        {
+            uint32 offset = be::swap(pTable->encoding[i].offset);
+            const uint8 * pRtn = reinterpret_cast<const uint8 *>(pCmap) + offset;
+            if (length)
+            {
+                if (offset > length - 2) return NULL;
+                uint16 format = be::read<uint16>(pRtn);
+                if (format == 4)
+                {
+                    if (offset > length - 4) return NULL;
+                    uint16 subTableLength = be::peek<uint16>(pRtn);
+                    if (i + 1 == csuPlatforms)
+                    {
+                        if (subTableLength > length - offset)
+                            return NULL;
+                    }
+                    else if (subTableLength > be::swap(pTable->encoding[i+1].offset))
+                        return NULL;
+                }
+                if (format == 12)
+                {
+                    if (offset > length - 6) return NULL;
+                    uint32 subTableLength = be::peek<uint32>(pRtn);
+                    if (i + 1 == csuPlatforms)
+                    {
+                        if (subTableLength > length - offset)
+                            return NULL;
+                    }
+                    else if (subTableLength > be::swap(pTable->encoding[i+1].offset))
+                        return NULL;
+                }
+            }
+            return reinterpret_cast<const uint8 *>(pCmap) + offset;
+        }
+    }
+
+    return 0;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Check the Microsoft Unicode subtable for expected values
+----------------------------------------------------------------------------------------------*/
+bool CheckCmapSubtable4(const void * pCmapSubtable4, const void * pCmapEnd /*, unsigned int maxgid*/)
+{
+    size_t table_len = (const byte *)pCmapEnd - (const byte *)pCmapSubtable4;
+    if (!pCmapSubtable4) return false;
+    const Sfnt::CmapSubTable * pTable = reinterpret_cast<const Sfnt::CmapSubTable *>(pCmapSubtable4);
+    // Bob H say some freeware TT fonts have version 1 (eg, CALIGULA.TTF)
+    // so don't check subtable version. 21 Mar 2002 spec changes version to language.
+    if (table_len < sizeof(*pTable) || be::swap(pTable->format) != 4) return false;
+    const Sfnt::CmapSubTableFormat4 * pTable4 = reinterpret_cast<const Sfnt::CmapSubTableFormat4 *>(pCmapSubtable4);
+    if (table_len < sizeof(*pTable4))
+        return false;
+    uint16 length = be::swap(pTable4->length);
+    if (length > table_len)
+        return false;
+    if (length < sizeof(Sfnt::CmapSubTableFormat4))
+        return false;
+    uint16 nRanges = be::swap(pTable4->seg_count_x2) >> 1;
+    if (!nRanges || length < sizeof(Sfnt::CmapSubTableFormat4) + 4 * nRanges * sizeof(uint16))
+        return false;
+    // check last range is properly terminated
+    uint16 chEnd = be::peek<uint16>(pTable4->end_code + nRanges - 1);
+    if (chEnd != 0xFFFF)
+        return false;
+#if 0
+    int lastend = -1;
+    for (int i = 0; i < nRanges; ++i)
+    {
+        uint16 end = be::peek<uint16>(pTable4->end_code + i);
+        uint16 start = be::peek<uint16>(pTable4->end_code + nRanges + 1 + i);
+        int16 delta = be::peek<int16>(pTable4->end_code + 2*nRanges + 1 + i);
+        uint16 offset = be::peek<uint16>(pTable4->end_code + 3*nRanges + 1 + i);
+        if (lastend >= end || lastend >= start)
+            return false;
+        if (offset)
+        {
+            const uint16 *gstart = pTable4->end_code + 3*nRanges + 1 + i + (offset >> 1);
+            const uint16 *gend = gstart + end - start;
+            if ((char *)gend >= (char *)pCmapSubtable4 + length)
+                return false;
+            while (gstart <= gend)
+            {
+                uint16 g = be::peek<uint16>(gstart++);
+                if (g && ((g + delta) & 0xFFFF) > maxgid)
+                    return false;
+            }
+        }
+        else if (((delta + end) & 0xFFFF) > maxgid)
+            return false;
+        lastend = end;
+    }
+#endif
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the Glyph ID for the given Unicode ID in the Microsoft Unicode subtable.
+    (Actually this code only depends on subtable being format 4.)
+    Return 0 if the Unicode ID is not in the subtable.
+----------------------------------------------------------------------------------------------*/
+gid16 CmapSubtable4Lookup(const void * pCmapSubtabel4, unsigned int nUnicodeId, int rangeKey)
+{
+    const Sfnt::CmapSubTableFormat4 * pTable = reinterpret_cast<const Sfnt::CmapSubTableFormat4 *>(pCmapSubtabel4);
+
+    uint16 nSeg = be::swap(pTable->seg_count_x2) >> 1;
+
+    uint16 n;
+    const uint16 * pLeft, * pMid;
+    uint16 cMid, chStart, chEnd;
+
+    if (rangeKey)
+    {
+        pMid = &(pTable->end_code[rangeKey]);
+        chEnd = be::peek<uint16>(pMid);
+    }
+    else
+    {
+        // Binary search of the endCode[] array
+        pLeft = &(pTable->end_code[0]);
+        n = nSeg;
+        while (n > 0)
+        {
+            cMid = n >> 1;           // Pick an element in the middle
+            pMid = pLeft + cMid;
+            chEnd = be::peek<uint16>(pMid);
+            if (nUnicodeId <= chEnd)
+            {
+                if (cMid == 0 || nUnicodeId > be::peek<uint16>(pMid -1))
+                        break;          // Must be this seg or none!
+                n = cMid;            // Continue on left side, omitting mid point
+            }
+            else
+            {
+                pLeft = pMid + 1;    // Continue on right side, omitting mid point
+                n -= (cMid + 1);
+            }
+        }
+
+        if (!n)
+        return 0;
+    }
+
+    // Ok, we're down to one segment and pMid points to the endCode element
+    // Either this is it or none is.
+
+    chStart = be::peek<uint16>(pMid += nSeg + 1);
+    if (chEnd >= nUnicodeId && nUnicodeId >= chStart)
+    {
+        // Found correct segment. Find Glyph Id
+        int16 idDelta = be::peek<uint16>(pMid += nSeg);
+        uint16 idRangeOffset = be::peek<uint16>(pMid += nSeg);
+
+        if (idRangeOffset == 0)
+            return (uint16)(idDelta + nUnicodeId); // must use modulus 2^16
+
+        // Look up value in glyphIdArray
+        const ptrdiff_t offset = (nUnicodeId - chStart) + (idRangeOffset >> 1) +
+                (pMid - reinterpret_cast<const uint16 *>(pTable));
+        if (offset * 2 + 1 >= be::swap<uint16>(pTable->length))
+            return 0;
+        gid16 nGlyphId = be::peek<uint16>(reinterpret_cast<const uint16 *>(pTable)+offset);
+        // If this value is 0, return 0. Else add the idDelta
+        return nGlyphId ? nGlyphId + idDelta : 0;
+    }
+
+    return 0;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the next Unicode value in the cmap. Pass 0 to obtain the first item.
+    Returns 0xFFFF as the last item.
+    pRangeKey is an optional key that is used to optimize the search; its value is the range
+    in which the character is found.
+----------------------------------------------------------------------------------------------*/
+unsigned int CmapSubtable4NextCodepoint(const void *pCmap31, unsigned int nUnicodeId, int * pRangeKey)
+{
+    const Sfnt::CmapSubTableFormat4 * pTable = reinterpret_cast<const Sfnt::CmapSubTableFormat4 *>(pCmap31);
+
+    uint16 nRange = be::swap(pTable->seg_count_x2) >> 1;
+
+    uint32 nUnicodePrev = (uint32)nUnicodeId;
+
+    const uint16 * pStartCode = &(pTable->end_code[0])
+        + nRange // length of end code array
+        + 1;   // reserved word
+
+    if (nUnicodePrev == 0)
+    {
+        // return the first codepoint.
+        if (pRangeKey)
+            *pRangeKey = 0;
+        return be::peek<uint16>(pStartCode);
+    }
+    else if (nUnicodePrev >= 0xFFFF)
+    {
+        if (pRangeKey)
+            *pRangeKey = nRange - 1;
+        return 0xFFFF;
+    }
+
+    int iRange = (pRangeKey) ? *pRangeKey : 0;
+    // Just in case we have a bad key:
+    while (iRange > 0 && be::peek<uint16>(pStartCode + iRange) > nUnicodePrev)
+        iRange--;
+    while (iRange < nRange - 1 && be::peek<uint16>(pTable->end_code + iRange) < nUnicodePrev)
+        iRange++;
+
+    // Now iRange is the range containing nUnicodePrev.
+    unsigned int nStartCode = be::peek<uint16>(pStartCode + iRange);
+    unsigned int nEndCode = be::peek<uint16>(pTable->end_code + iRange);
+
+    if (nStartCode > nUnicodePrev)
+        // Oops, nUnicodePrev is not in the cmap! Adjust so we get a reasonable
+        // answer this time around.
+        nUnicodePrev = nStartCode - 1;
+
+    if (nEndCode > nUnicodePrev)
+    {
+        // Next is in the same range; it is the next successive codepoint.
+        if (pRangeKey)
+            *pRangeKey = iRange;
+        return nUnicodePrev + 1;
+    }
+
+    // Otherwise the next codepoint is the first one in the next range.
+    // There is guaranteed to be a next range because there must be one that
+    // ends with 0xFFFF.
+    if (pRangeKey)
+        *pRangeKey = iRange + 1;
+    return (iRange + 1 >= nRange) ? 0xFFFF : be::peek<uint16>(pStartCode + iRange + 1);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Check the Microsoft UCS-4 subtable for expected values.
+----------------------------------------------------------------------------------------------*/
+bool CheckCmapSubtable12(const void *pCmapSubtable12, const void *pCmapEnd /*, unsigned int maxgid*/)
+{
+    size_t table_len = (const byte *)pCmapEnd - (const byte *)pCmapSubtable12;
+    if (!pCmapSubtable12)  return false;
+    const Sfnt::CmapSubTable * pTable = reinterpret_cast<const Sfnt::CmapSubTable *>(pCmapSubtable12);
+    if (table_len < sizeof(*pTable) || be::swap(pTable->format) != 12)
+        return false;
+    const Sfnt::CmapSubTableFormat12 * pTable12 = reinterpret_cast<const Sfnt::CmapSubTableFormat12 *>(pCmapSubtable12);
+    if (table_len < sizeof(*pTable12))
+        return false;
+    uint32 length = be::swap(pTable12->length);
+    if (length > table_len)
+        return false;
+    if (length < sizeof(Sfnt::CmapSubTableFormat12))
+        return false;
+    uint32 num_groups = be::swap(pTable12->num_groups);
+    if (num_groups > 0x10000000 || length != (sizeof(Sfnt::CmapSubTableFormat12) + (num_groups - 1) * sizeof(uint32) * 3))
+        return false;
+#if 0
+    for (unsigned int i = 0; i < num_groups; ++i)
+    {
+        if (be::swap(pTable12->group[i].end_char_code)  - be::swap(pTable12->group[i].start_char_code) + be::swap(pTable12->group[i].start_glyph_id) > maxgid)
+            return false;
+        if (i > 0 && be::swap(pTable12->group[i].start_char_code) <= be::swap(pTable12->group[i-1].end_char_code))
+            return false;
+    }
+#endif
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the Glyph ID for the given Unicode ID in the Microsoft UCS-4 subtable.
+    (Actually this code only depends on subtable being format 12.)
+    Return 0 if the Unicode ID is not in the subtable.
+----------------------------------------------------------------------------------------------*/
+gid16 CmapSubtable12Lookup(const void * pCmap310, unsigned int uUnicodeId, int rangeKey)
+{
+    const Sfnt::CmapSubTableFormat12 * pTable = reinterpret_cast<const Sfnt::CmapSubTableFormat12 *>(pCmap310);
+
+    //uint32 uLength = be::swap(pTable->length); //could use to test for premature end of table
+    uint32 ucGroups = be::swap(pTable->num_groups);
+
+    for (unsigned int i = rangeKey; i < ucGroups; i++)
+    {
+        uint32 uStartCode = be::swap(pTable->group[i].start_char_code);
+        uint32 uEndCode = be::swap(pTable->group[i].end_char_code);
+        if (uUnicodeId >= uStartCode && uUnicodeId <= uEndCode)
+        {
+            uint32 uDiff = uUnicodeId - uStartCode;
+            uint32 uStartGid = be::swap(pTable->group[i].start_glyph_id);
+            return static_cast<gid16>(uStartGid + uDiff);
+        }
+    }
+
+    return 0;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the next Unicode value in the cmap. Pass 0 to obtain the first item.
+    Returns 0x10FFFF as the last item.
+    pRangeKey is an optional key that is used to optimize the search; its value is the range
+    in which the character is found.
+----------------------------------------------------------------------------------------------*/
+unsigned int CmapSubtable12NextCodepoint(const void *pCmap310, unsigned int nUnicodeId, int * pRangeKey)
+{
+    const Sfnt::CmapSubTableFormat12 * pTable = reinterpret_cast<const Sfnt::CmapSubTableFormat12 *>(pCmap310);
+
+    int nRange = be::swap(pTable->num_groups);
+
+    uint32 nUnicodePrev = (uint32)nUnicodeId;
+
+    if (nUnicodePrev == 0)
+    {
+        // return the first codepoint.
+        if (pRangeKey)
+            *pRangeKey = 0;
+        return be::swap(pTable->group[0].start_char_code);
+    }
+    else if (nUnicodePrev >= 0x10FFFF)
+    {
+        if (pRangeKey)
+            *pRangeKey = nRange;
+        return 0x10FFFF;
+    }
+
+    int iRange = (pRangeKey) ? *pRangeKey : 0;
+    // Just in case we have a bad key:
+    while (iRange > 0 && be::swap(pTable->group[iRange].start_char_code) > nUnicodePrev)
+        iRange--;
+    while (iRange < nRange - 1 && be::swap(pTable->group[iRange].end_char_code) < nUnicodePrev)
+        iRange++;
+
+    // Now iRange is the range containing nUnicodePrev.
+
+    unsigned int nStartCode = be::swap(pTable->group[iRange].start_char_code);
+    unsigned int nEndCode = be::swap(pTable->group[iRange].end_char_code);
+
+    if (nStartCode > nUnicodePrev)
+        // Oops, nUnicodePrev is not in the cmap! Adjust so we get a reasonable
+        // answer this time around.
+        nUnicodePrev = nStartCode - 1;
+
+    if (nEndCode > nUnicodePrev)
+    {
+        // Next is in the same range; it is the next successive codepoint.
+        if (pRangeKey)
+            *pRangeKey = iRange;
+        return nUnicodePrev + 1;
+    }
+
+    // Otherwise the next codepoint is the first one in the next range, or 10FFFF if we're done.
+    if (pRangeKey)
+        *pRangeKey = iRange + 1;
+    return (iRange + 1 >= nRange) ? 0x10FFFF : be::swap(pTable->group[iRange + 1].start_char_code);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return the offset stored in the loca table for the given Glyph ID.
+    (This offset is into the glyf table.)
+    Return -1 if the lookup failed.
+    Technically this method should return an unsigned long but it is unlikely the offset will
+        exceed 2^31.
+----------------------------------------------------------------------------------------------*/
+size_t LocaLookup(gid16 nGlyphId,
+        const void * pLoca, size_t lLocaSize,
+        const void * pHead) // throw (std::out_of_range)
+{
+    const Sfnt::FontHeader * pTable = reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+    size_t res = -2;
+
+    // CheckTable verifies the index_to_loc_format is valid
+    if (be::swap(pTable->index_to_loc_format) == Sfnt::FontHeader::ShortIndexLocFormat)
+    { // loca entries are two bytes and have been divided by two
+        if (lLocaSize > 1 && nGlyphId + 1u < lLocaSize >> 1) // allow sentinel value to be accessed
+        {
+            const uint16 * pShortTable = reinterpret_cast<const uint16 *>(pLoca);
+            res = be::peek<uint16>(pShortTable + nGlyphId) << 1;
+            if (res == static_cast<size_t>(be::peek<uint16>(pShortTable + nGlyphId + 1) << 1))
+                return -1;
+        }
+    }
+    else if (be::swap(pTable->index_to_loc_format) == Sfnt::FontHeader::LongIndexLocFormat)
+    { // loca entries are four bytes
+        if (lLocaSize > 3 && nGlyphId + 1u < lLocaSize >> 2)
+        {
+            const uint32 * pLongTable = reinterpret_cast<const uint32 *>(pLoca);
+            res = be::peek<uint32>(pLongTable + nGlyphId);
+            if (res == static_cast<size_t>(be::peek<uint32>(pLongTable + nGlyphId + 1)))
+                return -1;
+        }
+    }
+
+    // only get here if glyph id was bad
+    return res;
+    //throw std::out_of_range("glyph id out of range for font");
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return a pointer into the glyf table based on the given offset (from LocaLookup).
+    Return NULL on error.
+----------------------------------------------------------------------------------------------*/
+void * GlyfLookup(const void * pGlyf, size_t nGlyfOffset, size_t nTableLen)
+{
+    const uint8 * pByte = reinterpret_cast<const uint8 *>(pGlyf);
+    if (OVERFLOW_OFFSET_CHECK(pByte, nGlyfOffset) || nGlyfOffset >= nTableLen - sizeof(Sfnt::Glyph))
+        return NULL;
+    return const_cast<uint8 *>(pByte + nGlyfOffset);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the bounding box coordinates for a simple glyf entry (non-composite).
+    Return true if successful, false otherwise.
+----------------------------------------------------------------------------------------------*/
+bool GlyfBox(const void * pSimpleGlyf, int & xMin, int & yMin,
+                      int & xMax, int & yMax)
+{
+    const Sfnt::Glyph * pGlyph = reinterpret_cast<const Sfnt::Glyph *>(pSimpleGlyf);
+
+    xMin = be::swap(pGlyph->x_min);
+    yMin = be::swap(pGlyph->y_min);
+    xMax = be::swap(pGlyph->x_max);
+    yMax = be::swap(pGlyph->y_max);
+
+    return true;
+}
+
+#ifdef ALL_TTFUTILS
+/*----------------------------------------------------------------------------------------------
+    Return the number of contours for a simple glyf entry (non-composite)
+    Returning -1 means this is a composite glyph
+----------------------------------------------------------------------------------------------*/
+int GlyfContourCount(const void * pSimpleGlyf)
+{
+    const Sfnt::Glyph * pGlyph = reinterpret_cast<const Sfnt::Glyph *>(pSimpleGlyf);
+    return be::swap(pGlyph->number_of_contours); // -1 means composite glyph
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the point numbers for the end points of the glyph contours for a simple
+    glyf entry (non-composite).
+    cnPointsTotal - count of contours from GlyfContourCount(); (same as number of end points)
+    prgnContourEndPoints - should point to a buffer large enough to hold cnPoints integers
+    cnPoints - count of points placed in above range
+    Return true if successful, false otherwise.
+        False could indicate a multi-level composite glyphs.
+----------------------------------------------------------------------------------------------*/
+bool GlyfContourEndPoints(const void * pSimpleGlyf, int * prgnContourEndPoint,
+                                   int cnPointsTotal, int & cnPoints)
+{
+    const Sfnt::SimpleGlyph * pGlyph = reinterpret_cast<const Sfnt::SimpleGlyph *>(pSimpleGlyf);
+
+    int cContours = be::swap(pGlyph->number_of_contours);
+    if (cContours < 0)
+        return false; // this method isn't supposed handle composite glyphs
+
+    for (int i = 0; i < cContours && i < cnPointsTotal; i++)
+    {
+        prgnContourEndPoint[i] = be::swap(pGlyph->end_pts_of_contours[i]);
+    }
+
+    cnPoints = cContours;
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the points for a simple glyf entry (non-composite)
+    cnPointsTotal - count of points from largest end point obtained from GlyfContourEndPoints
+    prgnX & prgnY - should point to buffers large enough to hold cnPointsTotal integers
+        The ranges are parallel so that coordinates for point(n) are found at offset n in both
+        ranges. This is raw point data with relative coordinates.
+    prgbFlag - should point to a buffer a large enough to hold cnPointsTotal bytes
+        This range is parallel to the prgnX & prgnY
+    cnPoints - count of points placed in above ranges
+    Return true if successful, false otherwise.
+        False could indicate a composite glyph
+----------------------------------------------------------------------------------------------*/
+bool GlyfPoints(const void * pSimpleGlyf, int * prgnX, int * prgnY,
+        char * prgbFlag, int cnPointsTotal, int & cnPoints)
+{
+    using namespace Sfnt;
+
+    const Sfnt::SimpleGlyph * pGlyph = reinterpret_cast<const Sfnt::SimpleGlyph *>(pSimpleGlyf);
+    int cContours = be::swap(pGlyph->number_of_contours);
+    // return false for composite glyph
+    if (cContours <= 0)
+        return false;
+    int cPts = be::swap(pGlyph->end_pts_of_contours[cContours - 1]) + 1;
+    if (cPts > cnPointsTotal)
+        return false;
+
+    // skip over bounding box data & point to byte count of instructions (hints)
+    const uint8 * pbGlyph = reinterpret_cast<const uint8 *>
+        (&pGlyph->end_pts_of_contours[cContours]);
+
+    // skip over hints & point to first flag
+    int cbHints = be::swap(*(uint16 *)pbGlyph);
+    pbGlyph += sizeof(uint16);
+    pbGlyph += cbHints;
+
+    // load flags & point to first x coordinate
+    int iFlag = 0;
+    while (iFlag < cPts)
+    {
+        if (!(*pbGlyph & SimpleGlyph::Repeat))
+        { // flag isn't repeated
+            prgbFlag[iFlag] = (char)*pbGlyph;
+            pbGlyph++;
+            iFlag++;
+        }
+        else
+        { // flag is repeated; count specified by next byte
+            char chFlag = (char)*pbGlyph;
+            pbGlyph++;
+            int cFlags = (int)*pbGlyph;
+            pbGlyph++;
+            prgbFlag[iFlag] = chFlag;
+            iFlag++;
+            for (int i = 0; i < cFlags; i++)
+            {
+                prgbFlag[iFlag + i] = chFlag;
+            }
+            iFlag += cFlags;
+        }
+    }
+    if (iFlag != cPts)
+        return false;
+
+    // load x coordinates
+    iFlag = 0;
+    while (iFlag < cPts)
+    {
+        if (prgbFlag[iFlag] & SimpleGlyph::XShort)
+        {
+            prgnX[iFlag] = *pbGlyph;
+            if (!(prgbFlag[iFlag] & SimpleGlyph::XIsPos))
+            {
+                prgnX[iFlag] = -prgnX[iFlag];
+            }
+            pbGlyph++;
+        }
+        else
+        {
+            if (prgbFlag[iFlag] & SimpleGlyph::XIsSame)
+            {
+                prgnX[iFlag] = 0;
+                // do NOT increment pbGlyph
+            }
+            else
+            {
+                prgnX[iFlag] = be::swap(*(int16 *)pbGlyph);
+                pbGlyph += sizeof(int16);
+            }
+        }
+        iFlag++;
+    }
+
+    // load y coordinates
+    iFlag = 0;
+    while (iFlag < cPts)
+    {
+        if (prgbFlag[iFlag] & SimpleGlyph::YShort)
+        {
+            prgnY[iFlag] = *pbGlyph;
+            if (!(prgbFlag[iFlag] & SimpleGlyph::YIsPos))
+            {
+                prgnY[iFlag] = -prgnY[iFlag];
+            }
+            pbGlyph++;
+        }
+        else
+        {
+            if (prgbFlag[iFlag] & SimpleGlyph::YIsSame)
+            {
+                prgnY[iFlag] = 0;
+                // do NOT increment pbGlyph
+            }
+            else
+            {
+                prgnY[iFlag] = be::swap(*(int16 *)pbGlyph);
+                pbGlyph += sizeof(int16);
+            }
+        }
+        iFlag++;
+    }
+
+    cnPoints = cPts;
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Fill prgnCompId with the component Glyph IDs from pSimpleGlyf.
+    Client must allocate space before calling.
+    pSimpleGlyf - assumed to point to a composite glyph
+    cCompIdTotal - the number of elements in prgnCompId
+    cCompId  - the total number of Glyph IDs stored in prgnCompId
+    Return true if successful, false otherwise
+        False could indicate a non-composite glyph or the input array was not big enough
+----------------------------------------------------------------------------------------------*/
+bool GetComponentGlyphIds(const void * pSimpleGlyf, int * prgnCompId,
+        size_t cnCompIdTotal, size_t & cnCompId)
+{
+    using namespace Sfnt;
+
+    if (GlyfContourCount(pSimpleGlyf) >= 0)
+        return false;
+
+    const Sfnt::SimpleGlyph * pGlyph = reinterpret_cast<const Sfnt::SimpleGlyph *>(pSimpleGlyf);
+    // for a composite glyph, the special data begins here
+    const uint8 * pbGlyph = reinterpret_cast<const uint8 *>(&pGlyph->end_pts_of_contours[0]);
+
+    uint16 GlyphFlags;
+    size_t iCurrentComp = 0;
+    do
+    {
+        GlyphFlags = be::swap(*((uint16 *)pbGlyph));
+        pbGlyph += sizeof(uint16);
+        prgnCompId[iCurrentComp++] = be::swap(*((uint16 *)pbGlyph));
+        pbGlyph += sizeof(uint16);
+        if (iCurrentComp >= cnCompIdTotal)
+            return false;
+        int nOffset = 0;
+        nOffset += GlyphFlags & CompoundGlyph::Arg1Arg2Words ? 4 : 2;
+        nOffset += GlyphFlags & CompoundGlyph::HaveScale ? 2 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveXAndYScale  ? 4 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveTwoByTwo  ? 8 :  0;
+        pbGlyph += nOffset;
+    } while (GlyphFlags & CompoundGlyph::MoreComponents);
+
+    cnCompId = iCurrentComp;
+
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return info on how a component glyph is to be placed
+    pSimpleGlyph - assumed to point to a composite glyph
+    nCompId - glyph id for component of interest
+    bOffset - if true, a & b are the x & y offsets for this component
+              if false, b is the point on this component that is attaching to point a on the
+                preceding glyph
+    Return true if successful, false otherwise
+        False could indicate a non-composite glyph or that component wasn't found
+----------------------------------------------------------------------------------------------*/
+bool GetComponentPlacement(const void * pSimpleGlyf, int nCompId,
+                                    bool fOffset, int & a, int & b)
+{
+    using namespace Sfnt;
+
+    if (GlyfContourCount(pSimpleGlyf) >= 0)
+        return false;
+
+    const Sfnt::SimpleGlyph * pGlyph = reinterpret_cast<const Sfnt::SimpleGlyph *>(pSimpleGlyf);
+    // for a composite glyph, the special data begins here
+    const uint8 * pbGlyph = reinterpret_cast<const uint8 *>(&pGlyph->end_pts_of_contours[0]);
+
+    uint16 GlyphFlags;
+    do
+    {
+        GlyphFlags = be::swap(*((uint16 *)pbGlyph));
+        pbGlyph += sizeof(uint16);
+        if (be::swap(*((uint16 *)pbGlyph)) == nCompId)
+        {
+            pbGlyph += sizeof(uint16); // skip over glyph id of component
+            fOffset = (GlyphFlags & CompoundGlyph::ArgsAreXYValues) == CompoundGlyph::ArgsAreXYValues;
+
+            if (GlyphFlags & CompoundGlyph::Arg1Arg2Words )
+            {
+                a = be::swap(*(int16 *)pbGlyph);
+                pbGlyph += sizeof(int16);
+                b = be::swap(*(int16 *)pbGlyph);
+                pbGlyph += sizeof(int16);
+            }
+            else
+            { // args are signed bytes
+                a = *pbGlyph++;
+                b = *pbGlyph++;
+            }
+            return true;
+        }
+        pbGlyph += sizeof(uint16); // skip over glyph id of component
+        int nOffset = 0;
+        nOffset += GlyphFlags & CompoundGlyph::Arg1Arg2Words  ? 4 : 2;
+        nOffset += GlyphFlags & CompoundGlyph::HaveScale ? 2 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveXAndYScale  ? 4 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveTwoByTwo  ? 8 :  0;
+        pbGlyph += nOffset;
+    } while (GlyphFlags & CompoundGlyph::MoreComponents);
+
+    // didn't find requested component
+    fOffset = true;
+    a = 0;
+    b = 0;
+    return false;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Return info on how a component glyph is to be transformed
+    pSimpleGlyph - assumed to point to a composite glyph
+    nCompId - glyph id for component of interest
+    flt11, flt11, flt11, flt11 - a 2x2 matrix giving the transform
+    bTransOffset - whether to transform the offset from above method
+        The spec is unclear about the meaning of this flag
+        Currently - initialize to true for MS rasterizer and false for Mac rasterizer, then
+            on return it will indicate whether transform should apply to offset (MSDN CD 10/99)
+    Return true if successful, false otherwise
+        False could indicate a non-composite glyph or that component wasn't found
+----------------------------------------------------------------------------------------------*/
+bool GetComponentTransform(const void * pSimpleGlyf, int nCompId,
+                                    float & flt11, float & flt12, float & flt21, float & flt22,
+                                    bool & fTransOffset)
+{
+    using namespace Sfnt;
+
+    if (GlyfContourCount(pSimpleGlyf) >= 0)
+        return false;
+
+    const Sfnt::SimpleGlyph * pGlyph = reinterpret_cast<const Sfnt::SimpleGlyph *>(pSimpleGlyf);
+    // for a composite glyph, the special data begins here
+    const uint8 * pbGlyph = reinterpret_cast<const uint8 *>(&pGlyph->end_pts_of_contours[0]);
+
+    uint16 GlyphFlags;
+    do
+    {
+        GlyphFlags = be::swap(*((uint16 *)pbGlyph));
+        pbGlyph += sizeof(uint16);
+        if (be::swap(*((uint16 *)pbGlyph)) == nCompId)
+        {
+            pbGlyph += sizeof(uint16); // skip over glyph id of component
+            pbGlyph += GlyphFlags & CompoundGlyph::Arg1Arg2Words  ? 4 : 2; // skip over placement data
+
+            if (fTransOffset) // MS rasterizer
+                fTransOffset = !(GlyphFlags & CompoundGlyph::UnscaledOffset);
+            else // Apple rasterizer
+                fTransOffset = (GlyphFlags & CompoundGlyph::ScaledOffset) != 0;
+
+            if (GlyphFlags & CompoundGlyph::HaveScale)
+            {
+                flt11 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+                flt12 = 0;
+                flt21 = 0;
+                flt22 = flt11;
+            }
+            else if (GlyphFlags & CompoundGlyph::HaveXAndYScale)
+            {
+                flt11 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+                flt12 = 0;
+                flt21 = 0;
+                flt22 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+            }
+            else if (GlyphFlags & CompoundGlyph::HaveTwoByTwo)
+            {
+                flt11 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+                flt12 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+                flt21 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+                flt22 = fixed_to_float<14>(be::swap(*(uint16 *)pbGlyph));
+                pbGlyph += sizeof(uint16);
+            }
+            else
+            { // identity transform
+                flt11 = 1.0;
+                flt12 = 0.0;
+                flt21 = 0.0;
+                flt22 = 1.0;
+            }
+            return true;
+        }
+        pbGlyph += sizeof(uint16); // skip over glyph id of component
+        int nOffset = 0;
+        nOffset += GlyphFlags & CompoundGlyph::Arg1Arg2Words  ? 4 : 2;
+        nOffset += GlyphFlags & CompoundGlyph::HaveScale ? 2 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveXAndYScale  ? 4 : 0;
+        nOffset += GlyphFlags & CompoundGlyph::HaveTwoByTwo  ? 8 :  0;
+        pbGlyph += nOffset;
+    } while (GlyphFlags & CompoundGlyph::MoreComponents);
+
+    // didn't find requested component
+    fTransOffset = false;
+    flt11 = 1;
+    flt12 = 0;
+    flt21 = 0;
+    flt22 = 1;
+    return false;
+}
+#endif
+
+/*----------------------------------------------------------------------------------------------
+    Return a pointer into the glyf table based on the given tables and Glyph ID
+    Since this method doesn't check for spaces, it is good to call IsSpace before using it.
+    Return NULL on error.
+----------------------------------------------------------------------------------------------*/
+void * GlyfLookup(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+                           size_t lGlyfSize, size_t lLocaSize, const void * pHead)
+{
+    // test for valid glyph id
+    // CheckTable verifies the index_to_loc_format is valid
+
+    const Sfnt::FontHeader * pTable
+        = reinterpret_cast<const Sfnt::FontHeader *>(pHead);
+
+    if (be::swap(pTable->index_to_loc_format) == Sfnt::FontHeader::ShortIndexLocFormat)
+    { // loca entries are two bytes (and have been divided by two)
+        if (nGlyphId >= (lLocaSize >> 1) - 1) // don't allow nGlyphId to access sentinel
+        {
+//          throw std::out_of_range("glyph id out of range for font");
+            return NULL;
+        }
+    }
+    if (be::swap(pTable->index_to_loc_format) == Sfnt::FontHeader::LongIndexLocFormat)
+    { // loca entries are four bytes
+        if (nGlyphId >= (lLocaSize >> 2) - 1)
+        {
+//          throw std::out_of_range("glyph id out of range for font");
+            return NULL;
+        }
+    }
+
+    size_t lGlyfOffset = LocaLookup(nGlyphId, pLoca, lLocaSize, pHead);
+    void * pSimpleGlyf = GlyfLookup(pGlyf, lGlyfOffset, lGlyfSize); // invalid loca offset returns null
+    return pSimpleGlyf;
+}
+
+#ifdef ALL_TTFUTILS
+/*----------------------------------------------------------------------------------------------
+    Determine if a particular Glyph ID has any data in the glyf table. If it is white space,
+    there will be no glyf data, though there will be metric data in hmtx, etc.
+----------------------------------------------------------------------------------------------*/
+bool IsSpace(gid16 nGlyphId, const void * pLoca, size_t lLocaSize, const void * pHead)
+{
+    size_t lGlyfOffset = LocaLookup(nGlyphId, pLoca, lLocaSize, pHead);
+
+    // the +1 should always work because there is a sentinel value at the end of the loca table
+    size_t lNextGlyfOffset = LocaLookup(nGlyphId + 1, pLoca, lLocaSize, pHead);
+
+    return (lNextGlyfOffset - lGlyfOffset) == 0;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Determine if a particular Glyph ID is a multi-level composite.
+----------------------------------------------------------------------------------------------*/
+bool IsDeepComposite(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+                    size_t lGlyfSize, long lLocaSize, const void * pHead)
+{
+    if (IsSpace(nGlyphId, pLoca, lLocaSize, pHead)) {return false;}
+
+    void * pSimpleGlyf = GlyfLookup(nGlyphId, pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+    if (pSimpleGlyf == NULL)
+        return false; // no way to really indicate an error occured here
+
+    if (GlyfContourCount(pSimpleGlyf) >= 0)
+        return false;
+
+    int rgnCompId[kMaxGlyphComponents]; // assumes only a limited number of glyph components
+    size_t cCompIdTotal = kMaxGlyphComponents;
+    size_t cCompId = 0;
+
+    if (!GetComponentGlyphIds(pSimpleGlyf, rgnCompId, cCompIdTotal, cCompId))
+        return false;
+
+    for (size_t i = 0; i < cCompId; i++)
+    {
+        pSimpleGlyf = GlyfLookup(static_cast<gid16>(rgnCompId[i]),
+                            pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+        if (pSimpleGlyf == NULL) {return false;}
+
+        if (GlyfContourCount(pSimpleGlyf) < 0)
+            return true;
+    }
+
+    return false;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the bounding box coordinates based on the given tables and Glyph ID
+    Handles both simple and composite glyphs.
+    Return true if successful, false otherwise. On false, all point values will be INT_MIN
+        False may indicate a white space glyph
+----------------------------------------------------------------------------------------------*/
+bool GlyfBox(gid16  nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void * pHead, int & xMin, int & yMin, int & xMax, int & yMax)
+{
+    xMin = yMin = xMax = yMax = INT_MIN;
+
+    if (IsSpace(nGlyphId, pLoca, lLocaSize, pHead)) {return false;}
+
+    void * pSimpleGlyf = GlyfLookup(nGlyphId, pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+    if (pSimpleGlyf == NULL) {return false;}
+
+    return GlyfBox(pSimpleGlyf, xMin, yMin, xMax, yMax);
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the number of contours based on the given tables and Glyph ID
+    Handles both simple and composite glyphs.
+    Return true if successful, false otherwise. On false, cnContours will be INT_MIN
+        False may indicate a white space glyph or a multi-level composite glyph.
+----------------------------------------------------------------------------------------------*/
+bool GlyfContourCount(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+    size_t lGlyfSize, size_t lLocaSize, const void * pHead, size_t & cnContours)
+{
+    cnContours = static_cast<size_t>(INT_MIN);
+
+    if (IsSpace(nGlyphId, pLoca, lLocaSize, pHead)) {return false;}
+
+    void * pSimpleGlyf = GlyfLookup(nGlyphId, pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+    if (pSimpleGlyf == NULL) {return false;}
+
+    int cRtnContours = GlyfContourCount(pSimpleGlyf);
+    if (cRtnContours >= 0)
+    {
+        cnContours = size_t(cRtnContours);
+        return true;
+    }
+
+    //handle composite glyphs
+
+    int rgnCompId[kMaxGlyphComponents]; // assumes no glyph will be made of more than 8 components
+    size_t cCompIdTotal = kMaxGlyphComponents;
+    size_t cCompId = 0;
+
+    if (!GetComponentGlyphIds(pSimpleGlyf, rgnCompId, cCompIdTotal, cCompId))
+        return false;
+
+    cRtnContours = 0;
+    int cTmp = 0;
+    for (size_t i = 0; i < cCompId; i++)
+    {
+        if (IsSpace(static_cast<gid16>(rgnCompId[i]), pLoca, lLocaSize, pHead)) {return false;}
+        pSimpleGlyf = GlyfLookup(static_cast<gid16>(rgnCompId[i]),
+                                 pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+        if (pSimpleGlyf == 0) {return false;}
+        // return false on multi-level composite
+        if ((cTmp = GlyfContourCount(pSimpleGlyf)) < 0)
+            return false;
+        cRtnContours += cTmp;
+    }
+
+    cnContours = size_t(cRtnContours);
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the point numbers for the end points of the glyph contours based on the given tables
+    and Glyph ID
+    Handles both simple and composite glyphs.
+    cnPoints - count of contours from GlyfContourCount (same as number of end points)
+    prgnContourEndPoints - should point to a buffer large enough to hold cnPoints integers
+    Return true if successful, false otherwise. On false, all end points are INT_MIN
+        False may indicate a white space glyph or a multi-level composite glyph.
+----------------------------------------------------------------------------------------------*/
+bool GlyfContourEndPoints(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+    size_t lGlyfSize, size_t lLocaSize, const void * pHead,
+    int * prgnContourEndPoint, size_t cnPoints)
+{
+        memset(prgnContourEndPoint, 0xFF, cnPoints * sizeof(int));
+    // std::fill_n(prgnContourEndPoint, cnPoints, INT_MIN);
+
+    if (IsSpace(nGlyphId, pLoca, lLocaSize, pHead)) {return false;}
+
+    void * pSimpleGlyf = GlyfLookup(nGlyphId, pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+    if (pSimpleGlyf == NULL) {return false;}
+
+    int cContours = GlyfContourCount(pSimpleGlyf);
+    int cActualPts = 0;
+    if (cContours > 0)
+        return GlyfContourEndPoints(pSimpleGlyf, prgnContourEndPoint, cnPoints, cActualPts);
+
+    // handle composite glyphs
+
+    int rgnCompId[kMaxGlyphComponents]; // assumes no glyph will be made of more than 8 components
+    size_t cCompIdTotal = kMaxGlyphComponents;
+    size_t cCompId = 0;
+
+    if (!GetComponentGlyphIds(pSimpleGlyf, rgnCompId, cCompIdTotal, cCompId))
+        return false;
+
+    int * prgnCurrentEndPoint = prgnContourEndPoint;
+    int cCurrentPoints = cnPoints;
+    int nPrevPt = 0;
+    for (size_t i = 0; i < cCompId; i++)
+    {
+        if (IsSpace(static_cast<gid16>(rgnCompId[i]), pLoca, lLocaSize, pHead)) {return false;}
+        pSimpleGlyf = GlyfLookup(static_cast<gid16>(rgnCompId[i]), pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+        if (pSimpleGlyf == NULL) {return false;}
+        // returns false on multi-level composite
+        if (!GlyfContourEndPoints(pSimpleGlyf, prgnCurrentEndPoint, cCurrentPoints, cActualPts))
+            return false;
+        // points in composite are numbered sequentially as components are added
+        //  must adjust end point numbers for new point numbers
+        for (int j = 0; j < cActualPts; j++)
+            prgnCurrentEndPoint[j] += nPrevPt;
+        nPrevPt = prgnCurrentEndPoint[cActualPts - 1] + 1;
+
+        prgnCurrentEndPoint += cActualPts;
+        cCurrentPoints -= cActualPts;
+    }
+
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Get the points for a glyph based on the given tables and Glyph ID
+    Handles both simple and composite glyphs.
+    cnPoints - count of points from largest end point obtained from GlyfContourEndPoints
+    prgnX & prgnY - should point to buffers large enough to hold cnPoints integers
+        The ranges are parallel so that coordinates for point(n) are found at offset n in
+        both ranges. These points are in absolute coordinates.
+    prgfOnCurve - should point to a buffer a large enough to hold cnPoints bytes (bool)
+        This range is parallel to the prgnX & prgnY
+    Return true if successful, false otherwise. On false, all points may be INT_MIN
+        False may indicate a white space glyph, a multi-level composite, or a corrupt font
+        It's not clear from the TTF spec when the transforms should be applied. Should the
+        transform be done before or after attachment point calcs? (current code - before)
+        Should the transform be applied to other offsets? (currently - no; however commented
+        out code is in place so that if CompoundGlyph::UnscaledOffset on the MS rasterizer is
+        clear (typical) then yes, and if CompoundGlyph::ScaledOffset on the Apple rasterizer is
+        clear (typical?) then no). See GetComponentTransform.
+        It's also unclear where point numbering with attachment poinst starts
+        (currently - first point number is relative to whole glyph, second point number is
+        relative to current glyph).
+----------------------------------------------------------------------------------------------*/
+bool GlyfPoints(gid16 nGlyphId, const void * pGlyf,
+        const void * pLoca, size_t lGlyfSize, size_t lLocaSize, const void * pHead,
+        const int * /*prgnContourEndPoint*/, size_t /*cnEndPoints*/,
+        int * prgnX, int * prgnY, bool * prgfOnCurve, size_t cnPoints)
+{
+        memset(prgnX, 0x7F, cnPoints * sizeof(int));
+        memset(prgnY, 0x7F, cnPoints * sizeof(int));
+
+    if (IsSpace(nGlyphId, pLoca, lLocaSize, pHead))
+        return false;
+
+    void * pSimpleGlyf = GlyfLookup(nGlyphId, pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+    if (pSimpleGlyf == NULL)
+        return false;
+
+    int cContours = GlyfContourCount(pSimpleGlyf);
+    int cActualPts;
+    if (cContours > 0)
+    {
+        if (!GlyfPoints(pSimpleGlyf, prgnX, prgnY, (char *)prgfOnCurve, cnPoints, cActualPts))
+            return false;
+        CalcAbsolutePoints(prgnX, prgnY, cnPoints);
+        SimplifyFlags((char *)prgfOnCurve, cnPoints);
+        return true;
+    }
+
+    // handle composite glyphs
+    int rgnCompId[kMaxGlyphComponents]; // assumes no glyph will be made of more than 8 components
+    size_t cCompIdTotal = kMaxGlyphComponents;
+    size_t cCompId = 0;
+
+    // this will fail if there are more components than there is room for
+    if (!GetComponentGlyphIds(pSimpleGlyf, rgnCompId, cCompIdTotal, cCompId))
+        return false;
+
+    int * prgnCurrentX = prgnX;
+    int * prgnCurrentY = prgnY;
+    char * prgbCurrentFlag = (char *)prgfOnCurve; // converting bool to char should be safe
+    int cCurrentPoints = cnPoints;
+    bool fOffset = true, fTransOff = true;
+    int a, b;
+    float flt11, flt12, flt21, flt22;
+    // int * prgnPrevX = prgnX; // in case first att pt number relative to preceding glyph
+    // int * prgnPrevY = prgnY;
+    for (size_t i = 0; i < cCompId; i++)
+    {
+        if (IsSpace(static_cast<gid16>(rgnCompId[i]), pLoca, lLocaSize, pHead)) {return false;}
+        void * pCompGlyf = GlyfLookup(static_cast<gid16>(rgnCompId[i]), pGlyf, pLoca, lGlyfSize, lLocaSize, pHead);
+        if (pCompGlyf == NULL) {return false;}
+        // returns false on multi-level composite
+        if (!GlyfPoints(pCompGlyf, prgnCurrentX, prgnCurrentY, prgbCurrentFlag,
+            cCurrentPoints, cActualPts))
+            return false;
+        if (!GetComponentPlacement(pSimpleGlyf, rgnCompId[i], fOffset, a, b))
+            return false;
+        if (!GetComponentTransform(pSimpleGlyf, rgnCompId[i],
+            flt11, flt12, flt21, flt22, fTransOff))
+            return false;
+        bool fIdTrans = flt11 == 1.0 && flt12 == 0.0 && flt21 == 0.0 && flt22 == 1.0;
+
+        // convert points to absolute coordinates
+        // do before transform and attachment point placement are applied
+        CalcAbsolutePoints(prgnCurrentX, prgnCurrentY, cActualPts);
+
+        // apply transform - see main method note above
+        // do before attachment point calcs
+        if (!fIdTrans)
+            for (int j = 0; j < cActualPts; j++)
+            {
+                int x = prgnCurrentX[j]; // store before transform applied
+                int y = prgnCurrentY[j];
+                prgnCurrentX[j] = (int)(x * flt11 + y * flt12);
+                prgnCurrentY[j] = (int)(x * flt21 + y * flt22);
+            }
+
+        // apply placement - see main method note above
+        int nXOff, nYOff;
+        if (fOffset) // explicit x & y offsets
+        {
+            /* ignore fTransOff for now
+            if (fTransOff && !fIdTrans)
+            {   // transform x & y offsets
+                nXOff = (int)(a * flt11 + b * flt12);
+                nYOff = (int)(a * flt21 + b * flt22);
+            }
+            else */
+            { // don't transform offset
+                nXOff = a;
+                nYOff = b;
+            }
+        }
+        else  // attachment points
+        {   // in case first point is relative to preceding glyph and second relative to current
+            // nXOff = prgnPrevX[a] - prgnCurrentX[b];
+            // nYOff = prgnPrevY[a] - prgnCurrentY[b];
+            // first point number relative to whole composite, second relative to current glyph
+            nXOff = prgnX[a] - prgnCurrentX[b];
+            nYOff = prgnY[a] - prgnCurrentY[b];
+        }
+        for (int j = 0; j < cActualPts; j++)
+        {
+            prgnCurrentX[j] += nXOff;
+            prgnCurrentY[j] += nYOff;
+        }
+
+        // prgnPrevX = prgnCurrentX;
+        // prgnPrevY = prgnCurrentY;
+        prgnCurrentX += cActualPts;
+        prgnCurrentY += cActualPts;
+        prgbCurrentFlag += cActualPts;
+        cCurrentPoints -= cActualPts;
+    }
+
+    SimplifyFlags((char *)prgfOnCurve, cnPoints);
+
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Simplify the meaning of flags to just indicate whether point is on-curve or off-curve.
+---------------------------------------------------------------------------------------------*/
+bool SimplifyFlags(char * prgbFlags, int cnPoints)
+{
+    for (int i = 0; i < cnPoints; i++)
+        prgbFlags[i] = static_cast<char>(prgbFlags[i] & Sfnt::SimpleGlyph::OnCurve);
+    return true;
+}
+
+/*----------------------------------------------------------------------------------------------
+    Convert relative point coordinates to absolute coordinates
+    Points are stored in the font such that they are offsets from one another except for the
+        first point of a glyph.
+---------------------------------------------------------------------------------------------*/
+bool CalcAbsolutePoints(int * prgnX, int * prgnY, int cnPoints)
+{
+    int nX = prgnX[0];
+    int nY = prgnY[0];
+    for (int i = 1; i < cnPoints; i++)
+    {
+        prgnX[i] += nX;
+        nX = prgnX[i];
+        prgnY[i] += nY;
+        nY = prgnY[i];
+    }
+
+    return true;
+}
+#endif
+
+/*----------------------------------------------------------------------------------------------
+    Return the length of the 'name' table in bytes.
+    Currently used.
+---------------------------------------------------------------------------------------------*/
+#if 0
+size_t NameTableLength(const byte * pTable)
+{
+    byte * pb = (const_cast<byte *>(pTable)) + 2; // skip format
+    size_t cRecords = *pb++ << 8; cRecords += *pb++;
+    int dbStringOffset0 = (*pb++) << 8; dbStringOffset0 += *pb++;
+    int dbMaxStringOffset = 0;
+    for (size_t irec = 0; irec < cRecords; irec++)
+    {
+        int nPlatform = (*pb++) << 8; nPlatform += *pb++;
+        int nEncoding = (*pb++) << 8; nEncoding += *pb++;
+        int nLanguage = (*pb++) << 8; nLanguage += *pb++;
+        int nName = (*pb++) << 8; nName += *pb++;
+        int cbStringLen = (*pb++) << 8; cbStringLen += *pb++;
+        int dbStringOffset = (*pb++) << 8; dbStringOffset += *pb++;
+        if (dbMaxStringOffset < dbStringOffset + cbStringLen)
+            dbMaxStringOffset = dbStringOffset + cbStringLen;
+    }
+    return dbStringOffset0 + dbMaxStringOffset;
+}
+#endif
+
+} // end of namespace TtfUtil
+} // end of namespace graphite
diff --git a/thirdparty/graphite/src/UtfCodec.cpp b/thirdparty/graphite/src/UtfCodec.cpp
new file mode 100644
index 0000000000..a944bbf9d0
--- /dev/null
+++ b/thirdparty/graphite/src/UtfCodec.cpp
@@ -0,0 +1,45 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "inc/UtfCodec.h"
+//using namespace graphite2;
+
+namespace graphite2 {
+
+}
+
+using namespace graphite2;
+
+const int8 _utf_codec<8>::sz_lut[16] =
+{
+        1,1,1,1,1,1,1,1,    // 1 byte
+        0,0,0,0,            // trailing byte
+        2,2,                // 2 bytes
+        3,                  // 3 bytes
+        4                   // 4 bytes
+};
+
+const byte  _utf_codec<8>::mask_lut[5] = {0x7f, 0xff, 0x3f, 0x1f, 0x0f};
diff --git a/thirdparty/graphite/src/call_machine.cpp b/thirdparty/graphite/src/call_machine.cpp
new file mode 100644
index 0000000000..fcd8a0c2c1
--- /dev/null
+++ b/thirdparty/graphite/src/call_machine.cpp
@@ -0,0 +1,138 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This call threaded interpreter implmentation for machine.h
+// Author: Tim Eves
+
+// Build either this interpreter or the direct_machine implementation.
+// The call threaded interpreter is portable across compilers and
+// architectures as well as being useful to debug (you can set breakpoints on
+// opcodes) but is slower that the direct threaded interpreter by a factor of 2
+
+#include <cassert>
+#include <cstring>
+#include <graphite2/Segment.h>
+#include "inc/Machine.h"
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/Rule.h"
+
+// Disable the unused parameter warning as th compiler is mistaken since dp
+// is always updated (even if by 0) on every opcode.
+#ifdef __GNUC__
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#endif
+
+#define registers           const byte * & dp, vm::Machine::stack_t * & sp, \
+                            vm::Machine::stack_t * const sb, regbank & reg
+
+// These are required by opcodes.h and should not be changed
+#define STARTOP(name)       bool name(registers) REGPARM(4);\
+                            bool name(registers) {
+#define ENDOP                   return (sp - sb)/Machine::STACK_MAX==0; \
+                            }
+
+#define EXIT(status)        { push(status); return false; }
+
+// This is required by opcode_table.h
+#define do_(name)           instr(name)
+
+
+using namespace graphite2;
+using namespace vm;
+
+struct regbank  {
+    slotref         is;
+    slotref *       map;
+    SlotMap       & smap;
+    slotref * const map_base;
+    const instr * & ip;
+    uint8           direction;
+    int8            flags;
+    Machine::status_t & status;
+};
+
+typedef bool        (* ip_t)(registers);
+
+// Pull in the opcode definitions
+// We pull these into a private namespace so these otherwise common names dont
+// pollute the toplevel namespace.
+namespace {
+#define smap    reg.smap
+#define seg     smap.segment
+#define is      reg.is
+#define ip      reg.ip
+#define map     reg.map
+#define mapb    reg.map_base
+#define flags   reg.flags
+#define dir     reg.direction
+#define status  reg.status
+
+#include "inc/opcodes.h"
+
+#undef smap
+#undef seg
+#undef is
+#undef ip
+#undef map
+#undef mapb
+#undef flags
+#undef dir
+}
+
+Machine::stack_t  Machine::run(const instr   * program,
+                               const byte    * data,
+                               slotref     * & map)
+
+{
+    assert(program != 0);
+
+    // Declare virtual machine registers
+    const instr   * ip = program-1;
+    const byte    * dp = data;
+    stack_t       * sp = _stack + Machine::STACK_GUARD,
+            * const sb = sp;
+    regbank         reg = {*map, map, _map, _map.begin()+_map.context(), ip, _map.dir(), 0, _status};
+
+    // Run the program
+    while ((reinterpret_cast<ip_t>(*++ip))(dp, sp, sb, reg)) {}
+    const stack_t ret = sp == _stack+STACK_GUARD+1 ? *sp-- : 0;
+
+    check_final_stack(sp);
+    map = reg.map;
+    *map = reg.is;
+    return ret;
+}
+
+// Pull in the opcode table
+namespace {
+#include "inc/opcode_table.h"
+}
+
+const opcode_t * Machine::getOpcodeTable() throw()
+{
+    return opcode_table;
+}
diff --git a/thirdparty/graphite/src/direct_machine.cpp b/thirdparty/graphite/src/direct_machine.cpp
new file mode 100644
index 0000000000..86206cfe37
--- /dev/null
+++ b/thirdparty/graphite/src/direct_machine.cpp
@@ -0,0 +1,140 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This direct threaded interpreter implmentation for machine.h
+// Author: Tim Eves
+
+// Build either this interpreter or the call_machine implementation.
+// The direct threaded interpreter is relies upon a gcc feature called
+// labels-as-values so is only portable to compilers that support the
+// extension (gcc only as far as I know) however it should build on any
+// architecture gcc supports.
+// This is twice as fast as the call threaded model and is likely faster on
+// inorder processors with short pipelines and little branch prediction such
+// as the ARM and possibly Atom chips.
+
+
+#include <cassert>
+#include <cstring>
+#include "inc/Machine.h"
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/Rule.h"
+
+#define STARTOP(name)           name: {
+#define ENDOP                   }; goto *((sp - sb)/Machine::STACK_MAX ? &&end : *++ip);
+#define EXIT(status)            { push(status); goto end; }
+
+#define do_(name)               &&name
+
+
+using namespace graphite2;
+using namespace vm;
+
+namespace {
+
+// The GCC manual has this to say about labels as values:
+//   The &&foo expressions for the same label might have different values
+//   if the containing function is inlined or cloned. If a program relies
+//   on them being always the same, __attribute__((__noinline__,__noclone__))
+//   should be used to prevent inlining and cloning.
+//
+// is_return in Code.cpp relies on being able to do comparisons, so it needs
+// them to be always the same.
+//
+// The GCC manual further adds:
+//   If &&foo is used in a static variable initializer, inlining and
+//   cloning is forbidden.
+//
+// In this file, &&foo *is* used in a static variable initializer, and it's not
+// entirely clear whether this should prevent inlining of the function or not.
+// In practice, though, clang 7 can end up inlining the function with ThinLTO,
+// which breaks at least is_return. https://bugs.llvm.org/show_bug.cgi?id=39241
+// So all in all, we need at least the __noinline__ attribute. __noclone__
+// is not supported by clang.
+__attribute__((__noinline__))
+const void * direct_run(const bool          get_table_mode,
+                        const instr       * program,
+                        const byte        * data,
+                        Machine::stack_t  * stack,
+                        slotref         * & __map,
+                        uint8                _dir,
+                        Machine::status_t & status,
+                        SlotMap           * __smap=0)
+{
+    // We need to define and return to opcode table from within this function
+    // other inorder to take the addresses of the instruction bodies.
+    #include "inc/opcode_table.h"
+    if (get_table_mode)
+        return opcode_table;
+
+    // Declare virtual machine registers
+    const instr           * ip = program;
+    const byte            * dp = data;
+    Machine::stack_t      * sp = stack + Machine::STACK_GUARD,
+                    * const sb = sp;
+    SlotMap             & smap = *__smap;
+    Segment              & seg = smap.segment;
+    slotref                 is = *__map,
+                         * map = __map,
+                  * const mapb = smap.begin()+smap.context();
+    uint8                  dir = _dir;
+    int8                 flags = 0;
+
+    // start the program
+    goto **ip;
+
+    // Pull in the opcode definitions
+    #include "inc/opcodes.h"
+
+    end:
+    __map  = map;
+    *__map = is;
+    return sp;
+}
+
+}
+
+const opcode_t * Machine::getOpcodeTable() throw()
+{
+    slotref * dummy;
+    Machine::status_t dumstat = Machine::finished;
+    return static_cast<const opcode_t *>(direct_run(true, 0, 0, 0, dummy, 0, dumstat));
+}
+
+
+Machine::stack_t  Machine::run(const instr   * program,
+                               const byte    * data,
+                               slotref     * & is)
+{
+    assert(program != 0);
+
+    const stack_t *sp = static_cast<const stack_t *>(
+                direct_run(false, program, data, _stack, is, _map.dir(), _status, &_map));
+    const stack_t ret = sp == _stack+STACK_GUARD+1 ? *sp-- : 0;
+    check_final_stack(sp);
+    return ret;
+}
diff --git a/thirdparty/graphite/src/gr_char_info.cpp b/thirdparty/graphite/src/gr_char_info.cpp
new file mode 100644
index 0000000000..612f9ba694
--- /dev/null
+++ b/thirdparty/graphite/src/gr_char_info.cpp
@@ -0,0 +1,65 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cassert>
+#include "graphite2/Segment.h"
+#include "inc/CharInfo.h"
+
+extern "C"
+{
+
+unsigned int gr_cinfo_unicode_char(const gr_char_info* p/*not NULL*/)
+{
+    assert(p);
+    return p->unicodeChar();
+}
+
+
+int gr_cinfo_break_weight(const gr_char_info* p/*not NULL*/)
+{
+    assert(p);
+    return p->breakWeight();
+}
+
+int gr_cinfo_after(const gr_char_info *p/*not NULL*/)
+{
+    assert(p);
+    return p->after();
+}
+
+int gr_cinfo_before(const gr_char_info *p/*not NULL*/)
+{
+    assert(p);
+    return p->before();
+}
+
+size_t gr_cinfo_base(const gr_char_info *p/*not NULL*/)
+{
+    assert(p);
+    return p->base();
+}
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/gr_face.cpp b/thirdparty/graphite/src/gr_face.cpp
new file mode 100644
index 0000000000..baa469727b
--- /dev/null
+++ b/thirdparty/graphite/src/gr_face.cpp
@@ -0,0 +1,267 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Font.h"
+#include "inc/Face.h"
+#include "inc/FileFace.h"
+#include "inc/GlyphCache.h"
+#include "inc/CmapCache.h"
+#include "inc/Silf.h"
+#include "inc/json.h"
+
+using namespace graphite2;
+
+#if !defined GRAPHITE2_NTRACING
+extern json *global_log;
+#endif
+
+namespace
+{
+    bool load_face(Face & face, unsigned int options)
+    {
+#ifdef GRAPHITE2_TELEMETRY
+        telemetry::category _misc_cat(face.tele.misc);
+#endif
+        Face::Table silf(face, Tag::Silf, 0x00050000);
+        if (!silf)
+            return false;
+
+        if (!face.readGlyphs(options))
+            return false;
+
+        if (silf)
+        {
+            if (!face.readFeatures() || !face.readGraphite(silf))
+            {
+#if !defined GRAPHITE2_NTRACING
+                if (global_log)
+                {
+                    *global_log << json::object
+                        << "type" << "fontload"
+                        << "failure" << face.error()
+                        << "context" << face.error_context()
+                    << json::close;
+                }
+#endif
+                return false;
+            }
+            else
+                return true;
+        }
+        else
+            return false;
+    }
+
+    inline
+    uint32 zeropad(const uint32 x)
+    {
+        if (x == 0x20202020)                    return 0;
+        if ((x & 0x00FFFFFF) == 0x00202020)     return x & 0xFF000000;
+        if ((x & 0x0000FFFF) == 0x00002020)     return x & 0xFFFF0000;
+        if ((x & 0x000000FF) == 0x00000020)     return x & 0xFFFFFF00;
+        return x;
+    }
+}
+
+extern "C" {
+
+gr_face* gr_make_face_with_ops(const void* appFaceHandle/*non-NULL*/, const gr_face_ops *ops, unsigned int faceOptions)
+                  //the appFaceHandle must stay alive all the time when the gr_face is alive. When finished with the gr_face, call destroy_face
+{
+    if (ops == 0)   return 0;
+
+    Face *res = new Face(appFaceHandle, *ops);
+    if (res && load_face(*res, faceOptions))
+        return static_cast<gr_face *>(res);
+
+    delete res;
+    return 0;
+}
+
+gr_face* gr_make_face(const void* appFaceHandle/*non-NULL*/, gr_get_table_fn tablefn, unsigned int faceOptions)
+{
+    const gr_face_ops ops = {sizeof(gr_face_ops), tablefn, NULL};
+    return gr_make_face_with_ops(appFaceHandle, &ops, faceOptions);
+}
+
+
+gr_face* gr_make_face_with_seg_cache_and_ops(const void* appFaceHandle/*non-NULL*/, const gr_face_ops *ops, unsigned int , unsigned int faceOptions)
+{
+  return gr_make_face_with_ops(appFaceHandle, ops, faceOptions);
+}
+
+gr_face* gr_make_face_with_seg_cache(const void* appFaceHandle/*non-NULL*/, gr_get_table_fn tablefn, unsigned int, unsigned int faceOptions)
+{
+  const gr_face_ops ops = {sizeof(gr_face_ops), tablefn, NULL};
+  return gr_make_face_with_ops(appFaceHandle, &ops, faceOptions);
+}
+
+gr_uint32 gr_str_to_tag(const char *str)
+{
+    uint32 res = 0;
+    switch(max(strlen(str),size_t(4)))
+    {
+        case 4: res |= str[3];       GR_FALLTHROUGH;
+        case 3: res |= str[2] << 8;  GR_FALLTHROUGH;
+        case 2: res |= str[1] << 16; GR_FALLTHROUGH;
+        case 1: res |= str[0] << 24; GR_FALLTHROUGH;
+        default:  break;
+    }
+    return res;
+}
+
+void gr_tag_to_str(gr_uint32 tag, char *str)
+{
+    if (!str) return;
+
+    *str++ = char(tag >> 24);
+    *str++ = char(tag >> 16);
+    *str++ = char(tag >> 8);
+    *str++ = char(tag);
+    *str = '\0';
+}
+
+gr_feature_val* gr_face_featureval_for_lang(const gr_face* pFace, gr_uint32 langname/*0 means clone default*/) //clones the features. if none for language, clones the default
+{
+    assert(pFace);
+    langname = zeropad(langname);
+    return static_cast<gr_feature_val *>(pFace->theSill().cloneFeatures(langname));
+}
+
+
+const gr_feature_ref* gr_face_find_fref(const gr_face* pFace, gr_uint32 featId)  //When finished with the FeatureRef, call destroy_FeatureRef
+{
+    assert(pFace);
+    featId = zeropad(featId);
+    const FeatureRef* pRef = pFace->featureById(featId);
+    return static_cast<const gr_feature_ref*>(pRef);
+}
+
+unsigned short gr_face_n_fref(const gr_face* pFace)
+{
+    assert(pFace);
+    int res = 0;
+    for (int i = 0; i < pFace->numFeatures(); ++i)
+        if (!(pFace->feature(i)->getFlags() & FeatureRef::HIDDEN))
+            ++res;
+    return res;
+}
+
+const gr_feature_ref* gr_face_fref(const gr_face* pFace, gr_uint16 i) //When finished with the FeatureRef, call destroy_FeatureRef
+{
+    assert(pFace);
+    int count = 0;
+    for (int j = 0; j < pFace->numFeatures(); ++j)
+    {
+        const FeatureRef* pRef = pFace->feature(j);
+        if (!(pRef->getFlags() & FeatureRef::HIDDEN))
+            if (count++ == i)
+                return static_cast<const gr_feature_ref*>(pRef);
+    }
+    return 0;
+}
+
+unsigned short gr_face_n_languages(const gr_face* pFace)
+{
+    assert(pFace);
+    return pFace->theSill().numLanguages();
+}
+
+gr_uint32 gr_face_lang_by_index(const gr_face* pFace, gr_uint16 i)
+{
+    assert(pFace);
+    return pFace->theSill().getLangName(i);
+}
+
+
+void gr_face_destroy(gr_face *face)
+{
+    delete static_cast<Face*>(face);
+}
+
+
+gr_uint16 gr_face_name_lang_for_locale(gr_face *face, const char * locale)
+{
+    if (face)
+    {
+        return face->languageForLocale(locale);
+    }
+    return 0;
+}
+
+unsigned short gr_face_n_glyphs(const gr_face* pFace)
+{
+    return pFace->glyphs().numGlyphs();
+}
+
+const gr_faceinfo *gr_face_info(const gr_face *pFace, gr_uint32 script)
+{
+    if (!pFace) return 0;
+    const Silf *silf = pFace->chooseSilf(script);
+    if (silf) return silf->silfInfo();
+    return 0;
+}
+
+int gr_face_is_char_supported(const gr_face* pFace, gr_uint32 usv, gr_uint32 script)
+{
+    const Cmap & cmap = pFace->cmap();
+    gr_uint16 gid = cmap[usv];
+    if (!gid)
+    {
+        const Silf * silf = pFace->chooseSilf(script);
+        gid = silf->findPseudo(usv);
+    }
+    return (gid != 0);
+}
+
+#ifndef GRAPHITE2_NFILEFACE
+gr_face* gr_make_file_face(const char *filename, unsigned int faceOptions)
+{
+    FileFace* pFileFace = new FileFace(filename);
+    if (*pFileFace)
+    {
+      gr_face* pRes = gr_make_face_with_ops(pFileFace, &FileFace::ops, faceOptions);
+      if (pRes)
+      {
+        pRes->takeFileFace(pFileFace);        //takes ownership
+        return pRes;
+      }
+    }
+
+    //error when loading
+
+    delete pFileFace;
+    return NULL;
+}
+
+gr_face* gr_make_file_face_with_seg_cache(const char* filename, unsigned int, unsigned int faceOptions)   //returns NULL on failure. //TBD better error handling
+                  //when finished with, call destroy_face
+{
+    return gr_make_file_face(filename, faceOptions);
+}
+#endif      //!GRAPHITE2_NFILEFACE
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/gr_features.cpp b/thirdparty/graphite/src/gr_features.cpp
new file mode 100644
index 0000000000..a560e053f2
--- /dev/null
+++ b/thirdparty/graphite/src/gr_features.cpp
@@ -0,0 +1,138 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Font.h"
+#include "inc/Face.h"
+#include "inc/FeatureMap.h"
+#include "inc/FeatureVal.h"
+#include "inc/NameTable.h"
+
+using namespace graphite2;
+
+extern "C" {
+
+
+gr_uint16 gr_fref_feature_value(const gr_feature_ref* pfeatureref, const gr_feature_val* feats)    //returns 0 if either pointer is NULL
+{
+    if (!pfeatureref || !feats) return 0;
+
+    return pfeatureref->getFeatureVal(*feats);
+}
+
+
+int gr_fref_set_feature_value(const gr_feature_ref* pfeatureref, gr_uint16 val, gr_feature_val* pDest)
+{
+    if (!pfeatureref || !pDest) return 0;
+
+    return pfeatureref->applyValToFeature(val, *pDest);
+}
+
+
+gr_uint32 gr_fref_id(const gr_feature_ref* pfeatureref)    //returns 0 if pointer is NULL
+{
+  if (!pfeatureref)
+    return 0;
+
+  return pfeatureref->getId();
+}
+
+
+gr_uint16 gr_fref_n_values(const gr_feature_ref* pfeatureref)
+{
+    if(!pfeatureref)
+        return 0;
+    return pfeatureref->getNumSettings();
+}
+
+
+gr_int16 gr_fref_value(const gr_feature_ref* pfeatureref, gr_uint16 settingno)
+{
+    if(!pfeatureref || (settingno >= pfeatureref->getNumSettings()))
+    {
+        return 0;
+    }
+    return pfeatureref->getSettingValue(settingno);
+}
+
+
+void* gr_fref_label(const gr_feature_ref* pfeatureref, gr_uint16 *langId, gr_encform utf, gr_uint32 *length)
+{
+    if(!pfeatureref)
+    {
+        langId = 0;
+        length = 0;
+        return NULL;
+    }
+    uint16 label = pfeatureref->getNameId();
+    NameTable * names = pfeatureref->getFace().nameTable();
+    if (!names)
+    {
+        langId = 0;
+        length = 0;
+        return NULL;
+    }
+    return names->getName(*langId, label, utf, *length);
+}
+
+
+void* gr_fref_value_label(const gr_feature_ref*pfeatureref, gr_uint16 setting,
+    gr_uint16 *langId, gr_encform utf, gr_uint32 *length)
+{
+    if(!pfeatureref || (setting >= pfeatureref->getNumSettings()))
+    {
+        langId = 0;
+        length = 0;
+        return NULL;
+    }
+    uint16 label = pfeatureref->getSettingName(setting);
+    NameTable * names = pfeatureref->getFace().nameTable();
+    if (!names)
+    {
+        langId = 0;
+        length = 0;
+        return NULL;
+    }
+    return names->getName(*langId, label, utf, *length);
+}
+
+
+void gr_label_destroy(void * label)
+{
+    free(label);
+}
+
+gr_feature_val* gr_featureval_clone(const gr_feature_val* pfeatures/*may be NULL*/)
+{                      //When finished with the Features, call features_destroy
+    return static_cast<gr_feature_val*>(pfeatures ? new Features(*pfeatures) : new Features);
+}
+
+void gr_featureval_destroy(gr_feature_val *p)
+{
+    delete static_cast<Features*>(p);
+}
+
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/gr_font.cpp b/thirdparty/graphite/src/gr_font.cpp
new file mode 100644
index 0000000000..724cc83c13
--- /dev/null
+++ b/thirdparty/graphite/src/gr_font.cpp
@@ -0,0 +1,74 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Font.h"
+#include "inc/Font.h"
+
+
+using namespace graphite2;
+
+extern "C" {
+
+void gr_engine_version(int *nMajor, int *nMinor, int *nBugFix)
+{
+    if (nMajor) *nMajor = GR2_VERSION_MAJOR;
+    if (nMinor) *nMinor = GR2_VERSION_MINOR;
+    if (nBugFix) *nBugFix = GR2_VERSION_BUGFIX;
+}
+
+gr_font* gr_make_font(float ppm/*pixels per em*/, const gr_face *face)
+{
+    return gr_make_font_with_advance_fn(ppm, 0, 0, face);
+}
+
+
+gr_font* gr_make_font_with_ops(float ppm/*pixels per em*/, const void* appFontHandle/*non-NULL*/, const gr_font_ops * font_ops, const gr_face * face/*needed for scaling*/)
+{                 //the appFontHandle must stay alive all the time when the gr_font is alive. When finished with the gr_font, call destroy_gr_font
+    if (face == 0 || ppm <= 0)  return 0;
+
+    Font * const res = new Font(ppm, *face, appFontHandle, font_ops);
+    if (*res)
+        return static_cast<gr_font*>(res);
+    else
+    {
+        delete res;
+        return 0;
+    }
+}
+
+gr_font* gr_make_font_with_advance_fn(float ppm/*pixels per em*/, const void* appFontHandle/*non-NULL*/, gr_advance_fn getAdvance, const gr_face * face/*needed for scaling*/)
+{
+    const gr_font_ops ops = {sizeof(gr_font_ops), getAdvance, NULL};
+    return gr_make_font_with_ops(ppm, appFontHandle, &ops, face);
+}
+
+void gr_font_destroy(gr_font *font)
+{
+    delete static_cast<Font*>(font);
+}
+
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/gr_logging.cpp b/thirdparty/graphite/src/gr_logging.cpp
new file mode 100644
index 0000000000..8f1e675e62
--- /dev/null
+++ b/thirdparty/graphite/src/gr_logging.cpp
@@ -0,0 +1,267 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include <cstdio>
+
+#include "graphite2/Log.h"
+#include "inc/debug.h"
+#include "inc/CharInfo.h"
+#include "inc/Slot.h"
+#include "inc/Segment.h"
+#include "inc/json.h"
+#include "inc/Collider.h"
+
+#if defined _WIN32
+#include "windows.h"
+#endif
+
+using namespace graphite2;
+
+#if !defined GRAPHITE2_NTRACING
+json *global_log = 0;
+#endif
+
+extern "C" {
+
+bool gr_start_logging(GR_MAYBE_UNUSED gr_face * face, const char *log_path)
+{
+    if (!log_path)  return false;
+
+#if !defined GRAPHITE2_NTRACING
+    gr_stop_logging(face);
+#if defined _WIN32
+    int n = MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, log_path, -1, 0, 0);
+    if (n == 0 || n > MAX_PATH - 12) return false;
+
+    LPWSTR wlog_path = gralloc<WCHAR>(n);
+    if (!wlog_path) return false;
+    FILE *log = 0;
+    if (wlog_path && MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, log_path, -1, wlog_path, n))
+        log = _wfopen(wlog_path, L"wt");
+
+    free(wlog_path);
+#else   // _WIN32
+    FILE *log = fopen(log_path, "wt");
+#endif  // _WIN32
+    if (!log)   return false;
+
+    if (face)
+    {
+        face->setLogger(log);
+        if (!face->logger()) return false;
+
+        *face->logger() << json::array;
+#ifdef GRAPHITE2_TELEMETRY
+        *face->logger() << face->tele;
+#endif
+    }
+    else
+    {
+        global_log = new json(log);
+        *global_log << json::array;
+    }
+
+    return true;
+#else   // GRAPHITE2_NTRACING
+    return false;
+#endif  // GRAPHITE2_NTRACING
+}
+
+bool graphite_start_logging(FILE * /* log */, GrLogMask /* mask */)
+{
+//#if !defined GRAPHITE2_NTRACING
+//  graphite_stop_logging();
+//
+//    if (!log) return false;
+//
+//    dbgout = new json(log);
+//    if (!dbgout) return false;
+//
+//    *dbgout << json::array;
+//    return true;
+//#else
+    return false;
+//#endif
+}
+
+void gr_stop_logging(GR_MAYBE_UNUSED gr_face * face)
+{
+#if !defined GRAPHITE2_NTRACING
+    if (face && face->logger())
+    {
+        FILE * log = face->logger()->stream();
+        face->setLogger(0);
+        fclose(log);
+    }
+    else if (!face && global_log)
+    {
+        FILE * log = global_log->stream();
+        delete global_log;
+        global_log = 0;
+        fclose(log);
+    }
+#endif
+}
+
+void graphite_stop_logging()
+{
+//    if (dbgout) delete dbgout;
+//    dbgout = 0;
+}
+
+} // extern "C"
+
+#ifdef GRAPHITE2_TELEMETRY
+size_t   * graphite2::telemetry::_category = 0UL;
+#endif
+
+#if !defined GRAPHITE2_NTRACING
+
+#ifdef GRAPHITE2_TELEMETRY
+
+json & graphite2::operator << (json & j, const telemetry & t) throw()
+{
+    j << json::object
+            << "type"   << "telemetry"
+            << "silf"   << t.silf
+            << "states" << t.states
+            << "starts" << t.starts
+            << "transitions" << t.transitions
+            << "glyphs" << t.glyph
+            << "code"   << t.code
+            << "misc"   << t.misc
+            << "total"  << (t.silf + t.states + t.starts + t.transitions + t.glyph + t.code + t.misc)
+        << json::close;
+    return j;
+}
+#else
+json & graphite2::operator << (json & j, const telemetry &) throw()
+{
+    return j;
+}
+#endif
+
+
+json & graphite2::operator << (json & j, const CharInfo & ci) throw()
+{
+    return j << json::object
+                << "offset"         << ci.base()
+                << "unicode"        << ci.unicodeChar()
+                << "break"          << ci.breakWeight()
+                << "flags"          << ci.flags()
+                << "slot" << json::flat << json::object
+                    << "before" << ci.before()
+                    << "after"  << ci.after()
+                    << json::close
+                << json::close;
+}
+
+
+json & graphite2::operator << (json & j, const dslot & ds) throw()
+{
+    assert(ds.first);
+    assert(ds.second);
+    const Segment & seg = *ds.first;
+    const Slot & s = *ds.second;
+    const SlotCollision *cslot = seg.collisionInfo(ds.second);
+
+    j << json::object
+        << "id"             << objectid(ds)
+        << "gid"            << s.gid()
+        << "charinfo" << json::flat << json::object
+            << "original"       << s.original()
+            << "before"         << s.before()
+            << "after"          << s.after()
+            << json::close
+        << "origin"         << s.origin()
+        << "shift"          << Position(float(s.getAttr(0, gr_slatShiftX, 0)),
+                                        float(s.getAttr(0, gr_slatShiftY, 0)))
+        << "advance"        << s.advancePos()
+        << "insert"         << s.isInsertBefore()
+        << "break"          << s.getAttr(&seg, gr_slatBreak, 0);
+    if (s.just() > 0)
+        j << "justification"    << s.just();
+    if (s.getBidiLevel() > 0)
+        j << "bidi"     << s.getBidiLevel();
+    if (!s.isBase())
+        j << "parent" << json::flat << json::object
+            << "id"             << objectid(dslot(&seg, s.attachedTo()))
+            << "level"          << s.getAttr(0, gr_slatAttLevel, 0)
+            << "offset"         << s.attachOffset()
+            << json::close;
+    j << "user" << json::flat << json::array;
+    for (int n = 0; n!= seg.numAttrs(); ++n)
+        j   << s.userAttrs()[n];
+    j       << json::close;
+    if (s.firstChild())
+    {
+        j   << "children" << json::flat << json::array;
+        for (const Slot *c = s.firstChild(); c; c = c->nextSibling())
+            j   << objectid(dslot(&seg, c));
+        j       << json::close;
+    }
+    if (cslot)
+    {
+		// Note: the reason for using Positions to lump together related attributes is to make the
+		// JSON output slightly more compact.
+        j << "collision" << json::flat << json::object
+//              << "shift" << cslot->shift() -- not used pass level, only within the collision routine itself
+              << "offset" << cslot->offset()
+              << "limit" << cslot->limit()
+              << "flags" << cslot->flags()
+              << "margin" << Position(cslot->margin(), cslot->marginWt())
+              << "exclude" << cslot->exclGlyph()
+              << "excludeoffset" << cslot->exclOffset();
+		if (cslot->seqOrder() != 0)
+		{
+			j << "seqclass" << Position(cslot->seqClass(), cslot->seqProxClass())
+				<< "seqorder" << cslot->seqOrder()
+				<< "seqabove" << Position(cslot->seqAboveXoff(), cslot->seqAboveWt())
+				<< "seqbelow" << Position(cslot->seqBelowXlim(), cslot->seqBelowWt())
+				<< "seqvalign" << Position(cslot->seqValignHt(), cslot->seqValignWt());
+		}
+        j << json::close;
+    }
+    return j << json::close;
+}
+
+
+graphite2::objectid::objectid(const dslot & ds) throw()
+{
+    const Slot * const p = ds.second;
+    uint32 s = uint32(reinterpret_cast<size_t>(p));
+    sprintf(name, "%.4x-%.2x-%.4hx", uint16(s >> 16), uint16(p ? p->userAttrs()[ds.first->silf()->numUser()] : 0), uint16(s));
+    name[sizeof name-1] = 0;
+}
+
+graphite2::objectid::objectid(const Segment * const p) throw()
+{
+    uint32 s = uint32(reinterpret_cast<size_t>(p));
+    sprintf(name, "%.4x-%.2x-%.4hx", uint16(s >> 16), 0, uint16(s));
+    name[sizeof name-1] = 0;
+}
+
+#endif
diff --git a/thirdparty/graphite/src/gr_segment.cpp b/thirdparty/graphite/src/gr_segment.cpp
new file mode 100644
index 0000000000..7a27e9c562
--- /dev/null
+++ b/thirdparty/graphite/src/gr_segment.cpp
@@ -0,0 +1,175 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Segment.h"
+#include "inc/UtfCodec.h"
+#include "inc/Segment.h"
+
+using namespace graphite2;
+
+namespace
+{
+
+  gr_segment* makeAndInitialize(const Font *font, const Face *face, uint32 script, const Features* pFeats/*must not be NULL*/, gr_encform enc, const void* pStart, size_t nChars, int dir)
+  {
+      if (script == 0x20202020) script = 0;
+      else if ((script & 0x00FFFFFF) == 0x00202020) script = script & 0xFF000000;
+      else if ((script & 0x0000FFFF) == 0x00002020) script = script & 0xFFFF0000;
+      else if ((script & 0x000000FF) == 0x00000020) script = script & 0xFFFFFF00;
+      // if (!font) return NULL;
+      Segment* pRes=new Segment(nChars, face, script, dir);
+
+
+      if (!pRes->read_text(face, pFeats, enc, pStart, nChars) || !pRes->runGraphite())
+      {
+        delete pRes;
+        return NULL;
+      }
+      pRes->finalise(font, true);
+
+      return static_cast<gr_segment*>(pRes);
+  }
+
+  template <typename utf_iter>
+  inline size_t count_unicode_chars(utf_iter first, const utf_iter last, const void **error)
+  {
+      size_t n_chars = 0;
+      uint32 usv = 0;
+
+      if (last)
+      {
+          if (!first.validate(last))
+          {
+              if (error)  *error = last - 1;
+              return 0;
+          }
+          for (;first != last; ++first, ++n_chars)
+              if ((usv = *first) == 0 || first.error()) break;
+      }
+      else
+      {
+          while ((usv = *first) != 0 && !first.error())
+          {
+              ++first;
+              ++n_chars;
+          }
+      }
+
+      if (error)  *error = first.error() ? first : 0;
+      return n_chars;
+  }
+}
+
+
+extern "C" {
+
+size_t gr_count_unicode_characters(gr_encform enc, const void* buffer_begin, const void* buffer_end/*don't go on or past end, If NULL then ignored*/, const void** pError)   //Also stops on nul. Any nul is not in the count
+{
+    assert(buffer_begin);
+
+    switch (enc)
+    {
+    case gr_utf8:   return count_unicode_chars<utf8::const_iterator>(buffer_begin, buffer_end, pError); break;
+    case gr_utf16:  return count_unicode_chars<utf16::const_iterator>(buffer_begin, buffer_end, pError); break;
+    case gr_utf32:  return count_unicode_chars<utf32::const_iterator>(buffer_begin, buffer_end, pError); break;
+    default:        return 0;
+    }
+}
+
+
+gr_segment* gr_make_seg(const gr_font *font, const gr_face *face, gr_uint32 script, const gr_feature_val* pFeats, gr_encform enc, const void* pStart, size_t nChars, int dir)
+{
+    if (!face) return nullptr;
+
+    const gr_feature_val * tmp_feats = 0;
+    if (pFeats == 0)
+        pFeats = tmp_feats = static_cast<const gr_feature_val*>(face->theSill().cloneFeatures(0));
+    gr_segment * seg = makeAndInitialize(font, face, script, pFeats, enc, pStart, nChars, dir);
+    delete static_cast<const FeatureVal*>(tmp_feats);
+
+    return seg;
+}
+
+
+void gr_seg_destroy(gr_segment* p)
+{
+    delete static_cast<Segment*>(p);
+}
+
+
+float gr_seg_advance_X(const gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return pSeg->advance().x;
+}
+
+
+float gr_seg_advance_Y(const gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return pSeg->advance().y;
+}
+
+
+unsigned int gr_seg_n_cinfo(const gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return static_cast<unsigned int>(pSeg->charInfoCount());
+}
+
+
+const gr_char_info* gr_seg_cinfo(const gr_segment* pSeg/*not NULL*/, unsigned int index/*must be <number_of_CharInfo*/)
+{
+    assert(pSeg);
+    return static_cast<const gr_char_info*>(pSeg->charinfo(index));
+}
+
+unsigned int gr_seg_n_slots(const gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return static_cast<unsigned int>(pSeg->slotCount());
+}
+
+const gr_slot* gr_seg_first_slot(gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return static_cast<const gr_slot*>(pSeg->first());
+}
+
+const gr_slot* gr_seg_last_slot(gr_segment* pSeg/*not NULL*/)
+{
+    assert(pSeg);
+    return static_cast<const gr_slot*>(pSeg->last());
+}
+
+float gr_seg_justify(gr_segment* pSeg/*not NULL*/, const gr_slot* pSlot/*not NULL*/, const gr_font *pFont, double width, enum gr_justFlags flags, const gr_slot *pFirst, const gr_slot *pLast)
+{
+    assert(pSeg);
+    assert(pSlot);
+    return pSeg->justify(const_cast<gr_slot *>(pSlot), pFont, float(width), justFlags(flags), const_cast<gr_slot *>(pFirst), const_cast<gr_slot *>(pLast));
+}
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/gr_slot.cpp b/thirdparty/graphite/src/gr_slot.cpp
new file mode 100644
index 0000000000..a3c6b46a7f
--- /dev/null
+++ b/thirdparty/graphite/src/gr_slot.cpp
@@ -0,0 +1,173 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#include "graphite2/Segment.h"
+#include "inc/Segment.h"
+#include "inc/Slot.h"
+#include "inc/Font.h"
+
+
+extern "C" {
+
+
+const gr_slot* gr_slot_next_in_segment(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return static_cast<const gr_slot*>(p->next());
+}
+
+const gr_slot* gr_slot_prev_in_segment(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return static_cast<const gr_slot*>(p->prev());
+}
+
+const gr_slot* gr_slot_attached_to(const gr_slot* p/*not NULL*/)        //returns NULL iff base. If called repeatedly on result, will get to a base
+{
+    assert(p);
+    return static_cast<const gr_slot*>(p->attachedTo());
+}
+
+
+const gr_slot* gr_slot_first_attachment(const gr_slot* p/*not NULL*/)        //returns NULL iff no attachments.
+{        //if slot_first_attachment(p) is not NULL, then slot_attached_to(slot_first_attachment(p))==p.
+    assert(p);
+    return static_cast<const gr_slot*>(p->firstChild());
+}
+
+
+const gr_slot* gr_slot_next_sibling_attachment(const gr_slot* p/*not NULL*/)        //returns NULL iff no more attachments.
+{        //if slot_next_sibling_attachment(p) is not NULL, then slot_attached_to(slot_next_sibling_attachment(p))==slot_attached_to(p).
+    assert(p);
+    return static_cast<const gr_slot*>(p->nextSibling());
+}
+
+
+unsigned short gr_slot_gid(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->glyph();
+}
+
+
+float gr_slot_origin_X(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->origin().x;
+}
+
+
+float gr_slot_origin_Y(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->origin().y;
+}
+
+
+float gr_slot_advance_X(const gr_slot* p/*not NULL*/, const gr_face *face, const gr_font *font)
+{
+    assert(p);
+    float scale = 1.0;
+    float res = p->advance();
+    if (font)
+    {
+        scale = font->scale();
+        int gid = p->glyph();
+        if (face && font->isHinted() && gid < face->glyphs().numGlyphs())
+            res = (res - face->glyphs().glyph(gid)->theAdvance().x) * scale + font->advance(gid);
+        else
+            res = res * scale;
+    }
+    return res;
+}
+
+float gr_slot_advance_Y(const gr_slot *p/*not NULL*/, GR_MAYBE_UNUSED const gr_face *face, const gr_font *font)
+{
+    assert(p);
+    float res = p->advancePos().y;
+    if (font)
+        return res * font->scale();
+    else
+        return res;
+}
+
+int gr_slot_before(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->before();
+}
+
+
+int gr_slot_after(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->after();
+}
+
+unsigned int gr_slot_index(const gr_slot *p/*not NULL*/)
+{
+    assert(p);
+    return p->index();
+}
+
+int gr_slot_attr(const gr_slot* p/*not NULL*/, const gr_segment* pSeg/*not NULL*/, gr_attrCode index, gr_uint8 subindex)
+{
+    assert(p);
+    return p->getAttr(pSeg, index, subindex);
+}
+
+
+int gr_slot_can_insert_before(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return (p->isInsertBefore())? 1 : 0;
+}
+
+
+int gr_slot_original(const gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    return p->original();
+}
+
+void gr_slot_linebreak_before(gr_slot* p/*not NULL*/)
+{
+    assert(p);
+    gr_slot *prev = (gr_slot *)p->prev();
+    prev->sibling(NULL);
+    prev->next(NULL);
+    p->prev(NULL);
+}
+
+#if 0       //what should this be
+size_t id(const gr_slot* p/*not NULL*/)
+{
+    return (size_t)p->id();
+}
+#endif
+
+
+} // extern "C"
diff --git a/thirdparty/graphite/src/inc/CharInfo.h b/thirdparty/graphite/src/inc/CharInfo.h
new file mode 100644
index 0000000000..01e7e31ac9
--- /dev/null
+++ b/thirdparty/graphite/src/inc/CharInfo.h
@@ -0,0 +1,66 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include "inc/Main.h"
+
+
+namespace graphite2 {
+
+class CharInfo
+{
+
+public:
+    CharInfo() : m_char(0), m_before(-1), m_after(-1), m_base(0), m_featureid(0), m_break(0), m_flags(0) {}
+    void init(int cid) { m_char = cid; }
+    unsigned int unicodeChar() const { return m_char; }
+    void feats(int offset) { m_featureid = offset; }
+    int fid() const { return m_featureid; }
+    int breakWeight() const { return m_break; }
+    void breakWeight(int val) { m_break = val; }
+    int after() const { return m_after; }
+    void after(int val) { m_after = val; }
+    int before() const { return m_before; }
+    void before(int val) { m_before = val; }
+    size_t base() const { return m_base; }
+    void base(size_t offset) { m_base = offset; }
+    void addflags(uint8 val) { m_flags |= val; }
+    uint8 flags() const { return m_flags; }
+
+    CLASS_NEW_DELETE
+private:
+    int m_char;     // Unicode character from character stream
+    int m_before;   // slot index before us, comes before
+    int m_after;    // slot index after us, comes after
+    size_t  m_base; // offset into input string corresponding to this charinfo
+    uint8 m_featureid;  // index into features list in the segment
+    int8 m_break;   // breakweight coming from lb table
+    uint8 m_flags;  // 0,1 segment split.
+};
+
+} // namespace graphite2
+
+struct gr_char_info : public graphite2::CharInfo {};
diff --git a/thirdparty/graphite/src/inc/CmapCache.h b/thirdparty/graphite/src/inc/CmapCache.h
new file mode 100644
index 0000000000..7820c958b0
--- /dev/null
+++ b/thirdparty/graphite/src/inc/CmapCache.h
@@ -0,0 +1,82 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "inc/Main.h"
+#include "inc/Face.h"
+
+namespace graphite2 {
+
+class Face;
+
+class Cmap
+{
+public:
+
+    virtual ~Cmap() throw() {}
+
+    virtual uint16 operator [] (const uint32) const throw() { return 0; }
+
+    virtual operator bool () const throw() { return false; }
+
+    CLASS_NEW_DELETE;
+};
+
+class DirectCmap : public Cmap
+{
+    DirectCmap(const DirectCmap &);
+    DirectCmap & operator = (const DirectCmap &);
+
+public:
+    DirectCmap(const Face &);
+    virtual uint16 operator [] (const uint32 usv) const throw();
+    virtual operator bool () const throw();
+
+    CLASS_NEW_DELETE;
+private:
+    const Face::Table   _cmap;
+    const void        * _smp,
+                      * _bmp;
+};
+
+class CachedCmap : public Cmap
+{
+    CachedCmap(const CachedCmap &);
+    CachedCmap & operator = (const CachedCmap &);
+
+public:
+    CachedCmap(const Face &);
+    virtual ~CachedCmap() throw();
+    virtual uint16 operator [] (const uint32 usv) const throw();
+    virtual operator bool () const throw();
+    CLASS_NEW_DELETE;
+private:
+    bool m_isBmpOnly;
+    uint16 ** m_blocks;
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Code.h b/thirdparty/graphite/src/inc/Code.h
new file mode 100644
index 0000000000..3cee67c81d
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Code.h
@@ -0,0 +1,171 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This class represents loaded graphite stack machine code.  It performs
+// basic sanity checks, on the incoming code to prevent more obvious problems
+// from crashing graphite.
+// Author: Tim Eves
+
+#pragma once
+
+#include <cassert>
+#include <graphite2/Types.h>
+#include "inc/Main.h"
+#include "inc/Machine.h"
+
+namespace graphite2 {
+
+class Silf;
+class Face;
+
+enum passtype {
+    PASS_TYPE_UNKNOWN = 0,
+    PASS_TYPE_LINEBREAK,
+    PASS_TYPE_SUBSTITUTE,
+    PASS_TYPE_POSITIONING,
+    PASS_TYPE_JUSTIFICATION
+};
+
+namespace vm {
+
+class Machine::Code
+{
+public:
+    enum status_t
+    {
+        loaded,
+        alloc_failed,
+        invalid_opcode,
+        unimplemented_opcode_used,
+        out_of_range_data,
+        jump_past_end,
+        arguments_exhausted,
+        missing_return,
+        nested_context_item,
+        underfull_stack
+    };
+
+private:
+    class decoder;
+
+    instr *     _code;
+    byte  *     _data;
+    size_t      _data_size,
+                _instr_count;
+    byte        _max_ref;
+    mutable status_t _status;
+    bool        _constraint,
+                _modify,
+                _delete;
+    mutable bool _own;
+
+    void release_buffers() throw ();
+    void failure(const status_t) throw();
+
+public:
+    static size_t estimateCodeDataOut(size_t num_bytecodes, int nRules, int nSlots);
+
+    Code() throw();
+    Code(bool is_constraint, const byte * bytecode_begin, const byte * const bytecode_end,
+         uint8 pre_context, uint16 rule_length, const Silf &, const Face &,
+         enum passtype pt, byte * * const _out = 0);
+    Code(const Machine::Code &) throw();
+    ~Code() throw();
+
+    Code & operator=(const Code &rhs) throw();
+    operator bool () const throw()                  { return _code && status() == loaded; }
+    status_t      status() const throw()            { return _status; }
+    bool          constraint() const throw()        { return _constraint; }
+    size_t        dataSize() const throw()          { return _data_size; }
+    size_t        instructionCount() const throw()  { return _instr_count; }
+    bool          immutable() const throw()         { return !(_delete || _modify); }
+    bool          deletes() const throw()           { return _delete; }
+    size_t        maxRef() const throw()            { return _max_ref; }
+    void          externalProgramMoved(ptrdiff_t) throw();
+
+    int32 run(Machine &m, slotref * & map) const;
+
+    CLASS_NEW_DELETE;
+};
+
+inline
+size_t  Machine::Code::estimateCodeDataOut(size_t n_bc, int nRules, int nSlots)
+{
+    // max is: all codes are instructions + 1 for each rule + max tempcopies
+    // allocate space for separate maximal code and data then merge them later
+    return (n_bc + nRules + nSlots) * sizeof(instr) + n_bc * sizeof(byte);
+}
+
+
+inline Machine::Code::Code() throw()
+: _code(0), _data(0), _data_size(0), _instr_count(0), _max_ref(0),
+  _status(loaded), _constraint(false), _modify(false), _delete(false),
+  _own(false)
+{
+}
+
+inline Machine::Code::Code(const Machine::Code &obj) throw ()
+ :  _code(obj._code),
+    _data(obj._data),
+    _data_size(obj._data_size),
+    _instr_count(obj._instr_count),
+    _max_ref(obj._max_ref),
+    _status(obj._status),
+    _constraint(obj._constraint),
+    _modify(obj._modify),
+    _delete(obj._delete),
+    _own(obj._own)
+{
+    obj._own = false;
+}
+
+inline Machine::Code & Machine::Code::operator=(const Machine::Code &rhs) throw() {
+    if (_instr_count > 0)
+        release_buffers();
+    _code        = rhs._code;
+    _data        = rhs._data;
+    _data_size   = rhs._data_size;
+    _instr_count = rhs._instr_count;
+    _status      = rhs._status;
+    _constraint  = rhs._constraint;
+    _modify      = rhs._modify;
+    _delete      = rhs._delete;
+    _own         = rhs._own;
+    rhs._own = false;
+    return *this;
+}
+
+inline void Machine::Code::externalProgramMoved(ptrdiff_t dist) throw()
+{
+    if (_code && !_own)
+    {
+        _code += dist / signed(sizeof(instr));
+        _data += dist;
+    }
+}
+
+} // namespace vm
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Collider.h b/thirdparty/graphite/src/inc/Collider.h
new file mode 100644
index 0000000000..71e8400501
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Collider.h
@@ -0,0 +1,245 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "inc/List.h"
+#include "inc/Position.h"
+#include "inc/Intervals.h"
+#include "inc/debug.h"
+
+namespace graphite2 {
+
+class json;
+class Slot;
+class Segment;
+
+#define SLOTCOLSETUINTPROP(x, y) uint16 x() const { return _ ##x; } void y (uint16 v) { _ ##x = v; }
+#define SLOTCOLSETINTPROP(x, y) int16 x() const { return _ ##x; } void y (int16 v) { _ ##x = v; }
+#define SLOTCOLSETPOSITIONPROP(x, y) const Position &x() const { return _ ##x; } void y (const Position &v) { _ ##x = v; }
+
+// Slot attributes related to collision-fixing
+class SlotCollision
+{
+public:
+    enum {
+    //  COLL_TESTONLY = 0,  // default - test other glyphs for collision with this one, but don't move this one
+        COLL_FIX = 1,       // fix collisions involving this glyph
+        COLL_IGNORE = 2,    // ignore this glyph altogether
+        COLL_START = 4,     // start of range of possible collisions
+        COLL_END = 8,       // end of range of possible collisions
+        COLL_KERN = 16,     // collisions with this glyph are fixed by adding kerning space after it
+        COLL_ISCOL = 32,    // this glyph has a collision
+        COLL_KNOWN = 64,    // we've figured out what's happening with this glyph
+        COLL_ISSPACE = 128,		// treat this glyph as a space with regard to kerning
+        COLL_TEMPLOCK = 256,    // Lock glyphs that have been given priority positioning
+        ////COLL_JUMPABLE = 128,    // moving glyphs may jump this stationary glyph in any direction - DELETE
+        ////COLL_OVERLAP = 256,    // use maxoverlap to restrict - DELETE
+    };
+
+    // Behavior for the collision.order attribute. To GDL this is an enum, to us it's a bitfield, with only 1 bit set
+    // Allows for easier inversion.
+    enum {
+        SEQ_ORDER_LEFTDOWN = 1,
+        SEQ_ORDER_RIGHTUP = 2,
+        SEQ_ORDER_NOABOVE = 4,
+        SEQ_ORDER_NOBELOW = 8,
+        SEQ_ORDER_NOLEFT = 16,
+        SEQ_ORDER_NORIGHT = 32
+    };
+
+    SlotCollision(Segment *seg, Slot *slot);
+    void initFromSlot(Segment *seg, Slot *slot);
+
+    const Rect &limit() const { return _limit; }
+    void setLimit(const Rect &r) { _limit = r; }
+    SLOTCOLSETPOSITIONPROP(shift, setShift)
+    SLOTCOLSETPOSITIONPROP(offset, setOffset)
+    SLOTCOLSETPOSITIONPROP(exclOffset, setExclOffset)
+    SLOTCOLSETUINTPROP(margin, setMargin)
+    SLOTCOLSETUINTPROP(marginWt, setMarginWt)
+    SLOTCOLSETUINTPROP(flags, setFlags)
+    SLOTCOLSETUINTPROP(exclGlyph, setExclGlyph)
+    SLOTCOLSETUINTPROP(seqClass, setSeqClass)
+    SLOTCOLSETUINTPROP(seqProxClass, setSeqProxClass)
+    SLOTCOLSETUINTPROP(seqOrder, setSeqOrder)
+    SLOTCOLSETINTPROP(seqAboveXoff, setSeqAboveXoff)
+    SLOTCOLSETUINTPROP(seqAboveWt, setSeqAboveWt)
+    SLOTCOLSETINTPROP(seqBelowXlim, setSeqBelowXlim)
+    SLOTCOLSETUINTPROP(seqBelowWt, setSeqBelowWt)
+    SLOTCOLSETUINTPROP(seqValignHt, setSeqValignHt)
+    SLOTCOLSETUINTPROP(seqValignWt, setSeqValignWt)
+
+    float getKern(int dir) const;
+    bool ignore() const;
+
+private:
+    Rect        _limit;
+    Position    _shift;     // adjustment within the given pass
+    Position    _offset;    // total adjustment for collisions
+    Position    _exclOffset;
+    uint16		_margin;
+    uint16		_marginWt;
+    uint16		_flags;
+    uint16		_exclGlyph;
+    uint16		_seqClass;
+	uint16		_seqProxClass;
+    uint16		_seqOrder;
+    int16		_seqAboveXoff;
+    uint16		_seqAboveWt;
+    int16		_seqBelowXlim;
+    uint16		_seqBelowWt;
+    uint16		_seqValignHt;
+    uint16		_seqValignWt;
+
+};  // end of class SlotColllision
+
+struct BBox;
+struct SlantBox;
+
+class ShiftCollider
+{
+public:
+    typedef std::pair<float, float> fpair;
+    typedef Vector<fpair> vfpairs;
+    typedef vfpairs::iterator ivfpairs;
+
+    ShiftCollider(json *dbgout);
+    ~ShiftCollider() throw() { };
+
+    bool initSlot(Segment *seg, Slot *aSlot, const Rect &constraint,
+                float margin, float marginMin, const Position &currShift,
+                const Position &currOffset, int dir, GR_MAYBE_UNUSED json * const dbgout);
+    bool mergeSlot(Segment *seg, Slot *slot, const SlotCollision *cinfo, const Position &currShift, bool isAfter,
+                bool sameCluster, bool &hasCol, bool isExclusion, GR_MAYBE_UNUSED json * const dbgout);
+    Position resolve(Segment *seg, bool &isCol, GR_MAYBE_UNUSED json * const dbgout);
+    void addBox_slope(bool isx, const Rect &box, const BBox &bb, const SlantBox &sb, const Position &org, float weight, float m, bool minright, int mode);
+    void removeBox(const Rect &box, const BBox &bb, const SlantBox &sb, const Position &org, int mode);
+    const Position &origin() const { return _origin; }
+
+#if !defined GRAPHITE2_NTRACING
+	void outputJsonDbg(json * const dbgout, Segment *seg, int axis);
+	void outputJsonDbgStartSlot(json * const dbgout, Segment *seg);
+	void outputJsonDbgEndSlot(json * const dbgout, Position resultPos, int bestAxis, bool isCol);
+	void outputJsonDbgOneVector(json * const dbgout, Segment *seg, int axis, float tleft, float bestCost, float bestVal);
+	void outputJsonDbgRawRanges(json * const dbgout, int axis);
+	void outputJsonDbgRemovals(json * const dbgout, int axis, Segment *seg);
+#endif
+
+    CLASS_NEW_DELETE;
+
+protected:
+    Zones _ranges[4];   // possible movements in 4 directions (horizontally, vertically, diagonally);
+    Slot *  _target;    // the glyph to fix
+    Rect    _limit;
+    Position _currShift;
+    Position _currOffset;
+    Position _origin;   // Base for all relative calculations
+    float   _margin;
+	float	_marginWt;
+    float   _len[4];
+    uint16  _seqClass;
+	uint16	_seqProxClass;
+    uint16  _seqOrder;
+
+	//bool _scraping[4];
+
+};	// end of class ShiftCollider
+
+inline
+ShiftCollider::ShiftCollider(GR_MAYBE_UNUSED json *dbgout)
+: _target(0),
+  _margin(0.0),
+  _marginWt(0.0),
+  _seqClass(0),
+  _seqProxClass(0),
+  _seqOrder(0)
+{
+#if !defined GRAPHITE2_NTRACING
+    for (int i = 0; i < 4; ++i)
+        _ranges[i].setdebug(dbgout);
+#endif
+}
+
+class KernCollider
+{
+public:
+    KernCollider(json *dbg);
+    ~KernCollider() throw() { };
+    bool initSlot(Segment *seg, Slot *aSlot, const Rect &constraint, float margin,
+            const Position &currShift, const Position &offsetPrev, int dir,
+            float ymin, float ymax, json * const dbgout);
+    bool mergeSlot(Segment *seg, Slot *slot, const Position &currShift, float currSpace, int dir, json * const dbgout);
+    Position resolve(Segment *seg, Slot *slot, int dir, json * const dbgout);
+    void shift(const Position &mv, int dir);
+
+    CLASS_NEW_DELETE;
+
+private:
+    Slot *  _target;        // the glyph to fix
+    Rect    _limit;
+    float   _margin;
+    Position _offsetPrev;   // kern from a previous pass
+    Position _currShift;    // NOT USED??
+    float _miny;	        // y-coordinates offset by global slot position
+    float _maxy;
+    Vector<float> _edges;   // edges of horizontal slices
+    float _sliceWidth;      // width of each slice
+    float _mingap;
+    float _xbound;        // max or min edge
+    bool  _hit;
+
+#if !defined GRAPHITE2_NTRACING
+    // Debugging
+    Segment * _seg;
+    Vector<float> _nearEdges; // closest potential collision in each slice
+    Vector<Slot*> _slotNear;
+#endif
+};	// end of class KernCollider
+
+
+inline
+float sqr(float x) {
+    return x * x;
+}
+
+inline
+KernCollider::KernCollider(GR_MAYBE_UNUSED json *dbg)
+: _target(0),
+  _margin(0.0f),
+  _miny(-1e38f),
+  _maxy(1e38f),
+  _sliceWidth(0.0f),
+  _mingap(0.0f),
+  _xbound(0.0),
+  _hit(false)
+{
+#if !defined GRAPHITE2_NTRACING
+    _seg = 0;
+#endif
+};
+
+};  // end of namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Compression.h b/thirdparty/graphite/src/inc/Compression.h
new file mode 100644
index 0000000000..9fe10e025d
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Compression.h
@@ -0,0 +1,104 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2015, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+#pragma once
+
+#include <cassert>
+#include <cstddef>
+#include <cstring>
+
+namespace
+{
+
+#if defined(_MSC_VER)
+typedef unsigned __int8 u8;
+typedef unsigned __int16 u16;
+typedef unsigned __int32 u32;
+typedef unsigned __int64 u64;
+#else
+#include <stdint.h>
+typedef uint8_t u8;
+typedef uint16_t u16;
+typedef uint32_t u32;
+typedef uint64_t u64;
+#endif
+
+ptrdiff_t const     MINMATCH = 4,
+                    LASTLITERALS = 5,
+                    MINCODA  = LASTLITERALS+1,
+                    MINSRCSIZE = 13;
+
+template<int S>
+inline
+void unaligned_copy(void * d, void const * s) {
+  ::memcpy(d, s, S);
+}
+
+inline
+size_t align(size_t p) {
+    return (p + sizeof(unsigned long)-1) & ~(sizeof(unsigned long)-1);
+}
+
+inline
+u8 * safe_copy(u8 * d, u8 const * s, size_t n) {
+    while (n--) *d++ = *s++;
+    return d;
+}
+
+inline
+u8 * overrun_copy(u8 * d, u8 const * s, size_t n) {
+    size_t const WS = sizeof(unsigned long);
+    u8 const * e = s + n;
+    do
+    {
+        unaligned_copy<WS>(d, s);
+        d += WS;
+        s += WS;
+    }
+    while (s < e);
+    d-=(s-e);
+
+    return d;
+}
+
+
+inline
+u8 * fast_copy(u8 * d, u8 const * s, size_t n) {
+    size_t const WS = sizeof(unsigned long);
+    size_t wn = n/WS;
+    while (wn--)
+    {
+        unaligned_copy<WS>(d, s);
+        d += WS;
+        s += WS;
+    }
+    n &= WS-1;
+    return safe_copy(d, s, n);
+}
+
+
+} // end of anonymous namespace
diff --git a/thirdparty/graphite/src/inc/Decompressor.h b/thirdparty/graphite/src/inc/Decompressor.h
new file mode 100644
index 0000000000..10f21b7af1
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Decompressor.h
@@ -0,0 +1,54 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2015, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+#pragma once
+
+#include <cstddef>
+
+namespace lz4
+{
+
+// decompress an LZ4 block
+// Parameters:
+//      @in         -   Input buffer containing an LZ4 block.
+//      @in_size    -   Size of the input LZ4 block in bytes.
+//      @out        -   Output buffer to hold decompressed results.
+//      @out_size   -   The size of the buffer pointed to by @out.
+// Invariants:
+//      @in         -   This buffer must be at least 1 machine word in length,
+//                      regardless of the actual LZ4 block size.
+//      @in_size    -   This must be at least 4 and must also be <= to the
+//                      allocated buffer @in.
+//      @out        -   This must be bigger than the input buffer and at least
+//                      13 bytes.
+//      @out_size   -   Must always be big enough to hold the expected size.
+// Return:
+//      -1          -  Decompression failed.
+//      size        -  Actual number of bytes decompressed.
+int decompress(void const *in, size_t in_size, void *out, size_t out_size);
+
+} // end of namespace shrinker
diff --git a/thirdparty/graphite/src/inc/Endian.h b/thirdparty/graphite/src/inc/Endian.h
new file mode 100644
index 0000000000..56ecfd8667
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Endian.h
@@ -0,0 +1,111 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+/*
+Description:
+    A set of fast template based decoders for decoding values of any C integer
+    type up to long int size laid out with most significant byte first or least
+    significant byte first (aka big endian or little endian).  These are CPU
+    byte order agnostic and will function the same regardless of the CPUs native
+    byte order.
+
+    Being template based means if the either le or be class is not used then
+    template code of unused functions will not be instantiated by the compiler
+    and thus shouldn't cause any overhead.
+*/
+
+#include <cstddef>
+
+#pragma once
+
+
+class be
+{
+    template<int S>
+    inline static unsigned long int _peek(const unsigned char * p) {
+        return _peek<S/2>(p) << (S/2)*8 | _peek<S/2>(p+S/2);
+    }
+public:
+    template<typename T>
+    inline static T peek(const void * p) {
+        return T(_peek<sizeof(T)>(static_cast<const unsigned char *>(p)));
+    }
+
+    template<typename T>
+    inline static T read(const unsigned char * &p) {
+        const T r = T(_peek<sizeof(T)>(p));
+        p += sizeof r;
+        return r;
+    }
+
+    template<typename T>
+    inline static T swap(const T x) {
+        return T(_peek<sizeof(T)>(reinterpret_cast<const unsigned char *>(&x)));
+    }
+
+    template<typename T>
+    inline static void skip(const unsigned char * &p, size_t n=1) {
+        p += sizeof(T)*n;
+    }
+};
+
+template<>
+inline unsigned long int be::_peek<1>(const unsigned char * p) { return *p; }
+
+
+class le
+{
+    template<int S>
+    inline static unsigned long int _peek(const unsigned char * p) {
+        return _peek<S/2>(p) | _peek<S/2>(p+S/2)  << (S/2)*8;
+    }
+public:
+    template<typename T>
+    inline static T peek(const void * p) {
+        return T(_peek<sizeof(T)>(static_cast<const unsigned char *>(p)));
+    }
+
+    template<typename T>
+    inline static T read(const unsigned char * &p) {
+        const T r = T(_peek<sizeof(T)>(p));
+        p += sizeof r;
+        return r;
+    }
+
+    template<typename T>
+    inline static T swap(const T x) {
+        return T(_peek<sizeof(T)>(reinterpret_cast<const unsigned char *>(&x)));
+    }
+
+    template<typename T>
+    inline static void skip(const unsigned char * &p, size_t n=1) {
+        p += sizeof(T)*n;
+    }
+};
+
+template<>
+inline unsigned long int le::_peek<1>(const unsigned char * p) { return *p; }
diff --git a/thirdparty/graphite/src/inc/Error.h b/thirdparty/graphite/src/inc/Error.h
new file mode 100644
index 0000000000..2b7ab763a2
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Error.h
@@ -0,0 +1,134 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2013, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+// numbers are explicitly assigned for future proofing
+
+namespace graphite2
+{
+
+class Error
+{
+public:
+    Error() : _e(0) {};
+    operator bool() { return (_e != 0); }
+    int error() { return _e; }
+    void error(int e) { _e = e; }
+    bool test(bool pr, int err) { return (_e = int(pr) * err); }
+
+private:
+    int _e;
+};
+
+enum errcontext {
+    EC_READGLYPHS = 1,      // while reading glyphs
+    EC_READSILF = 2,        // in Silf table
+    EC_ASILF = 3,           // in Silf %d
+    EC_APASS = 4,           // in Silf %d, pass %d
+    EC_PASSCCODE = 5,       // in pass constraint code for Silf %d, pass %d
+    EC_ARULE = 6,           // in Silf %d, pass %d, rule %d
+    EC_ASTARTS = 7,         // in Silf %d, pass %d, start state %d
+    EC_ATRANS = 8,          // in Silf %d, pass %d, fsm state %d
+    EC_ARULEMAP = 9         // in Silf %d, pass %d, state %d
+};
+
+enum errors {
+    E_OUTOFMEM = 1,         // Out of memory
+    E_NOGLYPHS = 2,         // There are no glyphs in the font
+    E_BADUPEM = 3,          // The units per em for the font is bad (0)
+    E_BADCMAP = 4,          // The font does not contain any useful cmaps
+    E_NOSILF = 5,           // Missing Silf table
+    E_TOOOLD = 6,           // Silf table version is too old
+    E_BADSIZE = 7,          // context object has the wrong structural size
+// Silf Subtable Errors take a Silf subtable number * 256 in the context
+    E_BADMAXGLYPH = 8,      // Silf max glyph id is too high
+    E_BADNUMJUSTS = 9,      // Number of Silf justification blocks is too high
+    E_BADENDJUSTS = 10,     // Silf justification blocks take too much of the Silf table space
+    E_BADCRITFEATURES = 11, // Critical features section in a Silf table is too big
+    E_BADSCRIPTTAGS = 12,   // Silf script tags area is too big
+    E_BADAPSEUDO = 13,      // The pseudo glyph attribute number is too high
+    E_BADABREAK = 14,       // The linebreak glyph attribute number is too high
+    E_BADABIDI = 15,        // The bidi glyph attribute number is too high
+    E_BADAMIRROR = 16,      // The mirrored glyph attribute number is too high
+    E_BADNUMPASSES = 17,    // The number of passes is > 128
+    E_BADPASSESSTART = 18,  // The Silf table is too small to hold any passes
+    E_BADPASSBOUND = 19,    // The positioning pass number is too low or the substitution pass number is too high
+    E_BADPPASS = 20,        // The positioning pass number is too high
+    E_BADSPASS = 21,        // the substitution pass number is too high
+    E_BADJPASSBOUND = 22,   // the justification pass must be higher than the positioning pass
+    E_BADJPASS = 23,        // the justification pass is too high
+    E_BADALIG = 24,         // the number of initial ligature component glyph attributes is too high
+    E_BADBPASS = 25,        // the bidi pass number is specified and is either too high or too low
+    E_BADNUMPSEUDO = 26,    // The number of pseudo glyphs is too high
+    E_BADCLASSSIZE = 27,    // The size of the classes block is bad
+    E_TOOMANYLINEAR = 28,   // The number of linear classes in the silf table is too high
+    E_CLASSESTOOBIG = 29,   // There are too many classes for the space allocated in the Silf subtable
+    E_MISALIGNEDCLASSES = 30,   // The class offsets in the class table don't line up with the number of classes
+    E_HIGHCLASSOFFSET = 31, // The class offsets point out of the class table
+    E_BADCLASSOFFSET = 32,  // A class offset is less than one following it
+    E_BADCLASSLOOKUPINFO = 33,  // The search header info for a non-linear class has wrong values in it
+// Pass subtable errors. Context has pass number * 65536
+    E_BADPASSSTART = 34,    // The start offset for a particular pass is bad
+    E_BADPASSEND = 35,      // The end offset for a particular pass is bad
+    E_BADPASSLENGTH = 36,   // The length of the pass is too small
+    E_BADNUMTRANS = 37,     // The number of transition states in the fsm is bad
+    E_BADNUMSUCCESS = 38,   // The number of success states in the fsm is bad
+    E_BADNUMSTATES = 39,    // The number of states in the fsm is bad
+    E_NORANGES = 40,        // There are no columns in the fsm
+    E_BADRULEMAPLEN = 41,   // The size of the success state to rule mapping is bad
+    E_BADCTXTLENBOUNDS = 42,    // The precontext maximum is greater than its minimum
+    E_BADCTXTLENS = 43,     // The lists of rule lengths or pre context lengths is bad
+    E_BADPASSCCODEPTR = 44, // The pass constraint code position does not align with where the forward reference says it should be
+    E_BADRULECCODEPTR = 45, // The rule constraint code position does not align with where the forward reference says it should be
+    E_BADCCODELEN = 46,     // Bad rule/pass constraint code length
+    E_BADACTIONCODEPTR = 47,    // The action code position does not align with where the forward reference says it should be
+    E_MUTABLECCODE = 48,    // Constraint code edits slots. It shouldn't.
+    E_BADSTATE = 49,        // Bad state transition referencing an illegal state
+    E_BADRULEMAPPING = 50,  // The structure of the rule mapping is bad
+    E_BADRANGE = 51,        // Bad column range structure including a glyph in more than one column
+    E_BADRULENUM = 52,      // A reference to a rule is out of range (too high)
+    E_BADACOLLISION = 53,   // Bad Silf table collision attribute number (too high)
+    E_BADEMPTYPASS = 54,    // Can't have empty passes (no rules) except for collision passes
+    E_BADSILFVERSION = 55,  // The Silf table has a bad version (probably too high)
+    E_BADCOLLISIONPASS = 56,    // Collision flags set on a non positioning pass
+    E_BADNUMCOLUMNS = 57,   // Arbitrarily limit number of columns in fsm
+// Code errors
+    E_CODEFAILURE = 60,     // Base code error. The following subcodes must align with Machine::Code::status_t in Code.h
+    E_CODEALLOC = 61,       // Out of memory
+    E_INVALIDOPCODE = 62,   // Invalid op code
+    E_UNIMPOPCODE = 63,     // Unimplemented op code encountered
+    E_OUTOFRANGECODE = 64,  // Code argument out of range
+    E_BADJUMPCODE = 65,     // Code jumps past end of op codes
+    E_CODEBADARGS = 66,     // Code arguments exhausted
+    E_CODENORETURN = 67,    // Missing return type op code at end of code
+    E_CODENESTEDCTXT = 68,   // Nested context encountered in code
+// Compression errors
+    E_BADSCHEME = 69,
+    E_SHRINKERFAILED = 70,
+};
+
+}
diff --git a/thirdparty/graphite/src/inc/Face.h b/thirdparty/graphite/src/inc/Face.h
new file mode 100644
index 0000000000..355c5aa0d3
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Face.h
@@ -0,0 +1,225 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <cstdio>
+
+#include "graphite2/Font.h"
+
+#include "inc/Main.h"
+#include "inc/FeatureMap.h"
+#include "inc/TtfUtil.h"
+#include "inc/Silf.h"
+#include "inc/Error.h"
+
+namespace graphite2 {
+
+class Cmap;
+class FileFace;
+class GlyphCache;
+class NameTable;
+class json;
+class Font;
+
+
+using TtfUtil::Tag;
+
+// These are the actual tags, as distinct from the consecutive IDs in TtfUtil.h
+
+class Face
+{
+    // Prevent any kind of copying
+    Face(const Face&);
+    Face& operator=(const Face&);
+
+public:
+    class Table;
+    static float default_glyph_advance(const void* face_ptr, gr_uint16 glyphid);
+
+    Face(const void* appFaceHandle/*non-NULL*/, const gr_face_ops & ops);
+    virtual ~Face();
+
+    virtual bool        runGraphite(Segment *seg, const Silf *silf) const;
+
+public:
+    bool                readGlyphs(uint32 faceOptions);
+    bool                readGraphite(const Table & silf);
+    bool                readFeatures();
+    void                takeFileFace(FileFace* pFileFace/*takes ownership*/);
+
+    const SillMap     & theSill() const;
+    const GlyphCache  & glyphs() const;
+    Cmap              & cmap() const;
+    NameTable         * nameTable() const;
+    void                setLogger(FILE *log_file);
+    json              * logger() const throw();
+
+    const Silf        * chooseSilf(uint32 script) const;
+    uint16              languageForLocale(const char * locale) const;
+
+    // Features
+    uint16              numFeatures() const;
+    const FeatureRef  * featureById(uint32 id) const;
+    const FeatureRef  * feature(uint16 index) const;
+
+    // Glyph related
+    int32  getGlyphMetric(uint16 gid, uint8 metric) const;
+    uint16 findPseudo(uint32 uid) const;
+
+    // Errors
+    unsigned int        error() const { return m_error; }
+    bool                error(Error e) { m_error = e.error(); return false; }
+    unsigned int        error_context() const { return m_error; }
+    void                error_context(unsigned int errcntxt) { m_errcntxt = errcntxt; }
+
+    CLASS_NEW_DELETE;
+private:
+    SillMap                 m_Sill;
+    gr_face_ops             m_ops;
+    const void            * m_appFaceHandle;    // non-NULL
+    FileFace              * m_pFileFace;        //owned
+    mutable GlyphCache    * m_pGlyphFaceCache;  // owned - never NULL
+    mutable Cmap          * m_cmap;             // cmap cache if available
+    mutable NameTable     * m_pNames;
+    mutable json          * m_logger;
+    unsigned int            m_error;
+    unsigned int            m_errcntxt;
+protected:
+    Silf                  * m_silfs;    // silf subtables.
+    uint16                  m_numSilf;  // num silf subtables in the silf table
+private:
+    uint16 m_ascent,
+           m_descent;
+#ifdef GRAPHITE2_TELEMETRY
+public:
+    mutable telemetry   tele;
+#endif
+};
+
+
+
+inline
+const SillMap & Face::theSill() const
+{
+    return m_Sill;
+}
+
+inline
+uint16 Face::numFeatures() const
+{
+    return m_Sill.theFeatureMap().numFeats();
+}
+
+inline
+const FeatureRef * Face::featureById(uint32 id) const
+{
+    return m_Sill.theFeatureMap().findFeatureRef(id);
+}
+
+inline
+const FeatureRef *Face::feature(uint16 index) const
+{
+    return m_Sill.theFeatureMap().feature(index);
+}
+
+inline
+const GlyphCache & Face::glyphs() const
+{
+    return *m_pGlyphFaceCache;
+}
+
+inline
+Cmap & Face::cmap() const
+{
+    return *m_cmap;
+};
+
+inline
+json * Face::logger() const throw()
+{
+    return m_logger;
+}
+
+
+
+class Face::Table
+{
+    const Face *            _f;
+    mutable const byte *    _p;
+    size_t                  _sz;
+    bool                    _compressed;
+
+    Error decompress();
+
+    void release();
+
+public:
+    Table() throw();
+    Table(const Face & face, const Tag n, uint32 version=0xffffffff) throw();
+    ~Table() throw();
+    Table(const Table && rhs) throw();
+
+    operator const byte * () const throw();
+
+    size_t  size() const throw();
+    Table & operator = (const Table && rhs) throw();
+};
+
+inline
+Face::Table::Table() throw()
+: _f(0), _p(0), _sz(0), _compressed(false)
+{
+}
+
+inline
+Face::Table::Table(const Table && rhs) throw()
+: _f(rhs._f), _p(rhs._p), _sz(rhs._sz), _compressed(rhs._compressed)
+{
+    rhs._p = 0;
+}
+
+inline
+Face::Table::~Table() throw()
+{
+    release();
+}
+
+inline
+Face::Table::operator const byte * () const throw()
+{
+    return _p;
+}
+
+inline
+size_t  Face::Table::size() const throw()
+{
+    return _sz;
+}
+
+} // namespace graphite2
+
+struct gr_face : public graphite2::Face {};
diff --git a/thirdparty/graphite/src/inc/FeatureMap.h b/thirdparty/graphite/src/inc/FeatureMap.h
new file mode 100644
index 0000000000..0f05e941a2
--- /dev/null
+++ b/thirdparty/graphite/src/inc/FeatureMap.h
@@ -0,0 +1,198 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include "inc/Main.h"
+#include "inc/FeatureVal.h"
+
+namespace graphite2 {
+
+// Forward declarations for implmentation types
+class FeatureMap;
+class Face;
+
+
+class FeatureSetting
+{
+public:
+    FeatureSetting(int16 theValue, uint16 labelId) : m_label(labelId), m_value(theValue) {};
+    uint16 label() const { return m_label; }
+    int16 value() const { return m_value; }
+
+    CLASS_NEW_DELETE;
+private:
+    FeatureSetting(const FeatureSetting & fs) : m_label(fs.m_label), m_value(fs.m_value) {};
+
+    uint16 m_label;
+    int16 m_value;
+};
+
+class FeatureRef
+{
+    typedef uint32      chunk_t;
+    static const uint8  SIZEOF_CHUNK = sizeof(chunk_t)*8;
+
+public:
+    enum flags_t : uint16 {
+        HIDDEN = 0x0800
+    };
+    FeatureRef() throw();
+    FeatureRef(const Face & face, unsigned short & bits_offset, uint32 max_val,
+               uint32 name, uint16 uiName, flags_t flags,
+               FeatureSetting *settings, uint16 num_set) throw();
+    ~FeatureRef() throw();
+
+    bool applyValToFeature(uint32 val, Features& pDest) const; //defined in GrFaceImp.h
+    void maskFeature(Features & pDest) const {
+    if (m_index < pDest.size())                 //defensive
+        pDest[m_index] |= m_mask;
+    }
+
+    uint32 getFeatureVal(const Features& feats) const; //defined in GrFaceImp.h
+
+    uint32 getId() const { return m_id; }
+    uint16 getNameId() const { return m_nameid; }
+    uint16 getNumSettings() const { return m_numSet; }
+    uint16 getSettingName(uint16 index) const { return m_nameValues[index].label(); }
+    int16  getSettingValue(uint16 index) const { return m_nameValues[index].value(); }
+    flags_t getFlags() const { return m_flags; }
+    uint32 maxVal() const { return m_max; }
+    const Face & getFace() const { assert(m_face); return *m_face;}
+    const FeatureMap* getFeatureMap() const;// { return m_pFace;}
+
+    CLASS_NEW_DELETE;
+private:
+    FeatureRef(const FeatureRef & rhs);
+
+    const Face     * m_face;
+    FeatureSetting * m_nameValues; // array of name table ids for feature values
+    chunk_t m_mask,             // bit mask to get the value from the vector
+            m_max;              // max value the value can take
+    uint32  m_id;               // feature identifier/name
+    uint16  m_nameid,           // Name table id for feature name
+            m_numSet;           // number of values (number of entries in m_nameValues)
+    flags_t m_flags;            // feature flags see FeatureRef::flags_t.
+    byte    m_bits,             // how many bits to shift the value into place
+            m_index;            // index into the array to find the ulong to mask
+
+private:        //unimplemented
+    FeatureRef& operator=(const FeatureRef&);
+};
+
+inline
+FeatureRef::FeatureRef() throw()
+: m_face(0),
+  m_nameValues(0),
+  m_mask(0), m_max(0),
+  m_id(0), m_nameid(0), m_numSet(0), 
+  m_flags(flags_t(0)),
+  m_bits(0), m_index(0)
+{
+}
+
+
+class NameAndFeatureRef
+{
+  public:
+    NameAndFeatureRef(uint32 name = 0) : m_name(name) , m_pFRef(NULL){}
+    NameAndFeatureRef(FeatureRef const & p) : m_name(p.getId()), m_pFRef(&p) {}
+
+    bool operator<(const NameAndFeatureRef& rhs) const //orders by m_name
+        {   return m_name<rhs.m_name; }
+
+    CLASS_NEW_DELETE
+
+    uint32 m_name;
+    const FeatureRef* m_pFRef;
+};
+
+class FeatureMap
+{
+public:
+    FeatureMap() : m_numFeats(0), m_feats(NULL), m_pNamedFeats(NULL) {}
+    ~FeatureMap() { delete[] m_feats; delete[] m_pNamedFeats; }
+
+    bool readFeats(const Face & face);
+    const FeatureRef *findFeatureRef(uint32 name) const;
+    FeatureRef *feature(uint16 index) const { return m_feats + index; }
+    //GrFeatureRef *featureRef(byte index) { return index < m_numFeats ? m_feats + index : NULL; }
+    const FeatureRef *featureRef(byte index) const { return index < m_numFeats ? m_feats + index : NULL; }
+    FeatureVal* cloneFeatures(uint32 langname/*0 means default*/) const;      //call destroy_Features when done.
+    uint16 numFeats() const { return m_numFeats; };
+    CLASS_NEW_DELETE
+private:
+friend class SillMap;
+    uint16 m_numFeats;
+
+    FeatureRef *m_feats;
+    NameAndFeatureRef* m_pNamedFeats;   //owned
+    FeatureVal m_defaultFeatures;        //owned
+
+private:        //defensive on m_feats, m_pNamedFeats, and m_defaultFeatures
+    FeatureMap(const FeatureMap&);
+    FeatureMap& operator=(const FeatureMap&);
+};
+
+
+class SillMap
+{
+private:
+    class LangFeaturePair
+    {
+        LangFeaturePair(const LangFeaturePair &);
+        LangFeaturePair & operator = (const LangFeaturePair &);
+
+    public:
+        LangFeaturePair() :  m_lang(0), m_pFeatures(0) {}
+        ~LangFeaturePair() { delete m_pFeatures; }
+
+        uint32 m_lang;
+        Features* m_pFeatures;      //owns
+        CLASS_NEW_DELETE
+    };
+public:
+    SillMap() : m_langFeats(NULL), m_numLanguages(0) {}
+    ~SillMap() { delete[] m_langFeats; }
+    bool readFace(const Face & face);
+    bool readSill(const Face & face);
+    FeatureVal* cloneFeatures(uint32 langname/*0 means default*/) const;      //call destroy_Features when done.
+    uint16 numLanguages() const { return m_numLanguages; };
+    uint32 getLangName(uint16 index) const { return (index < m_numLanguages)? m_langFeats[index].m_lang : 0; };
+
+    const FeatureMap & theFeatureMap() const { return m_FeatureMap; };
+private:
+    FeatureMap m_FeatureMap;        //of face
+    LangFeaturePair * m_langFeats;
+    uint16 m_numLanguages;
+
+private:        //defensive on m_langFeats
+    SillMap(const SillMap&);
+    SillMap& operator=(const SillMap&);
+};
+
+} // namespace graphite2
+
+struct gr_feature_ref : public graphite2::FeatureRef {};
diff --git a/thirdparty/graphite/src/inc/FeatureVal.h b/thirdparty/graphite/src/inc/FeatureVal.h
new file mode 100644
index 0000000000..cd3f93b2b5
--- /dev/null
+++ b/thirdparty/graphite/src/inc/FeatureVal.h
@@ -0,0 +1,68 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include <cstring>
+#include <cassert>
+#include "inc/Main.h"
+#include "inc/List.h"
+
+namespace graphite2 {
+
+class FeatureRef;
+class FeatureMap;
+
+class FeatureVal : public Vector<uint32>
+{
+public:
+    FeatureVal() : m_pMap(0) { }
+    FeatureVal(int num, const FeatureMap & pMap) : Vector<uint32>(num), m_pMap(&pMap) {}
+    FeatureVal(const FeatureVal & rhs) : Vector<uint32>(rhs), m_pMap(rhs.m_pMap) {}
+
+    FeatureVal & operator = (const FeatureVal & rhs) { Vector<uint32>::operator = (rhs); m_pMap = rhs.m_pMap; return *this; }
+
+    bool operator ==(const FeatureVal & b) const
+    {
+        size_t n = size();
+        if (n != b.size())      return false;
+
+        for(const_iterator l = begin(), r = b.begin(); n && *l == *r; --n, ++l, ++r);
+
+        return n == 0;
+    }
+
+    CLASS_NEW_DELETE
+private:
+    friend class FeatureRef;        //so that FeatureRefs can manipulate m_vec directly
+    const FeatureMap* m_pMap;
+};
+
+typedef FeatureVal Features;
+
+} // namespace graphite2
+
+
+struct gr_feature_val : public graphite2::FeatureVal {};
diff --git a/thirdparty/graphite/src/inc/FileFace.h b/thirdparty/graphite/src/inc/FileFace.h
new file mode 100644
index 0000000000..35927847f8
--- /dev/null
+++ b/thirdparty/graphite/src/inc/FileFace.h
@@ -0,0 +1,80 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+//#include "inc/FeatureMap.h"
+//#include "inc/GlyphsCache.h"
+//#include "inc/Silf.h"
+
+#ifndef GRAPHITE2_NFILEFACE
+
+#include <cstdio>
+#include <cassert>
+
+#include "graphite2/Font.h"
+
+#include "inc/Main.h"
+#include "inc/TtfTypes.h"
+#include "inc/TtfUtil.h"
+
+namespace graphite2 {
+
+
+class FileFace
+{
+    static const void * get_table_fn(const void* appFaceHandle, unsigned int name, size_t *len);
+    static void         rel_table_fn(const void* appFaceHandle, const void *table_buffer);
+
+public:
+    static const gr_face_ops ops;
+
+    FileFace(const char *filename);
+    ~FileFace();
+
+    operator bool () const throw();
+    CLASS_NEW_DELETE;
+
+private:        //defensive
+    FILE          * _file;
+    size_t          _file_len;
+
+    TtfUtil::Sfnt::OffsetSubTable         * _header_tbl;
+    TtfUtil::Sfnt::OffsetSubTable::Entry  * _table_dir;
+
+    FileFace(const FileFace&);
+    FileFace& operator=(const FileFace&);
+};
+
+inline
+FileFace::operator bool() const throw()
+{
+    return _file && _header_tbl && _table_dir;
+}
+
+} // namespace graphite2
+
+#endif      //!GRAPHITE2_NFILEFACE
diff --git a/thirdparty/graphite/src/inc/Font.h b/thirdparty/graphite/src/inc/Font.h
new file mode 100644
index 0000000000..9bc9ffb510
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Font.h
@@ -0,0 +1,90 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include <cassert>
+#include "graphite2/Font.h"
+#include "inc/Main.h"
+#include "inc/Face.h"
+
+namespace graphite2 {
+
+#define INVALID_ADVANCE -1e38f      // can't be a static const because non-integral
+
+class Font
+{
+public:
+    Font(float ppm, const Face & face, const void * appFontHandle=0, const gr_font_ops * ops=0);
+    virtual ~Font();
+
+    float advance(unsigned short glyphid) const;
+    float scale() const;
+    bool isHinted() const;
+    const Face & face() const;
+    operator bool () const throw()  { return m_advances; }
+
+    CLASS_NEW_DELETE;
+private:
+    gr_font_ops         m_ops;
+    const void  * const m_appFontHandle;
+    float             * m_advances;  // One advance per glyph in pixels. Nan if not defined
+    const Face        & m_face;
+    float               m_scale;      // scales from design units to ppm
+    bool                m_hinted;
+
+    Font(const Font&);
+    Font& operator=(const Font&);
+};
+
+inline
+float Font::advance(unsigned short glyphid) const
+{
+    if (m_advances[glyphid] == INVALID_ADVANCE)
+        m_advances[glyphid] = (*m_ops.glyph_advance_x)(m_appFontHandle, glyphid);
+    return m_advances[glyphid];
+}
+
+inline
+float Font::scale() const
+{
+    return m_scale;
+}
+
+inline
+bool Font::isHinted() const
+{
+    return m_hinted;
+}
+
+inline
+const Face & Font::face() const
+{
+    return m_face;
+}
+
+} // namespace graphite2
+
+struct gr_font : public graphite2::Font {};
diff --git a/thirdparty/graphite/src/inc/GlyphCache.h b/thirdparty/graphite/src/inc/GlyphCache.h
new file mode 100644
index 0000000000..7d5324e522
--- /dev/null
+++ b/thirdparty/graphite/src/inc/GlyphCache.h
@@ -0,0 +1,223 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+
+#include "graphite2/Font.h"
+#include "inc/Main.h"
+#include "inc/Position.h"
+#include "inc/GlyphFace.h"
+
+namespace graphite2 {
+
+class Face;
+class FeatureVal;
+class Segment;
+
+
+struct SlantBox
+{
+    static const SlantBox empty;
+
+//    SlantBox(float psi = 0., float pdi = 0., float psa = 0., float pda = 0.) : si(psi), di(pdi), sa(psa), da(pda) {};
+    float width() const { return sa - si; }
+    float height() const { return da - di; }
+    float si; // min
+    float di; // min
+    float sa; // max
+    float da; // max
+};
+
+
+struct BBox
+{
+    BBox(float pxi = 0, float pyi = 0., float pxa = 0., float pya = 0.) : xi(pxi), yi(pyi), xa(pxa), ya(pya) {};
+    float width() const { return xa - xi; }
+    float height() const { return ya - yi; }
+    float xi; // min
+    float yi; // min
+    float xa; // max
+    float ya; // max
+};
+
+
+class GlyphBox
+{
+    GlyphBox(const GlyphBox &);
+    GlyphBox & operator = (const GlyphBox &);
+
+public:
+    GlyphBox(uint8 numsubs, unsigned short bitmap, Rect *slanted) : _num(numsubs), _bitmap(bitmap), _slant(*slanted) {};
+
+    void addSubBox(int subindex, int boundary, Rect *val) { _subs[subindex * 2 + boundary] = *val; }
+    Rect &subVal(int subindex, int boundary) { return _subs[subindex * 2 + boundary]; }
+    const Rect &slant() const { return _slant; }
+    uint8 num() const { return _num; }
+    const Rect *subs() const { return _subs; }
+
+private:
+    uint8   _num;
+    unsigned short  _bitmap;
+    Rect    _slant;
+    Rect    _subs[1];
+};
+
+class GlyphCache
+{
+    class Loader;
+
+    GlyphCache(const GlyphCache&);
+    GlyphCache& operator=(const GlyphCache&);
+
+public:
+    GlyphCache(const Face & face, const uint32 face_options);
+    ~GlyphCache();
+
+    unsigned short  numGlyphs() const throw();
+    unsigned short  numAttrs() const throw();
+    unsigned short  unitsPerEm() const throw();
+
+    const GlyphFace *glyph(unsigned short glyphid) const;      //result may be changed by subsequent call with a different glyphid
+    const GlyphFace *glyphSafe(unsigned short glyphid) const;
+    float            getBoundingMetric(unsigned short glyphid, uint8 metric) const;
+    uint8            numSubBounds(unsigned short glyphid) const;
+    float            getSubBoundingMetric(unsigned short glyphid, uint8 subindex, uint8 metric) const;
+    const Rect &     slant(unsigned short glyphid) const { return _boxes[glyphid] ? _boxes[glyphid]->slant() : _empty_slant_box; }
+    const SlantBox & getBoundingSlantBox(unsigned short glyphid) const;
+    const BBox &     getBoundingBBox(unsigned short glyphid) const;
+    const SlantBox & getSubBoundingSlantBox(unsigned short glyphid, uint8 subindex) const;
+    const BBox &     getSubBoundingBBox(unsigned short glyphid, uint8 subindex) const;
+    bool             check(unsigned short glyphid) const;
+    bool             hasBoxes() const { return _boxes != 0; }
+
+    CLASS_NEW_DELETE;
+
+private:
+    const Rect            _empty_slant_box;
+    const Loader        * _glyph_loader;
+    const GlyphFace *   * _glyphs;
+    GlyphBox        *   * _boxes;
+    unsigned short        _num_glyphs,
+                          _num_attrs,
+                          _upem;
+};
+
+inline
+unsigned short GlyphCache::numGlyphs() const throw()
+{
+    return _num_glyphs;
+}
+
+inline
+unsigned short GlyphCache::numAttrs() const throw()
+{
+    return _num_attrs;
+}
+
+inline
+unsigned short  GlyphCache::unitsPerEm() const throw()
+{
+    return _upem;
+}
+
+inline
+bool GlyphCache::check(unsigned short glyphid) const
+{
+    return _boxes && glyphid < _num_glyphs;
+}
+
+inline
+const GlyphFace *GlyphCache::glyphSafe(unsigned short glyphid) const
+{
+    return glyphid < _num_glyphs ? glyph(glyphid) : NULL;
+}
+
+inline
+float GlyphCache::getBoundingMetric(unsigned short glyphid, uint8 metric) const
+{
+    if (glyphid >= _num_glyphs) return 0.;
+    switch (metric) {
+        case 0: return (float)(glyph(glyphid)->theBBox().bl.x);                          // x_min
+        case 1: return (float)(glyph(glyphid)->theBBox().bl.y);                          // y_min
+        case 2: return (float)(glyph(glyphid)->theBBox().tr.x);                          // x_max
+        case 3: return (float)(glyph(glyphid)->theBBox().tr.y);                          // y_max
+        case 4: return (float)(_boxes[glyphid] ? _boxes[glyphid]->slant().bl.x : 0.f);    // sum_min
+        case 5: return (float)(_boxes[glyphid] ? _boxes[glyphid]->slant().bl.y : 0.f);    // diff_min
+        case 6: return (float)(_boxes[glyphid] ? _boxes[glyphid]->slant().tr.x : 0.f);    // sum_max
+        case 7: return (float)(_boxes[glyphid] ? _boxes[glyphid]->slant().tr.y : 0.f);    // diff_max
+        default: return 0.;
+    }
+}
+
+inline const SlantBox &GlyphCache::getBoundingSlantBox(unsigned short glyphid) const
+{
+    return _boxes[glyphid] ? *(SlantBox *)(&(_boxes[glyphid]->slant())) : SlantBox::empty;
+}
+
+inline const BBox &GlyphCache::getBoundingBBox(unsigned short glyphid) const
+{
+    return *(BBox *)(&(glyph(glyphid)->theBBox()));
+}
+
+inline
+float GlyphCache::getSubBoundingMetric(unsigned short glyphid, uint8 subindex, uint8 metric) const
+{
+    GlyphBox *b = _boxes[glyphid];
+    if (b == NULL || subindex >= b->num()) return 0;
+
+    switch (metric) {
+        case 0: return b->subVal(subindex, 0).bl.x;
+        case 1: return b->subVal(subindex, 0).bl.y;
+        case 2: return b->subVal(subindex, 0).tr.x;
+        case 3: return b->subVal(subindex, 0).tr.y;
+        case 4: return b->subVal(subindex, 1).bl.x;
+        case 5: return b->subVal(subindex, 1).bl.y;
+        case 6: return b->subVal(subindex, 1).tr.x;
+        case 7: return b->subVal(subindex, 1).tr.y;
+        default: return 0.;
+    }
+}
+
+inline const SlantBox &GlyphCache::getSubBoundingSlantBox(unsigned short glyphid, uint8 subindex) const
+{
+    GlyphBox *b = _boxes[glyphid];
+    return *(SlantBox *)(b->subs() + 2 * subindex + 1);
+}
+
+inline const BBox &GlyphCache::getSubBoundingBBox(unsigned short glyphid, uint8 subindex) const
+{
+    GlyphBox *b = _boxes[glyphid];
+    return *(BBox *)(b->subs() + 2 * subindex);
+}
+
+inline
+uint8 GlyphCache::numSubBounds(unsigned short glyphid) const
+{
+    return _boxes[glyphid] ? _boxes[glyphid]->num() : 0;
+}
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/GlyphFace.h b/thirdparty/graphite/src/inc/GlyphFace.h
new file mode 100644
index 0000000000..fc29056146
--- /dev/null
+++ b/thirdparty/graphite/src/inc/GlyphFace.h
@@ -0,0 +1,83 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "inc/Main.h"
+#include "inc/Position.h"
+#include "inc/Sparse.h"
+
+namespace graphite2 {
+
+enum metrics {
+    kgmetLsb = 0, kgmetRsb,
+    kgmetBbTop, kgmetBbBottom, kgmetBbLeft, kgmetBbRight,
+    kgmetBbHeight, kgmetBbWidth,
+    kgmetAdvWidth, kgmetAdvHeight,
+    kgmetAscent, kgmetDescent
+};
+
+
+class GlyphFace
+{
+public:
+    GlyphFace();
+    template<typename I>
+    GlyphFace(const Rect & bbox, const Position & adv, I first, const I last);
+
+    const Position    & theAdvance() const;
+    const Rect        & theBBox() const { return m_bbox; }
+    const sparse      & attrs() const { return m_attrs; }
+    int32               getMetric(uint8 metric) const;
+
+    CLASS_NEW_DELETE;
+private:
+    Rect     m_bbox;        // bounding box metrics in design units
+    Position m_advance;     // Advance width and height in design units
+    sparse   m_attrs;
+};
+
+
+// Inlines: class GlyphFace
+//
+inline
+GlyphFace::GlyphFace()
+{}
+
+template<typename I>
+GlyphFace::GlyphFace(const Rect & bbox, const Position & adv, I first, const I last)
+: m_bbox(bbox),
+  m_advance(adv),
+  m_attrs(first, last)
+{
+}
+
+inline
+const Position & GlyphFace::theAdvance() const {
+    return m_advance;
+}
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Intervals.h b/thirdparty/graphite/src/inc/Intervals.h
new file mode 100644
index 0000000000..81d23187b6
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Intervals.h
@@ -0,0 +1,234 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <utility>
+
+#include "inc/Main.h"
+#include "inc/List.h"
+#include "inc/json.h"
+#include "inc/Position.h"
+
+// An IntervalSet represents the possible movement of a given glyph in a given direction
+// (horizontally, vertically, or diagonally).
+// A vector is needed to represent disjoint ranges, eg, -300..-150, 20..200, 500..750.
+// Each pair represents the min/max of a sub-range.
+
+namespace graphite2 {
+
+class Segment;
+
+enum zones_t {SD, XY};
+
+class Zones
+{
+    struct Exclusion
+    {
+        template<zones_t O>
+        static Exclusion weighted(float xmin, float xmax, float f, float a0,
+                float m, float xi, float ai, float c, bool nega);
+
+        float   x,  // x position
+                xm, // xmax position
+                c,  // constant + sum(MiXi^2)
+                sm, // sum(Mi)
+                smx; // sum(MiXi)
+        bool    open;
+
+        Exclusion(float x, float w, float smi, float smxi, float c);
+        Exclusion & operator += (Exclusion const & rhs);
+        uint8 outcode(float p) const;
+
+        Exclusion   split_at(float p);
+        void        left_trim(float p);
+
+        bool        track_cost(float & cost, float & x, float origin) const;
+
+    private:
+        float test_position(float origin) const;
+        float cost(float x) const;
+     };
+
+    typedef Vector<Exclusion>                   exclusions;
+
+    typedef exclusions::iterator                iterator;
+    typedef Exclusion *                         pointer;
+    typedef Exclusion &                         reference;
+    typedef std::reverse_iterator<iterator>     reverse_iterator;
+
+public:
+    typedef exclusions::const_iterator              const_iterator;
+    typedef Exclusion const *                       const_pointer;
+    typedef Exclusion const &                       const_reference;
+    typedef std::reverse_iterator<const_iterator>   const_reverse_iterator;
+
+#if !defined GRAPHITE2_NTRACING
+    struct Debug
+    {
+        Exclusion       _excl;
+        bool            _isdel;
+        Vector<void *>  _env;
+
+        Debug(Exclusion *e, bool isdel, json *dbg) : _excl(*e), _isdel(isdel), _env(dbg->getenvs()) { };
+    };
+
+    typedef Vector<Debug>                       debugs;
+    typedef debugs::const_iterator                    idebugs;
+    void addDebug(Exclusion *e);
+    void removeDebug(float pos, float posm);
+    void setdebug(json *dbgout) { _dbg = dbgout; }
+    idebugs dbgs_begin() const { return _dbgs.begin(); }
+    idebugs dbgs_end() const { return _dbgs.end(); }
+    void jsonDbgOut(Segment *seg) const;
+    Position position() const { return Position(_pos, _posm); }
+#endif
+
+    Zones();
+    template<zones_t O>
+    void initialise(float xmin, float xmax, float margin_len, float margin_weight, float ao);
+
+    void exclude(float xmin, float xmax);
+    void exclude_with_margins(float xmin, float xmax, int axis);
+
+    template<zones_t O>
+    void weighted(float xmin, float xmax, float f, float a0, float mi, float xi, float ai, float c, bool nega);
+    void weightedAxis(int axis, float xmin, float xmax, float f, float a0, float mi, float xi, float ai, float c, bool nega);
+
+    float closest( float origin, float &cost) const;
+
+    const_iterator begin() const { return _exclusions.begin(); }
+    const_iterator end() const { return _exclusions.end(); }
+
+private:
+    exclusions  _exclusions;
+#if !defined GRAPHITE2_NTRACING
+    json      * _dbg;
+    debugs      _dbgs;
+#endif
+    float       _margin_len,
+                _margin_weight,
+                _pos,
+                _posm;
+
+    void            insert(Exclusion e);
+    void            remove(float x, float xm);
+    const_iterator  find_exclusion_under(float x) const;
+};
+
+
+inline
+Zones::Zones()
+: _margin_len(0), _margin_weight(0), _pos(0), _posm(0)
+{
+#if !defined GRAPHITE2_NTRACING
+    _dbg = 0;
+#endif
+    _exclusions.reserve(8);
+}
+
+inline
+Zones::Exclusion::Exclusion(float x_, float xm_, float smi, float smxi, float c_)
+: x(x_), xm(xm_), c(c_), sm(smi), smx(smxi), open(false)
+{ }
+
+template<zones_t O>
+inline
+void Zones::initialise(float xmin, float xmax, float margin_len,
+        float margin_weight, float a0)
+{
+    _margin_len = margin_len;
+    _margin_weight = margin_weight;
+    _pos = xmin;
+    _posm = xmax;
+    _exclusions.clear();
+    _exclusions.push_back(Exclusion::weighted<O>(xmin, xmax, 1, a0, 0, 0, 0, 0, false));
+    _exclusions.front().open = true;
+#if !defined GRAPHITE2_NTRACING
+    _dbgs.clear();
+#endif
+}
+
+inline
+void Zones::exclude(float xmin, float xmax) {
+    remove(xmin, xmax);
+}
+
+template<zones_t O>
+inline
+void Zones::weighted(float xmin, float xmax, float f, float a0,
+        float m, float xi, float ai, float c, bool nega) {
+    insert(Exclusion::weighted<O>(xmin, xmax, f, a0, m, xi, ai, c, nega));
+}
+
+inline
+void Zones::weightedAxis(int axis, float xmin, float xmax, float f, float a0,
+        float m, float xi, float ai, float c, bool nega) {
+    if (axis < 2)
+        weighted<XY>(xmin, xmax, f, a0, m, xi, ai, c, nega);
+    else
+        weighted<SD>(xmin, xmax, f, a0, m, xi, ai, c, nega);
+}
+
+#if !defined GRAPHITE2_NTRACING
+inline
+void Zones::addDebug(Exclusion *e) {
+    if (_dbg)
+        _dbgs.push_back(Debug(e, false, _dbg));
+}
+
+inline
+void Zones::removeDebug(float pos, float posm) {
+    if (_dbg)
+    {
+        Exclusion e(pos, posm, 0, 0, 0);
+        _dbgs.push_back(Debug(&e, true, _dbg));
+    }
+}
+#endif
+
+template<>
+inline
+Zones::Exclusion Zones::Exclusion::weighted<XY>(float xmin, float xmax, float f, float a0,
+        float m, float xi, GR_MAYBE_UNUSED float ai, float c, GR_MAYBE_UNUSED bool nega) {
+    return Exclusion(xmin, xmax,
+            m + f,
+            m * xi,
+            m * xi * xi + f * a0 * a0 + c);
+}
+
+template<>
+inline
+Zones::Exclusion Zones::Exclusion::weighted<SD>(float xmin, float xmax, float f, float a0,
+        float m, float xi, float ai,float c, bool nega) {
+    float xia = nega ? xi - ai : xi + ai;
+    return Exclusion(xmin, xmax,
+            0.25f * (m + 2.f * f),
+            0.25f * m * xia,
+            0.25f * (m * xia * xia + 2.f * f * a0 * a0) + c);
+}
+
+} // end of namespace graphite2
diff --git a/thirdparty/graphite/src/inc/List.h b/thirdparty/graphite/src/inc/List.h
new file mode 100644
index 0000000000..a3b7a77961
--- /dev/null
+++ b/thirdparty/graphite/src/inc/List.h
@@ -0,0 +1,168 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+// designed to have a limited subset of the std::vector api
+#pragma once
+
+#include <cstddef>
+#include <cassert>
+#include <cstring>
+#include <cstdlib>
+#include <new>
+
+#include "Main.h"
+
+namespace graphite2 {
+
+template <typename T>
+inline
+ptrdiff_t distance(T* first, T* last) { return last-first; }
+
+
+template <typename T>
+class Vector
+{
+    T * m_first, *m_last, *m_end;
+public:
+    typedef       T &   reference;
+    typedef const T &   const_reference;
+    typedef       T *   iterator;
+    typedef const T *   const_iterator;
+
+    Vector() : m_first(0), m_last(0), m_end(0) {}
+    Vector(size_t n, const T& value = T())      : m_first(0), m_last(0), m_end(0) { insert(begin(), n, value); }
+    Vector(const Vector<T> &rhs)                : m_first(0), m_last(0), m_end(0) { insert(begin(), rhs.begin(), rhs.end()); }
+    template <typename I>
+    Vector(I first, const I last)               : m_first(0), m_last(0), m_end(0) { insert(begin(), first, last); }
+    ~Vector() { clear(); free(m_first); }
+
+    iterator            begin()         { return m_first; }
+    const_iterator      begin() const   { return m_first; }
+
+    iterator            end()           { return m_last; }
+    const_iterator      end() const     { return m_last; }
+
+    bool                empty() const   { return m_first == m_last; }
+    size_t              size() const    { return m_last - m_first; }
+    size_t              capacity() const{ return m_end - m_first; }
+
+    void                reserve(size_t n);
+    void                resize(size_t n, const T & v = T());
+
+    reference           front()         { assert(size() > 0); return *begin(); }
+    const_reference     front() const   { assert(size() > 0); return *begin(); }
+    reference           back()          { assert(size() > 0); return *(end()-1); }
+    const_reference     back() const    { assert(size() > 0); return *(end()-1); }
+
+    Vector<T>         & operator = (const Vector<T> & rhs) { assign(rhs.begin(), rhs.end()); return *this; }
+    reference           operator [] (size_t n)          { assert(size() > n); return m_first[n]; }
+    const_reference     operator [] (size_t n) const    { assert(size() > n); return m_first[n]; }
+
+    void                assign(size_t n, const T& u)    { clear(); insert(begin(), n, u); }
+    void                assign(const_iterator first, const_iterator last)      { clear(); insert(begin(), first, last); }
+    iterator            insert(iterator p, const T & x) { p = _insert_default(p, 1); new (p) T(x); return p; }
+    void                insert(iterator p, size_t n, const T & x);
+    void                insert(iterator p, const_iterator first, const_iterator last);
+    void                pop_back()              { assert(size() > 0); --m_last; }
+    void                push_back(const T &v)   { if (m_last == m_end) reserve(size()+1); new (m_last++) T(v); }
+
+    void                clear()                 { erase(begin(), end()); }
+    iterator            erase(iterator p)       { return erase(p, p+1); }
+    iterator            erase(iterator first, iterator last);
+
+private:
+    iterator            _insert_default(iterator p, size_t n);
+};
+
+template <typename T>
+inline
+void Vector<T>::reserve(size_t n)
+{
+    if (n > capacity())
+    {
+        const ptrdiff_t sz = size();
+        size_t requested;
+        if (checked_mul(n,sizeof(T), requested))  std::abort();
+        m_first = static_cast<T*>(realloc(m_first, requested));
+        if (!m_first)   std::abort();
+        m_last  = m_first + sz;
+        m_end   = m_first + n;
+    }
+}
+
+template <typename T>
+inline
+void Vector<T>::resize(size_t n, const T & v) {
+    const ptrdiff_t d = n-size();
+    if (d < 0)      erase(end()+d, end());
+    else if (d > 0) insert(end(), d, v);
+}
+
+template<typename T>
+inline
+typename Vector<T>::iterator Vector<T>::_insert_default(iterator p, size_t n)
+{
+    assert(begin() <= p && p <= end());
+    const ptrdiff_t i = p - begin();
+    reserve(((size() + n + 7) >> 3) << 3);
+    p = begin() + i;
+    // Move tail if there is one
+    if (p != end()) memmove(p + n, p, distance(p,end())*sizeof(T));
+    m_last += n;
+    return p;
+}
+
+template<typename T>
+inline
+void Vector<T>::insert(iterator p, size_t n, const T & x)
+{
+    p = _insert_default(p, n);
+    // Copy in elements
+    for (; n; --n, ++p) { new (p) T(x); }
+}
+
+template<typename T>
+inline
+void Vector<T>::insert(iterator p, const_iterator first, const_iterator last)
+{
+    p = _insert_default(p, distance(first, last));
+    // Copy in elements
+    for (;first != last; ++first, ++p) { new (p) T(*first); }
+}
+
+template<typename T>
+inline
+typename Vector<T>::iterator Vector<T>::erase(iterator first, iterator last)
+{
+    for (iterator e = first; e != last; ++e) e->~T();
+    const size_t sz = distance(first, last);
+    if (m_last != last) memmove(first, last, distance(last,end())*sizeof(T));
+    m_last -= sz;
+    return first;
+}
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Machine.h b/thirdparty/graphite/src/inc/Machine.h
new file mode 100644
index 0000000000..b23819fb98
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Machine.h
@@ -0,0 +1,207 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This general interpreter interface.
+// Author: Tim Eves
+
+// Build one of direct_machine.cpp or call_machine.cpp to implement this
+// interface.
+
+#pragma once
+#include <cstring>
+#include <limits>
+#include <graphite2/Types.h>
+#include "inc/Main.h"
+
+#if defined(__GNUC__)
+#if defined(__clang__) || (__GNUC__ * 100 + __GNUC_MINOR__ * 10) < 430
+#define     HOT
+#if defined(__x86_64)
+#define     REGPARM(n)      __attribute__((regparm(n)))
+#else
+#define     REGPARM(n)
+#endif
+#else
+#define     HOT             __attribute__((hot))
+#if defined(__x86_64)
+#define     REGPARM(n)      __attribute__((hot, regparm(n)))
+#else
+#define     REGPARM(n)
+#endif
+#endif
+#else
+#define     HOT
+#define     REGPARM(n)
+#endif
+
+#if defined(__MINGW32__)
+// MinGW's <limits> at some point includes winnt.h which #define's a
+// DELETE macro, which conflicts with enum opcode below, so we undefine
+// it here.
+#undef DELETE
+#endif
+
+namespace graphite2 {
+
+// Forward declarations
+class Segment;
+class Slot;
+class SlotMap;
+
+
+namespace vm
+{
+
+
+typedef void * instr;
+typedef Slot * slotref;
+
+enum {VARARGS = 0xff, MAX_NAME_LEN=32};
+
+enum opcode {
+    NOP = 0,
+
+    PUSH_BYTE,      PUSH_BYTEU,     PUSH_SHORT,     PUSH_SHORTU,    PUSH_LONG,
+
+    ADD,            SUB,            MUL,            DIV,
+    MIN_,           MAX_,
+    NEG,
+    TRUNC8,         TRUNC16,
+
+    COND,
+
+    AND,            OR,             NOT,
+    EQUAL,          NOT_EQ,
+    LESS,           GTR,            LESS_EQ,        GTR_EQ,
+
+    NEXT,           NEXT_N,         COPY_NEXT,
+    PUT_GLYPH_8BIT_OBS,              PUT_SUBS_8BIT_OBS,   PUT_COPY,
+    INSERT,         DELETE,
+    ASSOC,
+    CNTXT_ITEM,
+
+    ATTR_SET,       ATTR_ADD,       ATTR_SUB,
+    ATTR_SET_SLOT,
+    IATTR_SET_SLOT,
+    PUSH_SLOT_ATTR,                 PUSH_GLYPH_ATTR_OBS,
+    PUSH_GLYPH_METRIC,              PUSH_FEAT,
+    PUSH_ATT_TO_GATTR_OBS,          PUSH_ATT_TO_GLYPH_METRIC,
+    PUSH_ISLOT_ATTR,
+
+    PUSH_IGLYPH_ATTR,    // not implemented
+
+    POP_RET,                        RET_ZERO,           RET_TRUE,
+    IATTR_SET,                      IATTR_ADD,          IATTR_SUB,
+    PUSH_PROC_STATE,                PUSH_VERSION,
+    PUT_SUBS,                       PUT_SUBS2,          PUT_SUBS3,
+    PUT_GLYPH,                      PUSH_GLYPH_ATTR,    PUSH_ATT_TO_GLYPH_ATTR,
+    BITOR,                          BITAND,             BITNOT,
+    BITSET,                         SET_FEAT,
+    MAX_OPCODE,
+    // private opcodes for internal use only, comes after all other on disk opcodes
+    TEMP_COPY = MAX_OPCODE
+};
+
+struct opcode_t
+{
+    instr           impl[2];
+    uint8           param_sz;
+    char            name[MAX_NAME_LEN];
+};
+
+
+class Machine
+{
+public:
+    typedef int32  stack_t;
+    static size_t const STACK_ORDER  = 10,
+                        STACK_MAX    = 1 << STACK_ORDER,
+                        STACK_GUARD  = 2;
+
+    class Code;
+
+    enum status_t {
+        finished = 0,
+        stack_underflow,
+        stack_not_empty,
+        stack_overflow,
+        slot_offset_out_bounds,
+        died_early
+    };
+
+    Machine(SlotMap &) throw();
+    static const opcode_t *   getOpcodeTable() throw();
+
+    CLASS_NEW_DELETE;
+
+    SlotMap   & slotMap() const throw();
+    status_t    status() const throw();
+//    operator bool () const throw();
+
+private:
+    void    check_final_stack(const stack_t * const sp);
+    stack_t run(const instr * program, const byte * data,
+                slotref * & map) HOT;
+
+    SlotMap       & _map;
+    stack_t         _stack[STACK_MAX + 2*STACK_GUARD];
+    status_t        _status;
+};
+
+inline Machine::Machine(SlotMap & map) throw()
+: _map(map), _status(finished)
+{
+    // Initialise stack guard +1 entries as the stack pointer points to the
+    //  current top of stack, hence the first push will never write entry 0.
+    // Initialising the guard space like this is unnecessary and is only
+    //  done to keep valgrind happy during fuzz testing.  Hopefully loop
+    //  unrolling will flatten this.
+    for (size_t n = STACK_GUARD + 1; n; --n)  _stack[n-1] = 0;
+}
+
+inline SlotMap& Machine::slotMap() const throw()
+{
+    return _map;
+}
+
+inline Machine::status_t Machine::status() const throw()
+{
+    return _status;
+}
+
+inline void Machine::check_final_stack(const stack_t * const sp)
+{
+    if (_status != finished) return;
+
+    stack_t const * const base  = _stack + STACK_GUARD,
+                  * const limit = base + STACK_MAX;
+    if      (sp <  base)    _status = stack_underflow;       // This should be impossible now.
+    else if (sp >= limit)   _status = stack_overflow;        // So should this.
+    else if (sp != base)    _status = stack_not_empty;
+}
+
+} // namespace vm
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Main.h b/thirdparty/graphite/src/inc/Main.h
new file mode 100644
index 0000000000..ebf02dd553
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Main.h
@@ -0,0 +1,199 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <cstdlib>
+#include "graphite2/Types.h"
+
+#ifdef GRAPHITE2_CUSTOM_HEADER
+#include GRAPHITE2_CUSTOM_HEADER
+#endif
+
+namespace graphite2 {
+
+typedef gr_uint8        uint8;
+typedef gr_uint8        byte;
+typedef gr_uint16       uint16;
+typedef gr_uint32       uint32;
+typedef gr_int8         int8;
+typedef gr_int16        int16;
+typedef gr_int32        int32;
+typedef size_t          uintptr;
+
+#ifdef GRAPHITE2_TELEMETRY
+struct telemetry
+{
+    class category;
+
+    static size_t   * _category;
+    static void set_category(size_t & t) throw()    { _category = &t; }
+    static void stop() throw()                      { _category = 0; }
+    static void count_bytes(size_t n) throw()       { if (_category) *_category += n; }
+
+    size_t  misc,
+            silf,
+            glyph,
+            code,
+            states,
+            starts,
+            transitions;
+
+    telemetry() : misc(0), silf(0), glyph(0), code(0), states(0), starts(0), transitions(0) {}
+};
+
+class telemetry::category
+{
+    size_t * _prev;
+public:
+    category(size_t & t) : _prev(_category) { _category = &t; }
+    ~category() { _category = _prev; }
+};
+
+#else
+struct telemetry  {};
+#endif
+
+// Checked multiplaction to catch overflow or underflow when allocating memory
+#if defined(__has_builtin)
+  #if __has_builtin(__builtin_mul_overflow)
+    #define HAVE_BUILTIN_OVERFLOW
+  #endif
+#elif defined(__GNUC__) && (__GNUC__ >= 5) && !defined(__INTEL_COMPILER)
+  #define HAVE_BUILTIN_OVERFLOW
+#endif
+#if defined(__has_include)
+  #if __has_include(<intsafe.h>) && !defined(__CYGWIN__)
+    #define HAVE_INTSAFE_H
+  #endif
+#elif defined(_WIN32)
+  #define HAVE_INTSAFE_H
+#endif
+
+// Need to import intsafe into the top level namespace
+#if defined(HAVE_INTSAFE_H)
+} // namespace graphite2
+
+#include <intsafe.h>
+
+namespace graphite2 {
+#endif
+
+#if defined(HAVE_BUILTIN_OVERFLOW)
+inline
+bool checked_mul(const size_t a, const size_t b, size_t & t) {
+    return __builtin_mul_overflow(a, b, &t);
+}
+#elif defined(HAVE_INTSAFE_H)
+inline
+bool checked_mul(const size_t a, const size_t b, size_t & t) {
+    return SizeTMult(a, b, &t) == INTSAFE_E_ARITHMETIC_OVERFLOW;
+}
+#else
+inline
+bool checked_mul(const size_t a, const size_t b, size_t & t) {
+  t = a*b;
+  return (((a | b) & (~size_t(0) << (sizeof(size_t) << 2))) && (t / a != b));
+}
+#endif
+
+// typesafe wrapper around malloc for simple types
+// use free(pointer) to deallocate
+
+template <typename T> T * gralloc(size_t n)
+{
+    size_t total;
+    if (checked_mul(n, sizeof(T), total))
+      return 0;
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::count_bytes(total);
+#endif
+    return static_cast<T*>(malloc(total));
+}
+
+template <typename T> T * grzeroalloc(size_t n)
+{
+#ifdef GRAPHITE2_TELEMETRY
+    telemetry::count_bytes(sizeof(T) * n);
+#endif
+    return static_cast<T*>(calloc(n, sizeof(T)));
+}
+
+template <typename T>
+inline T min(const T a, const T b)
+{
+    return a < b ? a : b;
+}
+
+template <typename T>
+inline T max(const T a, const T b)
+{
+    return a > b ? a : b;
+}
+
+} // namespace graphite2
+
+#define CLASS_NEW_DELETE \
+    void * operator new   (size_t size){ return gralloc<byte>(size);} \
+    void * operator new   (size_t, void * p) throw() { return p; } \
+    void * operator new[] (size_t size) {return gralloc<byte>(size);} \
+    void * operator new[] (size_t, void * p) throw() { return p; } \
+    void operator delete   (void * p) throw() { free(p);} \
+    void operator delete   (void *, void *) throw() {} \
+    void operator delete[] (void * p)throw() { free(p); } \
+    void operator delete[] (void *, void *) throw() {}
+
+#if defined(__GNUC__)  || defined(__clang__)
+#define GR_MAYBE_UNUSED __attribute__((unused))
+#else
+#define GR_MAYBE_UNUSED
+#endif
+
+#ifndef __has_cpp_attribute
+#  define __has_cpp_attribute(x) 0
+#endif
+
+#if __has_cpp_attribute(clang::fallthrough)
+#  define GR_FALLTHROUGH [[clang::fallthrough]]
+#elif __has_cpp_attribute(gnu::fallthrough)
+#  define GR_FALLTHROUGH [[gnu::fallthrough]]
+#elif defined(_MSC_VER)
+   /*
+    * MSVC's __fallthrough annotations are checked by /analyze (Code Analysis):
+    * https://msdn.microsoft.com/en-us/library/ms235402%28VS.80%29.aspx
+    */
+    #include <sal.h>
+    #define GR_FALLTHROUGH __fallthrough
+#elif __GNUC__ >= 7
+    #define GR_FALLTHROUGH __attribute__ ((fallthrough))
+#else
+    #define GR_FALLTHROUGH /* fallthrough */
+#endif
+
+#ifdef _MSC_VER
+#pragma warning(disable: 4800)
+#pragma warning(disable: 4355)
+#endif
diff --git a/thirdparty/graphite/src/inc/NameTable.h b/thirdparty/graphite/src/inc/NameTable.h
new file mode 100644
index 0000000000..0fdbeb4d85
--- /dev/null
+++ b/thirdparty/graphite/src/inc/NameTable.h
@@ -0,0 +1,65 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <graphite2/Segment.h>
+#include "inc/TtfTypes.h"
+#include "inc/locale2lcid.h"
+
+namespace graphite2 {
+
+class NameTable
+{
+    NameTable(const NameTable &);
+    NameTable & operator = (const NameTable &);
+
+public:
+    NameTable(const void * data, size_t length, uint16 platfromId=3, uint16 encodingID = 1);
+    ~NameTable() { free(const_cast<TtfUtil::Sfnt::FontNames *>(m_table)); }
+    enum eNameFallback {
+        eNoFallback = 0,
+        eEnUSFallbackOnly = 1,
+        eEnOrAnyFallback = 2
+    };
+    uint16 setPlatformEncoding(uint16 platfromId=3, uint16 encodingID = 1);
+    void * getName(uint16 & languageId, uint16 nameId, gr_encform enc, uint32 & length);
+    uint16 getLanguageId(const char * bcp47Locale);
+
+    CLASS_NEW_DELETE
+private:
+    uint16 m_platformId;
+    uint16 m_encodingId;
+    uint16 m_languageCount;
+    uint16 m_platformOffset; // offset of first NameRecord with for platform 3, encoding 1
+    uint16 m_platformLastRecord;
+    uint16 m_nameDataLength;
+    const TtfUtil::Sfnt::FontNames * m_table;
+    const uint8 * m_nameData;
+    Locale2Lang m_locale2Lang;
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Pass.h b/thirdparty/graphite/src/inc/Pass.h
new file mode 100644
index 0000000000..e687a87d8c
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Pass.h
@@ -0,0 +1,118 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <cstdlib>
+#include "inc/Code.h"
+
+namespace graphite2 {
+
+class Segment;
+class Face;
+class Silf;
+struct Rule;
+struct RuleEntry;
+struct State;
+class FiniteStateMachine;
+class Error;
+class ShiftCollider;
+class KernCollider;
+class json;
+
+enum passtype;
+
+class Pass
+{
+public:
+    Pass();
+    ~Pass();
+
+    bool readPass(const byte * pPass, size_t pass_length, size_t subtable_base, Face & face,
+        enum passtype pt, uint32 version, Error &e);
+    bool runGraphite(vm::Machine & m, FiniteStateMachine & fsm, bool reverse) const;
+    void init(Silf *silf) { m_silf = silf; }
+    byte collisionLoops() const { return m_numCollRuns; }
+    bool reverseDir() const { return m_isReverseDir; }
+
+    CLASS_NEW_DELETE
+private:
+    void    findNDoRule(Slot* & iSlot, vm::Machine &, FiniteStateMachine& fsm) const;
+    int     doAction(const vm::Machine::Code* codeptr, Slot * & slot_out, vm::Machine &) const;
+    bool    testPassConstraint(vm::Machine & m) const;
+    bool    testConstraint(const Rule & r, vm::Machine &) const;
+    bool    readRules(const byte * rule_map, const size_t num_entries,
+                     const byte *precontext, const uint16 * sort_key,
+                     const uint16 * o_constraint, const byte *constraint_data,
+                     const uint16 * o_action, const byte * action_data,
+                     Face &, enum passtype pt, Error &e);
+    bool    readStates(const byte * starts, const byte * states, const byte * o_rule_map, Face &, Error &e);
+    bool    readRanges(const byte * ranges, size_t num_ranges, Error &e);
+    uint16  glyphToCol(const uint16 gid) const;
+    bool    runFSM(FiniteStateMachine & fsm, Slot * slot) const;
+    void    dumpRuleEventConsidered(const FiniteStateMachine & fsm, const RuleEntry & re) const;
+    void    dumpRuleEventOutput(const FiniteStateMachine & fsm, const Rule & r, Slot * os) const;
+    void    adjustSlot(int delta, Slot * & slot_out, SlotMap &) const;
+    bool    collisionShift(Segment *seg, int dir, json * const dbgout) const;
+    bool    collisionKern(Segment *seg, int dir, json * const dbgout) const;
+    bool    collisionFinish(Segment *seg, GR_MAYBE_UNUSED json * const dbgout) const;
+    bool    resolveCollisions(Segment *seg, Slot *slot, Slot *start, ShiftCollider &coll, bool isRev,
+                     int dir, bool &moved, bool &hasCol, json * const dbgout) const;
+    float   resolveKern(Segment *seg, Slot *slot, Slot *start, int dir,
+                     float &ymin, float &ymax, json *const dbgout) const;
+
+    const Silf        * m_silf;
+    uint16            * m_cols;
+    Rule              * m_rules; // rules
+    RuleEntry         * m_ruleMap;
+    uint16            * m_startStates; // prectxt length
+    uint16            * m_transitions;
+    State             * m_states;
+    vm::Machine::Code * m_codes;
+    byte              * m_progs;
+
+    byte   m_numCollRuns;
+    byte   m_kernColls;
+    byte   m_iMaxLoop;
+    uint16 m_numGlyphs;
+    uint16 m_numRules;
+    uint16 m_numStates;
+    uint16 m_numTransition;
+    uint16 m_numSuccess;
+    uint16 m_successStart;
+    uint16 m_numColumns;
+    byte m_minPreCtxt;
+    byte m_maxPreCtxt;
+    byte m_colThreshold;
+    bool m_isReverseDir;
+    vm::Machine::Code m_cPConstraint;
+
+private:        //defensive
+    Pass(const Pass&);
+    Pass& operator=(const Pass&);
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Position.h b/thirdparty/graphite/src/inc/Position.h
new file mode 100644
index 0000000000..510e4f4c41
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Position.h
@@ -0,0 +1,68 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+namespace graphite2 {
+
+class Position
+{
+public:
+    Position() : x(0), y(0) { }
+    Position(const float inx, const float iny) : x(inx), y(iny) {}
+    Position operator + (const Position& a) const { return Position(x + a.x, y + a.y); }
+    Position operator - (const Position& a) const { return Position(x - a.x, y - a.y); }
+    Position operator * (const float m) const { return Position(x * m, y * m); }
+    Position &operator += (const Position &a) { x += a.x; y += a.y; return *this; }
+    Position &operator *= (const float m) { x *= m; y *= m; return *this; }
+
+    float x;
+    float y;
+};
+
+class Rect
+{
+public :
+    Rect() {}
+    Rect(const Position& botLeft, const Position& topRight): bl(botLeft), tr(topRight) {}
+    Rect widen(const Rect& other) { return Rect(Position(bl.x > other.bl.x ? other.bl.x : bl.x, bl.y > other.bl.y ? other.bl.y : bl.y), Position(tr.x > other.tr.x ? tr.x : other.tr.x, tr.y > other.tr.y ? tr.y : other.tr.y)); }
+    Rect operator + (const Position &a) const { return Rect(Position(bl.x + a.x, bl.y + a.y), Position(tr.x + a.x, tr.y + a.y)); }
+    Rect operator - (const Position &a) const { return Rect(Position(bl.x - a.x, bl.y - a.y), Position(tr.x - a.x, tr.y - a.y)); }
+    Rect operator * (float m) const { return Rect(Position(bl.x, bl.y) * m, Position(tr.x, tr.y) * m); }
+    float width() const { return tr.x - bl.x; }
+    float height() const { return tr.y - bl.y; }
+
+    bool hitTest(Rect &other);
+
+    // returns Position(overlapx, overlapy) where overlap<0 if overlapping else positive)
+    Position overlap(Position &offset, Rect &other, Position &otherOffset);
+    //Position constrainedAvoid(Position &offset, Rect &box, Rect &sdbox, Position &other, Rect &obox, Rect &osdbox);
+
+    Position bl;
+    Position tr;
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Rule.h b/thirdparty/graphite/src/inc/Rule.h
new file mode 100644
index 0000000000..5964e003a6
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Rule.h
@@ -0,0 +1,305 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+
+#pragma once
+
+#include "inc/Code.h"
+#include "inc/Slot.h"
+
+namespace graphite2 {
+
+struct Rule {
+  const vm::Machine::Code * constraint,
+                 * action;
+  unsigned short   sort;
+  byte             preContext;
+#ifndef NDEBUG
+  uint16           rule_idx;
+#endif
+
+  Rule();
+  ~Rule() {}
+
+  CLASS_NEW_DELETE;
+
+private:
+  Rule(const Rule &);
+  Rule & operator = (const Rule &);
+};
+
+inline
+Rule::Rule()
+: constraint(0),
+  action(0),
+  sort(0),
+  preContext(0)
+{
+#ifndef NDEBUG
+  rule_idx = 0;
+#endif
+}
+
+
+struct RuleEntry
+{
+  const Rule   * rule;
+
+  inline
+  bool operator < (const RuleEntry &r) const
+  {
+    const unsigned short lsort = rule->sort, rsort = r.rule->sort;
+    return lsort > rsort || (lsort == rsort && rule < r.rule);
+  }
+
+  inline
+  bool operator == (const RuleEntry &r) const
+  {
+    return rule == r.rule;
+  }
+};
+
+
+struct State
+{
+  const RuleEntry     * rules,
+                      * rules_end;
+
+  bool   empty() const;
+};
+
+inline
+bool State::empty() const
+{
+    return rules_end == rules;
+}
+
+
+class SlotMap
+{
+public:
+  enum {MAX_SLOTS=64};
+  SlotMap(Segment & seg, uint8 direction, size_t maxSize);
+
+  Slot       * * begin();
+  Slot       * * end();
+  size_t         size() const;
+  unsigned short context() const;
+  void           reset(Slot &, unsigned short);
+
+  Slot * const & operator[](int n) const;
+  Slot       * & operator [] (int);
+  void           pushSlot(Slot * const slot);
+  void           collectGarbage(Slot *& aSlot);
+
+  Slot         * highwater() { return m_highwater; }
+  void           highwater(Slot *s) { m_highwater = s; m_highpassed = false; }
+  bool           highpassed() const { return m_highpassed; }
+  void           highpassed(bool v) { m_highpassed = v; }
+
+  uint8          dir() const { return m_dir; }
+  int            decMax() { return --m_maxSize; }
+
+  Segment &    segment;
+private:
+  Slot         * m_slot_map[MAX_SLOTS+1];
+  unsigned short m_size;
+  unsigned short m_precontext;
+  Slot         * m_highwater;
+  int            m_maxSize;
+  uint8          m_dir;
+  bool           m_highpassed;
+};
+
+
+class FiniteStateMachine
+{
+public:
+  enum {MAX_RULES=128};
+
+private:
+  class Rules
+  {
+  public:
+      Rules();
+      void              clear();
+      const RuleEntry * begin() const;
+      const RuleEntry * end() const;
+      size_t            size() const;
+
+      void accumulate_rules(const State &state);
+
+  private:
+      RuleEntry * m_begin,
+                * m_end,
+                  m_rules[MAX_RULES*2];
+  };
+
+public:
+  FiniteStateMachine(SlotMap & map, json * logger);
+  void      reset(Slot * & slot, const short unsigned int max_pre_ctxt);
+
+  Rules     rules;
+  SlotMap   & slots;
+  json    * const dbgout;
+};
+
+
+inline
+FiniteStateMachine::FiniteStateMachine(SlotMap& map, json * logger)
+: slots(map),
+  dbgout(logger)
+{
+}
+
+inline
+void FiniteStateMachine::reset(Slot * & slot, const short unsigned int max_pre_ctxt)
+{
+  rules.clear();
+  int ctxt = 0;
+  for (; ctxt != max_pre_ctxt && slot->prev(); ++ctxt, slot = slot->prev());
+  slots.reset(*slot, ctxt);
+}
+
+inline
+FiniteStateMachine::Rules::Rules()
+  : m_begin(m_rules), m_end(m_rules)
+{
+}
+
+inline
+void FiniteStateMachine::Rules::clear()
+{
+  m_end = m_begin;
+}
+
+inline
+const RuleEntry * FiniteStateMachine::Rules::begin() const
+{
+  return m_begin;
+}
+
+inline
+const RuleEntry * FiniteStateMachine::Rules::end() const
+{
+  return m_end;
+}
+
+inline
+size_t FiniteStateMachine::Rules::size() const
+{
+  return m_end - m_begin;
+}
+
+inline
+void FiniteStateMachine::Rules::accumulate_rules(const State &state)
+{
+  // Only bother if there are rules in the State object.
+  if (state.empty()) return;
+
+  // Merge the new sorted rules list into the current sorted result set.
+  const RuleEntry * lre = begin(), * rre = state.rules;
+  RuleEntry * out = m_rules + (m_begin == m_rules)*MAX_RULES;
+  const RuleEntry * const lrend = out + MAX_RULES,
+                  * const rrend = state.rules_end;
+  m_begin = out;
+  while (lre != end() && out != lrend)
+  {
+    if (*lre < *rre)      *out++ = *lre++;
+    else if (*rre < *lre) { *out++ = *rre++; }
+    else                { *out++ = *lre++; ++rre; }
+
+    if (rre == rrend)
+    {
+      while (lre != end() && out != lrend) { *out++ = *lre++; }
+      m_end = out;
+      return;
+    }
+  }
+  while (rre != rrend && out != lrend) { *out++ = *rre++; }
+  m_end = out;
+}
+
+inline
+SlotMap::SlotMap(Segment & seg, uint8 direction, size_t maxSize)
+: segment(seg), m_size(0), m_precontext(0), m_highwater(0),
+    m_maxSize(int(maxSize)), m_dir(direction), m_highpassed(false)
+{
+    m_slot_map[0] = 0;
+}
+
+inline
+Slot * * SlotMap::begin()
+{
+  return &m_slot_map[1]; // allow map to go 1 before slot_map when inserting
+                         // at start of segment.
+}
+
+inline
+Slot * * SlotMap::end()
+{
+  return m_slot_map + m_size + 1;
+}
+
+inline
+size_t SlotMap::size() const
+{
+  return m_size;
+}
+
+inline
+short unsigned int SlotMap::context() const
+{
+  return m_precontext;
+}
+
+inline
+void SlotMap::reset(Slot & slot, short unsigned int ctxt)
+{
+  m_size = 0;
+  m_precontext = ctxt;
+  *m_slot_map = slot.prev();
+}
+
+inline
+void SlotMap::pushSlot(Slot*const slot)
+{
+  m_slot_map[++m_size] = slot;
+}
+
+inline
+Slot * const & SlotMap::operator[](int n) const
+{
+  return m_slot_map[n + 1];
+}
+
+inline
+Slot * & SlotMap::operator[](int n)
+{
+  return m_slot_map[n + 1];
+}
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Segment.h b/thirdparty/graphite/src/inc/Segment.h
new file mode 100644
index 0000000000..6cf83408d4
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Segment.h
@@ -0,0 +1,236 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "inc/Main.h"
+
+#include <cassert>
+
+#include "inc/CharInfo.h"
+#include "inc/Face.h"
+#include "inc/FeatureVal.h"
+#include "inc/GlyphCache.h"
+#include "inc/GlyphFace.h"
+#include "inc/Slot.h"
+#include "inc/Position.h"
+#include "inc/List.h"
+#include "inc/Collider.h"
+
+#define MAX_SEG_GROWTH_FACTOR  64
+
+namespace graphite2 {
+
+typedef Vector<Features>        FeatureList;
+typedef Vector<Slot *>          SlotRope;
+typedef Vector<int16 *>         AttributeRope;
+typedef Vector<SlotJustify *>   JustifyRope;
+
+class Font;
+class Segment;
+class Silf;
+
+enum SpliceParam {
+/** sub-Segments longer than this are not cached
+ * (in Unicode code points) */
+    eMaxSpliceSize = 96
+};
+
+enum justFlags {
+    gr_justStartInline = 1,
+    gr_justEndInline = 2
+};
+
+class SegmentScopeState
+{
+private:
+    friend class Segment;
+    Slot * realFirstSlot;
+    Slot * slotBeforeScope;
+    Slot * slotAfterScope;
+    Slot * realLastSlot;
+    size_t numGlyphsOutsideScope;
+};
+
+class Segment
+{
+    // Prevent copying of any kind.
+    Segment(const Segment&);
+    Segment& operator=(const Segment&);
+
+public:
+
+    enum {
+        SEG_INITCOLLISIONS = 1,
+        SEG_HASCOLLISIONS = 2
+    };
+
+    size_t slotCount() const { return m_numGlyphs; }      //one slot per glyph
+    void extendLength(ptrdiff_t num) { m_numGlyphs += num; }
+    Position advance() const { return m_advance; }
+    bool runGraphite() { if (m_silf) return m_face->runGraphite(this, m_silf); else return true;};
+    void chooseSilf(uint32 script) { m_silf = m_face->chooseSilf(script); }
+    const Silf *silf() const { return m_silf; }
+    size_t charInfoCount() const { return m_numCharinfo; }
+    const CharInfo *charinfo(unsigned int index) const { return index < m_numCharinfo ? m_charinfo + index : NULL; }
+    CharInfo *charinfo(unsigned int index) { return index < m_numCharinfo ? m_charinfo + index : NULL; }
+
+    Segment(size_t numchars, const Face* face, uint32 script, int dir);
+    ~Segment();
+    uint8 flags() const { return m_flags; }
+    void flags(uint8 f) { m_flags = f; }
+    Slot *first() { return m_first; }
+    void first(Slot *p) { m_first = p; }
+    Slot *last() { return m_last; }
+    void last(Slot *p) { m_last = p; }
+    void appendSlot(int i, int cid, int gid, int fid, size_t coffset);
+    Slot *newSlot();
+    void freeSlot(Slot *);
+    SlotJustify *newJustify();
+    void freeJustify(SlotJustify *aJustify);
+    Position positionSlots(const Font *font=0, Slot *first=0, Slot *last=0, bool isRtl = false, bool isFinal = true);
+    void associateChars(int offset, size_t num);
+    void linkClusters(Slot *first, Slot *last);
+    uint16 getClassGlyph(uint16 cid, uint16 offset) const { return m_silf->getClassGlyph(cid, offset); }
+    uint16 findClassIndex(uint16 cid, uint16 gid) const { return m_silf->findClassIndex(cid, gid); }
+    int addFeatures(const Features& feats) { m_feats.push_back(feats); return int(m_feats.size()) - 1; }
+    uint32 getFeature(int index, uint8 findex) const { const FeatureRef* pFR=m_face->theSill().theFeatureMap().featureRef(findex); if (!pFR) return 0; else return pFR->getFeatureVal(m_feats[index]); }
+    void setFeature(int index, uint8 findex, uint32 val) {
+        const FeatureRef* pFR=m_face->theSill().theFeatureMap().featureRef(findex);
+        if (pFR)
+        {
+            if (val > pFR->maxVal()) val = pFR->maxVal();
+            pFR->applyValToFeature(val, m_feats[index]);
+        } }
+    int8 dir() const { return m_dir; }
+    void dir(int8 val) { m_dir = val; }
+    bool currdir() const { return ((m_dir >> 6) ^ m_dir) & 1; }
+    uint8 passBits() const { return m_passBits; }
+    void mergePassBits(const uint8 val) { m_passBits &= val; }
+    int16 glyphAttr(uint16 gid, uint16 gattr) const { const GlyphFace * p = m_face->glyphs().glyphSafe(gid); return p ? p->attrs()[gattr] : 0; }
+    int32 getGlyphMetric(Slot *iSlot, uint8 metric, uint8 attrLevel, bool rtl) const;
+    float glyphAdvance(uint16 gid) const { return m_face->glyphs().glyph(gid)->theAdvance().x; }
+    const Rect &theGlyphBBoxTemporary(uint16 gid) const { return m_face->glyphs().glyph(gid)->theBBox(); }   //warning value may become invalid when another glyph is accessed
+    Slot *findRoot(Slot *is) const { return is->attachedTo() ? findRoot(is->attachedTo()) : is; }
+    int numAttrs() const { return m_silf->numUser(); }
+    int defaultOriginal() const { return m_defaultOriginal; }
+    const Face * getFace() const { return m_face; }
+    const Features & getFeatures(unsigned int /*charIndex*/) { assert(m_feats.size() == 1); return m_feats[0]; }
+    void bidiPass(int paradir, uint8 aMirror);
+    int8 getSlotBidiClass(Slot *s) const;
+    void doMirror(uint16 aMirror);
+    Slot *addLineEnd(Slot *nSlot);
+    void delLineEnd(Slot *s);
+    bool hasJustification() const { return m_justifies.size() != 0; }
+    void reverseSlots();
+
+    bool isWhitespace(const int cid) const;
+    bool hasCollisionInfo() const { return (m_flags & SEG_HASCOLLISIONS) && m_collisions; }
+    SlotCollision *collisionInfo(const Slot *s) const { return m_collisions ? m_collisions + s->index() : 0; }
+    CLASS_NEW_DELETE
+
+public:       //only used by: GrSegment* makeAndInitialize(const GrFont *font, const GrFace *face, uint32 script, const FeaturesHandle& pFeats/*must not be IsNull*/, encform enc, const void* pStart, size_t nChars, int dir);
+    bool read_text(const Face *face, const Features* pFeats/*must not be NULL*/, gr_encform enc, const void*pStart, size_t nChars);
+    void finalise(const Font *font, bool reverse=false);
+    float justify(Slot *pSlot, const Font *font, float width, enum justFlags flags, Slot *pFirst, Slot *pLast);
+    bool initCollisions();
+
+private:
+    Position        m_advance;          // whole segment advance
+    SlotRope        m_slots;            // Vector of slot buffers
+    AttributeRope   m_userAttrs;        // Vector of userAttrs buffers
+    JustifyRope     m_justifies;        // Slot justification info buffers
+    FeatureList     m_feats;            // feature settings referenced by charinfos in this segment
+    Slot          * m_freeSlots;        // linked list of free slots
+    SlotJustify   * m_freeJustifies;    // Slot justification blocks free list
+    CharInfo      * m_charinfo;         // character info, one per input character
+    SlotCollision * m_collisions;
+    const Face    * m_face;             // GrFace
+    const Silf    * m_silf;
+    Slot          * m_first;            // first slot in segment
+    Slot          * m_last;             // last slot in segment
+    size_t          m_bufSize,          // how big a buffer to create when need more slots
+                    m_numGlyphs,
+                    m_numCharinfo;      // size of the array and number of input characters
+    int             m_defaultOriginal;  // number of whitespace chars in the string
+    int8            m_dir;
+    uint8           m_flags,            // General purpose flags
+                    m_passBits;         // if bit set then skip pass
+};
+
+inline
+int8 Segment::getSlotBidiClass(Slot *s) const
+{
+    int8 res = s->getBidiClass();
+    if (res != -1) return res;
+    res = int8(glyphAttr(s->gid(), m_silf->aBidi()));
+    s->setBidiClass(res);
+    return res;
+}
+
+inline
+void Segment::finalise(const Font *font, bool reverse)
+{
+    if (!m_first || !m_last) return;
+
+    m_advance = positionSlots(font, m_first, m_last, m_silf->dir(), true);
+    //associateChars(0, m_numCharinfo);
+    if (reverse && currdir() != (m_dir & 1))
+        reverseSlots();
+    linkClusters(m_first, m_last);
+}
+
+inline
+int32 Segment::getGlyphMetric(Slot *iSlot, uint8 metric, uint8 attrLevel, bool rtl) const {
+    if (attrLevel > 0)
+    {
+        Slot *is = findRoot(iSlot);
+        return is->clusterMetric(this, metric, attrLevel, rtl);
+    }
+    else
+        return m_face->getGlyphMetric(iSlot->gid(), metric);
+}
+
+inline
+bool Segment::isWhitespace(const int cid) const
+{
+    return ((cid >= 0x0009) * (cid <= 0x000D)
+         + (cid == 0x0020)
+         + (cid == 0x0085)
+         + (cid == 0x00A0)
+         + (cid == 0x1680)
+         + (cid == 0x180E)
+         + (cid >= 0x2000) * (cid <= 0x200A)
+         + (cid == 0x2028)
+         + (cid == 0x2029)
+         + (cid == 0x202F)
+         + (cid == 0x205F)
+         + (cid == 0x3000)) != 0;
+}
+
+} // namespace graphite2
+
+struct gr_segment : public graphite2::Segment {};
diff --git a/thirdparty/graphite/src/inc/Silf.h b/thirdparty/graphite/src/inc/Silf.h
new file mode 100644
index 0000000000..edc0c3a16d
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Silf.h
@@ -0,0 +1,128 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "graphite2/Font.h"
+#include "inc/Main.h"
+#include "inc/Pass.h"
+
+namespace graphite2 {
+
+class Face;
+class Segment;
+class FeatureVal;
+class VMScratch;
+class Error;
+
+class Pseudo
+{
+public:
+    uint32 uid;
+    uint32 gid;
+    CLASS_NEW_DELETE;
+};
+
+class Justinfo
+{
+public:
+    Justinfo(uint8 stretch, uint8 shrink, uint8 step, uint8 weight) :
+        m_astretch(stretch), m_ashrink(shrink), m_astep(step),
+        m_aweight(weight) {};
+    uint8 attrStretch() const { return m_astretch; }
+    uint8 attrShrink() const { return m_ashrink; }
+    uint8 attrStep() const { return m_astep; }
+    uint8 attrWeight() const { return m_aweight; }
+
+private:
+    uint8   m_astretch;
+    uint8   m_ashrink;
+    uint8   m_astep;
+    uint8   m_aweight;
+};
+
+class Silf
+{
+    // Prevent copying
+    Silf(const Silf&);
+    Silf& operator=(const Silf&);
+
+public:
+    Silf() throw();
+    ~Silf() throw();
+
+    bool readGraphite(const byte * const pSilf, size_t lSilf, Face &face, uint32 version);
+    bool runGraphite(Segment *seg, uint8 firstPass=0, uint8 lastPass=0, int dobidi = 0) const;
+    uint16 findClassIndex(uint16 cid, uint16 gid) const;
+    uint16 getClassGlyph(uint16 cid, unsigned int index) const;
+    uint16 findPseudo(uint32 uid) const;
+    uint8 numUser() const { return m_aUser; }
+    uint8 aPseudo() const { return m_aPseudo; }
+    uint8 aBreak() const { return m_aBreak; }
+    uint8 aMirror() const {return m_aMirror; }
+    uint8 aPassBits() const { return m_aPassBits; }
+    uint8 aBidi() const { return m_aBidi; }
+    uint8 aCollision() const { return m_aCollision; }
+    uint8 substitutionPass() const { return m_sPass; }
+    uint8 positionPass() const { return m_pPass; }
+    uint8 justificationPass() const { return m_jPass; }
+    uint8 bidiPass() const { return m_bPass; }
+    uint8 numPasses() const { return m_numPasses; }
+    uint8 maxCompPerLig() const { return m_iMaxComp; }
+    uint16 numClasses() const { return m_nClass; }
+    byte  flags() const { return m_flags; }
+    byte  dir() const { return m_dir; }
+    uint8 numJustLevels() const { return m_numJusts; }
+    Justinfo *justAttrs() const { return m_justs; }
+    uint16 endLineGlyphid() const { return m_gEndLine; }
+    const gr_faceinfo *silfInfo() const { return &m_silfinfo; }
+
+    CLASS_NEW_DELETE;
+
+private:
+    size_t readClassMap(const byte *p, size_t data_len, uint32 version, Error &e);
+    template<typename T> inline uint32 readClassOffsets(const byte *&p, size_t data_len, Error &e);
+
+    Pass          * m_passes;
+    Pseudo        * m_pseudos;
+    uint32        * m_classOffsets;
+    uint16        * m_classData;
+    Justinfo      * m_justs;
+    uint8           m_numPasses;
+    uint8           m_numJusts;
+    uint8           m_sPass, m_pPass, m_jPass, m_bPass,
+                    m_flags, m_dir;
+
+    uint8       m_aPseudo, m_aBreak, m_aUser, m_aBidi, m_aMirror, m_aPassBits,
+                m_iMaxComp, m_aCollision;
+    uint16      m_aLig, m_numPseudo, m_nClass, m_nLinear,
+                m_gEndLine;
+    gr_faceinfo m_silfinfo;
+
+    void releaseBuffers() throw();
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/Slot.h b/thirdparty/graphite/src/inc/Slot.h
new file mode 100644
index 0000000000..df39d9a3bb
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Slot.h
@@ -0,0 +1,170 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include "graphite2/Types.h"
+#include "graphite2/Segment.h"
+#include "inc/Main.h"
+#include "inc/Font.h"
+#include "inc/Position.h"
+
+namespace graphite2 {
+
+typedef gr_attrCode attrCode;
+
+class GlyphFace;
+class Segment;
+
+struct SlotJustify
+{
+    static const int NUMJUSTPARAMS = 5;
+
+    SlotJustify(const SlotJustify &);
+    SlotJustify & operator = (const SlotJustify &);
+
+public:
+    static size_t size_of(size_t levels) { return sizeof(SlotJustify) + ((levels > 1 ? levels : 1)*NUMJUSTPARAMS - 1)*sizeof(int16); }
+
+    void LoadSlot(const Slot *s, const Segment *seg);
+
+    SlotJustify *next;
+    int16 values[1];
+};
+
+class Slot
+{
+    enum Flag
+    {
+        DELETED     = 1,
+        INSERTED    = 2,
+        COPIED      = 4,
+        POSITIONED  = 8,
+        ATTACHED    = 16
+    };
+
+public:
+    struct iterator;
+
+    unsigned short gid() const { return m_glyphid; }
+    Position origin() const { return m_position; }
+    float advance() const { return m_advance.x; }
+    void advance(Position &val) { m_advance = val; }
+    Position advancePos() const { return m_advance; }
+    int before() const { return m_before; }
+    int after() const { return m_after; }
+    uint32 index() const { return m_index; }
+    void index(uint32 val) { m_index = val; }
+
+    Slot(int16 *m_userAttr = NULL);
+    void set(const Slot & slot, int charOffset, size_t numUserAttr, size_t justLevels, size_t numChars);
+    Slot *next() const { return m_next; }
+    void next(Slot *s) { m_next = s; }
+    Slot *prev() const { return m_prev; }
+    void prev(Slot *s) { m_prev = s; }
+    uint16 glyph() const { return m_realglyphid ? m_realglyphid : m_glyphid; }
+    void setGlyph(Segment *seg, uint16 glyphid, const GlyphFace * theGlyph = NULL);
+    void setRealGid(uint16 realGid) { m_realglyphid = realGid; }
+    void adjKern(const Position &pos) { m_shift = m_shift + pos; m_advance = m_advance + pos; }
+    void origin(const Position &pos) { m_position = pos + m_shift; }
+    void originate(int ind) { m_original = ind; }
+    int original() const { return m_original; }
+    void before(int ind) { m_before = ind; }
+    void after(int ind) { m_after = ind; }
+    bool isBase() const { return (!m_parent); }
+    void update(int numSlots, int numCharInfo, Position &relpos);
+    Position finalise(const Segment* seg, const Font* font, Position & base, Rect & bbox, uint8 attrLevel, float & clusterMin, bool rtl, bool isFinal, int depth = 0);
+    bool isDeleted() const { return (m_flags & DELETED) ? true : false; }
+    void markDeleted(bool state) { if (state) m_flags |= DELETED; else m_flags &= ~DELETED; }
+    bool isCopied() const { return (m_flags & COPIED) ? true : false; }
+    void markCopied(bool state) { if (state) m_flags |= COPIED; else m_flags &= ~COPIED; }
+    bool isPositioned() const { return (m_flags & POSITIONED) ? true : false; }
+    void markPositioned(bool state) { if (state) m_flags |= POSITIONED; else m_flags &= ~POSITIONED; }
+    bool isInsertBefore() const { return !(m_flags & INSERTED); }
+    uint8 getBidiLevel() const { return m_bidiLevel; }
+    void setBidiLevel(uint8 level) { m_bidiLevel = level; }
+    int8 getBidiClass(const Segment *seg);
+    int8 getBidiClass() const { return m_bidiCls; }
+    void setBidiClass(int8 cls) { m_bidiCls = cls; }
+    int16 *userAttrs() const { return m_userAttr; }
+    void userAttrs(int16 *p) { m_userAttr = p; }
+    void markInsertBefore(bool state) { if (!state) m_flags |= INSERTED; else m_flags &= ~INSERTED; }
+    void setAttr(Segment* seg, attrCode ind, uint8 subindex, int16 val, const SlotMap & map);
+    int getAttr(const Segment *seg, attrCode ind, uint8 subindex) const;
+    int getJustify(const Segment *seg, uint8 level, uint8 subindex) const;
+    void setJustify(Segment *seg, uint8 level, uint8 subindex, int16 value);
+    bool isLocalJustify() const { return m_justs != NULL; };
+    void attachTo(Slot *ap) { m_parent = ap; }
+    Slot *attachedTo() const { return m_parent; }
+    Position attachOffset() const { return m_attach - m_with; }
+    Slot* firstChild() const { return m_child; }
+    void firstChild(Slot *ap) { m_child = ap; }
+    bool child(Slot *ap);
+    Slot* nextSibling() const { return m_sibling; }
+    void nextSibling(Slot *ap) { m_sibling = ap; }
+    bool sibling(Slot *ap);
+    bool removeChild(Slot *ap);
+    int32 clusterMetric(const Segment* seg, uint8 metric, uint8 attrLevel, bool rtl);
+    void positionShift(Position a) { m_position += a; }
+    void floodShift(Position adj, int depth = 0);
+    float just() const { return m_just; }
+    void just(float j) { m_just = j; }
+    Slot *nextInCluster(const Slot *s) const;
+    bool isChildOf(const Slot *base) const;
+
+    CLASS_NEW_DELETE
+
+private:
+    Slot *m_next;           // linked list of slots
+    Slot *m_prev;
+    unsigned short m_glyphid;        // glyph id
+    uint16 m_realglyphid;
+    uint32 m_original;      // charinfo that originated this slot (e.g. for feature values)
+    uint32 m_before;        // charinfo index of before association
+    uint32 m_after;         // charinfo index of after association
+    uint32 m_index;         // slot index given to this slot during finalising
+    Slot *m_parent;         // index to parent we are attached to
+    Slot *m_child;          // index to first child slot that attaches to us
+    Slot *m_sibling;        // index to next child that attaches to our parent
+    Position m_position;    // absolute position of glyph
+    Position m_shift;       // .shift slot attribute
+    Position m_advance;     // .advance slot attribute
+    Position m_attach;      // attachment point on us
+    Position m_with;        // attachment point position on parent
+    float    m_just;        // Justification inserted space
+    uint8    m_flags;       // holds bit flags
+    byte     m_attLevel;    // attachment level
+    int8     m_bidiCls;     // bidirectional class
+    byte     m_bidiLevel;   // bidirectional level
+    int16   *m_userAttr;    // pointer to user attributes
+    SlotJustify *m_justs;   // pointer to justification parameters
+
+    friend class Segment;
+};
+
+} // namespace graphite2
+
+struct gr_slot : public graphite2::Slot {};
diff --git a/thirdparty/graphite/src/inc/Sparse.h b/thirdparty/graphite/src/inc/Sparse.h
new file mode 100644
index 0000000000..fcda890171
--- /dev/null
+++ b/thirdparty/graphite/src/inc/Sparse.h
@@ -0,0 +1,168 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include <iterator>
+#include <utility>
+
+#include "inc/Main.h"
+
+namespace graphite2 {
+
+
+// A read-only packed fast sparse array of uint16 with uint16 keys.
+// Like most container classes this has capacity and size properties and these
+// refer to the number of stored entries and the number of addressable entries
+// as normal. However due the sparse nature the capacity is always <= than the
+// size.
+class sparse
+{
+public:
+    typedef uint16  key_type;
+    typedef uint16  mapped_type;
+    typedef std::pair<const key_type, mapped_type> value_type;
+
+private:
+    typedef unsigned long   mask_t;
+
+    static const unsigned char  SIZEOF_CHUNK = (sizeof(mask_t) - sizeof(key_type))*8;
+
+    struct chunk
+    {
+        mask_t          mask:SIZEOF_CHUNK;
+        key_type        offset;
+    };
+
+    static const chunk  empty_chunk;
+    sparse(const sparse &);
+    sparse & operator = (const sparse &);
+
+public:
+    template<typename I>
+    sparse(I first, const I last);
+    sparse() throw();
+    ~sparse() throw();
+
+    operator bool () const throw();
+    mapped_type     operator [] (const key_type k) const throw();
+
+    size_t capacity() const throw();
+    size_t size()     const throw();
+
+    size_t _sizeof() const throw();
+
+    CLASS_NEW_DELETE;
+
+private:
+    union {
+        chunk         * map;
+        mapped_type   * values;
+    }           m_array;
+    key_type    m_nchunks;
+};
+
+
+inline
+sparse::sparse() throw() : m_nchunks(0)
+{
+    m_array.map = const_cast<graphite2::sparse::chunk *>(&empty_chunk);
+}
+
+
+template <typename I>
+sparse::sparse(I attr, const I last)
+: m_nchunks(0)
+{
+    m_array.map = 0;
+
+    // Find the maximum extent of the key space.
+    size_t n_values=0;
+    long lastkey = -1;
+    for (I i = attr; i != last; ++i, ++n_values)
+    {
+        const typename std::iterator_traits<I>::value_type v = *i;
+        if (v.second == 0)      { --n_values; continue; }
+        if (v.first <= lastkey) { m_nchunks = 0; return; }
+
+        lastkey = v.first;
+        const key_type k = v.first / SIZEOF_CHUNK;
+        if (k >= m_nchunks) m_nchunks = k+1;
+    }
+    if (m_nchunks == 0)
+    {
+        m_array.map=const_cast<graphite2::sparse::chunk *>(&empty_chunk);
+        return;
+    }
+
+    m_array.values = grzeroalloc<mapped_type>((m_nchunks*sizeof(chunk) + sizeof(mapped_type)-1)
+                                                 / sizeof(mapped_type)
+                                                 + n_values);
+
+    if (m_array.values == 0)
+        return;
+
+    // coverity[forward_null : FALSE] Since m_array is union and m_array.values is not NULL
+    chunk * ci = m_array.map;
+    ci->offset = (m_nchunks*sizeof(chunk) + sizeof(mapped_type)-1)/sizeof(mapped_type);
+    mapped_type * vi = m_array.values + ci->offset;
+    for (; attr != last; ++attr, ++vi)
+    {
+        const typename std::iterator_traits<I>::value_type v = *attr;
+        if (v.second == 0)  { --vi; continue; }
+
+        chunk * const ci_ = m_array.map + v.first/SIZEOF_CHUNK;
+
+        if (ci != ci_)
+        {
+            ci = ci_;
+            ci->offset = key_type(vi - m_array.values);
+        }
+
+        ci->mask |= 1UL << (SIZEOF_CHUNK - 1 - (v.first % SIZEOF_CHUNK));
+        *vi = v.second;
+    }
+}
+
+
+inline
+sparse::operator bool () const throw()
+{
+    return m_array.map != 0;
+}
+
+inline
+size_t sparse::size() const throw()
+{
+    return m_nchunks*SIZEOF_CHUNK;
+}
+
+inline
+size_t sparse::_sizeof() const throw()
+{
+    return sizeof(sparse) + capacity()*sizeof(mapped_type) + m_nchunks*sizeof(chunk);
+}
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/TtfTypes.h b/thirdparty/graphite/src/inc/TtfTypes.h
new file mode 100644
index 0000000000..ae67915304
--- /dev/null
+++ b/thirdparty/graphite/src/inc/TtfTypes.h
@@ -0,0 +1,419 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+/*--------------------------------------------------------------------*//*:Ignore this sentence.
+
+File: TtfTypes.h
+Responsibility: Tim Eves
+Last reviewed: Not yet.
+
+Description:
+Provides types required to represent the TTF basic types.
+-------------------------------------------------------------------------------*//*:End Ignore*/
+
+
+//**********************************************************************************************
+//  Include files
+//**********************************************************************************************
+namespace graphite2
+{
+namespace TtfUtil
+{
+//**********************************************************************************************
+//  Forward declarations
+//**********************************************************************************************
+
+
+//**********************************************************************************************
+//  Type declarations
+//**********************************************************************************************
+typedef unsigned char   uint8;
+typedef uint8           byte;
+typedef signed char     int8;
+typedef unsigned short  uint16;
+typedef short           int16;
+typedef unsigned int    uint32;
+typedef int             int32;
+
+typedef int16   short_frac;
+typedef int32   fixed;
+typedef int16   fword;
+typedef uint16  ufword;
+typedef int16   f2dot14;
+typedef uint32  long_date_time[2];
+
+//**********************************************************************************************
+//  Constants and enum types
+//**********************************************************************************************/
+enum
+{
+    OneFix = 1<<16
+};
+
+//**********************************************************************************************
+//  Table declarations
+//**********************************************************************************************
+namespace Sfnt
+{
+#pragma pack(push,1) // We need this or the structure members aren't aligned
+                        // correctly.  Fortunately this form of pragma is supposed
+                        // to be recognised by VS C++ too (at least according to
+                        // MSDN).
+
+    struct OffsetSubTable
+    {
+        uint32  scaler_type;
+        uint16  num_tables,
+            search_range,
+            entry_selector,
+            range_shift;
+        struct Entry
+        {
+            uint32  tag,
+                checksum,
+                offset,
+                length;
+        } table_directory[1];
+
+        enum ScalerType
+        {
+            TrueTypeMac = 0x74727565U,
+            TrueTypeWin = 0x00010000U,
+            Type1   = 0x74797031U
+        };
+    };
+
+
+
+
+    struct CharacterCodeMap
+    {
+        uint16  version,
+            num_subtables;
+        struct
+        {
+            uint16  platform_id,
+                platform_specific_id;
+            uint32  offset;
+        } encoding[1];
+    };
+
+    struct CmapSubTable
+    {
+        uint16  format,
+            length,
+            language;
+    };
+
+    struct CmapSubTableFormat4 : CmapSubTable
+    {
+        uint16  seg_count_x2,
+            search_range,
+            entry_selector,
+            range_shift,
+            end_code[1];
+        // There are arrarys after this which need their
+        // start positions calculated since end_code is
+        // seg_count uint16s long.
+    };
+
+    struct CmapSubTableFormat12
+    {
+        fixed   format;
+        uint32  length,
+            language,
+            num_groups;
+        struct
+        {
+            uint32  start_char_code,
+                end_char_code,
+                start_glyph_id;
+        } group[1];
+    };
+
+
+
+    struct FontHeader
+    {
+        fixed   version,
+            font_revision;
+        uint32  check_sum_adjustment,
+            magic_number;
+        uint16  flags,
+            units_per_em;
+        long_date_time  created,
+                modified;
+        fword   x_min,
+            y_min,
+            x_max,
+            y_max;
+        uint16  mac_style,
+            lowest_rec_ppem;
+        int16   font_direction_hint,
+            index_to_loc_format,
+            glyph_data_format;
+        enum
+        {
+            MagicNumber = 0x5F0F3CF5,
+            GlypDataFormat = 0
+        };
+        enum {ShortIndexLocFormat, LongIndexLocFormat};
+    };
+
+
+
+
+    struct PostScriptGlyphName
+    {
+        fixed   format,
+            italic_angle;
+        fword   underline_position,
+            underline_thickness;
+        uint32  is_fixed_pitch,
+            min_mem_type42,
+            max_mem_type42,
+            min_mem_type1,
+            max_mem_type1;
+        enum
+        {
+            Format1  = 0x10000,
+            Format2  = 0x20000,
+            Format25 = 0x28000,
+            Format3  = 0x30000,
+            Format4  = 0x40000
+        };
+    };
+
+    struct PostScriptGlyphName2 : PostScriptGlyphName
+    {
+        uint16  number_of_glyphs,
+            glyph_name_index[1];
+    };
+
+    struct PostScriptGlyphName25 : PostScriptGlyphName
+    {
+        uint16  number_of_glyphs;
+        int8    offset[1];
+    };
+
+    struct PostScriptGlyphName3 : PostScriptGlyphName {};
+
+    struct PostScriptGlyphName4 : PostScriptGlyphName
+    {
+        uint16 glyph_to_char_map[1];
+    };
+
+
+    struct HorizontalHeader
+    {
+        fixed   version;
+        fword   ascent,
+            descent,
+            line_gap;
+        ufword  advance_width_max;
+        fword   min_left_side_bearing,
+            max_left_side_bearing,
+            x_max_element;
+        int16   caret_slope_rise,
+            caret_slope_run;
+        fword   caret_offset;
+        int16   reserved[4],
+            metric_data_format;
+        uint16  num_long_hor_metrics;
+    };
+
+    struct MaximumProfile
+    {
+        fixed   version;
+        uint16  num_glyphs,
+            max_points,
+            max_contours,
+            max_component_points,
+            max_component_contours,
+            max_zones,
+            max_twilight_points,
+            max_storage,
+            max_function_defs,
+            max_instruction_defs,
+            max_stack_elements,
+            max_size_of_instructions,
+            max_component_elements,
+            max_component_depth;
+    };
+
+
+    typedef byte    Panose[10];
+
+    struct Compatibility0
+    {
+        uint16  version;
+        int16   x_avg_char_width;
+        uint16  weight_class,
+            width_class;
+        int16   fs_type,
+            y_subscript_x_size,
+            y_subscript_y_size,
+            y_subscript_x_offset,
+            y_subscript_y_offset,
+            y_superscript_x_size,
+            y_superscript_y_size,
+            y_superscript_x_offset,
+            y_superscript_y_offset,
+            y_strikeout_size,
+            y_strikeout_position,
+            family_class;
+        Panose  panose;
+        uint32  unicode_range[4];
+        int8    ach_vend_id[4];
+        uint16  fs_selection,
+            fs_first_char_index,
+            fs_last_char_index, // Acording to Apple's spec this is where v0 should end
+            typo_ascender,
+            typo_descender,
+            type_linegap,
+            win_ascent,
+            win_descent;
+
+        enum
+        {
+            Italic    =0x01,
+            Underscore=0x02,
+            Negative  =0x04,
+            Outlined  =0x08,
+            StrikeOut =0x10,
+            Bold      =0x20
+        };
+    };
+
+    struct Compatibility1 : Compatibility0
+    {
+        uint32  codepage_range[2];
+    };
+
+    struct Compatibility2 : Compatibility1
+    {
+            int16   x_height,
+                cap_height;
+            uint16  default_char,
+                break_char,
+                max_context;
+    };
+
+    struct Compatibility3 : Compatibility2 {};
+
+    typedef Compatibility3  Compatibility;
+
+
+    struct NameRecord
+    {
+        uint16  platform_id,
+            platform_specific_id,
+            language_id,
+            name_id,
+            length,
+            offset;
+        enum    {Unicode, Mactintosh, Reserved, Microsoft};
+        enum
+        {
+            Copyright, Family, Subfamily, UniqueSubfamily,
+            Fullname, Version, PostScript
+        };
+    };
+
+    struct LangTagRecord
+    {
+        uint16 length,
+            offset;
+    };
+
+    struct FontNames
+    {
+        uint16  format,
+            count,
+            string_offset;
+        NameRecord name_record[1];
+    };
+
+
+    struct HorizontalMetric
+    {
+        uint16  advance_width;
+        int16   left_side_bearing;
+    };
+
+
+    struct Glyph
+    {
+        int16   number_of_contours;
+        fword   x_min,
+            y_min,
+            x_max,
+            y_max;
+    };
+
+    struct SimpleGlyph : Glyph
+    {
+        uint16  end_pts_of_contours[1];
+        enum
+        {
+            OnCurve = 0x01,
+            XShort  = 0x02,
+            YShort  = 0x04,
+            Repeat  = 0x08,
+            XIsSame = 0x10,
+            XIsPos  = 0x10,
+            YIsSame = 0x20,
+            YIsPos  = 0x20
+        };
+    };
+
+    struct CompoundGlyph : Glyph
+    {
+        uint16  flags,
+            glyph_index;
+        enum
+        {
+            Arg1Arg2Words   = 0x01,
+            ArgsAreXYValues = 0x02,
+            RoundXYToGrid   = 0x04,
+            HaveScale       = 0x08,
+            MoreComponents  = 0x20,
+            HaveXAndYScale  = 0x40,
+            HaveTwoByTwo    = 0x80,
+            HaveInstructions = 0x100,
+            UseMyMetrics    = 0x200,
+            OverlapCompund  = 0x400,
+            ScaledOffset    = 0x800,
+            UnscaledOffset  = 0x1000
+        };
+    };
+
+#pragma pack(pop)
+} // end of namespace Sfnt
+
+} // end of namespace TtfUtil
+} // end of namespace graphite2
diff --git a/thirdparty/graphite/src/inc/TtfUtil.h b/thirdparty/graphite/src/inc/TtfUtil.h
new file mode 100644
index 0000000000..3952bc06fb
--- /dev/null
+++ b/thirdparty/graphite/src/inc/TtfUtil.h
@@ -0,0 +1,208 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+/*--------------------------------------------------------------------*//*:Ignore this sentence.
+
+File: TtfUtil.h
+Responsibility: Alan Ward
+Last reviewed: Not yet.
+
+Description:
+    Utility class for handling TrueType font files.
+----------------------------------------------------------------------------------------------*/
+
+
+#include <cstddef>
+
+namespace graphite2
+{
+namespace TtfUtil
+{
+
+#define OVERFLOW_OFFSET_CHECK(p, o) (o + reinterpret_cast<size_t>(p) < reinterpret_cast<size_t>(p))
+
+typedef long fontTableId32;
+typedef unsigned short gid16;
+
+#define TTF_TAG(a,b,c,d) ((a << 24UL) + (b << 16UL) + (c << 8UL) + (d))
+
+// Enumeration used to specify a table in a TTF file
+class Tag
+{
+    unsigned int _v;
+public:
+    Tag(const char n[5]) throw()            : _v(TTF_TAG(n[0],n[1],n[2],n[3])) {}
+    Tag(const unsigned int tag) throw()    : _v(tag) {}
+
+    operator unsigned int () const throw () { return _v; }
+
+    enum
+    {
+        Feat = TTF_TAG('F','e','a','t'),
+        Glat = TTF_TAG('G','l','a','t'),
+        Gloc = TTF_TAG('G','l','o','c'),
+        Sile = TTF_TAG('S','i','l','e'),
+        Silf = TTF_TAG('S','i','l','f'),
+        Sill = TTF_TAG('S','i','l','l'),
+        cmap = TTF_TAG('c','m','a','p'),
+        cvt  = TTF_TAG('c','v','t',' '),
+        cryp = TTF_TAG('c','r','y','p'),
+        head = TTF_TAG('h','e','a','d'),
+        fpgm = TTF_TAG('f','p','g','m'),
+        gdir = TTF_TAG('g','d','i','r'),
+        glyf = TTF_TAG('g','l','y','f'),
+        hdmx = TTF_TAG('h','d','m','x'),
+        hhea = TTF_TAG('h','h','e','a'),
+        hmtx = TTF_TAG('h','m','t','x'),
+        loca = TTF_TAG('l','o','c','a'),
+        kern = TTF_TAG('k','e','r','n'),
+        LTSH = TTF_TAG('L','T','S','H'),
+        maxp = TTF_TAG('m','a','x','p'),
+        name = TTF_TAG('n','a','m','e'),
+        OS_2 = TTF_TAG('O','S','/','2'),
+        post = TTF_TAG('p','o','s','t'),
+        prep = TTF_TAG('p','r','e','p')
+    };
+};
+
+/*----------------------------------------------------------------------------------------------
+    Class providing utility methods to parse a TrueType font file (TTF).
+    Callling application handles all file input and memory allocation.
+    Assumes minimal knowledge of TTF file format.
+----------------------------------------------------------------------------------------------*/
+    ////////////////////////////////// tools to find & check TTF tables
+    bool GetHeaderInfo(size_t & lOffset, size_t & lSize);
+    bool CheckHeader(const void * pHdr);
+    bool GetTableDirInfo(const void * pHdr, size_t & lOffset, size_t & lSize);
+    bool GetTableInfo(const Tag TableTag, const void * pHdr, const void * pTableDir,
+        size_t & lOffset, size_t & lSize);
+    bool CheckTable(const Tag TableId, const void * pTable, size_t lTableSize);
+
+    ////////////////////////////////// simple font wide info
+    size_t  GlyphCount(const void * pMaxp);
+#ifdef ALL_TTFUTILS
+    size_t  MaxCompositeComponentCount(const void * pMaxp);
+    size_t  MaxCompositeLevelCount(const void * pMaxp);
+    size_t  LocaGlyphCount(size_t lLocaSize, const void * pHead); // throw (std::domain_error);
+#endif
+    int DesignUnits(const void * pHead);
+#ifdef ALL_TTFUTILS
+    int HeadTableCheckSum(const void * pHead);
+    void HeadTableCreateTime(const void * pHead, unsigned int * pnDateBC, unsigned int * pnDateAD);
+    void HeadTableModifyTime(const void * pHead, unsigned int * pnDateBC, unsigned int * pnDateAD);
+    bool IsItalic(const void * pHead);
+    int FontAscent(const void * pOs2);
+    int FontDescent(const void * pOs2);
+    bool FontOs2Style(const void *pOs2, bool & fBold, bool & fItalic);
+    bool Get31EngFamilyInfo(const void * pName, size_t & lOffset, size_t & lSize);
+    bool Get31EngFullFontInfo(const void * pName, size_t & lOffset, size_t & lSize);
+    bool Get30EngFamilyInfo(const void * pName, size_t & lOffset, size_t & lSize);
+    bool Get30EngFullFontInfo(const void * pName, size_t & lOffset, size_t & lSize);
+    int PostLookup(const void * pPost, size_t lPostSize, const void * pMaxp,
+        const char * pPostName);
+#endif
+
+    ////////////////////////////////// utility methods helpful for name table
+    bool GetNameInfo(const void * pName, int nPlatformId, int nEncodingId,
+        int nLangId, int nNameId, size_t & lOffset, size_t & lSize);
+    //size_t NameTableLength(const byte * pTable);
+#ifdef ALL_TTFUTILS
+    int GetLangsForNames(const void * pName, int nPlatformId, int nEncodingId,
+        int *nameIdList, int cNameIds, short *langIdList);
+    void SwapWString(void * pWStr, size_t nSize = 0); // throw (std::invalid_argument);
+#endif
+
+    ////////////////////////////////// cmap lookup tools
+    const void * FindCmapSubtable(const void * pCmap, int nPlatformId = 3,
+        int nEncodingId = 1, size_t length = 0);
+    bool CheckCmapSubtable4(const void * pCmap31, const void * pCmapEnd /*, unsigned int maxgid*/);
+    gid16 CmapSubtable4Lookup(const void * pCmapSubtabel4, unsigned int nUnicodeId, int rangeKey = 0);
+    unsigned int CmapSubtable4NextCodepoint(const void *pCmap31, unsigned int nUnicodeId,
+        int * pRangeKey = 0);
+    bool CheckCmapSubtable12(const void *pCmap310, const void * pCmapEnd /*, unsigned int maxgid*/);
+    gid16 CmapSubtable12Lookup(const void * pCmap310, unsigned int uUnicodeId, int rangeKey = 0);
+    unsigned int CmapSubtable12NextCodepoint(const void *pCmap310, unsigned int nUnicodeId,
+        int * pRangeKey = 0);
+
+    ///////////////////////////////// horizontal metric data for a glyph
+    bool HorMetrics(gid16 nGlyphId, const void * pHmtx, size_t lHmtxSize,
+        const void * pHhea, int & nLsb, unsigned int & nAdvWid);
+
+    ////////////////////////////////// primitives for loca and glyf lookup
+    size_t LocaLookup(gid16 nGlyphId, const void * pLoca, size_t lLocaSize,
+        const void * pHead); // throw (std::out_of_range);
+    void * GlyfLookup(const void * pGlyf, size_t lGlyfOffset, size_t lTableLen);
+
+    ////////////////////////////////// primitves for simple glyph data
+    bool GlyfBox(const void * pSimpleGlyf, int & xMin, int & yMin,
+        int & xMax, int & yMax);
+
+#ifdef ALL_TTFUTILS
+    int GlyfContourCount(const void * pSimpleGlyf);
+    bool GlyfContourEndPoints(const void * pSimpleGlyf, int * prgnContourEndPoint,
+        int cnPointsTotal, size_t & cnPoints);
+    bool GlyfPoints(const void * pSimpleGlyf, int * prgnX, int * prgnY,
+        char * prgbFlag, int cnPointsTotal, int & cnPoints);
+
+    // primitive to find the glyph ids in a composite glyph
+    bool GetComponentGlyphIds(const void * pSimpleGlyf, int * prgnCompId,
+        size_t cnCompIdTotal, size_t & cnCompId);
+    // primitive to find the placement data for a component in a composite glyph
+    bool GetComponentPlacement(const void * pSimpleGlyf, int nCompId,
+        bool fOffset, int & a, int & b);
+    // primitive to find the transform data for a component in a composite glyph
+    bool GetComponentTransform(const void * pSimpleGlyf, int nCompId,
+        float & flt11, float & flt12, float & flt21, float & flt22, bool & fTransOffset);
+#endif
+
+    ////////////////////////////////// operate on composite or simple glyph (auto glyf lookup)
+    void * GlyfLookup(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void * pHead); // primitive used by below methods
+
+#ifdef ALL_TTFUTILS
+    // below are primary user methods for handling glyf data
+    bool IsSpace(gid16 nGlyphId, const void * pLoca, size_t lLocaSize, const void * pHead);
+    bool IsDeepComposite(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void * pHead);
+
+    bool GlyfBox(gid16 nGlyphId, const void * pGlyf, const void * pLoca, size_t lGlyfSize, size_t lLocaSize,
+        const void * pHead, int & xMin, int & yMin, int & xMax, int & yMax);
+    bool GlyfContourCount(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void *pHead, size_t & cnContours);
+    bool GlyfContourEndPoints(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void * pHead, int * prgnContourEndPoint, size_t cnPoints);
+    bool GlyfPoints(gid16 nGlyphId, const void * pGlyf, const void * pLoca,
+        size_t lGlyfSize, size_t lLocaSize, const void * pHead, const int * prgnContourEndPoint, size_t cnEndPoints,
+        int * prgnX, int * prgnY, bool * prgfOnCurve, size_t cnPoints);
+
+    // utitily method used by high-level GlyfPoints
+    bool SimplifyFlags(char * prgbFlags, int cnPoints);
+    bool CalcAbsolutePoints(int * prgnX, int * prgnY, int cnPoints);
+#endif
+
+} // end of namespace TtfUtil
+} // end of namespace graphite2
diff --git a/thirdparty/graphite/src/inc/UtfCodec.h b/thirdparty/graphite/src/inc/UtfCodec.h
new file mode 100644
index 0000000000..24a343d8d9
--- /dev/null
+++ b/thirdparty/graphite/src/inc/UtfCodec.h
@@ -0,0 +1,251 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+#include <cstdlib>
+#include "inc/Main.h"
+
+namespace graphite2 {
+
+typedef uint32  uchar_t;
+
+template <int N>
+struct _utf_codec
+{
+    typedef uchar_t codeunit_t;
+
+    static void     put(codeunit_t * cp, const uchar_t , int8 & len) throw();
+    static uchar_t  get(const codeunit_t * cp, int8 & len) throw();
+    static bool     validate(const codeunit_t * s, const codeunit_t * const e) throw();
+};
+
+
+template <>
+struct _utf_codec<32>
+{
+private:
+    static const uchar_t    limit = 0x110000;
+public:
+    typedef uint32  codeunit_t;
+
+    inline
+    static void put(codeunit_t * cp, const uchar_t usv, int8 & l) throw()
+    {
+        *cp = usv; l = 1;
+    }
+
+    inline
+    static uchar_t get(const codeunit_t * cp, int8 & l) throw()
+    {
+        if (cp[0] < limit)  { l = 1;  return cp[0]; }
+        else                { l = -1; return 0xFFFD; }
+    }
+
+    inline
+    static bool validate(const codeunit_t * s, const codeunit_t * const e) throw()
+    {
+        return s <= e;
+    }
+};
+
+
+template <>
+struct _utf_codec<16>
+{
+private:
+    static const int32  lead_offset      = 0xD800 - (0x10000 >> 10);
+    static const int32  surrogate_offset = 0x10000 - (0xD800 << 10) - 0xDC00;
+public:
+    typedef uint16  codeunit_t;
+
+    inline
+    static void put(codeunit_t * cp, const uchar_t usv, int8 & l) throw()
+    {
+        if (usv < 0x10000)  { l = 1; cp[0] = codeunit_t(usv); }
+        else
+        {
+            cp[0] = codeunit_t(lead_offset + (usv >> 10));
+            cp[1] = codeunit_t(0xDC00 + (usv & 0x3FF));
+            l = 2;
+        }
+    }
+
+    inline
+    static uchar_t get(const codeunit_t * cp, int8 & l) throw()
+    {
+        const uint32    uh = cp[0];
+        l = 1;
+
+        if (uh < 0xD800|| uh > 0xDFFF) { return uh; }
+        if (uh > 0xDBFF) { l = -1; return 0xFFFD; }
+        const uint32 ul = cp[1];
+        if (ul < 0xDC00 || ul > 0xDFFF) { l = -1; return 0xFFFD; }
+        ++l;
+        return (uh<<10) + ul + surrogate_offset;
+    }
+
+    inline
+    static bool validate(const codeunit_t * s, const codeunit_t * const e) throw()
+    {
+        const ptrdiff_t n = e-s;
+        if (n <= 0) return n == 0;
+        const uint32 u = *(e-1); // Get the last codepoint
+        return (u < 0xD800 || u > 0xDBFF);
+    }
+};
+
+
+template <>
+struct _utf_codec<8>
+{
+private:
+    static const int8 sz_lut[16];
+    static const byte mask_lut[5];
+    static const uchar_t    limit = 0x110000;
+
+public:
+    typedef uint8   codeunit_t;
+
+    inline
+    static void put(codeunit_t * cp, const uchar_t usv, int8 & l) throw()
+    {
+        if (usv < 0x80)     {l = 1; cp[0] = usv; return; }
+        if (usv < 0x0800)   {l = 2; cp[0] = 0xC0 + (usv >> 6);  cp[1] = 0x80 + (usv & 0x3F); return; }
+        if (usv < 0x10000)  {l = 3; cp[0] = 0xE0 + (usv >> 12); cp[1] = 0x80 + ((usv >> 6) & 0x3F);  cp[2] = 0x80 + (usv & 0x3F); return; }
+        else                {l = 4; cp[0] = 0xF0 + (usv >> 18); cp[1] = 0x80 + ((usv >> 12) & 0x3F); cp[2] = 0x80 + ((usv >> 6) & 0x3F); cp[3] = 0x80 + (usv & 0x3F); return; }
+    }
+
+    inline
+    static uchar_t get(const codeunit_t * cp, int8 & l) throw()
+    {
+        const int8 seq_sz = sz_lut[*cp >> 4];
+        uchar_t u = *cp & mask_lut[seq_sz];
+        l = 1;
+        bool toolong = false;
+
+        switch(seq_sz) {
+            case 4:     u <<= 6; u |= *++cp & 0x3F; if (*cp >> 6 != 2) break; ++l; toolong  = (u < 0x10); GR_FALLTHROUGH;
+                // no break
+            case 3:     u <<= 6; u |= *++cp & 0x3F; if (*cp >> 6 != 2) break; ++l; toolong |= (u < 0x20); GR_FALLTHROUGH;
+                // no break
+            case 2:     u <<= 6; u |= *++cp & 0x3F; if (*cp >> 6 != 2) break; ++l; toolong |= (u < 0x80); GR_FALLTHROUGH;
+                // no break
+            case 1:     break;
+            case 0:     l = -1; return 0xFFFD;
+        }
+
+        if (l != seq_sz || toolong  || u >= limit)
+        {
+            l = -l;
+            return 0xFFFD;
+        }
+        return u;
+    }
+
+    inline
+    static bool validate(const codeunit_t * s, const codeunit_t * const e) throw()
+    {
+        const ptrdiff_t n = e-s;
+        if (n <= 0) return n == 0;
+        s += (n-1);
+        if (*s < 0x80) return true;
+        if (*s >= 0xC0) return false;
+        if (n == 1) return true;
+        if (*--s < 0x80) return true;
+        if (*s >= 0xE0) return false;
+        if (n == 2 || *s >= 0xC0) return true;
+        if (*--s < 0x80) return true;
+        if (*s >= 0xF0) return false;
+        return true;
+    }
+
+};
+
+
+template <typename C>
+class _utf_iterator
+{
+    typedef _utf_codec<sizeof(C)*8> codec;
+
+    C             * cp;
+    mutable int8    sl;
+
+public:
+    typedef C           codeunit_type;
+    typedef uchar_t     value_type;
+    typedef uchar_t   * pointer;
+
+    class reference
+    {
+        const _utf_iterator & _i;
+
+        reference(const _utf_iterator & i): _i(i) {}
+    public:
+        operator value_type () const throw ()                   { return codec::get(_i.cp, _i.sl); }
+        reference & operator = (const value_type usv) throw()   { codec::put(_i.cp, usv, _i.sl); return *this; }
+
+        friend class _utf_iterator;
+    };
+
+
+    _utf_iterator(const void * us=0)    : cp(reinterpret_cast<C *>(const_cast<void *>(us))), sl(1) { }
+
+    _utf_iterator   & operator ++ ()    { cp += abs(sl); return *this; }
+    _utf_iterator   operator ++ (int)   { _utf_iterator tmp(*this); operator++(); return tmp; }
+
+    bool operator == (const _utf_iterator & rhs) const throw() { return cp >= rhs.cp; }
+    bool operator != (const _utf_iterator & rhs) const throw() { return !operator==(rhs); }
+
+    reference   operator * () const throw() { return *this; }
+    pointer     operator ->() const throw() { return &operator *(); }
+
+    operator codeunit_type * () const throw() { return cp; }
+
+    bool error() const throw()  { return sl < 1; }
+    bool validate(const _utf_iterator & e)  { return codec::validate(cp, e.cp); }
+};
+
+template <typename C>
+struct utf
+{
+    typedef typename _utf_codec<sizeof(C)*8>::codeunit_t codeunit_t;
+
+    typedef _utf_iterator<C>        iterator;
+    typedef _utf_iterator<const C>  const_iterator;
+
+    inline
+    static bool validate(codeunit_t * s, codeunit_t * e) throw() {
+        return _utf_codec<sizeof(C)*8>::validate(s,e);
+    }
+};
+
+
+typedef utf<uint32> utf32;
+typedef utf<uint16> utf16;
+typedef utf<uint8>  utf8;
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/bits.h b/thirdparty/graphite/src/inc/bits.h
new file mode 100644
index 0000000000..9365986a10
--- /dev/null
+++ b/thirdparty/graphite/src/inc/bits.h
@@ -0,0 +1,150 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2012, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+
+namespace graphite2
+{
+
+
+#if defined GRAPHITE2_BUILTINS && (defined __GNUC__ || defined __clang__)
+
+template<typename T>
+inline unsigned int bit_set_count(T v)
+{
+    return __builtin_popcount(v);
+}
+
+template<>
+inline unsigned int bit_set_count(int16 v)
+{
+    return __builtin_popcount(static_cast<uint16>(v));
+}
+
+template<>
+inline unsigned int bit_set_count(int8 v)
+{
+    return __builtin_popcount(static_cast<uint8>(v));
+}
+
+template<>
+inline unsigned int bit_set_count(unsigned long v)
+{
+    return __builtin_popcountl(v);
+}
+
+template<>
+inline unsigned int bit_set_count(signed long v)
+{
+    return __builtin_popcountl(v);
+}
+
+template<>
+inline unsigned int bit_set_count(unsigned long long v)
+{
+    return __builtin_popcountll(v);
+}
+
+template<>
+inline unsigned int bit_set_count(signed long long v)
+{
+    return __builtin_popcountll(v);
+}
+
+#else
+
+template<typename T>
+inline unsigned int bit_set_count(T v)
+{
+	static size_t const ONES = ~0;
+
+	v = v - ((v >> 1) & T(ONES/3));                      // temp
+    v = (v & T(ONES/15*3)) + ((v >> 2) & T(ONES/15*3));  // temp
+    v = (v + (v >> 4)) & T(ONES/255*15);                 // temp
+    return (T)(v * T(ONES/255)) >> (sizeof(T)-1)*8;      // count
+}
+
+#endif
+
+//TODO: Changed these to uintmax_t when we go to C++11
+template<int S>
+inline size_t _mask_over_val(size_t v)
+{
+    v = _mask_over_val<S/2>(v);
+    v |= v >> S*4;
+    return v;
+}
+
+//TODO: Changed these to uintmax_t when we go to C++11
+template<>
+inline size_t _mask_over_val<1>(size_t v)
+{
+    v |= v >> 1;
+    v |= v >> 2;
+    v |= v >> 4;
+    return v;
+}
+
+template<typename T>
+inline T mask_over_val(T v)
+{
+    return T(_mask_over_val<sizeof(T)>(v));
+}
+
+template<typename T>
+inline unsigned long next_highest_power2(T v)
+{
+    return _mask_over_val<sizeof(T)>(v-1)+1;
+}
+
+template<typename T>
+inline unsigned int log_binary(T v)
+{
+    return bit_set_count(mask_over_val(v))-1;
+}
+
+template<typename T>
+inline T has_zero(const T x)
+{
+    return (x - T(~T(0)/255)) & ~x & T(~T(0)/255*128);
+}
+
+template<typename T>
+inline T zero_bytes(const T x, unsigned char n)
+{
+    const T t = T(~T(0)/255*n);
+    return T((has_zero(x^t) >> 7)*n);
+}
+
+#if 0
+inline float float_round(float x, uint32 m)
+{
+    *reinterpret_cast<unsigned int *>(&x) &= m;
+    return *reinterpret_cast<float *>(&x);
+}
+#endif
+
+}
diff --git a/thirdparty/graphite/src/inc/debug.h b/thirdparty/graphite/src/inc/debug.h
new file mode 100644
index 0000000000..97175eb2cc
--- /dev/null
+++ b/thirdparty/graphite/src/inc/debug.h
@@ -0,0 +1,89 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+//  debug.h
+//
+//  Created on: 22 Dec 2011
+//      Author: tim
+
+#pragma once
+
+#if !defined GRAPHITE2_NTRACING
+
+#include <utility>
+#include "inc/json.h"
+#include "inc/Position.h"
+
+namespace graphite2
+{
+
+class CharInfo;
+class Segment;
+class Slot;
+
+typedef std::pair<const Segment * const, const Slot * const>    dslot;
+struct objectid
+{
+    char name[16];
+    objectid(const dslot &) throw();
+    objectid(const Segment * const p) throw();
+};
+
+
+json & operator << (json & j, const Position &) throw();
+json & operator << (json & j, const Rect &) throw();
+json & operator << (json & j, const CharInfo &) throw();
+json & operator << (json & j, const dslot &) throw();
+json & operator << (json & j, const objectid &) throw();
+json & operator << (json & j, const telemetry &) throw();
+
+
+
+inline
+json & operator << (json & j, const Position & p) throw()
+{
+    return j << json::flat << json::array << p.x << p.y << json::close;
+}
+
+
+inline
+json & operator << (json & j, const Rect & p) throw()
+{
+    return j << json::flat << json::array << p.bl.x << p.bl.y << p.tr.x << p.tr.y << json::close;
+}
+
+
+inline
+json & operator << (json & j, const objectid & sid) throw()
+{
+    return j << sid.name;
+}
+
+
+} // namespace graphite2
+
+#endif //!defined GRAPHITE2_NTRACING
+
diff --git a/thirdparty/graphite/src/inc/json.h b/thirdparty/graphite/src/inc/json.h
new file mode 100644
index 0000000000..554cd9a3d1
--- /dev/null
+++ b/thirdparty/graphite/src/inc/json.h
@@ -0,0 +1,178 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// JSON pretty printer for graphite font debug output logging.
+// Created on: 15 Dec 2011
+//     Author: Tim Eves
+
+#pragma once
+
+#include "inc/Main.h"
+#include <cassert>
+#include <cstdio>
+#include <cstdint>
+#include "inc/List.h"
+
+namespace graphite2 {
+
+class json
+{
+    // Prevent copying
+    json(const json &);
+    json & operator = (const json &);
+
+    typedef void (*_context_t)(json &);
+
+    FILE * const    _stream;
+    char            _contexts[128], // context stack
+                  * _context,       // current context (top of stack)
+                  * _flatten;       // if !0 points to context above which
+                                    //  pretty printed output should occur.
+    Vector<void *>  _env;
+
+    void context(const char current) throw();
+    void indent(const int d=0) throw();
+    void push_context(const char, const char) throw();
+    void pop_context() throw();
+
+public:
+    class closer;
+
+    using string = const char *;
+    using number = double;
+    enum class integer : std::intmax_t {};
+    enum class integer_u : std::uintmax_t {};
+    using boolean = bool;
+    static const std::nullptr_t  null;
+
+    void setenv(unsigned int index, void *val) { _env.reserve(index + 1); if (index >= _env.size()) _env.insert(_env.end(), _env.size() - index + 1, 0); _env[index] = val; }
+    void *getenv(unsigned int index) const { return _env[index]; }
+    const Vector<void *> &getenvs() const { return _env; }
+
+    static void flat(json &) throw();
+    static void close(json &) throw();
+    static void object(json &) throw();
+    static void array(json &) throw();
+    static void item(json &) throw();
+
+    json(FILE * stream) throw();
+    ~json() throw ();
+
+    FILE * stream() const throw();
+
+    json & operator << (string) throw();
+    json & operator << (number) throw();
+    json & operator << (integer) throw();
+    json & operator << (integer_u) throw();
+    json & operator << (boolean) throw();
+    json & operator << (std::nullptr_t) throw();
+    json & operator << (_context_t) throw();
+
+    operator bool() const throw();
+    bool good() const throw();
+    bool eof() const throw();
+
+    CLASS_NEW_DELETE;
+};
+
+class json::closer
+{
+    // Prevent copying.
+    closer(const closer &);
+    closer & operator = (const closer &);
+
+    json * const    _j;
+public:
+    closer(json * const j) : _j(j) {}
+    ~closer() throw() { if (_j)  *_j << close; }
+};
+
+inline
+json::json(FILE * s) throw()
+: _stream(s), _context(_contexts), _flatten(0)
+{
+    if (good())
+        fflush(s);
+}
+
+
+inline
+json::~json() throw ()
+{
+    while (_context > _contexts)    pop_context();
+}
+
+inline
+FILE * json::stream() const throw()     { return _stream; }
+
+
+inline
+json & json::operator << (json::_context_t ctxt) throw()
+{
+    ctxt(*this);
+    return *this;
+}
+
+inline
+json & operator << (json & j, signed char d) throw()   { return j << json::integer(d); }
+
+inline
+json & operator << (json & j, unsigned char d) throw() { return j << json::integer_u(d); }
+
+inline
+json & operator << (json & j, short int d) throw()   { return j << json::integer(d); }
+
+inline
+json & operator << (json & j, unsigned short int d) throw() { return j << json::integer_u(d); }
+
+inline
+json & operator << (json & j, int d) throw()         { return j << json::integer(d); }
+
+inline
+json & operator << (json & j, unsigned int d) throw()       { return j << json::integer_u(d); }
+
+inline
+json & operator << (json & j, long int d) throw()         { return j << json::integer(d); }
+
+inline
+json & operator << (json & j, unsigned long int d) throw()       { return j << json::integer_u(d); }
+
+inline
+json & operator << (json & j, long long int d) throw()         { return j << json::integer(d); }
+
+inline
+json & operator << (json & j, unsigned long long int d) throw()       { return j << json::integer_u(d); }
+
+inline
+json::operator bool() const throw()     { return good(); }
+
+inline
+bool json::good() const throw()         { return _stream && ferror(_stream) == 0; }
+
+inline
+bool json::eof() const throw()          { return feof(_stream) != 0; }
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/locale2lcid.h b/thirdparty/graphite/src/inc/locale2lcid.h
new file mode 100644
index 0000000000..25d5c0a3c8
--- /dev/null
+++ b/thirdparty/graphite/src/inc/locale2lcid.h
@@ -0,0 +1,450 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+#include <cstring>
+#include <cassert>
+
+#include "inc/Main.h"
+
+
+namespace graphite2 {
+
+struct IsoLangEntry
+{
+    unsigned short mnLang;
+    char maLangStr[4];
+    char maCountry[3];
+};
+
+// Windows Language ID, Locale ISO-639 language, country code as used in
+// naming table of OpenType fonts
+const IsoLangEntry LANG_ENTRIES[] = {
+    { 0x0401, "ar","SA" }, // Arabic Saudi Arabia
+    { 0x0402, "bg","BG" }, // Bulgarian Bulgaria
+    { 0x0403, "ca","ES" }, // Catalan Catalan
+    { 0x0404, "zh","TW" }, // Chinese Taiwan
+    { 0x0405, "cs","CZ" }, // Czech Czech Republic
+    { 0x0406, "da","DK" }, // Danish Denmark
+    { 0x0407, "de","DE" }, // German Germany
+    { 0x0408, "el","GR" }, // Greek Greece
+    { 0x0409, "en","US" }, // English United States
+    { 0x040A, "es","ES" }, // Spanish (Traditional Sort) Spain
+    { 0x040B, "fi","FI" }, // Finnish Finland
+    { 0x040C, "fr","FR" }, // French France
+    { 0x040D, "he","IL" }, // Hebrew Israel
+    { 0x040E, "hu","HU" }, // Hungarian Hungary
+    { 0x040F, "is","IS" }, // Icelandic Iceland
+    { 0x0410, "it","IT" }, // Italian Italy
+    { 0x0411, "jp","JP" }, // Japanese Japan
+    { 0x0412, "ko","KR" }, // Korean Korea
+    { 0x0413, "nl","NL" }, // Dutch Netherlands
+    { 0x0414, "no","NO" }, // Norwegian (Bokmal) Norway
+    { 0x0415, "pl","PL" }, // Polish Poland
+    { 0x0416, "pt","BR" }, // Portuguese Brazil
+    { 0x0417, "rm","CH" }, // Romansh Switzerland
+    { 0x0418, "ro","RO" }, // Romanian Romania
+    { 0x0419, "ru","RU" }, // Russian Russia
+    { 0x041A, "hr","HR" }, // Croatian Croatia
+    { 0x041B, "sk","SK" }, // Slovak Slovakia
+    { 0x041C, "sq","AL" }, // Albanian Albania
+    { 0x041D, "sv","SE" }, // Swedish Sweden
+    { 0x041E, "th","TH" }, // Thai Thailand
+    { 0x041F, "tr","TR" }, // Turkish Turkey
+    { 0x0420, "ur","PK" }, // Urdu Islamic Republic of Pakistan
+    { 0x0421, "id","ID" }, // Indonesian Indonesia
+    { 0x0422, "uk","UA" }, // Ukrainian Ukraine
+    { 0x0423, "be","BY" }, // Belarusian Belarus
+    { 0x0424, "sl","SI" }, // Slovenian Slovenia
+    { 0x0425, "et","EE" }, // Estonian Estonia
+    { 0x0426, "lv","LV" }, // Latvian Latvia
+    { 0x0427, "lt","LT" }, // Lithuanian Lithuania
+    { 0x0428, "tg","TJ" }, // Tajik (Cyrillic) Tajikistan
+    { 0x042A, "vi","VN" }, // Vietnamese Vietnam
+    { 0x042B, "hy","AM" }, // Armenian Armenia
+    { 0x042C, "az","AZ" }, // Azeri (Latin) Azerbaijan
+    { 0x042D, "eu","" }, // Basque Basque
+    { 0x042E, "hsb","DE" }, // Upper Sorbian Germany
+    { 0x042F, "mk","MK" }, // Macedonian (FYROM) Former Yugoslav Republic of Macedonia
+    { 0x0432, "tn","ZA" }, // Setswana South Africa
+    { 0x0434, "xh","ZA" }, // isiXhosa South Africa
+    { 0x0435, "zu","ZA" }, // isiZulu South Africa
+    { 0x0436, "af","ZA" }, // Afrikaans South Africa
+    { 0x0437, "ka","GE" }, // Georgian Georgia
+    { 0x0438, "fo","FO" }, // Faroese Faroe Islands
+    { 0x0439, "hi","IN" }, // Hindi India
+    { 0x043A, "mt","MT" }, // Maltese Malta
+    { 0x043B, "se","NO" }, // Sami (Northern) Norway
+    { 0x043E, "ms","MY" }, // Malay Malaysia
+    { 0x043F, "kk","KZ" }, // Kazakh Kazakhstan
+    { 0x0440, "ky","KG" }, // Kyrgyz Kyrgyzstan
+    { 0x0441, "sw","KE" }, // Kiswahili Kenya
+    { 0x0442, "tk","TM" }, // Turkmen Turkmenistan
+    { 0x0443, "uz","UZ" }, // Uzbek (Latin) Uzbekistan
+    { 0x0444, "tt","RU" }, // Tatar Russia
+    { 0x0445, "bn","IN" }, // Bengali India
+    { 0x0446, "pa","IN" }, // Punjabi India
+    { 0x0447, "gu","IN" }, // Gujarati India
+    { 0x0448, "or","IN" }, // Oriya India
+    { 0x0448, "wo","SN" }, // Wolof Senegal
+    { 0x0449, "ta","IN" }, // Tamil India
+    { 0x044A, "te","IN" }, // Telugu India
+    { 0x044B, "kn","IN" }, // Kannada India
+    { 0x044C, "ml","IN" }, // Malayalam India
+    { 0x044D, "as","IN" }, // Assamese India
+    { 0x044E, "mr","IN" }, // Marathi India
+    { 0x044F, "sa","IN" }, // Sanskrit India
+    { 0x0450, "mn","MN" }, // Mongolian (Cyrillic) Mongolia
+    { 0x0451, "bo","CN" }, // Tibetan PRC
+    { 0x0452, "cy","GB" }, // Welsh United Kingdom
+    { 0x0453, "km","KH" }, // Khmer Cambodia
+    { 0x0454, "lo","LA" }, // Lao Lao P.D.R.
+    { 0x0455, "my","MM" }, // Burmese Myanmar - not listed in Microsoft docs anymore
+    { 0x0456, "gl","ES" }, // Galician Galician
+    { 0x0457, "kok","IN" }, // Konkani India
+    { 0x045A, "syr","TR" }, // Syriac Syria
+    { 0x045B, "si","LK" }, // Sinhala Sri Lanka
+    { 0x045D, "iu","CA" }, // Inuktitut Canada
+    { 0x045E, "am","ET" }, // Amharic Ethiopia
+    { 0x0461, "ne","NP" }, // Nepali Nepal
+    { 0x0462, "fy","NL" }, // Frisian Netherlands
+    { 0x0463, "ps","AF" }, // Pashto Afghanistan
+    { 0x0464, "fil","PH" }, // Filipino Philippines
+    { 0x0465, "dv","MV" }, // Divehi Maldives
+    { 0x0468, "ha","NG" }, // Hausa (Latin) Nigeria
+    { 0x046A, "yo","NG" }, // Yoruba Nigeria
+    { 0x046B, "qu","BO" }, // Quechua Bolivia
+    { 0x046C, "st","ZA" }, // Sesotho sa Leboa South Africa
+    { 0x046D, "ba","RU" }, // Bashkir Russia
+    { 0x046E, "lb","LU" }, // Luxembourgish Luxembourg
+    { 0x046F, "kl","GL" }, // Greenlandic Greenland
+    { 0x0470, "ig","NG" }, // Igbo Nigeria
+    { 0x0478, "ii","CN" }, // Yi PRC
+    { 0x047A, "arn","CL" }, // Mapudungun Chile
+    { 0x047C, "moh","CA" }, // Mohawk Mohawk
+    { 0x047E, "br","FR" }, // Breton France
+    { 0x0480, "ug","CN" }, // Uighur PRC
+    { 0x0481, "mi","NZ" }, // Maori New Zealand
+    { 0x0482, "oc","FR" }, // Occitan France
+    { 0x0483, "co","FR" }, // Corsican France
+    { 0x0484, "gsw","FR" }, // Alsatian France
+    { 0x0485, "sah","RU" }, // Yakut Russia
+    { 0x0486, "qut","GT" }, // K'iche Guatemala
+    { 0x0487, "rw","RW" }, // Kinyarwanda Rwanda
+    { 0x048C, "gbz","AF" }, // Dari Afghanistan
+    { 0x0801, "ar","IQ" }, // Arabic Iraq
+    { 0x0804, "zn","CH" }, // Chinese People's Republic of China
+    { 0x0807, "de","CH" }, // German Switzerland
+    { 0x0809, "en","GB" }, // English United Kingdom
+    { 0x080A, "es","MX" }, // Spanish Mexico
+    { 0x080C, "fr","BE" }, // French Belgium
+    { 0x0810, "it","CH" }, // Italian Switzerland
+    { 0x0813, "nl","BE" }, // Dutch Belgium
+    { 0x0814, "nn","NO" }, // Norwegian (Nynorsk) Norway
+    { 0x0816, "pt","PT" }, // Portuguese Portugal
+    { 0x081A, "sh","RS" }, // Serbian (Latin) Serbia
+    { 0x081D, "sv","FI" }, // Sweden Finland
+    { 0x082C, "az","AZ" }, // Azeri (Cyrillic) Azerbaijan
+    { 0x082E, "dsb","DE" }, // Lower Sorbian Germany
+    { 0x083B, "se","SE" }, // Sami (Northern) Sweden
+    { 0x083C, "ga","IE" }, // Irish Ireland
+    { 0x083E, "ms","BN" }, // Malay Brunei Darussalam
+    { 0x0843, "uz","UZ" }, // Uzbek (Cyrillic) Uzbekistan
+    { 0x0845, "bn","BD" }, // Bengali Bangladesh
+    { 0x0850, "mn","MN" }, // Mongolian (Traditional) People's Republic of China
+    { 0x085D, "iu","CA" }, // Inuktitut (Latin) Canada
+    { 0x085F, "ber","DZ" }, // Tamazight (Latin) Algeria
+    { 0x086B, "es","EC" }, // Quechua Ecuador
+    { 0x0C01, "ar","EG" }, // Arabic Egypt
+    { 0x0C04, "zh","HK" }, // Chinese Hong Kong S.A.R.
+    { 0x0C07, "de","AT" }, // German Austria
+    { 0x0C09, "en","AU" }, // English Australia
+    { 0x0C0A, "es","ES" }, // Spanish (Modern Sort) Spain
+    { 0x0C0C, "fr","CA" }, // French Canada
+    { 0x0C1A, "sr","CS" }, // Serbian (Cyrillic) Serbia
+    { 0x0C3B, "se","FI" }, // Sami (Northern) Finland
+    { 0x0C6B, "qu","PE" }, // Quechua Peru
+    { 0x1001, "ar","LY" }, // Arabic Libya
+    { 0x1004, "zh","SG" }, // Chinese Singapore
+    { 0x1007, "de","LU" }, // German Luxembourg
+    { 0x1009, "en","CA" }, // English Canada
+    { 0x100A, "es","GT" }, // Spanish Guatemala
+    { 0x100C, "fr","CH" }, // French Switzerland
+    { 0x101A, "hr","BA" }, // Croatian (Latin) Bosnia and Herzegovina
+    { 0x103B, "smj","NO" }, // Sami (Lule) Norway
+    { 0x1401, "ar","DZ" }, // Arabic Algeria
+    { 0x1404, "zh","MO" }, // Chinese Macao S.A.R.
+    { 0x1407, "de","LI" }, // German Liechtenstein
+    { 0x1409, "en","NZ" }, // English New Zealand
+    { 0x140A, "es","CR" }, // Spanish Costa Rica
+    { 0x140C, "fr","LU" }, // French Luxembourg
+    { 0x141A, "bs","BA" }, // Bosnian (Latin) Bosnia and Herzegovina
+    { 0x143B, "smj","SE" }, // Sami (Lule) Sweden
+    { 0x1801, "ar","MA" }, // Arabic Morocco
+    { 0x1809, "en","IE" }, // English Ireland
+    { 0x180A, "es","PA" }, // Spanish Panama
+    { 0x180C, "fr","MC" }, // French Principality of Monoco
+    { 0x181A, "sh","BA" }, // Serbian (Latin) Bosnia and Herzegovina
+    { 0x183B, "sma","NO" }, // Sami (Southern) Norway
+    { 0x1C01, "ar","TN" }, // Arabic Tunisia
+    { 0x1C09, "en","ZA" }, // English South Africa
+    { 0x1C0A, "es","DO" }, // Spanish Dominican Republic
+    { 0x1C1A, "sr","BA" }, // Serbian (Cyrillic) Bosnia and Herzegovina
+    { 0x1C3B, "sma","SE" }, // Sami (Southern) Sweden
+    { 0x2001, "ar","OM" }, // Arabic Oman
+    { 0x2009, "en","JM" }, // English Jamaica
+    { 0x200A, "es","VE" }, // Spanish Venezuela
+    { 0x201A, "bs","BA" }, // Bosnian (Cyrillic) Bosnia and Herzegovina
+    { 0x203B, "sms","FI" }, // Sami (Skolt) Finland
+    { 0x2401, "ar","YE" }, // Arabic Yemen
+    { 0x2409, "en","BS" }, // English Caribbean
+    { 0x240A, "es","CO" }, // Spanish Colombia
+    { 0x243B, "smn","FI" }, // Sami (Inari) Finland
+    { 0x2801, "ar","SY" }, // Arabic Syria
+    { 0x2809, "en","BZ" }, // English Belize
+    { 0x280A, "es","PE" }, // Spanish Peru
+    { 0x2C01, "ar","JO" }, // Arabic Jordan
+    { 0x2C09, "en","TT" }, // English Trinidad and Tobago
+    { 0x2C0A, "es","AR" }, // Spanish Argentina
+    { 0x3001, "ar","LB" }, // Arabic Lebanon
+    { 0x3009, "en","ZW" }, // English Zimbabwe
+    { 0x300A, "es","EC" }, // Spanish Ecuador
+    { 0x3401, "ar","KW" }, // Arabic Kuwait
+    { 0x3409, "en","PH" }, // English Republic of the Philippines
+    { 0x340A, "es","CL" }, // Spanish Chile
+    { 0x3801, "ar","AE" }, // Arabic U.A.E.
+    { 0x380A, "es","UY" }, // Spanish Uruguay
+    { 0x3C01, "ar","BH" }, // Arabic Bahrain
+    { 0x3C0A, "es","PY" }, // Spanish Paraguay
+    { 0x4001, "ar","QA" }, // Arabic Qatar
+    { 0x4009, "en","IN" }, // English India
+    { 0x400A, "es","BO" }, // Spanish Bolivia
+    { 0x4409, "en","MY" }, // English Malaysia
+    { 0x440A, "es","SV" }, // Spanish El Salvador
+    { 0x4809, "en","SG" }, // English Singapore
+    { 0x480A, "es","HN" }, // Spanish Honduras
+    { 0x4C0A, "es","NI" }, // Spanish Nicaragua
+    { 0x500A, "es","PR" }, // Spanish Puerto Rico
+    { 0x540A, "es","US" } // Spanish United States
+};
+
+class Locale2Lang
+{
+    Locale2Lang(const Locale2Lang &);
+    Locale2Lang & operator = (const Locale2Lang &);
+
+public:
+    Locale2Lang() : mSeedPosition(128)
+    {
+        memset((void*)mLangLookup, 0, sizeof(mLangLookup));
+        // create a tri lookup on first 2 letters of language code
+        static const int maxIndex = sizeof(LANG_ENTRIES)/sizeof(IsoLangEntry);
+        for (int i = 0; i < maxIndex; i++)
+        {
+            size_t a = LANG_ENTRIES[i].maLangStr[0] - 'a';
+            size_t b = LANG_ENTRIES[i].maLangStr[1] - 'a';
+            if (mLangLookup[a][b])
+            {
+                const IsoLangEntry ** old = mLangLookup[a][b];
+                int len = 1;
+                while (old[len]) len++;
+                len += 2;
+                mLangLookup[a][b] = gralloc<const IsoLangEntry *>(len);
+                if (!mLangLookup[a][b])
+                {
+                    mLangLookup[a][b] = old;
+                    continue;
+                }
+                mLangLookup[a][b][--len] = NULL;
+                mLangLookup[a][b][--len] = &LANG_ENTRIES[i];
+                while (--len >= 0)
+                {
+                    assert(len >= 0);
+                    mLangLookup[a][b][len] = old[len];
+                }
+                free(old);
+            }
+            else
+            {
+                mLangLookup[a][b] = gralloc<const IsoLangEntry *>(2);
+                if (!mLangLookup[a][b]) continue;
+                mLangLookup[a][b][1] = NULL;
+                mLangLookup[a][b][0] = &LANG_ENTRIES[i];
+            }
+        }
+        while (2 * mSeedPosition < maxIndex)
+            mSeedPosition *= 2;
+    };
+    ~Locale2Lang()
+    {
+        for (int i = 0; i != 26; ++i)
+            for (int j = 0; j != 26; ++j)
+                free(mLangLookup[i][j]);
+    }
+    unsigned short getMsId(const char * locale) const
+    {
+        size_t length = strlen(locale);
+        size_t langLength = length;
+        const char * language = locale;
+        const char * script = NULL;
+        const char * region = NULL;
+        size_t regionLength = 0;
+        const char * dash = strchr(locale, '-');
+        if (dash && (dash != locale))
+        {
+            langLength = (dash - locale);
+            size_t nextPartLength = length - langLength - 1;
+            if (nextPartLength >= 2)
+            {
+                script = ++dash;
+                dash = strchr(dash, '-');
+                if (dash)
+                {
+                    nextPartLength = (dash - script);
+                    region = ++dash;
+                }
+                if (nextPartLength == 2 &&
+                    (locale[langLength+1] > 0x40) && (locale[langLength+1] < 0x5B) &&
+                    (locale[langLength+2] > 0x40) && (locale[langLength+2] < 0x5B))
+                {
+                    region = script;
+                    regionLength = nextPartLength;
+                    script = NULL;
+                }
+                else if (nextPartLength == 4)
+                {
+                    if (dash)
+                    {
+                        dash = strchr(dash, '-');
+                        if (dash)
+                        {
+                            nextPartLength = (dash - region);
+                        }
+                        else
+                        {
+                            nextPartLength = langLength - (region - locale);
+                        }
+                        regionLength = nextPartLength;
+                    }
+                }
+            }
+        }
+        size_t a = 'e' - 'a';
+        size_t b = 'n' - 'a';
+        unsigned short langId = 0;
+        int i = 0;
+        switch (langLength)
+        {
+            case 2:
+            {
+                a = language[0] - 'a';
+                b = language[1] - 'a';
+                if ((a < 26) && (b < 26) && mLangLookup[a][b])
+                {
+                    while (mLangLookup[a][b][i])
+                    {
+                        if (mLangLookup[a][b][i]->maLangStr[2] != '\0')
+                        {
+                            ++i;
+                            continue;
+                        }
+                        if (region && (strncmp(mLangLookup[a][b][i]->maCountry, region, regionLength) == 0))
+                        {
+                            langId = mLangLookup[a][b][i]->mnLang;
+                            break;
+                        }
+                        else if (langId == 0)
+                        {
+                            // possible fallback code
+                            langId = mLangLookup[a][b][i]->mnLang;
+                        }
+                        ++i;
+                    }
+                }
+            }
+            break;
+            case 3:
+            {
+                a = language[0] - 'a';
+                b = language[1] - 'a';
+                if (mLangLookup[a][b])
+                {
+                    while (mLangLookup[a][b][i])
+                    {
+                        if (mLangLookup[a][b][i]->maLangStr[2] != language[2])
+                        {
+                            ++i;
+                            continue;
+                        }
+                        if (region && (strncmp(mLangLookup[a][b][i]->maCountry, region, regionLength) == 0))
+                        {
+                            langId = mLangLookup[a][b][i]->mnLang;
+                            break;
+                        }
+                        else if (langId == 0)
+                        {
+                            // possible fallback code
+                            langId = mLangLookup[a][b][i]->mnLang;
+                        }
+                        ++i;
+                    }
+                }
+            }
+            break;
+            default:
+                break;
+        }
+        if (langId == 0) langId = 0x409;
+        return langId;
+    }
+    const IsoLangEntry * findEntryById(unsigned short langId) const
+    {
+        static const int maxIndex = sizeof(LANG_ENTRIES)/sizeof(IsoLangEntry);
+        int window = mSeedPosition;
+        int guess = mSeedPosition - 1;
+        while (LANG_ENTRIES[guess].mnLang != langId)
+        {
+            window /= 2;
+            if (window == 0) return NULL;
+            guess += (LANG_ENTRIES[guess].mnLang > langId)? -window : window;
+            while (guess >= maxIndex)
+            {
+                window /= 2;
+                guess -= window;
+                assert(window);
+            }
+        }
+        return &LANG_ENTRIES[guess];
+    }
+
+    CLASS_NEW_DELETE;
+
+private:
+    const IsoLangEntry ** mLangLookup[26][26];
+    int mSeedPosition;
+};
+
+} // namespace graphite2
diff --git a/thirdparty/graphite/src/inc/opcode_table.h b/thirdparty/graphite/src/inc/opcode_table.h
new file mode 100644
index 0000000000..cb5acde9a4
--- /dev/null
+++ b/thirdparty/graphite/src/inc/opcode_table.h
@@ -0,0 +1,124 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// This file will be pulled into and integrated into a machine implmentation
+// DO NOT build directly
+#pragma once
+
+#define do2(n)  do_(n) ,do_(n)
+#define NILOP   0U
+
+// types or parameters are: (.. is inclusive)
+//      number - any byte
+//      output_class - 0 .. silf.m_nClass
+//      input_class - 0 .. silf.m_nClass
+//      sattrnum - 0 .. 29 (gr_slatJWidth) , 55 (gr_slatUserDefn)
+//      attrid - 0 .. silf.numUser() where sattrnum == 55; 0..silf.m_iMaxComp where sattrnum == 15 otherwise 0
+//      gattrnum - 0 .. face->getGlyphFaceCache->numAttrs()
+//      gmetric - 0 .. 11 (kgmetDescent)
+//      featidx - 0 .. face.numFeatures()
+//      level - any byte
+static const opcode_t opcode_table[] =
+{
+    {{do2(nop)},                                    0, "NOP"},
+
+    {{do2(push_byte)},                              1, "PUSH_BYTE"},                // number
+    {{do2(push_byte_u)},                            1, "PUSH_BYTE_U"},              // number
+    {{do2(push_short)},                             2, "PUSH_SHORT"},               // number number
+    {{do2(push_short_u)},                           2, "PUSH_SHORT_U"},             // number number
+    {{do2(push_long)},                              4, "PUSH_LONG"},                // number number number number
+
+    {{do2(add)},                                    0, "ADD"},
+    {{do2(sub)},                                    0, "SUB"},
+    {{do2(mul)},                                    0, "MUL"},
+    {{do2(div_)},                                   0, "DIV"},
+    {{do2(min_)},                                   0, "MIN"},
+    {{do2(max_)},                                   0, "MAX"},
+    {{do2(neg)},                                    0, "NEG"},
+    {{do2(trunc8)},                                 0, "TRUNC8"},
+    {{do2(trunc16)},                                0, "TRUNC16"},
+
+    {{do2(cond)},                                   0, "COND"},
+    {{do2(and_)},                                   0, "AND"},      // 0x10
+    {{do2(or_)},                                    0, "OR"},
+    {{do2(not_)},                                   0, "NOT"},
+    {{do2(equal)},                                  0, "EQUAL"},
+    {{do2(not_eq_)},                                0, "NOT_EQ"},
+    {{do2(less)},                                   0, "LESS"},
+    {{do2(gtr)},                                    0, "GTR"},
+    {{do2(less_eq)},                                0, "LESS_EQ"},
+    {{do2(gtr_eq)},                                 0, "GTR_EQ"},   // 0x18
+
+    {{do_(next), NILOP},                            0, "NEXT"},
+    {{NILOP, NILOP},                                1, "NEXT_N"},                   // number <= smap.end - map
+    {{do_(next), NILOP},                            0, "COPY_NEXT"},
+    {{do_(put_glyph_8bit_obs), NILOP},              1, "PUT_GLYPH_8BIT_OBS"},       // output_class
+    {{do_(put_subs_8bit_obs), NILOP},               3, "PUT_SUBS_8BIT_OBS"},        // slot input_class output_class
+    {{do_(put_copy), NILOP},                        1, "PUT_COPY"},                 // slot
+    {{do_(insert), NILOP},                          0, "INSERT"},
+    {{do_(delete_), NILOP},                         0, "DELETE"},   // 0x20
+    {{do_(assoc), NILOP},                     VARARGS, "ASSOC"},
+    {{NILOP ,do_(cntxt_item)},                      2, "CNTXT_ITEM"},               // slot offset
+
+    {{do_(attr_set), NILOP},                        1, "ATTR_SET"},                 // sattrnum
+    {{do_(attr_add), NILOP},                        1, "ATTR_ADD"},                 // sattrnum
+    {{do_(attr_sub), NILOP},                        1, "ATTR_SUB"},                 // sattrnum
+    {{do_(attr_set_slot), NILOP},                   1, "ATTR_SET_SLOT"},            // sattrnum
+    {{do_(iattr_set_slot), NILOP},                  2, "IATTR_SET_SLOT"},           // sattrnum attrid
+    {{do2(push_slot_attr)},                         2, "PUSH_SLOT_ATTR"},           // sattrnum slot
+    {{do2(push_glyph_attr_obs)},                    2, "PUSH_GLYPH_ATTR_OBS"},      // gattrnum slot
+    {{do2(push_glyph_metric)},                      3, "PUSH_GLYPH_METRIC"},        // gmetric slot level
+    {{do2(push_feat)},                              2, "PUSH_FEAT"},                // featidx slot
+
+    {{do2(push_att_to_gattr_obs)},                  2, "PUSH_ATT_TO_GATTR_OBS"},    // gattrnum slot
+    {{do2(push_att_to_glyph_metric)},               3, "PUSH_ATT_TO_GLYPH_METRIC"}, // gmetric slot level
+    {{do2(push_islot_attr)},                        3, "PUSH_ISLOT_ATTR"},          // sattrnum slot attrid
+
+    {{NILOP,NILOP},                                 3, "PUSH_IGLYPH_ATTR"},
+
+    {{do2(pop_ret)},                                0, "POP_RET"},  // 0x30
+    {{do2(ret_zero)},                               0, "RET_ZERO"},
+    {{do2(ret_true)},                               0, "RET_TRUE"},
+
+    {{do_(iattr_set), NILOP},                       2, "IATTR_SET"},                // sattrnum attrid
+    {{do_(iattr_add), NILOP},                       2, "IATTR_ADD"},                // sattrnum attrid
+    {{do_(iattr_sub), NILOP},                       2, "IATTR_SUB"},                // sattrnum attrid
+    {{do2(push_proc_state)},                        1, "PUSH_PROC_STATE"},          // dummy
+    {{do2(push_version)},                           0, "PUSH_VERSION"},
+    {{do_(put_subs), NILOP},                        5, "PUT_SUBS"},                 // slot input_class input_class output_class output_class
+    {{NILOP,NILOP},                                 0, "PUT_SUBS2"},
+    {{NILOP,NILOP},                                 0, "PUT_SUBS3"},
+    {{do_(put_glyph), NILOP},                       2, "PUT_GLYPH"},                // output_class output_class
+    {{do2(push_glyph_attr)},                        3, "PUSH_GLYPH_ATTR"},          // gattrnum gattrnum slot
+    {{do2(push_att_to_glyph_attr)},                 3, "PUSH_ATT_TO_GLYPH_ATTR"},   // gattrnum gattrnum slot
+    {{do2(bor)},                                    0, "BITOR"},
+    {{do2(band)},                                   0, "BITAND"},
+    {{do2(bnot)},                                   0, "BITNOT"},   // 0x40
+    {{do2(setbits)},                                4, "BITSET"},
+    {{do_(set_feat), NILOP},                        2, "SET_FEAT"},                 // featidx slot
+    // private opcodes for internal use only, comes after all other on disk opcodes.
+    {{do_(temp_copy), NILOP},                       0, "TEMP_COPY"}
+};
diff --git a/thirdparty/graphite/src/inc/opcodes.h b/thirdparty/graphite/src/inc/opcodes.h
new file mode 100644
index 0000000000..ff2f1741e2
--- /dev/null
+++ b/thirdparty/graphite/src/inc/opcodes.h
@@ -0,0 +1,691 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2010, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+#pragma once
+// This file will be pulled into and integrated into a machine implmentation
+// DO NOT build directly and under no circumstances ever #include headers in
+// here or you will break the direct_machine.
+//
+// Implementers' notes
+// ==================
+// You have access to a few primitives and the full C++ code:
+//    declare_params(n) Tells the interpreter how many bytes of parameter
+//                      space to claim for this instruction uses and
+//                      initialises the param pointer.  You *must* before the
+//                      first use of param.
+//    use_params(n)     Claim n extra bytes of param space beyond what was
+//                      claimed using delcare_param.
+//    param             A const byte pointer for the parameter space claimed by
+//                      this instruction.
+//    binop(op)         Implement a binary operation on the stack using the
+//                      specified C++ operator.
+//    NOT_IMPLEMENTED   Any instruction body containing this will exit the
+//                      program with an assertion error.  Instructions that are
+//                      not implemented should also be marked NILOP in the
+//                      opcodes tables this will cause the code class to spot
+//                      them in a live code stream and throw a runtime_error
+//                      instead.
+//    push(n)           Push the value n onto the stack.
+//    pop()             Pop the top most value and return it.
+//
+//    You have access to the following named fast 'registers':
+//        sp        = The pointer to the current top of stack, the last value
+//                    pushed.
+//        seg       = A reference to the Segment this code is running over.
+//        is        = The current slot index
+//        isb       = The original base slot index at the start of this rule
+//        isf       = The first positioned slot
+//        isl       = The last positioned slot
+//        ip        = The current instruction pointer
+//        endPos    = Position of advance of last cluster
+//        dir       = writing system directionality of the font
+
+
+// #define NOT_IMPLEMENTED     assert(false)
+// #define NOT_IMPLEMENTED
+
+#define binop(op)           const uint32 a = pop(); *sp = uint32(*sp) op a
+#define sbinop(op)          const int32 a = pop(); *sp = int32(*sp) op a
+#define use_params(n)       dp += n
+
+#define declare_params(n)   const byte * param = dp; \
+                            use_params(n);
+
+#define push(n)             { *++sp = n; }
+#define pop()               (*sp--)
+#define slotat(x)           (map[(x)])
+#define DIE                 { is=seg.last(); status = Machine::died_early; EXIT(1); }
+#define POSITIONED          1
+
+STARTOP(nop)
+    do {} while (0);
+ENDOP
+
+STARTOP(push_byte)
+    declare_params(1);
+    push(int8(*param));
+ENDOP
+
+STARTOP(push_byte_u)
+    declare_params(1);
+    push(uint8(*param));
+ENDOP
+
+STARTOP(push_short)
+    declare_params(2);
+    const int16 r   = int16(param[0]) << 8
+                    | uint8(param[1]);
+    push(r);
+ENDOP
+
+STARTOP(push_short_u)
+    declare_params(2);
+    const uint16 r  = uint16(param[0]) << 8
+                    | uint8(param[1]);
+    push(r);
+ENDOP
+
+STARTOP(push_long)
+    declare_params(4);
+    const  int32 r  = int32(param[0]) << 24
+                    | uint32(param[1]) << 16
+                    | uint32(param[2]) << 8
+                    | uint8(param[3]);
+    push(r);
+ENDOP
+
+STARTOP(add)
+    binop(+);
+ENDOP
+
+STARTOP(sub)
+    binop(-);
+ENDOP
+
+STARTOP(mul)
+    binop(*);
+ENDOP
+
+STARTOP(div_)
+    const int32 b = pop();
+    const int32 a = int32(*sp);
+    if (b == 0 || (a == std::numeric_limits<int32>::min() && b == -1)) DIE;
+    *sp = int32(*sp) / b;
+ENDOP
+
+STARTOP(min_)
+    const int32 a = pop(), b = *sp;
+    if (a < b) *sp = a;
+ENDOP
+
+STARTOP(max_)
+    const int32 a = pop(), b = *sp;
+    if (a > b) *sp = a;
+ENDOP
+
+STARTOP(neg)
+    *sp = uint32(-int32(*sp));
+ENDOP
+
+STARTOP(trunc8)
+    *sp = uint8(*sp);
+ENDOP
+
+STARTOP(trunc16)
+    *sp = uint16(*sp);
+ENDOP
+
+STARTOP(cond)
+    const uint32 f = pop(), t = pop(), c = pop();
+    push(c ? t : f);
+ENDOP
+
+STARTOP(and_)
+    binop(&&);
+ENDOP
+
+STARTOP(or_)
+    binop(||);
+ENDOP
+
+STARTOP(not_)
+    *sp = !*sp;
+ENDOP
+
+STARTOP(equal)
+    binop(==);
+ENDOP
+
+STARTOP(not_eq_)
+    binop(!=);
+ENDOP
+
+STARTOP(less)
+    sbinop(<);
+ENDOP
+
+STARTOP(gtr)
+    sbinop(>);
+ENDOP
+
+STARTOP(less_eq)
+    sbinop(<=);
+ENDOP
+
+STARTOP(gtr_eq)
+    sbinop(>=);
+ENDOP
+
+STARTOP(next)
+    if (map - &smap[0] >= int(smap.size())) DIE
+    if (is)
+    {
+        if (is == smap.highwater())
+            smap.highpassed(true);
+        is = is->next();
+    }
+    ++map;
+ENDOP
+
+//STARTOP(next_n)
+//    use_params(1);
+//    NOT_IMPLEMENTED;
+    //declare_params(1);
+    //const size_t num = uint8(*param);
+//ENDOP
+
+//STARTOP(copy_next)
+//     if (is) is = is->next();
+//     ++map;
+// ENDOP
+
+STARTOP(put_glyph_8bit_obs)
+    declare_params(1);
+    const unsigned int output_class = uint8(*param);
+    is->setGlyph(&seg, seg.getClassGlyph(output_class, 0));
+ENDOP
+
+STARTOP(put_subs_8bit_obs)
+    declare_params(3);
+    const int           slot_ref     = int8(param[0]);
+    const unsigned int  input_class  = uint8(param[1]),
+                        output_class = uint8(param[2]);
+    uint16 index;
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        index = seg.findClassIndex(input_class, slot->gid());
+        is->setGlyph(&seg, seg.getClassGlyph(output_class, index));
+    }
+ENDOP
+
+STARTOP(put_copy)
+    declare_params(1);
+    const int  slot_ref = int8(*param);
+    if (is && !is->isDeleted())
+    {
+        slotref ref = slotat(slot_ref);
+        if (ref && ref != is)
+        {
+            int16 *tempUserAttrs = is->userAttrs();
+            if (is->attachedTo() || is->firstChild()) DIE
+            Slot *prev = is->prev();
+            Slot *next = is->next();
+            memcpy(tempUserAttrs, ref->userAttrs(), seg.numAttrs() * sizeof(uint16));
+            memcpy(is, ref, sizeof(Slot));
+            is->firstChild(NULL);
+            is->nextSibling(NULL);
+            is->userAttrs(tempUserAttrs);
+            is->next(next);
+            is->prev(prev);
+            if (is->attachedTo())
+                is->attachedTo()->child(is);
+        }
+        is->markCopied(false);
+        is->markDeleted(false);
+    }
+ENDOP
+
+STARTOP(insert)
+    if (smap.decMax() <= 0) DIE;
+    Slot *newSlot = seg.newSlot();
+    if (!newSlot) DIE;
+    Slot *iss = is;
+    while (iss && iss->isDeleted()) iss = iss->next();
+    if (!iss)
+    {
+        if (seg.last())
+        {
+            seg.last()->next(newSlot);
+            newSlot->prev(seg.last());
+            newSlot->before(seg.last()->before());
+            seg.last(newSlot);
+        }
+        else
+        {
+            seg.first(newSlot);
+            seg.last(newSlot);
+        }
+    }
+    else if (iss->prev())
+    {
+        iss->prev()->next(newSlot);
+        newSlot->prev(iss->prev());
+        newSlot->before(iss->prev()->after());
+    }
+    else
+    {
+        newSlot->prev(NULL);
+        newSlot->before(iss->before());
+        seg.first(newSlot);
+    }
+    newSlot->next(iss);
+    if (iss)
+    {
+        iss->prev(newSlot);
+        newSlot->originate(iss->original());
+        newSlot->after(iss->before());
+    }
+    else if (newSlot->prev())
+    {
+        newSlot->originate(newSlot->prev()->original());
+        newSlot->after(newSlot->prev()->after());
+    }
+    else
+    {
+        newSlot->originate(seg.defaultOriginal());
+    }
+    if (is == smap.highwater())
+        smap.highpassed(false);
+    is = newSlot;
+    seg.extendLength(1);
+    if (map != &smap[-1])
+        --map;
+ENDOP
+
+STARTOP(delete_)
+    if (!is || is->isDeleted()) DIE
+    is->markDeleted(true);
+    if (is->prev())
+        is->prev()->next(is->next());
+    else
+        seg.first(is->next());
+
+    if (is->next())
+        is->next()->prev(is->prev());
+    else
+        seg.last(is->prev());
+
+
+    if (is == smap.highwater())
+            smap.highwater(is->next());
+    if (is->prev())
+        is = is->prev();
+    seg.extendLength(-1);
+ENDOP
+
+STARTOP(assoc)
+    declare_params(1);
+    unsigned int  num = uint8(*param);
+    const int8 *  assocs = reinterpret_cast<const int8 *>(param+1);
+    use_params(num);
+    int max = -1;
+    int min = -1;
+
+    while (num-- > 0)
+    {
+        int sr = *assocs++;
+        slotref ts = slotat(sr);
+        if (ts && (min == -1 || ts->before() < min)) min = ts->before();
+        if (ts && ts->after() > max) max = ts->after();
+    }
+    if (min > -1)   // implies max > -1
+    {
+        is->before(min);
+        is->after(max);
+    }
+ENDOP
+
+STARTOP(cntxt_item)
+    // It turns out this is a cunningly disguised condition forward jump.
+    declare_params(3);
+    const int       is_arg = int8(param[0]);
+    const size_t    iskip  = uint8(param[1]),
+                    dskip  = uint8(param[2]);
+
+    if (mapb + is_arg != map)
+    {
+        ip += iskip;
+        dp += dskip;
+        push(true);
+    }
+ENDOP
+
+STARTOP(attr_set)
+    declare_params(1);
+    const attrCode      slat = attrCode(uint8(*param));
+    const          int  val  = pop();
+    is->setAttr(&seg, slat, 0, val, smap);
+ENDOP
+
+STARTOP(attr_add)
+    declare_params(1);
+    const attrCode      slat = attrCode(uint8(*param));
+    const     uint32_t  val  = pop();
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    uint32_t res = uint32_t(is->getAttr(&seg, slat, 0));
+    is->setAttr(&seg, slat, 0, int32_t(val + res), smap);
+ENDOP
+
+STARTOP(attr_sub)
+    declare_params(1);
+    const attrCode      slat = attrCode(uint8(*param));
+    const     uint32_t  val  = pop();
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    uint32_t res = uint32_t(is->getAttr(&seg, slat, 0));
+    is->setAttr(&seg, slat, 0, int32_t(res - val), smap);
+ENDOP
+
+STARTOP(attr_set_slot)
+    declare_params(1);
+    const attrCode  slat   = attrCode(uint8(*param));
+    const int       offset = int(map - smap.begin())*int(slat == gr_slatAttTo);
+    const int       val    = pop()  + offset;
+    is->setAttr(&seg, slat, offset, val, smap);
+ENDOP
+
+STARTOP(iattr_set_slot)
+    declare_params(2);
+    const attrCode  slat = attrCode(uint8(param[0]));
+    const uint8     idx  = uint8(param[1]);
+    const int       val  = int(pop()  + (map - smap.begin())*int(slat == gr_slatAttTo));
+    is->setAttr(&seg, slat, idx, val, smap);
+ENDOP
+
+STARTOP(push_slot_attr)
+    declare_params(2);
+    const attrCode      slat     = attrCode(uint8(param[0]));
+    const int           slot_ref = int8(param[1]);
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        int res = slot->getAttr(&seg, slat, 0);
+        push(res);
+    }
+ENDOP
+
+STARTOP(push_glyph_attr_obs)
+    declare_params(2);
+    const unsigned int  glyph_attr = uint8(param[0]);
+    const int           slot_ref   = int8(param[1]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+        push(int32(seg.glyphAttr(slot->gid(), glyph_attr)));
+ENDOP
+
+STARTOP(push_glyph_metric)
+    declare_params(3);
+    const unsigned int  glyph_attr  = uint8(param[0]);
+    const int           slot_ref    = int8(param[1]);
+    const signed int    attr_level  = uint8(param[2]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+        push(seg.getGlyphMetric(slot, glyph_attr, attr_level, dir));
+ENDOP
+
+STARTOP(push_feat)
+    declare_params(2);
+    const unsigned int  feat        = uint8(param[0]);
+    const int           slot_ref    = int8(param[1]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        uint8 fid = seg.charinfo(slot->original())->fid();
+        push(seg.getFeature(fid, feat));
+    }
+ENDOP
+
+STARTOP(push_att_to_gattr_obs)
+    declare_params(2);
+    const unsigned int  glyph_attr  = uint8(param[0]);
+    const int           slot_ref    = int8(param[1]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        slotref att = slot->attachedTo();
+        if (att) slot = att;
+        push(int32(seg.glyphAttr(slot->gid(), glyph_attr)));
+    }
+ENDOP
+
+STARTOP(push_att_to_glyph_metric)
+    declare_params(3);
+    const unsigned int  glyph_attr  = uint8(param[0]);
+    const int           slot_ref    = int8(param[1]);
+    const signed int    attr_level  = uint8(param[2]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        slotref att = slot->attachedTo();
+        if (att) slot = att;
+        push(int32(seg.getGlyphMetric(slot, glyph_attr, attr_level, dir)));
+    }
+ENDOP
+
+STARTOP(push_islot_attr)
+    declare_params(3);
+    const attrCode  slat     = attrCode(uint8(param[0]));
+    const int           slot_ref = int8(param[1]),
+                        idx      = uint8(param[2]);
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        int res = slot->getAttr(&seg, slat, idx);
+        push(res);
+    }
+ENDOP
+
+#if 0
+STARTOP(push_iglyph_attr) // not implemented
+    NOT_IMPLEMENTED;
+ENDOP
+#endif
+
+STARTOP(pop_ret)
+    const uint32 ret = pop();
+    EXIT(ret);
+ENDOP
+
+STARTOP(ret_zero)
+    EXIT(0);
+ENDOP
+
+STARTOP(ret_true)
+    EXIT(1);
+ENDOP
+
+STARTOP(iattr_set)
+    declare_params(2);
+    const attrCode      slat = attrCode(uint8(param[0]));
+    const uint8         idx  = uint8(param[1]);
+    const          int  val  = pop();
+    is->setAttr(&seg, slat, idx, val, smap);
+ENDOP
+
+STARTOP(iattr_add)
+    declare_params(2);
+    const attrCode      slat = attrCode(uint8(param[0]));
+    const uint8         idx  = uint8(param[1]);
+    const     uint32_t  val  = pop();
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    uint32_t res = uint32_t(is->getAttr(&seg, slat, idx));
+    is->setAttr(&seg, slat, idx, int32_t(val + res), smap);
+ENDOP
+
+STARTOP(iattr_sub)
+    declare_params(2);
+    const attrCode      slat = attrCode(uint8(param[0]));
+    const uint8         idx  = uint8(param[1]);
+    const     uint32_t  val  = pop();
+    if ((slat == gr_slatPosX || slat == gr_slatPosY) && (flags & POSITIONED) == 0)
+    {
+        seg.positionSlots(0, *smap.begin(), *(smap.end()-1), seg.currdir());
+        flags |= POSITIONED;
+    }
+    uint32_t res = uint32_t(is->getAttr(&seg, slat, idx));
+    is->setAttr(&seg, slat, idx, int32_t(res - val), smap);
+ENDOP
+
+STARTOP(push_proc_state)
+    use_params(1);
+    push(1);
+ENDOP
+
+STARTOP(push_version)
+    push(0x00030000);
+ENDOP
+
+STARTOP(put_subs)
+    declare_params(5);
+    const int        slot_ref     = int8(param[0]);
+    const unsigned int  input_class  = uint8(param[1]) << 8
+                                     | uint8(param[2]);
+    const unsigned int  output_class = uint8(param[3]) << 8
+                                     | uint8(param[4]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        int index = seg.findClassIndex(input_class, slot->gid());
+        is->setGlyph(&seg, seg.getClassGlyph(output_class, index));
+    }
+ENDOP
+
+#if 0
+STARTOP(put_subs2) // not implemented
+    NOT_IMPLEMENTED;
+ENDOP
+
+STARTOP(put_subs3) // not implemented
+    NOT_IMPLEMENTED;
+ENDOP
+#endif
+
+STARTOP(put_glyph)
+    declare_params(2);
+    const unsigned int output_class  = uint8(param[0]) << 8
+                                     | uint8(param[1]);
+    is->setGlyph(&seg, seg.getClassGlyph(output_class, 0));
+ENDOP
+
+STARTOP(push_glyph_attr)
+    declare_params(3);
+    const unsigned int  glyph_attr  = uint8(param[0]) << 8
+                                    | uint8(param[1]);
+    const int           slot_ref    = int8(param[2]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+        push(int32(seg.glyphAttr(slot->gid(), glyph_attr)));
+ENDOP
+
+STARTOP(push_att_to_glyph_attr)
+    declare_params(3);
+    const unsigned int  glyph_attr  = uint8(param[0]) << 8
+                                    | uint8(param[1]);
+    const int           slot_ref    = int8(param[2]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        slotref att = slot->attachedTo();
+        if (att) slot = att;
+        push(int32(seg.glyphAttr(slot->gid(), glyph_attr)));
+    }
+ENDOP
+
+STARTOP(temp_copy)
+    slotref newSlot = seg.newSlot();
+    if (!newSlot || !is) DIE;
+    int16 *tempUserAttrs = newSlot->userAttrs();
+    memcpy(newSlot, is, sizeof(Slot));
+    memcpy(tempUserAttrs, is->userAttrs(), seg.numAttrs() * sizeof(uint16));
+    newSlot->userAttrs(tempUserAttrs);
+    newSlot->markCopied(true);
+    *map = newSlot;
+ENDOP
+
+STARTOP(band)
+    binop(&);
+ENDOP
+
+STARTOP(bor)
+    binop(|);
+ENDOP
+
+STARTOP(bnot)
+    *sp = ~*sp;
+ENDOP
+
+STARTOP(setbits)
+    declare_params(4);
+    const uint16 m  = uint16(param[0]) << 8
+                    | uint8(param[1]);
+    const uint16 v  = uint16(param[2]) << 8
+                    | uint8(param[3]);
+    *sp = ((*sp) & ~m) | v;
+ENDOP
+
+STARTOP(set_feat)
+    declare_params(2);
+    const unsigned int  feat        = uint8(param[0]);
+    const int           slot_ref    = int8(param[1]);
+    slotref slot = slotat(slot_ref);
+    if (slot)
+    {
+        uint8 fid = seg.charinfo(slot->original())->fid();
+        seg.setFeature(fid, feat, pop());
+    }
+ENDOP
diff --git a/thirdparty/graphite/src/json.cpp b/thirdparty/graphite/src/json.cpp
new file mode 100644
index 0000000000..25f2190f71
--- /dev/null
+++ b/thirdparty/graphite/src/json.cpp
@@ -0,0 +1,147 @@
+/*  GRAPHITE2 LICENSING
+
+    Copyright 2011, SIL International
+    All rights reserved.
+
+    This library is free software; you can redistribute it and/or modify
+    it under the terms of the GNU Lesser General Public License as published
+    by the Free Software Foundation; either version 2.1 of License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    Lesser General Public License for more details.
+
+    You should also have received a copy of the GNU Lesser General Public
+    License along with this library in the file named "LICENSE".
+    If not, write to the Free Software Foundation, 51 Franklin Street,
+    Suite 500, Boston, MA 02110-1335, USA or visit their web page on the
+    internet at http://www.fsf.org/licenses/lgpl.html.
+
+Alternatively, the contents of this file may be used under the terms of the
+Mozilla Public License (http://mozilla.org/MPL) or the GNU General Public
+License, as published by the Free Software Foundation, either version 2
+of the License or (at your option) any later version.
+*/
+// JSON debug logging
+// Author: Tim Eves
+
+#if !defined GRAPHITE2_NTRACING
+
+#include <cstdio>
+#include <limits>
+#include "inc/json.h"
+
+#if defined(_MSC_VER)
+#define FORMAT_INTMAX "%lli"
+#define FORMAT_UINTMAX "%llu"
+#else
+#define FORMAT_INTMAX "%ji"
+#define FORMAT_UINTMAX "%ju"
+#endif
+
+using namespace graphite2;
+
+namespace
+{
+    enum
+    {
+        seq = ',',
+        obj='}', member=':', empty_obj='{',
+        arr=']', empty_arr='['
+    };
+}
+
+const std::nullptr_t json::null = nullptr;
+
+inline
+void json::context(const char current) throw()
+{
+    fprintf(_stream, "%c", *_context);
+    indent();
+    *_context = current;
+}
+
+
+void json::indent(const int d) throw()
+{
+    if (*_context == member || (_flatten  && _flatten < _context))
+        fputc(' ', _stream);
+    else
+        fprintf(_stream,  "\n%*s",  4*int(_context - _contexts + d), "");
+}
+
+
+inline
+void json::push_context(const char prefix, const char suffix) throw()
+{
+    assert(_context - _contexts < ptrdiff_t(sizeof _contexts));
+
+    if (_context == _contexts)
+        *_context = suffix;
+    else
+        context(suffix);
+    *++_context = prefix;
+}
+
+
+void json::pop_context() throw()
+{
+    assert(_context > _contexts);
+
+    if (*_context == seq)   indent(-1);
+    else                    fputc(*_context, _stream);
+
+    fputc(*--_context, _stream);
+    if (_context == _contexts)  fputc('\n', _stream);
+    fflush(_stream);
+
+    if (_flatten >= _context)   _flatten = 0;
+    *_context = seq;
+}
+
+
+// These four functions cannot be inlined as pointers to these
+// functions are needed for operator << (_context_t) to work.
+void json::flat(json & j) throw()   { if (!j._flatten)  j._flatten = j._context; }
+void json::close(json & j) throw()  { j.pop_context(); }
+void json::object(json & j) throw() { j.push_context('{', '}'); }
+void json::array(json & j) throw()  { j.push_context('[', ']'); }
+void json::item(json & j) throw()
+{
+    while (j._context > j._contexts+1 && j._context[-1] != arr)
+        j.pop_context();
+}
+
+
+json & json::operator << (json::string s) throw()
+{
+    const char ctxt = _context[-1] == obj ? *_context == member ? seq : member : seq;
+    context(ctxt);
+    fprintf(_stream, "\"%s\"", s);
+    if (ctxt == member) fputc(' ', _stream);
+
+    return *this;
+}
+
+json & json::operator << (json::number f) throw()
+{
+    context(seq);
+    if (std::numeric_limits<json::number>::infinity() == f)
+        fputs("Infinity", _stream);
+    else if (-std::numeric_limits<json::number>::infinity() == f)
+        fputs("-Infinity", _stream);
+    else if (std::numeric_limits<json::number>::quiet_NaN() == f ||
+            std::numeric_limits<json::number>::signaling_NaN() == f)
+        fputs("NaN", _stream);
+    else
+        fprintf(_stream, "%g", f);
+    return *this;
+}
+json & json::operator << (json::integer d) throw()  { context(seq); fprintf(_stream, FORMAT_INTMAX, intmax_t(d)); return *this; }
+json & json::operator << (json::integer_u d) throw()  { context(seq); fprintf(_stream, FORMAT_UINTMAX, uintmax_t(d)); return *this; }
+json & json::operator << (json::boolean b) throw()  { context(seq); fputs(b ? "true" : "false", _stream); return *this; }
+json & json::operator << (std::nullptr_t)  throw()  { context(seq); fputs("null",_stream); return *this; }
+
+#endif
diff --git a/thirdparty/harfbuzz/AUTHORS b/thirdparty/harfbuzz/AUTHORS
new file mode 100644
index 0000000000..83c0c66f99
--- /dev/null
+++ b/thirdparty/harfbuzz/AUTHORS
@@ -0,0 +1,14 @@
+Behdad Esfahbod
+David Corbett
+David Turner
+Ebrahim Byagowi
+Garret Rieger
+Jonathan Kew
+Khaled Hosny
+Lars Knoll
+Martin Hosken
+Owen Taylor
+Roderick Sheeter
+Roozbeh Pournader
+Simon Hausmann
+Werner Lemberg
diff --git a/thirdparty/harfbuzz/COPYING b/thirdparty/harfbuzz/COPYING
new file mode 100644
index 0000000000..57343164f2
--- /dev/null
+++ b/thirdparty/harfbuzz/COPYING
@@ -0,0 +1,38 @@
+HarfBuzz is licensed under the so-called "Old MIT" license.  Details follow.
+For parts of HarfBuzz that are licensed under different licenses see individual
+files names COPYING in subdirectories where applicable.
+
+Copyright © 2010,2011,2012,2013,2014,2015,2016,2017,2018,2019,2020  Google, Inc.
+Copyright © 2018,2019,2020  Ebrahim Byagowi
+Copyright © 2019,2020  Facebook, Inc. 
+Copyright © 2012  Mozilla Foundation
+Copyright © 2011  Codethink Limited
+Copyright © 2008,2010  Nokia Corporation and/or its subsidiary(-ies)
+Copyright © 2009  Keith Stribley
+Copyright © 2009  Martin Hosken and SIL International
+Copyright © 2007  Chris Wilson
+Copyright © 2006  Behdad Esfahbod
+Copyright © 2005  David Turner
+Copyright © 2004,2007,2008,2009,2010  Red Hat, Inc.
+Copyright © 1998-2004  David Turner and Werner Lemberg
+
+For full copyright notices consult the individual files in the package.
+
+
+Permission is hereby granted, without written agreement and without
+license or royalty fees, to use, copy, modify, and distribute this
+software and its documentation for any purpose, provided that the
+above copyright notice and the following two paragraphs appear in
+all copies of this software.
+
+IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+DAMAGE.
+
+THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
diff --git a/thirdparty/harfbuzz/NEWS b/thirdparty/harfbuzz/NEWS
new file mode 100644
index 0000000000..f211a3781c
--- /dev/null
+++ b/thirdparty/harfbuzz/NEWS
@@ -0,0 +1,2412 @@
+Overview of changes leading to 2.7.2
+Saturday, August 29, 2020
+====================================
+- Fix a regression in the previous release that caused a crash with Kaithi.
+- More OOM fixes.
+
+
+Overview of changes leading to 2.7.1
+Thursday, August 13, 2020
+====================================
+- ot-funcs now handles variable empty glyphs better when hvar/vvar isn't present.
+- Reverted a GDEF processing regression.
+- A couple of fixes to handle OOM better.
+
+
+Overview of changes leading to 2.7.0
+Saturday, July 25, 2020
+====================================
+- Use an implementation for round that always rounds up, some minor fluctuations
+  are expected on var font specially when hb-ot callback is used.
+- Fix an AAT's `kerx` issue on broken rendering of Devanagari Sangam MN.
+- Remove AAT's `lcar` table support from _get_ligature_carets API, not even much
+  use on macOS installed fonts (only two files).  GDEF support is the recommended
+  one and expected to work properly after issues fixed two releases ago.
+- Minor memory fixes to handle OOM better specially in hb-ft.
+- Minor .so files versioning scheme change and remove stable/unstable scheme
+  differences, was never used in practice (always default to stable scheme).
+- We are now suggesting careful packaging of the library using meson,
+  https://github.com/harfbuzz/harfbuzz/wiki/Notes-on-migration-to-meson
+  for more information.
+- Distribution package URL is changed, either use GitHub generated tarballs,
+  `https://github.com/harfbuzz/harfbuzz/archive/$pkgver.tar.gz`
+  or, even more preferably use commit hash of the release and git checkouts like,
+  `git+https://github.com/harfbuzz/harfbuzz#commit=$commit`
+
+
+Overview of changes leading to 2.6.8
+Monday, June 22, 2020
+====================================
+- New API to fetch glyph alternates from GSUB table.
+- hb-coretext build fix for macOS < 10.10.
+- Meson build fixes, cmake port removal is postponed but please prepare for
+  it and give us feedback.
+  Autotools is still our main build system however please consider
+  experimenting with meson also for packaging the library.
+- New API:
++hb_ot_layout_lookup_get_glyph_alternates()
+
+
+Overview of changes leading to 2.6.7
+Wednesday, June 3, 2020
+====================================
+- Update to Unicode 13.0.0.
+- Fix hb_ot_layout_get_ligature_carets for fonts without lcar table, it was
+  completely broken for all the other fonts since 2.1.2.
+- As a part of our migration to meson, this release will be the last one
+  to provide cmake port files but autotools still is our main build system.
+  There is a possibility that the next version or the after be released
+  using meson.
+
+
+Overview of changes leading to 2.6.6
+Tuesday, May 12, 2020
+====================================
+- A fix in AAT kerning for Geeza Pro.
+- Better support for resource fork fonts on macOS.
+
+
+Overview of changes leading to 2.6.5
+Friday, April 17, 2020
+====================================
+- Add experimental meson build system.  Autotools is still the primary
+  and supported build system.
+- AAT is now always preferred for horizontal scripts when both AAT and OT
+  layout tables exist at the same time.
+- Subsetter improvements.
+- New API:
++hb_ft_font_lock_face()
++hb_ft_font_unlock_face()
+
+
+Overview of changes leading to 2.6.4
+Monday, October 29, 2019
+====================================
+- Small bug fix.
+- Build fixes.
+
+
+Overview of changes leading to 2.6.3
+Monday, October 28, 2019
+====================================
+- Misc small fixes, mostly to build-related issues.
+- New API:
++hb_font_get_nominal_glyphs()
+
+
+Overview of changes leading to 2.6.2
+Monday, September 30, 2019
+====================================
+- Misc small fixes, mostly to build-related issues.
+
+
+Overview of changes leading to 2.6.1
+Thursday, August 22, 2019
+====================================
+- Fix regression with hb_font_create_sub_font scaling introduced in 2.6.0.
+- Change interpretation of font PTEM size / CoreText font size handling.
+  See https://github.com/harfbuzz/harfbuzz/pull/1484
+- hb-ot-font: Prefer symbol cmap subtable if present.
+- Apply 'dist'/'abvm'/'blwm' features to all scripts.
+- Drop experimental DirectWrite API.
+
+
+Overview of changes leading to 2.6.0
+Tuesday, August 13, 2019
+====================================
+- New OpenType metrics, baseline, and metadata table access APIs.
+- New API to set font variations to a named-instance.
+- New hb-gdi.h header and API for creating hb_face_t from HFONT.
+- Amalgam: Provide a single-file harfbuzz.cc file for easier alternate building.
+- More size-reduction configurable options, enabled by HB_TINY.
+- New API:
++hb_font_set_var_named_instance()
++hb_gdi_face_create()
++hb_ot_layout_baseline_tag_t
++hb_ot_layout_get_baseline()
++hb_ot_meta_tag_t
++hb_ot_meta_get_entry_tags()
++hb_ot_meta_reference_entry()
++hb_ot_metrics_tag_t
++hb_ot_metrics_get_position()
++hb_ot_metrics_get_variation()
++hb_ot_metrics_get_x_variation()
++hb_ot_metrics_get_y_variation()
+
+
+Overview of changes leading to 2.5.3
+Wednesday, June 26, 2019
+====================================
+- Fix UCD script data for Unicode 10+ scripts.  This was broken since 2.5.0.
+- More optimizations for HB_TINY.
+
+
+Overview of changes leading to 2.5.2
+Thursday, June 20, 2019
+====================================
+- More hb-config.hh facilities to shrink library size, namely when built as
+  HB_TINY.
+- New documentation of custom configurations in CONFIG.md.
+- Fix build on gcc 4.8.  That's supported again.
+- Universal Shaping Engine improvements thanks to David Corbett.
+- API Changes: Undeprecate some horizontal-kerning API and re-enable in hb-ft,
+  such that Type1 fonts will continue kerning.
+
+
+Overview of changes leading to 2.5.1
+Friday, May 31, 2019
+====================================
+- Fix build with various versions of Visual Studio.
+- Improved documentation, thanks to Nathan Willis.
+- Bugfix in subsetting glyf table.
+- Improved scripts for cross-compiling for Windows using mingw.
+- Rename HB_MATH_GLYPH_PART_FLAG_EXTENDER to HB_OT_MATH_GLYPH_PART_FLAG_EXTENDER.
+  A deprecated macro is added for backwards-compatibility.
+
+
+Overview of changes leading to 2.5.0
+Friday, May 24, 2019
+====================================
+- This release does not include much functional changes, but includes major internal
+  code-base changes.  We now require C++11.  Support for gcc 4.8 and earlier has been
+  dropped.
+- New hb-config.hh facility for compiling smaller library for embedded and web usecases.
+- New Unicode Character Databse implementation that is half the size of previously-used
+  UCDN.
+- Subsetter improvements.
+- Improved documentation, thanks to Nathan Willis.
+- Misc shaping fixes.
+
+
+Overview of changes leading to 2.4.0
+Monday, March 25, 2019
+====================================
+- Unicode 12.
+- Misc fixes.
+- Subsetter improvements.
+- New API:
+HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE
+hb_directwrite_face_create()
+
+
+Overview of changes leading to 2.3.1
+Wednesday, January 30, 2019
+====================================
+- AAT bug fixes.
+- Misc internal housekeeping cleanup.
+
+
+Overview of changes leading to 2.3.0
+Thursday, December 20, 2018
+====================================
+- Fix regression on big-endian architectures.  Ouch!
+- Misc bug and build fixes.
+- Fix subsetting of simple GSUB/GDEF.
+- Merge CFF / CFF2 support contributed by Adobe.  This mostly involves
+  the subsetter, but also get_glyph_extents on CFF fonts.
+
+New API in hb-aat.h:
++hb_aat_layout_has_substitution()
++hb_aat_layout_has_positioning()
++hb_aat_layout_has_tracking()
+
+
+Overview of changes leading to 2.2.0
+Thursday, November 29, 2018
+====================================
+- Misc shaping bug fixes.
+- Add font variations named-instance API.
+- Deprecate font variations axis enumeration API and add replacement.
+- AAT shaping improvements:
+  o Fixed 'kern' table Format 2 implementation.
+  o Implement 'feat' table API for feature detection.
+  o Blacklist 'GSUB' table of fonts from 'MUTF' foundry that also have 'morx'.
+
+New API:
++hb_aat_layout_feature_type_t
++hb_aat_layout_feature_selector_t
++hb_aat_layout_get_feature_types()
++hb_aat_layout_feature_type_get_name_id
++hb_aat_layout_feature_selector_info_t
++HB_AAT_LAYOUT_NO_SELECTOR_INDEX
++hb_aat_layout_feature_type_get_selector_infos()
++hb_ot_var_axis_flags_t
++hb_ot_var_axis_info_t
++hb_ot_var_get_axis_infos()
++hb_ot_var_find_axis_info()
++hb_ot_var_get_named_instance_count()
++hb_ot_var_named_instance_get_subfamily_name_id()
++hb_ot_var_named_instance_get_postscript_name_id()
++hb_ot_var_named_instance_get_design_coords()
+
+Deprecated API:
++HB_OT_VAR_NO_AXIS_INDEX
++hb_ot_var_axis_t
++hb_ot_var_get_axes()
++hb_ot_var_find_axis()
+
+
+Overview of changes leading to 2.1.3
+Friday, November 16, 2018
+====================================
+- Fix AAT 'mort' shaping, which was broken in 2.1.2
+
+
+Overview of changes leading to 2.1.2
+Friday, November 16, 2018
+====================================
+- Various internal changes.
+- AAT shaping improvements:
+  o Implement kern table Format 1 state-machine-based kerning.
+  o Implement cross-stream kerning (cursive positioning, etc).
+  o Ignore emptyish GSUB tables (zero scripts) if morx present.
+  o Don't apply GPOS if morx is being applied.  Matches Apple.
+
+
+-Overview of changes leading to 2.1.1
+Monday, November 5, 2018
+====================================
+- AAT improvements:
+  o Implement 'mort' table.
+  o Implement 'kern' subtables Format 1 and Format 3.
+
+
+Overview of changes leading to 2.1.0
+Tuesday, October 30, 2018
+====================================
+- AAT shaping improvements:
+  o Allow user controlling AAT features, for whole buffer only currently.
+  o Several 'morx' fixes.
+  o Implement tuple-kerns in 'kerx'; Fixes kerning with Apple default
+    San Francisco fonts.
+- Support for color fonts:
+  o COLR/CPAL API to fetch color layers.
+  o SVG table to fetch SVG documents.
+  o CBDT/sbix API to fetch PNG images.
+- New 'name' table API.
+- hb-ot-font now uses 'VORG' table to correctly position CFF glyphs
+  in vertical layout.
+- Various fuzzer-found bug fixes.
+
+Changed API:
+
+A type and a macro added in 2.0.0 were renamed:
+
+hb_name_id_t -> hb_ot_name_id_t
+HB_NAME_ID_INVALID -> HB_OT_NAME_ID_INVALID
+
+New API:
+
++hb_color_t
++HB_COLOR
++hb_color_get_alpha()
++hb_color_get_red()
++hb_color_get_green()
++hb_color_get_blue()
++hb_ot_color_has_palettes()
++hb_ot_color_palette_get_count()
++hb_ot_color_palette_get_name_id()
++hb_ot_color_palette_color_get_name_id()
++hb_ot_color_palette_flags_t
++hb_ot_color_palette_get_flags()
++hb_ot_color_palette_get_colors()
++hb_ot_color_has_layers()
++hb_ot_color_layer_t
++hb_ot_color_glyph_get_layers()
++hb_ot_color_has_svg()
++hb_ot_color_glyph_reference_svg()
++hb_ot_color_has_png()
++hb_ot_color_glyph_reference_png()
+
++hb_ot_name_id_t
++HB_OT_NAME_ID_INVALID
++HB_OT_NAME_ID_COPYRIGHT
++HB_OT_NAME_ID_FONT_FAMILY
++HB_OT_NAME_ID_FONT_SUBFAMILY
++HB_OT_NAME_ID_UNIQUE_ID
++HB_OT_NAME_ID_FULL_NAME
++HB_OT_NAME_ID_VERSION_STRING
++HB_OT_NAME_ID_POSTSCRIPT_NAME
++HB_OT_NAME_ID_TRADEMARK
++HB_OT_NAME_ID_MANUFACTURER
++HB_OT_NAME_ID_DESIGNER
++HB_OT_NAME_ID_DESCRIPTION
++HB_OT_NAME_ID_VENDOR_URL
++HB_OT_NAME_ID_DESIGNER_URL
++HB_OT_NAME_ID_LICENSE
++HB_OT_NAME_ID_LICENSE_URL
++HB_OT_NAME_ID_TYPOGRAPHIC_FAMILY
++HB_OT_NAME_ID_TYPOGRAPHIC_SUBFAMILY
++HB_OT_NAME_ID_MAC_FULL_NAME
++HB_OT_NAME_ID_SAMPLE_TEXT
++HB_OT_NAME_ID_CID_FINDFONT_NAME
++HB_OT_NAME_ID_WWS_FAMILY
++HB_OT_NAME_ID_WWS_SUBFAMILY
++HB_OT_NAME_ID_LIGHT_BACKGROUND
++HB_OT_NAME_ID_DARK_BACKGROUND
++HB_OT_NAME_ID_VARIATIONS_PS_PREFIX
++hb_ot_name_entry_t
++hb_ot_name_list_names()
++hb_ot_name_get_utf8()
++hb_ot_name_get_utf16()
++hb_ot_name_get_utf32()
+
+
+Overview of changes leading to 2.0.2
+Saturday, October 20, 2018
+====================================
+- Fix two minor memory access issues in AAT tables.
+
+
+Overview of changes leading to 2.0.1
+Friday, October 19, 2018
+====================================
+- Fix hb-version.h reported release version that went wrong (1.8.0)
+  with previous release.
+- Fix extrapolation in 'trak' table.
+- Fix hb-font infinite-recursion issue with some font funcs and
+  subclassed fonts.
+- Implement variation-kerning format in kerx table, although without
+  variation.
+- Fix return value of hb_map_is_empty().
+
+
+Overview of changes leading to 2.0.0
+Thursday, October 18, 2018
+====================================
+- Added AAT shaping support (morx/kerx/trak).
+  Automatically used if GSUB/GPOS are not available respectively.
+  Set HB_OPTIONS=aat env var to have morx/kerx preferred over
+  GSUB/GPOS.
+- Apply TrueType kern table internally, instead of relying on
+  hb_font_t callbacks.
+- Khmer shaper significantly rewritten to better match Uniscribe.
+- Indic3 tags ('dev3', etc) are passed to USE shaper.
+- .dfont Mac font containers implemented.
+- Script- and language-mapping revamped to better use BCP 47.
+- Misc USE and Indic fixes.
+- Misc everything fixes.
+- Too many things to list.  Biggest release since 0.9.1, with
+  over 500 commits in just over 5 weeks!  Didn't intend it to
+  be a big release.  Just happened to become.
+- hb-ft now locks underlying FT_Face during use.
+
+API changes:
+
+- Newly-created hb_font_t's now have our internal "hb-ot-font"
+  callbacks set on them, so they should work out of the box
+  without any callbacks set.  If callbacks are set, everything
+  is back to what it was before, the fallback callbacks are
+  null.  If you to get the internal implementation modified,
+  sub_font it.
+
+- New hb_font_funcs_set_nominal_glyphs_func() allows speeding
+  up character to glyph mapping.
+
+New API:
++HB_FEATURE_GLOBAL_START
++HB_FEATURE_GLOBAL_END
++hb_buffer_set_invisible_glyph()
++hb_buffer_get_invisible_glyph()
++hb_font_funcs_set_nominal_glyphs_func()
++hb_ot_layout_table_select_script()
++hb_ot_layout_script_select_language()
++hb_ot_layout_feature_get_name_ids()
++hb_ot_layout_feature_get_characters()
++hb_name_id_t
++HB_NAME_ID_INVALID
++HB_OT_MAX_TAGS_PER_SCRIPT
++hb_ot_tags_from_script_and_language()
++hb_ot_tags_to_script_and_language()
+
+Deprecated API:
+-hb_font_funcs_set_glyph_func()
+-hb_unicode_eastasian_width_func_t
+-hb_unicode_funcs_set_eastasian_width_func()
+-hb_unicode_eastasian_width()
+-hb_unicode_decompose_compatibility_func_t
+-HB_UNICODE_MAX_DECOMPOSITION_LEN
+-hb_unicode_funcs_set_decompose_compatibility_func()
+-hb_unicode_decompose_compatibility()
+-hb_font_funcs_set_glyph_h_kerning_func()
+-hb_font_funcs_set_glyph_v_kerning_func()
+-hb_font_get_glyph_h_kerning()
+-hb_font_get_glyph_v_kerning()
+-hb_font_get_glyph_kerning_for_direction()
+-hb_ot_layout_table_choose_script()
+-hb_ot_layout_script_find_language()
+-hb_ot_tags_from_script()
+-hb_ot_tag_from_language()
+
+
+Overview of changes leading to 1.9.0
+Monday, September 10, 2018
+====================================
+- Added 'cmap' API to hb_face_t.
+- Face-builder API.
+- hb-ot-font re-creation should be much leaner now, as the
+  font tables it uses are cached on hb_face_t now.
+- Internal source header file name changes:
+  hb-*-private.hh is renamed to hb-*.hh.
+
+New API:
++HB_UNICODE_MAX
++hb_face_collect_unicodes()
++hb_face_collect_variation_selectors()
++hb_face_collect_variation_unicodes()
++hb_face_builder_create()
++hb_face_builder_add_table()
+
+
+Overview of changes leading to 1.8.8
+Tuesday, August 14, 2018
+====================================
+- Fix hb-icu crash on architectures where compare_exchange_weak() can
+  fail falsely.  This bug was introduced in 1.8.4.
+  https://bugs.chromium.org/p/chromium/issues/detail?id=873568
+- More internal refactoring of atomic operations and singletons.
+- API changes:
+  The following functions do NOT reference their return value before
+  returning:
+  * hb_unicode_funcs_get_default()
+  * hb_glib_get_unicode_funcs()
+  * hb_icu_get_unicode_funcs()
+  This is consistent with their naming ("get", instead of "reference")
+  as well as how they are used in the wild (ie. no one calls destroy()
+  on their return value.)
+
+
+Overview of changes leading to 1.8.7
+Wednesday, August 8, 2018
+====================================
+- Fix assertion failure with GDEF-blacklisted fonts.
+
+
+Overview of changes leading to 1.8.6
+Tuesday, August 7, 2018
+====================================
+- Internal code shuffling.
+- New API to speed up getting advance widths for implementations
+  that have heavy overhead in get_h_advance callback:
++hb_font_funcs_set_glyph_h_advances_func
++hb_font_funcs_set_glyph_v_advances_func
++hb_font_get_glyph_advances_for_direction
++hb_font_get_glyph_h_advances
++hb_font_get_glyph_h_advances_func_t
++hb_font_get_glyph_v_advances
++hb_font_get_glyph_v_advances_func_t
+
+
+Overview of changes leading to 1.8.5
+Wednesday, August 1, 2018
+====================================
+- Major Khmer shaper improvements to better match Microsoft.
+- Indic bug fixes.
+- Internal improvements to atomic operations.
+
+
+Overview of changes leading to 1.8.4
+Tuesday, July 17, 2018
+====================================
+- Fix build on non-C++11.
+- Use C++-style GCC atomics and C++11 atomics.
+
+
+Overview of changes leading to 1.8.3
+Wednesday, July 11, 2018
+====================================
+- A couple of Indic / USE bug fixes.
+- Disable vectorization, as it was causing unaligned access bus error on
+  certain 32bit architectures.
+
+
+Overview of changes leading to 1.8.2
+Tuesday, July 3, 2018
+====================================
+- Fix infinite loop in Khmer shaper.
+- Improve hb_blob_create_from_file() for streams.
+
+
+Overview of changes leading to 1.8.1
+Tuesday, June 12, 2018
+====================================
+- Fix hb-version.h file generation; last two releases went out with wrong ones.
+- Add correctness bug in hb_set_t operations, introduced in 1.7.7.
+- Remove HB_SUBSET_BUILTIN build option.  Not necessary.
+
+
+Overview of changes leading to 1.8.0
+Tuesday, June 5, 2018
+====================================
+- Update to Unicode 11.0.0.
+
+
+Overview of changes leading to 1.7.7
+Tuesday, June 5, 2018
+====================================
+- Lots of internal changes, but not yet exposed externally.
+- All HarfBuzz objects are significantly smaller in size now.
+- Sinhala: Position repha on top of post-consonant, not base.
+  This better matches Windows 10 behavior, which was changed
+  from previous Windows versions.
+- New build options:
+  o New cpp macro HB_NO_ATEXIT
+  o New cpp macro HB_SUBSET_BUILTIN
+- Significant libharfbuzz-subset changes. API subject to change.
+- New API in libharfbuzz:
+
++hb_blob_create_from_file()
++hb_face_count()
+
+A hashmap implementation:
++hb-map.h
++HB_MAP_VALUE_INVALID
++hb_map_t
++hb_map_create()
++hb_map_get_empty()
++hb_map_reference()
++hb_map_destroy()
++hb_map_set_user_data()
++hb_map_get_user_data()
++hb_map_allocation_successful()
++hb_map_clear()
++hb_map_is_empty()
++hb_map_get_population()
++hb_map_set()
++hb_map_get()
++hb_map_del()
++hb_map_has()
+
+
+Overview of changes leading to 1.7.6
+Wednesday, March 7, 2018
+====================================
+
+- Fix to hb_set_t binary operations. Ouch.
+- New experimental harfbuzz-subset library. All of hb-subset.h
+  is experimental right now and API WILL change.
+
+- New API:
+hb_blob_copy_writable_or_fail()
+HB_OT_TAG_BASE
+hb_set_previous()
+hb_set_previous_range()
+
+
+Overview of changes leading to 1.7.5
+Tuesday, January 30, 2018
+====================================
+
+- Separate Khmer shaper from Indic.
+- First stab at AAT morx. Not hooked up.
+- Misc bug fixes.
+
+
+Overview of changes leading to 1.7.4
+Wednesday, December 20, 2017
+====================================
+
+- Fix collect_glyphs() regression caused by hb_set_t changes.
+
+
+Overview of changes leading to 1.7.3
+Monday, December 18, 2017
+====================================
+
+- hb_set_t performance tuning and optimizations.
+- Speed up collect_glyphs() and reject garbage data.
+- In hb_coretext_font_create() set font point-size (ptem).
+- Misc fixes.
+
+
+Overview of changes leading to 1.7.2
+Monday, December 4, 2017
+====================================
+
+- Optimize hb_set_add_range().
+- Misc fixes.
+- New API:
+hb_coretext_font_create()
+
+
+Overview of changes leading to 1.7.1
+Tuesday, November 14, 2017
+====================================
+
+- Fix atexit object destruction regression.
+- Fix minor integer-overflow.
+
+
+Overview of changes leading to 1.7.0
+Monday, November 13, 2017
+====================================
+
+- Minor Indic fixes.
+- Implement kerning and glyph names in hb-ot-font.
+- Various DSO optimization re .data and .bss sizes.
+- Make C++11 optional; build fixes.
+- Mark all other backends "unsafe-to-break".
+- Graphite fix.
+
+
+Overview of changes leading to 1.6.3
+Thursday, October 26th, 2017
+====================================
+
+- Fix hb_set_t some more.  Should be solid now.
+- Implement get_glyph_name() for hb-ot-font.
+- Misc fixes.
+
+
+Overview of changes leading to 1.6.2
+Monday, October 23nd, 2017
+====================================
+
+- Yesterday's release had a bad crasher; don't use it.  That's what
+  happens when one works on Sunday...
+  https://github.com/harfbuzz/harfbuzz/issues/578
+- Build fixes for FreeBSD and Chrome Android.
+
+
+Overview of changes leading to 1.6.1
+Sunday, October 22nd, 2017
+====================================
+
+- Don't skip over COMBINING GRAPHEME JOINER when ligating, etc.
+  To be refined: https://github.com/harfbuzz/harfbuzz/issues/554
+- Faster hb_set_t implementation.
+- Don't use deprecated ICU API.
+- Fix undefined-behavior in Myanmar shaper, introduced in 1.6.0
+- Deprecated API:
+  hb_set_invert()
+
+
+Overview of changes leading to 1.6.0
+Friday, October the 13th, 2017
+====================================
+
+- Update to Unicode 10.
+
+- Various Indic and Universal Shaping Engine fixes as a result of
+  HarfBuzz Hackfest with Jonathan Kew at Web Engines Hackfest at
+  the Igalia offices in A Coruña, Spain.  Thanks Igalia for having
+  us!
+
+- Implement Unicode Arabic Mark Ordering Algorithm UTR#53.
+
+- Implement optical sizing / tracking in CoreText backend, using
+  new API hb_font_set_ptem().
+
+- Allow notifying hb_font_t that underlying FT_Face changed sizing,
+  using new API hb_ft_font_changed().
+
+- More Graphite backend RTL fixes.
+
+- Fix caching of variable font shaping plans.
+
+- hb-view / hb-shape now accept following new arguments:
+
+  o --unicodes: takes a list of hex numbers that represent Unicode
+    codepoints.
+
+New API:
++hb_face_get_table_tags()
++hb_font_set_ptem()
++hb_font_get_ptem()
++hb_ft_font_changed()
+
+
+Overview of changes leading to 1.5.1
+Tuesday, September 5, 2017
+====================================
+
+- Fix "unsafe-to-break" in fallback shaping and other corner cases.
+  All our tests pass with --verify now, meaning unsafe-to-break API
+  works as expected.
+- Add --unicodes to hb-view / hb-shape.
+- [indic] Treat Consonant_With_Stacker as consonant.  This will need
+  further tweaking.
+- hb_buffer_diff() tweaks.
+
+
+Overview of changes leading to 1.5.0
+Wednesday, August 23, 2017
+====================================
+
+- Misc new API, for appending a buffer to another, and for comparing
+  contents of two buffers for types of differences.
+
+- New "unsafe-to-break" API.  Can be used to speed up reshaping
+  in line-breaking situations.  Essentially, after shaping, it returns
+  positions in the input string (some of the cluster boundaries) that
+  are "safe to break" in that if the text is segmented at that position
+  and two sides reshaped and concatenated, the shaping result is
+  exactly the same as shaping the text in one piece.
+
+  hb-view and hb-shape and hb-shape now take --verify, which verifies
+  the above property.
+
+  Some corner cases of the implementation are still not quite working.
+  Those will be fixed in subsequent releases.
+
+- New API:
+
+hb_buffer_append()
+
+hb_glyph_flags_t
+HB_GLYPH_FLAG_UNSAFE_TO_BREAK
+HB_GLYPH_FLAG_DEFINED
+hb_glyph_info_get_glyph_flags()
+
+HB_BUFFER_SERIALIZE_FLAG_GLYPH_FLAGS
+
+hb_buffer_diff_flags_t
+HB_BUFFER_DIFF_FLAG_EQUAL
+HB_BUFFER_DIFF_FLAG_CONTENT_TYPE_MISMATCH
+HB_BUFFER_DIFF_FLAG_LENGTH_MISMATCH
+HB_BUFFER_DIFF_FLAG_NOTDEF_PRESENT
+HB_BUFFER_DIFF_FLAG_DOTTED_CIRCLE_PRESENT
+HB_BUFFER_DIFF_FLAG_CODEPOINT_MISMATCH
+HB_BUFFER_DIFF_FLAG_CLUSTER_MISMATCH
+HB_BUFFER_DIFF_FLAG_GLYPH_FLAGS_MISMATCH
+HB_BUFFER_DIFF_FLAG_POSITION_MISMATCH
+hb_buffer_diff
+
+
+Overview of changes leading to 1.4.8
+Tuesday, August 8, 2017
+====================================
+
+- Major fix to avar table handling.
+- Rename hb-shape --show-message to --trace.
+- Build fixes.
+
+
+Overview of changes leading to 1.4.7
+Tuesday, July 18, 2017
+====================================
+
+- Multiple Indic, Tibetan, and Cham fixes.
+- CoreText: Allow disabling kerning.
+- Adjust Arabic feature order again.
+- Misc build fixes.
+
+
+Overview of changes leading to 1.4.6
+Sunday, April 23, 2017
+====================================
+
+- Graphite2: Fix RTL positioning issue.
+- Backlist GDEF of more versions of Padauk and Tahoma.
+- New, experimental, cmake alternative build system.
+
+
+Overview of changes leading to 1.4.5
+Friday, March 10, 2017
+====================================
+
+- Revert "Fix Context lookup application when moving back after a glyph..."
+  This introduced memory access problems.  To be fixed properly soon.
+
+
+Overview of changes leading to 1.4.4
+Sunday, March 5, 2017
+====================================
+
+- Fix Context lookup application when moving back after a glyph deletion.
+- Fix buffer-overrun in Bengali.
+
+
+Overview of changes leading to 1.4.3
+Saturday, February 25, 2017
+====================================
+
+- Route Adlam script to Arabic shaper.
+- Misc fixes.
+- New API:
+  hb_font_set_face()
+- Deprecate API:
+  hb_graphite2_font_get_gr_font()
+
+
+Overview of changes leading to 1.4.2
+Monday, January 23, 2017
+====================================
+
+- Implement OpenType Font Variation tables avar/fvar/HVAR/VVAR.
+- hb-shape and hb-view now accept --variations.
+- New API:
+
+hb_variation_t
+hb_variation_from_string()
+hb_variation_to_string()
+
+hb_font_set_variations()
+hb_font_set_var_coords_design()
+hb_font_get_var_coords_normalized()
+
+hb-ot-var.h:
+hb_ot_var_axis_t
+hb_ot_var_has_data()
+hb_ot_var_get_axis_count()
+hb_ot_var_get_axes()
+hb_ot_var_find_axis()
+hb_ot_var_normalize_variations()
+hb_ot_var_normalize_coords()
+
+- MVAR to be implemented later.  Access to named instances to be
+  implemented later as well.
+
+- Misc fixes.
+
+
+Overview of changes leading to 1.4.1
+Thursday, January 5, 2017
+====================================
+
+- Always build and use UCDN for Unicode data by default.
+  Reduces dependence on version of Unicode data in glib,
+  specially in the Windows bundles we are shipping, which
+  have very old glib.
+
+
+Overview of changes leading to 1.4.0
+Thursday, January 5, 2017
+====================================
+
+- Merged "OpenType GX" branch which adds core of support for
+  OpenType 1.8 Font Variations.  To that extent, the relevant
+  new API is:
+
+New API:
+hb_font_set_var_coords_normalized()
+
+  with supporting API:
+
+New API:
+HB_OT_LAYOUT_NO_VARIATIONS_INDEX
+hb_ot_layout_table_find_feature_variations()
+hb_ot_layout_feature_with_variations_get_lookups()
+hb_shape_plan_create2()
+hb_shape_plan_create_cached2()
+
+  Currently variations in GSUB/GPOS/GDEF are fully supported,
+  and no other tables are supported.  In particular, fvar/avar
+  are NOT supported, hence the hb_font_set_var_coords_normalized()
+  taking normalized coordinates.  API to take design coordinates
+  will be added in the future.
+
+  HVAR/VVAR/MVAR support will also be added to hb-ot-font in the
+  future.
+
+- Fix regression in GDEF glyph class processing.
+- Add decompositions for Chakma, Limbu, and Balinese in USE shaper.
+- Misc fixes.
+
+
+Overview of changes leading to 1.3.4
+Monday, December 5, 2016
+====================================
+
+- Fix vertical glyph origin in hb-ot-font.
+- Implement CBDT/CBLC color font glyph extents in hb-ot-font.
+
+
+Overview of changes leading to 1.3.3
+Wednesday, September 28, 2016
+====================================
+
+- Implement parsing of OpenType MATH table.
+New API:
+HB_OT_TAG_MATH
+HB_OT_MATH_SCRIPT
+hb_ot_math_constant_t
+hb_ot_math_kern_t
+hb_ot_math_glyph_variant_t
+hb_ot_math_glyph_part_flags_t
+hb_ot_math_glyph_part_t
+hb_ot_math_has_data
+hb_ot_math_get_constant
+hb_ot_math_get_glyph_italics_correction
+hb_ot_math_get_glyph_top_accent_attachment
+hb_ot_math_get_glyph_kerning
+hb_ot_math_is_glyph_extended_shape
+hb_ot_math_get_glyph_variants
+hb_ot_math_get_min_connector_overlap
+hb_ot_math_get_glyph_assembly
+
+
+Overview of changes leading to 1.3.2
+Wednesday, September 27, 2016
+====================================
+
+- Fix build of hb-coretext on older OS X versions.
+
+
+Overview of changes leading to 1.3.1
+Wednesday, September 7, 2016
+====================================
+
+- Blacklist bad GDEF of more fonts (Padauk).
+- More CoreText backend crash fixes with OS X 10.9.5.
+- Misc fixes.
+
+
+Overview of changes leading to 1.3.0
+Thursday, July 21, 2016
+====================================
+
+- Update to Unicode 9.0.0
+- Move Javanese from Indic shaper to Universal Shaping Engine.
+- Allow MultipleSubst to delete a glyph (matching Windows engine).
+- Update Universal Shaping Engine to latest draft from Microsoft.
+- DirectWrite backend improvements.  Note: this backend is for testing ONLY.
+- CoreText backend improvements with unreachable fonts.
+- Implement symbol fonts (cmap 3.0.0) in hb-ft and hb-ot-font.
+- Blacklist bad GDEF of more fonts (Tahoma & others).
+- Misc fixes.
+
+
+Overview of changes leading to 1.2.7
+Monday, May 2, 2016
+====================================
+
+- Blacklist another version of Times New Roman (Bold) Italic from Windows 7.
+- Fix Mongolian Free Variation Selectors shaping with certain fonts.
+- Fix Tibetan shorthand contractions shaping.
+- Improved list of language tag mappings.
+- Unbreak build on Windows CE.
+- Make 'glyf' table loading lazy in hb-ot-font.
+
+
+Overview of changes leading to 1.2.6
+Friday, April 8, 2016
+====================================
+
+- Blacklist GDEF table of another set of Times New Roman (Bold) Italic.
+- DirectWrite backend improvements.  Note: DirectWrite backend is
+  exclusively for our internal testing and should NOT be used in any
+  production system whatsoever.
+
+
+Overview of changes leading to 1.2.5
+Monday, April 4, 2016
+====================================
+
+- Fix GDEF mark-filtering-set, which was broken in 1.2.3.
+
+
+Overview of changes leading to 1.2.4
+Thursday, March 17, 2016
+====================================
+
+- Synthesize GDEF glyph class for any glyph that does not have one in GDEF.
+  I really hope we don't discover broken fonts that shape badly with this
+  change.
+- Misc build and other minor fixes.
+- API changes:
+  - Added HB_NDEBUG.  It's fine for production systems to define this to
+    disable high-overhead debugging checks.  However, I also reduced the
+    overhead of those checks, so it's a non-issue right now.  You can
+    forget it.  Just not defining anything at all is fine.
+
+
+Overview of changes leading to 1.2.3
+Thursday, February 25, 2016
+====================================
+
+- Blacklist GDEF table of certain versions of Times New Roman (Bold) Italic,
+  due to bug in glyph class of ASCII double-quote character.  This should
+  address "regression" introduced in 1.2.0 when we switched mark zeroing
+  in most shapers from BY_UNICODE_LATE to BY_GDEF_LATE.
+  This fourth release in a week should finally stablize things...
+
+- hb-ot-font's get_glyph() implementation saw some optimizations.  Though,
+  might be really hard to measure in real-world situations.
+
+- Also, two rather small API changes:
+
+We now disable some time-consuming internal bookkeeping if built with NDEBUG
+defined.  This is a first time that we use NDEBUG to disable debug code.  If
+there exist production systems that do NOT want to enable NDEBUG, please let
+me know and I'll add HB_NDEBUG.
+
+Added get_nominal_glyph() and get_variation_glyph() instead of get_glyph()
+
+New API:
+- hb_font_get_nominal_glyph_func_t
+- hb_font_get_variation_glyph_func_t
+- hb_font_funcs_set_nominal_glyph_func()
+- hb_font_funcs_set_variation_glyph_func()
+- hb_font_get_nominal_glyph()
+- hb_font_get_variation_glyph()
+
+Deprecated API:
+- hb_font_get_glyph_func_t
+- hb_font_funcs_set_glyph_func()
+
+Clients that implement their own font-funcs are encouraged to replace
+their get_glyph() implementation with a get_nominal_glyph() and
+get_variation_glyph() pair.  The variation version can assume that
+variation_selector argument is not zero.  Old (deprecated) functions
+will continue working indefinitely using internal gymnastics; it is
+just more efficient to use the new functions.
+
+
+Overview of changes leading to 1.2.2
+Wednesday, February 24, 2016
+====================================
+
+- Fix regression with mark positioning with fonts that have
+  non-zero mark advances.  This was introduced in 1.2.0 while
+  trying to make mark and cursive attachments to work together.
+  I have partially reverted that, so this version is much more
+  like what we had before.  All clients who updated to 1.2.0
+  should update to this version.
+
+
+Overview of changes leading to 1.2.1
+Tuesday, February 23, 2016
+====================================
+
+- CoreText: Fix bug with wrong scale if font scale was changed later.
+  https://github.com/libass/libass/issues/212
+- CoreText: Drastically speed up font initialization.
+- CoreText: Fix tiny leak.
+- Group ZWJ/ZWNJ with previous syllable under cluster-level=0.
+  https://github.com/harfbuzz/harfbuzz/issues/217
+- Add test/shaping/README.md about how to add tests to the suite.
+
+
+Overview of changes leading to 1.2.0
+Friday, February 19, 2016
+====================================
+
+- Fix various issues (hangs mostly) in case of memory allocation failure.
+- Change mark zeroing types of most shapers from BY_UNICODE_LATE to
+  BY_GDEF_LATE.  This seems to be what Uniscribe does.
+- Change mark zeroing of USE shaper from NONE to BY_GDEF_EARLY.  That's
+  what Windows does.
+- Allow GPOS cursive connection on marks, and fix the interaction with
+  mark attachment.  This work resulted in some changes to how mark
+  attachments work.  See:
+  https://github.com/harfbuzz/harfbuzz/issues/211
+  https://github.com/harfbuzz/harfbuzz/commit/86c68c7a2c971efe8e35b1f1bd99401dc8b688d2
+- Graphite2 shaper: improved negative advance handling (eg. Nastaliq).
+- Add nmake-based build system for Windows.
+- Minor speedup.
+- Misc. improvements.
+
+
+Overview of changes leading to 1.1.3
+Monday, January 11, 2016
+====================================
+
+- Ported Indic shaper to Unicode 8.0 data.
+- Universal Shaping Engine fixes.
+- Speed up CoreText shaper when font fallback happens in CoreText.
+- Documentation improvements, thanks to Khaled Hosny.
+- Very rough directwrite shaper for testing, thanks to Ebrahim Byagowi.
+- Misc bug fixes.
+- New API:
+
+  * Font extents:
+      hb_font_extents_t
+      hb_font_get_font_extents_func_t
+      hb_font_get_font_h_extents_func_t
+      hb_font_get_font_v_extents_func_t
+      hb_font_funcs_set_font_h_extents_func
+      hb_font_funcs_set_font_v_extents_func
+      hb_font_get_h_extents
+      hb_font_get_v_extents
+      hb_font_get_extents_for_direction
+
+  * Buffer message (aka debug):
+      hb_buffer_message_func_t
+      hb_buffer_set_message_func()
+    Actual message protocol to be fleshed out later.
+
+
+Overview of changes leading to 1.1.2
+Wednesday, November 26, 2015
+====================================
+
+- Fix badly-broken fallback shaper that affected terminology.
+  https://github.com/harfbuzz/harfbuzz/issues/187
+- Fix y_scaling in Graphite shaper.
+- API changes:
+  * An unset glyph_h_origin() function in font-funcs now (sensibly)
+    implies horizontal origin at 0,0.  Ie, the nil callback returns
+    true instead of false.  As such, implementations that have a
+    glyph_h_origin() that simply returns true, can remove that function
+    with HarfBuzz >= 1.1.2.  This results in a tiny speedup.
+
+
+Overview of changes leading to 1.1.1
+Wednesday, November 24, 2015
+====================================
+
+- Build fixes, specially for hb-coretext.
+
+
+Overview of changes leading to 1.1.0
+Wednesday, November 18, 2015
+====================================
+
+- Implement 'stch' stretch feature for Syriac Abbreviation Mark.
+  https://github.com/harfbuzz/harfbuzz/issues/141
+- Disable use of decompose_compatibility() callback.
+- Implement "shaping" of various Unicode space characters, even
+  if the font does not support them.
+  https://github.com/harfbuzz/harfbuzz/issues/153
+- If font does not support U+2011 NO-BREAK HYPHEN, fallback to
+  U+2010 HYPHEN.
+- Changes resulting from libFuzzer continuous fuzzing:
+  * Reject font tables that need more than 8 edits,
+  * Bound buffer growth during shaping to 32x,
+  * Fix assertions and other issues at OOM / buffer max-growth.
+- Misc fixes and optimizations.
+- API changes:
+  * All fonts created with hb_font_create() now inherit from
+    (ie. have parent) hb_font_get_empty().
+
+
+Overview of changes leading to 1.0.6
+Thursday, October 15, 2015
+====================================
+
+- Reduce max nesting level in OT lookups from 8 to 6.
+  Should not affect any real font as far as I know.
+- Fix memory access issue in ot-font.
+- Revert default load-flags of fonts created using hb_ft_font_create()
+  back to FT_LOAD_DEFAULT|FT_LOAD_NO_HINTING.  This was changed in
+  last release (1.0.5), but caused major issues, so revert.
+  https://github.com/harfbuzz/harfbuzz/issues/143
+
+
+Overview of changes leading to 1.0.5
+Tuesday, October 13, 2015
+====================================
+
+- Fix multiple memory access bugs discovered using libFuzzer.
+  https://github.com/harfbuzz/harfbuzz/issues/139
+  Everyone should upgrade to this version as soon as possible.
+  We now have continuous fuzzing set up, to avoid issues like
+  these creeping in again.
+- Misc fixes.
+
+- New API:
+  * hb_font_set_parent().
+  * hb_ft_font_[sg]et_load_flags()
+    The default flags for fonts created using hb_ft_font_create()
+    has changed to default to FT_LOAD_DEFAULT now.  Previously it
+    was defaulting to FT_LOAD_DFEAULT|FT_LOAD_NO_HINTING.
+
+- API changes:
+  * Fonts now default to units-per-EM as their scale, instead of 0.
+  * hb_font_create_sub_font() does NOT make parent font immutable
+    anymore.  hb_font_make_immutable() does.
+
+
+Overview of changes leading to 1.0.4
+Wednesday, September 30, 2015
+====================================
+
+- Fix minor out-of-bounds read error.
+
+
+Overview of changes leading to 1.0.3
+Tuesday, September 1, 2015
+====================================
+
+- Start of user documentation, from Simon Cozens!
+- Implement glyph_extents() for TrueType fonts in hb-ot-font.
+- Improve GPOS cursive attachments with conflicting lookups.
+- More fixes for cluster-level = 1.
+- Uniscribe positioning fix.
+
+
+Overview of changes leading to 1.0.2
+Wednesday, August 19, 2015
+====================================
+
+- Fix shaping with cluster-level > 0.
+- Fix Uniscribe backend font-size scaling.
+- Declare dependencies in harfbuzz.pc.
+  FreeType is not declared though, to avoid bugs in pkg-config
+  0.26 with recursive dependencies.
+- Slightly improved debug infrastructure.  More to come later.
+- Misc build fixes.
+
+
+Overview of changes leading to 1.0.1
+Monday, July 27, 2015
+====================================
+
+- Fix out-of-bounds access in USE shaper.
+
+
+Overview of changes leading to 1.0.0
+Sunday, July 26, 2015
+====================================
+
+- Implement Universal Shaping Engine:
+  https://www.microsoft.com/typography/OpenTypeDev/USE/intro.htm
+  http://blogs.windows.com/bloggingwindows/2015/02/23/windows-shapes-the-worlds-languages/
+- Bump version to 1.0.0.  The soname was NOT bumped.
+
+
+Overview of changes leading to 0.9.42
+Thursday, July 26, 2015
+=====================================
+
+- New API to allow for retrieving finer-grained cluster
+  mappings if the client desires to handle them.  Default
+  behavior is unchanged.
+- Fix cluster merging when removing default-ignorables.
+- Update to Unicode 8.0
+- hb-graphite2 fixes.
+- Misc fixes.
+- Removed HB_NO_MERGE_CLUSTERS hack.
+- New API:
+  hb_buffer_cluster_level_t enum
+  hb_buffer_get_cluster_level()
+  hb_buffer_set_cluster_level()
+  hb-shape / hb-view --cluster-level
+
+
+Overview of changes leading to 0.9.41
+Thursday, June 18, 2015
+=====================================
+
+- Fix hb-coretext with trailing whitespace in right-to-left.
+- New API: hb_buffer_reverse_range().
+- Allow implementing atomic ops in config.h.
+- Fix hb_language_t in language bindings.
+- Misc fixes.
+
+
+Overview of changes leading to 0.9.40
+Friday, March 20, 2015
+=====================================
+
+- Another hb-coretext crasher fix.  Ouch!
+- Happy Norouz!
+
+
+Overview of changes leading to 0.9.39
+Wednesday, March 4, 2015
+=====================================
+
+- Critical hb-coretext fixes.
+- Optimizations and refactoring; no functional change
+  expected.
+- Misc build fixes.
+
+
+Overview of changes leading to 0.9.38
+Friday, January 23, 2015
+=====================================
+
+- Fix minor out-of-bounds access in Indic shaper.
+- Change New Tai Lue shaping engine from South-East Asian to default,
+  reflecting change in Unicode encoding model.
+- Add hb-shape --font-size.  Can take up to two numbers for separate
+  x / y size.
+- Fix CoreText and FreeType scale issues with negative scales.
+- Reject blobs larger than 2GB.  This might break some icu-le-hb clients
+  that need security fixes.  See:
+  http://www.icu-project.org/trac/ticket/11450
+- Avoid accessing font tables during face destruction, in casce rogue
+  clients released face data already.
+- Fix up gobject-introspection a bit.  Python bindings kinda working.
+  See README.python.
+- Misc fixes.
+- API additions:
+  hb_ft_face_create_referenced()
+  hb_ft_font_create_referenced()
+
+
+Overview of changes leading to 0.9.37
+Wednesday, December 17, 2014
+=====================================
+
+- Fix out-of-bounds access in Context lookup format 3.
+- Indic: Allow ZWJ/ZWNJ before syllable modifiers.
+
+
+Overview of changes leading to 0.9.36
+Thursday, November 20, 2014
+=====================================
+
+- First time that three months went by without a release since
+  0.9.2 was released on August 10, 2012!
+- Fix performance bug in hb_ot_collect_glyphs():
+  https://bugzilla.mozilla.org/show_bug.cgi?id=1090869
+- Add basic vertical-text support to hb-ot-font.
+- Misc build fixes.
+
+
+Overview of changes leading to 0.9.35
+Saturday, August 13, 2014
+=====================================
+
+- Fix major shape-plan caching bug when more than one shaper were
+  provided to hb_shape_full() (as exercised by XeTeX).
+  http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1246370.html
+- Fix Arabic fallback shaping regression.  This was broken in 0.9.32.
+- Major hb-coretext fixes.  That backend is complete now, including
+  respecing buffer direction and language, down to vertical writing.
+- Build fixes for Windows CE.  Should build fine now.
+- Misc fixes:
+  Use atexit() only if it's safe to call from shared library
+  https://bugs.freedesktop.org/show_bug.cgi?id=82246
+  Mandaic had errors in its Unicode Joining_Type
+  https://bugs.freedesktop.org/show_bug.cgi?id=82306
+- API changes:
+
+  * hb_buffer_clear_contents() does not reset buffer flags now.
+
+    After 763e5466c0a03a7c27020e1e2598e488612529a7, one doesn't
+    need to set flags for different pieces of text.  The flags now
+    are something the client sets up once, depending on how it
+    actually uses the buffer.  As such, don't clear it in
+    clear_contents().
+
+    I don't expect any changes to be needed to any existing client.
+
+
+Overview of changes leading to 0.9.34
+Saturday, August 2, 2014
+=====================================
+
+- hb_feature_from_string() now accepts CSS font-feature-settings format.
+- As a result, hb-shape / hb-view --features also accept CSS-style strings.
+  Eg, "'liga' off" is accepted now.
+- Add old-spec Myanmar shaper:
+  https://bugs.freedesktop.org/show_bug.cgi?id=81775
+- Don't apply 'calt' in Hangul shaper.
+- Fix mark advance zeroing for Hebrew shaper:
+  https://bugs.freedesktop.org/show_bug.cgi?id=76767
+- Implement Windows-1256 custom Arabic shaping.  Only built on Windows,
+  and requires help from get_glyph().  Used by Firefox.
+  https://bugzilla.mozilla.org/show_bug.cgi?id=1045139
+- Disable 'liga' in vertical text.
+- Build fixes.
+- API changes:
+
+  * Make HB_BUFFER_FLAG_BOT/EOT easier to use.
+
+    Previously, we expected users to provide BOT/EOT flags when the
+    text *segment* was at paragraph boundaries.  This meant that for
+    clients that provide full paragraph to HarfBuzz (eg. Pango), they
+    had code like this:
+
+      hb_buffer_set_flags (hb_buffer,
+                           (item_offset == 0 ? HB_BUFFER_FLAG_BOT : 0) |
+                           (item_offset + item_length == paragraph_length ?
+                            HB_BUFFER_FLAG_EOT : 0));
+
+      hb_buffer_add_utf8 (hb_buffer,
+                          paragraph_text, paragraph_length,
+                          item_offset, item_length);
+
+    After this change such clients can simply say:
+
+      hb_buffer_set_flags (hb_buffer,
+                           HB_BUFFER_FLAG_BOT | HB_BUFFER_FLAG_EOT);
+
+      hb_buffer_add_utf8 (hb_buffer,
+                          paragraph_text, paragraph_length,
+                          item_offset, item_length);
+
+    Ie, HarfBuzz itself checks whether the segment is at the beginning/end
+    of the paragraph.  Clients that only pass item-at-a-time to HarfBuzz
+    continue not setting any flags whatsoever.
+
+    Another way to put it is: if there's pre-context text in the buffer,
+    HarfBuzz ignores the BOT flag.  If there's post-context, it ignores
+    EOT flag.
+
+
+Overview of changes leading to 0.9.33
+Tuesday, July 22, 2014
+=====================================
+
+- Turn off ARabic 'cswh' feature that was accidentally turned on.
+- Add HB_TAG_MAX_SIGNED.
+- Make hb_face_make_immutable() really make face immutable!
+- Windows build fixes.
+
+
+Overview of changes leading to 0.9.32
+Thursday, July 17, 2014
+=====================================
+
+- Apply Arabic shaping features in spec order exactly.
+- Another fix for Mongolian free variation selectors.
+- For non-Arabic scripts in Arabic shaper apply 'rlig' and 'calt'
+  together.
+- Minor adjustment to U+FFFD logic.
+- Fix hb-coretext build.
+
+
+Overview of changes leading to 0.9.31
+Wednesday, July 16, 2014
+=====================================
+
+- Only accept valid UTF-8/16/32; we missed many cases before.
+- Better shaping of invalid UTF-8/16/32.  Falls back to
+  U+FFFD REPLACEMENT CHARACTER now.
+- With all changes in this release, the buffer will contain fully
+  valid Unicode after hb_buffer_add_utf8/16/32 no matter how
+  broken the input is.  This can be overridden though.  See below.
+- Fix Mongolian Variation Selectors for fonts without GDEF.
+- Fix minor invalid buffer access.
+- Accept zh-Hant and zh-Hans language tags.  hb_ot_tag_to_language()
+  now uses these instead of private tags.
+- Build fixes.
+- New API:
+  * hb_buffer_add_codepoints().  This does what hb_buffer_add_utf32()
+    used to do, ie. no validity check on the input at all.  add_utf32
+    now replaces invalid Unicode codepoints with the replacement
+    character (see below).
+  * hb_buffer_set_replacement_codepoint()
+  * hb_buffer_get_replacement_codepoint()
+    Previously, in hb_buffer_add_utf8 and hb_buffer_add_utf16, when
+    we detected broken input, we replaced that with (hb_codepoint_t)-1.
+    This has changed to use U+FFFD now, but can be changed using these
+    new API.
+
+
+Overview of changes leading to 0.9.30
+Wednesday, July 9, 2014
+=====================================
+
+- Update to Unicode 7.0.0:
+  * New scripts Manichaean and Psalter Pahlavi are shaped using
+    Arabic shaper.
+  * All the other new scripts to through the generic shaper for
+    now.
+- Minor Indic improvements.
+- Fix graphite2 backend cluster mapping [crasher!]
+- API changes:
+  * New HB_SCRIPT_* values for Unicode 7.0 scripts.
+  * New function hb_ot_layout_language_get_required_feature().
+- Build fixes.
+
+
+Overview of changes leading to 0.9.29
+Thursday, May 29, 2014
+=====================================
+
+- Implement cmap in hb-ot-font.h.  No variation-selectors yet.
+- Myanmar: Allow MedialYa+Asat.
+- Various Indic fixes:
+  * Support most characters in Extended Devanagary and Vedic
+    Unicode blocks.
+  * Allow digits and a some punctuation as consonant placeholders.
+- Build fixes.
+
+
+Overview of changes leading to 0.9.28
+Monday, April 28, 2014
+=====================================
+
+- Unbreak old-spec Indic shaping. (bug 76705)
+- Fix shaping of U+17DD and U+0FC6.
+- Add HB_NO_MERGE_CLUSTERS build option.  NOT to be enabled by default
+  for shipping libraries.  It's an option for further experimentation
+  right now.  When we are sure how to do it properly, we will add
+  public run-time API for the functionality.
+- Build fixes.
+
+
+Overview of changes leading to 0.9.27
+Tuesday, March 18, 2014
+=====================================
+
+- Don't use "register" storage class specifier
+- Wrap definition of free_langs() with HAVE_ATEXIT
+- Add coretext_aat shaper and hb_coretext_face_create() constructor
+- If HAVE_ICU_BUILTIN is defined, use hb-icu Unicode callbacks
+- Add Myanmar test case from OpenType Myanmar spec
+- Only do fallback Hebrew composition if no GPOS 'mark' available
+- Allow bootstrapping without gtk-doc
+- Use AM_MISSING_PROG for ragel and git
+- Typo in ucdn's Makefile.am
+- Improve MemoryBarrier() implementation
+
+
+Overview of changes leading to 0.9.26
+Thursday, January 30, 2014
+=====================================
+
+- Misc fixes.
+- Fix application of 'rtlm' feature.
+- Automatically apply frac/numr/dnom around U+2044 FRACTION SLASH.
+- New header: hb-ot-shape.h
+- Uniscribe: fix scratch-buffer accounting.
+- Reorder Tai Tham SAKOT to after tone-marks.
+- Add Hangul shaper.
+- New files:
+  hb-ot-shape-complex-hangul.cc
+  hb-ot-shape-complex-hebrew.cc
+  hb-ot-shape-complex-tibetan.cc
+- Disable 'cswh' feature in Arabic shaper.
+- Coretext: better handle surrogate pairs.
+- Add HB_TAG_MAX and _HB_SCRIPT_MAX_VALUE.
+
+
+Overview of changes leading to 0.9.25
+Wednesday, December 4, 2013
+=====================================
+
+- Myanmar shaper improvements.
+- Avoid font fallback in CoreText backend.
+- Additional OpenType language tag mappiongs.
+- More aggressive shape-plan caching.
+- Build with / require automake 1.13.
+- Build with libtool 2.4.2.418 alpha to support ppc64le.
+
+
+Overview of changes leading to 0.9.24
+Tuesday, November 13, 2013
+=====================================
+
+- Misc compiler warning fixes with clang.
+- No functional changes.
+
+
+Overview of changes leading to 0.9.23
+Monday, October 28, 2013
+=====================================
+
+- "Udupi HarfBuzz Hackfest", Paris, October 14..18 2013.
+- Fix (Chain)Context recursion with non-monotone lookup positions.
+- Misc Indic bug fixes.
+- New Javanese / Buginese shaping, similar to Windows 8.1.
+
+
+Overview of changes leading to 0.9.22
+Thursday, October 3, 2013
+=====================================
+
+- Fix use-after-end-of-scope in hb_language_from_string().
+- Fix hiding of default_ignorables if font doesn't have space glyph.
+- Protect against out-of-range lookup indices.
+
+- API Changes:
+
+  * Added hb_ot_layout_table_get_lookup_count()
+
+
+Overview of changes leading to 0.9.21
+Monday, September 16, 2013
+=====================================
+
+- Rename gobject-introspection library name from harfbuzz to HarfBuzz.
+- Remove (long disabled) hb-old and hb-icu-le test shapers.
+- Misc gtk-doc and gobject-introspection annotations.
+- Misc fixes.
+- API changes:
+
+  * Add HB_SET_VALUE_INVALID
+
+Overview of changes leading to 0.9.20
+Thursday, August 29, 2013
+=====================================
+
+General:
+- Misc substitute_closure() fixes.
+- Build fixes.
+
+Documentation:
+- gtk-doc boilerplate integrated.  Docs are built now, but
+  contain no contents.  By next release hopefully we have
+  some content in.  Enable using --enable-gtk-doc.
+
+GObject and Introspection:
+- Added harfbuzz-gobject library (hb-gobject.h) that has type
+  bindings for all HarfBuzz objects and enums.  Enable using
+  --with-gobject.
+- Added gobject-introspection boilerplate.  Nothing useful
+  right now.  Work in progress.  Gets enabled automatically if
+  --with-gobject is used.  Override with --disable-introspection.
+
+OpenType shaper:
+- Apply 'mark' in Myanmar shaper.
+- Don't apply 'dlig' by default.
+
+Uniscribe shaper:
+- Support user features.
+- Fix loading of fonts that are also installed on the system.
+- Fix shaping of Arabic Presentation Forms.
+- Fix build with wide chars.
+
+CoreText shaper:
+- Support user features.
+
+Source changes:
+- hb_face_t code moved to hb-face.h / hb-face.cc.
+- Added hb-deprecated.h.
+
+API changes:
+- Added HB_DISABLE_DEPRECATED.
+- Deprecated HB_SCRIPT_CANADIAN_ABORIGINAL; replaced by
+  HB_SCRIPT_CANADIAN_SYLLABICS.
+- Deprecated HB_BUFFER_FLAGS_DEFAULT; replaced by
+  HB_BUFFER_FLAG_DEFAULT.
+- Deprecated HB_BUFFER_SERIALIZE_FLAGS_DEFAULT; replaced by
+  HB_BUFFER_SERIALIZE_FLAG_DEFAULT.
+
+
+Overview of changes leading to 0.9.19
+Tuesday, July 16, 2013
+=====================================
+
+- Build fixes.
+- Better handling of multiple variation selectors in a row.
+- Pass on variation selector to GSUB if not consumed by cmap.
+- Fix undefined memory access.
+- Add Javanese config to Indic shaper.
+- Misc bug fixes.
+
+Overview of changes leading to 0.9.18
+Tuesday, May 28, 2013
+=====================================
+
+New build system:
+
+- All unneeded code is all disabled by default,
+
+- Uniscribe and CoreText shapers can be enabled with their --with options,
+
+- icu_le and old shapers cannot be enabled for now,
+
+- glib, freetype, and cairo will be detected automatically.
+  They can be force on/off'ed with their --with options,
+
+- icu and graphite2 are default off, can be enabled with their --with
+  options,
+
+Moreover, ICU support is now build into a separate library:
+libharfbuzz-icu.so, and a new harfbuzz-icu.pc is shipped for it.
+Distros can enable ICU now without every application on earth
+getting linked to via libharfbuzz.so.
+
+For distros I recommend that they make sure they are building --with-glib
+--with-freetype --with-cairo, --with-icu, and optionally --with-graphite2;
+And package harfbuzz and harfbuzz-icu separately.
+
+
+Overview of changes leading to 0.9.17
+Monday, May 20, 2013
+=====================================
+
+- Build fixes.
+- Fix bug in hb_set_get_min().
+- Fix regression with Arabic mark positioning / width-zeroing.
+
+Overview of changes leading to 0.9.16
+Friday, April 19, 2013
+=====================================
+
+- Major speedup in OpenType lookup processing.  With the Amiri
+  Arabic font, this release is over 3x faster than previous
+  release.  All scripts / languages should see this speedup.
+
+- New --num-iterations option for hb-shape / hb-view; useful for
+  profiling.
+
+Overview of changes leading to 0.9.15
+Friday, April 05, 2013
+=====================================
+
+- Build fixes.
+- Fix crasher in graphite2 shaper.
+- Fix Arabic mark width zeroing regression.
+- Don't compose Hangul jamo into Unicode syllables.
+
+
+Overview of changes leading to 0.9.14
+Thursday, March 21, 2013
+=====================================
+
+- Build fixes.
+- Fix time-consuming sanitize with malicious fonts.
+- Implement hb_buffer_deserialize_glyphs() for both json and text.
+- Do not ignore Hangul filler characters.
+- Indic fixes:
+  * Fix Malayalam pre-base reordering interaction with post-forms.
+  * Further adjust ZWJ handling.  Should fix known regressions from
+    0.9.13.
+
+
+Overview of changes leading to 0.9.13
+Thursday, February 25, 2013
+=====================================
+
+- Build fixes.
+- Ngapi HarfBuzz Hackfest in London (February 2013):
+  * Fixed all known Indic bugs,
+  * New Win8-style Myanmar shaper,
+  * New South-East Asian shaper for Tai Tham, Cham, and New Tai Lue,
+  * Smartly ignore Default_Ignorable characters (joiners, etc) wheb
+    matching GSUB/GPOS lookups,
+  * Fix 'Phags-Pa U+A872 shaping,
+  * Fix partial disabling of default-on features,
+  * Allow disabling of TrueType kerning.
+- Fix possible crasher with broken fonts with overlapping tables.
+- Removed generated files from git again.  So, one needs ragel to
+  bootstrap from the git tree.
+
+API changes:
+- hb_shape() and related APIs now abort if buffer direction is
+  HB_DIRECTION_INVALID.  Previously, hb_shape() was calling
+  hb_buffer_guess_segment_properties() on the buffer before
+  shaping.  The heuristics in that function are fragile.  If the
+  user really wants the old behvaior, they can call that function
+  right before calling hb_shape() to get the old behavior.
+- hb_blob_create_sub_blob() always creates sub-blob with
+  HB_MEMORY_MODE_READONLY.  See comments for the reason.
+
+
+Overview of changes leading to 0.9.12
+Thursday, January 18, 2013
+=====================================
+
+- Build fixes for Sun compiler.
+- Minor bug fix.
+
+Overview of changes leading to 0.9.11
+Thursday, January 10, 2013
+=====================================
+
+- Build fixes.
+- Fix GPOS mark attachment with null Anchor offsets.
+- [Indic] Fix old-spec reordering of viramas if sequence ends in one.
+- Fix multi-threaded shaper data creation crash.
+- Add atomic ops for Solaris.
+
+API changes:
+- Rename hb_buffer_clear() to hb_buffer_clear_contents().
+
+
+Overview of changes leading to 0.9.10
+Thursday, January 3, 2013
+=====================================
+
+- [Indic] Fixed rendering of Malayalam dot-reph
+- Updated OT language tags.
+- Updated graphite2 backend.
+- Improved hb_ot_layout_get_size_params() logic.
+- Improve hb-shape/hb-view help output.
+- Fixed hb-set.h implementation to not crash.
+- Fixed various issues with hb_ot_layout_collect_lookups().
+- Various build fixes.
+
+New API:
+
+hb_graphite2_face_get_gr_face()
+hb_graphite2_font_get_gr_font()
+hb_coretext_face_get_cg_font()
+
+Modified API:
+
+hb_ot_layout_get_size_params()
+
+
+Overview of changes leading to 0.9.9
+Wednesday, December 5, 2012
+====================================
+
+- Fix build on Windows.
+- Minor improvements.
+
+
+Overview of changes leading to 0.9.8
+Tuesday, December 4, 2012
+====================================
+
+
+- Actually implement hb_shape_plan_get_shaper ().
+- Make UCDB data tables const.
+- Lots of internal refactoring in OTLayout tables.
+- Flesh out hb_ot_layout_lookup_collect_glyphs().
+
+New API:
+
+hb_ot_layout_collect_lookups()
+hb_ot_layout_get_size_params()
+
+
+Overview of changes leading to 0.9.7
+Sunday, November 21, 2012
+====================================
+
+
+HarfBuzz "All-You-Can-Eat-Sushi" (aka Vancouver) Hackfest and follow-on fixes.
+
+- Fix Arabic contextual joining using pre-context text.
+- Fix Sinhala "split matra" mess.
+- Fix Khmer shaping with broken fonts.
+- Implement Thai "PUA" shaping for old fonts.
+- Do NOT route Kharoshthi script through the Indic shaper.
+- Disable fallback positioning for Indic and Thai shapers.
+- Misc fixes.
+
+
+hb-shape / hb-view changes:
+
+- Add --text-before and --text-after
+- Add --bot / --eot / --preserve-default-ignorables
+- hb-shape --output-format=json
+
+
+New API:
+
+hb_buffer_clear()
+
+hb_buffer_flags_t
+
+HB_BUFFER_FLAGS_DEFAULT
+HB_BUFFER_FLAG_BOT
+HB_BUFFER_FLAG_EOT
+HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES
+
+hb_buffer_set_flags()
+hb_buffer_get_flags()
+
+HB_BUFFER_SERIALIZE_FLAGS
+hb_buffer_serialize_glyphs()
+hb_buffer_deserialize_glyphs()
+hb_buffer_serialize_list_formats()
+
+hb_set_add_range()
+hb_set_del_range()
+hb_set_get_population()
+hb_set_next_range()
+
+hb_face_[sg]et_glyph_count()
+
+hb_segment_properties_t
+HB_SEGMENT_PROPERTIES_DEFAULT
+hb_segment_properties_equal()
+hb_segment_properties_hash()
+
+hb_buffer_set_segment_properties()
+hb_buffer_get_segment_properties()
+
+hb_ot_layout_glyph_class_t
+hb_ot_layout_get_glyph_class()
+hb_ot_layout_get_glyphs_in_class()
+
+hb_shape_plan_t
+hb_shape_plan_create()
+hb_shape_plan_create_cached()
+hb_shape_plan_get_empty()
+hb_shape_plan_reference()
+hb_shape_plan_destroy()
+hb_shape_plan_set_user_data()
+hb_shape_plan_get_user_data()
+hb_shape_plan_execute()
+hb_shape_plan_get_shaper()
+
+hb_ot_shape_plan_collect_lookups()
+
+
+API changes:
+
+- Remove "mask" parameter from hb_buffer_add().
+- Rename hb_ot_layout_would_substitute_lookup() and hb_ot_layout_substitute_closure_lookup().
+- hb-set.h API const correction.
+- Renamed hb_set_min/max() to hb_set_get_min/max().
+- Rename hb_ot_layout_feature_get_lookup_indexes() to hb_ot_layout_feature_get_lookups().
+- Rename hb_buffer_guess_properties() to hb_buffer_guess_segment_properties().
+
+
+
+Overview of changes leading to 0.9.6
+Sunday, November 13, 2012
+====================================
+
+- Don't clear pre-context text if no new context is provided.
+- Fix ReverseChainingSubstLookup, which was totally borked.
+- Adjust output format of hb-shape a bit.
+- Include config.h.in in-tree.  Makes it easier for alternate build systems.
+- Fix hb_buffer_set_length(buffer, 0) invalid memory allocation.
+- Use ICU LayoutEngine's C API instead of C++.  Avoids much headache.
+- Drop glyphs for all of Unicode Default_Ignorable characters.
+- Misc build fixes.
+
+Arabic shaper:
+- Enable 'dlig' and 'mset' features in Arabic shaper.
+- Implement 'Phags-pa shaping, improve Mongolian.
+
+Indic shaper:
+- Decompose Sinhala split matras the way old HarfBuzz / Pango did.
+- Initial support for Consonant Medials.
+- Start adding new-style Myanmar shaping.
+- Make reph and 'pref' logic introspect the font.
+- Route Meetei-Mayek through the Indic shaper.
+- Don't apply 'liga' in Indic shaper.
+- Improve Malayalam pre-base reordering Ra interaction with Chillus.
+
+
+
+Overview of changes leading to 0.9.5
+Sunday, October 14, 2012
+====================================
+
+- Synthetic-GSUB Arabic fallback shaping.
+
+- Misc Indic improvements.
+
+- Add build system support for pthread.
+
+- Imported UCDN for in-tree Unicode callbacks implementation.
+
+- Context-aware Arabic joining.
+
+- Misc other fixes.
+
+- New API:
+
+  hb_feature_to/from-string()
+  hb_buffer_[sg]et_content_type()
+
+
+
+Overview of changes leading to 0.9.4
+Tuesday, Sep 03, 2012
+====================================
+
+- Indic improvements with old-spec Malayalam.
+
+- Better fallback glyph positioning, specially with Thai / Lao marks.
+
+- Implement dotted-circle insertion.
+
+- Better Arabic fallback shaping / ligation.
+
+- Added ICU LayoutEngine backend for testing.  Call it by the 'icu_le' name.
+
+- Misc fixes.
+
+
+
+Overview of changes leading to 0.9.3
+Friday, Aug 18, 2012
+====================================
+
+- Fixed fallback mark positioning for left-to-right text.
+
+- Improve mark positioning for the remaining combining classes.
+
+- Unbreak Thai and fallback Arabic shaping.
+
+- Port Arabic shaper to shape-plan caching.
+
+- Use new ICU normalizer functions.
+
+
+
+Overview of changes leading to 0.9.2
+Friday, Aug 10, 2012
+====================================
+
+- Over a thousand commits!  This is the first major release of HarfBuzz.
+
+- HarfBuzz is feature-complete now!  It should be in par, or better, than
+  both Pango's shapers and old HarfBuzz / Qt shapers.
+
+- New Indic shaper, supporting main Indic scripts, Sinhala, and Khmer.
+
+- Improved Arabic shaper, with fallback Arabic shaping, supporting Arabic,
+  Sinhala, N'ko, Mongolian, and Mandaic.
+
+- New Thai / Lao shaper.
+
+- Tibetan / Hangul support in the generic shaper.
+
+- Synthetic GDEF support for fonts without a GDEF table.
+
+- Fallback mark positioning for fonts without a GPOS table.
+
+- Unicode normalization shaping heuristic during glyph mapping.
+
+- New experimental Graphite2 backend.
+
+- New Uniscribe backend (primarily for testing).
+
+- New CoreText backend (primarily for testing).
+
+- Major optimization and speedup.
+
+- Test suites and testing infrastructure (work in progress).
+
+- Greatly improved hb-view cmdline tool.
+
+- hb-shape cmdline tool.
+
+- Unicode 6.1 support.
+
+Summary of API changes:
+
+o Changed API:
+
+  - Users are expected to only include main header files now (ie. hb.h,
+    hb-glib.h, hb-ft.h, ...)
+
+  - All struct tag names had their initial underscore removed.
+    Ie. "struct _hb_buffer_t" is "struct hb_buffer_t" now.
+
+  - All set_user_data() functions now take a "replace" boolean parameter.
+
+  - hb_buffer_create() takes zero arguments now.
+    Use hb_buffer_pre_allocate() to pre-allocate.
+
+  - hb_buffer_add_utf*() now accept -1 for length parameteres,
+    meaning "nul-terminated".
+
+  - hb_direction_t enum values changed.
+
+  - All *_from_string() APIs now take a length parameter to allow for
+    non-nul-terminated strings. A -1 length means "nul-terminated".
+
+  - Typedef for hb_language_t changed.
+
+  - hb_get_table_func_t renamed to hb_reference_table_func_t.
+
+  - hb_ot_layout_table_choose_script()
+
+  - Various renames in hb-unicode.h.
+
+o New API:
+
+  - hb_buffer_guess_properties()
+    Automatically called by hb_shape().
+
+  - hb_buffer_normalize_glyphs()
+
+  - hb_tag_from_string()
+
+  - hb-coretext.h
+
+  - hb-uniscribe.h
+
+  - hb_face_reference_blob()
+  - hb_face_[sg]et_index()
+  - hb_face_set_upem()
+
+  - hb_font_get_glyph_name_func_t
+    hb_font_get_glyph_from_name_func_t
+    hb_font_funcs_set_glyph_name_func()
+    hb_font_funcs_set_glyph_from_name_func()
+    hb_font_get_glyph_name()
+    hb_font_get_glyph_from_name()
+    hb_font_glyph_to_string()
+    hb_font_glyph_from_string()
+
+  - hb_font_set_funcs_data()
+
+  - hb_ft_font_set_funcs()
+  - hb_ft_font_get_face()
+
+  - hb-gobject.h (work in progress)
+
+  - hb_ot_shape_glyphs_closure()
+    hb_ot_layout_substitute_closure_lookup()
+
+  - hb-set.h
+
+  - hb_shape_full()
+
+  - hb_unicode_combining_class_t
+
+  - hb_unicode_compose_func_t
+    hb_unicode_decompose_func_t
+    hb_unicode_decompose_compatibility_func_t
+    hb_unicode_funcs_set_compose_func()
+    hb_unicode_funcs_set_decompose_func()
+    hb_unicode_funcs_set_decompose_compatibility_func()
+    hb_unicode_compose()
+    hb_unicode_decompose()
+    hb_unicode_decompose_compatibility()
+
+o Removed API:
+
+  - hb_ft_get_font_funcs()
+
+  - hb_ot_layout_substitute_start()
+    hb_ot_layout_substitute_lookup()
+    hb_ot_layout_substitute_finish()
+    hb_ot_layout_position_start()
+    hb_ot_layout_position_lookup()
+    hb_ot_layout_position_finish()
+
+
+
+Overview of changes leading to 0.6.0
+Friday, May 27, 2011
+====================================
+
+- Vertical text support in GPOS
+- Almost all API entries have unit tests now, under test/
+- All thread-safety issues are fixed
+
+Summary of API changes follows.
+
+
+* Simple Types API:
+
+  o New API:
+    HB_LANGUAGE_INVALID
+    hb_language_get_default()
+    hb_direction_to_string()
+    hb_direction_from_string()
+    hb_script_get_horizontal_direction()
+    HB_UNTAG()
+
+  o Renamed API:
+    hb_category_t renamed to hb_unicode_general_category_t
+
+  o Changed API:
+    hb_language_t is a typed pointers now
+
+  o Removed API:
+    HB_TAG_STR()
+
+
+* Use ISO 15924 tags for hb_script_t:
+
+  o New API:
+    hb_script_from_iso15924_tag()
+    hb_script_to_iso15924_tag()
+    hb_script_from_string()
+
+  o Changed API:
+    HB_SCRIPT_* enum members changed value.
+
+
+* Buffer API streamlined:
+
+  o New API:
+    hb_buffer_reset()
+    hb_buffer_set_length()
+    hb_buffer_allocation_successful()
+
+  o Renamed API:
+    hb_buffer_ensure() renamed to hb_buffer_pre_allocate()
+    hb_buffer_add_glyph() renamed to hb_buffer_add()
+
+  o Removed API:
+    hb_buffer_clear()
+    hb_buffer_clear_positions()
+
+  o Changed API:
+    hb_buffer_get_glyph_infos() takes an out length parameter now
+    hb_buffer_get_glyph_positions() takes an out length parameter now
+
+
+* Blob API streamlined:
+
+  o New API:
+    hb_blob_get_data()
+    hb_blob_get_data_writable()
+
+  o Renamed API:
+    hb_blob_create_empty() renamed to hb_blob_get_empty()
+
+  o Removed API:
+    hb_blob_lock()
+    hb_blob_unlock()
+    hb_blob_is_writable()
+    hb_blob_try_writable()
+
+  o Changed API:
+    hb_blob_create() takes user_data before destroy now
+
+
+* Unicode functions API:
+
+  o Unicode function vectors can subclass other unicode function vectors now.
+    Unimplemented callbacks in the subclass automatically chainup to the parent.
+
+  o All hb_unicode_funcs_t callbacks take a user_data now.  Their setters
+    take a user_data and its respective destroy callback.
+
+  o New API:
+    hb_unicode_funcs_get_empty()
+    hb_unicode_funcs_get_default()
+    hb_unicode_funcs_get_parent()
+
+  o Changed API:
+    hb_unicode_funcs_create() now takes a parent_funcs.
+
+  o Removed func getter functions:
+    hb_unicode_funcs_get_mirroring_func()
+    hb_unicode_funcs_get_general_category_func()
+    hb_unicode_funcs_get_script_func()
+    hb_unicode_funcs_get_combining_class_func()
+    hb_unicode_funcs_get_eastasian_width_func()
+
+
+* Face API:
+
+  o Renamed API:
+    hb_face_get_table() renamed to hb_face_reference_table()
+    hb_face_create_for_data() renamed to hb_face_create()
+
+  o Changed API:
+    hb_face_create_for_tables() takes user_data before destroy now
+    hb_face_reference_table() returns empty blob instead of NULL
+    hb_get_table_func_t accepts the face as first parameter now
+
+* Font API:
+
+  o Fonts can subclass other fonts now.  Unimplemented callbacks in the
+    subclass automatically chainup to the parent.  When chaining up,
+    scale is adjusted if the parent font has a different scale.
+
+  o All hb_font_funcs_t callbacks take a user_data now.  Their setters
+    take a user_data and its respective destroy callback.
+
+  o New API:
+    hb_font_get_parent()
+    hb_font_funcs_get_empty()
+    hb_font_create_sub_font()
+
+  o Removed API:
+    hb_font_funcs_copy()
+    hb_font_unset_funcs()
+
+  o Removed func getter functions:
+    hb_font_funcs_get_glyph_func()
+    hb_font_funcs_get_glyph_advance_func()
+    hb_font_funcs_get_glyph_extents_func()
+    hb_font_funcs_get_contour_point_func()
+    hb_font_funcs_get_kerning_func()
+
+  o Changed API:
+    hb_font_create() takes a face and references it now
+    hb_font_set_funcs() takes user_data before destroy now
+    hb_font_set_scale() accepts signed integers now
+    hb_font_get_contour_point_func_t now takes glyph first, then point_index
+    hb_font_get_glyph_func_t returns a success boolean now
+
+
+* Changed object model:
+
+  o All object types have a _get_empty() now:
+    hb_blob_get_empty()
+    hb_buffer_get_empty()
+    hb_face_get_empty()
+    hb_font_get_empty()
+    hb_font_funcs_get_empty()
+    hb_unicode_funcs_get_empty()
+
+  o Added _set_user_data() and _get_user_data() for all object types:
+    hb_blob_get_user_data()
+    hb_blob_set_user_data()
+    hb_buffer_get_user_data()
+    hb_buffer_set_user_data()
+    hb_face_get_user_data()
+    hb_face_set_user_data()
+    hb_font_funcs_get_user_data()
+    hb_font_funcs_set_user_data()
+    hb_font_get_user_data()
+    hb_font_set_user_data()
+    hb_unicode_funcs_get_user_data()
+    hb_unicode_funcs_set_user_data()
+
+  o Removed the _get_reference_count() from all object types:
+    hb_blob_get_reference_count()
+    hb_buffer_get_reference_count()
+    hb_face_get_reference_count()
+    hb_font_funcs_get_reference_count()
+    hb_font_get_reference_count()
+    hb_unicode_funcs_get_reference_count()
+
+  o Added _make_immutable() and _is_immutable() for all object types except for buffer:
+    hb_blob_make_immutable()
+    hb_blob_is_immutable()
+    hb_face_make_immutable()
+    hb_face_is_immutable()
+
+
+* Changed API for vertical text support
+
+  o The following callbacks where removed:
+    hb_font_get_glyph_advance_func_t
+    hb_font_get_kerning_func_t
+
+  o The following new callbacks added instead:
+    hb_font_get_glyph_h_advance_func_t
+    hb_font_get_glyph_v_advance_func_t
+    hb_font_get_glyph_h_origin_func_t
+    hb_font_get_glyph_v_origin_func_t
+    hb_font_get_glyph_h_kerning_func_t
+    hb_font_get_glyph_v_kerning_func_t
+
+  o The following API removed as such:
+    hb_font_funcs_set_glyph_advance_func()
+    hb_font_funcs_set_kerning_func()
+    hb_font_get_glyph_advance()
+    hb_font_get_kerning()
+
+  o New API added instead:
+    hb_font_funcs_set_glyph_h_advance_func()
+    hb_font_funcs_set_glyph_v_advance_func()
+    hb_font_funcs_set_glyph_h_origin_func()
+    hb_font_funcs_set_glyph_v_origin_func()
+    hb_font_funcs_set_glyph_h_kerning_func()
+    hb_font_funcs_set_glyph_v_kerning_func()
+    hb_font_get_glyph_h_advance()
+    hb_font_get_glyph_v_advance()
+    hb_font_get_glyph_h_origin()
+    hb_font_get_glyph_v_origin()
+    hb_font_get_glyph_h_kerning()
+    hb_font_get_glyph_v_kerning()
+
+  o The following higher-leve API added for convenience:
+    hb_font_get_glyph_advance_for_direction()
+    hb_font_get_glyph_origin_for_direction()
+    hb_font_add_glyph_origin_for_direction()
+    hb_font_subtract_glyph_origin_for_direction()
+    hb_font_get_glyph_kerning_for_direction()
+    hb_font_get_glyph_extents_for_origin()
+    hb_font_get_glyph_contour_point_for_origin()
+
+
+* OpenType Layout API:
+
+  o New API:
+    hb_ot_layout_position_start()
+    hb_ot_layout_substitute_start()
+    hb_ot_layout_substitute_finish()
+
+
+* Glue code:
+
+  o New API:
+    hb_glib_script_to_script()
+    hb_glib_script_from_script()
+    hb_icu_script_to_script()
+    hb_icu_script_from_script()
+
+
+* Version API added:
+
+  o New API:
+    HB_VERSION_MAJOR
+    HB_VERSION_MINOR
+    HB_VERSION_MICRO
+    HB_VERSION_STRING
+    HB_VERSION_CHECK()
+    hb_version()
+    hb_version_string()
+    hb_version_check()
+
+
diff --git a/thirdparty/harfbuzz/THANKS b/thirdparty/harfbuzz/THANKS
new file mode 100644
index 0000000000..88cb7e9ea1
--- /dev/null
+++ b/thirdparty/harfbuzz/THANKS
@@ -0,0 +1,7 @@
+Bradley Grainger
+Kenichi Ishibashi
+Ivan Kuckir <https://photopea.com/>
+Ryan Lortie
+Jeff Muizelaar
+suzuki toshiya
+Philip Withnall
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-ankr-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-ankr-table.hh
new file mode 100644
index 0000000000..f2785a6f58
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-ankr-table.hh
@@ -0,0 +1,98 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LAYOUT_ANKR_TABLE_HH
+#define HB_AAT_LAYOUT_ANKR_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+
+/*
+ * ankr -- Anchor Point
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6ankr.html
+ */
+#define HB_AAT_TAG_ankr HB_TAG('a','n','k','r')
+
+
+namespace AAT {
+
+using namespace OT;
+
+
+struct Anchor
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  FWORD		xCoordinate;
+  FWORD		yCoordinate;
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+typedef LArrayOf<Anchor> GlyphAnchors;
+
+struct ankr
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_ankr;
+
+  const Anchor &get_anchor (hb_codepoint_t glyph_id,
+			    unsigned int i,
+			    unsigned int num_glyphs) const
+  {
+    const NNOffsetTo<GlyphAnchors> *offset = (this+lookupTable).get_value (glyph_id, num_glyphs);
+    if (!offset)
+      return Null (Anchor);
+    const GlyphAnchors &anchors = &(this+anchorData) + *offset;
+    return anchors[i];
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version == 0 &&
+			  c->check_range (this, anchorData) &&
+			  lookupTable.sanitize (c, this, &(this+anchorData))));
+  }
+
+  protected:
+  HBUINT16	version;	/* Version number (set to zero) */
+  HBUINT16	flags;		/* Flags (currently unused; set to zero) */
+  LOffsetTo<Lookup<NNOffsetTo<GlyphAnchors>>>
+		lookupTable;	/* Offset to the table's lookup table */
+  LNNOffsetTo<HBUINT8>
+		anchorData;	/* Offset to the glyph data table */
+
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_ANKR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-bsln-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-bsln-table.hh
new file mode 100644
index 0000000000..cd36fc8953
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-bsln-table.hh
@@ -0,0 +1,158 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LAYOUT_BSLN_TABLE_HH
+#define HB_AAT_LAYOUT_BSLN_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+
+/*
+ * bsln -- Baseline
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6bsln.html
+ */
+#define HB_AAT_TAG_bsln HB_TAG('b','s','l','n')
+
+
+namespace AAT {
+
+
+struct BaselineTableFormat0Part
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  // Roman, Ideographic centered, Ideographic low, Hanging and Math
+  // are the default defined ones, but any other maybe accessed also.
+  HBINT16	deltas[32];	/* These are the FUnit distance deltas from
+				 * the font's natural baseline to the other
+				 * baselines used in the font. */
+  public:
+  DEFINE_SIZE_STATIC (64);
+};
+
+struct BaselineTableFormat1Part
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  lookupTable.sanitize (c)));
+  }
+
+  protected:
+  HBINT16	deltas[32];	/* ditto */
+  Lookup<HBUINT16>
+		lookupTable;	/* Lookup table that maps glyphs to their
+				 * baseline values. */
+  public:
+  DEFINE_SIZE_MIN (66);
+};
+
+struct BaselineTableFormat2Part
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBGlyphID	stdGlyph;	/* The specific glyph index number in this
+				 * font that is used to set the baseline values.
+				 * This is the standard glyph.
+				 * This glyph must contain a set of control points
+				 * (whose numbers are contained in the ctlPoints field)
+				 * that are used to determine baseline distances. */
+  HBUINT16	ctlPoints[32];	/* Set of control point numbers,
+				 * associated with the standard glyph.
+				 * A value of 0xFFFF means there is no corresponding
+				 * control point in the standard glyph. */
+  public:
+  DEFINE_SIZE_STATIC (66);
+};
+
+struct BaselineTableFormat3Part
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) && lookupTable.sanitize (c)));
+  }
+
+  protected:
+  HBGlyphID	stdGlyph;	/* ditto */
+  HBUINT16	ctlPoints[32];	/* ditto */
+  Lookup<HBUINT16>
+		lookupTable;	/* Lookup table that maps glyphs to their
+				 * baseline values. */
+  public:
+  DEFINE_SIZE_MIN (68);
+};
+
+struct bsln
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_bsln;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!(c->check_struct (this) && defaultBaseline < 32)))
+      return_trace (false);
+
+    switch (format)
+    {
+    case 0: return_trace (parts.format0.sanitize (c));
+    case 1: return_trace (parts.format1.sanitize (c));
+    case 2: return_trace (parts.format2.sanitize (c));
+    case 3: return_trace (parts.format3.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version number of the Baseline table. */
+  HBUINT16	format;		/* Format of the baseline table. Only one baseline
+				 * format may be selected for the font. */
+  HBUINT16	defaultBaseline;/* Default baseline value for all glyphs.
+				 * This value can be from 0 through 31. */
+  union {
+  // Distance-Based Formats
+  BaselineTableFormat0Part	format0;
+  BaselineTableFormat1Part	format1;
+  // Control Point-based Formats
+  BaselineTableFormat2Part	format2;
+  BaselineTableFormat3Part	format3;
+  } parts;
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_BSLN_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-common.hh b/thirdparty/harfbuzz/src/hb-aat-layout-common.hh
new file mode 100644
index 0000000000..75d523f5fc
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-common.hh
@@ -0,0 +1,840 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_LAYOUT_COMMON_HH
+#define HB_AAT_LAYOUT_COMMON_HH
+
+#include "hb-aat-layout.hh"
+#include "hb-open-type.hh"
+
+
+namespace AAT {
+
+using namespace OT;
+
+
+/*
+ * Lookup Table
+ */
+
+template <typename T> struct Lookup;
+
+template <typename T>
+struct LookupFormat0
+{
+  friend struct Lookup<T>;
+
+  private:
+  const T* get_value (hb_codepoint_t glyph_id, unsigned int num_glyphs) const
+  {
+    if (unlikely (glyph_id >= num_glyphs)) return nullptr;
+    return &arrayZ[glyph_id];
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (arrayZ.sanitize (c, c->get_num_glyphs ()));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (arrayZ.sanitize (c, c->get_num_glyphs (), base));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 0 */
+  UnsizedArrayOf<T>
+		arrayZ;		/* Array of lookup values, indexed by glyph index. */
+  public:
+  DEFINE_SIZE_UNBOUNDED (2);
+};
+
+
+template <typename T>
+struct LookupSegmentSingle
+{
+  static constexpr unsigned TerminationWordCount = 2u;
+
+  int cmp (hb_codepoint_t g) const
+  { return g < first ? -1 : g <= last ? 0 : +1 ; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && value.sanitize (c));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && value.sanitize (c, base));
+  }
+
+  HBGlyphID	last;		/* Last GlyphID in this segment */
+  HBGlyphID	first;		/* First GlyphID in this segment */
+  T		value;		/* The lookup value (only one) */
+  public:
+  DEFINE_SIZE_STATIC (4 + T::static_size);
+};
+
+template <typename T>
+struct LookupFormat2
+{
+  friend struct Lookup<T>;
+
+  private:
+  const T* get_value (hb_codepoint_t glyph_id) const
+  {
+    const LookupSegmentSingle<T> *v = segments.bsearch (glyph_id);
+    return v ? &v->value : nullptr;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (segments.sanitize (c));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (segments.sanitize (c, base));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 2 */
+  VarSizedBinSearchArrayOf<LookupSegmentSingle<T>>
+		segments;	/* The actual segments. These must already be sorted,
+				 * according to the first word in each one (the last
+				 * glyph in each segment). */
+  public:
+  DEFINE_SIZE_ARRAY (8, segments);
+};
+
+template <typename T>
+struct LookupSegmentArray
+{
+  static constexpr unsigned TerminationWordCount = 2u;
+
+  const T* get_value (hb_codepoint_t glyph_id, const void *base) const
+  {
+    return first <= glyph_id && glyph_id <= last ? &(base+valuesZ)[glyph_id - first] : nullptr;
+  }
+
+  int cmp (hb_codepoint_t g) const
+  { return g < first ? -1 : g <= last ? 0 : +1; }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  first <= last &&
+		  valuesZ.sanitize (c, base, last - first + 1));
+  }
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, const void *base, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  first <= last &&
+		  valuesZ.sanitize (c, base, last - first + 1, hb_forward<Ts> (ds)...));
+  }
+
+  HBGlyphID	last;		/* Last GlyphID in this segment */
+  HBGlyphID	first;		/* First GlyphID in this segment */
+  NNOffsetTo<UnsizedArrayOf<T>>
+		valuesZ;	/* A 16-bit offset from the start of
+				 * the table to the data. */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+template <typename T>
+struct LookupFormat4
+{
+  friend struct Lookup<T>;
+
+  private:
+  const T* get_value (hb_codepoint_t glyph_id) const
+  {
+    const LookupSegmentArray<T> *v = segments.bsearch (glyph_id);
+    return v ? v->get_value (glyph_id, this) : nullptr;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (segments.sanitize (c, this));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (segments.sanitize (c, this, base));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 4 */
+  VarSizedBinSearchArrayOf<LookupSegmentArray<T>>
+		segments;	/* The actual segments. These must already be sorted,
+				 * according to the first word in each one (the last
+				 * glyph in each segment). */
+  public:
+  DEFINE_SIZE_ARRAY (8, segments);
+};
+
+template <typename T>
+struct LookupSingle
+{
+  static constexpr unsigned TerminationWordCount = 1u;
+
+  int cmp (hb_codepoint_t g) const { return glyph.cmp (g); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && value.sanitize (c));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && value.sanitize (c, base));
+  }
+
+  HBGlyphID	glyph;		/* Last GlyphID */
+  T		value;		/* The lookup value (only one) */
+  public:
+  DEFINE_SIZE_STATIC (2 + T::static_size);
+};
+
+template <typename T>
+struct LookupFormat6
+{
+  friend struct Lookup<T>;
+
+  private:
+  const T* get_value (hb_codepoint_t glyph_id) const
+  {
+    const LookupSingle<T> *v = entries.bsearch (glyph_id);
+    return v ? &v->value : nullptr;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (entries.sanitize (c));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (entries.sanitize (c, base));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 6 */
+  VarSizedBinSearchArrayOf<LookupSingle<T>>
+		entries;	/* The actual entries, sorted by glyph index. */
+  public:
+  DEFINE_SIZE_ARRAY (8, entries);
+};
+
+template <typename T>
+struct LookupFormat8
+{
+  friend struct Lookup<T>;
+
+  private:
+  const T* get_value (hb_codepoint_t glyph_id) const
+  {
+    return firstGlyph <= glyph_id && glyph_id - firstGlyph < glyphCount ?
+	   &valueArrayZ[glyph_id - firstGlyph] : nullptr;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && valueArrayZ.sanitize (c, glyphCount));
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && valueArrayZ.sanitize (c, glyphCount, base));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 8 */
+  HBGlyphID	firstGlyph;	/* First glyph index included in the trimmed array. */
+  HBUINT16	glyphCount;	/* Total number of glyphs (equivalent to the last
+				 * glyph minus the value of firstGlyph plus 1). */
+  UnsizedArrayOf<T>
+		valueArrayZ;	/* The lookup values (indexed by the glyph index
+				 * minus the value of firstGlyph). */
+  public:
+  DEFINE_SIZE_ARRAY (6, valueArrayZ);
+};
+
+template <typename T>
+struct LookupFormat10
+{
+  friend struct Lookup<T>;
+
+  private:
+  const typename T::type get_value_or_null (hb_codepoint_t glyph_id) const
+  {
+    if (!(firstGlyph <= glyph_id && glyph_id - firstGlyph < glyphCount))
+      return Null (T);
+
+    const HBUINT8 *p = &valueArrayZ[(glyph_id - firstGlyph) * valueSize];
+
+    unsigned int v = 0;
+    unsigned int count = valueSize;
+    for (unsigned int i = 0; i < count; i++)
+      v = (v << 8) | *p++;
+
+    return v;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  valueSize <= 4 &&
+		  valueArrayZ.sanitize (c, glyphCount * valueSize));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 8 */
+  HBUINT16	valueSize;	/* Byte size of each value. */
+  HBGlyphID	firstGlyph;	/* First glyph index included in the trimmed array. */
+  HBUINT16	glyphCount;	/* Total number of glyphs (equivalent to the last
+				 * glyph minus the value of firstGlyph plus 1). */
+  UnsizedArrayOf<HBUINT8>
+		valueArrayZ;	/* The lookup values (indexed by the glyph index
+				 * minus the value of firstGlyph). */
+  public:
+  DEFINE_SIZE_ARRAY (8, valueArrayZ);
+};
+
+template <typename T>
+struct Lookup
+{
+  const T* get_value (hb_codepoint_t glyph_id, unsigned int num_glyphs) const
+  {
+    switch (u.format) {
+    case 0: return u.format0.get_value (glyph_id, num_glyphs);
+    case 2: return u.format2.get_value (glyph_id);
+    case 4: return u.format4.get_value (glyph_id);
+    case 6: return u.format6.get_value (glyph_id);
+    case 8: return u.format8.get_value (glyph_id);
+    default:return nullptr;
+    }
+  }
+
+  const typename T::type get_value_or_null (hb_codepoint_t glyph_id, unsigned int num_glyphs) const
+  {
+    switch (u.format) {
+      /* Format 10 cannot return a pointer. */
+      case 10: return u.format10.get_value_or_null (glyph_id);
+      default:
+      const T *v = get_value (glyph_id, num_glyphs);
+      return v ? *v : Null (T);
+    }
+  }
+
+  typename T::type get_class (hb_codepoint_t glyph_id,
+			      unsigned int num_glyphs,
+			      unsigned int outOfRange) const
+  {
+    const T *v = get_value (glyph_id, num_glyphs);
+    return v ? *v : outOfRange;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 0: return_trace (u.format0.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    case 4: return_trace (u.format4.sanitize (c));
+    case 6: return_trace (u.format6.sanitize (c));
+    case 8: return_trace (u.format8.sanitize (c));
+    case 10: return_trace (u.format10.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 0: return_trace (u.format0.sanitize (c, base));
+    case 2: return_trace (u.format2.sanitize (c, base));
+    case 4: return_trace (u.format4.sanitize (c, base));
+    case 6: return_trace (u.format6.sanitize (c, base));
+    case 8: return_trace (u.format8.sanitize (c, base));
+    case 10: return_trace (false); /* We don't support format10 here currently. */
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  LookupFormat0<T>	format0;
+  LookupFormat2<T>	format2;
+  LookupFormat4<T>	format4;
+  LookupFormat6<T>	format6;
+  LookupFormat8<T>	format8;
+  LookupFormat10<T>	format10;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+/* Lookup 0 has unbounded size (dependant on num_glyphs).  So we need to defined
+ * special NULL objects for Lookup<> objects, but since it's template our macros
+ * don't work.  So we have to hand-code them here.  UGLY. */
+} /* Close namespace. */
+/* Ugly hand-coded null objects for template Lookup<> :(. */
+extern HB_INTERNAL const unsigned char _hb_Null_AAT_Lookup[2];
+template <typename T>
+struct Null<AAT::Lookup<T>> {
+  static AAT::Lookup<T> const & get_null ()
+  { return *reinterpret_cast<const AAT::Lookup<T> *> (_hb_Null_AAT_Lookup); }
+};
+namespace AAT {
+
+enum { DELETED_GLYPH = 0xFFFF };
+
+/*
+ * (Extended) State Table
+ */
+
+template <typename T>
+struct Entry
+{
+  bool sanitize (hb_sanitize_context_t *c, unsigned int count) const
+  {
+    TRACE_SANITIZE (this);
+    /* Note, we don't recurse-sanitize data because we don't access it.
+     * That said, in our DEFINE_SIZE_STATIC we access T::static_size,
+     * which ensures that data has a simple sanitize(). To be determined
+     * if I need to remove that as well.
+     *
+     * HOWEVER! Because we are a template, our DEFINE_SIZE_STATIC
+     * assertion wouldn't be checked, hence the line below. */
+    static_assert (T::static_size, "");
+
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	newState;	/* Byte offset from beginning of state table
+				 * to the new state. Really?!?! Or just state
+				 * number?  The latter in morx for sure. */
+  HBUINT16	flags;		/* Table specific. */
+  T		data;		/* Optional offsets to per-glyph tables. */
+  public:
+  DEFINE_SIZE_STATIC (4 + T::static_size);
+};
+
+template <>
+struct Entry<void>
+{
+  bool sanitize (hb_sanitize_context_t *c, unsigned int count /*XXX Unused?*/) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	newState;	/* Byte offset from beginning of state table to the new state. */
+  HBUINT16	flags;		/* Table specific. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+template <typename Types, typename Extra>
+struct StateTable
+{
+  typedef typename Types::HBUINT HBUINT;
+  typedef typename Types::HBUSHORT HBUSHORT;
+  typedef typename Types::ClassTypeNarrow ClassType;
+
+  enum State
+  {
+    STATE_START_OF_TEXT = 0,
+    STATE_START_OF_LINE = 1,
+  };
+  enum Class
+  {
+    CLASS_END_OF_TEXT = 0,
+    CLASS_OUT_OF_BOUNDS = 1,
+    CLASS_DELETED_GLYPH = 2,
+    CLASS_END_OF_LINE = 3,
+  };
+
+  int new_state (unsigned int newState) const
+  { return Types::extended ? newState : ((int) newState - (int) stateArrayTable) / (int) nClasses; }
+
+  unsigned int get_class (hb_codepoint_t glyph_id, unsigned int num_glyphs) const
+  {
+    if (unlikely (glyph_id == DELETED_GLYPH)) return CLASS_DELETED_GLYPH;
+    return (this+classTable).get_class (glyph_id, num_glyphs, 1);
+  }
+
+  const Entry<Extra> *get_entries () const
+  { return (this+entryTable).arrayZ; }
+
+  const Entry<Extra> &get_entry (int state, unsigned int klass) const
+  {
+    if (unlikely (klass >= nClasses))
+      klass = StateTable<Types, Entry<Extra>>::CLASS_OUT_OF_BOUNDS;
+
+    const HBUSHORT *states = (this+stateArrayTable).arrayZ;
+    const Entry<Extra> *entries = (this+entryTable).arrayZ;
+
+    unsigned int entry = states[state * nClasses + klass];
+    DEBUG_MSG (APPLY, nullptr, "e%u", entry);
+
+    return entries[entry];
+  }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 unsigned int *num_entries_out = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!(c->check_struct (this) &&
+		    nClasses >= 4 /* Ensure pre-defined classes fit.  */ &&
+		    classTable.sanitize (c, this)))) return_trace (false);
+
+    const HBUSHORT *states = (this+stateArrayTable).arrayZ;
+    const Entry<Extra> *entries = (this+entryTable).arrayZ;
+
+    unsigned int num_classes = nClasses;
+    if (unlikely (hb_unsigned_mul_overflows (num_classes, states[0].static_size)))
+      return_trace (false);
+    unsigned int row_stride = num_classes * states[0].static_size;
+
+    /* Apple 'kern' table has this peculiarity:
+     *
+     * "Because the stateTableOffset in the state table header is (strictly
+     * speaking) redundant, some 'kern' tables use it to record an initial
+     * state where that should not be StartOfText. To determine if this is
+     * done, calculate what the stateTableOffset should be. If it's different
+     * from the actual stateTableOffset, use it as the initial state."
+     *
+     * We implement this by calling the initial state zero, but allow *negative*
+     * states if the start state indeed was not the first state.  Since the code
+     * is shared, this will also apply to 'mort' table.  The 'kerx' / 'morx'
+     * tables are not affected since those address states by index, not offset.
+     */
+
+    int min_state = 0;
+    int max_state = 0;
+    unsigned int num_entries = 0;
+
+    int state_pos = 0;
+    int state_neg = 0;
+    unsigned int entry = 0;
+    while (min_state < state_neg || state_pos <= max_state)
+    {
+      if (min_state < state_neg)
+      {
+	/* Negative states. */
+	if (unlikely (hb_unsigned_mul_overflows (min_state, num_classes)))
+	  return_trace (false);
+	if (unlikely (!c->check_range (&states[min_state * num_classes],
+				       -min_state,
+				       row_stride)))
+	  return_trace (false);
+	if ((c->max_ops -= state_neg - min_state) <= 0)
+	  return_trace (false);
+	{ /* Sweep new states. */
+	  const HBUSHORT *stop = &states[min_state * num_classes];
+	  if (unlikely (stop > states))
+	    return_trace (false);
+	  for (const HBUSHORT *p = states; stop < p; p--)
+	    num_entries = hb_max (num_entries, *(p - 1) + 1);
+	  state_neg = min_state;
+	}
+      }
+
+      if (state_pos <= max_state)
+      {
+	/* Positive states. */
+	if (unlikely (!c->check_range (states,
+				       max_state + 1,
+				       row_stride)))
+	  return_trace (false);
+	if ((c->max_ops -= max_state - state_pos + 1) <= 0)
+	  return_trace (false);
+	{ /* Sweep new states. */
+	  if (unlikely (hb_unsigned_mul_overflows ((max_state + 1), num_classes)))
+	    return_trace (false);
+	  const HBUSHORT *stop = &states[(max_state + 1) * num_classes];
+	  if (unlikely (stop < states))
+	    return_trace (false);
+	  for (const HBUSHORT *p = &states[state_pos * num_classes]; p < stop; p++)
+	    num_entries = hb_max (num_entries, *p + 1);
+	  state_pos = max_state + 1;
+	}
+      }
+
+      if (unlikely (!c->check_array (entries, num_entries)))
+	return_trace (false);
+      if ((c->max_ops -= num_entries - entry) <= 0)
+	return_trace (false);
+      { /* Sweep new entries. */
+	const Entry<Extra> *stop = &entries[num_entries];
+	for (const Entry<Extra> *p = &entries[entry]; p < stop; p++)
+	{
+	  int newState = new_state (p->newState);
+	  min_state = hb_min (min_state, newState);
+	  max_state = hb_max (max_state, newState);
+	}
+	entry = num_entries;
+      }
+    }
+
+    if (num_entries_out)
+      *num_entries_out = num_entries;
+
+    return_trace (true);
+  }
+
+  protected:
+  HBUINT	nClasses;	/* Number of classes, which is the number of indices
+				 * in a single line in the state array. */
+  NNOffsetTo<ClassType, HBUINT>
+		classTable;	/* Offset to the class table. */
+  NNOffsetTo<UnsizedArrayOf<HBUSHORT>, HBUINT>
+		stateArrayTable;/* Offset to the state array. */
+  NNOffsetTo<UnsizedArrayOf<Entry<Extra>>, HBUINT>
+		entryTable;	/* Offset to the entry array. */
+
+  public:
+  DEFINE_SIZE_STATIC (4 * sizeof (HBUINT));
+};
+
+template <typename HBUCHAR>
+struct ClassTable
+{
+  unsigned int get_class (hb_codepoint_t glyph_id, unsigned int outOfRange) const
+  {
+    unsigned int i = glyph_id - firstGlyph;
+    return i >= classArray.len ? outOfRange : classArray.arrayZ[i];
+  }
+  unsigned int get_class (hb_codepoint_t glyph_id,
+			  unsigned int num_glyphs HB_UNUSED,
+			  unsigned int outOfRange) const
+  {
+    return get_class (glyph_id, outOfRange);
+  }
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && classArray.sanitize (c));
+  }
+  protected:
+  HBGlyphID		firstGlyph;	/* First glyph index included in the trimmed array. */
+  ArrayOf<HBUCHAR>	classArray;	/* The class codes (indexed by glyph index minus
+					 * firstGlyph). */
+  public:
+  DEFINE_SIZE_ARRAY (4, classArray);
+};
+
+struct ObsoleteTypes
+{
+  static constexpr bool extended = false;
+  typedef HBUINT16 HBUINT;
+  typedef HBUINT8 HBUSHORT;
+  typedef ClassTable<HBUINT8> ClassTypeNarrow;
+  typedef ClassTable<HBUINT16> ClassTypeWide;
+
+  template <typename T>
+  static unsigned int offsetToIndex (unsigned int offset,
+				     const void *base,
+				     const T *array)
+  {
+    return (offset - ((const char *) array - (const char *) base)) / T::static_size;
+  }
+  template <typename T>
+  static unsigned int byteOffsetToIndex (unsigned int offset,
+					 const void *base,
+					 const T *array)
+  {
+    return offsetToIndex (offset, base, array);
+  }
+  template <typename T>
+  static unsigned int wordOffsetToIndex (unsigned int offset,
+					 const void *base,
+					 const T *array)
+  {
+    return offsetToIndex (2 * offset, base, array);
+  }
+};
+struct ExtendedTypes
+{
+  static constexpr bool extended = true;
+  typedef HBUINT32 HBUINT;
+  typedef HBUINT16 HBUSHORT;
+  typedef Lookup<HBUINT16> ClassTypeNarrow;
+  typedef Lookup<HBUINT16> ClassTypeWide;
+
+  template <typename T>
+  static unsigned int offsetToIndex (unsigned int offset,
+				     const void *base HB_UNUSED,
+				     const T *array HB_UNUSED)
+  {
+    return offset;
+  }
+  template <typename T>
+  static unsigned int byteOffsetToIndex (unsigned int offset,
+					 const void *base HB_UNUSED,
+					 const T *array HB_UNUSED)
+  {
+    return offset / 2;
+  }
+  template <typename T>
+  static unsigned int wordOffsetToIndex (unsigned int offset,
+					 const void *base HB_UNUSED,
+					 const T *array HB_UNUSED)
+  {
+    return offset;
+  }
+};
+
+template <typename Types, typename EntryData>
+struct StateTableDriver
+{
+  StateTableDriver (const StateTable<Types, EntryData> &machine_,
+		    hb_buffer_t *buffer_,
+		    hb_face_t *face_) :
+	      machine (machine_),
+	      buffer (buffer_),
+	      num_glyphs (face_->get_num_glyphs ()) {}
+
+  template <typename context_t>
+  void drive (context_t *c)
+  {
+    if (!c->in_place)
+      buffer->clear_output ();
+
+    int state = StateTable<Types, EntryData>::STATE_START_OF_TEXT;
+    for (buffer->idx = 0; buffer->successful;)
+    {
+      unsigned int klass = buffer->idx < buffer->len ?
+			   machine.get_class (buffer->info[buffer->idx].codepoint, num_glyphs) :
+			   (unsigned) StateTable<Types, EntryData>::CLASS_END_OF_TEXT;
+      DEBUG_MSG (APPLY, nullptr, "c%u at %u", klass, buffer->idx);
+      const Entry<EntryData> &entry = machine.get_entry (state, klass);
+
+      /* Unsafe-to-break before this if not in state 0, as things might
+       * go differently if we start from state 0 here.
+       *
+       * Ugh.  The indexing here is ugly... */
+      if (state && buffer->backtrack_len () && buffer->idx < buffer->len)
+      {
+	/* If there's no action and we're just epsilon-transitioning to state 0,
+	 * safe to break. */
+	if (c->is_actionable (this, entry) ||
+	    !(entry.newState == StateTable<Types, EntryData>::STATE_START_OF_TEXT &&
+	      entry.flags == context_t::DontAdvance))
+	  buffer->unsafe_to_break_from_outbuffer (buffer->backtrack_len () - 1, buffer->idx + 1);
+      }
+
+      /* Unsafe-to-break if end-of-text would kick in here. */
+      if (buffer->idx + 2 <= buffer->len)
+      {
+	const Entry<EntryData> &end_entry = machine.get_entry (state, StateTable<Types, EntryData>::CLASS_END_OF_TEXT);
+	if (c->is_actionable (this, end_entry))
+	  buffer->unsafe_to_break (buffer->idx, buffer->idx + 2);
+      }
+
+      c->transition (this, entry);
+
+      state = machine.new_state (entry.newState);
+      DEBUG_MSG (APPLY, nullptr, "s%d", state);
+
+      if (buffer->idx == buffer->len)
+	break;
+
+      if (!(entry.flags & context_t::DontAdvance) || buffer->max_ops-- <= 0)
+	buffer->next_glyph ();
+    }
+
+    if (!c->in_place)
+    {
+      for (; buffer->successful && buffer->idx < buffer->len;)
+	buffer->next_glyph ();
+      buffer->swap_buffers ();
+    }
+  }
+
+  public:
+  const StateTable<Types, EntryData> &machine;
+  hb_buffer_t *buffer;
+  unsigned int num_glyphs;
+};
+
+
+struct ankr;
+
+struct hb_aat_apply_context_t :
+       hb_dispatch_context_t<hb_aat_apply_context_t, bool, HB_DEBUG_APPLY>
+{
+  const char *get_name () { return "APPLY"; }
+  template <typename T>
+  return_t dispatch (const T &obj) { return obj.apply (this); }
+  static return_t default_return_value () { return false; }
+  bool stop_sublookup_iteration (return_t r) const { return r; }
+
+  const hb_ot_shape_plan_t *plan;
+  hb_font_t *font;
+  hb_face_t *face;
+  hb_buffer_t *buffer;
+  hb_sanitize_context_t sanitizer;
+  const ankr *ankr_table;
+
+  /* Unused. For debug tracing only. */
+  unsigned int lookup_index;
+
+  HB_INTERNAL hb_aat_apply_context_t (const hb_ot_shape_plan_t *plan_,
+				      hb_font_t *font_,
+				      hb_buffer_t *buffer_,
+				      hb_blob_t *blob = const_cast<hb_blob_t *> (&Null (hb_blob_t)));
+
+  HB_INTERNAL ~hb_aat_apply_context_t ();
+
+  HB_INTERNAL void set_ankr_table (const AAT::ankr *ankr_table_);
+
+  void set_lookup_index (unsigned int i) { lookup_index = i; }
+};
+
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-feat-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-feat-table.hh
new file mode 100644
index 0000000000..359e859cfc
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-feat-table.hh
@@ -0,0 +1,222 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LAYOUT_FEAT_TABLE_HH
+#define HB_AAT_LAYOUT_FEAT_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+
+/*
+ * feat -- Feature Name
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6feat.html
+ */
+#define HB_AAT_TAG_feat HB_TAG('f','e','a','t')
+
+
+namespace AAT {
+
+
+struct SettingName
+{
+  friend struct FeatureName;
+
+  int cmp (hb_aat_layout_feature_selector_t key) const
+  { return (int) key - (int) setting; }
+
+  hb_aat_layout_feature_selector_t get_selector () const
+  { return (hb_aat_layout_feature_selector_t) (unsigned) setting; }
+
+  hb_aat_layout_feature_selector_info_t get_info (hb_aat_layout_feature_selector_t default_selector) const
+  {
+    return {
+      nameIndex,
+      (hb_aat_layout_feature_selector_t) (unsigned int) setting,
+      default_selector == HB_AAT_LAYOUT_FEATURE_SELECTOR_INVALID
+	? (hb_aat_layout_feature_selector_t) (setting + 1)
+	: default_selector,
+      0
+    };
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	setting;	/* The setting. */
+  NameID	nameIndex;	/* The name table index for the setting's name. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+DECLARE_NULL_NAMESPACE_BYTES (AAT, SettingName);
+
+struct feat;
+
+struct FeatureName
+{
+  int cmp (hb_aat_layout_feature_type_t key) const
+  { return (int) key - (int) feature; }
+
+  enum {
+    Exclusive	= 0x8000,	/* If set, the feature settings are mutually exclusive. */
+    NotDefault	= 0x4000,	/* If clear, then the setting with an index of 0 in
+				 * the setting name array for this feature should
+				 * be taken as the default for the feature
+				 * (if one is required). If set, then bits 0-15 of this
+				 * featureFlags field contain the index of the setting
+				 * which is to be taken as the default. */
+    IndexMask	= 0x00FF	/* If bits 30 and 31 are set, then these sixteen bits
+				 * indicate the index of the setting in the setting name
+				 * array for this feature which should be taken
+				 * as the default. */
+  };
+
+  unsigned int get_selector_infos (unsigned int                           start_offset,
+				   unsigned int                          *selectors_count, /* IN/OUT.  May be NULL. */
+				   hb_aat_layout_feature_selector_info_t *selectors,       /* OUT.     May be NULL. */
+				   unsigned int                          *pdefault_index,  /* OUT.     May be NULL. */
+				   const void *base) const
+  {
+    hb_array_t< const SettingName> settings_table = (base+settingTableZ).as_array (nSettings);
+
+    static_assert (Index::NOT_FOUND_INDEX == HB_AAT_LAYOUT_NO_SELECTOR_INDEX, "");
+
+    hb_aat_layout_feature_selector_t default_selector = HB_AAT_LAYOUT_FEATURE_SELECTOR_INVALID;
+    unsigned int default_index = Index::NOT_FOUND_INDEX;
+    if (featureFlags & Exclusive)
+    {
+      default_index = (featureFlags & NotDefault) ? featureFlags & IndexMask : 0;
+      default_selector = settings_table[default_index].get_selector ();
+    }
+    if (pdefault_index)
+      *pdefault_index = default_index;
+
+    if (selectors_count)
+    {
+      + settings_table.sub_array (start_offset, selectors_count)
+      | hb_map ([=] (const SettingName& setting) { return setting.get_info (default_selector); })
+      | hb_sink (hb_array (selectors, *selectors_count))
+      ;
+    }
+    return settings_table.length;
+  }
+
+  hb_aat_layout_feature_type_t get_feature_type () const
+  { return (hb_aat_layout_feature_type_t) (unsigned int) feature; }
+
+  hb_ot_name_id_t get_feature_name_id () const { return nameIndex; }
+
+  bool is_exclusive () const { return featureFlags & Exclusive; }
+
+  /* A FeatureName with no settings is meaningless */
+  bool has_data () const { return nSettings; }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (base+settingTableZ).sanitize (c, nSettings)));
+  }
+
+  protected:
+  HBUINT16	feature;	/* Feature type. */
+  HBUINT16	nSettings;	/* The number of records in the setting name array. */
+  LNNOffsetTo<UnsizedArrayOf<SettingName>>
+		settingTableZ;	/* Offset in bytes from the beginning of this table to
+				 * this feature's setting name array. The actual type of
+				 * record this offset refers to will depend on the
+				 * exclusivity value, as described below. */
+  HBUINT16	featureFlags;	/* Single-bit flags associated with the feature type. */
+  HBINT16	nameIndex;	/* The name table index for the feature's name.
+				 * This index has values greater than 255 and
+				 * less than 32768. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct feat
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_feat;
+
+  bool has_data () const { return version.to_int (); }
+
+  unsigned int get_feature_types (unsigned int                  start_offset,
+				  unsigned int                 *count,
+				  hb_aat_layout_feature_type_t *features) const
+  {
+    if (count)
+    {
+      + namesZ.as_array (featureNameCount).sub_array (start_offset, count)
+      | hb_map (&FeatureName::get_feature_type)
+      | hb_sink (hb_array (features, *count))
+      ;
+    }
+    return featureNameCount;
+  }
+
+  bool exposes_feature (hb_aat_layout_feature_type_t feature_type) const
+  { return get_feature (feature_type).has_data (); }
+
+  const FeatureName& get_feature (hb_aat_layout_feature_type_t feature_type) const
+  { return namesZ.bsearch (featureNameCount, feature_type); }
+
+  hb_ot_name_id_t get_feature_name_id (hb_aat_layout_feature_type_t feature) const
+  { return get_feature (feature).get_feature_name_id (); }
+
+  unsigned int get_selector_infos (hb_aat_layout_feature_type_t           feature_type,
+				   unsigned int                           start_offset,
+				   unsigned int                          *selectors_count, /* IN/OUT.  May be NULL. */
+				   hb_aat_layout_feature_selector_info_t *selectors,       /* OUT.     May be NULL. */
+				   unsigned int                          *default_index    /* OUT.     May be NULL. */) const
+  {
+    return get_feature (feature_type).get_selector_infos (start_offset, selectors_count, selectors,
+							  default_index, this);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version.major == 1 &&
+			  namesZ.sanitize (c, featureNameCount, this)));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version number of the feature name table
+				 * (0x00010000 for the current version). */
+  HBUINT16	featureNameCount;
+				/* The number of entries in the feature name array. */
+  HBUINT16	reserved1;	/* Reserved (set to zero). */
+  HBUINT32	reserved2;	/* Reserved (set to zero). */
+  SortedUnsizedArrayOf<FeatureName>
+		namesZ;		/* The feature name array. */
+  public:
+  DEFINE_SIZE_ARRAY (12, namesZ);
+};
+
+} /* namespace AAT */
+
+#endif /* HB_AAT_LAYOUT_FEAT_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-just-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-just-table.hh
new file mode 100644
index 0000000000..49506e9f5a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-just-table.hh
@@ -0,0 +1,417 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LAYOUT_JUST_TABLE_HH
+#define HB_AAT_LAYOUT_JUST_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+#include "hb-ot-layout.hh"
+#include "hb-open-type.hh"
+
+#include "hb-aat-layout-morx-table.hh"
+
+/*
+ * just -- Justification
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6just.html
+ */
+#define HB_AAT_TAG_just HB_TAG('j','u','s','t')
+
+
+namespace AAT {
+
+using namespace OT;
+
+
+struct ActionSubrecordHeader
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  HBUINT16	actionClass;	/* The JustClass value associated with this
+				 * ActionSubrecord. */
+  HBUINT16	actionType;	/* The type of postcompensation action. */
+  HBUINT16	actionLength;	/* Length of this ActionSubrecord record, which
+				 * must be a multiple of 4. */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct DecompositionAction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  ActionSubrecordHeader
+		header;
+  HBFixed	lowerLimit;	/* If the distance factor is less than this value,
+				 * then the ligature is decomposed. */
+  HBFixed	upperLimit;	/* If the distance factor is greater than this value,
+				 * then the ligature is decomposed. */
+  HBUINT16	order;		/* Numerical order in which this ligature will
+				 * be decomposed; you may want infrequent ligatures
+				 * to decompose before more frequent ones. The ligatures
+				 * on the line of text will decompose in increasing
+				 * value of this field. */
+  ArrayOf<HBUINT16>
+		decomposedglyphs;
+				/* Number of 16-bit glyph indexes that follow;
+				 * the ligature will be decomposed into these glyphs.
+				 *
+				 * Array of decomposed glyphs. */
+  public:
+  DEFINE_SIZE_ARRAY (18, decomposedglyphs);
+};
+
+struct UnconditionalAddGlyphAction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  ActionSubrecordHeader
+		header;
+  HBGlyphID	addGlyph;	/* Glyph that should be added if the distance factor
+				 * is growing. */
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct ConditionalAddGlyphAction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  ActionSubrecordHeader
+		header;
+  HBFixed	substThreshold; /* Distance growth factor (in ems) at which
+				 * this glyph is replaced and the growth factor
+				 * recalculated. */
+  HBGlyphID	addGlyph;	/* Glyph to be added as kashida. If this value is
+				 * 0xFFFF, no extra glyph will be added. Note that
+				 * generally when a glyph is added, justification
+				 * will need to be redone. */
+  HBGlyphID	substGlyph;	/* Glyph to be substituted for this glyph if the
+				 * growth factor equals or exceeds the value of
+				 * substThreshold. */
+  public:
+  DEFINE_SIZE_STATIC (14);
+};
+
+struct DuctileGlyphAction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  ActionSubrecordHeader
+		header;
+  HBUINT32	variationAxis;	/* The 4-byte tag identifying the ductile axis.
+				 * This would normally be 0x64756374 ('duct'),
+				 * but you may use any axis the font contains. */
+  HBFixed	minimumLimit;	/* The lowest value for the ductility axis tha
+				 * still yields an acceptable appearance. Normally
+				 * this will be 1.0. */
+  HBFixed	noStretchValue; /* This is the default value that corresponds to
+				 * no change in appearance. Normally, this will
+				 * be 1.0. */
+  HBFixed	maximumLimit;	/* The highest value for the ductility axis that
+				 * still yields an acceptable appearance. */
+  public:
+  DEFINE_SIZE_STATIC (22);
+};
+
+struct RepeatedAddGlyphAction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  ActionSubrecordHeader
+		header;
+  HBUINT16	flags;		/* Currently unused; set to 0. */
+  HBGlyphID	glyph;		/* Glyph that should be added if the distance factor
+				 * is growing. */
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+struct ActionSubrecord
+{
+  unsigned int get_length () const { return u.header.actionLength; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (u.header.actionType)
+    {
+    case 0:  return_trace (u.decompositionAction.sanitize (c));
+    case 1:  return_trace (u.unconditionalAddGlyphAction.sanitize (c));
+    case 2:  return_trace (u.conditionalAddGlyphAction.sanitize (c));
+    // case 3: return_trace (u.stretchGlyphAction.sanitize (c));
+    case 4:  return_trace (u.decompositionAction.sanitize (c));
+    case 5:  return_trace (u.decompositionAction.sanitize (c));
+    default: return_trace (true);
+    }
+  }
+
+  protected:
+  union	{
+  ActionSubrecordHeader		header;
+  DecompositionAction		decompositionAction;
+  UnconditionalAddGlyphAction	unconditionalAddGlyphAction;
+  ConditionalAddGlyphAction	conditionalAddGlyphAction;
+  /* StretchGlyphAction stretchGlyphAction; -- Not supported by CoreText */
+  DuctileGlyphAction		ductileGlyphAction;
+  RepeatedAddGlyphAction	repeatedAddGlyphAction;
+  } u;				/* Data. The format of this data depends on
+				 * the value of the actionType field. */
+  public:
+  DEFINE_SIZE_UNION (6, header);
+};
+
+struct PostcompensationActionChain
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    unsigned int offset = min_size;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const ActionSubrecord& subrecord = StructAtOffset<ActionSubrecord> (this, offset);
+      if (unlikely (!subrecord.sanitize (c))) return_trace (false);
+      offset += subrecord.get_length ();
+    }
+
+    return_trace (true);
+  }
+
+  protected:
+  HBUINT32	count;
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct JustWidthDeltaEntry
+{
+  enum Flags
+  {
+    Reserved1		=0xE000,/* Reserved. You should set these bits to zero. */
+    UnlimiteGap		=0x1000,/* The glyph can take unlimited gap. When this
+				 * glyph participates in the justification process,
+				 * it and any other glyphs on the line having this
+				 * bit set absorb all the remaining gap. */
+    Reserved2		=0x0FF0,/* Reserved. You should set these bits to zero. */
+    Priority		=0x000F /* The justification priority of the glyph. */
+  };
+
+  enum Priority
+  {
+    Kashida		= 0,	/* Kashida priority. This is the highest priority
+				 * during justification. */
+    Whitespace		= 1,	/* Whitespace priority. Any whitespace glyphs (as
+				 * identified in the glyph properties table) will
+				 * get this priority. */
+    InterCharacter	= 2,	/* Inter-character priority. Give this to any
+				 * remaining glyphs. */
+    NullPriority	= 3	/* Null priority. You should set this priority for
+				 * glyphs that only participate in justification
+				 * after the above priorities. Normally all glyphs
+				 * have one of the previous three values. If you
+				 * don't want a glyph to participate in justification,
+				 * and you don't want to set its factors to zero,
+				 * you may instead assign it to the null priority. */
+  };
+
+  protected:
+  HBFixed	beforeGrowLimit;/* The ratio by which the advance width of the
+				 * glyph is permitted to grow on the left or top side. */
+  HBFixed	beforeShrinkLimit;
+				/* The ratio by which the advance width of the
+				 * glyph is permitted to shrink on the left or top side. */
+  HBFixed	afterGrowLimit;	/* The ratio by which the advance width of the glyph
+				 * is permitted to shrink on the left or top side. */
+  HBFixed	afterShrinkLimit;
+				/* The ratio by which the advance width of the glyph
+				 * is at most permitted to shrink on the right or
+				 * bottom side. */
+  HBUINT16	growFlags;	/* Flags controlling the grow case. */
+  HBUINT16	shrinkFlags;	/* Flags controlling the shrink case. */
+
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+struct WidthDeltaPair
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT32	justClass;	/* The justification category associated
+				 * with the wdRecord field. Only 7 bits of
+				 * this field are used. (The other bits are
+				 * used as padding to guarantee longword
+				 * alignment of the following record). */
+  JustWidthDeltaEntry
+		wdRecord;	/* The actual width delta record. */
+
+  public:
+  DEFINE_SIZE_STATIC (24);
+};
+
+typedef OT::LArrayOf<WidthDeltaPair> WidthDeltaCluster;
+
+struct JustificationCategory
+{
+  typedef void EntryData;
+
+  enum Flags
+  {
+    SetMark		=0x8000,/* If set, make the current glyph the marked
+				 * glyph. */
+    DontAdvance		=0x4000,/* If set, don't advance to the next glyph before
+				 * going to the new state. */
+    MarkCategory	=0x3F80,/* The justification category for the marked
+				 * glyph if nonzero. */
+    CurrentCategory	=0x007F /* The justification category for the current
+				 * glyph if nonzero. */
+  };
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  morphHeader.sanitize (c) &&
+			  stHeader.sanitize (c)));
+  }
+
+  protected:
+  ChainSubtable<ObsoleteTypes>
+		morphHeader;	/* Metamorphosis-style subtable header. */
+  StateTable<ObsoleteTypes, EntryData>
+		stHeader;	/* The justification insertion state table header */
+  public:
+  DEFINE_SIZE_STATIC (30);
+};
+
+struct JustificationHeader
+{
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  justClassTable.sanitize (c, base, base) &&
+			  wdcTable.sanitize (c, base) &&
+			  pcTable.sanitize (c, base) &&
+			  lookupTable.sanitize (c, base)));
+  }
+
+  protected:
+  OffsetTo<JustificationCategory>
+		justClassTable;	/* Offset to the justification category state table. */
+  OffsetTo<WidthDeltaCluster>
+		wdcTable;	/* Offset from start of justification table to start
+				 * of the subtable containing the width delta factors
+				 * for the glyphs in your font.
+				 *
+				 * The width delta clusters table. */
+  OffsetTo<PostcompensationActionChain>
+		pcTable;	/* Offset from start of justification table to start
+				 * of postcompensation subtable (set to zero if none).
+				 *
+				 * The postcompensation subtable, if present in the font. */
+  Lookup<OffsetTo<WidthDeltaCluster>>
+		lookupTable;	/* Lookup table associating glyphs with width delta
+				 * clusters. See the description of Width Delta Clusters
+				 * table for details on how to interpret the lookup values. */
+
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+struct just
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_just;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+
+    return_trace (likely (c->check_struct (this) &&
+			  version.major == 1 &&
+			  horizData.sanitize (c, this, this) &&
+			  vertData.sanitize (c, this, this)));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the justification table
+				 * (0x00010000u for version 1.0). */
+  HBUINT16	format;		/* Format of the justification table (set to 0). */
+  OffsetTo<JustificationHeader>
+		horizData;	/* Byte offset from the start of the justification table
+				 * to the header for tables that contain justification
+				 * information for horizontal text.
+				 * If you are not including this information,
+				 * store 0. */
+  OffsetTo<JustificationHeader>
+		vertData;	/* ditto, vertical */
+
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_JUST_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-kerx-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-kerx-table.hh
new file mode 100644
index 0000000000..1cd412164e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-kerx-table.hh
@@ -0,0 +1,999 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_LAYOUT_KERX_TABLE_HH
+#define HB_AAT_LAYOUT_KERX_TABLE_HH
+
+#include "hb-kern.hh"
+#include "hb-aat-layout-ankr-table.hh"
+
+/*
+ * kerx -- Extended Kerning
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6kerx.html
+ */
+#define HB_AAT_TAG_kerx HB_TAG('k','e','r','x')
+
+
+namespace AAT {
+
+using namespace OT;
+
+
+static inline int
+kerxTupleKern (int value,
+	       unsigned int tupleCount,
+	       const void *base,
+	       hb_aat_apply_context_t *c)
+{
+  if (likely (!tupleCount || !c)) return value;
+
+  unsigned int offset = value;
+  const FWORD *pv = &StructAtOffset<FWORD> (base, offset);
+  if (unlikely (!c->sanitizer.check_array (pv, tupleCount))) return 0;
+  return *pv;
+}
+
+
+struct hb_glyph_pair_t
+{
+  hb_codepoint_t left;
+  hb_codepoint_t right;
+};
+
+struct KernPair
+{
+  int get_kerning () const { return value; }
+
+  int cmp (const hb_glyph_pair_t &o) const
+  {
+    int ret = left.cmp (o.left);
+    if (ret) return ret;
+    return right.cmp (o.right);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBGlyphID	left;
+  HBGlyphID	right;
+  FWORD		value;
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+template <typename KernSubTableHeader>
+struct KerxSubTableFormat0
+{
+  int get_kerning (hb_codepoint_t left, hb_codepoint_t right,
+		   hb_aat_apply_context_t *c = nullptr) const
+  {
+    hb_glyph_pair_t pair = {left, right};
+    int v = pairs.bsearch (pair).get_kerning ();
+    return kerxTupleKern (v, header.tuple_count (), this, c);
+  }
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    if (!c->plan->requested_kerning)
+      return false;
+
+    if (header.coverage & header.Backwards)
+      return false;
+
+    accelerator_t accel (*this, c);
+    hb_kern_machine_t<accelerator_t> machine (accel, header.coverage & header.CrossStream);
+    machine.kern (c->font, c->buffer, c->plan->kern_mask);
+
+    return_trace (true);
+  }
+
+  struct accelerator_t
+  {
+    const KerxSubTableFormat0 &table;
+    hb_aat_apply_context_t *c;
+
+    accelerator_t (const KerxSubTableFormat0 &table_,
+		   hb_aat_apply_context_t *c_) :
+		     table (table_), c (c_) {}
+
+    int get_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+    { return table.get_kerning (left, right, c); }
+  };
+
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (pairs.sanitize (c)));
+  }
+
+  protected:
+  KernSubTableHeader	header;
+  BinSearchArrayOf<KernPair, typename KernSubTableHeader::Types::HBUINT>
+			pairs;	/* Sorted kern records. */
+  public:
+  DEFINE_SIZE_ARRAY (KernSubTableHeader::static_size + 16, pairs);
+};
+
+
+template <bool extended>
+struct Format1Entry;
+
+template <>
+struct Format1Entry<true>
+{
+  enum Flags
+  {
+    Push		= 0x8000,	/* If set, push this glyph on the kerning stack. */
+    DontAdvance		= 0x4000,	/* If set, don't advance to the next glyph
+					 * before going to the new state. */
+    Reset		= 0x2000,	/* If set, reset the kerning data (clear the stack) */
+    Reserved		= 0x1FFF,	/* Not used; set to 0. */
+  };
+
+  struct EntryData
+  {
+    HBUINT16	kernActionIndex;/* Index into the kerning value array. If
+				 * this index is 0xFFFF, then no kerning
+				 * is to be performed. */
+    public:
+    DEFINE_SIZE_STATIC (2);
+  };
+
+  static bool performAction (const Entry<EntryData> &entry)
+  { return entry.data.kernActionIndex != 0xFFFF; }
+
+  static unsigned int kernActionIndex (const Entry<EntryData> &entry)
+  { return entry.data.kernActionIndex; }
+};
+template <>
+struct Format1Entry<false>
+{
+  enum Flags
+  {
+    Push		= 0x8000,	/* If set, push this glyph on the kerning stack. */
+    DontAdvance		= 0x4000,	/* If set, don't advance to the next glyph
+					 * before going to the new state. */
+    Offset		= 0x3FFF,	/* Byte offset from beginning of subtable to the
+					 * value table for the glyphs on the kerning stack. */
+
+    Reset		= 0x0000,	/* Not supported? */
+  };
+
+  typedef void EntryData;
+
+  static bool performAction (const Entry<EntryData> &entry)
+  { return entry.flags & Offset; }
+
+  static unsigned int kernActionIndex (const Entry<EntryData> &entry)
+  { return entry.flags & Offset; }
+};
+
+template <typename KernSubTableHeader>
+struct KerxSubTableFormat1
+{
+  typedef typename KernSubTableHeader::Types Types;
+  typedef typename Types::HBUINT HBUINT;
+
+  typedef Format1Entry<Types::extended> Format1EntryT;
+  typedef typename Format1EntryT::EntryData EntryData;
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = true;
+    enum
+    {
+      DontAdvance	= Format1EntryT::DontAdvance,
+    };
+
+    driver_context_t (const KerxSubTableFormat1 *table_,
+		      hb_aat_apply_context_t *c_) :
+	c (c_),
+	table (table_),
+	/* Apparently the offset kernAction is from the beginning of the state-machine,
+	 * similar to offsets in morx table, NOT from beginning of this table, like
+	 * other subtables in kerx.  Discovered via testing. */
+	kernAction (&table->machine + table->kernAction),
+	depth (0),
+	crossStream (table->header.coverage & table->header.CrossStream) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver HB_UNUSED,
+			const Entry<EntryData> &entry)
+    { return Format1EntryT::performAction (entry); }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+      unsigned int flags = entry.flags;
+
+      if (flags & Format1EntryT::Reset)
+	depth = 0;
+
+      if (flags & Format1EntryT::Push)
+      {
+	if (likely (depth < ARRAY_LENGTH (stack)))
+	  stack[depth++] = buffer->idx;
+	else
+	  depth = 0; /* Probably not what CoreText does, but better? */
+      }
+
+      if (Format1EntryT::performAction (entry) && depth)
+      {
+	unsigned int tuple_count = hb_max (1u, table->header.tuple_count ());
+
+	unsigned int kern_idx = Format1EntryT::kernActionIndex (entry);
+	kern_idx = Types::byteOffsetToIndex (kern_idx, &table->machine, kernAction.arrayZ);
+	const FWORD *actions = &kernAction[kern_idx];
+	if (!c->sanitizer.check_array (actions, depth, tuple_count))
+	{
+	  depth = 0;
+	  return;
+	}
+
+	hb_mask_t kern_mask = c->plan->kern_mask;
+
+	/* From Apple 'kern' spec:
+	 * "Each pops one glyph from the kerning stack and applies the kerning value to it.
+	 * The end of the list is marked by an odd value... */
+	bool last = false;
+	while (!last && depth)
+	{
+	  unsigned int idx = stack[--depth];
+	  int v = *actions;
+	  actions += tuple_count;
+	  if (idx >= buffer->len) continue;
+
+	  /* "The end of the list is marked by an odd value..." */
+	  last = v & 1;
+	  v &= ~1;
+
+	  hb_glyph_position_t &o = buffer->pos[idx];
+
+	  if (HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction))
+	  {
+	    if (crossStream)
+	    {
+	      /* The following flag is undocumented in the spec, but described
+	       * in the 'kern' table example. */
+	      if (v == -0x8000)
+	      {
+		o.attach_type() = ATTACH_TYPE_NONE;
+		o.attach_chain() = 0;
+		o.y_offset = 0;
+	      }
+	      else if (o.attach_type())
+	      {
+		o.y_offset += c->font->em_scale_y (v);
+		buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+	      }
+	    }
+	    else if (buffer->info[idx].mask & kern_mask)
+	    {
+	      o.x_advance += c->font->em_scale_x (v);
+	      o.x_offset += c->font->em_scale_x (v);
+	    }
+	  }
+	  else
+	  {
+	    if (crossStream)
+	    {
+	      /* CoreText doesn't do crossStream kerning in vertical.  We do. */
+	      if (v == -0x8000)
+	      {
+		o.attach_type() = ATTACH_TYPE_NONE;
+		o.attach_chain() = 0;
+		o.x_offset = 0;
+	      }
+	      else if (o.attach_type())
+	      {
+		o.x_offset += c->font->em_scale_x (v);
+		buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+	      }
+	    }
+	    else if (buffer->info[idx].mask & kern_mask)
+	    {
+	      o.y_advance += c->font->em_scale_y (v);
+	      o.y_offset += c->font->em_scale_y (v);
+	    }
+	  }
+	}
+      }
+    }
+
+    private:
+    hb_aat_apply_context_t *c;
+    const KerxSubTableFormat1 *table;
+    const UnsizedArrayOf<FWORD> &kernAction;
+    unsigned int stack[8];
+    unsigned int depth;
+    bool crossStream;
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    if (!c->plan->requested_kerning &&
+	!(header.coverage & header.CrossStream))
+      return false;
+
+    driver_context_t dc (this, c);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->font->face);
+    driver.drive (&dc);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    /* The rest of array sanitizations are done at run-time. */
+    return_trace (likely (c->check_struct (this) &&
+			  machine.sanitize (c)));
+  }
+
+  protected:
+  KernSubTableHeader				header;
+  StateTable<Types, EntryData>			machine;
+  NNOffsetTo<UnsizedArrayOf<FWORD>, HBUINT>	kernAction;
+  public:
+  DEFINE_SIZE_STATIC (KernSubTableHeader::static_size + 5 * sizeof (HBUINT));
+};
+
+template <typename KernSubTableHeader>
+struct KerxSubTableFormat2
+{
+  typedef typename KernSubTableHeader::Types Types;
+  typedef typename Types::HBUINT HBUINT;
+
+  int get_kerning (hb_codepoint_t left, hb_codepoint_t right,
+		   hb_aat_apply_context_t *c) const
+  {
+    unsigned int num_glyphs = c->sanitizer.get_num_glyphs ();
+    unsigned int l = (this+leftClassTable).get_class (left, num_glyphs, 0);
+    unsigned int r = (this+rightClassTable).get_class (right, num_glyphs, 0);
+
+    const UnsizedArrayOf<FWORD> &arrayZ = this+array;
+    unsigned int kern_idx = l + r;
+    kern_idx = Types::offsetToIndex (kern_idx, this, arrayZ.arrayZ);
+    const FWORD *v = &arrayZ[kern_idx];
+    if (unlikely (!v->sanitize (&c->sanitizer))) return 0;
+
+    return kerxTupleKern (*v, header.tuple_count (), this, c);
+  }
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    if (!c->plan->requested_kerning)
+      return false;
+
+    if (header.coverage & header.Backwards)
+      return false;
+
+    accelerator_t accel (*this, c);
+    hb_kern_machine_t<accelerator_t> machine (accel, header.coverage & header.CrossStream);
+    machine.kern (c->font, c->buffer, c->plan->kern_mask);
+
+    return_trace (true);
+  }
+
+  struct accelerator_t
+  {
+    const KerxSubTableFormat2 &table;
+    hb_aat_apply_context_t *c;
+
+    accelerator_t (const KerxSubTableFormat2 &table_,
+		   hb_aat_apply_context_t *c_) :
+		     table (table_), c (c_) {}
+
+    int get_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+    { return table.get_kerning (left, right, c); }
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  leftClassTable.sanitize (c, this) &&
+			  rightClassTable.sanitize (c, this) &&
+			  c->check_range (this, array)));
+  }
+
+  protected:
+  KernSubTableHeader	header;
+  HBUINT		rowWidth;	/* The width, in bytes, of a row in the table. */
+  NNOffsetTo<typename Types::ClassTypeWide, HBUINT>
+			leftClassTable;	/* Offset from beginning of this subtable to
+					 * left-hand class table. */
+  NNOffsetTo<typename Types::ClassTypeWide, HBUINT>
+			rightClassTable;/* Offset from beginning of this subtable to
+					 * right-hand class table. */
+  NNOffsetTo<UnsizedArrayOf<FWORD>, HBUINT>
+			 array;		/* Offset from beginning of this subtable to
+					 * the start of the kerning array. */
+  public:
+  DEFINE_SIZE_STATIC (KernSubTableHeader::static_size + 4 * sizeof (HBUINT));
+};
+
+template <typename KernSubTableHeader>
+struct KerxSubTableFormat4
+{
+  typedef ExtendedTypes Types;
+
+  struct EntryData
+  {
+    HBUINT16	ankrActionIndex;/* Either 0xFFFF (for no action) or the index of
+				 * the action to perform. */
+    public:
+    DEFINE_SIZE_STATIC (2);
+  };
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = true;
+    enum Flags
+    {
+      Mark		= 0x8000,	/* If set, remember this glyph as the marked glyph. */
+      DontAdvance	= 0x4000,	/* If set, don't advance to the next glyph before
+					 * going to the new state. */
+      Reserved		= 0x3FFF,	/* Not used; set to 0. */
+    };
+
+    enum SubTableFlags
+    {
+      ActionType	= 0xC0000000,	/* A two-bit field containing the action type. */
+      Unused		= 0x3F000000,	/* Unused - must be zero. */
+      Offset		= 0x00FFFFFF,	/* Masks the offset in bytes from the beginning
+					 * of the subtable to the beginning of the control
+					 * point table. */
+    };
+
+    driver_context_t (const KerxSubTableFormat4 *table,
+		      hb_aat_apply_context_t *c_) :
+	c (c_),
+	action_type ((table->flags & ActionType) >> 30),
+	ankrData ((HBUINT16 *) ((const char *) &table->machine + (table->flags & Offset))),
+	mark_set (false),
+	mark (0) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver HB_UNUSED,
+			const Entry<EntryData> &entry)
+    { return entry.data.ankrActionIndex != 0xFFFF; }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+
+      if (mark_set && entry.data.ankrActionIndex != 0xFFFF && buffer->idx < buffer->len)
+      {
+	hb_glyph_position_t &o = buffer->cur_pos();
+	switch (action_type)
+	{
+	  case 0: /* Control Point Actions.*/
+	  {
+	    /* Indexed into glyph outline. */
+	    /* Each action (record in ankrData) contains two 16-bit fields, so we must
+	       double the ankrActionIndex to get the correct offset here. */
+	    const HBUINT16 *data = &ankrData[entry.data.ankrActionIndex * 2];
+	    if (!c->sanitizer.check_array (data, 2)) return;
+	    unsigned int markControlPoint = *data++;
+	    unsigned int currControlPoint = *data++;
+	    hb_position_t markX = 0;
+	    hb_position_t markY = 0;
+	    hb_position_t currX = 0;
+	    hb_position_t currY = 0;
+	    if (!c->font->get_glyph_contour_point_for_origin (c->buffer->info[mark].codepoint,
+							      markControlPoint,
+							      HB_DIRECTION_LTR /*XXX*/,
+							      &markX, &markY) ||
+		!c->font->get_glyph_contour_point_for_origin (c->buffer->cur ().codepoint,
+							      currControlPoint,
+							      HB_DIRECTION_LTR /*XXX*/,
+							      &currX, &currY))
+	      return;
+
+	    o.x_offset = markX - currX;
+	    o.y_offset = markY - currY;
+	  }
+	  break;
+
+	  case 1: /* Anchor Point Actions. */
+	  {
+	    /* Indexed into 'ankr' table. */
+	    /* Each action (record in ankrData) contains two 16-bit fields, so we must
+	       double the ankrActionIndex to get the correct offset here. */
+	    const HBUINT16 *data = &ankrData[entry.data.ankrActionIndex * 2];
+	    if (!c->sanitizer.check_array (data, 2)) return;
+	    unsigned int markAnchorPoint = *data++;
+	    unsigned int currAnchorPoint = *data++;
+	    const Anchor &markAnchor = c->ankr_table->get_anchor (c->buffer->info[mark].codepoint,
+								  markAnchorPoint,
+								  c->sanitizer.get_num_glyphs ());
+	    const Anchor &currAnchor = c->ankr_table->get_anchor (c->buffer->cur ().codepoint,
+								  currAnchorPoint,
+								  c->sanitizer.get_num_glyphs ());
+
+	    o.x_offset = c->font->em_scale_x (markAnchor.xCoordinate) - c->font->em_scale_x (currAnchor.xCoordinate);
+	    o.y_offset = c->font->em_scale_y (markAnchor.yCoordinate) - c->font->em_scale_y (currAnchor.yCoordinate);
+	  }
+	  break;
+
+	  case 2: /* Control Point Coordinate Actions. */
+	  {
+	    /* Each action contains four 16-bit fields, so we multiply the ankrActionIndex
+	       by 4 to get the correct offset for the given action. */
+	    const FWORD *data = (const FWORD *) &ankrData[entry.data.ankrActionIndex * 4];
+	    if (!c->sanitizer.check_array (data, 4)) return;
+	    int markX = *data++;
+	    int markY = *data++;
+	    int currX = *data++;
+	    int currY = *data++;
+
+	    o.x_offset = c->font->em_scale_x (markX) - c->font->em_scale_x (currX);
+	    o.y_offset = c->font->em_scale_y (markY) - c->font->em_scale_y (currY);
+	  }
+	  break;
+	}
+	o.attach_type() = ATTACH_TYPE_MARK;
+	o.attach_chain() = (int) mark - (int) buffer->idx;
+	buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+      }
+
+      if (entry.flags & Mark)
+      {
+	mark_set = true;
+	mark = buffer->idx;
+      }
+    }
+
+    private:
+    hb_aat_apply_context_t *c;
+    unsigned int action_type;
+    const HBUINT16 *ankrData;
+    bool mark_set;
+    unsigned int mark;
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    driver_context_t dc (this, c);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->font->face);
+    driver.drive (&dc);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    /* The rest of array sanitizations are done at run-time. */
+    return_trace (likely (c->check_struct (this) &&
+			  machine.sanitize (c)));
+  }
+
+  protected:
+  KernSubTableHeader		header;
+  StateTable<Types, EntryData>	machine;
+  HBUINT32			flags;
+  public:
+  DEFINE_SIZE_STATIC (KernSubTableHeader::static_size + 20);
+};
+
+template <typename KernSubTableHeader>
+struct KerxSubTableFormat6
+{
+  enum Flags
+  {
+    ValuesAreLong	= 0x00000001,
+  };
+
+  bool is_long () const { return flags & ValuesAreLong; }
+
+  int get_kerning (hb_codepoint_t left, hb_codepoint_t right,
+		   hb_aat_apply_context_t *c) const
+  {
+    unsigned int num_glyphs = c->sanitizer.get_num_glyphs ();
+    if (is_long ())
+    {
+      const typename U::Long &t = u.l;
+      unsigned int l = (this+t.rowIndexTable).get_value_or_null (left, num_glyphs);
+      unsigned int r = (this+t.columnIndexTable).get_value_or_null (right, num_glyphs);
+      unsigned int offset = l + r;
+      if (unlikely (offset < l)) return 0; /* Addition overflow. */
+      if (unlikely (hb_unsigned_mul_overflows (offset, sizeof (FWORD32)))) return 0;
+      const FWORD32 *v = &StructAtOffset<FWORD32> (&(this+t.array), offset * sizeof (FWORD32));
+      if (unlikely (!v->sanitize (&c->sanitizer))) return 0;
+      return kerxTupleKern (*v, header.tuple_count (), &(this+vector), c);
+    }
+    else
+    {
+      const typename U::Short &t = u.s;
+      unsigned int l = (this+t.rowIndexTable).get_value_or_null (left, num_glyphs);
+      unsigned int r = (this+t.columnIndexTable).get_value_or_null (right, num_glyphs);
+      unsigned int offset = l + r;
+      const FWORD *v = &StructAtOffset<FWORD> (&(this+t.array), offset * sizeof (FWORD));
+      if (unlikely (!v->sanitize (&c->sanitizer))) return 0;
+      return kerxTupleKern (*v, header.tuple_count (), &(this+vector), c);
+    }
+  }
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    if (!c->plan->requested_kerning)
+      return false;
+
+    if (header.coverage & header.Backwards)
+      return false;
+
+    accelerator_t accel (*this, c);
+    hb_kern_machine_t<accelerator_t> machine (accel, header.coverage & header.CrossStream);
+    machine.kern (c->font, c->buffer, c->plan->kern_mask);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (is_long () ?
+			   (
+			     u.l.rowIndexTable.sanitize (c, this) &&
+			     u.l.columnIndexTable.sanitize (c, this) &&
+			     c->check_range (this, u.l.array)
+			   ) : (
+			     u.s.rowIndexTable.sanitize (c, this) &&
+			     u.s.columnIndexTable.sanitize (c, this) &&
+			     c->check_range (this, u.s.array)
+			   )) &&
+			  (header.tuple_count () == 0 ||
+			   c->check_range (this, vector))));
+  }
+
+  struct accelerator_t
+  {
+    const KerxSubTableFormat6 &table;
+    hb_aat_apply_context_t *c;
+
+    accelerator_t (const KerxSubTableFormat6 &table_,
+		   hb_aat_apply_context_t *c_) :
+		     table (table_), c (c_) {}
+
+    int get_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+    { return table.get_kerning (left, right, c); }
+  };
+
+  protected:
+  KernSubTableHeader		header;
+  HBUINT32			flags;
+  HBUINT16			rowCount;
+  HBUINT16			columnCount;
+  union U
+  {
+    struct Long
+    {
+      LNNOffsetTo<Lookup<HBUINT32>>		rowIndexTable;
+      LNNOffsetTo<Lookup<HBUINT32>>		columnIndexTable;
+      LNNOffsetTo<UnsizedArrayOf<FWORD32>>	array;
+    } l;
+    struct Short
+    {
+      LNNOffsetTo<Lookup<HBUINT16>>		rowIndexTable;
+      LNNOffsetTo<Lookup<HBUINT16>>		columnIndexTable;
+      LNNOffsetTo<UnsizedArrayOf<FWORD>>	array;
+    } s;
+  } u;
+  LNNOffsetTo<UnsizedArrayOf<FWORD>>	vector;
+  public:
+  DEFINE_SIZE_STATIC (KernSubTableHeader::static_size + 24);
+};
+
+
+struct KerxSubTableHeader
+{
+  typedef ExtendedTypes Types;
+
+  unsigned   tuple_count () const { return tupleCount; }
+  bool     is_horizontal () const { return !(coverage & Vertical); }
+
+  enum Coverage
+  {
+    Vertical	= 0x80000000u,	/* Set if table has vertical kerning values. */
+    CrossStream	= 0x40000000u,	/* Set if table has cross-stream kerning values. */
+    Variation	= 0x20000000u,	/* Set if table has variation kerning values. */
+    Backwards	= 0x10000000u,	/* If clear, process the glyphs forwards, that
+				 * is, from first to last in the glyph stream.
+				 * If we, process them from last to first.
+				 * This flag only applies to state-table based
+				 * 'kerx' subtables (types 1 and 4). */
+    Reserved	= 0x0FFFFF00u,	/* Reserved, set to zero. */
+    SubtableType= 0x000000FFu,	/* Subtable type. */
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  public:
+  HBUINT32	length;
+  HBUINT32	coverage;
+  HBUINT32	tupleCount;
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct KerxSubTable
+{
+  friend struct kerx;
+
+  unsigned int get_size () const { return u.header.length; }
+  unsigned int get_type () const { return u.header.coverage & u.header.SubtableType; }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    unsigned int subtable_type = get_type ();
+    TRACE_DISPATCH (this, subtable_type);
+    switch (subtable_type) {
+    case 0:	return_trace (c->dispatch (u.format0, hb_forward<Ts> (ds)...));
+    case 1:	return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2:	return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    case 4:	return_trace (c->dispatch (u.format4, hb_forward<Ts> (ds)...));
+    case 6:	return_trace (c->dispatch (u.format6, hb_forward<Ts> (ds)...));
+    default:	return_trace (c->default_return_value ());
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.header.sanitize (c) ||
+	u.header.length <= u.header.static_size ||
+	!c->check_range (this, u.header.length))
+      return_trace (false);
+
+    return_trace (dispatch (c));
+  }
+
+  public:
+  union {
+  KerxSubTableHeader				header;
+  KerxSubTableFormat0<KerxSubTableHeader>	format0;
+  KerxSubTableFormat1<KerxSubTableHeader>	format1;
+  KerxSubTableFormat2<KerxSubTableHeader>	format2;
+  KerxSubTableFormat4<KerxSubTableHeader>	format4;
+  KerxSubTableFormat6<KerxSubTableHeader>	format6;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (12);
+};
+
+
+/*
+ * The 'kerx' Table
+ */
+
+template <typename T>
+struct KerxTable
+{
+  /* https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern */
+  const T* thiz () const { return static_cast<const T *> (this); }
+
+  bool has_state_machine () const
+  {
+    typedef typename T::SubTable SubTable;
+
+    const SubTable *st = &thiz()->firstSubTable;
+    unsigned int count = thiz()->tableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (st->get_type () == 1)
+	return true;
+      st = &StructAfter<SubTable> (*st);
+    }
+    return false;
+  }
+
+  bool has_cross_stream () const
+  {
+    typedef typename T::SubTable SubTable;
+
+    const SubTable *st = &thiz()->firstSubTable;
+    unsigned int count = thiz()->tableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (st->u.header.coverage & st->u.header.CrossStream)
+	return true;
+      st = &StructAfter<SubTable> (*st);
+    }
+    return false;
+  }
+
+  int get_h_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+  {
+    typedef typename T::SubTable SubTable;
+
+    int v = 0;
+    const SubTable *st = &thiz()->firstSubTable;
+    unsigned int count = thiz()->tableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if ((st->u.header.coverage & (st->u.header.Variation | st->u.header.CrossStream)) ||
+	  !st->u.header.is_horizontal ())
+	continue;
+      v += st->get_kerning (left, right);
+      st = &StructAfter<SubTable> (*st);
+    }
+    return v;
+  }
+
+  bool apply (AAT::hb_aat_apply_context_t *c) const
+  {
+    typedef typename T::SubTable SubTable;
+
+    bool ret = false;
+    bool seenCrossStream = false;
+    c->set_lookup_index (0);
+    const SubTable *st = &thiz()->firstSubTable;
+    unsigned int count = thiz()->tableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      bool reverse;
+
+      if (!T::Types::extended && (st->u.header.coverage & st->u.header.Variation))
+	goto skip;
+
+      if (HB_DIRECTION_IS_HORIZONTAL (c->buffer->props.direction) != st->u.header.is_horizontal ())
+	goto skip;
+
+      reverse = bool (st->u.header.coverage & st->u.header.Backwards) !=
+		HB_DIRECTION_IS_BACKWARD (c->buffer->props.direction);
+
+      if (!c->buffer->message (c->font, "start subtable %d", c->lookup_index))
+	goto skip;
+
+      if (!seenCrossStream &&
+	  (st->u.header.coverage & st->u.header.CrossStream))
+      {
+	/* Attach all glyphs into a chain. */
+	seenCrossStream = true;
+	hb_glyph_position_t *pos = c->buffer->pos;
+	unsigned int count = c->buffer->len;
+	for (unsigned int i = 0; i < count; i++)
+	{
+	  pos[i].attach_type() = ATTACH_TYPE_CURSIVE;
+	  pos[i].attach_chain() = HB_DIRECTION_IS_FORWARD (c->buffer->props.direction) ? -1 : +1;
+	  /* We intentionally don't set HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT,
+	   * since there needs to be a non-zero attachment for post-positioning to
+	   * be needed. */
+	}
+      }
+
+      if (reverse)
+	c->buffer->reverse ();
+
+      {
+	/* See comment in sanitize() for conditional here. */
+	hb_sanitize_with_object_t with (&c->sanitizer, i < count - 1 ? st : (const SubTable *) nullptr);
+	ret |= st->dispatch (c);
+      }
+
+      if (reverse)
+	c->buffer->reverse ();
+
+      (void) c->buffer->message (c->font, "end subtable %d", c->lookup_index);
+
+    skip:
+      st = &StructAfter<SubTable> (*st);
+      c->set_lookup_index (c->lookup_index + 1);
+    }
+
+    return ret;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!thiz()->version.sanitize (c) ||
+		  (unsigned) thiz()->version < (unsigned) T::minVersion ||
+		  !thiz()->tableCount.sanitize (c)))
+      return_trace (false);
+
+    typedef typename T::SubTable SubTable;
+
+    const SubTable *st = &thiz()->firstSubTable;
+    unsigned int count = thiz()->tableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (unlikely (!st->u.header.sanitize (c)))
+	return_trace (false);
+      /* OpenType kern table has 2-byte subtable lengths.  That's limiting.
+       * MS implementation also only supports one subtable, of format 0,
+       * anyway.  Certain versions of some fonts, like Calibry, contain
+       * kern subtable that exceeds 64kb.  Looks like, the subtable length
+       * is simply ignored.  Which makes sense.  It's only needed if you
+       * have multiple subtables.  To handle such fonts, we just ignore
+       * the length for the last subtable. */
+      hb_sanitize_with_object_t with (c, i < count - 1 ? st : (const SubTable *) nullptr);
+
+      if (unlikely (!st->sanitize (c)))
+	return_trace (false);
+
+      st = &StructAfter<SubTable> (*st);
+    }
+
+    return_trace (true);
+  }
+};
+
+struct kerx : KerxTable<kerx>
+{
+  friend struct KerxTable<kerx>;
+
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_kerx;
+  static constexpr unsigned minVersion = 2u;
+
+  typedef KerxSubTableHeader SubTableHeader;
+  typedef SubTableHeader::Types Types;
+  typedef KerxSubTable SubTable;
+
+  bool has_data () const { return version; }
+
+  protected:
+  HBUINT16	version;	/* The version number of the extended kerning table
+				 * (currently 2, 3, or 4). */
+  HBUINT16	unused;		/* Set to 0. */
+  HBUINT32	tableCount;	/* The number of subtables included in the extended kerning
+				 * table. */
+  SubTable	firstSubTable;	/* Subtables. */
+/*subtableGlyphCoverageArray*/	/* Only if version >= 3. We don't use. */
+
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_KERX_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-morx-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-morx-table.hh
new file mode 100644
index 0000000000..04027a61be
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-morx-table.hh
@@ -0,0 +1,1157 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_LAYOUT_MORX_TABLE_HH
+#define HB_AAT_LAYOUT_MORX_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-aat-layout-common.hh"
+#include "hb-ot-layout-common.hh"
+#include "hb-aat-map.hh"
+
+/*
+ * morx -- Extended Glyph Metamorphosis
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6morx.html
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6mort.html
+ */
+#define HB_AAT_TAG_morx HB_TAG('m','o','r','x')
+#define HB_AAT_TAG_mort HB_TAG('m','o','r','t')
+
+
+namespace AAT {
+
+using namespace OT;
+
+template <typename Types>
+struct RearrangementSubtable
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  typedef void EntryData;
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = true;
+    enum Flags
+    {
+      MarkFirst		= 0x8000,	/* If set, make the current glyph the first
+					 * glyph to be rearranged. */
+      DontAdvance	= 0x4000,	/* If set, don't advance to the next glyph
+					 * before going to the new state. This means
+					 * that the glyph index doesn't change, even
+					 * if the glyph at that index has changed. */
+      MarkLast		= 0x2000,	/* If set, make the current glyph the last
+					 * glyph to be rearranged. */
+      Reserved		= 0x1FF0,	/* These bits are reserved and should be set to 0. */
+      Verb		= 0x000F,	/* The type of rearrangement specified. */
+    };
+
+    driver_context_t (const RearrangementSubtable *table HB_UNUSED) :
+	ret (false),
+	start (0), end (0) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver HB_UNUSED,
+			const Entry<EntryData> &entry)
+    {
+      return (entry.flags & Verb) && start < end;
+    }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+      unsigned int flags = entry.flags;
+
+      if (flags & MarkFirst)
+	start = buffer->idx;
+
+      if (flags & MarkLast)
+	end = hb_min (buffer->idx + 1, buffer->len);
+
+      if ((flags & Verb) && start < end)
+      {
+	/* The following map has two nibbles, for start-side
+	 * and end-side. Values of 0,1,2 mean move that many
+	 * to the other side. Value of 3 means move 2 and
+	 * flip them. */
+	const unsigned char map[16] =
+	{
+	  0x00,	/* 0	no change */
+	  0x10,	/* 1	Ax => xA */
+	  0x01,	/* 2	xD => Dx */
+	  0x11,	/* 3	AxD => DxA */
+	  0x20,	/* 4	ABx => xAB */
+	  0x30,	/* 5	ABx => xBA */
+	  0x02,	/* 6	xCD => CDx */
+	  0x03,	/* 7	xCD => DCx */
+	  0x12,	/* 8	AxCD => CDxA */
+	  0x13,	/* 9	AxCD => DCxA */
+	  0x21,	/* 10	ABxD => DxAB */
+	  0x31,	/* 11	ABxD => DxBA */
+	  0x22,	/* 12	ABxCD => CDxAB */
+	  0x32,	/* 13	ABxCD => CDxBA */
+	  0x23,	/* 14	ABxCD => DCxAB */
+	  0x33,	/* 15	ABxCD => DCxBA */
+	};
+
+	unsigned int m = map[flags & Verb];
+	unsigned int l = hb_min (2u, m >> 4);
+	unsigned int r = hb_min (2u, m & 0x0F);
+	bool reverse_l = 3 == (m >> 4);
+	bool reverse_r = 3 == (m & 0x0F);
+
+	if (end - start >= l + r)
+	{
+	  buffer->merge_clusters (start, hb_min (buffer->idx + 1, buffer->len));
+	  buffer->merge_clusters (start, end);
+
+	  hb_glyph_info_t *info = buffer->info;
+	  hb_glyph_info_t buf[4];
+
+	  memcpy (buf, info + start, l * sizeof (buf[0]));
+	  memcpy (buf + 2, info + end - r, r * sizeof (buf[0]));
+
+	  if (l != r)
+	    memmove (info + start + r, info + start + l, (end - start - l - r) * sizeof (buf[0]));
+
+	  memcpy (info + start, buf + 2, r * sizeof (buf[0]));
+	  memcpy (info + end - l, buf, l * sizeof (buf[0]));
+	  if (reverse_l)
+	  {
+	    buf[0] = info[end - 1];
+	    info[end - 1] = info[end - 2];
+	    info[end - 2] = buf[0];
+	  }
+	  if (reverse_r)
+	  {
+	    buf[0] = info[start];
+	    info[start] = info[start + 1];
+	    info[start + 1] = buf[0];
+	  }
+	}
+      }
+    }
+
+    public:
+    bool ret;
+    private:
+    unsigned int start;
+    unsigned int end;
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    driver_context_t dc (this);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->face);
+    driver.drive (&dc);
+
+    return_trace (dc.ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (machine.sanitize (c));
+  }
+
+  protected:
+  StateTable<Types, EntryData>	machine;
+  public:
+  DEFINE_SIZE_STATIC (16);
+};
+
+template <typename Types>
+struct ContextualSubtable
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  struct EntryData
+  {
+    HBUINT16	markIndex;	/* Index of the substitution table for the
+				 * marked glyph (use 0xFFFF for none). */
+    HBUINT16	currentIndex;	/* Index of the substitution table for the
+				 * current glyph (use 0xFFFF for none). */
+    public:
+    DEFINE_SIZE_STATIC (4);
+  };
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = true;
+    enum Flags
+    {
+      SetMark		= 0x8000,	/* If set, make the current glyph the marked glyph. */
+      DontAdvance	= 0x4000,	/* If set, don't advance to the next glyph before
+					 * going to the new state. */
+      Reserved		= 0x3FFF,	/* These bits are reserved and should be set to 0. */
+    };
+
+    driver_context_t (const ContextualSubtable *table_,
+			     hb_aat_apply_context_t *c_) :
+	ret (false),
+	c (c_),
+	mark_set (false),
+	mark (0),
+	table (table_),
+	subs (table+table->substitutionTables) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver,
+			const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+
+      if (buffer->idx == buffer->len && !mark_set)
+	return false;
+
+      return entry.data.markIndex != 0xFFFF || entry.data.currentIndex != 0xFFFF;
+    }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+
+      /* Looks like CoreText applies neither mark nor current substitution for
+       * end-of-text if mark was not explicitly set. */
+      if (buffer->idx == buffer->len && !mark_set)
+	return;
+
+      const HBGlyphID *replacement;
+
+      replacement = nullptr;
+      if (Types::extended)
+      {
+	if (entry.data.markIndex != 0xFFFF)
+	{
+	  const Lookup<HBGlyphID> &lookup = subs[entry.data.markIndex];
+	  replacement = lookup.get_value (buffer->info[mark].codepoint, driver->num_glyphs);
+	}
+      }
+      else
+      {
+	unsigned int offset = entry.data.markIndex + buffer->info[mark].codepoint;
+	const UnsizedArrayOf<HBGlyphID> &subs_old = (const UnsizedArrayOf<HBGlyphID> &) subs;
+	replacement = &subs_old[Types::wordOffsetToIndex (offset, table, subs_old.arrayZ)];
+	if (!replacement->sanitize (&c->sanitizer) || !*replacement)
+	  replacement = nullptr;
+      }
+      if (replacement)
+      {
+	buffer->unsafe_to_break (mark, hb_min (buffer->idx + 1, buffer->len));
+	buffer->info[mark].codepoint = *replacement;
+	ret = true;
+      }
+
+      replacement = nullptr;
+      unsigned int idx = hb_min (buffer->idx, buffer->len - 1);
+      if (Types::extended)
+      {
+	if (entry.data.currentIndex != 0xFFFF)
+	{
+	  const Lookup<HBGlyphID> &lookup = subs[entry.data.currentIndex];
+	  replacement = lookup.get_value (buffer->info[idx].codepoint, driver->num_glyphs);
+	}
+      }
+      else
+      {
+	unsigned int offset = entry.data.currentIndex + buffer->info[idx].codepoint;
+	const UnsizedArrayOf<HBGlyphID> &subs_old = (const UnsizedArrayOf<HBGlyphID> &) subs;
+	replacement = &subs_old[Types::wordOffsetToIndex (offset, table, subs_old.arrayZ)];
+	if (!replacement->sanitize (&c->sanitizer) || !*replacement)
+	  replacement = nullptr;
+      }
+      if (replacement)
+      {
+	buffer->info[idx].codepoint = *replacement;
+	ret = true;
+      }
+
+      if (entry.flags & SetMark)
+      {
+	mark_set = true;
+	mark = buffer->idx;
+      }
+    }
+
+    public:
+    bool ret;
+    private:
+    hb_aat_apply_context_t *c;
+    bool mark_set;
+    unsigned int mark;
+    const ContextualSubtable *table;
+    const UnsizedOffsetListOf<Lookup<HBGlyphID>, HBUINT, false> &subs;
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    driver_context_t dc (this, c);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->face);
+    driver.drive (&dc);
+
+    return_trace (dc.ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+
+    unsigned int num_entries = 0;
+    if (unlikely (!machine.sanitize (c, &num_entries))) return_trace (false);
+
+    if (!Types::extended)
+      return_trace (substitutionTables.sanitize (c, this, 0));
+
+    unsigned int num_lookups = 0;
+
+    const Entry<EntryData> *entries = machine.get_entries ();
+    for (unsigned int i = 0; i < num_entries; i++)
+    {
+      const EntryData &data = entries[i].data;
+
+      if (data.markIndex != 0xFFFF)
+	num_lookups = hb_max (num_lookups, 1 + data.markIndex);
+      if (data.currentIndex != 0xFFFF)
+	num_lookups = hb_max (num_lookups, 1 + data.currentIndex);
+    }
+
+    return_trace (substitutionTables.sanitize (c, this, num_lookups));
+  }
+
+  protected:
+  StateTable<Types, EntryData>
+		machine;
+  NNOffsetTo<UnsizedOffsetListOf<Lookup<HBGlyphID>, HBUINT, false>, HBUINT>
+		substitutionTables;
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+
+template <bool extended>
+struct LigatureEntry;
+
+template <>
+struct LigatureEntry<true>
+{
+  enum Flags
+  {
+    SetComponent	= 0x8000,	/* Push this glyph onto the component stack for
+					 * eventual processing. */
+    DontAdvance		= 0x4000,	/* Leave the glyph pointer at this glyph for the
+					   next iteration. */
+    PerformAction	= 0x2000,	/* Use the ligActionIndex to process a ligature
+					 * group. */
+    Reserved		= 0x1FFF,	/* These bits are reserved and should be set to 0. */
+  };
+
+  struct EntryData
+  {
+    HBUINT16	ligActionIndex;	/* Index to the first ligActionTable entry
+				 * for processing this group, if indicated
+				 * by the flags. */
+    public:
+    DEFINE_SIZE_STATIC (2);
+  };
+
+  static bool performAction (const Entry<EntryData> &entry)
+  { return entry.flags & PerformAction; }
+
+  static unsigned int ligActionIndex (const Entry<EntryData> &entry)
+  { return entry.data.ligActionIndex; }
+};
+template <>
+struct LigatureEntry<false>
+{
+  enum Flags
+  {
+    SetComponent	= 0x8000,	/* Push this glyph onto the component stack for
+					 * eventual processing. */
+    DontAdvance		= 0x4000,	/* Leave the glyph pointer at this glyph for the
+					   next iteration. */
+    Offset		= 0x3FFF,	/* Byte offset from beginning of subtable to the
+					 * ligature action list. This value must be a
+					 * multiple of 4. */
+  };
+
+  typedef void EntryData;
+
+  static bool performAction (const Entry<EntryData> &entry)
+  { return entry.flags & Offset; }
+
+  static unsigned int ligActionIndex (const Entry<EntryData> &entry)
+  { return entry.flags & Offset; }
+};
+
+
+template <typename Types>
+struct LigatureSubtable
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  typedef LigatureEntry<Types::extended> LigatureEntryT;
+  typedef typename LigatureEntryT::EntryData EntryData;
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = false;
+    enum
+    {
+      DontAdvance	= LigatureEntryT::DontAdvance,
+    };
+    enum LigActionFlags
+    {
+      LigActionLast	= 0x80000000,	/* This is the last action in the list. This also
+					 * implies storage. */
+      LigActionStore	= 0x40000000,	/* Store the ligature at the current cumulated index
+					 * in the ligature table in place of the marked
+					 * (i.e. currently-popped) glyph. */
+      LigActionOffset	= 0x3FFFFFFF,	/* A 30-bit value which is sign-extended to 32-bits
+					 * and added to the glyph ID, resulting in an index
+					 * into the component table. */
+    };
+
+    driver_context_t (const LigatureSubtable *table_,
+		      hb_aat_apply_context_t *c_) :
+	ret (false),
+	c (c_),
+	table (table_),
+	ligAction (table+table->ligAction),
+	component (table+table->component),
+	ligature (table+table->ligature),
+	match_length (0) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver HB_UNUSED,
+			const Entry<EntryData> &entry)
+    {
+      return LigatureEntryT::performAction (entry);
+    }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+
+      DEBUG_MSG (APPLY, nullptr, "Ligature transition at %u", buffer->idx);
+      if (entry.flags & LigatureEntryT::SetComponent)
+      {
+	/* Never mark same index twice, in case DontAdvance was used... */
+	if (match_length && match_positions[(match_length - 1u) % ARRAY_LENGTH (match_positions)] == buffer->out_len)
+	  match_length--;
+
+	match_positions[match_length++ % ARRAY_LENGTH (match_positions)] = buffer->out_len;
+	DEBUG_MSG (APPLY, nullptr, "Set component at %u", buffer->out_len);
+      }
+
+      if (LigatureEntryT::performAction (entry))
+      {
+	DEBUG_MSG (APPLY, nullptr, "Perform action with %u", match_length);
+	unsigned int end = buffer->out_len;
+
+	if (unlikely (!match_length))
+	  return;
+
+	if (buffer->idx >= buffer->len)
+	  return; /* TODO Work on previous instead? */
+
+	unsigned int cursor = match_length;
+
+	unsigned int action_idx = LigatureEntryT::ligActionIndex (entry);
+	action_idx = Types::offsetToIndex (action_idx, table, ligAction.arrayZ);
+	const HBUINT32 *actionData = &ligAction[action_idx];
+
+	unsigned int ligature_idx = 0;
+	unsigned int action;
+	do
+	{
+	  if (unlikely (!cursor))
+	  {
+	    /* Stack underflow.  Clear the stack. */
+	    DEBUG_MSG (APPLY, nullptr, "Stack underflow");
+	    match_length = 0;
+	    break;
+	  }
+
+	  DEBUG_MSG (APPLY, nullptr, "Moving to stack position %u", cursor - 1);
+	  buffer->move_to (match_positions[--cursor % ARRAY_LENGTH (match_positions)]);
+
+	  if (unlikely (!actionData->sanitize (&c->sanitizer))) break;
+	  action = *actionData;
+
+	  uint32_t uoffset = action & LigActionOffset;
+	  if (uoffset & 0x20000000)
+	    uoffset |= 0xC0000000; /* Sign-extend. */
+	  int32_t offset = (int32_t) uoffset;
+	  unsigned int component_idx = buffer->cur().codepoint + offset;
+	  component_idx = Types::wordOffsetToIndex (component_idx, table, component.arrayZ);
+	  const HBUINT16 &componentData = component[component_idx];
+	  if (unlikely (!componentData.sanitize (&c->sanitizer))) break;
+	  ligature_idx += componentData;
+
+	  DEBUG_MSG (APPLY, nullptr, "Action store %u last %u",
+		     bool (action & LigActionStore),
+		     bool (action & LigActionLast));
+	  if (action & (LigActionStore | LigActionLast))
+	  {
+	    ligature_idx = Types::offsetToIndex (ligature_idx, table, ligature.arrayZ);
+	    const HBGlyphID &ligatureData = ligature[ligature_idx];
+	    if (unlikely (!ligatureData.sanitize (&c->sanitizer))) break;
+	    hb_codepoint_t lig = ligatureData;
+
+	    DEBUG_MSG (APPLY, nullptr, "Produced ligature %u", lig);
+	    buffer->replace_glyph (lig);
+
+	    unsigned int lig_end = match_positions[(match_length - 1u) % ARRAY_LENGTH (match_positions)] + 1u;
+	    /* Now go and delete all subsequent components. */
+	    while (match_length - 1u > cursor)
+	    {
+	      DEBUG_MSG (APPLY, nullptr, "Skipping ligature component");
+	      buffer->move_to (match_positions[--match_length % ARRAY_LENGTH (match_positions)]);
+	      buffer->replace_glyph (DELETED_GLYPH);
+	    }
+
+	    buffer->move_to (lig_end);
+	    buffer->merge_out_clusters (match_positions[cursor % ARRAY_LENGTH (match_positions)], buffer->out_len);
+	  }
+
+	  actionData++;
+	}
+	while (!(action & LigActionLast));
+	buffer->move_to (end);
+      }
+    }
+
+    public:
+    bool ret;
+    private:
+    hb_aat_apply_context_t *c;
+    const LigatureSubtable *table;
+    const UnsizedArrayOf<HBUINT32> &ligAction;
+    const UnsizedArrayOf<HBUINT16> &component;
+    const UnsizedArrayOf<HBGlyphID> &ligature;
+    unsigned int match_length;
+    unsigned int match_positions[HB_MAX_CONTEXT_LENGTH];
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    driver_context_t dc (this, c);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->face);
+    driver.drive (&dc);
+
+    return_trace (dc.ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    /* The rest of array sanitizations are done at run-time. */
+    return_trace (c->check_struct (this) && machine.sanitize (c) &&
+		  ligAction && component && ligature);
+  }
+
+  protected:
+  StateTable<Types, EntryData>
+		machine;
+  NNOffsetTo<UnsizedArrayOf<HBUINT32>, HBUINT>
+		ligAction;	/* Offset to the ligature action table. */
+  NNOffsetTo<UnsizedArrayOf<HBUINT16>, HBUINT>
+		component;	/* Offset to the component table. */
+  NNOffsetTo<UnsizedArrayOf<HBGlyphID>, HBUINT>
+		ligature;	/* Offset to the actual ligature lists. */
+  public:
+  DEFINE_SIZE_STATIC (28);
+};
+
+template <typename Types>
+struct NoncontextualSubtable
+{
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    bool ret = false;
+    unsigned int num_glyphs = c->face->get_num_glyphs ();
+
+    hb_glyph_info_t *info = c->buffer->info;
+    unsigned int count = c->buffer->len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const HBGlyphID *replacement = substitute.get_value (info[i].codepoint, num_glyphs);
+      if (replacement)
+      {
+	info[i].codepoint = *replacement;
+	ret = true;
+      }
+    }
+
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (substitute.sanitize (c));
+  }
+
+  protected:
+  Lookup<HBGlyphID>	substitute;
+  public:
+  DEFINE_SIZE_MIN (2);
+};
+
+template <typename Types>
+struct InsertionSubtable
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  struct EntryData
+  {
+    HBUINT16	currentInsertIndex;	/* Zero-based index into the insertion glyph table.
+					 * The number of glyphs to be inserted is contained
+					 * in the currentInsertCount field in the flags.
+					 * A value of 0xFFFF indicates no insertion is to
+					 * be done. */
+    HBUINT16	markedInsertIndex;	/* Zero-based index into the insertion glyph table.
+					 * The number of glyphs to be inserted is contained
+					 * in the markedInsertCount field in the flags.
+					 * A value of 0xFFFF indicates no insertion is to
+					 * be done. */
+    public:
+    DEFINE_SIZE_STATIC (4);
+  };
+
+  struct driver_context_t
+  {
+    static constexpr bool in_place = false;
+    enum Flags
+    {
+      SetMark		= 0x8000,	/* If set, mark the current glyph. */
+      DontAdvance	= 0x4000,	/* If set, don't advance to the next glyph before
+					 * going to the new state.  This does not mean
+					 * that the glyph pointed to is the same one as
+					 * before. If you've made insertions immediately
+					 * downstream of the current glyph, the next glyph
+					 * processed would in fact be the first one
+					 * inserted. */
+      CurrentIsKashidaLike= 0x2000,	/* If set, and the currentInsertList is nonzero,
+					 * then the specified glyph list will be inserted
+					 * as a kashida-like insertion, either before or
+					 * after the current glyph (depending on the state
+					 * of the currentInsertBefore flag). If clear, and
+					 * the currentInsertList is nonzero, then the
+					 * specified glyph list will be inserted as a
+					 * split-vowel-like insertion, either before or
+					 * after the current glyph (depending on the state
+					 * of the currentInsertBefore flag). */
+      MarkedIsKashidaLike= 0x1000,	/* If set, and the markedInsertList is nonzero,
+					 * then the specified glyph list will be inserted
+					 * as a kashida-like insertion, either before or
+					 * after the marked glyph (depending on the state
+					 * of the markedInsertBefore flag). If clear, and
+					 * the markedInsertList is nonzero, then the
+					 * specified glyph list will be inserted as a
+					 * split-vowel-like insertion, either before or
+					 * after the marked glyph (depending on the state
+					 * of the markedInsertBefore flag). */
+      CurrentInsertBefore= 0x0800,	/* If set, specifies that insertions are to be made
+					 * to the left of the current glyph. If clear,
+					 * they're made to the right of the current glyph. */
+      MarkedInsertBefore= 0x0400,	/* If set, specifies that insertions are to be
+					 * made to the left of the marked glyph. If clear,
+					 * they're made to the right of the marked glyph. */
+      CurrentInsertCount= 0x3E0,	/* This 5-bit field is treated as a count of the
+					 * number of glyphs to insert at the current
+					 * position. Since zero means no insertions, the
+					 * largest number of insertions at any given
+					 * current location is 31 glyphs. */
+      MarkedInsertCount= 0x001F,	/* This 5-bit field is treated as a count of the
+					 * number of glyphs to insert at the marked
+					 * position. Since zero means no insertions, the
+					 * largest number of insertions at any given
+					 * marked location is 31 glyphs. */
+    };
+
+    driver_context_t (const InsertionSubtable *table,
+		      hb_aat_apply_context_t *c_) :
+	ret (false),
+	c (c_),
+	mark (0),
+	insertionAction (table+table->insertionAction) {}
+
+    bool is_actionable (StateTableDriver<Types, EntryData> *driver HB_UNUSED,
+			const Entry<EntryData> &entry)
+    {
+      return (entry.flags & (CurrentInsertCount | MarkedInsertCount)) &&
+	     (entry.data.currentInsertIndex != 0xFFFF ||entry.data.markedInsertIndex != 0xFFFF);
+    }
+    void transition (StateTableDriver<Types, EntryData> *driver,
+		     const Entry<EntryData> &entry)
+    {
+      hb_buffer_t *buffer = driver->buffer;
+      unsigned int flags = entry.flags;
+
+      unsigned mark_loc = buffer->out_len;
+
+      if (entry.data.markedInsertIndex != 0xFFFF)
+      {
+	unsigned int count = (flags & MarkedInsertCount);
+	if (unlikely ((buffer->max_ops -= count) <= 0)) return;
+	unsigned int start = entry.data.markedInsertIndex;
+	const HBGlyphID *glyphs = &insertionAction[start];
+	if (unlikely (!c->sanitizer.check_array (glyphs, count))) count = 0;
+
+	bool before = flags & MarkedInsertBefore;
+
+	unsigned int end = buffer->out_len;
+	buffer->move_to (mark);
+
+	if (buffer->idx < buffer->len && !before)
+	  buffer->copy_glyph ();
+	/* TODO We ignore KashidaLike setting. */
+	for (unsigned int i = 0; i < count; i++)
+	  buffer->output_glyph (glyphs[i]);
+	if (buffer->idx < buffer->len && !before)
+	  buffer->skip_glyph ();
+
+	buffer->move_to (end + count);
+
+	buffer->unsafe_to_break_from_outbuffer (mark, hb_min (buffer->idx + 1, buffer->len));
+      }
+
+      if (flags & SetMark)
+	mark = mark_loc;
+
+      if (entry.data.currentInsertIndex != 0xFFFF)
+      {
+	unsigned int count = (flags & CurrentInsertCount) >> 5;
+	if (unlikely ((buffer->max_ops -= count) <= 0)) return;
+	unsigned int start = entry.data.currentInsertIndex;
+	const HBGlyphID *glyphs = &insertionAction[start];
+	if (unlikely (!c->sanitizer.check_array (glyphs, count))) count = 0;
+
+	bool before = flags & CurrentInsertBefore;
+
+	unsigned int end = buffer->out_len;
+
+	if (buffer->idx < buffer->len && !before)
+	  buffer->copy_glyph ();
+	/* TODO We ignore KashidaLike setting. */
+	for (unsigned int i = 0; i < count; i++)
+	  buffer->output_glyph (glyphs[i]);
+	if (buffer->idx < buffer->len && !before)
+	  buffer->skip_glyph ();
+
+	/* Humm. Not sure where to move to.  There's this wording under
+	 * DontAdvance flag:
+	 *
+	 * "If set, don't update the glyph index before going to the new state.
+	 * This does not mean that the glyph pointed to is the same one as
+	 * before. If you've made insertions immediately downstream of the
+	 * current glyph, the next glyph processed would in fact be the first
+	 * one inserted."
+	 *
+	 * This suggests that if DontAdvance is NOT set, we should move to
+	 * end+count.  If it *was*, then move to end, such that newly inserted
+	 * glyphs are now visible.
+	 *
+	 * https://github.com/harfbuzz/harfbuzz/issues/1224#issuecomment-427691417
+	 */
+	buffer->move_to ((flags & DontAdvance) ? end : end + count);
+      }
+    }
+
+    public:
+    bool ret;
+    private:
+    hb_aat_apply_context_t *c;
+    unsigned int mark;
+    const UnsizedArrayOf<HBGlyphID> &insertionAction;
+  };
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    driver_context_t dc (this, c);
+
+    StateTableDriver<Types, EntryData> driver (machine, c->buffer, c->face);
+    driver.drive (&dc);
+
+    return_trace (dc.ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    /* The rest of array sanitizations are done at run-time. */
+    return_trace (c->check_struct (this) && machine.sanitize (c) &&
+		  insertionAction);
+  }
+
+  protected:
+  StateTable<Types, EntryData>
+		machine;
+  NNOffsetTo<UnsizedArrayOf<HBGlyphID>, HBUINT>
+		insertionAction;	/* Byte offset from stateHeader to the start of
+					 * the insertion glyph table. */
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+
+struct Feature
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	featureType;	/* The type of feature. */
+  HBUINT16	featureSetting;	/* The feature's setting (aka selector). */
+  HBUINT32	enableFlags;	/* Flags for the settings that this feature
+				 * and setting enables. */
+  HBUINT32	disableFlags;	/* Complement of flags for the settings that this
+				 * feature and setting disable. */
+
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+template <typename Types>
+struct ChainSubtable
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  template <typename T>
+  friend struct Chain;
+
+  unsigned int get_size () const     { return length; }
+  unsigned int get_type () const     { return coverage & 0xFF; }
+  unsigned int get_coverage () const { return coverage >> (sizeof (HBUINT) * 8 - 8); }
+
+  enum Coverage
+  {
+    Vertical		= 0x80,	/* If set, this subtable will only be applied
+				 * to vertical text. If clear, this subtable
+				 * will only be applied to horizontal text. */
+    Backwards		= 0x40,	/* If set, this subtable will process glyphs
+				 * in descending order. If clear, it will
+				 * process the glyphs in ascending order. */
+    AllDirections	= 0x20,	/* If set, this subtable will be applied to
+				 * both horizontal and vertical text (i.e.
+				 * the state of bit 0x80000000 is ignored). */
+    Logical		= 0x10,	/* If set, this subtable will process glyphs
+				 * in logical order (or reverse logical order,
+				 * depending on the value of bit 0x80000000). */
+  };
+  enum Type
+  {
+    Rearrangement	= 0,
+    Contextual		= 1,
+    Ligature		= 2,
+    Noncontextual	= 4,
+    Insertion		= 5
+  };
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    unsigned int subtable_type = get_type ();
+    TRACE_DISPATCH (this, subtable_type);
+    switch (subtable_type) {
+    case Rearrangement:		return_trace (c->dispatch (u.rearrangement, hb_forward<Ts> (ds)...));
+    case Contextual:		return_trace (c->dispatch (u.contextual, hb_forward<Ts> (ds)...));
+    case Ligature:		return_trace (c->dispatch (u.ligature, hb_forward<Ts> (ds)...));
+    case Noncontextual:		return_trace (c->dispatch (u.noncontextual, hb_forward<Ts> (ds)...));
+    case Insertion:		return_trace (c->dispatch (u.insertion, hb_forward<Ts> (ds)...));
+    default:			return_trace (c->default_return_value ());
+    }
+  }
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_sanitize_with_object_t with (&c->sanitizer, this);
+    return_trace (dispatch (c));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!length.sanitize (c) ||
+	length <= min_size ||
+	!c->check_range (this, length))
+      return_trace (false);
+
+    hb_sanitize_with_object_t with (c, this);
+    return_trace (dispatch (c));
+  }
+
+  protected:
+  HBUINT	length;		/* Total subtable length, including this header. */
+  HBUINT	coverage;	/* Coverage flags and subtable type. */
+  HBUINT32	subFeatureFlags;/* The 32-bit mask identifying which subtable this is. */
+  union {
+  RearrangementSubtable<Types>	rearrangement;
+  ContextualSubtable<Types>	contextual;
+  LigatureSubtable<Types>	ligature;
+  NoncontextualSubtable<Types>	noncontextual;
+  InsertionSubtable<Types>	insertion;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (2 * sizeof (HBUINT) + 4);
+};
+
+template <typename Types>
+struct Chain
+{
+  typedef typename Types::HBUINT HBUINT;
+
+  hb_mask_t compile_flags (const hb_aat_map_builder_t *map) const
+  {
+    hb_mask_t flags = defaultFlags;
+    {
+      unsigned int count = featureCount;
+      for (unsigned i = 0; i < count; i++)
+      {
+	const Feature &feature = featureZ[i];
+	hb_aat_layout_feature_type_t type = (hb_aat_layout_feature_type_t) (unsigned int) feature.featureType;
+	hb_aat_layout_feature_selector_t setting = (hb_aat_layout_feature_selector_t) (unsigned int) feature.featureSetting;
+      retry:
+	// Check whether this type/setting pair was requested in the map, and if so, apply its flags.
+	// (The search here only looks at the type and setting fields of feature_info_t.)
+	hb_aat_map_builder_t::feature_info_t info = { type, setting, false, 0 };
+	if (map->features.bsearch (info))
+	{
+	  flags &= feature.disableFlags;
+	  flags |= feature.enableFlags;
+	}
+	else if (type == HB_AAT_LAYOUT_FEATURE_TYPE_LETTER_CASE && setting == HB_AAT_LAYOUT_FEATURE_SELECTOR_SMALL_CAPS)
+	{
+	  /* Deprecated. https://github.com/harfbuzz/harfbuzz/issues/1342 */
+	  type = HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE;
+	  setting = HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_SMALL_CAPS;
+	  goto retry;
+	}
+      }
+    }
+    return flags;
+  }
+
+  void apply (hb_aat_apply_context_t *c,
+	      hb_mask_t flags) const
+  {
+    const ChainSubtable<Types> *subtable = &StructAfter<ChainSubtable<Types>> (featureZ.as_array (featureCount));
+    unsigned int count = subtableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      bool reverse;
+
+      if (!(subtable->subFeatureFlags & flags))
+	goto skip;
+
+      if (!(subtable->get_coverage() & ChainSubtable<Types>::AllDirections) &&
+	  HB_DIRECTION_IS_VERTICAL (c->buffer->props.direction) !=
+	  bool (subtable->get_coverage() & ChainSubtable<Types>::Vertical))
+	goto skip;
+
+      /* Buffer contents is always in logical direction.  Determine if
+       * we need to reverse before applying this subtable.  We reverse
+       * back after if we did reverse indeed.
+       *
+       * Quoting the spac:
+       * """
+       * Bits 28 and 30 of the coverage field control the order in which
+       * glyphs are processed when the subtable is run by the layout engine.
+       * Bit 28 is used to indicate if the glyph processing direction is
+       * the same as logical order or layout order. Bit 30 is used to
+       * indicate whether glyphs are processed forwards or backwards within
+       * that order.
+
+		Bit 30	Bit 28	Interpretation for Horizontal Text
+		0	0	The subtable is processed in layout order
+				(the same order as the glyphs, which is
+				always left-to-right).
+		1	0	The subtable is processed in reverse layout order
+				(the order opposite that of the glyphs, which is
+				always right-to-left).
+		0	1	The subtable is processed in logical order
+				(the same order as the characters, which may be
+				left-to-right or right-to-left).
+		1	1	The subtable is processed in reverse logical order
+				(the order opposite that of the characters, which
+				may be right-to-left or left-to-right).
+       */
+      reverse = subtable->get_coverage () & ChainSubtable<Types>::Logical ?
+		bool (subtable->get_coverage () & ChainSubtable<Types>::Backwards) :
+		bool (subtable->get_coverage () & ChainSubtable<Types>::Backwards) !=
+		HB_DIRECTION_IS_BACKWARD (c->buffer->props.direction);
+
+      if (!c->buffer->message (c->font, "start chainsubtable %d", c->lookup_index))
+	goto skip;
+
+      if (reverse)
+	c->buffer->reverse ();
+
+      subtable->apply (c);
+
+      if (reverse)
+	c->buffer->reverse ();
+
+      (void) c->buffer->message (c->font, "end chainsubtable %d", c->lookup_index);
+
+      if (unlikely (!c->buffer->successful)) return;
+
+    skip:
+      subtable = &StructAfter<ChainSubtable<Types>> (*subtable);
+      c->set_lookup_index (c->lookup_index + 1);
+    }
+  }
+
+  unsigned int get_size () const { return length; }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int version HB_UNUSED) const
+  {
+    TRACE_SANITIZE (this);
+    if (!length.sanitize (c) ||
+	length < min_size ||
+	!c->check_range (this, length))
+      return_trace (false);
+
+    if (!c->check_array (featureZ.arrayZ, featureCount))
+      return_trace (false);
+
+    const ChainSubtable<Types> *subtable = &StructAfter<ChainSubtable<Types>> (featureZ.as_array (featureCount));
+    unsigned int count = subtableCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (!subtable->sanitize (c))
+	return_trace (false);
+      subtable = &StructAfter<ChainSubtable<Types>> (*subtable);
+    }
+
+    return_trace (true);
+  }
+
+  protected:
+  HBUINT32	defaultFlags;	/* The default specification for subtables. */
+  HBUINT32	length;		/* Total byte count, including this header. */
+  HBUINT	featureCount;	/* Number of feature subtable entries. */
+  HBUINT	subtableCount;	/* The number of subtables in the chain. */
+
+  UnsizedArrayOf<Feature>	featureZ;	/* Features. */
+/*ChainSubtable	firstSubtable;*//* Subtables. */
+/*subtableGlyphCoverageArray*/	/* Only if version >= 3. We don't use. */
+
+  public:
+  DEFINE_SIZE_MIN (8 + 2 * sizeof (HBUINT));
+};
+
+
+/*
+ * The 'mort'/'morx' Table
+ */
+
+template <typename Types, hb_tag_t TAG>
+struct mortmorx
+{
+  static constexpr hb_tag_t tableTag = TAG;
+
+  bool has_data () const { return version != 0; }
+
+  void compile_flags (const hb_aat_map_builder_t *mapper,
+		      hb_aat_map_t *map) const
+  {
+    const Chain<Types> *chain = &firstChain;
+    unsigned int count = chainCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      map->chain_flags.push (chain->compile_flags (mapper));
+      chain = &StructAfter<Chain<Types>> (*chain);
+    }
+  }
+
+  void apply (hb_aat_apply_context_t *c) const
+  {
+    if (unlikely (!c->buffer->successful)) return;
+    c->set_lookup_index (0);
+    const Chain<Types> *chain = &firstChain;
+    unsigned int count = chainCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      chain->apply (c, c->plan->aat_map.chain_flags[i]);
+      if (unlikely (!c->buffer->successful)) return;
+      chain = &StructAfter<Chain<Types>> (*chain);
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!version.sanitize (c) || !version || !chainCount.sanitize (c))
+      return_trace (false);
+
+    const Chain<Types> *chain = &firstChain;
+    unsigned int count = chainCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (!chain->sanitize (c, version))
+	return_trace (false);
+      chain = &StructAfter<Chain<Types>> (*chain);
+    }
+
+    return_trace (true);
+  }
+
+  protected:
+  HBUINT16	version;	/* Version number of the glyph metamorphosis table.
+				 * 1, 2, or 3. */
+  HBUINT16	unused;		/* Set to 0. */
+  HBUINT32	chainCount;	/* Number of metamorphosis chains contained in this
+				 * table. */
+  Chain<Types>	firstChain;	/* Chains. */
+
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+struct morx : mortmorx<ExtendedTypes, HB_AAT_TAG_morx> {};
+struct mort : mortmorx<ObsoleteTypes, HB_AAT_TAG_mort> {};
+
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_MORX_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-opbd-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-opbd-table.hh
new file mode 100644
index 0000000000..8c04a6482f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-opbd-table.hh
@@ -0,0 +1,173 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LAYOUT_OPBD_TABLE_HH
+#define HB_AAT_LAYOUT_OPBD_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+#include "hb-open-type.hh"
+
+/*
+ * opbd -- Optical Bounds
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6opbd.html
+ */
+#define HB_AAT_TAG_opbd HB_TAG('o','p','b','d')
+
+
+namespace AAT {
+
+struct OpticalBounds
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  FWORD		leftSide;
+  FWORD		topSide;
+  FWORD		rightSide;
+  FWORD		bottomSide;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct opbdFormat0
+{
+  bool get_bounds (hb_font_t *font, hb_codepoint_t glyph_id,
+		   hb_glyph_extents_t *extents, const void *base) const
+  {
+    const OffsetTo<OpticalBounds> *bounds_offset = lookupTable.get_value (glyph_id, font->face->get_num_glyphs ());
+    if (!bounds_offset) return false;
+    const OpticalBounds &bounds = base+*bounds_offset;
+
+    if (extents)
+      *extents = {
+	font->em_scale_x (bounds.leftSide),
+	font->em_scale_y (bounds.topSide),
+	font->em_scale_x (bounds.rightSide),
+	font->em_scale_y (bounds.bottomSide)
+      };
+    return true;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) && lookupTable.sanitize (c, base)));
+  }
+
+  protected:
+  Lookup<OffsetTo<OpticalBounds>>
+		lookupTable;	/* Lookup table associating glyphs with the four
+				 * int16 values for the left-side, top-side,
+				 * right-side, and bottom-side optical bounds. */
+  public:
+  DEFINE_SIZE_MIN (2);
+};
+
+struct opbdFormat1
+{
+  bool get_bounds (hb_font_t *font, hb_codepoint_t glyph_id,
+		   hb_glyph_extents_t *extents, const void *base) const
+  {
+    const OffsetTo<OpticalBounds> *bounds_offset = lookupTable.get_value (glyph_id, font->face->get_num_glyphs ());
+    if (!bounds_offset) return false;
+    const OpticalBounds &bounds = base+*bounds_offset;
+
+    hb_position_t left = 0, top = 0, right = 0, bottom = 0, ignore;
+    if (font->get_glyph_contour_point (glyph_id, bounds.leftSide, &left, &ignore) ||
+	font->get_glyph_contour_point (glyph_id, bounds.topSide, &ignore, &top) ||
+	font->get_glyph_contour_point (glyph_id, bounds.rightSide, &right, &ignore) ||
+	font->get_glyph_contour_point (glyph_id, bounds.bottomSide, &ignore, &bottom))
+    {
+      if (extents)
+	*extents = {left, top, right, bottom};
+      return true;
+    }
+    return false;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) && lookupTable.sanitize (c, base)));
+  }
+
+  protected:
+  Lookup<OffsetTo<OpticalBounds>>
+		lookupTable;	/* Lookup table associating glyphs with the four
+				 * int16 values for the left-side, top-side,
+				 * right-side, and bottom-side optical bounds. */
+  public:
+  DEFINE_SIZE_MIN (2);
+};
+
+struct opbd
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_opbd;
+
+  bool get_bounds (hb_font_t *font, hb_codepoint_t glyph_id,
+		   hb_glyph_extents_t *extents) const
+  {
+    switch (format)
+    {
+    case 0: return u.format0.get_bounds (font, glyph_id, extents, this);
+    case 1: return u.format1.get_bounds (font, glyph_id, extents, this);
+    default:return false;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this) || version.major != 1))
+      return_trace (false);
+
+    switch (format)
+    {
+    case 0: return_trace (u.format0.sanitize (c, this));
+    case 1: return_trace (u.format1.sanitize (c, this));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version number of the optical bounds
+				 * table (0x00010000 for the current version). */
+  HBUINT16	format;		/* Format of the optical bounds table.
+				 * Format 0 indicates distance and Format 1 indicates
+				 * control point. */
+  union {
+  opbdFormat0	format0;
+  opbdFormat1	format1;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_OPBD_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout-trak-table.hh b/thirdparty/harfbuzz/src/hb-aat-layout-trak-table.hh
new file mode 100644
index 0000000000..baa1c72020
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout-trak-table.hh
@@ -0,0 +1,230 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_LAYOUT_TRAK_TABLE_HH
+#define HB_AAT_LAYOUT_TRAK_TABLE_HH
+
+#include "hb-aat-layout-common.hh"
+#include "hb-ot-layout.hh"
+#include "hb-open-type.hh"
+
+/*
+ * trak -- Tracking
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6trak.html
+ */
+#define HB_AAT_TAG_trak HB_TAG('t','r','a','k')
+
+
+namespace AAT {
+
+
+struct TrackTableEntry
+{
+  friend struct TrackData;
+
+  float get_track_value () const { return track.to_float (); }
+
+  int get_value (const void *base, unsigned int index,
+		 unsigned int table_size) const
+  { return (base+valuesZ).as_array (table_size)[index]; }
+
+  public:
+  bool sanitize (hb_sanitize_context_t *c, const void *base,
+		 unsigned int table_size) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (valuesZ.sanitize (c, base, table_size))));
+  }
+
+  protected:
+  HBFixed	track;		/* Track value for this record. */
+  NameID	trackNameID;	/* The 'name' table index for this track.
+				 * (a short word or phrase like "loose"
+				 * or "very tight") */
+  NNOffsetTo<UnsizedArrayOf<FWORD>>
+		valuesZ;	/* Offset from start of tracking table to
+				 * per-size tracking values for this track. */
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct TrackData
+{
+  float interpolate_at (unsigned int idx,
+			float target_size,
+			const TrackTableEntry &trackTableEntry,
+			const void *base) const
+  {
+    unsigned int sizes = nSizes;
+    hb_array_t<const HBFixed> size_table ((base+sizeTable).arrayZ, sizes);
+
+    float s0 = size_table[idx].to_float ();
+    float s1 = size_table[idx + 1].to_float ();
+    float t = unlikely (s0 == s1) ? 0.f : (target_size - s0) / (s1 - s0);
+    return t * trackTableEntry.get_value (base, idx + 1, sizes) +
+	   (1.f - t) * trackTableEntry.get_value (base, idx, sizes);
+  }
+
+  int get_tracking (const void *base, float ptem) const
+  {
+    /*
+     * Choose track.
+     */
+    const TrackTableEntry *trackTableEntry = nullptr;
+    unsigned int count = nTracks;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      /* Note: Seems like the track entries are sorted by values.  But the
+       * spec doesn't explicitly say that.  It just mentions it in the example. */
+
+      /* For now we only seek for track entries with zero tracking value */
+
+      if (trackTable[i].get_track_value () == 0.f)
+      {
+	trackTableEntry = &trackTable[i];
+	break;
+      }
+    }
+    if (!trackTableEntry) return 0.;
+
+    /*
+     * Choose size.
+     */
+    unsigned int sizes = nSizes;
+    if (!sizes) return 0.;
+    if (sizes == 1) return trackTableEntry->get_value (base, 0, sizes);
+
+    hb_array_t<const HBFixed> size_table ((base+sizeTable).arrayZ, sizes);
+    unsigned int size_index;
+    for (size_index = 0; size_index < sizes - 1; size_index++)
+      if (size_table[size_index].to_float () >= ptem)
+	break;
+
+    return roundf (interpolate_at (size_index ? size_index - 1 : 0, ptem,
+				   *trackTableEntry, base));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  sizeTable.sanitize (c, base, nSizes) &&
+			  trackTable.sanitize (c, nTracks, base, nSizes)));
+  }
+
+  protected:
+  HBUINT16	nTracks;	/* Number of separate tracks included in this table. */
+  HBUINT16	nSizes;		/* Number of point sizes included in this table. */
+  LNNOffsetTo<UnsizedArrayOf<HBFixed>>
+		sizeTable;	/* Offset from start of the tracking table to
+				 * Array[nSizes] of size values.. */
+  UnsizedArrayOf<TrackTableEntry>
+		trackTable;	/* Array[nTracks] of TrackTableEntry records. */
+
+  public:
+  DEFINE_SIZE_ARRAY (8, trackTable);
+};
+
+struct trak
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_trak;
+
+  bool has_data () const { return version.to_int (); }
+
+  bool apply (hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    hb_mask_t trak_mask = c->plan->trak_mask;
+
+    const float ptem = c->font->ptem;
+    if (unlikely (ptem <= 0.f))
+      return_trace (false);
+
+    hb_buffer_t *buffer = c->buffer;
+    if (HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction))
+    {
+      const TrackData &trackData = this+horizData;
+      int tracking = trackData.get_tracking (this, ptem);
+      hb_position_t offset_to_add = c->font->em_scalef_x (tracking / 2);
+      hb_position_t advance_to_add = c->font->em_scalef_x (tracking);
+      foreach_grapheme (buffer, start, end)
+      {
+	if (!(buffer->info[start].mask & trak_mask)) continue;
+	buffer->pos[start].x_advance += advance_to_add;
+	buffer->pos[start].x_offset += offset_to_add;
+      }
+    }
+    else
+    {
+      const TrackData &trackData = this+vertData;
+      int tracking = trackData.get_tracking (this, ptem);
+      hb_position_t offset_to_add = c->font->em_scalef_y (tracking / 2);
+      hb_position_t advance_to_add = c->font->em_scalef_y (tracking);
+      foreach_grapheme (buffer, start, end)
+      {
+	if (!(buffer->info[start].mask & trak_mask)) continue;
+	buffer->pos[start].y_advance += advance_to_add;
+	buffer->pos[start].y_offset += offset_to_add;
+      }
+    }
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+
+    return_trace (likely (c->check_struct (this) &&
+			  version.major == 1 &&
+			  horizData.sanitize (c, this, this) &&
+			  vertData.sanitize (c, this, this)));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the tracking table
+				 * (0x00010000u for version 1.0). */
+  HBUINT16	format;		/* Format of the tracking table (set to 0). */
+  OffsetTo<TrackData>
+		horizData;	/* Offset from start of tracking table to TrackData
+				 * for horizontal text (or 0 if none). */
+  OffsetTo<TrackData>
+		vertData;	/* Offset from start of tracking table to TrackData
+				 * for vertical text (or 0 if none). */
+  HBUINT16	reserved;	/* Reserved. Set to 0. */
+
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LAYOUT_TRAK_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout.cc b/thirdparty/harfbuzz/src/hb-aat-layout.cc
new file mode 100644
index 0000000000..fac510e9e6
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout.cc
@@ -0,0 +1,382 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-aat-layout.hh"
+#include "hb-aat-layout-ankr-table.hh"
+#include "hb-aat-layout-bsln-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-aat-layout-feat-table.hh"
+#include "hb-aat-layout-just-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-aat-layout-kerx-table.hh"
+#include "hb-aat-layout-morx-table.hh"
+#include "hb-aat-layout-trak-table.hh"
+#include "hb-aat-ltag-table.hh"
+
+
+/*
+ * hb_aat_apply_context_t
+ */
+
+/* Note: This context is used for kerning, even without AAT, hence the condition. */
+#if !defined(HB_NO_AAT) || !defined(HB_NO_OT_KERN)
+
+AAT::hb_aat_apply_context_t::hb_aat_apply_context_t (const hb_ot_shape_plan_t *plan_,
+						     hb_font_t *font_,
+						     hb_buffer_t *buffer_,
+						     hb_blob_t *blob) :
+						       plan (plan_),
+						       font (font_),
+						       face (font->face),
+						       buffer (buffer_),
+						       sanitizer (),
+						       ankr_table (&Null (AAT::ankr)),
+						       lookup_index (0)
+{
+  sanitizer.init (blob);
+  sanitizer.set_num_glyphs (face->get_num_glyphs ());
+  sanitizer.start_processing ();
+  sanitizer.set_max_ops (HB_SANITIZE_MAX_OPS_MAX);
+}
+
+AAT::hb_aat_apply_context_t::~hb_aat_apply_context_t ()
+{ sanitizer.end_processing (); }
+
+void
+AAT::hb_aat_apply_context_t::set_ankr_table (const AAT::ankr *ankr_table_)
+{ ankr_table = ankr_table_; }
+
+#endif
+
+
+/**
+ * SECTION:hb-aat-layout
+ * @title: hb-aat-layout
+ * @short_description: Apple Advanced Typography Layout
+ * @include: hb-aat.h
+ *
+ * Functions for querying OpenType Layout features in the font face.
+ **/
+
+
+#if !defined(HB_NO_AAT) || defined(HAVE_CORETEXT)
+
+/* Table data courtesy of Apple.  Converted from mnemonics to integers
+ * when moving to this file. */
+static const hb_aat_feature_mapping_t feature_mappings[] =
+{
+  {HB_TAG ('a','f','r','c'), HB_AAT_LAYOUT_FEATURE_TYPE_FRACTIONS,               HB_AAT_LAYOUT_FEATURE_SELECTOR_VERTICAL_FRACTIONS,             HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_FRACTIONS},
+  {HB_TAG ('c','2','p','c'), HB_AAT_LAYOUT_FEATURE_TYPE_UPPER_CASE,              HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_PETITE_CAPS,         HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_UPPER_CASE},
+  {HB_TAG ('c','2','s','c'), HB_AAT_LAYOUT_FEATURE_TYPE_UPPER_CASE,              HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_SMALL_CAPS,          HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_UPPER_CASE},
+  {HB_TAG ('c','a','l','t'), HB_AAT_LAYOUT_FEATURE_TYPE_CONTEXTUAL_ALTERNATIVES, HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_ALTERNATES_ON,       HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_ALTERNATES_OFF},
+  {HB_TAG ('c','a','s','e'), HB_AAT_LAYOUT_FEATURE_TYPE_CASE_SENSITIVE_LAYOUT,   HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_LAYOUT_ON,       HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_LAYOUT_OFF},
+  {HB_TAG ('c','l','i','g'), HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES,               HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_LIGATURES_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_LIGATURES_OFF},
+  {HB_TAG ('c','p','s','p'), HB_AAT_LAYOUT_FEATURE_TYPE_CASE_SENSITIVE_LAYOUT,   HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_SPACING_ON,      HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_SPACING_OFF},
+  {HB_TAG ('c','s','w','h'), HB_AAT_LAYOUT_FEATURE_TYPE_CONTEXTUAL_ALTERNATIVES, HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_SWASH_ALTERNATES_ON, HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_SWASH_ALTERNATES_OFF},
+  {HB_TAG ('d','l','i','g'), HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES,               HB_AAT_LAYOUT_FEATURE_SELECTOR_RARE_LIGATURES_ON,              HB_AAT_LAYOUT_FEATURE_SELECTOR_RARE_LIGATURES_OFF},
+  {HB_TAG ('e','x','p','t'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_EXPERT_CHARACTERS,              (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('f','r','a','c'), HB_AAT_LAYOUT_FEATURE_TYPE_FRACTIONS,               HB_AAT_LAYOUT_FEATURE_SELECTOR_DIAGONAL_FRACTIONS,             HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_FRACTIONS},
+  {HB_TAG ('f','w','i','d'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_MONOSPACED_TEXT,                (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('h','a','l','t'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_HALF_WIDTH_TEXT,            (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('h','i','s','t'), HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES,               HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_OFF},
+  {HB_TAG ('h','k','n','a'), HB_AAT_LAYOUT_FEATURE_TYPE_ALTERNATE_KANA,          HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_HORIZ_KANA_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_HORIZ_KANA_OFF},
+  {HB_TAG ('h','l','i','g'), HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES,               HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_OFF},
+  {HB_TAG ('h','n','g','l'), HB_AAT_LAYOUT_FEATURE_TYPE_TRANSLITERATION,         HB_AAT_LAYOUT_FEATURE_SELECTOR_HANJA_TO_HANGUL,                HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_TRANSLITERATION},
+  {HB_TAG ('h','o','j','o'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_HOJO_CHARACTERS,                (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('h','w','i','d'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_HALF_WIDTH_TEXT,                (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('i','t','a','l'), HB_AAT_LAYOUT_FEATURE_TYPE_ITALIC_CJK_ROMAN,        HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_ON,            HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_OFF},
+  {HB_TAG ('j','p','0','4'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS2004_CHARACTERS,             (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('j','p','7','8'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1978_CHARACTERS,             (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('j','p','8','3'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1983_CHARACTERS,             (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('j','p','9','0'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1990_CHARACTERS,             (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('l','i','g','a'), HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES,               HB_AAT_LAYOUT_FEATURE_SELECTOR_COMMON_LIGATURES_ON,            HB_AAT_LAYOUT_FEATURE_SELECTOR_COMMON_LIGATURES_OFF},
+  {HB_TAG ('l','n','u','m'), HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_CASE,             HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_NUMBERS,             (hb_aat_layout_feature_selector_t) 2},
+  {HB_TAG ('m','g','r','k'), HB_AAT_LAYOUT_FEATURE_TYPE_MATHEMATICAL_EXTRAS,     HB_AAT_LAYOUT_FEATURE_SELECTOR_MATHEMATICAL_GREEK_ON,          HB_AAT_LAYOUT_FEATURE_SELECTOR_MATHEMATICAL_GREEK_OFF},
+  {HB_TAG ('n','l','c','k'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_NLCCHARACTERS,                  (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('o','n','u','m'), HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_CASE,             HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_NUMBERS,             (hb_aat_layout_feature_selector_t) 2},
+  {HB_TAG ('o','r','d','n'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION,       HB_AAT_LAYOUT_FEATURE_SELECTOR_ORDINALS,                       HB_AAT_LAYOUT_FEATURE_SELECTOR_NORMAL_POSITION},
+  {HB_TAG ('p','a','l','t'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_PROPORTIONAL_TEXT,          (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('p','c','a','p'), HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE,              HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_PETITE_CAPS,         HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_LOWER_CASE},
+  {HB_TAG ('p','k','n','a'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_TEXT,              (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('p','n','u','m'), HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_SPACING,          HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_NUMBERS,           (hb_aat_layout_feature_selector_t) 4},
+  {HB_TAG ('p','w','i','d'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_TEXT,              (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('q','w','i','d'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_QUARTER_WIDTH_TEXT,             (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('r','u','b','y'), HB_AAT_LAYOUT_FEATURE_TYPE_RUBY_KANA,               HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_ON,                   HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_OFF},
+  {HB_TAG ('s','i','n','f'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION,       HB_AAT_LAYOUT_FEATURE_SELECTOR_SCIENTIFIC_INFERIORS,           HB_AAT_LAYOUT_FEATURE_SELECTOR_NORMAL_POSITION},
+  {HB_TAG ('s','m','c','p'), HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE,              HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_SMALL_CAPS,          HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_LOWER_CASE},
+  {HB_TAG ('s','m','p','l'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_SIMPLIFIED_CHARACTERS,          (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('s','s','0','1'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ONE_ON,           HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ONE_OFF},
+  {HB_TAG ('s','s','0','2'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWO_ON,           HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWO_OFF},
+  {HB_TAG ('s','s','0','3'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THREE_ON,         HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THREE_OFF},
+  {HB_TAG ('s','s','0','4'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOUR_ON,          HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOUR_OFF},
+  {HB_TAG ('s','s','0','5'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIVE_ON,          HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIVE_OFF},
+  {HB_TAG ('s','s','0','6'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIX_ON,           HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIX_OFF},
+  {HB_TAG ('s','s','0','7'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVEN_ON,         HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVEN_OFF},
+  {HB_TAG ('s','s','0','8'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHT_ON,         HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHT_OFF},
+  {HB_TAG ('s','s','0','9'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINE_ON,          HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINE_OFF},
+  {HB_TAG ('s','s','1','0'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TEN_ON,           HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TEN_OFF},
+  {HB_TAG ('s','s','1','1'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ELEVEN_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ELEVEN_OFF},
+  {HB_TAG ('s','s','1','2'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWELVE_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWELVE_OFF},
+  {HB_TAG ('s','s','1','3'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THIRTEEN_ON,      HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THIRTEEN_OFF},
+  {HB_TAG ('s','s','1','4'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOURTEEN_ON,      HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOURTEEN_OFF},
+  {HB_TAG ('s','s','1','5'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIFTEEN_ON,       HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIFTEEN_OFF},
+  {HB_TAG ('s','s','1','6'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIXTEEN_ON,       HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIXTEEN_OFF},
+  {HB_TAG ('s','s','1','7'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVENTEEN_ON,     HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVENTEEN_OFF},
+  {HB_TAG ('s','s','1','8'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHTEEN_ON,      HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHTEEN_OFF},
+  {HB_TAG ('s','s','1','9'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINETEEN_ON,      HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINETEEN_OFF},
+  {HB_TAG ('s','s','2','0'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES,  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWENTY_ON,        HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWENTY_OFF},
+  {HB_TAG ('s','u','b','s'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION,       HB_AAT_LAYOUT_FEATURE_SELECTOR_INFERIORS,                      HB_AAT_LAYOUT_FEATURE_SELECTOR_NORMAL_POSITION},
+  {HB_TAG ('s','u','p','s'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION,       HB_AAT_LAYOUT_FEATURE_SELECTOR_SUPERIORS,                      HB_AAT_LAYOUT_FEATURE_SELECTOR_NORMAL_POSITION},
+  {HB_TAG ('s','w','s','h'), HB_AAT_LAYOUT_FEATURE_TYPE_CONTEXTUAL_ALTERNATIVES, HB_AAT_LAYOUT_FEATURE_SELECTOR_SWASH_ALTERNATES_ON,            HB_AAT_LAYOUT_FEATURE_SELECTOR_SWASH_ALTERNATES_OFF},
+  {HB_TAG ('t','i','t','l'), HB_AAT_LAYOUT_FEATURE_TYPE_STYLE_OPTIONS,           HB_AAT_LAYOUT_FEATURE_SELECTOR_TITLING_CAPS,                   HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_STYLE_OPTIONS},
+  {HB_TAG ('t','n','a','m'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_NAMES_CHARACTERS,   (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('t','n','u','m'), HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_SPACING,          HB_AAT_LAYOUT_FEATURE_SELECTOR_MONOSPACED_NUMBERS,             (hb_aat_layout_feature_selector_t) 4},
+  {HB_TAG ('t','r','a','d'), HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE,         HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_CHARACTERS,         (hb_aat_layout_feature_selector_t) 16},
+  {HB_TAG ('t','w','i','d'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_THIRD_WIDTH_TEXT,               (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('u','n','i','c'), HB_AAT_LAYOUT_FEATURE_TYPE_LETTER_CASE,             (hb_aat_layout_feature_selector_t) 14,                 (hb_aat_layout_feature_selector_t) 15},
+  {HB_TAG ('v','a','l','t'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_PROPORTIONAL_TEXT,          (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('v','e','r','t'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_SUBSTITUTION,   HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_ON,   HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_OFF},
+  {HB_TAG ('v','h','a','l'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_HALF_WIDTH_TEXT,            (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('v','k','n','a'), HB_AAT_LAYOUT_FEATURE_TYPE_ALTERNATE_KANA,          HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_VERT_KANA_ON,         HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_VERT_KANA_OFF},
+  {HB_TAG ('v','p','a','l'), HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING,            HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_PROPORTIONAL_TEXT,          (hb_aat_layout_feature_selector_t) 7},
+  {HB_TAG ('v','r','t','2'), HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_SUBSTITUTION,   HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_ON,   HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_OFF},
+  {HB_TAG ('z','e','r','o'), HB_AAT_LAYOUT_FEATURE_TYPE_TYPOGRAPHIC_EXTRAS,      HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASHED_ZERO_ON,                HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASHED_ZERO_OFF},
+};
+
+const hb_aat_feature_mapping_t *
+hb_aat_layout_find_feature_mapping (hb_tag_t tag)
+{
+  return hb_sorted_array (feature_mappings).bsearch (tag);
+}
+#endif
+
+
+#ifndef HB_NO_AAT
+
+/*
+ * mort/morx/kerx/trak
+ */
+
+
+void
+hb_aat_layout_compile_map (const hb_aat_map_builder_t *mapper,
+			   hb_aat_map_t *map)
+{
+  const AAT::morx& morx = *mapper->face->table.morx;
+  if (morx.has_data ())
+  {
+    morx.compile_flags (mapper, map);
+    return;
+  }
+
+  const AAT::mort& mort = *mapper->face->table.mort;
+  if (mort.has_data ())
+  {
+    mort.compile_flags (mapper, map);
+    return;
+  }
+}
+
+
+/*
+ * hb_aat_layout_has_substitution:
+ * @face:
+ *
+ * Returns:
+ * Since: 2.3.0
+ */
+hb_bool_t
+hb_aat_layout_has_substitution (hb_face_t *face)
+{
+  return face->table.morx->has_data () ||
+	 face->table.mort->has_data ();
+}
+
+void
+hb_aat_layout_substitute (const hb_ot_shape_plan_t *plan,
+			  hb_font_t *font,
+			  hb_buffer_t *buffer)
+{
+  hb_blob_t *morx_blob = font->face->table.morx.get_blob ();
+  const AAT::morx& morx = *morx_blob->as<AAT::morx> ();
+  if (morx.has_data ())
+  {
+    AAT::hb_aat_apply_context_t c (plan, font, buffer, morx_blob);
+    morx.apply (&c);
+    return;
+  }
+
+  hb_blob_t *mort_blob = font->face->table.mort.get_blob ();
+  const AAT::mort& mort = *mort_blob->as<AAT::mort> ();
+  if (mort.has_data ())
+  {
+    AAT::hb_aat_apply_context_t c (plan, font, buffer, mort_blob);
+    mort.apply (&c);
+    return;
+  }
+}
+
+void
+hb_aat_layout_zero_width_deleted_glyphs (hb_buffer_t *buffer)
+{
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pos = buffer->pos;
+  for (unsigned int i = 0; i < count; i++)
+    if (unlikely (info[i].codepoint == AAT::DELETED_GLYPH))
+      pos[i].x_advance = pos[i].y_advance = pos[i].x_offset = pos[i].y_offset = 0;
+}
+
+static bool
+is_deleted_glyph (const hb_glyph_info_t *info)
+{
+  return info->codepoint == AAT::DELETED_GLYPH;
+}
+
+void
+hb_aat_layout_remove_deleted_glyphs (hb_buffer_t *buffer)
+{
+  hb_ot_layout_delete_glyphs_inplace (buffer, is_deleted_glyph);
+}
+
+/*
+ * hb_aat_layout_has_positioning:
+ * @face:
+ *
+ * Returns:
+ * Since: 2.3.0
+ */
+hb_bool_t
+hb_aat_layout_has_positioning (hb_face_t *face)
+{
+  return face->table.kerx->has_data ();
+}
+
+void
+hb_aat_layout_position (const hb_ot_shape_plan_t *plan,
+			hb_font_t *font,
+			hb_buffer_t *buffer)
+{
+  hb_blob_t *kerx_blob = font->face->table.kerx.get_blob ();
+  const AAT::kerx& kerx = *kerx_blob->as<AAT::kerx> ();
+
+  AAT::hb_aat_apply_context_t c (plan, font, buffer, kerx_blob);
+  c.set_ankr_table (font->face->table.ankr.get ());
+  kerx.apply (&c);
+}
+
+
+/*
+ * hb_aat_layout_has_tracking:
+ * @face:
+ *
+ * Returns:
+ * Since: 2.3.0
+ */
+hb_bool_t
+hb_aat_layout_has_tracking (hb_face_t *face)
+{
+  return face->table.trak->has_data ();
+}
+
+void
+hb_aat_layout_track (const hb_ot_shape_plan_t *plan,
+		     hb_font_t *font,
+		     hb_buffer_t *buffer)
+{
+  const AAT::trak& trak = *font->face->table.trak;
+
+  AAT::hb_aat_apply_context_t c (plan, font, buffer);
+  trak.apply (&c);
+}
+
+/**
+ * hb_aat_layout_get_feature_types:
+ * @face: a face object
+ * @start_offset: iteration's start offset
+ * @feature_count:(inout) (allow-none): buffer size as input, filled size as output
+ * @features: (out caller-allocates) (array length=feature_count): features buffer
+ *
+ * Return value: Number of all available feature types.
+ *
+ * Since: 2.2.0
+ */
+unsigned int
+hb_aat_layout_get_feature_types (hb_face_t                    *face,
+				 unsigned int                  start_offset,
+				 unsigned int                 *feature_count, /* IN/OUT.  May be NULL. */
+				 hb_aat_layout_feature_type_t *features       /* OUT.     May be NULL. */)
+{
+  return face->table.feat->get_feature_types (start_offset, feature_count, features);
+}
+
+/**
+ * hb_aat_layout_feature_type_get_name_id:
+ * @face: a face object
+ * @feature_type: feature id
+ *
+ * Return value: Name ID index
+ *
+ * Since: 2.2.0
+ */
+hb_ot_name_id_t
+hb_aat_layout_feature_type_get_name_id (hb_face_t                    *face,
+					hb_aat_layout_feature_type_t  feature_type)
+{
+  return face->table.feat->get_feature_name_id (feature_type);
+}
+
+/**
+ * hb_aat_layout_feature_type_get_selectors:
+ * @face:    a face object
+ * @feature_type: feature id
+ * @start_offset:    iteration's start offset
+ * @selector_count: (inout) (allow-none): buffer size as input, filled size as output
+ * @selectors: (out caller-allocates) (array length=selector_count): settings buffer
+ * @default_index: (out) (allow-none): index of default selector if any
+ *
+ * If upon return, @default_index is set to #HB_AAT_LAYOUT_NO_SELECTOR_INDEX, then
+ * the feature type is non-exclusive.  Otherwise, @default_index is the index of
+ * the selector that is selected by default.
+ *
+ * Return value: Number of all available feature selectors.
+ *
+ * Since: 2.2.0
+ */
+unsigned int
+hb_aat_layout_feature_type_get_selector_infos (hb_face_t                             *face,
+					       hb_aat_layout_feature_type_t           feature_type,
+					       unsigned int                           start_offset,
+					       unsigned int                          *selector_count, /* IN/OUT.  May be NULL. */
+					       hb_aat_layout_feature_selector_info_t *selectors,      /* OUT.     May be NULL. */
+					       unsigned int                          *default_index   /* OUT.     May be NULL. */)
+{
+  return face->table.feat->get_selector_infos (feature_type, start_offset, selector_count, selectors, default_index);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout.h b/thirdparty/harfbuzz/src/hb-aat-layout.h
new file mode 100644
index 0000000000..b617e8b703
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout.h
@@ -0,0 +1,486 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_H_IN
+#error "Include <hb-aat.h> instead."
+#endif
+
+#ifndef HB_AAT_LAYOUT_H
+#define HB_AAT_LAYOUT_H
+
+#include "hb.h"
+
+#include "hb-ot.h"
+
+HB_BEGIN_DECLS
+
+/**
+ * hb_aat_layout_feature_type_t:
+ *
+ *
+ * Since: 2.2.0
+ */
+typedef enum
+{
+  HB_AAT_LAYOUT_FEATURE_TYPE_INVALID				= 0xFFFF,
+
+  HB_AAT_LAYOUT_FEATURE_TYPE_ALL_TYPOGRAPHIC			= 0,
+  HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES				= 1,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CURISVE_CONNECTION			= 2,
+  HB_AAT_LAYOUT_FEATURE_TYPE_LETTER_CASE			= 3,
+  HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_SUBSTITUTION		= 4,
+  HB_AAT_LAYOUT_FEATURE_TYPE_LINGUISTIC_REARRANGEMENT		= 5,
+  HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_SPACING			= 6,
+  HB_AAT_LAYOUT_FEATURE_TYPE_SMART_SWASH_TYPE			= 8,
+  HB_AAT_LAYOUT_FEATURE_TYPE_DIACRITICS_TYPE			= 9,
+  HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION			= 10,
+  HB_AAT_LAYOUT_FEATURE_TYPE_FRACTIONS				= 11,
+  HB_AAT_LAYOUT_FEATURE_TYPE_OVERLAPPING_CHARACTERS_TYPE	= 13,
+  HB_AAT_LAYOUT_FEATURE_TYPE_TYPOGRAPHIC_EXTRAS			= 14,
+  HB_AAT_LAYOUT_FEATURE_TYPE_MATHEMATICAL_EXTRAS		= 15,
+  HB_AAT_LAYOUT_FEATURE_TYPE_ORNAMENT_SETS_TYPE			= 16,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_ALTERNATIVES		= 17,
+  HB_AAT_LAYOUT_FEATURE_TYPE_DESIGN_COMPLEXITY_TYPE		= 18,
+  HB_AAT_LAYOUT_FEATURE_TYPE_STYLE_OPTIONS			= 19,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE			= 20,
+  HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_CASE			= 21,
+  HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING			= 22,
+  HB_AAT_LAYOUT_FEATURE_TYPE_TRANSLITERATION			= 23,
+  HB_AAT_LAYOUT_FEATURE_TYPE_ANNOTATION_TYPE			= 24,
+  HB_AAT_LAYOUT_FEATURE_TYPE_KANA_SPACING_TYPE			= 25,
+  HB_AAT_LAYOUT_FEATURE_TYPE_IDEOGRAPHIC_SPACING_TYPE		= 26,
+  HB_AAT_LAYOUT_FEATURE_TYPE_UNICODE_DECOMPOSITION_TYPE		= 27,
+  HB_AAT_LAYOUT_FEATURE_TYPE_RUBY_KANA				= 28,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CJK_SYMBOL_ALTERNATIVES_TYPE	= 29,
+  HB_AAT_LAYOUT_FEATURE_TYPE_IDEOGRAPHIC_ALTERNATIVES_TYPE	= 30,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CJK_VERTICAL_ROMAN_PLACEMENT_TYPE	= 31,
+  HB_AAT_LAYOUT_FEATURE_TYPE_ITALIC_CJK_ROMAN			= 32,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CASE_SENSITIVE_LAYOUT		= 33,
+  HB_AAT_LAYOUT_FEATURE_TYPE_ALTERNATE_KANA			= 34,
+  HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES		= 35,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CONTEXTUAL_ALTERNATIVES		= 36,
+  HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE				= 37,
+  HB_AAT_LAYOUT_FEATURE_TYPE_UPPER_CASE				= 38,
+  HB_AAT_LAYOUT_FEATURE_TYPE_LANGUAGE_TAG_TYPE			= 39,
+  HB_AAT_LAYOUT_FEATURE_TYPE_CJK_ROMAN_SPACING_TYPE		= 103,
+
+  _HB_AAT_LAYOUT_FEATURE_TYPE_MAX_VALUE = HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_aat_layout_feature_type_t;
+
+/**
+ * hb_aat_layout_feature_selector_t:
+ *
+ *
+ * Since: 2.2.0
+ */
+typedef enum
+{
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INVALID			= 0xFFFF,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_ALL_TYPOGRAPHIC */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALL_TYPE_FEATURES_ON		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALL_TYPE_FEATURES_OFF		= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_REQUIRED_LIGATURES_ON		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_REQUIRED_LIGATURES_OFF		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_COMMON_LIGATURES_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_COMMON_LIGATURES_OFF		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_RARE_LIGATURES_ON		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_RARE_LIGATURES_OFF		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LOGOS_ON			= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LOGOS_OFF			= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_REBUS_PICTURES_ON		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_REBUS_PICTURES_OFF		= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DIPHTHONG_LIGATURES_ON		= 10,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DIPHTHONG_LIGATURES_OFF	= 11,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SQUARED_LIGATURES_ON		= 12,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SQUARED_LIGATURES_OFF		= 13,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ABBREV_SQUARED_LIGATURES_ON	= 14,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ABBREV_SQUARED_LIGATURES_OFF	= 15,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SYMBOL_LIGATURES_ON		= 16,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SYMBOL_LIGATURES_OFF		= 17,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_LIGATURES_ON	= 18,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_LIGATURES_OFF	= 19,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_ON	= 20,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HISTORICAL_LIGATURES_OFF	= 21,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_LIGATURES */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_UNCONNECTED			= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PARTIALLY_CONNECTED		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CURSIVE			= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_LETTER_CASE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_AND_LOWER_CASE		= 0, /* deprecated */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALL_CAPS			= 1, /* deprecated */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALL_LOWER_CASE			= 2, /* deprecated */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SMALL_CAPS			= 3, /* deprecated */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INITIAL_CAPS			= 4, /* deprecated */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INITIAL_CAPS_AND_SMALL_CAPS	= 5, /* deprecated */
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_SUBSTITUTION */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SUBSTITUTE_VERTICAL_FORMS_OFF	= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_LINGUISTIC_REARRANGEMENT */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINGUISTIC_REARRANGEMENT_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINGUISTIC_REARRANGEMENT_OFF	= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_SPACING */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_MONOSPACED_NUMBERS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_NUMBERS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_THIRD_WIDTH_NUMBERS		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_QUARTER_WIDTH_NUMBERS		= 3,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_SMART_SWASH_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_WORD_INITIAL_SWASHES_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_WORD_INITIAL_SWASHES_OFF	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_WORD_FINAL_SWASHES_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_WORD_FINAL_SWASHES_OFF		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINE_INITIAL_SWASHES_ON	= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINE_INITIAL_SWASHES_OFF	= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINE_FINAL_SWASHES_ON		= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LINE_FINAL_SWASHES_OFF		= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NON_FINAL_SWASHES_ON		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NON_FINAL_SWASHES_OFF		= 9,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_DIACRITICS_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SHOW_DIACRITICS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HIDE_DIACRITICS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DECOMPOSE_DIACRITICS		= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_VERTICAL_POSITION */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NORMAL_POSITION		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SUPERIORS			= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INFERIORS			= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ORDINALS			= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SCIENTIFIC_INFERIORS		= 4,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_FRACTIONS */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_FRACTIONS			= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_VERTICAL_FRACTIONS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DIAGONAL_FRACTIONS		= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_OVERLAPPING_CHARACTERS_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PREVENT_OVERLAP_ON		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PREVENT_OVERLAP_OFF		= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_TYPOGRAPHIC_EXTRAS */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHENS_TO_EM_DASH_ON		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHENS_TO_EM_DASH_OFF		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHEN_TO_EN_DASH_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHEN_TO_EN_DASH_OFF		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASHED_ZERO_ON		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASHED_ZERO_OFF		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FORM_INTERROBANG_ON		= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FORM_INTERROBANG_OFF		= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SMART_QUOTES_ON		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SMART_QUOTES_OFF		= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PERIODS_TO_ELLIPSIS_ON		= 10,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PERIODS_TO_ELLIPSIS_OFF	= 11,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_MATHEMATICAL_EXTRAS */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHEN_TO_MINUS_ON		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HYPHEN_TO_MINUS_OFF		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ASTERISK_TO_MULTIPLY_ON	= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ASTERISK_TO_MULTIPLY_OFF	= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASH_TO_DIVIDE_ON		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SLASH_TO_DIVIDE_OFF		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INEQUALITY_LIGATURES_ON	= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INEQUALITY_LIGATURES_OFF	= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_EXPONENTS_ON			= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_EXPONENTS_OFF			= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_MATHEMATICAL_GREEK_ON		= 10,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_MATHEMATICAL_GREEK_OFF		= 11,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_ORNAMENT_SETS_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_ORNAMENTS			= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DINGBATS			= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PI_CHARACTERS			= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FLEURONS			= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DECORATIVE_BORDERS		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INTERNATIONAL_SYMBOLS		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_MATH_SYMBOLS			= 6,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_ALTERNATIVES */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_ALTERNATES			= 0,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_DESIGN_COMPLEXITY_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DESIGN_LEVEL1			= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DESIGN_LEVEL2			= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DESIGN_LEVEL3			= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DESIGN_LEVEL4			= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DESIGN_LEVEL5			= 4,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_STYLE_OPTIONS */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_STYLE_OPTIONS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DISPLAY_TEXT			= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ENGRAVED_TEXT			= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ILLUMINATED_CAPS		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TITLING_CAPS			= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TALL_CAPS			= 5,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_SHAPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_CHARACTERS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SIMPLIFIED_CHARACTERS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1978_CHARACTERS		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1983_CHARACTERS		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS1990_CHARACTERS		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_ALT_ONE		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_ALT_TWO		= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_ALT_THREE		= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_ALT_FOUR		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_ALT_FIVE		= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_EXPERT_CHARACTERS		= 10,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_JIS2004_CHARACTERS		= 11,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HOJO_CHARACTERS		= 12,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NLCCHARACTERS			= 13,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRADITIONAL_NAMES_CHARACTERS	= 14,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_NUMBER_CASE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_NUMBERS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_NUMBERS		= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_TEXT_SPACING */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_TEXT		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_MONOSPACED_TEXT		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HALF_WIDTH_TEXT		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_THIRD_WIDTH_TEXT		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_QUARTER_WIDTH_TEXT		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_PROPORTIONAL_TEXT		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALT_HALF_WIDTH_TEXT		= 6,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_TRANSLITERATION */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_TRANSLITERATION		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HANJA_TO_HANGUL		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HIRAGANA_TO_KATAKANA		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_KATAKANA_TO_HIRAGANA		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_KANA_TO_ROMANIZATION		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ROMANIZATION_TO_HIRAGANA	= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ROMANIZATION_TO_KATAKANA	= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HANJA_TO_HANGUL_ALT_ONE	= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HANJA_TO_HANGUL_ALT_TWO	= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HANJA_TO_HANGUL_ALT_THREE	= 9,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_ANNOTATION_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_ANNOTATION			= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_BOX_ANNOTATION			= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ROUNDED_BOX_ANNOTATION		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CIRCLE_ANNOTATION		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INVERTED_CIRCLE_ANNOTATION	= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PARENTHESIS_ANNOTATION		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PERIOD_ANNOTATION		= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ROMAN_NUMERAL_ANNOTATION	= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DIAMOND_ANNOTATION		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INVERTED_BOX_ANNOTATION	= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_INVERTED_ROUNDED_BOX_ANNOTATION= 10,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_KANA_SPACING_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FULL_WIDTH_KANA		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_KANA		= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_IDEOGRAPHIC_SPACING_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FULL_WIDTH_IDEOGRAPHS		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_IDEOGRAPHS	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HALF_WIDTH_IDEOGRAPHS		= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_UNICODE_DECOMPOSITION_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CANONICAL_COMPOSITION_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CANONICAL_COMPOSITION_OFF	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_COMPATIBILITY_COMPOSITION_ON	= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_COMPATIBILITY_COMPOSITION_OFF	= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRANSCODING_COMPOSITION_ON	= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_TRANSCODING_COMPOSITION_OFF	= 5,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_RUBY_KANA */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_RUBY_KANA			= 0, /* deprecated - use HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_OFF instead */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA			= 1, /* deprecated - use HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_ON instead */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_ON			= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_RUBY_KANA_OFF			= 3,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CJK_SYMBOL_ALTERNATIVES_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_CJK_SYMBOL_ALTERNATIVES	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_SYMBOL_ALT_ONE		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_SYMBOL_ALT_TWO		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_SYMBOL_ALT_THREE		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_SYMBOL_ALT_FOUR		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_SYMBOL_ALT_FIVE		= 5,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_IDEOGRAPHIC_ALTERNATIVES_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_IDEOGRAPHIC_ALTERNATIVES	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_IDEOGRAPHIC_ALT_ONE		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_IDEOGRAPHIC_ALT_TWO		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_IDEOGRAPHIC_ALT_THREE		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_IDEOGRAPHIC_ALT_FOUR		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_IDEOGRAPHIC_ALT_FIVE		= 5,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CJK_VERTICAL_ROMAN_PLACEMENT_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_VERTICAL_ROMAN_CENTERED	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_VERTICAL_ROMAN_HBASELINE	= 1,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_ITALIC_CJK_ROMAN */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_CJK_ITALIC_ROMAN		= 0,    /* deprecated - use HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_OFF instead */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN		= 1,    /* deprecated - use HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_ON instead */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CJK_ITALIC_ROMAN_OFF		= 3,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CASE_SENSITIVE_LAYOUT */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_LAYOUT_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_LAYOUT_OFF	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_SPACING_ON	= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CASE_SENSITIVE_SPACING_OFF	= 3,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_ALTERNATE_KANA */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_HORIZ_KANA_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_HORIZ_KANA_OFF	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_VERT_KANA_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_ALTERNATE_VERT_KANA_OFF	= 3,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_STYLISTIC_ALTERNATIVES */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_NO_STYLISTIC_ALTERNATES	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ONE_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ONE_OFF		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWO_ON		= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWO_OFF		= 5,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THREE_ON		= 6,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THREE_OFF	= 7,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOUR_ON		= 8,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOUR_OFF		= 9,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIVE_ON		= 10,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIVE_OFF		= 11,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIX_ON		= 12,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIX_OFF		= 13,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVEN_ON		= 14,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVEN_OFF	= 15,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHT_ON		= 16,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHT_OFF	= 17,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINE_ON		= 18,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINE_OFF		= 19,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TEN_ON		= 20,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TEN_OFF		= 21,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ELEVEN_ON	= 22,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_ELEVEN_OFF	= 23,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWELVE_ON	= 24,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWELVE_OFF	= 25,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THIRTEEN_ON	= 26,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_THIRTEEN_OFF	= 27,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOURTEEN_ON	= 28,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FOURTEEN_OFF	= 29,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIFTEEN_ON	= 30,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_FIFTEEN_OFF	= 31,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIXTEEN_ON	= 32,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SIXTEEN_OFF	= 33,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVENTEEN_ON	= 34,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_SEVENTEEN_OFF	= 35,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHTEEN_ON	= 36,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_EIGHTEEN_OFF	= 37,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINETEEN_ON	= 38,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_NINETEEN_OFF	= 39,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWENTY_ON	= 40,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_STYLISTIC_ALT_TWENTY_OFF	= 41,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CONTEXTUAL_ALTERNATIVES */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_ALTERNATES_ON	= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_ALTERNATES_OFF	= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SWASH_ALTERNATES_ON		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_SWASH_ALTERNATES_OFF		= 3,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_SWASH_ALTERNATES_ON	= 4,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_CONTEXTUAL_SWASH_ALTERNATES_OFF= 5,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_LOWER_CASE		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_SMALL_CAPS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_PETITE_CAPS		= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_UPPER_CASE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_UPPER_CASE		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_SMALL_CAPS		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_UPPER_CASE_PETITE_CAPS		= 2,
+
+  /* Selectors for #HB_AAT_LAYOUT_FEATURE_TYPE_CJK_ROMAN_SPACING_TYPE */
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_HALF_WIDTH_CJK_ROMAN		= 0,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_PROPORTIONAL_CJK_ROMAN		= 1,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_DEFAULT_CJK_ROMAN		= 2,
+  HB_AAT_LAYOUT_FEATURE_SELECTOR_FULL_WIDTH_CJK_ROMAN		= 3,
+
+  _HB_AAT_LAYOUT_FEATURE_SELECTOR_MAX_VALUE = HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_aat_layout_feature_selector_t;
+
+HB_EXTERN unsigned int
+hb_aat_layout_get_feature_types (hb_face_t                    *face,
+				 unsigned int                  start_offset,
+				 unsigned int                 *feature_count, /* IN/OUT.  May be NULL. */
+				 hb_aat_layout_feature_type_t *features       /* OUT.     May be NULL. */);
+
+HB_EXTERN hb_ot_name_id_t
+hb_aat_layout_feature_type_get_name_id (hb_face_t                    *face,
+					hb_aat_layout_feature_type_t  feature_type);
+
+typedef struct hb_aat_layout_feature_selector_info_t
+{
+  hb_ot_name_id_t			name_id;
+  hb_aat_layout_feature_selector_t	enable;
+  hb_aat_layout_feature_selector_t	disable;
+  /*< private >*/
+  unsigned int				reserved;
+} hb_aat_layout_feature_selector_info_t;
+
+#define HB_AAT_LAYOUT_NO_SELECTOR_INDEX		0xFFFFu
+
+HB_EXTERN unsigned int
+hb_aat_layout_feature_type_get_selector_infos (hb_face_t                             *face,
+					       hb_aat_layout_feature_type_t           feature_type,
+					       unsigned int                           start_offset,
+					       unsigned int                          *selector_count, /* IN/OUT.  May be NULL. */
+					       hb_aat_layout_feature_selector_info_t *selectors,      /* OUT.     May be NULL. */
+					       unsigned int                          *default_index   /* OUT.     May be NULL. */);
+
+
+/*
+ * morx/mort
+ */
+
+HB_EXTERN hb_bool_t
+hb_aat_layout_has_substitution (hb_face_t *face);
+
+
+/*
+ * kerx
+ */
+
+HB_EXTERN hb_bool_t
+hb_aat_layout_has_positioning (hb_face_t *face);
+
+
+/*
+ * trak
+ */
+
+HB_EXTERN hb_bool_t
+hb_aat_layout_has_tracking (hb_face_t *face);
+
+
+HB_END_DECLS
+
+#endif /* HB_AAT_LAYOUT_H */
diff --git a/thirdparty/harfbuzz/src/hb-aat-layout.hh b/thirdparty/harfbuzz/src/hb-aat-layout.hh
new file mode 100644
index 0000000000..5e4e3bda15
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-layout.hh
@@ -0,0 +1,75 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_LAYOUT_HH
+#define HB_AAT_LAYOUT_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape.hh"
+#include "hb-aat-ltag-table.hh"
+
+struct hb_aat_feature_mapping_t
+{
+  hb_tag_t otFeatureTag;
+  hb_aat_layout_feature_type_t aatFeatureType;
+  hb_aat_layout_feature_selector_t selectorToEnable;
+  hb_aat_layout_feature_selector_t selectorToDisable;
+
+  int cmp (hb_tag_t key) const
+  { return key < otFeatureTag ? -1 : key > otFeatureTag ? 1 : 0; }
+};
+
+HB_INTERNAL const hb_aat_feature_mapping_t *
+hb_aat_layout_find_feature_mapping (hb_tag_t tag);
+
+HB_INTERNAL void
+hb_aat_layout_compile_map (const hb_aat_map_builder_t *mapper,
+			   hb_aat_map_t *map);
+
+HB_INTERNAL void
+hb_aat_layout_substitute (const hb_ot_shape_plan_t *plan,
+			  hb_font_t *font,
+			  hb_buffer_t *buffer);
+
+HB_INTERNAL void
+hb_aat_layout_zero_width_deleted_glyphs (hb_buffer_t *buffer);
+
+HB_INTERNAL void
+hb_aat_layout_remove_deleted_glyphs (hb_buffer_t *buffer);
+
+HB_INTERNAL void
+hb_aat_layout_position (const hb_ot_shape_plan_t *plan,
+			hb_font_t *font,
+			hb_buffer_t *buffer);
+
+HB_INTERNAL void
+hb_aat_layout_track (const hb_ot_shape_plan_t *plan,
+		     hb_font_t *font,
+		     hb_buffer_t *buffer);
+
+
+#endif /* HB_AAT_LAYOUT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-ltag-table.hh b/thirdparty/harfbuzz/src/hb-aat-ltag-table.hh
new file mode 100644
index 0000000000..711f9aa6c1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-ltag-table.hh
@@ -0,0 +1,92 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_LTAG_TABLE_HH
+#define HB_AAT_LTAG_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * ltag -- Language Tag
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6ltag.html
+ */
+#define HB_AAT_TAG_ltag HB_TAG('l','t','a','g')
+
+
+namespace AAT {
+
+using namespace OT;
+
+
+struct FTStringRange
+{
+  friend struct ltag;
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && (base+tag).sanitize (c, length));
+  }
+
+  protected:
+  NNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		tag;		/* Offset from the start of the table to
+				 * the beginning of the string */
+  HBUINT16	length;		/* String length (in bytes) */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct ltag
+{
+  static constexpr hb_tag_t tableTag = HB_AAT_TAG_ltag;
+
+  hb_language_t get_language (unsigned int i) const
+  {
+    const FTStringRange &range = tagRanges[i];
+    return hb_language_from_string ((const char *) (this+range.tag).arrayZ,
+				    range.length);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version >= 1 &&
+			  tagRanges.sanitize (c, this)));
+  }
+
+  protected:
+  HBUINT32	version;	/* Table version; currently 1 */
+  HBUINT32	flags;		/* Table flags; currently none defined */
+  LArrayOf<FTStringRange>
+		tagRanges;	/* Range for each tag's string */
+  public:
+  DEFINE_SIZE_ARRAY (12, tagRanges);
+};
+
+} /* namespace AAT */
+
+
+#endif /* HB_AAT_LTAG_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat-map.cc b/thirdparty/harfbuzz/src/hb-aat-map.cc
new file mode 100644
index 0000000000..2c38c35029
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-map.cc
@@ -0,0 +1,102 @@
+/*
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2011,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_AAT_SHAPE
+
+#include "hb-aat-map.hh"
+
+#include "hb-aat-layout.hh"
+#include "hb-aat-layout-feat-table.hh"
+
+
+void hb_aat_map_builder_t::add_feature (hb_tag_t tag, unsigned value)
+{
+  if (!face->table.feat->has_data ()) return;
+
+  if (tag == HB_TAG ('a','a','l','t'))
+  {
+    if (!face->table.feat->exposes_feature (HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_ALTERNATIVES))
+      return;
+    feature_info_t *info = features.push();
+    info->type = HB_AAT_LAYOUT_FEATURE_TYPE_CHARACTER_ALTERNATIVES;
+    info->setting = (hb_aat_layout_feature_selector_t) value;
+    info->seq = features.length;
+    info->is_exclusive = true;
+    return;
+  }
+
+  const hb_aat_feature_mapping_t *mapping = hb_aat_layout_find_feature_mapping (tag);
+  if (!mapping) return;
+
+  const AAT::FeatureName* feature = &face->table.feat->get_feature (mapping->aatFeatureType);
+  if (!feature->has_data ())
+  {
+    /* Special case: Chain::compile_flags will fall back to the deprecated version of
+     * small-caps if necessary, so we need to check for that possibility.
+     * https://github.com/harfbuzz/harfbuzz/issues/2307 */
+    if (mapping->aatFeatureType == HB_AAT_LAYOUT_FEATURE_TYPE_LOWER_CASE &&
+	mapping->selectorToEnable == HB_AAT_LAYOUT_FEATURE_SELECTOR_LOWER_CASE_SMALL_CAPS)
+    {
+      feature = &face->table.feat->get_feature (HB_AAT_LAYOUT_FEATURE_TYPE_LETTER_CASE);
+      if (!feature->has_data ()) return;
+    }
+    else return;
+  }
+
+  feature_info_t *info = features.push();
+  info->type = mapping->aatFeatureType;
+  info->setting = value ? mapping->selectorToEnable : mapping->selectorToDisable;
+  info->seq = features.length;
+  info->is_exclusive = feature->is_exclusive ();
+}
+
+void
+hb_aat_map_builder_t::compile (hb_aat_map_t  &m)
+{
+  /* Sort features and merge duplicates */
+  if (features.length)
+  {
+    features.qsort ();
+    unsigned int j = 0;
+    for (unsigned int i = 1; i < features.length; i++)
+      if (features[i].type != features[j].type ||
+	  /* Nonexclusive feature selectors come in even/odd pairs to turn a setting on/off
+	   * respectively, so we mask out the low-order bit when checking for "duplicates"
+	   * (selectors referring to the same feature setting) here. */
+	  (!features[i].is_exclusive && ((features[i].setting & ~1) != (features[j].setting & ~1))))
+	features[++j] = features[i];
+    features.shrink (j + 1);
+  }
+
+  hb_aat_layout_compile_map (this, &m);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-aat-map.hh b/thirdparty/harfbuzz/src/hb-aat-map.hh
new file mode 100644
index 0000000000..5a0fa70544
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat-map.hh
@@ -0,0 +1,96 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_AAT_MAP_HH
+#define HB_AAT_MAP_HH
+
+#include "hb.hh"
+
+
+struct hb_aat_map_t
+{
+  friend struct hb_aat_map_builder_t;
+
+  public:
+
+  void init ()
+  {
+    memset (this, 0, sizeof (*this));
+    chain_flags.init ();
+  }
+  void fini () { chain_flags.fini (); }
+
+  public:
+  hb_vector_t<hb_mask_t> chain_flags;
+};
+
+struct hb_aat_map_builder_t
+{
+  public:
+
+  HB_INTERNAL hb_aat_map_builder_t (hb_face_t *face_,
+				    const hb_segment_properties_t *props_ HB_UNUSED) :
+				      face (face_) {}
+
+  HB_INTERNAL void add_feature (hb_tag_t tag, unsigned int value=1);
+
+  HB_INTERNAL void compile (hb_aat_map_t  &m);
+
+  public:
+  struct feature_info_t
+  {
+    hb_aat_layout_feature_type_t  type;
+    hb_aat_layout_feature_selector_t  setting;
+    bool is_exclusive;
+    unsigned  seq; /* For stable sorting only. */
+
+    HB_INTERNAL static int cmp (const void *pa, const void *pb)
+    {
+      const feature_info_t *a = (const feature_info_t *) pa;
+      const feature_info_t *b = (const feature_info_t *) pb;
+      if (a->type != b->type) return (a->type < b->type ? -1 : 1);
+      if (!a->is_exclusive &&
+	  (a->setting & ~1) != (b->setting & ~1)) return (a->setting < b->setting ? -1 : 1);
+	    return (a->seq < b->seq ? -1 : a->seq > b->seq ? 1 : 0);
+    }
+
+    /* compares type & setting only, not is_exclusive flag or seq number */
+    int cmp (const feature_info_t& f) const
+    {
+      return (f.type != type) ? (f.type < type ? -1 : 1) :
+	     (f.setting != setting) ? (f.setting < setting ? -1 : 1) : 0;
+    }
+  };
+
+  public:
+  hb_face_t *face;
+
+  public:
+  hb_sorted_vector_t<feature_info_t> features;
+};
+
+
+#endif /* HB_AAT_MAP_HH */
diff --git a/thirdparty/harfbuzz/src/hb-aat.h b/thirdparty/harfbuzz/src/hb-aat.h
new file mode 100644
index 0000000000..c14313d1e2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-aat.h
@@ -0,0 +1,38 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_AAT_H
+#define HB_AAT_H
+#define HB_AAT_H_IN
+
+#include "hb.h"
+
+#include "hb-aat-layout.h"
+
+HB_BEGIN_DECLS
+
+HB_END_DECLS
+
+#undef HB_AAT_H_IN
+#endif /* HB_AAT_H */
diff --git a/thirdparty/harfbuzz/src/hb-algs.hh b/thirdparty/harfbuzz/src/hb-algs.hh
new file mode 100644
index 0000000000..30b5812e12
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-algs.hh
@@ -0,0 +1,1127 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ * Copyright © 2019  Facebook, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_ALGS_HH
+#define HB_ALGS_HH
+
+#include "hb.hh"
+#include "hb-meta.hh"
+#include "hb-null.hh"
+#include "hb-number.hh"
+
+
+/* Encodes three unsigned integers in one 64-bit number.  If the inputs have more than 21 bits,
+ * values will be truncated / overlap, and might not decode exactly. */
+#define HB_CODEPOINT_ENCODE3(x,y,z) (((uint64_t) (x) << 42) | ((uint64_t) (y) << 21) | (uint64_t) (z))
+#define HB_CODEPOINT_DECODE3_1(v) ((hb_codepoint_t) ((v) >> 42))
+#define HB_CODEPOINT_DECODE3_2(v) ((hb_codepoint_t) ((v) >> 21) & 0x1FFFFFu)
+#define HB_CODEPOINT_DECODE3_3(v) ((hb_codepoint_t) (v) & 0x1FFFFFu)
+
+/* Custom encoding used by hb-ucd. */
+#define HB_CODEPOINT_ENCODE3_11_7_14(x,y,z) (((uint32_t) ((x) & 0x07FFu) << 21) | (((uint32_t) (y) & 0x007Fu) << 14) | (uint32_t) ((z) & 0x3FFFu))
+#define HB_CODEPOINT_DECODE3_11_7_14_1(v) ((hb_codepoint_t) ((v) >> 21))
+#define HB_CODEPOINT_DECODE3_11_7_14_2(v) ((hb_codepoint_t) (((v) >> 14) & 0x007Fu) | 0x0300)
+#define HB_CODEPOINT_DECODE3_11_7_14_3(v) ((hb_codepoint_t) (v) & 0x3FFFu)
+
+struct
+{
+  /* Note.  This is dangerous in that if it's passed an rvalue, it returns rvalue-reference. */
+  template <typename T> constexpr auto
+  operator () (T&& v) const HB_AUTO_RETURN ( hb_forward<T> (v) )
+}
+HB_FUNCOBJ (hb_identity);
+struct
+{
+  /* Like identity(), but only retains lvalue-references.  Rvalues are returned as rvalues. */
+  template <typename T> constexpr T&
+  operator () (T& v) const { return v; }
+
+  template <typename T> constexpr hb_remove_reference<T>
+  operator () (T&& v) const { return v; }
+}
+HB_FUNCOBJ (hb_lidentity);
+struct
+{
+  /* Like identity(), but always returns rvalue. */
+  template <typename T> constexpr hb_remove_reference<T>
+  operator () (T&& v) const { return v; }
+}
+HB_FUNCOBJ (hb_ridentity);
+
+struct
+{
+  template <typename T> constexpr bool
+  operator () (T&& v) const { return bool (hb_forward<T> (v)); }
+}
+HB_FUNCOBJ (hb_bool);
+
+struct
+{
+  private:
+
+  template <typename T> constexpr auto
+  impl (const T& v, hb_priority<1>) const HB_RETURN (uint32_t, hb_deref (v).hash ())
+
+  template <typename T,
+	    hb_enable_if (hb_is_integral (T))> constexpr auto
+  impl (const T& v, hb_priority<0>) const HB_AUTO_RETURN
+  (
+    /* Knuth's multiplicative method: */
+    (uint32_t) v * 2654435761u
+  )
+
+  public:
+
+  template <typename T> constexpr auto
+  operator () (const T& v) const HB_RETURN (uint32_t, impl (v, hb_prioritize))
+}
+HB_FUNCOBJ (hb_hash);
+
+
+struct
+{
+  private:
+
+  /* Pointer-to-member-function. */
+  template <typename Appl, typename T, typename ...Ts> auto
+  impl (Appl&& a, hb_priority<2>, T &&v, Ts&&... ds) const HB_AUTO_RETURN
+  ((hb_deref (hb_forward<T> (v)).*hb_forward<Appl> (a)) (hb_forward<Ts> (ds)...))
+
+  /* Pointer-to-member. */
+  template <typename Appl, typename T> auto
+  impl (Appl&& a, hb_priority<1>, T &&v) const HB_AUTO_RETURN
+  ((hb_deref (hb_forward<T> (v))).*hb_forward<Appl> (a))
+
+  /* Operator(). */
+  template <typename Appl, typename ...Ts> auto
+  impl (Appl&& a, hb_priority<0>, Ts&&... ds) const HB_AUTO_RETURN
+  (hb_deref (hb_forward<Appl> (a)) (hb_forward<Ts> (ds)...))
+
+  public:
+
+  template <typename Appl, typename ...Ts> auto
+  operator () (Appl&& a, Ts&&... ds) const HB_AUTO_RETURN
+  (
+    impl (hb_forward<Appl> (a),
+	  hb_prioritize,
+	  hb_forward<Ts> (ds)...)
+  )
+}
+HB_FUNCOBJ (hb_invoke);
+
+template <unsigned Pos, typename Appl, typename V>
+struct hb_partial_t
+{
+  hb_partial_t (Appl a, V v) : a (a), v (v) {}
+
+  static_assert (Pos > 0, "");
+
+  template <typename ...Ts,
+	    unsigned P = Pos,
+	    hb_enable_if (P == 1)> auto
+  operator () (Ts&& ...ds) -> decltype (hb_invoke (hb_declval (Appl),
+						   hb_declval (V),
+						   hb_declval (Ts)...))
+  {
+    return hb_invoke (hb_forward<Appl> (a),
+		      hb_forward<V> (v),
+		      hb_forward<Ts> (ds)...);
+  }
+  template <typename T0, typename ...Ts,
+	    unsigned P = Pos,
+	    hb_enable_if (P == 2)> auto
+  operator () (T0&& d0, Ts&& ...ds) -> decltype (hb_invoke (hb_declval (Appl),
+							    hb_declval (T0),
+							    hb_declval (V),
+							    hb_declval (Ts)...))
+  {
+    return hb_invoke (hb_forward<Appl> (a),
+		      hb_forward<T0> (d0),
+		      hb_forward<V> (v),
+		      hb_forward<Ts> (ds)...);
+  }
+
+  private:
+  hb_reference_wrapper<Appl> a;
+  V v;
+};
+template <unsigned Pos=1, typename Appl, typename V>
+auto hb_partial (Appl&& a, V&& v) HB_AUTO_RETURN
+(( hb_partial_t<Pos, Appl, V> (a, v) ))
+
+/* The following, HB_PARTIALIZE, macro uses a particular corner-case
+ * of C++11 that is not particularly well-supported by all compilers.
+ * What's happening is that it's using "this" in a trailing return-type
+ * via decltype().  Broken compilers deduce the type of "this" pointer
+ * in that context differently from what it resolves to in the body
+ * of the function.
+ *
+ * One probable cause of this is that at the time of trailing return
+ * type declaration, "this" points to an incomplete type, whereas in
+ * the function body the type is complete.  That doesn't justify the
+ * error in any way, but is probably what's happening.
+ *
+ * In the case of MSVC, we get around this by using C++14 "decltype(auto)"
+ * which deduces the type from the actual return statement.  For gcc 4.8
+ * we use "+this" instead of "this" which produces an rvalue that seems
+ * to be deduced as the same type with this particular compiler, and seem
+ * to be fine as default code path as well.
+ */
+#ifdef _MSC_VER
+/* https://github.com/harfbuzz/harfbuzz/issues/1730 */ \
+#define HB_PARTIALIZE(Pos) \
+  template <typename _T> \
+  decltype(auto) operator () (_T&& _v) const \
+  { return hb_partial<Pos> (this, hb_forward<_T> (_v)); } \
+  static_assert (true, "")
+#else
+/* https://github.com/harfbuzz/harfbuzz/issues/1724 */
+#define HB_PARTIALIZE(Pos) \
+  template <typename _T> \
+  auto operator () (_T&& _v) const HB_AUTO_RETURN \
+  (hb_partial<Pos> (+this, hb_forward<_T> (_v))) \
+  static_assert (true, "")
+#endif
+
+
+struct
+{
+  private:
+
+  template <typename Pred, typename Val> auto
+  impl (Pred&& p, Val &&v, hb_priority<1>) const HB_AUTO_RETURN
+  (hb_deref (hb_forward<Pred> (p)).has (hb_forward<Val> (v)))
+
+  template <typename Pred, typename Val> auto
+  impl (Pred&& p, Val &&v, hb_priority<0>) const HB_AUTO_RETURN
+  (
+    hb_invoke (hb_forward<Pred> (p),
+	       hb_forward<Val> (v))
+  )
+
+  public:
+
+  template <typename Pred, typename Val> auto
+  operator () (Pred&& p, Val &&v) const HB_RETURN (bool,
+    impl (hb_forward<Pred> (p),
+	  hb_forward<Val> (v),
+	  hb_prioritize)
+  )
+}
+HB_FUNCOBJ (hb_has);
+
+struct
+{
+  private:
+
+  template <typename Pred, typename Val> auto
+  impl (Pred&& p, Val &&v, hb_priority<1>) const HB_AUTO_RETURN
+  (
+    hb_has (hb_forward<Pred> (p),
+	    hb_forward<Val> (v))
+  )
+
+  template <typename Pred, typename Val> auto
+  impl (Pred&& p, Val &&v, hb_priority<0>) const HB_AUTO_RETURN
+  (
+    hb_forward<Pred> (p) == hb_forward<Val> (v)
+  )
+
+  public:
+
+  template <typename Pred, typename Val> auto
+  operator () (Pred&& p, Val &&v) const HB_RETURN (bool,
+    impl (hb_forward<Pred> (p),
+	  hb_forward<Val> (v),
+	  hb_prioritize)
+  )
+}
+HB_FUNCOBJ (hb_match);
+
+struct
+{
+  private:
+
+  template <typename Proj, typename Val> auto
+  impl (Proj&& f, Val &&v, hb_priority<2>) const HB_AUTO_RETURN
+  (hb_deref (hb_forward<Proj> (f)).get (hb_forward<Val> (v)))
+
+  template <typename Proj, typename Val> auto
+  impl (Proj&& f, Val &&v, hb_priority<1>) const HB_AUTO_RETURN
+  (
+    hb_invoke (hb_forward<Proj> (f),
+	       hb_forward<Val> (v))
+  )
+
+  template <typename Proj, typename Val> auto
+  impl (Proj&& f, Val &&v, hb_priority<0>) const HB_AUTO_RETURN
+  (
+    hb_forward<Proj> (f)[hb_forward<Val> (v)]
+  )
+
+  public:
+
+  template <typename Proj, typename Val> auto
+  operator () (Proj&& f, Val &&v) const HB_AUTO_RETURN
+  (
+    impl (hb_forward<Proj> (f),
+	  hb_forward<Val> (v),
+	  hb_prioritize)
+  )
+}
+HB_FUNCOBJ (hb_get);
+
+
+template <typename T1, typename T2>
+struct hb_pair_t
+{
+  typedef T1 first_t;
+  typedef T2 second_t;
+  typedef hb_pair_t<T1, T2> pair_t;
+
+  hb_pair_t (T1 a, T2 b) : first (a), second (b) {}
+
+  template <typename Q1, typename Q2,
+	    hb_enable_if (hb_is_convertible (T1, Q1) &&
+			  hb_is_convertible (T2, T2))>
+  operator hb_pair_t<Q1, Q2> () { return hb_pair_t<Q1, Q2> (first, second); }
+
+  hb_pair_t<T1, T2> reverse () const
+  { return hb_pair_t<T1, T2> (second, first); }
+
+  bool operator == (const pair_t& o) const { return first == o.first && second == o.second; }
+  bool operator != (const pair_t& o) const { return !(*this == o); }
+  bool operator < (const pair_t& o) const { return first < o.first || (first == o.first && second < o.second); }
+  bool operator >= (const pair_t& o) const { return !(*this < o); }
+  bool operator > (const pair_t& o) const { return first > o.first || (first == o.first && second > o.second); }
+  bool operator <= (const pair_t& o) const { return !(*this > o); }
+
+  T1 first;
+  T2 second;
+};
+#define hb_pair_t(T1,T2) hb_pair_t<T1, T2>
+template <typename T1, typename T2> static inline hb_pair_t<T1, T2>
+hb_pair (T1&& a, T2&& b) { return hb_pair_t<T1, T2> (a, b); }
+
+struct
+{
+  template <typename Pair> constexpr typename Pair::first_t
+  operator () (const Pair& pair) const { return pair.first; }
+}
+HB_FUNCOBJ (hb_first);
+
+struct
+{
+  template <typename Pair> constexpr typename Pair::second_t
+  operator () (const Pair& pair) const { return pair.second; }
+}
+HB_FUNCOBJ (hb_second);
+
+/* Note.  In min/max impl, we can use hb_type_identity<T> for second argument.
+ * However, that would silently convert between different-signedness integers.
+ * Instead we accept two different types, such that compiler can err if
+ * comparing integers of different signedness. */
+struct
+{
+  template <typename T, typename T2> constexpr auto
+  operator () (T&& a, T2&& b) const HB_AUTO_RETURN
+  (hb_forward<T> (a) <= hb_forward<T2> (b) ? hb_forward<T> (a) : hb_forward<T2> (b))
+}
+HB_FUNCOBJ (hb_min);
+struct
+{
+  template <typename T, typename T2> constexpr auto
+  operator () (T&& a, T2&& b) const HB_AUTO_RETURN
+  (hb_forward<T> (a) >= hb_forward<T2> (b) ? hb_forward<T> (a) : hb_forward<T2> (b))
+}
+HB_FUNCOBJ (hb_max);
+struct
+{
+  template <typename T, typename T2, typename T3> constexpr auto
+  operator () (T&& x, T2&& min, T3&& max) const HB_AUTO_RETURN
+  (hb_min (hb_max (hb_forward<T> (x), hb_forward<T2> (min)), hb_forward<T3> (max)))
+}
+HB_FUNCOBJ (hb_clamp);
+
+
+/*
+ * Bithacks.
+ */
+
+/* Return the number of 1 bits in v. */
+template <typename T>
+static inline HB_CONST_FUNC unsigned int
+hb_popcount (T v)
+{
+#if (defined(__GNUC__) && (__GNUC__ >= 4)) || defined(__clang__)
+  if (sizeof (T) <= sizeof (unsigned int))
+    return __builtin_popcount (v);
+
+  if (sizeof (T) <= sizeof (unsigned long))
+    return __builtin_popcountl (v);
+
+  if (sizeof (T) <= sizeof (unsigned long long))
+    return __builtin_popcountll (v);
+#endif
+
+  if (sizeof (T) <= 4)
+  {
+    /* "HACKMEM 169" */
+    uint32_t y;
+    y = (v >> 1) &033333333333;
+    y = v - y - ((y >>1) & 033333333333);
+    return (((y + (y >> 3)) & 030707070707) % 077);
+  }
+
+  if (sizeof (T) == 8)
+  {
+    unsigned int shift = 32;
+    return hb_popcount<uint32_t> ((uint32_t) v) + hb_popcount ((uint32_t) (v >> shift));
+  }
+
+  if (sizeof (T) == 16)
+  {
+    unsigned int shift = 64;
+    return hb_popcount<uint64_t> ((uint64_t) v) + hb_popcount ((uint64_t) (v >> shift));
+  }
+
+  assert (0);
+  return 0; /* Shut up stupid compiler. */
+}
+
+/* Returns the number of bits needed to store number */
+template <typename T>
+static inline HB_CONST_FUNC unsigned int
+hb_bit_storage (T v)
+{
+  if (unlikely (!v)) return 0;
+
+#if (defined(__GNUC__) && (__GNUC__ >= 4)) || defined(__clang__)
+  if (sizeof (T) <= sizeof (unsigned int))
+    return sizeof (unsigned int) * 8 - __builtin_clz (v);
+
+  if (sizeof (T) <= sizeof (unsigned long))
+    return sizeof (unsigned long) * 8 - __builtin_clzl (v);
+
+  if (sizeof (T) <= sizeof (unsigned long long))
+    return sizeof (unsigned long long) * 8 - __builtin_clzll (v);
+#endif
+
+#if (defined(_MSC_VER) && _MSC_VER >= 1500) || (defined(__MINGW32__) && (__GNUC__ < 4))
+  if (sizeof (T) <= sizeof (unsigned int))
+  {
+    unsigned long where;
+    _BitScanReverse (&where, v);
+    return 1 + where;
+  }
+# if defined(_WIN64)
+  if (sizeof (T) <= 8)
+  {
+    unsigned long where;
+    _BitScanReverse64 (&where, v);
+    return 1 + where;
+  }
+# endif
+#endif
+
+  if (sizeof (T) <= 4)
+  {
+    /* "bithacks" */
+    const unsigned int b[] = {0x2, 0xC, 0xF0, 0xFF00, 0xFFFF0000};
+    const unsigned int S[] = {1, 2, 4, 8, 16};
+    unsigned int r = 0;
+    for (int i = 4; i >= 0; i--)
+      if (v & b[i])
+      {
+	v >>= S[i];
+	r |= S[i];
+      }
+    return r + 1;
+  }
+  if (sizeof (T) <= 8)
+  {
+    /* "bithacks" */
+    const uint64_t b[] = {0x2ULL, 0xCULL, 0xF0ULL, 0xFF00ULL, 0xFFFF0000ULL, 0xFFFFFFFF00000000ULL};
+    const unsigned int S[] = {1, 2, 4, 8, 16, 32};
+    unsigned int r = 0;
+    for (int i = 5; i >= 0; i--)
+      if (v & b[i])
+      {
+	v >>= S[i];
+	r |= S[i];
+      }
+    return r + 1;
+  }
+  if (sizeof (T) == 16)
+  {
+    unsigned int shift = 64;
+    return (v >> shift) ? hb_bit_storage<uint64_t> ((uint64_t) (v >> shift)) + shift :
+			  hb_bit_storage<uint64_t> ((uint64_t) v);
+  }
+
+  assert (0);
+  return 0; /* Shut up stupid compiler. */
+}
+
+/* Returns the number of zero bits in the least significant side of v */
+template <typename T>
+static inline HB_CONST_FUNC unsigned int
+hb_ctz (T v)
+{
+  if (unlikely (!v)) return 8 * sizeof (T);
+
+#if (defined(__GNUC__) && (__GNUC__ >= 4)) || defined(__clang__)
+  if (sizeof (T) <= sizeof (unsigned int))
+    return __builtin_ctz (v);
+
+  if (sizeof (T) <= sizeof (unsigned long))
+    return __builtin_ctzl (v);
+
+  if (sizeof (T) <= sizeof (unsigned long long))
+    return __builtin_ctzll (v);
+#endif
+
+#if (defined(_MSC_VER) && _MSC_VER >= 1500) || (defined(__MINGW32__) && (__GNUC__ < 4))
+  if (sizeof (T) <= sizeof (unsigned int))
+  {
+    unsigned long where;
+    _BitScanForward (&where, v);
+    return where;
+  }
+# if defined(_WIN64)
+  if (sizeof (T) <= 8)
+  {
+    unsigned long where;
+    _BitScanForward64 (&where, v);
+    return where;
+  }
+# endif
+#endif
+
+  if (sizeof (T) <= 4)
+  {
+    /* "bithacks" */
+    unsigned int c = 32;
+    v &= - (int32_t) v;
+    if (v) c--;
+    if (v & 0x0000FFFF) c -= 16;
+    if (v & 0x00FF00FF) c -= 8;
+    if (v & 0x0F0F0F0F) c -= 4;
+    if (v & 0x33333333) c -= 2;
+    if (v & 0x55555555) c -= 1;
+    return c;
+  }
+  if (sizeof (T) <= 8)
+  {
+    /* "bithacks" */
+    unsigned int c = 64;
+    v &= - (int64_t) (v);
+    if (v) c--;
+    if (v & 0x00000000FFFFFFFFULL) c -= 32;
+    if (v & 0x0000FFFF0000FFFFULL) c -= 16;
+    if (v & 0x00FF00FF00FF00FFULL) c -= 8;
+    if (v & 0x0F0F0F0F0F0F0F0FULL) c -= 4;
+    if (v & 0x3333333333333333ULL) c -= 2;
+    if (v & 0x5555555555555555ULL) c -= 1;
+    return c;
+  }
+  if (sizeof (T) == 16)
+  {
+    unsigned int shift = 64;
+    return (uint64_t) v ? hb_bit_storage<uint64_t> ((uint64_t) v) :
+			  hb_bit_storage<uint64_t> ((uint64_t) (v >> shift)) + shift;
+  }
+
+  assert (0);
+  return 0; /* Shut up stupid compiler. */
+}
+
+
+/*
+ * Tiny stuff.
+ */
+
+/* ASCII tag/character handling */
+static inline bool ISALPHA (unsigned char c)
+{ return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'); }
+static inline bool ISALNUM (unsigned char c)
+{ return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || (c >= '0' && c <= '9'); }
+static inline bool ISSPACE (unsigned char c)
+{ return c == ' ' || c =='\f'|| c =='\n'|| c =='\r'|| c =='\t'|| c =='\v'; }
+static inline unsigned char TOUPPER (unsigned char c)
+{ return (c >= 'a' && c <= 'z') ? c - 'a' + 'A' : c; }
+static inline unsigned char TOLOWER (unsigned char c)
+{ return (c >= 'A' && c <= 'Z') ? c - 'A' + 'a' : c; }
+static inline bool ISHEX (unsigned char c)
+{ return (c >= '0' && c <= '9') || (c >= 'a' && c <= 'f') || (c >= 'A' && c <= 'F'); }
+static inline unsigned char TOHEX (uint8_t c)
+{ return (c & 0xF) <= 9 ? (c & 0xF) + '0' : (c & 0xF) + 'a' - 10; }
+static inline uint8_t FROMHEX (unsigned char c)
+{ return (c >= '0' && c <= '9') ? c - '0' : TOLOWER (c) - 'a' + 10; }
+
+static inline unsigned int DIV_CEIL (const unsigned int a, unsigned int b)
+{ return (a + (b - 1)) / b; }
+
+
+#undef  ARRAY_LENGTH
+template <typename Type, unsigned int n>
+static inline unsigned int ARRAY_LENGTH (const Type (&)[n]) { return n; }
+/* A const version, but does not detect erratically being called on pointers. */
+#define ARRAY_LENGTH_CONST(__array) ((signed int) (sizeof (__array) / sizeof (__array[0])))
+
+
+static inline int
+hb_memcmp (const void *a, const void *b, unsigned int len)
+{
+  /* It's illegal to pass NULL to memcmp(), even if len is zero.
+   * So, wrap it.
+   * https://sourceware.org/bugzilla/show_bug.cgi?id=23878 */
+  if (unlikely (!len)) return 0;
+  return memcmp (a, b, len);
+}
+
+static inline void *
+hb_memset (void *s, int c, unsigned int n)
+{
+  /* It's illegal to pass NULL to memset(), even if n is zero. */
+  if (unlikely (!n)) return 0;
+  return memset (s, c, n);
+}
+
+static inline unsigned int
+hb_ceil_to_4 (unsigned int v)
+{
+  return ((v - 1) | 3) + 1;
+}
+
+template <typename T> static inline bool
+hb_in_range (T u, T lo, T hi)
+{
+  static_assert (!hb_is_signed<T>::value, "");
+
+  /* The casts below are important as if T is smaller than int,
+   * the subtract results will become a signed int! */
+  return (T)(u - lo) <= (T)(hi - lo);
+}
+template <typename T> static inline bool
+hb_in_ranges (T u, T lo1, T hi1, T lo2, T hi2)
+{
+  return hb_in_range (u, lo1, hi1) || hb_in_range (u, lo2, hi2);
+}
+template <typename T> static inline bool
+hb_in_ranges (T u, T lo1, T hi1, T lo2, T hi2, T lo3, T hi3)
+{
+  return hb_in_range (u, lo1, hi1) || hb_in_range (u, lo2, hi2) || hb_in_range (u, lo3, hi3);
+}
+
+
+/*
+ * Overflow checking.
+ */
+
+/* Consider __builtin_mul_overflow use here also */
+static inline bool
+hb_unsigned_mul_overflows (unsigned int count, unsigned int size)
+{
+  return (size > 0) && (count >= ((unsigned int) -1) / size);
+}
+
+
+/*
+ * Sort and search.
+ */
+
+template <typename K, typename V, typename ...Ts>
+static int
+_hb_cmp_method (const void *pkey, const void *pval, Ts... ds)
+{
+  const K& key = * (const K*) pkey;
+  const V& val = * (const V*) pval;
+
+  return val.cmp (key, ds...);
+}
+
+template <typename V, typename K, typename ...Ts>
+static inline bool
+hb_bsearch_impl (unsigned *pos, /* Out */
+		 const K& key,
+		 V* base, size_t nmemb, size_t stride,
+		 int (*compar)(const void *_key, const void *_item, Ts... _ds),
+		 Ts... ds)
+{
+  /* This is our *only* bsearch implementation. */
+
+  int min = 0, max = (int) nmemb - 1;
+  while (min <= max)
+  {
+    int mid = ((unsigned int) min + (unsigned int) max) / 2;
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+    V* p = (V*) (((const char *) base) + (mid * stride));
+#pragma GCC diagnostic pop
+    int c = compar ((const void *) hb_addressof (key), (const void *) p, ds...);
+    if (c < 0)
+      max = mid - 1;
+    else if (c > 0)
+      min = mid + 1;
+    else
+    {
+      *pos = mid;
+      return true;
+    }
+  }
+  *pos = min;
+  return false;
+}
+
+template <typename V, typename K>
+static inline V*
+hb_bsearch (const K& key, V* base,
+	    size_t nmemb, size_t stride = sizeof (V),
+	    int (*compar)(const void *_key, const void *_item) = _hb_cmp_method<K, V>)
+{
+  unsigned pos;
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+  return hb_bsearch_impl (&pos, key, base, nmemb, stride, compar) ?
+	 (V*) (((const char *) base) + (pos * stride)) : nullptr;
+#pragma GCC diagnostic pop
+}
+template <typename V, typename K, typename ...Ts>
+static inline V*
+hb_bsearch (const K& key, V* base,
+	    size_t nmemb, size_t stride,
+	    int (*compar)(const void *_key, const void *_item, Ts... _ds),
+	    Ts... ds)
+{
+  unsigned pos;
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+  return hb_bsearch_impl (&pos, key, base, nmemb, stride, compar, ds...) ?
+	 (V*) (((const char *) base) + (pos * stride)) : nullptr;
+#pragma GCC diagnostic pop
+}
+
+
+/* From https://github.com/noporpoise/sort_r
+   Feb 5, 2019 (c8c65c1e)
+   Modified to support optional argument using templates */
+
+/* Isaac Turner 29 April 2014 Public Domain */
+
+/*
+hb_qsort function to be exported.
+Parameters:
+  base is the array to be sorted
+  nel is the number of elements in the array
+  width is the size in bytes of each element of the array
+  compar is the comparison function
+  arg (optional) is a pointer to be passed to the comparison function
+
+void hb_qsort(void *base, size_t nel, size_t width,
+              int (*compar)(const void *_a, const void *_b, [void *_arg]),
+              [void *arg]);
+*/
+
+#define SORT_R_SWAP(a,b,tmp) ((tmp) = (a), (a) = (b), (b) = (tmp))
+
+/* swap a and b */
+/* a and b must not be equal! */
+static inline void sort_r_swap(char *__restrict a, char *__restrict b,
+                               size_t w)
+{
+  char tmp, *end = a+w;
+  for(; a < end; a++, b++) { SORT_R_SWAP(*a, *b, tmp); }
+}
+
+/* swap a, b iff a>b */
+/* a and b must not be equal! */
+/* __restrict is same as restrict but better support on old machines */
+template <typename ...Ts>
+static inline int sort_r_cmpswap(char *__restrict a,
+                                 char *__restrict b, size_t w,
+                                 int (*compar)(const void *_a,
+                                               const void *_b,
+                                               Ts... _ds),
+                                 Ts... ds)
+{
+  if(compar(a, b, ds...) > 0) {
+    sort_r_swap(a, b, w);
+    return 1;
+  }
+  return 0;
+}
+
+/*
+Swap consecutive blocks of bytes of size na and nb starting at memory addr ptr,
+with the smallest swap so that the blocks are in the opposite order. Blocks may
+be internally re-ordered e.g.
+  12345ab  ->   ab34512
+  123abc   ->   abc123
+  12abcde  ->   deabc12
+*/
+static inline void sort_r_swap_blocks(char *ptr, size_t na, size_t nb)
+{
+  if(na > 0 && nb > 0) {
+    if(na > nb) { sort_r_swap(ptr, ptr+na, nb); }
+    else { sort_r_swap(ptr, ptr+nb, na); }
+  }
+}
+
+/* Implement recursive quicksort ourselves */
+/* Note: quicksort is not stable, equivalent values may be swapped */
+template <typename ...Ts>
+static inline void sort_r_simple(void *base, size_t nel, size_t w,
+                                 int (*compar)(const void *_a,
+                                               const void *_b,
+                                               Ts... _ds),
+                                 Ts... ds)
+{
+  char *b = (char *)base, *end = b + nel*w;
+
+  /* for(size_t i=0; i<nel; i++) {printf("%4i", *(int*)(b + i*sizeof(int)));}
+  printf("\n"); */
+
+  if(nel < 10) {
+    /* Insertion sort for arbitrarily small inputs */
+    char *pi, *pj;
+    for(pi = b+w; pi < end; pi += w) {
+      for(pj = pi; pj > b && sort_r_cmpswap(pj-w,pj,w,compar,ds...); pj -= w) {}
+    }
+  }
+  else
+  {
+    /* nel > 9; Quicksort */
+
+    int cmp;
+    char *pl, *ple, *pr, *pre, *pivot;
+    char *last = b+w*(nel-1), *tmp;
+
+    /*
+    Use median of second, middle and second-last items as pivot.
+    First and last may have been swapped with pivot and therefore be extreme
+    */
+    char *l[3];
+    l[0] = b + w;
+    l[1] = b+w*(nel/2);
+    l[2] = last - w;
+
+    /* printf("pivots: %i, %i, %i\n", *(int*)l[0], *(int*)l[1], *(int*)l[2]); */
+
+    if(compar(l[0],l[1],ds...) > 0) { SORT_R_SWAP(l[0], l[1], tmp); }
+    if(compar(l[1],l[2],ds...) > 0) {
+      SORT_R_SWAP(l[1], l[2], tmp);
+      if(compar(l[0],l[1],ds...) > 0) { SORT_R_SWAP(l[0], l[1], tmp); }
+    }
+
+    /* swap mid value (l[1]), and last element to put pivot as last element */
+    if(l[1] != last) { sort_r_swap(l[1], last, w); }
+
+    /*
+    pl is the next item on the left to be compared to the pivot
+    pr is the last item on the right that was compared to the pivot
+    ple is the left position to put the next item that equals the pivot
+    ple is the last right position where we put an item that equals the pivot
+                                           v- end (beyond the array)
+      EEEEEELLLLLLLLuuuuuuuuGGGGGGGEEEEEEEE.
+      ^- b  ^- ple  ^- pl   ^- pr  ^- pre ^- last (where the pivot is)
+    Pivot comparison key:
+      E = equal, L = less than, u = unknown, G = greater than, E = equal
+    */
+    pivot = last;
+    ple = pl = b;
+    pre = pr = last;
+
+    /*
+    Strategy:
+    Loop into the list from the left and right at the same time to find:
+    - an item on the left that is greater than the pivot
+    - an item on the right that is less than the pivot
+    Once found, they are swapped and the loop continues.
+    Meanwhile items that are equal to the pivot are moved to the edges of the
+    array.
+    */
+    while(pl < pr) {
+      /* Move left hand items which are equal to the pivot to the far left.
+         break when we find an item that is greater than the pivot */
+      for(; pl < pr; pl += w) {
+        cmp = compar(pl, pivot, ds...);
+        if(cmp > 0) { break; }
+        else if(cmp == 0) {
+          if(ple < pl) { sort_r_swap(ple, pl, w); }
+          ple += w;
+        }
+      }
+      /* break if last batch of left hand items were equal to pivot */
+      if(pl >= pr) { break; }
+      /* Move right hand items which are equal to the pivot to the far right.
+         break when we find an item that is less than the pivot */
+      for(; pl < pr; ) {
+        pr -= w; /* Move right pointer onto an unprocessed item */
+        cmp = compar(pr, pivot, ds...);
+        if(cmp == 0) {
+          pre -= w;
+          if(pr < pre) { sort_r_swap(pr, pre, w); }
+        }
+        else if(cmp < 0) {
+          if(pl < pr) { sort_r_swap(pl, pr, w); }
+          pl += w;
+          break;
+        }
+      }
+    }
+
+    pl = pr; /* pr may have gone below pl */
+
+    /*
+    Now we need to go from: EEELLLGGGGEEEE
+                        to: LLLEEEEEEEGGGG
+    Pivot comparison key:
+      E = equal, L = less than, u = unknown, G = greater than, E = equal
+    */
+    sort_r_swap_blocks(b, ple-b, pl-ple);
+    sort_r_swap_blocks(pr, pre-pr, end-pre);
+
+    /*for(size_t i=0; i<nel; i++) {printf("%4i", *(int*)(b + i*sizeof(int)));}
+    printf("\n");*/
+
+    sort_r_simple(b, (pl-ple)/w, w, compar, ds...);
+    sort_r_simple(end-(pre-pr), (pre-pr)/w, w, compar, ds...);
+  }
+}
+
+static inline void
+hb_qsort (void *base, size_t nel, size_t width,
+	  int (*compar)(const void *_a, const void *_b))
+{
+#if defined(__OPTIMIZE_SIZE__) && !defined(HB_USE_INTERNAL_QSORT)
+  qsort (base, nel, width, compar);
+#else
+  sort_r_simple (base, nel, width, compar);
+#endif
+}
+
+static inline void
+hb_qsort (void *base, size_t nel, size_t width,
+	  int (*compar)(const void *_a, const void *_b, void *_arg),
+	  void *arg)
+{
+#ifdef HAVE_GNU_QSORT_R
+  qsort_r (base, nel, width, compar, arg);
+#else
+  sort_r_simple (base, nel, width, compar, arg);
+#endif
+}
+
+
+template <typename T, typename T2, typename T3> static inline void
+hb_stable_sort (T *array, unsigned int len, int(*compar)(const T2 *, const T2 *), T3 *array2)
+{
+  for (unsigned int i = 1; i < len; i++)
+  {
+    unsigned int j = i;
+    while (j && compar (&array[j - 1], &array[i]) > 0)
+      j--;
+    if (i == j)
+      continue;
+    /* Move item i to occupy place for item j, shift what's in between. */
+    {
+      T t = array[i];
+      memmove (&array[j + 1], &array[j], (i - j) * sizeof (T));
+      array[j] = t;
+    }
+    if (array2)
+    {
+      T3 t = array2[i];
+      memmove (&array2[j + 1], &array2[j], (i - j) * sizeof (T3));
+      array2[j] = t;
+    }
+  }
+}
+
+template <typename T> static inline void
+hb_stable_sort (T *array, unsigned int len, int(*compar)(const T *, const T *))
+{
+  hb_stable_sort (array, len, compar, (int *) nullptr);
+}
+
+static inline hb_bool_t
+hb_codepoint_parse (const char *s, unsigned int len, int base, hb_codepoint_t *out)
+{
+  unsigned int v;
+  const char *p = s;
+  const char *end = p + len;
+  if (unlikely (!hb_parse_uint (&p, end, &v, true/* whole buffer */, base)))
+    return false;
+
+  *out = v;
+  return true;
+}
+
+
+/* Operators. */
+
+struct hb_bitwise_and
+{ HB_PARTIALIZE(2);
+  static constexpr bool passthru_left = false;
+  static constexpr bool passthru_right = false;
+  template <typename T> constexpr auto
+  operator () (const T &a, const T &b) const HB_AUTO_RETURN (a & b)
+}
+HB_FUNCOBJ (hb_bitwise_and);
+struct hb_bitwise_or
+{ HB_PARTIALIZE(2);
+  static constexpr bool passthru_left = true;
+  static constexpr bool passthru_right = true;
+  template <typename T> constexpr auto
+  operator () (const T &a, const T &b) const HB_AUTO_RETURN (a | b)
+}
+HB_FUNCOBJ (hb_bitwise_or);
+struct hb_bitwise_xor
+{ HB_PARTIALIZE(2);
+  static constexpr bool passthru_left = true;
+  static constexpr bool passthru_right = true;
+  template <typename T> constexpr auto
+  operator () (const T &a, const T &b) const HB_AUTO_RETURN (a ^ b)
+}
+HB_FUNCOBJ (hb_bitwise_xor);
+struct hb_bitwise_sub
+{ HB_PARTIALIZE(2);
+  static constexpr bool passthru_left = true;
+  static constexpr bool passthru_right = false;
+  template <typename T> constexpr auto
+  operator () (const T &a, const T &b) const HB_AUTO_RETURN (a & ~b)
+}
+HB_FUNCOBJ (hb_bitwise_sub);
+struct
+{
+  template <typename T> constexpr auto
+  operator () (const T &a) const HB_AUTO_RETURN (~a)
+}
+HB_FUNCOBJ (hb_bitwise_neg);
+
+struct
+{ HB_PARTIALIZE(2);
+  template <typename T, typename T2> constexpr auto
+  operator () (const T &a, const T2 &b) const HB_AUTO_RETURN (a + b)
+}
+HB_FUNCOBJ (hb_add);
+struct
+{ HB_PARTIALIZE(2);
+  template <typename T, typename T2> constexpr auto
+  operator () (const T &a, const T2 &b) const HB_AUTO_RETURN (a - b)
+}
+HB_FUNCOBJ (hb_sub);
+struct
+{ HB_PARTIALIZE(2);
+  template <typename T, typename T2> constexpr auto
+  operator () (const T &a, const T2 &b) const HB_AUTO_RETURN (a * b)
+}
+HB_FUNCOBJ (hb_mul);
+struct
+{ HB_PARTIALIZE(2);
+  template <typename T, typename T2> constexpr auto
+  operator () (const T &a, const T2 &b) const HB_AUTO_RETURN (a / b)
+}
+HB_FUNCOBJ (hb_div);
+struct
+{ HB_PARTIALIZE(2);
+  template <typename T, typename T2> constexpr auto
+  operator () (const T &a, const T2 &b) const HB_AUTO_RETURN (a % b)
+}
+HB_FUNCOBJ (hb_mod);
+struct
+{
+  template <typename T> constexpr auto
+  operator () (const T &a) const HB_AUTO_RETURN (+a)
+}
+HB_FUNCOBJ (hb_pos);
+struct
+{
+  template <typename T> constexpr auto
+  operator () (const T &a) const HB_AUTO_RETURN (-a)
+}
+HB_FUNCOBJ (hb_neg);
+struct
+{
+  template <typename T> constexpr auto
+  operator () (T &a) const HB_AUTO_RETURN (++a)
+}
+HB_FUNCOBJ (hb_inc);
+struct
+{
+  template <typename T> constexpr auto
+  operator () (T &a) const HB_AUTO_RETURN (--a)
+}
+HB_FUNCOBJ (hb_dec);
+
+
+/* Compiler-assisted vectorization. */
+
+/* Type behaving similar to vectorized vars defined using __attribute__((vector_size(...))),
+ * basically a fixed-size bitset. */
+template <typename elt_t, unsigned int byte_size>
+struct hb_vector_size_t
+{
+  elt_t& operator [] (unsigned int i) { return v[i]; }
+  const elt_t& operator [] (unsigned int i) const { return v[i]; }
+
+  void clear (unsigned char v = 0) { memset (this, v, sizeof (*this)); }
+
+  template <typename Op>
+  hb_vector_size_t process (const Op& op) const
+  {
+    hb_vector_size_t r;
+    for (unsigned int i = 0; i < ARRAY_LENGTH (v); i++)
+      r.v[i] = op (v[i]);
+    return r;
+  }
+  template <typename Op>
+  hb_vector_size_t process (const Op& op, const hb_vector_size_t &o) const
+  {
+    hb_vector_size_t r;
+    for (unsigned int i = 0; i < ARRAY_LENGTH (v); i++)
+      r.v[i] = op (v[i], o.v[i]);
+    return r;
+  }
+  hb_vector_size_t operator | (const hb_vector_size_t &o) const
+  { return process (hb_bitwise_or, o); }
+  hb_vector_size_t operator & (const hb_vector_size_t &o) const
+  { return process (hb_bitwise_and, o); }
+  hb_vector_size_t operator ^ (const hb_vector_size_t &o) const
+  { return process (hb_bitwise_xor, o); }
+  hb_vector_size_t operator ~ () const
+  { return process (hb_bitwise_neg); }
+
+  private:
+  static_assert (0 == byte_size % sizeof (elt_t), "");
+  elt_t v[byte_size / sizeof (elt_t)];
+};
+
+
+#endif /* HB_ALGS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-array.hh b/thirdparty/harfbuzz/src/hb-array.hh
new file mode 100644
index 0000000000..568cd02c79
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-array.hh
@@ -0,0 +1,408 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_ARRAY_HH
+#define HB_ARRAY_HH
+
+#include "hb.hh"
+#include "hb-algs.hh"
+#include "hb-iter.hh"
+#include "hb-null.hh"
+
+
+template <typename Type>
+struct hb_sorted_array_t;
+
+template <typename Type>
+struct hb_array_t : hb_iter_with_fallback_t<hb_array_t<Type>, Type&>
+{
+  /*
+   * Constructors.
+   */
+  hb_array_t () : arrayZ (nullptr), length (0), backwards_length (0) {}
+  hb_array_t (Type *array_, unsigned int length_) : arrayZ (array_), length (length_), backwards_length (0) {}
+  template <unsigned int length_>
+  hb_array_t (Type (&array_)[length_]) : arrayZ (array_), length (length_), backwards_length (0) {}
+
+  template <typename U,
+	    hb_enable_if (hb_is_cr_convertible(U, Type))>
+  hb_array_t (const hb_array_t<U> &o) :
+    hb_iter_with_fallback_t<hb_array_t, Type&> (),
+    arrayZ (o.arrayZ), length (o.length), backwards_length (o.backwards_length) {}
+  template <typename U,
+	    hb_enable_if (hb_is_cr_convertible(U, Type))>
+  hb_array_t& operator = (const hb_array_t<U> &o)
+  { arrayZ = o.arrayZ; length = o.length; backwards_length = o.backwards_length; return *this; }
+
+  /*
+   * Iterator implementation.
+   */
+  typedef Type& __item_t__;
+  static constexpr bool is_random_access_iterator = true;
+  Type& __item_at__ (unsigned i) const
+  {
+    if (unlikely (i >= length)) return CrapOrNull (Type);
+    return arrayZ[i];
+  }
+  void __forward__ (unsigned n)
+  {
+    if (unlikely (n > length))
+      n = length;
+    length -= n;
+    backwards_length += n;
+    arrayZ += n;
+  }
+  void __rewind__ (unsigned n)
+  {
+    if (unlikely (n > backwards_length))
+      n = backwards_length;
+    length += n;
+    backwards_length -= n;
+    arrayZ -= n;
+  }
+  unsigned __len__ () const { return length; }
+  /* Ouch. The operator== compares the contents of the array.  For range-based for loops,
+   * it's best if we can just compare arrayZ, though comparing contents is still fast,
+   * but also would require that Type has operator==.  As such, we optimize this operator
+   * for range-based for loop and just compare arrayZ.  No need to compare length, as we
+   * assume we're only compared to .end(). */
+  bool operator != (const hb_array_t& o) const
+  { return arrayZ != o.arrayZ; }
+
+  /* Extra operators.
+   */
+  Type * operator & () const { return arrayZ; }
+  operator hb_array_t<const Type> () { return hb_array_t<const Type> (arrayZ, length); }
+  template <typename T> operator T * () const { return arrayZ; }
+
+  HB_INTERNAL bool operator == (const hb_array_t &o) const;
+
+  uint32_t hash () const {
+    uint32_t current = 0;
+    for (unsigned int i = 0; i < this->length; i++) {
+      current = current * 31 + hb_hash (this->arrayZ[i]);
+    }
+    return current;
+  }
+
+  /*
+   * Compare, Sort, and Search.
+   */
+
+  /* Note: our compare is NOT lexicographic; it also does NOT call Type::cmp. */
+  int cmp (const hb_array_t &a) const
+  {
+    if (length != a.length)
+      return (int) a.length - (int) length;
+    return hb_memcmp (a.arrayZ, arrayZ, get_size ());
+  }
+  HB_INTERNAL static int cmp (const void *pa, const void *pb)
+  {
+    hb_array_t *a = (hb_array_t *) pa;
+    hb_array_t *b = (hb_array_t *) pb;
+    return b->cmp (*a);
+  }
+
+  template <typename T>
+  Type *lsearch (const T &x, Type *not_found = nullptr)
+  {
+    unsigned i;
+    return lfind (x, &i) ? &this->arrayZ[i] : not_found;
+  }
+  template <typename T>
+  const Type *lsearch (const T &x, const Type *not_found = nullptr) const
+  {
+    unsigned i;
+    return lfind (x, &i) ? &this->arrayZ[i] : not_found;
+  }
+  template <typename T>
+  bool lfind (const T &x, unsigned *pos = nullptr) const
+  {
+    for (unsigned i = 0; i < length; ++i)
+      if (!this->arrayZ[i].cmp (x))
+      {
+	if (pos)
+	  *pos = i;
+	return true;
+      }
+
+    return false;
+  }
+
+  hb_sorted_array_t<Type> qsort (int (*cmp_)(const void*, const void*))
+  {
+    if (likely (length))
+      hb_qsort (arrayZ, length, this->get_item_size (), cmp_);
+    return hb_sorted_array_t<Type> (*this);
+  }
+  hb_sorted_array_t<Type> qsort ()
+  {
+    if (likely (length))
+      hb_qsort (arrayZ, length, this->get_item_size (), Type::cmp);
+    return hb_sorted_array_t<Type> (*this);
+  }
+  void qsort (unsigned int start, unsigned int end)
+  {
+    end = hb_min (end, length);
+    assert (start <= end);
+    if (likely (start < end))
+      hb_qsort (arrayZ + start, end - start, this->get_item_size (), Type::cmp);
+  }
+
+  /*
+   * Other methods.
+   */
+
+  unsigned int get_size () const { return length * this->get_item_size (); }
+
+  /*
+   * Reverse the order of items in this array in the range [start, end).
+   */
+  void reverse (unsigned start = 0, unsigned end = -1)
+  {
+    start = hb_min (start, length);
+    end = hb_min (end, length);
+
+    if (end < start + 2)
+      return;
+
+    for (unsigned lhs = start, rhs = end - 1; lhs < rhs; lhs++, rhs--) {
+      Type temp = arrayZ[rhs];
+      arrayZ[rhs] = arrayZ[lhs];
+      arrayZ[lhs] = temp;
+    }
+  }
+
+  hb_array_t sub_array (unsigned int start_offset = 0, unsigned int *seg_count = nullptr /* IN/OUT */) const
+  {
+    if (!start_offset && !seg_count)
+      return *this;
+
+    unsigned int count = length;
+    if (unlikely (start_offset > count))
+      count = 0;
+    else
+      count -= start_offset;
+    if (seg_count)
+      count = *seg_count = hb_min (count, *seg_count);
+    return hb_array_t (arrayZ + start_offset, count);
+  }
+  hb_array_t sub_array (unsigned int start_offset, unsigned int seg_count) const
+  { return sub_array (start_offset, &seg_count); }
+
+  hb_array_t truncate (unsigned length) const { return sub_array (0, length); }
+
+  template <typename T,
+	    unsigned P = sizeof (Type),
+	    hb_enable_if (P == 1)>
+  const T *as () const
+  { return length < hb_null_size (T) ? &Null (T) : reinterpret_cast<const T *> (arrayZ); }
+
+  template <typename T,
+	    unsigned P = sizeof (Type),
+	    hb_enable_if (P == 1)>
+  bool check_range (const T *p, unsigned int size = T::static_size) const
+  {
+    return arrayZ <= ((const char *) p)
+	&& ((const char *) p) <= arrayZ + length
+	&& (unsigned int) (arrayZ + length - (const char *) p) >= size;
+  }
+
+  /* Only call if you allocated the underlying array using malloc() or similar. */
+  void free ()
+  { ::free ((void *) arrayZ); arrayZ = nullptr; length = 0; }
+
+  template <typename hb_serialize_context_t>
+  hb_array_t copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    auto* out = c->start_embed (arrayZ);
+    if (unlikely (!c->extend_size (out, get_size ()))) return_trace (hb_array_t ());
+    for (unsigned i = 0; i < length; i++)
+      out[i] = arrayZ[i]; /* TODO: add version that calls c->copy() */
+    return_trace (hb_array_t (out, length));
+  }
+
+  template <typename hb_sanitize_context_t>
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return c->check_array (arrayZ, length); }
+
+  /*
+   * Members
+   */
+
+  public:
+  Type *arrayZ;
+  unsigned int length;
+  unsigned int backwards_length;
+};
+template <typename T> inline hb_array_t<T>
+hb_array (T *array, unsigned int length)
+{ return hb_array_t<T> (array, length); }
+template <typename T, unsigned int length_> inline hb_array_t<T>
+hb_array (T (&array_)[length_])
+{ return hb_array_t<T> (array_); }
+
+enum hb_bfind_not_found_t
+{
+  HB_BFIND_NOT_FOUND_DONT_STORE,
+  HB_BFIND_NOT_FOUND_STORE,
+  HB_BFIND_NOT_FOUND_STORE_CLOSEST,
+};
+
+template <typename Type>
+struct hb_sorted_array_t :
+	hb_iter_t<hb_sorted_array_t<Type>, Type&>,
+	hb_array_t<Type>
+{
+  typedef hb_iter_t<hb_sorted_array_t, Type&> iter_base_t;
+  HB_ITER_USING (iter_base_t);
+  static constexpr bool is_random_access_iterator = true;
+  static constexpr bool is_sorted_iterator = true;
+
+  hb_sorted_array_t () : hb_array_t<Type> () {}
+  hb_sorted_array_t (Type *array_, unsigned int length_) : hb_array_t<Type> (array_, length_) {}
+  template <unsigned int length_>
+  hb_sorted_array_t (Type (&array_)[length_]) : hb_array_t<Type> (array_) {}
+
+  template <typename U,
+	    hb_enable_if (hb_is_cr_convertible(U, Type))>
+  hb_sorted_array_t (const hb_array_t<U> &o) :
+    hb_iter_t<hb_sorted_array_t, Type&> (),
+    hb_array_t<Type> (o) {}
+  template <typename U,
+	    hb_enable_if (hb_is_cr_convertible(U, Type))>
+  hb_sorted_array_t& operator = (const hb_array_t<U> &o)
+  { hb_array_t<Type> (*this) = o; return *this; }
+
+  /* Iterator implementation. */
+  bool operator != (const hb_sorted_array_t& o) const
+  { return this->arrayZ != o.arrayZ || this->length != o.length; }
+
+  hb_sorted_array_t sub_array (unsigned int start_offset, unsigned int *seg_count /* IN/OUT */) const
+  { return hb_sorted_array_t (((const hb_array_t<Type> *) (this))->sub_array (start_offset, seg_count)); }
+  hb_sorted_array_t sub_array (unsigned int start_offset, unsigned int seg_count) const
+  { return sub_array (start_offset, &seg_count); }
+
+  hb_sorted_array_t truncate (unsigned length) const { return sub_array (0, length); }
+
+  template <typename T>
+  Type *bsearch (const T &x, Type *not_found = nullptr)
+  {
+    unsigned int i;
+    return bfind (x, &i) ? &this->arrayZ[i] : not_found;
+  }
+  template <typename T>
+  const Type *bsearch (const T &x, const Type *not_found = nullptr) const
+  {
+    unsigned int i;
+    return bfind (x, &i) ? &this->arrayZ[i] : not_found;
+  }
+  template <typename T>
+  bool bfind (const T &x, unsigned int *i = nullptr,
+	      hb_bfind_not_found_t not_found = HB_BFIND_NOT_FOUND_DONT_STORE,
+	      unsigned int to_store = (unsigned int) -1) const
+  {
+    unsigned pos;
+
+    if (bsearch_impl (x, &pos))
+    {
+      if (i)
+	*i = pos;
+      return true;
+    }
+
+    if (i)
+    {
+      switch (not_found)
+      {
+	case HB_BFIND_NOT_FOUND_DONT_STORE:
+	  break;
+
+	case HB_BFIND_NOT_FOUND_STORE:
+	  *i = to_store;
+	  break;
+
+	case HB_BFIND_NOT_FOUND_STORE_CLOSEST:
+	  *i = pos;
+	  break;
+      }
+    }
+    return false;
+  }
+  template <typename T>
+  bool bsearch_impl (const T &x, unsigned *pos) const
+  {
+    return hb_bsearch_impl (pos,
+			    x,
+			    this->arrayZ,
+			    this->length,
+			    sizeof (Type),
+			    _hb_cmp_method<T, Type>);
+  }
+};
+template <typename T> inline hb_sorted_array_t<T>
+hb_sorted_array (T *array, unsigned int length)
+{ return hb_sorted_array_t<T> (array, length); }
+template <typename T, unsigned int length_> inline hb_sorted_array_t<T>
+hb_sorted_array (T (&array_)[length_])
+{ return hb_sorted_array_t<T> (array_); }
+
+template <typename T>
+bool hb_array_t<T>::operator == (const hb_array_t<T> &o) const
+{
+  if (o.length != this->length) return false;
+  for (unsigned int i = 0; i < this->length; i++) {
+    if (this->arrayZ[i] != o.arrayZ[i]) return false;
+  }
+  return true;
+}
+
+/* TODO Specialize opeator== for hb_bytes_t and hb_ubytes_t. */
+
+template <>
+inline uint32_t hb_array_t<const char>::hash () const {
+  uint32_t current = 0;
+  for (unsigned int i = 0; i < this->length; i++)
+    current = current * 31 + (uint32_t) (this->arrayZ[i] * 2654435761u);
+  return current;
+}
+
+template <>
+inline uint32_t hb_array_t<const unsigned char>::hash () const {
+  uint32_t current = 0;
+  for (unsigned int i = 0; i < this->length; i++)
+    current = current * 31 + (uint32_t) (this->arrayZ[i] * 2654435761u);
+  return current;
+}
+
+
+typedef hb_array_t<const char> hb_bytes_t;
+typedef hb_array_t<const unsigned char> hb_ubytes_t;
+
+
+
+#endif /* HB_ARRAY_HH */
diff --git a/thirdparty/harfbuzz/src/hb-atomic.hh b/thirdparty/harfbuzz/src/hb-atomic.hh
new file mode 100644
index 0000000000..b3fb296b4e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-atomic.hh
@@ -0,0 +1,295 @@
+/*
+ * Copyright © 2007  Chris Wilson
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Contributor(s):
+ *	Chris Wilson <chris@chris-wilson.co.uk>
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_ATOMIC_HH
+#define HB_ATOMIC_HH
+
+#include "hb.hh"
+#include "hb-meta.hh"
+
+
+/*
+ * Atomic integers and pointers.
+ */
+
+
+/* We need external help for these */
+
+#if defined(hb_atomic_int_impl_add) \
+ && defined(hb_atomic_ptr_impl_get) \
+ && defined(hb_atomic_ptr_impl_cmpexch)
+
+/* Defined externally, i.e. in config.h. */
+
+
+#elif !defined(HB_NO_MT) && defined(__ATOMIC_ACQUIRE)
+
+/* C++11-style GCC primitives. */
+
+#define _hb_memory_barrier()			__sync_synchronize ()
+
+#define hb_atomic_int_impl_add(AI, V)		__atomic_fetch_add ((AI), (V), __ATOMIC_ACQ_REL)
+#define hb_atomic_int_impl_set_relaxed(AI, V)	__atomic_store_n ((AI), (V), __ATOMIC_RELAXED)
+#define hb_atomic_int_impl_set(AI, V)		__atomic_store_n ((AI), (V), __ATOMIC_RELEASE)
+#define hb_atomic_int_impl_get_relaxed(AI)	__atomic_load_n ((AI), __ATOMIC_RELAXED)
+#define hb_atomic_int_impl_get(AI)		__atomic_load_n ((AI), __ATOMIC_ACQUIRE)
+
+#define hb_atomic_ptr_impl_set_relaxed(P, V)	__atomic_store_n ((P), (V), __ATOMIC_RELAXED)
+#define hb_atomic_ptr_impl_get_relaxed(P)	__atomic_load_n ((P), __ATOMIC_RELAXED)
+#define hb_atomic_ptr_impl_get(P)		__atomic_load_n ((P), __ATOMIC_ACQUIRE)
+static inline bool
+_hb_atomic_ptr_impl_cmplexch (const void **P, const void *O_, const void *N)
+{
+  const void *O = O_; // Need lvalue
+  return __atomic_compare_exchange_n ((void **) P, (void **) &O, (void *) N, true, __ATOMIC_ACQ_REL, __ATOMIC_RELAXED);
+}
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	_hb_atomic_ptr_impl_cmplexch ((const void **) (P), (O), (N))
+
+#elif !defined(HB_NO_MT) && __cplusplus >= 201103L
+
+/* C++11 atomics. */
+
+#include <atomic>
+
+#define _hb_memory_barrier()			std::atomic_thread_fence(std::memory_order_ack_rel)
+#define _hb_memory_r_barrier()			std::atomic_thread_fence(std::memory_order_acquire)
+#define _hb_memory_w_barrier()			std::atomic_thread_fence(std::memory_order_release)
+
+#define hb_atomic_int_impl_add(AI, V)		(reinterpret_cast<std::atomic<int> *> (AI)->fetch_add ((V), std::memory_order_acq_rel))
+#define hb_atomic_int_impl_set_relaxed(AI, V)	(reinterpret_cast<std::atomic<int> *> (AI)->store ((V), std::memory_order_relaxed))
+#define hb_atomic_int_impl_set(AI, V)		(reinterpret_cast<std::atomic<int> *> (AI)->store ((V), std::memory_order_release))
+#define hb_atomic_int_impl_get_relaxed(AI)	(reinterpret_cast<std::atomic<int> const *> (AI)->load (std::memory_order_relaxed))
+#define hb_atomic_int_impl_get(AI)		(reinterpret_cast<std::atomic<int> const *> (AI)->load (std::memory_order_acquire))
+
+#define hb_atomic_ptr_impl_set_relaxed(P, V)	(reinterpret_cast<std::atomic<void*> *> (P)->store ((V), std::memory_order_relaxed))
+#define hb_atomic_ptr_impl_get_relaxed(P)	(reinterpret_cast<std::atomic<void*> const *> (P)->load (std::memory_order_relaxed))
+#define hb_atomic_ptr_impl_get(P)		(reinterpret_cast<std::atomic<void*> *> (P)->load (std::memory_order_acquire))
+static inline bool
+_hb_atomic_ptr_impl_cmplexch (const void **P, const void *O_, const void *N)
+{
+  const void *O = O_; // Need lvalue
+  return reinterpret_cast<std::atomic<const void*> *> (P)->compare_exchange_weak (O, N, std::memory_order_acq_rel, std::memory_order_relaxed);
+}
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	_hb_atomic_ptr_impl_cmplexch ((const void **) (P), (O), (N))
+
+
+#elif !defined(HB_NO_MT) && defined(_WIN32)
+
+#include <windows.h>
+
+static inline void _hb_memory_barrier ()
+{
+#if !defined(MemoryBarrier) && !defined(__MINGW32_VERSION)
+  /* MinGW has a convoluted history of supporting MemoryBarrier. */
+  LONG dummy = 0;
+  InterlockedExchange (&dummy, 1);
+#else
+  MemoryBarrier ();
+#endif
+}
+#define _hb_memory_barrier()			_hb_memory_barrier ()
+
+#define hb_atomic_int_impl_add(AI, V)		InterlockedExchangeAdd ((LONG *) (AI), (V))
+static_assert ((sizeof (LONG) == sizeof (int)), "");
+
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	(InterlockedCompareExchangePointer ((P), (N), (O)) == (O))
+
+
+#elif !defined(HB_NO_MT) && defined(HAVE_INTEL_ATOMIC_PRIMITIVES)
+
+#define _hb_memory_barrier()			__sync_synchronize ()
+
+#define hb_atomic_int_impl_add(AI, V)		__sync_fetch_and_add ((AI), (V))
+
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	__sync_bool_compare_and_swap ((P), (O), (N))
+
+
+#elif !defined(HB_NO_MT) && defined(HAVE_SOLARIS_ATOMIC_OPS)
+
+#include <atomic.h>
+#include <mbarrier.h>
+
+#define _hb_memory_r_barrier()			__machine_r_barrier ()
+#define _hb_memory_w_barrier()			__machine_w_barrier ()
+#define _hb_memory_barrier()			__machine_rw_barrier ()
+
+static inline int _hb_fetch_and_add (int *AI, int V)
+{
+  _hb_memory_w_barrier ();
+  int result = atomic_add_int_nv ((uint_t *) AI, V) - V;
+  _hb_memory_r_barrier ();
+  return result;
+}
+static inline bool _hb_compare_and_swap_ptr (void **P, void *O, void *N)
+{
+  _hb_memory_w_barrier ();
+  bool result = atomic_cas_ptr (P, O, N) == O;
+  _hb_memory_r_barrier ();
+  return result;
+}
+
+#define hb_atomic_int_impl_add(AI, V)           _hb_fetch_and_add ((AI), (V))
+
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)       _hb_compare_and_swap_ptr ((P), (O), (N))
+
+
+#elif !defined(HB_NO_MT) && defined(__APPLE__)
+
+#include <libkern/OSAtomic.h>
+#ifdef __MAC_OS_X_MIN_REQUIRED
+#include <AvailabilityMacros.h>
+#elif defined(__IPHONE_OS_MIN_REQUIRED)
+#include <Availability.h>
+#endif
+
+#define _hb_memory_barrier()			OSMemoryBarrier ()
+
+#define hb_atomic_int_impl_add(AI, V)		(OSAtomicAdd32Barrier ((V), (AI)) - (V))
+
+#if (MAC_OS_X_VERSION_MIN_REQUIRED > MAC_OS_X_VERSION_10_4 || __IPHONE_VERSION_MIN_REQUIRED >= 20100)
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	OSAtomicCompareAndSwapPtrBarrier ((O), (N), (P))
+#else
+#if __ppc64__ || __x86_64__ || __aarch64__
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	OSAtomicCompareAndSwap64Barrier ((int64_t) (O), (int64_t) (N), (int64_t*) (P))
+#else
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	OSAtomicCompareAndSwap32Barrier ((int32_t) (O), (int32_t) (N), (int32_t*) (P))
+#endif
+#endif
+
+
+#elif !defined(HB_NO_MT) && defined(_AIX) && (defined(__IBMCPP__) || defined(__ibmxl__))
+
+#include <builtins.h>
+
+#define _hb_memory_barrier()			__lwsync ()
+
+static inline int _hb_fetch_and_add (int *AI, int V)
+{
+  _hb_memory_barrier ();
+  int result = __fetch_and_add (AI, V);
+  _hb_memory_barrier ();
+  return result;
+}
+static inline bool _hb_compare_and_swaplp (long *P, long O, long N)
+{
+  _hb_memory_barrier ();
+  bool result = __compare_and_swaplp (P, &O, N);
+  _hb_memory_barrier ();
+  return result;
+}
+
+#define hb_atomic_int_impl_add(AI, V)           _hb_fetch_and_add ((AI), (V))
+
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)       _hb_compare_and_swaplp ((long *) (P), (long) (O), (long) (N))
+static_assert ((sizeof (long) == sizeof (void *)), "");
+
+
+#elif defined(HB_NO_MT)
+
+#define hb_atomic_int_impl_add(AI, V)		((*(AI) += (V)) - (V))
+
+#define _hb_memory_barrier()			do {} while (0)
+
+#define hb_atomic_ptr_impl_cmpexch(P,O,N)	(* (void **) (P) == (void *) (O) ? (* (void **) (P) = (void *) (N), true) : false)
+
+
+#else
+
+#error "Could not find any system to define atomic_int macros."
+#error "Check hb-atomic.hh for possible resolutions."
+
+#endif
+
+
+#ifndef _hb_memory_r_barrier
+#define _hb_memory_r_barrier()			_hb_memory_barrier ()
+#endif
+#ifndef _hb_memory_w_barrier
+#define _hb_memory_w_barrier()			_hb_memory_barrier ()
+#endif
+#ifndef hb_atomic_int_impl_set_relaxed
+#define hb_atomic_int_impl_set_relaxed(AI, V)	(*(AI) = (V))
+#endif
+#ifndef hb_atomic_int_impl_get_relaxed
+#define hb_atomic_int_impl_get_relaxed(AI)	(*(AI))
+#endif
+
+#ifndef hb_atomic_ptr_impl_set_relaxed
+#define hb_atomic_ptr_impl_set_relaxed(P, V)	(*(P) = (V))
+#endif
+#ifndef hb_atomic_ptr_impl_get_relaxed
+#define hb_atomic_ptr_impl_get_relaxed(P)	(*(P))
+#endif
+#ifndef hb_atomic_int_impl_set
+inline void hb_atomic_int_impl_set (int *AI, int v)	{ _hb_memory_w_barrier (); *AI = v; }
+#endif
+#ifndef hb_atomic_int_impl_get
+inline int hb_atomic_int_impl_get (const int *AI)	{ int v = *AI; _hb_memory_r_barrier (); return v; }
+#endif
+#ifndef hb_atomic_ptr_impl_get
+inline void *hb_atomic_ptr_impl_get (void ** const P)	{ void *v = *P; _hb_memory_r_barrier (); return v; }
+#endif
+
+
+#define HB_ATOMIC_INT_INIT(V)          {V}
+struct hb_atomic_int_t
+{
+  void set_relaxed (int v_) { hb_atomic_int_impl_set_relaxed (&v, v_); }
+  void set (int v_) { hb_atomic_int_impl_set (&v, v_); }
+  int get_relaxed () const { return hb_atomic_int_impl_get_relaxed (&v); }
+  int get () const { return hb_atomic_int_impl_get (&v); }
+  int inc () { return hb_atomic_int_impl_add (&v,  1); }
+  int dec () { return hb_atomic_int_impl_add (&v, -1); }
+
+  int v;
+};
+
+
+#define HB_ATOMIC_PTR_INIT(V)          {V}
+template <typename P>
+struct hb_atomic_ptr_t
+{
+  typedef hb_remove_pointer<P> T;
+
+  void init (T* v_ = nullptr) { set_relaxed (v_); }
+  void set_relaxed (T* v_) { hb_atomic_ptr_impl_set_relaxed (&v, v_); }
+  T *get_relaxed () const { return (T *) hb_atomic_ptr_impl_get_relaxed (&v); }
+  T *get () const { return (T *) hb_atomic_ptr_impl_get ((void **) &v); }
+  bool cmpexch (const T *old, T *new_) const { return hb_atomic_ptr_impl_cmpexch ((void **) &v, (void *) old, (void *) new_); }
+
+  T * operator -> () const                    { return get (); }
+  template <typename C> operator C * () const { return get (); }
+
+  T *v;
+};
+
+
+#endif /* HB_ATOMIC_HH */
diff --git a/thirdparty/harfbuzz/src/hb-bimap.hh b/thirdparty/harfbuzz/src/hb-bimap.hh
new file mode 100644
index 0000000000..e9f3a6a52d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-bimap.hh
@@ -0,0 +1,166 @@
+/*
+ * Copyright © 2019 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_BIMAP_HH
+#define HB_BIMAP_HH
+
+#include "hb.hh"
+#include "hb-map.hh"
+
+/* Bi-directional map */
+struct hb_bimap_t
+{
+  hb_bimap_t () { init (); }
+  ~hb_bimap_t () { fini (); }
+
+  void init ()
+  {
+    forw_map.init ();
+    back_map.init ();
+  }
+
+  void fini ()
+  {
+    forw_map.fini ();
+    back_map.fini ();
+  }
+
+  void reset ()
+  {
+    forw_map.reset ();
+    back_map.reset ();
+  }
+
+  bool in_error () const { return forw_map.in_error () || back_map.in_error (); }
+
+  void set (hb_codepoint_t lhs, hb_codepoint_t rhs)
+  {
+    if (unlikely (lhs == HB_MAP_VALUE_INVALID)) return;
+    if (unlikely (rhs == HB_MAP_VALUE_INVALID)) { del (lhs); return; }
+    forw_map.set (lhs, rhs);
+    back_map.set (rhs, lhs);
+  }
+
+  hb_codepoint_t get (hb_codepoint_t lhs) const { return forw_map.get (lhs); }
+  hb_codepoint_t backward (hb_codepoint_t rhs) const { return back_map.get (rhs); }
+
+  hb_codepoint_t operator [] (hb_codepoint_t lhs) const { return get (lhs); }
+  bool has (hb_codepoint_t lhs, hb_codepoint_t *vp = nullptr) const { return forw_map.has (lhs, vp); }
+
+  void del (hb_codepoint_t lhs)
+  {
+    back_map.del (get (lhs));
+    forw_map.del (lhs);
+  }
+
+  void clear ()
+  {
+    forw_map.clear ();
+    back_map.clear ();
+  }
+
+  bool is_empty () const { return get_population () == 0; }
+
+  unsigned int get_population () const { return forw_map.get_population (); }
+
+  protected:
+  hb_map_t  forw_map;
+  hb_map_t  back_map;
+};
+
+/* Inremental bimap: only lhs is given, rhs is incrementally assigned */
+struct hb_inc_bimap_t : hb_bimap_t
+{
+  hb_inc_bimap_t () { init (); }
+
+  void init ()
+  {
+    hb_bimap_t::init ();
+    next_value = 0;
+  }
+
+  /* Add a mapping from lhs to rhs with a unique value if lhs is unknown.
+   * Return the rhs value as the result.
+   */
+  hb_codepoint_t add (hb_codepoint_t lhs)
+  {
+    hb_codepoint_t  rhs = forw_map[lhs];
+    if (rhs == HB_MAP_VALUE_INVALID)
+    {
+      rhs = next_value++;
+      set (lhs, rhs);
+    }
+    return rhs;
+  }
+
+  hb_codepoint_t skip ()
+  { return next_value++; }
+
+  hb_codepoint_t get_next_value () const
+  { return next_value; }
+
+  void add_set (const hb_set_t *set)
+  {
+    hb_codepoint_t i = HB_SET_VALUE_INVALID;
+    while (hb_set_next (set, &i)) add (i);
+  }
+
+  /* Create an identity map. */
+  bool identity (unsigned int size)
+  {
+    clear ();
+    for (hb_codepoint_t i = 0; i < size; i++) set (i, i);
+    return !in_error ();
+  }
+
+  protected:
+  static int cmp_id (const void* a, const void* b)
+  { return (int)*(const hb_codepoint_t *)a - (int)*(const hb_codepoint_t *)b; }
+
+  public:
+  /* Optional: after finished adding all mappings in a random order,
+   * reassign rhs to lhs so that they are in the same order. */
+  void sort ()
+  {
+    hb_codepoint_t  count = get_population ();
+    hb_vector_t <hb_codepoint_t> work;
+    work.resize (count);
+
+    for (hb_codepoint_t rhs = 0; rhs < count; rhs++)
+      work[rhs] = back_map[rhs];
+
+    work.qsort (cmp_id);
+
+    clear ();
+    for (hb_codepoint_t rhs = 0; rhs < count; rhs++)
+      set (work[rhs], rhs);
+  }
+
+  protected:
+  unsigned int	next_value;
+};
+
+#endif /* HB_BIMAP_HH */
diff --git a/thirdparty/harfbuzz/src/hb-blob.cc b/thirdparty/harfbuzz/src/hb-blob.cc
new file mode 100644
index 0000000000..94ed50fd3c
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-blob.cc
@@ -0,0 +1,717 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+#include "hb-blob.hh"
+
+#ifdef HAVE_SYS_MMAN_H
+#ifdef HAVE_UNISTD_H
+#include <unistd.h>
+#endif /* HAVE_UNISTD_H */
+#include <sys/mman.h>
+#endif /* HAVE_SYS_MMAN_H */
+
+#include <stdio.h>
+#include <stdlib.h>
+
+
+/**
+ * SECTION: hb-blob
+ * @title: hb-blob
+ * @short_description: Binary data containers
+ * @include: hb.h
+ *
+ * Blobs wrap a chunk of binary data to handle lifecycle management of data
+ * while it is passed between client and HarfBuzz.  Blobs are primarily used
+ * to create font faces, but also to access font face tables, as well as
+ * pass around other binary data.
+ **/
+
+
+/**
+ * hb_blob_create: (skip)
+ * @data: Pointer to blob data.
+ * @length: Length of @data in bytes.
+ * @mode: Memory mode for @data.
+ * @user_data: Data parameter to pass to @destroy.
+ * @destroy: Callback to call when @data is not needed anymore.
+ *
+ * Creates a new "blob" object wrapping @data.  The @mode parameter is used
+ * to negotiate ownership and lifecycle of @data.
+ *
+ * Return value: New blob, or the empty blob if something failed or if @length is
+ * zero.  Destroy with hb_blob_destroy().
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_blob_create (const char        *data,
+		unsigned int       length,
+		hb_memory_mode_t   mode,
+		void              *user_data,
+		hb_destroy_func_t  destroy)
+{
+  hb_blob_t *blob;
+
+  if (!length ||
+      length >= 1u << 31 ||
+      !(blob = hb_object_create<hb_blob_t> ())) {
+    if (destroy)
+      destroy (user_data);
+    return hb_blob_get_empty ();
+  }
+
+  blob->data = data;
+  blob->length = length;
+  blob->mode = mode;
+
+  blob->user_data = user_data;
+  blob->destroy = destroy;
+
+  if (blob->mode == HB_MEMORY_MODE_DUPLICATE) {
+    blob->mode = HB_MEMORY_MODE_READONLY;
+    if (!blob->try_make_writable ()) {
+      hb_blob_destroy (blob);
+      return hb_blob_get_empty ();
+    }
+  }
+
+  return blob;
+}
+
+static void
+_hb_blob_destroy (void *data)
+{
+  hb_blob_destroy ((hb_blob_t *) data);
+}
+
+/**
+ * hb_blob_create_sub_blob:
+ * @parent: Parent blob.
+ * @offset: Start offset of sub-blob within @parent, in bytes.
+ * @length: Length of sub-blob.
+ *
+ * Returns a blob that represents a range of bytes in @parent.  The new
+ * blob is always created with %HB_MEMORY_MODE_READONLY, meaning that it
+ * will never modify data in the parent blob.  The parent data is not
+ * expected to be modified, and will result in undefined behavior if it
+ * is.
+ *
+ * Makes @parent immutable.
+ *
+ * Return value: New blob, or the empty blob if something failed or if
+ * @length is zero or @offset is beyond the end of @parent's data.  Destroy
+ * with hb_blob_destroy().
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_blob_create_sub_blob (hb_blob_t    *parent,
+			 unsigned int  offset,
+			 unsigned int  length)
+{
+  hb_blob_t *blob;
+
+  if (!length || !parent || offset >= parent->length)
+    return hb_blob_get_empty ();
+
+  hb_blob_make_immutable (parent);
+
+  blob = hb_blob_create (parent->data + offset,
+			 hb_min (length, parent->length - offset),
+			 HB_MEMORY_MODE_READONLY,
+			 hb_blob_reference (parent),
+			 _hb_blob_destroy);
+
+  return blob;
+}
+
+/**
+ * hb_blob_copy_writable_or_fail:
+ * @blob: A blob.
+ *
+ * Makes a writable copy of @blob.
+ *
+ * Return value: New blob, or nullptr if allocation failed.
+ *
+ * Since: 1.8.0
+ **/
+hb_blob_t *
+hb_blob_copy_writable_or_fail (hb_blob_t *blob)
+{
+  blob = hb_blob_create (blob->data,
+			 blob->length,
+			 HB_MEMORY_MODE_DUPLICATE,
+			 nullptr,
+			 nullptr);
+
+  if (unlikely (blob == hb_blob_get_empty ()))
+    blob = nullptr;
+
+  return blob;
+}
+
+/**
+ * hb_blob_get_empty:
+ *
+ * Returns the singleton empty blob.
+ *
+ * See TODO:link object types for more information.
+ *
+ * Return value: (transfer full): the empty blob.
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_blob_get_empty ()
+{
+  return const_cast<hb_blob_t *> (&Null (hb_blob_t));
+}
+
+/**
+ * hb_blob_reference: (skip)
+ * @blob: a blob.
+ *
+ * Increases the reference count on @blob.
+ *
+ * See TODO:link object types for more information.
+ *
+ * Return value: @blob.
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_blob_reference (hb_blob_t *blob)
+{
+  return hb_object_reference (blob);
+}
+
+/**
+ * hb_blob_destroy: (skip)
+ * @blob: a blob.
+ *
+ * Decreases the reference count on @blob, and if it reaches zero, destroys
+ * @blob, freeing all memory, possibly calling the destroy-callback the blob
+ * was created for if it has not been called already.
+ *
+ * See TODO:link object types for more information.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_blob_destroy (hb_blob_t *blob)
+{
+  if (!hb_object_destroy (blob)) return;
+
+  blob->fini_shallow ();
+
+  free (blob);
+}
+
+/**
+ * hb_blob_set_user_data: (skip)
+ * @blob: a blob.
+ * @key: key for data to set.
+ * @data: data to set.
+ * @destroy: callback to call when @data is not needed anymore.
+ * @replace: whether to replace an existing data with the same key.
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_blob_set_user_data (hb_blob_t          *blob,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace)
+{
+  return hb_object_set_user_data (blob, key, data, destroy, replace);
+}
+
+/**
+ * hb_blob_get_user_data: (skip)
+ * @blob: a blob.
+ * @key: key for data to get.
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_blob_get_user_data (hb_blob_t          *blob,
+		       hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (blob, key);
+}
+
+
+/**
+ * hb_blob_make_immutable:
+ * @blob: a blob.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_blob_make_immutable (hb_blob_t *blob)
+{
+  if (hb_object_is_immutable (blob))
+    return;
+
+  hb_object_make_immutable (blob);
+}
+
+/**
+ * hb_blob_is_immutable:
+ * @blob: a blob.
+ *
+ *
+ *
+ * Return value: TODO
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_blob_is_immutable (hb_blob_t *blob)
+{
+  return hb_object_is_immutable (blob);
+}
+
+
+/**
+ * hb_blob_get_length:
+ * @blob: a blob.
+ *
+ *
+ *
+ * Return value: the length of blob data in bytes.
+ *
+ * Since: 0.9.2
+ **/
+unsigned int
+hb_blob_get_length (hb_blob_t *blob)
+{
+  return blob->length;
+}
+
+/**
+ * hb_blob_get_data:
+ * @blob: a blob.
+ * @length: (out):
+ *
+ *
+ *
+ * Returns: (transfer none) (array length=length):
+ *
+ * Since: 0.9.2
+ **/
+const char *
+hb_blob_get_data (hb_blob_t *blob, unsigned int *length)
+{
+  if (length)
+    *length = blob->length;
+
+  return blob->data;
+}
+
+/**
+ * hb_blob_get_data_writable:
+ * @blob: a blob.
+ * @length: (out): output length of the writable data.
+ *
+ * Tries to make blob data writable (possibly copying it) and
+ * return pointer to data.
+ *
+ * Fails if blob has been made immutable, or if memory allocation
+ * fails.
+ *
+ * Returns: (transfer none) (array length=length): Writable blob data,
+ * or %NULL if failed.
+ *
+ * Since: 0.9.2
+ **/
+char *
+hb_blob_get_data_writable (hb_blob_t *blob, unsigned int *length)
+{
+  if (!blob->try_make_writable ()) {
+    if (length)
+      *length = 0;
+
+    return nullptr;
+  }
+
+  if (length)
+    *length = blob->length;
+
+  return const_cast<char *> (blob->data);
+}
+
+
+bool
+hb_blob_t::try_make_writable_inplace_unix ()
+{
+#if defined(HAVE_SYS_MMAN_H) && defined(HAVE_MPROTECT)
+  uintptr_t pagesize = -1, mask, length;
+  const char *addr;
+
+#if defined(HAVE_SYSCONF) && defined(_SC_PAGE_SIZE)
+  pagesize = (uintptr_t) sysconf (_SC_PAGE_SIZE);
+#elif defined(HAVE_SYSCONF) && defined(_SC_PAGESIZE)
+  pagesize = (uintptr_t) sysconf (_SC_PAGESIZE);
+#elif defined(HAVE_GETPAGESIZE)
+  pagesize = (uintptr_t) getpagesize ();
+#endif
+
+  if ((uintptr_t) -1L == pagesize) {
+    DEBUG_MSG_FUNC (BLOB, this, "failed to get pagesize: %s", strerror (errno));
+    return false;
+  }
+  DEBUG_MSG_FUNC (BLOB, this, "pagesize is %lu", (unsigned long) pagesize);
+
+  mask = ~(pagesize-1);
+  addr = (const char *) (((uintptr_t) this->data) & mask);
+  length = (const char *) (((uintptr_t) this->data + this->length + pagesize-1) & mask)  - addr;
+  DEBUG_MSG_FUNC (BLOB, this,
+		  "calling mprotect on [%p..%p] (%lu bytes)",
+		  addr, addr+length, (unsigned long) length);
+  if (-1 == mprotect ((void *) addr, length, PROT_READ | PROT_WRITE)) {
+    DEBUG_MSG_FUNC (BLOB, this, "mprotect failed: %s", strerror (errno));
+    return false;
+  }
+
+  this->mode = HB_MEMORY_MODE_WRITABLE;
+
+  DEBUG_MSG_FUNC (BLOB, this,
+		  "successfully made [%p..%p] (%lu bytes) writable\n",
+		  addr, addr+length, (unsigned long) length);
+  return true;
+#else
+  return false;
+#endif
+}
+
+bool
+hb_blob_t::try_make_writable_inplace ()
+{
+  DEBUG_MSG_FUNC (BLOB, this, "making writable inplace\n");
+
+  if (this->try_make_writable_inplace_unix ())
+    return true;
+
+  DEBUG_MSG_FUNC (BLOB, this, "making writable -> FAILED\n");
+
+  /* Failed to make writable inplace, mark that */
+  this->mode = HB_MEMORY_MODE_READONLY;
+  return false;
+}
+
+bool
+hb_blob_t::try_make_writable ()
+{
+  if (hb_object_is_immutable (this))
+    return false;
+
+  if (this->mode == HB_MEMORY_MODE_WRITABLE)
+    return true;
+
+  if (this->mode == HB_MEMORY_MODE_READONLY_MAY_MAKE_WRITABLE && this->try_make_writable_inplace ())
+    return true;
+
+  if (this->mode == HB_MEMORY_MODE_WRITABLE)
+    return true;
+
+
+  DEBUG_MSG_FUNC (BLOB, this, "current data is -> %p\n", this->data);
+
+  char *new_data;
+
+  new_data = (char *) malloc (this->length);
+  if (unlikely (!new_data))
+    return false;
+
+  DEBUG_MSG_FUNC (BLOB, this, "dupped successfully -> %p\n", this->data);
+
+  memcpy (new_data, this->data, this->length);
+  this->destroy_user_data ();
+  this->mode = HB_MEMORY_MODE_WRITABLE;
+  this->data = new_data;
+  this->user_data = new_data;
+  this->destroy = free;
+
+  return true;
+}
+
+/*
+ * Mmap
+ */
+
+#ifndef HB_NO_OPEN
+#ifdef HAVE_MMAP
+# if !defined(HB_NO_RESOURCE_FORK) && defined(__APPLE__)
+#  include <sys/paths.h>
+# endif
+# include <sys/types.h>
+# include <sys/stat.h>
+# include <fcntl.h>
+#endif
+
+#ifdef _WIN32
+# include <windows.h>
+#else
+# ifndef O_BINARY
+#  define O_BINARY 0
+# endif
+#endif
+
+#ifndef MAP_NORESERVE
+# define MAP_NORESERVE 0
+#endif
+
+struct hb_mapped_file_t
+{
+  char *contents;
+  unsigned long length;
+#ifdef _WIN32
+  HANDLE mapping;
+#endif
+};
+
+#if (defined(HAVE_MMAP) || defined(_WIN32)) && !defined(HB_NO_MMAP)
+static void
+_hb_mapped_file_destroy (void *file_)
+{
+  hb_mapped_file_t *file = (hb_mapped_file_t *) file_;
+#ifdef HAVE_MMAP
+  munmap (file->contents, file->length);
+#elif defined(_WIN32)
+  UnmapViewOfFile (file->contents);
+  CloseHandle (file->mapping);
+#else
+  assert (0); // If we don't have mmap we shouldn't reach here
+#endif
+
+  free (file);
+}
+#endif
+
+#ifdef _PATH_RSRCFORKSPEC
+static int
+_open_resource_fork (const char *file_name, hb_mapped_file_t *file)
+{
+  size_t name_len = strlen (file_name);
+  size_t len = name_len + sizeof (_PATH_RSRCFORKSPEC);
+
+  char *rsrc_name = (char *) malloc (len);
+  if (unlikely (!rsrc_name)) return -1;
+
+  strncpy (rsrc_name, file_name, name_len);
+  strncpy (rsrc_name + name_len, _PATH_RSRCFORKSPEC,
+	   sizeof (_PATH_RSRCFORKSPEC) - 1);
+
+  int fd = open (rsrc_name, O_RDONLY | O_BINARY, 0);
+  free (rsrc_name);
+
+  if (fd != -1)
+  {
+    struct stat st;
+    if (fstat (fd, &st) != -1)
+      file->length = (unsigned long) st.st_size;
+    else
+    {
+      close (fd);
+      fd = -1;
+    }
+  }
+
+  return fd;
+}
+#endif
+
+/**
+ * hb_blob_create_from_file:
+ * @file_name: font filename.
+ *
+ * Returns: A hb_blob_t pointer with the content of the file
+ *
+ * Since: 1.7.7
+ **/
+hb_blob_t *
+hb_blob_create_from_file (const char *file_name)
+{
+  /* Adopted from glib's gmappedfile.c with Matthias Clasen and
+     Allison Lortie permission but changed a lot to suit our need. */
+#if defined(HAVE_MMAP) && !defined(HB_NO_MMAP)
+  hb_mapped_file_t *file = (hb_mapped_file_t *) calloc (1, sizeof (hb_mapped_file_t));
+  if (unlikely (!file)) return hb_blob_get_empty ();
+
+  int fd = open (file_name, O_RDONLY | O_BINARY, 0);
+  if (unlikely (fd == -1)) goto fail_without_close;
+
+  struct stat st;
+  if (unlikely (fstat (fd, &st) == -1)) goto fail;
+
+  file->length = (unsigned long) st.st_size;
+
+#ifdef _PATH_RSRCFORKSPEC
+  if (unlikely (file->length == 0))
+  {
+    int rfd = _open_resource_fork (file_name, file);
+    if (rfd != -1)
+    {
+      close (fd);
+      fd = rfd;
+    }
+  }
+#endif
+
+  file->contents = (char *) mmap (nullptr, file->length, PROT_READ,
+				  MAP_PRIVATE | MAP_NORESERVE, fd, 0);
+
+  if (unlikely (file->contents == MAP_FAILED)) goto fail;
+
+  close (fd);
+
+  return hb_blob_create (file->contents, file->length,
+			 HB_MEMORY_MODE_READONLY_MAY_MAKE_WRITABLE, (void *) file,
+			 (hb_destroy_func_t) _hb_mapped_file_destroy);
+
+fail:
+  close (fd);
+fail_without_close:
+  free (file);
+
+#elif defined(_WIN32) && !defined(HB_NO_MMAP)
+  hb_mapped_file_t *file = (hb_mapped_file_t *) calloc (1, sizeof (hb_mapped_file_t));
+  if (unlikely (!file)) return hb_blob_get_empty ();
+
+  HANDLE fd;
+  unsigned int size = strlen (file_name) + 1;
+  wchar_t * wchar_file_name = (wchar_t *) malloc (sizeof (wchar_t) * size);
+  if (unlikely (!wchar_file_name)) goto fail_without_close;
+  mbstowcs (wchar_file_name, file_name, size);
+#if !WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+  {
+    CREATEFILE2_EXTENDED_PARAMETERS ceparams = { 0 };
+    ceparams.dwSize = sizeof(CREATEFILE2_EXTENDED_PARAMETERS);
+    ceparams.dwFileAttributes = FILE_ATTRIBUTE_NORMAL | FILE_FLAG_OVERLAPPED & 0xFFFF;
+    ceparams.dwFileFlags = FILE_ATTRIBUTE_NORMAL | FILE_FLAG_OVERLAPPED & 0xFFF00000;
+    ceparams.dwSecurityQosFlags = FILE_ATTRIBUTE_NORMAL | FILE_FLAG_OVERLAPPED & 0x000F0000;
+    ceparams.lpSecurityAttributes = nullptr;
+    ceparams.hTemplateFile = nullptr;
+    fd = CreateFile2 (wchar_file_name, GENERIC_READ, FILE_SHARE_READ,
+		      OPEN_EXISTING, &ceparams);
+  }
+#else
+  fd = CreateFileW (wchar_file_name, GENERIC_READ, FILE_SHARE_READ, nullptr,
+		    OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL|FILE_FLAG_OVERLAPPED,
+		    nullptr);
+#endif
+  free (wchar_file_name);
+
+  if (unlikely (fd == INVALID_HANDLE_VALUE)) goto fail_without_close;
+
+#if !WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+  {
+    LARGE_INTEGER length;
+    GetFileSizeEx (fd, &length);
+    file->length = length.LowPart;
+    file->mapping = CreateFileMappingFromApp (fd, nullptr, PAGE_READONLY, length.QuadPart, nullptr);
+  }
+#else
+  file->length = (unsigned long) GetFileSize (fd, nullptr);
+  file->mapping = CreateFileMapping (fd, nullptr, PAGE_READONLY, 0, 0, nullptr);
+#endif
+  if (unlikely (!file->mapping)) goto fail;
+
+#if !WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+  file->contents = (char *) MapViewOfFileFromApp (file->mapping, FILE_MAP_READ, 0, 0);
+#else
+  file->contents = (char *) MapViewOfFile (file->mapping, FILE_MAP_READ, 0, 0, 0);
+#endif
+  if (unlikely (!file->contents)) goto fail;
+
+  CloseHandle (fd);
+  return hb_blob_create (file->contents, file->length,
+			 HB_MEMORY_MODE_READONLY_MAY_MAKE_WRITABLE, (void *) file,
+			 (hb_destroy_func_t) _hb_mapped_file_destroy);
+
+fail:
+  CloseHandle (fd);
+fail_without_close:
+  free (file);
+
+#endif
+
+  /* The following tries to read a file without knowing its size beforehand
+     It's used as a fallback for systems without mmap or to read from pipes */
+  unsigned long len = 0, allocated = BUFSIZ * 16;
+  char *data = (char *) malloc (allocated);
+  if (unlikely (!data)) return hb_blob_get_empty ();
+
+  FILE *fp = fopen (file_name, "rb");
+  if (unlikely (!fp)) goto fread_fail_without_close;
+
+  while (!feof (fp))
+  {
+    if (allocated - len < BUFSIZ)
+    {
+      allocated *= 2;
+      /* Don't allocate and go more than ~536MB, our mmap reader still
+	 can cover files like that but lets limit our fallback reader */
+      if (unlikely (allocated > (2 << 28))) goto fread_fail;
+      char *new_data = (char *) realloc (data, allocated);
+      if (unlikely (!new_data)) goto fread_fail;
+      data = new_data;
+    }
+
+    unsigned long addition = fread (data + len, 1, allocated - len, fp);
+
+    int err = ferror (fp);
+#ifdef EINTR // armcc doesn't have it
+    if (unlikely (err == EINTR)) continue;
+#endif
+    if (unlikely (err)) goto fread_fail;
+
+    len += addition;
+  }
+	fclose (fp);
+
+  return hb_blob_create (data, len, HB_MEMORY_MODE_WRITABLE, data,
+			 (hb_destroy_func_t) free);
+
+fread_fail:
+  fclose (fp);
+fread_fail_without_close:
+  free (data);
+  return hb_blob_get_empty ();
+}
+#endif /* !HB_NO_OPEN */
diff --git a/thirdparty/harfbuzz/src/hb-blob.h b/thirdparty/harfbuzz/src/hb-blob.h
new file mode 100644
index 0000000000..f80e9af2d9
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-blob.h
@@ -0,0 +1,131 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_BLOB_H
+#define HB_BLOB_H
+
+#include "hb-common.h"
+
+HB_BEGIN_DECLS
+
+
+/*
+ * Note re various memory-modes:
+ *
+ * - In no case shall the HarfBuzz client modify memory
+ *   that is passed to HarfBuzz in a blob.  If there is
+ *   any such possibility, MODE_DUPLICATE should be used
+ *   such that HarfBuzz makes a copy immediately,
+ *
+ * - Use MODE_READONLY otherwise, unless you really really
+ *   really know what you are doing,
+ *
+ * - MODE_WRITABLE is appropriate if you really made a
+ *   copy of data solely for the purpose of passing to
+ *   HarfBuzz and doing that just once (no reuse!),
+ *
+ * - If the font is mmap()ed, it's ok to use
+ *   READONLY_MAY_MAKE_WRITABLE, however, using that mode
+ *   correctly is very tricky.  Use MODE_READONLY instead.
+ */
+typedef enum {
+  HB_MEMORY_MODE_DUPLICATE,
+  HB_MEMORY_MODE_READONLY,
+  HB_MEMORY_MODE_WRITABLE,
+  HB_MEMORY_MODE_READONLY_MAY_MAKE_WRITABLE
+} hb_memory_mode_t;
+
+typedef struct hb_blob_t hb_blob_t;
+
+HB_EXTERN hb_blob_t *
+hb_blob_create (const char        *data,
+		unsigned int       length,
+		hb_memory_mode_t   mode,
+		void              *user_data,
+		hb_destroy_func_t  destroy);
+
+HB_EXTERN hb_blob_t *
+hb_blob_create_from_file (const char *file_name);
+
+/* Always creates with MEMORY_MODE_READONLY.
+ * Even if the parent blob is writable, we don't
+ * want the user of the sub-blob to be able to
+ * modify the parent data as that data may be
+ * shared among multiple sub-blobs.
+ */
+HB_EXTERN hb_blob_t *
+hb_blob_create_sub_blob (hb_blob_t    *parent,
+			 unsigned int  offset,
+			 unsigned int  length);
+
+HB_EXTERN hb_blob_t *
+hb_blob_copy_writable_or_fail (hb_blob_t *blob);
+
+HB_EXTERN hb_blob_t *
+hb_blob_get_empty (void);
+
+HB_EXTERN hb_blob_t *
+hb_blob_reference (hb_blob_t *blob);
+
+HB_EXTERN void
+hb_blob_destroy (hb_blob_t *blob);
+
+HB_EXTERN hb_bool_t
+hb_blob_set_user_data (hb_blob_t          *blob,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace);
+
+
+HB_EXTERN void *
+hb_blob_get_user_data (hb_blob_t          *blob,
+		       hb_user_data_key_t *key);
+
+
+HB_EXTERN void
+hb_blob_make_immutable (hb_blob_t *blob);
+
+HB_EXTERN hb_bool_t
+hb_blob_is_immutable (hb_blob_t *blob);
+
+
+HB_EXTERN unsigned int
+hb_blob_get_length (hb_blob_t *blob);
+
+HB_EXTERN const char *
+hb_blob_get_data (hb_blob_t *blob, unsigned int *length);
+
+HB_EXTERN char *
+hb_blob_get_data_writable (hb_blob_t *blob, unsigned int *length);
+
+HB_END_DECLS
+
+#endif /* HB_BLOB_H */
diff --git a/thirdparty/harfbuzz/src/hb-blob.hh b/thirdparty/harfbuzz/src/hb-blob.hh
new file mode 100644
index 0000000000..d85bd823b0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-blob.hh
@@ -0,0 +1,97 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_BLOB_HH
+#define HB_BLOB_HH
+
+#include "hb.hh"
+
+
+/*
+ * hb_blob_t
+ */
+
+struct hb_blob_t
+{
+  void fini_shallow () { destroy_user_data (); }
+
+  void destroy_user_data ()
+  {
+    if (destroy)
+    {
+      destroy (user_data);
+      user_data = nullptr;
+      destroy = nullptr;
+    }
+  }
+
+  HB_INTERNAL bool try_make_writable ();
+  HB_INTERNAL bool try_make_writable_inplace ();
+  HB_INTERNAL bool try_make_writable_inplace_unix ();
+
+  hb_bytes_t as_bytes () const { return hb_bytes_t (data, length); }
+  template <typename Type>
+  const Type* as () const { return as_bytes ().as<Type> (); }
+
+  public:
+  hb_object_header_t header;
+
+  const char *data;
+  unsigned int length;
+  hb_memory_mode_t mode;
+
+  void *user_data;
+  hb_destroy_func_t destroy;
+};
+
+
+/*
+ * hb_blob_ptr_t
+ */
+
+template <typename P>
+struct hb_blob_ptr_t
+{
+  typedef hb_remove_pointer<P> T;
+
+  hb_blob_ptr_t (hb_blob_t *b_ = nullptr) : b (b_) {}
+  hb_blob_t * operator = (hb_blob_t *b_) { return b = b_; }
+  const T * operator -> () const { return get (); }
+  const T & operator * () const  { return *get (); }
+  template <typename C> operator const C * () const { return get (); }
+  operator const char * () const { return (const char *) get (); }
+  const T * get () const { return b->as<T> (); }
+  hb_blob_t * get_blob () const { return b.get_raw (); }
+  unsigned int get_length () const { return b.get ()->length; }
+  void destroy () { hb_blob_destroy (b.get ()); b = nullptr; }
+
+  hb_nonnull_ptr_t<hb_blob_t> b;
+};
+
+
+#endif /* HB_BLOB_HH */
diff --git a/thirdparty/harfbuzz/src/hb-buffer-deserialize-json.hh b/thirdparty/harfbuzz/src/hb-buffer-deserialize-json.hh
new file mode 100644
index 0000000000..1f9e2e91db
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer-deserialize-json.hh
@@ -0,0 +1,643 @@
+
+#line 1 "hb-buffer-deserialize-json.rl"
+/*
+ * Copyright © 2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_BUFFER_DESERIALIZE_JSON_HH
+#define HB_BUFFER_DESERIALIZE_JSON_HH
+
+#include "hb.hh"
+
+
+#line 36 "hb-buffer-deserialize-json.hh"
+static const unsigned char _deserialize_json_trans_keys[] = {
+	0u, 0u, 9u, 123u, 9u, 34u, 97u, 103u, 120u, 121u, 34u, 34u, 9u, 58u, 9u, 57u, 
+	48u, 57u, 9u, 125u, 9u, 125u, 9u, 125u, 34u, 34u, 9u, 58u, 9u, 57u, 48u, 57u, 
+	9u, 125u, 9u, 125u, 108u, 108u, 34u, 34u, 9u, 58u, 9u, 57u, 9u, 125u, 9u, 125u, 
+	120u, 121u, 34u, 34u, 9u, 58u, 9u, 57u, 48u, 57u, 9u, 125u, 9u, 125u, 34u, 34u, 
+	9u, 58u, 9u, 57u, 48u, 57u, 9u, 125u, 9u, 125u, 34u, 34u, 9u, 58u, 9u, 57u, 
+	65u, 122u, 34u, 122u, 9u, 125u, 9u, 125u, 9u, 93u, 9u, 123u, 0u, 0u, 0
+};
+
+static const char _deserialize_json_key_spans[] = {
+	0, 115, 26, 7, 2, 1, 50, 49, 
+	10, 117, 117, 117, 1, 50, 49, 10, 
+	117, 117, 1, 1, 50, 49, 117, 117, 
+	2, 1, 50, 49, 10, 117, 117, 1, 
+	50, 49, 10, 117, 117, 1, 50, 49, 
+	58, 89, 117, 117, 85, 115, 0
+};
+
+static const short _deserialize_json_index_offsets[] = {
+	0, 0, 116, 143, 151, 154, 156, 207, 
+	257, 268, 386, 504, 622, 624, 675, 725, 
+	736, 854, 972, 974, 976, 1027, 1077, 1195, 
+	1313, 1316, 1318, 1369, 1419, 1430, 1548, 1666, 
+	1668, 1719, 1769, 1780, 1898, 2016, 2018, 2069, 
+	2119, 2178, 2268, 2386, 2504, 2590, 2706
+};
+
+static const char _deserialize_json_indicies[] = {
+	0, 0, 0, 0, 0, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	0, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 2, 1, 3, 3, 3, 
+	3, 3, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 3, 1, 4, 1, 
+	5, 1, 6, 7, 1, 1, 8, 1, 
+	9, 10, 1, 11, 1, 11, 11, 11, 
+	11, 11, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 11, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 12, 1, 
+	12, 12, 12, 12, 12, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 12, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 13, 1, 1, 14, 
+	15, 15, 15, 15, 15, 15, 15, 15, 
+	15, 1, 16, 17, 17, 17, 17, 17, 
+	17, 17, 17, 17, 1, 18, 18, 18, 
+	18, 18, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 18, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	19, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 20, 1, 21, 21, 21, 21, 21, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 21, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 3, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 22, 
+	1, 18, 18, 18, 18, 18, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	18, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 19, 1, 1, 1, 
+	17, 17, 17, 17, 17, 17, 17, 17, 
+	17, 17, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 20, 1, 23, 
+	1, 23, 23, 23, 23, 23, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	23, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 24, 1, 24, 24, 24, 24, 
+	24, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 24, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	25, 1, 1, 26, 27, 27, 27, 27, 
+	27, 27, 27, 27, 27, 1, 28, 29, 
+	29, 29, 29, 29, 29, 29, 29, 29, 
+	1, 30, 30, 30, 30, 30, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	30, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 31, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 32, 1, 30, 
+	30, 30, 30, 30, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 30, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 31, 1, 1, 1, 29, 29, 
+	29, 29, 29, 29, 29, 29, 29, 29, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 32, 1, 33, 1, 34, 
+	1, 34, 34, 34, 34, 34, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	34, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 35, 1, 35, 35, 35, 35, 
+	35, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 35, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 36, 37, 37, 37, 37, 
+	37, 37, 37, 37, 37, 1, 38, 38, 
+	38, 38, 38, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 38, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 39, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 40, 1, 38, 38, 38, 38, 
+	38, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 38, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 39, 
+	1, 1, 1, 41, 41, 41, 41, 41, 
+	41, 41, 41, 41, 41, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	40, 1, 42, 43, 1, 44, 1, 44, 
+	44, 44, 44, 44, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 44, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	45, 1, 45, 45, 45, 45, 45, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 45, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 46, 1, 
+	1, 47, 48, 48, 48, 48, 48, 48, 
+	48, 48, 48, 1, 49, 50, 50, 50, 
+	50, 50, 50, 50, 50, 50, 1, 51, 
+	51, 51, 51, 51, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 51, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 52, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 53, 1, 51, 51, 51, 
+	51, 51, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 51, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	52, 1, 1, 1, 50, 50, 50, 50, 
+	50, 50, 50, 50, 50, 50, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 53, 1, 54, 1, 54, 54, 54, 
+	54, 54, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 54, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 55, 1, 
+	55, 55, 55, 55, 55, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 55, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 56, 1, 1, 57, 
+	58, 58, 58, 58, 58, 58, 58, 58, 
+	58, 1, 59, 60, 60, 60, 60, 60, 
+	60, 60, 60, 60, 1, 61, 61, 61, 
+	61, 61, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 61, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	62, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 63, 1, 61, 61, 61, 61, 61, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 61, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 62, 1, 
+	1, 1, 60, 60, 60, 60, 60, 60, 
+	60, 60, 60, 60, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 63, 
+	1, 64, 1, 64, 64, 64, 64, 64, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 64, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 65, 1, 65, 65, 
+	65, 65, 65, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 65, 1, 66, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 67, 68, 68, 
+	68, 68, 68, 68, 68, 68, 68, 1, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 1, 1, 1, 1, 1, 1, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 69, 69, 69, 69, 69, 69, 
+	69, 69, 1, 70, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 71, 71, 
+	1, 71, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 1, 1, 1, 1, 1, 
+	1, 1, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 1, 1, 1, 1, 
+	71, 1, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 71, 71, 71, 71, 
+	71, 71, 71, 71, 1, 72, 72, 72, 
+	72, 72, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 72, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	73, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 74, 1, 72, 72, 72, 72, 72, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 72, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 73, 1, 
+	1, 1, 75, 75, 75, 75, 75, 75, 
+	75, 75, 75, 75, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 74, 
+	1, 76, 76, 76, 76, 76, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	76, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 77, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 78, 1, 0, 
+	0, 0, 0, 0, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 0, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 2, 1, 1, 0
+};
+
+static const char _deserialize_json_trans_targs[] = {
+	1, 0, 2, 2, 3, 4, 18, 24, 
+	37, 5, 12, 6, 7, 8, 9, 11, 
+	9, 11, 10, 2, 44, 10, 44, 13, 
+	14, 15, 16, 17, 16, 17, 10, 2, 
+	44, 19, 20, 21, 22, 23, 10, 2, 
+	44, 23, 25, 31, 26, 27, 28, 29, 
+	30, 29, 30, 10, 2, 44, 32, 33, 
+	34, 35, 36, 35, 36, 10, 2, 44, 
+	38, 39, 40, 42, 43, 41, 10, 41, 
+	10, 2, 44, 43, 44, 45, 46
+};
+
+static const char _deserialize_json_trans_actions[] = {
+	0, 0, 1, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 2, 2, 2, 
+	0, 0, 3, 3, 4, 0, 5, 0, 
+	0, 2, 2, 2, 0, 0, 6, 6, 
+	7, 0, 0, 0, 2, 2, 8, 8, 
+	9, 0, 0, 0, 0, 0, 2, 2, 
+	2, 0, 0, 10, 10, 11, 0, 0, 
+	2, 2, 2, 0, 0, 12, 12, 13, 
+	0, 0, 0, 2, 2, 2, 14, 0, 
+	15, 15, 16, 0, 0, 0, 0
+};
+
+static const int deserialize_json_start = 1;
+static const int deserialize_json_first_final = 44;
+static const int deserialize_json_error = 0;
+
+static const int deserialize_json_en_main = 1;
+
+
+#line 97 "hb-buffer-deserialize-json.rl"
+
+
+static hb_bool_t
+_hb_buffer_deserialize_glyphs_json (hb_buffer_t *buffer,
+				    const char *buf,
+				    unsigned int buf_len,
+				    const char **end_ptr,
+				    hb_font_t *font)
+{
+  const char *p = buf, *pe = buf + buf_len;
+
+  /* Ensure we have positions. */
+  (void) hb_buffer_get_glyph_positions (buffer, nullptr);
+
+  while (p < pe && ISSPACE (*p))
+    p++;
+  if (p < pe && *p == (buffer->len ? ',' : '['))
+  {
+    *end_ptr = ++p;
+  }
+
+  const char *tok = nullptr;
+  int cs;
+  hb_glyph_info_t info = {0};
+  hb_glyph_position_t pos = {0};
+  
+#line 466 "hb-buffer-deserialize-json.hh"
+	{
+	cs = deserialize_json_start;
+	}
+
+#line 471 "hb-buffer-deserialize-json.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+	if ( cs == 0 )
+		goto _out;
+_resume:
+	_keys = _deserialize_json_trans_keys + (cs<<1);
+	_inds = _deserialize_json_indicies + _deserialize_json_index_offsets[cs];
+
+	_slen = _deserialize_json_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=(*p) &&
+		(*p) <= _keys[1] ?
+		(*p) - _keys[0] : _slen ];
+
+	cs = _deserialize_json_trans_targs[_trans];
+
+	if ( _deserialize_json_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _deserialize_json_trans_actions[_trans] ) {
+	case 1:
+#line 38 "hb-buffer-deserialize-json.rl"
+	{
+	memset (&info, 0, sizeof (info));
+	memset (&pos , 0, sizeof (pos ));
+}
+	break;
+	case 5:
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 2:
+#line 51 "hb-buffer-deserialize-json.rl"
+	{
+	tok = p;
+}
+	break;
+	case 14:
+#line 55 "hb-buffer-deserialize-json.rl"
+	{
+	if (!hb_font_glyph_from_string (font,
+					tok, p - tok,
+					&info.codepoint))
+	  return false;
+}
+	break;
+	case 15:
+#line 62 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_uint (tok, p, &info.codepoint)) return false; }
+	break;
+	case 8:
+#line 63 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_uint (tok, p, &info.cluster )) return false; }
+	break;
+	case 10:
+#line 64 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.x_offset )) return false; }
+	break;
+	case 12:
+#line 65 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.y_offset )) return false; }
+	break;
+	case 3:
+#line 66 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.x_advance)) return false; }
+	break;
+	case 6:
+#line 67 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.y_advance)) return false; }
+	break;
+	case 16:
+#line 62 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_uint (tok, p, &info.codepoint)) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 9:
+#line 63 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_uint (tok, p, &info.cluster )) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 11:
+#line 64 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.x_offset )) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 13:
+#line 65 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.y_offset )) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 4:
+#line 66 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.x_advance)) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 7:
+#line 67 "hb-buffer-deserialize-json.rl"
+	{ if (!parse_int  (tok, p, &pos.y_advance)) return false; }
+#line 43 "hb-buffer-deserialize-json.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+#line 624 "hb-buffer-deserialize-json.hh"
+	}
+
+_again:
+	if ( cs == 0 )
+		goto _out;
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	_out: {}
+	}
+
+#line 125 "hb-buffer-deserialize-json.rl"
+
+
+  *end_ptr = p;
+
+  return p == pe && *(p-1) != ']';
+}
+
+#endif /* HB_BUFFER_DESERIALIZE_JSON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-buffer-deserialize-text.hh b/thirdparty/harfbuzz/src/hb-buffer-deserialize-text.hh
new file mode 100644
index 0000000000..67f0a1252f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer-deserialize-text.hh
@@ -0,0 +1,571 @@
+
+#line 1 "hb-buffer-deserialize-text.rl"
+/*
+ * Copyright © 2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_BUFFER_DESERIALIZE_TEXT_HH
+#define HB_BUFFER_DESERIALIZE_TEXT_HH
+
+#include "hb.hh"
+
+
+#line 36 "hb-buffer-deserialize-text.hh"
+static const unsigned char _deserialize_text_trans_keys[] = {
+	0u, 0u, 9u, 122u, 45u, 57u, 48u, 57u, 45u, 57u, 48u, 57u, 48u, 57u, 45u, 57u, 
+	48u, 57u, 44u, 44u, 45u, 57u, 48u, 57u, 44u, 57u, 9u, 124u, 9u, 124u, 0u, 0u, 
+	9u, 122u, 9u, 124u, 9u, 124u, 9u, 124u, 9u, 124u, 9u, 124u, 9u, 124u, 9u, 124u, 
+	9u, 124u, 9u, 124u, 9u, 124u, 0
+};
+
+static const char _deserialize_text_key_spans[] = {
+	0, 114, 13, 10, 13, 10, 10, 13, 
+	10, 1, 13, 10, 14, 116, 116, 0, 
+	114, 116, 116, 116, 116, 116, 116, 116, 
+	116, 116, 116
+};
+
+static const short _deserialize_text_index_offsets[] = {
+	0, 0, 115, 129, 140, 154, 165, 176, 
+	190, 201, 203, 217, 228, 243, 360, 477, 
+	478, 593, 710, 827, 944, 1061, 1178, 1295, 
+	1412, 1529, 1646
+};
+
+static const char _deserialize_text_indicies[] = {
+	0, 0, 0, 0, 0, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	0, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	2, 3, 3, 3, 3, 3, 3, 3, 
+	3, 3, 1, 1, 1, 1, 1, 1, 
+	1, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 1, 1, 1, 1, 1, 
+	1, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 1, 5, 1, 1, 6, 
+	7, 7, 7, 7, 7, 7, 7, 7, 
+	7, 1, 8, 9, 9, 9, 9, 9, 
+	9, 9, 9, 9, 1, 10, 1, 1, 
+	11, 12, 12, 12, 12, 12, 12, 12, 
+	12, 12, 1, 13, 14, 14, 14, 14, 
+	14, 14, 14, 14, 14, 1, 15, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	1, 17, 1, 1, 18, 19, 19, 19, 
+	19, 19, 19, 19, 19, 19, 1, 20, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 1, 22, 1, 23, 1, 1, 24, 
+	25, 25, 25, 25, 25, 25, 25, 25, 
+	25, 1, 26, 27, 27, 27, 27, 27, 
+	27, 27, 27, 27, 1, 22, 1, 1, 
+	1, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 1, 28, 28, 28, 28, 
+	28, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 28, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 29, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	30, 1, 1, 31, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	32, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 33, 
+	1, 34, 34, 34, 34, 34, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	34, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 35, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 36, 1, 1, 0, 
+	0, 0, 0, 0, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 0, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 2, 3, 
+	3, 3, 3, 3, 3, 3, 3, 3, 
+	1, 1, 1, 1, 1, 1, 1, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 1, 1, 1, 1, 1, 1, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 1, 28, 28, 28, 28, 28, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 28, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 29, 1, 1, 1, 
+	1, 37, 37, 37, 37, 37, 37, 37, 
+	37, 37, 37, 1, 1, 1, 30, 1, 
+	1, 31, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 32, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 33, 1, 38, 
+	38, 38, 38, 38, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 38, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 39, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 40, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 41, 1, 42, 42, 42, 42, 
+	42, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 42, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	43, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 44, 
+	1, 42, 42, 42, 42, 42, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	42, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	14, 14, 14, 14, 14, 14, 14, 14, 
+	14, 14, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 43, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 44, 1, 38, 38, 
+	38, 38, 38, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 38, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 39, 1, 1, 1, 9, 9, 9, 
+	9, 9, 9, 9, 9, 9, 9, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 40, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 41, 1, 45, 45, 45, 45, 45, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 45, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 46, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 47, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 48, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 49, 1, 
+	50, 50, 50, 50, 50, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 50, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 51, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 52, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 53, 1, 50, 50, 50, 
+	50, 50, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 50, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 51, 
+	1, 1, 1, 1, 27, 27, 27, 27, 
+	27, 27, 27, 27, 27, 27, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 52, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	53, 1, 45, 45, 45, 45, 45, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 45, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 46, 1, 1, 1, 
+	1, 54, 54, 54, 54, 54, 54, 54, 
+	54, 54, 54, 1, 1, 1, 1, 1, 
+	1, 47, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 48, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 49, 1, 28, 
+	28, 28, 28, 28, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 28, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 29, 1, 55, 55, 1, 55, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	1, 1, 1, 30, 1, 1, 31, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 1, 1, 32, 1, 55, 1, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 55, 55, 55, 55, 55, 55, 55, 
+	55, 1, 33, 1, 0
+};
+
+static const char _deserialize_text_trans_targs[] = {
+	1, 0, 13, 17, 26, 3, 18, 21, 
+	18, 21, 5, 19, 20, 19, 20, 22, 
+	25, 8, 9, 12, 9, 12, 10, 11, 
+	23, 24, 23, 24, 14, 2, 6, 7, 
+	15, 16, 14, 15, 16, 17, 14, 4, 
+	15, 16, 14, 15, 16, 14, 2, 7, 
+	15, 16, 14, 2, 15, 16, 25, 26
+};
+
+static const char _deserialize_text_trans_actions[] = {
+	0, 0, 1, 1, 1, 2, 2, 2, 
+	0, 0, 2, 2, 2, 0, 0, 2, 
+	2, 2, 2, 2, 0, 0, 3, 2, 
+	2, 2, 0, 0, 4, 5, 5, 5, 
+	4, 4, 0, 0, 0, 0, 6, 7, 
+	6, 6, 8, 8, 8, 9, 10, 10, 
+	9, 9, 11, 12, 11, 11, 0, 0
+};
+
+static const char _deserialize_text_eof_actions[] = {
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 4, 0, 0, 
+	0, 4, 6, 8, 8, 6, 9, 11, 
+	11, 9, 4
+};
+
+static const int deserialize_text_start = 1;
+static const int deserialize_text_first_final = 13;
+static const int deserialize_text_error = 0;
+
+static const int deserialize_text_en_main = 1;
+
+
+#line 91 "hb-buffer-deserialize-text.rl"
+
+
+static hb_bool_t
+_hb_buffer_deserialize_glyphs_text (hb_buffer_t *buffer,
+				    const char *buf,
+				    unsigned int buf_len,
+				    const char **end_ptr,
+				    hb_font_t *font)
+{
+  const char *p = buf, *pe = buf + buf_len;
+
+  /* Ensure we have positions. */
+  (void) hb_buffer_get_glyph_positions (buffer, nullptr);
+
+  while (p < pe && ISSPACE (*p))
+    p++;
+  if (p < pe && *p == (buffer->len ? '|' : '['))
+  {
+    *end_ptr = ++p;
+  }
+
+  const char *eof = pe, *tok = nullptr;
+  int cs;
+  hb_glyph_info_t info = {0};
+  hb_glyph_position_t pos = {0};
+  
+#line 343 "hb-buffer-deserialize-text.hh"
+	{
+	cs = deserialize_text_start;
+	}
+
+#line 348 "hb-buffer-deserialize-text.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+	if ( cs == 0 )
+		goto _out;
+_resume:
+	_keys = _deserialize_text_trans_keys + (cs<<1);
+	_inds = _deserialize_text_indicies + _deserialize_text_index_offsets[cs];
+
+	_slen = _deserialize_text_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=(*p) &&
+		(*p) <= _keys[1] ?
+		(*p) - _keys[0] : _slen ];
+
+	cs = _deserialize_text_trans_targs[_trans];
+
+	if ( _deserialize_text_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _deserialize_text_trans_actions[_trans] ) {
+	case 2:
+#line 51 "hb-buffer-deserialize-text.rl"
+	{
+	tok = p;
+}
+	break;
+	case 5:
+#line 55 "hb-buffer-deserialize-text.rl"
+	{
+	if (!hb_font_glyph_from_string (font,
+					tok, p - tok,
+					&info.codepoint))
+	  return false;
+}
+	break;
+	case 10:
+#line 62 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_uint (tok, p, &info.cluster )) return false; }
+	break;
+	case 3:
+#line 63 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.x_offset )) return false; }
+	break;
+	case 12:
+#line 64 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.y_offset )) return false; }
+	break;
+	case 7:
+#line 65 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.x_advance)) return false; }
+	break;
+	case 1:
+#line 38 "hb-buffer-deserialize-text.rl"
+	{
+	memset (&info, 0, sizeof (info));
+	memset (&pos , 0, sizeof (pos ));
+}
+#line 51 "hb-buffer-deserialize-text.rl"
+	{
+	tok = p;
+}
+	break;
+	case 4:
+#line 55 "hb-buffer-deserialize-text.rl"
+	{
+	if (!hb_font_glyph_from_string (font,
+					tok, p - tok,
+					&info.codepoint))
+	  return false;
+}
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 9:
+#line 62 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_uint (tok, p, &info.cluster )) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 11:
+#line 64 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.y_offset )) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 6:
+#line 65 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.x_advance)) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 8:
+#line 66 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.y_advance)) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+#line 480 "hb-buffer-deserialize-text.hh"
+	}
+
+_again:
+	if ( cs == 0 )
+		goto _out;
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	if ( p == eof )
+	{
+	switch ( _deserialize_text_eof_actions[cs] ) {
+	case 4:
+#line 55 "hb-buffer-deserialize-text.rl"
+	{
+	if (!hb_font_glyph_from_string (font,
+					tok, p - tok,
+					&info.codepoint))
+	  return false;
+}
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 9:
+#line 62 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_uint (tok, p, &info.cluster )) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 11:
+#line 64 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.y_offset )) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 6:
+#line 65 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.x_advance)) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+	case 8:
+#line 66 "hb-buffer-deserialize-text.rl"
+	{ if (!parse_int  (tok, p, &pos.y_advance)) return false; }
+#line 43 "hb-buffer-deserialize-text.rl"
+	{
+	buffer->add_info (info);
+	if (unlikely (!buffer->successful))
+	  return false;
+	buffer->pos[buffer->len - 1] = pos;
+	*end_ptr = p;
+}
+	break;
+#line 557 "hb-buffer-deserialize-text.hh"
+	}
+	}
+
+	_out: {}
+	}
+
+#line 119 "hb-buffer-deserialize-text.rl"
+
+
+  *end_ptr = p;
+
+  return p == pe && *(p-1) != ']';
+}
+
+#endif /* HB_BUFFER_DESERIALIZE_TEXT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-buffer-serialize.cc b/thirdparty/harfbuzz/src/hb-buffer-serialize.cc
new file mode 100644
index 0000000000..bc6c978b38
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer-serialize.cc
@@ -0,0 +1,474 @@
+/*
+ * Copyright © 2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_BUFFER_SERIALIZE
+
+#include "hb-buffer.hh"
+
+
+static const char *serialize_formats[] = {
+  "text",
+  "json",
+  nullptr
+};
+
+/**
+ * hb_buffer_serialize_list_formats:
+ *
+ * Returns a list of supported buffer serialization formats.
+ *
+ * Return value: (transfer none):
+ * A string array of buffer serialization formats. Should not be freed.
+ *
+ * Since: 0.9.7
+ **/
+const char **
+hb_buffer_serialize_list_formats ()
+{
+  return serialize_formats;
+}
+
+/**
+ * hb_buffer_serialize_format_from_string:
+ * @str: (array length=len) (element-type uint8_t): a string to parse
+ * @len: length of @str, or -1 if string is %NULL terminated
+ *
+ * Parses a string into an #hb_buffer_serialize_format_t. Does not check if
+ * @str is a valid buffer serialization format, use
+ * hb_buffer_serialize_list_formats() to get the list of supported formats.
+ *
+ * Return value:
+ * The parsed #hb_buffer_serialize_format_t.
+ *
+ * Since: 0.9.7
+ **/
+hb_buffer_serialize_format_t
+hb_buffer_serialize_format_from_string (const char *str, int len)
+{
+  /* Upper-case it. */
+  return (hb_buffer_serialize_format_t) (hb_tag_from_string (str, len) & ~0x20202020u);
+}
+
+/**
+ * hb_buffer_serialize_format_to_string:
+ * @format: an #hb_buffer_serialize_format_t to convert.
+ *
+ * Converts @format to the string corresponding it, or %NULL if it is not a valid
+ * #hb_buffer_serialize_format_t.
+ *
+ * Return value: (transfer none):
+ * A %NULL terminated string corresponding to @format. Should not be freed.
+ *
+ * Since: 0.9.7
+ **/
+const char *
+hb_buffer_serialize_format_to_string (hb_buffer_serialize_format_t format)
+{
+  switch ((unsigned) format)
+  {
+    case HB_BUFFER_SERIALIZE_FORMAT_TEXT:	return serialize_formats[0];
+    case HB_BUFFER_SERIALIZE_FORMAT_JSON:	return serialize_formats[1];
+    default:
+    case HB_BUFFER_SERIALIZE_FORMAT_INVALID:	return nullptr;
+  }
+}
+
+static unsigned int
+_hb_buffer_serialize_glyphs_json (hb_buffer_t *buffer,
+				  unsigned int start,
+				  unsigned int end,
+				  char *buf,
+				  unsigned int buf_size,
+				  unsigned int *buf_consumed,
+				  hb_font_t *font,
+				  hb_buffer_serialize_flags_t flags)
+{
+  hb_glyph_info_t *info = hb_buffer_get_glyph_infos (buffer, nullptr);
+  hb_glyph_position_t *pos = (flags & HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS) ?
+			     nullptr : hb_buffer_get_glyph_positions (buffer, nullptr);
+
+  *buf_consumed = 0;
+  hb_position_t x = 0, y = 0;
+  for (unsigned int i = start; i < end; i++)
+  {
+    char b[1024];
+    char *p = b;
+
+    /* In the following code, we know b is large enough that no overflow can happen. */
+
+#define APPEND(s) HB_STMT_START { strcpy (p, s); p += strlen (s); } HB_STMT_END
+
+    if (i)
+      *p++ = ',';
+
+    *p++ = '{';
+
+    APPEND ("\"g\":");
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_GLYPH_NAMES))
+    {
+      char g[128];
+      hb_font_glyph_to_string (font, info[i].codepoint, g, sizeof (g));
+      *p++ = '"';
+      for (char *q = g; *q; q++) {
+	if (*q == '"')
+	  *p++ = '\\';
+	*p++ = *q;
+      }
+      *p++ = '"';
+    }
+    else
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "%u", info[i].codepoint));
+
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_CLUSTERS)) {
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"cl\":%u", info[i].cluster));
+    }
+
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS))
+    {
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"dx\":%d,\"dy\":%d",
+			     x+pos[i].x_offset, y+pos[i].y_offset));
+      if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES))
+	p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"ax\":%d,\"ay\":%d",
+			       pos[i].x_advance, pos[i].y_advance));
+    }
+
+    if (flags & HB_BUFFER_SERIALIZE_FLAG_GLYPH_FLAGS)
+    {
+      if (info[i].mask & HB_GLYPH_FLAG_DEFINED)
+	p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"fl\":%u", info[i].mask & HB_GLYPH_FLAG_DEFINED));
+    }
+
+    if (flags & HB_BUFFER_SERIALIZE_FLAG_GLYPH_EXTENTS)
+    {
+      hb_glyph_extents_t extents;
+      hb_font_get_glyph_extents(font, info[i].codepoint, &extents);
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"xb\":%d,\"yb\":%d",
+		extents.x_bearing, extents.y_bearing));
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",\"w\":%d,\"h\":%d",
+		extents.width, extents.height));
+    }
+
+    *p++ = '}';
+
+    unsigned int l = p - b;
+    if (buf_size > l)
+    {
+      memcpy (buf, b, l);
+      buf += l;
+      buf_size -= l;
+      *buf_consumed += l;
+      *buf = '\0';
+    } else
+      return i - start;
+
+    if (pos && (flags & HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES))
+    {
+      x += pos[i].x_advance;
+      y += pos[i].y_advance;
+    }
+  }
+
+  return end - start;
+}
+
+static unsigned int
+_hb_buffer_serialize_glyphs_text (hb_buffer_t *buffer,
+				  unsigned int start,
+				  unsigned int end,
+				  char *buf,
+				  unsigned int buf_size,
+				  unsigned int *buf_consumed,
+				  hb_font_t *font,
+				  hb_buffer_serialize_flags_t flags)
+{
+  hb_glyph_info_t *info = hb_buffer_get_glyph_infos (buffer, nullptr);
+  hb_glyph_position_t *pos = (flags & HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS) ?
+			     nullptr : hb_buffer_get_glyph_positions (buffer, nullptr);
+
+  *buf_consumed = 0;
+  hb_position_t x = 0, y = 0;
+  for (unsigned int i = start; i < end; i++)
+  {
+    char b[1024];
+    char *p = b;
+
+    /* In the following code, we know b is large enough that no overflow can happen. */
+
+    if (i)
+      *p++ = '|';
+
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_GLYPH_NAMES))
+    {
+      hb_font_glyph_to_string (font, info[i].codepoint, p, 128);
+      p += strlen (p);
+    }
+    else
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "%u", info[i].codepoint));
+
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_CLUSTERS)) {
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "=%u", info[i].cluster));
+    }
+
+    if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS))
+    {
+      if (x+pos[i].x_offset || y+pos[i].y_offset)
+	p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "@%d,%d", x+pos[i].x_offset, y+pos[i].y_offset));
+
+      if (!(flags & HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES))
+      {
+	*p++ = '+';
+	p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "%d", pos[i].x_advance));
+	if (pos[i].y_advance)
+	  p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), ",%d", pos[i].y_advance));
+      }
+    }
+
+    if (flags & HB_BUFFER_SERIALIZE_FLAG_GLYPH_FLAGS)
+    {
+      if (info[i].mask & HB_GLYPH_FLAG_DEFINED)
+	p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "#%X", info[i].mask &HB_GLYPH_FLAG_DEFINED));
+    }
+
+    if (flags & HB_BUFFER_SERIALIZE_FLAG_GLYPH_EXTENTS)
+    {
+      hb_glyph_extents_t extents;
+      hb_font_get_glyph_extents(font, info[i].codepoint, &extents);
+      p += hb_max (0, snprintf (p, ARRAY_LENGTH (b) - (p - b), "<%d,%d,%d,%d>", extents.x_bearing, extents.y_bearing, extents.width, extents.height));
+    }
+
+    unsigned int l = p - b;
+    if (buf_size > l)
+    {
+      memcpy (buf, b, l);
+      buf += l;
+      buf_size -= l;
+      *buf_consumed += l;
+      *buf = '\0';
+    } else
+      return i - start;
+
+    if (pos && (flags & HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES))
+    {
+      x += pos[i].x_advance;
+      y += pos[i].y_advance;
+    }
+  }
+
+  return end - start;
+}
+
+/**
+ * hb_buffer_serialize_glyphs:
+ * @buffer: an #hb_buffer_t buffer.
+ * @start: the first item in @buffer to serialize.
+ * @end: the last item in @buffer to serialize.
+ * @buf: (out) (array length=buf_size) (element-type uint8_t): output string to
+ *       write serialized buffer into.
+ * @buf_size: the size of @buf.
+ * @buf_consumed: (out) (allow-none): if not %NULL, will be set to the number of byes written into @buf.
+ * @font: (allow-none): the #hb_font_t used to shape this buffer, needed to
+ *        read glyph names and extents. If %NULL, and empty font will be used.
+ * @format: the #hb_buffer_serialize_format_t to use for formatting the output.
+ * @flags: the #hb_buffer_serialize_flags_t that control what glyph properties
+ *         to serialize.
+ *
+ * Serializes @buffer into a textual representation of its glyph content,
+ * useful for showing the contents of the buffer, for example during debugging.
+ * There are currently two supported serialization formats:
+ *
+ * ## text
+ * A human-readable, plain text format.
+ * The serialized glyphs will look something like:
+ *
+ * ```
+ * [uni0651=0@518,0+0|uni0628=0+1897]
+ * ```
+ * - The serialized glyphs are delimited with `[` and `]`.
+ * - Glyphs are separated with `|`
+ * - Each glyph starts with glyph name, or glyph index if
+ *   #HB_BUFFER_SERIALIZE_FLAG_NO_GLYPH_NAMES flag is set. Then,
+ *   - If #HB_BUFFER_SERIALIZE_FLAG_NO_CLUSTERS is not set, `=` then #hb_glyph_info_t.cluster.
+ *   - If #HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS is not set, the #hb_glyph_position_t in the format:
+ *     - If both #hb_glyph_position_t.x_offset and #hb_glyph_position_t.y_offset are not 0, `@x_offset,y_offset`. Then,
+ *     - `+x_advance`, then `,y_advance` if #hb_glyph_position_t.y_advance is not 0. Then,
+ *   - If #HB_BUFFER_SERIALIZE_FLAG_GLYPH_EXTENTS is set, the
+ *     #hb_glyph_extents_t in the format
+ *     `&lt;x_bearing,y_bearing,width,height&gt;`
+ *
+ * ## json
+ * TODO.
+ *
+ * Return value:
+ * The number of serialized items.
+ *
+ * Since: 0.9.7
+ **/
+unsigned int
+hb_buffer_serialize_glyphs (hb_buffer_t *buffer,
+			    unsigned int start,
+			    unsigned int end,
+			    char *buf,
+			    unsigned int buf_size,
+			    unsigned int *buf_consumed,
+			    hb_font_t *font,
+			    hb_buffer_serialize_format_t format,
+			    hb_buffer_serialize_flags_t flags)
+{
+  assert (start <= end && end <= buffer->len);
+
+  unsigned int sconsumed;
+  if (!buf_consumed)
+    buf_consumed = &sconsumed;
+  *buf_consumed = 0;
+  if (buf_size)
+    *buf = '\0';
+
+  assert ((!buffer->len && (buffer->content_type == HB_BUFFER_CONTENT_TYPE_INVALID)) ||
+	  (buffer->content_type == HB_BUFFER_CONTENT_TYPE_GLYPHS));
+
+  if (!buffer->have_positions)
+    flags |= HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS;
+
+  if (unlikely (start == end))
+    return 0;
+
+  if (!font)
+    font = hb_font_get_empty ();
+
+  switch (format)
+  {
+    case HB_BUFFER_SERIALIZE_FORMAT_TEXT:
+      return _hb_buffer_serialize_glyphs_text (buffer, start, end,
+					       buf, buf_size, buf_consumed,
+					       font, flags);
+
+    case HB_BUFFER_SERIALIZE_FORMAT_JSON:
+      return _hb_buffer_serialize_glyphs_json (buffer, start, end,
+					       buf, buf_size, buf_consumed,
+					       font, flags);
+
+    default:
+    case HB_BUFFER_SERIALIZE_FORMAT_INVALID:
+      return 0;
+
+  }
+}
+
+static bool
+parse_int (const char *pp, const char *end, int32_t *pv)
+{
+  int v;
+  const char *p = pp;
+  if (unlikely (!hb_parse_int (&p, end, &v, true/* whole buffer */)))
+    return false;
+
+  *pv = v;
+  return true;
+}
+
+static bool
+parse_uint (const char *pp, const char *end, uint32_t *pv)
+{
+  unsigned int v;
+  const char *p = pp;
+  if (unlikely (!hb_parse_uint (&p, end, &v, true/* whole buffer */)))
+    return false;
+
+  *pv = v;
+  return true;
+}
+
+#include "hb-buffer-deserialize-json.hh"
+#include "hb-buffer-deserialize-text.hh"
+
+/**
+ * hb_buffer_deserialize_glyphs:
+ * @buffer: an #hb_buffer_t buffer.
+ * @buf: (array length=buf_len):
+ * @buf_len:
+ * @end_ptr: (out):
+ * @font:
+ * @format:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_buffer_deserialize_glyphs (hb_buffer_t *buffer,
+			      const char *buf,
+			      int buf_len, /* -1 means nul-terminated */
+			      const char **end_ptr, /* May be NULL */
+			      hb_font_t *font, /* May be NULL */
+			      hb_buffer_serialize_format_t format)
+{
+  const char *end;
+  if (!end_ptr)
+    end_ptr = &end;
+  *end_ptr = buf;
+
+  assert ((!buffer->len && (buffer->content_type == HB_BUFFER_CONTENT_TYPE_INVALID)) ||
+	  (buffer->content_type == HB_BUFFER_CONTENT_TYPE_GLYPHS));
+
+  if (buf_len == -1)
+    buf_len = strlen (buf);
+
+  if (!buf_len)
+  {
+    *end_ptr = buf;
+    return false;
+  }
+
+  hb_buffer_set_content_type (buffer, HB_BUFFER_CONTENT_TYPE_GLYPHS);
+
+  if (!font)
+    font = hb_font_get_empty ();
+
+  switch (format)
+  {
+    case HB_BUFFER_SERIALIZE_FORMAT_TEXT:
+      return _hb_buffer_deserialize_glyphs_text (buffer,
+						 buf, buf_len, end_ptr,
+						 font);
+
+    case HB_BUFFER_SERIALIZE_FORMAT_JSON:
+      return _hb_buffer_deserialize_glyphs_json (buffer,
+						 buf, buf_len, end_ptr,
+						 font);
+
+    default:
+    case HB_BUFFER_SERIALIZE_FORMAT_INVALID:
+      return false;
+
+  }
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-buffer.cc b/thirdparty/harfbuzz/src/hb-buffer.cc
new file mode 100644
index 0000000000..4fadbb78d2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer.cc
@@ -0,0 +1,2004 @@
+/*
+ * Copyright © 1998-2004  David Turner and Werner Lemberg
+ * Copyright © 2004,2007,2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Owen Taylor, Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb-buffer.hh"
+#include "hb-utf.hh"
+
+
+/**
+ * SECTION: hb-buffer
+ * @title: hb-buffer
+ * @short_description: Input and output buffers
+ * @include: hb.h
+ *
+ * Buffers serve dual role in HarfBuzz; they hold the input characters that are
+ * passed to hb_shape(), and after shaping they hold the output glyphs.
+ **/
+
+
+/**
+ * hb_segment_properties_equal:
+ * @a: first #hb_segment_properties_t to compare.
+ * @b: second #hb_segment_properties_t to compare.
+ *
+ * Checks the equality of two #hb_segment_properties_t's.
+ *
+ * Return value:
+ * %true if all properties of @a equal those of @b, false otherwise.
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_segment_properties_equal (const hb_segment_properties_t *a,
+			     const hb_segment_properties_t *b)
+{
+  return a->direction == b->direction &&
+	 a->script    == b->script    &&
+	 a->language  == b->language  &&
+	 a->reserved1 == b->reserved1 &&
+	 a->reserved2 == b->reserved2;
+
+}
+
+/**
+ * hb_segment_properties_hash:
+ * @p: #hb_segment_properties_t to hash.
+ *
+ * Creates a hash representing @p.
+ *
+ * Return value:
+ * A hash of @p.
+ *
+ * Since: 0.9.7
+ **/
+unsigned int
+hb_segment_properties_hash (const hb_segment_properties_t *p)
+{
+  return (unsigned int) p->direction ^
+	 (unsigned int) p->script ^
+	 (intptr_t) (p->language);
+}
+
+
+
+/* Here is how the buffer works internally:
+ *
+ * There are two info pointers: info and out_info.  They always have
+ * the same allocated size, but different lengths.
+ *
+ * As an optimization, both info and out_info may point to the
+ * same piece of memory, which is owned by info.  This remains the
+ * case as long as out_len doesn't exceed i at any time.
+ * In that case, swap_buffers() is no-op and the glyph operations operate
+ * mostly in-place.
+ *
+ * As soon as out_info gets longer than info, out_info is moved over
+ * to an alternate buffer (which we reuse the pos buffer for!), and its
+ * current contents (out_len entries) are copied to the new place.
+ * This should all remain transparent to the user.  swap_buffers() then
+ * switches info and out_info.
+ */
+
+
+
+/* Internal API */
+
+bool
+hb_buffer_t::enlarge (unsigned int size)
+{
+  if (unlikely (!successful))
+    return false;
+  if (unlikely (size > max_len))
+  {
+    successful = false;
+    return false;
+  }
+
+  unsigned int new_allocated = allocated;
+  hb_glyph_position_t *new_pos = nullptr;
+  hb_glyph_info_t *new_info = nullptr;
+  bool separate_out = out_info != info;
+
+  if (unlikely (hb_unsigned_mul_overflows (size, sizeof (info[0]))))
+    goto done;
+
+  while (size >= new_allocated)
+    new_allocated += (new_allocated >> 1) + 32;
+
+  static_assert ((sizeof (info[0]) == sizeof (pos[0])), "");
+  if (unlikely (hb_unsigned_mul_overflows (new_allocated, sizeof (info[0]))))
+    goto done;
+
+  new_pos = (hb_glyph_position_t *) realloc (pos, new_allocated * sizeof (pos[0]));
+  new_info = (hb_glyph_info_t *) realloc (info, new_allocated * sizeof (info[0]));
+
+done:
+  if (unlikely (!new_pos || !new_info))
+    successful = false;
+
+  if (likely (new_pos))
+    pos = new_pos;
+
+  if (likely (new_info))
+    info = new_info;
+
+  out_info = separate_out ? (hb_glyph_info_t *) pos : info;
+  if (likely (successful))
+    allocated = new_allocated;
+
+  return likely (successful);
+}
+
+bool
+hb_buffer_t::make_room_for (unsigned int num_in,
+			    unsigned int num_out)
+{
+  if (unlikely (!ensure (out_len + num_out))) return false;
+
+  if (out_info == info &&
+      out_len + num_out > idx + num_in)
+  {
+    assert (have_output);
+
+    out_info = (hb_glyph_info_t *) pos;
+    memcpy (out_info, info, out_len * sizeof (out_info[0]));
+  }
+
+  return true;
+}
+
+bool
+hb_buffer_t::shift_forward (unsigned int count)
+{
+  assert (have_output);
+  if (unlikely (!ensure (len + count))) return false;
+
+  memmove (info + idx + count, info + idx, (len - idx) * sizeof (info[0]));
+  if (idx + count > len)
+  {
+    /* Under memory failure we might expose this area.  At least
+     * clean it up.  Oh well...
+     *
+     * Ideally, we should at least set Default_Ignorable bits on
+     * these, as well as consistent cluster values.  But the former
+     * is layering violation... */
+    memset (info + len, 0, (idx + count - len) * sizeof (info[0]));
+  }
+  len += count;
+  idx += count;
+
+  return true;
+}
+
+hb_buffer_t::scratch_buffer_t *
+hb_buffer_t::get_scratch_buffer (unsigned int *size)
+{
+  have_output = false;
+  have_positions = false;
+
+  out_len = 0;
+  out_info = info;
+
+  assert ((uintptr_t) pos % sizeof (scratch_buffer_t) == 0);
+  *size = allocated * sizeof (pos[0]) / sizeof (scratch_buffer_t);
+  return (scratch_buffer_t *) (void *) pos;
+}
+
+
+
+/* HarfBuzz-Internal API */
+
+void
+hb_buffer_t::reset ()
+{
+  if (unlikely (hb_object_is_immutable (this)))
+    return;
+
+  hb_unicode_funcs_destroy (unicode);
+  unicode = hb_unicode_funcs_reference (hb_unicode_funcs_get_default ());
+  flags = HB_BUFFER_FLAG_DEFAULT;
+  replacement = HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT;
+  invisible = 0;
+
+  clear ();
+}
+
+void
+hb_buffer_t::clear ()
+{
+  if (unlikely (hb_object_is_immutable (this)))
+    return;
+
+  hb_segment_properties_t default_props = HB_SEGMENT_PROPERTIES_DEFAULT;
+  props = default_props;
+  scratch_flags = HB_BUFFER_SCRATCH_FLAG_DEFAULT;
+
+  content_type = HB_BUFFER_CONTENT_TYPE_INVALID;
+  successful = true;
+  have_output = false;
+  have_positions = false;
+
+  idx = 0;
+  len = 0;
+  out_len = 0;
+  out_info = info;
+
+  serial = 0;
+
+  memset (context, 0, sizeof context);
+  memset (context_len, 0, sizeof context_len);
+
+  deallocate_var_all ();
+}
+
+void
+hb_buffer_t::add (hb_codepoint_t  codepoint,
+		  unsigned int    cluster)
+{
+  hb_glyph_info_t *glyph;
+
+  if (unlikely (!ensure (len + 1))) return;
+
+  glyph = &info[len];
+
+  memset (glyph, 0, sizeof (*glyph));
+  glyph->codepoint = codepoint;
+  glyph->mask = 0;
+  glyph->cluster = cluster;
+
+  len++;
+}
+
+void
+hb_buffer_t::add_info (const hb_glyph_info_t &glyph_info)
+{
+  if (unlikely (!ensure (len + 1))) return;
+
+  info[len] = glyph_info;
+
+  len++;
+}
+
+
+void
+hb_buffer_t::remove_output ()
+{
+  if (unlikely (hb_object_is_immutable (this)))
+    return;
+
+  have_output = false;
+  have_positions = false;
+
+  out_len = 0;
+  out_info = info;
+}
+
+void
+hb_buffer_t::clear_output ()
+{
+  if (unlikely (hb_object_is_immutable (this)))
+    return;
+
+  have_output = true;
+  have_positions = false;
+
+  out_len = 0;
+  out_info = info;
+}
+
+void
+hb_buffer_t::clear_positions ()
+{
+  if (unlikely (hb_object_is_immutable (this)))
+    return;
+
+  have_output = false;
+  have_positions = true;
+
+  out_len = 0;
+  out_info = info;
+
+  hb_memset (pos, 0, sizeof (pos[0]) * len);
+}
+
+void
+hb_buffer_t::swap_buffers ()
+{
+  if (unlikely (!successful)) return;
+
+  assert (have_output);
+  have_output = false;
+
+  if (out_info != info)
+  {
+    hb_glyph_info_t *tmp_string;
+    tmp_string = info;
+    info = out_info;
+    out_info = tmp_string;
+    pos = (hb_glyph_position_t *) out_info;
+  }
+
+  unsigned int tmp;
+  tmp = len;
+  len = out_len;
+  out_len = tmp;
+
+  idx = 0;
+}
+
+
+void
+hb_buffer_t::replace_glyphs (unsigned int num_in,
+			     unsigned int num_out,
+			     const uint32_t *glyph_data)
+{
+  if (unlikely (!make_room_for (num_in, num_out))) return;
+
+  assert (idx + num_in <= len);
+
+  merge_clusters (idx, idx + num_in);
+
+  hb_glyph_info_t orig_info = info[idx];
+  hb_glyph_info_t *pinfo = &out_info[out_len];
+  for (unsigned int i = 0; i < num_out; i++)
+  {
+    *pinfo = orig_info;
+    pinfo->codepoint = glyph_data[i];
+    pinfo++;
+  }
+
+  idx  += num_in;
+  out_len += num_out;
+}
+
+bool
+hb_buffer_t::move_to (unsigned int i)
+{
+  if (!have_output)
+  {
+    assert (i <= len);
+    idx = i;
+    return true;
+  }
+  if (unlikely (!successful))
+    return false;
+
+  assert (i <= out_len + (len - idx));
+
+  if (out_len < i)
+  {
+    unsigned int count = i - out_len;
+    if (unlikely (!make_room_for (count, count))) return false;
+
+    memmove (out_info + out_len, info + idx, count * sizeof (out_info[0]));
+    idx += count;
+    out_len += count;
+  }
+  else if (out_len > i)
+  {
+    /* Tricky part: rewinding... */
+    unsigned int count = out_len - i;
+
+    /* This will blow in our face if memory allocation fails later
+     * in this same lookup...
+     *
+     * We used to shift with extra 32 items, instead of the 0 below.
+     * But that would leave empty slots in the buffer in case of allocation
+     * failures.  Setting to zero for now to avoid other problems (see
+     * comments in shift_forward().  This can cause O(N^2) behavior more
+     * severely than adding 32 empty slots can... */
+    if (unlikely (idx < count && !shift_forward (count + 0))) return false;
+
+    assert (idx >= count);
+
+    idx -= count;
+    out_len -= count;
+    memmove (info + idx, out_info + out_len, count * sizeof (out_info[0]));
+  }
+
+  return true;
+}
+
+
+void
+hb_buffer_t::set_masks (hb_mask_t    value,
+			hb_mask_t    mask,
+			unsigned int cluster_start,
+			unsigned int cluster_end)
+{
+  hb_mask_t not_mask = ~mask;
+  value &= mask;
+
+  if (!mask)
+    return;
+
+  unsigned int count = len;
+  for (unsigned int i = 0; i < count; i++)
+    if (cluster_start <= info[i].cluster && info[i].cluster < cluster_end)
+      info[i].mask = (info[i].mask & not_mask) | value;
+}
+
+void
+hb_buffer_t::reverse_range (unsigned int start,
+			    unsigned int end)
+{
+  if (end - start < 2)
+    return;
+
+  hb_array_t<hb_glyph_info_t> (info, len).reverse (start, end);
+
+  if (have_positions) {
+    hb_array_t<hb_glyph_position_t> (pos, len).reverse (start, end);
+  }
+}
+
+void
+hb_buffer_t::reverse ()
+{
+  if (unlikely (!len))
+    return;
+
+  reverse_range (0, len);
+}
+
+void
+hb_buffer_t::reverse_clusters ()
+{
+  unsigned int i, start, count, last_cluster;
+
+  if (unlikely (!len))
+    return;
+
+  reverse ();
+
+  count = len;
+  start = 0;
+  last_cluster = info[0].cluster;
+  for (i = 1; i < count; i++) {
+    if (last_cluster != info[i].cluster) {
+      reverse_range (start, i);
+      start = i;
+      last_cluster = info[i].cluster;
+    }
+  }
+  reverse_range (start, i);
+}
+
+void
+hb_buffer_t::merge_clusters_impl (unsigned int start,
+				  unsigned int end)
+{
+  if (cluster_level == HB_BUFFER_CLUSTER_LEVEL_CHARACTERS)
+  {
+    unsafe_to_break (start, end);
+    return;
+  }
+
+  unsigned int cluster = info[start].cluster;
+
+  for (unsigned int i = start + 1; i < end; i++)
+    cluster = hb_min (cluster, info[i].cluster);
+
+  /* Extend end */
+  while (end < len && info[end - 1].cluster == info[end].cluster)
+    end++;
+
+  /* Extend start */
+  while (idx < start && info[start - 1].cluster == info[start].cluster)
+    start--;
+
+  /* If we hit the start of buffer, continue in out-buffer. */
+  if (idx == start)
+    for (unsigned int i = out_len; i && out_info[i - 1].cluster == info[start].cluster; i--)
+      set_cluster (out_info[i - 1], cluster);
+
+  for (unsigned int i = start; i < end; i++)
+    set_cluster (info[i], cluster);
+}
+void
+hb_buffer_t::merge_out_clusters (unsigned int start,
+				 unsigned int end)
+{
+  if (cluster_level == HB_BUFFER_CLUSTER_LEVEL_CHARACTERS)
+    return;
+
+  if (unlikely (end - start < 2))
+    return;
+
+  unsigned int cluster = out_info[start].cluster;
+
+  for (unsigned int i = start + 1; i < end; i++)
+    cluster = hb_min (cluster, out_info[i].cluster);
+
+  /* Extend start */
+  while (start && out_info[start - 1].cluster == out_info[start].cluster)
+    start--;
+
+  /* Extend end */
+  while (end < out_len && out_info[end - 1].cluster == out_info[end].cluster)
+    end++;
+
+  /* If we hit the end of out-buffer, continue in buffer. */
+  if (end == out_len)
+    for (unsigned int i = idx; i < len && info[i].cluster == out_info[end - 1].cluster; i++)
+      set_cluster (info[i], cluster);
+
+  for (unsigned int i = start; i < end; i++)
+    set_cluster (out_info[i], cluster);
+}
+void
+hb_buffer_t::delete_glyph ()
+{
+  /* The logic here is duplicated in hb_ot_hide_default_ignorables(). */
+
+  unsigned int cluster = info[idx].cluster;
+  if (idx + 1 < len && cluster == info[idx + 1].cluster)
+  {
+    /* Cluster survives; do nothing. */
+    goto done;
+  }
+
+  if (out_len)
+  {
+    /* Merge cluster backward. */
+    if (cluster < out_info[out_len - 1].cluster)
+    {
+      unsigned int mask = info[idx].mask;
+      unsigned int old_cluster = out_info[out_len - 1].cluster;
+      for (unsigned i = out_len; i && out_info[i - 1].cluster == old_cluster; i--)
+	set_cluster (out_info[i - 1], cluster, mask);
+    }
+    goto done;
+  }
+
+  if (idx + 1 < len)
+  {
+    /* Merge cluster forward. */
+    merge_clusters (idx, idx + 2);
+    goto done;
+  }
+
+done:
+  skip_glyph ();
+}
+
+void
+hb_buffer_t::unsafe_to_break_impl (unsigned int start, unsigned int end)
+{
+  unsigned int cluster = UINT_MAX;
+  cluster = _unsafe_to_break_find_min_cluster (info, start, end, cluster);
+  _unsafe_to_break_set_mask (info, start, end, cluster);
+}
+void
+hb_buffer_t::unsafe_to_break_from_outbuffer (unsigned int start, unsigned int end)
+{
+  if (!have_output)
+  {
+    unsafe_to_break_impl (start, end);
+    return;
+  }
+
+  assert (start <= out_len);
+  assert (idx <= end);
+
+  unsigned int cluster = UINT_MAX;
+  cluster = _unsafe_to_break_find_min_cluster (out_info, start, out_len, cluster);
+  cluster = _unsafe_to_break_find_min_cluster (info, idx, end, cluster);
+  _unsafe_to_break_set_mask (out_info, start, out_len, cluster);
+  _unsafe_to_break_set_mask (info, idx, end, cluster);
+}
+
+void
+hb_buffer_t::guess_segment_properties ()
+{
+  assert ((content_type == HB_BUFFER_CONTENT_TYPE_UNICODE) ||
+	  (!len && (content_type == HB_BUFFER_CONTENT_TYPE_INVALID)));
+
+  /* If script is set to INVALID, guess from buffer contents */
+  if (props.script == HB_SCRIPT_INVALID) {
+    for (unsigned int i = 0; i < len; i++) {
+      hb_script_t script = unicode->script (info[i].codepoint);
+      if (likely (script != HB_SCRIPT_COMMON &&
+		  script != HB_SCRIPT_INHERITED &&
+		  script != HB_SCRIPT_UNKNOWN)) {
+	props.script = script;
+	break;
+      }
+    }
+  }
+
+  /* If direction is set to INVALID, guess from script */
+  if (props.direction == HB_DIRECTION_INVALID) {
+    props.direction = hb_script_get_horizontal_direction (props.script);
+    if (props.direction == HB_DIRECTION_INVALID)
+      props.direction = HB_DIRECTION_LTR;
+  }
+
+  /* If language is not set, use default language from locale */
+  if (props.language == HB_LANGUAGE_INVALID) {
+    /* TODO get_default_for_script? using $LANGUAGE */
+    props.language = hb_language_get_default ();
+  }
+}
+
+
+/* Public API */
+
+DEFINE_NULL_INSTANCE (hb_buffer_t) =
+{
+  HB_OBJECT_HEADER_STATIC,
+
+  const_cast<hb_unicode_funcs_t *> (&_hb_Null_hb_unicode_funcs_t),
+  HB_BUFFER_FLAG_DEFAULT,
+  HB_BUFFER_CLUSTER_LEVEL_DEFAULT,
+  HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT,
+  0, /* invisible */
+  HB_BUFFER_SCRATCH_FLAG_DEFAULT,
+  HB_BUFFER_MAX_LEN_DEFAULT,
+  HB_BUFFER_MAX_OPS_DEFAULT,
+
+  HB_BUFFER_CONTENT_TYPE_INVALID,
+  HB_SEGMENT_PROPERTIES_DEFAULT,
+  false, /* successful */
+  true, /* have_output */
+  true  /* have_positions */
+
+  /* Zero is good enough for everything else. */
+};
+
+
+/**
+ * hb_buffer_create: (Xconstructor)
+ *
+ * Creates a new #hb_buffer_t with all properties to defaults.
+ *
+ * Return value: (transfer full):
+ * A newly allocated #hb_buffer_t with a reference count of 1. The initial
+ * reference count should be released with hb_buffer_destroy() when you are done
+ * using the #hb_buffer_t. This function never returns %NULL. If memory cannot
+ * be allocated, a special #hb_buffer_t object will be returned on which
+ * hb_buffer_allocation_successful() returns %false.
+ *
+ * Since: 0.9.2
+ **/
+hb_buffer_t *
+hb_buffer_create ()
+{
+  hb_buffer_t *buffer;
+
+  if (!(buffer = hb_object_create<hb_buffer_t> ()))
+    return hb_buffer_get_empty ();
+
+  buffer->max_len = HB_BUFFER_MAX_LEN_DEFAULT;
+  buffer->max_ops = HB_BUFFER_MAX_OPS_DEFAULT;
+
+  buffer->reset ();
+
+  return buffer;
+}
+
+/**
+ * hb_buffer_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_buffer_t *
+hb_buffer_get_empty ()
+{
+  return const_cast<hb_buffer_t *> (&Null (hb_buffer_t));
+}
+
+/**
+ * hb_buffer_reference: (skip)
+ * @buffer: an #hb_buffer_t.
+ *
+ * Increases the reference count on @buffer by one. This prevents @buffer from
+ * being destroyed until a matching call to hb_buffer_destroy() is made.
+ *
+ * Return value: (transfer full):
+ * The referenced #hb_buffer_t.
+ *
+ * Since: 0.9.2
+ **/
+hb_buffer_t *
+hb_buffer_reference (hb_buffer_t *buffer)
+{
+  return hb_object_reference (buffer);
+}
+
+/**
+ * hb_buffer_destroy: (skip)
+ * @buffer: an #hb_buffer_t.
+ *
+ * Deallocate the @buffer.
+ * Decreases the reference count on @buffer by one. If the result is zero, then
+ * @buffer and all associated resources are freed. See hb_buffer_reference().
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_destroy (hb_buffer_t *buffer)
+{
+  if (!hb_object_destroy (buffer)) return;
+
+  hb_unicode_funcs_destroy (buffer->unicode);
+
+  free (buffer->info);
+  free (buffer->pos);
+#ifndef HB_NO_BUFFER_MESSAGE
+  if (buffer->message_destroy)
+    buffer->message_destroy (buffer->message_data);
+#endif
+
+  free (buffer);
+}
+
+/**
+ * hb_buffer_set_user_data: (skip)
+ * @buffer: an #hb_buffer_t.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_buffer_set_user_data (hb_buffer_t        *buffer,
+			 hb_user_data_key_t *key,
+			 void *              data,
+			 hb_destroy_func_t   destroy,
+			 hb_bool_t           replace)
+{
+  return hb_object_set_user_data (buffer, key, data, destroy, replace);
+}
+
+/**
+ * hb_buffer_get_user_data: (skip)
+ * @buffer: an #hb_buffer_t.
+ * @key:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_buffer_get_user_data (hb_buffer_t        *buffer,
+			 hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (buffer, key);
+}
+
+
+/**
+ * hb_buffer_set_content_type:
+ * @buffer: an #hb_buffer_t.
+ * @content_type: the type of buffer contents to set
+ *
+ * Sets the type of @buffer contents, buffers are either empty, contain
+ * characters (before shaping) or glyphs (the result of shaping).
+ *
+ * Since: 0.9.5
+ **/
+void
+hb_buffer_set_content_type (hb_buffer_t              *buffer,
+			    hb_buffer_content_type_t  content_type)
+{
+  buffer->content_type = content_type;
+}
+
+/**
+ * hb_buffer_get_content_type:
+ * @buffer: an #hb_buffer_t.
+ *
+ * see hb_buffer_set_content_type().
+ *
+ * Return value:
+ * The type of @buffer contents.
+ *
+ * Since: 0.9.5
+ **/
+hb_buffer_content_type_t
+hb_buffer_get_content_type (hb_buffer_t *buffer)
+{
+  return buffer->content_type;
+}
+
+
+/**
+ * hb_buffer_set_unicode_funcs:
+ * @buffer: an #hb_buffer_t.
+ * @unicode_funcs:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_set_unicode_funcs (hb_buffer_t        *buffer,
+			     hb_unicode_funcs_t *unicode_funcs)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  if (!unicode_funcs)
+    unicode_funcs = hb_unicode_funcs_get_default ();
+
+  hb_unicode_funcs_reference (unicode_funcs);
+  hb_unicode_funcs_destroy (buffer->unicode);
+  buffer->unicode = unicode_funcs;
+}
+
+/**
+ * hb_buffer_get_unicode_funcs:
+ * @buffer: an #hb_buffer_t.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_unicode_funcs_t *
+hb_buffer_get_unicode_funcs (hb_buffer_t        *buffer)
+{
+  return buffer->unicode;
+}
+
+/**
+ * hb_buffer_set_direction:
+ * @buffer: an #hb_buffer_t.
+ * @direction: the #hb_direction_t of the @buffer
+ *
+ * Set the text flow direction of the buffer. No shaping can happen without
+ * setting @buffer direction, and it controls the visual direction for the
+ * output glyphs; for RTL direction the glyphs will be reversed. Many layout
+ * features depend on the proper setting of the direction, for example,
+ * reversing RTL text before shaping, then shaping with LTR direction is not
+ * the same as keeping the text in logical order and shaping with RTL
+ * direction.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_set_direction (hb_buffer_t    *buffer,
+			 hb_direction_t  direction)
+
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->props.direction = direction;
+}
+
+/**
+ * hb_buffer_get_direction:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_direction()
+ *
+ * Return value:
+ * The direction of the @buffer.
+ *
+ * Since: 0.9.2
+ **/
+hb_direction_t
+hb_buffer_get_direction (hb_buffer_t    *buffer)
+{
+  return buffer->props.direction;
+}
+
+/**
+ * hb_buffer_set_script:
+ * @buffer: an #hb_buffer_t.
+ * @script: an #hb_script_t to set.
+ *
+ * Sets the script of @buffer to @script.
+ *
+ * Script is crucial for choosing the proper shaping behaviour for scripts that
+ * require it (e.g. Arabic) and the which OpenType features defined in the font
+ * to be applied.
+ *
+ * You can pass one of the predefined #hb_script_t values, or use
+ * hb_script_from_string() or hb_script_from_iso15924_tag() to get the
+ * corresponding script from an ISO 15924 script tag.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_set_script (hb_buffer_t *buffer,
+		      hb_script_t  script)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->props.script = script;
+}
+
+/**
+ * hb_buffer_get_script:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_script().
+ *
+ * Return value:
+ * The #hb_script_t of the @buffer.
+ *
+ * Since: 0.9.2
+ **/
+hb_script_t
+hb_buffer_get_script (hb_buffer_t *buffer)
+{
+  return buffer->props.script;
+}
+
+/**
+ * hb_buffer_set_language:
+ * @buffer: an #hb_buffer_t.
+ * @language: an hb_language_t to set.
+ *
+ * Sets the language of @buffer to @language.
+ *
+ * Languages are crucial for selecting which OpenType feature to apply to the
+ * buffer which can result in applying language-specific behaviour. Languages
+ * are orthogonal to the scripts, and though they are related, they are
+ * different concepts and should not be confused with each other.
+ *
+ * Use hb_language_from_string() to convert from BCP 47 language tags to
+ * #hb_language_t.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_set_language (hb_buffer_t   *buffer,
+			hb_language_t  language)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->props.language = language;
+}
+
+/**
+ * hb_buffer_get_language:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_language().
+ *
+ * Return value: (transfer none):
+ * The #hb_language_t of the buffer. Must not be freed by the caller.
+ *
+ * Since: 0.9.2
+ **/
+hb_language_t
+hb_buffer_get_language (hb_buffer_t *buffer)
+{
+  return buffer->props.language;
+}
+
+/**
+ * hb_buffer_set_segment_properties:
+ * @buffer: an #hb_buffer_t.
+ * @props: an #hb_segment_properties_t to use.
+ *
+ * Sets the segment properties of the buffer, a shortcut for calling
+ * hb_buffer_set_direction(), hb_buffer_set_script() and
+ * hb_buffer_set_language() individually.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_buffer_set_segment_properties (hb_buffer_t *buffer,
+				  const hb_segment_properties_t *props)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->props = *props;
+}
+
+/**
+ * hb_buffer_get_segment_properties:
+ * @buffer: an #hb_buffer_t.
+ * @props: (out): the output #hb_segment_properties_t.
+ *
+ * Sets @props to the #hb_segment_properties_t of @buffer.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_buffer_get_segment_properties (hb_buffer_t *buffer,
+				  hb_segment_properties_t *props)
+{
+  *props = buffer->props;
+}
+
+
+/**
+ * hb_buffer_set_flags:
+ * @buffer: an #hb_buffer_t.
+ * @flags: the buffer flags to set.
+ *
+ * Sets @buffer flags to @flags. See #hb_buffer_flags_t.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_buffer_set_flags (hb_buffer_t       *buffer,
+		     hb_buffer_flags_t  flags)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->flags = flags;
+}
+
+/**
+ * hb_buffer_get_flags:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_flags().
+ *
+ * Return value:
+ * The @buffer flags.
+ *
+ * Since: 0.9.7
+ **/
+hb_buffer_flags_t
+hb_buffer_get_flags (hb_buffer_t *buffer)
+{
+  return buffer->flags;
+}
+
+/**
+ * hb_buffer_set_cluster_level:
+ * @buffer: an #hb_buffer_t.
+ * @cluster_level:
+ *
+ *
+ *
+ * Since: 0.9.42
+ **/
+void
+hb_buffer_set_cluster_level (hb_buffer_t               *buffer,
+			     hb_buffer_cluster_level_t  cluster_level)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->cluster_level = cluster_level;
+}
+
+/**
+ * hb_buffer_get_cluster_level:
+ * @buffer: an #hb_buffer_t.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.42
+ **/
+hb_buffer_cluster_level_t
+hb_buffer_get_cluster_level (hb_buffer_t *buffer)
+{
+  return buffer->cluster_level;
+}
+
+
+/**
+ * hb_buffer_set_replacement_codepoint:
+ * @buffer: an #hb_buffer_t.
+ * @replacement: the replacement #hb_codepoint_t
+ *
+ * Sets the #hb_codepoint_t that replaces invalid entries for a given encoding
+ * when adding text to @buffer.
+ *
+ * Default is %HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT.
+ *
+ * Since: 0.9.31
+ **/
+void
+hb_buffer_set_replacement_codepoint (hb_buffer_t    *buffer,
+				     hb_codepoint_t  replacement)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->replacement = replacement;
+}
+
+/**
+ * hb_buffer_get_replacement_codepoint:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_replacement_codepoint().
+ *
+ * Return value:
+ * The @buffer replacement #hb_codepoint_t.
+ *
+ * Since: 0.9.31
+ **/
+hb_codepoint_t
+hb_buffer_get_replacement_codepoint (hb_buffer_t    *buffer)
+{
+  return buffer->replacement;
+}
+
+
+/**
+ * hb_buffer_set_invisible_glyph:
+ * @buffer: an #hb_buffer_t.
+ * @invisible: the invisible #hb_codepoint_t
+ *
+ * Sets the #hb_codepoint_t that replaces invisible characters in
+ * the shaping result.  If set to zero (default), the glyph for the
+ * U+0020 SPACE character is used.  Otherwise, this value is used
+ * verbatim.
+ *
+ * Since: 2.0.0
+ **/
+void
+hb_buffer_set_invisible_glyph (hb_buffer_t    *buffer,
+			       hb_codepoint_t  invisible)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  buffer->invisible = invisible;
+}
+
+/**
+ * hb_buffer_get_invisible_glyph:
+ * @buffer: an #hb_buffer_t.
+ *
+ * See hb_buffer_set_invisible_glyph().
+ *
+ * Return value:
+ * The @buffer invisible #hb_codepoint_t.
+ *
+ * Since: 2.0.0
+ **/
+hb_codepoint_t
+hb_buffer_get_invisible_glyph (hb_buffer_t    *buffer)
+{
+  return buffer->invisible;
+}
+
+
+/**
+ * hb_buffer_reset:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Resets the buffer to its initial status, as if it was just newly created
+ * with hb_buffer_create().
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_reset (hb_buffer_t *buffer)
+{
+  buffer->reset ();
+}
+
+/**
+ * hb_buffer_clear_contents:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Similar to hb_buffer_reset(), but does not clear the Unicode functions and
+ * the replacement code point.
+ *
+ * Since: 0.9.11
+ **/
+void
+hb_buffer_clear_contents (hb_buffer_t *buffer)
+{
+  buffer->clear ();
+}
+
+/**
+ * hb_buffer_pre_allocate:
+ * @buffer: an #hb_buffer_t.
+ * @size: number of items to pre allocate.
+ *
+ * Pre allocates memory for @buffer to fit at least @size number of items.
+ *
+ * Return value:
+ * %true if @buffer memory allocation succeeded, %false otherwise.
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_buffer_pre_allocate (hb_buffer_t *buffer, unsigned int size)
+{
+  return buffer->ensure (size);
+}
+
+/**
+ * hb_buffer_allocation_successful:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Check if allocating memory for the buffer succeeded.
+ *
+ * Return value:
+ * %true if @buffer memory allocation succeeded, %false otherwise.
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_buffer_allocation_successful (hb_buffer_t  *buffer)
+{
+  return buffer->successful;
+}
+
+/**
+ * hb_buffer_add:
+ * @buffer: an #hb_buffer_t.
+ * @codepoint: a Unicode code point.
+ * @cluster: the cluster value of @codepoint.
+ *
+ * Appends a character with the Unicode value of @codepoint to @buffer, and
+ * gives it the initial cluster value of @cluster. Clusters can be any thing
+ * the client wants, they are usually used to refer to the index of the
+ * character in the input text stream and are output in
+ * #hb_glyph_info_t.cluster field.
+ *
+ * This function does not check the validity of @codepoint, it is up to the
+ * caller to ensure it is a valid Unicode code point.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_buffer_add (hb_buffer_t    *buffer,
+	       hb_codepoint_t  codepoint,
+	       unsigned int    cluster)
+{
+  buffer->add (codepoint, cluster);
+  buffer->clear_context (1);
+}
+
+/**
+ * hb_buffer_set_length:
+ * @buffer: an #hb_buffer_t.
+ * @length: the new length of @buffer.
+ *
+ * Similar to hb_buffer_pre_allocate(), but clears any new items added at the
+ * end.
+ *
+ * Return value:
+ * %true if @buffer memory allocation succeeded, %false otherwise.
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_buffer_set_length (hb_buffer_t  *buffer,
+		      unsigned int  length)
+{
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return length == 0;
+
+  if (!buffer->ensure (length))
+    return false;
+
+  /* Wipe the new space */
+  if (length > buffer->len) {
+    memset (buffer->info + buffer->len, 0, sizeof (buffer->info[0]) * (length - buffer->len));
+    if (buffer->have_positions)
+      memset (buffer->pos + buffer->len, 0, sizeof (buffer->pos[0]) * (length - buffer->len));
+  }
+
+  buffer->len = length;
+
+  if (!length)
+  {
+    buffer->content_type = HB_BUFFER_CONTENT_TYPE_INVALID;
+    buffer->clear_context (0);
+  }
+  buffer->clear_context (1);
+
+  return true;
+}
+
+/**
+ * hb_buffer_get_length:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Returns the number of items in the buffer.
+ *
+ * Return value:
+ * The @buffer length.
+ * The value valid as long as buffer has not been modified.
+ *
+ * Since: 0.9.2
+ **/
+unsigned int
+hb_buffer_get_length (hb_buffer_t *buffer)
+{
+  return buffer->len;
+}
+
+/**
+ * hb_buffer_get_glyph_infos:
+ * @buffer: an #hb_buffer_t.
+ * @length: (out): output array length.
+ *
+ * Returns @buffer glyph information array.  Returned pointer
+ * is valid as long as @buffer contents are not modified.
+ *
+ * Return value: (transfer none) (array length=length):
+ * The @buffer glyph information array.
+ * The value valid as long as buffer has not been modified.
+ *
+ * Since: 0.9.2
+ **/
+hb_glyph_info_t *
+hb_buffer_get_glyph_infos (hb_buffer_t  *buffer,
+			   unsigned int *length)
+{
+  if (length)
+    *length = buffer->len;
+
+  return (hb_glyph_info_t *) buffer->info;
+}
+
+/**
+ * hb_buffer_get_glyph_positions:
+ * @buffer: an #hb_buffer_t.
+ * @length: (out): output length.
+ *
+ * Returns @buffer glyph position array.  Returned pointer
+ * is valid as long as @buffer contents are not modified.
+ *
+ * Return value: (transfer none) (array length=length):
+ * The @buffer glyph position array.
+ * The value valid as long as buffer has not been modified.
+ *
+ * Since: 0.9.2
+ **/
+hb_glyph_position_t *
+hb_buffer_get_glyph_positions (hb_buffer_t  *buffer,
+			       unsigned int *length)
+{
+  if (!buffer->have_positions)
+    buffer->clear_positions ();
+
+  if (length)
+    *length = buffer->len;
+
+  return (hb_glyph_position_t *) buffer->pos;
+}
+
+/**
+ * hb_glyph_info_get_glyph_flags:
+ * @info: a #hb_glyph_info_t.
+ *
+ * Returns glyph flags encoded within a #hb_glyph_info_t.
+ *
+ * Return value:
+ * The #hb_glyph_flags_t encoded within @info.
+ *
+ * Since: 1.5.0
+ **/
+hb_glyph_flags_t
+(hb_glyph_info_get_glyph_flags) (const hb_glyph_info_t *info)
+{
+  return hb_glyph_info_get_glyph_flags (info);
+}
+
+/**
+ * hb_buffer_reverse:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Reverses buffer contents.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_reverse (hb_buffer_t *buffer)
+{
+  buffer->reverse ();
+}
+
+/**
+ * hb_buffer_reverse_range:
+ * @buffer: an #hb_buffer_t.
+ * @start: start index.
+ * @end: end index.
+ *
+ * Reverses buffer contents between start to end.
+ *
+ * Since: 0.9.41
+ **/
+void
+hb_buffer_reverse_range (hb_buffer_t *buffer,
+			 unsigned int start, unsigned int end)
+{
+  buffer->reverse_range (start, end);
+}
+
+/**
+ * hb_buffer_reverse_clusters:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Reverses buffer clusters.  That is, the buffer contents are
+ * reversed, then each cluster (consecutive items having the
+ * same cluster number) are reversed again.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_reverse_clusters (hb_buffer_t *buffer)
+{
+  buffer->reverse_clusters ();
+}
+
+/**
+ * hb_buffer_guess_segment_properties:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Sets unset buffer segment properties based on buffer Unicode
+ * contents.  If buffer is not empty, it must have content type
+ * %HB_BUFFER_CONTENT_TYPE_UNICODE.
+ *
+ * If buffer script is not set (ie. is %HB_SCRIPT_INVALID), it
+ * will be set to the Unicode script of the first character in
+ * the buffer that has a script other than %HB_SCRIPT_COMMON,
+ * %HB_SCRIPT_INHERITED, and %HB_SCRIPT_UNKNOWN.
+ *
+ * Next, if buffer direction is not set (ie. is %HB_DIRECTION_INVALID),
+ * it will be set to the natural horizontal direction of the
+ * buffer script as returned by hb_script_get_horizontal_direction().
+ * If hb_script_get_horizontal_direction() returns %HB_DIRECTION_INVALID,
+ * then %HB_DIRECTION_LTR is used.
+ *
+ * Finally, if buffer language is not set (ie. is %HB_LANGUAGE_INVALID),
+ * it will be set to the process's default language as returned by
+ * hb_language_get_default().  This may change in the future by
+ * taking buffer script into consideration when choosing a language.
+ * Note that hb_language_get_default() is NOT threadsafe the first time
+ * it is called.  See documentation for that function for details.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_buffer_guess_segment_properties (hb_buffer_t *buffer)
+{
+  buffer->guess_segment_properties ();
+}
+
+template <typename utf_t>
+static inline void
+hb_buffer_add_utf (hb_buffer_t  *buffer,
+		   const typename utf_t::codepoint_t *text,
+		   int           text_length,
+		   unsigned int  item_offset,
+		   int           item_length)
+{
+  typedef typename utf_t::codepoint_t T;
+  const hb_codepoint_t replacement = buffer->replacement;
+
+  assert ((buffer->content_type == HB_BUFFER_CONTENT_TYPE_UNICODE) ||
+	  (!buffer->len && (buffer->content_type == HB_BUFFER_CONTENT_TYPE_INVALID)));
+
+  if (unlikely (hb_object_is_immutable (buffer)))
+    return;
+
+  if (text_length == -1)
+    text_length = utf_t::strlen (text);
+
+  if (item_length == -1)
+    item_length = text_length - item_offset;
+
+  buffer->ensure (buffer->len + item_length * sizeof (T) / 4);
+
+  /* If buffer is empty and pre-context provided, install it.
+   * This check is written this way, to make sure people can
+   * provide pre-context in one add_utf() call, then provide
+   * text in a follow-up call.  See:
+   *
+   * https://bugzilla.mozilla.org/show_bug.cgi?id=801410#c13
+   */
+  if (!buffer->len && item_offset > 0)
+  {
+    /* Add pre-context */
+    buffer->clear_context (0);
+    const T *prev = text + item_offset;
+    const T *start = text;
+    while (start < prev && buffer->context_len[0] < buffer->CONTEXT_LENGTH)
+    {
+      hb_codepoint_t u;
+      prev = utf_t::prev (prev, start, &u, replacement);
+      buffer->context[0][buffer->context_len[0]++] = u;
+    }
+  }
+
+  const T *next = text + item_offset;
+  const T *end = next + item_length;
+  while (next < end)
+  {
+    hb_codepoint_t u;
+    const T *old_next = next;
+    next = utf_t::next (next, end, &u, replacement);
+    buffer->add (u, old_next - (const T *) text);
+  }
+
+  /* Add post-context */
+  buffer->clear_context (1);
+  end = text + text_length;
+  while (next < end && buffer->context_len[1] < buffer->CONTEXT_LENGTH)
+  {
+    hb_codepoint_t u;
+    next = utf_t::next (next, end, &u, replacement);
+    buffer->context[1][buffer->context_len[1]++] = u;
+  }
+
+  buffer->content_type = HB_BUFFER_CONTENT_TYPE_UNICODE;
+}
+
+/**
+ * hb_buffer_add_utf8:
+ * @buffer: an #hb_buffer_t.
+ * @text: (array length=text_length) (element-type uint8_t): an array of UTF-8
+ *               characters to append.
+ * @text_length: the length of the @text, or -1 if it is %NULL terminated.
+ * @item_offset: the offset of the first character to add to the @buffer.
+ * @item_length: the number of characters to add to the @buffer, or -1 for the
+ *               end of @text (assuming it is %NULL terminated).
+ *
+ * See hb_buffer_add_codepoints().
+ *
+ * Replaces invalid UTF-8 characters with the @buffer replacement code point,
+ * see hb_buffer_set_replacement_codepoint().
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_add_utf8 (hb_buffer_t  *buffer,
+		    const char   *text,
+		    int           text_length,
+		    unsigned int  item_offset,
+		    int           item_length)
+{
+  hb_buffer_add_utf<hb_utf8_t> (buffer, (const uint8_t *) text, text_length, item_offset, item_length);
+}
+
+/**
+ * hb_buffer_add_utf16:
+ * @buffer: an #hb_buffer_t.
+ * @text: (array length=text_length): an array of UTF-16 characters to append.
+ * @text_length: the length of the @text, or -1 if it is %NULL terminated.
+ * @item_offset: the offset of the first character to add to the @buffer.
+ * @item_length: the number of characters to add to the @buffer, or -1 for the
+ *               end of @text (assuming it is %NULL terminated).
+ *
+ * See hb_buffer_add_codepoints().
+ *
+ * Replaces invalid UTF-16 characters with the @buffer replacement code point,
+ * see hb_buffer_set_replacement_codepoint().
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_add_utf16 (hb_buffer_t    *buffer,
+		     const uint16_t *text,
+		     int             text_length,
+		     unsigned int    item_offset,
+		     int             item_length)
+{
+  hb_buffer_add_utf<hb_utf16_t> (buffer, text, text_length, item_offset, item_length);
+}
+
+/**
+ * hb_buffer_add_utf32:
+ * @buffer: an #hb_buffer_t.
+ * @text: (array length=text_length): an array of UTF-32 characters to append.
+ * @text_length: the length of the @text, or -1 if it is %NULL terminated.
+ * @item_offset: the offset of the first character to add to the @buffer.
+ * @item_length: the number of characters to add to the @buffer, or -1 for the
+ *               end of @text (assuming it is %NULL terminated).
+ *
+ * See hb_buffer_add_codepoints().
+ *
+ * Replaces invalid UTF-32 characters with the @buffer replacement code point,
+ * see hb_buffer_set_replacement_codepoint().
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_add_utf32 (hb_buffer_t    *buffer,
+		     const uint32_t *text,
+		     int             text_length,
+		     unsigned int    item_offset,
+		     int             item_length)
+{
+  hb_buffer_add_utf<hb_utf32_t> (buffer, text, text_length, item_offset, item_length);
+}
+
+/**
+ * hb_buffer_add_latin1:
+ * @buffer: an #hb_buffer_t.
+ * @text: (array length=text_length) (element-type uint8_t): an array of UTF-8
+ *               characters to append.
+ * @text_length: the length of the @text, or -1 if it is %NULL terminated.
+ * @item_offset: the offset of the first character to add to the @buffer.
+ * @item_length: the number of characters to add to the @buffer, or -1 for the
+ *               end of @text (assuming it is %NULL terminated).
+ *
+ * Similar to hb_buffer_add_codepoints(), but allows only access to first 256
+ * Unicode code points that can fit in 8-bit strings.
+ *
+ * <note>Has nothing to do with non-Unicode Latin-1 encoding.</note>
+ *
+ * Since: 0.9.39
+ **/
+void
+hb_buffer_add_latin1 (hb_buffer_t   *buffer,
+		      const uint8_t *text,
+		      int            text_length,
+		      unsigned int   item_offset,
+		      int            item_length)
+{
+  hb_buffer_add_utf<hb_latin1_t> (buffer, text, text_length, item_offset, item_length);
+}
+
+/**
+ * hb_buffer_add_codepoints:
+ * @buffer: a #hb_buffer_t to append characters to.
+ * @text: (array length=text_length): an array of Unicode code points to append.
+ * @text_length: the length of the @text, or -1 if it is %NULL terminated.
+ * @item_offset: the offset of the first code point to add to the @buffer.
+ * @item_length: the number of code points to add to the @buffer, or -1 for the
+ *               end of @text (assuming it is %NULL terminated).
+ *
+ * Appends characters from @text array to @buffer. The @item_offset is the
+ * position of the first character from @text that will be appended, and
+ * @item_length is the number of character. When shaping part of a larger text
+ * (e.g. a run of text from a paragraph), instead of passing just the substring
+ * corresponding to the run, it is preferable to pass the whole
+ * paragraph and specify the run start and length as @item_offset and
+ * @item_length, respectively, to give HarfBuzz the full context to be able,
+ * for example, to do cross-run Arabic shaping or properly handle combining
+ * marks at stat of run.
+ *
+ * This function does not check the validity of @text, it is up to the caller
+ * to ensure it contains a valid Unicode code points.
+ *
+ * Since: 0.9.31
+ **/
+void
+hb_buffer_add_codepoints (hb_buffer_t          *buffer,
+			  const hb_codepoint_t *text,
+			  int                   text_length,
+			  unsigned int          item_offset,
+			  int                   item_length)
+{
+  hb_buffer_add_utf<hb_utf32_novalidate_t> (buffer, text, text_length, item_offset, item_length);
+}
+
+
+/**
+ * hb_buffer_append:
+ * @buffer: an #hb_buffer_t.
+ * @source: source #hb_buffer_t.
+ * @start: start index into source buffer to copy.  Use 0 to copy from start of buffer.
+ * @end: end index into source buffer to copy.  Use @HB_FEATURE_GLOBAL_END to copy to end of buffer.
+ *
+ * Append (part of) contents of another buffer to this buffer.
+ *
+ * Since: 1.5.0
+ **/
+HB_EXTERN void
+hb_buffer_append (hb_buffer_t *buffer,
+		  hb_buffer_t *source,
+		  unsigned int start,
+		  unsigned int end)
+{
+  assert (!buffer->have_output && !source->have_output);
+  assert (buffer->have_positions == source->have_positions ||
+	  !buffer->len || !source->len);
+  assert (buffer->content_type == source->content_type ||
+	  !buffer->len || !source->len);
+
+  if (end > source->len)
+    end = source->len;
+  if (start > end)
+    start = end;
+  if (start == end)
+    return;
+
+  if (!buffer->len)
+    buffer->content_type = source->content_type;
+  if (!buffer->have_positions && source->have_positions)
+    buffer->clear_positions ();
+
+  if (buffer->len + (end - start) < buffer->len) /* Overflows. */
+  {
+    buffer->successful = false;
+    return;
+  }
+
+  unsigned int orig_len = buffer->len;
+  hb_buffer_set_length (buffer, buffer->len + (end - start));
+  if (unlikely (!buffer->successful))
+    return;
+
+  memcpy (buffer->info + orig_len, source->info + start, (end - start) * sizeof (buffer->info[0]));
+  if (buffer->have_positions)
+    memcpy (buffer->pos + orig_len, source->pos + start, (end - start) * sizeof (buffer->pos[0]));
+}
+
+
+static int
+compare_info_codepoint (const hb_glyph_info_t *pa,
+			const hb_glyph_info_t *pb)
+{
+  return (int) pb->codepoint - (int) pa->codepoint;
+}
+
+static inline void
+normalize_glyphs_cluster (hb_buffer_t *buffer,
+			  unsigned int start,
+			  unsigned int end,
+			  bool backward)
+{
+  hb_glyph_position_t *pos = buffer->pos;
+
+  /* Total cluster advance */
+  hb_position_t total_x_advance = 0, total_y_advance = 0;
+  for (unsigned int i = start; i < end; i++)
+  {
+    total_x_advance += pos[i].x_advance;
+    total_y_advance += pos[i].y_advance;
+  }
+
+  hb_position_t x_advance = 0, y_advance = 0;
+  for (unsigned int i = start; i < end; i++)
+  {
+    pos[i].x_offset += x_advance;
+    pos[i].y_offset += y_advance;
+
+    x_advance += pos[i].x_advance;
+    y_advance += pos[i].y_advance;
+
+    pos[i].x_advance = 0;
+    pos[i].y_advance = 0;
+  }
+
+  if (backward)
+  {
+    /* Transfer all cluster advance to the last glyph. */
+    pos[end - 1].x_advance = total_x_advance;
+    pos[end - 1].y_advance = total_y_advance;
+
+    hb_stable_sort (buffer->info + start, end - start - 1, compare_info_codepoint, buffer->pos + start);
+  } else {
+    /* Transfer all cluster advance to the first glyph. */
+    pos[start].x_advance += total_x_advance;
+    pos[start].y_advance += total_y_advance;
+    for (unsigned int i = start + 1; i < end; i++) {
+      pos[i].x_offset -= total_x_advance;
+      pos[i].y_offset -= total_y_advance;
+    }
+    hb_stable_sort (buffer->info + start + 1, end - start - 1, compare_info_codepoint, buffer->pos + start + 1);
+  }
+}
+
+/**
+ * hb_buffer_normalize_glyphs:
+ * @buffer: an #hb_buffer_t.
+ *
+ * Reorders a glyph buffer to have canonical in-cluster glyph order / position.
+ * The resulting clusters should behave identical to pre-reordering clusters.
+ *
+ * <note>This has nothing to do with Unicode normalization.</note>
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_buffer_normalize_glyphs (hb_buffer_t *buffer)
+{
+  assert (buffer->have_positions);
+  assert ((buffer->content_type == HB_BUFFER_CONTENT_TYPE_GLYPHS) ||
+	  (!buffer->len && (buffer->content_type == HB_BUFFER_CONTENT_TYPE_INVALID)));
+
+  bool backward = HB_DIRECTION_IS_BACKWARD (buffer->props.direction);
+
+  foreach_cluster (buffer, start, end)
+    normalize_glyphs_cluster (buffer, start, end, backward);
+}
+
+void
+hb_buffer_t::sort (unsigned int start, unsigned int end, int(*compar)(const hb_glyph_info_t *, const hb_glyph_info_t *))
+{
+  assert (!have_positions);
+  for (unsigned int i = start + 1; i < end; i++)
+  {
+    unsigned int j = i;
+    while (j > start && compar (&info[j - 1], &info[i]) > 0)
+      j--;
+    if (i == j)
+      continue;
+    /* Move item i to occupy place for item j, shift what's in between. */
+    merge_clusters (j, i + 1);
+    {
+      hb_glyph_info_t t = info[i];
+      memmove (&info[j + 1], &info[j], (i - j) * sizeof (hb_glyph_info_t));
+      info[j] = t;
+    }
+  }
+}
+
+
+/*
+ * Comparing buffers.
+ */
+
+/**
+ * hb_buffer_diff:
+ * @buffer: a buffer.
+ * @reference: other buffer to compare to.
+ * @dottedcircle_glyph: glyph id of U+25CC DOTTED CIRCLE, or (hb_codepont_t) -1.
+ * @position_fuzz: allowed absolute difference in position values.
+ *
+ * If dottedcircle_glyph is (hb_codepoint_t) -1 then %HB_BUFFER_DIFF_FLAG_DOTTED_CIRCLE_PRESENT
+ * and %HB_BUFFER_DIFF_FLAG_NOTDEF_PRESENT are never returned.  This should be used by most
+ * callers if just comparing two buffers is needed.
+ *
+ * Since: 1.5.0
+ **/
+hb_buffer_diff_flags_t
+hb_buffer_diff (hb_buffer_t *buffer,
+		hb_buffer_t *reference,
+		hb_codepoint_t dottedcircle_glyph,
+		unsigned int position_fuzz)
+{
+  if (buffer->content_type != reference->content_type && buffer->len && reference->len)
+    return HB_BUFFER_DIFF_FLAG_CONTENT_TYPE_MISMATCH;
+
+  hb_buffer_diff_flags_t result = HB_BUFFER_DIFF_FLAG_EQUAL;
+  bool contains = dottedcircle_glyph != (hb_codepoint_t) -1;
+
+  unsigned int count = reference->len;
+
+  if (buffer->len != count)
+  {
+    /*
+     * we can't compare glyph-by-glyph, but we do want to know if there
+     * are .notdef or dottedcircle glyphs present in the reference buffer
+     */
+    const hb_glyph_info_t *info = reference->info;
+    unsigned int i;
+    for (i = 0; i < count; i++)
+    {
+      if (contains && info[i].codepoint == dottedcircle_glyph)
+	result |= HB_BUFFER_DIFF_FLAG_DOTTED_CIRCLE_PRESENT;
+      if (contains && info[i].codepoint == 0)
+	result |= HB_BUFFER_DIFF_FLAG_NOTDEF_PRESENT;
+    }
+    result |= HB_BUFFER_DIFF_FLAG_LENGTH_MISMATCH;
+    return hb_buffer_diff_flags_t (result);
+  }
+
+  if (!count)
+    return hb_buffer_diff_flags_t (result);
+
+  const hb_glyph_info_t *buf_info = buffer->info;
+  const hb_glyph_info_t *ref_info = reference->info;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    if (buf_info->codepoint != ref_info->codepoint)
+      result |= HB_BUFFER_DIFF_FLAG_CODEPOINT_MISMATCH;
+    if (buf_info->cluster != ref_info->cluster)
+      result |= HB_BUFFER_DIFF_FLAG_CLUSTER_MISMATCH;
+    if ((buf_info->mask & ~ref_info->mask & HB_GLYPH_FLAG_DEFINED))
+      result |= HB_BUFFER_DIFF_FLAG_GLYPH_FLAGS_MISMATCH;
+    if (contains && ref_info->codepoint == dottedcircle_glyph)
+      result |= HB_BUFFER_DIFF_FLAG_DOTTED_CIRCLE_PRESENT;
+    if (contains && ref_info->codepoint == 0)
+      result |= HB_BUFFER_DIFF_FLAG_NOTDEF_PRESENT;
+    buf_info++;
+    ref_info++;
+  }
+
+  if (buffer->content_type == HB_BUFFER_CONTENT_TYPE_GLYPHS)
+  {
+    assert (buffer->have_positions);
+    const hb_glyph_position_t *buf_pos = buffer->pos;
+    const hb_glyph_position_t *ref_pos = reference->pos;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if ((unsigned int) abs (buf_pos->x_advance - ref_pos->x_advance) > position_fuzz ||
+	  (unsigned int) abs (buf_pos->y_advance - ref_pos->y_advance) > position_fuzz ||
+	  (unsigned int) abs (buf_pos->x_offset - ref_pos->x_offset) > position_fuzz ||
+	  (unsigned int) abs (buf_pos->y_offset - ref_pos->y_offset) > position_fuzz)
+      {
+	result |= HB_BUFFER_DIFF_FLAG_POSITION_MISMATCH;
+	break;
+      }
+      buf_pos++;
+      ref_pos++;
+    }
+  }
+
+  return result;
+}
+
+
+/*
+ * Debugging.
+ */
+
+#ifndef HB_NO_BUFFER_MESSAGE
+/**
+ * hb_buffer_set_message_func:
+ * @buffer: an #hb_buffer_t.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.1.3
+ **/
+void
+hb_buffer_set_message_func (hb_buffer_t *buffer,
+			    hb_buffer_message_func_t func,
+			    void *user_data, hb_destroy_func_t destroy)
+{
+  if (buffer->message_destroy)
+    buffer->message_destroy (buffer->message_data);
+
+  if (func) {
+    buffer->message_func = func;
+    buffer->message_data = user_data;
+    buffer->message_destroy = destroy;
+  } else {
+    buffer->message_func = nullptr;
+    buffer->message_data = nullptr;
+    buffer->message_destroy = nullptr;
+  }
+}
+bool
+hb_buffer_t::message_impl (hb_font_t *font, const char *fmt, va_list ap)
+{
+  char buf[100];
+  vsnprintf (buf, sizeof (buf), fmt, ap);
+  return (bool) this->message_func (this, font, buf, this->message_data);
+}
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-buffer.h b/thirdparty/harfbuzz/src/hb-buffer.h
new file mode 100644
index 0000000000..2f581f3c73
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer.h
@@ -0,0 +1,586 @@
+/*
+ * Copyright © 1998-2004  David Turner and Werner Lemberg
+ * Copyright © 2004,2007,2009  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Owen Taylor, Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_BUFFER_H
+#define HB_BUFFER_H
+
+#include "hb-common.h"
+#include "hb-unicode.h"
+#include "hb-font.h"
+
+HB_BEGIN_DECLS
+
+/**
+ * hb_glyph_info_t:
+ * @codepoint: either a Unicode code point (before shaping) or a glyph index
+ *             (after shaping).
+ * @cluster: the index of the character in the original text that corresponds
+ *           to this #hb_glyph_info_t, or whatever the client passes to
+ *           hb_buffer_add(). More than one #hb_glyph_info_t can have the same
+ *           @cluster value, if they resulted from the same character (e.g. one
+ *           to many glyph substitution), and when more than one character gets
+ *           merged in the same glyph (e.g. many to one glyph substitution) the
+ *           #hb_glyph_info_t will have the smallest cluster value of them.
+ *           By default some characters are merged into the same cluster
+ *           (e.g. combining marks have the same cluster as their bases)
+ *           even if they are separate glyphs, hb_buffer_set_cluster_level()
+ *           allow selecting more fine-grained cluster handling.
+ *
+ * The #hb_glyph_info_t is the structure that holds information about the
+ * glyphs and their relation to input text.
+ */
+typedef struct hb_glyph_info_t
+{
+  hb_codepoint_t codepoint;
+  /*< private >*/
+  hb_mask_t      mask;
+  /*< public >*/
+  uint32_t       cluster;
+
+  /*< private >*/
+  hb_var_int_t   var1;
+  hb_var_int_t   var2;
+} hb_glyph_info_t;
+
+/**
+ * hb_glyph_flags_t:
+ * @HB_GLYPH_FLAG_UNSAFE_TO_BREAK: Indicates that if input text is broken at the
+ * 				   beginning of the cluster this glyph is part of,
+ * 				   then both sides need to be re-shaped, as the
+ * 				   result might be different.  On the flip side,
+ * 				   it means that when this flag is not present,
+ * 				   then it's safe to break the glyph-run at the
+ * 				   beginning of this cluster, and the two sides
+ * 				   represent the exact same result one would get
+ * 				   if breaking input text at the beginning of
+ * 				   this cluster and shaping the two sides
+ * 				   separately.  This can be used to optimize
+ * 				   paragraph layout, by avoiding re-shaping
+ * 				   of each line after line-breaking, or limiting
+ * 				   the reshaping to a small piece around the
+ * 				   breaking point only.
+ * @HB_GLYPH_FLAG_DEFINED: All the currently defined flags.
+ *
+ * Since: 1.5.0
+ */
+typedef enum { /*< flags >*/
+  HB_GLYPH_FLAG_UNSAFE_TO_BREAK		= 0x00000001,
+
+  HB_GLYPH_FLAG_DEFINED			= 0x00000001 /* OR of all defined flags */
+} hb_glyph_flags_t;
+
+HB_EXTERN hb_glyph_flags_t
+hb_glyph_info_get_glyph_flags (const hb_glyph_info_t *info);
+
+#define hb_glyph_info_get_glyph_flags(info) \
+	((hb_glyph_flags_t) ((unsigned int) (info)->mask & HB_GLYPH_FLAG_DEFINED))
+
+
+/**
+ * hb_glyph_position_t:
+ * @x_advance: how much the line advances after drawing this glyph when setting
+ *             text in horizontal direction.
+ * @y_advance: how much the line advances after drawing this glyph when setting
+ *             text in vertical direction.
+ * @x_offset: how much the glyph moves on the X-axis before drawing it, this
+ *            should not affect how much the line advances.
+ * @y_offset: how much the glyph moves on the Y-axis before drawing it, this
+ *            should not affect how much the line advances.
+ *
+ * The #hb_glyph_position_t is the structure that holds the positions of the
+ * glyph in both horizontal and vertical directions. All positions in
+ * #hb_glyph_position_t are relative to the current point.
+ *
+ */
+typedef struct hb_glyph_position_t {
+  hb_position_t  x_advance;
+  hb_position_t  y_advance;
+  hb_position_t  x_offset;
+  hb_position_t  y_offset;
+
+  /*< private >*/
+  hb_var_int_t   var;
+} hb_glyph_position_t;
+
+/**
+ * hb_segment_properties_t:
+ * @direction: the #hb_direction_t of the buffer, see hb_buffer_set_direction().
+ * @script: the #hb_script_t of the buffer, see hb_buffer_set_script().
+ * @language: the #hb_language_t of the buffer, see hb_buffer_set_language().
+ *
+ * The structure that holds various text properties of an #hb_buffer_t. Can be
+ * set and retrieved using hb_buffer_set_segment_properties() and
+ * hb_buffer_get_segment_properties(), respectively.
+ */
+typedef struct hb_segment_properties_t {
+  hb_direction_t  direction;
+  hb_script_t     script;
+  hb_language_t   language;
+  /*< private >*/
+  void           *reserved1;
+  void           *reserved2;
+} hb_segment_properties_t;
+
+#define HB_SEGMENT_PROPERTIES_DEFAULT {HB_DIRECTION_INVALID, \
+				       HB_SCRIPT_INVALID, \
+				       HB_LANGUAGE_INVALID, \
+				       (void *) 0, \
+				       (void *) 0}
+
+HB_EXTERN hb_bool_t
+hb_segment_properties_equal (const hb_segment_properties_t *a,
+			     const hb_segment_properties_t *b);
+
+HB_EXTERN unsigned int
+hb_segment_properties_hash (const hb_segment_properties_t *p);
+
+
+
+/**
+ * hb_buffer_t:
+ *
+ * The main structure holding the input text and its properties before shaping,
+ * and output glyphs and their information after shaping.
+ */
+
+typedef struct hb_buffer_t hb_buffer_t;
+
+HB_EXTERN hb_buffer_t *
+hb_buffer_create (void);
+
+HB_EXTERN hb_buffer_t *
+hb_buffer_get_empty (void);
+
+HB_EXTERN hb_buffer_t *
+hb_buffer_reference (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_destroy (hb_buffer_t *buffer);
+
+HB_EXTERN hb_bool_t
+hb_buffer_set_user_data (hb_buffer_t        *buffer,
+			 hb_user_data_key_t *key,
+			 void *              data,
+			 hb_destroy_func_t   destroy,
+			 hb_bool_t           replace);
+
+HB_EXTERN void *
+hb_buffer_get_user_data (hb_buffer_t        *buffer,
+			 hb_user_data_key_t *key);
+
+
+/**
+ * hb_buffer_content_type_t:
+ * @HB_BUFFER_CONTENT_TYPE_INVALID: Initial value for new buffer.
+ * @HB_BUFFER_CONTENT_TYPE_UNICODE: The buffer contains input characters (before shaping).
+ * @HB_BUFFER_CONTENT_TYPE_GLYPHS: The buffer contains output glyphs (after shaping).
+ */
+typedef enum {
+  HB_BUFFER_CONTENT_TYPE_INVALID = 0,
+  HB_BUFFER_CONTENT_TYPE_UNICODE,
+  HB_BUFFER_CONTENT_TYPE_GLYPHS
+} hb_buffer_content_type_t;
+
+HB_EXTERN void
+hb_buffer_set_content_type (hb_buffer_t              *buffer,
+			    hb_buffer_content_type_t  content_type);
+
+HB_EXTERN hb_buffer_content_type_t
+hb_buffer_get_content_type (hb_buffer_t *buffer);
+
+
+HB_EXTERN void
+hb_buffer_set_unicode_funcs (hb_buffer_t        *buffer,
+			     hb_unicode_funcs_t *unicode_funcs);
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_buffer_get_unicode_funcs (hb_buffer_t        *buffer);
+
+HB_EXTERN void
+hb_buffer_set_direction (hb_buffer_t    *buffer,
+			 hb_direction_t  direction);
+
+HB_EXTERN hb_direction_t
+hb_buffer_get_direction (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_set_script (hb_buffer_t *buffer,
+		      hb_script_t  script);
+
+HB_EXTERN hb_script_t
+hb_buffer_get_script (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_set_language (hb_buffer_t   *buffer,
+			hb_language_t  language);
+
+
+HB_EXTERN hb_language_t
+hb_buffer_get_language (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_set_segment_properties (hb_buffer_t *buffer,
+				  const hb_segment_properties_t *props);
+
+HB_EXTERN void
+hb_buffer_get_segment_properties (hb_buffer_t *buffer,
+				  hb_segment_properties_t *props);
+
+HB_EXTERN void
+hb_buffer_guess_segment_properties (hb_buffer_t *buffer);
+
+
+/**
+ * hb_buffer_flags_t:
+ * @HB_BUFFER_FLAG_DEFAULT: the default buffer flag.
+ * @HB_BUFFER_FLAG_BOT: flag indicating that special handling of the beginning
+ *                      of text paragraph can be applied to this buffer. Should usually
+ *                      be set, unless you are passing to the buffer only part
+ *                      of the text without the full context.
+ * @HB_BUFFER_FLAG_EOT: flag indicating that special handling of the end of text
+ *                      paragraph can be applied to this buffer, similar to
+ *                      @HB_BUFFER_FLAG_BOT.
+ * @HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES:
+ *                      flag indication that character with Default_Ignorable
+ *                      Unicode property should use the corresponding glyph
+ *                      from the font, instead of hiding them (done by
+ *                      replacing them with the space glyph and zeroing the
+ *                      advance width.)  This flag takes precedence over
+ *                      @HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES.
+ * @HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES:
+ *                      flag indication that character with Default_Ignorable
+ *                      Unicode property should be removed from glyph string
+ *                      instead of hiding them (done by replacing them with the
+ *                      space glyph and zeroing the advance width.)
+ *                      @HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES takes
+ *                      precedence over this flag. Since: 1.8.0
+ * @HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE:
+ *                      flag indicating that a dotted circle should
+ *                      not be inserted in the rendering of incorrect
+ *                      character sequences (such at <0905 093E>). Since: 2.4
+ *
+ * Since: 0.9.20
+ */
+typedef enum { /*< flags >*/
+  HB_BUFFER_FLAG_DEFAULT			= 0x00000000u,
+  HB_BUFFER_FLAG_BOT				= 0x00000001u, /* Beginning-of-text */
+  HB_BUFFER_FLAG_EOT				= 0x00000002u, /* End-of-text */
+  HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES	= 0x00000004u,
+  HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES	= 0x00000008u,
+  HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE	= 0x00000010u
+} hb_buffer_flags_t;
+
+HB_EXTERN void
+hb_buffer_set_flags (hb_buffer_t       *buffer,
+		     hb_buffer_flags_t  flags);
+
+HB_EXTERN hb_buffer_flags_t
+hb_buffer_get_flags (hb_buffer_t *buffer);
+
+/**
+ * hb_buffer_cluster_level_t:
+ * @HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES: Return cluster values grouped by graphemes into
+ *   monotone order.
+ * @HB_BUFFER_CLUSTER_LEVEL_MONOTONE_CHARACTERS: Return cluster values grouped into monotone order.
+ * @HB_BUFFER_CLUSTER_LEVEL_CHARACTERS: Don't group cluster values.
+ * @HB_BUFFER_CLUSTER_LEVEL_DEFAULT: Default cluster level,
+ *   equal to @HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES.
+ *
+ * Since: 0.9.42
+ */
+typedef enum {
+  HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES	= 0,
+  HB_BUFFER_CLUSTER_LEVEL_MONOTONE_CHARACTERS	= 1,
+  HB_BUFFER_CLUSTER_LEVEL_CHARACTERS		= 2,
+  HB_BUFFER_CLUSTER_LEVEL_DEFAULT = HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES
+} hb_buffer_cluster_level_t;
+
+HB_EXTERN void
+hb_buffer_set_cluster_level (hb_buffer_t               *buffer,
+			     hb_buffer_cluster_level_t  cluster_level);
+
+HB_EXTERN hb_buffer_cluster_level_t
+hb_buffer_get_cluster_level (hb_buffer_t *buffer);
+
+/**
+ * HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT:
+ *
+ * The default code point for replacing invalid characters in a given encoding.
+ * Set to U+FFFD REPLACEMENT CHARACTER.
+ *
+ * Since: 0.9.31
+ */
+#define HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT 0xFFFDu
+
+HB_EXTERN void
+hb_buffer_set_replacement_codepoint (hb_buffer_t    *buffer,
+				     hb_codepoint_t  replacement);
+
+HB_EXTERN hb_codepoint_t
+hb_buffer_get_replacement_codepoint (hb_buffer_t    *buffer);
+
+HB_EXTERN void
+hb_buffer_set_invisible_glyph (hb_buffer_t    *buffer,
+			       hb_codepoint_t  invisible);
+
+HB_EXTERN hb_codepoint_t
+hb_buffer_get_invisible_glyph (hb_buffer_t    *buffer);
+
+
+HB_EXTERN void
+hb_buffer_reset (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_clear_contents (hb_buffer_t *buffer);
+
+HB_EXTERN hb_bool_t
+hb_buffer_pre_allocate (hb_buffer_t  *buffer,
+			unsigned int  size);
+
+
+HB_EXTERN hb_bool_t
+hb_buffer_allocation_successful (hb_buffer_t  *buffer);
+
+HB_EXTERN void
+hb_buffer_reverse (hb_buffer_t *buffer);
+
+HB_EXTERN void
+hb_buffer_reverse_range (hb_buffer_t *buffer,
+			 unsigned int start, unsigned int end);
+
+HB_EXTERN void
+hb_buffer_reverse_clusters (hb_buffer_t *buffer);
+
+
+/* Filling the buffer in */
+
+HB_EXTERN void
+hb_buffer_add (hb_buffer_t    *buffer,
+	       hb_codepoint_t  codepoint,
+	       unsigned int    cluster);
+
+HB_EXTERN void
+hb_buffer_add_utf8 (hb_buffer_t  *buffer,
+		    const char   *text,
+		    int           text_length,
+		    unsigned int  item_offset,
+		    int           item_length);
+
+HB_EXTERN void
+hb_buffer_add_utf16 (hb_buffer_t    *buffer,
+		     const uint16_t *text,
+		     int             text_length,
+		     unsigned int    item_offset,
+		     int             item_length);
+
+HB_EXTERN void
+hb_buffer_add_utf32 (hb_buffer_t    *buffer,
+		     const uint32_t *text,
+		     int             text_length,
+		     unsigned int    item_offset,
+		     int             item_length);
+
+HB_EXTERN void
+hb_buffer_add_latin1 (hb_buffer_t   *buffer,
+		      const uint8_t *text,
+		      int            text_length,
+		      unsigned int   item_offset,
+		      int            item_length);
+
+HB_EXTERN void
+hb_buffer_add_codepoints (hb_buffer_t          *buffer,
+			  const hb_codepoint_t *text,
+			  int                   text_length,
+			  unsigned int          item_offset,
+			  int                   item_length);
+
+HB_EXTERN void
+hb_buffer_append (hb_buffer_t *buffer,
+		  hb_buffer_t *source,
+		  unsigned int start,
+		  unsigned int end);
+
+HB_EXTERN hb_bool_t
+hb_buffer_set_length (hb_buffer_t  *buffer,
+		      unsigned int  length);
+
+HB_EXTERN unsigned int
+hb_buffer_get_length (hb_buffer_t *buffer);
+
+/* Getting glyphs out of the buffer */
+
+HB_EXTERN hb_glyph_info_t *
+hb_buffer_get_glyph_infos (hb_buffer_t  *buffer,
+			   unsigned int *length);
+
+HB_EXTERN hb_glyph_position_t *
+hb_buffer_get_glyph_positions (hb_buffer_t  *buffer,
+			       unsigned int *length);
+
+
+HB_EXTERN void
+hb_buffer_normalize_glyphs (hb_buffer_t *buffer);
+
+
+/*
+ * Serialize
+ */
+
+/**
+ * hb_buffer_serialize_flags_t:
+ * @HB_BUFFER_SERIALIZE_FLAG_DEFAULT: serialize glyph names, clusters and positions.
+ * @HB_BUFFER_SERIALIZE_FLAG_NO_CLUSTERS: do not serialize glyph cluster.
+ * @HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS: do not serialize glyph position information.
+ * @HB_BUFFER_SERIALIZE_FLAG_NO_GLYPH_NAMES: do no serialize glyph name.
+ * @HB_BUFFER_SERIALIZE_FLAG_GLYPH_EXTENTS: serialize glyph extents.
+ * @HB_BUFFER_SERIALIZE_FLAG_GLYPH_FLAGS: serialize glyph flags. Since: 1.5.0
+ * @HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES: do not serialize glyph advances,
+ *  glyph offsets will reflect absolute glyph positions. Since: 1.8.0
+ *
+ * Flags that control what glyph information are serialized in hb_buffer_serialize_glyphs().
+ *
+ * Since: 0.9.20
+ */
+typedef enum { /*< flags >*/
+  HB_BUFFER_SERIALIZE_FLAG_DEFAULT		= 0x00000000u,
+  HB_BUFFER_SERIALIZE_FLAG_NO_CLUSTERS		= 0x00000001u,
+  HB_BUFFER_SERIALIZE_FLAG_NO_POSITIONS		= 0x00000002u,
+  HB_BUFFER_SERIALIZE_FLAG_NO_GLYPH_NAMES	= 0x00000004u,
+  HB_BUFFER_SERIALIZE_FLAG_GLYPH_EXTENTS	= 0x00000008u,
+  HB_BUFFER_SERIALIZE_FLAG_GLYPH_FLAGS		= 0x00000010u,
+  HB_BUFFER_SERIALIZE_FLAG_NO_ADVANCES		= 0x00000020u
+} hb_buffer_serialize_flags_t;
+
+/**
+ * hb_buffer_serialize_format_t:
+ * @HB_BUFFER_SERIALIZE_FORMAT_TEXT: a human-readable, plain text format.
+ * @HB_BUFFER_SERIALIZE_FORMAT_JSON: a machine-readable JSON format.
+ * @HB_BUFFER_SERIALIZE_FORMAT_INVALID: invalid format.
+ *
+ * The buffer serialization and de-serialization format used in
+ * hb_buffer_serialize_glyphs() and hb_buffer_deserialize_glyphs().
+ *
+ * Since: 0.9.2
+ */
+typedef enum {
+  HB_BUFFER_SERIALIZE_FORMAT_TEXT	= HB_TAG('T','E','X','T'),
+  HB_BUFFER_SERIALIZE_FORMAT_JSON	= HB_TAG('J','S','O','N'),
+  HB_BUFFER_SERIALIZE_FORMAT_INVALID	= HB_TAG_NONE
+} hb_buffer_serialize_format_t;
+
+HB_EXTERN hb_buffer_serialize_format_t
+hb_buffer_serialize_format_from_string (const char *str, int len);
+
+HB_EXTERN const char *
+hb_buffer_serialize_format_to_string (hb_buffer_serialize_format_t format);
+
+HB_EXTERN const char **
+hb_buffer_serialize_list_formats (void);
+
+HB_EXTERN unsigned int
+hb_buffer_serialize_glyphs (hb_buffer_t *buffer,
+			    unsigned int start,
+			    unsigned int end,
+			    char *buf,
+			    unsigned int buf_size,
+			    unsigned int *buf_consumed,
+			    hb_font_t *font,
+			    hb_buffer_serialize_format_t format,
+			    hb_buffer_serialize_flags_t flags);
+
+HB_EXTERN hb_bool_t
+hb_buffer_deserialize_glyphs (hb_buffer_t *buffer,
+			      const char *buf,
+			      int buf_len,
+			      const char **end_ptr,
+			      hb_font_t *font,
+			      hb_buffer_serialize_format_t format);
+
+
+/*
+ * Compare buffers
+ */
+
+typedef enum { /*< flags >*/
+  HB_BUFFER_DIFF_FLAG_EQUAL			= 0x0000,
+
+  /* Buffers with different content_type cannot be meaningfully compared
+   * in any further detail. */
+  HB_BUFFER_DIFF_FLAG_CONTENT_TYPE_MISMATCH	= 0x0001,
+
+  /* For buffers with differing length, the per-glyph comparison is not
+   * attempted, though we do still scan reference for dottedcircle / .notdef
+   * glyphs. */
+  HB_BUFFER_DIFF_FLAG_LENGTH_MISMATCH		= 0x0002,
+
+  /* We want to know if dottedcircle / .notdef glyphs are present in the
+   * reference, as we may not care so much about other differences in this
+   * case. */
+  HB_BUFFER_DIFF_FLAG_NOTDEF_PRESENT		= 0x0004,
+  HB_BUFFER_DIFF_FLAG_DOTTED_CIRCLE_PRESENT	= 0x0008,
+
+  /* If the buffers have the same length, we compare them glyph-by-glyph
+   * and report which aspect(s) of the glyph info/position are different. */
+  HB_BUFFER_DIFF_FLAG_CODEPOINT_MISMATCH	= 0x0010,
+  HB_BUFFER_DIFF_FLAG_CLUSTER_MISMATCH		= 0x0020,
+  HB_BUFFER_DIFF_FLAG_GLYPH_FLAGS_MISMATCH	= 0x0040,
+  HB_BUFFER_DIFF_FLAG_POSITION_MISMATCH		= 0x0080
+
+} hb_buffer_diff_flags_t;
+
+/* Compare the contents of two buffers, report types of differences. */
+HB_EXTERN hb_buffer_diff_flags_t
+hb_buffer_diff (hb_buffer_t *buffer,
+		hb_buffer_t *reference,
+		hb_codepoint_t dottedcircle_glyph,
+		unsigned int position_fuzz);
+
+
+/*
+ * Debugging.
+ */
+
+typedef hb_bool_t	(*hb_buffer_message_func_t)	(hb_buffer_t *buffer,
+							 hb_font_t   *font,
+							 const char  *message,
+							 void        *user_data);
+
+HB_EXTERN void
+hb_buffer_set_message_func (hb_buffer_t *buffer,
+			    hb_buffer_message_func_t func,
+			    void *user_data, hb_destroy_func_t destroy);
+
+
+HB_END_DECLS
+
+#endif /* HB_BUFFER_H */
diff --git a/thirdparty/harfbuzz/src/hb-buffer.hh b/thirdparty/harfbuzz/src/hb-buffer.hh
new file mode 100644
index 0000000000..3420ba434a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-buffer.hh
@@ -0,0 +1,451 @@
+/*
+ * Copyright © 1998-2004  David Turner and Werner Lemberg
+ * Copyright © 2004,2007,2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Owen Taylor, Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_BUFFER_HH
+#define HB_BUFFER_HH
+
+#include "hb.hh"
+#include "hb-unicode.hh"
+
+
+#ifndef HB_BUFFER_MAX_LEN_FACTOR
+#define HB_BUFFER_MAX_LEN_FACTOR 32
+#endif
+#ifndef HB_BUFFER_MAX_LEN_MIN
+#define HB_BUFFER_MAX_LEN_MIN 8192
+#endif
+#ifndef HB_BUFFER_MAX_LEN_DEFAULT
+#define HB_BUFFER_MAX_LEN_DEFAULT 0x3FFFFFFF /* Shaping more than a billion chars? Let us know! */
+#endif
+
+#ifndef HB_BUFFER_MAX_OPS_FACTOR
+#define HB_BUFFER_MAX_OPS_FACTOR 64
+#endif
+#ifndef HB_BUFFER_MAX_OPS_MIN
+#define HB_BUFFER_MAX_OPS_MIN 1024
+#endif
+#ifndef HB_BUFFER_MAX_OPS_DEFAULT
+#define HB_BUFFER_MAX_OPS_DEFAULT 0x1FFFFFFF /* Shaping more than a billion operations? Let us know! */
+#endif
+
+static_assert ((sizeof (hb_glyph_info_t) == 20), "");
+static_assert ((sizeof (hb_glyph_info_t) == sizeof (hb_glyph_position_t)), "");
+
+HB_MARK_AS_FLAG_T (hb_buffer_flags_t);
+HB_MARK_AS_FLAG_T (hb_buffer_serialize_flags_t);
+HB_MARK_AS_FLAG_T (hb_buffer_diff_flags_t);
+
+enum hb_buffer_scratch_flags_t {
+  HB_BUFFER_SCRATCH_FLAG_DEFAULT			= 0x00000000u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_NON_ASCII			= 0x00000001u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_DEFAULT_IGNORABLES		= 0x00000002u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_SPACE_FALLBACK		= 0x00000004u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT		= 0x00000008u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_UNSAFE_TO_BREAK		= 0x00000010u,
+  HB_BUFFER_SCRATCH_FLAG_HAS_CGJ			= 0x00000020u,
+
+  /* Reserved for complex shapers' internal use. */
+  HB_BUFFER_SCRATCH_FLAG_COMPLEX0			= 0x01000000u,
+  HB_BUFFER_SCRATCH_FLAG_COMPLEX1			= 0x02000000u,
+  HB_BUFFER_SCRATCH_FLAG_COMPLEX2			= 0x04000000u,
+  HB_BUFFER_SCRATCH_FLAG_COMPLEX3			= 0x08000000u,
+};
+HB_MARK_AS_FLAG_T (hb_buffer_scratch_flags_t);
+
+
+/*
+ * hb_buffer_t
+ */
+
+struct hb_buffer_t
+{
+  hb_object_header_t header;
+
+  /* Information about how the text in the buffer should be treated */
+  hb_unicode_funcs_t *unicode; /* Unicode functions */
+  hb_buffer_flags_t flags; /* BOT / EOT / etc. */
+  hb_buffer_cluster_level_t cluster_level;
+  hb_codepoint_t replacement; /* U+FFFD or something else. */
+  hb_codepoint_t invisible; /* 0 or something else. */
+  hb_buffer_scratch_flags_t scratch_flags; /* Have space-fallback, etc. */
+  unsigned int max_len; /* Maximum allowed len. */
+  int max_ops; /* Maximum allowed operations. */
+
+  /* Buffer contents */
+  hb_buffer_content_type_t content_type;
+  hb_segment_properties_t props; /* Script, language, direction */
+
+  bool successful; /* Allocations successful */
+  bool have_output; /* Whether we have an output buffer going on */
+  bool have_positions; /* Whether we have positions */
+
+  unsigned int idx; /* Cursor into ->info and ->pos arrays */
+  unsigned int len; /* Length of ->info and ->pos arrays */
+  unsigned int out_len; /* Length of ->out array if have_output */
+
+  unsigned int allocated; /* Length of allocated arrays */
+  hb_glyph_info_t     *info;
+  hb_glyph_info_t     *out_info;
+  hb_glyph_position_t *pos;
+
+  unsigned int serial;
+
+  /* Text before / after the main buffer contents.
+   * Always in Unicode, and ordered outward.
+   * Index 0 is for "pre-context", 1 for "post-context". */
+  static constexpr unsigned CONTEXT_LENGTH = 5u;
+  hb_codepoint_t context[2][CONTEXT_LENGTH];
+  unsigned int context_len[2];
+
+  /* Debugging API */
+#ifndef HB_NO_BUFFER_MESSAGE
+  hb_buffer_message_func_t message_func;
+  void *message_data;
+  hb_destroy_func_t message_destroy;
+#endif
+
+  /* Internal debugging. */
+  /* The bits here reflect current allocations of the bytes in glyph_info_t's var1 and var2. */
+#ifndef HB_NDEBUG
+  uint8_t allocated_var_bits;
+#endif
+
+
+  /* Methods */
+
+  bool in_error () const { return !successful; }
+
+  void allocate_var (unsigned int start, unsigned int count)
+  {
+#ifndef HB_NDEBUG
+    unsigned int end = start + count;
+    assert (end <= 8);
+    unsigned int bits = (1u<<end) - (1u<<start);
+    assert (0 == (allocated_var_bits & bits));
+    allocated_var_bits |= bits;
+#endif
+  }
+  void deallocate_var (unsigned int start, unsigned int count)
+  {
+#ifndef HB_NDEBUG
+    unsigned int end = start + count;
+    assert (end <= 8);
+    unsigned int bits = (1u<<end) - (1u<<start);
+    assert (bits == (allocated_var_bits & bits));
+    allocated_var_bits &= ~bits;
+#endif
+  }
+  void assert_var (unsigned int start, unsigned int count)
+  {
+#ifndef HB_NDEBUG
+    unsigned int end = start + count;
+    assert (end <= 8);
+    unsigned int bits = (1u<<end) - (1u<<start);
+    assert (bits == (allocated_var_bits & bits));
+#endif
+  }
+  void deallocate_var_all ()
+  {
+#ifndef HB_NDEBUG
+    allocated_var_bits = 0;
+#endif
+  }
+
+  hb_glyph_info_t &cur (unsigned int i = 0) { return info[idx + i]; }
+  hb_glyph_info_t cur (unsigned int i = 0) const { return info[idx + i]; }
+
+  hb_glyph_position_t &cur_pos (unsigned int i = 0) { return pos[idx + i]; }
+  hb_glyph_position_t cur_pos (unsigned int i = 0) const { return pos[idx + i]; }
+
+  hb_glyph_info_t &prev ()      { return out_info[out_len ? out_len - 1 : 0]; }
+  hb_glyph_info_t prev () const { return out_info[out_len ? out_len - 1 : 0]; }
+
+  bool has_separate_output () const { return info != out_info; }
+
+
+  HB_INTERNAL void reset ();
+  HB_INTERNAL void clear ();
+
+  unsigned int backtrack_len () const { return have_output? out_len : idx; }
+  unsigned int lookahead_len () const { return len - idx; }
+  unsigned int next_serial () { return serial++; }
+
+  HB_INTERNAL void add (hb_codepoint_t  codepoint,
+			unsigned int    cluster);
+  HB_INTERNAL void add_info (const hb_glyph_info_t &glyph_info);
+
+  HB_INTERNAL void reverse_range (unsigned int start, unsigned int end);
+  HB_INTERNAL void reverse ();
+  HB_INTERNAL void reverse_clusters ();
+  HB_INTERNAL void guess_segment_properties ();
+
+  HB_INTERNAL void swap_buffers ();
+  HB_INTERNAL void remove_output ();
+  HB_INTERNAL void clear_output ();
+  HB_INTERNAL void clear_positions ();
+
+  HB_INTERNAL void replace_glyphs (unsigned int num_in,
+				   unsigned int num_out,
+				   const hb_codepoint_t *glyph_data);
+
+  void replace_glyph (hb_codepoint_t glyph_index)
+  {
+    if (unlikely (out_info != info || out_len != idx)) {
+      if (unlikely (!make_room_for (1, 1))) return;
+      out_info[out_len] = info[idx];
+    }
+    out_info[out_len].codepoint = glyph_index;
+
+    idx++;
+    out_len++;
+  }
+  /* Makes a copy of the glyph at idx to output and replace glyph_index */
+  hb_glyph_info_t & output_glyph (hb_codepoint_t glyph_index)
+  {
+    if (unlikely (!make_room_for (0, 1))) return Crap (hb_glyph_info_t);
+
+    if (unlikely (idx == len && !out_len))
+      return Crap (hb_glyph_info_t);
+
+    out_info[out_len] = idx < len ? info[idx] : out_info[out_len - 1];
+    out_info[out_len].codepoint = glyph_index;
+
+    out_len++;
+
+    return out_info[out_len - 1];
+  }
+  void output_info (const hb_glyph_info_t &glyph_info)
+  {
+    if (unlikely (!make_room_for (0, 1))) return;
+
+    out_info[out_len] = glyph_info;
+
+    out_len++;
+  }
+  /* Copies glyph at idx to output but doesn't advance idx */
+  void copy_glyph ()
+  {
+    if (unlikely (!make_room_for (0, 1))) return;
+
+    out_info[out_len] = info[idx];
+
+    out_len++;
+  }
+  /* Copies glyph at idx to output and advance idx.
+   * If there's no output, just advance idx. */
+  void
+  next_glyph ()
+  {
+    if (have_output)
+    {
+      if (out_info != info || out_len != idx)
+      {
+	if (unlikely (!make_room_for (1, 1))) return;
+	out_info[out_len] = info[idx];
+      }
+      out_len++;
+    }
+
+    idx++;
+  }
+  /* Copies n glyphs at idx to output and advance idx.
+   * If there's no output, just advance idx. */
+  void
+  next_glyphs (unsigned int n)
+  {
+    if (have_output)
+    {
+      if (out_info != info || out_len != idx)
+      {
+	if (unlikely (!make_room_for (n, n))) return;
+	memmove (out_info + out_len, info + idx, n * sizeof (out_info[0]));
+      }
+      out_len += n;
+    }
+
+    idx += n;
+  }
+  /* Advance idx without copying to output. */
+  void skip_glyph () { idx++; }
+  void reset_masks (hb_mask_t mask)
+  {
+    for (unsigned int j = 0; j < len; j++)
+      info[j].mask = mask;
+  }
+  void add_masks (hb_mask_t mask)
+  {
+    for (unsigned int j = 0; j < len; j++)
+      info[j].mask |= mask;
+  }
+  HB_INTERNAL void set_masks (hb_mask_t value, hb_mask_t mask,
+			      unsigned int cluster_start, unsigned int cluster_end);
+
+  void merge_clusters (unsigned int start, unsigned int end)
+  {
+    if (end - start < 2)
+      return;
+    merge_clusters_impl (start, end);
+  }
+  HB_INTERNAL void merge_clusters_impl (unsigned int start, unsigned int end);
+  HB_INTERNAL void merge_out_clusters (unsigned int start, unsigned int end);
+  /* Merge clusters for deleting current glyph, and skip it. */
+  HB_INTERNAL void delete_glyph ();
+
+  void unsafe_to_break (unsigned int start,
+			unsigned int end)
+  {
+    if (end - start < 2)
+      return;
+    unsafe_to_break_impl (start, end);
+  }
+  HB_INTERNAL void unsafe_to_break_impl (unsigned int start, unsigned int end);
+  HB_INTERNAL void unsafe_to_break_from_outbuffer (unsigned int start, unsigned int end);
+
+
+  /* Internal methods */
+  HB_INTERNAL bool move_to (unsigned int i); /* i is output-buffer index. */
+
+  HB_INTERNAL bool enlarge (unsigned int size);
+
+  bool ensure (unsigned int size)
+  { return likely (!size || size < allocated) ? true : enlarge (size); }
+
+  bool ensure_inplace (unsigned int size)
+  { return likely (!size || size < allocated); }
+
+  HB_INTERNAL bool make_room_for (unsigned int num_in, unsigned int num_out);
+  HB_INTERNAL bool shift_forward (unsigned int count);
+
+  typedef long scratch_buffer_t;
+  HB_INTERNAL scratch_buffer_t *get_scratch_buffer (unsigned int *size);
+
+  void clear_context (unsigned int side) { context_len[side] = 0; }
+
+  HB_INTERNAL void sort (unsigned int start, unsigned int end, int(*compar)(const hb_glyph_info_t *, const hb_glyph_info_t *));
+
+  bool messaging ()
+  {
+#ifdef HB_NO_BUFFER_MESSAGE
+    return false;
+#else
+    return unlikely (message_func);
+#endif
+  }
+  bool message (hb_font_t *font, const char *fmt, ...) HB_PRINTF_FUNC(3, 4)
+  {
+#ifdef HB_NO_BUFFER_MESSAGE
+   return true;
+#else
+    if (!messaging ())
+      return true;
+    va_list ap;
+    va_start (ap, fmt);
+    bool ret = message_impl (font, fmt, ap);
+    va_end (ap);
+    return ret;
+#endif
+  }
+  HB_INTERNAL bool message_impl (hb_font_t *font, const char *fmt, va_list ap) HB_PRINTF_FUNC(3, 0);
+
+  static void
+  set_cluster (hb_glyph_info_t &inf, unsigned int cluster, unsigned int mask = 0)
+  {
+    if (inf.cluster != cluster)
+    {
+      if (mask & HB_GLYPH_FLAG_UNSAFE_TO_BREAK)
+	inf.mask |= HB_GLYPH_FLAG_UNSAFE_TO_BREAK;
+      else
+	inf.mask &= ~HB_GLYPH_FLAG_UNSAFE_TO_BREAK;
+    }
+    inf.cluster = cluster;
+  }
+
+  unsigned int
+  _unsafe_to_break_find_min_cluster (const hb_glyph_info_t *infos,
+				     unsigned int start, unsigned int end,
+				     unsigned int cluster) const
+  {
+    for (unsigned int i = start; i < end; i++)
+      cluster = hb_min (cluster, infos[i].cluster);
+    return cluster;
+  }
+  void
+  _unsafe_to_break_set_mask (hb_glyph_info_t *infos,
+			     unsigned int start, unsigned int end,
+			     unsigned int cluster)
+  {
+    for (unsigned int i = start; i < end; i++)
+      if (cluster != infos[i].cluster)
+      {
+	scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_UNSAFE_TO_BREAK;
+	infos[i].mask |= HB_GLYPH_FLAG_UNSAFE_TO_BREAK;
+      }
+  }
+
+  void unsafe_to_break_all () { unsafe_to_break_impl (0, len); }
+  void safe_to_break_all ()
+  {
+    for (unsigned int i = 0; i < len; i++)
+      info[i].mask &= ~HB_GLYPH_FLAG_UNSAFE_TO_BREAK;
+  }
+};
+DECLARE_NULL_INSTANCE (hb_buffer_t);
+
+
+/* Loop over clusters. Duplicated in foreach_syllable(). */
+#define foreach_cluster(buffer, start, end) \
+  for (unsigned int \
+       _count = buffer->len, \
+       start = 0, end = _count ? _next_cluster (buffer, 0) : 0; \
+       start < _count; \
+       start = end, end = _next_cluster (buffer, start))
+
+static inline unsigned int
+_next_cluster (hb_buffer_t *buffer, unsigned int start)
+{
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+
+  unsigned int cluster = info[start].cluster;
+  while (++start < count && cluster == info[start].cluster)
+    ;
+
+  return start;
+}
+
+
+#define HB_BUFFER_XALLOCATE_VAR(b, func, var) \
+  b->func (offsetof (hb_glyph_info_t, var) - offsetof(hb_glyph_info_t, var1), \
+	   sizeof (b->info[0].var))
+#define HB_BUFFER_ALLOCATE_VAR(b, var)		HB_BUFFER_XALLOCATE_VAR (b, allocate_var,   var ())
+#define HB_BUFFER_DEALLOCATE_VAR(b, var)	HB_BUFFER_XALLOCATE_VAR (b, deallocate_var, var ())
+#define HB_BUFFER_ASSERT_VAR(b, var)		HB_BUFFER_XALLOCATE_VAR (b, assert_var,     var ())
+
+
+#endif /* HB_BUFFER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cache.hh b/thirdparty/harfbuzz/src/hb-cache.hh
new file mode 100644
index 0000000000..bf26d96be4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cache.hh
@@ -0,0 +1,80 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_CACHE_HH
+#define HB_CACHE_HH
+
+#include "hb.hh"
+
+
+/* Implements a lock-free cache for int->int functions. */
+
+template <unsigned int key_bits, unsigned int value_bits, unsigned int cache_bits>
+struct hb_cache_t
+{
+  static_assert ((key_bits >= cache_bits), "");
+  static_assert ((key_bits + value_bits - cache_bits <= 8 * sizeof (hb_atomic_int_t)), "");
+  static_assert (sizeof (hb_atomic_int_t) == sizeof (unsigned int), "");
+
+  void init () { clear (); }
+  void fini () {}
+
+  void clear ()
+  {
+    for (unsigned i = 0; i < ARRAY_LENGTH (values); i++)
+      values[i].set_relaxed (-1);
+  }
+
+  bool get (unsigned int key, unsigned int *value) const
+  {
+    unsigned int k = key & ((1u<<cache_bits)-1);
+    unsigned int v = values[k].get_relaxed ();
+    if ((key_bits + value_bits - cache_bits == 8 * sizeof (hb_atomic_int_t) && v == (unsigned int) -1) ||
+	(v >> value_bits) != (key >> cache_bits))
+      return false;
+    *value = v & ((1u<<value_bits)-1);
+    return true;
+  }
+
+  bool set (unsigned int key, unsigned int value)
+  {
+    if (unlikely ((key >> key_bits) || (value >> value_bits)))
+      return false; /* Overflows */
+    unsigned int k = key & ((1u<<cache_bits)-1);
+    unsigned int v = ((key>>cache_bits)<<value_bits) | value;
+    values[k].set_relaxed (v);
+    return true;
+  }
+
+  private:
+  hb_atomic_int_t values[1u<<cache_bits];
+};
+
+typedef hb_cache_t<21, 16, 8> hb_cmap_cache_t;
+typedef hb_cache_t<16, 24, 8> hb_advance_cache_t;
+
+
+#endif /* HB_CACHE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cff-interp-common.hh b/thirdparty/harfbuzz/src/hb-cff-interp-common.hh
new file mode 100644
index 0000000000..91a9b7d0d1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cff-interp-common.hh
@@ -0,0 +1,688 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_CFF_INTERP_COMMON_HH
+#define HB_CFF_INTERP_COMMON_HH
+
+namespace CFF {
+
+using namespace OT;
+
+typedef unsigned int op_code_t;
+
+
+/* === Dict operators === */
+
+/* One byte operators (0-31) */
+#define OpCode_version		  0 /* CFF Top */
+#define OpCode_Notice		  1 /* CFF Top */
+#define OpCode_FullName		  2 /* CFF Top */
+#define OpCode_FamilyName	  3 /* CFF Top */
+#define OpCode_Weight		  4 /* CFF Top */
+#define OpCode_FontBBox		  5 /* CFF Top */
+#define OpCode_BlueValues	  6 /* CFF Private, CFF2 Private */
+#define OpCode_OtherBlues	  7 /* CFF Private, CFF2 Private */
+#define OpCode_FamilyBlues	  8 /* CFF Private, CFF2 Private */
+#define OpCode_FamilyOtherBlues	  9 /* CFF Private, CFF2 Private */
+#define OpCode_StdHW		 10 /* CFF Private, CFF2 Private */
+#define OpCode_StdVW		 11 /* CFF Private, CFF2 Private */
+#define OpCode_escape		 12 /* All. Shared with CS */
+#define OpCode_UniqueID		 13 /* CFF Top */
+#define OpCode_XUID		 14 /* CFF Top */
+#define OpCode_charset		 15 /* CFF Top (0) */
+#define OpCode_Encoding		 16 /* CFF Top (0) */
+#define OpCode_CharStrings	 17 /* CFF Top, CFF2 Top */
+#define OpCode_Private		 18 /* CFF Top, CFF2 FD */
+#define OpCode_Subrs		 19 /* CFF Private, CFF2 Private */
+#define OpCode_defaultWidthX	 20 /* CFF Private (0) */
+#define OpCode_nominalWidthX	 21 /* CFF Private (0) */
+#define OpCode_vsindexdict	 22 /* CFF2 Private/CS */
+#define OpCode_blenddict	 23 /* CFF2 Private/CS */
+#define OpCode_vstore		 24 /* CFF2 Top */
+#define OpCode_reserved25	 25
+#define OpCode_reserved26	 26
+#define OpCode_reserved27	 27
+
+/* Numbers */
+#define OpCode_shortint		 28 /* 16-bit integer, All */
+#define OpCode_longintdict	 29 /* 32-bit integer, All */
+#define OpCode_BCD		 30 /* Real number, CFF2 Top/FD */
+#define OpCode_reserved31	 31
+
+/* 1-byte integers */
+#define OpCode_OneByteIntFirst	 32 /* All. beginning of the range of first byte ints */
+#define OpCode_OneByteIntLast	246 /* All. ending of the range of first byte int */
+
+/* 2-byte integers */
+#define OpCode_TwoBytePosInt0	247 /* All. first byte of two byte positive int (+108 to +1131) */
+#define OpCode_TwoBytePosInt1	248
+#define OpCode_TwoBytePosInt2	249
+#define OpCode_TwoBytePosInt3	250
+
+#define OpCode_TwoByteNegInt0	251 /* All. first byte of two byte negative int (-1131 to -108) */
+#define OpCode_TwoByteNegInt1	252
+#define OpCode_TwoByteNegInt2	253
+#define OpCode_TwoByteNegInt3	254
+
+/* Two byte escape operators 12, (0-41) */
+#define OpCode_ESC_Base		256
+#define Make_OpCode_ESC(byte2)	((op_code_t)(OpCode_ESC_Base + (byte2)))
+
+inline op_code_t Unmake_OpCode_ESC (op_code_t op)  { return (op_code_t)(op - OpCode_ESC_Base); }
+inline bool Is_OpCode_ESC (op_code_t op) { return op >= OpCode_ESC_Base; }
+inline unsigned int OpCode_Size (op_code_t op) { return Is_OpCode_ESC (op) ? 2: 1; }
+
+#define OpCode_Copyright	Make_OpCode_ESC(0) /* CFF Top */
+#define OpCode_isFixedPitch	Make_OpCode_ESC(1) /* CFF Top (false) */
+#define OpCode_ItalicAngle	Make_OpCode_ESC(2) /* CFF Top (0) */
+#define OpCode_UnderlinePosition Make_OpCode_ESC(3) /* CFF Top (-100) */
+#define OpCode_UnderlineThickness Make_OpCode_ESC(4) /* CFF Top (50) */
+#define OpCode_PaintType	Make_OpCode_ESC(5) /* CFF Top (0) */
+#define OpCode_CharstringType	Make_OpCode_ESC(6) /* CFF Top (2) */
+#define OpCode_FontMatrix	Make_OpCode_ESC(7) /* CFF Top, CFF2 Top (.001 0 0 .001 0 0)*/
+#define OpCode_StrokeWidth	Make_OpCode_ESC(8) /* CFF Top (0) */
+#define OpCode_BlueScale	Make_OpCode_ESC(9) /* CFF Private, CFF2 Private (0.039625) */
+#define OpCode_BlueShift	Make_OpCode_ESC(10) /* CFF Private, CFF2 Private (7) */
+#define OpCode_BlueFuzz		Make_OpCode_ESC(11) /* CFF Private, CFF2 Private (1) */
+#define OpCode_StemSnapH	Make_OpCode_ESC(12) /* CFF Private, CFF2 Private */
+#define OpCode_StemSnapV	Make_OpCode_ESC(13) /* CFF Private, CFF2 Private */
+#define OpCode_ForceBold	Make_OpCode_ESC(14) /* CFF Private (false) */
+#define OpCode_reservedESC15	Make_OpCode_ESC(15)
+#define OpCode_reservedESC16	Make_OpCode_ESC(16)
+#define OpCode_LanguageGroup	Make_OpCode_ESC(17) /* CFF Private, CFF2 Private (0) */
+#define OpCode_ExpansionFactor	Make_OpCode_ESC(18) /* CFF Private, CFF2 Private (0.06) */
+#define OpCode_initialRandomSeed Make_OpCode_ESC(19) /* CFF Private (0) */
+#define OpCode_SyntheticBase	Make_OpCode_ESC(20) /* CFF Top */
+#define OpCode_PostScript	Make_OpCode_ESC(21) /* CFF Top */
+#define OpCode_BaseFontName	Make_OpCode_ESC(22) /* CFF Top */
+#define OpCode_BaseFontBlend	Make_OpCode_ESC(23) /* CFF Top */
+#define OpCode_reservedESC24	Make_OpCode_ESC(24)
+#define OpCode_reservedESC25	Make_OpCode_ESC(25)
+#define OpCode_reservedESC26	Make_OpCode_ESC(26)
+#define OpCode_reservedESC27	Make_OpCode_ESC(27)
+#define OpCode_reservedESC28	Make_OpCode_ESC(28)
+#define OpCode_reservedESC29	Make_OpCode_ESC(29)
+#define OpCode_ROS		Make_OpCode_ESC(30) /* CFF Top_CID */
+#define OpCode_CIDFontVersion	Make_OpCode_ESC(31) /* CFF Top_CID (0) */
+#define OpCode_CIDFontRevision	Make_OpCode_ESC(32) /* CFF Top_CID (0) */
+#define OpCode_CIDFontType	Make_OpCode_ESC(33) /* CFF Top_CID (0) */
+#define OpCode_CIDCount		Make_OpCode_ESC(34) /* CFF Top_CID (8720) */
+#define OpCode_UIDBase		Make_OpCode_ESC(35) /* CFF Top_CID */
+#define OpCode_FDArray		Make_OpCode_ESC(36) /* CFF Top_CID, CFF2 Top */
+#define OpCode_FDSelect		Make_OpCode_ESC(37) /* CFF Top_CID, CFF2 Top */
+#define OpCode_FontName		Make_OpCode_ESC(38) /* CFF Top_CID */
+
+
+/* === CharString operators === */
+
+#define OpCode_hstem		  1 /* CFF, CFF2 */
+#define OpCode_Reserved2	  2
+#define OpCode_vstem		  3 /* CFF, CFF2 */
+#define OpCode_vmoveto		  4 /* CFF, CFF2 */
+#define OpCode_rlineto		  5 /* CFF, CFF2 */
+#define OpCode_hlineto		  6 /* CFF, CFF2 */
+#define OpCode_vlineto		  7 /* CFF, CFF2 */
+#define OpCode_rrcurveto	  8 /* CFF, CFF2 */
+#define OpCode_Reserved9	  9
+#define OpCode_callsubr		 10 /* CFF, CFF2 */
+#define OpCode_return		 11 /* CFF */
+//#define OpCode_escape		 12 /* CFF, CFF2 */
+#define OpCode_Reserved13	 13
+#define OpCode_endchar		 14 /* CFF */
+#define OpCode_vsindexcs	 15 /* CFF2 */
+#define OpCode_blendcs		 16 /* CFF2 */
+#define OpCode_Reserved17	 17
+#define OpCode_hstemhm		 18 /* CFF, CFF2 */
+#define OpCode_hintmask		 19 /* CFF, CFF2 */
+#define OpCode_cntrmask		 20 /* CFF, CFF2 */
+#define OpCode_rmoveto		 21 /* CFF, CFF2 */
+#define OpCode_hmoveto		 22 /* CFF, CFF2 */
+#define OpCode_vstemhm		 23 /* CFF, CFF2 */
+#define OpCode_rcurveline	 24 /* CFF, CFF2 */
+#define OpCode_rlinecurve	 25 /* CFF, CFF2 */
+#define OpCode_vvcurveto	 26 /* CFF, CFF2 */
+#define OpCode_hhcurveto	 27 /* CFF, CFF2 */
+//#define OpCode_shortint	 28 /* CFF, CFF2 */
+#define OpCode_callgsubr	 29 /* CFF, CFF2 */
+#define OpCode_vhcurveto	 30 /* CFF, CFF2 */
+#define OpCode_hvcurveto	 31 /* CFF, CFF2 */
+
+#define OpCode_fixedcs		255 /* 32-bit fixed */
+
+/* Two byte escape operators 12, (0-41) */
+#define OpCode_dotsection	Make_OpCode_ESC(0) /* CFF (obsoleted) */
+#define OpCode_ReservedESC1	Make_OpCode_ESC(1)
+#define OpCode_ReservedESC2	Make_OpCode_ESC(2)
+#define OpCode_and		Make_OpCode_ESC(3) /* CFF */
+#define OpCode_or		Make_OpCode_ESC(4) /* CFF */
+#define OpCode_not		Make_OpCode_ESC(5) /* CFF */
+#define OpCode_ReservedESC6	Make_OpCode_ESC(6)
+#define OpCode_ReservedESC7	Make_OpCode_ESC(7)
+#define OpCode_ReservedESC8	Make_OpCode_ESC(8)
+#define OpCode_abs		Make_OpCode_ESC(9) /* CFF */
+#define OpCode_add		Make_OpCode_ESC(10) /* CFF */
+#define OpCode_sub		Make_OpCode_ESC(11) /* CFF */
+#define OpCode_div		Make_OpCode_ESC(12) /* CFF */
+#define OpCode_ReservedESC13	Make_OpCode_ESC(13)
+#define OpCode_neg		Make_OpCode_ESC(14) /* CFF */
+#define OpCode_eq		Make_OpCode_ESC(15) /* CFF */
+#define OpCode_ReservedESC16	Make_OpCode_ESC(16)
+#define OpCode_ReservedESC17	Make_OpCode_ESC(17)
+#define OpCode_drop		Make_OpCode_ESC(18) /* CFF */
+#define OpCode_ReservedESC19	Make_OpCode_ESC(19)
+#define OpCode_put		Make_OpCode_ESC(20) /* CFF */
+#define OpCode_get		Make_OpCode_ESC(21) /* CFF */
+#define OpCode_ifelse		Make_OpCode_ESC(22) /* CFF */
+#define OpCode_random		Make_OpCode_ESC(23) /* CFF */
+#define OpCode_mul		Make_OpCode_ESC(24) /* CFF */
+//#define OpCode_reservedESC25	Make_OpCode_ESC(25)
+#define OpCode_sqrt		Make_OpCode_ESC(26) /* CFF */
+#define OpCode_dup		Make_OpCode_ESC(27) /* CFF */
+#define OpCode_exch		Make_OpCode_ESC(28) /* CFF */
+#define OpCode_index		Make_OpCode_ESC(29) /* CFF */
+#define OpCode_roll		Make_OpCode_ESC(30) /* CFF */
+#define OpCode_reservedESC31	Make_OpCode_ESC(31)
+#define OpCode_reservedESC32	Make_OpCode_ESC(32)
+#define OpCode_reservedESC33	Make_OpCode_ESC(33)
+#define OpCode_hflex		Make_OpCode_ESC(34) /* CFF, CFF2 */
+#define OpCode_flex		Make_OpCode_ESC(35) /* CFF, CFF2 */
+#define OpCode_hflex1		Make_OpCode_ESC(36) /* CFF, CFF2 */
+#define OpCode_flex1		Make_OpCode_ESC(37) /* CFF, CFF2 */
+
+
+#define OpCode_Invalid		0xFFFFu
+
+
+struct number_t
+{
+  void init () { set_real (0.0); }
+  void fini () {}
+
+  void set_int (int v)       { value = v; }
+  int to_int () const        { return value; }
+
+  void set_fixed (int32_t v) { value = v / 65536.0; }
+  int32_t to_fixed () const  { return value * 65536.0; }
+
+  void set_real (double v)   { value = v; }
+  double to_real () const    { return value; }
+
+  bool in_int_range () const
+  { return ((double) (int16_t) to_int () == value); }
+
+  bool operator >  (const number_t &n) const { return value > n.to_real (); }
+  bool operator <  (const number_t &n) const { return n > *this; }
+  bool operator >= (const number_t &n) const { return !(*this < n); }
+  bool operator <= (const number_t &n) const { return !(*this > n); }
+
+  const number_t &operator += (const number_t &n)
+  {
+    set_real (to_real () + n.to_real ());
+
+    return *this;
+  }
+
+  protected:
+  double value;
+};
+
+/* byte string */
+struct UnsizedByteStr : UnsizedArrayOf <HBUINT8>
+{
+  // encode 2-byte int (Dict/CharString) or 4-byte int (Dict)
+  template <typename T, typename V>
+  static bool serialize_int (hb_serialize_context_t *c, op_code_t intOp, V value)
+  {
+    TRACE_SERIALIZE (this);
+
+    HBUINT8 *p = c->allocate_size<HBUINT8> (1);
+    if (unlikely (!p)) return_trace (false);
+    *p = intOp;
+
+    T *ip = c->allocate_size<T> (T::static_size);
+    if (unlikely (!ip)) return_trace (false);
+    return_trace (c->check_assign (*ip, value));
+  }
+
+  template <typename V>
+  static bool serialize_int4 (hb_serialize_context_t *c, V value)
+  { return serialize_int<HBINT32> (c, OpCode_longintdict, value); }
+
+  template <typename V>
+  static bool serialize_int2 (hb_serialize_context_t *c, V value)
+  { return serialize_int<HBINT16> (c, OpCode_shortint, value); }
+
+  /* Defining null_size allows a Null object may be created. Should be safe because:
+   * A descendent struct Dict uses a Null pointer to indicate a missing table,
+   * checked before access.
+   * byte_str_t, a wrapper struct pairing a byte pointer along with its length, always
+   * checks the length before access. A Null pointer is used as the initial pointer
+   * along with zero length by the default ctor.
+   */
+  DEFINE_SIZE_MIN(0);
+};
+
+/* Holder of a section of byte string within a CFFIndex entry */
+struct byte_str_t : hb_ubytes_t
+{
+  byte_str_t ()
+    : hb_ubytes_t () {}
+  byte_str_t (const UnsizedByteStr& s, unsigned int l)
+    : hb_ubytes_t ((const unsigned char*)&s, l) {}
+  byte_str_t (const unsigned char *s, unsigned int l)
+    : hb_ubytes_t (s, l) {}
+  byte_str_t (const hb_ubytes_t &ub)	/* conversion from hb_ubytes_t */
+    : hb_ubytes_t (ub) {}
+
+  /* sub-string */
+  byte_str_t sub_str (unsigned int offset, unsigned int len_) const
+  { return byte_str_t (hb_ubytes_t::sub_array (offset, len_)); }
+
+  bool check_limit (unsigned int offset, unsigned int count) const
+  { return (offset + count <= length); }
+};
+
+/* A byte string associated with the current offset and an error condition */
+struct byte_str_ref_t
+{
+  byte_str_ref_t () { init (); }
+
+  void init ()
+  {
+    str = byte_str_t ();
+    offset = 0;
+    error = false;
+  }
+
+  void fini () {}
+
+  byte_str_ref_t (const byte_str_t &str_, unsigned int offset_ = 0)
+    : str (str_), offset (offset_), error (false) {}
+
+  void reset (const byte_str_t &str_, unsigned int offset_ = 0)
+  {
+    str = str_;
+    offset = offset_;
+    error = false;
+  }
+
+  const unsigned char& operator [] (int i) {
+    if (unlikely ((unsigned int) (offset + i) >= str.length))
+    {
+      set_error ();
+      return Null (unsigned char);
+    }
+    return str[offset + i];
+  }
+
+  /* Conversion to byte_str_t */
+  operator byte_str_t () const { return str.sub_str (offset, str.length - offset); }
+
+  byte_str_t sub_str (unsigned int offset_, unsigned int len_) const
+  { return str.sub_str (offset_, len_); }
+
+  bool avail (unsigned int count=1) const
+  { return (!in_error () && str.check_limit (offset, count)); }
+  void inc (unsigned int count=1)
+  {
+    if (likely (!in_error () && (offset <= str.length) && (offset + count <= str.length)))
+    {
+      offset += count;
+    }
+    else
+    {
+      offset = str.length;
+      set_error ();
+    }
+  }
+
+  void set_error ()      { error = true; }
+  bool in_error () const { return error; }
+
+  byte_str_t       str;
+  unsigned int  offset; /* beginning of the sub-string within str */
+
+  protected:
+  bool	  error;
+};
+
+typedef hb_vector_t<byte_str_t> byte_str_array_t;
+
+/* stack */
+template <typename ELEM, int LIMIT>
+struct cff_stack_t
+{
+  void init ()
+  {
+    error = false;
+    count = 0;
+    elements.init ();
+    elements.resize (kSizeLimit);
+    for (unsigned int i = 0; i < elements.length; i++)
+      elements[i].init ();
+  }
+  void fini () { elements.fini_deep (); }
+
+  ELEM& operator [] (unsigned int i)
+  {
+    if (unlikely (i >= count)) set_error ();
+    return elements[i];
+  }
+
+  void push (const ELEM &v)
+  {
+    if (likely (count < elements.length))
+      elements[count++] = v;
+    else
+      set_error ();
+  }
+  ELEM &push ()
+  {
+    if (likely (count < elements.length))
+      return elements[count++];
+    else
+    {
+      set_error ();
+      return Crap (ELEM);
+    }
+  }
+
+  ELEM& pop ()
+  {
+    if (likely (count > 0))
+      return elements[--count];
+    else
+    {
+      set_error ();
+      return Crap (ELEM);
+    }
+  }
+  void pop (unsigned int n)
+  {
+    if (likely (count >= n))
+      count -= n;
+    else
+      set_error ();
+  }
+
+  const ELEM& peek ()
+  {
+    if (unlikely (count < 0))
+    {
+      set_error ();
+      return Null (ELEM);
+    }
+    return elements[count - 1];
+  }
+
+  void unpop ()
+  {
+    if (likely (count < elements.length))
+      count++;
+    else
+      set_error ();
+  }
+
+  void clear () { count = 0; }
+
+  bool in_error () const { return (error || elements.in_error ()); }
+  void set_error ()      { error = true; }
+
+  unsigned int get_count () const { return count; }
+  bool is_empty () const          { return !count; }
+
+  static constexpr unsigned kSizeLimit = LIMIT;
+
+  protected:
+  bool error;
+  unsigned int count;
+  hb_vector_t<ELEM> elements;
+};
+
+/* argument stack */
+template <typename ARG=number_t>
+struct arg_stack_t : cff_stack_t<ARG, 513>
+{
+  void push_int (int v)
+  {
+    ARG &n = S::push ();
+    n.set_int (v);
+  }
+
+  void push_fixed (int32_t v)
+  {
+    ARG &n = S::push ();
+    n.set_fixed (v);
+  }
+
+  void push_real (double v)
+  {
+    ARG &n = S::push ();
+    n.set_real (v);
+  }
+
+  ARG& pop_num () { return this->pop (); }
+
+  int pop_int ()  { return this->pop ().to_int (); }
+
+  unsigned int pop_uint ()
+  {
+    int i = pop_int ();
+    if (unlikely (i < 0))
+    {
+      i = 0;
+      S::set_error ();
+    }
+    return (unsigned) i;
+  }
+
+  void push_longint_from_substr (byte_str_ref_t& str_ref)
+  {
+    push_int ((str_ref[0] << 24) | (str_ref[1] << 16) | (str_ref[2] << 8) | (str_ref[3]));
+    str_ref.inc (4);
+  }
+
+  bool push_fixed_from_substr (byte_str_ref_t& str_ref)
+  {
+    if (unlikely (!str_ref.avail (4)))
+      return false;
+    push_fixed ((int32_t)*(const HBUINT32*)&str_ref[0]);
+    str_ref.inc (4);
+    return true;
+  }
+
+  hb_array_t<const ARG> get_subarray (unsigned int start) const
+  { return S::elements.sub_array (start); }
+
+  private:
+  typedef cff_stack_t<ARG, 513> S;
+};
+
+/* an operator prefixed by its operands in a byte string */
+struct op_str_t
+{
+  void init () {}
+  void fini () {}
+
+  op_code_t  op;
+  byte_str_t str;
+};
+
+/* base of OP_SERIALIZER */
+struct op_serializer_t
+{
+  protected:
+  bool copy_opstr (hb_serialize_context_t *c, const op_str_t& opstr) const
+  {
+    TRACE_SERIALIZE (this);
+
+    HBUINT8 *d = c->allocate_size<HBUINT8> (opstr.str.length);
+    if (unlikely (!d)) return_trace (false);
+    memcpy (d, &opstr.str[0], opstr.str.length);
+    return_trace (true);
+  }
+};
+
+template <typename VAL>
+struct parsed_values_t
+{
+  void init ()
+  {
+    opStart = 0;
+    values.init ();
+  }
+  void fini () { values.fini_deep (); }
+
+  void add_op (op_code_t op, const byte_str_ref_t& str_ref = byte_str_ref_t ())
+  {
+    VAL *val = values.push ();
+    val->op = op;
+    val->str = str_ref.str.sub_str (opStart, str_ref.offset - opStart);
+    opStart = str_ref.offset;
+  }
+
+  void add_op (op_code_t op, const byte_str_ref_t& str_ref, const VAL &v)
+  {
+    VAL *val = values.push (v);
+    val->op = op;
+    val->str = str_ref.sub_str ( opStart, str_ref.offset - opStart);
+    opStart = str_ref.offset;
+  }
+
+  bool has_op (op_code_t op) const
+  {
+    for (unsigned int i = 0; i < get_count (); i++)
+      if (get_value (i).op == op) return true;
+    return false;
+  }
+
+  unsigned get_count () const { return values.length; }
+  const VAL &get_value (unsigned int i)   const { return values[i]; }
+  const VAL &operator [] (unsigned int i) const { return get_value (i); }
+
+  unsigned int       opStart;
+  hb_vector_t<VAL>   values;
+};
+
+template <typename ARG=number_t>
+struct interp_env_t
+{
+  void init (const byte_str_t &str_)
+  {
+    str_ref.reset (str_);
+    argStack.init ();
+    error = false;
+  }
+  void fini () { argStack.fini (); }
+
+  bool in_error () const
+  { return error || str_ref.in_error () || argStack.in_error (); }
+
+  void set_error () { error = true; }
+
+  op_code_t fetch_op ()
+  {
+    op_code_t  op = OpCode_Invalid;
+    if (unlikely (!str_ref.avail ()))
+      return OpCode_Invalid;
+    op = (op_code_t)(unsigned char)str_ref[0];
+    if (op == OpCode_escape) {
+      if (unlikely (!str_ref.avail ()))
+	return OpCode_Invalid;
+      op = Make_OpCode_ESC(str_ref[1]);
+      str_ref.inc ();
+    }
+    str_ref.inc ();
+    return op;
+  }
+
+  const ARG& eval_arg (unsigned int i) { return argStack[i]; }
+
+  ARG& pop_arg () { return argStack.pop (); }
+  void pop_n_args (unsigned int n) { argStack.pop (n); }
+
+  void clear_args () { pop_n_args (argStack.get_count ()); }
+
+  byte_str_ref_t
+		str_ref;
+  arg_stack_t<ARG>
+		argStack;
+  protected:
+  bool		error;
+};
+
+typedef interp_env_t<> num_interp_env_t;
+
+template <typename ARG=number_t>
+struct opset_t
+{
+  static void process_op (op_code_t op, interp_env_t<ARG>& env)
+  {
+    switch (op) {
+      case OpCode_shortint:
+	env.argStack.push_int ((int16_t)((env.str_ref[0] << 8) | env.str_ref[1]));
+	env.str_ref.inc (2);
+	break;
+
+      case OpCode_TwoBytePosInt0: case OpCode_TwoBytePosInt1:
+      case OpCode_TwoBytePosInt2: case OpCode_TwoBytePosInt3:
+	env.argStack.push_int ((int16_t)((op - OpCode_TwoBytePosInt0) * 256 + env.str_ref[0] + 108));
+	env.str_ref.inc ();
+	break;
+
+      case OpCode_TwoByteNegInt0: case OpCode_TwoByteNegInt1:
+      case OpCode_TwoByteNegInt2: case OpCode_TwoByteNegInt3:
+	env.argStack.push_int ((-(int16_t)(op - OpCode_TwoByteNegInt0) * 256 - env.str_ref[0] - 108));
+	env.str_ref.inc ();
+	break;
+
+      default:
+	/* 1-byte integer */
+	if (likely ((OpCode_OneByteIntFirst <= op) && (op <= OpCode_OneByteIntLast)))
+	{
+	  env.argStack.push_int ((int)op - 139);
+	} else {
+	  /* invalid unknown operator */
+	  env.clear_args ();
+	  env.set_error ();
+	}
+	break;
+    }
+  }
+};
+
+template <typename ENV>
+struct interpreter_t
+{
+  ~interpreter_t() { fini (); }
+
+  void fini () { env.fini (); }
+
+  ENV env;
+};
+
+} /* namespace CFF */
+
+#endif /* HB_CFF_INTERP_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cff-interp-cs-common.hh b/thirdparty/harfbuzz/src/hb-cff-interp-cs-common.hh
new file mode 100644
index 0000000000..52d778ffe2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cff-interp-cs-common.hh
@@ -0,0 +1,911 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_CFF_INTERP_CS_COMMON_HH
+#define HB_CFF_INTERP_CS_COMMON_HH
+
+#include "hb.hh"
+#include "hb-cff-interp-common.hh"
+
+namespace CFF {
+
+using namespace OT;
+
+enum cs_type_t {
+  CSType_CharString,
+  CSType_GlobalSubr,
+  CSType_LocalSubr
+};
+
+struct call_context_t
+{
+  void init (const byte_str_ref_t substr_=byte_str_ref_t (), cs_type_t type_=CSType_CharString, unsigned int subr_num_=0)
+  {
+    str_ref = substr_;
+    type = type_;
+    subr_num = subr_num_;
+  }
+
+  void fini () {}
+
+  byte_str_ref_t  str_ref;
+  cs_type_t	  type;
+  unsigned int    subr_num;
+};
+
+/* call stack */
+const unsigned int kMaxCallLimit = 10;
+struct call_stack_t : cff_stack_t<call_context_t, kMaxCallLimit> {};
+
+template <typename SUBRS>
+struct biased_subrs_t
+{
+  void init (const SUBRS *subrs_)
+  {
+    subrs = subrs_;
+    unsigned int  nSubrs = get_count ();
+    if (nSubrs < 1240)
+      bias = 107;
+    else if (nSubrs < 33900)
+      bias = 1131;
+    else
+      bias = 32768;
+  }
+
+  void fini () {}
+
+  unsigned int get_count () const { return subrs ? subrs->count : 0; }
+  unsigned int get_bias () const  { return bias; }
+
+  byte_str_t operator [] (unsigned int index) const
+  {
+    if (unlikely (!subrs || index >= subrs->count))
+      return Null (byte_str_t);
+    else
+      return (*subrs)[index];
+  }
+
+  protected:
+  unsigned int  bias;
+  const SUBRS   *subrs;
+};
+
+struct point_t
+{
+  void init ()
+  {
+    x.init ();
+    y.init ();
+  }
+
+  void set_int (int _x, int _y)
+  {
+    x.set_int (_x);
+    y.set_int (_y);
+  }
+
+  void move_x (const number_t &dx) { x += dx; }
+  void move_y (const number_t &dy) { y += dy; }
+  void move (const number_t &dx, const number_t &dy) { move_x (dx); move_y (dy); }
+  void move (const point_t &d) { move_x (d.x); move_y (d.y); }
+
+  number_t  x;
+  number_t  y;
+};
+
+template <typename ARG, typename SUBRS>
+struct cs_interp_env_t : interp_env_t<ARG>
+{
+  void init (const byte_str_t &str, const SUBRS *globalSubrs_, const SUBRS *localSubrs_)
+  {
+    interp_env_t<ARG>::init (str);
+
+    context.init (str, CSType_CharString);
+    seen_moveto = true;
+    seen_hintmask = false;
+    hstem_count = 0;
+    vstem_count = 0;
+    hintmask_size = 0;
+    pt.init ();
+    callStack.init ();
+    globalSubrs.init (globalSubrs_);
+    localSubrs.init (localSubrs_);
+  }
+  void fini ()
+  {
+    interp_env_t<ARG>::fini ();
+
+    callStack.fini ();
+    globalSubrs.fini ();
+    localSubrs.fini ();
+  }
+
+  bool in_error () const
+  {
+    return callStack.in_error () || SUPER::in_error ();
+  }
+
+  bool pop_subr_num (const biased_subrs_t<SUBRS>& biasedSubrs, unsigned int &subr_num)
+  {
+    subr_num = 0;
+    int n = SUPER::argStack.pop_int ();
+    n += biasedSubrs.get_bias ();
+    if (unlikely ((n < 0) || ((unsigned int)n >= biasedSubrs.get_count ())))
+      return false;
+
+    subr_num = (unsigned int)n;
+    return true;
+  }
+
+  void call_subr (const biased_subrs_t<SUBRS>& biasedSubrs, cs_type_t type)
+  {
+    unsigned int subr_num = 0;
+
+    if (unlikely (!pop_subr_num (biasedSubrs, subr_num)
+		 || callStack.get_count () >= kMaxCallLimit))
+    {
+      SUPER::set_error ();
+      return;
+    }
+    context.str_ref = SUPER::str_ref;
+    callStack.push (context);
+
+    context.init ( biasedSubrs[subr_num], type, subr_num);
+    SUPER::str_ref = context.str_ref;
+  }
+
+  void return_from_subr ()
+  {
+    if (unlikely (SUPER::str_ref.in_error ()))
+      SUPER::set_error ();
+    context = callStack.pop ();
+    SUPER::str_ref = context.str_ref;
+  }
+
+  void determine_hintmask_size ()
+  {
+    if (!seen_hintmask)
+    {
+      vstem_count += SUPER::argStack.get_count() / 2;
+      hintmask_size = (hstem_count + vstem_count + 7) >> 3;
+      seen_hintmask = true;
+    }
+  }
+
+  void set_endchar (bool endchar_flag_) { endchar_flag = endchar_flag_; }
+  bool is_endchar () const { return endchar_flag; }
+
+  const number_t &get_x () const { return pt.x; }
+  const number_t &get_y () const { return pt.y; }
+  const point_t &get_pt () const { return pt; }
+
+  void moveto (const point_t &pt_ ) { pt = pt_; }
+
+  public:
+  call_context_t   context;
+  bool	  endchar_flag;
+  bool	  seen_moveto;
+  bool	  seen_hintmask;
+
+  unsigned int  hstem_count;
+  unsigned int  vstem_count;
+  unsigned int  hintmask_size;
+  call_stack_t	callStack;
+  biased_subrs_t<SUBRS>   globalSubrs;
+  biased_subrs_t<SUBRS>   localSubrs;
+
+  private:
+  point_t	 pt;
+
+  typedef interp_env_t<ARG> SUPER;
+};
+
+template <typename ENV, typename PARAM>
+struct path_procs_null_t
+{
+  static void rmoveto (ENV &env, PARAM& param) {}
+  static void hmoveto (ENV &env, PARAM& param) {}
+  static void vmoveto (ENV &env, PARAM& param) {}
+  static void rlineto (ENV &env, PARAM& param) {}
+  static void hlineto (ENV &env, PARAM& param) {}
+  static void vlineto (ENV &env, PARAM& param) {}
+  static void rrcurveto (ENV &env, PARAM& param) {}
+  static void rcurveline (ENV &env, PARAM& param) {}
+  static void rlinecurve (ENV &env, PARAM& param) {}
+  static void vvcurveto (ENV &env, PARAM& param) {}
+  static void hhcurveto (ENV &env, PARAM& param) {}
+  static void vhcurveto (ENV &env, PARAM& param) {}
+  static void hvcurveto (ENV &env, PARAM& param) {}
+  static void moveto (ENV &env, PARAM& param, const point_t &pt) {}
+  static void line (ENV &env, PARAM& param, const point_t &pt1) {}
+  static void curve (ENV &env, PARAM& param, const point_t &pt1, const point_t &pt2, const point_t &pt3) {}
+  static void hflex (ENV &env, PARAM& param) {}
+  static void flex (ENV &env, PARAM& param) {}
+  static void hflex1 (ENV &env, PARAM& param) {}
+  static void flex1 (ENV &env, PARAM& param) {}
+};
+
+template <typename ARG, typename OPSET, typename ENV, typename PARAM, typename PATH=path_procs_null_t<ENV, PARAM>>
+struct cs_opset_t : opset_t<ARG>
+{
+  static void process_op (op_code_t op, ENV &env, PARAM& param)
+  {
+    switch (op) {
+
+      case OpCode_return:
+	env.return_from_subr ();
+	break;
+      case OpCode_endchar:
+	OPSET::check_width (op, env, param);
+	env.set_endchar (true);
+	OPSET::flush_args_and_op (op, env, param);
+	break;
+
+      case OpCode_fixedcs:
+	env.argStack.push_fixed_from_substr (env.str_ref);
+	break;
+
+      case OpCode_callsubr:
+	env.call_subr (env.localSubrs, CSType_LocalSubr);
+	break;
+
+      case OpCode_callgsubr:
+	env.call_subr (env.globalSubrs, CSType_GlobalSubr);
+	break;
+
+      case OpCode_hstem:
+      case OpCode_hstemhm:
+	OPSET::check_width (op, env, param);
+	OPSET::process_hstem (op, env, param);
+	break;
+      case OpCode_vstem:
+      case OpCode_vstemhm:
+	OPSET::check_width (op, env, param);
+	OPSET::process_vstem (op, env, param);
+	break;
+      case OpCode_hintmask:
+      case OpCode_cntrmask:
+	OPSET::check_width (op, env, param);
+	OPSET::process_hintmask (op, env, param);
+	break;
+      case OpCode_rmoveto:
+	OPSET::check_width (op, env, param);
+	PATH::rmoveto (env, param);
+	OPSET::process_post_move (op, env, param);
+	break;
+      case OpCode_hmoveto:
+	OPSET::check_width (op, env, param);
+	PATH::hmoveto (env, param);
+	OPSET::process_post_move (op, env, param);
+	break;
+      case OpCode_vmoveto:
+	OPSET::check_width (op, env, param);
+	PATH::vmoveto (env, param);
+	OPSET::process_post_move (op, env, param);
+	break;
+      case OpCode_rlineto:
+	PATH::rlineto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_hlineto:
+	PATH::hlineto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_vlineto:
+	PATH::vlineto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_rrcurveto:
+	PATH::rrcurveto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_rcurveline:
+	PATH::rcurveline (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_rlinecurve:
+	PATH::rlinecurve (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_vvcurveto:
+	PATH::vvcurveto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_hhcurveto:
+	PATH::hhcurveto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_vhcurveto:
+	PATH::vhcurveto (env, param);
+	process_post_path (op, env, param);
+	break;
+      case OpCode_hvcurveto:
+	PATH::hvcurveto (env, param);
+	process_post_path (op, env, param);
+	break;
+
+      case OpCode_hflex:
+	PATH::hflex (env, param);
+	OPSET::process_post_flex (op, env, param);
+	break;
+
+      case OpCode_flex:
+	PATH::flex (env, param);
+	OPSET::process_post_flex (op, env, param);
+	break;
+
+      case OpCode_hflex1:
+	PATH::hflex1 (env, param);
+	OPSET::process_post_flex (op, env, param);
+	break;
+
+      case OpCode_flex1:
+	PATH::flex1 (env, param);
+	OPSET::process_post_flex (op, env, param);
+	break;
+
+      default:
+	SUPER::process_op (op, env);
+	break;
+    }
+  }
+
+  static void process_hstem (op_code_t op, ENV &env, PARAM& param)
+  {
+    env.hstem_count += env.argStack.get_count () / 2;
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static void process_vstem (op_code_t op, ENV &env, PARAM& param)
+  {
+    env.vstem_count += env.argStack.get_count () / 2;
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static void process_hintmask (op_code_t op, ENV &env, PARAM& param)
+  {
+    env.determine_hintmask_size ();
+    if (likely (env.str_ref.avail (env.hintmask_size)))
+    {
+      OPSET::flush_hintmask (op, env, param);
+      env.str_ref.inc (env.hintmask_size);
+    }
+  }
+
+  static void process_post_flex (op_code_t op, ENV &env, PARAM& param)
+  {
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static void check_width (op_code_t op, ENV &env, PARAM& param)
+  {}
+
+  static void process_post_move (op_code_t op, ENV &env, PARAM& param)
+  {
+    if (!env.seen_moveto)
+    {
+      env.determine_hintmask_size ();
+      env.seen_moveto = true;
+    }
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static void process_post_path (op_code_t op, ENV &env, PARAM& param)
+  {
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static void flush_args_and_op (op_code_t op, ENV &env, PARAM& param)
+  {
+    OPSET::flush_args (env, param);
+    OPSET::flush_op (op, env, param);
+  }
+
+  static void flush_args (ENV &env, PARAM& param)
+  {
+    env.pop_n_args (env.argStack.get_count ());
+  }
+
+  static void flush_op (op_code_t op, ENV &env, PARAM& param)
+  {
+  }
+
+  static void flush_hintmask (op_code_t op, ENV &env, PARAM& param)
+  {
+    OPSET::flush_args_and_op (op, env, param);
+  }
+
+  static bool is_number_op (op_code_t op)
+  {
+    switch (op)
+    {
+      case OpCode_shortint:
+      case OpCode_fixedcs:
+      case OpCode_TwoBytePosInt0: case OpCode_TwoBytePosInt1:
+      case OpCode_TwoBytePosInt2: case OpCode_TwoBytePosInt3:
+      case OpCode_TwoByteNegInt0: case OpCode_TwoByteNegInt1:
+      case OpCode_TwoByteNegInt2: case OpCode_TwoByteNegInt3:
+	return true;
+
+      default:
+	/* 1-byte integer */
+	return (OpCode_OneByteIntFirst <= op) && (op <= OpCode_OneByteIntLast);
+    }
+  }
+
+  protected:
+  typedef opset_t<ARG>  SUPER;
+};
+
+template <typename PATH, typename ENV, typename PARAM>
+struct path_procs_t
+{
+  static void rmoveto (ENV &env, PARAM& param)
+  {
+    point_t pt1 = env.get_pt ();
+    const number_t &dy = env.pop_arg ();
+    const number_t &dx = env.pop_arg ();
+    pt1.move (dx, dy);
+    PATH::moveto (env, param, pt1);
+  }
+
+  static void hmoveto (ENV &env, PARAM& param)
+  {
+    point_t pt1 = env.get_pt ();
+    pt1.move_x (env.pop_arg ());
+    PATH::moveto (env, param, pt1);
+  }
+
+  static void vmoveto (ENV &env, PARAM& param)
+  {
+    point_t pt1 = env.get_pt ();
+    pt1.move_y (env.pop_arg ());
+    PATH::moveto (env, param, pt1);
+  }
+
+  static void rlineto (ENV &env, PARAM& param)
+  {
+    for (unsigned int i = 0; i + 2 <= env.argStack.get_count (); i += 2)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+      PATH::line (env, param, pt1);
+    }
+  }
+
+  static void hlineto (ENV &env, PARAM& param)
+  {
+    point_t pt1;
+    unsigned int i = 0;
+    for (; i + 2 <= env.argStack.get_count (); i += 2)
+    {
+      pt1 = env.get_pt ();
+      pt1.move_x (env.eval_arg (i));
+      PATH::line (env, param, pt1);
+      pt1.move_y (env.eval_arg (i+1));
+      PATH::line (env, param, pt1);
+    }
+    if (i < env.argStack.get_count ())
+    {
+      pt1 = env.get_pt ();
+      pt1.move_x (env.eval_arg (i));
+      PATH::line (env, param, pt1);
+    }
+  }
+
+  static void vlineto (ENV &env, PARAM& param)
+  {
+    point_t pt1;
+    unsigned int i = 0;
+    for (; i + 2 <= env.argStack.get_count (); i += 2)
+    {
+      pt1 = env.get_pt ();
+      pt1.move_y (env.eval_arg (i));
+      PATH::line (env, param, pt1);
+      pt1.move_x (env.eval_arg (i+1));
+      PATH::line (env, param, pt1);
+    }
+    if (i < env.argStack.get_count ())
+    {
+      pt1 = env.get_pt ();
+      pt1.move_y (env.eval_arg (i));
+      PATH::line (env, param, pt1);
+    }
+  }
+
+  static void rrcurveto (ENV &env, PARAM& param)
+  {
+    for (unsigned int i = 0; i + 6 <= env.argStack.get_count (); i += 6)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+2), env.eval_arg (i+3));
+      point_t pt3 = pt2;
+      pt3.move (env.eval_arg (i+4), env.eval_arg (i+5));
+      PATH::curve (env, param, pt1, pt2, pt3);
+    }
+  }
+
+  static void rcurveline (ENV &env, PARAM& param)
+  {
+    unsigned int arg_count = env.argStack.get_count ();
+    if (unlikely (arg_count < 8))
+      return;
+
+    unsigned int i = 0;
+    unsigned int curve_limit = arg_count - 2;
+    for (; i + 6 <= curve_limit; i += 6)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+2), env.eval_arg (i+3));
+      point_t pt3 = pt2;
+      pt3.move (env.eval_arg (i+4), env.eval_arg (i+5));
+      PATH::curve (env, param, pt1, pt2, pt3);
+    }
+
+    point_t pt1 = env.get_pt ();
+    pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+    PATH::line (env, param, pt1);
+  }
+
+  static void rlinecurve (ENV &env, PARAM& param)
+  {
+    unsigned int arg_count = env.argStack.get_count ();
+    if (unlikely (arg_count < 8))
+      return;
+
+    unsigned int i = 0;
+    unsigned int line_limit = arg_count - 6;
+    for (; i + 2 <= line_limit; i += 2)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+      PATH::line (env, param, pt1);
+    }
+
+    point_t pt1 = env.get_pt ();
+    pt1.move (env.eval_arg (i), env.eval_arg (i+1));
+    point_t pt2 = pt1;
+    pt2.move (env.eval_arg (i+2), env.eval_arg (i+3));
+    point_t pt3 = pt2;
+    pt3.move (env.eval_arg (i+4), env.eval_arg (i+5));
+    PATH::curve (env, param, pt1, pt2, pt3);
+  }
+
+  static void vvcurveto (ENV &env, PARAM& param)
+  {
+    unsigned int i = 0;
+    point_t pt1 = env.get_pt ();
+    if ((env.argStack.get_count () & 1) != 0)
+      pt1.move_x (env.eval_arg (i++));
+    for (; i + 4 <= env.argStack.get_count (); i += 4)
+    {
+      pt1.move_y (env.eval_arg (i));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+      point_t pt3 = pt2;
+      pt3.move_y (env.eval_arg (i+3));
+      PATH::curve (env, param, pt1, pt2, pt3);
+      pt1 = env.get_pt ();
+    }
+  }
+
+  static void hhcurveto (ENV &env, PARAM& param)
+  {
+    unsigned int i = 0;
+    point_t pt1 = env.get_pt ();
+    if ((env.argStack.get_count () & 1) != 0)
+      pt1.move_y (env.eval_arg (i++));
+    for (; i + 4 <= env.argStack.get_count (); i += 4)
+    {
+      pt1.move_x (env.eval_arg (i));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+      point_t pt3 = pt2;
+      pt3.move_x (env.eval_arg (i+3));
+      PATH::curve (env, param, pt1, pt2, pt3);
+      pt1 = env.get_pt ();
+    }
+  }
+
+  static void vhcurveto (ENV &env, PARAM& param)
+  {
+    point_t pt1, pt2, pt3;
+    unsigned int i = 0;
+    if ((env.argStack.get_count () % 8) >= 4)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move_y (env.eval_arg (i));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+      point_t pt3 = pt2;
+      pt3.move_x (env.eval_arg (i+3));
+      i += 4;
+
+      for (; i + 8 <= env.argStack.get_count (); i += 8)
+      {
+	PATH::curve (env, param, pt1, pt2, pt3);
+	pt1 = env.get_pt ();
+	pt1.move_x (env.eval_arg (i));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+	pt3 = pt2;
+	pt3.move_y (env.eval_arg (i+3));
+	PATH::curve (env, param, pt1, pt2, pt3);
+
+	pt1 = pt3;
+	pt1.move_y (env.eval_arg (i+4));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+5), env.eval_arg (i+6));
+	pt3 = pt2;
+	pt3.move_x (env.eval_arg (i+7));
+      }
+      if (i < env.argStack.get_count ())
+	pt3.move_y (env.eval_arg (i));
+      PATH::curve (env, param, pt1, pt2, pt3);
+    }
+    else
+    {
+      for (; i + 8 <= env.argStack.get_count (); i += 8)
+      {
+	pt1 = env.get_pt ();
+	pt1.move_y (env.eval_arg (i));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+	pt3 = pt2;
+	pt3.move_x (env.eval_arg (i+3));
+	PATH::curve (env, param, pt1, pt2, pt3);
+
+	pt1 = pt3;
+	pt1.move_x (env.eval_arg (i+4));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+5), env.eval_arg (i+6));
+	pt3 = pt2;
+	pt3.move_y (env.eval_arg (i+7));
+	if ((env.argStack.get_count () - i < 16) && ((env.argStack.get_count () & 1) != 0))
+	  pt3.move_x (env.eval_arg (i+8));
+	PATH::curve (env, param, pt1, pt2, pt3);
+      }
+    }
+  }
+
+  static void hvcurveto (ENV &env, PARAM& param)
+  {
+    point_t pt1, pt2, pt3;
+    unsigned int i = 0;
+    if ((env.argStack.get_count () % 8) >= 4)
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move_x (env.eval_arg (i));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+      point_t pt3 = pt2;
+      pt3.move_y (env.eval_arg (i+3));
+      i += 4;
+
+      for (; i + 8 <= env.argStack.get_count (); i += 8)
+      {
+	PATH::curve (env, param, pt1, pt2, pt3);
+	pt1 = env.get_pt ();
+	pt1.move_y (env.eval_arg (i));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+	pt3 = pt2;
+	pt3.move_x (env.eval_arg (i+3));
+	PATH::curve (env, param, pt1, pt2, pt3);
+
+	pt1 = pt3;
+	pt1.move_x (env.eval_arg (i+4));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+5), env.eval_arg (i+6));
+	pt3 = pt2;
+	pt3.move_y (env.eval_arg (i+7));
+      }
+      if (i < env.argStack.get_count ())
+	pt3.move_x (env.eval_arg (i));
+      PATH::curve (env, param, pt1, pt2, pt3);
+    }
+    else
+    {
+      for (; i + 8 <= env.argStack.get_count (); i += 8)
+      {
+	pt1 = env.get_pt ();
+	pt1.move_x (env.eval_arg (i));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+1), env.eval_arg (i+2));
+	pt3 = pt2;
+	pt3.move_y (env.eval_arg (i+3));
+	PATH::curve (env, param, pt1, pt2, pt3);
+
+	pt1 = pt3;
+	pt1.move_y (env.eval_arg (i+4));
+	pt2 = pt1;
+	pt2.move (env.eval_arg (i+5), env.eval_arg (i+6));
+	pt3 = pt2;
+	pt3.move_x (env.eval_arg (i+7));
+	if ((env.argStack.get_count () - i < 16) && ((env.argStack.get_count () & 1) != 0))
+	  pt3.move_y (env.eval_arg (i+8));
+	PATH::curve (env, param, pt1, pt2, pt3);
+      }
+    }
+  }
+
+  /* default actions to be overridden */
+  static void moveto (ENV &env, PARAM& param, const point_t &pt)
+  { env.moveto (pt); }
+
+  static void line (ENV &env, PARAM& param, const point_t &pt1)
+  { PATH::moveto (env, param, pt1); }
+
+  static void curve (ENV &env, PARAM& param, const point_t &pt1, const point_t &pt2, const point_t &pt3)
+  { PATH::moveto (env, param, pt3); }
+
+  static void hflex (ENV &env, PARAM& param)
+  {
+    if (likely (env.argStack.get_count () == 7))
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move_x (env.eval_arg (0));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (1), env.eval_arg (2));
+      point_t pt3 = pt2;
+      pt3.move_x (env.eval_arg (3));
+      point_t pt4 = pt3;
+      pt4.move_x (env.eval_arg (4));
+      point_t pt5 = pt4;
+      pt5.move_x (env.eval_arg (5));
+      pt5.y = pt1.y;
+      point_t pt6 = pt5;
+      pt6.move_x (env.eval_arg (6));
+
+      curve2 (env, param, pt1, pt2, pt3, pt4, pt5, pt6);
+    }
+    else
+      env.set_error ();
+  }
+
+  static void flex (ENV &env, PARAM& param)
+  {
+    if (likely (env.argStack.get_count () == 13))
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (0), env.eval_arg (1));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (2), env.eval_arg (3));
+      point_t pt3 = pt2;
+      pt3.move (env.eval_arg (4), env.eval_arg (5));
+      point_t pt4 = pt3;
+      pt4.move (env.eval_arg (6), env.eval_arg (7));
+      point_t pt5 = pt4;
+      pt5.move (env.eval_arg (8), env.eval_arg (9));
+      point_t pt6 = pt5;
+      pt6.move (env.eval_arg (10), env.eval_arg (11));
+
+      curve2 (env, param, pt1, pt2, pt3, pt4, pt5, pt6);
+    }
+    else
+      env.set_error ();
+  }
+
+  static void hflex1 (ENV &env, PARAM& param)
+  {
+    if (likely (env.argStack.get_count () == 9))
+    {
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (0), env.eval_arg (1));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (2), env.eval_arg (3));
+      point_t pt3 = pt2;
+      pt3.move_x (env.eval_arg (4));
+      point_t pt4 = pt3;
+      pt4.move_x (env.eval_arg (5));
+      point_t pt5 = pt4;
+      pt5.move (env.eval_arg (6), env.eval_arg (7));
+      point_t pt6 = pt5;
+      pt6.move_x (env.eval_arg (8));
+      pt6.y = env.get_pt ().y;
+
+      curve2 (env, param, pt1, pt2, pt3, pt4, pt5, pt6);
+    }
+    else
+      env.set_error ();
+  }
+
+  static void flex1 (ENV &env, PARAM& param)
+  {
+    if (likely (env.argStack.get_count () == 11))
+    {
+      point_t d;
+      d.init ();
+      for (unsigned int i = 0; i < 10; i += 2)
+	d.move (env.eval_arg (i), env.eval_arg (i+1));
+
+      point_t pt1 = env.get_pt ();
+      pt1.move (env.eval_arg (0), env.eval_arg (1));
+      point_t pt2 = pt1;
+      pt2.move (env.eval_arg (2), env.eval_arg (3));
+      point_t pt3 = pt2;
+      pt3.move (env.eval_arg (4), env.eval_arg (5));
+      point_t pt4 = pt3;
+      pt4.move (env.eval_arg (6), env.eval_arg (7));
+      point_t pt5 = pt4;
+      pt5.move (env.eval_arg (8), env.eval_arg (9));
+      point_t pt6 = pt5;
+
+      if (fabs (d.x.to_real ()) > fabs (d.y.to_real ()))
+      {
+	pt6.move_x (env.eval_arg (10));
+	pt6.y = env.get_pt ().y;
+      }
+      else
+      {
+	pt6.x = env.get_pt ().x;
+	pt6.move_y (env.eval_arg (10));
+      }
+
+      curve2 (env, param, pt1, pt2, pt3, pt4, pt5, pt6);
+    }
+    else
+      env.set_error ();
+  }
+
+  protected:
+  static void curve2 (ENV &env, PARAM& param,
+		      const point_t &pt1, const point_t &pt2, const point_t &pt3,
+		      const point_t &pt4, const point_t &pt5, const point_t &pt6)
+  {
+    PATH::curve (env, param, pt1, pt2, pt3);
+    PATH::curve (env, param, pt4, pt5, pt6);
+  }
+};
+
+template <typename ENV, typename OPSET, typename PARAM>
+struct cs_interpreter_t : interpreter_t<ENV>
+{
+  bool interpret (PARAM& param)
+  {
+    SUPER::env.set_endchar (false);
+
+    for (;;) {
+      OPSET::process_op (SUPER::env.fetch_op (), SUPER::env, param);
+      if (unlikely (SUPER::env.in_error ()))
+	return false;
+      if (SUPER::env.is_endchar ())
+	break;
+    }
+
+    return true;
+  }
+
+  private:
+  typedef interpreter_t<ENV> SUPER;
+};
+
+} /* namespace CFF */
+
+#endif /* HB_CFF_INTERP_CS_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cff-interp-dict-common.hh b/thirdparty/harfbuzz/src/hb-cff-interp-dict-common.hh
new file mode 100644
index 0000000000..a520ca3bce
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cff-interp-dict-common.hh
@@ -0,0 +1,201 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_CFF_INTERP_DICT_COMMON_HH
+#define HB_CFF_INTERP_DICT_COMMON_HH
+
+#include "hb-cff-interp-common.hh"
+
+namespace CFF {
+
+using namespace OT;
+
+/* an opstr and the parsed out dict value(s) */
+struct dict_val_t : op_str_t
+{
+  void init () { single_val.set_int (0); }
+  void fini () {}
+
+  number_t	      single_val;
+};
+
+typedef dict_val_t num_dict_val_t;
+
+template <typename VAL> struct dict_values_t : parsed_values_t<VAL> {};
+
+template <typename OPSTR=op_str_t>
+struct top_dict_values_t : dict_values_t<OPSTR>
+{
+  void init ()
+  {
+    dict_values_t<OPSTR>::init ();
+    charStringsOffset = 0;
+    FDArrayOffset = 0;
+  }
+  void fini () { dict_values_t<OPSTR>::fini (); }
+
+  unsigned int  charStringsOffset;
+  unsigned int  FDArrayOffset;
+};
+
+struct dict_opset_t : opset_t<number_t>
+{
+  static void process_op (op_code_t op, interp_env_t<number_t>& env)
+  {
+    switch (op) {
+      case OpCode_longintdict:  /* 5-byte integer */
+	env.argStack.push_longint_from_substr (env.str_ref);
+	break;
+
+      case OpCode_BCD:  /* real number */
+	env.argStack.push_real (parse_bcd (env.str_ref));
+	break;
+
+      default:
+	opset_t<number_t>::process_op (op, env);
+	break;
+    }
+  }
+
+  /* Turns CFF's BCD format into strtod understandable string */
+  static double parse_bcd (byte_str_ref_t& str_ref)
+  {
+    if (unlikely (str_ref.in_error ())) return .0;
+
+    enum Nibble { DECIMAL=10, EXP_POS, EXP_NEG, RESERVED, NEG, END };
+
+    char buf[32];
+    unsigned char byte = 0;
+    for (unsigned i = 0, count = 0; count < ARRAY_LENGTH (buf); ++i, ++count)
+    {
+      unsigned nibble;
+      if (!(i & 1))
+      {
+	if (unlikely (!str_ref.avail ())) break;
+
+	byte = str_ref[0];
+	str_ref.inc ();
+	nibble = byte >> 4;
+      }
+      else
+	nibble = byte & 0x0F;
+
+      if (unlikely (nibble == RESERVED)) break;
+      else if (nibble == END)
+      {
+	const char *p = buf;
+	double pv;
+	if (unlikely (!hb_parse_double (&p, p + count, &pv, true/* whole buffer */)))
+	  break;
+	return pv;
+      }
+      else
+      {
+	buf[count] = "0123456789.EE?-?"[nibble];
+	if (nibble == EXP_NEG)
+	{
+	  ++count;
+	  if (unlikely (count == ARRAY_LENGTH (buf))) break;
+	  buf[count] = '-';
+	}
+      }
+    }
+
+    str_ref.set_error ();
+    return .0;
+  }
+
+  static bool is_hint_op (op_code_t op)
+  {
+    switch (op)
+    {
+      case OpCode_BlueValues:
+      case OpCode_OtherBlues:
+      case OpCode_FamilyBlues:
+      case OpCode_FamilyOtherBlues:
+      case OpCode_StemSnapH:
+      case OpCode_StemSnapV:
+      case OpCode_StdHW:
+      case OpCode_StdVW:
+      case OpCode_BlueScale:
+      case OpCode_BlueShift:
+      case OpCode_BlueFuzz:
+      case OpCode_ForceBold:
+      case OpCode_LanguageGroup:
+      case OpCode_ExpansionFactor:
+	return true;
+      default:
+	return false;
+    }
+  }
+};
+
+template <typename VAL=op_str_t>
+struct top_dict_opset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, interp_env_t<number_t>& env, top_dict_values_t<VAL> & dictval)
+  {
+    switch (op) {
+      case OpCode_CharStrings:
+	dictval.charStringsOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_FDArray:
+	dictval.FDArrayOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_FontMatrix:
+	env.clear_args ();
+	break;
+      default:
+	dict_opset_t::process_op (op, env);
+	break;
+    }
+  }
+};
+
+template <typename OPSET, typename PARAM, typename ENV=num_interp_env_t>
+struct dict_interpreter_t : interpreter_t<ENV>
+{
+  bool interpret (PARAM& param)
+  {
+    param.init ();
+    while (SUPER::env.str_ref.avail ())
+    {
+      OPSET::process_op (SUPER::env.fetch_op (), SUPER::env, param);
+      if (unlikely (SUPER::env.in_error ()))
+	return false;
+    }
+
+    return true;
+  }
+
+  private:
+  typedef interpreter_t<ENV> SUPER;
+};
+
+} /* namespace CFF */
+
+#endif /* HB_CFF_INTERP_DICT_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cff1-interp-cs.hh b/thirdparty/harfbuzz/src/hb-cff1-interp-cs.hh
new file mode 100644
index 0000000000..1c8762c172
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cff1-interp-cs.hh
@@ -0,0 +1,161 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_CFF1_INTERP_CS_HH
+#define HB_CFF1_INTERP_CS_HH
+
+#include "hb.hh"
+#include "hb-cff-interp-cs-common.hh"
+
+namespace CFF {
+
+using namespace OT;
+
+typedef biased_subrs_t<CFF1Subrs>   cff1_biased_subrs_t;
+
+struct cff1_cs_interp_env_t : cs_interp_env_t<number_t, CFF1Subrs>
+{
+  template <typename ACC>
+  void init (const byte_str_t &str, ACC &acc, unsigned int fd)
+  {
+    SUPER::init (str, acc.globalSubrs, acc.privateDicts[fd].localSubrs);
+    processed_width = false;
+    has_width = false;
+    arg_start = 0;
+    in_seac = false;
+  }
+
+  void fini () { SUPER::fini (); }
+
+  void set_width (bool has_width_)
+  {
+    if (likely (!processed_width && (SUPER::argStack.get_count () > 0)))
+    {
+      if (has_width_)
+      {
+	width = SUPER::argStack[0];
+	has_width = true;
+	arg_start = 1;
+      }
+    }
+    processed_width = true;
+  }
+
+  void clear_args ()
+  {
+    arg_start = 0;
+    SUPER::clear_args ();
+  }
+
+  void set_in_seac (bool _in_seac) { in_seac = _in_seac; }
+
+  bool	  processed_width;
+  bool	  has_width;
+  unsigned int  arg_start;
+  number_t	width;
+  bool	  in_seac;
+
+  private:
+  typedef cs_interp_env_t<number_t, CFF1Subrs> SUPER;
+};
+
+template <typename OPSET, typename PARAM, typename PATH=path_procs_null_t<cff1_cs_interp_env_t, PARAM>>
+struct cff1_cs_opset_t : cs_opset_t<number_t, OPSET, cff1_cs_interp_env_t, PARAM, PATH>
+{
+  /* PostScript-originated legacy opcodes (OpCode_add etc) are unsupported */
+  /* Type 1-originated deprecated opcodes, seac behavior of endchar and dotsection are supported */
+
+  static void process_op (op_code_t op, cff1_cs_interp_env_t &env, PARAM& param)
+  {
+    switch (op) {
+      case OpCode_dotsection:
+	SUPER::flush_args_and_op (op, env, param);
+	break;
+
+      case OpCode_endchar:
+	OPSET::check_width (op, env, param);
+	if (env.argStack.get_count () >= 4)
+	{
+	  OPSET::process_seac (env, param);
+	}
+	OPSET::flush_args_and_op (op, env, param);
+	env.set_endchar (true);
+	break;
+
+      default:
+	SUPER::process_op (op, env, param);
+    }
+  }
+
+  static void check_width (op_code_t op, cff1_cs_interp_env_t &env, PARAM& param)
+  {
+    if (!env.processed_width)
+    {
+      bool  has_width = false;
+      switch (op)
+      {
+	case OpCode_endchar:
+	case OpCode_hstem:
+	case OpCode_hstemhm:
+	case OpCode_vstem:
+	case OpCode_vstemhm:
+	case OpCode_hintmask:
+	case OpCode_cntrmask:
+	  has_width = ((env.argStack.get_count () & 1) != 0);
+	  break;
+	case OpCode_hmoveto:
+	case OpCode_vmoveto:
+	  has_width = (env.argStack.get_count () > 1);
+	  break;
+	case OpCode_rmoveto:
+	  has_width = (env.argStack.get_count () > 2);
+	  break;
+	default:
+	  return;
+      }
+      env.set_width (has_width);
+    }
+  }
+
+  static void process_seac (cff1_cs_interp_env_t &env, PARAM& param)
+  {
+  }
+
+  static void flush_args (cff1_cs_interp_env_t &env, PARAM& param)
+  {
+    SUPER::flush_args (env, param);
+    env.clear_args ();  /* pop off width */
+  }
+
+  private:
+  typedef cs_opset_t<number_t, OPSET, cff1_cs_interp_env_t, PARAM, PATH>  SUPER;
+};
+
+template <typename OPSET, typename PARAM>
+struct cff1_cs_interpreter_t : cs_interpreter_t<cff1_cs_interp_env_t, OPSET, PARAM> {};
+
+} /* namespace CFF */
+
+#endif /* HB_CFF1_INTERP_CS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-cff2-interp-cs.hh b/thirdparty/harfbuzz/src/hb-cff2-interp-cs.hh
new file mode 100644
index 0000000000..332ece31cd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-cff2-interp-cs.hh
@@ -0,0 +1,272 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_CFF2_INTERP_CS_HH
+#define HB_CFF2_INTERP_CS_HH
+
+#include "hb.hh"
+#include "hb-cff-interp-cs-common.hh"
+
+namespace CFF {
+
+using namespace OT;
+
+struct blend_arg_t : number_t
+{
+  void init ()
+  {
+    number_t::init ();
+    deltas.init ();
+  }
+
+  void fini ()
+  {
+    number_t::fini ();
+    deltas.fini_deep ();
+  }
+
+  void set_int (int v) { reset_blends (); number_t::set_int (v); }
+  void set_fixed (int32_t v) { reset_blends (); number_t::set_fixed (v); }
+  void set_real (double v) { reset_blends (); number_t::set_real (v); }
+
+  void set_blends (unsigned int numValues_, unsigned int valueIndex_,
+		   unsigned int numBlends, hb_array_t<const blend_arg_t> blends_)
+  {
+    numValues = numValues_;
+    valueIndex = valueIndex_;
+    deltas.resize (numBlends);
+    for (unsigned int i = 0; i < numBlends; i++)
+      deltas[i] = blends_[i];
+  }
+
+  bool blending () const { return deltas.length > 0; }
+  void reset_blends ()
+  {
+    numValues = valueIndex = 0;
+    deltas.resize (0);
+  }
+
+  unsigned int numValues;
+  unsigned int valueIndex;
+  hb_vector_t<number_t> deltas;
+};
+
+typedef interp_env_t<blend_arg_t> BlendInterpEnv;
+typedef biased_subrs_t<CFF2Subrs>   cff2_biased_subrs_t;
+
+struct cff2_cs_interp_env_t : cs_interp_env_t<blend_arg_t, CFF2Subrs>
+{
+  template <typename ACC>
+  void init (const byte_str_t &str, ACC &acc, unsigned int fd,
+	     const int *coords_=nullptr, unsigned int num_coords_=0)
+  {
+    SUPER::init (str, acc.globalSubrs, acc.privateDicts[fd].localSubrs);
+
+    coords = coords_;
+    num_coords = num_coords_;
+    varStore = acc.varStore;
+    seen_blend = false;
+    seen_vsindex_ = false;
+    scalars.init ();
+    do_blend = num_coords && coords && varStore->size;
+    set_ivs (acc.privateDicts[fd].ivs);
+  }
+
+  void fini ()
+  {
+    scalars.fini ();
+    SUPER::fini ();
+  }
+
+  op_code_t fetch_op ()
+  {
+    if (this->str_ref.avail ())
+      return SUPER::fetch_op ();
+
+    /* make up return or endchar op */
+    if (this->callStack.is_empty ())
+      return OpCode_endchar;
+    else
+      return OpCode_return;
+  }
+
+  const blend_arg_t& eval_arg (unsigned int i)
+  {
+    blend_arg_t  &arg = argStack[i];
+    blend_arg (arg);
+    return arg;
+  }
+
+  const blend_arg_t& pop_arg ()
+  {
+    blend_arg_t  &arg = argStack.pop ();
+    blend_arg (arg);
+    return arg;
+  }
+
+  void process_blend ()
+  {
+    if (!seen_blend)
+    {
+      region_count = varStore->varStore.get_region_index_count (get_ivs ());
+      if (do_blend)
+      {
+	if (unlikely (!scalars.resize (region_count)))
+	  set_error ();
+	else
+	  varStore->varStore.get_scalars (get_ivs (), coords, num_coords,
+					  &scalars[0], region_count);
+      }
+      seen_blend = true;
+    }
+  }
+
+  void process_vsindex ()
+  {
+    unsigned int  index = argStack.pop_uint ();
+    if (unlikely (seen_vsindex () || seen_blend))
+    {
+      set_error ();
+    }
+    else
+    {
+      set_ivs (index);
+    }
+    seen_vsindex_ = true;
+  }
+
+  unsigned int get_region_count () const { return region_count; }
+  void	 set_region_count (unsigned int region_count_) { region_count = region_count_; }
+  unsigned int get_ivs () const { return ivs; }
+  void	 set_ivs (unsigned int ivs_) { ivs = ivs_; }
+  bool	 seen_vsindex () const { return seen_vsindex_; }
+
+  protected:
+  void blend_arg (blend_arg_t &arg)
+  {
+    if (do_blend && arg.blending ())
+    {
+      if (likely (scalars.length == arg.deltas.length))
+      {
+	double v = arg.to_real ();
+	for (unsigned int i = 0; i < scalars.length; i++)
+	{
+	  v += (double)scalars[i] * arg.deltas[i].to_real ();
+	}
+	arg.set_real (v);
+	arg.deltas.resize (0);
+      }
+    }
+  }
+
+  protected:
+  const int     *coords;
+  unsigned int  num_coords;
+  const	 CFF2VariationStore *varStore;
+  unsigned int  region_count;
+  unsigned int  ivs;
+  hb_vector_t<float>  scalars;
+  bool	  do_blend;
+  bool	  seen_vsindex_;
+  bool	  seen_blend;
+
+  typedef cs_interp_env_t<blend_arg_t, CFF2Subrs> SUPER;
+};
+template <typename OPSET, typename PARAM, typename PATH=path_procs_null_t<cff2_cs_interp_env_t, PARAM>>
+struct cff2_cs_opset_t : cs_opset_t<blend_arg_t, OPSET, cff2_cs_interp_env_t, PARAM, PATH>
+{
+  static void process_op (op_code_t op, cff2_cs_interp_env_t &env, PARAM& param)
+  {
+    switch (op) {
+      case OpCode_callsubr:
+      case OpCode_callgsubr:
+	/* a subroutine number shoudln't be a blended value */
+	if (unlikely (env.argStack.peek ().blending ()))
+	{
+	  env.set_error ();
+	  break;
+	}
+	SUPER::process_op (op, env, param);
+	break;
+
+      case OpCode_blendcs:
+	OPSET::process_blend (env, param);
+	break;
+
+      case OpCode_vsindexcs:
+	if (unlikely (env.argStack.peek ().blending ()))
+	{
+	  env.set_error ();
+	  break;
+	}
+	OPSET::process_vsindex (env, param);
+	break;
+
+      default:
+	SUPER::process_op (op, env, param);
+    }
+  }
+
+  static void process_blend (cff2_cs_interp_env_t &env, PARAM& param)
+  {
+    unsigned int n, k;
+
+    env.process_blend ();
+    k = env.get_region_count ();
+    n = env.argStack.pop_uint ();
+    /* copy the blend values into blend array of the default values */
+    unsigned int start = env.argStack.get_count () - ((k+1) * n);
+    /* let an obvious error case fail, but note CFF2 spec doesn't forbid n==0 */
+    if (unlikely (start > env.argStack.get_count ()))
+    {
+      env.set_error ();
+      return;
+    }
+    for (unsigned int i = 0; i < n; i++)
+    {
+      const hb_array_t<const blend_arg_t>	blends = env.argStack.get_subarray (start + n + (i * k));
+      env.argStack[start + i].set_blends (n, i, k, blends);
+    }
+
+    /* pop off blend values leaving default values now adorned with blend values */
+    env.argStack.pop (k * n);
+  }
+
+  static void process_vsindex (cff2_cs_interp_env_t &env, PARAM& param)
+  {
+    env.process_vsindex ();
+    env.clear_args ();
+  }
+
+  private:
+  typedef cs_opset_t<blend_arg_t, OPSET, cff2_cs_interp_env_t, PARAM, PATH>  SUPER;
+};
+
+template <typename OPSET, typename PARAM>
+struct cff2_cs_interpreter_t : cs_interpreter_t<cff2_cs_interp_env_t, OPSET, PARAM> {};
+
+} /* namespace CFF */
+
+#endif /* HB_CFF2_INTERP_CS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-common.cc b/thirdparty/harfbuzz/src/hb-common.cc
new file mode 100644
index 0000000000..5acfa78431
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-common.cc
@@ -0,0 +1,1098 @@
+/*
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+#include "hb-machinery.hh"
+
+#include <locale.h>
+
+#ifdef HB_NO_SETLOCALE
+#define setlocale(Category, Locale) "C"
+#endif
+
+/**
+ * SECTION:hb-common
+ * @title: hb-common
+ * @short_description: Common data types
+ * @include: hb.h
+ *
+ * Common data types used across HarfBuzz are defined here.
+ **/
+
+
+/* hb_options_t */
+
+hb_atomic_int_t _hb_options;
+
+void
+_hb_options_init ()
+{
+  hb_options_union_t u;
+  u.i = 0;
+  u.opts.initialized = true;
+
+  const char *c = getenv ("HB_OPTIONS");
+  if (c)
+  {
+    while (*c)
+    {
+      const char *p = strchr (c, ':');
+      if (!p)
+	p = c + strlen (c);
+
+#define OPTION(name, symbol) \
+	if (0 == strncmp (c, name, p - c) && strlen (name) == static_cast<size_t>(p - c)) do { u.opts.symbol = true; } while (0)
+
+      OPTION ("uniscribe-bug-compatible", uniscribe_bug_compatible);
+
+#undef OPTION
+
+      c = *p ? p + 1 : p;
+    }
+
+  }
+
+  /* This is idempotent and threadsafe. */
+  _hb_options.set_relaxed (u.i);
+}
+
+
+/* hb_tag_t */
+
+/**
+ * hb_tag_from_string:
+ * @str: (array length=len) (element-type uint8_t):
+ * @len:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_tag_t
+hb_tag_from_string (const char *str, int len)
+{
+  char tag[4];
+  unsigned int i;
+
+  if (!str || !len || !*str)
+    return HB_TAG_NONE;
+
+  if (len < 0 || len > 4)
+    len = 4;
+  for (i = 0; i < (unsigned) len && str[i]; i++)
+    tag[i] = str[i];
+  for (; i < 4; i++)
+    tag[i] = ' ';
+
+  return HB_TAG (tag[0], tag[1], tag[2], tag[3]);
+}
+
+/**
+ * hb_tag_to_string:
+ * @tag:
+ * @buf: (out caller-allocates) (array fixed-size=4) (element-type uint8_t):
+ *
+ *
+ *
+ * Since: 0.9.5
+ **/
+void
+hb_tag_to_string (hb_tag_t tag, char *buf)
+{
+  buf[0] = (char) (uint8_t) (tag >> 24);
+  buf[1] = (char) (uint8_t) (tag >> 16);
+  buf[2] = (char) (uint8_t) (tag >>  8);
+  buf[3] = (char) (uint8_t) (tag >>  0);
+}
+
+
+/* hb_direction_t */
+
+const char direction_strings[][4] = {
+  "ltr",
+  "rtl",
+  "ttb",
+  "btt"
+};
+
+/**
+ * hb_direction_from_string:
+ * @str: (array length=len) (element-type uint8_t):
+ * @len:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_direction_t
+hb_direction_from_string (const char *str, int len)
+{
+  if (unlikely (!str || !len || !*str))
+    return HB_DIRECTION_INVALID;
+
+  /* Lets match loosely: just match the first letter, such that
+   * all of "ltr", "left-to-right", etc work!
+   */
+  char c = TOLOWER (str[0]);
+  for (unsigned int i = 0; i < ARRAY_LENGTH (direction_strings); i++)
+    if (c == direction_strings[i][0])
+      return (hb_direction_t) (HB_DIRECTION_LTR + i);
+
+  return HB_DIRECTION_INVALID;
+}
+
+/**
+ * hb_direction_to_string:
+ * @direction:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+const char *
+hb_direction_to_string (hb_direction_t direction)
+{
+  if (likely ((unsigned int) (direction - HB_DIRECTION_LTR)
+	      < ARRAY_LENGTH (direction_strings)))
+    return direction_strings[direction - HB_DIRECTION_LTR];
+
+  return "invalid";
+}
+
+
+/* hb_language_t */
+
+struct hb_language_impl_t {
+  const char s[1];
+};
+
+static const char canon_map[256] = {
+   0,   0,   0,   0,   0,   0,   0,   0,    0,   0,   0,   0,   0,   0,   0,   0,
+   0,   0,   0,   0,   0,   0,   0,   0,    0,   0,   0,   0,   0,   0,   0,   0,
+   0,   0,   0,   0,   0,   0,   0,   0,    0,   0,   0,   0,   0,  '-',  0,   0,
+  '0', '1', '2', '3', '4', '5', '6', '7',  '8', '9',  0,   0,   0,   0,   0,   0,
+   0,  'a', 'b', 'c', 'd', 'e', 'f', 'g',  'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
+  'p', 'q', 'r', 's', 't', 'u', 'v', 'w',  'x', 'y', 'z',  0,   0,   0,   0,  '-',
+   0,  'a', 'b', 'c', 'd', 'e', 'f', 'g',  'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
+  'p', 'q', 'r', 's', 't', 'u', 'v', 'w',  'x', 'y', 'z',  0,   0,   0,   0,   0
+};
+
+static bool
+lang_equal (hb_language_t  v1,
+	    const void    *v2)
+{
+  const unsigned char *p1 = (const unsigned char *) v1;
+  const unsigned char *p2 = (const unsigned char *) v2;
+
+  while (*p1 && *p1 == canon_map[*p2]) {
+    p1++;
+    p2++;
+  }
+
+  return *p1 == canon_map[*p2];
+}
+
+#if 0
+static unsigned int
+lang_hash (const void *key)
+{
+  const unsigned char *p = key;
+  unsigned int h = 0;
+  while (canon_map[*p])
+    {
+      h = (h << 5) - h + canon_map[*p];
+      p++;
+    }
+
+  return h;
+}
+#endif
+
+
+struct hb_language_item_t {
+
+  struct hb_language_item_t *next;
+  hb_language_t lang;
+
+  bool operator == (const char *s) const
+  { return lang_equal (lang, s); }
+
+  hb_language_item_t & operator = (const char *s) {
+    /* If a custom allocated is used calling strdup() pairs
+    badly with a call to the custom free() in fini() below.
+    Therefore don't call strdup(), implement its behavior.
+    */
+    size_t len = strlen(s) + 1;
+    lang = (hb_language_t) malloc(len);
+    if (likely (lang))
+    {
+      memcpy((unsigned char *) lang, s, len);
+      for (unsigned char *p = (unsigned char *) lang; *p; p++)
+	*p = canon_map[*p];
+    }
+
+    return *this;
+  }
+
+  void fini () { free ((void *) lang); }
+};
+
+
+/* Thread-safe lock-free language list */
+
+static hb_atomic_ptr_t <hb_language_item_t> langs;
+
+#if HB_USE_ATEXIT
+static void
+free_langs ()
+{
+retry:
+  hb_language_item_t *first_lang = langs;
+  if (unlikely (!langs.cmpexch (first_lang, nullptr)))
+    goto retry;
+
+  while (first_lang) {
+    hb_language_item_t *next = first_lang->next;
+    first_lang->fini ();
+    free (first_lang);
+    first_lang = next;
+  }
+}
+#endif
+
+static hb_language_item_t *
+lang_find_or_insert (const char *key)
+{
+retry:
+  hb_language_item_t *first_lang = langs;
+
+  for (hb_language_item_t *lang = first_lang; lang; lang = lang->next)
+    if (*lang == key)
+      return lang;
+
+  /* Not found; allocate one. */
+  hb_language_item_t *lang = (hb_language_item_t *) calloc (1, sizeof (hb_language_item_t));
+  if (unlikely (!lang))
+    return nullptr;
+  lang->next = first_lang;
+  *lang = key;
+  if (unlikely (!lang->lang))
+  {
+    free (lang);
+    return nullptr;
+  }
+
+  if (unlikely (!langs.cmpexch (first_lang, lang)))
+  {
+    lang->fini ();
+    free (lang);
+    goto retry;
+  }
+
+#if HB_USE_ATEXIT
+  if (!first_lang)
+    atexit (free_langs); /* First person registers atexit() callback. */
+#endif
+
+  return lang;
+}
+
+
+/**
+ * hb_language_from_string:
+ * @str: (array length=len) (element-type uint8_t): a string representing
+ *       a BCP 47 language tag
+ * @len: length of the @str, or -1 if it is %NULL-terminated.
+ *
+ * Converts @str representing a BCP 47 language tag to the corresponding
+ * #hb_language_t.
+ *
+ * Return value: (transfer none):
+ * The #hb_language_t corresponding to the BCP 47 language tag.
+ *
+ * Since: 0.9.2
+ **/
+hb_language_t
+hb_language_from_string (const char *str, int len)
+{
+  if (!str || !len || !*str)
+    return HB_LANGUAGE_INVALID;
+
+  hb_language_item_t *item = nullptr;
+  if (len >= 0)
+  {
+    /* NUL-terminate it. */
+    char strbuf[64];
+    len = hb_min (len, (int) sizeof (strbuf) - 1);
+    memcpy (strbuf, str, len);
+    strbuf[len] = '\0';
+    item = lang_find_or_insert (strbuf);
+  }
+  else
+    item = lang_find_or_insert (str);
+
+  return likely (item) ? item->lang : HB_LANGUAGE_INVALID;
+}
+
+/**
+ * hb_language_to_string:
+ * @language: an #hb_language_t to convert.
+ *
+ * See hb_language_from_string().
+ *
+ * Return value: (transfer none):
+ * A %NULL-terminated string representing the @language. Must not be freed by
+ * the caller.
+ *
+ * Since: 0.9.2
+ **/
+const char *
+hb_language_to_string (hb_language_t language)
+{
+  if (unlikely (!language)) return nullptr;
+
+  return language->s;
+}
+
+/**
+ * hb_language_get_default:
+ *
+ * Get default language from current locale.
+ *
+ * Note that the first time this function is called, it calls
+ * "setlocale (LC_CTYPE, nullptr)" to fetch current locale.  The underlying
+ * setlocale function is, in many implementations, NOT threadsafe.  To avoid
+ * problems, call this function once before multiple threads can call it.
+ * This function is only used from hb_buffer_guess_segment_properties() by
+ * HarfBuzz itself.
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+hb_language_t
+hb_language_get_default ()
+{
+  static hb_atomic_ptr_t <hb_language_t> default_language;
+
+  hb_language_t language = default_language;
+  if (unlikely (language == HB_LANGUAGE_INVALID))
+  {
+    language = hb_language_from_string (setlocale (LC_CTYPE, nullptr), -1);
+    (void) default_language.cmpexch (HB_LANGUAGE_INVALID, language);
+  }
+
+  return language;
+}
+
+
+/* hb_script_t */
+
+/**
+ * hb_script_from_iso15924_tag:
+ * @tag: an #hb_tag_t representing an ISO 15924 tag.
+ *
+ * Converts an ISO 15924 script tag to a corresponding #hb_script_t.
+ *
+ * Return value:
+ * An #hb_script_t corresponding to the ISO 15924 tag.
+ *
+ * Since: 0.9.2
+ **/
+hb_script_t
+hb_script_from_iso15924_tag (hb_tag_t tag)
+{
+  if (unlikely (tag == HB_TAG_NONE))
+    return HB_SCRIPT_INVALID;
+
+  /* Be lenient, adjust case (one capital letter followed by three small letters) */
+  tag = (tag & 0xDFDFDFDFu) | 0x00202020u;
+
+  switch (tag) {
+
+    /* These graduated from the 'Q' private-area codes, but
+     * the old code is still aliased by Unicode, and the Qaai
+     * one in use by ICU. */
+    case HB_TAG('Q','a','a','i'): return HB_SCRIPT_INHERITED;
+    case HB_TAG('Q','a','a','c'): return HB_SCRIPT_COPTIC;
+
+    /* Script variants from https://unicode.org/iso15924/ */
+    case HB_TAG('C','y','r','s'): return HB_SCRIPT_CYRILLIC;
+    case HB_TAG('L','a','t','f'): return HB_SCRIPT_LATIN;
+    case HB_TAG('L','a','t','g'): return HB_SCRIPT_LATIN;
+    case HB_TAG('S','y','r','e'): return HB_SCRIPT_SYRIAC;
+    case HB_TAG('S','y','r','j'): return HB_SCRIPT_SYRIAC;
+    case HB_TAG('S','y','r','n'): return HB_SCRIPT_SYRIAC;
+  }
+
+  /* If it looks right, just use the tag as a script */
+  if (((uint32_t) tag & 0xE0E0E0E0u) == 0x40606060u)
+    return (hb_script_t) tag;
+
+  /* Otherwise, return unknown */
+  return HB_SCRIPT_UNKNOWN;
+}
+
+/**
+ * hb_script_from_string:
+ * @str: (array length=len) (element-type uint8_t): a string representing an
+ *       ISO 15924 tag.
+ * @len: length of the @str, or -1 if it is %NULL-terminated.
+ *
+ * Converts a string @str representing an ISO 15924 script tag to a
+ * corresponding #hb_script_t. Shorthand for hb_tag_from_string() then
+ * hb_script_from_iso15924_tag().
+ *
+ * Return value:
+ * An #hb_script_t corresponding to the ISO 15924 tag.
+ *
+ * Since: 0.9.2
+ **/
+hb_script_t
+hb_script_from_string (const char *str, int len)
+{
+  return hb_script_from_iso15924_tag (hb_tag_from_string (str, len));
+}
+
+/**
+ * hb_script_to_iso15924_tag:
+ * @script: an #hb_script_t to convert.
+ *
+ * See hb_script_from_iso15924_tag().
+ *
+ * Return value:
+ * An #hb_tag_t representing an ISO 15924 script tag.
+ *
+ * Since: 0.9.2
+ **/
+hb_tag_t
+hb_script_to_iso15924_tag (hb_script_t script)
+{
+  return (hb_tag_t) script;
+}
+
+/**
+ * hb_script_get_horizontal_direction:
+ * @script:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_direction_t
+hb_script_get_horizontal_direction (hb_script_t script)
+{
+  /* https://docs.google.com/spreadsheets/d/1Y90M0Ie3MUJ6UVCRDOypOtijlMDLNNyyLk36T6iMu0o */
+  switch ((hb_tag_t) script)
+  {
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_ARABIC:
+    case HB_SCRIPT_HEBREW:
+
+    /* Unicode-3.0 additions */
+    case HB_SCRIPT_SYRIAC:
+    case HB_SCRIPT_THAANA:
+
+    /* Unicode-4.0 additions */
+    case HB_SCRIPT_CYPRIOT:
+
+    /* Unicode-4.1 additions */
+    case HB_SCRIPT_KHAROSHTHI:
+
+    /* Unicode-5.0 additions */
+    case HB_SCRIPT_PHOENICIAN:
+    case HB_SCRIPT_NKO:
+
+    /* Unicode-5.1 additions */
+    case HB_SCRIPT_LYDIAN:
+
+    /* Unicode-5.2 additions */
+    case HB_SCRIPT_AVESTAN:
+    case HB_SCRIPT_IMPERIAL_ARAMAIC:
+    case HB_SCRIPT_INSCRIPTIONAL_PAHLAVI:
+    case HB_SCRIPT_INSCRIPTIONAL_PARTHIAN:
+    case HB_SCRIPT_OLD_SOUTH_ARABIAN:
+    case HB_SCRIPT_OLD_TURKIC:
+    case HB_SCRIPT_SAMARITAN:
+
+    /* Unicode-6.0 additions */
+    case HB_SCRIPT_MANDAIC:
+
+    /* Unicode-6.1 additions */
+    case HB_SCRIPT_MEROITIC_CURSIVE:
+    case HB_SCRIPT_MEROITIC_HIEROGLYPHS:
+
+    /* Unicode-7.0 additions */
+    case HB_SCRIPT_MANICHAEAN:
+    case HB_SCRIPT_MENDE_KIKAKUI:
+    case HB_SCRIPT_NABATAEAN:
+    case HB_SCRIPT_OLD_NORTH_ARABIAN:
+    case HB_SCRIPT_PALMYRENE:
+    case HB_SCRIPT_PSALTER_PAHLAVI:
+
+    /* Unicode-8.0 additions */
+    case HB_SCRIPT_HATRAN:
+
+    /* Unicode-9.0 additions */
+    case HB_SCRIPT_ADLAM:
+
+    /* Unicode-11.0 additions */
+    case HB_SCRIPT_HANIFI_ROHINGYA:
+    case HB_SCRIPT_OLD_SOGDIAN:
+    case HB_SCRIPT_SOGDIAN:
+
+    /* Unicode-12.0 additions */
+    case HB_SCRIPT_ELYMAIC:
+
+    /* Unicode-13.0 additions */
+    case HB_SCRIPT_CHORASMIAN:
+    case HB_SCRIPT_YEZIDI:
+
+      return HB_DIRECTION_RTL;
+
+
+    /* https://github.com/harfbuzz/harfbuzz/issues/1000 */
+    case HB_SCRIPT_OLD_HUNGARIAN:
+    case HB_SCRIPT_OLD_ITALIC:
+    case HB_SCRIPT_RUNIC:
+
+      return HB_DIRECTION_INVALID;
+  }
+
+  return HB_DIRECTION_LTR;
+}
+
+
+/* hb_version */
+
+
+/**
+ * SECTION:hb-version
+ * @title: hb-version
+ * @short_description: Information about the version of HarfBuzz in use
+ * @include: hb.h
+ *
+ * These functions and macros allow accessing version of the HarfBuzz
+ * library used at compile- as well as run-time, and to direct code
+ * conditionally based on those versions, again, at compile- or run-time.
+ **/
+
+
+/**
+ * hb_version:
+ * @major: (out): Library major version component.
+ * @minor: (out): Library minor version component.
+ * @micro: (out): Library micro version component.
+ *
+ * Returns library version as three integer components.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_version (unsigned int *major,
+	    unsigned int *minor,
+	    unsigned int *micro)
+{
+  *major = HB_VERSION_MAJOR;
+  *minor = HB_VERSION_MINOR;
+  *micro = HB_VERSION_MICRO;
+}
+
+/**
+ * hb_version_string:
+ *
+ * Returns library version as a string with three components.
+ *
+ * Return value: library version string.
+ *
+ * Since: 0.9.2
+ **/
+const char *
+hb_version_string ()
+{
+  return HB_VERSION_STRING;
+}
+
+/**
+ * hb_version_atleast:
+ * @major:
+ * @minor:
+ * @micro:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.30
+ **/
+hb_bool_t
+hb_version_atleast (unsigned int major,
+		    unsigned int minor,
+		    unsigned int micro)
+{
+  return HB_VERSION_ATLEAST (major, minor, micro);
+}
+
+
+
+/* hb_feature_t and hb_variation_t */
+
+static bool
+parse_space (const char **pp, const char *end)
+{
+  while (*pp < end && ISSPACE (**pp))
+    (*pp)++;
+  return true;
+}
+
+static bool
+parse_char (const char **pp, const char *end, char c)
+{
+  parse_space (pp, end);
+
+  if (*pp == end || **pp != c)
+    return false;
+
+  (*pp)++;
+  return true;
+}
+
+static bool
+parse_uint (const char **pp, const char *end, unsigned int *pv)
+{
+  /* Intentionally use hb_parse_int inside instead of hb_parse_uint,
+   * such that -1 turns into "big number"... */
+  int v;
+  if (unlikely (!hb_parse_int (pp, end, &v))) return false;
+
+  *pv = v;
+  return true;
+}
+
+static bool
+parse_uint32 (const char **pp, const char *end, uint32_t *pv)
+{
+  /* Intentionally use hb_parse_int inside instead of hb_parse_uint,
+   * such that -1 turns into "big number"... */
+  int v;
+  if (unlikely (!hb_parse_int (pp, end, &v))) return false;
+
+  *pv = v;
+  return true;
+}
+
+static bool
+parse_bool (const char **pp, const char *end, uint32_t *pv)
+{
+  parse_space (pp, end);
+
+  const char *p = *pp;
+  while (*pp < end && ISALPHA(**pp))
+    (*pp)++;
+
+  /* CSS allows on/off as aliases 1/0. */
+  if (*pp - p == 2
+      && TOLOWER (p[0]) == 'o'
+      && TOLOWER (p[1]) == 'n')
+    *pv = 1;
+  else if (*pp - p == 3
+	   && TOLOWER (p[0]) == 'o'
+	   && TOLOWER (p[1]) == 'f'
+	   && TOLOWER (p[2]) == 'f')
+    *pv = 0;
+  else
+    return false;
+
+  return true;
+}
+
+/* hb_feature_t */
+
+static bool
+parse_feature_value_prefix (const char **pp, const char *end, hb_feature_t *feature)
+{
+  if (parse_char (pp, end, '-'))
+    feature->value = 0;
+  else {
+    parse_char (pp, end, '+');
+    feature->value = 1;
+  }
+
+  return true;
+}
+
+static bool
+parse_tag (const char **pp, const char *end, hb_tag_t *tag)
+{
+  parse_space (pp, end);
+
+  char quote = 0;
+
+  if (*pp < end && (**pp == '\'' || **pp == '"'))
+  {
+    quote = **pp;
+    (*pp)++;
+  }
+
+  const char *p = *pp;
+  while (*pp < end && (ISALNUM(**pp) || **pp == '_'))
+    (*pp)++;
+
+  if (p == *pp || *pp - p > 4)
+    return false;
+
+  *tag = hb_tag_from_string (p, *pp - p);
+
+  if (quote)
+  {
+    /* CSS expects exactly four bytes.  And we only allow quotations for
+     * CSS compatibility.  So, enforce the length. */
+     if (*pp - p != 4)
+       return false;
+    if (*pp == end || **pp != quote)
+      return false;
+    (*pp)++;
+  }
+
+  return true;
+}
+
+static bool
+parse_feature_indices (const char **pp, const char *end, hb_feature_t *feature)
+{
+  parse_space (pp, end);
+
+  bool has_start;
+
+  feature->start = HB_FEATURE_GLOBAL_START;
+  feature->end = HB_FEATURE_GLOBAL_END;
+
+  if (!parse_char (pp, end, '['))
+    return true;
+
+  has_start = parse_uint (pp, end, &feature->start);
+
+  if (parse_char (pp, end, ':') || parse_char (pp, end, ';')) {
+    parse_uint (pp, end, &feature->end);
+  } else {
+    if (has_start)
+      feature->end = feature->start + 1;
+  }
+
+  return parse_char (pp, end, ']');
+}
+
+static bool
+parse_feature_value_postfix (const char **pp, const char *end, hb_feature_t *feature)
+{
+  bool had_equal = parse_char (pp, end, '=');
+  bool had_value = parse_uint32 (pp, end, &feature->value) ||
+		   parse_bool (pp, end, &feature->value);
+  /* CSS doesn't use equal-sign between tag and value.
+   * If there was an equal-sign, then there *must* be a value.
+   * A value without an equal-sign is ok, but not required. */
+  return !had_equal || had_value;
+}
+
+static bool
+parse_one_feature (const char **pp, const char *end, hb_feature_t *feature)
+{
+  return parse_feature_value_prefix (pp, end, feature) &&
+	 parse_tag (pp, end, &feature->tag) &&
+	 parse_feature_indices (pp, end, feature) &&
+	 parse_feature_value_postfix (pp, end, feature) &&
+	 parse_space (pp, end) &&
+	 *pp == end;
+}
+
+/**
+ * hb_feature_from_string:
+ * @str: (array length=len) (element-type uint8_t): a string to parse
+ * @len: length of @str, or -1 if string is %NULL terminated
+ * @feature: (out): the #hb_feature_t to initialize with the parsed values
+ *
+ * Parses a string into a #hb_feature_t.
+ *
+ * The format for specifying feature strings follows. All valid CSS
+ * font-feature-settings values other than 'normal' and the global values are
+ * also accepted, though not documented below. CSS string escapes are not
+ * supported.
+ *
+ * The range indices refer to the positions between Unicode characters. The
+ * position before the first character is always 0.
+ *
+ * The format is Python-esque.  Here is how it all works:
+ *
+ * <informaltable pgwide='1' align='left' frame='none'>
+ * <tgroup cols='5'>
+ * <thead>
+ * <row><entry>Syntax</entry>    <entry>Value</entry> <entry>Start</entry> <entry>End</entry></row>
+ * </thead>
+ * <tbody>
+ * <row><entry>Setting value:</entry></row>
+ * <row><entry>kern</entry>      <entry>1</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature on</entry></row>
+ * <row><entry>+kern</entry>     <entry>1</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature on</entry></row>
+ * <row><entry>-kern</entry>     <entry>0</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature off</entry></row>
+ * <row><entry>kern=0</entry>    <entry>0</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature off</entry></row>
+ * <row><entry>kern=1</entry>    <entry>1</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature on</entry></row>
+ * <row><entry>aalt=2</entry>    <entry>2</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Choose 2nd alternate</entry></row>
+ * <row><entry>Setting index:</entry></row>
+ * <row><entry>kern[]</entry>    <entry>1</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature on</entry></row>
+ * <row><entry>kern[:]</entry>   <entry>1</entry>     <entry>0</entry>      <entry>∞</entry>   <entry>Turn feature on</entry></row>
+ * <row><entry>kern[5:]</entry>  <entry>1</entry>     <entry>5</entry>      <entry>∞</entry>   <entry>Turn feature on, partial</entry></row>
+ * <row><entry>kern[:5]</entry>  <entry>1</entry>     <entry>0</entry>      <entry>5</entry>   <entry>Turn feature on, partial</entry></row>
+ * <row><entry>kern[3:5]</entry> <entry>1</entry>     <entry>3</entry>      <entry>5</entry>   <entry>Turn feature on, range</entry></row>
+ * <row><entry>kern[3]</entry>   <entry>1</entry>     <entry>3</entry>      <entry>3+1</entry> <entry>Turn feature on, single char</entry></row>
+ * <row><entry>Mixing it all:</entry></row>
+ * <row><entry>aalt[3:5]=2</entry> <entry>2</entry>   <entry>3</entry>      <entry>5</entry>   <entry>Turn 2nd alternate on for range</entry></row>
+ * </tbody>
+ * </tgroup>
+ * </informaltable>
+ *
+ * Return value:
+ * %true if @str is successfully parsed, %false otherwise.
+ *
+ * Since: 0.9.5
+ **/
+hb_bool_t
+hb_feature_from_string (const char *str, int len,
+			hb_feature_t *feature)
+{
+  hb_feature_t feat;
+
+  if (len < 0)
+    len = strlen (str);
+
+  if (likely (parse_one_feature (&str, str + len, &feat)))
+  {
+    if (feature)
+      *feature = feat;
+    return true;
+  }
+
+  if (feature)
+    memset (feature, 0, sizeof (*feature));
+  return false;
+}
+
+/**
+ * hb_feature_to_string:
+ * @feature: an #hb_feature_t to convert
+ * @buf: (array length=size) (out): output string
+ * @size: the allocated size of @buf
+ *
+ * Converts a #hb_feature_t into a %NULL-terminated string in the format
+ * understood by hb_feature_from_string(). The client in responsible for
+ * allocating big enough size for @buf, 128 bytes is more than enough.
+ *
+ * Since: 0.9.5
+ **/
+void
+hb_feature_to_string (hb_feature_t *feature,
+		      char *buf, unsigned int size)
+{
+  if (unlikely (!size)) return;
+
+  char s[128];
+  unsigned int len = 0;
+  if (feature->value == 0)
+    s[len++] = '-';
+  hb_tag_to_string (feature->tag, s + len);
+  len += 4;
+  while (len && s[len - 1] == ' ')
+    len--;
+  if (feature->start != HB_FEATURE_GLOBAL_START || feature->end != HB_FEATURE_GLOBAL_END)
+  {
+    s[len++] = '[';
+    if (feature->start)
+      len += hb_max (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u", feature->start));
+    if (feature->end != feature->start + 1) {
+      s[len++] = ':';
+      if (feature->end != HB_FEATURE_GLOBAL_END)
+	len += hb_max (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u", feature->end));
+    }
+    s[len++] = ']';
+  }
+  if (feature->value > 1)
+  {
+    s[len++] = '=';
+    len += hb_max (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u", feature->value));
+  }
+  assert (len < ARRAY_LENGTH (s));
+  len = hb_min (len, size - 1);
+  memcpy (buf, s, len);
+  buf[len] = '\0';
+}
+
+/* hb_variation_t */
+
+static bool
+parse_variation_value (const char **pp, const char *end, hb_variation_t *variation)
+{
+  parse_char (pp, end, '='); /* Optional. */
+  double v;
+  if (unlikely (!hb_parse_double (pp, end, &v))) return false;
+
+  variation->value = v;
+  return true;
+}
+
+static bool
+parse_one_variation (const char **pp, const char *end, hb_variation_t *variation)
+{
+  return parse_tag (pp, end, &variation->tag) &&
+	 parse_variation_value (pp, end, variation) &&
+	 parse_space (pp, end) &&
+	 *pp == end;
+}
+
+/**
+ * hb_variation_from_string:
+ *
+ * Since: 1.4.2
+ */
+hb_bool_t
+hb_variation_from_string (const char *str, int len,
+			  hb_variation_t *variation)
+{
+  hb_variation_t var;
+
+  if (len < 0)
+    len = strlen (str);
+
+  if (likely (parse_one_variation (&str, str + len, &var)))
+  {
+    if (variation)
+      *variation = var;
+    return true;
+  }
+
+  if (variation)
+    memset (variation, 0, sizeof (*variation));
+  return false;
+}
+
+/**
+ * hb_variation_to_string:
+ *
+ * Since: 1.4.2
+ */
+void
+hb_variation_to_string (hb_variation_t *variation,
+			char *buf, unsigned int size)
+{
+  if (unlikely (!size)) return;
+
+  char s[128];
+  unsigned int len = 0;
+  hb_tag_to_string (variation->tag, s + len);
+  len += 4;
+  while (len && s[len - 1] == ' ')
+    len--;
+  s[len++] = '=';
+  len += hb_max (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%g", (double) variation->value));
+
+  assert (len < ARRAY_LENGTH (s));
+  len = hb_min (len, size - 1);
+  memcpy (buf, s, len);
+  buf[len] = '\0';
+}
+
+/**
+ * hb_color_get_alpha:
+ * color: a #hb_color_t we are interested in its channels.
+ *
+ * Return value: Alpha channel value of the given color
+ *
+ * Since: 2.1.0
+ */
+uint8_t
+(hb_color_get_alpha) (hb_color_t color)
+{
+  return hb_color_get_alpha (color);
+}
+
+/**
+ * hb_color_get_red:
+ * color: a #hb_color_t we are interested in its channels.
+ *
+ * Return value: Red channel value of the given color
+ *
+ * Since: 2.1.0
+ */
+uint8_t
+(hb_color_get_red) (hb_color_t color)
+{
+  return hb_color_get_red (color);
+}
+
+/**
+ * hb_color_get_green:
+ * color: a #hb_color_t we are interested in its channels.
+ *
+ * Return value: Green channel value of the given color
+ *
+ * Since: 2.1.0
+ */
+uint8_t
+(hb_color_get_green) (hb_color_t color)
+{
+  return hb_color_get_green (color);
+}
+
+/**
+ * hb_color_get_blue:
+ * color: a #hb_color_t we are interested in its channels.
+ *
+ * Return value: Blue channel value of the given color
+ *
+ * Since: 2.1.0
+ */
+uint8_t
+(hb_color_get_blue) (hb_color_t color)
+{
+  return hb_color_get_blue (color);
+}
+
+
+/* If there is no visibility control, then hb-static.cc will NOT
+ * define anything.  Instead, we get it to define one set in here
+ * only, so only libharfbuzz.so defines them, not other libs. */
+#ifdef HB_NO_VISIBILITY
+#undef HB_NO_VISIBILITY
+#include "hb-static.cc"
+#define HB_NO_VISIBILITY 1
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-common.h b/thirdparty/harfbuzz/src/hb-common.h
new file mode 100644
index 0000000000..a97a5f5a04
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-common.h
@@ -0,0 +1,513 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_COMMON_H
+#define HB_COMMON_H
+
+#ifndef HB_EXTERN
+#define HB_EXTERN extern
+#endif
+
+#ifndef HB_BEGIN_DECLS
+# ifdef __cplusplus
+#  define HB_BEGIN_DECLS	extern "C" {
+#  define HB_END_DECLS		}
+# else /* !__cplusplus */
+#  define HB_BEGIN_DECLS
+#  define HB_END_DECLS
+# endif /* !__cplusplus */
+#endif
+
+#if defined (_SVR4) || defined (SVR4) || defined (__OpenBSD__) || \
+    defined (_sgi) || defined (__sun) || defined (sun) || \
+    defined (__digital__) || defined (__HP_cc)
+#  include <inttypes.h>
+#elif defined (_AIX)
+#  include <sys/inttypes.h>
+#elif defined (_MSC_VER) && _MSC_VER < 1600
+/* VS 2010 (_MSC_VER 1600) has stdint.h */
+typedef __int8 int8_t;
+typedef unsigned __int8 uint8_t;
+typedef __int16 int16_t;
+typedef unsigned __int16 uint16_t;
+typedef __int32 int32_t;
+typedef unsigned __int32 uint32_t;
+typedef __int64 int64_t;
+typedef unsigned __int64 uint64_t;
+#elif defined (__KERNEL__)
+#  include <linux/types.h>
+#else
+#  include <stdint.h>
+#endif
+
+#if defined(__GNUC__) && ((__GNUC__ > 3) || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1))
+#define HB_DEPRECATED __attribute__((__deprecated__))
+#elif defined(_MSC_VER) && (_MSC_VER >= 1300)
+#define HB_DEPRECATED __declspec(deprecated)
+#else
+#define HB_DEPRECATED
+#endif
+
+#if defined(__GNUC__) && ((__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ >= 5))
+#define HB_DEPRECATED_FOR(f) __attribute__((__deprecated__("Use '" #f "' instead")))
+#elif defined(_MSC_FULL_VER) && (_MSC_FULL_VER > 140050320)
+#define HB_DEPRECATED_FOR(f) __declspec(deprecated("is deprecated. Use '" #f "' instead"))
+#else
+#define HB_DEPRECATED_FOR(f) HB_DEPRECATED
+#endif
+
+
+HB_BEGIN_DECLS
+
+
+typedef int hb_bool_t;
+
+typedef uint32_t hb_codepoint_t;
+typedef int32_t hb_position_t;
+typedef uint32_t hb_mask_t;
+
+typedef union _hb_var_int_t {
+  uint32_t u32;
+  int32_t i32;
+  uint16_t u16[2];
+  int16_t i16[2];
+  uint8_t u8[4];
+  int8_t i8[4];
+} hb_var_int_t;
+
+
+/* hb_tag_t */
+
+typedef uint32_t hb_tag_t;
+
+#define HB_TAG(c1,c2,c3,c4) ((hb_tag_t)((((uint32_t)(c1)&0xFF)<<24)|(((uint32_t)(c2)&0xFF)<<16)|(((uint32_t)(c3)&0xFF)<<8)|((uint32_t)(c4)&0xFF)))
+#define HB_UNTAG(tag)   (uint8_t)(((tag)>>24)&0xFF), (uint8_t)(((tag)>>16)&0xFF), (uint8_t)(((tag)>>8)&0xFF), (uint8_t)((tag)&0xFF)
+
+#define HB_TAG_NONE HB_TAG(0,0,0,0)
+#define HB_TAG_MAX HB_TAG(0xff,0xff,0xff,0xff)
+#define HB_TAG_MAX_SIGNED HB_TAG(0x7f,0xff,0xff,0xff)
+
+/* len=-1 means str is NUL-terminated. */
+HB_EXTERN hb_tag_t
+hb_tag_from_string (const char *str, int len);
+
+/* buf should have 4 bytes. */
+HB_EXTERN void
+hb_tag_to_string (hb_tag_t tag, char *buf);
+
+
+/**
+ * hb_direction_t:
+ * @HB_DIRECTION_INVALID: Initial, unset direction.
+ * @HB_DIRECTION_LTR: Text is set horizontally from left to right.
+ * @HB_DIRECTION_RTL: Text is set horizontally from right to left.
+ * @HB_DIRECTION_TTB: Text is set vertically from top to bottom.
+ * @HB_DIRECTION_BTT: Text is set vertically from bottom to top.
+ */
+typedef enum {
+  HB_DIRECTION_INVALID = 0,
+  HB_DIRECTION_LTR = 4,
+  HB_DIRECTION_RTL,
+  HB_DIRECTION_TTB,
+  HB_DIRECTION_BTT
+} hb_direction_t;
+
+/* len=-1 means str is NUL-terminated */
+HB_EXTERN hb_direction_t
+hb_direction_from_string (const char *str, int len);
+
+HB_EXTERN const char *
+hb_direction_to_string (hb_direction_t direction);
+
+#define HB_DIRECTION_IS_VALID(dir)	((((unsigned int) (dir)) & ~3U) == 4)
+/* Direction must be valid for the following */
+#define HB_DIRECTION_IS_HORIZONTAL(dir)	((((unsigned int) (dir)) & ~1U) == 4)
+#define HB_DIRECTION_IS_VERTICAL(dir)	((((unsigned int) (dir)) & ~1U) == 6)
+#define HB_DIRECTION_IS_FORWARD(dir)	((((unsigned int) (dir)) & ~2U) == 4)
+#define HB_DIRECTION_IS_BACKWARD(dir)	((((unsigned int) (dir)) & ~2U) == 5)
+#define HB_DIRECTION_REVERSE(dir)	((hb_direction_t) (((unsigned int) (dir)) ^ 1))
+
+
+/* hb_language_t */
+
+typedef const struct hb_language_impl_t *hb_language_t;
+
+HB_EXTERN hb_language_t
+hb_language_from_string (const char *str, int len);
+
+HB_EXTERN const char *
+hb_language_to_string (hb_language_t language);
+
+#define HB_LANGUAGE_INVALID ((hb_language_t) 0)
+
+HB_EXTERN hb_language_t
+hb_language_get_default (void);
+
+
+/* hb_script_t */
+
+/* https://unicode.org/iso15924/ */
+/* https://docs.google.com/spreadsheets/d/1Y90M0Ie3MUJ6UVCRDOypOtijlMDLNNyyLk36T6iMu0o */
+/* Unicode Character Database property: Script (sc) */
+typedef enum
+{
+  /*1.1*/ HB_SCRIPT_COMMON			= HB_TAG ('Z','y','y','y'),
+  /*1.1*/ HB_SCRIPT_INHERITED			= HB_TAG ('Z','i','n','h'),
+  /*5.0*/ HB_SCRIPT_UNKNOWN			= HB_TAG ('Z','z','z','z'),
+
+  /*1.1*/ HB_SCRIPT_ARABIC			= HB_TAG ('A','r','a','b'),
+  /*1.1*/ HB_SCRIPT_ARMENIAN			= HB_TAG ('A','r','m','n'),
+  /*1.1*/ HB_SCRIPT_BENGALI			= HB_TAG ('B','e','n','g'),
+  /*1.1*/ HB_SCRIPT_CYRILLIC			= HB_TAG ('C','y','r','l'),
+  /*1.1*/ HB_SCRIPT_DEVANAGARI			= HB_TAG ('D','e','v','a'),
+  /*1.1*/ HB_SCRIPT_GEORGIAN			= HB_TAG ('G','e','o','r'),
+  /*1.1*/ HB_SCRIPT_GREEK			= HB_TAG ('G','r','e','k'),
+  /*1.1*/ HB_SCRIPT_GUJARATI			= HB_TAG ('G','u','j','r'),
+  /*1.1*/ HB_SCRIPT_GURMUKHI			= HB_TAG ('G','u','r','u'),
+  /*1.1*/ HB_SCRIPT_HANGUL			= HB_TAG ('H','a','n','g'),
+  /*1.1*/ HB_SCRIPT_HAN				= HB_TAG ('H','a','n','i'),
+  /*1.1*/ HB_SCRIPT_HEBREW			= HB_TAG ('H','e','b','r'),
+  /*1.1*/ HB_SCRIPT_HIRAGANA			= HB_TAG ('H','i','r','a'),
+  /*1.1*/ HB_SCRIPT_KANNADA			= HB_TAG ('K','n','d','a'),
+  /*1.1*/ HB_SCRIPT_KATAKANA			= HB_TAG ('K','a','n','a'),
+  /*1.1*/ HB_SCRIPT_LAO				= HB_TAG ('L','a','o','o'),
+  /*1.1*/ HB_SCRIPT_LATIN			= HB_TAG ('L','a','t','n'),
+  /*1.1*/ HB_SCRIPT_MALAYALAM			= HB_TAG ('M','l','y','m'),
+  /*1.1*/ HB_SCRIPT_ORIYA			= HB_TAG ('O','r','y','a'),
+  /*1.1*/ HB_SCRIPT_TAMIL			= HB_TAG ('T','a','m','l'),
+  /*1.1*/ HB_SCRIPT_TELUGU			= HB_TAG ('T','e','l','u'),
+  /*1.1*/ HB_SCRIPT_THAI			= HB_TAG ('T','h','a','i'),
+
+  /*2.0*/ HB_SCRIPT_TIBETAN			= HB_TAG ('T','i','b','t'),
+
+  /*3.0*/ HB_SCRIPT_BOPOMOFO			= HB_TAG ('B','o','p','o'),
+  /*3.0*/ HB_SCRIPT_BRAILLE			= HB_TAG ('B','r','a','i'),
+  /*3.0*/ HB_SCRIPT_CANADIAN_SYLLABICS		= HB_TAG ('C','a','n','s'),
+  /*3.0*/ HB_SCRIPT_CHEROKEE			= HB_TAG ('C','h','e','r'),
+  /*3.0*/ HB_SCRIPT_ETHIOPIC			= HB_TAG ('E','t','h','i'),
+  /*3.0*/ HB_SCRIPT_KHMER			= HB_TAG ('K','h','m','r'),
+  /*3.0*/ HB_SCRIPT_MONGOLIAN			= HB_TAG ('M','o','n','g'),
+  /*3.0*/ HB_SCRIPT_MYANMAR			= HB_TAG ('M','y','m','r'),
+  /*3.0*/ HB_SCRIPT_OGHAM			= HB_TAG ('O','g','a','m'),
+  /*3.0*/ HB_SCRIPT_RUNIC			= HB_TAG ('R','u','n','r'),
+  /*3.0*/ HB_SCRIPT_SINHALA			= HB_TAG ('S','i','n','h'),
+  /*3.0*/ HB_SCRIPT_SYRIAC			= HB_TAG ('S','y','r','c'),
+  /*3.0*/ HB_SCRIPT_THAANA			= HB_TAG ('T','h','a','a'),
+  /*3.0*/ HB_SCRIPT_YI				= HB_TAG ('Y','i','i','i'),
+
+  /*3.1*/ HB_SCRIPT_DESERET			= HB_TAG ('D','s','r','t'),
+  /*3.1*/ HB_SCRIPT_GOTHIC			= HB_TAG ('G','o','t','h'),
+  /*3.1*/ HB_SCRIPT_OLD_ITALIC			= HB_TAG ('I','t','a','l'),
+
+  /*3.2*/ HB_SCRIPT_BUHID			= HB_TAG ('B','u','h','d'),
+  /*3.2*/ HB_SCRIPT_HANUNOO			= HB_TAG ('H','a','n','o'),
+  /*3.2*/ HB_SCRIPT_TAGALOG			= HB_TAG ('T','g','l','g'),
+  /*3.2*/ HB_SCRIPT_TAGBANWA			= HB_TAG ('T','a','g','b'),
+
+  /*4.0*/ HB_SCRIPT_CYPRIOT			= HB_TAG ('C','p','r','t'),
+  /*4.0*/ HB_SCRIPT_LIMBU			= HB_TAG ('L','i','m','b'),
+  /*4.0*/ HB_SCRIPT_LINEAR_B			= HB_TAG ('L','i','n','b'),
+  /*4.0*/ HB_SCRIPT_OSMANYA			= HB_TAG ('O','s','m','a'),
+  /*4.0*/ HB_SCRIPT_SHAVIAN			= HB_TAG ('S','h','a','w'),
+  /*4.0*/ HB_SCRIPT_TAI_LE			= HB_TAG ('T','a','l','e'),
+  /*4.0*/ HB_SCRIPT_UGARITIC			= HB_TAG ('U','g','a','r'),
+
+  /*4.1*/ HB_SCRIPT_BUGINESE			= HB_TAG ('B','u','g','i'),
+  /*4.1*/ HB_SCRIPT_COPTIC			= HB_TAG ('C','o','p','t'),
+  /*4.1*/ HB_SCRIPT_GLAGOLITIC			= HB_TAG ('G','l','a','g'),
+  /*4.1*/ HB_SCRIPT_KHAROSHTHI			= HB_TAG ('K','h','a','r'),
+  /*4.1*/ HB_SCRIPT_NEW_TAI_LUE			= HB_TAG ('T','a','l','u'),
+  /*4.1*/ HB_SCRIPT_OLD_PERSIAN			= HB_TAG ('X','p','e','o'),
+  /*4.1*/ HB_SCRIPT_SYLOTI_NAGRI		= HB_TAG ('S','y','l','o'),
+  /*4.1*/ HB_SCRIPT_TIFINAGH			= HB_TAG ('T','f','n','g'),
+
+  /*5.0*/ HB_SCRIPT_BALINESE			= HB_TAG ('B','a','l','i'),
+  /*5.0*/ HB_SCRIPT_CUNEIFORM			= HB_TAG ('X','s','u','x'),
+  /*5.0*/ HB_SCRIPT_NKO				= HB_TAG ('N','k','o','o'),
+  /*5.0*/ HB_SCRIPT_PHAGS_PA			= HB_TAG ('P','h','a','g'),
+  /*5.0*/ HB_SCRIPT_PHOENICIAN			= HB_TAG ('P','h','n','x'),
+
+  /*5.1*/ HB_SCRIPT_CARIAN			= HB_TAG ('C','a','r','i'),
+  /*5.1*/ HB_SCRIPT_CHAM			= HB_TAG ('C','h','a','m'),
+  /*5.1*/ HB_SCRIPT_KAYAH_LI			= HB_TAG ('K','a','l','i'),
+  /*5.1*/ HB_SCRIPT_LEPCHA			= HB_TAG ('L','e','p','c'),
+  /*5.1*/ HB_SCRIPT_LYCIAN			= HB_TAG ('L','y','c','i'),
+  /*5.1*/ HB_SCRIPT_LYDIAN			= HB_TAG ('L','y','d','i'),
+  /*5.1*/ HB_SCRIPT_OL_CHIKI			= HB_TAG ('O','l','c','k'),
+  /*5.1*/ HB_SCRIPT_REJANG			= HB_TAG ('R','j','n','g'),
+  /*5.1*/ HB_SCRIPT_SAURASHTRA			= HB_TAG ('S','a','u','r'),
+  /*5.1*/ HB_SCRIPT_SUNDANESE			= HB_TAG ('S','u','n','d'),
+  /*5.1*/ HB_SCRIPT_VAI				= HB_TAG ('V','a','i','i'),
+
+  /*5.2*/ HB_SCRIPT_AVESTAN			= HB_TAG ('A','v','s','t'),
+  /*5.2*/ HB_SCRIPT_BAMUM			= HB_TAG ('B','a','m','u'),
+  /*5.2*/ HB_SCRIPT_EGYPTIAN_HIEROGLYPHS	= HB_TAG ('E','g','y','p'),
+  /*5.2*/ HB_SCRIPT_IMPERIAL_ARAMAIC		= HB_TAG ('A','r','m','i'),
+  /*5.2*/ HB_SCRIPT_INSCRIPTIONAL_PAHLAVI	= HB_TAG ('P','h','l','i'),
+  /*5.2*/ HB_SCRIPT_INSCRIPTIONAL_PARTHIAN	= HB_TAG ('P','r','t','i'),
+  /*5.2*/ HB_SCRIPT_JAVANESE			= HB_TAG ('J','a','v','a'),
+  /*5.2*/ HB_SCRIPT_KAITHI			= HB_TAG ('K','t','h','i'),
+  /*5.2*/ HB_SCRIPT_LISU			= HB_TAG ('L','i','s','u'),
+  /*5.2*/ HB_SCRIPT_MEETEI_MAYEK		= HB_TAG ('M','t','e','i'),
+  /*5.2*/ HB_SCRIPT_OLD_SOUTH_ARABIAN		= HB_TAG ('S','a','r','b'),
+  /*5.2*/ HB_SCRIPT_OLD_TURKIC			= HB_TAG ('O','r','k','h'),
+  /*5.2*/ HB_SCRIPT_SAMARITAN			= HB_TAG ('S','a','m','r'),
+  /*5.2*/ HB_SCRIPT_TAI_THAM			= HB_TAG ('L','a','n','a'),
+  /*5.2*/ HB_SCRIPT_TAI_VIET			= HB_TAG ('T','a','v','t'),
+
+  /*6.0*/ HB_SCRIPT_BATAK			= HB_TAG ('B','a','t','k'),
+  /*6.0*/ HB_SCRIPT_BRAHMI			= HB_TAG ('B','r','a','h'),
+  /*6.0*/ HB_SCRIPT_MANDAIC			= HB_TAG ('M','a','n','d'),
+
+  /*6.1*/ HB_SCRIPT_CHAKMA			= HB_TAG ('C','a','k','m'),
+  /*6.1*/ HB_SCRIPT_MEROITIC_CURSIVE		= HB_TAG ('M','e','r','c'),
+  /*6.1*/ HB_SCRIPT_MEROITIC_HIEROGLYPHS	= HB_TAG ('M','e','r','o'),
+  /*6.1*/ HB_SCRIPT_MIAO			= HB_TAG ('P','l','r','d'),
+  /*6.1*/ HB_SCRIPT_SHARADA			= HB_TAG ('S','h','r','d'),
+  /*6.1*/ HB_SCRIPT_SORA_SOMPENG		= HB_TAG ('S','o','r','a'),
+  /*6.1*/ HB_SCRIPT_TAKRI			= HB_TAG ('T','a','k','r'),
+
+  /*
+   * Since: 0.9.30
+   */
+  /*7.0*/ HB_SCRIPT_BASSA_VAH			= HB_TAG ('B','a','s','s'),
+  /*7.0*/ HB_SCRIPT_CAUCASIAN_ALBANIAN		= HB_TAG ('A','g','h','b'),
+  /*7.0*/ HB_SCRIPT_DUPLOYAN			= HB_TAG ('D','u','p','l'),
+  /*7.0*/ HB_SCRIPT_ELBASAN			= HB_TAG ('E','l','b','a'),
+  /*7.0*/ HB_SCRIPT_GRANTHA			= HB_TAG ('G','r','a','n'),
+  /*7.0*/ HB_SCRIPT_KHOJKI			= HB_TAG ('K','h','o','j'),
+  /*7.0*/ HB_SCRIPT_KHUDAWADI			= HB_TAG ('S','i','n','d'),
+  /*7.0*/ HB_SCRIPT_LINEAR_A			= HB_TAG ('L','i','n','a'),
+  /*7.0*/ HB_SCRIPT_MAHAJANI			= HB_TAG ('M','a','h','j'),
+  /*7.0*/ HB_SCRIPT_MANICHAEAN			= HB_TAG ('M','a','n','i'),
+  /*7.0*/ HB_SCRIPT_MENDE_KIKAKUI		= HB_TAG ('M','e','n','d'),
+  /*7.0*/ HB_SCRIPT_MODI			= HB_TAG ('M','o','d','i'),
+  /*7.0*/ HB_SCRIPT_MRO				= HB_TAG ('M','r','o','o'),
+  /*7.0*/ HB_SCRIPT_NABATAEAN			= HB_TAG ('N','b','a','t'),
+  /*7.0*/ HB_SCRIPT_OLD_NORTH_ARABIAN		= HB_TAG ('N','a','r','b'),
+  /*7.0*/ HB_SCRIPT_OLD_PERMIC			= HB_TAG ('P','e','r','m'),
+  /*7.0*/ HB_SCRIPT_PAHAWH_HMONG		= HB_TAG ('H','m','n','g'),
+  /*7.0*/ HB_SCRIPT_PALMYRENE			= HB_TAG ('P','a','l','m'),
+  /*7.0*/ HB_SCRIPT_PAU_CIN_HAU			= HB_TAG ('P','a','u','c'),
+  /*7.0*/ HB_SCRIPT_PSALTER_PAHLAVI		= HB_TAG ('P','h','l','p'),
+  /*7.0*/ HB_SCRIPT_SIDDHAM			= HB_TAG ('S','i','d','d'),
+  /*7.0*/ HB_SCRIPT_TIRHUTA			= HB_TAG ('T','i','r','h'),
+  /*7.0*/ HB_SCRIPT_WARANG_CITI			= HB_TAG ('W','a','r','a'),
+
+  /*8.0*/ HB_SCRIPT_AHOM			= HB_TAG ('A','h','o','m'),
+  /*8.0*/ HB_SCRIPT_ANATOLIAN_HIEROGLYPHS	= HB_TAG ('H','l','u','w'),
+  /*8.0*/ HB_SCRIPT_HATRAN			= HB_TAG ('H','a','t','r'),
+  /*8.0*/ HB_SCRIPT_MULTANI			= HB_TAG ('M','u','l','t'),
+  /*8.0*/ HB_SCRIPT_OLD_HUNGARIAN		= HB_TAG ('H','u','n','g'),
+  /*8.0*/ HB_SCRIPT_SIGNWRITING			= HB_TAG ('S','g','n','w'),
+
+  /*
+   * Since 1.3.0
+   */
+  /*9.0*/ HB_SCRIPT_ADLAM			= HB_TAG ('A','d','l','m'),
+  /*9.0*/ HB_SCRIPT_BHAIKSUKI			= HB_TAG ('B','h','k','s'),
+  /*9.0*/ HB_SCRIPT_MARCHEN			= HB_TAG ('M','a','r','c'),
+  /*9.0*/ HB_SCRIPT_OSAGE			= HB_TAG ('O','s','g','e'),
+  /*9.0*/ HB_SCRIPT_TANGUT			= HB_TAG ('T','a','n','g'),
+  /*9.0*/ HB_SCRIPT_NEWA			= HB_TAG ('N','e','w','a'),
+
+  /*
+   * Since 1.6.0
+   */
+  /*10.0*/HB_SCRIPT_MASARAM_GONDI		= HB_TAG ('G','o','n','m'),
+  /*10.0*/HB_SCRIPT_NUSHU			= HB_TAG ('N','s','h','u'),
+  /*10.0*/HB_SCRIPT_SOYOMBO			= HB_TAG ('S','o','y','o'),
+  /*10.0*/HB_SCRIPT_ZANABAZAR_SQUARE		= HB_TAG ('Z','a','n','b'),
+
+  /*
+   * Since 1.8.0
+   */
+  /*11.0*/HB_SCRIPT_DOGRA			= HB_TAG ('D','o','g','r'),
+  /*11.0*/HB_SCRIPT_GUNJALA_GONDI		= HB_TAG ('G','o','n','g'),
+  /*11.0*/HB_SCRIPT_HANIFI_ROHINGYA		= HB_TAG ('R','o','h','g'),
+  /*11.0*/HB_SCRIPT_MAKASAR			= HB_TAG ('M','a','k','a'),
+  /*11.0*/HB_SCRIPT_MEDEFAIDRIN			= HB_TAG ('M','e','d','f'),
+  /*11.0*/HB_SCRIPT_OLD_SOGDIAN			= HB_TAG ('S','o','g','o'),
+  /*11.0*/HB_SCRIPT_SOGDIAN			= HB_TAG ('S','o','g','d'),
+
+  /*
+   * Since 2.4.0
+   */
+  /*12.0*/HB_SCRIPT_ELYMAIC			= HB_TAG ('E','l','y','m'),
+  /*12.0*/HB_SCRIPT_NANDINAGARI			= HB_TAG ('N','a','n','d'),
+  /*12.0*/HB_SCRIPT_NYIAKENG_PUACHUE_HMONG	= HB_TAG ('H','m','n','p'),
+  /*12.0*/HB_SCRIPT_WANCHO			= HB_TAG ('W','c','h','o'),
+
+  /*
+   * Since 2.6.7
+   */
+  /*13.0*/HB_SCRIPT_CHORASMIAN			= HB_TAG ('C','h','r','s'),
+  /*13.0*/HB_SCRIPT_DIVES_AKURU			= HB_TAG ('D','i','a','k'),
+  /*13.0*/HB_SCRIPT_KHITAN_SMALL_SCRIPT		= HB_TAG ('K','i','t','s'),
+  /*13.0*/HB_SCRIPT_YEZIDI			= HB_TAG ('Y','e','z','i'),
+
+  /* No script set. */
+  HB_SCRIPT_INVALID				= HB_TAG_NONE,
+
+  /* Dummy values to ensure any hb_tag_t value can be passed/stored as hb_script_t
+   * without risking undefined behavior.  We have two, for historical reasons.
+   * HB_TAG_MAX used to be unsigned, but that was invalid Ansi C, so was changed
+   * to _HB_SCRIPT_MAX_VALUE to be equal to HB_TAG_MAX_SIGNED as well.
+   *
+   * See this thread for technicalities:
+   *
+   *   https://lists.freedesktop.org/archives/harfbuzz/2014-March/004150.html
+   */
+  _HB_SCRIPT_MAX_VALUE				= HB_TAG_MAX_SIGNED, /*< skip >*/
+  _HB_SCRIPT_MAX_VALUE_SIGNED			= HB_TAG_MAX_SIGNED /*< skip >*/
+
+} hb_script_t;
+
+
+/* Script functions */
+
+HB_EXTERN hb_script_t
+hb_script_from_iso15924_tag (hb_tag_t tag);
+
+HB_EXTERN hb_script_t
+hb_script_from_string (const char *str, int len);
+
+HB_EXTERN hb_tag_t
+hb_script_to_iso15924_tag (hb_script_t script);
+
+HB_EXTERN hb_direction_t
+hb_script_get_horizontal_direction (hb_script_t script);
+
+
+/* User data */
+
+typedef struct hb_user_data_key_t {
+  /*< private >*/
+  char unused;
+} hb_user_data_key_t;
+
+typedef void (*hb_destroy_func_t) (void *user_data);
+
+
+/* Font features and variations. */
+
+/**
+ * HB_FEATURE_GLOBAL_START
+ *
+ * Since: 2.0.0
+ */
+#define HB_FEATURE_GLOBAL_START	0
+/**
+ * HB_FEATURE_GLOBAL_END
+ *
+ * Since: 2.0.0
+ */
+#define HB_FEATURE_GLOBAL_END	((unsigned int) -1)
+
+/**
+ * hb_feature_t:
+ * @tag: a feature tag
+ * @value: 0 disables the feature, non-zero (usually 1) enables the feature.
+ * For features implemented as lookup type 3 (like 'salt') the @value is a one
+ * based index into the alternates.
+ * @start: the cluster to start applying this feature setting (inclusive).
+ * @end: the cluster to end applying this feature setting (exclusive).
+ *
+ * The #hb_feature_t is the structure that holds information about requested
+ * feature application. The feature will be applied with the given value to all
+ * glyphs which are in clusters between @start (inclusive) and @end (exclusive).
+ * Setting start to @HB_FEATURE_GLOBAL_START and end to @HB_FEATURE_GLOBAL_END
+ * specifies that the feature always applies to the entire buffer.
+ */
+typedef struct hb_feature_t {
+  hb_tag_t      tag;
+  uint32_t      value;
+  unsigned int  start;
+  unsigned int  end;
+} hb_feature_t;
+
+HB_EXTERN hb_bool_t
+hb_feature_from_string (const char *str, int len,
+			hb_feature_t *feature);
+
+HB_EXTERN void
+hb_feature_to_string (hb_feature_t *feature,
+		      char *buf, unsigned int size);
+
+/**
+ * hb_variation_t:
+ *
+ * Since: 1.4.2
+ */
+typedef struct hb_variation_t {
+  hb_tag_t tag;
+  float    value;
+} hb_variation_t;
+
+HB_EXTERN hb_bool_t
+hb_variation_from_string (const char *str, int len,
+			  hb_variation_t *variation);
+
+HB_EXTERN void
+hb_variation_to_string (hb_variation_t *variation,
+			char *buf, unsigned int size);
+
+/**
+ * hb_color_t:
+ *
+ * Data type for holding color values.
+ *
+ * Since: 2.1.0
+ */
+typedef uint32_t hb_color_t;
+
+#define HB_COLOR(b,g,r,a) ((hb_color_t) HB_TAG ((b),(g),(r),(a)))
+
+HB_EXTERN uint8_t
+hb_color_get_alpha (hb_color_t color);
+#define hb_color_get_alpha(color)	((color) & 0xFF)
+
+HB_EXTERN uint8_t
+hb_color_get_red (hb_color_t color);
+#define hb_color_get_red(color)		(((color) >> 8) & 0xFF)
+
+HB_EXTERN uint8_t
+hb_color_get_green (hb_color_t color);
+#define hb_color_get_green(color)	(((color) >> 16) & 0xFF)
+
+HB_EXTERN uint8_t
+hb_color_get_blue (hb_color_t color);
+#define hb_color_get_blue(color)	(((color) >> 24) & 0xFF)
+
+HB_END_DECLS
+
+#endif /* HB_COMMON_H */
diff --git a/thirdparty/harfbuzz/src/hb-config.hh b/thirdparty/harfbuzz/src/hb-config.hh
new file mode 100644
index 0000000000..fc8d424bfb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-config.hh
@@ -0,0 +1,163 @@
+/*
+ * Copyright © 2019  Facebook, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_CONFIG_HH
+#define HB_CONFIG_HH
+
+#if 0 /* Make test happy. */
+#include "hb.hh"
+#endif
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+
+#ifdef HB_TINY
+#define HB_LEAN
+#define HB_MINI
+#define HB_NO_MT
+#define HB_NO_UCD_UNASSIGNED
+#ifndef NDEBUG
+#define NDEBUG
+#endif
+#ifndef __OPTIMIZE_SIZE__
+#define __OPTIMIZE_SIZE__
+#endif
+#endif
+
+#ifdef HB_LEAN
+#define HB_DISABLE_DEPRECATED
+#define HB_NDEBUG
+#define HB_NO_ATEXIT
+#define HB_NO_BUFFER_MESSAGE
+#define HB_NO_BUFFER_SERIALIZE
+#define HB_NO_BITMAP
+#define HB_NO_CFF
+#define HB_NO_COLOR
+#define HB_NO_DRAW
+#define HB_NO_ERRNO
+#define HB_NO_FACE_COLLECT_UNICODES
+#define HB_NO_GETENV
+#define HB_NO_HINTING
+#define HB_NO_LANGUAGE_PRIVATE_SUBTAG
+#define HB_NO_LAYOUT_FEATURE_PARAMS
+#define HB_NO_LAYOUT_COLLECT_GLYPHS
+#define HB_NO_LAYOUT_UNUSED
+#define HB_NO_MATH
+#define HB_NO_META
+#define HB_NO_METRICS
+#define HB_NO_MMAP
+#define HB_NO_NAME
+#define HB_NO_OPEN
+#define HB_NO_SETLOCALE
+#define HB_NO_OT_FONT_GLYPH_NAMES
+#define HB_NO_OT_SHAPE_FRACTIONS
+#define HB_NO_STYLE
+#define HB_NO_SUBSET_LAYOUT
+#define HB_NO_VAR
+#endif
+
+#ifdef HB_MINI
+#define HB_NO_AAT
+#define HB_NO_LEGACY
+#endif
+
+
+/* Closure of options. */
+
+#ifdef HB_DISABLE_DEPRECATED
+#define HB_IF_NOT_DEPRECATED(x)
+#else
+#define HB_IF_NOT_DEPRECATED(x) x
+#endif
+
+#ifdef HB_NO_AAT
+#define HB_NO_OT_NAME_LANGUAGE_AAT
+#define HB_NO_AAT_SHAPE
+#endif
+
+#ifdef HB_NO_BITMAP
+#define HB_NO_OT_FONT_BITMAP
+#endif
+
+#ifdef HB_NO_CFF
+#define HB_NO_OT_FONT_CFF
+#define HB_NO_SUBSET_CFF
+#endif
+
+#ifdef HB_NO_GETENV
+#define HB_NO_UNISCRIBE_BUG_COMPATIBLE
+#endif
+
+#ifdef HB_NO_LEGACY
+#define HB_NO_CMAP_LEGACY_SUBTABLES
+#define HB_NO_FALLBACK_SHAPE
+#define HB_NO_OT_KERN
+#define HB_NO_OT_LAYOUT_BLACKLIST
+#define HB_NO_OT_SHAPE_FALLBACK
+#endif
+
+#ifdef HB_NO_NAME
+#define HB_NO_OT_NAME_LANGUAGE
+#endif
+
+#ifdef HB_NO_OT
+#define HB_NO_OT_FONT
+#define HB_NO_OT_LAYOUT
+#define HB_NO_OT_TAG
+#define HB_NO_OT_SHAPE
+#endif
+
+#ifdef HB_NO_OT_SHAPE
+#define HB_NO_AAT_SHAPE
+#endif
+
+#ifdef HB_NO_OT_SHAPE_FALLBACK
+#define HB_NO_OT_SHAPE_COMPLEX_ARABIC_FALLBACK
+#define HB_NO_OT_SHAPE_COMPLEX_HEBREW_FALLBACK
+#define HB_NO_OT_SHAPE_COMPLEX_THAI_FALLBACK
+#define HB_NO_OT_SHAPE_COMPLEX_VOWEL_CONSTRAINTS
+#endif
+
+#ifdef NDEBUG
+#ifndef HB_NDEBUG
+#define HB_NDEBUG
+#endif
+#endif
+
+#ifdef __OPTIMIZE_SIZE__
+#ifndef HB_OPTIMIZE_SIZE
+#define HB_OPTIMIZE_SIZE
+#endif
+#endif
+
+#ifdef HAVE_CONFIG_OVERRIDE_H
+#include "config-override.h"
+#endif
+
+
+#endif /* HB_CONFIG_HH */
diff --git a/thirdparty/harfbuzz/src/hb-coretext.cc b/thirdparty/harfbuzz/src/hb-coretext.cc
new file mode 100644
index 0000000000..7b6b2bd5ef
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-coretext.cc
@@ -0,0 +1,1194 @@
+/*
+ * Copyright © 2012,2013  Mozilla Foundation.
+ * Copyright © 2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_CORETEXT
+
+#include "hb-shaper-impl.hh"
+
+#include "hb-coretext.h"
+#include "hb-aat-layout.hh"
+#include <math.h>
+
+
+/**
+ * SECTION:hb-coretext
+ * @title: hb-coretext
+ * @short_description: CoreText integration
+ * @include: hb-coretext.h
+ *
+ * Functions for using HarfBuzz with the CoreText fonts.
+ **/
+
+/* https://developer.apple.com/documentation/coretext/1508745-ctfontcreatewithgraphicsfont */
+#define HB_CORETEXT_DEFAULT_FONT_SIZE 12.f
+
+static void
+release_table_data (void *user_data)
+{
+  CFDataRef cf_data = reinterpret_cast<CFDataRef> (user_data);
+  CFRelease(cf_data);
+}
+
+static hb_blob_t *
+_hb_cg_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  CGFontRef cg_font = reinterpret_cast<CGFontRef> (user_data);
+  CFDataRef cf_data = CGFontCopyTableForTag (cg_font, tag);
+  if (unlikely (!cf_data))
+    return nullptr;
+
+  const char *data = reinterpret_cast<const char*> (CFDataGetBytePtr (cf_data));
+  const size_t length = CFDataGetLength (cf_data);
+  if (!data || !length)
+  {
+    CFRelease (cf_data);
+    return nullptr;
+  }
+
+  return hb_blob_create (data, length, HB_MEMORY_MODE_READONLY,
+			 reinterpret_cast<void *> (const_cast<__CFData *> (cf_data)),
+			 release_table_data);
+}
+
+static void
+_hb_cg_font_release (void *data)
+{
+  CGFontRelease ((CGFontRef) data);
+}
+
+
+static CTFontDescriptorRef
+get_last_resort_font_desc ()
+{
+  // TODO Handle allocation failures?
+  CTFontDescriptorRef last_resort = CTFontDescriptorCreateWithNameAndSize (CFSTR("LastResort"), 0);
+  CFArrayRef cascade_list = CFArrayCreate (kCFAllocatorDefault,
+					   (const void **) &last_resort,
+					   1,
+					   &kCFTypeArrayCallBacks);
+  CFRelease (last_resort);
+  CFDictionaryRef attributes = CFDictionaryCreate (kCFAllocatorDefault,
+						   (const void **) &kCTFontCascadeListAttribute,
+						   (const void **) &cascade_list,
+						   1,
+						   &kCFTypeDictionaryKeyCallBacks,
+						   &kCFTypeDictionaryValueCallBacks);
+  CFRelease (cascade_list);
+
+  CTFontDescriptorRef font_desc = CTFontDescriptorCreateWithAttributes (attributes);
+  CFRelease (attributes);
+  return font_desc;
+}
+
+static void
+release_data (void *info, const void *data, size_t size)
+{
+  assert (hb_blob_get_length ((hb_blob_t *) info) == size &&
+	  hb_blob_get_data ((hb_blob_t *) info, nullptr) == data);
+
+  hb_blob_destroy ((hb_blob_t *) info);
+}
+
+static CGFontRef
+create_cg_font (hb_face_t *face)
+{
+  CGFontRef cg_font = nullptr;
+  if (face->destroy == _hb_cg_font_release)
+  {
+    cg_font = CGFontRetain ((CGFontRef) face->user_data);
+  }
+  else
+  {
+    hb_blob_t *blob = hb_face_reference_blob (face);
+    unsigned int blob_length;
+    const char *blob_data = hb_blob_get_data (blob, &blob_length);
+    if (unlikely (!blob_length))
+      DEBUG_MSG (CORETEXT, face, "Face has empty blob");
+
+    CGDataProviderRef provider = CGDataProviderCreateWithData (blob, blob_data, blob_length, &release_data);
+    if (likely (provider))
+    {
+      cg_font = CGFontCreateWithDataProvider (provider);
+      if (unlikely (!cg_font))
+	DEBUG_MSG (CORETEXT, face, "Face CGFontCreateWithDataProvider() failed");
+      CGDataProviderRelease (provider);
+    }
+  }
+  return cg_font;
+}
+
+static CTFontRef
+create_ct_font (CGFontRef cg_font, CGFloat font_size)
+{
+  CTFontRef ct_font = nullptr;
+
+  /* CoreText does not enable trak table usage / tracking when creating a CTFont
+   * using CTFontCreateWithGraphicsFont. The only way of enabling tracking seems
+   * to be through the CTFontCreateUIFontForLanguage call. */
+  CFStringRef cg_postscript_name = CGFontCopyPostScriptName (cg_font);
+  if (CFStringHasPrefix (cg_postscript_name, CFSTR (".SFNSText")) ||
+      CFStringHasPrefix (cg_postscript_name, CFSTR (".SFNSDisplay")))
+  {
+#if !(defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE) && MAC_OS_X_VERSION_MIN_REQUIRED < 1080
+# define kCTFontUIFontSystem kCTFontSystemFontType
+# define kCTFontUIFontEmphasizedSystem kCTFontEmphasizedSystemFontType
+#endif
+    CTFontUIFontType font_type = kCTFontUIFontSystem;
+    if (CFStringHasSuffix (cg_postscript_name, CFSTR ("-Bold")))
+      font_type = kCTFontUIFontEmphasizedSystem;
+
+    ct_font = CTFontCreateUIFontForLanguage (font_type, font_size, nullptr);
+    CFStringRef ct_result_name = CTFontCopyPostScriptName(ct_font);
+    if (CFStringCompare (ct_result_name, cg_postscript_name, 0) != kCFCompareEqualTo)
+    {
+      CFRelease(ct_font);
+      ct_font = nullptr;
+    }
+    CFRelease (ct_result_name);
+  }
+  CFRelease (cg_postscript_name);
+
+  if (!ct_font)
+    ct_font = CTFontCreateWithGraphicsFont (cg_font, font_size, nullptr, nullptr);
+
+  if (unlikely (!ct_font)) {
+    DEBUG_MSG (CORETEXT, cg_font, "Font CTFontCreateWithGraphicsFont() failed");
+    return nullptr;
+  }
+
+  /* crbug.com/576941 and crbug.com/625902 and the investigation in the latter
+   * bug indicate that the cascade list reconfiguration occasionally causes
+   * crashes in CoreText on OS X 10.9, thus let's skip this step on older
+   * operating system versions. Except for the emoji font, where _not_
+   * reconfiguring the cascade list causes CoreText crashes. For details, see
+   * crbug.com/549610 */
+  // 0x00070000 stands for "kCTVersionNumber10_10", see CoreText.h
+  if (&CTGetCoreTextVersion != nullptr && CTGetCoreTextVersion() < 0x00070000) {
+    CFStringRef fontName = CTFontCopyPostScriptName (ct_font);
+    bool isEmojiFont = CFStringCompare (fontName, CFSTR("AppleColorEmoji"), 0) == kCFCompareEqualTo;
+    CFRelease (fontName);
+    if (!isEmojiFont)
+      return ct_font;
+  }
+
+  CFURLRef original_url = nullptr;
+#if !(defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE) && MAC_OS_X_VERSION_MIN_REQUIRED < 1060
+  ATSFontRef atsFont;
+  FSRef fsref;
+  OSStatus status;
+  atsFont = CTFontGetPlatformFont (ct_font, NULL);
+  status = ATSFontGetFileReference (atsFont, &fsref);
+  if (status == noErr)
+    original_url = CFURLCreateFromFSRef (NULL, &fsref);
+#else
+  original_url = (CFURLRef) CTFontCopyAttribute (ct_font, kCTFontURLAttribute);
+#endif
+
+  /* Create font copy with cascade list that has LastResort first; this speeds up CoreText
+   * font fallback which we don't need anyway. */
+  {
+    CTFontDescriptorRef last_resort_font_desc = get_last_resort_font_desc ();
+    CTFontRef new_ct_font = CTFontCreateCopyWithAttributes (ct_font, 0.0, nullptr, last_resort_font_desc);
+    CFRelease (last_resort_font_desc);
+    if (new_ct_font)
+    {
+      /* The CTFontCreateCopyWithAttributes call fails to stay on the same font
+       * when reconfiguring the cascade list and may switch to a different font
+       * when there are fonts that go by the same name, since the descriptor is
+       * just name and size.
+       *
+       * Avoid reconfiguring the cascade lists if the new font is outside the
+       * system locations that we cannot access from the sandboxed renderer
+       * process in Blink. This can be detected by the new file URL location
+       * that the newly found font points to. */
+      CFURLRef new_url = nullptr;
+#if !(defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE) && MAC_OS_X_VERSION_MIN_REQUIRED < 1060
+      atsFont = CTFontGetPlatformFont (new_ct_font, NULL);
+      status = ATSFontGetFileReference (atsFont, &fsref);
+      if (status == noErr)
+	new_url = CFURLCreateFromFSRef (NULL, &fsref);
+#else
+      new_url = (CFURLRef) CTFontCopyAttribute (new_ct_font, kCTFontURLAttribute);
+#endif
+      // Keep reconfigured font if URL cannot be retrieved (seems to be the case
+      // on Mac OS 10.12 Sierra), speculative fix for crbug.com/625606
+      if (!original_url || !new_url || CFEqual (original_url, new_url)) {
+	CFRelease (ct_font);
+	ct_font = new_ct_font;
+      } else {
+	CFRelease (new_ct_font);
+	DEBUG_MSG (CORETEXT, ct_font, "Discarding reconfigured CTFont, location changed.");
+      }
+      if (new_url)
+	CFRelease (new_url);
+    }
+    else
+      DEBUG_MSG (CORETEXT, ct_font, "Font copy with empty cascade list failed");
+  }
+
+  if (original_url)
+    CFRelease (original_url);
+  return ct_font;
+}
+
+hb_coretext_face_data_t *
+_hb_coretext_shaper_face_data_create (hb_face_t *face)
+{
+  CGFontRef cg_font = create_cg_font (face);
+
+  if (unlikely (!cg_font))
+  {
+    DEBUG_MSG (CORETEXT, face, "CGFont creation failed..");
+    return nullptr;
+  }
+
+  return (hb_coretext_face_data_t *) cg_font;
+}
+
+void
+_hb_coretext_shaper_face_data_destroy (hb_coretext_face_data_t *data)
+{
+  CFRelease ((CGFontRef) data);
+}
+
+/**
+ * hb_coretext_face_create:
+ * @cg_font: The CGFontRef to work upon
+ *
+ * Creates an #hb_face_t face object from the specified
+ * CGFontRef.
+ *
+ * Return value: the new #hb_face_t face object
+ *
+ * Since: 0.9.10
+ */
+hb_face_t *
+hb_coretext_face_create (CGFontRef cg_font)
+{
+  return hb_face_create_for_tables (_hb_cg_reference_table, CGFontRetain (cg_font), _hb_cg_font_release);
+}
+
+/**
+ * hb_coretext_face_get_cg_font:
+ * @face: The #hb_face_t to work upon
+ *
+ * Fetches the CGFontRef associated with an #hb_face_t
+ * face object
+ *
+ * Return value: the CGFontRef found
+ *
+ * Since: 0.9.10
+ */
+CGFontRef
+hb_coretext_face_get_cg_font (hb_face_t *face)
+{
+  return (CGFontRef) (const void *) face->data.coretext;
+}
+
+
+hb_coretext_font_data_t *
+_hb_coretext_shaper_font_data_create (hb_font_t *font)
+{
+  hb_face_t *face = font->face;
+  const hb_coretext_face_data_t *face_data = face->data.coretext;
+  if (unlikely (!face_data)) return nullptr;
+  CGFontRef cg_font = (CGFontRef) (const void *) face->data.coretext;
+
+  CGFloat font_size = (CGFloat) (font->ptem <= 0.f ? HB_CORETEXT_DEFAULT_FONT_SIZE : font->ptem);
+  CTFontRef ct_font = create_ct_font (cg_font, font_size);
+
+  if (unlikely (!ct_font))
+  {
+    DEBUG_MSG (CORETEXT, font, "CGFont creation failed..");
+    return nullptr;
+  }
+
+  return (hb_coretext_font_data_t *) ct_font;
+}
+
+void
+_hb_coretext_shaper_font_data_destroy (hb_coretext_font_data_t *data)
+{
+  CFRelease ((CTFontRef) data);
+}
+
+static const hb_coretext_font_data_t *
+hb_coretext_font_data_sync (hb_font_t *font)
+{
+retry:
+  const hb_coretext_font_data_t *data = font->data.coretext;
+  if (unlikely (!data)) return nullptr;
+
+  if (fabs (CTFontGetSize ((CTFontRef) data) - (CGFloat) font->ptem) > .5)
+  {
+    /* XXX-MT-bug
+     * Note that evaluating condition above can be dangerous if another thread
+     * got here first and destructed data.  That's, as always, bad use pattern.
+     * If you modify the font (change font size), other threads must not be
+     * using it at the same time.  However, since this check is delayed to
+     * when one actually tries to shape something, this is a XXX race condition
+     * (and the only one we have that I know of) right now.  Ie. you modify the
+     * font size in one thread, then (supposedly safely) try to use it from two
+     * or more threads and BOOM!  I'm not sure how to fix this.  We want RCU.
+     */
+
+    /* Drop and recreate. */
+    /* If someone dropped it in the mean time, throw it away and don't touch it.
+     * Otherwise, destruct it. */
+    if (likely (font->data.coretext.cmpexch (const_cast<hb_coretext_font_data_t *> (data), nullptr)))
+      _hb_coretext_shaper_font_data_destroy (const_cast<hb_coretext_font_data_t *> (data));
+    else
+      goto retry;
+  }
+  return font->data.coretext;
+}
+
+/**
+ * hb_coretext_font_create:
+ * @ct_font: The CTFontRef to work upon
+ *
+ * Creates an #hb_font_t font object from the specified
+ * CTFontRef.
+ *
+ * Return value: the new #hb_font_t font object
+ *
+ * Since: 1.7.2
+ **/
+hb_font_t *
+hb_coretext_font_create (CTFontRef ct_font)
+{
+  CGFontRef cg_font = CTFontCopyGraphicsFont (ct_font, nullptr);
+  hb_face_t *face = hb_coretext_face_create (cg_font);
+  CFRelease (cg_font);
+  hb_font_t *font = hb_font_create (face);
+  hb_face_destroy (face);
+
+  if (unlikely (hb_object_is_immutable (font)))
+    return font;
+
+  hb_font_set_ptem (font, CTFontGetSize (ct_font));
+
+  /* Let there be dragons here... */
+  font->data.coretext.cmpexch (nullptr, (hb_coretext_font_data_t *) CFRetain (ct_font));
+
+  return font;
+}
+
+/**
+ * hb_coretext_face_get_ct_font:
+ * @font: #hb_font_t to work upon
+ *
+ * Fetches the CTFontRef associated with the specified
+ * #hb_font_t font object.
+ *
+ * Return value: the CTFontRef found
+ *
+ * Since: 0.9.10
+ */
+CTFontRef
+hb_coretext_font_get_ct_font (hb_font_t *font)
+{
+  const hb_coretext_font_data_t *data = hb_coretext_font_data_sync (font);
+  return data ? (CTFontRef) data : nullptr;
+}
+
+
+/*
+ * shaper
+ */
+
+struct feature_record_t {
+  unsigned int feature;
+  unsigned int setting;
+};
+
+struct active_feature_t {
+  feature_record_t rec;
+  unsigned int order;
+
+  HB_INTERNAL static int cmp (const void *pa, const void *pb) {
+    const active_feature_t *a = (const active_feature_t *) pa;
+    const active_feature_t *b = (const active_feature_t *) pb;
+    return a->rec.feature < b->rec.feature ? -1 : a->rec.feature > b->rec.feature ? 1 :
+	   a->order < b->order ? -1 : a->order > b->order ? 1 :
+	   a->rec.setting < b->rec.setting ? -1 : a->rec.setting > b->rec.setting ? 1 :
+	   0;
+  }
+  bool operator== (const active_feature_t *f) {
+    return cmp (this, f) == 0;
+  }
+};
+
+struct feature_event_t {
+  unsigned int index;
+  bool start;
+  active_feature_t feature;
+
+  HB_INTERNAL static int cmp (const void *pa, const void *pb) {
+    const feature_event_t *a = (const feature_event_t *) pa;
+    const feature_event_t *b = (const feature_event_t *) pb;
+    return a->index < b->index ? -1 : a->index > b->index ? 1 :
+	   a->start < b->start ? -1 : a->start > b->start ? 1 :
+	   active_feature_t::cmp (&a->feature, &b->feature);
+  }
+};
+
+struct range_record_t {
+  CTFontRef font;
+  unsigned int index_first; /* == start */
+  unsigned int index_last;  /* == end - 1 */
+};
+
+
+hb_bool_t
+_hb_coretext_shape (hb_shape_plan_t    *shape_plan,
+		    hb_font_t          *font,
+		    hb_buffer_t        *buffer,
+		    const hb_feature_t *features,
+		    unsigned int        num_features)
+{
+  hb_face_t *face = font->face;
+  CGFontRef cg_font = (CGFontRef) (const void *) face->data.coretext;
+  CTFontRef ct_font = (CTFontRef) hb_coretext_font_data_sync (font);
+
+  CGFloat ct_font_size = CTFontGetSize (ct_font);
+  CGFloat x_mult = (CGFloat) font->x_scale / ct_font_size;
+  CGFloat y_mult = (CGFloat) font->y_scale / ct_font_size;
+
+  /* Attach marks to their bases, to match the 'ot' shaper.
+   * Adapted from a very old version of hb-ot-shape:hb_form_clusters().
+   * Note that this only makes us be closer to the 'ot' shaper,
+   * but by no means the same.  For example, if there's
+   * B1 M1 B2 M2, and B1-B2 form a ligature, M2's cluster will
+   * continue pointing to B2 even though B2 was merged into B1's
+   * cluster... */
+  if (buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
+  {
+    hb_unicode_funcs_t *unicode = buffer->unicode;
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    for (unsigned int i = 1; i < count; i++)
+      if (HB_UNICODE_GENERAL_CATEGORY_IS_MARK (unicode->general_category (info[i].codepoint)))
+	buffer->merge_clusters (i - 1, i + 1);
+  }
+
+  hb_vector_t<feature_record_t> feature_records;
+  hb_vector_t<range_record_t> range_records;
+
+  /*
+   * Set up features.
+   * (copied + modified from code from hb-uniscribe.cc)
+   */
+  if (num_features)
+  {
+    /* Sort features by start/end events. */
+    hb_vector_t<feature_event_t> feature_events;
+    for (unsigned int i = 0; i < num_features; i++)
+    {
+      active_feature_t feature;
+
+#if MAC_OS_X_VERSION_MIN_REQUIRED < 101000
+      const hb_aat_feature_mapping_t * mapping = hb_aat_layout_find_feature_mapping (features[i].tag);
+      if (!mapping)
+	continue;
+
+      feature.rec.feature = mapping->aatFeatureType;
+      feature.rec.setting = features[i].value ? mapping->selectorToEnable : mapping->selectorToDisable;
+#else
+      feature.rec.feature = features[i].tag;
+      feature.rec.setting = features[i].value;
+#endif
+      feature.order = i;
+
+      feature_event_t *event;
+
+      event = feature_events.push ();
+      event->index = features[i].start;
+      event->start = true;
+      event->feature = feature;
+
+      event = feature_events.push ();
+      event->index = features[i].end;
+      event->start = false;
+      event->feature = feature;
+    }
+    feature_events.qsort ();
+    /* Add a strategic final event. */
+    {
+      active_feature_t feature;
+      feature.rec.feature = HB_TAG_NONE;
+      feature.rec.setting = 0;
+      feature.order = num_features + 1;
+
+      feature_event_t *event = feature_events.push ();
+      event->index = 0; /* This value does magic. */
+      event->start = false;
+      event->feature = feature;
+    }
+
+    /* Scan events and save features for each range. */
+    hb_vector_t<active_feature_t> active_features;
+    unsigned int last_index = 0;
+    for (unsigned int i = 0; i < feature_events.length; i++)
+    {
+      feature_event_t *event = &feature_events[i];
+
+      if (event->index != last_index)
+      {
+	/* Save a snapshot of active features and the range. */
+	range_record_t *range = range_records.push ();
+
+	if (active_features.length)
+	{
+	  CFMutableArrayRef features_array = CFArrayCreateMutable(kCFAllocatorDefault, 0, &kCFTypeArrayCallBacks);
+
+	  /* TODO sort and resolve conflicting features? */
+	  /* active_features.qsort (); */
+	  for (unsigned int j = 0; j < active_features.length; j++)
+	  {
+#if MAC_OS_X_VERSION_MIN_REQUIRED < 101000
+	    CFStringRef keys[] = {
+	      kCTFontFeatureTypeIdentifierKey,
+	      kCTFontFeatureSelectorIdentifierKey
+	    };
+	    CFNumberRef values[] = {
+	      CFNumberCreate (kCFAllocatorDefault, kCFNumberIntType, &active_features[j].rec.feature),
+	      CFNumberCreate (kCFAllocatorDefault, kCFNumberIntType, &active_features[j].rec.setting)
+	    };
+#else
+	    char tag[5] = {HB_UNTAG (active_features[j].rec.feature)};
+	    CFTypeRef keys[] = {
+	      kCTFontOpenTypeFeatureTag,
+	      kCTFontOpenTypeFeatureValue
+	    };
+	    CFTypeRef values[] = {
+	      CFStringCreateWithCString (kCFAllocatorDefault, tag, kCFStringEncodingASCII),
+	      CFNumberCreate (kCFAllocatorDefault, kCFNumberIntType, &active_features[j].rec.setting)
+	    };
+#endif
+	    static_assert ((ARRAY_LENGTH_CONST (keys) == ARRAY_LENGTH_CONST (values)), "");
+	    CFDictionaryRef dict = CFDictionaryCreate (kCFAllocatorDefault,
+						       (const void **) keys,
+						       (const void **) values,
+						       ARRAY_LENGTH (keys),
+						       &kCFTypeDictionaryKeyCallBacks,
+						       &kCFTypeDictionaryValueCallBacks);
+	    for (unsigned int i = 0; i < ARRAY_LENGTH (values); i++)
+	      CFRelease (values[i]);
+
+	    CFArrayAppendValue (features_array, dict);
+	    CFRelease (dict);
+
+	  }
+
+	  CFDictionaryRef attributes = CFDictionaryCreate (kCFAllocatorDefault,
+							   (const void **) &kCTFontFeatureSettingsAttribute,
+							   (const void **) &features_array,
+							   1,
+							   &kCFTypeDictionaryKeyCallBacks,
+							   &kCFTypeDictionaryValueCallBacks);
+	  CFRelease (features_array);
+
+	  CTFontDescriptorRef font_desc = CTFontDescriptorCreateWithAttributes (attributes);
+	  CFRelease (attributes);
+
+	  range->font = CTFontCreateCopyWithAttributes (ct_font, 0.0, nullptr, font_desc);
+	  CFRelease (font_desc);
+	}
+	else
+	{
+	  range->font = nullptr;
+	}
+
+	range->index_first = last_index;
+	range->index_last  = event->index - 1;
+
+	last_index = event->index;
+      }
+
+      if (event->start)
+      {
+	active_features.push (event->feature);
+      } else {
+	active_feature_t *feature = active_features.find (&event->feature);
+	if (feature)
+	  active_features.remove (feature - active_features.arrayZ);
+      }
+    }
+  }
+
+  unsigned int scratch_size;
+  hb_buffer_t::scratch_buffer_t *scratch = buffer->get_scratch_buffer (&scratch_size);
+
+#define ALLOCATE_ARRAY(Type, name, len, on_no_room) \
+  Type *name = (Type *) scratch; \
+  do { \
+    unsigned int _consumed = DIV_CEIL ((len) * sizeof (Type), sizeof (*scratch)); \
+    if (unlikely (_consumed > scratch_size)) \
+    { \
+      on_no_room; \
+      assert (0); \
+    } \
+    scratch += _consumed; \
+    scratch_size -= _consumed; \
+  } while (0)
+
+  ALLOCATE_ARRAY (UniChar, pchars, buffer->len * 2, ((void)nullptr) /*nothing*/);
+  unsigned int chars_len = 0;
+  for (unsigned int i = 0; i < buffer->len; i++) {
+    hb_codepoint_t c = buffer->info[i].codepoint;
+    if (likely (c <= 0xFFFFu))
+      pchars[chars_len++] = c;
+    else if (unlikely (c > 0x10FFFFu))
+      pchars[chars_len++] = 0xFFFDu;
+    else {
+      pchars[chars_len++] = 0xD800u + ((c - 0x10000u) >> 10);
+      pchars[chars_len++] = 0xDC00u + ((c - 0x10000u) & ((1u << 10) - 1));
+    }
+  }
+
+  ALLOCATE_ARRAY (unsigned int, log_clusters, chars_len, ((void)nullptr) /*nothing*/);
+  chars_len = 0;
+  for (unsigned int i = 0; i < buffer->len; i++)
+  {
+    hb_codepoint_t c = buffer->info[i].codepoint;
+    unsigned int cluster = buffer->info[i].cluster;
+    log_clusters[chars_len++] = cluster;
+    if (hb_in_range (c, 0x10000u, 0x10FFFFu))
+      log_clusters[chars_len++] = cluster; /* Surrogates. */
+  }
+
+#define FAIL(...) \
+  HB_STMT_START { \
+    DEBUG_MSG (CORETEXT, nullptr, __VA_ARGS__); \
+    ret = false; \
+    goto fail; \
+  } HB_STMT_END
+
+  bool ret = true;
+  CFStringRef string_ref = nullptr;
+  CTLineRef line = nullptr;
+
+  if (false)
+  {
+resize_and_retry:
+    DEBUG_MSG (CORETEXT, buffer, "Buffer resize");
+    /* string_ref uses the scratch-buffer for backing store, and line references
+     * string_ref (via attr_string).  We must release those before resizing buffer. */
+    assert (string_ref);
+    assert (line);
+    CFRelease (string_ref);
+    CFRelease (line);
+    string_ref = nullptr;
+    line = nullptr;
+
+    /* Get previous start-of-scratch-area, that we use later for readjusting
+     * our existing scratch arrays. */
+    unsigned int old_scratch_used;
+    hb_buffer_t::scratch_buffer_t *old_scratch;
+    old_scratch = buffer->get_scratch_buffer (&old_scratch_used);
+    old_scratch_used = scratch - old_scratch;
+
+    if (unlikely (!buffer->ensure (buffer->allocated * 2)))
+      FAIL ("Buffer resize failed");
+
+    /* Adjust scratch, pchars, and log_cluster arrays.  This is ugly, but really the
+     * cleanest way to do without completely restructuring the rest of this shaper. */
+    scratch = buffer->get_scratch_buffer (&scratch_size);
+    pchars = reinterpret_cast<UniChar *> (((char *) scratch + ((char *) pchars - (char *) old_scratch)));
+    log_clusters = reinterpret_cast<unsigned int *> (((char *) scratch + ((char *) log_clusters - (char *) old_scratch)));
+    scratch += old_scratch_used;
+    scratch_size -= old_scratch_used;
+  }
+  {
+    string_ref = CFStringCreateWithCharactersNoCopy (nullptr,
+						     pchars, chars_len,
+						     kCFAllocatorNull);
+    if (unlikely (!string_ref))
+      FAIL ("CFStringCreateWithCharactersNoCopy failed");
+
+    /* Create an attributed string, populate it, and create a line from it, then release attributed string. */
+    {
+      CFMutableAttributedStringRef attr_string = CFAttributedStringCreateMutable (kCFAllocatorDefault,
+										  chars_len);
+      if (unlikely (!attr_string))
+	FAIL ("CFAttributedStringCreateMutable failed");
+      CFAttributedStringReplaceString (attr_string, CFRangeMake (0, 0), string_ref);
+      if (HB_DIRECTION_IS_VERTICAL (buffer->props.direction))
+      {
+	CFAttributedStringSetAttribute (attr_string, CFRangeMake (0, chars_len),
+					kCTVerticalFormsAttributeName, kCFBooleanTrue);
+      }
+
+      if (buffer->props.language)
+      {
+/* What's the iOS equivalent of this check?
+ * The symbols was introduced in iOS 7.0.
+ * At any rate, our fallback is safe and works fine. */
+#if !(defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE) && MAC_OS_X_VERSION_MIN_REQUIRED < 1090
+#  define kCTLanguageAttributeName CFSTR ("NSLanguage")
+#endif
+	CFStringRef lang = CFStringCreateWithCStringNoCopy (kCFAllocatorDefault,
+							    hb_language_to_string (buffer->props.language),
+							    kCFStringEncodingUTF8,
+							    kCFAllocatorNull);
+	if (unlikely (!lang))
+	{
+	  CFRelease (attr_string);
+	  FAIL ("CFStringCreateWithCStringNoCopy failed");
+	}
+	CFAttributedStringSetAttribute (attr_string, CFRangeMake (0, chars_len),
+					kCTLanguageAttributeName, lang);
+	CFRelease (lang);
+      }
+      CFAttributedStringSetAttribute (attr_string, CFRangeMake (0, chars_len),
+				      kCTFontAttributeName, ct_font);
+
+      if (num_features && range_records.length)
+      {
+	unsigned int start = 0;
+	range_record_t *last_range = &range_records[0];
+	for (unsigned int k = 0; k < chars_len; k++)
+	{
+	  range_record_t *range = last_range;
+	  while (log_clusters[k] < range->index_first)
+	    range--;
+	  while (log_clusters[k] > range->index_last)
+	    range++;
+	  if (range != last_range)
+	  {
+	    if (last_range->font)
+	      CFAttributedStringSetAttribute (attr_string, CFRangeMake (start, k - start),
+					      kCTFontAttributeName, last_range->font);
+
+	    start = k;
+	  }
+
+	  last_range = range;
+	}
+	if (start != chars_len && last_range->font)
+	  CFAttributedStringSetAttribute (attr_string, CFRangeMake (start, chars_len - start),
+					  kCTFontAttributeName, last_range->font);
+      }
+      /* Enable/disable kern if requested.
+       *
+       * Note: once kern is disabled, reenabling it doesn't currently seem to work in CoreText.
+       */
+      if (num_features)
+      {
+	unsigned int zeroint = 0;
+	CFNumberRef zero = CFNumberCreate (kCFAllocatorDefault, kCFNumberIntType, &zeroint);
+	for (unsigned int i = 0; i < num_features; i++)
+	{
+	  const hb_feature_t &feature = features[i];
+	  if (feature.tag == HB_TAG('k','e','r','n') &&
+	      feature.start < chars_len && feature.start < feature.end)
+	  {
+	    CFRange feature_range = CFRangeMake (feature.start,
+						 hb_min (feature.end, chars_len) - feature.start);
+	    if (feature.value)
+	      CFAttributedStringRemoveAttribute (attr_string, feature_range, kCTKernAttributeName);
+	    else
+	      CFAttributedStringSetAttribute (attr_string, feature_range, kCTKernAttributeName, zero);
+	  }
+	}
+	CFRelease (zero);
+      }
+
+      int level = HB_DIRECTION_IS_FORWARD (buffer->props.direction) ? 0 : 1;
+      CFNumberRef level_number = CFNumberCreate (kCFAllocatorDefault, kCFNumberIntType, &level);
+#if !(defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE) && MAC_OS_X_VERSION_MIN_REQUIRED < 1060
+      extern const CFStringRef kCTTypesetterOptionForcedEmbeddingLevel;
+#endif
+      CFDictionaryRef options = CFDictionaryCreate (kCFAllocatorDefault,
+						    (const void **) &kCTTypesetterOptionForcedEmbeddingLevel,
+						    (const void **) &level_number,
+						    1,
+						    &kCFTypeDictionaryKeyCallBacks,
+						    &kCFTypeDictionaryValueCallBacks);
+      CFRelease (level_number);
+      if (unlikely (!options))
+      {
+	CFRelease (attr_string);
+	FAIL ("CFDictionaryCreate failed");
+      }
+
+      CTTypesetterRef typesetter = CTTypesetterCreateWithAttributedStringAndOptions (attr_string, options);
+      CFRelease (options);
+      CFRelease (attr_string);
+      if (unlikely (!typesetter))
+	FAIL ("CTTypesetterCreateWithAttributedStringAndOptions failed");
+
+      line = CTTypesetterCreateLine (typesetter, CFRangeMake(0, 0));
+      CFRelease (typesetter);
+      if (unlikely (!line))
+	FAIL ("CTTypesetterCreateLine failed");
+    }
+
+    CFArrayRef glyph_runs = CTLineGetGlyphRuns (line);
+    unsigned int num_runs = CFArrayGetCount (glyph_runs);
+    DEBUG_MSG (CORETEXT, nullptr, "Num runs: %d", num_runs);
+
+    buffer->len = 0;
+    uint32_t status_and = ~0, status_or = 0;
+    double advances_so_far = 0;
+    /* For right-to-left runs, CoreText returns the glyphs positioned such that
+     * any trailing whitespace is to the left of (0,0).  Adjust coordinate system
+     * to fix for that.  Test with any RTL string with trailing spaces.
+     * https://crbug.com/469028
+     */
+    if (HB_DIRECTION_IS_BACKWARD (buffer->props.direction))
+    {
+      advances_so_far -= CTLineGetTrailingWhitespaceWidth (line);
+      if (HB_DIRECTION_IS_VERTICAL (buffer->props.direction))
+	  advances_so_far = -advances_so_far;
+    }
+
+    const CFRange range_all = CFRangeMake (0, 0);
+
+    for (unsigned int i = 0; i < num_runs; i++)
+    {
+      CTRunRef run = static_cast<CTRunRef>(CFArrayGetValueAtIndex (glyph_runs, i));
+      CTRunStatus run_status = CTRunGetStatus (run);
+      status_or  |= run_status;
+      status_and &= run_status;
+      DEBUG_MSG (CORETEXT, run, "CTRunStatus: %x", run_status);
+      double run_advance = CTRunGetTypographicBounds (run, range_all, nullptr, nullptr, nullptr);
+      if (HB_DIRECTION_IS_VERTICAL (buffer->props.direction))
+	  run_advance = -run_advance;
+      DEBUG_MSG (CORETEXT, run, "Run advance: %g", run_advance);
+
+      /* CoreText does automatic font fallback (AKA "cascading") for  characters
+       * not supported by the requested font, and provides no way to turn it off,
+       * so we must detect if the returned run uses a font other than the requested
+       * one and fill in the buffer with .notdef glyphs instead of random glyph
+       * indices from a different font.
+       */
+      CFDictionaryRef attributes = CTRunGetAttributes (run);
+      CTFontRef run_ct_font = static_cast<CTFontRef>(CFDictionaryGetValue (attributes, kCTFontAttributeName));
+      if (!CFEqual (run_ct_font, ct_font))
+      {
+	/* The run doesn't use our main font instance.  We have to figure out
+	 * whether font fallback happened, or this is just CoreText giving us
+	 * another CTFont using the same underlying CGFont.  CoreText seems
+	 * to do that in a variety of situations, one of which being vertical
+	 * text, but also perhaps for caching reasons.
+	 *
+	 * First, see if it uses any of our subfonts created to set font features...
+	 *
+	 * Next, compare the CGFont to the one we used to create our fonts.
+	 * Even this doesn't work all the time.
+	 *
+	 * Finally, we compare PS names, which I don't think are unique...
+	 *
+	 * Looks like if we really want to be sure here we have to modify the
+	 * font to change the name table, similar to what we do in the uniscribe
+	 * backend.
+	 *
+	 * However, even that wouldn't work if we were passed in the CGFont to
+	 * construct a hb_face to begin with.
+	 *
+	 * See: https://github.com/harfbuzz/harfbuzz/pull/36
+	 *
+	 * Also see: https://bugs.chromium.org/p/chromium/issues/detail?id=597098
+	 */
+	bool matched = false;
+	for (unsigned int i = 0; i < range_records.length; i++)
+	  if (range_records[i].font && CFEqual (run_ct_font, range_records[i].font))
+	  {
+	    matched = true;
+	    break;
+	  }
+	if (!matched)
+	{
+	  CGFontRef run_cg_font = CTFontCopyGraphicsFont (run_ct_font, nullptr);
+	  if (run_cg_font)
+	  {
+	    matched = CFEqual (run_cg_font, cg_font);
+	    CFRelease (run_cg_font);
+	  }
+	}
+	if (!matched)
+	{
+	  CFStringRef font_ps_name = CTFontCopyName (ct_font, kCTFontPostScriptNameKey);
+	  CFStringRef run_ps_name = CTFontCopyName (run_ct_font, kCTFontPostScriptNameKey);
+	  CFComparisonResult result = CFStringCompare (run_ps_name, font_ps_name, 0);
+	  CFRelease (run_ps_name);
+	  CFRelease (font_ps_name);
+	  if (result == kCFCompareEqualTo)
+	    matched = true;
+	}
+	if (!matched)
+	{
+	  CFRange range = CTRunGetStringRange (run);
+	  DEBUG_MSG (CORETEXT, run, "Run used fallback font: %ld..%ld",
+		     range.location, range.location + range.length);
+	  if (!buffer->ensure_inplace (buffer->len + range.length))
+	    goto resize_and_retry;
+	  hb_glyph_info_t *info = buffer->info + buffer->len;
+
+	  hb_codepoint_t notdef = 0;
+	  hb_direction_t dir = buffer->props.direction;
+	  hb_position_t x_advance, y_advance, x_offset, y_offset;
+	  hb_font_get_glyph_advance_for_direction (font, notdef, dir, &x_advance, &y_advance);
+	  hb_font_get_glyph_origin_for_direction (font, notdef, dir, &x_offset, &y_offset);
+	  hb_position_t advance = x_advance + y_advance;
+	  x_offset = -x_offset;
+	  y_offset = -y_offset;
+
+	  unsigned int old_len = buffer->len;
+	  for (CFIndex j = range.location; j < range.location + range.length; j++)
+	  {
+	      UniChar ch = CFStringGetCharacterAtIndex (string_ref, j);
+	      if (hb_in_range<UniChar> (ch, 0xDC00u, 0xDFFFu) && range.location < j)
+	      {
+		ch = CFStringGetCharacterAtIndex (string_ref, j - 1);
+		if (hb_in_range<UniChar> (ch, 0xD800u, 0xDBFFu))
+		  /* This is the second of a surrogate pair.  Don't need .notdef
+		   * for this one. */
+		  continue;
+	      }
+	      if (buffer->unicode->is_default_ignorable (ch))
+		continue;
+
+	      info->codepoint = notdef;
+	      info->cluster = log_clusters[j];
+
+	      info->mask = advance;
+	      info->var1.i32 = x_offset;
+	      info->var2.i32 = y_offset;
+
+	      info++;
+	      buffer->len++;
+	  }
+	  if (HB_DIRECTION_IS_BACKWARD (buffer->props.direction))
+	    buffer->reverse_range (old_len, buffer->len);
+	  advances_so_far += run_advance;
+	  continue;
+	}
+      }
+
+      unsigned int num_glyphs = CTRunGetGlyphCount (run);
+      if (num_glyphs == 0)
+	continue;
+
+      if (!buffer->ensure_inplace (buffer->len + num_glyphs))
+	goto resize_and_retry;
+
+      hb_glyph_info_t *run_info = buffer->info + buffer->len;
+
+      /* Testing used to indicate that CTRunGetGlyphsPtr, etc (almost?) always
+       * succeed, and so copying data to our own buffer will be rare.  Reports
+       * have it that this changed in OS X 10.10 Yosemite, and nullptr is returned
+       * frequently.  At any rate, we can test that codepath by setting USE_PTR
+       * to false. */
+
+#define USE_PTR true
+
+#define SCRATCH_SAVE() \
+  unsigned int scratch_size_saved = scratch_size; \
+  hb_buffer_t::scratch_buffer_t *scratch_saved = scratch
+
+#define SCRATCH_RESTORE() \
+  scratch_size = scratch_size_saved; \
+  scratch = scratch_saved
+
+      { /* Setup glyphs */
+	SCRATCH_SAVE();
+	const CGGlyph* glyphs = USE_PTR ? CTRunGetGlyphsPtr (run) : nullptr;
+	if (!glyphs) {
+	  ALLOCATE_ARRAY (CGGlyph, glyph_buf, num_glyphs, goto resize_and_retry);
+	  CTRunGetGlyphs (run, range_all, glyph_buf);
+	  glyphs = glyph_buf;
+	}
+	const CFIndex* string_indices = USE_PTR ? CTRunGetStringIndicesPtr (run) : nullptr;
+	if (!string_indices) {
+	  ALLOCATE_ARRAY (CFIndex, index_buf, num_glyphs, goto resize_and_retry);
+	  CTRunGetStringIndices (run, range_all, index_buf);
+	  string_indices = index_buf;
+	}
+	hb_glyph_info_t *info = run_info;
+	for (unsigned int j = 0; j < num_glyphs; j++)
+	{
+	  info->codepoint = glyphs[j];
+	  info->cluster = log_clusters[string_indices[j]];
+	  info++;
+	}
+	SCRATCH_RESTORE();
+      }
+      {
+	/* Setup positions.
+	 * Note that CoreText does not return advances for glyphs.  As such,
+	 * for all but last glyph, we use the delta position to next glyph as
+	 * advance (in the advance direction only), and for last glyph we set
+	 * whatever is needed to make the whole run's advance add up. */
+	SCRATCH_SAVE();
+	const CGPoint* positions = USE_PTR ? CTRunGetPositionsPtr (run) : nullptr;
+	if (!positions) {
+	  ALLOCATE_ARRAY (CGPoint, position_buf, num_glyphs, goto resize_and_retry);
+	  CTRunGetPositions (run, range_all, position_buf);
+	  positions = position_buf;
+	}
+	hb_glyph_info_t *info = run_info;
+	if (HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction))
+	{
+	  hb_position_t x_offset = (positions[0].x - advances_so_far) * x_mult;
+	  for (unsigned int j = 0; j < num_glyphs; j++)
+	  {
+	    double advance;
+	    if (likely (j + 1 < num_glyphs))
+	      advance = positions[j + 1].x - positions[j].x;
+	    else /* last glyph */
+	      advance = run_advance - (positions[j].x - positions[0].x);
+	    info->mask = advance * x_mult;
+	    info->var1.i32 = x_offset;
+	    info->var2.i32 = positions[j].y * y_mult;
+	    info++;
+	  }
+	}
+	else
+	{
+	  hb_position_t y_offset = (positions[0].y - advances_so_far) * y_mult;
+	  for (unsigned int j = 0; j < num_glyphs; j++)
+	  {
+	    double advance;
+	    if (likely (j + 1 < num_glyphs))
+	      advance = positions[j + 1].y - positions[j].y;
+	    else /* last glyph */
+	      advance = run_advance - (positions[j].y - positions[0].y);
+	    info->mask = advance * y_mult;
+	    info->var1.i32 = positions[j].x * x_mult;
+	    info->var2.i32 = y_offset;
+	    info++;
+	  }
+	}
+	SCRATCH_RESTORE();
+	advances_so_far += run_advance;
+      }
+#undef SCRATCH_RESTORE
+#undef SCRATCH_SAVE
+#undef USE_PTR
+#undef ALLOCATE_ARRAY
+
+      buffer->len += num_glyphs;
+    }
+
+    /* Mac OS 10.6 doesn't have kCTTypesetterOptionForcedEmbeddingLevel,
+     * or if it does, it doesn't respect it.  So we get runs with wrong
+     * directions.  As such, disable the assert...  It wouldn't crash, but
+     * cursoring will be off...
+     *
+     * https://crbug.com/419769
+     */
+    if (false)
+    {
+      /* Make sure all runs had the expected direction. */
+      HB_UNUSED bool backward = HB_DIRECTION_IS_BACKWARD (buffer->props.direction);
+      assert (bool (status_and & kCTRunStatusRightToLeft) == backward);
+      assert (bool (status_or  & kCTRunStatusRightToLeft) == backward);
+    }
+
+    buffer->clear_positions ();
+
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    hb_glyph_position_t *pos = buffer->pos;
+    if (HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction))
+      for (unsigned int i = 0; i < count; i++)
+      {
+	pos->x_advance = info->mask;
+	pos->x_offset = info->var1.i32;
+	pos->y_offset = info->var2.i32;
+
+	info++, pos++;
+      }
+    else
+      for (unsigned int i = 0; i < count; i++)
+      {
+	pos->y_advance = info->mask;
+	pos->x_offset = info->var1.i32;
+	pos->y_offset = info->var2.i32;
+
+	info++, pos++;
+      }
+
+    /* Fix up clusters so that we never return out-of-order indices;
+     * if core text has reordered glyphs, we'll merge them to the
+     * beginning of the reordered cluster.  CoreText is nice enough
+     * to tell us whenever it has produced nonmonotonic results...
+     * Note that we assume the input clusters were nonmonotonic to
+     * begin with.
+     *
+     * This does *not* mean we'll form the same clusters as Uniscribe
+     * or the native OT backend, only that the cluster indices will be
+     * monotonic in the output buffer. */
+    if (count > 1 && (status_or & kCTRunStatusNonMonotonic))
+    {
+      hb_glyph_info_t *info = buffer->info;
+      if (HB_DIRECTION_IS_FORWARD (buffer->props.direction))
+      {
+	unsigned int cluster = info[count - 1].cluster;
+	for (unsigned int i = count - 1; i > 0; i--)
+	{
+	  cluster = hb_min (cluster, info[i - 1].cluster);
+	  info[i - 1].cluster = cluster;
+	}
+      }
+      else
+      {
+	unsigned int cluster = info[0].cluster;
+	for (unsigned int i = 1; i < count; i++)
+	{
+	  cluster = hb_min (cluster, info[i].cluster);
+	  info[i].cluster = cluster;
+	}
+      }
+    }
+  }
+
+  buffer->unsafe_to_break_all ();
+
+#undef FAIL
+
+fail:
+  if (string_ref)
+    CFRelease (string_ref);
+  if (line)
+    CFRelease (line);
+
+  for (unsigned int i = 0; i < range_records.length; i++)
+    if (range_records[i].font)
+      CFRelease (range_records[i].font);
+
+  return ret;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-coretext.h b/thirdparty/harfbuzz/src/hb-coretext.h
new file mode 100644
index 0000000000..e53dbaf2c7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-coretext.h
@@ -0,0 +1,96 @@
+/*
+ * Copyright © 2012  Mozilla Foundation.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ */
+
+#ifndef HB_CORETEXT_H
+#define HB_CORETEXT_H
+
+#include "hb.h"
+
+#include <TargetConditionals.h>
+#if TARGET_OS_IPHONE
+#  include <CoreText/CoreText.h>
+#  include <CoreGraphics/CoreGraphics.h>
+#else
+#  include <ApplicationServices/ApplicationServices.h>
+#endif
+
+HB_BEGIN_DECLS
+
+
+/**
+ * HB_CORETEXT_TAG_MORT:
+ *
+ * The #hb_tag_t tag for the `mort` (glyph metamorphosis) table,
+ * which holds AAT features. 
+ *
+ * For more information, see 
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6mort.html
+ *
+ **/
+#define HB_CORETEXT_TAG_MORT HB_TAG('m','o','r','t')
+
+/**
+ * HB_CORETEXT_TAG_MORX:
+ *
+ * The #hb_tag_t tag for the `morx` (extended glyph metamorphosis)
+ * table, which holds AAT features. 
+ *
+ * For more information, see
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6morx.html
+ *
+ **/
+#define HB_CORETEXT_TAG_MORX HB_TAG('m','o','r','x')
+
+/**
+ * HB_CORETEXT_TAG_KERX:
+ *
+ * The #hb_tag_t tag for the `kerx` (extended kerning) table, which
+ * holds AAT kerning information. 
+ *
+ * For more information, see 
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6kerx.html
+ *
+ **/
+#define HB_CORETEXT_TAG_KERX HB_TAG('k','e','r','x')
+
+
+HB_EXTERN hb_face_t *
+hb_coretext_face_create (CGFontRef cg_font);
+
+HB_EXTERN hb_font_t *
+hb_coretext_font_create (CTFontRef ct_font);
+
+
+HB_EXTERN CGFontRef
+hb_coretext_face_get_cg_font (hb_face_t *face);
+
+HB_EXTERN CTFontRef
+hb_coretext_font_get_ct_font (hb_font_t *font);
+
+
+HB_END_DECLS
+
+#endif /* HB_CORETEXT_H */
diff --git a/thirdparty/harfbuzz/src/hb-debug.hh b/thirdparty/harfbuzz/src/hb-debug.hh
new file mode 100644
index 0000000000..ec3a1ff211
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-debug.hh
@@ -0,0 +1,459 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_DEBUG_HH
+#define HB_DEBUG_HH
+
+#include "hb.hh"
+#include "hb-atomic.hh"
+#include "hb-algs.hh"
+
+
+#ifndef HB_DEBUG
+#define HB_DEBUG 0
+#endif
+
+
+/*
+ * Global runtime options.
+ */
+
+struct hb_options_t
+{
+  bool unused : 1; /* In-case sign bit is here. */
+  bool initialized : 1;
+  bool uniscribe_bug_compatible : 1;
+};
+
+union hb_options_union_t {
+  int i;
+  hb_options_t opts;
+};
+static_assert ((sizeof (hb_atomic_int_t) >= sizeof (hb_options_union_t)), "");
+
+HB_INTERNAL void
+_hb_options_init ();
+
+extern HB_INTERNAL hb_atomic_int_t _hb_options;
+
+static inline hb_options_t
+hb_options ()
+{
+#ifdef HB_NO_GETENV
+  return hb_options_t ();
+#endif
+  /* Make a local copy, so we can access bitfield threadsafely. */
+  hb_options_union_t u;
+  u.i = _hb_options.get_relaxed ();
+
+  if (unlikely (!u.i))
+  {
+    _hb_options_init ();
+    u.i = _hb_options.get_relaxed ();
+  }
+
+  return u.opts;
+}
+
+
+/*
+ * Debug output (needs enabling at compile time.)
+ */
+
+static inline bool
+_hb_debug (unsigned int level,
+	   unsigned int max_level)
+{
+  return level < max_level;
+}
+
+#define DEBUG_LEVEL_ENABLED(WHAT, LEVEL) (_hb_debug ((LEVEL), HB_DEBUG_##WHAT))
+#define DEBUG_ENABLED(WHAT) (DEBUG_LEVEL_ENABLED (WHAT, 0))
+
+static inline void
+_hb_print_func (const char *func)
+{
+  if (func)
+  {
+    unsigned int func_len = strlen (func);
+    /* Skip "static" */
+    if (0 == strncmp (func, "static ", 7))
+      func += 7;
+    /* Skip "typename" */
+    if (0 == strncmp (func, "typename ", 9))
+      func += 9;
+    /* Skip return type */
+    const char *space = strchr (func, ' ');
+    if (space)
+      func = space + 1;
+    /* Skip parameter list */
+    const char *paren = strchr (func, '(');
+    if (paren)
+      func_len = paren - func;
+    fprintf (stderr, "%.*s", func_len, func);
+  }
+}
+
+template <int max_level> static inline void
+_hb_debug_msg_va (const char *what,
+		  const void *obj,
+		  const char *func,
+		  bool indented,
+		  unsigned int level,
+		  int level_dir,
+		  const char *message,
+		  va_list ap) HB_PRINTF_FUNC(7, 0);
+template <int max_level> static inline void
+_hb_debug_msg_va (const char *what,
+		  const void *obj,
+		  const char *func,
+		  bool indented,
+		  unsigned int level,
+		  int level_dir,
+		  const char *message,
+		  va_list ap)
+{
+  if (!_hb_debug (level, max_level))
+    return;
+
+  fprintf (stderr, "%-10s", what ? what : "");
+
+  if (obj)
+    fprintf (stderr, "(%*p) ", (unsigned int) (2 * sizeof (void *)), obj);
+  else
+    fprintf (stderr, " %*s  ", (unsigned int) (2 * sizeof (void *)), "");
+
+  if (indented) {
+#define VBAR	"\342\224\202"	/* U+2502 BOX DRAWINGS LIGHT VERTICAL */
+#define VRBAR	"\342\224\234"	/* U+251C BOX DRAWINGS LIGHT VERTICAL AND RIGHT */
+#define DLBAR	"\342\225\256"	/* U+256E BOX DRAWINGS LIGHT ARC DOWN AND LEFT */
+#define ULBAR	"\342\225\257"	/* U+256F BOX DRAWINGS LIGHT ARC UP AND LEFT */
+#define LBAR	"\342\225\264"	/* U+2574 BOX DRAWINGS LIGHT LEFT */
+    static const char bars[] =
+      VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR
+      VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR
+      VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR
+      VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR
+      VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR VBAR;
+    fprintf (stderr, "%2u %s" VRBAR "%s",
+	     level,
+	     bars + sizeof (bars) - 1 - hb_min ((unsigned int) sizeof (bars) - 1, (unsigned int) (sizeof (VBAR) - 1) * level),
+	     level_dir ? (level_dir > 0 ? DLBAR : ULBAR) : LBAR);
+  } else
+    fprintf (stderr, "   " VRBAR LBAR);
+
+  _hb_print_func (func);
+
+  if (message)
+  {
+    fprintf (stderr, ": ");
+    vfprintf (stderr, message, ap);
+  }
+
+  fprintf (stderr, "\n");
+}
+template <> inline void HB_PRINTF_FUNC(7, 0)
+_hb_debug_msg_va<0> (const char *what HB_UNUSED,
+		     const void *obj HB_UNUSED,
+		     const char *func HB_UNUSED,
+		     bool indented HB_UNUSED,
+		     unsigned int level HB_UNUSED,
+		     int level_dir HB_UNUSED,
+		     const char *message HB_UNUSED,
+		     va_list ap HB_UNUSED) {}
+
+template <int max_level> static inline void
+_hb_debug_msg (const char *what,
+	       const void *obj,
+	       const char *func,
+	       bool indented,
+	       unsigned int level,
+	       int level_dir,
+	       const char *message,
+	       ...) HB_PRINTF_FUNC(7, 8);
+template <int max_level> static inline void HB_PRINTF_FUNC(7, 8)
+_hb_debug_msg (const char *what,
+	       const void *obj,
+	       const char *func,
+	       bool indented,
+	       unsigned int level,
+	       int level_dir,
+	       const char *message,
+	       ...)
+{
+  va_list ap;
+  va_start (ap, message);
+  _hb_debug_msg_va<max_level> (what, obj, func, indented, level, level_dir, message, ap);
+  va_end (ap);
+}
+template <> inline void
+_hb_debug_msg<0> (const char *what HB_UNUSED,
+		  const void *obj HB_UNUSED,
+		  const char *func HB_UNUSED,
+		  bool indented HB_UNUSED,
+		  unsigned int level HB_UNUSED,
+		  int level_dir HB_UNUSED,
+		  const char *message HB_UNUSED,
+		  ...) HB_PRINTF_FUNC(7, 8);
+template <> inline void HB_PRINTF_FUNC(7, 8)
+_hb_debug_msg<0> (const char *what HB_UNUSED,
+		  const void *obj HB_UNUSED,
+		  const char *func HB_UNUSED,
+		  bool indented HB_UNUSED,
+		  unsigned int level HB_UNUSED,
+		  int level_dir HB_UNUSED,
+		  const char *message HB_UNUSED,
+		  ...) {}
+
+#define DEBUG_MSG_LEVEL(WHAT, OBJ, LEVEL, LEVEL_DIR, ...)	_hb_debug_msg<HB_DEBUG_##WHAT> (#WHAT, (OBJ), nullptr,    true, (LEVEL), (LEVEL_DIR), __VA_ARGS__)
+#define DEBUG_MSG(WHAT, OBJ, ...)				_hb_debug_msg<HB_DEBUG_##WHAT> (#WHAT, (OBJ), nullptr,    false, 0, 0, __VA_ARGS__)
+#define DEBUG_MSG_FUNC(WHAT, OBJ, ...)				_hb_debug_msg<HB_DEBUG_##WHAT> (#WHAT, (OBJ), HB_FUNC, false, 0, 0, __VA_ARGS__)
+
+
+/*
+ * Printer
+ */
+
+template <typename T>
+struct hb_printer_t {
+  const char *print (const T&) { return "something"; }
+};
+
+template <>
+struct hb_printer_t<bool> {
+  const char *print (bool v) { return v ? "true" : "false"; }
+};
+
+template <>
+struct hb_printer_t<hb_empty_t> {
+  const char *print (hb_empty_t) { return ""; }
+};
+
+
+/*
+ * Trace
+ */
+
+template <typename T>
+static inline void _hb_warn_no_return (bool returned)
+{
+  if (unlikely (!returned)) {
+    fprintf (stderr, "OUCH, returned with no call to return_trace().  This is a bug, please report.\n");
+  }
+}
+template <>
+/*static*/ inline void _hb_warn_no_return<hb_empty_t> (bool returned HB_UNUSED)
+{}
+
+template <int max_level, typename ret_t>
+struct hb_auto_trace_t
+{
+  explicit inline hb_auto_trace_t (unsigned int *plevel_,
+				   const char *what_,
+				   const void *obj_,
+				   const char *func,
+				   const char *message,
+				   ...) HB_PRINTF_FUNC(6, 7)
+				   : plevel (plevel_), what (what_), obj (obj_), returned (false)
+  {
+    if (plevel) ++*plevel;
+
+    va_list ap;
+    va_start (ap, message);
+    _hb_debug_msg_va<max_level> (what, obj, func, true, plevel ? *plevel : 0, +1, message, ap);
+    va_end (ap);
+  }
+  ~hb_auto_trace_t ()
+  {
+    _hb_warn_no_return<ret_t> (returned);
+    if (!returned) {
+      _hb_debug_msg<max_level> (what, obj, nullptr, true, plevel ? *plevel : 1, -1, " ");
+    }
+    if (plevel) --*plevel;
+  }
+
+  template <typename T>
+  T ret (T&& v,
+	 const char *func = "",
+	 unsigned int line = 0)
+  {
+    if (unlikely (returned)) {
+      fprintf (stderr, "OUCH, double calls to return_trace().  This is a bug, please report.\n");
+      return hb_forward<T> (v);
+    }
+
+    _hb_debug_msg<max_level> (what, obj, func, true, plevel ? *plevel : 1, -1,
+			      "return %s (line %d)",
+			      hb_printer_t<decltype (v)>().print (v), line);
+    if (plevel) --*plevel;
+    plevel = nullptr;
+    returned = true;
+    return hb_forward<T> (v);
+  }
+
+  private:
+  unsigned int *plevel;
+  const char *what;
+  const void *obj;
+  bool returned;
+};
+template <typename ret_t> /* Make sure we don't use hb_auto_trace_t when not tracing. */
+struct hb_auto_trace_t<0, ret_t>
+{
+  explicit inline hb_auto_trace_t (unsigned int *plevel_,
+				   const char *what_,
+				   const void *obj_,
+				   const char *func,
+				   const char *message,
+				   ...) HB_PRINTF_FUNC(6, 7) {}
+
+  template <typename T>
+  T ret (T&& v,
+	 const char *func HB_UNUSED = nullptr,
+	 unsigned int line HB_UNUSED = 0) { return hb_forward<T> (v); }
+};
+
+/* For disabled tracing; optimize out everything.
+ * https://github.com/harfbuzz/harfbuzz/pull/605 */
+template <typename ret_t>
+struct hb_no_trace_t {
+  template <typename T>
+  T ret (T&& v,
+	 const char *func HB_UNUSED = nullptr,
+	 unsigned int line HB_UNUSED = 0) { return hb_forward<T> (v); }
+};
+
+#define return_trace(RET) return trace.ret (RET, HB_FUNC, __LINE__)
+
+
+/*
+ * Instances.
+ */
+
+#ifndef HB_DEBUG_ARABIC
+#define HB_DEBUG_ARABIC (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_BLOB
+#define HB_DEBUG_BLOB (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_CORETEXT
+#define HB_DEBUG_CORETEXT (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_DIRECTWRITE
+#define HB_DEBUG_DIRECTWRITE (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_FT
+#define HB_DEBUG_FT (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_OBJECT
+#define HB_DEBUG_OBJECT (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_SHAPE_PLAN
+#define HB_DEBUG_SHAPE_PLAN (HB_DEBUG+0)
+#endif
+
+#ifndef HB_DEBUG_UNISCRIBE
+#define HB_DEBUG_UNISCRIBE (HB_DEBUG+0)
+#endif
+
+/*
+ * With tracing.
+ */
+
+#ifndef HB_DEBUG_APPLY
+#define HB_DEBUG_APPLY (HB_DEBUG+0)
+#endif
+#if HB_DEBUG_APPLY
+#define TRACE_APPLY(this) \
+	hb_auto_trace_t<HB_DEBUG_APPLY, bool> trace \
+	(&c->debug_depth, c->get_name (), this, HB_FUNC, \
+	 "idx %d gid %u lookup %d", \
+	 c->buffer->idx, c->buffer->cur().codepoint, (int) c->lookup_index)
+#else
+#define TRACE_APPLY(this) hb_no_trace_t<bool> trace
+#endif
+
+#ifndef HB_DEBUG_SANITIZE
+#define HB_DEBUG_SANITIZE (HB_DEBUG+0)
+#endif
+#if HB_DEBUG_SANITIZE
+#define TRACE_SANITIZE(this) \
+	hb_auto_trace_t<HB_DEBUG_SANITIZE, bool> trace \
+	(&c->debug_depth, c->get_name (), this, HB_FUNC, \
+	 " ")
+#else
+#define TRACE_SANITIZE(this) hb_no_trace_t<bool> trace
+#endif
+
+#ifndef HB_DEBUG_SERIALIZE
+#define HB_DEBUG_SERIALIZE (HB_DEBUG+0)
+#endif
+#if HB_DEBUG_SERIALIZE
+#define TRACE_SERIALIZE(this) \
+	hb_auto_trace_t<HB_DEBUG_SERIALIZE, bool> trace \
+	(&c->debug_depth, "SERIALIZE", c, HB_FUNC, \
+	 " ")
+#else
+#define TRACE_SERIALIZE(this) hb_no_trace_t<bool> trace
+#endif
+
+#ifndef HB_DEBUG_SUBSET
+#define HB_DEBUG_SUBSET (HB_DEBUG+0)
+#endif
+#if HB_DEBUG_SUBSET
+#define TRACE_SUBSET(this) \
+  hb_auto_trace_t<HB_DEBUG_SUBSET, bool> trace \
+  (&c->debug_depth, c->get_name (), this, HB_FUNC, \
+   " ")
+#else
+#define TRACE_SUBSET(this) hb_no_trace_t<bool> trace
+#endif
+
+#ifndef HB_DEBUG_DISPATCH
+#define HB_DEBUG_DISPATCH ( \
+	HB_DEBUG_APPLY + \
+	HB_DEBUG_SANITIZE + \
+	HB_DEBUG_SERIALIZE + \
+	HB_DEBUG_SUBSET + \
+	0)
+#endif
+#if HB_DEBUG_DISPATCH
+#define TRACE_DISPATCH(this, format) \
+	hb_auto_trace_t<context_t::max_debug_depth, typename context_t::return_t> trace \
+	(&c->debug_depth, c->get_name (), this, HB_FUNC, \
+	 "format %d", (int) format)
+#else
+#define TRACE_DISPATCH(this, format) hb_no_trace_t<typename context_t::return_t> trace
+#endif
+
+
+#endif /* HB_DEBUG_HH */
diff --git a/thirdparty/harfbuzz/src/hb-deprecated.h b/thirdparty/harfbuzz/src/hb-deprecated.h
new file mode 100644
index 0000000000..43f89a4c4e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-deprecated.h
@@ -0,0 +1,195 @@
+/*
+ * Copyright © 2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_DEPRECATED_H
+#define HB_DEPRECATED_H
+
+#include "hb-common.h"
+#include "hb-unicode.h"
+#include "hb-font.h"
+#include "hb-set.h"
+
+
+/**
+ * SECTION:hb-deprecated
+ * @title: hb-deprecated
+ * @short_description: Deprecated API
+ * @include: hb.h
+ *
+ * These API have been deprecated in favor of newer API, or because they
+ * were deemed unnecessary.
+ **/
+
+
+HB_BEGIN_DECLS
+
+#ifndef HB_DISABLE_DEPRECATED
+
+
+#define HB_SCRIPT_CANADIAN_ABORIGINAL		HB_SCRIPT_CANADIAN_SYLLABICS
+
+#define HB_BUFFER_FLAGS_DEFAULT			HB_BUFFER_FLAG_DEFAULT
+#define HB_BUFFER_SERIALIZE_FLAGS_DEFAULT	HB_BUFFER_SERIALIZE_FLAG_DEFAULT
+
+typedef hb_bool_t (*hb_font_get_glyph_func_t) (hb_font_t *font, void *font_data,
+					       hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+					       hb_codepoint_t *glyph,
+					       void *user_data);
+
+HB_EXTERN HB_DEPRECATED_FOR(hb_font_funcs_set_nominal_glyph_func and hb_font_funcs_set_variation_glyph_func) void
+hb_font_funcs_set_glyph_func (hb_font_funcs_t *ffuncs,
+			      hb_font_get_glyph_func_t func,
+			      void *user_data, hb_destroy_func_t destroy);
+
+HB_EXTERN HB_DEPRECATED void
+hb_set_invert (hb_set_t *set);
+
+/**
+ * hb_unicode_eastasian_width_func_t:
+ *
+ * Deprecated: 2.0.0
+ */
+typedef unsigned int			(*hb_unicode_eastasian_width_func_t)	(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      unicode,
+										 void               *user_data);
+
+/**
+ * hb_unicode_funcs_set_eastasian_width_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ * 
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+HB_EXTERN HB_DEPRECATED void
+hb_unicode_funcs_set_eastasian_width_func (hb_unicode_funcs_t *ufuncs,
+					   hb_unicode_eastasian_width_func_t func,
+					   void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_eastasian_width:
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+HB_EXTERN HB_DEPRECATED unsigned int
+hb_unicode_eastasian_width (hb_unicode_funcs_t *ufuncs,
+			    hb_codepoint_t unicode);
+
+
+/**
+ * hb_unicode_decompose_compatibility_func_t:
+ * @ufuncs: a Unicode function structure
+ * @u: codepoint to decompose
+ * @decomposed: address of codepoint array (of length %HB_UNICODE_MAX_DECOMPOSITION_LEN) to write decomposition into
+ * @user_data: user data pointer as passed to hb_unicode_funcs_set_decompose_compatibility_func()
+ *
+ * Fully decompose @u to its Unicode compatibility decomposition. The codepoints of the decomposition will be written to @decomposed.
+ * The complete length of the decomposition will be returned.
+ *
+ * If @u has no compatibility decomposition, zero should be returned.
+ *
+ * The Unicode standard guarantees that a buffer of length %HB_UNICODE_MAX_DECOMPOSITION_LEN codepoints will always be sufficient for any
+ * compatibility decomposition plus an terminating value of 0.  Consequently, @decompose must be allocated by the caller to be at least this length.  Implementations
+ * of this function type must ensure that they do not write past the provided array.
+ *
+ * Return value: number of codepoints in the full compatibility decomposition of @u, or 0 if no decomposition available.
+ *
+ * Deprecated: 2.0.0
+ */
+typedef unsigned int			(*hb_unicode_decompose_compatibility_func_t)	(hb_unicode_funcs_t *ufuncs,
+											 hb_codepoint_t      u,
+											 hb_codepoint_t     *decomposed,
+											 void               *user_data);
+
+/**
+ * HB_UNICODE_MAX_DECOMPOSITION_LEN:
+ *
+ * See Unicode 6.1 for details on the maximum decomposition length.
+ *
+ * Deprecated: 2.0.0
+ */
+#define HB_UNICODE_MAX_DECOMPOSITION_LEN (18+1) /* codepoints */
+
+/**
+ * hb_unicode_funcs_set_decompose_compatibility_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ * 
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+HB_EXTERN HB_DEPRECATED void
+hb_unicode_funcs_set_decompose_compatibility_func (hb_unicode_funcs_t *ufuncs,
+						   hb_unicode_decompose_compatibility_func_t func,
+						   void *user_data, hb_destroy_func_t destroy);
+
+HB_EXTERN HB_DEPRECATED unsigned int
+hb_unicode_decompose_compatibility (hb_unicode_funcs_t *ufuncs,
+				    hb_codepoint_t      u,
+				    hb_codepoint_t     *decomposed);
+
+
+typedef hb_font_get_glyph_kerning_func_t hb_font_get_glyph_v_kerning_func_t;
+
+/**
+ * hb_font_funcs_set_glyph_v_kerning_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ * 
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_v_kerning_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_v_kerning_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+HB_EXTERN hb_position_t
+hb_font_get_glyph_v_kerning (hb_font_t *font,
+			     hb_codepoint_t top_glyph, hb_codepoint_t bottom_glyph);
+
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_DEPRECATED_H */
diff --git a/thirdparty/harfbuzz/src/hb-directwrite.cc b/thirdparty/harfbuzz/src/hb-directwrite.cc
new file mode 100644
index 0000000000..f2fce073e0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-directwrite.cc
@@ -0,0 +1,979 @@
+/*
+ * Copyright © 2015-2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_DIRECTWRITE
+
+#include "hb-shaper-impl.hh"
+
+#include <dwrite_1.h>
+
+#include "hb-directwrite.h"
+
+
+/* Declare object creator for dynamic support of DWRITE */
+typedef HRESULT (* WINAPI t_DWriteCreateFactory)(
+  DWRITE_FACTORY_TYPE factoryType,
+  REFIID              iid,
+  IUnknown            **factory
+);
+
+/*
+ * hb-directwrite uses new/delete syntatically but as we let users
+ * to override malloc/free, we will redefine new/delete so users
+ * won't need to do that by their own.
+ */
+void* operator new (size_t size)        { return malloc (size); }
+void* operator new [] (size_t size)     { return malloc (size); }
+void operator delete (void* pointer)    { free (pointer); }
+void operator delete [] (void* pointer) { free (pointer); }
+
+
+/*
+ * DirectWrite font stream helpers
+ */
+
+// This is a font loader which provides only one font (unlike its original design).
+// For a better implementation which was also source of this
+// and DWriteFontFileStream, have a look at to NativeFontResourceDWrite.cpp in Mozilla
+class DWriteFontFileLoader : public IDWriteFontFileLoader
+{
+private:
+  IDWriteFontFileStream *mFontFileStream;
+public:
+  DWriteFontFileLoader (IDWriteFontFileStream *fontFileStream)
+  { mFontFileStream = fontFileStream; }
+
+  // IUnknown interface
+  IFACEMETHOD (QueryInterface) (IID const& iid, OUT void** ppObject)
+  { return S_OK; }
+  IFACEMETHOD_ (ULONG, AddRef) ()  { return 1; }
+  IFACEMETHOD_ (ULONG, Release) () { return 1; }
+
+  // IDWriteFontFileLoader methods
+  virtual HRESULT STDMETHODCALLTYPE
+  CreateStreamFromKey (void const* fontFileReferenceKey,
+		       uint32_t fontFileReferenceKeySize,
+		       OUT IDWriteFontFileStream** fontFileStream)
+  {
+    *fontFileStream = mFontFileStream;
+    return S_OK;
+  }
+
+  virtual ~DWriteFontFileLoader() {}
+};
+
+class DWriteFontFileStream : public IDWriteFontFileStream
+{
+private:
+  uint8_t *mData;
+  uint32_t mSize;
+public:
+  DWriteFontFileStream (uint8_t *aData, uint32_t aSize)
+  {
+    mData = aData;
+    mSize = aSize;
+  }
+
+  // IUnknown interface
+  IFACEMETHOD (QueryInterface) (IID const& iid, OUT void** ppObject)
+  { return S_OK; }
+  IFACEMETHOD_ (ULONG, AddRef) ()  { return 1; }
+  IFACEMETHOD_ (ULONG, Release) () { return 1; }
+
+  // IDWriteFontFileStream methods
+  virtual HRESULT STDMETHODCALLTYPE
+  ReadFileFragment (void const** fragmentStart,
+		    UINT64 fileOffset,
+		    UINT64 fragmentSize,
+		    OUT void** fragmentContext)
+  {
+    // We are required to do bounds checking.
+    if (fileOffset + fragmentSize > mSize) return E_FAIL;
+
+    // truncate the 64 bit fileOffset to size_t sized index into mData
+    size_t index = static_cast<size_t> (fileOffset);
+
+    // We should be alive for the duration of this.
+    *fragmentStart = &mData[index];
+    *fragmentContext = nullptr;
+    return S_OK;
+  }
+
+  virtual void STDMETHODCALLTYPE
+  ReleaseFileFragment (void* fragmentContext) {}
+
+  virtual HRESULT STDMETHODCALLTYPE
+  GetFileSize (OUT UINT64* fileSize)
+  {
+    *fileSize = mSize;
+    return S_OK;
+  }
+
+  virtual HRESULT STDMETHODCALLTYPE
+  GetLastWriteTime (OUT UINT64* lastWriteTime) { return E_NOTIMPL; }
+
+  virtual ~DWriteFontFileStream() {}
+};
+
+
+/*
+* shaper face data
+*/
+
+struct hb_directwrite_face_data_t
+{
+  HMODULE dwrite_dll;
+  IDWriteFactory *dwriteFactory;
+  IDWriteFontFile *fontFile;
+  DWriteFontFileStream *fontFileStream;
+  DWriteFontFileLoader *fontFileLoader;
+  IDWriteFontFace *fontFace;
+  hb_blob_t *faceBlob;
+};
+
+hb_directwrite_face_data_t *
+_hb_directwrite_shaper_face_data_create (hb_face_t *face)
+{
+  hb_directwrite_face_data_t *data = new hb_directwrite_face_data_t;
+  if (unlikely (!data))
+    return nullptr;
+
+#define FAIL(...) \
+  HB_STMT_START { \
+    DEBUG_MSG (DIRECTWRITE, nullptr, __VA_ARGS__); \
+    return nullptr; \
+  } HB_STMT_END
+
+  data->dwrite_dll = LoadLibrary (TEXT ("DWRITE"));
+  if (unlikely (!data->dwrite_dll))
+    FAIL ("Cannot find DWrite.DLL");
+
+  t_DWriteCreateFactory p_DWriteCreateFactory;
+
+#if defined(__GNUC__)
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-function-type"
+#endif
+
+  p_DWriteCreateFactory = (t_DWriteCreateFactory)
+			  GetProcAddress (data->dwrite_dll, "DWriteCreateFactory");
+
+#if defined(__GNUC__)
+#pragma GCC diagnostic pop
+#endif
+
+  if (unlikely (!p_DWriteCreateFactory))
+    FAIL ("Cannot find DWriteCreateFactory().");
+
+  HRESULT hr;
+
+  // TODO: factory and fontFileLoader should be cached separately
+  IDWriteFactory* dwriteFactory;
+  hr = p_DWriteCreateFactory (DWRITE_FACTORY_TYPE_SHARED, __uuidof (IDWriteFactory),
+			      (IUnknown**) &dwriteFactory);
+
+  if (unlikely (hr != S_OK))
+    FAIL ("Failed to run DWriteCreateFactory().");
+
+  hb_blob_t *blob = hb_face_reference_blob (face);
+  DWriteFontFileStream *fontFileStream;
+  fontFileStream = new DWriteFontFileStream ((uint8_t *) hb_blob_get_data (blob, nullptr),
+					     hb_blob_get_length (blob));
+
+  DWriteFontFileLoader *fontFileLoader = new DWriteFontFileLoader (fontFileStream);
+  dwriteFactory->RegisterFontFileLoader (fontFileLoader);
+
+  IDWriteFontFile *fontFile;
+  uint64_t fontFileKey = 0;
+  hr = dwriteFactory->CreateCustomFontFileReference (&fontFileKey, sizeof (fontFileKey),
+						     fontFileLoader, &fontFile);
+
+  if (FAILED (hr))
+    FAIL ("Failed to load font file from data!");
+
+  BOOL isSupported;
+  DWRITE_FONT_FILE_TYPE fileType;
+  DWRITE_FONT_FACE_TYPE faceType;
+  uint32_t numberOfFaces;
+  hr = fontFile->Analyze (&isSupported, &fileType, &faceType, &numberOfFaces);
+  if (FAILED (hr) || !isSupported)
+    FAIL ("Font file is not supported.");
+
+#undef FAIL
+
+  IDWriteFontFace *fontFace;
+  dwriteFactory->CreateFontFace (faceType, 1, &fontFile, 0,
+				 DWRITE_FONT_SIMULATIONS_NONE, &fontFace);
+
+  data->dwriteFactory = dwriteFactory;
+  data->fontFile = fontFile;
+  data->fontFileStream = fontFileStream;
+  data->fontFileLoader = fontFileLoader;
+  data->fontFace = fontFace;
+  data->faceBlob = blob;
+
+  return data;
+}
+
+void
+_hb_directwrite_shaper_face_data_destroy (hb_directwrite_face_data_t *data)
+{
+  if (data->fontFace)
+    data->fontFace->Release ();
+  if (data->fontFile)
+    data->fontFile->Release ();
+  if (data->dwriteFactory)
+  {
+    if (data->fontFileLoader)
+      data->dwriteFactory->UnregisterFontFileLoader (data->fontFileLoader);
+    data->dwriteFactory->Release ();
+  }
+  if (data->fontFileLoader)
+    delete data->fontFileLoader;
+  if (data->fontFileStream)
+    delete data->fontFileStream;
+  if (data->faceBlob)
+    hb_blob_destroy (data->faceBlob);
+  if (data->dwrite_dll)
+    FreeLibrary (data->dwrite_dll);
+  if (data)
+    delete data;
+}
+
+
+/*
+ * shaper font data
+ */
+
+struct hb_directwrite_font_data_t {};
+
+hb_directwrite_font_data_t *
+_hb_directwrite_shaper_font_data_create (hb_font_t *font)
+{
+  hb_directwrite_font_data_t *data = new hb_directwrite_font_data_t;
+  if (unlikely (!data))
+    return nullptr;
+
+  return data;
+}
+
+void
+_hb_directwrite_shaper_font_data_destroy (hb_directwrite_font_data_t *data)
+{
+  delete data;
+}
+
+
+// Most of TextAnalysis is originally written by Bas Schouten for Mozilla project
+// but now is relicensed to MIT for HarfBuzz use
+class TextAnalysis : public IDWriteTextAnalysisSource, public IDWriteTextAnalysisSink
+{
+public:
+
+  IFACEMETHOD (QueryInterface) (IID const& iid, OUT void** ppObject)
+  { return S_OK; }
+  IFACEMETHOD_ (ULONG, AddRef) () { return 1; }
+  IFACEMETHOD_ (ULONG, Release) () { return 1; }
+
+  // A single contiguous run of characters containing the same analysis
+  // results.
+  struct Run
+  {
+    uint32_t mTextStart;   // starting text position of this run
+    uint32_t mTextLength;  // number of contiguous code units covered
+    uint32_t mGlyphStart;  // starting glyph in the glyphs array
+    uint32_t mGlyphCount;  // number of glyphs associated with this run
+    // text
+    DWRITE_SCRIPT_ANALYSIS mScript;
+    uint8_t mBidiLevel;
+    bool mIsSideways;
+
+    bool ContainsTextPosition (uint32_t aTextPosition) const
+    {
+      return aTextPosition >= mTextStart &&
+	     aTextPosition <  mTextStart + mTextLength;
+    }
+
+    Run *nextRun;
+  };
+
+public:
+  TextAnalysis (const wchar_t* text, uint32_t textLength,
+		const wchar_t* localeName, DWRITE_READING_DIRECTION readingDirection)
+	       : mTextLength (textLength), mText (text), mLocaleName (localeName),
+		 mReadingDirection (readingDirection), mCurrentRun (nullptr) {}
+  ~TextAnalysis ()
+  {
+    // delete runs, except mRunHead which is part of the TextAnalysis object
+    for (Run *run = mRunHead.nextRun; run;)
+    {
+      Run *origRun = run;
+      run = run->nextRun;
+      delete origRun;
+    }
+  }
+
+  STDMETHODIMP
+  GenerateResults (IDWriteTextAnalyzer* textAnalyzer, Run **runHead)
+  {
+    // Analyzes the text using the script analyzer and returns
+    // the result as a series of runs.
+
+    HRESULT hr = S_OK;
+
+    // Initially start out with one result that covers the entire range.
+    // This result will be subdivided by the analysis processes.
+    mRunHead.mTextStart = 0;
+    mRunHead.mTextLength = mTextLength;
+    mRunHead.mBidiLevel =
+      (mReadingDirection == DWRITE_READING_DIRECTION_RIGHT_TO_LEFT);
+    mRunHead.nextRun = nullptr;
+    mCurrentRun = &mRunHead;
+
+    // Call each of the analyzers in sequence, recording their results.
+    if (SUCCEEDED (hr = textAnalyzer->AnalyzeScript (this, 0, mTextLength, this)))
+      *runHead = &mRunHead;
+
+    return hr;
+  }
+
+  // IDWriteTextAnalysisSource implementation
+
+  IFACEMETHODIMP
+  GetTextAtPosition (uint32_t textPosition,
+		     OUT wchar_t const** textString,
+		     OUT uint32_t* textLength)
+  {
+    if (textPosition >= mTextLength)
+    {
+      // No text at this position, valid query though.
+      *textString = nullptr;
+      *textLength = 0;
+    }
+    else
+    {
+      *textString = mText + textPosition;
+      *textLength = mTextLength - textPosition;
+    }
+    return S_OK;
+  }
+
+  IFACEMETHODIMP
+  GetTextBeforePosition (uint32_t textPosition,
+			 OUT wchar_t const** textString,
+			 OUT uint32_t* textLength)
+  {
+    if (textPosition == 0 || textPosition > mTextLength)
+    {
+      // Either there is no text before here (== 0), or this
+      // is an invalid position. The query is considered valid though.
+      *textString = nullptr;
+      *textLength = 0;
+    }
+    else
+    {
+      *textString = mText;
+      *textLength = textPosition;
+    }
+    return S_OK;
+  }
+
+  IFACEMETHODIMP_ (DWRITE_READING_DIRECTION)
+  GetParagraphReadingDirection () { return mReadingDirection; }
+
+  IFACEMETHODIMP GetLocaleName (uint32_t textPosition, uint32_t* textLength,
+				wchar_t const** localeName)
+  { return S_OK; }
+
+  IFACEMETHODIMP
+  GetNumberSubstitution (uint32_t textPosition,
+			 OUT uint32_t* textLength,
+			 OUT IDWriteNumberSubstitution** numberSubstitution)
+  {
+    // We do not support number substitution.
+    *numberSubstitution = nullptr;
+    *textLength = mTextLength - textPosition;
+
+    return S_OK;
+  }
+
+  // IDWriteTextAnalysisSink implementation
+
+  IFACEMETHODIMP
+  SetScriptAnalysis (uint32_t textPosition, uint32_t textLength,
+		     DWRITE_SCRIPT_ANALYSIS const* scriptAnalysis)
+  {
+    SetCurrentRun (textPosition);
+    SplitCurrentRun (textPosition);
+    while (textLength > 0)
+    {
+      Run *run = FetchNextRun (&textLength);
+      run->mScript = *scriptAnalysis;
+    }
+
+    return S_OK;
+  }
+
+  IFACEMETHODIMP
+  SetLineBreakpoints (uint32_t textPosition,
+		      uint32_t textLength,
+		      const DWRITE_LINE_BREAKPOINT* lineBreakpoints)
+  { return S_OK; }
+
+  IFACEMETHODIMP SetBidiLevel (uint32_t textPosition, uint32_t textLength,
+			       uint8_t explicitLevel, uint8_t resolvedLevel)
+  { return S_OK; }
+
+  IFACEMETHODIMP
+  SetNumberSubstitution (uint32_t textPosition, uint32_t textLength,
+			 IDWriteNumberSubstitution* numberSubstitution)
+  { return S_OK; }
+
+protected:
+  Run *FetchNextRun (IN OUT uint32_t* textLength)
+  {
+    // Used by the sink setters, this returns a reference to the next run.
+    // Position and length are adjusted to now point after the current run
+    // being returned.
+
+    Run *origRun = mCurrentRun;
+    // Split the tail if needed (the length remaining is less than the
+    // current run's size).
+    if (*textLength < mCurrentRun->mTextLength)
+      SplitCurrentRun (mCurrentRun->mTextStart + *textLength);
+    else
+      // Just advance the current run.
+      mCurrentRun = mCurrentRun->nextRun;
+    *textLength -= origRun->mTextLength;
+
+    // Return a reference to the run that was just current.
+    return origRun;
+  }
+
+  void SetCurrentRun (uint32_t textPosition)
+  {
+    // Move the current run to the given position.
+    // Since the analyzers generally return results in a forward manner,
+    // this will usually just return early. If not, find the
+    // corresponding run for the text position.
+
+    if (mCurrentRun && mCurrentRun->ContainsTextPosition (textPosition))
+      return;
+
+    for (Run *run = &mRunHead; run; run = run->nextRun)
+      if (run->ContainsTextPosition (textPosition))
+      {
+	mCurrentRun = run;
+	return;
+      }
+    assert (0); // We should always be able to find the text position in one of our runs
+  }
+
+  void SplitCurrentRun (uint32_t splitPosition)
+  {
+    if (!mCurrentRun)
+    {
+      assert (0); // SplitCurrentRun called without current run
+      // Shouldn't be calling this when no current run is set!
+      return;
+    }
+    // Split the current run.
+    if (splitPosition <= mCurrentRun->mTextStart)
+    {
+      // No need to split, already the start of a run
+      // or before it. Usually the first.
+      return;
+    }
+    Run *newRun = new Run;
+
+    *newRun = *mCurrentRun;
+
+    // Insert the new run in our linked list.
+    newRun->nextRun = mCurrentRun->nextRun;
+    mCurrentRun->nextRun = newRun;
+
+    // Adjust runs' text positions and lengths.
+    uint32_t splitPoint = splitPosition - mCurrentRun->mTextStart;
+    newRun->mTextStart += splitPoint;
+    newRun->mTextLength -= splitPoint;
+    mCurrentRun->mTextLength = splitPoint;
+    mCurrentRun = newRun;
+  }
+
+protected:
+  // Input
+  // (weak references are fine here, since this class is a transient
+  //  stack-based helper that doesn't need to copy data)
+  uint32_t mTextLength;
+  const wchar_t* mText;
+  const wchar_t* mLocaleName;
+  DWRITE_READING_DIRECTION mReadingDirection;
+
+  // Current processing state.
+  Run *mCurrentRun;
+
+  // Output is a list of runs starting here
+  Run  mRunHead;
+};
+
+/*
+ * shaper
+ */
+
+static hb_bool_t
+_hb_directwrite_shape_full (hb_shape_plan_t    *shape_plan,
+			    hb_font_t          *font,
+			    hb_buffer_t        *buffer,
+			    const hb_feature_t *features,
+			    unsigned int        num_features,
+			    float               lineWidth)
+{
+  hb_face_t *face = font->face;
+  const hb_directwrite_face_data_t *face_data = face->data.directwrite;
+  IDWriteFactory *dwriteFactory = face_data->dwriteFactory;
+  IDWriteFontFace *fontFace = face_data->fontFace;
+
+  IDWriteTextAnalyzer* analyzer;
+  dwriteFactory->CreateTextAnalyzer (&analyzer);
+
+  unsigned int scratch_size;
+  hb_buffer_t::scratch_buffer_t *scratch = buffer->get_scratch_buffer (&scratch_size);
+#define ALLOCATE_ARRAY(Type, name, len) \
+  Type *name = (Type *) scratch; \
+  do { \
+    unsigned int _consumed = DIV_CEIL ((len) * sizeof (Type), sizeof (*scratch)); \
+    assert (_consumed <= scratch_size); \
+    scratch += _consumed; \
+    scratch_size -= _consumed; \
+  } while (0)
+
+#define utf16_index() var1.u32
+
+  ALLOCATE_ARRAY (wchar_t, textString, buffer->len * 2);
+
+  unsigned int chars_len = 0;
+  for (unsigned int i = 0; i < buffer->len; i++)
+  {
+    hb_codepoint_t c = buffer->info[i].codepoint;
+    buffer->info[i].utf16_index () = chars_len;
+    if (likely (c <= 0xFFFFu))
+      textString[chars_len++] = c;
+    else if (unlikely (c > 0x10FFFFu))
+      textString[chars_len++] = 0xFFFDu;
+    else
+    {
+      textString[chars_len++] = 0xD800u + ((c - 0x10000u) >> 10);
+      textString[chars_len++] = 0xDC00u + ((c - 0x10000u) & ((1u << 10) - 1));
+    }
+  }
+
+  ALLOCATE_ARRAY (WORD, log_clusters, chars_len);
+  /* Need log_clusters to assign features. */
+  chars_len = 0;
+  for (unsigned int i = 0; i < buffer->len; i++)
+  {
+    hb_codepoint_t c = buffer->info[i].codepoint;
+    unsigned int cluster = buffer->info[i].cluster;
+    log_clusters[chars_len++] = cluster;
+    if (hb_in_range (c, 0x10000u, 0x10FFFFu))
+      log_clusters[chars_len++] = cluster; /* Surrogates. */
+  }
+
+  // TODO: Handle TEST_DISABLE_OPTIONAL_LIGATURES
+
+  DWRITE_READING_DIRECTION readingDirection;
+  readingDirection = buffer->props.direction ?
+		     DWRITE_READING_DIRECTION_RIGHT_TO_LEFT :
+		     DWRITE_READING_DIRECTION_LEFT_TO_RIGHT;
+
+  /*
+  * There's an internal 16-bit limit on some things inside the analyzer,
+  * but we never attempt to shape a word longer than 64K characters
+  * in a single gfxShapedWord, so we cannot exceed that limit.
+  */
+  uint32_t textLength = buffer->len;
+
+  TextAnalysis analysis (textString, textLength, nullptr, readingDirection);
+  TextAnalysis::Run *runHead;
+  HRESULT hr;
+  hr = analysis.GenerateResults (analyzer, &runHead);
+
+#define FAIL(...) \
+  HB_STMT_START { \
+    DEBUG_MSG (DIRECTWRITE, nullptr, __VA_ARGS__); \
+    return false; \
+  } HB_STMT_END
+
+  if (FAILED (hr))
+    FAIL ("Analyzer failed to generate results.");
+
+  uint32_t maxGlyphCount = 3 * textLength / 2 + 16;
+  uint32_t glyphCount;
+  bool isRightToLeft = HB_DIRECTION_IS_BACKWARD (buffer->props.direction);
+
+  const wchar_t localeName[20] = {0};
+  if (buffer->props.language)
+    mbstowcs ((wchar_t*) localeName,
+	      hb_language_to_string (buffer->props.language), 20);
+
+  // TODO: it does work but doesn't care about ranges
+  DWRITE_TYPOGRAPHIC_FEATURES typographic_features;
+  typographic_features.featureCount = num_features;
+  if (num_features)
+  {
+    typographic_features.features = new DWRITE_FONT_FEATURE[num_features];
+    for (unsigned int i = 0; i < num_features; ++i)
+    {
+      typographic_features.features[i].nameTag = (DWRITE_FONT_FEATURE_TAG)
+						 hb_uint32_swap (features[i].tag);
+      typographic_features.features[i].parameter = features[i].value;
+    }
+  }
+  const DWRITE_TYPOGRAPHIC_FEATURES* dwFeatures;
+  dwFeatures = (const DWRITE_TYPOGRAPHIC_FEATURES*) &typographic_features;
+  const uint32_t featureRangeLengths[] = { textLength };
+  //
+
+  uint16_t* clusterMap;
+  clusterMap = new uint16_t[textLength];
+  DWRITE_SHAPING_TEXT_PROPERTIES* textProperties;
+  textProperties = new DWRITE_SHAPING_TEXT_PROPERTIES[textLength];
+retry_getglyphs:
+  uint16_t* glyphIndices = new uint16_t[maxGlyphCount];
+  DWRITE_SHAPING_GLYPH_PROPERTIES* glyphProperties;
+  glyphProperties = new DWRITE_SHAPING_GLYPH_PROPERTIES[maxGlyphCount];
+
+  hr = analyzer->GetGlyphs (textString, textLength, fontFace, false,
+			    isRightToLeft, &runHead->mScript, localeName,
+			    nullptr, &dwFeatures, featureRangeLengths, 1,
+			    maxGlyphCount, clusterMap, textProperties,
+			    glyphIndices, glyphProperties, &glyphCount);
+
+  if (unlikely (hr == HRESULT_FROM_WIN32 (ERROR_INSUFFICIENT_BUFFER)))
+  {
+    delete [] glyphIndices;
+    delete [] glyphProperties;
+
+    maxGlyphCount *= 2;
+
+    goto retry_getglyphs;
+  }
+  if (FAILED (hr))
+    FAIL ("Analyzer failed to get glyphs.");
+
+  float* glyphAdvances = new float[maxGlyphCount];
+  DWRITE_GLYPH_OFFSET* glyphOffsets = new DWRITE_GLYPH_OFFSET[maxGlyphCount];
+
+  /* The -2 in the following is to compensate for possible
+   * alignment needed after the WORD array.  sizeof (WORD) == 2. */
+  unsigned int glyphs_size = (scratch_size * sizeof (int) - 2)
+			     / (sizeof (WORD) +
+				sizeof (DWRITE_SHAPING_GLYPH_PROPERTIES) +
+				sizeof (int) +
+				sizeof (DWRITE_GLYPH_OFFSET) +
+				sizeof (uint32_t));
+  ALLOCATE_ARRAY (uint32_t, vis_clusters, glyphs_size);
+
+#undef ALLOCATE_ARRAY
+
+  int fontEmSize = font->face->get_upem ();
+  if (fontEmSize < 0) fontEmSize = -fontEmSize;
+
+  if (fontEmSize < 0) fontEmSize = -fontEmSize;
+  double x_mult = (double) font->x_scale / fontEmSize;
+  double y_mult = (double) font->y_scale / fontEmSize;
+
+  hr = analyzer->GetGlyphPlacements (textString, clusterMap, textProperties,
+				     textLength, glyphIndices, glyphProperties,
+				     glyphCount, fontFace, fontEmSize,
+				     false, isRightToLeft, &runHead->mScript, localeName,
+				     &dwFeatures, featureRangeLengths, 1,
+				     glyphAdvances, glyphOffsets);
+
+  if (FAILED (hr))
+    FAIL ("Analyzer failed to get glyph placements.");
+
+  IDWriteTextAnalyzer1* analyzer1;
+  analyzer->QueryInterface (&analyzer1);
+
+  if (analyzer1 && lineWidth)
+  {
+    DWRITE_JUSTIFICATION_OPPORTUNITY* justificationOpportunities =
+      new DWRITE_JUSTIFICATION_OPPORTUNITY[maxGlyphCount];
+    hr = analyzer1->GetJustificationOpportunities (fontFace, fontEmSize, runHead->mScript,
+						   textLength, glyphCount, textString,
+						   clusterMap, glyphProperties,
+						   justificationOpportunities);
+
+    if (FAILED (hr))
+      FAIL ("Analyzer failed to get justification opportunities.");
+
+    float* justifiedGlyphAdvances = new float[maxGlyphCount];
+    DWRITE_GLYPH_OFFSET* justifiedGlyphOffsets = new DWRITE_GLYPH_OFFSET[glyphCount];
+    hr = analyzer1->JustifyGlyphAdvances (lineWidth, glyphCount, justificationOpportunities,
+					  glyphAdvances, glyphOffsets, justifiedGlyphAdvances,
+					  justifiedGlyphOffsets);
+
+    if (FAILED (hr)) FAIL ("Analyzer failed to get justify glyph advances.");
+
+    DWRITE_SCRIPT_PROPERTIES scriptProperties;
+    hr = analyzer1->GetScriptProperties (runHead->mScript, &scriptProperties);
+    if (FAILED (hr)) FAIL ("Analyzer failed to get script properties.");
+    uint32_t justificationCharacter = scriptProperties.justificationCharacter;
+
+    // if a script justificationCharacter is not space, it can have GetJustifiedGlyphs
+    if (justificationCharacter != 32)
+    {
+      uint16_t* modifiedClusterMap = new uint16_t[textLength];
+    retry_getjustifiedglyphs:
+      uint16_t* modifiedGlyphIndices = new uint16_t[maxGlyphCount];
+      float* modifiedGlyphAdvances = new float[maxGlyphCount];
+      DWRITE_GLYPH_OFFSET* modifiedGlyphOffsets = new DWRITE_GLYPH_OFFSET[maxGlyphCount];
+      uint32_t actualGlyphsCount;
+      hr = analyzer1->GetJustifiedGlyphs (fontFace, fontEmSize, runHead->mScript,
+					  textLength, glyphCount, maxGlyphCount,
+					  clusterMap, glyphIndices, glyphAdvances,
+					  justifiedGlyphAdvances, justifiedGlyphOffsets,
+					  glyphProperties, &actualGlyphsCount,
+					  modifiedClusterMap, modifiedGlyphIndices,
+					  modifiedGlyphAdvances, modifiedGlyphOffsets);
+
+      if (hr == HRESULT_FROM_WIN32 (ERROR_INSUFFICIENT_BUFFER))
+      {
+	maxGlyphCount = actualGlyphsCount;
+	delete [] modifiedGlyphIndices;
+	delete [] modifiedGlyphAdvances;
+	delete [] modifiedGlyphOffsets;
+
+	maxGlyphCount = actualGlyphsCount;
+
+	goto retry_getjustifiedglyphs;
+      }
+      if (FAILED (hr))
+	FAIL ("Analyzer failed to get justified glyphs.");
+
+      delete [] clusterMap;
+      delete [] glyphIndices;
+      delete [] glyphAdvances;
+      delete [] glyphOffsets;
+
+      glyphCount = actualGlyphsCount;
+      clusterMap = modifiedClusterMap;
+      glyphIndices = modifiedGlyphIndices;
+      glyphAdvances = modifiedGlyphAdvances;
+      glyphOffsets = modifiedGlyphOffsets;
+
+      delete [] justifiedGlyphAdvances;
+      delete [] justifiedGlyphOffsets;
+    }
+    else
+    {
+      delete [] glyphAdvances;
+      delete [] glyphOffsets;
+
+      glyphAdvances = justifiedGlyphAdvances;
+      glyphOffsets = justifiedGlyphOffsets;
+    }
+
+    delete [] justificationOpportunities;
+  }
+
+  /* Ok, we've got everything we need, now compose output buffer,
+   * very, *very*, carefully! */
+
+  /* Calculate visual-clusters.  That's what we ship. */
+  for (unsigned int i = 0; i < glyphCount; i++)
+    vis_clusters[i] = (uint32_t) -1;
+  for (unsigned int i = 0; i < buffer->len; i++)
+  {
+    uint32_t *p =
+      &vis_clusters[log_clusters[buffer->info[i].utf16_index ()]];
+    *p = hb_min (*p, buffer->info[i].cluster);
+  }
+  for (unsigned int i = 1; i < glyphCount; i++)
+    if (vis_clusters[i] == (uint32_t) -1)
+      vis_clusters[i] = vis_clusters[i - 1];
+
+#undef utf16_index
+
+  if (unlikely (!buffer->ensure (glyphCount)))
+    FAIL ("Buffer in error");
+
+#undef FAIL
+
+  /* Set glyph infos */
+  buffer->len = 0;
+  for (unsigned int i = 0; i < glyphCount; i++)
+  {
+    hb_glyph_info_t *info = &buffer->info[buffer->len++];
+
+    info->codepoint = glyphIndices[i];
+    info->cluster = vis_clusters[i];
+
+    /* The rest is crap.  Let's store position info there for now. */
+    info->mask = glyphAdvances[i];
+    info->var1.i32 = glyphOffsets[i].advanceOffset;
+    info->var2.i32 = glyphOffsets[i].ascenderOffset;
+  }
+
+  /* Set glyph positions */
+  buffer->clear_positions ();
+  for (unsigned int i = 0; i < glyphCount; i++)
+  {
+    hb_glyph_info_t *info = &buffer->info[i];
+    hb_glyph_position_t *pos = &buffer->pos[i];
+
+    /* TODO vertical */
+    pos->x_advance = x_mult * (int32_t) info->mask;
+    pos->x_offset = x_mult * (isRightToLeft ? -info->var1.i32 : info->var1.i32);
+    pos->y_offset = y_mult * info->var2.i32;
+  }
+
+  if (isRightToLeft) hb_buffer_reverse (buffer);
+
+  delete [] clusterMap;
+  delete [] glyphIndices;
+  delete [] textProperties;
+  delete [] glyphProperties;
+  delete [] glyphAdvances;
+  delete [] glyphOffsets;
+
+  if (num_features)
+    delete [] typographic_features.features;
+
+  /* Wow, done! */
+  return true;
+}
+
+hb_bool_t
+_hb_directwrite_shape (hb_shape_plan_t    *shape_plan,
+		       hb_font_t          *font,
+		       hb_buffer_t        *buffer,
+		       const hb_feature_t *features,
+		       unsigned int        num_features)
+{
+  return _hb_directwrite_shape_full (shape_plan, font, buffer,
+				     features, num_features, 0);
+}
+
+HB_UNUSED static bool
+_hb_directwrite_shape_experimental_width (hb_font_t          *font,
+					  hb_buffer_t        *buffer,
+					  const hb_feature_t *features,
+					  unsigned int        num_features,
+					  float               width)
+{
+  static const char *shapers = "directwrite";
+  hb_shape_plan_t *shape_plan;
+  shape_plan = hb_shape_plan_create_cached (font->face, &buffer->props,
+					    features, num_features, &shapers);
+  hb_bool_t res = _hb_directwrite_shape_full (shape_plan, font, buffer,
+					      features, num_features, width);
+
+  buffer->unsafe_to_break_all ();
+
+  return res;
+}
+
+struct _hb_directwrite_font_table_context {
+  IDWriteFontFace *face;
+  void *table_context;
+};
+
+static void
+_hb_directwrite_table_data_release (void *data)
+{
+  _hb_directwrite_font_table_context *context = (_hb_directwrite_font_table_context *) data;
+  context->face->ReleaseFontTable (context->table_context);
+  delete context;
+}
+
+static hb_blob_t *
+_hb_directwrite_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  IDWriteFontFace *dw_face = ((IDWriteFontFace *) user_data);
+  const void *data;
+  uint32_t length;
+  void *table_context;
+  BOOL exists;
+  if (!dw_face || FAILED (dw_face->TryGetFontTable (hb_uint32_swap (tag), &data,
+						    &length, &table_context, &exists)))
+    return nullptr;
+
+  if (!data || !exists || !length)
+  {
+    dw_face->ReleaseFontTable (table_context);
+    return nullptr;
+  }
+
+  _hb_directwrite_font_table_context *context = new _hb_directwrite_font_table_context;
+  context->face = dw_face;
+  context->table_context = table_context;
+
+  return hb_blob_create ((const char *) data, length, HB_MEMORY_MODE_READONLY,
+			 context, _hb_directwrite_table_data_release);
+}
+
+static void
+_hb_directwrite_font_release (void *data)
+{
+  if (data)
+    ((IDWriteFontFace *) data)->Release ();
+}
+
+/**
+ * hb_directwrite_face_create:
+ * @font_face: a DirectWrite IDWriteFontFace object.
+ *
+ * Return value: #hb_face_t object corresponding to the given input
+ *
+ * Since: 2.4.0
+ **/
+hb_face_t *
+hb_directwrite_face_create (IDWriteFontFace *font_face)
+{
+  if (font_face)
+    font_face->AddRef ();
+  return hb_face_create_for_tables (_hb_directwrite_reference_table, font_face,
+				    _hb_directwrite_font_release);
+}
+
+/**
+* hb_directwrite_face_get_font_face:
+* @face: a #hb_face_t object
+*
+* Return value: DirectWrite IDWriteFontFace object corresponding to the given input
+*
+* Since: 2.5.0
+**/
+IDWriteFontFace *
+hb_directwrite_face_get_font_face (hb_face_t *face)
+{
+  return face->data.directwrite->fontFace;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-directwrite.h b/thirdparty/harfbuzz/src/hb-directwrite.h
new file mode 100644
index 0000000000..f837627a28
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-directwrite.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright © 2015-2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_DIRECTWRITE_H
+#define HB_DIRECTWRITE_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+HB_EXTERN hb_face_t *
+hb_directwrite_face_create (IDWriteFontFace *font_face);
+
+HB_EXTERN IDWriteFontFace *
+hb_directwrite_face_get_font_face (hb_face_t *face);
+
+HB_END_DECLS
+
+#endif /* HB_DIRECTWRITE_H */
diff --git a/thirdparty/harfbuzz/src/hb-dispatch.hh b/thirdparty/harfbuzz/src/hb-dispatch.hh
new file mode 100644
index 0000000000..7eace86e54
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-dispatch.hh
@@ -0,0 +1,61 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2012,2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_DISPATCH_HH
+#define HB_DISPATCH_HH
+
+#include "hb.hh"
+
+/*
+ * Dispatch
+ */
+
+template <typename Context, typename Return=hb_empty_t, unsigned int MaxDebugDepth=0>
+struct hb_dispatch_context_t
+{
+  hb_dispatch_context_t () : debug_depth (0) {}
+  private:
+  /* https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern */
+  const Context* thiz () const { return static_cast<const Context *> (this); }
+	Context* thiz ()       { return static_cast<      Context *> (this); }
+  public:
+  const char *get_name () { return "UNKNOWN"; }
+  static constexpr unsigned max_debug_depth = MaxDebugDepth;
+  typedef Return return_t;
+  template <typename T, typename F>
+  bool may_dispatch (const T *obj HB_UNUSED, const F *format HB_UNUSED) { return true; }
+  template <typename T, typename ...Ts>
+  return_t dispatch (const T &obj, Ts&&... ds)
+  { return obj.dispatch (thiz (), hb_forward<Ts> (ds)...); }
+  static return_t no_dispatch_return_value () { return Context::default_return_value (); }
+  static bool stop_sublookup_iteration (const return_t r HB_UNUSED) { return false; }
+  unsigned debug_depth;
+};
+
+
+#endif /* HB_DISPATCH_HH */
diff --git a/thirdparty/harfbuzz/src/hb-draw.cc b/thirdparty/harfbuzz/src/hb-draw.cc
new file mode 100644
index 0000000000..1a5f9c8c6b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-draw.cc
@@ -0,0 +1,261 @@
+/*
+ * Copyright © 2019-2020  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_DRAW
+#ifdef HB_EXPERIMENTAL_API
+
+#include "hb-draw.hh"
+#include "hb-ot.h"
+#include "hb-ot-glyf-table.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-ot-cff2-table.hh"
+
+/**
+ * hb_draw_funcs_set_move_to_func:
+ * @funcs: draw functions object
+ * @move_to: move-to callback
+ *
+ * Sets move-to callback to the draw functions object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_set_move_to_func (hb_draw_funcs_t        *funcs,
+				hb_draw_move_to_func_t  move_to)
+{
+  if (unlikely (hb_object_is_immutable (funcs))) return;
+  funcs->move_to = move_to;
+}
+
+/**
+ * hb_draw_funcs_set_line_to_func:
+ * @funcs: draw functions object
+ * @line_to: line-to callback
+ *
+ * Sets line-to callback to the draw functions object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_set_line_to_func (hb_draw_funcs_t        *funcs,
+				hb_draw_line_to_func_t  line_to)
+{
+  if (unlikely (hb_object_is_immutable (funcs))) return;
+  funcs->line_to = line_to;
+}
+
+/**
+ * hb_draw_funcs_set_quadratic_to_func:
+ * @funcs: draw functions object
+ * @move_to: quadratic-to callback
+ *
+ * Sets quadratic-to callback to the draw functions object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_set_quadratic_to_func (hb_draw_funcs_t             *funcs,
+				     hb_draw_quadratic_to_func_t  quadratic_to)
+{
+  if (unlikely (hb_object_is_immutable (funcs))) return;
+  funcs->quadratic_to = quadratic_to;
+  funcs->is_quadratic_to_set = true;
+}
+
+/**
+ * hb_draw_funcs_set_cubic_to_func:
+ * @funcs: draw functions
+ * @cubic_to: cubic-to callback
+ *
+ * Sets cubic-to callback to the draw functions object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_set_cubic_to_func (hb_draw_funcs_t         *funcs,
+				 hb_draw_cubic_to_func_t  cubic_to)
+{
+  if (unlikely (hb_object_is_immutable (funcs))) return;
+  funcs->cubic_to = cubic_to;
+}
+
+/**
+ * hb_draw_funcs_set_close_path_func:
+ * @funcs: draw functions object
+ * @close_path: close-path callback
+ *
+ * Sets close-path callback to the draw functions object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_set_close_path_func (hb_draw_funcs_t           *funcs,
+				   hb_draw_close_path_func_t  close_path)
+{
+  if (unlikely (hb_object_is_immutable (funcs))) return;
+  funcs->close_path = close_path;
+}
+
+static void
+_move_to_nil (hb_position_t to_x HB_UNUSED, hb_position_t to_y HB_UNUSED, void *user_data HB_UNUSED) {}
+
+static void
+_line_to_nil (hb_position_t to_x HB_UNUSED, hb_position_t to_y HB_UNUSED, void *user_data HB_UNUSED) {}
+
+static void
+_quadratic_to_nil (hb_position_t control_x HB_UNUSED, hb_position_t control_y HB_UNUSED,
+		   hb_position_t to_x HB_UNUSED, hb_position_t to_y HB_UNUSED,
+		   void *user_data HB_UNUSED) {}
+
+static void
+_cubic_to_nil (hb_position_t control1_x HB_UNUSED, hb_position_t control1_y HB_UNUSED,
+	       hb_position_t control2_x HB_UNUSED, hb_position_t control2_y HB_UNUSED,
+	       hb_position_t to_x HB_UNUSED, hb_position_t to_y HB_UNUSED,
+	       void *user_data HB_UNUSED) {}
+
+static void
+_close_path_nil (void *user_data HB_UNUSED) {}
+
+/**
+ * hb_draw_funcs_create:
+ *
+ * Creates a new draw callbacks object.
+ *
+ * Since: EXPERIMENTAL
+ **/
+hb_draw_funcs_t *
+hb_draw_funcs_create ()
+{
+  hb_draw_funcs_t *funcs;
+  if (unlikely (!(funcs = hb_object_create<hb_draw_funcs_t> ())))
+    return const_cast<hb_draw_funcs_t *> (&Null (hb_draw_funcs_t));
+
+  funcs->move_to = (hb_draw_move_to_func_t) _move_to_nil;
+  funcs->line_to = (hb_draw_line_to_func_t) _line_to_nil;
+  funcs->quadratic_to = (hb_draw_quadratic_to_func_t) _quadratic_to_nil;
+  funcs->is_quadratic_to_set = false;
+  funcs->cubic_to = (hb_draw_cubic_to_func_t) _cubic_to_nil;
+  funcs->close_path = (hb_draw_close_path_func_t) _close_path_nil;
+  return funcs;
+}
+
+/**
+ * hb_draw_funcs_reference:
+ * @funcs: draw functions
+ *
+ * Add to callbacks object refcount.
+ *
+ * Returns: The same object.
+ * Since: EXPERIMENTAL
+ **/
+hb_draw_funcs_t *
+hb_draw_funcs_reference (hb_draw_funcs_t *funcs)
+{
+  return hb_object_reference (funcs);
+}
+
+/**
+ * hb_draw_funcs_destroy:
+ * @funcs: draw functions
+ *
+ * Decreases refcount of callbacks object and deletes the object if it reaches
+ * to zero.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_destroy (hb_draw_funcs_t *funcs)
+{
+  if (!hb_object_destroy (funcs)) return;
+
+  free (funcs);
+}
+
+/**
+ * hb_draw_funcs_make_immutable:
+ * @funcs: draw functions
+ *
+ * Makes funcs object immutable.
+ *
+ * Since: EXPERIMENTAL
+ **/
+void
+hb_draw_funcs_make_immutable (hb_draw_funcs_t *funcs)
+{
+  if (hb_object_is_immutable (funcs))
+    return;
+
+  hb_object_make_immutable (funcs);
+}
+
+/**
+ * hb_draw_funcs_is_immutable:
+ * @funcs: draw functions
+ *
+ * Checks whether funcs is immutable.
+ *
+ * Returns: If is immutable.
+ * Since: EXPERIMENTAL
+ **/
+hb_bool_t
+hb_draw_funcs_is_immutable (hb_draw_funcs_t *funcs)
+{
+  return hb_object_is_immutable (funcs);
+}
+
+/**
+ * hb_font_draw_glyph:
+ * @font: a font object
+ * @glyph: a glyph id
+ * @funcs: draw callbacks object
+ * @user_data: parameter you like be passed to the callbacks when are called
+ *
+ * Draw a glyph.
+ *
+ * Returns: Whether the font had the glyph and the operation completed successfully.
+ * Since: EXPERIMENTAL
+ **/
+hb_bool_t
+hb_font_draw_glyph (hb_font_t *font, hb_codepoint_t glyph,
+		    const hb_draw_funcs_t *funcs,
+		    void *user_data)
+{
+  if (unlikely (funcs == &Null (hb_draw_funcs_t) ||
+		glyph >= font->face->get_num_glyphs ()))
+    return false;
+
+  draw_helper_t draw_helper (funcs, user_data);
+  if (font->face->table.glyf->get_path (font, glyph, draw_helper)) return true;
+#ifndef HB_NO_CFF
+  if (font->face->table.cff1->get_path (font, glyph, draw_helper)) return true;
+  if (font->face->table.cff2->get_path (font, glyph, draw_helper)) return true;
+#endif
+
+  return false;
+}
+
+#endif
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-draw.h b/thirdparty/harfbuzz/src/hb-draw.h
new file mode 100644
index 0000000000..98eccf4c0c
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-draw.h
@@ -0,0 +1,98 @@
+/*
+ * Copyright © 2019-2020  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_DRAW_H
+#define HB_DRAW_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+#ifdef HB_EXPERIMENTAL_API
+typedef void (*hb_draw_move_to_func_t) (hb_position_t to_x, hb_position_t to_y, void *user_data);
+typedef void (*hb_draw_line_to_func_t) (hb_position_t to_x, hb_position_t to_y, void *user_data);
+typedef void (*hb_draw_quadratic_to_func_t) (hb_position_t control_x, hb_position_t control_y,
+					     hb_position_t to_x, hb_position_t to_y,
+					     void *user_data);
+typedef void (*hb_draw_cubic_to_func_t) (hb_position_t control1_x, hb_position_t control1_y,
+					 hb_position_t control2_x, hb_position_t control2_y,
+					 hb_position_t to_x, hb_position_t to_y,
+					 void *user_data);
+typedef void (*hb_draw_close_path_func_t) (void *user_data);
+
+/**
+ * hb_draw_funcs_t:
+ *
+ * Glyph draw callbacks.
+ *
+ * _move_to, _line_to and _cubic_to calls are nessecary to be defined but we
+ * translate _quadratic_to calls to _cubic_to if the callback isn't defined.
+ *
+ * Since: EXPERIMENTAL
+ **/
+typedef struct hb_draw_funcs_t hb_draw_funcs_t;
+
+HB_EXTERN void
+hb_draw_funcs_set_move_to_func (hb_draw_funcs_t        *funcs,
+				hb_draw_move_to_func_t  move_to);
+
+HB_EXTERN void
+hb_draw_funcs_set_line_to_func (hb_draw_funcs_t        *funcs,
+				hb_draw_line_to_func_t  line_to);
+
+HB_EXTERN void
+hb_draw_funcs_set_quadratic_to_func (hb_draw_funcs_t             *funcs,
+				     hb_draw_quadratic_to_func_t  quadratic_to);
+
+HB_EXTERN void
+hb_draw_funcs_set_cubic_to_func (hb_draw_funcs_t         *funcs,
+				 hb_draw_cubic_to_func_t  cubic_to);
+
+HB_EXTERN void
+hb_draw_funcs_set_close_path_func (hb_draw_funcs_t           *funcs,
+				   hb_draw_close_path_func_t  close_path);
+
+HB_EXTERN hb_draw_funcs_t *
+hb_draw_funcs_create (void);
+
+HB_EXTERN hb_draw_funcs_t *
+hb_draw_funcs_reference (hb_draw_funcs_t *funcs);
+
+HB_EXTERN void
+hb_draw_funcs_destroy (hb_draw_funcs_t *funcs);
+
+HB_EXTERN void
+hb_draw_funcs_make_immutable (hb_draw_funcs_t *funcs);
+
+HB_EXTERN hb_bool_t
+hb_draw_funcs_is_immutable (hb_draw_funcs_t *funcs);
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_DRAW_H */
diff --git a/thirdparty/harfbuzz/src/hb-draw.hh b/thirdparty/harfbuzz/src/hb-draw.hh
new file mode 100644
index 0000000000..2aa0a5b4db
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-draw.hh
@@ -0,0 +1,139 @@
+/*
+ * Copyright © 2020  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_DRAW_HH
+#define HB_DRAW_HH
+
+#include "hb.hh"
+
+#ifdef HB_EXPERIMENTAL_API
+struct hb_draw_funcs_t
+{
+  hb_object_header_t header;
+
+  hb_draw_move_to_func_t move_to;
+  hb_draw_line_to_func_t line_to;
+  hb_draw_quadratic_to_func_t quadratic_to;
+  bool is_quadratic_to_set;
+  hb_draw_cubic_to_func_t cubic_to;
+  hb_draw_close_path_func_t close_path;
+};
+
+struct draw_helper_t
+{
+  draw_helper_t (const hb_draw_funcs_t *funcs_, void *user_data_)
+  {
+    funcs = funcs_;
+    user_data = user_data_;
+    path_open = false;
+    path_start_x = current_x = path_start_y = current_y = 0;
+  }
+  ~draw_helper_t () { end_path (); }
+
+  void move_to (hb_position_t x, hb_position_t y)
+  {
+    if (path_open) end_path ();
+    current_x = path_start_x = x;
+    current_y = path_start_y = y;
+  }
+
+  void line_to (hb_position_t x, hb_position_t y)
+  {
+    if (equal_to_current (x, y)) return;
+    if (!path_open) start_path ();
+    funcs->line_to (x, y, user_data);
+    current_x = x;
+    current_y = y;
+  }
+
+  void
+  quadratic_to (hb_position_t control_x, hb_position_t control_y,
+		hb_position_t to_x, hb_position_t to_y)
+  {
+    if (equal_to_current (control_x, control_y) && equal_to_current (to_x, to_y))
+      return;
+    if (!path_open) start_path ();
+    if (funcs->is_quadratic_to_set)
+      funcs->quadratic_to (control_x, control_y, to_x, to_y, user_data);
+    else
+      funcs->cubic_to (roundf ((current_x + 2.f * control_x) / 3.f),
+		       roundf ((current_y + 2.f * control_y) / 3.f),
+		       roundf ((to_x + 2.f * control_x) / 3.f),
+		       roundf ((to_y + 2.f * control_y) / 3.f),
+		       to_x, to_y, user_data);
+    current_x = to_x;
+    current_y = to_y;
+  }
+
+  void
+  cubic_to (hb_position_t control1_x, hb_position_t control1_y,
+	    hb_position_t control2_x, hb_position_t control2_y,
+	    hb_position_t to_x, hb_position_t to_y)
+  {
+    if (equal_to_current (control1_x, control1_y) &&
+	equal_to_current (control2_x, control2_y) &&
+	equal_to_current (to_x, to_y))
+      return;
+    if (!path_open) start_path ();
+    funcs->cubic_to (control1_x, control1_y, control2_x, control2_y, to_x, to_y, user_data);
+    current_x = to_x;
+    current_y = to_y;
+  }
+
+  void end_path ()
+  {
+    if (path_open)
+    {
+      if ((path_start_x != current_x) || (path_start_y != current_y))
+	funcs->line_to (path_start_x, path_start_y, user_data);
+      funcs->close_path (user_data);
+    }
+    path_open = false;
+    path_start_x = current_x = path_start_y = current_y = 0;
+  }
+
+  protected:
+  bool equal_to_current (hb_position_t x, hb_position_t y)
+  { return current_x == x && current_y == y; }
+
+  void start_path ()
+  {
+    if (path_open) end_path ();
+    path_open = true;
+    funcs->move_to (path_start_x, path_start_y, user_data);
+  }
+
+  hb_position_t path_start_x;
+  hb_position_t path_start_y;
+
+  hb_position_t current_x;
+  hb_position_t current_y;
+
+  bool path_open;
+  const hb_draw_funcs_t *funcs;
+  void *user_data;
+};
+#endif
+
+#endif /* HB_DRAW_HH */
diff --git a/thirdparty/harfbuzz/src/hb-face.cc b/thirdparty/harfbuzz/src/hb-face.cc
new file mode 100644
index 0000000000..7bde50df5b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-face.cc
@@ -0,0 +1,733 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-face.hh"
+#include "hb-blob.hh"
+#include "hb-open-file.hh"
+#include "hb-ot-face.hh"
+#include "hb-ot-cmap-table.hh"
+
+
+/**
+ * SECTION:hb-face
+ * @title: hb-face
+ * @short_description: Font face objects
+ * @include: hb.h
+ *
+ * Font face is objects represent a single face in a font family.
+ * More exactly, a font face represents a single face in a binary font file.
+ * Font faces are typically built from a binary blob and a face index.
+ * Font faces are used to create fonts.
+ **/
+
+
+/**
+ * hb_face_count:
+ * @blob: a blob.
+ *
+ * Get number of faces in a blob.
+ *
+ * Return value: Number of faces in @blob
+ *
+ * Since: 1.7.7
+ **/
+unsigned int
+hb_face_count (hb_blob_t *blob)
+{
+  if (unlikely (!blob))
+    return 0;
+
+  /* TODO We shouldn't be sanitizing blob.  Port to run sanitizer and return if not sane. */
+  /* Make API signature const after. */
+  hb_blob_t *sanitized = hb_sanitize_context_t ().sanitize_blob<OT::OpenTypeFontFile> (hb_blob_reference (blob));
+  const OT::OpenTypeFontFile& ot = *sanitized->as<OT::OpenTypeFontFile> ();
+  unsigned int ret = ot.get_face_count ();
+  hb_blob_destroy (sanitized);
+
+  return ret;
+}
+
+/*
+ * hb_face_t
+ */
+
+DEFINE_NULL_INSTANCE (hb_face_t) =
+{
+  HB_OBJECT_HEADER_STATIC,
+
+  nullptr, /* reference_table_func */
+  nullptr, /* user_data */
+  nullptr, /* destroy */
+
+  0,    /* index */
+  HB_ATOMIC_INT_INIT (1000), /* upem */
+  HB_ATOMIC_INT_INIT (0),    /* num_glyphs */
+
+  /* Zero for the rest is fine. */
+};
+
+
+/**
+ * hb_face_create_for_tables:
+ * @reference_table_func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Return value: (transfer full)
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_face_create_for_tables (hb_reference_table_func_t  reference_table_func,
+			   void                      *user_data,
+			   hb_destroy_func_t          destroy)
+{
+  hb_face_t *face;
+
+  if (!reference_table_func || !(face = hb_object_create<hb_face_t> ())) {
+    if (destroy)
+      destroy (user_data);
+    return hb_face_get_empty ();
+  }
+
+  face->reference_table_func = reference_table_func;
+  face->user_data = user_data;
+  face->destroy = destroy;
+
+  face->num_glyphs.set_relaxed (-1);
+
+  face->data.init0 (face);
+  face->table.init0 (face);
+
+  return face;
+}
+
+
+typedef struct hb_face_for_data_closure_t {
+  hb_blob_t *blob;
+  unsigned int  index;
+} hb_face_for_data_closure_t;
+
+static hb_face_for_data_closure_t *
+_hb_face_for_data_closure_create (hb_blob_t *blob, unsigned int index)
+{
+  hb_face_for_data_closure_t *closure;
+
+  closure = (hb_face_for_data_closure_t *) calloc (1, sizeof (hb_face_for_data_closure_t));
+  if (unlikely (!closure))
+    return nullptr;
+
+  closure->blob = blob;
+  closure->index = index;
+
+  return closure;
+}
+
+static void
+_hb_face_for_data_closure_destroy (void *data)
+{
+  hb_face_for_data_closure_t *closure = (hb_face_for_data_closure_t *) data;
+
+  hb_blob_destroy (closure->blob);
+  free (closure);
+}
+
+static hb_blob_t *
+_hb_face_for_data_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  hb_face_for_data_closure_t *data = (hb_face_for_data_closure_t *) user_data;
+
+  if (tag == HB_TAG_NONE)
+    return hb_blob_reference (data->blob);
+
+  const OT::OpenTypeFontFile &ot_file = *data->blob->as<OT::OpenTypeFontFile> ();
+  unsigned int base_offset;
+  const OT::OpenTypeFontFace &ot_face = ot_file.get_face (data->index, &base_offset);
+
+  const OT::OpenTypeTable &table = ot_face.get_table_by_tag (tag);
+
+  hb_blob_t *blob = hb_blob_create_sub_blob (data->blob, base_offset + table.offset, table.length);
+
+  return blob;
+}
+
+/**
+ * hb_face_create: (Xconstructor)
+ * @blob:
+ * @index:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_face_create (hb_blob_t    *blob,
+		unsigned int  index)
+{
+  hb_face_t *face;
+
+  if (unlikely (!blob))
+    blob = hb_blob_get_empty ();
+
+  blob = hb_sanitize_context_t ().sanitize_blob<OT::OpenTypeFontFile> (hb_blob_reference (blob));
+
+  hb_face_for_data_closure_t *closure = _hb_face_for_data_closure_create (blob, index);
+
+  if (unlikely (!closure))
+  {
+    hb_blob_destroy (blob);
+    return hb_face_get_empty ();
+  }
+
+  face = hb_face_create_for_tables (_hb_face_for_data_reference_table,
+				    closure,
+				    _hb_face_for_data_closure_destroy);
+
+  face->index = index;
+
+  return face;
+}
+
+/**
+ * hb_face_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full)
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_face_get_empty ()
+{
+  return const_cast<hb_face_t *> (&Null (hb_face_t));
+}
+
+
+/**
+ * hb_face_reference: (skip)
+ * @face: a face.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_face_reference (hb_face_t *face)
+{
+  return hb_object_reference (face);
+}
+
+/**
+ * hb_face_destroy: (skip)
+ * @face: a face.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_face_destroy (hb_face_t *face)
+{
+  if (!hb_object_destroy (face)) return;
+
+  for (hb_face_t::plan_node_t *node = face->shape_plans; node; )
+  {
+    hb_face_t::plan_node_t *next = node->next;
+    hb_shape_plan_destroy (node->shape_plan);
+    free (node);
+    node = next;
+  }
+
+  face->data.fini ();
+  face->table.fini ();
+
+  if (face->destroy)
+    face->destroy (face->user_data);
+
+  free (face);
+}
+
+/**
+ * hb_face_set_user_data: (skip)
+ * @face: a face.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_face_set_user_data (hb_face_t          *face,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace)
+{
+  return hb_object_set_user_data (face, key, data, destroy, replace);
+}
+
+/**
+ * hb_face_get_user_data: (skip)
+ * @face: a face.
+ * @key:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_face_get_user_data (const hb_face_t    *face,
+		       hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (face, key);
+}
+
+/**
+ * hb_face_make_immutable:
+ * @face: a face.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_face_make_immutable (hb_face_t *face)
+{
+  if (hb_object_is_immutable (face))
+    return;
+
+  hb_object_make_immutable (face);
+}
+
+/**
+ * hb_face_is_immutable:
+ * @face: a face.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_face_is_immutable (const hb_face_t *face)
+{
+  return hb_object_is_immutable (face);
+}
+
+
+/**
+ * hb_face_reference_table:
+ * @face: a face.
+ * @tag:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_face_reference_table (const hb_face_t *face,
+			 hb_tag_t tag)
+{
+  if (unlikely (tag == HB_TAG_NONE))
+    return hb_blob_get_empty ();
+
+  return face->reference_table (tag);
+}
+
+/**
+ * hb_face_reference_blob:
+ * @face: a face.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_blob_t *
+hb_face_reference_blob (hb_face_t *face)
+{
+  return face->reference_table (HB_TAG_NONE);
+}
+
+/**
+ * hb_face_set_index:
+ * @face: a face.
+ * @index:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_face_set_index (hb_face_t    *face,
+		   unsigned int  index)
+{
+  if (hb_object_is_immutable (face))
+    return;
+
+  face->index = index;
+}
+
+/**
+ * hb_face_get_index:
+ * @face: a face.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+unsigned int
+hb_face_get_index (const hb_face_t *face)
+{
+  return face->index;
+}
+
+/**
+ * hb_face_set_upem:
+ * @face: a face.
+ * @upem:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_face_set_upem (hb_face_t    *face,
+		  unsigned int  upem)
+{
+  if (hb_object_is_immutable (face))
+    return;
+
+  face->upem.set_relaxed (upem);
+}
+
+/**
+ * hb_face_get_upem:
+ * @face: a face.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+unsigned int
+hb_face_get_upem (const hb_face_t *face)
+{
+  return face->get_upem ();
+}
+
+/**
+ * hb_face_set_glyph_count:
+ * @face: a face.
+ * @glyph_count:
+ *
+ *
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_face_set_glyph_count (hb_face_t    *face,
+			 unsigned int  glyph_count)
+{
+  if (hb_object_is_immutable (face))
+    return;
+
+  face->num_glyphs.set_relaxed (glyph_count);
+}
+
+/**
+ * hb_face_get_glyph_count:
+ * @face: a face.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.7
+ **/
+unsigned int
+hb_face_get_glyph_count (const hb_face_t *face)
+{
+  return face->get_num_glyphs ();
+}
+
+/**
+ * hb_face_get_table_tags:
+ * @face: a face.
+ * @start_offset: index of first tag to return.
+ * @table_count: input length of @table_tags array, output number of items written.
+ * @table_tags: array to write tags into.
+ *
+ * Retrieves table tags for a face, if possible.
+ *
+ * Return value: total number of tables, or 0 if not possible to list.
+ *
+ * Since: 1.6.0
+ **/
+unsigned int
+hb_face_get_table_tags (const hb_face_t *face,
+			unsigned int  start_offset,
+			unsigned int *table_count, /* IN/OUT */
+			hb_tag_t     *table_tags /* OUT */)
+{
+  if (face->destroy != (hb_destroy_func_t) _hb_face_for_data_closure_destroy)
+  {
+    if (table_count)
+      *table_count = 0;
+    return 0;
+  }
+
+  hb_face_for_data_closure_t *data = (hb_face_for_data_closure_t *) face->user_data;
+
+  const OT::OpenTypeFontFile &ot_file = *data->blob->as<OT::OpenTypeFontFile> ();
+  const OT::OpenTypeFontFace &ot_face = ot_file.get_face (data->index);
+
+  return ot_face.get_table_tags (start_offset, table_count, table_tags);
+}
+
+
+/*
+ * Character set.
+ */
+
+
+#ifndef HB_NO_FACE_COLLECT_UNICODES
+/**
+ * hb_face_collect_unicodes:
+ * @face: font face.
+ * @out: set to add Unicode characters covered by @face to.
+ *
+ * Since: 1.9.0
+ */
+void
+hb_face_collect_unicodes (hb_face_t *face,
+			  hb_set_t  *out)
+{
+  face->table.cmap->collect_unicodes (out, face->get_num_glyphs ());
+}
+/**
+ * hb_face_collect_variation_selectors:
+ * @face: font face.
+ * @out: set to add Variation Selector characters covered by @face to.
+ *
+ *
+ *
+ * Since: 1.9.0
+ */
+void
+hb_face_collect_variation_selectors (hb_face_t *face,
+				     hb_set_t  *out)
+{
+  face->table.cmap->collect_variation_selectors (out);
+}
+/**
+ * hb_face_collect_variation_unicodes:
+ * @face: font face.
+ * @out: set to add Unicode characters for @variation_selector covered by @face to.
+ *
+ *
+ *
+ * Since: 1.9.0
+ */
+void
+hb_face_collect_variation_unicodes (hb_face_t *face,
+				    hb_codepoint_t variation_selector,
+				    hb_set_t  *out)
+{
+  face->table.cmap->collect_variation_unicodes (variation_selector, out);
+}
+#endif
+
+
+/*
+ * face-builder: A face that has add_table().
+ */
+
+struct hb_face_builder_data_t
+{
+  struct table_entry_t
+  {
+    int cmp (hb_tag_t t) const
+    {
+      if (t < tag) return -1;
+      if (t > tag) return -1;
+      return 0;
+    }
+
+    hb_tag_t   tag;
+    hb_blob_t *blob;
+  };
+
+  hb_vector_t<table_entry_t> tables;
+};
+
+static hb_face_builder_data_t *
+_hb_face_builder_data_create ()
+{
+  hb_face_builder_data_t *data = (hb_face_builder_data_t *) calloc (1, sizeof (hb_face_builder_data_t));
+  if (unlikely (!data))
+    return nullptr;
+
+  data->tables.init ();
+
+  return data;
+}
+
+static void
+_hb_face_builder_data_destroy (void *user_data)
+{
+  hb_face_builder_data_t *data = (hb_face_builder_data_t *) user_data;
+
+  for (unsigned int i = 0; i < data->tables.length; i++)
+    hb_blob_destroy (data->tables[i].blob);
+
+  data->tables.fini ();
+
+  free (data);
+}
+
+static hb_blob_t *
+_hb_face_builder_data_reference_blob (hb_face_builder_data_t *data)
+{
+
+  unsigned int table_count = data->tables.length;
+  unsigned int face_length = table_count * 16 + 12;
+
+  for (unsigned int i = 0; i < table_count; i++)
+    face_length += hb_ceil_to_4 (hb_blob_get_length (data->tables[i].blob));
+
+  char *buf = (char *) malloc (face_length);
+  if (unlikely (!buf))
+    return nullptr;
+
+  hb_serialize_context_t c (buf, face_length);
+  c.propagate_error (data->tables);
+  OT::OpenTypeFontFile *f = c.start_serialize<OT::OpenTypeFontFile> ();
+
+  bool is_cff = data->tables.lsearch (HB_TAG ('C','F','F',' ')) || data->tables.lsearch (HB_TAG ('C','F','F','2'));
+  hb_tag_t sfnt_tag = is_cff ? OT::OpenTypeFontFile::CFFTag : OT::OpenTypeFontFile::TrueTypeTag;
+
+  bool ret = f->serialize_single (&c, sfnt_tag, data->tables.as_array ());
+
+  c.end_serialize ();
+
+  if (unlikely (!ret))
+  {
+    free (buf);
+    return nullptr;
+  }
+
+  return hb_blob_create (buf, face_length, HB_MEMORY_MODE_WRITABLE, buf, free);
+}
+
+static hb_blob_t *
+_hb_face_builder_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  hb_face_builder_data_t *data = (hb_face_builder_data_t *) user_data;
+
+  if (!tag)
+    return _hb_face_builder_data_reference_blob (data);
+
+  hb_face_builder_data_t::table_entry_t *entry = data->tables.lsearch (tag);
+  if (entry)
+    return hb_blob_reference (entry->blob);
+
+  return nullptr;
+}
+
+
+/**
+ * hb_face_builder_create:
+ *
+ * Creates a #hb_face_t that can be used with hb_face_builder_add_table().
+ * After tables are added to the face, it can be compiled to a binary
+ * font file by calling hb_face_reference_blob().
+ *
+ * Return value: (transfer full): New face.
+ *
+ * Since: 1.9.0
+ **/
+hb_face_t *
+hb_face_builder_create ()
+{
+  hb_face_builder_data_t *data = _hb_face_builder_data_create ();
+  if (unlikely (!data)) return hb_face_get_empty ();
+
+  return hb_face_create_for_tables (_hb_face_builder_reference_table,
+				    data,
+				    _hb_face_builder_data_destroy);
+}
+
+/**
+ * hb_face_builder_add_table:
+ *
+ * Add table for @tag with data provided by @blob to the face.  @face must
+ * be created using hb_face_builder_create().
+ *
+ * Since: 1.9.0
+ **/
+hb_bool_t
+hb_face_builder_add_table (hb_face_t *face, hb_tag_t tag, hb_blob_t *blob)
+{
+  if (unlikely (face->destroy != (hb_destroy_func_t) _hb_face_builder_data_destroy))
+    return false;
+
+  hb_face_builder_data_t *data = (hb_face_builder_data_t *) face->user_data;
+
+  hb_face_builder_data_t::table_entry_t *entry = data->tables.push ();
+  if (data->tables.in_error())
+    return false;
+
+  entry->tag = tag;
+  entry->blob = hb_blob_reference (blob);
+
+  return true;
+}
diff --git a/thirdparty/harfbuzz/src/hb-face.h b/thirdparty/harfbuzz/src/hb-face.h
new file mode 100644
index 0000000000..e8ff090d55
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-face.h
@@ -0,0 +1,158 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_FACE_H
+#define HB_FACE_H
+
+#include "hb-common.h"
+#include "hb-blob.h"
+#include "hb-set.h"
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN unsigned int
+hb_face_count (hb_blob_t *blob);
+
+
+/*
+ * hb_face_t
+ */
+
+typedef struct hb_face_t hb_face_t;
+
+HB_EXTERN hb_face_t *
+hb_face_create (hb_blob_t    *blob,
+		unsigned int  index);
+
+typedef hb_blob_t * (*hb_reference_table_func_t)  (hb_face_t *face, hb_tag_t tag, void *user_data);
+
+/* calls destroy() when not needing user_data anymore */
+HB_EXTERN hb_face_t *
+hb_face_create_for_tables (hb_reference_table_func_t  reference_table_func,
+			   void                      *user_data,
+			   hb_destroy_func_t          destroy);
+
+HB_EXTERN hb_face_t *
+hb_face_get_empty (void);
+
+HB_EXTERN hb_face_t *
+hb_face_reference (hb_face_t *face);
+
+HB_EXTERN void
+hb_face_destroy (hb_face_t *face);
+
+HB_EXTERN hb_bool_t
+hb_face_set_user_data (hb_face_t          *face,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace);
+
+HB_EXTERN void *
+hb_face_get_user_data (const hb_face_t    *face,
+		       hb_user_data_key_t *key);
+
+HB_EXTERN void
+hb_face_make_immutable (hb_face_t *face);
+
+HB_EXTERN hb_bool_t
+hb_face_is_immutable (const hb_face_t *face);
+
+
+HB_EXTERN hb_blob_t *
+hb_face_reference_table (const hb_face_t *face,
+			 hb_tag_t tag);
+
+HB_EXTERN hb_blob_t *
+hb_face_reference_blob (hb_face_t *face);
+
+HB_EXTERN void
+hb_face_set_index (hb_face_t    *face,
+		   unsigned int  index);
+
+HB_EXTERN unsigned int
+hb_face_get_index (const hb_face_t *face);
+
+HB_EXTERN void
+hb_face_set_upem (hb_face_t    *face,
+		  unsigned int  upem);
+
+HB_EXTERN unsigned int
+hb_face_get_upem (const hb_face_t *face);
+
+HB_EXTERN void
+hb_face_set_glyph_count (hb_face_t    *face,
+			 unsigned int  glyph_count);
+
+HB_EXTERN unsigned int
+hb_face_get_glyph_count (const hb_face_t *face);
+
+HB_EXTERN unsigned int
+hb_face_get_table_tags (const hb_face_t *face,
+			unsigned int  start_offset,
+			unsigned int *table_count, /* IN/OUT */
+			hb_tag_t     *table_tags /* OUT */);
+
+
+/*
+ * Character set.
+ */
+
+HB_EXTERN void
+hb_face_collect_unicodes (hb_face_t *face,
+			  hb_set_t  *out);
+
+HB_EXTERN void
+hb_face_collect_variation_selectors (hb_face_t *face,
+				     hb_set_t  *out);
+
+HB_EXTERN void
+hb_face_collect_variation_unicodes (hb_face_t *face,
+				    hb_codepoint_t variation_selector,
+				    hb_set_t  *out);
+
+
+/*
+ * Builder face.
+ */
+
+HB_EXTERN hb_face_t *
+hb_face_builder_create (void);
+
+HB_EXTERN hb_bool_t
+hb_face_builder_add_table (hb_face_t *face,
+			   hb_tag_t   tag,
+			   hb_blob_t *blob);
+
+
+HB_END_DECLS
+
+#endif /* HB_FACE_H */
diff --git a/thirdparty/harfbuzz/src/hb-face.hh b/thirdparty/harfbuzz/src/hb-face.hh
new file mode 100644
index 0000000000..f1b472ccf3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-face.hh
@@ -0,0 +1,109 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_FACE_HH
+#define HB_FACE_HH
+
+#include "hb.hh"
+
+#include "hb-shaper.hh"
+#include "hb-shape-plan.hh"
+#include "hb-ot-face.hh"
+
+
+/*
+ * hb_face_t
+ */
+
+#define HB_SHAPER_IMPLEMENT(shaper) HB_SHAPER_DATA_INSTANTIATE_SHAPERS(shaper, face);
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+
+struct hb_face_t
+{
+  hb_object_header_t header;
+
+  hb_reference_table_func_t  reference_table_func;
+  void                      *user_data;
+  hb_destroy_func_t          destroy;
+
+  unsigned int index;			/* Face index in a collection, zero-based. */
+  mutable hb_atomic_int_t upem;		/* Units-per-EM. */
+  mutable hb_atomic_int_t num_glyphs;	/* Number of glyphs. */
+
+  hb_shaper_object_dataset_t<hb_face_t> data;/* Various shaper data. */
+  hb_ot_face_t table;			/* All the face's tables. */
+
+  /* Cache */
+  struct plan_node_t
+  {
+    hb_shape_plan_t *shape_plan;
+    plan_node_t *next;
+  };
+  hb_atomic_ptr_t<plan_node_t> shape_plans;
+
+  hb_blob_t *reference_table (hb_tag_t tag) const
+  {
+    hb_blob_t *blob;
+
+    if (unlikely (!reference_table_func))
+      return hb_blob_get_empty ();
+
+    blob = reference_table_func (/*XXX*/const_cast<hb_face_t *> (this), tag, user_data);
+    if (unlikely (!blob))
+      return hb_blob_get_empty ();
+
+    return blob;
+  }
+
+  HB_PURE_FUNC unsigned int get_upem () const
+  {
+    unsigned int ret = upem.get_relaxed ();
+    if (unlikely (!ret))
+    {
+      return load_upem ();
+    }
+    return ret;
+  }
+
+  unsigned int get_num_glyphs () const
+  {
+    unsigned int ret = num_glyphs.get_relaxed ();
+    if (unlikely (ret == UINT_MAX))
+      return load_num_glyphs ();
+    return ret;
+  }
+
+  private:
+  HB_INTERNAL unsigned int load_upem () const;
+  HB_INTERNAL unsigned int load_num_glyphs () const;
+};
+DECLARE_NULL_INSTANCE (hb_face_t);
+
+
+#endif /* HB_FACE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-fallback-shape.cc b/thirdparty/harfbuzz/src/hb-fallback-shape.cc
new file mode 100644
index 0000000000..c5b7c2c230
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-fallback-shape.cc
@@ -0,0 +1,125 @@
+/*
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb-shaper-impl.hh"
+
+#ifndef HB_NO_FALLBACK_SHAPE
+
+/*
+ * shaper face data
+ */
+
+struct hb_fallback_face_data_t {};
+
+hb_fallback_face_data_t *
+_hb_fallback_shaper_face_data_create (hb_face_t *face HB_UNUSED)
+{
+  return (hb_fallback_face_data_t *) HB_SHAPER_DATA_SUCCEEDED;
+}
+
+void
+_hb_fallback_shaper_face_data_destroy (hb_fallback_face_data_t *data HB_UNUSED)
+{
+}
+
+
+/*
+ * shaper font data
+ */
+
+struct hb_fallback_font_data_t {};
+
+hb_fallback_font_data_t *
+_hb_fallback_shaper_font_data_create (hb_font_t *font HB_UNUSED)
+{
+  return (hb_fallback_font_data_t *) HB_SHAPER_DATA_SUCCEEDED;
+}
+
+void
+_hb_fallback_shaper_font_data_destroy (hb_fallback_font_data_t *data HB_UNUSED)
+{
+}
+
+
+/*
+ * shaper
+ */
+
+hb_bool_t
+_hb_fallback_shape (hb_shape_plan_t    *shape_plan HB_UNUSED,
+		    hb_font_t          *font,
+		    hb_buffer_t        *buffer,
+		    const hb_feature_t *features HB_UNUSED,
+		    unsigned int        num_features HB_UNUSED)
+{
+  /* TODO
+   *
+   * - Apply fallback kern.
+   * - Handle Variation Selectors?
+   * - Apply normalization?
+   *
+   * This will make the fallback shaper into a dumb "TrueType"
+   * shaper which many people unfortunately still request.
+   */
+
+  hb_codepoint_t space;
+  bool has_space = (bool) font->get_nominal_glyph (' ', &space);
+
+  buffer->clear_positions ();
+
+  hb_direction_t direction = buffer->props.direction;
+  hb_unicode_funcs_t *unicode = buffer->unicode;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pos = buffer->pos;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    if (has_space && unicode->is_default_ignorable (info[i].codepoint)) {
+      info[i].codepoint = space;
+      pos[i].x_advance = 0;
+      pos[i].y_advance = 0;
+      continue;
+    }
+    (void) font->get_nominal_glyph (info[i].codepoint, &info[i].codepoint);
+    font->get_glyph_advance_for_direction (info[i].codepoint,
+					   direction,
+					   &pos[i].x_advance,
+					   &pos[i].y_advance);
+    font->subtract_glyph_origin_for_direction (info[i].codepoint,
+					       direction,
+					       &pos[i].x_offset,
+					       &pos[i].y_offset);
+  }
+
+  if (HB_DIRECTION_IS_BACKWARD (direction))
+    hb_buffer_reverse (buffer);
+
+  buffer->safe_to_break_all ();
+
+  return true;
+}
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-font.cc b/thirdparty/harfbuzz/src/hb-font.cc
new file mode 100644
index 0000000000..27959487dc
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-font.cc
@@ -0,0 +1,2186 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-font.hh"
+#include "hb-machinery.hh"
+
+#include "hb-ot.h"
+
+#include "hb-ot-var-avar-table.hh"
+#include "hb-ot-var-fvar-table.hh"
+
+
+/**
+ * SECTION:hb-font
+ * @title: hb-font
+ * @short_description: Font objects
+ * @include: hb.h
+ *
+ * Font objects represent a font face at a certain size and other
+ * parameters (pixels per EM, points per EM, variation settings.)
+ * Fonts are created from font faces, and are used as input to
+ * hb_shape() among other things.
+ **/
+
+
+/*
+ * hb_font_funcs_t
+ */
+
+static hb_bool_t
+hb_font_get_font_h_extents_nil (hb_font_t *font HB_UNUSED,
+				void *font_data HB_UNUSED,
+				hb_font_extents_t *extents,
+				void *user_data HB_UNUSED)
+{
+  memset (extents, 0, sizeof (*extents));
+  return false;
+}
+static hb_bool_t
+hb_font_get_font_h_extents_default (hb_font_t *font,
+				    void *font_data HB_UNUSED,
+				    hb_font_extents_t *extents,
+				    void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_font_h_extents (extents);
+  if (ret) {
+    extents->ascender = font->parent_scale_y_distance (extents->ascender);
+    extents->descender = font->parent_scale_y_distance (extents->descender);
+    extents->line_gap = font->parent_scale_y_distance (extents->line_gap);
+  }
+  return ret;
+}
+
+static hb_bool_t
+hb_font_get_font_v_extents_nil (hb_font_t *font HB_UNUSED,
+				void *font_data HB_UNUSED,
+				hb_font_extents_t *extents,
+				void *user_data HB_UNUSED)
+{
+  memset (extents, 0, sizeof (*extents));
+  return false;
+}
+static hb_bool_t
+hb_font_get_font_v_extents_default (hb_font_t *font,
+				    void *font_data HB_UNUSED,
+				    hb_font_extents_t *extents,
+				    void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_font_v_extents (extents);
+  if (ret) {
+    extents->ascender = font->parent_scale_x_distance (extents->ascender);
+    extents->descender = font->parent_scale_x_distance (extents->descender);
+    extents->line_gap = font->parent_scale_x_distance (extents->line_gap);
+  }
+  return ret;
+}
+
+static hb_bool_t
+hb_font_get_nominal_glyph_nil (hb_font_t *font HB_UNUSED,
+			       void *font_data HB_UNUSED,
+			       hb_codepoint_t unicode HB_UNUSED,
+			       hb_codepoint_t *glyph,
+			       void *user_data HB_UNUSED)
+{
+  *glyph = 0;
+  return false;
+}
+static hb_bool_t
+hb_font_get_nominal_glyph_default (hb_font_t *font,
+				   void *font_data HB_UNUSED,
+				   hb_codepoint_t unicode,
+				   hb_codepoint_t *glyph,
+				   void *user_data HB_UNUSED)
+{
+  if (font->has_nominal_glyphs_func_set ())
+  {
+    return font->get_nominal_glyphs (1, &unicode, 0, glyph, 0);
+  }
+  return font->parent->get_nominal_glyph (unicode, glyph);
+}
+
+#define hb_font_get_nominal_glyphs_nil hb_font_get_nominal_glyphs_default
+static unsigned int
+hb_font_get_nominal_glyphs_default (hb_font_t *font,
+				    void *font_data HB_UNUSED,
+				    unsigned int count,
+				    const hb_codepoint_t *first_unicode,
+				    unsigned int unicode_stride,
+				    hb_codepoint_t *first_glyph,
+				    unsigned int glyph_stride,
+				    void *user_data HB_UNUSED)
+{
+  if (font->has_nominal_glyph_func_set ())
+  {
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (!font->get_nominal_glyph (*first_unicode, first_glyph))
+	return i;
+
+      first_unicode = &StructAtOffsetUnaligned<hb_codepoint_t> (first_unicode, unicode_stride);
+      first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+    }
+    return count;
+  }
+
+  return font->parent->get_nominal_glyphs (count,
+					   first_unicode, unicode_stride,
+					   first_glyph, glyph_stride);
+}
+
+static hb_bool_t
+hb_font_get_variation_glyph_nil (hb_font_t *font HB_UNUSED,
+				 void *font_data HB_UNUSED,
+				 hb_codepoint_t unicode HB_UNUSED,
+				 hb_codepoint_t variation_selector HB_UNUSED,
+				 hb_codepoint_t *glyph,
+				 void *user_data HB_UNUSED)
+{
+  *glyph = 0;
+  return false;
+}
+static hb_bool_t
+hb_font_get_variation_glyph_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t unicode,
+				     hb_codepoint_t variation_selector,
+				     hb_codepoint_t *glyph,
+				     void *user_data HB_UNUSED)
+{
+  return font->parent->get_variation_glyph (unicode, variation_selector, glyph);
+}
+
+
+static hb_position_t
+hb_font_get_glyph_h_advance_nil (hb_font_t *font,
+				 void *font_data HB_UNUSED,
+				 hb_codepoint_t glyph HB_UNUSED,
+				 void *user_data HB_UNUSED)
+{
+  return font->x_scale;
+}
+static hb_position_t
+hb_font_get_glyph_h_advance_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t glyph,
+				     void *user_data HB_UNUSED)
+{
+  if (font->has_glyph_h_advances_func_set ())
+  {
+    hb_position_t ret;
+    font->get_glyph_h_advances (1, &glyph, 0, &ret, 0);
+    return ret;
+  }
+  return font->parent_scale_x_distance (font->parent->get_glyph_h_advance (glyph));
+}
+
+static hb_position_t
+hb_font_get_glyph_v_advance_nil (hb_font_t *font,
+				 void *font_data HB_UNUSED,
+				 hb_codepoint_t glyph HB_UNUSED,
+				 void *user_data HB_UNUSED)
+{
+  /* TODO use font_extents.ascender+descender */
+  return font->y_scale;
+}
+static hb_position_t
+hb_font_get_glyph_v_advance_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t glyph,
+				     void *user_data HB_UNUSED)
+{
+  if (font->has_glyph_v_advances_func_set ())
+  {
+    hb_position_t ret;
+    font->get_glyph_v_advances (1, &glyph, 0, &ret, 0);
+    return ret;
+  }
+  return font->parent_scale_y_distance (font->parent->get_glyph_v_advance (glyph));
+}
+
+#define hb_font_get_glyph_h_advances_nil hb_font_get_glyph_h_advances_default
+static void
+hb_font_get_glyph_h_advances_default (hb_font_t* font,
+				      void* font_data HB_UNUSED,
+				      unsigned int count,
+				      const hb_codepoint_t *first_glyph,
+				      unsigned int glyph_stride,
+				      hb_position_t *first_advance,
+				      unsigned int advance_stride,
+				      void *user_data HB_UNUSED)
+{
+  if (font->has_glyph_h_advance_func_set ())
+  {
+    for (unsigned int i = 0; i < count; i++)
+    {
+      *first_advance = font->get_glyph_h_advance (*first_glyph);
+      first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+      first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+    }
+    return;
+  }
+
+  font->parent->get_glyph_h_advances (count,
+				      first_glyph, glyph_stride,
+				      first_advance, advance_stride);
+  for (unsigned int i = 0; i < count; i++)
+  {
+    *first_advance = font->parent_scale_x_distance (*first_advance);
+    first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+  }
+}
+
+#define hb_font_get_glyph_v_advances_nil hb_font_get_glyph_v_advances_default
+static void
+hb_font_get_glyph_v_advances_default (hb_font_t* font,
+				      void* font_data HB_UNUSED,
+				      unsigned int count,
+				      const hb_codepoint_t *first_glyph,
+				      unsigned int glyph_stride,
+				      hb_position_t *first_advance,
+				      unsigned int advance_stride,
+				      void *user_data HB_UNUSED)
+{
+  if (font->has_glyph_v_advance_func_set ())
+  {
+    for (unsigned int i = 0; i < count; i++)
+    {
+      *first_advance = font->get_glyph_v_advance (*first_glyph);
+      first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+      first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+    }
+    return;
+  }
+
+  font->parent->get_glyph_v_advances (count,
+				      first_glyph, glyph_stride,
+				      first_advance, advance_stride);
+  for (unsigned int i = 0; i < count; i++)
+  {
+    *first_advance = font->parent_scale_y_distance (*first_advance);
+    first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+  }
+}
+
+static hb_bool_t
+hb_font_get_glyph_h_origin_nil (hb_font_t *font HB_UNUSED,
+				void *font_data HB_UNUSED,
+				hb_codepoint_t glyph HB_UNUSED,
+				hb_position_t *x,
+				hb_position_t *y,
+				void *user_data HB_UNUSED)
+{
+  *x = *y = 0;
+  return true;
+}
+static hb_bool_t
+hb_font_get_glyph_h_origin_default (hb_font_t *font,
+				    void *font_data HB_UNUSED,
+				    hb_codepoint_t glyph,
+				    hb_position_t *x,
+				    hb_position_t *y,
+				    void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_glyph_h_origin (glyph, x, y);
+  if (ret)
+    font->parent_scale_position (x, y);
+  return ret;
+}
+
+static hb_bool_t
+hb_font_get_glyph_v_origin_nil (hb_font_t *font HB_UNUSED,
+				void *font_data HB_UNUSED,
+				hb_codepoint_t glyph HB_UNUSED,
+				hb_position_t *x,
+				hb_position_t *y,
+				void *user_data HB_UNUSED)
+{
+  *x = *y = 0;
+  return false;
+}
+static hb_bool_t
+hb_font_get_glyph_v_origin_default (hb_font_t *font,
+				    void *font_data HB_UNUSED,
+				    hb_codepoint_t glyph,
+				    hb_position_t *x,
+				    hb_position_t *y,
+				    void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_glyph_v_origin (glyph, x, y);
+  if (ret)
+    font->parent_scale_position (x, y);
+  return ret;
+}
+
+static hb_position_t
+hb_font_get_glyph_h_kerning_nil (hb_font_t *font HB_UNUSED,
+				 void *font_data HB_UNUSED,
+				 hb_codepoint_t left_glyph HB_UNUSED,
+				 hb_codepoint_t right_glyph HB_UNUSED,
+				 void *user_data HB_UNUSED)
+{
+  return 0;
+}
+static hb_position_t
+hb_font_get_glyph_h_kerning_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t left_glyph,
+				     hb_codepoint_t right_glyph,
+				     void *user_data HB_UNUSED)
+{
+  return font->parent_scale_x_distance (font->parent->get_glyph_h_kerning (left_glyph, right_glyph));
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+static hb_position_t
+hb_font_get_glyph_v_kerning_nil (hb_font_t *font HB_UNUSED,
+				 void *font_data HB_UNUSED,
+				 hb_codepoint_t top_glyph HB_UNUSED,
+				 hb_codepoint_t bottom_glyph HB_UNUSED,
+				 void *user_data HB_UNUSED)
+{
+  return 0;
+}
+static hb_position_t
+hb_font_get_glyph_v_kerning_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t top_glyph,
+				     hb_codepoint_t bottom_glyph,
+				     void *user_data HB_UNUSED)
+{
+  return font->parent_scale_y_distance (font->parent->get_glyph_v_kerning (top_glyph, bottom_glyph));
+}
+#endif
+
+static hb_bool_t
+hb_font_get_glyph_extents_nil (hb_font_t *font HB_UNUSED,
+			       void *font_data HB_UNUSED,
+			       hb_codepoint_t glyph HB_UNUSED,
+			       hb_glyph_extents_t *extents,
+			       void *user_data HB_UNUSED)
+{
+  memset (extents, 0, sizeof (*extents));
+  return false;
+}
+static hb_bool_t
+hb_font_get_glyph_extents_default (hb_font_t *font,
+				   void *font_data HB_UNUSED,
+				   hb_codepoint_t glyph,
+				   hb_glyph_extents_t *extents,
+				   void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_glyph_extents (glyph, extents);
+  if (ret) {
+    font->parent_scale_position (&extents->x_bearing, &extents->y_bearing);
+    font->parent_scale_distance (&extents->width, &extents->height);
+  }
+  return ret;
+}
+
+static hb_bool_t
+hb_font_get_glyph_contour_point_nil (hb_font_t *font HB_UNUSED,
+				     void *font_data HB_UNUSED,
+				     hb_codepoint_t glyph HB_UNUSED,
+				     unsigned int point_index HB_UNUSED,
+				     hb_position_t *x,
+				     hb_position_t *y,
+				     void *user_data HB_UNUSED)
+{
+  *x = *y = 0;
+  return false;
+}
+static hb_bool_t
+hb_font_get_glyph_contour_point_default (hb_font_t *font,
+					 void *font_data HB_UNUSED,
+					 hb_codepoint_t glyph,
+					 unsigned int point_index,
+					 hb_position_t *x,
+					 hb_position_t *y,
+					 void *user_data HB_UNUSED)
+{
+  hb_bool_t ret = font->parent->get_glyph_contour_point (glyph, point_index, x, y);
+  if (ret)
+    font->parent_scale_position (x, y);
+  return ret;
+}
+
+static hb_bool_t
+hb_font_get_glyph_name_nil (hb_font_t *font HB_UNUSED,
+			    void *font_data HB_UNUSED,
+			    hb_codepoint_t glyph HB_UNUSED,
+			    char *name, unsigned int size,
+			    void *user_data HB_UNUSED)
+{
+  if (size) *name = '\0';
+  return false;
+}
+static hb_bool_t
+hb_font_get_glyph_name_default (hb_font_t *font,
+				void *font_data HB_UNUSED,
+				hb_codepoint_t glyph,
+				char *name, unsigned int size,
+				void *user_data HB_UNUSED)
+{
+  return font->parent->get_glyph_name (glyph, name, size);
+}
+
+static hb_bool_t
+hb_font_get_glyph_from_name_nil (hb_font_t *font HB_UNUSED,
+				 void *font_data HB_UNUSED,
+				 const char *name HB_UNUSED,
+				 int len HB_UNUSED, /* -1 means nul-terminated */
+				 hb_codepoint_t *glyph,
+				 void *user_data HB_UNUSED)
+{
+  *glyph = 0;
+  return false;
+}
+static hb_bool_t
+hb_font_get_glyph_from_name_default (hb_font_t *font,
+				     void *font_data HB_UNUSED,
+				     const char *name, int len, /* -1 means nul-terminated */
+				     hb_codepoint_t *glyph,
+				     void *user_data HB_UNUSED)
+{
+  return font->parent->get_glyph_from_name (name, len, glyph);
+}
+
+DEFINE_NULL_INSTANCE (hb_font_funcs_t) =
+{
+  HB_OBJECT_HEADER_STATIC,
+
+  {
+#define HB_FONT_FUNC_IMPLEMENT(name) nullptr,
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  },
+  {
+#define HB_FONT_FUNC_IMPLEMENT(name) nullptr,
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  },
+  {
+    {
+#define HB_FONT_FUNC_IMPLEMENT(name) hb_font_get_##name##_nil,
+      HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+    }
+  }
+};
+
+static const hb_font_funcs_t _hb_font_funcs_default = {
+  HB_OBJECT_HEADER_STATIC,
+
+  {
+#define HB_FONT_FUNC_IMPLEMENT(name) nullptr,
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  },
+  {
+#define HB_FONT_FUNC_IMPLEMENT(name) nullptr,
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  },
+  {
+    {
+#define HB_FONT_FUNC_IMPLEMENT(name) hb_font_get_##name##_default,
+      HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+    }
+  }
+};
+
+
+/**
+ * hb_font_funcs_create: (Xconstructor)
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_funcs_t *
+hb_font_funcs_create ()
+{
+  hb_font_funcs_t *ffuncs;
+
+  if (!(ffuncs = hb_object_create<hb_font_funcs_t> ()))
+    return hb_font_funcs_get_empty ();
+
+  ffuncs->get = _hb_font_funcs_default.get;
+
+  return ffuncs;
+}
+
+/**
+ * hb_font_funcs_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_funcs_t *
+hb_font_funcs_get_empty ()
+{
+  return const_cast<hb_font_funcs_t *> (&_hb_font_funcs_default);
+}
+
+/**
+ * hb_font_funcs_reference: (skip)
+ * @ffuncs: font functions.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_font_funcs_t *
+hb_font_funcs_reference (hb_font_funcs_t *ffuncs)
+{
+  return hb_object_reference (ffuncs);
+}
+
+/**
+ * hb_font_funcs_destroy: (skip)
+ * @ffuncs: font functions.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_funcs_destroy (hb_font_funcs_t *ffuncs)
+{
+  if (!hb_object_destroy (ffuncs)) return;
+
+#define HB_FONT_FUNC_IMPLEMENT(name) if (ffuncs->destroy.name) \
+  ffuncs->destroy.name (ffuncs->user_data.name);
+  HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+
+  free (ffuncs);
+}
+
+/**
+ * hb_font_funcs_set_user_data: (skip)
+ * @ffuncs: font functions.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_funcs_set_user_data (hb_font_funcs_t    *ffuncs,
+			     hb_user_data_key_t *key,
+			     void *              data,
+			     hb_destroy_func_t   destroy,
+			     hb_bool_t           replace)
+{
+  return hb_object_set_user_data (ffuncs, key, data, destroy, replace);
+}
+
+/**
+ * hb_font_funcs_get_user_data: (skip)
+ * @ffuncs: font functions.
+ * @key:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_font_funcs_get_user_data (hb_font_funcs_t    *ffuncs,
+			     hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (ffuncs, key);
+}
+
+
+/**
+ * hb_font_funcs_make_immutable:
+ * @ffuncs: font functions.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_funcs_make_immutable (hb_font_funcs_t *ffuncs)
+{
+  if (hb_object_is_immutable (ffuncs))
+    return;
+
+  hb_object_make_immutable (ffuncs);
+}
+
+/**
+ * hb_font_funcs_is_immutable:
+ * @ffuncs: font functions.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_funcs_is_immutable (hb_font_funcs_t *ffuncs)
+{
+  return hb_object_is_immutable (ffuncs);
+}
+
+
+#define HB_FONT_FUNC_IMPLEMENT(name) \
+									 \
+void                                                                     \
+hb_font_funcs_set_##name##_func (hb_font_funcs_t             *ffuncs,    \
+				 hb_font_get_##name##_func_t  func,      \
+				 void                        *user_data, \
+				 hb_destroy_func_t            destroy)   \
+{                                                                        \
+  if (hb_object_is_immutable (ffuncs))                                   \
+  {                                                                      \
+    if (destroy)                                                         \
+      destroy (user_data);                                               \
+    return;                                                              \
+  }                                                                      \
+									 \
+  if (ffuncs->destroy.name)                                              \
+    ffuncs->destroy.name (ffuncs->user_data.name);                       \
+									 \
+  if (func) {                                                            \
+    ffuncs->get.f.name = func;                                           \
+    ffuncs->user_data.name = user_data;                                  \
+    ffuncs->destroy.name = destroy;                                      \
+  } else {                                                               \
+    ffuncs->get.f.name = hb_font_get_##name##_default;                   \
+    ffuncs->user_data.name = nullptr;                                    \
+    ffuncs->destroy.name = nullptr;                                      \
+  }                                                                      \
+}
+
+HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+
+bool
+hb_font_t::has_func_set (unsigned int i)
+{
+  return this->klass->get.array[i] != _hb_font_funcs_default.get.array[i];
+}
+
+bool
+hb_font_t::has_func (unsigned int i)
+{
+  return has_func_set (i) ||
+	 (parent && parent != &_hb_Null_hb_font_t && parent->has_func (i));
+}
+
+/* Public getters */
+
+/**
+ * hb_font_get_h_extents:
+ * @font: a font.
+ * @extents: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.1.3
+ **/
+hb_bool_t
+hb_font_get_h_extents (hb_font_t *font,
+		       hb_font_extents_t *extents)
+{
+  return font->get_font_h_extents (extents);
+}
+
+/**
+ * hb_font_get_v_extents:
+ * @font: a font.
+ * @extents: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.1.3
+ **/
+hb_bool_t
+hb_font_get_v_extents (hb_font_t *font,
+		       hb_font_extents_t *extents)
+{
+  return font->get_font_v_extents (extents);
+}
+
+/**
+ * hb_font_get_glyph:
+ * @font: a font.
+ * @unicode:
+ * @variation_selector:
+ * @glyph: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph (hb_font_t *font,
+		   hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+		   hb_codepoint_t *glyph)
+{
+  if (unlikely (variation_selector))
+    return font->get_variation_glyph (unicode, variation_selector, glyph);
+  return font->get_nominal_glyph (unicode, glyph);
+}
+
+/**
+ * hb_font_get_nominal_glyph:
+ * @font: a font.
+ * @unicode:
+ * @glyph: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.2.3
+ **/
+hb_bool_t
+hb_font_get_nominal_glyph (hb_font_t *font,
+			   hb_codepoint_t unicode,
+			   hb_codepoint_t *glyph)
+{
+  return font->get_nominal_glyph (unicode, glyph);
+}
+
+/**
+ * hb_font_get_nominal_glyphs:
+ * @font: a font.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 2.6.3
+ **/
+unsigned int
+hb_font_get_nominal_glyphs (hb_font_t *font,
+			    unsigned int count,
+			    const hb_codepoint_t *first_unicode,
+			    unsigned int unicode_stride,
+			    hb_codepoint_t *first_glyph,
+			    unsigned int glyph_stride)
+{
+  return font->get_nominal_glyphs (count,
+				   first_unicode, unicode_stride,
+				   first_glyph, glyph_stride);
+}
+
+/**
+ * hb_font_get_variation_glyph:
+ * @font: a font.
+ * @unicode:
+ * @variation_selector:
+ * @glyph: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.2.3
+ **/
+hb_bool_t
+hb_font_get_variation_glyph (hb_font_t *font,
+			     hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+			     hb_codepoint_t *glyph)
+{
+  return font->get_variation_glyph (unicode, variation_selector, glyph);
+}
+
+/**
+ * hb_font_get_glyph_h_advance:
+ * @font: a font.
+ * @glyph:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_position_t
+hb_font_get_glyph_h_advance (hb_font_t *font,
+			     hb_codepoint_t glyph)
+{
+  return font->get_glyph_h_advance (glyph);
+}
+
+/**
+ * hb_font_get_glyph_v_advance:
+ * @font: a font.
+ * @glyph:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_position_t
+hb_font_get_glyph_v_advance (hb_font_t *font,
+			     hb_codepoint_t glyph)
+{
+  return font->get_glyph_v_advance (glyph);
+}
+
+/**
+ * hb_font_get_glyph_h_advances:
+ * @font: a font.
+ *
+ *
+ *
+ * Since: 1.8.6
+ **/
+void
+hb_font_get_glyph_h_advances (hb_font_t* font,
+			      unsigned int count,
+			      const hb_codepoint_t *first_glyph,
+			      unsigned glyph_stride,
+			      hb_position_t *first_advance,
+			      unsigned advance_stride)
+{
+  font->get_glyph_h_advances (count, first_glyph, glyph_stride, first_advance, advance_stride);
+}
+/**
+ * hb_font_get_glyph_v_advances:
+ * @font: a font.
+ *
+ *
+ *
+ * Since: 1.8.6
+ **/
+void
+hb_font_get_glyph_v_advances (hb_font_t* font,
+			      unsigned int count,
+			      const hb_codepoint_t *first_glyph,
+			      unsigned glyph_stride,
+			      hb_position_t *first_advance,
+			      unsigned advance_stride)
+{
+  font->get_glyph_v_advances (count, first_glyph, glyph_stride, first_advance, advance_stride);
+}
+
+/**
+ * hb_font_get_glyph_h_origin:
+ * @font: a font.
+ * @glyph:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_h_origin (hb_font_t *font,
+			    hb_codepoint_t glyph,
+			    hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_h_origin (glyph, x, y);
+}
+
+/**
+ * hb_font_get_glyph_v_origin:
+ * @font: a font.
+ * @glyph:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_v_origin (hb_font_t *font,
+			    hb_codepoint_t glyph,
+			    hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_v_origin (glyph, x, y);
+}
+
+/**
+ * hb_font_get_glyph_h_kerning:
+ * @font: a font.
+ * @left_glyph:
+ * @right_glyph:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_position_t
+hb_font_get_glyph_h_kerning (hb_font_t *font,
+			     hb_codepoint_t left_glyph, hb_codepoint_t right_glyph)
+{
+  return font->get_glyph_h_kerning (left_glyph, right_glyph);
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_font_get_glyph_v_kerning:
+ * @font: a font.
+ * @top_glyph:
+ * @bottom_glyph:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+hb_position_t
+hb_font_get_glyph_v_kerning (hb_font_t *font,
+			     hb_codepoint_t top_glyph, hb_codepoint_t bottom_glyph)
+{
+  return font->get_glyph_v_kerning (top_glyph, bottom_glyph);
+}
+#endif
+
+/**
+ * hb_font_get_glyph_extents:
+ * @font: a font.
+ * @glyph:
+ * @extents: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_extents (hb_font_t *font,
+			   hb_codepoint_t glyph,
+			   hb_glyph_extents_t *extents)
+{
+  return font->get_glyph_extents (glyph, extents);
+}
+
+/**
+ * hb_font_get_glyph_contour_point:
+ * @font: a font.
+ * @glyph:
+ * @point_index:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_contour_point (hb_font_t *font,
+				 hb_codepoint_t glyph, unsigned int point_index,
+				 hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_contour_point (glyph, point_index, x, y);
+}
+
+/**
+ * hb_font_get_glyph_name:
+ * @font: a font.
+ * @glyph:
+ * @name: (array length=size):
+ * @size:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_name (hb_font_t *font,
+			hb_codepoint_t glyph,
+			char *name, unsigned int size)
+{
+  return font->get_glyph_name (glyph, name, size);
+}
+
+/**
+ * hb_font_get_glyph_from_name:
+ * @font: a font.
+ * @name: (array length=len):
+ * @len:
+ * @glyph: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_from_name (hb_font_t *font,
+			     const char *name, int len, /* -1 means nul-terminated */
+			     hb_codepoint_t *glyph)
+{
+  return font->get_glyph_from_name (name, len, glyph);
+}
+
+
+/* A bit higher-level, and with fallback */
+
+/**
+ * hb_font_get_extents_for_direction:
+ * @font: a font.
+ * @direction:
+ * @extents: (out):
+ *
+ *
+ *
+ * Since: 1.1.3
+ **/
+void
+hb_font_get_extents_for_direction (hb_font_t *font,
+				   hb_direction_t direction,
+				   hb_font_extents_t *extents)
+{
+  return font->get_extents_for_direction (direction, extents);
+}
+/**
+ * hb_font_get_glyph_advance_for_direction:
+ * @font: a font.
+ * @glyph:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_get_glyph_advance_for_direction (hb_font_t *font,
+					 hb_codepoint_t glyph,
+					 hb_direction_t direction,
+					 hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_advance_for_direction (glyph, direction, x, y);
+}
+/**
+ * hb_font_get_glyph_advances_for_direction:
+ * @font: a font.
+ * @direction:
+ *
+ *
+ *
+ * Since: 1.8.6
+ **/
+HB_EXTERN void
+hb_font_get_glyph_advances_for_direction (hb_font_t* font,
+					  hb_direction_t direction,
+					  unsigned int count,
+					  const hb_codepoint_t *first_glyph,
+					  unsigned glyph_stride,
+					  hb_position_t *first_advance,
+					  unsigned advance_stride)
+{
+  font->get_glyph_advances_for_direction (direction, count, first_glyph, glyph_stride, first_advance, advance_stride);
+}
+
+/**
+ * hb_font_get_glyph_origin_for_direction:
+ * @font: a font.
+ * @glyph:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_get_glyph_origin_for_direction (hb_font_t *font,
+					hb_codepoint_t glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_origin_for_direction (glyph, direction, x, y);
+}
+
+/**
+ * hb_font_add_glyph_origin_for_direction:
+ * @font: a font.
+ * @glyph:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_add_glyph_origin_for_direction (hb_font_t *font,
+					hb_codepoint_t glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y)
+{
+  return font->add_glyph_origin_for_direction (glyph, direction, x, y);
+}
+
+/**
+ * hb_font_subtract_glyph_origin_for_direction:
+ * @font: a font.
+ * @glyph:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_subtract_glyph_origin_for_direction (hb_font_t *font,
+					     hb_codepoint_t glyph,
+					     hb_direction_t direction,
+					     hb_position_t *x, hb_position_t *y)
+{
+  return font->subtract_glyph_origin_for_direction (glyph, direction, x, y);
+}
+
+/**
+ * hb_font_get_glyph_kerning_for_direction:
+ * @font: a font.
+ * @first_glyph:
+ * @second_glyph:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_get_glyph_kerning_for_direction (hb_font_t *font,
+					 hb_codepoint_t first_glyph, hb_codepoint_t second_glyph,
+					 hb_direction_t direction,
+					 hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_kerning_for_direction (first_glyph, second_glyph, direction, x, y);
+}
+
+/**
+ * hb_font_get_glyph_extents_for_origin:
+ * @font: a font.
+ * @glyph:
+ * @direction:
+ * @extents: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_extents_for_origin (hb_font_t *font,
+				      hb_codepoint_t glyph,
+				      hb_direction_t direction,
+				      hb_glyph_extents_t *extents)
+{
+  return font->get_glyph_extents_for_origin (glyph, direction, extents);
+}
+
+/**
+ * hb_font_get_glyph_contour_point_for_origin:
+ * @font: a font.
+ * @glyph:
+ * @point_index:
+ * @direction:
+ * @x: (out):
+ * @y: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_get_glyph_contour_point_for_origin (hb_font_t *font,
+					    hb_codepoint_t glyph, unsigned int point_index,
+					    hb_direction_t direction,
+					    hb_position_t *x, hb_position_t *y)
+{
+  return font->get_glyph_contour_point_for_origin (glyph, point_index, direction, x, y);
+}
+
+/* Generates gidDDD if glyph has no name. */
+/**
+ * hb_font_glyph_to_string:
+ * @font: a font.
+ * @glyph:
+ * @s: (array length=size):
+ * @size:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_glyph_to_string (hb_font_t *font,
+			 hb_codepoint_t glyph,
+			 char *s, unsigned int size)
+{
+  font->glyph_to_string (glyph, s, size);
+}
+
+/* Parses gidDDD and uniUUUU strings automatically. */
+/**
+ * hb_font_glyph_from_string:
+ * @font: a font.
+ * @s: (array length=len) (element-type uint8_t):
+ * @len:
+ * @glyph: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_glyph_from_string (hb_font_t *font,
+			   const char *s, int len, /* -1 means nul-terminated */
+			   hb_codepoint_t *glyph)
+{
+  return font->glyph_from_string (s, len, glyph);
+}
+
+
+/*
+ * hb_font_t
+ */
+
+DEFINE_NULL_INSTANCE (hb_font_t) =
+{
+  HB_OBJECT_HEADER_STATIC,
+
+  nullptr, /* parent */
+  const_cast<hb_face_t *> (&_hb_Null_hb_face_t),
+
+  1000, /* x_scale */
+  1000, /* y_scale */
+  1<<16, /* x_mult */
+  1<<16, /* y_mult */
+
+  0, /* x_ppem */
+  0, /* y_ppem */
+  0, /* ptem */
+
+  0, /* num_coords */
+  nullptr, /* coords */
+  nullptr, /* design_coords */
+
+  const_cast<hb_font_funcs_t *> (&_hb_Null_hb_font_funcs_t),
+
+  /* Zero for the rest is fine. */
+};
+
+
+static hb_font_t *
+_hb_font_create (hb_face_t *face)
+{
+  hb_font_t *font;
+
+  if (unlikely (!face))
+    face = hb_face_get_empty ();
+  if (!(font = hb_object_create<hb_font_t> ()))
+    return hb_font_get_empty ();
+
+  hb_face_make_immutable (face);
+  font->parent = hb_font_get_empty ();
+  font->face = hb_face_reference (face);
+  font->klass = hb_font_funcs_get_empty ();
+  font->data.init0 (font);
+  font->x_scale = font->y_scale = hb_face_get_upem (face);
+  font->x_mult = font->y_mult = 1 << 16;
+
+  return font;
+}
+
+/**
+ * hb_font_create: (Xconstructor)
+ * @face: a face.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_font_create (hb_face_t *face)
+{
+  hb_font_t *font = _hb_font_create (face);
+
+#ifndef HB_NO_OT_FONT
+  /* Install our in-house, very lightweight, funcs. */
+  hb_ot_font_set_funcs (font);
+#endif
+
+  return font;
+}
+
+static void
+_hb_font_adopt_var_coords (hb_font_t *font,
+			   int *coords, /* 2.14 normalized */
+			   float *design_coords,
+			   unsigned int coords_length)
+{
+  free (font->coords);
+  free (font->design_coords);
+
+  font->coords = coords;
+  font->design_coords = design_coords;
+  font->num_coords = coords_length;
+}
+
+/**
+ * hb_font_create_sub_font:
+ * @parent: parent font.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_font_create_sub_font (hb_font_t *parent)
+{
+  if (unlikely (!parent))
+    parent = hb_font_get_empty ();
+
+  hb_font_t *font = _hb_font_create (parent->face);
+
+  if (unlikely (hb_object_is_immutable (font)))
+    return font;
+
+  font->parent = hb_font_reference (parent);
+
+  font->x_scale = parent->x_scale;
+  font->y_scale = parent->y_scale;
+  font->mults_changed ();
+  font->x_ppem = parent->x_ppem;
+  font->y_ppem = parent->y_ppem;
+  font->ptem = parent->ptem;
+
+  unsigned int num_coords = parent->num_coords;
+  if (num_coords)
+  {
+    int *coords = (int *) calloc (num_coords, sizeof (parent->coords[0]));
+    float *design_coords = (float *) calloc (num_coords, sizeof (parent->design_coords[0]));
+    if (likely (coords && design_coords))
+    {
+      memcpy (coords, parent->coords, num_coords * sizeof (parent->coords[0]));
+      memcpy (design_coords, parent->design_coords, num_coords * sizeof (parent->design_coords[0]));
+      _hb_font_adopt_var_coords (font, coords, design_coords, num_coords);
+    }
+    else
+    {
+      free (coords);
+      free (design_coords);
+    }
+  }
+
+  return font;
+}
+
+/**
+ * hb_font_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full)
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_font_get_empty ()
+{
+  return const_cast<hb_font_t *> (&Null (hb_font_t));
+}
+
+/**
+ * hb_font_reference: (skip)
+ * @font: a font.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_font_reference (hb_font_t *font)
+{
+  return hb_object_reference (font);
+}
+
+/**
+ * hb_font_destroy: (skip)
+ * @font: a font.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_destroy (hb_font_t *font)
+{
+  if (!hb_object_destroy (font)) return;
+
+  font->data.fini ();
+
+  if (font->destroy)
+    font->destroy (font->user_data);
+
+  hb_font_destroy (font->parent);
+  hb_face_destroy (font->face);
+  hb_font_funcs_destroy (font->klass);
+
+  free (font->coords);
+  free (font->design_coords);
+
+  free (font);
+}
+
+/**
+ * hb_font_set_user_data: (skip)
+ * @font: a font.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_set_user_data (hb_font_t          *font,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace)
+{
+  return hb_object_set_user_data (font, key, data, destroy, replace);
+}
+
+/**
+ * hb_font_get_user_data: (skip)
+ * @font: a font.
+ * @key:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_font_get_user_data (hb_font_t          *font,
+		       hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (font, key);
+}
+
+/**
+ * hb_font_make_immutable:
+ * @font: a font.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_make_immutable (hb_font_t *font)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  if (font->parent)
+    hb_font_make_immutable (font->parent);
+
+  hb_object_make_immutable (font);
+}
+
+/**
+ * hb_font_is_immutable:
+ * @font: a font.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_font_is_immutable (hb_font_t *font)
+{
+  return hb_object_is_immutable (font);
+}
+
+/**
+ * hb_font_set_parent:
+ * @font: a font.
+ * @parent: new parent.
+ *
+ * Sets parent font of @font.
+ *
+ * Since: 1.0.5
+ **/
+void
+hb_font_set_parent (hb_font_t *font,
+		    hb_font_t *parent)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  if (!parent)
+    parent = hb_font_get_empty ();
+
+  hb_font_t *old = font->parent;
+
+  font->parent = hb_font_reference (parent);
+
+  hb_font_destroy (old);
+}
+
+/**
+ * hb_font_get_parent:
+ * @font: a font.
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_font_get_parent (hb_font_t *font)
+{
+  return font->parent;
+}
+
+/**
+ * hb_font_set_face:
+ * @font: a font.
+ * @face: new face.
+ *
+ * Sets font-face of @font.
+ *
+ * Since: 1.4.3
+ **/
+void
+hb_font_set_face (hb_font_t *font,
+		  hb_face_t *face)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  if (unlikely (!face))
+    face = hb_face_get_empty ();
+
+  hb_face_t *old = font->face;
+
+  hb_face_make_immutable (face);
+  font->face = hb_face_reference (face);
+  font->mults_changed ();
+
+  hb_face_destroy (old);
+}
+
+/**
+ * hb_font_get_face:
+ * @font: a font.
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_font_get_face (hb_font_t *font)
+{
+  return font->face;
+}
+
+
+/**
+ * hb_font_set_funcs:
+ * @font: a font.
+ * @klass: (closure font_data) (destroy destroy) (scope notified):
+ * @font_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_set_funcs (hb_font_t         *font,
+		   hb_font_funcs_t   *klass,
+		   void              *font_data,
+		   hb_destroy_func_t  destroy)
+{
+  if (hb_object_is_immutable (font))
+  {
+    if (destroy)
+      destroy (font_data);
+    return;
+  }
+
+  if (font->destroy)
+    font->destroy (font->user_data);
+
+  if (!klass)
+    klass = hb_font_funcs_get_empty ();
+
+  hb_font_funcs_reference (klass);
+  hb_font_funcs_destroy (font->klass);
+  font->klass = klass;
+  font->user_data = font_data;
+  font->destroy = destroy;
+}
+
+/**
+ * hb_font_set_funcs_data:
+ * @font: a font.
+ * @font_data: (destroy destroy) (scope notified):
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_set_funcs_data (hb_font_t         *font,
+			void              *font_data,
+			hb_destroy_func_t  destroy)
+{
+  /* Destroy user_data? */
+  if (hb_object_is_immutable (font))
+  {
+    if (destroy)
+      destroy (font_data);
+    return;
+  }
+
+  if (font->destroy)
+    font->destroy (font->user_data);
+
+  font->user_data = font_data;
+  font->destroy = destroy;
+}
+
+
+/**
+ * hb_font_set_scale:
+ * @font: a font.
+ * @x_scale:
+ * @y_scale:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_set_scale (hb_font_t *font,
+		   int x_scale,
+		   int y_scale)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  font->x_scale = x_scale;
+  font->y_scale = y_scale;
+  font->mults_changed ();
+}
+
+/**
+ * hb_font_get_scale:
+ * @font: a font.
+ * @x_scale: (out):
+ * @y_scale: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_get_scale (hb_font_t *font,
+		   int *x_scale,
+		   int *y_scale)
+{
+  if (x_scale) *x_scale = font->x_scale;
+  if (y_scale) *y_scale = font->y_scale;
+}
+
+/**
+ * hb_font_set_ppem:
+ * @font: a font.
+ * @x_ppem:
+ * @y_ppem:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_set_ppem (hb_font_t *font,
+		  unsigned int x_ppem,
+		  unsigned int y_ppem)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  font->x_ppem = x_ppem;
+  font->y_ppem = y_ppem;
+}
+
+/**
+ * hb_font_get_ppem:
+ * @font: a font.
+ * @x_ppem: (out):
+ * @y_ppem: (out):
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_font_get_ppem (hb_font_t *font,
+		  unsigned int *x_ppem,
+		  unsigned int *y_ppem)
+{
+  if (x_ppem) *x_ppem = font->x_ppem;
+  if (y_ppem) *y_ppem = font->y_ppem;
+}
+
+/**
+ * hb_font_set_ptem:
+ * @font: a font.
+ * @ptem: font size in points.
+ *
+ * Sets "point size" of the font.  Set to 0 to unset.
+ *
+ * There are 72 points in an inch.
+ *
+ * Since: 1.6.0
+ **/
+void
+hb_font_set_ptem (hb_font_t *font, float ptem)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  font->ptem = ptem;
+}
+
+/**
+ * hb_font_get_ptem:
+ * @font: a font.
+ *
+ * Gets the "point size" of the font.  A value of 0 means unset.
+ *
+ * Return value: Point size.
+ *
+ * Since: 0.9.2
+ **/
+float
+hb_font_get_ptem (hb_font_t *font)
+{
+  return font->ptem;
+}
+
+#ifndef HB_NO_VAR
+/*
+ * Variations
+ */
+
+/**
+ * hb_font_set_variations:
+ *
+ * Since: 1.4.2
+ */
+void
+hb_font_set_variations (hb_font_t *font,
+			const hb_variation_t *variations,
+			unsigned int variations_length)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  if (!variations_length)
+  {
+    hb_font_set_var_coords_normalized (font, nullptr, 0);
+    return;
+  }
+
+  unsigned int coords_length = hb_ot_var_get_axis_count (font->face);
+
+  int *normalized = coords_length ? (int *) calloc (coords_length, sizeof (int)) : nullptr;
+  float *design_coords = coords_length ? (float *) calloc (coords_length, sizeof (float)) : nullptr;
+
+  if (unlikely (coords_length && !(normalized && design_coords)))
+  {
+    free (normalized);
+    free (design_coords);
+    return;
+  }
+
+  const OT::fvar &fvar = *font->face->table.fvar;
+  for (unsigned int i = 0; i < variations_length; i++)
+  {
+    hb_ot_var_axis_info_t info;
+    if (hb_ot_var_find_axis_info (font->face, variations[i].tag, &info) &&
+	info.axis_index < coords_length)
+    {
+      float v = variations[i].value;
+      design_coords[info.axis_index] = v;
+      normalized[info.axis_index] = fvar.normalize_axis_value (info.axis_index, v);
+    }
+  }
+  font->face->table.avar->map_coords (normalized, coords_length);
+
+  _hb_font_adopt_var_coords (font, normalized, design_coords, coords_length);
+}
+
+/**
+ * hb_font_set_var_coords_design:
+ *
+ * Since: 1.4.2
+ */
+void
+hb_font_set_var_coords_design (hb_font_t *font,
+			       const float *coords,
+			       unsigned int coords_length)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  int *normalized = coords_length ? (int *) calloc (coords_length, sizeof (int)) : nullptr;
+  float *design_coords = coords_length ? (float *) calloc (coords_length, sizeof (float)) : nullptr;
+
+  if (unlikely (coords_length && !(normalized && design_coords)))
+  {
+    free (normalized);
+    free (design_coords);
+    return;
+  }
+
+  if (coords_length)
+    memcpy (design_coords, coords, coords_length * sizeof (font->design_coords[0]));
+
+  hb_ot_var_normalize_coords (font->face, coords_length, coords, normalized);
+  _hb_font_adopt_var_coords (font, normalized, design_coords, coords_length);
+}
+
+/**
+ * hb_font_set_var_named_instance:
+ * @font: a font.
+ * @instance_index: named instance index.
+ *
+ * Sets design coords of a font from a named instance index.
+ *
+ * Since: 2.6.0
+ */
+void
+hb_font_set_var_named_instance (hb_font_t *font,
+				unsigned instance_index)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  unsigned int coords_length = hb_ot_var_named_instance_get_design_coords (font->face, instance_index, nullptr, nullptr);
+
+  float *coords = coords_length ? (float *) calloc (coords_length, sizeof (float)) : nullptr;
+  if (unlikely (coords_length && !coords))
+    return;
+
+  hb_ot_var_named_instance_get_design_coords (font->face, instance_index, &coords_length, coords);
+  hb_font_set_var_coords_design (font, coords, coords_length);
+  free (coords);
+}
+
+/**
+ * hb_font_set_var_coords_normalized:
+ *
+ * Since: 1.4.2
+ */
+void
+hb_font_set_var_coords_normalized (hb_font_t *font,
+				   const int *coords, /* 2.14 normalized */
+				   unsigned int coords_length)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  int *copy = coords_length ? (int *) calloc (coords_length, sizeof (coords[0])) : nullptr;
+  int *unmapped = coords_length ? (int *) calloc (coords_length, sizeof (coords[0])) : nullptr;
+  float *design_coords = coords_length ? (float *) calloc (coords_length, sizeof (design_coords[0])) : nullptr;
+
+  if (unlikely (coords_length && !(copy && unmapped && design_coords)))
+  {
+    free (copy);
+    free (unmapped);
+    free (design_coords);
+    return;
+  }
+
+  if (coords_length)
+  {
+    memcpy (copy, coords, coords_length * sizeof (coords[0]));
+    memcpy (unmapped, coords, coords_length * sizeof (coords[0]));
+  }
+
+  /* Best effort design coords simulation */
+  font->face->table.avar->unmap_coords (unmapped, coords_length);
+  for (unsigned int i = 0; i < coords_length; ++i)
+    design_coords[i] = font->face->table.fvar->unnormalize_axis_value (i, unmapped[i]);
+  free (unmapped);
+
+  _hb_font_adopt_var_coords (font, copy, design_coords, coords_length);
+}
+
+/**
+ * hb_font_get_var_coords_normalized:
+ *
+ * Return value is valid as long as variation coordinates of the font
+ * are not modified.
+ *
+ * Since: 1.4.2
+ */
+const int *
+hb_font_get_var_coords_normalized (hb_font_t *font,
+				   unsigned int *length)
+{
+  if (length)
+    *length = font->num_coords;
+
+  return font->coords;
+}
+
+#ifdef HB_EXPERIMENTAL_API
+/**
+ * hb_font_get_var_coords_design:
+ *
+ * Return value is valid as long as variation coordinates of the font
+ * are not modified.
+ *
+ * Since: EXPERIMENTAL
+ */
+const float *
+hb_font_get_var_coords_design (hb_font_t *font,
+			       unsigned int *length)
+{
+  if (length)
+    *length = font->num_coords;
+
+  return font->design_coords;
+}
+#endif
+#endif
+
+#ifndef HB_DISABLE_DEPRECATED
+/*
+ * Deprecated get_glyph_func():
+ */
+
+struct hb_trampoline_closure_t
+{
+  void *user_data;
+  hb_destroy_func_t destroy;
+  unsigned int ref_count;
+};
+
+template <typename FuncType>
+struct hb_trampoline_t
+{
+  hb_trampoline_closure_t closure; /* Must be first. */
+  FuncType func;
+};
+
+template <typename FuncType>
+static hb_trampoline_t<FuncType> *
+trampoline_create (FuncType           func,
+		   void              *user_data,
+		   hb_destroy_func_t  destroy)
+{
+  typedef hb_trampoline_t<FuncType> trampoline_t;
+
+  trampoline_t *trampoline = (trampoline_t *) calloc (1, sizeof (trampoline_t));
+
+  if (unlikely (!trampoline))
+    return nullptr;
+
+  trampoline->closure.user_data = user_data;
+  trampoline->closure.destroy = destroy;
+  trampoline->closure.ref_count = 1;
+  trampoline->func = func;
+
+  return trampoline;
+}
+
+static void
+trampoline_reference (hb_trampoline_closure_t *closure)
+{
+  closure->ref_count++;
+}
+
+static void
+trampoline_destroy (void *user_data)
+{
+  hb_trampoline_closure_t *closure = (hb_trampoline_closure_t *) user_data;
+
+  if (--closure->ref_count)
+    return;
+
+  if (closure->destroy)
+    closure->destroy (closure->user_data);
+  free (closure);
+}
+
+typedef hb_trampoline_t<hb_font_get_glyph_func_t> hb_font_get_glyph_trampoline_t;
+
+static hb_bool_t
+hb_font_get_nominal_glyph_trampoline (hb_font_t *font,
+				      void *font_data,
+				      hb_codepoint_t unicode,
+				      hb_codepoint_t *glyph,
+				      void *user_data)
+{
+  hb_font_get_glyph_trampoline_t *trampoline = (hb_font_get_glyph_trampoline_t *) user_data;
+  return trampoline->func (font, font_data, unicode, 0, glyph, trampoline->closure.user_data);
+}
+
+static hb_bool_t
+hb_font_get_variation_glyph_trampoline (hb_font_t *font,
+					void *font_data,
+					hb_codepoint_t unicode,
+					hb_codepoint_t variation_selector,
+					hb_codepoint_t *glyph,
+					void *user_data)
+{
+  hb_font_get_glyph_trampoline_t *trampoline = (hb_font_get_glyph_trampoline_t *) user_data;
+  return trampoline->func (font, font_data, unicode, variation_selector, glyph, trampoline->closure.user_data);
+}
+
+/**
+ * hb_font_funcs_set_glyph_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified): callback function.
+ * @user_data: data to pass to @func.
+ * @destroy: function to call when @user_data is not needed anymore.
+ *
+ * Deprecated.  Use hb_font_funcs_set_nominal_glyph_func() and
+ * hb_font_funcs_set_variation_glyph_func() instead.
+ *
+ * Since: 0.9.2
+ * Deprecated: 1.2.3
+ **/
+void
+hb_font_funcs_set_glyph_func (hb_font_funcs_t *ffuncs,
+			      hb_font_get_glyph_func_t func,
+			      void *user_data, hb_destroy_func_t destroy)
+{
+  if (hb_object_is_immutable (ffuncs))
+  {
+    if (destroy)
+      destroy (user_data);
+    return;
+  }
+
+  hb_font_get_glyph_trampoline_t *trampoline;
+
+  trampoline = trampoline_create (func, user_data, destroy);
+  if (unlikely (!trampoline))
+  {
+    if (destroy)
+      destroy (user_data);
+    return;
+  }
+
+  hb_font_funcs_set_nominal_glyph_func (ffuncs,
+					hb_font_get_nominal_glyph_trampoline,
+					trampoline,
+					trampoline_destroy);
+
+  trampoline_reference (&trampoline->closure);
+  hb_font_funcs_set_variation_glyph_func (ffuncs,
+					  hb_font_get_variation_glyph_trampoline,
+					  trampoline,
+					  trampoline_destroy);
+}
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-font.h b/thirdparty/harfbuzz/src/hb-font.h
new file mode 100644
index 0000000000..e1a5719f1d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-font.h
@@ -0,0 +1,735 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_FONT_H
+#define HB_FONT_H
+
+#include "hb-common.h"
+#include "hb-face.h"
+#include "hb-draw.h"
+
+HB_BEGIN_DECLS
+
+
+typedef struct hb_font_t hb_font_t;
+
+
+/*
+ * hb_font_funcs_t
+ */
+
+typedef struct hb_font_funcs_t hb_font_funcs_t;
+
+HB_EXTERN hb_font_funcs_t *
+hb_font_funcs_create (void);
+
+HB_EXTERN hb_font_funcs_t *
+hb_font_funcs_get_empty (void);
+
+HB_EXTERN hb_font_funcs_t *
+hb_font_funcs_reference (hb_font_funcs_t *ffuncs);
+
+HB_EXTERN void
+hb_font_funcs_destroy (hb_font_funcs_t *ffuncs);
+
+HB_EXTERN hb_bool_t
+hb_font_funcs_set_user_data (hb_font_funcs_t    *ffuncs,
+			     hb_user_data_key_t *key,
+			     void *              data,
+			     hb_destroy_func_t   destroy,
+			     hb_bool_t           replace);
+
+
+HB_EXTERN void *
+hb_font_funcs_get_user_data (hb_font_funcs_t    *ffuncs,
+			     hb_user_data_key_t *key);
+
+
+HB_EXTERN void
+hb_font_funcs_make_immutable (hb_font_funcs_t *ffuncs);
+
+HB_EXTERN hb_bool_t
+hb_font_funcs_is_immutable (hb_font_funcs_t *ffuncs);
+
+
+/* font and glyph extents */
+
+/* Note that typically ascender is positive and descender negative in coordinate systems that grow up. */
+typedef struct hb_font_extents_t
+{
+  hb_position_t ascender; /* typographic ascender. */
+  hb_position_t descender; /* typographic descender. */
+  hb_position_t line_gap; /* suggested line spacing gap. */
+  /*< private >*/
+  hb_position_t reserved9;
+  hb_position_t reserved8;
+  hb_position_t reserved7;
+  hb_position_t reserved6;
+  hb_position_t reserved5;
+  hb_position_t reserved4;
+  hb_position_t reserved3;
+  hb_position_t reserved2;
+  hb_position_t reserved1;
+} hb_font_extents_t;
+
+/* Note that height is negative in coordinate systems that grow up. */
+typedef struct hb_glyph_extents_t
+{
+  hb_position_t x_bearing; /* left side of glyph from origin. */
+  hb_position_t y_bearing; /* top side of glyph from origin. */
+  hb_position_t width; /* distance from left to right side. */
+  hb_position_t height; /* distance from top to bottom side. */
+} hb_glyph_extents_t;
+
+/* func types */
+
+typedef hb_bool_t (*hb_font_get_font_extents_func_t) (hb_font_t *font, void *font_data,
+						       hb_font_extents_t *extents,
+						       void *user_data);
+typedef hb_font_get_font_extents_func_t hb_font_get_font_h_extents_func_t;
+typedef hb_font_get_font_extents_func_t hb_font_get_font_v_extents_func_t;
+
+
+typedef hb_bool_t (*hb_font_get_nominal_glyph_func_t) (hb_font_t *font, void *font_data,
+						       hb_codepoint_t unicode,
+						       hb_codepoint_t *glyph,
+						       void *user_data);
+typedef hb_bool_t (*hb_font_get_variation_glyph_func_t) (hb_font_t *font, void *font_data,
+							 hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+							 hb_codepoint_t *glyph,
+							 void *user_data);
+
+typedef unsigned int (*hb_font_get_nominal_glyphs_func_t) (hb_font_t *font, void *font_data,
+							   unsigned int count,
+							   const hb_codepoint_t *first_unicode,
+							   unsigned int unicode_stride,
+							   hb_codepoint_t *first_glyph,
+							   unsigned int glyph_stride,
+							   void *user_data);
+
+
+typedef hb_position_t (*hb_font_get_glyph_advance_func_t) (hb_font_t *font, void *font_data,
+							   hb_codepoint_t glyph,
+							   void *user_data);
+typedef hb_font_get_glyph_advance_func_t hb_font_get_glyph_h_advance_func_t;
+typedef hb_font_get_glyph_advance_func_t hb_font_get_glyph_v_advance_func_t;
+
+typedef void (*hb_font_get_glyph_advances_func_t) (hb_font_t* font, void* font_data,
+						   unsigned int count,
+						   const hb_codepoint_t *first_glyph,
+						   unsigned glyph_stride,
+						   hb_position_t *first_advance,
+						   unsigned advance_stride,
+						   void *user_data);
+typedef hb_font_get_glyph_advances_func_t hb_font_get_glyph_h_advances_func_t;
+typedef hb_font_get_glyph_advances_func_t hb_font_get_glyph_v_advances_func_t;
+
+typedef hb_bool_t (*hb_font_get_glyph_origin_func_t) (hb_font_t *font, void *font_data,
+						      hb_codepoint_t glyph,
+						      hb_position_t *x, hb_position_t *y,
+						      void *user_data);
+typedef hb_font_get_glyph_origin_func_t hb_font_get_glyph_h_origin_func_t;
+typedef hb_font_get_glyph_origin_func_t hb_font_get_glyph_v_origin_func_t;
+
+typedef hb_position_t (*hb_font_get_glyph_kerning_func_t) (hb_font_t *font, void *font_data,
+							   hb_codepoint_t first_glyph, hb_codepoint_t second_glyph,
+							   void *user_data);
+typedef hb_font_get_glyph_kerning_func_t hb_font_get_glyph_h_kerning_func_t;
+
+
+typedef hb_bool_t (*hb_font_get_glyph_extents_func_t) (hb_font_t *font, void *font_data,
+						       hb_codepoint_t glyph,
+						       hb_glyph_extents_t *extents,
+						       void *user_data);
+typedef hb_bool_t (*hb_font_get_glyph_contour_point_func_t) (hb_font_t *font, void *font_data,
+							     hb_codepoint_t glyph, unsigned int point_index,
+							     hb_position_t *x, hb_position_t *y,
+							     void *user_data);
+
+
+typedef hb_bool_t (*hb_font_get_glyph_name_func_t) (hb_font_t *font, void *font_data,
+						    hb_codepoint_t glyph,
+						    char *name, unsigned int size,
+						    void *user_data);
+typedef hb_bool_t (*hb_font_get_glyph_from_name_func_t) (hb_font_t *font, void *font_data,
+							 const char *name, int len, /* -1 means nul-terminated */
+							 hb_codepoint_t *glyph,
+							 void *user_data);
+
+
+/* func setters */
+
+/**
+ * hb_font_funcs_set_font_h_extents_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.1.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_font_h_extents_func (hb_font_funcs_t *ffuncs,
+				       hb_font_get_font_h_extents_func_t func,
+				       void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_font_v_extents_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.1.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_font_v_extents_func (hb_font_funcs_t *ffuncs,
+				       hb_font_get_font_v_extents_func_t func,
+				       void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_nominal_glyph_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.2.3
+ **/
+HB_EXTERN void
+hb_font_funcs_set_nominal_glyph_func (hb_font_funcs_t *ffuncs,
+				      hb_font_get_nominal_glyph_func_t func,
+				      void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_nominal_glyphs_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 2.0.0
+ **/
+HB_EXTERN void
+hb_font_funcs_set_nominal_glyphs_func (hb_font_funcs_t *ffuncs,
+				       hb_font_get_nominal_glyphs_func_t func,
+				       void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_variation_glyph_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.2.3
+ **/
+HB_EXTERN void
+hb_font_funcs_set_variation_glyph_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_variation_glyph_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_h_advance_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_h_advance_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_h_advance_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_v_advance_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_v_advance_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_v_advance_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_h_advances_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.8.6
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_h_advances_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_h_advances_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_v_advances_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 1.8.6
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_v_advances_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_v_advances_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_h_origin_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_h_origin_func (hb_font_funcs_t *ffuncs,
+				       hb_font_get_glyph_h_origin_func_t func,
+				       void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_v_origin_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_v_origin_func (hb_font_funcs_t *ffuncs,
+				       hb_font_get_glyph_v_origin_func_t func,
+				       void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_h_kerning_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_h_kerning_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_h_kerning_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_extents_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_extents_func (hb_font_funcs_t *ffuncs,
+				      hb_font_get_glyph_extents_func_t func,
+				      void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_contour_point_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_contour_point_func (hb_font_funcs_t *ffuncs,
+					    hb_font_get_glyph_contour_point_func_t func,
+					    void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_name_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_name_func (hb_font_funcs_t *ffuncs,
+				   hb_font_get_glyph_name_func_t func,
+				   void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_font_funcs_set_glyph_from_name_func:
+ * @ffuncs: font functions.
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_font_funcs_set_glyph_from_name_func (hb_font_funcs_t *ffuncs,
+					hb_font_get_glyph_from_name_func_t func,
+					void *user_data, hb_destroy_func_t destroy);
+
+/* func dispatch */
+
+HB_EXTERN hb_bool_t
+hb_font_get_h_extents (hb_font_t *font,
+		       hb_font_extents_t *extents);
+HB_EXTERN hb_bool_t
+hb_font_get_v_extents (hb_font_t *font,
+		       hb_font_extents_t *extents);
+
+HB_EXTERN hb_bool_t
+hb_font_get_nominal_glyph (hb_font_t *font,
+			   hb_codepoint_t unicode,
+			   hb_codepoint_t *glyph);
+HB_EXTERN hb_bool_t
+hb_font_get_variation_glyph (hb_font_t *font,
+			     hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+			     hb_codepoint_t *glyph);
+
+HB_EXTERN unsigned int
+hb_font_get_nominal_glyphs (hb_font_t *font,
+			    unsigned int count,
+			    const hb_codepoint_t *first_unicode,
+			    unsigned int unicode_stride,
+			    hb_codepoint_t *first_glyph,
+			    unsigned int glyph_stride);
+
+HB_EXTERN hb_position_t
+hb_font_get_glyph_h_advance (hb_font_t *font,
+			     hb_codepoint_t glyph);
+HB_EXTERN hb_position_t
+hb_font_get_glyph_v_advance (hb_font_t *font,
+			     hb_codepoint_t glyph);
+
+HB_EXTERN void
+hb_font_get_glyph_h_advances (hb_font_t* font,
+			      unsigned int count,
+			      const hb_codepoint_t *first_glyph,
+			      unsigned glyph_stride,
+			      hb_position_t *first_advance,
+			      unsigned advance_stride);
+HB_EXTERN void
+hb_font_get_glyph_v_advances (hb_font_t* font,
+			      unsigned int count,
+			      const hb_codepoint_t *first_glyph,
+			      unsigned glyph_stride,
+			      hb_position_t *first_advance,
+			      unsigned advance_stride);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_h_origin (hb_font_t *font,
+			    hb_codepoint_t glyph,
+			    hb_position_t *x, hb_position_t *y);
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_v_origin (hb_font_t *font,
+			    hb_codepoint_t glyph,
+			    hb_position_t *x, hb_position_t *y);
+
+HB_EXTERN hb_position_t
+hb_font_get_glyph_h_kerning (hb_font_t *font,
+			     hb_codepoint_t left_glyph, hb_codepoint_t right_glyph);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_extents (hb_font_t *font,
+			   hb_codepoint_t glyph,
+			   hb_glyph_extents_t *extents);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_contour_point (hb_font_t *font,
+				 hb_codepoint_t glyph, unsigned int point_index,
+				 hb_position_t *x, hb_position_t *y);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_name (hb_font_t *font,
+			hb_codepoint_t glyph,
+			char *name, unsigned int size);
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_from_name (hb_font_t *font,
+			     const char *name, int len, /* -1 means nul-terminated */
+			     hb_codepoint_t *glyph);
+
+
+/* high-level funcs, with fallback */
+
+/* Calls either hb_font_get_nominal_glyph() if variation_selector is 0,
+ * otherwise calls hb_font_get_variation_glyph(). */
+HB_EXTERN hb_bool_t
+hb_font_get_glyph (hb_font_t *font,
+		   hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+		   hb_codepoint_t *glyph);
+
+HB_EXTERN void
+hb_font_get_extents_for_direction (hb_font_t *font,
+				   hb_direction_t direction,
+				   hb_font_extents_t *extents);
+HB_EXTERN void
+hb_font_get_glyph_advance_for_direction (hb_font_t *font,
+					 hb_codepoint_t glyph,
+					 hb_direction_t direction,
+					 hb_position_t *x, hb_position_t *y);
+HB_EXTERN void
+hb_font_get_glyph_advances_for_direction (hb_font_t* font,
+					  hb_direction_t direction,
+					  unsigned int count,
+					  const hb_codepoint_t *first_glyph,
+					  unsigned glyph_stride,
+					  hb_position_t *first_advance,
+					  unsigned advance_stride);
+HB_EXTERN void
+hb_font_get_glyph_origin_for_direction (hb_font_t *font,
+					hb_codepoint_t glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y);
+HB_EXTERN void
+hb_font_add_glyph_origin_for_direction (hb_font_t *font,
+					hb_codepoint_t glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y);
+HB_EXTERN void
+hb_font_subtract_glyph_origin_for_direction (hb_font_t *font,
+					     hb_codepoint_t glyph,
+					     hb_direction_t direction,
+					     hb_position_t *x, hb_position_t *y);
+
+HB_EXTERN void
+hb_font_get_glyph_kerning_for_direction (hb_font_t *font,
+					 hb_codepoint_t first_glyph, hb_codepoint_t second_glyph,
+					 hb_direction_t direction,
+					 hb_position_t *x, hb_position_t *y);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_extents_for_origin (hb_font_t *font,
+				      hb_codepoint_t glyph,
+				      hb_direction_t direction,
+				      hb_glyph_extents_t *extents);
+
+HB_EXTERN hb_bool_t
+hb_font_get_glyph_contour_point_for_origin (hb_font_t *font,
+					    hb_codepoint_t glyph, unsigned int point_index,
+					    hb_direction_t direction,
+					    hb_position_t *x, hb_position_t *y);
+
+/* Generates gidDDD if glyph has no name. */
+HB_EXTERN void
+hb_font_glyph_to_string (hb_font_t *font,
+			 hb_codepoint_t glyph,
+			 char *s, unsigned int size);
+/* Parses gidDDD and uniUUUU strings automatically. */
+HB_EXTERN hb_bool_t
+hb_font_glyph_from_string (hb_font_t *font,
+			   const char *s, int len, /* -1 means nul-terminated */
+			   hb_codepoint_t *glyph);
+
+
+/*
+ * hb_font_t
+ */
+
+/* Fonts are very light-weight objects */
+
+HB_EXTERN hb_font_t *
+hb_font_create (hb_face_t *face);
+
+HB_EXTERN hb_font_t *
+hb_font_create_sub_font (hb_font_t *parent);
+
+HB_EXTERN hb_font_t *
+hb_font_get_empty (void);
+
+HB_EXTERN hb_font_t *
+hb_font_reference (hb_font_t *font);
+
+HB_EXTERN void
+hb_font_destroy (hb_font_t *font);
+
+HB_EXTERN hb_bool_t
+hb_font_set_user_data (hb_font_t          *font,
+		       hb_user_data_key_t *key,
+		       void *              data,
+		       hb_destroy_func_t   destroy,
+		       hb_bool_t           replace);
+
+
+HB_EXTERN void *
+hb_font_get_user_data (hb_font_t          *font,
+		       hb_user_data_key_t *key);
+
+HB_EXTERN void
+hb_font_make_immutable (hb_font_t *font);
+
+HB_EXTERN hb_bool_t
+hb_font_is_immutable (hb_font_t *font);
+
+HB_EXTERN void
+hb_font_set_parent (hb_font_t *font,
+		    hb_font_t *parent);
+
+HB_EXTERN hb_font_t *
+hb_font_get_parent (hb_font_t *font);
+
+HB_EXTERN void
+hb_font_set_face (hb_font_t *font,
+		  hb_face_t *face);
+
+HB_EXTERN hb_face_t *
+hb_font_get_face (hb_font_t *font);
+
+
+HB_EXTERN void
+hb_font_set_funcs (hb_font_t         *font,
+		   hb_font_funcs_t   *klass,
+		   void              *font_data,
+		   hb_destroy_func_t  destroy);
+
+/* Be *very* careful with this function! */
+HB_EXTERN void
+hb_font_set_funcs_data (hb_font_t         *font,
+			void              *font_data,
+			hb_destroy_func_t  destroy);
+
+
+HB_EXTERN void
+hb_font_set_scale (hb_font_t *font,
+		   int x_scale,
+		   int y_scale);
+
+HB_EXTERN void
+hb_font_get_scale (hb_font_t *font,
+		   int *x_scale,
+		   int *y_scale);
+
+/*
+ * A zero value means "no hinting in that direction"
+ */
+HB_EXTERN void
+hb_font_set_ppem (hb_font_t *font,
+		  unsigned int x_ppem,
+		  unsigned int y_ppem);
+
+HB_EXTERN void
+hb_font_get_ppem (hb_font_t *font,
+		  unsigned int *x_ppem,
+		  unsigned int *y_ppem);
+
+/*
+ * Point size per EM.  Used for optical-sizing in CoreText.
+ * A value of zero means "not set".
+ */
+HB_EXTERN void
+hb_font_set_ptem (hb_font_t *font, float ptem);
+
+HB_EXTERN float
+hb_font_get_ptem (hb_font_t *font);
+
+HB_EXTERN void
+hb_font_set_variations (hb_font_t *font,
+			const hb_variation_t *variations,
+			unsigned int variations_length);
+
+HB_EXTERN void
+hb_font_set_var_coords_design (hb_font_t *font,
+			       const float *coords,
+			       unsigned int coords_length);
+
+#ifdef HB_EXPERIMENTAL_API
+HB_EXTERN const float *
+hb_font_get_var_coords_design (hb_font_t *font,
+			       unsigned int *length);
+#endif
+
+HB_EXTERN void
+hb_font_set_var_coords_normalized (hb_font_t *font,
+				   const int *coords, /* 2.14 normalized */
+				   unsigned int coords_length);
+
+HB_EXTERN const int *
+hb_font_get_var_coords_normalized (hb_font_t *font,
+				   unsigned int *length);
+
+HB_EXTERN void
+hb_font_set_var_named_instance (hb_font_t *font,
+				unsigned instance_index);
+
+#ifdef HB_EXPERIMENTAL_API
+HB_EXTERN hb_bool_t
+hb_font_draw_glyph (hb_font_t *font, hb_codepoint_t glyph,
+		    const hb_draw_funcs_t *funcs, void *user_data);
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_FONT_H */
diff --git a/thirdparty/harfbuzz/src/hb-font.hh b/thirdparty/harfbuzz/src/hb-font.hh
new file mode 100644
index 0000000000..8fc7f44d44
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-font.hh
@@ -0,0 +1,632 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_FONT_HH
+#define HB_FONT_HH
+
+#include "hb.hh"
+
+#include "hb-face.hh"
+#include "hb-shaper.hh"
+
+
+/*
+ * hb_font_funcs_t
+ */
+
+#define HB_FONT_FUNCS_IMPLEMENT_CALLBACKS \
+  HB_FONT_FUNC_IMPLEMENT (font_h_extents) \
+  HB_FONT_FUNC_IMPLEMENT (font_v_extents) \
+  HB_FONT_FUNC_IMPLEMENT (nominal_glyph) \
+  HB_FONT_FUNC_IMPLEMENT (nominal_glyphs) \
+  HB_FONT_FUNC_IMPLEMENT (variation_glyph) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_h_advance) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_v_advance) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_h_advances) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_v_advances) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_h_origin) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_v_origin) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_h_kerning) \
+  HB_IF_NOT_DEPRECATED (HB_FONT_FUNC_IMPLEMENT (glyph_v_kerning)) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_extents) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_contour_point) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_name) \
+  HB_FONT_FUNC_IMPLEMENT (glyph_from_name) \
+  /* ^--- Add new callbacks here */
+
+struct hb_font_funcs_t
+{
+  hb_object_header_t header;
+
+  struct {
+#define HB_FONT_FUNC_IMPLEMENT(name) void *name;
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  } user_data;
+
+  struct {
+#define HB_FONT_FUNC_IMPLEMENT(name) hb_destroy_func_t name;
+    HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+  } destroy;
+
+  /* Don't access these directly.  Call font->get_*() instead. */
+  union get_t {
+    struct get_funcs_t {
+#define HB_FONT_FUNC_IMPLEMENT(name) hb_font_get_##name##_func_t name;
+      HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+    } f;
+    void (*array[0
+#define HB_FONT_FUNC_IMPLEMENT(name) +1
+      HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+		]) ();
+  } get;
+};
+DECLARE_NULL_INSTANCE (hb_font_funcs_t);
+
+
+/*
+ * hb_font_t
+ */
+
+#define HB_SHAPER_IMPLEMENT(shaper) HB_SHAPER_DATA_INSTANTIATE_SHAPERS(shaper, font);
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+
+struct hb_font_t
+{
+  hb_object_header_t header;
+
+  hb_font_t *parent;
+  hb_face_t *face;
+
+  int32_t x_scale;
+  int32_t y_scale;
+  int64_t x_mult;
+  int64_t y_mult;
+
+  unsigned int x_ppem;
+  unsigned int y_ppem;
+
+  float ptem;
+
+  /* Font variation coordinates. */
+  unsigned int num_coords;
+  int *coords;
+  float *design_coords;
+
+  hb_font_funcs_t   *klass;
+  void              *user_data;
+  hb_destroy_func_t  destroy;
+
+  hb_shaper_object_dataset_t<hb_font_t> data; /* Various shaper data. */
+
+
+  /* Convert from font-space to user-space */
+  int64_t dir_mult (hb_direction_t direction)
+  { return HB_DIRECTION_IS_VERTICAL(direction) ? y_mult : x_mult; }
+  hb_position_t em_scale_x (int16_t v) { return em_mult (v, x_mult); }
+  hb_position_t em_scale_y (int16_t v) { return em_mult (v, y_mult); }
+  hb_position_t em_scalef_x (float v) { return em_scalef (v, x_scale); }
+  hb_position_t em_scalef_y (float v) { return em_scalef (v, y_scale); }
+  float em_fscale_x (int16_t v) { return em_fscale (v, x_scale); }
+  float em_fscale_y (int16_t v) { return em_fscale (v, y_scale); }
+  hb_position_t em_scale_dir (int16_t v, hb_direction_t direction)
+  { return em_mult (v, dir_mult (direction)); }
+
+  /* Convert from parent-font user-space to our user-space */
+  hb_position_t parent_scale_x_distance (hb_position_t v)
+  {
+    if (unlikely (parent && parent->x_scale != x_scale))
+      return (hb_position_t) (v * (int64_t) this->x_scale / this->parent->x_scale);
+    return v;
+  }
+  hb_position_t parent_scale_y_distance (hb_position_t v)
+  {
+    if (unlikely (parent && parent->y_scale != y_scale))
+      return (hb_position_t) (v * (int64_t) this->y_scale / this->parent->y_scale);
+    return v;
+  }
+  hb_position_t parent_scale_x_position (hb_position_t v)
+  { return parent_scale_x_distance (v); }
+  hb_position_t parent_scale_y_position (hb_position_t v)
+  { return parent_scale_y_distance (v); }
+
+  void parent_scale_distance (hb_position_t *x, hb_position_t *y)
+  {
+    *x = parent_scale_x_distance (*x);
+    *y = parent_scale_y_distance (*y);
+  }
+  void parent_scale_position (hb_position_t *x, hb_position_t *y)
+  {
+    *x = parent_scale_x_position (*x);
+    *y = parent_scale_y_position (*y);
+  }
+
+
+  /* Public getters */
+
+  HB_INTERNAL bool has_func (unsigned int i);
+  HB_INTERNAL bool has_func_set (unsigned int i);
+
+  /* has_* ... */
+#define HB_FONT_FUNC_IMPLEMENT(name) \
+  bool \
+  has_##name##_func () \
+  { \
+    hb_font_funcs_t *funcs = this->klass; \
+    unsigned int i = offsetof (hb_font_funcs_t::get_t::get_funcs_t, name) / sizeof (funcs->get.array[0]); \
+    return has_func (i); \
+  } \
+  bool \
+  has_##name##_func_set () \
+  { \
+    hb_font_funcs_t *funcs = this->klass; \
+    unsigned int i = offsetof (hb_font_funcs_t::get_t::get_funcs_t, name) / sizeof (funcs->get.array[0]); \
+    return has_func_set (i); \
+  }
+  HB_FONT_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_FONT_FUNC_IMPLEMENT
+
+  hb_bool_t get_font_h_extents (hb_font_extents_t *extents)
+  {
+    memset (extents, 0, sizeof (*extents));
+    return klass->get.f.font_h_extents (this, user_data,
+					extents,
+					klass->user_data.font_h_extents);
+  }
+  hb_bool_t get_font_v_extents (hb_font_extents_t *extents)
+  {
+    memset (extents, 0, sizeof (*extents));
+    return klass->get.f.font_v_extents (this, user_data,
+					extents,
+					klass->user_data.font_v_extents);
+  }
+
+  bool has_glyph (hb_codepoint_t unicode)
+  {
+    hb_codepoint_t glyph;
+    return get_nominal_glyph (unicode, &glyph);
+  }
+
+  hb_bool_t get_nominal_glyph (hb_codepoint_t unicode,
+			       hb_codepoint_t *glyph)
+  {
+    *glyph = 0;
+    return klass->get.f.nominal_glyph (this, user_data,
+				       unicode, glyph,
+				       klass->user_data.nominal_glyph);
+  }
+  unsigned int get_nominal_glyphs (unsigned int count,
+				   const hb_codepoint_t *first_unicode,
+				   unsigned int unicode_stride,
+				   hb_codepoint_t *first_glyph,
+				   unsigned int glyph_stride)
+  {
+    return klass->get.f.nominal_glyphs (this, user_data,
+					count,
+					first_unicode, unicode_stride,
+					first_glyph, glyph_stride,
+					klass->user_data.nominal_glyphs);
+  }
+
+  hb_bool_t get_variation_glyph (hb_codepoint_t unicode, hb_codepoint_t variation_selector,
+				 hb_codepoint_t *glyph)
+  {
+    *glyph = 0;
+    return klass->get.f.variation_glyph (this, user_data,
+					 unicode, variation_selector, glyph,
+					 klass->user_data.variation_glyph);
+  }
+
+  hb_position_t get_glyph_h_advance (hb_codepoint_t glyph)
+  {
+    return klass->get.f.glyph_h_advance (this, user_data,
+					 glyph,
+					 klass->user_data.glyph_h_advance);
+  }
+
+  hb_position_t get_glyph_v_advance (hb_codepoint_t glyph)
+  {
+    return klass->get.f.glyph_v_advance (this, user_data,
+					 glyph,
+					 klass->user_data.glyph_v_advance);
+  }
+
+  void get_glyph_h_advances (unsigned int count,
+			     const hb_codepoint_t *first_glyph,
+			     unsigned int glyph_stride,
+			     hb_position_t *first_advance,
+			     unsigned int advance_stride)
+  {
+    return klass->get.f.glyph_h_advances (this, user_data,
+					  count,
+					  first_glyph, glyph_stride,
+					  first_advance, advance_stride,
+					  klass->user_data.glyph_h_advances);
+  }
+
+  void get_glyph_v_advances (unsigned int count,
+			     const hb_codepoint_t *first_glyph,
+			     unsigned int glyph_stride,
+			     hb_position_t *first_advance,
+			     unsigned int advance_stride)
+  {
+    return klass->get.f.glyph_v_advances (this, user_data,
+					  count,
+					  first_glyph, glyph_stride,
+					  first_advance, advance_stride,
+					  klass->user_data.glyph_v_advances);
+  }
+
+  hb_bool_t get_glyph_h_origin (hb_codepoint_t glyph,
+				hb_position_t *x, hb_position_t *y)
+  {
+    *x = *y = 0;
+    return klass->get.f.glyph_h_origin (this, user_data,
+					glyph, x, y,
+					klass->user_data.glyph_h_origin);
+  }
+
+  hb_bool_t get_glyph_v_origin (hb_codepoint_t glyph,
+				hb_position_t *x, hb_position_t *y)
+  {
+    *x = *y = 0;
+    return klass->get.f.glyph_v_origin (this, user_data,
+					glyph, x, y,
+					klass->user_data.glyph_v_origin);
+  }
+
+  hb_position_t get_glyph_h_kerning (hb_codepoint_t left_glyph,
+				     hb_codepoint_t right_glyph)
+  {
+#ifdef HB_DISABLE_DEPRECATED
+    return 0;
+#else
+    return klass->get.f.glyph_h_kerning (this, user_data,
+					 left_glyph, right_glyph,
+					 klass->user_data.glyph_h_kerning);
+#endif
+  }
+
+  hb_position_t get_glyph_v_kerning (hb_codepoint_t top_glyph,
+				     hb_codepoint_t bottom_glyph)
+  {
+#ifdef HB_DISABLE_DEPRECATED
+    return 0;
+#else
+    return klass->get.f.glyph_v_kerning (this, user_data,
+					 top_glyph, bottom_glyph,
+					 klass->user_data.glyph_v_kerning);
+#endif
+  }
+
+  hb_bool_t get_glyph_extents (hb_codepoint_t glyph,
+			       hb_glyph_extents_t *extents)
+  {
+    memset (extents, 0, sizeof (*extents));
+    return klass->get.f.glyph_extents (this, user_data,
+				       glyph,
+				       extents,
+				       klass->user_data.glyph_extents);
+  }
+
+  hb_bool_t get_glyph_contour_point (hb_codepoint_t glyph, unsigned int point_index,
+				     hb_position_t *x, hb_position_t *y)
+  {
+    *x = *y = 0;
+    return klass->get.f.glyph_contour_point (this, user_data,
+					     glyph, point_index,
+					     x, y,
+					     klass->user_data.glyph_contour_point);
+  }
+
+  hb_bool_t get_glyph_name (hb_codepoint_t glyph,
+			    char *name, unsigned int size)
+  {
+    if (size) *name = '\0';
+    return klass->get.f.glyph_name (this, user_data,
+				    glyph,
+				    name, size,
+				    klass->user_data.glyph_name);
+  }
+
+  hb_bool_t get_glyph_from_name (const char *name, int len, /* -1 means nul-terminated */
+				 hb_codepoint_t *glyph)
+  {
+    *glyph = 0;
+    if (len == -1) len = strlen (name);
+    return klass->get.f.glyph_from_name (this, user_data,
+					 name, len,
+					 glyph,
+					 klass->user_data.glyph_from_name);
+  }
+
+
+  /* A bit higher-level, and with fallback */
+
+  void get_h_extents_with_fallback (hb_font_extents_t *extents)
+  {
+    if (!get_font_h_extents (extents))
+    {
+      extents->ascender = y_scale * .8;
+      extents->descender = extents->ascender - y_scale;
+      extents->line_gap = 0;
+    }
+  }
+  void get_v_extents_with_fallback (hb_font_extents_t *extents)
+  {
+    if (!get_font_v_extents (extents))
+    {
+      extents->ascender = x_scale / 2;
+      extents->descender = extents->ascender - x_scale;
+      extents->line_gap = 0;
+    }
+  }
+
+  void get_extents_for_direction (hb_direction_t direction,
+				  hb_font_extents_t *extents)
+  {
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (direction)))
+      get_h_extents_with_fallback (extents);
+    else
+      get_v_extents_with_fallback (extents);
+  }
+
+  void get_glyph_advance_for_direction (hb_codepoint_t glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y)
+  {
+    *x = *y = 0;
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (direction)))
+      *x = get_glyph_h_advance (glyph);
+    else
+      *y = get_glyph_v_advance (glyph);
+  }
+  void get_glyph_advances_for_direction (hb_direction_t direction,
+					 unsigned int count,
+					 const hb_codepoint_t *first_glyph,
+					 unsigned glyph_stride,
+					 hb_position_t *first_advance,
+					 unsigned advance_stride)
+  {
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (direction)))
+      get_glyph_h_advances (count, first_glyph, glyph_stride, first_advance, advance_stride);
+    else
+      get_glyph_v_advances (count, first_glyph, glyph_stride, first_advance, advance_stride);
+  }
+
+  void guess_v_origin_minus_h_origin (hb_codepoint_t glyph,
+				      hb_position_t *x, hb_position_t *y)
+  {
+    *x = get_glyph_h_advance (glyph) / 2;
+
+    /* TODO cache this somehow?! */
+    hb_font_extents_t extents;
+    get_h_extents_with_fallback (&extents);
+    *y = extents.ascender;
+  }
+
+  void get_glyph_h_origin_with_fallback (hb_codepoint_t glyph,
+					 hb_position_t *x, hb_position_t *y)
+  {
+    if (!get_glyph_h_origin (glyph, x, y) &&
+	 get_glyph_v_origin (glyph, x, y))
+    {
+      hb_position_t dx, dy;
+      guess_v_origin_minus_h_origin (glyph, &dx, &dy);
+      *x -= dx; *y -= dy;
+    }
+  }
+  void get_glyph_v_origin_with_fallback (hb_codepoint_t glyph,
+					 hb_position_t *x, hb_position_t *y)
+  {
+    if (!get_glyph_v_origin (glyph, x, y) &&
+	 get_glyph_h_origin (glyph, x, y))
+    {
+      hb_position_t dx, dy;
+      guess_v_origin_minus_h_origin (glyph, &dx, &dy);
+      *x += dx; *y += dy;
+    }
+  }
+
+  void get_glyph_origin_for_direction (hb_codepoint_t glyph,
+				       hb_direction_t direction,
+				       hb_position_t *x, hb_position_t *y)
+  {
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (direction)))
+      get_glyph_h_origin_with_fallback (glyph, x, y);
+    else
+      get_glyph_v_origin_with_fallback (glyph, x, y);
+  }
+
+  void add_glyph_h_origin (hb_codepoint_t glyph,
+			   hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_h_origin_with_fallback (glyph, &origin_x, &origin_y);
+
+    *x += origin_x;
+    *y += origin_y;
+  }
+  void add_glyph_v_origin (hb_codepoint_t glyph,
+			   hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_v_origin_with_fallback (glyph, &origin_x, &origin_y);
+
+    *x += origin_x;
+    *y += origin_y;
+  }
+  void add_glyph_origin_for_direction (hb_codepoint_t glyph,
+				       hb_direction_t direction,
+				       hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_origin_for_direction (glyph, direction, &origin_x, &origin_y);
+
+    *x += origin_x;
+    *y += origin_y;
+  }
+
+  void subtract_glyph_h_origin (hb_codepoint_t glyph,
+				hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_h_origin_with_fallback (glyph, &origin_x, &origin_y);
+
+    *x -= origin_x;
+    *y -= origin_y;
+  }
+  void subtract_glyph_v_origin (hb_codepoint_t glyph,
+				hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_v_origin_with_fallback (glyph, &origin_x, &origin_y);
+
+    *x -= origin_x;
+    *y -= origin_y;
+  }
+  void subtract_glyph_origin_for_direction (hb_codepoint_t glyph,
+					    hb_direction_t direction,
+					    hb_position_t *x, hb_position_t *y)
+  {
+    hb_position_t origin_x, origin_y;
+
+    get_glyph_origin_for_direction (glyph, direction, &origin_x, &origin_y);
+
+    *x -= origin_x;
+    *y -= origin_y;
+  }
+
+  void get_glyph_kerning_for_direction (hb_codepoint_t first_glyph, hb_codepoint_t second_glyph,
+					hb_direction_t direction,
+					hb_position_t *x, hb_position_t *y)
+  {
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (direction))) {
+      *y = 0;
+      *x = get_glyph_h_kerning (first_glyph, second_glyph);
+    } else {
+      *x = 0;
+      *y = get_glyph_v_kerning (first_glyph, second_glyph);
+    }
+  }
+
+  hb_bool_t get_glyph_extents_for_origin (hb_codepoint_t glyph,
+					  hb_direction_t direction,
+					  hb_glyph_extents_t *extents)
+  {
+    hb_bool_t ret = get_glyph_extents (glyph, extents);
+
+    if (ret)
+      subtract_glyph_origin_for_direction (glyph, direction, &extents->x_bearing, &extents->y_bearing);
+
+    return ret;
+  }
+
+  hb_bool_t get_glyph_contour_point_for_origin (hb_codepoint_t glyph, unsigned int point_index,
+						hb_direction_t direction,
+						hb_position_t *x, hb_position_t *y)
+  {
+    hb_bool_t ret = get_glyph_contour_point (glyph, point_index, x, y);
+
+    if (ret)
+      subtract_glyph_origin_for_direction (glyph, direction, x, y);
+
+    return ret;
+  }
+
+  /* Generates gidDDD if glyph has no name. */
+  void
+  glyph_to_string (hb_codepoint_t glyph,
+		   char *s, unsigned int size)
+  {
+    if (get_glyph_name (glyph, s, size)) return;
+
+    if (size && snprintf (s, size, "gid%u", glyph) < 0)
+      *s = '\0';
+  }
+
+  /* Parses gidDDD and uniUUUU strings automatically. */
+  hb_bool_t
+  glyph_from_string (const char *s, int len, /* -1 means nul-terminated */
+		     hb_codepoint_t *glyph)
+  {
+    if (get_glyph_from_name (s, len, glyph)) return true;
+
+    if (len == -1) len = strlen (s);
+
+    /* Straight glyph index. */
+    if (hb_codepoint_parse (s, len, 10, glyph))
+      return true;
+
+    if (len > 3)
+    {
+      /* gidDDD syntax for glyph indices. */
+      if (0 == strncmp (s, "gid", 3) &&
+	  hb_codepoint_parse (s + 3, len - 3, 10, glyph))
+	return true;
+
+      /* uniUUUU and other Unicode character indices. */
+      hb_codepoint_t unichar;
+      if (0 == strncmp (s, "uni", 3) &&
+	  hb_codepoint_parse (s + 3, len - 3, 16, &unichar) &&
+	  get_nominal_glyph (unichar, glyph))
+	return true;
+    }
+
+    return false;
+  }
+
+  void mults_changed ()
+  {
+    signed upem = face->get_upem ();
+    x_mult = ((int64_t) x_scale << 16) / upem;
+    y_mult = ((int64_t) y_scale << 16) / upem;
+  }
+
+  hb_position_t em_mult (int16_t v, int64_t mult)
+  {
+    return (hb_position_t) ((v * mult) >> 16);
+  }
+  hb_position_t em_scalef (float v, int scale)
+  { return (hb_position_t) roundf (v * scale / face->get_upem ()); }
+  float em_fscale (int16_t v, int scale)
+  { return (float) v * scale / face->get_upem (); }
+};
+DECLARE_NULL_INSTANCE (hb_font_t);
+
+
+#endif /* HB_FONT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ft.cc b/thirdparty/harfbuzz/src/hb-ft.cc
new file mode 100644
index 0000000000..2680873c27
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ft.cc
@@ -0,0 +1,1042 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2009  Keith Stribley
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_FREETYPE
+
+#include "hb-ft.h"
+
+#include "hb-font.hh"
+#include "hb-machinery.hh"
+#include "hb-cache.hh"
+
+#include FT_ADVANCES_H
+#include FT_MULTIPLE_MASTERS_H
+#include FT_TRUETYPE_TABLES_H
+
+
+/**
+ * SECTION:hb-ft
+ * @title: hb-ft
+ * @short_description: FreeType integration
+ * @include: hb-ft.h
+ *
+ * Functions for using HarfBuzz with the FreeType library.
+ *
+ * HarfBuzz supports using FreeType to provide face and
+ * font data.
+ *
+ * <note>Note that FreeType is not thread-safe, therefore these
+ * functions are not thread-safe either.</note>
+ **/
+
+
+/* TODO:
+ *
+ * In general, this file does a fine job of what it's supposed to do.
+ * There are, however, things that need more work:
+ *
+ *   - FreeType works in 26.6 mode.  Clients can decide to use that mode, and everything
+ *     would work fine.  However, we also abuse this API for performing in font-space,
+ *     but don't pass the correct flags to FreeType.  We just abuse the no-hinting mode
+ *     for that, such that no rounding etc happens.  As such, we don't set ppem, and
+ *     pass NO_HINTING as load_flags.  Would be much better to use NO_SCALE, and scale
+ *     ourselves.
+ *
+ *   - We don't handle / allow for emboldening / obliqueing.
+ *
+ *   - In the future, we should add constructors to create fonts in font space?
+ */
+
+
+struct hb_ft_font_t
+{
+  mutable hb_mutex_t lock;
+  FT_Face ft_face;
+  int load_flags;
+  bool symbol; /* Whether selected cmap is symbol cmap. */
+  bool unref; /* Whether to destroy ft_face when done. */
+
+  mutable hb_atomic_int_t cached_x_scale;
+  mutable hb_advance_cache_t advance_cache;
+};
+
+static hb_ft_font_t *
+_hb_ft_font_create (FT_Face ft_face, bool symbol, bool unref)
+{
+  hb_ft_font_t *ft_font = (hb_ft_font_t *) calloc (1, sizeof (hb_ft_font_t));
+  if (unlikely (!ft_font)) return nullptr;
+
+  ft_font->lock.init ();
+  ft_font->ft_face = ft_face;
+  ft_font->symbol = symbol;
+  ft_font->unref = unref;
+
+  ft_font->load_flags = FT_LOAD_DEFAULT | FT_LOAD_NO_HINTING;
+
+  ft_font->cached_x_scale.set_relaxed (0);
+  ft_font->advance_cache.init ();
+
+  return ft_font;
+}
+
+static void
+_hb_ft_face_destroy (void *data)
+{
+  FT_Done_Face ((FT_Face) data);
+}
+
+static void
+_hb_ft_font_destroy (void *data)
+{
+  hb_ft_font_t *ft_font = (hb_ft_font_t *) data;
+
+  ft_font->advance_cache.fini ();
+
+  if (ft_font->unref)
+    _hb_ft_face_destroy (ft_font->ft_face);
+
+  ft_font->lock.fini ();
+
+  free (ft_font);
+}
+
+/**
+ * hb_ft_font_set_load_flags:
+ * @font: #hb_font_t to work upon
+ * @load_flags: The FreeType load flags to set
+ *
+ * Sets the FT_Load_Glyph load flags for the specified #hb_font_t.
+ *
+ * For more information, see 
+ * https://www.freetype.org/freetype2/docs/reference/ft2-base_interface.html#ft_load_xxx
+ *
+ * Since: 1.0.5
+ **/
+void
+hb_ft_font_set_load_flags (hb_font_t *font, int load_flags)
+{
+  if (hb_object_is_immutable (font))
+    return;
+
+  if (unlikely (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy))
+    return;
+
+  hb_ft_font_t *ft_font = (hb_ft_font_t *) font->user_data;
+
+  ft_font->load_flags = load_flags;
+}
+
+/**
+ * hb_ft_font_get_load_flags:
+ * @font: #hb_font_t to work upon
+ *
+ * Fetches the FT_Load_Glyph load flags of the specified #hb_font_t.
+ *
+ * For more information, see 
+ * https://www.freetype.org/freetype2/docs/reference/ft2-base_interface.html#ft_load_xxx
+ *
+ * Return value: FT_Load_Glyph flags found
+ *
+ * Since: 1.0.5
+ **/
+int
+hb_ft_font_get_load_flags (hb_font_t *font)
+{
+  if (unlikely (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy))
+    return 0;
+
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font->user_data;
+
+  return ft_font->load_flags;
+}
+
+/**
+ * hb_ft_get_face:
+ * @font: #hb_font_t to work upon
+ *
+ * Fetches the FT_Face associated with the specified #hb_font_t
+ * font object.
+ *
+ * Return value: the FT_Face found
+ *
+ * Since: 0.9.2
+ **/
+FT_Face
+hb_ft_font_get_face (hb_font_t *font)
+{
+  if (unlikely (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy))
+    return nullptr;
+
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font->user_data;
+
+  return ft_font->ft_face;
+}
+
+/**
+ * hb_ft_font_lock_face:
+ * @font:
+ *
+ *
+ *
+ * Return value:
+ * Since: 2.6.5
+ **/
+FT_Face
+hb_ft_font_lock_face (hb_font_t *font)
+{
+  if (unlikely (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy))
+    return nullptr;
+
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font->user_data;
+
+  ft_font->lock.lock ();
+
+  return ft_font->ft_face;
+}
+
+/**
+ * hb_ft_font_unlock_face:
+ * @font:
+ *
+ *
+ *
+ * Return value:
+ * Since: 2.6.5
+ **/
+void
+hb_ft_font_unlock_face (hb_font_t *font)
+{
+  if (unlikely (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy))
+    return;
+
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font->user_data;
+
+  ft_font->lock.unlock ();
+}
+
+
+static hb_bool_t
+hb_ft_get_nominal_glyph (hb_font_t *font HB_UNUSED,
+			 void *font_data,
+			 hb_codepoint_t unicode,
+			 hb_codepoint_t *glyph,
+			 void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  unsigned int g = FT_Get_Char_Index (ft_font->ft_face, unicode);
+
+  if (unlikely (!g))
+  {
+    if (unlikely (ft_font->symbol) && unicode <= 0x00FFu)
+    {
+      /* For symbol-encoded OpenType fonts, we duplicate the
+       * U+F000..F0FF range at U+0000..U+00FF.  That's what
+       * Windows seems to do, and that's hinted about at:
+       * https://docs.microsoft.com/en-us/typography/opentype/spec/recom
+       * under "Non-Standard (Symbol) Fonts". */
+      g = FT_Get_Char_Index (ft_font->ft_face, 0xF000u + unicode);
+      if (!g)
+	return false;
+    }
+    else
+      return false;
+  }
+
+  *glyph = g;
+  return true;
+}
+
+static unsigned int
+hb_ft_get_nominal_glyphs (hb_font_t *font HB_UNUSED,
+			  void *font_data,
+			  unsigned int count,
+			  const hb_codepoint_t *first_unicode,
+			  unsigned int unicode_stride,
+			  hb_codepoint_t *first_glyph,
+			  unsigned int glyph_stride,
+			  void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  unsigned int done;
+  for (done = 0;
+       done < count && (*first_glyph = FT_Get_Char_Index (ft_font->ft_face, *first_unicode));
+       done++)
+  {
+    first_unicode = &StructAtOffsetUnaligned<hb_codepoint_t> (first_unicode, unicode_stride);
+    first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+  }
+  /* We don't need to do ft_font->symbol dance here, since HB calls the singular
+   * nominal_glyph() for what we don't handle here. */
+  return done;
+}
+
+
+static hb_bool_t
+hb_ft_get_variation_glyph (hb_font_t *font HB_UNUSED,
+			   void *font_data,
+			   hb_codepoint_t unicode,
+			   hb_codepoint_t variation_selector,
+			   hb_codepoint_t *glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  unsigned int g = FT_Face_GetCharVariantIndex (ft_font->ft_face, unicode, variation_selector);
+
+  if (unlikely (!g))
+    return false;
+
+  *glyph = g;
+  return true;
+}
+
+static void
+hb_ft_get_glyph_h_advances (hb_font_t* font, void* font_data,
+			    unsigned count,
+			    const hb_codepoint_t *first_glyph,
+			    unsigned glyph_stride,
+			    hb_position_t *first_advance,
+			    unsigned advance_stride,
+			    void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+  int load_flags = ft_font->load_flags;
+  int mult = font->x_scale < 0 ? -1 : +1;
+
+  if (font->x_scale != ft_font->cached_x_scale.get ())
+  {
+    ft_font->advance_cache.clear ();
+    ft_font->cached_x_scale.set (font->x_scale);
+  }
+
+  for (unsigned int i = 0; i < count; i++)
+  {
+    FT_Fixed v = 0;
+    hb_codepoint_t glyph = *first_glyph;
+
+    unsigned int cv;
+    if (ft_font->advance_cache.get (glyph, &cv))
+      v = cv;
+    else
+    {
+      FT_Get_Advance (ft_face, glyph, load_flags, &v);
+      ft_font->advance_cache.set (glyph, v);
+    }
+
+    *first_advance = (v * mult + (1<<9)) >> 10;
+    first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+    first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+  }
+}
+
+static hb_position_t
+hb_ft_get_glyph_v_advance (hb_font_t *font,
+			   void *font_data,
+			   hb_codepoint_t glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Fixed v;
+
+  if (unlikely (FT_Get_Advance (ft_font->ft_face, glyph, ft_font->load_flags | FT_LOAD_VERTICAL_LAYOUT, &v)))
+    return 0;
+
+  if (font->y_scale < 0)
+    v = -v;
+
+  /* Note: FreeType's vertical metrics grows downward while other FreeType coordinates
+   * have a Y growing upward.  Hence the extra negation. */
+  return (-v + (1<<9)) >> 10;
+}
+
+static hb_bool_t
+hb_ft_get_glyph_v_origin (hb_font_t *font,
+			  void *font_data,
+			  hb_codepoint_t glyph,
+			  hb_position_t *x,
+			  hb_position_t *y,
+			  void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+
+  if (unlikely (FT_Load_Glyph (ft_face, glyph, ft_font->load_flags)))
+    return false;
+
+  /* Note: FreeType's vertical metrics grows downward while other FreeType coordinates
+   * have a Y growing upward.  Hence the extra negation. */
+  *x = ft_face->glyph->metrics.horiBearingX -   ft_face->glyph->metrics.vertBearingX;
+  *y = ft_face->glyph->metrics.horiBearingY - (-ft_face->glyph->metrics.vertBearingY);
+
+  if (font->x_scale < 0)
+    *x = -*x;
+  if (font->y_scale < 0)
+    *y = -*y;
+
+  return true;
+}
+
+#ifndef HB_NO_OT_SHAPE_FALLBACK
+static hb_position_t
+hb_ft_get_glyph_h_kerning (hb_font_t *font,
+			   void *font_data,
+			   hb_codepoint_t left_glyph,
+			   hb_codepoint_t right_glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  FT_Vector kerningv;
+
+  FT_Kerning_Mode mode = font->x_ppem ? FT_KERNING_DEFAULT : FT_KERNING_UNFITTED;
+  if (FT_Get_Kerning (ft_font->ft_face, left_glyph, right_glyph, mode, &kerningv))
+    return 0;
+
+  return kerningv.x;
+}
+#endif
+
+static hb_bool_t
+hb_ft_get_glyph_extents (hb_font_t *font,
+			 void *font_data,
+			 hb_codepoint_t glyph,
+			 hb_glyph_extents_t *extents,
+			 void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+
+  if (unlikely (FT_Load_Glyph (ft_face, glyph, ft_font->load_flags)))
+    return false;
+
+  extents->x_bearing = ft_face->glyph->metrics.horiBearingX;
+  extents->y_bearing = ft_face->glyph->metrics.horiBearingY;
+  extents->width = ft_face->glyph->metrics.width;
+  extents->height = -ft_face->glyph->metrics.height;
+  if (font->x_scale < 0)
+  {
+    extents->x_bearing = -extents->x_bearing;
+    extents->width = -extents->width;
+  }
+  if (font->y_scale < 0)
+  {
+    extents->y_bearing = -extents->y_bearing;
+    extents->height = -extents->height;
+  }
+  return true;
+}
+
+static hb_bool_t
+hb_ft_get_glyph_contour_point (hb_font_t *font HB_UNUSED,
+			       void *font_data,
+			       hb_codepoint_t glyph,
+			       unsigned int point_index,
+			       hb_position_t *x,
+			       hb_position_t *y,
+			       void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+
+  if (unlikely (FT_Load_Glyph (ft_face, glyph, ft_font->load_flags)))
+      return false;
+
+  if (unlikely (ft_face->glyph->format != FT_GLYPH_FORMAT_OUTLINE))
+      return false;
+
+  if (unlikely (point_index >= (unsigned int) ft_face->glyph->outline.n_points))
+      return false;
+
+  *x = ft_face->glyph->outline.points[point_index].x;
+  *y = ft_face->glyph->outline.points[point_index].y;
+
+  return true;
+}
+
+static hb_bool_t
+hb_ft_get_glyph_name (hb_font_t *font HB_UNUSED,
+		      void *font_data,
+		      hb_codepoint_t glyph,
+		      char *name, unsigned int size,
+		      void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+
+  hb_bool_t ret = !FT_Get_Glyph_Name (ft_face, glyph, name, size);
+  if (ret && (size && !*name))
+    ret = false;
+
+  return ret;
+}
+
+static hb_bool_t
+hb_ft_get_glyph_from_name (hb_font_t *font HB_UNUSED,
+			   void *font_data,
+			   const char *name, int len, /* -1 means nul-terminated */
+			   hb_codepoint_t *glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+
+  if (len < 0)
+    *glyph = FT_Get_Name_Index (ft_face, (FT_String *) name);
+  else {
+    /* Make a nul-terminated version. */
+    char buf[128];
+    len = hb_min (len, (int) sizeof (buf) - 1);
+    strncpy (buf, name, len);
+    buf[len] = '\0';
+    *glyph = FT_Get_Name_Index (ft_face, buf);
+  }
+
+  if (*glyph == 0)
+  {
+    /* Check whether the given name was actually the name of glyph 0. */
+    char buf[128];
+    if (!FT_Get_Glyph_Name(ft_face, 0, buf, sizeof (buf)) &&
+	len < 0 ? !strcmp (buf, name) : !strncmp (buf, name, len))
+      return true;
+  }
+
+  return *glyph != 0;
+}
+
+static hb_bool_t
+hb_ft_get_font_h_extents (hb_font_t *font HB_UNUSED,
+			  void *font_data,
+			  hb_font_extents_t *metrics,
+			  void *user_data HB_UNUSED)
+{
+  const hb_ft_font_t *ft_font = (const hb_ft_font_t *) font_data;
+  hb_lock_t lock (ft_font->lock);
+  FT_Face ft_face = ft_font->ft_face;
+  metrics->ascender = FT_MulFix(ft_face->ascender, ft_face->size->metrics.y_scale);
+  metrics->descender = FT_MulFix(ft_face->descender, ft_face->size->metrics.y_scale);
+  metrics->line_gap = FT_MulFix( ft_face->height, ft_face->size->metrics.y_scale ) - (metrics->ascender - metrics->descender);
+  if (font->y_scale < 0)
+  {
+    metrics->ascender = -metrics->ascender;
+    metrics->descender = -metrics->descender;
+    metrics->line_gap = -metrics->line_gap;
+  }
+  return true;
+}
+
+#if HB_USE_ATEXIT
+static void free_static_ft_funcs ();
+#endif
+
+static struct hb_ft_font_funcs_lazy_loader_t : hb_font_funcs_lazy_loader_t<hb_ft_font_funcs_lazy_loader_t>
+{
+  static hb_font_funcs_t *create ()
+  {
+    hb_font_funcs_t *funcs = hb_font_funcs_create ();
+
+    hb_font_funcs_set_font_h_extents_func (funcs, hb_ft_get_font_h_extents, nullptr, nullptr);
+    //hb_font_funcs_set_font_v_extents_func (funcs, hb_ft_get_font_v_extents, nullptr, nullptr);
+    hb_font_funcs_set_nominal_glyph_func (funcs, hb_ft_get_nominal_glyph, nullptr, nullptr);
+    hb_font_funcs_set_nominal_glyphs_func (funcs, hb_ft_get_nominal_glyphs, nullptr, nullptr);
+    hb_font_funcs_set_variation_glyph_func (funcs, hb_ft_get_variation_glyph, nullptr, nullptr);
+    hb_font_funcs_set_glyph_h_advances_func (funcs, hb_ft_get_glyph_h_advances, nullptr, nullptr);
+    hb_font_funcs_set_glyph_v_advance_func (funcs, hb_ft_get_glyph_v_advance, nullptr, nullptr);
+    //hb_font_funcs_set_glyph_h_origin_func (funcs, hb_ft_get_glyph_h_origin, nullptr, nullptr);
+    hb_font_funcs_set_glyph_v_origin_func (funcs, hb_ft_get_glyph_v_origin, nullptr, nullptr);
+#ifndef HB_NO_OT_SHAPE_FALLBACK
+    hb_font_funcs_set_glyph_h_kerning_func (funcs, hb_ft_get_glyph_h_kerning, nullptr, nullptr);
+#endif
+    //hb_font_funcs_set_glyph_v_kerning_func (funcs, hb_ft_get_glyph_v_kerning, nullptr, nullptr);
+    hb_font_funcs_set_glyph_extents_func (funcs, hb_ft_get_glyph_extents, nullptr, nullptr);
+    hb_font_funcs_set_glyph_contour_point_func (funcs, hb_ft_get_glyph_contour_point, nullptr, nullptr);
+    hb_font_funcs_set_glyph_name_func (funcs, hb_ft_get_glyph_name, nullptr, nullptr);
+    hb_font_funcs_set_glyph_from_name_func (funcs, hb_ft_get_glyph_from_name, nullptr, nullptr);
+
+    hb_font_funcs_make_immutable (funcs);
+
+#if HB_USE_ATEXIT
+    atexit (free_static_ft_funcs);
+#endif
+
+    return funcs;
+  }
+} static_ft_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_ft_funcs ()
+{
+  static_ft_funcs.free_instance ();
+}
+#endif
+
+static hb_font_funcs_t *
+_hb_ft_get_font_funcs ()
+{
+  return static_ft_funcs.get_unconst ();
+}
+
+static void
+_hb_ft_font_set_funcs (hb_font_t *font, FT_Face ft_face, bool unref)
+{
+  bool symbol = ft_face->charmap && ft_face->charmap->encoding == FT_ENCODING_MS_SYMBOL;
+
+  hb_ft_font_t *ft_font = _hb_ft_font_create (ft_face, symbol, unref);
+  if (unlikely (!ft_font)) return;
+
+  hb_font_set_funcs (font,
+		     _hb_ft_get_font_funcs (),
+		     ft_font,
+		     _hb_ft_font_destroy);
+}
+
+
+static hb_blob_t *
+_hb_ft_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  FT_Face ft_face = (FT_Face) user_data;
+  FT_Byte *buffer;
+  FT_ULong  length = 0;
+  FT_Error error;
+
+  /* Note: FreeType like HarfBuzz uses the NONE tag for fetching the entire blob */
+
+  error = FT_Load_Sfnt_Table (ft_face, tag, 0, nullptr, &length);
+  if (error)
+    return nullptr;
+
+  buffer = (FT_Byte *) malloc (length);
+  if (!buffer)
+    return nullptr;
+
+  error = FT_Load_Sfnt_Table (ft_face, tag, 0, buffer, &length);
+  if (error)
+  {
+    free (buffer);
+    return nullptr;
+  }
+
+  return hb_blob_create ((const char *) buffer, length,
+			 HB_MEMORY_MODE_WRITABLE,
+			 buffer, free);
+}
+
+/**
+ * hb_ft_face_create:
+ * @ft_face: (destroy destroy) (scope notified): FT_Face to work upon
+ * @destroy: A callback to call when the face object is not needed anymore
+ *
+ * Creates an #hb_face_t face object from the specified FT_Face.
+ *
+ * This variant of the function does not provide any life-cycle management.
+ *
+ * Most client programs should use hb_ft_face_create_referenced()
+ * (or, perhaps, hb_ft_face_create_cached()) instead. 
+ *
+ * If you know you have valid reasons not to use hb_ft_face_create_referenced(),
+ * then it is the client program's responsibility to destroy @ft_face 
+ * after the #hb_face_t face object has been destroyed.
+ *
+ * Return value: (transfer full): the new #hb_face_t face object
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_ft_face_create (FT_Face           ft_face,
+		   hb_destroy_func_t destroy)
+{
+  hb_face_t *face;
+
+  if (!ft_face->stream->read) {
+    hb_blob_t *blob;
+
+    blob = hb_blob_create ((const char *) ft_face->stream->base,
+			   (unsigned int) ft_face->stream->size,
+			   HB_MEMORY_MODE_READONLY,
+			   ft_face, destroy);
+    face = hb_face_create (blob, ft_face->face_index);
+    hb_blob_destroy (blob);
+  } else {
+    face = hb_face_create_for_tables (_hb_ft_reference_table, ft_face, destroy);
+  }
+
+  hb_face_set_index (face, ft_face->face_index);
+  hb_face_set_upem (face, ft_face->units_per_EM);
+
+  return face;
+}
+
+/**
+ * hb_ft_face_create_referenced:
+ * @ft_face: FT_Face to work upon
+ *
+ * Creates an #hb_face_t face object from the specified FT_Face.
+ *
+ * This is the preferred variant of the hb_ft_face_create*
+ * function family, because it calls FT_Reference_Face() on @ft_face,
+ * ensuring that @ft_face remains alive as long as the resulting
+ * #hb_face_t face object remains alive. Also calls FT_Done_Face()
+ * when the #hb_face_t face object is destroyed.
+ *
+ * Use this version unless you know you have good reasons not to.
+ *
+ * Return value: (transfer full): the new #hb_face_t face object
+ *
+ * Since: 0.9.38
+ **/
+hb_face_t *
+hb_ft_face_create_referenced (FT_Face ft_face)
+{
+  FT_Reference_Face (ft_face);
+  return hb_ft_face_create (ft_face, _hb_ft_face_destroy);
+}
+
+static void
+hb_ft_face_finalize (FT_Face ft_face)
+{
+  hb_face_destroy ((hb_face_t *) ft_face->generic.data);
+}
+
+/**
+ * hb_ft_face_create_cached:
+ * @ft_face: FT_Face to work upon
+ *
+ * Creates an #hb_face_t face object from the specified FT_Face.
+ *
+ * This variant of the function caches the newly created #hb_face_t
+ * face object, using the @generic pointer of @ft_face. Subsequent function
+ * calls that are passed the same @ft_face parameter will have the same
+ * #hb_face_t returned to them, and that #hb_face_t will be correctly
+ * reference counted.
+ *
+ * However, client programs are still responsible for destroying
+ * @ft_face after the last #hb_face_t face object has been destroyed.
+ *
+ * Return value: (transfer full): the new #hb_face_t face object
+ *
+ * Since: 0.9.2
+ **/
+hb_face_t *
+hb_ft_face_create_cached (FT_Face ft_face)
+{
+  if (unlikely (!ft_face->generic.data || ft_face->generic.finalizer != (FT_Generic_Finalizer) hb_ft_face_finalize))
+  {
+    if (ft_face->generic.finalizer)
+      ft_face->generic.finalizer (ft_face);
+
+    ft_face->generic.data = hb_ft_face_create (ft_face, nullptr);
+    ft_face->generic.finalizer = (FT_Generic_Finalizer) hb_ft_face_finalize;
+  }
+
+  return hb_face_reference ((hb_face_t *) ft_face->generic.data);
+}
+
+/**
+ * hb_ft_font_create:
+ * @ft_face: (destroy destroy) (scope notified): FT_Face to work upon
+ * @destroy: (optional): A callback to call when the font object is not needed anymore
+ *
+ * Creates an #hb_font_t font object from the specified FT_Face.
+ *
+ * <note>Note: You must set the face size on @ft_face before calling
+ * hb_ft_font_create() on it. Otherwise, HarfBuzz will not pick up
+ * the face size.</note>
+ *
+ * This variant of the function does not provide any life-cycle management.
+ *
+ * Most client programs should use hb_ft_font_create_referenced()
+ * instead. 
+ *
+ * If you know you have valid reasons not to use hb_ft_font_create_referenced(),
+ * then it is the client program's responsibility to destroy @ft_face 
+ * after the #hb_font_t font object has been destroyed.
+ *
+ * HarfBuzz will use the @destroy callback on the #hb_font_t font object 
+ * if it is supplied when you use this function. However, even if @destroy
+ * is provided, it is the client program's responsibility to destroy @ft_face,
+ * and it is the client program's responsibility to ensure that @ft_face is
+ * destroyed only after the #hb_font_t font object has been destroyed.
+ *
+ * Return value: (transfer full): the new #hb_font_t font object
+ *
+ * Since: 0.9.2
+ **/
+hb_font_t *
+hb_ft_font_create (FT_Face           ft_face,
+		   hb_destroy_func_t destroy)
+{
+  hb_font_t *font;
+  hb_face_t *face;
+
+  face = hb_ft_face_create (ft_face, destroy);
+  font = hb_font_create (face);
+  hb_face_destroy (face);
+  _hb_ft_font_set_funcs (font, ft_face, false);
+  hb_ft_font_changed (font);
+  return font;
+}
+
+/**
+ * hb_ft_font_has_changed:
+ * @font: #hb_font_t to work upon
+ *
+ * Refreshes the state of @font when the underlying FT_Face has changed.
+ * This function should be called after changing the size or
+ * variation-axis settings on the FT_Face.
+ *
+ * Since: 1.0.5
+ **/
+void
+hb_ft_font_changed (hb_font_t *font)
+{
+  if (font->destroy != (hb_destroy_func_t) _hb_ft_font_destroy)
+    return;
+
+  hb_ft_font_t *ft_font = (hb_ft_font_t *) font->user_data;
+
+  FT_Face ft_face = ft_font->ft_face;
+
+  hb_font_set_scale (font,
+		     (int) (((uint64_t) ft_face->size->metrics.x_scale * (uint64_t) ft_face->units_per_EM + (1u<<15)) >> 16),
+		     (int) (((uint64_t) ft_face->size->metrics.y_scale * (uint64_t) ft_face->units_per_EM + (1u<<15)) >> 16));
+#if 0 /* hb-ft works in no-hinting model */
+  hb_font_set_ppem (font,
+		    ft_face->size->metrics.x_ppem,
+		    ft_face->size->metrics.y_ppem);
+#endif
+
+#if defined(HAVE_FT_GET_VAR_BLEND_COORDINATES) && !defined(HB_NO_VAR)
+  FT_MM_Var *mm_var = nullptr;
+  if (!FT_Get_MM_Var (ft_face, &mm_var))
+  {
+    FT_Fixed *ft_coords = (FT_Fixed *) calloc (mm_var->num_axis, sizeof (FT_Fixed));
+    int *coords = (int *) calloc (mm_var->num_axis, sizeof (int));
+    if (coords && ft_coords)
+    {
+      if (!FT_Get_Var_Blend_Coordinates (ft_face, mm_var->num_axis, ft_coords))
+      {
+	bool nonzero = false;
+
+	for (unsigned int i = 0; i < mm_var->num_axis; ++i)
+	 {
+	  coords[i] = ft_coords[i] >>= 2;
+	  nonzero = nonzero || coords[i];
+	 }
+
+	if (nonzero)
+	  hb_font_set_var_coords_normalized (font, coords, mm_var->num_axis);
+	else
+	  hb_font_set_var_coords_normalized (font, nullptr, 0);
+      }
+    }
+    free (coords);
+    free (ft_coords);
+#ifdef HAVE_FT_DONE_MM_VAR
+    FT_Done_MM_Var (ft_face->glyph->library, mm_var);
+#else
+    free (mm_var);
+#endif
+  }
+#endif
+}
+
+/**
+ * hb_ft_font_create_referenced:
+ * @ft_face: FT_Face to work upon
+ *
+ * Creates an #hb_font_t font object from the specified FT_Face.
+ *
+ * <note>Note: You must set the face size on @ft_face before calling
+ * hb_ft_font_create_references() on it. Otherwise, HarfBuzz will not pick up
+ * the face size.</note>
+ *
+ * This is the preferred variant of the hb_ft_font_create*
+ * function family, because it calls FT_Reference_Face() on @ft_face,
+ * ensuring that @ft_face remains alive as long as the resulting
+ * #hb_font_t font object remains alive.
+ *
+ * Use this version unless you know you have good reasons not to.
+ *
+ * Return value: (transfer full): the new #hb_font_t font object
+ *
+ * Since: 0.9.38
+ **/
+hb_font_t *
+hb_ft_font_create_referenced (FT_Face ft_face)
+{
+  FT_Reference_Face (ft_face);
+  return hb_ft_font_create (ft_face, _hb_ft_face_destroy);
+}
+
+#if HB_USE_ATEXIT
+static void free_static_ft_library ();
+#endif
+
+static struct hb_ft_library_lazy_loader_t : hb_lazy_loader_t<hb_remove_pointer<FT_Library>,
+							     hb_ft_library_lazy_loader_t>
+{
+  static FT_Library create ()
+  {
+    FT_Library l;
+    if (FT_Init_FreeType (&l))
+      return nullptr;
+
+#if HB_USE_ATEXIT
+    atexit (free_static_ft_library);
+#endif
+
+    return l;
+  }
+  static void destroy (FT_Library l)
+  {
+    FT_Done_FreeType (l);
+  }
+  static FT_Library get_null ()
+  {
+    return nullptr;
+  }
+} static_ft_library;
+
+#if HB_USE_ATEXIT
+static
+void free_static_ft_library ()
+{
+  static_ft_library.free_instance ();
+}
+#endif
+
+static FT_Library
+get_ft_library ()
+{
+  return static_ft_library.get_unconst ();
+}
+
+static void
+_release_blob (FT_Face ft_face)
+{
+  hb_blob_destroy ((hb_blob_t *) ft_face->generic.data);
+}
+
+/**
+ * hb_ft_font_set_funcs:
+ * @font: #hb_font_t to work upon
+ *
+ * Configures the font-functions structure of the specified
+ * #hb_font_t font object to use FreeType font functions.
+ *
+ * In particular, you can use this function to configure an
+ * existing #hb_face_t face object for use with FreeType font
+ * functions even if that #hb_face_t face object was initially
+ * created with hb_face_create(), and therefore was not
+ * initially configured to use FreeType font functions.
+ *
+ * An #hb_face_t face object created with hb_ft_face_create()
+ * is preconfigured for FreeType font functions and does not
+ * require this function to be used.
+ *
+ * <note>Note: Internally, this function creates an FT_Face.
+* </note>
+ *
+ * Since: 1.0.5
+ **/
+void
+hb_ft_font_set_funcs (hb_font_t *font)
+{
+  hb_blob_t *blob = hb_face_reference_blob (font->face);
+  unsigned int blob_length;
+  const char *blob_data = hb_blob_get_data (blob, &blob_length);
+  if (unlikely (!blob_length))
+    DEBUG_MSG (FT, font, "Font face has empty blob");
+
+  FT_Face ft_face = nullptr;
+  FT_Error err = FT_New_Memory_Face (get_ft_library (),
+				     (const FT_Byte *) blob_data,
+				     blob_length,
+				     hb_face_get_index (font->face),
+				     &ft_face);
+
+  if (unlikely (err)) {
+    hb_blob_destroy (blob);
+    DEBUG_MSG (FT, font, "Font face FT_New_Memory_Face() failed");
+    return;
+  }
+
+  if (FT_Select_Charmap (ft_face, FT_ENCODING_MS_SYMBOL))
+    FT_Select_Charmap (ft_face, FT_ENCODING_UNICODE);
+
+  FT_Set_Char_Size (ft_face,
+		    abs (font->x_scale), abs (font->y_scale),
+		    0, 0);
+#if 0
+		    font->x_ppem * 72 * 64 / font->x_scale,
+		    font->y_ppem * 72 * 64 / font->y_scale);
+#endif
+  if (font->x_scale < 0 || font->y_scale < 0)
+  {
+    FT_Matrix matrix = { font->x_scale < 0 ? -1 : +1, 0,
+			  0, font->y_scale < 0 ? -1 : +1};
+    FT_Set_Transform (ft_face, &matrix, nullptr);
+  }
+
+#if defined(HAVE_FT_GET_VAR_BLEND_COORDINATES) && !defined(HB_NO_VAR)
+  unsigned int num_coords;
+  const int *coords = hb_font_get_var_coords_normalized (font, &num_coords);
+  if (num_coords)
+  {
+    FT_Fixed *ft_coords = (FT_Fixed *) calloc (num_coords, sizeof (FT_Fixed));
+    if (ft_coords)
+    {
+      for (unsigned int i = 0; i < num_coords; i++)
+	ft_coords[i] = coords[i] * 4;
+      FT_Set_Var_Blend_Coordinates (ft_face, num_coords, ft_coords);
+      free (ft_coords);
+    }
+  }
+#endif
+
+  ft_face->generic.data = blob;
+  ft_face->generic.finalizer = (FT_Generic_Finalizer) _release_blob;
+
+  _hb_ft_font_set_funcs (font, ft_face, true);
+  hb_ft_font_set_load_flags (font, FT_LOAD_DEFAULT | FT_LOAD_NO_HINTING);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ft.h b/thirdparty/harfbuzz/src/hb-ft.h
new file mode 100644
index 0000000000..bf07115ab9
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ft.h
@@ -0,0 +1,138 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_FT_H
+#define HB_FT_H
+
+#include "hb.h"
+
+#include <ft2build.h>
+#include FT_FREETYPE_H
+
+HB_BEGIN_DECLS
+
+/*
+ * Note: FreeType is not thread-safe.
+ * Hence, these functions are not either.
+ */
+
+/*
+ * hb-face from ft-face.
+ */
+
+/* This one creates a new hb-face for given ft-face.
+ * When the returned hb-face is destroyed, the destroy
+ * callback is called (if not NULL), with the ft-face passed
+ * to it.
+ *
+ * The client is responsible to make sure that ft-face is
+ * destroyed after hb-face is destroyed.
+ *
+ * Most often you don't want this function.  You should use either
+ * hb_ft_face_create_cached(), or hb_ft_face_create_referenced().
+ * In particular, if you are going to pass NULL as destroy, you
+ * probably should use (the more recent) hb_ft_face_create_referenced()
+ * instead.
+ */
+HB_EXTERN hb_face_t *
+hb_ft_face_create (FT_Face           ft_face,
+		   hb_destroy_func_t destroy);
+
+/* This version is like hb_ft_face_create(), except that it caches
+ * the hb-face using the generic pointer of the ft-face.  This means
+ * that subsequent calls to this function with the same ft-face will
+ * return the same hb-face (correctly referenced).
+ *
+ * Client is still responsible for making sure that ft-face is destroyed
+ * after hb-face is.
+ */
+HB_EXTERN hb_face_t *
+hb_ft_face_create_cached (FT_Face ft_face);
+
+/* This version is like hb_ft_face_create(), except that it calls
+ * FT_Reference_Face() on ft-face, as such keeping ft-face alive
+ * as long as the hb-face is.
+ *
+ * This is the most convenient version to use.  Use it unless you have
+ * very good reasons not to.
+ */
+HB_EXTERN hb_face_t *
+hb_ft_face_create_referenced (FT_Face ft_face);
+
+
+/*
+ * hb-font from ft-face.
+ */
+
+/*
+ * Note:
+ *
+ * Set face size on ft-face before creating hb-font from it.
+ * Otherwise hb-ft would NOT pick up the font size correctly.
+ */
+
+/* See notes on hb_ft_face_create().  Same issues re lifecycle-management
+ * apply here.  Use hb_ft_font_create_referenced() if you can. */
+HB_EXTERN hb_font_t *
+hb_ft_font_create (FT_Face           ft_face,
+		   hb_destroy_func_t destroy);
+
+/* See notes on hb_ft_face_create_referenced() re lifecycle-management
+ * issues. */
+HB_EXTERN hb_font_t *
+hb_ft_font_create_referenced (FT_Face ft_face);
+
+HB_EXTERN FT_Face
+hb_ft_font_get_face (hb_font_t *font);
+
+HB_EXTERN FT_Face
+hb_ft_font_lock_face (hb_font_t *font);
+
+HB_EXTERN void
+hb_ft_font_unlock_face (hb_font_t *font);
+
+HB_EXTERN void
+hb_ft_font_set_load_flags (hb_font_t *font, int load_flags);
+
+HB_EXTERN int
+hb_ft_font_get_load_flags (hb_font_t *font);
+
+/* Call when size or variations settings on underlying FT_Face change. */
+HB_EXTERN void
+hb_ft_font_changed (hb_font_t *font);
+
+/* Makes an hb_font_t use FreeType internally to implement font functions.
+ * Note: this internally creates an FT_Face.  Use it when you create your
+ * hb_face_t using hb_face_create(). */
+HB_EXTERN void
+hb_ft_font_set_funcs (hb_font_t *font);
+
+
+HB_END_DECLS
+
+#endif /* HB_FT_H */
diff --git a/thirdparty/harfbuzz/src/hb-gdi.cc b/thirdparty/harfbuzz/src/hb-gdi.cc
new file mode 100644
index 0000000000..f6306ef89f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-gdi.cc
@@ -0,0 +1,73 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_GDI
+
+#include "hb-gdi.h"
+
+static hb_blob_t *
+_hb_gdi_reference_table (hb_face_t *face HB_UNUSED, hb_tag_t tag, void *user_data)
+{
+  char *buffer = nullptr;
+  DWORD length = 0;
+
+  HDC hdc = GetDC (nullptr);
+  if (unlikely (!SelectObject (hdc, (HFONT) user_data))) goto fail;
+
+  length = GetFontData (hdc, hb_uint32_swap (tag), 0, buffer, length);
+  if (unlikely (length == GDI_ERROR)) goto fail_with_releasedc;
+
+  buffer = (char *) malloc (length);
+  if (unlikely (!buffer)) goto fail_with_releasedc;
+  length = GetFontData (hdc, hb_uint32_swap (tag), 0, buffer, length);
+  if (unlikely (length == GDI_ERROR)) goto fail_with_releasedc_and_free;
+  ReleaseDC (nullptr, hdc);
+
+  return hb_blob_create ((const char *) buffer, length, HB_MEMORY_MODE_WRITABLE, buffer, free);
+
+fail_with_releasedc_and_free:
+  free (buffer);
+fail_with_releasedc:
+  ReleaseDC (nullptr, hdc);
+fail:
+  return hb_blob_get_empty ();
+}
+
+/**
+ * hb_gdi_face_create:
+ * @hfont: a HFONT object.
+ *
+ * Return value: #hb_face_t object corresponding to the given input
+ *
+ * Since: 2.6.0
+ **/
+hb_face_t *
+hb_gdi_face_create (HFONT hfont)
+{
+  return hb_face_create_for_tables (_hb_gdi_reference_table, (void *) hfont, nullptr);
+}
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-gdi.h b/thirdparty/harfbuzz/src/hb-gdi.h
new file mode 100644
index 0000000000..68cc43917e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-gdi.h
@@ -0,0 +1,39 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_GDI_H
+#define HB_GDI_H
+
+#include "hb.h"
+
+#include <windows.h>
+
+HB_BEGIN_DECLS
+
+HB_EXTERN hb_face_t *
+hb_gdi_face_create (HFONT hfont);
+
+HB_END_DECLS
+
+#endif /* HB_GDI_H */
diff --git a/thirdparty/harfbuzz/src/hb-glib.cc b/thirdparty/harfbuzz/src/hb-glib.cc
new file mode 100644
index 0000000000..f93bb8853c
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-glib.cc
@@ -0,0 +1,307 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_GLIB
+
+#include "hb-glib.h"
+
+#include "hb-machinery.hh"
+
+
+/**
+ * SECTION:hb-glib
+ * @title: hb-glib
+ * @short_description: GLib integration
+ * @include: hb-glib.h
+ *
+ * Functions for using HarfBuzz with the GLib library.
+ *
+ * HarfBuzz supports using GLib to provide Unicode data, by attaching
+ * GLib functions to the virtual methods in a #hb_unicode_funcs_t function
+ * structure.
+ **/
+
+
+/**
+ * hb_glib_script_to_script:
+ * @script: The GUnicodeScript identifier to query
+ *
+ * Fetches the #hb_script_t script that corresponds to the
+ * specified GUnicodeScript identifier.
+ *
+ * Return value: the #hb_script_t script found
+ *
+ * Since: 0.9.38
+ **/
+hb_script_t
+hb_glib_script_to_script (GUnicodeScript script)
+{
+  return (hb_script_t) g_unicode_script_to_iso15924 (script);
+}
+
+/**
+ * hb_glib_script_from_script:
+ * @script: The #hb_script_t to query
+ *
+ * Fetches the GUnicodeScript identifier that corresponds to the
+ * specified #hb_script_t script.
+ *
+ * Return value: the GUnicodeScript identifier found
+ *
+ * Since: 0.9.38
+ **/
+GUnicodeScript
+hb_glib_script_from_script (hb_script_t script)
+{
+  return g_unicode_script_from_iso15924 (script);
+}
+
+
+static hb_unicode_combining_class_t
+hb_glib_unicode_combining_class (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+				 hb_codepoint_t      unicode,
+				 void               *user_data HB_UNUSED)
+
+{
+  return (hb_unicode_combining_class_t) g_unichar_combining_class (unicode);
+}
+
+static hb_unicode_general_category_t
+hb_glib_unicode_general_category (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+				  hb_codepoint_t      unicode,
+				  void               *user_data HB_UNUSED)
+
+{
+  /* hb_unicode_general_category_t and GUnicodeType are identical */
+  return (hb_unicode_general_category_t) g_unichar_type (unicode);
+}
+
+static hb_codepoint_t
+hb_glib_unicode_mirroring (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			   hb_codepoint_t      unicode,
+			   void               *user_data HB_UNUSED)
+{
+  g_unichar_get_mirror_char (unicode, &unicode);
+  return unicode;
+}
+
+static hb_script_t
+hb_glib_unicode_script (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			hb_codepoint_t      unicode,
+			void               *user_data HB_UNUSED)
+{
+  return hb_glib_script_to_script (g_unichar_get_script (unicode));
+}
+
+static hb_bool_t
+hb_glib_unicode_compose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			 hb_codepoint_t      a,
+			 hb_codepoint_t      b,
+			 hb_codepoint_t     *ab,
+			 void               *user_data HB_UNUSED)
+{
+#if GLIB_CHECK_VERSION(2,29,12)
+  return g_unichar_compose (a, b, ab);
+#endif
+
+  /* We don't ifdef-out the fallback code such that compiler always
+   * sees it and makes sure it's compilable. */
+
+  gchar utf8[12];
+  gchar *normalized;
+  int len;
+  hb_bool_t ret;
+
+  len = g_unichar_to_utf8 (a, utf8);
+  len += g_unichar_to_utf8 (b, utf8 + len);
+  normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFC);
+  len = g_utf8_strlen (normalized, -1);
+  if (unlikely (!len))
+    return false;
+
+  if (len == 1) {
+    *ab = g_utf8_get_char (normalized);
+    ret = true;
+  } else {
+    ret = false;
+  }
+
+  g_free (normalized);
+  return ret;
+}
+
+static hb_bool_t
+hb_glib_unicode_decompose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			   hb_codepoint_t      ab,
+			   hb_codepoint_t     *a,
+			   hb_codepoint_t     *b,
+			   void               *user_data HB_UNUSED)
+{
+#if GLIB_CHECK_VERSION(2,29,12)
+  return g_unichar_decompose (ab, a, b);
+#endif
+
+  /* We don't ifdef-out the fallback code such that compiler always
+   * sees it and makes sure it's compilable. */
+
+  gchar utf8[6];
+  gchar *normalized;
+  int len;
+  hb_bool_t ret;
+
+  len = g_unichar_to_utf8 (ab, utf8);
+  normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFD);
+  len = g_utf8_strlen (normalized, -1);
+  if (unlikely (!len))
+    return false;
+
+  if (len == 1) {
+    *a = g_utf8_get_char (normalized);
+    *b = 0;
+    ret = *a != ab;
+  } else if (len == 2) {
+    *a = g_utf8_get_char (normalized);
+    *b = g_utf8_get_char (g_utf8_next_char (normalized));
+    /* Here's the ugly part: if ab decomposes to a single character and
+     * that character decomposes again, we have to detect that and undo
+     * the second part :-(. */
+    gchar *recomposed = g_utf8_normalize (normalized, -1, G_NORMALIZE_NFC);
+    hb_codepoint_t c = g_utf8_get_char (recomposed);
+    if (c != ab && c != *a) {
+      *a = c;
+      *b = 0;
+    }
+    g_free (recomposed);
+    ret = true;
+  } else {
+    /* If decomposed to more than two characters, take the last one,
+     * and recompose the rest to get the first component. */
+    gchar *end = g_utf8_offset_to_pointer (normalized, len - 1);
+    gchar *recomposed;
+    *b = g_utf8_get_char (end);
+    recomposed = g_utf8_normalize (normalized, end - normalized, G_NORMALIZE_NFC);
+    /* We expect that recomposed has exactly one character now. */
+    *a = g_utf8_get_char (recomposed);
+    g_free (recomposed);
+    ret = true;
+  }
+
+  g_free (normalized);
+  return ret;
+}
+
+
+#if HB_USE_ATEXIT
+static void free_static_glib_funcs ();
+#endif
+
+static struct hb_glib_unicode_funcs_lazy_loader_t : hb_unicode_funcs_lazy_loader_t<hb_glib_unicode_funcs_lazy_loader_t>
+{
+  static hb_unicode_funcs_t *create ()
+  {
+    hb_unicode_funcs_t *funcs = hb_unicode_funcs_create (nullptr);
+
+    hb_unicode_funcs_set_combining_class_func (funcs, hb_glib_unicode_combining_class, nullptr, nullptr);
+    hb_unicode_funcs_set_general_category_func (funcs, hb_glib_unicode_general_category, nullptr, nullptr);
+    hb_unicode_funcs_set_mirroring_func (funcs, hb_glib_unicode_mirroring, nullptr, nullptr);
+    hb_unicode_funcs_set_script_func (funcs, hb_glib_unicode_script, nullptr, nullptr);
+    hb_unicode_funcs_set_compose_func (funcs, hb_glib_unicode_compose, nullptr, nullptr);
+    hb_unicode_funcs_set_decompose_func (funcs, hb_glib_unicode_decompose, nullptr, nullptr);
+
+    hb_unicode_funcs_make_immutable (funcs);
+
+#if HB_USE_ATEXIT
+    atexit (free_static_glib_funcs);
+#endif
+
+    return funcs;
+  }
+} static_glib_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_glib_funcs ()
+{
+  static_glib_funcs.free_instance ();
+}
+#endif
+
+/**
+ * hb_glib_get_unicode_funcs:
+ *
+ * Fetches a Unicode-functions structure that is populated
+ * with the appropriate GLib function for each method.
+ *
+ * Return value: (transfer none): a pointer to the #hb_unicode_funcs_t Unicode-functions structure
+ *
+ * Since: 0.9.38
+ **/
+hb_unicode_funcs_t *
+hb_glib_get_unicode_funcs ()
+{
+  return static_glib_funcs.get_unconst ();
+}
+
+
+
+#if GLIB_CHECK_VERSION(2,31,10)
+
+static void
+_hb_g_bytes_unref (void *data)
+{
+  g_bytes_unref ((GBytes *) data);
+}
+
+/**
+ * hb_glib_blob_create:
+ * @gbytes: the GBytes structure to work upon
+ *
+ * Creates an #hb_blob_t blob from the specified
+ * GBytes data structure.
+ *
+ * Return value: (transfer full): the new #hb_blob_t blob object
+ *
+ * Since: 0.9.38
+ **/
+hb_blob_t *
+hb_glib_blob_create (GBytes *gbytes)
+{
+  gsize size = 0;
+  gconstpointer data = g_bytes_get_data (gbytes, &size);
+  return hb_blob_create ((const char *) data,
+			 size,
+			 HB_MEMORY_MODE_READONLY,
+			 g_bytes_ref (gbytes),
+			 _hb_g_bytes_unref);
+}
+#endif
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-glib.h b/thirdparty/harfbuzz/src/hb-glib.h
new file mode 100644
index 0000000000..5f04183ba1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-glib.h
@@ -0,0 +1,56 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_GLIB_H
+#define HB_GLIB_H
+
+#include "hb.h"
+
+#include <glib.h>
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN hb_script_t
+hb_glib_script_to_script (GUnicodeScript script);
+
+HB_EXTERN GUnicodeScript
+hb_glib_script_from_script (hb_script_t script);
+
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_glib_get_unicode_funcs (void);
+
+#if GLIB_CHECK_VERSION(2,31,10)
+HB_EXTERN hb_blob_t *
+hb_glib_blob_create (GBytes *gbytes);
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_GLIB_H */
diff --git a/thirdparty/harfbuzz/src/hb-gobject-structs.cc b/thirdparty/harfbuzz/src/hb-gobject-structs.cc
new file mode 100644
index 0000000000..7c46e26400
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-gobject-structs.cc
@@ -0,0 +1,110 @@
+/*
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_GOBJECT
+
+
+/**
+ * SECTION:hb-gobject
+ * @title: hb-gobject
+ * @short_description: GObject integration support
+ * @include: hb-gobject.h
+ *
+ * Support for using HarfBuzz with the GObject library to provide
+ * type data.
+ *
+ * The types and functions listed here are solely a linkage between
+ * HarfBuzz's public data types and the GTypes used by the GObject framework.
+ * HarfBuzz uses GObject introspection to generate its Python bindings 
+ * (and potentially other language bindings); client programs should never need
+ * to access the GObject-integration mechanics.
+ *
+ * For client programs using the GNOME and GTK software stack, please see the
+ * GLib and FreeType integration pages.
+ **/
+
+
+/* g++ didn't like older gtype.h gcc-only code path. */
+#include <glib.h>
+#if !GLIB_CHECK_VERSION(2,29,16)
+#undef __GNUC__
+#undef __GNUC_MINOR__
+#define __GNUC__ 2
+#define __GNUC_MINOR__ 6
+#endif
+
+#include "hb-gobject.h"
+
+#define HB_DEFINE_BOXED_TYPE(name,copy_func,free_func) \
+GType \
+hb_gobject_##name##_get_type () \
+{ \
+   static gsize type_id = 0; \
+   if (g_once_init_enter (&type_id)) { \
+      GType id = g_boxed_type_register_static (g_intern_static_string ("hb_" #name "_t"), \
+					       (GBoxedCopyFunc) copy_func, \
+					       (GBoxedFreeFunc) free_func); \
+      g_once_init_leave (&type_id, id); \
+   } \
+   return type_id; \
+}
+
+#define HB_DEFINE_OBJECT_TYPE(name) \
+	HB_DEFINE_BOXED_TYPE (name, hb_##name##_reference, hb_##name##_destroy)
+
+#define HB_DEFINE_VALUE_TYPE(name) \
+	static hb_##name##_t *_hb_##name##_reference (const hb_##name##_t *l) \
+	{ \
+	  hb_##name##_t *c = (hb_##name##_t *) calloc (1, sizeof (hb_##name##_t)); \
+	  if (unlikely (!c)) return nullptr; \
+	  *c = *l; \
+	  return c; \
+	} \
+	static void _hb_##name##_destroy (hb_##name##_t *l) { free (l); } \
+	HB_DEFINE_BOXED_TYPE (name, _hb_##name##_reference, _hb_##name##_destroy)
+
+HB_DEFINE_OBJECT_TYPE (buffer)
+HB_DEFINE_OBJECT_TYPE (blob)
+HB_DEFINE_OBJECT_TYPE (face)
+HB_DEFINE_OBJECT_TYPE (font)
+HB_DEFINE_OBJECT_TYPE (font_funcs)
+HB_DEFINE_OBJECT_TYPE (set)
+HB_DEFINE_OBJECT_TYPE (map)
+HB_DEFINE_OBJECT_TYPE (shape_plan)
+HB_DEFINE_OBJECT_TYPE (unicode_funcs)
+HB_DEFINE_VALUE_TYPE (feature)
+HB_DEFINE_VALUE_TYPE (glyph_info)
+HB_DEFINE_VALUE_TYPE (glyph_position)
+HB_DEFINE_VALUE_TYPE (segment_properties)
+HB_DEFINE_VALUE_TYPE (user_data_key)
+
+HB_DEFINE_VALUE_TYPE (ot_math_glyph_variant)
+HB_DEFINE_VALUE_TYPE (ot_math_glyph_part)
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-gobject-structs.h b/thirdparty/harfbuzz/src/hb-gobject-structs.h
new file mode 100644
index 0000000000..6fad8d7019
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-gobject-structs.h
@@ -0,0 +1,142 @@
+/*
+ * Copyright (C) 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_GOBJECT_H_IN
+#error "Include <hb-gobject.h> instead."
+#endif
+
+#ifndef HB_GOBJECT_STRUCTS_H
+#define HB_GOBJECT_STRUCTS_H
+
+#include "hb.h"
+
+#include <glib-object.h>
+
+HB_BEGIN_DECLS
+
+
+/* Object types */
+
+/**
+ * hb_gobject_blob_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_blob_get_type (void);
+#define HB_GOBJECT_TYPE_BLOB (hb_gobject_blob_get_type ())
+
+/**
+ * hb_gobject_buffer_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_buffer_get_type (void);
+#define HB_GOBJECT_TYPE_BUFFER (hb_gobject_buffer_get_type ())
+
+/**
+ * hb_gobject_face_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_face_get_type (void);
+#define HB_GOBJECT_TYPE_FACE (hb_gobject_face_get_type ())
+
+/**
+ * hb_gobject_font_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_font_get_type (void);
+#define HB_GOBJECT_TYPE_FONT (hb_gobject_font_get_type ())
+
+/**
+ * hb_gobject_font_funcs_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_font_funcs_get_type (void);
+#define HB_GOBJECT_TYPE_FONT_FUNCS (hb_gobject_font_funcs_get_type ())
+
+HB_EXTERN GType
+hb_gobject_set_get_type (void);
+#define HB_GOBJECT_TYPE_SET (hb_gobject_set_get_type ())
+
+HB_EXTERN GType
+hb_gobject_map_get_type (void);
+#define HB_GOBJECT_TYPE_MAP (hb_gobject_map_get_type ())
+
+HB_EXTERN GType
+hb_gobject_shape_plan_get_type (void);
+#define HB_GOBJECT_TYPE_SHAPE_PLAN (hb_gobject_shape_plan_get_type ())
+
+/**
+ * hb_gobject_unicode_funcs_get_type:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN GType
+hb_gobject_unicode_funcs_get_type (void);
+#define HB_GOBJECT_TYPE_UNICODE_FUNCS (hb_gobject_unicode_funcs_get_type ())
+
+/* Value types */
+
+HB_EXTERN GType
+hb_gobject_feature_get_type (void);
+#define HB_GOBJECT_TYPE_FEATURE (hb_gobject_feature_get_type ())
+
+HB_EXTERN GType
+hb_gobject_glyph_info_get_type (void);
+#define HB_GOBJECT_TYPE_GLYPH_INFO (hb_gobject_glyph_info_get_type ())
+
+HB_EXTERN GType
+hb_gobject_glyph_position_get_type (void);
+#define HB_GOBJECT_TYPE_GLYPH_POSITION (hb_gobject_glyph_position_get_type ())
+
+HB_EXTERN GType
+hb_gobject_segment_properties_get_type (void);
+#define HB_GOBJECT_TYPE_SEGMENT_PROPERTIES (hb_gobject_segment_properties_get_type ())
+
+HB_EXTERN GType
+hb_gobject_user_data_key_get_type (void);
+#define HB_GOBJECT_TYPE_USER_DATA_KEY (hb_gobject_user_data_key_get_type ())
+
+HB_EXTERN GType
+hb_gobject_ot_math_glyph_variant_get_type (void);
+#define HB_GOBJECT_TYPE_OT_MATH_GLYPH_VARIANT (hb_gobject_ot_math_glyph_variant_get_type ())
+
+HB_EXTERN GType
+hb_gobject_ot_math_glyph_part_get_type (void);
+#define HB_GOBJECT_TYPE_OT_MATH_GLYPH_PART (hb_gobject_ot_math_glyph_part_get_type ())
+
+
+HB_END_DECLS
+
+#endif /* HB_GOBJECT_H */
diff --git a/thirdparty/harfbuzz/src/hb-gobject.h b/thirdparty/harfbuzz/src/hb-gobject.h
new file mode 100644
index 0000000000..8891aa0ee7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-gobject.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright (C) 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_GOBJECT_H
+#define HB_GOBJECT_H
+#define HB_GOBJECT_H_IN
+
+#include "hb.h"
+
+#include "hb-gobject-enums.h"
+#include "hb-gobject-structs.h"
+
+HB_BEGIN_DECLS
+HB_END_DECLS
+
+#undef HB_GOBJECT_H_IN
+#endif /* HB_GOBJECT_H */
diff --git a/thirdparty/harfbuzz/src/hb-graphite2.cc b/thirdparty/harfbuzz/src/hb-graphite2.cc
new file mode 100644
index 0000000000..d8a72dc2f1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-graphite2.cc
@@ -0,0 +1,442 @@
+/*
+ * Copyright © 2011  Martin Hosken
+ * Copyright © 2011  SIL International
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_GRAPHITE2
+
+#include "hb-shaper-impl.hh"
+
+#include "hb-graphite2.h"
+
+#include <graphite2/Segment.h>
+
+#include "hb-ot-layout.h"
+
+
+/**
+ * SECTION:hb-graphite2
+ * @title: hb-graphite2
+ * @short_description: Graphite2 integration
+ * @include: hb-graphite2.h
+ *
+ * Functions for using HarfBuzz with fonts that include Graphite features.
+ * 
+ * For Graphite features to work, you must be sure that HarfBuzz was compiled
+ * with the `graphite2` shaping engine enabled. Currently, the default is to
+ * not enable `graphite2` shaping.
+ **/
+
+
+/*
+ * shaper face data
+ */
+
+typedef struct hb_graphite2_tablelist_t
+{
+  struct hb_graphite2_tablelist_t *next;
+  hb_blob_t *blob;
+  unsigned int tag;
+} hb_graphite2_tablelist_t;
+
+struct hb_graphite2_face_data_t
+{
+  hb_face_t *face;
+  gr_face   *grface;
+  hb_atomic_ptr_t<hb_graphite2_tablelist_t> tlist;
+};
+
+static const void *hb_graphite2_get_table (const void *data, unsigned int tag, size_t *len)
+{
+  hb_graphite2_face_data_t *face_data = (hb_graphite2_face_data_t *) data;
+  hb_graphite2_tablelist_t *tlist = face_data->tlist;
+
+  hb_blob_t *blob = nullptr;
+
+  for (hb_graphite2_tablelist_t *p = tlist; p; p = p->next)
+    if (p->tag == tag) {
+      blob = p->blob;
+      break;
+    }
+
+  if (unlikely (!blob))
+  {
+    blob = face_data->face->reference_table (tag);
+
+    hb_graphite2_tablelist_t *p = (hb_graphite2_tablelist_t *) calloc (1, sizeof (hb_graphite2_tablelist_t));
+    if (unlikely (!p)) {
+      hb_blob_destroy (blob);
+      return nullptr;
+    }
+    p->blob = blob;
+    p->tag = tag;
+
+retry:
+    hb_graphite2_tablelist_t *tlist = face_data->tlist;
+    p->next = tlist;
+
+    if (unlikely (!face_data->tlist.cmpexch (tlist, p)))
+      goto retry;
+  }
+
+  unsigned int tlen;
+  const char *d = hb_blob_get_data (blob, &tlen);
+  *len = tlen;
+  return d;
+}
+
+hb_graphite2_face_data_t *
+_hb_graphite2_shaper_face_data_create (hb_face_t *face)
+{
+  hb_blob_t *silf_blob = face->reference_table (HB_GRAPHITE2_TAG_SILF);
+  /* Umm, we just reference the table to check whether it exists.
+   * Maybe add better API for this? */
+  if (!hb_blob_get_length (silf_blob))
+  {
+    hb_blob_destroy (silf_blob);
+    return nullptr;
+  }
+  hb_blob_destroy (silf_blob);
+
+  hb_graphite2_face_data_t *data = (hb_graphite2_face_data_t *) calloc (1, sizeof (hb_graphite2_face_data_t));
+  if (unlikely (!data))
+    return nullptr;
+
+  data->face = face;
+  data->grface = gr_make_face (data, &hb_graphite2_get_table, gr_face_preloadAll);
+
+  if (unlikely (!data->grface)) {
+    free (data);
+    return nullptr;
+  }
+
+  return data;
+}
+
+void
+_hb_graphite2_shaper_face_data_destroy (hb_graphite2_face_data_t *data)
+{
+  hb_graphite2_tablelist_t *tlist = data->tlist;
+
+  while (tlist)
+  {
+    hb_graphite2_tablelist_t *old = tlist;
+    hb_blob_destroy (tlist->blob);
+    tlist = tlist->next;
+    free (old);
+  }
+
+  gr_face_destroy (data->grface);
+
+  free (data);
+}
+
+/**
+ * hb_graphite2_face_get_gr_face:
+ * @face: @hb_face_t to query
+ *
+ * Fetches the Graphite2 gr_face corresponding to the specified
+ * #hb_face_t face object.
+ *
+ * Return value: the gr_face found
+ *
+ * Since: 0.9.10
+ */
+gr_face *
+hb_graphite2_face_get_gr_face (hb_face_t *face)
+{
+  const hb_graphite2_face_data_t *data = face->data.graphite2;
+  return data ? data->grface : nullptr;
+}
+
+
+/*
+ * shaper font data
+ */
+
+struct hb_graphite2_font_data_t {};
+
+hb_graphite2_font_data_t *
+_hb_graphite2_shaper_font_data_create (hb_font_t *font HB_UNUSED)
+{
+  return (hb_graphite2_font_data_t *) HB_SHAPER_DATA_SUCCEEDED;
+}
+
+void
+_hb_graphite2_shaper_font_data_destroy (hb_graphite2_font_data_t *data HB_UNUSED)
+{
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_graphite2_font_get_gr_font:
+ *
+ * Since: 0.9.10
+ * Deprecated: 1.4.2
+ */
+gr_font *
+hb_graphite2_font_get_gr_font (hb_font_t *font HB_UNUSED)
+{
+  return nullptr;
+}
+#endif
+
+
+/*
+ * shaper
+ */
+
+struct hb_graphite2_cluster_t {
+  unsigned int base_char;
+  unsigned int num_chars;
+  unsigned int base_glyph;
+  unsigned int num_glyphs;
+  unsigned int cluster;
+  unsigned int advance;
+};
+
+hb_bool_t
+_hb_graphite2_shape (hb_shape_plan_t    *shape_plan HB_UNUSED,
+		     hb_font_t          *font,
+		     hb_buffer_t        *buffer,
+		     const hb_feature_t *features,
+		     unsigned int        num_features)
+{
+  hb_face_t *face = font->face;
+  gr_face *grface = face->data.graphite2->grface;
+
+  const char *lang = hb_language_to_string (hb_buffer_get_language (buffer));
+  const char *lang_end = lang ? strchr (lang, '-') : nullptr;
+  int lang_len = lang_end ? lang_end - lang : -1;
+  gr_feature_val *feats = gr_face_featureval_for_lang (grface, lang ? hb_tag_from_string (lang, lang_len) : 0);
+
+  for (unsigned int i = 0; i < num_features; i++)
+  {
+    const gr_feature_ref *fref = gr_face_find_fref (grface, features[i].tag);
+    if (fref)
+      gr_fref_set_feature_value (fref, features[i].value, feats);
+  }
+
+  gr_segment *seg = nullptr;
+  const gr_slot *is;
+  unsigned int ci = 0, ic = 0;
+  unsigned int curradvx = 0, curradvy = 0;
+
+  unsigned int scratch_size;
+  hb_buffer_t::scratch_buffer_t *scratch = buffer->get_scratch_buffer (&scratch_size);
+
+  uint32_t *chars = (uint32_t *) scratch;
+
+  for (unsigned int i = 0; i < buffer->len; ++i)
+    chars[i] = buffer->info[i].codepoint;
+
+  /* TODO ensure_native_direction. */
+
+  hb_tag_t script_tag[HB_OT_MAX_TAGS_PER_SCRIPT];
+  unsigned int count = HB_OT_MAX_TAGS_PER_SCRIPT;
+  hb_ot_tags_from_script_and_language (hb_buffer_get_script (buffer),
+				       HB_LANGUAGE_INVALID,
+				       &count,
+				       script_tag,
+				       nullptr, nullptr);
+
+  seg = gr_make_seg (nullptr, grface,
+		     count ? script_tag[count - 1] : HB_OT_TAG_DEFAULT_SCRIPT,
+		     feats,
+		     gr_utf32, chars, buffer->len,
+		     2 | (hb_buffer_get_direction (buffer) == HB_DIRECTION_RTL ? 1 : 0));
+
+  if (unlikely (!seg)) {
+    if (feats) gr_featureval_destroy (feats);
+    return false;
+  }
+
+  unsigned int glyph_count = gr_seg_n_slots (seg);
+  if (unlikely (!glyph_count)) {
+    if (feats) gr_featureval_destroy (feats);
+    gr_seg_destroy (seg);
+    buffer->len = 0;
+    return true;
+  }
+
+  buffer->ensure (glyph_count);
+  scratch = buffer->get_scratch_buffer (&scratch_size);
+  while ((DIV_CEIL (sizeof (hb_graphite2_cluster_t) * buffer->len, sizeof (*scratch)) +
+	  DIV_CEIL (sizeof (hb_codepoint_t) * glyph_count, sizeof (*scratch))) > scratch_size)
+  {
+    if (unlikely (!buffer->ensure (buffer->allocated * 2)))
+    {
+      if (feats) gr_featureval_destroy (feats);
+      gr_seg_destroy (seg);
+      return false;
+    }
+    scratch = buffer->get_scratch_buffer (&scratch_size);
+  }
+
+#define ALLOCATE_ARRAY(Type, name, len) \
+  Type *name = (Type *) scratch; \
+  do { \
+    unsigned int _consumed = DIV_CEIL ((len) * sizeof (Type), sizeof (*scratch)); \
+    assert (_consumed <= scratch_size); \
+    scratch += _consumed; \
+    scratch_size -= _consumed; \
+  } while (0)
+
+  ALLOCATE_ARRAY (hb_graphite2_cluster_t, clusters, buffer->len);
+  ALLOCATE_ARRAY (hb_codepoint_t, gids, glyph_count);
+
+#undef ALLOCATE_ARRAY
+
+  memset (clusters, 0, sizeof (clusters[0]) * buffer->len);
+
+  hb_codepoint_t *pg = gids;
+  clusters[0].cluster = buffer->info[0].cluster;
+  unsigned int upem = hb_face_get_upem (face);
+  float xscale = (float) font->x_scale / upem;
+  float yscale = (float) font->y_scale / upem;
+  yscale *= yscale / xscale;
+  unsigned int curradv = 0;
+  if (HB_DIRECTION_IS_BACKWARD(buffer->props.direction))
+  {
+    curradv = gr_slot_origin_X(gr_seg_first_slot(seg)) * xscale;
+    clusters[0].advance = gr_seg_advance_X(seg) * xscale - curradv;
+  }
+  else
+    clusters[0].advance = 0;
+  for (is = gr_seg_first_slot (seg), ic = 0; is; is = gr_slot_next_in_segment (is), ic++)
+  {
+    unsigned int before = gr_slot_before (is);
+    unsigned int after = gr_slot_after (is);
+    *pg = gr_slot_gid (is);
+    pg++;
+    while (clusters[ci].base_char > before && ci)
+    {
+      clusters[ci-1].num_chars += clusters[ci].num_chars;
+      clusters[ci-1].num_glyphs += clusters[ci].num_glyphs;
+      clusters[ci-1].advance += clusters[ci].advance;
+      ci--;
+    }
+
+    if (gr_slot_can_insert_before (is) && clusters[ci].num_chars && before >= clusters[ci].base_char + clusters[ci].num_chars)
+    {
+      hb_graphite2_cluster_t *c = clusters + ci + 1;
+      c->base_char = clusters[ci].base_char + clusters[ci].num_chars;
+      c->cluster = buffer->info[c->base_char].cluster;
+      c->num_chars = before - c->base_char;
+      c->base_glyph = ic;
+      c->num_glyphs = 0;
+      if (HB_DIRECTION_IS_BACKWARD(buffer->props.direction))
+      {
+	c->advance = curradv - gr_slot_origin_X(is) * xscale;
+	curradv -= c->advance;
+      }
+      else
+      {
+	c->advance = 0;
+	clusters[ci].advance += gr_slot_origin_X(is) * xscale - curradv;
+	curradv += clusters[ci].advance;
+      }
+      ci++;
+    }
+    clusters[ci].num_glyphs++;
+
+    if (clusters[ci].base_char + clusters[ci].num_chars < after + 1)
+	clusters[ci].num_chars = after + 1 - clusters[ci].base_char;
+  }
+
+  if (HB_DIRECTION_IS_BACKWARD(buffer->props.direction))
+    clusters[ci].advance += curradv;
+  else
+    clusters[ci].advance += gr_seg_advance_X(seg) * xscale - curradv;
+  ci++;
+
+  for (unsigned int i = 0; i < ci; ++i)
+  {
+    for (unsigned int j = 0; j < clusters[i].num_glyphs; ++j)
+    {
+      hb_glyph_info_t *info = &buffer->info[clusters[i].base_glyph + j];
+      info->codepoint = gids[clusters[i].base_glyph + j];
+      info->cluster = clusters[i].cluster;
+      info->var1.i32 = clusters[i].advance;     // all glyphs in the cluster get the same advance
+    }
+  }
+  buffer->len = glyph_count;
+
+  /* Positioning. */
+  unsigned int currclus = UINT_MAX;
+  const hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pPos = hb_buffer_get_glyph_positions (buffer, nullptr);
+  if (!HB_DIRECTION_IS_BACKWARD(buffer->props.direction))
+  {
+    curradvx = 0;
+    for (is = gr_seg_first_slot (seg); is; pPos++, ++info, is = gr_slot_next_in_segment (is))
+    {
+      pPos->x_offset = gr_slot_origin_X (is) * xscale - curradvx;
+      pPos->y_offset = gr_slot_origin_Y (is) * yscale - curradvy;
+      if (info->cluster != currclus) {
+	pPos->x_advance = info->var1.i32;
+	curradvx += pPos->x_advance;
+	currclus = info->cluster;
+      } else
+	pPos->x_advance = 0.;
+
+      pPos->y_advance = gr_slot_advance_Y (is, grface, nullptr) * yscale;
+      curradvy += pPos->y_advance;
+    }
+  }
+  else
+  {
+    curradvx = gr_seg_advance_X(seg) * xscale;
+    for (is = gr_seg_first_slot (seg); is; pPos++, info++, is = gr_slot_next_in_segment (is))
+    {
+      if (info->cluster != currclus)
+      {
+	pPos->x_advance = info->var1.i32;
+	curradvx -= pPos->x_advance;
+	currclus = info->cluster;
+      } else
+	pPos->x_advance = 0.;
+
+      pPos->y_advance = gr_slot_advance_Y (is, grface, nullptr) * yscale;
+      curradvy -= pPos->y_advance;
+      pPos->x_offset = gr_slot_origin_X (is) * xscale - info->var1.i32 - curradvx + pPos->x_advance;
+      pPos->y_offset = gr_slot_origin_Y (is) * yscale - curradvy;
+    }
+    hb_buffer_reverse_clusters (buffer);
+  }
+
+  if (feats) gr_featureval_destroy (feats);
+  gr_seg_destroy (seg);
+
+  buffer->unsafe_to_break_all ();
+
+  return true;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-graphite2.h b/thirdparty/harfbuzz/src/hb-graphite2.h
new file mode 100644
index 0000000000..f299da9f71
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-graphite2.h
@@ -0,0 +1,60 @@
+/*
+ * Copyright © 2011  Martin Hosken
+ * Copyright © 2011  SIL International
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_GRAPHITE2_H
+#define HB_GRAPHITE2_H
+
+#include "hb.h"
+
+#include <graphite2/Font.h>
+
+HB_BEGIN_DECLS
+
+/**
+ * HB_GRAPHITE2_TAG_SILF:
+ *
+ * The #hb_tag_t tag for the `Silf` table, which holds Graphite
+ * features. 
+ *
+ * For more information, see http://graphite.sil.org/
+ *
+ **/
+#define HB_GRAPHITE2_TAG_SILF HB_TAG('S','i','l','f')
+
+
+HB_EXTERN gr_face *
+hb_graphite2_face_get_gr_face (hb_face_t *face);
+
+#ifndef HB_DISABLE_DEPRECATED
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_graphite2_face_get_gr_face) gr_font *
+hb_graphite2_font_get_gr_font (hb_font_t *font);
+
+#endif
+
+
+HB_END_DECLS
+
+#endif /* HB_GRAPHITE2_H */
diff --git a/thirdparty/harfbuzz/src/hb-icu.cc b/thirdparty/harfbuzz/src/hb-icu.cc
new file mode 100644
index 0000000000..008a39e414
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-icu.cc
@@ -0,0 +1,296 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2009  Keith Stribley
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_ICU
+
+#include "hb-icu.h"
+
+#include "hb-machinery.hh"
+
+#include <unicode/uchar.h>
+#include <unicode/unorm2.h>
+#include <unicode/ustring.h>
+#include <unicode/utf16.h>
+#include <unicode/uversion.h>
+
+/* ICU extra semicolon, fixed since 65, https://github.com/unicode-org/icu/commit/480bec3 */
+#if U_ICU_VERSION_MAJOR_NUM < 65 && (defined(__GNUC__) || defined(__clang__))
+#define HB_ICU_EXTRA_SEMI_IGNORED
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wextra-semi-stmt"
+#endif
+
+/**
+ * SECTION:hb-icu
+ * @title: hb-icu
+ * @short_description: ICU integration
+ * @include: hb-icu.h
+ *
+ * Functions for using HarfBuzz with the International Components for Unicode
+ * (ICU) library. HarfBuzz supports using ICU to provide Unicode data, by attaching
+ * ICU functions to the virtual methods in a #hb_unicode_funcs_t function
+ * structure.
+ **/
+
+/**
+ * hb_icu_script_to_script:
+ * @script: The UScriptCode identifier to query
+ *
+ * Fetches the #hb_script_t script that corresponds to the
+ * specified UScriptCode identifier.
+ *
+ * Return value: the #hb_script_t script found
+ *
+ **/
+
+hb_script_t
+hb_icu_script_to_script (UScriptCode script)
+{
+  if (unlikely (script == USCRIPT_INVALID_CODE))
+    return HB_SCRIPT_INVALID;
+
+  return hb_script_from_string (uscript_getShortName (script), -1);
+}
+
+/**
+ * hb_icu_script_from_script:
+ * @script: The #hb_script_t script to query
+ *
+ * Fetches the UScriptCode identifier that corresponds to the
+ * specified #hb_script_t script.
+ *
+ * Return value: the UScriptCode identifier found
+ *
+ **/
+UScriptCode
+hb_icu_script_from_script (hb_script_t script)
+{
+  if (unlikely (script == HB_SCRIPT_INVALID))
+    return USCRIPT_INVALID_CODE;
+
+  unsigned int numScriptCode = 1 + u_getIntPropertyMaxValue (UCHAR_SCRIPT);
+  for (unsigned int i = 0; i < numScriptCode; i++)
+    if (unlikely (hb_icu_script_to_script ((UScriptCode) i) == script))
+      return (UScriptCode) i;
+
+  return USCRIPT_UNKNOWN;
+}
+
+
+static hb_unicode_combining_class_t
+hb_icu_unicode_combining_class (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+				hb_codepoint_t      unicode,
+				void               *user_data HB_UNUSED)
+
+{
+  return (hb_unicode_combining_class_t) u_getCombiningClass (unicode);
+}
+
+static hb_unicode_general_category_t
+hb_icu_unicode_general_category (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+				 hb_codepoint_t      unicode,
+				 void               *user_data HB_UNUSED)
+{
+  switch (u_getIntPropertyValue(unicode, UCHAR_GENERAL_CATEGORY))
+  {
+  case U_UNASSIGNED:			return HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED;
+
+  case U_UPPERCASE_LETTER:		return HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER;
+  case U_LOWERCASE_LETTER:		return HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER;
+  case U_TITLECASE_LETTER:		return HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER;
+  case U_MODIFIER_LETTER:		return HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER;
+  case U_OTHER_LETTER:			return HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER;
+
+  case U_NON_SPACING_MARK:		return HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK;
+  case U_ENCLOSING_MARK:		return HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK;
+  case U_COMBINING_SPACING_MARK:	return HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK;
+
+  case U_DECIMAL_DIGIT_NUMBER:		return HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER;
+  case U_LETTER_NUMBER:			return HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER;
+  case U_OTHER_NUMBER:			return HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER;
+
+  case U_SPACE_SEPARATOR:		return HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR;
+  case U_LINE_SEPARATOR:		return HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR;
+  case U_PARAGRAPH_SEPARATOR:		return HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR;
+
+  case U_CONTROL_CHAR:			return HB_UNICODE_GENERAL_CATEGORY_CONTROL;
+  case U_FORMAT_CHAR:			return HB_UNICODE_GENERAL_CATEGORY_FORMAT;
+  case U_PRIVATE_USE_CHAR:		return HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE;
+  case U_SURROGATE:			return HB_UNICODE_GENERAL_CATEGORY_SURROGATE;
+
+
+  case U_DASH_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION;
+  case U_START_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION;
+  case U_END_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION;
+  case U_CONNECTOR_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION;
+  case U_OTHER_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION;
+
+  case U_MATH_SYMBOL:			return HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL;
+  case U_CURRENCY_SYMBOL:		return HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL;
+  case U_MODIFIER_SYMBOL:		return HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL;
+  case U_OTHER_SYMBOL:			return HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL;
+
+  case U_INITIAL_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION;
+  case U_FINAL_PUNCTUATION:		return HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION;
+  }
+
+  return HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED;
+}
+
+static hb_codepoint_t
+hb_icu_unicode_mirroring (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			  hb_codepoint_t      unicode,
+			  void               *user_data HB_UNUSED)
+{
+  return u_charMirror(unicode);
+}
+
+static hb_script_t
+hb_icu_unicode_script (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+		       hb_codepoint_t      unicode,
+		       void               *user_data HB_UNUSED)
+{
+  UErrorCode status = U_ZERO_ERROR;
+  UScriptCode scriptCode = uscript_getScript(unicode, &status);
+
+  if (unlikely (U_FAILURE (status)))
+    return HB_SCRIPT_UNKNOWN;
+
+  return hb_icu_script_to_script (scriptCode);
+}
+
+static hb_bool_t
+hb_icu_unicode_compose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			hb_codepoint_t      a,
+			hb_codepoint_t      b,
+			hb_codepoint_t     *ab,
+			void               *user_data)
+{
+  const UNormalizer2 *normalizer = (const UNormalizer2 *) user_data;
+  UChar32 ret = unorm2_composePair (normalizer, a, b);
+  if (ret < 0) return false;
+  *ab = ret;
+  return true;
+}
+
+static hb_bool_t
+hb_icu_unicode_decompose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			  hb_codepoint_t      ab,
+			  hb_codepoint_t     *a,
+			  hb_codepoint_t     *b,
+			  void               *user_data)
+{
+  const UNormalizer2 *normalizer = (const UNormalizer2 *) user_data;
+  UChar decomposed[4];
+  int len;
+  UErrorCode icu_err = U_ZERO_ERROR;
+  len = unorm2_getRawDecomposition (normalizer, ab, decomposed,
+				    ARRAY_LENGTH (decomposed), &icu_err);
+  if (U_FAILURE (icu_err) || len < 0) return false;
+
+  len = u_countChar32 (decomposed, len);
+  if (len == 1)
+  {
+    U16_GET_UNSAFE (decomposed, 0, *a);
+    *b = 0;
+    return *a != ab;
+  }
+  else if (len == 2)
+  {
+    len = 0;
+    U16_NEXT_UNSAFE (decomposed, len, *a);
+    U16_NEXT_UNSAFE (decomposed, len, *b);
+  }
+  return true;
+}
+
+
+#if HB_USE_ATEXIT
+static void free_static_icu_funcs ();
+#endif
+
+static struct hb_icu_unicode_funcs_lazy_loader_t : hb_unicode_funcs_lazy_loader_t<hb_icu_unicode_funcs_lazy_loader_t>
+{
+  static hb_unicode_funcs_t *create ()
+  {
+    void *user_data = nullptr;
+    UErrorCode icu_err = U_ZERO_ERROR;
+    user_data = (void *) unorm2_getNFCInstance (&icu_err);
+    assert (user_data);
+
+    hb_unicode_funcs_t *funcs = hb_unicode_funcs_create (nullptr);
+
+    hb_unicode_funcs_set_combining_class_func (funcs, hb_icu_unicode_combining_class, nullptr, nullptr);
+    hb_unicode_funcs_set_general_category_func (funcs, hb_icu_unicode_general_category, nullptr, nullptr);
+    hb_unicode_funcs_set_mirroring_func (funcs, hb_icu_unicode_mirroring, nullptr, nullptr);
+    hb_unicode_funcs_set_script_func (funcs, hb_icu_unicode_script, nullptr, nullptr);
+    hb_unicode_funcs_set_compose_func (funcs, hb_icu_unicode_compose, user_data, nullptr);
+    hb_unicode_funcs_set_decompose_func (funcs, hb_icu_unicode_decompose, user_data, nullptr);
+
+    hb_unicode_funcs_make_immutable (funcs);
+
+#if HB_USE_ATEXIT
+    atexit (free_static_icu_funcs);
+#endif
+
+    return funcs;
+  }
+} static_icu_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_icu_funcs ()
+{
+  static_icu_funcs.free_instance ();
+}
+#endif
+
+/**
+ * hb_icu_get_unicode_funcs:
+ *
+ * Fetches a Unicode-functions structure that is populated
+ * with the appropriate ICU function for each method.
+ *
+ * Return value: (transfer none): a pointer to the #hb_unicode_funcs_t Unicode-functions structure
+ *
+ * Since: 0.9.38
+ **/
+hb_unicode_funcs_t *
+hb_icu_get_unicode_funcs ()
+{
+  return static_icu_funcs.get_unconst ();
+}
+
+#ifdef HB_ICU_EXTRA_SEMI_IGNORED
+#pragma GCC diagnostic pop
+#endif
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-icu.h b/thirdparty/harfbuzz/src/hb-icu.h
new file mode 100644
index 0000000000..2db6a7b679
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-icu.h
@@ -0,0 +1,52 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_ICU_H
+#define HB_ICU_H
+
+#include "hb.h"
+
+#include <unicode/uscript.h>
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN hb_script_t
+hb_icu_script_to_script (UScriptCode script);
+
+HB_EXTERN UScriptCode
+hb_icu_script_from_script (hb_script_t script);
+
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_icu_get_unicode_funcs (void);
+
+
+HB_END_DECLS
+
+#endif /* HB_ICU_H */
diff --git a/thirdparty/harfbuzz/src/hb-iter.hh b/thirdparty/harfbuzz/src/hb-iter.hh
new file mode 100644
index 0000000000..981c5c218c
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-iter.hh
@@ -0,0 +1,939 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ * Copyright © 2019  Facebook, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_ITER_HH
+#define HB_ITER_HH
+
+#include "hb.hh"
+#include "hb-algs.hh"
+#include "hb-meta.hh"
+
+
+/* Unified iterator object.
+ *
+ * The goal of this template is to make the same iterator interface
+ * available to all types, and make it very easy and compact to use.
+ * hb_iter_tator objects are small, light-weight, objects that can be
+ * copied by value.  If the collection / object being iterated on
+ * is writable, then the iterator returns lvalues, otherwise it
+ * returns rvalues.
+ *
+ * TODO Document more.
+ *
+ * If iterator implementation implements operator!=, then can be
+ * used in range-based for loop.  That comes free if the iterator
+ * is random-access.  Otherwise, the range-based for loop incurs
+ * one traversal to find end(), which can be avoided if written
+ * as a while-style for loop, or if iterator implements a faster
+ * __end__() method.
+ * TODO When opting in for C++17, address this by changing return
+ * type of .end()?
+ */
+
+/*
+ * Base classes for iterators.
+ */
+
+/* Base class for all iterators. */
+template <typename iter_t, typename Item = typename iter_t::__item_t__>
+struct hb_iter_t
+{
+  typedef Item item_t;
+  constexpr unsigned get_item_size () const { return hb_static_size (Item); }
+  static constexpr bool is_iterator = true;
+  static constexpr bool is_random_access_iterator = false;
+  static constexpr bool is_sorted_iterator = false;
+
+  private:
+  /* https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern */
+  const iter_t* thiz () const { return static_cast<const iter_t *> (this); }
+	iter_t* thiz ()       { return static_cast<      iter_t *> (this); }
+  public:
+
+  /* TODO:
+   * Port operators below to use hb_enable_if to sniff which method implements
+   * an operator and use it, and remove hb_iter_fallback_mixin_t completely. */
+
+  /* Operators. */
+  iter_t iter () const { return *thiz(); }
+  iter_t operator + () const { return *thiz(); }
+  iter_t begin () const { return *thiz(); }
+  iter_t end () const { return thiz()->__end__ (); }
+  explicit operator bool () const { return thiz()->__more__ (); }
+  unsigned len () const { return thiz()->__len__ (); }
+  /* The following can only be enabled if item_t is reference type.  Otherwise
+   * it will be returning pointer to temporary rvalue.
+   * TODO Use a wrapper return type to fix for non-reference type. */
+  template <typename T = item_t,
+	    hb_enable_if (hb_is_reference (T))>
+  hb_remove_reference<item_t>* operator -> () const { return hb_addressof (**thiz()); }
+  item_t operator * () const { return thiz()->__item__ (); }
+  item_t operator * () { return thiz()->__item__ (); }
+  item_t operator [] (unsigned i) const { return thiz()->__item_at__ (i); }
+  item_t operator [] (unsigned i) { return thiz()->__item_at__ (i); }
+  iter_t& operator += (unsigned count) &  { thiz()->__forward__ (count); return *thiz(); }
+  iter_t  operator += (unsigned count) && { thiz()->__forward__ (count); return *thiz(); }
+  iter_t& operator ++ () &  { thiz()->__next__ (); return *thiz(); }
+  iter_t  operator ++ () && { thiz()->__next__ (); return *thiz(); }
+  iter_t& operator -= (unsigned count) &  { thiz()->__rewind__ (count); return *thiz(); }
+  iter_t  operator -= (unsigned count) && { thiz()->__rewind__ (count); return *thiz(); }
+  iter_t& operator -- () &  { thiz()->__prev__ (); return *thiz(); }
+  iter_t  operator -- () && { thiz()->__prev__ (); return *thiz(); }
+  iter_t operator + (unsigned count) const { auto c = thiz()->iter (); c += count; return c; }
+  friend iter_t operator + (unsigned count, const iter_t &it) { return it + count; }
+  iter_t operator ++ (int) { iter_t c (*thiz()); ++*thiz(); return c; }
+  iter_t operator - (unsigned count) const { auto c = thiz()->iter (); c -= count; return c; }
+  iter_t operator -- (int) { iter_t c (*thiz()); --*thiz(); return c; }
+  template <typename T>
+  iter_t& operator >> (T &v) &  { v = **thiz(); ++*thiz(); return *thiz(); }
+  template <typename T>
+  iter_t  operator >> (T &v) && { v = **thiz(); ++*thiz(); return *thiz(); }
+  template <typename T>
+  iter_t& operator << (const T v) &  { **thiz() = v; ++*thiz(); return *thiz(); }
+  template <typename T>
+  iter_t  operator << (const T v) && { **thiz() = v; ++*thiz(); return *thiz(); }
+
+  protected:
+  hb_iter_t () = default;
+  hb_iter_t (const hb_iter_t &o HB_UNUSED) = default;
+  hb_iter_t (hb_iter_t &&o HB_UNUSED) = default;
+  hb_iter_t& operator = (const hb_iter_t &o HB_UNUSED) = default;
+  hb_iter_t& operator = (hb_iter_t &&o HB_UNUSED) = default;
+};
+
+#define HB_ITER_USING(Name) \
+  using item_t = typename Name::item_t; \
+  using Name::begin; \
+  using Name::end; \
+  using Name::get_item_size; \
+  using Name::is_iterator; \
+  using Name::iter; \
+  using Name::operator bool; \
+  using Name::len; \
+  using Name::operator ->; \
+  using Name::operator *; \
+  using Name::operator []; \
+  using Name::operator +=; \
+  using Name::operator ++; \
+  using Name::operator -=; \
+  using Name::operator --; \
+  using Name::operator +; \
+  using Name::operator -; \
+  using Name::operator >>; \
+  using Name::operator <<; \
+  static_assert (true, "")
+
+/* Returns iterator / item type of a type. */
+template <typename Iterable>
+using hb_iter_type = decltype (hb_deref (hb_declval (Iterable)).iter ());
+template <typename Iterable>
+using hb_item_type = decltype (*hb_deref (hb_declval (Iterable)).iter ());
+
+
+template <typename> struct hb_array_t;
+template <typename> struct hb_sorted_array_t;
+
+struct
+{
+  template <typename T> hb_iter_type<T>
+  operator () (T&& c) const
+  { return hb_deref (hb_forward<T> (c)).iter (); }
+
+  /* Specialization for C arrays. */
+
+  template <typename Type> inline hb_array_t<Type>
+  operator () (Type *array, unsigned int length) const
+  { return hb_array_t<Type> (array, length); }
+
+  template <typename Type, unsigned int length> hb_array_t<Type>
+  operator () (Type (&array)[length]) const
+  { return hb_array_t<Type> (array, length); }
+
+}
+HB_FUNCOBJ (hb_iter);
+struct
+{
+  template <typename T> unsigned
+  operator () (T&& c) const
+  { return c.len (); }
+
+}
+HB_FUNCOBJ (hb_len);
+
+/* Mixin to fill in what the subclass doesn't provide. */
+template <typename iter_t, typename item_t = typename iter_t::__item_t__>
+struct hb_iter_fallback_mixin_t
+{
+  private:
+  /* https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern */
+  const iter_t* thiz () const { return static_cast<const iter_t *> (this); }
+	iter_t* thiz ()       { return static_cast<      iter_t *> (this); }
+  public:
+
+  /* Access: Implement __item__(), or __item_at__() if random-access. */
+  item_t __item__ () const { return (*thiz())[0]; }
+  item_t __item_at__ (unsigned i) const { return *(*thiz() + i); }
+
+  /* Termination: Implement __more__(), or __len__() if random-access. */
+  bool __more__ () const { return bool (thiz()->len ()); }
+  unsigned __len__ () const
+  { iter_t c (*thiz()); unsigned l = 0; while (c) { c++; l++; } return l; }
+
+  /* Advancing: Implement __next__(), or __forward__() if random-access. */
+  void __next__ () { *thiz() += 1; }
+  void __forward__ (unsigned n) { while (*thiz() && n--) ++*thiz(); }
+
+  /* Rewinding: Implement __prev__() or __rewind__() if bidirectional. */
+  void __prev__ () { *thiz() -= 1; }
+  void __rewind__ (unsigned n) { while (*thiz() && n--) --*thiz(); }
+
+  /* Range-based for: Implement __end__() if can be done faster,
+   * and operator!=. */
+  iter_t __end__ () const
+  {
+    if (thiz()->is_random_access_iterator)
+      return *thiz() + thiz()->len ();
+    /* Above expression loops twice. Following loops once. */
+    auto it = *thiz();
+    while (it) ++it;
+    return it;
+  }
+
+  protected:
+  hb_iter_fallback_mixin_t () = default;
+  hb_iter_fallback_mixin_t (const hb_iter_fallback_mixin_t &o HB_UNUSED) = default;
+  hb_iter_fallback_mixin_t (hb_iter_fallback_mixin_t &&o HB_UNUSED) = default;
+  hb_iter_fallback_mixin_t& operator = (const hb_iter_fallback_mixin_t &o HB_UNUSED) = default;
+  hb_iter_fallback_mixin_t& operator = (hb_iter_fallback_mixin_t &&o HB_UNUSED) = default;
+};
+
+template <typename iter_t, typename item_t = typename iter_t::__item_t__>
+struct hb_iter_with_fallback_t :
+  hb_iter_t<iter_t, item_t>,
+  hb_iter_fallback_mixin_t<iter_t, item_t>
+{
+  protected:
+  hb_iter_with_fallback_t () = default;
+  hb_iter_with_fallback_t (const hb_iter_with_fallback_t &o HB_UNUSED) = default;
+  hb_iter_with_fallback_t (hb_iter_with_fallback_t &&o HB_UNUSED) = default;
+  hb_iter_with_fallback_t& operator = (const hb_iter_with_fallback_t &o HB_UNUSED) = default;
+  hb_iter_with_fallback_t& operator = (hb_iter_with_fallback_t &&o HB_UNUSED) = default;
+};
+
+/*
+ * Meta-programming predicates.
+ */
+
+/* hb_is_iterator() / hb_is_iterator_of() */
+
+template<typename Iter, typename Item>
+struct hb_is_iterator_of
+{
+  template <typename Item2 = Item>
+  static hb_true_type impl (hb_priority<2>, hb_iter_t<Iter, hb_type_identity<Item2>> *);
+  static hb_false_type impl (hb_priority<0>, const void *);
+
+  public:
+  static constexpr bool value = decltype (impl (hb_prioritize, hb_declval (Iter*)))::value;
+};
+#define hb_is_iterator_of(Iter, Item) hb_is_iterator_of<Iter, Item>::value
+#define hb_is_iterator(Iter) hb_is_iterator_of (Iter, typename Iter::item_t)
+
+/* hb_is_iterable() */
+
+template <typename T>
+struct hb_is_iterable
+{
+  private:
+
+  template <typename U>
+  static auto impl (hb_priority<1>) -> decltype (hb_declval (U).iter (), hb_true_type ());
+
+  template <typename>
+  static hb_false_type impl (hb_priority<0>);
+
+  public:
+  static constexpr bool value = decltype (impl<T> (hb_prioritize))::value;
+};
+#define hb_is_iterable(Iterable) hb_is_iterable<Iterable>::value
+
+/* hb_is_source_of() / hb_is_sink_of() */
+
+template<typename Iter, typename Item>
+struct hb_is_source_of
+{
+  private:
+  template <typename Iter2 = Iter,
+	    hb_enable_if (hb_is_convertible (typename Iter2::item_t, hb_add_lvalue_reference<hb_add_const<Item>>))>
+  static hb_true_type impl (hb_priority<2>);
+  template <typename Iter2 = Iter>
+  static auto impl (hb_priority<1>) -> decltype (hb_declval (Iter2) >> hb_declval (Item &), hb_true_type ());
+  static hb_false_type impl (hb_priority<0>);
+
+  public:
+  static constexpr bool value = decltype (impl (hb_prioritize))::value;
+};
+#define hb_is_source_of(Iter, Item) hb_is_source_of<Iter, Item>::value
+
+template<typename Iter, typename Item>
+struct hb_is_sink_of
+{
+  private:
+  template <typename Iter2 = Iter,
+	    hb_enable_if (hb_is_convertible (typename Iter2::item_t, hb_add_lvalue_reference<Item>))>
+  static hb_true_type impl (hb_priority<2>);
+  template <typename Iter2 = Iter>
+  static auto impl (hb_priority<1>) -> decltype (hb_declval (Iter2) << hb_declval (Item), hb_true_type ());
+  static hb_false_type impl (hb_priority<0>);
+
+  public:
+  static constexpr bool value = decltype (impl (hb_prioritize))::value;
+};
+#define hb_is_sink_of(Iter, Item) hb_is_sink_of<Iter, Item>::value
+
+/* This is commonly used, so define: */
+#define hb_is_sorted_source_of(Iter, Item) \
+	(hb_is_source_of(Iter, Item) && Iter::is_sorted_iterator)
+
+
+/* Range-based 'for' for iterables. */
+
+template <typename Iterable,
+	  hb_requires (hb_is_iterable (Iterable))>
+static inline auto begin (Iterable&& iterable) HB_AUTO_RETURN (hb_iter (iterable).begin ())
+
+template <typename Iterable,
+	  hb_requires (hb_is_iterable (Iterable))>
+static inline auto end (Iterable&& iterable) HB_AUTO_RETURN (hb_iter (iterable).end ())
+
+/* begin()/end() are NOT looked up non-ADL.  So each namespace must declare them.
+ * Do it for namespace OT. */
+namespace OT {
+
+template <typename Iterable,
+	  hb_requires (hb_is_iterable (Iterable))>
+static inline auto begin (Iterable&& iterable) HB_AUTO_RETURN (hb_iter (iterable).begin ())
+
+template <typename Iterable,
+	  hb_requires (hb_is_iterable (Iterable))>
+static inline auto end (Iterable&& iterable) HB_AUTO_RETURN (hb_iter (iterable).end ())
+
+}
+
+
+/*
+ * Adaptors, combiners, etc.
+ */
+
+template <typename Lhs, typename Rhs,
+	  hb_requires (hb_is_iterator (Lhs))>
+static inline auto
+operator | (Lhs&& lhs, Rhs&& rhs) HB_AUTO_RETURN (hb_forward<Rhs> (rhs) (hb_forward<Lhs> (lhs)))
+
+/* hb_map(), hb_filter(), hb_reduce() */
+
+enum  class hb_function_sortedness_t {
+  NOT_SORTED,
+  RETAINS_SORTING,
+  SORTED,
+};
+
+template <typename Iter, typename Proj, hb_function_sortedness_t Sorted,
+	 hb_requires (hb_is_iterator (Iter))>
+struct hb_map_iter_t :
+  hb_iter_t<hb_map_iter_t<Iter, Proj, Sorted>,
+	    decltype (hb_get (hb_declval (Proj), *hb_declval (Iter)))>
+{
+  hb_map_iter_t (const Iter& it, Proj f_) : it (it), f (f_) {}
+
+  typedef decltype (hb_get (hb_declval (Proj), *hb_declval (Iter))) __item_t__;
+  static constexpr bool is_random_access_iterator = Iter::is_random_access_iterator;
+  static constexpr bool is_sorted_iterator =
+    Sorted == hb_function_sortedness_t::SORTED ? true :
+    Sorted == hb_function_sortedness_t::RETAINS_SORTING ? Iter::is_sorted_iterator :
+    false;
+  __item_t__ __item__ () const { return hb_get (f.get (), *it); }
+  __item_t__ __item_at__ (unsigned i) const { return hb_get (f.get (), it[i]); }
+  bool __more__ () const { return bool (it); }
+  unsigned __len__ () const { return it.len (); }
+  void __next__ () { ++it; }
+  void __forward__ (unsigned n) { it += n; }
+  void __prev__ () { --it; }
+  void __rewind__ (unsigned n) { it -= n; }
+  hb_map_iter_t __end__ () const { return hb_map_iter_t (it.end (), f); }
+  bool operator != (const hb_map_iter_t& o) const
+  { return it != o.it; }
+
+  private:
+  Iter it;
+  hb_reference_wrapper<Proj> f;
+};
+
+template <typename Proj, hb_function_sortedness_t Sorted>
+struct hb_map_iter_factory_t
+{
+  hb_map_iter_factory_t (Proj f) : f (f) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  hb_map_iter_t<Iter, Proj, Sorted>
+  operator () (Iter it)
+  { return hb_map_iter_t<Iter, Proj, Sorted> (it, f); }
+
+  private:
+  Proj f;
+};
+struct
+{
+  template <typename Proj>
+  hb_map_iter_factory_t<Proj, hb_function_sortedness_t::NOT_SORTED>
+  operator () (Proj&& f) const
+  { return hb_map_iter_factory_t<Proj, hb_function_sortedness_t::NOT_SORTED> (f); }
+}
+HB_FUNCOBJ (hb_map);
+struct
+{
+  template <typename Proj>
+  hb_map_iter_factory_t<Proj, hb_function_sortedness_t::RETAINS_SORTING>
+  operator () (Proj&& f) const
+  { return hb_map_iter_factory_t<Proj, hb_function_sortedness_t::RETAINS_SORTING> (f); }
+}
+HB_FUNCOBJ (hb_map_retains_sorting);
+struct
+{
+  template <typename Proj>
+  hb_map_iter_factory_t<Proj, hb_function_sortedness_t::SORTED>
+  operator () (Proj&& f) const
+  { return hb_map_iter_factory_t<Proj, hb_function_sortedness_t::SORTED> (f); }
+}
+HB_FUNCOBJ (hb_map_sorted);
+
+template <typename Iter, typename Pred, typename Proj,
+	 hb_requires (hb_is_iterator (Iter))>
+struct hb_filter_iter_t :
+  hb_iter_with_fallback_t<hb_filter_iter_t<Iter, Pred, Proj>,
+			  typename Iter::item_t>
+{
+  hb_filter_iter_t (const Iter& it_, Pred p_, Proj f_) : it (it_), p (p_), f (f_)
+  { while (it && !hb_has (p.get (), hb_get (f.get (), *it))) ++it; }
+
+  typedef typename Iter::item_t __item_t__;
+  static constexpr bool is_sorted_iterator = Iter::is_sorted_iterator;
+  __item_t__ __item__ () const { return *it; }
+  bool __more__ () const { return bool (it); }
+  void __next__ () { do ++it; while (it && !hb_has (p.get (), hb_get (f.get (), *it))); }
+  void __prev__ () { do --it; while (it && !hb_has (p.get (), hb_get (f.get (), *it))); }
+  hb_filter_iter_t __end__ () const { return hb_filter_iter_t (it.end (), p, f); }
+  bool operator != (const hb_filter_iter_t& o) const
+  { return it != o.it; }
+
+  private:
+  Iter it;
+  hb_reference_wrapper<Pred> p;
+  hb_reference_wrapper<Proj> f;
+};
+template <typename Pred, typename Proj>
+struct hb_filter_iter_factory_t
+{
+  hb_filter_iter_factory_t (Pred p, Proj f) : p (p), f (f) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  hb_filter_iter_t<Iter, Pred, Proj>
+  operator () (Iter it)
+  { return hb_filter_iter_t<Iter, Pred, Proj> (it, p, f); }
+
+  private:
+  Pred p;
+  Proj f;
+};
+struct
+{
+  template <typename Pred = decltype ((hb_identity)),
+	    typename Proj = decltype ((hb_identity))>
+  hb_filter_iter_factory_t<Pred, Proj>
+  operator () (Pred&& p = hb_identity, Proj&& f = hb_identity) const
+  { return hb_filter_iter_factory_t<Pred, Proj> (p, f); }
+}
+HB_FUNCOBJ (hb_filter);
+
+template <typename Redu, typename InitT>
+struct hb_reduce_t
+{
+  hb_reduce_t (Redu r, InitT init_value) : r (r), init_value (init_value) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter)),
+	    typename AccuT = hb_decay<decltype (hb_declval (Redu) (hb_declval (InitT), hb_declval (typename Iter::item_t)))>>
+  AccuT
+  operator () (Iter it)
+  {
+    AccuT value = init_value;
+    for (; it; ++it)
+      value = r (value, *it);
+    return value;
+  }
+
+  private:
+  Redu r;
+  InitT init_value;
+};
+struct
+{
+  template <typename Redu, typename InitT>
+  hb_reduce_t<Redu, InitT>
+  operator () (Redu&& r, InitT init_value) const
+  { return hb_reduce_t<Redu, InitT> (r, init_value); }
+}
+HB_FUNCOBJ (hb_reduce);
+
+
+/* hb_zip() */
+
+template <typename A, typename B>
+struct hb_zip_iter_t :
+  hb_iter_t<hb_zip_iter_t<A, B>,
+	    hb_pair_t<typename A::item_t, typename B::item_t>>
+{
+  hb_zip_iter_t () {}
+  hb_zip_iter_t (const A& a, const B& b) : a (a), b (b) {}
+
+  typedef hb_pair_t<typename A::item_t, typename B::item_t> __item_t__;
+  static constexpr bool is_random_access_iterator =
+    A::is_random_access_iterator &&
+    B::is_random_access_iterator;
+  /* Note.  The following categorization is only valid if A is strictly sorted,
+   * ie. does NOT have duplicates.  Previously I tried to categorize sortedness
+   * more granularly, see commits:
+   *
+   *   513762849a683914fc266a17ddf38f133cccf072
+   *   4d3cf2adb669c345cc43832d11689271995e160a
+   *
+   * However, that was not enough, since hb_sorted_array_t, hb_sorted_vector_t,
+   * SortedArrayOf, etc all needed to be updated to add more variants.  At that
+   * point I saw it not worth the effort, and instead we now deem all sorted
+   * collections as essentially strictly-sorted for the purposes of zip.
+   *
+   * The above assumption is not as bad as it sounds.  Our "sorted" comes with
+   * no guarantees.  It's just a contract, put in place to help you remember,
+   * and think about, whether an iterator you receive is expected to be
+   * sorted or not.  As such, it's not perfect by definition, and should not
+   * be treated so.  The inaccuracy here just errs in the direction of being
+   * more permissive, so your code compiles instead of erring on the side of
+   * marking your zipped iterator unsorted in which case your code won't
+   * compile.
+   *
+   * This semantical limitation does NOT affect logic in any other place I
+   * know of as of this writing.
+   */
+  static constexpr bool is_sorted_iterator = A::is_sorted_iterator;
+
+  __item_t__ __item__ () const { return __item_t__ (*a, *b); }
+  __item_t__ __item_at__ (unsigned i) const { return __item_t__ (a[i], b[i]); }
+  bool __more__ () const { return bool (a) && bool (b); }
+  unsigned __len__ () const { return hb_min (a.len (), b.len ()); }
+  void __next__ () { ++a; ++b; }
+  void __forward__ (unsigned n) { a += n; b += n; }
+  void __prev__ () { --a; --b; }
+  void __rewind__ (unsigned n) { a -= n; b -= n; }
+  hb_zip_iter_t __end__ () const { return hb_zip_iter_t (a.end (), b.end ()); }
+  /* Note, we should stop if ANY of the iters reaches end.  As such two compare
+   * unequal if both items are unequal, NOT if either is unequal. */
+  bool operator != (const hb_zip_iter_t& o) const
+  { return a != o.a && b != o.b; }
+
+  private:
+  A a;
+  B b;
+};
+struct
+{ HB_PARTIALIZE(2);
+  template <typename A, typename B,
+	    hb_requires (hb_is_iterable (A) && hb_is_iterable (B))>
+  hb_zip_iter_t<hb_iter_type<A>, hb_iter_type<B>>
+  operator () (A&& a, B&& b) const
+  { return hb_zip_iter_t<hb_iter_type<A>, hb_iter_type<B>> (hb_iter (a), hb_iter (b)); }
+}
+HB_FUNCOBJ (hb_zip);
+
+/* hb_apply() */
+
+template <typename Appl>
+struct hb_apply_t
+{
+  hb_apply_t (Appl a) : a (a) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  void operator () (Iter it)
+  {
+    for (; it; ++it)
+      (void) hb_invoke (a, *it);
+  }
+
+  private:
+  Appl a;
+};
+struct
+{
+  template <typename Appl> hb_apply_t<Appl>
+  operator () (Appl&& a) const
+  { return hb_apply_t<Appl> (a); }
+
+  template <typename Appl> hb_apply_t<Appl&>
+  operator () (Appl *a) const
+  { return hb_apply_t<Appl&> (*a); }
+}
+HB_FUNCOBJ (hb_apply);
+
+/* hb_range()/hb_iota()/hb_repeat() */
+
+template <typename T, typename S>
+struct hb_range_iter_t :
+  hb_iter_t<hb_range_iter_t<T, S>, T>
+{
+  hb_range_iter_t (T start, T end_, S step) : v (start), end_ (end_for (start, end_, step)), step (step) {}
+
+  typedef T __item_t__;
+  static constexpr bool is_random_access_iterator = true;
+  static constexpr bool is_sorted_iterator = true;
+  __item_t__ __item__ () const { return hb_ridentity (v); }
+  __item_t__ __item_at__ (unsigned j) const { return v + j * step; }
+  bool __more__ () const { return v != end_; }
+  unsigned __len__ () const { return !step ? UINT_MAX : (end_ - v) / step; }
+  void __next__ () { v += step; }
+  void __forward__ (unsigned n) { v += n * step; }
+  void __prev__ () { v -= step; }
+  void __rewind__ (unsigned n) { v -= n * step; }
+  hb_range_iter_t __end__ () const { return hb_range_iter_t (end_, end_, step); }
+  bool operator != (const hb_range_iter_t& o) const
+  { return v != o.v; }
+
+  private:
+  static inline T end_for (T start, T end_, S step)
+  {
+    if (!step)
+      return end_;
+    auto res = (end_ - start) % step;
+    if (!res)
+      return end_;
+    end_ += step - res;
+    return end_;
+  }
+
+  private:
+  T v;
+  T end_;
+  S step;
+};
+struct
+{
+  template <typename T = unsigned> hb_range_iter_t<T, unsigned>
+  operator () (T end = (unsigned) -1) const
+  { return hb_range_iter_t<T, unsigned> (0, end, 1u); }
+
+  template <typename T, typename S = unsigned> hb_range_iter_t<T, S>
+  operator () (T start, T end, S step = 1u) const
+  { return hb_range_iter_t<T, S> (start, end, step); }
+}
+HB_FUNCOBJ (hb_range);
+
+template <typename T, typename S>
+struct hb_iota_iter_t :
+  hb_iter_with_fallback_t<hb_iota_iter_t<T, S>, T>
+{
+  hb_iota_iter_t (T start, S step) : v (start), step (step) {}
+
+  private:
+
+  template <typename S2 = S>
+  auto
+  inc (hb_type_identity<S2> s, hb_priority<1>)
+    -> hb_void_t<decltype (hb_invoke (hb_forward<S2> (s), hb_declval<T&> ()))>
+  { v = hb_invoke (hb_forward<S2> (s), v); }
+
+  void
+  inc (S s, hb_priority<0>)
+  { v += s; }
+
+  public:
+
+  typedef T __item_t__;
+  static constexpr bool is_random_access_iterator = true;
+  static constexpr bool is_sorted_iterator = true;
+  __item_t__ __item__ () const { return hb_ridentity (v); }
+  bool __more__ () const { return true; }
+  unsigned __len__ () const { return UINT_MAX; }
+  void __next__ () { inc (step, hb_prioritize); }
+  void __prev__ () { v -= step; }
+  hb_iota_iter_t __end__ () const { return *this; }
+  bool operator != (const hb_iota_iter_t& o) const { return true; }
+
+  private:
+  T v;
+  S step;
+};
+struct
+{
+  template <typename T = unsigned, typename S = unsigned> hb_iota_iter_t<T, S>
+  operator () (T start = 0u, S step = 1u) const
+  { return hb_iota_iter_t<T, S> (start, step); }
+}
+HB_FUNCOBJ (hb_iota);
+
+template <typename T>
+struct hb_repeat_iter_t :
+  hb_iter_t<hb_repeat_iter_t<T>, T>
+{
+  hb_repeat_iter_t (T value) : v (value) {}
+
+  typedef T __item_t__;
+  static constexpr bool is_random_access_iterator = true;
+  static constexpr bool is_sorted_iterator = true;
+  __item_t__ __item__ () const { return v; }
+  __item_t__ __item_at__ (unsigned j) const { return v; }
+  bool __more__ () const { return true; }
+  unsigned __len__ () const { return UINT_MAX; }
+  void __next__ () {}
+  void __forward__ (unsigned) {}
+  void __prev__ () {}
+  void __rewind__ (unsigned) {}
+  hb_repeat_iter_t __end__ () const { return *this; }
+  bool operator != (const hb_repeat_iter_t& o) const { return true; }
+
+  private:
+  T v;
+};
+struct
+{
+  template <typename T> hb_repeat_iter_t<T>
+  operator () (T value) const
+  { return hb_repeat_iter_t<T> (value); }
+}
+HB_FUNCOBJ (hb_repeat);
+
+/* hb_enumerate()/hb_take() */
+
+struct
+{
+  template <typename Iterable,
+	    typename Index = unsigned,
+	    hb_requires (hb_is_iterable (Iterable))>
+  auto operator () (Iterable&& it, Index start = 0u) const HB_AUTO_RETURN
+  ( hb_zip (hb_iota (start), it) )
+}
+HB_FUNCOBJ (hb_enumerate);
+
+struct
+{ HB_PARTIALIZE(2);
+  template <typename Iterable,
+	    hb_requires (hb_is_iterable (Iterable))>
+  auto operator () (Iterable&& it, unsigned count) const HB_AUTO_RETURN
+  ( hb_zip (hb_range (count), it) | hb_map (hb_second) )
+
+  /* Specialization arrays. */
+
+  template <typename Type> inline hb_array_t<Type>
+  operator () (hb_array_t<Type> array, unsigned count) const
+  { return array.sub_array (0, count); }
+
+  template <typename Type> inline hb_sorted_array_t<Type>
+  operator () (hb_sorted_array_t<Type> array, unsigned count) const
+  { return array.sub_array (0, count); }
+}
+HB_FUNCOBJ (hb_take);
+
+struct
+{ HB_PARTIALIZE(2);
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  auto operator () (Iter it, unsigned count) const HB_AUTO_RETURN
+  (
+    + hb_iota (it, hb_add (count))
+    | hb_map (hb_take (count))
+    | hb_take ((hb_len (it) + count - 1) / count)
+  )
+}
+HB_FUNCOBJ (hb_chop);
+
+/* hb_sink() */
+
+template <typename Sink>
+struct hb_sink_t
+{
+  hb_sink_t (Sink s) : s (s) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  void operator () (Iter it)
+  {
+    for (; it; ++it)
+      s << *it;
+  }
+
+  private:
+  Sink s;
+};
+struct
+{
+  template <typename Sink> hb_sink_t<Sink>
+  operator () (Sink&& s) const
+  { return hb_sink_t<Sink> (s); }
+
+  template <typename Sink> hb_sink_t<Sink&>
+  operator () (Sink *s) const
+  { return hb_sink_t<Sink&> (*s); }
+}
+HB_FUNCOBJ (hb_sink);
+
+/* hb-drain: hb_sink to void / blackhole / /dev/null. */
+
+struct
+{
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  void operator () (Iter it) const
+  {
+    for (; it; ++it)
+      (void) *it;
+  }
+}
+HB_FUNCOBJ (hb_drain);
+
+/* hb_unzip(): unzip and sink to two sinks. */
+
+template <typename Sink1, typename Sink2>
+struct hb_unzip_t
+{
+  hb_unzip_t (Sink1 s1, Sink2 s2) : s1 (s1), s2 (s2) {}
+
+  template <typename Iter,
+	    hb_requires (hb_is_iterator (Iter))>
+  void operator () (Iter it)
+  {
+    for (; it; ++it)
+    {
+      const auto &v = *it;
+      s1 << v.first;
+      s2 << v.second;
+    }
+  }
+
+  private:
+  Sink1 s1;
+  Sink2 s2;
+};
+struct
+{
+  template <typename Sink1, typename Sink2> hb_unzip_t<Sink1, Sink2>
+  operator () (Sink1&& s1, Sink2&& s2) const
+  { return hb_unzip_t<Sink1, Sink2> (s1, s2); }
+
+  template <typename Sink1, typename Sink2> hb_unzip_t<Sink1&, Sink2&>
+  operator () (Sink1 *s1, Sink2 *s2) const
+  { return hb_unzip_t<Sink1&, Sink2&> (*s1, *s2); }
+}
+HB_FUNCOBJ (hb_unzip);
+
+
+/* hb-all, hb-any, hb-none. */
+
+struct
+{
+  template <typename Iterable,
+	    typename Pred = decltype ((hb_identity)),
+	    typename Proj = decltype ((hb_identity)),
+	    hb_requires (hb_is_iterable (Iterable))>
+  bool operator () (Iterable&& c,
+		    Pred&& p = hb_identity,
+		    Proj&& f = hb_identity) const
+  {
+    for (auto it = hb_iter (c); it; ++it)
+      if (!hb_match (hb_forward<Pred> (p), hb_get (hb_forward<Proj> (f), *it)))
+	return false;
+    return true;
+  }
+}
+HB_FUNCOBJ (hb_all);
+struct
+{
+  template <typename Iterable,
+	    typename Pred = decltype ((hb_identity)),
+	    typename Proj = decltype ((hb_identity)),
+	    hb_requires (hb_is_iterable (Iterable))>
+  bool operator () (Iterable&& c,
+		    Pred&& p = hb_identity,
+		    Proj&& f = hb_identity) const
+  {
+    for (auto it = hb_iter (c); it; ++it)
+      if (hb_match (hb_forward<Pred> (p), hb_get (hb_forward<Proj> (f), *it)))
+	return true;
+    return false;
+  }
+}
+HB_FUNCOBJ (hb_any);
+struct
+{
+  template <typename Iterable,
+	    typename Pred = decltype ((hb_identity)),
+	    typename Proj = decltype ((hb_identity)),
+	    hb_requires (hb_is_iterable (Iterable))>
+  bool operator () (Iterable&& c,
+		    Pred&& p = hb_identity,
+		    Proj&& f = hb_identity) const
+  {
+    for (auto it = hb_iter (c); it; ++it)
+      if (hb_match (hb_forward<Pred> (p), hb_get (hb_forward<Proj> (f), *it)))
+	return false;
+    return true;
+  }
+}
+HB_FUNCOBJ (hb_none);
+
+/*
+ * Algorithms operating on iterators.
+ */
+
+template <typename C, typename V,
+	  hb_requires (hb_is_iterable (C))>
+inline void
+hb_fill (C& c, const V &v)
+{
+  for (auto i = hb_iter (c); i; i++)
+    *i = v;
+}
+
+template <typename S, typename D>
+inline void
+hb_copy (S&& is, D&& id)
+{
+  hb_iter (is) | hb_sink (id);
+}
+
+
+#endif /* HB_ITER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-kern.hh b/thirdparty/harfbuzz/src/hb-kern.hh
new file mode 100644
index 0000000000..3f952fe7fc
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-kern.hh
@@ -0,0 +1,138 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_KERN_HH
+#define HB_KERN_HH
+
+#include "hb-open-type.hh"
+#include "hb-aat-layout-common.hh"
+#include "hb-ot-layout-gpos-table.hh"
+
+
+namespace OT {
+
+
+template <typename Driver>
+struct hb_kern_machine_t
+{
+  hb_kern_machine_t (const Driver &driver_,
+		     bool crossStream_ = false) :
+		       driver (driver_),
+		       crossStream (crossStream_) {}
+
+  HB_NO_SANITIZE_SIGNED_INTEGER_OVERFLOW
+  void kern (hb_font_t   *font,
+	     hb_buffer_t *buffer,
+	     hb_mask_t    kern_mask,
+	     bool         scale = true) const
+  {
+    OT::hb_ot_apply_context_t c (1, font, buffer);
+    c.set_lookup_mask (kern_mask);
+    c.set_lookup_props (OT::LookupFlag::IgnoreMarks);
+    auto &skippy_iter = c.iter_input;
+
+    bool horizontal = HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction);
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    hb_glyph_position_t *pos = buffer->pos;
+    for (unsigned int idx = 0; idx < count;)
+    {
+      if (!(info[idx].mask & kern_mask))
+      {
+	idx++;
+	continue;
+      }
+
+      skippy_iter.reset (idx, 1);
+      if (!skippy_iter.next ())
+      {
+	idx++;
+	continue;
+      }
+
+      unsigned int i = idx;
+      unsigned int j = skippy_iter.idx;
+
+      hb_position_t kern = driver.get_kerning (info[i].codepoint,
+					       info[j].codepoint);
+
+
+      if (likely (!kern))
+	goto skip;
+
+      if (horizontal)
+      {
+	if (scale)
+	  kern = font->em_scale_x (kern);
+	if (crossStream)
+	{
+	  pos[j].y_offset = kern;
+	  buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+	}
+	else
+	{
+	  hb_position_t kern1 = kern >> 1;
+	  hb_position_t kern2 = kern - kern1;
+	  pos[i].x_advance += kern1;
+	  pos[j].x_advance += kern2;
+	  pos[j].x_offset += kern2;
+	}
+      }
+      else
+      {
+	if (scale)
+	  kern = font->em_scale_y (kern);
+	if (crossStream)
+	{
+	  pos[j].x_offset = kern;
+	  buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+	}
+	else
+	{
+	  hb_position_t kern1 = kern >> 1;
+	  hb_position_t kern2 = kern - kern1;
+	  pos[i].y_advance += kern1;
+	  pos[j].y_advance += kern2;
+	  pos[j].y_offset += kern2;
+	}
+      }
+
+      buffer->unsafe_to_break (i, j + 1);
+
+    skip:
+      idx = skippy_iter.idx;
+    }
+  }
+
+  const Driver &driver;
+  bool crossStream;
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_KERN_HH */
diff --git a/thirdparty/harfbuzz/src/hb-machinery.hh b/thirdparty/harfbuzz/src/hb-machinery.hh
new file mode 100644
index 0000000000..54bc60d4c8
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-machinery.hh
@@ -0,0 +1,307 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2012,2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_MACHINERY_HH
+#define HB_MACHINERY_HH
+
+#include "hb.hh"
+#include "hb-blob.hh"
+
+#include "hb-dispatch.hh"
+#include "hb-sanitize.hh"
+#include "hb-serialize.hh"
+
+
+/*
+ * Casts
+ */
+
+/* StructAtOffset<T>(P,Ofs) returns the struct T& that is placed at memory
+ * location pointed to by P plus Ofs bytes. */
+template<typename Type>
+static inline const Type& StructAtOffset(const void *P, unsigned int offset)
+{ return * reinterpret_cast<const Type*> ((const char *) P + offset); }
+template<typename Type>
+static inline Type& StructAtOffset(void *P, unsigned int offset)
+{ return * reinterpret_cast<Type*> ((char *) P + offset); }
+template<typename Type>
+static inline const Type& StructAtOffsetUnaligned(const void *P, unsigned int offset)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+  return * reinterpret_cast<const Type*> ((const char *) P + offset);
+#pragma GCC diagnostic pop
+}
+template<typename Type>
+static inline Type& StructAtOffsetUnaligned(void *P, unsigned int offset)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+  return * reinterpret_cast<Type*> ((char *) P + offset);
+#pragma GCC diagnostic pop
+}
+
+/* StructAfter<T>(X) returns the struct T& that is placed after X.
+ * Works with X of variable size also.  X must implement get_size() */
+template<typename Type, typename TObject>
+static inline const Type& StructAfter(const TObject &X)
+{ return StructAtOffset<Type>(&X, X.get_size()); }
+template<typename Type, typename TObject>
+static inline Type& StructAfter(TObject &X)
+{ return StructAtOffset<Type>(&X, X.get_size()); }
+
+
+/*
+ * Size checking
+ */
+
+/* Check _assertion in a method environment */
+#define _DEFINE_INSTANCE_ASSERTION1(_line, _assertion) \
+  void _instance_assertion_on_line_##_line () const \
+  { static_assert ((_assertion), ""); }
+# define _DEFINE_INSTANCE_ASSERTION0(_line, _assertion) _DEFINE_INSTANCE_ASSERTION1 (_line, _assertion)
+# define DEFINE_INSTANCE_ASSERTION(_assertion) _DEFINE_INSTANCE_ASSERTION0 (__LINE__, _assertion)
+
+/* Check that _code compiles in a method environment */
+#define _DEFINE_COMPILES_ASSERTION1(_line, _code) \
+  void _compiles_assertion_on_line_##_line () const \
+  { _code; }
+# define _DEFINE_COMPILES_ASSERTION0(_line, _code) _DEFINE_COMPILES_ASSERTION1 (_line, _code)
+# define DEFINE_COMPILES_ASSERTION(_code) _DEFINE_COMPILES_ASSERTION0 (__LINE__, _code)
+
+
+#define DEFINE_SIZE_STATIC(size) \
+  DEFINE_INSTANCE_ASSERTION (sizeof (*this) == (size)) \
+  unsigned int get_size () const { return (size); } \
+  static constexpr unsigned null_size = (size); \
+  static constexpr unsigned min_size = (size); \
+  static constexpr unsigned static_size = (size)
+
+#define DEFINE_SIZE_UNION(size, _member) \
+  DEFINE_COMPILES_ASSERTION ((void) this->u._member.static_size) \
+  DEFINE_INSTANCE_ASSERTION (sizeof(this->u._member) == (size)) \
+  static constexpr unsigned null_size = (size); \
+  static constexpr unsigned min_size = (size)
+
+#define DEFINE_SIZE_MIN(size) \
+  DEFINE_INSTANCE_ASSERTION (sizeof (*this) >= (size)) \
+  static constexpr unsigned null_size = (size); \
+  static constexpr unsigned min_size = (size)
+
+#define DEFINE_SIZE_UNBOUNDED(size) \
+  DEFINE_INSTANCE_ASSERTION (sizeof (*this) >= (size)) \
+  static constexpr unsigned min_size = (size)
+
+#define DEFINE_SIZE_ARRAY(size, array) \
+  DEFINE_COMPILES_ASSERTION ((void) (array)[0].static_size) \
+  DEFINE_INSTANCE_ASSERTION (sizeof (*this) == (size) + (HB_VAR_ARRAY+0) * sizeof ((array)[0])) \
+  static constexpr unsigned null_size = (size); \
+  static constexpr unsigned min_size = (size)
+
+#define DEFINE_SIZE_ARRAY_SIZED(size, array) \
+  unsigned int get_size () const { return (size - (array).min_size + (array).get_size ()); } \
+  DEFINE_SIZE_ARRAY(size, array)
+
+
+
+/*
+ * Lazy loaders.
+ */
+
+template <typename Data, unsigned int WheresData>
+struct hb_data_wrapper_t
+{
+  static_assert (WheresData > 0, "");
+
+  Data * get_data () const
+  { return *(((Data **) (void *) this) - WheresData); }
+
+  bool is_inert () const { return !get_data (); }
+
+  template <typename Stored, typename Subclass>
+  Stored * call_create () const { return Subclass::create (get_data ()); }
+};
+template <>
+struct hb_data_wrapper_t<void, 0>
+{
+  bool is_inert () const { return false; }
+
+  template <typename Stored, typename Funcs>
+  Stored * call_create () const { return Funcs::create (); }
+};
+
+template <typename T1, typename T2> struct hb_non_void_t { typedef T1 value; };
+template <typename T2> struct hb_non_void_t<void, T2> { typedef T2 value; };
+
+template <typename Returned,
+	  typename Subclass = void,
+	  typename Data = void,
+	  unsigned int WheresData = 0,
+	  typename Stored = Returned>
+struct hb_lazy_loader_t : hb_data_wrapper_t<Data, WheresData>
+{
+  typedef typename hb_non_void_t<Subclass,
+				 hb_lazy_loader_t<Returned,Subclass,Data,WheresData,Stored>
+				>::value Funcs;
+
+  void init0 () {} /* Init, when memory is already set to 0. No-op for us. */
+  void init ()  { instance.set_relaxed (nullptr); }
+  void fini ()  { do_destroy (instance.get ()); }
+
+  void free_instance ()
+  {
+  retry:
+    Stored *p = instance.get ();
+    if (unlikely (p && !cmpexch (p, nullptr)))
+      goto retry;
+    do_destroy (p);
+  }
+
+  static void do_destroy (Stored *p)
+  {
+    if (p && p != const_cast<Stored *> (Funcs::get_null ()))
+      Funcs::destroy (p);
+  }
+
+  const Returned * operator -> () const { return get (); }
+  const Returned & operator * () const  { return *get (); }
+  explicit operator bool () const
+  { return get_stored () != Funcs::get_null (); }
+  template <typename C> operator const C * () const { return get (); }
+
+  Stored * get_stored () const
+  {
+  retry:
+    Stored *p = this->instance.get ();
+    if (unlikely (!p))
+    {
+      if (unlikely (this->is_inert ()))
+	return const_cast<Stored *> (Funcs::get_null ());
+
+      p = this->template call_create<Stored, Funcs> ();
+      if (unlikely (!p))
+	p = const_cast<Stored *> (Funcs::get_null ());
+
+      if (unlikely (!cmpexch (nullptr, p)))
+      {
+	do_destroy (p);
+	goto retry;
+      }
+    }
+    return p;
+  }
+  Stored * get_stored_relaxed () const
+  {
+    return this->instance.get_relaxed ();
+  }
+
+  bool cmpexch (Stored *current, Stored *value) const
+  {
+    /* This *must* be called when there are no other threads accessing. */
+    return this->instance.cmpexch (current, value);
+  }
+
+  const Returned * get () const { return Funcs::convert (get_stored ()); }
+  const Returned * get_relaxed () const { return Funcs::convert (get_stored_relaxed ()); }
+  Returned * get_unconst () const { return const_cast<Returned *> (Funcs::convert (get_stored ())); }
+
+  /* To be possibly overloaded by subclasses. */
+  static Returned* convert (Stored *p) { return p; }
+
+  /* By default null/init/fini the object. */
+  static const Stored* get_null () { return &Null (Stored); }
+  static Stored *create (Data *data)
+  {
+    Stored *p = (Stored *) calloc (1, sizeof (Stored));
+    if (likely (p))
+      p->init (data);
+    return p;
+  }
+  static Stored *create ()
+  {
+    Stored *p = (Stored *) calloc (1, sizeof (Stored));
+    if (likely (p))
+      p->init ();
+    return p;
+  }
+  static void destroy (Stored *p)
+  {
+    p->fini ();
+    free (p);
+  }
+
+//  private:
+  /* Must only have one pointer. */
+  hb_atomic_ptr_t<Stored *> instance;
+};
+
+/* Specializations. */
+
+template <typename T, unsigned int WheresFace>
+struct hb_face_lazy_loader_t : hb_lazy_loader_t<T,
+						hb_face_lazy_loader_t<T, WheresFace>,
+						hb_face_t, WheresFace> {};
+
+template <typename T, unsigned int WheresFace>
+struct hb_table_lazy_loader_t : hb_lazy_loader_t<T,
+						 hb_table_lazy_loader_t<T, WheresFace>,
+						 hb_face_t, WheresFace,
+						 hb_blob_t>
+{
+  static hb_blob_t *create (hb_face_t *face)
+  { return hb_sanitize_context_t ().reference_table<T> (face); }
+  static void destroy (hb_blob_t *p) { hb_blob_destroy (p); }
+
+  static const hb_blob_t *get_null ()
+  { return hb_blob_get_empty (); }
+
+  static const T* convert (const hb_blob_t *blob)
+  { return blob->as<T> (); }
+
+  hb_blob_t* get_blob () const { return this->get_stored (); }
+};
+
+template <typename Subclass>
+struct hb_font_funcs_lazy_loader_t : hb_lazy_loader_t<hb_font_funcs_t, Subclass>
+{
+  static void destroy (hb_font_funcs_t *p)
+  { hb_font_funcs_destroy (p); }
+  static const hb_font_funcs_t *get_null ()
+  { return hb_font_funcs_get_empty (); }
+};
+template <typename Subclass>
+struct hb_unicode_funcs_lazy_loader_t : hb_lazy_loader_t<hb_unicode_funcs_t, Subclass>
+{
+  static void destroy (hb_unicode_funcs_t *p)
+  { hb_unicode_funcs_destroy (p); }
+  static const hb_unicode_funcs_t *get_null ()
+  { return hb_unicode_funcs_get_empty (); }
+};
+
+
+#endif /* HB_MACHINERY_HH */
diff --git a/thirdparty/harfbuzz/src/hb-map.cc b/thirdparty/harfbuzz/src/hb-map.cc
new file mode 100644
index 0000000000..191be14372
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-map.cc
@@ -0,0 +1,268 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb-map.hh"
+
+
+/**
+ * SECTION:hb-map
+ * @title: hb-map
+ * @short_description: Object representing integer to integer mapping
+ * @include: hb.h
+ *
+ * Map objects are integer-to-integer hash-maps.  Currently they are
+ * not used in the HarfBuzz public API, but are provided for client's
+ * use if desired.
+ **/
+
+
+/**
+ * hb_map_create: (Xconstructor)
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 1.7.7
+ **/
+hb_map_t *
+hb_map_create ()
+{
+  hb_map_t *map;
+
+  if (!(map = hb_object_create<hb_map_t> ()))
+    return hb_map_get_empty ();
+
+  map->init_shallow ();
+
+  return map;
+}
+
+/**
+ * hb_map_get_empty:
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 1.7.7
+ **/
+hb_map_t *
+hb_map_get_empty ()
+{
+  return const_cast<hb_map_t *> (&Null (hb_map_t));
+}
+
+/**
+ * hb_map_reference: (skip)
+ * @map: a map.
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 1.7.7
+ **/
+hb_map_t *
+hb_map_reference (hb_map_t *map)
+{
+  return hb_object_reference (map);
+}
+
+/**
+ * hb_map_destroy: (skip)
+ * @map: a map.
+ *
+ * Since: 1.7.7
+ **/
+void
+hb_map_destroy (hb_map_t *map)
+{
+  if (!hb_object_destroy (map)) return;
+
+  map->fini_shallow ();
+
+  free (map);
+}
+
+/**
+ * hb_map_set_user_data: (skip)
+ * @map: a map.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ * Return value:
+ *
+ * Since: 1.7.7
+ **/
+hb_bool_t
+hb_map_set_user_data (hb_map_t           *map,
+		      hb_user_data_key_t *key,
+		      void *              data,
+		      hb_destroy_func_t   destroy,
+		      hb_bool_t           replace)
+{
+  return hb_object_set_user_data (map, key, data, destroy, replace);
+}
+
+/**
+ * hb_map_get_user_data: (skip)
+ * @map: a map.
+ * @key:
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 1.7.7
+ **/
+void *
+hb_map_get_user_data (hb_map_t           *map,
+		      hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (map, key);
+}
+
+
+/**
+ * hb_map_allocation_successful:
+ * @map: a map.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.7.7
+ **/
+hb_bool_t
+hb_map_allocation_successful (const hb_map_t  *map)
+{
+  return map->successful;
+}
+
+
+/**
+ * hb_map_set:
+ * @map: a map.
+ * @key:
+ * @value:
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+void
+hb_map_set (hb_map_t       *map,
+	    hb_codepoint_t  key,
+	    hb_codepoint_t  value)
+{
+  map->set (key, value);
+}
+
+/**
+ * hb_map_get:
+ * @map: a map.
+ * @key:
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+hb_codepoint_t
+hb_map_get (const hb_map_t *map,
+	    hb_codepoint_t  key)
+{
+  return map->get (key);
+}
+
+/**
+ * hb_map_del:
+ * @map: a map.
+ * @key:
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+void
+hb_map_del (hb_map_t       *map,
+	    hb_codepoint_t  key)
+{
+  map->del (key);
+}
+
+/**
+ * hb_map_has:
+ * @map: a map.
+ * @key:
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+hb_bool_t
+hb_map_has (const hb_map_t *map,
+	    hb_codepoint_t  key)
+{
+  return map->has (key);
+}
+
+
+/**
+ * hb_map_clear:
+ * @map: a map.
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+void
+hb_map_clear (hb_map_t *map)
+{
+  return map->clear ();
+}
+
+/**
+ * hb_map_is_empty:
+ * @map: a map.
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+hb_bool_t
+hb_map_is_empty (const hb_map_t *map)
+{
+  return map->is_empty ();
+}
+
+/**
+ * hb_map_get_population:
+ * @map: a map.
+ *
+ *
+ *
+ * Since: 1.7.7
+ **/
+unsigned int
+hb_map_get_population (const hb_map_t *map)
+{
+  return map->get_population ();
+}
diff --git a/thirdparty/harfbuzz/src/hb-map.h b/thirdparty/harfbuzz/src/hb-map.h
new file mode 100644
index 0000000000..b77843c2ba
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-map.h
@@ -0,0 +1,104 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_MAP_H
+#define HB_MAP_H
+
+#include "hb-common.h"
+
+HB_BEGIN_DECLS
+
+
+/*
+ * Since: 1.7.7
+ */
+#define HB_MAP_VALUE_INVALID ((hb_codepoint_t) -1)
+
+typedef struct hb_map_t hb_map_t;
+
+
+HB_EXTERN hb_map_t *
+hb_map_create (void);
+
+HB_EXTERN hb_map_t *
+hb_map_get_empty (void);
+
+HB_EXTERN hb_map_t *
+hb_map_reference (hb_map_t *map);
+
+HB_EXTERN void
+hb_map_destroy (hb_map_t *map);
+
+HB_EXTERN hb_bool_t
+hb_map_set_user_data (hb_map_t           *map,
+		      hb_user_data_key_t *key,
+		      void *              data,
+		      hb_destroy_func_t   destroy,
+		      hb_bool_t           replace);
+
+HB_EXTERN void *
+hb_map_get_user_data (hb_map_t           *map,
+		      hb_user_data_key_t *key);
+
+
+/* Returns false if allocation has failed before */
+HB_EXTERN hb_bool_t
+hb_map_allocation_successful (const hb_map_t *map);
+
+HB_EXTERN void
+hb_map_clear (hb_map_t *map);
+
+HB_EXTERN hb_bool_t
+hb_map_is_empty (const hb_map_t *map);
+
+HB_EXTERN unsigned int
+hb_map_get_population (const hb_map_t *map);
+
+HB_EXTERN void
+hb_map_set (hb_map_t       *map,
+	    hb_codepoint_t  key,
+	    hb_codepoint_t  value);
+
+HB_EXTERN hb_codepoint_t
+hb_map_get (const hb_map_t *map,
+	    hb_codepoint_t  key);
+
+HB_EXTERN void
+hb_map_del (hb_map_t       *map,
+	    hb_codepoint_t  key);
+
+HB_EXTERN hb_bool_t
+hb_map_has (const hb_map_t *map,
+	    hb_codepoint_t  key);
+
+
+HB_END_DECLS
+
+#endif /* HB_MAP_H */
diff --git a/thirdparty/harfbuzz/src/hb-map.hh b/thirdparty/harfbuzz/src/hb-map.hh
new file mode 100644
index 0000000000..92c1bd67e5
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-map.hh
@@ -0,0 +1,326 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_MAP_HH
+#define HB_MAP_HH
+
+#include "hb.hh"
+
+
+/*
+ * hb_hashmap_t
+ */
+
+template <typename K, typename V,
+	  K kINVALID = hb_is_pointer (K) ? 0 : hb_is_signed (K) ? hb_int_min (K) : (K) -1,
+	  V vINVALID = hb_is_pointer (V) ? 0 : hb_is_signed (V) ? hb_int_min (V) : (V) -1>
+struct hb_hashmap_t
+{
+  HB_DELETE_COPY_ASSIGN (hb_hashmap_t);
+  hb_hashmap_t ()  { init (); }
+  ~hb_hashmap_t () { fini (); }
+
+  static_assert (hb_is_integral (K) || hb_is_pointer (K), "");
+  static_assert (hb_is_integral (V) || hb_is_pointer (V), "");
+
+  struct item_t
+  {
+    K key;
+    V value;
+    uint32_t hash;
+
+    void clear () { key = kINVALID; value = vINVALID; hash = 0; }
+
+    bool operator == (const K &o) { return hb_deref (key) == hb_deref (o); }
+    bool operator == (const item_t &o) { return *this == o.key; }
+    bool is_unused () const    { return key == kINVALID; }
+    bool is_tombstone () const { return key != kINVALID && value == vINVALID; }
+    bool is_real () const { return key != kINVALID && value != vINVALID; }
+    hb_pair_t<K, V> get_pair() const { return hb_pair_t<K, V> (key, value); }
+  };
+
+  hb_object_header_t header;
+  bool successful; /* Allocations successful */
+  unsigned int population; /* Not including tombstones. */
+  unsigned int occupancy; /* Including tombstones. */
+  unsigned int mask;
+  unsigned int prime;
+  item_t *items;
+
+  void init_shallow ()
+  {
+    successful = true;
+    population = occupancy = 0;
+    mask = 0;
+    prime = 0;
+    items = nullptr;
+  }
+  void init ()
+  {
+    hb_object_init (this);
+    init_shallow ();
+  }
+  void fini_shallow ()
+  {
+    free (items);
+    items = nullptr;
+    population = occupancy = 0;
+  }
+  void fini ()
+  {
+    hb_object_fini (this);
+    fini_shallow ();
+  }
+
+  void reset ()
+  {
+    if (unlikely (hb_object_is_immutable (this)))
+      return;
+    successful = true;
+    clear ();
+  }
+
+  bool in_error () const { return !successful; }
+
+  bool resize ()
+  {
+    if (unlikely (!successful)) return false;
+
+    unsigned int power = hb_bit_storage (population * 2 + 8);
+    unsigned int new_size = 1u << power;
+    item_t *new_items = (item_t *) malloc ((size_t) new_size * sizeof (item_t));
+    if (unlikely (!new_items))
+    {
+      successful = false;
+      return false;
+    }
+    for (auto &_ : hb_iter (new_items, new_size))
+      _.clear ();
+
+    unsigned int old_size = mask + 1;
+    item_t *old_items = items;
+
+    /* Switch to new, empty, array. */
+    population = occupancy = 0;
+    mask = new_size - 1;
+    prime = prime_for (power);
+    items = new_items;
+
+    /* Insert back old items. */
+    if (old_items)
+      for (unsigned int i = 0; i < old_size; i++)
+	if (old_items[i].is_real ())
+	  set_with_hash (old_items[i].key,
+			 old_items[i].hash,
+			 old_items[i].value);
+
+    free (old_items);
+
+    return true;
+  }
+
+  void set (K key, V value)
+  {
+    set_with_hash (key, hb_hash (key), value);
+  }
+
+  V get (K key) const
+  {
+    if (unlikely (!items)) return vINVALID;
+    unsigned int i = bucket_for (key);
+    return items[i].is_real () && items[i] == key ? items[i].value : vINVALID;
+  }
+
+  void del (K key) { set (key, vINVALID); }
+
+  /* Has interface. */
+  static constexpr V SENTINEL = vINVALID;
+  typedef V value_t;
+  value_t operator [] (K k) const { return get (k); }
+  bool has (K k, V *vp = nullptr) const
+  {
+    V v = (*this)[k];
+    if (vp) *vp = v;
+    return v != SENTINEL;
+  }
+  /* Projection. */
+  V operator () (K k) const { return get (k); }
+
+  void clear ()
+  {
+    if (unlikely (hb_object_is_immutable (this)))
+      return;
+    if (items)
+      for (auto &_ : hb_iter (items, mask + 1))
+	_.clear ();
+
+    population = occupancy = 0;
+  }
+
+  bool is_empty () const { return population == 0; }
+
+  unsigned int get_population () const { return population; }
+
+  /*
+   * Iterator
+   */
+  auto iter () const HB_AUTO_RETURN
+  (
+    + hb_array (items, mask ? mask + 1 : 0)
+    | hb_filter (&item_t::is_real)
+    | hb_map (&item_t::get_pair)
+  )
+  auto keys () const HB_AUTO_RETURN
+  (
+    + hb_array (items, mask ? mask + 1 : 0)
+    | hb_filter (&item_t::is_real)
+    | hb_map (&item_t::key)
+    | hb_map (hb_ridentity)
+  )
+  auto values () const HB_AUTO_RETURN
+  (
+    + hb_array (items, mask ? mask + 1 : 0)
+    | hb_filter (&item_t::is_real)
+    | hb_map (&item_t::value)
+    | hb_map (hb_ridentity)
+  )
+
+  /* Sink interface. */
+  hb_hashmap_t& operator << (const hb_pair_t<K, V>& v)
+  { set (v.first, v.second); return *this; }
+
+  protected:
+
+  void set_with_hash (K key, uint32_t hash, V value)
+  {
+    if (unlikely (!successful)) return;
+    if (unlikely (key == kINVALID)) return;
+    if ((occupancy + occupancy / 2) >= mask && !resize ()) return;
+    unsigned int i = bucket_for_hash (key, hash);
+
+    if (value == vINVALID && items[i].key != key)
+      return; /* Trying to delete non-existent key. */
+
+    if (!items[i].is_unused ())
+    {
+      occupancy--;
+      if (items[i].is_tombstone ())
+	population--;
+    }
+
+    items[i].key = key;
+    items[i].value = value;
+    items[i].hash = hash;
+
+    occupancy++;
+    if (!items[i].is_tombstone ())
+      population++;
+  }
+
+  unsigned int bucket_for (K key) const
+  {
+    return bucket_for_hash (key, hb_hash (key));
+  }
+
+  unsigned int bucket_for_hash (K key, uint32_t hash) const
+  {
+    unsigned int i = hash % prime;
+    unsigned int step = 0;
+    unsigned int tombstone = (unsigned) -1;
+    while (!items[i].is_unused ())
+    {
+      if (items[i].hash == hash && items[i] == key)
+	return i;
+      if (tombstone == (unsigned) -1 && items[i].is_tombstone ())
+	tombstone = i;
+      i = (i + ++step) & mask;
+    }
+    return tombstone == (unsigned) -1 ? i : tombstone;
+  }
+
+  static unsigned int prime_for (unsigned int shift)
+  {
+    /* Following comment and table copied from glib. */
+    /* Each table size has an associated prime modulo (the first prime
+     * lower than the table size) used to find the initial bucket. Probing
+     * then works modulo 2^n. The prime modulo is necessary to get a
+     * good distribution with poor hash functions.
+     */
+    /* Not declaring static to make all kinds of compilers happy... */
+    /*static*/ const unsigned int prime_mod [32] =
+    {
+      1,          /* For 1 << 0 */
+      2,
+      3,
+      7,
+      13,
+      31,
+      61,
+      127,
+      251,
+      509,
+      1021,
+      2039,
+      4093,
+      8191,
+      16381,
+      32749,
+      65521,      /* For 1 << 16 */
+      131071,
+      262139,
+      524287,
+      1048573,
+      2097143,
+      4194301,
+      8388593,
+      16777213,
+      33554393,
+      67108859,
+      134217689,
+      268435399,
+      536870909,
+      1073741789,
+      2147483647  /* For 1 << 31 */
+    };
+
+    if (unlikely (shift >= ARRAY_LENGTH (prime_mod)))
+      return prime_mod[ARRAY_LENGTH (prime_mod) - 1];
+
+    return prime_mod[shift];
+  }
+};
+
+/*
+ * hb_map_t
+ */
+
+struct hb_map_t : hb_hashmap_t<hb_codepoint_t,
+			       hb_codepoint_t,
+			       HB_MAP_VALUE_INVALID,
+			       HB_MAP_VALUE_INVALID> {};
+
+
+#endif /* HB_MAP_HH */
diff --git a/thirdparty/harfbuzz/src/hb-meta.hh b/thirdparty/harfbuzz/src/hb-meta.hh
new file mode 100644
index 0000000000..4c0898b1b7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-meta.hh
@@ -0,0 +1,410 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_META_HH
+#define HB_META_HH
+
+#include "hb.hh"
+
+
+/*
+ * C++ template meta-programming & fundamentals used with them.
+ */
+
+/* Void!  For when we need a expression-type of void. */
+struct hb_empty_t {};
+
+/* https://en.cppreference.com/w/cpp/types/void_t */
+template<typename... Ts> struct _hb_void_t { typedef void type; };
+template<typename... Ts> using hb_void_t = typename _hb_void_t<Ts...>::type;
+
+template<typename Head, typename... Ts> struct _hb_head_t { typedef Head type; };
+template<typename... Ts> using hb_head_t = typename _hb_head_t<Ts...>::type;
+
+template <typename T, T v> struct hb_integral_constant { static constexpr T value = v; };
+template <bool b> using hb_bool_constant = hb_integral_constant<bool, b>;
+using hb_true_type = hb_bool_constant<true>;
+using hb_false_type = hb_bool_constant<false>;
+
+
+/* Basic type SFINAE. */
+
+template <bool B, typename T = void> struct hb_enable_if {};
+template <typename T>                struct hb_enable_if<true, T> { typedef T type; };
+#define hb_enable_if(Cond) typename hb_enable_if<(Cond)>::type* = nullptr
+/* Concepts/Requires alias: */
+#define hb_requires(Cond) hb_enable_if((Cond))
+
+template <typename T, typename T2> struct hb_is_same : hb_false_type {};
+template <typename T>              struct hb_is_same<T, T> : hb_true_type {};
+#define hb_is_same(T, T2) hb_is_same<T, T2>::value
+
+/* Function overloading SFINAE and priority. */
+
+#define HB_RETURN(Ret, E) -> hb_head_t<Ret, decltype ((E))> { return (E); }
+#define HB_AUTO_RETURN(E) -> decltype ((E)) { return (E); }
+#define HB_VOID_RETURN(E) -> hb_void_t<decltype ((E))> { (E); }
+
+template <unsigned Pri> struct hb_priority : hb_priority<Pri - 1> {};
+template <>             struct hb_priority<0> {};
+#define hb_prioritize hb_priority<16> ()
+
+#define HB_FUNCOBJ(x) static_const x HB_UNUSED
+
+
+template <typename T> struct hb_type_identity_t { typedef T type; };
+template <typename T> using hb_type_identity = typename hb_type_identity_t<T>::type;
+
+struct
+{
+  template <typename T> constexpr T*
+  operator () (T& arg) const
+  {
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-align"
+    /* https://en.cppreference.com/w/cpp/memory/addressof */
+    return reinterpret_cast<T*> (
+	     &const_cast<char&> (
+		reinterpret_cast<const volatile char&> (arg)));
+#pragma GCC diagnostic pop
+  }
+}
+HB_FUNCOBJ (hb_addressof);
+
+template <typename T> static inline T hb_declval ();
+#define hb_declval(T) (hb_declval<T> ())
+
+template <typename T> struct hb_match_const		: hb_type_identity_t<T>, hb_bool_constant<false>{};
+template <typename T> struct hb_match_const<const T>	: hb_type_identity_t<T>, hb_bool_constant<true>	{};
+template <typename T> using hb_remove_const = typename hb_match_const<T>::type;
+template <typename T> using hb_add_const = const T;
+#define hb_is_const(T) hb_match_const<T>::value
+template <typename T> struct hb_match_reference		: hb_type_identity_t<T>, hb_bool_constant<false>{};
+template <typename T> struct hb_match_reference<T &>	: hb_type_identity_t<T>, hb_bool_constant<true>	{};
+template <typename T> struct hb_match_reference<T &&>	: hb_type_identity_t<T>, hb_bool_constant<true>	{};
+template <typename T> using hb_remove_reference = typename hb_match_reference<T>::type;
+template <typename T> auto _hb_try_add_lvalue_reference (hb_priority<1>) -> hb_type_identity<T&>;
+template <typename T> auto _hb_try_add_lvalue_reference (hb_priority<0>) -> hb_type_identity<T>;
+template <typename T> using hb_add_lvalue_reference = decltype (_hb_try_add_lvalue_reference<T> (hb_prioritize));
+template <typename T> auto _hb_try_add_rvalue_reference (hb_priority<1>) -> hb_type_identity<T&&>;
+template <typename T> auto _hb_try_add_rvalue_reference (hb_priority<0>) -> hb_type_identity<T>;
+template <typename T> using hb_add_rvalue_reference = decltype (_hb_try_add_rvalue_reference<T> (hb_prioritize));
+#define hb_is_reference(T) hb_match_reference<T>::value
+template <typename T> struct hb_match_pointer		: hb_type_identity_t<T>, hb_bool_constant<false>{};
+template <typename T> struct hb_match_pointer<T *>	: hb_type_identity_t<T>, hb_bool_constant<true>	{};
+template <typename T> using hb_remove_pointer = typename hb_match_pointer<T>::type;
+template <typename T> auto _hb_try_add_pointer (hb_priority<1>) -> hb_type_identity<hb_remove_reference<T>*>;
+template <typename T> auto _hb_try_add_pointer (hb_priority<1>) -> hb_type_identity<T>;
+template <typename T> using hb_add_pointer = decltype (_hb_try_add_pointer<T> (hb_prioritize));
+#define hb_is_pointer(T) hb_match_pointer<T>::value
+
+
+/* TODO Add feature-parity to std::decay. */
+template <typename T> using hb_decay = hb_remove_const<hb_remove_reference<T>>;
+
+
+template<bool B, class T, class F>
+struct _hb_conditional { typedef T type; };
+template<class T, class F>
+struct _hb_conditional<false, T, F> { typedef F type; };
+template<bool B, class T, class F>
+using hb_conditional = typename _hb_conditional<B, T, F>::type;
+
+
+template <typename From, typename To>
+struct hb_is_convertible
+{
+  private:
+  static constexpr bool   from_void = hb_is_same (void, hb_decay<From>);
+  static constexpr bool     to_void = hb_is_same (void, hb_decay<To>  );
+  static constexpr bool either_void = from_void || to_void;
+  static constexpr bool   both_void = from_void && to_void;
+
+  static hb_true_type impl2 (hb_conditional<to_void, int, To>);
+
+  template <typename T>
+  static auto impl (hb_priority<1>) -> decltype (impl2 (hb_declval (T)));
+  template <typename T>
+  static hb_false_type impl (hb_priority<0>);
+  public:
+  static constexpr bool value = both_void ||
+		       (!either_void &&
+			decltype (impl<hb_conditional<from_void, int, From>> (hb_prioritize))::value);
+};
+#define hb_is_convertible(From,To) hb_is_convertible<From, To>::value
+
+template <typename Base, typename Derived>
+using hb_is_base_of = hb_is_convertible<hb_decay<Derived> *, hb_decay<Base> *>;
+#define hb_is_base_of(Base,Derived) hb_is_base_of<Base, Derived>::value
+
+template <typename From, typename To>
+using hb_is_cr_convertible = hb_bool_constant<
+  hb_is_same (hb_decay<From>, hb_decay<To>) &&
+  (!hb_is_const (From) || hb_is_const (To)) &&
+  (!hb_is_reference (To) || hb_is_const (To) || hb_is_reference (To))
+>;
+#define hb_is_cr_convertible(From,To) hb_is_cr_convertible<From, To>::value
+
+/* std::move and std::forward */
+
+template <typename T>
+static constexpr hb_remove_reference<T>&& hb_move (T&& t) { return (hb_remove_reference<T>&&) (t); }
+
+template <typename T>
+static constexpr T&& hb_forward (hb_remove_reference<T>& t) { return (T&&) t; }
+template <typename T>
+static constexpr T&& hb_forward (hb_remove_reference<T>&& t) { return (T&&) t; }
+
+struct
+{
+  template <typename T> constexpr auto
+  operator () (T&& v) const HB_AUTO_RETURN (hb_forward<T> (v))
+
+  template <typename T> constexpr auto
+  operator () (T *v) const HB_AUTO_RETURN (*v)
+}
+HB_FUNCOBJ (hb_deref);
+
+struct
+{
+  template <typename T> constexpr auto
+  operator () (T&& v) const HB_AUTO_RETURN (hb_forward<T> (v))
+
+  template <typename T> constexpr auto
+  operator () (T& v) const HB_AUTO_RETURN (hb_addressof (v))
+}
+HB_FUNCOBJ (hb_ref);
+
+template <typename T>
+struct hb_reference_wrapper
+{
+  hb_reference_wrapper (T v) : v (v) {}
+  bool operator == (const hb_reference_wrapper& o) const { return v == o.v; }
+  bool operator != (const hb_reference_wrapper& o) const { return v != o.v; }
+  operator T () const { return v; }
+  T get () const { return v; }
+  T v;
+};
+template <typename T>
+struct hb_reference_wrapper<T&>
+{
+  hb_reference_wrapper (T& v) : v (hb_addressof (v)) {}
+  bool operator == (const hb_reference_wrapper& o) const { return v == o.v; }
+  bool operator != (const hb_reference_wrapper& o) const { return v != o.v; }
+  operator T& () const { return *v; }
+  T& get () const { return *v; }
+  T* v;
+};
+
+
+template <typename T>
+using hb_is_integral = hb_bool_constant<
+  hb_is_same (hb_decay<T>, char) ||
+  hb_is_same (hb_decay<T>, signed char) ||
+  hb_is_same (hb_decay<T>, unsigned char) ||
+  hb_is_same (hb_decay<T>, signed int) ||
+  hb_is_same (hb_decay<T>, unsigned int) ||
+  hb_is_same (hb_decay<T>, signed short) ||
+  hb_is_same (hb_decay<T>, unsigned short) ||
+  hb_is_same (hb_decay<T>, signed long) ||
+  hb_is_same (hb_decay<T>, unsigned long) ||
+  hb_is_same (hb_decay<T>, signed long long) ||
+  hb_is_same (hb_decay<T>, unsigned long long) ||
+  false
+>;
+#define hb_is_integral(T) hb_is_integral<T>::value
+template <typename T>
+using hb_is_floating_point = hb_bool_constant<
+  hb_is_same (hb_decay<T>, float) ||
+  hb_is_same (hb_decay<T>, double) ||
+  hb_is_same (hb_decay<T>, long double) ||
+  false
+>;
+#define hb_is_floating_point(T) hb_is_floating_point<T>::value
+template <typename T>
+using hb_is_arithmetic = hb_bool_constant<
+  hb_is_integral (T) ||
+  hb_is_floating_point (T) ||
+  false
+>;
+#define hb_is_arithmetic(T) hb_is_arithmetic<T>::value
+
+
+template <typename T>
+using hb_is_signed = hb_conditional<hb_is_arithmetic (T),
+				    hb_bool_constant<(T) -1 < (T) 0>,
+				    hb_false_type>;
+#define hb_is_signed(T) hb_is_signed<T>::value
+template <typename T>
+using hb_is_unsigned = hb_conditional<hb_is_arithmetic (T),
+				      hb_bool_constant<(T) 0 < (T) -1>,
+				      hb_false_type>;
+#define hb_is_unsigned(T) hb_is_unsigned<T>::value
+
+template <typename T> struct hb_int_min;
+template <> struct hb_int_min<char>			: hb_integral_constant<char,			CHAR_MIN>	{};
+template <> struct hb_int_min<signed char>		: hb_integral_constant<signed char,		SCHAR_MIN>	{};
+template <> struct hb_int_min<unsigned char>		: hb_integral_constant<unsigned char,		0>		{};
+template <> struct hb_int_min<signed short>		: hb_integral_constant<signed short,		SHRT_MIN>	{};
+template <> struct hb_int_min<unsigned short>		: hb_integral_constant<unsigned short,		0>		{};
+template <> struct hb_int_min<signed int>		: hb_integral_constant<signed int,		INT_MIN>	{};
+template <> struct hb_int_min<unsigned int>		: hb_integral_constant<unsigned int,		0>		{};
+template <> struct hb_int_min<signed long>		: hb_integral_constant<signed long,		LONG_MIN>	{};
+template <> struct hb_int_min<unsigned long>		: hb_integral_constant<unsigned long,		0>		{};
+template <> struct hb_int_min<signed long long>		: hb_integral_constant<signed long long,	LLONG_MIN>	{};
+template <> struct hb_int_min<unsigned long long>	: hb_integral_constant<unsigned long long,	0>		{};
+#define hb_int_min(T) hb_int_min<T>::value
+template <typename T> struct hb_int_max;
+template <> struct hb_int_max<char>			: hb_integral_constant<char,			CHAR_MAX>	{};
+template <> struct hb_int_max<signed char>		: hb_integral_constant<signed char,		SCHAR_MAX>	{};
+template <> struct hb_int_max<unsigned char>		: hb_integral_constant<unsigned char,		UCHAR_MAX>	{};
+template <> struct hb_int_max<signed short>		: hb_integral_constant<signed short,		SHRT_MAX>	{};
+template <> struct hb_int_max<unsigned short>		: hb_integral_constant<unsigned short,		USHRT_MAX>	{};
+template <> struct hb_int_max<signed int>		: hb_integral_constant<signed int,		INT_MAX>	{};
+template <> struct hb_int_max<unsigned int>		: hb_integral_constant<unsigned int,		UINT_MAX>	{};
+template <> struct hb_int_max<signed long>		: hb_integral_constant<signed long,		LONG_MAX>	{};
+template <> struct hb_int_max<unsigned long>		: hb_integral_constant<unsigned long,		ULONG_MAX>	{};
+template <> struct hb_int_max<signed long long>		: hb_integral_constant<signed long long,	LLONG_MAX>	{};
+template <> struct hb_int_max<unsigned long long>	: hb_integral_constant<unsigned long long,	ULLONG_MAX>	{};
+#define hb_int_max(T) hb_int_max<T>::value
+
+
+
+template <typename T, typename>
+struct _hb_is_destructible : hb_false_type {};
+template <typename T>
+struct _hb_is_destructible<T, hb_void_t<decltype (hb_declval (T).~T ())>> : hb_true_type {};
+template <typename T>
+using hb_is_destructible = _hb_is_destructible<T, void>;
+#define hb_is_destructible(T) hb_is_destructible<T>::value
+
+template <typename T, typename, typename ...Ts>
+struct _hb_is_constructible : hb_false_type {};
+template <typename T, typename ...Ts>
+struct _hb_is_constructible<T, hb_void_t<decltype (T (hb_declval (Ts)...))>, Ts...> : hb_true_type {};
+template <typename T, typename ...Ts>
+using hb_is_constructible = _hb_is_constructible<T, void, Ts...>;
+#define hb_is_constructible(...) hb_is_constructible<__VA_ARGS__>::value
+
+template <typename T>
+using hb_is_default_constructible = hb_is_constructible<T>;
+#define hb_is_default_constructible(T) hb_is_default_constructible<T>::value
+
+template <typename T>
+using hb_is_copy_constructible = hb_is_constructible<T, hb_add_lvalue_reference<hb_add_const<T>>>;
+#define hb_is_copy_constructible(T) hb_is_copy_constructible<T>::value
+
+template <typename T>
+using hb_is_move_constructible = hb_is_constructible<T, hb_add_rvalue_reference<hb_add_const<T>>>;
+#define hb_is_move_constructible(T) hb_is_move_constructible<T>::value
+
+template <typename T, typename U, typename>
+struct _hb_is_assignable : hb_false_type {};
+template <typename T, typename U>
+struct _hb_is_assignable<T, U, hb_void_t<decltype (hb_declval (T) = hb_declval (U))>> : hb_true_type {};
+template <typename T, typename U>
+using hb_is_assignable = _hb_is_assignable<T, U, void>;
+#define hb_is_assignable(T,U) hb_is_assignable<T, U>::value
+
+template <typename T>
+using hb_is_copy_assignable = hb_is_assignable<hb_add_lvalue_reference<T>,
+					       hb_add_lvalue_reference<hb_add_const<T>>>;
+#define hb_is_copy_assignable(T) hb_is_copy_assignable<T>::value
+
+template <typename T>
+using hb_is_move_assignable = hb_is_assignable<hb_add_lvalue_reference<T>,
+					       hb_add_rvalue_reference<T>>;
+#define hb_is_move_assignable(T) hb_is_move_assignable<T>::value
+
+/* Trivial versions. */
+
+template <typename T> union hb_trivial { T value; };
+
+template <typename T>
+using hb_is_trivially_destructible= hb_is_destructible<hb_trivial<T>>;
+#define hb_is_trivially_destructible(T) hb_is_trivially_destructible<T>::value
+
+/* Don't know how to do the following. */
+//template <typename T, typename ...Ts>
+//using hb_is_trivially_constructible= hb_is_constructible<hb_trivial<T>, hb_trivial<Ts>...>;
+//#define hb_is_trivially_constructible(...) hb_is_trivially_constructible<__VA_ARGS__>::value
+
+template <typename T>
+using hb_is_trivially_default_constructible= hb_is_default_constructible<hb_trivial<T>>;
+#define hb_is_trivially_default_constructible(T) hb_is_trivially_default_constructible<T>::value
+
+template <typename T>
+using hb_is_trivially_copy_constructible= hb_is_copy_constructible<hb_trivial<T>>;
+#define hb_is_trivially_copy_constructible(T) hb_is_trivially_copy_constructible<T>::value
+
+template <typename T>
+using hb_is_trivially_move_constructible= hb_is_move_constructible<hb_trivial<T>>;
+#define hb_is_trivially_move_constructible(T) hb_is_trivially_move_constructible<T>::value
+
+/* Don't know how to do the following. */
+//template <typename T, typename U>
+//using hb_is_trivially_assignable= hb_is_assignable<hb_trivial<T>, hb_trivial<U>>;
+//#define hb_is_trivially_assignable(T,U) hb_is_trivially_assignable<T, U>::value
+
+template <typename T>
+using hb_is_trivially_copy_assignable= hb_is_copy_assignable<hb_trivial<T>>;
+#define hb_is_trivially_copy_assignable(T) hb_is_trivially_copy_assignable<T>::value
+
+template <typename T>
+using hb_is_trivially_move_assignable= hb_is_move_assignable<hb_trivial<T>>;
+#define hb_is_trivially_move_assignable(T) hb_is_trivially_move_assignable<T>::value
+
+template <typename T>
+using hb_is_trivially_copyable= hb_bool_constant<
+  hb_is_trivially_destructible (T) &&
+  (!hb_is_move_assignable (T) || hb_is_trivially_move_assignable (T)) &&
+  (!hb_is_move_constructible (T) || hb_is_trivially_move_constructible (T)) &&
+  (!hb_is_copy_assignable (T) || hb_is_trivially_copy_assignable (T)) &&
+  (!hb_is_copy_constructible (T) || hb_is_trivially_copy_constructible (T)) &&
+  true
+>;
+#define hb_is_trivially_copyable(T) hb_is_trivially_copyable<T>::value
+
+template <typename T>
+using hb_is_trivial= hb_bool_constant<
+  hb_is_trivially_copyable (T) &&
+  hb_is_trivially_default_constructible (T)
+>;
+#define hb_is_trivial(T) hb_is_trivial<T>::value
+
+/* hb_unwrap_type (T)
+ * If T has no T::type, returns T. Otherwise calls itself on T::type recursively.
+ */
+
+template <typename T, typename>
+struct _hb_unwrap_type : hb_type_identity_t<T> {};
+template <typename T>
+struct _hb_unwrap_type<T, hb_void_t<typename T::type>> : _hb_unwrap_type<typename T::type, void> {};
+template <typename T>
+using hb_unwrap_type = _hb_unwrap_type<T, void>;
+#define hb_unwrap_type(T) typename hb_unwrap_type<T>::type
+
+#endif /* HB_META_HH */
diff --git a/thirdparty/harfbuzz/src/hb-mutex.hh b/thirdparty/harfbuzz/src/hb-mutex.hh
new file mode 100644
index 0000000000..56392d049b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-mutex.hh
@@ -0,0 +1,133 @@
+/*
+ * Copyright © 2007  Chris Wilson
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Contributor(s):
+ *	Chris Wilson <chris@chris-wilson.co.uk>
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_MUTEX_HH
+#define HB_MUTEX_HH
+
+#include "hb.hh"
+
+
+/* mutex */
+
+/* We need external help for these */
+
+#if defined(HB_MUTEX_IMPL_INIT) \
+ && defined(hb_mutex_impl_init) \
+ && defined(hb_mutex_impl_lock) \
+ && defined(hb_mutex_impl_unlock) \
+ && defined(hb_mutex_impl_finish)
+
+/* Defined externally, i.e. in config.h; must have typedef'ed hb_mutex_impl_t as well. */
+
+
+#elif !defined(HB_NO_MT) && (defined(HAVE_PTHREAD) || defined(__APPLE__))
+
+#include <pthread.h>
+typedef pthread_mutex_t hb_mutex_impl_t;
+#define HB_MUTEX_IMPL_INIT	PTHREAD_MUTEX_INITIALIZER
+#define hb_mutex_impl_init(M)	pthread_mutex_init (M, nullptr)
+#define hb_mutex_impl_lock(M)	pthread_mutex_lock (M)
+#define hb_mutex_impl_unlock(M)	pthread_mutex_unlock (M)
+#define hb_mutex_impl_finish(M)	pthread_mutex_destroy (M)
+
+
+#elif !defined(HB_NO_MT) && defined(_WIN32)
+
+typedef CRITICAL_SECTION hb_mutex_impl_t;
+#define HB_MUTEX_IMPL_INIT	{0}
+#if !WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+#define hb_mutex_impl_init(M)	InitializeCriticalSectionEx (M, 0, 0)
+#else
+#define hb_mutex_impl_init(M)	InitializeCriticalSection (M)
+#endif
+#define hb_mutex_impl_lock(M)	EnterCriticalSection (M)
+#define hb_mutex_impl_unlock(M)	LeaveCriticalSection (M)
+#define hb_mutex_impl_finish(M)	DeleteCriticalSection (M)
+
+
+#elif !defined(HB_NO_MT) && defined(HAVE_INTEL_ATOMIC_PRIMITIVES)
+
+#if defined(HAVE_SCHED_H) && defined(HAVE_SCHED_YIELD)
+# include <sched.h>
+# define HB_SCHED_YIELD() sched_yield ()
+#else
+# define HB_SCHED_YIELD() HB_STMT_START {} HB_STMT_END
+#endif
+
+/* This actually is not a totally awful implementation. */
+typedef volatile int hb_mutex_impl_t;
+#define HB_MUTEX_IMPL_INIT	0
+#define hb_mutex_impl_init(M)	*(M) = 0
+#define hb_mutex_impl_lock(M)	HB_STMT_START { while (__sync_lock_test_and_set((M), 1)) HB_SCHED_YIELD (); } HB_STMT_END
+#define hb_mutex_impl_unlock(M)	__sync_lock_release (M)
+#define hb_mutex_impl_finish(M)	HB_STMT_START {} HB_STMT_END
+
+
+#elif defined(HB_NO_MT)
+
+typedef int hb_mutex_impl_t;
+#define HB_MUTEX_IMPL_INIT	0
+#define hb_mutex_impl_init(M)	HB_STMT_START {} HB_STMT_END
+#define hb_mutex_impl_lock(M)	HB_STMT_START {} HB_STMT_END
+#define hb_mutex_impl_unlock(M)	HB_STMT_START {} HB_STMT_END
+#define hb_mutex_impl_finish(M)	HB_STMT_START {} HB_STMT_END
+
+
+#else
+
+#error "Could not find any system to define mutex macros."
+#error "Check hb-mutex.hh for possible resolutions."
+
+#endif
+
+
+#define HB_MUTEX_INIT		{HB_MUTEX_IMPL_INIT}
+
+struct hb_mutex_t
+{
+  hb_mutex_impl_t m;
+
+  void init   () { hb_mutex_impl_init   (&m); }
+  void lock   () { hb_mutex_impl_lock   (&m); }
+  void unlock () { hb_mutex_impl_unlock (&m); }
+  void fini ()   { hb_mutex_impl_finish (&m); }
+};
+
+struct hb_lock_t
+{
+  hb_lock_t (hb_mutex_t &mutex_) : mutex (mutex_) { mutex.lock (); }
+  ~hb_lock_t () { mutex.unlock (); }
+  private:
+  hb_mutex_t &mutex;
+};
+
+
+#endif /* HB_MUTEX_HH */
diff --git a/thirdparty/harfbuzz/src/hb-null.hh b/thirdparty/harfbuzz/src/hb-null.hh
new file mode 100644
index 0000000000..9853939b07
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-null.hh
@@ -0,0 +1,184 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_NULL_HH
+#define HB_NULL_HH
+
+#include "hb.hh"
+#include "hb-meta.hh"
+
+
+/*
+ * Static pools
+ */
+
+/* Global nul-content Null pool.  Enlarge as necessary. */
+
+#define HB_NULL_POOL_SIZE 384
+
+/* Use SFINAE to sniff whether T has min_size; in which case return T::null_size,
+ * otherwise return sizeof(T). */
+
+/* The hard way...
+ * https://stackoverflow.com/questions/7776448/sfinae-tried-with-bool-gives-compiler-error-template-argument-tvalue-invol
+ */
+
+template <typename T, typename>
+struct _hb_null_size : hb_integral_constant<unsigned, sizeof (T)> {};
+template <typename T>
+struct _hb_null_size<T, hb_void_t<decltype (T::min_size)>> : hb_integral_constant<unsigned, T::null_size> {};
+
+template <typename T>
+using hb_null_size = _hb_null_size<T, void>;
+#define hb_null_size(T) hb_null_size<T>::value
+
+/* These doesn't belong here, but since is copy/paste from above, put it here. */
+
+/* hb_static_size (T)
+ * Returns T::static_size if T::min_size is defined, or sizeof (T) otherwise. */
+
+template <typename T, typename>
+struct _hb_static_size : hb_integral_constant<unsigned, sizeof (T)> {};
+template <typename T>
+struct _hb_static_size<T, hb_void_t<decltype (T::min_size)>> : hb_integral_constant<unsigned, T::static_size> {};
+template <typename T>
+using hb_static_size = _hb_static_size<T, void>;
+#define hb_static_size(T) hb_static_size<T>::value
+
+
+/*
+ * Null()
+ */
+
+extern HB_INTERNAL
+uint64_t const _hb_NullPool[(HB_NULL_POOL_SIZE + sizeof (uint64_t) - 1) / sizeof (uint64_t)];
+
+/* Generic nul-content Null objects. */
+template <typename Type>
+struct Null {
+  static Type const & get_null ()
+  {
+    static_assert (hb_null_size (Type) <= HB_NULL_POOL_SIZE, "Increase HB_NULL_POOL_SIZE.");
+    return *reinterpret_cast<Type const *> (_hb_NullPool);
+  }
+};
+template <typename QType>
+struct NullHelper
+{
+  typedef hb_remove_const<hb_remove_reference<QType>> Type;
+  static const Type & get_null () { return Null<Type>::get_null (); }
+};
+#define Null(Type) NullHelper<Type>::get_null ()
+
+/* Specializations for arbitrary-content Null objects expressed in bytes. */
+#define DECLARE_NULL_NAMESPACE_BYTES(Namespace, Type) \
+	} /* Close namespace. */ \
+	extern HB_INTERNAL const unsigned char _hb_Null_##Namespace##_##Type[Namespace::Type::null_size]; \
+	template <> \
+	struct Null<Namespace::Type> { \
+	  static Namespace::Type const & get_null () { \
+	    return *reinterpret_cast<const Namespace::Type *> (_hb_Null_##Namespace##_##Type); \
+	  } \
+	}; \
+	namespace Namespace { \
+	static_assert (true, "") /* Require semicolon after. */
+#define DEFINE_NULL_NAMESPACE_BYTES(Namespace, Type) \
+	const unsigned char _hb_Null_##Namespace##_##Type[Namespace::Type::null_size]
+
+/* Specializations for arbitrary-content Null objects expressed as struct initializer. */
+#define DECLARE_NULL_INSTANCE(Type) \
+	extern HB_INTERNAL const Type _hb_Null_##Type; \
+	template <> \
+	struct Null<Type> { \
+	  static Type const & get_null () { \
+	    return _hb_Null_##Type; \
+	  } \
+	}; \
+	static_assert (true, "") /* Require semicolon after. */
+#define DEFINE_NULL_INSTANCE(Type) \
+	const Type _hb_Null_##Type
+
+/* Global writable pool.  Enlarge as necessary. */
+
+/* To be fully correct, CrapPool must be thread_local. However, we do not rely on CrapPool
+ * for correct operation. It only exist to catch and divert program logic bugs instead of
+ * causing bad memory access. So, races there are not actually introducing incorrectness
+ * in the code. Has ~12kb binary size overhead to have it, also clang build fails with it. */
+extern HB_INTERNAL
+/*thread_local*/ uint64_t _hb_CrapPool[(HB_NULL_POOL_SIZE + sizeof (uint64_t) - 1) / sizeof (uint64_t)];
+
+/* CRAP pool: Common Region for Access Protection. */
+template <typename Type>
+static inline Type& Crap () {
+  static_assert (hb_null_size (Type) <= HB_NULL_POOL_SIZE, "Increase HB_NULL_POOL_SIZE.");
+  Type *obj = reinterpret_cast<Type *> (_hb_CrapPool);
+  memcpy (obj, &Null (Type), sizeof (*obj));
+  return *obj;
+}
+template <typename QType>
+struct CrapHelper
+{
+  typedef hb_remove_const<hb_remove_reference<QType>> Type;
+  static Type & get_crap () { return Crap<Type> (); }
+};
+#define Crap(Type) CrapHelper<Type>::get_crap ()
+
+template <typename Type>
+struct CrapOrNullHelper {
+  static Type & get () { return Crap (Type); }
+};
+template <typename Type>
+struct CrapOrNullHelper<const Type> {
+  static const Type & get () { return Null (Type); }
+};
+#define CrapOrNull(Type) CrapOrNullHelper<Type>::get ()
+
+
+/*
+ * hb_nonnull_ptr_t
+ */
+
+template <typename P>
+struct hb_nonnull_ptr_t
+{
+  typedef hb_remove_pointer<P> T;
+
+  hb_nonnull_ptr_t (T *v_ = nullptr) : v (v_) {}
+  T * operator = (T *v_)   { return v = v_; }
+  T * operator -> () const { return get (); }
+  T & operator * () const  { return *get (); }
+  T ** operator & () const { return &v; }
+  /* Only auto-cast to const types. */
+  template <typename C> operator const C * () const { return get (); }
+  operator const char * () const { return (const char *) get (); }
+  T * get () const { return v ? v : const_cast<T *> (&Null (T)); }
+  T * get_raw () const { return v; }
+
+  T *v;
+};
+
+
+#endif /* HB_NULL_HH */
diff --git a/thirdparty/harfbuzz/src/hb-number-parser.hh b/thirdparty/harfbuzz/src/hb-number-parser.hh
new file mode 100644
index 0000000000..1a9dbba6dd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-number-parser.hh
@@ -0,0 +1,237 @@
+
+#line 1 "hb-number-parser.rl"
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ */
+
+#ifndef HB_NUMBER_PARSER_HH
+#define HB_NUMBER_PARSER_HH
+
+#include "hb.hh"
+
+
+#line 35 "hb-number-parser.hh"
+static const unsigned char _double_parser_trans_keys[] = {
+	0u, 0u, 43u, 57u, 46u, 57u, 48u, 57u, 43u, 57u, 48u, 57u, 48u, 101u, 48u, 57u, 
+	46u, 101u, 0
+};
+
+static const char _double_parser_key_spans[] = {
+	0, 15, 12, 10, 15, 10, 54, 10, 
+	56
+};
+
+static const unsigned char _double_parser_index_offsets[] = {
+	0, 0, 16, 29, 40, 56, 67, 122, 
+	133
+};
+
+static const char _double_parser_indicies[] = {
+	0, 1, 2, 3, 1, 4, 4, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	1, 3, 1, 4, 4, 4, 4, 4, 
+	4, 4, 4, 4, 4, 1, 5, 5, 
+	5, 5, 5, 5, 5, 5, 5, 5, 
+	1, 6, 1, 7, 1, 1, 8, 8, 
+	8, 8, 8, 8, 8, 8, 8, 8, 
+	1, 8, 8, 8, 8, 8, 8, 8, 
+	8, 8, 8, 1, 5, 5, 5, 5, 
+	5, 5, 5, 5, 5, 5, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 9, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 9, 1, 8, 8, 8, 8, 8, 
+	8, 8, 8, 8, 8, 1, 3, 1, 
+	4, 4, 4, 4, 4, 4, 4, 4, 
+	4, 4, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 9, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 1, 1, 1, 9, 1, 0
+};
+
+static const char _double_parser_trans_targs[] = {
+	2, 0, 2, 3, 8, 6, 5, 5, 
+	7, 4
+};
+
+static const char _double_parser_trans_actions[] = {
+	0, 0, 1, 0, 2, 3, 0, 4, 
+	5, 0
+};
+
+static const int double_parser_start = 1;
+static const int double_parser_first_final = 6;
+static const int double_parser_error = 0;
+
+static const int double_parser_en_main = 1;
+
+
+#line 68 "hb-number-parser.rl"
+
+
+/* Works only for n < 512 */
+static inline double
+_pow10 (unsigned exponent)
+{
+  static const double _powers_of_10[] =
+  {
+    1.0e+256,
+    1.0e+128,
+    1.0e+64,
+    1.0e+32,
+    1.0e+16,
+    1.0e+8,
+    10000.,
+    100.,
+    10.
+  };
+  unsigned mask = 1 << (ARRAY_LENGTH (_powers_of_10) - 1);
+  double result = 1;
+  for (const double *power = _powers_of_10; mask; ++power, mask >>= 1)
+    if (exponent & mask) result *= *power;
+  return result;
+}
+
+/* a variant of strtod that also gets end of buffer in its second argument */
+static inline double
+strtod_rl (const char *p, const char **end_ptr /* IN/OUT */)
+{
+  double value = 0;
+  double frac = 0;
+  double frac_count = 0;
+  unsigned exp = 0;
+  bool neg = false, exp_neg = false, exp_overflow = false;
+  const unsigned long long MAX_FRACT = 0xFFFFFFFFFFFFFull; /* 2^52-1 */
+  const unsigned MAX_EXP = 0x7FFu; /* 2^11-1 */
+
+  const char *pe = *end_ptr;
+  while (p < pe && ISSPACE (*p))
+    p++;
+
+  int cs;
+  
+#line 139 "hb-number-parser.hh"
+	{
+	cs = double_parser_start;
+	}
+
+#line 144 "hb-number-parser.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+	if ( cs == 0 )
+		goto _out;
+_resume:
+	_keys = _double_parser_trans_keys + (cs<<1);
+	_inds = _double_parser_indicies + _double_parser_index_offsets[cs];
+
+	_slen = _double_parser_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=(*p) &&
+		(*p) <= _keys[1] ?
+		(*p) - _keys[0] : _slen ];
+
+	cs = _double_parser_trans_targs[_trans];
+
+	if ( _double_parser_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _double_parser_trans_actions[_trans] ) {
+	case 1:
+#line 37 "hb-number-parser.rl"
+	{ neg = true; }
+	break;
+	case 4:
+#line 38 "hb-number-parser.rl"
+	{ exp_neg = true; }
+	break;
+	case 2:
+#line 40 "hb-number-parser.rl"
+	{
+	value = value * 10. + ((*p) - '0');
+}
+	break;
+	case 3:
+#line 43 "hb-number-parser.rl"
+	{
+	if (likely (frac <= MAX_FRACT / 10))
+	{
+	  frac = frac * 10. + ((*p) - '0');
+	  ++frac_count;
+	}
+}
+	break;
+	case 5:
+#line 50 "hb-number-parser.rl"
+	{
+	if (likely (exp * 10 + ((*p) - '0') <= MAX_EXP))
+	  exp = exp * 10 + ((*p) - '0');
+	else
+	  exp_overflow = true;
+}
+	break;
+#line 202 "hb-number-parser.hh"
+	}
+
+_again:
+	if ( cs == 0 )
+		goto _out;
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	_out: {}
+	}
+
+#line 113 "hb-number-parser.rl"
+
+
+  *end_ptr = p;
+
+  if (frac_count) value += frac / _pow10 (frac_count);
+  if (neg) value *= -1.;
+
+  if (unlikely (exp_overflow))
+  {
+    if (value == 0) return value;
+    if (exp_neg)    return neg ? -DBL_MIN : DBL_MIN;
+    else            return neg ? -DBL_MAX : DBL_MAX;
+  }
+
+  if (exp)
+  {
+    if (exp_neg) value /= _pow10 (exp);
+    else         value *= _pow10 (exp);
+  }
+
+  return value;
+}
+
+#endif /* HB_NUMBER_PARSER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-number.cc b/thirdparty/harfbuzz/src/hb-number.cc
new file mode 100644
index 0000000000..6e4f3f7ebd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-number.cc
@@ -0,0 +1,80 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ */
+
+#include "hb.hh"
+#include "hb-machinery.hh"
+#include "hb-number.hh"
+#include "hb-number-parser.hh"
+
+template<typename T, typename Func>
+static bool
+_parse_number (const char **pp, const char *end, T *pv,
+	       bool whole_buffer, Func f)
+{
+  char buf[32];
+  unsigned len = hb_min (ARRAY_LENGTH (buf) - 1, (unsigned) (end - *pp));
+  strncpy (buf, *pp, len);
+  buf[len] = '\0';
+
+  char *p = buf;
+  char *pend = p;
+
+  errno = 0;
+  *pv = f (p, &pend);
+  if (unlikely (errno || p == pend ||
+		/* Check if consumed whole buffer if is requested */
+		(whole_buffer && pend - p != end - *pp)))
+    return false;
+
+  *pp += pend - p;
+  return true;
+}
+
+bool
+hb_parse_int (const char **pp, const char *end, int *pv, bool whole_buffer)
+{
+  return _parse_number<int> (pp, end, pv, whole_buffer,
+			     [] (const char *p, char **end)
+			     { return strtol (p, end, 10); });
+}
+
+bool
+hb_parse_uint (const char **pp, const char *end, unsigned *pv,
+	       bool whole_buffer, int base)
+{
+  return _parse_number<unsigned> (pp, end, pv, whole_buffer,
+				  [base] (const char *p, char **end)
+				  { return strtoul (p, end, base); });
+}
+
+bool
+hb_parse_double (const char **pp, const char *end, double *pv, bool whole_buffer)
+{
+  const char *pend = end;
+  *pv = strtod_rl (*pp, &pend);
+  if (unlikely (*pp == pend)) return false;
+  *pp = pend;
+  return !whole_buffer || end == pend;
+}
diff --git a/thirdparty/harfbuzz/src/hb-number.hh b/thirdparty/harfbuzz/src/hb-number.hh
new file mode 100644
index 0000000000..14d1260aa3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-number.hh
@@ -0,0 +1,41 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ */
+
+#ifndef HB_NUMBER_HH
+#define HB_NUMBER_HH
+
+HB_INTERNAL bool
+hb_parse_int (const char **pp, const char *end, int *pv,
+	      bool whole_buffer = false);
+
+HB_INTERNAL bool
+hb_parse_uint (const char **pp, const char *end, unsigned int *pv,
+	       bool whole_buffer = false, int base = 10);
+
+HB_INTERNAL bool
+hb_parse_double (const char **pp, const char *end, double *pv,
+		 bool whole_buffer = false);
+
+#endif /* HB_NUMBER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-object.hh b/thirdparty/harfbuzz/src/hb-object.hh
new file mode 100644
index 0000000000..39845a70e7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-object.hh
@@ -0,0 +1,342 @@
+/*
+ * Copyright © 2007  Chris Wilson
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Contributor(s):
+ *	Chris Wilson <chris@chris-wilson.co.uk>
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OBJECT_HH
+#define HB_OBJECT_HH
+
+#include "hb.hh"
+#include "hb-atomic.hh"
+#include "hb-mutex.hh"
+#include "hb-vector.hh"
+
+
+/*
+ * Lockable set
+ */
+
+template <typename item_t, typename lock_t>
+struct hb_lockable_set_t
+{
+  hb_vector_t<item_t> items;
+
+  void init () { items.init (); }
+
+  template <typename T>
+  item_t *replace_or_insert (T v, lock_t &l, bool replace)
+  {
+    l.lock ();
+    item_t *item = items.find (v);
+    if (item) {
+      if (replace) {
+	item_t old = *item;
+	*item = v;
+	l.unlock ();
+	old.fini ();
+      }
+      else {
+	item = nullptr;
+	l.unlock ();
+      }
+    } else {
+      item = items.push (v);
+      l.unlock ();
+    }
+    return item;
+  }
+
+  template <typename T>
+  void remove (T v, lock_t &l)
+  {
+    l.lock ();
+    item_t *item = items.find (v);
+    if (item)
+    {
+      item_t old = *item;
+      *item = items[items.length - 1];
+      items.pop ();
+      l.unlock ();
+      old.fini ();
+    } else {
+      l.unlock ();
+    }
+  }
+
+  template <typename T>
+  bool find (T v, item_t *i, lock_t &l)
+  {
+    l.lock ();
+    item_t *item = items.find (v);
+    if (item)
+      *i = *item;
+    l.unlock ();
+    return !!item;
+  }
+
+  template <typename T>
+  item_t *find_or_insert (T v, lock_t &l)
+  {
+    l.lock ();
+    item_t *item = items.find (v);
+    if (!item) {
+      item = items.push (v);
+    }
+    l.unlock ();
+    return item;
+  }
+
+  void fini (lock_t &l)
+  {
+    if (!items.length)
+    {
+      /* No need to lock. */
+      items.fini ();
+      return;
+    }
+    l.lock ();
+    while (items.length)
+    {
+      item_t old = items[items.length - 1];
+      items.pop ();
+      l.unlock ();
+      old.fini ();
+      l.lock ();
+    }
+    items.fini ();
+    l.unlock ();
+  }
+
+};
+
+
+/*
+ * Reference-count.
+ */
+
+#define HB_REFERENCE_COUNT_INERT_VALUE 0
+#define HB_REFERENCE_COUNT_POISON_VALUE -0x0000DEAD
+#define HB_REFERENCE_COUNT_INIT {HB_ATOMIC_INT_INIT (HB_REFERENCE_COUNT_INERT_VALUE)}
+
+struct hb_reference_count_t
+{
+  mutable hb_atomic_int_t ref_count;
+
+  void init (int v = 1) { ref_count.set_relaxed (v); }
+  int get_relaxed () const { return ref_count.get_relaxed (); }
+  int inc () const { return ref_count.inc (); }
+  int dec () const { return ref_count.dec (); }
+  void fini () { ref_count.set_relaxed (HB_REFERENCE_COUNT_POISON_VALUE); }
+
+  bool is_inert () const { return ref_count.get_relaxed () == HB_REFERENCE_COUNT_INERT_VALUE; }
+  bool is_valid () const { return ref_count.get_relaxed () > 0; }
+};
+
+
+/* user_data */
+
+struct hb_user_data_array_t
+{
+  struct hb_user_data_item_t {
+    hb_user_data_key_t *key;
+    void *data;
+    hb_destroy_func_t destroy;
+
+    bool operator == (const hb_user_data_key_t *other_key) const { return key == other_key; }
+    bool operator == (const hb_user_data_item_t &other) const { return key == other.key; }
+
+    void fini () { if (destroy) destroy (data); }
+  };
+
+  hb_mutex_t lock;
+  hb_lockable_set_t<hb_user_data_item_t, hb_mutex_t> items;
+
+  void init () { lock.init (); items.init (); }
+
+  HB_INTERNAL bool set (hb_user_data_key_t *key,
+			void *              data,
+			hb_destroy_func_t   destroy,
+			hb_bool_t           replace);
+
+  HB_INTERNAL void *get (hb_user_data_key_t *key);
+
+  void fini () { items.fini (lock); lock.fini (); }
+};
+
+
+/*
+ * Object header
+ */
+
+struct hb_object_header_t
+{
+  hb_reference_count_t ref_count;
+  mutable hb_atomic_int_t writable;
+  hb_atomic_ptr_t<hb_user_data_array_t> user_data;
+};
+#define HB_OBJECT_HEADER_STATIC \
+	{ \
+	  HB_REFERENCE_COUNT_INIT, \
+	  HB_ATOMIC_INT_INIT (false), \
+	  HB_ATOMIC_PTR_INIT (nullptr) \
+	}
+
+
+/*
+ * Object
+ */
+
+template <typename Type>
+static inline void hb_object_trace (const Type *obj, const char *function)
+{
+  DEBUG_MSG (OBJECT, (void *) obj,
+	     "%s refcount=%d",
+	     function,
+	     obj ? obj->header.ref_count.get_relaxed () : 0);
+}
+
+template <typename Type>
+static inline Type *hb_object_create ()
+{
+  Type *obj = (Type *) calloc (1, sizeof (Type));
+
+  if (unlikely (!obj))
+    return obj;
+
+  hb_object_init (obj);
+  hb_object_trace (obj, HB_FUNC);
+  return obj;
+}
+template <typename Type>
+static inline void hb_object_init (Type *obj)
+{
+  obj->header.ref_count.init ();
+  obj->header.writable.set_relaxed (true);
+  obj->header.user_data.init ();
+}
+template <typename Type>
+static inline bool hb_object_is_inert (const Type *obj)
+{
+  return unlikely (obj->header.ref_count.is_inert ());
+}
+template <typename Type>
+static inline bool hb_object_is_valid (const Type *obj)
+{
+  return likely (obj->header.ref_count.is_valid ());
+}
+template <typename Type>
+static inline bool hb_object_is_immutable (const Type *obj)
+{
+  return !obj->header.writable.get_relaxed ();
+}
+template <typename Type>
+static inline void hb_object_make_immutable (const Type *obj)
+{
+  obj->header.writable.set_relaxed (false);
+}
+template <typename Type>
+static inline Type *hb_object_reference (Type *obj)
+{
+  hb_object_trace (obj, HB_FUNC);
+  if (unlikely (!obj || hb_object_is_inert (obj)))
+    return obj;
+  assert (hb_object_is_valid (obj));
+  obj->header.ref_count.inc ();
+  return obj;
+}
+template <typename Type>
+static inline bool hb_object_destroy (Type *obj)
+{
+  hb_object_trace (obj, HB_FUNC);
+  if (unlikely (!obj || hb_object_is_inert (obj)))
+    return false;
+  assert (hb_object_is_valid (obj));
+  if (obj->header.ref_count.dec () != 1)
+    return false;
+
+  hb_object_fini (obj);
+  return true;
+}
+template <typename Type>
+static inline void hb_object_fini (Type *obj)
+{
+  obj->header.ref_count.fini (); /* Do this before user_data */
+  hb_user_data_array_t *user_data = obj->header.user_data.get ();
+  if (user_data)
+  {
+    user_data->fini ();
+    free (user_data);
+    user_data = nullptr;
+  }
+}
+template <typename Type>
+static inline bool hb_object_set_user_data (Type               *obj,
+					    hb_user_data_key_t *key,
+					    void *              data,
+					    hb_destroy_func_t   destroy,
+					    hb_bool_t           replace)
+{
+  if (unlikely (!obj || hb_object_is_inert (obj)))
+    return false;
+  assert (hb_object_is_valid (obj));
+
+retry:
+  hb_user_data_array_t *user_data = obj->header.user_data.get ();
+  if (unlikely (!user_data))
+  {
+    user_data = (hb_user_data_array_t *) calloc (sizeof (hb_user_data_array_t), 1);
+    if (unlikely (!user_data))
+      return false;
+    user_data->init ();
+    if (unlikely (!obj->header.user_data.cmpexch (nullptr, user_data)))
+    {
+      user_data->fini ();
+      free (user_data);
+      goto retry;
+    }
+  }
+
+  return user_data->set (key, data, destroy, replace);
+}
+
+template <typename Type>
+static inline void *hb_object_get_user_data (Type               *obj,
+					     hb_user_data_key_t *key)
+{
+  if (unlikely (!obj || hb_object_is_inert (obj)))
+    return nullptr;
+  assert (hb_object_is_valid (obj));
+  hb_user_data_array_t *user_data = obj->header.user_data.get ();
+  if (!user_data)
+    return nullptr;
+  return user_data->get (key);
+}
+
+
+#endif /* HB_OBJECT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-open-file.hh b/thirdparty/harfbuzz/src/hb-open-file.hh
new file mode 100644
index 0000000000..ac13dd23c3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-open-file.hh
@@ -0,0 +1,521 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OPEN_FILE_HH
+#define HB_OPEN_FILE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-head-table.hh"
+
+
+namespace OT {
+
+
+/*
+ *
+ * The OpenType Font File
+ *
+ */
+
+
+/*
+ * Organization of an OpenType Font
+ */
+
+struct OpenTypeFontFile;
+struct OffsetTable;
+struct TTCHeader;
+
+
+typedef struct TableRecord
+{
+  int cmp (Tag t) const { return -t.cmp (tag); }
+
+  HB_INTERNAL static int cmp (const void *pa, const void *pb)
+  {
+    const TableRecord *a = (const TableRecord *) pa;
+    const TableRecord *b = (const TableRecord *) pb;
+    return b->cmp (a->tag);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  Tag		tag;		/* 4-byte identifier. */
+  CheckSum	checkSum;	/* CheckSum for this table. */
+  Offset32	offset;		/* Offset from beginning of TrueType font
+				 * file. */
+  HBUINT32	length;		/* Length of this table. */
+  public:
+  DEFINE_SIZE_STATIC (16);
+} OpenTypeTable;
+
+typedef struct OffsetTable
+{
+  friend struct OpenTypeFontFile;
+
+  unsigned int get_table_count () const { return tables.len; }
+  const TableRecord& get_table (unsigned int i) const
+  { return tables[i]; }
+  unsigned int get_table_tags (unsigned int  start_offset,
+			       unsigned int *table_count, /* IN/OUT */
+			       hb_tag_t     *table_tags /* OUT */) const
+  {
+    if (table_count)
+    {
+      + tables.sub_array (start_offset, table_count)
+      | hb_map (&TableRecord::tag)
+      | hb_sink (hb_array (table_tags, *table_count))
+      ;
+    }
+    return tables.len;
+  }
+  bool find_table_index (hb_tag_t tag, unsigned int *table_index) const
+  {
+    Tag t;
+    t = tag;
+    return tables.bfind (t, table_index, HB_BFIND_NOT_FOUND_STORE, Index::NOT_FOUND_INDEX);
+  }
+  const TableRecord& get_table_by_tag (hb_tag_t tag) const
+  {
+    unsigned int table_index;
+    find_table_index (tag, &table_index);
+    return get_table (table_index);
+  }
+
+  public:
+
+  template <typename item_t>
+  bool serialize (hb_serialize_context_t *c,
+		  hb_tag_t sfnt_tag,
+		  hb_array_t<item_t> items)
+  {
+    TRACE_SERIALIZE (this);
+    /* Alloc 12 for the OTHeader. */
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    /* Write sfntVersion (bytes 0..3). */
+    sfnt_version = sfnt_tag;
+    /* Take space for numTables, searchRange, entrySelector, RangeShift
+     * and the TableRecords themselves.  */
+    if (unlikely (!tables.serialize (c, items.length))) return_trace (false);
+
+    const char *dir_end = (const char *) c->head;
+    HBUINT32 *checksum_adjustment = nullptr;
+
+    /* Write OffsetTables, alloc for and write actual table blobs. */
+    for (unsigned int i = 0; i < tables.len; i++)
+    {
+      TableRecord &rec = tables.arrayZ[i];
+      hb_blob_t *blob = items[i].blob;
+      rec.tag = items[i].tag;
+      rec.length = blob->length;
+      rec.offset.serialize (c, this);
+
+      /* Allocate room for the table and copy it. */
+      char *start = (char *) c->allocate_size<void> (rec.length);
+      if (unlikely (!start)) return false;
+
+      if (likely (rec.length))
+	memcpy (start, blob->data, rec.length);
+
+      /* 4-byte alignment. */
+      c->align (4);
+      const char *end = (const char *) c->head;
+
+      if (items[i].tag == HB_OT_TAG_head &&
+	  (unsigned) (end - start) >= head::static_size)
+      {
+	head *h = (head *) start;
+	checksum_adjustment = &h->checkSumAdjustment;
+	*checksum_adjustment = 0;
+      }
+
+      rec.checkSum.set_for_data (start, end - start);
+    }
+
+    tables.qsort ();
+
+    if (checksum_adjustment)
+    {
+      CheckSum checksum;
+
+      /* The following line is a slower version of the following block. */
+      //checksum.set_for_data (this, (const char *) c->head - (const char *) this);
+      checksum.set_for_data (this, dir_end - (const char *) this);
+      for (unsigned int i = 0; i < items.length; i++)
+      {
+	TableRecord &rec = tables.arrayZ[i];
+	checksum = checksum + rec.checkSum;
+      }
+
+      *checksum_adjustment = 0xB1B0AFBAu - checksum;
+    }
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && tables.sanitize (c));
+  }
+
+  protected:
+  Tag		sfnt_version;	/* '\0\001\0\00' if TrueType / 'OTTO' if CFF */
+  BinSearchArrayOf<TableRecord>
+		tables;
+  public:
+  DEFINE_SIZE_ARRAY (12, tables);
+} OpenTypeFontFace;
+
+
+/*
+ * TrueType Collections
+ */
+
+struct TTCHeaderVersion1
+{
+  friend struct TTCHeader;
+
+  unsigned int get_face_count () const { return table.len; }
+  const OpenTypeFontFace& get_face (unsigned int i) const { return this+table[i]; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (table.sanitize (c, this));
+  }
+
+  protected:
+  Tag		ttcTag;		/* TrueType Collection ID string: 'ttcf' */
+  FixedVersion<>version;	/* Version of the TTC Header (1.0),
+				 * 0x00010000u */
+  LArrayOf<LOffsetTo<OffsetTable>>
+		table;		/* Array of offsets to the OffsetTable for each font
+				 * from the beginning of the file */
+  public:
+  DEFINE_SIZE_ARRAY (12, table);
+};
+
+struct TTCHeader
+{
+  friend struct OpenTypeFontFile;
+
+  private:
+
+  unsigned int get_face_count () const
+  {
+    switch (u.header.version.major) {
+    case 2: /* version 2 is compatible with version 1 */
+    case 1: return u.version1.get_face_count ();
+    default:return 0;
+    }
+  }
+  const OpenTypeFontFace& get_face (unsigned int i) const
+  {
+    switch (u.header.version.major) {
+    case 2: /* version 2 is compatible with version 1 */
+    case 1: return u.version1.get_face (i);
+    default:return Null (OpenTypeFontFace);
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!u.header.version.sanitize (c))) return_trace (false);
+    switch (u.header.version.major) {
+    case 2: /* version 2 is compatible with version 1 */
+    case 1: return_trace (u.version1.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  struct {
+  Tag		ttcTag;		/* TrueType Collection ID string: 'ttcf' */
+  FixedVersion<>version;	/* Version of the TTC Header (1.0 or 2.0),
+				 * 0x00010000u or 0x00020000u */
+  }			header;
+  TTCHeaderVersion1	version1;
+  } u;
+};
+
+/*
+ * Mac Resource Fork
+ *
+ * http://mirror.informatimago.com/next/developer.apple.com/documentation/mac/MoreToolbox/MoreToolbox-99.html
+ */
+
+struct ResourceRecord
+{
+  const OpenTypeFontFace & get_face (const void *data_base) const
+  { return * reinterpret_cast<const OpenTypeFontFace *> ((data_base+offset).arrayZ); }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const void *data_base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  offset.sanitize (c, data_base) &&
+		  get_face (data_base).sanitize (c));
+  }
+
+  protected:
+  HBUINT16	id;		/* Resource ID. */
+  HBINT16	nameOffset;	/* Offset from beginning of resource name list
+				 * to resource name, -1 means there is none. */
+  HBUINT8	attrs;		/* Resource attributes */
+  NNOffsetTo<LArrayOf<HBUINT8>, HBUINT24>
+		offset;		/* Offset from beginning of data block to
+				 * data for this resource */
+  HBUINT32	reserved;	/* Reserved for handle to resource */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+#define HB_TAG_sfnt HB_TAG ('s','f','n','t')
+
+struct ResourceTypeRecord
+{
+  unsigned int get_resource_count () const
+  { return tag == HB_TAG_sfnt ? resCountM1 + 1 : 0; }
+
+  bool is_sfnt () const { return tag == HB_TAG_sfnt; }
+
+  const ResourceRecord& get_resource_record (unsigned int i,
+					     const void *type_base) const
+  { return (type_base+resourcesZ).as_array (get_resource_count ())[i]; }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const void *type_base,
+		 const void *data_base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  resourcesZ.sanitize (c, type_base,
+				       get_resource_count (),
+				       data_base));
+  }
+
+  protected:
+  Tag		tag;		/* Resource type. */
+  HBUINT16	resCountM1;	/* Number of resources minus 1. */
+  NNOffsetTo<UnsizedArrayOf<ResourceRecord>>
+		resourcesZ;	/* Offset from beginning of resource type list
+				 * to reference item list for this type. */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct ResourceMap
+{
+  unsigned int get_face_count () const
+  {
+    unsigned int count = get_type_count ();
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const ResourceTypeRecord& type = get_type_record (i);
+      if (type.is_sfnt ())
+	return type.get_resource_count ();
+    }
+    return 0;
+  }
+
+  const OpenTypeFontFace& get_face (unsigned int idx,
+				    const void *data_base) const
+  {
+    unsigned int count = get_type_count ();
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const ResourceTypeRecord& type = get_type_record (i);
+      /* The check for idx < count is here because ResourceRecord is NOT null-safe.
+       * Because an offset of 0 there does NOT mean null. */
+      if (type.is_sfnt () && idx < type.get_resource_count ())
+	return type.get_resource_record (idx, &(this+typeList)).get_face (data_base);
+    }
+    return Null (OpenTypeFontFace);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *data_base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  typeList.sanitize (c, this,
+				     &(this+typeList),
+				     data_base));
+  }
+
+  private:
+  unsigned int get_type_count () const { return (this+typeList).lenM1 + 1; }
+
+  const ResourceTypeRecord& get_type_record (unsigned int i) const
+  { return (this+typeList)[i]; }
+
+  protected:
+  HBUINT8	reserved0[16];	/* Reserved for copy of resource header */
+  HBUINT32	reserved1;	/* Reserved for handle to next resource map */
+  HBUINT16	resreved2;	/* Reserved for file reference number */
+  HBUINT16	attrs;		/* Resource fork attribute */
+  NNOffsetTo<ArrayOfM1<ResourceTypeRecord>>
+		typeList;	/* Offset from beginning of map to
+				 * resource type list */
+  Offset16	nameList;	/* Offset from beginning of map to
+				 * resource name list */
+  public:
+  DEFINE_SIZE_STATIC (28);
+};
+
+struct ResourceForkHeader
+{
+  unsigned int get_face_count () const
+  { return (this+map).get_face_count (); }
+
+  const OpenTypeFontFace& get_face (unsigned int idx,
+				    unsigned int *base_offset = nullptr) const
+  {
+    const OpenTypeFontFace &face = (this+map).get_face (idx, &(this+data));
+    if (base_offset)
+      *base_offset = (const char *) &face - (const char *) this;
+    return face;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  data.sanitize (c, this, dataLen) &&
+		  map.sanitize (c, this, &(this+data)));
+  }
+
+  protected:
+  LNNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		data;		/* Offset from beginning of resource fork
+				 * to resource data */
+  LNNOffsetTo<ResourceMap >
+		map;		/* Offset from beginning of resource fork
+				 * to resource map */
+  HBUINT32	dataLen;	/* Length of resource data */
+  HBUINT32	mapLen;		/* Length of resource map */
+  public:
+  DEFINE_SIZE_STATIC (16);
+};
+
+/*
+ * OpenType Font File
+ */
+
+struct OpenTypeFontFile
+{
+  enum {
+    CFFTag		= HB_TAG ('O','T','T','O'), /* OpenType with Postscript outlines */
+    TrueTypeTag		= HB_TAG ( 0 , 1 , 0 , 0 ), /* OpenType with TrueType outlines */
+    TTCTag		= HB_TAG ('t','t','c','f'), /* TrueType Collection */
+    DFontTag		= HB_TAG ( 0 , 0 , 1 , 0 ), /* DFont Mac Resource Fork */
+    TrueTag		= HB_TAG ('t','r','u','e'), /* Obsolete Apple TrueType */
+    Typ1Tag		= HB_TAG ('t','y','p','1')  /* Obsolete Apple Type1 font in SFNT container */
+  };
+
+  hb_tag_t get_tag () const { return u.tag; }
+
+  unsigned int get_face_count () const
+  {
+    switch (u.tag) {
+    case CFFTag:	/* All the non-collection tags */
+    case TrueTag:
+    case Typ1Tag:
+    case TrueTypeTag:	return 1;
+    case TTCTag:	return u.ttcHeader.get_face_count ();
+    case DFontTag:	return u.rfHeader.get_face_count ();
+    default:		return 0;
+    }
+  }
+  const OpenTypeFontFace& get_face (unsigned int i, unsigned int *base_offset = nullptr) const
+  {
+    if (base_offset)
+      *base_offset = 0;
+    switch (u.tag) {
+    /* Note: for non-collection SFNT data we ignore index.  This is because
+     * Apple dfont container is a container of SFNT's.  So each SFNT is a
+     * non-TTC, but the index is more than zero. */
+    case CFFTag:	/* All the non-collection tags */
+    case TrueTag:
+    case Typ1Tag:
+    case TrueTypeTag:	return u.fontFace;
+    case TTCTag:	return u.ttcHeader.get_face (i);
+    case DFontTag:	return u.rfHeader.get_face (i, base_offset);
+    default:		return Null (OpenTypeFontFace);
+    }
+  }
+
+  template <typename item_t>
+  bool serialize_single (hb_serialize_context_t *c,
+			 hb_tag_t sfnt_tag,
+			 hb_array_t<item_t> items)
+  {
+    TRACE_SERIALIZE (this);
+    assert (sfnt_tag != TTCTag);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    return_trace (u.fontFace.serialize (c, sfnt_tag, items));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!u.tag.sanitize (c))) return_trace (false);
+    switch (u.tag) {
+    case CFFTag:	/* All the non-collection tags */
+    case TrueTag:
+    case Typ1Tag:
+    case TrueTypeTag:	return_trace (u.fontFace.sanitize (c));
+    case TTCTag:	return_trace (u.ttcHeader.sanitize (c));
+    case DFontTag:	return_trace (u.rfHeader.sanitize (c));
+    default:		return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  Tag			tag;		/* 4-byte identifier. */
+  OpenTypeFontFace	fontFace;
+  TTCHeader		ttcHeader;
+  ResourceForkHeader	rfHeader;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (4, tag);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OPEN_FILE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-open-type.hh b/thirdparty/harfbuzz/src/hb-open-type.hh
new file mode 100644
index 0000000000..50558cf8d3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-open-type.hh
@@ -0,0 +1,1078 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OPEN_TYPE_HH
+#define HB_OPEN_TYPE_HH
+
+#include "hb.hh"
+#include "hb-blob.hh"
+#include "hb-face.hh"
+#include "hb-machinery.hh"
+#include "hb-subset.hh"
+
+
+namespace OT {
+
+
+/*
+ *
+ * The OpenType Font File: Data Types
+ */
+
+
+/* "The following data types are used in the OpenType font file.
+ *  All OpenType fonts use Motorola-style byte ordering (Big Endian):" */
+
+/*
+ * Int types
+ */
+
+/* Integer types in big-endian order and no alignment requirement */
+template <typename Type, unsigned int Size>
+struct IntType
+{
+  typedef Type type;
+  typedef hb_conditional<hb_is_signed (Type), signed, unsigned> wide_type;
+
+  IntType& operator = (wide_type i) { v = i; return *this; }
+  operator wide_type () const { return v; }
+  bool operator == (const IntType &o) const { return (Type) v == (Type) o.v; }
+  bool operator != (const IntType &o) const { return !(*this == o); }
+
+  IntType& operator += (unsigned count) { *this = *this + count; return *this; }
+  IntType& operator -= (unsigned count) { *this = *this - count; return *this; }
+  IntType& operator ++ () { *this += 1; return *this; }
+  IntType& operator -- () { *this -= 1; return *this; }
+  IntType operator ++ (int) { IntType c (*this); ++*this; return c; }
+  IntType operator -- (int) { IntType c (*this); --*this; return c; }
+
+  HB_INTERNAL static int cmp (const IntType *a, const IntType *b)
+  { return b->cmp (*a); }
+  HB_INTERNAL static int cmp (const void *a, const void *b)
+  {
+    IntType *pa = (IntType *) a;
+    IntType *pb = (IntType *) b;
+
+    return pb->cmp (*pa);
+  }
+  template <typename Type2>
+  int cmp (Type2 a) const
+  {
+    Type b = v;
+    if (sizeof (Type) < sizeof (int) && sizeof (Type2) < sizeof (int))
+      return (int) a - (int) b;
+    else
+      return a < b ? -1 : a == b ? 0 : +1;
+  }
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+  protected:
+  BEInt<Type, Size> v;
+  public:
+  DEFINE_SIZE_STATIC (Size);
+};
+
+typedef IntType<uint8_t,  1> HBUINT8;	/* 8-bit unsigned integer. */
+typedef IntType<int8_t,   1> HBINT8;	/* 8-bit signed integer. */
+typedef IntType<uint16_t, 2> HBUINT16;	/* 16-bit unsigned integer. */
+typedef IntType<int16_t,  2> HBINT16;	/* 16-bit signed integer. */
+typedef IntType<uint32_t, 4> HBUINT32;	/* 32-bit unsigned integer. */
+typedef IntType<int32_t,  4> HBINT32;	/* 32-bit signed integer. */
+/* Note: we cannot defined a signed HBINT24 because there's no corresponding C type.
+ * Works for unsigned, but not signed, since we rely on compiler for sign-extension. */
+typedef IntType<uint32_t, 3> HBUINT24;	/* 24-bit unsigned integer. */
+
+/* 16-bit signed integer (HBINT16) that describes a quantity in FUnits. */
+typedef HBINT16 FWORD;
+
+/* 32-bit signed integer (HBINT32) that describes a quantity in FUnits. */
+typedef HBINT32 FWORD32;
+
+/* 16-bit unsigned integer (HBUINT16) that describes a quantity in FUnits. */
+typedef HBUINT16 UFWORD;
+
+/* 16-bit signed fixed number with the low 14 bits of fraction (2.14). */
+struct F2DOT14 : HBINT16
+{
+  F2DOT14& operator = (uint16_t i ) { HBINT16::operator= (i); return *this; }
+  // 16384 means 1<<14
+  float to_float () const  { return ((int32_t) v) / 16384.f; }
+  void set_float (float f) { v = roundf (f * 16384.f); }
+  public:
+  DEFINE_SIZE_STATIC (2);
+};
+
+/* 32-bit signed fixed-point number (16.16). */
+struct HBFixed : HBINT32
+{
+  HBFixed& operator = (uint32_t i) { HBINT32::operator= (i); return *this; }
+  // 65536 means 1<<16
+  float to_float () const  { return ((int32_t) v) / 65536.f; }
+  void set_float (float f) { v = roundf (f * 65536.f); }
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+/* Date represented in number of seconds since 12:00 midnight, January 1,
+ * 1904. The value is represented as a signed 64-bit integer. */
+struct LONGDATETIME
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+  protected:
+  HBINT32 major;
+  HBUINT32 minor;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+/* Array of four uint8s (length = 32 bits) used to identify a script, language
+ * system, feature, or baseline */
+struct Tag : HBUINT32
+{
+  Tag& operator = (hb_tag_t i) { HBUINT32::operator= (i); return *this; }
+  /* What the char* converters return is NOT nul-terminated.  Print using "%.4s" */
+  operator const char* () const { return reinterpret_cast<const char *> (&this->v); }
+  operator char* ()             { return reinterpret_cast<char *> (&this->v); }
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+/* Glyph index number, same as uint16 (length = 16 bits) */
+struct HBGlyphID : HBUINT16
+{
+  HBGlyphID& operator = (uint16_t i) { HBUINT16::operator= (i); return *this; }
+};
+
+/* Script/language-system/feature index */
+struct Index : HBUINT16 {
+  static constexpr unsigned NOT_FOUND_INDEX = 0xFFFFu;
+  Index& operator = (uint16_t i) { HBUINT16::operator= (i); return *this; }
+};
+DECLARE_NULL_NAMESPACE_BYTES (OT, Index);
+
+typedef Index NameID;
+
+/* Offset, Null offset = 0 */
+template <typename Type, bool has_null=true>
+struct Offset : Type
+{
+  Offset& operator = (typename Type::type i) { Type::operator= (i); return *this; }
+
+  typedef Type type;
+
+  bool is_null () const { return has_null && 0 == *this; }
+
+  void *serialize (hb_serialize_context_t *c, const void *base)
+  {
+    void *t = c->start_embed<void> ();
+    c->check_assign (*this, (unsigned) ((char *) t - (char *) base));
+    return t;
+  }
+
+  public:
+  DEFINE_SIZE_STATIC (sizeof (Type));
+};
+
+typedef Offset<HBUINT16> Offset16;
+typedef Offset<HBUINT32> Offset32;
+
+
+/* CheckSum */
+struct CheckSum : HBUINT32
+{
+  CheckSum& operator = (uint32_t i) { HBUINT32::operator= (i); return *this; }
+
+  /* This is reference implementation from the spec. */
+  static uint32_t CalcTableChecksum (const HBUINT32 *Table, uint32_t Length)
+  {
+    uint32_t Sum = 0L;
+    assert (0 == (Length & 3));
+    const HBUINT32 *EndPtr = Table + Length / HBUINT32::static_size;
+
+    while (Table < EndPtr)
+      Sum += *Table++;
+    return Sum;
+  }
+
+  /* Note: data should be 4byte aligned and have 4byte padding at the end. */
+  void set_for_data (const void *data, unsigned int length)
+  { *this = CalcTableChecksum ((const HBUINT32 *) data, length); }
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+
+/*
+ * Version Numbers
+ */
+
+template <typename FixedType=HBUINT16>
+struct FixedVersion
+{
+  uint32_t to_int () const { return (major << (sizeof (FixedType) * 8)) + minor; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  FixedType major;
+  FixedType minor;
+  public:
+  DEFINE_SIZE_STATIC (2 * sizeof (FixedType));
+};
+
+
+/*
+ * Template subclasses of Offset that do the dereferencing.
+ * Use: (base+offset)
+ */
+
+template <typename Type, bool has_null>
+struct _hb_has_null
+{
+  static const Type *get_null () { return nullptr; }
+  static Type *get_crap ()       { return nullptr; }
+};
+template <typename Type>
+struct _hb_has_null<Type, true>
+{
+  static const Type *get_null () { return &Null (Type); }
+  static       Type *get_crap () { return &Crap (Type); }
+};
+
+template <typename Type, typename OffsetType=HBUINT16, bool has_null=true>
+struct OffsetTo : Offset<OffsetType, has_null>
+{
+  HB_DELETE_COPY_ASSIGN (OffsetTo);
+  OffsetTo () = default;
+
+  OffsetTo& operator = (typename OffsetType::type i) { OffsetType::operator= (i); return *this; }
+
+  const Type& operator () (const void *base) const
+  {
+    if (unlikely (this->is_null ())) return *_hb_has_null<Type, has_null>::get_null ();
+    return StructAtOffset<const Type> (base, *this);
+  }
+  Type& operator () (void *base) const
+  {
+    if (unlikely (this->is_null ())) return *_hb_has_null<Type, has_null>::get_crap ();
+    return StructAtOffset<Type> (base, *this);
+  }
+
+  template <typename Base,
+	    hb_enable_if (hb_is_convertible (const Base, const void *))>
+  friend const Type& operator + (const Base &base, const OffsetTo &offset) { return offset ((const void *) base); }
+  template <typename Base,
+	    hb_enable_if (hb_is_convertible (const Base, const void *))>
+  friend const Type& operator + (const OffsetTo &offset, const Base &base) { return offset ((const void *) base); }
+  template <typename Base,
+	    hb_enable_if (hb_is_convertible (Base, void *))>
+  friend Type& operator + (Base &&base, OffsetTo &offset) { return offset ((void *) base); }
+  template <typename Base,
+	    hb_enable_if (hb_is_convertible (Base, void *))>
+  friend Type& operator + (OffsetTo &offset, Base &&base) { return offset ((void *) base); }
+
+  Type& serialize (hb_serialize_context_t *c, const void *base)
+  {
+    return * (Type *) Offset<OffsetType>::serialize (c, base);
+  }
+
+  template <typename ...Ts>
+  bool serialize_subset (hb_subset_context_t *c, const OffsetTo& src,
+			 const void *src_base, Ts&&... ds)
+  {
+    *this = 0;
+    if (src.is_null ())
+      return false;
+
+    auto *s = c->serializer;
+
+    s->push ();
+
+    bool ret = c->dispatch (src_base+src, hb_forward<Ts> (ds)...);
+
+    if (ret || !has_null)
+      s->add_link (*this, s->pop_pack ());
+    else
+      s->pop_discard ();
+
+    return ret;
+  }
+
+  /* TODO: Somehow merge this with previous function into a serialize_dispatch(). */
+  /* Workaround clang bug: https://bugs.llvm.org/show_bug.cgi?id=23029
+   * Can't compile: whence = hb_serialize_context_t::Head followed by Ts&&...
+   */
+  template <typename ...Ts>
+  bool serialize_copy (hb_serialize_context_t *c, const OffsetTo& src,
+		       const void *src_base, unsigned dst_bias,
+		       hb_serialize_context_t::whence_t whence,
+		       Ts&&... ds)
+  {
+    *this = 0;
+    if (src.is_null ())
+      return false;
+
+    c->push ();
+
+    bool ret = c->copy (src_base+src, hb_forward<Ts> (ds)...);
+
+    c->add_link (*this, c->pop_pack (), whence, dst_bias);
+
+    return ret;
+  }
+
+  bool serialize_copy (hb_serialize_context_t *c, const OffsetTo& src,
+		       const void *src_base, unsigned dst_bias = 0)
+  { return serialize_copy (c, src, src_base, dst_bias, hb_serialize_context_t::Head); }
+
+  bool sanitize_shallow (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this))) return_trace (false);
+    if (unlikely (this->is_null ())) return_trace (true);
+    if (unlikely (!c->check_range (base, *this))) return_trace (false);
+    return_trace (true);
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, const void *base, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (sanitize_shallow (c, base) &&
+		  (this->is_null () ||
+		   c->dispatch (StructAtOffset<Type> (base, *this), hb_forward<Ts> (ds)...) ||
+		   neuter (c)));
+  }
+
+  /* Set the offset to Null */
+  bool neuter (hb_sanitize_context_t *c) const
+  {
+    if (!has_null) return false;
+    return c->try_set (this, 0);
+  }
+  DEFINE_SIZE_STATIC (sizeof (OffsetType));
+};
+/* Partial specializations. */
+template <typename Type, bool has_null=true>
+using LOffsetTo = OffsetTo<Type, HBUINT32, has_null>;
+template <typename Type, typename OffsetType=HBUINT16>
+using NNOffsetTo = OffsetTo<Type, OffsetType, false>;
+template <typename Type>
+using LNNOffsetTo = LOffsetTo<Type, false>;
+
+
+/*
+ * Array Types
+ */
+
+template <typename Type>
+struct UnsizedArrayOf
+{
+  typedef Type item_t;
+  static constexpr unsigned item_size = hb_static_size (Type);
+
+  HB_DELETE_CREATE_COPY_ASSIGN (UnsizedArrayOf);
+
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    const Type *p = &arrayZ[i];
+    if (unlikely (p < arrayZ)) return Null (Type); /* Overflowed. */
+    return *p;
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    Type *p = &arrayZ[i];
+    if (unlikely (p < arrayZ)) return Crap (Type); /* Overflowed. */
+    return *p;
+  }
+
+  unsigned int get_size (unsigned int len) const
+  { return len * Type::static_size; }
+
+  template <typename T> operator T * () { return arrayZ; }
+  template <typename T> operator const T * () const { return arrayZ; }
+  hb_array_t<Type> as_array (unsigned int len)
+  { return hb_array (arrayZ, len); }
+  hb_array_t<const Type> as_array (unsigned int len) const
+  { return hb_array (arrayZ, len); }
+  operator hb_array_t<      Type> ()       { return as_array (); }
+  operator hb_array_t<const Type> () const { return as_array (); }
+
+  template <typename T>
+  Type &lsearch (unsigned int len, const T &x, Type &not_found = Crap (Type))
+  { return *as_array (len).lsearch (x, &not_found); }
+  template <typename T>
+  const Type &lsearch (unsigned int len, const T &x, const Type &not_found = Null (Type)) const
+  { return *as_array (len).lsearch (x, &not_found); }
+  template <typename T>
+  bool lfind (unsigned int len, const T &x, unsigned *pos = nullptr) const
+  { return as_array (len).lfind (x, pos); }
+
+  void qsort (unsigned int len, unsigned int start = 0, unsigned int end = (unsigned int) -1)
+  { as_array (len).qsort (start, end); }
+
+  bool serialize (hb_serialize_context_t *c, unsigned int items_len)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend (*this, items_len))) return_trace (false);
+    return_trace (true);
+  }
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, Type))>
+  bool serialize (hb_serialize_context_t *c, Iterator items)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned count = items.len ();
+    if (unlikely (!serialize (c, count))) return_trace (false);
+    /* TODO Umm. Just exhaust the iterator instead?  Being extra
+     * cautious right now.. */
+    for (unsigned i = 0; i < count; i++, ++items)
+      arrayZ[i] = *items;
+    return_trace (true);
+  }
+
+  UnsizedArrayOf* copy (hb_serialize_context_t *c, unsigned count) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->start_embed (this);
+    if (unlikely (!as_array (count).copy (c))) return_trace (nullptr);
+    return_trace (out);
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, unsigned int count, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!sanitize_shallow (c, count))) return_trace (false);
+    if (!sizeof... (Ts) && hb_is_trivially_copyable (Type)) return_trace (true);
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!c->dispatch (arrayZ[i], hb_forward<Ts> (ds)...)))
+	return_trace (false);
+    return_trace (true);
+  }
+
+  bool sanitize_shallow (hb_sanitize_context_t *c, unsigned int count) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_array (arrayZ, count));
+  }
+
+  public:
+  Type		arrayZ[HB_VAR_ARRAY];
+  public:
+  DEFINE_SIZE_UNBOUNDED (0);
+};
+
+/* Unsized array of offset's */
+template <typename Type, typename OffsetType, bool has_null=true>
+using UnsizedOffsetArrayOf = UnsizedArrayOf<OffsetTo<Type, OffsetType, has_null>>;
+
+/* Unsized array of offsets relative to the beginning of the array itself. */
+template <typename Type, typename OffsetType, bool has_null=true>
+struct UnsizedOffsetListOf : UnsizedOffsetArrayOf<Type, OffsetType, has_null>
+{
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    const OffsetTo<Type, OffsetType, has_null> *p = &this->arrayZ[i];
+    if (unlikely (p < this->arrayZ)) return Null (Type); /* Overflowed. */
+    return this+*p;
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    const OffsetTo<Type, OffsetType, has_null> *p = &this->arrayZ[i];
+    if (unlikely (p < this->arrayZ)) return Crap (Type); /* Overflowed. */
+    return this+*p;
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, unsigned int count, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace ((UnsizedOffsetArrayOf<Type, OffsetType, has_null>
+		   ::sanitize (c, count, this, hb_forward<Ts> (ds)...)));
+  }
+};
+
+/* An array with sorted elements.  Supports binary searching. */
+template <typename Type>
+struct SortedUnsizedArrayOf : UnsizedArrayOf<Type>
+{
+  hb_sorted_array_t<Type> as_array (unsigned int len)
+  { return hb_sorted_array (this->arrayZ, len); }
+  hb_sorted_array_t<const Type> as_array (unsigned int len) const
+  { return hb_sorted_array (this->arrayZ, len); }
+  operator hb_sorted_array_t<Type> ()             { return as_array (); }
+  operator hb_sorted_array_t<const Type> () const { return as_array (); }
+
+  template <typename T>
+  Type &bsearch (unsigned int len, const T &x, Type &not_found = Crap (Type))
+  { return *as_array (len).bsearch (x, &not_found); }
+  template <typename T>
+  const Type &bsearch (unsigned int len, const T &x, const Type &not_found = Null (Type)) const
+  { return *as_array (len).bsearch (x, &not_found); }
+  template <typename T>
+  bool bfind (unsigned int len, const T &x, unsigned int *i = nullptr,
+	      hb_bfind_not_found_t not_found = HB_BFIND_NOT_FOUND_DONT_STORE,
+	      unsigned int to_store = (unsigned int) -1) const
+  { return as_array (len).bfind (x, i, not_found, to_store); }
+};
+
+
+/* An array with a number of elements. */
+template <typename Type, typename LenType=HBUINT16>
+struct ArrayOf
+{
+  typedef Type item_t;
+  static constexpr unsigned item_size = hb_static_size (Type);
+
+  HB_DELETE_CREATE_COPY_ASSIGN (ArrayOf);
+
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= len)) return Null (Type);
+    return arrayZ[i];
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= len)) return Crap (Type);
+    return arrayZ[i];
+  }
+
+  unsigned int get_size () const
+  { return len.static_size + len * Type::static_size; }
+
+  explicit operator bool () const { return len; }
+
+  void pop () { len--; }
+
+  hb_array_t<      Type> as_array ()       { return hb_array (arrayZ, len); }
+  hb_array_t<const Type> as_array () const { return hb_array (arrayZ, len); }
+
+  /* Iterator. */
+  typedef hb_array_t<const Type>   iter_t;
+  typedef hb_array_t<      Type> writer_t;
+    iter_t   iter () const { return as_array (); }
+  writer_t writer ()       { return as_array (); }
+  operator   iter_t () const { return   iter (); }
+  operator writer_t ()       { return writer (); }
+
+  hb_array_t<const Type> sub_array (unsigned int start_offset, unsigned int count) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<const Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<Type> sub_array (unsigned int start_offset, unsigned int count)
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */)
+  { return as_array ().sub_array (start_offset, count); }
+
+  hb_success_t serialize (hb_serialize_context_t *c, unsigned items_len)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    c->check_assign (len, items_len);
+    if (unlikely (!c->extend (*this))) return_trace (false);
+    return_trace (true);
+  }
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, Type))>
+  hb_success_t serialize (hb_serialize_context_t *c, Iterator items)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned count = items.len ();
+    if (unlikely (!serialize (c, count))) return_trace (false);
+    /* TODO Umm. Just exhaust the iterator instead?  Being extra
+     * cautious right now.. */
+    for (unsigned i = 0; i < count; i++, ++items)
+      arrayZ[i] = *items;
+    return_trace (true);
+  }
+
+  Type* serialize_append (hb_serialize_context_t *c)
+  {
+    TRACE_SERIALIZE (this);
+    len++;
+    if (unlikely (!len || !c->extend (*this)))
+    {
+      len--;
+      return_trace (nullptr);
+    }
+    return_trace (&arrayZ[len - 1]);
+  }
+
+  ArrayOf* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->start_embed (this);
+    if (unlikely (!c->extend_min (out))) return_trace (nullptr);
+    c->check_assign (out->len, len);
+    if (unlikely (!as_array ().copy (c))) return_trace (nullptr);
+    return_trace (out);
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!sanitize_shallow (c))) return_trace (false);
+    if (!sizeof... (Ts) && hb_is_trivially_copyable (Type)) return_trace (true);
+    unsigned int count = len;
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!c->dispatch (arrayZ[i], hb_forward<Ts> (ds)...)))
+	return_trace (false);
+    return_trace (true);
+  }
+
+  template <typename T>
+  Type &lsearch (const T &x, Type &not_found = Crap (Type))
+  { return *as_array ().lsearch (x, &not_found); }
+  template <typename T>
+  const Type &lsearch (const T &x, const Type &not_found = Null (Type)) const
+  { return *as_array ().lsearch (x, &not_found); }
+  template <typename T>
+  bool lfind (const T &x, unsigned *pos = nullptr) const
+  { return as_array ().lfind (x, pos); }
+
+  void qsort (unsigned int start = 0, unsigned int end = (unsigned int) -1)
+  { as_array ().qsort (start, end); }
+
+  bool sanitize_shallow (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (len.sanitize (c) && c->check_array (arrayZ, len));
+  }
+
+  public:
+  LenType	len;
+  Type		arrayZ[HB_VAR_ARRAY];
+  public:
+  DEFINE_SIZE_ARRAY (sizeof (LenType), arrayZ);
+};
+template <typename Type>
+using LArrayOf = ArrayOf<Type, HBUINT32>;
+using PString = ArrayOf<HBUINT8, HBUINT8>;
+
+/* Array of Offset's */
+template <typename Type>
+using OffsetArrayOf = ArrayOf<OffsetTo<Type, HBUINT16>>;
+template <typename Type>
+using LOffsetArrayOf = ArrayOf<OffsetTo<Type, HBUINT32>>;
+template <typename Type>
+using LOffsetLArrayOf = ArrayOf<OffsetTo<Type, HBUINT32>, HBUINT32>;
+
+/* Array of offsets relative to the beginning of the array itself. */
+template <typename Type>
+struct OffsetListOf : OffsetArrayOf<Type>
+{
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= this->len)) return Null (Type);
+    return this+this->arrayZ[i];
+  }
+  const Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= this->len)) return Crap (Type);
+    return this+this->arrayZ[i];
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    struct OffsetListOf<Type> *out = c->serializer->embed (*this);
+    if (unlikely (!out)) return_trace (false);
+    unsigned int count = this->len;
+    for (unsigned int i = 0; i < count; i++)
+      out->arrayZ[i].serialize_subset (c, this->arrayZ[i], this, out);
+    return_trace (true);
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (OffsetArrayOf<Type>::sanitize (c, this, hb_forward<Ts> (ds)...));
+  }
+};
+
+/* An array starting at second element. */
+template <typename Type, typename LenType=HBUINT16>
+struct HeadlessArrayOf
+{
+  static constexpr unsigned item_size = Type::static_size;
+
+  HB_DELETE_CREATE_COPY_ASSIGN (HeadlessArrayOf);
+
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= lenP1 || !i)) return Null (Type);
+    return arrayZ[i-1];
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= lenP1 || !i)) return Crap (Type);
+    return arrayZ[i-1];
+  }
+  unsigned int get_size () const
+  { return lenP1.static_size + get_length () * Type::static_size; }
+
+  unsigned get_length () const { return lenP1 ? lenP1 - 1 : 0; }
+
+  hb_array_t<      Type> as_array ()       { return hb_array (arrayZ, get_length ()); }
+  hb_array_t<const Type> as_array () const { return hb_array (arrayZ, get_length ()); }
+
+  /* Iterator. */
+  typedef hb_array_t<const Type>   iter_t;
+  typedef hb_array_t<      Type> writer_t;
+    iter_t   iter () const { return as_array (); }
+  writer_t writer ()       { return as_array (); }
+  operator   iter_t () const { return   iter (); }
+  operator writer_t ()       { return writer (); }
+
+  bool serialize (hb_serialize_context_t *c, unsigned int items_len)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    c->check_assign (lenP1, items_len + 1);
+    if (unlikely (!c->extend (*this))) return_trace (false);
+    return_trace (true);
+  }
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, Type))>
+  bool serialize (hb_serialize_context_t *c, Iterator items)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned count = items.len ();
+    if (unlikely (!serialize (c, count))) return_trace (false);
+    /* TODO Umm. Just exhaust the iterator instead?  Being extra
+     * cautious right now.. */
+    for (unsigned i = 0; i < count; i++, ++items)
+      arrayZ[i] = *items;
+    return_trace (true);
+  }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!sanitize_shallow (c))) return_trace (false);
+    if (!sizeof... (Ts) && hb_is_trivially_copyable (Type)) return_trace (true);
+    unsigned int count = get_length ();
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!c->dispatch (arrayZ[i], hb_forward<Ts> (ds)...)))
+	return_trace (false);
+    return_trace (true);
+  }
+
+  private:
+  bool sanitize_shallow (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (lenP1.sanitize (c) &&
+		  (!lenP1 || c->check_array (arrayZ, lenP1 - 1)));
+  }
+
+  public:
+  LenType	lenP1;
+  Type		arrayZ[HB_VAR_ARRAY];
+  public:
+  DEFINE_SIZE_ARRAY (sizeof (LenType), arrayZ);
+};
+
+/* An array storing length-1. */
+template <typename Type, typename LenType=HBUINT16>
+struct ArrayOfM1
+{
+  HB_DELETE_CREATE_COPY_ASSIGN (ArrayOfM1);
+
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i > lenM1)) return Null (Type);
+    return arrayZ[i];
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i > lenM1)) return Crap (Type);
+    return arrayZ[i];
+  }
+  unsigned int get_size () const
+  { return lenM1.static_size + (lenM1 + 1) * Type::static_size; }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!sanitize_shallow (c))) return_trace (false);
+    unsigned int count = lenM1 + 1;
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!c->dispatch (arrayZ[i], hb_forward<Ts> (ds)...)))
+	return_trace (false);
+    return_trace (true);
+  }
+
+  private:
+  bool sanitize_shallow (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (lenM1.sanitize (c) &&
+		  (c->check_array (arrayZ, lenM1 + 1)));
+  }
+
+  public:
+  LenType	lenM1;
+  Type		arrayZ[HB_VAR_ARRAY];
+  public:
+  DEFINE_SIZE_ARRAY (sizeof (LenType), arrayZ);
+};
+
+/* An array with sorted elements.  Supports binary searching. */
+template <typename Type, typename LenType=HBUINT16>
+struct SortedArrayOf : ArrayOf<Type, LenType>
+{
+  hb_sorted_array_t<      Type> as_array ()       { return hb_sorted_array (this->arrayZ, this->len); }
+  hb_sorted_array_t<const Type> as_array () const { return hb_sorted_array (this->arrayZ, this->len); }
+
+  /* Iterator. */
+  typedef hb_sorted_array_t<const Type>   iter_t;
+  typedef hb_sorted_array_t<      Type> writer_t;
+    iter_t   iter () const { return as_array (); }
+  writer_t writer ()       { return as_array (); }
+  operator   iter_t () const { return   iter (); }
+  operator writer_t ()       { return writer (); }
+
+  hb_sorted_array_t<const Type> sub_array (unsigned int start_offset, unsigned int count) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_sorted_array_t<const Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_sorted_array_t<Type> sub_array (unsigned int start_offset, unsigned int count)
+  { return as_array ().sub_array (start_offset, count); }
+  hb_sorted_array_t<Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */)
+  { return as_array ().sub_array (start_offset, count); }
+
+  bool serialize (hb_serialize_context_t *c, unsigned int items_len)
+  {
+    TRACE_SERIALIZE (this);
+    bool ret = ArrayOf<Type, LenType>::serialize (c, items_len);
+    return_trace (ret);
+  }
+  template <typename Iterator,
+	    hb_requires (hb_is_sorted_source_of (Iterator, Type))>
+  bool serialize (hb_serialize_context_t *c, Iterator items)
+  {
+    TRACE_SERIALIZE (this);
+    bool ret = ArrayOf<Type, LenType>::serialize (c, items);
+    return_trace (ret);
+  }
+
+  template <typename T>
+  Type &bsearch (const T &x, Type &not_found = Crap (Type))
+  { return *as_array ().bsearch (x, &not_found); }
+  template <typename T>
+  const Type &bsearch (const T &x, const Type &not_found = Null (Type)) const
+  { return *as_array ().bsearch (x, &not_found); }
+  template <typename T>
+  bool bfind (const T &x, unsigned int *i = nullptr,
+	      hb_bfind_not_found_t not_found = HB_BFIND_NOT_FOUND_DONT_STORE,
+	      unsigned int to_store = (unsigned int) -1) const
+  { return as_array ().bfind (x, i, not_found, to_store); }
+};
+
+/*
+ * Binary-search arrays
+ */
+
+template <typename LenType=HBUINT16>
+struct BinSearchHeader
+{
+  operator uint32_t () const { return len; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  BinSearchHeader& operator = (unsigned int v)
+  {
+    len = v;
+    assert (len == v);
+    entrySelector = hb_max (1u, hb_bit_storage (v)) - 1;
+    searchRange = 16 * (1u << entrySelector);
+    rangeShift = v * 16 > searchRange
+		 ? 16 * v - searchRange
+		 : 0;
+    return *this;
+  }
+
+  protected:
+  LenType	len;
+  LenType	searchRange;
+  LenType	entrySelector;
+  LenType	rangeShift;
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+template <typename Type, typename LenType=HBUINT16>
+using BinSearchArrayOf = SortedArrayOf<Type, BinSearchHeader<LenType>>;
+
+
+struct VarSizedBinSearchHeader
+{
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT16	unitSize;	/* Size of a lookup unit for this search in bytes. */
+  HBUINT16	nUnits;		/* Number of units of the preceding size to be searched. */
+  HBUINT16	searchRange;	/* The value of unitSize times the largest power of 2
+				 * that is less than or equal to the value of nUnits. */
+  HBUINT16	entrySelector;	/* The log base 2 of the largest power of 2 less than
+				 * or equal to the value of nUnits. */
+  HBUINT16	rangeShift;	/* The value of unitSize times the difference of the
+				 * value of nUnits minus the largest power of 2 less
+				 * than or equal to the value of nUnits. */
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+template <typename Type>
+struct VarSizedBinSearchArrayOf
+{
+  static constexpr unsigned item_size = Type::static_size;
+
+  HB_DELETE_CREATE_COPY_ASSIGN (VarSizedBinSearchArrayOf);
+
+  bool last_is_terminator () const
+  {
+    if (unlikely (!header.nUnits)) return false;
+
+    /* Gah.
+     *
+     * "The number of termination values that need to be included is table-specific.
+     * The value that indicates binary search termination is 0xFFFF." */
+    const HBUINT16 *words = &StructAtOffset<HBUINT16> (&bytesZ, (header.nUnits - 1) * header.unitSize);
+    unsigned int count = Type::TerminationWordCount;
+    for (unsigned int i = 0; i < count; i++)
+      if (words[i] != 0xFFFFu)
+	return false;
+    return true;
+  }
+
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= get_length ())) return Null (Type);
+    return StructAtOffset<Type> (&bytesZ, i * header.unitSize);
+  }
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= get_length ())) return Crap (Type);
+    return StructAtOffset<Type> (&bytesZ, i * header.unitSize);
+  }
+  unsigned int get_length () const
+  { return header.nUnits - last_is_terminator (); }
+  unsigned int get_size () const
+  { return header.static_size + header.nUnits * header.unitSize; }
+
+  template <typename ...Ts>
+  bool sanitize (hb_sanitize_context_t *c, Ts&&... ds) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!sanitize_shallow (c))) return_trace (false);
+    if (!sizeof... (Ts) && hb_is_trivially_copyable (Type)) return_trace (true);
+    unsigned int count = get_length ();
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!(*this)[i].sanitize (c, hb_forward<Ts> (ds)...)))
+	return_trace (false);
+    return_trace (true);
+  }
+
+  template <typename T>
+  const Type *bsearch (const T &key) const
+  {
+    unsigned pos;
+    return hb_bsearch_impl (&pos,
+			    key,
+			    (const void *) bytesZ,
+			    get_length (),
+			    header.unitSize,
+			    _hb_cmp_method<T, Type>)
+	   ? (const Type *) (((const char *) &bytesZ) + (pos * header.unitSize))
+	   : nullptr;
+  }
+
+  private:
+  bool sanitize_shallow (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (header.sanitize (c) &&
+		  Type::static_size <= header.unitSize &&
+		  c->check_range (bytesZ.arrayZ,
+				  header.nUnits,
+				  header.unitSize));
+  }
+
+  protected:
+  VarSizedBinSearchHeader	header;
+  UnsizedArrayOf<HBUINT8>	bytesZ;
+  public:
+  DEFINE_SIZE_ARRAY (10, bytesZ);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OPEN_TYPE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff-common.hh b/thirdparty/harfbuzz/src/hb-ot-cff-common.hh
new file mode 100644
index 0000000000..e5286cd792
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff-common.hh
@@ -0,0 +1,622 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+#ifndef HB_OT_CFF_COMMON_HH
+#define HB_OT_CFF_COMMON_HH
+
+#include "hb-open-type.hh"
+#include "hb-bimap.hh"
+#include "hb-ot-layout-common.hh"
+#include "hb-cff-interp-dict-common.hh"
+#include "hb-subset-plan.hh"
+
+namespace CFF {
+
+using namespace OT;
+
+#define CFF_UNDEF_CODE  0xFFFFFFFF
+
+using objidx_t = hb_serialize_context_t::objidx_t;
+using whence_t = hb_serialize_context_t::whence_t;
+
+/* utility macro */
+template<typename Type>
+static inline const Type& StructAtOffsetOrNull (const void *P, unsigned int offset)
+{ return offset ? StructAtOffset<Type> (P, offset) : Null (Type); }
+
+inline unsigned int calcOffSize (unsigned int dataSize)
+{
+  unsigned int size = 1;
+  unsigned int offset = dataSize + 1;
+  while (offset & ~0xFF)
+  {
+    size++;
+    offset >>= 8;
+  }
+  /* format does not support size > 4; caller should handle it as an error */
+  return size;
+}
+
+struct code_pair_t
+{
+  hb_codepoint_t code;
+  hb_codepoint_t glyph;
+};
+
+typedef hb_vector_t<unsigned char> str_buff_t;
+struct str_buff_vec_t : hb_vector_t<str_buff_t>
+{
+  void fini () { SUPER::fini_deep (); }
+
+  unsigned int total_size () const
+  {
+    unsigned int size = 0;
+    for (unsigned int i = 0; i < length; i++)
+      size += (*this)[i].length;
+    return size;
+  }
+
+  private:
+  typedef hb_vector_t<str_buff_t> SUPER;
+};
+
+/* CFF INDEX */
+template <typename COUNT>
+struct CFFIndex
+{
+  static unsigned int calculate_offset_array_size (unsigned int offSize, unsigned int count)
+  { return offSize * (count + 1); }
+
+  unsigned int offset_array_size () const
+  { return calculate_offset_array_size (offSize, count); }
+
+  CFFIndex *copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = get_size ();
+    CFFIndex *out = c->allocate_size<CFFIndex> (size);
+    if (likely (out))
+      memcpy (out, this, size);
+    return_trace (out);
+  }
+
+  bool serialize (hb_serialize_context_t *c, const CFFIndex &src)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = src.get_size ();
+    CFFIndex *dest = c->allocate_size<CFFIndex> (size);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, &src, size);
+    return_trace (true);
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned int offSize_,
+		  const byte_str_array_t &byteArray)
+  {
+    TRACE_SERIALIZE (this);
+    if (byteArray.length == 0)
+    {
+      COUNT *dest = c->allocate_min<COUNT> ();
+      if (unlikely (!dest)) return_trace (false);
+      *dest = 0;
+    }
+    else
+    {
+      /* serialize CFFIndex header */
+      if (unlikely (!c->extend_min (*this))) return_trace (false);
+      this->count = byteArray.length;
+      this->offSize = offSize_;
+      if (unlikely (!c->allocate_size<HBUINT8> (offSize_ * (byteArray.length + 1))))
+	return_trace (false);
+
+      /* serialize indices */
+      unsigned int  offset = 1;
+      unsigned int  i = 0;
+      for (; i < byteArray.length; i++)
+      {
+	set_offset_at (i, offset);
+	offset += byteArray[i].get_size ();
+      }
+      set_offset_at (i, offset);
+
+      /* serialize data */
+      for (unsigned int i = 0; i < byteArray.length; i++)
+      {
+	const byte_str_t &bs = byteArray[i];
+	unsigned char *dest = c->allocate_size<unsigned char> (bs.length);
+	if (unlikely (!dest)) return_trace (false);
+	memcpy (dest, &bs[0], bs.length);
+      }
+    }
+    return_trace (true);
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned int offSize_,
+		  const str_buff_vec_t &buffArray)
+  {
+    byte_str_array_t  byteArray;
+    byteArray.init ();
+    byteArray.resize (buffArray.length);
+    for (unsigned int i = 0; i < byteArray.length; i++)
+      byteArray[i] = byte_str_t (buffArray[i].arrayZ, buffArray[i].length);
+    bool result = this->serialize (c, offSize_, byteArray);
+    byteArray.fini ();
+    return result;
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+    if (it.len () == 0)
+    {
+      COUNT *dest = c->allocate_min<COUNT> ();
+      if (unlikely (!dest)) return_trace (false);
+      *dest = 0;
+    }
+    else
+    {
+      serialize_header(c, + it | hb_map ([] (const byte_str_t &_) { return _.length; }));
+      for (const byte_str_t &_ : +it)
+	_.copy (c);
+    }
+    return_trace (true);
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  const byte_str_array_t &byteArray)
+  { return serialize (c, + hb_iter (byteArray)); }
+
+  bool serialize (hb_serialize_context_t *c,
+		  const str_buff_vec_t &buffArray)
+  {
+    auto it =
+    + hb_iter (buffArray)
+    | hb_map ([] (const str_buff_t &_) { return byte_str_t (_.arrayZ, _.length); })
+    ;
+    return serialize (c, it);
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  bool serialize_header (hb_serialize_context_t *c,
+			Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+
+    unsigned total = + it | hb_reduce (hb_add, 0);
+    unsigned off_size = calcOffSize (total);
+
+    /* serialize CFFIndex header */
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    this->count = it.len ();
+    this->offSize = off_size;
+    if (unlikely (!c->allocate_size<HBUINT8> (off_size * (it.len () + 1))))
+      return_trace (false);
+
+    /* serialize indices */
+    unsigned int offset = 1;
+    unsigned int i = 0;
+    for (unsigned _ : +it)
+    {
+      CFFIndex<COUNT>::set_offset_at (i++, offset);
+      offset += _;
+    }
+    CFFIndex<COUNT>::set_offset_at (i, offset);
+
+    return_trace (true);
+  }
+
+  void set_offset_at (unsigned int index, unsigned int offset)
+  {
+    HBUINT8 *p = offsets + offSize * index + offSize;
+    unsigned int size = offSize;
+    for (; size; size--)
+    {
+      --p;
+      *p = offset & 0xFF;
+      offset >>= 8;
+    }
+  }
+
+  unsigned int offset_at (unsigned int index) const
+  {
+    assert (index <= count);
+    const HBUINT8 *p = offsets + offSize * index;
+    unsigned int size = offSize;
+    unsigned int offset = 0;
+    for (; size; size--)
+      offset = (offset << 8) + *p++;
+    return offset;
+  }
+
+  unsigned int length_at (unsigned int index) const
+  {
+    if (unlikely ((offset_at (index + 1) < offset_at (index)) ||
+		  (offset_at (index + 1) > offset_at (count))))
+      return 0;
+    return offset_at (index + 1) - offset_at (index);
+  }
+
+  const unsigned char *data_base () const
+  { return (const unsigned char *) this + min_size + offset_array_size (); }
+
+  unsigned int data_size () const { return HBINT8::static_size; }
+
+  byte_str_t operator [] (unsigned int index) const
+  {
+    if (unlikely (index >= count)) return Null (byte_str_t);
+    return byte_str_t (data_base () + offset_at (index) - 1, length_at (index));
+  }
+
+  unsigned int get_size () const
+  {
+    if (this == &Null (CFFIndex)) return 0;
+    if (count > 0)
+      return min_size + offset_array_size () + (offset_at (count) - 1);
+    return count.static_size;  /* empty CFFIndex contains count only */
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely ((c->check_struct (this) && count == 0) || /* empty INDEX */
+			  (c->check_struct (this) && offSize >= 1 && offSize <= 4 &&
+			   c->check_array (offsets, offSize, count + 1) &&
+			   c->check_array ((const HBUINT8*) data_base (), 1, max_offset () - 1))));
+  }
+
+  protected:
+  unsigned int max_offset () const
+  {
+    unsigned int max = 0;
+    for (unsigned int i = 0; i < count + 1u; i++)
+    {
+      unsigned int off = offset_at (i);
+      if (off > max) max = off;
+    }
+    return max;
+  }
+
+  public:
+  COUNT		count;		/* Number of object data. Note there are (count+1) offsets */
+  HBUINT8	offSize;	/* The byte size of each offset in the offsets array. */
+  HBUINT8	offsets[HB_VAR_ARRAY];
+				/* The array of (count + 1) offsets into objects array (1-base). */
+  /* HBUINT8 data[HB_VAR_ARRAY];	Object data */
+  public:
+  DEFINE_SIZE_ARRAY (COUNT::static_size + HBUINT8::static_size, offsets);
+};
+
+template <typename COUNT, typename TYPE>
+struct CFFIndexOf : CFFIndex<COUNT>
+{
+  const byte_str_t operator [] (unsigned int index) const
+  {
+    if (likely (index < CFFIndex<COUNT>::count))
+      return byte_str_t (CFFIndex<COUNT>::data_base () + CFFIndex<COUNT>::offset_at (index) - 1, CFFIndex<COUNT>::length_at (index));
+    return Null (byte_str_t);
+  }
+
+  template <typename DATA, typename PARAM1, typename PARAM2>
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned int offSize_,
+		  const DATA *dataArray,
+		  unsigned int dataArrayLen,
+		  const hb_vector_t<unsigned int> &dataSizeArray,
+		  const PARAM1 &param1,
+		  const PARAM2 &param2)
+  {
+    TRACE_SERIALIZE (this);
+    /* serialize CFFIndex header */
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    this->count = dataArrayLen;
+    this->offSize = offSize_;
+    if (unlikely (!c->allocate_size<HBUINT8> (offSize_ * (dataArrayLen + 1))))
+      return_trace (false);
+
+    /* serialize indices */
+    unsigned int  offset = 1;
+    unsigned int  i = 0;
+    for (; i < dataArrayLen; i++)
+    {
+      CFFIndex<COUNT>::set_offset_at (i, offset);
+      offset += dataSizeArray[i];
+    }
+    CFFIndex<COUNT>::set_offset_at (i, offset);
+
+    /* serialize data */
+    for (unsigned int i = 0; i < dataArrayLen; i++)
+    {
+      TYPE *dest = c->start_embed<TYPE> ();
+      if (unlikely (!dest || !dest->serialize (c, dataArray[i], param1, param2)))
+	return_trace (false);
+    }
+    return_trace (true);
+  }
+};
+
+/* Top Dict, Font Dict, Private Dict */
+struct Dict : UnsizedByteStr
+{
+  template <typename DICTVAL, typename OP_SERIALIZER, typename ...Ts>
+  bool serialize (hb_serialize_context_t *c,
+		  const DICTVAL &dictval,
+		  OP_SERIALIZER& opszr,
+		  Ts&&... ds)
+  {
+    TRACE_SERIALIZE (this);
+    for (unsigned int i = 0; i < dictval.get_count (); i++)
+      if (unlikely (!opszr.serialize (c, dictval[i], hb_forward<Ts> (ds)...)))
+	return_trace (false);
+
+    return_trace (true);
+  }
+
+  template <typename T, typename V>
+  static bool serialize_int_op (hb_serialize_context_t *c, op_code_t op, V value, op_code_t intOp)
+  {
+    // XXX: not sure why but LLVM fails to compile the following 'unlikely' macro invocation
+    if (/*unlikely*/ (!serialize_int<T, V> (c, intOp, value)))
+      return false;
+
+    TRACE_SERIALIZE (this);
+    /* serialize the opcode */
+    HBUINT8 *p = c->allocate_size<HBUINT8> (OpCode_Size (op));
+    if (unlikely (!p)) return_trace (false);
+    if (Is_OpCode_ESC (op))
+    {
+      *p = OpCode_escape;
+      op = Unmake_OpCode_ESC (op);
+      p++;
+    }
+    *p = op;
+    return_trace (true);
+  }
+
+  template <typename V>
+  static bool serialize_int4_op (hb_serialize_context_t *c, op_code_t op, V value)
+  { return serialize_int_op<HBINT32> (c, op, value, OpCode_longintdict); }
+
+  template <typename V>
+  static bool serialize_int2_op (hb_serialize_context_t *c, op_code_t op, V value)
+  { return serialize_int_op<HBINT16> (c, op, value, OpCode_shortint); }
+
+  template <typename T, int int_op>
+  static bool serialize_link_op (hb_serialize_context_t *c, op_code_t op, objidx_t link, whence_t whence)
+  {
+    T &ofs = *(T *) (c->head + OpCode_Size (int_op));
+    if (unlikely (!serialize_int_op<T> (c, op, 0, int_op))) return false;
+    c->add_link (ofs, link, whence);
+    return true;
+  }
+
+  static bool serialize_link4_op (hb_serialize_context_t *c, op_code_t op, objidx_t link, whence_t whence = whence_t::Head)
+  { return serialize_link_op<HBINT32, OpCode_longintdict> (c, op, link, whence); }
+
+  static bool serialize_link2_op (hb_serialize_context_t *c, op_code_t op, objidx_t link, whence_t whence = whence_t::Head)
+  { return serialize_link_op<HBINT16, OpCode_shortint> (c, op, link, whence); }
+};
+
+struct TopDict : Dict {};
+struct FontDict : Dict {};
+struct PrivateDict : Dict {};
+
+struct table_info_t
+{
+  void init () { offset = size = 0; link = 0; }
+
+  unsigned int    offset;
+  unsigned int    size;
+  objidx_t	  link;
+};
+
+template <typename COUNT>
+struct FDArray : CFFIndexOf<COUNT, FontDict>
+{
+  template <typename DICTVAL, typename INFO, typename Iterator, typename OP_SERIALIZER>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  OP_SERIALIZER& opszr)
+  {
+    TRACE_SERIALIZE (this);
+
+    /* serialize INDEX data */
+    hb_vector_t<unsigned> sizes;
+    c->push ();
+    + it
+    | hb_map ([&] (const hb_pair_t<const DICTVAL&, const INFO&> &_)
+    {
+      FontDict *dict = c->start_embed<FontDict> ();
+		dict->serialize (c, _.first, opszr, _.second);
+		return c->head - (const char*)dict;
+	      })
+    | hb_sink (sizes)
+    ;
+    c->pop_pack (false);
+
+    /* serialize INDEX header */
+    return_trace (CFFIndex<COUNT>::serialize_header (c, hb_iter (sizes)));
+  }
+};
+
+/* FDSelect */
+struct FDSelect0 {
+  bool sanitize (hb_sanitize_context_t *c, unsigned int fdcount) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!(c->check_struct (this))))
+      return_trace (false);
+    for (unsigned int i = 0; i < c->get_num_glyphs (); i++)
+      if (unlikely (!fds[i].sanitize (c)))
+	return_trace (false);
+
+    return_trace (true);
+  }
+
+  hb_codepoint_t get_fd (hb_codepoint_t glyph) const
+  { return (hb_codepoint_t) fds[glyph]; }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  { return HBUINT8::static_size * num_glyphs; }
+
+  HBUINT8     fds[HB_VAR_ARRAY];
+
+  DEFINE_SIZE_MIN (0);
+};
+
+template <typename GID_TYPE, typename FD_TYPE>
+struct FDSelect3_4_Range
+{
+  bool sanitize (hb_sanitize_context_t *c, const void * /*nullptr*/, unsigned int fdcount) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (first < c->get_num_glyphs () && (fd < fdcount));
+  }
+
+  GID_TYPE    first;
+  FD_TYPE     fd;
+  public:
+  DEFINE_SIZE_STATIC (GID_TYPE::static_size + FD_TYPE::static_size);
+};
+
+template <typename GID_TYPE, typename FD_TYPE>
+struct FDSelect3_4
+{
+  unsigned int get_size () const
+  { return GID_TYPE::static_size * 2 + ranges.get_size (); }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int fdcount) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this) || !ranges.sanitize (c, nullptr, fdcount) ||
+		  (nRanges () == 0) || ranges[0].first != 0))
+      return_trace (false);
+
+    for (unsigned int i = 1; i < nRanges (); i++)
+      if (unlikely (ranges[i - 1].first >= ranges[i].first))
+	return_trace (false);
+
+    if (unlikely (!sentinel().sanitize (c) || (sentinel() != c->get_num_glyphs ())))
+      return_trace (false);
+
+    return_trace (true);
+  }
+
+  hb_codepoint_t get_fd (hb_codepoint_t glyph) const
+  {
+    unsigned int i;
+    for (i = 1; i < nRanges (); i++)
+      if (glyph < ranges[i].first)
+	break;
+
+    return (hb_codepoint_t) ranges[i - 1].fd;
+  }
+
+  GID_TYPE        &nRanges ()       { return ranges.len; }
+  GID_TYPE         nRanges () const { return ranges.len; }
+  GID_TYPE       &sentinel ()       { return StructAfter<GID_TYPE> (ranges[nRanges () - 1]); }
+  const GID_TYPE &sentinel () const { return StructAfter<GID_TYPE> (ranges[nRanges () - 1]); }
+
+  ArrayOf<FDSelect3_4_Range<GID_TYPE, FD_TYPE>, GID_TYPE> ranges;
+  /* GID_TYPE sentinel */
+
+  DEFINE_SIZE_ARRAY (GID_TYPE::static_size, ranges);
+};
+
+typedef FDSelect3_4<HBUINT16, HBUINT8> FDSelect3;
+typedef FDSelect3_4_Range<HBUINT16, HBUINT8> FDSelect3_Range;
+
+struct FDSelect
+{
+  bool serialize (hb_serialize_context_t *c, const FDSelect &src, unsigned int num_glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = src.get_size (num_glyphs);
+    FDSelect *dest = c->allocate_size<FDSelect> (size);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, &src, size);
+    return_trace (true);
+  }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  {
+    switch (format)
+    {
+    case 0: return format.static_size + u.format0.get_size (num_glyphs);
+    case 3: return format.static_size + u.format3.get_size ();
+    default:return 0;
+    }
+  }
+
+  hb_codepoint_t get_fd (hb_codepoint_t glyph) const
+  {
+    if (this == &Null (FDSelect)) return 0;
+
+    switch (format)
+    {
+    case 0: return u.format0.get_fd (glyph);
+    case 3: return u.format3.get_fd (glyph);
+    default:return 0;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int fdcount) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (format)
+    {
+    case 0: return_trace (u.format0.sanitize (c, fdcount));
+    case 3: return_trace (u.format3.sanitize (c, fdcount));
+    default:return_trace (false);
+    }
+  }
+
+  HBUINT8	format;
+  union {
+  FDSelect0	format0;
+  FDSelect3	format3;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (1);
+};
+
+template <typename COUNT>
+struct Subrs : CFFIndex<COUNT>
+{
+  typedef COUNT count_type;
+  typedef CFFIndex<COUNT> SUPER;
+};
+
+} /* namespace CFF */
+
+#endif /* HB_OT_CFF_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff1-std-str.hh b/thirdparty/harfbuzz/src/hb-ot-cff1-std-str.hh
new file mode 100644
index 0000000000..65d56ae18b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff1-std-str.hh
@@ -0,0 +1,425 @@
+/*
+ * Copyright © 2019  Adobe, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_CFF1_STD_STR_HH
+#if 0 /* Make checks happy. */
+#define HB_OT_CFF1_STD_STR_HH
+#include "hb.hh"
+#endif
+
+_S(".notdef")
+_S("space")
+_S("exclam")
+_S("quotedbl")
+_S("numbersign")
+_S("dollar")
+_S("percent")
+_S("ampersand")
+_S("quoteright")
+_S("parenleft")
+_S("parenright")
+_S("asterisk")
+_S("plus")
+_S("comma")
+_S("hyphen")
+_S("period")
+_S("slash")
+_S("zero")
+_S("one")
+_S("two")
+_S("three")
+_S("four")
+_S("five")
+_S("six")
+_S("seven")
+_S("eight")
+_S("nine")
+_S("colon")
+_S("semicolon")
+_S("less")
+_S("equal")
+_S("greater")
+_S("question")
+_S("at")
+_S("A")
+_S("B")
+_S("C")
+_S("D")
+_S("E")
+_S("F")
+_S("G")
+_S("H")
+_S("I")
+_S("J")
+_S("K")
+_S("L")
+_S("M")
+_S("N")
+_S("O")
+_S("P")
+_S("Q")
+_S("R")
+_S("S")
+_S("T")
+_S("U")
+_S("V")
+_S("W")
+_S("X")
+_S("Y")
+_S("Z")
+_S("bracketleft")
+_S("backslash")
+_S("bracketright")
+_S("asciicircum")
+_S("underscore")
+_S("quoteleft")
+_S("a")
+_S("b")
+_S("c")
+_S("d")
+_S("e")
+_S("f")
+_S("g")
+_S("h")
+_S("i")
+_S("j")
+_S("k")
+_S("l")
+_S("m")
+_S("n")
+_S("o")
+_S("p")
+_S("q")
+_S("r")
+_S("s")
+_S("t")
+_S("u")
+_S("v")
+_S("w")
+_S("x")
+_S("y")
+_S("z")
+_S("braceleft")
+_S("bar")
+_S("braceright")
+_S("asciitilde")
+_S("exclamdown")
+_S("cent")
+_S("sterling")
+_S("fraction")
+_S("yen")
+_S("florin")
+_S("section")
+_S("currency")
+_S("quotesingle")
+_S("quotedblleft")
+_S("guillemotleft")
+_S("guilsinglleft")
+_S("guilsinglright")
+_S("fi")
+_S("fl")
+_S("endash")
+_S("dagger")
+_S("daggerdbl")
+_S("periodcentered")
+_S("paragraph")
+_S("bullet")
+_S("quotesinglbase")
+_S("quotedblbase")
+_S("quotedblright")
+_S("guillemotright")
+_S("ellipsis")
+_S("perthousand")
+_S("questiondown")
+_S("grave")
+_S("acute")
+_S("circumflex")
+_S("tilde")
+_S("macron")
+_S("breve")
+_S("dotaccent")
+_S("dieresis")
+_S("ring")
+_S("cedilla")
+_S("hungarumlaut")
+_S("ogonek")
+_S("caron")
+_S("emdash")
+_S("AE")
+_S("ordfeminine")
+_S("Lslash")
+_S("Oslash")
+_S("OE")
+_S("ordmasculine")
+_S("ae")
+_S("dotlessi")
+_S("lslash")
+_S("oslash")
+_S("oe")
+_S("germandbls")
+_S("onesuperior")
+_S("logicalnot")
+_S("mu")
+_S("trademark")
+_S("Eth")
+_S("onehalf")
+_S("plusminus")
+_S("Thorn")
+_S("onequarter")
+_S("divide")
+_S("brokenbar")
+_S("degree")
+_S("thorn")
+_S("threequarters")
+_S("twosuperior")
+_S("registered")
+_S("minus")
+_S("eth")
+_S("multiply")
+_S("threesuperior")
+_S("copyright")
+_S("Aacute")
+_S("Acircumflex")
+_S("Adieresis")
+_S("Agrave")
+_S("Aring")
+_S("Atilde")
+_S("Ccedilla")
+_S("Eacute")
+_S("Ecircumflex")
+_S("Edieresis")
+_S("Egrave")
+_S("Iacute")
+_S("Icircumflex")
+_S("Idieresis")
+_S("Igrave")
+_S("Ntilde")
+_S("Oacute")
+_S("Ocircumflex")
+_S("Odieresis")
+_S("Ograve")
+_S("Otilde")
+_S("Scaron")
+_S("Uacute")
+_S("Ucircumflex")
+_S("Udieresis")
+_S("Ugrave")
+_S("Yacute")
+_S("Ydieresis")
+_S("Zcaron")
+_S("aacute")
+_S("acircumflex")
+_S("adieresis")
+_S("agrave")
+_S("aring")
+_S("atilde")
+_S("ccedilla")
+_S("eacute")
+_S("ecircumflex")
+_S("edieresis")
+_S("egrave")
+_S("iacute")
+_S("icircumflex")
+_S("idieresis")
+_S("igrave")
+_S("ntilde")
+_S("oacute")
+_S("ocircumflex")
+_S("odieresis")
+_S("ograve")
+_S("otilde")
+_S("scaron")
+_S("uacute")
+_S("ucircumflex")
+_S("udieresis")
+_S("ugrave")
+_S("yacute")
+_S("ydieresis")
+_S("zcaron")
+_S("exclamsmall")
+_S("Hungarumlautsmall")
+_S("dollaroldstyle")
+_S("dollarsuperior")
+_S("ampersandsmall")
+_S("Acutesmall")
+_S("parenleftsuperior")
+_S("parenrightsuperior")
+_S("twodotenleader")
+_S("onedotenleader")
+_S("zerooldstyle")
+_S("oneoldstyle")
+_S("twooldstyle")
+_S("threeoldstyle")
+_S("fouroldstyle")
+_S("fiveoldstyle")
+_S("sixoldstyle")
+_S("sevenoldstyle")
+_S("eightoldstyle")
+_S("nineoldstyle")
+_S("commasuperior")
+_S("threequartersemdash")
+_S("periodsuperior")
+_S("questionsmall")
+_S("asuperior")
+_S("bsuperior")
+_S("centsuperior")
+_S("dsuperior")
+_S("esuperior")
+_S("isuperior")
+_S("lsuperior")
+_S("msuperior")
+_S("nsuperior")
+_S("osuperior")
+_S("rsuperior")
+_S("ssuperior")
+_S("tsuperior")
+_S("ff")
+_S("ffi")
+_S("ffl")
+_S("parenleftinferior")
+_S("parenrightinferior")
+_S("Circumflexsmall")
+_S("hyphensuperior")
+_S("Gravesmall")
+_S("Asmall")
+_S("Bsmall")
+_S("Csmall")
+_S("Dsmall")
+_S("Esmall")
+_S("Fsmall")
+_S("Gsmall")
+_S("Hsmall")
+_S("Ismall")
+_S("Jsmall")
+_S("Ksmall")
+_S("Lsmall")
+_S("Msmall")
+_S("Nsmall")
+_S("Osmall")
+_S("Psmall")
+_S("Qsmall")
+_S("Rsmall")
+_S("Ssmall")
+_S("Tsmall")
+_S("Usmall")
+_S("Vsmall")
+_S("Wsmall")
+_S("Xsmall")
+_S("Ysmall")
+_S("Zsmall")
+_S("colonmonetary")
+_S("onefitted")
+_S("rupiah")
+_S("Tildesmall")
+_S("exclamdownsmall")
+_S("centoldstyle")
+_S("Lslashsmall")
+_S("Scaronsmall")
+_S("Zcaronsmall")
+_S("Dieresissmall")
+_S("Brevesmall")
+_S("Caronsmall")
+_S("Dotaccentsmall")
+_S("Macronsmall")
+_S("figuredash")
+_S("hypheninferior")
+_S("Ogoneksmall")
+_S("Ringsmall")
+_S("Cedillasmall")
+_S("questiondownsmall")
+_S("oneeighth")
+_S("threeeighths")
+_S("fiveeighths")
+_S("seveneighths")
+_S("onethird")
+_S("twothirds")
+_S("zerosuperior")
+_S("foursuperior")
+_S("fivesuperior")
+_S("sixsuperior")
+_S("sevensuperior")
+_S("eightsuperior")
+_S("ninesuperior")
+_S("zeroinferior")
+_S("oneinferior")
+_S("twoinferior")
+_S("threeinferior")
+_S("fourinferior")
+_S("fiveinferior")
+_S("sixinferior")
+_S("seveninferior")
+_S("eightinferior")
+_S("nineinferior")
+_S("centinferior")
+_S("dollarinferior")
+_S("periodinferior")
+_S("commainferior")
+_S("Agravesmall")
+_S("Aacutesmall")
+_S("Acircumflexsmall")
+_S("Atildesmall")
+_S("Adieresissmall")
+_S("Aringsmall")
+_S("AEsmall")
+_S("Ccedillasmall")
+_S("Egravesmall")
+_S("Eacutesmall")
+_S("Ecircumflexsmall")
+_S("Edieresissmall")
+_S("Igravesmall")
+_S("Iacutesmall")
+_S("Icircumflexsmall")
+_S("Idieresissmall")
+_S("Ethsmall")
+_S("Ntildesmall")
+_S("Ogravesmall")
+_S("Oacutesmall")
+_S("Ocircumflexsmall")
+_S("Otildesmall")
+_S("Odieresissmall")
+_S("OEsmall")
+_S("Oslashsmall")
+_S("Ugravesmall")
+_S("Uacutesmall")
+_S("Ucircumflexsmall")
+_S("Udieresissmall")
+_S("Yacutesmall")
+_S("Thornsmall")
+_S("Ydieresissmall")
+_S("001.000")
+_S("001.001")
+_S("001.002")
+_S("001.003")
+_S("Black")
+_S("Bold")
+_S("Book")
+_S("Light")
+_S("Medium")
+_S("Regular")
+_S("Roman")
+_S("Semibold")
+
+#endif /* HB_OT_CFF1_STD_STR_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff1-table.cc b/thirdparty/harfbuzz/src/hb-ot-cff1-table.cc
new file mode 100644
index 0000000000..66b9c8c907
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff1-table.cc
@@ -0,0 +1,620 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_CFF
+
+#include "hb-draw.hh"
+#include "hb-algs.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-cff1-interp-cs.hh"
+
+using namespace CFF;
+
+struct sid_to_gid_t
+{
+  uint16_t  sid;
+  uint8_t   gid;
+
+  int cmp (uint16_t a) const
+  {
+    if (a == sid) return 0;
+    return (a < sid) ? -1 : 1;
+  }
+};
+
+/* SID to code */
+static const uint8_t standard_encoding_to_code [] =
+{
+    0,   32,   33,   34,   35,   36,   37,   38,  39,   40,   41,   42,   43,   44,   45,   46,
+   47,   48,   49,   50,   51,   52,   53,   54,  55,   56,   57,   58,   59,   60,   61,   62,
+   63,   64,   65,   66,   67,   68,   69,   70,  71,   72,   73,   74,   75,   76,   77,   78,
+   79,   80,   81,   82,   83,   84,   85,   86,  87,   88,   89,   90,   91,   92,   93,   94,
+   95,   96,   97,   98,   99,  100,  101,  102, 103,  104,  105,  106,  107,  108,  109,  110,
+  111,  112,  113,  114,  115,  116,  117,  118, 119,  120,  121,  122,  123,  124,  125,  126,
+  161,  162,  163,  164,  165,  166,  167,  168, 169,  170,  171,  172,  173,  174,  175,  177,
+  178,  179,  180,  182,  183,  184,  185,  186, 187,  188,  189,  191,  193,  194,  195,  196,
+  197,  198,  199,  200,  202,  203,  205,  206, 207,  208,  225,  227,  232,  233,  234,  235,
+  241,  245,  248,  249,  250,  251
+};
+
+/* SID to code */
+static const uint8_t expert_encoding_to_code [] =
+{
+    0,   32,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,   44,   45,   46,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,   58,   59,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,   47,    0,    0,    0,    0,    0,    0,    0,    0,    0,   87,   88,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,  201,    0,    0,    0,    0,  189,    0,    0,  188,    0,
+    0,    0,    0,  190,  202,    0,    0,    0,    0,  203,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,   33,   34,   36,   37,   38,   39,   40,   41,   42,   43,   48,
+   49,   50,   51,   52,   53,   54,   55,   56,   57,   60,   61,   62,   63,   65,   66,   67,
+   68,   69,   73,   76,   77,   78,   79,   82,   83,   84,   86,   89,   90,   91,   93,   94,
+   95,   96,   97,   98,   99,  100,  101,  102,  103,  104,  105,  106,  107,  108,  109,  110,
+  111,  112,  113,  114,  115,  116,  117,  118,  119,  120,  121,  122,  123,  124,  125,  126,
+  161,  162,  163,  166,  167,  168,  169,  170,  172,  175,  178,  179,  182,  183,  184,  191,
+  192,  193,  194,  195,  196,  197,  200,  204,  205,  206,  207,  208,  209,  210,  211,  212,
+  213,  214,  215,  216,  217,  218,  219,  220,  221,  222,  223,  224,  225,  226,  227,  228,
+  229,  230,  231,  232,  233,  234,  235,  236,  237,  238,  239,  240,  241,  242,  243,  244,
+  245,  246,  247,  248,  249,  250,  251,  252,  253,  254,  255
+};
+
+/* glyph ID to SID */
+static const uint16_t expert_charset_to_sid [] =
+{
+    0,    1,  229,  230,  231,  232,  233,  234,  235,  236,  237,  238,   13,   14,   15,   99,
+  239,  240,  241,  242,  243,  244,  245,  246,  247,  248,   27,   28,  249,  250,  251,  252,
+  253,  254,  255,  256,  257,  258,  259,  260,  261,  262,  263,  264,  265,  266,  109,  110,
+  267,  268,  269,  270,  271,  272,  273,  274,  275,  276,  277,  278,  279,  280,  281,  282,
+  283,  284,  285,  286,  287,  288,  289,  290,  291,  292,  293,  294,  295,  296,  297,  298,
+  299,  300,  301,  302,  303,  304,  305,  306,  307,  308,  309,  310,  311,  312,  313,  314,
+  315,  316,  317,  318,  158,  155,  163,  319,  320,  321,  322,  323,  324,  325,  326,  150,
+  164,  169,  327,  328,  329,  330,  331,  332,  333,  334,  335,  336,  337,  338,  339,  340,
+  341,  342,  343,  344,  345,  346,  347,  348,  349,  350,  351,  352,  353,  354,  355,  356,
+  357,  358,  359,  360,  361,  362,  363,  364,  365,  366,  367,  368,  369,  370,  371,  372,
+  373,  374,  375,  376,  377,  378
+};
+
+/* glyph ID to SID */
+static const uint16_t expert_subset_charset_to_sid [] =
+{
+    0,    1,  231,  232,  235,  236,  237,  238,   13,   14,   15,   99,  239,  240,  241,  242,
+  243,  244,  245,  246,  247,  248,   27,   28,  249,  250,  251,  253,  254,  255,  256,  257,
+  258,  259,  260,  261,  262,  263,  264,  265,  266,  109,  110,  267,  268,  269,  270,  272,
+  300,  301,  302,  305,  314,  315,  158,  155,  163,  320,  321,  322,  323,  324,  325,  326,
+  150,  164,  169,  327,  328,  329,  330,  331,  332,  333,  334,  335,  336,  337,  338,  339,
+  340,  341,  342,  343,  344,  345,  346
+};
+
+/* SID to glyph ID */
+static const sid_to_gid_t expert_charset_sid_to_gid [] =
+{
+    { 1, 1 },     { 13, 12 },   { 14, 13 },   { 15, 14 },
+    { 27, 26 },   { 28, 27 },   { 99, 15 },   { 109, 46 },
+    { 110, 47 },  { 150, 111 }, { 155, 101 }, { 158, 100 },
+    { 163, 102 }, { 164, 112 }, { 169, 113 }, { 229, 2 },
+    { 230, 3 },   { 231, 4 },   { 232, 5 },   { 233, 6 },
+    { 234, 7 },   { 235, 8 },   { 236, 9 },   { 237, 10 },
+    { 238, 11 },  { 239, 16 },  { 240, 17 },  { 241, 18 },
+    { 242, 19 },  { 243, 20 },  { 244, 21 },  { 245, 22 },
+    { 246, 23 },  { 247, 24 },  { 248, 25 },  { 249, 28 },
+    { 250, 29 },  { 251, 30 },  { 252, 31 },  { 253, 32 },
+    { 254, 33 },  { 255, 34 },  { 256, 35 },  { 257, 36 },
+    { 258, 37 },  { 259, 38 },  { 260, 39 },  { 261, 40 },
+    { 262, 41 },  { 263, 42 },  { 264, 43 },  { 265, 44 },
+    { 266, 45 },  { 267, 48 },  { 268, 49 },  { 269, 50 },
+    { 270, 51 },  { 271, 52 },  { 272, 53 },  { 273, 54 },
+    { 274, 55 },  { 275, 56 },  { 276, 57 },  { 277, 58 },
+    { 278, 59 },  { 279, 60 },  { 280, 61 },  { 281, 62 },
+    { 282, 63 },  { 283, 64 },  { 284, 65 },  { 285, 66 },
+    { 286, 67 },  { 287, 68 },  { 288, 69 },  { 289, 70 },
+    { 290, 71 },  { 291, 72 },  { 292, 73 },  { 293, 74 },
+    { 294, 75 },  { 295, 76 },  { 296, 77 },  { 297, 78 },
+    { 298, 79 },  { 299, 80 },  { 300, 81 },  { 301, 82 },
+    { 302, 83 },  { 303, 84 },  { 304, 85 },  { 305, 86 },
+    { 306, 87 },  { 307, 88 },  { 308, 89 },  { 309, 90 },
+    { 310, 91 },  { 311, 92 },  { 312, 93 },  { 313, 94 },
+    { 314, 95 },  { 315, 96 },  { 316, 97 },  { 317, 98 },
+    { 318, 99 },  { 319, 103 }, { 320, 104 }, { 321, 105 },
+    { 322, 106 }, { 323, 107 }, { 324, 108 }, { 325, 109 },
+    { 326, 110 }, { 327, 114 }, { 328, 115 }, { 329, 116 },
+    { 330, 117 }, { 331, 118 }, { 332, 119 }, { 333, 120 },
+    { 334, 121 }, { 335, 122 }, { 336, 123 }, { 337, 124 },
+    { 338, 125 }, { 339, 126 }, { 340, 127 }, { 341, 128 },
+    { 342, 129 }, { 343, 130 }, { 344, 131 }, { 345, 132 },
+    { 346, 133 }, { 347, 134 }, { 348, 135 }, { 349, 136 },
+    { 350, 137 }, { 351, 138 }, { 352, 139 }, { 353, 140 },
+    { 354, 141 }, { 355, 142 }, { 356, 143 }, { 357, 144 },
+    { 358, 145 }, { 359, 146 }, { 360, 147 }, { 361, 148 },
+    { 362, 149 }, { 363, 150 }, { 364, 151 }, { 365, 152 },
+    { 366, 153 }, { 367, 154 }, { 368, 155 }, { 369, 156 },
+    { 370, 157 }, { 371, 158 }, { 372, 159 }, { 373, 160 },
+    { 374, 161 }, { 375, 162 }, { 376, 163 }, { 377, 164 },
+    { 378, 165 }
+};
+
+/* SID to glyph ID */
+static const sid_to_gid_t expert_subset_charset_sid_to_gid [] =
+{
+  { 1, 1 },       { 13, 8 },      { 14, 9 },      { 15, 10 },
+  { 27, 22 },     { 28, 23 },     { 99, 11 },     { 109, 41 },
+  { 110, 42 },    { 150, 64 },    { 155, 55 },    { 158, 54 },
+  { 163, 56 },    { 164, 65 },    { 169, 66 },    { 231, 2 },
+  { 232, 3 },     { 235, 4 },     { 236, 5 },     { 237, 6 },
+  { 238, 7 },     { 239, 12 },    { 240, 13 },    { 241, 14 },
+  { 242, 15 },    { 243, 16 },    { 244, 17 },    { 245, 18 },
+  { 246, 19 },    { 247, 20 },    { 248, 21 },    { 249, 24 },
+  { 250, 25 },    { 251, 26 },    { 253, 27 },    { 254, 28 },
+  { 255, 29 },    { 256, 30 },    { 257, 31 },    { 258, 32 },
+  { 259, 33 },    { 260, 34 },    { 261, 35 },    { 262, 36 },
+  { 263, 37 },    { 264, 38 },    { 265, 39 },    { 266, 40 },
+  { 267, 43 },    { 268, 44 },    { 269, 45 },    { 270, 46 },
+  { 272, 47 },    { 300, 48 },    { 301, 49 },    { 302, 50 },
+  { 305, 51 },    { 314, 52 },    { 315, 53 },    { 320, 57 },
+  { 321, 58 },    { 322, 59 },    { 323, 60 },    { 324, 61 },
+  { 325, 62 },    { 326, 63 },    { 327, 67 },    { 328, 68 },
+  { 329, 69 },    { 330, 70 },    { 331, 71 },    { 332, 72 },
+  { 333, 73 },    { 334, 74 },    { 335, 75 },    { 336, 76 },
+  { 337, 77 },    { 338, 78 },    { 339, 79 },    { 340, 80 },
+  { 341, 81 },    { 342, 82 },    { 343, 83 },    { 344, 84 },
+  { 345, 85 },    { 346, 86 }
+};
+
+/* code to SID */
+static const uint8_t standard_encoding_to_sid [] =
+{
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    1,    2,    3,    4,    5,    6,    7,    8,    9,   10,   11,   12,   13,   14,   15,   16,
+    17,  18,   19,   20,   21,   22,   23,   24,   25,   26,   27,   28,   29,   30,   31,   32,
+    33,  34,   35,   36,   37,   38,   39,   40,   41,   42,   43,   44,   45,   46,   47,   48,
+    49,  50,   51,   52,   53,   54,   55,   56,   57,   58,   59,   60,   61,   62,   63,   64,
+    65,  66,   67,   68,   69,   70,   71,   72,   73,   74,   75,   76,   77,   78,   79,   80,
+    81,  82,   83,   84,   85,   86,   87,   88,   89,   90,   91,   92,   93,   94,   95,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,   96,   97,   98,   99,  100,  101,  102,  103,  104,  105,  106,  107,  108,  109,  110,
+    0,  111,  112,  113,  114,    0,  115,  116,  117,  118,  119,  120,  121,  122,    0,  123,
+    0,  124,  125,  126,  127,  128,  129,  130,  131,    0,  132,  133,    0,  134,  135,  136,
+  137,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+    0,   138,   0,  139,    0,    0,    0,    0,  140,  141,  142,  143,    0,    0,    0,    0,
+    0,   144,   0,    0,    0,  145,    0,    0,  146,  147,  148,  149,    0,    0,    0,    0
+};
+
+hb_codepoint_t OT::cff1::lookup_standard_encoding_for_code (hb_codepoint_t sid)
+{
+  if (sid < ARRAY_LENGTH (standard_encoding_to_code))
+    return (hb_codepoint_t)standard_encoding_to_code[sid];
+  else
+    return 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_expert_encoding_for_code (hb_codepoint_t sid)
+{
+  if (sid < ARRAY_LENGTH (expert_encoding_to_code))
+    return (hb_codepoint_t)expert_encoding_to_code[sid];
+  else
+    return 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_expert_charset_for_sid (hb_codepoint_t glyph)
+{
+  if (glyph < ARRAY_LENGTH (expert_charset_to_sid))
+    return (hb_codepoint_t)expert_charset_to_sid[glyph];
+  else
+    return 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_expert_subset_charset_for_sid (hb_codepoint_t glyph)
+{
+  if (glyph < ARRAY_LENGTH (expert_subset_charset_to_sid))
+    return (hb_codepoint_t)expert_subset_charset_to_sid[glyph];
+  else
+    return 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_expert_charset_for_glyph (hb_codepoint_t sid)
+{
+  const auto *pair = hb_sorted_array (expert_charset_sid_to_gid).bsearch (sid);
+  return pair ? pair->gid : 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_expert_subset_charset_for_glyph (hb_codepoint_t sid)
+{
+  const auto *pair = hb_sorted_array (expert_subset_charset_sid_to_gid).bsearch (sid);
+  return pair ? pair->gid : 0;
+}
+
+hb_codepoint_t OT::cff1::lookup_standard_encoding_for_sid (hb_codepoint_t code)
+{
+  if (code < ARRAY_LENGTH (standard_encoding_to_sid))
+    return (hb_codepoint_t)standard_encoding_to_sid[code];
+  else
+    return CFF_UNDEF_SID;
+}
+
+struct bounds_t
+{
+  void init ()
+  {
+    min.set_int (INT_MAX, INT_MAX);
+    max.set_int (INT_MIN, INT_MIN);
+  }
+
+  void update (const point_t &pt)
+  {
+    if (pt.x < min.x) min.x = pt.x;
+    if (pt.x > max.x) max.x = pt.x;
+    if (pt.y < min.y) min.y = pt.y;
+    if (pt.y > max.y) max.y = pt.y;
+  }
+
+  void merge (const bounds_t &b)
+  {
+    if (empty ())
+      *this = b;
+    else if (!b.empty ())
+    {
+      if (b.min.x < min.x) min.x = b.min.x;
+      if (b.max.x > max.x) max.x = b.max.x;
+      if (b.min.y < min.y) min.y = b.min.y;
+      if (b.max.y > max.y) max.y = b.max.y;
+    }
+  }
+
+  void offset (const point_t &delta)
+  {
+    if (!empty ())
+    {
+      min.move (delta);
+      max.move (delta);
+    }
+  }
+
+  bool empty () const { return (min.x >= max.x) || (min.y >= max.y); }
+
+  point_t min;
+  point_t max;
+};
+
+struct cff1_extents_param_t
+{
+  void init (const OT::cff1::accelerator_t *_cff)
+  {
+    path_open = false;
+    cff = _cff;
+    bounds.init ();
+  }
+
+  void start_path   ()       { path_open = true; }
+  void end_path     ()       { path_open = false; }
+  bool is_path_open () const { return path_open; }
+
+  bool path_open;
+  bounds_t bounds;
+
+  const OT::cff1::accelerator_t *cff;
+};
+
+struct cff1_path_procs_extents_t : path_procs_t<cff1_path_procs_extents_t, cff1_cs_interp_env_t, cff1_extents_param_t>
+{
+  static void moveto (cff1_cs_interp_env_t &env, cff1_extents_param_t& param, const point_t &pt)
+  {
+    param.end_path ();
+    env.moveto (pt);
+  }
+
+  static void line (cff1_cs_interp_env_t &env, cff1_extents_param_t& param, const point_t &pt1)
+  {
+    if (!param.is_path_open ())
+    {
+      param.start_path ();
+      param.bounds.update (env.get_pt ());
+    }
+    env.moveto (pt1);
+    param.bounds.update (env.get_pt ());
+  }
+
+  static void curve (cff1_cs_interp_env_t &env, cff1_extents_param_t& param, const point_t &pt1, const point_t &pt2, const point_t &pt3)
+  {
+    if (!param.is_path_open ())
+    {
+      param.start_path ();
+      param.bounds.update (env.get_pt ());
+    }
+    /* include control points */
+    param.bounds.update (pt1);
+    param.bounds.update (pt2);
+    env.moveto (pt3);
+    param.bounds.update (env.get_pt ());
+  }
+};
+
+static bool _get_bounds (const OT::cff1::accelerator_t *cff, hb_codepoint_t glyph, bounds_t &bounds, bool in_seac=false);
+
+struct cff1_cs_opset_extents_t : cff1_cs_opset_t<cff1_cs_opset_extents_t, cff1_extents_param_t, cff1_path_procs_extents_t>
+{
+  static void process_seac (cff1_cs_interp_env_t &env, cff1_extents_param_t& param)
+  {
+    unsigned int  n = env.argStack.get_count ();
+    point_t delta;
+    delta.x = env.argStack[n-4];
+    delta.y = env.argStack[n-3];
+    hb_codepoint_t base = param.cff->std_code_to_glyph (env.argStack[n-2].to_int ());
+    hb_codepoint_t accent = param.cff->std_code_to_glyph (env.argStack[n-1].to_int ());
+
+    bounds_t  base_bounds, accent_bounds;
+    if (likely (!env.in_seac && base && accent
+	       && _get_bounds (param.cff, base, base_bounds, true)
+	       && _get_bounds (param.cff, accent, accent_bounds, true)))
+    {
+      param.bounds.merge (base_bounds);
+      accent_bounds.offset (delta);
+      param.bounds.merge (accent_bounds);
+    }
+    else
+      env.set_error ();
+  }
+};
+
+bool _get_bounds (const OT::cff1::accelerator_t *cff, hb_codepoint_t glyph, bounds_t &bounds, bool in_seac)
+{
+  bounds.init ();
+  if (unlikely (!cff->is_valid () || (glyph >= cff->num_glyphs))) return false;
+
+  unsigned int fd = cff->fdSelect->get_fd (glyph);
+  cff1_cs_interpreter_t<cff1_cs_opset_extents_t, cff1_extents_param_t> interp;
+  const byte_str_t str = (*cff->charStrings)[glyph];
+  interp.env.init (str, *cff, fd);
+  interp.env.set_in_seac (in_seac);
+  cff1_extents_param_t  param;
+  param.init (cff);
+  if (unlikely (!interp.interpret (param))) return false;
+  bounds = param.bounds;
+  return true;
+}
+
+bool OT::cff1::accelerator_t::get_extents (hb_font_t *font, hb_codepoint_t glyph, hb_glyph_extents_t *extents) const
+{
+#ifdef HB_NO_OT_FONT_CFF
+  /* XXX Remove check when this code moves to .hh file. */
+  return true;
+#endif
+
+  bounds_t bounds;
+
+  if (!_get_bounds (this, glyph, bounds))
+    return false;
+
+  if (bounds.min.x >= bounds.max.x)
+  {
+    extents->width = 0;
+    extents->x_bearing = 0;
+  }
+  else
+  {
+    extents->x_bearing = font->em_scalef_x (bounds.min.x.to_real ());
+    extents->width = font->em_scalef_x (bounds.max.x.to_real () - bounds.min.x.to_real ());
+  }
+  if (bounds.min.y >= bounds.max.y)
+  {
+    extents->height = 0;
+    extents->y_bearing = 0;
+  }
+  else
+  {
+    extents->y_bearing = font->em_scalef_y (bounds.max.y.to_real ());
+    extents->height = font->em_scalef_y (bounds.min.y.to_real () - bounds.max.y.to_real ());
+  }
+
+  return true;
+}
+
+#ifdef HB_EXPERIMENTAL_API
+struct cff1_path_param_t
+{
+  cff1_path_param_t (const OT::cff1::accelerator_t *cff_, hb_font_t *font_,
+		     draw_helper_t &draw_helper_, point_t *delta_)
+  {
+    draw_helper = &draw_helper_;
+    cff = cff_;
+    font = font_;
+    delta = delta_;
+  }
+
+  void move_to (const point_t &p)
+  {
+    point_t point = p;
+    if (delta) point.move (*delta);
+    draw_helper->move_to (font->em_scalef_x (point.x.to_real ()), font->em_scalef_y (point.y.to_real ()));
+  }
+
+  void line_to (const point_t &p)
+  {
+    point_t point = p;
+    if (delta) point.move (*delta);
+    draw_helper->line_to (font->em_scalef_x (point.x.to_real ()), font->em_scalef_y (point.y.to_real ()));
+  }
+
+  void cubic_to (const point_t &p1, const point_t &p2, const point_t &p3)
+  {
+    point_t point1 = p1, point2 = p2, point3 = p3;
+    if (delta)
+    {
+      point1.move (*delta);
+      point2.move (*delta);
+      point3.move (*delta);
+    }
+    draw_helper->cubic_to (font->em_scalef_x (point1.x.to_real ()), font->em_scalef_y (point1.y.to_real ()),
+			   font->em_scalef_x (point2.x.to_real ()), font->em_scalef_y (point2.y.to_real ()),
+			   font->em_scalef_x (point3.x.to_real ()), font->em_scalef_y (point3.y.to_real ()));
+  }
+
+  void end_path () { draw_helper->end_path (); }
+
+  hb_font_t *font;
+  draw_helper_t *draw_helper;
+  point_t *delta;
+
+  const OT::cff1::accelerator_t *cff;
+};
+
+struct cff1_path_procs_path_t : path_procs_t<cff1_path_procs_path_t, cff1_cs_interp_env_t, cff1_path_param_t>
+{
+  static void moveto (cff1_cs_interp_env_t &env, cff1_path_param_t& param, const point_t &pt)
+  {
+    param.move_to (pt);
+    env.moveto (pt);
+  }
+
+  static void line (cff1_cs_interp_env_t &env, cff1_path_param_t &param, const point_t &pt1)
+  {
+    param.line_to (pt1);
+    env.moveto (pt1);
+  }
+
+  static void curve (cff1_cs_interp_env_t &env, cff1_path_param_t &param, const point_t &pt1, const point_t &pt2, const point_t &pt3)
+  {
+    param.cubic_to (pt1, pt2, pt3);
+    env.moveto (pt3);
+  }
+};
+
+static bool _get_path (const OT::cff1::accelerator_t *cff, hb_font_t *font, hb_codepoint_t glyph,
+		       draw_helper_t &draw_helper, bool in_seac = false, point_t *delta = nullptr);
+
+struct cff1_cs_opset_path_t : cff1_cs_opset_t<cff1_cs_opset_path_t, cff1_path_param_t, cff1_path_procs_path_t>
+{
+  static void process_seac (cff1_cs_interp_env_t &env, cff1_path_param_t& param)
+  {
+    /* End previous path */
+    param.end_path ();
+
+    unsigned int n = env.argStack.get_count ();
+    point_t delta;
+    delta.x = env.argStack[n-4];
+    delta.y = env.argStack[n-3];
+    hb_codepoint_t base = param.cff->std_code_to_glyph (env.argStack[n-2].to_int ());
+    hb_codepoint_t accent = param.cff->std_code_to_glyph (env.argStack[n-1].to_int ());
+
+    if (unlikely (!(!env.in_seac && base && accent
+		    && _get_path (param.cff, param.font, base, *param.draw_helper, true)
+		    && _get_path (param.cff, param.font, accent, *param.draw_helper, true, &delta))))
+      env.set_error ();
+  }
+};
+
+bool _get_path (const OT::cff1::accelerator_t *cff, hb_font_t *font, hb_codepoint_t glyph,
+		draw_helper_t &draw_helper, bool in_seac, point_t *delta)
+{
+  if (unlikely (!cff->is_valid () || (glyph >= cff->num_glyphs))) return false;
+
+  unsigned int fd = cff->fdSelect->get_fd (glyph);
+  cff1_cs_interpreter_t<cff1_cs_opset_path_t, cff1_path_param_t> interp;
+  const byte_str_t str = (*cff->charStrings)[glyph];
+  interp.env.init (str, *cff, fd);
+  interp.env.set_in_seac (in_seac);
+  cff1_path_param_t param (cff, font, draw_helper, delta);
+  if (unlikely (!interp.interpret (param))) return false;
+
+  /* Let's end the path specially since it is called inside seac also */
+  param.end_path ();
+
+  return true;
+}
+
+bool OT::cff1::accelerator_t::get_path (hb_font_t *font, hb_codepoint_t glyph, draw_helper_t &draw_helper) const
+{
+#ifdef HB_NO_OT_FONT_CFF
+  /* XXX Remove check when this code moves to .hh file. */
+  return true;
+#endif
+
+  return _get_path (this, font, glyph, draw_helper);
+}
+#endif
+
+struct get_seac_param_t
+{
+  void init (const OT::cff1::accelerator_t *_cff)
+  {
+    cff = _cff;
+    base = 0;
+    accent = 0;
+  }
+
+  bool has_seac () const { return base && accent; }
+
+  const OT::cff1::accelerator_t *cff;
+  hb_codepoint_t  base;
+  hb_codepoint_t  accent;
+};
+
+struct cff1_cs_opset_seac_t : cff1_cs_opset_t<cff1_cs_opset_seac_t, get_seac_param_t>
+{
+  static void process_seac (cff1_cs_interp_env_t &env, get_seac_param_t& param)
+  {
+    unsigned int  n = env.argStack.get_count ();
+    hb_codepoint_t  base_char = (hb_codepoint_t)env.argStack[n-2].to_int ();
+    hb_codepoint_t  accent_char = (hb_codepoint_t)env.argStack[n-1].to_int ();
+
+    param.base = param.cff->std_code_to_glyph (base_char);
+    param.accent = param.cff->std_code_to_glyph (accent_char);
+  }
+};
+
+bool OT::cff1::accelerator_t::get_seac_components (hb_codepoint_t glyph, hb_codepoint_t *base, hb_codepoint_t *accent) const
+{
+  if (unlikely (!is_valid () || (glyph >= num_glyphs))) return false;
+
+  unsigned int fd = fdSelect->get_fd (glyph);
+  cff1_cs_interpreter_t<cff1_cs_opset_seac_t, get_seac_param_t> interp;
+  const byte_str_t str = (*charStrings)[glyph];
+  interp.env.init (str, *this, fd);
+  get_seac_param_t  param;
+  param.init (this);
+  if (unlikely (!interp.interpret (param))) return false;
+
+  if (param.has_seac ())
+  {
+    *base = param.base;
+    *accent = param.accent;
+    return true;
+  }
+  return false;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff1-table.hh b/thirdparty/harfbuzz/src/hb-ot-cff1-table.hh
new file mode 100644
index 0000000000..7228f77727
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff1-table.hh
@@ -0,0 +1,1403 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_CFF1_TABLE_HH
+#define HB_OT_CFF1_TABLE_HH
+
+#include "hb-ot-cff-common.hh"
+#include "hb-subset-cff1.hh"
+#include "hb-draw.hh"
+
+#define HB_STRING_ARRAY_NAME cff1_std_strings
+#define HB_STRING_ARRAY_LIST "hb-ot-cff1-std-str.hh"
+#include "hb-string-array.hh"
+#undef HB_STRING_ARRAY_LIST
+#undef HB_STRING_ARRAY_NAME
+
+namespace CFF {
+
+/*
+ * CFF -- Compact Font Format (CFF)
+ * https://www.adobe.com/content/dam/acom/en/devnet/font/pdfs/5176.CFF.pdf
+ */
+#define HB_OT_TAG_cff1 HB_TAG('C','F','F',' ')
+
+#define CFF_UNDEF_SID   CFF_UNDEF_CODE
+
+enum EncodingID { StandardEncoding = 0, ExpertEncoding = 1 };
+enum CharsetID { ISOAdobeCharset = 0, ExpertCharset = 1, ExpertSubsetCharset = 2 };
+
+typedef CFFIndex<HBUINT16>  CFF1Index;
+template <typename Type> struct CFF1IndexOf : CFFIndexOf<HBUINT16, Type> {};
+
+typedef CFFIndex<HBUINT16> CFF1Index;
+typedef CFF1Index          CFF1CharStrings;
+typedef Subrs<HBUINT16>    CFF1Subrs;
+
+struct CFF1FDSelect : FDSelect {};
+
+/* Encoding */
+struct Encoding0 {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (codes.sanitize (c));
+  }
+
+  hb_codepoint_t get_code (hb_codepoint_t glyph) const
+  {
+    assert (glyph > 0);
+    glyph--;
+    if (glyph < nCodes ())
+    {
+      return (hb_codepoint_t)codes[glyph];
+    }
+    else
+      return CFF_UNDEF_CODE;
+  }
+
+  HBUINT8 &nCodes () { return codes.len; }
+  HBUINT8 nCodes () const { return codes.len; }
+
+  ArrayOf<HBUINT8, HBUINT8> codes;
+
+  DEFINE_SIZE_ARRAY_SIZED (1, codes);
+};
+
+struct Encoding1_Range {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT8   first;
+  HBUINT8   nLeft;
+
+  DEFINE_SIZE_STATIC (2);
+};
+
+struct Encoding1 {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (ranges.sanitize (c));
+  }
+
+  hb_codepoint_t get_code (hb_codepoint_t glyph) const
+  {
+    assert (glyph > 0);
+    glyph--;
+    for (unsigned int i = 0; i < nRanges (); i++)
+    {
+      if (glyph <= ranges[i].nLeft)
+      {
+	hb_codepoint_t code = (hb_codepoint_t) ranges[i].first + glyph;
+	return (likely (code < 0x100) ? code: CFF_UNDEF_CODE);
+      }
+      glyph -= (ranges[i].nLeft + 1);
+    }
+    return CFF_UNDEF_CODE;
+  }
+
+  HBUINT8 &nRanges () { return ranges.len; }
+  HBUINT8 nRanges () const { return ranges.len; }
+
+  ArrayOf<Encoding1_Range, HBUINT8> ranges;
+
+  DEFINE_SIZE_ARRAY_SIZED (1, ranges);
+};
+
+struct SuppEncoding {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT8   code;
+  HBUINT16  glyph;
+
+  DEFINE_SIZE_STATIC (3);
+};
+
+struct CFF1SuppEncData {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (supps.sanitize (c));
+  }
+
+  void get_codes (hb_codepoint_t sid, hb_vector_t<hb_codepoint_t> &codes) const
+  {
+    for (unsigned int i = 0; i < nSups (); i++)
+      if (sid == supps[i].glyph)
+	codes.push (supps[i].code);
+  }
+
+  HBUINT8 &nSups () { return supps.len; }
+  HBUINT8 nSups () const { return supps.len; }
+
+  ArrayOf<SuppEncoding, HBUINT8> supps;
+
+  DEFINE_SIZE_ARRAY_SIZED (1, supps);
+};
+
+struct Encoding
+{
+  /* serialize a fullset Encoding */
+  bool serialize (hb_serialize_context_t *c, const Encoding &src)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = src.get_size ();
+    Encoding *dest = c->allocate_size<Encoding> (size);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, &src, size);
+    return_trace (true);
+  }
+
+  /* serialize a subset Encoding */
+  bool serialize (hb_serialize_context_t *c,
+		  uint8_t format,
+		  unsigned int enc_count,
+		  const hb_vector_t<code_pair_t>& code_ranges,
+		  const hb_vector_t<code_pair_t>& supp_codes)
+  {
+    TRACE_SERIALIZE (this);
+    Encoding *dest = c->extend_min (*this);
+    if (unlikely (!dest)) return_trace (false);
+    dest->format = format | ((supp_codes.length > 0) ? 0x80 : 0);
+    switch (format) {
+    case 0:
+    {
+      Encoding0 *fmt0 = c->allocate_size<Encoding0> (Encoding0::min_size + HBUINT8::static_size * enc_count);
+      if (unlikely (!fmt0)) return_trace (false);
+      fmt0->nCodes () = enc_count;
+      unsigned int glyph = 0;
+      for (unsigned int i = 0; i < code_ranges.length; i++)
+      {
+	hb_codepoint_t code = code_ranges[i].code;
+	for (int left = (int)code_ranges[i].glyph; left >= 0; left--)
+	  fmt0->codes[glyph++] = code++;
+	if (unlikely (!((glyph <= 0x100) && (code <= 0x100))))
+	  return_trace (false);
+      }
+    }
+    break;
+
+    case 1:
+    {
+      Encoding1 *fmt1 = c->allocate_size<Encoding1> (Encoding1::min_size + Encoding1_Range::static_size * code_ranges.length);
+      if (unlikely (!fmt1)) return_trace (false);
+      fmt1->nRanges () = code_ranges.length;
+      for (unsigned int i = 0; i < code_ranges.length; i++)
+      {
+	if (unlikely (!((code_ranges[i].code <= 0xFF) && (code_ranges[i].glyph <= 0xFF))))
+	  return_trace (false);
+	fmt1->ranges[i].first = code_ranges[i].code;
+	fmt1->ranges[i].nLeft = code_ranges[i].glyph;
+      }
+    }
+    break;
+
+    }
+
+    if (supp_codes.length)
+    {
+      CFF1SuppEncData *suppData = c->allocate_size<CFF1SuppEncData> (CFF1SuppEncData::min_size + SuppEncoding::static_size * supp_codes.length);
+      if (unlikely (!suppData)) return_trace (false);
+      suppData->nSups () = supp_codes.length;
+      for (unsigned int i = 0; i < supp_codes.length; i++)
+      {
+	suppData->supps[i].code = supp_codes[i].code;
+	suppData->supps[i].glyph = supp_codes[i].glyph; /* actually SID */
+      }
+    }
+
+    return_trace (true);
+  }
+
+  unsigned int get_size () const
+  {
+    unsigned int size = min_size;
+    switch (table_format ())
+    {
+    case 0: size += u.format0.get_size (); break;
+    case 1: size += u.format1.get_size (); break;
+    }
+    if (has_supplement ())
+      size += suppEncData ().get_size ();
+    return size;
+  }
+
+  hb_codepoint_t get_code (hb_codepoint_t glyph) const
+  {
+    switch (table_format ())
+    {
+    case 0: return u.format0.get_code (glyph);
+    case 1: return u.format1.get_code (glyph);
+    default:return 0;
+    }
+  }
+
+  uint8_t table_format () const { return format & 0x7F; }
+  bool  has_supplement () const { return format & 0x80; }
+
+  void get_supplement_codes (hb_codepoint_t sid, hb_vector_t<hb_codepoint_t> &codes) const
+  {
+    codes.resize (0);
+    if (has_supplement ())
+      suppEncData().get_codes (sid, codes);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (table_format ())
+    {
+    case 0: if (unlikely (!u.format0.sanitize (c))) { return_trace (false); } break;
+    case 1: if (unlikely (!u.format1.sanitize (c))) { return_trace (false); } break;
+    default:return_trace (false);
+    }
+    return_trace (likely (!has_supplement () || suppEncData ().sanitize (c)));
+  }
+
+  protected:
+  const CFF1SuppEncData &suppEncData () const
+  {
+    switch (table_format ())
+    {
+    case 0: return StructAfter<CFF1SuppEncData> (u.format0.codes[u.format0.nCodes ()-1]);
+    case 1: return StructAfter<CFF1SuppEncData> (u.format1.ranges[u.format1.nRanges ()-1]);
+    default:return Null (CFF1SuppEncData);
+    }
+  }
+
+  public:
+  HBUINT8	format;
+  union {
+  Encoding0	format0;
+  Encoding1	format1;
+  } u;
+  /* CFF1SuppEncData  suppEncData; */
+
+  DEFINE_SIZE_MIN (1);
+};
+
+/* Charset */
+struct Charset0 {
+  bool sanitize (hb_sanitize_context_t *c, unsigned int num_glyphs) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && sids[num_glyphs - 1].sanitize (c));
+  }
+
+  hb_codepoint_t get_sid (hb_codepoint_t glyph) const
+  {
+    if (glyph == 0)
+      return 0;
+    else
+      return sids[glyph - 1];
+  }
+
+  hb_codepoint_t get_glyph (hb_codepoint_t sid, unsigned int num_glyphs) const
+  {
+    if (sid == 0)
+      return 0;
+
+    for (unsigned int glyph = 1; glyph < num_glyphs; glyph++)
+    {
+      if (sids[glyph-1] == sid)
+	return glyph;
+    }
+    return 0;
+  }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  {
+    assert (num_glyphs > 0);
+    return HBUINT16::static_size * (num_glyphs - 1);
+  }
+
+  HBUINT16  sids[HB_VAR_ARRAY];
+
+  DEFINE_SIZE_ARRAY(0, sids);
+};
+
+template <typename TYPE>
+struct Charset_Range {
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT16  first;
+  TYPE      nLeft;
+
+  DEFINE_SIZE_STATIC (HBUINT16::static_size + TYPE::static_size);
+};
+
+template <typename TYPE>
+struct Charset1_2 {
+  bool sanitize (hb_sanitize_context_t *c, unsigned int num_glyphs) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+    num_glyphs--;
+    for (unsigned int i = 0; num_glyphs > 0; i++)
+    {
+      if (unlikely (!ranges[i].sanitize (c) || (num_glyphs < ranges[i].nLeft + 1)))
+	return_trace (false);
+      num_glyphs -= (ranges[i].nLeft + 1);
+    }
+    return_trace (true);
+  }
+
+  hb_codepoint_t get_sid (hb_codepoint_t glyph) const
+  {
+    if (glyph == 0) return 0;
+    glyph--;
+    for (unsigned int i = 0;; i++)
+    {
+      if (glyph <= ranges[i].nLeft)
+	return (hb_codepoint_t)ranges[i].first + glyph;
+      glyph -= (ranges[i].nLeft + 1);
+    }
+
+    return 0;
+  }
+
+  hb_codepoint_t get_glyph (hb_codepoint_t sid, unsigned int num_glyphs) const
+  {
+    if (sid == 0) return 0;
+    hb_codepoint_t  glyph = 1;
+    for (unsigned int i = 0;; i++)
+    {
+      if (glyph >= num_glyphs)
+	return 0;
+      if ((ranges[i].first <= sid) && (sid <= ranges[i].first + ranges[i].nLeft))
+	return glyph + (sid - ranges[i].first);
+      glyph += (ranges[i].nLeft + 1);
+    }
+
+    return 0;
+  }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  {
+    unsigned int size = HBUINT8::static_size;
+    int glyph = (int)num_glyphs;
+
+    assert (glyph > 0);
+    glyph--;
+    for (unsigned int i = 0; glyph > 0; i++)
+    {
+      glyph -= (ranges[i].nLeft + 1);
+      size += Charset_Range<TYPE>::static_size;
+    }
+
+    return size;
+  }
+
+  Charset_Range<TYPE>   ranges[HB_VAR_ARRAY];
+
+  DEFINE_SIZE_ARRAY (0, ranges);
+};
+
+typedef Charset1_2<HBUINT8>     Charset1;
+typedef Charset1_2<HBUINT16>    Charset2;
+typedef Charset_Range<HBUINT8>  Charset1_Range;
+typedef Charset_Range<HBUINT16> Charset2_Range;
+
+struct Charset
+{
+  /* serialize a fullset Charset */
+  bool serialize (hb_serialize_context_t *c, const Charset &src, unsigned int num_glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = src.get_size (num_glyphs);
+    Charset *dest = c->allocate_size<Charset> (size);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, &src, size);
+    return_trace (true);
+  }
+
+  /* serialize a subset Charset */
+  bool serialize (hb_serialize_context_t *c,
+		  uint8_t format,
+		  unsigned int num_glyphs,
+		  const hb_vector_t<code_pair_t>& sid_ranges)
+  {
+    TRACE_SERIALIZE (this);
+    Charset *dest = c->extend_min (*this);
+    if (unlikely (!dest)) return_trace (false);
+    dest->format = format;
+    switch (format)
+    {
+    case 0:
+    {
+      Charset0 *fmt0 = c->allocate_size<Charset0> (Charset0::min_size + HBUINT16::static_size * (num_glyphs - 1));
+      if (unlikely (!fmt0)) return_trace (false);
+      unsigned int glyph = 0;
+      for (unsigned int i = 0; i < sid_ranges.length; i++)
+      {
+	hb_codepoint_t sid = sid_ranges[i].code;
+	for (int left = (int)sid_ranges[i].glyph; left >= 0; left--)
+	  fmt0->sids[glyph++] = sid++;
+      }
+    }
+    break;
+
+    case 1:
+    {
+      Charset1 *fmt1 = c->allocate_size<Charset1> (Charset1::min_size + Charset1_Range::static_size * sid_ranges.length);
+      if (unlikely (!fmt1)) return_trace (false);
+      for (unsigned int i = 0; i < sid_ranges.length; i++)
+      {
+	if (unlikely (!(sid_ranges[i].glyph <= 0xFF)))
+	  return_trace (false);
+	fmt1->ranges[i].first = sid_ranges[i].code;
+	fmt1->ranges[i].nLeft = sid_ranges[i].glyph;
+      }
+    }
+    break;
+
+    case 2:
+    {
+      Charset2 *fmt2 = c->allocate_size<Charset2> (Charset2::min_size + Charset2_Range::static_size * sid_ranges.length);
+      if (unlikely (!fmt2)) return_trace (false);
+      for (unsigned int i = 0; i < sid_ranges.length; i++)
+      {
+	if (unlikely (!(sid_ranges[i].glyph <= 0xFFFF)))
+	  return_trace (false);
+	fmt2->ranges[i].first = sid_ranges[i].code;
+	fmt2->ranges[i].nLeft = sid_ranges[i].glyph;
+      }
+    }
+    break;
+
+    }
+    return_trace (true);
+  }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  {
+    switch (format)
+    {
+    case 0: return min_size + u.format0.get_size (num_glyphs);
+    case 1: return min_size + u.format1.get_size (num_glyphs);
+    case 2: return min_size + u.format2.get_size (num_glyphs);
+    default:return 0;
+    }
+  }
+
+  hb_codepoint_t get_sid (hb_codepoint_t glyph, unsigned int num_glyphs) const
+  {
+    if (unlikely (glyph >= num_glyphs)) return 0;
+    switch (format)
+    {
+    case 0: return u.format0.get_sid (glyph);
+    case 1: return u.format1.get_sid (glyph);
+    case 2: return u.format2.get_sid (glyph);
+    default:return 0;
+    }
+  }
+
+  hb_codepoint_t get_glyph (hb_codepoint_t sid, unsigned int num_glyphs) const
+  {
+    switch (format)
+    {
+    case 0: return u.format0.get_glyph (sid, num_glyphs);
+    case 1: return u.format1.get_glyph (sid, num_glyphs);
+    case 2: return u.format2.get_glyph (sid, num_glyphs);
+    default:return 0;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (format)
+    {
+    case 0: return_trace (u.format0.sanitize (c, c->get_num_glyphs ()));
+    case 1: return_trace (u.format1.sanitize (c, c->get_num_glyphs ()));
+    case 2: return_trace (u.format2.sanitize (c, c->get_num_glyphs ()));
+    default:return_trace (false);
+    }
+  }
+
+  HBUINT8       format;
+  union {
+    Charset0    format0;
+    Charset1    format1;
+    Charset2    format2;
+  } u;
+
+  DEFINE_SIZE_MIN (1);
+};
+
+struct CFF1StringIndex : CFF1Index
+{
+  bool serialize (hb_serialize_context_t *c, const CFF1StringIndex &strings,
+		  const hb_inc_bimap_t &sidmap)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely ((strings.count == 0) || (sidmap.get_population () == 0)))
+    {
+      if (unlikely (!c->extend_min (this->count)))
+	return_trace (false);
+      count = 0;
+      return_trace (true);
+    }
+
+    byte_str_array_t bytesArray;
+    bytesArray.init ();
+    if (!bytesArray.resize (sidmap.get_population ()))
+      return_trace (false);
+    for (unsigned int i = 0; i < strings.count; i++)
+    {
+      hb_codepoint_t  j = sidmap[i];
+      if (j != HB_MAP_VALUE_INVALID)
+	bytesArray[j] = strings[i];
+    }
+
+    bool result = CFF1Index::serialize (c, bytesArray);
+    bytesArray.fini ();
+    return_trace (result);
+  }
+};
+
+struct cff1_top_dict_interp_env_t : num_interp_env_t
+{
+  cff1_top_dict_interp_env_t ()
+    : num_interp_env_t(), prev_offset(0), last_offset(0) {}
+
+  unsigned int prev_offset;
+  unsigned int last_offset;
+};
+
+struct name_dict_values_t
+{
+  enum name_dict_val_index_t
+  {
+      version,
+      notice,
+      copyright,
+      fullName,
+      familyName,
+      weight,
+      postscript,
+      fontName,
+      baseFontName,
+      registry,
+      ordering,
+
+      ValCount
+  };
+
+  void init ()
+  {
+    for (unsigned int i = 0; i < ValCount; i++)
+      values[i] = CFF_UNDEF_SID;
+  }
+
+  unsigned int& operator[] (unsigned int i)
+  { assert (i < ValCount); return values[i]; }
+
+  unsigned int operator[] (unsigned int i) const
+  { assert (i < ValCount); return values[i]; }
+
+  static enum name_dict_val_index_t name_op_to_index (op_code_t op)
+  {
+    switch (op) {
+      default: // can't happen - just make some compiler happy
+      case OpCode_version:
+	return version;
+      case OpCode_Notice:
+	return notice;
+      case OpCode_Copyright:
+	return copyright;
+      case OpCode_FullName:
+	return fullName;
+      case OpCode_FamilyName:
+	return familyName;
+      case OpCode_Weight:
+	return weight;
+      case OpCode_PostScript:
+	return postscript;
+      case OpCode_FontName:
+	return fontName;
+      case OpCode_BaseFontName:
+	return baseFontName;
+    }
+  }
+
+  unsigned int  values[ValCount];
+};
+
+struct cff1_top_dict_val_t : op_str_t
+{
+  unsigned int  last_arg_offset;
+};
+
+struct cff1_top_dict_values_t : top_dict_values_t<cff1_top_dict_val_t>
+{
+  void init ()
+  {
+    top_dict_values_t<cff1_top_dict_val_t>::init ();
+
+    nameSIDs.init ();
+    ros_supplement = 0;
+    cidCount = 8720;
+    EncodingOffset = 0;
+    CharsetOffset = 0;
+    FDSelectOffset = 0;
+    privateDictInfo.init ();
+  }
+  void fini () { top_dict_values_t<cff1_top_dict_val_t>::fini (); }
+
+  bool is_CID () const
+  { return nameSIDs[name_dict_values_t::registry] != CFF_UNDEF_SID; }
+
+  name_dict_values_t  nameSIDs;
+  unsigned int    ros_supplement_offset;
+  unsigned int    ros_supplement;
+  unsigned int    cidCount;
+
+  unsigned int    EncodingOffset;
+  unsigned int    CharsetOffset;
+  unsigned int    FDSelectOffset;
+  table_info_t    privateDictInfo;
+};
+
+struct cff1_top_dict_opset_t : top_dict_opset_t<cff1_top_dict_val_t>
+{
+  static void process_op (op_code_t op, cff1_top_dict_interp_env_t& env, cff1_top_dict_values_t& dictval)
+  {
+    cff1_top_dict_val_t  val;
+    val.last_arg_offset = (env.last_offset-1) - dictval.opStart;  /* offset to the last argument */
+
+    switch (op) {
+      case OpCode_version:
+      case OpCode_Notice:
+      case OpCode_Copyright:
+      case OpCode_FullName:
+      case OpCode_FamilyName:
+      case OpCode_Weight:
+      case OpCode_PostScript:
+      case OpCode_BaseFontName:
+	dictval.nameSIDs[name_dict_values_t::name_op_to_index (op)] = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_isFixedPitch:
+      case OpCode_ItalicAngle:
+      case OpCode_UnderlinePosition:
+      case OpCode_UnderlineThickness:
+      case OpCode_PaintType:
+      case OpCode_CharstringType:
+      case OpCode_UniqueID:
+      case OpCode_StrokeWidth:
+      case OpCode_SyntheticBase:
+      case OpCode_CIDFontVersion:
+      case OpCode_CIDFontRevision:
+      case OpCode_CIDFontType:
+      case OpCode_UIDBase:
+      case OpCode_FontBBox:
+      case OpCode_XUID:
+      case OpCode_BaseFontBlend:
+	env.clear_args ();
+	break;
+
+      case OpCode_CIDCount:
+	dictval.cidCount = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      case OpCode_ROS:
+	dictval.ros_supplement = env.argStack.pop_uint ();
+	dictval.nameSIDs[name_dict_values_t::ordering] = env.argStack.pop_uint ();
+	dictval.nameSIDs[name_dict_values_t::registry] = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      case OpCode_Encoding:
+	dictval.EncodingOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	if (unlikely (dictval.EncodingOffset == 0)) return;
+	break;
+
+      case OpCode_charset:
+	dictval.CharsetOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	if (unlikely (dictval.CharsetOffset == 0)) return;
+	break;
+
+      case OpCode_FDSelect:
+	dictval.FDSelectOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      case OpCode_Private:
+	dictval.privateDictInfo.offset = env.argStack.pop_uint ();
+	dictval.privateDictInfo.size = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	env.last_offset = env.str_ref.offset;
+	top_dict_opset_t<cff1_top_dict_val_t>::process_op (op, env, dictval);
+	/* Record this operand below if stack is empty, otherwise done */
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref, val);
+  }
+};
+
+struct cff1_font_dict_values_t : dict_values_t<op_str_t>
+{
+  void init ()
+  {
+    dict_values_t<op_str_t>::init ();
+    privateDictInfo.init ();
+    fontName = CFF_UNDEF_SID;
+  }
+  void fini () { dict_values_t<op_str_t>::fini (); }
+
+  table_info_t       privateDictInfo;
+  unsigned int    fontName;
+};
+
+struct cff1_font_dict_opset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, num_interp_env_t& env, cff1_font_dict_values_t& dictval)
+  {
+    switch (op) {
+      case OpCode_FontName:
+	dictval.fontName = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_FontMatrix:
+      case OpCode_PaintType:
+	env.clear_args ();
+	break;
+      case OpCode_Private:
+	dictval.privateDictInfo.offset = env.argStack.pop_uint ();
+	dictval.privateDictInfo.size = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	dict_opset_t::process_op (op, env);
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref);
+  }
+};
+
+template <typename VAL>
+struct cff1_private_dict_values_base_t : dict_values_t<VAL>
+{
+  void init ()
+  {
+    dict_values_t<VAL>::init ();
+    subrsOffset = 0;
+    localSubrs = &Null (CFF1Subrs);
+  }
+  void fini () { dict_values_t<VAL>::fini (); }
+
+  unsigned int      subrsOffset;
+  const CFF1Subrs    *localSubrs;
+};
+
+typedef cff1_private_dict_values_base_t<op_str_t> cff1_private_dict_values_subset_t;
+typedef cff1_private_dict_values_base_t<num_dict_val_t> cff1_private_dict_values_t;
+
+struct cff1_private_dict_opset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, num_interp_env_t& env, cff1_private_dict_values_t& dictval)
+  {
+    num_dict_val_t val;
+    val.init ();
+
+    switch (op) {
+      case OpCode_BlueValues:
+      case OpCode_OtherBlues:
+      case OpCode_FamilyBlues:
+      case OpCode_FamilyOtherBlues:
+      case OpCode_StemSnapH:
+      case OpCode_StemSnapV:
+	env.clear_args ();
+	break;
+      case OpCode_StdHW:
+      case OpCode_StdVW:
+      case OpCode_BlueScale:
+      case OpCode_BlueShift:
+      case OpCode_BlueFuzz:
+      case OpCode_ForceBold:
+      case OpCode_LanguageGroup:
+      case OpCode_ExpansionFactor:
+      case OpCode_initialRandomSeed:
+      case OpCode_defaultWidthX:
+      case OpCode_nominalWidthX:
+	val.single_val = env.argStack.pop_num ();
+	env.clear_args ();
+	break;
+      case OpCode_Subrs:
+	dictval.subrsOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	dict_opset_t::process_op (op, env);
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref, val);
+  }
+};
+
+struct cff1_private_dict_opset_subset : dict_opset_t
+{
+  static void process_op (op_code_t op, num_interp_env_t& env, cff1_private_dict_values_subset_t& dictval)
+  {
+    switch (op) {
+      case OpCode_BlueValues:
+      case OpCode_OtherBlues:
+      case OpCode_FamilyBlues:
+      case OpCode_FamilyOtherBlues:
+      case OpCode_StemSnapH:
+      case OpCode_StemSnapV:
+      case OpCode_StdHW:
+      case OpCode_StdVW:
+      case OpCode_BlueScale:
+      case OpCode_BlueShift:
+      case OpCode_BlueFuzz:
+      case OpCode_ForceBold:
+      case OpCode_LanguageGroup:
+      case OpCode_ExpansionFactor:
+      case OpCode_initialRandomSeed:
+      case OpCode_defaultWidthX:
+      case OpCode_nominalWidthX:
+	env.clear_args ();
+	break;
+
+      case OpCode_Subrs:
+	dictval.subrsOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	dict_opset_t::process_op (op, env);
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref);
+  }
+};
+
+typedef dict_interpreter_t<cff1_top_dict_opset_t, cff1_top_dict_values_t, cff1_top_dict_interp_env_t> cff1_top_dict_interpreter_t;
+typedef dict_interpreter_t<cff1_font_dict_opset_t, cff1_font_dict_values_t> cff1_font_dict_interpreter_t;
+
+typedef CFF1Index CFF1NameIndex;
+typedef CFF1IndexOf<TopDict> CFF1TopDictIndex;
+
+struct cff1_font_dict_values_mod_t
+{
+  cff1_font_dict_values_mod_t() { init (); }
+
+  void init () { init ( &Null (cff1_font_dict_values_t), CFF_UNDEF_SID ); }
+
+  void init (const cff1_font_dict_values_t *base_,
+	     unsigned int fontName_)
+  {
+    base = base_;
+    fontName = fontName_;
+    privateDictInfo.init ();
+  }
+
+  unsigned get_count () const { return base->get_count (); }
+
+  const op_str_t &operator [] (unsigned int i) const { return (*base)[i]; }
+
+  const cff1_font_dict_values_t    *base;
+  table_info_t		   privateDictInfo;
+  unsigned int		fontName;
+};
+
+struct CFF1FDArray : FDArray<HBUINT16>
+{
+  /* FDArray::serialize() requires this partial specialization to compile */
+  template <typename ITER, typename OP_SERIALIZER>
+  bool serialize (hb_serialize_context_t *c, ITER it, OP_SERIALIZER& opszr)
+  { return FDArray<HBUINT16>::serialize<cff1_font_dict_values_mod_t, cff1_font_dict_values_mod_t> (c, it, opszr); }
+};
+
+} /* namespace CFF */
+
+namespace OT {
+
+using namespace CFF;
+
+struct cff1
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_cff1;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (version.major == 1));
+  }
+
+  template <typename PRIVOPSET, typename PRIVDICTVAL>
+  struct accelerator_templ_t
+  {
+    void init (hb_face_t *face)
+    {
+      topDict.init ();
+      fontDicts.init ();
+      privateDicts.init ();
+
+      this->blob = sc.reference_table<cff1> (face);
+
+      /* setup for run-time santization */
+      sc.init (this->blob);
+      sc.start_processing ();
+
+      const OT::cff1 *cff = this->blob->template as<OT::cff1> ();
+
+      if (cff == &Null (OT::cff1))
+      { fini (); return; }
+
+      nameIndex = &cff->nameIndex (cff);
+      if ((nameIndex == &Null (CFF1NameIndex)) || !nameIndex->sanitize (&sc))
+      { fini (); return; }
+
+      topDictIndex = &StructAtOffset<CFF1TopDictIndex> (nameIndex, nameIndex->get_size ());
+      if ((topDictIndex == &Null (CFF1TopDictIndex)) || !topDictIndex->sanitize (&sc) || (topDictIndex->count == 0))
+      { fini (); return; }
+
+      { /* parse top dict */
+	const byte_str_t topDictStr = (*topDictIndex)[0];
+	if (unlikely (!topDictStr.sanitize (&sc))) { fini (); return; }
+	cff1_top_dict_interpreter_t top_interp;
+	top_interp.env.init (topDictStr);
+	topDict.init ();
+	if (unlikely (!top_interp.interpret (topDict))) { fini (); return; }
+      }
+
+      if (is_predef_charset ())
+	charset = &Null (Charset);
+      else
+      {
+	charset = &StructAtOffsetOrNull<Charset> (cff, topDict.CharsetOffset);
+	if (unlikely ((charset == &Null (Charset)) || !charset->sanitize (&sc))) { fini (); return; }
+      }
+
+      fdCount = 1;
+      if (is_CID ())
+      {
+	fdArray = &StructAtOffsetOrNull<CFF1FDArray> (cff, topDict.FDArrayOffset);
+	fdSelect = &StructAtOffsetOrNull<CFF1FDSelect> (cff, topDict.FDSelectOffset);
+	if (unlikely ((fdArray == &Null (CFF1FDArray)) || !fdArray->sanitize (&sc) ||
+	    (fdSelect == &Null (CFF1FDSelect)) || !fdSelect->sanitize (&sc, fdArray->count)))
+	{ fini (); return; }
+
+	fdCount = fdArray->count;
+      }
+      else
+      {
+	fdArray = &Null (CFF1FDArray);
+	fdSelect = &Null (CFF1FDSelect);
+      }
+
+      encoding = &Null (Encoding);
+      if (is_CID ())
+      {
+	if (unlikely (charset == &Null (Charset))) { fini (); return; }
+      }
+      else
+      {
+	if (!is_predef_encoding ())
+	{
+	  encoding = &StructAtOffsetOrNull<Encoding> (cff, topDict.EncodingOffset);
+	  if (unlikely ((encoding == &Null (Encoding)) || !encoding->sanitize (&sc))) { fini (); return; }
+	}
+      }
+
+      stringIndex = &StructAtOffset<CFF1StringIndex> (topDictIndex, topDictIndex->get_size ());
+      if ((stringIndex == &Null (CFF1StringIndex)) || !stringIndex->sanitize (&sc))
+      { fini (); return; }
+
+      globalSubrs = &StructAtOffset<CFF1Subrs> (stringIndex, stringIndex->get_size ());
+      if ((globalSubrs != &Null (CFF1Subrs)) && !globalSubrs->sanitize (&sc))
+      { fini (); return; }
+
+      charStrings = &StructAtOffsetOrNull<CFF1CharStrings> (cff, topDict.charStringsOffset);
+
+      if ((charStrings == &Null (CFF1CharStrings)) || unlikely (!charStrings->sanitize (&sc)))
+      { fini (); return; }
+
+      num_glyphs = charStrings->count;
+      if (num_glyphs != sc.get_num_glyphs ())
+      { fini (); return; }
+
+      if (unlikely (!privateDicts.resize (fdCount)))
+      { fini (); return; }
+      for (unsigned int i = 0; i < fdCount; i++)
+	privateDicts[i].init ();
+
+      // parse CID font dicts and gather private dicts
+      if (is_CID ())
+      {
+	for (unsigned int i = 0; i < fdCount; i++)
+	{
+	  byte_str_t fontDictStr = (*fdArray)[i];
+	  if (unlikely (!fontDictStr.sanitize (&sc))) { fini (); return; }
+	  cff1_font_dict_values_t *font;
+	  cff1_font_dict_interpreter_t font_interp;
+	  font_interp.env.init (fontDictStr);
+	  font = fontDicts.push ();
+	  if (unlikely (font == &Crap (cff1_font_dict_values_t))) { fini (); return; }
+	  font->init ();
+	  if (unlikely (!font_interp.interpret (*font))) { fini (); return; }
+	  PRIVDICTVAL *priv = &privateDicts[i];
+	  const byte_str_t privDictStr (StructAtOffset<UnsizedByteStr> (cff, font->privateDictInfo.offset), font->privateDictInfo.size);
+	  if (unlikely (!privDictStr.sanitize (&sc))) { fini (); return; }
+	  dict_interpreter_t<PRIVOPSET, PRIVDICTVAL> priv_interp;
+	  priv_interp.env.init (privDictStr);
+	  priv->init ();
+	  if (unlikely (!priv_interp.interpret (*priv))) { fini (); return; }
+
+	  priv->localSubrs = &StructAtOffsetOrNull<CFF1Subrs> (&privDictStr, priv->subrsOffset);
+	  if (priv->localSubrs != &Null (CFF1Subrs) &&
+	      unlikely (!priv->localSubrs->sanitize (&sc)))
+	  { fini (); return; }
+	}
+      }
+      else  /* non-CID */
+      {
+	cff1_top_dict_values_t *font = &topDict;
+	PRIVDICTVAL *priv = &privateDicts[0];
+
+	const byte_str_t privDictStr (StructAtOffset<UnsizedByteStr> (cff, font->privateDictInfo.offset), font->privateDictInfo.size);
+	if (unlikely (!privDictStr.sanitize (&sc))) { fini (); return; }
+	dict_interpreter_t<PRIVOPSET, PRIVDICTVAL> priv_interp;
+	priv_interp.env.init (privDictStr);
+	priv->init ();
+	if (unlikely (!priv_interp.interpret (*priv))) { fini (); return; }
+
+	priv->localSubrs = &StructAtOffsetOrNull<CFF1Subrs> (&privDictStr, priv->subrsOffset);
+	if (priv->localSubrs != &Null (CFF1Subrs) &&
+	    unlikely (!priv->localSubrs->sanitize (&sc)))
+	{ fini (); return; }
+      }
+    }
+
+    void fini ()
+    {
+      sc.end_processing ();
+      topDict.fini ();
+      fontDicts.fini_deep ();
+      privateDicts.fini_deep ();
+      hb_blob_destroy (blob);
+      blob = nullptr;
+    }
+
+    bool is_valid () const { return blob; }
+    bool   is_CID () const { return topDict.is_CID (); }
+
+    bool is_predef_charset () const { return topDict.CharsetOffset <= ExpertSubsetCharset; }
+
+    unsigned int std_code_to_glyph (hb_codepoint_t code) const
+    {
+      hb_codepoint_t sid = lookup_standard_encoding_for_sid (code);
+      if (unlikely (sid == CFF_UNDEF_SID))
+	return 0;
+
+      if (charset != &Null (Charset))
+	return charset->get_glyph (sid, num_glyphs);
+      else if ((topDict.CharsetOffset == ISOAdobeCharset)
+	      && (code <= 228 /*zcaron*/)) return sid;
+      return 0;
+    }
+
+    bool is_predef_encoding () const { return topDict.EncodingOffset <= ExpertEncoding; }
+
+    hb_codepoint_t glyph_to_code (hb_codepoint_t glyph) const
+    {
+      if (encoding != &Null (Encoding))
+	return encoding->get_code (glyph);
+      else
+      {
+	hb_codepoint_t sid = glyph_to_sid (glyph);
+	if (sid == 0) return 0;
+	hb_codepoint_t code = 0;
+	switch (topDict.EncodingOffset)
+	{
+	case StandardEncoding:
+	  code = lookup_standard_encoding_for_code (sid);
+	  break;
+	case ExpertEncoding:
+	  code = lookup_expert_encoding_for_code (sid);
+	  break;
+	default:
+	  break;
+	}
+	return code;
+      }
+    }
+
+    hb_codepoint_t glyph_to_sid (hb_codepoint_t glyph) const
+    {
+      if (charset != &Null (Charset))
+	return charset->get_sid (glyph, num_glyphs);
+      else
+      {
+	hb_codepoint_t sid = 0;
+	switch (topDict.CharsetOffset)
+	{
+	  case ISOAdobeCharset:
+	    if (glyph <= 228 /*zcaron*/) sid = glyph;
+	    break;
+	  case ExpertCharset:
+	    sid = lookup_expert_charset_for_sid (glyph);
+	    break;
+	  case ExpertSubsetCharset:
+	      sid = lookup_expert_subset_charset_for_sid (glyph);
+	    break;
+	  default:
+	    break;
+	}
+	return sid;
+      }
+    }
+
+    hb_codepoint_t sid_to_glyph (hb_codepoint_t sid) const
+    {
+      if (charset != &Null (Charset))
+	return charset->get_glyph (sid, num_glyphs);
+      else
+      {
+	hb_codepoint_t glyph = 0;
+	switch (topDict.CharsetOffset)
+	{
+	  case ISOAdobeCharset:
+	    if (sid <= 228 /*zcaron*/) glyph = sid;
+	    break;
+	  case ExpertCharset:
+	    glyph = lookup_expert_charset_for_glyph (sid);
+	    break;
+	  case ExpertSubsetCharset:
+	    glyph = lookup_expert_subset_charset_for_glyph (sid);
+	    break;
+	  default:
+	    break;
+	}
+	return glyph;
+      }
+    }
+
+    protected:
+    hb_blob_t	           *blob;
+    hb_sanitize_context_t   sc;
+
+    public:
+    const Encoding	    *encoding;
+    const Charset	    *charset;
+    const CFF1NameIndex     *nameIndex;
+    const CFF1TopDictIndex  *topDictIndex;
+    const CFF1StringIndex   *stringIndex;
+    const CFF1Subrs	    *globalSubrs;
+    const CFF1CharStrings   *charStrings;
+    const CFF1FDArray       *fdArray;
+    const CFF1FDSelect      *fdSelect;
+    unsigned int	     fdCount;
+
+    cff1_top_dict_values_t   topDict;
+    hb_vector_t<cff1_font_dict_values_t>
+			     fontDicts;
+    hb_vector_t<PRIVDICTVAL> privateDicts;
+
+    unsigned int	     num_glyphs;
+  };
+
+  struct accelerator_t : accelerator_templ_t<cff1_private_dict_opset_t, cff1_private_dict_values_t>
+  {
+    void init (hb_face_t *face)
+    {
+      SUPER::init (face);
+
+      if (!is_valid ()) return;
+      if (is_CID ()) return;
+
+      /* fill glyph_names */
+      for (hb_codepoint_t gid = 0; gid < num_glyphs; gid++)
+      {
+	hb_codepoint_t	sid = glyph_to_sid (gid);
+	gname_t	gname;
+	gname.sid = sid;
+	if (sid < cff1_std_strings_length)
+	  gname.name = cff1_std_strings (sid);
+	else
+	{
+	  byte_str_t	ustr = (*stringIndex)[sid - cff1_std_strings_length];
+	  gname.name = hb_bytes_t ((const char*)ustr.arrayZ, ustr.length);
+	}
+	if (unlikely (!gname.name.arrayZ)) { fini (); return; }
+	glyph_names.push (gname);
+      }
+      glyph_names.qsort ();
+    }
+
+    void fini ()
+    {
+      glyph_names.fini ();
+
+      SUPER::fini ();
+    }
+
+    bool get_glyph_name (hb_codepoint_t glyph,
+			 char *buf, unsigned int buf_len) const
+    {
+      if (!buf) return true;
+      if (unlikely (!is_valid ())) return false;
+      if (is_CID()) return false;
+      hb_codepoint_t sid = glyph_to_sid (glyph);
+      const char *str;
+      size_t str_len;
+      if (sid < cff1_std_strings_length)
+      {
+	hb_bytes_t byte_str = cff1_std_strings (sid);
+	str = byte_str.arrayZ;
+	str_len = byte_str.length;
+      }
+      else
+      {
+	byte_str_t ubyte_str = (*stringIndex)[sid - cff1_std_strings_length];
+	str = (const char *)ubyte_str.arrayZ;
+	str_len = ubyte_str.length;
+      }
+      if (!str_len) return false;
+      unsigned int len = hb_min (buf_len - 1, str_len);
+      strncpy (buf, (const char*)str, len);
+      buf[len] = '\0';
+      return true;
+    }
+
+    bool get_glyph_from_name (const char *name, int len,
+			      hb_codepoint_t *glyph) const
+    {
+      if (len < 0) len = strlen (name);
+      if (unlikely (!len)) return false;
+
+      gname_t key = { hb_bytes_t (name, len), 0 };
+      const gname_t *gname = glyph_names.bsearch (key);
+      if (!gname) return false;
+      hb_codepoint_t gid = sid_to_glyph (gname->sid);
+      if (!gid && gname->sid) return false;
+      *glyph = gid;
+      return true;
+    }
+
+    HB_INTERNAL bool get_extents (hb_font_t *font, hb_codepoint_t glyph, hb_glyph_extents_t *extents) const;
+    HB_INTERNAL bool get_seac_components (hb_codepoint_t glyph, hb_codepoint_t *base, hb_codepoint_t *accent) const;
+#ifdef HB_EXPERIMENTAL_API
+    HB_INTERNAL bool get_path (hb_font_t *font, hb_codepoint_t glyph, draw_helper_t &draw_helper) const;
+#endif
+
+    private:
+    struct gname_t
+    {
+      hb_bytes_t	name;
+      uint16_t		sid;
+
+      static int cmp (const void *a_, const void *b_)
+      {
+	const gname_t *a = (const gname_t *)a_;
+	const gname_t *b = (const gname_t *)b_;
+	int minlen = hb_min (a->name.length, b->name.length);
+	int ret = strncmp (a->name.arrayZ, b->name.arrayZ, minlen);
+	if (ret) return ret;
+	return a->name.length - b->name.length;
+      }
+
+      int cmp (const gname_t &a) const { return cmp (&a, this); }
+    };
+
+    hb_sorted_vector_t<gname_t>	glyph_names;
+
+    typedef accelerator_templ_t<cff1_private_dict_opset_t, cff1_private_dict_values_t> SUPER;
+  };
+
+  struct accelerator_subset_t : accelerator_templ_t<cff1_private_dict_opset_subset, cff1_private_dict_values_subset_t> {};
+
+  bool subset (hb_subset_context_t *c) const { return hb_subset_cff1 (c); }
+
+  protected:
+  HB_INTERNAL static hb_codepoint_t lookup_standard_encoding_for_code (hb_codepoint_t sid);
+  HB_INTERNAL static hb_codepoint_t lookup_expert_encoding_for_code (hb_codepoint_t sid);
+  HB_INTERNAL static hb_codepoint_t lookup_expert_charset_for_sid (hb_codepoint_t glyph);
+  HB_INTERNAL static hb_codepoint_t lookup_expert_subset_charset_for_sid (hb_codepoint_t glyph);
+  HB_INTERNAL static hb_codepoint_t lookup_expert_charset_for_glyph (hb_codepoint_t sid);
+  HB_INTERNAL static hb_codepoint_t lookup_expert_subset_charset_for_glyph (hb_codepoint_t sid);
+  HB_INTERNAL static hb_codepoint_t lookup_standard_encoding_for_sid (hb_codepoint_t code);
+
+  public:
+  FixedVersion<HBUINT8> version;	  /* Version of CFF table. set to 0x0100u */
+  OffsetTo<CFF1NameIndex, HBUINT8> nameIndex; /* headerSize = Offset to Name INDEX. */
+  HBUINT8	       offSize;	  /* offset size (unused?) */
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct cff1_accelerator_t : cff1::accelerator_t {};
+} /* namespace OT */
+
+#endif /* HB_OT_CFF1_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff2-table.cc b/thirdparty/harfbuzz/src/hb-ot-cff2-table.cc
new file mode 100644
index 0000000000..ac0feeee21
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff2-table.cc
@@ -0,0 +1,215 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_FONT_CFF
+
+#include "hb-ot-cff2-table.hh"
+#include "hb-cff2-interp-cs.hh"
+#include "hb-draw.hh"
+
+using namespace CFF;
+
+struct cff2_extents_param_t
+{
+  void init ()
+  {
+    path_open = false;
+    min_x.set_int (INT_MAX);
+    min_y.set_int (INT_MAX);
+    max_x.set_int (INT_MIN);
+    max_y.set_int (INT_MIN);
+  }
+
+  void   start_path ()       { path_open = true; }
+  void     end_path ()       { path_open = false; }
+  bool is_path_open () const { return path_open; }
+
+  void update_bounds (const point_t &pt)
+  {
+    if (pt.x < min_x) min_x = pt.x;
+    if (pt.x > max_x) max_x = pt.x;
+    if (pt.y < min_y) min_y = pt.y;
+    if (pt.y > max_y) max_y = pt.y;
+  }
+
+  bool  path_open;
+  number_t min_x;
+  number_t min_y;
+  number_t max_x;
+  number_t max_y;
+};
+
+struct cff2_path_procs_extents_t : path_procs_t<cff2_path_procs_extents_t, cff2_cs_interp_env_t, cff2_extents_param_t>
+{
+  static void moveto (cff2_cs_interp_env_t &env, cff2_extents_param_t& param, const point_t &pt)
+  {
+    param.end_path ();
+    env.moveto (pt);
+  }
+
+  static void line (cff2_cs_interp_env_t &env, cff2_extents_param_t& param, const point_t &pt1)
+  {
+    if (!param.is_path_open ())
+    {
+      param.start_path ();
+      param.update_bounds (env.get_pt ());
+    }
+    env.moveto (pt1);
+    param.update_bounds (env.get_pt ());
+  }
+
+  static void curve (cff2_cs_interp_env_t &env, cff2_extents_param_t& param, const point_t &pt1, const point_t &pt2, const point_t &pt3)
+  {
+    if (!param.is_path_open ())
+    {
+      param.start_path ();
+      param.update_bounds (env.get_pt ());
+    }
+    /* include control points */
+    param.update_bounds (pt1);
+    param.update_bounds (pt2);
+    env.moveto (pt3);
+    param.update_bounds (env.get_pt ());
+  }
+};
+
+struct cff2_cs_opset_extents_t : cff2_cs_opset_t<cff2_cs_opset_extents_t, cff2_extents_param_t, cff2_path_procs_extents_t> {};
+
+bool OT::cff2::accelerator_t::get_extents (hb_font_t *font,
+					   hb_codepoint_t glyph,
+					   hb_glyph_extents_t *extents) const
+{
+#ifdef HB_NO_OT_FONT_CFF
+  /* XXX Remove check when this code moves to .hh file. */
+  return true;
+#endif
+
+  if (unlikely (!is_valid () || (glyph >= num_glyphs))) return false;
+
+  unsigned int fd = fdSelect->get_fd (glyph);
+  cff2_cs_interpreter_t<cff2_cs_opset_extents_t, cff2_extents_param_t> interp;
+  const byte_str_t str = (*charStrings)[glyph];
+  interp.env.init (str, *this, fd, font->coords, font->num_coords);
+  cff2_extents_param_t  param;
+  param.init ();
+  if (unlikely (!interp.interpret (param))) return false;
+
+  if (param.min_x >= param.max_x)
+  {
+    extents->width = 0;
+    extents->x_bearing = 0;
+  }
+  else
+  {
+    extents->x_bearing = font->em_scalef_x (param.min_x.to_real ());
+    extents->width = font->em_scalef_x (param.max_x.to_real () - param.min_x.to_real ());
+  }
+  if (param.min_y >= param.max_y)
+  {
+    extents->height = 0;
+    extents->y_bearing = 0;
+  }
+  else
+  {
+    extents->y_bearing = font->em_scalef_y (param.max_y.to_real ());
+    extents->height = font->em_scalef_y (param.min_y.to_real () - param.max_y.to_real ());
+  }
+
+  return true;
+}
+
+#ifdef HB_EXPERIMENTAL_API
+struct cff2_path_param_t
+{
+  cff2_path_param_t (hb_font_t *font_, draw_helper_t &draw_helper_)
+  {
+    draw_helper = &draw_helper_;
+    font = font_;
+  }
+
+  void move_to (const point_t &p)
+  { draw_helper->move_to (font->em_scalef_x (p.x.to_real ()), font->em_scalef_y (p.y.to_real ())); }
+
+  void line_to (const point_t &p)
+  { draw_helper->line_to (font->em_scalef_x (p.x.to_real ()), font->em_scalef_y (p.y.to_real ())); }
+
+  void cubic_to (const point_t &p1, const point_t &p2, const point_t &p3)
+  {
+    draw_helper->cubic_to (font->em_scalef_x (p1.x.to_real ()), font->em_scalef_y (p1.y.to_real ()),
+			   font->em_scalef_x (p2.x.to_real ()), font->em_scalef_y (p2.y.to_real ()),
+			   font->em_scalef_x (p3.x.to_real ()), font->em_scalef_y (p3.y.to_real ()));
+  }
+
+  protected:
+  draw_helper_t *draw_helper;
+  hb_font_t *font;
+};
+
+struct cff2_path_procs_path_t : path_procs_t<cff2_path_procs_path_t, cff2_cs_interp_env_t, cff2_path_param_t>
+{
+  static void moveto (cff2_cs_interp_env_t &env, cff2_path_param_t& param, const point_t &pt)
+  {
+    param.move_to (pt);
+    env.moveto (pt);
+  }
+
+  static void line (cff2_cs_interp_env_t &env, cff2_path_param_t& param, const point_t &pt1)
+  {
+    param.line_to (pt1);
+    env.moveto (pt1);
+  }
+
+  static void curve (cff2_cs_interp_env_t &env, cff2_path_param_t& param, const point_t &pt1, const point_t &pt2, const point_t &pt3)
+  {
+    param.cubic_to (pt1, pt2, pt3);
+    env.moveto (pt3);
+  }
+};
+
+struct cff2_cs_opset_path_t : cff2_cs_opset_t<cff2_cs_opset_path_t, cff2_path_param_t, cff2_path_procs_path_t> {};
+
+bool OT::cff2::accelerator_t::get_path (hb_font_t *font, hb_codepoint_t glyph, draw_helper_t &draw_helper) const
+{
+#ifdef HB_NO_OT_FONT_CFF
+  /* XXX Remove check when this code moves to .hh file. */
+  return true;
+#endif
+
+  if (unlikely (!is_valid () || (glyph >= num_glyphs))) return false;
+
+  unsigned int fd = fdSelect->get_fd (glyph);
+  cff2_cs_interpreter_t<cff2_cs_opset_path_t, cff2_path_param_t> interp;
+  const byte_str_t str = (*charStrings)[glyph];
+  interp.env.init (str, *this, fd, font->coords, font->num_coords);
+  cff2_path_param_t param (font, draw_helper);
+  if (unlikely (!interp.interpret (param))) return false;
+  return true;
+}
+#endif
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-cff2-table.hh b/thirdparty/harfbuzz/src/hb-ot-cff2-table.hh
new file mode 100644
index 0000000000..829217feaa
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cff2-table.hh
@@ -0,0 +1,531 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_CFF2_TABLE_HH
+#define HB_OT_CFF2_TABLE_HH
+
+#include "hb-ot-cff-common.hh"
+#include "hb-subset-cff2.hh"
+#include "hb-draw.hh"
+
+namespace CFF {
+
+/*
+ * CFF2 -- Compact Font Format (CFF) Version 2
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/cff2
+ */
+#define HB_OT_TAG_cff2 HB_TAG('C','F','F','2')
+
+typedef CFFIndex<HBUINT32>  CFF2Index;
+template <typename Type> struct CFF2IndexOf : CFFIndexOf<HBUINT32, Type> {};
+
+typedef CFF2Index         CFF2CharStrings;
+typedef Subrs<HBUINT32>   CFF2Subrs;
+
+typedef FDSelect3_4<HBUINT32, HBUINT16> FDSelect4;
+typedef FDSelect3_4_Range<HBUINT32, HBUINT16> FDSelect4_Range;
+
+struct CFF2FDSelect
+{
+  bool serialize (hb_serialize_context_t *c, const CFF2FDSelect &src, unsigned int num_glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size = src.get_size (num_glyphs);
+    CFF2FDSelect *dest = c->allocate_size<CFF2FDSelect> (size);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, &src, size);
+    return_trace (true);
+  }
+
+  unsigned int get_size (unsigned int num_glyphs) const
+  {
+    switch (format)
+    {
+    case 0: return format.static_size + u.format0.get_size (num_glyphs);
+    case 3: return format.static_size + u.format3.get_size ();
+    case 4: return format.static_size + u.format4.get_size ();
+    default:return 0;
+    }
+  }
+
+  hb_codepoint_t get_fd (hb_codepoint_t glyph) const
+  {
+    if (this == &Null (CFF2FDSelect))
+      return 0;
+
+    switch (format)
+    {
+    case 0: return u.format0.get_fd (glyph);
+    case 3: return u.format3.get_fd (glyph);
+    case 4: return u.format4.get_fd (glyph);
+    default:return 0;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int fdcount) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (format)
+    {
+    case 0: return_trace (u.format0.sanitize (c, fdcount));
+    case 3: return_trace (u.format3.sanitize (c, fdcount));
+    case 4: return_trace (u.format4.sanitize (c, fdcount));
+    default:return_trace (false);
+    }
+  }
+
+  HBUINT8	format;
+  union {
+  FDSelect0	format0;
+  FDSelect3	format3;
+  FDSelect4	format4;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (2);
+};
+
+struct CFF2VariationStore
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)) && c->check_range (&varStore, size) && varStore.sanitize (c));
+  }
+
+  bool serialize (hb_serialize_context_t *c, const CFF2VariationStore *varStore)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int size_ = varStore->get_size ();
+    CFF2VariationStore *dest = c->allocate_size<CFF2VariationStore> (size_);
+    if (unlikely (!dest)) return_trace (false);
+    memcpy (dest, varStore, size_);
+    return_trace (true);
+  }
+
+  unsigned int get_size () const { return HBUINT16::static_size + size; }
+
+  HBUINT16	size;
+  VariationStore  varStore;
+
+  DEFINE_SIZE_MIN (2 + VariationStore::min_size);
+};
+
+struct cff2_top_dict_values_t : top_dict_values_t<>
+{
+  void init ()
+  {
+    top_dict_values_t<>::init ();
+    vstoreOffset = 0;
+    FDSelectOffset = 0;
+  }
+  void fini () { top_dict_values_t<>::fini (); }
+
+  unsigned int  vstoreOffset;
+  unsigned int  FDSelectOffset;
+};
+
+struct cff2_top_dict_opset_t : top_dict_opset_t<>
+{
+  static void process_op (op_code_t op, num_interp_env_t& env, cff2_top_dict_values_t& dictval)
+  {
+    switch (op) {
+      case OpCode_FontMatrix:
+	{
+	  dict_val_t val;
+	  val.init ();
+	  dictval.add_op (op, env.str_ref);
+	  env.clear_args ();
+	}
+	break;
+
+      case OpCode_vstore:
+	dictval.vstoreOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_FDSelect:
+	dictval.FDSelectOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	SUPER::process_op (op, env, dictval);
+	/* Record this operand below if stack is empty, otherwise done */
+	if (!env.argStack.is_empty ()) return;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref);
+  }
+
+  typedef top_dict_opset_t<> SUPER;
+};
+
+struct cff2_font_dict_values_t : dict_values_t<op_str_t>
+{
+  void init ()
+  {
+    dict_values_t<op_str_t>::init ();
+    privateDictInfo.init ();
+  }
+  void fini () { dict_values_t<op_str_t>::fini (); }
+
+  table_info_t    privateDictInfo;
+};
+
+struct cff2_font_dict_opset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, num_interp_env_t& env, cff2_font_dict_values_t& dictval)
+  {
+    switch (op) {
+      case OpCode_Private:
+	dictval.privateDictInfo.offset = env.argStack.pop_uint ();
+	dictval.privateDictInfo.size = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	SUPER::process_op (op, env);
+	if (!env.argStack.is_empty ())
+	  return;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref);
+  }
+
+  private:
+  typedef dict_opset_t SUPER;
+};
+
+template <typename VAL>
+struct cff2_private_dict_values_base_t : dict_values_t<VAL>
+{
+  void init ()
+  {
+    dict_values_t<VAL>::init ();
+    subrsOffset = 0;
+    localSubrs = &Null (CFF2Subrs);
+    ivs = 0;
+  }
+  void fini () { dict_values_t<VAL>::fini (); }
+
+  unsigned int      subrsOffset;
+  const CFF2Subrs   *localSubrs;
+  unsigned int      ivs;
+};
+
+typedef cff2_private_dict_values_base_t<op_str_t> cff2_private_dict_values_subset_t;
+typedef cff2_private_dict_values_base_t<num_dict_val_t> cff2_private_dict_values_t;
+
+struct cff2_priv_dict_interp_env_t : num_interp_env_t
+{
+  void init (const byte_str_t &str)
+  {
+    num_interp_env_t::init (str);
+    ivs = 0;
+    seen_vsindex = false;
+  }
+
+  void process_vsindex ()
+  {
+    if (likely (!seen_vsindex))
+    {
+      set_ivs (argStack.pop_uint ());
+    }
+    seen_vsindex = true;
+  }
+
+  unsigned int get_ivs () const { return ivs; }
+  void	 set_ivs (unsigned int ivs_) { ivs = ivs_; }
+
+  protected:
+  unsigned int  ivs;
+  bool	  seen_vsindex;
+};
+
+struct cff2_private_dict_opset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, cff2_priv_dict_interp_env_t& env, cff2_private_dict_values_t& dictval)
+  {
+    num_dict_val_t val;
+    val.init ();
+
+    switch (op) {
+      case OpCode_StdHW:
+      case OpCode_StdVW:
+      case OpCode_BlueScale:
+      case OpCode_BlueShift:
+      case OpCode_BlueFuzz:
+      case OpCode_ExpansionFactor:
+      case OpCode_LanguageGroup:
+	val.single_val = env.argStack.pop_num ();
+	env.clear_args ();
+	break;
+      case OpCode_BlueValues:
+      case OpCode_OtherBlues:
+      case OpCode_FamilyBlues:
+      case OpCode_FamilyOtherBlues:
+      case OpCode_StemSnapH:
+      case OpCode_StemSnapV:
+	env.clear_args ();
+	break;
+      case OpCode_Subrs:
+	dictval.subrsOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+      case OpCode_vsindexdict:
+	env.process_vsindex ();
+	dictval.ivs = env.get_ivs ();
+	env.clear_args ();
+	break;
+      case OpCode_blenddict:
+	break;
+
+      default:
+	dict_opset_t::process_op (op, env);
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref, val);
+  }
+};
+
+struct cff2_private_dict_opset_subset_t : dict_opset_t
+{
+  static void process_op (op_code_t op, cff2_priv_dict_interp_env_t& env, cff2_private_dict_values_subset_t& dictval)
+  {
+    switch (op) {
+      case OpCode_BlueValues:
+      case OpCode_OtherBlues:
+      case OpCode_FamilyBlues:
+      case OpCode_FamilyOtherBlues:
+      case OpCode_StdHW:
+      case OpCode_StdVW:
+      case OpCode_BlueScale:
+      case OpCode_BlueShift:
+      case OpCode_BlueFuzz:
+      case OpCode_StemSnapH:
+      case OpCode_StemSnapV:
+      case OpCode_LanguageGroup:
+      case OpCode_ExpansionFactor:
+	env.clear_args ();
+	break;
+
+      case OpCode_blenddict:
+	env.clear_args ();
+	return;
+
+      case OpCode_Subrs:
+	dictval.subrsOffset = env.argStack.pop_uint ();
+	env.clear_args ();
+	break;
+
+      default:
+	SUPER::process_op (op, env);
+	if (!env.argStack.is_empty ()) return;
+	break;
+    }
+
+    if (unlikely (env.in_error ())) return;
+
+    dictval.add_op (op, env.str_ref);
+  }
+
+  private:
+  typedef dict_opset_t SUPER;
+};
+
+typedef dict_interpreter_t<cff2_top_dict_opset_t, cff2_top_dict_values_t> cff2_top_dict_interpreter_t;
+typedef dict_interpreter_t<cff2_font_dict_opset_t, cff2_font_dict_values_t> cff2_font_dict_interpreter_t;
+
+struct CFF2FDArray : FDArray<HBUINT32>
+{
+  /* FDArray::serialize does not compile without this partial specialization */
+  template <typename ITER, typename OP_SERIALIZER>
+  bool serialize (hb_serialize_context_t *c, ITER it, OP_SERIALIZER& opszr)
+  { return FDArray<HBUINT32>::serialize<cff2_font_dict_values_t, table_info_t> (c, it, opszr); }
+};
+
+} /* namespace CFF */
+
+namespace OT {
+
+using namespace CFF;
+
+struct cff2
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_cff2;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (version.major == 2));
+  }
+
+  template <typename PRIVOPSET, typename PRIVDICTVAL>
+  struct accelerator_templ_t
+  {
+    void init (hb_face_t *face)
+    {
+      topDict.init ();
+      fontDicts.init ();
+      privateDicts.init ();
+
+      this->blob = sc.reference_table<cff2> (face);
+
+      /* setup for run-time santization */
+      sc.init (this->blob);
+      sc.start_processing ();
+
+      const OT::cff2 *cff2 = this->blob->template as<OT::cff2> ();
+
+      if (cff2 == &Null (OT::cff2))
+      { fini (); return; }
+
+      { /* parse top dict */
+	byte_str_t topDictStr (cff2 + cff2->topDict, cff2->topDictSize);
+	if (unlikely (!topDictStr.sanitize (&sc))) { fini (); return; }
+	cff2_top_dict_interpreter_t top_interp;
+	top_interp.env.init (topDictStr);
+	topDict.init ();
+	if (unlikely (!top_interp.interpret (topDict))) { fini (); return; }
+      }
+
+      globalSubrs = &StructAtOffset<CFF2Subrs> (cff2, cff2->topDict + cff2->topDictSize);
+      varStore = &StructAtOffsetOrNull<CFF2VariationStore> (cff2, topDict.vstoreOffset);
+      charStrings = &StructAtOffsetOrNull<CFF2CharStrings> (cff2, topDict.charStringsOffset);
+      fdArray = &StructAtOffsetOrNull<CFF2FDArray> (cff2, topDict.FDArrayOffset);
+      fdSelect = &StructAtOffsetOrNull<CFF2FDSelect> (cff2, topDict.FDSelectOffset);
+
+      if (((varStore != &Null (CFF2VariationStore)) && unlikely (!varStore->sanitize (&sc))) ||
+	  (charStrings == &Null (CFF2CharStrings)) || unlikely (!charStrings->sanitize (&sc)) ||
+	  (globalSubrs == &Null (CFF2Subrs)) || unlikely (!globalSubrs->sanitize (&sc)) ||
+	  (fdArray == &Null (CFF2FDArray)) || unlikely (!fdArray->sanitize (&sc)) ||
+	  (((fdSelect != &Null (CFF2FDSelect)) && unlikely (!fdSelect->sanitize (&sc, fdArray->count)))))
+      { fini (); return; }
+
+      num_glyphs = charStrings->count;
+      if (num_glyphs != sc.get_num_glyphs ())
+      { fini (); return; }
+
+      fdCount = fdArray->count;
+      if (!privateDicts.resize (fdCount))
+      { fini (); return; }
+
+      /* parse font dicts and gather private dicts */
+      for (unsigned int i = 0; i < fdCount; i++)
+      {
+	const byte_str_t fontDictStr = (*fdArray)[i];
+	if (unlikely (!fontDictStr.sanitize (&sc))) { fini (); return; }
+	cff2_font_dict_values_t  *font;
+	cff2_font_dict_interpreter_t font_interp;
+	font_interp.env.init (fontDictStr);
+	font = fontDicts.push ();
+	if (unlikely (font == &Crap (cff2_font_dict_values_t))) { fini (); return; }
+	font->init ();
+	if (unlikely (!font_interp.interpret (*font))) { fini (); return; }
+
+	const byte_str_t privDictStr (StructAtOffsetOrNull<UnsizedByteStr> (cff2, font->privateDictInfo.offset), font->privateDictInfo.size);
+	if (unlikely (!privDictStr.sanitize (&sc))) { fini (); return; }
+	dict_interpreter_t<PRIVOPSET, PRIVDICTVAL, cff2_priv_dict_interp_env_t>  priv_interp;
+	priv_interp.env.init(privDictStr);
+	privateDicts[i].init ();
+	if (unlikely (!priv_interp.interpret (privateDicts[i]))) { fini (); return; }
+
+	privateDicts[i].localSubrs = &StructAtOffsetOrNull<CFF2Subrs> (&privDictStr[0], privateDicts[i].subrsOffset);
+	if (privateDicts[i].localSubrs != &Null (CFF2Subrs) &&
+	  unlikely (!privateDicts[i].localSubrs->sanitize (&sc)))
+	{ fini (); return; }
+      }
+    }
+
+    void fini ()
+    {
+      sc.end_processing ();
+      topDict.fini ();
+      fontDicts.fini_deep ();
+      privateDicts.fini_deep ();
+      hb_blob_destroy (blob);
+      blob = nullptr;
+    }
+
+    bool is_valid () const { return blob; }
+
+    protected:
+    hb_blob_t			*blob;
+    hb_sanitize_context_t	sc;
+
+    public:
+    cff2_top_dict_values_t	topDict;
+    const CFF2Subrs		*globalSubrs;
+    const CFF2VariationStore	*varStore;
+    const CFF2CharStrings	*charStrings;
+    const CFF2FDArray		*fdArray;
+    const CFF2FDSelect		*fdSelect;
+    unsigned int		fdCount;
+
+    hb_vector_t<cff2_font_dict_values_t>     fontDicts;
+    hb_vector_t<PRIVDICTVAL>  privateDicts;
+
+    unsigned int	      num_glyphs;
+  };
+
+  struct accelerator_t : accelerator_templ_t<cff2_private_dict_opset_t, cff2_private_dict_values_t>
+  {
+    HB_INTERNAL bool get_extents (hb_font_t *font,
+				  hb_codepoint_t glyph,
+				  hb_glyph_extents_t *extents) const;
+#ifdef HB_EXPERIMENTAL_API
+    HB_INTERNAL bool get_path (hb_font_t *font, hb_codepoint_t glyph, draw_helper_t &draw_helper) const;
+#endif
+  };
+
+  typedef accelerator_templ_t<cff2_private_dict_opset_subset_t, cff2_private_dict_values_subset_t> accelerator_subset_t;
+
+  bool subset (hb_subset_context_t *c) const { return hb_subset_cff2 (c); }
+
+  public:
+  FixedVersion<HBUINT8>		version;	/* Version of CFF2 table. set to 0x0200u */
+  NNOffsetTo<TopDict, HBUINT8>	topDict;	/* headerSize = Offset to Top DICT. */
+  HBUINT16			topDictSize;	/* Top DICT size */
+
+  public:
+  DEFINE_SIZE_STATIC (5);
+};
+
+struct cff2_accelerator_t : cff2::accelerator_t {};
+} /* namespace OT */
+
+#endif /* HB_OT_CFF2_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-cmap-table.hh b/thirdparty/harfbuzz/src/hb-ot-cmap-table.hh
new file mode 100644
index 0000000000..cc48379bb8
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-cmap-table.hh
@@ -0,0 +1,1711 @@
+/*
+ * Copyright © 2014  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_CMAP_TABLE_HH
+#define HB_OT_CMAP_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-set.hh"
+
+/*
+ * cmap -- Character to Glyph Index Mapping
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/cmap
+ */
+#define HB_OT_TAG_cmap HB_TAG('c','m','a','p')
+
+namespace OT {
+
+
+struct CmapSubtableFormat0
+{
+  bool get_glyph (hb_codepoint_t codepoint, hb_codepoint_t *glyph) const
+  {
+    hb_codepoint_t gid = codepoint < 256 ? glyphIdArray[codepoint] : 0;
+    if (!gid)
+      return false;
+    *glyph = gid;
+    return true;
+  }
+  void collect_unicodes (hb_set_t *out) const
+  {
+    for (unsigned int i = 0; i < 256; i++)
+      if (glyphIdArray[i])
+	out->add (i);
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    for (unsigned i = 0; i < 256; i++)
+      if (glyphIdArray[i])
+      {
+	hb_codepoint_t glyph = glyphIdArray[i];
+	unicodes->add (i);
+	mapping->set (i, glyph);
+      }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format number is set to 0. */
+  HBUINT16	length;		/* Byte length of this subtable. */
+  HBUINT16	language;	/* Ignore. */
+  HBUINT8	glyphIdArray[256];/* An array that maps character
+				 * code to glyph index values. */
+  public:
+  DEFINE_SIZE_STATIC (6 + 256);
+};
+
+struct CmapSubtableFormat4
+{
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  HBUINT16* serialize_endcode_array (hb_serialize_context_t *c,
+				     Iterator it)
+  {
+    HBUINT16 *endCode = c->start_embed<HBUINT16> ();
+    hb_codepoint_t prev_endcp = 0xFFFF;
+
+    for (const hb_item_type<Iterator> _ : +it)
+    {
+      if (prev_endcp != 0xFFFF && prev_endcp + 1u != _.first)
+      {
+	HBUINT16 end_code;
+	end_code = prev_endcp;
+	c->copy<HBUINT16> (end_code);
+      }
+      prev_endcp = _.first;
+    }
+
+    {
+      // last endCode
+      HBUINT16 endcode;
+      endcode = prev_endcp;
+      if (unlikely (!c->copy<HBUINT16> (endcode))) return nullptr;
+      // There must be a final entry with end_code == 0xFFFF.
+      if (prev_endcp != 0xFFFF)
+      {
+	HBUINT16 finalcode;
+	finalcode = 0xFFFF;
+	if (unlikely (!c->copy<HBUINT16> (finalcode))) return nullptr;
+      }
+    }
+
+    return endCode;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  HBUINT16* serialize_startcode_array (hb_serialize_context_t *c,
+				       Iterator it)
+  {
+    HBUINT16 *startCode = c->start_embed<HBUINT16> ();
+    hb_codepoint_t prev_cp = 0xFFFF;
+
+    for (const hb_item_type<Iterator> _ : +it)
+    {
+      if (prev_cp == 0xFFFF || prev_cp + 1u != _.first)
+      {
+	HBUINT16 start_code;
+	start_code = _.first;
+	c->copy<HBUINT16> (start_code);
+      }
+
+      prev_cp = _.first;
+    }
+
+    // There must be a final entry with end_code == 0xFFFF.
+    if (it.len () == 0 || prev_cp != 0xFFFF)
+    {
+      HBUINT16 finalcode;
+      finalcode = 0xFFFF;
+      if (unlikely (!c->copy<HBUINT16> (finalcode))) return nullptr;
+    }
+
+    return startCode;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  HBINT16* serialize_idDelta_array (hb_serialize_context_t *c,
+				    Iterator it,
+				    HBUINT16 *endCode,
+				    HBUINT16 *startCode,
+				    unsigned segcount)
+  {
+    unsigned i = 0;
+    hb_codepoint_t last_gid = 0, start_gid = 0, last_cp = 0xFFFF;
+    bool use_delta = true;
+
+    HBINT16 *idDelta = c->start_embed<HBINT16> ();
+    if ((char *)idDelta - (char *)startCode != (int) segcount * (int) HBINT16::static_size)
+      return nullptr;
+
+    for (const hb_item_type<Iterator> _ : +it)
+    {
+      if (_.first == startCode[i])
+      {
+	use_delta = true;
+	start_gid = _.second;
+      }
+      else if (_.second != last_gid + 1) use_delta = false;
+
+      if (_.first == endCode[i])
+      {
+	HBINT16 delta;
+	if (use_delta) delta = (int)start_gid - (int)startCode[i];
+	else delta = 0;
+	c->copy<HBINT16> (delta);
+
+	i++;
+      }
+
+      last_gid = _.second;
+      last_cp = _.first;
+    }
+
+    if (it.len () == 0 || last_cp != 0xFFFF)
+    {
+      HBINT16 delta;
+      delta = 1;
+      if (unlikely (!c->copy<HBINT16> (delta))) return nullptr;
+    }
+
+    return idDelta;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  HBUINT16* serialize_rangeoffset_glyid (hb_serialize_context_t *c,
+					 Iterator it,
+					 HBUINT16 *endCode,
+					 HBUINT16 *startCode,
+					 HBINT16 *idDelta,
+					 unsigned segcount)
+  {
+    HBUINT16 *idRangeOffset = c->allocate_size<HBUINT16> (HBUINT16::static_size * segcount);
+    if (unlikely (!c->check_success (idRangeOffset))) return nullptr;
+    if (unlikely ((char *)idRangeOffset - (char *)idDelta != (int) segcount * (int) HBINT16::static_size)) return nullptr;
+
+    + hb_range (segcount)
+    | hb_filter ([&] (const unsigned _) { return idDelta[_] == 0; })
+    | hb_apply ([&] (const unsigned i)
+		{
+		  idRangeOffset[i] = 2 * (c->start_embed<HBUINT16> () - idRangeOffset - i);
+
+		  + it
+		  | hb_filter ([&] (const hb_item_type<Iterator> _) { return _.first >= startCode[i] && _.first <= endCode[i]; })
+		  | hb_apply ([&] (const hb_item_type<Iterator> _)
+			      {
+				HBUINT16 glyID;
+				glyID = _.second;
+				c->copy<HBUINT16> (glyID);
+			      })
+		  ;
+
+
+		})
+    ;
+
+    return idRangeOffset;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    auto format4_iter =
+    + it
+    | hb_filter ([&] (const hb_pair_t<hb_codepoint_t, hb_codepoint_t> _)
+		 { return _.first <= 0xFFFF; })
+    ;
+
+    if (format4_iter.len () == 0) return;
+
+    unsigned table_initpos = c->length ();
+    if (unlikely (!c->extend_min (*this))) return;
+    this->format = 4;
+
+    //serialize endCode[]
+    HBUINT16 *endCode = serialize_endcode_array (c, format4_iter);
+    if (unlikely (!endCode)) return;
+
+    unsigned segcount = (c->length () - min_size) / HBUINT16::static_size;
+
+    // 2 bytes of padding.
+    if (unlikely (!c->allocate_size<HBUINT16> (HBUINT16::static_size))) return; // 2 bytes of padding.
+
+   // serialize startCode[]
+    HBUINT16 *startCode = serialize_startcode_array (c, format4_iter);
+    if (unlikely (!startCode)) return;
+
+    //serialize idDelta[]
+    HBINT16 *idDelta = serialize_idDelta_array (c, format4_iter, endCode, startCode, segcount);
+    if (unlikely (!idDelta)) return;
+
+    HBUINT16 *idRangeOffset = serialize_rangeoffset_glyid (c, format4_iter, endCode, startCode, idDelta, segcount);
+    if (unlikely (!c->check_success (idRangeOffset))) return;
+
+    if (unlikely (!c->check_assign(this->length, c->length () - table_initpos))) return;
+    this->segCountX2 = segcount * 2;
+    this->entrySelector = hb_max (1u, hb_bit_storage (segcount)) - 1;
+    this->searchRange = 2 * (1u << this->entrySelector);
+    this->rangeShift = segcount * 2 > this->searchRange
+		       ? 2 * segcount - this->searchRange
+		       : 0;
+  }
+
+  struct accelerator_t
+  {
+    accelerator_t () {}
+    accelerator_t (const CmapSubtableFormat4 *subtable) { init (subtable); }
+    ~accelerator_t () { fini (); }
+
+    void init (const CmapSubtableFormat4 *subtable)
+    {
+      segCount = subtable->segCountX2 / 2;
+      endCount = subtable->values.arrayZ;
+      startCount = endCount + segCount + 1;
+      idDelta = startCount + segCount;
+      idRangeOffset = idDelta + segCount;
+      glyphIdArray = idRangeOffset + segCount;
+      glyphIdArrayLength = (subtable->length - 16 - 8 * segCount) / 2;
+    }
+    void fini () {}
+
+    bool get_glyph (hb_codepoint_t codepoint, hb_codepoint_t *glyph) const
+    {
+      struct CustomRange
+      {
+	int cmp (hb_codepoint_t k,
+		 unsigned distance) const
+	{
+	  if (k > last) return +1;
+	  if (k < (&last)[distance]) return -1;
+	  return 0;
+	}
+	HBUINT16 last;
+      };
+
+      const HBUINT16 *found = hb_bsearch (codepoint,
+					  this->endCount,
+					  this->segCount,
+					  2,
+					  _hb_cmp_method<hb_codepoint_t, CustomRange, unsigned>,
+					  this->segCount + 1);
+      if (!found)
+	return false;
+      unsigned int i = found - endCount;
+
+      hb_codepoint_t gid;
+      unsigned int rangeOffset = this->idRangeOffset[i];
+      if (rangeOffset == 0)
+	gid = codepoint + this->idDelta[i];
+      else
+      {
+	/* Somebody has been smoking... */
+	unsigned int index = rangeOffset / 2 + (codepoint - this->startCount[i]) + i - this->segCount;
+	if (unlikely (index >= this->glyphIdArrayLength))
+	  return false;
+	gid = this->glyphIdArray[index];
+	if (unlikely (!gid))
+	  return false;
+	gid += this->idDelta[i];
+      }
+      gid &= 0xFFFFu;
+      if (!gid)
+	return false;
+      *glyph = gid;
+      return true;
+    }
+
+    HB_INTERNAL static bool get_glyph_func (const void *obj, hb_codepoint_t codepoint, hb_codepoint_t *glyph)
+    { return ((const accelerator_t *) obj)->get_glyph (codepoint, glyph); }
+
+    void collect_unicodes (hb_set_t *out) const
+    {
+      unsigned int count = this->segCount;
+      if (count && this->startCount[count - 1] == 0xFFFFu)
+	count--; /* Skip sentinel segment. */
+      for (unsigned int i = 0; i < count; i++)
+      {
+	hb_codepoint_t start = this->startCount[i];
+	hb_codepoint_t end = this->endCount[i];
+	unsigned int rangeOffset = this->idRangeOffset[i];
+	if (rangeOffset == 0)
+	{
+	  for (hb_codepoint_t codepoint = start; codepoint <= end; codepoint++)
+	  {
+	    hb_codepoint_t gid = (codepoint + this->idDelta[i]) & 0xFFFFu;
+	    if (unlikely (!gid))
+	      continue;
+	    out->add (codepoint);
+	  }
+	}
+	else
+	{
+	  for (hb_codepoint_t codepoint = start; codepoint <= end; codepoint++)
+	  {
+	    unsigned int index = rangeOffset / 2 + (codepoint - this->startCount[i]) + i - this->segCount;
+	    if (unlikely (index >= this->glyphIdArrayLength))
+	      break;
+	    hb_codepoint_t gid = this->glyphIdArray[index];
+	    if (unlikely (!gid))
+	      continue;
+	    out->add (codepoint);
+	  }
+	}
+      }
+    }
+
+    void collect_mapping (hb_set_t *unicodes, /* OUT */
+			  hb_map_t *mapping /* OUT */) const
+    {
+      unsigned count = this->segCount;
+      if (count && this->startCount[count - 1] == 0xFFFFu)
+	count--; /* Skip sentinel segment. */
+      for (unsigned i = 0; i < count; i++)
+      {
+	hb_codepoint_t start = this->startCount[i];
+	hb_codepoint_t end = this->endCount[i];
+	unsigned rangeOffset = this->idRangeOffset[i];
+	if (rangeOffset == 0)
+	{
+	  for (hb_codepoint_t codepoint = start; codepoint <= end; codepoint++)
+	  {
+	    hb_codepoint_t gid = (codepoint + this->idDelta[i]) & 0xFFFFu;
+	    if (unlikely (!gid))
+	      continue;
+	    unicodes->add (codepoint);
+	    mapping->set (codepoint, gid);
+	  }
+	}
+	else
+	{
+	  for (hb_codepoint_t codepoint = start; codepoint <= end; codepoint++)
+	  {
+	    unsigned index = rangeOffset / 2 + (codepoint - this->startCount[i]) + i - this->segCount;
+	    if (unlikely (index >= this->glyphIdArrayLength))
+	      break;
+	    hb_codepoint_t gid = this->glyphIdArray[index];
+	    if (unlikely (!gid))
+	      continue;
+	    unicodes->add (codepoint);
+	    mapping->set (codepoint, gid);
+	  }
+	}
+      }
+    }
+
+    const HBUINT16 *endCount;
+    const HBUINT16 *startCount;
+    const HBUINT16 *idDelta;
+    const HBUINT16 *idRangeOffset;
+    const HBUINT16 *glyphIdArray;
+    unsigned int segCount;
+    unsigned int glyphIdArrayLength;
+  };
+
+  bool get_glyph (hb_codepoint_t codepoint, hb_codepoint_t *glyph) const
+  {
+    accelerator_t accel (this);
+    return accel.get_glyph_func (&accel, codepoint, glyph);
+  }
+  void collect_unicodes (hb_set_t *out) const
+  {
+    accelerator_t accel (this);
+    accel.collect_unicodes (out);
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    accelerator_t accel (this);
+    accel.collect_mapping (unicodes, mapping);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    if (unlikely (!c->check_range (this, length)))
+    {
+      /* Some broken fonts have too long of a "length" value.
+       * If that is the case, just change the value to truncate
+       * the subtable at the end of the blob. */
+      uint16_t new_length = (uint16_t) hb_min ((uintptr_t) 65535,
+					       (uintptr_t) (c->end -
+							    (char *) this));
+      if (!c->try_set (&length, new_length))
+	return_trace (false);
+    }
+
+    return_trace (16 + 4 * (unsigned int) segCountX2 <= length);
+  }
+
+
+
+  protected:
+  HBUINT16	format;		/* Format number is set to 4. */
+  HBUINT16	length;		/* This is the length in bytes of the
+				 * subtable. */
+  HBUINT16	language;	/* Ignore. */
+  HBUINT16	segCountX2;	/* 2 x segCount. */
+  HBUINT16	searchRange;	/* 2 * (2**floor(log2(segCount))) */
+  HBUINT16	entrySelector;	/* log2(searchRange/2) */
+  HBUINT16	rangeShift;	/* 2 x segCount - searchRange */
+
+  UnsizedArrayOf<HBUINT16>
+		values;
+#if 0
+  HBUINT16	endCount[segCount];	/* End characterCode for each segment,
+					 * last=0xFFFFu. */
+  HBUINT16	reservedPad;		/* Set to 0. */
+  HBUINT16	startCount[segCount];	/* Start character code for each segment. */
+  HBINT16		idDelta[segCount];	/* Delta for all character codes in segment. */
+  HBUINT16	idRangeOffset[segCount];/* Offsets into glyphIdArray or 0 */
+  UnsizedArrayOf<HBUINT16>
+		glyphIdArray;	/* Glyph index array (arbitrary length) */
+#endif
+
+  public:
+  DEFINE_SIZE_ARRAY (14, values);
+};
+
+struct CmapSubtableLongGroup
+{
+  friend struct CmapSubtableFormat12;
+  friend struct CmapSubtableFormat13;
+  template<typename U>
+  friend struct CmapSubtableLongSegmented;
+  friend struct cmap;
+
+  int cmp (hb_codepoint_t codepoint) const
+  {
+    if (codepoint < startCharCode) return -1;
+    if (codepoint > endCharCode)   return +1;
+    return 0;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  private:
+  HBUINT32		startCharCode;	/* First character code in this group. */
+  HBUINT32		endCharCode;	/* Last character code in this group. */
+  HBUINT32		glyphID;	/* Glyph index; interpretation depends on
+					 * subtable format. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+DECLARE_NULL_NAMESPACE_BYTES (OT, CmapSubtableLongGroup);
+
+template <typename UINT>
+struct CmapSubtableTrimmed
+{
+  bool get_glyph (hb_codepoint_t codepoint, hb_codepoint_t *glyph) const
+  {
+    /* Rely on our implicit array bound-checking. */
+    hb_codepoint_t gid = glyphIdArray[codepoint - startCharCode];
+    if (!gid)
+      return false;
+    *glyph = gid;
+    return true;
+  }
+  void collect_unicodes (hb_set_t *out) const
+  {
+    hb_codepoint_t start = startCharCode;
+    unsigned int count = glyphIdArray.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (glyphIdArray[i])
+	out->add (start + i);
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    hb_codepoint_t start_cp = startCharCode;
+    unsigned count = glyphIdArray.len;
+    for (unsigned i = 0; i < count; i++)
+      if (glyphIdArray[i])
+      {
+	hb_codepoint_t unicode = start_cp + i;
+	hb_codepoint_t glyphid = glyphIdArray[i];
+	unicodes->add (unicode);
+	mapping->set (unicode, glyphid);
+      }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && glyphIdArray.sanitize (c));
+  }
+
+  protected:
+  UINT		formatReserved;	/* Subtable format and (maybe) padding. */
+  UINT		length;		/* Byte length of this subtable. */
+  UINT		language;	/* Ignore. */
+  UINT		startCharCode;	/* First character code covered. */
+  ArrayOf<HBGlyphID, UINT>
+		glyphIdArray;	/* Array of glyph index values for character
+				 * codes in the range. */
+  public:
+  DEFINE_SIZE_ARRAY (5 * sizeof (UINT), glyphIdArray);
+};
+
+struct CmapSubtableFormat6  : CmapSubtableTrimmed<HBUINT16> {};
+struct CmapSubtableFormat10 : CmapSubtableTrimmed<HBUINT32 > {};
+
+template <typename T>
+struct CmapSubtableLongSegmented
+{
+  friend struct cmap;
+
+  bool get_glyph (hb_codepoint_t codepoint, hb_codepoint_t *glyph) const
+  {
+    hb_codepoint_t gid = T::group_get_glyph (groups.bsearch (codepoint), codepoint);
+    if (!gid)
+      return false;
+    *glyph = gid;
+    return true;
+  }
+
+  void collect_unicodes (hb_set_t *out, unsigned int num_glyphs) const
+  {
+    for (unsigned int i = 0; i < this->groups.len; i++)
+    {
+      hb_codepoint_t start = this->groups[i].startCharCode;
+      hb_codepoint_t end = hb_min ((hb_codepoint_t) this->groups[i].endCharCode,
+				   (hb_codepoint_t) HB_UNICODE_MAX);
+      hb_codepoint_t gid = this->groups[i].glyphID;
+      if (!gid)
+      {
+	/* Intention is: if (hb_is_same (T, CmapSubtableFormat13)) continue; */
+	if (! T::group_get_glyph (this->groups[i], end)) continue;
+	start++;
+	gid++;
+      }
+      if (unlikely ((unsigned int) gid >= num_glyphs)) continue;
+      if (unlikely ((unsigned int) (gid + end - start) >= num_glyphs))
+	end = start + (hb_codepoint_t) num_glyphs - gid;
+
+      out->add_range (start, end);
+    }
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping, /* OUT */
+			unsigned num_glyphs) const
+  {
+    for (unsigned i = 0; i < this->groups.len; i++)
+    {
+      hb_codepoint_t start = this->groups[i].startCharCode;
+      hb_codepoint_t end = hb_min ((hb_codepoint_t) this->groups[i].endCharCode,
+				   (hb_codepoint_t) HB_UNICODE_MAX);
+      hb_codepoint_t gid = this->groups[i].glyphID;
+      if (!gid)
+      {
+	/* Intention is: if (hb_is_same (T, CmapSubtableFormat13)) continue; */
+	if (! T::group_get_glyph (this->groups[i], end)) continue;
+	start++;
+	gid++;
+      }
+      if (unlikely ((unsigned int) gid >= num_glyphs)) continue;
+      if (unlikely ((unsigned int) (gid + end - start) >= num_glyphs))
+	end = start + (hb_codepoint_t) num_glyphs - gid;
+
+      for (unsigned cp = start; cp <= end; cp++)
+      {
+	unicodes->add (cp);
+	mapping->set (cp, gid);
+	gid++;
+      }
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && groups.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	format;		/* Subtable format; set to 12. */
+  HBUINT16	reserved;	/* Reserved; set to 0. */
+  HBUINT32	length;		/* Byte length of this subtable. */
+  HBUINT32	language;	/* Ignore. */
+  SortedArrayOf<CmapSubtableLongGroup, HBUINT32>
+		groups;		/* Groupings. */
+  public:
+  DEFINE_SIZE_ARRAY (16, groups);
+};
+
+struct CmapSubtableFormat12 : CmapSubtableLongSegmented<CmapSubtableFormat12>
+{
+  static hb_codepoint_t group_get_glyph (const CmapSubtableLongGroup &group,
+					 hb_codepoint_t u)
+  { return likely (group.startCharCode <= group.endCharCode) ?
+	   group.glyphID + (u - group.startCharCode) : 0; }
+
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    if (it.len () == 0) return;
+    unsigned table_initpos = c->length ();
+    if (unlikely (!c->extend_min (*this))) return;
+
+    hb_codepoint_t startCharCode = 0xFFFF, endCharCode = 0xFFFF;
+    hb_codepoint_t glyphID = 0;
+
+    for (const hb_item_type<Iterator> _ : +it)
+    {
+      if (startCharCode == 0xFFFF)
+      {
+	startCharCode = _.first;
+	endCharCode = _.first;
+	glyphID = _.second;
+      }
+      else if (!_is_gid_consecutive (endCharCode, startCharCode, glyphID, _.first, _.second))
+      {
+	CmapSubtableLongGroup  grouprecord;
+	grouprecord.startCharCode = startCharCode;
+	grouprecord.endCharCode = endCharCode;
+	grouprecord.glyphID = glyphID;
+	c->copy<CmapSubtableLongGroup> (grouprecord);
+
+	startCharCode = _.first;
+	endCharCode = _.first;
+	glyphID = _.second;
+      }
+      else
+	endCharCode = _.first;
+    }
+
+    CmapSubtableLongGroup record;
+    record.startCharCode = startCharCode;
+    record.endCharCode = endCharCode;
+    record.glyphID = glyphID;
+    c->copy<CmapSubtableLongGroup> (record);
+
+    this->format = 12;
+    this->reserved = 0;
+    this->length = c->length () - table_initpos;
+    this->groups.len = (this->length - min_size)/CmapSubtableLongGroup::static_size;
+  }
+
+  static size_t get_sub_table_size (const hb_sorted_vector_t<CmapSubtableLongGroup> &groups_data)
+  { return 16 + 12 * groups_data.length; }
+
+  private:
+  static bool _is_gid_consecutive (hb_codepoint_t endCharCode,
+				   hb_codepoint_t startCharCode,
+				   hb_codepoint_t glyphID,
+				   hb_codepoint_t cp,
+				   hb_codepoint_t new_gid)
+  {
+    return (cp - 1 == endCharCode) &&
+	new_gid == glyphID + (cp - startCharCode);
+  }
+
+};
+
+struct CmapSubtableFormat13 : CmapSubtableLongSegmented<CmapSubtableFormat13>
+{
+  static hb_codepoint_t group_get_glyph (const CmapSubtableLongGroup &group,
+					 hb_codepoint_t u HB_UNUSED)
+  { return group.glyphID; }
+};
+
+typedef enum
+{
+  GLYPH_VARIANT_NOT_FOUND = 0,
+  GLYPH_VARIANT_FOUND = 1,
+  GLYPH_VARIANT_USE_DEFAULT = 2
+} glyph_variant_t;
+
+struct UnicodeValueRange
+{
+  int cmp (const hb_codepoint_t &codepoint) const
+  {
+    if (codepoint < startUnicodeValue) return -1;
+    if (codepoint > startUnicodeValue + additionalCount) return +1;
+    return 0;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT24	startUnicodeValue;	/* First value in this range. */
+  HBUINT8	additionalCount;	/* Number of additional values in this
+					 * range. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct DefaultUVS : SortedArrayOf<UnicodeValueRange, HBUINT32>
+{
+  void collect_unicodes (hb_set_t *out) const
+  {
+    unsigned int count = len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      hb_codepoint_t first = arrayZ[i].startUnicodeValue;
+      hb_codepoint_t last = hb_min ((hb_codepoint_t) (first + arrayZ[i].additionalCount),
+				    (hb_codepoint_t) HB_UNICODE_MAX);
+      out->add_range (first, last);
+    }
+  }
+
+  DefaultUVS* copy (hb_serialize_context_t *c,
+		    const hb_set_t *unicodes) const
+  {
+    DefaultUVS *out = c->start_embed<DefaultUVS> ();
+    if (unlikely (!out)) return nullptr;
+    auto snap = c->snapshot ();
+
+    HBUINT32 len;
+    len = 0;
+    if (unlikely (!c->copy<HBUINT32> (len))) return nullptr;
+    unsigned init_len = c->length ();
+
+    hb_codepoint_t lastCode = HB_MAP_VALUE_INVALID;
+    int count = -1;
+
+    for (const UnicodeValueRange& _ : as_array ())
+    {
+      for (const unsigned addcnt : hb_range ((unsigned) _.additionalCount + 1))
+      {
+	unsigned curEntry = (unsigned) _.startUnicodeValue + addcnt;
+	if (!unicodes->has (curEntry)) continue;
+	count += 1;
+	if (lastCode == HB_MAP_VALUE_INVALID)
+	  lastCode = curEntry;
+	else if (lastCode + count != curEntry)
+	{
+	  UnicodeValueRange rec;
+	  rec.startUnicodeValue = lastCode;
+	  rec.additionalCount = count - 1;
+	  c->copy<UnicodeValueRange> (rec);
+
+	  lastCode = curEntry;
+	  count = 0;
+	}
+      }
+    }
+
+    if (lastCode != HB_MAP_VALUE_INVALID)
+    {
+      UnicodeValueRange rec;
+      rec.startUnicodeValue = lastCode;
+      rec.additionalCount = count;
+      c->copy<UnicodeValueRange> (rec);
+    }
+
+    if (c->length () - init_len == 0)
+    {
+      c->revert (snap);
+      return nullptr;
+    }
+    else
+    {
+      if (unlikely (!c->check_assign (out->len, (c->length () - init_len) / UnicodeValueRange::static_size))) return nullptr;
+      return out;
+    }
+  }
+
+  public:
+  DEFINE_SIZE_ARRAY (4, *this);
+};
+
+struct UVSMapping
+{
+  int cmp (const hb_codepoint_t &codepoint) const
+  { return unicodeValue.cmp (codepoint); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT24	unicodeValue;	/* Base Unicode value of the UVS */
+  HBGlyphID	glyphID;	/* Glyph ID of the UVS */
+  public:
+  DEFINE_SIZE_STATIC (5);
+};
+
+struct NonDefaultUVS : SortedArrayOf<UVSMapping, HBUINT32>
+{
+  void collect_unicodes (hb_set_t *out) const
+  {
+    unsigned int count = len;
+    for (unsigned int i = 0; i < count; i++)
+      out->add (arrayZ[i].unicodeValue);
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    unsigned count = len;
+    for (unsigned i = 0; i < count; i++)
+    {
+      hb_codepoint_t unicode = arrayZ[i].unicodeValue;
+      hb_codepoint_t glyphid = arrayZ[i].glyphID;
+      unicodes->add (unicode);
+      mapping->set (unicode, glyphid);
+    }
+  }
+
+  void closure_glyphs (const hb_set_t      *unicodes,
+		       hb_set_t            *glyphset) const
+  {
+    + as_array ()
+    | hb_filter (unicodes, &UVSMapping::unicodeValue)
+    | hb_map (&UVSMapping::glyphID)
+    | hb_sink (glyphset)
+    ;
+  }
+
+  NonDefaultUVS* copy (hb_serialize_context_t *c,
+		       const hb_set_t *unicodes,
+		       const hb_set_t *glyphs_requested,
+		       const hb_map_t *glyph_map) const
+  {
+    NonDefaultUVS *out = c->start_embed<NonDefaultUVS> ();
+    if (unlikely (!out)) return nullptr;
+
+    auto it =
+    + as_array ()
+    | hb_filter ([&] (const UVSMapping& _)
+		 {
+		   return unicodes->has (_.unicodeValue) || glyphs_requested->has (_.glyphID);
+		 })
+    ;
+
+    if (!it) return nullptr;
+
+    HBUINT32 len;
+    len = it.len ();
+    if (unlikely (!c->copy<HBUINT32> (len))) return nullptr;
+
+    for (const UVSMapping& _ : it)
+    {
+      UVSMapping mapping;
+      mapping.unicodeValue = _.unicodeValue;
+      mapping.glyphID = glyph_map->get (_.glyphID);
+      c->copy<UVSMapping> (mapping);
+    }
+
+    return out;
+  }
+
+  public:
+  DEFINE_SIZE_ARRAY (4, *this);
+};
+
+struct VariationSelectorRecord
+{
+  glyph_variant_t get_glyph (hb_codepoint_t codepoint,
+			     hb_codepoint_t *glyph,
+			     const void *base) const
+  {
+    if ((base+defaultUVS).bfind (codepoint))
+      return GLYPH_VARIANT_USE_DEFAULT;
+    const UVSMapping &nonDefault = (base+nonDefaultUVS).bsearch (codepoint);
+    if (nonDefault.glyphID)
+    {
+      *glyph = nonDefault.glyphID;
+       return GLYPH_VARIANT_FOUND;
+    }
+    return GLYPH_VARIANT_NOT_FOUND;
+  }
+
+  VariationSelectorRecord(const VariationSelectorRecord& other)
+  {
+    *this = other;
+  }
+
+  void operator= (const VariationSelectorRecord& other)
+  {
+    varSelector = other.varSelector;
+    HBUINT32 offset = other.defaultUVS;
+    defaultUVS = offset;
+    offset = other.nonDefaultUVS;
+    nonDefaultUVS = offset;
+  }
+
+  void collect_unicodes (hb_set_t *out, const void *base) const
+  {
+    (base+defaultUVS).collect_unicodes (out);
+    (base+nonDefaultUVS).collect_unicodes (out);
+  }
+
+  void collect_mapping (const void *base,
+			hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    (base+defaultUVS).collect_unicodes (unicodes);
+    (base+nonDefaultUVS).collect_mapping (unicodes, mapping);
+  }
+
+  int cmp (const hb_codepoint_t &variation_selector) const
+  { return varSelector.cmp (variation_selector); }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  defaultUVS.sanitize (c, base) &&
+		  nonDefaultUVS.sanitize (c, base));
+  }
+
+  hb_pair_t<unsigned, unsigned>
+  copy (hb_serialize_context_t *c,
+	const hb_set_t *unicodes,
+	const hb_set_t *glyphs_requested,
+	const hb_map_t *glyph_map,
+	const void *base) const
+  {
+    auto snap = c->snapshot ();
+    auto *out = c->embed<VariationSelectorRecord> (*this);
+    if (unlikely (!out)) return hb_pair (0, 0);
+
+    out->defaultUVS = 0;
+    out->nonDefaultUVS = 0;
+
+    unsigned non_default_uvs_objidx = 0;
+    if (nonDefaultUVS != 0)
+    {
+      c->push ();
+      if (c->copy (base+nonDefaultUVS, unicodes, glyphs_requested, glyph_map))
+	non_default_uvs_objidx = c->pop_pack ();
+      else c->pop_discard ();
+    }
+
+    unsigned default_uvs_objidx = 0;
+    if (defaultUVS != 0)
+    {
+      c->push ();
+      if (c->copy (base+defaultUVS, unicodes))
+	default_uvs_objidx = c->pop_pack ();
+      else c->pop_discard ();
+    }
+
+
+    if (!default_uvs_objidx && !non_default_uvs_objidx)
+      c->revert (snap);
+
+    return hb_pair (default_uvs_objidx, non_default_uvs_objidx);
+  }
+
+  HBUINT24	varSelector;	/* Variation selector. */
+  LOffsetTo<DefaultUVS>
+		defaultUVS;	/* Offset to Default UVS Table.  May be 0. */
+  LOffsetTo<NonDefaultUVS>
+		nonDefaultUVS;	/* Offset to Non-Default UVS Table.  May be 0. */
+  public:
+  DEFINE_SIZE_STATIC (11);
+};
+
+struct CmapSubtableFormat14
+{
+  glyph_variant_t get_glyph_variant (hb_codepoint_t codepoint,
+				     hb_codepoint_t variation_selector,
+				     hb_codepoint_t *glyph) const
+  { return record.bsearch (variation_selector).get_glyph (codepoint, glyph, this); }
+
+  void collect_variation_selectors (hb_set_t *out) const
+  {
+    unsigned int count = record.len;
+    for (unsigned int i = 0; i < count; i++)
+      out->add (record.arrayZ[i].varSelector);
+  }
+  void collect_variation_unicodes (hb_codepoint_t variation_selector,
+				   hb_set_t *out) const
+  { record.bsearch (variation_selector).collect_unicodes (out, this); }
+
+  void serialize (hb_serialize_context_t *c,
+		  const hb_set_t *unicodes,
+		  const hb_set_t *glyphs_requested,
+		  const hb_map_t *glyph_map,
+		  const void *base)
+  {
+    auto snap = c->snapshot ();
+    unsigned table_initpos = c->length ();
+    const char* init_tail = c->tail;
+
+    if (unlikely (!c->extend_min (*this))) return;
+    this->format = 14;
+
+    auto src_tbl = reinterpret_cast<const CmapSubtableFormat14*> (base);
+
+    /*
+     * Some versions of OTS require that offsets are in order. Due to the use
+     * of push()/pop_pack() serializing the variation records in order results
+     * in the offsets being in reverse order (first record has the largest
+     * offset). While this is perfectly valid, it will cause some versions of
+     * OTS to consider this table bad.
+     *
+     * So to prevent this issue we serialize the variation records in reverse
+     * order, so that the offsets are ordered from small to large. Since
+     * variation records are supposed to be in increasing order of varSelector
+     * we then have to reverse the order of the written variation selector
+     * records after everything is finalized.
+     */
+    hb_vector_t<hb_pair_t<unsigned, unsigned>> obj_indices;
+    for (int i = src_tbl->record.len - 1; i >= 0; i--)
+    {
+      hb_pair_t<unsigned, unsigned> result = src_tbl->record[i].copy (c, unicodes, glyphs_requested, glyph_map, base);
+      if (result.first || result.second)
+	obj_indices.push (result);
+    }
+
+    if (c->length () - table_initpos == CmapSubtableFormat14::min_size)
+    {
+      c->revert (snap);
+      return;
+    }
+
+    if (unlikely (!c->check_success (!obj_indices.in_error ())))
+      return;
+
+    int tail_len = init_tail - c->tail;
+    c->check_assign (this->length, c->length () - table_initpos + tail_len);
+    c->check_assign (this->record.len,
+		     (c->length () - table_initpos - CmapSubtableFormat14::min_size) /
+		     VariationSelectorRecord::static_size);
+
+    /* Correct the incorrect write order by reversing the order of the variation
+       records array. */
+    _reverse_variation_records ();
+
+    /* Now that records are in the right order, we can set up the offsets. */
+    _add_links_to_variation_records (c, obj_indices);
+  }
+
+  void _reverse_variation_records ()
+  {
+    record.as_array ().reverse ();
+  }
+
+  void _add_links_to_variation_records (hb_serialize_context_t *c,
+					const hb_vector_t<hb_pair_t<unsigned, unsigned>>& obj_indices)
+  {
+    for (unsigned i = 0; i < obj_indices.length; i++)
+    {
+      /*
+       * Since the record array has been reversed (see comments in copy())
+       * but obj_indices has not been, the indices at obj_indices[i]
+       * are for the variation record at record[j].
+       */
+      int j = obj_indices.length - 1 - i;
+      c->add_link (record[j].defaultUVS, obj_indices[i].first);
+      c->add_link (record[j].nonDefaultUVS, obj_indices[i].second);
+    }
+  }
+
+  void closure_glyphs (const hb_set_t      *unicodes,
+		       hb_set_t            *glyphset) const
+  {
+    + hb_iter (record)
+    | hb_filter (hb_bool, &VariationSelectorRecord::nonDefaultUVS)
+    | hb_map (&VariationSelectorRecord::nonDefaultUVS)
+    | hb_map (hb_add (this))
+    | hb_apply ([=] (const NonDefaultUVS& _) { _.closure_glyphs (unicodes, glyphset); })
+    ;
+  }
+
+  void collect_unicodes (hb_set_t *out) const
+  {
+    for (const VariationSelectorRecord& _ : record)
+      _.collect_unicodes (out, this);
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping /* OUT */) const
+  {
+    for (const VariationSelectorRecord& _ : record)
+      _.collect_mapping (this, unicodes, mapping);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  record.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format number is set to 14. */
+  HBUINT32	length;		/* Byte length of this subtable. */
+  SortedArrayOf<VariationSelectorRecord, HBUINT32>
+		record;		/* Variation selector records; sorted
+				 * in increasing order of `varSelector'. */
+  public:
+  DEFINE_SIZE_ARRAY (10, record);
+};
+
+struct CmapSubtable
+{
+  /* Note: We intentionally do NOT implement subtable formats 2 and 8. */
+
+  bool get_glyph (hb_codepoint_t codepoint,
+		  hb_codepoint_t *glyph) const
+  {
+    switch (u.format) {
+    case  0: return u.format0 .get_glyph (codepoint, glyph);
+    case  4: return u.format4 .get_glyph (codepoint, glyph);
+    case  6: return u.format6 .get_glyph (codepoint, glyph);
+    case 10: return u.format10.get_glyph (codepoint, glyph);
+    case 12: return u.format12.get_glyph (codepoint, glyph);
+    case 13: return u.format13.get_glyph (codepoint, glyph);
+    case 14:
+    default: return false;
+    }
+  }
+  void collect_unicodes (hb_set_t *out, unsigned int num_glyphs = UINT_MAX) const
+  {
+    switch (u.format) {
+    case  0: u.format0 .collect_unicodes (out); return;
+    case  4: u.format4 .collect_unicodes (out); return;
+    case  6: u.format6 .collect_unicodes (out); return;
+    case 10: u.format10.collect_unicodes (out); return;
+    case 12: u.format12.collect_unicodes (out, num_glyphs); return;
+    case 13: u.format13.collect_unicodes (out, num_glyphs); return;
+    case 14:
+    default: return;
+    }
+  }
+
+  void collect_mapping (hb_set_t *unicodes, /* OUT */
+			hb_map_t *mapping, /* OUT */
+			unsigned num_glyphs = UINT_MAX) const
+  {
+    switch (u.format) {
+    case  0: u.format0 .collect_mapping (unicodes, mapping); return;
+    case  4: u.format4 .collect_mapping (unicodes, mapping); return;
+    case  6: u.format6 .collect_mapping (unicodes, mapping); return;
+    case 10: u.format10.collect_mapping (unicodes, mapping); return;
+    case 12: u.format12.collect_mapping (unicodes, mapping, num_glyphs); return;
+    case 13: u.format13.collect_mapping (unicodes, mapping, num_glyphs); return;
+    case 14:
+    default: return;
+    }
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  unsigned format,
+		  const hb_subset_plan_t *plan,
+		  const void *base)
+  {
+    switch (format) {
+    case  4: return u.format4.serialize (c, it);
+    case 12: return u.format12.serialize (c, it);
+    case 14: return u.format14.serialize (c, plan->unicodes, plan->glyphs_requested, plan->glyph_map, base);
+    default: return;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case  0: return_trace (u.format0 .sanitize (c));
+    case  4: return_trace (u.format4 .sanitize (c));
+    case  6: return_trace (u.format6 .sanitize (c));
+    case 10: return_trace (u.format10.sanitize (c));
+    case 12: return_trace (u.format12.sanitize (c));
+    case 13: return_trace (u.format13.sanitize (c));
+    case 14: return_trace (u.format14.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  public:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  CmapSubtableFormat0	format0;
+  CmapSubtableFormat4	format4;
+  CmapSubtableFormat6	format6;
+  CmapSubtableFormat10	format10;
+  CmapSubtableFormat12	format12;
+  CmapSubtableFormat13	format13;
+  CmapSubtableFormat14	format14;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+
+struct EncodingRecord
+{
+  int cmp (const EncodingRecord &other) const
+  {
+    int ret;
+    ret = platformID.cmp (other.platformID);
+    if (ret) return ret;
+    ret = encodingID.cmp (other.encodingID);
+    if (ret) return ret;
+    return 0;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  subtable.sanitize (c, base));
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  EncodingRecord* copy (hb_serialize_context_t *c,
+			Iterator it,
+			unsigned format,
+			const void *base,
+			const hb_subset_plan_t *plan,
+			/* INOUT */ unsigned *objidx) const
+  {
+    TRACE_SERIALIZE (this);
+    auto snap = c->snapshot ();
+    auto *out = c->embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+    out->subtable = 0;
+
+    if (*objidx == 0)
+    {
+      CmapSubtable *cmapsubtable = c->push<CmapSubtable> ();
+      unsigned origin_length = c->length ();
+      cmapsubtable->serialize (c, it, format, plan, &(base+subtable));
+      if (c->length () - origin_length > 0) *objidx = c->pop_pack ();
+      else c->pop_discard ();
+    }
+
+    if (*objidx == 0)
+    {
+      c->revert (snap);
+      return_trace (nullptr);
+    }
+
+    c->add_link (out->subtable, *objidx);
+    return_trace (out);
+  }
+
+  HBUINT16	platformID;	/* Platform ID. */
+  HBUINT16	encodingID;	/* Platform-specific encoding ID. */
+  LOffsetTo<CmapSubtable>
+		subtable;	/* Byte offset from beginning of table to the subtable for this encoding. */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct cmap
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_cmap;
+
+  template<typename Iterator, typename EncodingRecIter,
+	   hb_requires (hb_is_iterator (EncodingRecIter))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  EncodingRecIter encodingrec_iter,
+		  const void *base,
+		  const hb_subset_plan_t *plan)
+  {
+    if (unlikely (!c->extend_min ((*this))))  return;
+    this->version = 0;
+
+    unsigned format4objidx = 0, format12objidx = 0, format14objidx = 0;
+
+    for (const EncodingRecord& _ : encodingrec_iter)
+    {
+      unsigned format = (base+_.subtable).u.format;
+      if (!plan->glyphs_requested->is_empty ())
+      {
+	hb_set_t unicodes_set;
+	hb_map_t cp_glyphid_map;
+	(base+_.subtable).collect_mapping (&unicodes_set, &cp_glyphid_map);
+
+	auto table_iter =
+	+ hb_zip (unicodes_set.iter(), unicodes_set.iter() | hb_map(cp_glyphid_map))
+	| hb_filter (plan->_glyphset, hb_second)
+	| hb_filter ([plan] (const hb_pair_t<hb_codepoint_t, hb_codepoint_t>& p)
+		     {
+		       return plan->unicodes->has (p.first) ||
+			      plan->glyphs_requested->has (p.second);
+		     })
+	| hb_map ([plan] (const hb_pair_t<hb_codepoint_t, hb_codepoint_t>& p_org)
+		  {
+		    return hb_pair_t<hb_codepoint_t, hb_codepoint_t> (p_org.first, plan->glyph_map->get(p_org.second));
+		  })
+	;
+
+	if (format == 4) c->copy (_, table_iter, 4u, base, plan, &format4objidx);
+	else if (format == 12) c->copy (_, table_iter, 12u, base, plan, &format12objidx);
+	else if (format == 14) c->copy (_, table_iter, 14u, base, plan, &format14objidx);
+      }
+      /* when --gids option is not used, we iterate input unicodes instead of
+       * all codepoints in each subtable, which is more efficient */
+      else
+      {
+	hb_set_t unicodes_set;
+	(base+_.subtable).collect_unicodes (&unicodes_set);
+
+	if (format == 4) c->copy (_, + it | hb_filter (unicodes_set, hb_first), 4u, base, plan, &format4objidx);
+	else if (format == 12) c->copy (_, + it | hb_filter (unicodes_set, hb_first), 12u, base, plan, &format12objidx);
+	else if (format == 14) c->copy (_, it, 14u, base, plan, &format14objidx);
+      }
+    }
+
+    c->check_assign(this->encodingRecord.len, (c->length () - cmap::min_size)/EncodingRecord::static_size);
+  }
+
+  void closure_glyphs (const hb_set_t      *unicodes,
+		       hb_set_t            *glyphset) const
+  {
+    + hb_iter (encodingRecord)
+    | hb_map (&EncodingRecord::subtable)
+    | hb_map (hb_add (this))
+    | hb_filter ([&] (const CmapSubtable& _) { return _.u.format == 14; })
+    | hb_apply ([=] (const CmapSubtable& _) { _.u.format14.closure_glyphs (unicodes, glyphset); })
+    ;
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    cmap *cmap_prime = c->serializer->start_embed<cmap> ();
+    if (unlikely (!c->serializer->check_success (cmap_prime))) return_trace (false);
+
+    auto encodingrec_iter =
+    + hb_iter (encodingRecord)
+    | hb_filter ([&] (const EncodingRecord& _)
+		{
+		  if ((_.platformID == 0 && _.encodingID == 3) ||
+		      (_.platformID == 0 && _.encodingID == 4) ||
+		      (_.platformID == 3 && _.encodingID == 1) ||
+		      (_.platformID == 3 && _.encodingID == 10) ||
+		      (this + _.subtable).u.format == 14)
+		    return true;
+
+		  return false;
+		})
+    ;
+
+    if (unlikely (!encodingrec_iter.len ())) return_trace (false);
+
+    const EncodingRecord *unicode_bmp= nullptr, *unicode_ucs4 = nullptr, *ms_bmp = nullptr, *ms_ucs4 = nullptr;
+    bool has_format12 = false;
+
+    for (const EncodingRecord& _ : encodingrec_iter)
+    {
+      unsigned format = (this + _.subtable).u.format;
+      if (format == 12) has_format12 = true;
+
+      const EncodingRecord *table = hb_addressof (_);
+      if      (_.platformID == 0 && _.encodingID ==  3) unicode_bmp = table;
+      else if (_.platformID == 0 && _.encodingID ==  4) unicode_ucs4 = table;
+      else if (_.platformID == 3 && _.encodingID ==  1) ms_bmp = table;
+      else if (_.platformID == 3 && _.encodingID == 10) ms_ucs4 = table;
+    }
+
+    if (unlikely (!has_format12 && !unicode_bmp && !ms_bmp)) return_trace (false);
+    if (unlikely (has_format12 && (!unicode_ucs4 && !ms_ucs4))) return_trace (false);
+
+    auto it =
+    + hb_iter (c->plan->unicodes)
+    | hb_map ([&] (hb_codepoint_t _)
+	      {
+		hb_codepoint_t new_gid = HB_MAP_VALUE_INVALID;
+		c->plan->new_gid_for_codepoint (_, &new_gid);
+		return hb_pair_t<hb_codepoint_t, hb_codepoint_t> (_, new_gid);
+	      })
+    | hb_filter ([&] (const hb_pair_t<hb_codepoint_t, hb_codepoint_t> _)
+		 { return (_.second != HB_MAP_VALUE_INVALID); })
+    ;
+    cmap_prime->serialize (c->serializer, it, encodingrec_iter, this, c->plan);
+    return_trace (true);
+  }
+
+  const CmapSubtable *find_best_subtable (bool *symbol = nullptr) const
+  {
+    if (symbol) *symbol = false;
+
+    const CmapSubtable *subtable;
+
+    /* Symbol subtable.
+     * Prefer symbol if available.
+     * https://github.com/harfbuzz/harfbuzz/issues/1918 */
+    if ((subtable = this->find_subtable (3, 0)))
+    {
+      if (symbol) *symbol = true;
+      return subtable;
+    }
+
+    /* 32-bit subtables. */
+    if ((subtable = this->find_subtable (3, 10))) return subtable;
+    if ((subtable = this->find_subtable (0, 6))) return subtable;
+    if ((subtable = this->find_subtable (0, 4))) return subtable;
+
+    /* 16-bit subtables. */
+    if ((subtable = this->find_subtable (3, 1))) return subtable;
+    if ((subtable = this->find_subtable (0, 3))) return subtable;
+    if ((subtable = this->find_subtable (0, 2))) return subtable;
+    if ((subtable = this->find_subtable (0, 1))) return subtable;
+    if ((subtable = this->find_subtable (0, 0))) return subtable;
+
+    /* Meh. */
+    return &Null (CmapSubtable);
+  }
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      this->table = hb_sanitize_context_t ().reference_table<cmap> (face);
+      bool symbol;
+      this->subtable = table->find_best_subtable (&symbol);
+      this->subtable_uvs = &Null (CmapSubtableFormat14);
+      {
+	const CmapSubtable *st = table->find_subtable (0, 5);
+	if (st && st->u.format == 14)
+	  subtable_uvs = &st->u.format14;
+      }
+
+      this->get_glyph_data = subtable;
+      if (unlikely (symbol))
+	this->get_glyph_funcZ = get_glyph_from_symbol<CmapSubtable>;
+      else
+      {
+	switch (subtable->u.format) {
+	/* Accelerate format 4 and format 12. */
+	default:
+	  this->get_glyph_funcZ = get_glyph_from<CmapSubtable>;
+	  break;
+	case 12:
+	  this->get_glyph_funcZ = get_glyph_from<CmapSubtableFormat12>;
+	  break;
+	case  4:
+	{
+	  this->format4_accel.init (&subtable->u.format4);
+	  this->get_glyph_data = &this->format4_accel;
+	  this->get_glyph_funcZ = this->format4_accel.get_glyph_func;
+	  break;
+	}
+	}
+      }
+    }
+
+    void fini () { this->table.destroy (); }
+
+    bool get_nominal_glyph (hb_codepoint_t  unicode,
+			    hb_codepoint_t *glyph) const
+    {
+      if (unlikely (!this->get_glyph_funcZ)) return false;
+      return this->get_glyph_funcZ (this->get_glyph_data, unicode, glyph);
+    }
+    unsigned int get_nominal_glyphs (unsigned int count,
+				     const hb_codepoint_t *first_unicode,
+				     unsigned int unicode_stride,
+				     hb_codepoint_t *first_glyph,
+				     unsigned int glyph_stride) const
+    {
+      if (unlikely (!this->get_glyph_funcZ)) return 0;
+
+      hb_cmap_get_glyph_func_t get_glyph_funcZ = this->get_glyph_funcZ;
+      const void *get_glyph_data = this->get_glyph_data;
+
+      unsigned int done;
+      for (done = 0;
+	   done < count && get_glyph_funcZ (get_glyph_data, *first_unicode, first_glyph);
+	   done++)
+      {
+	first_unicode = &StructAtOffsetUnaligned<hb_codepoint_t> (first_unicode, unicode_stride);
+	first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+      }
+      return done;
+    }
+
+    bool get_variation_glyph (hb_codepoint_t  unicode,
+			      hb_codepoint_t  variation_selector,
+			      hb_codepoint_t *glyph) const
+    {
+      switch (this->subtable_uvs->get_glyph_variant (unicode,
+						     variation_selector,
+						     glyph))
+      {
+	case GLYPH_VARIANT_NOT_FOUND:	return false;
+	case GLYPH_VARIANT_FOUND:	return true;
+	case GLYPH_VARIANT_USE_DEFAULT:	break;
+      }
+
+      return get_nominal_glyph (unicode, glyph);
+    }
+
+    void collect_unicodes (hb_set_t *out, unsigned int num_glyphs) const
+    { subtable->collect_unicodes (out, num_glyphs); }
+    void collect_mapping (hb_set_t *unicodes, hb_map_t *mapping,
+			  unsigned num_glyphs = UINT_MAX) const
+    { subtable->collect_mapping (unicodes, mapping, num_glyphs); }
+    void collect_variation_selectors (hb_set_t *out) const
+    { subtable_uvs->collect_variation_selectors (out); }
+    void collect_variation_unicodes (hb_codepoint_t variation_selector,
+				     hb_set_t *out) const
+    { subtable_uvs->collect_variation_unicodes (variation_selector, out); }
+
+    protected:
+    typedef bool (*hb_cmap_get_glyph_func_t) (const void *obj,
+					      hb_codepoint_t codepoint,
+					      hb_codepoint_t *glyph);
+
+    template <typename Type>
+    HB_INTERNAL static bool get_glyph_from (const void *obj,
+					    hb_codepoint_t codepoint,
+					    hb_codepoint_t *glyph)
+    {
+      const Type *typed_obj = (const Type *) obj;
+      return typed_obj->get_glyph (codepoint, glyph);
+    }
+
+    template <typename Type>
+    HB_INTERNAL static bool get_glyph_from_symbol (const void *obj,
+						   hb_codepoint_t codepoint,
+						   hb_codepoint_t *glyph)
+    {
+      const Type *typed_obj = (const Type *) obj;
+      if (likely (typed_obj->get_glyph (codepoint, glyph)))
+	return true;
+
+      if (codepoint <= 0x00FFu)
+      {
+	/* For symbol-encoded OpenType fonts, we duplicate the
+	 * U+F000..F0FF range at U+0000..U+00FF.  That's what
+	 * Windows seems to do, and that's hinted about at:
+	 * https://docs.microsoft.com/en-us/typography/opentype/spec/recom
+	 * under "Non-Standard (Symbol) Fonts". */
+	return typed_obj->get_glyph (0xF000u + codepoint, glyph);
+      }
+
+      return false;
+    }
+
+    private:
+    hb_nonnull_ptr_t<const CmapSubtable> subtable;
+    hb_nonnull_ptr_t<const CmapSubtableFormat14> subtable_uvs;
+
+    hb_cmap_get_glyph_func_t get_glyph_funcZ;
+    const void *get_glyph_data;
+
+    CmapSubtableFormat4::accelerator_t format4_accel;
+
+    public:
+    hb_blob_ptr_t<cmap> table;
+  };
+
+  protected:
+
+  const CmapSubtable *find_subtable (unsigned int platform_id,
+				     unsigned int encoding_id) const
+  {
+    EncodingRecord key;
+    key.platformID = platform_id;
+    key.encodingID = encoding_id;
+
+    const EncodingRecord &result = encodingRecord.bsearch (key);
+    if (!result.subtable)
+      return nullptr;
+
+    return &(this+result.subtable);
+  }
+
+  const EncodingRecord *find_encodingrec (unsigned int platform_id,
+					  unsigned int encoding_id) const
+  {
+    EncodingRecord key;
+    key.platformID = platform_id;
+    key.encodingID = encoding_id;
+
+    return encodingRecord.as_array ().bsearch (key);
+  }
+
+  bool find_subtable (unsigned format) const
+  {
+    auto it =
+    + hb_iter (encodingRecord)
+    | hb_map (&EncodingRecord::subtable)
+    | hb_map (hb_add (this))
+    | hb_filter ([&] (const CmapSubtable& _) { return _.u.format == format; })
+    ;
+
+    return it.len ();
+  }
+
+  public:
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (version == 0) &&
+		  encodingRecord.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	version;	/* Table version number (0). */
+  SortedArrayOf<EncodingRecord>
+		encodingRecord;	/* Encoding tables. */
+  public:
+  DEFINE_SIZE_ARRAY (4, encodingRecord);
+};
+
+struct cmap_accelerator_t : cmap::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_CMAP_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color-cbdt-table.hh b/thirdparty/harfbuzz/src/hb-ot-color-cbdt-table.hh
new file mode 100644
index 0000000000..aaa1c37c64
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color-cbdt-table.hh
@@ -0,0 +1,985 @@
+/*
+ * Copyright © 2016  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Seigo Nonaka, Calder Kitagawa
+ */
+
+#ifndef HB_OT_COLOR_CBDT_TABLE_HH
+#define HB_OT_COLOR_CBDT_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * CBLC -- Color Bitmap Location
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/cblc
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/eblc
+ * CBDT -- Color Bitmap Data
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/cbdt
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/ebdt
+ */
+#define HB_OT_TAG_CBLC HB_TAG('C','B','L','C')
+#define HB_OT_TAG_CBDT HB_TAG('C','B','D','T')
+
+
+namespace OT {
+
+struct cblc_bitmap_size_subset_context_t
+{
+  const char *cbdt;
+  unsigned int cbdt_length;
+  hb_vector_t<char> *cbdt_prime;
+  unsigned int size;		/* INOUT
+				 *  Input: old size of IndexSubtable
+				 *  Output: new size of IndexSubtable
+				 */
+  unsigned int num_tables;	/* INOUT
+				 *  Input: old number of subtables.
+				 *  Output: new number of subtables.
+				 */
+  hb_codepoint_t start_glyph;	/* OUT */
+  hb_codepoint_t end_glyph;	/* OUT */
+};
+
+static inline bool
+_copy_data_to_cbdt (hb_vector_t<char> *cbdt_prime,
+		    const void        *data,
+		    unsigned           length)
+{
+  unsigned int new_len = cbdt_prime->length + length;
+  if (unlikely (!cbdt_prime->alloc (new_len))) return false;
+  memcpy (cbdt_prime->arrayZ + cbdt_prime->length, data, length);
+  cbdt_prime->length = new_len;
+  return true;
+}
+
+struct SmallGlyphMetrics
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  void get_extents (hb_font_t *font, hb_glyph_extents_t *extents) const
+  {
+    extents->x_bearing = font->em_scale_x (bearingX);
+    extents->y_bearing = font->em_scale_y (bearingY);
+    extents->width = font->em_scale_x (width);
+    extents->height = font->em_scale_y (-static_cast<int>(height));
+  }
+
+  HBUINT8	height;
+  HBUINT8	width;
+  HBINT8	bearingX;
+  HBINT8	bearingY;
+  HBUINT8	advance;
+  public:
+  DEFINE_SIZE_STATIC (5);
+};
+
+struct BigGlyphMetrics : SmallGlyphMetrics
+{
+  HBINT8	vertBearingX;
+  HBINT8	vertBearingY;
+  HBUINT8	vertAdvance;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct SBitLineMetrics
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBINT8	ascender;
+  HBINT8	decender;
+  HBUINT8	widthMax;
+  HBINT8	caretSlopeNumerator;
+  HBINT8	caretSlopeDenominator;
+  HBINT8	caretOffset;
+  HBINT8	minOriginSB;
+  HBINT8	minAdvanceSB;
+  HBINT8	maxBeforeBL;
+  HBINT8	minAfterBL;
+  HBINT8	padding1;
+  HBINT8	padding2;
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+
+/*
+ * Index Subtables.
+ */
+
+struct IndexSubtableHeader
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT16	indexFormat;
+  HBUINT16	imageFormat;
+  HBUINT32	imageDataOffset;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+template <typename OffsetType>
+struct IndexSubtableFormat1Or3
+{
+  bool sanitize (hb_sanitize_context_t *c, unsigned int glyph_count) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  offsetArrayZ.sanitize (c, glyph_count + 1));
+  }
+
+  bool get_image_data (unsigned int idx,
+		       unsigned int *offset,
+		       unsigned int *length) const
+  {
+    if (unlikely (offsetArrayZ[idx + 1] <= offsetArrayZ[idx]))
+      return false;
+
+    *offset = header.imageDataOffset + offsetArrayZ[idx];
+    *length = offsetArrayZ[idx + 1] - offsetArrayZ[idx];
+    return true;
+  }
+
+  bool add_offset (hb_serialize_context_t *c,
+		   unsigned int offset,
+		   unsigned int *size /* OUT (accumulated) */)
+  {
+    TRACE_SERIALIZE (this);
+    Offset<OffsetType> embedded_offset;
+    embedded_offset = offset;
+    *size += sizeof (OffsetType);
+    auto *o = c->embed (embedded_offset);
+    return_trace ((bool) o);
+  }
+
+  IndexSubtableHeader	header;
+  UnsizedArrayOf<Offset<OffsetType>>
+			offsetArrayZ;
+  public:
+  DEFINE_SIZE_ARRAY (8, offsetArrayZ);
+};
+
+struct IndexSubtableFormat1 : IndexSubtableFormat1Or3<HBUINT32> {};
+struct IndexSubtableFormat3 : IndexSubtableFormat1Or3<HBUINT16> {};
+
+struct IndexSubtable
+{
+  bool sanitize (hb_sanitize_context_t *c, unsigned int glyph_count) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.header.sanitize (c)) return_trace (false);
+    switch (u.header.indexFormat)
+    {
+    case 1: return_trace (u.format1.sanitize (c, glyph_count));
+    case 3: return_trace (u.format3.sanitize (c, glyph_count));
+    default:return_trace (true);
+    }
+  }
+
+  bool
+  finish_subtable (hb_serialize_context_t *c,
+		   unsigned int cbdt_prime_len,
+		   unsigned int num_glyphs,
+		   unsigned int *size /* OUT (accumulated) */)
+  {
+    TRACE_SERIALIZE (this);
+
+    unsigned int local_offset = cbdt_prime_len - u.header.imageDataOffset;
+    switch (u.header.indexFormat)
+    {
+    case 1: return_trace (u.format1.add_offset (c, local_offset, size));
+    case 3: {
+      if (!u.format3.add_offset (c, local_offset, size))
+	return_trace (false);
+      if (!(num_glyphs & 0x01))  // Pad to 32-bit alignment if needed.
+	return_trace (u.format3.add_offset (c, 0, size));
+      return_trace (true);
+    }
+    // TODO: implement 2, 4, 5.
+    case 2: case 4:  // No-op.
+    case 5:  // Pad to 32-bit aligned.
+    default: return_trace (false);
+    }
+  }
+
+  bool
+  fill_missing_glyphs (hb_serialize_context_t *c,
+		       unsigned int cbdt_prime_len,
+		       unsigned int num_missing,
+		       unsigned int *size /* OUT (accumulated) */,
+		       unsigned int *num_glyphs /* OUT (accumulated) */)
+  {
+    TRACE_SERIALIZE (this);
+
+    unsigned int local_offset = cbdt_prime_len - u.header.imageDataOffset;
+    switch (u.header.indexFormat)
+    {
+    case 1: {
+      for (unsigned int i = 0; i < num_missing; i++)
+      {
+	if (unlikely (!u.format1.add_offset (c, local_offset, size)))
+	  return_trace (false);
+	*num_glyphs += 1;
+      }
+      return_trace (true);
+    }
+    case 3: {
+      for (unsigned int i = 0; i < num_missing; i++)
+      {
+	if (unlikely (!u.format3.add_offset (c, local_offset, size)))
+	  return_trace (false);
+	*num_glyphs += 1;
+      }
+      return_trace (true);
+    }
+    // TODO: implement 2, 4, 5.
+    case 2:  // Add empty space in cbdt_prime?.
+    case 4: case 5:  // No-op as sparse is supported.
+    default: return_trace (false);
+    }
+  }
+
+  bool
+  copy_glyph_at_idx (hb_serialize_context_t *c, unsigned int idx,
+		     const char *cbdt, unsigned int cbdt_length,
+		     hb_vector_t<char> *cbdt_prime /* INOUT */,
+		     IndexSubtable *subtable_prime /* INOUT */,
+		     unsigned int *size /* OUT (accumulated) */) const
+  {
+    TRACE_SERIALIZE (this);
+
+    unsigned int offset, length, format;
+    if (unlikely (!get_image_data (idx, &offset, &length, &format))) return_trace (false);
+    if (unlikely (offset > cbdt_length || cbdt_length - offset < length)) return_trace (false);
+
+    auto *header_prime = subtable_prime->get_header ();
+    unsigned int new_local_offset = cbdt_prime->length - (unsigned int) header_prime->imageDataOffset;
+    if (unlikely (!_copy_data_to_cbdt (cbdt_prime, cbdt + offset, length))) return_trace (false);
+
+    return_trace (subtable_prime->add_offset (c, new_local_offset, size));
+  }
+
+  bool
+  add_offset (hb_serialize_context_t *c, unsigned int local_offset,
+	      unsigned int *size /* OUT (accumulated) */)
+  {
+    TRACE_SERIALIZE (this);
+    switch (u.header.indexFormat)
+    {
+    case 1: return_trace (u.format1.add_offset (c, local_offset, size));
+    case 3: return_trace (u.format3.add_offset (c, local_offset, size));
+    // TODO: Implement tables 2, 4, 5
+    case 2:  // Should be a no-op.
+    case 4: case 5:  // Handle sparse cases.
+    default: return_trace (false);
+    }
+  }
+
+  bool get_extents (hb_glyph_extents_t *extents HB_UNUSED) const
+  {
+    switch (u.header.indexFormat)
+    {
+    case 2: case 5: /* TODO */
+    case 1: case 3: case 4: /* Variable-metrics formats do not have metrics here. */
+    default:return (false);
+    }
+  }
+
+  bool
+  get_image_data (unsigned int idx, unsigned int *offset,
+		  unsigned int *length, unsigned int *format) const
+  {
+    *format = u.header.imageFormat;
+    switch (u.header.indexFormat)
+    {
+    case 1: return u.format1.get_image_data (idx, offset, length);
+    case 3: return u.format3.get_image_data (idx, offset, length);
+    default: return false;
+    }
+  }
+
+  const IndexSubtableHeader* get_header () const { return &u.header; }
+
+  void populate_header (unsigned index_format,
+			unsigned image_format,
+			unsigned int image_data_offset,
+			unsigned int *size)
+  {
+    u.header.indexFormat = index_format;
+    u.header.imageFormat = image_format;
+    u.header.imageDataOffset = image_data_offset;
+    switch (u.header.indexFormat)
+    {
+    case 1: *size += IndexSubtableFormat1::min_size; break;
+    case 3: *size += IndexSubtableFormat3::min_size; break;
+    }
+  }
+
+  protected:
+  union {
+  IndexSubtableHeader	header;
+  IndexSubtableFormat1	format1;
+  IndexSubtableFormat3	format3;
+  /* TODO: Format 2, 4, 5. */
+  } u;
+  public:
+  DEFINE_SIZE_UNION (8, header);
+};
+
+struct IndexSubtableRecord
+{
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  firstGlyphIndex <= lastGlyphIndex &&
+		  offsetToSubtable.sanitize (c, base, lastGlyphIndex - firstGlyphIndex + 1));
+  }
+
+  const IndexSubtable* get_subtable (const void *base) const
+  {
+    return &(base+offsetToSubtable);
+  }
+
+  bool add_new_subtable (hb_subset_context_t* c,
+			 cblc_bitmap_size_subset_context_t *bitmap_size_context,
+			 IndexSubtableRecord *record,
+			 const hb_vector_t<hb_pair_t<hb_codepoint_t, const IndexSubtableRecord*>> *lookup, /* IN */
+			 const void *base,
+			 unsigned int *start /* INOUT */) const
+  {
+    TRACE_SERIALIZE (this);
+
+    auto *subtable = c->serializer->start_embed<IndexSubtable> ();
+    if (unlikely (!subtable)) return_trace (false);
+    if (unlikely (!c->serializer->extend_min (subtable))) return_trace (false);
+
+    auto *old_subtable = get_subtable (base);
+    auto *old_header = old_subtable->get_header ();
+
+    subtable->populate_header (old_header->indexFormat,
+			       old_header->imageFormat,
+			       bitmap_size_context->cbdt_prime->length,
+			       &bitmap_size_context->size);
+
+    unsigned int num_glyphs = 0;
+    bool early_exit = false;
+    for (unsigned int i = *start; i < lookup->length; i++)
+    {
+      hb_codepoint_t new_gid = (*lookup)[i].first;
+      const IndexSubtableRecord *next_record = (*lookup)[i].second;
+      const IndexSubtable *next_subtable = next_record->get_subtable (base);
+      auto *next_header = next_subtable->get_header ();
+      if (next_header != old_header)
+      {
+	*start = i;
+	early_exit = true;
+	break;
+      }
+      unsigned int num_missing = record->add_glyph_for_subset (new_gid);
+      if (unlikely (!subtable->fill_missing_glyphs (c->serializer,
+						    bitmap_size_context->cbdt_prime->length,
+						    num_missing,
+						    &bitmap_size_context->size,
+						    &num_glyphs)))
+	return_trace (false);
+
+      hb_codepoint_t old_gid = 0;
+      c->plan->old_gid_for_new_gid (new_gid, &old_gid);
+      if (old_gid < next_record->firstGlyphIndex)
+	return_trace (false);
+
+      unsigned int old_idx = (unsigned int) old_gid - next_record->firstGlyphIndex;
+      if (unlikely (!next_subtable->copy_glyph_at_idx (c->serializer,
+						       old_idx,
+						       bitmap_size_context->cbdt,
+						       bitmap_size_context->cbdt_length,
+						       bitmap_size_context->cbdt_prime,
+						       subtable,
+						       &bitmap_size_context->size)))
+	return_trace (false);
+      num_glyphs += 1;
+    }
+    if (!early_exit)
+      *start = lookup->length;
+    if (unlikely (!subtable->finish_subtable (c->serializer,
+					      bitmap_size_context->cbdt_prime->length,
+					      num_glyphs,
+					      &bitmap_size_context->size)))
+      return_trace (false);
+    return_trace (true);
+  }
+
+  bool add_new_record (hb_subset_context_t *c,
+		       cblc_bitmap_size_subset_context_t *bitmap_size_context,
+		       const hb_vector_t<hb_pair_t<hb_codepoint_t, const IndexSubtableRecord*>> *lookup, /* IN */
+		       const void *base,
+		       unsigned int *start, /* INOUT */
+		       hb_vector_t<IndexSubtableRecord>* records /* INOUT */) const
+  {
+    TRACE_SERIALIZE (this);
+    auto snap = c->serializer->snapshot ();
+    unsigned int old_size = bitmap_size_context->size;
+    unsigned int old_cbdt_prime_length = bitmap_size_context->cbdt_prime->length;
+
+    // Set to invalid state to indicate filling glyphs is not yet started.
+    if (unlikely (!records->resize (records->length + 1)))
+      return_trace (c->serializer->check_success (false));
+
+    (*records)[records->length - 1].firstGlyphIndex = 1;
+    (*records)[records->length - 1].lastGlyphIndex = 0;
+    bitmap_size_context->size += IndexSubtableRecord::min_size;
+
+    c->serializer->push ();
+
+    if (unlikely (!add_new_subtable (c, bitmap_size_context, &((*records)[records->length - 1]), lookup, base, start)))
+    {
+      c->serializer->pop_discard ();
+      c->serializer->revert (snap);
+      bitmap_size_context->cbdt_prime->shrink (old_cbdt_prime_length);
+      bitmap_size_context->size = old_size;
+      records->resize (records->length - 1);
+      return_trace (false);
+    }
+
+    bitmap_size_context->num_tables += 1;
+    return_trace (true);
+  }
+
+  unsigned int add_glyph_for_subset (hb_codepoint_t gid)
+  {
+    if (firstGlyphIndex > lastGlyphIndex)
+    {
+      firstGlyphIndex = gid;
+      lastGlyphIndex = gid;
+      return 0;
+    }
+    // TODO maybe assert? this shouldn't occur.
+    if (lastGlyphIndex > gid)
+      return 0;
+    unsigned int num_missing = (unsigned int) (gid - lastGlyphIndex - 1);
+    lastGlyphIndex = gid;
+    return num_missing;
+  }
+
+  bool get_extents (hb_glyph_extents_t *extents, const void *base) const
+  { return (base+offsetToSubtable).get_extents (extents); }
+
+  bool get_image_data (unsigned int  gid,
+		       const void   *base,
+		       unsigned int *offset,
+		       unsigned int *length,
+		       unsigned int *format) const
+  {
+    if (gid < firstGlyphIndex || gid > lastGlyphIndex) return false;
+    return (base+offsetToSubtable).get_image_data (gid - firstGlyphIndex,
+						   offset, length, format);
+  }
+
+  HBGlyphID			firstGlyphIndex;
+  HBGlyphID			lastGlyphIndex;
+  LOffsetTo<IndexSubtable>	offsetToSubtable;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct IndexSubtableArray
+{
+  friend struct CBDT;
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int count) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (indexSubtablesZ.sanitize (c, count, this));
+  }
+
+  void
+  build_lookup (hb_subset_context_t *c, cblc_bitmap_size_subset_context_t *bitmap_size_context,
+		hb_vector_t<hb_pair_t<hb_codepoint_t,
+		const IndexSubtableRecord*>> *lookup /* OUT */) const
+  {
+    bool start_glyph_is_set = false;
+    for (hb_codepoint_t new_gid = 0; new_gid < c->plan->num_output_glyphs (); new_gid++)
+    {
+      hb_codepoint_t old_gid;
+      if (unlikely (!c->plan->old_gid_for_new_gid (new_gid, &old_gid))) continue;
+
+      const IndexSubtableRecord* record = find_table (old_gid, bitmap_size_context->num_tables);
+      if (unlikely (!record)) continue;
+
+      // Don't add gaps to the lookup. The best way to determine if a glyph is a
+      // gap is that it has no image data.
+      unsigned int offset, length, format;
+      if (unlikely (!record->get_image_data (old_gid, this, &offset, &length, &format))) continue;
+
+      lookup->push (hb_pair_t<hb_codepoint_t, const IndexSubtableRecord*> (new_gid, record));
+
+      if (!start_glyph_is_set)
+      {
+	bitmap_size_context->start_glyph = new_gid;
+	start_glyph_is_set = true;
+      }
+
+      bitmap_size_context->end_glyph = new_gid;
+    }
+  }
+
+  bool
+  subset (hb_subset_context_t *c,
+	  cblc_bitmap_size_subset_context_t *bitmap_size_context) const
+  {
+    TRACE_SUBSET (this);
+
+    auto *dst = c->serializer->start_embed<IndexSubtableArray> ();
+    if (unlikely (!dst)) return_trace (false);
+
+    hb_vector_t<hb_pair_t<hb_codepoint_t, const IndexSubtableRecord*>> lookup;
+    build_lookup (c, bitmap_size_context, &lookup);
+    if (unlikely (lookup.in_error ()))
+      return c->serializer->check_success (false);
+
+    bitmap_size_context->size = 0;
+    bitmap_size_context->num_tables = 0;
+    hb_vector_t<IndexSubtableRecord> records;
+    for (unsigned int start = 0; start < lookup.length;)
+    {
+      if (unlikely (!lookup[start].second->add_new_record (c, bitmap_size_context, &lookup, this, &start, &records)))
+      {
+	// Discard any leftover pushes to the serializer from successful records.
+	for (unsigned int i = 0; i < records.length; i++)
+	  c->serializer->pop_discard ();
+	return_trace (false);
+      }
+    }
+
+    /* Workaround to ensure offset ordering is from least to greatest when
+     * resolving links. */
+    hb_vector_t<hb_serialize_context_t::objidx_t> objidxs;
+    for (unsigned int i = 0; i < records.length; i++)
+      objidxs.push (c->serializer->pop_pack ());
+    for (unsigned int i = 0; i < records.length; i++)
+    {
+      IndexSubtableRecord* record = c->serializer->embed (records[i]);
+      if (unlikely (!record)) return_trace (false);
+      c->serializer->add_link (record->offsetToSubtable, objidxs[records.length - 1 - i]);
+    }
+    return_trace (true);
+  }
+
+  public:
+  const IndexSubtableRecord* find_table (hb_codepoint_t glyph, unsigned int numTables) const
+  {
+    for (unsigned int i = 0; i < numTables; ++i)
+    {
+      unsigned int firstGlyphIndex = indexSubtablesZ[i].firstGlyphIndex;
+      unsigned int lastGlyphIndex = indexSubtablesZ[i].lastGlyphIndex;
+      if (firstGlyphIndex <= glyph && glyph <= lastGlyphIndex)
+	return &indexSubtablesZ[i];
+    }
+    return nullptr;
+  }
+
+  protected:
+  UnsizedArrayOf<IndexSubtableRecord>	indexSubtablesZ;
+};
+
+struct BitmapSizeTable
+{
+  friend struct CBLC;
+  friend struct CBDT;
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  indexSubtableArrayOffset.sanitize (c, base, numberOfIndexSubtables) &&
+		  horizontal.sanitize (c) &&
+		  vertical.sanitize (c));
+  }
+
+  const IndexSubtableRecord *
+  find_table (hb_codepoint_t glyph, const void *base, const void **out_base) const
+  {
+    *out_base = &(base+indexSubtableArrayOffset);
+    return (base+indexSubtableArrayOffset).find_table (glyph, numberOfIndexSubtables);
+  }
+
+  bool
+  subset (hb_subset_context_t *c, const void *base,
+	  const char *cbdt, unsigned int cbdt_length,
+	  hb_vector_t<char> *cbdt_prime /* INOUT */) const
+  {
+    TRACE_SUBSET (this);
+    auto *out_table = c->serializer->embed (this);
+    if (unlikely (!out_table)) return_trace (false);
+
+    cblc_bitmap_size_subset_context_t bitmap_size_context;
+    bitmap_size_context.cbdt = cbdt;
+    bitmap_size_context.cbdt_length = cbdt_length;
+    bitmap_size_context.cbdt_prime = cbdt_prime;
+    bitmap_size_context.size = indexTablesSize;
+    bitmap_size_context.num_tables = numberOfIndexSubtables;
+    bitmap_size_context.start_glyph = 1;
+    bitmap_size_context.end_glyph = 0;
+
+    if (!out_table->indexSubtableArrayOffset.serialize_subset (c,
+							       indexSubtableArrayOffset,
+							       base,
+							       &bitmap_size_context))
+      return_trace (false);
+    if (!bitmap_size_context.size ||
+	!bitmap_size_context.num_tables ||
+	bitmap_size_context.start_glyph > bitmap_size_context.end_glyph)
+      return_trace (false);
+
+    out_table->indexTablesSize = bitmap_size_context.size;
+    out_table->numberOfIndexSubtables = bitmap_size_context.num_tables;
+    out_table->startGlyphIndex = bitmap_size_context.start_glyph;
+    out_table->endGlyphIndex = bitmap_size_context.end_glyph;
+    return_trace (true);
+  }
+
+  protected:
+  LNNOffsetTo<IndexSubtableArray>
+			indexSubtableArrayOffset;
+  HBUINT32		indexTablesSize;
+  HBUINT32		numberOfIndexSubtables;
+  HBUINT32		colorRef;
+  SBitLineMetrics	horizontal;
+  SBitLineMetrics	vertical;
+  HBGlyphID		startGlyphIndex;
+  HBGlyphID		endGlyphIndex;
+  HBUINT8		ppemX;
+  HBUINT8		ppemY;
+  HBUINT8		bitDepth;
+  HBINT8		flags;
+  public:
+  DEFINE_SIZE_STATIC (48);
+};
+
+
+/*
+ * Glyph Bitmap Data Formats.
+ */
+
+struct GlyphBitmapDataFormat17
+{
+  SmallGlyphMetrics	glyphMetrics;
+  LArrayOf<HBUINT8>	data;
+  public:
+  DEFINE_SIZE_ARRAY (9, data);
+};
+
+struct GlyphBitmapDataFormat18
+{
+  BigGlyphMetrics	glyphMetrics;
+  LArrayOf<HBUINT8>	data;
+  public:
+  DEFINE_SIZE_ARRAY (12, data);
+};
+
+struct GlyphBitmapDataFormat19
+{
+  LArrayOf<HBUINT8>	data;
+  public:
+  DEFINE_SIZE_ARRAY (4, data);
+};
+
+struct CBLC
+{
+  friend struct CBDT;
+
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_CBLC;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (version.major == 2 || version.major == 3) &&
+		  sizeTables.sanitize (c, this));
+  }
+
+  static bool
+  sink_cbdt (hb_subset_context_t *c, hb_vector_t<char>* cbdt_prime)
+  {
+    hb_blob_t *cbdt_prime_blob = hb_blob_create (cbdt_prime->arrayZ,
+						 cbdt_prime->length,
+						 HB_MEMORY_MODE_WRITABLE,
+						 cbdt_prime->arrayZ,
+						 free);
+    cbdt_prime->init ();  // Leak arrayZ to the blob.
+    bool ret = c->plan->add_table (HB_OT_TAG_CBDT, cbdt_prime_blob);
+    hb_blob_destroy (cbdt_prime_blob);
+    return ret;
+  }
+
+  bool
+  subset_size_table (hb_subset_context_t *c, const BitmapSizeTable& table,
+		     const char *cbdt /* IN */, unsigned int cbdt_length,
+		     CBLC *cblc_prime /* INOUT */, hb_vector_t<char> *cbdt_prime /* INOUT */) const
+  {
+    TRACE_SUBSET (this);
+    cblc_prime->sizeTables.len++;
+
+    auto snap = c->serializer->snapshot ();
+    auto cbdt_prime_len = cbdt_prime->length;
+
+    if (!table.subset (c, this, cbdt, cbdt_length, cbdt_prime))
+    {
+      cblc_prime->sizeTables.len--;
+      c->serializer->revert (snap);
+      cbdt_prime->shrink (cbdt_prime_len);
+      return_trace (false);
+    }
+    return_trace (true);
+  }
+
+  // Implemented in cc file as it depends on definition of CBDT.
+  HB_INTERNAL bool subset (hb_subset_context_t *c) const;
+
+  protected:
+  const BitmapSizeTable &choose_strike (hb_font_t *font) const
+  {
+    unsigned count = sizeTables.len;
+    if (unlikely (!count))
+      return Null (BitmapSizeTable);
+
+    unsigned int requested_ppem = hb_max (font->x_ppem, font->y_ppem);
+    if (!requested_ppem)
+      requested_ppem = 1<<30; /* Choose largest strike. */
+    unsigned int best_i = 0;
+    unsigned int best_ppem = hb_max (sizeTables[0].ppemX, sizeTables[0].ppemY);
+
+    for (unsigned int i = 1; i < count; i++)
+    {
+      unsigned int ppem = hb_max (sizeTables[i].ppemX, sizeTables[i].ppemY);
+      if ((requested_ppem <= ppem && ppem < best_ppem) ||
+	  (requested_ppem > best_ppem && ppem > best_ppem))
+      {
+	best_i = i;
+	best_ppem = ppem;
+      }
+    }
+
+    return sizeTables[best_i];
+  }
+
+  protected:
+  FixedVersion<>		version;
+  LArrayOf<BitmapSizeTable>	sizeTables;
+  public:
+  DEFINE_SIZE_ARRAY (8, sizeTables);
+};
+
+struct CBDT
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_CBDT;
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      cblc = hb_sanitize_context_t ().reference_table<CBLC> (face);
+      cbdt = hb_sanitize_context_t ().reference_table<CBDT> (face);
+
+      upem = hb_face_get_upem (face);
+    }
+
+    void fini ()
+    {
+      this->cblc.destroy ();
+      this->cbdt.destroy ();
+    }
+
+    bool
+    get_extents (hb_font_t *font, hb_codepoint_t glyph, hb_glyph_extents_t *extents) const
+    {
+      const void *base;
+      const BitmapSizeTable &strike = this->cblc->choose_strike (font);
+      const IndexSubtableRecord *subtable_record = strike.find_table (glyph, cblc, &base);
+      if (!subtable_record || !strike.ppemX || !strike.ppemY)
+	return false;
+
+      if (subtable_record->get_extents (extents, base))
+	return true;
+
+      unsigned int image_offset = 0, image_length = 0, image_format = 0;
+      if (!subtable_record->get_image_data (glyph, base, &image_offset, &image_length, &image_format))
+	return false;
+
+      unsigned int cbdt_len = cbdt.get_length ();
+      if (unlikely (image_offset > cbdt_len || cbdt_len - image_offset < image_length))
+	return false;
+
+      switch (image_format)
+      {
+      case 17: {
+	if (unlikely (image_length < GlyphBitmapDataFormat17::min_size))
+	  return false;
+	auto &glyphFormat17 = StructAtOffset<GlyphBitmapDataFormat17> (this->cbdt, image_offset);
+	glyphFormat17.glyphMetrics.get_extents (font, extents);
+	break;
+      }
+      case 18: {
+	if (unlikely (image_length < GlyphBitmapDataFormat18::min_size))
+	  return false;
+	auto &glyphFormat18 = StructAtOffset<GlyphBitmapDataFormat18> (this->cbdt, image_offset);
+	glyphFormat18.glyphMetrics.get_extents (font, extents);
+	break;
+      }
+      default: return false; /* TODO: Support other image formats. */
+      }
+
+      /* Convert to font units. */
+      float x_scale = upem / (float) strike.ppemX;
+      float y_scale = upem / (float) strike.ppemY;
+      extents->x_bearing = roundf (extents->x_bearing * x_scale);
+      extents->y_bearing = roundf (extents->y_bearing * y_scale);
+      extents->width = roundf (extents->width * x_scale);
+      extents->height = roundf (extents->height * y_scale);
+
+      return true;
+    }
+
+    hb_blob_t*
+    reference_png (hb_font_t *font, hb_codepoint_t glyph) const
+    {
+      const void *base;
+      const BitmapSizeTable &strike = this->cblc->choose_strike (font);
+      const IndexSubtableRecord *subtable_record = strike.find_table (glyph, cblc, &base);
+      if (!subtable_record || !strike.ppemX || !strike.ppemY)
+	return hb_blob_get_empty ();
+
+      unsigned int image_offset = 0, image_length = 0, image_format = 0;
+      if (!subtable_record->get_image_data (glyph, base, &image_offset, &image_length, &image_format))
+	return hb_blob_get_empty ();
+
+      unsigned int cbdt_len = cbdt.get_length ();
+      if (unlikely (image_offset > cbdt_len || cbdt_len - image_offset < image_length))
+	return hb_blob_get_empty ();
+
+      switch (image_format)
+      {
+      case 17:
+      {
+	if (unlikely (image_length < GlyphBitmapDataFormat17::min_size))
+	  return hb_blob_get_empty ();
+	auto &glyphFormat17 = StructAtOffset<GlyphBitmapDataFormat17> (this->cbdt, image_offset);
+	return hb_blob_create_sub_blob (cbdt.get_blob (),
+					image_offset + GlyphBitmapDataFormat17::min_size,
+					glyphFormat17.data.len);
+      }
+      case 18:
+      {
+	if (unlikely (image_length < GlyphBitmapDataFormat18::min_size))
+	  return hb_blob_get_empty ();
+	auto &glyphFormat18 = StructAtOffset<GlyphBitmapDataFormat18> (this->cbdt, image_offset);
+	return hb_blob_create_sub_blob (cbdt.get_blob (),
+					image_offset + GlyphBitmapDataFormat18::min_size,
+					glyphFormat18.data.len);
+      }
+      case 19:
+      {
+	if (unlikely (image_length < GlyphBitmapDataFormat19::min_size))
+	  return hb_blob_get_empty ();
+	auto &glyphFormat19 = StructAtOffset<GlyphBitmapDataFormat19> (this->cbdt, image_offset);
+	return hb_blob_create_sub_blob (cbdt.get_blob (),
+					image_offset + GlyphBitmapDataFormat19::min_size,
+					glyphFormat19.data.len);
+      }
+      default: return hb_blob_get_empty (); /* TODO: Support other image formats. */
+      }
+    }
+
+    bool has_data () const { return cbdt.get_length (); }
+
+    private:
+    hb_blob_ptr_t<CBLC> cblc;
+    hb_blob_ptr_t<CBDT> cbdt;
+
+    unsigned int upem;
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (version.major == 2 || version.major == 3));
+  }
+
+  protected:
+  FixedVersion<>		version;
+  UnsizedArrayOf<HBUINT8>	dataZ;
+  public:
+  DEFINE_SIZE_ARRAY (4, dataZ);
+};
+
+inline bool
+CBLC::subset (hb_subset_context_t *c) const
+{
+  TRACE_SUBSET (this);
+
+  auto *cblc_prime = c->serializer->start_embed<CBLC> ();
+
+  // Use a vector as a secondary buffer as the tables need to be built in parallel.
+  hb_vector_t<char> cbdt_prime;
+
+  if (unlikely (!cblc_prime)) return_trace (false);
+  if (unlikely (!c->serializer->extend_min (cblc_prime))) return_trace (false);
+  cblc_prime->version = version;
+
+  hb_blob_t* cbdt_blob = hb_sanitize_context_t ().reference_table<CBDT> (c->plan->source);
+  unsigned int cbdt_length;
+  CBDT* cbdt = (CBDT *) hb_blob_get_data (cbdt_blob, &cbdt_length);
+  if (unlikely (cbdt_length < CBDT::min_size))
+  {
+    hb_blob_destroy (cbdt_blob);
+    return_trace (false);
+  }
+  _copy_data_to_cbdt (&cbdt_prime, cbdt, CBDT::min_size);
+
+  for (const BitmapSizeTable& table : + sizeTables.iter ())
+    subset_size_table (c, table, (const char *) cbdt, cbdt_length, cblc_prime, &cbdt_prime);
+
+  hb_blob_destroy (cbdt_blob);
+
+  return_trace (CBLC::sink_cbdt (c, &cbdt_prime));
+}
+
+struct CBDT_accelerator_t : CBDT::accelerator_t {};
+
+} /* namespace OT */
+
+#endif /* HB_OT_COLOR_CBDT_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color-colr-table.hh b/thirdparty/harfbuzz/src/hb-ot-color-colr-table.hh
new file mode 100644
index 0000000000..92a49bb4f4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color-colr-table.hh
@@ -0,0 +1,278 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ * Copyright © 2020  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Calder Kitagawa
+ */
+
+#ifndef HB_OT_COLOR_COLR_TABLE_HH
+#define HB_OT_COLOR_COLR_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * COLR -- Color
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/colr
+ */
+#define HB_OT_TAG_COLR HB_TAG('C','O','L','R')
+
+
+namespace OT {
+
+
+struct LayerRecord
+{
+  operator hb_ot_color_layer_t () const { return {glyphId, colorIdx}; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBGlyphID	glyphId;	/* Glyph ID of layer glyph */
+  Index		colorIdx;	/* Index value to use with a
+				 * selected color palette.
+				 * An index value of 0xFFFF
+				 * is a special case indicating
+				 * that the text foreground
+				 * color (defined by a
+				 * higher-level client) should
+				 * be used and shall not be
+				 * treated as actual index
+				 * into CPAL ColorRecord array. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct BaseGlyphRecord
+{
+  int cmp (hb_codepoint_t g) const
+  { return g < glyphId ? -1 : g > glyphId ? 1 : 0; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  public:
+  HBGlyphID	glyphId;	/* Glyph ID of reference glyph */
+  HBUINT16	firstLayerIdx;	/* Index (from beginning of
+				 * the Layer Records) to the
+				 * layer record. There will be
+				 * numLayers consecutive entries
+				 * for this base glyph. */
+  HBUINT16	numLayers;	/* Number of color layers
+				 * associated with this glyph */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct COLR
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_COLR;
+
+  bool has_data () const { return numBaseGlyphs; }
+
+  unsigned int get_glyph_layers (hb_codepoint_t       glyph,
+				 unsigned int         start_offset,
+				 unsigned int        *count, /* IN/OUT.  May be NULL. */
+				 hb_ot_color_layer_t *layers /* OUT.     May be NULL. */) const
+  {
+    const BaseGlyphRecord &record = (this+baseGlyphsZ).bsearch (numBaseGlyphs, glyph);
+
+    hb_array_t<const LayerRecord> all_layers = (this+layersZ).as_array (numLayers);
+    hb_array_t<const LayerRecord> glyph_layers = all_layers.sub_array (record.firstLayerIdx,
+								       record.numLayers);
+    if (count)
+    {
+      + glyph_layers.sub_array (start_offset, count)
+      | hb_sink (hb_array (layers, *count))
+      ;
+    }
+    return glyph_layers.length;
+  }
+
+  struct accelerator_t
+  {
+    accelerator_t () {}
+    ~accelerator_t () { fini (); }
+
+    void init (hb_face_t *face)
+    { colr = hb_sanitize_context_t ().reference_table<COLR> (face); }
+
+    void fini () { this->colr.destroy (); }
+
+    bool is_valid () { return colr.get_blob ()->length; }
+
+    void closure_glyphs (hb_codepoint_t glyph,
+			 hb_set_t *related_ids /* OUT */) const
+    { colr->closure_glyphs (glyph, related_ids); }
+
+    private:
+    hb_blob_ptr_t<COLR> colr;
+  };
+
+  void closure_glyphs (hb_codepoint_t glyph,
+		       hb_set_t *related_ids /* OUT */) const
+  {
+    const BaseGlyphRecord *record = get_base_glyph_record (glyph);
+    if (!record) return;
+
+    auto glyph_layers = (this+layersZ).as_array (numLayers).sub_array (record->firstLayerIdx,
+								       record->numLayers);
+    if (!glyph_layers.length) return;
+    related_ids->add_array (&glyph_layers[0].glyphId, glyph_layers.length, LayerRecord::min_size);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (this+baseGlyphsZ).sanitize (c, numBaseGlyphs) &&
+			  (this+layersZ).sanitize (c, numLayers)));
+  }
+
+  template<typename BaseIterator, typename LayerIterator,
+	   hb_requires (hb_is_iterator (BaseIterator)),
+	   hb_requires (hb_is_iterator (LayerIterator))>
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned version,
+		  BaseIterator base_it,
+		  LayerIterator layer_it)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (base_it.len () != layer_it.len ()))
+      return_trace (false);
+
+    if (unlikely (!c->extend_min (this))) return_trace (false);
+    this->version = version;
+    numLayers = 0;
+    numBaseGlyphs = base_it.len ();
+    baseGlyphsZ = COLR::min_size;
+    layersZ = COLR::min_size + numBaseGlyphs * BaseGlyphRecord::min_size;
+
+    for (const hb_item_type<BaseIterator> _ : + base_it.iter ())
+    {
+      auto* record = c->embed (_);
+      if (unlikely (!record)) return_trace (false);
+      record->firstLayerIdx = numLayers;
+      numLayers += record->numLayers;
+    }
+
+    for (const hb_item_type<LayerIterator>& _ : + layer_it.iter ())
+      _.as_array ().copy (c);
+
+    return_trace (true);
+  }
+
+  const BaseGlyphRecord* get_base_glyph_record (hb_codepoint_t gid) const
+  {
+    if ((unsigned int) gid == 0) // Ignore notdef.
+      return nullptr;
+    const BaseGlyphRecord* record = &(this+baseGlyphsZ).bsearch (numBaseGlyphs, (unsigned int) gid);
+    if ((record && (hb_codepoint_t) record->glyphId != gid))
+      record = nullptr;
+    return record;
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    const hb_map_t &reverse_glyph_map = *c->plan->reverse_glyph_map;
+
+    auto base_it =
+    + hb_range (c->plan->num_output_glyphs ())
+    | hb_map_retains_sorting ([&](hb_codepoint_t new_gid)
+			      {
+				hb_codepoint_t old_gid = reverse_glyph_map.get (new_gid);
+
+				const BaseGlyphRecord* old_record = get_base_glyph_record (old_gid);
+				if (unlikely (!old_record))
+				  return hb_pair_t<bool, BaseGlyphRecord> (false, Null (BaseGlyphRecord));
+
+				BaseGlyphRecord new_record;
+				new_record.glyphId = new_gid;
+				new_record.numLayers = old_record->numLayers;
+				return hb_pair_t<bool, BaseGlyphRecord> (true, new_record);
+			      })
+    | hb_filter (hb_first)
+    | hb_map_retains_sorting (hb_second)
+    ;
+
+    auto layer_it =
+    + hb_range (c->plan->num_output_glyphs ())
+    | hb_map (reverse_glyph_map)
+    | hb_map_retains_sorting ([&](hb_codepoint_t old_gid)
+			      {
+				const BaseGlyphRecord* old_record = get_base_glyph_record (old_gid);
+				hb_vector_t<LayerRecord> out_layers;
+
+				if (unlikely (!old_record ||
+					      old_record->firstLayerIdx >= numLayers ||
+					      old_record->firstLayerIdx + old_record->numLayers > numLayers))
+				  return hb_pair_t<bool, hb_vector_t<LayerRecord>> (false, out_layers);
+
+				auto layers = (this+layersZ).as_array (numLayers).sub_array (old_record->firstLayerIdx,
+											     old_record->numLayers);
+				out_layers.resize (layers.length);
+				for (unsigned int i = 0; i < layers.length; i++) {
+				  out_layers[i] = layers[i];
+				  hb_codepoint_t new_gid = 0;
+				  if (unlikely (!c->plan->new_gid_for_old_gid (out_layers[i].glyphId, &new_gid)))
+				    return hb_pair_t<bool, hb_vector_t<LayerRecord>> (false, out_layers);
+				  out_layers[i].glyphId = new_gid;
+				}
+
+				return hb_pair_t<bool, hb_vector_t<LayerRecord>> (true, out_layers);
+			      })
+    | hb_filter (hb_first)
+    | hb_map_retains_sorting (hb_second)
+    ;
+
+    if (unlikely (!base_it || !layer_it || base_it.len () != layer_it.len ()))
+      return_trace (false);
+
+    COLR *colr_prime = c->serializer->start_embed<COLR> ();
+    return_trace (colr_prime->serialize (c->serializer, version, base_it, layer_it));
+  }
+
+  protected:
+  HBUINT16	version;	/* Table version number (starts at 0). */
+  HBUINT16	numBaseGlyphs;	/* Number of Base Glyph Records. */
+  LNNOffsetTo<SortedUnsizedArrayOf<BaseGlyphRecord>>
+		baseGlyphsZ;	/* Offset to Base Glyph records. */
+  LNNOffsetTo<UnsizedArrayOf<LayerRecord>>
+		layersZ;	/* Offset to Layer Records. */
+  HBUINT16	numLayers;	/* Number of Layer Records. */
+  public:
+  DEFINE_SIZE_STATIC (14);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_COLOR_COLR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color-cpal-table.hh b/thirdparty/harfbuzz/src/hb-ot-color-cpal-table.hh
new file mode 100644
index 0000000000..fa7d3207be
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color-cpal-table.hh
@@ -0,0 +1,190 @@
+/*
+ * Copyright © 2016  Google, Inc.
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Sascha Brawer
+ */
+
+#ifndef HB_OT_COLOR_CPAL_TABLE_HH
+#define HB_OT_COLOR_CPAL_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-color.h"
+#include "hb-ot-name.h"
+
+
+/*
+ * CPAL -- Color Palette
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/cpal
+ */
+#define HB_OT_TAG_CPAL HB_TAG('C','P','A','L')
+
+
+namespace OT {
+
+
+struct CPALV1Tail
+{
+  friend struct CPAL;
+
+  private:
+  hb_ot_color_palette_flags_t get_palette_flags (const void *base,
+						 unsigned int palette_index,
+						 unsigned int palette_count) const
+  {
+    if (!paletteFlagsZ) return HB_OT_COLOR_PALETTE_FLAG_DEFAULT;
+    return (hb_ot_color_palette_flags_t) (uint32_t)
+	   (base+paletteFlagsZ).as_array (palette_count)[palette_index];
+  }
+
+  hb_ot_name_id_t get_palette_name_id (const void *base,
+				       unsigned int palette_index,
+				       unsigned int palette_count) const
+  {
+    if (!paletteLabelsZ) return HB_OT_NAME_ID_INVALID;
+    return (base+paletteLabelsZ).as_array (palette_count)[palette_index];
+  }
+
+  hb_ot_name_id_t get_color_name_id (const void *base,
+				     unsigned int color_index,
+				     unsigned int color_count) const
+  {
+    if (!colorLabelsZ) return HB_OT_NAME_ID_INVALID;
+    return (base+colorLabelsZ).as_array (color_count)[color_index];
+  }
+
+  public:
+  bool sanitize (hb_sanitize_context_t *c,
+		 const void *base,
+		 unsigned int palette_count,
+		 unsigned int color_count) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  (!paletteFlagsZ  || (base+paletteFlagsZ).sanitize (c, palette_count)) &&
+		  (!paletteLabelsZ || (base+paletteLabelsZ).sanitize (c, palette_count)) &&
+		  (!colorLabelsZ   || (base+colorLabelsZ).sanitize (c, color_count)));
+  }
+
+  protected:
+  LNNOffsetTo<UnsizedArrayOf<HBUINT32>>
+		paletteFlagsZ;		/* Offset from the beginning of CPAL table to
+					 * the Palette Type Array. Set to 0 if no array
+					 * is provided. */
+  LNNOffsetTo<UnsizedArrayOf<NameID>>
+		paletteLabelsZ;		/* Offset from the beginning of CPAL table to
+					 * the palette labels array. Set to 0 if no
+					 * array is provided. */
+  LNNOffsetTo<UnsizedArrayOf<NameID>>
+		colorLabelsZ;		/* Offset from the beginning of CPAL table to
+					 * the color labels array. Set to 0
+					 * if no array is provided. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+typedef HBUINT32 BGRAColor;
+
+struct CPAL
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_CPAL;
+
+  bool has_data () const { return numPalettes; }
+
+  unsigned int get_size () const
+  { return min_size + numPalettes * sizeof (colorRecordIndicesZ[0]); }
+
+  unsigned int get_palette_count () const { return numPalettes; }
+  unsigned int   get_color_count () const { return numColors; }
+
+  hb_ot_color_palette_flags_t get_palette_flags (unsigned int palette_index) const
+  { return v1 ().get_palette_flags (this, palette_index, numPalettes); }
+
+  hb_ot_name_id_t get_palette_name_id (unsigned int palette_index) const
+  { return v1 ().get_palette_name_id (this, palette_index, numPalettes); }
+
+  hb_ot_name_id_t get_color_name_id (unsigned int color_index) const
+  { return v1 ().get_color_name_id (this, color_index, numColors); }
+
+  unsigned int get_palette_colors (unsigned int  palette_index,
+				   unsigned int  start_offset,
+				   unsigned int *color_count, /* IN/OUT.  May be NULL. */
+				   hb_color_t   *colors       /* OUT.     May be NULL. */) const
+  {
+    if (unlikely (palette_index >= numPalettes))
+    {
+      if (color_count) *color_count = 0;
+      return 0;
+    }
+    unsigned int start_index = colorRecordIndicesZ[palette_index];
+    hb_array_t<const BGRAColor> all_colors ((this+colorRecordsZ).arrayZ, numColorRecords);
+    hb_array_t<const BGRAColor> palette_colors = all_colors.sub_array (start_index,
+								       numColors);
+    if (color_count)
+    {
+      + palette_colors.sub_array (start_offset, color_count)
+      | hb_sink (hb_array (colors, *color_count))
+      ;
+    }
+    return numColors;
+  }
+
+  private:
+  const CPALV1Tail& v1 () const
+  {
+    if (version == 0) return Null (CPALV1Tail);
+    return StructAfter<CPALV1Tail> (*this);
+  }
+
+  public:
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  (this+colorRecordsZ).sanitize (c, numColorRecords) &&
+		  colorRecordIndicesZ.sanitize (c, numPalettes) &&
+		  (version == 0 || v1 ().sanitize (c, this, numPalettes, numColors)));
+  }
+
+  protected:
+  HBUINT16	version;		/* Table version number */
+  /* Version 0 */
+  HBUINT16	numColors;		/* Number of colors in each palette. */
+  HBUINT16	numPalettes;		/* Number of palettes in the table. */
+  HBUINT16	numColorRecords;	/* Total number of color records, combined for
+					 * all palettes. */
+  LNNOffsetTo<UnsizedArrayOf<BGRAColor>>
+		colorRecordsZ;		/* Offset from the beginning of CPAL table to
+					 * the first ColorRecord. */
+  UnsizedArrayOf<HBUINT16>
+		colorRecordIndicesZ;	/* Index of each palette’s first color record in
+					 * the combined color record array. */
+/*CPALV1Tail	v1;*/
+  public:
+  DEFINE_SIZE_ARRAY (12, colorRecordIndicesZ);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_COLOR_CPAL_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color-sbix-table.hh b/thirdparty/harfbuzz/src/hb-ot-color-sbix-table.hh
new file mode 100644
index 0000000000..09da11597d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color-sbix-table.hh
@@ -0,0 +1,414 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ * Copyright © 2020  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Calder Kitagawa
+ */
+
+#ifndef HB_OT_COLOR_SBIX_TABLE_HH
+#define HB_OT_COLOR_SBIX_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout-common.hh"
+
+/*
+ * sbix -- Standard Bitmap Graphics
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/sbix
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6sbix.html
+ */
+#define HB_OT_TAG_sbix HB_TAG('s','b','i','x')
+
+
+namespace OT {
+
+
+struct SBIXGlyph
+{
+  SBIXGlyph* copy (hb_serialize_context_t *c, unsigned int data_length) const
+  {
+    TRACE_SERIALIZE (this);
+    SBIXGlyph* new_glyph = c->start_embed<SBIXGlyph> ();
+    if (unlikely (!new_glyph)) return_trace (nullptr);
+    if (unlikely (!c->extend_min (new_glyph))) return_trace (nullptr);
+
+    new_glyph->xOffset = xOffset;
+    new_glyph->yOffset = yOffset;
+    new_glyph->graphicType = graphicType;
+    data.copy (c, data_length);
+    return_trace (new_glyph);
+  }
+
+  HBINT16	xOffset;	/* The horizontal (x-axis) offset from the left
+				 * edge of the graphic to the glyph’s origin.
+				 * That is, the x-coordinate of the point on the
+				 * baseline at the left edge of the glyph. */
+  HBINT16	yOffset;	/* The vertical (y-axis) offset from the bottom
+				 * edge of the graphic to the glyph’s origin.
+				 * That is, the y-coordinate of the point on the
+				 * baseline at the left edge of the glyph. */
+  Tag		graphicType;	/* Indicates the format of the embedded graphic
+				 * data: one of 'jpg ', 'png ' or 'tiff', or the
+				 * special format 'dupe'. */
+  UnsizedArrayOf<HBUINT8>
+		data;		/* The actual embedded graphic data. The total
+				 * length is inferred from sequential entries in
+				 * the glyphDataOffsets array and the fixed size
+				 * (8 bytes) of the preceding fields. */
+  public:
+  DEFINE_SIZE_ARRAY (8, data);
+};
+
+struct SBIXStrike
+{
+  static unsigned int get_size (unsigned num_glyphs)
+  { return min_size + num_glyphs * HBUINT32::static_size; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  imageOffsetsZ.sanitize_shallow (c, c->get_num_glyphs () + 1));
+  }
+
+  hb_blob_t *get_glyph_blob (unsigned int  glyph_id,
+			     hb_blob_t    *sbix_blob,
+			     hb_tag_t      file_type,
+			     int          *x_offset,
+			     int          *y_offset,
+			     unsigned int  num_glyphs,
+			     unsigned int *strike_ppem) const
+  {
+    if (unlikely (!ppem)) return hb_blob_get_empty (); /* To get Null() object out of the way. */
+
+    unsigned int retry_count = 8;
+    unsigned int sbix_len = sbix_blob->length;
+    unsigned int strike_offset = (const char *) this - (const char *) sbix_blob->data;
+    assert (strike_offset < sbix_len);
+
+  retry:
+    if (unlikely (glyph_id >= num_glyphs ||
+		  imageOffsetsZ[glyph_id + 1] <= imageOffsetsZ[glyph_id] ||
+		  imageOffsetsZ[glyph_id + 1] - imageOffsetsZ[glyph_id] <= SBIXGlyph::min_size ||
+		  (unsigned int) imageOffsetsZ[glyph_id + 1] > sbix_len - strike_offset))
+      return hb_blob_get_empty ();
+
+    unsigned int glyph_offset = strike_offset + (unsigned int) imageOffsetsZ[glyph_id] + SBIXGlyph::min_size;
+    unsigned int glyph_length = imageOffsetsZ[glyph_id + 1] - imageOffsetsZ[glyph_id] - SBIXGlyph::min_size;
+
+    const SBIXGlyph *glyph = &(this+imageOffsetsZ[glyph_id]);
+
+    if (glyph->graphicType == HB_TAG ('d','u','p','e'))
+    {
+      if (glyph_length >= 2)
+      {
+	glyph_id = *((HBUINT16 *) &glyph->data);
+	if (retry_count--)
+	  goto retry;
+      }
+      return hb_blob_get_empty ();
+    }
+
+    if (unlikely (file_type != glyph->graphicType))
+      return hb_blob_get_empty ();
+
+    if (strike_ppem) *strike_ppem = ppem;
+    if (x_offset) *x_offset = glyph->xOffset;
+    if (y_offset) *y_offset = glyph->yOffset;
+    return hb_blob_create_sub_blob (sbix_blob, glyph_offset, glyph_length);
+  }
+
+  bool subset (hb_subset_context_t *c, unsigned int available_len) const
+  {
+    TRACE_SUBSET (this);
+    unsigned int num_output_glyphs = c->plan->num_output_glyphs ();
+
+    auto* out = c->serializer->start_embed<SBIXStrike> ();
+    if (unlikely (!out)) return_trace (false);
+    auto snap = c->serializer->snapshot ();
+    if (unlikely (!c->serializer->extend (*out, num_output_glyphs + 1))) return_trace (false);
+    out->ppem = ppem;
+    out->resolution = resolution;
+    HBUINT32 head;
+    head = get_size (num_output_glyphs + 1);
+
+    bool has_glyphs = false;
+    for (unsigned new_gid = 0; new_gid < num_output_glyphs; new_gid++)
+    {
+      hb_codepoint_t old_gid;
+      if (!c->plan->old_gid_for_new_gid (new_gid, &old_gid) ||
+	  unlikely (imageOffsetsZ[old_gid].is_null () ||
+		    imageOffsetsZ[old_gid + 1].is_null () ||
+		    imageOffsetsZ[old_gid + 1] <= imageOffsetsZ[old_gid] ||
+		    imageOffsetsZ[old_gid + 1] - imageOffsetsZ[old_gid] <= SBIXGlyph::min_size) ||
+		    (unsigned int) imageOffsetsZ[old_gid + 1] > available_len)
+      {
+	out->imageOffsetsZ[new_gid] = head;
+	continue;
+      }
+      has_glyphs = true;
+      unsigned int delta = imageOffsetsZ[old_gid + 1] - imageOffsetsZ[old_gid];
+      unsigned int glyph_data_length = delta - SBIXGlyph::min_size;
+      if (!(this+imageOffsetsZ[old_gid]).copy (c->serializer, glyph_data_length))
+	return_trace (false);
+      out->imageOffsetsZ[new_gid] = head;
+      head += delta;
+    }
+    if (has_glyphs)
+      out->imageOffsetsZ[num_output_glyphs] = head;
+    else
+      c->serializer->revert (snap);
+    return_trace (has_glyphs);
+  }
+
+  public:
+  HBUINT16	ppem;		/* The PPEM size for which this strike was designed. */
+  HBUINT16	resolution;	/* The device pixel density (in PPI) for which this
+				 * strike was designed. (E.g., 96 PPI, 192 PPI.) */
+  protected:
+  UnsizedArrayOf<LOffsetTo<SBIXGlyph>>
+		imageOffsetsZ;	/* Offset from the beginning of the strike data header
+				 * to bitmap data for an individual glyph ID. */
+  public:
+  DEFINE_SIZE_ARRAY (4, imageOffsetsZ);
+};
+
+struct sbix
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_sbix;
+
+  bool has_data () const { return version; }
+
+  const SBIXStrike &get_strike (unsigned int i) const { return this+strikes[i]; }
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      table = hb_sanitize_context_t ().reference_table<sbix> (face);
+      num_glyphs = face->get_num_glyphs ();
+    }
+    void fini () { table.destroy (); }
+
+    bool has_data () const { return table->has_data (); }
+
+    bool get_extents (hb_font_t          *font,
+		      hb_codepoint_t      glyph,
+		      hb_glyph_extents_t *extents) const
+    {
+      /* We only support PNG right now, and following function checks type. */
+      return get_png_extents (font, glyph, extents);
+    }
+
+    hb_blob_t *reference_png (hb_font_t      *font,
+			      hb_codepoint_t  glyph_id,
+			      int            *x_offset,
+			      int            *y_offset,
+			      unsigned int   *available_ppem) const
+    {
+      return choose_strike (font).get_glyph_blob (glyph_id, table.get_blob (),
+						  HB_TAG ('p','n','g',' '),
+						  x_offset, y_offset,
+						  num_glyphs, available_ppem);
+    }
+
+    private:
+
+    const SBIXStrike &choose_strike (hb_font_t *font) const
+    {
+      unsigned count = table->strikes.len;
+      if (unlikely (!count))
+	return Null (SBIXStrike);
+
+      unsigned int requested_ppem = hb_max (font->x_ppem, font->y_ppem);
+      if (!requested_ppem)
+	requested_ppem = 1<<30; /* Choose largest strike. */
+      /* TODO Add DPI sensitivity as well? */
+      unsigned int best_i = 0;
+      unsigned int best_ppem = table->get_strike (0).ppem;
+
+      for (unsigned int i = 1; i < count; i++)
+      {
+	unsigned int ppem = (table->get_strike (i)).ppem;
+	if ((requested_ppem <= ppem && ppem < best_ppem) ||
+	    (requested_ppem > best_ppem && ppem > best_ppem))
+	{
+	  best_i = i;
+	  best_ppem = ppem;
+	}
+      }
+
+      return table->get_strike (best_i);
+    }
+
+    struct PNGHeader
+    {
+      HBUINT8	signature[8];
+      struct
+      {
+	struct
+	{
+	  HBUINT32	length;
+	  Tag		type;
+	}		header;
+	HBUINT32	width;
+	HBUINT32	height;
+	HBUINT8		bitDepth;
+	HBUINT8		colorType;
+	HBUINT8		compressionMethod;
+	HBUINT8		filterMethod;
+	HBUINT8		interlaceMethod;
+      } IHDR;
+
+      public:
+      DEFINE_SIZE_STATIC (29);
+    };
+
+    bool get_png_extents (hb_font_t          *font,
+			  hb_codepoint_t      glyph,
+			  hb_glyph_extents_t *extents) const
+    {
+      /* Following code is safe to call even without data.
+       * But faster to short-circuit. */
+      if (!has_data ())
+	return false;
+
+      int x_offset = 0, y_offset = 0;
+      unsigned int strike_ppem = 0;
+      hb_blob_t *blob = reference_png (font, glyph, &x_offset, &y_offset, &strike_ppem);
+
+      const PNGHeader &png = *blob->as<PNGHeader>();
+
+      extents->x_bearing = x_offset;
+      extents->y_bearing = png.IHDR.height + y_offset;
+      extents->width     = png.IHDR.width;
+      extents->height    = -1 * png.IHDR.height;
+
+      /* Convert to font units. */
+      if (strike_ppem)
+      {
+	float scale = font->face->get_upem () / (float) strike_ppem;
+	extents->x_bearing = font->em_scalef_x (extents->x_bearing * scale);
+	extents->y_bearing = font->em_scalef_y (extents->y_bearing * scale);
+	extents->width = font->em_scalef_x (extents->width * scale);
+	extents->height = font->em_scalef_y (extents->height * scale);
+      }
+      else
+      {
+	extents->x_bearing = font->em_scale_x (extents->x_bearing);
+	extents->y_bearing = font->em_scale_y (extents->y_bearing);
+	extents->width = font->em_scale_x (extents->width);
+	extents->height = font->em_scale_y (extents->height);
+      }
+
+      hb_blob_destroy (blob);
+
+      return strike_ppem;
+    }
+
+    private:
+    hb_blob_ptr_t<sbix> table;
+
+    unsigned int num_glyphs;
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version >= 1 &&
+			  strikes.sanitize (c, this)));
+  }
+
+  bool
+  add_strike (hb_subset_context_t *c, unsigned i) const
+  {
+    if (strikes[i].is_null () || c->source_blob->length < (unsigned) strikes[i])
+      return false;
+
+    return (this+strikes[i]).subset (c, c->source_blob->length - (unsigned) strikes[i]);
+  }
+
+  bool serialize_strike_offsets (hb_subset_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+
+    auto *out = c->serializer->start_embed<LOffsetLArrayOf<SBIXStrike>> ();
+    if (unlikely (!out)) return_trace (false);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    hb_vector_t<LOffsetTo<SBIXStrike>*> new_strikes;
+    hb_vector_t<hb_serialize_context_t::objidx_t> objidxs;
+    for (int i = strikes.len - 1; i >= 0; --i)
+    {
+      auto* o = out->serialize_append (c->serializer);
+      if (unlikely (!o)) return_trace (false);
+      *o = 0;
+      auto snap = c->serializer->snapshot ();
+      c->serializer->push ();
+      bool ret = add_strike (c, i);
+      if (!ret)
+      {
+	c->serializer->pop_discard ();
+	out->pop ();
+	c->serializer->revert (snap);
+      }
+      else
+      {
+	objidxs.push (c->serializer->pop_pack ());
+	new_strikes.push (o);
+      }
+    }
+    for (unsigned int i = 0; i < new_strikes.length; ++i)
+      c->serializer->add_link (*new_strikes[i], objidxs[new_strikes.length - 1 - i]);
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t* c) const
+  {
+    TRACE_SUBSET (this);
+
+    sbix *sbix_prime = c->serializer->start_embed<sbix> ();
+    if (unlikely (!sbix_prime)) return_trace (false);
+    if (unlikely (!c->serializer->embed (this->version))) return_trace (false);
+    if (unlikely (!c->serializer->embed (this->flags))) return_trace (false);
+
+    return_trace (serialize_strike_offsets (c));
+  }
+
+  protected:
+  HBUINT16	version;	/* Table version number — set to 1 */
+  HBUINT16	flags;		/* Bit 0: Set to 1. Bit 1: Draw outlines.
+				 * Bits 2 to 15: reserved (set to 0). */
+  LOffsetLArrayOf<SBIXStrike>
+		strikes;	/* Offsets from the beginning of the 'sbix'
+				 * table to data for each individual bitmap strike. */
+  public:
+  DEFINE_SIZE_ARRAY (8, strikes);
+};
+
+struct sbix_accelerator_t : sbix::accelerator_t {};
+
+} /* namespace OT */
+
+#endif /* HB_OT_COLOR_SBIX_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color-svg-table.hh b/thirdparty/harfbuzz/src/hb-ot-color-svg-table.hh
new file mode 100644
index 0000000000..1cc40ae53f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color-svg-table.hh
@@ -0,0 +1,124 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_COLOR_SVG_TABLE_HH
+#define HB_OT_COLOR_SVG_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * SVG -- SVG (Scalable Vector Graphics)
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/svg
+ */
+
+#define HB_OT_TAG_SVG HB_TAG('S','V','G',' ')
+
+
+namespace OT {
+
+
+struct SVGDocumentIndexEntry
+{
+  int cmp (hb_codepoint_t g) const
+  { return g < startGlyphID ? -1 : g > endGlyphID ? 1 : 0; }
+
+  hb_blob_t *reference_blob (hb_blob_t *svg_blob, unsigned int index_offset) const
+  {
+    return hb_blob_create_sub_blob (svg_blob,
+				    index_offset + (unsigned int) svgDoc,
+				    svgDocLength);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  svgDoc.sanitize (c, base, svgDocLength));
+  }
+
+  protected:
+  HBUINT16	startGlyphID;	/* The first glyph ID in the range described by
+				 * this index entry. */
+  HBUINT16	endGlyphID;	/* The last glyph ID in the range described by
+				 * this index entry. Must be >= startGlyphID. */
+  LNNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		svgDoc;		/* Offset from the beginning of the SVG Document Index
+				 * to an SVG document. Must be non-zero. */
+  HBUINT32	svgDocLength;	/* Length of the SVG document.
+				 * Must be non-zero. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct SVG
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_SVG;
+
+  bool has_data () const { return svgDocEntries; }
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    { table = hb_sanitize_context_t ().reference_table<SVG> (face); }
+    void fini () { table.destroy (); }
+
+    hb_blob_t *reference_blob_for_glyph (hb_codepoint_t glyph_id) const
+    {
+      return table->get_glyph_entry (glyph_id).reference_blob (table.get_blob (),
+							       table->svgDocEntries);
+    }
+
+    bool has_data () const { return table->has_data (); }
+
+    private:
+    hb_blob_ptr_t<SVG> table;
+  };
+
+  const SVGDocumentIndexEntry &get_glyph_entry (hb_codepoint_t glyph_id) const
+  { return (this+svgDocEntries).bsearch (glyph_id); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (this+svgDocEntries).sanitize_shallow (c)));
+  }
+
+  protected:
+  HBUINT16	version;	/* Table version (starting at 0). */
+  LOffsetTo<SortedArrayOf<SVGDocumentIndexEntry>>
+		svgDocEntries;	/* Offset (relative to the start of the SVG table) to the
+				 * SVG Documents Index. Must be non-zero. */
+				/* Array of SVG Document Index Entries. */
+  HBUINT32	reserved;	/* Set to 0. */
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+struct SVG_accelerator_t : SVG::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_COLOR_SVG_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-color.cc b/thirdparty/harfbuzz/src/hb-ot-color.cc
new file mode 100644
index 0000000000..0e7203a88b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color.cc
@@ -0,0 +1,321 @@
+/*
+ * Copyright © 2016  Google, Inc.
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Sascha Brawer, Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_COLOR
+
+#include "hb-ot.h"
+
+#include "hb-ot-color-cbdt-table.hh"
+#include "hb-ot-color-colr-table.hh"
+#include "hb-ot-color-cpal-table.hh"
+#include "hb-ot-color-sbix-table.hh"
+#include "hb-ot-color-svg-table.hh"
+
+#include <stdlib.h>
+#include <string.h>
+
+
+/**
+ * SECTION:hb-ot-color
+ * @title: hb-ot-color
+ * @short_description: OpenType Color Fonts
+ * @include: hb-ot.h
+ *
+ * Functions for fetching color-font information from OpenType font faces.
+ *
+ * HarfBuzz supports `COLR`/`CPAL`, `sbix`, `CBDT`, and `SVG` color fonts.
+ **/
+
+
+/*
+ * CPAL
+ */
+
+
+/**
+ * hb_ot_color_has_palettes:
+ * @face: #hb_face_t to work upon
+ *
+ * Tests whether a face includes a `CPAL` color-palette table.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ * Since: 2.1.0
+ */
+hb_bool_t
+hb_ot_color_has_palettes (hb_face_t *face)
+{
+  return face->table.CPAL->has_data ();
+}
+
+/**
+ * hb_ot_color_palette_get_count:
+ * @face: #hb_face_t to work upon
+ *
+ * Fetches the number of color palettes in a face.
+ *
+ * Return value: the number of palettes found
+ *
+ * Since: 2.1.0
+ */
+unsigned int
+hb_ot_color_palette_get_count (hb_face_t *face)
+{
+  return face->table.CPAL->get_palette_count ();
+}
+
+/**
+ * hb_ot_color_palette_get_name_id:
+ * @face: #hb_face_t to work upon
+ * @palette_index: The index of the color palette 
+ *
+ * Fetches the `name` table Name ID that provides display names for
+ * a `CPAL` color palette. 
+ *
+ * Palette display names can be generic (e.g., "Default") or provide
+ * specific, themed names (e.g., "Spring", "Summer", "Fall", and "Winter").
+ *
+ * Return value: the Named ID found for the palette. 
+ * If the requested palette has no name the result is #HB_OT_NAME_ID_INVALID.
+ *
+ * Since: 2.1.0
+ */
+hb_ot_name_id_t
+hb_ot_color_palette_get_name_id (hb_face_t *face,
+				 unsigned int palette_index)
+{
+  return face->table.CPAL->get_palette_name_id (palette_index);
+}
+
+/**
+ * hb_ot_color_palette_color_get_name_id:
+ * @face: #hb_face_t to work upon
+ * @color_index: The index of the color
+ *
+ * Fetches the `name` table Name ID that provides display names for
+ * the specificed color in a face's `CPAL` color palette. 
+ *
+ * Display names can be generic (e.g., "Background") or specific
+ * (e.g., "Eye color").
+ *
+ * Return value: the Name ID found for the color.
+ *
+ * Since: 2.1.0
+ */
+hb_ot_name_id_t
+hb_ot_color_palette_color_get_name_id (hb_face_t *face,
+				       unsigned int color_index)
+{
+  return face->table.CPAL->get_color_name_id (color_index);
+}
+
+/**
+ * hb_ot_color_palette_get_flags:
+ * @face: #hb_face_t to work upon
+ * @palette_index: The index of the color palette
+ *
+ * Fetches the flags defined for a color palette.
+ *
+ * Return value: the #hb_ot_color_palette_flags_t of the requested color palette
+ *
+ * Since: 2.1.0
+ */
+hb_ot_color_palette_flags_t
+hb_ot_color_palette_get_flags (hb_face_t *face,
+			       unsigned int palette_index)
+{
+  return face->table.CPAL->get_palette_flags (palette_index);
+}
+
+/**
+ * hb_ot_color_palette_get_colors:
+ * @face: #hb_face_t to work upon
+ * @palette_index: the index of the color palette to query
+ * @start_offset: offset of the first color to retrieve
+ * @color_count: (inout) (optional): Input = the maximum number of colors to return;
+ *               Output = the actual number of colors returned (may be zero)
+ * @colors: (out) (array length=color_count) (nullable): The array of #hb_color_t records found
+ *
+ * Fetches a list of the colors in a color palette.
+ *
+ * After calling this function, @colors will be filled with the palette
+ * colors. If @colors is NULL, the function will just return the number
+ * of total colors without storing any actual colors; this can be used
+ * for allocating a buffer of suitable size before calling
+ * hb_ot_color_palette_get_colors() a second time.
+ *
+ * Return value: the total number of colors in the palette
+ *
+ * Since: 2.1.0
+ */
+unsigned int
+hb_ot_color_palette_get_colors (hb_face_t     *face,
+				unsigned int   palette_index,
+				unsigned int   start_offset,
+				unsigned int  *colors_count  /* IN/OUT.  May be NULL. */,
+				hb_color_t    *colors        /* OUT.     May be NULL. */)
+{
+  return face->table.CPAL->get_palette_colors (palette_index, start_offset, colors_count, colors);
+}
+
+
+/*
+ * COLR
+ */
+
+/**
+ * hb_ot_color_has_layers:
+ * @face: #hb_face_t to work upon
+ *
+ * Tests whether a face includes any `COLR` color layers.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ * Since: 2.1.0
+ */
+hb_bool_t
+hb_ot_color_has_layers (hb_face_t *face)
+{
+  return face->table.COLR->has_data ();
+}
+
+/**
+ * hb_ot_color_glyph_get_layers:
+ * @face: #hb_face_t to work upon
+ * @glyph: The glyph index to query
+ * @start_offset: offset of the first layer to retrieve
+ * @layer_count: (inout) (optional): Input = the maximum number of layers to return;
+ *         Output = the actual number of layers returned (may be zero)
+ * @layers: (out) (array length=layer_count) (nullable): The array of layers found
+ *
+ * Fetches a list of all color layers for the specified glyph index in the specified
+ * face. The list returned will begin at the offset provided.
+ *
+ * Return value: Total number of layers available for the glyph index queried
+ *
+ * Since: 2.1.0
+ */
+unsigned int
+hb_ot_color_glyph_get_layers (hb_face_t           *face,
+			      hb_codepoint_t       glyph,
+			      unsigned int         start_offset,
+			      unsigned int        *layer_count, /* IN/OUT.  May be NULL. */
+			      hb_ot_color_layer_t *layers /* OUT.     May be NULL. */)
+{
+  return face->table.COLR->get_glyph_layers (glyph, start_offset, layer_count, layers);
+}
+
+
+/*
+ * SVG
+ */
+
+/**
+ * hb_ot_color_has_svg:
+ * @face: #hb_face_t to work upon.
+ *
+ * Tests whether a face includes any `SVG` glyph images.
+ *
+ * Return value: true if data found, false otherwise.
+ *
+ * Since: 2.1.0
+ */
+hb_bool_t
+hb_ot_color_has_svg (hb_face_t *face)
+{
+  return face->table.SVG->has_data ();
+}
+
+/**
+ * hb_ot_color_glyph_reference_svg:
+ * @face: #hb_face_t to work upon
+ * @glyph: a svg glyph index
+ *
+ * Fetches the SVG document for a glyph. The blob may be either plain text or gzip-encoded.
+ *
+ * Return value: (transfer full): An #hb_blob_t containing the SVG document of the glyph, if available
+ *
+ * Since: 2.1.0
+ */
+hb_blob_t *
+hb_ot_color_glyph_reference_svg (hb_face_t *face, hb_codepoint_t glyph)
+{
+  return face->table.SVG->reference_blob_for_glyph (glyph);
+}
+
+
+/*
+ * PNG: CBDT or sbix
+ */
+
+/**
+ * hb_ot_color_has_png:
+ * @face: #hb_face_t to work upon
+ *
+ * Tests whether a face has PNG glyph images (either in `CBDT` or `sbix` tables).
+ *
+ * Return value: true if data found, false otherwise
+ *
+ * Since: 2.1.0
+ */
+hb_bool_t
+hb_ot_color_has_png (hb_face_t *face)
+{
+  return face->table.CBDT->has_data () || face->table.sbix->has_data ();
+}
+
+/**
+ * hb_ot_color_glyph_reference_png:
+ * @font: #hb_font_t to work upon
+ * @glyph: a glyph index
+ *
+ * Fetches the PNG image for a glyph. This function takes a font object, not a face object,
+ * as input. To get an optimally sized PNG blob, the UPEM value must be set on the @font
+ * object. If UPEM is unset, the blob returned will be the largest PNG available.
+ *
+ * Return value: (transfer full): An #hb_blob_t containing the PNG image for the glyph, if available
+ *
+ * Since: 2.1.0
+ */
+hb_blob_t *
+hb_ot_color_glyph_reference_png (hb_font_t *font, hb_codepoint_t  glyph)
+{
+  hb_blob_t *blob = hb_blob_get_empty ();
+
+  if (font->face->table.sbix->has_data ())
+    blob = font->face->table.sbix->reference_png (font, glyph, nullptr, nullptr, nullptr);
+
+  if (!blob->length && font->face->table.CBDT->has_data ())
+    blob = font->face->table.CBDT->reference_png (font, glyph);
+
+  return blob;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-color.h b/thirdparty/harfbuzz/src/hb-ot-color.h
new file mode 100644
index 0000000000..63ef20a1a0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-color.h
@@ -0,0 +1,139 @@
+/*
+ * Copyright © 2016  Google, Inc.
+ * Copyright © 2018  Khaled Hosny
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Sascha Brawer, Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_COLOR_H
+#define HB_OT_COLOR_H
+
+#include "hb.h"
+#include "hb-ot-name.h"
+
+HB_BEGIN_DECLS
+
+
+/*
+ * Color palettes.
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_color_has_palettes (hb_face_t *face);
+
+HB_EXTERN unsigned int
+hb_ot_color_palette_get_count (hb_face_t *face);
+
+HB_EXTERN hb_ot_name_id_t
+hb_ot_color_palette_get_name_id (hb_face_t *face,
+				 unsigned int palette_index);
+
+HB_EXTERN hb_ot_name_id_t
+hb_ot_color_palette_color_get_name_id (hb_face_t *face,
+				       unsigned int color_index);
+
+/**
+ * hb_ot_color_palette_flags_t:
+ * @HB_OT_COLOR_PALETTE_FLAG_DEFAULT: Default indicating that there is nothing special
+ *   to note about a color palette.
+ * @HB_OT_COLOR_PALETTE_FLAG_USABLE_WITH_LIGHT_BACKGROUND: Flag indicating that the color
+ *   palette is appropriate to use when displaying the font on a light background such as white.
+ * @HB_OT_COLOR_PALETTE_FLAG_USABLE_WITH_DARK_BACKGROUND: Flag indicating that the color
+ *   palette is appropriate to use when displaying the font on a dark background such as black.
+ *
+ * Since: 2.1.0
+ */
+typedef enum { /*< flags >*/
+  HB_OT_COLOR_PALETTE_FLAG_DEFAULT			= 0x00000000u,
+  HB_OT_COLOR_PALETTE_FLAG_USABLE_WITH_LIGHT_BACKGROUND	= 0x00000001u,
+  HB_OT_COLOR_PALETTE_FLAG_USABLE_WITH_DARK_BACKGROUND	= 0x00000002u
+} hb_ot_color_palette_flags_t;
+
+HB_EXTERN hb_ot_color_palette_flags_t
+hb_ot_color_palette_get_flags (hb_face_t *face,
+			       unsigned int palette_index);
+
+HB_EXTERN unsigned int
+hb_ot_color_palette_get_colors (hb_face_t    *face,
+				unsigned int  palette_index,
+				unsigned int  start_offset,
+				unsigned int *color_count,  /* IN/OUT.  May be NULL. */
+				hb_color_t   *colors        /* OUT.     May be NULL. */);
+
+
+/*
+ * Color layers.
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_color_has_layers (hb_face_t *face);
+
+/**
+ * hb_ot_color_layer_t:
+ *
+ * Pairs of glyph and color index.
+ *
+ * Since: 2.1.0
+ **/
+typedef struct hb_ot_color_layer_t
+{
+  hb_codepoint_t glyph;
+  unsigned int   color_index;
+} hb_ot_color_layer_t;
+
+HB_EXTERN unsigned int
+hb_ot_color_glyph_get_layers (hb_face_t           *face,
+			      hb_codepoint_t       glyph,
+			      unsigned int         start_offset,
+			      unsigned int        *layer_count, /* IN/OUT.  May be NULL. */
+			      hb_ot_color_layer_t *layers /* OUT.     May be NULL. */);
+
+/*
+ * SVG
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_color_has_svg (hb_face_t *face);
+
+HB_EXTERN hb_blob_t *
+hb_ot_color_glyph_reference_svg (hb_face_t *face, hb_codepoint_t glyph);
+
+/*
+ * PNG: CBDT or sbix
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_color_has_png (hb_face_t *face);
+
+HB_EXTERN hb_blob_t *
+hb_ot_color_glyph_reference_png (hb_font_t *font, hb_codepoint_t glyph);
+
+
+HB_END_DECLS
+
+#endif /* HB_OT_COLOR_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-deprecated.h b/thirdparty/harfbuzz/src/hb-ot-deprecated.h
new file mode 100644
index 0000000000..bc72f8a701
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-deprecated.h
@@ -0,0 +1,111 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_DEPRECATED_H
+#define HB_OT_DEPRECATED_H
+
+#include "hb.h"
+#include "hb-ot-name.h"
+
+
+HB_BEGIN_DECLS
+
+#ifndef HB_DISABLE_DEPRECATED
+
+
+/* https://github.com/harfbuzz/harfbuzz/issues/1734 */
+#define HB_MATH_GLYPH_PART_FLAG_EXTENDER HB_OT_MATH_GLYPH_PART_FLAG_EXTENDER
+
+
+/* Like hb_ot_layout_table_find_script, but takes zero-terminated array of scripts to test */
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_layout_table_select_script) hb_bool_t
+hb_ot_layout_table_choose_script (hb_face_t      *face,
+				  hb_tag_t        table_tag,
+				  const hb_tag_t *script_tags,
+				  unsigned int   *script_index,
+				  hb_tag_t       *chosen_script);
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_layout_script_select_language) hb_bool_t
+hb_ot_layout_script_find_language (hb_face_t    *face,
+				   hb_tag_t      table_tag,
+				   unsigned int  script_index,
+				   hb_tag_t      language_tag,
+				   unsigned int *language_index);
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_tags_from_script_and_language) void
+hb_ot_tags_from_script (hb_script_t  script,
+			hb_tag_t    *script_tag_1,
+			hb_tag_t    *script_tag_2);
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_tags_from_script_and_language) hb_tag_t
+hb_ot_tag_from_language (hb_language_t language);
+
+
+/**
+ * HB_OT_VAR_NO_AXIS_INDEX:
+ *
+ * Since: 1.4.2
+ * Deprecated: 2.2.0
+ */
+#define HB_OT_VAR_NO_AXIS_INDEX		0xFFFFFFFFu
+
+/**
+ * hb_ot_var_axis_t:
+ *
+ * Since: 1.4.2
+ * Deprecated: 2.2.0
+ */
+typedef struct hb_ot_var_axis_t
+{
+  hb_tag_t tag;
+  hb_ot_name_id_t name_id;
+  float min_value;
+  float default_value;
+  float max_value;
+} hb_ot_var_axis_t;
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_var_get_axis_infos) unsigned int
+hb_ot_var_get_axes (hb_face_t        *face,
+		    unsigned int      start_offset,
+		    unsigned int     *axes_count /* IN/OUT */,
+		    hb_ot_var_axis_t *axes_array /* OUT */);
+
+HB_EXTERN HB_DEPRECATED_FOR (hb_ot_var_find_axis_info) hb_bool_t
+hb_ot_var_find_axis (hb_face_t        *face,
+		     hb_tag_t          axis_tag,
+		     unsigned int     *axis_index,
+		     hb_ot_var_axis_t *axis_info);
+
+
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_OT_DEPRECATED_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-face-table-list.hh b/thirdparty/harfbuzz/src/hb-ot-face-table-list.hh
new file mode 100644
index 0000000000..367e143fdf
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-face-table-list.hh
@@ -0,0 +1,138 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2012,2013  Google, Inc.
+ * Copyright © 2019, Facebook Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_FACE_TABLE_LIST_HH
+#define HB_OT_FACE_TABLE_LIST_HH
+#endif /* HB_OT_FACE_TABLE_LIST_HH */ /* Dummy header guards */
+
+#ifndef HB_OT_ACCELERATOR
+#define HB_OT_ACCELERATOR(Namespace, Type) HB_OT_TABLE (Namespace, Type)
+#define _HB_OT_ACCELERATOR_UNDEF
+#endif
+
+
+/* This lists font tables that the hb_face_t will contain and lazily
+ * load.  Don't add a table unless it's used though.  This is not
+ * exactly free. */
+
+/* v--- Add new tables in the right place here. */
+
+
+/* OpenType fundamentals. */
+HB_OT_TABLE (OT, head)
+#if !defined(HB_NO_FACE_COLLECT_UNICODES) || !defined(HB_NO_OT_FONT)
+HB_OT_ACCELERATOR (OT, cmap)
+#endif
+HB_OT_TABLE (OT, hhea)
+HB_OT_ACCELERATOR (OT, hmtx)
+HB_OT_TABLE (OT, OS2)
+#if !defined(HB_NO_OT_FONT_GLYPH_NAMES) || !defined(HB_NO_METRICS) || !defined(HB_NO_STYLE)
+HB_OT_ACCELERATOR (OT, post)
+#endif
+#ifndef HB_NO_NAME
+HB_OT_ACCELERATOR (OT, name)
+#endif
+#ifndef HB_NO_STYLE
+HB_OT_TABLE (OT, STAT)
+#endif
+#ifndef HB_NO_META
+HB_OT_ACCELERATOR (OT, meta)
+#endif
+
+/* Vertical layout. */
+HB_OT_TABLE (OT, vhea)
+HB_OT_ACCELERATOR (OT, vmtx)
+
+/* TrueType outlines. */
+HB_OT_ACCELERATOR (OT, glyf)
+
+/* CFF outlines. */
+#ifndef HB_NO_CFF
+HB_OT_ACCELERATOR (OT, cff1)
+HB_OT_ACCELERATOR (OT, cff2)
+HB_OT_TABLE (OT, VORG)
+#endif
+
+/* OpenType variations. */
+#ifndef HB_NO_VAR
+HB_OT_TABLE (OT, fvar)
+HB_OT_TABLE (OT, avar)
+HB_OT_ACCELERATOR (OT, gvar)
+HB_OT_TABLE (OT, MVAR)
+#endif
+
+/* Legacy kern. */
+#ifndef HB_NO_OT_KERN
+HB_OT_TABLE (OT, kern)
+#endif
+
+/* OpenType shaping. */
+#ifndef HB_NO_OT_LAYOUT
+HB_OT_ACCELERATOR (OT, GDEF)
+HB_OT_ACCELERATOR (OT, GSUB)
+HB_OT_ACCELERATOR (OT, GPOS)
+//HB_OT_TABLE (OT, JSTF)
+#endif
+
+/* OpenType baseline. */
+#ifndef HB_NO_BASE
+HB_OT_TABLE (OT, BASE)
+#endif
+
+/* AAT shaping. */
+#ifndef HB_NO_AAT
+HB_OT_TABLE (AAT, morx)
+HB_OT_TABLE (AAT, mort)
+HB_OT_TABLE (AAT, kerx)
+HB_OT_TABLE (AAT, ankr)
+HB_OT_TABLE (AAT, trak)
+HB_OT_TABLE (AAT, ltag)
+HB_OT_TABLE (AAT, feat)
+// HB_OT_TABLE (AAT, opbd)
+#endif
+
+/* OpenType color fonts. */
+#ifndef HB_NO_COLOR
+HB_OT_TABLE (OT, COLR)
+HB_OT_TABLE (OT, CPAL)
+HB_OT_ACCELERATOR (OT, CBDT)
+HB_OT_ACCELERATOR (OT, sbix)
+HB_OT_ACCELERATOR (OT, SVG)
+#endif
+
+/* OpenType math. */
+#ifndef HB_NO_MATH
+HB_OT_TABLE (OT, MATH)
+#endif
+
+
+#ifdef _HB_OT_ACCELERATOR_UNDEF
+#undef HB_OT_ACCELERATOR
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-face.cc b/thirdparty/harfbuzz/src/hb-ot-face.cc
new file mode 100644
index 0000000000..5ef8df43ce
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-face.cc
@@ -0,0 +1,58 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb-ot-face.hh"
+
+#include "hb-ot-cmap-table.hh"
+#include "hb-ot-glyf-table.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-ot-cff2-table.hh"
+#include "hb-ot-hmtx-table.hh"
+#include "hb-ot-kern-table.hh"
+#include "hb-ot-meta-table.hh"
+#include "hb-ot-name-table.hh"
+#include "hb-ot-post-table.hh"
+#include "hb-ot-color-cbdt-table.hh"
+#include "hb-ot-color-sbix-table.hh"
+#include "hb-ot-color-svg-table.hh"
+#include "hb-ot-layout-gdef-table.hh"
+#include "hb-ot-layout-gsub-table.hh"
+#include "hb-ot-layout-gpos-table.hh"
+
+
+void hb_ot_face_t::init0 (hb_face_t *face)
+{
+  this->face = face;
+#define HB_OT_TABLE(Namespace, Type) Type.init0 ();
+#include "hb-ot-face-table-list.hh"
+#undef HB_OT_TABLE
+}
+void hb_ot_face_t::fini ()
+{
+#define HB_OT_TABLE(Namespace, Type) Type.fini ();
+#include "hb-ot-face-table-list.hh"
+#undef HB_OT_TABLE
+}
diff --git a/thirdparty/harfbuzz/src/hb-ot-face.hh b/thirdparty/harfbuzz/src/hb-ot-face.hh
new file mode 100644
index 0000000000..e24d380bca
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-face.hh
@@ -0,0 +1,74 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_FACE_HH
+#define HB_OT_FACE_HH
+
+#include "hb.hh"
+
+#include "hb-machinery.hh"
+
+
+/*
+ * hb_ot_face_t
+ */
+
+/* Declare tables. */
+#define HB_OT_TABLE(Namespace, Type) namespace Namespace { struct Type; }
+#define HB_OT_ACCELERATOR(Namespace, Type) HB_OT_TABLE (Namespace, Type##_accelerator_t)
+#include "hb-ot-face-table-list.hh"
+#undef HB_OT_ACCELERATOR
+#undef HB_OT_TABLE
+
+struct hb_ot_face_t
+{
+  HB_INTERNAL void init0 (hb_face_t *face);
+  HB_INTERNAL void fini ();
+
+#define HB_OT_TABLE_ORDER(Namespace, Type) \
+    HB_PASTE (ORDER_, HB_PASTE (Namespace, HB_PASTE (_, Type)))
+  enum order_t
+  {
+    ORDER_ZERO,
+#define HB_OT_TABLE(Namespace, Type) HB_OT_TABLE_ORDER (Namespace, Type),
+#include "hb-ot-face-table-list.hh"
+#undef HB_OT_TABLE
+  };
+
+  hb_face_t *face; /* MUST be JUST before the lazy loaders. */
+#define HB_OT_TABLE(Namespace, Type) \
+  hb_table_lazy_loader_t<Namespace::Type, HB_OT_TABLE_ORDER (Namespace, Type)> Type;
+#define HB_OT_ACCELERATOR(Namespace, Type) \
+  hb_face_lazy_loader_t<Namespace::Type##_accelerator_t, HB_OT_TABLE_ORDER (Namespace, Type)> Type;
+#include "hb-ot-face-table-list.hh"
+#undef HB_OT_ACCELERATOR
+#undef HB_OT_TABLE
+};
+
+
+#endif /* HB_OT_FACE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-font.cc b/thirdparty/harfbuzz/src/hb-ot-font.cc
new file mode 100644
index 0000000000..a1dc88603a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-font.cc
@@ -0,0 +1,336 @@
+/*
+ * Copyright © 2011,2014  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod, Roozbeh Pournader
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_FONT
+
+#include "hb-ot.h"
+
+#include "hb-font.hh"
+#include "hb-machinery.hh"
+#include "hb-ot-face.hh"
+
+#include "hb-ot-cmap-table.hh"
+#include "hb-ot-glyf-table.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-ot-cff2-table.hh"
+#include "hb-ot-hmtx-table.hh"
+#include "hb-ot-os2-table.hh"
+#include "hb-ot-post-table.hh"
+#include "hb-ot-stat-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-ot-vorg-table.hh"
+#include "hb-ot-color-cbdt-table.hh"
+#include "hb-ot-color-sbix-table.hh"
+
+
+/**
+ * SECTION:hb-ot-font
+ * @title: hb-ot-font
+ * @short_description: OpenType font implementation
+ * @include: hb-ot.h
+ *
+ * Functions for using OpenType fonts with hb_shape().  Note that fonts returned
+ * by hb_font_create() default to using these functions, so most clients would
+ * never need to call these functions directly.
+ **/
+
+
+static hb_bool_t
+hb_ot_get_nominal_glyph (hb_font_t *font HB_UNUSED,
+			 void *font_data,
+			 hb_codepoint_t unicode,
+			 hb_codepoint_t *glyph,
+			 void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  return ot_face->cmap->get_nominal_glyph (unicode, glyph);
+}
+
+static unsigned int
+hb_ot_get_nominal_glyphs (hb_font_t *font HB_UNUSED,
+			  void *font_data,
+			  unsigned int count,
+			  const hb_codepoint_t *first_unicode,
+			  unsigned int unicode_stride,
+			  hb_codepoint_t *first_glyph,
+			  unsigned int glyph_stride,
+			  void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  return ot_face->cmap->get_nominal_glyphs (count,
+					    first_unicode, unicode_stride,
+					    first_glyph, glyph_stride);
+}
+
+static hb_bool_t
+hb_ot_get_variation_glyph (hb_font_t *font HB_UNUSED,
+			   void *font_data,
+			   hb_codepoint_t unicode,
+			   hb_codepoint_t variation_selector,
+			   hb_codepoint_t *glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  return ot_face->cmap->get_variation_glyph (unicode, variation_selector, glyph);
+}
+
+static void
+hb_ot_get_glyph_h_advances (hb_font_t* font, void* font_data,
+			    unsigned count,
+			    const hb_codepoint_t *first_glyph,
+			    unsigned glyph_stride,
+			    hb_position_t *first_advance,
+			    unsigned advance_stride,
+			    void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  const OT::hmtx_accelerator_t &hmtx = *ot_face->hmtx;
+
+  for (unsigned int i = 0; i < count; i++)
+  {
+    *first_advance = font->em_scale_x (hmtx.get_advance (*first_glyph, font));
+    first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+    first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+  }
+}
+
+static void
+hb_ot_get_glyph_v_advances (hb_font_t* font, void* font_data,
+			    unsigned count,
+			    const hb_codepoint_t *first_glyph,
+			    unsigned glyph_stride,
+			    hb_position_t *first_advance,
+			    unsigned advance_stride,
+			    void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  const OT::vmtx_accelerator_t &vmtx = *ot_face->vmtx;
+
+  for (unsigned int i = 0; i < count; i++)
+  {
+    *first_advance = font->em_scale_y (-(int) vmtx.get_advance (*first_glyph, font));
+    first_glyph = &StructAtOffsetUnaligned<hb_codepoint_t> (first_glyph, glyph_stride);
+    first_advance = &StructAtOffsetUnaligned<hb_position_t> (first_advance, advance_stride);
+  }
+}
+
+static hb_bool_t
+hb_ot_get_glyph_v_origin (hb_font_t *font,
+			  void *font_data,
+			  hb_codepoint_t glyph,
+			  hb_position_t *x,
+			  hb_position_t *y,
+			  void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+
+  *x = font->get_glyph_h_advance (glyph) / 2;
+
+#ifndef HB_NO_OT_FONT_CFF
+  const OT::VORG &VORG = *ot_face->VORG;
+  if (VORG.has_data ())
+  {
+    *y = font->em_scale_y (VORG.get_y_origin (glyph));
+    return true;
+  }
+#endif
+
+  hb_glyph_extents_t extents = {0};
+  if (ot_face->glyf->get_extents (font, glyph, &extents))
+  {
+    const OT::vmtx_accelerator_t &vmtx = *ot_face->vmtx;
+    hb_position_t tsb = vmtx.get_side_bearing (font, glyph);
+    *y = extents.y_bearing + font->em_scale_y (tsb);
+    return true;
+  }
+
+  hb_font_extents_t font_extents;
+  font->get_h_extents_with_fallback (&font_extents);
+  *y = font_extents.ascender;
+
+  return true;
+}
+
+static hb_bool_t
+hb_ot_get_glyph_extents (hb_font_t *font,
+			 void *font_data,
+			 hb_codepoint_t glyph,
+			 hb_glyph_extents_t *extents,
+			 void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+
+#if !defined(HB_NO_OT_FONT_BITMAP) && !defined(HB_NO_COLOR)
+  if (ot_face->sbix->get_extents (font, glyph, extents)) return true;
+#endif
+  if (ot_face->glyf->get_extents (font, glyph, extents)) return true;
+#ifndef HB_NO_OT_FONT_CFF
+  if (ot_face->cff1->get_extents (font, glyph, extents)) return true;
+  if (ot_face->cff2->get_extents (font, glyph, extents)) return true;
+#endif
+#if !defined(HB_NO_OT_FONT_BITMAP) && !defined(HB_NO_COLOR)
+  if (ot_face->CBDT->get_extents (font, glyph, extents)) return true;
+#endif
+
+  // TODO Hook up side-bearings variations.
+  return false;
+}
+
+#ifndef HB_NO_OT_FONT_GLYPH_NAMES
+static hb_bool_t
+hb_ot_get_glyph_name (hb_font_t *font HB_UNUSED,
+		      void *font_data,
+		      hb_codepoint_t glyph,
+		      char *name, unsigned int size,
+		      void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  if (ot_face->post->get_glyph_name (glyph, name, size)) return true;
+#ifndef HB_NO_OT_FONT_CFF
+  if (ot_face->cff1->get_glyph_name (glyph, name, size)) return true;
+#endif
+  return false;
+}
+static hb_bool_t
+hb_ot_get_glyph_from_name (hb_font_t *font HB_UNUSED,
+			   void *font_data,
+			   const char *name, int len,
+			   hb_codepoint_t *glyph,
+			   void *user_data HB_UNUSED)
+{
+  const hb_ot_face_t *ot_face = (const hb_ot_face_t *) font_data;
+  if (ot_face->post->get_glyph_from_name (name, len, glyph)) return true;
+#ifndef HB_NO_OT_FONT_CFF
+    if (ot_face->cff1->get_glyph_from_name (name, len, glyph)) return true;
+#endif
+  return false;
+}
+#endif
+
+static hb_bool_t
+hb_ot_get_font_h_extents (hb_font_t *font,
+			  void *font_data HB_UNUSED,
+			  hb_font_extents_t *metrics,
+			  void *user_data HB_UNUSED)
+{
+  return _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER, &metrics->ascender) &&
+	 _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER, &metrics->descender) &&
+	 _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP, &metrics->line_gap);
+}
+
+static hb_bool_t
+hb_ot_get_font_v_extents (hb_font_t *font,
+			  void *font_data HB_UNUSED,
+			  hb_font_extents_t *metrics,
+			  void *user_data HB_UNUSED)
+{
+  return _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_VERTICAL_ASCENDER, &metrics->ascender) &&
+	 _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_VERTICAL_DESCENDER, &metrics->descender) &&
+	 _hb_ot_metrics_get_position_common (font, HB_OT_METRICS_TAG_VERTICAL_LINE_GAP, &metrics->line_gap);
+}
+
+#if HB_USE_ATEXIT
+static void free_static_ot_funcs ();
+#endif
+
+static struct hb_ot_font_funcs_lazy_loader_t : hb_font_funcs_lazy_loader_t<hb_ot_font_funcs_lazy_loader_t>
+{
+  static hb_font_funcs_t *create ()
+  {
+    hb_font_funcs_t *funcs = hb_font_funcs_create ();
+
+    hb_font_funcs_set_font_h_extents_func (funcs, hb_ot_get_font_h_extents, nullptr, nullptr);
+    hb_font_funcs_set_font_v_extents_func (funcs, hb_ot_get_font_v_extents, nullptr, nullptr);
+    hb_font_funcs_set_nominal_glyph_func (funcs, hb_ot_get_nominal_glyph, nullptr, nullptr);
+    hb_font_funcs_set_nominal_glyphs_func (funcs, hb_ot_get_nominal_glyphs, nullptr, nullptr);
+    hb_font_funcs_set_variation_glyph_func (funcs, hb_ot_get_variation_glyph, nullptr, nullptr);
+    hb_font_funcs_set_glyph_h_advances_func (funcs, hb_ot_get_glyph_h_advances, nullptr, nullptr);
+    hb_font_funcs_set_glyph_v_advances_func (funcs, hb_ot_get_glyph_v_advances, nullptr, nullptr);
+    //hb_font_funcs_set_glyph_h_origin_func (funcs, hb_ot_get_glyph_h_origin, nullptr, nullptr);
+    hb_font_funcs_set_glyph_v_origin_func (funcs, hb_ot_get_glyph_v_origin, nullptr, nullptr);
+    hb_font_funcs_set_glyph_extents_func (funcs, hb_ot_get_glyph_extents, nullptr, nullptr);
+    //hb_font_funcs_set_glyph_contour_point_func (funcs, hb_ot_get_glyph_contour_point, nullptr, nullptr);
+#ifndef HB_NO_OT_FONT_GLYPH_NAMES
+    hb_font_funcs_set_glyph_name_func (funcs, hb_ot_get_glyph_name, nullptr, nullptr);
+    hb_font_funcs_set_glyph_from_name_func (funcs, hb_ot_get_glyph_from_name, nullptr, nullptr);
+#endif
+
+    hb_font_funcs_make_immutable (funcs);
+
+#if HB_USE_ATEXIT
+    atexit (free_static_ot_funcs);
+#endif
+
+    return funcs;
+  }
+} static_ot_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_ot_funcs ()
+{
+  static_ot_funcs.free_instance ();
+}
+#endif
+
+static hb_font_funcs_t *
+_hb_ot_get_font_funcs ()
+{
+  return static_ot_funcs.get_unconst ();
+}
+
+
+/**
+ * hb_ot_font_set_funcs:
+ *
+ * Since: 0.9.28
+ **/
+void
+hb_ot_font_set_funcs (hb_font_t *font)
+{
+  hb_font_set_funcs (font,
+		     _hb_ot_get_font_funcs (),
+		     &font->face->table,
+		     nullptr);
+}
+
+#ifndef HB_NO_VAR
+int
+_glyf_get_side_bearing_var (hb_font_t *font, hb_codepoint_t glyph, bool is_vertical)
+{
+  return font->face->table.glyf->get_side_bearing_var (font, glyph, is_vertical);
+}
+
+unsigned
+_glyf_get_advance_var (hb_font_t *font, hb_codepoint_t glyph, bool is_vertical)
+{
+  return font->face->table.glyf->get_advance_var (font, glyph, is_vertical);
+}
+#endif
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-font.h b/thirdparty/harfbuzz/src/hb-ot-font.h
new file mode 100644
index 0000000000..80eaa54b1a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-font.h
@@ -0,0 +1,45 @@
+/*
+ * Copyright © 2014  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod, Roozbeh Pournader
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_FONT_H
+#define HB_OT_FONT_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN void
+hb_ot_font_set_funcs (hb_font_t *font);
+
+
+HB_END_DECLS
+
+#endif /* HB_OT_FONT_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-gasp-table.hh b/thirdparty/harfbuzz/src/hb-ot-gasp-table.hh
new file mode 100644
index 0000000000..4f291924af
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-gasp-table.hh
@@ -0,0 +1,84 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_GASP_TABLE_HH
+#define HB_OT_GASP_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-hhea-table.hh"
+#include "hb-ot-os2-table.hh"
+#include "hb-ot-var-hvar-table.hh"
+
+/*
+ * gasp -- Grid-fitting and Scan-conversion Procedure
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/gasp
+ */
+#define HB_OT_TAG_gasp HB_TAG('g','a','s','p')
+
+
+namespace OT {
+
+struct GaspRange
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	rangeMaxPPEM;	/* Upper limit of range, in PPEM */
+  HBUINT16	rangeGaspBehavior;
+				/* Flags describing desired rasterizer behavior. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct gasp
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_gasp;
+
+  const GaspRange &get_gasp_range (unsigned int i) const
+  { return gaspRanges[i]; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  gaspRanges.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	version;	/* Version number (set to 1) */
+  ArrayOf<GaspRange>
+		gaspRanges;	/* Number of records to follow
+				 * Sorted by ppem */
+  public:
+  DEFINE_SIZE_ARRAY (4, gaspRanges);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_GASP_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-glyf-table.hh b/thirdparty/harfbuzz/src/hb-ot-glyf-table.hh
new file mode 100644
index 0000000000..5470bd96da
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-glyf-table.hh
@@ -0,0 +1,1261 @@
+/*
+ * Copyright © 2015  Google, Inc.
+ * Copyright © 2019  Adobe Inc.
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod, Garret Rieger, Roderick Sheeter
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_GLYF_TABLE_HH
+#define HB_OT_GLYF_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-head-table.hh"
+#include "hb-ot-hmtx-table.hh"
+#include "hb-ot-var-gvar-table.hh"
+#include "hb-draw.hh"
+
+namespace OT {
+
+
+/*
+ * loca -- Index to Location
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/loca
+ */
+#define HB_OT_TAG_loca HB_TAG('l','o','c','a')
+
+
+struct loca
+{
+  friend struct glyf;
+
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_loca;
+
+  bool sanitize (hb_sanitize_context_t *c HB_UNUSED) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (true);
+  }
+
+  protected:
+  UnsizedArrayOf<HBUINT8>
+		dataZ;	/* Location data. */
+  public:
+  DEFINE_SIZE_MIN (0);	/* In reality, this is UNBOUNDED() type; but since we always
+			 * check the size externally, allow Null() object of it by
+			 * defining it _MIN instead. */
+};
+
+
+/*
+ * glyf -- TrueType Glyph Data
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/glyf
+ */
+#define HB_OT_TAG_glyf HB_TAG('g','l','y','f')
+
+
+struct glyf
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_glyf;
+
+  bool sanitize (hb_sanitize_context_t *c HB_UNUSED) const
+  {
+    TRACE_SANITIZE (this);
+    /* Runtime checks as eager sanitizing each glyph is costy */
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_source_of (Iterator, unsigned int))>
+  static bool
+  _add_loca_and_head (hb_subset_plan_t * plan, Iterator padded_offsets)
+  {
+    unsigned max_offset =
+    + padded_offsets
+    | hb_reduce (hb_add, 0)
+    ;
+    unsigned num_offsets = padded_offsets.len () + 1;
+    bool use_short_loca = max_offset < 0x1FFFF;
+    unsigned entry_size = use_short_loca ? 2 : 4;
+    char *loca_prime_data = (char *) calloc (entry_size, num_offsets);
+
+    if (unlikely (!loca_prime_data)) return false;
+
+    DEBUG_MSG (SUBSET, nullptr, "loca entry_size %d num_offsets %d "
+				"max_offset %d size %d",
+	       entry_size, num_offsets, max_offset, entry_size * num_offsets);
+
+    if (use_short_loca)
+      _write_loca (padded_offsets, 1, hb_array ((HBUINT16 *) loca_prime_data, num_offsets));
+    else
+      _write_loca (padded_offsets, 0, hb_array ((HBUINT32 *) loca_prime_data, num_offsets));
+
+    hb_blob_t *loca_blob = hb_blob_create (loca_prime_data,
+					   entry_size * num_offsets,
+					   HB_MEMORY_MODE_WRITABLE,
+					   loca_prime_data,
+					   free);
+
+    bool result = plan->add_table (HB_OT_TAG_loca, loca_blob)
+	       && _add_head_and_set_loca_version (plan, use_short_loca);
+
+    hb_blob_destroy (loca_blob);
+    return result;
+  }
+
+  template<typename IteratorIn, typename IteratorOut,
+	   hb_requires (hb_is_source_of (IteratorIn, unsigned int)),
+	   hb_requires (hb_is_sink_of (IteratorOut, unsigned))>
+  static void
+  _write_loca (IteratorIn it, unsigned right_shift, IteratorOut dest)
+  {
+    unsigned int offset = 0;
+    dest << 0;
+    + it
+    | hb_map ([=, &offset] (unsigned int padded_size)
+	      {
+		offset += padded_size;
+		DEBUG_MSG (SUBSET, nullptr, "loca entry offset %d", offset);
+		return offset >> right_shift;
+	      })
+    | hb_sink (dest)
+    ;
+  }
+
+  /* requires source of SubsetGlyph complains the identifier isn't declared */
+  template <typename Iterator>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  const hb_subset_plan_t *plan)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned init_len = c->length ();
+    for (const auto &_ : it) _.serialize (c, plan);
+
+    /* As a special case when all glyph in the font are empty, add a zero byte
+     * to the table, so that OTS doesn’t reject it, and to make the table work
+     * on Windows as well.
+     * See https://github.com/khaledhosny/ots/issues/52 */
+    if (init_len == c->length ())
+    {
+      HBUINT8 empty_byte;
+      empty_byte = 0;
+      c->copy (empty_byte);
+    }
+    return_trace (true);
+  }
+
+  /* Byte region(s) per glyph to output
+     unpadded, hints removed if so requested
+     If we fail to process a glyph we produce an empty (0-length) glyph */
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    glyf *glyf_prime = c->serializer->start_embed <glyf> ();
+    if (unlikely (!c->serializer->check_success (glyf_prime))) return_trace (false);
+
+    hb_vector_t<SubsetGlyph> glyphs;
+    _populate_subset_glyphs (c->plan, &glyphs);
+
+    glyf_prime->serialize (c->serializer, hb_iter (glyphs), c->plan);
+
+    auto padded_offsets =
+    + hb_iter (glyphs)
+    | hb_map (&SubsetGlyph::padded_size)
+    ;
+
+    if (c->serializer->in_error ()) return_trace (false);
+    return_trace (c->serializer->check_success (_add_loca_and_head (c->plan,
+								    padded_offsets)));
+  }
+
+  template <typename SubsetGlyph>
+  void
+  _populate_subset_glyphs (const hb_subset_plan_t   *plan,
+			   hb_vector_t<SubsetGlyph> *glyphs /* OUT */) const
+  {
+    OT::glyf::accelerator_t glyf;
+    glyf.init (plan->source);
+
+    + hb_range (plan->num_output_glyphs ())
+    | hb_map ([&] (hb_codepoint_t new_gid)
+	      {
+		SubsetGlyph subset_glyph = {0};
+		subset_glyph.new_gid = new_gid;
+
+		/* should never fail: all old gids should be mapped */
+		if (!plan->old_gid_for_new_gid (new_gid, &subset_glyph.old_gid))
+		  return subset_glyph;
+
+		subset_glyph.source_glyph = glyf.glyph_for_gid (subset_glyph.old_gid, true);
+		if (plan->drop_hints) subset_glyph.drop_hints_bytes ();
+		else subset_glyph.dest_start = subset_glyph.source_glyph.get_bytes ();
+
+		return subset_glyph;
+	      })
+    | hb_sink (glyphs)
+    ;
+
+    glyf.fini ();
+  }
+
+  static bool
+  _add_head_and_set_loca_version (hb_subset_plan_t *plan, bool use_short_loca)
+  {
+    hb_blob_t *head_blob = hb_sanitize_context_t ().reference_table<head> (plan->source);
+    hb_blob_t *head_prime_blob = hb_blob_copy_writable_or_fail (head_blob);
+    hb_blob_destroy (head_blob);
+
+    if (unlikely (!head_prime_blob))
+      return false;
+
+    head *head_prime = (head *) hb_blob_get_data_writable (head_prime_blob, nullptr);
+    head_prime->indexToLocFormat = use_short_loca ? 0 : 1;
+    bool success = plan->add_table (HB_OT_TAG_head, head_prime_blob);
+
+    hb_blob_destroy (head_prime_blob);
+    return success;
+  }
+
+  struct CompositeGlyphChain
+  {
+    protected:
+    enum composite_glyph_flag_t
+    {
+      ARG_1_AND_2_ARE_WORDS	= 0x0001,
+      ARGS_ARE_XY_VALUES	= 0x0002,
+      ROUND_XY_TO_GRID		= 0x0004,
+      WE_HAVE_A_SCALE		= 0x0008,
+      MORE_COMPONENTS		= 0x0020,
+      WE_HAVE_AN_X_AND_Y_SCALE	= 0x0040,
+      WE_HAVE_A_TWO_BY_TWO	= 0x0080,
+      WE_HAVE_INSTRUCTIONS	= 0x0100,
+      USE_MY_METRICS		= 0x0200,
+      OVERLAP_COMPOUND		= 0x0400,
+      SCALED_COMPONENT_OFFSET	= 0x0800,
+      UNSCALED_COMPONENT_OFFSET = 0x1000
+    };
+
+    public:
+    unsigned int get_size () const
+    {
+      unsigned int size = min_size;
+      /* arg1 and 2 are int16 */
+      if (flags & ARG_1_AND_2_ARE_WORDS) size += 4;
+      /* arg1 and 2 are int8 */
+      else size += 2;
+
+      /* One x 16 bit (scale) */
+      if (flags & WE_HAVE_A_SCALE) size += 2;
+      /* Two x 16 bit (xscale, yscale) */
+      else if (flags & WE_HAVE_AN_X_AND_Y_SCALE) size += 4;
+      /* Four x 16 bit (xscale, scale01, scale10, yscale) */
+      else if (flags & WE_HAVE_A_TWO_BY_TWO) size += 8;
+
+      return size;
+    }
+
+    void set_glyph_index (hb_codepoint_t new_gid) { glyphIndex = new_gid; }
+    hb_codepoint_t get_glyph_index ()       const { return glyphIndex; }
+
+    void drop_instructions_flag ()  { flags = (uint16_t) flags & ~WE_HAVE_INSTRUCTIONS; }
+    bool has_instructions ()  const { return   flags & WE_HAVE_INSTRUCTIONS; }
+
+    bool has_more ()          const { return   flags & MORE_COMPONENTS; }
+    bool is_use_my_metrics () const { return   flags & USE_MY_METRICS; }
+    bool is_anchored ()       const { return !(flags & ARGS_ARE_XY_VALUES); }
+    void get_anchor_points (unsigned int &point1, unsigned int &point2) const
+    {
+      const HBUINT8 *p = &StructAfter<const HBUINT8> (glyphIndex);
+      if (flags & ARG_1_AND_2_ARE_WORDS)
+      {
+	point1 = ((const HBUINT16 *) p)[0];
+	point2 = ((const HBUINT16 *) p)[1];
+      }
+      else
+      {
+	point1 = p[0];
+	point2 = p[1];
+      }
+    }
+
+    void transform_points (contour_point_vector_t &points) const
+    {
+      float matrix[4];
+      contour_point_t trans;
+      if (get_transformation (matrix, trans))
+      {
+	if (scaled_offsets ())
+	{
+	  points.translate (trans);
+	  points.transform (matrix);
+	}
+	else
+	{
+	  points.transform (matrix);
+	  points.translate (trans);
+	}
+      }
+    }
+
+    protected:
+    bool scaled_offsets () const
+    { return (flags & (SCALED_COMPONENT_OFFSET | UNSCALED_COMPONENT_OFFSET)) == SCALED_COMPONENT_OFFSET; }
+
+    bool get_transformation (float (&matrix)[4], contour_point_t &trans) const
+    {
+      matrix[0] = matrix[3] = 1.f;
+      matrix[1] = matrix[2] = 0.f;
+
+      int tx, ty;
+      const HBINT8 *p = &StructAfter<const HBINT8> (glyphIndex);
+      if (flags & ARG_1_AND_2_ARE_WORDS)
+      {
+	tx = *(const HBINT16 *) p;
+	p += HBINT16::static_size;
+	ty = *(const HBINT16 *) p;
+	p += HBINT16::static_size;
+      }
+      else
+      {
+	tx = *p++;
+	ty = *p++;
+      }
+      if (is_anchored ()) tx = ty = 0;
+
+      trans.init ((float) tx, (float) ty);
+
+      {
+	const F2DOT14 *points = (const F2DOT14 *) p;
+	if (flags & WE_HAVE_A_SCALE)
+	{
+	  matrix[0] = matrix[3] = points[0].to_float ();
+	  return true;
+	}
+	else if (flags & WE_HAVE_AN_X_AND_Y_SCALE)
+	{
+	  matrix[0] = points[0].to_float ();
+	  matrix[3] = points[1].to_float ();
+	  return true;
+	}
+	else if (flags & WE_HAVE_A_TWO_BY_TWO)
+	{
+	  matrix[0] = points[0].to_float ();
+	  matrix[1] = points[1].to_float ();
+	  matrix[2] = points[2].to_float ();
+	  matrix[3] = points[3].to_float ();
+	  return true;
+	}
+      }
+      return tx || ty;
+    }
+
+    protected:
+    HBUINT16	flags;
+    HBGlyphID	glyphIndex;
+    public:
+    DEFINE_SIZE_MIN (4);
+  };
+
+  struct composite_iter_t : hb_iter_with_fallback_t<composite_iter_t, const CompositeGlyphChain &>
+  {
+    typedef const CompositeGlyphChain *__item_t__;
+    composite_iter_t (hb_bytes_t glyph_, __item_t__ current_) :
+      glyph (glyph_), current (current_)
+    { if (!check_range (current)) current = nullptr; }
+    composite_iter_t () : glyph (hb_bytes_t ()), current (nullptr) {}
+
+    const CompositeGlyphChain &__item__ () const { return *current; }
+    bool __more__ () const { return current; }
+    void __next__ ()
+    {
+      if (!current->has_more ()) { current = nullptr; return; }
+
+      const CompositeGlyphChain *possible = &StructAfter<CompositeGlyphChain,
+							 CompositeGlyphChain> (*current);
+      if (!check_range (possible)) { current = nullptr; return; }
+      current = possible;
+    }
+    bool operator != (const composite_iter_t& o) const
+    { return glyph != o.glyph || current != o.current; }
+
+    bool check_range (const CompositeGlyphChain *composite) const
+    {
+      return glyph.check_range (composite, CompositeGlyphChain::min_size)
+	  && glyph.check_range (composite, composite->get_size ());
+    }
+
+    private:
+    hb_bytes_t glyph;
+    __item_t__ current;
+  };
+
+  enum phantom_point_index_t
+  {
+    PHANTOM_LEFT   = 0,
+    PHANTOM_RIGHT  = 1,
+    PHANTOM_TOP    = 2,
+    PHANTOM_BOTTOM = 3,
+    PHANTOM_COUNT  = 4
+  };
+
+  struct accelerator_t;
+
+  struct Glyph
+  {
+    enum simple_glyph_flag_t
+    {
+      FLAG_ON_CURVE  = 0x01,
+      FLAG_X_SHORT   = 0x02,
+      FLAG_Y_SHORT   = 0x04,
+      FLAG_REPEAT    = 0x08,
+      FLAG_X_SAME    = 0x10,
+      FLAG_Y_SAME    = 0x20,
+      FLAG_RESERVED1 = 0x40,
+      FLAG_RESERVED2 = 0x80
+    };
+
+    private:
+    struct GlyphHeader
+    {
+      bool has_data () const { return numberOfContours; }
+
+      bool get_extents (hb_font_t *font, const accelerator_t &glyf_accelerator,
+		        hb_codepoint_t gid, hb_glyph_extents_t *extents) const
+      {
+	/* Undocumented rasterizer behavior: shift glyph to the left by (lsb - xMin), i.e., xMin = lsb */
+	/* extents->x_bearing = hb_min (glyph_header.xMin, glyph_header.xMax); */
+	extents->x_bearing = font->em_scale_x (glyf_accelerator.hmtx->get_side_bearing (gid));
+	extents->y_bearing = font->em_scale_y (hb_max (yMin, yMax));
+	extents->width     = font->em_scale_x (hb_max (xMin, xMax) - hb_min (xMin, xMax));
+	extents->height    = font->em_scale_y (hb_min (yMin, yMax) - hb_max (yMin, yMax));
+
+	return true;
+      }
+
+      HBINT16	numberOfContours;
+			/* If the number of contours is
+			 * greater than or equal to zero,
+			 * this is a simple glyph; if negative,
+			 * this is a composite glyph. */
+      FWORD	xMin;	/* Minimum x for coordinate data. */
+      FWORD	yMin;	/* Minimum y for coordinate data. */
+      FWORD	xMax;	/* Maximum x for coordinate data. */
+      FWORD	yMax;	/* Maximum y for coordinate data. */
+      public:
+      DEFINE_SIZE_STATIC (10);
+    };
+
+    struct SimpleGlyph
+    {
+      const GlyphHeader &header;
+      hb_bytes_t bytes;
+      SimpleGlyph (const GlyphHeader &header_, hb_bytes_t bytes_) :
+	header (header_), bytes (bytes_) {}
+
+      unsigned int instruction_len_offset () const
+      { return GlyphHeader::static_size + 2 * header.numberOfContours; }
+
+      unsigned int length (unsigned int instruction_len) const
+      { return instruction_len_offset () + 2 + instruction_len; }
+
+      unsigned int instructions_length () const
+      {
+	unsigned int instruction_length_offset = instruction_len_offset ();
+	if (unlikely (instruction_length_offset + 2 > bytes.length)) return 0;
+
+	const HBUINT16 &instructionLength = StructAtOffset<HBUINT16> (&bytes, instruction_length_offset);
+	/* Out of bounds of the current glyph */
+	if (unlikely (length (instructionLength) > bytes.length)) return 0;
+	return instructionLength;
+      }
+
+      const Glyph trim_padding () const
+      {
+	/* based on FontTools _g_l_y_f.py::trim */
+	const char *glyph = bytes.arrayZ;
+	const char *glyph_end = glyph + bytes.length;
+	/* simple glyph w/contours, possibly trimmable */
+	glyph += instruction_len_offset ();
+
+	if (unlikely (glyph + 2 >= glyph_end)) return Glyph ();
+	unsigned int num_coordinates = StructAtOffset<HBUINT16> (glyph - 2, 0) + 1;
+	unsigned int num_instructions = StructAtOffset<HBUINT16> (glyph, 0);
+
+	glyph += 2 + num_instructions;
+
+	unsigned int coord_bytes = 0;
+	unsigned int coords_with_flags = 0;
+	while (glyph < glyph_end)
+	{
+	  uint8_t flag = *glyph;
+	  glyph++;
+
+	  unsigned int repeat = 1;
+	  if (flag & FLAG_REPEAT)
+	  {
+	    if (unlikely (glyph >= glyph_end)) return Glyph ();
+	    repeat = *glyph + 1;
+	    glyph++;
+	  }
+
+	  unsigned int xBytes, yBytes;
+	  xBytes = yBytes = 0;
+	  if (flag & FLAG_X_SHORT) xBytes = 1;
+	  else if ((flag & FLAG_X_SAME) == 0) xBytes = 2;
+
+	  if (flag & FLAG_Y_SHORT) yBytes = 1;
+	  else if ((flag & FLAG_Y_SAME) == 0) yBytes = 2;
+
+	  coord_bytes += (xBytes + yBytes) * repeat;
+	  coords_with_flags += repeat;
+	  if (coords_with_flags >= num_coordinates) break;
+	}
+
+	if (unlikely (coords_with_flags != num_coordinates)) return Glyph ();
+	return Glyph (bytes.sub_array (0, bytes.length + coord_bytes - (glyph_end - glyph)));
+      }
+
+      /* zero instruction length */
+      void drop_hints ()
+      {
+	GlyphHeader &glyph_header = const_cast<GlyphHeader &> (header);
+	(HBUINT16 &) StructAtOffset<HBUINT16> (&glyph_header, instruction_len_offset ()) = 0;
+      }
+
+      void drop_hints_bytes (hb_bytes_t &dest_start, hb_bytes_t &dest_end) const
+      {
+	unsigned int instructions_len = instructions_length ();
+	unsigned int glyph_length = length (instructions_len);
+	dest_start = bytes.sub_array (0, glyph_length - instructions_len);
+	dest_end = bytes.sub_array (glyph_length, bytes.length - glyph_length);
+      }
+
+      static bool read_points (const HBUINT8 *&p /* IN/OUT */,
+			       contour_point_vector_t &points_ /* IN/OUT */,
+			       const hb_bytes_t &bytes,
+			       void (* setter) (contour_point_t &_, float v),
+			       const simple_glyph_flag_t short_flag,
+			       const simple_glyph_flag_t same_flag)
+      {
+	float v = 0;
+	for (unsigned i = 0; i < points_.length; i++)
+	{
+	  uint8_t flag = points_[i].flag;
+	  if (flag & short_flag)
+	  {
+	    if (unlikely (!bytes.check_range (p))) return false;
+	    if (flag & same_flag)
+	      v += *p++;
+	    else
+	      v -= *p++;
+	  }
+	  else
+	  {
+	    if (!(flag & same_flag))
+	    {
+	      if (unlikely (!bytes.check_range ((const HBUINT16 *) p))) return false;
+	      v += *(const HBINT16 *) p;
+	      p += HBINT16::static_size;
+	    }
+	  }
+	  setter (points_[i], v);
+	}
+	return true;
+      }
+
+      bool get_contour_points (contour_point_vector_t &points_ /* OUT */,
+			       bool phantom_only = false) const
+      {
+	const HBUINT16 *endPtsOfContours = &StructAfter<HBUINT16> (header);
+	int num_contours = header.numberOfContours;
+	if (unlikely (!bytes.check_range (&endPtsOfContours[num_contours + 1]))) return false;
+	unsigned int num_points = endPtsOfContours[num_contours - 1] + 1;
+
+	points_.resize (num_points);
+	for (unsigned int i = 0; i < points_.length; i++) points_[i].init ();
+	if (phantom_only) return true;
+
+	for (int i = 0; i < num_contours; i++)
+	  points_[endPtsOfContours[i]].is_end_point = true;
+
+	/* Skip instructions */
+	const HBUINT8 *p = &StructAtOffset<HBUINT8> (&endPtsOfContours[num_contours + 1],
+						     endPtsOfContours[num_contours]);
+
+	/* Read flags */
+	for (unsigned int i = 0; i < num_points; i++)
+	{
+	  if (unlikely (!bytes.check_range (p))) return false;
+	  uint8_t flag = *p++;
+	  points_[i].flag = flag;
+	  if (flag & FLAG_REPEAT)
+	  {
+	    if (unlikely (!bytes.check_range (p))) return false;
+	    unsigned int repeat_count = *p++;
+	    while ((repeat_count-- > 0) && (++i < num_points))
+	      points_[i].flag = flag;
+	  }
+	}
+
+	/* Read x & y coordinates */
+	return read_points (p, points_, bytes, [] (contour_point_t &p, float v) { p.x = v; },
+			    FLAG_X_SHORT, FLAG_X_SAME)
+	    && read_points (p, points_, bytes, [] (contour_point_t &p, float v) { p.y = v; },
+			    FLAG_Y_SHORT, FLAG_Y_SAME);
+      }
+    };
+
+    struct CompositeGlyph
+    {
+      const GlyphHeader &header;
+      hb_bytes_t bytes;
+      CompositeGlyph (const GlyphHeader &header_, hb_bytes_t bytes_) :
+	header (header_), bytes (bytes_) {}
+
+      composite_iter_t get_iterator () const
+      { return composite_iter_t (bytes, &StructAfter<CompositeGlyphChain, GlyphHeader> (header)); }
+
+      unsigned int instructions_length (hb_bytes_t bytes) const
+      {
+	unsigned int start = bytes.length;
+	unsigned int end = bytes.length;
+	const CompositeGlyphChain *last = nullptr;
+	for (auto &item : get_iterator ())
+	  last = &item;
+	if (unlikely (!last)) return 0;
+
+	if (last->has_instructions ())
+	  start = (char *) last - &bytes + last->get_size ();
+	if (unlikely (start > end)) return 0;
+	return end - start;
+      }
+
+      /* Trimming for composites not implemented.
+       * If removing hints it falls out of that. */
+      const Glyph trim_padding () const { return Glyph (bytes); }
+
+      void drop_hints ()
+      {
+	for (const auto &_ : get_iterator ())
+	  const_cast<CompositeGlyphChain &> (_).drop_instructions_flag ();
+      }
+
+      /* Chop instructions off the end */
+      void drop_hints_bytes (hb_bytes_t &dest_start) const
+      { dest_start = bytes.sub_array (0, bytes.length - instructions_length (bytes)); }
+    };
+
+    enum glyph_type_t { EMPTY, SIMPLE, COMPOSITE };
+
+    public:
+    composite_iter_t get_composite_iterator () const
+    {
+      if (type != COMPOSITE) return composite_iter_t ();
+      return CompositeGlyph (*header, bytes).get_iterator ();
+    }
+
+    const Glyph trim_padding () const
+    {
+      switch (type) {
+      case COMPOSITE: return CompositeGlyph (*header, bytes).trim_padding ();
+      case SIMPLE:    return SimpleGlyph (*header, bytes).trim_padding ();
+      default:        return bytes;
+      }
+    }
+
+    void drop_hints ()
+    {
+      switch (type) {
+      case COMPOSITE: CompositeGlyph (*header, bytes).drop_hints (); return;
+      case SIMPLE:    SimpleGlyph (*header, bytes).drop_hints (); return;
+      default:        return;
+      }
+    }
+
+    void drop_hints_bytes (hb_bytes_t &dest_start, hb_bytes_t &dest_end) const
+    {
+      switch (type) {
+      case COMPOSITE: CompositeGlyph (*header, bytes).drop_hints_bytes (dest_start); return;
+      case SIMPLE:    SimpleGlyph (*header, bytes).drop_hints_bytes (dest_start, dest_end); return;
+      default:        return;
+      }
+    }
+
+    /* Note: Recursively calls itself.
+     * all_points includes phantom points
+     */
+    bool get_points (hb_font_t *font, const accelerator_t &glyf_accelerator,
+		     contour_point_vector_t &all_points /* OUT */,
+		     bool phantom_only = false,
+		     unsigned int depth = 0) const
+    {
+      if (unlikely (depth > HB_MAX_NESTING_LEVEL)) return false;
+      contour_point_vector_t points;
+
+      switch (type) {
+      case COMPOSITE:
+      {
+	/* pseudo component points for each component in composite glyph */
+	unsigned num_points = hb_len (CompositeGlyph (*header, bytes).get_iterator ());
+	if (unlikely (!points.resize (num_points))) return false;
+	for (unsigned i = 0; i < points.length; i++)
+	  points[i].init ();
+	break;
+      }
+      case SIMPLE:
+	if (unlikely (!SimpleGlyph (*header, bytes).get_contour_points (points, phantom_only)))
+	  return false;
+	break;
+      }
+
+      /* Init phantom points */
+      if (unlikely (!points.resize (points.length + PHANTOM_COUNT))) return false;
+      hb_array_t<contour_point_t> phantoms = points.sub_array (points.length - PHANTOM_COUNT, PHANTOM_COUNT);
+      {
+	for (unsigned i = 0; i < PHANTOM_COUNT; ++i) phantoms[i].init ();
+	int h_delta = (int) header->xMin - glyf_accelerator.hmtx->get_side_bearing (gid);
+	int v_orig  = (int) header->yMax + glyf_accelerator.vmtx->get_side_bearing (gid);
+	unsigned h_adv = glyf_accelerator.hmtx->get_advance (gid);
+	unsigned v_adv = glyf_accelerator.vmtx->get_advance (gid);
+	phantoms[PHANTOM_LEFT].x = h_delta;
+	phantoms[PHANTOM_RIGHT].x = h_adv + h_delta;
+	phantoms[PHANTOM_TOP].y = v_orig;
+	phantoms[PHANTOM_BOTTOM].y = v_orig - (int) v_adv;
+      }
+
+#ifndef HB_NO_VAR
+      if (unlikely (!glyf_accelerator.gvar->apply_deltas_to_points (gid, font, points.as_array ())))
+	return false;
+#endif
+
+      switch (type) {
+      case SIMPLE:
+	all_points.extend (points.as_array ());
+	break;
+      case COMPOSITE:
+      {
+	unsigned int comp_index = 0;
+	for (auto &item : get_composite_iterator ())
+	{
+	  contour_point_vector_t comp_points;
+	  if (unlikely (!glyf_accelerator.glyph_for_gid (item.get_glyph_index ())
+					 .get_points (font, glyf_accelerator, comp_points,
+			 			      phantom_only, depth + 1)
+			|| comp_points.length < PHANTOM_COUNT))
+	    return false;
+
+	  /* Copy phantom points from component if USE_MY_METRICS flag set */
+	  if (item.is_use_my_metrics ())
+	    for (unsigned int i = 0; i < PHANTOM_COUNT; i++)
+	      phantoms[i] = comp_points[comp_points.length - PHANTOM_COUNT + i];
+
+	  /* Apply component transformation & translation */
+	  item.transform_points (comp_points);
+
+	  /* Apply translation from gvar */
+	  comp_points.translate (points[comp_index]);
+
+	  if (item.is_anchored ())
+	  {
+	    unsigned int p1, p2;
+	    item.get_anchor_points (p1, p2);
+	    if (likely (p1 < all_points.length && p2 < comp_points.length))
+	    {
+	      contour_point_t delta;
+	      delta.init (all_points[p1].x - comp_points[p2].x,
+			  all_points[p1].y - comp_points[p2].y);
+
+	      comp_points.translate (delta);
+	    }
+	  }
+
+	  all_points.extend (comp_points.sub_array (0, comp_points.length - PHANTOM_COUNT));
+
+	  comp_index++;
+	}
+
+	all_points.extend (phantoms);
+      } break;
+      default:
+	all_points.extend (phantoms);
+      }
+
+      if (depth == 0) /* Apply at top level */
+      {
+	/* Undocumented rasterizer behavior:
+	 * Shift points horizontally by the updated left side bearing
+	 */
+	contour_point_t delta;
+	delta.init (-phantoms[PHANTOM_LEFT].x, 0.f);
+	if (delta.x) all_points.translate (delta);
+      }
+
+      return true;
+    }
+
+    bool get_extents (hb_font_t *font, const accelerator_t &glyf_accelerator,
+		      hb_glyph_extents_t *extents) const
+    {
+      if (type == EMPTY) return true; /* Empty glyph; zero extents. */
+      return header->get_extents (font, glyf_accelerator, gid, extents);
+    }
+
+    hb_bytes_t get_bytes () const { return bytes; }
+
+    Glyph (hb_bytes_t bytes_ = hb_bytes_t (),
+	   hb_codepoint_t gid_ = (hb_codepoint_t) -1) : bytes (bytes_), gid (gid_),
+							header (bytes.as<GlyphHeader> ())
+    {
+      int num_contours = header->numberOfContours;
+      if (unlikely (num_contours == 0)) type = EMPTY;
+      else if (num_contours > 0) type = SIMPLE;
+      else type = COMPOSITE; /* negative numbers */
+    }
+
+    protected:
+    hb_bytes_t bytes;
+    hb_codepoint_t gid;
+    const GlyphHeader *header;
+    unsigned type;
+  };
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face_)
+    {
+      short_offset = false;
+      num_glyphs = 0;
+      loca_table = nullptr;
+      glyf_table = nullptr;
+#ifndef HB_NO_VAR
+      gvar = nullptr;
+#endif
+      hmtx = nullptr;
+      vmtx = nullptr;
+      face = face_;
+      const OT::head &head = *face->table.head;
+      if (head.indexToLocFormat > 1 || head.glyphDataFormat > 0)
+	/* Unknown format.  Leave num_glyphs=0, that takes care of disabling us. */
+	return;
+      short_offset = 0 == head.indexToLocFormat;
+
+      loca_table = hb_sanitize_context_t ().reference_table<loca> (face);
+      glyf_table = hb_sanitize_context_t ().reference_table<glyf> (face);
+#ifndef HB_NO_VAR
+      gvar = face->table.gvar;
+#endif
+      hmtx = face->table.hmtx;
+      vmtx = face->table.vmtx;
+
+      num_glyphs = hb_max (1u, loca_table.get_length () / (short_offset ? 2 : 4)) - 1;
+      num_glyphs = hb_min (num_glyphs, face->get_num_glyphs ());
+    }
+
+    void fini ()
+    {
+      loca_table.destroy ();
+      glyf_table.destroy ();
+    }
+
+    protected:
+    template<typename T>
+    bool get_points (hb_font_t *font, hb_codepoint_t gid, T consumer) const
+    {
+      if (gid >= num_glyphs) return false;
+
+      /* Making this alloc free is not that easy
+	 https://github.com/harfbuzz/harfbuzz/issues/2095
+	 mostly because of gvar handling in VF fonts,
+	 perhaps a separate path for non-VF fonts can be considered */
+      contour_point_vector_t all_points;
+
+      bool phantom_only = !consumer.is_consuming_contour_points ();
+      if (unlikely (!glyph_for_gid (gid).get_points (font, *this, all_points, phantom_only)))
+	return false;
+
+      if (consumer.is_consuming_contour_points ())
+      {
+	for (unsigned point_index = 0; point_index + 4 < all_points.length; ++point_index)
+	  consumer.consume_point (all_points[point_index]);
+	consumer.points_end ();
+      }
+
+      /* Where to write phantoms, nullptr if not requested */
+      contour_point_t *phantoms = consumer.get_phantoms_sink ();
+      if (phantoms)
+	for (unsigned i = 0; i < PHANTOM_COUNT; ++i)
+	  phantoms[i] = all_points[all_points.length - PHANTOM_COUNT + i];
+
+      return true;
+    }
+
+#ifndef HB_NO_VAR
+    struct points_aggregator_t
+    {
+      hb_font_t *font;
+      hb_glyph_extents_t *extents;
+      contour_point_t *phantoms;
+
+      struct contour_bounds_t
+      {
+	contour_bounds_t () { min_x = min_y = FLT_MAX; max_x = max_y = -FLT_MAX; }
+
+	void add (const contour_point_t &p)
+	{
+	  min_x = hb_min (min_x, p.x);
+	  min_y = hb_min (min_y, p.y);
+	  max_x = hb_max (max_x, p.x);
+	  max_y = hb_max (max_y, p.y);
+	}
+
+	bool empty () const { return (min_x >= max_x) || (min_y >= max_y); }
+
+	void get_extents (hb_font_t *font, hb_glyph_extents_t *extents)
+	{
+	  if (unlikely (empty ()))
+	  {
+	    extents->width = 0;
+	    extents->x_bearing = 0;
+	    extents->height = 0;
+	    extents->y_bearing = 0;
+	    return;
+	  }
+	  extents->x_bearing = font->em_scalef_x (min_x);
+	  extents->width = font->em_scalef_x (max_x - min_x);
+	  extents->y_bearing = font->em_scalef_y (max_y);
+	  extents->height = font->em_scalef_y (min_y - max_y);
+	}
+
+	protected:
+	float min_x, min_y, max_x, max_y;
+      } bounds;
+
+      points_aggregator_t (hb_font_t *font_, hb_glyph_extents_t *extents_, contour_point_t *phantoms_)
+      {
+	font = font_;
+	extents = extents_;
+	phantoms = phantoms_;
+	if (extents) bounds = contour_bounds_t ();
+      }
+
+      void consume_point (const contour_point_t &point) { bounds.add (point); }
+      void points_end () { bounds.get_extents (font, extents); }
+
+      bool is_consuming_contour_points () { return extents; }
+      contour_point_t *get_phantoms_sink () { return phantoms; }
+    };
+
+    public:
+    unsigned
+    get_advance_var (hb_font_t *font, hb_codepoint_t gid, bool is_vertical) const
+    {
+      if (unlikely (gid >= num_glyphs)) return 0;
+
+      bool success = false;
+
+      contour_point_t phantoms[PHANTOM_COUNT];
+      if (likely (font->num_coords == gvar->get_axis_count ()))
+	success = get_points (font, gid, points_aggregator_t (font, nullptr, phantoms));
+
+      if (unlikely (!success))
+	return is_vertical ? vmtx->get_advance (gid) : hmtx->get_advance (gid);
+
+      float result = is_vertical
+		   ? phantoms[PHANTOM_TOP].y - phantoms[PHANTOM_BOTTOM].y
+		   : phantoms[PHANTOM_RIGHT].x - phantoms[PHANTOM_LEFT].x;
+      return hb_clamp (roundf (result), 0.f, (float) UINT_MAX / 2);
+    }
+
+    int get_side_bearing_var (hb_font_t *font, hb_codepoint_t gid, bool is_vertical) const
+    {
+      if (unlikely (gid >= num_glyphs)) return 0;
+
+      hb_glyph_extents_t extents;
+
+      contour_point_t phantoms[PHANTOM_COUNT];
+      if (unlikely (!get_points (font, gid, points_aggregator_t (font, &extents, phantoms))))
+	return is_vertical ? vmtx->get_side_bearing (gid) : hmtx->get_side_bearing (gid);
+
+      return is_vertical
+	   ? ceilf (phantoms[PHANTOM_TOP].y) - extents.y_bearing
+	   : floorf (phantoms[PHANTOM_LEFT].x);
+    }
+#endif
+
+    public:
+    bool get_extents (hb_font_t *font, hb_codepoint_t gid, hb_glyph_extents_t *extents) const
+    {
+      if (unlikely (gid >= num_glyphs)) return false;
+
+#ifndef HB_NO_VAR
+      if (font->num_coords && font->num_coords == gvar->get_axis_count ())
+	return get_points (font, gid, points_aggregator_t (font, extents, nullptr));
+#endif
+      return glyph_for_gid (gid).get_extents (font, *this, extents);
+    }
+
+    const Glyph
+    glyph_for_gid (hb_codepoint_t gid, bool needs_padding_removal = false) const
+    {
+      if (unlikely (gid >= num_glyphs)) return Glyph ();
+
+      unsigned int start_offset, end_offset;
+
+      if (short_offset)
+      {
+	const HBUINT16 *offsets = (const HBUINT16 *) loca_table->dataZ.arrayZ;
+	start_offset = 2 * offsets[gid];
+	end_offset   = 2 * offsets[gid + 1];
+      }
+      else
+      {
+	const HBUINT32 *offsets = (const HBUINT32 *) loca_table->dataZ.arrayZ;
+	start_offset = offsets[gid];
+	end_offset   = offsets[gid + 1];
+      }
+
+      if (unlikely (start_offset > end_offset || end_offset > glyf_table.get_length ()))
+	return Glyph ();
+
+      Glyph glyph (hb_bytes_t ((const char *) this->glyf_table + start_offset,
+			       end_offset - start_offset), gid);
+      return needs_padding_removal ? glyph.trim_padding () : glyph;
+    }
+
+    void
+    add_gid_and_children (hb_codepoint_t gid, hb_set_t *gids_to_retain,
+			  unsigned int depth = 0) const
+    {
+      if (unlikely (depth++ > HB_MAX_NESTING_LEVEL)) return;
+      /* Check if is already visited */
+      if (gids_to_retain->has (gid)) return;
+
+      gids_to_retain->add (gid);
+
+      for (auto &item : glyph_for_gid (gid).get_composite_iterator ())
+	add_gid_and_children (item.get_glyph_index (), gids_to_retain, depth);
+    }
+
+#ifdef HB_EXPERIMENTAL_API
+    struct path_builder_t
+    {
+      hb_font_t *font;
+      draw_helper_t *draw_helper;
+
+      struct optional_point_t
+      {
+	optional_point_t () { has_data = false; }
+	optional_point_t (float x_, float y_) { x = x_; y = y_; has_data = true; }
+
+	bool has_data;
+	float x;
+	float y;
+
+	optional_point_t lerp (optional_point_t p, float t)
+	{ return optional_point_t (x + t * (p.x - x), y + t * (p.y - y)); }
+      } first_oncurve, first_offcurve, last_offcurve;
+
+      path_builder_t (hb_font_t *font_, draw_helper_t &draw_helper_)
+      {
+	font = font_;
+	draw_helper = &draw_helper_;
+	first_oncurve = first_offcurve = last_offcurve = optional_point_t ();
+      }
+
+      /* based on https://github.com/RazrFalcon/ttf-parser/blob/4f32821/src/glyf.rs#L287
+	 See also:
+	 * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM01/Chap1.html
+	 * https://stackoverflow.com/a/20772557 */
+      void consume_point (const contour_point_t &point)
+      {
+	/* Skip empty contours */
+	if (unlikely (point.is_end_point && !first_oncurve.has_data && !first_offcurve.has_data))
+	  return;
+
+	bool is_on_curve = point.flag & Glyph::FLAG_ON_CURVE;
+	optional_point_t p (point.x, point.y);
+	if (!first_oncurve.has_data)
+	{
+	  if (is_on_curve)
+	  {
+	    first_oncurve = p;
+	    draw_helper->move_to (font->em_scalef_x (p.x), font->em_scalef_y (p.y));
+	  }
+	  else
+	  {
+	    if (first_offcurve.has_data)
+	    {
+	      optional_point_t mid = first_offcurve.lerp (p, .5f);
+	      first_oncurve = mid;
+	      last_offcurve = p;
+	      draw_helper->move_to (font->em_scalef_x (mid.x), font->em_scalef_y (mid.y));
+	    }
+	    else
+	      first_offcurve = p;
+	  }
+	}
+	else
+	{
+	  if (last_offcurve.has_data)
+	  {
+	    if (is_on_curve)
+	    {
+	      draw_helper->quadratic_to (font->em_scalef_x (last_offcurve.x), font->em_scalef_y (last_offcurve.y),
+					 font->em_scalef_x (p.x), font->em_scalef_y (p.y));
+	      last_offcurve = optional_point_t ();
+	    }
+	    else
+	    {
+	      optional_point_t mid = last_offcurve.lerp (p, .5f);
+	      draw_helper->quadratic_to (font->em_scalef_x (last_offcurve.x), font->em_scalef_y (last_offcurve.y),
+					 font->em_scalef_x (mid.x), font->em_scalef_y (mid.y));
+	      last_offcurve = p;
+	    }
+	  }
+	  else
+	  {
+	    if (is_on_curve)
+	      draw_helper->line_to (font->em_scalef_x (p.x), font->em_scalef_y (p.y));
+	    else
+	      last_offcurve = p;
+	  }
+	}
+
+	if (point.is_end_point)
+	{
+	  if (first_offcurve.has_data && last_offcurve.has_data)
+	  {
+	    optional_point_t mid = last_offcurve.lerp (first_offcurve, .5f);
+	    draw_helper->quadratic_to (font->em_scalef_x (last_offcurve.x), font->em_scalef_y (last_offcurve.y),
+				       font->em_scalef_x (mid.x), font->em_scalef_y (mid.y));
+	    last_offcurve = optional_point_t ();
+	    /* now check the rest */
+	  }
+
+	  if (first_offcurve.has_data && first_oncurve.has_data)
+	    draw_helper->quadratic_to (font->em_scalef_x (first_offcurve.x), font->em_scalef_y (first_offcurve.y),
+				       font->em_scalef_x (first_oncurve.x), font->em_scalef_y (first_oncurve.y));
+	  else if (last_offcurve.has_data && first_oncurve.has_data)
+	    draw_helper->quadratic_to (font->em_scalef_x (last_offcurve.x), font->em_scalef_y (last_offcurve.y),
+				       font->em_scalef_x (first_oncurve.x), font->em_scalef_y (first_oncurve.y));
+	  else if (first_oncurve.has_data)
+	    draw_helper->line_to (font->em_scalef_x (first_oncurve.x), font->em_scalef_y (first_oncurve.y));
+
+	  /* Getting ready for the next contour */
+	  first_oncurve = first_offcurve = last_offcurve = optional_point_t ();
+	  draw_helper->end_path ();
+	}
+      }
+      void points_end () {}
+
+      bool is_consuming_contour_points () { return true; }
+      contour_point_t *get_phantoms_sink () { return nullptr; }
+    };
+
+    bool
+    get_path (hb_font_t *font, hb_codepoint_t gid, draw_helper_t &draw_helper) const
+    { return get_points (font, gid, path_builder_t (font, draw_helper)); }
+#endif
+
+#ifndef HB_NO_VAR
+    const gvar_accelerator_t *gvar;
+#endif
+    const hmtx_accelerator_t *hmtx;
+    const vmtx_accelerator_t *vmtx;
+
+    private:
+    bool short_offset;
+    unsigned int num_glyphs;
+    hb_blob_ptr_t<loca> loca_table;
+    hb_blob_ptr_t<glyf> glyf_table;
+    hb_face_t *face;
+  };
+
+  struct SubsetGlyph
+  {
+    hb_codepoint_t new_gid;
+    hb_codepoint_t old_gid;
+    Glyph source_glyph;
+    hb_bytes_t dest_start;  /* region of source_glyph to copy first */
+    hb_bytes_t dest_end;    /* region of source_glyph to copy second */
+
+    bool serialize (hb_serialize_context_t *c,
+		    const hb_subset_plan_t *plan) const
+    {
+      TRACE_SERIALIZE (this);
+
+      hb_bytes_t dest_glyph = dest_start.copy (c);
+      dest_glyph = hb_bytes_t (&dest_glyph, dest_glyph.length + dest_end.copy (c).length);
+      unsigned int pad_length = padding ();
+      DEBUG_MSG (SUBSET, nullptr, "serialize %d byte glyph, width %d pad %d", dest_glyph.length, dest_glyph.length + pad_length, pad_length);
+
+      HBUINT8 pad;
+      pad = 0;
+      while (pad_length > 0)
+      {
+	c->embed (pad);
+	pad_length--;
+      }
+
+      if (unlikely (!dest_glyph.length)) return_trace (true);
+
+      /* update components gids */
+      for (auto &_ : Glyph (dest_glyph).get_composite_iterator ())
+      {
+	hb_codepoint_t new_gid;
+	if (plan->new_gid_for_old_gid (_.get_glyph_index (), &new_gid))
+	  const_cast<CompositeGlyphChain &> (_).set_glyph_index (new_gid);
+      }
+
+      if (plan->drop_hints) Glyph (dest_glyph).drop_hints ();
+
+      return_trace (true);
+    }
+
+    void drop_hints_bytes ()
+    { source_glyph.drop_hints_bytes (dest_start, dest_end); }
+
+    unsigned int      length () const { return dest_start.length + dest_end.length; }
+    /* pad to 2 to ensure 2-byte loca will be ok */
+    unsigned int     padding () const { return length () % 2; }
+    unsigned int padded_size () const { return length () + padding (); }
+  };
+
+  protected:
+  UnsizedArrayOf<HBUINT8>
+		dataZ;	/* Glyphs data. */
+  public:
+  DEFINE_SIZE_MIN (0);	/* In reality, this is UNBOUNDED() type; but since we always
+			 * check the size externally, allow Null() object of it by
+			 * defining it _MIN instead. */
+};
+
+struct glyf_accelerator_t : glyf::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_GLYF_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-hdmx-table.hh b/thirdparty/harfbuzz/src/hb-ot-hdmx-table.hh
new file mode 100644
index 0000000000..c9c391bad5
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-hdmx-table.hh
@@ -0,0 +1,177 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger
+ */
+
+#ifndef HB_OT_HDMX_TABLE_HH
+#define HB_OT_HDMX_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * hdmx -- Horizontal Device Metrics
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/hdmx
+ */
+#define HB_OT_TAG_hdmx HB_TAG('h','d','m','x')
+
+
+namespace OT {
+
+
+struct DeviceRecord
+{
+  static unsigned int get_size (unsigned count)
+  { return hb_ceil_to_4 (min_size + count * HBUINT8::static_size); }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c, unsigned pixelSize, Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+
+    unsigned length = it.len ();
+
+    if (unlikely (!c->extend (*this, length)))  return_trace (false);
+
+    this->pixelSize = pixelSize;
+    this->maxWidth =
+    + it
+    | hb_reduce (hb_max, 0u);
+
+    + it
+    | hb_sink (widthsZ.as_array (length));
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned sizeDeviceRecord) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  c->check_range (this, sizeDeviceRecord)));
+  }
+
+  HBUINT8			pixelSize;	/* Pixel size for following widths (as ppem). */
+  HBUINT8			maxWidth;	/* Maximum width. */
+  UnsizedArrayOf<HBUINT8>	widthsZ;	/* Array of widths (numGlyphs is from the 'maxp' table). */
+  public:
+  DEFINE_SIZE_ARRAY (2, widthsZ);
+};
+
+
+struct hdmx
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_hdmx;
+
+  unsigned int get_size () const
+  { return min_size + numRecords * sizeDeviceRecord; }
+
+  const DeviceRecord& operator [] (unsigned int i) const
+  {
+    /* XXX Null(DeviceRecord) is NOT safe as it's num-glyphs lengthed.
+     * https://github.com/harfbuzz/harfbuzz/issues/1300 */
+    if (unlikely (i >= numRecords)) return Null (DeviceRecord);
+    return StructAtOffset<DeviceRecord> (&this->firstDeviceRecord, i * sizeDeviceRecord);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c, unsigned version, Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+
+    if (unlikely (!c->extend_min ((*this))))  return_trace (false);
+
+    this->version = version;
+    this->numRecords = it.len ();
+    this->sizeDeviceRecord = DeviceRecord::get_size (it ? (*it).second.len () : 0);
+
+    for (const hb_item_type<Iterator>& _ : +it)
+      c->start_embed<DeviceRecord> ()->serialize (c, _.first, _.second);
+
+    return_trace (c->successful);
+  }
+
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    hdmx *hdmx_prime = c->serializer->start_embed <hdmx> ();
+    if (unlikely (!hdmx_prime)) return_trace (false);
+
+    auto it =
+    + hb_range ((unsigned) numRecords)
+    | hb_map ([c, this] (unsigned _)
+	{
+	  const DeviceRecord *device_record =
+	    &StructAtOffset<DeviceRecord> (&firstDeviceRecord,
+					   _ * sizeDeviceRecord);
+	  auto row =
+	    + hb_range (c->plan->num_output_glyphs ())
+	    | hb_map (c->plan->reverse_glyph_map)
+	    | hb_map ([this, c, device_record] (hb_codepoint_t _)
+		      {
+			if (c->plan->is_empty_glyph (_))
+			  return Null (HBUINT8);
+			return device_record->widthsZ.as_array (get_num_glyphs ()) [_];
+		      })
+	    ;
+	  return hb_pair ((unsigned) device_record->pixelSize, +row);
+	})
+    ;
+
+    hdmx_prime->serialize (c->serializer, version, it);
+    return_trace (true);
+  }
+
+  unsigned get_num_glyphs () const
+  {
+    return sizeDeviceRecord - DeviceRecord::min_size;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  !hb_unsigned_mul_overflows (numRecords, sizeDeviceRecord) &&
+		  sizeDeviceRecord >= DeviceRecord::min_size &&
+		  c->check_range (this, get_size ()));
+  }
+
+  protected:
+  HBUINT16	version;	/* Table version number (0) */
+  HBUINT16	numRecords;	/* Number of device records. */
+  HBUINT32	sizeDeviceRecord;
+				/* Size of a device record, 32-bit aligned. */
+  DeviceRecord	firstDeviceRecord;
+				/* Array of device records. */
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_HDMX_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-head-table.hh b/thirdparty/harfbuzz/src/hb-ot-head-table.hh
new file mode 100644
index 0000000000..5613a96dbf
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-head-table.hh
@@ -0,0 +1,179 @@
+/*
+ * Copyright © 2010  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_HEAD_TABLE_HH
+#define HB_OT_HEAD_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * head -- Font Header
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/head
+ */
+#define HB_OT_TAG_head HB_TAG('h','e','a','d')
+
+
+namespace OT {
+
+
+struct head
+{
+  friend struct OffsetTable;
+
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_head;
+
+  unsigned int get_upem () const
+  {
+    unsigned int upem = unitsPerEm;
+    /* If no valid head table found, assume 1000, which matches typical Type1 usage. */
+    return 16 <= upem && upem <= 16384 ? upem : 1000;
+  }
+
+  bool serialize (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace ((bool) c->embed (this));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    return_trace (serialize (c->serializer));
+  }
+
+  enum mac_style_flag_t {
+    BOLD	= 1u<<0,
+    ITALIC	= 1u<<1,
+    UNDERLINE	= 1u<<2,
+    OUTLINE	= 1u<<3,
+    SHADOW	= 1u<<4,
+    CONDENSED	= 1u<<5
+  };
+
+  bool is_bold () const      { return macStyle & BOLD; }
+  bool is_italic () const    { return macStyle & ITALIC; }
+  bool is_condensed () const { return macStyle & CONDENSED; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  version.major == 1 &&
+		  magicNumber == 0x5F0F3CF5u);
+  }
+
+  protected:
+  FixedVersion<>version;		/* Version of the head table--currently
+					 * 0x00010000u for version 1.0. */
+  FixedVersion<>fontRevision;		/* Set by font manufacturer. */
+  HBUINT32	checkSumAdjustment;	/* To compute: set it to 0, sum the
+					 * entire font as HBUINT32, then store
+					 * 0xB1B0AFBAu - sum. */
+  HBUINT32	magicNumber;		/* Set to 0x5F0F3CF5u. */
+  HBUINT16	flags;			/* Bit 0: Baseline for font at y=0;
+					 * Bit 1: Left sidebearing point at x=0;
+					 * Bit 2: Instructions may depend on point size;
+					 * Bit 3: Force ppem to integer values for all
+					 *   internal scaler math; may use fractional
+					 *   ppem sizes if this bit is clear;
+					 * Bit 4: Instructions may alter advance width
+					 *   (the advance widths might not scale linearly);
+					 * Bits 5-10: These should be set according to
+					 *   Apple's specification. However, they are not
+					 *   implemented in OpenType.
+					 * Bit 5: This bit should be set in fonts that are
+					 *   intended to e laid out vertically, and in
+					 *   which the glyphs have been drawn such that an
+					 *   x-coordinate of 0 corresponds to the desired
+					 *   vertical baseline.
+					 * Bit 6: This bit must be set to zero.
+					 * Bit 7: This bit should be set if the font
+					 *   requires layout for correct linguistic
+					 *   rendering (e.g. Arabic fonts).
+					 * Bit 8: This bit should be set for a GX font
+					 *   which has one or more metamorphosis effects
+					 *   designated as happening by default.
+					 * Bit 9: This bit should be set if the font
+					 *   contains any strong right-to-left glyphs.
+					 * Bit 10: This bit should be set if the font
+					 *   contains Indic-style rearrangement effects.
+					 * Bit 11: Font data is 'lossless,' as a result
+					 *   of having been compressed and decompressed
+					 *   with the Agfa MicroType Express engine.
+					 * Bit 12: Font converted (produce compatible metrics)
+					 * Bit 13: Font optimized for ClearType™.
+					 *   Note, fonts that rely on embedded bitmaps (EBDT)
+					 *   for rendering should not be considered optimized
+					 *   for ClearType, and therefore should keep this bit
+					 *   cleared.
+					 * Bit 14: Last Resort font. If set, indicates that
+					 * the glyphs encoded in the cmap subtables are simply
+					 * generic symbolic representations of code point
+					 * ranges and don’t truly represent support for those
+					 * code points. If unset, indicates that the glyphs
+					 * encoded in the cmap subtables represent proper
+					 * support for those code points.
+					 * Bit 15: Reserved, set to 0. */
+  HBUINT16	unitsPerEm;		/* Valid range is from 16 to 16384. This value
+					 * should be a power of 2 for fonts that have
+					 * TrueType outlines. */
+  LONGDATETIME	created;		/* Number of seconds since 12:00 midnight,
+					   January 1, 1904. 64-bit integer */
+  LONGDATETIME	modified;		/* Number of seconds since 12:00 midnight,
+					   January 1, 1904. 64-bit integer */
+  HBINT16	xMin;			/* For all glyph bounding boxes. */
+  HBINT16	yMin;			/* For all glyph bounding boxes. */
+  HBINT16	xMax;			/* For all glyph bounding boxes. */
+  HBINT16	yMax;			/* For all glyph bounding boxes. */
+  HBUINT16	macStyle;		/* Bit 0: Bold (if set to 1);
+					 * Bit 1: Italic (if set to 1)
+					 * Bit 2: Underline (if set to 1)
+					 * Bit 3: Outline (if set to 1)
+					 * Bit 4: Shadow (if set to 1)
+					 * Bit 5: Condensed (if set to 1)
+					 * Bit 6: Extended (if set to 1)
+					 * Bits 7-15: Reserved (set to 0). */
+  HBUINT16	lowestRecPPEM;		/* Smallest readable size in pixels. */
+  HBINT16	fontDirectionHint;	/* Deprecated (Set to 2).
+					 * 0: Fully mixed directional glyphs;
+					 * 1: Only strongly left to right;
+					 * 2: Like 1 but also contains neutrals;
+					 * -1: Only strongly right to left;
+					 * -2: Like -1 but also contains neutrals. */
+  public:
+  HBUINT16	indexToLocFormat;	/* 0 for short offsets, 1 for long. */
+  HBUINT16	glyphDataFormat;	/* 0 for current format. */
+
+  DEFINE_SIZE_STATIC (54);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_HEAD_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-hhea-table.hh b/thirdparty/harfbuzz/src/hb-ot-hhea-table.hh
new file mode 100644
index 0000000000..d9c9bd3537
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-hhea-table.hh
@@ -0,0 +1,104 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_HHEA_TABLE_HH
+#define HB_OT_HHEA_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * hhea -- Horizontal Header
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/hhea
+ * vhea -- Vertical Header
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/vhea
+ */
+#define HB_OT_TAG_hhea HB_TAG('h','h','e','a')
+#define HB_OT_TAG_vhea HB_TAG('v','h','e','a')
+
+
+namespace OT {
+
+
+template <typename T>
+struct _hea
+{
+  bool has_data () const { return version.major; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && likely (version.major == 1));
+  }
+
+  public:
+  FixedVersion<>version;	/* 0x00010000u for version 1.0. */
+  FWORD		ascender;	/* Typographic ascent. */
+  FWORD		descender;	/* Typographic descent. */
+  FWORD		lineGap;	/* Typographic line gap. */
+  UFWORD	advanceMax;	/* Maximum advance width/height value in
+				 * metrics table. */
+  FWORD		minLeadingBearing;
+				/* Minimum left/top sidebearing value in
+				 * metrics table. */
+  FWORD		minTrailingBearing;
+				/* Minimum right/bottom sidebearing value;
+				 * calculated as Min(aw - lsb -
+				 * (xMax - xMin)) for horizontal. */
+  FWORD		maxExtent;	/* horizontal: Max(lsb + (xMax - xMin)),
+				 * vertical: minLeadingBearing+(yMax-yMin). */
+  HBINT16	caretSlopeRise;	/* Used to calculate the slope of the
+				 * cursor (rise/run); 1 for vertical caret,
+				 * 0 for horizontal.*/
+  HBINT16	caretSlopeRun;	/* 0 for vertical caret, 1 for horizontal. */
+  HBINT16	caretOffset;	/* The amount by which a slanted
+				 * highlight on a glyph needs
+				 * to be shifted to produce the
+				 * best appearance. Set to 0 for
+				 * non-slanted fonts. */
+  HBINT16	reserved1;	/* Set to 0. */
+  HBINT16	reserved2;	/* Set to 0. */
+  HBINT16	reserved3;	/* Set to 0. */
+  HBINT16	reserved4;	/* Set to 0. */
+  HBINT16	metricDataFormat;/* 0 for current format. */
+  HBUINT16	numberOfLongMetrics;
+				/* Number of LongMetric entries in metric
+				 * table. */
+  public:
+  DEFINE_SIZE_STATIC (36);
+};
+
+struct hhea : _hea<hhea> {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_hhea;
+};
+struct vhea : _hea<vhea> {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_vhea;
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_HHEA_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-hmtx-table.hh b/thirdparty/harfbuzz/src/hb-ot-hmtx-table.hh
new file mode 100644
index 0000000000..d06c0fa4a4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-hmtx-table.hh
@@ -0,0 +1,340 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod, Roderick Sheeter
+ */
+
+#ifndef HB_OT_HMTX_TABLE_HH
+#define HB_OT_HMTX_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-hhea-table.hh"
+#include "hb-ot-var-hvar-table.hh"
+#include "hb-ot-metrics.hh"
+
+/*
+ * hmtx -- Horizontal Metrics
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/hmtx
+ * vmtx -- Vertical Metrics
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/vmtx
+ */
+#define HB_OT_TAG_hmtx HB_TAG('h','m','t','x')
+#define HB_OT_TAG_vmtx HB_TAG('v','m','t','x')
+
+
+HB_INTERNAL int
+_glyf_get_side_bearing_var (hb_font_t *font, hb_codepoint_t glyph, bool is_vertical);
+
+HB_INTERNAL unsigned
+_glyf_get_advance_var (hb_font_t *font, hb_codepoint_t glyph, bool is_vertical);
+
+
+namespace OT {
+
+
+struct LongMetric
+{
+  UFWORD	advance; /* Advance width/height. */
+  FWORD		sb; /* Leading (left/top) side bearing. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+
+template <typename T, typename H>
+struct hmtxvmtx
+{
+  bool sanitize (hb_sanitize_context_t *c HB_UNUSED) const
+  {
+    TRACE_SANITIZE (this);
+    /* We don't check for anything specific here.  The users of the
+     * struct do all the hard work... */
+    return_trace (true);
+  }
+
+
+  bool subset_update_header (hb_subset_plan_t *plan,
+			     unsigned int num_hmetrics) const
+  {
+    hb_blob_t *src_blob = hb_sanitize_context_t ().reference_table<H> (plan->source, H::tableTag);
+    hb_blob_t *dest_blob = hb_blob_copy_writable_or_fail (src_blob);
+    hb_blob_destroy (src_blob);
+
+    if (unlikely (!dest_blob)) {
+      return false;
+    }
+
+    unsigned int length;
+    H *table = (H *) hb_blob_get_data (dest_blob, &length);
+    table->numberOfLongMetrics = num_hmetrics;
+
+    bool result = plan->add_table (H::tableTag, dest_blob);
+    hb_blob_destroy (dest_blob);
+
+    return result;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  unsigned num_advances)
+  {
+    unsigned idx = 0;
+    for (auto _ : it)
+    {
+      if (idx < num_advances)
+      {
+	LongMetric lm;
+	lm.advance = _.first;
+	lm.sb = _.second;
+	if (unlikely (!c->embed<LongMetric> (&lm))) return;
+      }
+      else
+      {
+	FWORD *sb = c->allocate_size<FWORD> (FWORD::static_size);
+	if (unlikely (!sb)) return;
+	*sb = _.second;
+      }
+      idx++;
+    }
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    T *table_prime = c->serializer->start_embed <T> ();
+    if (unlikely (!table_prime)) return_trace (false);
+
+    accelerator_t _mtx;
+    _mtx.init (c->plan->source);
+    unsigned num_advances = _mtx.num_advances_for_subset (c->plan);
+
+    auto it =
+    + hb_range (c->plan->num_output_glyphs ())
+    | hb_map ([c, &_mtx] (unsigned _)
+	      {
+		hb_codepoint_t old_gid;
+		if (!c->plan->old_gid_for_new_gid (_, &old_gid))
+		  return hb_pair (0u, 0);
+		return hb_pair (_mtx.get_advance (old_gid), _mtx.get_side_bearing (old_gid));
+	      })
+    ;
+
+    table_prime->serialize (c->serializer, it, num_advances);
+
+    _mtx.fini ();
+
+    if (unlikely (c->serializer->ran_out_of_room || c->serializer->in_error ()))
+      return_trace (false);
+
+    // Amend header num hmetrics
+    if (unlikely (!subset_update_header (c->plan, num_advances)))
+      return_trace (false);
+
+    return_trace (true);
+  }
+
+  struct accelerator_t
+  {
+    friend struct hmtxvmtx;
+
+    void init (hb_face_t *face,
+	       unsigned int default_advance_ = 0)
+    {
+      default_advance = default_advance_ ? default_advance_ : hb_face_get_upem (face);
+
+      num_advances = T::is_horizontal ? face->table.hhea->numberOfLongMetrics : face->table.vhea->numberOfLongMetrics;
+
+      table = hb_sanitize_context_t ().reference_table<hmtxvmtx> (face, T::tableTag);
+
+      /* Cap num_metrics() and num_advances() based on table length. */
+      unsigned int len = table.get_length ();
+      if (unlikely (num_advances * 4 > len))
+	num_advances = len / 4;
+      num_metrics = num_advances + (len - 4 * num_advances) / 2;
+
+      /* We MUST set num_metrics to zero if num_advances is zero.
+       * Our get_advance() depends on that. */
+      if (unlikely (!num_advances))
+      {
+	num_metrics = num_advances = 0;
+	table.destroy ();
+	table = hb_blob_get_empty ();
+      }
+
+      var_table = hb_sanitize_context_t ().reference_table<HVARVVAR> (face, T::variationsTag);
+    }
+
+    void fini ()
+    {
+      table.destroy ();
+      var_table.destroy ();
+    }
+
+    int get_side_bearing (hb_codepoint_t glyph) const
+    {
+      if (glyph < num_advances)
+	return table->longMetricZ[glyph].sb;
+
+      if (unlikely (glyph >= num_metrics))
+	return 0;
+
+      const FWORD *bearings = (const FWORD *) &table->longMetricZ[num_advances];
+      return bearings[glyph - num_advances];
+    }
+
+    int get_side_bearing (hb_font_t *font, hb_codepoint_t glyph) const
+    {
+      int side_bearing = get_side_bearing (glyph);
+
+#ifndef HB_NO_VAR
+      if (unlikely (glyph >= num_metrics) || !font->num_coords)
+	return side_bearing;
+
+      if (var_table.get_length ())
+	return side_bearing + var_table->get_side_bearing_var (glyph, font->coords, font->num_coords); // TODO Optimize?!
+
+      return _glyf_get_side_bearing_var (font, glyph, T::tableTag == HB_OT_TAG_vmtx);
+#else
+      return side_bearing;
+#endif
+    }
+
+    unsigned int get_advance (hb_codepoint_t glyph) const
+    {
+      if (unlikely (glyph >= num_metrics))
+      {
+	/* If num_metrics is zero, it means we don't have the metrics table
+	 * for this direction: return default advance.  Otherwise, it means that the
+	 * glyph index is out of bound: return zero. */
+	if (num_metrics)
+	  return 0;
+	else
+	  return default_advance;
+      }
+
+      return table->longMetricZ[hb_min (glyph, (uint32_t) num_advances - 1)].advance;
+    }
+
+    unsigned int get_advance (hb_codepoint_t  glyph,
+			      hb_font_t      *font) const
+    {
+      unsigned int advance = get_advance (glyph);
+
+#ifndef HB_NO_VAR
+      if (unlikely (glyph >= num_metrics) || !font->num_coords)
+	return advance;
+
+      if (var_table.get_length ())
+	return advance + roundf (var_table->get_advance_var (glyph, font)); // TODO Optimize?!
+
+      return _glyf_get_advance_var (font, glyph, T::tableTag == HB_OT_TAG_vmtx);
+#else
+      return advance;
+#endif
+    }
+
+    unsigned int num_advances_for_subset (const hb_subset_plan_t *plan) const
+    {
+      unsigned int num_advances = plan->num_output_glyphs ();
+      unsigned int last_advance = _advance_for_new_gid (plan,
+							num_advances - 1);
+      while (num_advances > 1 &&
+	     last_advance == _advance_for_new_gid (plan,
+						   num_advances - 2))
+      {
+	num_advances--;
+      }
+
+      return num_advances;
+    }
+
+    private:
+    unsigned int _advance_for_new_gid (const hb_subset_plan_t *plan,
+				       hb_codepoint_t new_gid) const
+    {
+      hb_codepoint_t old_gid;
+      if (!plan->old_gid_for_new_gid (new_gid, &old_gid))
+	return 0;
+
+      return get_advance (old_gid);
+    }
+
+    protected:
+    unsigned int num_metrics;
+    unsigned int num_advances;
+    unsigned int default_advance;
+
+    private:
+    hb_blob_ptr_t<hmtxvmtx> table;
+    hb_blob_ptr_t<HVARVVAR> var_table;
+  };
+
+  protected:
+  UnsizedArrayOf<LongMetric>
+		longMetricZ;	/* Paired advance width and leading
+				 * bearing values for each glyph. The
+				 * value numOfHMetrics comes from
+				 * the 'hhea' table. If the font is
+				 * monospaced, only one entry need
+				 * be in the array, but that entry is
+				 * required. The last entry applies to
+				 * all subsequent glyphs. */
+/*UnsizedArrayOf<FWORD>	leadingBearingX;*/
+				/* Here the advance is assumed
+				 * to be the same as the advance
+				 * for the last entry above. The
+				 * number of entries in this array is
+				 * derived from numGlyphs (from 'maxp'
+				 * table) minus numberOfLongMetrics.
+				 * This generally is used with a run
+				 * of monospaced glyphs (e.g., Kanji
+				 * fonts or Courier fonts). Only one
+				 * run is allowed and it must be at
+				 * the end. This allows a monospaced
+				 * font to vary the side bearing
+				 * values for each glyph. */
+  public:
+  DEFINE_SIZE_ARRAY (0, longMetricZ);
+};
+
+struct hmtx : hmtxvmtx<hmtx, hhea> {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_hmtx;
+  static constexpr hb_tag_t variationsTag = HB_OT_TAG_HVAR;
+  static constexpr bool is_horizontal = true;
+};
+struct vmtx : hmtxvmtx<vmtx, vhea> {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_vmtx;
+  static constexpr hb_tag_t variationsTag = HB_OT_TAG_VVAR;
+  static constexpr bool is_horizontal = false;
+};
+
+struct hmtx_accelerator_t : hmtx::accelerator_t {};
+struct vmtx_accelerator_t : vmtx::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_HMTX_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-kern-table.hh b/thirdparty/harfbuzz/src/hb-ot-kern-table.hh
new file mode 100644
index 0000000000..3563cab8bd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-kern-table.hh
@@ -0,0 +1,359 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_KERN_TABLE_HH
+#define HB_OT_KERN_TABLE_HH
+
+#include "hb-aat-layout-kerx-table.hh"
+
+
+/*
+ * kern -- Kerning
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/kern
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6kern.html
+ */
+#define HB_OT_TAG_kern HB_TAG('k','e','r','n')
+
+
+namespace OT {
+
+
+template <typename KernSubTableHeader>
+struct KernSubTableFormat3
+{
+  int get_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+  {
+    hb_array_t<const FWORD> kernValue = kernValueZ.as_array (kernValueCount);
+    hb_array_t<const HBUINT8> leftClass = StructAfter<const UnsizedArrayOf<HBUINT8>> (kernValue).as_array (glyphCount);
+    hb_array_t<const HBUINT8> rightClass = StructAfter<const UnsizedArrayOf<HBUINT8>> (leftClass).as_array (glyphCount);
+    hb_array_t<const HBUINT8> kernIndex = StructAfter<const UnsizedArrayOf<HBUINT8>> (rightClass).as_array (leftClassCount * rightClassCount);
+
+    unsigned int leftC = leftClass[left];
+    unsigned int rightC = rightClass[right];
+    if (unlikely (leftC >= leftClassCount || rightC >= rightClassCount))
+      return 0;
+    unsigned int i = leftC * rightClassCount + rightC;
+    return kernValue[kernIndex[i]];
+  }
+
+  bool apply (AAT::hb_aat_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    if (!c->plan->requested_kerning)
+      return false;
+
+    if (header.coverage & header.Backwards)
+      return false;
+
+    hb_kern_machine_t<KernSubTableFormat3> machine (*this, header.coverage & header.CrossStream);
+    machine.kern (c->font, c->buffer, c->plan->kern_mask);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  c->check_range (kernValueZ,
+				  kernValueCount * sizeof (FWORD) +
+				  glyphCount * 2 +
+				  leftClassCount * rightClassCount));
+  }
+
+  protected:
+  KernSubTableHeader
+		header;
+  HBUINT16	glyphCount;	/* The number of glyphs in this font. */
+  HBUINT8	kernValueCount;	/* The number of kerning values. */
+  HBUINT8	leftClassCount;	/* The number of left-hand classes. */
+  HBUINT8	rightClassCount;/* The number of right-hand classes. */
+  HBUINT8	flags;		/* Set to zero (reserved for future use). */
+  UnsizedArrayOf<FWORD>
+		kernValueZ;	/* The kerning values.
+				 * Length kernValueCount. */
+#if 0
+  UnsizedArrayOf<HBUINT8>
+		leftClass;	/* The left-hand classes.
+				 * Length glyphCount. */
+  UnsizedArrayOf<HBUINT8>
+		rightClass;	/* The right-hand classes.
+				 * Length glyphCount. */
+  UnsizedArrayOf<HBUINT8>kernIndex;
+				/* The indices into the kernValue array.
+				 * Length leftClassCount * rightClassCount */
+#endif
+  public:
+  DEFINE_SIZE_ARRAY (KernSubTableHeader::static_size + 6, kernValueZ);
+};
+
+template <typename KernSubTableHeader>
+struct KernSubTable
+{
+  unsigned int get_size () const { return u.header.length; }
+  unsigned int get_type () const { return u.header.format; }
+
+  int get_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+  {
+    switch (get_type ()) {
+    /* This method hooks up to hb_font_t's get_h_kerning.  Only support Format0. */
+    case 0: return u.format0.get_kerning (left, right);
+    default:return 0;
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    unsigned int subtable_type = get_type ();
+    TRACE_DISPATCH (this, subtable_type);
+    switch (subtable_type) {
+    case 0:	return_trace (c->dispatch (u.format0));
+#ifndef HB_NO_AAT_SHAPE
+    case 1:	return_trace (u.header.apple ? c->dispatch (u.format1, hb_forward<Ts> (ds)...) : c->default_return_value ());
+#endif
+    case 2:	return_trace (c->dispatch (u.format2));
+#ifndef HB_NO_AAT_SHAPE
+    case 3:	return_trace (u.header.apple ? c->dispatch (u.format3, hb_forward<Ts> (ds)...) : c->default_return_value ());
+#endif
+    default:	return_trace (c->default_return_value ());
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!u.header.sanitize (c) ||
+		  u.header.length < u.header.min_size ||
+		  !c->check_range (this, u.header.length))) return_trace (false);
+
+    return_trace (dispatch (c));
+  }
+
+  public:
+  union {
+  KernSubTableHeader				header;
+  AAT::KerxSubTableFormat0<KernSubTableHeader>	format0;
+  AAT::KerxSubTableFormat1<KernSubTableHeader>	format1;
+  AAT::KerxSubTableFormat2<KernSubTableHeader>	format2;
+  KernSubTableFormat3<KernSubTableHeader>	format3;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (KernSubTableHeader::static_size);
+};
+
+
+struct KernOTSubTableHeader
+{
+  static constexpr bool apple = false;
+  typedef AAT::ObsoleteTypes Types;
+
+  unsigned   tuple_count () const { return 0; }
+  bool     is_horizontal () const { return (coverage & Horizontal); }
+
+  enum Coverage
+  {
+    Horizontal	= 0x01u,
+    Minimum	= 0x02u,
+    CrossStream	= 0x04u,
+    Override	= 0x08u,
+
+    /* Not supported: */
+    Backwards	= 0x00u,
+    Variation	= 0x00u,
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	versionZ;	/* Unused. */
+  HBUINT16	length;		/* Length of the subtable (including this header). */
+  HBUINT8	format;		/* Subtable format. */
+  HBUINT8	coverage;	/* Coverage bits. */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct KernOT : AAT::KerxTable<KernOT>
+{
+  friend struct AAT::KerxTable<KernOT>;
+
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_kern;
+  static constexpr unsigned minVersion = 0u;
+
+  typedef KernOTSubTableHeader SubTableHeader;
+  typedef SubTableHeader::Types Types;
+  typedef KernSubTable<SubTableHeader> SubTable;
+
+  protected:
+  HBUINT16	version;	/* Version--0x0000u */
+  HBUINT16	tableCount;	/* Number of subtables in the kerning table. */
+  SubTable	firstSubTable;	/* Subtables. */
+  public:
+  DEFINE_SIZE_MIN (4);
+};
+
+
+struct KernAATSubTableHeader
+{
+  static constexpr bool apple = true;
+  typedef AAT::ObsoleteTypes Types;
+
+  unsigned   tuple_count () const { return 0; }
+  bool     is_horizontal () const { return !(coverage & Vertical); }
+
+  enum Coverage
+  {
+    Vertical	= 0x80u,
+    CrossStream	= 0x40u,
+    Variation	= 0x20u,
+
+    /* Not supported: */
+    Backwards	= 0x00u,
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT32	length;		/* Length of the subtable (including this header). */
+  HBUINT8	coverage;	/* Coverage bits. */
+  HBUINT8	format;		/* Subtable format. */
+  HBUINT16	tupleIndex;	/* The tuple index (used for variations fonts).
+				 * This value specifies which tuple this subtable covers.
+				 * Note: We don't implement. */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct KernAAT : AAT::KerxTable<KernAAT>
+{
+  friend struct AAT::KerxTable<KernAAT>;
+
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_kern;
+  static constexpr unsigned minVersion = 0x00010000u;
+
+  typedef KernAATSubTableHeader SubTableHeader;
+  typedef SubTableHeader::Types Types;
+  typedef KernSubTable<SubTableHeader> SubTable;
+
+  protected:
+  HBUINT32	version;	/* Version--0x00010000u */
+  HBUINT32	tableCount;	/* Number of subtables in the kerning table. */
+  SubTable	firstSubTable;	/* Subtables. */
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+struct kern
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_kern;
+
+  bool     has_data () const { return u.version32; }
+  unsigned get_type () const { return u.major; }
+
+  bool has_state_machine () const
+  {
+    switch (get_type ()) {
+    case 0: return u.ot.has_state_machine ();
+#ifndef HB_NO_AAT_SHAPE
+    case 1: return u.aat.has_state_machine ();
+#endif
+    default:return false;
+    }
+  }
+
+  bool has_cross_stream () const
+  {
+    switch (get_type ()) {
+    case 0: return u.ot.has_cross_stream ();
+#ifndef HB_NO_AAT_SHAPE
+    case 1: return u.aat.has_cross_stream ();
+#endif
+    default:return false;
+    }
+  }
+
+  int get_h_kerning (hb_codepoint_t left, hb_codepoint_t right) const
+  {
+    switch (get_type ()) {
+    case 0: return u.ot.get_h_kerning (left, right);
+#ifndef HB_NO_AAT_SHAPE
+    case 1: return u.aat.get_h_kerning (left, right);
+#endif
+    default:return 0;
+    }
+  }
+
+  bool apply (AAT::hb_aat_apply_context_t *c) const
+  { return dispatch (c); }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    unsigned int subtable_type = get_type ();
+    TRACE_DISPATCH (this, subtable_type);
+    switch (subtable_type) {
+    case 0:	return_trace (c->dispatch (u.ot, hb_forward<Ts> (ds)...));
+#ifndef HB_NO_AAT_SHAPE
+    case 1:	return_trace (c->dispatch (u.aat, hb_forward<Ts> (ds)...));
+#endif
+    default:	return_trace (c->default_return_value ());
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.version32.sanitize (c)) return_trace (false);
+    return_trace (dispatch (c));
+  }
+
+  protected:
+  union {
+  HBUINT32		version32;
+  HBUINT16		major;
+  KernOT		ot;
+#ifndef HB_NO_AAT_SHAPE
+  KernAAT		aat;
+#endif
+  } u;
+  public:
+  DEFINE_SIZE_UNION (4, version32);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_KERN_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-base-table.hh b/thirdparty/harfbuzz/src/hb-ot-layout-base-table.hh
new file mode 100644
index 0000000000..02fe14fa06
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-base-table.hh
@@ -0,0 +1,509 @@
+/*
+ * Copyright © 2016  Elie Roux <elie.roux@telecom-bretagne.eu>
+ * Copyright © 2018  Google, Inc.
+ * Copyright © 2018-2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_BASE_TABLE_HH
+#define HB_OT_LAYOUT_BASE_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout-common.hh"
+
+namespace OT {
+
+/*
+ * BASE -- Baseline
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/base
+ */
+
+struct BaseCoordFormat1
+{
+  hb_position_t get_coord () const { return coordinate; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 1 */
+  FWORD		coordinate;	/* X or Y value, in design units */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct BaseCoordFormat2
+{
+  hb_position_t get_coord () const
+  {
+    /* TODO */
+    return coordinate;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 2 */
+  FWORD		coordinate;	/* X or Y value, in design units */
+  HBGlyphID	referenceGlyph;	/* Glyph ID of control glyph */
+  HBUINT16	coordPoint;	/* Index of contour point on the
+				 * reference glyph */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct BaseCoordFormat3
+{
+  hb_position_t get_coord (hb_font_t *font,
+			   const VariationStore &var_store,
+			   hb_direction_t direction) const
+  {
+    const Device &device = this+deviceTable;
+    return coordinate + (HB_DIRECTION_IS_VERTICAL (direction) ?
+			 device.get_y_delta (font, var_store) :
+			 device.get_x_delta (font, var_store));
+  }
+
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  deviceTable.sanitize (c, this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 3 */
+  FWORD		coordinate;	/* X or Y value, in design units */
+  OffsetTo<Device>
+		deviceTable;	/* Offset to Device table for X or
+				 * Y value, from beginning of
+				 * BaseCoord table (may be NULL). */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct BaseCoord
+{
+  bool has_data () const { return u.format; }
+
+  hb_position_t get_coord (hb_font_t            *font,
+			   const VariationStore &var_store,
+			   hb_direction_t        direction) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.get_coord ();
+    case 2: return u.format2.get_coord ();
+    case 3: return u.format3.get_coord (font, var_store, direction);
+    default:return 0;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!u.format.sanitize (c))) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    case 3: return_trace (u.format3.sanitize (c));
+    default:return_trace (false);
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;
+  BaseCoordFormat1	format1;
+  BaseCoordFormat2	format2;
+  BaseCoordFormat3	format3;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+struct FeatMinMaxRecord
+{
+  int cmp (hb_tag_t key) const { return tag.cmp (key); }
+
+  bool has_data () const { return tag; }
+
+  void get_min_max (const BaseCoord **min, const BaseCoord **max) const
+  {
+    if (likely (min)) *min = &(this+minCoord);
+    if (likely (max)) *max = &(this+maxCoord);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  minCoord.sanitize (c, this) &&
+			  maxCoord.sanitize (c, this)));
+  }
+
+  protected:
+  Tag		tag;		/* 4-byte feature identification tag--must
+				 * match feature tag in FeatureList */
+  OffsetTo<BaseCoord>
+		minCoord;	/* Offset to BaseCoord table that defines
+				 * the minimum extent value, from beginning
+				 * of MinMax table (may be NULL) */
+  OffsetTo<BaseCoord>
+		maxCoord;	/* Offset to BaseCoord table that defines
+				 * the maximum extent value, from beginning
+				 * of MinMax table (may be NULL) */
+  public:
+  DEFINE_SIZE_STATIC (8);
+
+};
+
+struct MinMax
+{
+  void get_min_max (hb_tag_t          feature_tag,
+		    const BaseCoord **min,
+		    const BaseCoord **max) const
+  {
+    const FeatMinMaxRecord &minMaxCoord = featMinMaxRecords.bsearch (feature_tag);
+    if (minMaxCoord.has_data ())
+      minMaxCoord.get_min_max (min, max);
+    else
+    {
+      if (likely (min)) *min = &(this+minCoord);
+      if (likely (max)) *max = &(this+maxCoord);
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  minCoord.sanitize (c, this) &&
+			  maxCoord.sanitize (c, this) &&
+			  featMinMaxRecords.sanitize (c, this)));
+  }
+
+  protected:
+  OffsetTo<BaseCoord>
+		minCoord;	/* Offset to BaseCoord table that defines
+				 * minimum extent value, from the beginning
+				 * of MinMax table (may be NULL) */
+  OffsetTo<BaseCoord>
+		maxCoord;	/* Offset to BaseCoord table that defines
+				 * maximum extent value, from the beginning
+				 * of MinMax table (may be NULL) */
+  SortedArrayOf<FeatMinMaxRecord>
+		featMinMaxRecords;
+				/* Array of FeatMinMaxRecords, in alphabetical
+				 * order by featureTableTag */
+  public:
+  DEFINE_SIZE_ARRAY (6, featMinMaxRecords);
+};
+
+struct BaseValues
+{
+  const BaseCoord &get_base_coord (int baseline_tag_index) const
+  {
+    if (baseline_tag_index == -1) baseline_tag_index = defaultIndex;
+    return this+baseCoords[baseline_tag_index];
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  baseCoords.sanitize (c, this)));
+  }
+
+  protected:
+  Index		defaultIndex;	/* Index number of default baseline for this
+				 * script — equals index position of baseline tag
+				 * in baselineTags array of the BaseTagList */
+  OffsetArrayOf<BaseCoord>
+		baseCoords;	/* Number of BaseCoord tables defined — should equal
+				 * baseTagCount in the BaseTagList
+				 *
+				 * Array of offsets to BaseCoord tables, from beginning of
+				 * BaseValues table — order matches baselineTags array in
+				 * the BaseTagList */
+  public:
+  DEFINE_SIZE_ARRAY (4, baseCoords);
+};
+
+struct BaseLangSysRecord
+{
+  int cmp (hb_tag_t key) const { return baseLangSysTag.cmp (key); }
+
+  bool has_data () const { return baseLangSysTag; }
+
+  const MinMax &get_min_max () const { return this+minMax; }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  minMax.sanitize (c, this)));
+  }
+
+  protected:
+  Tag		baseLangSysTag;	/* 4-byte language system identification tag */
+  OffsetTo<MinMax>
+		minMax;		/* Offset to MinMax table, from beginning
+				 * of BaseScript table */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct BaseScript
+{
+  const MinMax &get_min_max (hb_tag_t language_tag) const
+  {
+    const BaseLangSysRecord& record = baseLangSysRecords.bsearch (language_tag);
+    return record.has_data () ? record.get_min_max () : this+defaultMinMax;
+  }
+
+  const BaseCoord &get_base_coord (int baseline_tag_index) const
+  { return (this+baseValues).get_base_coord (baseline_tag_index); }
+
+  bool has_data () const { return baseValues; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  baseValues.sanitize (c, this) &&
+			  defaultMinMax.sanitize (c, this) &&
+			  baseLangSysRecords.sanitize (c, this)));
+  }
+
+  protected:
+  OffsetTo<BaseValues>
+		baseValues;	/* Offset to BaseValues table, from beginning
+				 * of BaseScript table (may be NULL) */
+  OffsetTo<MinMax>
+		defaultMinMax;	/* Offset to MinMax table, from beginning of
+				 * BaseScript table (may be NULL) */
+  SortedArrayOf<BaseLangSysRecord>
+		baseLangSysRecords;
+				/* Number of BaseLangSysRecords
+				 * defined — may be zero (0) */
+
+  public:
+  DEFINE_SIZE_ARRAY (6, baseLangSysRecords);
+};
+
+struct BaseScriptList;
+struct BaseScriptRecord
+{
+  int cmp (hb_tag_t key) const { return baseScriptTag.cmp (key); }
+
+  bool has_data () const { return baseScriptTag; }
+
+  const BaseScript &get_base_script (const BaseScriptList *list) const
+  { return list+baseScript; }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  baseScript.sanitize (c, base)));
+  }
+
+  protected:
+  Tag		baseScriptTag;	/* 4-byte script identification tag */
+  OffsetTo<BaseScript>
+		baseScript;	/* Offset to BaseScript table, from beginning
+				 * of BaseScriptList */
+
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct BaseScriptList
+{
+  const BaseScript &get_base_script (hb_tag_t script) const
+  {
+    const BaseScriptRecord *record = &baseScriptRecords.bsearch (script);
+    if (!record->has_data ()) record = &baseScriptRecords.bsearch (HB_TAG ('D','F','L','T'));
+    return record->has_data () ? record->get_base_script (this) : Null (BaseScript);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  baseScriptRecords.sanitize (c, this));
+  }
+
+  protected:
+  SortedArrayOf<BaseScriptRecord>
+			baseScriptRecords;
+
+  public:
+  DEFINE_SIZE_ARRAY (2, baseScriptRecords);
+};
+
+struct Axis
+{
+  bool get_baseline (hb_tag_t          baseline_tag,
+		     hb_tag_t          script_tag,
+		     hb_tag_t          language_tag,
+		     const BaseCoord **coord) const
+  {
+    const BaseScript &base_script = (this+baseScriptList).get_base_script (script_tag);
+    if (!base_script.has_data ()) return false;
+
+    if (likely (coord))
+    {
+      unsigned int tag_index = 0;
+      (this+baseTagList).bfind (baseline_tag, &tag_index);
+      *coord = &base_script.get_base_coord (tag_index);
+    }
+
+    return true;
+  }
+
+  bool get_min_max (hb_tag_t          script_tag,
+		    hb_tag_t          language_tag,
+		    hb_tag_t          feature_tag,
+		    const BaseCoord **min_coord,
+		    const BaseCoord **max_coord) const
+  {
+    const BaseScript &base_script = (this+baseScriptList).get_base_script (script_tag);
+    if (!base_script.has_data ()) return false;
+
+    base_script.get_min_max (language_tag).get_min_max (feature_tag, min_coord, max_coord);
+
+    return true;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (this+baseTagList).sanitize (c) &&
+			  (this+baseScriptList).sanitize (c)));
+  }
+
+  protected:
+  OffsetTo<SortedArrayOf<Tag>>
+		baseTagList;	/* Offset to BaseTagList table, from beginning
+				 * of Axis table (may be NULL)
+				 * Array of 4-byte baseline identification tags — must
+				 * be in alphabetical order */
+  OffsetTo<BaseScriptList>
+		baseScriptList;	/* Offset to BaseScriptList table, from beginning
+				 * of Axis table
+				 * Array of BaseScriptRecords, in alphabetical order
+				 * by baseScriptTag */
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct BASE
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_BASE;
+
+  const Axis &get_axis (hb_direction_t direction) const
+  { return HB_DIRECTION_IS_VERTICAL (direction) ? this+vAxis : this+hAxis; }
+
+  const VariationStore &get_var_store () const
+  { return version.to_int () < 0x00010001u ? Null (VariationStore) : this+varStore; }
+
+  bool get_baseline (hb_font_t      *font,
+		     hb_tag_t        baseline_tag,
+		     hb_direction_t  direction,
+		     hb_tag_t        script_tag,
+		     hb_tag_t        language_tag,
+		     hb_position_t  *base) const
+  {
+    const BaseCoord *base_coord = nullptr;
+    if (unlikely (!get_axis (direction).get_baseline (baseline_tag, script_tag, language_tag, &base_coord) ||
+		  !base_coord || !base_coord->has_data ()))
+      return false;
+
+    if (likely (base))
+      *base = base_coord->get_coord (font, get_var_store (), direction);
+
+    return true;
+  }
+
+  /* TODO: Expose this separately sometime? */
+  bool get_min_max (hb_font_t      *font,
+		    hb_direction_t  direction,
+		    hb_tag_t        script_tag,
+		    hb_tag_t        language_tag,
+		    hb_tag_t        feature_tag,
+		    hb_position_t  *min,
+		    hb_position_t  *max)
+  {
+    const BaseCoord *min_coord, *max_coord;
+    if (!get_axis (direction).get_min_max (script_tag, language_tag, feature_tag,
+					   &min_coord, &max_coord))
+      return false;
+
+    const VariationStore &var_store = get_var_store ();
+    if (likely (min && min_coord)) *min = min_coord->get_coord (font, var_store, direction);
+    if (likely (max && max_coord)) *max = max_coord->get_coord (font, var_store, direction);
+    return true;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  likely (version.major == 1) &&
+			  hAxis.sanitize (c, this) &&
+			  vAxis.sanitize (c, this) &&
+			  (version.to_int () < 0x00010001u || varStore.sanitize (c, this))));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the BASE table */
+  OffsetTo<Axis>hAxis;		/* Offset to horizontal Axis table, from beginning
+				 * of BASE table (may be NULL) */
+  OffsetTo<Axis>vAxis;		/* Offset to vertical Axis table, from beginning
+				 * of BASE table (may be NULL) */
+  LOffsetTo<VariationStore>
+		varStore;	/* Offset to the table of Item Variation
+				 * Store--from beginning of BASE
+				 * header (may be NULL).  Introduced
+				 * in version 0x00010001. */
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_BASE_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-common.hh b/thirdparty/harfbuzz/src/hb-ot-layout-common.hh
new file mode 100644
index 0000000000..3140dd6328
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-common.hh
@@ -0,0 +1,3178 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_COMMON_HH
+#define HB_OT_LAYOUT_COMMON_HH
+
+#include "hb.hh"
+#include "hb-ot-layout.hh"
+#include "hb-open-type.hh"
+#include "hb-set.hh"
+#include "hb-bimap.hh"
+
+
+#ifndef HB_MAX_NESTING_LEVEL
+#define HB_MAX_NESTING_LEVEL	6
+#endif
+#ifndef HB_MAX_CONTEXT_LENGTH
+#define HB_MAX_CONTEXT_LENGTH	64
+#endif
+#ifndef HB_CLOSURE_MAX_STAGES
+/*
+ * The maximum number of times a lookup can be applied during shaping.
+ * Used to limit the number of iterations of the closure algorithm.
+ * This must be larger than the number of times add_pause() is
+ * called in a collect_features call of any shaper.
+ */
+#define HB_CLOSURE_MAX_STAGES	32
+#endif
+
+#ifndef HB_MAX_SCRIPTS
+#define HB_MAX_SCRIPTS	500
+#endif
+
+#ifndef HB_MAX_LANGSYS
+#define HB_MAX_LANGSYS	2000
+#endif
+
+#ifndef HB_MAX_FEATURES
+#define HB_MAX_FEATURES 750
+#endif
+
+#ifndef HB_MAX_FEATURE_INDICES
+#define HB_MAX_FEATURE_INDICES	1500
+#endif
+
+#ifndef HB_MAX_LOOKUP_INDICES
+#define HB_MAX_LOOKUP_INDICES	20000
+#endif
+
+
+namespace OT {
+
+
+#define NOT_COVERED		((unsigned int) -1)
+
+
+template<typename Iterator>
+static inline void Coverage_serialize (hb_serialize_context_t *c,
+				       Iterator it);
+
+template<typename Iterator>
+static inline void ClassDef_serialize (hb_serialize_context_t *c,
+				       Iterator it);
+
+static void ClassDef_remap_and_serialize (hb_serialize_context_t *c,
+					  const hb_set_t &glyphset,
+					  const hb_map_t &gid_klass_map,
+					  hb_sorted_vector_t<HBGlyphID> &glyphs,
+					  const hb_set_t &klasses,
+					  hb_map_t *klass_map /*INOUT*/);
+
+struct hb_subset_layout_context_t :
+  hb_dispatch_context_t<hb_subset_layout_context_t, hb_empty_t, HB_DEBUG_SUBSET>
+{
+  const char *get_name () { return "SUBSET_LAYOUT"; }
+  static return_t default_return_value () { return hb_empty_t (); }
+
+  bool visitScript ()
+  {
+    return script_count++ < HB_MAX_SCRIPTS;
+  }
+
+  bool visitLangSys ()
+  {
+    return langsys_count++ < HB_MAX_LANGSYS;
+  }
+
+  bool visitFeatureIndex (int count)
+  {
+    feature_index_count += count;
+    return feature_index_count < HB_MAX_FEATURE_INDICES;
+  }
+
+  bool visitLookupIndex()
+  {
+    lookup_index_count++;
+    return lookup_index_count < HB_MAX_LOOKUP_INDICES;
+  }
+
+  hb_subset_context_t *subset_context;
+  const hb_tag_t table_tag;
+  const hb_map_t *lookup_index_map;
+  const hb_map_t *feature_index_map;
+
+  hb_subset_layout_context_t (hb_subset_context_t *c_,
+			      hb_tag_t tag_,
+			      hb_map_t *lookup_map_,
+			      hb_map_t *feature_map_) :
+				subset_context (c_),
+				table_tag (tag_),
+				lookup_index_map (lookup_map_),
+				feature_index_map (feature_map_),
+				script_count (0),
+				langsys_count (0),
+				feature_index_count (0),
+				lookup_index_count (0)
+  {}
+
+  private:
+  unsigned script_count;
+  unsigned langsys_count;
+  unsigned feature_index_count;
+  unsigned lookup_index_count;
+};
+
+struct hb_collect_variation_indices_context_t :
+       hb_dispatch_context_t<hb_collect_variation_indices_context_t>
+{
+  template <typename T>
+  return_t dispatch (const T &obj) { obj.collect_variation_indices (this); return hb_empty_t (); }
+  static return_t default_return_value () { return hb_empty_t (); }
+
+  hb_set_t *layout_variation_indices;
+  const hb_set_t *glyph_set;
+  const hb_map_t *gpos_lookups;
+
+  hb_collect_variation_indices_context_t (hb_set_t *layout_variation_indices_,
+					  const hb_set_t *glyph_set_,
+					  const hb_map_t *gpos_lookups_) :
+					layout_variation_indices (layout_variation_indices_),
+					glyph_set (glyph_set_),
+					gpos_lookups (gpos_lookups_) {}
+};
+
+template<typename OutputArray>
+struct subset_offset_array_t
+{
+  subset_offset_array_t (hb_subset_context_t *subset_context_,
+			 OutputArray& out_,
+			 const void *base_) : subset_context (subset_context_),
+					      out (out_), base (base_) {}
+
+  template <typename T>
+  bool operator () (T&& offset)
+  {
+    auto *o = out.serialize_append (subset_context->serializer);
+    if (unlikely (!o)) return false;
+    auto snap = subset_context->serializer->snapshot ();
+    bool ret = o->serialize_subset (subset_context, offset, base);
+    if (!ret)
+    {
+      out.pop ();
+      subset_context->serializer->revert (snap);
+    }
+    return ret;
+  }
+
+  private:
+  hb_subset_context_t *subset_context;
+  OutputArray &out;
+  const void *base;
+};
+
+
+template<typename OutputArray, typename Arg>
+struct subset_offset_array_arg_t
+{
+  subset_offset_array_arg_t (hb_subset_context_t *subset_context_,
+			     OutputArray& out_,
+			     const void *base_,
+			     Arg &&arg_) : subset_context (subset_context_), out (out_),
+					  base (base_), arg (arg_) {}
+
+  template <typename T>
+  bool operator () (T&& offset)
+  {
+    auto *o = out.serialize_append (subset_context->serializer);
+    if (unlikely (!o)) return false;
+    auto snap = subset_context->serializer->snapshot ();
+    bool ret = o->serialize_subset (subset_context, offset, base, arg);
+    if (!ret)
+    {
+      out.pop ();
+      subset_context->serializer->revert (snap);
+    }
+    return ret;
+  }
+
+  private:
+  hb_subset_context_t *subset_context;
+  OutputArray &out;
+  const void *base;
+  Arg &&arg;
+};
+
+/*
+ * Helper to subset an array of offsets. Subsets the thing pointed to by each offset
+ * and discards the offset in the array if the subset operation results in an empty
+ * thing.
+ */
+struct
+{
+  template<typename OutputArray>
+  subset_offset_array_t<OutputArray>
+  operator () (hb_subset_context_t *subset_context, OutputArray& out,
+	       const void *base) const
+  { return subset_offset_array_t<OutputArray> (subset_context, out, base); }
+
+  /* Variant with one extra argument passed to serialize_subset */
+  template<typename OutputArray, typename Arg>
+  subset_offset_array_arg_t<OutputArray, Arg>
+  operator () (hb_subset_context_t *subset_context, OutputArray& out,
+	       const void *base, Arg &&arg) const
+  { return subset_offset_array_arg_t<OutputArray, Arg> (subset_context, out, base, arg); }
+}
+HB_FUNCOBJ (subset_offset_array);
+
+template<typename OutputArray>
+struct subset_record_array_t
+{
+  subset_record_array_t (hb_subset_layout_context_t *c_, OutputArray* out_,
+			 const void *base_) : subset_layout_context (c_),
+					      out (out_), base (base_) {}
+
+  template <typename T>
+  void
+  operator () (T&& record)
+  {
+    auto snap = subset_layout_context->subset_context->serializer->snapshot ();
+    bool ret = record.subset (subset_layout_context, base);
+    if (!ret) subset_layout_context->subset_context->serializer->revert (snap);
+    else out->len++;
+  }
+
+  private:
+  hb_subset_layout_context_t *subset_layout_context;
+  OutputArray *out;
+  const void *base;
+};
+
+/*
+ * Helper to subset a RecordList/record array. Subsets each Record in the array and
+ * discards the record if the subset operation returns false.
+ */
+struct
+{
+  template<typename OutputArray>
+  subset_record_array_t<OutputArray>
+  operator () (hb_subset_layout_context_t *c, OutputArray* out,
+	       const void *base) const
+  { return subset_record_array_t<OutputArray> (c, out, base); }
+}
+HB_FUNCOBJ (subset_record_array);
+
+/*
+ *
+ * OpenType Layout Common Table Formats
+ *
+ */
+
+
+/*
+ * Script, ScriptList, LangSys, Feature, FeatureList, Lookup, LookupList
+ */
+
+struct Record_sanitize_closure_t {
+  hb_tag_t tag;
+  const void *list_base;
+};
+
+template <typename Type>
+struct Record
+{
+  int cmp (hb_tag_t a) const { return tag.cmp (a); }
+
+  bool subset (hb_subset_layout_context_t *c, const void *base) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->subset_context->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+    bool ret = out->offset.serialize_subset (c->subset_context, offset, base, c, &tag);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    const Record_sanitize_closure_t closure = {tag, base};
+    return_trace (c->check_struct (this) && offset.sanitize (c, base, &closure));
+  }
+
+  Tag		tag;		/* 4-byte Tag identifier */
+  OffsetTo<Type>
+		offset;		/* Offset from beginning of object holding
+				 * the Record */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+template <typename Type>
+struct RecordArrayOf : SortedArrayOf<Record<Type>>
+{
+  const OffsetTo<Type>& get_offset (unsigned int i) const
+  { return (*this)[i].offset; }
+  OffsetTo<Type>& get_offset (unsigned int i)
+  { return (*this)[i].offset; }
+  const Tag& get_tag (unsigned int i) const
+  { return (*this)[i].tag; }
+  unsigned int get_tags (unsigned int start_offset,
+			 unsigned int *record_count /* IN/OUT */,
+			 hb_tag_t     *record_tags /* OUT */) const
+  {
+    if (record_count)
+    {
+      + this->sub_array (start_offset, record_count)
+      | hb_map (&Record<Type>::tag)
+      | hb_sink (hb_array (record_tags, *record_count))
+      ;
+    }
+    return this->len;
+  }
+  bool find_index (hb_tag_t tag, unsigned int *index) const
+  {
+    return this->bfind (tag, index, HB_BFIND_NOT_FOUND_STORE, Index::NOT_FOUND_INDEX);
+  }
+};
+
+template <typename Type>
+struct RecordListOf : RecordArrayOf<Type>
+{
+  const Type& operator [] (unsigned int i) const
+  { return this+this->get_offset (i); }
+
+  bool subset (hb_subset_context_t *c,
+	       hb_subset_layout_context_t *l) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    + this->iter ()
+    | hb_apply (subset_record_array (l, out, this))
+    ;
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (RecordArrayOf<Type>::sanitize (c, this));
+  }
+};
+
+struct Feature;
+
+struct RecordListOfFeature : RecordListOf<Feature>
+{
+  bool subset (hb_subset_context_t *c,
+	       hb_subset_layout_context_t *l) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    unsigned count = this->len;
+    + hb_zip (*this, hb_range (count))
+    | hb_filter (l->feature_index_map, hb_second)
+    | hb_map (hb_first)
+    | hb_apply (subset_record_array (l, out, this))
+    ;
+    return_trace (true);
+  }
+};
+
+struct RangeRecord
+{
+  int cmp (hb_codepoint_t g) const
+  { return g < first ? -1 : g <= last ? 0 : +1; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  { return glyphs->intersects (first, last); }
+
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  { return glyphs->add_range (first, last); }
+
+  HBGlyphID	first;		/* First GlyphID in the range */
+  HBGlyphID	last;		/* Last GlyphID in the range */
+  HBUINT16	value;		/* Value */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+DECLARE_NULL_NAMESPACE_BYTES (OT, RangeRecord);
+
+
+struct IndexArray : ArrayOf<Index>
+{
+  bool intersects (const hb_map_t *indexes) const
+  { return hb_any (*this, indexes); }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  hb_subset_layout_context_t *l,
+		  Iterator it)
+  {
+    if (!it) return;
+    if (unlikely (!c->extend_min ((*this)))) return;
+
+    for (const auto _ : it)
+    {
+      if (!l->visitLookupIndex()) break;
+
+      Index i;
+      i = _;
+      c->copy (i);
+      this->len++;
+    }
+  }
+
+  unsigned int get_indexes (unsigned int start_offset,
+			    unsigned int *_count /* IN/OUT */,
+			    unsigned int *_indexes /* OUT */) const
+  {
+    if (_count)
+    {
+      + this->sub_array (start_offset, _count)
+      | hb_sink (hb_array (_indexes, *_count))
+      ;
+    }
+    return this->len;
+  }
+
+  void add_indexes_to (hb_set_t* output /* OUT */) const
+  {
+    output->add_array (arrayZ, len);
+  }
+};
+
+
+struct LangSys
+{
+  unsigned int get_feature_count () const
+  { return featureIndex.len; }
+  hb_tag_t get_feature_index (unsigned int i) const
+  { return featureIndex[i]; }
+  unsigned int get_feature_indexes (unsigned int start_offset,
+				    unsigned int *feature_count /* IN/OUT */,
+				    unsigned int *feature_indexes /* OUT */) const
+  { return featureIndex.get_indexes (start_offset, feature_count, feature_indexes); }
+  void add_feature_indexes_to (hb_set_t *feature_indexes) const
+  { featureIndex.add_indexes_to (feature_indexes); }
+
+  bool has_required_feature () const { return reqFeatureIndex != 0xFFFFu; }
+  unsigned int get_required_feature_index () const
+  {
+    if (reqFeatureIndex == 0xFFFFu)
+      return Index::NOT_FOUND_INDEX;
+   return reqFeatureIndex;
+  }
+
+  LangSys* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (c->embed (*this));
+  }
+
+  bool operator == (const LangSys& o) const
+  {
+    if (featureIndex.len != o.featureIndex.len ||
+	reqFeatureIndex != o.reqFeatureIndex)
+      return false;
+
+    for (const auto _ : + hb_zip (featureIndex, o.featureIndex))
+      if (_.first != _.second) return false;
+
+    return true;
+  }
+
+  bool subset (hb_subset_context_t        *c,
+	       hb_subset_layout_context_t *l,
+	       const Tag                  *tag = nullptr) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    out->reqFeatureIndex = l->feature_index_map->has (reqFeatureIndex) ? l->feature_index_map->get (reqFeatureIndex) : 0xFFFFu;
+
+    if (!l->visitFeatureIndex (featureIndex.len))
+      return_trace (false);
+
+    auto it =
+    + hb_iter (featureIndex)
+    | hb_filter (l->feature_index_map)
+    | hb_map (l->feature_index_map)
+    ;
+
+    bool ret = bool (it);
+    out->featureIndex.serialize (c->serializer, l, it);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const Record_sanitize_closure_t * = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && featureIndex.sanitize (c));
+  }
+
+  Offset16	lookupOrderZ;	/* = Null (reserved for an offset to a
+				 * reordering table) */
+  HBUINT16	reqFeatureIndex;/* Index of a feature required for this
+				 * language system--if no required features
+				 * = 0xFFFFu */
+  IndexArray	featureIndex;	/* Array of indices into the FeatureList */
+  public:
+  DEFINE_SIZE_ARRAY_SIZED (6, featureIndex);
+};
+DECLARE_NULL_NAMESPACE_BYTES (OT, LangSys);
+
+struct Script
+{
+  unsigned int get_lang_sys_count () const
+  { return langSys.len; }
+  const Tag& get_lang_sys_tag (unsigned int i) const
+  { return langSys.get_tag (i); }
+  unsigned int get_lang_sys_tags (unsigned int start_offset,
+				  unsigned int *lang_sys_count /* IN/OUT */,
+				  hb_tag_t     *lang_sys_tags /* OUT */) const
+  { return langSys.get_tags (start_offset, lang_sys_count, lang_sys_tags); }
+  const LangSys& get_lang_sys (unsigned int i) const
+  {
+    if (i == Index::NOT_FOUND_INDEX) return get_default_lang_sys ();
+    return this+langSys[i].offset;
+  }
+  bool find_lang_sys_index (hb_tag_t tag, unsigned int *index) const
+  { return langSys.find_index (tag, index); }
+
+  bool has_default_lang_sys () const           { return defaultLangSys != 0; }
+  const LangSys& get_default_lang_sys () const { return this+defaultLangSys; }
+
+  bool subset (hb_subset_context_t         *c,
+	       hb_subset_layout_context_t  *l,
+	       const Tag                   *tag) const
+  {
+    TRACE_SUBSET (this);
+    if (!l->visitScript ()) return_trace (false);
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    bool defaultLang = false;
+    if (has_default_lang_sys ())
+    {
+      c->serializer->push ();
+      const LangSys& ls = this+defaultLangSys;
+      bool ret = ls.subset (c, l);
+      if (!ret && tag && *tag != HB_TAG ('D', 'F', 'L', 'T'))
+      {
+	c->serializer->pop_discard ();
+	out->defaultLangSys = 0;
+      }
+      else
+      {
+	c->serializer->add_link (out->defaultLangSys, c->serializer->pop_pack ());
+	defaultLang = true;
+      }
+    }
+
+    + langSys.iter ()
+    | hb_filter ([=] (const Record<LangSys>& record) {return l->visitLangSys (); })
+    | hb_filter ([&] (const Record<LangSys>& record)
+		 {
+		   const LangSys& d = this+defaultLangSys;
+		   const LangSys& l = this+record.offset;
+		   return !(l == d);
+		 })
+    | hb_apply (subset_record_array (l, &(out->langSys), this))
+    ;
+
+    return_trace (bool (out->langSys.len) || defaultLang || l->table_tag == HB_OT_TAG_GSUB);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const Record_sanitize_closure_t * = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (defaultLangSys.sanitize (c, this) && langSys.sanitize (c, this));
+  }
+
+  protected:
+  OffsetTo<LangSys>
+		defaultLangSys;	/* Offset to DefaultLangSys table--from
+				 * beginning of Script table--may be Null */
+  RecordArrayOf<LangSys>
+		langSys;	/* Array of LangSysRecords--listed
+				 * alphabetically by LangSysTag */
+  public:
+  DEFINE_SIZE_ARRAY_SIZED (4, langSys);
+};
+
+typedef RecordListOf<Script> ScriptList;
+
+
+/* https://docs.microsoft.com/en-us/typography/opentype/spec/features_pt#size */
+struct FeatureParamsSize
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this))) return_trace (false);
+
+    /* This subtable has some "history", if you will.  Some earlier versions of
+     * Adobe tools calculated the offset of the FeatureParams sutable from the
+     * beginning of the FeatureList table!  Now, that is dealt with in the
+     * Feature implementation.  But we still need to be able to tell junk from
+     * real data.  Note: We don't check that the nameID actually exists.
+     *
+     * Read Roberts wrote on 9/15/06 on opentype-list@indx.co.uk :
+     *
+     * Yes, it is correct that a new version of the AFDKO (version 2.0) will be
+     * coming out soon, and that the makeotf program will build a font with a
+     * 'size' feature that is correct by the specification.
+     *
+     * The specification for this feature tag is in the "OpenType Layout Tag
+     * Registry". You can see a copy of this at:
+     * https://docs.microsoft.com/en-us/typography/opentype/spec/features_pt#tag-size
+     *
+     * Here is one set of rules to determine if the 'size' feature is built
+     * correctly, or as by the older versions of MakeOTF. You may be able to do
+     * better.
+     *
+     * Assume that the offset to the size feature is according to specification,
+     * and make the following value checks. If it fails, assume the size
+     * feature is calculated as versions of MakeOTF before the AFDKO 2.0 built it.
+     * If this fails, reject the 'size' feature. The older makeOTF's calculated the
+     * offset from the beginning of the FeatureList table, rather than from the
+     * beginning of the 'size' Feature table.
+     *
+     * If "design size" == 0:
+     *     fails check
+     *
+     * Else if ("subfamily identifier" == 0 and
+     *     "range start" == 0 and
+     *     "range end" == 0 and
+     *     "range start" == 0 and
+     *     "menu name ID" == 0)
+     *     passes check: this is the format used when there is a design size
+     * specified, but there is no recommended size range.
+     *
+     * Else if ("design size" <  "range start" or
+     *     "design size" >   "range end" or
+     *     "range end" <= "range start" or
+     *     "menu name ID"  < 256 or
+     *     "menu name ID"  > 32767 or
+     *     menu name ID is not a name ID which is actually in the name table)
+     *     fails test
+     * Else
+     *     passes test.
+     */
+
+    if (!designSize)
+      return_trace (false);
+    else if (subfamilyID == 0 &&
+	     subfamilyNameID == 0 &&
+	     rangeStart == 0 &&
+	     rangeEnd == 0)
+      return_trace (true);
+    else if (designSize < rangeStart ||
+	     designSize > rangeEnd ||
+	     subfamilyNameID < 256 ||
+	     subfamilyNameID > 32767)
+      return_trace (false);
+    else
+      return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    return_trace ((bool) c->serializer->embed (*this));
+  }
+
+  HBUINT16	designSize;	/* Represents the design size in 720/inch
+				 * units (decipoints).  The design size entry
+				 * must be non-zero.  When there is a design
+				 * size but no recommended size range, the
+				 * rest of the array will consist of zeros. */
+  HBUINT16	subfamilyID;	/* Has no independent meaning, but serves
+				 * as an identifier that associates fonts
+				 * in a subfamily. All fonts which share a
+				 * Preferred or Font Family name and which
+				 * differ only by size range shall have the
+				 * same subfamily value, and no fonts which
+				 * differ in weight or style shall have the
+				 * same subfamily value. If this value is
+				 * zero, the remaining fields in the array
+				 * will be ignored. */
+  NameID	subfamilyNameID;/* If the preceding value is non-zero, this
+				 * value must be set in the range 256 - 32767
+				 * (inclusive). It records the value of a
+				 * field in the name table, which must
+				 * contain English-language strings encoded
+				 * in Windows Unicode and Macintosh Roman,
+				 * and may contain additional strings
+				 * localized to other scripts and languages.
+				 * Each of these strings is the name an
+				 * application should use, in combination
+				 * with the family name, to represent the
+				 * subfamily in a menu.  Applications will
+				 * choose the appropriate version based on
+				 * their selection criteria. */
+  HBUINT16	rangeStart;	/* Large end of the recommended usage range
+				 * (inclusive), stored in 720/inch units
+				 * (decipoints). */
+  HBUINT16	rangeEnd;	/* Small end of the recommended usage range
+				   (exclusive), stored in 720/inch units
+				 * (decipoints). */
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+/* https://docs.microsoft.com/en-us/typography/opentype/spec/features_pt#ssxx */
+struct FeatureParamsStylisticSet
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    /* Right now minorVersion is at zero.  Which means, any table supports
+     * the uiNameID field. */
+    return_trace (c->check_struct (this));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    return_trace ((bool) c->serializer->embed (*this));
+  }
+
+  HBUINT16	version;	/* (set to 0): This corresponds to a “minor”
+				 * version number. Additional data may be
+				 * added to the end of this Feature Parameters
+				 * table in the future. */
+
+  NameID	uiNameID;	/* The 'name' table name ID that specifies a
+				 * string (or strings, for multiple languages)
+				 * for a user-interface label for this
+				 * feature.  The values of uiLabelNameId and
+				 * sampleTextNameId are expected to be in the
+				 * font-specific name ID range (256-32767),
+				 * though that is not a requirement in this
+				 * Feature Parameters specification. The
+				 * user-interface label for the feature can
+				 * be provided in multiple languages. An
+				 * English string should be included as a
+				 * fallback. The string should be kept to a
+				 * minimal length to fit comfortably with
+				 * different application interfaces. */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+/* https://docs.microsoft.com/en-us/typography/opentype/spec/features_ae#cv01-cv99 */
+struct FeatureParamsCharacterVariants
+{
+  unsigned
+  get_characters (unsigned start_offset, unsigned *char_count, hb_codepoint_t *chars) const
+  {
+    if (char_count)
+    {
+      + characters.sub_array (start_offset, char_count)
+      | hb_sink (hb_array (chars, *char_count))
+      ;
+    }
+    return characters.len;
+  }
+
+  unsigned get_size () const
+  { return min_size + characters.len * HBUINT24::static_size; }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    return_trace ((bool) c->serializer->embed (*this));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  characters.sanitize (c));
+  }
+
+  HBUINT16	format;			/* Format number is set to 0. */
+  NameID	featUILableNameID;	/* The ‘name’ table name ID that
+					 * specifies a string (or strings,
+					 * for multiple languages) for a
+					 * user-interface label for this
+					 * feature. (May be NULL.) */
+  NameID	featUITooltipTextNameID;/* The ‘name’ table name ID that
+					 * specifies a string (or strings,
+					 * for multiple languages) that an
+					 * application can use for tooltip
+					 * text for this feature. (May be
+					 * nullptr.) */
+  NameID	sampleTextNameID;	/* The ‘name’ table name ID that
+					 * specifies sample text that
+					 * illustrates the effect of this
+					 * feature. (May be NULL.) */
+  HBUINT16	numNamedParameters;	/* Number of named parameters. (May
+					 * be zero.) */
+  NameID	firstParamUILabelNameID;/* The first ‘name’ table name ID
+					 * used to specify strings for
+					 * user-interface labels for the
+					 * feature parameters. (Must be zero
+					 * if numParameters is zero.) */
+  ArrayOf<HBUINT24>
+		characters;		/* Array of the Unicode Scalar Value
+					 * of the characters for which this
+					 * feature provides glyph variants.
+					 * (May be zero.) */
+  public:
+  DEFINE_SIZE_ARRAY (14, characters);
+};
+
+struct FeatureParams
+{
+  bool sanitize (hb_sanitize_context_t *c, hb_tag_t tag) const
+  {
+#ifdef HB_NO_LAYOUT_FEATURE_PARAMS
+    return true;
+#endif
+    TRACE_SANITIZE (this);
+    if (tag == HB_TAG ('s','i','z','e'))
+      return_trace (u.size.sanitize (c));
+    if ((tag & 0xFFFF0000u) == HB_TAG ('s','s','\0','\0')) /* ssXX */
+      return_trace (u.stylisticSet.sanitize (c));
+    if ((tag & 0xFFFF0000u) == HB_TAG ('c','v','\0','\0')) /* cvXX */
+      return_trace (u.characterVariants.sanitize (c));
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c, const Tag* tag) const
+  {
+    TRACE_SUBSET (this);
+    if (!tag) return_trace (false);
+    if (*tag == HB_TAG ('s','i','z','e'))
+      return_trace (u.size.subset (c));
+    if ((*tag & 0xFFFF0000u) == HB_TAG ('s','s','\0','\0')) /* ssXX */
+      return_trace (u.stylisticSet.subset (c));
+    if ((*tag & 0xFFFF0000u) == HB_TAG ('c','v','\0','\0')) /* cvXX */
+      return_trace (u.characterVariants.subset (c));
+    return_trace (false);
+  }
+
+#ifndef HB_NO_LAYOUT_FEATURE_PARAMS
+  const FeatureParamsSize& get_size_params (hb_tag_t tag) const
+  {
+    if (tag == HB_TAG ('s','i','z','e'))
+      return u.size;
+    return Null (FeatureParamsSize);
+  }
+  const FeatureParamsStylisticSet& get_stylistic_set_params (hb_tag_t tag) const
+  {
+    if ((tag & 0xFFFF0000u) == HB_TAG ('s','s','\0','\0')) /* ssXX */
+      return u.stylisticSet;
+    return Null (FeatureParamsStylisticSet);
+  }
+  const FeatureParamsCharacterVariants& get_character_variants_params (hb_tag_t tag) const
+  {
+    if ((tag & 0xFFFF0000u) == HB_TAG ('c','v','\0','\0')) /* cvXX */
+      return u.characterVariants;
+    return Null (FeatureParamsCharacterVariants);
+  }
+#endif
+
+  private:
+  union {
+  FeatureParamsSize			size;
+  FeatureParamsStylisticSet		stylisticSet;
+  FeatureParamsCharacterVariants	characterVariants;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (0);
+};
+
+struct Feature
+{
+  unsigned int get_lookup_count () const
+  { return lookupIndex.len; }
+  hb_tag_t get_lookup_index (unsigned int i) const
+  { return lookupIndex[i]; }
+  unsigned int get_lookup_indexes (unsigned int start_index,
+				   unsigned int *lookup_count /* IN/OUT */,
+				   unsigned int *lookup_tags /* OUT */) const
+  { return lookupIndex.get_indexes (start_index, lookup_count, lookup_tags); }
+  void add_lookup_indexes_to (hb_set_t *lookup_indexes) const
+  { lookupIndex.add_indexes_to (lookup_indexes); }
+
+  const FeatureParams &get_feature_params () const
+  { return this+featureParams; }
+
+  bool intersects_lookup_indexes (const hb_map_t *lookup_indexes) const
+  { return lookupIndex.intersects (lookup_indexes); }
+
+  bool subset (hb_subset_context_t         *c,
+	       hb_subset_layout_context_t  *l,
+	       const Tag                   *tag = nullptr) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    bool subset_featureParams = out->featureParams.serialize_subset (c, featureParams, this, tag);
+
+    auto it =
+    + hb_iter (lookupIndex)
+    | hb_filter (l->lookup_index_map)
+    | hb_map (l->lookup_index_map)
+    ;
+
+    out->lookupIndex.serialize (c->serializer, l, it);
+    return_trace (bool (it) || subset_featureParams
+		  || (tag && *tag == HB_TAG ('p', 'r', 'e', 'f')));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const Record_sanitize_closure_t *closure = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!(c->check_struct (this) && lookupIndex.sanitize (c))))
+      return_trace (false);
+
+    /* Some earlier versions of Adobe tools calculated the offset of the
+     * FeatureParams subtable from the beginning of the FeatureList table!
+     *
+     * If sanitizing "failed" for the FeatureParams subtable, try it with the
+     * alternative location.  We would know sanitize "failed" if old value
+     * of the offset was non-zero, but it's zeroed now.
+     *
+     * Only do this for the 'size' feature, since at the time of the faulty
+     * Adobe tools, only the 'size' feature had FeatureParams defined.
+     */
+
+    if (likely (featureParams.is_null ()))
+      return_trace (true);
+
+    unsigned int orig_offset = featureParams;
+    if (unlikely (!featureParams.sanitize (c, this, closure ? closure->tag : HB_TAG_NONE)))
+      return_trace (false);
+
+    if (featureParams == 0 && closure &&
+	closure->tag == HB_TAG ('s','i','z','e') &&
+	closure->list_base && closure->list_base < this)
+    {
+      unsigned int new_offset_int = orig_offset -
+				    (((char *) this) - ((char *) closure->list_base));
+
+      OffsetTo<FeatureParams> new_offset;
+      /* Check that it would not overflow. */
+      new_offset = new_offset_int;
+      if (new_offset == new_offset_int &&
+	  c->try_set (&featureParams, new_offset_int) &&
+	  !featureParams.sanitize (c, this, closure ? closure->tag : HB_TAG_NONE))
+	return_trace (false);
+    }
+
+    return_trace (true);
+  }
+
+  OffsetTo<FeatureParams>
+		 featureParams;	/* Offset to Feature Parameters table (if one
+				 * has been defined for the feature), relative
+				 * to the beginning of the Feature Table; = Null
+				 * if not required */
+  IndexArray	 lookupIndex;	/* Array of LookupList indices */
+  public:
+  DEFINE_SIZE_ARRAY_SIZED (4, lookupIndex);
+};
+
+typedef RecordListOf<Feature> FeatureList;
+
+
+struct LookupFlag : HBUINT16
+{
+  enum Flags {
+    RightToLeft		= 0x0001u,
+    IgnoreBaseGlyphs	= 0x0002u,
+    IgnoreLigatures	= 0x0004u,
+    IgnoreMarks		= 0x0008u,
+    IgnoreFlags		= 0x000Eu,
+    UseMarkFilteringSet	= 0x0010u,
+    Reserved		= 0x00E0u,
+    MarkAttachmentType	= 0xFF00u
+  };
+  public:
+  DEFINE_SIZE_STATIC (2);
+};
+
+} /* namespace OT */
+/* This has to be outside the namespace. */
+HB_MARK_AS_FLAG_T (OT::LookupFlag::Flags);
+namespace OT {
+
+struct Lookup
+{
+  unsigned int get_subtable_count () const { return subTable.len; }
+
+  template <typename TSubTable>
+  const OffsetArrayOf<TSubTable>& get_subtables () const
+  { return reinterpret_cast<const OffsetArrayOf<TSubTable> &> (subTable); }
+  template <typename TSubTable>
+  OffsetArrayOf<TSubTable>& get_subtables ()
+  { return reinterpret_cast<OffsetArrayOf<TSubTable> &> (subTable); }
+
+  template <typename TSubTable>
+  const TSubTable& get_subtable (unsigned int i) const
+  { return this+get_subtables<TSubTable> ()[i]; }
+  template <typename TSubTable>
+  TSubTable& get_subtable (unsigned int i)
+  { return this+get_subtables<TSubTable> ()[i]; }
+
+  unsigned int get_size () const
+  {
+    const HBUINT16 &markFilteringSet = StructAfter<const HBUINT16> (subTable);
+    if (lookupFlag & LookupFlag::UseMarkFilteringSet)
+      return (const char *) &StructAfter<const char> (markFilteringSet) - (const char *) this;
+    return (const char *) &markFilteringSet - (const char *) this;
+  }
+
+  unsigned int get_type () const { return lookupType; }
+
+  /* lookup_props is a 32-bit integer where the lower 16-bit is LookupFlag and
+   * higher 16-bit is mark-filtering-set if the lookup uses one.
+   * Not to be confused with glyph_props which is very similar. */
+  uint32_t get_props () const
+  {
+    unsigned int flag = lookupFlag;
+    if (unlikely (flag & LookupFlag::UseMarkFilteringSet))
+    {
+      const HBUINT16 &markFilteringSet = StructAfter<HBUINT16> (subTable);
+      flag += (markFilteringSet << 16);
+    }
+    return flag;
+  }
+
+  template <typename TSubTable, typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    unsigned int lookup_type = get_type ();
+    TRACE_DISPATCH (this, lookup_type);
+    unsigned int count = get_subtable_count ();
+    for (unsigned int i = 0; i < count; i++) {
+      typename context_t::return_t r = get_subtable<TSubTable> (i).dispatch (c, lookup_type, hb_forward<Ts> (ds)...);
+      if (c->stop_sublookup_iteration (r))
+	return_trace (r);
+    }
+    return_trace (c->default_return_value ());
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned int lookup_type,
+		  uint32_t lookup_props,
+		  unsigned int num_subtables)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    lookupType = lookup_type;
+    lookupFlag = lookup_props & 0xFFFFu;
+    if (unlikely (!subTable.serialize (c, num_subtables))) return_trace (false);
+    if (lookupFlag & LookupFlag::UseMarkFilteringSet)
+    {
+      if (unlikely (!c->extend (*this))) return_trace (false);
+      HBUINT16 &markFilteringSet = StructAfter<HBUINT16> (subTable);
+      markFilteringSet = lookup_props >> 16;
+    }
+    return_trace (true);
+  }
+
+  template <typename TSubTable>
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+    out->lookupType = lookupType;
+    out->lookupFlag = lookupFlag;
+
+    const hb_set_t *glyphset = c->plan->glyphset ();
+    unsigned int lookup_type = get_type ();
+    + hb_iter (get_subtables <TSubTable> ())
+    | hb_filter ([this, glyphset, lookup_type] (const OffsetTo<TSubTable> &_) { return (this+_).intersects (glyphset, lookup_type); })
+    | hb_apply (subset_offset_array (c, out->get_subtables<TSubTable> (), this, lookup_type))
+    ;
+
+    return_trace (true);
+  }
+
+  template <typename TSubTable>
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!(c->check_struct (this) && subTable.sanitize (c))) return_trace (false);
+
+    unsigned subtables = get_subtable_count ();
+    if (unlikely (!c->visit_subtables (subtables))) return_trace (false);
+
+    if (lookupFlag & LookupFlag::UseMarkFilteringSet)
+    {
+      const HBUINT16 &markFilteringSet = StructAfter<HBUINT16> (subTable);
+      if (!markFilteringSet.sanitize (c)) return_trace (false);
+    }
+
+    if (unlikely (!get_subtables<TSubTable> ().sanitize (c, this, get_type ())))
+      return_trace (false);
+
+    if (unlikely (get_type () == TSubTable::Extension && !c->get_edit_count ()))
+    {
+      /* The spec says all subtables of an Extension lookup should
+       * have the same type, which shall not be the Extension type
+       * itself (but we already checked for that).
+       * This is specially important if one has a reverse type!
+       *
+       * We only do this if sanitizer edit_count is zero.  Otherwise,
+       * some of the subtables might have become insane after they
+       * were sanity-checked by the edits of subsequent subtables.
+       * https://bugs.chromium.org/p/chromium/issues/detail?id=960331
+       */
+      unsigned int type = get_subtable<TSubTable> (0).u.extension.get_type ();
+      for (unsigned int i = 1; i < subtables; i++)
+	if (get_subtable<TSubTable> (i).u.extension.get_type () != type)
+	  return_trace (false);
+    }
+    return_trace (true);
+  }
+
+  private:
+  HBUINT16	lookupType;		/* Different enumerations for GSUB and GPOS */
+  HBUINT16	lookupFlag;		/* Lookup qualifiers */
+  ArrayOf<Offset16>
+		subTable;		/* Array of SubTables */
+/*HBUINT16	markFilteringSetX[HB_VAR_ARRAY];*//* Index (base 0) into GDEF mark glyph sets
+					 * structure. This field is only present if bit
+					 * UseMarkFilteringSet of lookup flags is set. */
+  public:
+  DEFINE_SIZE_ARRAY (6, subTable);
+};
+
+typedef OffsetListOf<Lookup> LookupList;
+
+template <typename TLookup>
+struct LookupOffsetList : OffsetListOf<TLookup>
+{
+  bool subset (hb_subset_context_t        *c,
+	       hb_subset_layout_context_t *l) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    unsigned count = this->len;
+    + hb_zip (*this, hb_range (count))
+    | hb_filter (l->lookup_index_map, hb_second)
+    | hb_map (hb_first)
+    | hb_apply (subset_offset_array (c, *out, this))
+    ;
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (OffsetListOf<TLookup>::sanitize (c, this));
+  }
+};
+
+
+/*
+ * Coverage Table
+ */
+
+struct CoverageFormat1
+{
+  friend struct Coverage;
+
+  private:
+  unsigned int get_coverage (hb_codepoint_t glyph_id) const
+  {
+    unsigned int i;
+    glyphArray.bfind (glyph_id, &i, HB_BFIND_NOT_FOUND_STORE, NOT_COVERED);
+    return i;
+  }
+
+  template <typename Iterator,
+      hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c, Iterator glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (glyphArray.serialize (c, glyphs));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (glyphArray.sanitize (c));
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    /* TODO Speed up, using hb_set_next() and bsearch()? */
+    unsigned int count = glyphArray.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (glyphs->has (glyphArray[i]))
+	return true;
+    return false;
+  }
+  bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
+  { return glyphs->has (glyphArray[index]); }
+
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  { return glyphs->add_sorted_array (glyphArray.arrayZ, glyphArray.len); }
+
+  public:
+  /* Older compilers need this to be public. */
+  struct iter_t
+  {
+    void init (const struct CoverageFormat1 &c_) { c = &c_; i = 0; }
+    void fini () {}
+    bool more () const { return i < c->glyphArray.len; }
+    void next () { i++; }
+    hb_codepoint_t get_glyph () const { return c->glyphArray[i]; }
+    bool operator != (const iter_t& o) const
+    { return i != o.i || c != o.c; }
+
+    private:
+    const struct CoverageFormat1 *c;
+    unsigned int i;
+  };
+  private:
+
+  protected:
+  HBUINT16	coverageFormat;	/* Format identifier--format = 1 */
+  SortedArrayOf<HBGlyphID>
+		glyphArray;	/* Array of GlyphIDs--in numerical order */
+  public:
+  DEFINE_SIZE_ARRAY (4, glyphArray);
+};
+
+struct CoverageFormat2
+{
+  friend struct Coverage;
+
+  private:
+  unsigned int get_coverage (hb_codepoint_t glyph_id) const
+  {
+    const RangeRecord &range = rangeRecord.bsearch (glyph_id);
+    return likely (range.first <= range.last)
+	 ? (unsigned int) range.value + (glyph_id - range.first)
+	 : NOT_COVERED;
+  }
+
+  template <typename Iterator,
+      hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c, Iterator glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    if (unlikely (!glyphs))
+    {
+      rangeRecord.len = 0;
+      return_trace (true);
+    }
+
+    /* TODO(iter) Write more efficiently? */
+
+    unsigned num_ranges = 0;
+    hb_codepoint_t last = (hb_codepoint_t) -2;
+    for (auto g: glyphs)
+    {
+      if (last + 1 != g)
+	num_ranges++;
+      last = g;
+    }
+
+    if (unlikely (!rangeRecord.serialize (c, num_ranges))) return_trace (false);
+
+    unsigned count = 0;
+    unsigned range = (unsigned) -1;
+    last = (hb_codepoint_t) -2;
+    for (auto g: glyphs)
+    {
+      if (last + 1 != g)
+      {
+	range++;
+	rangeRecord[range].first = g;
+	rangeRecord[range].value = count;
+      }
+      rangeRecord[range].last = g;
+      last = g;
+      count++;
+    }
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (rangeRecord.sanitize (c));
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    /* TODO Speed up, using hb_set_next() and bsearch()? */
+    unsigned int count = rangeRecord.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (rangeRecord[i].intersects (glyphs))
+	return true;
+    return false;
+  }
+  bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
+  {
+    unsigned int i;
+    unsigned int count = rangeRecord.len;
+    for (i = 0; i < count; i++) {
+      const RangeRecord &range = rangeRecord[i];
+      if (range.value <= index &&
+	  index < (unsigned int) range.value + (range.last - range.first) &&
+	  range.intersects (glyphs))
+	return true;
+      else if (index < range.value)
+	return false;
+    }
+    return false;
+  }
+
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  {
+    unsigned int count = rangeRecord.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!rangeRecord[i].collect_coverage (glyphs)))
+	return false;
+    return true;
+  }
+
+  public:
+  /* Older compilers need this to be public. */
+  struct iter_t
+  {
+    void init (const CoverageFormat2 &c_)
+    {
+      c = &c_;
+      coverage = 0;
+      i = 0;
+      j = c->rangeRecord.len ? c->rangeRecord[0].first : 0;
+      if (unlikely (c->rangeRecord[0].first > c->rangeRecord[0].last))
+      {
+	/* Broken table. Skip. */
+	i = c->rangeRecord.len;
+      }
+    }
+    void fini () {}
+    bool more () const { return i < c->rangeRecord.len; }
+    void next ()
+    {
+      if (j >= c->rangeRecord[i].last)
+      {
+	i++;
+	if (more ())
+	{
+	  unsigned int old = coverage;
+	  j = c->rangeRecord[i].first;
+	  coverage = c->rangeRecord[i].value;
+	  if (unlikely (coverage != old + 1))
+	  {
+	    /* Broken table. Skip. Important to avoid DoS.
+	     * Also, our callers depend on coverage being
+	     * consecutive and monotonically increasing,
+	     * ie. iota(). */
+	   i = c->rangeRecord.len;
+	   return;
+	  }
+	}
+	return;
+      }
+      coverage++;
+      j++;
+    }
+    hb_codepoint_t get_glyph () const { return j; }
+    bool operator != (const iter_t& o) const
+    { return i != o.i || j != o.j || c != o.c; }
+
+    private:
+    const struct CoverageFormat2 *c;
+    unsigned int i, coverage;
+    hb_codepoint_t j;
+  };
+  private:
+
+  protected:
+  HBUINT16	coverageFormat;	/* Format identifier--format = 2 */
+  SortedArrayOf<RangeRecord>
+		rangeRecord;	/* Array of glyph ranges--ordered by
+				 * Start GlyphID. rangeCount entries
+				 * long */
+  public:
+  DEFINE_SIZE_ARRAY (4, rangeRecord);
+};
+
+struct Coverage
+{
+  /* Has interface. */
+  static constexpr unsigned SENTINEL = NOT_COVERED;
+  typedef unsigned int value_t;
+  value_t operator [] (hb_codepoint_t k) const { return get (k); }
+  bool has (hb_codepoint_t k) const { return (*this)[k] != SENTINEL; }
+  /* Predicate. */
+  bool operator () (hb_codepoint_t k) const { return has (k); }
+
+  unsigned int get (hb_codepoint_t k) const { return get_coverage (k); }
+  unsigned int get_coverage (hb_codepoint_t glyph_id) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.get_coverage (glyph_id);
+    case 2: return u.format2.get_coverage (glyph_id);
+    default:return NOT_COVERED;
+    }
+  }
+
+  template <typename Iterator,
+      hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c, Iterator glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    unsigned count = 0;
+    unsigned num_ranges = 0;
+    hb_codepoint_t last = (hb_codepoint_t) -2;
+    for (auto g: glyphs)
+    {
+      if (last + 1 != g)
+	num_ranges++;
+      last = g;
+      count++;
+    }
+    u.format = count <= num_ranges * 3 ? 1 : 2;
+
+    switch (u.format)
+    {
+    case 1: return_trace (u.format1.serialize (c, glyphs));
+    case 2: return_trace (u.format2.serialize (c, glyphs));
+    default:return_trace (false);
+    }
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto it =
+    + iter ()
+    | hb_filter (glyphset)
+    | hb_map_retains_sorting (glyph_map)
+    ;
+
+    bool ret = bool (it);
+    Coverage_serialize (c->serializer, it);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format)
+    {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.intersects (glyphs);
+    case 2: return u.format2.intersects (glyphs);
+    default:return false;
+    }
+  }
+  bool intersects_coverage (const hb_set_t *glyphs, unsigned int index) const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.intersects_coverage (glyphs, index);
+    case 2: return u.format2.intersects_coverage (glyphs, index);
+    default:return false;
+    }
+  }
+
+  /* Might return false if array looks unsorted.
+   * Used for faster rejection of corrupt data. */
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.collect_coverage (glyphs);
+    case 2: return u.format2.collect_coverage (glyphs);
+    default:return false;
+    }
+  }
+
+  struct iter_t : hb_iter_with_fallback_t<iter_t, hb_codepoint_t>
+  {
+    static constexpr bool is_sorted_iterator = true;
+    iter_t (const Coverage &c_ = Null (Coverage))
+    {
+      memset (this, 0, sizeof (*this));
+      format = c_.u.format;
+      switch (format)
+      {
+      case 1: u.format1.init (c_.u.format1); return;
+      case 2: u.format2.init (c_.u.format2); return;
+      default:				     return;
+      }
+    }
+    bool __more__ () const
+    {
+      switch (format)
+      {
+      case 1: return u.format1.more ();
+      case 2: return u.format2.more ();
+      default:return false;
+      }
+    }
+    void __next__ ()
+    {
+      switch (format)
+      {
+      case 1: u.format1.next (); break;
+      case 2: u.format2.next (); break;
+      default:			 break;
+      }
+    }
+    typedef hb_codepoint_t __item_t__;
+    __item_t__ __item__ () const { return get_glyph (); }
+
+    hb_codepoint_t get_glyph () const
+    {
+      switch (format)
+      {
+      case 1: return u.format1.get_glyph ();
+      case 2: return u.format2.get_glyph ();
+      default:return 0;
+      }
+    }
+    bool operator != (const iter_t& o) const
+    {
+      if (format != o.format) return true;
+      switch (format)
+      {
+      case 1: return u.format1 != o.u.format1;
+      case 2: return u.format2 != o.u.format2;
+      default:return false;
+      }
+    }
+
+    private:
+    unsigned int format;
+    union {
+    CoverageFormat2::iter_t	format2; /* Put this one first since it's larger; helps shut up compiler. */
+    CoverageFormat1::iter_t	format1;
+    } u;
+  };
+  iter_t iter () const { return iter_t (*this); }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  CoverageFormat1	format1;
+  CoverageFormat2	format2;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+template<typename Iterator>
+static inline void
+Coverage_serialize (hb_serialize_context_t *c,
+		    Iterator it)
+{ c->start_embed<Coverage> ()->serialize (c, it); }
+
+static void ClassDef_remap_and_serialize (hb_serialize_context_t *c,
+					  const hb_set_t &glyphset,
+					  const hb_map_t &gid_klass_map,
+					  hb_sorted_vector_t<HBGlyphID> &glyphs,
+					  const hb_set_t &klasses,
+					  hb_map_t *klass_map /*INOUT*/)
+{
+  if (!klass_map)
+  {
+    ClassDef_serialize (c, hb_zip (glyphs.iter (), + glyphs.iter ()
+						   | hb_map (gid_klass_map)));
+    return;
+  }
+
+  /* any glyph not assigned a class value falls into Class zero (0),
+   * if any glyph assigned to class 0, remapping must start with 0->0*/
+  if (glyphset.get_population () > gid_klass_map.get_population ())
+    klass_map->set (0, 0);
+
+  unsigned idx = klass_map->has (0) ? 1 : 0;
+  for (const unsigned k: klasses.iter ())
+  {
+    if (klass_map->has (k)) continue;
+    klass_map->set (k, idx);
+    idx++;
+  }
+
+  auto it =
+  + glyphs.iter ()
+  | hb_map_retains_sorting ([&] (const HBGlyphID& gid) -> hb_pair_t<hb_codepoint_t, unsigned>
+			    {
+			      unsigned new_klass = klass_map->get (gid_klass_map[gid]);
+			      return hb_pair ((hb_codepoint_t)gid, new_klass);
+			    })
+  ;
+
+  c->propagate_error (glyphs, klasses);
+  ClassDef_serialize (c, it);
+}
+
+/*
+ * Class Definition Table
+ */
+
+struct ClassDefFormat1
+{
+  friend struct ClassDef;
+
+  private:
+  unsigned int get_class (hb_codepoint_t glyph_id) const
+  {
+    return classValue[(unsigned int) (glyph_id - startGlyph)];
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    if (unlikely (!it))
+    {
+      startGlyph = 0;
+      classValue.len = 0;
+      return_trace (true);
+    }
+
+    hb_codepoint_t glyph_min = (*it).first;
+    hb_codepoint_t glyph_max = + it
+			       | hb_map (hb_first)
+			       | hb_reduce (hb_max, 0u);
+    unsigned glyph_count = glyph_max - glyph_min + 1;
+
+    startGlyph = glyph_min;
+    if (unlikely (!classValue.serialize (c, glyph_count))) return_trace (false);
+    for (const hb_pair_t<hb_codepoint_t, unsigned> gid_klass_pair : + it)
+    {
+      unsigned idx = gid_klass_pair.first - glyph_min;
+      classValue[idx] = gid_klass_pair.second;
+    }
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       hb_map_t *klass_map = nullptr /*OUT*/) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->_glyphset_gsub;
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    hb_sorted_vector_t<HBGlyphID> glyphs;
+    hb_set_t orig_klasses;
+    hb_map_t gid_org_klass_map;
+
+    hb_codepoint_t start = startGlyph;
+    hb_codepoint_t end   = start + classValue.len;
+    for (const hb_codepoint_t gid : + hb_range (start, end)
+				    | hb_filter (glyphset))
+    {
+      unsigned klass = classValue[gid - start];
+      if (!klass) continue;
+
+      glyphs.push (glyph_map[gid]);
+      gid_org_klass_map.set (glyph_map[gid], klass);
+      orig_klasses.add (klass);
+    }
+
+    ClassDef_remap_and_serialize (c->serializer, glyphset, gid_org_klass_map,
+				  glyphs, orig_klasses, klass_map);
+    return_trace ((bool) glyphs);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && classValue.sanitize (c));
+  }
+
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  {
+    unsigned int start = 0;
+    unsigned int count = classValue.len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (classValue[i])
+	continue;
+
+      if (start != i)
+	if (unlikely (!glyphs->add_range (startGlyph + start, startGlyph + i)))
+	  return false;
+
+      start = i + 1;
+    }
+    if (start != count)
+      if (unlikely (!glyphs->add_range (startGlyph + start, startGlyph + count)))
+	return false;
+
+    return true;
+  }
+
+  template <typename set_t>
+  bool collect_class (set_t *glyphs, unsigned int klass) const
+  {
+    unsigned int count = classValue.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (classValue[i] == klass) glyphs->add (startGlyph + i);
+    return true;
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    /* TODO Speed up, using hb_set_next()? */
+    hb_codepoint_t start = startGlyph;
+    hb_codepoint_t end = startGlyph + classValue.len;
+    for (hb_codepoint_t iter = startGlyph - 1;
+	 hb_set_next (glyphs, &iter) && iter < end;)
+      if (classValue[iter - start]) return true;
+    return false;
+  }
+  bool intersects_class (const hb_set_t *glyphs, unsigned int klass) const
+  {
+    unsigned int count = classValue.len;
+    if (klass == 0)
+    {
+      /* Match if there's any glyph that is not listed! */
+      hb_codepoint_t g = HB_SET_VALUE_INVALID;
+      if (!hb_set_next (glyphs, &g)) return false;
+      if (g < startGlyph) return true;
+      g = startGlyph + count - 1;
+      if (hb_set_next (glyphs, &g)) return true;
+      /* Fall through. */
+    }
+    for (unsigned int i = 0; i < count; i++)
+      if (classValue[i] == klass && glyphs->has (startGlyph + i))
+	return true;
+    return false;
+  }
+
+  protected:
+  HBUINT16	classFormat;	/* Format identifier--format = 1 */
+  HBGlyphID	startGlyph;	/* First GlyphID of the classValueArray */
+  ArrayOf<HBUINT16>
+		classValue;	/* Array of Class Values--one per GlyphID */
+  public:
+  DEFINE_SIZE_ARRAY (6, classValue);
+};
+
+struct ClassDefFormat2
+{
+  friend struct ClassDef;
+
+  private:
+  unsigned int get_class (hb_codepoint_t glyph_id) const
+  {
+    return rangeRecord.bsearch (glyph_id).value;
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    if (unlikely (!it))
+    {
+      rangeRecord.len = 0;
+      return_trace (true);
+    }
+
+    unsigned num_ranges = 1;
+    hb_codepoint_t prev_gid = (*it).first;
+    unsigned prev_klass = (*it).second;
+
+    RangeRecord range_rec;
+    range_rec.first = prev_gid;
+    range_rec.last = prev_gid;
+    range_rec.value = prev_klass;
+
+    RangeRecord *record = c->copy (range_rec);
+    if (unlikely (!record)) return_trace (false);
+
+    for (const auto gid_klass_pair : + (++it))
+    {
+      hb_codepoint_t cur_gid = gid_klass_pair.first;
+      unsigned cur_klass = gid_klass_pair.second;
+
+      if (cur_gid != prev_gid + 1 ||
+	  cur_klass != prev_klass)
+      {
+	if (unlikely (!record)) break;
+	record->last = prev_gid;
+	num_ranges++;
+
+	range_rec.first = cur_gid;
+	range_rec.last = cur_gid;
+	range_rec.value = cur_klass;
+
+	record = c->copy (range_rec);
+      }
+
+      prev_klass = cur_klass;
+      prev_gid = cur_gid;
+    }
+
+    if (likely (record)) record->last = prev_gid;
+    rangeRecord.len = num_ranges;
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       hb_map_t *klass_map = nullptr /*OUT*/) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->_glyphset_gsub;
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    hb_sorted_vector_t<HBGlyphID> glyphs;
+    hb_set_t orig_klasses;
+    hb_map_t gid_org_klass_map;
+
+    unsigned count = rangeRecord.len;
+    for (unsigned i = 0; i < count; i++)
+    {
+      unsigned klass = rangeRecord[i].value;
+      if (!klass) continue;
+      hb_codepoint_t start = rangeRecord[i].first;
+      hb_codepoint_t end   = rangeRecord[i].last + 1;
+      for (hb_codepoint_t g = start; g < end; g++)
+      {
+	if (!glyphset.has (g)) continue;
+	glyphs.push (glyph_map[g]);
+	gid_org_klass_map.set (glyph_map[g], klass);
+	orig_klasses.add (klass);
+      }
+    }
+
+    ClassDef_remap_and_serialize (c->serializer, glyphset, gid_org_klass_map,
+				  glyphs, orig_klasses, klass_map);
+    return_trace ((bool) glyphs);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (rangeRecord.sanitize (c));
+  }
+
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  {
+    unsigned int count = rangeRecord.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (rangeRecord[i].value)
+	if (unlikely (!rangeRecord[i].collect_coverage (glyphs)))
+	  return false;
+    return true;
+  }
+
+  template <typename set_t>
+  bool collect_class (set_t *glyphs, unsigned int klass) const
+  {
+    unsigned int count = rangeRecord.len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (rangeRecord[i].value == klass)
+	if (unlikely (!rangeRecord[i].collect_coverage (glyphs)))
+	  return false;
+    }
+    return true;
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    /* TODO Speed up, using hb_set_next() and bsearch()? */
+    unsigned int count = rangeRecord.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (rangeRecord[i].intersects (glyphs))
+	return true;
+    return false;
+  }
+  bool intersects_class (const hb_set_t *glyphs, unsigned int klass) const
+  {
+    unsigned int count = rangeRecord.len;
+    if (klass == 0)
+    {
+      /* Match if there's any glyph that is not listed! */
+      hb_codepoint_t g = HB_SET_VALUE_INVALID;
+      for (unsigned int i = 0; i < count; i++)
+      {
+	if (!hb_set_next (glyphs, &g))
+	  break;
+	if (g < rangeRecord[i].first)
+	  return true;
+	g = rangeRecord[i].last;
+      }
+      if (g != HB_SET_VALUE_INVALID && hb_set_next (glyphs, &g))
+	return true;
+      /* Fall through. */
+    }
+    for (unsigned int i = 0; i < count; i++)
+      if (rangeRecord[i].value == klass && rangeRecord[i].intersects (glyphs))
+	return true;
+    return false;
+  }
+
+  protected:
+  HBUINT16	classFormat;	/* Format identifier--format = 2 */
+  SortedArrayOf<RangeRecord>
+		rangeRecord;	/* Array of glyph ranges--ordered by
+				 * Start GlyphID */
+  public:
+  DEFINE_SIZE_ARRAY (4, rangeRecord);
+};
+
+struct ClassDef
+{
+  /* Has interface. */
+  static constexpr unsigned SENTINEL = 0;
+  typedef unsigned int value_t;
+  value_t operator [] (hb_codepoint_t k) const { return get (k); }
+  bool has (hb_codepoint_t k) const { return (*this)[k] != SENTINEL; }
+  /* Projection. */
+  hb_codepoint_t operator () (hb_codepoint_t k) const { return get (k); }
+
+  unsigned int get (hb_codepoint_t k) const { return get_class (k); }
+  unsigned int get_class (hb_codepoint_t glyph_id) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.get_class (glyph_id);
+    case 2: return u.format2.get_class (glyph_id);
+    default:return 0;
+    }
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c, Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    unsigned format = 2;
+    if (likely (it))
+    {
+      hb_codepoint_t glyph_min = (*it).first;
+      hb_codepoint_t glyph_max = + it
+				 | hb_map (hb_first)
+				 | hb_reduce (hb_max, 0u);
+
+      unsigned num_ranges = 1;
+      hb_codepoint_t prev_gid = glyph_min;
+      unsigned prev_klass = (*it).second;
+
+      for (const auto gid_klass_pair : it)
+      {
+	hb_codepoint_t cur_gid = gid_klass_pair.first;
+	unsigned cur_klass = gid_klass_pair.second;
+	if (cur_gid == glyph_min || !cur_klass) continue;
+	if (cur_gid != prev_gid + 1 ||
+	    cur_klass != prev_klass)
+	  num_ranges++;
+
+	prev_gid = cur_gid;
+	prev_klass = cur_klass;
+      }
+
+      if (1 + (glyph_max - glyph_min + 1) <= num_ranges * 3)
+	format = 1;
+    }
+    u.format = format;
+
+    switch (u.format)
+    {
+    case 1: return_trace (u.format1.serialize (c, it));
+    case 2: return_trace (u.format2.serialize (c, it));
+    default:return_trace (false);
+    }
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       hb_map_t *klass_map = nullptr /*OUT*/) const
+  {
+    TRACE_SUBSET (this);
+    switch (u.format) {
+    case 1: return_trace (u.format1.subset (c, klass_map));
+    case 2: return_trace (u.format2.subset (c, klass_map));
+    default:return_trace (false);
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  /* Might return false if array looks unsorted.
+   * Used for faster rejection of corrupt data. */
+  template <typename set_t>
+  bool collect_coverage (set_t *glyphs) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.collect_coverage (glyphs);
+    case 2: return u.format2.collect_coverage (glyphs);
+    default:return false;
+    }
+  }
+
+  /* Might return false if array looks unsorted.
+   * Used for faster rejection of corrupt data. */
+  template <typename set_t>
+  bool collect_class (set_t *glyphs, unsigned int klass) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.collect_class (glyphs, klass);
+    case 2: return u.format2.collect_class (glyphs, klass);
+    default:return false;
+    }
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.intersects (glyphs);
+    case 2: return u.format2.intersects (glyphs);
+    default:return false;
+    }
+  }
+  bool intersects_class (const hb_set_t *glyphs, unsigned int klass) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.intersects_class (glyphs, klass);
+    case 2: return u.format2.intersects_class (glyphs, klass);
+    default:return false;
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  ClassDefFormat1	format1;
+  ClassDefFormat2	format2;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+template<typename Iterator>
+static inline void ClassDef_serialize (hb_serialize_context_t *c,
+				       Iterator it)
+{ c->start_embed<ClassDef> ()->serialize (c, it); }
+
+
+/*
+ * Item Variation Store
+ */
+
+struct VarRegionAxis
+{
+  float evaluate (int coord) const
+  {
+    int start = startCoord, peak = peakCoord, end = endCoord;
+
+    /* TODO Move these to sanitize(). */
+    if (unlikely (start > peak || peak > end))
+      return 1.;
+    if (unlikely (start < 0 && end > 0 && peak != 0))
+      return 1.;
+
+    if (peak == 0 || coord == peak)
+      return 1.;
+
+    if (coord <= start || end <= coord)
+      return 0.;
+
+    /* Interpolate */
+    if (coord < peak)
+      return float (coord - start) / (peak - start);
+    else
+      return float (end - coord) / (end - peak);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+    /* TODO Handle invalid start/peak/end configs, so we don't
+     * have to do that at runtime. */
+  }
+
+  public:
+  F2DOT14	startCoord;
+  F2DOT14	peakCoord;
+  F2DOT14	endCoord;
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct VarRegionList
+{
+  float evaluate (unsigned int region_index,
+		  const int *coords, unsigned int coord_len) const
+  {
+    if (unlikely (region_index >= regionCount))
+      return 0.;
+
+    const VarRegionAxis *axes = axesZ.arrayZ + (region_index * axisCount);
+
+    float v = 1.;
+    unsigned int count = axisCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      int coord = i < coord_len ? coords[i] : 0;
+      float factor = axes[i].evaluate (coord);
+      if (factor == 0.f)
+	return 0.;
+      v *= factor;
+    }
+    return v;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  axesZ.sanitize (c, (unsigned int) axisCount * (unsigned int) regionCount));
+  }
+
+  bool serialize (hb_serialize_context_t *c, const VarRegionList *src, const hb_bimap_t &region_map)
+  {
+    TRACE_SERIALIZE (this);
+    VarRegionList *out = c->allocate_min<VarRegionList> ();
+    if (unlikely (!out)) return_trace (false);
+    axisCount = src->axisCount;
+    regionCount = region_map.get_population ();
+    if (unlikely (!c->allocate_size<VarRegionList> (get_size () - min_size))) return_trace (false);
+    unsigned int region_count = src->get_region_count ();
+    for (unsigned int r = 0; r < regionCount; r++)
+    {
+      unsigned int backward = region_map.backward (r);
+      if (backward >= region_count) return_trace (false);
+      memcpy (&axesZ[axisCount * r], &src->axesZ[axisCount * backward], VarRegionAxis::static_size * axisCount);
+    }
+
+    return_trace (true);
+  }
+
+  unsigned int get_size () const { return min_size + VarRegionAxis::static_size * axisCount * regionCount; }
+  unsigned int get_region_count () const { return regionCount; }
+
+  protected:
+  HBUINT16	axisCount;
+  HBUINT16	regionCount;
+  UnsizedArrayOf<VarRegionAxis>
+		axesZ;
+  public:
+  DEFINE_SIZE_ARRAY (4, axesZ);
+};
+
+struct VarData
+{
+  unsigned int get_region_index_count () const
+  { return regionIndices.len; }
+
+  unsigned int get_row_size () const
+  { return shortCount + regionIndices.len; }
+
+  unsigned int get_size () const
+  { return itemCount * get_row_size (); }
+
+  float get_delta (unsigned int inner,
+		   const int *coords, unsigned int coord_count,
+		   const VarRegionList &regions) const
+  {
+    if (unlikely (inner >= itemCount))
+      return 0.;
+
+   unsigned int count = regionIndices.len;
+   unsigned int scount = shortCount;
+
+   const HBUINT8 *bytes = get_delta_bytes ();
+   const HBUINT8 *row = bytes + inner * (scount + count);
+
+   float delta = 0.;
+   unsigned int i = 0;
+
+   const HBINT16 *scursor = reinterpret_cast<const HBINT16 *> (row);
+   for (; i < scount; i++)
+   {
+     float scalar = regions.evaluate (regionIndices.arrayZ[i], coords, coord_count);
+     delta += scalar * *scursor++;
+   }
+   const HBINT8 *bcursor = reinterpret_cast<const HBINT8 *> (scursor);
+   for (; i < count; i++)
+   {
+     float scalar = regions.evaluate (regionIndices.arrayZ[i], coords, coord_count);
+     delta += scalar * *bcursor++;
+   }
+
+   return delta;
+  }
+
+  void get_scalars (const int *coords, unsigned int coord_count,
+		    const VarRegionList &regions,
+		    float *scalars /*OUT */,
+		    unsigned int num_scalars) const
+  {
+    unsigned count = hb_min (num_scalars, regionIndices.len);
+    for (unsigned int i = 0; i < count; i++)
+      scalars[i] = regions.evaluate (regionIndices.arrayZ[i], coords, coord_count);
+    for (unsigned int i = count; i < num_scalars; i++)
+      scalars[i] = 0.f;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  regionIndices.sanitize (c) &&
+		  shortCount <= regionIndices.len &&
+		  c->check_range (get_delta_bytes (),
+				  itemCount,
+				  get_row_size ()));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  const VarData *src,
+		  const hb_inc_bimap_t &inner_map,
+		  const hb_bimap_t &region_map)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    itemCount = inner_map.get_next_value ();
+
+    /* Optimize short count */
+    unsigned short ri_count = src->regionIndices.len;
+    enum delta_size_t { kZero=0, kByte, kShort };
+    hb_vector_t<delta_size_t> delta_sz;
+    hb_vector_t<unsigned int> ri_map;	/* maps old index to new index */
+    delta_sz.resize (ri_count);
+    ri_map.resize (ri_count);
+    unsigned int new_short_count = 0;
+    unsigned int r;
+    for (r = 0; r < ri_count; r++)
+    {
+      delta_sz[r] = kZero;
+      for (unsigned int i = 0; i < inner_map.get_next_value (); i++)
+      {
+	unsigned int old = inner_map.backward (i);
+	int16_t delta = src->get_item_delta (old, r);
+	if (delta < -128 || 127 < delta)
+	{
+	  delta_sz[r] = kShort;
+	  new_short_count++;
+	  break;
+	}
+	else if (delta != 0)
+	  delta_sz[r] = kByte;
+      }
+    }
+    unsigned int short_index = 0;
+    unsigned int byte_index = new_short_count;
+    unsigned int new_ri_count = 0;
+    for (r = 0; r < ri_count; r++)
+      if (delta_sz[r])
+      {
+	ri_map[r] = (delta_sz[r] == kShort)? short_index++ : byte_index++;
+	new_ri_count++;
+      }
+
+    shortCount = new_short_count;
+    regionIndices.len = new_ri_count;
+
+    unsigned int size = regionIndices.get_size () - HBUINT16::static_size/*regionIndices.len*/ + (get_row_size () * itemCount);
+    if (unlikely (!c->allocate_size<HBUINT8> (size)))
+      return_trace (false);
+
+    for (r = 0; r < ri_count; r++)
+      if (delta_sz[r]) regionIndices[ri_map[r]] = region_map[src->regionIndices[r]];
+
+    for (unsigned int i = 0; i < itemCount; i++)
+    {
+      unsigned int	old = inner_map.backward (i);
+      for (unsigned int r = 0; r < ri_count; r++)
+	if (delta_sz[r]) set_item_delta (i, ri_map[r], src->get_item_delta (old, r));
+    }
+
+    return_trace (true);
+  }
+
+  void collect_region_refs (hb_inc_bimap_t &region_map, const hb_inc_bimap_t &inner_map) const
+  {
+    for (unsigned int r = 0; r < regionIndices.len; r++)
+    {
+      unsigned int region = regionIndices[r];
+      if (region_map.has (region)) continue;
+      for (unsigned int i = 0; i < inner_map.get_next_value (); i++)
+	if (get_item_delta (inner_map.backward (i), r) != 0)
+	{
+	  region_map.add (region);
+	  break;
+	}
+    }
+  }
+
+  protected:
+  const HBUINT8 *get_delta_bytes () const
+  { return &StructAfter<HBUINT8> (regionIndices); }
+
+  HBUINT8 *get_delta_bytes ()
+  { return &StructAfter<HBUINT8> (regionIndices); }
+
+  int16_t get_item_delta (unsigned int item, unsigned int region) const
+  {
+    if ( item >= itemCount || unlikely (region >= regionIndices.len)) return 0;
+    const HBINT8 *p = (const HBINT8 *)get_delta_bytes () + item * get_row_size ();
+    if (region < shortCount)
+      return ((const HBINT16 *)p)[region];
+    else
+      return (p + HBINT16::static_size * shortCount)[region - shortCount];
+  }
+
+  void set_item_delta (unsigned int item, unsigned int region, int16_t delta)
+  {
+    HBINT8 *p = (HBINT8 *)get_delta_bytes () + item * get_row_size ();
+    if (region < shortCount)
+      ((HBINT16 *)p)[region] = delta;
+    else
+      (p + HBINT16::static_size * shortCount)[region - shortCount] = delta;
+  }
+
+  protected:
+  HBUINT16		itemCount;
+  HBUINT16		shortCount;
+  ArrayOf<HBUINT16>	regionIndices;
+/*UnsizedArrayOf<HBUINT8>bytesX;*/
+  public:
+  DEFINE_SIZE_ARRAY (6, regionIndices);
+};
+
+struct VariationStore
+{
+  float get_delta (unsigned int outer, unsigned int inner,
+		   const int *coords, unsigned int coord_count) const
+  {
+#ifdef HB_NO_VAR
+    return 0.f;
+#endif
+
+    if (unlikely (outer >= dataSets.len))
+      return 0.f;
+
+    return (this+dataSets[outer]).get_delta (inner,
+					     coords, coord_count,
+					     this+regions);
+  }
+
+  float get_delta (unsigned int index,
+		   const int *coords, unsigned int coord_count) const
+  {
+    unsigned int outer = index >> 16;
+    unsigned int inner = index & 0xFFFF;
+    return get_delta (outer, inner, coords, coord_count);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+#ifdef HB_NO_VAR
+    return true;
+#endif
+
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  format == 1 &&
+		  regions.sanitize (c, this) &&
+		  dataSets.sanitize (c, this));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  const VariationStore *src,
+		  const hb_array_t <hb_inc_bimap_t> &inner_maps)
+  {
+    TRACE_SERIALIZE (this);
+    unsigned int set_count = 0;
+    for (unsigned int i = 0; i < inner_maps.length; i++)
+      if (inner_maps[i].get_population () > 0) set_count++;
+
+    unsigned int size = min_size + HBUINT32::static_size * set_count;
+    if (unlikely (!c->allocate_size<HBUINT32> (size))) return_trace (false);
+    format = 1;
+
+    hb_inc_bimap_t region_map;
+    for (unsigned int i = 0; i < inner_maps.length; i++)
+      (src+src->dataSets[i]).collect_region_refs (region_map, inner_maps[i]);
+    region_map.sort ();
+
+    if (unlikely (!regions.serialize (c, this)
+		  .serialize (c, &(src+src->regions), region_map))) return_trace (false);
+
+    /* TODO: The following code could be simplified when
+     * OffsetListOf::subset () can take a custom param to be passed to VarData::serialize ()
+     */
+    dataSets.len = set_count;
+    unsigned int set_index = 0;
+    for (unsigned int i = 0; i < inner_maps.length; i++)
+    {
+      if (inner_maps[i].get_population () == 0) continue;
+      if (unlikely (!dataSets[set_index++].serialize (c, this)
+		      .serialize (c, &(src+src->dataSets[i]), inner_maps[i], region_map)))
+	return_trace (false);
+    }
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    VariationStore *varstore_prime = c->serializer->start_embed<VariationStore> ();
+    if (unlikely (!varstore_prime)) return_trace (false);
+
+    const hb_set_t *variation_indices = c->plan->layout_variation_indices;
+    if (variation_indices->is_empty ()) return_trace (false);
+
+    hb_vector_t<hb_inc_bimap_t> inner_maps;
+    inner_maps.resize ((unsigned) dataSets.len);
+    for (unsigned i = 0; i < inner_maps.length; i++)
+      inner_maps[i].init ();
+
+    for (unsigned idx : c->plan->layout_variation_indices->iter ())
+    {
+      uint16_t major = idx >> 16;
+      uint16_t minor = idx & 0xFFFF;
+
+      if (major >= inner_maps.length)
+      {
+	for (unsigned i = 0; i < inner_maps.length; i++)
+	  inner_maps[i].fini ();
+	return_trace (false);
+      }
+      inner_maps[major].add (minor);
+    }
+    varstore_prime->serialize (c->serializer, this, inner_maps.as_array ());
+
+    for (unsigned i = 0; i < inner_maps.length; i++)
+      inner_maps[i].fini ();
+    return_trace (bool (varstore_prime->dataSets));
+  }
+
+  unsigned int get_region_index_count (unsigned int ivs) const
+  { return (this+dataSets[ivs]).get_region_index_count (); }
+
+  void get_scalars (unsigned int ivs,
+		    const int *coords, unsigned int coord_count,
+		    float *scalars /*OUT*/,
+		    unsigned int num_scalars) const
+  {
+#ifdef HB_NO_VAR
+    for (unsigned i = 0; i < num_scalars; i++)
+      scalars[i] = 0.f;
+    return;
+#endif
+
+    (this+dataSets[ivs]).get_scalars (coords, coord_count, this+regions,
+				      &scalars[0], num_scalars);
+  }
+
+  unsigned int get_sub_table_count () const { return dataSets.len; }
+
+  protected:
+  HBUINT16				format;
+  LOffsetTo<VarRegionList>		regions;
+  LOffsetArrayOf<VarData>		dataSets;
+  public:
+  DEFINE_SIZE_ARRAY (8, dataSets);
+};
+
+/*
+ * Feature Variations
+ */
+
+struct ConditionFormat1
+{
+  friend struct Condition;
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+    return_trace (true);
+  }
+
+  private:
+  bool evaluate (const int *coords, unsigned int coord_len) const
+  {
+    int coord = axisIndex < coord_len ? coords[axisIndex] : 0;
+    return filterRangeMinValue <= coord && coord <= filterRangeMaxValue;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier--format = 1 */
+  HBUINT16	axisIndex;
+  F2DOT14	filterRangeMinValue;
+  F2DOT14	filterRangeMaxValue;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct Condition
+{
+  bool evaluate (const int *coords, unsigned int coord_len) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.evaluate (coords, coord_len);
+    default:return false;
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  ConditionFormat1	format1;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+struct ConditionSet
+{
+  bool evaluate (const int *coords, unsigned int coord_len) const
+  {
+    unsigned int count = conditions.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (!(this+conditions.arrayZ[i]).evaluate (coords, coord_len))
+	return false;
+    return true;
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    + conditions.iter ()
+    | hb_apply (subset_offset_array (c, out->conditions, this))
+    ;
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (conditions.sanitize (c, this));
+  }
+
+  protected:
+  LOffsetArrayOf<Condition>	conditions;
+  public:
+  DEFINE_SIZE_ARRAY (2, conditions);
+};
+
+struct FeatureTableSubstitutionRecord
+{
+  friend struct FeatureTableSubstitution;
+
+  void collect_lookups (const void *base, hb_set_t *lookup_indexes /* OUT */) const
+  {
+    return (base+feature).add_lookup_indexes_to (lookup_indexes);
+  }
+
+  void closure_features (const void *base,
+			 const hb_map_t *lookup_indexes,
+			 hb_set_t       *feature_indexes /* OUT */) const
+  {
+    if ((base+feature).intersects_lookup_indexes (lookup_indexes))
+      feature_indexes->add (featureIndex);
+  }
+
+  bool subset (hb_subset_layout_context_t *c, const void *base) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->subset_context->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+
+    out->featureIndex = c->feature_index_map->get (featureIndex);
+    bool ret = out->feature.serialize_subset (c->subset_context, feature, base, c);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && feature.sanitize (c, base));
+  }
+
+  protected:
+  HBUINT16		featureIndex;
+  LOffsetTo<Feature>	feature;
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct FeatureTableSubstitution
+{
+  const Feature *find_substitute (unsigned int feature_index) const
+  {
+    unsigned int count = substitutions.len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const FeatureTableSubstitutionRecord &record = substitutions.arrayZ[i];
+      if (record.featureIndex == feature_index)
+	return &(this+record.feature);
+    }
+    return nullptr;
+  }
+
+  void collect_lookups (const hb_set_t *feature_indexes,
+			hb_set_t       *lookup_indexes /* OUT */) const
+  {
+    + hb_iter (substitutions)
+    | hb_filter (feature_indexes, &FeatureTableSubstitutionRecord::featureIndex)
+    | hb_apply ([this, lookup_indexes] (const FeatureTableSubstitutionRecord& r)
+		{ r.collect_lookups (this, lookup_indexes); })
+    ;
+  }
+
+  void closure_features (const hb_map_t *lookup_indexes,
+			 hb_set_t       *feature_indexes /* OUT */) const
+  {
+    for (const FeatureTableSubstitutionRecord& record : substitutions)
+      record.closure_features (this, lookup_indexes, feature_indexes);
+  }
+
+  bool subset (hb_subset_context_t        *c,
+	       hb_subset_layout_context_t *l) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    out->version.major = version.major;
+    out->version.minor = version.minor;
+
+    + substitutions.iter ()
+    | hb_apply (subset_record_array (l, &(out->substitutions), this))
+    ;
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  substitutions.sanitize (c, this));
+  }
+
+  protected:
+  FixedVersion<>	version;	/* Version--0x00010000u */
+  ArrayOf<FeatureTableSubstitutionRecord>
+			substitutions;
+  public:
+  DEFINE_SIZE_ARRAY (6, substitutions);
+};
+
+struct FeatureVariationRecord
+{
+  friend struct FeatureVariations;
+
+  void collect_lookups (const void     *base,
+			const hb_set_t *feature_indexes,
+			hb_set_t       *lookup_indexes /* OUT */) const
+  {
+    return (base+substitutions).collect_lookups (feature_indexes, lookup_indexes);
+  }
+
+  void closure_features (const void     *base,
+			 const hb_map_t *lookup_indexes,
+			 hb_set_t       *feature_indexes /* OUT */) const
+  {
+    (base+substitutions).closure_features (lookup_indexes, feature_indexes);
+  }
+
+  bool subset (hb_subset_layout_context_t *c, const void *base) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->subset_context->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+
+    out->conditions.serialize_subset (c->subset_context, conditions, base);
+    out->substitutions.serialize_subset (c->subset_context, substitutions, base, c);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (conditions.sanitize (c, base) &&
+		  substitutions.sanitize (c, base));
+  }
+
+  protected:
+  LOffsetTo<ConditionSet>
+			conditions;
+  LOffsetTo<FeatureTableSubstitution>
+			substitutions;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct FeatureVariations
+{
+  static constexpr unsigned NOT_FOUND_INDEX = 0xFFFFFFFFu;
+
+  bool find_index (const int *coords, unsigned int coord_len,
+		   unsigned int *index) const
+  {
+    unsigned int count = varRecords.len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      const FeatureVariationRecord &record = varRecords.arrayZ[i];
+      if ((this+record.conditions).evaluate (coords, coord_len))
+      {
+	*index = i;
+	return true;
+      }
+    }
+    *index = NOT_FOUND_INDEX;
+    return false;
+  }
+
+  const Feature *find_substitute (unsigned int variations_index,
+				  unsigned int feature_index) const
+  {
+    const FeatureVariationRecord &record = varRecords[variations_index];
+    return (this+record.substitutions).find_substitute (feature_index);
+  }
+
+  FeatureVariations* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (c->embed (*this));
+  }
+
+  void collect_lookups (const hb_set_t *feature_indexes,
+			hb_set_t       *lookup_indexes /* OUT */) const
+  {
+    for (const FeatureVariationRecord& r : varRecords)
+      r.collect_lookups (this, feature_indexes, lookup_indexes);
+  }
+
+  void closure_features (const hb_map_t *lookup_indexes,
+			 hb_set_t       *feature_indexes /* OUT */) const
+  {
+    for (const FeatureVariationRecord& record : varRecords)
+      record.closure_features (this, lookup_indexes, feature_indexes);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       hb_subset_layout_context_t *l) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out || !c->serializer->extend_min (out))) return_trace (false);
+
+    out->version.major = version.major;
+    out->version.minor = version.minor;
+
+    + varRecords.iter ()
+    | hb_apply (subset_record_array (l, &(out->varRecords), this))
+    ;
+    return_trace (bool (out->varRecords));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  varRecords.sanitize (c, this));
+  }
+
+  protected:
+  FixedVersion<>	version;	/* Version--0x00010000u */
+  LArrayOf<FeatureVariationRecord>
+			varRecords;
+  public:
+  DEFINE_SIZE_ARRAY_SIZED (8, varRecords);
+};
+
+
+/*
+ * Device Tables
+ */
+
+struct HintingDevice
+{
+  friend struct Device;
+
+  private:
+
+  hb_position_t get_x_delta (hb_font_t *font) const
+  { return get_delta (font->x_ppem, font->x_scale); }
+
+  hb_position_t get_y_delta (hb_font_t *font) const
+  { return get_delta (font->y_ppem, font->y_scale); }
+
+  public:
+
+  unsigned int get_size () const
+  {
+    unsigned int f = deltaFormat;
+    if (unlikely (f < 1 || f > 3 || startSize > endSize)) return 3 * HBUINT16::static_size;
+    return HBUINT16::static_size * (4 + ((endSize - startSize) >> (4 - f)));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && c->check_range (this, this->get_size ()));
+  }
+
+  HintingDevice* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (c->embed<HintingDevice> (this));
+  }
+
+  private:
+
+  int get_delta (unsigned int ppem, int scale) const
+  {
+    if (!ppem) return 0;
+
+    int pixels = get_delta_pixels (ppem);
+
+    if (!pixels) return 0;
+
+    return (int) (pixels * (int64_t) scale / ppem);
+  }
+  int get_delta_pixels (unsigned int ppem_size) const
+  {
+    unsigned int f = deltaFormat;
+    if (unlikely (f < 1 || f > 3))
+      return 0;
+
+    if (ppem_size < startSize || ppem_size > endSize)
+      return 0;
+
+    unsigned int s = ppem_size - startSize;
+
+    unsigned int byte = deltaValueZ[s >> (4 - f)];
+    unsigned int bits = (byte >> (16 - (((s & ((1 << (4 - f)) - 1)) + 1) << f)));
+    unsigned int mask = (0xFFFFu >> (16 - (1 << f)));
+
+    int delta = bits & mask;
+
+    if ((unsigned int) delta >= ((mask + 1) >> 1))
+      delta -= mask + 1;
+
+    return delta;
+  }
+
+  protected:
+  HBUINT16	startSize;		/* Smallest size to correct--in ppem */
+  HBUINT16	endSize;		/* Largest size to correct--in ppem */
+  HBUINT16	deltaFormat;		/* Format of DeltaValue array data: 1, 2, or 3
+					 * 1	Signed 2-bit value, 8 values per uint16
+					 * 2	Signed 4-bit value, 4 values per uint16
+					 * 3	Signed 8-bit value, 2 values per uint16
+					 */
+  UnsizedArrayOf<HBUINT16>
+		deltaValueZ;		/* Array of compressed data */
+  public:
+  DEFINE_SIZE_ARRAY (6, deltaValueZ);
+};
+
+struct VariationDevice
+{
+  friend struct Device;
+
+  private:
+
+  hb_position_t get_x_delta (hb_font_t *font, const VariationStore &store) const
+  { return font->em_scalef_x (get_delta (font, store)); }
+
+  hb_position_t get_y_delta (hb_font_t *font, const VariationStore &store) const
+  { return font->em_scalef_y (get_delta (font, store)); }
+
+  VariationDevice* copy (hb_serialize_context_t *c, const hb_map_t *layout_variation_idx_map) const
+  {
+    TRACE_SERIALIZE (this);
+    auto snap = c->snapshot ();
+    auto *out = c->embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+    if (!layout_variation_idx_map || layout_variation_idx_map->is_empty ()) return_trace (out);
+
+    unsigned org_idx = (outerIndex << 16) + innerIndex;
+    if (!layout_variation_idx_map->has (org_idx))
+    {
+      c->revert (snap);
+      return_trace (nullptr);
+    }
+    unsigned new_idx = layout_variation_idx_map->get (org_idx);
+    out->outerIndex = new_idx >> 16;
+    out->innerIndex = new_idx & 0xFFFF;
+    return_trace (out);
+  }
+
+  void record_variation_index (hb_set_t *layout_variation_indices) const
+  {
+    unsigned var_idx = (outerIndex << 16) + innerIndex;
+    layout_variation_indices->add (var_idx);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  private:
+
+  float get_delta (hb_font_t *font, const VariationStore &store) const
+  {
+    return store.get_delta (outerIndex, innerIndex, font->coords, font->num_coords);
+  }
+
+  protected:
+  HBUINT16	outerIndex;
+  HBUINT16	innerIndex;
+  HBUINT16	deltaFormat;	/* Format identifier for this table: 0x0x8000 */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct DeviceHeader
+{
+  protected:
+  HBUINT16		reserved1;
+  HBUINT16		reserved2;
+  public:
+  HBUINT16		format;		/* Format identifier */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct Device
+{
+  hb_position_t get_x_delta (hb_font_t *font, const VariationStore &store=Null (VariationStore)) const
+  {
+    switch (u.b.format)
+    {
+#ifndef HB_NO_HINTING
+    case 1: case 2: case 3:
+      return u.hinting.get_x_delta (font);
+#endif
+#ifndef HB_NO_VAR
+    case 0x8000:
+      return u.variation.get_x_delta (font, store);
+#endif
+    default:
+      return 0;
+    }
+  }
+  hb_position_t get_y_delta (hb_font_t *font, const VariationStore &store=Null (VariationStore)) const
+  {
+    switch (u.b.format)
+    {
+    case 1: case 2: case 3:
+#ifndef HB_NO_HINTING
+      return u.hinting.get_y_delta (font);
+#endif
+#ifndef HB_NO_VAR
+    case 0x8000:
+      return u.variation.get_y_delta (font, store);
+#endif
+    default:
+      return 0;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.b.format.sanitize (c)) return_trace (false);
+    switch (u.b.format) {
+#ifndef HB_NO_HINTING
+    case 1: case 2: case 3:
+      return_trace (u.hinting.sanitize (c));
+#endif
+#ifndef HB_NO_VAR
+    case 0x8000:
+      return_trace (u.variation.sanitize (c));
+#endif
+    default:
+      return_trace (true);
+    }
+  }
+
+  Device* copy (hb_serialize_context_t *c, const hb_map_t *layout_variation_idx_map=nullptr) const
+  {
+    TRACE_SERIALIZE (this);
+    switch (u.b.format) {
+#ifndef HB_NO_HINTING
+    case 1:
+    case 2:
+    case 3:
+      return_trace (reinterpret_cast<Device *> (u.hinting.copy (c)));
+#endif
+#ifndef HB_NO_VAR
+    case 0x8000:
+      return_trace (reinterpret_cast<Device *> (u.variation.copy (c, layout_variation_idx_map)));
+#endif
+    default:
+      return_trace (nullptr);
+    }
+  }
+
+  void collect_variation_indices (hb_set_t *layout_variation_indices) const
+  {
+    switch (u.b.format) {
+#ifndef HB_NO_HINTING
+    case 1:
+    case 2:
+    case 3:
+      return;
+#endif
+#ifndef HB_NO_VAR
+    case 0x8000:
+      u.variation.record_variation_index (layout_variation_indices);
+      return;
+#endif
+    default:
+      return;
+    }
+  }
+
+  protected:
+  union {
+  DeviceHeader		b;
+  HintingDevice		hinting;
+#ifndef HB_NO_VAR
+  VariationDevice	variation;
+#endif
+  } u;
+  public:
+  DEFINE_SIZE_UNION (6, b);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-gdef-table.hh b/thirdparty/harfbuzz/src/hb-ot-layout-gdef-table.hh
new file mode 100644
index 0000000000..437e760f64
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-gdef-table.hh
@@ -0,0 +1,725 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2010,2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_GDEF_TABLE_HH
+#define HB_OT_LAYOUT_GDEF_TABLE_HH
+
+#include "hb-ot-layout-common.hh"
+
+#include "hb-font.hh"
+
+
+namespace OT {
+
+
+/*
+ * Attachment List Table
+ */
+
+/* Array of contour point indices--in increasing numerical order */
+struct AttachPoint : ArrayOf<HBUINT16>
+{
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out)) return_trace (false);
+
+    return_trace (out->serialize (c->serializer, + iter ()));
+  }
+};
+
+struct AttachList
+{
+  unsigned int get_attach_points (hb_codepoint_t glyph_id,
+				  unsigned int start_offset,
+				  unsigned int *point_count /* IN/OUT */,
+				  unsigned int *point_array /* OUT */) const
+  {
+    unsigned int index = (this+coverage).get_coverage (glyph_id);
+    if (index == NOT_COVERED)
+    {
+      if (point_count)
+	*point_count = 0;
+      return 0;
+    }
+
+    const AttachPoint &points = this+attachPoint[index];
+
+    if (point_count)
+    {
+      + points.sub_array (start_offset, point_count)
+      | hb_sink (hb_array (point_array, *point_count))
+      ;
+    }
+
+    return points.len;
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, attachPoint)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->attachPoint, this), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && attachPoint.sanitize (c, this));
+  }
+
+  protected:
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table -- from
+					 * beginning of AttachList table */
+  OffsetArrayOf<AttachPoint>
+		attachPoint;		/* Array of AttachPoint tables
+					 * in Coverage Index order */
+  public:
+  DEFINE_SIZE_ARRAY (4, attachPoint);
+};
+
+/*
+ * Ligature Caret Table
+ */
+
+struct CaretValueFormat1
+{
+  friend struct CaretValue;
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+    return_trace (true);
+  }
+
+  private:
+  hb_position_t get_caret_value (hb_font_t *font, hb_direction_t direction) const
+  {
+    return HB_DIRECTION_IS_HORIZONTAL (direction) ? font->em_scale_x (coordinate) : font->em_scale_y (coordinate);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBUINT16	caretValueFormat;	/* Format identifier--format = 1 */
+  FWORD		coordinate;		/* X or Y value, in design units */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct CaretValueFormat2
+{
+  friend struct CaretValue;
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+    return_trace (true);
+  }
+
+  private:
+  hb_position_t get_caret_value (hb_font_t *font, hb_direction_t direction, hb_codepoint_t glyph_id) const
+  {
+    hb_position_t x, y;
+    font->get_glyph_contour_point_for_origin (glyph_id, caretValuePoint, direction, &x, &y);
+    return HB_DIRECTION_IS_HORIZONTAL (direction) ? x : y;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBUINT16	caretValueFormat;	/* Format identifier--format = 2 */
+  HBUINT16	caretValuePoint;	/* Contour point index on glyph */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct CaretValueFormat3
+{
+  friend struct CaretValue;
+
+  hb_position_t get_caret_value (hb_font_t *font, hb_direction_t direction,
+				 const VariationStore &var_store) const
+  {
+    return HB_DIRECTION_IS_HORIZONTAL (direction) ?
+	   font->em_scale_x (coordinate) + (this+deviceTable).get_x_delta (font, var_store) :
+	   font->em_scale_y (coordinate) + (this+deviceTable).get_y_delta (font, var_store);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->embed (this);
+    if (unlikely (!out)) return_trace (false);
+
+    return_trace (out->deviceTable.serialize_copy (c->serializer, deviceTable, this, c->serializer->to_bias (out),
+						   hb_serialize_context_t::Head, c->plan->layout_variation_idx_map));
+  }
+
+  void collect_variation_indices (hb_set_t *layout_variation_indices) const
+  { (this+deviceTable).collect_variation_indices (layout_variation_indices); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && deviceTable.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	caretValueFormat;	/* Format identifier--format = 3 */
+  FWORD		coordinate;		/* X or Y value, in design units */
+  OffsetTo<Device>
+		deviceTable;		/* Offset to Device table for X or Y
+					 * value--from beginning of CaretValue
+					 * table */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct CaretValue
+{
+  hb_position_t get_caret_value (hb_font_t *font,
+				 hb_direction_t direction,
+				 hb_codepoint_t glyph_id,
+				 const VariationStore &var_store) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.get_caret_value (font, direction);
+    case 2: return u.format2.get_caret_value (font, direction, glyph_id);
+    case 3: return u.format3.get_caret_value (font, direction, var_store);
+    default:return 0;
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    case 3: return_trace (c->dispatch (u.format3, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  void collect_variation_indices (hb_set_t *layout_variation_indices) const
+  {
+    switch (u.format) {
+    case 1:
+    case 2:
+      return;
+    case 3:
+      u.format3.collect_variation_indices (layout_variation_indices);
+      return;
+    default: return;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    case 3: return_trace (u.format3.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  CaretValueFormat1	format1;
+  CaretValueFormat2	format2;
+  CaretValueFormat3	format3;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+struct LigGlyph
+{
+  unsigned get_lig_carets (hb_font_t            *font,
+			   hb_direction_t        direction,
+			   hb_codepoint_t        glyph_id,
+			   const VariationStore &var_store,
+			   unsigned              start_offset,
+			   unsigned             *caret_count /* IN/OUT */,
+			   hb_position_t        *caret_array /* OUT */) const
+  {
+    if (caret_count)
+    {
+      + carets.sub_array (start_offset, caret_count)
+      | hb_map (hb_add (this))
+      | hb_map ([&] (const CaretValue &value) { return value.get_caret_value (font, direction, glyph_id, var_store); })
+      | hb_sink (hb_array (caret_array, *caret_count))
+      ;
+    }
+
+    return carets.len;
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    + hb_iter (carets)
+    | hb_apply (subset_offset_array (c, out->carets, this))
+    ;
+
+    return_trace (bool (out->carets));
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    for (const OffsetTo<CaretValue>& offset : carets.iter ())
+      (this+offset).collect_variation_indices (c->layout_variation_indices);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (carets.sanitize (c, this));
+  }
+
+  protected:
+  OffsetArrayOf<CaretValue>
+		carets;			/* Offset array of CaretValue tables
+					 * --from beginning of LigGlyph table
+					 * --in increasing coordinate order */
+  public:
+  DEFINE_SIZE_ARRAY (2, carets);
+};
+
+struct LigCaretList
+{
+  unsigned int get_lig_carets (hb_font_t *font,
+			       hb_direction_t direction,
+			       hb_codepoint_t glyph_id,
+			       const VariationStore &var_store,
+			       unsigned int start_offset,
+			       unsigned int *caret_count /* IN/OUT */,
+			       hb_position_t *caret_array /* OUT */) const
+  {
+    unsigned int index = (this+coverage).get_coverage (glyph_id);
+    if (index == NOT_COVERED)
+    {
+      if (caret_count)
+	*caret_count = 0;
+      return 0;
+    }
+    const LigGlyph &lig_glyph = this+ligGlyph[index];
+    return lig_glyph.get_lig_carets (font, direction, glyph_id, var_store, start_offset, caret_count, caret_array);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, ligGlyph)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->ligGlyph, this), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    + hb_zip (this+coverage, ligGlyph)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const LigGlyph& _) { _.collect_variation_indices (c); })
+    ;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && ligGlyph.sanitize (c, this));
+  }
+
+  protected:
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of LigCaretList table */
+  OffsetArrayOf<LigGlyph>
+		ligGlyph;		/* Array of LigGlyph tables
+					 * in Coverage Index order */
+  public:
+  DEFINE_SIZE_ARRAY (4, ligGlyph);
+};
+
+
+struct MarkGlyphSetsFormat1
+{
+  bool covers (unsigned int set_index, hb_codepoint_t glyph_id) const
+  { return (this+coverage[set_index]).get_coverage (glyph_id) != NOT_COVERED; }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    bool ret = true;
+    for (const LOffsetTo<Coverage>& offset : coverage.iter ())
+    {
+      auto *o = out->coverage.serialize_append (c->serializer);
+      if (unlikely (!o))
+      {
+	ret = false;
+	break;
+      }
+
+      //not using o->serialize_subset (c, offset, this, out) here because
+      //OTS doesn't allow null offset.
+      //See issue: https://github.com/khaledhosny/ots/issues/172
+      c->serializer->push ();
+      c->dispatch (this+offset);
+      c->serializer->add_link (*o, c->serializer->pop_pack ());
+    }
+
+    return_trace (ret && out->coverage.len);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  ArrayOf<LOffsetTo<Coverage>>
+		coverage;		/* Array of long offsets to mark set
+					 * coverage tables */
+  public:
+  DEFINE_SIZE_ARRAY (4, coverage);
+};
+
+struct MarkGlyphSets
+{
+  bool covers (unsigned int set_index, hb_codepoint_t glyph_id) const
+  {
+    switch (u.format) {
+    case 1: return u.format1.covers (set_index, glyph_id);
+    default:return false;
+    }
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    switch (u.format) {
+    case 1: return_trace (u.format1.subset (c));
+    default:return_trace (false);
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  MarkGlyphSetsFormat1	format1;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+
+/*
+ * GDEF -- Glyph Definition
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/gdef
+ */
+
+
+struct GDEF
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_GDEF;
+
+  enum GlyphClasses {
+    UnclassifiedGlyph	= 0,
+    BaseGlyph		= 1,
+    LigatureGlyph	= 2,
+    MarkGlyph		= 3,
+    ComponentGlyph	= 4
+  };
+
+  bool has_data () const { return version.to_int (); }
+  bool has_glyph_classes () const { return glyphClassDef != 0; }
+  unsigned int get_glyph_class (hb_codepoint_t glyph) const
+  { return (this+glyphClassDef).get_class (glyph); }
+  void get_glyphs_in_class (unsigned int klass, hb_set_t *glyphs) const
+  { (this+glyphClassDef).collect_class (glyphs, klass); }
+
+  bool has_mark_attachment_types () const { return markAttachClassDef != 0; }
+  unsigned int get_mark_attachment_type (hb_codepoint_t glyph) const
+  { return (this+markAttachClassDef).get_class (glyph); }
+
+  bool has_attach_points () const { return attachList != 0; }
+  unsigned int get_attach_points (hb_codepoint_t glyph_id,
+				  unsigned int start_offset,
+				  unsigned int *point_count /* IN/OUT */,
+				  unsigned int *point_array /* OUT */) const
+  { return (this+attachList).get_attach_points (glyph_id, start_offset, point_count, point_array); }
+
+  bool has_lig_carets () const { return ligCaretList != 0; }
+  unsigned int get_lig_carets (hb_font_t *font,
+			       hb_direction_t direction,
+			       hb_codepoint_t glyph_id,
+			       unsigned int start_offset,
+			       unsigned int *caret_count /* IN/OUT */,
+			       hb_position_t *caret_array /* OUT */) const
+  { return (this+ligCaretList).get_lig_carets (font,
+					       direction, glyph_id, get_var_store(),
+					       start_offset, caret_count, caret_array); }
+
+  bool has_mark_sets () const { return version.to_int () >= 0x00010002u && markGlyphSetsDef != 0; }
+  bool mark_set_covers (unsigned int set_index, hb_codepoint_t glyph_id) const
+  { return version.to_int () >= 0x00010002u && (this+markGlyphSetsDef).covers (set_index, glyph_id); }
+
+  bool has_var_store () const { return version.to_int () >= 0x00010003u && varStore != 0; }
+  const VariationStore &get_var_store () const
+  { return version.to_int () >= 0x00010003u ? this+varStore : Null (VariationStore); }
+
+  /* glyph_props is a 16-bit integer where the lower 8-bit have bits representing
+   * glyph class and other bits, and high 8-bit the mark attachment type (if any).
+   * Not to be confused with lookup_props which is very similar. */
+  unsigned int get_glyph_props (hb_codepoint_t glyph) const
+  {
+    unsigned int klass = get_glyph_class (glyph);
+
+    static_assert (((unsigned int) HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH == (unsigned int) LookupFlag::IgnoreBaseGlyphs), "");
+    static_assert (((unsigned int) HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE == (unsigned int) LookupFlag::IgnoreLigatures), "");
+    static_assert (((unsigned int) HB_OT_LAYOUT_GLYPH_PROPS_MARK == (unsigned int) LookupFlag::IgnoreMarks), "");
+
+    switch (klass) {
+    default:			return 0;
+    case BaseGlyph:		return HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH;
+    case LigatureGlyph:		return HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE;
+    case MarkGlyph:
+	  klass = get_mark_attachment_type (glyph);
+	  return HB_OT_LAYOUT_GLYPH_PROPS_MARK | (klass << 8);
+    }
+  }
+
+  HB_INTERNAL bool is_blocklisted (hb_blob_t *blob,
+				   hb_face_t *face) const;
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      this->table = hb_sanitize_context_t ().reference_table<GDEF> (face);
+      if (unlikely (this->table->is_blocklisted (this->table.get_blob (), face)))
+      {
+	hb_blob_destroy (this->table.get_blob ());
+	this->table = hb_blob_get_empty ();
+      }
+    }
+
+    void fini () { this->table.destroy (); }
+
+    hb_blob_ptr_t<GDEF> table;
+  };
+
+  unsigned int get_size () const
+  {
+    return min_size +
+	   (version.to_int () >= 0x00010002u ? markGlyphSetsDef.static_size : 0) +
+	   (version.to_int () >= 0x00010003u ? varStore.static_size : 0);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  { (this+ligCaretList).collect_variation_indices (c); }
+
+  void remap_layout_variation_indices (const hb_set_t *layout_variation_indices,
+				       hb_map_t *layout_variation_idx_map /* OUT */) const
+  {
+    if (version.to_int () < 0x00010003u || !varStore) return;
+    if (layout_variation_indices->is_empty ()) return;
+
+    unsigned new_major = 0, new_minor = 0;
+    unsigned last_major = (layout_variation_indices->get_min ()) >> 16;
+    for (unsigned idx : layout_variation_indices->iter ())
+    {
+      uint16_t major = idx >> 16;
+      if (major >= (this+varStore).get_sub_table_count ()) break;
+      if (major != last_major)
+      {
+	new_minor = 0;
+	++new_major;
+      }
+
+      unsigned new_idx = (new_major << 16) + new_minor;
+      layout_variation_idx_map->set (idx, new_idx);
+      ++new_minor;
+      last_major = major;
+    }
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->embed (*this);
+    if (unlikely (!out)) return_trace (false);
+
+    bool subset_glyphclassdef = out->glyphClassDef.serialize_subset (c, glyphClassDef, this);
+    bool subset_attachlist = out->attachList.serialize_subset (c, attachList, this);
+    bool subset_ligcaretlist = out->ligCaretList.serialize_subset (c, ligCaretList, this);
+    bool subset_markattachclassdef = out->markAttachClassDef.serialize_subset (c, markAttachClassDef, this);
+
+    bool subset_markglyphsetsdef = true;
+    if (version.to_int () >= 0x00010002u)
+    {
+      subset_markglyphsetsdef = out->markGlyphSetsDef.serialize_subset (c, markGlyphSetsDef, this);
+      if (!subset_markglyphsetsdef &&
+	  version.to_int () == 0x00010002u)
+	out->version.minor = 0;
+    }
+
+    bool subset_varstore = true;
+    if (version.to_int () >= 0x00010003u)
+    {
+      subset_varstore = out->varStore.serialize_subset (c, varStore, this);
+      if (!subset_varstore && version.to_int () == 0x00010003u)
+	out->version.minor = 2;
+    }
+
+    return_trace (subset_glyphclassdef || subset_attachlist ||
+		  subset_ligcaretlist || subset_markattachclassdef ||
+		  (out->version.to_int () >= 0x00010002u && subset_markglyphsetsdef) ||
+		  (out->version.to_int () >= 0x00010003u && subset_varstore));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  glyphClassDef.sanitize (c, this) &&
+		  attachList.sanitize (c, this) &&
+		  ligCaretList.sanitize (c, this) &&
+		  markAttachClassDef.sanitize (c, this) &&
+		  (version.to_int () < 0x00010002u || markGlyphSetsDef.sanitize (c, this)) &&
+		  (version.to_int () < 0x00010003u || varStore.sanitize (c, this)));
+  }
+
+  protected:
+  FixedVersion<>version;		/* Version of the GDEF table--currently
+					 * 0x00010003u */
+  OffsetTo<ClassDef>
+		glyphClassDef;		/* Offset to class definition table
+					 * for glyph type--from beginning of
+					 * GDEF header (may be Null) */
+  OffsetTo<AttachList>
+		attachList;		/* Offset to list of glyphs with
+					 * attachment points--from beginning
+					 * of GDEF header (may be Null) */
+  OffsetTo<LigCaretList>
+		ligCaretList;		/* Offset to list of positioning points
+					 * for ligature carets--from beginning
+					 * of GDEF header (may be Null) */
+  OffsetTo<ClassDef>
+		markAttachClassDef;	/* Offset to class definition table for
+					 * mark attachment type--from beginning
+					 * of GDEF header (may be Null) */
+  OffsetTo<MarkGlyphSets>
+		markGlyphSetsDef;	/* Offset to the table of mark set
+					 * definitions--from beginning of GDEF
+					 * header (may be NULL).  Introduced
+					 * in version 0x00010002. */
+  LOffsetTo<VariationStore>
+		varStore;		/* Offset to the table of Item Variation
+					 * Store--from beginning of GDEF
+					 * header (may be NULL).  Introduced
+					 * in version 0x00010003. */
+  public:
+  DEFINE_SIZE_MIN (12);
+};
+
+struct GDEF_accelerator_t : GDEF::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_GDEF_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-gpos-table.hh b/thirdparty/harfbuzz/src/hb-ot-layout-gpos-table.hh
new file mode 100644
index 0000000000..2217d298fb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-gpos-table.hh
@@ -0,0 +1,2740 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_GPOS_TABLE_HH
+#define HB_OT_LAYOUT_GPOS_TABLE_HH
+
+#include "hb-ot-layout-gsubgpos.hh"
+
+
+namespace OT {
+
+struct MarkArray;
+static void Markclass_closure_and_remap_indexes (const Coverage  &mark_coverage,
+						 const MarkArray &mark_array,
+						 const hb_set_t  &glyphset,
+						 hb_map_t*        klass_mapping /* INOUT */);
+
+/* buffer **position** var allocations */
+#define attach_chain() var.i16[0] /* glyph to which this attaches to, relative to current glyphs; negative for going back, positive for forward. */
+#define attach_type() var.u8[2] /* attachment type */
+/* Note! if attach_chain() is zero, the value of attach_type() is irrelevant. */
+
+enum attach_type_t {
+  ATTACH_TYPE_NONE	= 0X00,
+
+  /* Each attachment should be either a mark or a cursive; can't be both. */
+  ATTACH_TYPE_MARK	= 0X01,
+  ATTACH_TYPE_CURSIVE	= 0X02,
+};
+
+
+/* Shared Tables: ValueRecord, Anchor Table, and MarkArray */
+
+typedef HBUINT16 Value;
+
+typedef UnsizedArrayOf<Value> ValueRecord;
+
+struct ValueFormat : HBUINT16
+{
+  enum Flags {
+    xPlacement	= 0x0001u,	/* Includes horizontal adjustment for placement */
+    yPlacement	= 0x0002u,	/* Includes vertical adjustment for placement */
+    xAdvance	= 0x0004u,	/* Includes horizontal adjustment for advance */
+    yAdvance	= 0x0008u,	/* Includes vertical adjustment for advance */
+    xPlaDevice	= 0x0010u,	/* Includes horizontal Device table for placement */
+    yPlaDevice	= 0x0020u,	/* Includes vertical Device table for placement */
+    xAdvDevice	= 0x0040u,	/* Includes horizontal Device table for advance */
+    yAdvDevice	= 0x0080u,	/* Includes vertical Device table for advance */
+    ignored	= 0x0F00u,	/* Was used in TrueType Open for MM fonts */
+    reserved	= 0xF000u,	/* For future use */
+
+    devices	= 0x00F0u	/* Mask for having any Device table */
+  };
+
+/* All fields are options.  Only those available advance the value pointer. */
+#if 0
+  HBINT16		xPlacement;	/* Horizontal adjustment for
+					 * placement--in design units */
+  HBINT16		yPlacement;	/* Vertical adjustment for
+					 * placement--in design units */
+  HBINT16		xAdvance;	/* Horizontal adjustment for
+					 * advance--in design units (only used
+					 * for horizontal writing) */
+  HBINT16		yAdvance;	/* Vertical adjustment for advance--in
+					 * design units (only used for vertical
+					 * writing) */
+  OffsetTo<Device>	xPlaDevice;	/* Offset to Device table for
+					 * horizontal placement--measured from
+					 * beginning of PosTable (may be NULL) */
+  OffsetTo<Device>	yPlaDevice;	/* Offset to Device table for vertical
+					 * placement--measured from beginning
+					 * of PosTable (may be NULL) */
+  OffsetTo<Device>	xAdvDevice;	/* Offset to Device table for
+					 * horizontal advance--measured from
+					 * beginning of PosTable (may be NULL) */
+  OffsetTo<Device>	yAdvDevice;	/* Offset to Device table for vertical
+					 * advance--measured from beginning of
+					 * PosTable (may be NULL) */
+#endif
+
+  unsigned int get_len () const  { return hb_popcount ((unsigned int) *this); }
+  unsigned int get_size () const { return get_len () * Value::static_size; }
+
+  bool apply_value (hb_ot_apply_context_t *c,
+		    const void            *base,
+		    const Value           *values,
+		    hb_glyph_position_t   &glyph_pos) const
+  {
+    bool ret = false;
+    unsigned int format = *this;
+    if (!format) return ret;
+
+    hb_font_t *font = c->font;
+    bool horizontal = HB_DIRECTION_IS_HORIZONTAL (c->direction);
+
+    if (format & xPlacement) glyph_pos.x_offset  += font->em_scale_x (get_short (values++, &ret));
+    if (format & yPlacement) glyph_pos.y_offset  += font->em_scale_y (get_short (values++, &ret));
+    if (format & xAdvance) {
+      if (likely (horizontal)) glyph_pos.x_advance += font->em_scale_x (get_short (values, &ret));
+      values++;
+    }
+    /* y_advance values grow downward but font-space grows upward, hence negation */
+    if (format & yAdvance) {
+      if (unlikely (!horizontal)) glyph_pos.y_advance -= font->em_scale_y (get_short (values, &ret));
+      values++;
+    }
+
+    if (!has_device ()) return ret;
+
+    bool use_x_device = font->x_ppem || font->num_coords;
+    bool use_y_device = font->y_ppem || font->num_coords;
+
+    if (!use_x_device && !use_y_device) return ret;
+
+    const VariationStore &store = c->var_store;
+
+    /* pixel -> fractional pixel */
+    if (format & xPlaDevice) {
+      if (use_x_device) glyph_pos.x_offset  += (base + get_device (values, &ret)).get_x_delta (font, store);
+      values++;
+    }
+    if (format & yPlaDevice) {
+      if (use_y_device) glyph_pos.y_offset  += (base + get_device (values, &ret)).get_y_delta (font, store);
+      values++;
+    }
+    if (format & xAdvDevice) {
+      if (horizontal && use_x_device) glyph_pos.x_advance += (base + get_device (values, &ret)).get_x_delta (font, store);
+      values++;
+    }
+    if (format & yAdvDevice) {
+      /* y_advance values grow downward but font-space grows upward, hence negation */
+      if (!horizontal && use_y_device) glyph_pos.y_advance -= (base + get_device (values, &ret)).get_y_delta (font, store);
+      values++;
+    }
+    return ret;
+  }
+
+  void serialize_copy (hb_serialize_context_t *c, const void *base,
+		       const Value *values, const hb_map_t *layout_variation_idx_map) const
+  {
+    unsigned int format = *this;
+    if (!format) return;
+
+    if (format & xPlacement) c->copy (*values++);
+    if (format & yPlacement) c->copy (*values++);
+    if (format & xAdvance)   c->copy (*values++);
+    if (format & yAdvance)   c->copy (*values++);
+
+    if (format & xPlaDevice) copy_device (c, base, values++, layout_variation_idx_map);
+    if (format & yPlaDevice) copy_device (c, base, values++, layout_variation_idx_map);
+    if (format & xAdvDevice) copy_device (c, base, values++, layout_variation_idx_map);
+    if (format & yAdvDevice) copy_device (c, base, values++, layout_variation_idx_map);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  const void *base,
+				  const hb_array_t<const Value>& values) const
+  {
+    unsigned format = *this;
+    unsigned i = 0;
+    if (format & xPlacement) i++;
+    if (format & yPlacement) i++;
+    if (format & xAdvance) i++;
+    if (format & yAdvance) i++;
+    if (format & xPlaDevice)
+    {
+      (base + get_device (&(values[i]))).collect_variation_indices (c->layout_variation_indices);
+      i++;
+    }
+
+    if (format & ValueFormat::yPlaDevice)
+    {
+      (base + get_device (&(values[i]))).collect_variation_indices (c->layout_variation_indices);
+      i++;
+    }
+
+    if (format & ValueFormat::xAdvDevice)
+    {
+
+      (base + get_device (&(values[i]))).collect_variation_indices (c->layout_variation_indices);
+      i++;
+    }
+
+    if (format & ValueFormat::yAdvDevice)
+    {
+
+      (base + get_device (&(values[i]))).collect_variation_indices (c->layout_variation_indices);
+      i++;
+    }
+  }
+
+  private:
+  bool sanitize_value_devices (hb_sanitize_context_t *c, const void *base, const Value *values) const
+  {
+    unsigned int format = *this;
+
+    if (format & xPlacement) values++;
+    if (format & yPlacement) values++;
+    if (format & xAdvance)   values++;
+    if (format & yAdvance)   values++;
+
+    if ((format & xPlaDevice) && !get_device (values++).sanitize (c, base)) return false;
+    if ((format & yPlaDevice) && !get_device (values++).sanitize (c, base)) return false;
+    if ((format & xAdvDevice) && !get_device (values++).sanitize (c, base)) return false;
+    if ((format & yAdvDevice) && !get_device (values++).sanitize (c, base)) return false;
+
+    return true;
+  }
+
+  static inline OffsetTo<Device>& get_device (Value* value)
+  {
+    return *static_cast<OffsetTo<Device> *> (value);
+  }
+  static inline const OffsetTo<Device>& get_device (const Value* value, bool *worked=nullptr)
+  {
+    if (worked) *worked |= bool (*value);
+    return *static_cast<const OffsetTo<Device> *> (value);
+  }
+
+  bool copy_device (hb_serialize_context_t *c, const void *base,
+		    const Value *src_value, const hb_map_t *layout_variation_idx_map) const
+  {
+    Value	*dst_value = c->copy (*src_value);
+
+    if (!dst_value) return false;
+    if (*dst_value == 0) return true;
+
+    *dst_value = 0;
+    c->push ();
+    if ((base + get_device (src_value)).copy (c, layout_variation_idx_map))
+    {
+      c->add_link (*dst_value, c->pop_pack ());
+      return true;
+    }
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  static inline const HBINT16& get_short (const Value* value, bool *worked=nullptr)
+  {
+    if (worked) *worked |= bool (*value);
+    return *reinterpret_cast<const HBINT16 *> (value);
+  }
+
+  public:
+
+  bool has_device () const
+  {
+    unsigned int format = *this;
+    return (format & devices) != 0;
+  }
+
+  bool sanitize_value (hb_sanitize_context_t *c, const void *base, const Value *values) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_range (values, get_size ()) && (!has_device () || sanitize_value_devices (c, base, values)));
+  }
+
+  bool sanitize_values (hb_sanitize_context_t *c, const void *base, const Value *values, unsigned int count) const
+  {
+    TRACE_SANITIZE (this);
+    unsigned int len = get_len ();
+
+    if (!c->check_range (values, count, get_size ())) return_trace (false);
+
+    if (!has_device ()) return_trace (true);
+
+    for (unsigned int i = 0; i < count; i++) {
+      if (!sanitize_value_devices (c, base, values))
+	return_trace (false);
+      values += len;
+    }
+
+    return_trace (true);
+  }
+
+  /* Just sanitize referenced Device tables.  Doesn't check the values themselves. */
+  bool sanitize_values_stride_unsafe (hb_sanitize_context_t *c, const void *base, const Value *values, unsigned int count, unsigned int stride) const
+  {
+    TRACE_SANITIZE (this);
+
+    if (!has_device ()) return_trace (true);
+
+    for (unsigned int i = 0; i < count; i++) {
+      if (!sanitize_value_devices (c, base, values))
+	return_trace (false);
+      values += stride;
+    }
+
+    return_trace (true);
+  }
+};
+
+template<typename Iterator>
+static void SinglePos_serialize (hb_serialize_context_t *c,
+				 const void *src,
+				 Iterator it,
+				 ValueFormat valFormat,
+				 const hb_map_t *layout_variation_idx_map);
+
+
+struct AnchorFormat1
+{
+  void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id HB_UNUSED,
+		   float *x, float *y) const
+  {
+    hb_font_t *font = c->font;
+    *x = font->em_fscale_x (xCoordinate);
+    *y = font->em_fscale_y (yCoordinate);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  AnchorFormat1* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (c->embed<AnchorFormat1> (this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  FWORD		xCoordinate;		/* Horizontal value--in design units */
+  FWORD		yCoordinate;		/* Vertical value--in design units */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct AnchorFormat2
+{
+  void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id,
+		   float *x, float *y) const
+  {
+    hb_font_t *font = c->font;
+
+#ifdef HB_NO_HINTING
+    *x = font->em_fscale_x (xCoordinate);
+    *y = font->em_fscale_y (yCoordinate);
+    return;
+#endif
+
+    unsigned int x_ppem = font->x_ppem;
+    unsigned int y_ppem = font->y_ppem;
+    hb_position_t cx = 0, cy = 0;
+    bool ret;
+
+    ret = (x_ppem || y_ppem) &&
+	  font->get_glyph_contour_point_for_origin (glyph_id, anchorPoint, HB_DIRECTION_LTR, &cx, &cy);
+    *x = ret && x_ppem ? cx : font->em_fscale_x (xCoordinate);
+    *y = ret && y_ppem ? cy : font->em_fscale_y (yCoordinate);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  AnchorFormat2* copy (hb_serialize_context_t *c) const
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (c->embed<AnchorFormat2> (this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  FWORD		xCoordinate;		/* Horizontal value--in design units */
+  FWORD		yCoordinate;		/* Vertical value--in design units */
+  HBUINT16	anchorPoint;		/* Index to glyph contour point */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct AnchorFormat3
+{
+  void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id HB_UNUSED,
+		   float *x, float *y) const
+  {
+    hb_font_t *font = c->font;
+    *x = font->em_fscale_x (xCoordinate);
+    *y = font->em_fscale_y (yCoordinate);
+
+    if (font->x_ppem || font->num_coords)
+      *x += (this+xDeviceTable).get_x_delta (font, c->var_store);
+    if (font->y_ppem || font->num_coords)
+      *y += (this+yDeviceTable).get_y_delta (font, c->var_store);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && xDeviceTable.sanitize (c, this) && yDeviceTable.sanitize (c, this));
+  }
+
+  AnchorFormat3* copy (hb_serialize_context_t *c,
+		       const hb_map_t *layout_variation_idx_map) const
+  {
+    TRACE_SERIALIZE (this);
+    if (!layout_variation_idx_map) return_trace (nullptr);
+
+    auto *out = c->embed<AnchorFormat3> (this);
+    if (unlikely (!out)) return_trace (nullptr);
+
+    out->xDeviceTable.serialize_copy (c, xDeviceTable, this, 0, hb_serialize_context_t::Head, layout_variation_idx_map);
+    out->yDeviceTable.serialize_copy (c, yDeviceTable, this, 0, hb_serialize_context_t::Head, layout_variation_idx_map);
+    return_trace (out);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    (this+xDeviceTable).collect_variation_indices (c->layout_variation_indices);
+    (this+yDeviceTable).collect_variation_indices (c->layout_variation_indices);
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 3 */
+  FWORD		xCoordinate;		/* Horizontal value--in design units */
+  FWORD		yCoordinate;		/* Vertical value--in design units */
+  OffsetTo<Device>
+		xDeviceTable;		/* Offset to Device table for X
+					 * coordinate-- from beginning of
+					 * Anchor table (may be NULL) */
+  OffsetTo<Device>
+		yDeviceTable;		/* Offset to Device table for Y
+					 * coordinate-- from beginning of
+					 * Anchor table (may be NULL) */
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+struct Anchor
+{
+  void get_anchor (hb_ot_apply_context_t *c, hb_codepoint_t glyph_id,
+		   float *x, float *y) const
+  {
+    *x = *y = 0;
+    switch (u.format) {
+    case 1: u.format1.get_anchor (c, glyph_id, x, y); return;
+    case 2: u.format2.get_anchor (c, glyph_id, x, y); return;
+    case 3: u.format3.get_anchor (c, glyph_id, x, y); return;
+    default:					      return;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!u.format.sanitize (c)) return_trace (false);
+    switch (u.format) {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    case 3: return_trace (u.format3.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  Anchor* copy (hb_serialize_context_t *c, const hb_map_t *layout_variation_idx_map) const
+  {
+    TRACE_SERIALIZE (this);
+    switch (u.format) {
+    case 1: return_trace (reinterpret_cast<Anchor *> (u.format1.copy (c)));
+    case 2: return_trace (reinterpret_cast<Anchor *> (u.format2.copy (c)));
+    case 3: return_trace (reinterpret_cast<Anchor *> (u.format3.copy (c, layout_variation_idx_map)));
+    default:return_trace (nullptr);
+    }
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    switch (u.format) {
+    case 1: case 2:
+      return;
+    case 3:
+      u.format3.collect_variation_indices (c);
+      return;
+    default: return;
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  AnchorFormat1		format1;
+  AnchorFormat2		format2;
+  AnchorFormat3		format3;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+
+struct AnchorMatrix
+{
+  const Anchor& get_anchor (unsigned int row, unsigned int col,
+			    unsigned int cols, bool *found) const
+  {
+    *found = false;
+    if (unlikely (row >= rows || col >= cols)) return Null (Anchor);
+    *found = !matrixZ[row * cols + col].is_null ();
+    return this+matrixZ[row * cols + col];
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  Iterator index_iter) const
+  {
+    for (unsigned i : index_iter)
+      (this+matrixZ[i]).collect_variation_indices (c);
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  bool serialize (hb_serialize_context_t *c,
+		  unsigned                num_rows,
+		  AnchorMatrix const     *offset_matrix,
+		  const hb_map_t         *layout_variation_idx_map,
+		  Iterator                index_iter)
+  {
+    TRACE_SERIALIZE (this);
+    if (!index_iter) return_trace (false);
+    if (unlikely (!c->extend_min ((*this))))  return_trace (false);
+
+    this->rows = num_rows;
+    for (const unsigned i : index_iter)
+    {
+      auto *offset = c->embed (offset_matrix->matrixZ[i]);
+      if (!offset) return_trace (false);
+      offset->serialize_copy (c, offset_matrix->matrixZ[i],
+			      offset_matrix, c->to_bias (this),
+			      hb_serialize_context_t::Head,
+			      layout_variation_idx_map);
+    }
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int cols) const
+  {
+    TRACE_SANITIZE (this);
+    if (!c->check_struct (this)) return_trace (false);
+    if (unlikely (hb_unsigned_mul_overflows (rows, cols))) return_trace (false);
+    unsigned int count = rows * cols;
+    if (!c->check_array (matrixZ.arrayZ, count)) return_trace (false);
+    for (unsigned int i = 0; i < count; i++)
+      if (!matrixZ[i].sanitize (c, this)) return_trace (false);
+    return_trace (true);
+  }
+
+  HBUINT16	rows;			/* Number of rows */
+  UnsizedArrayOf<OffsetTo<Anchor>>
+		matrixZ;		/* Matrix of offsets to Anchor tables--
+					 * from beginning of AnchorMatrix table */
+  public:
+  DEFINE_SIZE_ARRAY (2, matrixZ);
+};
+
+
+struct MarkRecord
+{
+  friend struct MarkArray;
+
+  unsigned get_class () const { return (unsigned) klass; }
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && markAnchor.sanitize (c, base));
+  }
+
+  MarkRecord *copy (hb_serialize_context_t *c,
+		    const void             *src_base,
+		    unsigned                dst_bias,
+		    const hb_map_t         *klass_mapping,
+		    const hb_map_t         *layout_variation_idx_map) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+
+    out->klass = klass_mapping->get (klass);
+    out->markAnchor.serialize_copy (c, markAnchor, src_base, dst_bias, hb_serialize_context_t::Head, layout_variation_idx_map);
+    return_trace (out);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  const void *src_base) const
+  {
+    (src_base+markAnchor).collect_variation_indices (c);
+  }
+
+  protected:
+  HBUINT16	klass;			/* Class defined for this mark */
+  OffsetTo<Anchor>
+		markAnchor;		/* Offset to Anchor table--from
+					 * beginning of MarkArray table */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct MarkArray : ArrayOf<MarkRecord>	/* Array of MarkRecords--in Coverage order */
+{
+  bool apply (hb_ot_apply_context_t *c,
+	      unsigned int mark_index, unsigned int glyph_index,
+	      const AnchorMatrix &anchors, unsigned int class_count,
+	      unsigned int glyph_pos) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    const MarkRecord &record = ArrayOf<MarkRecord>::operator[](mark_index);
+    unsigned int mark_class = record.klass;
+
+    const Anchor& mark_anchor = this + record.markAnchor;
+    bool found;
+    const Anchor& glyph_anchor = anchors.get_anchor (glyph_index, mark_class, class_count, &found);
+    /* If this subtable doesn't have an anchor for this base and this class,
+     * return false such that the subsequent subtables have a chance at it. */
+    if (unlikely (!found)) return_trace (false);
+
+    float mark_x, mark_y, base_x, base_y;
+
+    buffer->unsafe_to_break (glyph_pos, buffer->idx);
+    mark_anchor.get_anchor (c, buffer->cur().codepoint, &mark_x, &mark_y);
+    glyph_anchor.get_anchor (c, buffer->info[glyph_pos].codepoint, &base_x, &base_y);
+
+    hb_glyph_position_t &o = buffer->cur_pos();
+    o.x_offset = roundf (base_x - mark_x);
+    o.y_offset = roundf (base_y - mark_y);
+    o.attach_type() = ATTACH_TYPE_MARK;
+    o.attach_chain() = (int) glyph_pos - (int) buffer->idx;
+    buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+
+    buffer->idx++;
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_source_of (Iterator, MarkRecord))>
+  bool serialize (hb_serialize_context_t *c,
+		  const hb_map_t         *klass_mapping,
+		  const hb_map_t         *layout_variation_idx_map,
+		  const void             *base,
+		  Iterator                it)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!c->check_assign (len, it.len ()))) return_trace (false);
+    c->copy_all (it, base, c->to_bias (this), klass_mapping, layout_variation_idx_map);
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (ArrayOf<MarkRecord>::sanitize (c, this));
+  }
+};
+
+
+/* Lookups */
+
+struct SinglePosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    if (!valueFormat.has_device ()) return;
+
+    auto it =
+    + hb_iter (this+coverage)
+    | hb_filter (c->glyph_set)
+    ;
+
+    if (!it) return;
+    valueFormat.collect_variation_indices (c, this, values.as_array (valueFormat.get_len ()));
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int index = (this+coverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    valueFormat.apply_value (c, this, values, buffer->cur_pos());
+
+    buffer->idx++;
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  const void *src,
+		  Iterator it,
+		  ValueFormat valFormat,
+		  const hb_map_t *layout_variation_idx_map)
+  {
+    auto out = c->extend_min (*this);
+    if (unlikely (!out)) return;
+    if (unlikely (!c->check_assign (valueFormat, valFormat))) return;
+
+    + it
+    | hb_map (hb_second)
+    | hb_apply ([&] (hb_array_t<const Value> _)
+		{ valFormat.serialize_copy (c, src, &_, layout_variation_idx_map); })
+    ;
+
+    auto glyphs =
+    + it
+    | hb_map_retains_sorting (hb_first)
+    ;
+
+    coverage.serialize (c, this).serialize (c, glyphs);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto it =
+    + hb_iter (this+coverage)
+    | hb_filter (glyphset)
+    | hb_map_retains_sorting (glyph_map)
+    | hb_zip (hb_repeat (values.as_array (valueFormat.get_len ())))
+    ;
+
+    bool ret = bool (it);
+    SinglePos_serialize (c->serializer, this, it, valueFormat, c->plan->layout_variation_idx_map);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  coverage.sanitize (c, this) &&
+		  valueFormat.sanitize_value (c, this, values));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of subtable */
+  ValueFormat	valueFormat;		/* Defines the types of data in the
+					 * ValueRecord */
+  ValueRecord	values;			/* Defines positioning
+					 * value(s)--applied to all glyphs in
+					 * the Coverage table */
+  public:
+  DEFINE_SIZE_ARRAY (6, values);
+};
+
+struct SinglePosFormat2
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    if (!valueFormat.has_device ()) return;
+
+    auto it =
+    + hb_zip (this+coverage, hb_range ((unsigned) valueCount))
+    | hb_filter (c->glyph_set, hb_first)
+    ;
+
+    if (!it) return;
+
+    unsigned sub_length = valueFormat.get_len ();
+    const hb_array_t<const Value> values_array = values.as_array (valueCount * sub_length);
+
+    for (unsigned i : + it
+		      | hb_map (hb_second))
+      valueFormat.collect_variation_indices (c, this, values_array.sub_array (i * sub_length, sub_length));
+
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int index = (this+coverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    if (likely (index >= valueCount)) return_trace (false);
+
+    valueFormat.apply_value (c, this,
+			     &values[index * valueFormat.get_len ()],
+			     buffer->cur_pos());
+
+    buffer->idx++;
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  const void *src,
+		  Iterator it,
+		  ValueFormat valFormat,
+		  const hb_map_t *layout_variation_idx_map)
+  {
+    auto out = c->extend_min (*this);
+    if (unlikely (!out)) return;
+    if (unlikely (!c->check_assign (valueFormat, valFormat))) return;
+    if (unlikely (!c->check_assign (valueCount, it.len ()))) return;
+
+    + it
+    | hb_map (hb_second)
+    | hb_apply ([&] (hb_array_t<const Value> _)
+		{ valFormat.serialize_copy (c, src, &_, layout_variation_idx_map); })
+    ;
+
+    auto glyphs =
+    + it
+    | hb_map_retains_sorting (hb_first)
+    ;
+
+    coverage.serialize (c, this).serialize (c, glyphs);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    unsigned sub_length = valueFormat.get_len ();
+    auto values_array = values.as_array (valueCount * sub_length);
+
+    auto it =
+    + hb_zip (this+coverage, hb_range ((unsigned) valueCount))
+    | hb_filter (glyphset, hb_first)
+    | hb_map_retains_sorting ([&] (const hb_pair_t<hb_codepoint_t, unsigned>& _)
+			      {
+				return hb_pair (glyph_map[_.first],
+						values_array.sub_array (_.second * sub_length,
+									sub_length));
+			      })
+    ;
+
+    bool ret = bool (it);
+    SinglePos_serialize (c->serializer, this, it, valueFormat, c->plan->layout_variation_idx_map);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  coverage.sanitize (c, this) &&
+		  valueFormat.sanitize_values (c, this, values, valueCount));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of subtable */
+  ValueFormat	valueFormat;		/* Defines the types of data in the
+					 * ValueRecord */
+  HBUINT16	valueCount;		/* Number of ValueRecords */
+  ValueRecord	values;			/* Array of ValueRecords--positioning
+					 * values applied to glyphs */
+  public:
+  DEFINE_SIZE_ARRAY (8, values);
+};
+
+struct SinglePos
+{
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  unsigned get_format (Iterator glyph_val_iter_pairs)
+  {
+    hb_array_t<const Value> first_val_iter = hb_second (*glyph_val_iter_pairs);
+
+    for (const auto iter : glyph_val_iter_pairs)
+      for (const auto _ : hb_zip (iter.second, first_val_iter))
+	if (_.first != _.second)
+	  return 2;
+
+    return 1;
+  }
+
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  const void *src,
+		  Iterator glyph_val_iter_pairs,
+		  ValueFormat valFormat,
+		  const hb_map_t *layout_variation_idx_map)
+  {
+    if (unlikely (!c->extend_min (u.format))) return;
+    unsigned format = 2;
+
+    if (glyph_val_iter_pairs) format = get_format (glyph_val_iter_pairs);
+
+    u.format = format;
+    switch (u.format) {
+    case 1: u.format1.serialize (c, src, glyph_val_iter_pairs, valFormat, layout_variation_idx_map);
+	    return;
+    case 2: u.format2.serialize (c, src, glyph_val_iter_pairs, valFormat, layout_variation_idx_map);
+	    return;
+    default:return;
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  SinglePosFormat1	format1;
+  SinglePosFormat2	format2;
+  } u;
+};
+
+template<typename Iterator>
+static void
+SinglePos_serialize (hb_serialize_context_t *c,
+		     const void *src,
+		     Iterator it,
+		     ValueFormat valFormat,
+		     const hb_map_t *layout_variation_idx_map)
+{ c->start_embed<SinglePos> ()->serialize (c, src, it, valFormat, layout_variation_idx_map); }
+
+
+struct PairValueRecord
+{
+  friend struct PairSet;
+
+  int cmp (hb_codepoint_t k) const
+  { return secondGlyph.cmp (k); }
+
+  struct serialize_closure_t
+  {
+    const void 		*base;
+    const ValueFormat	*valueFormats;
+    unsigned		len1; /* valueFormats[0].get_len() */
+    const hb_map_t 	*glyph_map;
+    const hb_map_t      *layout_variation_idx_map;
+  };
+
+  bool serialize (hb_serialize_context_t *c,
+		  serialize_closure_t *closure) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->start_embed (*this);
+    if (unlikely (!c->extend_min (out))) return_trace (false);
+
+    out->secondGlyph = (*closure->glyph_map)[secondGlyph];
+
+    closure->valueFormats[0].serialize_copy (c, closure->base, &values[0], closure->layout_variation_idx_map);
+    closure->valueFormats[1].serialize_copy (c, closure->base, &values[closure->len1], closure->layout_variation_idx_map);
+
+    return_trace (true);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  const ValueFormat *valueFormats,
+				  const void *base) const
+  {
+    unsigned record1_len = valueFormats[0].get_len ();
+    unsigned record2_len = valueFormats[1].get_len ();
+    const hb_array_t<const Value> values_array = values.as_array (record1_len + record2_len);
+
+    if (valueFormats[0].has_device ())
+      valueFormats[0].collect_variation_indices (c, base, values_array.sub_array (0, record1_len));
+
+    if (valueFormats[1].has_device ())
+      valueFormats[1].collect_variation_indices (c, base, values_array.sub_array (record1_len, record2_len));
+  }
+
+  protected:
+  HBGlyphID	secondGlyph;		/* GlyphID of second glyph in the
+					 * pair--first glyph is listed in the
+					 * Coverage table */
+  ValueRecord	values;			/* Positioning data for the first glyph
+					 * followed by for second glyph */
+  public:
+  DEFINE_SIZE_ARRAY (2, values);
+};
+
+struct PairSet
+{
+  friend struct PairPosFormat1;
+
+  bool intersects (const hb_set_t *glyphs,
+		   const ValueFormat *valueFormats) const
+  {
+    unsigned int len1 = valueFormats[0].get_len ();
+    unsigned int len2 = valueFormats[1].get_len ();
+    unsigned int record_size = HBUINT16::static_size * (1 + len1 + len2);
+
+    const PairValueRecord *record = &firstPairValueRecord;
+    unsigned int count = len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (glyphs->has (record->secondGlyph))
+	return true;
+      record = &StructAtOffset<const PairValueRecord> (record, record_size);
+    }
+    return false;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c,
+		       const ValueFormat *valueFormats) const
+  {
+    unsigned int len1 = valueFormats[0].get_len ();
+    unsigned int len2 = valueFormats[1].get_len ();
+    unsigned int record_size = HBUINT16::static_size * (1 + len1 + len2);
+
+    const PairValueRecord *record = &firstPairValueRecord;
+    c->input->add_array (&record->secondGlyph, len, record_size);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  const ValueFormat *valueFormats) const
+  {
+    unsigned len1 = valueFormats[0].get_len ();
+    unsigned len2 = valueFormats[1].get_len ();
+    unsigned record_size = HBUINT16::static_size * (1 + len1 + len2);
+
+    const PairValueRecord *record = &firstPairValueRecord;
+    unsigned count = len;
+    for (unsigned i = 0; i < count; i++)
+    {
+      if (c->glyph_set->has (record->secondGlyph))
+      { record->collect_variation_indices (c, valueFormats, this); }
+
+      record = &StructAtOffset<const PairValueRecord> (record, record_size);
+    }
+  }
+
+  bool apply (hb_ot_apply_context_t *c,
+	      const ValueFormat *valueFormats,
+	      unsigned int pos) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int len1 = valueFormats[0].get_len ();
+    unsigned int len2 = valueFormats[1].get_len ();
+    unsigned int record_size = HBUINT16::static_size * (1 + len1 + len2);
+
+    const PairValueRecord *record = hb_bsearch (buffer->info[pos].codepoint,
+						&firstPairValueRecord,
+						len,
+						record_size);
+    if (record)
+    {
+      /* Note the intentional use of "|" instead of short-circuit "||". */
+      if (valueFormats[0].apply_value (c, this, &record->values[0], buffer->cur_pos()) |
+	  valueFormats[1].apply_value (c, this, &record->values[len1], buffer->pos[pos]))
+	buffer->unsafe_to_break (buffer->idx, pos + 1);
+      if (len2)
+	pos++;
+      buffer->idx = pos;
+      return_trace (true);
+    }
+    return_trace (false);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       const ValueFormat valueFormats[2]) const
+  {
+    TRACE_SUBSET (this);
+    auto snap = c->serializer->snapshot ();
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->len = 0;
+
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    unsigned len1 = valueFormats[0].get_len ();
+    unsigned len2 = valueFormats[1].get_len ();
+    unsigned record_size = HBUINT16::static_size + Value::static_size * (len1 + len2);
+
+    PairValueRecord::serialize_closure_t closure =
+    {
+      this,
+      valueFormats,
+      len1,
+      &glyph_map,
+      c->plan->layout_variation_idx_map
+    };
+
+    const PairValueRecord *record = &firstPairValueRecord;
+    unsigned count = len, num = 0;
+    for (unsigned i = 0; i < count; i++)
+    {
+      if (glyphset.has (record->secondGlyph)
+	 && record->serialize (c->serializer, &closure)) num++;
+      record = &StructAtOffset<const PairValueRecord> (record, record_size);
+    }
+
+    out->len = num;
+    if (!num) c->serializer->revert (snap);
+    return_trace (num);
+  }
+
+  struct sanitize_closure_t
+  {
+    const ValueFormat *valueFormats;
+    unsigned int len1; /* valueFormats[0].get_len() */
+    unsigned int stride; /* 1 + len1 + len2 */
+  };
+
+  bool sanitize (hb_sanitize_context_t *c, const sanitize_closure_t *closure) const
+  {
+    TRACE_SANITIZE (this);
+    if (!(c->check_struct (this)
+       && c->check_range (&firstPairValueRecord,
+			  len,
+			  HBUINT16::static_size,
+			  closure->stride))) return_trace (false);
+
+    unsigned int count = len;
+    const PairValueRecord *record = &firstPairValueRecord;
+    return_trace (closure->valueFormats[0].sanitize_values_stride_unsafe (c, this, &record->values[0], count, closure->stride) &&
+		  closure->valueFormats[1].sanitize_values_stride_unsafe (c, this, &record->values[closure->len1], count, closure->stride));
+  }
+
+  protected:
+  HBUINT16		len;	/* Number of PairValueRecords */
+  PairValueRecord	firstPairValueRecord;
+				/* Array of PairValueRecords--ordered
+				 * by GlyphID of the second glyph */
+  public:
+  DEFINE_SIZE_MIN (2);
+};
+
+struct PairPosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return
+    + hb_zip (this+coverage, pairSet)
+    | hb_filter (*glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map ([glyphs, this] (const OffsetTo<PairSet> &_)
+	      { return (this+_).intersects (glyphs, valueFormat); })
+    | hb_any
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    if ((!valueFormat[0].has_device ()) && (!valueFormat[1].has_device ())) return;
+
+    auto it =
+    + hb_zip (this+coverage, pairSet)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    ;
+
+    if (!it) return;
+    + it
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const PairSet& _) { _.collect_variation_indices (c, valueFormat); })
+    ;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    unsigned int count = pairSet.len;
+    for (unsigned int i = 0; i < count; i++)
+      (this+pairSet[i]).collect_glyphs (c, valueFormat);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int index = (this+coverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    if (!skippy_iter.next ()) return_trace (false);
+
+    return_trace ((this+pairSet[index]).apply (c, valueFormat, skippy_iter.idx));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+    out->valueFormat[0] = valueFormat[0];
+    out->valueFormat[1] = valueFormat[1];
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+
+    + hb_zip (this+coverage, pairSet)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter ([this, c, out] (const OffsetTo<PairSet>& _)
+		 {
+		   auto *o = out->pairSet.serialize_append (c->serializer);
+		   if (unlikely (!o)) return false;
+		   auto snap = c->serializer->snapshot ();
+		   bool ret = o->serialize_subset (c, _, this, valueFormat);
+		   if (!ret)
+		   {
+		     out->pairSet.pop ();
+		     c->serializer->revert (snap);
+		   }
+		   return ret;
+		 },
+		 hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+
+    if (!c->check_struct (this)) return_trace (false);
+
+    unsigned int len1 = valueFormat[0].get_len ();
+    unsigned int len2 = valueFormat[1].get_len ();
+    PairSet::sanitize_closure_t closure =
+    {
+      valueFormat,
+      len1,
+      1 + len1 + len2
+    };
+
+    return_trace (coverage.sanitize (c, this) && pairSet.sanitize (c, this, &closure));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of subtable */
+  ValueFormat	valueFormat[2];		/* [0] Defines the types of data in
+					 * ValueRecord1--for the first glyph
+					 * in the pair--may be zero (0) */
+					/* [1] Defines the types of data in
+					 * ValueRecord2--for the second glyph
+					 * in the pair--may be zero (0) */
+  OffsetArrayOf<PairSet>
+		pairSet;		/* Array of PairSet tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (10, pairSet);
+};
+
+struct PairPosFormat2
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return (this+coverage).intersects (glyphs) &&
+	   (this+classDef2).intersects (glyphs);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    if ((!valueFormat1.has_device ()) && (!valueFormat2.has_device ())) return;
+
+    hb_set_t class1_set, class2_set;
+    for (const unsigned cp : c->glyph_set->iter ())
+    {
+      unsigned klass1 = (this+classDef1).get (cp);
+      unsigned klass2 = (this+classDef2).get (cp);
+      class1_set.add (klass1);
+      class2_set.add (klass2);
+    }
+
+    if (class1_set.is_empty () || class2_set.is_empty ()) return;
+
+    unsigned len1 = valueFormat1.get_len ();
+    unsigned len2 = valueFormat2.get_len ();
+    const hb_array_t<const Value> values_array = values.as_array ((unsigned)class1Count * (unsigned) class2Count * (len1 + len2));
+    for (const unsigned class1_idx : class1_set.iter ())
+    {
+      for (const unsigned class2_idx : class2_set.iter ())
+      {
+	unsigned start_offset = (class1_idx * (unsigned) class2Count + class2_idx) * (len1 + len2);
+	if (valueFormat1.has_device ())
+	  valueFormat1.collect_variation_indices (c, this, values_array.sub_array (start_offset, len1));
+
+	if (valueFormat2.has_device ())
+	  valueFormat2.collect_variation_indices (c, this, values_array.sub_array (start_offset+len1, len2));
+      }
+    }
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    if (unlikely (!(this+classDef2).collect_coverage (c->input))) return;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int index = (this+coverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    if (!skippy_iter.next ()) return_trace (false);
+
+    unsigned int len1 = valueFormat1.get_len ();
+    unsigned int len2 = valueFormat2.get_len ();
+    unsigned int record_len = len1 + len2;
+
+    unsigned int klass1 = (this+classDef1).get_class (buffer->cur().codepoint);
+    unsigned int klass2 = (this+classDef2).get_class (buffer->info[skippy_iter.idx].codepoint);
+    if (unlikely (klass1 >= class1Count || klass2 >= class2Count)) return_trace (false);
+
+    const Value *v = &values[record_len * (klass1 * class2Count + klass2)];
+    /* Note the intentional use of "|" instead of short-circuit "||". */
+    if (valueFormat1.apply_value (c, this, v, buffer->cur_pos()) |
+	valueFormat2.apply_value (c, this, v + len1, buffer->pos[skippy_iter.idx]))
+      buffer->unsafe_to_break (buffer->idx, skippy_iter.idx + 1);
+
+    buffer->idx = skippy_iter.idx;
+    if (len2)
+      buffer->idx++;
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+    out->valueFormat1 = valueFormat1;
+    out->valueFormat2 = valueFormat2;
+
+    hb_map_t klass1_map;
+    out->classDef1.serialize_subset (c, classDef1, this, &klass1_map);
+    out->class1Count = klass1_map.get_population ();
+
+    hb_map_t klass2_map;
+    out->classDef2.serialize_subset (c, classDef2, this, &klass2_map);
+    out->class2Count = klass2_map.get_population ();
+
+    unsigned len1 = valueFormat1.get_len ();
+    unsigned len2 = valueFormat2.get_len ();
+
+    + hb_range ((unsigned) class1Count)
+    | hb_filter (klass1_map)
+    | hb_apply ([&] (const unsigned class1_idx)
+		{
+		  + hb_range ((unsigned) class2Count)
+		  | hb_filter (klass2_map)
+		  | hb_apply ([&] (const unsigned class2_idx)
+			      {
+				unsigned idx = (class1_idx * (unsigned) class2Count + class2_idx) * (len1 + len2);
+				valueFormat1.serialize_copy (c->serializer, this, &values[idx], c->plan->layout_variation_idx_map);
+				valueFormat2.serialize_copy (c->serializer, this, &values[idx + len1], c->plan->layout_variation_idx_map);
+			      })
+		  ;
+		})
+    ;
+
+    const hb_set_t &glyphset = *c->plan->_glyphset_gsub;
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto it =
+    + hb_iter (this+coverage)
+    | hb_filter (glyphset)
+    | hb_map_retains_sorting (glyph_map)
+    ;
+
+    out->coverage.serialize (c->serializer, out).serialize (c->serializer, it);
+    return_trace (out->class1Count && out->class2Count && bool (it));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!(c->check_struct (this)
+       && coverage.sanitize (c, this)
+       && classDef1.sanitize (c, this)
+       && classDef2.sanitize (c, this))) return_trace (false);
+
+    unsigned int len1 = valueFormat1.get_len ();
+    unsigned int len2 = valueFormat2.get_len ();
+    unsigned int stride = len1 + len2;
+    unsigned int record_size = valueFormat1.get_size () + valueFormat2.get_size ();
+    unsigned int count = (unsigned int) class1Count * (unsigned int) class2Count;
+    return_trace (c->check_range ((const void *) values,
+				  count,
+				  record_size) &&
+		  valueFormat1.sanitize_values_stride_unsafe (c, this, &values[0], count, stride) &&
+		  valueFormat2.sanitize_values_stride_unsafe (c, this, &values[len1], count, stride));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of subtable */
+  ValueFormat	valueFormat1;		/* ValueRecord definition--for the
+					 * first glyph of the pair--may be zero
+					 * (0) */
+  ValueFormat	valueFormat2;		/* ValueRecord definition--for the
+					 * second glyph of the pair--may be
+					 * zero (0) */
+  OffsetTo<ClassDef>
+		classDef1;		/* Offset to ClassDef table--from
+					 * beginning of PairPos subtable--for
+					 * the first glyph of the pair */
+  OffsetTo<ClassDef>
+		classDef2;		/* Offset to ClassDef table--from
+					 * beginning of PairPos subtable--for
+					 * the second glyph of the pair */
+  HBUINT16	class1Count;		/* Number of classes in ClassDef1
+					 * table--includes Class0 */
+  HBUINT16	class2Count;		/* Number of classes in ClassDef2
+					 * table--includes Class0 */
+  ValueRecord	values;			/* Matrix of value pairs:
+					 * class1-major, class2-minor,
+					 * Each entry has value1 and value2 */
+  public:
+  DEFINE_SIZE_ARRAY (16, values);
+};
+
+struct PairPos
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  PairPosFormat1	format1;
+  PairPosFormat2	format2;
+  } u;
+};
+
+
+struct EntryExitRecord
+{
+  friend struct CursivePosFormat1;
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (entryAnchor.sanitize (c, base) && exitAnchor.sanitize (c, base));
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c,
+				  const void *src_base) const
+  {
+    (src_base+entryAnchor).collect_variation_indices (c);
+    (src_base+exitAnchor).collect_variation_indices (c);
+  }
+
+  EntryExitRecord* copy (hb_serialize_context_t *c,
+			 const void *src_base,
+			 const void *dst_base,
+			 const hb_map_t *layout_variation_idx_map) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+
+    out->entryAnchor.serialize_copy (c, entryAnchor, src_base, c->to_bias (dst_base), hb_serialize_context_t::Head, layout_variation_idx_map);
+    out->exitAnchor.serialize_copy (c, exitAnchor, src_base, c->to_bias (dst_base), hb_serialize_context_t::Head, layout_variation_idx_map);
+    return_trace (out);
+  }
+
+  protected:
+  OffsetTo<Anchor>
+		entryAnchor;		/* Offset to EntryAnchor table--from
+					 * beginning of CursivePos
+					 * subtable--may be NULL */
+  OffsetTo<Anchor>
+		exitAnchor;		/* Offset to ExitAnchor table--from
+					 * beginning of CursivePos
+					 * subtable--may be NULL */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+static void
+reverse_cursive_minor_offset (hb_glyph_position_t *pos, unsigned int i, hb_direction_t direction, unsigned int new_parent);
+
+struct CursivePosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    + hb_zip (this+coverage, entryExitRecord)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    | hb_apply ([&] (const EntryExitRecord& record) { record.collect_variation_indices (c, this); })
+    ;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { if (unlikely (!(this+coverage).collect_coverage (c->input))) return; }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+
+    const EntryExitRecord &this_record = entryExitRecord[(this+coverage).get_coverage  (buffer->cur().codepoint)];
+    if (!this_record.entryAnchor) return_trace (false);
+
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    if (!skippy_iter.prev ()) return_trace (false);
+
+    const EntryExitRecord &prev_record = entryExitRecord[(this+coverage).get_coverage  (buffer->info[skippy_iter.idx].codepoint)];
+    if (!prev_record.exitAnchor) return_trace (false);
+
+    unsigned int i = skippy_iter.idx;
+    unsigned int j = buffer->idx;
+
+    buffer->unsafe_to_break (i, j);
+    float entry_x, entry_y, exit_x, exit_y;
+    (this+prev_record.exitAnchor).get_anchor (c, buffer->info[i].codepoint, &exit_x, &exit_y);
+    (this+this_record.entryAnchor).get_anchor (c, buffer->info[j].codepoint, &entry_x, &entry_y);
+
+    hb_glyph_position_t *pos = buffer->pos;
+
+    hb_position_t d;
+    /* Main-direction adjustment */
+    switch (c->direction) {
+      case HB_DIRECTION_LTR:
+	pos[i].x_advance  = roundf (exit_x) + pos[i].x_offset;
+
+	d = roundf (entry_x) + pos[j].x_offset;
+	pos[j].x_advance -= d;
+	pos[j].x_offset  -= d;
+	break;
+      case HB_DIRECTION_RTL:
+	d = roundf (exit_x) + pos[i].x_offset;
+	pos[i].x_advance -= d;
+	pos[i].x_offset  -= d;
+
+	pos[j].x_advance  = roundf (entry_x) + pos[j].x_offset;
+	break;
+      case HB_DIRECTION_TTB:
+	pos[i].y_advance  = roundf (exit_y) + pos[i].y_offset;
+
+	d = roundf (entry_y) + pos[j].y_offset;
+	pos[j].y_advance -= d;
+	pos[j].y_offset  -= d;
+	break;
+      case HB_DIRECTION_BTT:
+	d = roundf (exit_y) + pos[i].y_offset;
+	pos[i].y_advance -= d;
+	pos[i].y_offset  -= d;
+
+	pos[j].y_advance  = roundf (entry_y);
+	break;
+      case HB_DIRECTION_INVALID:
+      default:
+	break;
+    }
+
+    /* Cross-direction adjustment */
+
+    /* We attach child to parent (think graph theory and rooted trees whereas
+     * the root stays on baseline and each node aligns itself against its
+     * parent.
+     *
+     * Optimize things for the case of RightToLeft, as that's most common in
+     * Arabic. */
+    unsigned int child  = i;
+    unsigned int parent = j;
+    hb_position_t x_offset = entry_x - exit_x;
+    hb_position_t y_offset = entry_y - exit_y;
+    if  (!(c->lookup_props & LookupFlag::RightToLeft))
+    {
+      unsigned int k = child;
+      child = parent;
+      parent = k;
+      x_offset = -x_offset;
+      y_offset = -y_offset;
+    }
+
+    /* If child was already connected to someone else, walk through its old
+     * chain and reverse the link direction, such that the whole tree of its
+     * previous connection now attaches to new parent.  Watch out for case
+     * where new parent is on the path from old chain...
+     */
+    reverse_cursive_minor_offset (pos, child, c->direction, parent);
+
+    pos[child].attach_type() = ATTACH_TYPE_CURSIVE;
+    pos[child].attach_chain() = (int) parent - (int) child;
+    buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT;
+    if (likely (HB_DIRECTION_IS_HORIZONTAL (c->direction)))
+      pos[child].y_offset = y_offset;
+    else
+      pos[child].x_offset = x_offset;
+
+    /* If parent was attached to child, break them free.
+     * https://github.com/harfbuzz/harfbuzz/issues/2469
+     */
+    if (unlikely (pos[parent].attach_chain() == -pos[child].attach_chain()))
+      pos[parent].attach_chain() = 0;
+
+    buffer->idx++;
+    return_trace (true);
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  const void *src_base,
+		  const hb_map_t *layout_variation_idx_map)
+  {
+    if (unlikely (!c->extend_min ((*this)))) return;
+    this->format = 1;
+    this->entryExitRecord.len = it.len ();
+
+    for (const EntryExitRecord& entry_record : + it
+					       | hb_map (hb_second))
+      c->copy (entry_record, src_base, this, layout_variation_idx_map);
+
+    auto glyphs =
+    + it
+    | hb_map_retains_sorting (hb_first)
+    ;
+
+    coverage.serialize (c, this).serialize (c, glyphs);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!out)) return_trace (false);
+
+    auto it =
+    + hb_zip (this+coverage, entryExitRecord)
+    | hb_filter (glyphset, hb_first)
+    | hb_map_retains_sorting ([&] (hb_pair_t<hb_codepoint_t, const EntryExitRecord&> p) -> hb_pair_t<hb_codepoint_t, const EntryExitRecord&>
+			      { return hb_pair (glyph_map[p.first], p.second);})
+    ;
+
+    bool ret = bool (it);
+    out->serialize (c->serializer, it, this, c->plan->layout_variation_idx_map);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && entryExitRecord.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of subtable */
+  ArrayOf<EntryExitRecord>
+		entryExitRecord;	/* Array of EntryExit records--in
+					 * Coverage Index order */
+  public:
+  DEFINE_SIZE_ARRAY (6, entryExitRecord);
+};
+
+struct CursivePos
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  CursivePosFormat1	format1;
+  } u;
+};
+
+
+typedef AnchorMatrix BaseArray;		/* base-major--
+					 * in order of BaseCoverage Index--,
+					 * mark-minor--
+					 * ordered by class--zero-based. */
+
+static void Markclass_closure_and_remap_indexes (const Coverage  &mark_coverage,
+						 const MarkArray &mark_array,
+						 const hb_set_t  &glyphset,
+						 hb_map_t*        klass_mapping /* INOUT */)
+{
+  hb_set_t orig_classes;
+
+  + hb_zip (mark_coverage, mark_array)
+  | hb_filter (glyphset, hb_first)
+  | hb_map (hb_second)
+  | hb_map (&MarkRecord::get_class)
+  | hb_sink (orig_classes)
+  ;
+
+  unsigned idx = 0;
+  for (auto klass : orig_classes.iter ())
+  {
+    if (klass_mapping->has (klass)) continue;
+    klass_mapping->set (klass, idx);
+    idx++;
+  }
+}
+
+struct MarkBasePosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return (this+markCoverage).intersects (glyphs) &&
+	   (this+baseCoverage).intersects (glyphs);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    + hb_zip (this+markCoverage, this+markArray)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    | hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+markArray)); })
+    ;
+
+    hb_map_t klass_mapping;
+    Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, *c->glyph_set, &klass_mapping);
+
+    unsigned basecount = (this+baseArray).rows;
+    auto base_iter =
+    + hb_zip (this+baseCoverage, hb_range (basecount))
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    ;
+
+    hb_sorted_vector_t<unsigned> base_indexes;
+    for (const unsigned row : base_iter)
+    {
+      + hb_range ((unsigned) classCount)
+      | hb_filter (klass_mapping)
+      | hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
+      | hb_sink (base_indexes)
+      ;
+    }
+    (this+baseArray).collect_variation_indices (c, base_indexes.iter ());
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+markCoverage).collect_coverage (c->input))) return;
+    if (unlikely (!(this+baseCoverage).collect_coverage (c->input))) return;
+  }
+
+  const Coverage &get_coverage () const { return this+markCoverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int mark_index = (this+markCoverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (mark_index == NOT_COVERED)) return_trace (false);
+
+    /* Now we search backwards for a non-mark glyph */
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    skippy_iter.set_lookup_props (LookupFlag::IgnoreMarks);
+    do {
+      if (!skippy_iter.prev ()) return_trace (false);
+      /* We only want to attach to the first of a MultipleSubst sequence.
+       * https://github.com/harfbuzz/harfbuzz/issues/740
+       * Reject others...
+       * ...but stop if we find a mark in the MultipleSubst sequence:
+       * https://github.com/harfbuzz/harfbuzz/issues/1020 */
+      if (!_hb_glyph_info_multiplied (&buffer->info[skippy_iter.idx]) ||
+	  0 == _hb_glyph_info_get_lig_comp (&buffer->info[skippy_iter.idx]) ||
+	  (skippy_iter.idx == 0 ||
+	   _hb_glyph_info_is_mark (&buffer->info[skippy_iter.idx - 1]) ||
+	   _hb_glyph_info_get_lig_id (&buffer->info[skippy_iter.idx]) !=
+	   _hb_glyph_info_get_lig_id (&buffer->info[skippy_iter.idx - 1]) ||
+	   _hb_glyph_info_get_lig_comp (&buffer->info[skippy_iter.idx]) !=
+	   _hb_glyph_info_get_lig_comp (&buffer->info[skippy_iter.idx - 1]) + 1
+	   ))
+	break;
+      skippy_iter.reject ();
+    } while (true);
+
+    /* Checking that matched glyph is actually a base glyph by GDEF is too strong; disabled */
+    //if (!_hb_glyph_info_is_base_glyph (&buffer->info[skippy_iter.idx])) { return_trace (false); }
+
+    unsigned int base_index = (this+baseCoverage).get_coverage  (buffer->info[skippy_iter.idx].codepoint);
+    if (base_index == NOT_COVERED) return_trace (false);
+
+    return_trace ((this+markArray).apply (c, mark_index, base_index, this+baseArray, classCount, skippy_iter.idx));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    hb_map_t klass_mapping;
+    Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, glyphset, &klass_mapping);
+
+    if (!klass_mapping.get_population ()) return_trace (false);
+    out->classCount = klass_mapping.get_population ();
+
+    auto mark_iter =
+    + hb_zip (this+markCoverage, this+markArray)
+    | hb_filter (glyphset, hb_first)
+    ;
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + mark_iter
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    if (!out->markCoverage.serialize (c->serializer, out)
+			  .serialize (c->serializer, new_coverage.iter ()))
+      return_trace (false);
+
+    out->markArray.serialize (c->serializer, out)
+		  .serialize (c->serializer, &klass_mapping, c->plan->layout_variation_idx_map, &(this+markArray), + mark_iter
+										                                   | hb_map (hb_second));
+
+    unsigned basecount = (this+baseArray).rows;
+    auto base_iter =
+    + hb_zip (this+baseCoverage, hb_range (basecount))
+    | hb_filter (glyphset, hb_first)
+    ;
+
+    new_coverage.reset ();
+    + base_iter
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    if (!out->baseCoverage.serialize (c->serializer, out)
+			  .serialize (c->serializer, new_coverage.iter ()))
+      return_trace (false);
+
+    hb_sorted_vector_t<unsigned> base_indexes;
+    for (const unsigned row : + base_iter
+			      | hb_map (hb_second))
+    {
+      + hb_range ((unsigned) classCount)
+      | hb_filter (klass_mapping)
+      | hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
+      | hb_sink (base_indexes)
+      ;
+    }
+    out->baseArray.serialize (c->serializer, out)
+		  .serialize (c->serializer, base_iter.len (), &(this+baseArray), c->plan->layout_variation_idx_map, base_indexes.iter ());
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  markCoverage.sanitize (c, this) &&
+		  baseCoverage.sanitize (c, this) &&
+		  markArray.sanitize (c, this) &&
+		  baseArray.sanitize (c, this, (unsigned int) classCount));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		markCoverage;		/* Offset to MarkCoverage table--from
+					 * beginning of MarkBasePos subtable */
+  OffsetTo<Coverage>
+		baseCoverage;		/* Offset to BaseCoverage table--from
+					 * beginning of MarkBasePos subtable */
+  HBUINT16	classCount;		/* Number of classes defined for marks */
+  OffsetTo<MarkArray>
+		markArray;		/* Offset to MarkArray table--from
+					 * beginning of MarkBasePos subtable */
+  OffsetTo<BaseArray>
+		baseArray;		/* Offset to BaseArray table--from
+					 * beginning of MarkBasePos subtable */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct MarkBasePos
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  MarkBasePosFormat1	format1;
+  } u;
+};
+
+
+typedef AnchorMatrix LigatureAttach;	/* component-major--
+					 * in order of writing direction--,
+					 * mark-minor--
+					 * ordered by class--zero-based. */
+
+typedef OffsetListOf<LigatureAttach> LigatureArray;
+					/* Array of LigatureAttach
+					 * tables ordered by
+					 * LigatureCoverage Index */
+
+struct MarkLigPosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return (this+markCoverage).intersects (glyphs) &&
+	   (this+ligatureCoverage).intersects (glyphs);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    + hb_zip (this+markCoverage, this+markArray)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    | hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+markArray)); })
+    ;
+
+    hb_map_t klass_mapping;
+    Markclass_closure_and_remap_indexes (this+markCoverage, this+markArray, *c->glyph_set, &klass_mapping);
+
+    unsigned ligcount = (this+ligatureArray).len;
+    auto lig_iter =
+    + hb_zip (this+ligatureCoverage, hb_range (ligcount))
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    ;
+
+    const LigatureArray& lig_array = this+ligatureArray;
+    for (const unsigned i : lig_iter)
+    {
+      hb_sorted_vector_t<unsigned> lig_indexes;
+      unsigned row_count = lig_array[i].rows;
+      for (unsigned row : + hb_range (row_count))
+      {
+	+ hb_range ((unsigned) classCount)
+	| hb_filter (klass_mapping)
+	| hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
+	| hb_sink (lig_indexes)
+	;
+      }
+
+      lig_array[i].collect_variation_indices (c, lig_indexes.iter ());
+    }
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+markCoverage).collect_coverage (c->input))) return;
+    if (unlikely (!(this+ligatureCoverage).collect_coverage (c->input))) return;
+  }
+
+  const Coverage &get_coverage () const { return this+markCoverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int mark_index = (this+markCoverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (mark_index == NOT_COVERED)) return_trace (false);
+
+    /* Now we search backwards for a non-mark glyph */
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    skippy_iter.set_lookup_props (LookupFlag::IgnoreMarks);
+    if (!skippy_iter.prev ()) return_trace (false);
+
+    /* Checking that matched glyph is actually a ligature by GDEF is too strong; disabled */
+    //if (!_hb_glyph_info_is_ligature (&buffer->info[skippy_iter.idx])) { return_trace (false); }
+
+    unsigned int j = skippy_iter.idx;
+    unsigned int lig_index = (this+ligatureCoverage).get_coverage  (buffer->info[j].codepoint);
+    if (lig_index == NOT_COVERED) return_trace (false);
+
+    const LigatureArray& lig_array = this+ligatureArray;
+    const LigatureAttach& lig_attach = lig_array[lig_index];
+
+    /* Find component to attach to */
+    unsigned int comp_count = lig_attach.rows;
+    if (unlikely (!comp_count)) return_trace (false);
+
+    /* We must now check whether the ligature ID of the current mark glyph
+     * is identical to the ligature ID of the found ligature.  If yes, we
+     * can directly use the component index.  If not, we attach the mark
+     * glyph to the last component of the ligature. */
+    unsigned int comp_index;
+    unsigned int lig_id = _hb_glyph_info_get_lig_id (&buffer->info[j]);
+    unsigned int mark_id = _hb_glyph_info_get_lig_id (&buffer->cur());
+    unsigned int mark_comp = _hb_glyph_info_get_lig_comp (&buffer->cur());
+    if (lig_id && lig_id == mark_id && mark_comp > 0)
+      comp_index = hb_min (comp_count, _hb_glyph_info_get_lig_comp (&buffer->cur())) - 1;
+    else
+      comp_index = comp_count - 1;
+
+    return_trace ((this+markArray).apply (c, mark_index, comp_index, lig_attach, classCount, j));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    // TODO(subset)
+    return_trace (false);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  markCoverage.sanitize (c, this) &&
+		  ligatureCoverage.sanitize (c, this) &&
+		  markArray.sanitize (c, this) &&
+		  ligatureArray.sanitize (c, this, (unsigned int) classCount));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		markCoverage;		/* Offset to Mark Coverage table--from
+					 * beginning of MarkLigPos subtable */
+  OffsetTo<Coverage>
+		ligatureCoverage;	/* Offset to Ligature Coverage
+					 * table--from beginning of MarkLigPos
+					 * subtable */
+  HBUINT16	classCount;		/* Number of defined mark classes */
+  OffsetTo<MarkArray>
+		markArray;		/* Offset to MarkArray table--from
+					 * beginning of MarkLigPos subtable */
+  OffsetTo<LigatureArray>
+		ligatureArray;		/* Offset to LigatureArray table--from
+					 * beginning of MarkLigPos subtable */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct MarkLigPos
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  MarkLigPosFormat1	format1;
+  } u;
+};
+
+
+typedef AnchorMatrix Mark2Array;	/* mark2-major--
+					 * in order of Mark2Coverage Index--,
+					 * mark1-minor--
+					 * ordered by class--zero-based. */
+
+struct MarkMarkPosFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return (this+mark1Coverage).intersects (glyphs) &&
+	   (this+mark2Coverage).intersects (glyphs);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    + hb_zip (this+mark1Coverage, this+mark1Array)
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    | hb_apply ([&] (const MarkRecord& record) { record.collect_variation_indices (c, &(this+mark1Array)); })
+    ;
+
+    hb_map_t klass_mapping;
+    Markclass_closure_and_remap_indexes (this+mark1Coverage, this+mark1Array, *c->glyph_set, &klass_mapping);
+
+    unsigned mark2_count = (this+mark2Array).rows;
+    auto mark2_iter =
+    + hb_zip (this+mark2Coverage, hb_range (mark2_count))
+    | hb_filter (c->glyph_set, hb_first)
+    | hb_map (hb_second)
+    ;
+
+    hb_sorted_vector_t<unsigned> mark2_indexes;
+    for (const unsigned row : mark2_iter)
+    {
+      + hb_range ((unsigned) classCount)
+      | hb_filter (klass_mapping)
+      | hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
+      | hb_sink (mark2_indexes)
+      ;
+    }
+    (this+mark2Array).collect_variation_indices (c, mark2_indexes.iter ());
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+mark1Coverage).collect_coverage (c->input))) return;
+    if (unlikely (!(this+mark2Coverage).collect_coverage (c->input))) return;
+  }
+
+  const Coverage &get_coverage () const { return this+mark1Coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_buffer_t *buffer = c->buffer;
+    unsigned int mark1_index = (this+mark1Coverage).get_coverage  (buffer->cur().codepoint);
+    if (likely (mark1_index == NOT_COVERED)) return_trace (false);
+
+    /* now we search backwards for a suitable mark glyph until a non-mark glyph */
+    hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+    skippy_iter.reset (buffer->idx, 1);
+    skippy_iter.set_lookup_props (c->lookup_props & ~LookupFlag::IgnoreFlags);
+    if (!skippy_iter.prev ()) return_trace (false);
+
+    if (!_hb_glyph_info_is_mark (&buffer->info[skippy_iter.idx])) { return_trace (false); }
+
+    unsigned int j = skippy_iter.idx;
+
+    unsigned int id1 = _hb_glyph_info_get_lig_id (&buffer->cur());
+    unsigned int id2 = _hb_glyph_info_get_lig_id (&buffer->info[j]);
+    unsigned int comp1 = _hb_glyph_info_get_lig_comp (&buffer->cur());
+    unsigned int comp2 = _hb_glyph_info_get_lig_comp (&buffer->info[j]);
+
+    if (likely (id1 == id2))
+    {
+      if (id1 == 0) /* Marks belonging to the same base. */
+	goto good;
+      else if (comp1 == comp2) /* Marks belonging to the same ligature component. */
+	goto good;
+    }
+    else
+    {
+      /* If ligature ids don't match, it may be the case that one of the marks
+       * itself is a ligature.  In which case match. */
+      if ((id1 > 0 && !comp1) || (id2 > 0 && !comp2))
+	goto good;
+    }
+
+    /* Didn't match. */
+    return_trace (false);
+
+    good:
+    unsigned int mark2_index = (this+mark2Coverage).get_coverage  (buffer->info[j].codepoint);
+    if (mark2_index == NOT_COVERED) return_trace (false);
+
+    return_trace ((this+mark1Array).apply (c, mark1_index, mark2_index, this+mark2Array, classCount, j));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    hb_map_t klass_mapping;
+    Markclass_closure_and_remap_indexes (this+mark1Coverage, this+mark1Array, glyphset, &klass_mapping);
+
+    if (!klass_mapping.get_population ()) return_trace (false);
+    out->classCount = klass_mapping.get_population ();
+
+    auto mark1_iter =
+    + hb_zip (this+mark1Coverage, this+mark1Array)
+    | hb_filter (glyphset, hb_first)
+    ;
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + mark1_iter
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    if (!out->mark1Coverage.serialize (c->serializer, out)
+			   .serialize (c->serializer, new_coverage.iter ()))
+      return_trace (false);
+
+    out->mark1Array.serialize (c->serializer, out)
+		   .serialize (c->serializer, &klass_mapping, c->plan->layout_variation_idx_map, &(this+mark1Array), + mark1_iter
+										                                     | hb_map (hb_second));
+
+    unsigned mark2count = (this+mark2Array).rows;
+    auto mark2_iter =
+    + hb_zip (this+mark2Coverage, hb_range (mark2count))
+    | hb_filter (glyphset, hb_first)
+    ;
+
+    new_coverage.reset ();
+    + mark2_iter
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    if (!out->mark2Coverage.serialize (c->serializer, out)
+			   .serialize (c->serializer, new_coverage.iter ()))
+      return_trace (false);
+
+    hb_sorted_vector_t<unsigned> mark2_indexes;
+    for (const unsigned row : + mark2_iter
+			      | hb_map (hb_second))
+    {
+      + hb_range ((unsigned) classCount)
+      | hb_filter (klass_mapping)
+      | hb_map ([&] (const unsigned col) { return row * (unsigned) classCount + col; })
+      | hb_sink (mark2_indexes)
+      ;
+    }
+    out->mark2Array.serialize (c->serializer, out)
+		   .serialize (c->serializer, mark2_iter.len (), &(this+mark2Array), c->plan->layout_variation_idx_map, mark2_indexes.iter ());
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  mark1Coverage.sanitize (c, this) &&
+		  mark2Coverage.sanitize (c, this) &&
+		  mark1Array.sanitize (c, this) &&
+		  mark2Array.sanitize (c, this, (unsigned int) classCount));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		mark1Coverage;		/* Offset to Combining Mark1 Coverage
+					 * table--from beginning of MarkMarkPos
+					 * subtable */
+  OffsetTo<Coverage>
+		mark2Coverage;		/* Offset to Combining Mark2 Coverage
+					 * table--from beginning of MarkMarkPos
+					 * subtable */
+  HBUINT16	classCount;		/* Number of defined mark classes */
+  OffsetTo<MarkArray>
+		mark1Array;		/* Offset to Mark1Array table--from
+					 * beginning of MarkMarkPos subtable */
+  OffsetTo<Mark2Array>
+		mark2Array;		/* Offset to Mark2Array table--from
+					 * beginning of MarkMarkPos subtable */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct MarkMarkPos
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  MarkMarkPosFormat1	format1;
+  } u;
+};
+
+
+struct ContextPos : Context {};
+
+struct ChainContextPos : ChainContext {};
+
+struct ExtensionPos : Extension<ExtensionPos>
+{
+  typedef struct PosLookupSubTable SubTable;
+};
+
+
+
+/*
+ * PosLookup
+ */
+
+
+struct PosLookupSubTable
+{
+  friend struct Lookup;
+  friend struct PosLookup;
+
+  enum Type {
+    Single		= 1,
+    Pair		= 2,
+    Cursive		= 3,
+    MarkBase		= 4,
+    MarkLig		= 5,
+    MarkMark		= 6,
+    Context		= 7,
+    ChainContext	= 8,
+    Extension		= 9
+  };
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, unsigned int lookup_type, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, lookup_type);
+    switch (lookup_type) {
+    case Single:		return_trace (u.single.dispatch (c, hb_forward<Ts> (ds)...));
+    case Pair:			return_trace (u.pair.dispatch (c, hb_forward<Ts> (ds)...));
+    case Cursive:		return_trace (u.cursive.dispatch (c, hb_forward<Ts> (ds)...));
+    case MarkBase:		return_trace (u.markBase.dispatch (c, hb_forward<Ts> (ds)...));
+    case MarkLig:		return_trace (u.markLig.dispatch (c, hb_forward<Ts> (ds)...));
+    case MarkMark:		return_trace (u.markMark.dispatch (c, hb_forward<Ts> (ds)...));
+    case Context:		return_trace (u.context.dispatch (c, hb_forward<Ts> (ds)...));
+    case ChainContext:		return_trace (u.chainContext.dispatch (c, hb_forward<Ts> (ds)...));
+    case Extension:		return_trace (u.extension.dispatch (c, hb_forward<Ts> (ds)...));
+    default:			return_trace (c->default_return_value ());
+    }
+  }
+
+  bool intersects (const hb_set_t *glyphs, unsigned int lookup_type) const
+  {
+    hb_intersects_context_t c (glyphs);
+    return dispatch (&c, lookup_type);
+  }
+
+  protected:
+  union {
+  SinglePos		single;
+  PairPos		pair;
+  CursivePos		cursive;
+  MarkBasePos		markBase;
+  MarkLigPos		markLig;
+  MarkMarkPos		markMark;
+  ContextPos		context;
+  ChainContextPos	chainContext;
+  ExtensionPos		extension;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (0);
+};
+
+
+struct PosLookup : Lookup
+{
+  typedef struct PosLookupSubTable SubTable;
+
+  const SubTable& get_subtable (unsigned int i) const
+  { return Lookup::get_subtable<SubTable> (i); }
+
+  bool is_reverse () const
+  {
+    return false;
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    return_trace (dispatch (c));
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    hb_intersects_context_t c (glyphs);
+    return dispatch (&c);
+  }
+
+  hb_collect_glyphs_context_t::return_t collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { return dispatch (c); }
+
+  hb_closure_lookups_context_t::return_t closure_lookups (hb_closure_lookups_context_t *c, unsigned this_index) const
+  {
+    if (c->is_lookup_visited (this_index))
+      return hb_closure_lookups_context_t::default_return_value ();
+
+    c->set_lookup_visited (this_index);
+    if (!intersects (c->glyphs))
+    {
+      c->set_lookup_inactive (this_index);
+      return hb_closure_lookups_context_t::default_return_value ();
+    }
+    c->set_recurse_func (dispatch_closure_lookups_recurse_func);
+
+    hb_closure_lookups_context_t::return_t ret = dispatch (c);
+    return ret;
+  }
+
+  template <typename set_t>
+  void collect_coverage (set_t *glyphs) const
+  {
+    hb_collect_coverage_context_t<set_t> c (glyphs);
+    dispatch (&c);
+  }
+
+  static inline bool apply_recurse_func (hb_ot_apply_context_t *c, unsigned int lookup_index);
+
+  template <typename context_t>
+  static typename context_t::return_t dispatch_recurse_func (context_t *c, unsigned int lookup_index);
+
+  HB_INTERNAL static hb_closure_lookups_context_t::return_t dispatch_closure_lookups_recurse_func (hb_closure_lookups_context_t *c, unsigned this_index);
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  { return Lookup::dispatch<SubTable> (c, hb_forward<Ts> (ds)...); }
+
+  bool subset (hb_subset_context_t *c) const
+  { return Lookup::subset<SubTable> (c); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return Lookup::sanitize<SubTable> (c); }
+};
+
+/*
+ * GPOS -- Glyph Positioning
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/gpos
+ */
+
+struct GPOS : GSUBGPOS
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_GPOS;
+
+  const PosLookup& get_lookup (unsigned int i) const
+  { return static_cast<const PosLookup &> (GSUBGPOS::get_lookup (i)); }
+
+  static inline void position_start (hb_font_t *font, hb_buffer_t *buffer);
+  static inline void position_finish_advances (hb_font_t *font, hb_buffer_t *buffer);
+  static inline void position_finish_offsets (hb_font_t *font, hb_buffer_t *buffer);
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    hb_subset_layout_context_t l (c, tableTag, c->plan->gpos_lookups, c->plan->gpos_features);
+    return GSUBGPOS::subset<PosLookup> (&l);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return GSUBGPOS::sanitize<PosLookup> (c); }
+
+  HB_INTERNAL bool is_blocklisted (hb_blob_t *blob,
+				   hb_face_t *face) const;
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  {
+    for (unsigned i = 0; i < GSUBGPOS::get_lookup_count (); i++)
+    {
+      if (!c->gpos_lookups->has (i)) continue;
+      const PosLookup &l = get_lookup (i);
+      l.dispatch (c);
+    }
+  }
+
+  void closure_lookups (hb_face_t      *face,
+			const hb_set_t *glyphs,
+			hb_set_t       *lookup_indexes /* IN/OUT */) const
+  { GSUBGPOS::closure_lookups<PosLookup> (face, glyphs, lookup_indexes); }
+
+  typedef GSUBGPOS::accelerator_t<GPOS> accelerator_t;
+};
+
+
+static void
+reverse_cursive_minor_offset (hb_glyph_position_t *pos, unsigned int i, hb_direction_t direction, unsigned int new_parent)
+{
+  int chain = pos[i].attach_chain(), type = pos[i].attach_type();
+  if (likely (!chain || 0 == (type & ATTACH_TYPE_CURSIVE)))
+    return;
+
+  pos[i].attach_chain() = 0;
+
+  unsigned int j = (int) i + chain;
+
+  /* Stop if we see new parent in the chain. */
+  if (j == new_parent)
+    return;
+
+  reverse_cursive_minor_offset (pos, j, direction, new_parent);
+
+  if (HB_DIRECTION_IS_HORIZONTAL (direction))
+    pos[j].y_offset = -pos[i].y_offset;
+  else
+    pos[j].x_offset = -pos[i].x_offset;
+
+  pos[j].attach_chain() = -chain;
+  pos[j].attach_type() = type;
+}
+static void
+propagate_attachment_offsets (hb_glyph_position_t *pos,
+			      unsigned int len,
+			      unsigned int i,
+			      hb_direction_t direction)
+{
+  /* Adjusts offsets of attached glyphs (both cursive and mark) to accumulate
+   * offset of glyph they are attached to. */
+  int chain = pos[i].attach_chain(), type = pos[i].attach_type();
+  if (likely (!chain))
+    return;
+
+  pos[i].attach_chain() = 0;
+
+  unsigned int j = (int) i + chain;
+
+  if (unlikely (j >= len))
+    return;
+
+  propagate_attachment_offsets (pos, len, j, direction);
+
+  assert (!!(type & ATTACH_TYPE_MARK) ^ !!(type & ATTACH_TYPE_CURSIVE));
+
+  if (type & ATTACH_TYPE_CURSIVE)
+  {
+    if (HB_DIRECTION_IS_HORIZONTAL (direction))
+      pos[i].y_offset += pos[j].y_offset;
+    else
+      pos[i].x_offset += pos[j].x_offset;
+  }
+  else /*if (type & ATTACH_TYPE_MARK)*/
+  {
+    pos[i].x_offset += pos[j].x_offset;
+    pos[i].y_offset += pos[j].y_offset;
+
+    assert (j < i);
+    if (HB_DIRECTION_IS_FORWARD (direction))
+      for (unsigned int k = j; k < i; k++) {
+	pos[i].x_offset -= pos[k].x_advance;
+	pos[i].y_offset -= pos[k].y_advance;
+      }
+    else
+      for (unsigned int k = j + 1; k < i + 1; k++) {
+	pos[i].x_offset += pos[k].x_advance;
+	pos[i].y_offset += pos[k].y_advance;
+      }
+  }
+}
+
+void
+GPOS::position_start (hb_font_t *font HB_UNUSED, hb_buffer_t *buffer)
+{
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+    buffer->pos[i].attach_chain() = buffer->pos[i].attach_type() = 0;
+}
+
+void
+GPOS::position_finish_advances (hb_font_t *font HB_UNUSED, hb_buffer_t *buffer HB_UNUSED)
+{
+  //_hb_buffer_assert_gsubgpos_vars (buffer);
+}
+
+void
+GPOS::position_finish_offsets (hb_font_t *font HB_UNUSED, hb_buffer_t *buffer)
+{
+  _hb_buffer_assert_gsubgpos_vars (buffer);
+
+  unsigned int len;
+  hb_glyph_position_t *pos = hb_buffer_get_glyph_positions (buffer, &len);
+  hb_direction_t direction = buffer->props.direction;
+
+  /* Handle attachments */
+  if (buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_GPOS_ATTACHMENT)
+    for (unsigned int i = 0; i < len; i++)
+      propagate_attachment_offsets (pos, len, i, direction);
+}
+
+
+struct GPOS_accelerator_t : GPOS::accelerator_t {};
+
+
+/* Out-of-class implementation for methods recursing */
+
+#ifndef HB_NO_OT_LAYOUT
+template <typename context_t>
+/*static*/ typename context_t::return_t PosLookup::dispatch_recurse_func (context_t *c, unsigned int lookup_index)
+{
+  const PosLookup &l = c->face->table.GPOS.get_relaxed ()->table->get_lookup (lookup_index);
+  return l.dispatch (c);
+}
+
+/*static*/ inline hb_closure_lookups_context_t::return_t PosLookup::dispatch_closure_lookups_recurse_func (hb_closure_lookups_context_t *c, unsigned this_index)
+{
+  const PosLookup &l = c->face->table.GPOS.get_relaxed ()->table->get_lookup (this_index);
+  return l.closure_lookups (c, this_index);
+}
+
+/*static*/ bool PosLookup::apply_recurse_func (hb_ot_apply_context_t *c, unsigned int lookup_index)
+{
+  const PosLookup &l = c->face->table.GPOS.get_relaxed ()->table->get_lookup (lookup_index);
+  unsigned int saved_lookup_props = c->lookup_props;
+  unsigned int saved_lookup_index = c->lookup_index;
+  c->set_lookup_index (lookup_index);
+  c->set_lookup_props (l.get_props ());
+  bool ret = l.dispatch (c);
+  c->set_lookup_index (saved_lookup_index);
+  c->set_lookup_props (saved_lookup_props);
+  return ret;
+}
+#endif
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_GPOS_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-gsub-table.hh b/thirdparty/harfbuzz/src/hb-ot-layout-gsub-table.hh
new file mode 100644
index 0000000000..2f41d67819
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-gsub-table.hh
@@ -0,0 +1,1627 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_GSUB_TABLE_HH
+#define HB_OT_LAYOUT_GSUB_TABLE_HH
+
+#include "hb-ot-layout-gsubgpos.hh"
+
+
+namespace OT {
+
+typedef hb_pair_t<hb_codepoint_t, hb_codepoint_t> hb_codepoint_pair_t;
+
+template<typename Iterator>
+static void SingleSubst_serialize (hb_serialize_context_t *c,
+				   Iterator it);
+
+
+struct SingleSubstFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    unsigned d = deltaGlyphID;
+    + hb_iter (this+coverage)
+    | hb_filter (*c->glyphs)
+    | hb_map ([d] (hb_codepoint_t g) { return (g + d) & 0xFFFFu; })
+    | hb_sink (c->output)
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    unsigned d = deltaGlyphID;
+    + hb_iter (this+coverage)
+    | hb_map ([d] (hb_codepoint_t g) { return (g + d) & 0xFFFFu; })
+    | hb_sink (c->output)
+    ;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  { return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    hb_codepoint_t glyph_id = c->buffer->cur().codepoint;
+    unsigned int index = (this+coverage).get_coverage (glyph_id);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    /* According to the Adobe Annotated OpenType Suite, result is always
+     * limited to 16bit. */
+    glyph_id = (glyph_id + deltaGlyphID) & 0xFFFFu;
+    c->replace_glyph (glyph_id);
+
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_sorted_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator glyphs,
+		  unsigned delta)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!coverage.serialize (c, this).serialize (c, glyphs))) return_trace (false);
+    c->check_assign (deltaGlyphID, delta);
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    hb_codepoint_t delta = deltaGlyphID;
+
+    auto it =
+    + hb_iter (this+coverage)
+    | hb_filter (glyphset)
+    | hb_map_retains_sorting ([&] (hb_codepoint_t g) {
+				return hb_codepoint_pair_t (g,
+							    (g + delta) & 0xFFFF); })
+    | hb_filter (glyphset, hb_second)
+    | hb_map_retains_sorting ([&] (hb_codepoint_pair_t p) -> hb_codepoint_pair_t
+			      { return hb_pair (glyph_map[p.first], glyph_map[p.second]); })
+    ;
+
+    bool ret = bool (it);
+    SingleSubst_serialize (c->serializer, it);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && deltaGlyphID.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of Substitution table */
+  HBUINT16	deltaGlyphID;		/* Add to original GlyphID to get
+					 * substitute GlyphID, modulo 0x10000 */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct SingleSubstFormat2
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    + hb_zip (this+coverage, substitute)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_sink (c->output)
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    + hb_zip (this+coverage, substitute)
+    | hb_map (hb_second)
+    | hb_sink (c->output)
+    ;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  { return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    if (unlikely (index >= substitute.len)) return_trace (false);
+
+    c->replace_glyph (substitute[index]);
+
+    return_trace (true);
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_sorted_source_of (Iterator,
+						hb_codepoint_pair_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it)
+  {
+    TRACE_SERIALIZE (this);
+    auto substitutes =
+      + it
+      | hb_map (hb_second)
+      ;
+    auto glyphs =
+      + it
+      | hb_map_retains_sorting (hb_first)
+      ;
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!substitute.serialize (c, substitutes))) return_trace (false);
+    if (unlikely (!coverage.serialize (c, this).serialize (c, glyphs))) return_trace (false);
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset_gsub ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto it =
+    + hb_zip (this+coverage, substitute)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (glyphset, hb_second)
+    | hb_map_retains_sorting ([&] (hb_pair_t<hb_codepoint_t, const HBGlyphID &> p) -> hb_codepoint_pair_t
+			      { return hb_pair (glyph_map[p.first], glyph_map[p.second]); })
+    ;
+
+    bool ret = bool (it);
+    SingleSubst_serialize (c->serializer, it);
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && substitute.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of Substitution table */
+  ArrayOf<HBGlyphID>
+		substitute;		/* Array of substitute
+					 * GlyphIDs--ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, substitute);
+};
+
+struct SingleSubst
+{
+
+  template<typename Iterator,
+	   hb_requires (hb_is_sorted_source_of (Iterator,
+						const hb_codepoint_pair_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator glyphs)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (u.format))) return_trace (false);
+    unsigned format = 2;
+    unsigned delta = 0;
+    if (glyphs)
+    {
+      format = 1;
+      auto get_delta = [=] (hb_codepoint_pair_t _)
+		       { return (unsigned) (_.second - _.first) & 0xFFFF; };
+      delta = get_delta (*glyphs);
+      if (!hb_all (++(+glyphs), delta, get_delta)) format = 2;
+    }
+    u.format = format;
+    switch (u.format) {
+    case 1: return_trace (u.format1.serialize (c,
+					       + glyphs
+					       | hb_map_retains_sorting (hb_first),
+					       delta));
+    case 2: return_trace (u.format2.serialize (c, glyphs));
+    default:return_trace (false);
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  SingleSubstFormat1	format1;
+  SingleSubstFormat2	format2;
+  } u;
+};
+
+template<typename Iterator>
+static void
+SingleSubst_serialize (hb_serialize_context_t *c,
+		       Iterator it)
+{ c->start_embed<SingleSubst> ()->serialize (c, it); }
+
+struct Sequence
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return hb_all (substitute, glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  { c->output->add_array (substitute.arrayZ, substitute.len); }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { c->output->add_array (substitute.arrayZ, substitute.len); }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int count = substitute.len;
+
+    /* Special-case to make it in-place and not consider this
+     * as a "multiplied" substitution. */
+    if (unlikely (count == 1))
+    {
+      c->replace_glyph (substitute.arrayZ[0]);
+      return_trace (true);
+    }
+    /* Spec disallows this, but Uniscribe allows it.
+     * https://github.com/harfbuzz/harfbuzz/issues/253 */
+    else if (unlikely (count == 0))
+    {
+      c->buffer->delete_glyph ();
+      return_trace (true);
+    }
+
+    unsigned int klass = _hb_glyph_info_is_ligature (&c->buffer->cur()) ?
+			 HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH : 0;
+
+    for (unsigned int i = 0; i < count; i++) {
+      _hb_glyph_info_set_lig_props_for_component (&c->buffer->cur(), i);
+      c->output_glyph_for_component (substitute.arrayZ[i], klass);
+    }
+    c->buffer->skip_glyph ();
+
+    return_trace (true);
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator subst)
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (substitute.serialize (c, subst));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    if (!intersects (&glyphset)) return_trace (false);
+
+    auto it =
+    + hb_iter (substitute)
+    | hb_map (glyph_map)
+    ;
+
+    auto *out = c->serializer->start_embed (*this);
+    return_trace (out->serialize (c->serializer, it));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (substitute.sanitize (c));
+  }
+
+  protected:
+  ArrayOf<HBGlyphID>
+		substitute;		/* String of GlyphIDs to substitute */
+  public:
+  DEFINE_SIZE_ARRAY (2, substitute);
+};
+
+struct MultipleSubstFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    + hb_zip (this+coverage, sequence)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const Sequence &_) { _.closure (c); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    + hb_zip (this+coverage, sequence)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const Sequence &_) { _.collect_glyphs (c); })
+    ;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  { return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    return_trace ((this+sequence[index]).apply (c));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> glyphs,
+		  hb_array_t<const unsigned int> substitute_len_list,
+		  hb_array_t<const HBGlyphID> substitute_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!sequence.serialize (c, glyphs.length))) return_trace (false);
+    for (unsigned int i = 0; i < glyphs.length; i++)
+    {
+      unsigned int substitute_len = substitute_len_list[i];
+      if (unlikely (!sequence[i].serialize (c, this)
+				.serialize (c, substitute_glyphs_list.sub_array (0, substitute_len))))
+	return_trace (false);
+      substitute_glyphs_list += substitute_len;
+    }
+    return_trace (coverage.serialize (c, this).serialize (c, glyphs));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, sequence)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->sequence, this), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && sequence.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of Substitution table */
+  OffsetArrayOf<Sequence>
+		sequence;		/* Array of Sequence tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, sequence);
+};
+
+struct MultipleSubst
+{
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> glyphs,
+		  hb_array_t<const unsigned int> substitute_len_list,
+		  hb_array_t<const HBGlyphID> substitute_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (u.format))) return_trace (false);
+    unsigned int format = 1;
+    u.format = format;
+    switch (u.format) {
+    case 1: return_trace (u.format1.serialize (c, glyphs, substitute_len_list, substitute_glyphs_list));
+    default:return_trace (false);
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  MultipleSubstFormat1	format1;
+  } u;
+};
+
+struct AlternateSet
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return hb_any (alternates, glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  { c->output->add_array (alternates.arrayZ, alternates.len); }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  { c->output->add_array (alternates.arrayZ, alternates.len); }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int count = alternates.len;
+
+    if (unlikely (!count)) return_trace (false);
+
+    hb_mask_t glyph_mask = c->buffer->cur().mask;
+    hb_mask_t lookup_mask = c->lookup_mask;
+
+    /* Note: This breaks badly if two features enabled this lookup together. */
+    unsigned int shift = hb_ctz (lookup_mask);
+    unsigned int alt_index = ((lookup_mask & glyph_mask) >> shift);
+
+    /* If alt_index is MAX_VALUE, randomize feature if it is the rand feature. */
+    if (alt_index == HB_OT_MAP_MAX_VALUE && c->random)
+      alt_index = c->random_number () % count + 1;
+
+    if (unlikely (alt_index > count || alt_index == 0)) return_trace (false);
+
+    c->replace_glyph (alternates[alt_index - 1]);
+
+    return_trace (true);
+  }
+
+  unsigned
+  get_alternates (unsigned        start_offset,
+		  unsigned       *alternate_count  /* IN/OUT.  May be NULL. */,
+		  hb_codepoint_t *alternate_glyphs /* OUT.     May be NULL. */) const
+  {
+    if (alternates.len && alternate_count)
+    {
+      + alternates.sub_array (start_offset, alternate_count)
+      | hb_sink (hb_array (alternate_glyphs, *alternate_count))
+      ;
+    }
+    return alternates.len;
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator alts)
+  {
+    TRACE_SERIALIZE (this);
+    return_trace (alternates.serialize (c, alts));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto it =
+      + hb_iter (alternates)
+      | hb_filter (glyphset)
+      | hb_map (glyph_map)
+      ;
+
+    auto *out = c->serializer->start_embed (*this);
+    return_trace (out->serialize (c->serializer, it) &&
+		  out->alternates);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (alternates.sanitize (c));
+  }
+
+  protected:
+  ArrayOf<HBGlyphID>
+		alternates;		/* Array of alternate GlyphIDs--in
+					 * arbitrary order */
+  public:
+  DEFINE_SIZE_ARRAY (2, alternates);
+};
+
+struct AlternateSubstFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return (this+coverage).intersects (glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    + hb_zip (this+coverage, alternateSet)
+    | hb_filter (c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const AlternateSet &_) { _.closure (c); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+    + hb_zip (this+coverage, alternateSet)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const AlternateSet &_) { _.collect_glyphs (c); })
+    ;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  { return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
+
+  unsigned
+  get_glyph_alternates (hb_codepoint_t  gid,
+			unsigned        start_offset,
+			unsigned       *alternate_count  /* IN/OUT.  May be NULL. */,
+			hb_codepoint_t *alternate_glyphs /* OUT.     May be NULL. */) const
+  { return (this+alternateSet[(this+coverage).get_coverage (gid)])
+	   .get_alternates (start_offset, alternate_count, alternate_glyphs); }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    return_trace ((this+alternateSet[index]).apply (c));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> glyphs,
+		  hb_array_t<const unsigned int> alternate_len_list,
+		  hb_array_t<const HBGlyphID> alternate_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!alternateSet.serialize (c, glyphs.length))) return_trace (false);
+    for (unsigned int i = 0; i < glyphs.length; i++)
+    {
+      unsigned int alternate_len = alternate_len_list[i];
+      if (unlikely (!alternateSet[i].serialize (c, this)
+				    .serialize (c, alternate_glyphs_list.sub_array (0, alternate_len))))
+	return_trace (false);
+      alternate_glyphs_list += alternate_len;
+    }
+    return_trace (coverage.serialize (c, this).serialize (c, glyphs));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, alternateSet)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->alternateSet, this), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && alternateSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of Substitution table */
+  OffsetArrayOf<AlternateSet>
+		alternateSet;		/* Array of AlternateSet tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, alternateSet);
+};
+
+struct AlternateSubst
+{
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> glyphs,
+		  hb_array_t<const unsigned int> alternate_len_list,
+		  hb_array_t<const HBGlyphID> alternate_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (u.format))) return_trace (false);
+    unsigned int format = 1;
+    u.format = format;
+    switch (u.format) {
+    case 1: return_trace (u.format1.serialize (c, glyphs, alternate_len_list, alternate_glyphs_list));
+    default:return_trace (false);
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  AlternateSubstFormat1	format1;
+  } u;
+};
+
+
+struct Ligature
+{
+  bool intersects (const hb_set_t *glyphs) const
+  { return hb_all (component, glyphs); }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    if (!intersects (c->glyphs)) return;
+    c->output->add (ligGlyph);
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    c->input->add_array (component.arrayZ, component.get_length ());
+    c->output->add (ligGlyph);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    if (c->len != component.lenP1)
+      return false;
+
+    for (unsigned int i = 1; i < c->len; i++)
+      if (likely (c->glyphs[i] != component[i]))
+	return false;
+
+    return true;
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int count = component.lenP1;
+
+    if (unlikely (!count)) return_trace (false);
+
+    /* Special-case to make it in-place and not consider this
+     * as a "ligated" substitution. */
+    if (unlikely (count == 1))
+    {
+      c->replace_glyph (ligGlyph);
+      return_trace (true);
+    }
+
+    unsigned int total_component_count = 0;
+
+    unsigned int match_length = 0;
+    unsigned int match_positions[HB_MAX_CONTEXT_LENGTH];
+
+    if (likely (!match_input (c, count,
+			      &component[1],
+			      match_glyph,
+			      nullptr,
+			      &match_length,
+			      match_positions,
+			      &total_component_count)))
+      return_trace (false);
+
+    ligate_input (c,
+		  count,
+		  match_positions,
+		  match_length,
+		  ligGlyph,
+		  total_component_count);
+
+    return_trace (true);
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, hb_codepoint_t))>
+  bool serialize (hb_serialize_context_t *c,
+		  hb_codepoint_t ligature,
+		  Iterator components /* Starting from second */)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    ligGlyph = ligature;
+    if (unlikely (!component.serialize (c, components))) return_trace (false);
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    if (!intersects (&glyphset) || !glyphset.has (ligGlyph)) return_trace (false);
+
+    auto it =
+      + hb_iter (component)
+      | hb_map (glyph_map)
+      ;
+
+    auto *out = c->serializer->start_embed (*this);
+    return_trace (out->serialize (c->serializer,
+				  glyph_map[ligGlyph],
+				  it));
+  }
+
+  public:
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (ligGlyph.sanitize (c) && component.sanitize (c));
+  }
+
+  protected:
+  HBGlyphID	ligGlyph;		/* GlyphID of ligature to substitute */
+  HeadlessArrayOf<HBGlyphID>
+		component;		/* Array of component GlyphIDs--start
+					 * with the second  component--ordered
+					 * in writing direction */
+  public:
+  DEFINE_SIZE_ARRAY (4, component);
+};
+
+struct LigatureSet
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return
+    + hb_iter (ligature)
+    | hb_map (hb_add (this))
+    | hb_map ([glyphs] (const Ligature &_) { return _.intersects (glyphs); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    + hb_iter (ligature)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const Ligature &_) { _.closure (c); })
+    ;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    + hb_iter (ligature)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const Ligature &_) { _.collect_glyphs (c); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    return
+    + hb_iter (ligature)
+    | hb_map (hb_add (this))
+    | hb_map ([c] (const Ligature &_) { return _.would_apply (c); })
+    | hb_any
+    ;
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int num_ligs = ligature.len;
+    for (unsigned int i = 0; i < num_ligs; i++)
+    {
+      const Ligature &lig = this+ligature[i];
+      if (lig.apply (c)) return_trace (true);
+    }
+
+    return_trace (false);
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  hb_array_t<const HBGlyphID> ligatures,
+		  hb_array_t<const unsigned int> component_count_list,
+		  hb_array_t<const HBGlyphID> &component_list /* Starting from second for each ligature */)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!ligature.serialize (c, ligatures.length))) return_trace (false);
+    for (unsigned int i = 0; i < ligatures.length; i++)
+    {
+      unsigned int component_count = (unsigned) hb_max ((int) component_count_list[i] - 1, 0);
+      if (unlikely (!ligature[i].serialize (c, this)
+				.serialize (c,
+					    ligatures[i],
+					    component_list.sub_array (0, component_count))))
+	return_trace (false);
+      component_list += component_count;
+    }
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    + hb_iter (ligature)
+    | hb_filter (subset_offset_array (c, out->ligature, this))
+    | hb_drain
+    ;
+    return_trace (bool (out->ligature));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (ligature.sanitize (c, this));
+  }
+
+  protected:
+  OffsetArrayOf<Ligature>
+		ligature;		/* Array LigatureSet tables
+					 * ordered by preference */
+  public:
+  DEFINE_SIZE_ARRAY (2, ligature);
+};
+
+struct LigatureSubstFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    return
+    + hb_zip (this+coverage, ligatureSet)
+    | hb_filter (*glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map ([this, glyphs] (const OffsetTo<LigatureSet> &_)
+	      { return (this+_).intersects (glyphs); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    + hb_zip (this+coverage, ligatureSet)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const LigatureSet &_) { _.closure (c); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+
+    + hb_zip (this+coverage, ligatureSet)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([c] (const LigatureSet &_) { _.collect_glyphs (c); })
+    ;
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    unsigned int index = (this+coverage).get_coverage (c->glyphs[0]);
+    if (likely (index == NOT_COVERED)) return false;
+
+    const LigatureSet &lig_set = this+ligatureSet[index];
+    return lig_set.would_apply (c);
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur ().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const LigatureSet &lig_set = this+ligatureSet[index];
+    return_trace (lig_set.apply (c));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> first_glyphs,
+		  hb_array_t<const unsigned int> ligature_per_first_glyph_count_list,
+		  hb_array_t<const HBGlyphID> ligatures_list,
+		  hb_array_t<const unsigned int> component_count_list,
+		  hb_array_t<const HBGlyphID> component_list /* Starting from second for each ligature */)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+    if (unlikely (!ligatureSet.serialize (c, first_glyphs.length))) return_trace (false);
+    for (unsigned int i = 0; i < first_glyphs.length; i++)
+    {
+      unsigned int ligature_count = ligature_per_first_glyph_count_list[i];
+      if (unlikely (!ligatureSet[i].serialize (c, this)
+				   .serialize (c,
+					       ligatures_list.sub_array (0, ligature_count),
+					       component_count_list.sub_array (0, ligature_count),
+					       component_list))) return_trace (false);
+      ligatures_list += ligature_count;
+      component_count_list += ligature_count;
+    }
+    return_trace (coverage.serialize (c, this).serialize (c, first_glyphs));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, ligatureSet)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->ligatureSet, this), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && ligatureSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of Substitution table */
+  OffsetArrayOf<LigatureSet>
+		ligatureSet;		/* Array LigatureSet tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, ligatureSet);
+};
+
+struct LigatureSubst
+{
+  bool serialize (hb_serialize_context_t *c,
+		  hb_sorted_array_t<const HBGlyphID> first_glyphs,
+		  hb_array_t<const unsigned int> ligature_per_first_glyph_count_list,
+		  hb_array_t<const HBGlyphID> ligatures_list,
+		  hb_array_t<const unsigned int> component_count_list,
+		  hb_array_t<const HBGlyphID> component_list /* Starting from second for each ligature */)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!c->extend_min (u.format))) return_trace (false);
+    unsigned int format = 1;
+    u.format = format;
+    switch (u.format) {
+    case 1: return_trace (u.format1.serialize (c,
+					       first_glyphs,
+					       ligature_per_first_glyph_count_list,
+					       ligatures_list,
+					       component_count_list,
+					       component_list));
+    default:return_trace (false);
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  LigatureSubstFormat1	format1;
+  } u;
+};
+
+
+struct ContextSubst : Context {};
+
+struct ChainContextSubst : ChainContext {};
+
+struct ExtensionSubst : Extension<ExtensionSubst>
+{
+  typedef struct SubstLookupSubTable SubTable;
+  bool is_reverse () const;
+};
+
+
+struct ReverseChainSingleSubstFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    if (!(this+coverage).intersects (glyphs))
+      return false;
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+
+    unsigned int count;
+
+    count = backtrack.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (!(this+backtrack[i]).intersects (glyphs))
+	return false;
+
+    count = lookahead.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (!(this+lookahead[i]).intersects (glyphs))
+	return false;
+
+    return true;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    if (!intersects (c->glyphs)) return;
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    const ArrayOf<HBGlyphID> &substitute = StructAfter<ArrayOf<HBGlyphID>> (lookahead);
+
+    + hb_zip (this+coverage, substitute)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_sink (c->output)
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    if (unlikely (!(this+coverage).collect_coverage (c->input))) return;
+
+    unsigned int count;
+
+    count = backtrack.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!(this+backtrack[i]).collect_coverage (c->before))) return;
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    count = lookahead.len;
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!(this+lookahead[i]).collect_coverage (c->after))) return;
+
+    const ArrayOf<HBGlyphID> &substitute = StructAfter<ArrayOf<HBGlyphID>> (lookahead);
+    count = substitute.len;
+    c->output->add_array (substitute.arrayZ, substitute.len);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  { return c->len == 1 && (this+coverage).get_coverage (c->glyphs[0]) != NOT_COVERED; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    if (unlikely (c->nesting_level_left != HB_MAX_NESTING_LEVEL))
+      return_trace (false); /* No chaining to this type */
+
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur ().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    const ArrayOf<HBGlyphID> &substitute = StructAfter<ArrayOf<HBGlyphID>> (lookahead);
+
+    if (unlikely (index >= substitute.len)) return_trace (false);
+
+    unsigned int start_index = 0, end_index = 0;
+    if (match_backtrack (c,
+			 backtrack.len, (HBUINT16 *) backtrack.arrayZ,
+			 match_coverage, this,
+			 &start_index) &&
+	match_lookahead (c,
+			 lookahead.len, (HBUINT16 *) lookahead.arrayZ,
+			 match_coverage, this,
+			 1, &end_index))
+    {
+      c->buffer->unsafe_to_break_from_outbuffer (start_index, end_index);
+      c->replace_glyph_inplace (substitute[index]);
+      /* Note: We DON'T decrease buffer->idx.  The main loop does it
+       * for us.  This is useful for preventing surprises if someone
+       * calls us through a Context lookup. */
+      return_trace (true);
+    }
+
+    return_trace (false);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    // TODO(subset)
+    return_trace (false);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!(coverage.sanitize (c, this) && backtrack.sanitize (c, this)))
+      return_trace (false);
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    if (!lookahead.sanitize (c, this))
+      return_trace (false);
+    const ArrayOf<HBGlyphID> &substitute = StructAfter<ArrayOf<HBGlyphID>> (lookahead);
+    return_trace (substitute.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of table */
+  OffsetArrayOf<Coverage>
+		backtrack;		/* Array of coverage tables
+					 * in backtracking sequence, in glyph
+					 * sequence order */
+  OffsetArrayOf<Coverage>
+		lookaheadX;		/* Array of coverage tables
+					 * in lookahead sequence, in glyph
+					 * sequence order */
+  ArrayOf<HBGlyphID>
+		substituteX;		/* Array of substitute
+					 * GlyphIDs--ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_MIN (10);
+};
+
+struct ReverseChainSingleSubst
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16				format;		/* Format identifier */
+  ReverseChainSingleSubstFormat1	format1;
+  } u;
+};
+
+
+
+/*
+ * SubstLookup
+ */
+
+struct SubstLookupSubTable
+{
+  friend struct Lookup;
+  friend struct SubstLookup;
+
+  enum Type {
+    Single		= 1,
+    Multiple		= 2,
+    Alternate		= 3,
+    Ligature		= 4,
+    Context		= 5,
+    ChainContext	= 6,
+    Extension		= 7,
+    ReverseChainSingle	= 8
+  };
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, unsigned int lookup_type, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, lookup_type);
+    switch (lookup_type) {
+    case Single:		return_trace (u.single.dispatch (c, hb_forward<Ts> (ds)...));
+    case Multiple:		return_trace (u.multiple.dispatch (c, hb_forward<Ts> (ds)...));
+    case Alternate:		return_trace (u.alternate.dispatch (c, hb_forward<Ts> (ds)...));
+    case Ligature:		return_trace (u.ligature.dispatch (c, hb_forward<Ts> (ds)...));
+    case Context:		return_trace (u.context.dispatch (c, hb_forward<Ts> (ds)...));
+    case ChainContext:		return_trace (u.chainContext.dispatch (c, hb_forward<Ts> (ds)...));
+    case Extension:		return_trace (u.extension.dispatch (c, hb_forward<Ts> (ds)...));
+    case ReverseChainSingle:	return_trace (u.reverseChainContextSingle.dispatch (c, hb_forward<Ts> (ds)...));
+    default:			return_trace (c->default_return_value ());
+    }
+  }
+
+  bool intersects (const hb_set_t *glyphs, unsigned int lookup_type) const
+  {
+    hb_intersects_context_t c (glyphs);
+    return dispatch (&c, lookup_type);
+  }
+
+  protected:
+  union {
+  SingleSubst			single;
+  MultipleSubst			multiple;
+  AlternateSubst		alternate;
+  LigatureSubst			ligature;
+  ContextSubst			context;
+  ChainContextSubst		chainContext;
+  ExtensionSubst		extension;
+  ReverseChainSingleSubst	reverseChainContextSingle;
+  } u;
+  public:
+  DEFINE_SIZE_MIN (0);
+};
+
+
+struct SubstLookup : Lookup
+{
+  typedef SubstLookupSubTable SubTable;
+
+  const SubTable& get_subtable (unsigned int i) const
+  { return Lookup::get_subtable<SubTable> (i); }
+
+  static inline bool lookup_type_is_reverse (unsigned int lookup_type)
+  { return lookup_type == SubTable::ReverseChainSingle; }
+
+  bool is_reverse () const
+  {
+    unsigned int type = get_type ();
+    if (unlikely (type == SubTable::Extension))
+      return reinterpret_cast<const ExtensionSubst &> (get_subtable (0)).is_reverse ();
+    return lookup_type_is_reverse (type);
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    return_trace (dispatch (c));
+  }
+
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    hb_intersects_context_t c (glyphs);
+    return dispatch (&c);
+  }
+
+  hb_closure_context_t::return_t closure (hb_closure_context_t *c, unsigned int this_index) const
+  {
+    if (!c->should_visit_lookup (this_index))
+      return hb_closure_context_t::default_return_value ();
+
+    c->set_recurse_func (dispatch_closure_recurse_func);
+
+    hb_closure_context_t::return_t ret = dispatch (c);
+
+    c->flush ();
+
+    return ret;
+  }
+
+  hb_closure_lookups_context_t::return_t closure_lookups (hb_closure_lookups_context_t *c, unsigned this_index) const
+  {
+    if (c->is_lookup_visited (this_index))
+      return hb_closure_lookups_context_t::default_return_value ();
+
+    c->set_lookup_visited (this_index);
+    if (!intersects (c->glyphs))
+    {
+      c->set_lookup_inactive (this_index);
+      return hb_closure_lookups_context_t::default_return_value ();
+    }
+
+    c->set_recurse_func (dispatch_closure_lookups_recurse_func);
+
+    hb_closure_lookups_context_t::return_t ret = dispatch (c);
+    return ret;
+  }
+
+  hb_collect_glyphs_context_t::return_t collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    c->set_recurse_func (dispatch_recurse_func<hb_collect_glyphs_context_t>);
+    return dispatch (c);
+  }
+
+  template <typename set_t>
+  void collect_coverage (set_t *glyphs) const
+  {
+    hb_collect_coverage_context_t<set_t> c (glyphs);
+    dispatch (&c);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c,
+		    const hb_ot_layout_lookup_accelerator_t *accel) const
+  {
+    if (unlikely (!c->len)) return false;
+    if (!accel->may_have (c->glyphs[0])) return false;
+      return dispatch (c);
+  }
+
+  static inline bool apply_recurse_func (hb_ot_apply_context_t *c, unsigned int lookup_index);
+
+  SubTable& serialize_subtable (hb_serialize_context_t *c,
+				unsigned int i)
+  { return get_subtables<SubTable> ()[i].serialize (c, this); }
+
+  bool serialize_single (hb_serialize_context_t *c,
+			 uint32_t lookup_props,
+			 hb_sorted_array_t<const HBGlyphID> glyphs,
+			 hb_array_t<const HBGlyphID> substitutes)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!Lookup::serialize (c, SubTable::Single, lookup_props, 1))) return_trace (false);
+    return_trace (serialize_subtable (c, 0).u.single.
+		  serialize (c, hb_zip (glyphs, substitutes)));
+  }
+
+  bool serialize_multiple (hb_serialize_context_t *c,
+			   uint32_t lookup_props,
+			   hb_sorted_array_t<const HBGlyphID> glyphs,
+			   hb_array_t<const unsigned int> substitute_len_list,
+			   hb_array_t<const HBGlyphID> substitute_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!Lookup::serialize (c, SubTable::Multiple, lookup_props, 1))) return_trace (false);
+    return_trace (serialize_subtable (c, 0).u.multiple.
+		  serialize (c,
+			     glyphs,
+			     substitute_len_list,
+			     substitute_glyphs_list));
+  }
+
+  bool serialize_alternate (hb_serialize_context_t *c,
+			    uint32_t lookup_props,
+			    hb_sorted_array_t<const HBGlyphID> glyphs,
+			    hb_array_t<const unsigned int> alternate_len_list,
+			    hb_array_t<const HBGlyphID> alternate_glyphs_list)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!Lookup::serialize (c, SubTable::Alternate, lookup_props, 1))) return_trace (false);
+    return_trace (serialize_subtable (c, 0).u.alternate.
+		  serialize (c,
+			     glyphs,
+			     alternate_len_list,
+			     alternate_glyphs_list));
+  }
+
+  bool serialize_ligature (hb_serialize_context_t *c,
+			   uint32_t lookup_props,
+			   hb_sorted_array_t<const HBGlyphID> first_glyphs,
+			   hb_array_t<const unsigned int> ligature_per_first_glyph_count_list,
+			   hb_array_t<const HBGlyphID> ligatures_list,
+			   hb_array_t<const unsigned int> component_count_list,
+			   hb_array_t<const HBGlyphID> component_list /* Starting from second for each ligature */)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!Lookup::serialize (c, SubTable::Ligature, lookup_props, 1))) return_trace (false);
+    return_trace (serialize_subtable (c, 0).u.ligature.
+		  serialize (c,
+			     first_glyphs,
+			     ligature_per_first_glyph_count_list,
+			     ligatures_list,
+			     component_count_list,
+			     component_list));
+  }
+
+  template <typename context_t>
+  static inline typename context_t::return_t dispatch_recurse_func (context_t *c, unsigned int lookup_index);
+
+  static inline hb_closure_context_t::return_t dispatch_closure_recurse_func (hb_closure_context_t *c, unsigned int lookup_index)
+  {
+    if (!c->should_visit_lookup (lookup_index))
+      return hb_empty_t ();
+
+    hb_closure_context_t::return_t ret = dispatch_recurse_func (c, lookup_index);
+
+    /* While in theory we should flush here, it will cause timeouts because a recursive
+     * lookup can keep growing the glyph set.  Skip, and outer loop will retry up to
+     * HB_CLOSURE_MAX_STAGES time, which should be enough for every realistic font. */
+    //c->flush ();
+
+    return ret;
+  }
+
+  HB_INTERNAL static hb_closure_lookups_context_t::return_t dispatch_closure_lookups_recurse_func (hb_closure_lookups_context_t *c, unsigned lookup_index);
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  { return Lookup::dispatch<SubTable> (c, hb_forward<Ts> (ds)...); }
+
+  bool subset (hb_subset_context_t *c) const
+  { return Lookup::subset<SubTable> (c); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return Lookup::sanitize<SubTable> (c); }
+};
+
+/*
+ * GSUB -- Glyph Substitution
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/gsub
+ */
+
+struct GSUB : GSUBGPOS
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_GSUB;
+
+  const SubstLookup& get_lookup (unsigned int i) const
+  { return static_cast<const SubstLookup &> (GSUBGPOS::get_lookup (i)); }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    hb_subset_layout_context_t l (c, tableTag, c->plan->gsub_lookups, c->plan->gsub_features);
+    return GSUBGPOS::subset<SubstLookup> (&l);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return GSUBGPOS::sanitize<SubstLookup> (c); }
+
+  HB_INTERNAL bool is_blocklisted (hb_blob_t *blob,
+				   hb_face_t *face) const;
+
+  void closure_lookups (hb_face_t      *face,
+			const hb_set_t *glyphs,
+			hb_set_t       *lookup_indexes /* IN/OUT */) const
+  { GSUBGPOS::closure_lookups<SubstLookup> (face, glyphs, lookup_indexes); }
+
+  typedef GSUBGPOS::accelerator_t<GSUB> accelerator_t;
+};
+
+
+struct GSUB_accelerator_t : GSUB::accelerator_t {};
+
+
+/* Out-of-class implementation for methods recursing */
+
+#ifndef HB_NO_OT_LAYOUT
+/*static*/ inline bool ExtensionSubst::is_reverse () const
+{
+  return SubstLookup::lookup_type_is_reverse (get_type ());
+}
+template <typename context_t>
+/*static*/ typename context_t::return_t SubstLookup::dispatch_recurse_func (context_t *c, unsigned int lookup_index)
+{
+  const SubstLookup &l = c->face->table.GSUB.get_relaxed ()->table->get_lookup (lookup_index);
+  return l.dispatch (c);
+}
+
+/*static*/ inline hb_closure_lookups_context_t::return_t SubstLookup::dispatch_closure_lookups_recurse_func (hb_closure_lookups_context_t *c, unsigned this_index)
+{
+  const SubstLookup &l = c->face->table.GSUB.get_relaxed ()->table->get_lookup (this_index);
+  return l.closure_lookups (c, this_index);
+}
+
+/*static*/ bool SubstLookup::apply_recurse_func (hb_ot_apply_context_t *c, unsigned int lookup_index)
+{
+  const SubstLookup &l = c->face->table.GSUB.get_relaxed ()->table->get_lookup (lookup_index);
+  unsigned int saved_lookup_props = c->lookup_props;
+  unsigned int saved_lookup_index = c->lookup_index;
+  c->set_lookup_index (lookup_index);
+  c->set_lookup_props (l.get_props ());
+  bool ret = l.dispatch (c);
+  c->set_lookup_index (saved_lookup_index);
+  c->set_lookup_props (saved_lookup_props);
+  return ret;
+}
+#endif
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_GSUB_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-gsubgpos.hh b/thirdparty/harfbuzz/src/hb-ot-layout-gsubgpos.hh
new file mode 100644
index 0000000000..03a578cec0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-gsubgpos.hh
@@ -0,0 +1,3422 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_GSUBGPOS_HH
+#define HB_OT_LAYOUT_GSUBGPOS_HH
+
+#include "hb.hh"
+#include "hb-buffer.hh"
+#include "hb-map.hh"
+#include "hb-set.hh"
+#include "hb-ot-map.hh"
+#include "hb-ot-layout-common.hh"
+#include "hb-ot-layout-gdef-table.hh"
+
+
+namespace OT {
+
+
+struct hb_intersects_context_t :
+       hb_dispatch_context_t<hb_intersects_context_t, bool>
+{
+  template <typename T>
+  return_t dispatch (const T &obj) { return obj.intersects (this->glyphs); }
+  static return_t default_return_value () { return false; }
+  bool stop_sublookup_iteration (return_t r) const { return r; }
+
+  const hb_set_t *glyphs;
+
+  hb_intersects_context_t (const hb_set_t *glyphs_) :
+			     glyphs (glyphs_) {}
+};
+
+struct hb_closure_context_t :
+       hb_dispatch_context_t<hb_closure_context_t>
+{
+  typedef return_t (*recurse_func_t) (hb_closure_context_t *c, unsigned int lookup_index);
+  template <typename T>
+  return_t dispatch (const T &obj) { obj.closure (this); return hb_empty_t (); }
+  static return_t default_return_value () { return hb_empty_t (); }
+  void recurse (unsigned int lookup_index)
+  {
+    if (unlikely (nesting_level_left == 0 || !recurse_func))
+      return;
+
+    nesting_level_left--;
+    recurse_func (this, lookup_index);
+    nesting_level_left++;
+  }
+
+  bool lookup_limit_exceeded ()
+  { return lookup_count > HB_MAX_LOOKUP_INDICES; }
+
+  bool should_visit_lookup (unsigned int lookup_index)
+  {
+    if (lookup_count++ > HB_MAX_LOOKUP_INDICES)
+      return false;
+
+    if (is_lookup_done (lookup_index))
+      return false;
+
+    done_lookups->set (lookup_index, glyphs->get_population ());
+    return true;
+  }
+
+  bool is_lookup_done (unsigned int lookup_index)
+  {
+    if (done_lookups->in_error ())
+      return true;
+
+    /* Have we visited this lookup with the current set of glyphs? */
+    return done_lookups->get (lookup_index) == glyphs->get_population ();
+  }
+
+  hb_face_t *face;
+  hb_set_t *glyphs;
+  hb_set_t output[1];
+  recurse_func_t recurse_func;
+  unsigned int nesting_level_left;
+
+  hb_closure_context_t (hb_face_t *face_,
+			hb_set_t *glyphs_,
+			hb_map_t *done_lookups_,
+			unsigned int nesting_level_left_ = HB_MAX_NESTING_LEVEL) :
+			  face (face_),
+			  glyphs (glyphs_),
+			  recurse_func (nullptr),
+			  nesting_level_left (nesting_level_left_),
+			  done_lookups (done_lookups_),
+			  lookup_count (0)
+  {}
+
+  ~hb_closure_context_t () { flush (); }
+
+  void set_recurse_func (recurse_func_t func) { recurse_func = func; }
+
+  void flush ()
+  {
+    hb_set_del_range (output, face->get_num_glyphs (), hb_set_get_max (output));	/* Remove invalid glyphs. */
+    hb_set_union (glyphs, output);
+    hb_set_clear (output);
+  }
+
+  private:
+  hb_map_t *done_lookups;
+  unsigned int lookup_count;
+};
+
+struct hb_closure_lookups_context_t :
+       hb_dispatch_context_t<hb_closure_lookups_context_t>
+{
+  typedef return_t (*recurse_func_t) (hb_closure_lookups_context_t *c, unsigned lookup_index);
+  template <typename T>
+  return_t dispatch (const T &obj) { obj.closure_lookups (this); return hb_empty_t (); }
+  static return_t default_return_value () { return hb_empty_t (); }
+  void recurse (unsigned lookup_index)
+  {
+    if (unlikely (nesting_level_left == 0 || !recurse_func))
+      return;
+
+    /* Return if new lookup was recursed to before. */
+    if (is_lookup_visited (lookup_index))
+      return;
+
+    set_lookup_visited (lookup_index);
+    nesting_level_left--;
+    recurse_func (this, lookup_index);
+    nesting_level_left++;
+  }
+
+  void set_lookup_visited (unsigned lookup_index)
+  { visited_lookups->add (lookup_index); }
+
+  void set_lookup_inactive (unsigned lookup_index)
+  { inactive_lookups->add (lookup_index); }
+
+  bool lookup_limit_exceeded ()
+  { return lookup_count > HB_MAX_LOOKUP_INDICES; }
+
+  bool is_lookup_visited (unsigned lookup_index)
+  {
+    if (lookup_count++ > HB_MAX_LOOKUP_INDICES)
+      return true;
+
+    if (visited_lookups->in_error ())
+      return true;
+
+    return visited_lookups->has (lookup_index);
+  }
+
+  hb_face_t *face;
+  const hb_set_t *glyphs;
+  recurse_func_t recurse_func;
+  unsigned int nesting_level_left;
+
+  hb_closure_lookups_context_t (hb_face_t *face_,
+				const hb_set_t *glyphs_,
+				hb_set_t *visited_lookups_,
+				hb_set_t *inactive_lookups_,
+				unsigned nesting_level_left_ = HB_MAX_NESTING_LEVEL) :
+				face (face_),
+				glyphs (glyphs_),
+				recurse_func (nullptr),
+				nesting_level_left (nesting_level_left_),
+				visited_lookups (visited_lookups_),
+				inactive_lookups (inactive_lookups_),
+				lookup_count (0) {}
+
+  void set_recurse_func (recurse_func_t func) { recurse_func = func; }
+
+  private:
+  hb_set_t *visited_lookups;
+  hb_set_t *inactive_lookups;
+  unsigned int lookup_count;
+};
+
+struct hb_would_apply_context_t :
+       hb_dispatch_context_t<hb_would_apply_context_t, bool>
+{
+  template <typename T>
+  return_t dispatch (const T &obj) { return obj.would_apply (this); }
+  static return_t default_return_value () { return false; }
+  bool stop_sublookup_iteration (return_t r) const { return r; }
+
+  hb_face_t *face;
+  const hb_codepoint_t *glyphs;
+  unsigned int len;
+  bool zero_context;
+
+  hb_would_apply_context_t (hb_face_t *face_,
+			    const hb_codepoint_t *glyphs_,
+			    unsigned int len_,
+			    bool zero_context_) :
+			      face (face_),
+			      glyphs (glyphs_),
+			      len (len_),
+			      zero_context (zero_context_) {}
+};
+
+struct hb_collect_glyphs_context_t :
+       hb_dispatch_context_t<hb_collect_glyphs_context_t>
+{
+  typedef return_t (*recurse_func_t) (hb_collect_glyphs_context_t *c, unsigned int lookup_index);
+  template <typename T>
+  return_t dispatch (const T &obj) { obj.collect_glyphs (this); return hb_empty_t (); }
+  static return_t default_return_value () { return hb_empty_t (); }
+  void recurse (unsigned int lookup_index)
+  {
+    if (unlikely (nesting_level_left == 0 || !recurse_func))
+      return;
+
+    /* Note that GPOS sets recurse_func to nullptr already, so it doesn't get
+     * past the previous check.  For GSUB, we only want to collect the output
+     * glyphs in the recursion.  If output is not requested, we can go home now.
+     *
+     * Note further, that the above is not exactly correct.  A recursed lookup
+     * is allowed to match input that is not matched in the context, but that's
+     * not how most fonts are built.  It's possible to relax that and recurse
+     * with all sets here if it proves to be an issue.
+     */
+
+    if (output == hb_set_get_empty ())
+      return;
+
+    /* Return if new lookup was recursed to before. */
+    if (recursed_lookups->has (lookup_index))
+      return;
+
+    hb_set_t *old_before = before;
+    hb_set_t *old_input  = input;
+    hb_set_t *old_after  = after;
+    before = input = after = hb_set_get_empty ();
+
+    nesting_level_left--;
+    recurse_func (this, lookup_index);
+    nesting_level_left++;
+
+    before = old_before;
+    input  = old_input;
+    after  = old_after;
+
+    recursed_lookups->add (lookup_index);
+  }
+
+  hb_face_t *face;
+  hb_set_t *before;
+  hb_set_t *input;
+  hb_set_t *after;
+  hb_set_t *output;
+  recurse_func_t recurse_func;
+  hb_set_t *recursed_lookups;
+  unsigned int nesting_level_left;
+
+  hb_collect_glyphs_context_t (hb_face_t *face_,
+			       hb_set_t  *glyphs_before, /* OUT.  May be NULL */
+			       hb_set_t  *glyphs_input,  /* OUT.  May be NULL */
+			       hb_set_t  *glyphs_after,  /* OUT.  May be NULL */
+			       hb_set_t  *glyphs_output, /* OUT.  May be NULL */
+			       unsigned int nesting_level_left_ = HB_MAX_NESTING_LEVEL) :
+			      face (face_),
+			      before (glyphs_before ? glyphs_before : hb_set_get_empty ()),
+			      input  (glyphs_input  ? glyphs_input  : hb_set_get_empty ()),
+			      after  (glyphs_after  ? glyphs_after  : hb_set_get_empty ()),
+			      output (glyphs_output ? glyphs_output : hb_set_get_empty ()),
+			      recurse_func (nullptr),
+			      recursed_lookups (hb_set_create ()),
+			      nesting_level_left (nesting_level_left_) {}
+  ~hb_collect_glyphs_context_t () { hb_set_destroy (recursed_lookups); }
+
+  void set_recurse_func (recurse_func_t func) { recurse_func = func; }
+};
+
+
+
+template <typename set_t>
+struct hb_collect_coverage_context_t :
+       hb_dispatch_context_t<hb_collect_coverage_context_t<set_t>, const Coverage &>
+{
+  typedef const Coverage &return_t; // Stoopid that we have to dupe this here.
+  template <typename T>
+  return_t dispatch (const T &obj) { return obj.get_coverage (); }
+  static return_t default_return_value () { return Null (Coverage); }
+  bool stop_sublookup_iteration (return_t r) const
+  {
+    r.collect_coverage (set);
+    return false;
+  }
+
+  hb_collect_coverage_context_t (set_t *set_) :
+				   set (set_) {}
+
+  set_t *set;
+};
+
+
+struct hb_ot_apply_context_t :
+       hb_dispatch_context_t<hb_ot_apply_context_t, bool, HB_DEBUG_APPLY>
+{
+  struct matcher_t
+  {
+    matcher_t () :
+	     lookup_props (0),
+	     ignore_zwnj (false),
+	     ignore_zwj (false),
+	     mask (-1),
+#define arg1(arg) (arg) /* Remove the macro to see why it's needed! */
+	     syllable arg1(0),
+#undef arg1
+	     match_func (nullptr),
+	     match_data (nullptr) {}
+
+    typedef bool (*match_func_t) (hb_codepoint_t glyph_id, const HBUINT16 &value, const void *data);
+
+    void set_ignore_zwnj (bool ignore_zwnj_) { ignore_zwnj = ignore_zwnj_; }
+    void set_ignore_zwj (bool ignore_zwj_) { ignore_zwj = ignore_zwj_; }
+    void set_lookup_props (unsigned int lookup_props_) { lookup_props = lookup_props_; }
+    void set_mask (hb_mask_t mask_) { mask = mask_; }
+    void set_syllable (uint8_t syllable_)  { syllable = syllable_; }
+    void set_match_func (match_func_t match_func_,
+			 const void *match_data_)
+    { match_func = match_func_; match_data = match_data_; }
+
+    enum may_match_t {
+      MATCH_NO,
+      MATCH_YES,
+      MATCH_MAYBE
+    };
+
+    may_match_t may_match (const hb_glyph_info_t &info,
+			   const HBUINT16        *glyph_data) const
+    {
+      if (!(info.mask & mask) ||
+	  (syllable && syllable != info.syllable ()))
+	return MATCH_NO;
+
+      if (match_func)
+	return match_func (info.codepoint, *glyph_data, match_data) ? MATCH_YES : MATCH_NO;
+
+      return MATCH_MAYBE;
+    }
+
+    enum may_skip_t {
+      SKIP_NO,
+      SKIP_YES,
+      SKIP_MAYBE
+    };
+
+    may_skip_t may_skip (const hb_ot_apply_context_t *c,
+			 const hb_glyph_info_t       &info) const
+    {
+      if (!c->check_glyph_property (&info, lookup_props))
+	return SKIP_YES;
+
+      if (unlikely (_hb_glyph_info_is_default_ignorable_and_not_hidden (&info) &&
+		    (ignore_zwnj || !_hb_glyph_info_is_zwnj (&info)) &&
+		    (ignore_zwj || !_hb_glyph_info_is_zwj (&info))))
+	return SKIP_MAYBE;
+
+      return SKIP_NO;
+    }
+
+    protected:
+    unsigned int lookup_props;
+    bool ignore_zwnj;
+    bool ignore_zwj;
+    hb_mask_t mask;
+    uint8_t syllable;
+    match_func_t match_func;
+    const void *match_data;
+  };
+
+  struct skipping_iterator_t
+  {
+    void init (hb_ot_apply_context_t *c_, bool context_match = false)
+    {
+      c = c_;
+      match_glyph_data = nullptr;
+      matcher.set_match_func (nullptr, nullptr);
+      matcher.set_lookup_props (c->lookup_props);
+      /* Ignore ZWNJ if we are matching GPOS, or matching GSUB context and asked to. */
+      matcher.set_ignore_zwnj (c->table_index == 1 || (context_match && c->auto_zwnj));
+      /* Ignore ZWJ if we are matching context, or asked to. */
+      matcher.set_ignore_zwj  (context_match || c->auto_zwj);
+      matcher.set_mask (context_match ? -1 : c->lookup_mask);
+    }
+    void set_lookup_props (unsigned int lookup_props)
+    {
+      matcher.set_lookup_props (lookup_props);
+    }
+    void set_match_func (matcher_t::match_func_t match_func_,
+			 const void *match_data_,
+			 const HBUINT16 glyph_data[])
+    {
+      matcher.set_match_func (match_func_, match_data_);
+      match_glyph_data = glyph_data;
+    }
+
+    void reset (unsigned int start_index_,
+		unsigned int num_items_)
+    {
+      idx = start_index_;
+      num_items = num_items_;
+      end = c->buffer->len;
+      matcher.set_syllable (start_index_ == c->buffer->idx ? c->buffer->cur().syllable () : 0);
+    }
+
+    void reject ()
+    {
+      num_items++;
+      if (match_glyph_data) match_glyph_data--;
+    }
+
+    matcher_t::may_skip_t
+    may_skip (const hb_glyph_info_t &info) const
+    { return matcher.may_skip (c, info); }
+
+    bool next ()
+    {
+      assert (num_items > 0);
+      while (idx + num_items < end)
+      {
+	idx++;
+	const hb_glyph_info_t &info = c->buffer->info[idx];
+
+	matcher_t::may_skip_t skip = matcher.may_skip (c, info);
+	if (unlikely (skip == matcher_t::SKIP_YES))
+	  continue;
+
+	matcher_t::may_match_t match = matcher.may_match (info, match_glyph_data);
+	if (match == matcher_t::MATCH_YES ||
+	    (match == matcher_t::MATCH_MAYBE &&
+	     skip == matcher_t::SKIP_NO))
+	{
+	  num_items--;
+	  if (match_glyph_data) match_glyph_data++;
+	  return true;
+	}
+
+	if (skip == matcher_t::SKIP_NO)
+	  return false;
+      }
+      return false;
+    }
+    bool prev ()
+    {
+      assert (num_items > 0);
+      while (idx > num_items - 1)
+      {
+	idx--;
+	const hb_glyph_info_t &info = c->buffer->out_info[idx];
+
+	matcher_t::may_skip_t skip = matcher.may_skip (c, info);
+	if (unlikely (skip == matcher_t::SKIP_YES))
+	  continue;
+
+	matcher_t::may_match_t match = matcher.may_match (info, match_glyph_data);
+	if (match == matcher_t::MATCH_YES ||
+	    (match == matcher_t::MATCH_MAYBE &&
+	     skip == matcher_t::SKIP_NO))
+	{
+	  num_items--;
+	  if (match_glyph_data) match_glyph_data++;
+	  return true;
+	}
+
+	if (skip == matcher_t::SKIP_NO)
+	  return false;
+      }
+      return false;
+    }
+
+    unsigned int idx;
+    protected:
+    hb_ot_apply_context_t *c;
+    matcher_t matcher;
+    const HBUINT16 *match_glyph_data;
+
+    unsigned int num_items;
+    unsigned int end;
+  };
+
+
+  const char *get_name () { return "APPLY"; }
+  typedef return_t (*recurse_func_t) (hb_ot_apply_context_t *c, unsigned int lookup_index);
+  template <typename T>
+  return_t dispatch (const T &obj) { return obj.apply (this); }
+  static return_t default_return_value () { return false; }
+  bool stop_sublookup_iteration (return_t r) const { return r; }
+  return_t recurse (unsigned int sub_lookup_index)
+  {
+    if (unlikely (nesting_level_left == 0 || !recurse_func || buffer->max_ops-- <= 0))
+      return default_return_value ();
+
+    nesting_level_left--;
+    bool ret = recurse_func (this, sub_lookup_index);
+    nesting_level_left++;
+    return ret;
+  }
+
+  skipping_iterator_t iter_input, iter_context;
+
+  hb_font_t *font;
+  hb_face_t *face;
+  hb_buffer_t *buffer;
+  recurse_func_t recurse_func;
+  const GDEF &gdef;
+  const VariationStore &var_store;
+
+  hb_direction_t direction;
+  hb_mask_t lookup_mask;
+  unsigned int table_index; /* GSUB/GPOS */
+  unsigned int lookup_index;
+  unsigned int lookup_props;
+  unsigned int nesting_level_left;
+
+  bool has_glyph_classes;
+  bool auto_zwnj;
+  bool auto_zwj;
+  bool random;
+
+  uint32_t random_state;
+
+
+  hb_ot_apply_context_t (unsigned int table_index_,
+			 hb_font_t *font_,
+			 hb_buffer_t *buffer_) :
+			iter_input (), iter_context (),
+			font (font_), face (font->face), buffer (buffer_),
+			recurse_func (nullptr),
+			gdef (
+#ifndef HB_NO_OT_LAYOUT
+			      *face->table.GDEF->table
+#else
+			      Null (GDEF)
+#endif
+			     ),
+			var_store (gdef.get_var_store ()),
+			direction (buffer_->props.direction),
+			lookup_mask (1),
+			table_index (table_index_),
+			lookup_index ((unsigned int) -1),
+			lookup_props (0),
+			nesting_level_left (HB_MAX_NESTING_LEVEL),
+			has_glyph_classes (gdef.has_glyph_classes ()),
+			auto_zwnj (true),
+			auto_zwj (true),
+			random (false),
+			random_state (1) { init_iters (); }
+
+  void init_iters ()
+  {
+    iter_input.init (this, false);
+    iter_context.init (this, true);
+  }
+
+  void set_lookup_mask (hb_mask_t mask) { lookup_mask = mask; init_iters (); }
+  void set_auto_zwj (bool auto_zwj_) { auto_zwj = auto_zwj_; init_iters (); }
+  void set_auto_zwnj (bool auto_zwnj_) { auto_zwnj = auto_zwnj_; init_iters (); }
+  void set_random (bool random_) { random = random_; }
+  void set_recurse_func (recurse_func_t func) { recurse_func = func; }
+  void set_lookup_index (unsigned int lookup_index_) { lookup_index = lookup_index_; }
+  void set_lookup_props (unsigned int lookup_props_) { lookup_props = lookup_props_; init_iters (); }
+
+  uint32_t random_number ()
+  {
+    /* http://www.cplusplus.com/reference/random/minstd_rand/ */
+    random_state = random_state * 48271 % 2147483647;
+    return random_state;
+  }
+
+  bool match_properties_mark (hb_codepoint_t  glyph,
+			      unsigned int    glyph_props,
+			      unsigned int    match_props) const
+  {
+    /* If using mark filtering sets, the high short of
+     * match_props has the set index.
+     */
+    if (match_props & LookupFlag::UseMarkFilteringSet)
+      return gdef.mark_set_covers (match_props >> 16, glyph);
+
+    /* The second byte of match_props has the meaning
+     * "ignore marks of attachment type different than
+     * the attachment type specified."
+     */
+    if (match_props & LookupFlag::MarkAttachmentType)
+      return (match_props & LookupFlag::MarkAttachmentType) == (glyph_props & LookupFlag::MarkAttachmentType);
+
+    return true;
+  }
+
+  bool check_glyph_property (const hb_glyph_info_t *info,
+			     unsigned int  match_props) const
+  {
+    hb_codepoint_t glyph = info->codepoint;
+    unsigned int glyph_props = _hb_glyph_info_get_glyph_props (info);
+
+    /* Not covered, if, for example, glyph class is ligature and
+     * match_props includes LookupFlags::IgnoreLigatures
+     */
+    if (glyph_props & match_props & LookupFlag::IgnoreFlags)
+      return false;
+
+    if (unlikely (glyph_props & HB_OT_LAYOUT_GLYPH_PROPS_MARK))
+      return match_properties_mark (glyph, glyph_props, match_props);
+
+    return true;
+  }
+
+  void _set_glyph_props (hb_codepoint_t glyph_index,
+			  unsigned int class_guess = 0,
+			  bool ligature = false,
+			  bool component = false) const
+  {
+    unsigned int add_in = _hb_glyph_info_get_glyph_props (&buffer->cur()) &
+			  HB_OT_LAYOUT_GLYPH_PROPS_PRESERVE;
+    add_in |= HB_OT_LAYOUT_GLYPH_PROPS_SUBSTITUTED;
+    if (ligature)
+    {
+      add_in |= HB_OT_LAYOUT_GLYPH_PROPS_LIGATED;
+      /* In the only place that the MULTIPLIED bit is used, Uniscribe
+       * seems to only care about the "last" transformation between
+       * Ligature and Multiple substitutions.  Ie. if you ligate, expand,
+       * and ligate again, it forgives the multiplication and acts as
+       * if only ligation happened.  As such, clear MULTIPLIED bit.
+       */
+      add_in &= ~HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED;
+    }
+    if (component)
+      add_in |= HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED;
+    if (likely (has_glyph_classes))
+      _hb_glyph_info_set_glyph_props (&buffer->cur(), add_in | gdef.get_glyph_props (glyph_index));
+    else if (class_guess)
+      _hb_glyph_info_set_glyph_props (&buffer->cur(), add_in | class_guess);
+  }
+
+  void replace_glyph (hb_codepoint_t glyph_index) const
+  {
+    _set_glyph_props (glyph_index);
+    buffer->replace_glyph (glyph_index);
+  }
+  void replace_glyph_inplace (hb_codepoint_t glyph_index) const
+  {
+    _set_glyph_props (glyph_index);
+    buffer->cur().codepoint = glyph_index;
+  }
+  void replace_glyph_with_ligature (hb_codepoint_t glyph_index,
+				    unsigned int class_guess) const
+  {
+    _set_glyph_props (glyph_index, class_guess, true);
+    buffer->replace_glyph (glyph_index);
+  }
+  void output_glyph_for_component (hb_codepoint_t glyph_index,
+				   unsigned int class_guess) const
+  {
+    _set_glyph_props (glyph_index, class_guess, false, true);
+    buffer->output_glyph (glyph_index);
+  }
+};
+
+
+struct hb_get_subtables_context_t :
+       hb_dispatch_context_t<hb_get_subtables_context_t>
+{
+  template <typename Type>
+  static inline bool apply_to (const void *obj, OT::hb_ot_apply_context_t *c)
+  {
+    const Type *typed_obj = (const Type *) obj;
+    return typed_obj->apply (c);
+  }
+
+  typedef bool (*hb_apply_func_t) (const void *obj, OT::hb_ot_apply_context_t *c);
+
+  struct hb_applicable_t
+  {
+    template <typename T>
+    void init (const T &obj_, hb_apply_func_t apply_func_)
+    {
+      obj = &obj_;
+      apply_func = apply_func_;
+      digest.init ();
+      obj_.get_coverage ().collect_coverage (&digest);
+    }
+
+    bool apply (OT::hb_ot_apply_context_t *c) const
+    {
+      return digest.may_have (c->buffer->cur().codepoint) && apply_func (obj, c);
+    }
+
+    private:
+    const void *obj;
+    hb_apply_func_t apply_func;
+    hb_set_digest_t digest;
+  };
+
+  typedef hb_vector_t<hb_applicable_t> array_t;
+
+  /* Dispatch interface. */
+  template <typename T>
+  return_t dispatch (const T &obj)
+  {
+    hb_applicable_t *entry = array.push();
+    entry->init (obj, apply_to<T>);
+    return hb_empty_t ();
+  }
+  static return_t default_return_value () { return hb_empty_t (); }
+
+  hb_get_subtables_context_t (array_t &array_) :
+			      array (array_) {}
+
+  array_t &array;
+};
+
+
+
+
+typedef bool (*intersects_func_t) (const hb_set_t *glyphs, const HBUINT16 &value, const void *data);
+typedef void (*collect_glyphs_func_t) (hb_set_t *glyphs, const HBUINT16 &value, const void *data);
+typedef bool (*match_func_t) (hb_codepoint_t glyph_id, const HBUINT16 &value, const void *data);
+
+struct ContextClosureFuncs
+{
+  intersects_func_t intersects;
+};
+struct ContextCollectGlyphsFuncs
+{
+  collect_glyphs_func_t collect;
+};
+struct ContextApplyFuncs
+{
+  match_func_t match;
+};
+
+
+static inline bool intersects_glyph (const hb_set_t *glyphs, const HBUINT16 &value, const void *data HB_UNUSED)
+{
+  return glyphs->has (value);
+}
+static inline bool intersects_class (const hb_set_t *glyphs, const HBUINT16 &value, const void *data)
+{
+  const ClassDef &class_def = *reinterpret_cast<const ClassDef *>(data);
+  return class_def.intersects_class (glyphs, value);
+}
+static inline bool intersects_coverage (const hb_set_t *glyphs, const HBUINT16 &value, const void *data)
+{
+  const OffsetTo<Coverage> &coverage = (const OffsetTo<Coverage>&)value;
+  return (data+coverage).intersects (glyphs);
+}
+
+static inline bool array_is_subset_of (const hb_set_t *glyphs,
+				       unsigned int count,
+				       const HBUINT16 values[],
+				       intersects_func_t intersects_func,
+				       const void *intersects_data)
+{
+  for (const HBUINT16 &_ : + hb_iter (values, count))
+    if (!intersects_func (glyphs, _, intersects_data)) return false;
+  return true;
+}
+
+
+static inline void collect_glyph (hb_set_t *glyphs, const HBUINT16 &value, const void *data HB_UNUSED)
+{
+  glyphs->add (value);
+}
+static inline void collect_class (hb_set_t *glyphs, const HBUINT16 &value, const void *data)
+{
+  const ClassDef &class_def = *reinterpret_cast<const ClassDef *>(data);
+  class_def.collect_class (glyphs, value);
+}
+static inline void collect_coverage (hb_set_t *glyphs, const HBUINT16 &value, const void *data)
+{
+  const OffsetTo<Coverage> &coverage = (const OffsetTo<Coverage>&)value;
+  (data+coverage).collect_coverage (glyphs);
+}
+static inline void collect_array (hb_collect_glyphs_context_t *c HB_UNUSED,
+				  hb_set_t *glyphs,
+				  unsigned int count,
+				  const HBUINT16 values[],
+				  collect_glyphs_func_t collect_func,
+				  const void *collect_data)
+{
+  return
+  + hb_iter (values, count)
+  | hb_apply ([&] (const HBUINT16 &_) { collect_func (glyphs, _, collect_data); })
+  ;
+}
+
+
+static inline bool match_glyph (hb_codepoint_t glyph_id, const HBUINT16 &value, const void *data HB_UNUSED)
+{
+  return glyph_id == value;
+}
+static inline bool match_class (hb_codepoint_t glyph_id, const HBUINT16 &value, const void *data)
+{
+  const ClassDef &class_def = *reinterpret_cast<const ClassDef *>(data);
+  return class_def.get_class (glyph_id) == value;
+}
+static inline bool match_coverage (hb_codepoint_t glyph_id, const HBUINT16 &value, const void *data)
+{
+  const OffsetTo<Coverage> &coverage = (const OffsetTo<Coverage>&)value;
+  return (data+coverage).get_coverage (glyph_id) != NOT_COVERED;
+}
+
+static inline bool would_match_input (hb_would_apply_context_t *c,
+				      unsigned int count, /* Including the first glyph (not matched) */
+				      const HBUINT16 input[], /* Array of input values--start with second glyph */
+				      match_func_t match_func,
+				      const void *match_data)
+{
+  if (count != c->len)
+    return false;
+
+  for (unsigned int i = 1; i < count; i++)
+    if (likely (!match_func (c->glyphs[i], input[i - 1], match_data)))
+      return false;
+
+  return true;
+}
+static inline bool match_input (hb_ot_apply_context_t *c,
+				unsigned int count, /* Including the first glyph (not matched) */
+				const HBUINT16 input[], /* Array of input values--start with second glyph */
+				match_func_t match_func,
+				const void *match_data,
+				unsigned int *end_offset,
+				unsigned int match_positions[HB_MAX_CONTEXT_LENGTH],
+				unsigned int *p_total_component_count = nullptr)
+{
+  TRACE_APPLY (nullptr);
+
+  if (unlikely (count > HB_MAX_CONTEXT_LENGTH)) return_trace (false);
+
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_input;
+  skippy_iter.reset (buffer->idx, count - 1);
+  skippy_iter.set_match_func (match_func, match_data, input);
+
+  /*
+   * This is perhaps the trickiest part of OpenType...  Remarks:
+   *
+   * - If all components of the ligature were marks, we call this a mark ligature.
+   *
+   * - If there is no GDEF, and the ligature is NOT a mark ligature, we categorize
+   *   it as a ligature glyph.
+   *
+   * - Ligatures cannot be formed across glyphs attached to different components
+   *   of previous ligatures.  Eg. the sequence is LAM,SHADDA,LAM,FATHA,HEH, and
+   *   LAM,LAM,HEH form a ligature, leaving SHADDA,FATHA next to eachother.
+   *   However, it would be wrong to ligate that SHADDA,FATHA sequence.
+   *   There are a couple of exceptions to this:
+   *
+   *   o If a ligature tries ligating with marks that belong to it itself, go ahead,
+   *     assuming that the font designer knows what they are doing (otherwise it can
+   *     break Indic stuff when a matra wants to ligate with a conjunct,
+   *
+   *   o If two marks want to ligate and they belong to different components of the
+   *     same ligature glyph, and said ligature glyph is to be ignored according to
+   *     mark-filtering rules, then allow.
+   *     https://github.com/harfbuzz/harfbuzz/issues/545
+   */
+
+  unsigned int total_component_count = 0;
+  total_component_count += _hb_glyph_info_get_lig_num_comps (&buffer->cur());
+
+  unsigned int first_lig_id = _hb_glyph_info_get_lig_id (&buffer->cur());
+  unsigned int first_lig_comp = _hb_glyph_info_get_lig_comp (&buffer->cur());
+
+  enum {
+    LIGBASE_NOT_CHECKED,
+    LIGBASE_MAY_NOT_SKIP,
+    LIGBASE_MAY_SKIP
+  } ligbase = LIGBASE_NOT_CHECKED;
+
+  match_positions[0] = buffer->idx;
+  for (unsigned int i = 1; i < count; i++)
+  {
+    if (!skippy_iter.next ()) return_trace (false);
+
+    match_positions[i] = skippy_iter.idx;
+
+    unsigned int this_lig_id = _hb_glyph_info_get_lig_id (&buffer->info[skippy_iter.idx]);
+    unsigned int this_lig_comp = _hb_glyph_info_get_lig_comp (&buffer->info[skippy_iter.idx]);
+
+    if (first_lig_id && first_lig_comp)
+    {
+      /* If first component was attached to a previous ligature component,
+       * all subsequent components should be attached to the same ligature
+       * component, otherwise we shouldn't ligate them... */
+      if (first_lig_id != this_lig_id || first_lig_comp != this_lig_comp)
+      {
+	/* ...unless, we are attached to a base ligature and that base
+	 * ligature is ignorable. */
+	if (ligbase == LIGBASE_NOT_CHECKED)
+	{
+	  bool found = false;
+	  const auto *out = buffer->out_info;
+	  unsigned int j = buffer->out_len;
+	  while (j && _hb_glyph_info_get_lig_id (&out[j - 1]) == first_lig_id)
+	  {
+	    if (_hb_glyph_info_get_lig_comp (&out[j - 1]) == 0)
+	    {
+	      j--;
+	      found = true;
+	      break;
+	    }
+	    j--;
+	  }
+
+	  if (found && skippy_iter.may_skip (out[j]) == hb_ot_apply_context_t::matcher_t::SKIP_YES)
+	    ligbase = LIGBASE_MAY_SKIP;
+	  else
+	    ligbase = LIGBASE_MAY_NOT_SKIP;
+	}
+
+	if (ligbase == LIGBASE_MAY_NOT_SKIP)
+	  return_trace (false);
+      }
+    }
+    else
+    {
+      /* If first component was NOT attached to a previous ligature component,
+       * all subsequent components should also NOT be attached to any ligature
+       * component, unless they are attached to the first component itself! */
+      if (this_lig_id && this_lig_comp && (this_lig_id != first_lig_id))
+	return_trace (false);
+    }
+
+    total_component_count += _hb_glyph_info_get_lig_num_comps (&buffer->info[skippy_iter.idx]);
+  }
+
+  *end_offset = skippy_iter.idx - buffer->idx + 1;
+
+  if (p_total_component_count)
+    *p_total_component_count = total_component_count;
+
+  return_trace (true);
+}
+static inline bool ligate_input (hb_ot_apply_context_t *c,
+				 unsigned int count, /* Including the first glyph */
+				 const unsigned int match_positions[HB_MAX_CONTEXT_LENGTH], /* Including the first glyph */
+				 unsigned int match_length,
+				 hb_codepoint_t lig_glyph,
+				 unsigned int total_component_count)
+{
+  TRACE_APPLY (nullptr);
+
+  hb_buffer_t *buffer = c->buffer;
+
+  buffer->merge_clusters (buffer->idx, buffer->idx + match_length);
+
+  /* - If a base and one or more marks ligate, consider that as a base, NOT
+   *   ligature, such that all following marks can still attach to it.
+   *   https://github.com/harfbuzz/harfbuzz/issues/1109
+   *
+   * - If all components of the ligature were marks, we call this a mark ligature.
+   *   If it *is* a mark ligature, we don't allocate a new ligature id, and leave
+   *   the ligature to keep its old ligature id.  This will allow it to attach to
+   *   a base ligature in GPOS.  Eg. if the sequence is: LAM,LAM,SHADDA,FATHA,HEH,
+   *   and LAM,LAM,HEH for a ligature, they will leave SHADDA and FATHA with a
+   *   ligature id and component value of 2.  Then if SHADDA,FATHA form a ligature
+   *   later, we don't want them to lose their ligature id/component, otherwise
+   *   GPOS will fail to correctly position the mark ligature on top of the
+   *   LAM,LAM,HEH ligature.  See:
+   *     https://bugzilla.gnome.org/show_bug.cgi?id=676343
+   *
+   * - If a ligature is formed of components that some of which are also ligatures
+   *   themselves, and those ligature components had marks attached to *their*
+   *   components, we have to attach the marks to the new ligature component
+   *   positions!  Now *that*'s tricky!  And these marks may be following the
+   *   last component of the whole sequence, so we should loop forward looking
+   *   for them and update them.
+   *
+   *   Eg. the sequence is LAM,LAM,SHADDA,FATHA,HEH, and the font first forms a
+   *   'calt' ligature of LAM,HEH, leaving the SHADDA and FATHA with a ligature
+   *   id and component == 1.  Now, during 'liga', the LAM and the LAM-HEH ligature
+   *   form a LAM-LAM-HEH ligature.  We need to reassign the SHADDA and FATHA to
+   *   the new ligature with a component value of 2.
+   *
+   *   This in fact happened to a font...  See:
+   *   https://bugzilla.gnome.org/show_bug.cgi?id=437633
+   */
+
+  bool is_base_ligature = _hb_glyph_info_is_base_glyph (&buffer->info[match_positions[0]]);
+  bool is_mark_ligature = _hb_glyph_info_is_mark (&buffer->info[match_positions[0]]);
+  for (unsigned int i = 1; i < count; i++)
+    if (!_hb_glyph_info_is_mark (&buffer->info[match_positions[i]]))
+    {
+      is_base_ligature = false;
+      is_mark_ligature = false;
+      break;
+    }
+  bool is_ligature = !is_base_ligature && !is_mark_ligature;
+
+  unsigned int klass = is_ligature ? HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE : 0;
+  unsigned int lig_id = is_ligature ? _hb_allocate_lig_id (buffer) : 0;
+  unsigned int last_lig_id = _hb_glyph_info_get_lig_id (&buffer->cur());
+  unsigned int last_num_components = _hb_glyph_info_get_lig_num_comps (&buffer->cur());
+  unsigned int components_so_far = last_num_components;
+
+  if (is_ligature)
+  {
+    _hb_glyph_info_set_lig_props_for_ligature (&buffer->cur(), lig_id, total_component_count);
+    if (_hb_glyph_info_get_general_category (&buffer->cur()) == HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)
+    {
+      _hb_glyph_info_set_general_category (&buffer->cur(), HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER);
+    }
+  }
+  c->replace_glyph_with_ligature (lig_glyph, klass);
+
+  for (unsigned int i = 1; i < count; i++)
+  {
+    while (buffer->idx < match_positions[i] && buffer->successful)
+    {
+      if (is_ligature)
+      {
+	unsigned int this_comp = _hb_glyph_info_get_lig_comp (&buffer->cur());
+	if (this_comp == 0)
+	  this_comp = last_num_components;
+	unsigned int new_lig_comp = components_so_far - last_num_components +
+				    hb_min (this_comp, last_num_components);
+	  _hb_glyph_info_set_lig_props_for_mark (&buffer->cur(), lig_id, new_lig_comp);
+      }
+      buffer->next_glyph ();
+    }
+
+    last_lig_id = _hb_glyph_info_get_lig_id (&buffer->cur());
+    last_num_components = _hb_glyph_info_get_lig_num_comps (&buffer->cur());
+    components_so_far += last_num_components;
+
+    /* Skip the base glyph */
+    buffer->idx++;
+  }
+
+  if (!is_mark_ligature && last_lig_id)
+  {
+    /* Re-adjust components for any marks following. */
+    for (unsigned i = buffer->idx; i < buffer->len; ++i)
+    {
+      if (last_lig_id != _hb_glyph_info_get_lig_id (&buffer->info[i])) break;
+
+      unsigned this_comp = _hb_glyph_info_get_lig_comp (&buffer->info[i]);
+      if (!this_comp) break;
+
+      unsigned new_lig_comp = components_so_far - last_num_components +
+			      hb_min (this_comp, last_num_components);
+      _hb_glyph_info_set_lig_props_for_mark (&buffer->info[i], lig_id, new_lig_comp);
+    }
+  }
+  return_trace (true);
+}
+
+static inline bool match_backtrack (hb_ot_apply_context_t *c,
+				    unsigned int count,
+				    const HBUINT16 backtrack[],
+				    match_func_t match_func,
+				    const void *match_data,
+				    unsigned int *match_start)
+{
+  TRACE_APPLY (nullptr);
+
+  hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_context;
+  skippy_iter.reset (c->buffer->backtrack_len (), count);
+  skippy_iter.set_match_func (match_func, match_data, backtrack);
+
+  for (unsigned int i = 0; i < count; i++)
+    if (!skippy_iter.prev ())
+      return_trace (false);
+
+  *match_start = skippy_iter.idx;
+
+  return_trace (true);
+}
+
+static inline bool match_lookahead (hb_ot_apply_context_t *c,
+				    unsigned int count,
+				    const HBUINT16 lookahead[],
+				    match_func_t match_func,
+				    const void *match_data,
+				    unsigned int offset,
+				    unsigned int *end_index)
+{
+  TRACE_APPLY (nullptr);
+
+  hb_ot_apply_context_t::skipping_iterator_t &skippy_iter = c->iter_context;
+  skippy_iter.reset (c->buffer->idx + offset - 1, count);
+  skippy_iter.set_match_func (match_func, match_data, lookahead);
+
+  for (unsigned int i = 0; i < count; i++)
+    if (!skippy_iter.next ())
+      return_trace (false);
+
+  *end_index = skippy_iter.idx + 1;
+
+  return_trace (true);
+}
+
+
+
+struct LookupRecord
+{
+  LookupRecord* copy (hb_serialize_context_t *c,
+		      const hb_map_t         *lookup_map) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->embed (*this);
+    if (unlikely (!out)) return_trace (nullptr);
+
+    out->lookupListIndex = hb_map_get (lookup_map, lookupListIndex);
+    return_trace (out);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT16	sequenceIndex;		/* Index into current glyph
+					 * sequence--first glyph = 0 */
+  HBUINT16	lookupListIndex;	/* Lookup to apply to that
+					 * position--zero--based */
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+template <typename context_t>
+static inline void recurse_lookups (context_t *c,
+				    unsigned int lookupCount,
+				    const LookupRecord lookupRecord[] /* Array of LookupRecords--in design order */)
+{
+  for (unsigned int i = 0; i < lookupCount; i++)
+    c->recurse (lookupRecord[i].lookupListIndex);
+}
+
+static inline bool apply_lookup (hb_ot_apply_context_t *c,
+				 unsigned int count, /* Including the first glyph */
+				 unsigned int match_positions[HB_MAX_CONTEXT_LENGTH], /* Including the first glyph */
+				 unsigned int lookupCount,
+				 const LookupRecord lookupRecord[], /* Array of LookupRecords--in design order */
+				 unsigned int match_length)
+{
+  TRACE_APPLY (nullptr);
+
+  hb_buffer_t *buffer = c->buffer;
+  int end;
+
+  /* All positions are distance from beginning of *output* buffer.
+   * Adjust. */
+  {
+    unsigned int bl = buffer->backtrack_len ();
+    end = bl + match_length;
+
+    int delta = bl - buffer->idx;
+    /* Convert positions to new indexing. */
+    for (unsigned int j = 0; j < count; j++)
+      match_positions[j] += delta;
+  }
+
+  for (unsigned int i = 0; i < lookupCount && buffer->successful; i++)
+  {
+    unsigned int idx = lookupRecord[i].sequenceIndex;
+    if (idx >= count)
+      continue;
+
+    /* Don't recurse to ourself at same position.
+     * Note that this test is too naive, it doesn't catch longer loops. */
+    if (idx == 0 && lookupRecord[i].lookupListIndex == c->lookup_index)
+      continue;
+
+    if (unlikely (!buffer->move_to (match_positions[idx])))
+      break;
+
+    if (unlikely (buffer->max_ops <= 0))
+      break;
+
+    unsigned int orig_len = buffer->backtrack_len () + buffer->lookahead_len ();
+    if (!c->recurse (lookupRecord[i].lookupListIndex))
+      continue;
+
+    unsigned int new_len = buffer->backtrack_len () + buffer->lookahead_len ();
+    int delta = new_len - orig_len;
+
+    if (!delta)
+      continue;
+
+    /* Recursed lookup changed buffer len.  Adjust.
+     *
+     * TODO:
+     *
+     * Right now, if buffer length increased by n, we assume n new glyphs
+     * were added right after the current position, and if buffer length
+     * was decreased by n, we assume n match positions after the current
+     * one where removed.  The former (buffer length increased) case is
+     * fine, but the decrease case can be improved in at least two ways,
+     * both of which are significant:
+     *
+     *   - If recursed-to lookup is MultipleSubst and buffer length
+     *     decreased, then it's current match position that was deleted,
+     *     NOT the one after it.
+     *
+     *   - If buffer length was decreased by n, it does not necessarily
+     *     mean that n match positions where removed, as there might
+     *     have been marks and default-ignorables in the sequence.  We
+     *     should instead drop match positions between current-position
+     *     and current-position + n instead.
+     *
+     * It should be possible to construct tests for both of these cases.
+     */
+
+    end += delta;
+    if (end <= int (match_positions[idx]))
+    {
+      /* End might end up being smaller than match_positions[idx] if the recursed
+       * lookup ended up removing many items, more than we have had matched.
+       * Just never rewind end back and get out of here.
+       * https://bugs.chromium.org/p/chromium/issues/detail?id=659496 */
+      end = match_positions[idx];
+      /* There can't be any further changes. */
+      break;
+    }
+
+    unsigned int next = idx + 1; /* next now is the position after the recursed lookup. */
+
+    if (delta > 0)
+    {
+      if (unlikely (delta + count > HB_MAX_CONTEXT_LENGTH))
+	break;
+    }
+    else
+    {
+      /* NOTE: delta is negative. */
+      delta = hb_max (delta, (int) next - (int) count);
+      next -= delta;
+    }
+
+    /* Shift! */
+    memmove (match_positions + next + delta, match_positions + next,
+	     (count - next) * sizeof (match_positions[0]));
+    next += delta;
+    count += delta;
+
+    /* Fill in new entries. */
+    for (unsigned int j = idx + 1; j < next; j++)
+      match_positions[j] = match_positions[j - 1] + 1;
+
+    /* And fixup the rest. */
+    for (; next < count; next++)
+      match_positions[next] += delta;
+  }
+
+  buffer->move_to (end);
+
+  return_trace (true);
+}
+
+
+
+/* Contextual lookups */
+
+struct ContextClosureLookupContext
+{
+  ContextClosureFuncs funcs;
+  const void *intersects_data;
+};
+
+struct ContextCollectGlyphsLookupContext
+{
+  ContextCollectGlyphsFuncs funcs;
+  const void *collect_data;
+};
+
+struct ContextApplyLookupContext
+{
+  ContextApplyFuncs funcs;
+  const void *match_data;
+};
+
+static inline bool context_intersects (const hb_set_t *glyphs,
+				       unsigned int inputCount, /* Including the first glyph (not matched) */
+				       const HBUINT16 input[], /* Array of input values--start with second glyph */
+				       ContextClosureLookupContext &lookup_context)
+{
+  return array_is_subset_of (glyphs,
+			     inputCount ? inputCount - 1 : 0, input,
+			     lookup_context.funcs.intersects, lookup_context.intersects_data);
+}
+
+static inline void context_closure_lookup (hb_closure_context_t *c,
+					   unsigned int inputCount, /* Including the first glyph (not matched) */
+					   const HBUINT16 input[], /* Array of input values--start with second glyph */
+					   unsigned int lookupCount,
+					   const LookupRecord lookupRecord[],
+					   ContextClosureLookupContext &lookup_context)
+{
+  if (context_intersects (c->glyphs,
+			  inputCount, input,
+			  lookup_context))
+    recurse_lookups (c,
+		     lookupCount, lookupRecord);
+}
+
+static inline void context_collect_glyphs_lookup (hb_collect_glyphs_context_t *c,
+						  unsigned int inputCount, /* Including the first glyph (not matched) */
+						  const HBUINT16 input[], /* Array of input values--start with second glyph */
+						  unsigned int lookupCount,
+						  const LookupRecord lookupRecord[],
+						  ContextCollectGlyphsLookupContext &lookup_context)
+{
+  collect_array (c, c->input,
+		 inputCount ? inputCount - 1 : 0, input,
+		 lookup_context.funcs.collect, lookup_context.collect_data);
+  recurse_lookups (c,
+		   lookupCount, lookupRecord);
+}
+
+static inline bool context_would_apply_lookup (hb_would_apply_context_t *c,
+					       unsigned int inputCount, /* Including the first glyph (not matched) */
+					       const HBUINT16 input[], /* Array of input values--start with second glyph */
+					       unsigned int lookupCount HB_UNUSED,
+					       const LookupRecord lookupRecord[] HB_UNUSED,
+					       ContextApplyLookupContext &lookup_context)
+{
+  return would_match_input (c,
+			    inputCount, input,
+			    lookup_context.funcs.match, lookup_context.match_data);
+}
+static inline bool context_apply_lookup (hb_ot_apply_context_t *c,
+					 unsigned int inputCount, /* Including the first glyph (not matched) */
+					 const HBUINT16 input[], /* Array of input values--start with second glyph */
+					 unsigned int lookupCount,
+					 const LookupRecord lookupRecord[],
+					 ContextApplyLookupContext &lookup_context)
+{
+  unsigned int match_length = 0;
+  unsigned int match_positions[HB_MAX_CONTEXT_LENGTH];
+  return match_input (c,
+		      inputCount, input,
+		      lookup_context.funcs.match, lookup_context.match_data,
+		      &match_length, match_positions)
+      && (c->buffer->unsafe_to_break (c->buffer->idx, c->buffer->idx + match_length),
+	  apply_lookup (c,
+		       inputCount, match_positions,
+		       lookupCount, lookupRecord,
+		       match_length));
+}
+
+struct Rule
+{
+  bool intersects (const hb_set_t *glyphs, ContextClosureLookupContext &lookup_context) const
+  {
+    return context_intersects (glyphs,
+			       inputCount, inputZ.arrayZ,
+			       lookup_context);
+  }
+
+  void closure (hb_closure_context_t *c, ContextClosureLookupContext &lookup_context) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array ((inputCount ? inputCount - 1 : 0)));
+    context_closure_lookup (c,
+			    inputCount, inputZ.arrayZ,
+			    lookupCount, lookupRecord.arrayZ,
+			    lookup_context);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array (inputCount ? inputCount - 1 : 0));
+    recurse_lookups (c, lookupCount, lookupRecord.arrayZ);
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c,
+		       ContextCollectGlyphsLookupContext &lookup_context) const
+  {
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array (inputCount ? inputCount - 1 : 0));
+    context_collect_glyphs_lookup (c,
+				   inputCount, inputZ.arrayZ,
+				   lookupCount, lookupRecord.arrayZ,
+				   lookup_context);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c,
+		    ContextApplyLookupContext &lookup_context) const
+  {
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array (inputCount ? inputCount - 1 : 0));
+    return context_would_apply_lookup (c,
+				       inputCount, inputZ.arrayZ,
+				       lookupCount, lookupRecord.arrayZ,
+				       lookup_context);
+  }
+
+  bool apply (hb_ot_apply_context_t *c,
+	      ContextApplyLookupContext &lookup_context) const
+  {
+    TRACE_APPLY (this);
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array (inputCount ? inputCount - 1 : 0));
+    return_trace (context_apply_lookup (c, inputCount, inputZ.arrayZ, lookupCount, lookupRecord.arrayZ, lookup_context));
+  }
+
+  bool serialize (hb_serialize_context_t *c,
+		  const hb_map_t *input_mapping, /* old->new glyphid or class mapping */
+		  const hb_map_t *lookup_map) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->start_embed (this);
+    if (unlikely (!c->extend_min (out))) return_trace (false);
+
+    out->inputCount = inputCount;
+    out->lookupCount = lookupCount;
+
+    const hb_array_t<const HBUINT16> input = inputZ.as_array (inputCount - 1);
+    for (const auto org : input)
+    {
+      HBUINT16 d;
+      d = input_mapping->get (org);
+      c->copy (d);
+    }
+
+    const UnsizedArrayOf<LookupRecord> &lookupRecord = StructAfter<UnsizedArrayOf<LookupRecord>>
+						       (inputZ.as_array ((inputCount ? inputCount - 1 : 0)));
+    for (unsigned i = 0; i < (unsigned) lookupCount; i++)
+      c->copy (lookupRecord[i], lookup_map);
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       const hb_map_t *lookup_map,
+	       const hb_map_t *klass_map = nullptr) const
+  {
+    TRACE_SUBSET (this);
+
+    const hb_array_t<const HBUINT16> input = inputZ.as_array ((inputCount ? inputCount - 1 : 0));
+    if (!input.length) return_trace (false);
+
+    const hb_map_t *mapping = klass_map == nullptr ? c->plan->glyph_map : klass_map;
+    if (!hb_all (input, mapping)) return_trace (false);
+    return_trace (serialize (c->serializer, mapping, lookup_map));
+  }
+
+  public:
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (inputCount.sanitize (c) &&
+		  lookupCount.sanitize (c) &&
+		  c->check_range (inputZ.arrayZ,
+				  inputZ.item_size * (inputCount ? inputCount - 1 : 0) +
+				  LookupRecord::static_size * lookupCount));
+  }
+
+  protected:
+  HBUINT16	inputCount;		/* Total number of glyphs in input
+					 * glyph sequence--includes the first
+					 * glyph */
+  HBUINT16	lookupCount;		/* Number of LookupRecords */
+  UnsizedArrayOf<HBUINT16>
+		inputZ;			/* Array of match inputs--start with
+					 * second glyph */
+/*UnsizedArrayOf<LookupRecord>
+		lookupRecordX;*/	/* Array of LookupRecords--in
+					 * design order */
+  public:
+  DEFINE_SIZE_ARRAY (4, inputZ);
+};
+
+struct RuleSet
+{
+  bool intersects (const hb_set_t *glyphs,
+		   ContextClosureLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const Rule &_) { return _.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c,
+		ContextClosureLookupContext &lookup_context) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const Rule &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const Rule &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c,
+		       ContextCollectGlyphsLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const Rule &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c,
+		    ContextApplyLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const Rule &_) { return _.would_apply (c, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  bool apply (hb_ot_apply_context_t *c,
+	      ContextApplyLookupContext &lookup_context) const
+  {
+    TRACE_APPLY (this);
+    return_trace (
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const Rule &_) { return _.apply (c, lookup_context); })
+    | hb_any
+    )
+    ;
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       const hb_map_t *lookup_map,
+	       const hb_map_t *klass_map = nullptr) const
+  {
+    TRACE_SUBSET (this);
+
+    auto snap = c->serializer->snapshot ();
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    for (const OffsetTo<Rule>& _ : rule)
+    {
+      if (!_) continue;
+      auto *o = out->rule.serialize_append (c->serializer);
+      if (unlikely (!o)) continue;
+
+      auto o_snap = c->serializer->snapshot ();
+      if (!o->serialize_subset (c, _, this, lookup_map, klass_map))
+      {
+	out->rule.pop ();
+	c->serializer->revert (o_snap);
+      }
+    }
+
+    bool ret = bool (out->rule);
+    if (!ret) c->serializer->revert (snap);
+
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (rule.sanitize (c, this));
+  }
+
+  protected:
+  OffsetArrayOf<Rule>
+		rule;			/* Array of Rule tables
+					 * ordered by preference */
+  public:
+  DEFINE_SIZE_ARRAY (2, rule);
+};
+
+
+struct ContextFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_glyph},
+      nullptr
+    };
+
+    return
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (*glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const RuleSet &_) { return _.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_glyph},
+      nullptr
+    };
+
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    (this+coverage).collect_coverage (c->input);
+
+    struct ContextCollectGlyphsLookupContext lookup_context = {
+      {collect_glyph},
+      nullptr
+    };
+
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const RuleSet &rule_set = this+ruleSet[(this+coverage).get_coverage (c->glyphs[0])];
+    struct ContextApplyLookupContext lookup_context = {
+      {match_glyph},
+      nullptr
+    };
+    return rule_set.would_apply (c, lookup_context);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED))
+      return_trace (false);
+
+    const RuleSet &rule_set = this+ruleSet[index];
+    struct ContextApplyLookupContext lookup_context = {
+      {match_glyph},
+      nullptr
+    };
+    return_trace (rule_set.apply (c, lookup_context));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->ruleSet, this, lookup_map), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && ruleSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of table */
+  OffsetArrayOf<RuleSet>
+		ruleSet;		/* Array of RuleSet tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, ruleSet);
+};
+
+
+struct ContextFormat2
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    if (!(this+coverage).intersects (glyphs))
+      return false;
+
+    const ClassDef &class_def = this+classDef;
+
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_class},
+      &class_def
+    };
+
+    return
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_enumerate
+    | hb_map ([&] (const hb_pair_t<unsigned, const RuleSet &> p)
+	      { return class_def.intersects_class (glyphs, p.first) &&
+		       p.second.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    if (!(this+coverage).intersects (c->glyphs))
+      return;
+
+    const ClassDef &class_def = this+classDef;
+
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_class},
+      &class_def
+    };
+
+    return
+    + hb_enumerate (ruleSet)
+    | hb_filter ([&] (unsigned _)
+		 { return class_def.intersects_class (c->glyphs, _); },
+		 hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    (this+coverage).collect_coverage (c->input);
+
+    const ClassDef &class_def = this+classDef;
+    struct ContextCollectGlyphsLookupContext lookup_context = {
+      {collect_class},
+      &class_def
+    };
+
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const RuleSet &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const ClassDef &class_def = this+classDef;
+    unsigned int index = class_def.get_class (c->glyphs[0]);
+    const RuleSet &rule_set = this+ruleSet[index];
+    struct ContextApplyLookupContext lookup_context = {
+      {match_class},
+      &class_def
+    };
+    return rule_set.would_apply (c, lookup_context);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const ClassDef &class_def = this+classDef;
+    index = class_def.get_class (c->buffer->cur().codepoint);
+    const RuleSet &rule_set = this+ruleSet[index];
+    struct ContextApplyLookupContext lookup_context = {
+      {match_class},
+      &class_def
+    };
+    return_trace (rule_set.apply (c, lookup_context));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+    if (unlikely (!out->coverage.serialize_subset (c, coverage, this)))
+      return_trace (false);
+
+    hb_map_t klass_map;
+    out->classDef.serialize_subset (c, classDef, this, &klass_map);
+
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    bool ret = true;
+    unsigned non_zero_index = 0, index = 0;
+    for (const hb_pair_t<unsigned, const OffsetTo<RuleSet>&> _ : + hb_enumerate (ruleSet)
+								 | hb_filter (klass_map, hb_first))
+    {
+      auto *o = out->ruleSet.serialize_append (c->serializer);
+      if (unlikely (!o))
+      {
+	ret = false;
+	break;
+      }
+
+      if (o->serialize_subset (c, _.second, this, lookup_map, &klass_map))
+	non_zero_index = index;
+
+      index++;
+    }
+
+    if (!ret) return_trace (ret);
+
+    //prune empty trailing ruleSets
+    --index;
+    while (index > non_zero_index)
+    {
+      out->ruleSet.pop ();
+      index--;
+    }
+
+    return_trace (bool (out->ruleSet));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && classDef.sanitize (c, this) && ruleSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of table */
+  OffsetTo<ClassDef>
+		classDef;		/* Offset to glyph ClassDef table--from
+					 * beginning of table */
+  OffsetArrayOf<RuleSet>
+		ruleSet;		/* Array of RuleSet tables
+					 * ordered by class */
+  public:
+  DEFINE_SIZE_ARRAY (8, ruleSet);
+};
+
+
+struct ContextFormat3
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    if (!(this+coverageZ[0]).intersects (glyphs))
+      return false;
+
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_coverage},
+      this
+    };
+    return context_intersects (glyphs,
+			       glyphCount, (const HBUINT16 *) (coverageZ.arrayZ + 1),
+			       lookup_context);
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    if (!(this+coverageZ[0]).intersects (c->glyphs))
+      return;
+
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    struct ContextClosureLookupContext lookup_context = {
+      {intersects_coverage},
+      this
+    };
+    context_closure_lookup (c,
+			    glyphCount, (const HBUINT16 *) (coverageZ.arrayZ + 1),
+			    lookupCount, lookupRecord,
+			    lookup_context);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    recurse_lookups (c, lookupCount, lookupRecord);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    (this+coverageZ[0]).collect_coverage (c->input);
+
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    struct ContextCollectGlyphsLookupContext lookup_context = {
+      {collect_coverage},
+      this
+    };
+
+    context_collect_glyphs_lookup (c,
+				   glyphCount, (const HBUINT16 *) (coverageZ.arrayZ + 1),
+				   lookupCount, lookupRecord,
+				   lookup_context);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    struct ContextApplyLookupContext lookup_context = {
+      {match_coverage},
+      this
+    };
+    return context_would_apply_lookup (c,
+				       glyphCount, (const HBUINT16 *) (coverageZ.arrayZ + 1),
+				       lookupCount, lookupRecord,
+				       lookup_context);
+  }
+
+  const Coverage &get_coverage () const { return this+coverageZ[0]; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverageZ[0]).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    struct ContextApplyLookupContext lookup_context = {
+      {match_coverage},
+      this
+    };
+    return_trace (context_apply_lookup (c, glyphCount, (const HBUINT16 *) (coverageZ.arrayZ + 1), lookupCount, lookupRecord, lookup_context));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    out->format = format;
+    out->glyphCount = glyphCount;
+    out->lookupCount = lookupCount;
+
+    auto coverages = coverageZ.as_array (glyphCount);
+
+    for (const OffsetTo<Coverage>& offset : coverages)
+    {
+      auto *o = c->serializer->allocate_size<OffsetTo<Coverage>> (OffsetTo<Coverage>::static_size);
+      if (unlikely (!o)) return_trace (false);
+      if (!o->serialize_subset (c, offset, this)) return_trace (false);
+    }
+
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    for (unsigned i = 0; i < (unsigned) lookupCount; i++)
+      c->serializer->copy (lookupRecord[i], lookup_map);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!c->check_struct (this)) return_trace (false);
+    unsigned int count = glyphCount;
+    if (!count) return_trace (false); /* We want to access coverageZ[0] freely. */
+    if (!c->check_array (coverageZ.arrayZ, count)) return_trace (false);
+    for (unsigned int i = 0; i < count; i++)
+      if (!coverageZ[i].sanitize (c, this)) return_trace (false);
+    const LookupRecord *lookupRecord = &StructAfter<LookupRecord> (coverageZ.as_array (glyphCount));
+    return_trace (c->check_array (lookupRecord, lookupCount));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 3 */
+  HBUINT16	glyphCount;		/* Number of glyphs in the input glyph
+					 * sequence */
+  HBUINT16	lookupCount;		/* Number of LookupRecords */
+  UnsizedArrayOf<OffsetTo<Coverage>>
+		coverageZ;		/* Array of offsets to Coverage
+					 * table in glyph sequence order */
+/*UnsizedArrayOf<LookupRecord>
+		lookupRecordX;*/	/* Array of LookupRecords--in
+					 * design order */
+  public:
+  DEFINE_SIZE_ARRAY (6, coverageZ);
+};
+
+struct Context
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    case 3: return_trace (c->dispatch (u.format3, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  ContextFormat1	format1;
+  ContextFormat2	format2;
+  ContextFormat3	format3;
+  } u;
+};
+
+
+/* Chaining Contextual lookups */
+
+struct ChainContextClosureLookupContext
+{
+  ContextClosureFuncs funcs;
+  const void *intersects_data[3];
+};
+
+struct ChainContextCollectGlyphsLookupContext
+{
+  ContextCollectGlyphsFuncs funcs;
+  const void *collect_data[3];
+};
+
+struct ChainContextApplyLookupContext
+{
+  ContextApplyFuncs funcs;
+  const void *match_data[3];
+};
+
+static inline bool chain_context_intersects (const hb_set_t *glyphs,
+					     unsigned int backtrackCount,
+					     const HBUINT16 backtrack[],
+					     unsigned int inputCount, /* Including the first glyph (not matched) */
+					     const HBUINT16 input[], /* Array of input values--start with second glyph */
+					     unsigned int lookaheadCount,
+					     const HBUINT16 lookahead[],
+					     ChainContextClosureLookupContext &lookup_context)
+{
+  return array_is_subset_of (glyphs,
+			     backtrackCount, backtrack,
+			     lookup_context.funcs.intersects, lookup_context.intersects_data[0])
+      && array_is_subset_of (glyphs,
+			     inputCount ? inputCount - 1 : 0, input,
+			     lookup_context.funcs.intersects, lookup_context.intersects_data[1])
+      && array_is_subset_of (glyphs,
+			     lookaheadCount, lookahead,
+			     lookup_context.funcs.intersects, lookup_context.intersects_data[2]);
+}
+
+static inline void chain_context_closure_lookup (hb_closure_context_t *c,
+						 unsigned int backtrackCount,
+						 const HBUINT16 backtrack[],
+						 unsigned int inputCount, /* Including the first glyph (not matched) */
+						 const HBUINT16 input[], /* Array of input values--start with second glyph */
+						 unsigned int lookaheadCount,
+						 const HBUINT16 lookahead[],
+						 unsigned int lookupCount,
+						 const LookupRecord lookupRecord[],
+						 ChainContextClosureLookupContext &lookup_context)
+{
+  if (chain_context_intersects (c->glyphs,
+				backtrackCount, backtrack,
+				inputCount, input,
+				lookaheadCount, lookahead,
+				lookup_context))
+    recurse_lookups (c,
+		     lookupCount, lookupRecord);
+}
+
+static inline void chain_context_collect_glyphs_lookup (hb_collect_glyphs_context_t *c,
+							unsigned int backtrackCount,
+							const HBUINT16 backtrack[],
+							unsigned int inputCount, /* Including the first glyph (not matched) */
+							const HBUINT16 input[], /* Array of input values--start with second glyph */
+							unsigned int lookaheadCount,
+							const HBUINT16 lookahead[],
+							unsigned int lookupCount,
+							const LookupRecord lookupRecord[],
+							ChainContextCollectGlyphsLookupContext &lookup_context)
+{
+  collect_array (c, c->before,
+		 backtrackCount, backtrack,
+		 lookup_context.funcs.collect, lookup_context.collect_data[0]);
+  collect_array (c, c->input,
+		 inputCount ? inputCount - 1 : 0, input,
+		 lookup_context.funcs.collect, lookup_context.collect_data[1]);
+  collect_array (c, c->after,
+		 lookaheadCount, lookahead,
+		 lookup_context.funcs.collect, lookup_context.collect_data[2]);
+  recurse_lookups (c,
+		   lookupCount, lookupRecord);
+}
+
+static inline bool chain_context_would_apply_lookup (hb_would_apply_context_t *c,
+						     unsigned int backtrackCount,
+						     const HBUINT16 backtrack[] HB_UNUSED,
+						     unsigned int inputCount, /* Including the first glyph (not matched) */
+						     const HBUINT16 input[], /* Array of input values--start with second glyph */
+						     unsigned int lookaheadCount,
+						     const HBUINT16 lookahead[] HB_UNUSED,
+						     unsigned int lookupCount HB_UNUSED,
+						     const LookupRecord lookupRecord[] HB_UNUSED,
+						     ChainContextApplyLookupContext &lookup_context)
+{
+  return (c->zero_context ? !backtrackCount && !lookaheadCount : true)
+      && would_match_input (c,
+			    inputCount, input,
+			    lookup_context.funcs.match, lookup_context.match_data[1]);
+}
+
+static inline bool chain_context_apply_lookup (hb_ot_apply_context_t *c,
+					       unsigned int backtrackCount,
+					       const HBUINT16 backtrack[],
+					       unsigned int inputCount, /* Including the first glyph (not matched) */
+					       const HBUINT16 input[], /* Array of input values--start with second glyph */
+					       unsigned int lookaheadCount,
+					       const HBUINT16 lookahead[],
+					       unsigned int lookupCount,
+					       const LookupRecord lookupRecord[],
+					       ChainContextApplyLookupContext &lookup_context)
+{
+  unsigned int start_index = 0, match_length = 0, end_index = 0;
+  unsigned int match_positions[HB_MAX_CONTEXT_LENGTH];
+  return match_input (c,
+		      inputCount, input,
+		      lookup_context.funcs.match, lookup_context.match_data[1],
+		      &match_length, match_positions)
+      && match_backtrack (c,
+			  backtrackCount, backtrack,
+			  lookup_context.funcs.match, lookup_context.match_data[0],
+			  &start_index)
+      && match_lookahead (c,
+			  lookaheadCount, lookahead,
+			  lookup_context.funcs.match, lookup_context.match_data[2],
+			  match_length, &end_index)
+      && (c->buffer->unsafe_to_break_from_outbuffer (start_index, end_index),
+	  apply_lookup (c,
+			inputCount, match_positions,
+			lookupCount, lookupRecord,
+			match_length));
+}
+
+struct ChainRule
+{
+  bool intersects (const hb_set_t *glyphs, ChainContextClosureLookupContext &lookup_context) const
+  {
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    return chain_context_intersects (glyphs,
+				     backtrack.len, backtrack.arrayZ,
+				     input.lenP1, input.arrayZ,
+				     lookahead.len, lookahead.arrayZ,
+				     lookup_context);
+  }
+
+  void closure (hb_closure_context_t *c,
+		ChainContextClosureLookupContext &lookup_context) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    chain_context_closure_lookup (c,
+				  backtrack.len, backtrack.arrayZ,
+				  input.lenP1, input.arrayZ,
+				  lookahead.len, lookahead.arrayZ,
+				  lookup.len, lookup.arrayZ,
+				  lookup_context);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    recurse_lookups (c, lookup.len, lookup.arrayZ);
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c,
+		       ChainContextCollectGlyphsLookupContext &lookup_context) const
+  {
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    chain_context_collect_glyphs_lookup (c,
+					 backtrack.len, backtrack.arrayZ,
+					 input.lenP1, input.arrayZ,
+					 lookahead.len, lookahead.arrayZ,
+					 lookup.len, lookup.arrayZ,
+					 lookup_context);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c,
+		    ChainContextApplyLookupContext &lookup_context) const
+  {
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    return chain_context_would_apply_lookup (c,
+					     backtrack.len, backtrack.arrayZ,
+					     input.lenP1, input.arrayZ,
+					     lookahead.len, lookahead.arrayZ, lookup.len,
+					     lookup.arrayZ, lookup_context);
+  }
+
+  bool apply (hb_ot_apply_context_t *c, ChainContextApplyLookupContext &lookup_context) const
+  {
+    TRACE_APPLY (this);
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    return_trace (chain_context_apply_lookup (c,
+					      backtrack.len, backtrack.arrayZ,
+					      input.lenP1, input.arrayZ,
+					      lookahead.len, lookahead.arrayZ, lookup.len,
+					      lookup.arrayZ, lookup_context));
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  void serialize_array (hb_serialize_context_t *c,
+			HBUINT16 len,
+			Iterator it) const
+  {
+    c->copy (len);
+    for (const auto g : it)
+    {
+      HBUINT16 gid;
+      gid = g;
+      c->copy (gid);
+    }
+  }
+
+  ChainRule* copy (hb_serialize_context_t *c,
+		   const hb_map_t *lookup_map,
+		   const hb_map_t *backtrack_map,
+		   const hb_map_t *input_map = nullptr,
+		   const hb_map_t *lookahead_map = nullptr) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->start_embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+
+    const hb_map_t *mapping = backtrack_map;
+    serialize_array (c, backtrack.len, + backtrack.iter ()
+				       | hb_map (mapping));
+
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    if (input_map) mapping = input_map;
+    serialize_array (c, input.lenP1, + input.iter ()
+				     | hb_map (mapping));
+
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    if (lookahead_map) mapping = lookahead_map;
+    serialize_array (c, lookahead.len, + lookahead.iter ()
+				       | hb_map (mapping));
+
+    const ArrayOf<LookupRecord> &lookupRecord = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    HBUINT16 lookupCount;
+    lookupCount = lookupRecord.len;
+    if (!c->copy (lookupCount)) return_trace (nullptr);
+
+    for (unsigned i = 0; i < (unsigned) lookupCount; i++)
+      if (!c->copy (lookupRecord[i], lookup_map)) return_trace (nullptr);
+
+    return_trace (out);
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       const hb_map_t *lookup_map,
+	       const hb_map_t *backtrack_map = nullptr,
+	       const hb_map_t *input_map = nullptr,
+	       const hb_map_t *lookahead_map = nullptr) const
+  {
+    TRACE_SUBSET (this);
+
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+
+    if (!backtrack_map)
+    {
+      const hb_set_t &glyphset = *c->plan->glyphset ();
+      if (!hb_all (backtrack, glyphset) ||
+	  !hb_all (input, glyphset) ||
+	  !hb_all (lookahead, glyphset))
+	return_trace (false);
+
+      copy (c->serializer, lookup_map, c->plan->glyph_map);
+    }
+    else
+    {
+      if (!hb_all (backtrack, backtrack_map) ||
+	  !hb_all (input, input_map) ||
+	  !hb_all (lookahead, lookahead_map))
+	return_trace (false);
+
+      copy (c->serializer, lookup_map, backtrack_map, input_map, lookahead_map);
+    }
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!backtrack.sanitize (c)) return_trace (false);
+    const HeadlessArrayOf<HBUINT16> &input = StructAfter<HeadlessArrayOf<HBUINT16>> (backtrack);
+    if (!input.sanitize (c)) return_trace (false);
+    const ArrayOf<HBUINT16> &lookahead = StructAfter<ArrayOf<HBUINT16>> (input);
+    if (!lookahead.sanitize (c)) return_trace (false);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    return_trace (lookup.sanitize (c));
+  }
+
+  protected:
+  ArrayOf<HBUINT16>
+		backtrack;		/* Array of backtracking values
+					 * (to be matched before the input
+					 * sequence) */
+  HeadlessArrayOf<HBUINT16>
+		inputX;			/* Array of input values (start with
+					 * second glyph) */
+  ArrayOf<HBUINT16>
+		lookaheadX;		/* Array of lookahead values's (to be
+					 * matched after the input sequence) */
+  ArrayOf<LookupRecord>
+		lookupX;		/* Array of LookupRecords--in
+					 * design order) */
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+struct ChainRuleSet
+{
+  bool intersects (const hb_set_t *glyphs, ChainContextClosureLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const ChainRule &_) { return _.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+  void closure (hb_closure_context_t *c, ChainContextClosureLookupContext &lookup_context) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRule &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    if (unlikely (c->lookup_limit_exceeded ())) return;
+
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRule &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c, ChainContextCollectGlyphsLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRule &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c, ChainContextApplyLookupContext &lookup_context) const
+  {
+    return
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const ChainRule &_) { return _.would_apply (c, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  bool apply (hb_ot_apply_context_t *c, ChainContextApplyLookupContext &lookup_context) const
+  {
+    TRACE_APPLY (this);
+    return_trace (
+    + hb_iter (rule)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const ChainRule &_) { return _.apply (c, lookup_context); })
+    | hb_any
+    )
+    ;
+  }
+
+  bool subset (hb_subset_context_t *c,
+	       const hb_map_t *lookup_map,
+	       const hb_map_t *backtrack_klass_map = nullptr,
+	       const hb_map_t *input_klass_map = nullptr,
+	       const hb_map_t *lookahead_klass_map = nullptr) const
+  {
+    TRACE_SUBSET (this);
+
+    auto snap = c->serializer->snapshot ();
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+
+    for (const OffsetTo<ChainRule>& _ : rule)
+    {
+      if (!_) continue;
+      auto *o = out->rule.serialize_append (c->serializer);
+      if (unlikely (!o)) continue;
+
+      auto o_snap = c->serializer->snapshot ();
+      if (!o->serialize_subset (c, _, this,
+				lookup_map,
+				backtrack_klass_map,
+				input_klass_map,
+				lookahead_klass_map))
+      {
+	out->rule.pop ();
+	c->serializer->revert (o_snap);
+      }
+    }
+
+    bool ret = bool (out->rule);
+    if (!ret) c->serializer->revert (snap);
+
+    return_trace (ret);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (rule.sanitize (c, this));
+  }
+
+  protected:
+  OffsetArrayOf<ChainRule>
+		rule;			/* Array of ChainRule tables
+					 * ordered by preference */
+  public:
+  DEFINE_SIZE_ARRAY (2, rule);
+};
+
+struct ChainContextFormat1
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_glyph},
+      {nullptr, nullptr, nullptr}
+    };
+
+    return
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (*glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_map ([&] (const ChainRuleSet &_) { return _.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_glyph},
+      {nullptr, nullptr, nullptr}
+    };
+
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (*c->glyphs, hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    (this+coverage).collect_coverage (c->input);
+
+    struct ChainContextCollectGlyphsLookupContext lookup_context = {
+      {collect_glyph},
+      {nullptr, nullptr, nullptr}
+    };
+
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const ChainRuleSet &rule_set = this+ruleSet[(this+coverage).get_coverage (c->glyphs[0])];
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_glyph},
+      {nullptr, nullptr, nullptr}
+    };
+    return rule_set.would_apply (c, lookup_context);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const ChainRuleSet &rule_set = this+ruleSet[index];
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_glyph},
+      {nullptr, nullptr, nullptr}
+    };
+    return_trace (rule_set.apply (c, lookup_context));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    const hb_set_t &glyphset = *c->plan->glyphset ();
+    const hb_map_t &glyph_map = *c->plan->glyph_map;
+
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    hb_sorted_vector_t<hb_codepoint_t> new_coverage;
+    + hb_zip (this+coverage, ruleSet)
+    | hb_filter (glyphset, hb_first)
+    | hb_filter (subset_offset_array (c, out->ruleSet, this, lookup_map), hb_second)
+    | hb_map (hb_first)
+    | hb_map (glyph_map)
+    | hb_sink (new_coverage)
+    ;
+
+    out->coverage.serialize (c->serializer, out)
+		 .serialize (c->serializer, new_coverage.iter ());
+    return_trace (bool (new_coverage));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) && ruleSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 1 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of table */
+  OffsetArrayOf<ChainRuleSet>
+		ruleSet;		/* Array of ChainRuleSet tables
+					 * ordered by Coverage Index */
+  public:
+  DEFINE_SIZE_ARRAY (6, ruleSet);
+};
+
+struct ChainContextFormat2
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    if (!(this+coverage).intersects (glyphs))
+      return false;
+
+    const ClassDef &backtrack_class_def = this+backtrackClassDef;
+    const ClassDef &input_class_def = this+inputClassDef;
+    const ClassDef &lookahead_class_def = this+lookaheadClassDef;
+
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_class},
+      {&backtrack_class_def,
+       &input_class_def,
+       &lookahead_class_def}
+    };
+
+    return
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_enumerate
+    | hb_map ([&] (const hb_pair_t<unsigned, const ChainRuleSet &> p)
+	      { return input_class_def.intersects_class (glyphs, p.first) &&
+		       p.second.intersects (glyphs, lookup_context); })
+    | hb_any
+    ;
+  }
+  void closure (hb_closure_context_t *c) const
+  {
+    if (!(this+coverage).intersects (c->glyphs))
+      return;
+
+    const ClassDef &backtrack_class_def = this+backtrackClassDef;
+    const ClassDef &input_class_def = this+inputClassDef;
+    const ClassDef &lookahead_class_def = this+lookaheadClassDef;
+
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_class},
+      {&backtrack_class_def,
+       &input_class_def,
+       &lookahead_class_def}
+    };
+
+    return
+    + hb_enumerate (ruleSet)
+    | hb_filter ([&] (unsigned _)
+		 { return input_class_def.intersects_class (c->glyphs, _); },
+		 hb_first)
+    | hb_map (hb_second)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.closure (c, lookup_context); })
+    ;
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.closure_lookups (c); })
+    ;
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    (this+coverage).collect_coverage (c->input);
+
+    const ClassDef &backtrack_class_def = this+backtrackClassDef;
+    const ClassDef &input_class_def = this+inputClassDef;
+    const ClassDef &lookahead_class_def = this+lookaheadClassDef;
+
+    struct ChainContextCollectGlyphsLookupContext lookup_context = {
+      {collect_class},
+      {&backtrack_class_def,
+       &input_class_def,
+       &lookahead_class_def}
+    };
+
+    + hb_iter (ruleSet)
+    | hb_map (hb_add (this))
+    | hb_apply ([&] (const ChainRuleSet &_) { _.collect_glyphs (c, lookup_context); })
+    ;
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const ClassDef &backtrack_class_def = this+backtrackClassDef;
+    const ClassDef &input_class_def = this+inputClassDef;
+    const ClassDef &lookahead_class_def = this+lookaheadClassDef;
+
+    unsigned int index = input_class_def.get_class (c->glyphs[0]);
+    const ChainRuleSet &rule_set = this+ruleSet[index];
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_class},
+      {&backtrack_class_def,
+       &input_class_def,
+       &lookahead_class_def}
+    };
+    return rule_set.would_apply (c, lookup_context);
+  }
+
+  const Coverage &get_coverage () const { return this+coverage; }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    unsigned int index = (this+coverage).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const ClassDef &backtrack_class_def = this+backtrackClassDef;
+    const ClassDef &input_class_def = this+inputClassDef;
+    const ClassDef &lookahead_class_def = this+lookaheadClassDef;
+
+    index = input_class_def.get_class (c->buffer->cur().codepoint);
+    const ChainRuleSet &rule_set = this+ruleSet[index];
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_class},
+      {&backtrack_class_def,
+       &input_class_def,
+       &lookahead_class_def}
+    };
+    return_trace (rule_set.apply (c, lookup_context));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->serializer->start_embed (*this);
+    if (unlikely (!c->serializer->extend_min (out))) return_trace (false);
+    out->format = format;
+    out->coverage.serialize_subset (c, coverage, this);
+
+    hb_map_t backtrack_klass_map;
+    out->backtrackClassDef.serialize_subset (c, backtrackClassDef, this, &backtrack_klass_map);
+    if (unlikely (!c->serializer->check_success (!backtrack_klass_map.in_error ())))
+      return_trace (false);
+
+    // subset inputClassDef based on glyphs survived in Coverage subsetting
+    hb_map_t input_klass_map;
+    out->inputClassDef.serialize_subset (c, inputClassDef, this, &input_klass_map);
+    if (unlikely (!c->serializer->check_success (!input_klass_map.in_error ())))
+      return_trace (false);
+
+    hb_map_t lookahead_klass_map;
+    out->lookaheadClassDef.serialize_subset (c, lookaheadClassDef, this, &lookahead_klass_map);
+    if (unlikely (!c->serializer->check_success (!lookahead_klass_map.in_error ())))
+      return_trace (false);
+
+    unsigned non_zero_index = 0, index = 0;
+    bool ret = true;
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    for (const OffsetTo<ChainRuleSet>& _ : + hb_enumerate (ruleSet)
+					   | hb_filter (input_klass_map, hb_first)
+					   | hb_map (hb_second))
+    {
+      auto *o = out->ruleSet.serialize_append (c->serializer);
+      if (unlikely (!o))
+      {
+	ret = false;
+	break;
+      }
+      if (o->serialize_subset (c, _, this,
+			       lookup_map,
+			       &backtrack_klass_map,
+			       &input_klass_map,
+			       &lookahead_klass_map))
+	non_zero_index = index;
+
+      index++;
+    }
+
+    if (!ret) return_trace (ret);
+
+    //prune empty trailing ruleSets
+    --index;
+    while (index > non_zero_index)
+    {
+      out->ruleSet.pop ();
+      index--;
+    }
+
+    return_trace (bool (out->ruleSet));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (coverage.sanitize (c, this) &&
+		  backtrackClassDef.sanitize (c, this) &&
+		  inputClassDef.sanitize (c, this) &&
+		  lookaheadClassDef.sanitize (c, this) &&
+		  ruleSet.sanitize (c, this));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 2 */
+  OffsetTo<Coverage>
+		coverage;		/* Offset to Coverage table--from
+					 * beginning of table */
+  OffsetTo<ClassDef>
+		backtrackClassDef;	/* Offset to glyph ClassDef table
+					 * containing backtrack sequence
+					 * data--from beginning of table */
+  OffsetTo<ClassDef>
+		inputClassDef;		/* Offset to glyph ClassDef
+					 * table containing input sequence
+					 * data--from beginning of table */
+  OffsetTo<ClassDef>
+		lookaheadClassDef;	/* Offset to glyph ClassDef table
+					 * containing lookahead sequence
+					 * data--from beginning of table */
+  OffsetArrayOf<ChainRuleSet>
+		ruleSet;		/* Array of ChainRuleSet tables
+					 * ordered by class */
+  public:
+  DEFINE_SIZE_ARRAY (12, ruleSet);
+};
+
+struct ChainContextFormat3
+{
+  bool intersects (const hb_set_t *glyphs) const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+
+    if (!(this+input[0]).intersects (glyphs))
+      return false;
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_coverage},
+      {this, this, this}
+    };
+    return chain_context_intersects (glyphs,
+				     backtrack.len, (const HBUINT16 *) backtrack.arrayZ,
+				     input.len, (const HBUINT16 *) input.arrayZ + 1,
+				     lookahead.len, (const HBUINT16 *) lookahead.arrayZ,
+				     lookup_context);
+  }
+
+  void closure (hb_closure_context_t *c) const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+
+    if (!(this+input[0]).intersects (c->glyphs))
+      return;
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    struct ChainContextClosureLookupContext lookup_context = {
+      {intersects_coverage},
+      {this, this, this}
+    };
+    chain_context_closure_lookup (c,
+				  backtrack.len, (const HBUINT16 *) backtrack.arrayZ,
+				  input.len, (const HBUINT16 *) input.arrayZ + 1,
+				  lookahead.len, (const HBUINT16 *) lookahead.arrayZ,
+				  lookup.len, lookup.arrayZ,
+				  lookup_context);
+  }
+
+  void closure_lookups (hb_closure_lookups_context_t *c) const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    recurse_lookups (c, lookup.len, lookup.arrayZ);
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const {}
+
+  void collect_glyphs (hb_collect_glyphs_context_t *c) const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+
+    (this+input[0]).collect_coverage (c->input);
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    struct ChainContextCollectGlyphsLookupContext lookup_context = {
+      {collect_coverage},
+      {this, this, this}
+    };
+    chain_context_collect_glyphs_lookup (c,
+					 backtrack.len, (const HBUINT16 *) backtrack.arrayZ,
+					 input.len, (const HBUINT16 *) input.arrayZ + 1,
+					 lookahead.len, (const HBUINT16 *) lookahead.arrayZ,
+					 lookup.len, lookup.arrayZ,
+					 lookup_context);
+  }
+
+  bool would_apply (hb_would_apply_context_t *c) const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_coverage},
+      {this, this, this}
+    };
+    return chain_context_would_apply_lookup (c,
+					     backtrack.len, (const HBUINT16 *) backtrack.arrayZ,
+					     input.len, (const HBUINT16 *) input.arrayZ + 1,
+					     lookahead.len, (const HBUINT16 *) lookahead.arrayZ,
+					     lookup.len, lookup.arrayZ, lookup_context);
+  }
+
+  const Coverage &get_coverage () const
+  {
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    return this+input[0];
+  }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    TRACE_APPLY (this);
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+
+    unsigned int index = (this+input[0]).get_coverage (c->buffer->cur().codepoint);
+    if (likely (index == NOT_COVERED)) return_trace (false);
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    struct ChainContextApplyLookupContext lookup_context = {
+      {match_coverage},
+      {this, this, this}
+    };
+    return_trace (chain_context_apply_lookup (c,
+					      backtrack.len, (const HBUINT16 *) backtrack.arrayZ,
+					      input.len, (const HBUINT16 *) input.arrayZ + 1,
+					      lookahead.len, (const HBUINT16 *) lookahead.arrayZ,
+					      lookup.len, lookup.arrayZ, lookup_context));
+  }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator))>
+  bool serialize_coverage_offsets (hb_subset_context_t *c, Iterator it, const void* base) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->serializer->start_embed<OffsetArrayOf<Coverage>> ();
+
+    if (unlikely (!c->serializer->allocate_size<HBUINT16> (HBUINT16::static_size))) return_trace (false);
+
+    + it
+    | hb_apply (subset_offset_array (c, *out, base))
+    ;
+
+    return_trace (out->len);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    auto *out = c->serializer->start_embed (this);
+    if (unlikely (!out)) return_trace (false);
+    if (unlikely (!c->serializer->embed (this->format))) return_trace (false);
+
+    if (!serialize_coverage_offsets (c, backtrack.iter (), this))
+      return_trace (false);
+
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    if (!serialize_coverage_offsets (c, input.iter (), this))
+      return_trace (false);
+
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    if (!serialize_coverage_offsets (c, lookahead.iter (), this))
+      return_trace (false);
+
+    const ArrayOf<LookupRecord> &lookupRecord = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    HBUINT16 lookupCount;
+    lookupCount = lookupRecord.len;
+    if (!c->serializer->copy (lookupCount)) return_trace (false);
+
+    const hb_map_t *lookup_map = c->table_tag == HB_OT_TAG_GSUB ? c->plan->gsub_lookups : c->plan->gpos_lookups;
+    for (unsigned i = 0; i < (unsigned) lookupCount; i++)
+      if (!c->serializer->copy (lookupRecord[i], lookup_map)) return_trace (false);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (!backtrack.sanitize (c, this)) return_trace (false);
+    const OffsetArrayOf<Coverage> &input = StructAfter<OffsetArrayOf<Coverage>> (backtrack);
+    if (!input.sanitize (c, this)) return_trace (false);
+    if (!input.len) return_trace (false); /* To be consistent with Context. */
+    const OffsetArrayOf<Coverage> &lookahead = StructAfter<OffsetArrayOf<Coverage>> (input);
+    if (!lookahead.sanitize (c, this)) return_trace (false);
+    const ArrayOf<LookupRecord> &lookup = StructAfter<ArrayOf<LookupRecord>> (lookahead);
+    return_trace (lookup.sanitize (c));
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier--format = 3 */
+  OffsetArrayOf<Coverage>
+		backtrack;		/* Array of coverage tables
+					 * in backtracking sequence, in  glyph
+					 * sequence order */
+  OffsetArrayOf<Coverage>
+		inputX		;	/* Array of coverage
+					 * tables in input sequence, in glyph
+					 * sequence order */
+  OffsetArrayOf<Coverage>
+		lookaheadX;		/* Array of coverage tables
+					 * in lookahead sequence, in glyph
+					 * sequence order */
+  ArrayOf<LookupRecord>
+		lookupX;		/* Array of LookupRecords--in
+					 * design order) */
+  public:
+  DEFINE_SIZE_MIN (10);
+};
+
+struct ChainContext
+{
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (c->dispatch (u.format1, hb_forward<Ts> (ds)...));
+    case 2: return_trace (c->dispatch (u.format2, hb_forward<Ts> (ds)...));
+    case 3: return_trace (c->dispatch (u.format3, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;	/* Format identifier */
+  ChainContextFormat1	format1;
+  ChainContextFormat2	format2;
+  ChainContextFormat3	format3;
+  } u;
+};
+
+
+template <typename T>
+struct ExtensionFormat1
+{
+  unsigned int get_type () const { return extensionLookupType; }
+
+  template <typename X>
+  const X& get_subtable () const
+  { return this + reinterpret_cast<const LOffsetTo<typename T::SubTable> &> (extensionOffset); }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, format);
+    if (unlikely (!c->may_dispatch (this, this))) return_trace (c->no_dispatch_return_value ());
+    return_trace (get_subtable<typename T::SubTable> ().dispatch (c, get_type (), hb_forward<Ts> (ds)...));
+  }
+
+  void collect_variation_indices (hb_collect_variation_indices_context_t *c) const
+  { dispatch (c); }
+
+  /* This is called from may_dispatch() above with hb_sanitize_context_t. */
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  extensionLookupType != T::SubTable::Extension);
+  }
+
+  protected:
+  HBUINT16	format;			/* Format identifier. Set to 1. */
+  HBUINT16	extensionLookupType;	/* Lookup type of subtable referenced
+					 * by ExtensionOffset (i.e. the
+					 * extension subtable). */
+  Offset32	extensionOffset;	/* Offset to the extension subtable,
+					 * of lookup type subtable. */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+template <typename T>
+struct Extension
+{
+  unsigned int get_type () const
+  {
+    switch (u.format) {
+    case 1: return u.format1.get_type ();
+    default:return 0;
+    }
+  }
+  template <typename X>
+  const X& get_subtable () const
+  {
+    switch (u.format) {
+    case 1: return u.format1.template get_subtable<typename T::SubTable> ();
+    default:return Null (typename T::SubTable);
+    }
+  }
+
+  template <typename context_t, typename ...Ts>
+  typename context_t::return_t dispatch (context_t *c, Ts&&... ds) const
+  {
+    TRACE_DISPATCH (this, u.format);
+    if (unlikely (!c->may_dispatch (this, &u.format))) return_trace (c->no_dispatch_return_value ());
+    switch (u.format) {
+    case 1: return_trace (u.format1.dispatch (c, hb_forward<Ts> (ds)...));
+    default:return_trace (c->default_return_value ());
+    }
+  }
+
+  protected:
+  union {
+  HBUINT16		format;		/* Format identifier */
+  ExtensionFormat1<T>	format1;
+  } u;
+};
+
+
+/*
+ * GSUB/GPOS Common
+ */
+
+struct hb_ot_layout_lookup_accelerator_t
+{
+  template <typename TLookup>
+  void init (const TLookup &lookup)
+  {
+    digest.init ();
+    lookup.collect_coverage (&digest);
+
+    subtables.init ();
+    OT::hb_get_subtables_context_t c_get_subtables (subtables);
+    lookup.dispatch (&c_get_subtables);
+  }
+  void fini () { subtables.fini (); }
+
+  bool may_have (hb_codepoint_t g) const
+  { return digest.may_have (g); }
+
+  bool apply (hb_ot_apply_context_t *c) const
+  {
+    for (unsigned int i = 0; i < subtables.length; i++)
+      if (subtables[i].apply (c))
+	return true;
+    return false;
+  }
+
+  private:
+  hb_set_digest_t digest;
+  hb_get_subtables_context_t::array_t subtables;
+};
+
+struct GSUBGPOS
+{
+  bool has_data () const { return version.to_int (); }
+  unsigned int get_script_count () const
+  { return (this+scriptList).len; }
+  const Tag& get_script_tag (unsigned int i) const
+  { return (this+scriptList).get_tag (i); }
+  unsigned int get_script_tags (unsigned int start_offset,
+				unsigned int *script_count /* IN/OUT */,
+				hb_tag_t     *script_tags /* OUT */) const
+  { return (this+scriptList).get_tags (start_offset, script_count, script_tags); }
+  const Script& get_script (unsigned int i) const
+  { return (this+scriptList)[i]; }
+  bool find_script_index (hb_tag_t tag, unsigned int *index) const
+  { return (this+scriptList).find_index (tag, index); }
+
+  unsigned int get_feature_count () const
+  { return (this+featureList).len; }
+  hb_tag_t get_feature_tag (unsigned int i) const
+  { return i == Index::NOT_FOUND_INDEX ? HB_TAG_NONE : (this+featureList).get_tag (i); }
+  unsigned int get_feature_tags (unsigned int start_offset,
+				 unsigned int *feature_count /* IN/OUT */,
+				 hb_tag_t     *feature_tags /* OUT */) const
+  { return (this+featureList).get_tags (start_offset, feature_count, feature_tags); }
+  const Feature& get_feature (unsigned int i) const
+  { return (this+featureList)[i]; }
+  bool find_feature_index (hb_tag_t tag, unsigned int *index) const
+  { return (this+featureList).find_index (tag, index); }
+
+  unsigned int get_lookup_count () const
+  { return (this+lookupList).len; }
+  const Lookup& get_lookup (unsigned int i) const
+  { return (this+lookupList)[i]; }
+
+  bool find_variations_index (const int *coords, unsigned int num_coords,
+			      unsigned int *index) const
+  {
+#ifdef HB_NO_VAR
+    *index = FeatureVariations::NOT_FOUND_INDEX;
+    return false;
+#endif
+    return (version.to_int () >= 0x00010001u ? this+featureVars : Null (FeatureVariations))
+	    .find_index (coords, num_coords, index);
+  }
+  const Feature& get_feature_variation (unsigned int feature_index,
+					unsigned int variations_index) const
+  {
+#ifndef HB_NO_VAR
+    if (FeatureVariations::NOT_FOUND_INDEX != variations_index &&
+	version.to_int () >= 0x00010001u)
+    {
+      const Feature *feature = (this+featureVars).find_substitute (variations_index,
+								   feature_index);
+      if (feature)
+	return *feature;
+    }
+#endif
+    return get_feature (feature_index);
+  }
+
+  void feature_variation_collect_lookups (const hb_set_t *feature_indexes,
+					  hb_set_t       *lookup_indexes /* OUT */) const
+  {
+#ifndef HB_NO_VAR
+    if (version.to_int () >= 0x00010001u)
+      (this+featureVars).collect_lookups (feature_indexes, lookup_indexes);
+#endif
+  }
+
+  template <typename TLookup>
+  void closure_lookups (hb_face_t      *face,
+			const hb_set_t *glyphs,
+			hb_set_t       *lookup_indexes /* IN/OUT */) const
+  {
+    hb_set_t visited_lookups, inactive_lookups;
+    OT::hb_closure_lookups_context_t c (face, glyphs, &visited_lookups, &inactive_lookups);
+
+    for (unsigned lookup_index : + hb_iter (lookup_indexes))
+      reinterpret_cast<const TLookup &> (get_lookup (lookup_index)).closure_lookups (&c, lookup_index);
+
+    hb_set_union (lookup_indexes, &visited_lookups);
+    hb_set_subtract (lookup_indexes, &inactive_lookups);
+  }
+
+  template <typename TLookup>
+  bool subset (hb_subset_layout_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    auto *out = c->subset_context->serializer->embed (*this);
+    if (unlikely (!out)) return_trace (false);
+
+    typedef LookupOffsetList<TLookup> TLookupList;
+    reinterpret_cast<OffsetTo<TLookupList> &> (out->lookupList)
+	.serialize_subset (c->subset_context,
+			   reinterpret_cast<const OffsetTo<TLookupList> &> (lookupList),
+			   this,
+			   c);
+
+    reinterpret_cast<OffsetTo<RecordListOfFeature> &> (out->featureList)
+	.serialize_subset (c->subset_context,
+			   reinterpret_cast<const OffsetTo<RecordListOfFeature> &> (featureList),
+			   this,
+			   c);
+
+    out->scriptList.serialize_subset (c->subset_context,
+				      scriptList,
+				      this,
+				      c);
+
+#ifndef HB_NO_VAR
+    if (version.to_int () >= 0x00010001u)
+    {
+      bool ret = out->featureVars.serialize_subset (c->subset_context, featureVars, this, c);
+      if (!ret)
+      {
+	out->version.major = 1;
+	out->version.minor = 0;
+      }
+    }
+#endif
+
+    return_trace (true);
+  }
+
+  void closure_features (const hb_map_t *lookup_indexes, /* IN */
+			 hb_set_t       *feature_indexes /* OUT */) const
+  {
+    unsigned int feature_count = hb_min (get_feature_count (), (unsigned) HB_MAX_FEATURES);
+    for (unsigned i = 0; i < feature_count; i++)
+    {
+      const Feature& f = get_feature (i);
+      if ((!f.featureParams.is_null ()) || f.intersects_lookup_indexes (lookup_indexes))
+	feature_indexes->add (i);
+    }
+#ifndef HB_NO_VAR
+    if (version.to_int () >= 0x00010001u)
+      (this+featureVars).closure_features (lookup_indexes, feature_indexes);
+#endif
+  }
+
+  unsigned int get_size () const
+  {
+    return min_size +
+	   (version.to_int () >= 0x00010001u ? featureVars.static_size : 0);
+  }
+
+  template <typename TLookup>
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    typedef OffsetListOf<TLookup> TLookupList;
+    if (unlikely (!(version.sanitize (c) &&
+		    likely (version.major == 1) &&
+		    scriptList.sanitize (c, this) &&
+		    featureList.sanitize (c, this) &&
+		    reinterpret_cast<const OffsetTo<TLookupList> &> (lookupList).sanitize (c, this))))
+      return_trace (false);
+
+#ifndef HB_NO_VAR
+    if (unlikely (!(version.to_int () < 0x00010001u || featureVars.sanitize (c, this))))
+      return_trace (false);
+#endif
+
+    return_trace (true);
+  }
+
+  template <typename T>
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      this->table = hb_sanitize_context_t ().reference_table<T> (face);
+      if (unlikely (this->table->is_blocklisted (this->table.get_blob (), face)))
+      {
+	hb_blob_destroy (this->table.get_blob ());
+	this->table = hb_blob_get_empty ();
+      }
+
+      this->lookup_count = table->get_lookup_count ();
+
+      this->accels = (hb_ot_layout_lookup_accelerator_t *) calloc (this->lookup_count, sizeof (hb_ot_layout_lookup_accelerator_t));
+      if (unlikely (!this->accels))
+	this->lookup_count = 0;
+
+      for (unsigned int i = 0; i < this->lookup_count; i++)
+	this->accels[i].init (table->get_lookup (i));
+    }
+
+    void fini ()
+    {
+      for (unsigned int i = 0; i < this->lookup_count; i++)
+	this->accels[i].fini ();
+      free (this->accels);
+      this->table.destroy ();
+    }
+
+    hb_blob_ptr_t<T> table;
+    unsigned int lookup_count;
+    hb_ot_layout_lookup_accelerator_t *accels;
+  };
+
+  protected:
+  FixedVersion<>version;	/* Version of the GSUB/GPOS table--initially set
+				 * to 0x00010000u */
+  OffsetTo<ScriptList>
+		scriptList;	/* ScriptList table */
+  OffsetTo<FeatureList>
+		featureList;	/* FeatureList table */
+  OffsetTo<LookupList>
+		lookupList;	/* LookupList table */
+  LOffsetTo<FeatureVariations>
+		featureVars;	/* Offset to Feature Variations
+				   table--from beginning of table
+				 * (may be NULL).  Introduced
+				 * in version 0x00010001. */
+  public:
+  DEFINE_SIZE_MIN (10);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_GSUBGPOS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout-jstf-table.hh b/thirdparty/harfbuzz/src/hb-ot-layout-jstf-table.hh
new file mode 100644
index 0000000000..ffd2bf4574
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout-jstf-table.hh
@@ -0,0 +1,235 @@
+/*
+ * Copyright © 2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_JSTF_TABLE_HH
+#define HB_OT_LAYOUT_JSTF_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout-gpos-table.hh"
+
+
+namespace OT {
+
+
+/*
+ * JstfModList -- Justification Modification List Tables
+ */
+
+typedef IndexArray JstfModList;
+
+
+/*
+ * JstfMax -- Justification Maximum Table
+ */
+
+typedef OffsetListOf<PosLookup> JstfMax;
+
+
+/*
+ * JstfPriority -- Justification Priority Table
+ */
+
+struct JstfPriority
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  shrinkageEnableGSUB.sanitize (c, this) &&
+		  shrinkageDisableGSUB.sanitize (c, this) &&
+		  shrinkageEnableGPOS.sanitize (c, this) &&
+		  shrinkageDisableGPOS.sanitize (c, this) &&
+		  shrinkageJstfMax.sanitize (c, this) &&
+		  extensionEnableGSUB.sanitize (c, this) &&
+		  extensionDisableGSUB.sanitize (c, this) &&
+		  extensionEnableGPOS.sanitize (c, this) &&
+		  extensionDisableGPOS.sanitize (c, this) &&
+		  extensionJstfMax.sanitize (c, this));
+  }
+
+  protected:
+  OffsetTo<JstfModList>
+		shrinkageEnableGSUB;	/* Offset to Shrinkage Enable GSUB
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		shrinkageDisableGSUB;	/* Offset to Shrinkage Disable GSUB
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		shrinkageEnableGPOS;	/* Offset to Shrinkage Enable GPOS
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		shrinkageDisableGPOS;	/* Offset to Shrinkage Disable GPOS
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfMax>
+		shrinkageJstfMax;	/* Offset to Shrinkage JstfMax table--
+					 * from beginning of JstfPriority table
+					 * --may be NULL */
+  OffsetTo<JstfModList>
+		extensionEnableGSUB;	/* Offset to Extension Enable GSUB
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		extensionDisableGSUB;	/* Offset to Extension Disable GSUB
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		extensionEnableGPOS;	/* Offset to Extension Enable GPOS
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfModList>
+		extensionDisableGPOS;	/* Offset to Extension Disable GPOS
+					 * JstfModList table--from beginning of
+					 * JstfPriority table--may be NULL */
+  OffsetTo<JstfMax>
+		extensionJstfMax;	/* Offset to Extension JstfMax table--
+					 * from beginning of JstfPriority table
+					 * --may be NULL */
+
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+
+/*
+ * JstfLangSys -- Justification Language System Table
+ */
+
+struct JstfLangSys : OffsetListOf<JstfPriority>
+{
+  bool sanitize (hb_sanitize_context_t *c,
+		 const Record_sanitize_closure_t * = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (OffsetListOf<JstfPriority>::sanitize (c));
+  }
+};
+
+
+/*
+ * ExtenderGlyphs -- Extender Glyph Table
+ */
+
+typedef SortedArrayOf<HBGlyphID> ExtenderGlyphs;
+
+
+/*
+ * JstfScript -- The Justification Table
+ */
+
+struct JstfScript
+{
+  unsigned int get_lang_sys_count () const
+  { return langSys.len; }
+  const Tag& get_lang_sys_tag (unsigned int i) const
+  { return langSys.get_tag (i); }
+  unsigned int get_lang_sys_tags (unsigned int start_offset,
+				  unsigned int *lang_sys_count /* IN/OUT */,
+				  hb_tag_t     *lang_sys_tags /* OUT */) const
+  { return langSys.get_tags (start_offset, lang_sys_count, lang_sys_tags); }
+  const JstfLangSys& get_lang_sys (unsigned int i) const
+  {
+    if (i == Index::NOT_FOUND_INDEX) return get_default_lang_sys ();
+    return this+langSys[i].offset;
+  }
+  bool find_lang_sys_index (hb_tag_t tag, unsigned int *index) const
+  { return langSys.find_index (tag, index); }
+
+  bool has_default_lang_sys () const               { return defaultLangSys != 0; }
+  const JstfLangSys& get_default_lang_sys () const { return this+defaultLangSys; }
+
+  bool sanitize (hb_sanitize_context_t *c,
+		 const Record_sanitize_closure_t * = nullptr) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (extenderGlyphs.sanitize (c, this) &&
+		  defaultLangSys.sanitize (c, this) &&
+		  langSys.sanitize (c, this));
+  }
+
+  protected:
+  OffsetTo<ExtenderGlyphs>
+		extenderGlyphs;	/* Offset to ExtenderGlyph table--from beginning
+				 * of JstfScript table-may be NULL */
+  OffsetTo<JstfLangSys>
+		defaultLangSys;	/* Offset to DefaultJstfLangSys table--from
+				 * beginning of JstfScript table--may be Null */
+  RecordArrayOf<JstfLangSys>
+		langSys;	/* Array of JstfLangSysRecords--listed
+				 * alphabetically by LangSysTag */
+  public:
+  DEFINE_SIZE_ARRAY (6, langSys);
+};
+
+
+/*
+ * JSTF -- Justification
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/jstf
+ */
+
+struct JSTF
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_JSTF;
+
+  unsigned int get_script_count () const
+  { return scriptList.len; }
+  const Tag& get_script_tag (unsigned int i) const
+  { return scriptList.get_tag (i); }
+  unsigned int get_script_tags (unsigned int start_offset,
+				unsigned int *script_count /* IN/OUT */,
+				hb_tag_t     *script_tags /* OUT */) const
+  { return scriptList.get_tags (start_offset, script_count, script_tags); }
+  const JstfScript& get_script (unsigned int i) const
+  { return this+scriptList[i].offset; }
+  bool find_script_index (hb_tag_t tag, unsigned int *index) const
+  { return scriptList.find_index (tag, index); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  scriptList.sanitize (c, this));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the JSTF table--initially set
+				 * to 0x00010000u */
+  RecordArrayOf<JstfScript>
+		scriptList;	/* Array of JstfScripts--listed
+				 * alphabetically by ScriptTag */
+  public:
+  DEFINE_SIZE_ARRAY (6, scriptList);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_LAYOUT_JSTF_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout.cc b/thirdparty/harfbuzz/src/hb-ot-layout.cc
new file mode 100644
index 0000000000..46408bb9d3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout.cc
@@ -0,0 +1,1993 @@
+/*
+ * Copyright © 1998-2004  David Turner and Werner Lemberg
+ * Copyright © 2006  Behdad Esfahbod
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_LAYOUT
+
+#ifdef HB_NO_OT_TAG
+#error "Cannot compile hb-ot-layout.cc with HB_NO_OT_TAG."
+#endif
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout.hh"
+#include "hb-ot-face.hh"
+#include "hb-ot-map.hh"
+#include "hb-map.hh"
+
+#include "hb-ot-kern-table.hh"
+#include "hb-ot-layout-gdef-table.hh"
+#include "hb-ot-layout-gsub-table.hh"
+#include "hb-ot-layout-gpos-table.hh"
+#include "hb-ot-layout-base-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-ot-layout-jstf-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-ot-name-table.hh"
+#include "hb-ot-os2-table.hh"
+
+#include "hb-aat-layout-morx-table.hh"
+#include "hb-aat-layout-opbd-table.hh" // Just so we compile it; unused otherwise.
+
+/**
+ * SECTION:hb-ot-layout
+ * @title: hb-ot-layout
+ * @short_description: OpenType Layout
+ * @include: hb-ot.h
+ *
+ * Functions for querying OpenType Layout features in the font face.
+ **/
+
+
+/*
+ * kern
+ */
+
+#ifndef HB_NO_OT_KERN
+/**
+ * hb_ot_layout_has_kerning:
+ * @face: The #hb_face_t to work on
+ *
+ * Tests whether a face includes any kerning data in the 'kern' table.
+ * Does NOT test for kerning lookups in the GPOS table.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ **/
+bool
+hb_ot_layout_has_kerning (hb_face_t *face)
+{
+  return face->table.kern->has_data ();
+}
+
+/**
+ * hb_ot_layout_has_machine_kerning:
+ * @face: The #hb_face_t to work on
+ *
+ * Tests whether a face includes any state-machine kerning in the 'kern' table.
+ * Does NOT examine the GPOS table.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ **/
+bool
+hb_ot_layout_has_machine_kerning (hb_face_t *face)
+{
+  return face->table.kern->has_state_machine ();
+}
+
+/**
+ * hb_ot_layout_has_cross_kerning:
+ * @face: The #hb_face_t to work on
+ *
+ * Tests whether a face has any cross-stream kerning (i.e., kerns
+ * that make adjustments perpendicular to the direction of the text
+ * flow: Y adjustments in horizontal text or X adjustments in
+ * vertical text) in the 'kern' table.
+ *
+ * Does NOT examine the GPOS table.
+ *
+ * Return value: true is data found, false otherwise
+ *
+ **/
+bool
+hb_ot_layout_has_cross_kerning (hb_face_t *face)
+{
+  return face->table.kern->has_cross_stream ();
+}
+
+void
+hb_ot_layout_kern (const hb_ot_shape_plan_t *plan,
+		   hb_font_t *font,
+		   hb_buffer_t  *buffer)
+{
+  hb_blob_t *blob = font->face->table.kern.get_blob ();
+  const AAT::kern& kern = *blob->as<AAT::kern> ();
+
+  AAT::hb_aat_apply_context_t c (plan, font, buffer, blob);
+
+  kern.apply (&c);
+}
+#endif
+
+
+/*
+ * GDEF
+ */
+
+bool
+OT::GDEF::is_blocklisted (hb_blob_t *blob,
+			  hb_face_t *face) const
+{
+#ifdef HB_NO_OT_LAYOUT_BLACKLIST
+  return false;
+#endif
+  /* The ugly business of blocklisting individual fonts' tables happen here!
+   * See this thread for why we finally had to bend in and do this:
+   * https://lists.freedesktop.org/archives/harfbuzz/2016-February/005489.html
+   *
+   * In certain versions of Times New Roman Italic and Bold Italic,
+   * ASCII double quotation mark U+0022 has wrong glyph class 3 (mark)
+   * in GDEF.  Many versions of Tahoma have bad GDEF tables that
+   * incorrectly classify some spacing marks such as certain IPA
+   * symbols as glyph class 3. So do older versions of Microsoft
+   * Himalaya, and the version of Cantarell shipped by Ubuntu 16.04.
+   *
+   * Nuke the GDEF tables of to avoid unwanted width-zeroing.
+   *
+   * See https://bugzilla.mozilla.org/show_bug.cgi?id=1279925
+   *     https://bugzilla.mozilla.org/show_bug.cgi?id=1279693
+   *     https://bugzilla.mozilla.org/show_bug.cgi?id=1279875
+   */
+  switch HB_CODEPOINT_ENCODE3(blob->length,
+			      face->table.GSUB->table.get_length (),
+			      face->table.GPOS->table.get_length ())
+  {
+    /* sha1sum:c5ee92f0bca4bfb7d06c4d03e8cf9f9cf75d2e8a Windows 7? timesi.ttf */
+    case HB_CODEPOINT_ENCODE3 (442, 2874, 42038):
+    /* sha1sum:37fc8c16a0894ab7b749e35579856c73c840867b Windows 7? timesbi.ttf */
+    case HB_CODEPOINT_ENCODE3 (430, 2874, 40662):
+    /* sha1sum:19fc45110ea6cd3cdd0a5faca256a3797a069a80 Windows 7 timesi.ttf */
+    case HB_CODEPOINT_ENCODE3 (442, 2874, 39116):
+    /* sha1sum:6d2d3c9ed5b7de87bc84eae0df95ee5232ecde26 Windows 7 timesbi.ttf */
+    case HB_CODEPOINT_ENCODE3 (430, 2874, 39374):
+    /* sha1sum:8583225a8b49667c077b3525333f84af08c6bcd8 OS X 10.11.3 Times New Roman Italic.ttf */
+    case HB_CODEPOINT_ENCODE3 (490, 3046, 41638):
+    /* sha1sum:ec0f5a8751845355b7c3271d11f9918a966cb8c9 OS X 10.11.3 Times New Roman Bold Italic.ttf */
+    case HB_CODEPOINT_ENCODE3 (478, 3046, 41902):
+    /* sha1sum:96eda93f7d33e79962451c6c39a6b51ee893ce8c  tahoma.ttf from Windows 8 */
+    case HB_CODEPOINT_ENCODE3 (898, 12554, 46470):
+    /* sha1sum:20928dc06014e0cd120b6fc942d0c3b1a46ac2bc  tahomabd.ttf from Windows 8 */
+    case HB_CODEPOINT_ENCODE3 (910, 12566, 47732):
+    /* sha1sum:4f95b7e4878f60fa3a39ca269618dfde9721a79e  tahoma.ttf from Windows 8.1 */
+    case HB_CODEPOINT_ENCODE3 (928, 23298, 59332):
+    /* sha1sum:6d400781948517c3c0441ba42acb309584b73033  tahomabd.ttf from Windows 8.1 */
+    case HB_CODEPOINT_ENCODE3 (940, 23310, 60732):
+    /* tahoma.ttf v6.04 from Windows 8.1 x64, see https://bugzilla.mozilla.org/show_bug.cgi?id=1279925 */
+    case HB_CODEPOINT_ENCODE3 (964, 23836, 60072):
+    /* tahomabd.ttf v6.04 from Windows 8.1 x64, see https://bugzilla.mozilla.org/show_bug.cgi?id=1279925 */
+    case HB_CODEPOINT_ENCODE3 (976, 23832, 61456):
+    /* sha1sum:e55fa2dfe957a9f7ec26be516a0e30b0c925f846  tahoma.ttf from Windows 10 */
+    case HB_CODEPOINT_ENCODE3 (994, 24474, 60336):
+    /* sha1sum:7199385abb4c2cc81c83a151a7599b6368e92343  tahomabd.ttf from Windows 10 */
+    case HB_CODEPOINT_ENCODE3 (1006, 24470, 61740):
+    /* tahoma.ttf v6.91 from Windows 10 x64, see https://bugzilla.mozilla.org/show_bug.cgi?id=1279925 */
+    case HB_CODEPOINT_ENCODE3 (1006, 24576, 61346):
+    /* tahomabd.ttf v6.91 from Windows 10 x64, see https://bugzilla.mozilla.org/show_bug.cgi?id=1279925 */
+    case HB_CODEPOINT_ENCODE3 (1018, 24572, 62828):
+    /* sha1sum:b9c84d820c49850d3d27ec498be93955b82772b5  tahoma.ttf from Windows 10 AU */
+    case HB_CODEPOINT_ENCODE3 (1006, 24576, 61352):
+    /* sha1sum:2bdfaab28174bdadd2f3d4200a30a7ae31db79d2  tahomabd.ttf from Windows 10 AU */
+    case HB_CODEPOINT_ENCODE3 (1018, 24572, 62834):
+    /* sha1sum:b0d36cf5a2fbe746a3dd277bffc6756a820807a7  Tahoma.ttf from Mac OS X 10.9 */
+    case HB_CODEPOINT_ENCODE3 (832, 7324, 47162):
+    /* sha1sum:12fc4538e84d461771b30c18b5eb6bd434e30fba  Tahoma Bold.ttf from Mac OS X 10.9 */
+    case HB_CODEPOINT_ENCODE3 (844, 7302, 45474):
+    /* sha1sum:eb8afadd28e9cf963e886b23a30b44ab4fd83acc  himalaya.ttf from Windows 7 */
+    case HB_CODEPOINT_ENCODE3 (180, 13054, 7254):
+    /* sha1sum:73da7f025b238a3f737aa1fde22577a6370f77b0  himalaya.ttf from Windows 8 */
+    case HB_CODEPOINT_ENCODE3 (192, 12638, 7254):
+    /* sha1sum:6e80fd1c0b059bbee49272401583160dc1e6a427  himalaya.ttf from Windows 8.1 */
+    case HB_CODEPOINT_ENCODE3 (192, 12690, 7254):
+    /* 8d9267aea9cd2c852ecfb9f12a6e834bfaeafe44  cantarell-fonts-0.0.21/otf/Cantarell-Regular.otf */
+    /* 983988ff7b47439ab79aeaf9a45bd4a2c5b9d371  cantarell-fonts-0.0.21/otf/Cantarell-Oblique.otf */
+    case HB_CODEPOINT_ENCODE3 (188, 248, 3852):
+    /* 2c0c90c6f6087ffbfea76589c93113a9cbb0e75f  cantarell-fonts-0.0.21/otf/Cantarell-Bold.otf */
+    /* 55461f5b853c6da88069ffcdf7f4dd3f8d7e3e6b  cantarell-fonts-0.0.21/otf/Cantarell-Bold-Oblique.otf */
+    case HB_CODEPOINT_ENCODE3 (188, 264, 3426):
+    /* d125afa82a77a6475ac0e74e7c207914af84b37a padauk-2.80/Padauk.ttf RHEL 7.2 */
+    case HB_CODEPOINT_ENCODE3 (1058, 47032, 11818):
+    /* 0f7b80437227b90a577cc078c0216160ae61b031 padauk-2.80/Padauk-Bold.ttf RHEL 7.2*/
+    case HB_CODEPOINT_ENCODE3 (1046, 47030, 12600):
+    /* d3dde9aa0a6b7f8f6a89ef1002e9aaa11b882290 padauk-2.80/Padauk.ttf Ubuntu 16.04 */
+    case HB_CODEPOINT_ENCODE3 (1058, 71796, 16770):
+    /* 5f3c98ccccae8a953be2d122c1b3a77fd805093f padauk-2.80/Padauk-Bold.ttf Ubuntu 16.04 */
+    case HB_CODEPOINT_ENCODE3 (1046, 71790, 17862):
+    /* 6c93b63b64e8b2c93f5e824e78caca555dc887c7 padauk-2.80/Padauk-book.ttf */
+    case HB_CODEPOINT_ENCODE3 (1046, 71788, 17112):
+    /* d89b1664058359b8ec82e35d3531931125991fb9 padauk-2.80/Padauk-bookbold.ttf */
+    case HB_CODEPOINT_ENCODE3 (1058, 71794, 17514):
+    /* 824cfd193aaf6234b2b4dc0cf3c6ef576c0d00ef padauk-3.0/Padauk-book.ttf */
+    case HB_CODEPOINT_ENCODE3 (1330, 109904, 57938):
+    /* 91fcc10cf15e012d27571e075b3b4dfe31754a8a padauk-3.0/Padauk-bookbold.ttf */
+    case HB_CODEPOINT_ENCODE3 (1330, 109904, 58972):
+    /* sha1sum: c26e41d567ed821bed997e937bc0c41435689e85  Padauk.ttf
+     *  "Padauk Regular" "Version 2.5", see https://crbug.com/681813 */
+    case HB_CODEPOINT_ENCODE3 (1004, 59092, 14836):
+      return true;
+  }
+  return false;
+}
+
+static void
+_hb_ot_layout_set_glyph_props (hb_font_t *font,
+			       hb_buffer_t *buffer)
+{
+  _hb_buffer_assert_gsubgpos_vars (buffer);
+
+  const OT::GDEF &gdef = *font->face->table.GDEF->table;
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    _hb_glyph_info_set_glyph_props (&buffer->info[i], gdef.get_glyph_props (buffer->info[i].codepoint));
+    _hb_glyph_info_clear_lig_props (&buffer->info[i]);
+    buffer->info[i].syllable() = 0;
+  }
+}
+
+/* Public API */
+
+/**
+ * hb_ot_layout_has_glyph_classes:
+ * @face: #hb_face_t to work upon
+ *
+ * Tests whether a face has any glyph classes defined in its GDEF table.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_has_glyph_classes (hb_face_t *face)
+{
+  return face->table.GDEF->table->has_glyph_classes ();
+}
+
+/**
+ * hb_ot_layout_get_glyph_class:
+ * @face: The #hb_face_t to work on
+ * @glyph: The #hb_codepoint_t code point to query
+ *
+ * Fetches the GDEF class of the requested glyph in the specified face.
+ *
+ * Return value: The #hb_ot_layout_glyph_class_t glyph class of the given code
+ * point in the GDEF table of the face.
+ *
+ * Since: 0.9.7
+ **/
+hb_ot_layout_glyph_class_t
+hb_ot_layout_get_glyph_class (hb_face_t      *face,
+			      hb_codepoint_t  glyph)
+{
+  return (hb_ot_layout_glyph_class_t) face->table.GDEF->table->get_glyph_class (glyph);
+}
+
+/**
+ * hb_ot_layout_get_glyphs_in_class:
+ * @face: The #hb_face_t to work on
+ * @klass: The #hb_ot_layout_glyph_class_t GDEF class to retrieve
+ * @glyphs: (out): The #hb_set_t set of all glyphs belonging to the requested
+ *          class.
+ *
+ * Retrieves the set of all glyphs from the face that belong to the requested
+ * glyph class in the face's GDEF table.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_ot_layout_get_glyphs_in_class (hb_face_t                  *face,
+				  hb_ot_layout_glyph_class_t  klass,
+				  hb_set_t                   *glyphs /* OUT */)
+{
+  return face->table.GDEF->table->get_glyphs_in_class (klass, glyphs);
+}
+
+#ifndef HB_NO_LAYOUT_UNUSED
+/**
+ * hb_ot_layout_get_attach_points:
+ * @face: The #hb_face_t to work on
+ * @glyph: The #hb_codepoint_t code point to query
+ * @start_offset: offset of the first attachment point to retrieve
+ * @point_count: (inout) (allow-none): Input = the maximum number of attachment points to return;
+ *               Output = the actual number of attachment points returned (may be zero)
+ * @point_array: (out) (array length=point_count): The array of attachment points found for the query
+ *
+ * Fetches a list of all attachment points for the specified glyph in the GDEF
+ * table of the face. The list returned will begin at the offset provided.
+ *
+ * Useful if the client program wishes to cache the list.
+ *
+ **/
+unsigned int
+hb_ot_layout_get_attach_points (hb_face_t      *face,
+				hb_codepoint_t  glyph,
+				unsigned int    start_offset,
+				unsigned int   *point_count /* IN/OUT */,
+				unsigned int   *point_array /* OUT */)
+{
+  return face->table.GDEF->table->get_attach_points (glyph,
+						     start_offset,
+						     point_count,
+						     point_array);
+}
+/**
+ * hb_ot_layout_get_ligature_carets:
+ * @font: The #hb_font_t to work on
+ * @direction: The #hb_direction_t text direction to use
+ * @glyph: The #hb_codepoint_t code point to query
+ * @start_offset: offset of the first caret position to retrieve
+ * @caret_count: (inout) (allow-none): Input = the maximum number of caret positions to return;
+ *               Output = the actual number of caret positions returned (may be zero)
+ * @caret_array: (out) (array length=caret_count): The array of caret positions found for the query
+ *
+ * Fetches a list of the caret positions defined for a ligature glyph in the GDEF
+ * table of the font. The list returned will begin at the offset provided.
+ *
+ **/
+unsigned int
+hb_ot_layout_get_ligature_carets (hb_font_t      *font,
+				  hb_direction_t  direction,
+				  hb_codepoint_t  glyph,
+				  unsigned int    start_offset,
+				  unsigned int   *caret_count /* IN/OUT */,
+				  hb_position_t  *caret_array /* OUT */)
+{
+  return font->face->table.GDEF->table->get_lig_carets (font, direction, glyph, start_offset, caret_count, caret_array);
+}
+#endif
+
+
+/*
+ * GSUB/GPOS
+ */
+
+bool
+OT::GSUB::is_blocklisted (hb_blob_t *blob HB_UNUSED,
+			  hb_face_t *face) const
+{
+#ifdef HB_NO_OT_LAYOUT_BLACKLIST
+  return false;
+#endif
+  return false;
+}
+
+bool
+OT::GPOS::is_blocklisted (hb_blob_t *blob HB_UNUSED,
+			  hb_face_t *face HB_UNUSED) const
+{
+#ifdef HB_NO_OT_LAYOUT_BLACKLIST
+  return false;
+#endif
+  return false;
+}
+
+static const OT::GSUBGPOS&
+get_gsubgpos_table (hb_face_t *face,
+		    hb_tag_t   table_tag)
+{
+  switch (table_tag) {
+    case HB_OT_TAG_GSUB: return *face->table.GSUB->table;
+    case HB_OT_TAG_GPOS: return *face->table.GPOS->table;
+    default:             return Null (OT::GSUBGPOS);
+  }
+}
+
+
+/**
+ * hb_ot_layout_table_get_script_tags:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @start_offset: offset of the first script tag to retrieve
+ * @script_count: (inout) (allow-none): Input = the maximum number of script tags to return;
+ *                Output = the actual number of script tags returned (may be zero)
+ * @script_tags: (out) (array length=script_count): The array of #hb_tag_t script tags found for the query
+ *
+ * Fetches a list of all scripts enumerated in the specified face's GSUB table
+ * or GPOS table. The list returned will begin at the offset provided.
+ *
+ **/
+unsigned int
+hb_ot_layout_table_get_script_tags (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  start_offset,
+				    unsigned int *script_count /* IN/OUT */,
+				    hb_tag_t     *script_tags  /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  return g.get_script_tags (start_offset, script_count, script_tags);
+}
+
+#define HB_OT_TAG_LATIN_SCRIPT		HB_TAG ('l', 'a', 't', 'n')
+
+/**
+ * hb_ot_layout_table_find_script:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_tag: #hb_tag_t of the script tag requested
+ * @script_index: (out): The index of the requested script tag
+ *
+ * Fetches the index if a given script tag in the specified face's GSUB table
+ * or GPOS table.
+ *
+ * Return value: true if the script is found, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_table_find_script (hb_face_t    *face,
+				hb_tag_t      table_tag,
+				hb_tag_t      script_tag,
+				unsigned int *script_index /* OUT */)
+{
+  static_assert ((OT::Index::NOT_FOUND_INDEX == HB_OT_LAYOUT_NO_SCRIPT_INDEX), "");
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  if (g.find_script_index (script_tag, script_index))
+    return true;
+
+  /* try finding 'DFLT' */
+  if (g.find_script_index (HB_OT_TAG_DEFAULT_SCRIPT, script_index))
+    return false;
+
+  /* try with 'dflt'; MS site has had typos and many fonts use it now :(.
+   * including many versions of DejaVu Sans Mono! */
+  if (g.find_script_index (HB_OT_TAG_DEFAULT_LANGUAGE, script_index))
+    return false;
+
+  /* try with 'latn'; some old fonts put their features there even though
+     they're really trying to support Thai, for example :( */
+  if (g.find_script_index (HB_OT_TAG_LATIN_SCRIPT, script_index))
+    return false;
+
+  if (script_index) *script_index = HB_OT_LAYOUT_NO_SCRIPT_INDEX;
+  return false;
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_ot_layout_table_choose_script:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_tags: Array of #hb_tag_t script tags
+ * @script_index: (out): The index of the requested script tag
+ * @chosen_script: (out): #hb_tag_t of the script tag requested
+ *
+ * Deprecated since 2.0.0
+ **/
+hb_bool_t
+hb_ot_layout_table_choose_script (hb_face_t      *face,
+				  hb_tag_t        table_tag,
+				  const hb_tag_t *script_tags,
+				  unsigned int   *script_index  /* OUT */,
+				  hb_tag_t       *chosen_script /* OUT */)
+{
+  const hb_tag_t *t;
+  for (t = script_tags; *t; t++);
+  return hb_ot_layout_table_select_script (face, table_tag, t - script_tags, script_tags, script_index, chosen_script);
+}
+#endif
+
+/**
+ * hb_ot_layout_table_select_script:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_count: Number of script tags in the array
+ * @script_tags: Array of #hb_tag_t script tags
+ * @script_index: (out): The index of the requested script
+ * @chosen_script: (out): #hb_tag_t of the requested script
+ *
+ * Since: 2.0.0
+ **/
+hb_bool_t
+hb_ot_layout_table_select_script (hb_face_t      *face,
+				  hb_tag_t        table_tag,
+				  unsigned int    script_count,
+				  const hb_tag_t *script_tags,
+				  unsigned int   *script_index  /* OUT */,
+				  hb_tag_t       *chosen_script /* OUT */)
+{
+  static_assert ((OT::Index::NOT_FOUND_INDEX == HB_OT_LAYOUT_NO_SCRIPT_INDEX), "");
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  unsigned int i;
+
+  for (i = 0; i < script_count; i++)
+  {
+    if (g.find_script_index (script_tags[i], script_index))
+    {
+      if (chosen_script)
+	*chosen_script = script_tags[i];
+      return true;
+    }
+  }
+
+  /* try finding 'DFLT' */
+  if (g.find_script_index (HB_OT_TAG_DEFAULT_SCRIPT, script_index)) {
+    if (chosen_script)
+      *chosen_script = HB_OT_TAG_DEFAULT_SCRIPT;
+    return false;
+  }
+
+  /* try with 'dflt'; MS site has had typos and many fonts use it now :( */
+  if (g.find_script_index (HB_OT_TAG_DEFAULT_LANGUAGE, script_index)) {
+    if (chosen_script)
+      *chosen_script = HB_OT_TAG_DEFAULT_LANGUAGE;
+    return false;
+  }
+
+  /* try with 'latn'; some old fonts put their features there even though
+     they're really trying to support Thai, for example :( */
+  if (g.find_script_index (HB_OT_TAG_LATIN_SCRIPT, script_index)) {
+    if (chosen_script)
+      *chosen_script = HB_OT_TAG_LATIN_SCRIPT;
+    return false;
+  }
+
+  if (script_index) *script_index = HB_OT_LAYOUT_NO_SCRIPT_INDEX;
+  if (chosen_script)
+    *chosen_script = HB_OT_LAYOUT_NO_SCRIPT_INDEX;
+  return false;
+}
+
+
+/**
+ * hb_ot_layout_table_get_feature_tags:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @start_offset: offset of the first feature tag to retrieve
+ * @feature_count: (inout) (allow-none): Input = the maximum number of feature tags to return;
+ *                 Output = the actual number of feature tags returned (may be zero)
+ * @feature_tags: (out) (array length=feature_count): Array of feature tags found in the table
+ *
+ * Fetches a list of all feature tags in the given face's GSUB or GPOS table.
+ *
+ **/
+unsigned int
+hb_ot_layout_table_get_feature_tags (hb_face_t    *face,
+				     hb_tag_t      table_tag,
+				     unsigned int  start_offset,
+				     unsigned int *feature_count /* IN/OUT */,
+				     hb_tag_t     *feature_tags  /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  return g.get_feature_tags (start_offset, feature_count, feature_tags);
+}
+
+
+/**
+ * hb_ot_layout_table_find_feature:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @feature_tag: The #hb_tag_t og the requested feature tag
+ * @feature_index: (out): The index of the requested feature
+ *
+ * Fetches the index for a given feature tag in the specified face's GSUB table
+ * or GPOS table.
+ *
+ * Return value: true if the feature is found, false otherwise
+ **/
+bool
+hb_ot_layout_table_find_feature (hb_face_t    *face,
+				 hb_tag_t      table_tag,
+				 hb_tag_t      feature_tag,
+				 unsigned int *feature_index /* OUT */)
+{
+  static_assert ((OT::Index::NOT_FOUND_INDEX == HB_OT_LAYOUT_NO_FEATURE_INDEX), "");
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  unsigned int num_features = g.get_feature_count ();
+  for (unsigned int i = 0; i < num_features; i++)
+  {
+    if (feature_tag == g.get_feature_tag (i)) {
+      if (feature_index) *feature_index = i;
+      return true;
+    }
+  }
+
+  if (feature_index) *feature_index = HB_OT_LAYOUT_NO_FEATURE_INDEX;
+  return false;
+}
+
+
+/**
+ * hb_ot_layout_script_get_language_tags:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @start_offset: offset of the first language tag to retrieve
+ * @language_count: (inout) (allow-none): Input = the maximum number of language tags to return;
+ *                  Output = the actual number of language tags returned (may be zero)
+ * @language_tags: (out) (array length=language_count): Array of language tags found in the table
+ *
+ * Fetches a list of language tags in the given face's GSUB or GPOS table, underneath
+ * the specified script index. The list returned will begin at the offset provided.
+ *
+ **/
+unsigned int
+hb_ot_layout_script_get_language_tags (hb_face_t    *face,
+				       hb_tag_t      table_tag,
+				       unsigned int  script_index,
+				       unsigned int  start_offset,
+				       unsigned int *language_count /* IN/OUT */,
+				       hb_tag_t     *language_tags  /* OUT */)
+{
+  const OT::Script &s = get_gsubgpos_table (face, table_tag).get_script (script_index);
+
+  return s.get_lang_sys_tags (start_offset, language_count, language_tags);
+}
+
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_ot_layout_script_find_language:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_tag: The #hb_tag_t of the requested language
+ * @language_index: The index of the requested language
+ *
+ * Fetches the index of a given language tag in the specified face's GSUB table
+ * or GPOS table, underneath the specified script tag.
+ *
+ * Return value: true if the language tag is found, false otherwise
+ *
+ * Since: ??
+ * Deprecated: ??
+ **/
+hb_bool_t
+hb_ot_layout_script_find_language (hb_face_t    *face,
+				   hb_tag_t      table_tag,
+				   unsigned int  script_index,
+				   hb_tag_t      language_tag,
+				   unsigned int *language_index)
+{
+  return hb_ot_layout_script_select_language (face,
+					      table_tag,
+					      script_index,
+					      1,
+					      &language_tag,
+					      language_index);
+}
+#endif
+
+
+/**
+ * hb_ot_layout_script_select_language:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_count: The number of languages in the specified script
+ * @language_tags: The array of language tags
+ * @language_index: (out): The index of the requested language
+ *
+ * Fetches the index of a given language tag in the specified face's GSUB table
+ * or GPOS table, underneath the specified script index.
+ *
+ * Return value: true if the language tag is found, false otherwise
+ *
+ * Since: 2.0.0
+ **/
+hb_bool_t
+hb_ot_layout_script_select_language (hb_face_t      *face,
+				     hb_tag_t        table_tag,
+				     unsigned int    script_index,
+				     unsigned int    language_count,
+				     const hb_tag_t *language_tags,
+				     unsigned int   *language_index /* OUT */)
+{
+  static_assert ((OT::Index::NOT_FOUND_INDEX == HB_OT_LAYOUT_DEFAULT_LANGUAGE_INDEX), "");
+  const OT::Script &s = get_gsubgpos_table (face, table_tag).get_script (script_index);
+  unsigned int i;
+
+  for (i = 0; i < language_count; i++)
+  {
+    if (s.find_lang_sys_index (language_tags[i], language_index))
+      return true;
+  }
+
+  /* try finding 'dflt' */
+  if (s.find_lang_sys_index (HB_OT_TAG_DEFAULT_LANGUAGE, language_index))
+    return false;
+
+  if (language_index) *language_index = HB_OT_LAYOUT_DEFAULT_LANGUAGE_INDEX;
+  return false;
+}
+
+
+/**
+ * hb_ot_layout_language_get_required_feature_index:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_index: The index of the requested language tag
+ * @feature_index: (out): The index of the requested feature
+ *
+ * Fetches the index of a requested feature in the given face's GSUB or GPOS table,
+ * underneath the specified script and language.
+ *
+ * Return value: true if the feature is found, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_language_get_required_feature_index (hb_face_t    *face,
+						  hb_tag_t      table_tag,
+						  unsigned int  script_index,
+						  unsigned int  language_index,
+						  unsigned int *feature_index /* OUT */)
+{
+  return hb_ot_layout_language_get_required_feature (face,
+						     table_tag,
+						     script_index,
+						     language_index,
+						     feature_index,
+						     nullptr);
+}
+
+
+/**
+ * hb_ot_layout_language_get_required_feature:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_index: The index of the requested language tag
+ * @feature_index: (out): The index of the requested feature
+ * @feature_tag: (out): The #hb_tag_t of the requested feature
+ *
+ * Fetches the tag of a requested feature index in the given face's GSUB or GPOS table,
+ * underneath the specified script and language.
+ *
+ * Return value: true if the feature is found, false otherwise
+ *
+ * Since: 0.9.30
+ **/
+hb_bool_t
+hb_ot_layout_language_get_required_feature (hb_face_t    *face,
+					    hb_tag_t      table_tag,
+					    unsigned int  script_index,
+					    unsigned int  language_index,
+					    unsigned int *feature_index /* OUT */,
+					    hb_tag_t     *feature_tag   /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  const OT::LangSys &l = g.get_script (script_index).get_lang_sys (language_index);
+
+  unsigned int index = l.get_required_feature_index ();
+  if (feature_index) *feature_index = index;
+  if (feature_tag) *feature_tag = g.get_feature_tag (index);
+
+  return l.has_required_feature ();
+}
+
+
+/**
+ * hb_ot_layout_language_get_feature_indexes:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_index: The index of the requested language tag
+ * @start_offset: offset of the first feature tag to retrieve
+ * @feature_count: (inout) (allow-none): Input = the maximum number of feature tags to return;
+ *                 Output: the actual number of feature tags returned (may be zero)
+ * @feature_indexes: (out) (array length=feature_count): The array of feature indexes found for the query
+ *
+ * Fetches a list of all features in the specified face's GSUB table
+ * or GPOS table, underneath the specified script and language. The list
+ * returned will begin at the offset provided.
+ **/
+unsigned int
+hb_ot_layout_language_get_feature_indexes (hb_face_t    *face,
+					   hb_tag_t      table_tag,
+					   unsigned int  script_index,
+					   unsigned int  language_index,
+					   unsigned int  start_offset,
+					   unsigned int *feature_count   /* IN/OUT */,
+					   unsigned int *feature_indexes /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  const OT::LangSys &l = g.get_script (script_index).get_lang_sys (language_index);
+
+  return l.get_feature_indexes (start_offset, feature_count, feature_indexes);
+}
+
+
+/**
+ * hb_ot_layout_language_get_feature_tags:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_index: The index of the requested language tag
+ * @start_offset: offset of the first feature tag to retrieve
+ * @feature_count: (inout) (allow-none): Input = the maximum number of feature tags to return;
+ *                 Output = the actual number of feature tags returned (may be zero)
+ * @feature_tags: (out) (array length=feature_count): The array of #hb_tag_t feature tags found for the query
+ *
+ * Fetches a list of all features in the specified face's GSUB table
+ * or GPOS table, underneath the specified script and language. The list
+ * returned will begin at the offset provided.
+ *
+ **/
+unsigned int
+hb_ot_layout_language_get_feature_tags (hb_face_t    *face,
+					hb_tag_t      table_tag,
+					unsigned int  script_index,
+					unsigned int  language_index,
+					unsigned int  start_offset,
+					unsigned int *feature_count /* IN/OUT */,
+					hb_tag_t     *feature_tags  /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  const OT::LangSys &l = g.get_script (script_index).get_lang_sys (language_index);
+
+  static_assert ((sizeof (unsigned int) == sizeof (hb_tag_t)), "");
+  unsigned int ret = l.get_feature_indexes (start_offset, feature_count, (unsigned int *) feature_tags);
+
+  if (feature_tags) {
+    unsigned int count = *feature_count;
+    for (unsigned int i = 0; i < count; i++)
+      feature_tags[i] = g.get_feature_tag ((unsigned int) feature_tags[i]);
+  }
+
+  return ret;
+}
+
+
+/**
+ * hb_ot_layout_language_find_feature:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @script_index: The index of the requested script tag
+ * @language_index: The index of the requested language tag
+ * @feature_tag: #hb_tag_t of the feature tag requested
+ * @feature_index: (out): The index of the requested feature
+ *
+ * Fetches the index of a given feature tag in the specified face's GSUB table
+ * or GPOS table, underneath the specified script and language.
+ *
+ * Return value: true if the feature is found, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_language_find_feature (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  script_index,
+				    unsigned int  language_index,
+				    hb_tag_t      feature_tag,
+				    unsigned int *feature_index /* OUT */)
+{
+  static_assert ((OT::Index::NOT_FOUND_INDEX == HB_OT_LAYOUT_NO_FEATURE_INDEX), "");
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  const OT::LangSys &l = g.get_script (script_index).get_lang_sys (language_index);
+
+  unsigned int num_features = l.get_feature_count ();
+  for (unsigned int i = 0; i < num_features; i++) {
+    unsigned int f_index = l.get_feature_index (i);
+
+    if (feature_tag == g.get_feature_tag (f_index)) {
+      if (feature_index) *feature_index = f_index;
+      return true;
+    }
+  }
+
+  if (feature_index) *feature_index = HB_OT_LAYOUT_NO_FEATURE_INDEX;
+  return false;
+}
+
+
+/**
+ * hb_ot_layout_feature_get_lookups:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @feature_index: The index of the requested feature
+ * @start_offset: offset of the first lookup to retrieve
+ * @lookup_count: (inout) (allow-none): Input = the maximum number of lookups to return;
+ *                Output = the actual number of lookups returned (may be zero)
+ * @lookup_indexes: (out) (array length=lookup_count): The array of lookup indexes found for the query
+ *
+ * Fetches a list of all lookups enumerated for the specified feature, in
+ * the specified face's GSUB table or GPOS table. The list returned will
+ * begin at the offset provided.
+ *
+ * Since: 0.9.7
+ **/
+unsigned int
+hb_ot_layout_feature_get_lookups (hb_face_t    *face,
+				  hb_tag_t      table_tag,
+				  unsigned int  feature_index,
+				  unsigned int  start_offset,
+				  unsigned int *lookup_count   /* IN/OUT */,
+				  unsigned int *lookup_indexes /* OUT */)
+{
+  return hb_ot_layout_feature_with_variations_get_lookups (face,
+							   table_tag,
+							   feature_index,
+							   HB_OT_LAYOUT_NO_VARIATIONS_INDEX,
+							   start_offset,
+							   lookup_count,
+							   lookup_indexes);
+}
+
+
+/**
+ * hb_ot_layout_table_get_lookup_count:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ *
+ * Fetches the total number of lookups enumerated in the specified
+ * face's GSUB table or GPOS table.
+ *
+ * Since: 0.9.22
+ **/
+unsigned int
+hb_ot_layout_table_get_lookup_count (hb_face_t    *face,
+				     hb_tag_t      table_tag)
+{
+  return get_gsubgpos_table (face, table_tag).get_lookup_count ();
+}
+
+
+struct hb_collect_features_context_t
+{
+  hb_collect_features_context_t (hb_face_t *face,
+				 hb_tag_t   table_tag,
+				 hb_set_t  *feature_indexes_)
+    : g (get_gsubgpos_table (face, table_tag)),
+      feature_indexes (feature_indexes_),
+      script_count (0),langsys_count (0), feature_index_count (0) {}
+
+  bool visited (const OT::Script &s)
+  {
+    /* We might have Null() object here.  Don't want to involve
+     * that in the memoize.  So, detect empty objects and return. */
+    if (unlikely (!s.has_default_lang_sys () &&
+		  !s.get_lang_sys_count ()))
+      return true;
+
+    if (script_count++ > HB_MAX_SCRIPTS)
+      return true;
+
+    return visited (s, visited_script);
+  }
+  bool visited (const OT::LangSys &l)
+  {
+    /* We might have Null() object here.  Don't want to involve
+     * that in the memoize.  So, detect empty objects and return. */
+    if (unlikely (!l.has_required_feature () &&
+		  !l.get_feature_count ()))
+      return true;
+
+    if (langsys_count++ > HB_MAX_LANGSYS)
+      return true;
+
+    return visited (l, visited_langsys);
+  }
+
+  bool visited_feature_indices (unsigned count)
+  {
+    feature_index_count += count;
+    return feature_index_count > HB_MAX_FEATURE_INDICES;
+  }
+
+  private:
+  template <typename T>
+  bool visited (const T &p, hb_set_t &visited_set)
+  {
+    hb_codepoint_t delta = (hb_codepoint_t) ((uintptr_t) &p - (uintptr_t) &g);
+     if (visited_set.has (delta))
+      return true;
+
+    visited_set.add (delta);
+    return false;
+  }
+
+  public:
+  const OT::GSUBGPOS &g;
+  hb_set_t           *feature_indexes;
+
+  private:
+  hb_set_t visited_script;
+  hb_set_t visited_langsys;
+  unsigned int script_count;
+  unsigned int langsys_count;
+  unsigned int feature_index_count;
+};
+
+static void
+langsys_collect_features (hb_collect_features_context_t *c,
+			  const OT::LangSys  &l,
+			  const hb_tag_t     *features)
+{
+  if (c->visited (l)) return;
+
+  if (!features)
+  {
+    /* All features. */
+    if (l.has_required_feature () && !c->visited_feature_indices (1))
+      c->feature_indexes->add (l.get_required_feature_index ());
+
+    if (!c->visited_feature_indices (l.featureIndex.len))
+      l.add_feature_indexes_to (c->feature_indexes);
+  }
+  else
+  {
+    /* Ugh. Any faster way? */
+    for (; *features; features++)
+    {
+      hb_tag_t feature_tag = *features;
+      unsigned int num_features = l.get_feature_count ();
+      for (unsigned int i = 0; i < num_features; i++)
+      {
+	unsigned int feature_index = l.get_feature_index (i);
+
+	if (feature_tag == c->g.get_feature_tag (feature_index))
+	{
+	  c->feature_indexes->add (feature_index);
+	  break;
+	}
+      }
+    }
+  }
+}
+
+static void
+script_collect_features (hb_collect_features_context_t *c,
+			 const OT::Script   &s,
+			 const hb_tag_t *languages,
+			 const hb_tag_t *features)
+{
+  if (c->visited (s)) return;
+
+  if (!languages)
+  {
+    /* All languages. */
+    if (s.has_default_lang_sys ())
+      langsys_collect_features (c,
+				s.get_default_lang_sys (),
+				features);
+
+    unsigned int count = s.get_lang_sys_count ();
+    for (unsigned int language_index = 0; language_index < count; language_index++)
+      langsys_collect_features (c,
+				s.get_lang_sys (language_index),
+				features);
+  }
+  else
+  {
+    for (; *languages; languages++)
+    {
+      unsigned int language_index;
+      if (s.find_lang_sys_index (*languages, &language_index))
+	langsys_collect_features (c,
+				  s.get_lang_sys (language_index),
+				  features);
+    }
+  }
+}
+
+
+/**
+ * hb_ot_layout_collect_features:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @scripts: The array of scripts to collect features for
+ * @languages: The array of languages to collect features for
+ * @features: The array of features to collect
+ * @feature_indexes: (out): The array of feature indexes found for the query
+ *
+ * Fetches a list of all feature indexes in the specified face's GSUB table
+ * or GPOS table, underneath the specified scripts, languages, and features.
+ * If no list of scripts is provided, all scripts will be queried. If no list
+ * of languages is provided, all languages will be queried. If no list of
+ * features is provided, all features will be queried.
+ *
+ * Since: 1.8.5
+ **/
+void
+hb_ot_layout_collect_features (hb_face_t      *face,
+			       hb_tag_t        table_tag,
+			       const hb_tag_t *scripts,
+			       const hb_tag_t *languages,
+			       const hb_tag_t *features,
+			       hb_set_t       *feature_indexes /* OUT */)
+{
+  hb_collect_features_context_t c (face, table_tag, feature_indexes);
+  if (!scripts)
+  {
+    /* All scripts. */
+    unsigned int count = c.g.get_script_count ();
+    for (unsigned int script_index = 0; script_index < count; script_index++)
+      script_collect_features (&c,
+			       c.g.get_script (script_index),
+			       languages,
+			       features);
+  }
+  else
+  {
+    for (; *scripts; scripts++)
+    {
+      unsigned int script_index;
+      if (c.g.find_script_index (*scripts, &script_index))
+	script_collect_features (&c,
+				 c.g.get_script (script_index),
+				 languages,
+				 features);
+    }
+  }
+}
+
+
+/**
+ * hb_ot_layout_collect_lookups:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @scripts: The array of scripts to collect lookups for
+ * @languages: The array of languages to collect lookups for
+ * @features: The array of features to collect lookups for
+ * @lookup_indexes: (out): The array of lookup indexes found for the query
+ *
+ * Fetches a list of all feature-lookup indexes in the specified face's GSUB
+ * table or GPOS table, underneath the specified scripts, languages, and
+ * features. If no list of scripts is provided, all scripts will be queried.
+ * If no list of languages is provided, all languages will be queried. If no
+ * list of features is provided, all features will be queried.
+ *
+ * Since: 0.9.8
+ **/
+void
+hb_ot_layout_collect_lookups (hb_face_t      *face,
+			      hb_tag_t        table_tag,
+			      const hb_tag_t *scripts,
+			      const hb_tag_t *languages,
+			      const hb_tag_t *features,
+			      hb_set_t       *lookup_indexes /* OUT */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  hb_set_t feature_indexes;
+  hb_ot_layout_collect_features (face, table_tag, scripts, languages, features, &feature_indexes);
+
+  for (hb_codepoint_t feature_index = HB_SET_VALUE_INVALID;
+       hb_set_next (&feature_indexes, &feature_index);)
+    g.get_feature (feature_index).add_lookup_indexes_to (lookup_indexes);
+
+  g.feature_variation_collect_lookups (&feature_indexes, lookup_indexes);
+}
+
+
+#ifndef HB_NO_LAYOUT_COLLECT_GLYPHS
+/**
+ * hb_ot_layout_lookup_collect_glyphs:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @lookup_index: The index of the feature lookup to query
+ * @glyphs_before: (out): Array of glyphs preceding the substitution range
+ * @glyphs_input: (out): Array of input glyphs that would be substituted by the lookup
+ * @glyphs_after: (out): Array of glyphs following the substitution range
+ * @glyphs_output: (out): Array of glyphs that would be the substitued output of the lookup
+ *
+ * Fetches a list of all glyphs affected by the specified lookup in the
+ * specified face's GSUB table or GPOS table.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_ot_layout_lookup_collect_glyphs (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  lookup_index,
+				    hb_set_t     *glyphs_before, /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_input,  /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_after,  /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_output  /* OUT.  May be NULL */)
+{
+  OT::hb_collect_glyphs_context_t c (face,
+				     glyphs_before,
+				     glyphs_input,
+				     glyphs_after,
+				     glyphs_output);
+
+  switch (table_tag)
+  {
+    case HB_OT_TAG_GSUB:
+    {
+      const OT::SubstLookup& l = face->table.GSUB->table->get_lookup (lookup_index);
+      l.collect_glyphs (&c);
+      return;
+    }
+    case HB_OT_TAG_GPOS:
+    {
+      const OT::PosLookup& l = face->table.GPOS->table->get_lookup (lookup_index);
+      l.collect_glyphs (&c);
+      return;
+    }
+  }
+}
+#endif
+
+
+/* Variations support */
+
+
+/**
+ * hb_ot_layout_table_find_feature_variations:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @coords: The variation coordinates to query
+ * @num_coords: The number of variation coorinates
+ * @variations_index: (out): The array of feature variations found for the query
+ *
+ * Fetches a list of feature variations in the specified face's GSUB table
+ * or GPOS table, at the specified variation coordinates.
+ *
+ **/
+hb_bool_t
+hb_ot_layout_table_find_feature_variations (hb_face_t    *face,
+					    hb_tag_t      table_tag,
+					    const int    *coords,
+					    unsigned int  num_coords,
+					    unsigned int *variations_index /* out */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  return g.find_variations_index (coords, num_coords, variations_index);
+}
+
+
+/**
+ * hb_ot_layout_feature_with_variations_get_lookups:
+ * @face: #hb_face_t to work upon
+ * @table_tag: HB_OT_TAG_GSUB or HB_OT_TAG_GPOS
+ * @feature_index: The index of the feature to query
+ * @variations_index: The index of the feature variation to query
+ * @start_offset: offset of the first lookup to retrieve
+ * @lookup_count: (inout) (allow-none): Input = the maximum number of lookups to return;
+ *                Output = the actual number of lookups returned (may be zero)
+ * @lookup_indexes: (out) (array length=lookup_count): The array of lookups found for the query
+ *
+ * Fetches a list of all lookups enumerated for the specified feature, in
+ * the specified face's GSUB table or GPOS table, enabled at the specified
+ * variations index. The list returned will begin at the offset provided.
+ *
+ **/
+unsigned int
+hb_ot_layout_feature_with_variations_get_lookups (hb_face_t    *face,
+						  hb_tag_t      table_tag,
+						  unsigned int  feature_index,
+						  unsigned int  variations_index,
+						  unsigned int  start_offset,
+						  unsigned int *lookup_count /* IN/OUT */,
+						  unsigned int *lookup_indexes /* OUT */)
+{
+  static_assert ((OT::FeatureVariations::NOT_FOUND_INDEX == HB_OT_LAYOUT_NO_VARIATIONS_INDEX), "");
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  const OT::Feature &f = g.get_feature_variation (feature_index, variations_index);
+
+  return f.get_lookup_indexes (start_offset, lookup_count, lookup_indexes);
+}
+
+
+/*
+ * OT::GSUB
+ */
+
+
+/**
+ * hb_ot_layout_has_substitution:
+ * @face: #hb_face_t to work upon
+ *
+ * Tests whether the specified face includes any GSUB substitutions.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_has_substitution (hb_face_t *face)
+{
+  return face->table.GSUB->table->has_data ();
+}
+
+
+/**
+ * hb_ot_layout_lookup_would_substitute:
+ * @face: #hb_face_t to work upon
+ * @lookup_index: The index of the lookup to query
+ * @glyphs: The sequence of glyphs to query for substitution
+ * @glyphs_length: The length of the glyph sequence
+ * @zero_context: #hb_bool_t indicating whether substitutions should be context-free
+ *
+ * Tests whether a specified lookup in the specified face would
+ * trigger a substitution on the given glyph sequence.
+ *
+ * Return value: true if a substitution would be triggered, false otherwise
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_ot_layout_lookup_would_substitute (hb_face_t            *face,
+				      unsigned int          lookup_index,
+				      const hb_codepoint_t *glyphs,
+				      unsigned int          glyphs_length,
+				      hb_bool_t             zero_context)
+{
+  if (unlikely (lookup_index >= face->table.GSUB->lookup_count)) return false;
+  OT::hb_would_apply_context_t c (face, glyphs, glyphs_length, (bool) zero_context);
+
+  const OT::SubstLookup& l = face->table.GSUB->table->get_lookup (lookup_index);
+  return l.would_apply (&c, &face->table.GSUB->accels[lookup_index]);
+}
+
+
+/**
+ * hb_ot_layout_substitute_start:
+ * @font: #hb_font_t to use
+ * @buffer: #hb_buffer_t buffer to work upon
+ *
+ * Called before substitution lookups are performed, to ensure that glyph
+ * class and other properties are set on the glyphs in the buffer.
+ *
+ **/
+void
+hb_ot_layout_substitute_start (hb_font_t    *font,
+			       hb_buffer_t  *buffer)
+{
+  _hb_ot_layout_set_glyph_props (font, buffer);
+}
+
+void
+hb_ot_layout_delete_glyphs_inplace (hb_buffer_t *buffer,
+				    bool (*filter) (const hb_glyph_info_t *info))
+{
+  /* Merge clusters and delete filtered glyphs.
+   * NOTE! We can't use out-buffer as we have positioning data. */
+  unsigned int j = 0;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pos = buffer->pos;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    if (filter (&info[i]))
+    {
+      /* Merge clusters.
+       * Same logic as buffer->delete_glyph(), but for in-place removal. */
+
+      unsigned int cluster = info[i].cluster;
+      if (i + 1 < count && cluster == info[i + 1].cluster)
+	continue; /* Cluster survives; do nothing. */
+
+      if (j)
+      {
+	/* Merge cluster backward. */
+	if (cluster < info[j - 1].cluster)
+	{
+	  unsigned int mask = info[i].mask;
+	  unsigned int old_cluster = info[j - 1].cluster;
+	  for (unsigned k = j; k && info[k - 1].cluster == old_cluster; k--)
+	    buffer->set_cluster (info[k - 1], cluster, mask);
+	}
+	continue;
+      }
+
+      if (i + 1 < count)
+	buffer->merge_clusters (i, i + 2); /* Merge cluster forward. */
+
+      continue;
+    }
+
+    if (j != i)
+    {
+      info[j] = info[i];
+      pos[j] = pos[i];
+    }
+    j++;
+  }
+  buffer->len = j;
+}
+
+/**
+ * hb_ot_layout_lookup_substitute_closure:
+ * @face: #hb_face_t to work upon
+ * @lookup_index: index of the feature lookup to query
+ * @glyphs: (out): Array of glyphs comprising the transitive closure of the lookup
+ *
+ * Compute the transitive closure of glyphs needed for a
+ * specified lookup.
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_ot_layout_lookup_substitute_closure (hb_face_t    *face,
+					unsigned int  lookup_index,
+					hb_set_t     *glyphs /* OUT */)
+{
+  hb_map_t done_lookups;
+  OT::hb_closure_context_t c (face, glyphs, &done_lookups);
+
+  const OT::SubstLookup& l = face->table.GSUB->table->get_lookup (lookup_index);
+
+  l.closure (&c, lookup_index);
+}
+
+/**
+ * hb_ot_layout_lookups_substitute_closure:
+ * @face: #hb_face_t to work upon
+ * @lookups: The set of lookups to query
+ * @glyphs: (out): Array of glyphs comprising the transitive closure of the lookups
+ *
+ * Compute the transitive closure of glyphs needed for all of the
+ * provided lookups.
+ *
+ * Since: 1.8.1
+ **/
+void
+hb_ot_layout_lookups_substitute_closure (hb_face_t      *face,
+					 const hb_set_t *lookups,
+					 hb_set_t       *glyphs /* OUT */)
+{
+  hb_map_t done_lookups;
+  OT::hb_closure_context_t c (face, glyphs, &done_lookups);
+  const OT::GSUB& gsub = *face->table.GSUB->table;
+
+  unsigned int iteration_count = 0;
+  unsigned int glyphs_length;
+  do
+  {
+    glyphs_length = glyphs->get_population ();
+    if (lookups)
+    {
+      for (hb_codepoint_t lookup_index = HB_SET_VALUE_INVALID; hb_set_next (lookups, &lookup_index);)
+	gsub.get_lookup (lookup_index).closure (&c, lookup_index);
+    }
+    else
+    {
+      for (unsigned int i = 0; i < gsub.get_lookup_count (); i++)
+	gsub.get_lookup (i).closure (&c, i);
+    }
+  } while (iteration_count++ <= HB_CLOSURE_MAX_STAGES &&
+	   glyphs_length != glyphs->get_population ());
+}
+
+/*
+ * OT::GPOS
+ */
+
+
+/**
+ * hb_ot_layout_has_positioning:
+ * @face: #hb_face_t to work upon
+ *
+ * Return value: true if the face has GPOS data, false otherwise
+ *
+ **/
+hb_bool_t
+hb_ot_layout_has_positioning (hb_face_t *face)
+{
+  return face->table.GPOS->table->has_data ();
+}
+
+/**
+ * hb_ot_layout_position_start:
+ * @font: #hb_font_t to use
+ * @buffer: #hb_buffer_t buffer to work upon
+ *
+ * Called before positioning lookups are performed, to ensure that glyph
+ * attachment types and glyph-attachment chains are set for the glyphs in the buffer.
+ *
+ **/
+void
+hb_ot_layout_position_start (hb_font_t *font, hb_buffer_t *buffer)
+{
+  OT::GPOS::position_start (font, buffer);
+}
+
+
+/**
+ * hb_ot_layout_position_finish_advances:
+ * @font: #hb_font_t to use
+ * @buffer: #hb_buffer_t buffer to work upon
+ *
+ * Called after positioning lookups are performed, to finish glyph advances.
+ *
+ **/
+void
+hb_ot_layout_position_finish_advances (hb_font_t *font, hb_buffer_t *buffer)
+{
+  OT::GPOS::position_finish_advances (font, buffer);
+}
+
+/**
+ * hb_ot_layout_position_finish_offsets:
+ * @font: #hb_font_t to use
+ * @buffer: #hb_buffer_t buffer to work upon
+ *
+ * Called after positioning lookups are performed, to finish glyph offsets.
+ *
+ **/
+void
+hb_ot_layout_position_finish_offsets (hb_font_t *font, hb_buffer_t *buffer)
+{
+  OT::GPOS::position_finish_offsets (font, buffer);
+}
+
+
+#ifndef HB_NO_LAYOUT_FEATURE_PARAMS
+/**
+ * hb_ot_layout_get_size_params:
+ * @face: #hb_face_t to work upon
+ * @design_size: (out): The design size of the face
+ * @subfamily_id: (out): The identifier of the face within the font subfamily
+ * @subfamily_name_id: (out): The ‘name’ table name ID of the face within the font subfamily
+ * @range_start: (out): The minimum size of the recommended size range for the face
+ * @range_end: (out): The maximum size of the recommended size range for the face
+ *
+ * Fetches optical-size feature data (i.e., the `size` feature from GPOS). Note that
+ * the subfamily_id and the subfamily name string (accessible via the subfamily_name_id)
+ * as used here are defined as pertaining only to fonts within a font family that differ
+ * specifically in their respective size ranges; other ways to differentiate fonts within
+ * a subfamily are not covered by the `size` feature.
+ *
+ * For more information on this distinction, see the [`size` feature documentation](
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/features_pt#tag-size).
+ *
+ * Return value: true if data found, false otherwise
+ *
+ * Since: 0.9.10
+ **/
+hb_bool_t
+hb_ot_layout_get_size_params (hb_face_t       *face,
+			      unsigned int    *design_size,       /* OUT.  May be NULL */
+			      unsigned int    *subfamily_id,      /* OUT.  May be NULL */
+			      hb_ot_name_id_t *subfamily_name_id, /* OUT.  May be NULL */
+			      unsigned int    *range_start,       /* OUT.  May be NULL */
+			      unsigned int    *range_end          /* OUT.  May be NULL */)
+{
+  const OT::GPOS &gpos = *face->table.GPOS->table;
+  const hb_tag_t tag = HB_TAG ('s','i','z','e');
+
+  unsigned int num_features = gpos.get_feature_count ();
+  for (unsigned int i = 0; i < num_features; i++)
+  {
+    if (tag == gpos.get_feature_tag (i))
+    {
+      const OT::Feature &f = gpos.get_feature (i);
+      const OT::FeatureParamsSize &params = f.get_feature_params ().get_size_params (tag);
+
+      if (params.designSize)
+      {
+	if (design_size) *design_size = params.designSize;
+	if (subfamily_id) *subfamily_id = params.subfamilyID;
+	if (subfamily_name_id) *subfamily_name_id = params.subfamilyNameID;
+	if (range_start) *range_start = params.rangeStart;
+	if (range_end) *range_end = params.rangeEnd;
+
+	return true;
+      }
+    }
+  }
+
+  if (design_size) *design_size = 0;
+  if (subfamily_id) *subfamily_id = 0;
+  if (subfamily_name_id) *subfamily_name_id = HB_OT_NAME_ID_INVALID;
+  if (range_start) *range_start = 0;
+  if (range_end) *range_end = 0;
+
+  return false;
+}
+/**
+ * hb_ot_layout_feature_get_name_ids:
+ * @face: #hb_face_t to work upon
+ * @table_tag: table tag to query, "GSUB" or "GPOS".
+ * @feature_index: index of feature to query.
+ * @label_id: (out) (allow-none): The ‘name’ table name ID that specifies a string
+ *            for a user-interface label for this feature. (May be NULL.)
+ * @tooltip_id: (out) (allow-none): The ‘name’ table name ID that specifies a string
+ *              that an application can use for tooltip text for this
+ *              feature. (May be NULL.)
+ * @sample_id: (out) (allow-none): The ‘name’ table name ID that specifies sample text
+ *             that illustrates the effect of this feature. (May be NULL.)
+ * @num_named_parameters: (out) (allow-none):  Number of named parameters. (May be zero.)
+ * @first_param_id: (out) (allow-none): The first ‘name’ table name ID used to specify
+ *                  strings for user-interface labels for the feature
+ *                  parameters. (Must be zero if numParameters is zero.)
+ *
+ * Fetches name indices from feature parameters for "Stylistic Set" ('ssXX') or
+ * "Character Variant" ('cvXX') features.
+ *
+ * Return value: true if data found, false otherwise
+ *
+ * Since: 2.0.0
+ **/
+hb_bool_t
+hb_ot_layout_feature_get_name_ids (hb_face_t       *face,
+				   hb_tag_t         table_tag,
+				   unsigned int     feature_index,
+				   hb_ot_name_id_t *label_id,             /* OUT.  May be NULL */
+				   hb_ot_name_id_t *tooltip_id,           /* OUT.  May be NULL */
+				   hb_ot_name_id_t *sample_id,            /* OUT.  May be NULL */
+				   unsigned int    *num_named_parameters, /* OUT.  May be NULL */
+				   hb_ot_name_id_t *first_param_id        /* OUT.  May be NULL */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+
+  hb_tag_t feature_tag = g.get_feature_tag (feature_index);
+  const OT::Feature &f = g.get_feature (feature_index);
+
+  const OT::FeatureParams &feature_params = f.get_feature_params ();
+  if (&feature_params != &Null (OT::FeatureParams))
+  {
+    const OT::FeatureParamsStylisticSet& ss_params =
+      feature_params.get_stylistic_set_params (feature_tag);
+    if (&ss_params != &Null (OT::FeatureParamsStylisticSet)) /* ssXX */
+    {
+      if (label_id) *label_id = ss_params.uiNameID;
+      // ssXX features don't have the rest
+      if (tooltip_id) *tooltip_id = HB_OT_NAME_ID_INVALID;
+      if (sample_id) *sample_id = HB_OT_NAME_ID_INVALID;
+      if (num_named_parameters) *num_named_parameters = 0;
+      if (first_param_id) *first_param_id = HB_OT_NAME_ID_INVALID;
+      return true;
+    }
+    const OT::FeatureParamsCharacterVariants& cv_params =
+      feature_params.get_character_variants_params (feature_tag);
+    if (&cv_params != &Null (OT::FeatureParamsCharacterVariants)) /* cvXX */
+    {
+      if (label_id) *label_id = cv_params.featUILableNameID;
+      if (tooltip_id) *tooltip_id = cv_params.featUITooltipTextNameID;
+      if (sample_id) *sample_id = cv_params.sampleTextNameID;
+      if (num_named_parameters) *num_named_parameters = cv_params.numNamedParameters;
+      if (first_param_id) *first_param_id = cv_params.firstParamUILabelNameID;
+      return true;
+    }
+  }
+
+  if (label_id) *label_id = HB_OT_NAME_ID_INVALID;
+  if (tooltip_id) *tooltip_id = HB_OT_NAME_ID_INVALID;
+  if (sample_id) *sample_id = HB_OT_NAME_ID_INVALID;
+  if (num_named_parameters) *num_named_parameters = 0;
+  if (first_param_id) *first_param_id = HB_OT_NAME_ID_INVALID;
+  return false;
+}
+/**
+ * hb_ot_layout_feature_get_characters:
+ * @face: #hb_face_t to work upon
+ * @table_tag: table tag to query, "GSUB" or "GPOS".
+ * @feature_index: index of feature to query.
+ * @start_offset: offset of the first character to retrieve
+ * @char_count: (inout) (allow-none): Input = the maximum number of characters to return;
+ *              Output = the actual number of characters returned (may be zero)
+ * @characters: (out caller-allocates) (array length=char_count): A buffer pointer.
+ *              The Unicode codepoints of the characters for which this feature provides
+ *               glyph variants.
+ *
+ * Fetches a list of the characters defined as having a variant under the specified
+ * "Character Variant" ("cvXX") feature tag.
+ *
+ * Return value: Number of total sample characters in the cvXX feature.
+ *
+ * Since: 2.0.0
+ **/
+unsigned int
+hb_ot_layout_feature_get_characters (hb_face_t      *face,
+				     hb_tag_t        table_tag,
+				     unsigned int    feature_index,
+				     unsigned int    start_offset,
+				     unsigned int   *char_count, /* IN/OUT.  May be NULL */
+				     hb_codepoint_t *characters  /* OUT.     May be NULL */)
+{
+  const OT::GSUBGPOS &g = get_gsubgpos_table (face, table_tag);
+  return g.get_feature (feature_index)
+	  .get_feature_params ()
+	  .get_character_variants_params(g.get_feature_tag (feature_index))
+	  .get_characters (start_offset, char_count, characters);
+}
+#endif
+
+
+/*
+ * Parts of different types are implemented here such that they have direct
+ * access to GSUB/GPOS lookups.
+ */
+
+
+struct GSUBProxy
+{
+  static constexpr unsigned table_index = 0u;
+  static constexpr bool inplace = false;
+  typedef OT::SubstLookup Lookup;
+
+  GSUBProxy (hb_face_t *face) :
+    table (*face->table.GSUB->table),
+    accels (face->table.GSUB->accels) {}
+
+  const OT::GSUB &table;
+  const OT::hb_ot_layout_lookup_accelerator_t *accels;
+};
+
+struct GPOSProxy
+{
+  static constexpr unsigned table_index = 1u;
+  static constexpr bool inplace = true;
+  typedef OT::PosLookup Lookup;
+
+  GPOSProxy (hb_face_t *face) :
+    table (*face->table.GPOS->table),
+    accels (face->table.GPOS->accels) {}
+
+  const OT::GPOS &table;
+  const OT::hb_ot_layout_lookup_accelerator_t *accels;
+};
+
+
+static inline bool
+apply_forward (OT::hb_ot_apply_context_t *c,
+	       const OT::hb_ot_layout_lookup_accelerator_t &accel)
+{
+  bool ret = false;
+  hb_buffer_t *buffer = c->buffer;
+  while (buffer->idx < buffer->len && buffer->successful)
+  {
+    bool applied = false;
+    if (accel.may_have (buffer->cur().codepoint) &&
+	(buffer->cur().mask & c->lookup_mask) &&
+	c->check_glyph_property (&buffer->cur(), c->lookup_props))
+     {
+       applied = accel.apply (c);
+     }
+
+    if (applied)
+      ret = true;
+    else
+      buffer->next_glyph ();
+  }
+  return ret;
+}
+
+static inline bool
+apply_backward (OT::hb_ot_apply_context_t *c,
+	       const OT::hb_ot_layout_lookup_accelerator_t &accel)
+{
+  bool ret = false;
+  hb_buffer_t *buffer = c->buffer;
+  do
+  {
+    if (accel.may_have (buffer->cur().codepoint) &&
+	(buffer->cur().mask & c->lookup_mask) &&
+	c->check_glyph_property (&buffer->cur(), c->lookup_props))
+     ret |= accel.apply (c);
+
+    /* The reverse lookup doesn't "advance" cursor (for good reason). */
+    buffer->idx--;
+
+  }
+  while ((int) buffer->idx >= 0);
+  return ret;
+}
+
+template <typename Proxy>
+static inline void
+apply_string (OT::hb_ot_apply_context_t *c,
+	      const typename Proxy::Lookup &lookup,
+	      const OT::hb_ot_layout_lookup_accelerator_t &accel)
+{
+  hb_buffer_t *buffer = c->buffer;
+
+  if (unlikely (!buffer->len || !c->lookup_mask))
+    return;
+
+  c->set_lookup_props (lookup.get_props ());
+
+  if (likely (!lookup.is_reverse ()))
+  {
+    /* in/out forward substitution/positioning */
+    if (Proxy::table_index == 0u)
+      buffer->clear_output ();
+    buffer->idx = 0;
+
+    bool ret;
+    ret = apply_forward (c, accel);
+    if (ret)
+    {
+      if (!Proxy::inplace)
+	buffer->swap_buffers ();
+      else
+	assert (!buffer->has_separate_output ());
+    }
+  }
+  else
+  {
+    /* in-place backward substitution/positioning */
+    if (Proxy::table_index == 0u)
+      buffer->remove_output ();
+    buffer->idx = buffer->len - 1;
+
+    apply_backward (c, accel);
+  }
+}
+
+template <typename Proxy>
+inline void hb_ot_map_t::apply (const Proxy &proxy,
+				const hb_ot_shape_plan_t *plan,
+				hb_font_t *font,
+				hb_buffer_t *buffer) const
+{
+  const unsigned int table_index = proxy.table_index;
+  unsigned int i = 0;
+  OT::hb_ot_apply_context_t c (table_index, font, buffer);
+  c.set_recurse_func (Proxy::Lookup::apply_recurse_func);
+
+  for (unsigned int stage_index = 0; stage_index < stages[table_index].length; stage_index++) {
+    const stage_map_t *stage = &stages[table_index][stage_index];
+    for (; i < stage->last_lookup; i++)
+    {
+      unsigned int lookup_index = lookups[table_index][i].index;
+      if (!buffer->message (font, "start lookup %d", lookup_index)) continue;
+      c.set_lookup_index (lookup_index);
+      c.set_lookup_mask (lookups[table_index][i].mask);
+      c.set_auto_zwj (lookups[table_index][i].auto_zwj);
+      c.set_auto_zwnj (lookups[table_index][i].auto_zwnj);
+      if (lookups[table_index][i].random)
+      {
+	c.set_random (true);
+	buffer->unsafe_to_break_all ();
+      }
+      apply_string<Proxy> (&c,
+			   proxy.table.get_lookup (lookup_index),
+			   proxy.accels[lookup_index]);
+      (void) buffer->message (font, "end lookup %d", lookup_index);
+    }
+
+    if (stage->pause_func)
+    {
+      buffer->clear_output ();
+      stage->pause_func (plan, font, buffer);
+    }
+  }
+}
+
+void hb_ot_map_t::substitute (const hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer) const
+{
+  GSUBProxy proxy (font->face);
+  if (!buffer->message (font, "start table GSUB")) return;
+  apply (proxy, plan, font, buffer);
+  (void)buffer->message (font, "end table GSUB");
+}
+
+void hb_ot_map_t::position (const hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer) const
+{
+  GPOSProxy proxy (font->face);
+  if (!buffer->message (font, "start table GPOS")) return;
+  apply (proxy, plan, font, buffer);
+  (void)buffer->message (font, "end table GPOS");
+}
+
+void
+hb_ot_layout_substitute_lookup (OT::hb_ot_apply_context_t *c,
+				const OT::SubstLookup &lookup,
+				const OT::hb_ot_layout_lookup_accelerator_t &accel)
+{
+  apply_string<GSUBProxy> (c, lookup, accel);
+}
+
+#ifndef HB_NO_BASE
+/**
+ * hb_ot_layout_get_baseline:
+ * @font: a font
+ * @baseline_tag: a baseline tag
+ * @direction: text direction.
+ * @script_tag:  script tag.
+ * @language_tag: language tag.
+ * @coord: (out): baseline value if found.
+ *
+ * Fetches a baseline value from the face.
+ *
+ * Return value: if found baseline value in the font.
+ *
+ * Since: 2.6.0
+ **/
+hb_bool_t
+hb_ot_layout_get_baseline (hb_font_t                   *font,
+			   hb_ot_layout_baseline_tag_t  baseline_tag,
+			   hb_direction_t               direction,
+			   hb_tag_t                     script_tag,
+			   hb_tag_t                     language_tag,
+			   hb_position_t               *coord        /* OUT.  May be NULL. */)
+{
+  bool result = font->face->table.BASE->get_baseline (font, baseline_tag, direction, script_tag, language_tag, coord);
+
+  if (result && coord)
+    *coord = HB_DIRECTION_IS_HORIZONTAL (direction) ? font->em_scale_y (*coord) : font->em_scale_x (*coord);
+
+  return result;
+}
+#endif
+
+
+struct hb_get_glyph_alternates_dispatch_t :
+       hb_dispatch_context_t<hb_get_glyph_alternates_dispatch_t, unsigned>
+{
+  static return_t default_return_value () { return 0; }
+  bool stop_sublookup_iteration (return_t r) const { return r; }
+
+  hb_face_t *face;
+
+  hb_get_glyph_alternates_dispatch_t (hb_face_t *face) :
+					face (face) {}
+
+  private:
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<1>, Ts&&... ds) HB_AUTO_RETURN
+  ( obj.get_glyph_alternates (hb_forward<Ts> (ds)...) )
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<0>, Ts&&... ds) HB_AUTO_RETURN
+  ( default_return_value () )
+  public:
+  template <typename T, typename ...Ts> auto
+  dispatch (const T &obj, Ts&&... ds) HB_AUTO_RETURN
+  ( _dispatch (obj, hb_prioritize, hb_forward<Ts> (ds)...) )
+};
+
+/**
+ * hb_ot_layout_lookup_get_glyph_alternates:
+ * @face: a face.
+ * @lookup_index: index of the feature lookup to query.
+ * @glyph: a glyph id.
+ * @start_offset: starting offset.
+ * @alternate_count: (inout) (allow-none): Input = the maximum number of alternate glyphs to return;
+ *                   Output = the actual number of alternate glyphs returned (may be zero).
+ * @alternate_glyphs: (out caller-allocates) (array length=alternate_count): A glyphs buffer.
+ *                    Alternate glyphs associated with the glyph id.
+ *
+ * Fetches alternates of a glyph from a given GSUB lookup index.
+ *
+ * Return value: total number of alternates found in the specific lookup index for the given glyph id.
+ *
+ * Since: 2.6.8
+ **/
+HB_EXTERN unsigned
+hb_ot_layout_lookup_get_glyph_alternates (hb_face_t      *face,
+					  unsigned        lookup_index,
+					  hb_codepoint_t  glyph,
+					  unsigned        start_offset,
+					  unsigned       *alternate_count  /* IN/OUT.  May be NULL. */,
+					  hb_codepoint_t *alternate_glyphs /* OUT.     May be NULL. */)
+{
+  hb_get_glyph_alternates_dispatch_t c (face);
+  const OT::SubstLookup &lookup = face->table.GSUB->table->get_lookup (lookup_index);
+  auto ret = lookup.dispatch (&c, glyph, start_offset, alternate_count, alternate_glyphs);
+  if (!ret && alternate_count) *alternate_count = 0;
+  return ret;
+}
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout.h b/thirdparty/harfbuzz/src/hb-ot-layout.h
new file mode 100644
index 0000000000..545d5f7fc4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout.h
@@ -0,0 +1,462 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_LAYOUT_H
+#define HB_OT_LAYOUT_H
+
+#include "hb.h"
+
+#include "hb-ot-name.h"
+
+HB_BEGIN_DECLS
+
+
+#define HB_OT_TAG_BASE HB_TAG('B','A','S','E')
+#define HB_OT_TAG_GDEF HB_TAG('G','D','E','F')
+#define HB_OT_TAG_GSUB HB_TAG('G','S','U','B')
+#define HB_OT_TAG_GPOS HB_TAG('G','P','O','S')
+#define HB_OT_TAG_JSTF HB_TAG('J','S','T','F')
+
+
+/*
+ * Script & Language tags.
+ */
+
+#define HB_OT_TAG_DEFAULT_SCRIPT	HB_TAG ('D', 'F', 'L', 'T')
+#define HB_OT_TAG_DEFAULT_LANGUAGE	HB_TAG ('d', 'f', 'l', 't')
+
+/**
+ * HB_OT_MAX_TAGS_PER_SCRIPT:
+ *
+ * Since: 2.0.0
+ **/
+#define HB_OT_MAX_TAGS_PER_SCRIPT	3u
+/**
+ * HB_OT_MAX_TAGS_PER_LANGUAGE:
+ *
+ * Since: 2.0.0
+ **/
+#define HB_OT_MAX_TAGS_PER_LANGUAGE	3u
+
+HB_EXTERN void
+hb_ot_tags_from_script_and_language (hb_script_t   script,
+				     hb_language_t language,
+				     unsigned int *script_count /* IN/OUT */,
+				     hb_tag_t     *script_tags /* OUT */,
+				     unsigned int *language_count /* IN/OUT */,
+				     hb_tag_t     *language_tags /* OUT */);
+
+HB_EXTERN hb_script_t
+hb_ot_tag_to_script (hb_tag_t tag);
+
+HB_EXTERN hb_language_t
+hb_ot_tag_to_language (hb_tag_t tag);
+
+HB_EXTERN void
+hb_ot_tags_to_script_and_language (hb_tag_t       script_tag,
+				   hb_tag_t       language_tag,
+				   hb_script_t   *script /* OUT */,
+				   hb_language_t *language /* OUT */);
+
+
+/*
+ * GDEF
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_has_glyph_classes (hb_face_t *face);
+
+/**
+ * hb_ot_layout_glyph_class_t:
+ * @HB_OT_LAYOUT_GLYPH_CLASS_UNCLASSIFIED: Glyphs not matching the other classifications
+ * @HB_OT_LAYOUT_GLYPH_CLASS_BASE_GLYPH: Spacing, single characters, capable of accepting marks
+ * @HB_OT_LAYOUT_GLYPH_CLASS_LIGATURE: Glyphs that represent ligation of multiple characters
+ * @HB_OT_LAYOUT_GLYPH_CLASS_MARK: Non-spacing, combining glyphs that represent marks
+ * @HB_OT_LAYOUT_GLYPH_CLASS_COMPONENT: Spacing glyphs that represent part of a single character
+ *
+ * The GDEF classes defined for glyphs.
+ *
+ **/
+typedef enum {
+  HB_OT_LAYOUT_GLYPH_CLASS_UNCLASSIFIED	= 0,
+  HB_OT_LAYOUT_GLYPH_CLASS_BASE_GLYPH	= 1,
+  HB_OT_LAYOUT_GLYPH_CLASS_LIGATURE	= 2,
+  HB_OT_LAYOUT_GLYPH_CLASS_MARK		= 3,
+  HB_OT_LAYOUT_GLYPH_CLASS_COMPONENT	= 4
+} hb_ot_layout_glyph_class_t;
+
+HB_EXTERN hb_ot_layout_glyph_class_t
+hb_ot_layout_get_glyph_class (hb_face_t      *face,
+			      hb_codepoint_t  glyph);
+
+HB_EXTERN void
+hb_ot_layout_get_glyphs_in_class (hb_face_t                  *face,
+				  hb_ot_layout_glyph_class_t  klass,
+				  hb_set_t                   *glyphs /* OUT */);
+
+/* Not that useful.  Provides list of attach points for a glyph that a
+ * client may want to cache */
+HB_EXTERN unsigned int
+hb_ot_layout_get_attach_points (hb_face_t      *face,
+				hb_codepoint_t  glyph,
+				unsigned int    start_offset,
+				unsigned int   *point_count /* IN/OUT */,
+				unsigned int   *point_array /* OUT */);
+
+/* Ligature caret positions */
+HB_EXTERN unsigned int
+hb_ot_layout_get_ligature_carets (hb_font_t      *font,
+				  hb_direction_t  direction,
+				  hb_codepoint_t  glyph,
+				  unsigned int    start_offset,
+				  unsigned int   *caret_count /* IN/OUT */,
+				  hb_position_t  *caret_array /* OUT */);
+
+
+/*
+ * GSUB/GPOS feature query and enumeration interface
+ */
+
+#define HB_OT_LAYOUT_NO_SCRIPT_INDEX		0xFFFFu
+#define HB_OT_LAYOUT_NO_FEATURE_INDEX		0xFFFFu
+#define HB_OT_LAYOUT_DEFAULT_LANGUAGE_INDEX	0xFFFFu
+#define HB_OT_LAYOUT_NO_VARIATIONS_INDEX	0xFFFFFFFFu
+
+HB_EXTERN unsigned int
+hb_ot_layout_table_get_script_tags (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  start_offset,
+				    unsigned int *script_count /* IN/OUT */,
+				    hb_tag_t     *script_tags /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_table_find_script (hb_face_t    *face,
+				hb_tag_t      table_tag,
+				hb_tag_t      script_tag,
+				unsigned int *script_index /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_table_select_script (hb_face_t      *face,
+				  hb_tag_t        table_tag,
+				  unsigned int    script_count,
+				  const hb_tag_t *script_tags,
+				  unsigned int   *script_index /* OUT */,
+				  hb_tag_t       *chosen_script /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_table_get_feature_tags (hb_face_t    *face,
+				     hb_tag_t      table_tag,
+				     unsigned int  start_offset,
+				     unsigned int *feature_count /* IN/OUT */,
+				     hb_tag_t     *feature_tags /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_script_get_language_tags (hb_face_t    *face,
+				       hb_tag_t      table_tag,
+				       unsigned int  script_index,
+				       unsigned int  start_offset,
+				       unsigned int *language_count /* IN/OUT */,
+				       hb_tag_t     *language_tags /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_script_select_language (hb_face_t      *face,
+				     hb_tag_t        table_tag,
+				     unsigned int    script_index,
+				     unsigned int    language_count,
+				     const hb_tag_t *language_tags,
+				     unsigned int   *language_index /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_language_get_required_feature_index (hb_face_t    *face,
+						  hb_tag_t      table_tag,
+						  unsigned int  script_index,
+						  unsigned int  language_index,
+						  unsigned int *feature_index /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_language_get_required_feature (hb_face_t    *face,
+					    hb_tag_t      table_tag,
+					    unsigned int  script_index,
+					    unsigned int  language_index,
+					    unsigned int *feature_index /* OUT */,
+					    hb_tag_t     *feature_tag /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_language_get_feature_indexes (hb_face_t    *face,
+					   hb_tag_t      table_tag,
+					   unsigned int  script_index,
+					   unsigned int  language_index,
+					   unsigned int  start_offset,
+					   unsigned int *feature_count /* IN/OUT */,
+					   unsigned int *feature_indexes /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_language_get_feature_tags (hb_face_t    *face,
+					hb_tag_t      table_tag,
+					unsigned int  script_index,
+					unsigned int  language_index,
+					unsigned int  start_offset,
+					unsigned int *feature_count /* IN/OUT */,
+					hb_tag_t     *feature_tags /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_language_find_feature (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  script_index,
+				    unsigned int  language_index,
+				    hb_tag_t      feature_tag,
+				    unsigned int *feature_index /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_feature_get_lookups (hb_face_t    *face,
+				  hb_tag_t      table_tag,
+				  unsigned int  feature_index,
+				  unsigned int  start_offset,
+				  unsigned int *lookup_count /* IN/OUT */,
+				  unsigned int *lookup_indexes /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_table_get_lookup_count (hb_face_t    *face,
+				     hb_tag_t      table_tag);
+
+HB_EXTERN void
+hb_ot_layout_collect_features (hb_face_t      *face,
+			       hb_tag_t        table_tag,
+			       const hb_tag_t *scripts,
+			       const hb_tag_t *languages,
+			       const hb_tag_t *features,
+			       hb_set_t       *feature_indexes /* OUT */);
+
+HB_EXTERN void
+hb_ot_layout_collect_lookups (hb_face_t      *face,
+			      hb_tag_t        table_tag,
+			      const hb_tag_t *scripts,
+			      const hb_tag_t *languages,
+			      const hb_tag_t *features,
+			      hb_set_t       *lookup_indexes /* OUT */);
+
+HB_EXTERN void
+hb_ot_layout_lookup_collect_glyphs (hb_face_t    *face,
+				    hb_tag_t      table_tag,
+				    unsigned int  lookup_index,
+				    hb_set_t     *glyphs_before, /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_input,  /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_after,  /* OUT.  May be NULL */
+				    hb_set_t     *glyphs_output  /* OUT.  May be NULL */);
+
+#ifdef HB_NOT_IMPLEMENTED
+typedef struct
+{
+  const hb_codepoint_t *before,
+  unsigned int          before_length,
+  const hb_codepoint_t *input,
+  unsigned int          input_length,
+  const hb_codepoint_t *after,
+  unsigned int          after_length,
+} hb_ot_layout_glyph_sequence_t;
+
+typedef hb_bool_t
+(*hb_ot_layout_glyph_sequence_func_t) (hb_font_t    *font,
+				       hb_tag_t      table_tag,
+				       unsigned int  lookup_index,
+				       const hb_ot_layout_glyph_sequence_t *sequence,
+				       void         *user_data);
+
+HB_EXTERN void
+Xhb_ot_layout_lookup_enumerate_sequences (hb_face_t    *face,
+					 hb_tag_t      table_tag,
+					 unsigned int  lookup_index,
+					 hb_ot_layout_glyph_sequence_func_t callback,
+					 void         *user_data);
+#endif
+
+/* Variations support */
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_table_find_feature_variations (hb_face_t    *face,
+					    hb_tag_t      table_tag,
+					    const int    *coords,
+					    unsigned int  num_coords,
+					    unsigned int *variations_index /* out */);
+
+HB_EXTERN unsigned int
+hb_ot_layout_feature_with_variations_get_lookups (hb_face_t    *face,
+						  hb_tag_t      table_tag,
+						  unsigned int  feature_index,
+						  unsigned int  variations_index,
+						  unsigned int  start_offset,
+						  unsigned int *lookup_count /* IN/OUT */,
+						  unsigned int *lookup_indexes /* OUT */);
+
+
+/*
+ * GSUB
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_has_substitution (hb_face_t *face);
+
+HB_EXTERN unsigned
+hb_ot_layout_lookup_get_glyph_alternates (hb_face_t      *face,
+					  unsigned        lookup_index,
+					  hb_codepoint_t  glyph,
+					  unsigned        start_offset,
+					  unsigned       *alternate_count /* IN/OUT */,
+					  hb_codepoint_t *alternate_glyphs /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_lookup_would_substitute (hb_face_t            *face,
+				      unsigned int          lookup_index,
+				      const hb_codepoint_t *glyphs,
+				      unsigned int          glyphs_length,
+				      hb_bool_t             zero_context);
+
+HB_EXTERN void
+hb_ot_layout_lookup_substitute_closure (hb_face_t    *face,
+					unsigned int  lookup_index,
+					hb_set_t     *glyphs
+					/*TODO , hb_bool_t  inclusive */);
+
+HB_EXTERN void
+hb_ot_layout_lookups_substitute_closure (hb_face_t      *face,
+					 const hb_set_t *lookups,
+					 hb_set_t       *glyphs);
+
+
+#ifdef HB_NOT_IMPLEMENTED
+/* Note: You better have GDEF when using this API, or marks won't do much. */
+HB_EXTERN hb_bool_t
+Xhb_ot_layout_lookup_substitute (hb_font_t            *font,
+				unsigned int          lookup_index,
+				const hb_ot_layout_glyph_sequence_t *sequence,
+				unsigned int          out_size,
+				hb_codepoint_t       *glyphs_out,   /* OUT */
+				unsigned int         *clusters_out, /* OUT */
+				unsigned int         *out_length    /* OUT */);
+#endif
+
+
+/*
+ * GPOS
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_has_positioning (hb_face_t *face);
+
+#ifdef HB_NOT_IMPLEMENTED
+/* Note: You better have GDEF when using this API, or marks won't do much. */
+HB_EXTERN hb_bool_t
+Xhb_ot_layout_lookup_position (hb_font_t            *font,
+			      unsigned int          lookup_index,
+			      const hb_ot_layout_glyph_sequence_t *sequence,
+			      hb_glyph_position_t  *positions /* IN / OUT */);
+#endif
+
+/* Optical 'size' feature info.  Returns true if found.
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/features_pt#size */
+HB_EXTERN hb_bool_t
+hb_ot_layout_get_size_params (hb_face_t       *face,
+			      unsigned int    *design_size,       /* OUT.  May be NULL */
+			      unsigned int    *subfamily_id,      /* OUT.  May be NULL */
+			      hb_ot_name_id_t *subfamily_name_id, /* OUT.  May be NULL */
+			      unsigned int    *range_start,       /* OUT.  May be NULL */
+			      unsigned int    *range_end          /* OUT.  May be NULL */);
+
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_feature_get_name_ids (hb_face_t       *face,
+				   hb_tag_t         table_tag,
+				   unsigned int     feature_index,
+				   hb_ot_name_id_t *label_id             /* OUT.  May be NULL */,
+				   hb_ot_name_id_t *tooltip_id           /* OUT.  May be NULL */,
+				   hb_ot_name_id_t *sample_id            /* OUT.  May be NULL */,
+				   unsigned int    *num_named_parameters /* OUT.  May be NULL */,
+				   hb_ot_name_id_t *first_param_id       /* OUT.  May be NULL */);
+
+
+HB_EXTERN unsigned int
+hb_ot_layout_feature_get_characters (hb_face_t      *face,
+				     hb_tag_t        table_tag,
+				     unsigned int    feature_index,
+				     unsigned int    start_offset,
+				     unsigned int   *char_count    /* IN/OUT.  May be NULL */,
+				     hb_codepoint_t *characters    /* OUT.     May be NULL */);
+
+/*
+ * BASE
+ */
+
+/**
+ * hb_ot_layout_baseline_tag_t:
+ * @HB_OT_LAYOUT_BASELINE_TAG_ROMAN: The baseline used by alphabetic scripts such as Latin, Cyrillic and Greek.
+ * In vertical writing mode, the alphabetic baseline for characters rotated 90 degrees clockwise.
+ * (This would not apply to alphabetic characters that remain upright in vertical writing mode, since these
+ * characters are not rotated.)
+ * @HB_OT_LAYOUT_BASELINE_TAG_HANGING: The hanging baseline. In horizontal direction, this is the horizontal
+ * line from which syllables seem, to hang in Tibetan and other similar scripts. In vertical writing mode,
+ * for Tibetan (or some other similar script) characters rotated 90 degrees clockwise.
+ * @HB_OT_LAYOUT_BASELINE_TAG_IDEO_FACE_BOTTOM_OR_LEFT: Ideographic character face bottom or left edge,
+ * if the direction is horizontal or vertical, respectively.
+ * @HB_OT_LAYOUT_BASELINE_TAG_IDEO_FACE_TOP_OR_RIGHT: Ideographic character face top or right edge,
+ * if the direction is horizontal or vertical, respectively.
+ * @HB_OT_LAYOUT_BASELINE_TAG_IDEO_EMBOX_BOTTOM_OR_LEFT: Ideographic em-box bottom or left edge,
+ * if the direction is horizontal or vertical, respectively.
+ * @HB_OT_LAYOUT_BASELINE_TAG_IDEO_EMBOX_TOP_OR_RIGHT: Ideographic em-box top or right edge baseline,
+ * if the direction is horizontal or vertical, respectively.
+ * @HB_OT_LAYOUT_BASELINE_TAG_MATH: The baseline about which mathematical characters are centered.
+ * In vertical writing mode when mathematical characters rotated 90 degrees clockwise, are centered.
+ *
+ * Baseline tags from https://docs.microsoft.com/en-us/typography/opentype/spec/baselinetags
+ *
+ * Since: 2.6.0
+ */
+typedef enum {
+  HB_OT_LAYOUT_BASELINE_TAG_ROMAN			= HB_TAG ('r','o','m','n'),
+  HB_OT_LAYOUT_BASELINE_TAG_HANGING			= HB_TAG ('h','a','n','g'),
+  HB_OT_LAYOUT_BASELINE_TAG_IDEO_FACE_BOTTOM_OR_LEFT	= HB_TAG ('i','c','f','b'),
+  HB_OT_LAYOUT_BASELINE_TAG_IDEO_FACE_TOP_OR_RIGHT	= HB_TAG ('i','c','f','t'),
+  HB_OT_LAYOUT_BASELINE_TAG_IDEO_EMBOX_BOTTOM_OR_LEFT	= HB_TAG ('i','d','e','o'),
+  HB_OT_LAYOUT_BASELINE_TAG_IDEO_EMBOX_TOP_OR_RIGHT	= HB_TAG ('i','d','t','p'),
+  HB_OT_LAYOUT_BASELINE_TAG_MATH			= HB_TAG ('m','a','t','h'),
+
+  _HB_OT_LAYOUT_BASELINE_TAG_MAX_VALUE = HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_ot_layout_baseline_tag_t;
+
+HB_EXTERN hb_bool_t
+hb_ot_layout_get_baseline (hb_font_t                   *font,
+			   hb_ot_layout_baseline_tag_t  baseline_tag,
+			   hb_direction_t               direction,
+			   hb_tag_t                     script_tag,
+			   hb_tag_t                     language_tag,
+			   hb_position_t               *coord        /* OUT.  May be NULL. */);
+
+HB_END_DECLS
+
+#endif /* HB_OT_LAYOUT_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-layout.hh b/thirdparty/harfbuzz/src/hb-ot-layout.hh
new file mode 100644
index 0000000000..f3bb15581a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-layout.hh
@@ -0,0 +1,627 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_LAYOUT_HH
+#define HB_OT_LAYOUT_HH
+
+#include "hb.hh"
+
+#include "hb-font.hh"
+#include "hb-buffer.hh"
+#include "hb-open-type.hh"
+#include "hb-ot-shape.hh"
+#include "hb-set-digest.hh"
+
+
+struct hb_ot_shape_plan_t;
+
+
+/*
+ * kern
+ */
+
+HB_INTERNAL bool
+hb_ot_layout_has_kerning (hb_face_t *face);
+
+HB_INTERNAL bool
+hb_ot_layout_has_machine_kerning (hb_face_t *face);
+
+HB_INTERNAL bool
+hb_ot_layout_has_cross_kerning (hb_face_t *face);
+
+HB_INTERNAL void
+hb_ot_layout_kern (const hb_ot_shape_plan_t *plan,
+		   hb_font_t *font,
+		   hb_buffer_t  *buffer);
+
+
+/* Private API corresponding to hb-ot-layout.h: */
+
+HB_INTERNAL bool
+hb_ot_layout_table_find_feature (hb_face_t    *face,
+				 hb_tag_t      table_tag,
+				 hb_tag_t      feature_tag,
+				 unsigned int *feature_index);
+
+
+/*
+ * GDEF
+ */
+
+enum hb_ot_layout_glyph_props_flags_t
+{
+  /* The following three match LookupFlags::Ignore* numbers. */
+  HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH	= 0x02u,
+  HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE	= 0x04u,
+  HB_OT_LAYOUT_GLYPH_PROPS_MARK		= 0x08u,
+
+  /* The following are used internally; not derived from GDEF. */
+  HB_OT_LAYOUT_GLYPH_PROPS_SUBSTITUTED	= 0x10u,
+  HB_OT_LAYOUT_GLYPH_PROPS_LIGATED	= 0x20u,
+  HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED	= 0x40u,
+
+  HB_OT_LAYOUT_GLYPH_PROPS_PRESERVE     = HB_OT_LAYOUT_GLYPH_PROPS_SUBSTITUTED |
+					  HB_OT_LAYOUT_GLYPH_PROPS_LIGATED |
+					  HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED
+};
+HB_MARK_AS_FLAG_T (hb_ot_layout_glyph_props_flags_t);
+
+
+/*
+ * GSUB/GPOS
+ */
+
+
+/* Should be called before all the substitute_lookup's are done. */
+HB_INTERNAL void
+hb_ot_layout_substitute_start (hb_font_t    *font,
+			       hb_buffer_t  *buffer);
+
+HB_INTERNAL void
+hb_ot_layout_delete_glyphs_inplace (hb_buffer_t *buffer,
+				    bool (*filter) (const hb_glyph_info_t *info));
+
+namespace OT {
+  struct hb_ot_apply_context_t;
+  struct SubstLookup;
+  struct hb_ot_layout_lookup_accelerator_t;
+}
+
+HB_INTERNAL void
+hb_ot_layout_substitute_lookup (OT::hb_ot_apply_context_t *c,
+				const OT::SubstLookup &lookup,
+				const OT::hb_ot_layout_lookup_accelerator_t &accel);
+
+
+/* Should be called before all the position_lookup's are done. */
+HB_INTERNAL void
+hb_ot_layout_position_start (hb_font_t    *font,
+			     hb_buffer_t  *buffer);
+
+/* Should be called after all the position_lookup's are done, to fini advances. */
+HB_INTERNAL void
+hb_ot_layout_position_finish_advances (hb_font_t    *font,
+				       hb_buffer_t  *buffer);
+
+/* Should be called after hb_ot_layout_position_finish_advances, to fini offsets. */
+HB_INTERNAL void
+hb_ot_layout_position_finish_offsets (hb_font_t    *font,
+				      hb_buffer_t  *buffer);
+
+
+/*
+ * Buffer var routines.
+ */
+
+/* buffer var allocations, used during the entire shaping process */
+#define unicode_props()		var2.u16[0]
+
+/* buffer var allocations, used during the GSUB/GPOS processing */
+#define glyph_props()		var1.u16[0] /* GDEF glyph properties */
+#define lig_props()		var1.u8[2] /* GSUB/GPOS ligature tracking */
+#define syllable()		var1.u8[3] /* GSUB/GPOS shaping boundaries */
+
+
+/* Loop over syllables. Based on foreach_cluster(). */
+#define foreach_syllable(buffer, start, end) \
+  for (unsigned int \
+       _count = buffer->len, \
+       start = 0, end = _count ? _hb_next_syllable (buffer, 0) : 0; \
+       start < _count; \
+       start = end, end = _hb_next_syllable (buffer, start))
+
+static inline unsigned int
+_hb_next_syllable (hb_buffer_t *buffer, unsigned int start)
+{
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+
+  unsigned int syllable = info[start].syllable();
+  while (++start < count && syllable == info[start].syllable())
+    ;
+
+  return start;
+}
+
+static inline void
+_hb_clear_syllables (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		     hb_font_t *font HB_UNUSED,
+		     hb_buffer_t *buffer)
+{
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+    info[i].syllable() = 0;
+}
+
+
+/* unicode_props */
+
+/* Design:
+ * unicode_props() is a two-byte number.  The low byte includes:
+ * - General_Category: 5 bits.
+ * - A bit each for:
+ *   * Is it Default_Ignorable(); we have a modified Default_Ignorable().
+ *   * Whether it's one of the three Mongolian Free Variation Selectors,
+ *     CGJ, or other characters that are hidden but should not be ignored
+ *     like most other Default_Ignorable()s do during matching.
+ *   * Whether it's a grapheme continuation.
+ *
+ * The high-byte has different meanings, switched by the Gen-Cat:
+ * - For Mn,Mc,Me: the modified Combining_Class.
+ * - For Cf: whether it's ZWJ, ZWNJ, or something else.
+ * - For Ws: index of which space character this is, if space fallback
+ *   is needed, ie. we don't set this by default, only if asked to.
+ */
+
+enum hb_unicode_props_flags_t {
+  UPROPS_MASK_GEN_CAT	= 0x001Fu,
+  UPROPS_MASK_IGNORABLE	= 0x0020u,
+  UPROPS_MASK_HIDDEN	= 0x0040u, /* MONGOLIAN FREE VARIATION SELECTOR 1..3, or TAG characters */
+  UPROPS_MASK_CONTINUATION=0x0080u,
+
+  /* If GEN_CAT=FORMAT, top byte masks: */
+  UPROPS_MASK_Cf_ZWJ	= 0x0100u,
+  UPROPS_MASK_Cf_ZWNJ	= 0x0200u
+};
+HB_MARK_AS_FLAG_T (hb_unicode_props_flags_t);
+
+static inline void
+_hb_glyph_info_set_unicode_props (hb_glyph_info_t *info, hb_buffer_t *buffer)
+{
+  hb_unicode_funcs_t *unicode = buffer->unicode;
+  unsigned int u = info->codepoint;
+  unsigned int gen_cat = (unsigned int) unicode->general_category (u);
+  unsigned int props = gen_cat;
+
+  if (u >= 0x80u)
+  {
+    buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_NON_ASCII;
+
+    if (unlikely (unicode->is_default_ignorable (u)))
+    {
+      buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_DEFAULT_IGNORABLES;
+      props |=  UPROPS_MASK_IGNORABLE;
+      if (u == 0x200Cu) props |= UPROPS_MASK_Cf_ZWNJ;
+      else if (u == 0x200Du) props |= UPROPS_MASK_Cf_ZWJ;
+      /* Mongolian Free Variation Selectors need to be remembered
+       * because although we need to hide them like default-ignorables,
+       * they need to non-ignorable during shaping.  This is similar to
+       * what we do for joiners in Indic-like shapers, but since the
+       * FVSes are GC=Mn, we have use a separate bit to remember them.
+       * Fixes:
+       * https://github.com/harfbuzz/harfbuzz/issues/234 */
+      else if (unlikely (hb_in_range<hb_codepoint_t> (u, 0x180Bu, 0x180Du))) props |= UPROPS_MASK_HIDDEN;
+      /* TAG characters need similar treatment. Fixes:
+       * https://github.com/harfbuzz/harfbuzz/issues/463 */
+      else if (unlikely (hb_in_range<hb_codepoint_t> (u, 0xE0020u, 0xE007Fu))) props |= UPROPS_MASK_HIDDEN;
+      /* COMBINING GRAPHEME JOINER should not be skipped; at least some times.
+       * https://github.com/harfbuzz/harfbuzz/issues/554 */
+      else if (unlikely (u == 0x034Fu))
+      {
+	buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_CGJ;
+	props |= UPROPS_MASK_HIDDEN;
+      }
+    }
+
+    if (unlikely (HB_UNICODE_GENERAL_CATEGORY_IS_MARK (gen_cat)))
+    {
+      props |= UPROPS_MASK_CONTINUATION;
+      props |= unicode->modified_combining_class (u)<<8;
+    }
+  }
+
+  info->unicode_props() = props;
+}
+
+static inline void
+_hb_glyph_info_set_general_category (hb_glyph_info_t *info,
+				     hb_unicode_general_category_t gen_cat)
+{
+  /* Clears top-byte. */
+  info->unicode_props() = (unsigned int) gen_cat | (info->unicode_props() & (0xFF & ~UPROPS_MASK_GEN_CAT));
+}
+
+static inline hb_unicode_general_category_t
+_hb_glyph_info_get_general_category (const hb_glyph_info_t *info)
+{
+  return (hb_unicode_general_category_t) (info->unicode_props() & UPROPS_MASK_GEN_CAT);
+}
+
+static inline bool
+_hb_glyph_info_is_unicode_mark (const hb_glyph_info_t *info)
+{
+  return HB_UNICODE_GENERAL_CATEGORY_IS_MARK (info->unicode_props() & UPROPS_MASK_GEN_CAT);
+}
+static inline void
+_hb_glyph_info_set_modified_combining_class (hb_glyph_info_t *info,
+					     unsigned int modified_class)
+{
+  if (unlikely (!_hb_glyph_info_is_unicode_mark (info)))
+    return;
+  info->unicode_props() = (modified_class<<8) | (info->unicode_props() & 0xFF);
+}
+static inline unsigned int
+_hb_glyph_info_get_modified_combining_class (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_is_unicode_mark (info) ? info->unicode_props()>>8 : 0;
+}
+#define info_cc(info) (_hb_glyph_info_get_modified_combining_class (&(info)))
+
+static inline bool
+_hb_glyph_info_is_unicode_space (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_get_general_category (info) ==
+	 HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR;
+}
+static inline void
+_hb_glyph_info_set_unicode_space_fallback_type (hb_glyph_info_t *info, hb_unicode_funcs_t::space_t s)
+{
+  if (unlikely (!_hb_glyph_info_is_unicode_space (info)))
+    return;
+  info->unicode_props() = (((unsigned int) s)<<8) | (info->unicode_props() & 0xFF);
+}
+static inline hb_unicode_funcs_t::space_t
+_hb_glyph_info_get_unicode_space_fallback_type (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_is_unicode_space (info) ?
+	 (hb_unicode_funcs_t::space_t) (info->unicode_props()>>8) :
+	 hb_unicode_funcs_t::NOT_SPACE;
+}
+
+static inline bool _hb_glyph_info_ligated (const hb_glyph_info_t *info);
+
+static inline bool
+_hb_glyph_info_is_default_ignorable (const hb_glyph_info_t *info)
+{
+  return (info->unicode_props() & UPROPS_MASK_IGNORABLE) &&
+	 !_hb_glyph_info_ligated (info);
+}
+static inline bool
+_hb_glyph_info_is_default_ignorable_and_not_hidden (const hb_glyph_info_t *info)
+{
+  return ((info->unicode_props() & (UPROPS_MASK_IGNORABLE|UPROPS_MASK_HIDDEN))
+	  == UPROPS_MASK_IGNORABLE) &&
+	 !_hb_glyph_info_ligated (info);
+}
+static inline void
+_hb_glyph_info_unhide (hb_glyph_info_t *info)
+{
+  info->unicode_props() &= ~ UPROPS_MASK_HIDDEN;
+}
+
+static inline void
+_hb_glyph_info_set_continuation (hb_glyph_info_t *info)
+{
+  info->unicode_props() |= UPROPS_MASK_CONTINUATION;
+}
+static inline void
+_hb_glyph_info_reset_continuation (hb_glyph_info_t *info)
+{
+  info->unicode_props() &= ~ UPROPS_MASK_CONTINUATION;
+}
+static inline bool
+_hb_glyph_info_is_continuation (const hb_glyph_info_t *info)
+{
+  return info->unicode_props() & UPROPS_MASK_CONTINUATION;
+}
+/* Loop over grapheme. Based on foreach_cluster(). */
+#define foreach_grapheme(buffer, start, end) \
+  for (unsigned int \
+       _count = buffer->len, \
+       start = 0, end = _count ? _hb_next_grapheme (buffer, 0) : 0; \
+       start < _count; \
+       start = end, end = _hb_next_grapheme (buffer, start))
+
+static inline unsigned int
+_hb_next_grapheme (hb_buffer_t *buffer, unsigned int start)
+{
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+
+  while (++start < count && _hb_glyph_info_is_continuation (&info[start]))
+    ;
+
+  return start;
+}
+
+static inline bool
+_hb_glyph_info_is_unicode_format (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_get_general_category (info) ==
+	 HB_UNICODE_GENERAL_CATEGORY_FORMAT;
+}
+static inline bool
+_hb_glyph_info_is_zwnj (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_is_unicode_format (info) && (info->unicode_props() & UPROPS_MASK_Cf_ZWNJ);
+}
+static inline bool
+_hb_glyph_info_is_zwj (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_is_unicode_format (info) && (info->unicode_props() & UPROPS_MASK_Cf_ZWJ);
+}
+static inline bool
+_hb_glyph_info_is_joiner (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_is_unicode_format (info) && (info->unicode_props() & (UPROPS_MASK_Cf_ZWNJ|UPROPS_MASK_Cf_ZWJ));
+}
+static inline void
+_hb_glyph_info_flip_joiners (hb_glyph_info_t *info)
+{
+  if (!_hb_glyph_info_is_unicode_format (info))
+    return;
+  info->unicode_props() ^= UPROPS_MASK_Cf_ZWNJ | UPROPS_MASK_Cf_ZWJ;
+}
+
+/* lig_props: aka lig_id / lig_comp
+ *
+ * When a ligature is formed:
+ *
+ *   - The ligature glyph and any marks in between all the same newly allocated
+ *     lig_id,
+ *   - The ligature glyph will get lig_num_comps set to the number of components
+ *   - The marks get lig_comp > 0, reflecting which component of the ligature
+ *     they were applied to.
+ *   - This is used in GPOS to attach marks to the right component of a ligature
+ *     in MarkLigPos,
+ *   - Note that when marks are ligated together, much of the above is skipped
+ *     and the current lig_id reused.
+ *
+ * When a multiple-substitution is done:
+ *
+ *   - All resulting glyphs will have lig_id = 0,
+ *   - The resulting glyphs will have lig_comp = 0, 1, 2, ... respectively.
+ *   - This is used in GPOS to attach marks to the first component of a
+ *     multiple substitution in MarkBasePos.
+ *
+ * The numbers are also used in GPOS to do mark-to-mark positioning only
+ * to marks that belong to the same component of the same ligature.
+ */
+
+static inline void
+_hb_glyph_info_clear_lig_props (hb_glyph_info_t *info)
+{
+  info->lig_props() = 0;
+}
+
+#define IS_LIG_BASE 0x10
+
+static inline void
+_hb_glyph_info_set_lig_props_for_ligature (hb_glyph_info_t *info,
+					   unsigned int lig_id,
+					   unsigned int lig_num_comps)
+{
+  info->lig_props() = (lig_id << 5) | IS_LIG_BASE | (lig_num_comps & 0x0F);
+}
+
+static inline void
+_hb_glyph_info_set_lig_props_for_mark (hb_glyph_info_t *info,
+				       unsigned int lig_id,
+				       unsigned int lig_comp)
+{
+  info->lig_props() = (lig_id << 5) | (lig_comp & 0x0F);
+}
+
+static inline void
+_hb_glyph_info_set_lig_props_for_component (hb_glyph_info_t *info, unsigned int comp)
+{
+  _hb_glyph_info_set_lig_props_for_mark (info, 0, comp);
+}
+
+static inline unsigned int
+_hb_glyph_info_get_lig_id (const hb_glyph_info_t *info)
+{
+  return info->lig_props() >> 5;
+}
+
+static inline bool
+_hb_glyph_info_ligated_internal (const hb_glyph_info_t *info)
+{
+  return !!(info->lig_props() & IS_LIG_BASE);
+}
+
+static inline unsigned int
+_hb_glyph_info_get_lig_comp (const hb_glyph_info_t *info)
+{
+  if (_hb_glyph_info_ligated_internal (info))
+    return 0;
+  else
+    return info->lig_props() & 0x0F;
+}
+
+static inline unsigned int
+_hb_glyph_info_get_lig_num_comps (const hb_glyph_info_t *info)
+{
+  if ((info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE) &&
+      _hb_glyph_info_ligated_internal (info))
+    return info->lig_props() & 0x0F;
+  else
+    return 1;
+}
+
+static inline uint8_t
+_hb_allocate_lig_id (hb_buffer_t *buffer) {
+  uint8_t lig_id = buffer->next_serial () & 0x07;
+  if (unlikely (!lig_id))
+    lig_id = _hb_allocate_lig_id (buffer); /* in case of overflow */
+  return lig_id;
+}
+
+/* glyph_props: */
+
+static inline void
+_hb_glyph_info_set_glyph_props (hb_glyph_info_t *info, unsigned int props)
+{
+  info->glyph_props() = props;
+}
+
+static inline unsigned int
+_hb_glyph_info_get_glyph_props (const hb_glyph_info_t *info)
+{
+  return info->glyph_props();
+}
+
+static inline bool
+_hb_glyph_info_is_base_glyph (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH);
+}
+
+static inline bool
+_hb_glyph_info_is_ligature (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_LIGATURE);
+}
+
+static inline bool
+_hb_glyph_info_is_mark (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_MARK);
+}
+
+static inline bool
+_hb_glyph_info_substituted (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_SUBSTITUTED);
+}
+
+static inline bool
+_hb_glyph_info_ligated (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_LIGATED);
+}
+
+static inline bool
+_hb_glyph_info_multiplied (const hb_glyph_info_t *info)
+{
+  return !!(info->glyph_props() & HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED);
+}
+
+static inline bool
+_hb_glyph_info_ligated_and_didnt_multiply (const hb_glyph_info_t *info)
+{
+  return _hb_glyph_info_ligated (info) && !_hb_glyph_info_multiplied (info);
+}
+
+static inline void
+_hb_glyph_info_clear_ligated_and_multiplied (hb_glyph_info_t *info)
+{
+  info->glyph_props() &= ~(HB_OT_LAYOUT_GLYPH_PROPS_LIGATED |
+			   HB_OT_LAYOUT_GLYPH_PROPS_MULTIPLIED);
+}
+
+static inline void
+_hb_glyph_info_clear_substituted (hb_glyph_info_t *info)
+{
+  info->glyph_props() &= ~(HB_OT_LAYOUT_GLYPH_PROPS_SUBSTITUTED);
+}
+
+static inline void
+_hb_clear_substitution_flags (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			      hb_font_t *font HB_UNUSED,
+			      hb_buffer_t *buffer)
+{
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+    _hb_glyph_info_clear_substituted (&info[i]);
+}
+
+
+/* Allocation / deallocation. */
+
+static inline void
+_hb_buffer_allocate_unicode_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, unicode_props);
+}
+
+static inline void
+_hb_buffer_deallocate_unicode_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_DEALLOCATE_VAR (buffer, unicode_props);
+}
+
+static inline void
+_hb_buffer_assert_unicode_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_ASSERT_VAR (buffer, unicode_props);
+}
+
+static inline void
+_hb_buffer_allocate_gsubgpos_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, glyph_props);
+  HB_BUFFER_ALLOCATE_VAR (buffer, lig_props);
+  HB_BUFFER_ALLOCATE_VAR (buffer, syllable);
+}
+
+static inline void
+_hb_buffer_deallocate_gsubgpos_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_DEALLOCATE_VAR (buffer, syllable);
+  HB_BUFFER_DEALLOCATE_VAR (buffer, lig_props);
+  HB_BUFFER_DEALLOCATE_VAR (buffer, glyph_props);
+}
+
+static inline void
+_hb_buffer_assert_gsubgpos_vars (hb_buffer_t *buffer)
+{
+  HB_BUFFER_ASSERT_VAR (buffer, glyph_props);
+  HB_BUFFER_ASSERT_VAR (buffer, lig_props);
+  HB_BUFFER_ASSERT_VAR (buffer, syllable);
+}
+
+/* Make sure no one directly touches our props... */
+#undef unicode_props0
+#undef unicode_props1
+#undef lig_props
+#undef glyph_props
+
+#endif /* HB_OT_LAYOUT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-map.cc b/thirdparty/harfbuzz/src/hb-ot-map.cc
new file mode 100644
index 0000000000..e4bb4b6366
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-map.cc
@@ -0,0 +1,342 @@
+/*
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2011,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-map.hh"
+#include "hb-ot-shape.hh"
+#include "hb-ot-layout.hh"
+
+
+void hb_ot_map_t::collect_lookups (unsigned int table_index, hb_set_t *lookups_out) const
+{
+  for (unsigned int i = 0; i < lookups[table_index].length; i++)
+    lookups_out->add (lookups[table_index][i].index);
+}
+
+
+hb_ot_map_builder_t::hb_ot_map_builder_t (hb_face_t *face_,
+					  const hb_segment_properties_t *props_)
+{
+  memset (this, 0, sizeof (*this));
+
+  feature_infos.init ();
+  for (unsigned int table_index = 0; table_index < 2; table_index++)
+    stages[table_index].init ();
+
+  face = face_;
+  props = *props_;
+
+
+  /* Fetch script/language indices for GSUB/GPOS.  We need these later to skip
+   * features not available in either table and not waste precious bits for them. */
+
+  unsigned int script_count = HB_OT_MAX_TAGS_PER_SCRIPT;
+  unsigned int language_count = HB_OT_MAX_TAGS_PER_LANGUAGE;
+  hb_tag_t script_tags[HB_OT_MAX_TAGS_PER_SCRIPT];
+  hb_tag_t language_tags[HB_OT_MAX_TAGS_PER_LANGUAGE];
+
+  hb_ot_tags_from_script_and_language (props.script, props.language, &script_count, script_tags, &language_count, language_tags);
+
+  for (unsigned int table_index = 0; table_index < 2; table_index++) {
+    hb_tag_t table_tag = table_tags[table_index];
+    found_script[table_index] = (bool) hb_ot_layout_table_select_script (face, table_tag, script_count, script_tags, &script_index[table_index], &chosen_script[table_index]);
+    hb_ot_layout_script_select_language (face, table_tag, script_index[table_index], language_count, language_tags, &language_index[table_index]);
+  }
+}
+
+hb_ot_map_builder_t::~hb_ot_map_builder_t ()
+{
+  feature_infos.fini ();
+  for (unsigned int table_index = 0; table_index < 2; table_index++)
+    stages[table_index].fini ();
+}
+
+void hb_ot_map_builder_t::add_feature (hb_tag_t tag,
+				       hb_ot_map_feature_flags_t flags,
+				       unsigned int value)
+{
+  if (unlikely (!tag)) return;
+  feature_info_t *info = feature_infos.push();
+  info->tag = tag;
+  info->seq = feature_infos.length;
+  info->max_value = value;
+  info->flags = flags;
+  info->default_value = (flags & F_GLOBAL) ? value : 0;
+  info->stage[0] = current_stage[0];
+  info->stage[1] = current_stage[1];
+}
+
+void
+hb_ot_map_builder_t::add_lookups (hb_ot_map_t  &m,
+				  unsigned int  table_index,
+				  unsigned int  feature_index,
+				  unsigned int  variations_index,
+				  hb_mask_t     mask,
+				  bool          auto_zwnj,
+				  bool          auto_zwj,
+				  bool          random)
+{
+  unsigned int lookup_indices[32];
+  unsigned int offset, len;
+  unsigned int table_lookup_count;
+
+  table_lookup_count = hb_ot_layout_table_get_lookup_count (face, table_tags[table_index]);
+
+  offset = 0;
+  do {
+    len = ARRAY_LENGTH (lookup_indices);
+    hb_ot_layout_feature_with_variations_get_lookups (face,
+						      table_tags[table_index],
+						      feature_index,
+						      variations_index,
+						      offset, &len,
+						      lookup_indices);
+
+    for (unsigned int i = 0; i < len; i++)
+    {
+      if (lookup_indices[i] >= table_lookup_count)
+	continue;
+      hb_ot_map_t::lookup_map_t *lookup = m.lookups[table_index].push ();
+      lookup->mask = mask;
+      lookup->index = lookup_indices[i];
+      lookup->auto_zwnj = auto_zwnj;
+      lookup->auto_zwj = auto_zwj;
+      lookup->random = random;
+    }
+
+    offset += len;
+  } while (len == ARRAY_LENGTH (lookup_indices));
+}
+
+
+void hb_ot_map_builder_t::add_pause (unsigned int table_index, hb_ot_map_t::pause_func_t pause_func)
+{
+  stage_info_t *s = stages[table_index].push ();
+  s->index = current_stage[table_index];
+  s->pause_func = pause_func;
+
+  current_stage[table_index]++;
+}
+
+void
+hb_ot_map_builder_t::compile (hb_ot_map_t                  &m,
+			      const hb_ot_shape_plan_key_t &key)
+{
+  static_assert ((!(HB_GLYPH_FLAG_DEFINED & (HB_GLYPH_FLAG_DEFINED + 1))), "");
+  unsigned int global_bit_mask = HB_GLYPH_FLAG_DEFINED + 1;
+  unsigned int global_bit_shift = hb_popcount (HB_GLYPH_FLAG_DEFINED);
+
+  m.global_mask = global_bit_mask;
+
+  unsigned int required_feature_index[2];
+  hb_tag_t required_feature_tag[2];
+  /* We default to applying required feature in stage 0.  If the required
+   * feature has a tag that is known to the shaper, we apply required feature
+   * in the stage for that tag.
+   */
+  unsigned int required_feature_stage[2] = {0, 0};
+
+  for (unsigned int table_index = 0; table_index < 2; table_index++)
+  {
+    m.chosen_script[table_index] = chosen_script[table_index];
+    m.found_script[table_index] = found_script[table_index];
+
+    hb_ot_layout_language_get_required_feature (face,
+						table_tags[table_index],
+						script_index[table_index],
+						language_index[table_index],
+						&required_feature_index[table_index],
+						&required_feature_tag[table_index]);
+  }
+
+  /* Sort features and merge duplicates */
+  if (feature_infos.length)
+  {
+    feature_infos.qsort ();
+    unsigned int j = 0;
+    for (unsigned int i = 1; i < feature_infos.length; i++)
+      if (feature_infos[i].tag != feature_infos[j].tag)
+	feature_infos[++j] = feature_infos[i];
+      else {
+	if (feature_infos[i].flags & F_GLOBAL) {
+	  feature_infos[j].flags |= F_GLOBAL;
+	  feature_infos[j].max_value = feature_infos[i].max_value;
+	  feature_infos[j].default_value = feature_infos[i].default_value;
+	} else {
+	  if (feature_infos[j].flags & F_GLOBAL)
+	    feature_infos[j].flags ^= F_GLOBAL;
+	  feature_infos[j].max_value = hb_max (feature_infos[j].max_value, feature_infos[i].max_value);
+	  /* Inherit default_value from j */
+	}
+	feature_infos[j].flags |= (feature_infos[i].flags & F_HAS_FALLBACK);
+	feature_infos[j].stage[0] = hb_min (feature_infos[j].stage[0], feature_infos[i].stage[0]);
+	feature_infos[j].stage[1] = hb_min (feature_infos[j].stage[1], feature_infos[i].stage[1]);
+      }
+    feature_infos.shrink (j + 1);
+  }
+
+
+  /* Allocate bits now */
+  unsigned int next_bit = global_bit_shift + 1;
+
+  for (unsigned int i = 0; i < feature_infos.length; i++)
+  {
+    const feature_info_t *info = &feature_infos[i];
+
+    unsigned int bits_needed;
+
+    if ((info->flags & F_GLOBAL) && info->max_value == 1)
+      /* Uses the global bit */
+      bits_needed = 0;
+    else
+      /* Limit bits per feature. */
+      bits_needed = hb_min (HB_OT_MAP_MAX_BITS, hb_bit_storage (info->max_value));
+
+    if (!info->max_value || next_bit + bits_needed > 8 * sizeof (hb_mask_t))
+      continue; /* Feature disabled, or not enough bits. */
+
+
+    bool found = false;
+    unsigned int feature_index[2];
+    for (unsigned int table_index = 0; table_index < 2; table_index++)
+    {
+      if (required_feature_tag[table_index] == info->tag)
+	required_feature_stage[table_index] = info->stage[table_index];
+
+      found |= (bool) hb_ot_layout_language_find_feature (face,
+							  table_tags[table_index],
+							  script_index[table_index],
+							  language_index[table_index],
+							  info->tag,
+							  &feature_index[table_index]);
+    }
+    if (!found && (info->flags & F_GLOBAL_SEARCH))
+    {
+      for (unsigned int table_index = 0; table_index < 2; table_index++)
+      {
+	found |= (bool) hb_ot_layout_table_find_feature (face,
+							 table_tags[table_index],
+							 info->tag,
+							 &feature_index[table_index]);
+      }
+    }
+    if (!found && !(info->flags & F_HAS_FALLBACK))
+      continue;
+
+
+    hb_ot_map_t::feature_map_t *map = m.features.push ();
+
+    map->tag = info->tag;
+    map->index[0] = feature_index[0];
+    map->index[1] = feature_index[1];
+    map->stage[0] = info->stage[0];
+    map->stage[1] = info->stage[1];
+    map->auto_zwnj = !(info->flags & F_MANUAL_ZWNJ);
+    map->auto_zwj = !(info->flags & F_MANUAL_ZWJ);
+    map->random = !!(info->flags & F_RANDOM);
+    if ((info->flags & F_GLOBAL) && info->max_value == 1) {
+      /* Uses the global bit */
+      map->shift = global_bit_shift;
+      map->mask = global_bit_mask;
+    } else {
+      map->shift = next_bit;
+      map->mask = (1u << (next_bit + bits_needed)) - (1u << next_bit);
+      next_bit += bits_needed;
+      m.global_mask |= (info->default_value << map->shift) & map->mask;
+    }
+    map->_1_mask = (1u << map->shift) & map->mask;
+    map->needs_fallback = !found;
+
+  }
+  feature_infos.shrink (0); /* Done with these */
+
+
+  add_gsub_pause (nullptr);
+  add_gpos_pause (nullptr);
+
+  for (unsigned int table_index = 0; table_index < 2; table_index++)
+  {
+    /* Collect lookup indices for features */
+
+    unsigned int stage_index = 0;
+    unsigned int last_num_lookups = 0;
+    for (unsigned stage = 0; stage < current_stage[table_index]; stage++)
+    {
+      if (required_feature_index[table_index] != HB_OT_LAYOUT_NO_FEATURE_INDEX &&
+	  required_feature_stage[table_index] == stage)
+	add_lookups (m, table_index,
+		     required_feature_index[table_index],
+		     key.variations_index[table_index],
+		     global_bit_mask);
+
+      for (unsigned i = 0; i < m.features.length; i++)
+	if (m.features[i].stage[table_index] == stage)
+	  add_lookups (m, table_index,
+		       m.features[i].index[table_index],
+		       key.variations_index[table_index],
+		       m.features[i].mask,
+		       m.features[i].auto_zwnj,
+		       m.features[i].auto_zwj,
+		       m.features[i].random);
+
+      /* Sort lookups and merge duplicates */
+      if (last_num_lookups < m.lookups[table_index].length)
+      {
+	m.lookups[table_index].qsort (last_num_lookups, m.lookups[table_index].length);
+
+	unsigned int j = last_num_lookups;
+	for (unsigned int i = j + 1; i < m.lookups[table_index].length; i++)
+	  if (m.lookups[table_index][i].index != m.lookups[table_index][j].index)
+	    m.lookups[table_index][++j] = m.lookups[table_index][i];
+	  else
+	  {
+	    m.lookups[table_index][j].mask |= m.lookups[table_index][i].mask;
+	    m.lookups[table_index][j].auto_zwnj &= m.lookups[table_index][i].auto_zwnj;
+	    m.lookups[table_index][j].auto_zwj &= m.lookups[table_index][i].auto_zwj;
+	  }
+	m.lookups[table_index].shrink (j + 1);
+      }
+
+      last_num_lookups = m.lookups[table_index].length;
+
+      if (stage_index < stages[table_index].length && stages[table_index][stage_index].index == stage) {
+	hb_ot_map_t::stage_map_t *stage_map = m.stages[table_index].push ();
+	stage_map->last_lookup = last_num_lookups;
+	stage_map->pause_func = stages[table_index][stage_index].pause_func;
+
+	stage_index++;
+      }
+    }
+  }
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-map.hh b/thirdparty/harfbuzz/src/hb-ot-map.hh
new file mode 100644
index 0000000000..7629a869bd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-map.hh
@@ -0,0 +1,284 @@
+/*
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2011,2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_MAP_HH
+#define HB_OT_MAP_HH
+
+#include "hb-buffer.hh"
+
+
+#define HB_OT_MAP_MAX_BITS 8u
+#define HB_OT_MAP_MAX_VALUE ((1u << HB_OT_MAP_MAX_BITS) - 1u)
+
+struct hb_ot_shape_plan_t;
+
+static const hb_tag_t table_tags[2] = {HB_OT_TAG_GSUB, HB_OT_TAG_GPOS};
+
+struct hb_ot_map_t
+{
+  friend struct hb_ot_map_builder_t;
+
+  public:
+
+  struct feature_map_t {
+    hb_tag_t tag; /* should be first for our bsearch to work */
+    unsigned int index[2]; /* GSUB/GPOS */
+    unsigned int stage[2]; /* GSUB/GPOS */
+    unsigned int shift;
+    hb_mask_t mask;
+    hb_mask_t _1_mask; /* mask for value=1, for quick access */
+    unsigned int needs_fallback : 1;
+    unsigned int auto_zwnj : 1;
+    unsigned int auto_zwj : 1;
+    unsigned int random : 1;
+
+    int cmp (const hb_tag_t tag_) const
+    { return tag_ < tag ? -1 : tag_ > tag ? 1 : 0; }
+  };
+
+  struct lookup_map_t {
+    unsigned short index;
+    unsigned short auto_zwnj : 1;
+    unsigned short auto_zwj : 1;
+    unsigned short random : 1;
+    hb_mask_t mask;
+
+    HB_INTERNAL static int cmp (const void *pa, const void *pb)
+    {
+      const lookup_map_t *a = (const lookup_map_t *) pa;
+      const lookup_map_t *b = (const lookup_map_t *) pb;
+      return a->index < b->index ? -1 : a->index > b->index ? 1 : 0;
+    }
+  };
+
+  typedef void (*pause_func_t) (const struct hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer);
+
+  struct stage_map_t {
+    unsigned int last_lookup; /* Cumulative */
+    pause_func_t pause_func;
+  };
+
+  void init ()
+  {
+    memset (this, 0, sizeof (*this));
+
+    features.init ();
+    for (unsigned int table_index = 0; table_index < 2; table_index++)
+    {
+      lookups[table_index].init ();
+      stages[table_index].init ();
+    }
+  }
+  void fini ()
+  {
+    features.fini ();
+    for (unsigned int table_index = 0; table_index < 2; table_index++)
+    {
+      lookups[table_index].fini ();
+      stages[table_index].fini ();
+    }
+  }
+
+  hb_mask_t get_global_mask () const { return global_mask; }
+
+  hb_mask_t get_mask (hb_tag_t feature_tag, unsigned int *shift = nullptr) const
+  {
+    const feature_map_t *map = features.bsearch (feature_tag);
+    if (shift) *shift = map ? map->shift : 0;
+    return map ? map->mask : 0;
+  }
+
+  bool needs_fallback (hb_tag_t feature_tag) const
+  {
+    const feature_map_t *map = features.bsearch (feature_tag);
+    return map ? map->needs_fallback : false;
+  }
+
+  hb_mask_t get_1_mask (hb_tag_t feature_tag) const
+  {
+    const feature_map_t *map = features.bsearch (feature_tag);
+    return map ? map->_1_mask : 0;
+  }
+
+  unsigned int get_feature_index (unsigned int table_index, hb_tag_t feature_tag) const
+  {
+    const feature_map_t *map = features.bsearch (feature_tag);
+    return map ? map->index[table_index] : HB_OT_LAYOUT_NO_FEATURE_INDEX;
+  }
+
+  unsigned int get_feature_stage (unsigned int table_index, hb_tag_t feature_tag) const
+  {
+    const feature_map_t *map = features.bsearch (feature_tag);
+    return map ? map->stage[table_index] : UINT_MAX;
+  }
+
+  void get_stage_lookups (unsigned int table_index, unsigned int stage,
+			  const struct lookup_map_t **plookups, unsigned int *lookup_count) const
+  {
+    if (unlikely (stage == UINT_MAX)) {
+      *plookups = nullptr;
+      *lookup_count = 0;
+      return;
+    }
+    assert (stage <= stages[table_index].length);
+    unsigned int start = stage ? stages[table_index][stage - 1].last_lookup : 0;
+    unsigned int end   = stage < stages[table_index].length ? stages[table_index][stage].last_lookup : lookups[table_index].length;
+    *plookups = end == start ? nullptr : &lookups[table_index][start];
+    *lookup_count = end - start;
+  }
+
+  HB_INTERNAL void collect_lookups (unsigned int table_index, hb_set_t *lookups) const;
+  template <typename Proxy>
+  HB_INTERNAL void apply (const Proxy &proxy,
+			  const struct hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer) const;
+  HB_INTERNAL void substitute (const struct hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer) const;
+  HB_INTERNAL void position (const struct hb_ot_shape_plan_t *plan, hb_font_t *font, hb_buffer_t *buffer) const;
+
+  public:
+  hb_tag_t chosen_script[2];
+  bool found_script[2];
+
+  private:
+
+  hb_mask_t global_mask;
+
+  hb_sorted_vector_t<feature_map_t> features;
+  hb_vector_t<lookup_map_t> lookups[2]; /* GSUB/GPOS */
+  hb_vector_t<stage_map_t> stages[2]; /* GSUB/GPOS */
+};
+
+enum hb_ot_map_feature_flags_t
+{
+  F_NONE		= 0x0000u,
+  F_GLOBAL		= 0x0001u, /* Feature applies to all characters; results in no mask allocated for it. */
+  F_HAS_FALLBACK	= 0x0002u, /* Has fallback implementation, so include mask bit even if feature not found. */
+  F_MANUAL_ZWNJ		= 0x0004u, /* Don't skip over ZWNJ when matching **context**. */
+  F_MANUAL_ZWJ		= 0x0008u, /* Don't skip over ZWJ when matching **input**. */
+  F_MANUAL_JOINERS	= F_MANUAL_ZWNJ | F_MANUAL_ZWJ,
+  F_GLOBAL_MANUAL_JOINERS= F_GLOBAL | F_MANUAL_JOINERS,
+  F_GLOBAL_HAS_FALLBACK = F_GLOBAL | F_HAS_FALLBACK,
+  F_GLOBAL_SEARCH	= 0x0010u, /* If feature not found in LangSys, look for it in global feature list and pick one. */
+  F_RANDOM		= 0x0020u  /* Randomly select a glyph from an AlternateSubstFormat1 subtable. */
+};
+HB_MARK_AS_FLAG_T (hb_ot_map_feature_flags_t);
+
+
+struct hb_ot_map_feature_t
+{
+  hb_tag_t tag;
+  hb_ot_map_feature_flags_t flags;
+};
+
+struct hb_ot_shape_plan_key_t;
+
+struct hb_ot_map_builder_t
+{
+  public:
+
+  HB_INTERNAL hb_ot_map_builder_t (hb_face_t *face_,
+				   const hb_segment_properties_t *props_);
+
+  HB_INTERNAL ~hb_ot_map_builder_t ();
+
+  HB_INTERNAL void add_feature (hb_tag_t tag,
+				hb_ot_map_feature_flags_t flags=F_NONE,
+				unsigned int value=1);
+
+  void add_feature (const hb_ot_map_feature_t &feat)
+  { add_feature (feat.tag, feat.flags); }
+
+  void enable_feature (hb_tag_t tag,
+		       hb_ot_map_feature_flags_t flags=F_NONE,
+		       unsigned int value=1)
+  { add_feature (tag, F_GLOBAL | flags, value); }
+
+  void disable_feature (hb_tag_t tag)
+  { add_feature (tag, F_GLOBAL, 0); }
+
+  void add_gsub_pause (hb_ot_map_t::pause_func_t pause_func)
+  { add_pause (0, pause_func); }
+  void add_gpos_pause (hb_ot_map_t::pause_func_t pause_func)
+  { add_pause (1, pause_func); }
+
+  HB_INTERNAL void compile (hb_ot_map_t                  &m,
+			    const hb_ot_shape_plan_key_t &key);
+
+  private:
+
+  HB_INTERNAL void add_lookups (hb_ot_map_t  &m,
+				unsigned int  table_index,
+				unsigned int  feature_index,
+				unsigned int  variations_index,
+				hb_mask_t     mask,
+				bool          auto_zwnj = true,
+				bool          auto_zwj = true,
+				bool          random = false);
+
+  struct feature_info_t {
+    hb_tag_t tag;
+    unsigned int seq; /* sequence#, used for stable sorting only */
+    unsigned int max_value;
+    hb_ot_map_feature_flags_t flags;
+    unsigned int default_value; /* for non-global features, what should the unset glyphs take */
+    unsigned int stage[2]; /* GSUB/GPOS */
+
+    HB_INTERNAL static int cmp (const void *pa, const void *pb)
+    {
+      const feature_info_t *a = (const feature_info_t *) pa;
+      const feature_info_t *b = (const feature_info_t *) pb;
+      return (a->tag != b->tag) ?  (a->tag < b->tag ? -1 : 1) :
+	     (a->seq < b->seq ? -1 : a->seq > b->seq ? 1 : 0);
+    }
+  };
+
+  struct stage_info_t {
+    unsigned int index;
+    hb_ot_map_t::pause_func_t pause_func;
+  };
+
+  HB_INTERNAL void add_pause (unsigned int table_index, hb_ot_map_t::pause_func_t pause_func);
+
+  public:
+
+  hb_face_t *face;
+  hb_segment_properties_t props;
+
+  hb_tag_t chosen_script[2];
+  bool found_script[2];
+  unsigned int script_index[2], language_index[2];
+
+  private:
+
+  unsigned int current_stage[2]; /* GSUB/GPOS */
+  hb_vector_t<feature_info_t> feature_infos;
+  hb_vector_t<stage_info_t> stages[2]; /* GSUB/GPOS */
+};
+
+
+
+#endif /* HB_OT_MAP_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-math-table.hh b/thirdparty/harfbuzz/src/hb-ot-math-table.hh
new file mode 100644
index 0000000000..26aa080603
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-math-table.hh
@@ -0,0 +1,728 @@
+/*
+ * Copyright © 2016  Igalia S.L.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Igalia Author(s): Frédéric Wang
+ */
+
+#ifndef HB_OT_MATH_TABLE_HH
+#define HB_OT_MATH_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout-common.hh"
+#include "hb-ot-math.h"
+
+namespace OT {
+
+
+struct MathValueRecord
+{
+  hb_position_t get_x_value (hb_font_t *font, const void *base) const
+  { return font->em_scale_x (value) + (base+deviceTable).get_x_delta (font); }
+  hb_position_t get_y_value (hb_font_t *font, const void *base) const
+  { return font->em_scale_y (value) + (base+deviceTable).get_y_delta (font); }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && deviceTable.sanitize (c, base));
+  }
+
+  protected:
+  HBINT16		value;		/* The X or Y value in design units */
+  OffsetTo<Device>	deviceTable;	/* Offset to the device table - from the
+					 * beginning of parent table.  May be NULL.
+					 * Suggested format for device table is 1. */
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct MathConstants
+{
+  bool sanitize_math_value_records (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+
+    unsigned int count = ARRAY_LENGTH (mathValueRecords);
+    for (unsigned int i = 0; i < count; i++)
+      if (!mathValueRecords[i].sanitize (c, this))
+	return_trace (false);
+
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && sanitize_math_value_records (c));
+  }
+
+  hb_position_t get_value (hb_ot_math_constant_t constant,
+			   hb_font_t *font) const
+  {
+    switch (constant) {
+
+    case HB_OT_MATH_CONSTANT_SCRIPT_PERCENT_SCALE_DOWN:
+    case HB_OT_MATH_CONSTANT_SCRIPT_SCRIPT_PERCENT_SCALE_DOWN:
+      return percentScaleDown[constant - HB_OT_MATH_CONSTANT_SCRIPT_PERCENT_SCALE_DOWN];
+
+    case HB_OT_MATH_CONSTANT_DELIMITED_SUB_FORMULA_MIN_HEIGHT:
+    case HB_OT_MATH_CONSTANT_DISPLAY_OPERATOR_MIN_HEIGHT:
+      return font->em_scale_y (minHeight[constant - HB_OT_MATH_CONSTANT_DELIMITED_SUB_FORMULA_MIN_HEIGHT]);
+
+    case HB_OT_MATH_CONSTANT_RADICAL_KERN_AFTER_DEGREE:
+    case HB_OT_MATH_CONSTANT_RADICAL_KERN_BEFORE_DEGREE:
+    case HB_OT_MATH_CONSTANT_SKEWED_FRACTION_HORIZONTAL_GAP:
+    case HB_OT_MATH_CONSTANT_SPACE_AFTER_SCRIPT:
+      return mathValueRecords[constant - HB_OT_MATH_CONSTANT_MATH_LEADING].get_x_value (font, this);
+
+    case HB_OT_MATH_CONSTANT_ACCENT_BASE_HEIGHT:
+    case HB_OT_MATH_CONSTANT_AXIS_HEIGHT:
+    case HB_OT_MATH_CONSTANT_FLATTENED_ACCENT_BASE_HEIGHT:
+    case HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_DISPLAY_STYLE_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_FRACTION_DENOM_DISPLAY_STYLE_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_DISPLAY_STYLE_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_FRACTION_NUM_DISPLAY_STYLE_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_FRACTION_RULE_THICKNESS:
+    case HB_OT_MATH_CONSTANT_LOWER_LIMIT_BASELINE_DROP_MIN:
+    case HB_OT_MATH_CONSTANT_LOWER_LIMIT_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_MATH_LEADING:
+    case HB_OT_MATH_CONSTANT_OVERBAR_EXTRA_ASCENDER:
+    case HB_OT_MATH_CONSTANT_OVERBAR_RULE_THICKNESS:
+    case HB_OT_MATH_CONSTANT_OVERBAR_VERTICAL_GAP:
+    case HB_OT_MATH_CONSTANT_RADICAL_DISPLAY_STYLE_VERTICAL_GAP:
+    case HB_OT_MATH_CONSTANT_RADICAL_EXTRA_ASCENDER:
+    case HB_OT_MATH_CONSTANT_RADICAL_RULE_THICKNESS:
+    case HB_OT_MATH_CONSTANT_RADICAL_VERTICAL_GAP:
+    case HB_OT_MATH_CONSTANT_SKEWED_FRACTION_VERTICAL_GAP:
+    case HB_OT_MATH_CONSTANT_STACK_BOTTOM_DISPLAY_STYLE_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_STACK_BOTTOM_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_STACK_DISPLAY_STYLE_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_STACK_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_STACK_TOP_DISPLAY_STYLE_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_STACK_TOP_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_STRETCH_STACK_BOTTOM_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_STRETCH_STACK_GAP_ABOVE_MIN:
+    case HB_OT_MATH_CONSTANT_STRETCH_STACK_GAP_BELOW_MIN:
+    case HB_OT_MATH_CONSTANT_STRETCH_STACK_TOP_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_SUBSCRIPT_BASELINE_DROP_MIN:
+    case HB_OT_MATH_CONSTANT_SUBSCRIPT_SHIFT_DOWN:
+    case HB_OT_MATH_CONSTANT_SUBSCRIPT_TOP_MAX:
+    case HB_OT_MATH_CONSTANT_SUB_SUPERSCRIPT_GAP_MIN:
+    case HB_OT_MATH_CONSTANT_SUPERSCRIPT_BASELINE_DROP_MAX:
+    case HB_OT_MATH_CONSTANT_SUPERSCRIPT_BOTTOM_MAX_WITH_SUBSCRIPT:
+    case HB_OT_MATH_CONSTANT_SUPERSCRIPT_BOTTOM_MIN:
+    case HB_OT_MATH_CONSTANT_SUPERSCRIPT_SHIFT_UP:
+    case HB_OT_MATH_CONSTANT_SUPERSCRIPT_SHIFT_UP_CRAMPED:
+    case HB_OT_MATH_CONSTANT_UNDERBAR_EXTRA_DESCENDER:
+    case HB_OT_MATH_CONSTANT_UNDERBAR_RULE_THICKNESS:
+    case HB_OT_MATH_CONSTANT_UNDERBAR_VERTICAL_GAP:
+    case HB_OT_MATH_CONSTANT_UPPER_LIMIT_BASELINE_RISE_MIN:
+    case HB_OT_MATH_CONSTANT_UPPER_LIMIT_GAP_MIN:
+      return mathValueRecords[constant - HB_OT_MATH_CONSTANT_MATH_LEADING].get_y_value (font, this);
+
+    case HB_OT_MATH_CONSTANT_RADICAL_DEGREE_BOTTOM_RAISE_PERCENT:
+      return radicalDegreeBottomRaisePercent;
+
+    default:
+      return 0;
+    }
+  }
+
+  protected:
+  HBINT16 percentScaleDown[2];
+  HBUINT16 minHeight[2];
+  MathValueRecord mathValueRecords[51];
+  HBINT16 radicalDegreeBottomRaisePercent;
+
+  public:
+  DEFINE_SIZE_STATIC (214);
+};
+
+struct MathItalicsCorrectionInfo
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  coverage.sanitize (c, this) &&
+		  italicsCorrection.sanitize (c, this));
+  }
+
+  hb_position_t get_value (hb_codepoint_t glyph,
+			   hb_font_t *font) const
+  {
+    unsigned int index = (this+coverage).get_coverage (glyph);
+    return italicsCorrection[index].get_x_value (font, this);
+  }
+
+  protected:
+  OffsetTo<Coverage>       coverage;		/* Offset to Coverage table -
+						 * from the beginning of
+						 * MathItalicsCorrectionInfo
+						 * table. */
+  ArrayOf<MathValueRecord> italicsCorrection;	/* Array of MathValueRecords
+						 * defining italics correction
+						 * values for each
+						 * covered glyph. */
+
+  public:
+  DEFINE_SIZE_ARRAY (4, italicsCorrection);
+};
+
+struct MathTopAccentAttachment
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  topAccentCoverage.sanitize (c, this) &&
+		  topAccentAttachment.sanitize (c, this));
+  }
+
+  hb_position_t get_value (hb_codepoint_t glyph,
+			   hb_font_t *font) const
+  {
+    unsigned int index = (this+topAccentCoverage).get_coverage (glyph);
+    if (index == NOT_COVERED)
+      return font->get_glyph_h_advance (glyph) / 2;
+    return topAccentAttachment[index].get_x_value (font, this);
+  }
+
+  protected:
+  OffsetTo<Coverage>       topAccentCoverage;   /* Offset to Coverage table -
+						 * from the beginning of
+						 * MathTopAccentAttachment
+						 * table. */
+  ArrayOf<MathValueRecord> topAccentAttachment; /* Array of MathValueRecords
+						 * defining top accent
+						 * attachment points for each
+						 * covered glyph. */
+
+  public:
+  DEFINE_SIZE_ARRAY (2 + 2, topAccentAttachment);
+};
+
+struct MathKern
+{
+  bool sanitize_math_value_records (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    unsigned int count = 2 * heightCount + 1;
+    for (unsigned int i = 0; i < count; i++)
+      if (!mathValueRecordsZ.arrayZ[i].sanitize (c, this)) return_trace (false);
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  c->check_array (mathValueRecordsZ.arrayZ, 2 * heightCount + 1) &&
+		  sanitize_math_value_records (c));
+  }
+
+  hb_position_t get_value (hb_position_t correction_height, hb_font_t *font) const
+  {
+    const MathValueRecord* correctionHeight = mathValueRecordsZ.arrayZ;
+    const MathValueRecord* kernValue = mathValueRecordsZ.arrayZ + heightCount;
+    int sign = font->y_scale < 0 ? -1 : +1;
+
+    /* The description of the MathKern table is a ambiguous, but interpreting
+     * "between the two heights found at those indexes" for 0 < i < len as
+     *
+     *   correctionHeight[i-1] < correction_height <= correctionHeight[i]
+     *
+     * makes the result consistent with the limit cases and we can just use the
+     * binary search algorithm of std::upper_bound:
+     */
+    unsigned int i = 0;
+    unsigned int count = heightCount;
+    while (count > 0)
+    {
+      unsigned int half = count / 2;
+      hb_position_t height = correctionHeight[i + half].get_y_value (font, this);
+      if (sign * height < sign * correction_height)
+      {
+	i += half + 1;
+	count -= half + 1;
+      } else
+	count = half;
+    }
+    return kernValue[i].get_x_value (font, this);
+  }
+
+  protected:
+  HBUINT16	heightCount;
+  UnsizedArrayOf<MathValueRecord>
+		mathValueRecordsZ;
+				/* Array of correction heights at
+				 * which the kern value changes.
+				 * Sorted by the height value in
+				 * design units (heightCount entries),
+				 * Followed by:
+				 * Array of kern values corresponding
+				 * to heights. (heightCount+1 entries).
+				 */
+
+  public:
+  DEFINE_SIZE_ARRAY (2, mathValueRecordsZ);
+};
+
+struct MathKernInfoRecord
+{
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+
+    unsigned int count = ARRAY_LENGTH (mathKern);
+    for (unsigned int i = 0; i < count; i++)
+      if (unlikely (!mathKern[i].sanitize (c, base)))
+	return_trace (false);
+
+    return_trace (true);
+  }
+
+  hb_position_t get_kerning (hb_ot_math_kern_t kern,
+			     hb_position_t correction_height,
+			     hb_font_t *font,
+			     const void *base) const
+  {
+    unsigned int idx = kern;
+    if (unlikely (idx >= ARRAY_LENGTH (mathKern))) return 0;
+    return (base+mathKern[idx]).get_value (correction_height, font);
+  }
+
+  protected:
+  /* Offset to MathKern table for each corner -
+   * from the beginning of MathKernInfo table.  May be NULL. */
+  OffsetTo<MathKern> mathKern[4];
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct MathKernInfo
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  mathKernCoverage.sanitize (c, this) &&
+		  mathKernInfoRecords.sanitize (c, this));
+  }
+
+  hb_position_t get_kerning (hb_codepoint_t glyph,
+			     hb_ot_math_kern_t kern,
+			     hb_position_t correction_height,
+			     hb_font_t *font) const
+  {
+    unsigned int index = (this+mathKernCoverage).get_coverage (glyph);
+    return mathKernInfoRecords[index].get_kerning (kern, correction_height, font, this);
+  }
+
+  protected:
+  OffsetTo<Coverage>
+		mathKernCoverage;
+				/* Offset to Coverage table -
+				 * from the beginning of the
+				 * MathKernInfo table. */
+  ArrayOf<MathKernInfoRecord>
+		mathKernInfoRecords;
+				/* Array of MathKernInfoRecords,
+				 * per-glyph information for
+				 * mathematical positioning
+				 * of subscripts and
+				 * superscripts. */
+
+  public:
+  DEFINE_SIZE_ARRAY (4, mathKernInfoRecords);
+};
+
+struct MathGlyphInfo
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  mathItalicsCorrectionInfo.sanitize (c, this) &&
+		  mathTopAccentAttachment.sanitize (c, this) &&
+		  extendedShapeCoverage.sanitize (c, this) &&
+		  mathKernInfo.sanitize (c, this));
+  }
+
+  hb_position_t
+  get_italics_correction (hb_codepoint_t  glyph, hb_font_t *font) const
+  { return (this+mathItalicsCorrectionInfo).get_value (glyph, font); }
+
+  hb_position_t
+  get_top_accent_attachment (hb_codepoint_t  glyph, hb_font_t *font) const
+  { return (this+mathTopAccentAttachment).get_value (glyph, font); }
+
+  bool is_extended_shape (hb_codepoint_t glyph) const
+  { return (this+extendedShapeCoverage).get_coverage (glyph) != NOT_COVERED; }
+
+  hb_position_t get_kerning (hb_codepoint_t glyph,
+			     hb_ot_math_kern_t kern,
+			     hb_position_t correction_height,
+			     hb_font_t *font) const
+  { return (this+mathKernInfo).get_kerning (glyph, kern, correction_height, font); }
+
+  protected:
+  /* Offset to MathItalicsCorrectionInfo table -
+   * from the beginning of MathGlyphInfo table. */
+  OffsetTo<MathItalicsCorrectionInfo> mathItalicsCorrectionInfo;
+
+  /* Offset to MathTopAccentAttachment table -
+   * from the beginning of MathGlyphInfo table. */
+  OffsetTo<MathTopAccentAttachment> mathTopAccentAttachment;
+
+  /* Offset to coverage table for Extended Shape glyphs -
+   * from the beginning of MathGlyphInfo table. When the left or right glyph of
+   * a box is an extended shape variant, the (ink) box (and not the default
+   * position defined by values in MathConstants table) should be used for
+   * vertical positioning purposes.  May be NULL.. */
+  OffsetTo<Coverage> extendedShapeCoverage;
+
+   /* Offset to MathKernInfo table -
+    * from the beginning of MathGlyphInfo table. */
+  OffsetTo<MathKernInfo> mathKernInfo;
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct MathGlyphVariantRecord
+{
+  friend struct MathGlyphConstruction;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  HBGlyphID variantGlyph;       /* Glyph ID for the variant. */
+  HBUINT16  advanceMeasurement; /* Advance width/height, in design units, of the
+				 * variant, in the direction of requested
+				 * glyph extension. */
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct PartFlags : HBUINT16
+{
+  enum Flags {
+    Extender	= 0x0001u, /* If set, the part can be skipped or repeated. */
+
+    Defined	= 0x0001u, /* All defined flags. */
+  };
+
+  public:
+  DEFINE_SIZE_STATIC (2);
+};
+
+struct MathGlyphPartRecord
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  void extract (hb_ot_math_glyph_part_t &out,
+		int64_t mult,
+		hb_font_t *font) const
+  {
+    out.glyph			= glyph;
+
+    out.start_connector_length	= font->em_mult (startConnectorLength, mult);
+    out.end_connector_length	= font->em_mult (endConnectorLength, mult);
+    out.full_advance		= font->em_mult (fullAdvance, mult);
+
+    static_assert ((unsigned int) HB_OT_MATH_GLYPH_PART_FLAG_EXTENDER ==
+		   (unsigned int) PartFlags::Extender, "");
+
+    out.flags = (hb_ot_math_glyph_part_flags_t)
+		(unsigned int)
+		(partFlags & PartFlags::Defined);
+  }
+
+  protected:
+  HBGlyphID	glyph;		/* Glyph ID for the part. */
+  HBUINT16	startConnectorLength;
+				/* Advance width/ height of the straight bar
+				 * connector material, in design units, is at
+				 * the beginning of the glyph, in the
+				 * direction of the extension. */
+  HBUINT16	endConnectorLength;
+				/* Advance width/ height of the straight bar
+				 * connector material, in design units, is at
+				 * the end of the glyph, in the direction of
+				 * the extension. */
+  HBUINT16	fullAdvance;	/* Full advance width/height for this part,
+				 * in the direction of the extension.
+				 * In design units. */
+  PartFlags	partFlags;	/* Part qualifiers. */
+
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+struct MathGlyphAssembly
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  italicsCorrection.sanitize (c, this) &&
+		  partRecords.sanitize (c));
+  }
+
+  unsigned int get_parts (hb_direction_t direction,
+			  hb_font_t *font,
+			  unsigned int start_offset,
+			  unsigned int *parts_count, /* IN/OUT */
+			  hb_ot_math_glyph_part_t *parts /* OUT */,
+			  hb_position_t *italics_correction /* OUT */) const
+  {
+    if (parts_count)
+    {
+      int64_t mult = font->dir_mult (direction);
+      for (auto _ : hb_zip (partRecords.sub_array (start_offset, parts_count),
+			    hb_array (parts, *parts_count)))
+	_.first.extract (_.second, mult, font);
+    }
+
+    if (italics_correction)
+      *italics_correction = italicsCorrection.get_x_value (font, this);
+
+    return partRecords.len;
+  }
+
+  protected:
+  MathValueRecord
+		italicsCorrection;
+				/* Italics correction of this
+				 * MathGlyphAssembly. Should not
+				 * depend on the assembly size. */
+  ArrayOf<MathGlyphPartRecord>
+		partRecords;	/* Array of part records, from
+				 * left to right and bottom to
+				 * top. */
+
+  public:
+  DEFINE_SIZE_ARRAY (6, partRecords);
+};
+
+struct MathGlyphConstruction
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  glyphAssembly.sanitize (c, this) &&
+		  mathGlyphVariantRecord.sanitize (c));
+  }
+
+  const MathGlyphAssembly &get_assembly () const { return this+glyphAssembly; }
+
+  unsigned int get_variants (hb_direction_t direction,
+			     hb_font_t *font,
+			     unsigned int start_offset,
+			     unsigned int *variants_count, /* IN/OUT */
+			     hb_ot_math_glyph_variant_t *variants /* OUT */) const
+  {
+    if (variants_count)
+    {
+      int64_t mult = font->dir_mult (direction);
+      for (auto _ : hb_zip (mathGlyphVariantRecord.sub_array (start_offset, variants_count),
+			    hb_array (variants, *variants_count)))
+	_.second = {_.first.variantGlyph, font->em_mult (_.first.advanceMeasurement, mult)};
+    }
+    return mathGlyphVariantRecord.len;
+  }
+
+  protected:
+  /* Offset to MathGlyphAssembly table for this shape - from the beginning of
+     MathGlyphConstruction table.  May be NULL. */
+  OffsetTo<MathGlyphAssembly>	  glyphAssembly;
+
+  /* MathGlyphVariantRecords for alternative variants of the glyphs. */
+  ArrayOf<MathGlyphVariantRecord> mathGlyphVariantRecord;
+
+  public:
+  DEFINE_SIZE_ARRAY (4, mathGlyphVariantRecord);
+};
+
+struct MathVariants
+{
+  bool sanitize_offsets (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    unsigned int count = vertGlyphCount + horizGlyphCount;
+    for (unsigned int i = 0; i < count; i++)
+      if (!glyphConstruction.arrayZ[i].sanitize (c, this)) return_trace (false);
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  vertGlyphCoverage.sanitize (c, this) &&
+		  horizGlyphCoverage.sanitize (c, this) &&
+		  c->check_array (glyphConstruction.arrayZ, vertGlyphCount + horizGlyphCount) &&
+		  sanitize_offsets (c));
+  }
+
+  hb_position_t get_min_connector_overlap (hb_direction_t direction,
+						  hb_font_t *font) const
+  { return font->em_scale_dir (minConnectorOverlap, direction); }
+
+  unsigned int get_glyph_variants (hb_codepoint_t glyph,
+				   hb_direction_t direction,
+				   hb_font_t *font,
+				   unsigned int start_offset,
+				   unsigned int *variants_count, /* IN/OUT */
+				   hb_ot_math_glyph_variant_t *variants /* OUT */) const
+  { return get_glyph_construction (glyph, direction, font)
+	   .get_variants (direction, font, start_offset, variants_count, variants); }
+
+  unsigned int get_glyph_parts (hb_codepoint_t glyph,
+				hb_direction_t direction,
+				hb_font_t *font,
+				unsigned int start_offset,
+				unsigned int *parts_count, /* IN/OUT */
+				hb_ot_math_glyph_part_t *parts /* OUT */,
+				hb_position_t *italics_correction /* OUT */) const
+  { return get_glyph_construction (glyph, direction, font)
+	   .get_assembly ()
+	   .get_parts (direction, font,
+		       start_offset, parts_count, parts,
+		       italics_correction); }
+
+  private:
+  const MathGlyphConstruction &
+  get_glyph_construction (hb_codepoint_t glyph,
+			  hb_direction_t direction,
+			  hb_font_t *font HB_UNUSED) const
+  {
+    bool vertical = HB_DIRECTION_IS_VERTICAL (direction);
+    unsigned int count = vertical ? vertGlyphCount : horizGlyphCount;
+    const OffsetTo<Coverage> &coverage = vertical ? vertGlyphCoverage
+						  : horizGlyphCoverage;
+
+    unsigned int index = (this+coverage).get_coverage (glyph);
+    if (unlikely (index >= count)) return Null (MathGlyphConstruction);
+
+    if (!vertical)
+      index += vertGlyphCount;
+
+    return this+glyphConstruction[index];
+  }
+
+  protected:
+  HBUINT16	minConnectorOverlap;
+				/* Minimum overlap of connecting
+				 * glyphs during glyph construction,
+				 * in design units. */
+  OffsetTo<Coverage> vertGlyphCoverage;
+				/* Offset to Coverage table -
+				 * from the beginning of MathVariants
+				 * table. */
+  OffsetTo<Coverage> horizGlyphCoverage;
+				/* Offset to Coverage table -
+				 * from the beginning of MathVariants
+				 * table. */
+  HBUINT16	vertGlyphCount;	/* Number of glyphs for which
+				 * information is provided for
+				 * vertically growing variants. */
+  HBUINT16	horizGlyphCount;/* Number of glyphs for which
+				 * information is provided for
+				 * horizontally growing variants. */
+
+  /* Array of offsets to MathGlyphConstruction tables - from the beginning of
+     the MathVariants table, for shapes growing in vertical/horizontal
+     direction. */
+  UnsizedArrayOf<OffsetTo<MathGlyphConstruction>>
+			glyphConstruction;
+
+  public:
+  DEFINE_SIZE_ARRAY (10, glyphConstruction);
+};
+
+
+/*
+ * MATH -- Mathematical typesetting
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/math
+ */
+
+struct MATH
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_MATH;
+
+  bool has_data () const { return version.to_int (); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  mathConstants.sanitize (c, this) &&
+		  mathGlyphInfo.sanitize (c, this) &&
+		  mathVariants.sanitize (c, this));
+  }
+
+  hb_position_t get_constant (hb_ot_math_constant_t  constant,
+			      hb_font_t		   *font) const
+  { return (this+mathConstants).get_value (constant, font); }
+
+  const MathGlyphInfo &get_glyph_info () const { return this+mathGlyphInfo; }
+
+  const MathVariants &get_variants () const    { return this+mathVariants; }
+
+  protected:
+  FixedVersion<>version;	/* Version of the MATH table
+				 * initially set to 0x00010000u */
+  OffsetTo<MathConstants>
+		mathConstants;	/* MathConstants table */
+  OffsetTo<MathGlyphInfo>
+		mathGlyphInfo;	/* MathGlyphInfo table */
+  OffsetTo<MathVariants>
+		mathVariants;	/* MathVariants table */
+
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_MATH_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-math.cc b/thirdparty/harfbuzz/src/hb-ot-math.cc
new file mode 100644
index 0000000000..9d8c6e735a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-math.cc
@@ -0,0 +1,293 @@
+/*
+ * Copyright © 2016  Igalia S.L.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Igalia Author(s): Frédéric Wang
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_MATH
+
+#include "hb-ot-math-table.hh"
+
+
+/**
+ * SECTION:hb-ot-math
+ * @title: hb-ot-math
+ * @short_description: OpenType Math information
+ * @include: hb-ot.h
+ *
+ * Functions for fetching mathematics layout data from OpenType fonts.
+ *
+ * HarfBuzz itself does not implement a math layout solution. The
+ * functions and types provided can be used by client programs to access
+ * the font data necessary for typesetting OpenType Math layout.
+ *
+ **/
+
+
+/*
+ * OT::MATH
+ */
+
+/**
+ * hb_ot_math_has_data:
+ * @face: #hb_face_t to test
+ *
+ * Tests whether a face has a `MATH` table.
+ *
+ * Return value: true if the table is found, false otherwise
+ *
+ * Since: 1.3.3
+ **/
+hb_bool_t
+hb_ot_math_has_data (hb_face_t *face)
+{
+  return face->table.MATH->has_data ();
+}
+
+/**
+ * hb_ot_math_get_constant:
+ * @font: #hb_font_t to work upon
+ * @constant: #hb_ot_math_constant_t the constant to retrieve
+ *
+ * Fetches the specified math constant. For most constants, the value returned
+ * is an #hb_position_t.
+ *
+ * However, if the requested constant is #HB_OT_MATH_CONSTANT_SCRIPT_PERCENT_SCALE_DOWN,
+ * #HB_OT_MATH_CONSTANT_SCRIPT_SCRIPT_PERCENT_SCALE_DOWN or
+ * #HB_OT_MATH_CONSTANT_SCRIPT_PERCENT_SCALE_DOWN, then the return value is
+ * an integer between 0 and 100 representing that percentage.
+ *
+ * Return value: the requested constant or zero
+ *
+ * Since: 1.3.3
+ **/
+hb_position_t
+hb_ot_math_get_constant (hb_font_t *font,
+			 hb_ot_math_constant_t constant)
+{
+  return font->face->table.MATH->get_constant(constant, font);
+}
+
+/**
+ * hb_ot_math_get_glyph_italics_correction:
+ * @font: #hb_font_t to work upon
+ * @glyph: The glyph index from which to retrieve the value
+ *
+ * Fetches an italics-correction value (if one exists) for the specified
+ * glyph index.
+ *
+  * Return value: the italics correction of the glyph or zero
+ *
+ * Since: 1.3.3
+ **/
+hb_position_t
+hb_ot_math_get_glyph_italics_correction (hb_font_t *font,
+					 hb_codepoint_t glyph)
+{
+  return font->face->table.MATH->get_glyph_info().get_italics_correction (glyph, font);
+}
+
+/**
+ * hb_ot_math_get_glyph_top_accent_attachment:
+ * @font: #hb_font_t to work upon
+ * @glyph: The glyph index from which to retrieve the value
+ *
+ * Fetches a top-accent-attachment value (if one exists) for the specified
+ * glyph index.
+ *
+ * For any glyph that does not have a top-accent-attachment value - that is,
+ * a glyph not covered by the `MathTopAccentAttachment` table (or, when
+ * @font has no `MathTopAccentAttachment` table or no `MATH` table, any
+ * glyph) - the function synthesizes a value, returning the position at
+ * one-half the glyph's advance width.
+ *
+ * Return value: the top accent attachment of the glyph or 0.5 * the advance
+ *               width of @glyph
+ *
+ * Since: 1.3.3
+ **/
+hb_position_t
+hb_ot_math_get_glyph_top_accent_attachment (hb_font_t *font,
+					    hb_codepoint_t glyph)
+{
+  return font->face->table.MATH->get_glyph_info().get_top_accent_attachment (glyph, font);
+}
+
+/**
+ * hb_ot_math_is_glyph_extended_shape:
+ * @face: #hb_face_t to work upon
+ * @glyph: The glyph index to test
+ *
+ * Tests whether the given glyph index is an extended shape in the face.
+ *
+ * Return value: true if the glyph is an extended shape, false otherwise
+ *
+ * Since: 1.3.3
+ **/
+hb_bool_t
+hb_ot_math_is_glyph_extended_shape (hb_face_t *face,
+				    hb_codepoint_t glyph)
+{
+  return face->table.MATH->get_glyph_info().is_extended_shape (glyph);
+}
+
+/**
+ * hb_ot_math_get_glyph_kerning:
+ * @font: #hb_font_t to work upon
+ * @glyph: The glyph index from which to retrieve the value
+ * @kern: The #hb_ot_math_kern_t from which to retrieve the value
+ * @correction_height: the correction height to use to determine the kerning.
+ *
+ * Fetches the math kerning (cut-ins) value for the specified font, glyph index, and
+ * @kern. 
+ *
+ * If the MathKern table is found, the function examines it to find a height
+ * value that is greater or equal to @correction_height. If such a height
+ * value is found, corresponding kerning value from the table is returned. If
+ * no such height value is found, the last kerning value is returned.
+ *
+ * Return value: requested kerning value or zero
+ *
+ * Since: 1.3.3
+ **/
+hb_position_t
+hb_ot_math_get_glyph_kerning (hb_font_t *font,
+			      hb_codepoint_t glyph,
+			      hb_ot_math_kern_t kern,
+			      hb_position_t correction_height)
+{
+  return font->face->table.MATH->get_glyph_info().get_kerning (glyph,
+							       kern,
+							       correction_height,
+							       font);
+}
+
+/**
+ * hb_ot_math_get_glyph_variants:
+ * @font: #hb_font_t to work upon
+ * @glyph: The index of the glyph to stretch
+ * @direction: The direction of the stretching (horizontal or vertical)
+ * @start_offset: offset of the first variant to retrieve
+ * @variants_count: (inout): Input = the maximum number of variants to return;
+ *                           Output = the actual number of variants returned
+ * @variants: (out) (array length=variants_count): array of variants returned
+ *
+ * Fetches the MathGlyphConstruction for the specified font, glyph index, and
+ * direction. The corresponding list of size variants is returned as a list of
+ * #hb_ot_math_glyph_variant_t structs.
+ *
+ * <note>The @direction parameter is only used to select between horizontal
+ * or vertical directions for the construction. Even though all #hb_direction_t
+ * values are accepted, only the result of #HB_DIRECTION_IS_HORIZONTAL is
+ * considered.</note> 
+ *
+ * Return value: the total number of size variants available or zero
+ *
+ * Since: 1.3.3
+ **/
+unsigned int
+hb_ot_math_get_glyph_variants (hb_font_t *font,
+			       hb_codepoint_t glyph,
+			       hb_direction_t direction,
+			       unsigned int start_offset,
+			       unsigned int *variants_count, /* IN/OUT */
+			       hb_ot_math_glyph_variant_t *variants /* OUT */)
+{
+  return font->face->table.MATH->get_variants().get_glyph_variants (glyph, direction, font,
+								    start_offset,
+								    variants_count,
+								    variants);
+}
+
+/**
+ * hb_ot_math_get_min_connector_overlap:
+ * @font: #hb_font_t to work upon
+ * @direction: direction of the stretching (horizontal or vertical)
+ *
+ * Fetches the MathVariants table for the specified font and returns the
+ * minimum overlap of connecting glyphs that are required to draw a glyph
+ * assembly in the specified direction.
+ *
+ * <note>The @direction parameter is only used to select between horizontal
+ * or vertical directions for the construction. Even though all #hb_direction_t
+ * values are accepted, only the result of #HB_DIRECTION_IS_HORIZONTAL is
+ * considered.</note> 
+ *
+ * Return value: requested minimum connector overlap or zero
+ *
+ * Since: 1.3.3
+ **/
+hb_position_t
+hb_ot_math_get_min_connector_overlap (hb_font_t *font,
+				      hb_direction_t direction)
+{
+  return font->face->table.MATH->get_variants().get_min_connector_overlap (direction, font);
+}
+
+/**
+ * hb_ot_math_get_glyph_assembly:
+ * @font: #hb_font_t to work upon
+ * @glyph: The index of the glyph to stretch
+ * @direction: direction of the stretching (horizontal or vertical)
+ * @start_offset: offset of the first glyph part to retrieve
+ * @parts_count: (inout): Input = maximum number of glyph parts to return;
+ *               Output = actual number of parts returned
+ * @parts: (out) (array length=parts_count): the glyph parts returned
+ * @italics_correction: (out): italics correction of the glyph assembly
+ *
+ * Fetches the GlyphAssembly for the specified font, glyph index, and direction.
+ * Returned are a list of #hb_ot_math_glyph_part_t glyph parts that can be
+ * used to draw the glyph and an italics-correction value (if one is defined
+ * in the font).
+ *
+ * <note>The @direction parameter is only used to select between horizontal
+ * or vertical directions for the construction. Even though all #hb_direction_t
+ * values are accepted, only the result of #HB_DIRECTION_IS_HORIZONTAL is
+ * considered.</note> 
+ *
+ * Return value: the total number of parts in the glyph assembly
+ *
+ * Since: 1.3.3
+ **/
+unsigned int
+hb_ot_math_get_glyph_assembly (hb_font_t *font,
+			       hb_codepoint_t glyph,
+			       hb_direction_t direction,
+			       unsigned int start_offset,
+			       unsigned int *parts_count, /* IN/OUT */
+			       hb_ot_math_glyph_part_t *parts, /* OUT */
+			       hb_position_t *italics_correction /* OUT */)
+{
+  return font->face->table.MATH->get_variants().get_glyph_parts (glyph,
+								 direction,
+								 font,
+								 start_offset,
+								 parts_count,
+								 parts,
+								 italics_correction);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-math.h b/thirdparty/harfbuzz/src/hb-ot-math.h
new file mode 100644
index 0000000000..ad864a762d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-math.h
@@ -0,0 +1,230 @@
+/*
+ * Copyright © 2016  Igalia S.L.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Igalia Author(s): Frédéric Wang
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_MATH_H
+#define HB_OT_MATH_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+
+/*
+ * MATH
+ */
+
+#define HB_OT_TAG_MATH HB_TAG('M','A','T','H')
+
+/* Use with hb_buffer_set_script() for math shaping. */
+#define HB_OT_MATH_SCRIPT HB_TAG('m','a','t','h')
+
+/* Types */
+
+/**
+ * hb_ot_math_constant_t:
+ *
+ * The 'MATH' table constants specified at
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/math
+ *
+ * Since: 1.3.3
+ */
+typedef enum {
+  HB_OT_MATH_CONSTANT_SCRIPT_PERCENT_SCALE_DOWN = 0,
+  HB_OT_MATH_CONSTANT_SCRIPT_SCRIPT_PERCENT_SCALE_DOWN = 1,
+  HB_OT_MATH_CONSTANT_DELIMITED_SUB_FORMULA_MIN_HEIGHT = 2,
+  HB_OT_MATH_CONSTANT_DISPLAY_OPERATOR_MIN_HEIGHT = 3,
+  HB_OT_MATH_CONSTANT_MATH_LEADING = 4,
+  HB_OT_MATH_CONSTANT_AXIS_HEIGHT = 5,
+  HB_OT_MATH_CONSTANT_ACCENT_BASE_HEIGHT = 6,
+  HB_OT_MATH_CONSTANT_FLATTENED_ACCENT_BASE_HEIGHT = 7,
+  HB_OT_MATH_CONSTANT_SUBSCRIPT_SHIFT_DOWN = 8,
+  HB_OT_MATH_CONSTANT_SUBSCRIPT_TOP_MAX = 9,
+  HB_OT_MATH_CONSTANT_SUBSCRIPT_BASELINE_DROP_MIN = 10,
+  HB_OT_MATH_CONSTANT_SUPERSCRIPT_SHIFT_UP = 11,
+  HB_OT_MATH_CONSTANT_SUPERSCRIPT_SHIFT_UP_CRAMPED = 12,
+  HB_OT_MATH_CONSTANT_SUPERSCRIPT_BOTTOM_MIN = 13,
+  HB_OT_MATH_CONSTANT_SUPERSCRIPT_BASELINE_DROP_MAX = 14,
+  HB_OT_MATH_CONSTANT_SUB_SUPERSCRIPT_GAP_MIN = 15,
+  HB_OT_MATH_CONSTANT_SUPERSCRIPT_BOTTOM_MAX_WITH_SUBSCRIPT = 16,
+  HB_OT_MATH_CONSTANT_SPACE_AFTER_SCRIPT = 17,
+  HB_OT_MATH_CONSTANT_UPPER_LIMIT_GAP_MIN = 18,
+  HB_OT_MATH_CONSTANT_UPPER_LIMIT_BASELINE_RISE_MIN = 19,
+  HB_OT_MATH_CONSTANT_LOWER_LIMIT_GAP_MIN = 20,
+  HB_OT_MATH_CONSTANT_LOWER_LIMIT_BASELINE_DROP_MIN = 21,
+  HB_OT_MATH_CONSTANT_STACK_TOP_SHIFT_UP = 22,
+  HB_OT_MATH_CONSTANT_STACK_TOP_DISPLAY_STYLE_SHIFT_UP = 23,
+  HB_OT_MATH_CONSTANT_STACK_BOTTOM_SHIFT_DOWN = 24,
+  HB_OT_MATH_CONSTANT_STACK_BOTTOM_DISPLAY_STYLE_SHIFT_DOWN = 25,
+  HB_OT_MATH_CONSTANT_STACK_GAP_MIN = 26,
+  HB_OT_MATH_CONSTANT_STACK_DISPLAY_STYLE_GAP_MIN = 27,
+  HB_OT_MATH_CONSTANT_STRETCH_STACK_TOP_SHIFT_UP = 28,
+  HB_OT_MATH_CONSTANT_STRETCH_STACK_BOTTOM_SHIFT_DOWN = 29,
+  HB_OT_MATH_CONSTANT_STRETCH_STACK_GAP_ABOVE_MIN = 30,
+  HB_OT_MATH_CONSTANT_STRETCH_STACK_GAP_BELOW_MIN = 31,
+  HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_SHIFT_UP = 32,
+  HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_DISPLAY_STYLE_SHIFT_UP = 33,
+  HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_SHIFT_DOWN = 34,
+  HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_DISPLAY_STYLE_SHIFT_DOWN = 35,
+  HB_OT_MATH_CONSTANT_FRACTION_NUMERATOR_GAP_MIN = 36,
+  HB_OT_MATH_CONSTANT_FRACTION_NUM_DISPLAY_STYLE_GAP_MIN = 37,
+  HB_OT_MATH_CONSTANT_FRACTION_RULE_THICKNESS = 38,
+  HB_OT_MATH_CONSTANT_FRACTION_DENOMINATOR_GAP_MIN = 39,
+  HB_OT_MATH_CONSTANT_FRACTION_DENOM_DISPLAY_STYLE_GAP_MIN = 40,
+  HB_OT_MATH_CONSTANT_SKEWED_FRACTION_HORIZONTAL_GAP = 41,
+  HB_OT_MATH_CONSTANT_SKEWED_FRACTION_VERTICAL_GAP = 42,
+  HB_OT_MATH_CONSTANT_OVERBAR_VERTICAL_GAP = 43,
+  HB_OT_MATH_CONSTANT_OVERBAR_RULE_THICKNESS = 44,
+  HB_OT_MATH_CONSTANT_OVERBAR_EXTRA_ASCENDER = 45,
+  HB_OT_MATH_CONSTANT_UNDERBAR_VERTICAL_GAP = 46,
+  HB_OT_MATH_CONSTANT_UNDERBAR_RULE_THICKNESS = 47,
+  HB_OT_MATH_CONSTANT_UNDERBAR_EXTRA_DESCENDER = 48,
+  HB_OT_MATH_CONSTANT_RADICAL_VERTICAL_GAP = 49,
+  HB_OT_MATH_CONSTANT_RADICAL_DISPLAY_STYLE_VERTICAL_GAP = 50,
+  HB_OT_MATH_CONSTANT_RADICAL_RULE_THICKNESS = 51,
+  HB_OT_MATH_CONSTANT_RADICAL_EXTRA_ASCENDER = 52,
+  HB_OT_MATH_CONSTANT_RADICAL_KERN_BEFORE_DEGREE = 53,
+  HB_OT_MATH_CONSTANT_RADICAL_KERN_AFTER_DEGREE = 54,
+  HB_OT_MATH_CONSTANT_RADICAL_DEGREE_BOTTOM_RAISE_PERCENT = 55
+} hb_ot_math_constant_t;
+
+/**
+ * hb_ot_math_kern_t:
+ *
+ * The math kerning-table types defined for the four corners
+ * of a glyph.
+ *
+ * Since: 1.3.3
+ */
+typedef enum {
+  HB_OT_MATH_KERN_TOP_RIGHT = 0,
+  HB_OT_MATH_KERN_TOP_LEFT = 1,
+  HB_OT_MATH_KERN_BOTTOM_RIGHT = 2,
+  HB_OT_MATH_KERN_BOTTOM_LEFT = 3
+} hb_ot_math_kern_t;
+
+/**
+ * hb_ot_math_glyph_variant_t:
+ * @glyph: The glyph index of the variant
+ * @advance: The advance width of the variant
+ *
+ * Data type to hold math-variant information for a glyph.
+ *
+ * Since: 1.3.3
+ */
+typedef struct hb_ot_math_glyph_variant_t {
+  hb_codepoint_t glyph;
+  hb_position_t advance;
+} hb_ot_math_glyph_variant_t;
+
+/**
+ * hb_ot_math_glyph_part_flags_t:
+ *
+ * Flags for math glyph parts.
+ *
+ * Since: 1.3.3
+ */
+typedef enum { /*< flags >*/
+  HB_OT_MATH_GLYPH_PART_FLAG_EXTENDER	= 0x00000001u  /* Extender glyph */
+} hb_ot_math_glyph_part_flags_t;
+
+/**
+ * hb_ot_math_glyph_part_t:
+ * @glyph: The glyph index of the variant part
+ * @start_connector_length: The length of the connector on the starting side of the variant part
+ * @end_connector_length: The length of the connector on the ending side of the variant part
+ * @full_advance: The total advance of the part
+ * @flags: #hb_ot_math_glyph_part_flags_t flags for the part
+ * 
+ * Data type to hold information for a "part" component of a math-variant glyph.
+ * Large variants for stretchable math glyphs (such as parentheses) can be constructed
+ * on the fly from parts.
+ *
+ * Since: 1.3.3
+ */
+typedef struct hb_ot_math_glyph_part_t {
+  hb_codepoint_t glyph;
+  hb_position_t start_connector_length;
+  hb_position_t end_connector_length;
+  hb_position_t full_advance;
+  hb_ot_math_glyph_part_flags_t flags;
+} hb_ot_math_glyph_part_t;
+
+/* Methods */
+
+HB_EXTERN hb_bool_t
+hb_ot_math_has_data (hb_face_t *face);
+
+HB_EXTERN hb_position_t
+hb_ot_math_get_constant (hb_font_t *font,
+			 hb_ot_math_constant_t constant);
+
+HB_EXTERN hb_position_t
+hb_ot_math_get_glyph_italics_correction (hb_font_t *font,
+					 hb_codepoint_t glyph);
+
+HB_EXTERN hb_position_t
+hb_ot_math_get_glyph_top_accent_attachment (hb_font_t *font,
+					    hb_codepoint_t glyph);
+
+HB_EXTERN hb_bool_t
+hb_ot_math_is_glyph_extended_shape (hb_face_t *face,
+				    hb_codepoint_t glyph);
+
+HB_EXTERN hb_position_t
+hb_ot_math_get_glyph_kerning (hb_font_t *font,
+			      hb_codepoint_t glyph,
+			      hb_ot_math_kern_t kern,
+			      hb_position_t correction_height);
+
+HB_EXTERN unsigned int
+hb_ot_math_get_glyph_variants (hb_font_t *font,
+			       hb_codepoint_t glyph,
+			       hb_direction_t direction,
+			       unsigned int start_offset,
+			       unsigned int *variants_count, /* IN/OUT */
+			       hb_ot_math_glyph_variant_t *variants /* OUT */);
+
+HB_EXTERN hb_position_t
+hb_ot_math_get_min_connector_overlap (hb_font_t *font,
+				      hb_direction_t direction);
+
+HB_EXTERN unsigned int
+hb_ot_math_get_glyph_assembly (hb_font_t *font,
+			       hb_codepoint_t glyph,
+			       hb_direction_t direction,
+			       unsigned int start_offset,
+			       unsigned int *parts_count, /* IN/OUT */
+			       hb_ot_math_glyph_part_t *parts, /* OUT */
+			       hb_position_t *italics_correction /* OUT */);
+
+
+HB_END_DECLS
+
+#endif /* HB_OT_MATH_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-maxp-table.hh b/thirdparty/harfbuzz/src/hb-ot-maxp-table.hh
new file mode 100644
index 0000000000..929956d12b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-maxp-table.hh
@@ -0,0 +1,142 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_MAXP_TABLE_HH
+#define HB_OT_MAXP_TABLE_HH
+
+#include "hb-open-type.hh"
+
+namespace OT {
+
+
+/*
+ * maxp -- Maximum Profile
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/maxp
+ */
+
+#define HB_OT_TAG_maxp HB_TAG('m','a','x','p')
+
+struct maxpV1Tail
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  HBUINT16 maxPoints;		  /* Maximum points in a non-composite glyph. */
+  HBUINT16 maxContours;		  /* Maximum contours in a non-composite glyph. */
+  HBUINT16 maxCompositePoints;	  /* Maximum points in a composite glyph. */
+  HBUINT16 maxCompositeContours;  /* Maximum contours in a composite glyph. */
+  HBUINT16 maxZones;		  /* 1 if instructions do not use the twilight zone (Z0),
+				   * or 2 if instructions do use Z0; should be set to 2 in
+				   * most cases. */
+  HBUINT16 maxTwilightPoints;	  /* Maximum points used in Z0. */
+  HBUINT16 maxStorage;		  /* Number of Storage Area locations. */
+  HBUINT16 maxFunctionDefs;	  /* Number of FDEFs, equal to the highest function number + 1. */
+  HBUINT16 maxInstructionDefs;	  /* Number of IDEFs. */
+  HBUINT16 maxStackElements;	  /* Maximum stack depth. (This includes Font and CVT
+				   * Programs, as well as the instructions for each glyph.) */
+  HBUINT16 maxSizeOfInstructions; /* Maximum byte count for glyph instructions. */
+  HBUINT16 maxComponentElements;  /* Maximum number of components referenced at
+				   * "top level" for any composite glyph. */
+  HBUINT16 maxComponentDepth;	  /* Maximum levels of recursion; 1 for simple components. */
+ public:
+  DEFINE_SIZE_STATIC (26);
+};
+
+
+struct maxp
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_maxp;
+
+  unsigned int get_num_glyphs () const { return numGlyphs; }
+
+  void set_num_glyphs (unsigned int count)
+  {
+    numGlyphs = count;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    if (version.major == 1)
+    {
+      const maxpV1Tail &v1 = StructAfter<maxpV1Tail> (*this);
+      return_trace (v1.sanitize (c));
+    }
+    return_trace (likely (version.major == 0 && version.minor == 0x5000u));
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    maxp *maxp_prime = c->serializer->embed (this);
+    if (unlikely (!maxp_prime)) return_trace (false);
+
+    maxp_prime->numGlyphs = c->plan->num_output_glyphs ();
+    if (maxp_prime->version.major == 1)
+    {
+      const maxpV1Tail *src_v1 = &StructAfter<maxpV1Tail> (*this);
+      maxpV1Tail *dest_v1 = c->serializer->embed<maxpV1Tail> (src_v1);
+      if (unlikely (!dest_v1)) return_trace (false);
+
+      if (c->plan->drop_hints)
+	drop_hint_fields (dest_v1);
+    }
+
+    return_trace (true);
+  }
+
+  static void drop_hint_fields (maxpV1Tail* dest_v1)
+  {
+    dest_v1->maxZones = 1;
+    dest_v1->maxTwilightPoints = 0;
+    dest_v1->maxStorage = 0;
+    dest_v1->maxFunctionDefs = 0;
+    dest_v1->maxInstructionDefs = 0;
+    dest_v1->maxStackElements = 0;
+    dest_v1->maxSizeOfInstructions = 0;
+  }
+
+  protected:
+  FixedVersion<>version;/* Version of the maxp table (0.5 or 1.0),
+			 * 0x00005000u or 0x00010000u. */
+  HBUINT16	numGlyphs;
+			/* The number of glyphs in the font. */
+/*maxpV1Tail	v1Tail[HB_VAR_ARRAY]; */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_MAXP_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-meta-table.hh b/thirdparty/harfbuzz/src/hb-ot-meta-table.hh
new file mode 100644
index 0000000000..1225e26ce1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-meta-table.hh
@@ -0,0 +1,127 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_META_TABLE_HH
+#define HB_OT_META_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * meta -- Metadata Table
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/meta
+ * https://developer.apple.com/fonts/TrueType-Reference-Manual/RM06/Chap6meta.html
+ */
+#define HB_OT_TAG_meta HB_TAG ('m','e','t','a')
+
+
+namespace OT {
+
+
+struct DataMap
+{
+  int cmp (hb_tag_t a) const { return tag.cmp (a); }
+
+  hb_tag_t get_tag () const { return tag; }
+
+  hb_blob_t *reference_entry (hb_blob_t *meta_blob) const
+  { return hb_blob_create_sub_blob (meta_blob, dataZ, dataLength); }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  dataZ.sanitize (c, base, dataLength)));
+  }
+
+  protected:
+  Tag		tag;		/* A tag indicating the type of metadata. */
+  LNNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		dataZ;		/* Offset in bytes from the beginning of the
+				 * metadata table to the data for this tag. */
+  HBUINT32	dataLength;	/* Length of the data. The data is not required to
+				 * be padded to any byte boundary. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct meta
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_meta;
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    { table = hb_sanitize_context_t ().reference_table<meta> (face); }
+    void fini () { table.destroy (); }
+
+    hb_blob_t *reference_entry (hb_tag_t tag) const
+    { return table->dataMaps.lsearch (tag).reference_entry (table.get_blob ()); }
+
+    unsigned int get_entries (unsigned int      start_offset,
+			      unsigned int     *count,
+			      hb_ot_meta_tag_t *entries) const
+    {
+      if (count)
+      {
+	+ table->dataMaps.sub_array (start_offset, count)
+	| hb_map (&DataMap::get_tag)
+	| hb_map ([](hb_tag_t tag) { return (hb_ot_meta_tag_t) tag; })
+	| hb_sink (hb_array (entries, *count))
+	;
+      }
+      return table->dataMaps.len;
+    }
+
+    private:
+    hb_blob_ptr_t<meta> table;
+  };
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version == 1 &&
+			  dataMaps.sanitize (c, this)));
+  }
+
+  protected:
+  HBUINT32	version;	/* Version number of the metadata table — set to 1. */
+  HBUINT32	flags;		/* Flags — currently unused; set to 0. */
+  HBUINT32	dataOffset;
+				/* Per Apple specification:
+				 * Offset from the beginning of the table to the data.
+				 * Per OT specification:
+				 * Reserved. Not used; should be set to 0. */
+  LArrayOf<DataMap>
+		dataMaps;/* Array of data map records. */
+  public:
+  DEFINE_SIZE_ARRAY (16, dataMaps);
+};
+
+struct meta_accelerator_t : meta::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_META_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-meta.cc b/thirdparty/harfbuzz/src/hb-ot-meta.cc
new file mode 100644
index 0000000000..54a0e10f9b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-meta.cc
@@ -0,0 +1,77 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_META
+
+#include "hb-ot-meta-table.hh"
+
+/**
+ * SECTION:hb-ot-meta
+ * @title: hb-ot-meta
+ * @short_description: OpenType Metadata
+ * @include: hb-ot.h
+ *
+ * Functions for fetching metadata from fonts.
+ **/
+
+/**
+ * hb_ot_meta_get_entry_tags:
+ * @face: a face object
+ * @start_offset: iteration's start offset
+ * @entries_count:(inout) (allow-none): buffer size as input, filled size as output
+ * @entries: (out caller-allocates) (array length=entries_count): entries tags buffer
+ *
+ * Return value: Number of all available feature types.
+ *
+ * Since: 2.6.0
+ **/
+unsigned int
+hb_ot_meta_get_entry_tags (hb_face_t        *face,
+			   unsigned int      start_offset,
+			   unsigned int     *entries_count, /* IN/OUT.  May be NULL. */
+			   hb_ot_meta_tag_t *entries        /* OUT.     May be NULL. */)
+{
+  return face->table.meta->get_entries (start_offset, entries_count, entries);
+}
+
+/**
+ * hb_ot_meta_reference_entry:
+ * @face: a #hb_face_t object.
+ * @meta_tag: tag of metadata you like to have.
+ *
+ * It fetches metadata entry of a given tag from a font.
+ *
+ * Returns: (transfer full): A blob containing the blob.
+ *
+ * Since: 2.6.0
+ **/
+hb_blob_t *
+hb_ot_meta_reference_entry (hb_face_t *face, hb_ot_meta_tag_t meta_tag)
+{
+  return face->table.meta->reference_entry (meta_tag);
+}
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-meta.h b/thirdparty/harfbuzz/src/hb-ot-meta.h
new file mode 100644
index 0000000000..0278d84148
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-meta.h
@@ -0,0 +1,71 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_META_H
+#define HB_OT_META_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+/**
+ * hb_ot_meta_tag_t:
+ * @HB_OT_META_TAG_DESIGN_LANGUAGES: Design languages. Text, using only
+ * Basic Latin (ASCII) characters. Indicates languages and/or scripts
+ * for the user audiences that the font was primarily designed for.
+ * @HB_OT_META_TAG_SUPPORTED_LANGUAGES: Supported languages. Text, using
+ * only Basic Latin (ASCII) characters. Indicates languages and/or scripts
+ * that the font is declared to be capable of supporting.
+ *
+ * Known metadata tags from https://docs.microsoft.com/en-us/typography/opentype/spec/meta
+ *
+ * Since: 2.6.0
+ **/
+typedef enum {
+/*
+   HB_OT_META_TAG_APPL		= HB_TAG ('a','p','p','l'),
+   HB_OT_META_TAG_BILD		= HB_TAG ('b','i','l','d'),
+*/
+  HB_OT_META_TAG_DESIGN_LANGUAGES	= HB_TAG ('d','l','n','g'),
+  HB_OT_META_TAG_SUPPORTED_LANGUAGES	= HB_TAG ('s','l','n','g'),
+
+  _HB_OT_META_TAG_MAX_VALUE = HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_ot_meta_tag_t;
+
+HB_EXTERN unsigned int
+hb_ot_meta_get_entry_tags (hb_face_t        *face,
+			   unsigned int      start_offset,
+			   unsigned int     *entries_count, /* IN/OUT.  May be NULL. */
+			   hb_ot_meta_tag_t *entries        /* OUT.     May be NULL. */);
+
+HB_EXTERN hb_blob_t *
+hb_ot_meta_reference_entry (hb_face_t *face, hb_ot_meta_tag_t meta_tag);
+
+HB_END_DECLS
+
+#endif /* HB_OT_META_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-metrics.cc b/thirdparty/harfbuzz/src/hb-ot-metrics.cc
new file mode 100644
index 0000000000..181ac4d57e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-metrics.cc
@@ -0,0 +1,231 @@
+/*
+ * Copyright © 2018-2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#include "hb-ot-var-mvar-table.hh"
+#include "hb-ot-gasp-table.hh" // Just so we compile it; unused otherwise.
+#include "hb-ot-os2-table.hh"
+#include "hb-ot-post-table.hh"
+#include "hb-ot-hhea-table.hh"
+#include "hb-ot-metrics.hh"
+#include "hb-ot-face.hh"
+
+
+static float
+_fix_ascender_descender (float value, hb_ot_metrics_tag_t metrics_tag)
+{
+  if (metrics_tag == HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER ||
+      metrics_tag == HB_OT_METRICS_TAG_VERTICAL_ASCENDER)
+    return fabs ((double) value);
+  if (metrics_tag == HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER ||
+      metrics_tag == HB_OT_METRICS_TAG_VERTICAL_DESCENDER)
+    return -fabs ((double) value);
+  return value;
+}
+
+/* The common part of _get_position logic needed on hb-ot-font and here
+   to be able to have slim builds without the not always needed parts */
+bool
+_hb_ot_metrics_get_position_common (hb_font_t           *font,
+				    hb_ot_metrics_tag_t  metrics_tag,
+				    hb_position_t       *position     /* OUT.  May be NULL. */)
+{
+  hb_face_t *face = font->face;
+  switch ((unsigned) metrics_tag)
+  {
+#ifndef HB_NO_VAR
+#define GET_VAR face->table.MVAR->get_var (metrics_tag, font->coords, font->num_coords)
+#else
+#define GET_VAR .0f
+#endif
+#define GET_METRIC_X(TABLE, ATTR) \
+  (face->table.TABLE->has_data () && \
+    (position && (*position = font->em_scalef_x (_fix_ascender_descender ( \
+      face->table.TABLE->ATTR + GET_VAR, metrics_tag))), true))
+#define GET_METRIC_Y(TABLE, ATTR) \
+  (face->table.TABLE->has_data () && \
+    (position && (*position = font->em_scalef_y (_fix_ascender_descender ( \
+      face->table.TABLE->ATTR + GET_VAR, metrics_tag))), true))
+  case HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER:
+    return (face->table.OS2->use_typo_metrics () && GET_METRIC_Y (OS2, sTypoAscender)) ||
+	   GET_METRIC_Y (hhea, ascender);
+  case HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER:
+    return (face->table.OS2->use_typo_metrics () && GET_METRIC_Y (OS2, sTypoDescender)) ||
+	   GET_METRIC_Y (hhea, descender);
+  case HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP:
+    return (face->table.OS2->use_typo_metrics () && GET_METRIC_Y (OS2, sTypoLineGap)) ||
+	   GET_METRIC_Y (hhea, lineGap);
+  case HB_OT_METRICS_TAG_VERTICAL_ASCENDER:  return GET_METRIC_X (vhea, ascender);
+  case HB_OT_METRICS_TAG_VERTICAL_DESCENDER: return GET_METRIC_X (vhea, descender);
+  case HB_OT_METRICS_TAG_VERTICAL_LINE_GAP:  return GET_METRIC_X (vhea, lineGap);
+#undef GET_METRIC_Y
+#undef GET_METRIC_X
+#undef GET_VAR
+  default:                               assert (0); return false;
+  }
+}
+
+#ifndef HB_NO_METRICS
+
+#if 0
+static bool
+_get_gasp (hb_face_t *face, float *result, hb_ot_metrics_tag_t metrics_tag)
+{
+  const OT::GaspRange& range = face->table.gasp->get_gasp_range (metrics_tag - HB_TAG ('g','s','p','0'));
+  if (&range == &Null (OT::GaspRange)) return false;
+  if (result) *result = range.rangeMaxPPEM + font->face->table.MVAR->get_var (metrics_tag, font->coords, font->num_coords);
+  return true;
+}
+#endif
+
+/* Private tags for https://github.com/harfbuzz/harfbuzz/issues/1866 */
+#define _HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER_OS2   HB_TAG ('O','a','s','c')
+#define _HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER_HHEA  HB_TAG ('H','a','s','c')
+#define _HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER_OS2  HB_TAG ('O','d','s','c')
+#define _HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER_HHEA HB_TAG ('H','d','s','c')
+#define _HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP_OS2   HB_TAG ('O','l','g','p')
+#define _HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP_HHEA  HB_TAG ('H','l','g','p')
+
+/**
+ * hb_ot_metrics_get_position:
+ * @font: a #hb_font_t object.
+ * @metrics_tag: tag of metrics value you like to fetch.
+ * @position: (out) (optional): result of metrics value from the font.
+ *
+ * It fetches metrics value corresponding to a given tag from a font.
+ *
+ * Returns: Whether found the requested metrics in the font.
+ * Since: 2.6.0
+ **/
+hb_bool_t
+hb_ot_metrics_get_position (hb_font_t           *font,
+			    hb_ot_metrics_tag_t  metrics_tag,
+			    hb_position_t       *position     /* OUT.  May be NULL. */)
+{
+  hb_face_t *face = font->face;
+  switch ((unsigned) metrics_tag)
+  {
+  case HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER:
+  case HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER:
+  case HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP:
+  case HB_OT_METRICS_TAG_VERTICAL_ASCENDER:
+  case HB_OT_METRICS_TAG_VERTICAL_DESCENDER:
+  case HB_OT_METRICS_TAG_VERTICAL_LINE_GAP:           return _hb_ot_metrics_get_position_common (font, metrics_tag, position);
+#ifndef HB_NO_VAR
+#define GET_VAR hb_ot_metrics_get_variation (font, metrics_tag)
+#else
+#define GET_VAR 0
+#endif
+#define GET_METRIC_X(TABLE, ATTR) \
+  (face->table.TABLE->has_data () && \
+    (position && (*position = font->em_scalef_x (face->table.TABLE->ATTR + GET_VAR)), true))
+#define GET_METRIC_Y(TABLE, ATTR) \
+  (face->table.TABLE->has_data () && \
+    (position && (*position = font->em_scalef_y (face->table.TABLE->ATTR + GET_VAR)), true))
+  case HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_ASCENT:  return GET_METRIC_Y (OS2, usWinAscent);
+  case HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_DESCENT: return GET_METRIC_Y (OS2, usWinDescent);
+  case HB_OT_METRICS_TAG_HORIZONTAL_CARET_RISE:       return GET_METRIC_Y (hhea, caretSlopeRise);
+  case HB_OT_METRICS_TAG_HORIZONTAL_CARET_RUN:        return GET_METRIC_X (hhea, caretSlopeRun);
+  case HB_OT_METRICS_TAG_HORIZONTAL_CARET_OFFSET:     return GET_METRIC_X (hhea, caretOffset);
+  case HB_OT_METRICS_TAG_VERTICAL_CARET_RISE:         return GET_METRIC_X (vhea, caretSlopeRise);
+  case HB_OT_METRICS_TAG_VERTICAL_CARET_RUN:          return GET_METRIC_Y (vhea, caretSlopeRun);
+  case HB_OT_METRICS_TAG_VERTICAL_CARET_OFFSET:       return GET_METRIC_Y (vhea, caretOffset);
+  case HB_OT_METRICS_TAG_X_HEIGHT:                    return GET_METRIC_Y (OS2->v2 (), sxHeight);
+  case HB_OT_METRICS_TAG_CAP_HEIGHT:                  return GET_METRIC_Y (OS2->v2 (), sCapHeight);
+  case HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_SIZE:         return GET_METRIC_X (OS2, ySubscriptXSize);
+  case HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_SIZE:         return GET_METRIC_Y (OS2, ySubscriptYSize);
+  case HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_OFFSET:       return GET_METRIC_X (OS2, ySubscriptXOffset);
+  case HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_OFFSET:       return GET_METRIC_Y (OS2, ySubscriptYOffset);
+  case HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_SIZE:       return GET_METRIC_X (OS2, ySuperscriptXSize);
+  case HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_SIZE:       return GET_METRIC_Y (OS2, ySuperscriptYSize);
+  case HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_OFFSET:     return GET_METRIC_X (OS2, ySuperscriptXOffset);
+  case HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_OFFSET:     return GET_METRIC_Y (OS2, ySuperscriptYOffset);
+  case HB_OT_METRICS_TAG_STRIKEOUT_SIZE:              return GET_METRIC_Y (OS2, yStrikeoutSize);
+  case HB_OT_METRICS_TAG_STRIKEOUT_OFFSET:            return GET_METRIC_Y (OS2, yStrikeoutPosition);
+  case HB_OT_METRICS_TAG_UNDERLINE_SIZE:              return GET_METRIC_Y (post->table, underlineThickness);
+  case HB_OT_METRICS_TAG_UNDERLINE_OFFSET:            return GET_METRIC_Y (post->table, underlinePosition);
+
+  /* Private tags */
+  case _HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER_OS2:    return GET_METRIC_Y (OS2, sTypoAscender);
+  case _HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER_HHEA:   return GET_METRIC_Y (hhea, ascender);
+  case _HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER_OS2:   return GET_METRIC_Y (OS2, sTypoDescender);
+  case _HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER_HHEA:  return GET_METRIC_Y (hhea, descender);
+  case _HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP_OS2:    return GET_METRIC_Y (OS2, sTypoLineGap);
+  case _HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP_HHEA:   return GET_METRIC_Y (hhea, lineGap);
+#undef GET_METRIC_Y
+#undef GET_METRIC_X
+#undef GET_VAR
+  default:                                        return false;
+  }
+}
+
+#ifndef HB_NO_VAR
+/**
+ * hb_ot_metrics_get_variation:
+ * @font:
+ * @metrics_tag:
+ *
+ * Returns:
+ *
+ * Since: 2.6.0
+ **/
+float
+hb_ot_metrics_get_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag)
+{
+  return font->face->table.MVAR->get_var (metrics_tag, font->coords, font->num_coords);
+}
+
+/**
+ * hb_ot_metrics_get_x_variation:
+ * @font:
+ * @metrics_tag:
+ *
+ * Returns:
+ *
+ * Since: 2.6.0
+ **/
+hb_position_t
+hb_ot_metrics_get_x_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag)
+{
+  return font->em_scalef_x (hb_ot_metrics_get_variation (font, metrics_tag));
+}
+
+/**
+ * hb_ot_metrics_get_y_variation:
+ * @font:
+ * @metrics_tag:
+ *
+ * Returns:
+ *
+ * Since: 2.6.0
+ **/
+hb_position_t
+hb_ot_metrics_get_y_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag)
+{
+  return font->em_scalef_y (hb_ot_metrics_get_variation (font, metrics_tag));
+}
+#endif
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-metrics.h b/thirdparty/harfbuzz/src/hb-ot-metrics.h
new file mode 100644
index 0000000000..42c7363c03
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-metrics.h
@@ -0,0 +1,122 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_METRICS_H
+#define HB_OT_METRICS_H
+
+#include "hb.h"
+#include "hb-ot-name.h"
+
+HB_BEGIN_DECLS
+
+
+/**
+ * hb_ot_metrics_tag_t:
+ * @HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER: horizontal ascender.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER: horizontal descender.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP: horizontal line gap.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_ASCENT: horizontal clipping ascent.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_DESCENT: horizontal clipping descent.
+ * @HB_OT_METRICS_TAG_VERTICAL_ASCENDER: vertical ascender.
+ * @HB_OT_METRICS_TAG_VERTICAL_DESCENDER: vertical descender.
+ * @HB_OT_METRICS_TAG_VERTICAL_LINE_GAP: vertical line gap.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_CARET_RISE: horizontal caret rise.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_CARET_RUN: horizontal caret run.
+ * @HB_OT_METRICS_TAG_HORIZONTAL_CARET_OFFSET: horizontal caret offset.
+ * @HB_OT_METRICS_TAG_VERTICAL_CARET_RISE: vertical caret rise.
+ * @HB_OT_METRICS_TAG_VERTICAL_CARET_RUN: vertical caret run.
+ * @HB_OT_METRICS_TAG_VERTICAL_CARET_OFFSET: vertical caret offset.
+ * @HB_OT_METRICS_TAG_X_HEIGHT: x height.
+ * @HB_OT_METRICS_TAG_CAP_HEIGHT: cap height.
+ * @HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_SIZE: subscript em x size.
+ * @HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_SIZE: subscript em y size.
+ * @HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_OFFSET: subscript em x offset.
+ * @HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_OFFSET: subscript em y offset.
+ * @HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_SIZE: superscript em x size.
+ * @HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_SIZE: superscript em y size.
+ * @HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_OFFSET: superscript em x offset.
+ * @HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_OFFSET: superscript em y offset.
+ * @HB_OT_METRICS_TAG_STRIKEOUT_SIZE: strikeout size.
+ * @HB_OT_METRICS_TAG_STRIKEOUT_OFFSET: strikeout offset.
+ * @HB_OT_METRICS_TAG_UNDERLINE_SIZE: underline size.
+ * @HB_OT_METRICS_TAG_UNDERLINE_OFFSET: underline offset.
+ *
+ * From https://docs.microsoft.com/en-us/typography/opentype/spec/mvar#value-tags
+ *
+ * Since: 2.6.0
+ **/
+typedef enum {
+  HB_OT_METRICS_TAG_HORIZONTAL_ASCENDER		= HB_TAG ('h','a','s','c'),
+  HB_OT_METRICS_TAG_HORIZONTAL_DESCENDER	= HB_TAG ('h','d','s','c'),
+  HB_OT_METRICS_TAG_HORIZONTAL_LINE_GAP		= HB_TAG ('h','l','g','p'),
+  HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_ASCENT	= HB_TAG ('h','c','l','a'),
+  HB_OT_METRICS_TAG_HORIZONTAL_CLIPPING_DESCENT	= HB_TAG ('h','c','l','d'),
+  HB_OT_METRICS_TAG_VERTICAL_ASCENDER		= HB_TAG ('v','a','s','c'),
+  HB_OT_METRICS_TAG_VERTICAL_DESCENDER		= HB_TAG ('v','d','s','c'),
+  HB_OT_METRICS_TAG_VERTICAL_LINE_GAP		= HB_TAG ('v','l','g','p'),
+  HB_OT_METRICS_TAG_HORIZONTAL_CARET_RISE	= HB_TAG ('h','c','r','s'),
+  HB_OT_METRICS_TAG_HORIZONTAL_CARET_RUN	= HB_TAG ('h','c','r','n'),
+  HB_OT_METRICS_TAG_HORIZONTAL_CARET_OFFSET	= HB_TAG ('h','c','o','f'),
+  HB_OT_METRICS_TAG_VERTICAL_CARET_RISE		= HB_TAG ('v','c','r','s'),
+  HB_OT_METRICS_TAG_VERTICAL_CARET_RUN		= HB_TAG ('v','c','r','n'),
+  HB_OT_METRICS_TAG_VERTICAL_CARET_OFFSET	= HB_TAG ('v','c','o','f'),
+  HB_OT_METRICS_TAG_X_HEIGHT			= HB_TAG ('x','h','g','t'),
+  HB_OT_METRICS_TAG_CAP_HEIGHT			= HB_TAG ('c','p','h','t'),
+  HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_SIZE		= HB_TAG ('s','b','x','s'),
+  HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_SIZE		= HB_TAG ('s','b','y','s'),
+  HB_OT_METRICS_TAG_SUBSCRIPT_EM_X_OFFSET	= HB_TAG ('s','b','x','o'),
+  HB_OT_METRICS_TAG_SUBSCRIPT_EM_Y_OFFSET	= HB_TAG ('s','b','y','o'),
+  HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_SIZE	= HB_TAG ('s','p','x','s'),
+  HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_SIZE	= HB_TAG ('s','p','y','s'),
+  HB_OT_METRICS_TAG_SUPERSCRIPT_EM_X_OFFSET	= HB_TAG ('s','p','x','o'),
+  HB_OT_METRICS_TAG_SUPERSCRIPT_EM_Y_OFFSET	= HB_TAG ('s','p','y','o'),
+  HB_OT_METRICS_TAG_STRIKEOUT_SIZE		= HB_TAG ('s','t','r','s'),
+  HB_OT_METRICS_TAG_STRIKEOUT_OFFSET		= HB_TAG ('s','t','r','o'),
+  HB_OT_METRICS_TAG_UNDERLINE_SIZE		= HB_TAG ('u','n','d','s'),
+  HB_OT_METRICS_TAG_UNDERLINE_OFFSET		= HB_TAG ('u','n','d','o'),
+
+  _HB_OT_METRICS_TAG_MAX_VALUE = HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_ot_metrics_tag_t;
+
+HB_EXTERN hb_bool_t
+hb_ot_metrics_get_position (hb_font_t           *font,
+			    hb_ot_metrics_tag_t  metrics_tag,
+			    hb_position_t       *position     /* OUT.  May be NULL. */);
+
+HB_EXTERN float
+hb_ot_metrics_get_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag);
+
+HB_EXTERN hb_position_t
+hb_ot_metrics_get_x_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag);
+
+HB_EXTERN hb_position_t
+hb_ot_metrics_get_y_variation (hb_font_t *font, hb_ot_metrics_tag_t metrics_tag);
+
+HB_END_DECLS
+
+#endif /* HB_OT_METRICS_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-metrics.hh b/thirdparty/harfbuzz/src/hb-ot-metrics.hh
new file mode 100644
index 0000000000..19a5e9ed41
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-metrics.hh
@@ -0,0 +1,35 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_METRICS_HH
+#define HB_OT_METRICS_HH
+
+#include "hb.hh"
+
+HB_INTERNAL bool
+_hb_ot_metrics_get_position_common (hb_font_t           *font,
+				    hb_ot_metrics_tag_t  metrics_tag,
+				    hb_position_t       *position     /* OUT.  May be NULL. */);
+
+#endif /* HB_OT_METRICS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-name-language-static.hh b/thirdparty/harfbuzz/src/hb-ot-name-language-static.hh
new file mode 100644
index 0000000000..c496dc2981
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-name-language-static.hh
@@ -0,0 +1,456 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_NAME_LANGUAGE_STATIC_HH
+#define HB_OT_NAME_LANGUAGE_STATIC_HH
+
+#include "hb-ot-name-language.hh"
+
+/* Following two tables were generated by joining FreeType, FontConfig,
+ * and OpenType specification language lists, then filled in missing
+ * entries using:
+ * https://docs.microsoft.com/en-us/windows/desktop/intl/language-identifier-constants-and-strings
+ */
+
+struct hb_ot_language_map_t
+{
+  int cmp (unsigned int key) const
+  { return key < code ? -1 : key > code ? +1 : 0; }
+
+  uint16_t	code;
+  char		lang[6];
+};
+
+static const hb_ot_language_map_t
+hb_ms_language_map[] =
+{
+  {0x0001,	"ar"},	/* ??? */
+  {0x0004,	"zh"},	/* ??? */
+  {0x0009,	"en"},	/* ??? */
+  {0x0401,	"ar"},	/* Arabic (Saudi Arabia) */
+  {0x0402,	"bg"},	/* Bulgarian (Bulgaria) */
+  {0x0403,	"ca"},	/* Catalan (Catalan) */
+  {0x0404,	"zh-tw"},	/* Chinese (Taiwan) */
+  {0x0405,	"cs"},	/* Czech (Czech Republic) */
+  {0x0406,	"da"},	/* Danish (Denmark) */
+  {0x0407,	"de"},	/* German (Germany) */
+  {0x0408,	"el"},	/* Greek (Greece) */
+  {0x0409,	"en"},	/* English (United States) */
+  {0x040A,	"es"},	/* Spanish (Traditional Sort) (Spain) */
+  {0x040B,	"fi"},	/* Finnish (Finland) */
+  {0x040C,	"fr"},	/* French (France) */
+  {0x040D,	"he"},	/* Hebrew (Israel) */
+  {0x040E,	"hu"},	/* Hungarian (Hungary) */
+  {0x040F,	"is"},	/* Icelandic (Iceland) */
+  {0x0410,	"it"},	/* Italian (Italy) */
+  {0x0411,	"ja"},	/* Japanese (Japan) */
+  {0x0412,	"ko"},	/* Korean (Korea) */
+  {0x0413,	"nl"},	/* Dutch (Netherlands) */
+  {0x0414,	"no"},	/* Norwegian (Bokmal) (Norway) */
+  {0x0415,	"pl"},	/* Polish (Poland) */
+  {0x0416,	"pt"},	/* Portuguese (Brazil) */
+  {0x0417,	"rm"},	/* Romansh (Switzerland) */
+  {0x0418,	"ro"},	/* Romanian (Romania) */
+  {0x0419,	"ru"},	/* Russian (Russia) */
+  {0x041A,	"hr"},	/* Croatian (Croatia) */
+  {0x041B,	"sk"},	/* Slovak (Slovakia) */
+  {0x041C,	"sq"},	/* Albanian (Albania) */
+  {0x041D,	"sv"},	/* Swedish (Sweden) */
+  {0x041E,	"th"},	/* Thai (Thailand) */
+  {0x041F,	"tr"},	/* Turkish (Turkey) */
+  {0x0420,	"ur"},	/* Urdu (Islamic Republic of Pakistan) */
+  {0x0421,	"id"},	/* Indonesian (Indonesia) */
+  {0x0422,	"uk"},	/* Ukrainian (Ukraine) */
+  {0x0423,	"be"},	/* Belarusian (Belarus) */
+  {0x0424,	"sl"},	/* Slovenian (Slovenia) */
+  {0x0425,	"et"},	/* Estonian (Estonia) */
+  {0x0426,	"lv"},	/* Latvian (Latvia) */
+  {0x0427,	"lt"},	/* Lithuanian (Lithuania) */
+  {0x0428,	"tg"},	/* Tajik (Cyrillic) (Tajikistan) */
+  {0x0429,	"fa"},	/* Persian (Iran) */
+  {0x042A,	"vi"},	/* Vietnamese (Vietnam) */
+  {0x042B,	"hy"},	/* Armenian (Armenia) */
+  {0x042C,	"az"},	/* Azeri (Latin) (Azerbaijan) */
+  {0x042D,	"eu"},	/* Basque (Basque) */
+  {0x042E,	"hsb"},	/* Upper Sorbian (Germany) */
+  {0x042F,	"mk"},	/* Macedonian (FYROM) (Former Yugoslav Republic of Macedonia) */
+  {0x0430,	"st"},	/* ??? */
+  {0x0431,	"ts"},	/* ??? */
+  {0x0432,	"tn"},	/* Setswana (South Africa) */
+  {0x0433,	"ven"},	/* ??? */
+  {0x0434,	"xh"},	/* isiXhosa (South Africa) */
+  {0x0435,	"zu"},	/* isiZulu (South Africa) */
+  {0x0436,	"af"},	/* Afrikaans (South Africa) */
+  {0x0437,	"ka"},	/* Georgian (Georgia) */
+  {0x0438,	"fo"},	/* Faroese (Faroe Islands) */
+  {0x0439,	"hi"},	/* Hindi (India) */
+  {0x043A,	"mt"},	/* Maltese (Malta) */
+  {0x043B,	"se"},	/* Sami (Northern) (Norway) */
+  {0x043C,	"ga"},	/* ??? */
+  {0x043D,	"yi"},	/* ??? */
+  {0x043E,	"ms"},	/* Malay (Malaysia) */
+  {0x043F,	"kk"},	/* Kazakh (Kazakhstan) */
+  {0x0440,	"ky"},	/* Kyrgyz (Kyrgyzstan) */
+  {0x0441,	"sw"},	/* Kiswahili (Kenya) */
+  {0x0442,	"tk"},	/* Turkmen (Turkmenistan) */
+  {0x0443,	"uz"},	/* Uzbek (Latin) (Uzbekistan) */
+  {0x0444,	"tt"},	/* Tatar (Russia) */
+  {0x0445,	"bn"},	/* Bengali (India) */
+  {0x0446,	"pa"},	/* Punjabi (India) */
+  {0x0447,	"gu"},	/* Gujarati (India) */
+  {0x0448,	"or"},	/* Odia (formerly Oriya) (India) */
+  {0x0449,	"ta"},	/* Tamil (India) */
+  {0x044A,	"te"},	/* Telugu (India) */
+  {0x044B,	"kn"},	/* Kannada (India) */
+  {0x044C,	"ml"},	/* Malayalam (India) */
+  {0x044D,	"as"},	/* Assamese (India) */
+  {0x044E,	"mr"},	/* Marathi (India) */
+  {0x044F,	"sa"},	/* Sanskrit (India) */
+  {0x0450,	"mn"},	/* Mongolian (Cyrillic) (Mongolia) */
+  {0x0451,	"bo"},	/* Tibetan (PRC) */
+  {0x0452,	"cy"},	/* Welsh (United Kingdom) */
+  {0x0453,	"km"},	/* Khmer (Cambodia) */
+  {0x0454,	"lo"},	/* Lao (Lao P.D.R.) */
+  {0x0455,	"my"},	/* ??? */
+  {0x0456,	"gl"},	/* Galician (Galician) */
+  {0x0457,	"kok"},	/* Konkani (India) */
+  {0x0458,	"mni"},	/* ??? */
+  {0x0459,	"sd"},	/* ??? */
+  {0x045A,	"syr"},	/* Syriac (Syria) */
+  {0x045B,	"si"},	/* Sinhala (Sri Lanka) */
+  {0x045C,	"chr"},	/* ??? */
+  {0x045D,	"iu"},	/* Inuktitut (Canada) */
+  {0x045E,	"am"},	/* Amharic (Ethiopia) */
+  {0x0460,	"ks"},	/* ??? */
+  {0x0461,	"ne"},	/* Nepali (Nepal) */
+  {0x0462,	"fy"},	/* Frisian (Netherlands) */
+  {0x0463,	"ps"},	/* Pashto (Afghanistan) */
+  {0x0464,	"phi"},	/* Filipino (Philippines) */
+  {0x0465,	"div"},	/* Divehi (Maldives) */
+  {0x0468,	"ha"},	/* Hausa (Latin) (Nigeria) */
+  {0x046A,	"yo"},	/* Yoruba (Nigeria) */
+  {0x046B,	"quz"},	/* Quechua (Bolivia) */
+  {0x046C,	"nso"},	/* Sesotho sa Leboa (South Africa) */
+  {0x046D,	"ba"},	/* Bashkir (Russia) */
+  {0x046E,	"lb"},	/* Luxembourgish (Luxembourg) */
+  {0x046F,	"kl"},	/* Greenlandic (Greenland) */
+  {0x0470,	"ibo"},	/* Igbo (Nigeria) */
+  {0x0471,	"kau"},	/* ??? */
+  {0x0472,	"om"},	/* ??? */
+  {0x0473,	"ti"},	/* ??? */
+  {0x0474,	"gn"},	/* ??? */
+  {0x0475,	"haw"},	/* ??? */
+  {0x0476,	"la"},	/* ??? */
+  {0x0477,	"so"},	/* ??? */
+  {0x0478,	"ii"},	/* Yi (PRC) */
+  {0x0479,	"pap"},	/* ??? */
+  {0x047A,	"arn"},	/* Mapudungun (Chile) */
+  {0x047C,	"moh"},	/* Mohawk (Mohawk) */
+  {0x047E,	"br"},	/* Breton (France) */
+  {0x0480,	"ug"},	/* Uighur (PRC) */
+  {0x0481,	"mi"},	/* Maori (New Zealand) */
+  {0x0482,	"oc"},	/* Occitan (France) */
+  {0x0483,	"co"},	/* Corsican (France) */
+  {0x0484,	"gsw"},	/* Alsatian (France) */
+  {0x0485,	"sah"},	/* Yakut (Russia) */
+  {0x0486,	"qut"},	/* K'iche (Guatemala) */
+  {0x0487,	"rw"},	/* Kinyarwanda (Rwanda) */
+  {0x0488,	"wo"},	/* Wolof (Senegal) */
+  {0x048C,	"fa"},	/* Dari (Afghanistan) */
+  {0x0801,	"ar"},	/* Arabic (Iraq) */
+  {0x0804,	"zh-cn"},	/* Chinese (People’s Republic of China) */
+  {0x0807,	"de"},	/* German (Switzerland) */
+  {0x0809,	"en"},	/* English (United Kingdom) */
+  {0x080A,	"es"},	/* Spanish (Mexico) */
+  {0x080C,	"fr"},	/* French (Belgium) */
+  {0x0810,	"it"},	/* Italian (Switzerland) */
+  {0x0812,	"ko"},	/* ??? */
+  {0x0813,	"nl"},	/* Dutch (Belgium) */
+  {0x0814,	"nn"},	/* Norwegian (Nynorsk) (Norway) */
+  {0x0816,	"pt"},	/* Portuguese (Portugal) */
+  {0x0818,	"mo"},	/* ??? */
+  {0x0819,	"ru"},	/* ??? */
+  {0x081A,	"sr"},	/* Serbian (Latin) (Serbia) */
+  {0x081D,	"sv"},	/* Sweden (Finland) */
+  {0x0820,	"ur"},	/* ??? */
+  {0x0827,	"lt"},	/* ??? */
+  {0x082C,	"az"},	/* Azeri (Cyrillic) (Azerbaijan) */
+  {0x082E,	"dsb"},	/* Lower Sorbian (Germany) */
+//{0x083B,	""},	/* Sami (Northern) (Sweden) */
+  {0x083C,	"gd"},	/* Irish (Ireland) */
+  {0x083E,	"ms"},	/* Malay (Brunei Darussalam) */
+  {0x0843,	"uz"},	/* Uzbek (Cyrillic) (Uzbekistan) */
+  {0x0845,	"bn"},	/* Bengali (Bangladesh) */
+  {0x0846,	"ar"},	/* ??? */
+  {0x0850,	"mn"},	/* Mongolian (Traditional) (People’s Republic of China) */
+  {0x0851,	"dz"},	/* ??? */
+  {0x085D,	"iu"},	/* Inuktitut (Latin) (Canada) */
+  {0x085F,	"tzm"},	/* Tamazight (Latin) (Algeria) */
+  {0x0861,	"ne"},	/* ??? */
+//{0x086B,	""},	/* Quechua (Ecuador) */
+  {0x0873,	"ti"},	/* ??? */
+  {0x0C01,	"ar"},	/* Arabic (Egypt) */
+  {0x0C04,	"zh-hk"},	/* Chinese (Hong Kong S.A.R.) */
+  {0x0C07,	"de"},	/* German (Austria) */
+  {0x0C09,	"en"},	/* English (Australia) */
+  {0x0C0A,	"es"},	/* Spanish (Modern Sort) (Spain) */
+  {0x0C0C,	"fr"},	/* French (Canada) */
+  {0x0C1A,	"sr"},	/* Serbian (Cyrillic) (Serbia) */
+  {0x0C3B,	"se"},	/* Sami (Northern) (Finland) */
+//{0x0C6B,	""},	/* Quechua (Peru) */
+  {0x1001,	"ar"},	/* Arabic (Libya) */
+  {0x1004,	"zh-sg"},	/* Chinese (Singapore) */
+  {0x1007,	"de"},	/* German (Luxembourg) */
+  {0x1009,	"en"},	/* English (Canada) */
+  {0x100A,	"es"},	/* Spanish (Guatemala) */
+  {0x100C,	"fr"},	/* French (Switzerland) */
+  {0x101A,	"hr"},	/* Croatian (Latin) (Bosnia and Herzegovina) */
+  {0x103B,	"smj"},	/* Sami (Lule) (Norway) */
+  {0x1401,	"ar"},	/* Arabic (Algeria) */
+//{0x1404,	""},	/* Chinese (Macao S.A.R.) */
+  {0x1407,	"de"},	/* German (Liechtenstein) */
+  {0x1409,	"en"},	/* English (New Zealand) */
+  {0x140A,	"es"},	/* Spanish (Costa Rica) */
+  {0x140C,	"fr"},	/* French (Luxembourg) */
+  {0x141A,	"bs"},	/* Bosnian (Latin) (Bosnia and Herzegovina) */
+//{0x143B,	""},	/* Sami (Lule) (Sweden) */
+  {0x1801,	"ar"},	/* Arabic (Morocco) */
+  {0x1809,	"en"},	/* English (Ireland) */
+  {0x180A,	"es"},	/* Spanish (Panama) */
+  {0x180C,	"fr"},	/* French (Principality of Monaco) */
+//{0x181A,	""},	/* Serbian (Latin) (Bosnia and Herzegovina) */
+  {0x183B,	"sma"},	/* Sami (Southern) (Norway) */
+  {0x1C01,	"ar"},	/* Arabic (Tunisia) */
+  {0x1C09,	"en"},	/* English (South Africa) */
+  {0x1C0A,	"es"},	/* Spanish (Dominican Republic) */
+  {0x1C0C,	"fr"},	/* ??? */
+//{0x1C1A,	""},	/* Serbian (Cyrillic) (Bosnia and Herzegovina) */
+//{0x1C3B,	""},	/* Sami (Southern) (Sweden) */
+  {0x2001,	"ar"},	/* Arabic (Oman) */
+  {0x2009,	"en"},	/* English (Jamaica) */
+  {0x200A,	"es"},	/* Spanish (Venezuela) */
+  {0x200C,	"fr"},	/* ??? */
+  {0x201A,	"bs"},	/* Bosnian (Cyrillic) (Bosnia and Herzegovina) */
+  {0x203B,	"sms"},	/* Sami (Skolt) (Finland) */
+  {0x2401,	"ar"},	/* Arabic (Yemen) */
+  {0x2409,	"en"},	/* English (Caribbean) */
+  {0x240A,	"es"},	/* Spanish (Colombia) */
+  {0x240C,	"fr"},	/* ??? */
+  {0x243B,	"smn"},	/* Sami (Inari) (Finland) */
+  {0x2801,	"ar"},	/* Arabic (Syria) */
+  {0x2809,	"en"},	/* English (Belize) */
+  {0x280A,	"es"},	/* Spanish (Peru) */
+  {0x280C,	"fr"},	/* ??? */
+  {0x2C01,	"ar"},	/* Arabic (Jordan) */
+  {0x2C09,	"en"},	/* English (Trinidad and Tobago) */
+  {0x2C0A,	"es"},	/* Spanish (Argentina) */
+  {0x2C0C,	"fr"},	/* ??? */
+  {0x3001,	"ar"},	/* Arabic (Lebanon) */
+  {0x3009,	"en"},	/* English (Zimbabwe) */
+  {0x300A,	"es"},	/* Spanish (Ecuador) */
+  {0x300C,	"fr"},	/* ??? */
+  {0x3401,	"ar"},	/* Arabic (Kuwait) */
+  {0x3409,	"en"},	/* English (Republic of the Philippines) */
+  {0x340A,	"es"},	/* Spanish (Chile) */
+  {0x340C,	"fr"},	/* ??? */
+  {0x3801,	"ar"},	/* Arabic (U.A.E.) */
+  {0x380A,	"es"},	/* Spanish (Uruguay) */
+  {0x380C,	"fr"},	/* ??? */
+  {0x3C01,	"ar"},	/* Arabic (Bahrain) */
+  {0x3C09,	"en"},	/* ??? */
+  {0x3C0A,	"es"},	/* Spanish (Paraguay) */
+  {0x3C0C,	"fr"},	/* ??? */
+  {0x4001,	"ar"},	/* Arabic (Qatar) */
+  {0x4009,	"en"},	/* English (India) */
+  {0x400A,	"es"},	/* Spanish (Bolivia) */
+  {0x4409,	"en"},	/* English (Malaysia) */
+  {0x440A,	"es"},	/* Spanish (El Salvador) */
+  {0x4809,	"en"},	/* English (Singapore) */
+  {0x480A,	"es"},	/* Spanish (Honduras) */
+  {0x4C0A,	"es"},	/* Spanish (Nicaragua) */
+  {0x500A,	"es"},	/* Spanish (Puerto Rico) */
+  {0x540A,	"es"},	/* Spanish (United States) */
+  {0xE40A,	"es"},	/* ??? */
+  {0xE40C,	"fr"},	/* ??? */
+};
+
+static const hb_ot_language_map_t
+hb_mac_language_map[] =
+{
+  {  0,	"en"},	/* English */
+  {  1,	"fr"},	/* French */
+  {  2,	"de"},	/* German */
+  {  3,	"it"},	/* Italian */
+  {  4,	"nl"},	/* Dutch */
+  {  5,	"sv"},	/* Swedish */
+  {  6,	"es"},	/* Spanish */
+  {  7,	"da"},	/* Danish */
+  {  8,	"pt"},	/* Portuguese */
+  {  9,	"no"},	/* Norwegian */
+  { 10,	"he"},	/* Hebrew */
+  { 11,	"ja"},	/* Japanese */
+  { 12,	"ar"},	/* Arabic */
+  { 13,	"fi"},	/* Finnish */
+  { 14,	"el"},	/* Greek */
+  { 15,	"is"},	/* Icelandic */
+  { 16,	"mt"},	/* Maltese */
+  { 17,	"tr"},	/* Turkish */
+  { 18,	"hr"},	/* Croatian */
+  { 19,	"zh-tw"},	/* Chinese (Traditional) */
+  { 20,	"ur"},	/* Urdu */
+  { 21,	"hi"},	/* Hindi */
+  { 22,	"th"},	/* Thai */
+  { 23,	"ko"},	/* Korean */
+  { 24,	"lt"},	/* Lithuanian */
+  { 25,	"pl"},	/* Polish */
+  { 26,	"hu"},	/* Hungarian */
+  { 27,	"et"},	/* Estonian */
+  { 28,	"lv"},	/* Latvian */
+//{ 29,	""},	/* Sami */
+  { 30,	"fo"},	/* Faroese */
+  { 31,	"fa"},	/* Farsi/Persian */
+  { 32,	"ru"},	/* Russian */
+  { 33,	"zh-cn"},	/* Chinese (Simplified) */
+  { 34,	"nl"},	/* Flemish */
+  { 35,	"ga"},	/* Irish Gaelic */
+  { 36,	"sq"},	/* Albanian */
+  { 37,	"ro"},	/* Romanian */
+  { 38,	"cs"},	/* Czech */
+  { 39,	"sk"},	/* Slovak */
+  { 40,	"sl"},	/* Slovenian */
+  { 41,	"yi"},	/* Yiddish */
+  { 42,	"sr"},	/* Serbian */
+  { 43,	"mk"},	/* Macedonian */
+  { 44,	"bg"},	/* Bulgarian */
+  { 45,	"uk"},	/* Ukrainian */
+  { 46,	"be"},	/* Byelorussian */
+  { 47,	"uz"},	/* Uzbek */
+  { 48,	"kk"},	/* Kazakh */
+  { 49,	"az"},	/* Azerbaijani (Cyrillic script) */
+  { 50,	"az"},	/* Azerbaijani (Arabic script) */
+  { 51,	"hy"},	/* Armenian */
+  { 52,	"ka"},	/* Georgian */
+  { 53,	"mo"},	/* Moldavian */
+  { 54,	"ky"},	/* Kirghiz */
+  { 55,	"tg"},	/* Tajiki */
+  { 56,	"tk"},	/* Turkmen */
+  { 57,	"mn"},	/* Mongolian (Mongolian script) */
+  { 58,	"mn"},	/* Mongolian (Cyrillic script) */
+  { 59,	"ps"},	/* Pashto */
+  { 60,	"ku"},	/* Kurdish */
+  { 61,	"ks"},	/* Kashmiri */
+  { 62,	"sd"},	/* Sindhi */
+  { 63,	"bo"},	/* Tibetan */
+  { 64,	"ne"},	/* Nepali */
+  { 65,	"sa"},	/* Sanskrit */
+  { 66,	"mr"},	/* Marathi */
+  { 67,	"bn"},	/* Bengali */
+  { 68,	"as"},	/* Assamese */
+  { 69,	"gu"},	/* Gujarati */
+  { 70,	"pa"},	/* Punjabi */
+  { 71,	"or"},	/* Oriya */
+  { 72,	"ml"},	/* Malayalam */
+  { 73,	"kn"},	/* Kannada */
+  { 74,	"ta"},	/* Tamil */
+  { 75,	"te"},	/* Telugu */
+  { 76,	"si"},	/* Sinhalese */
+  { 77,	"my"},	/* Burmese */
+  { 78,	"km"},	/* Khmer */
+  { 79,	"lo"},	/* Lao */
+  { 80,	"vi"},	/* Vietnamese */
+  { 81,	"id"},	/* Indonesian */
+  { 82,	"tl"},	/* Tagalog */
+  { 83,	"ms"},	/* Malay (Roman script) */
+  { 84,	"ms"},	/* Malay (Arabic script) */
+  { 85,	"am"},	/* Amharic */
+  { 86,	"ti"},	/* Tigrinya */
+  { 87,	"om"},	/* Galla */
+  { 88,	"so"},	/* Somali */
+  { 89,	"sw"},	/* Swahili */
+  { 90,	"rw"},	/* Kinyarwanda/Ruanda */
+  { 91,	"rn"},	/* Rundi */
+  { 92,	"ny"},	/* Nyanja/Chewa */
+  { 93,	"mg"},	/* Malagasy */
+  { 94,	"eo"},	/* Esperanto */
+  {128,	"cy"},	/* Welsh */
+  {129,	"eu"},	/* Basque */
+  {130,	"ca"},	/* Catalan */
+  {131,	"la"},	/* Latin */
+  {132,	"qu"},	/* Quechua */
+  {133,	"gn"},	/* Guarani */
+  {134,	"ay"},	/* Aymara */
+  {135,	"tt"},	/* Tatar */
+  {136,	"ug"},	/* Uighur */
+  {137,	"dz"},	/* Dzongkha */
+  {138,	"jw"},	/* Javanese (Roman script) */
+  {139,	"su"},	/* Sundanese (Roman script) */
+  {140,	"gl"},	/* Galician */
+  {141,	"af"},	/* Afrikaans */
+  {142,	"br"},	/* Breton */
+  {143,	"iu"},	/* Inuktitut */
+  {144,	"gd"},	/* Scottish Gaelic */
+  {145,	"gv"},	/* Manx Gaelic */
+  {146,	"ga"},	/* Irish Gaelic (with dot above) */
+  {147,	"to"},	/* Tongan */
+  {148,	"el"},	/* Greek (polytonic) */
+  {149,	"ik"},	/* Greenlandic */
+  {150,	"az"},	/* Azerbaijani (Roman script) */
+};
+
+
+static hb_language_t
+_hb_ot_name_language_for (unsigned int code,
+			  const hb_ot_language_map_t *array,
+			  unsigned int len)
+{
+#ifdef HB_NO_OT_NAME_LANGUAGE
+  return HB_LANGUAGE_INVALID;
+#endif
+  auto *entry = hb_bsearch (code, array, len);
+
+  if (entry)
+    return hb_language_from_string (entry->lang, -1);
+
+  return HB_LANGUAGE_INVALID;
+}
+
+hb_language_t
+_hb_ot_name_language_for_ms_code (unsigned int code)
+{
+  return _hb_ot_name_language_for (code,
+				   hb_ms_language_map,
+				   ARRAY_LENGTH (hb_ms_language_map));
+}
+
+hb_language_t
+_hb_ot_name_language_for_mac_code (unsigned int code)
+{
+  return _hb_ot_name_language_for (code,
+				   hb_mac_language_map,
+				   ARRAY_LENGTH (hb_mac_language_map));
+}
+
+#endif /* HB_OT_NAME_LANGUAGE_STATIC_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-name-language.hh b/thirdparty/harfbuzz/src/hb-ot-name-language.hh
new file mode 100644
index 0000000000..903076c0d5
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-name-language.hh
@@ -0,0 +1,40 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_NAME_LANGUAGE_HH
+#define HB_OT_NAME_LANGUAGE_HH
+
+#include "hb.hh"
+
+
+HB_INTERNAL hb_language_t
+_hb_ot_name_language_for_ms_code (unsigned int code);
+
+HB_INTERNAL hb_language_t
+_hb_ot_name_language_for_mac_code (unsigned int code);
+
+
+#endif /* HB_OT_NAME_LANGUAGE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-name-table.hh b/thirdparty/harfbuzz/src/hb-ot-name-table.hh
new file mode 100644
index 0000000000..ece3c28466
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-name-table.hh
@@ -0,0 +1,376 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_NAME_TABLE_HH
+#define HB_OT_NAME_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-name-language.hh"
+#include "hb-aat-layout.hh"
+
+
+namespace OT {
+
+
+#define entry_score var.u16[0]
+#define entry_index var.u16[1]
+
+
+/*
+ * name -- Naming
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/name
+ */
+#define HB_OT_TAG_name HB_TAG('n','a','m','e')
+
+#define UNSUPPORTED	42
+
+struct NameRecord
+{
+  hb_language_t language (hb_face_t *face) const
+  {
+#ifndef HB_NO_OT_NAME_LANGUAGE
+    unsigned int p = platformID;
+    unsigned int l = languageID;
+
+    if (p == 3)
+      return _hb_ot_name_language_for_ms_code (l);
+
+    if (p == 1)
+      return _hb_ot_name_language_for_mac_code (l);
+
+#ifndef HB_NO_OT_NAME_LANGUAGE_AAT
+    if (p == 0)
+      return face->table.ltag->get_language (l);
+#endif
+
+#endif
+    return HB_LANGUAGE_INVALID;
+  }
+
+  uint16_t score () const
+  {
+    /* Same order as in cmap::find_best_subtable(). */
+    unsigned int p = platformID;
+    unsigned int e = encodingID;
+
+    /* 32-bit. */
+    if (p == 3 && e == 10) return 0;
+    if (p == 0 && e ==  6) return 1;
+    if (p == 0 && e ==  4) return 2;
+
+    /* 16-bit. */
+    if (p == 3 && e ==  1) return 3;
+    if (p == 0 && e ==  3) return 4;
+    if (p == 0 && e ==  2) return 5;
+    if (p == 0 && e ==  1) return 6;
+    if (p == 0 && e ==  0) return 7;
+
+    /* Symbol. */
+    if (p == 3 && e ==  0) return 8;
+
+    /* We treat all Mac Latin names as ASCII only. */
+    if (p == 1 && e ==  0) return 10; /* 10 is magic number :| */
+
+    return UNSUPPORTED;
+  }
+
+  NameRecord* copy (hb_serialize_context_t *c, const void *base) const
+  {
+    TRACE_SERIALIZE (this);
+    auto *out = c->embed (this);
+    if (unlikely (!out)) return_trace (nullptr);
+    out->offset.serialize_copy (c, offset, base, 0, hb_serialize_context_t::Tail, length);
+    return_trace (out);
+  }
+
+  bool isUnicode () const
+  {
+    unsigned int p = platformID;
+    unsigned int e = encodingID;
+
+    return (p == 0 ||
+	    (p == 3 && (e == 0 || e == 1 || e == 10)));
+  }
+
+  static int cmp (const void *pa, const void *pb)
+  {
+    const NameRecord *a = (const NameRecord *)pa;
+    const NameRecord *b = (const NameRecord *)pb;
+
+    if (a->platformID != b->platformID)
+      return a->platformID - b->platformID;
+
+    if (a->encodingID != b->encodingID)
+      return a->encodingID - b->encodingID;
+
+    if (a->languageID != b->languageID)
+      return a->languageID - b->languageID;
+
+    if (a->nameID != b->nameID)
+      return a->nameID - b->nameID;
+
+    if (a->length != b->length)
+      return a->length - b->length;
+
+    return 0;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c, const void *base) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && offset.sanitize (c, base, length));
+  }
+
+  HBUINT16	platformID;	/* Platform ID. */
+  HBUINT16	encodingID;	/* Platform-specific encoding ID. */
+  HBUINT16	languageID;	/* Language ID. */
+  HBUINT16	nameID;		/* Name ID. */
+  HBUINT16	length;		/* String length (in bytes). */
+  NNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		offset;		/* String offset from start of storage area (in bytes). */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+static int
+_hb_ot_name_entry_cmp_key (const void *pa, const void *pb)
+{
+  const hb_ot_name_entry_t *a = (const hb_ot_name_entry_t *) pa;
+  const hb_ot_name_entry_t *b = (const hb_ot_name_entry_t *) pb;
+
+  /* Compare by name_id, then language. */
+
+  if (a->name_id != b->name_id)
+    return a->name_id - b->name_id;
+
+  if (a->language == b->language) return 0;
+  if (!a->language) return -1;
+  if (!b->language) return +1;
+  return strcmp (hb_language_to_string (a->language),
+		 hb_language_to_string (b->language));
+}
+
+static int
+_hb_ot_name_entry_cmp (const void *pa, const void *pb)
+{
+  /* Compare by name_id, then language, then score, then index. */
+
+  int v = _hb_ot_name_entry_cmp_key (pa, pb);
+  if (v)
+    return v;
+
+  const hb_ot_name_entry_t *a = (const hb_ot_name_entry_t *) pa;
+  const hb_ot_name_entry_t *b = (const hb_ot_name_entry_t *) pb;
+
+  if (a->entry_score != b->entry_score)
+    return a->entry_score - b->entry_score;
+
+  if (a->entry_index != b->entry_index)
+    return a->entry_index - b->entry_index;
+
+  return 0;
+}
+
+struct name
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_name;
+
+  unsigned int get_size () const
+  { return min_size + count * nameRecordZ.item_size; }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_source_of (Iterator, const NameRecord &))>
+  bool serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  const void *src_string_pool)
+  {
+    TRACE_SERIALIZE (this);
+
+    if (unlikely (!c->extend_min ((*this))))  return_trace (false);
+
+    this->format = 0;
+    this->count = it.len ();
+
+    NameRecord *name_records = (NameRecord *) calloc (it.len (), NameRecord::static_size);
+    if (unlikely (!name_records)) return_trace (false);
+
+    hb_array_t<NameRecord> records (name_records, it.len ());
+
+    for (const NameRecord& record : it)
+    {
+      memcpy (name_records, &record, NameRecord::static_size);
+      name_records++;
+    }
+
+    records.qsort ();
+
+    c->copy_all (records, src_string_pool);
+    free (records.arrayZ);
+
+    if (unlikely (c->ran_out_of_room)) return_trace (false);
+
+    this->stringOffset = c->length ();
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    name *name_prime = c->serializer->start_embed<name> ();
+    if (unlikely (!name_prime)) return_trace (false);
+
+    auto it =
+    + nameRecordZ.as_array (count)
+    | hb_filter (c->plan->name_ids, &NameRecord::nameID)
+    | hb_filter (c->plan->name_languages, &NameRecord::languageID)
+    | hb_filter ([&] (const NameRecord& namerecord) { return c->plan->name_legacy || namerecord.isUnicode (); })
+    ;
+
+    name_prime->serialize (c->serializer, it, hb_addressof (this + stringOffset));
+    return_trace (name_prime->count);
+  }
+
+  bool sanitize_records (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    const void *string_pool = (this+stringOffset).arrayZ;
+    return_trace (nameRecordZ.sanitize (c, count, string_pool));
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  likely (format == 0 || format == 1) &&
+		  c->check_array (nameRecordZ.arrayZ, count) &&
+		  c->check_range (this, stringOffset) &&
+		  sanitize_records (c));
+  }
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      this->table = hb_sanitize_context_t ().reference_table<name> (face);
+      assert (this->table.get_length () >= this->table->stringOffset);
+      this->pool = (const char *) (const void *) (this->table+this->table->stringOffset);
+      this->pool_len = this->table.get_length () - this->table->stringOffset;
+      const hb_array_t<const NameRecord> all_names (this->table->nameRecordZ.arrayZ,
+						    this->table->count);
+
+      this->names.init ();
+      this->names.alloc (all_names.length);
+
+      for (unsigned int i = 0; i < all_names.length; i++)
+      {
+	hb_ot_name_entry_t *entry = this->names.push ();
+
+	entry->name_id = all_names[i].nameID;
+	entry->language = all_names[i].language (face);
+	entry->entry_score =  all_names[i].score ();
+	entry->entry_index = i;
+      }
+
+      this->names.qsort (_hb_ot_name_entry_cmp);
+      /* Walk and pick best only for each name_id,language pair,
+       * while dropping unsupported encodings. */
+      unsigned int j = 0;
+      for (unsigned int i = 0; i < this->names.length; i++)
+      {
+	if (this->names[i].entry_score == UNSUPPORTED ||
+	    this->names[i].language == HB_LANGUAGE_INVALID)
+	  continue;
+	if (i &&
+	    this->names[i - 1].name_id  == this->names[i].name_id &&
+	    this->names[i - 1].language == this->names[i].language)
+	  continue;
+	this->names[j++] = this->names[i];
+      }
+      this->names.resize (j);
+    }
+
+    void fini ()
+    {
+      this->names.fini ();
+      this->table.destroy ();
+    }
+
+    int get_index (hb_ot_name_id_t  name_id,
+		   hb_language_t    language,
+		   unsigned int    *width=nullptr) const
+    {
+      const hb_ot_name_entry_t key = {name_id, {0}, language};
+      const hb_ot_name_entry_t *entry = hb_bsearch (key, (const hb_ot_name_entry_t *) this->names,
+						    this->names.length,
+						    sizeof (hb_ot_name_entry_t),
+						    _hb_ot_name_entry_cmp_key);
+      if (!entry)
+	return -1;
+
+      if (width)
+	*width = entry->entry_score < 10 ? 2 : 1;
+
+      return entry->entry_index;
+    }
+
+    hb_bytes_t get_name (unsigned int idx) const
+    {
+      const hb_array_t<const NameRecord> all_names (table->nameRecordZ.arrayZ, table->count);
+      const NameRecord &record = all_names[idx];
+      const hb_bytes_t string_pool (pool, pool_len);
+      return string_pool.sub_array (record.offset, record.length);
+    }
+
+    private:
+    const char *pool;
+    unsigned int pool_len;
+    public:
+    hb_blob_ptr_t<name> table;
+    hb_vector_t<hb_ot_name_entry_t> names;
+  };
+
+  /* We only implement format 0 for now. */
+  HBUINT16	format;		/* Format selector (=0/1). */
+  HBUINT16	count;		/* Number of name records. */
+  NNOffsetTo<UnsizedArrayOf<HBUINT8>>
+		stringOffset;	/* Offset to start of string storage (from start of table). */
+  UnsizedArrayOf<NameRecord>
+		nameRecordZ;	/* The name records where count is the number of records. */
+  public:
+  DEFINE_SIZE_ARRAY (6, nameRecordZ);
+};
+
+#undef entry_index
+#undef entry_score
+
+struct name_accelerator_t : name::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_NAME_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-name.cc b/thirdparty/harfbuzz/src/hb-ot-name.cc
new file mode 100644
index 0000000000..10122b8c2e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-name.cc
@@ -0,0 +1,228 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_NAME
+
+#include "hb-ot-name-table.hh"
+
+#include "hb-utf.hh"
+
+
+/**
+ * SECTION:hb-ot-name
+ * @title: hb-ot-name
+ * @short_description: OpenType font name information
+ * @include: hb-ot.h
+ *
+ * Functions for fetching name strings from OpenType fonts.
+ **/
+
+
+/**
+ * hb_ot_name_list_names:
+ * @face: font face.
+ * @num_entries: (out) (allow-none): number of returned entries.
+ *
+ * Enumerates all available name IDs and language combinations. Returned
+ * array is owned by the @face and should not be modified.  It can be
+ * used as long as @face is alive.
+ *
+ * Returns: (out) (transfer none) (array length=num_entries): Array of available name entries.
+ * Since: 2.1.0
+ **/
+const hb_ot_name_entry_t *
+hb_ot_name_list_names (hb_face_t    *face,
+		       unsigned int *num_entries /* OUT */)
+{
+  const OT::name_accelerator_t &name = *face->table.name;
+  if (num_entries) *num_entries = name.names.length;
+  return (const hb_ot_name_entry_t *) name.names;
+}
+
+
+template <typename in_utf_t, typename out_utf_t>
+static inline unsigned int
+hb_ot_name_convert_utf (hb_bytes_t                       bytes,
+			unsigned int                    *text_size /* IN/OUT */,
+			typename out_utf_t::codepoint_t *text /* OUT */)
+{
+  unsigned int src_len = bytes.length / sizeof (typename in_utf_t::codepoint_t);
+  const typename in_utf_t::codepoint_t *src = (const typename in_utf_t::codepoint_t *) bytes.arrayZ;
+  const typename in_utf_t::codepoint_t *src_end = src + src_len;
+
+  typename out_utf_t::codepoint_t *dst = text;
+
+  hb_codepoint_t unicode;
+  const hb_codepoint_t replacement = HB_BUFFER_REPLACEMENT_CODEPOINT_DEFAULT;
+
+  if (text_size && *text_size)
+  {
+    (*text_size)--; /* Same room for NUL-termination. */
+    const typename out_utf_t::codepoint_t *dst_end = text + *text_size;
+
+    while (src < src_end && dst < dst_end)
+    {
+      const typename in_utf_t::codepoint_t *src_next = in_utf_t::next (src, src_end, &unicode, replacement);
+      typename out_utf_t::codepoint_t *dst_next = out_utf_t::encode (dst, dst_end, unicode);
+      if (dst_next == dst)
+	break; /* Out-of-room. */
+
+      dst = dst_next;
+      src = src_next;
+    }
+
+    *text_size = dst - text;
+    *dst = 0; /* NUL-terminate. */
+  }
+
+  /* Accumulate length of rest. */
+  unsigned int dst_len = dst - text;
+  while (src < src_end)
+  {
+    src = in_utf_t::next (src, src_end, &unicode, replacement);
+    dst_len += out_utf_t::encode_len (unicode);
+  }
+  return dst_len;
+}
+
+template <typename utf_t>
+static inline unsigned int
+hb_ot_name_get_utf (hb_face_t       *face,
+		    hb_ot_name_id_t  name_id,
+		    hb_language_t    language,
+		    unsigned int    *text_size /* IN/OUT */,
+		    typename utf_t::codepoint_t *text /* OUT */)
+{
+  const OT::name_accelerator_t &name = *face->table.name;
+
+  if (!language)
+    language = hb_language_from_string ("en", 2);
+
+  unsigned int width;
+  int idx = name.get_index (name_id, language, &width);
+  if (idx != -1)
+  {
+    hb_bytes_t bytes = name.get_name (idx);
+
+    if (width == 2) /* UTF16-BE */
+      return hb_ot_name_convert_utf<hb_utf16_be_t, utf_t> (bytes, text_size, text);
+
+    if (width == 1) /* ASCII */
+      return hb_ot_name_convert_utf<hb_ascii_t, utf_t> (bytes, text_size, text);
+  }
+
+  if (text_size)
+  {
+    if (*text_size)
+      *text = 0;
+    *text_size = 0;
+  }
+  return 0;
+}
+
+/**
+ * hb_ot_name_get_utf8:
+ * @face: font face.
+ * @name_id: OpenType name identifier to fetch.
+ * @language: language to fetch the name for.
+ * @text_size: (inout) (allow-none): input size of @text buffer, and output size of
+ *                                   text written to buffer.
+ * @text: (out caller-allocates) (array length=text_size): buffer to write fetched name into.
+ *
+ * Fetches a font name from the OpenType 'name' table.
+ * If @language is #HB_LANGUAGE_INVALID, English ("en") is assumed.
+ * Returns string in UTF-8 encoding.
+ *
+ * Returns: full length of the requested string, or 0 if not found.
+ * Since: 2.1.0
+ **/
+unsigned int
+hb_ot_name_get_utf8 (hb_face_t       *face,
+		     hb_ot_name_id_t  name_id,
+		     hb_language_t    language,
+		     unsigned int    *text_size /* IN/OUT */,
+		     char            *text      /* OUT */)
+{
+  return hb_ot_name_get_utf<hb_utf8_t> (face, name_id, language, text_size,
+					(hb_utf8_t::codepoint_t *) text);
+}
+
+/**
+ * hb_ot_name_get_utf16:
+ * @face: font face.
+ * @name_id: OpenType name identifier to fetch.
+ * @language: language to fetch the name for.
+ * @text_size: (inout) (allow-none): input size of @text buffer, and output size of
+ *                                   text written to buffer.
+ * @text: (out caller-allocates) (array length=text_size): buffer to write fetched name into.
+ *
+ * Fetches a font name from the OpenType 'name' table.
+ * If @language is #HB_LANGUAGE_INVALID, English ("en") is assumed.
+ * Returns string in UTF-16 encoding.
+ *
+ * Returns: full length of the requested string, or 0 if not found.
+ * Since: 2.1.0
+ **/
+unsigned int
+hb_ot_name_get_utf16 (hb_face_t       *face,
+		      hb_ot_name_id_t  name_id,
+		      hb_language_t    language,
+		      unsigned int    *text_size /* IN/OUT */,
+		      uint16_t        *text      /* OUT */)
+{
+  return hb_ot_name_get_utf<hb_utf16_t> (face, name_id, language, text_size, text);
+}
+
+/**
+ * hb_ot_name_get_utf32:
+ * @face: font face.
+ * @name_id: OpenType name identifier to fetch.
+ * @language: language to fetch the name for.
+ * @text_size: (inout) (allow-none): input size of @text buffer, and output size of
+ *                                   text written to buffer.
+ * @text: (out caller-allocates) (array length=text_size): buffer to write fetched name into.
+ *
+ * Fetches a font name from the OpenType 'name' table.
+ * If @language is #HB_LANGUAGE_INVALID, English ("en") is assumed.
+ * Returns string in UTF-32 encoding.
+ *
+ * Returns: full length of the requested string, or 0 if not found.
+ * Since: 2.1.0
+ **/
+unsigned int
+hb_ot_name_get_utf32 (hb_face_t       *face,
+		      hb_ot_name_id_t  name_id,
+		      hb_language_t    language,
+		      unsigned int    *text_size /* IN/OUT */,
+		      uint32_t        *text      /* OUT */)
+{
+  return hb_ot_name_get_utf<hb_utf32_t> (face, name_id, language, text_size, text);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-name.h b/thirdparty/harfbuzz/src/hb-ot-name.h
new file mode 100644
index 0000000000..3b4ad581c7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-name.h
@@ -0,0 +1,129 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_NAME_H
+#define HB_OT_NAME_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+
+/**
+ * hb_ot_name_id_t:
+ * @HB_OT_NAME_ID_INVALID: Value to represent a nonexistent name ID.
+ *
+ * An integral type representing an OpenType 'name' table name identifier.
+ * There are predefined name IDs, as well as name IDs return from other
+ * API.  These can be used to fetch name strings from a font face.
+ *
+ * Since: 2.0.0
+ **/
+enum
+{
+  HB_OT_NAME_ID_COPYRIGHT		= 0,
+  HB_OT_NAME_ID_FONT_FAMILY		= 1,
+  HB_OT_NAME_ID_FONT_SUBFAMILY		= 2,
+  HB_OT_NAME_ID_UNIQUE_ID		= 3,
+  HB_OT_NAME_ID_FULL_NAME		= 4,
+  HB_OT_NAME_ID_VERSION_STRING		= 5,
+  HB_OT_NAME_ID_POSTSCRIPT_NAME		= 6,
+  HB_OT_NAME_ID_TRADEMARK		= 7,
+  HB_OT_NAME_ID_MANUFACTURER		= 8,
+  HB_OT_NAME_ID_DESIGNER		= 9,
+  HB_OT_NAME_ID_DESCRIPTION		= 10,
+  HB_OT_NAME_ID_VENDOR_URL		= 11,
+  HB_OT_NAME_ID_DESIGNER_URL		= 12,
+  HB_OT_NAME_ID_LICENSE			= 13,
+  HB_OT_NAME_ID_LICENSE_URL		= 14,
+/*HB_OT_NAME_ID_RESERVED		= 15,*/
+  HB_OT_NAME_ID_TYPOGRAPHIC_FAMILY	= 16,
+  HB_OT_NAME_ID_TYPOGRAPHIC_SUBFAMILY	= 17,
+  HB_OT_NAME_ID_MAC_FULL_NAME		= 18,
+  HB_OT_NAME_ID_SAMPLE_TEXT		= 19,
+  HB_OT_NAME_ID_CID_FINDFONT_NAME	= 20,
+  HB_OT_NAME_ID_WWS_FAMILY		= 21,
+  HB_OT_NAME_ID_WWS_SUBFAMILY		= 22,
+  HB_OT_NAME_ID_LIGHT_BACKGROUND	= 23,
+  HB_OT_NAME_ID_DARK_BACKGROUND		= 24,
+  HB_OT_NAME_ID_VARIATIONS_PS_PREFIX	= 25,
+
+  HB_OT_NAME_ID_INVALID			= 0xFFFF
+};
+
+typedef unsigned int hb_ot_name_id_t;
+
+
+/**
+ * hb_ot_name_entry_t:
+ * @name_id: name ID
+ * @language: language
+ *
+ * Structure representing a name ID in a particular language.
+ *
+ * Since: 2.1.0
+ **/
+typedef struct hb_ot_name_entry_t
+{
+  hb_ot_name_id_t name_id;
+  /*< private >*/
+  hb_var_int_t    var;
+  /*< public >*/
+  hb_language_t   language;
+} hb_ot_name_entry_t;
+
+HB_EXTERN const hb_ot_name_entry_t *
+hb_ot_name_list_names (hb_face_t    *face,
+		       unsigned int *num_entries /* OUT */);
+
+
+HB_EXTERN unsigned int
+hb_ot_name_get_utf8 (hb_face_t       *face,
+		     hb_ot_name_id_t  name_id,
+		     hb_language_t    language,
+		     unsigned int    *text_size /* IN/OUT */,
+		     char            *text      /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_name_get_utf16 (hb_face_t       *face,
+		      hb_ot_name_id_t  name_id,
+		      hb_language_t    language,
+		      unsigned int    *text_size /* IN/OUT */,
+		      uint16_t        *text      /* OUT */);
+
+HB_EXTERN unsigned int
+hb_ot_name_get_utf32 (hb_face_t       *face,
+		      hb_ot_name_id_t  name_id,
+		      hb_language_t    language,
+		      unsigned int    *text_size /* IN/OUT */,
+		      uint32_t        *text      /* OUT */);
+
+
+HB_END_DECLS
+
+#endif /* HB_OT_NAME_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-os2-table.hh b/thirdparty/harfbuzz/src/hb-ot-os2-table.hh
new file mode 100644
index 0000000000..7d31b712c4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-os2-table.hh
@@ -0,0 +1,316 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_OS2_TABLE_HH
+#define HB_OT_OS2_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-os2-unicode-ranges.hh"
+#include "hb-ot-cmap-table.hh"
+
+#include "hb-set.hh"
+
+/*
+ * OS/2 and Windows Metrics
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/os2
+ */
+#define HB_OT_TAG_OS2 HB_TAG('O','S','/','2')
+
+
+namespace OT {
+
+struct OS2V1Tail
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT32	ulCodePageRange1;
+  HBUINT32	ulCodePageRange2;
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct OS2V2Tail
+{
+  bool has_data () const { return sxHeight || sCapHeight; }
+
+  const OS2V2Tail * operator -> () const { return this; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBINT16	sxHeight;
+  HBINT16	sCapHeight;
+  HBUINT16	usDefaultChar;
+  HBUINT16	usBreakChar;
+  HBUINT16	usMaxContext;
+  public:
+  DEFINE_SIZE_STATIC (10);
+};
+
+struct OS2V5Tail
+{
+  inline bool get_optical_size (unsigned int *lower, unsigned int *upper) const
+  {
+    unsigned int lower_optical_size = usLowerOpticalPointSize;
+    unsigned int upper_optical_size = usUpperOpticalPointSize;
+
+    /* Per https://docs.microsoft.com/en-us/typography/opentype/spec/os2#lps */
+    if (lower_optical_size < upper_optical_size &&
+	lower_optical_size >= 1 && lower_optical_size <= 0xFFFE &&
+	upper_optical_size >= 2 && upper_optical_size <= 0xFFFF)
+    {
+      *lower = lower_optical_size;
+      *upper = upper_optical_size;
+      return true;
+    }
+    return false;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBUINT16	usLowerOpticalPointSize;
+  HBUINT16	usUpperOpticalPointSize;
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct OS2
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_OS2;
+
+  bool has_data () const { return usWeightClass || usWidthClass || usFirstCharIndex || usLastCharIndex; }
+
+  const OS2V1Tail &v1 () const { return version >= 1 ? v1X : Null (OS2V1Tail); }
+  const OS2V2Tail &v2 () const { return version >= 2 ? v2X : Null (OS2V2Tail); }
+  const OS2V5Tail &v5 () const { return version >= 5 ? v5X : Null (OS2V5Tail); }
+
+  enum selection_flag_t {
+    ITALIC		= 1u<<0,
+    UNDERSCORE		= 1u<<1,
+    NEGATIVE		= 1u<<2,
+    OUTLINED		= 1u<<3,
+    STRIKEOUT		= 1u<<4,
+    BOLD		= 1u<<5,
+    REGULAR		= 1u<<6,
+    USE_TYPO_METRICS	= 1u<<7,
+    WWS			= 1u<<8,
+    OBLIQUE		= 1u<<9
+  };
+
+  bool        is_italic () const { return fsSelection & ITALIC; }
+  bool       is_oblique () const { return fsSelection & OBLIQUE; }
+  bool use_typo_metrics () const { return fsSelection & USE_TYPO_METRICS; }
+
+  enum width_class_t {
+    FWIDTH_ULTRA_CONDENSED	= 1, /* 50% */
+    FWIDTH_EXTRA_CONDENSED	= 2, /* 62.5% */
+    FWIDTH_CONDENSED		= 3, /* 75% */
+    FWIDTH_SEMI_CONDENSED	= 4, /* 87.5% */
+    FWIDTH_NORMAL		= 5, /* 100% */
+    FWIDTH_SEMI_EXPANDED	= 6, /* 112.5% */
+    FWIDTH_EXPANDED		= 7, /* 125% */
+    FWIDTH_EXTRA_EXPANDED	= 8, /* 150% */
+    FWIDTH_ULTRA_EXPANDED	= 9  /* 200% */
+  };
+
+  float get_width () const
+  {
+    switch (usWidthClass) {
+    case FWIDTH_ULTRA_CONDENSED:return 50.f;
+    case FWIDTH_EXTRA_CONDENSED:return 62.5f;
+    case FWIDTH_CONDENSED:	return 75.f;
+    case FWIDTH_SEMI_CONDENSED:	return 87.5f;
+    default:
+    case FWIDTH_NORMAL:		return 100.f;
+    case FWIDTH_SEMI_EXPANDED:	return 112.5f;
+    case FWIDTH_EXPANDED:	return 125.f;
+    case FWIDTH_EXTRA_EXPANDED:	return 150.f;
+    case FWIDTH_ULTRA_EXPANDED:	return 200.f;
+    }
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    OS2 *os2_prime = c->serializer->embed (this);
+    if (unlikely (!os2_prime)) return_trace (false);
+
+    hb_set_t unicodes;
+    if (!c->plan->glyphs_requested->is_empty ())
+    {
+      hb_map_t unicode_glyphid_map;
+      
+      OT::cmap::accelerator_t cmap;
+      cmap.init (c->plan->source);
+      cmap.collect_mapping (&unicodes, &unicode_glyphid_map);
+      cmap.fini ();
+      
+      if (c->plan->unicodes->is_empty ()) unicodes.clear ();
+      else hb_set_set (&unicodes, c->plan->unicodes);
+  
+      + unicode_glyphid_map.iter ()
+      | hb_filter (c->plan->glyphs_requested, hb_second)
+      | hb_map (hb_first)
+      | hb_sink (unicodes)
+      ;
+    }
+    /* when --gids option is not used, no need to do collect_mapping that is
+       * iterating all codepoints in each subtable, which is not efficient */
+    uint16_t min_cp, max_cp;
+    find_min_and_max_codepoint (unicodes.is_empty () ? c->plan->unicodes : &unicodes, &min_cp, &max_cp);
+    os2_prime->usFirstCharIndex = min_cp;
+    os2_prime->usLastCharIndex = max_cp;
+
+    _update_unicode_ranges (unicodes.is_empty () ? c->plan->unicodes : &unicodes, os2_prime->ulUnicodeRange);
+
+    return_trace (true);
+  }
+
+  void _update_unicode_ranges (const hb_set_t *codepoints,
+			       HBUINT32 ulUnicodeRange[4]) const
+  {
+    HBUINT32	newBits[4];
+    for (unsigned int i = 0; i < 4; i++)
+      newBits[i] = 0;
+
+    hb_codepoint_t cp = HB_SET_VALUE_INVALID;
+    while (codepoints->next (&cp)) {
+      unsigned int bit = _hb_ot_os2_get_unicode_range_bit (cp);
+      if (bit < 128)
+      {
+	unsigned int block = bit / 32;
+	unsigned int bit_in_block = bit % 32;
+	unsigned int mask = 1 << bit_in_block;
+	newBits[block] = newBits[block] | mask;
+      }
+      if (cp >= 0x10000 && cp <= 0x110000)
+      {
+	/* the spec says that bit 57 ("Non Plane 0") implies that there's
+	   at least one codepoint beyond the BMP; so I also include all
+	   the non-BMP codepoints here */
+	newBits[1] = newBits[1] | (1 << 25);
+      }
+    }
+
+    for (unsigned int i = 0; i < 4; i++)
+      ulUnicodeRange[i] = ulUnicodeRange[i] & newBits[i]; // set bits only if set in the original
+  }
+
+  static void find_min_and_max_codepoint (const hb_set_t *codepoints,
+					  uint16_t *min_cp, /* OUT */
+					  uint16_t *max_cp  /* OUT */)
+  {
+    *min_cp = hb_min (0xFFFFu, codepoints->get_min ());
+    *max_cp = hb_min (0xFFFFu, codepoints->get_max ());
+  }
+
+  /* https://github.com/Microsoft/Font-Validator/blob/520aaae/OTFontFileVal/val_OS2.cs#L644-L681 */
+  enum font_page_t
+  {
+    FONT_PAGE_HEBREW		= 0xB100, /* Hebrew Windows 3.1 font page */
+    FONT_PAGE_SIMP_ARABIC	= 0xB200, /* Simplified Arabic Windows 3.1 font page */
+    FONT_PAGE_TRAD_ARABIC	= 0xB300, /* Traditional Arabic Windows 3.1 font page */
+    FONT_PAGE_OEM_ARABIC	= 0xB400, /* OEM Arabic Windows 3.1 font page */
+    FONT_PAGE_SIMP_FARSI	= 0xBA00, /* Simplified Farsi Windows 3.1 font page */
+    FONT_PAGE_TRAD_FARSI	= 0xBB00, /* Traditional Farsi Windows 3.1 font page */
+    FONT_PAGE_THAI		= 0xDE00  /* Thai Windows 3.1 font page */
+  };
+  font_page_t get_font_page () const
+  { return (font_page_t) (version == 0 ? fsSelection & 0xFF00 : 0); }
+
+  unsigned get_size () const
+  {
+    unsigned result = min_size;
+    if (version >= 1) result += v1X.get_size ();
+    if (version >= 2) result += v2X.get_size ();
+    if (version >= 5) result += v5X.get_size ();
+    return result;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this))) return_trace (false);
+    if (unlikely (version >= 1 && !v1X.sanitize (c))) return_trace (false);
+    if (unlikely (version >= 2 && !v2X.sanitize (c))) return_trace (false);
+    if (unlikely (version >= 5 && !v5X.sanitize (c))) return_trace (false);
+    return_trace (true);
+  }
+
+  public:
+  HBUINT16	version;
+  HBINT16	xAvgCharWidth;
+  HBUINT16	usWeightClass;
+  HBUINT16	usWidthClass;
+  HBUINT16	fsType;
+  HBINT16	ySubscriptXSize;
+  HBINT16	ySubscriptYSize;
+  HBINT16	ySubscriptXOffset;
+  HBINT16	ySubscriptYOffset;
+  HBINT16	ySuperscriptXSize;
+  HBINT16	ySuperscriptYSize;
+  HBINT16	ySuperscriptXOffset;
+  HBINT16	ySuperscriptYOffset;
+  HBINT16	yStrikeoutSize;
+  HBINT16	yStrikeoutPosition;
+  HBINT16	sFamilyClass;
+  HBUINT8	panose[10];
+  HBUINT32	ulUnicodeRange[4];
+  Tag		achVendID;
+  HBUINT16	fsSelection;
+  HBUINT16	usFirstCharIndex;
+  HBUINT16	usLastCharIndex;
+  HBINT16	sTypoAscender;
+  HBINT16	sTypoDescender;
+  HBINT16	sTypoLineGap;
+  HBUINT16	usWinAscent;
+  HBUINT16	usWinDescent;
+  OS2V1Tail	v1X;
+  OS2V2Tail	v2X;
+  OS2V5Tail	v5X;
+  public:
+  DEFINE_SIZE_MIN (78);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_OS2_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-os2-unicode-ranges.hh b/thirdparty/harfbuzz/src/hb-ot-os2-unicode-ranges.hh
new file mode 100644
index 0000000000..9613d2d186
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-os2-unicode-ranges.hh
@@ -0,0 +1,231 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger
+ */
+
+#ifndef HB_OT_OS2_UNICODE_RANGES_HH
+#define HB_OT_OS2_UNICODE_RANGES_HH
+
+#include "hb.hh"
+
+namespace OT {
+
+struct OS2Range
+{
+  int cmp (hb_codepoint_t key) const
+  { return (key < start) ? -1 : key <= end ? 0 : +1; }
+
+  hb_codepoint_t start;
+  hb_codepoint_t end;
+  unsigned int bit;
+};
+
+/* Note: The contents of this array was generated using gen-os2-unicode-ranges.py. */
+static const OS2Range _hb_os2_unicode_ranges[] =
+{
+  {     0x0,     0x7F,   0}, // Basic Latin
+  {    0x80,     0xFF,   1}, // Latin-1 Supplement
+  {   0x100,    0x17F,   2}, // Latin Extended-A
+  {   0x180,    0x24F,   3}, // Latin Extended-B
+  {   0x250,    0x2AF,   4}, // IPA Extensions
+  {   0x2B0,    0x2FF,   5}, // Spacing Modifier Letters
+  {   0x300,    0x36F,   6}, // Combining Diacritical Marks
+  {   0x370,    0x3FF,   7}, // Greek and Coptic
+  {   0x400,    0x4FF,   9}, // Cyrillic
+  {   0x500,    0x52F,   9}, // Cyrillic Supplement
+  {   0x530,    0x58F,  10}, // Armenian
+  {   0x590,    0x5FF,  11}, // Hebrew
+  {   0x600,    0x6FF,  13}, // Arabic
+  {   0x700,    0x74F,  71}, // Syriac
+  {   0x750,    0x77F,  13}, // Arabic Supplement
+  {   0x780,    0x7BF,  72}, // Thaana
+  {   0x7C0,    0x7FF,  14}, // NKo
+  {   0x900,    0x97F,  15}, // Devanagari
+  {   0x980,    0x9FF,  16}, // Bengali
+  {   0xA00,    0xA7F,  17}, // Gurmukhi
+  {   0xA80,    0xAFF,  18}, // Gujarati
+  {   0xB00,    0xB7F,  19}, // Oriya
+  {   0xB80,    0xBFF,  20}, // Tamil
+  {   0xC00,    0xC7F,  21}, // Telugu
+  {   0xC80,    0xCFF,  22}, // Kannada
+  {   0xD00,    0xD7F,  23}, // Malayalam
+  {   0xD80,    0xDFF,  73}, // Sinhala
+  {   0xE00,    0xE7F,  24}, // Thai
+  {   0xE80,    0xEFF,  25}, // Lao
+  {   0xF00,    0xFFF,  70}, // Tibetan
+  {  0x1000,   0x109F,  74}, // Myanmar
+  {  0x10A0,   0x10FF,  26}, // Georgian
+  {  0x1100,   0x11FF,  28}, // Hangul Jamo
+  {  0x1200,   0x137F,  75}, // Ethiopic
+  {  0x1380,   0x139F,  75}, // Ethiopic Supplement
+  {  0x13A0,   0x13FF,  76}, // Cherokee
+  {  0x1400,   0x167F,  77}, // Unified Canadian Aboriginal Syllabics
+  {  0x1680,   0x169F,  78}, // Ogham
+  {  0x16A0,   0x16FF,  79}, // Runic
+  {  0x1700,   0x171F,  84}, // Tagalog
+  {  0x1720,   0x173F,  84}, // Hanunoo
+  {  0x1740,   0x175F,  84}, // Buhid
+  {  0x1760,   0x177F,  84}, // Tagbanwa
+  {  0x1780,   0x17FF,  80}, // Khmer
+  {  0x1800,   0x18AF,  81}, // Mongolian
+  {  0x1900,   0x194F,  93}, // Limbu
+  {  0x1950,   0x197F,  94}, // Tai Le
+  {  0x1980,   0x19DF,  95}, // New Tai Lue
+  {  0x19E0,   0x19FF,  80}, // Khmer Symbols
+  {  0x1A00,   0x1A1F,  96}, // Buginese
+  {  0x1B00,   0x1B7F,  27}, // Balinese
+  {  0x1B80,   0x1BBF, 112}, // Sundanese
+  {  0x1C00,   0x1C4F, 113}, // Lepcha
+  {  0x1C50,   0x1C7F, 114}, // Ol Chiki
+  {  0x1D00,   0x1D7F,   4}, // Phonetic Extensions
+  {  0x1D80,   0x1DBF,   4}, // Phonetic Extensions Supplement
+  {  0x1DC0,   0x1DFF,   6}, // Combining Diacritical Marks Supplement
+  {  0x1E00,   0x1EFF,  29}, // Latin Extended Additional
+  {  0x1F00,   0x1FFF,  30}, // Greek Extended
+  {  0x2000,   0x206F,  31}, // General Punctuation
+  {  0x2070,   0x209F,  32}, // Superscripts And Subscripts
+  {  0x20A0,   0x20CF,  33}, // Currency Symbols
+  {  0x20D0,   0x20FF,  34}, // Combining Diacritical Marks For Symbols
+  {  0x2100,   0x214F,  35}, // Letterlike Symbols
+  {  0x2150,   0x218F,  36}, // Number Forms
+  {  0x2190,   0x21FF,  37}, // Arrows
+  {  0x2200,   0x22FF,  38}, // Mathematical Operators
+  {  0x2300,   0x23FF,  39}, // Miscellaneous Technical
+  {  0x2400,   0x243F,  40}, // Control Pictures
+  {  0x2440,   0x245F,  41}, // Optical Character Recognition
+  {  0x2460,   0x24FF,  42}, // Enclosed Alphanumerics
+  {  0x2500,   0x257F,  43}, // Box Drawing
+  {  0x2580,   0x259F,  44}, // Block Elements
+  {  0x25A0,   0x25FF,  45}, // Geometric Shapes
+  {  0x2600,   0x26FF,  46}, // Miscellaneous Symbols
+  {  0x2700,   0x27BF,  47}, // Dingbats
+  {  0x27C0,   0x27EF,  38}, // Miscellaneous Mathematical Symbols-A
+  {  0x27F0,   0x27FF,  37}, // Supplemental Arrows-A
+  {  0x2800,   0x28FF,  82}, // Braille Patterns
+  {  0x2900,   0x297F,  37}, // Supplemental Arrows-B
+  {  0x2980,   0x29FF,  38}, // Miscellaneous Mathematical Symbols-B
+  {  0x2A00,   0x2AFF,  38}, // Supplemental Mathematical Operators
+  {  0x2B00,   0x2BFF,  37}, // Miscellaneous Symbols and Arrows
+  {  0x2C00,   0x2C5F,  97}, // Glagolitic
+  {  0x2C60,   0x2C7F,  29}, // Latin Extended-C
+  {  0x2C80,   0x2CFF,   8}, // Coptic
+  {  0x2D00,   0x2D2F,  26}, // Georgian Supplement
+  {  0x2D30,   0x2D7F,  98}, // Tifinagh
+  {  0x2D80,   0x2DDF,  75}, // Ethiopic Extended
+  {  0x2DE0,   0x2DFF,   9}, // Cyrillic Extended-A
+  {  0x2E00,   0x2E7F,  31}, // Supplemental Punctuation
+  {  0x2E80,   0x2EFF,  59}, // CJK Radicals Supplement
+  {  0x2F00,   0x2FDF,  59}, // Kangxi Radicals
+  {  0x2FF0,   0x2FFF,  59}, // Ideographic Description Characters
+  {  0x3000,   0x303F,  48}, // CJK Symbols And Punctuation
+  {  0x3040,   0x309F,  49}, // Hiragana
+  {  0x30A0,   0x30FF,  50}, // Katakana
+  {  0x3100,   0x312F,  51}, // Bopomofo
+  {  0x3130,   0x318F,  52}, // Hangul Compatibility Jamo
+  {  0x3190,   0x319F,  59}, // Kanbun
+  {  0x31A0,   0x31BF,  51}, // Bopomofo Extended
+  {  0x31C0,   0x31EF,  61}, // CJK Strokes
+  {  0x31F0,   0x31FF,  50}, // Katakana Phonetic Extensions
+  {  0x3200,   0x32FF,  54}, // Enclosed CJK Letters And Months
+  {  0x3300,   0x33FF,  55}, // CJK Compatibility
+  {  0x3400,   0x4DBF,  59}, // CJK Unified Ideographs Extension A
+  {  0x4DC0,   0x4DFF,  99}, // Yijing Hexagram Symbols
+  {  0x4E00,   0x9FFF,  59}, // CJK Unified Ideographs
+  {  0xA000,   0xA48F,  83}, // Yi Syllables
+  {  0xA490,   0xA4CF,  83}, // Yi Radicals
+  {  0xA500,   0xA63F,  12}, // Vai
+  {  0xA640,   0xA69F,   9}, // Cyrillic Extended-B
+  {  0xA700,   0xA71F,   5}, // Modifier Tone Letters
+  {  0xA720,   0xA7FF,  29}, // Latin Extended-D
+  {  0xA800,   0xA82F, 100}, // Syloti Nagri
+  {  0xA840,   0xA87F,  53}, // Phags-pa
+  {  0xA880,   0xA8DF, 115}, // Saurashtra
+  {  0xA900,   0xA92F, 116}, // Kayah Li
+  {  0xA930,   0xA95F, 117}, // Rejang
+  {  0xAA00,   0xAA5F, 118}, // Cham
+  {  0xAC00,   0xD7AF,  56}, // Hangul Syllables
+  {  0xD800,   0xDFFF,  57}, // Non-Plane 0 *
+  {  0xE000,   0xF8FF,  60}, // Private Use Area (plane 0)
+  {  0xF900,   0xFAFF,  61}, // CJK Compatibility Ideographs
+  {  0xFB00,   0xFB4F,  62}, // Alphabetic Presentation Forms
+  {  0xFB50,   0xFDFF,  63}, // Arabic Presentation Forms-A
+  {  0xFE00,   0xFE0F,  91}, // Variation Selectors
+  {  0xFE10,   0xFE1F,  65}, // Vertical Forms
+  {  0xFE20,   0xFE2F,  64}, // Combining Half Marks
+  {  0xFE30,   0xFE4F,  65}, // CJK Compatibility Forms
+  {  0xFE50,   0xFE6F,  66}, // Small Form Variants
+  {  0xFE70,   0xFEFF,  67}, // Arabic Presentation Forms-B
+  {  0xFF00,   0xFFEF,  68}, // Halfwidth And Fullwidth Forms
+  {  0xFFF0,   0xFFFF,  69}, // Specials
+  { 0x10000,  0x1007F, 101}, // Linear B Syllabary
+  { 0x10080,  0x100FF, 101}, // Linear B Ideograms
+  { 0x10100,  0x1013F, 101}, // Aegean Numbers
+  { 0x10140,  0x1018F, 102}, // Ancient Greek Numbers
+  { 0x10190,  0x101CF, 119}, // Ancient Symbols
+  { 0x101D0,  0x101FF, 120}, // Phaistos Disc
+  { 0x10280,  0x1029F, 121}, // Lycian
+  { 0x102A0,  0x102DF, 121}, // Carian
+  { 0x10300,  0x1032F,  85}, // Old Italic
+  { 0x10330,  0x1034F,  86}, // Gothic
+  { 0x10380,  0x1039F, 103}, // Ugaritic
+  { 0x103A0,  0x103DF, 104}, // Old Persian
+  { 0x10400,  0x1044F,  87}, // Deseret
+  { 0x10450,  0x1047F, 105}, // Shavian
+  { 0x10480,  0x104AF, 106}, // Osmanya
+  { 0x10800,  0x1083F, 107}, // Cypriot Syllabary
+  { 0x10900,  0x1091F,  58}, // Phoenician
+  { 0x10920,  0x1093F, 121}, // Lydian
+  { 0x10A00,  0x10A5F, 108}, // Kharoshthi
+  { 0x12000,  0x123FF, 110}, // Cuneiform
+  { 0x12400,  0x1247F, 110}, // Cuneiform Numbers and Punctuation
+  { 0x1D000,  0x1D0FF,  88}, // Byzantine Musical Symbols
+  { 0x1D100,  0x1D1FF,  88}, // Musical Symbols
+  { 0x1D200,  0x1D24F,  88}, // Ancient Greek Musical Notation
+  { 0x1D300,  0x1D35F, 109}, // Tai Xuan Jing Symbols
+  { 0x1D360,  0x1D37F, 111}, // Counting Rod Numerals
+  { 0x1D400,  0x1D7FF,  89}, // Mathematical Alphanumeric Symbols
+  { 0x1F000,  0x1F02F, 122}, // Mahjong Tiles
+  { 0x1F030,  0x1F09F, 122}, // Domino Tiles
+  { 0x20000,  0x2A6DF,  59}, // CJK Unified Ideographs Extension B
+  { 0x2F800,  0x2FA1F,  61}, // CJK Compatibility Ideographs Supplement
+  { 0xE0000,  0xE007F,  92}, // Tags
+  { 0xE0100,  0xE01EF,  91}, // Variation Selectors Supplement
+  { 0xF0000,  0xFFFFD,  90}, // Private Use (plane 15)
+  {0x100000, 0x10FFFD,  90}, // Private Use (plane 16)
+};
+
+/**
+ * _hb_ot_os2_get_unicode_range_bit:
+ * Returns the bit to be set in os/2 ulUnicodeOS2Range for a given codepoint.
+ **/
+static unsigned int
+_hb_ot_os2_get_unicode_range_bit (hb_codepoint_t cp)
+{
+  auto *range = hb_sorted_array (_hb_os2_unicode_ranges).bsearch (cp);
+  return range ? range->bit : -1;
+}
+
+} /* namespace OT */
+
+#endif /* HB_OT_OS2_UNICODE_RANGES_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-post-macroman.hh b/thirdparty/harfbuzz/src/hb-ot-post-macroman.hh
new file mode 100644
index 0000000000..b4df8aaeea
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-post-macroman.hh
@@ -0,0 +1,294 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_POST_MACROMAN_HH
+#if 0 /* Make checks happy. */
+#define HB_OT_POST_MACROMAN_HH
+#include "hb.hh"
+#endif
+
+
+_S(".notdef")
+_S(".null")
+_S("nonmarkingreturn")
+_S("space")
+_S("exclam")
+_S("quotedbl")
+_S("numbersign")
+_S("dollar")
+_S("percent")
+_S("ampersand")
+_S("quotesingle")
+_S("parenleft")
+_S("parenright")
+_S("asterisk")
+_S("plus")
+_S("comma")
+_S("hyphen")
+_S("period")
+_S("slash")
+_S("zero")
+_S("one")
+_S("two")
+_S("three")
+_S("four")
+_S("five")
+_S("six")
+_S("seven")
+_S("eight")
+_S("nine")
+_S("colon")
+_S("semicolon")
+_S("less")
+_S("equal")
+_S("greater")
+_S("question")
+_S("at")
+_S("A")
+_S("B")
+_S("C")
+_S("D")
+_S("E")
+_S("F")
+_S("G")
+_S("H")
+_S("I")
+_S("J")
+_S("K")
+_S("L")
+_S("M")
+_S("N")
+_S("O")
+_S("P")
+_S("Q")
+_S("R")
+_S("S")
+_S("T")
+_S("U")
+_S("V")
+_S("W")
+_S("X")
+_S("Y")
+_S("Z")
+_S("bracketleft")
+_S("backslash")
+_S("bracketright")
+_S("asciicircum")
+_S("underscore")
+_S("grave")
+_S("a")
+_S("b")
+_S("c")
+_S("d")
+_S("e")
+_S("f")
+_S("g")
+_S("h")
+_S("i")
+_S("j")
+_S("k")
+_S("l")
+_S("m")
+_S("n")
+_S("o")
+_S("p")
+_S("q")
+_S("r")
+_S("s")
+_S("t")
+_S("u")
+_S("v")
+_S("w")
+_S("x")
+_S("y")
+_S("z")
+_S("braceleft")
+_S("bar")
+_S("braceright")
+_S("asciitilde")
+_S("Adieresis")
+_S("Aring")
+_S("Ccedilla")
+_S("Eacute")
+_S("Ntilde")
+_S("Odieresis")
+_S("Udieresis")
+_S("aacute")
+_S("agrave")
+_S("acircumflex")
+_S("adieresis")
+_S("atilde")
+_S("aring")
+_S("ccedilla")
+_S("eacute")
+_S("egrave")
+_S("ecircumflex")
+_S("edieresis")
+_S("iacute")
+_S("igrave")
+_S("icircumflex")
+_S("idieresis")
+_S("ntilde")
+_S("oacute")
+_S("ograve")
+_S("ocircumflex")
+_S("odieresis")
+_S("otilde")
+_S("uacute")
+_S("ugrave")
+_S("ucircumflex")
+_S("udieresis")
+_S("dagger")
+_S("degree")
+_S("cent")
+_S("sterling")
+_S("section")
+_S("bullet")
+_S("paragraph")
+_S("germandbls")
+_S("registered")
+_S("copyright")
+_S("trademark")
+_S("acute")
+_S("dieresis")
+_S("notequal")
+_S("AE")
+_S("Oslash")
+_S("infinity")
+_S("plusminus")
+_S("lessequal")
+_S("greaterequal")
+_S("yen")
+_S("mu")
+_S("partialdiff")
+_S("summation")
+_S("product")
+_S("pi")
+_S("integral")
+_S("ordfeminine")
+_S("ordmasculine")
+_S("Omega")
+_S("ae")
+_S("oslash")
+_S("questiondown")
+_S("exclamdown")
+_S("logicalnot")
+_S("radical")
+_S("florin")
+_S("approxequal")
+_S("Delta")
+_S("guillemotleft")
+_S("guillemotright")
+_S("ellipsis")
+_S("nonbreakingspace")
+_S("Agrave")
+_S("Atilde")
+_S("Otilde")
+_S("OE")
+_S("oe")
+_S("endash")
+_S("emdash")
+_S("quotedblleft")
+_S("quotedblright")
+_S("quoteleft")
+_S("quoteright")
+_S("divide")
+_S("lozenge")
+_S("ydieresis")
+_S("Ydieresis")
+_S("fraction")
+_S("currency")
+_S("guilsinglleft")
+_S("guilsinglright")
+_S("fi")
+_S("fl")
+_S("daggerdbl")
+_S("periodcentered")
+_S("quotesinglbase")
+_S("quotedblbase")
+_S("perthousand")
+_S("Acircumflex")
+_S("Ecircumflex")
+_S("Aacute")
+_S("Edieresis")
+_S("Egrave")
+_S("Iacute")
+_S("Icircumflex")
+_S("Idieresis")
+_S("Igrave")
+_S("Oacute")
+_S("Ocircumflex")
+_S("apple")
+_S("Ograve")
+_S("Uacute")
+_S("Ucircumflex")
+_S("Ugrave")
+_S("dotlessi")
+_S("circumflex")
+_S("tilde")
+_S("macron")
+_S("breve")
+_S("dotaccent")
+_S("ring")
+_S("cedilla")
+_S("hungarumlaut")
+_S("ogonek")
+_S("caron")
+_S("Lslash")
+_S("lslash")
+_S("Scaron")
+_S("scaron")
+_S("Zcaron")
+_S("zcaron")
+_S("brokenbar")
+_S("Eth")
+_S("eth")
+_S("Yacute")
+_S("yacute")
+_S("Thorn")
+_S("thorn")
+_S("minus")
+_S("multiply")
+_S("onesuperior")
+_S("twosuperior")
+_S("threesuperior")
+_S("onehalf")
+_S("onequarter")
+_S("threequarters")
+_S("franc")
+_S("Gbreve")
+_S("gbreve")
+_S("Idotaccent")
+_S("Scedilla")
+_S("scedilla")
+_S("Cacute")
+_S("cacute")
+_S("Ccaron")
+_S("ccaron")
+_S("dcroat")
+
+
+#endif /* HB_OT_POST_MACROMAN_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-post-table.hh b/thirdparty/harfbuzz/src/hb-ot-post-table.hh
new file mode 100644
index 0000000000..8586331cd4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-post-table.hh
@@ -0,0 +1,298 @@
+/*
+ * Copyright © 2016  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_POST_TABLE_HH
+#define HB_OT_POST_TABLE_HH
+
+#include "hb-open-type.hh"
+
+#define HB_STRING_ARRAY_NAME format1_names
+#define HB_STRING_ARRAY_LIST "hb-ot-post-macroman.hh"
+#include "hb-string-array.hh"
+#undef HB_STRING_ARRAY_LIST
+#undef HB_STRING_ARRAY_NAME
+
+/*
+ * post -- PostScript
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/post
+ */
+#define HB_OT_TAG_post HB_TAG('p','o','s','t')
+
+
+namespace OT {
+
+
+struct postV2Tail
+{
+  friend struct post;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (glyphNameIndex.sanitize (c));
+  }
+
+  protected:
+  ArrayOf<HBUINT16>	glyphNameIndex;	/* This is not an offset, but is the
+					 * ordinal number of the glyph in 'post'
+					 * string tables. */
+/*UnsizedArrayOf<HBUINT8>
+			namesX;*/	/* Glyph names with length bytes [variable]
+					 * (a Pascal string). */
+
+  public:
+  DEFINE_SIZE_ARRAY (2, glyphNameIndex);
+};
+
+struct post
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_post;
+
+  void serialize (hb_serialize_context_t *c) const
+  {
+    post *post_prime = c->allocate_min<post> ();
+    if (unlikely (!post_prime))  return;
+
+    memcpy (post_prime, this, post::min_size);
+    post_prime->version.major = 3; // Version 3 does not have any glyph names.
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    post *post_prime = c->serializer->start_embed<post> ();
+    if (unlikely (!post_prime)) return_trace (false);
+
+    serialize (c->serializer);
+    if (c->serializer->in_error () || c->serializer->ran_out_of_room) return_trace (false);
+
+    return_trace (true);
+  }
+
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    {
+      index_to_offset.init ();
+
+      table = hb_sanitize_context_t ().reference_table<post> (face);
+      unsigned int table_length = table.get_length ();
+
+      version = table->version.to_int ();
+      if (version != 0x00020000) return;
+
+      const postV2Tail &v2 = table->v2X;
+
+      glyphNameIndex = &v2.glyphNameIndex;
+      pool = &StructAfter<uint8_t> (v2.glyphNameIndex);
+
+      const uint8_t *end = (const uint8_t *) (const void *) table + table_length;
+      for (const uint8_t *data = pool;
+	   index_to_offset.length < 65535 && data < end && data + *data < end;
+	   data += 1 + *data)
+	index_to_offset.push (data - pool);
+    }
+    void fini ()
+    {
+      index_to_offset.fini ();
+      free (gids_sorted_by_name.get ());
+      table.destroy ();
+    }
+
+    bool get_glyph_name (hb_codepoint_t glyph,
+			 char *buf, unsigned int buf_len) const
+    {
+      hb_bytes_t s = find_glyph_name (glyph);
+      if (!s.length) return false;
+      if (!buf_len) return true;
+      unsigned int len = hb_min (buf_len - 1, s.length);
+      strncpy (buf, s.arrayZ, len);
+      buf[len] = '\0';
+      return true;
+    }
+
+    bool get_glyph_from_name (const char *name, int len,
+			      hb_codepoint_t *glyph) const
+    {
+      unsigned int count = get_glyph_count ();
+      if (unlikely (!count)) return false;
+
+      if (len < 0) len = strlen (name);
+
+      if (unlikely (!len)) return false;
+
+    retry:
+      uint16_t *gids = gids_sorted_by_name.get ();
+
+      if (unlikely (!gids))
+      {
+	gids = (uint16_t *) malloc (count * sizeof (gids[0]));
+	if (unlikely (!gids))
+	  return false; /* Anything better?! */
+
+	for (unsigned int i = 0; i < count; i++)
+	  gids[i] = i;
+	hb_qsort (gids, count, sizeof (gids[0]), cmp_gids, (void *) this);
+
+	if (unlikely (!gids_sorted_by_name.cmpexch (nullptr, gids)))
+	{
+	  free (gids);
+	  goto retry;
+	}
+      }
+
+      hb_bytes_t st (name, len);
+      auto* gid = hb_bsearch (st, gids, count, sizeof (gids[0]), cmp_key, (void *) this);
+      if (gid)
+      {
+	*glyph = *gid;
+	return true;
+      }
+
+      return false;
+    }
+
+    hb_blob_ptr_t<post> table;
+
+    protected:
+
+    unsigned int get_glyph_count () const
+    {
+      if (version == 0x00010000)
+	return format1_names_length;
+
+      if (version == 0x00020000)
+	return glyphNameIndex->len;
+
+      return 0;
+    }
+
+    static int cmp_gids (const void *pa, const void *pb, void *arg)
+    {
+      const accelerator_t *thiz = (const accelerator_t *) arg;
+      uint16_t a = * (const uint16_t *) pa;
+      uint16_t b = * (const uint16_t *) pb;
+      return thiz->find_glyph_name (b).cmp (thiz->find_glyph_name (a));
+    }
+
+    static int cmp_key (const void *pk, const void *po, void *arg)
+    {
+      const accelerator_t *thiz = (const accelerator_t *) arg;
+      const hb_bytes_t *key = (const hb_bytes_t *) pk;
+      uint16_t o = * (const uint16_t *) po;
+      return thiz->find_glyph_name (o).cmp (*key);
+    }
+
+    hb_bytes_t find_glyph_name (hb_codepoint_t glyph) const
+    {
+      if (version == 0x00010000)
+      {
+	if (glyph >= format1_names_length)
+	  return hb_bytes_t ();
+
+	return format1_names (glyph);
+      }
+
+      if (version != 0x00020000 || glyph >= glyphNameIndex->len)
+	return hb_bytes_t ();
+
+      unsigned int index = glyphNameIndex->arrayZ[glyph];
+      if (index < format1_names_length)
+	return format1_names (index);
+      index -= format1_names_length;
+
+      if (index >= index_to_offset.length)
+	return hb_bytes_t ();
+      unsigned int offset = index_to_offset[index];
+
+      const uint8_t *data = pool + offset;
+      unsigned int name_length = *data;
+      data++;
+
+      return hb_bytes_t ((const char *) data, name_length);
+    }
+
+    private:
+    uint32_t version;
+    const ArrayOf<HBUINT16> *glyphNameIndex;
+    hb_vector_t<uint32_t> index_to_offset;
+    const uint8_t *pool;
+    hb_atomic_ptr_t<uint16_t *> gids_sorted_by_name;
+  };
+
+  bool has_data () const { return version.to_int (); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  (version.to_int () == 0x00010000 ||
+			   (version.to_int () == 0x00020000 && v2X.sanitize (c)) ||
+			   version.to_int () == 0x00030000)));
+  }
+
+  public:
+  FixedVersion<>version;		/* 0x00010000 for version 1.0
+					 * 0x00020000 for version 2.0
+					 * 0x00025000 for version 2.5 (deprecated)
+					 * 0x00030000 for version 3.0 */
+  HBFixed	italicAngle;		/* Italic angle in counter-clockwise degrees
+					 * from the vertical. Zero for upright text,
+					 * negative for text that leans to the right
+					 * (forward). */
+  FWORD		underlinePosition;	/* This is the suggested distance of the top
+					 * of the underline from the baseline
+					 * (negative values indicate below baseline).
+					 * The PostScript definition of this FontInfo
+					 * dictionary key (the y coordinate of the
+					 * center of the stroke) is not used for
+					 * historical reasons. The value of the
+					 * PostScript key may be calculated by
+					 * subtracting half the underlineThickness
+					 * from the value of this field. */
+  FWORD		underlineThickness;	/* Suggested values for the underline
+					   thickness. */
+  HBUINT32	isFixedPitch;		/* Set to 0 if the font is proportionally
+					 * spaced, non-zero if the font is not
+					 * proportionally spaced (i.e. monospaced). */
+  HBUINT32	minMemType42;		/* Minimum memory usage when an OpenType font
+					 * is downloaded. */
+  HBUINT32	maxMemType42;		/* Maximum memory usage when an OpenType font
+					 * is downloaded. */
+  HBUINT32	minMemType1;		/* Minimum memory usage when an OpenType font
+					 * is downloaded as a Type 1 font. */
+  HBUINT32	maxMemType1;		/* Maximum memory usage when an OpenType font
+					 * is downloaded as a Type 1 font. */
+  postV2Tail	v2X;
+  DEFINE_SIZE_MIN (32);
+};
+
+struct post_accelerator_t : post::accelerator_t {};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_POST_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-fallback.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-fallback.hh
new file mode 100644
index 0000000000..244e967b12
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-fallback.hh
@@ -0,0 +1,348 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_ARABIC_FALLBACK_HH
+#define HB_OT_SHAPE_COMPLEX_ARABIC_FALLBACK_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape.hh"
+#include "hb-ot-layout-gsub-table.hh"
+
+
+/* Features ordered the same as the entries in shaping_table rows,
+ * followed by rlig.  Don't change. */
+static const hb_tag_t arabic_fallback_features[] =
+{
+  HB_TAG('i','n','i','t'),
+  HB_TAG('m','e','d','i'),
+  HB_TAG('f','i','n','a'),
+  HB_TAG('i','s','o','l'),
+  HB_TAG('r','l','i','g'),
+};
+
+static OT::SubstLookup *
+arabic_fallback_synthesize_lookup_single (const hb_ot_shape_plan_t *plan HB_UNUSED,
+					  hb_font_t *font,
+					  unsigned int feature_index)
+{
+  OT::HBGlyphID glyphs[SHAPING_TABLE_LAST - SHAPING_TABLE_FIRST + 1];
+  OT::HBGlyphID substitutes[SHAPING_TABLE_LAST - SHAPING_TABLE_FIRST + 1];
+  unsigned int num_glyphs = 0;
+
+  /* Populate arrays */
+  for (hb_codepoint_t u = SHAPING_TABLE_FIRST; u < SHAPING_TABLE_LAST + 1; u++)
+  {
+    hb_codepoint_t s = shaping_table[u - SHAPING_TABLE_FIRST][feature_index];
+    hb_codepoint_t u_glyph, s_glyph;
+
+    if (!s ||
+	!hb_font_get_glyph (font, u, 0, &u_glyph) ||
+	!hb_font_get_glyph (font, s, 0, &s_glyph) ||
+	u_glyph == s_glyph ||
+	u_glyph > 0xFFFFu || s_glyph > 0xFFFFu)
+      continue;
+
+    glyphs[num_glyphs] = u_glyph;
+    substitutes[num_glyphs] = s_glyph;
+
+    num_glyphs++;
+  }
+
+  if (!num_glyphs)
+    return nullptr;
+
+  /* Bubble-sort or something equally good!
+   * May not be good-enough for presidential candidate interviews, but good-enough for us... */
+  hb_stable_sort (&glyphs[0], num_glyphs,
+		  (int(*)(const OT::HBUINT16*, const OT::HBUINT16 *)) OT::HBGlyphID::cmp,
+		  &substitutes[0]);
+
+
+  /* Each glyph takes four bytes max, and there's some overhead. */
+  char buf[(SHAPING_TABLE_LAST - SHAPING_TABLE_FIRST + 1) * 4 + 128];
+  hb_serialize_context_t c (buf, sizeof (buf));
+  OT::SubstLookup *lookup = c.start_serialize<OT::SubstLookup> ();
+  bool ret = lookup->serialize_single (&c,
+				       OT::LookupFlag::IgnoreMarks,
+				       hb_sorted_array (glyphs, num_glyphs),
+				       hb_array (substitutes, num_glyphs));
+  c.end_serialize ();
+
+  return ret && !c.in_error () ? c.copy<OT::SubstLookup> () : nullptr;
+}
+
+static OT::SubstLookup *
+arabic_fallback_synthesize_lookup_ligature (const hb_ot_shape_plan_t *plan HB_UNUSED,
+					    hb_font_t *font)
+{
+  OT::HBGlyphID first_glyphs[ARRAY_LENGTH_CONST (ligature_table)];
+  unsigned int first_glyphs_indirection[ARRAY_LENGTH_CONST (ligature_table)];
+  unsigned int ligature_per_first_glyph_count_list[ARRAY_LENGTH_CONST (first_glyphs)];
+  unsigned int num_first_glyphs = 0;
+
+  /* We know that all our ligatures are 2-component */
+  OT::HBGlyphID ligature_list[ARRAY_LENGTH_CONST (first_glyphs) * ARRAY_LENGTH_CONST(ligature_table[0].ligatures)];
+  unsigned int component_count_list[ARRAY_LENGTH_CONST (ligature_list)];
+  OT::HBGlyphID component_list[ARRAY_LENGTH_CONST (ligature_list) * 1/* One extra component per ligature */];
+  unsigned int num_ligatures = 0;
+
+  /* Populate arrays */
+
+  /* Sort out the first-glyphs */
+  for (unsigned int first_glyph_idx = 0; first_glyph_idx < ARRAY_LENGTH (first_glyphs); first_glyph_idx++)
+  {
+    hb_codepoint_t first_u = ligature_table[first_glyph_idx].first;
+    hb_codepoint_t first_glyph;
+    if (!hb_font_get_glyph (font, first_u, 0, &first_glyph))
+      continue;
+    first_glyphs[num_first_glyphs] = first_glyph;
+    ligature_per_first_glyph_count_list[num_first_glyphs] = 0;
+    first_glyphs_indirection[num_first_glyphs] = first_glyph_idx;
+    num_first_glyphs++;
+  }
+  hb_stable_sort (&first_glyphs[0], num_first_glyphs,
+		  (int(*)(const OT::HBUINT16*, const OT::HBUINT16 *)) OT::HBGlyphID::cmp,
+		  &first_glyphs_indirection[0]);
+
+  /* Now that the first-glyphs are sorted, walk again, populate ligatures. */
+  for (unsigned int i = 0; i < num_first_glyphs; i++)
+  {
+    unsigned int first_glyph_idx = first_glyphs_indirection[i];
+
+    for (unsigned int second_glyph_idx = 0; second_glyph_idx < ARRAY_LENGTH (ligature_table[0].ligatures); second_glyph_idx++)
+    {
+      hb_codepoint_t second_u   = ligature_table[first_glyph_idx].ligatures[second_glyph_idx].second;
+      hb_codepoint_t ligature_u = ligature_table[first_glyph_idx].ligatures[second_glyph_idx].ligature;
+      hb_codepoint_t second_glyph, ligature_glyph;
+      if (!second_u ||
+	  !hb_font_get_glyph (font, second_u,   0, &second_glyph) ||
+	  !hb_font_get_glyph (font, ligature_u, 0, &ligature_glyph))
+	continue;
+
+      ligature_per_first_glyph_count_list[i]++;
+
+      ligature_list[num_ligatures] = ligature_glyph;
+      component_count_list[num_ligatures] = 2;
+      component_list[num_ligatures] = second_glyph;
+      num_ligatures++;
+    }
+  }
+
+  if (!num_ligatures)
+    return nullptr;
+
+
+  /* 16 bytes per ligature ought to be enough... */
+  char buf[ARRAY_LENGTH_CONST (ligature_list) * 16 + 128];
+  hb_serialize_context_t c (buf, sizeof (buf));
+  OT::SubstLookup *lookup = c.start_serialize<OT::SubstLookup> ();
+  bool ret = lookup->serialize_ligature (&c,
+					 OT::LookupFlag::IgnoreMarks,
+					 hb_sorted_array (first_glyphs, num_first_glyphs),
+					 hb_array (ligature_per_first_glyph_count_list, num_first_glyphs),
+					 hb_array (ligature_list, num_ligatures),
+					 hb_array (component_count_list, num_ligatures),
+					 hb_array (component_list, num_ligatures));
+  c.end_serialize ();
+  /* TODO sanitize the results? */
+
+  return ret && !c.in_error () ? c.copy<OT::SubstLookup> () : nullptr;
+}
+
+static OT::SubstLookup *
+arabic_fallback_synthesize_lookup (const hb_ot_shape_plan_t *plan,
+				   hb_font_t *font,
+				   unsigned int feature_index)
+{
+  if (feature_index < 4)
+    return arabic_fallback_synthesize_lookup_single (plan, font, feature_index);
+  else
+    return arabic_fallback_synthesize_lookup_ligature (plan, font);
+}
+
+#define ARABIC_FALLBACK_MAX_LOOKUPS 5
+
+struct arabic_fallback_plan_t
+{
+  unsigned int num_lookups;
+  bool free_lookups;
+
+  hb_mask_t mask_array[ARABIC_FALLBACK_MAX_LOOKUPS];
+  OT::SubstLookup *lookup_array[ARABIC_FALLBACK_MAX_LOOKUPS];
+  OT::hb_ot_layout_lookup_accelerator_t accel_array[ARABIC_FALLBACK_MAX_LOOKUPS];
+};
+
+#if defined(_WIN32) && !defined(HB_NO_WIN1256)
+#define HB_WITH_WIN1256
+#endif
+
+#ifdef HB_WITH_WIN1256
+#include "hb-ot-shape-complex-arabic-win1256.hh"
+#endif
+
+struct ManifestLookup
+{
+  public:
+  OT::Tag tag;
+  OT::OffsetTo<OT::SubstLookup> lookupOffset;
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+typedef OT::ArrayOf<ManifestLookup> Manifest;
+
+static bool
+arabic_fallback_plan_init_win1256 (arabic_fallback_plan_t *fallback_plan HB_UNUSED,
+				   const hb_ot_shape_plan_t *plan HB_UNUSED,
+				   hb_font_t *font HB_UNUSED)
+{
+#ifdef HB_WITH_WIN1256
+  /* Does this font look like it's Windows-1256-encoded? */
+  hb_codepoint_t g;
+  if (!(hb_font_get_glyph (font, 0x0627u, 0, &g) && g == 199 /* ALEF */ &&
+	hb_font_get_glyph (font, 0x0644u, 0, &g) && g == 225 /* LAM */ &&
+	hb_font_get_glyph (font, 0x0649u, 0, &g) && g == 236 /* ALEF MAKSURA */ &&
+	hb_font_get_glyph (font, 0x064Au, 0, &g) && g == 237 /* YEH */ &&
+	hb_font_get_glyph (font, 0x0652u, 0, &g) && g == 250 /* SUKUN */))
+    return false;
+
+  const Manifest &manifest = reinterpret_cast<const Manifest&> (arabic_win1256_gsub_lookups.manifest);
+  static_assert (sizeof (arabic_win1256_gsub_lookups.manifestData) ==
+		 ARABIC_FALLBACK_MAX_LOOKUPS * sizeof (ManifestLookup), "");
+  /* TODO sanitize the table? */
+
+  unsigned j = 0;
+  unsigned int count = manifest.len;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    fallback_plan->mask_array[j] = plan->map.get_1_mask (manifest[i].tag);
+    if (fallback_plan->mask_array[j])
+    {
+      fallback_plan->lookup_array[j] = const_cast<OT::SubstLookup*> (&(&manifest+manifest[i].lookupOffset));
+      if (fallback_plan->lookup_array[j])
+      {
+	fallback_plan->accel_array[j].init (*fallback_plan->lookup_array[j]);
+	j++;
+      }
+    }
+  }
+
+  fallback_plan->num_lookups = j;
+  fallback_plan->free_lookups = false;
+
+  return j > 0;
+#else
+  return false;
+#endif
+}
+
+static bool
+arabic_fallback_plan_init_unicode (arabic_fallback_plan_t *fallback_plan,
+				   const hb_ot_shape_plan_t *plan,
+				   hb_font_t *font)
+{
+  static_assert ((ARRAY_LENGTH_CONST(arabic_fallback_features) <= ARABIC_FALLBACK_MAX_LOOKUPS), "");
+  unsigned int j = 0;
+  for (unsigned int i = 0; i < ARRAY_LENGTH(arabic_fallback_features) ; i++)
+  {
+    fallback_plan->mask_array[j] = plan->map.get_1_mask (arabic_fallback_features[i]);
+    if (fallback_plan->mask_array[j])
+    {
+      fallback_plan->lookup_array[j] = arabic_fallback_synthesize_lookup (plan, font, i);
+      if (fallback_plan->lookup_array[j])
+      {
+	fallback_plan->accel_array[j].init (*fallback_plan->lookup_array[j]);
+	j++;
+      }
+    }
+  }
+
+  fallback_plan->num_lookups = j;
+  fallback_plan->free_lookups = true;
+
+  return j > 0;
+}
+
+static arabic_fallback_plan_t *
+arabic_fallback_plan_create (const hb_ot_shape_plan_t *plan,
+			     hb_font_t *font)
+{
+  arabic_fallback_plan_t *fallback_plan = (arabic_fallback_plan_t *) calloc (1, sizeof (arabic_fallback_plan_t));
+  if (unlikely (!fallback_plan))
+    return const_cast<arabic_fallback_plan_t *> (&Null (arabic_fallback_plan_t));
+
+  fallback_plan->num_lookups = 0;
+  fallback_plan->free_lookups = false;
+
+  /* Try synthesizing GSUB table using Unicode Arabic Presentation Forms,
+   * in case the font has cmap entries for the presentation-forms characters. */
+  if (arabic_fallback_plan_init_unicode (fallback_plan, plan, font))
+    return fallback_plan;
+
+  /* See if this looks like a Windows-1256-encoded font.  If it does, use a
+   * hand-coded GSUB table. */
+  if (arabic_fallback_plan_init_win1256 (fallback_plan, plan, font))
+    return fallback_plan;
+
+  assert (fallback_plan->num_lookups == 0);
+  free (fallback_plan);
+  return const_cast<arabic_fallback_plan_t *> (&Null (arabic_fallback_plan_t));
+}
+
+static void
+arabic_fallback_plan_destroy (arabic_fallback_plan_t *fallback_plan)
+{
+  if (!fallback_plan || fallback_plan->num_lookups == 0)
+    return;
+
+  for (unsigned int i = 0; i < fallback_plan->num_lookups; i++)
+    if (fallback_plan->lookup_array[i])
+    {
+      fallback_plan->accel_array[i].fini ();
+      if (fallback_plan->free_lookups)
+	free (fallback_plan->lookup_array[i]);
+    }
+
+  free (fallback_plan);
+}
+
+static void
+arabic_fallback_plan_shape (arabic_fallback_plan_t *fallback_plan,
+			    hb_font_t *font,
+			    hb_buffer_t *buffer)
+{
+  OT::hb_ot_apply_context_t c (0, font, buffer);
+  for (unsigned int i = 0; i < fallback_plan->num_lookups; i++)
+    if (fallback_plan->lookup_array[i]) {
+      c.set_lookup_mask (fallback_plan->mask_array[i]);
+      hb_ot_layout_substitute_lookup (&c,
+				      *fallback_plan->lookup_array[i],
+				      fallback_plan->accel_array[i]);
+    }
+}
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_ARABIC_FALLBACK_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-joining-list.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-joining-list.hh
new file mode 100644
index 0000000000..c022d4bb06
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-joining-list.hh
@@ -0,0 +1,46 @@
+/* == Start of generated function == */
+/*
+ * The following function is generated by running:
+ *
+ *   ./gen-arabic-joining-list.py ArabicShaping.txt Scripts.txt
+ *
+ * on files with these headers:
+ *
+ * # ArabicShaping-13.0.0.txt
+ * # Date: 2020-01-31, 23:55:00 GMT [KW, RP]
+ * # Scripts-13.0.0.txt
+ * # Date: 2020-01-22, 00:07:43 GMT
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_ARABIC_JOINING_LIST_HH
+#define HB_OT_SHAPE_COMPLEX_ARABIC_JOINING_LIST_HH
+
+static bool
+has_arabic_joining (hb_script_t script)
+{
+  /* List of scripts that have data in arabic-table. */
+  switch ((int) script)
+  {
+    case HB_SCRIPT_ADLAM:
+    case HB_SCRIPT_ARABIC:
+    case HB_SCRIPT_CHORASMIAN:
+    case HB_SCRIPT_HANIFI_ROHINGYA:
+    case HB_SCRIPT_MANDAIC:
+    case HB_SCRIPT_MANICHAEAN:
+    case HB_SCRIPT_MONGOLIAN:
+    case HB_SCRIPT_NKO:
+    case HB_SCRIPT_PHAGS_PA:
+    case HB_SCRIPT_PSALTER_PAHLAVI:
+    case HB_SCRIPT_SOGDIAN:
+    case HB_SCRIPT_SYRIAC:
+      return true;
+
+    default:
+      return false;
+  }
+}
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_ARABIC_JOINING_LIST_HH */
+
+/* == End of generated function == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-table.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-table.hh
new file mode 100644
index 0000000000..70ffe623c0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-table.hh
@@ -0,0 +1,433 @@
+/* == Start of generated table == */
+/*
+ * The following table is generated by running:
+ *
+ *   ./gen-arabic-table.py ArabicShaping.txt UnicodeData.txt Blocks.txt
+ *
+ * on files with these headers:
+ *
+ * # ArabicShaping-13.0.0.txt
+ * # Date: 2020-01-31, 23:55:00 GMT [KW, RP]
+ * # Blocks-13.0.0.txt
+ * # Date: 2019-07-10, 19:06:00 GMT [KW]
+ * UnicodeData.txt does not have a header.
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_ARABIC_TABLE_HH
+#define HB_OT_SHAPE_COMPLEX_ARABIC_TABLE_HH
+
+
+#define A	JOINING_GROUP_ALAPH
+#define DR	JOINING_GROUP_DALATH_RISH
+#define C	JOINING_TYPE_C
+#define D	JOINING_TYPE_D
+#define L	JOINING_TYPE_L
+#define R	JOINING_TYPE_R
+#define T	JOINING_TYPE_T
+#define U	JOINING_TYPE_U
+#define X	JOINING_TYPE_X
+
+static const uint8_t joining_table[] =
+{
+
+#define joining_offset_0x0600u 0
+
+  /* Arabic */
+
+  /* 0600 */ U,U,U,U,U,U,X,X,U,X,X,U,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 0620 */ D,U,R,R,R,R,D,R,D,R,D,D,D,D,D,R,R,R,R,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 0640 */ C,D,D,D,D,D,D,D,R,D,D,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 0660 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,D,D,X,R,R,R,U,R,R,R,D,D,D,D,D,D,D,D,
+  /* 0680 */ D,D,D,D,D,D,D,D,R,R,R,R,R,R,R,R,R,R,R,R,R,R,R,R,R,R,D,D,D,D,D,D,
+  /* 06A0 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 06C0 */ R,D,D,R,R,R,R,R,R,R,R,R,D,R,D,R,D,D,R,R,X,R,X,X,X,X,X,X,X,U,X,X,
+  /* 06E0 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,R,R,X,X,X,X,X,X,X,X,X,X,D,D,D,X,X,D,
+
+  /* Syriac */
+
+  /* 0700 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,T,A,X,D,D,D,DR,DR,R,R,R,D,D,D,D,R,D,
+  /* 0720 */ D,D,D,D,D,D,D,D,R,D,DR,D,R,D,D,DR,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 0740 */ X,X,X,X,X,X,X,X,X,X,X,X,X,R,D,D,
+
+  /* Arabic Supplement */
+
+  /* 0740 */                                 D,D,D,D,D,D,D,D,D,R,R,R,D,D,D,D,
+  /* 0760 */ D,D,D,D,D,D,D,D,D,D,D,R,R,D,D,D,D,R,D,R,R,D,D,D,R,R,D,D,D,D,D,D,
+
+  /* FILLER */
+
+  /* 0780 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 07A0 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+
+  /* NKo */
+
+  /* 07C0 */ X,X,X,X,X,X,X,X,X,X,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 07E0 */ D,D,D,D,D,D,D,D,D,D,D,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,C,X,X,X,X,X,
+
+  /* FILLER */
+
+  /* 0800 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 0820 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+
+  /* Mandaic */
+
+  /* 0840 */ R,D,D,D,D,D,R,R,D,R,D,D,D,D,D,D,D,D,D,D,R,D,R,R,R,X,X,X,X,X,X,X,
+
+  /* Syriac Supplement */
+
+  /* 0860 */ D,U,D,D,D,D,U,R,D,R,R,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 0880 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+
+  /* Arabic Extended-A */
+
+  /* 08A0 */ D,D,D,D,D,D,D,D,D,D,R,R,R,U,R,D,D,R,R,D,D,X,D,D,D,R,D,D,D,D,D,D,
+  /* 08C0 */ D,D,D,D,D,D,D,D,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 08E0 */ X,X,U,
+
+#define joining_offset_0x1806u 739
+
+  /* Mongolian */
+
+  /* 1800 */             U,D,X,X,C,X,X,X,U,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 1820 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 1840 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 1860 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,X,X,X,X,X,X,X,
+  /* 1880 */ U,U,U,U,U,T,T,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 18A0 */ D,D,D,D,D,D,D,D,D,X,D,
+
+#define joining_offset_0x200cu 904
+
+  /* General Punctuation */
+
+  /* 2000 */                         U,C,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 2020 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,U,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 2040 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 2060 */ X,X,X,X,X,X,U,U,U,U,
+
+#define joining_offset_0xa840u 998
+
+  /* Phags-pa */
+
+  /* A840 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* A860 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,L,U,
+
+#define joining_offset_0x10ac0u 1050
+
+  /* Manichaean */
+
+  /* 10AC0 */ D,D,D,D,D,R,U,R,U,R,R,U,U,L,R,R,R,R,R,D,D,D,D,L,D,D,D,D,D,R,D,D,
+  /* 10AE0 */ D,R,U,U,R,X,X,X,X,X,X,D,D,D,D,R,
+
+#define joining_offset_0x10b80u 1098
+
+  /* Psalter Pahlavi */
+
+  /* 10B80 */ D,R,D,R,R,R,D,D,D,R,D,D,R,D,R,R,D,R,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+  /* 10BA0 */ X,X,X,X,X,X,X,X,X,R,R,R,R,D,D,U,
+
+#define joining_offset_0x10d00u 1146
+
+  /* Hanifi Rohingya */
+
+  /* 10D00 */ L,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 10D20 */ D,D,R,D,
+
+#define joining_offset_0x10f30u 1182
+
+  /* Sogdian */
+
+  /* 10F20 */                                 D,D,D,R,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 10F40 */ D,D,D,D,D,U,X,X,X,X,X,X,X,X,X,X,X,D,D,D,R,
+
+#define joining_offset_0x10fb0u 1219
+
+  /* Chorasmian */
+
+  /* 10FA0 */                                 D,U,D,D,R,R,R,U,D,R,R,D,D,R,D,D,
+  /* 10FC0 */ U,D,R,R,D,U,U,U,U,R,D,L,
+
+#define joining_offset_0x110bdu 1247
+
+  /* Kaithi */
+
+  /* 110A0 */                                                           U,X,X,
+  /* 110C0 */ X,X,X,X,X,X,X,X,X,X,X,X,X,U,
+
+#define joining_offset_0x1e900u 1264
+
+  /* Adlam */
+
+  /* 1E900 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 1E920 */ D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 1E940 */ D,D,D,D,X,X,X,X,X,X,X,T,
+
+}; /* Table items: 1340; occupancy: 57% */
+
+
+static unsigned int
+joining_type (hb_codepoint_t u)
+{
+  switch (u >> 12)
+  {
+    case 0x0u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x0600u, 0x08E2u)) return joining_table[u - 0x0600u + joining_offset_0x0600u];
+      break;
+
+    case 0x1u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x1806u, 0x18AAu)) return joining_table[u - 0x1806u + joining_offset_0x1806u];
+      break;
+
+    case 0x2u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x200Cu, 0x2069u)) return joining_table[u - 0x200Cu + joining_offset_0x200cu];
+      break;
+
+    case 0xAu:
+      if (hb_in_range<hb_codepoint_t> (u, 0xA840u, 0xA873u)) return joining_table[u - 0xA840u + joining_offset_0xa840u];
+      break;
+
+    case 0x10u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x10AC0u, 0x10AEFu)) return joining_table[u - 0x10AC0u + joining_offset_0x10ac0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x10B80u, 0x10BAFu)) return joining_table[u - 0x10B80u + joining_offset_0x10b80u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x10D00u, 0x10D23u)) return joining_table[u - 0x10D00u + joining_offset_0x10d00u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x10F30u, 0x10F54u)) return joining_table[u - 0x10F30u + joining_offset_0x10f30u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x10FB0u, 0x10FCBu)) return joining_table[u - 0x10FB0u + joining_offset_0x10fb0u];
+      break;
+
+    case 0x11u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x110BDu, 0x110CDu)) return joining_table[u - 0x110BDu + joining_offset_0x110bdu];
+      break;
+
+    case 0x1Eu:
+      if (hb_in_range<hb_codepoint_t> (u, 0x1E900u, 0x1E94Bu)) return joining_table[u - 0x1E900u + joining_offset_0x1e900u];
+      break;
+
+    default:
+      break;
+  }
+  return X;
+}
+
+#undef A
+#undef DR
+#undef C
+#undef D
+#undef L
+#undef R
+#undef T
+#undef U
+#undef X
+
+
+static const uint16_t shaping_table[][4] =
+{
+  {0x0000u, 0x0000u, 0x0000u, 0xFE80u}, /* U+0621 ARABIC LETTER HAMZA ISOLATED FORM */
+  {0x0000u, 0x0000u, 0xFE82u, 0xFE81u}, /* U+0622 ARABIC LETTER ALEF WITH MADDA ABOVE */
+  {0x0000u, 0x0000u, 0xFE84u, 0xFE83u}, /* U+0623 ARABIC LETTER ALEF WITH HAMZA ABOVE */
+  {0x0000u, 0x0000u, 0xFE86u, 0xFE85u}, /* U+0624 ARABIC LETTER WAW WITH HAMZA ABOVE */
+  {0x0000u, 0x0000u, 0xFE88u, 0xFE87u}, /* U+0625 ARABIC LETTER ALEF WITH HAMZA BELOW */
+  {0xFE8Bu, 0xFE8Cu, 0xFE8Au, 0xFE89u}, /* U+0626 ARABIC LETTER YEH WITH HAMZA ABOVE */
+  {0x0000u, 0x0000u, 0xFE8Eu, 0xFE8Du}, /* U+0627 ARABIC LETTER ALEF */
+  {0xFE91u, 0xFE92u, 0xFE90u, 0xFE8Fu}, /* U+0628 ARABIC LETTER BEH */
+  {0x0000u, 0x0000u, 0xFE94u, 0xFE93u}, /* U+0629 ARABIC LETTER TEH MARBUTA */
+  {0xFE97u, 0xFE98u, 0xFE96u, 0xFE95u}, /* U+062A ARABIC LETTER TEH */
+  {0xFE9Bu, 0xFE9Cu, 0xFE9Au, 0xFE99u}, /* U+062B ARABIC LETTER THEH */
+  {0xFE9Fu, 0xFEA0u, 0xFE9Eu, 0xFE9Du}, /* U+062C ARABIC LETTER JEEM */
+  {0xFEA3u, 0xFEA4u, 0xFEA2u, 0xFEA1u}, /* U+062D ARABIC LETTER HAH */
+  {0xFEA7u, 0xFEA8u, 0xFEA6u, 0xFEA5u}, /* U+062E ARABIC LETTER KHAH */
+  {0x0000u, 0x0000u, 0xFEAAu, 0xFEA9u}, /* U+062F ARABIC LETTER DAL */
+  {0x0000u, 0x0000u, 0xFEACu, 0xFEABu}, /* U+0630 ARABIC LETTER THAL */
+  {0x0000u, 0x0000u, 0xFEAEu, 0xFEADu}, /* U+0631 ARABIC LETTER REH */
+  {0x0000u, 0x0000u, 0xFEB0u, 0xFEAFu}, /* U+0632 ARABIC LETTER ZAIN */
+  {0xFEB3u, 0xFEB4u, 0xFEB2u, 0xFEB1u}, /* U+0633 ARABIC LETTER SEEN */
+  {0xFEB7u, 0xFEB8u, 0xFEB6u, 0xFEB5u}, /* U+0634 ARABIC LETTER SHEEN */
+  {0xFEBBu, 0xFEBCu, 0xFEBAu, 0xFEB9u}, /* U+0635 ARABIC LETTER SAD */
+  {0xFEBFu, 0xFEC0u, 0xFEBEu, 0xFEBDu}, /* U+0636 ARABIC LETTER DAD */
+  {0xFEC3u, 0xFEC4u, 0xFEC2u, 0xFEC1u}, /* U+0637 ARABIC LETTER TAH */
+  {0xFEC7u, 0xFEC8u, 0xFEC6u, 0xFEC5u}, /* U+0638 ARABIC LETTER ZAH */
+  {0xFECBu, 0xFECCu, 0xFECAu, 0xFEC9u}, /* U+0639 ARABIC LETTER AIN */
+  {0xFECFu, 0xFED0u, 0xFECEu, 0xFECDu}, /* U+063A ARABIC LETTER GHAIN */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+063B  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+063C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+063D  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+063E  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+063F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0640  */
+  {0xFED3u, 0xFED4u, 0xFED2u, 0xFED1u}, /* U+0641 ARABIC LETTER FEH */
+  {0xFED7u, 0xFED8u, 0xFED6u, 0xFED5u}, /* U+0642 ARABIC LETTER QAF */
+  {0xFEDBu, 0xFEDCu, 0xFEDAu, 0xFED9u}, /* U+0643 ARABIC LETTER KAF */
+  {0xFEDFu, 0xFEE0u, 0xFEDEu, 0xFEDDu}, /* U+0644 ARABIC LETTER LAM */
+  {0xFEE3u, 0xFEE4u, 0xFEE2u, 0xFEE1u}, /* U+0645 ARABIC LETTER MEEM */
+  {0xFEE7u, 0xFEE8u, 0xFEE6u, 0xFEE5u}, /* U+0646 ARABIC LETTER NOON */
+  {0xFEEBu, 0xFEECu, 0xFEEAu, 0xFEE9u}, /* U+0647 ARABIC LETTER HEH */
+  {0x0000u, 0x0000u, 0xFEEEu, 0xFEEDu}, /* U+0648 ARABIC LETTER WAW */
+  {0xFBE8u, 0xFBE9u, 0xFEF0u, 0xFEEFu}, /* U+0649 ARABIC LETTER */
+  {0xFEF3u, 0xFEF4u, 0xFEF2u, 0xFEF1u}, /* U+064A ARABIC LETTER YEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+064B  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+064C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+064D  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+064E  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+064F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0650  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0651  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0652  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0653  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0654  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0655  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0656  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0657  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0658  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0659  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065A  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065B  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065D  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065E  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+065F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0660  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0661  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0662  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0663  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0664  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0665  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0666  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0667  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0668  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0669  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066A  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066B  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066D  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066E  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+066F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0670  */
+  {0x0000u, 0x0000u, 0xFB51u, 0xFB50u}, /* U+0671 ARABIC LETTER ALEF WASLA */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0672  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0673  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0674  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0675  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0676  */
+  {0x0000u, 0x0000u, 0x0000u, 0xFBDDu}, /* U+0677 ARABIC LETTER U WITH HAMZA ABOVE ISOLATED FORM */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0678  */
+  {0xFB68u, 0xFB69u, 0xFB67u, 0xFB66u}, /* U+0679 ARABIC LETTER TTEH */
+  {0xFB60u, 0xFB61u, 0xFB5Fu, 0xFB5Eu}, /* U+067A ARABIC LETTER TTEHEH */
+  {0xFB54u, 0xFB55u, 0xFB53u, 0xFB52u}, /* U+067B ARABIC LETTER BEEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+067C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+067D  */
+  {0xFB58u, 0xFB59u, 0xFB57u, 0xFB56u}, /* U+067E ARABIC LETTER PEH */
+  {0xFB64u, 0xFB65u, 0xFB63u, 0xFB62u}, /* U+067F ARABIC LETTER TEHEH */
+  {0xFB5Cu, 0xFB5Du, 0xFB5Bu, 0xFB5Au}, /* U+0680 ARABIC LETTER BEHEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0681  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0682  */
+  {0xFB78u, 0xFB79u, 0xFB77u, 0xFB76u}, /* U+0683 ARABIC LETTER NYEH */
+  {0xFB74u, 0xFB75u, 0xFB73u, 0xFB72u}, /* U+0684 ARABIC LETTER DYEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0685  */
+  {0xFB7Cu, 0xFB7Du, 0xFB7Bu, 0xFB7Au}, /* U+0686 ARABIC LETTER TCHEH */
+  {0xFB80u, 0xFB81u, 0xFB7Fu, 0xFB7Eu}, /* U+0687 ARABIC LETTER TCHEHEH */
+  {0x0000u, 0x0000u, 0xFB89u, 0xFB88u}, /* U+0688 ARABIC LETTER DDAL */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0689  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+068A  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+068B  */
+  {0x0000u, 0x0000u, 0xFB85u, 0xFB84u}, /* U+068C ARABIC LETTER DAHAL */
+  {0x0000u, 0x0000u, 0xFB83u, 0xFB82u}, /* U+068D ARABIC LETTER DDAHAL */
+  {0x0000u, 0x0000u, 0xFB87u, 0xFB86u}, /* U+068E ARABIC LETTER DUL */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+068F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0690  */
+  {0x0000u, 0x0000u, 0xFB8Du, 0xFB8Cu}, /* U+0691 ARABIC LETTER RREH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0692  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0693  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0694  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0695  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0696  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0697  */
+  {0x0000u, 0x0000u, 0xFB8Bu, 0xFB8Au}, /* U+0698 ARABIC LETTER JEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+0699  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069A  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069B  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069C  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069D  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069E  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+069F  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A0  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A1  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A2  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A3  */
+  {0xFB6Cu, 0xFB6Du, 0xFB6Bu, 0xFB6Au}, /* U+06A4 ARABIC LETTER VEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A5  */
+  {0xFB70u, 0xFB71u, 0xFB6Fu, 0xFB6Eu}, /* U+06A6 ARABIC LETTER PEHEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A7  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06A8  */
+  {0xFB90u, 0xFB91u, 0xFB8Fu, 0xFB8Eu}, /* U+06A9 ARABIC LETTER KEHEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06AA  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06AB  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06AC  */
+  {0xFBD5u, 0xFBD6u, 0xFBD4u, 0xFBD3u}, /* U+06AD ARABIC LETTER NG */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06AE  */
+  {0xFB94u, 0xFB95u, 0xFB93u, 0xFB92u}, /* U+06AF ARABIC LETTER GAF */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B0  */
+  {0xFB9Cu, 0xFB9Du, 0xFB9Bu, 0xFB9Au}, /* U+06B1 ARABIC LETTER NGOEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B2  */
+  {0xFB98u, 0xFB99u, 0xFB97u, 0xFB96u}, /* U+06B3 ARABIC LETTER GUEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B4  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B5  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B6  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B7  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B8  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06B9  */
+  {0x0000u, 0x0000u, 0xFB9Fu, 0xFB9Eu}, /* U+06BA ARABIC LETTER NOON GHUNNA */
+  {0xFBA2u, 0xFBA3u, 0xFBA1u, 0xFBA0u}, /* U+06BB ARABIC LETTER RNOON */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06BC  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06BD  */
+  {0xFBACu, 0xFBADu, 0xFBABu, 0xFBAAu}, /* U+06BE ARABIC LETTER HEH DOACHASHMEE */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06BF  */
+  {0x0000u, 0x0000u, 0xFBA5u, 0xFBA4u}, /* U+06C0 ARABIC LETTER HEH WITH YEH ABOVE */
+  {0xFBA8u, 0xFBA9u, 0xFBA7u, 0xFBA6u}, /* U+06C1 ARABIC LETTER HEH GOAL */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06C2  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06C3  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06C4  */
+  {0x0000u, 0x0000u, 0xFBE1u, 0xFBE0u}, /* U+06C5 ARABIC LETTER KIRGHIZ OE */
+  {0x0000u, 0x0000u, 0xFBDAu, 0xFBD9u}, /* U+06C6 ARABIC LETTER OE */
+  {0x0000u, 0x0000u, 0xFBD8u, 0xFBD7u}, /* U+06C7 ARABIC LETTER U */
+  {0x0000u, 0x0000u, 0xFBDCu, 0xFBDBu}, /* U+06C8 ARABIC LETTER YU */
+  {0x0000u, 0x0000u, 0xFBE3u, 0xFBE2u}, /* U+06C9 ARABIC LETTER KIRGHIZ YU */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06CA  */
+  {0x0000u, 0x0000u, 0xFBDFu, 0xFBDEu}, /* U+06CB ARABIC LETTER VE */
+  {0xFBFEu, 0xFBFFu, 0xFBFDu, 0xFBFCu}, /* U+06CC ARABIC LETTER FARSI YEH */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06CD  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06CE  */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06CF  */
+  {0xFBE6u, 0xFBE7u, 0xFBE5u, 0xFBE4u}, /* U+06D0 ARABIC LETTER E */
+  {0x0000u, 0x0000u, 0x0000u, 0x0000u}, /* U+06D1  */
+  {0x0000u, 0x0000u, 0xFBAFu, 0xFBAEu}, /* U+06D2 ARABIC LETTER YEH BARREE */
+  {0x0000u, 0x0000u, 0xFBB1u, 0xFBB0u}, /* U+06D3 ARABIC LETTER YEH BARREE WITH HAMZA ABOVE */
+};
+
+#define SHAPING_TABLE_FIRST	0x0621u
+#define SHAPING_TABLE_LAST	0x06D3u
+
+
+static const struct ligature_set_t {
+ uint16_t first;
+ struct ligature_pairs_t {
+   uint16_t second;
+   uint16_t ligature;
+ } ligatures[4];
+} ligature_table[] =
+{
+  { 0xFEDFu, {
+    { 0xFE82u, 0xFEF5u }, /* ARABIC LIGATURE LAM WITH ALEF WITH MADDA ABOVE ISOLATED FORM */
+    { 0xFE84u, 0xFEF7u }, /* ARABIC LIGATURE LAM WITH ALEF WITH HAMZA ABOVE ISOLATED FORM */
+    { 0xFE88u, 0xFEF9u }, /* ARABIC LIGATURE LAM WITH ALEF WITH HAMZA BELOW ISOLATED FORM */
+    { 0xFE8Eu, 0xFEFBu }, /* ARABIC LIGATURE LAM WITH ALEF ISOLATED FORM */
+  }},
+  { 0xFEE0u, {
+    { 0xFE82u, 0xFEF6u }, /* ARABIC LIGATURE LAM WITH ALEF WITH MADDA ABOVE FINAL FORM */
+    { 0xFE84u, 0xFEF8u }, /* ARABIC LIGATURE LAM WITH ALEF WITH HAMZA ABOVE FINAL FORM */
+    { 0xFE88u, 0xFEFAu }, /* ARABIC LIGATURE LAM WITH ALEF WITH HAMZA BELOW FINAL FORM */
+    { 0xFE8Eu, 0xFEFCu }, /* ARABIC LIGATURE LAM WITH ALEF FINAL FORM */
+  }},
+};
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_ARABIC_TABLE_HH */
+
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-win1256.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-win1256.hh
new file mode 100644
index 0000000000..b15e145f2f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic-win1256.hh
@@ -0,0 +1,323 @@
+/*
+ * Copyright © 2014  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_ARABIC_WIN1256_HH
+
+
+/*
+ * The macros in the first part of this file are generic macros that can
+ * be used to define the bytes for OpenType table data in code in a
+ * readable manner.  We can move the macros to reside with their respective
+ * struct types, but since we only use these to define one data table, the
+ * Windows-1256 Arabic shaping table in this file, we keep them here.
+ */
+
+
+/* First we measure, then we cut. */
+#ifndef OT_MEASURE
+#define OT_MEASURE
+#define OT_TABLE_START			static const struct TABLE_NAME {
+#define OT_TABLE_END			}
+#define OT_LABEL_START(Name)		unsigned char Name[
+#define OT_LABEL_END			];
+#define OT_UINT8(u8)			+1/*byte*/
+#define OT_UINT16(u16)			+2/*bytes*/
+#else
+#undef  OT_MEASURE
+#define OT_TABLE_START			TABLE_NAME = {
+#define OT_TABLE_END			};
+#define OT_LABEL_START(Name)		{
+#define OT_LABEL_END			},
+#define OT_UINT8(u8)			(u8),
+#define OT_UINT16(u16)			(unsigned char)((u16)>>8), (unsigned char)((u16)&0xFFu),
+#define OT_COUNT(Name, ItemSize)	((unsigned int) sizeof(((struct TABLE_NAME*)0)->Name) \
+					 / (unsigned int)(ItemSize) \
+					 /* OT_ASSERT it's divisible (and positive). */)
+#define OT_DISTANCE(From,To)		((unsigned int) \
+					 ((char*)(&((struct TABLE_NAME*)0)->To) - \
+					  (char*)(&((struct TABLE_NAME*)0)->From)) \
+					 /* OT_ASSERT it's positive. */)
+#endif
+
+
+#define OT_LABEL(Name) \
+	OT_LABEL_END \
+	OT_LABEL_START(Name)
+
+/* Whenever we receive an argument that is a list, it will expand to
+ * contain commas.  That cannot be passed to another macro because the
+ * commas will throw off the preprocessor.  The solution is to wrap
+ * the passed-in argument in OT_LIST() before passing to the next macro.
+ * Unfortunately this trick requires vararg macros. */
+#define OT_LIST(...) __VA_ARGS__
+
+
+/*
+ * Basic Types
+ */
+
+#define OT_TAG(a,b,c,d) \
+	OT_UINT8(a) OT_UINT8(b) OT_UINT8(c) OT_UINT8(d)
+
+#define OT_OFFSET(From, To) /* Offset from From to To in bytes */ \
+	OT_UINT16(OT_DISTANCE(From, To))
+
+#define OT_GLYPHID /* GlyphID */ \
+	OT_UINT16
+
+#define OT_UARRAY(Name, Items) \
+	OT_LABEL_START(Name) \
+	OT_UINT16(OT_COUNT(Name##Data, 2)) \
+	OT_LABEL(Name##Data) \
+	Items \
+	OT_LABEL_END
+
+#define OT_UHEADLESSARRAY(Name, Items) \
+	OT_LABEL_START(Name) \
+	OT_UINT16(OT_COUNT(Name##Data, 2) + 1) \
+	OT_LABEL(Name##Data) \
+	Items \
+	OT_LABEL_END
+
+
+/*
+ * Common Types
+ */
+
+#define OT_LOOKUP_FLAG_IGNORE_MARKS	0x08u
+
+#define OT_LOOKUP(Name, LookupType, LookupFlag, SubLookupOffsets) \
+	OT_LABEL_START(Name) \
+	OT_UINT16(LookupType) \
+	OT_UINT16(LookupFlag) \
+	OT_LABEL_END \
+	OT_UARRAY(Name##SubLookupOffsetsArray, OT_LIST(SubLookupOffsets))
+
+#define OT_SUBLOOKUP(Name, SubFormat, Items) \
+	OT_LABEL_START(Name) \
+	OT_UINT16(SubFormat) \
+	Items
+
+#define OT_COVERAGE1(Name, Items) \
+	OT_LABEL_START(Name) \
+	OT_UINT16(1) \
+	OT_LABEL_END \
+	OT_UARRAY(Name##Glyphs, OT_LIST(Items))
+
+
+/*
+ * GSUB
+ */
+
+#define OT_LOOKUP_TYPE_SUBST_SINGLE	1u
+#define OT_LOOKUP_TYPE_SUBST_LIGATURE	4u
+
+#define OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(Name, FromGlyphs, ToGlyphs) \
+	OT_SUBLOOKUP(Name, 2, \
+		OT_OFFSET(Name, Name##Coverage) \
+		OT_LABEL_END \
+		OT_UARRAY(Name##Substitute, OT_LIST(ToGlyphs)) \
+	) \
+	OT_COVERAGE1(Name##Coverage, OT_LIST(FromGlyphs)) \
+	/* ASSERT_STATIC_EXPR_ZERO (len(FromGlyphs) == len(ToGlyphs)) */
+
+#define OT_SUBLOOKUP_LIGATURE_SUBST_FORMAT1(Name, FirstGlyphs, LigatureSetOffsets) \
+	OT_SUBLOOKUP(Name, 1, \
+		OT_OFFSET(Name, Name##Coverage) \
+		OT_LABEL_END \
+		OT_UARRAY(Name##LigatureSetOffsetsArray, OT_LIST(LigatureSetOffsets)) \
+	) \
+	OT_COVERAGE1(Name##Coverage, OT_LIST(FirstGlyphs)) \
+	/* ASSERT_STATIC_EXPR_ZERO (len(FirstGlyphs) == len(LigatureSetOffsets)) */
+
+#define OT_LIGATURE_SET(Name, LigatureSetOffsets) \
+	OT_UARRAY(Name, OT_LIST(LigatureSetOffsets))
+
+#define OT_LIGATURE(Name, Components, LigGlyph) \
+	OT_LABEL_START(Name) \
+	LigGlyph \
+	OT_LABEL_END \
+	OT_UHEADLESSARRAY(Name##ComponentsArray, OT_LIST(Components))
+
+/*
+ *
+ * Start of Windows-1256 shaping table.
+ *
+ */
+
+/* Table name. */
+#define TABLE_NAME arabic_win1256_gsub_lookups
+
+/* Table manifest. */
+#define MANIFEST(Items) \
+	OT_LABEL_START(manifest) \
+	OT_UINT16(OT_COUNT(manifestData, 6)) \
+	OT_LABEL(manifestData) \
+	Items \
+	OT_LABEL_END
+
+#define MANIFEST_LOOKUP(Tag, Name) \
+	Tag \
+	OT_OFFSET(manifest, Name)
+
+/* Shorthand. */
+#define G	OT_GLYPHID
+
+/*
+ * Table Start
+ */
+OT_TABLE_START
+
+
+/*
+ * Manifest
+ */
+MANIFEST(
+	MANIFEST_LOOKUP(OT_TAG('r','l','i','g'), rligLookup)
+	MANIFEST_LOOKUP(OT_TAG('i','n','i','t'), initLookup)
+	MANIFEST_LOOKUP(OT_TAG('m','e','d','i'), mediLookup)
+	MANIFEST_LOOKUP(OT_TAG('f','i','n','a'), finaLookup)
+	MANIFEST_LOOKUP(OT_TAG('r','l','i','g'), rligMarksLookup)
+)
+
+/*
+ * Lookups
+ */
+OT_LOOKUP(initLookup, OT_LOOKUP_TYPE_SUBST_SINGLE, OT_LOOKUP_FLAG_IGNORE_MARKS,
+	OT_OFFSET(initLookup, initmediSubLookup)
+	OT_OFFSET(initLookup, initSubLookup)
+)
+OT_LOOKUP(mediLookup, OT_LOOKUP_TYPE_SUBST_SINGLE, OT_LOOKUP_FLAG_IGNORE_MARKS,
+	OT_OFFSET(mediLookup, initmediSubLookup)
+	OT_OFFSET(mediLookup, mediSubLookup)
+	OT_OFFSET(mediLookup, medifinaLamAlefSubLookup)
+)
+OT_LOOKUP(finaLookup, OT_LOOKUP_TYPE_SUBST_SINGLE, OT_LOOKUP_FLAG_IGNORE_MARKS,
+	OT_OFFSET(finaLookup, finaSubLookup)
+	/* We don't need this one currently as the sequence inherits masks
+	 * from the first item.  Just in case we change that in the future
+	 * to be smart about Arabic masks when ligating... */
+	OT_OFFSET(finaLookup, medifinaLamAlefSubLookup)
+)
+OT_LOOKUP(rligLookup, OT_LOOKUP_TYPE_SUBST_LIGATURE, OT_LOOKUP_FLAG_IGNORE_MARKS,
+	OT_OFFSET(rligLookup, lamAlefLigaturesSubLookup)
+)
+OT_LOOKUP(rligMarksLookup, OT_LOOKUP_TYPE_SUBST_LIGATURE, 0,
+	OT_OFFSET(rligMarksLookup, shaddaLigaturesSubLookup)
+)
+
+/*
+ * init/medi/fina forms
+ */
+OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(initmediSubLookup,
+	G(198)	G(200)	G(201)	G(202)	G(203)	G(204)	G(205)	G(206)	G(211)
+	G(212)	G(213)	G(214)	G(223)	G(225)	G(227)	G(228)	G(236)	G(237),
+	G(162)	G(4)	G(5)	G(5)	G(6)	G(7)	G(9)	G(11)	G(13)
+	G(14)	G(15)	G(26)	G(140)	G(141)	G(142)	G(143)	G(154)	G(154)
+)
+OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(initSubLookup,
+	G(218)	G(219)	G(221)	G(222)	G(229),
+	G(27)	G(30)	G(128)	G(131)	G(144)
+)
+OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(mediSubLookup,
+	G(218)	G(219)	G(221)	G(222)	G(229),
+	G(28)	G(31)	G(129)	G(138)	G(149)
+)
+OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(finaSubLookup,
+	G(194)	G(195)	G(197)	G(198)	G(199)	G(201)	G(204)	G(205)	G(206)
+	G(218)	G(219)	G(229)	G(236)	G(237),
+	G(2)	G(1)	G(3)	G(181)	G(0)	G(159)	G(8)	G(10)	G(12)
+	G(29)	G(127)	G(152) G(160)	G(156)
+)
+OT_SUBLOOKUP_SINGLE_SUBST_FORMAT2(medifinaLamAlefSubLookup,
+	G(165)	G(178)	G(180)	G(252),
+	G(170)	G(179)	G(185)	G(255)
+)
+
+/*
+ * Lam+Alef ligatures
+ */
+OT_SUBLOOKUP_LIGATURE_SUBST_FORMAT1(lamAlefLigaturesSubLookup,
+	G(225),
+	OT_OFFSET(lamAlefLigaturesSubLookup, lamLigatureSet)
+)
+OT_LIGATURE_SET(lamLigatureSet,
+	OT_OFFSET(lamLigatureSet, lamInitLigature1)
+	OT_OFFSET(lamLigatureSet, lamInitLigature2)
+	OT_OFFSET(lamLigatureSet, lamInitLigature3)
+	OT_OFFSET(lamLigatureSet, lamInitLigature4)
+)
+OT_LIGATURE(lamInitLigature1, G(199), G(165))
+OT_LIGATURE(lamInitLigature2, G(195), G(178))
+OT_LIGATURE(lamInitLigature3, G(194), G(180))
+OT_LIGATURE(lamInitLigature4, G(197), G(252))
+
+/*
+ * Shadda ligatures
+ */
+OT_SUBLOOKUP_LIGATURE_SUBST_FORMAT1(shaddaLigaturesSubLookup,
+	G(248),
+	OT_OFFSET(shaddaLigaturesSubLookup, shaddaLigatureSet)
+)
+OT_LIGATURE_SET(shaddaLigatureSet,
+	OT_OFFSET(shaddaLigatureSet, shaddaLigature1)
+	OT_OFFSET(shaddaLigatureSet, shaddaLigature2)
+	OT_OFFSET(shaddaLigatureSet, shaddaLigature3)
+)
+OT_LIGATURE(shaddaLigature1, G(243), G(172))
+OT_LIGATURE(shaddaLigature2, G(245), G(173))
+OT_LIGATURE(shaddaLigature3, G(246), G(175))
+
+/*
+ * Table end
+ */
+OT_TABLE_END
+
+
+/*
+ * Clean up
+ */
+#undef OT_TABLE_START
+#undef OT_TABLE_END
+#undef OT_LABEL_START
+#undef OT_LABEL_END
+#undef OT_UINT8
+#undef OT_UINT16
+#undef OT_DISTANCE
+#undef OT_COUNT
+
+/*
+ * Include a second time to get the table data...
+ */
+#if 0
+#include "hb.hh" /* Make check-includes.sh happy. */
+#endif
+#ifdef OT_MEASURE
+#include "hb-ot-shape-complex-arabic-win1256.hh"
+#endif
+
+#define HB_OT_SHAPE_COMPLEX_ARABIC_WIN1256_HH
+#endif /* HB_OT_SHAPE_COMPLEX_ARABIC_WIN1256_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.cc
new file mode 100644
index 0000000000..1e93f0efd5
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.cc
@@ -0,0 +1,716 @@
+/*
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-arabic.hh"
+#include "hb-ot-shape.hh"
+
+
+/* buffer var allocations */
+#define arabic_shaping_action() complex_var_u8_0() /* arabic shaping action */
+
+#define HB_BUFFER_SCRATCH_FLAG_ARABIC_HAS_STCH HB_BUFFER_SCRATCH_FLAG_COMPLEX0
+
+/* See:
+ * https://github.com/harfbuzz/harfbuzz/commit/6e6f82b6f3dde0fc6c3c7d991d9ec6cfff57823d#commitcomment-14248516 */
+#define HB_ARABIC_GENERAL_CATEGORY_IS_WORD(gen_cat) \
+	(FLAG_UNSAFE (gen_cat) & \
+	 (FLAG (HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE) | \
+	  /*FLAG (HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER) |*/ \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER) | \
+	  /*FLAG (HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER) |*/ \
+	  /*FLAG (HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER) |*/ \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL)))
+
+
+/*
+ * Joining types:
+ */
+
+/*
+ * Bits used in the joining tables
+ */
+enum hb_arabic_joining_type_t {
+  JOINING_TYPE_U		= 0,
+  JOINING_TYPE_L		= 1,
+  JOINING_TYPE_R		= 2,
+  JOINING_TYPE_D		= 3,
+  JOINING_TYPE_C		= JOINING_TYPE_D,
+  JOINING_GROUP_ALAPH		= 4,
+  JOINING_GROUP_DALATH_RISH	= 5,
+  NUM_STATE_MACHINE_COLS	= 6,
+
+  JOINING_TYPE_T = 7,
+  JOINING_TYPE_X = 8  /* means: use general-category to choose between U or T. */
+};
+
+#include "hb-ot-shape-complex-arabic-table.hh"
+
+static unsigned int get_joining_type (hb_codepoint_t u, hb_unicode_general_category_t gen_cat)
+{
+  unsigned int j_type = joining_type(u);
+  if (likely (j_type != JOINING_TYPE_X))
+    return j_type;
+
+  return (FLAG_UNSAFE(gen_cat) &
+	  (FLAG(HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK) |
+	   FLAG(HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK) |
+	   FLAG(HB_UNICODE_GENERAL_CATEGORY_FORMAT))
+	 ) ?  JOINING_TYPE_T : JOINING_TYPE_U;
+}
+
+#define FEATURE_IS_SYRIAC(tag) hb_in_range<unsigned char> ((unsigned char) (tag), '2', '3')
+
+static const hb_tag_t arabic_features[] =
+{
+  HB_TAG('i','s','o','l'),
+  HB_TAG('f','i','n','a'),
+  HB_TAG('f','i','n','2'),
+  HB_TAG('f','i','n','3'),
+  HB_TAG('m','e','d','i'),
+  HB_TAG('m','e','d','2'),
+  HB_TAG('i','n','i','t'),
+  HB_TAG_NONE
+};
+
+
+/* Same order as the feature array */
+enum arabic_action_t {
+  ISOL,
+  FINA,
+  FIN2,
+  FIN3,
+  MEDI,
+  MED2,
+  INIT,
+
+  NONE,
+
+  ARABIC_NUM_FEATURES = NONE,
+
+  /* We abuse the same byte for other things... */
+  STCH_FIXED,
+  STCH_REPEATING,
+};
+
+static const struct arabic_state_table_entry {
+	uint8_t prev_action;
+	uint8_t curr_action;
+	uint16_t next_state;
+} arabic_state_table[][NUM_STATE_MACHINE_COLS] =
+{
+  /*   jt_U,          jt_L,          jt_R,          jt_D,          jg_ALAPH,      jg_DALATH_RISH */
+
+  /* State 0: prev was U, not willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {NONE,ISOL,1}, {NONE,ISOL,2}, {NONE,ISOL,1}, {NONE,ISOL,6}, },
+
+  /* State 1: prev was R or ISOL/ALAPH, not willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {NONE,ISOL,1}, {NONE,ISOL,2}, {NONE,FIN2,5}, {NONE,ISOL,6}, },
+
+  /* State 2: prev was D/L in ISOL form, willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {INIT,FINA,1}, {INIT,FINA,3}, {INIT,FINA,4}, {INIT,FINA,6}, },
+
+  /* State 3: prev was D in FINA form, willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {MEDI,FINA,1}, {MEDI,FINA,3}, {MEDI,FINA,4}, {MEDI,FINA,6}, },
+
+  /* State 4: prev was FINA ALAPH, not willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {MED2,ISOL,1}, {MED2,ISOL,2}, {MED2,FIN2,5}, {MED2,ISOL,6}, },
+
+  /* State 5: prev was FIN2/FIN3 ALAPH, not willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {ISOL,ISOL,1}, {ISOL,ISOL,2}, {ISOL,FIN2,5}, {ISOL,ISOL,6}, },
+
+  /* State 6: prev was DALATH/RISH, not willing to join. */
+  { {NONE,NONE,0}, {NONE,ISOL,2}, {NONE,ISOL,1}, {NONE,ISOL,2}, {NONE,FIN3,5}, {NONE,ISOL,6}, }
+};
+
+
+static void
+arabic_fallback_shape (const hb_ot_shape_plan_t *plan,
+		       hb_font_t *font,
+		       hb_buffer_t *buffer);
+
+static void
+record_stch (const hb_ot_shape_plan_t *plan,
+	     hb_font_t *font,
+	     hb_buffer_t *buffer);
+
+static void
+collect_features_arabic (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* We apply features according to the Arabic spec, with pauses
+   * in between most.
+   *
+   * The pause between init/medi/... and rlig is required.  See eg:
+   * https://bugzilla.mozilla.org/show_bug.cgi?id=644184
+   *
+   * The pauses between init/medi/... themselves are not necessarily
+   * needed as only one of those features is applied to any character.
+   * The only difference it makes is when fonts have contextual
+   * substitutions.  We now follow the order of the spec, which makes
+   * for better experience if that's what Uniscribe is doing.
+   *
+   * At least for Arabic, looks like Uniscribe has a pause between
+   * rlig and calt.  Otherwise the IranNastaliq's ALLAH ligature won't
+   * work.  However, testing shows that rlig and calt are applied
+   * together for Mongolian in Uniscribe.  As such, we only add a
+   * pause for Arabic, not other scripts.
+   *
+   * A pause after calt is required to make KFGQPC Uthmanic Script HAFS
+   * work correctly.  See https://github.com/harfbuzz/harfbuzz/issues/505
+   */
+
+
+  map->enable_feature (HB_TAG('s','t','c','h'));
+  map->add_gsub_pause (record_stch);
+
+  map->enable_feature (HB_TAG('c','c','m','p'));
+  map->enable_feature (HB_TAG('l','o','c','l'));
+
+  map->add_gsub_pause (nullptr);
+
+  for (unsigned int i = 0; i < ARABIC_NUM_FEATURES; i++)
+  {
+    bool has_fallback = plan->props.script == HB_SCRIPT_ARABIC && !FEATURE_IS_SYRIAC (arabic_features[i]);
+    map->add_feature (arabic_features[i], has_fallback ? F_HAS_FALLBACK : F_NONE);
+    map->add_gsub_pause (nullptr);
+  }
+
+  /* Normally, Unicode says a ZWNJ means "don't ligate".  In Arabic script
+   * however, it says a ZWJ should also mean "don't ligate".  So we run
+   * the main ligating features as MANUAL_ZWJ. */
+
+  map->enable_feature (HB_TAG('r','l','i','g'), F_MANUAL_ZWJ | F_HAS_FALLBACK);
+
+  if (plan->props.script == HB_SCRIPT_ARABIC)
+    map->add_gsub_pause (arabic_fallback_shape);
+
+  /* No pause after rclt.  See 98460779bae19e4d64d29461ff154b3527bf8420. */
+  map->enable_feature (HB_TAG('r','c','l','t'), F_MANUAL_ZWJ);
+  map->enable_feature (HB_TAG('c','a','l','t'), F_MANUAL_ZWJ);
+  map->add_gsub_pause (nullptr);
+
+  /* The spec includes 'cswh'.  Earlier versions of Windows
+   * used to enable this by default, but testing suggests
+   * that Windows 8 and later do not enable it by default,
+   * and spec now says 'Off by default'.
+   * We disabled this in ae23c24c32.
+   * Note that IranNastaliq uses this feature extensively
+   * to fixup broken glyph sequences.  Oh well...
+   * Test case: U+0643,U+0640,U+0631. */
+  //map->enable_feature (HB_TAG('c','s','w','h'));
+  map->enable_feature (HB_TAG('m','s','e','t'));
+}
+
+#include "hb-ot-shape-complex-arabic-fallback.hh"
+
+struct arabic_shape_plan_t
+{
+  /* The "+ 1" in the next array is to accommodate for the "NONE" command,
+   * which is not an OpenType feature, but this simplifies the code by not
+   * having to do a "if (... < NONE) ..." and just rely on the fact that
+   * mask_array[NONE] == 0. */
+  hb_mask_t mask_array[ARABIC_NUM_FEATURES + 1];
+
+  hb_atomic_ptr_t<arabic_fallback_plan_t> fallback_plan;
+
+  unsigned int do_fallback : 1;
+  unsigned int has_stch : 1;
+};
+
+void *
+data_create_arabic (const hb_ot_shape_plan_t *plan)
+{
+  arabic_shape_plan_t *arabic_plan = (arabic_shape_plan_t *) calloc (1, sizeof (arabic_shape_plan_t));
+  if (unlikely (!arabic_plan))
+    return nullptr;
+
+  arabic_plan->do_fallback = plan->props.script == HB_SCRIPT_ARABIC;
+  arabic_plan->has_stch = !!plan->map.get_1_mask (HB_TAG ('s','t','c','h'));
+  for (unsigned int i = 0; i < ARABIC_NUM_FEATURES; i++) {
+    arabic_plan->mask_array[i] = plan->map.get_1_mask (arabic_features[i]);
+    arabic_plan->do_fallback = arabic_plan->do_fallback &&
+			       (FEATURE_IS_SYRIAC (arabic_features[i]) ||
+				plan->map.needs_fallback (arabic_features[i]));
+  }
+
+  return arabic_plan;
+}
+
+void
+data_destroy_arabic (void *data)
+{
+  arabic_shape_plan_t *arabic_plan = (arabic_shape_plan_t *) data;
+
+  arabic_fallback_plan_destroy (arabic_plan->fallback_plan);
+
+  free (data);
+}
+
+static void
+arabic_joining (hb_buffer_t *buffer)
+{
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int prev = UINT_MAX, state = 0;
+
+  /* Check pre-context */
+  for (unsigned int i = 0; i < buffer->context_len[0]; i++)
+  {
+    unsigned int this_type = get_joining_type (buffer->context[0][i], buffer->unicode->general_category (buffer->context[0][i]));
+
+    if (unlikely (this_type == JOINING_TYPE_T))
+      continue;
+
+    const arabic_state_table_entry *entry = &arabic_state_table[state][this_type];
+    state = entry->next_state;
+    break;
+  }
+
+  for (unsigned int i = 0; i < count; i++)
+  {
+    unsigned int this_type = get_joining_type (info[i].codepoint, _hb_glyph_info_get_general_category (&info[i]));
+
+    if (unlikely (this_type == JOINING_TYPE_T)) {
+      info[i].arabic_shaping_action() = NONE;
+      continue;
+    }
+
+    const arabic_state_table_entry *entry = &arabic_state_table[state][this_type];
+
+    if (entry->prev_action != NONE && prev != UINT_MAX)
+    {
+      info[prev].arabic_shaping_action() = entry->prev_action;
+      buffer->unsafe_to_break (prev, i + 1);
+    }
+
+    info[i].arabic_shaping_action() = entry->curr_action;
+
+    prev = i;
+    state = entry->next_state;
+  }
+
+  for (unsigned int i = 0; i < buffer->context_len[1]; i++)
+  {
+    unsigned int this_type = get_joining_type (buffer->context[1][i], buffer->unicode->general_category (buffer->context[1][i]));
+
+    if (unlikely (this_type == JOINING_TYPE_T))
+      continue;
+
+    const arabic_state_table_entry *entry = &arabic_state_table[state][this_type];
+    if (entry->prev_action != NONE && prev != UINT_MAX)
+      info[prev].arabic_shaping_action() = entry->prev_action;
+    break;
+  }
+}
+
+static void
+mongolian_variation_selectors (hb_buffer_t *buffer)
+{
+  /* Copy arabic_shaping_action() from base to Mongolian variation selectors. */
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 1; i < count; i++)
+    if (unlikely (hb_in_range<hb_codepoint_t> (info[i].codepoint, 0x180Bu, 0x180Du)))
+      info[i].arabic_shaping_action() = info[i - 1].arabic_shaping_action();
+}
+
+void
+setup_masks_arabic_plan (const arabic_shape_plan_t *arabic_plan,
+			 hb_buffer_t               *buffer,
+			 hb_script_t                script)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, arabic_shaping_action);
+
+  arabic_joining (buffer);
+  if (script == HB_SCRIPT_MONGOLIAN)
+    mongolian_variation_selectors (buffer);
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    info[i].mask |= arabic_plan->mask_array[info[i].arabic_shaping_action()];
+}
+
+static void
+setup_masks_arabic (const hb_ot_shape_plan_t *plan,
+		    hb_buffer_t              *buffer,
+		    hb_font_t                *font HB_UNUSED)
+{
+  const arabic_shape_plan_t *arabic_plan = (const arabic_shape_plan_t *) plan->data;
+  setup_masks_arabic_plan (arabic_plan, buffer, plan->props.script);
+}
+
+static void
+arabic_fallback_shape (const hb_ot_shape_plan_t *plan,
+		       hb_font_t *font,
+		       hb_buffer_t *buffer)
+{
+#ifdef HB_NO_OT_SHAPE_COMPLEX_ARABIC_FALLBACK
+  return;
+#endif
+
+  const arabic_shape_plan_t *arabic_plan = (const arabic_shape_plan_t *) plan->data;
+
+  if (!arabic_plan->do_fallback)
+    return;
+
+retry:
+  arabic_fallback_plan_t *fallback_plan = arabic_plan->fallback_plan;
+  if (unlikely (!fallback_plan))
+  {
+    /* This sucks.  We need a font to build the fallback plan... */
+    fallback_plan = arabic_fallback_plan_create (plan, font);
+    if (unlikely (!arabic_plan->fallback_plan.cmpexch (nullptr, fallback_plan)))
+    {
+      arabic_fallback_plan_destroy (fallback_plan);
+      goto retry;
+    }
+  }
+
+  arabic_fallback_plan_shape (fallback_plan, font, buffer);
+}
+
+/*
+ * Stretch feature: "stch".
+ * See example here:
+ * https://docs.microsoft.com/en-us/typography/script-development/syriac
+ * We implement this in a generic way, such that the Arabic subtending
+ * marks can use it as well.
+ */
+
+static void
+record_stch (const hb_ot_shape_plan_t *plan,
+	     hb_font_t *font HB_UNUSED,
+	     hb_buffer_t *buffer)
+{
+  const arabic_shape_plan_t *arabic_plan = (const arabic_shape_plan_t *) plan->data;
+  if (!arabic_plan->has_stch)
+    return;
+
+  /* 'stch' feature was just applied.  Look for anything that multiplied,
+   * and record it for stch treatment later.  Note that rtlm, frac, etc
+   * are applied before stch, but we assume that they didn't result in
+   * anything multiplying into 5 pieces, so it's safe-ish... */
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if (unlikely (_hb_glyph_info_multiplied (&info[i])))
+    {
+      unsigned int comp = _hb_glyph_info_get_lig_comp (&info[i]);
+      info[i].arabic_shaping_action() = comp % 2 ? STCH_REPEATING : STCH_FIXED;
+      buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_ARABIC_HAS_STCH;
+    }
+}
+
+static void
+apply_stch (const hb_ot_shape_plan_t *plan HB_UNUSED,
+	    hb_buffer_t              *buffer,
+	    hb_font_t                *font)
+{
+  if (likely (!(buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_ARABIC_HAS_STCH)))
+    return;
+
+  /* The Arabic shaper currently always processes in RTL mode, so we should
+   * stretch / position the stretched pieces to the left / preceding glyphs. */
+
+  /* We do a two pass implementation:
+   * First pass calculates the exact number of extra glyphs we need,
+   * We then enlarge buffer to have that much room,
+   * Second pass applies the stretch, copying things to the end of buffer.
+   */
+
+  int sign = font->x_scale < 0 ? -1 : +1;
+  unsigned int extra_glyphs_needed = 0; // Set during MEASURE, used during CUT
+  enum { MEASURE, CUT } /* step_t */;
+
+  for (unsigned int step = MEASURE; step <= CUT; step = step + 1)
+  {
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    hb_glyph_position_t *pos = buffer->pos;
+    unsigned int new_len = count + extra_glyphs_needed; // write head during CUT
+    unsigned int j = new_len;
+    for (unsigned int i = count; i; i--)
+    {
+      if (!hb_in_range<uint8_t> (info[i - 1].arabic_shaping_action(), STCH_FIXED, STCH_REPEATING))
+      {
+	if (step == CUT)
+	{
+	  --j;
+	  info[j] = info[i - 1];
+	  pos[j] = pos[i - 1];
+	}
+	continue;
+      }
+
+      /* Yay, justification! */
+
+      hb_position_t w_total = 0; // Total to be filled
+      hb_position_t w_fixed = 0; // Sum of fixed tiles
+      hb_position_t w_repeating = 0; // Sum of repeating tiles
+      int n_fixed = 0;
+      int n_repeating = 0;
+
+      unsigned int end = i;
+      while (i &&
+	     hb_in_range<uint8_t> (info[i - 1].arabic_shaping_action(), STCH_FIXED, STCH_REPEATING))
+      {
+	i--;
+	hb_position_t width = font->get_glyph_h_advance (info[i].codepoint);
+	if (info[i].arabic_shaping_action() == STCH_FIXED)
+	{
+	  w_fixed += width;
+	  n_fixed++;
+	}
+	else
+	{
+	  w_repeating += width;
+	  n_repeating++;
+	}
+      }
+      unsigned int start = i;
+      unsigned int context = i;
+      while (context &&
+	     !hb_in_range<uint8_t> (info[context - 1].arabic_shaping_action(), STCH_FIXED, STCH_REPEATING) &&
+	     (_hb_glyph_info_is_default_ignorable (&info[context - 1]) ||
+	      HB_ARABIC_GENERAL_CATEGORY_IS_WORD (_hb_glyph_info_get_general_category (&info[context - 1]))))
+      {
+	context--;
+	w_total += pos[context].x_advance;
+      }
+      i++; // Don't touch i again.
+
+      DEBUG_MSG (ARABIC, nullptr, "%s stretch at (%d,%d,%d)",
+		 step == MEASURE ? "measuring" : "cutting", context, start, end);
+      DEBUG_MSG (ARABIC, nullptr, "rest of word:    count=%d width %d", start - context, w_total);
+      DEBUG_MSG (ARABIC, nullptr, "fixed tiles:     count=%d width=%d", n_fixed, w_fixed);
+      DEBUG_MSG (ARABIC, nullptr, "repeating tiles: count=%d width=%d", n_repeating, w_repeating);
+
+      /* Number of additional times to repeat each repeating tile. */
+      int n_copies = 0;
+
+      hb_position_t w_remaining = w_total - w_fixed;
+      if (sign * w_remaining > sign * w_repeating && sign * w_repeating > 0)
+	n_copies = (sign * w_remaining) / (sign * w_repeating) - 1;
+
+      /* See if we can improve the fit by adding an extra repeat and squeezing them together a bit. */
+      hb_position_t extra_repeat_overlap = 0;
+      hb_position_t shortfall = sign * w_remaining - sign * w_repeating * (n_copies + 1);
+      if (shortfall > 0 && n_repeating > 0)
+      {
+	++n_copies;
+	hb_position_t excess = (n_copies + 1) * sign * w_repeating - sign * w_remaining;
+	if (excess > 0)
+	  extra_repeat_overlap = excess / (n_copies * n_repeating);
+      }
+
+      if (step == MEASURE)
+      {
+	extra_glyphs_needed += n_copies * n_repeating;
+	DEBUG_MSG (ARABIC, nullptr, "will add extra %d copies of repeating tiles", n_copies);
+      }
+      else
+      {
+	buffer->unsafe_to_break (context, end);
+	hb_position_t x_offset = 0;
+	for (unsigned int k = end; k > start; k--)
+	{
+	  hb_position_t width = font->get_glyph_h_advance (info[k - 1].codepoint);
+
+	  unsigned int repeat = 1;
+	  if (info[k - 1].arabic_shaping_action() == STCH_REPEATING)
+	    repeat += n_copies;
+
+	  DEBUG_MSG (ARABIC, nullptr, "appending %d copies of glyph %d; j=%d",
+		     repeat, info[k - 1].codepoint, j);
+	  for (unsigned int n = 0; n < repeat; n++)
+	  {
+	    x_offset -= width;
+	    if (n > 0)
+	      x_offset += extra_repeat_overlap;
+	    pos[k - 1].x_offset = x_offset;
+	    /* Append copy. */
+	    --j;
+	    info[j] = info[k - 1];
+	    pos[j] = pos[k - 1];
+	  }
+	}
+      }
+    }
+
+    if (step == MEASURE)
+    {
+      if (unlikely (!buffer->ensure (count + extra_glyphs_needed)))
+	break;
+    }
+    else
+    {
+      assert (j == 0);
+      buffer->len = new_len;
+    }
+  }
+}
+
+
+static void
+postprocess_glyphs_arabic (const hb_ot_shape_plan_t *plan,
+			   hb_buffer_t              *buffer,
+			   hb_font_t                *font)
+{
+  apply_stch (plan, buffer, font);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, arabic_shaping_action);
+}
+
+/* https://www.unicode.org/reports/tr53/ */
+
+static hb_codepoint_t
+modifier_combining_marks[] =
+{
+  0x0654u, /* ARABIC HAMZA ABOVE */
+  0x0655u, /* ARABIC HAMZA BELOW */
+  0x0658u, /* ARABIC MARK NOON GHUNNA */
+  0x06DCu, /* ARABIC SMALL HIGH SEEN */
+  0x06E3u, /* ARABIC SMALL LOW SEEN */
+  0x06E7u, /* ARABIC SMALL HIGH YEH */
+  0x06E8u, /* ARABIC SMALL HIGH NOON */
+  0x08D3u, /* ARABIC SMALL LOW WAW */
+  0x08F3u, /* ARABIC SMALL HIGH WAW */
+};
+
+static inline bool
+info_is_mcm (const hb_glyph_info_t &info)
+{
+  hb_codepoint_t u = info.codepoint;
+  for (unsigned int i = 0; i < ARRAY_LENGTH (modifier_combining_marks); i++)
+    if (u == modifier_combining_marks[i])
+      return true;
+  return false;
+}
+
+static void
+reorder_marks_arabic (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		      hb_buffer_t              *buffer,
+		      unsigned int              start,
+		      unsigned int              end)
+{
+  hb_glyph_info_t *info = buffer->info;
+
+  DEBUG_MSG (ARABIC, buffer, "Reordering marks from %d to %d", start, end);
+
+  unsigned int i = start;
+  for (unsigned int cc = 220; cc <= 230; cc += 10)
+  {
+    DEBUG_MSG (ARABIC, buffer, "Looking for %d's starting at %d", cc, i);
+    while (i < end && info_cc(info[i]) < cc)
+      i++;
+    DEBUG_MSG (ARABIC, buffer, "Looking for %d's stopped at %d", cc, i);
+
+    if (i == end)
+      break;
+
+    if (info_cc(info[i]) > cc)
+      continue;
+
+    unsigned int j = i;
+    while (j < end && info_cc(info[j]) == cc && info_is_mcm (info[j]))
+      j++;
+
+    if (i == j)
+      continue;
+
+    DEBUG_MSG (ARABIC, buffer, "Found %d's from %d to %d", cc, i, j);
+
+    /* Shift it! */
+    DEBUG_MSG (ARABIC, buffer, "Shifting %d's: %d %d", cc, i, j);
+    hb_glyph_info_t temp[HB_OT_SHAPE_COMPLEX_MAX_COMBINING_MARKS];
+    assert (j - i <= ARRAY_LENGTH (temp));
+    buffer->merge_clusters (start, j);
+    memmove (temp, &info[i], (j - i) * sizeof (hb_glyph_info_t));
+    memmove (&info[start + j - i], &info[start], (i - start) * sizeof (hb_glyph_info_t));
+    memmove (&info[start], temp, (j - i) * sizeof (hb_glyph_info_t));
+
+    /* Renumber CC such that the reordered sequence is still sorted.
+     * 22 and 26 are chosen because they are smaller than all Arabic categories,
+     * and are folded back to 220/230 respectively during fallback mark positioning.
+     *
+     * We do this because the CGJ-handling logic in the normalizer relies on
+     * mark sequences having an increasing order even after this reordering.
+     * https://github.com/harfbuzz/harfbuzz/issues/554
+     * This, however, does break some obscure sequences, where the normalizer
+     * might compose a sequence that it should not.  For example, in the seequence
+     * ALEF, HAMZAH, MADDAH, we should NOT try to compose ALEF+MADDAH, but with this
+     * renumbering, we will.
+     */
+    unsigned int new_start = start + j - i;
+    unsigned int new_cc = cc == 220 ? HB_MODIFIED_COMBINING_CLASS_CCC22 : HB_MODIFIED_COMBINING_CLASS_CCC26;
+    while (start < new_start)
+    {
+      _hb_glyph_info_set_modified_combining_class (&info[start], new_cc);
+      start++;
+    }
+
+    i = j;
+  }
+}
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_arabic =
+{
+  collect_features_arabic,
+  nullptr, /* override_features */
+  data_create_arabic,
+  data_destroy_arabic,
+  nullptr, /* preprocess_text */
+  postprocess_glyphs_arabic,
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  setup_masks_arabic,
+  HB_TAG_NONE, /* gpos_tag */
+  reorder_marks_arabic,
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
+  true, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.hh
new file mode 100644
index 0000000000..5bf6ff6338
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-arabic.hh
@@ -0,0 +1,50 @@
+/*
+ * Copyright © 2015  Mozilla Foundation.
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_ARABIC_HH
+#define HB_OT_SHAPE_COMPLEX_ARABIC_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape-complex.hh"
+
+
+struct arabic_shape_plan_t;
+
+HB_INTERNAL void *
+data_create_arabic (const hb_ot_shape_plan_t *plan);
+
+HB_INTERNAL void
+data_destroy_arabic (void *data);
+
+HB_INTERNAL void
+setup_masks_arabic_plan (const arabic_shape_plan_t *arabic_plan,
+			 hb_buffer_t               *buffer,
+			 hb_script_t                script);
+
+#endif /* HB_OT_SHAPE_COMPLEX_ARABIC_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-default.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-default.cc
new file mode 100644
index 0000000000..a755aea098
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-default.cc
@@ -0,0 +1,73 @@
+/*
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex.hh"
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_default =
+{
+  nullptr, /* collect_features */
+  nullptr, /* override_features */
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  nullptr, /* setup_masks */
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
+  true, /* fallback_position */
+};
+
+/* Same as default but no mark advance zeroing / fallback positioning.
+ * Dumbest shaper ever, basically. */
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_dumber =
+{
+  nullptr, /* collect_features */
+  nullptr, /* override_features */
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  nullptr, /* setup_masks */
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-hangul.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-hangul.cc
new file mode 100644
index 0000000000..f5915f43ae
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-hangul.cc
@@ -0,0 +1,439 @@
+/*
+ * Copyright © 2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex.hh"
+
+
+/* Hangul shaper */
+
+
+/* Same order as the feature array below */
+enum {
+  _JMO,
+
+  LJMO,
+  VJMO,
+  TJMO,
+
+  FIRST_HANGUL_FEATURE = LJMO,
+  HANGUL_FEATURE_COUNT = TJMO + 1
+};
+
+static const hb_tag_t hangul_features[HANGUL_FEATURE_COUNT] =
+{
+  HB_TAG_NONE,
+  HB_TAG('l','j','m','o'),
+  HB_TAG('v','j','m','o'),
+  HB_TAG('t','j','m','o')
+};
+
+static void
+collect_features_hangul (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  for (unsigned int i = FIRST_HANGUL_FEATURE; i < HANGUL_FEATURE_COUNT; i++)
+    map->add_feature (hangul_features[i]);
+}
+
+static void
+override_features_hangul (hb_ot_shape_planner_t *plan)
+{
+  /* Uniscribe does not apply 'calt' for Hangul, and certain fonts
+   * (Noto Sans CJK, Source Sans Han, etc) apply all of jamo lookups
+   * in calt, which is not desirable. */
+  plan->map.disable_feature (HB_TAG('c','a','l','t'));
+}
+
+struct hangul_shape_plan_t
+{
+  hb_mask_t mask_array[HANGUL_FEATURE_COUNT];
+};
+
+static void *
+data_create_hangul (const hb_ot_shape_plan_t *plan)
+{
+  hangul_shape_plan_t *hangul_plan = (hangul_shape_plan_t *) calloc (1, sizeof (hangul_shape_plan_t));
+  if (unlikely (!hangul_plan))
+    return nullptr;
+
+  for (unsigned int i = 0; i < HANGUL_FEATURE_COUNT; i++)
+    hangul_plan->mask_array[i] = plan->map.get_1_mask (hangul_features[i]);
+
+  return hangul_plan;
+}
+
+static void
+data_destroy_hangul (void *data)
+{
+  free (data);
+}
+
+/* Constants for algorithmic hangul syllable [de]composition. */
+#define LBase 0x1100u
+#define VBase 0x1161u
+#define TBase 0x11A7u
+#define LCount 19u
+#define VCount 21u
+#define TCount 28u
+#define SBase 0xAC00u
+#define NCount (VCount * TCount)
+#define SCount (LCount * NCount)
+
+#define isCombiningL(u) (hb_in_range<hb_codepoint_t> ((u), LBase, LBase+LCount-1))
+#define isCombiningV(u) (hb_in_range<hb_codepoint_t> ((u), VBase, VBase+VCount-1))
+#define isCombiningT(u) (hb_in_range<hb_codepoint_t> ((u), TBase+1, TBase+TCount-1))
+#define isCombinedS(u) (hb_in_range<hb_codepoint_t> ((u), SBase, SBase+SCount-1))
+
+#define isL(u) (hb_in_ranges<hb_codepoint_t> ((u), 0x1100u, 0x115Fu, 0xA960u, 0xA97Cu))
+#define isV(u) (hb_in_ranges<hb_codepoint_t> ((u), 0x1160u, 0x11A7u, 0xD7B0u, 0xD7C6u))
+#define isT(u) (hb_in_ranges<hb_codepoint_t> ((u), 0x11A8u, 0x11FFu, 0xD7CBu, 0xD7FBu))
+
+#define isHangulTone(u) (hb_in_range<hb_codepoint_t> ((u), 0x302Eu, 0x302Fu))
+
+/* buffer var allocations */
+#define hangul_shaping_feature() complex_var_u8_0() /* hangul jamo shaping feature */
+
+static bool
+is_zero_width_char (hb_font_t *font,
+		    hb_codepoint_t unicode)
+{
+  hb_codepoint_t glyph;
+  return hb_font_get_glyph (font, unicode, 0, &glyph) && hb_font_get_glyph_h_advance (font, glyph) == 0;
+}
+
+static void
+preprocess_text_hangul (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			hb_buffer_t              *buffer,
+			hb_font_t                *font)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, hangul_shaping_feature);
+
+  /* Hangul syllables come in two shapes: LV, and LVT.  Of those:
+   *
+   *   - LV can be precomposed, or decomposed.  Lets call those
+   *     <LV> and <L,V>,
+   *   - LVT can be fully precomposed, partically precomposed, or
+   *     fully decomposed.  Ie. <LVT>, <LV,T>, or <L,V,T>.
+   *
+   * The composition / decomposition is mechanical.  However, not
+   * all <L,V> sequences compose, and not all <LV,T> sequences
+   * compose.
+   *
+   * Here are the specifics:
+   *
+   *   - <L>: U+1100..115F, U+A960..A97F
+   *   - <V>: U+1160..11A7, U+D7B0..D7C7
+   *   - <T>: U+11A8..11FF, U+D7CB..D7FB
+   *
+   *   - Only the <L,V> sequences for some of the U+11xx ranges combine.
+   *   - Only <LV,T> sequences for some of the Ts in U+11xx range combine.
+   *
+   * Here is what we want to accomplish in this shaper:
+   *
+   *   - If the whole syllable can be precomposed, do that,
+   *   - Otherwise, fully decompose and apply ljmo/vjmo/tjmo features.
+   *   - If a valid syllable is followed by a Hangul tone mark, reorder the tone
+   *     mark to precede the whole syllable - unless it is a zero-width glyph, in
+   *     which case we leave it untouched, assuming it's designed to overstrike.
+   *
+   * That is, of the different possible syllables:
+   *
+   *   <L>
+   *   <L,V>
+   *   <L,V,T>
+   *   <LV>
+   *   <LVT>
+   *   <LV, T>
+   *
+   * - <L> needs no work.
+   *
+   * - <LV> and <LVT> can stay the way they are if the font supports them, otherwise we
+   *   should fully decompose them if font supports.
+   *
+   * - <L,V> and <L,V,T> we should compose if the whole thing can be composed.
+   *
+   * - <LV,T> we should compose if the whole thing can be composed, otherwise we should
+   *   decompose.
+   */
+
+  buffer->clear_output ();
+  unsigned int start = 0, end = 0; /* Extent of most recently seen syllable;
+				    * valid only if start < end
+				    */
+  unsigned int count = buffer->len;
+
+  for (buffer->idx = 0; buffer->idx < count && buffer->successful;)
+  {
+    hb_codepoint_t u = buffer->cur().codepoint;
+
+    if (isHangulTone (u))
+    {
+      /*
+       * We could cache the width of the tone marks and the existence of dotted-circle,
+       * but the use of the Hangul tone mark characters seems to be rare enough that
+       * I didn't bother for now.
+       */
+      if (start < end && end == buffer->out_len)
+      {
+	/* Tone mark follows a valid syllable; move it in front, unless it's zero width. */
+	buffer->unsafe_to_break_from_outbuffer (start, buffer->idx);
+	buffer->next_glyph ();
+	if (!is_zero_width_char (font, u))
+	{
+	  buffer->merge_out_clusters (start, end + 1);
+	  hb_glyph_info_t *info = buffer->out_info;
+	  hb_glyph_info_t tone = info[end];
+	  memmove (&info[start + 1], &info[start], (end - start) * sizeof (hb_glyph_info_t));
+	  info[start] = tone;
+	}
+      }
+      else
+      {
+	/* No valid syllable as base for tone mark; try to insert dotted circle. */
+      if (!(buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE) &&
+	  font->has_glyph (0x25CCu))
+	{
+	  hb_codepoint_t chars[2];
+	  if (!is_zero_width_char (font, u)) {
+	    chars[0] = u;
+	    chars[1] = 0x25CCu;
+	  } else {
+	    chars[0] = 0x25CCu;
+	    chars[1] = u;
+	  }
+	  buffer->replace_glyphs (1, 2, chars);
+	}
+	else
+	{
+	  /* No dotted circle available in the font; just leave tone mark untouched. */
+	  buffer->next_glyph ();
+	}
+      }
+      start = end = buffer->out_len;
+      continue;
+    }
+
+    start = buffer->out_len; /* Remember current position as a potential syllable start;
+			      * will only be used if we set end to a later position.
+			      */
+
+    if (isL (u) && buffer->idx + 1 < count)
+    {
+      hb_codepoint_t l = u;
+      hb_codepoint_t v = buffer->cur(+1).codepoint;
+      if (isV (v))
+      {
+	/* Have <L,V> or <L,V,T>. */
+	hb_codepoint_t t = 0;
+	unsigned int tindex = 0;
+	if (buffer->idx + 2 < count)
+	{
+	  t = buffer->cur(+2).codepoint;
+	  if (isT (t))
+	    tindex = t - TBase; /* Only used if isCombiningT (t); otherwise invalid. */
+	  else
+	    t = 0; /* The next character was not a trailing jamo. */
+	}
+	buffer->unsafe_to_break (buffer->idx, buffer->idx + (t ? 3 : 2));
+
+	/* We've got a syllable <L,V,T?>; see if it can potentially be composed. */
+	if (isCombiningL (l) && isCombiningV (v) && (t == 0 || isCombiningT (t)))
+	{
+	  /* Try to compose; if this succeeds, end is set to start+1. */
+	  hb_codepoint_t s = SBase + (l - LBase) * NCount + (v - VBase) * TCount + tindex;
+	  if (font->has_glyph (s))
+	  {
+	    buffer->replace_glyphs (t ? 3 : 2, 1, &s);
+	    if (unlikely (!buffer->successful))
+	      return;
+	    end = start + 1;
+	    continue;
+	  }
+	}
+
+	/* We didn't compose, either because it's an Old Hangul syllable without a
+	 * precomposed character in Unicode, or because the font didn't support the
+	 * necessary precomposed glyph.
+	 * Set jamo features on the individual glyphs, and advance past them.
+	 */
+	buffer->cur().hangul_shaping_feature() = LJMO;
+	buffer->next_glyph ();
+	buffer->cur().hangul_shaping_feature() = VJMO;
+	buffer->next_glyph ();
+	if (t)
+	{
+	  buffer->cur().hangul_shaping_feature() = TJMO;
+	  buffer->next_glyph ();
+	  end = start + 3;
+	}
+	else
+	  end = start + 2;
+	if (buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
+	  buffer->merge_out_clusters (start, end);
+	continue;
+      }
+    }
+
+    else if (isCombinedS (u))
+    {
+      /* Have <LV>, <LVT>, or <LV,T> */
+      hb_codepoint_t s = u;
+      bool has_glyph = font->has_glyph (s);
+      unsigned int lindex = (s - SBase) / NCount;
+      unsigned int nindex = (s - SBase) % NCount;
+      unsigned int vindex = nindex / TCount;
+      unsigned int tindex = nindex % TCount;
+
+      if (!tindex &&
+	  buffer->idx + 1 < count &&
+	  isCombiningT (buffer->cur(+1).codepoint))
+      {
+	/* <LV,T>, try to combine. */
+	unsigned int new_tindex = buffer->cur(+1).codepoint - TBase;
+	hb_codepoint_t new_s = s + new_tindex;
+	if (font->has_glyph (new_s))
+	{
+	  buffer->replace_glyphs (2, 1, &new_s);
+	  if (unlikely (!buffer->successful))
+	    return;
+	  end = start + 1;
+	  continue;
+	}
+	else
+	  buffer->unsafe_to_break (buffer->idx, buffer->idx + 2); /* Mark unsafe between LV and T. */
+      }
+
+      /* Otherwise, decompose if font doesn't support <LV> or <LVT>,
+       * or if having non-combining <LV,T>.  Note that we already handled
+       * combining <LV,T> above. */
+      if (!has_glyph ||
+	  (!tindex &&
+	   buffer->idx + 1 < count &&
+	   isT (buffer->cur(+1).codepoint)))
+      {
+	hb_codepoint_t decomposed[3] = {LBase + lindex,
+					VBase + vindex,
+					TBase + tindex};
+	if (font->has_glyph (decomposed[0]) &&
+	    font->has_glyph (decomposed[1]) &&
+	    (!tindex || font->has_glyph (decomposed[2])))
+	{
+	  unsigned int s_len = tindex ? 3 : 2;
+	  buffer->replace_glyphs (1, s_len, decomposed);
+
+	  /* If we decomposed an LV because of a non-combining T following,
+	   * we want to include this T in the syllable.
+	   */
+	  if (has_glyph && !tindex)
+	  {
+	    buffer->next_glyph ();
+	    s_len++;
+	  }
+
+	  if (unlikely (!buffer->successful))
+	    return;
+
+	  /* We decomposed S: apply jamo features to the individual glyphs
+	   * that are now in buffer->out_info.
+	   */
+	  hb_glyph_info_t *info = buffer->out_info;
+	  end = start + s_len;
+
+	  unsigned int i = start;
+	  info[i++].hangul_shaping_feature() = LJMO;
+	  info[i++].hangul_shaping_feature() = VJMO;
+	  if (i < end)
+	    info[i++].hangul_shaping_feature() = TJMO;
+
+	  if (buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
+	    buffer->merge_out_clusters (start, end);
+	  continue;
+	}
+	else if ((!tindex && buffer->idx + 1 < count && isT (buffer->cur(+1).codepoint)))
+	  buffer->unsafe_to_break (buffer->idx, buffer->idx + 2); /* Mark unsafe between LV and T. */
+      }
+
+      if (has_glyph)
+      {
+	/* We didn't decompose the S, so just advance past it. */
+	end = start + 1;
+	buffer->next_glyph ();
+	continue;
+      }
+    }
+
+    /* Didn't find a recognizable syllable, so we leave end <= start;
+     * this will prevent tone-mark reordering happening.
+     */
+    buffer->next_glyph ();
+  }
+  buffer->swap_buffers ();
+}
+
+static void
+setup_masks_hangul (const hb_ot_shape_plan_t *plan,
+		    hb_buffer_t              *buffer,
+		    hb_font_t                *font HB_UNUSED)
+{
+  const hangul_shape_plan_t *hangul_plan = (const hangul_shape_plan_t *) plan->data;
+
+  if (likely (hangul_plan))
+  {
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    for (unsigned int i = 0; i < count; i++, info++)
+      info->mask |= hangul_plan->mask_array[info->hangul_shaping_feature()];
+  }
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, hangul_shaping_feature);
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_hangul =
+{
+  collect_features_hangul,
+  override_features_hangul,
+  data_create_hangul,
+  data_destroy_hangul,
+  preprocess_text_hangul,
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_NONE,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  setup_masks_hangul,
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-hebrew.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-hebrew.cc
new file mode 100644
index 0000000000..334d3ded82
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-hebrew.cc
@@ -0,0 +1,185 @@
+/*
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex.hh"
+
+
+static bool
+compose_hebrew (const hb_ot_shape_normalize_context_t *c,
+		hb_codepoint_t  a,
+		hb_codepoint_t  b,
+		hb_codepoint_t *ab)
+{
+  /* Hebrew presentation-form shaping.
+   * https://bugzilla.mozilla.org/show_bug.cgi?id=728866
+   * Hebrew presentation forms with dagesh, for characters U+05D0..05EA;
+   * Note that some letters do not have a dagesh presForm encoded.
+   */
+  static const hb_codepoint_t sDageshForms[0x05EAu - 0x05D0u + 1] = {
+    0xFB30u, /* ALEF */
+    0xFB31u, /* BET */
+    0xFB32u, /* GIMEL */
+    0xFB33u, /* DALET */
+    0xFB34u, /* HE */
+    0xFB35u, /* VAV */
+    0xFB36u, /* ZAYIN */
+    0x0000u, /* HET */
+    0xFB38u, /* TET */
+    0xFB39u, /* YOD */
+    0xFB3Au, /* FINAL KAF */
+    0xFB3Bu, /* KAF */
+    0xFB3Cu, /* LAMED */
+    0x0000u, /* FINAL MEM */
+    0xFB3Eu, /* MEM */
+    0x0000u, /* FINAL NUN */
+    0xFB40u, /* NUN */
+    0xFB41u, /* SAMEKH */
+    0x0000u, /* AYIN */
+    0xFB43u, /* FINAL PE */
+    0xFB44u, /* PE */
+    0x0000u, /* FINAL TSADI */
+    0xFB46u, /* TSADI */
+    0xFB47u, /* QOF */
+    0xFB48u, /* RESH */
+    0xFB49u, /* SHIN */
+    0xFB4Au /* TAV */
+  };
+
+  bool found = (bool) c->unicode->compose (a, b, ab);
+
+#ifdef HB_NO_OT_SHAPE_COMPLEX_HEBREW_FALLBACK
+  return found;
+#endif
+
+  if (!found && !c->plan->has_gpos_mark)
+  {
+      /* Special-case Hebrew presentation forms that are excluded from
+       * standard normalization, but wanted for old fonts. */
+      switch (b) {
+      case 0x05B4u: /* HIRIQ */
+	  if (a == 0x05D9u) { /* YOD */
+	      *ab = 0xFB1Du;
+	      found = true;
+	  }
+	  break;
+      case 0x05B7u: /* patah */
+	  if (a == 0x05F2u) { /* YIDDISH YOD YOD */
+	      *ab = 0xFB1Fu;
+	      found = true;
+	  } else if (a == 0x05D0u) { /* ALEF */
+	      *ab = 0xFB2Eu;
+	      found = true;
+	  }
+	  break;
+      case 0x05B8u: /* QAMATS */
+	  if (a == 0x05D0u) { /* ALEF */
+	      *ab = 0xFB2Fu;
+	      found = true;
+	  }
+	  break;
+      case 0x05B9u: /* HOLAM */
+	  if (a == 0x05D5u) { /* VAV */
+	      *ab = 0xFB4Bu;
+	      found = true;
+	  }
+	  break;
+      case 0x05BCu: /* DAGESH */
+	  if (a >= 0x05D0u && a <= 0x05EAu) {
+	      *ab = sDageshForms[a - 0x05D0u];
+	      found = (*ab != 0);
+	  } else if (a == 0xFB2Au) { /* SHIN WITH SHIN DOT */
+	      *ab = 0xFB2Cu;
+	      found = true;
+	  } else if (a == 0xFB2Bu) { /* SHIN WITH SIN DOT */
+	      *ab = 0xFB2Du;
+	      found = true;
+	  }
+	  break;
+      case 0x05BFu: /* RAFE */
+	  switch (a) {
+	  case 0x05D1u: /* BET */
+	      *ab = 0xFB4Cu;
+	      found = true;
+	      break;
+	  case 0x05DBu: /* KAF */
+	      *ab = 0xFB4Du;
+	      found = true;
+	      break;
+	  case 0x05E4u: /* PE */
+	      *ab = 0xFB4Eu;
+	      found = true;
+	      break;
+	  }
+	  break;
+      case 0x05C1u: /* SHIN DOT */
+	  if (a == 0x05E9u) { /* SHIN */
+	      *ab = 0xFB2Au;
+	      found = true;
+	  } else if (a == 0xFB49u) { /* SHIN WITH DAGESH */
+	      *ab = 0xFB2Cu;
+	      found = true;
+	  }
+	  break;
+      case 0x05C2u: /* SIN DOT */
+	  if (a == 0x05E9u) { /* SHIN */
+	      *ab = 0xFB2Bu;
+	      found = true;
+	  } else if (a == 0xFB49u) { /* SHIN WITH DAGESH */
+	      *ab = 0xFB2Du;
+	      found = true;
+	  }
+	  break;
+      }
+  }
+
+  return found;
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_hebrew =
+{
+  nullptr, /* collect_features */
+  nullptr, /* override_features */
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
+  nullptr, /* decompose */
+  compose_hebrew,
+  nullptr, /* setup_masks */
+  HB_TAG ('h','e','b','r'), /* gpos_tag. https://github.com/harfbuzz/harfbuzz/issues/347#issuecomment-267838368 */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
+  true, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-machine.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-machine.hh
new file mode 100644
index 0000000000..670b6bf486
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-machine.hh
@@ -0,0 +1,574 @@
+
+#line 1 "hb-ot-shape-complex-indic-machine.rl"
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH
+#define HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH
+
+#include "hb.hh"
+
+
+#line 36 "hb-ot-shape-complex-indic-machine.hh"
+static const unsigned char _indic_syllable_machine_trans_keys[] = {
+	8u, 8u, 4u, 8u, 5u, 7u, 5u, 8u, 4u, 8u, 6u, 6u, 16u, 16u, 4u, 8u, 
+	4u, 13u, 4u, 8u, 8u, 8u, 5u, 7u, 5u, 8u, 4u, 8u, 6u, 6u, 16u, 16u, 
+	4u, 8u, 4u, 13u, 4u, 13u, 4u, 13u, 8u, 8u, 5u, 7u, 5u, 8u, 4u, 8u, 
+	6u, 6u, 16u, 16u, 4u, 8u, 4u, 8u, 4u, 13u, 8u, 8u, 5u, 7u, 5u, 8u, 
+	4u, 8u, 6u, 6u, 16u, 16u, 4u, 8u, 4u, 8u, 5u, 8u, 8u, 8u, 1u, 19u, 
+	3u, 17u, 3u, 17u, 4u, 17u, 1u, 16u, 5u, 10u, 5u, 10u, 10u, 10u, 5u, 10u, 
+	1u, 16u, 1u, 16u, 1u, 16u, 3u, 10u, 4u, 10u, 5u, 10u, 4u, 10u, 5u, 10u, 
+	3u, 10u, 5u, 10u, 3u, 17u, 3u, 17u, 3u, 17u, 3u, 17u, 4u, 17u, 1u, 16u, 
+	3u, 17u, 3u, 17u, 4u, 17u, 1u, 16u, 5u, 10u, 10u, 10u, 5u, 10u, 1u, 16u, 
+	1u, 16u, 3u, 10u, 4u, 10u, 5u, 10u, 4u, 10u, 5u, 10u, 5u, 10u, 3u, 10u, 
+	5u, 10u, 3u, 17u, 3u, 17u, 4u, 8u, 3u, 17u, 3u, 17u, 4u, 17u, 1u, 16u, 
+	3u, 17u, 1u, 16u, 5u, 10u, 10u, 10u, 5u, 10u, 1u, 16u, 1u, 16u, 3u, 10u, 
+	4u, 10u, 5u, 10u, 3u, 17u, 4u, 10u, 5u, 10u, 5u, 10u, 3u, 10u, 5u, 10u, 
+	3u, 17u, 4u, 13u, 4u, 8u, 3u, 17u, 3u, 17u, 4u, 17u, 1u, 16u, 3u, 17u, 
+	1u, 16u, 5u, 10u, 10u, 10u, 5u, 10u, 1u, 16u, 1u, 16u, 3u, 10u, 4u, 10u, 
+	5u, 10u, 3u, 17u, 4u, 10u, 5u, 10u, 5u, 10u, 3u, 10u, 5u, 10u, 1u, 17u, 
+	3u, 17u, 1u, 17u, 4u, 13u, 5u, 10u, 10u, 10u, 5u, 10u, 1u, 16u, 3u, 10u, 
+	5u, 10u, 5u, 10u, 10u, 10u, 5u, 10u, 1u, 16u, 0
+};
+
+static const char _indic_syllable_machine_key_spans[] = {
+	1, 5, 3, 4, 5, 1, 1, 5, 
+	10, 5, 1, 3, 4, 5, 1, 1, 
+	5, 10, 10, 10, 1, 3, 4, 5, 
+	1, 1, 5, 5, 10, 1, 3, 4, 
+	5, 1, 1, 5, 5, 4, 1, 19, 
+	15, 15, 14, 16, 6, 6, 1, 6, 
+	16, 16, 16, 8, 7, 6, 7, 6, 
+	8, 6, 15, 15, 15, 15, 14, 16, 
+	15, 15, 14, 16, 6, 1, 6, 16, 
+	16, 8, 7, 6, 7, 6, 6, 8, 
+	6, 15, 15, 5, 15, 15, 14, 16, 
+	15, 16, 6, 1, 6, 16, 16, 8, 
+	7, 6, 15, 7, 6, 6, 8, 6, 
+	15, 10, 5, 15, 15, 14, 16, 15, 
+	16, 6, 1, 6, 16, 16, 8, 7, 
+	6, 15, 7, 6, 6, 8, 6, 17, 
+	15, 17, 10, 6, 1, 6, 16, 8, 
+	6, 6, 1, 6, 16
+};
+
+static const short _indic_syllable_machine_index_offsets[] = {
+	0, 2, 8, 12, 17, 23, 25, 27, 
+	33, 44, 50, 52, 56, 61, 67, 69, 
+	71, 77, 88, 99, 110, 112, 116, 121, 
+	127, 129, 131, 137, 143, 154, 156, 160, 
+	165, 171, 173, 175, 181, 187, 192, 194, 
+	214, 230, 246, 261, 278, 285, 292, 294, 
+	301, 318, 335, 352, 361, 369, 376, 384, 
+	391, 400, 407, 423, 439, 455, 471, 486, 
+	503, 519, 535, 550, 567, 574, 576, 583, 
+	600, 617, 626, 634, 641, 649, 656, 663, 
+	672, 679, 695, 711, 717, 733, 749, 764, 
+	781, 797, 814, 821, 823, 830, 847, 864, 
+	873, 881, 888, 904, 912, 919, 926, 935, 
+	942, 958, 969, 975, 991, 1007, 1022, 1039, 
+	1055, 1072, 1079, 1081, 1088, 1105, 1122, 1131, 
+	1139, 1146, 1162, 1170, 1177, 1184, 1193, 1200, 
+	1218, 1234, 1252, 1263, 1270, 1272, 1279, 1296, 
+	1305, 1312, 1319, 1321, 1328
+};
+
+static const unsigned char _indic_syllable_machine_indicies[] = {
+	1, 0, 2, 3, 3, 4, 1, 0, 
+	3, 3, 4, 0, 3, 3, 4, 1, 
+	0, 5, 3, 3, 4, 1, 0, 6, 
+	0, 7, 0, 8, 3, 3, 4, 1, 
+	0, 2, 3, 3, 4, 1, 0, 0, 
+	0, 0, 9, 0, 11, 12, 12, 13, 
+	14, 10, 14, 10, 12, 12, 13, 10, 
+	12, 12, 13, 14, 10, 15, 12, 12, 
+	13, 14, 10, 16, 10, 17, 10, 18, 
+	12, 12, 13, 14, 10, 11, 12, 12, 
+	13, 14, 10, 10, 10, 10, 19, 10, 
+	11, 12, 12, 13, 14, 10, 10, 10, 
+	10, 20, 10, 22, 23, 23, 24, 25, 
+	21, 21, 21, 21, 26, 21, 25, 21, 
+	23, 23, 24, 27, 23, 23, 24, 25, 
+	21, 28, 23, 23, 24, 25, 21, 29, 
+	21, 30, 21, 22, 23, 23, 24, 25, 
+	21, 31, 23, 23, 24, 25, 21, 33, 
+	34, 34, 35, 36, 32, 32, 32, 32, 
+	37, 32, 36, 32, 34, 34, 35, 32, 
+	34, 34, 35, 36, 32, 38, 34, 34, 
+	35, 36, 32, 39, 32, 40, 32, 33, 
+	34, 34, 35, 36, 32, 41, 34, 34, 
+	35, 36, 32, 23, 23, 24, 1, 0, 
+	43, 42, 45, 46, 47, 48, 49, 50, 
+	24, 25, 44, 51, 52, 52, 26, 44, 
+	53, 54, 55, 56, 57, 44, 59, 60, 
+	61, 62, 4, 1, 58, 63, 58, 58, 
+	9, 58, 58, 58, 64, 58, 65, 60, 
+	66, 66, 4, 1, 58, 63, 58, 58, 
+	58, 58, 58, 58, 64, 58, 60, 66, 
+	66, 4, 1, 58, 63, 58, 58, 58, 
+	58, 58, 58, 64, 58, 45, 58, 58, 
+	58, 67, 68, 58, 1, 58, 63, 58, 
+	58, 58, 58, 58, 45, 58, 69, 69, 
+	58, 1, 58, 63, 58, 63, 58, 58, 
+	70, 58, 63, 58, 63, 58, 63, 58, 
+	58, 58, 58, 63, 58, 45, 58, 71, 
+	58, 69, 69, 58, 1, 58, 63, 58, 
+	58, 58, 58, 58, 45, 58, 45, 58, 
+	58, 58, 69, 69, 58, 1, 58, 63, 
+	58, 58, 58, 58, 58, 45, 58, 45, 
+	58, 58, 58, 69, 68, 58, 1, 58, 
+	63, 58, 58, 58, 58, 58, 45, 58, 
+	72, 7, 73, 74, 4, 1, 58, 63, 
+	58, 7, 73, 74, 4, 1, 58, 63, 
+	58, 73, 73, 4, 1, 58, 63, 58, 
+	75, 76, 76, 4, 1, 58, 63, 58, 
+	67, 77, 58, 1, 58, 63, 58, 67, 
+	58, 69, 69, 58, 1, 58, 63, 58, 
+	69, 77, 58, 1, 58, 63, 58, 59, 
+	60, 66, 66, 4, 1, 58, 63, 58, 
+	58, 58, 58, 58, 58, 64, 58, 59, 
+	60, 61, 66, 4, 1, 58, 63, 58, 
+	58, 9, 58, 58, 58, 64, 58, 79, 
+	80, 81, 82, 13, 14, 78, 83, 78, 
+	78, 20, 78, 78, 78, 84, 78, 85, 
+	80, 86, 82, 13, 14, 78, 83, 78, 
+	78, 78, 78, 78, 78, 84, 78, 80, 
+	86, 82, 13, 14, 78, 83, 78, 78, 
+	78, 78, 78, 78, 84, 78, 87, 78, 
+	78, 78, 88, 89, 78, 14, 78, 83, 
+	78, 78, 78, 78, 78, 87, 78, 90, 
+	80, 91, 92, 13, 14, 78, 83, 78, 
+	78, 19, 78, 78, 78, 84, 78, 93, 
+	80, 86, 86, 13, 14, 78, 83, 78, 
+	78, 78, 78, 78, 78, 84, 78, 80, 
+	86, 86, 13, 14, 78, 83, 78, 78, 
+	78, 78, 78, 78, 84, 78, 87, 78, 
+	78, 78, 94, 89, 78, 14, 78, 83, 
+	78, 78, 78, 78, 78, 87, 78, 83, 
+	78, 78, 95, 78, 83, 78, 83, 78, 
+	83, 78, 78, 78, 78, 83, 78, 87, 
+	78, 96, 78, 94, 94, 78, 14, 78, 
+	83, 78, 78, 78, 78, 78, 87, 78, 
+	87, 78, 78, 78, 94, 94, 78, 14, 
+	78, 83, 78, 78, 78, 78, 78, 87, 
+	78, 97, 17, 98, 99, 13, 14, 78, 
+	83, 78, 17, 98, 99, 13, 14, 78, 
+	83, 78, 98, 98, 13, 14, 78, 83, 
+	78, 100, 101, 101, 13, 14, 78, 83, 
+	78, 88, 102, 78, 14, 78, 83, 78, 
+	94, 94, 78, 14, 78, 83, 78, 88, 
+	78, 94, 94, 78, 14, 78, 83, 78, 
+	94, 102, 78, 14, 78, 83, 78, 90, 
+	80, 86, 86, 13, 14, 78, 83, 78, 
+	78, 78, 78, 78, 78, 84, 78, 90, 
+	80, 91, 86, 13, 14, 78, 83, 78, 
+	78, 19, 78, 78, 78, 84, 78, 11, 
+	12, 12, 13, 14, 78, 79, 80, 86, 
+	82, 13, 14, 78, 83, 78, 78, 78, 
+	78, 78, 78, 84, 78, 104, 48, 105, 
+	105, 24, 25, 103, 51, 103, 103, 103, 
+	103, 103, 103, 55, 103, 48, 105, 105, 
+	24, 25, 103, 51, 103, 103, 103, 103, 
+	103, 103, 55, 103, 106, 103, 103, 103, 
+	107, 108, 103, 25, 103, 51, 103, 103, 
+	103, 103, 103, 106, 103, 47, 48, 109, 
+	110, 24, 25, 103, 51, 103, 103, 26, 
+	103, 103, 103, 55, 103, 106, 103, 103, 
+	103, 111, 108, 103, 25, 103, 51, 103, 
+	103, 103, 103, 103, 106, 103, 51, 103, 
+	103, 112, 103, 51, 103, 51, 103, 51, 
+	103, 103, 103, 103, 51, 103, 106, 103, 
+	113, 103, 111, 111, 103, 25, 103, 51, 
+	103, 103, 103, 103, 103, 106, 103, 106, 
+	103, 103, 103, 111, 111, 103, 25, 103, 
+	51, 103, 103, 103, 103, 103, 106, 103, 
+	114, 30, 115, 116, 24, 25, 103, 51, 
+	103, 30, 115, 116, 24, 25, 103, 51, 
+	103, 115, 115, 24, 25, 103, 51, 103, 
+	47, 48, 105, 105, 24, 25, 103, 51, 
+	103, 103, 103, 103, 103, 103, 55, 103, 
+	117, 118, 118, 24, 25, 103, 51, 103, 
+	107, 119, 103, 25, 103, 51, 103, 111, 
+	111, 103, 25, 103, 51, 103, 107, 103, 
+	111, 111, 103, 25, 103, 51, 103, 111, 
+	119, 103, 25, 103, 51, 103, 47, 48, 
+	109, 105, 24, 25, 103, 51, 103, 103, 
+	26, 103, 103, 103, 55, 103, 22, 23, 
+	23, 24, 25, 120, 120, 120, 120, 26, 
+	120, 22, 23, 23, 24, 25, 120, 122, 
+	123, 124, 125, 35, 36, 121, 126, 121, 
+	121, 37, 121, 121, 121, 127, 121, 128, 
+	123, 125, 125, 35, 36, 121, 126, 121, 
+	121, 121, 121, 121, 121, 127, 121, 123, 
+	125, 125, 35, 36, 121, 126, 121, 121, 
+	121, 121, 121, 121, 127, 121, 129, 121, 
+	121, 121, 130, 131, 121, 36, 121, 126, 
+	121, 121, 121, 121, 121, 129, 121, 122, 
+	123, 124, 52, 35, 36, 121, 126, 121, 
+	121, 37, 121, 121, 121, 127, 121, 129, 
+	121, 121, 121, 132, 131, 121, 36, 121, 
+	126, 121, 121, 121, 121, 121, 129, 121, 
+	126, 121, 121, 133, 121, 126, 121, 126, 
+	121, 126, 121, 121, 121, 121, 126, 121, 
+	129, 121, 134, 121, 132, 132, 121, 36, 
+	121, 126, 121, 121, 121, 121, 121, 129, 
+	121, 129, 121, 121, 121, 132, 132, 121, 
+	36, 121, 126, 121, 121, 121, 121, 121, 
+	129, 121, 135, 40, 136, 137, 35, 36, 
+	121, 126, 121, 40, 136, 137, 35, 36, 
+	121, 126, 121, 136, 136, 35, 36, 121, 
+	126, 121, 122, 123, 125, 125, 35, 36, 
+	121, 126, 121, 121, 121, 121, 121, 121, 
+	127, 121, 138, 139, 139, 35, 36, 121, 
+	126, 121, 130, 140, 121, 36, 121, 126, 
+	121, 132, 132, 121, 36, 121, 126, 121, 
+	130, 121, 132, 132, 121, 36, 121, 126, 
+	121, 132, 140, 121, 36, 121, 126, 121, 
+	45, 46, 47, 48, 109, 105, 24, 25, 
+	103, 51, 52, 52, 26, 103, 103, 45, 
+	55, 103, 59, 141, 61, 62, 4, 1, 
+	58, 63, 58, 58, 9, 58, 58, 58, 
+	64, 58, 45, 46, 47, 48, 142, 143, 
+	24, 144, 58, 145, 58, 52, 26, 58, 
+	58, 45, 55, 58, 22, 146, 146, 24, 
+	144, 58, 63, 58, 58, 26, 58, 145, 
+	58, 58, 147, 58, 145, 58, 145, 58, 
+	145, 58, 58, 58, 58, 145, 58, 45, 
+	58, 71, 22, 146, 146, 24, 144, 58, 
+	63, 58, 58, 58, 58, 58, 45, 58, 
+	149, 148, 150, 150, 148, 43, 148, 151, 
+	148, 150, 150, 148, 43, 148, 151, 148, 
+	151, 148, 148, 152, 148, 151, 148, 151, 
+	148, 151, 148, 148, 148, 148, 151, 148, 
+	45, 120, 120, 120, 120, 120, 120, 120, 
+	120, 120, 52, 120, 120, 120, 120, 45, 
+	120, 0
+};
+
+static const unsigned char _indic_syllable_machine_trans_targs[] = {
+	39, 45, 50, 2, 51, 5, 6, 53, 
+	57, 58, 39, 67, 11, 73, 68, 14, 
+	15, 75, 80, 81, 84, 39, 89, 21, 
+	95, 90, 98, 39, 24, 25, 97, 103, 
+	39, 112, 30, 118, 113, 121, 33, 34, 
+	120, 126, 39, 137, 39, 40, 60, 85, 
+	87, 105, 106, 91, 107, 127, 128, 99, 
+	135, 140, 39, 41, 43, 8, 59, 46, 
+	54, 42, 1, 44, 48, 0, 47, 49, 
+	52, 3, 4, 55, 7, 56, 39, 61, 
+	63, 18, 83, 69, 76, 62, 9, 64, 
+	78, 71, 65, 17, 82, 66, 10, 70, 
+	72, 74, 12, 13, 77, 16, 79, 39, 
+	86, 26, 88, 101, 93, 19, 104, 20, 
+	92, 94, 96, 22, 23, 100, 27, 102, 
+	39, 39, 108, 110, 28, 35, 114, 122, 
+	109, 111, 124, 116, 29, 115, 117, 119, 
+	31, 32, 123, 36, 125, 129, 130, 134, 
+	131, 132, 37, 133, 39, 136, 38, 138, 
+	139
+};
+
+static const char _indic_syllable_machine_trans_actions[] = {
+	1, 0, 2, 0, 2, 0, 0, 2, 
+	2, 2, 3, 2, 0, 2, 0, 0, 
+	0, 2, 2, 2, 2, 4, 2, 0, 
+	5, 0, 5, 6, 0, 0, 5, 2, 
+	7, 2, 0, 2, 0, 2, 0, 0, 
+	2, 2, 8, 0, 11, 2, 2, 5, 
+	0, 12, 12, 0, 2, 5, 2, 5, 
+	2, 0, 13, 2, 0, 0, 2, 0, 
+	2, 2, 0, 2, 2, 0, 0, 2, 
+	2, 0, 0, 0, 0, 2, 14, 2, 
+	0, 0, 2, 0, 2, 2, 0, 2, 
+	2, 2, 2, 0, 2, 2, 0, 0, 
+	2, 2, 0, 0, 0, 0, 2, 15, 
+	5, 0, 5, 2, 2, 0, 5, 0, 
+	0, 2, 5, 0, 0, 0, 0, 2, 
+	16, 17, 2, 0, 0, 0, 0, 2, 
+	2, 2, 2, 2, 0, 0, 2, 2, 
+	0, 0, 0, 0, 2, 0, 18, 18, 
+	0, 0, 0, 0, 19, 2, 0, 0, 
+	0
+};
+
+static const char _indic_syllable_machine_to_state_actions[] = {
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 9, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0
+};
+
+static const char _indic_syllable_machine_from_state_actions[] = {
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 10, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0
+};
+
+static const short _indic_syllable_machine_eof_trans[] = {
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 11, 11, 11, 11, 11, 11, 11, 
+	11, 11, 11, 22, 22, 28, 22, 22, 
+	22, 22, 22, 22, 33, 33, 33, 33, 
+	33, 33, 33, 33, 33, 1, 43, 0, 
+	59, 59, 59, 59, 59, 59, 59, 59, 
+	59, 59, 59, 59, 59, 59, 59, 59, 
+	59, 59, 59, 59, 79, 79, 79, 79, 
+	79, 79, 79, 79, 79, 79, 79, 79, 
+	79, 79, 79, 79, 79, 79, 79, 79, 
+	79, 79, 79, 79, 79, 104, 104, 104, 
+	104, 104, 104, 104, 104, 104, 104, 104, 
+	104, 104, 104, 104, 104, 104, 104, 104, 
+	104, 121, 121, 122, 122, 122, 122, 122, 
+	122, 122, 122, 122, 122, 122, 122, 122, 
+	122, 122, 122, 122, 122, 122, 122, 104, 
+	59, 59, 59, 59, 59, 59, 59, 149, 
+	149, 149, 149, 149, 121
+};
+
+static const int indic_syllable_machine_start = 39;
+static const int indic_syllable_machine_first_final = 39;
+static const int indic_syllable_machine_error = -1;
+
+static const int indic_syllable_machine_en_main = 39;
+
+
+#line 36 "hb-ot-shape-complex-indic-machine.rl"
+
+
+
+#line 93 "hb-ot-shape-complex-indic-machine.rl"
+
+
+#define found_syllable(syllable_type) \
+  HB_STMT_START { \
+    if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \
+    for (unsigned int i = ts; i < te; i++) \
+      info[i].syllable() = (syllable_serial << 4) | indic_##syllable_type; \
+    syllable_serial++; \
+    if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
+  } HB_STMT_END
+
+static void
+find_syllables_indic (hb_buffer_t *buffer)
+{
+  unsigned int p, pe, eof, ts, te, act;
+  int cs;
+  hb_glyph_info_t *info = buffer->info;
+  
+#line 411 "hb-ot-shape-complex-indic-machine.hh"
+	{
+	cs = indic_syllable_machine_start;
+	ts = 0;
+	te = 0;
+	act = 0;
+	}
+
+#line 113 "hb-ot-shape-complex-indic-machine.rl"
+
+
+  p = 0;
+  pe = eof = buffer->len;
+
+  unsigned int syllable_serial = 1;
+  
+#line 427 "hb-ot-shape-complex-indic-machine.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const unsigned char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+_resume:
+	switch ( _indic_syllable_machine_from_state_actions[cs] ) {
+	case 10:
+#line 1 "NONE"
+	{ts = p;}
+	break;
+#line 441 "hb-ot-shape-complex-indic-machine.hh"
+	}
+
+	_keys = _indic_syllable_machine_trans_keys + (cs<<1);
+	_inds = _indic_syllable_machine_indicies + _indic_syllable_machine_index_offsets[cs];
+
+	_slen = _indic_syllable_machine_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=( info[p].indic_category()) &&
+		( info[p].indic_category()) <= _keys[1] ?
+		( info[p].indic_category()) - _keys[0] : _slen ];
+
+_eof_trans:
+	cs = _indic_syllable_machine_trans_targs[_trans];
+
+	if ( _indic_syllable_machine_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _indic_syllable_machine_trans_actions[_trans] ) {
+	case 2:
+#line 1 "NONE"
+	{te = p+1;}
+	break;
+	case 11:
+#line 89 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p+1;{ found_syllable (non_indic_cluster); }}
+	break;
+	case 13:
+#line 84 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (consonant_syllable); }}
+	break;
+	case 14:
+#line 85 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (vowel_syllable); }}
+	break;
+	case 17:
+#line 86 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (standalone_cluster); }}
+	break;
+	case 19:
+#line 87 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (symbol_cluster); }}
+	break;
+	case 15:
+#line 88 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (broken_cluster); }}
+	break;
+	case 16:
+#line 89 "hb-ot-shape-complex-indic-machine.rl"
+	{te = p;p--;{ found_syllable (non_indic_cluster); }}
+	break;
+	case 1:
+#line 84 "hb-ot-shape-complex-indic-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (consonant_syllable); }}
+	break;
+	case 3:
+#line 85 "hb-ot-shape-complex-indic-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (vowel_syllable); }}
+	break;
+	case 7:
+#line 86 "hb-ot-shape-complex-indic-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (standalone_cluster); }}
+	break;
+	case 8:
+#line 87 "hb-ot-shape-complex-indic-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (symbol_cluster); }}
+	break;
+	case 4:
+#line 88 "hb-ot-shape-complex-indic-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (broken_cluster); }}
+	break;
+	case 6:
+#line 1 "NONE"
+	{	switch( act ) {
+	case 1:
+	{{p = ((te))-1;} found_syllable (consonant_syllable); }
+	break;
+	case 5:
+	{{p = ((te))-1;} found_syllable (broken_cluster); }
+	break;
+	case 6:
+	{{p = ((te))-1;} found_syllable (non_indic_cluster); }
+	break;
+	}
+	}
+	break;
+	case 18:
+#line 1 "NONE"
+	{te = p+1;}
+#line 84 "hb-ot-shape-complex-indic-machine.rl"
+	{act = 1;}
+	break;
+	case 5:
+#line 1 "NONE"
+	{te = p+1;}
+#line 88 "hb-ot-shape-complex-indic-machine.rl"
+	{act = 5;}
+	break;
+	case 12:
+#line 1 "NONE"
+	{te = p+1;}
+#line 89 "hb-ot-shape-complex-indic-machine.rl"
+	{act = 6;}
+	break;
+#line 544 "hb-ot-shape-complex-indic-machine.hh"
+	}
+
+_again:
+	switch ( _indic_syllable_machine_to_state_actions[cs] ) {
+	case 9:
+#line 1 "NONE"
+	{ts = 0;}
+	break;
+#line 553 "hb-ot-shape-complex-indic-machine.hh"
+	}
+
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	if ( p == eof )
+	{
+	if ( _indic_syllable_machine_eof_trans[cs] > 0 ) {
+		_trans = _indic_syllable_machine_eof_trans[cs] - 1;
+		goto _eof_trans;
+	}
+	}
+
+	}
+
+#line 121 "hb-ot-shape-complex-indic-machine.rl"
+
+}
+
+#undef found_syllable
+
+#endif /* HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-table.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-table.cc
new file mode 100644
index 0000000000..a150fd2486
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic-table.cc
@@ -0,0 +1,501 @@
+/* == Start of generated table == */
+/*
+ * The following table is generated by running:
+ *
+ *   ./gen-indic-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt Blocks.txt
+ *
+ * on files with these headers:
+ *
+ * # IndicSyllabicCategory-13.0.0.txt
+ * # Date: 2019-07-22, 19:55:00 GMT [KW, RP]
+ * # IndicPositionalCategory-13.0.0.txt
+ * # Date: 2019-07-23, 00:01:00 GMT [KW, RP]
+ * # Blocks-13.0.0.txt
+ * # Date: 2019-07-10, 19:06:00 GMT [KW]
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-indic.hh"
+
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-macros"
+
+#define ISC_A    INDIC_SYLLABIC_CATEGORY_AVAGRAHA                    /*   17 chars; Avagraha */
+#define ISC_Bi   INDIC_SYLLABIC_CATEGORY_BINDU                       /*   91 chars; Bindu */
+#define ISC_BJN  INDIC_SYLLABIC_CATEGORY_BRAHMI_JOINING_NUMBER       /*   20 chars; Brahmi_Joining_Number */
+#define ISC_Ca   INDIC_SYLLABIC_CATEGORY_CANTILLATION_MARK           /*   59 chars; Cantillation_Mark */
+#define ISC_C    INDIC_SYLLABIC_CATEGORY_CONSONANT                   /* 2195 chars; Consonant */
+#define ISC_CD   INDIC_SYLLABIC_CATEGORY_CONSONANT_DEAD              /*   12 chars; Consonant_Dead */
+#define ISC_CF   INDIC_SYLLABIC_CATEGORY_CONSONANT_FINAL             /*   67 chars; Consonant_Final */
+#define ISC_CHL  INDIC_SYLLABIC_CATEGORY_CONSONANT_HEAD_LETTER       /*    5 chars; Consonant_Head_Letter */
+#define ISC_CIP  INDIC_SYLLABIC_CATEGORY_CONSONANT_INITIAL_POSTFIXED /*    1 chars; Consonant_Initial_Postfixed */
+#define ISC_CK   INDIC_SYLLABIC_CATEGORY_CONSONANT_KILLER            /*    2 chars; Consonant_Killer */
+#define ISC_CM   INDIC_SYLLABIC_CATEGORY_CONSONANT_MEDIAL            /*   31 chars; Consonant_Medial */
+#define ISC_CP   INDIC_SYLLABIC_CATEGORY_CONSONANT_PLACEHOLDER       /*   22 chars; Consonant_Placeholder */
+#define ISC_CPR  INDIC_SYLLABIC_CATEGORY_CONSONANT_PRECEDING_REPHA   /*    3 chars; Consonant_Preceding_Repha */
+#define ISC_CPrf INDIC_SYLLABIC_CATEGORY_CONSONANT_PREFIXED          /*   10 chars; Consonant_Prefixed */
+#define ISC_CS   INDIC_SYLLABIC_CATEGORY_CONSONANT_SUBJOINED         /*   94 chars; Consonant_Subjoined */
+#define ISC_CSR  INDIC_SYLLABIC_CATEGORY_CONSONANT_SUCCEEDING_REPHA  /*    4 chars; Consonant_Succeeding_Repha */
+#define ISC_CWS  INDIC_SYLLABIC_CATEGORY_CONSONANT_WITH_STACKER      /*    8 chars; Consonant_With_Stacker */
+#define ISC_GM   INDIC_SYLLABIC_CATEGORY_GEMINATION_MARK             /*    3 chars; Gemination_Mark */
+#define ISC_IS   INDIC_SYLLABIC_CATEGORY_INVISIBLE_STACKER           /*   12 chars; Invisible_Stacker */
+#define ISC_ZWJ  INDIC_SYLLABIC_CATEGORY_JOINER                      /*    1 chars; Joiner */
+#define ISC_ML   INDIC_SYLLABIC_CATEGORY_MODIFYING_LETTER            /*    1 chars; Modifying_Letter */
+#define ISC_ZWNJ INDIC_SYLLABIC_CATEGORY_NON_JOINER                  /*    1 chars; Non_Joiner */
+#define ISC_N    INDIC_SYLLABIC_CATEGORY_NUKTA                       /*   31 chars; Nukta */
+#define ISC_Nd   INDIC_SYLLABIC_CATEGORY_NUMBER                      /*  491 chars; Number */
+#define ISC_NJ   INDIC_SYLLABIC_CATEGORY_NUMBER_JOINER               /*    1 chars; Number_Joiner */
+#define ISC_x    INDIC_SYLLABIC_CATEGORY_OTHER                       /*    1 chars; Other */
+#define ISC_PK   INDIC_SYLLABIC_CATEGORY_PURE_KILLER                 /*   23 chars; Pure_Killer */
+#define ISC_RS   INDIC_SYLLABIC_CATEGORY_REGISTER_SHIFTER            /*    2 chars; Register_Shifter */
+#define ISC_SM   INDIC_SYLLABIC_CATEGORY_SYLLABLE_MODIFIER           /*   25 chars; Syllable_Modifier */
+#define ISC_TL   INDIC_SYLLABIC_CATEGORY_TONE_LETTER                 /*    7 chars; Tone_Letter */
+#define ISC_TM   INDIC_SYLLABIC_CATEGORY_TONE_MARK                   /*   42 chars; Tone_Mark */
+#define ISC_V    INDIC_SYLLABIC_CATEGORY_VIRAMA                      /*   27 chars; Virama */
+#define ISC_Vs   INDIC_SYLLABIC_CATEGORY_VISARGA                     /*   35 chars; Visarga */
+#define ISC_Vo   INDIC_SYLLABIC_CATEGORY_VOWEL                       /*   30 chars; Vowel */
+#define ISC_M    INDIC_SYLLABIC_CATEGORY_VOWEL_DEPENDENT             /*  683 chars; Vowel_Dependent */
+#define ISC_VI   INDIC_SYLLABIC_CATEGORY_VOWEL_INDEPENDENT           /*  484 chars; Vowel_Independent */
+
+#define IMC_B    INDIC_MATRA_CATEGORY_BOTTOM                         /*  351 chars; Bottom */
+#define IMC_BL   INDIC_MATRA_CATEGORY_BOTTOM_AND_LEFT                /*    1 chars; Bottom_And_Left */
+#define IMC_BR   INDIC_MATRA_CATEGORY_BOTTOM_AND_RIGHT               /*    4 chars; Bottom_And_Right */
+#define IMC_L    INDIC_MATRA_CATEGORY_LEFT                           /*   64 chars; Left */
+#define IMC_LR   INDIC_MATRA_CATEGORY_LEFT_AND_RIGHT                 /*   22 chars; Left_And_Right */
+#define IMC_x    INDIC_MATRA_CATEGORY_NOT_APPLICABLE                 /*    1 chars; Not_Applicable */
+#define IMC_O    INDIC_MATRA_CATEGORY_OVERSTRUCK                     /*   10 chars; Overstruck */
+#define IMC_R    INDIC_MATRA_CATEGORY_RIGHT                          /*  288 chars; Right */
+#define IMC_T    INDIC_MATRA_CATEGORY_TOP                            /*  415 chars; Top */
+#define IMC_TB   INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM                 /*   10 chars; Top_And_Bottom */
+#define IMC_TBL  INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM_AND_LEFT        /*    2 chars; Top_And_Bottom_And_Left */
+#define IMC_TBR  INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM_AND_RIGHT       /*    1 chars; Top_And_Bottom_And_Right */
+#define IMC_TL   INDIC_MATRA_CATEGORY_TOP_AND_LEFT                   /*    6 chars; Top_And_Left */
+#define IMC_TLR  INDIC_MATRA_CATEGORY_TOP_AND_LEFT_AND_RIGHT         /*    4 chars; Top_And_Left_And_Right */
+#define IMC_TR   INDIC_MATRA_CATEGORY_TOP_AND_RIGHT                  /*   13 chars; Top_And_Right */
+#define IMC_VOL  INDIC_MATRA_CATEGORY_VISUAL_ORDER_LEFT              /*   19 chars; Visual_Order_Left */
+
+#pragma GCC diagnostic pop
+
+#define _(S,M) INDIC_COMBINE_CATEGORIES (ISC_##S, IMC_##M)
+
+
+static const INDIC_TABLE_ELEMENT_TYPE indic_table[] = {
+
+
+#define indic_offset_0x0028u 0
+
+
+  /* Basic Latin */
+
+  /* 0028 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(CP,x),  _(x,x),  _(x,x),
+  /* 0030 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0038 */ _(Nd,x), _(Nd,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0x00b0u 24
+
+
+  /* Latin-1 Supplement */
+
+  /* 00B0 */  _(x,x),  _(x,x), _(SM,x), _(SM,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 00B8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 00C0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 00C8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 00D0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(CP,x),
+
+#define indic_offset_0x0900u 64
+
+
+  /* Devanagari */
+
+  /* 0900 */ _(Bi,T), _(Bi,T), _(Bi,T), _(Vs,R), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0908 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0910 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0918 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0920 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0928 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0930 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0938 */  _(C,x),  _(C,x),  _(M,T),  _(M,R),  _(N,B),  _(A,x),  _(M,R),  _(M,L),
+  /* 0940 */  _(M,R),  _(M,B),  _(M,B),  _(M,B),  _(M,B),  _(M,T),  _(M,T),  _(M,T),
+  /* 0948 */  _(M,T),  _(M,R),  _(M,R),  _(M,R),  _(M,R),  _(V,B),  _(M,L),  _(M,R),
+  /* 0950 */  _(x,x), _(Ca,T), _(Ca,B),  _(x,T),  _(x,T),  _(M,T),  _(M,B),  _(M,B),
+  /* 0958 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0960 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0968 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0970 */  _(x,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0978 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+
+  /* Bengali */
+
+  /* 0980 */ _(CP,x), _(Bi,T), _(Bi,R), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0988 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x),  _(x,x), _(VI,x),
+  /* 0990 */ _(VI,x),  _(x,x),  _(x,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0998 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 09A0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 09A8 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 09B0 */  _(C,x),  _(x,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(C,x),
+  /* 09B8 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(N,B),  _(A,x),  _(M,R),  _(M,L),
+  /* 09C0 */  _(M,R),  _(M,B),  _(M,B),  _(M,B),  _(M,B),  _(x,x),  _(x,x),  _(M,L),
+  /* 09C8 */  _(M,L),  _(x,x),  _(x,x), _(M,LR), _(M,LR),  _(V,B), _(CD,x),  _(x,x),
+  /* 09D0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,R),
+  /* 09D8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),
+  /* 09E0 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 09E8 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 09F0 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 09F8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(Bi,x),  _(x,x), _(SM,T),  _(x,x),
+
+  /* Gurmukhi */
+
+  /* 0A00 */  _(x,x), _(Bi,T), _(Bi,T), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0A08 */ _(VI,x), _(VI,x), _(VI,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(VI,x),
+  /* 0A10 */ _(VI,x),  _(x,x),  _(x,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0A18 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0A20 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0A28 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0A30 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),
+  /* 0A38 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(N,B),  _(x,x),  _(M,R),  _(M,L),
+  /* 0A40 */  _(M,R),  _(M,B),  _(M,B),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,T),
+  /* 0A48 */  _(M,T),  _(x,x),  _(x,x),  _(M,T),  _(M,T),  _(V,B),  _(x,x),  _(x,x),
+  /* 0A50 */  _(x,x), _(Ca,B),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0A58 */  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(x,x),
+  /* 0A60 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0A68 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0A70 */ _(Bi,T), _(GM,T), _(CP,x), _(CP,x),  _(x,x), _(CM,B),  _(x,x),  _(x,x),
+  /* 0A78 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+  /* Gujarati */
+
+  /* 0A80 */  _(x,x), _(Bi,T), _(Bi,T), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0A88 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x), _(VI,x),
+  /* 0A90 */ _(VI,x), _(VI,x),  _(x,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0A98 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0AA0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0AA8 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0AB0 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0AB8 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(N,B),  _(A,x),  _(M,R),  _(M,L),
+  /* 0AC0 */  _(M,R),  _(M,B),  _(M,B),  _(M,B),  _(M,B),  _(M,T),  _(x,x),  _(M,T),
+  /* 0AC8 */  _(M,T), _(M,TR),  _(x,x),  _(M,R),  _(M,R),  _(V,B),  _(x,x),  _(x,x),
+  /* 0AD0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0AD8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0AE0 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0AE8 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0AF0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0AF8 */  _(x,x),  _(C,x), _(Ca,T), _(Ca,T), _(Ca,T),  _(N,T),  _(N,T),  _(N,T),
+
+  /* Oriya */
+
+  /* 0B00 */  _(x,x), _(Bi,T), _(Bi,R), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0B08 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x),  _(x,x), _(VI,x),
+  /* 0B10 */ _(VI,x),  _(x,x),  _(x,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0B18 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0B20 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0B28 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0B30 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0B38 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(N,B),  _(A,x),  _(M,R),  _(M,T),
+  /* 0B40 */  _(M,R),  _(M,B),  _(M,B),  _(M,B),  _(M,B),  _(x,x),  _(x,x),  _(M,L),
+  /* 0B48 */ _(M,TL),  _(x,x),  _(x,x), _(M,LR),_(M,TLR),  _(V,B),  _(x,x),  _(x,x),
+  /* 0B50 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,T),  _(M,T), _(M,TR),
+  /* 0B58 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),
+  /* 0B60 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0B68 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0B70 */  _(x,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0B78 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+  /* Tamil */
+
+  /* 0B80 */  _(x,x),  _(x,x), _(Bi,T), _(ML,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0B88 */ _(VI,x), _(VI,x), _(VI,x),  _(x,x),  _(x,x),  _(x,x), _(VI,x), _(VI,x),
+  /* 0B90 */ _(VI,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x),  _(C,x),  _(x,x),  _(x,x),
+  /* 0B98 */  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),
+  /* 0BA0 */  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0BA8 */  _(C,x),  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(C,x),
+  /* 0BB0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0BB8 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,R),  _(M,R),
+  /* 0BC0 */  _(M,T),  _(M,R),  _(M,R),  _(x,x),  _(x,x),  _(x,x),  _(M,L),  _(M,L),
+  /* 0BC8 */  _(M,L),  _(x,x), _(M,LR), _(M,LR), _(M,LR),  _(V,T),  _(x,x),  _(x,x),
+  /* 0BD0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,R),
+  /* 0BD8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0BE0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0BE8 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0BF0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0BF8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+  /* Telugu */
+
+  /* 0C00 */ _(Bi,T), _(Bi,R), _(Bi,R), _(Vs,R), _(Bi,T), _(VI,x), _(VI,x), _(VI,x),
+  /* 0C08 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x), _(VI,x), _(VI,x),
+  /* 0C10 */ _(VI,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C18 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C20 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C28 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C30 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C38 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(A,x),  _(M,T),  _(M,T),
+  /* 0C40 */  _(M,T),  _(M,R),  _(M,R),  _(M,R),  _(M,R),  _(x,x),  _(M,T),  _(M,T),
+  /* 0C48 */ _(M,TB),  _(x,x),  _(M,T),  _(M,T),  _(M,T),  _(V,T),  _(x,x),  _(x,x),
+  /* 0C50 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,T),  _(M,B),  _(x,x),
+  /* 0C58 */  _(C,x),  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0C60 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0C68 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0C70 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0C78 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+  /* Kannada */
+
+  /* 0C80 */ _(Bi,x), _(Bi,T), _(Bi,R), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0C88 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x), _(VI,x), _(VI,x),
+  /* 0C90 */ _(VI,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0C98 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0CA0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0CA8 */  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0CB0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0CB8 */  _(C,x),  _(C,x),  _(x,x),  _(x,x),  _(N,B),  _(A,x),  _(M,R),  _(M,T),
+  /* 0CC0 */ _(M,TR),  _(M,R),  _(M,R),  _(M,R),  _(M,R),  _(x,x),  _(M,T), _(M,TR),
+  /* 0CC8 */ _(M,TR),  _(x,x), _(M,TR), _(M,TR),  _(M,T),  _(V,T),  _(x,x),  _(x,x),
+  /* 0CD0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,R),  _(M,R),  _(x,x),
+  /* 0CD8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(C,x),  _(x,x),
+  /* 0CE0 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0CE8 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0CF0 */  _(x,x),_(CWS,x),_(CWS,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0CF8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+  /* Malayalam */
+
+  /* 0D00 */ _(Bi,T), _(Bi,T), _(Bi,R), _(Vs,R), _(Bi,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0D08 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x), _(VI,x), _(VI,x),
+  /* 0D10 */ _(VI,x),  _(x,x), _(VI,x), _(VI,x), _(VI,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0D18 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0D20 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0D28 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0D30 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0D38 */  _(C,x),  _(C,x),  _(C,x), _(PK,T), _(PK,T),  _(A,x),  _(M,R),  _(M,R),
+  /* 0D40 */  _(M,R),  _(M,R),  _(M,R),  _(M,B),  _(M,B),  _(x,x),  _(M,L),  _(M,L),
+  /* 0D48 */  _(M,L),  _(x,x), _(M,LR), _(M,LR), _(M,LR),  _(V,T),_(CPR,T),  _(x,x),
+  /* 0D50 */  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(CD,x), _(CD,x), _(CD,x),  _(M,R),
+  /* 0D58 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(VI,x),
+  /* 0D60 */ _(VI,x), _(VI,x),  _(M,B),  _(M,B),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0D68 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0D70 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 0D78 */  _(x,x),  _(x,x), _(CD,x), _(CD,x), _(CD,x), _(CD,x), _(CD,x), _(CD,x),
+
+  /* Sinhala */
+
+  /* 0D80 */  _(x,x), _(Bi,T), _(Bi,R), _(Vs,R),  _(x,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0D88 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 0D90 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x),
+  /* 0D98 */  _(x,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0DA0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0DA8 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0DB0 */  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 0DB8 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(x,x),  _(C,x),  _(x,x),  _(x,x),
+  /* 0DC0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(x,x),
+  /* 0DC8 */  _(x,x),  _(x,x),  _(V,T),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(M,R),
+  /* 0DD0 */  _(M,R),  _(M,R),  _(M,T),  _(M,T),  _(M,B),  _(x,x),  _(M,B),  _(x,x),
+  /* 0DD8 */  _(M,R),  _(M,L), _(M,TL),  _(M,L), _(M,LR),_(M,TLR), _(M,LR),  _(M,R),
+  /* 0DE0 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(Nd,x), _(Nd,x),
+  /* 0DE8 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 0DF0 */  _(x,x),  _(x,x),  _(M,R),  _(M,R),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0x1000u 1336
+
+
+  /* Myanmar */
+
+  /* 1000 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1008 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1010 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1018 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1020 */  _(C,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 1028 */ _(VI,x), _(VI,x), _(VI,x),  _(M,R),  _(M,R),  _(M,T),  _(M,T),  _(M,B),
+  /* 1030 */  _(M,B),  _(M,L),  _(M,T),  _(M,T),  _(M,T),  _(M,T), _(Bi,T), _(TM,B),
+  /* 1038 */ _(Vs,R), _(IS,x), _(PK,T), _(CM,R),_(CM,TBL), _(CM,B), _(CM,B),  _(C,x),
+  /* 1040 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 1048 */ _(Nd,x), _(Nd,x),  _(x,x), _(CP,x),  _(x,x),  _(x,x), _(CP,x),  _(x,x),
+  /* 1050 */  _(C,x),  _(C,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(M,R),  _(M,R),
+  /* 1058 */  _(M,B),  _(M,B),  _(C,x),  _(C,x),  _(C,x),  _(C,x), _(CM,B), _(CM,B),
+  /* 1060 */ _(CM,B),  _(C,x),  _(M,R), _(TM,R), _(TM,R),  _(C,x),  _(C,x),  _(M,R),
+  /* 1068 */  _(M,R), _(TM,R), _(TM,R), _(TM,R), _(TM,R), _(TM,R),  _(C,x),  _(C,x),
+  /* 1070 */  _(C,x),  _(M,T),  _(M,T),  _(M,T),  _(M,T),  _(C,x),  _(C,x),  _(C,x),
+  /* 1078 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1080 */  _(C,x),  _(C,x), _(CM,B),  _(M,R),  _(M,L),  _(M,T),  _(M,T), _(TM,R),
+  /* 1088 */ _(TM,R), _(TM,R), _(TM,R), _(TM,R), _(TM,R), _(TM,B),  _(C,x), _(TM,R),
+  /* 1090 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 1098 */ _(Nd,x), _(Nd,x), _(TM,R), _(TM,R),  _(M,R),  _(M,T),  _(x,x),  _(x,x),
+
+#define indic_offset_0x1780u 1496
+
+
+  /* Khmer */
+
+  /* 1780 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1788 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1790 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 1798 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* 17A0 */  _(C,x),  _(C,x),  _(C,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 17A8 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x), _(VI,x),
+  /* 17B0 */ _(VI,x), _(VI,x), _(VI,x), _(VI,x),  _(x,x),  _(x,x),  _(M,R),  _(M,T),
+  /* 17B8 */  _(M,T),  _(M,T),  _(M,T),  _(M,B),  _(M,B),  _(M,B), _(M,TL),_(M,TLR),
+  /* 17C0 */ _(M,LR),  _(M,L),  _(M,L),  _(M,L), _(M,LR), _(M,LR), _(Bi,T), _(Vs,R),
+  /* 17C8 */  _(M,R), _(RS,T), _(RS,T), _(SM,T),_(CSR,T), _(CK,T), _(SM,T), _(SM,T),
+  /* 17D0 */ _(SM,T), _(PK,T), _(IS,x), _(SM,T),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 17D8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(A,x), _(SM,T),  _(x,x),  _(x,x),
+  /* 17E0 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* 17E8 */ _(Nd,x), _(Nd,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0x1cd0u 1608
+
+
+  /* Vedic Extensions */
+
+  /* 1CD0 */ _(Ca,T), _(Ca,T), _(Ca,T),  _(x,x), _(Ca,O), _(Ca,B), _(Ca,B), _(Ca,B),
+  /* 1CD8 */ _(Ca,B), _(Ca,B), _(Ca,T), _(Ca,T), _(Ca,B), _(Ca,B), _(Ca,B), _(Ca,B),
+  /* 1CE0 */ _(Ca,T), _(Ca,R),  _(x,O),  _(x,O),  _(x,O),  _(x,O),  _(x,O),  _(x,O),
+  /* 1CE8 */  _(x,O),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,B),  _(x,x),  _(x,x),
+  /* 1CF0 */  _(x,x),  _(x,x), _(CD,x), _(CD,x), _(Ca,T),_(CWS,x),_(CWS,x), _(Ca,R),
+  /* 1CF8 */ _(Ca,x), _(Ca,x), _(CP,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0x2008u 1656
+
+
+  /* General Punctuation */
+
+  /* 2008 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),_(ZWNJ,x),_(ZWJ,x),  _(x,x),  _(x,x),
+  /* 2010 */ _(CP,x), _(CP,x), _(CP,x), _(CP,x), _(CP,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0x2070u 1672
+
+
+  /* Superscripts and Subscripts */
+
+  /* 2070 */  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(SM,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 2078 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* 2080 */  _(x,x),  _(x,x), _(SM,x), _(SM,x), _(SM,x),  _(x,x),  _(x,x),  _(x,x),
+
+#define indic_offset_0xa8e0u 1696
+
+
+  /* Devanagari Extended */
+
+  /* A8E0 */ _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T),
+  /* A8E8 */ _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T), _(Ca,T),
+  /* A8F0 */ _(Ca,T), _(Ca,T), _(Bi,x), _(Bi,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),
+  /* A8F8 */  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x),  _(x,x), _(VI,x),  _(M,T),
+
+#define indic_offset_0xa9e0u 1728
+
+
+  /* Myanmar Extended-B */
+
+  /* A9E0 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(M,T),  _(x,x),  _(C,x),
+  /* A9E8 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* A9F0 */ _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x), _(Nd,x),
+  /* A9F8 */ _(Nd,x), _(Nd,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(x,x),
+
+#define indic_offset_0xaa60u 1760
+
+
+  /* Myanmar Extended-A */
+
+  /* AA60 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* AA68 */  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),  _(C,x),
+  /* AA70 */  _(x,x),  _(C,x),  _(C,x),  _(C,x), _(CP,x), _(CP,x), _(CP,x),  _(x,x),
+  /* AA78 */  _(x,x),  _(x,x),  _(C,x), _(TM,R), _(TM,T), _(TM,R),  _(C,x),  _(C,x),
+
+}; /* Table items: 1792; occupancy: 70% */
+
+INDIC_TABLE_ELEMENT_TYPE
+hb_indic_get_categories (hb_codepoint_t u)
+{
+  switch (u >> 12)
+  {
+    case 0x0u:
+      if (unlikely (u == 0x00A0u)) return _(CP,x);
+      if (hb_in_range<hb_codepoint_t> (u, 0x0028u, 0x003Fu)) return indic_table[u - 0x0028u + indic_offset_0x0028u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x00B0u, 0x00D7u)) return indic_table[u - 0x00B0u + indic_offset_0x00b0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x0900u, 0x0DF7u)) return indic_table[u - 0x0900u + indic_offset_0x0900u];
+      break;
+
+    case 0x1u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x1000u, 0x109Fu)) return indic_table[u - 0x1000u + indic_offset_0x1000u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1780u, 0x17EFu)) return indic_table[u - 0x1780u + indic_offset_0x1780u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1CD0u, 0x1CFFu)) return indic_table[u - 0x1CD0u + indic_offset_0x1cd0u];
+      break;
+
+    case 0x2u:
+      if (unlikely (u == 0x25CCu)) return _(CP,x);
+      if (hb_in_range<hb_codepoint_t> (u, 0x2008u, 0x2017u)) return indic_table[u - 0x2008u + indic_offset_0x2008u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x2070u, 0x2087u)) return indic_table[u - 0x2070u + indic_offset_0x2070u];
+      break;
+
+    case 0xAu:
+      if (hb_in_range<hb_codepoint_t> (u, 0xA8E0u, 0xA8FFu)) return indic_table[u - 0xA8E0u + indic_offset_0xa8e0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0xA9E0u, 0xA9FFu)) return indic_table[u - 0xA9E0u + indic_offset_0xa9e0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0xAA60u, 0xAA7Fu)) return indic_table[u - 0xAA60u + indic_offset_0xaa60u];
+      break;
+
+    default:
+      break;
+  }
+  return _(x,x);
+}
+
+#undef _
+
+#undef ISC_A
+#undef ISC_Bi
+#undef ISC_BJN
+#undef ISC_Ca
+#undef ISC_C
+#undef ISC_CD
+#undef ISC_CF
+#undef ISC_CHL
+#undef ISC_CIP
+#undef ISC_CK
+#undef ISC_CM
+#undef ISC_CP
+#undef ISC_CPR
+#undef ISC_CPrf
+#undef ISC_CS
+#undef ISC_CSR
+#undef ISC_CWS
+#undef ISC_GM
+#undef ISC_IS
+#undef ISC_ZWJ
+#undef ISC_ML
+#undef ISC_ZWNJ
+#undef ISC_N
+#undef ISC_Nd
+#undef ISC_NJ
+#undef ISC_x
+#undef ISC_PK
+#undef ISC_RS
+#undef ISC_SM
+#undef ISC_TL
+#undef ISC_TM
+#undef ISC_V
+#undef ISC_Vs
+#undef ISC_Vo
+#undef ISC_M
+#undef ISC_VI
+
+#undef IMC_B
+#undef IMC_BL
+#undef IMC_BR
+#undef IMC_L
+#undef IMC_LR
+#undef IMC_x
+#undef IMC_O
+#undef IMC_R
+#undef IMC_T
+#undef IMC_TB
+#undef IMC_TBL
+#undef IMC_TBR
+#undef IMC_TL
+#undef IMC_TLR
+#undef IMC_TR
+#undef IMC_VOL
+
+#endif
+
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.cc
new file mode 100644
index 0000000000..34972f81e2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.cc
@@ -0,0 +1,1615 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-indic.hh"
+#include "hb-ot-shape-complex-vowel-constraints.hh"
+#include "hb-ot-layout.hh"
+
+
+/*
+ * Indic shaper.
+ */
+
+
+/*
+ * Indic configurations.  Note that we do not want to keep every single script-specific
+ * behavior in these tables necessarily.  This should mainly be used for per-script
+ * properties that are cheaper keeping here, than in the code.  Ie. if, say, one and
+ * only one script has an exception, that one script can be if'ed directly in the code,
+ * instead of adding a new flag in these structs.
+ */
+
+enum base_position_t {
+  BASE_POS_LAST_SINHALA,
+  BASE_POS_LAST
+};
+enum reph_position_t {
+  REPH_POS_AFTER_MAIN  = POS_AFTER_MAIN,
+  REPH_POS_BEFORE_SUB  = POS_BEFORE_SUB,
+  REPH_POS_AFTER_SUB   = POS_AFTER_SUB,
+  REPH_POS_BEFORE_POST = POS_BEFORE_POST,
+  REPH_POS_AFTER_POST  = POS_AFTER_POST
+};
+enum reph_mode_t {
+  REPH_MODE_IMPLICIT,  /* Reph formed out of initial Ra,H sequence. */
+  REPH_MODE_EXPLICIT,  /* Reph formed out of initial Ra,H,ZWJ sequence. */
+  REPH_MODE_LOG_REPHA  /* Encoded Repha character, needs reordering. */
+};
+enum blwf_mode_t {
+  BLWF_MODE_PRE_AND_POST, /* Below-forms feature applied to pre-base and post-base. */
+  BLWF_MODE_POST_ONLY     /* Below-forms feature applied to post-base only. */
+};
+struct indic_config_t
+{
+  hb_script_t     script;
+  bool            has_old_spec;
+  hb_codepoint_t  virama;
+  base_position_t base_pos;
+  reph_position_t reph_pos;
+  reph_mode_t     reph_mode;
+  blwf_mode_t     blwf_mode;
+};
+
+static const indic_config_t indic_configs[] =
+{
+  /* Default.  Should be first. */
+  {HB_SCRIPT_INVALID,	false,      0,BASE_POS_LAST, REPH_POS_BEFORE_POST,REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_DEVANAGARI,true, 0x094Du,BASE_POS_LAST, REPH_POS_BEFORE_POST,REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_BENGALI,	true, 0x09CDu,BASE_POS_LAST, REPH_POS_AFTER_SUB,  REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_GURMUKHI,	true, 0x0A4Du,BASE_POS_LAST, REPH_POS_BEFORE_SUB, REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_GUJARATI,	true, 0x0ACDu,BASE_POS_LAST, REPH_POS_BEFORE_POST,REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_ORIYA,	true, 0x0B4Du,BASE_POS_LAST, REPH_POS_AFTER_MAIN, REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_TAMIL,	true, 0x0BCDu,BASE_POS_LAST, REPH_POS_AFTER_POST, REPH_MODE_IMPLICIT, BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_TELUGU,	true, 0x0C4Du,BASE_POS_LAST, REPH_POS_AFTER_POST, REPH_MODE_EXPLICIT, BLWF_MODE_POST_ONLY},
+  {HB_SCRIPT_KANNADA,	true, 0x0CCDu,BASE_POS_LAST, REPH_POS_AFTER_POST, REPH_MODE_IMPLICIT, BLWF_MODE_POST_ONLY},
+  {HB_SCRIPT_MALAYALAM,	true, 0x0D4Du,BASE_POS_LAST, REPH_POS_AFTER_MAIN, REPH_MODE_LOG_REPHA,BLWF_MODE_PRE_AND_POST},
+  {HB_SCRIPT_SINHALA,	false,0x0DCAu,BASE_POS_LAST_SINHALA,
+						     REPH_POS_AFTER_POST, REPH_MODE_EXPLICIT, BLWF_MODE_PRE_AND_POST},
+};
+
+
+
+/*
+ * Indic shaper.
+ */
+
+static const hb_ot_map_feature_t
+indic_features[] =
+{
+  /*
+   * Basic features.
+   * These features are applied in order, one at a time, after initial_reordering.
+   */
+  {HB_TAG('n','u','k','t'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('a','k','h','n'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('r','p','h','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('r','k','r','f'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('p','r','e','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('b','l','w','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('a','b','v','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('h','a','l','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('p','s','t','f'),        F_MANUAL_JOINERS},
+  {HB_TAG('v','a','t','u'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('c','j','c','t'), F_GLOBAL_MANUAL_JOINERS},
+  /*
+   * Other features.
+   * These features are applied all at once, after final_reordering
+   * but before clearing syllables.
+   * Default Bengali font in Windows for example has intermixed
+   * lookups for init,pres,abvs,blws features.
+   */
+  {HB_TAG('i','n','i','t'),        F_MANUAL_JOINERS},
+  {HB_TAG('p','r','e','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('a','b','v','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('b','l','w','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('p','s','t','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('h','a','l','n'), F_GLOBAL_MANUAL_JOINERS},
+};
+
+/*
+ * Must be in the same order as the indic_features array.
+ */
+enum {
+  _INDIC_NUKT,
+  _INDIC_AKHN,
+  INDIC_RPHF,
+  _INDIC_RKRF,
+  INDIC_PREF,
+  INDIC_BLWF,
+  INDIC_ABVF,
+  INDIC_HALF,
+  INDIC_PSTF,
+  _INDIC_VATU,
+  _INDIC_CJCT,
+
+  INDIC_INIT,
+  _INDIC_PRES,
+  _INDIC_ABVS,
+  _INDIC_BLWS,
+  _INDIC_PSTS,
+  _INDIC_HALN,
+
+  INDIC_NUM_FEATURES,
+  INDIC_BASIC_FEATURES = INDIC_INIT, /* Don't forget to update this! */
+};
+
+static void
+setup_syllables_indic (const hb_ot_shape_plan_t *plan,
+		       hb_font_t *font,
+		       hb_buffer_t *buffer);
+static void
+initial_reordering_indic (const hb_ot_shape_plan_t *plan,
+			  hb_font_t *font,
+			  hb_buffer_t *buffer);
+static void
+final_reordering_indic (const hb_ot_shape_plan_t *plan,
+			hb_font_t *font,
+			hb_buffer_t *buffer);
+
+static void
+collect_features_indic (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* Do this before any lookups have been applied. */
+  map->add_gsub_pause (setup_syllables_indic);
+
+  map->enable_feature (HB_TAG('l','o','c','l'));
+  /* The Indic specs do not require ccmp, but we apply it here since if
+   * there is a use of it, it's typically at the beginning. */
+  map->enable_feature (HB_TAG('c','c','m','p'));
+
+
+  unsigned int i = 0;
+  map->add_gsub_pause (initial_reordering_indic);
+
+  for (; i < INDIC_BASIC_FEATURES; i++) {
+    map->add_feature (indic_features[i]);
+    map->add_gsub_pause (nullptr);
+  }
+
+  map->add_gsub_pause (final_reordering_indic);
+
+  for (; i < INDIC_NUM_FEATURES; i++)
+    map->add_feature (indic_features[i]);
+
+  map->enable_feature (HB_TAG('c','a','l','t'));
+  map->enable_feature (HB_TAG('c','l','i','g'));
+
+  map->add_gsub_pause (_hb_clear_syllables);
+}
+
+static void
+override_features_indic (hb_ot_shape_planner_t *plan)
+{
+  plan->map.disable_feature (HB_TAG('l','i','g','a'));
+}
+
+
+struct indic_shape_plan_t
+{
+  bool load_virama_glyph (hb_font_t *font, hb_codepoint_t *pglyph) const
+  {
+    hb_codepoint_t glyph = virama_glyph.get_relaxed ();
+    if (unlikely (glyph == (hb_codepoint_t) -1))
+    {
+      if (!config->virama || !font->get_nominal_glyph (config->virama, &glyph))
+	glyph = 0;
+      /* Technically speaking, the spec says we should apply 'locl' to virama too.
+       * Maybe one day... */
+
+      /* Our get_nominal_glyph() function needs a font, so we can't get the virama glyph
+       * during shape planning...  Instead, overwrite it here. */
+      virama_glyph.set_relaxed ((int) glyph);
+    }
+
+    *pglyph = glyph;
+    return glyph != 0;
+  }
+
+  const indic_config_t *config;
+
+  bool is_old_spec;
+#ifndef HB_NO_UNISCRIBE_BUG_COMPATIBLE
+  bool uniscribe_bug_compatible;
+#else
+  static constexpr bool uniscribe_bug_compatible = false;
+#endif
+  mutable hb_atomic_int_t virama_glyph;
+
+  hb_indic_would_substitute_feature_t rphf;
+  hb_indic_would_substitute_feature_t pref;
+  hb_indic_would_substitute_feature_t blwf;
+  hb_indic_would_substitute_feature_t pstf;
+  hb_indic_would_substitute_feature_t vatu;
+
+  hb_mask_t mask_array[INDIC_NUM_FEATURES];
+};
+
+static void *
+data_create_indic (const hb_ot_shape_plan_t *plan)
+{
+  indic_shape_plan_t *indic_plan = (indic_shape_plan_t *) calloc (1, sizeof (indic_shape_plan_t));
+  if (unlikely (!indic_plan))
+    return nullptr;
+
+  indic_plan->config = &indic_configs[0];
+  for (unsigned int i = 1; i < ARRAY_LENGTH (indic_configs); i++)
+    if (plan->props.script == indic_configs[i].script) {
+      indic_plan->config = &indic_configs[i];
+      break;
+    }
+
+  indic_plan->is_old_spec = indic_plan->config->has_old_spec && ((plan->map.chosen_script[0] & 0x000000FFu) != '2');
+#ifndef HB_NO_UNISCRIBE_BUG_COMPATIBLE
+  indic_plan->uniscribe_bug_compatible = hb_options ().uniscribe_bug_compatible;
+#endif
+  indic_plan->virama_glyph.set_relaxed (-1);
+
+  /* Use zero-context would_substitute() matching for new-spec of the main
+   * Indic scripts, and scripts with one spec only, but not for old-specs.
+   * The new-spec for all dual-spec scripts says zero-context matching happens.
+   *
+   * However, testing with Malayalam shows that old and new spec both allow
+   * context.  Testing with Bengali new-spec however shows that it doesn't.
+   * So, the heuristic here is the way it is.  It should *only* be changed,
+   * as we discover more cases of what Windows does.  DON'T TOUCH OTHERWISE.
+   */
+  bool zero_context = !indic_plan->is_old_spec && plan->props.script != HB_SCRIPT_MALAYALAM;
+  indic_plan->rphf.init (&plan->map, HB_TAG('r','p','h','f'), zero_context);
+  indic_plan->pref.init (&plan->map, HB_TAG('p','r','e','f'), zero_context);
+  indic_plan->blwf.init (&plan->map, HB_TAG('b','l','w','f'), zero_context);
+  indic_plan->pstf.init (&plan->map, HB_TAG('p','s','t','f'), zero_context);
+  indic_plan->vatu.init (&plan->map, HB_TAG('v','a','t','u'), zero_context);
+
+  for (unsigned int i = 0; i < ARRAY_LENGTH (indic_plan->mask_array); i++)
+    indic_plan->mask_array[i] = (indic_features[i].flags & F_GLOBAL) ?
+				 0 : plan->map.get_1_mask (indic_features[i].tag);
+
+  return indic_plan;
+}
+
+static void
+data_destroy_indic (void *data)
+{
+  free (data);
+}
+
+static indic_position_t
+consonant_position_from_face (const indic_shape_plan_t *indic_plan,
+			      const hb_codepoint_t consonant,
+			      const hb_codepoint_t virama,
+			      hb_face_t *face)
+{
+  /* For old-spec, the order of glyphs is Consonant,Virama,
+   * whereas for new-spec, it's Virama,Consonant.  However,
+   * some broken fonts (like Free Sans) simply copied lookups
+   * from old-spec to new-spec without modification.
+   * And oddly enough, Uniscribe seems to respect those lookups.
+   * Eg. in the sequence U+0924,U+094D,U+0930, Uniscribe finds
+   * base at 0.  The font however, only has lookups matching
+   * 930,94D in 'blwf', not the expected 94D,930 (with new-spec
+   * table).  As such, we simply match both sequences.  Seems
+   * to work.
+   *
+   * Vatu is done as well, for:
+   * https://github.com/harfbuzz/harfbuzz/issues/1587
+   */
+  hb_codepoint_t glyphs[3] = {virama, consonant, virama};
+  if (indic_plan->blwf.would_substitute (glyphs  , 2, face) ||
+      indic_plan->blwf.would_substitute (glyphs+1, 2, face) ||
+      indic_plan->vatu.would_substitute (glyphs  , 2, face) ||
+      indic_plan->vatu.would_substitute (glyphs+1, 2, face))
+    return POS_BELOW_C;
+  if (indic_plan->pstf.would_substitute (glyphs  , 2, face) ||
+      indic_plan->pstf.would_substitute (glyphs+1, 2, face))
+    return POS_POST_C;
+  if (indic_plan->pref.would_substitute (glyphs  , 2, face) ||
+      indic_plan->pref.would_substitute (glyphs+1, 2, face))
+    return POS_POST_C;
+  return POS_BASE_C;
+}
+
+
+enum indic_syllable_type_t {
+  indic_consonant_syllable,
+  indic_vowel_syllable,
+  indic_standalone_cluster,
+  indic_symbol_cluster,
+  indic_broken_cluster,
+  indic_non_indic_cluster,
+};
+
+#include "hb-ot-shape-complex-indic-machine.hh"
+
+
+static void
+setup_masks_indic (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		   hb_buffer_t              *buffer,
+		   hb_font_t                *font HB_UNUSED)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, indic_category);
+  HB_BUFFER_ALLOCATE_VAR (buffer, indic_position);
+
+  /* We cannot setup masks here.  We save information about characters
+   * and setup masks later on in a pause-callback. */
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    set_indic_properties (info[i]);
+}
+
+static void
+setup_syllables_indic (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		       hb_font_t *font HB_UNUSED,
+		       hb_buffer_t *buffer)
+{
+  find_syllables_indic (buffer);
+  foreach_syllable (buffer, start, end)
+    buffer->unsafe_to_break (start, end);
+}
+
+static int
+compare_indic_order (const hb_glyph_info_t *pa, const hb_glyph_info_t *pb)
+{
+  int a = pa->indic_position();
+  int b = pb->indic_position();
+
+  return a < b ? -1 : a == b ? 0 : +1;
+}
+
+
+
+static void
+update_consonant_positions_indic (const hb_ot_shape_plan_t *plan,
+				  hb_font_t         *font,
+				  hb_buffer_t       *buffer)
+{
+  const indic_shape_plan_t *indic_plan = (const indic_shape_plan_t *) plan->data;
+
+  if (indic_plan->config->base_pos != BASE_POS_LAST)
+    return;
+
+  hb_codepoint_t virama;
+  if (indic_plan->load_virama_glyph (font, &virama))
+  {
+    hb_face_t *face = font->face;
+    unsigned int count = buffer->len;
+    hb_glyph_info_t *info = buffer->info;
+    for (unsigned int i = 0; i < count; i++)
+      if (info[i].indic_position() == POS_BASE_C)
+      {
+	hb_codepoint_t consonant = info[i].codepoint;
+	info[i].indic_position() = consonant_position_from_face (indic_plan, consonant, virama, face);
+      }
+  }
+}
+
+
+/* Rules from:
+ * https://docs.microsqoft.com/en-us/typography/script-development/devanagari */
+
+static void
+initial_reordering_consonant_syllable (const hb_ot_shape_plan_t *plan,
+				       hb_face_t *face,
+				       hb_buffer_t *buffer,
+				       unsigned int start, unsigned int end)
+{
+  const indic_shape_plan_t *indic_plan = (const indic_shape_plan_t *) plan->data;
+  hb_glyph_info_t *info = buffer->info;
+
+  /* https://github.com/harfbuzz/harfbuzz/issues/435#issuecomment-335560167
+   * // For compatibility with legacy usage in Kannada,
+   * // Ra+h+ZWJ must behave like Ra+ZWJ+h...
+   */
+  if (buffer->props.script == HB_SCRIPT_KANNADA &&
+      start + 3 <= end &&
+      is_one_of (info[start  ], FLAG (OT_Ra)) &&
+      is_one_of (info[start+1], FLAG (OT_H)) &&
+      is_one_of (info[start+2], FLAG (OT_ZWJ)))
+  {
+    buffer->merge_clusters (start+1, start+3);
+    hb_glyph_info_t tmp = info[start+1];
+    info[start+1] = info[start+2];
+    info[start+2] = tmp;
+  }
+
+  /* 1. Find base consonant:
+   *
+   * The shaping engine finds the base consonant of the syllable, using the
+   * following algorithm: starting from the end of the syllable, move backwards
+   * until a consonant is found that does not have a below-base or post-base
+   * form (post-base forms have to follow below-base forms), or that is not a
+   * pre-base-reordering Ra, or arrive at the first consonant. The consonant
+   * stopped at will be the base.
+   *
+   *   o If the syllable starts with Ra + Halant (in a script that has Reph)
+   *     and has more than one consonant, Ra is excluded from candidates for
+   *     base consonants.
+   */
+
+  unsigned int base = end;
+  bool has_reph = false;
+
+  {
+    /* -> If the syllable starts with Ra + Halant (in a script that has Reph)
+     *    and has more than one consonant, Ra is excluded from candidates for
+     *    base consonants. */
+    unsigned int limit = start;
+    if (indic_plan->mask_array[INDIC_RPHF] &&
+	start + 3 <= end &&
+	(
+	 (indic_plan->config->reph_mode == REPH_MODE_IMPLICIT && !is_joiner (info[start + 2])) ||
+	 (indic_plan->config->reph_mode == REPH_MODE_EXPLICIT && info[start + 2].indic_category() == OT_ZWJ)
+	))
+    {
+      /* See if it matches the 'rphf' feature. */
+      hb_codepoint_t glyphs[3] = {info[start].codepoint,
+				  info[start + 1].codepoint,
+				  indic_plan->config->reph_mode == REPH_MODE_EXPLICIT ?
+				    info[start + 2].codepoint : 0};
+      if (indic_plan->rphf.would_substitute (glyphs, 2, face) ||
+	  (indic_plan->config->reph_mode == REPH_MODE_EXPLICIT &&
+	   indic_plan->rphf.would_substitute (glyphs, 3, face)))
+      {
+	limit += 2;
+	while (limit < end && is_joiner (info[limit]))
+	  limit++;
+	base = start;
+	has_reph = true;
+      }
+    } else if (indic_plan->config->reph_mode == REPH_MODE_LOG_REPHA && info[start].indic_category() == OT_Repha)
+    {
+	limit += 1;
+	while (limit < end && is_joiner (info[limit]))
+	  limit++;
+	base = start;
+	has_reph = true;
+    }
+
+    switch (indic_plan->config->base_pos)
+    {
+      case BASE_POS_LAST:
+      {
+	/* -> starting from the end of the syllable, move backwards */
+	unsigned int i = end;
+	bool seen_below = false;
+	do {
+	  i--;
+	  /* -> until a consonant is found */
+	  if (is_consonant (info[i]))
+	  {
+	    /* -> that does not have a below-base or post-base form
+	     * (post-base forms have to follow below-base forms), */
+	    if (info[i].indic_position() != POS_BELOW_C &&
+		(info[i].indic_position() != POS_POST_C || seen_below))
+	    {
+	      base = i;
+	      break;
+	    }
+	    if (info[i].indic_position() == POS_BELOW_C)
+	      seen_below = true;
+
+	    /* -> or that is not a pre-base-reordering Ra,
+	     *
+	     * IMPLEMENTATION NOTES:
+	     *
+	     * Our pre-base-reordering Ra's are marked POS_POST_C, so will be skipped
+	     * by the logic above already.
+	     */
+
+	    /* -> or arrive at the first consonant. The consonant stopped at will
+	     * be the base. */
+	    base = i;
+	  }
+	  else
+	  {
+	    /* A ZWJ after a Halant stops the base search, and requests an explicit
+	     * half form.
+	     * A ZWJ before a Halant, requests a subjoined form instead, and hence
+	     * search continues.  This is particularly important for Bengali
+	     * sequence Ra,H,Ya that should form Ya-Phalaa by subjoining Ya. */
+	    if (start < i &&
+		info[i].indic_category() == OT_ZWJ &&
+		info[i - 1].indic_category() == OT_H)
+	      break;
+	  }
+	} while (i > limit);
+      }
+      break;
+
+      case BASE_POS_LAST_SINHALA:
+      {
+	/* Sinhala base positioning is slightly different from main Indic, in that:
+	 * 1. Its ZWJ behavior is different,
+	 * 2. We don't need to look into the font for consonant positions.
+	 */
+
+	if (!has_reph)
+	  base = limit;
+
+	/* Find the last base consonant that is not blocked by ZWJ.  If there is
+	 * a ZWJ right before a base consonant, that would request a subjoined form. */
+	for (unsigned int i = limit; i < end; i++)
+	  if (is_consonant (info[i]))
+	  {
+	    if (limit < i && info[i - 1].indic_category() == OT_ZWJ)
+	      break;
+	    else
+	      base = i;
+	  }
+
+	/* Mark all subsequent consonants as below. */
+	for (unsigned int i = base + 1; i < end; i++)
+	  if (is_consonant (info[i]))
+	    info[i].indic_position() = POS_BELOW_C;
+      }
+      break;
+    }
+
+    /* -> If the syllable starts with Ra + Halant (in a script that has Reph)
+     *    and has more than one consonant, Ra is excluded from candidates for
+     *    base consonants.
+     *
+     *  Only do this for unforced Reph. (ie. not for Ra,H,ZWJ. */
+    if (has_reph && base == start && limit - base <= 2) {
+      /* Have no other consonant, so Reph is not formed and Ra becomes base. */
+      has_reph = false;
+    }
+  }
+
+
+  /* 2. Decompose and reorder Matras:
+   *
+   * Each matra and any syllable modifier sign in the syllable are moved to the
+   * appropriate position relative to the consonant(s) in the syllable. The
+   * shaping engine decomposes two- or three-part matras into their constituent
+   * parts before any repositioning. Matra characters are classified by which
+   * consonant in a conjunct they have affinity for and are reordered to the
+   * following positions:
+   *
+   *   o Before first half form in the syllable
+   *   o After subjoined consonants
+   *   o After post-form consonant
+   *   o After main consonant (for above marks)
+   *
+   * IMPLEMENTATION NOTES:
+   *
+   * The normalize() routine has already decomposed matras for us, so we don't
+   * need to worry about that.
+   */
+
+
+  /* 3.  Reorder marks to canonical order:
+   *
+   * Adjacent nukta and halant or nukta and vedic sign are always repositioned
+   * if necessary, so that the nukta is first.
+   *
+   * IMPLEMENTATION NOTES:
+   *
+   * We don't need to do this: the normalize() routine already did this for us.
+   */
+
+
+  /* Reorder characters */
+
+  for (unsigned int i = start; i < base; i++)
+    info[i].indic_position() = hb_min (POS_PRE_C, (indic_position_t) info[i].indic_position());
+
+  if (base < end)
+    info[base].indic_position() = POS_BASE_C;
+
+  /* Mark final consonants.  A final consonant is one appearing after a matra.
+   * Happens in Sinhala. */
+  for (unsigned int i = base + 1; i < end; i++)
+    if (info[i].indic_category() == OT_M) {
+      for (unsigned int j = i + 1; j < end; j++)
+	if (is_consonant (info[j])) {
+	 info[j].indic_position() = POS_FINAL_C;
+	 break;
+       }
+      break;
+    }
+
+  /* Handle beginning Ra */
+  if (has_reph)
+    info[start].indic_position() = POS_RA_TO_BECOME_REPH;
+
+  /* For old-style Indic script tags, move the first post-base Halant after
+   * last consonant.
+   *
+   * Reports suggest that in some scripts Uniscribe does this only if there
+   * is *not* a Halant after last consonant already.  We know that is the
+   * case for Kannada, while it reorders unconditionally in other scripts,
+   * eg. Malayalam, Bengali, and Devanagari.  We don't currently know about
+   * other scripts, so we block Kannada.
+   *
+   * Kannada test case:
+   * U+0C9A,U+0CCD,U+0C9A,U+0CCD
+   * With some versions of Lohit Kannada.
+   * https://bugs.freedesktop.org/show_bug.cgi?id=59118
+   *
+   * Malayalam test case:
+   * U+0D38,U+0D4D,U+0D31,U+0D4D,U+0D31,U+0D4D
+   * With lohit-ttf-20121122/Lohit-Malayalam.ttf
+   *
+   * Bengali test case:
+   * U+0998,U+09CD,U+09AF,U+09CD
+   * With Windows XP vrinda.ttf
+   * https://github.com/harfbuzz/harfbuzz/issues/1073
+   *
+   * Devanagari test case:
+   * U+091F,U+094D,U+0930,U+094D
+   * With chandas.ttf
+   * https://github.com/harfbuzz/harfbuzz/issues/1071
+   */
+  if (indic_plan->is_old_spec)
+  {
+    bool disallow_double_halants = buffer->props.script == HB_SCRIPT_KANNADA;
+    for (unsigned int i = base + 1; i < end; i++)
+      if (info[i].indic_category() == OT_H)
+      {
+	unsigned int j;
+	for (j = end - 1; j > i; j--)
+	  if (is_consonant (info[j]) ||
+	      (disallow_double_halants && info[j].indic_category() == OT_H))
+	    break;
+	if (info[j].indic_category() != OT_H && j > i) {
+	  /* Move Halant to after last consonant. */
+	  hb_glyph_info_t t = info[i];
+	  memmove (&info[i], &info[i + 1], (j - i) * sizeof (info[0]));
+	  info[j] = t;
+	}
+	break;
+      }
+  }
+
+  /* Attach misc marks to previous char to move with them. */
+  {
+    indic_position_t last_pos = POS_START;
+    for (unsigned int i = start; i < end; i++)
+    {
+      if ((FLAG_UNSAFE (info[i].indic_category()) & (JOINER_FLAGS | FLAG (OT_N) | FLAG (OT_RS) | MEDIAL_FLAGS | FLAG (OT_H))))
+      {
+	info[i].indic_position() = last_pos;
+	if (unlikely (info[i].indic_category() == OT_H &&
+		      info[i].indic_position() == POS_PRE_M))
+	{
+	  /*
+	   * Uniscribe doesn't move the Halant with Left Matra.
+	   * TEST: U+092B,U+093F,U+094DE
+	   * We follow.  This is important for the Sinhala
+	   * U+0DDA split matra since it decomposes to U+0DD9,U+0DCA
+	   * where U+0DD9 is a left matra and U+0DCA is the virama.
+	   * We don't want to move the virama with the left matra.
+	   * TEST: U+0D9A,U+0DDA
+	   */
+	  for (unsigned int j = i; j > start; j--)
+	    if (info[j - 1].indic_position() != POS_PRE_M) {
+	      info[i].indic_position() = info[j - 1].indic_position();
+	      break;
+	    }
+	}
+      } else if (info[i].indic_position() != POS_SMVD) {
+	last_pos = (indic_position_t) info[i].indic_position();
+      }
+    }
+  }
+  /* For post-base consonants let them own anything before them
+   * since the last consonant or matra. */
+  {
+    unsigned int last = base;
+    for (unsigned int i = base + 1; i < end; i++)
+      if (is_consonant (info[i]))
+      {
+	for (unsigned int j = last + 1; j < i; j++)
+	  if (info[j].indic_position() < POS_SMVD)
+	    info[j].indic_position() = info[i].indic_position();
+	last = i;
+      } else if (info[i].indic_category() == OT_M)
+	last = i;
+  }
+
+
+  {
+    /* Use syllable() for sort accounting temporarily. */
+    unsigned int syllable = info[start].syllable();
+    for (unsigned int i = start; i < end; i++)
+      info[i].syllable() = i - start;
+
+    /* Sit tight, rock 'n roll! */
+    hb_stable_sort (info + start, end - start, compare_indic_order);
+    /* Find base again */
+    base = end;
+    for (unsigned int i = start; i < end; i++)
+      if (info[i].indic_position() == POS_BASE_C)
+      {
+	base = i;
+	break;
+      }
+    /* Things are out-of-control for post base positions, they may shuffle
+     * around like crazy.  In old-spec mode, we move halants around, so in
+     * that case merge all clusters after base.  Otherwise, check the sort
+     * order and merge as needed.
+     * For pre-base stuff, we handle cluster issues in final reordering.
+     *
+     * We could use buffer->sort() for this, if there was no special
+     * reordering of pre-base stuff happening later...
+     * We don't want to merge_clusters all of that, which buffer->sort()
+     * would.
+     */
+    if (indic_plan->is_old_spec || end - start > 127)
+      buffer->merge_clusters (base, end);
+    else
+    {
+      /* Note!  syllable() is a one-byte field. */
+      for (unsigned int i = base; i < end; i++)
+	if (info[i].syllable() != 255)
+	{
+	  unsigned int max = i;
+	  unsigned int j = start + info[i].syllable();
+	  while (j != i)
+	  {
+	    max = hb_max (max, j);
+	    unsigned int next = start + info[j].syllable();
+	    info[j].syllable() = 255; /* So we don't process j later again. */
+	    j = next;
+	  }
+	  if (i != max)
+	    buffer->merge_clusters (i, max + 1);
+	}
+    }
+
+    /* Put syllable back in. */
+    for (unsigned int i = start; i < end; i++)
+      info[i].syllable() = syllable;
+  }
+
+  /* Setup masks now */
+
+  {
+    hb_mask_t mask;
+
+    /* Reph */
+    for (unsigned int i = start; i < end && info[i].indic_position() == POS_RA_TO_BECOME_REPH; i++)
+      info[i].mask |= indic_plan->mask_array[INDIC_RPHF];
+
+    /* Pre-base */
+    mask = indic_plan->mask_array[INDIC_HALF];
+    if (!indic_plan->is_old_spec &&
+	indic_plan->config->blwf_mode == BLWF_MODE_PRE_AND_POST)
+      mask |= indic_plan->mask_array[INDIC_BLWF];
+    for (unsigned int i = start; i < base; i++)
+      info[i].mask  |= mask;
+    /* Base */
+    mask = 0;
+    if (base < end)
+      info[base].mask |= mask;
+    /* Post-base */
+    mask = indic_plan->mask_array[INDIC_BLWF] |
+	   indic_plan->mask_array[INDIC_ABVF] |
+	   indic_plan->mask_array[INDIC_PSTF];
+    for (unsigned int i = base + 1; i < end; i++)
+      info[i].mask  |= mask;
+  }
+
+  if (indic_plan->is_old_spec &&
+      buffer->props.script == HB_SCRIPT_DEVANAGARI)
+  {
+    /* Old-spec eye-lash Ra needs special handling.  From the
+     * spec:
+     *
+     * "The feature 'below-base form' is applied to consonants
+     * having below-base forms and following the base consonant.
+     * The exception is vattu, which may appear below half forms
+     * as well as below the base glyph. The feature 'below-base
+     * form' will be applied to all such occurrences of Ra as well."
+     *
+     * Test case: U+0924,U+094D,U+0930,U+094d,U+0915
+     * with Sanskrit 2003 font.
+     *
+     * However, note that Ra,Halant,ZWJ is the correct way to
+     * request eyelash form of Ra, so we wouldbn't inhibit it
+     * in that sequence.
+     *
+     * Test case: U+0924,U+094D,U+0930,U+094d,U+200D,U+0915
+     */
+    for (unsigned int i = start; i + 1 < base; i++)
+      if (info[i  ].indic_category() == OT_Ra &&
+	  info[i+1].indic_category() == OT_H  &&
+	  (i + 2 == base ||
+	   info[i+2].indic_category() != OT_ZWJ))
+      {
+	info[i  ].mask |= indic_plan->mask_array[INDIC_BLWF];
+	info[i+1].mask |= indic_plan->mask_array[INDIC_BLWF];
+      }
+  }
+
+  unsigned int pref_len = 2;
+  if (indic_plan->mask_array[INDIC_PREF] && base + pref_len < end)
+  {
+    /* Find a Halant,Ra sequence and mark it for pre-base-reordering processing. */
+    for (unsigned int i = base + 1; i + pref_len - 1 < end; i++) {
+      hb_codepoint_t glyphs[2];
+      for (unsigned int j = 0; j < pref_len; j++)
+	glyphs[j] = info[i + j].codepoint;
+      if (indic_plan->pref.would_substitute (glyphs, pref_len, face))
+      {
+	for (unsigned int j = 0; j < pref_len; j++)
+	  info[i++].mask |= indic_plan->mask_array[INDIC_PREF];
+	break;
+      }
+    }
+  }
+
+  /* Apply ZWJ/ZWNJ effects */
+  for (unsigned int i = start + 1; i < end; i++)
+    if (is_joiner (info[i])) {
+      bool non_joiner = info[i].indic_category() == OT_ZWNJ;
+      unsigned int j = i;
+
+      do {
+	j--;
+
+	/* ZWJ/ZWNJ should disable CJCT.  They do that by simply
+	 * being there, since we don't skip them for the CJCT
+	 * feature (ie. F_MANUAL_ZWJ) */
+
+	/* A ZWNJ disables HALF. */
+	if (non_joiner)
+	  info[j].mask &= ~indic_plan->mask_array[INDIC_HALF];
+
+      } while (j > start && !is_consonant (info[j]));
+    }
+}
+
+static void
+initial_reordering_standalone_cluster (const hb_ot_shape_plan_t *plan,
+				       hb_face_t *face,
+				       hb_buffer_t *buffer,
+				       unsigned int start, unsigned int end)
+{
+  /* We treat placeholder/dotted-circle as if they are consonants, so we
+   * should just chain.  Only if not in compatibility mode that is... */
+
+  const indic_shape_plan_t *indic_plan = (const indic_shape_plan_t *) plan->data;
+  if (indic_plan->uniscribe_bug_compatible)
+  {
+    /* For dotted-circle, this is what Uniscribe does:
+     * If dotted-circle is the last glyph, it just does nothing.
+     * Ie. It doesn't form Reph. */
+    if (buffer->info[end - 1].indic_category() == OT_DOTTEDCIRCLE)
+      return;
+  }
+
+  initial_reordering_consonant_syllable (plan, face, buffer, start, end);
+}
+
+static void
+initial_reordering_syllable_indic (const hb_ot_shape_plan_t *plan,
+				   hb_face_t *face,
+				   hb_buffer_t *buffer,
+				   unsigned int start, unsigned int end)
+{
+  indic_syllable_type_t syllable_type = (indic_syllable_type_t) (buffer->info[start].syllable() & 0x0F);
+  switch (syllable_type)
+  {
+    case indic_vowel_syllable: /* We made the vowels look like consonants.  So let's call the consonant logic! */
+    case indic_consonant_syllable:
+     initial_reordering_consonant_syllable (plan, face, buffer, start, end);
+     break;
+
+    case indic_broken_cluster: /* We already inserted dotted-circles, so just call the standalone_cluster. */
+    case indic_standalone_cluster:
+     initial_reordering_standalone_cluster (plan, face, buffer, start, end);
+     break;
+
+    case indic_symbol_cluster:
+    case indic_non_indic_cluster:
+      break;
+  }
+}
+
+static inline void
+insert_dotted_circles_indic (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			     hb_font_t *font,
+			     hb_buffer_t *buffer)
+{
+  if (unlikely (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE))
+    return;
+
+  /* Note: This loop is extra overhead, but should not be measurable.
+   * TODO Use a buffer scratch flag to remove the loop. */
+  bool has_broken_syllables = false;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if ((info[i].syllable() & 0x0F) == indic_broken_cluster)
+    {
+      has_broken_syllables = true;
+      break;
+    }
+  if (likely (!has_broken_syllables))
+    return;
+
+
+  hb_codepoint_t dottedcircle_glyph;
+  if (!font->get_nominal_glyph (0x25CCu, &dottedcircle_glyph))
+    return;
+
+  hb_glyph_info_t dottedcircle = {0};
+  dottedcircle.codepoint = 0x25CCu;
+  set_indic_properties (dottedcircle);
+  dottedcircle.codepoint = dottedcircle_glyph;
+
+  buffer->clear_output ();
+
+  buffer->idx = 0;
+  unsigned int last_syllable = 0;
+  while (buffer->idx < buffer->len && buffer->successful)
+  {
+    unsigned int syllable = buffer->cur().syllable();
+    indic_syllable_type_t syllable_type = (indic_syllable_type_t) (syllable & 0x0F);
+    if (unlikely (last_syllable != syllable && syllable_type == indic_broken_cluster))
+    {
+      last_syllable = syllable;
+
+      hb_glyph_info_t ginfo = dottedcircle;
+      ginfo.cluster = buffer->cur().cluster;
+      ginfo.mask = buffer->cur().mask;
+      ginfo.syllable() = buffer->cur().syllable();
+
+      /* Insert dottedcircle after possible Repha. */
+      while (buffer->idx < buffer->len && buffer->successful &&
+	     last_syllable == buffer->cur().syllable() &&
+	     buffer->cur().indic_category() == OT_Repha)
+	buffer->next_glyph ();
+
+      buffer->output_info (ginfo);
+    }
+    else
+      buffer->next_glyph ();
+  }
+  buffer->swap_buffers ();
+}
+
+static void
+initial_reordering_indic (const hb_ot_shape_plan_t *plan,
+			  hb_font_t *font,
+			  hb_buffer_t *buffer)
+{
+  update_consonant_positions_indic (plan, font, buffer);
+  insert_dotted_circles_indic (plan, font, buffer);
+
+  foreach_syllable (buffer, start, end)
+    initial_reordering_syllable_indic (plan, font->face, buffer, start, end);
+}
+
+static void
+final_reordering_syllable_indic (const hb_ot_shape_plan_t *plan,
+				 hb_buffer_t *buffer,
+				 unsigned int start, unsigned int end)
+{
+  const indic_shape_plan_t *indic_plan = (const indic_shape_plan_t *) plan->data;
+  hb_glyph_info_t *info = buffer->info;
+
+
+  /* This function relies heavily on halant glyphs.  Lots of ligation
+   * and possibly multiple substitutions happened prior to this
+   * phase, and that might have messed up our properties.  Recover
+   * from a particular case of that where we're fairly sure that a
+   * class of OT_H is desired but has been lost. */
+  /* We don't call load_virama_glyph(), since we know it's already
+   * loaded. */
+  hb_codepoint_t virama_glyph = indic_plan->virama_glyph.get_relaxed ();
+  if (virama_glyph)
+  {
+    for (unsigned int i = start; i < end; i++)
+      if (info[i].codepoint == virama_glyph &&
+	  _hb_glyph_info_ligated (&info[i]) &&
+	  _hb_glyph_info_multiplied (&info[i]))
+      {
+	/* This will make sure that this glyph passes is_halant() test. */
+	info[i].indic_category() = OT_H;
+	_hb_glyph_info_clear_ligated_and_multiplied (&info[i]);
+      }
+  }
+
+
+  /* 4. Final reordering:
+   *
+   * After the localized forms and basic shaping forms GSUB features have been
+   * applied (see below), the shaping engine performs some final glyph
+   * reordering before applying all the remaining font features to the entire
+   * syllable.
+   */
+
+  bool try_pref = !!indic_plan->mask_array[INDIC_PREF];
+
+  /* Find base again */
+  unsigned int base;
+  for (base = start; base < end; base++)
+    if (info[base].indic_position() >= POS_BASE_C)
+    {
+      if (try_pref && base + 1 < end)
+      {
+	for (unsigned int i = base + 1; i < end; i++)
+	  if ((info[i].mask & indic_plan->mask_array[INDIC_PREF]) != 0)
+	  {
+	    if (!(_hb_glyph_info_substituted (&info[i]) &&
+		  _hb_glyph_info_ligated_and_didnt_multiply (&info[i])))
+	    {
+	      /* Ok, this was a 'pref' candidate but didn't form any.
+	       * Base is around here... */
+	      base = i;
+	      while (base < end && is_halant (info[base]))
+		base++;
+	      info[base].indic_position() = POS_BASE_C;
+
+	      try_pref = false;
+	    }
+	    break;
+	  }
+      }
+      /* For Malayalam, skip over unformed below- (but NOT post-) forms. */
+      if (buffer->props.script == HB_SCRIPT_MALAYALAM)
+      {
+	for (unsigned int i = base + 1; i < end; i++)
+	{
+	  while (i < end && is_joiner (info[i]))
+	    i++;
+	  if (i == end || !is_halant (info[i]))
+	    break;
+	  i++; /* Skip halant. */
+	  while (i < end && is_joiner (info[i]))
+	    i++;
+	  if (i < end && is_consonant (info[i]) && info[i].indic_position() == POS_BELOW_C)
+	  {
+	    base = i;
+	    info[base].indic_position() = POS_BASE_C;
+	  }
+	}
+      }
+
+      if (start < base && info[base].indic_position() > POS_BASE_C)
+	base--;
+      break;
+    }
+  if (base == end && start < base &&
+      is_one_of (info[base - 1], FLAG (OT_ZWJ)))
+    base--;
+  if (base < end)
+    while (start < base &&
+	   is_one_of (info[base], (FLAG (OT_N) | FLAG (OT_H))))
+      base--;
+
+
+  /*   o Reorder matras:
+   *
+   *     If a pre-base matra character had been reordered before applying basic
+   *     features, the glyph can be moved closer to the main consonant based on
+   *     whether half-forms had been formed. Actual position for the matra is
+   *     defined as “after last standalone halant glyph, after initial matra
+   *     position and before the main consonant”. If ZWJ or ZWNJ follow this
+   *     halant, position is moved after it.
+   *
+   * IMPLEMENTATION NOTES:
+   *
+   * It looks like the last sentence is wrong.  Testing, with Windows 7 Uniscribe
+   * and Devanagari shows that the behavior is best described as:
+   *
+   * "If ZWJ follows this halant, matra is NOT repositioned after this halant.
+   *  If ZWNJ follows this halant, position is moved after it."
+   *
+   * Test case, with Adobe Devanagari or Nirmala UI:
+   *
+   *   U+091F,U+094D,U+200C,U+092F,U+093F
+   *   (Matra moves to the middle, after ZWNJ.)
+   *
+   *   U+091F,U+094D,U+200D,U+092F,U+093F
+   *   (Matra does NOT move, stays to the left.)
+   *
+   * https://github.com/harfbuzz/harfbuzz/issues/1070
+   */
+
+  if (start + 1 < end && start < base) /* Otherwise there can't be any pre-base matra characters. */
+  {
+    /* If we lost track of base, alas, position before last thingy. */
+    unsigned int new_pos = base == end ? base - 2 : base - 1;
+
+    /* Malayalam / Tamil do not have "half" forms or explicit virama forms.
+     * The glyphs formed by 'half' are Chillus or ligated explicit viramas.
+     * We want to position matra after them.
+     */
+    if (buffer->props.script != HB_SCRIPT_MALAYALAM && buffer->props.script != HB_SCRIPT_TAMIL)
+    {
+    search:
+      while (new_pos > start &&
+	     !(is_one_of (info[new_pos], (FLAG (OT_M) | FLAG (OT_H)))))
+	new_pos--;
+
+      /* If we found no Halant we are done.
+       * Otherwise only proceed if the Halant does
+       * not belong to the Matra itself! */
+      if (is_halant (info[new_pos]) &&
+	  info[new_pos].indic_position() != POS_PRE_M)
+      {
+#if 0 // See comment above
+	/* -> If ZWJ or ZWNJ follow this halant, position is moved after it. */
+	if (new_pos + 1 < end && is_joiner (info[new_pos + 1]))
+	  new_pos++;
+#endif
+	if (new_pos + 1 < end)
+	{
+	  /* -> If ZWJ follows this halant, matra is NOT repositioned after this halant. */
+	  if (info[new_pos + 1].indic_category() == OT_ZWJ)
+	  {
+	    /* Keep searching. */
+	    if (new_pos > start)
+	    {
+	      new_pos--;
+	      goto search;
+	    }
+	  }
+	  /* -> If ZWNJ follows this halant, position is moved after it.
+	   *
+	   * IMPLEMENTATION NOTES:
+	   *
+	   * This is taken care of by the state-machine. A Halant,ZWNJ is a terminating
+	   * sequence for a consonant syllable; any pre-base matras occurring after it
+	   * will belong to the subsequent syllable.
+	   */
+	}
+      }
+      else
+	new_pos = start; /* No move. */
+    }
+
+    if (start < new_pos && info[new_pos].indic_position () != POS_PRE_M)
+    {
+      /* Now go see if there's actually any matras... */
+      for (unsigned int i = new_pos; i > start; i--)
+	if (info[i - 1].indic_position () == POS_PRE_M)
+	{
+	  unsigned int old_pos = i - 1;
+	  if (old_pos < base && base <= new_pos) /* Shouldn't actually happen. */
+	    base--;
+
+	  hb_glyph_info_t tmp = info[old_pos];
+	  memmove (&info[old_pos], &info[old_pos + 1], (new_pos - old_pos) * sizeof (info[0]));
+	  info[new_pos] = tmp;
+
+	  /* Note: this merge_clusters() is intentionally *after* the reordering.
+	   * Indic matra reordering is special and tricky... */
+	  buffer->merge_clusters (new_pos, hb_min (end, base + 1));
+
+	  new_pos--;
+	}
+    } else {
+      for (unsigned int i = start; i < base; i++)
+	if (info[i].indic_position () == POS_PRE_M) {
+	  buffer->merge_clusters (i, hb_min (end, base + 1));
+	  break;
+	}
+    }
+  }
+
+
+  /*   o Reorder reph:
+   *
+   *     Reph’s original position is always at the beginning of the syllable,
+   *     (i.e. it is not reordered at the character reordering stage). However,
+   *     it will be reordered according to the basic-forms shaping results.
+   *     Possible positions for reph, depending on the script, are; after main,
+   *     before post-base consonant forms, and after post-base consonant forms.
+   */
+
+  /* Two cases:
+   *
+   * - If repha is encoded as a sequence of characters (Ra,H or Ra,H,ZWJ), then
+   *   we should only move it if the sequence ligated to the repha form.
+   *
+   * - If repha is encoded separately and in the logical position, we should only
+   *   move it if it did NOT ligate.  If it ligated, it's probably the font trying
+   *   to make it work without the reordering.
+   */
+  if (start + 1 < end &&
+      info[start].indic_position() == POS_RA_TO_BECOME_REPH &&
+      ((info[start].indic_category() == OT_Repha) ^
+       _hb_glyph_info_ligated_and_didnt_multiply (&info[start])))
+  {
+    unsigned int new_reph_pos;
+    reph_position_t reph_pos = indic_plan->config->reph_pos;
+
+    /*       1. If reph should be positioned after post-base consonant forms,
+     *          proceed to step 5.
+     */
+    if (reph_pos == REPH_POS_AFTER_POST)
+    {
+      goto reph_step_5;
+    }
+
+    /*       2. If the reph repositioning class is not after post-base: target
+     *          position is after the first explicit halant glyph between the
+     *          first post-reph consonant and last main consonant. If ZWJ or ZWNJ
+     *          are following this halant, position is moved after it. If such
+     *          position is found, this is the target position. Otherwise,
+     *          proceed to the next step.
+     *
+     *          Note: in old-implementation fonts, where classifications were
+     *          fixed in shaping engine, there was no case where reph position
+     *          will be found on this step.
+     */
+    {
+      new_reph_pos = start + 1;
+      while (new_reph_pos < base && !is_halant (info[new_reph_pos]))
+	new_reph_pos++;
+
+      if (new_reph_pos < base && is_halant (info[new_reph_pos]))
+      {
+	/* ->If ZWJ or ZWNJ are following this halant, position is moved after it. */
+	if (new_reph_pos + 1 < base && is_joiner (info[new_reph_pos + 1]))
+	  new_reph_pos++;
+	goto reph_move;
+      }
+    }
+
+    /*       3. If reph should be repositioned after the main consonant: find the
+     *          first consonant not ligated with main, or find the first
+     *          consonant that is not a potential pre-base-reordering Ra.
+     */
+    if (reph_pos == REPH_POS_AFTER_MAIN)
+    {
+      new_reph_pos = base;
+      while (new_reph_pos + 1 < end && info[new_reph_pos + 1].indic_position() <= POS_AFTER_MAIN)
+	new_reph_pos++;
+      if (new_reph_pos < end)
+	goto reph_move;
+    }
+
+    /*       4. If reph should be positioned before post-base consonant, find
+     *          first post-base classified consonant not ligated with main. If no
+     *          consonant is found, the target position should be before the
+     *          first matra, syllable modifier sign or vedic sign.
+     */
+    /* This is our take on what step 4 is trying to say (and failing, BADLY). */
+    if (reph_pos == REPH_POS_AFTER_SUB)
+    {
+      new_reph_pos = base;
+      while (new_reph_pos + 1 < end &&
+	     !( FLAG_UNSAFE (info[new_reph_pos + 1].indic_position()) & (FLAG (POS_POST_C) | FLAG (POS_AFTER_POST) | FLAG (POS_SMVD))))
+	new_reph_pos++;
+      if (new_reph_pos < end)
+	goto reph_move;
+    }
+
+    /*       5. If no consonant is found in steps 3 or 4, move reph to a position
+     *          immediately before the first post-base matra, syllable modifier
+     *          sign or vedic sign that has a reordering class after the intended
+     *          reph position. For example, if the reordering position for reph
+     *          is post-main, it will skip above-base matras that also have a
+     *          post-main position.
+     */
+    reph_step_5:
+    {
+      /* Copied from step 2. */
+      new_reph_pos = start + 1;
+      while (new_reph_pos < base && !is_halant (info[new_reph_pos]))
+	new_reph_pos++;
+
+      if (new_reph_pos < base && is_halant (info[new_reph_pos]))
+      {
+	/* ->If ZWJ or ZWNJ are following this halant, position is moved after it. */
+	if (new_reph_pos + 1 < base && is_joiner (info[new_reph_pos + 1]))
+	  new_reph_pos++;
+	goto reph_move;
+      }
+    }
+    /* See https://github.com/harfbuzz/harfbuzz/issues/2298#issuecomment-615318654 */
+
+    /*       6. Otherwise, reorder reph to the end of the syllable.
+     */
+    {
+      new_reph_pos = end - 1;
+      while (new_reph_pos > start && info[new_reph_pos].indic_position() == POS_SMVD)
+	new_reph_pos--;
+
+      /*
+       * If the Reph is to be ending up after a Matra,Halant sequence,
+       * position it before that Halant so it can interact with the Matra.
+       * However, if it's a plain Consonant,Halant we shouldn't do that.
+       * Uniscribe doesn't do this.
+       * TEST: U+0930,U+094D,U+0915,U+094B,U+094D
+       */
+      if (!indic_plan->uniscribe_bug_compatible &&
+	  unlikely (is_halant (info[new_reph_pos])))
+      {
+	for (unsigned int i = base + 1; i < new_reph_pos; i++)
+	  if (info[i].indic_category() == OT_M) {
+	    /* Ok, got it. */
+	    new_reph_pos--;
+	  }
+      }
+
+      goto reph_move;
+    }
+
+    reph_move:
+    {
+      /* Move */
+      buffer->merge_clusters (start, new_reph_pos + 1);
+      hb_glyph_info_t reph = info[start];
+      memmove (&info[start], &info[start + 1], (new_reph_pos - start) * sizeof (info[0]));
+      info[new_reph_pos] = reph;
+
+      if (start < base && base <= new_reph_pos)
+	base--;
+    }
+  }
+
+
+  /*   o Reorder pre-base-reordering consonants:
+   *
+   *     If a pre-base-reordering consonant is found, reorder it according to
+   *     the following rules:
+   */
+
+  if (try_pref && base + 1 < end) /* Otherwise there can't be any pre-base-reordering Ra. */
+  {
+    for (unsigned int i = base + 1; i < end; i++)
+      if ((info[i].mask & indic_plan->mask_array[INDIC_PREF]) != 0)
+      {
+	/*       1. Only reorder a glyph produced by substitution during application
+	 *          of the <pref> feature. (Note that a font may shape a Ra consonant with
+	 *          the feature generally but block it in certain contexts.)
+	 */
+	/* Note: We just check that something got substituted.  We don't check that
+	 * the <pref> feature actually did it...
+	 *
+	 * Reorder pref only if it ligated. */
+	if (_hb_glyph_info_ligated_and_didnt_multiply (&info[i]))
+	{
+	  /*
+	   *       2. Try to find a target position the same way as for pre-base matra.
+	   *          If it is found, reorder pre-base consonant glyph.
+	   *
+	   *       3. If position is not found, reorder immediately before main
+	   *          consonant.
+	   */
+
+	  unsigned int new_pos = base;
+	  /* Malayalam / Tamil do not have "half" forms or explicit virama forms.
+	   * The glyphs formed by 'half' are Chillus or ligated explicit viramas.
+	   * We want to position matra after them.
+	   */
+	  if (buffer->props.script != HB_SCRIPT_MALAYALAM && buffer->props.script != HB_SCRIPT_TAMIL)
+	  {
+	    while (new_pos > start &&
+		   !(is_one_of (info[new_pos - 1], FLAG(OT_M) | FLAG (OT_H))))
+	      new_pos--;
+	  }
+
+	  if (new_pos > start && is_halant (info[new_pos - 1]))
+	  {
+	    /* -> If ZWJ or ZWNJ follow this halant, position is moved after it. */
+	    if (new_pos < end && is_joiner (info[new_pos]))
+	      new_pos++;
+	  }
+
+	  {
+	    unsigned int old_pos = i;
+
+	    buffer->merge_clusters (new_pos, old_pos + 1);
+	    hb_glyph_info_t tmp = info[old_pos];
+	    memmove (&info[new_pos + 1], &info[new_pos], (old_pos - new_pos) * sizeof (info[0]));
+	    info[new_pos] = tmp;
+
+	    if (new_pos <= base && base < old_pos)
+	      base++;
+	  }
+	}
+
+	break;
+      }
+  }
+
+
+  /* Apply 'init' to the Left Matra if it's a word start. */
+  if (info[start].indic_position () == POS_PRE_M)
+  {
+    if (!start ||
+	!(FLAG_UNSAFE (_hb_glyph_info_get_general_category (&info[start - 1])) &
+	 FLAG_RANGE (HB_UNICODE_GENERAL_CATEGORY_FORMAT, HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)))
+      info[start].mask |= indic_plan->mask_array[INDIC_INIT];
+    else
+      buffer->unsafe_to_break (start - 1, start + 1);
+  }
+
+
+  /*
+   * Finish off the clusters and go home!
+   */
+  if (indic_plan->uniscribe_bug_compatible)
+  {
+    switch ((hb_tag_t) plan->props.script)
+    {
+      case HB_SCRIPT_TAMIL:
+      case HB_SCRIPT_SINHALA:
+	break;
+
+      default:
+	/* Uniscribe merges the entire syllable into a single cluster... Except for Tamil & Sinhala.
+	 * This means, half forms are submerged into the main consonant's cluster.
+	 * This is unnecessary, and makes cursor positioning harder, but that's what
+	 * Uniscribe does. */
+	buffer->merge_clusters (start, end);
+	break;
+    }
+  }
+}
+
+
+static void
+final_reordering_indic (const hb_ot_shape_plan_t *plan,
+			hb_font_t *font HB_UNUSED,
+			hb_buffer_t *buffer)
+{
+  unsigned int count = buffer->len;
+  if (unlikely (!count)) return;
+
+  foreach_syllable (buffer, start, end)
+    final_reordering_syllable_indic (plan, buffer, start, end);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, indic_category);
+  HB_BUFFER_DEALLOCATE_VAR (buffer, indic_position);
+}
+
+
+static void
+preprocess_text_indic (const hb_ot_shape_plan_t *plan,
+		       hb_buffer_t              *buffer,
+		       hb_font_t                *font)
+{
+  _hb_preprocess_text_vowel_constraints (plan, buffer, font);
+}
+
+static bool
+decompose_indic (const hb_ot_shape_normalize_context_t *c,
+		 hb_codepoint_t  ab,
+		 hb_codepoint_t *a,
+		 hb_codepoint_t *b)
+{
+  switch (ab)
+  {
+    /* Don't decompose these. */
+    case 0x0931u  : return false; /* DEVANAGARI LETTER RRA */
+    // https://github.com/harfbuzz/harfbuzz/issues/779
+    case 0x09DCu  : return false; /* BENGALI LETTER RRA */
+    case 0x09DDu  : return false; /* BENGALI LETTER RHA */
+    case 0x0B94u  : return false; /* TAMIL LETTER AU */
+
+
+    /*
+     * Decompose split matras that don't have Unicode decompositions.
+     */
+
+#if 0
+    /* Gujarati */
+    /* This one has no decomposition in Unicode, but needs no decomposition either. */
+    /* case 0x0AC9u  : return false; */
+
+    /* Oriya */
+    case 0x0B57u  : *a = no decomp, -> RIGHT; return true;
+#endif
+  }
+
+  if ((ab == 0x0DDAu || hb_in_range<hb_codepoint_t> (ab, 0x0DDCu, 0x0DDEu)))
+  {
+    /*
+     * Sinhala split matras...  Let the fun begin.
+     *
+     * These four characters have Unicode decompositions.  However, Uniscribe
+     * decomposes them "Khmer-style", that is, it uses the character itself to
+     * get the second half.  The first half of all four decompositions is always
+     * U+0DD9.
+     *
+     * Now, there are buggy fonts, namely, the widely used lklug.ttf, that are
+     * broken with Uniscribe.  But we need to support them.  As such, we only
+     * do the Uniscribe-style decomposition if the character is transformed into
+     * its "sec.half" form by the 'pstf' feature.  Otherwise, we fall back to
+     * Unicode decomposition.
+     *
+     * Note that we can't unconditionally use Unicode decomposition.  That would
+     * break some other fonts, that are designed to work with Uniscribe, and
+     * don't have positioning features for the Unicode-style decomposition.
+     *
+     * Argh...
+     *
+     * The Uniscribe behavior is now documented in the newly published Sinhala
+     * spec in 2012:
+     *
+     *   https://docs.microsoft.com/en-us/typography/script-development/sinhala#shaping
+     */
+
+
+    const indic_shape_plan_t *indic_plan = (const indic_shape_plan_t *) c->plan->data;
+    hb_codepoint_t glyph;
+    if (indic_plan->uniscribe_bug_compatible ||
+	(c->font->get_nominal_glyph (ab, &glyph) &&
+	 indic_plan->pstf.would_substitute (&glyph, 1, c->font->face)))
+    {
+      /* Ok, safe to use Uniscribe-style decomposition. */
+      *a = 0x0DD9u;
+      *b = ab;
+      return true;
+    }
+  }
+
+  return (bool) c->unicode->decompose (ab, a, b);
+}
+
+static bool
+compose_indic (const hb_ot_shape_normalize_context_t *c,
+	       hb_codepoint_t  a,
+	       hb_codepoint_t  b,
+	       hb_codepoint_t *ab)
+{
+  /* Avoid recomposing split matras. */
+  if (HB_UNICODE_GENERAL_CATEGORY_IS_MARK (c->unicode->general_category (a)))
+    return false;
+
+  /* Composition-exclusion exceptions that we want to recompose. */
+  if (a == 0x09AFu && b == 0x09BCu) { *ab = 0x09DFu; return true; }
+
+  return (bool) c->unicode->compose (a, b, ab);
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_indic =
+{
+  collect_features_indic,
+  override_features_indic,
+  data_create_indic,
+  data_destroy_indic,
+  preprocess_text_indic,
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
+  decompose_indic,
+  compose_indic,
+  setup_masks_indic,
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.hh
new file mode 100644
index 0000000000..41bd8bd6cc
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-indic.hh
@@ -0,0 +1,436 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_INDIC_HH
+#define HB_OT_SHAPE_COMPLEX_INDIC_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape-complex.hh"
+
+
+/* buffer var allocations */
+#define indic_category() complex_var_u8_0() /* indic_category_t */
+#define indic_position() complex_var_u8_1() /* indic_position_t */
+
+
+#define INDIC_TABLE_ELEMENT_TYPE uint16_t
+
+/* Cateories used in the OpenType spec:
+ * https://docs.microsoft.com/en-us/typography/script-development/devanagari
+ */
+/* Note: This enum is duplicated in the -machine.rl source file.
+ * Not sure how to avoid duplication. */
+enum indic_category_t {
+  OT_X = 0,
+  OT_C = 1,
+  OT_V = 2,
+  OT_N = 3,
+  OT_H = 4,
+  OT_ZWNJ = 5,
+  OT_ZWJ = 6,
+  OT_M = 7,
+  OT_SM = 8,
+  /* OT_VD = 9, UNUSED; we use OT_A instead. */
+  OT_A = 10,
+  OT_PLACEHOLDER = 11,
+  OT_DOTTEDCIRCLE = 12,
+  OT_RS = 13, /* Register Shifter, used in Khmer OT spec. */
+  OT_Coeng = 14, /* Khmer-style Virama. */
+  OT_Repha = 15, /* Atomically-encoded logical or visual repha. */
+  OT_Ra = 16,
+  OT_CM = 17,  /* Consonant-Medial. */
+  OT_Symbol = 18, /* Avagraha, etc that take marks (SM,A,VD). */
+  OT_CS = 19,
+
+  /* The following are used by Khmer & Myanmar shapers.  Defined
+   * here for them to share. */
+  OT_VAbv    = 26,
+  OT_VBlw    = 27,
+  OT_VPre    = 28,
+  OT_VPst    = 29,
+};
+
+#define MEDIAL_FLAGS (FLAG (OT_CM))
+
+/* Note:
+ *
+ * We treat Vowels and placeholders as if they were consonants.  This is safe because Vowels
+ * cannot happen in a consonant syllable.  The plus side however is, we can call the
+ * consonant syllable logic from the vowel syllable function and get it all right! */
+#define CONSONANT_FLAGS (FLAG (OT_C) | FLAG (OT_CS) | FLAG (OT_Ra) | MEDIAL_FLAGS | FLAG (OT_V) | FLAG (OT_PLACEHOLDER) | FLAG (OT_DOTTEDCIRCLE))
+#define JOINER_FLAGS (FLAG (OT_ZWJ) | FLAG (OT_ZWNJ))
+
+
+/* Visual positions in a syllable from left to right. */
+enum indic_position_t {
+  POS_START = 0,
+
+  POS_RA_TO_BECOME_REPH = 1,
+  POS_PRE_M = 2,
+  POS_PRE_C = 3,
+
+  POS_BASE_C = 4,
+  POS_AFTER_MAIN = 5,
+
+  POS_ABOVE_C = 6,
+
+  POS_BEFORE_SUB = 7,
+  POS_BELOW_C = 8,
+  POS_AFTER_SUB = 9,
+
+  POS_BEFORE_POST = 10,
+  POS_POST_C = 11,
+  POS_AFTER_POST = 12,
+
+  POS_FINAL_C = 13,
+  POS_SMVD = 14,
+
+  POS_END = 15
+};
+
+/* Categories used in IndicSyllabicCategory.txt from UCD. */
+enum indic_syllabic_category_t {
+  INDIC_SYLLABIC_CATEGORY_OTHER				= OT_X,
+
+  INDIC_SYLLABIC_CATEGORY_AVAGRAHA			= OT_Symbol,
+  INDIC_SYLLABIC_CATEGORY_BINDU				= OT_SM,
+  INDIC_SYLLABIC_CATEGORY_BRAHMI_JOINING_NUMBER		= OT_PLACEHOLDER, /* Don't care. */
+  INDIC_SYLLABIC_CATEGORY_CANTILLATION_MARK		= OT_A,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT			= OT_C,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_DEAD		= OT_C,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_FINAL		= OT_CM,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_HEAD_LETTER		= OT_C,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_KILLER		= OT_M, /* U+17CD only. */
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_MEDIAL		= OT_CM,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_PLACEHOLDER		= OT_PLACEHOLDER,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_PRECEDING_REPHA	= OT_Repha,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_PREFIXED		= OT_X, /* Don't care. */
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_SUBJOINED		= OT_CM,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_SUCCEEDING_REPHA	= OT_CM,
+  INDIC_SYLLABIC_CATEGORY_CONSONANT_WITH_STACKER	= OT_CS,
+  INDIC_SYLLABIC_CATEGORY_GEMINATION_MARK		= OT_SM, /* https://github.com/harfbuzz/harfbuzz/issues/552 */
+  INDIC_SYLLABIC_CATEGORY_INVISIBLE_STACKER		= OT_Coeng,
+  INDIC_SYLLABIC_CATEGORY_JOINER			= OT_ZWJ,
+  INDIC_SYLLABIC_CATEGORY_MODIFYING_LETTER		= OT_X,
+  INDIC_SYLLABIC_CATEGORY_NON_JOINER			= OT_ZWNJ,
+  INDIC_SYLLABIC_CATEGORY_NUKTA				= OT_N,
+  INDIC_SYLLABIC_CATEGORY_NUMBER			= OT_PLACEHOLDER,
+  INDIC_SYLLABIC_CATEGORY_NUMBER_JOINER			= OT_PLACEHOLDER, /* Don't care. */
+  INDIC_SYLLABIC_CATEGORY_PURE_KILLER			= OT_M, /* Is like a vowel matra. */
+  INDIC_SYLLABIC_CATEGORY_REGISTER_SHIFTER		= OT_RS,
+  INDIC_SYLLABIC_CATEGORY_SYLLABLE_MODIFIER		= OT_SM,
+  INDIC_SYLLABIC_CATEGORY_TONE_LETTER			= OT_X,
+  INDIC_SYLLABIC_CATEGORY_TONE_MARK			= OT_N,
+  INDIC_SYLLABIC_CATEGORY_VIRAMA			= OT_H,
+  INDIC_SYLLABIC_CATEGORY_VISARGA			= OT_SM,
+  INDIC_SYLLABIC_CATEGORY_VOWEL				= OT_V,
+  INDIC_SYLLABIC_CATEGORY_VOWEL_DEPENDENT		= OT_M,
+  INDIC_SYLLABIC_CATEGORY_VOWEL_INDEPENDENT		= OT_V
+};
+
+/* Categories used in IndicSMatraCategory.txt from UCD */
+enum indic_matra_category_t {
+  INDIC_MATRA_CATEGORY_NOT_APPLICABLE			= POS_END,
+
+  INDIC_MATRA_CATEGORY_LEFT				= POS_PRE_C,
+  INDIC_MATRA_CATEGORY_TOP				= POS_ABOVE_C,
+  INDIC_MATRA_CATEGORY_BOTTOM				= POS_BELOW_C,
+  INDIC_MATRA_CATEGORY_RIGHT				= POS_POST_C,
+
+  /* These should resolve to the position of the last part of the split sequence. */
+  INDIC_MATRA_CATEGORY_BOTTOM_AND_RIGHT			= INDIC_MATRA_CATEGORY_RIGHT,
+  INDIC_MATRA_CATEGORY_LEFT_AND_RIGHT			= INDIC_MATRA_CATEGORY_RIGHT,
+  INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM			= INDIC_MATRA_CATEGORY_BOTTOM,
+  INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM_AND_LEFT		= INDIC_MATRA_CATEGORY_BOTTOM,
+  INDIC_MATRA_CATEGORY_TOP_AND_BOTTOM_AND_RIGHT		= INDIC_MATRA_CATEGORY_RIGHT,
+  INDIC_MATRA_CATEGORY_TOP_AND_LEFT			= INDIC_MATRA_CATEGORY_TOP,
+  INDIC_MATRA_CATEGORY_TOP_AND_LEFT_AND_RIGHT		= INDIC_MATRA_CATEGORY_RIGHT,
+  INDIC_MATRA_CATEGORY_TOP_AND_RIGHT			= INDIC_MATRA_CATEGORY_RIGHT,
+
+  INDIC_MATRA_CATEGORY_OVERSTRUCK			= POS_AFTER_MAIN,
+  INDIC_MATRA_CATEGORY_VISUAL_ORDER_LEFT		= POS_PRE_M
+};
+
+#define INDIC_COMBINE_CATEGORIES(S,M) \
+  ( \
+    ASSERT_STATIC_EXPR_ZERO (S < 255 && M < 255) + \
+    ( S | \
+     ( \
+      ( \
+       S == INDIC_SYLLABIC_CATEGORY_CONSONANT_MEDIAL || \
+       S == INDIC_SYLLABIC_CATEGORY_GEMINATION_MARK || \
+       S == INDIC_SYLLABIC_CATEGORY_REGISTER_SHIFTER || \
+       S == INDIC_SYLLABIC_CATEGORY_CONSONANT_SUCCEEDING_REPHA || \
+       S == INDIC_SYLLABIC_CATEGORY_VIRAMA || \
+       S == INDIC_SYLLABIC_CATEGORY_VOWEL_DEPENDENT || \
+       false \
+       ? M : INDIC_MATRA_CATEGORY_NOT_APPLICABLE \
+      ) << 8 \
+     ) \
+    ) \
+   )
+
+HB_INTERNAL INDIC_TABLE_ELEMENT_TYPE
+hb_indic_get_categories (hb_codepoint_t u);
+
+
+static inline bool
+is_one_of (const hb_glyph_info_t &info, unsigned int flags)
+{
+  /* If it ligated, all bets are off. */
+  if (_hb_glyph_info_ligated (&info)) return false;
+  return !!(FLAG_UNSAFE (info.indic_category()) & flags);
+}
+
+static inline bool
+is_joiner (const hb_glyph_info_t &info)
+{
+  return is_one_of (info, JOINER_FLAGS);
+}
+
+static inline bool
+is_consonant (const hb_glyph_info_t &info)
+{
+  return is_one_of (info, CONSONANT_FLAGS);
+}
+
+static inline bool
+is_halant (const hb_glyph_info_t &info)
+{
+  return is_one_of (info, FLAG (OT_H));
+}
+
+#define IN_HALF_BLOCK(u, Base) (((u) & ~0x7Fu) == (Base))
+
+#define IS_DEVA(u) (IN_HALF_BLOCK (u, 0x0900u))
+#define IS_BENG(u) (IN_HALF_BLOCK (u, 0x0980u))
+#define IS_GURU(u) (IN_HALF_BLOCK (u, 0x0A00u))
+#define IS_GUJR(u) (IN_HALF_BLOCK (u, 0x0A80u))
+#define IS_ORYA(u) (IN_HALF_BLOCK (u, 0x0B00u))
+#define IS_TAML(u) (IN_HALF_BLOCK (u, 0x0B80u))
+#define IS_TELU(u) (IN_HALF_BLOCK (u, 0x0C00u))
+#define IS_KNDA(u) (IN_HALF_BLOCK (u, 0x0C80u))
+#define IS_MLYM(u) (IN_HALF_BLOCK (u, 0x0D00u))
+#define IS_SINH(u) (IN_HALF_BLOCK (u, 0x0D80u))
+
+
+#define MATRA_POS_LEFT(u)	POS_PRE_M
+#define MATRA_POS_RIGHT(u)	( \
+				  IS_DEVA(u) ? POS_AFTER_SUB  : \
+				  IS_BENG(u) ? POS_AFTER_POST : \
+				  IS_GURU(u) ? POS_AFTER_POST : \
+				  IS_GUJR(u) ? POS_AFTER_POST : \
+				  IS_ORYA(u) ? POS_AFTER_POST : \
+				  IS_TAML(u) ? POS_AFTER_POST : \
+				  IS_TELU(u) ? (u <= 0x0C42u ? POS_BEFORE_SUB : POS_AFTER_SUB) : \
+				  IS_KNDA(u) ? (u < 0x0CC3u || u > 0xCD6u ? POS_BEFORE_SUB : POS_AFTER_SUB) : \
+				  IS_MLYM(u) ? POS_AFTER_POST : \
+				  IS_SINH(u) ? POS_AFTER_SUB  : \
+				  /*default*/  POS_AFTER_SUB    \
+				)
+#define MATRA_POS_TOP(u)	( /* BENG and MLYM don't have top matras. */ \
+				  IS_DEVA(u) ? POS_AFTER_SUB  : \
+				  IS_GURU(u) ? POS_AFTER_POST : /* Deviate from spec */ \
+				  IS_GUJR(u) ? POS_AFTER_SUB  : \
+				  IS_ORYA(u) ? POS_AFTER_MAIN : \
+				  IS_TAML(u) ? POS_AFTER_SUB  : \
+				  IS_TELU(u) ? POS_BEFORE_SUB : \
+				  IS_KNDA(u) ? POS_BEFORE_SUB : \
+				  IS_SINH(u) ? POS_AFTER_SUB  : \
+				  /*default*/  POS_AFTER_SUB    \
+				)
+#define MATRA_POS_BOTTOM(u)	( \
+				  IS_DEVA(u) ? POS_AFTER_SUB  : \
+				  IS_BENG(u) ? POS_AFTER_SUB  : \
+				  IS_GURU(u) ? POS_AFTER_POST : \
+				  IS_GUJR(u) ? POS_AFTER_POST : \
+				  IS_ORYA(u) ? POS_AFTER_SUB  : \
+				  IS_TAML(u) ? POS_AFTER_POST : \
+				  IS_TELU(u) ? POS_BEFORE_SUB : \
+				  IS_KNDA(u) ? POS_BEFORE_SUB : \
+				  IS_MLYM(u) ? POS_AFTER_POST : \
+				  IS_SINH(u) ? POS_AFTER_SUB  : \
+				  /*default*/  POS_AFTER_SUB    \
+				)
+
+static inline indic_position_t
+matra_position_indic (hb_codepoint_t u, indic_position_t side)
+{
+  switch ((int) side)
+  {
+    case POS_PRE_C:	return MATRA_POS_LEFT (u);
+    case POS_POST_C:	return MATRA_POS_RIGHT (u);
+    case POS_ABOVE_C:	return MATRA_POS_TOP (u);
+    case POS_BELOW_C:	return MATRA_POS_BOTTOM (u);
+  }
+  return side;
+}
+
+/* XXX
+ * This is a hack for now.  We should move this data into the main Indic table.
+ * Or completely remove it and just check in the tables.
+ */
+static const hb_codepoint_t ra_chars[] = {
+  0x0930u, /* Devanagari */
+  0x09B0u, /* Bengali */
+  0x09F0u, /* Bengali */
+  0x0A30u, /* Gurmukhi */	/* No Reph */
+  0x0AB0u, /* Gujarati */
+  0x0B30u, /* Oriya */
+  0x0BB0u, /* Tamil */		/* No Reph */
+  0x0C30u, /* Telugu */		/* Reph formed only with ZWJ */
+  0x0CB0u, /* Kannada */
+  0x0D30u, /* Malayalam */	/* No Reph, Logical Repha */
+
+  0x0DBBu, /* Sinhala */	/* Reph formed only with ZWJ */
+
+  0x179Au, /* Khmer */
+};
+
+static inline bool
+is_ra (hb_codepoint_t u)
+{
+  for (unsigned int i = 0; i < ARRAY_LENGTH (ra_chars); i++)
+    if (u == ra_chars[i])
+      return true;
+  return false;
+}
+
+static inline void
+set_indic_properties (hb_glyph_info_t &info)
+{
+  hb_codepoint_t u = info.codepoint;
+  unsigned int type = hb_indic_get_categories (u);
+  indic_category_t cat = (indic_category_t) (type & 0x7Fu);
+  indic_position_t pos = (indic_position_t) (type >> 8);
+
+
+  /*
+   * Re-assign category
+   */
+
+  /* The following act more like the Bindus. */
+  if (unlikely (hb_in_range<hb_codepoint_t> (u, 0x0953u, 0x0954u)))
+    cat = OT_SM;
+  /* The following act like consonants. */
+  else if (unlikely (hb_in_ranges<hb_codepoint_t> (u, 0x0A72u, 0x0A73u,
+				      0x1CF5u, 0x1CF6u)))
+    cat = OT_C;
+  /* TODO: The following should only be allowed after a Visarga.
+   * For now, just treat them like regular tone marks. */
+  else if (unlikely (hb_in_range<hb_codepoint_t> (u, 0x1CE2u, 0x1CE8u)))
+    cat = OT_A;
+  /* TODO: The following should only be allowed after some of
+   * the nasalization marks, maybe only for U+1CE9..U+1CF1.
+   * For now, just treat them like tone marks. */
+  else if (unlikely (u == 0x1CEDu))
+    cat = OT_A;
+  /* The following take marks in standalone clusters, similar to Avagraha. */
+  else if (unlikely (hb_in_ranges<hb_codepoint_t> (u, 0xA8F2u, 0xA8F7u,
+				      0x1CE9u, 0x1CECu,
+				      0x1CEEu, 0x1CF1u)))
+  {
+    cat = OT_Symbol;
+    static_assert (((int) INDIC_SYLLABIC_CATEGORY_AVAGRAHA == OT_Symbol), "");
+  }
+  else if (unlikely (u == 0x0A51u))
+  {
+    /* https://github.com/harfbuzz/harfbuzz/issues/524 */
+    cat = OT_M;
+    pos = POS_BELOW_C;
+  }
+
+  /* According to ScriptExtensions.txt, these Grantha marks may also be used in Tamil,
+   * so the Indic shaper needs to know their categories. */
+  else if (unlikely (u == 0x11301u || u == 0x11303u)) cat = OT_SM;
+  else if (unlikely (u == 0x1133Bu || u == 0x1133Cu)) cat = OT_N;
+
+  else if (unlikely (u == 0x0AFBu)) cat = OT_N; /* https://github.com/harfbuzz/harfbuzz/issues/552 */
+
+  else if (unlikely (u == 0x0980u)) cat = OT_PLACEHOLDER; /* https://github.com/harfbuzz/harfbuzz/issues/538 */
+  else if (unlikely (u == 0x09FCu)) cat = OT_PLACEHOLDER; /* https://github.com/harfbuzz/harfbuzz/pull/1613 */
+  else if (unlikely (u == 0x0C80u)) cat = OT_PLACEHOLDER; /* https://github.com/harfbuzz/harfbuzz/pull/623 */
+  else if (unlikely (hb_in_range<hb_codepoint_t> (u, 0x2010u, 0x2011u)))
+				    cat = OT_PLACEHOLDER;
+  else if (unlikely (u == 0x25CCu)) cat = OT_DOTTEDCIRCLE;
+
+
+  /*
+   * Re-assign position.
+   */
+
+  if ((FLAG_UNSAFE (cat) & CONSONANT_FLAGS))
+  {
+    pos = POS_BASE_C;
+    if (is_ra (u))
+      cat = OT_Ra;
+  }
+  else if (cat == OT_M)
+  {
+    pos = matra_position_indic (u, pos);
+  }
+  else if ((FLAG_UNSAFE (cat) & (FLAG (OT_SM) /* | FLAG (OT_VD) */ | FLAG (OT_A) | FLAG (OT_Symbol))))
+  {
+    pos = POS_SMVD;
+  }
+
+  if (unlikely (u == 0x0B01u)) pos = POS_BEFORE_SUB; /* Oriya Bindu is BeforeSub in the spec. */
+
+
+
+  info.indic_category() = cat;
+  info.indic_position() = pos;
+}
+
+struct hb_indic_would_substitute_feature_t
+{
+  void init (const hb_ot_map_t *map, hb_tag_t feature_tag, bool zero_context_)
+  {
+    zero_context = zero_context_;
+    map->get_stage_lookups (0/*GSUB*/,
+			    map->get_feature_stage (0/*GSUB*/, feature_tag),
+			    &lookups, &count);
+  }
+
+  bool would_substitute (const hb_codepoint_t *glyphs,
+			 unsigned int          glyphs_count,
+			 hb_face_t            *face) const
+  {
+    for (unsigned int i = 0; i < count; i++)
+      if (hb_ot_layout_lookup_would_substitute (face, lookups[i].index, glyphs, glyphs_count, zero_context))
+	return true;
+    return false;
+  }
+
+  private:
+  const hb_ot_map_t::lookup_map_t *lookups;
+  unsigned int count;
+  bool zero_context;
+};
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_INDIC_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.hh
new file mode 100644
index 0000000000..a040318d34
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.hh
@@ -0,0 +1,372 @@
+
+#line 1 "hb-ot-shape-complex-khmer-machine.rl"
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH
+#define HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH
+
+#include "hb.hh"
+
+
+#line 36 "hb-ot-shape-complex-khmer-machine.hh"
+static const unsigned char _khmer_syllable_machine_trans_keys[] = {
+	5u, 26u, 5u, 21u, 5u, 26u, 5u, 21u, 1u, 16u, 5u, 21u, 5u, 26u, 5u, 21u, 
+	5u, 26u, 5u, 21u, 5u, 21u, 5u, 26u, 5u, 21u, 1u, 16u, 5u, 21u, 5u, 26u, 
+	5u, 21u, 5u, 26u, 5u, 21u, 5u, 26u, 1u, 29u, 5u, 29u, 5u, 29u, 5u, 29u, 
+	22u, 22u, 5u, 22u, 5u, 29u, 5u, 29u, 5u, 29u, 1u, 16u, 5u, 26u, 5u, 29u, 
+	5u, 29u, 22u, 22u, 5u, 22u, 5u, 29u, 5u, 29u, 1u, 16u, 5u, 29u, 5u, 29u, 
+	0
+};
+
+static const char _khmer_syllable_machine_key_spans[] = {
+	22, 17, 22, 17, 16, 17, 22, 17, 
+	22, 17, 17, 22, 17, 16, 17, 22, 
+	17, 22, 17, 22, 29, 25, 25, 25, 
+	1, 18, 25, 25, 25, 16, 22, 25, 
+	25, 1, 18, 25, 25, 16, 25, 25
+};
+
+static const short _khmer_syllable_machine_index_offsets[] = {
+	0, 23, 41, 64, 82, 99, 117, 140, 
+	158, 181, 199, 217, 240, 258, 275, 293, 
+	316, 334, 357, 375, 398, 428, 454, 480, 
+	506, 508, 527, 553, 579, 605, 622, 645, 
+	671, 697, 699, 718, 744, 770, 787, 813
+};
+
+static const char _khmer_syllable_machine_indicies[] = {
+	1, 1, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 2, 
+	3, 0, 0, 0, 0, 4, 0, 1, 
+	1, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 3, 
+	0, 1, 1, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 3, 0, 0, 0, 0, 4, 0, 
+	5, 5, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	4, 0, 6, 6, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 6, 0, 7, 7, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 8, 0, 9, 9, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 10, 0, 0, 
+	0, 0, 4, 0, 9, 9, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 10, 0, 11, 11, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 12, 0, 
+	0, 0, 0, 4, 0, 11, 11, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 12, 0, 14, 
+	14, 13, 13, 13, 13, 13, 13, 13, 
+	13, 13, 13, 13, 13, 13, 13, 15, 
+	13, 14, 14, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 15, 16, 16, 16, 16, 17, 16, 
+	18, 18, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	17, 16, 19, 19, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 19, 16, 20, 20, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 21, 16, 22, 22, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 23, 16, 16, 
+	16, 16, 17, 16, 22, 22, 16, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 23, 16, 24, 24, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 25, 16, 
+	16, 16, 16, 17, 16, 24, 24, 16, 
+	16, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 25, 16, 14, 
+	14, 16, 16, 16, 16, 16, 16, 16, 
+	16, 16, 16, 16, 16, 16, 26, 15, 
+	16, 16, 16, 16, 17, 16, 28, 28, 
+	27, 27, 29, 29, 27, 27, 27, 27, 
+	2, 2, 27, 30, 27, 28, 27, 27, 
+	27, 27, 15, 19, 27, 27, 27, 17, 
+	23, 25, 21, 27, 32, 32, 31, 31, 
+	31, 31, 31, 31, 31, 33, 31, 31, 
+	31, 31, 31, 2, 3, 6, 31, 31, 
+	31, 4, 10, 12, 8, 31, 34, 34, 
+	31, 31, 31, 31, 31, 31, 31, 35, 
+	31, 31, 31, 31, 31, 31, 3, 6, 
+	31, 31, 31, 4, 10, 12, 8, 31, 
+	5, 5, 31, 31, 31, 31, 31, 31, 
+	31, 35, 31, 31, 31, 31, 31, 31, 
+	4, 6, 31, 31, 31, 31, 31, 31, 
+	8, 31, 6, 31, 7, 7, 31, 31, 
+	31, 31, 31, 31, 31, 35, 31, 31, 
+	31, 31, 31, 31, 8, 6, 31, 36, 
+	36, 31, 31, 31, 31, 31, 31, 31, 
+	35, 31, 31, 31, 31, 31, 31, 10, 
+	6, 31, 31, 31, 4, 31, 31, 8, 
+	31, 37, 37, 31, 31, 31, 31, 31, 
+	31, 31, 35, 31, 31, 31, 31, 31, 
+	31, 12, 6, 31, 31, 31, 4, 10, 
+	31, 8, 31, 34, 34, 31, 31, 31, 
+	31, 31, 31, 31, 33, 31, 31, 31, 
+	31, 31, 31, 3, 6, 31, 31, 31, 
+	4, 10, 12, 8, 31, 28, 28, 31, 
+	31, 31, 31, 31, 31, 31, 31, 31, 
+	31, 31, 31, 31, 28, 31, 14, 14, 
+	38, 38, 38, 38, 38, 38, 38, 38, 
+	38, 38, 38, 38, 38, 38, 15, 38, 
+	38, 38, 38, 17, 38, 40, 40, 39, 
+	39, 39, 39, 39, 39, 39, 41, 39, 
+	39, 39, 39, 39, 39, 15, 19, 39, 
+	39, 39, 17, 23, 25, 21, 39, 18, 
+	18, 39, 39, 39, 39, 39, 39, 39, 
+	41, 39, 39, 39, 39, 39, 39, 17, 
+	19, 39, 39, 39, 39, 39, 39, 21, 
+	39, 19, 39, 20, 20, 39, 39, 39, 
+	39, 39, 39, 39, 41, 39, 39, 39, 
+	39, 39, 39, 21, 19, 39, 42, 42, 
+	39, 39, 39, 39, 39, 39, 39, 41, 
+	39, 39, 39, 39, 39, 39, 23, 19, 
+	39, 39, 39, 17, 39, 39, 21, 39, 
+	43, 43, 39, 39, 39, 39, 39, 39, 
+	39, 41, 39, 39, 39, 39, 39, 39, 
+	25, 19, 39, 39, 39, 17, 23, 39, 
+	21, 39, 44, 44, 39, 39, 39, 39, 
+	39, 39, 39, 39, 39, 39, 39, 39, 
+	39, 44, 39, 45, 45, 39, 39, 39, 
+	39, 39, 39, 39, 30, 39, 39, 39, 
+	39, 39, 26, 15, 19, 39, 39, 39, 
+	17, 23, 25, 21, 39, 40, 40, 39, 
+	39, 39, 39, 39, 39, 39, 30, 39, 
+	39, 39, 39, 39, 39, 15, 19, 39, 
+	39, 39, 17, 23, 25, 21, 39, 0
+};
+
+static const char _khmer_syllable_machine_trans_targs[] = {
+	20, 1, 28, 22, 23, 3, 24, 5, 
+	25, 7, 26, 9, 27, 20, 10, 31, 
+	20, 32, 12, 33, 14, 34, 16, 35, 
+	18, 36, 39, 20, 21, 30, 37, 20, 
+	0, 29, 2, 4, 6, 8, 20, 20, 
+	11, 13, 15, 17, 38, 19
+};
+
+static const char _khmer_syllable_machine_trans_actions[] = {
+	1, 0, 2, 2, 2, 0, 0, 0, 
+	2, 0, 2, 0, 2, 3, 0, 4, 
+	5, 2, 0, 0, 0, 2, 0, 2, 
+	0, 2, 4, 8, 2, 9, 0, 10, 
+	0, 0, 0, 0, 0, 0, 11, 12, 
+	0, 0, 0, 0, 4, 0
+};
+
+static const char _khmer_syllable_machine_to_state_actions[] = {
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 6, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0
+};
+
+static const char _khmer_syllable_machine_from_state_actions[] = {
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 7, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0
+};
+
+static const unsigned char _khmer_syllable_machine_eof_trans[] = {
+	1, 1, 1, 1, 1, 1, 1, 1, 
+	1, 1, 14, 17, 17, 17, 17, 17, 
+	17, 17, 17, 17, 0, 32, 32, 32, 
+	32, 32, 32, 32, 32, 32, 39, 40, 
+	40, 40, 40, 40, 40, 40, 40, 40
+};
+
+static const int khmer_syllable_machine_start = 20;
+static const int khmer_syllable_machine_first_final = 20;
+static const int khmer_syllable_machine_error = -1;
+
+static const int khmer_syllable_machine_en_main = 20;
+
+
+#line 36 "hb-ot-shape-complex-khmer-machine.rl"
+
+
+
+#line 80 "hb-ot-shape-complex-khmer-machine.rl"
+
+
+#define found_syllable(syllable_type) \
+  HB_STMT_START { \
+    if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \
+    for (unsigned int i = ts; i < te; i++) \
+      info[i].syllable() = (syllable_serial << 4) | khmer_##syllable_type; \
+    syllable_serial++; \
+    if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
+  } HB_STMT_END
+
+static void
+find_syllables_khmer (hb_buffer_t *buffer)
+{
+  unsigned int p, pe, eof, ts, te, act HB_UNUSED;
+  int cs;
+  hb_glyph_info_t *info = buffer->info;
+  
+#line 242 "hb-ot-shape-complex-khmer-machine.hh"
+	{
+	cs = khmer_syllable_machine_start;
+	ts = 0;
+	te = 0;
+	act = 0;
+	}
+
+#line 100 "hb-ot-shape-complex-khmer-machine.rl"
+
+
+  p = 0;
+  pe = eof = buffer->len;
+
+  unsigned int syllable_serial = 1;
+  
+#line 258 "hb-ot-shape-complex-khmer-machine.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+_resume:
+	switch ( _khmer_syllable_machine_from_state_actions[cs] ) {
+	case 7:
+#line 1 "NONE"
+	{ts = p;}
+	break;
+#line 272 "hb-ot-shape-complex-khmer-machine.hh"
+	}
+
+	_keys = _khmer_syllable_machine_trans_keys + (cs<<1);
+	_inds = _khmer_syllable_machine_indicies + _khmer_syllable_machine_index_offsets[cs];
+
+	_slen = _khmer_syllable_machine_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=( info[p].khmer_category()) &&
+		( info[p].khmer_category()) <= _keys[1] ?
+		( info[p].khmer_category()) - _keys[0] : _slen ];
+
+_eof_trans:
+	cs = _khmer_syllable_machine_trans_targs[_trans];
+
+	if ( _khmer_syllable_machine_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _khmer_syllable_machine_trans_actions[_trans] ) {
+	case 2:
+#line 1 "NONE"
+	{te = p+1;}
+	break;
+	case 8:
+#line 76 "hb-ot-shape-complex-khmer-machine.rl"
+	{te = p+1;{ found_syllable (non_khmer_cluster); }}
+	break;
+	case 10:
+#line 74 "hb-ot-shape-complex-khmer-machine.rl"
+	{te = p;p--;{ found_syllable (consonant_syllable); }}
+	break;
+	case 12:
+#line 75 "hb-ot-shape-complex-khmer-machine.rl"
+	{te = p;p--;{ found_syllable (broken_cluster); }}
+	break;
+	case 11:
+#line 76 "hb-ot-shape-complex-khmer-machine.rl"
+	{te = p;p--;{ found_syllable (non_khmer_cluster); }}
+	break;
+	case 1:
+#line 74 "hb-ot-shape-complex-khmer-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (consonant_syllable); }}
+	break;
+	case 5:
+#line 75 "hb-ot-shape-complex-khmer-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (broken_cluster); }}
+	break;
+	case 3:
+#line 1 "NONE"
+	{	switch( act ) {
+	case 2:
+	{{p = ((te))-1;} found_syllable (broken_cluster); }
+	break;
+	case 3:
+	{{p = ((te))-1;} found_syllable (non_khmer_cluster); }
+	break;
+	}
+	}
+	break;
+	case 4:
+#line 1 "NONE"
+	{te = p+1;}
+#line 75 "hb-ot-shape-complex-khmer-machine.rl"
+	{act = 2;}
+	break;
+	case 9:
+#line 1 "NONE"
+	{te = p+1;}
+#line 76 "hb-ot-shape-complex-khmer-machine.rl"
+	{act = 3;}
+	break;
+#line 342 "hb-ot-shape-complex-khmer-machine.hh"
+	}
+
+_again:
+	switch ( _khmer_syllable_machine_to_state_actions[cs] ) {
+	case 6:
+#line 1 "NONE"
+	{ts = 0;}
+	break;
+#line 351 "hb-ot-shape-complex-khmer-machine.hh"
+	}
+
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	if ( p == eof )
+	{
+	if ( _khmer_syllable_machine_eof_trans[cs] > 0 ) {
+		_trans = _khmer_syllable_machine_eof_trans[cs] - 1;
+		goto _eof_trans;
+	}
+	}
+
+	}
+
+#line 108 "hb-ot-shape-complex-khmer-machine.rl"
+
+}
+
+#undef found_syllable
+
+#endif /* HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.rl b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.rl
new file mode 100644
index 0000000000..e7f14533dd
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer-machine.rl
@@ -0,0 +1,113 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH
+#define HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH
+
+#include "hb.hh"
+
+%%{
+  machine khmer_syllable_machine;
+  alphtype unsigned char;
+  write data;
+}%%
+
+%%{
+
+# Same order as enum khmer_category_t.  Not sure how to avoid duplication.
+C    = 1;
+V    = 2;
+ZWNJ = 5;
+ZWJ  = 6;
+PLACEHOLDER = 11;
+DOTTEDCIRCLE = 12;
+Coeng= 14;
+Ra   = 16;
+Robatic = 20;
+Xgroup  = 21;
+Ygroup  = 22;
+VAbv = 26;
+VBlw = 27;
+VPre = 28;
+VPst = 29;
+
+c = (C | Ra | V);
+cn = c.((ZWJ|ZWNJ)?.Robatic)?;
+joiner = (ZWJ | ZWNJ);
+xgroup = (joiner*.Xgroup)*;
+ygroup = Ygroup*;
+
+# This grammar was experimentally extracted from what Uniscribe allows.
+
+matra_group = VPre? xgroup VBlw? xgroup (joiner?.VAbv)? xgroup VPst?;
+syllable_tail = xgroup matra_group xgroup (Coeng.c)? ygroup;
+
+
+broken_cluster =	(Coeng.cn)* (Coeng | syllable_tail);
+consonant_syllable =	(cn|PLACEHOLDER|DOTTEDCIRCLE) broken_cluster;
+other =			any;
+
+main := |*
+	consonant_syllable	=> { found_syllable (consonant_syllable); };
+	broken_cluster		=> { found_syllable (broken_cluster); };
+	other			=> { found_syllable (non_khmer_cluster); };
+*|;
+
+
+}%%
+
+#define found_syllable(syllable_type) \
+  HB_STMT_START { \
+    if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \
+    for (unsigned int i = ts; i < te; i++) \
+      info[i].syllable() = (syllable_serial << 4) | khmer_##syllable_type; \
+    syllable_serial++; \
+    if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
+  } HB_STMT_END
+
+static void
+find_syllables_khmer (hb_buffer_t *buffer)
+{
+  unsigned int p, pe, eof, ts, te, act HB_UNUSED;
+  int cs;
+  hb_glyph_info_t *info = buffer->info;
+  %%{
+    write init;
+    getkey info[p].khmer_category();
+  }%%
+
+  p = 0;
+  pe = eof = buffer->len;
+
+  unsigned int syllable_serial = 1;
+  %%{
+    write exec;
+  }%%
+}
+
+#undef found_syllable
+
+#endif /* HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.cc
new file mode 100644
index 0000000000..3da8374899
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.cc
@@ -0,0 +1,457 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-khmer.hh"
+#include "hb-ot-layout.hh"
+
+
+/*
+ * Khmer shaper.
+ */
+
+static const hb_ot_map_feature_t
+khmer_features[] =
+{
+  /*
+   * Basic features.
+   * These features are applied in order, one at a time, after reordering.
+   */
+  {HB_TAG('p','r','e','f'), F_MANUAL_JOINERS},
+  {HB_TAG('b','l','w','f'), F_MANUAL_JOINERS},
+  {HB_TAG('a','b','v','f'), F_MANUAL_JOINERS},
+  {HB_TAG('p','s','t','f'), F_MANUAL_JOINERS},
+  {HB_TAG('c','f','a','r'), F_MANUAL_JOINERS},
+  /*
+   * Other features.
+   * These features are applied all at once after clearing syllables.
+   */
+  {HB_TAG('p','r','e','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('a','b','v','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('b','l','w','s'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('p','s','t','s'), F_GLOBAL_MANUAL_JOINERS},
+};
+
+/*
+ * Must be in the same order as the khmer_features array.
+ */
+enum {
+  KHMER_PREF,
+  KHMER_BLWF,
+  KHMER_ABVF,
+  KHMER_PSTF,
+  KHMER_CFAR,
+
+  _KHMER_PRES,
+  _KHMER_ABVS,
+  _KHMER_BLWS,
+  _KHMER_PSTS,
+
+  KHMER_NUM_FEATURES,
+  KHMER_BASIC_FEATURES = _KHMER_PRES, /* Don't forget to update this! */
+};
+
+static void
+setup_syllables_khmer (const hb_ot_shape_plan_t *plan,
+		       hb_font_t *font,
+		       hb_buffer_t *buffer);
+static void
+reorder_khmer (const hb_ot_shape_plan_t *plan,
+	       hb_font_t *font,
+	       hb_buffer_t *buffer);
+
+static void
+collect_features_khmer (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* Do this before any lookups have been applied. */
+  map->add_gsub_pause (setup_syllables_khmer);
+  map->add_gsub_pause (reorder_khmer);
+
+  /* Testing suggests that Uniscribe does NOT pause between basic
+   * features.  Test with KhmerUI.ttf and the following three
+   * sequences:
+   *
+   *   U+1789,U+17BC
+   *   U+1789,U+17D2,U+1789
+   *   U+1789,U+17D2,U+1789,U+17BC
+   *
+   * https://github.com/harfbuzz/harfbuzz/issues/974
+   */
+  map->enable_feature (HB_TAG('l','o','c','l'));
+  map->enable_feature (HB_TAG('c','c','m','p'));
+
+  unsigned int i = 0;
+  for (; i < KHMER_BASIC_FEATURES; i++)
+    map->add_feature (khmer_features[i]);
+
+  map->add_gsub_pause (_hb_clear_syllables);
+
+  for (; i < KHMER_NUM_FEATURES; i++)
+    map->add_feature (khmer_features[i]);
+}
+
+static void
+override_features_khmer (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* Khmer spec has 'clig' as part of required shaping features:
+   * "Apply feature 'clig' to form ligatures that are desired for
+   * typographical correctness.", hence in overrides... */
+  map->enable_feature (HB_TAG('c','l','i','g'));
+
+  /* Uniscribe does not apply 'kern' in Khmer. */
+  if (hb_options ().uniscribe_bug_compatible)
+  {
+    map->disable_feature (HB_TAG('k','e','r','n'));
+  }
+
+  map->disable_feature (HB_TAG('l','i','g','a'));
+}
+
+
+struct khmer_shape_plan_t
+{
+  bool get_virama_glyph (hb_font_t *font, hb_codepoint_t *pglyph) const
+  {
+    hb_codepoint_t glyph = virama_glyph;
+    if (unlikely (virama_glyph == (hb_codepoint_t) -1))
+    {
+      if (!font->get_nominal_glyph (0x17D2u, &glyph))
+	glyph = 0;
+      /* Technically speaking, the spec says we should apply 'locl' to virama too.
+       * Maybe one day... */
+
+      /* Our get_nominal_glyph() function needs a font, so we can't get the virama glyph
+       * during shape planning...  Instead, overwrite it here.  It's safe.  Don't worry! */
+      virama_glyph = glyph;
+    }
+
+    *pglyph = glyph;
+    return glyph != 0;
+  }
+
+  mutable hb_codepoint_t virama_glyph;
+
+  hb_mask_t mask_array[KHMER_NUM_FEATURES];
+};
+
+static void *
+data_create_khmer (const hb_ot_shape_plan_t *plan)
+{
+  khmer_shape_plan_t *khmer_plan = (khmer_shape_plan_t *) calloc (1, sizeof (khmer_shape_plan_t));
+  if (unlikely (!khmer_plan))
+    return nullptr;
+
+  khmer_plan->virama_glyph = (hb_codepoint_t) -1;
+
+  for (unsigned int i = 0; i < ARRAY_LENGTH (khmer_plan->mask_array); i++)
+    khmer_plan->mask_array[i] = (khmer_features[i].flags & F_GLOBAL) ?
+				 0 : plan->map.get_1_mask (khmer_features[i].tag);
+
+  return khmer_plan;
+}
+
+static void
+data_destroy_khmer (void *data)
+{
+  free (data);
+}
+
+
+enum khmer_syllable_type_t {
+  khmer_consonant_syllable,
+  khmer_broken_cluster,
+  khmer_non_khmer_cluster,
+};
+
+#include "hb-ot-shape-complex-khmer-machine.hh"
+
+static void
+setup_masks_khmer (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		   hb_buffer_t              *buffer,
+		   hb_font_t                *font HB_UNUSED)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, khmer_category);
+
+  /* We cannot setup masks here.  We save information about characters
+   * and setup masks later on in a pause-callback. */
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    set_khmer_properties (info[i]);
+}
+
+static void
+setup_syllables_khmer (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		       hb_font_t *font HB_UNUSED,
+		       hb_buffer_t *buffer)
+{
+  find_syllables_khmer (buffer);
+  foreach_syllable (buffer, start, end)
+    buffer->unsafe_to_break (start, end);
+}
+
+
+/* Rules from:
+ * https://docs.microsoft.com/en-us/typography/script-development/devanagari */
+
+static void
+reorder_consonant_syllable (const hb_ot_shape_plan_t *plan,
+			    hb_face_t *face HB_UNUSED,
+			    hb_buffer_t *buffer,
+			    unsigned int start, unsigned int end)
+{
+  const khmer_shape_plan_t *khmer_plan = (const khmer_shape_plan_t *) plan->data;
+  hb_glyph_info_t *info = buffer->info;
+
+  /* Setup masks. */
+  {
+    /* Post-base */
+    hb_mask_t mask = khmer_plan->mask_array[KHMER_BLWF] |
+		     khmer_plan->mask_array[KHMER_ABVF] |
+		     khmer_plan->mask_array[KHMER_PSTF];
+    for (unsigned int i = start + 1; i < end; i++)
+      info[i].mask  |= mask;
+  }
+
+  unsigned int num_coengs = 0;
+  for (unsigned int i = start + 1; i < end; i++)
+  {
+    /* """
+     * When a COENG + (Cons | IndV) combination are found (and subscript count
+     * is less than two) the character combination is handled according to the
+     * subscript type of the character following the COENG.
+     *
+     * ...
+     *
+     * Subscript Type 2 - The COENG + RO characters are reordered to immediately
+     * before the base glyph. Then the COENG + RO characters are assigned to have
+     * the 'pref' OpenType feature applied to them.
+     * """
+     */
+    if (info[i].khmer_category() == OT_Coeng && num_coengs <= 2 && i + 1 < end)
+    {
+      num_coengs++;
+
+      if (info[i + 1].khmer_category() == OT_Ra)
+      {
+	for (unsigned int j = 0; j < 2; j++)
+	  info[i + j].mask |= khmer_plan->mask_array[KHMER_PREF];
+
+	/* Move the Coeng,Ro sequence to the start. */
+	buffer->merge_clusters (start, i + 2);
+	hb_glyph_info_t t0 = info[i];
+	hb_glyph_info_t t1 = info[i + 1];
+	memmove (&info[start + 2], &info[start], (i - start) * sizeof (info[0]));
+	info[start] = t0;
+	info[start + 1] = t1;
+
+	/* Mark the subsequent stuff with 'cfar'.  Used in Khmer.
+	 * Read the feature spec.
+	 * This allows distinguishing the following cases with MS Khmer fonts:
+	 * U+1784,U+17D2,U+179A,U+17D2,U+1782
+	 * U+1784,U+17D2,U+1782,U+17D2,U+179A
+	 */
+	if (khmer_plan->mask_array[KHMER_CFAR])
+	  for (unsigned int j = i + 2; j < end; j++)
+	    info[j].mask |= khmer_plan->mask_array[KHMER_CFAR];
+
+	num_coengs = 2; /* Done. */
+      }
+    }
+
+    /* Reorder left matra piece. */
+    else if (info[i].khmer_category() == OT_VPre)
+    {
+      /* Move to the start. */
+      buffer->merge_clusters (start, i + 1);
+      hb_glyph_info_t t = info[i];
+      memmove (&info[start + 1], &info[start], (i - start) * sizeof (info[0]));
+      info[start] = t;
+    }
+  }
+}
+
+static void
+reorder_syllable_khmer (const hb_ot_shape_plan_t *plan,
+			hb_face_t *face,
+			hb_buffer_t *buffer,
+			unsigned int start, unsigned int end)
+{
+  khmer_syllable_type_t syllable_type = (khmer_syllable_type_t) (buffer->info[start].syllable() & 0x0F);
+  switch (syllable_type)
+  {
+    case khmer_broken_cluster: /* We already inserted dotted-circles, so just call the consonant_syllable. */
+    case khmer_consonant_syllable:
+     reorder_consonant_syllable (plan, face, buffer, start, end);
+     break;
+
+    case khmer_non_khmer_cluster:
+      break;
+  }
+}
+
+static inline void
+insert_dotted_circles_khmer (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			     hb_font_t *font,
+			     hb_buffer_t *buffer)
+{
+  if (unlikely (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE))
+    return;
+
+  /* Note: This loop is extra overhead, but should not be measurable.
+   * TODO Use a buffer scratch flag to remove the loop. */
+  bool has_broken_syllables = false;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if ((info[i].syllable() & 0x0F) == khmer_broken_cluster)
+    {
+      has_broken_syllables = true;
+      break;
+    }
+  if (likely (!has_broken_syllables))
+    return;
+
+
+  hb_codepoint_t dottedcircle_glyph;
+  if (!font->get_nominal_glyph (0x25CCu, &dottedcircle_glyph))
+    return;
+
+  hb_glyph_info_t dottedcircle = {0};
+  dottedcircle.codepoint = 0x25CCu;
+  set_khmer_properties (dottedcircle);
+  dottedcircle.codepoint = dottedcircle_glyph;
+
+  buffer->clear_output ();
+
+  buffer->idx = 0;
+  unsigned int last_syllable = 0;
+  while (buffer->idx < buffer->len && buffer->successful)
+  {
+    unsigned int syllable = buffer->cur().syllable();
+    khmer_syllable_type_t syllable_type = (khmer_syllable_type_t) (syllable & 0x0F);
+    if (unlikely (last_syllable != syllable && syllable_type == khmer_broken_cluster))
+    {
+      last_syllable = syllable;
+
+      hb_glyph_info_t ginfo = dottedcircle;
+      ginfo.cluster = buffer->cur().cluster;
+      ginfo.mask = buffer->cur().mask;
+      ginfo.syllable() = buffer->cur().syllable();
+
+      /* Insert dottedcircle after possible Repha. */
+      while (buffer->idx < buffer->len && buffer->successful &&
+	     last_syllable == buffer->cur().syllable() &&
+	     buffer->cur().khmer_category() == OT_Repha)
+	buffer->next_glyph ();
+
+      buffer->output_info (ginfo);
+    }
+    else
+      buffer->next_glyph ();
+  }
+  buffer->swap_buffers ();
+}
+
+static void
+reorder_khmer (const hb_ot_shape_plan_t *plan,
+	       hb_font_t *font,
+	       hb_buffer_t *buffer)
+{
+  insert_dotted_circles_khmer (plan, font, buffer);
+
+  foreach_syllable (buffer, start, end)
+    reorder_syllable_khmer (plan, font->face, buffer, start, end);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, khmer_category);
+}
+
+
+static bool
+decompose_khmer (const hb_ot_shape_normalize_context_t *c,
+		 hb_codepoint_t  ab,
+		 hb_codepoint_t *a,
+		 hb_codepoint_t *b)
+{
+  switch (ab)
+  {
+    /*
+     * Decompose split matras that don't have Unicode decompositions.
+     */
+
+    /* Khmer */
+    case 0x17BEu  : *a = 0x17C1u; *b= 0x17BEu; return true;
+    case 0x17BFu  : *a = 0x17C1u; *b= 0x17BFu; return true;
+    case 0x17C0u  : *a = 0x17C1u; *b= 0x17C0u; return true;
+    case 0x17C4u  : *a = 0x17C1u; *b= 0x17C4u; return true;
+    case 0x17C5u  : *a = 0x17C1u; *b= 0x17C5u; return true;
+  }
+
+  return (bool) c->unicode->decompose (ab, a, b);
+}
+
+static bool
+compose_khmer (const hb_ot_shape_normalize_context_t *c,
+	       hb_codepoint_t  a,
+	       hb_codepoint_t  b,
+	       hb_codepoint_t *ab)
+{
+  /* Avoid recomposing split matras. */
+  if (HB_UNICODE_GENERAL_CATEGORY_IS_MARK (c->unicode->general_category (a)))
+    return false;
+
+  return (bool) c->unicode->compose (a, b, ab);
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_khmer =
+{
+  collect_features_khmer,
+  override_features_khmer,
+  data_create_khmer,
+  data_destroy_khmer,
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
+  decompose_khmer,
+  compose_khmer,
+  setup_masks_khmer,
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.hh
new file mode 100644
index 0000000000..11a77bfd4b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-khmer.hh
@@ -0,0 +1,113 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_KHMER_HH
+#define HB_OT_SHAPE_COMPLEX_KHMER_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape-complex-indic.hh"
+
+
+/* buffer var allocations */
+#define khmer_category() indic_category() /* khmer_category_t */
+
+
+/* Note: This enum is duplicated in the -machine.rl source file.
+ * Not sure how to avoid duplication. */
+enum khmer_category_t
+{
+  OT_Robatic = 20,
+  OT_Xgroup  = 21,
+  OT_Ygroup  = 22,
+  //OT_VAbv = 26,
+  //OT_VBlw = 27,
+  //OT_VPre = 28,
+  //OT_VPst = 29,
+};
+
+static inline void
+set_khmer_properties (hb_glyph_info_t &info)
+{
+  hb_codepoint_t u = info.codepoint;
+  unsigned int type = hb_indic_get_categories (u);
+  khmer_category_t cat = (khmer_category_t) (type & 0x7Fu);
+  indic_position_t pos = (indic_position_t) (type >> 8);
+
+
+  /*
+   * Re-assign category
+   *
+   * These categories are experimentally extracted from what Uniscribe allows.
+   */
+  switch (u)
+  {
+    case 0x179Au:
+      cat = (khmer_category_t) OT_Ra;
+      break;
+
+    case 0x17CCu:
+    case 0x17C9u:
+    case 0x17CAu:
+      cat = OT_Robatic;
+      break;
+
+    case 0x17C6u:
+    case 0x17CBu:
+    case 0x17CDu:
+    case 0x17CEu:
+    case 0x17CFu:
+    case 0x17D0u:
+    case 0x17D1u:
+      cat = OT_Xgroup;
+      break;
+
+    case 0x17C7u:
+    case 0x17C8u:
+    case 0x17DDu:
+    case 0x17D3u: /* Just guessing. Uniscribe doesn't categorize it. */
+      cat = OT_Ygroup;
+      break;
+  }
+
+  /*
+   * Re-assign position.
+   */
+  if (cat == (khmer_category_t) OT_M)
+    switch ((int) pos)
+    {
+      case POS_PRE_C:	cat = (khmer_category_t) OT_VPre; break;
+      case POS_BELOW_C:	cat = (khmer_category_t) OT_VBlw; break;
+      case POS_ABOVE_C:	cat = (khmer_category_t) OT_VAbv; break;
+      case POS_POST_C:	cat = (khmer_category_t) OT_VPst; break;
+      default: assert (0);
+    }
+
+  info.khmer_category() = cat;
+}
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_KHMER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar-machine.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar-machine.hh
new file mode 100644
index 0000000000..c2f4c0045c
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar-machine.hh
@@ -0,0 +1,430 @@
+
+#line 1 "hb-ot-shape-complex-myanmar-machine.rl"
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_MYANMAR_MACHINE_HH
+#define HB_OT_SHAPE_COMPLEX_MYANMAR_MACHINE_HH
+
+#include "hb.hh"
+
+
+#line 36 "hb-ot-shape-complex-myanmar-machine.hh"
+static const unsigned char _myanmar_syllable_machine_trans_keys[] = {
+	1u, 32u, 3u, 30u, 5u, 29u, 5u, 8u, 5u, 29u, 3u, 25u, 5u, 25u, 5u, 25u, 
+	3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 1u, 16u, 3u, 29u, 3u, 29u, 3u, 29u, 
+	3u, 29u, 3u, 29u, 3u, 30u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 
+	5u, 29u, 5u, 8u, 5u, 29u, 3u, 25u, 5u, 25u, 5u, 25u, 3u, 29u, 3u, 29u, 
+	3u, 29u, 3u, 29u, 1u, 16u, 3u, 30u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 
+	3u, 29u, 3u, 30u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 29u, 3u, 30u, 
+	3u, 29u, 1u, 32u, 1u, 32u, 8u, 8u, 0
+};
+
+static const char _myanmar_syllable_machine_key_spans[] = {
+	32, 28, 25, 4, 25, 23, 21, 21, 
+	27, 27, 27, 27, 16, 27, 27, 27, 
+	27, 27, 28, 27, 27, 27, 27, 27, 
+	25, 4, 25, 23, 21, 21, 27, 27, 
+	27, 27, 16, 28, 27, 27, 27, 27, 
+	27, 28, 27, 27, 27, 27, 27, 28, 
+	27, 32, 32, 1
+};
+
+static const short _myanmar_syllable_machine_index_offsets[] = {
+	0, 33, 62, 88, 93, 119, 143, 165, 
+	187, 215, 243, 271, 299, 316, 344, 372, 
+	400, 428, 456, 485, 513, 541, 569, 597, 
+	625, 651, 656, 682, 706, 728, 750, 778, 
+	806, 834, 862, 879, 908, 936, 964, 992, 
+	1020, 1048, 1077, 1105, 1133, 1161, 1189, 1217, 
+	1246, 1274, 1307, 1340
+};
+
+static const char _myanmar_syllable_machine_indicies[] = {
+	1, 1, 2, 3, 4, 4, 0, 5, 
+	0, 6, 1, 0, 0, 0, 0, 7, 
+	0, 8, 9, 0, 10, 11, 12, 13, 
+	14, 15, 16, 17, 18, 19, 20, 1, 
+	0, 22, 23, 24, 24, 21, 25, 21, 
+	26, 21, 21, 21, 21, 21, 21, 21, 
+	27, 21, 21, 28, 29, 30, 31, 32, 
+	33, 34, 35, 36, 37, 21, 24, 24, 
+	21, 25, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 38, 21, 21, 21, 21, 
+	21, 21, 32, 21, 21, 21, 36, 21, 
+	24, 24, 21, 25, 21, 24, 24, 21, 
+	25, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 32, 21, 21, 21, 36, 21, 39, 
+	21, 24, 24, 21, 25, 21, 32, 21, 
+	21, 21, 21, 21, 21, 21, 40, 21, 
+	21, 21, 21, 21, 21, 32, 21, 24, 
+	24, 21, 25, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 40, 21, 21, 21, 
+	21, 21, 21, 32, 21, 24, 24, 21, 
+	25, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 32, 21, 22, 21, 24, 24, 21, 
+	25, 21, 26, 21, 21, 21, 21, 21, 
+	21, 21, 41, 21, 21, 41, 21, 21, 
+	21, 32, 42, 21, 21, 36, 21, 22, 
+	21, 24, 24, 21, 25, 21, 26, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 32, 21, 21, 
+	21, 36, 21, 22, 21, 24, 24, 21, 
+	25, 21, 26, 21, 21, 21, 21, 21, 
+	21, 21, 41, 21, 21, 21, 21, 21, 
+	21, 32, 42, 21, 21, 36, 21, 22, 
+	21, 24, 24, 21, 25, 21, 26, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 32, 42, 21, 
+	21, 36, 21, 1, 1, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 1, 21, 22, 21, 24, 24, 
+	21, 25, 21, 26, 21, 21, 21, 21, 
+	21, 21, 21, 27, 21, 21, 28, 29, 
+	30, 31, 32, 33, 34, 35, 36, 21, 
+	22, 21, 24, 24, 21, 25, 21, 26, 
+	21, 21, 21, 21, 21, 21, 21, 43, 
+	21, 21, 21, 21, 21, 21, 32, 33, 
+	34, 35, 36, 21, 22, 21, 24, 24, 
+	21, 25, 21, 26, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 32, 33, 34, 35, 36, 21, 
+	22, 21, 24, 24, 21, 25, 21, 26, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 32, 33, 
+	34, 21, 36, 21, 22, 21, 24, 24, 
+	21, 25, 21, 26, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 32, 21, 34, 21, 36, 21, 
+	22, 21, 24, 24, 21, 25, 21, 26, 
+	21, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 32, 33, 
+	34, 35, 36, 43, 21, 22, 21, 24, 
+	24, 21, 25, 21, 26, 21, 21, 21, 
+	21, 21, 21, 21, 21, 21, 21, 28, 
+	21, 30, 21, 32, 33, 34, 35, 36, 
+	21, 22, 21, 24, 24, 21, 25, 21, 
+	26, 21, 21, 21, 21, 21, 21, 21, 
+	43, 21, 21, 28, 21, 21, 21, 32, 
+	33, 34, 35, 36, 21, 22, 21, 24, 
+	24, 21, 25, 21, 26, 21, 21, 21, 
+	21, 21, 21, 21, 44, 21, 21, 28, 
+	29, 30, 21, 32, 33, 34, 35, 36, 
+	21, 22, 21, 24, 24, 21, 25, 21, 
+	26, 21, 21, 21, 21, 21, 21, 21, 
+	21, 21, 21, 28, 29, 30, 21, 32, 
+	33, 34, 35, 36, 21, 22, 23, 24, 
+	24, 21, 25, 21, 26, 21, 21, 21, 
+	21, 21, 21, 21, 27, 21, 21, 28, 
+	29, 30, 31, 32, 33, 34, 35, 36, 
+	21, 46, 46, 45, 5, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 47, 45, 
+	45, 45, 45, 45, 45, 14, 45, 45, 
+	45, 18, 45, 46, 46, 45, 5, 45, 
+	46, 46, 45, 5, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 14, 45, 45, 45, 
+	18, 45, 48, 45, 46, 46, 45, 5, 
+	45, 14, 45, 45, 45, 45, 45, 45, 
+	45, 49, 45, 45, 45, 45, 45, 45, 
+	14, 45, 46, 46, 45, 5, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 49, 
+	45, 45, 45, 45, 45, 45, 14, 45, 
+	46, 46, 45, 5, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 14, 45, 2, 45, 
+	46, 46, 45, 5, 45, 6, 45, 45, 
+	45, 45, 45, 45, 45, 50, 45, 45, 
+	50, 45, 45, 45, 14, 51, 45, 45, 
+	18, 45, 2, 45, 46, 46, 45, 5, 
+	45, 6, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	14, 45, 45, 45, 18, 45, 2, 45, 
+	46, 46, 45, 5, 45, 6, 45, 45, 
+	45, 45, 45, 45, 45, 50, 45, 45, 
+	45, 45, 45, 45, 14, 51, 45, 45, 
+	18, 45, 2, 45, 46, 46, 45, 5, 
+	45, 6, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	14, 51, 45, 45, 18, 45, 52, 52, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 52, 45, 2, 
+	3, 46, 46, 45, 5, 45, 6, 45, 
+	45, 45, 45, 45, 45, 45, 8, 45, 
+	45, 10, 11, 12, 13, 14, 15, 16, 
+	17, 18, 19, 45, 2, 45, 46, 46, 
+	45, 5, 45, 6, 45, 45, 45, 45, 
+	45, 45, 45, 8, 45, 45, 10, 11, 
+	12, 13, 14, 15, 16, 17, 18, 45, 
+	2, 45, 46, 46, 45, 5, 45, 6, 
+	45, 45, 45, 45, 45, 45, 45, 53, 
+	45, 45, 45, 45, 45, 45, 14, 15, 
+	16, 17, 18, 45, 2, 45, 46, 46, 
+	45, 5, 45, 6, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 14, 15, 16, 17, 18, 45, 
+	2, 45, 46, 46, 45, 5, 45, 6, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 14, 15, 
+	16, 45, 18, 45, 2, 45, 46, 46, 
+	45, 5, 45, 6, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 14, 45, 16, 45, 18, 45, 
+	2, 45, 46, 46, 45, 5, 45, 6, 
+	45, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 14, 15, 
+	16, 17, 18, 53, 45, 2, 45, 46, 
+	46, 45, 5, 45, 6, 45, 45, 45, 
+	45, 45, 45, 45, 45, 45, 45, 10, 
+	45, 12, 45, 14, 15, 16, 17, 18, 
+	45, 2, 45, 46, 46, 45, 5, 45, 
+	6, 45, 45, 45, 45, 45, 45, 45, 
+	53, 45, 45, 10, 45, 45, 45, 14, 
+	15, 16, 17, 18, 45, 2, 45, 46, 
+	46, 45, 5, 45, 6, 45, 45, 45, 
+	45, 45, 45, 45, 54, 45, 45, 10, 
+	11, 12, 45, 14, 15, 16, 17, 18, 
+	45, 2, 45, 46, 46, 45, 5, 45, 
+	6, 45, 45, 45, 45, 45, 45, 45, 
+	45, 45, 45, 10, 11, 12, 45, 14, 
+	15, 16, 17, 18, 45, 2, 3, 46, 
+	46, 45, 5, 45, 6, 45, 45, 45, 
+	45, 45, 45, 45, 8, 45, 45, 10, 
+	11, 12, 13, 14, 15, 16, 17, 18, 
+	45, 22, 23, 24, 24, 21, 25, 21, 
+	26, 21, 21, 21, 21, 21, 21, 21, 
+	55, 21, 21, 28, 29, 30, 31, 32, 
+	33, 34, 35, 36, 37, 21, 22, 56, 
+	24, 24, 21, 25, 21, 26, 21, 21, 
+	21, 21, 21, 21, 21, 27, 21, 21, 
+	28, 29, 30, 31, 32, 33, 34, 35, 
+	36, 21, 1, 1, 2, 3, 46, 46, 
+	45, 5, 45, 6, 1, 45, 45, 45, 
+	45, 1, 45, 8, 45, 45, 10, 11, 
+	12, 13, 14, 15, 16, 17, 18, 19, 
+	45, 1, 45, 1, 1, 57, 57, 57, 
+	57, 57, 57, 57, 57, 1, 57, 57, 
+	57, 57, 1, 57, 57, 57, 57, 57, 
+	57, 57, 57, 57, 57, 57, 57, 57, 
+	57, 57, 1, 57, 58, 57, 0
+};
+
+static const char _myanmar_syllable_machine_trans_targs[] = {
+	0, 1, 24, 34, 0, 25, 31, 47, 
+	36, 50, 37, 42, 43, 44, 27, 39, 
+	40, 41, 30, 46, 51, 0, 2, 12, 
+	0, 3, 9, 13, 14, 19, 20, 21, 
+	5, 16, 17, 18, 8, 23, 4, 6, 
+	7, 10, 11, 15, 22, 0, 0, 26, 
+	28, 29, 32, 33, 35, 38, 45, 48, 
+	49, 0, 0
+};
+
+static const char _myanmar_syllable_machine_trans_actions[] = {
+	3, 0, 0, 0, 4, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 5, 0, 0, 
+	6, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 7, 8, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 9, 10
+};
+
+static const char _myanmar_syllable_machine_to_state_actions[] = {
+	1, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0
+};
+
+static const char _myanmar_syllable_machine_from_state_actions[] = {
+	2, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0
+};
+
+static const short _myanmar_syllable_machine_eof_trans[] = {
+	0, 22, 22, 22, 22, 22, 22, 22, 
+	22, 22, 22, 22, 22, 22, 22, 22, 
+	22, 22, 22, 22, 22, 22, 22, 22, 
+	46, 46, 46, 46, 46, 46, 46, 46, 
+	46, 46, 46, 46, 46, 46, 46, 46, 
+	46, 46, 46, 46, 46, 46, 46, 22, 
+	22, 46, 58, 58
+};
+
+static const int myanmar_syllable_machine_start = 0;
+static const int myanmar_syllable_machine_first_final = 0;
+static const int myanmar_syllable_machine_error = -1;
+
+static const int myanmar_syllable_machine_en_main = 0;
+
+
+#line 36 "hb-ot-shape-complex-myanmar-machine.rl"
+
+
+
+#line 94 "hb-ot-shape-complex-myanmar-machine.rl"
+
+
+#define found_syllable(syllable_type) \
+  HB_STMT_START { \
+    if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \
+    for (unsigned int i = ts; i < te; i++) \
+      info[i].syllable() = (syllable_serial << 4) | myanmar_##syllable_type; \
+    syllable_serial++; \
+    if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
+  } HB_STMT_END
+
+static void
+find_syllables_myanmar (hb_buffer_t *buffer)
+{
+  unsigned int p, pe, eof, ts, te, act HB_UNUSED;
+  int cs;
+  hb_glyph_info_t *info = buffer->info;
+  
+#line 320 "hb-ot-shape-complex-myanmar-machine.hh"
+	{
+	cs = myanmar_syllable_machine_start;
+	ts = 0;
+	te = 0;
+	act = 0;
+	}
+
+#line 114 "hb-ot-shape-complex-myanmar-machine.rl"
+
+
+  p = 0;
+  pe = eof = buffer->len;
+
+  unsigned int syllable_serial = 1;
+  
+#line 336 "hb-ot-shape-complex-myanmar-machine.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+_resume:
+	switch ( _myanmar_syllable_machine_from_state_actions[cs] ) {
+	case 2:
+#line 1 "NONE"
+	{ts = p;}
+	break;
+#line 350 "hb-ot-shape-complex-myanmar-machine.hh"
+	}
+
+	_keys = _myanmar_syllable_machine_trans_keys + (cs<<1);
+	_inds = _myanmar_syllable_machine_indicies + _myanmar_syllable_machine_index_offsets[cs];
+
+	_slen = _myanmar_syllable_machine_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=( info[p].myanmar_category()) &&
+		( info[p].myanmar_category()) <= _keys[1] ?
+		( info[p].myanmar_category()) - _keys[0] : _slen ];
+
+_eof_trans:
+	cs = _myanmar_syllable_machine_trans_targs[_trans];
+
+	if ( _myanmar_syllable_machine_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _myanmar_syllable_machine_trans_actions[_trans] ) {
+	case 6:
+#line 86 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p+1;{ found_syllable (consonant_syllable); }}
+	break;
+	case 4:
+#line 87 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p+1;{ found_syllable (non_myanmar_cluster); }}
+	break;
+	case 10:
+#line 88 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p+1;{ found_syllable (punctuation_cluster); }}
+	break;
+	case 8:
+#line 89 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p+1;{ found_syllable (broken_cluster); }}
+	break;
+	case 3:
+#line 90 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p+1;{ found_syllable (non_myanmar_cluster); }}
+	break;
+	case 5:
+#line 86 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p;p--;{ found_syllable (consonant_syllable); }}
+	break;
+	case 7:
+#line 89 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p;p--;{ found_syllable (broken_cluster); }}
+	break;
+	case 9:
+#line 90 "hb-ot-shape-complex-myanmar-machine.rl"
+	{te = p;p--;{ found_syllable (non_myanmar_cluster); }}
+	break;
+#line 400 "hb-ot-shape-complex-myanmar-machine.hh"
+	}
+
+_again:
+	switch ( _myanmar_syllable_machine_to_state_actions[cs] ) {
+	case 1:
+#line 1 "NONE"
+	{ts = 0;}
+	break;
+#line 409 "hb-ot-shape-complex-myanmar-machine.hh"
+	}
+
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	if ( p == eof )
+	{
+	if ( _myanmar_syllable_machine_eof_trans[cs] > 0 ) {
+		_trans = _myanmar_syllable_machine_eof_trans[cs] - 1;
+		goto _eof_trans;
+	}
+	}
+
+	}
+
+#line 122 "hb-ot-shape-complex-myanmar-machine.rl"
+
+}
+
+#undef found_syllable
+
+#endif /* HB_OT_SHAPE_COMPLEX_MYANMAR_MACHINE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.cc
new file mode 100644
index 0000000000..fc3490d716
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.cc
@@ -0,0 +1,387 @@
+/*
+ * Copyright © 2011,2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-myanmar.hh"
+
+
+/*
+ * Myanmar shaper.
+ */
+
+static const hb_tag_t
+myanmar_basic_features[] =
+{
+  /*
+   * Basic features.
+   * These features are applied in order, one at a time, after reordering.
+   */
+  HB_TAG('r','p','h','f'),
+  HB_TAG('p','r','e','f'),
+  HB_TAG('b','l','w','f'),
+  HB_TAG('p','s','t','f'),
+};
+static const hb_tag_t
+myanmar_other_features[] =
+{
+  /*
+   * Other features.
+   * These features are applied all at once, after clearing syllables.
+   */
+  HB_TAG('p','r','e','s'),
+  HB_TAG('a','b','v','s'),
+  HB_TAG('b','l','w','s'),
+  HB_TAG('p','s','t','s'),
+};
+
+static void
+setup_syllables_myanmar (const hb_ot_shape_plan_t *plan,
+			 hb_font_t *font,
+			 hb_buffer_t *buffer);
+static void
+reorder_myanmar (const hb_ot_shape_plan_t *plan,
+		 hb_font_t *font,
+		 hb_buffer_t *buffer);
+
+static void
+collect_features_myanmar (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* Do this before any lookups have been applied. */
+  map->add_gsub_pause (setup_syllables_myanmar);
+
+  map->enable_feature (HB_TAG('l','o','c','l'));
+  /* The Indic specs do not require ccmp, but we apply it here since if
+   * there is a use of it, it's typically at the beginning. */
+  map->enable_feature (HB_TAG('c','c','m','p'));
+
+
+  map->add_gsub_pause (reorder_myanmar);
+
+  for (unsigned int i = 0; i < ARRAY_LENGTH (myanmar_basic_features); i++)
+  {
+    map->enable_feature (myanmar_basic_features[i], F_MANUAL_ZWJ);
+    map->add_gsub_pause (nullptr);
+  }
+
+  map->add_gsub_pause (_hb_clear_syllables);
+
+  for (unsigned int i = 0; i < ARRAY_LENGTH (myanmar_other_features); i++)
+    map->enable_feature (myanmar_other_features[i], F_MANUAL_ZWJ);
+}
+
+static void
+override_features_myanmar (hb_ot_shape_planner_t *plan)
+{
+  plan->map.disable_feature (HB_TAG('l','i','g','a'));
+}
+
+
+enum myanmar_syllable_type_t {
+  myanmar_consonant_syllable,
+  myanmar_punctuation_cluster,
+  myanmar_broken_cluster,
+  myanmar_non_myanmar_cluster,
+};
+
+#include "hb-ot-shape-complex-myanmar-machine.hh"
+
+
+static void
+setup_masks_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		     hb_buffer_t              *buffer,
+		     hb_font_t                *font HB_UNUSED)
+{
+  HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_category);
+  HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_position);
+
+  /* We cannot setup masks here.  We save information about characters
+   * and setup masks later on in a pause-callback. */
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    set_myanmar_properties (info[i]);
+}
+
+static void
+setup_syllables_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			 hb_font_t *font HB_UNUSED,
+			 hb_buffer_t *buffer)
+{
+  find_syllables_myanmar (buffer);
+  foreach_syllable (buffer, start, end)
+    buffer->unsafe_to_break (start, end);
+}
+
+static int
+compare_myanmar_order (const hb_glyph_info_t *pa, const hb_glyph_info_t *pb)
+{
+  int a = pa->myanmar_position();
+  int b = pb->myanmar_position();
+
+  return a < b ? -1 : a == b ? 0 : +1;
+}
+
+
+/* Rules from:
+ * https://docs.microsoft.com/en-us/typography/script-development/myanmar */
+
+static void
+initial_reordering_consonant_syllable (hb_buffer_t *buffer,
+				       unsigned int start, unsigned int end)
+{
+  hb_glyph_info_t *info = buffer->info;
+
+  unsigned int base = end;
+  bool has_reph = false;
+
+  {
+    unsigned int limit = start;
+    if (start + 3 <= end &&
+	info[start  ].myanmar_category() == OT_Ra &&
+	info[start+1].myanmar_category() == OT_As &&
+	info[start+2].myanmar_category() == OT_H)
+    {
+      limit += 3;
+      base = start;
+      has_reph = true;
+    }
+
+    {
+      if (!has_reph)
+	base = limit;
+
+      for (unsigned int i = limit; i < end; i++)
+	if (is_consonant (info[i]))
+	{
+	  base = i;
+	  break;
+	}
+    }
+  }
+
+  /* Reorder! */
+  {
+    unsigned int i = start;
+    for (; i < start + (has_reph ? 3 : 0); i++)
+      info[i].myanmar_position() = POS_AFTER_MAIN;
+    for (; i < base; i++)
+      info[i].myanmar_position() = POS_PRE_C;
+    if (i < end)
+    {
+      info[i].myanmar_position() = POS_BASE_C;
+      i++;
+    }
+    indic_position_t pos = POS_AFTER_MAIN;
+    /* The following loop may be ugly, but it implements all of
+     * Myanmar reordering! */
+    for (; i < end; i++)
+    {
+      if (info[i].myanmar_category() == OT_MR) /* Pre-base reordering */
+      {
+	info[i].myanmar_position() = POS_PRE_C;
+	continue;
+      }
+      if (info[i].myanmar_position() < POS_BASE_C) /* Left matra */
+      {
+	continue;
+      }
+      if (info[i].myanmar_category() == OT_VS)
+      {
+	info[i].myanmar_position() = info[i - 1].myanmar_position();
+	continue;
+      }
+
+      if (pos == POS_AFTER_MAIN && info[i].myanmar_category() == OT_VBlw)
+      {
+	pos = POS_BELOW_C;
+	info[i].myanmar_position() = pos;
+	continue;
+      }
+
+      if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_A)
+      {
+	info[i].myanmar_position() = POS_BEFORE_SUB;
+	continue;
+      }
+      if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_VBlw)
+      {
+	info[i].myanmar_position() = pos;
+	continue;
+      }
+      if (pos == POS_BELOW_C && info[i].myanmar_category() != OT_A)
+      {
+	pos = POS_AFTER_SUB;
+	info[i].myanmar_position() = pos;
+	continue;
+      }
+      info[i].myanmar_position() = pos;
+    }
+  }
+
+  /* Sit tight, rock 'n roll! */
+  buffer->sort (start, end, compare_myanmar_order);
+}
+
+static void
+reorder_syllable_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			  hb_face_t *face HB_UNUSED,
+			  hb_buffer_t *buffer,
+			  unsigned int start, unsigned int end)
+{
+  myanmar_syllable_type_t syllable_type = (myanmar_syllable_type_t) (buffer->info[start].syllable() & 0x0F);
+  switch (syllable_type) {
+
+    case myanmar_broken_cluster: /* We already inserted dotted-circles, so just call the consonant_syllable. */
+    case myanmar_consonant_syllable:
+      initial_reordering_consonant_syllable  (buffer, start, end);
+      break;
+
+    case myanmar_punctuation_cluster:
+    case myanmar_non_myanmar_cluster:
+      break;
+  }
+}
+
+static inline void
+insert_dotted_circles_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			       hb_font_t *font,
+			       hb_buffer_t *buffer)
+{
+  if (unlikely (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE))
+    return;
+
+  /* Note: This loop is extra overhead, but should not be measurable.
+   * TODO Use a buffer scratch flag to remove the loop. */
+  bool has_broken_syllables = false;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if ((info[i].syllable() & 0x0F) == myanmar_broken_cluster)
+    {
+      has_broken_syllables = true;
+      break;
+    }
+  if (likely (!has_broken_syllables))
+    return;
+
+
+  hb_codepoint_t dottedcircle_glyph;
+  if (!font->get_nominal_glyph (0x25CCu, &dottedcircle_glyph))
+    return;
+
+  hb_glyph_info_t dottedcircle = {0};
+  dottedcircle.codepoint = 0x25CCu;
+  set_myanmar_properties (dottedcircle);
+  dottedcircle.codepoint = dottedcircle_glyph;
+
+  buffer->clear_output ();
+
+  buffer->idx = 0;
+  unsigned int last_syllable = 0;
+  while (buffer->idx < buffer->len && buffer->successful)
+  {
+    unsigned int syllable = buffer->cur().syllable();
+    myanmar_syllable_type_t syllable_type = (myanmar_syllable_type_t) (syllable & 0x0F);
+    if (unlikely (last_syllable != syllable && syllable_type == myanmar_broken_cluster))
+    {
+      last_syllable = syllable;
+
+      hb_glyph_info_t ginfo = dottedcircle;
+      ginfo.cluster = buffer->cur().cluster;
+      ginfo.mask = buffer->cur().mask;
+      ginfo.syllable() = buffer->cur().syllable();
+
+      buffer->output_info (ginfo);
+    }
+    else
+      buffer->next_glyph ();
+  }
+  buffer->swap_buffers ();
+}
+
+static void
+reorder_myanmar (const hb_ot_shape_plan_t *plan,
+		 hb_font_t *font,
+		 hb_buffer_t *buffer)
+{
+  insert_dotted_circles_myanmar (plan, font, buffer);
+
+  foreach_syllable (buffer, start, end)
+    reorder_syllable_myanmar (plan, font->face, buffer, start, end);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_category);
+  HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_position);
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar =
+{
+  collect_features_myanmar,
+  override_features_myanmar,
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  setup_masks_myanmar,
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY,
+  false, /* fallback_position */
+};
+
+
+/* Ugly Zawgyi encoding.
+ * Disable all auto processing.
+ * https://github.com/harfbuzz/harfbuzz/issues/1162 */
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar_zawgyi =
+{
+  nullptr, /* collect_features */
+  nullptr, /* override_features */
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  nullptr, /* preprocess_text */
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_NONE,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  nullptr, /* setup_masks */
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.hh
new file mode 100644
index 0000000000..7b9821e6ba
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-myanmar.hh
@@ -0,0 +1,171 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_MYANMAR_HH
+#define HB_OT_SHAPE_COMPLEX_MYANMAR_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape-complex-indic.hh"
+
+
+/* buffer var allocations */
+#define myanmar_category() indic_category() /* myanmar_category_t */
+#define myanmar_position() indic_position() /* myanmar_position_t */
+
+
+/* Note: This enum is duplicated in the -machine.rl source file.
+ * Not sure how to avoid duplication. */
+enum myanmar_category_t {
+  OT_As  = 18,  /* Asat */
+  OT_D0  = 20, /* Digit zero */
+  OT_DB  = OT_N, /* Dot below */
+  OT_GB  = OT_PLACEHOLDER,
+  OT_MH  = 21, /* Various consonant medial types */
+  OT_MR  = 22, /* Various consonant medial types */
+  OT_MW  = 23, /* Various consonant medial types */
+  OT_MY  = 24, /* Various consonant medial types */
+  OT_PT  = 25, /* Pwo and other tones */
+  //OT_VAbv = 26,
+  //OT_VBlw = 27,
+  //OT_VPre = 28,
+  //OT_VPst = 29,
+  OT_VS   = 30, /* Variation selectors */
+  OT_P    = 31, /* Punctuation */
+  OT_D    = 32, /* Digits except zero */
+};
+
+
+static inline void
+set_myanmar_properties (hb_glyph_info_t &info)
+{
+  hb_codepoint_t u = info.codepoint;
+  unsigned int type = hb_indic_get_categories (u);
+  unsigned int cat = type & 0x7Fu;
+  indic_position_t pos = (indic_position_t) (type >> 8);
+
+  /* Myanmar
+   * https://docs.microsoft.com/en-us/typography/script-development/myanmar#analyze
+   */
+  if (unlikely (hb_in_range<hb_codepoint_t> (u, 0xFE00u, 0xFE0Fu)))
+    cat = OT_VS;
+
+  switch (u)
+  {
+    case 0x104Eu:
+      cat = OT_C; /* The spec says C, IndicSyllableCategory doesn't have. */
+      break;
+
+    case 0x002Du: case 0x00A0u: case 0x00D7u: case 0x2012u:
+    case 0x2013u: case 0x2014u: case 0x2015u: case 0x2022u:
+    case 0x25CCu: case 0x25FBu: case 0x25FCu: case 0x25FDu:
+    case 0x25FEu:
+      cat = OT_GB;
+      break;
+
+    case 0x1004u: case 0x101Bu: case 0x105Au:
+      cat = OT_Ra;
+      break;
+
+    case 0x1032u: case 0x1036u:
+      cat = OT_A;
+      break;
+
+    case 0x1039u:
+      cat = OT_H;
+      break;
+
+    case 0x103Au:
+      cat = OT_As;
+      break;
+
+    case 0x1041u: case 0x1042u: case 0x1043u: case 0x1044u:
+    case 0x1045u: case 0x1046u: case 0x1047u: case 0x1048u:
+    case 0x1049u: case 0x1090u: case 0x1091u: case 0x1092u:
+    case 0x1093u: case 0x1094u: case 0x1095u: case 0x1096u:
+    case 0x1097u: case 0x1098u: case 0x1099u:
+      cat = OT_D;
+      break;
+
+    case 0x1040u:
+      cat = OT_D; /* XXX The spec says D0, but Uniscribe doesn't seem to do. */
+      break;
+
+    case 0x103Eu: case 0x1060u:
+      cat = OT_MH;
+      break;
+
+    case 0x103Cu:
+      cat = OT_MR;
+      break;
+
+    case 0x103Du: case 0x1082u:
+      cat = OT_MW;
+      break;
+
+    case 0x103Bu: case 0x105Eu: case 0x105Fu:
+      cat = OT_MY;
+      break;
+
+    case 0x1063u: case 0x1064u: case 0x1069u: case 0x106Au:
+    case 0x106Bu: case 0x106Cu: case 0x106Du: case 0xAA7Bu:
+      cat = OT_PT;
+      break;
+
+    case 0x1038u: case 0x1087u: case 0x1088u: case 0x1089u:
+    case 0x108Au: case 0x108Bu: case 0x108Cu: case 0x108Du:
+    case 0x108Fu: case 0x109Au: case 0x109Bu: case 0x109Cu:
+      cat = OT_SM;
+      break;
+
+    case 0x104Au: case 0x104Bu:
+      cat = OT_P;
+      break;
+
+    case 0xAA74u: case 0xAA75u: case 0xAA76u:
+      /* https://github.com/harfbuzz/harfbuzz/issues/218 */
+      cat = OT_C;
+      break;
+  }
+
+  if (cat == OT_M)
+  {
+    switch ((int) pos)
+    {
+      case POS_PRE_C:	cat = (myanmar_category_t) OT_VPre;
+			pos = POS_PRE_M; break;
+      case POS_ABOVE_C:	cat = (myanmar_category_t) OT_VAbv;   break;
+      case POS_BELOW_C:	cat = (myanmar_category_t) OT_VBlw;   break;
+      case POS_POST_C:	cat = (myanmar_category_t) OT_VPst;   break;
+    }
+  }
+
+  info.myanmar_category() = cat;
+  info.myanmar_position() = pos;
+}
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_MYANMAR_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-thai.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-thai.cc
new file mode 100644
index 0000000000..347ea2e7ac
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-thai.cc
@@ -0,0 +1,394 @@
+/*
+ * Copyright © 2010,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex.hh"
+
+
+/* Thai / Lao shaper */
+
+
+/* PUA shaping */
+
+
+enum thai_consonant_type_t
+{
+  NC,
+  AC,
+  RC,
+  DC,
+  NOT_CONSONANT,
+  NUM_CONSONANT_TYPES = NOT_CONSONANT
+};
+
+static thai_consonant_type_t
+get_consonant_type (hb_codepoint_t u)
+{
+  if (u == 0x0E1Bu || u == 0x0E1Du || u == 0x0E1Fu/* || u == 0x0E2Cu*/)
+    return AC;
+  if (u == 0x0E0Du || u == 0x0E10u)
+    return RC;
+  if (u == 0x0E0Eu || u == 0x0E0Fu)
+    return DC;
+  if (hb_in_range<hb_codepoint_t> (u, 0x0E01u, 0x0E2Eu))
+    return NC;
+  return NOT_CONSONANT;
+}
+
+
+enum thai_mark_type_t
+{
+  AV,
+  BV,
+  T,
+  NOT_MARK,
+  NUM_MARK_TYPES = NOT_MARK
+};
+
+static thai_mark_type_t
+get_mark_type (hb_codepoint_t u)
+{
+  if (u == 0x0E31u || hb_in_range<hb_codepoint_t> (u, 0x0E34u, 0x0E37u) ||
+      u == 0x0E47u || hb_in_range<hb_codepoint_t> (u, 0x0E4Du, 0x0E4Eu))
+    return AV;
+  if (hb_in_range<hb_codepoint_t> (u, 0x0E38u, 0x0E3Au))
+    return BV;
+  if (hb_in_range<hb_codepoint_t> (u, 0x0E48u, 0x0E4Cu))
+    return T;
+  return NOT_MARK;
+}
+
+
+enum thai_action_t
+{
+  NOP,
+  SD,  /* Shift combining-mark down */
+  SL,  /* Shift combining-mark left */
+  SDL, /* Shift combining-mark down-left */
+  RD   /* Remove descender from base */
+};
+
+static hb_codepoint_t
+thai_pua_shape (hb_codepoint_t u, thai_action_t action, hb_font_t *font)
+{
+  struct thai_pua_mapping_t {
+    hb_codepoint_t u;
+    hb_codepoint_t win_pua;
+    hb_codepoint_t mac_pua;
+  } const *pua_mappings = nullptr;
+  static const thai_pua_mapping_t SD_mappings[] = {
+    {0x0E48u, 0xF70Au, 0xF88Bu}, /* MAI EK */
+    {0x0E49u, 0xF70Bu, 0xF88Eu}, /* MAI THO */
+    {0x0E4Au, 0xF70Cu, 0xF891u}, /* MAI TRI */
+    {0x0E4Bu, 0xF70Du, 0xF894u}, /* MAI CHATTAWA */
+    {0x0E4Cu, 0xF70Eu, 0xF897u}, /* THANTHAKHAT */
+    {0x0E38u, 0xF718u, 0xF89Bu}, /* SARA U */
+    {0x0E39u, 0xF719u, 0xF89Cu}, /* SARA UU */
+    {0x0E3Au, 0xF71Au, 0xF89Du}, /* PHINTHU */
+    {0x0000u, 0x0000u, 0x0000u}
+  };
+  static const thai_pua_mapping_t SDL_mappings[] = {
+    {0x0E48u, 0xF705u, 0xF88Cu}, /* MAI EK */
+    {0x0E49u, 0xF706u, 0xF88Fu}, /* MAI THO */
+    {0x0E4Au, 0xF707u, 0xF892u}, /* MAI TRI */
+    {0x0E4Bu, 0xF708u, 0xF895u}, /* MAI CHATTAWA */
+    {0x0E4Cu, 0xF709u, 0xF898u}, /* THANTHAKHAT */
+    {0x0000u, 0x0000u, 0x0000u}
+  };
+  static const thai_pua_mapping_t SL_mappings[] = {
+    {0x0E48u, 0xF713u, 0xF88Au}, /* MAI EK */
+    {0x0E49u, 0xF714u, 0xF88Du}, /* MAI THO */
+    {0x0E4Au, 0xF715u, 0xF890u}, /* MAI TRI */
+    {0x0E4Bu, 0xF716u, 0xF893u}, /* MAI CHATTAWA */
+    {0x0E4Cu, 0xF717u, 0xF896u}, /* THANTHAKHAT */
+    {0x0E31u, 0xF710u, 0xF884u}, /* MAI HAN-AKAT */
+    {0x0E34u, 0xF701u, 0xF885u}, /* SARA I */
+    {0x0E35u, 0xF702u, 0xF886u}, /* SARA II */
+    {0x0E36u, 0xF703u, 0xF887u}, /* SARA UE */
+    {0x0E37u, 0xF704u, 0xF888u}, /* SARA UEE */
+    {0x0E47u, 0xF712u, 0xF889u}, /* MAITAIKHU */
+    {0x0E4Du, 0xF711u, 0xF899u}, /* NIKHAHIT */
+    {0x0000u, 0x0000u, 0x0000u}
+  };
+  static const thai_pua_mapping_t RD_mappings[] = {
+    {0x0E0Du, 0xF70Fu, 0xF89Au}, /* YO YING */
+    {0x0E10u, 0xF700u, 0xF89Eu}, /* THO THAN */
+    {0x0000u, 0x0000u, 0x0000u}
+  };
+
+  switch (action) {
+    case NOP: return u;
+    case SD:  pua_mappings = SD_mappings; break;
+    case SDL: pua_mappings = SDL_mappings; break;
+    case SL:  pua_mappings = SL_mappings; break;
+    case RD:  pua_mappings = RD_mappings; break;
+  }
+  for (; pua_mappings->u; pua_mappings++)
+    if (pua_mappings->u == u)
+    {
+      hb_codepoint_t glyph;
+      if (hb_font_get_glyph (font, pua_mappings->win_pua, 0, &glyph))
+	return pua_mappings->win_pua;
+      if (hb_font_get_glyph (font, pua_mappings->mac_pua, 0, &glyph))
+	return pua_mappings->mac_pua;
+      break;
+    }
+  return u;
+}
+
+
+static enum thai_above_state_t
+{     /* Cluster above looks like: */
+  T0, /*  ⣤                      */
+  T1, /*     ⣼                   */
+  T2, /*        ⣾                */
+  T3, /*           ⣿             */
+  NUM_ABOVE_STATES
+} thai_above_start_state[NUM_CONSONANT_TYPES + 1/* For NOT_CONSONANT */] =
+{
+  T0, /* NC */
+  T1, /* AC */
+  T0, /* RC */
+  T0, /* DC */
+  T3, /* NOT_CONSONANT */
+};
+
+static const struct thai_above_state_machine_edge_t {
+  thai_action_t action;
+  thai_above_state_t next_state;
+} thai_above_state_machine[NUM_ABOVE_STATES][NUM_MARK_TYPES] =
+{        /*AV*/    /*BV*/    /*T*/
+/*T0*/ {{NOP,T3}, {NOP,T0}, {SD, T3}},
+/*T1*/ {{SL, T2}, {NOP,T1}, {SDL,T2}},
+/*T2*/ {{NOP,T3}, {NOP,T2}, {SL, T3}},
+/*T3*/ {{NOP,T3}, {NOP,T3}, {NOP,T3}},
+};
+
+
+static enum thai_below_state_t
+{
+  B0, /* No descender */
+  B1, /* Removable descender */
+  B2, /* Strict descender */
+  NUM_BELOW_STATES
+} thai_below_start_state[NUM_CONSONANT_TYPES + 1/* For NOT_CONSONANT */] =
+{
+  B0, /* NC */
+  B0, /* AC */
+  B1, /* RC */
+  B2, /* DC */
+  B2, /* NOT_CONSONANT */
+};
+
+static const struct thai_below_state_machine_edge_t {
+  thai_action_t action;
+  thai_below_state_t next_state;
+} thai_below_state_machine[NUM_BELOW_STATES][NUM_MARK_TYPES] =
+{        /*AV*/    /*BV*/    /*T*/
+/*B0*/ {{NOP,B0}, {NOP,B2}, {NOP, B0}},
+/*B1*/ {{NOP,B1}, {RD, B2}, {NOP, B1}},
+/*B2*/ {{NOP,B2}, {SD, B2}, {NOP, B2}},
+};
+
+
+static void
+do_thai_pua_shaping (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		     hb_buffer_t              *buffer,
+		     hb_font_t                *font)
+{
+#ifdef HB_NO_OT_SHAPE_COMPLEX_THAI_FALLBACK
+  return;
+#endif
+
+  thai_above_state_t above_state = thai_above_start_state[NOT_CONSONANT];
+  thai_below_state_t below_state = thai_below_start_state[NOT_CONSONANT];
+  unsigned int base = 0;
+
+  hb_glyph_info_t *info = buffer->info;
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    thai_mark_type_t mt = get_mark_type (info[i].codepoint);
+
+    if (mt == NOT_MARK) {
+      thai_consonant_type_t ct = get_consonant_type (info[i].codepoint);
+      above_state = thai_above_start_state[ct];
+      below_state = thai_below_start_state[ct];
+      base = i;
+      continue;
+    }
+
+    const thai_above_state_machine_edge_t &above_edge = thai_above_state_machine[above_state][mt];
+    const thai_below_state_machine_edge_t &below_edge = thai_below_state_machine[below_state][mt];
+    above_state = above_edge.next_state;
+    below_state = below_edge.next_state;
+
+    /* At least one of the above/below actions is NOP. */
+    thai_action_t action = above_edge.action != NOP ? above_edge.action : below_edge.action;
+
+    buffer->unsafe_to_break (base, i);
+    if (action == RD)
+      info[base].codepoint = thai_pua_shape (info[base].codepoint, action, font);
+    else
+      info[i].codepoint = thai_pua_shape (info[i].codepoint, action, font);
+  }
+}
+
+
+static void
+preprocess_text_thai (const hb_ot_shape_plan_t *plan,
+		      hb_buffer_t              *buffer,
+		      hb_font_t                *font)
+{
+  /* This function implements the shaping logic documented here:
+   *
+   *   https://linux.thai.net/~thep/th-otf/shaping.html
+   *
+   * The first shaping rule listed there is needed even if the font has Thai
+   * OpenType tables.  The rest do fallback positioning based on PUA codepoints.
+   * We implement that only if there exist no Thai GSUB in the font.
+   */
+
+  /* The following is NOT specified in the MS OT Thai spec, however, it seems
+   * to be what Uniscribe and other engines implement.  According to Eric Muller:
+   *
+   * When you have a SARA AM, decompose it in NIKHAHIT + SARA AA, *and* move the
+   * NIKHAHIT backwards over any tone mark (0E48-0E4B).
+   *
+   * <0E14, 0E4B, 0E33> -> <0E14, 0E4D, 0E4B, 0E32>
+   *
+   * This reordering is legit only when the NIKHAHIT comes from a SARA AM, not
+   * when it's there to start with. The string <0E14, 0E4B, 0E4D> is probably
+   * not what a user wanted, but the rendering is nevertheless nikhahit above
+   * chattawa.
+   *
+   * Same for Lao.
+   *
+   * Note:
+   *
+   * Uniscribe also does some below-marks reordering.  Namely, it positions U+0E3A
+   * after U+0E38 and U+0E39.  We do that by modifying the ccc for U+0E3A.
+   * See unicode->modified_combining_class ().  Lao does NOT have a U+0E3A
+   * equivalent.
+   */
+
+
+  /*
+   * Here are the characters of significance:
+   *
+   *			Thai	Lao
+   * SARA AM:		U+0E33	U+0EB3
+   * SARA AA:		U+0E32	U+0EB2
+   * Nikhahit:		U+0E4D	U+0ECD
+   *
+   * Testing shows that Uniscribe reorder the following marks:
+   * Thai:	<0E31,0E34..0E37,0E47..0E4E>
+   * Lao:	<0EB1,0EB4..0EB7,0EC7..0ECE>
+   *
+   * Note how the Lao versions are the same as Thai + 0x80.
+   */
+
+  /* We only get one script at a time, so a script-agnostic implementation
+   * is adequate here. */
+#define IS_SARA_AM(x) (((x) & ~0x0080u) == 0x0E33u)
+#define NIKHAHIT_FROM_SARA_AM(x) ((x) - 0x0E33u + 0x0E4Du)
+#define SARA_AA_FROM_SARA_AM(x) ((x) - 1)
+#define IS_TONE_MARK(x) (hb_in_ranges<hb_codepoint_t> ((x) & ~0x0080u, 0x0E34u, 0x0E37u, 0x0E47u, 0x0E4Eu, 0x0E31u, 0x0E31u))
+
+  buffer->clear_output ();
+  unsigned int count = buffer->len;
+  for (buffer->idx = 0; buffer->idx < count && buffer->successful;)
+  {
+    hb_codepoint_t u = buffer->cur().codepoint;
+    if (likely (!IS_SARA_AM (u))) {
+      buffer->next_glyph ();
+      continue;
+    }
+
+    /* Is SARA AM. Decompose and reorder. */
+    hb_glyph_info_t &nikhahit = buffer->output_glyph (NIKHAHIT_FROM_SARA_AM (u));
+    _hb_glyph_info_set_continuation (&nikhahit);
+    buffer->replace_glyph (SARA_AA_FROM_SARA_AM (u));
+    if (unlikely (!buffer->successful))
+      return;
+
+    /* Make Nikhahit be recognized as a ccc=0 mark when zeroing widths. */
+    unsigned int end = buffer->out_len;
+    _hb_glyph_info_set_general_category (&buffer->out_info[end - 2], HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK);
+
+    /* Ok, let's see... */
+    unsigned int start = end - 2;
+    while (start > 0 && IS_TONE_MARK (buffer->out_info[start - 1].codepoint))
+      start--;
+
+    if (start + 2 < end)
+    {
+      /* Move Nikhahit (end-2) to the beginning */
+      buffer->merge_out_clusters (start, end);
+      hb_glyph_info_t t = buffer->out_info[end - 2];
+      memmove (buffer->out_info + start + 1,
+	       buffer->out_info + start,
+	       sizeof (buffer->out_info[0]) * (end - start - 2));
+      buffer->out_info[start] = t;
+    }
+    else
+    {
+      /* Since we decomposed, and NIKHAHIT is combining, merge clusters with the
+       * previous cluster. */
+      if (start && buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
+	buffer->merge_out_clusters (start - 1, end);
+    }
+  }
+  buffer->swap_buffers ();
+
+  /* If font has Thai GSUB, we are done. */
+  if (plan->props.script == HB_SCRIPT_THAI && !plan->map.found_script[0])
+    do_thai_pua_shaping (plan, buffer, font);
+}
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_thai =
+{
+  nullptr, /* collect_features */
+  nullptr, /* override_features */
+  nullptr, /* data_create */
+  nullptr, /* data_destroy */
+  preprocess_text_thai,
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
+  nullptr, /* decompose */
+  nullptr, /* compose */
+  nullptr, /* setup_masks */
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
+  false,/* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-machine.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-machine.hh
new file mode 100644
index 0000000000..462342c618
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-machine.hh
@@ -0,0 +1,562 @@
+
+#line 1 "hb-ot-shape-complex-use-machine.rl"
+/*
+ * Copyright © 2015  Mozilla Foundation.
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH
+#define HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH
+
+#include "hb.hh"
+
+
+#line 38 "hb-ot-shape-complex-use-machine.hh"
+static const unsigned char _use_syllable_machine_trans_keys[] = {
+	12u, 48u, 1u, 15u, 1u, 1u, 12u, 48u, 1u, 1u, 0u, 48u, 21u, 21u, 11u, 48u, 
+	11u, 48u, 1u, 15u, 1u, 1u, 11u, 48u, 22u, 48u, 23u, 48u, 24u, 47u, 25u, 47u, 
+	26u, 47u, 45u, 46u, 46u, 46u, 24u, 48u, 24u, 48u, 24u, 48u, 1u, 1u, 24u, 48u, 
+	23u, 48u, 23u, 48u, 23u, 48u, 22u, 48u, 22u, 48u, 22u, 48u, 22u, 48u, 11u, 48u, 
+	1u, 48u, 11u, 48u, 13u, 21u, 4u, 4u, 13u, 13u, 11u, 48u, 11u, 48u, 41u, 42u, 
+	42u, 42u, 11u, 48u, 11u, 48u, 22u, 48u, 23u, 48u, 24u, 47u, 25u, 47u, 26u, 47u, 
+	45u, 46u, 46u, 46u, 24u, 48u, 24u, 48u, 24u, 48u, 24u, 48u, 23u, 48u, 23u, 48u, 
+	23u, 48u, 22u, 48u, 22u, 48u, 22u, 48u, 22u, 48u, 11u, 48u, 1u, 48u, 1u, 15u, 
+	4u, 4u, 13u, 21u, 13u, 13u, 12u, 48u, 1u, 48u, 11u, 48u, 41u, 42u, 42u, 42u, 
+	21u, 42u, 1u, 5u, 0
+};
+
+static const char _use_syllable_machine_key_spans[] = {
+	37, 15, 1, 37, 1, 49, 1, 38, 
+	38, 15, 1, 38, 27, 26, 24, 23, 
+	22, 2, 1, 25, 25, 25, 1, 25, 
+	26, 26, 26, 27, 27, 27, 27, 38, 
+	48, 38, 9, 1, 1, 38, 38, 2, 
+	1, 38, 38, 27, 26, 24, 23, 22, 
+	2, 1, 25, 25, 25, 25, 26, 26, 
+	26, 27, 27, 27, 27, 38, 48, 15, 
+	1, 9, 1, 37, 48, 38, 2, 1, 
+	22, 5
+};
+
+static const short _use_syllable_machine_index_offsets[] = {
+	0, 38, 54, 56, 94, 96, 146, 148, 
+	187, 226, 242, 244, 283, 311, 338, 363, 
+	387, 410, 413, 415, 441, 467, 493, 495, 
+	521, 548, 575, 602, 630, 658, 686, 714, 
+	753, 802, 841, 851, 853, 855, 894, 933, 
+	936, 938, 977, 1016, 1044, 1071, 1096, 1120, 
+	1143, 1146, 1148, 1174, 1200, 1226, 1252, 1279, 
+	1306, 1333, 1361, 1389, 1417, 1445, 1484, 1533, 
+	1549, 1551, 1561, 1563, 1601, 1650, 1689, 1692, 
+	1694, 1717
+};
+
+static const char _use_syllable_machine_indicies[] = {
+	1, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	1, 0, 0, 0, 1, 0, 3, 2, 
+	2, 2, 2, 2, 2, 2, 2, 2, 
+	2, 2, 2, 2, 4, 2, 3, 2, 
+	6, 5, 5, 5, 5, 5, 5, 5, 
+	5, 5, 5, 5, 5, 5, 5, 5, 
+	5, 5, 5, 5, 5, 5, 5, 5, 
+	5, 5, 5, 5, 5, 5, 5, 5, 
+	6, 5, 5, 5, 6, 5, 7, 5, 
+	8, 9, 10, 8, 11, 12, 10, 10, 
+	10, 10, 10, 3, 13, 14, 10, 15, 
+	8, 8, 16, 17, 10, 10, 18, 19, 
+	20, 21, 22, 23, 24, 18, 25, 26, 
+	27, 28, 29, 30, 10, 31, 32, 33, 
+	10, 34, 35, 36, 37, 38, 39, 40, 
+	13, 10, 42, 41, 44, 1, 43, 43, 
+	45, 43, 43, 43, 43, 43, 46, 47, 
+	48, 49, 50, 51, 52, 53, 47, 54, 
+	46, 55, 56, 57, 58, 43, 59, 60, 
+	61, 43, 43, 43, 43, 62, 63, 64, 
+	65, 1, 43, 44, 1, 43, 43, 45, 
+	43, 43, 43, 43, 43, 66, 47, 48, 
+	49, 50, 51, 52, 53, 47, 54, 55, 
+	55, 56, 57, 58, 43, 59, 60, 61, 
+	43, 43, 43, 43, 62, 63, 64, 65, 
+	1, 43, 44, 67, 67, 67, 67, 67, 
+	67, 67, 67, 67, 67, 67, 67, 67, 
+	68, 67, 44, 67, 44, 1, 43, 43, 
+	45, 43, 43, 43, 43, 43, 43, 47, 
+	48, 49, 50, 51, 52, 53, 47, 54, 
+	55, 55, 56, 57, 58, 43, 59, 60, 
+	61, 43, 43, 43, 43, 62, 63, 64, 
+	65, 1, 43, 47, 48, 49, 50, 51, 
+	43, 43, 43, 43, 43, 43, 56, 57, 
+	58, 43, 59, 60, 61, 43, 43, 43, 
+	43, 48, 63, 64, 65, 69, 43, 48, 
+	49, 50, 51, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 59, 60, 61, 
+	43, 43, 43, 43, 43, 63, 64, 65, 
+	69, 43, 49, 50, 51, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 63, 
+	64, 65, 43, 50, 51, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 63, 
+	64, 65, 43, 51, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 63, 64, 
+	65, 43, 63, 64, 43, 64, 43, 49, 
+	50, 51, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 59, 60, 61, 43, 
+	43, 43, 43, 43, 63, 64, 65, 69, 
+	43, 49, 50, 51, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 60, 
+	61, 43, 43, 43, 43, 43, 63, 64, 
+	65, 69, 43, 49, 50, 51, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 43, 
+	43, 43, 61, 43, 43, 43, 43, 43, 
+	63, 64, 65, 69, 43, 71, 70, 49, 
+	50, 51, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 43, 43, 43, 43, 
+	43, 43, 43, 43, 63, 64, 65, 69, 
+	43, 48, 49, 50, 51, 43, 43, 43, 
+	43, 43, 43, 56, 57, 58, 43, 59, 
+	60, 61, 43, 43, 43, 43, 48, 63, 
+	64, 65, 69, 43, 48, 49, 50, 51, 
+	43, 43, 43, 43, 43, 43, 43, 57, 
+	58, 43, 59, 60, 61, 43, 43, 43, 
+	43, 48, 63, 64, 65, 69, 43, 48, 
+	49, 50, 51, 43, 43, 43, 43, 43, 
+	43, 43, 43, 58, 43, 59, 60, 61, 
+	43, 43, 43, 43, 48, 63, 64, 65, 
+	69, 43, 47, 48, 49, 50, 51, 43, 
+	53, 47, 43, 43, 43, 56, 57, 58, 
+	43, 59, 60, 61, 43, 43, 43, 43, 
+	48, 63, 64, 65, 69, 43, 47, 48, 
+	49, 50, 51, 43, 72, 47, 43, 43, 
+	43, 56, 57, 58, 43, 59, 60, 61, 
+	43, 43, 43, 43, 48, 63, 64, 65, 
+	69, 43, 47, 48, 49, 50, 51, 43, 
+	43, 47, 43, 43, 43, 56, 57, 58, 
+	43, 59, 60, 61, 43, 43, 43, 43, 
+	48, 63, 64, 65, 69, 43, 47, 48, 
+	49, 50, 51, 52, 53, 47, 43, 43, 
+	43, 56, 57, 58, 43, 59, 60, 61, 
+	43, 43, 43, 43, 48, 63, 64, 65, 
+	69, 43, 44, 1, 43, 43, 45, 43, 
+	43, 43, 43, 43, 43, 47, 48, 49, 
+	50, 51, 52, 53, 47, 54, 43, 55, 
+	56, 57, 58, 43, 59, 60, 61, 43, 
+	43, 43, 43, 62, 63, 64, 65, 1, 
+	43, 44, 67, 67, 67, 67, 67, 67, 
+	67, 67, 67, 67, 67, 67, 67, 68, 
+	67, 67, 67, 67, 67, 67, 67, 48, 
+	49, 50, 51, 67, 67, 67, 67, 67, 
+	67, 67, 67, 67, 67, 59, 60, 61, 
+	67, 67, 67, 67, 67, 63, 64, 65, 
+	69, 67, 44, 1, 43, 43, 45, 43, 
+	43, 43, 43, 43, 43, 47, 48, 49, 
+	50, 51, 52, 53, 47, 54, 46, 55, 
+	56, 57, 58, 43, 59, 60, 61, 43, 
+	43, 43, 43, 62, 63, 64, 65, 1, 
+	43, 74, 73, 73, 73, 73, 73, 73, 
+	73, 75, 73, 11, 76, 74, 73, 44, 
+	1, 43, 43, 45, 43, 43, 43, 43, 
+	43, 77, 47, 48, 49, 50, 51, 52, 
+	53, 47, 54, 46, 55, 56, 57, 58, 
+	43, 59, 60, 61, 43, 78, 79, 43, 
+	62, 63, 64, 65, 1, 43, 44, 1, 
+	43, 43, 45, 43, 43, 43, 43, 43, 
+	43, 47, 48, 49, 50, 51, 52, 53, 
+	47, 54, 46, 55, 56, 57, 58, 43, 
+	59, 60, 61, 43, 78, 79, 43, 62, 
+	63, 64, 65, 1, 43, 78, 79, 80, 
+	79, 80, 3, 6, 81, 81, 82, 81, 
+	81, 81, 81, 81, 83, 18, 19, 20, 
+	21, 22, 23, 24, 18, 25, 27, 27, 
+	28, 29, 30, 81, 31, 32, 33, 81, 
+	81, 81, 81, 37, 38, 39, 40, 6, 
+	81, 3, 6, 81, 81, 82, 81, 81, 
+	81, 81, 81, 81, 18, 19, 20, 21, 
+	22, 23, 24, 18, 25, 27, 27, 28, 
+	29, 30, 81, 31, 32, 33, 81, 81, 
+	81, 81, 37, 38, 39, 40, 6, 81, 
+	18, 19, 20, 21, 22, 81, 81, 81, 
+	81, 81, 81, 28, 29, 30, 81, 31, 
+	32, 33, 81, 81, 81, 81, 19, 38, 
+	39, 40, 84, 81, 19, 20, 21, 22, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 31, 32, 33, 81, 81, 81, 
+	81, 81, 38, 39, 40, 84, 81, 20, 
+	21, 22, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 38, 39, 40, 81, 
+	21, 22, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 38, 39, 40, 81, 
+	22, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 38, 39, 40, 81, 38, 
+	39, 81, 39, 81, 20, 21, 22, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 31, 32, 33, 81, 81, 81, 81, 
+	81, 38, 39, 40, 84, 81, 20, 21, 
+	22, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 32, 33, 81, 81, 
+	81, 81, 81, 38, 39, 40, 84, 81, 
+	20, 21, 22, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 33, 
+	81, 81, 81, 81, 81, 38, 39, 40, 
+	84, 81, 20, 21, 22, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 38, 
+	39, 40, 84, 81, 19, 20, 21, 22, 
+	81, 81, 81, 81, 81, 81, 28, 29, 
+	30, 81, 31, 32, 33, 81, 81, 81, 
+	81, 19, 38, 39, 40, 84, 81, 19, 
+	20, 21, 22, 81, 81, 81, 81, 81, 
+	81, 81, 29, 30, 81, 31, 32, 33, 
+	81, 81, 81, 81, 19, 38, 39, 40, 
+	84, 81, 19, 20, 21, 22, 81, 81, 
+	81, 81, 81, 81, 81, 81, 30, 81, 
+	31, 32, 33, 81, 81, 81, 81, 19, 
+	38, 39, 40, 84, 81, 18, 19, 20, 
+	21, 22, 81, 24, 18, 81, 81, 81, 
+	28, 29, 30, 81, 31, 32, 33, 81, 
+	81, 81, 81, 19, 38, 39, 40, 84, 
+	81, 18, 19, 20, 21, 22, 81, 85, 
+	18, 81, 81, 81, 28, 29, 30, 81, 
+	31, 32, 33, 81, 81, 81, 81, 19, 
+	38, 39, 40, 84, 81, 18, 19, 20, 
+	21, 22, 81, 81, 18, 81, 81, 81, 
+	28, 29, 30, 81, 31, 32, 33, 81, 
+	81, 81, 81, 19, 38, 39, 40, 84, 
+	81, 18, 19, 20, 21, 22, 23, 24, 
+	18, 81, 81, 81, 28, 29, 30, 81, 
+	31, 32, 33, 81, 81, 81, 81, 19, 
+	38, 39, 40, 84, 81, 3, 6, 81, 
+	81, 82, 81, 81, 81, 81, 81, 81, 
+	18, 19, 20, 21, 22, 23, 24, 18, 
+	25, 81, 27, 28, 29, 30, 81, 31, 
+	32, 33, 81, 81, 81, 81, 37, 38, 
+	39, 40, 6, 81, 3, 81, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	81, 81, 4, 81, 81, 81, 81, 81, 
+	81, 81, 19, 20, 21, 22, 81, 81, 
+	81, 81, 81, 81, 81, 81, 81, 81, 
+	31, 32, 33, 81, 81, 81, 81, 81, 
+	38, 39, 40, 84, 81, 3, 86, 86, 
+	86, 86, 86, 86, 86, 86, 86, 86, 
+	86, 86, 86, 4, 86, 87, 81, 14, 
+	81, 81, 81, 81, 81, 81, 81, 88, 
+	81, 14, 81, 6, 86, 86, 86, 86, 
+	86, 86, 86, 86, 86, 86, 86, 86, 
+	86, 86, 86, 86, 86, 86, 86, 86, 
+	86, 86, 86, 86, 86, 86, 86, 86, 
+	86, 86, 86, 6, 86, 86, 86, 6, 
+	86, 9, 81, 81, 81, 9, 81, 81, 
+	81, 81, 81, 3, 6, 14, 81, 82, 
+	81, 81, 81, 81, 81, 81, 18, 19, 
+	20, 21, 22, 23, 24, 18, 25, 26, 
+	27, 28, 29, 30, 81, 31, 32, 33, 
+	81, 34, 35, 81, 37, 38, 39, 40, 
+	6, 81, 3, 6, 81, 81, 82, 81, 
+	81, 81, 81, 81, 81, 18, 19, 20, 
+	21, 22, 23, 24, 18, 25, 26, 27, 
+	28, 29, 30, 81, 31, 32, 33, 81, 
+	81, 81, 81, 37, 38, 39, 40, 6, 
+	81, 34, 35, 81, 35, 81, 78, 80, 
+	80, 80, 80, 80, 80, 80, 80, 80, 
+	80, 80, 80, 80, 80, 80, 80, 80, 
+	80, 80, 78, 79, 80, 9, 86, 86, 
+	86, 9, 86, 0
+};
+
+static const char _use_syllable_machine_trans_targs[] = {
+	5, 9, 5, 41, 2, 5, 1, 53, 
+	6, 7, 5, 34, 37, 63, 64, 67, 
+	68, 72, 43, 44, 45, 46, 47, 57, 
+	58, 60, 69, 61, 54, 55, 56, 50, 
+	51, 52, 70, 71, 73, 62, 48, 49, 
+	5, 5, 5, 5, 8, 0, 33, 12, 
+	13, 14, 15, 16, 27, 28, 30, 31, 
+	24, 25, 26, 19, 20, 21, 32, 17, 
+	18, 5, 11, 5, 10, 22, 5, 23, 
+	29, 5, 35, 36, 5, 38, 39, 40, 
+	5, 5, 3, 42, 4, 59, 5, 65, 
+	66
+};
+
+static const char _use_syllable_machine_trans_actions[] = {
+	1, 0, 2, 3, 0, 4, 0, 5, 
+	0, 5, 8, 0, 5, 9, 0, 9, 
+	3, 0, 5, 5, 0, 0, 0, 5, 
+	5, 5, 3, 3, 5, 5, 5, 5, 
+	5, 5, 0, 0, 0, 3, 0, 0, 
+	10, 11, 12, 13, 5, 0, 5, 0, 
+	0, 0, 0, 0, 0, 0, 0, 5, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 14, 5, 15, 0, 0, 16, 0, 
+	0, 17, 0, 0, 18, 5, 0, 0, 
+	19, 20, 0, 3, 0, 5, 21, 0, 
+	0
+};
+
+static const char _use_syllable_machine_to_state_actions[] = {
+	0, 0, 0, 0, 0, 6, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0
+};
+
+static const char _use_syllable_machine_from_state_actions[] = {
+	0, 0, 0, 0, 0, 7, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0, 0, 0, 0, 0, 0, 0, 
+	0, 0
+};
+
+static const short _use_syllable_machine_eof_trans[] = {
+	1, 3, 3, 6, 6, 0, 42, 44, 
+	44, 68, 68, 44, 44, 44, 44, 44, 
+	44, 44, 44, 44, 44, 44, 71, 44, 
+	44, 44, 44, 44, 44, 44, 44, 44, 
+	68, 44, 74, 77, 74, 44, 44, 81, 
+	81, 82, 82, 82, 82, 82, 82, 82, 
+	82, 82, 82, 82, 82, 82, 82, 82, 
+	82, 82, 82, 82, 82, 82, 82, 87, 
+	82, 82, 82, 87, 82, 82, 82, 82, 
+	81, 87
+};
+
+static const int use_syllable_machine_start = 5;
+static const int use_syllable_machine_first_final = 5;
+static const int use_syllable_machine_error = -1;
+
+static const int use_syllable_machine_en_main = 5;
+
+
+#line 38 "hb-ot-shape-complex-use-machine.rl"
+
+
+
+#line 162 "hb-ot-shape-complex-use-machine.rl"
+
+
+#define found_syllable(syllable_type) \
+  HB_STMT_START { \
+    if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \
+    for (unsigned int i = ts; i < te; i++) \
+      info[i].syllable() = (syllable_serial << 4) | use_##syllable_type; \
+    syllable_serial++; \
+    if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
+  } HB_STMT_END
+
+static void
+find_syllables_use (hb_buffer_t *buffer)
+{
+  unsigned int p, pe, eof, ts, te, act;
+  int cs;
+  hb_glyph_info_t *info = buffer->info;
+  
+#line 396 "hb-ot-shape-complex-use-machine.hh"
+	{
+	cs = use_syllable_machine_start;
+	ts = 0;
+	te = 0;
+	act = 0;
+	}
+
+#line 182 "hb-ot-shape-complex-use-machine.rl"
+
+
+  p = 0;
+  pe = eof = buffer->len;
+
+  unsigned int syllable_serial = 1;
+  
+#line 412 "hb-ot-shape-complex-use-machine.hh"
+	{
+	int _slen;
+	int _trans;
+	const unsigned char *_keys;
+	const char *_inds;
+	if ( p == pe )
+		goto _test_eof;
+_resume:
+	switch ( _use_syllable_machine_from_state_actions[cs] ) {
+	case 7:
+#line 1 "NONE"
+	{ts = p;}
+	break;
+#line 426 "hb-ot-shape-complex-use-machine.hh"
+	}
+
+	_keys = _use_syllable_machine_trans_keys + (cs<<1);
+	_inds = _use_syllable_machine_indicies + _use_syllable_machine_index_offsets[cs];
+
+	_slen = _use_syllable_machine_key_spans[cs];
+	_trans = _inds[ _slen > 0 && _keys[0] <=( info[p].use_category()) &&
+		( info[p].use_category()) <= _keys[1] ?
+		( info[p].use_category()) - _keys[0] : _slen ];
+
+_eof_trans:
+	cs = _use_syllable_machine_trans_targs[_trans];
+
+	if ( _use_syllable_machine_trans_actions[_trans] == 0 )
+		goto _again;
+
+	switch ( _use_syllable_machine_trans_actions[_trans] ) {
+	case 5:
+#line 1 "NONE"
+	{te = p+1;}
+	break;
+	case 12:
+#line 150 "hb-ot-shape-complex-use-machine.rl"
+	{te = p+1;{ found_syllable (independent_cluster); }}
+	break;
+	case 14:
+#line 153 "hb-ot-shape-complex-use-machine.rl"
+	{te = p+1;{ found_syllable (standard_cluster); }}
+	break;
+	case 10:
+#line 157 "hb-ot-shape-complex-use-machine.rl"
+	{te = p+1;{ found_syllable (broken_cluster); }}
+	break;
+	case 8:
+#line 158 "hb-ot-shape-complex-use-machine.rl"
+	{te = p+1;{ found_syllable (non_cluster); }}
+	break;
+	case 11:
+#line 150 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (independent_cluster); }}
+	break;
+	case 15:
+#line 151 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (virama_terminated_cluster); }}
+	break;
+	case 16:
+#line 152 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (sakot_terminated_cluster); }}
+	break;
+	case 13:
+#line 153 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (standard_cluster); }}
+	break;
+	case 18:
+#line 154 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (number_joiner_terminated_cluster); }}
+	break;
+	case 17:
+#line 155 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (numeral_cluster); }}
+	break;
+	case 19:
+#line 156 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (symbol_cluster); }}
+	break;
+	case 20:
+#line 157 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (broken_cluster); }}
+	break;
+	case 21:
+#line 158 "hb-ot-shape-complex-use-machine.rl"
+	{te = p;p--;{ found_syllable (non_cluster); }}
+	break;
+	case 1:
+#line 153 "hb-ot-shape-complex-use-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (standard_cluster); }}
+	break;
+	case 4:
+#line 157 "hb-ot-shape-complex-use-machine.rl"
+	{{p = ((te))-1;}{ found_syllable (broken_cluster); }}
+	break;
+	case 2:
+#line 1 "NONE"
+	{	switch( act ) {
+	case 8:
+	{{p = ((te))-1;} found_syllable (broken_cluster); }
+	break;
+	case 9:
+	{{p = ((te))-1;} found_syllable (non_cluster); }
+	break;
+	}
+	}
+	break;
+	case 3:
+#line 1 "NONE"
+	{te = p+1;}
+#line 157 "hb-ot-shape-complex-use-machine.rl"
+	{act = 8;}
+	break;
+	case 9:
+#line 1 "NONE"
+	{te = p+1;}
+#line 158 "hb-ot-shape-complex-use-machine.rl"
+	{act = 9;}
+	break;
+#line 532 "hb-ot-shape-complex-use-machine.hh"
+	}
+
+_again:
+	switch ( _use_syllable_machine_to_state_actions[cs] ) {
+	case 6:
+#line 1 "NONE"
+	{ts = 0;}
+	break;
+#line 541 "hb-ot-shape-complex-use-machine.hh"
+	}
+
+	if ( ++p != pe )
+		goto _resume;
+	_test_eof: {}
+	if ( p == eof )
+	{
+	if ( _use_syllable_machine_eof_trans[cs] > 0 ) {
+		_trans = _use_syllable_machine_eof_trans[cs] - 1;
+		goto _eof_trans;
+	}
+	}
+
+	}
+
+#line 190 "hb-ot-shape-complex-use-machine.rl"
+
+}
+
+#undef found_syllable
+
+#endif /* HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-table.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-table.cc
new file mode 100644
index 0000000000..aa9c350862
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use-table.cc
@@ -0,0 +1,873 @@
+/* == Start of generated table == */
+/*
+ * The following table is generated by running:
+ *
+ *   ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt UnicodeData.txt Blocks.txt
+ *
+ * on files with these headers:
+ *
+ * # IndicSyllabicCategory-13.0.0.txt
+ * # Date: 2019-07-22, 19:55:00 GMT [KW, RP]
+ * # IndicPositionalCategory-13.0.0.txt
+ * # Date: 2019-07-23, 00:01:00 GMT [KW, RP]
+ * # Blocks-13.0.0.txt
+ * # Date: 2019-07-10, 19:06:00 GMT [KW]
+ * UnicodeData.txt does not have a header.
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-use.hh"
+
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-macros"
+#define B	USE_B	/* BASE */
+#define CGJ	USE_CGJ	/* CGJ */
+#define CS	USE_CS	/* CONS_WITH_STACKER */
+#define GB	USE_GB	/* BASE_OTHER */
+#define H	USE_H	/* HALANT */
+#define HN	USE_HN	/* HALANT_NUM */
+#define HVM	USE_HVM	/* HALANT_OR_VOWEL_MODIFIER */
+#define IND	USE_IND	/* BASE_IND */
+#define N	USE_N	/* BASE_NUM */
+#define O	USE_O	/* OTHER */
+#define R	USE_R	/* REPHA */
+#define Rsv	USE_Rsv	/* Reserved */
+#define S	USE_S	/* SYM */
+#define SUB	USE_SUB	/* CONS_SUB */
+#define Sk	USE_Sk	/* SAKOT */
+#define VS	USE_VS	/* VARIATION_SELECTOR */
+#define WJ	USE_WJ	/* Word_Joiner */
+#define ZWJ	USE_ZWJ	/* ZWJ */
+#define ZWNJ	USE_ZWNJ	/* ZWNJ */
+#define CMAbv	USE_CMAbv
+#define CMBlw	USE_CMBlw
+#define FAbv	USE_FAbv
+#define FBlw	USE_FBlw
+#define FPst	USE_FPst
+#define FMAbv	USE_FMAbv
+#define FMBlw	USE_FMBlw
+#define FMPst	USE_FMPst
+#define MAbv	USE_MAbv
+#define MBlw	USE_MBlw
+#define MPst	USE_MPst
+#define MPre	USE_MPre
+#define SMAbv	USE_SMAbv
+#define SMBlw	USE_SMBlw
+#define VAbv	USE_VAbv
+#define VBlw	USE_VBlw
+#define VPst	USE_VPst
+#define VPre	USE_VPre
+#define VMAbv	USE_VMAbv
+#define VMBlw	USE_VMBlw
+#define VMPst	USE_VMPst
+#define VMPre	USE_VMPre
+#pragma GCC diagnostic pop
+
+static const USE_TABLE_ELEMENT_TYPE use_table[] = {
+
+
+#define use_offset_0x0028u 0
+
+
+  /* Basic Latin */
+                                                                         O,     O,     O,     O,     O,    GB,     O,     O,
+  /* 0030 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x00a0u 24
+
+
+  /* Latin-1 Supplement */
+
+  /* 00A0 */    GB,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 00B0 */     O,     O, FMPst, FMPst,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 00C0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 00D0 */     O,     O,     O,     O,     O,     O,     O,    GB,
+
+#define use_offset_0x0348u 80
+
+
+  /* Combining Diacritical Marks */
+                                                                         O,     O,     O,     O,     O,     O,     O,   CGJ,
+
+#define use_offset_0x0900u 88
+
+
+  /* Devanagari */
+
+  /* 0900 */ VMAbv, VMAbv, VMAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0910 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0920 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0930 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VAbv,  VPst, CMBlw,     B,  VPst,  VPre,
+  /* 0940 */  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,  VPst,  VPst,  VPst,  VPst,     H,  VPre,  VPst,
+  /* 0950 */     O, VMAbv, VMBlw,     O,     O,  VAbv,  VBlw,  VBlw,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0960 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0970 */     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+
+  /* Bengali */
+
+  /* 0980 */    GB, VMAbv, VMPst, VMPst,     O,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     B,
+  /* 0990 */     B,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 09A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 09B0 */     B,     O,     B,     O,     O,     O,     B,     B,     B,     B,     O,     O, CMBlw,     B,  VPst,  VPre,
+  /* 09C0 */  VPst,  VBlw,  VBlw,  VBlw,  VBlw,     O,     O,  VPre,  VPre,     O,     O,  VPst,  VPst,     H,   IND,     O,
+  /* 09D0 */     O,     O,     O,     O,     O,     O,     O,  VPst,     O,     O,     O,     O,     B,     B,     O,     B,
+  /* 09E0 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 09F0 */     B,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     B,     O, FMAbv,     O,
+
+  /* Gurmukhi */
+
+  /* 0A00 */     O, VMAbv, VMAbv, VMPst,     O,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     B,
+  /* 0A10 */     B,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0A20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 0A30 */     B,     O,     B,     B,     O,     B,     B,     O,     B,     B,     O,     O, CMBlw,     O,  VPst,  VPre,
+  /* 0A40 */  VPst,  VBlw,  VBlw,     O,     O,     O,     O,  VAbv,  VAbv,     O,     O,  VAbv,  VAbv,     H,     O,     O,
+  /* 0A50 */     O, VMBlw,     O,     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,     O,     B,     O,
+  /* 0A60 */     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0A70 */ VMAbv, CMAbv,    GB,    GB,     O,  MBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Gujarati */
+
+  /* 0A80 */     O, VMAbv, VMAbv, VMPst,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,
+  /* 0A90 */     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0AA0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 0AB0 */     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,     O,     O, CMBlw,     B,  VPst,  VPre,
+  /* 0AC0 */  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,     O,  VAbv,  VAbv,  VAbv,     O,  VPst,  VPst,     H,     O,     O,
+  /* 0AD0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 0AE0 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0AF0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     B, VMAbv, VMAbv, VMAbv, CMAbv, CMAbv, CMAbv,
+
+  /* Oriya */
+
+  /* 0B00 */     O, VMAbv, VMPst, VMPst,     O,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     B,
+  /* 0B10 */     B,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0B20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 0B30 */     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,     O,     O, CMBlw,     B,  VPst,  VAbv,
+  /* 0B40 */  VPst,  VBlw,  VBlw,  VBlw,  VBlw,     O,     O,  VPre,  VPst,     O,     O,  VPst,  VPst,     H,     O,     O,
+  /* 0B50 */     O,     O,     O,     O,     O,  VAbv,  VAbv,  VAbv,     O,     O,     O,     O,     B,     B,     O,     B,
+  /* 0B60 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0B70 */     O,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Tamil */
+
+  /* 0B80 */     O,     O, VMAbv,   IND,     O,     B,     B,     B,     B,     B,     B,     O,     O,     O,     B,     B,
+  /* 0B90 */     B,     O,     B,     B,     B,     B,     O,     O,     O,     B,     B,     O,     B,     O,     B,     B,
+  /* 0BA0 */     O,     O,     O,     B,     B,     O,     O,     O,     B,     B,     B,     O,     O,     O,     B,     B,
+  /* 0BB0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,  VPst,  VPst,
+  /* 0BC0 */  VAbv,  VPst,  VPst,     O,     O,     O,  VPre,  VPre,  VPre,     O,  VPst,  VPst,  VPst,     H,     O,     O,
+  /* 0BD0 */     O,     O,     O,     O,     O,     O,     O,  VPst,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 0BE0 */     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0BF0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Telugu */
+
+  /* 0C00 */ VMAbv, VMPst, VMPst, VMPst, VMAbv,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,
+  /* 0C10 */     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0C20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 0C30 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     B,  VAbv,  VAbv,
+  /* 0C40 */  VAbv,  VPst,  VPst,  VPst,  VPst,     O,  VAbv,  VAbv,  VAbv,     O,  VAbv,  VAbv,  VAbv,     H,     O,     O,
+  /* 0C50 */     O,     O,     O,     O,     O,  VAbv,  VBlw,     O,     B,     B,     B,     O,     O,     O,     O,     O,
+  /* 0C60 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0C70 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Kannada */
+
+  /* 0C80 */     B, VMAbv, VMPst, VMPst,     O,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,
+  /* 0C90 */     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0CA0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 0CB0 */     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     O,     O, CMBlw,     B,  VPst,  VAbv,
+  /* 0CC0 */  VAbv,  VPst,  VPst,  VPst,  VPst,     O,  VAbv,  VAbv,  VAbv,     O,  VAbv,  VAbv,  VAbv,     H,     O,     O,
+  /* 0CD0 */     O,     O,     O,     O,     O,  VPst,  VPst,     O,     O,     O,     O,     O,     O,     O,     B,     O,
+  /* 0CE0 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0CF0 */     O,    CS,    CS,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Malayalam */
+
+  /* 0D00 */ VMAbv, VMAbv, VMPst, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,
+  /* 0D10 */     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0D20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0D30 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VAbv,  VAbv,     B,  VPst,  VPst,
+  /* 0D40 */  VPst,  VPst,  VPst,  VBlw,  VBlw,     O,  VPre,  VPre,  VPre,     O,  VPst,  VPst,  VPst,     H,     R,     O,
+  /* 0D50 */     O,     O,     O,     O,   IND,   IND,   IND,  VPst,     O,     O,     O,     O,     O,     O,     O,     B,
+  /* 0D60 */     B,     B,  VBlw,  VBlw,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0D70 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,   IND,   IND,   IND,   IND,   IND,   IND,
+
+  /* Sinhala */
+
+  /* 0D80 */     O, VMAbv, VMPst, VMPst,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0D90 */     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     B,     B,     B,     B,     B,     B,
+  /* 0DA0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0DB0 */     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     O,     O,
+  /* 0DC0 */     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     H,     O,     O,     O,     O,  VPst,
+  /* 0DD0 */  VPst,  VPst,  VAbv,  VAbv,  VBlw,     O,  VBlw,     O,  VPst,  VPre,  VPst,  VPre,  VPst,  VPst,  VPst,  VPst,
+  /* 0DE0 */     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0DF0 */     O,     O,  VPst,  VPst,     O,     O,     O,     O,
+
+#define use_offset_0x0f18u 1360
+
+
+  /* Tibetan */
+                                                                      VBlw,  VBlw,     O,     O,     O,     O,     O,     O,
+  /* 0F20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0F30 */     B,     B,     B,     B,     O, FMBlw,     O, FMBlw,     O, CMAbv,     O,     O,     O,     O,  VPst,  VPre,
+  /* 0F40 */     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,
+  /* 0F50 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 0F60 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,
+  /* 0F70 */     O,  VBlw,  VBlw,  VAbv,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,  VBlw,  VBlw,  VBlw, VMAbv, VMPst,
+  /* 0F80 */  VBlw,  VAbv, VMAbv, VMAbv,  VBlw,   IND, VMAbv, VMAbv,     B,     B,     B,     B,     B,   SUB,   SUB,   SUB,
+  /* 0F90 */   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,     O,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,
+  /* 0FA0 */   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,
+  /* 0FB0 */   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,     O,     O,     O,
+  /* 0FC0 */     O,     O,     O,     O,     O,     O, FMBlw,     O,
+
+#define use_offset_0x1000u 1536
+
+
+  /* Myanmar */
+
+  /* 1000 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1010 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1020 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,  VPst,  VAbv,  VAbv,  VBlw,
+  /* 1030 */  VBlw,  VPre,  VAbv,  VAbv,  VAbv,  VAbv, VMAbv, VMBlw, VMPst,     H,  VAbv,  MPst,  MPre,  MBlw,  MBlw,     B,
+  /* 1040 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,    GB,     O,     O,    GB,     O,
+  /* 1050 */     B,     B,     B,     B,     B,     B,  VPst,  VPst,  VBlw,  VBlw,     B,     B,     B,     B,  MBlw,  MBlw,
+  /* 1060 */  MBlw,     B,  VPst, VMPst, VMPst,     B,     B,  VPst,  VPst, VMPst, VMPst, VMPst, VMPst, VMPst,     B,     B,
+  /* 1070 */     B,  VAbv,  VAbv,  VAbv,  VAbv,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1080 */     B,     B,  MBlw,  VPst,  VPre,  VAbv,  VAbv, VMPst, VMPst, VMPst, VMPst, VMPst, VMPst, VMBlw,     B, VMPst,
+  /* 1090 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B, VMPst, VMPst,  VPst,  VAbv,     O,     O,
+
+#define use_offset_0x1700u 1696
+
+
+  /* Tagalog */
+
+  /* 1700 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,
+  /* 1710 */     B,     B,  VAbv,  VBlw,  VBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Hanunoo */
+
+  /* 1720 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1730 */     B,     B,  VAbv,  VBlw,  VBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Buhid */
+
+  /* 1740 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1750 */     B,     B,  VAbv,  VBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Tagbanwa */
+
+  /* 1760 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,
+  /* 1770 */     B,     O,  VAbv,  VBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Khmer */
+
+  /* 1780 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1790 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 17A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 17B0 */     B,     B,     B,     B,     O,     O,  VPst,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,  VBlw,  VBlw,  VPst,  VPst,
+  /* 17C0 */  VPst,  VPre,  VPre,  VPre,  VPst,  VPst, VMAbv, VMPst,  VPst, VMAbv, VMAbv, FMAbv,  FAbv, CMAbv, FMAbv, FMAbv,
+  /* 17D0 */ FMAbv,  VAbv,     H, FMAbv,     O,     O,     O,     O,     O,     O,     O,     O,     B, FMAbv,     O,     O,
+  /* 17E0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x1900u 1936
+
+
+  /* Limbu */
+
+  /* 1900 */    GB,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1910 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,
+  /* 1920 */  VAbv,  VAbv,  VBlw,  VPst,  VPst,  VAbv,  VAbv,  VAbv,  VAbv,   SUB,   SUB,   SUB,     O,     O,     O,     O,
+  /* 1930 */  FPst,  FPst, VMBlw,  FPst,  FPst,  FPst,  FPst,  FPst,  FPst,  FBlw,  VAbv, FMBlw,     O,     O,     O,     O,
+  /* 1940 */     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+
+  /* Tai Le */
+
+  /* 1950 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1960 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,
+  /* 1970 */     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* New Tai Lue */
+
+  /* 1980 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1990 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 19A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,
+  /* 19B0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 19C0 */     B,     B,     B,     B,     B,     B,     B,     B, VMPst, VMPst,     O,     O,     O,     O,     O,     O,
+  /* 19D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,
+  /* 19E0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 19F0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Buginese */
+
+  /* 1A00 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1A10 */     B,     B,     B,     B,     B,     B,     B,  VAbv,  VBlw,  VPre,  VPst,  VAbv,     O,     O,     O,     O,
+
+  /* Tai Tham */
+
+  /* 1A20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1A30 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1A40 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1A50 */     B,     B,     B,     B,     B,  MPre,  MBlw,   SUB,  FAbv,  FAbv,  MAbv,   SUB,   SUB,   SUB,   SUB,     O,
+  /* 1A60 */    Sk,  VPst,  VAbv,  VPst,  VPst,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,  VBlw,  VAbv,  VBlw,  VPst,  VPre,  VPre,
+  /* 1A70 */  VPre,  VPre,  VPre,  VAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv,  VAbv, FMAbv, FMAbv,     O,     O, FMBlw,
+  /* 1A80 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+  /* 1A90 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x1b00u 2352
+
+
+  /* Balinese */
+
+  /* 1B00 */ VMAbv, VMAbv, VMAbv,  FAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1B10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1B20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1B30 */     B,     B,     B,     B, CMAbv,  VPst,  VAbv,  VAbv,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VPre,  VPre,
+  /* 1B40 */  VPst,  VPst,  VAbv,  VAbv,     H,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,
+  /* 1B50 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,    GB,    GB,     O,     O,    GB,
+  /* 1B60 */     O,     S,    GB,     S,     S,     S,     S,     S,    GB,     S,     S, SMAbv, SMBlw, SMAbv, SMAbv, SMAbv,
+  /* 1B70 */ SMAbv, SMAbv, SMAbv, SMAbv,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Sundanese */
+
+  /* 1B80 */ VMAbv,  FAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1B90 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1BA0 */     B,   SUB,   SUB,   SUB,  VAbv,  VBlw,  VPre,  VPst,  VAbv,  VAbv,  VPst,     H,   SUB,   SUB,     B,     B,
+  /* 1BB0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+
+  /* Batak */
+
+  /* 1BC0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1BD0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1BE0 */     B,     B,     B,     B,     B,     B, CMAbv,  VPst,  VAbv,  VAbv,  VPst,  VPst,  VPst,  VAbv,  VPst,  VAbv,
+  /* 1BF0 */  FAbv,  FAbv, CMBlw, CMBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Lepcha */
+
+  /* 1C00 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1C10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 1C20 */     B,     B,     B,     B,   SUB,   SUB,  VPst,  VPre,  VPre,  VPre,  VPst,  VPst,  VBlw,  FAbv,  FAbv,  FAbv,
+  /* 1C30 */  FAbv,  FAbv,  FAbv,  FAbv, VMPre, VMPre, FMAbv, CMBlw,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 1C40 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     B,     B,     B,
+
+#define use_offset_0x1cd0u 2688
+
+
+  /* Vedic Extensions */
+
+  /* 1CD0 */ VMAbv, VMAbv, VMAbv,     O, VMBlw, VMBlw, VMBlw, VMBlw, VMBlw, VMBlw, VMAbv, VMAbv, VMBlw, VMBlw, VMBlw, VMBlw,
+  /* 1CE0 */ VMAbv, VMPst, VMBlw, VMBlw, VMBlw, VMBlw, VMBlw, VMBlw, VMBlw,     O,     O,     O,     O, VMBlw,     O,     O,
+  /* 1CF0 */     O,     O,   IND,   IND, VMAbv,    CS,    CS, VMPst, VMAbv, VMAbv,    GB,     O,     O,     O,     O,     O,
+
+#define use_offset_0x1df8u 2736
+
+
+  /* Combining Diacritical Marks Supplement */
+                                                                         O,     O,     O, FMAbv,     O,     O,     O,     O,
+
+#define use_offset_0x2008u 2744
+
+
+  /* General Punctuation */
+                                                                         O,     O,     O,     O,  ZWNJ,   ZWJ,     O,     O,
+  /* 2010 */    GB,    GB,    GB,    GB,    GB,     O,     O,     O,
+
+#define use_offset_0x2060u 2760
+
+  /* 2060 */    WJ,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Superscripts and Subscripts */
+
+  /* 2070 */     O,     O,     O,     O, FMPst,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 2080 */     O,     O, FMPst, FMPst, FMPst,     O,     O,     O,
+
+#define use_offset_0x20f0u 2800
+
+
+  /* Combining Diacritical Marks for Symbols */
+
+  /* 20F0 */ VMAbv,     O,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x25c8u 2808
+
+
+  /* Geometric Shapes */
+                                                                         O,     O,     O,     O,    GB,     O,     O,     O,
+
+#define use_offset_0xa800u 2816
+
+
+  /* Syloti Nagri */
+
+  /* A800 */     B,     B,  VAbv,     B,     B,     B,     H,     B,     B,     B,     B, VMAbv,     B,     B,     B,     B,
+  /* A810 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A820 */     B,     B,     B,  VPst,  VPst,  VBlw,  VAbv,  VPst,     O,     O,     O,     O,  VBlw,     O,     O,     O,
+  /* A830 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Phags-pa */
+
+  /* A840 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A850 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A860 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A870 */     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Saurashtra */
+
+  /* A880 */ VMPst, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A890 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A8A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A8B0 */     B,     B,     B,     B,  MPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,  VPst,
+  /* A8C0 */  VPst,  VPst,  VPst,  VPst,     H, VMAbv,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* A8D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Devanagari Extended */
+
+  /* A8E0 */ VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv,
+  /* A8F0 */ VMAbv, VMAbv,     B,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     B,  VAbv,
+
+  /* Kayah Li */
+
+  /* A900 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A910 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A920 */     B,     B,     B,     B,     B,     B,  VAbv,  VAbv,  VAbv,  VAbv,  VAbv, VMBlw, VMBlw, VMBlw,     O,     O,
+
+  /* Rejang */
+
+  /* A930 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A940 */     B,     B,     B,     B,     B,     B,     B,  VBlw,  VBlw,  VBlw,  VAbv,  VBlw,  VBlw,  VBlw,  VBlw,  FAbv,
+  /* A950 */  FAbv,  FAbv,  FPst,  VPst,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* A960 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* A970 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Javanese */
+
+  /* A980 */ VMAbv, VMAbv,  FAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A990 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A9A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A9B0 */     B,     B,     B, CMAbv,  VPst,  VPst,  VAbv,  VAbv,  VBlw,  VBlw,  VPre,  VPre,  VAbv,  MBlw,  MBlw,  MBlw,
+  /* A9C0 */     H,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* A9D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Myanmar Extended-B */
+
+  /* A9E0 */     B,     B,     B,     B,     B,  VAbv,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* A9F0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,
+
+  /* Cham */
+
+  /* AA00 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AA10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AA20 */     B,     B,     B,     B,     B,     B,     B,     B,     B, VMAbv,  VAbv,  VAbv,  VAbv,  VBlw,  VAbv,  VPre,
+  /* AA30 */  VPre,  VAbv,  VBlw,  MPst,  MPre,  MBlw,  MBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* AA40 */     B,     B,     B,  FAbv,     B,     B,     B,     B,     B,     B,     B,     B,  FAbv,  FPst,     O,     O,
+  /* AA50 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Myanmar Extended-A */
+
+  /* AA60 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AA70 */     O,     B,     B,     B,    GB,    GB,    GB,     O,     O,     O,     B, VMPst, VMAbv, VMPst,     B,     B,
+
+  /* Tai Viet */
+
+  /* AA80 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AA90 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AAA0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* AAB0 */  VAbv,     B,  VAbv,  VAbv,  VBlw,     B,     B,  VAbv,  VAbv,     B,     B,     B,     B,     B,  VAbv, VMAbv,
+  /* AAC0 */     B, VMAbv,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* AAD0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Meetei Mayek Extensions */
+
+  /* AAE0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPre,  VBlw,  VAbv,  VPre,  VPst,
+  /* AAF0 */     O,     O,     O,     O,     O, VMPst,     H,     O,
+
+#define use_offset_0xabc0u 3576
+
+
+  /* Meetei Mayek */
+
+  /* ABC0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* ABD0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* ABE0 */     B,     B,     B,  VPst,  VPst,  VAbv,  VPst,  VPst,  VBlw,  VPst,  VPst,     O, VMPst,  VBlw,     O,     O,
+  /* ABF0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0xfe00u 3640
+
+
+  /* Variation Selectors */
+
+  /* FE00 */    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,    VS,
+
+#define use_offset_0x10a00u 3656
+
+
+  /* Kharoshthi */
+
+  /* 10A00 */     B,  VBlw,  VBlw,  VBlw,     O,  VAbv,  VBlw,     O,     O,     O,     O,     O,  VBlw,  VBlw, VMBlw, VMAbv,
+  /* 10A10 */     B,     B,     B,     B,     O,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,
+  /* 10A20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 10A30 */     B,     B,     B,     B,     B,     B,     O,     O, CMAbv, CMBlw, CMBlw,     O,     O,     O,     O,     H,
+  /* 10A40 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x11000u 3736
+
+
+  /* Brahmi */
+
+  /* 11000 */ VMPst, VMAbv, VMPst,    CS,    CS,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11010 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11020 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11030 */     B,     B,     B,     B,     B,     B,     B,     B,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,  VBlw,  VBlw,  VBlw,
+  /* 11040 */  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,   HVM,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11050 */     O,     O,     N,     N,     N,     N,     N,     N,     N,     N,     N,     N,     N,     N,     N,     N,
+  /* 11060 */     N,     N,     N,     N,     N,     N,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11070 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,    HN,
+
+  /* Kaithi */
+
+  /* 11080 */ VMAbv, VMAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11090 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 110A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 110B0 */  VPst,  VPre,  VPst,  VBlw,  VBlw,  VAbv,  VAbv,  VPst,  VPst,     H, CMBlw,     O,     O,     O,     O,     O,
+
+#define use_offset_0x11100u 3928
+
+
+  /* Chakma */
+
+  /* 11100 */ VMAbv, VMAbv, VMAbv,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11110 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11120 */     B,     B,     B,     B,     B,     B,     B,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VPre,  VBlw,  VAbv,  VAbv,
+  /* 11130 */  VBlw,  VAbv,  VAbv,     H, CMBlw,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11140 */     O,     O,     O,     O,     B,  VPst,  VPst,     B,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Mahajani */
+
+  /* 11150 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11160 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11170 */     B,     B,     B, CMBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Sharada */
+
+  /* 11180 */ VMAbv, VMAbv, VMPst,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11190 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 111A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 111B0 */     B,     B,     B,  VPst,  VPre,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,
+  /* 111C0 */     H,     B,     R,     R,     O,     O,     O,     O,    GB, FMBlw, CMBlw,  VAbv,  VBlw,     O,  VPre, VMAbv,
+  /* 111D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Sinhala Archaic Numbers */
+
+  /* 111E0 */     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 111F0 */     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Khojki */
+
+  /* 11200 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11210 */     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11220 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,  VPst,  VPst,  VBlw,
+  /* 11230 */  VAbv,  VAbv,  VAbv,  VAbv, VMAbv,     H, CMAbv, CMAbv,     O,     O,     O,     O,     O,     O, VMAbv,     O,
+
+#define use_offset_0x11280u 4248
+
+
+  /* Multani */
+
+  /* 11280 */     B,     B,     B,     B,     B,     B,     B,     O,     B,     O,     B,     B,     B,     B,     O,     B,
+  /* 11290 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,
+  /* 112A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Khudawadi */
+
+  /* 112B0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 112C0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 112D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B, VMAbv,
+  /* 112E0 */  VPst,  VPre,  VPst,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv, CMBlw,  VBlw,     O,     O,     O,     O,     O,
+  /* 112F0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Grantha */
+
+  /* 11300 */ VMAbv, VMAbv, VMAbv, VMAbv,     O,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     B,
+  /* 11310 */     B,     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11320 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 11330 */     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,     O, CMBlw, CMBlw,     B,  VPst,  VPst,
+  /* 11340 */  VAbv,  VPst,  VPst,  VPst,  VPst,     O,     O,  VPre,  VPre,     O,     O,  VPst,  VPst,   HVM,     O,     O,
+  /* 11350 */     O,     O,     O,     O,     O,     O,     O,  VPst,     O,     O,     O,     O,     O,     O,     B,     B,
+  /* 11360 */     B,     B,  VPst,  VPst,     O,     O, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv, VMAbv,     O,     O,     O,
+  /* 11370 */ VMAbv, VMAbv, VMAbv, VMAbv, VMAbv,     O,     O,     O,
+
+#define use_offset_0x11400u 4496
+
+
+  /* Newa */
+
+  /* 11400 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11410 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11420 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11430 */     B,     B,     B,     B,     B,  VPst,  VPre,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,
+  /* 11440 */  VPst,  VPst,     H, VMAbv, VMAbv, VMPst, CMBlw,     B,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11450 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O, FMAbv,     B,
+  /* 11460 */    CS,    CS,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11470 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Tirhuta */
+
+  /* 11480 */     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11490 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 114A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 114B0 */  VPst,  VPre,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VPre,  VAbv,  VPst,  VPst,  VPst,  VPst, VMAbv,
+  /* 114C0 */ VMAbv, VMAbv,     H, CMBlw,     B,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 114D0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x11580u 4720
+
+
+  /* Siddham */
+
+  /* 11580 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11590 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 115A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,
+  /* 115B0 */  VPre,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,     O,     O,  VPre,  VPst,  VPst,  VPst, VMAbv, VMAbv, VMPst,     H,
+  /* 115C0 */ CMBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 115D0 */     O,     O,     O,     O,     O,     O,     O,     O,     B,     B,     B,     B,  VBlw,  VBlw,     O,     O,
+  /* 115E0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 115F0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Modi */
+
+  /* 11600 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11610 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11620 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11630 */  VPst,  VPst,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VPst,  VPst, VMAbv, VMPst,     H,
+  /* 11640 */  VAbv,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11650 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+  /* 11660 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11670 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Takri */
+
+  /* 11680 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11690 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 116A0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B, VMAbv, VMPst,  VAbv,  VPre,  VPst,
+  /* 116B0 */  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,     H, CMBlw,     B,     O,     O,     O,     O,     O,     O,     O,
+  /* 116C0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+  /* 116D0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 116E0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 116F0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Ahom */
+
+  /* 11700 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11710 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,  MBlw,  MPre,  MAbv,
+  /* 11720 */  VPst,  VPst,  VAbv,  VAbv,  VBlw,  VBlw,  VPre,  VAbv,  VBlw,  VAbv,  VAbv,  VAbv,     O,     O,     O,     O,
+  /* 11730 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,
+
+#define use_offset_0x11800u 5168
+
+
+  /* Dogra */
+
+  /* 11800 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11810 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11820 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,  VPre,  VPst,  VBlw,
+  /* 11830 */  VBlw,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv, VMAbv, VMPst,     H, CMBlw,     O,     O,     O,     O,     O,
+
+#define use_offset_0x11900u 5232
+
+
+  /* Dives Akuru */
+
+  /* 11900 */     B,     B,     B,     B,     B,     B,     B,     O,     O,     B,     O,     O,     B,     B,     B,     B,
+  /* 11910 */     B,     B,     B,     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11920 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11930 */  VPst,  VPst,  VPst,  VPst,  VPst,  VPre,     O,  VPre,  VPst,     O,     O, VMAbv, VMAbv,  VPst,     H,     R,
+  /* 11940 */  MPst,     R,  MBlw, CMBlw,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11950 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x119a0u 5328
+
+
+  /* Nandinagari */
+
+  /* 119A0 */     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     B,     B,     B,     B,     B,     B,
+  /* 119B0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 119C0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 119D0 */     B,  VPst,  VPre,  VPst,  VBlw,  VBlw,  VBlw,  VBlw,     O,     O,  VAbv,  VAbv,  VPst,  VPst, VMPst, VMPst,
+  /* 119E0 */     H,     B,     O,     O,  VPre,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 119F0 */     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Zanabazar Square */
+
+  /* 11A00 */     B,  VAbv,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,     B,     B,     B,     B,     B,
+  /* 11A10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11A20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11A30 */     B,     B,     B, FMBlw,  VBlw, VMAbv, VMAbv, VMAbv, VMAbv, VMPst,     R,  MBlw,  MBlw,  MBlw,  MBlw,    GB,
+  /* 11A40 */     O,     O,     O,     O,     O,    GB,     O,     H,     O,     O,     O,     O,     O,     O,     O,     O,
+
+  /* Soyombo */
+
+  /* 11A50 */     B,  VAbv,  VBlw,  VBlw,  VAbv,  VAbv,  VAbv,  VPst,  VPst,  VBlw,  VBlw,  VBlw,     B,     B,     B,     B,
+  /* 11A60 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11A70 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11A80 */     B,     B,     B,     B,     R,     R,     R,     R,     R,     R,  FBlw,  FBlw,  FBlw,  FBlw,  FBlw,  FBlw,
+  /* 11A90 */  FBlw,  FBlw,  FBlw,  FBlw,  FBlw,  FBlw, VMAbv, VMPst, CMAbv,     H,     O,     O,     O,     B,     O,     O,
+
+#define use_offset_0x11c00u 5584
+
+
+  /* Bhaiksuki */
+
+  /* 11C00 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 11C10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11C20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,
+  /* 11C30 */  VAbv,  VAbv,  VBlw,  VBlw,  VBlw,  VBlw,  VBlw,     O,  VAbv,  VAbv,  VAbv,  VAbv, VMAbv, VMAbv, VMPst,     H,
+  /* 11C40 */     B,     O,     O,     O,    GB,    GB,     O,     O,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11C50 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11C60 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,
+
+  /* Marchen */
+
+  /* 11C70 */     O,     O,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11C80 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11C90 */     O,     O,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,
+  /* 11CA0 */   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,     O,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,   SUB,
+  /* 11CB0 */  VBlw,  VPre,  VBlw,  VAbv,  VPst, VMAbv, VMAbv,     O,
+
+#define use_offset_0x11d00u 5768
+
+
+  /* Masaram Gondi */
+
+  /* 11D00 */     B,     B,     B,     B,     B,     B,     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,
+  /* 11D10 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11D20 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11D30 */     B,  VAbv,  VAbv,  VAbv,  VAbv,  VAbv,  VBlw,     O,     O,     O,  VAbv,     O,  VAbv,  VAbv,     O,  VAbv,
+  /* 11D40 */ VMAbv, VMAbv, CMBlw,  VAbv,  VBlw,     H,     R,  MBlw,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11D50 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+  /* Gunjala Gondi */
+
+  /* 11D60 */     B,     B,     B,     B,     B,     B,     O,     B,     B,     O,     B,     B,     B,     B,     B,     B,
+  /* 11D70 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11D80 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,  VPst,  VPst,  VPst,  VPst,  VPst,     O,
+  /* 11D90 */  VAbv,  VAbv,     O,  VPst,  VPst, VMAbv, VMPst,     H,     O,     O,     O,     O,     O,     O,     O,     O,
+  /* 11DA0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     O,     O,     O,     O,     O,     O,
+
+#define use_offset_0x11ee0u 5944
+
+
+  /* Makasar */
+
+  /* 11EE0 */     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,     B,
+  /* 11EF0 */     B,     B,    GB,  VAbv,  VBlw,  VPre,  VPst,     O,
+
+}; /* Table items: 5968; occupancy: 74% */
+
+USE_TABLE_ELEMENT_TYPE
+hb_use_get_category (hb_codepoint_t u)
+{
+  switch (u >> 12)
+  {
+    case 0x0u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x0028u, 0x003Fu)) return use_table[u - 0x0028u + use_offset_0x0028u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x00A0u, 0x00D7u)) return use_table[u - 0x00A0u + use_offset_0x00a0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x0348u, 0x034Fu)) return use_table[u - 0x0348u + use_offset_0x0348u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x0900u, 0x0DF7u)) return use_table[u - 0x0900u + use_offset_0x0900u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x0F18u, 0x0FC7u)) return use_table[u - 0x0F18u + use_offset_0x0f18u];
+      break;
+
+    case 0x1u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x1000u, 0x109Fu)) return use_table[u - 0x1000u + use_offset_0x1000u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1700u, 0x17EFu)) return use_table[u - 0x1700u + use_offset_0x1700u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1900u, 0x1A9Fu)) return use_table[u - 0x1900u + use_offset_0x1900u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1B00u, 0x1C4Fu)) return use_table[u - 0x1B00u + use_offset_0x1b00u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1CD0u, 0x1CFFu)) return use_table[u - 0x1CD0u + use_offset_0x1cd0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x1DF8u, 0x1DFFu)) return use_table[u - 0x1DF8u + use_offset_0x1df8u];
+      break;
+
+    case 0x2u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x2008u, 0x2017u)) return use_table[u - 0x2008u + use_offset_0x2008u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x2060u, 0x2087u)) return use_table[u - 0x2060u + use_offset_0x2060u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x20F0u, 0x20F7u)) return use_table[u - 0x20F0u + use_offset_0x20f0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x25C8u, 0x25CFu)) return use_table[u - 0x25C8u + use_offset_0x25c8u];
+      break;
+
+    case 0xAu:
+      if (hb_in_range<hb_codepoint_t> (u, 0xA800u, 0xAAF7u)) return use_table[u - 0xA800u + use_offset_0xa800u];
+      if (hb_in_range<hb_codepoint_t> (u, 0xABC0u, 0xABFFu)) return use_table[u - 0xABC0u + use_offset_0xabc0u];
+      break;
+
+    case 0xFu:
+      if (hb_in_range<hb_codepoint_t> (u, 0xFE00u, 0xFE0Fu)) return use_table[u - 0xFE00u + use_offset_0xfe00u];
+      break;
+
+    case 0x10u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x10A00u, 0x10A4Fu)) return use_table[u - 0x10A00u + use_offset_0x10a00u];
+      break;
+
+    case 0x11u:
+      if (hb_in_range<hb_codepoint_t> (u, 0x11000u, 0x110BFu)) return use_table[u - 0x11000u + use_offset_0x11000u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11100u, 0x1123Fu)) return use_table[u - 0x11100u + use_offset_0x11100u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11280u, 0x11377u)) return use_table[u - 0x11280u + use_offset_0x11280u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11400u, 0x114DFu)) return use_table[u - 0x11400u + use_offset_0x11400u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11580u, 0x1173Fu)) return use_table[u - 0x11580u + use_offset_0x11580u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11800u, 0x1183Fu)) return use_table[u - 0x11800u + use_offset_0x11800u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11900u, 0x1195Fu)) return use_table[u - 0x11900u + use_offset_0x11900u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x119A0u, 0x11A9Fu)) return use_table[u - 0x119A0u + use_offset_0x119a0u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11C00u, 0x11CB7u)) return use_table[u - 0x11C00u + use_offset_0x11c00u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11D00u, 0x11DAFu)) return use_table[u - 0x11D00u + use_offset_0x11d00u];
+      if (hb_in_range<hb_codepoint_t> (u, 0x11EE0u, 0x11EF7u)) return use_table[u - 0x11EE0u + use_offset_0x11ee0u];
+      break;
+
+    default:
+      break;
+  }
+  return USE_O;
+}
+
+#undef B
+#undef CGJ
+#undef CS
+#undef GB
+#undef H
+#undef HN
+#undef HVM
+#undef IND
+#undef N
+#undef O
+#undef R
+#undef Rsv
+#undef S
+#undef SUB
+#undef Sk
+#undef VS
+#undef WJ
+#undef ZWJ
+#undef ZWNJ
+#undef CMAbv
+#undef CMBlw
+#undef FAbv
+#undef FBlw
+#undef FPst
+#undef FMAbv
+#undef FMBlw
+#undef FMPst
+#undef MAbv
+#undef MBlw
+#undef MPst
+#undef MPre
+#undef SMAbv
+#undef SMBlw
+#undef VAbv
+#undef VBlw
+#undef VPst
+#undef VPre
+#undef VMAbv
+#undef VMBlw
+#undef VMPst
+#undef VMPre
+
+
+#endif
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.cc
new file mode 100644
index 0000000000..a1e25bdd80
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.cc
@@ -0,0 +1,569 @@
+/*
+ * Copyright © 2015  Mozilla Foundation.
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-use.hh"
+#include "hb-ot-shape-complex-arabic.hh"
+#include "hb-ot-shape-complex-arabic-joining-list.hh"
+#include "hb-ot-shape-complex-vowel-constraints.hh"
+
+/* buffer var allocations */
+#define use_category() complex_var_u8_1()
+
+
+/*
+ * Universal Shaping Engine.
+ * https://docs.microsoft.com/en-us/typography/script-development/use
+ */
+
+static const hb_tag_t
+use_basic_features[] =
+{
+  /*
+   * Basic features.
+   * These features are applied all at once, before reordering.
+   */
+  HB_TAG('r','k','r','f'),
+  HB_TAG('a','b','v','f'),
+  HB_TAG('b','l','w','f'),
+  HB_TAG('h','a','l','f'),
+  HB_TAG('p','s','t','f'),
+  HB_TAG('v','a','t','u'),
+  HB_TAG('c','j','c','t'),
+};
+static const hb_tag_t
+use_topographical_features[] =
+{
+  HB_TAG('i','s','o','l'),
+  HB_TAG('i','n','i','t'),
+  HB_TAG('m','e','d','i'),
+  HB_TAG('f','i','n','a'),
+};
+/* Same order as use_topographical_features. */
+enum joining_form_t {
+  USE_ISOL,
+  USE_INIT,
+  USE_MEDI,
+  USE_FINA,
+  _USE_NONE
+};
+static const hb_tag_t
+use_other_features[] =
+{
+  /*
+   * Other features.
+   * These features are applied all at once, after reordering and
+   * clearing syllables.
+   */
+  HB_TAG('a','b','v','s'),
+  HB_TAG('b','l','w','s'),
+  HB_TAG('h','a','l','n'),
+  HB_TAG('p','r','e','s'),
+  HB_TAG('p','s','t','s'),
+};
+
+static void
+setup_syllables_use (const hb_ot_shape_plan_t *plan,
+		     hb_font_t *font,
+		     hb_buffer_t *buffer);
+static void
+record_rphf_use (const hb_ot_shape_plan_t *plan,
+		 hb_font_t *font,
+		 hb_buffer_t *buffer);
+static void
+record_pref_use (const hb_ot_shape_plan_t *plan,
+		 hb_font_t *font,
+		 hb_buffer_t *buffer);
+static void
+reorder_use (const hb_ot_shape_plan_t *plan,
+	     hb_font_t *font,
+	     hb_buffer_t *buffer);
+
+static void
+collect_features_use (hb_ot_shape_planner_t *plan)
+{
+  hb_ot_map_builder_t *map = &plan->map;
+
+  /* Do this before any lookups have been applied. */
+  map->add_gsub_pause (setup_syllables_use);
+
+  /* "Default glyph pre-processing group" */
+  map->enable_feature (HB_TAG('l','o','c','l'));
+  map->enable_feature (HB_TAG('c','c','m','p'));
+  map->enable_feature (HB_TAG('n','u','k','t'));
+  map->enable_feature (HB_TAG('a','k','h','n'), F_MANUAL_ZWJ);
+
+  /* "Reordering group" */
+  map->add_gsub_pause (_hb_clear_substitution_flags);
+  map->add_feature (HB_TAG('r','p','h','f'), F_MANUAL_ZWJ);
+  map->add_gsub_pause (record_rphf_use);
+  map->add_gsub_pause (_hb_clear_substitution_flags);
+  map->enable_feature (HB_TAG('p','r','e','f'), F_MANUAL_ZWJ);
+  map->add_gsub_pause (record_pref_use);
+
+  /* "Orthographic unit shaping group" */
+  for (unsigned int i = 0; i < ARRAY_LENGTH (use_basic_features); i++)
+    map->enable_feature (use_basic_features[i], F_MANUAL_ZWJ);
+
+  map->add_gsub_pause (reorder_use);
+  map->add_gsub_pause (_hb_clear_syllables);
+
+  /* "Topographical features" */
+  for (unsigned int i = 0; i < ARRAY_LENGTH (use_topographical_features); i++)
+    map->add_feature (use_topographical_features[i]);
+  map->add_gsub_pause (nullptr);
+
+  /* "Standard typographic presentation" */
+  for (unsigned int i = 0; i < ARRAY_LENGTH (use_other_features); i++)
+    map->enable_feature (use_other_features[i], F_MANUAL_ZWJ);
+}
+
+struct use_shape_plan_t
+{
+  hb_mask_t rphf_mask;
+
+  arabic_shape_plan_t *arabic_plan;
+};
+
+static void *
+data_create_use (const hb_ot_shape_plan_t *plan)
+{
+  use_shape_plan_t *use_plan = (use_shape_plan_t *) calloc (1, sizeof (use_shape_plan_t));
+  if (unlikely (!use_plan))
+    return nullptr;
+
+  use_plan->rphf_mask = plan->map.get_1_mask (HB_TAG('r','p','h','f'));
+
+  if (has_arabic_joining (plan->props.script))
+  {
+    use_plan->arabic_plan = (arabic_shape_plan_t *) data_create_arabic (plan);
+    if (unlikely (!use_plan->arabic_plan))
+    {
+      free (use_plan);
+      return nullptr;
+    }
+  }
+
+  return use_plan;
+}
+
+static void
+data_destroy_use (void *data)
+{
+  use_shape_plan_t *use_plan = (use_shape_plan_t *) data;
+
+  if (use_plan->arabic_plan)
+    data_destroy_arabic (use_plan->arabic_plan);
+
+  free (data);
+}
+
+enum use_syllable_type_t {
+  use_independent_cluster,
+  use_virama_terminated_cluster,
+  use_sakot_terminated_cluster,
+  use_standard_cluster,
+  use_number_joiner_terminated_cluster,
+  use_numeral_cluster,
+  use_symbol_cluster,
+  use_broken_cluster,
+  use_non_cluster,
+};
+
+#include "hb-ot-shape-complex-use-machine.hh"
+
+
+static void
+setup_masks_use (const hb_ot_shape_plan_t *plan,
+		 hb_buffer_t              *buffer,
+		 hb_font_t                *font HB_UNUSED)
+{
+  const use_shape_plan_t *use_plan = (const use_shape_plan_t *) plan->data;
+
+  /* Do this before allocating use_category(). */
+  if (use_plan->arabic_plan)
+  {
+    setup_masks_arabic_plan (use_plan->arabic_plan, buffer, plan->props.script);
+  }
+
+  HB_BUFFER_ALLOCATE_VAR (buffer, use_category);
+
+  /* We cannot setup masks here.  We save information about characters
+   * and setup masks later on in a pause-callback. */
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    info[i].use_category() = hb_use_get_category (info[i].codepoint);
+}
+
+static void
+setup_rphf_mask (const hb_ot_shape_plan_t *plan,
+		 hb_buffer_t *buffer)
+{
+  const use_shape_plan_t *use_plan = (const use_shape_plan_t *) plan->data;
+
+  hb_mask_t mask = use_plan->rphf_mask;
+  if (!mask) return;
+
+  hb_glyph_info_t *info = buffer->info;
+
+  foreach_syllable (buffer, start, end)
+  {
+    unsigned int limit = info[start].use_category() == USE_R ? 1 : hb_min (3u, end - start);
+    for (unsigned int i = start; i < start + limit; i++)
+      info[i].mask |= mask;
+  }
+}
+
+static void
+setup_topographical_masks (const hb_ot_shape_plan_t *plan,
+			   hb_buffer_t *buffer)
+{
+  const use_shape_plan_t *use_plan = (const use_shape_plan_t *) plan->data;
+  if (use_plan->arabic_plan)
+    return;
+
+  static_assert ((USE_INIT < 4 && USE_ISOL < 4 && USE_MEDI < 4 && USE_FINA < 4), "");
+  hb_mask_t masks[4], all_masks = 0;
+  for (unsigned int i = 0; i < 4; i++)
+  {
+    masks[i] = plan->map.get_1_mask (use_topographical_features[i]);
+    if (masks[i] == plan->map.get_global_mask ())
+      masks[i] = 0;
+    all_masks |= masks[i];
+  }
+  if (!all_masks)
+    return;
+  hb_mask_t other_masks = ~all_masks;
+
+  unsigned int last_start = 0;
+  joining_form_t last_form = _USE_NONE;
+  hb_glyph_info_t *info = buffer->info;
+  foreach_syllable (buffer, start, end)
+  {
+    use_syllable_type_t syllable_type = (use_syllable_type_t) (info[start].syllable() & 0x0F);
+    switch (syllable_type)
+    {
+      case use_independent_cluster:
+      case use_symbol_cluster:
+      case use_non_cluster:
+	/* These don't join.  Nothing to do. */
+	last_form = _USE_NONE;
+	break;
+
+      case use_virama_terminated_cluster:
+      case use_sakot_terminated_cluster:
+      case use_standard_cluster:
+      case use_number_joiner_terminated_cluster:
+      case use_numeral_cluster:
+      case use_broken_cluster:
+
+	bool join = last_form == USE_FINA || last_form == USE_ISOL;
+
+	if (join)
+	{
+	  /* Fixup previous syllable's form. */
+	  last_form = last_form == USE_FINA ? USE_MEDI : USE_INIT;
+	  for (unsigned int i = last_start; i < start; i++)
+	    info[i].mask = (info[i].mask & other_masks) | masks[last_form];
+	}
+
+	/* Form for this syllable. */
+	last_form = join ? USE_FINA : USE_ISOL;
+	for (unsigned int i = start; i < end; i++)
+	  info[i].mask = (info[i].mask & other_masks) | masks[last_form];
+
+	break;
+    }
+
+    last_start = start;
+  }
+}
+
+static void
+setup_syllables_use (const hb_ot_shape_plan_t *plan,
+		     hb_font_t *font HB_UNUSED,
+		     hb_buffer_t *buffer)
+{
+  find_syllables_use (buffer);
+  foreach_syllable (buffer, start, end)
+    buffer->unsafe_to_break (start, end);
+  setup_rphf_mask (plan, buffer);
+  setup_topographical_masks (plan, buffer);
+}
+
+static void
+record_rphf_use (const hb_ot_shape_plan_t *plan,
+		 hb_font_t *font HB_UNUSED,
+		 hb_buffer_t *buffer)
+{
+  const use_shape_plan_t *use_plan = (const use_shape_plan_t *) plan->data;
+
+  hb_mask_t mask = use_plan->rphf_mask;
+  if (!mask) return;
+  hb_glyph_info_t *info = buffer->info;
+
+  foreach_syllable (buffer, start, end)
+  {
+    /* Mark a substituted repha as USE_R. */
+    for (unsigned int i = start; i < end && (info[i].mask & mask); i++)
+      if (_hb_glyph_info_substituted (&info[i]))
+      {
+	info[i].use_category() = USE_R;
+	break;
+      }
+  }
+}
+
+static void
+record_pref_use (const hb_ot_shape_plan_t *plan HB_UNUSED,
+		 hb_font_t *font HB_UNUSED,
+		 hb_buffer_t *buffer)
+{
+  hb_glyph_info_t *info = buffer->info;
+
+  foreach_syllable (buffer, start, end)
+  {
+    /* Mark a substituted pref as VPre, as they behave the same way. */
+    for (unsigned int i = start; i < end; i++)
+      if (_hb_glyph_info_substituted (&info[i]))
+      {
+	info[i].use_category() = USE_VPre;
+	break;
+      }
+  }
+}
+
+static inline bool
+is_halant_use (const hb_glyph_info_t &info)
+{
+  return (info.use_category() == USE_H || info.use_category() == USE_HVM) &&
+	 !_hb_glyph_info_ligated (&info);
+}
+
+static void
+reorder_syllable_use (hb_buffer_t *buffer, unsigned int start, unsigned int end)
+{
+  use_syllable_type_t syllable_type = (use_syllable_type_t) (buffer->info[start].syllable() & 0x0F);
+  /* Only a few syllable types need reordering. */
+  if (unlikely (!(FLAG_UNSAFE (syllable_type) &
+		  (FLAG (use_virama_terminated_cluster) |
+		   FLAG (use_sakot_terminated_cluster) |
+		   FLAG (use_standard_cluster) |
+		   FLAG (use_broken_cluster) |
+		   0))))
+    return;
+
+  hb_glyph_info_t *info = buffer->info;
+
+#define POST_BASE_FLAGS64 (FLAG64 (USE_FM) | \
+			   FLAG64 (USE_FAbv) | \
+			   FLAG64 (USE_FBlw) | \
+			   FLAG64 (USE_FPst) | \
+			   FLAG64 (USE_MAbv) | \
+			   FLAG64 (USE_MBlw) | \
+			   FLAG64 (USE_MPst) | \
+			   FLAG64 (USE_MPre) | \
+			   FLAG64 (USE_VAbv) | \
+			   FLAG64 (USE_VBlw) | \
+			   FLAG64 (USE_VPst) | \
+			   FLAG64 (USE_VPre) | \
+			   FLAG64 (USE_VMAbv) | \
+			   FLAG64 (USE_VMBlw) | \
+			   FLAG64 (USE_VMPst) | \
+			   FLAG64 (USE_VMPre))
+
+  /* Move things forward. */
+  if (info[start].use_category() == USE_R && end - start > 1)
+  {
+    /* Got a repha.  Reorder it towards the end, but before the first post-base
+     * glyph. */
+    for (unsigned int i = start + 1; i < end; i++)
+    {
+      bool is_post_base_glyph = (FLAG64_UNSAFE (info[i].use_category()) & POST_BASE_FLAGS64) ||
+				is_halant_use (info[i]);
+      if (is_post_base_glyph || i == end - 1)
+      {
+	/* If we hit a post-base glyph, move before it; otherwise move to the
+	 * end. Shift things in between backward. */
+
+	if (is_post_base_glyph)
+	  i--;
+
+	buffer->merge_clusters (start, i + 1);
+	hb_glyph_info_t t = info[start];
+	memmove (&info[start], &info[start + 1], (i - start) * sizeof (info[0]));
+	info[i] = t;
+
+	break;
+      }
+    }
+  }
+
+  /* Move things back. */
+  unsigned int j = start;
+  for (unsigned int i = start; i < end; i++)
+  {
+    uint32_t flag = FLAG_UNSAFE (info[i].use_category());
+    if (is_halant_use (info[i]))
+    {
+      /* If we hit a halant, move after it; otherwise move to the beginning, and
+       * shift things in between forward. */
+      j = i + 1;
+    }
+    else if (((flag) & (FLAG (USE_VPre) | FLAG (USE_VMPre))) &&
+	     /* Only move the first component of a MultipleSubst. */
+	     0 == _hb_glyph_info_get_lig_comp (&info[i]) &&
+	     j < i)
+    {
+      buffer->merge_clusters (j, i + 1);
+      hb_glyph_info_t t = info[i];
+      memmove (&info[j + 1], &info[j], (i - j) * sizeof (info[0]));
+      info[j] = t;
+    }
+  }
+}
+
+static inline void
+insert_dotted_circles_use (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			   hb_font_t *font,
+			   hb_buffer_t *buffer)
+{
+  if (unlikely (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE))
+    return;
+
+  /* Note: This loop is extra overhead, but should not be measurable.
+   * TODO Use a buffer scratch flag to remove the loop. */
+  bool has_broken_syllables = false;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if ((info[i].syllable() & 0x0F) == use_broken_cluster)
+    {
+      has_broken_syllables = true;
+      break;
+    }
+  if (likely (!has_broken_syllables))
+    return;
+
+  hb_glyph_info_t dottedcircle = {0};
+  if (!font->get_nominal_glyph (0x25CCu, &dottedcircle.codepoint))
+    return;
+  dottedcircle.use_category() = hb_use_get_category (0x25CC);
+
+  buffer->clear_output ();
+
+  buffer->idx = 0;
+  unsigned int last_syllable = 0;
+  while (buffer->idx < buffer->len && buffer->successful)
+  {
+    unsigned int syllable = buffer->cur().syllable();
+    use_syllable_type_t syllable_type = (use_syllable_type_t) (syllable & 0x0F);
+    if (unlikely (last_syllable != syllable && syllable_type == use_broken_cluster))
+    {
+      last_syllable = syllable;
+
+      hb_glyph_info_t ginfo = dottedcircle;
+      ginfo.cluster = buffer->cur().cluster;
+      ginfo.mask = buffer->cur().mask;
+      ginfo.syllable() = buffer->cur().syllable();
+
+      /* Insert dottedcircle after possible Repha. */
+      while (buffer->idx < buffer->len && buffer->successful &&
+	     last_syllable == buffer->cur().syllable() &&
+	     buffer->cur().use_category() == USE_R)
+	buffer->next_glyph ();
+
+      buffer->output_info (ginfo);
+    }
+    else
+      buffer->next_glyph ();
+  }
+  buffer->swap_buffers ();
+}
+
+static void
+reorder_use (const hb_ot_shape_plan_t *plan,
+	     hb_font_t *font,
+	     hb_buffer_t *buffer)
+{
+  insert_dotted_circles_use (plan, font, buffer);
+
+  foreach_syllable (buffer, start, end)
+    reorder_syllable_use (buffer, start, end);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, use_category);
+}
+
+
+static void
+preprocess_text_use (const hb_ot_shape_plan_t *plan,
+		     hb_buffer_t              *buffer,
+		     hb_font_t                *font)
+{
+  _hb_preprocess_text_vowel_constraints (plan, buffer, font);
+}
+
+static bool
+compose_use (const hb_ot_shape_normalize_context_t *c,
+	     hb_codepoint_t  a,
+	     hb_codepoint_t  b,
+	     hb_codepoint_t *ab)
+{
+  /* Avoid recomposing split matras. */
+  if (HB_UNICODE_GENERAL_CATEGORY_IS_MARK (c->unicode->general_category (a)))
+    return false;
+
+  return (bool)c->unicode->compose (a, b, ab);
+}
+
+
+const hb_ot_complex_shaper_t _hb_ot_complex_shaper_use =
+{
+  collect_features_use,
+  nullptr, /* override_features */
+  data_create_use,
+  data_destroy_use,
+  preprocess_text_use,
+  nullptr, /* postprocess_glyphs */
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
+  nullptr, /* decompose */
+  compose_use,
+  setup_masks_use,
+  HB_TAG_NONE, /* gpos_tag */
+  nullptr, /* reorder_marks */
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY,
+  false, /* fallback_position */
+};
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.hh
new file mode 100644
index 0000000000..ce6645ecd3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-use.hh
@@ -0,0 +1,105 @@
+/*
+ * Copyright © 2015  Mozilla Foundation.
+ * Copyright © 2015  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Mozilla Author(s): Jonathan Kew
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_USE_HH
+#define HB_OT_SHAPE_COMPLEX_USE_HH
+
+#include "hb.hh"
+
+
+#include "hb-ot-shape-complex.hh"
+
+
+#define USE_TABLE_ELEMENT_TYPE uint8_t
+
+/* Cateories used in the Universal Shaping Engine spec:
+ * https://docs.microsoft.com/en-us/typography/script-development/use
+ */
+/* Note: This enum is duplicated in the -machine.rl source file.
+ * Not sure how to avoid duplication. */
+enum use_category_t {
+  USE_O		= 0,	/* OTHER */
+
+  USE_B		= 1,	/* BASE */
+  USE_IND	= 3,	/* BASE_IND */
+  USE_N		= 4,	/* BASE_NUM */
+  USE_GB	= 5,	/* BASE_OTHER */
+  USE_CGJ	= 6,	/* CGJ */
+//  USE_F		= 7,	/* CONS_FINAL */
+  USE_FM	= 8,	/* CONS_FINAL_MOD */
+//  USE_M		= 9,	/* CONS_MED */
+//  USE_CM	= 10,	/* CONS_MOD */
+  USE_SUB	= 11,	/* CONS_SUB */
+  USE_H		= 12,	/* HALANT */
+
+  USE_HN	= 13,	/* HALANT_NUM */
+  USE_ZWNJ	= 14,	/* Zero width non-joiner */
+  USE_ZWJ	= 15,	/* Zero width joiner */
+  USE_WJ	= 16,	/* Word joiner */
+  USE_Rsv	= 17,	/* Reserved characters */
+  USE_R		= 18,	/* REPHA */
+  USE_S		= 19,	/* SYM */
+//  USE_SM	= 20,	/* SYM_MOD */
+  USE_VS	= 21,	/* VARIATION_SELECTOR */
+//  USE_V	= 36,	/* VOWEL */
+//  USE_VM	= 40,	/* VOWEL_MOD */
+  USE_CS	= 43,	/* CONS_WITH_STACKER */
+
+  /* https://github.com/harfbuzz/harfbuzz/issues/1102 */
+  USE_HVM	= 44,	/* HALANT_OR_VOWEL_MODIFIER */
+
+  USE_Sk	= 48,	/* SAKOT */
+
+  USE_FAbv	= 24,	/* CONS_FINAL_ABOVE */
+  USE_FBlw	= 25,	/* CONS_FINAL_BELOW */
+  USE_FPst	= 26,	/* CONS_FINAL_POST */
+  USE_MAbv	= 27,	/* CONS_MED_ABOVE */
+  USE_MBlw	= 28,	/* CONS_MED_BELOW */
+  USE_MPst	= 29,	/* CONS_MED_POST */
+  USE_MPre	= 30,	/* CONS_MED_PRE */
+  USE_CMAbv	= 31,	/* CONS_MOD_ABOVE */
+  USE_CMBlw	= 32,	/* CONS_MOD_BELOW */
+  USE_VAbv	= 33,	/* VOWEL_ABOVE / VOWEL_ABOVE_BELOW / VOWEL_ABOVE_BELOW_POST / VOWEL_ABOVE_POST */
+  USE_VBlw	= 34,	/* VOWEL_BELOW / VOWEL_BELOW_POST */
+  USE_VPst	= 35,	/* VOWEL_POST	UIPC = Right */
+  USE_VPre	= 22,	/* VOWEL_PRE / VOWEL_PRE_ABOVE / VOWEL_PRE_ABOVE_POST / VOWEL_PRE_POST */
+  USE_VMAbv	= 37,	/* VOWEL_MOD_ABOVE */
+  USE_VMBlw	= 38,	/* VOWEL_MOD_BELOW */
+  USE_VMPst	= 39,	/* VOWEL_MOD_POST */
+  USE_VMPre	= 23,	/* VOWEL_MOD_PRE */
+  USE_SMAbv	= 41,	/* SYM_MOD_ABOVE */
+  USE_SMBlw	= 42,	/* SYM_MOD_BELOW */
+  USE_FMAbv	= 45,	/* CONS_FINAL_MOD	UIPC = Top */
+  USE_FMBlw	= 46,	/* CONS_FINAL_MOD	UIPC = Bottom */
+  USE_FMPst	= 47,	/* CONS_FINAL_MOD	UIPC = Not_Applicable */
+};
+
+HB_INTERNAL USE_TABLE_ELEMENT_TYPE
+hb_use_get_category (hb_codepoint_t u);
+
+#endif /* HB_OT_SHAPE_COMPLEX_USE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.cc b/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.cc
new file mode 100644
index 0000000000..c3368c6ec2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.cc
@@ -0,0 +1,464 @@
+/* == Start of generated functions == */
+/*
+ * The following functions are generated by running:
+ *
+ *   ./gen-vowel-constraints.py ms-use/IndicShapingInvalidCluster.txt Scripts.txt
+ *
+ * on files with these headers:
+ *
+ * # IndicShapingInvalidCluster.txt
+ * # Date: 2015-03-12, 21:17:00 GMT [AG]
+ * # Date: 2019-11-08, 23:22:00 GMT [AG]
+ *
+ * # Scripts-13.0.0.txt
+ * # Date: 2020-01-22, 00:07:43 GMT
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-complex-vowel-constraints.hh"
+
+static void
+_output_dotted_circle (hb_buffer_t *buffer)
+{
+  hb_glyph_info_t &dottedcircle = buffer->output_glyph (0x25CCu);
+  _hb_glyph_info_reset_continuation (&dottedcircle);
+}
+
+static void
+_output_with_dotted_circle (hb_buffer_t *buffer)
+{
+  _output_dotted_circle (buffer);
+  buffer->next_glyph ();
+}
+
+void
+_hb_preprocess_text_vowel_constraints (const hb_ot_shape_plan_t *plan HB_UNUSED,
+				       hb_buffer_t              *buffer,
+				       hb_font_t                *font HB_UNUSED)
+{
+#ifdef HB_NO_OT_SHAPE_COMPLEX_VOWEL_CONSTRAINTS
+  return;
+#endif
+  if (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE)
+    return;
+
+  /* UGLY UGLY UGLY business of adding dotted-circle in the middle of
+   * vowel-sequences that look like another vowel.  Data for each script
+   * collected from the USE script development spec.
+   *
+   * https://github.com/harfbuzz/harfbuzz/issues/1019
+   */
+  bool processed = false;
+  buffer->clear_output ();
+  unsigned int count = buffer->len;
+  switch ((unsigned) buffer->props.script)
+  {
+    case HB_SCRIPT_DEVANAGARI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0905u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x093Au: case 0x093Bu: case 0x093Eu: case 0x0945u:
+	      case 0x0946u: case 0x0949u: case 0x094Au: case 0x094Bu:
+	      case 0x094Cu: case 0x094Fu: case 0x0956u: case 0x0957u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0906u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x093Au: case 0x0945u: case 0x0946u: case 0x0947u:
+	      case 0x0948u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0909u:
+	    matched = 0x0941u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x090Fu:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0945u: case 0x0946u: case 0x0947u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0930u:
+	    if (0x094Du == buffer->cur (1).codepoint &&
+		buffer->idx + 2 < count &&
+		0x0907u == buffer->cur (2).codepoint)
+	    {
+	      buffer->next_glyph ();
+	      matched = true;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_BENGALI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0985u:
+	    matched = 0x09BEu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x098Bu:
+	    matched = 0x09C3u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x098Cu:
+	    matched = 0x09E2u == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_GURMUKHI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0A05u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0A3Eu: case 0x0A48u: case 0x0A4Cu:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0A72u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0A3Fu: case 0x0A40u: case 0x0A47u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0A73u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0A41u: case 0x0A42u: case 0x0A4Bu:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_GUJARATI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0A85u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0ABEu: case 0x0AC5u: case 0x0AC7u: case 0x0AC8u:
+	      case 0x0AC9u: case 0x0ACBu: case 0x0ACCu:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0AC5u:
+	    matched = 0x0ABEu == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_ORIYA:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0B05u:
+	    matched = 0x0B3Eu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0B0Fu: case 0x0B13u:
+	    matched = 0x0B57u == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_TAMIL:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	if (0x0B85u == buffer->cur ().codepoint &&
+	    0x0BC2u == buffer->cur (1).codepoint)
+	{
+	  matched = true;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_TELUGU:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0C12u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0C4Cu: case 0x0C55u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0C3Fu: case 0x0C46u: case 0x0C4Au:
+	    matched = 0x0C55u == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_KANNADA:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0C89u: case 0x0C8Bu:
+	    matched = 0x0CBEu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0C92u:
+	    matched = 0x0CCCu == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_MALAYALAM:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0D07u: case 0x0D09u:
+	    matched = 0x0D57u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0D0Eu:
+	    matched = 0x0D46u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0D12u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0D3Eu: case 0x0D57u:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_SINHALA:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x0D85u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0DCFu: case 0x0DD0u: case 0x0DD1u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x0D8Bu: case 0x0D8Fu: case 0x0D94u:
+	    matched = 0x0DDFu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0D8Du:
+	    matched = 0x0DD8u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x0D91u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x0DCAu: case 0x0DD9u: case 0x0DDAu: case 0x0DDCu:
+	      case 0x0DDDu:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_BRAHMI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x11005u:
+	    matched = 0x11038u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x1100Bu:
+	    matched = 0x1103Eu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x1100Fu:
+	    matched = 0x11042u == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_KHUDAWADI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x112B0u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x112E0u: case 0x112E5u: case 0x112E6u: case 0x112E7u:
+	      case 0x112E8u:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_TIRHUTA:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x11481u:
+	    matched = 0x114B0u == buffer->cur (1).codepoint;
+	    break;
+	  case 0x1148Bu: case 0x1148Du:
+	    matched = 0x114BAu == buffer->cur (1).codepoint;
+	    break;
+	  case 0x114AAu:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x114B5u: case 0x114B6u:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_MODI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x11600u: case 0x11601u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x11639u: case 0x1163Au:
+		matched = true;
+		break;
+	    }
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    case HB_SCRIPT_TAKRI:
+      for (buffer->idx = 0; buffer->idx + 1 < count && buffer->successful;)
+      {
+	bool matched = false;
+	switch (buffer->cur ().codepoint)
+	{
+	  case 0x11680u:
+	    switch (buffer->cur (1).codepoint)
+	    {
+	      case 0x116ADu: case 0x116B4u: case 0x116B5u:
+		matched = true;
+		break;
+	    }
+	    break;
+	  case 0x11686u:
+	    matched = 0x116B2u == buffer->cur (1).codepoint;
+	    break;
+	}
+	buffer->next_glyph ();
+	if (matched) _output_with_dotted_circle (buffer);
+      }
+      processed = true;
+      break;
+
+    default:
+      break;
+  }
+  if (processed)
+  {
+    if (buffer->idx < count)
+      buffer->next_glyph ();
+    buffer->swap_buffers ();
+  }
+}
+
+
+#endif
+/* == End of generated functions == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.hh
new file mode 100644
index 0000000000..d9082d4ead
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex-vowel-constraints.hh
@@ -0,0 +1,39 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_VOWEL_CONSTRAINTS_HH
+#define HB_OT_SHAPE_COMPLEX_VOWEL_CONSTRAINTS_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape-complex.hh"
+
+HB_INTERNAL void
+_hb_preprocess_text_vowel_constraints (const hb_ot_shape_plan_t *plan,
+				       hb_buffer_t              *buffer,
+				       hb_font_t                *font);
+
+#endif /* HB_OT_SHAPE_COMPLEX_VOWEL_CONSTRAINTS_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-complex.hh b/thirdparty/harfbuzz/src/hb-ot-shape-complex.hh
new file mode 100644
index 0000000000..61f4c0e158
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-complex.hh
@@ -0,0 +1,402 @@
+/*
+ * Copyright © 2010,2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_COMPLEX_HH
+#define HB_OT_SHAPE_COMPLEX_HH
+
+#include "hb.hh"
+
+#include "hb-ot-layout.hh"
+#include "hb-ot-shape.hh"
+#include "hb-ot-shape-normalize.hh"
+
+
+/* buffer var allocations, used by complex shapers */
+#define complex_var_u8_0()	var2.u8[2]
+#define complex_var_u8_1()	var2.u8[3]
+
+
+#define HB_OT_SHAPE_COMPLEX_MAX_COMBINING_MARKS 32
+
+enum hb_ot_shape_zero_width_marks_type_t {
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE,
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY,
+  HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE
+};
+
+
+/* Master OT shaper list */
+#define HB_COMPLEX_SHAPERS_IMPLEMENT_SHAPERS \
+  HB_COMPLEX_SHAPER_IMPLEMENT (arabic) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (default) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (dumber) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (hangul) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (hebrew) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (indic) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (khmer) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (myanmar) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (myanmar_zawgyi) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (thai) \
+  HB_COMPLEX_SHAPER_IMPLEMENT (use) \
+  /* ^--- Add new shapers here; keep sorted. */
+
+
+struct hb_ot_complex_shaper_t
+{
+  /* collect_features()
+   * Called during shape_plan().
+   * Shapers should use plan->map to add their features and callbacks.
+   * May be NULL.
+   */
+  void (*collect_features) (hb_ot_shape_planner_t *plan);
+
+  /* override_features()
+   * Called during shape_plan().
+   * Shapers should use plan->map to override features and add callbacks after
+   * common features are added.
+   * May be NULL.
+   */
+  void (*override_features) (hb_ot_shape_planner_t *plan);
+
+
+  /* data_create()
+   * Called at the end of shape_plan().
+   * Whatever shapers return will be accessible through plan->data later.
+   * If nullptr is returned, means a plan failure.
+   */
+  void *(*data_create) (const hb_ot_shape_plan_t *plan);
+
+  /* data_destroy()
+   * Called when the shape_plan is being destroyed.
+   * plan->data is passed here for destruction.
+   * If nullptr is returned, means a plan failure.
+   * May be NULL.
+   */
+  void (*data_destroy) (void *data);
+
+
+  /* preprocess_text()
+   * Called during shape().
+   * Shapers can use to modify text before shaping starts.
+   * May be NULL.
+   */
+  void (*preprocess_text) (const hb_ot_shape_plan_t *plan,
+			   hb_buffer_t              *buffer,
+			   hb_font_t                *font);
+
+  /* postprocess_glyphs()
+   * Called during shape().
+   * Shapers can use to modify glyphs after shaping ends.
+   * May be NULL.
+   */
+  void (*postprocess_glyphs) (const hb_ot_shape_plan_t *plan,
+			      hb_buffer_t              *buffer,
+			      hb_font_t                *font);
+
+
+  hb_ot_shape_normalization_mode_t normalization_preference;
+
+  /* decompose()
+   * Called during shape()'s normalization.
+   * May be NULL.
+   */
+  bool (*decompose) (const hb_ot_shape_normalize_context_t *c,
+		     hb_codepoint_t  ab,
+		     hb_codepoint_t *a,
+		     hb_codepoint_t *b);
+
+  /* compose()
+   * Called during shape()'s normalization.
+   * May be NULL.
+   */
+  bool (*compose) (const hb_ot_shape_normalize_context_t *c,
+		   hb_codepoint_t  a,
+		   hb_codepoint_t  b,
+		   hb_codepoint_t *ab);
+
+  /* setup_masks()
+   * Called during shape().
+   * Shapers should use map to get feature masks and set on buffer.
+   * Shapers may NOT modify characters.
+   * May be NULL.
+   */
+  void (*setup_masks) (const hb_ot_shape_plan_t *plan,
+		       hb_buffer_t              *buffer,
+		       hb_font_t                *font);
+
+  /* gpos_tag()
+   * If not HB_TAG_NONE, then must match found GPOS script tag for
+   * GPOS to be applied.  Otherwise, fallback positioning will be used.
+   */
+  hb_tag_t gpos_tag;
+
+  /* reorder_marks()
+   * Called during shape().
+   * Shapers can use to modify ordering of combining marks.
+   * May be NULL.
+   */
+  void (*reorder_marks) (const hb_ot_shape_plan_t *plan,
+			 hb_buffer_t              *buffer,
+			 unsigned int              start,
+			 unsigned int              end);
+
+  hb_ot_shape_zero_width_marks_type_t zero_width_marks;
+
+  bool fallback_position;
+};
+
+#define HB_COMPLEX_SHAPER_IMPLEMENT(name) extern HB_INTERNAL const hb_ot_complex_shaper_t _hb_ot_complex_shaper_##name;
+HB_COMPLEX_SHAPERS_IMPLEMENT_SHAPERS
+#undef HB_COMPLEX_SHAPER_IMPLEMENT
+
+
+static inline const hb_ot_complex_shaper_t *
+hb_ot_shape_complex_categorize (const hb_ot_shape_planner_t *planner)
+{
+  switch ((hb_tag_t) planner->props.script)
+  {
+    default:
+      return &_hb_ot_complex_shaper_default;
+
+
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_ARABIC:
+
+    /* Unicode-3.0 additions */
+    case HB_SCRIPT_MONGOLIAN:
+    case HB_SCRIPT_SYRIAC:
+
+    /* Unicode-5.0 additions */
+    case HB_SCRIPT_NKO:
+    case HB_SCRIPT_PHAGS_PA:
+
+    /* Unicode-6.0 additions */
+    case HB_SCRIPT_MANDAIC:
+
+    /* Unicode-7.0 additions */
+    case HB_SCRIPT_MANICHAEAN:
+    case HB_SCRIPT_PSALTER_PAHLAVI:
+
+    /* Unicode-9.0 additions */
+    case HB_SCRIPT_ADLAM:
+
+    /* Unicode-11.0 additions */
+    case HB_SCRIPT_HANIFI_ROHINGYA:
+    case HB_SCRIPT_SOGDIAN:
+
+      /* For Arabic script, use the Arabic shaper even if no OT script tag was found.
+       * This is because we do fallback shaping for Arabic script (and not others).
+       * But note that Arabic shaping is applicable only to horizontal layout; for
+       * vertical text, just use the generic shaper instead. */
+      if ((planner->map.chosen_script[0] != HB_OT_TAG_DEFAULT_SCRIPT ||
+	   planner->props.script == HB_SCRIPT_ARABIC) &&
+	  HB_DIRECTION_IS_HORIZONTAL(planner->props.direction))
+	return &_hb_ot_complex_shaper_arabic;
+      else
+	return &_hb_ot_complex_shaper_default;
+
+
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_THAI:
+    case HB_SCRIPT_LAO:
+
+      return &_hb_ot_complex_shaper_thai;
+
+
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_HANGUL:
+
+      return &_hb_ot_complex_shaper_hangul;
+
+
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_HEBREW:
+
+      return &_hb_ot_complex_shaper_hebrew;
+
+
+    /* Unicode-1.1 additions */
+    case HB_SCRIPT_BENGALI:
+    case HB_SCRIPT_DEVANAGARI:
+    case HB_SCRIPT_GUJARATI:
+    case HB_SCRIPT_GURMUKHI:
+    case HB_SCRIPT_KANNADA:
+    case HB_SCRIPT_MALAYALAM:
+    case HB_SCRIPT_ORIYA:
+    case HB_SCRIPT_TAMIL:
+    case HB_SCRIPT_TELUGU:
+
+    /* Unicode-3.0 additions */
+    case HB_SCRIPT_SINHALA:
+
+      /* If the designer designed the font for the 'DFLT' script,
+       * (or we ended up arbitrarily pick 'latn'), use the default shaper.
+       * Otherwise, use the specific shaper.
+       *
+       * If it's indy3 tag, send to USE. */
+      if (planner->map.chosen_script[0] == HB_TAG ('D','F','L','T') ||
+	  planner->map.chosen_script[0] == HB_TAG ('l','a','t','n'))
+	return &_hb_ot_complex_shaper_default;
+      else if ((planner->map.chosen_script[0] & 0x000000FF) == '3')
+	return &_hb_ot_complex_shaper_use;
+      else
+	return &_hb_ot_complex_shaper_indic;
+
+    case HB_SCRIPT_KHMER:
+	return &_hb_ot_complex_shaper_khmer;
+
+    case HB_SCRIPT_MYANMAR:
+      /* If the designer designed the font for the 'DFLT' script,
+       * (or we ended up arbitrarily pick 'latn'), use the default shaper.
+       * Otherwise, use the specific shaper.
+       *
+       * If designer designed for 'mymr' tag, also send to default
+       * shaper.  That's tag used from before Myanmar shaping spec
+       * was developed.  The shaping spec uses 'mym2' tag. */
+      if (planner->map.chosen_script[0] == HB_TAG ('D','F','L','T') ||
+	  planner->map.chosen_script[0] == HB_TAG ('l','a','t','n') ||
+	  planner->map.chosen_script[0] == HB_TAG ('m','y','m','r'))
+	return &_hb_ot_complex_shaper_default;
+      else
+	return &_hb_ot_complex_shaper_myanmar;
+
+
+    /* https://github.com/harfbuzz/harfbuzz/issues/1162 */
+    case HB_SCRIPT_MYANMAR_ZAWGYI:
+
+      return &_hb_ot_complex_shaper_myanmar_zawgyi;
+
+
+    /* Unicode-2.0 additions */
+    case HB_SCRIPT_TIBETAN:
+
+    /* Unicode-3.0 additions */
+    //case HB_SCRIPT_MONGOLIAN:
+    //case HB_SCRIPT_SINHALA:
+
+    /* Unicode-3.2 additions */
+    case HB_SCRIPT_BUHID:
+    case HB_SCRIPT_HANUNOO:
+    case HB_SCRIPT_TAGALOG:
+    case HB_SCRIPT_TAGBANWA:
+
+    /* Unicode-4.0 additions */
+    case HB_SCRIPT_LIMBU:
+    case HB_SCRIPT_TAI_LE:
+
+    /* Unicode-4.1 additions */
+    case HB_SCRIPT_BUGINESE:
+    case HB_SCRIPT_KHAROSHTHI:
+    case HB_SCRIPT_SYLOTI_NAGRI:
+    case HB_SCRIPT_TIFINAGH:
+
+    /* Unicode-5.0 additions */
+    case HB_SCRIPT_BALINESE:
+    //case HB_SCRIPT_NKO:
+    //case HB_SCRIPT_PHAGS_PA:
+
+    /* Unicode-5.1 additions */
+    case HB_SCRIPT_CHAM:
+    case HB_SCRIPT_KAYAH_LI:
+    case HB_SCRIPT_LEPCHA:
+    case HB_SCRIPT_REJANG:
+    case HB_SCRIPT_SAURASHTRA:
+    case HB_SCRIPT_SUNDANESE:
+
+    /* Unicode-5.2 additions */
+    case HB_SCRIPT_EGYPTIAN_HIEROGLYPHS:
+    case HB_SCRIPT_JAVANESE:
+    case HB_SCRIPT_KAITHI:
+    case HB_SCRIPT_MEETEI_MAYEK:
+    case HB_SCRIPT_TAI_THAM:
+    case HB_SCRIPT_TAI_VIET:
+
+    /* Unicode-6.0 additions */
+    case HB_SCRIPT_BATAK:
+    case HB_SCRIPT_BRAHMI:
+    //case HB_SCRIPT_MANDAIC:
+
+    /* Unicode-6.1 additions */
+    case HB_SCRIPT_CHAKMA:
+    case HB_SCRIPT_SHARADA:
+    case HB_SCRIPT_TAKRI:
+
+    /* Unicode-7.0 additions */
+    case HB_SCRIPT_DUPLOYAN:
+    case HB_SCRIPT_GRANTHA:
+    case HB_SCRIPT_KHOJKI:
+    case HB_SCRIPT_KHUDAWADI:
+    case HB_SCRIPT_MAHAJANI:
+    //case HB_SCRIPT_MANICHAEAN:
+    case HB_SCRIPT_MODI:
+    case HB_SCRIPT_PAHAWH_HMONG:
+    //case HB_SCRIPT_PSALTER_PAHLAVI:
+    case HB_SCRIPT_SIDDHAM:
+    case HB_SCRIPT_TIRHUTA:
+
+    /* Unicode-8.0 additions */
+    case HB_SCRIPT_AHOM:
+
+    /* Unicode-9.0 additions */
+    //case HB_SCRIPT_ADLAM:
+    case HB_SCRIPT_BHAIKSUKI:
+    case HB_SCRIPT_MARCHEN:
+    case HB_SCRIPT_NEWA:
+
+    /* Unicode-10.0 additions */
+    case HB_SCRIPT_MASARAM_GONDI:
+    case HB_SCRIPT_SOYOMBO:
+    case HB_SCRIPT_ZANABAZAR_SQUARE:
+
+    /* Unicode-11.0 additions */
+    case HB_SCRIPT_DOGRA:
+    case HB_SCRIPT_GUNJALA_GONDI:
+    //case HB_SCRIPT_HANIFI_ROHINGYA:
+    case HB_SCRIPT_MAKASAR:
+    //case HB_SCRIPT_SOGDIAN:
+
+    /* Unicode-12.0 additions */
+    case HB_SCRIPT_NANDINAGARI:
+
+    /* Unicode-13.0 additions */
+    case HB_SCRIPT_CHORASMIAN:
+    case HB_SCRIPT_DIVES_AKURU:
+
+      /* If the designer designed the font for the 'DFLT' script,
+       * (or we ended up arbitrarily pick 'latn'), use the default shaper.
+       * Otherwise, use the specific shaper.
+       * Note that for some simple scripts, there may not be *any*
+       * GSUB/GPOS needed, so there may be no scripts found! */
+      if (planner->map.chosen_script[0] == HB_TAG ('D','F','L','T') ||
+	  planner->map.chosen_script[0] == HB_TAG ('l','a','t','n'))
+	return &_hb_ot_complex_shaper_default;
+      else
+	return &_hb_ot_complex_shaper_use;
+  }
+}
+
+
+#endif /* HB_OT_SHAPE_COMPLEX_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-fallback.cc b/thirdparty/harfbuzz/src/hb-ot-shape-fallback.cc
new file mode 100644
index 0000000000..42bf524d16
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-fallback.cc
@@ -0,0 +1,596 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-fallback.hh"
+#include "hb-kern.hh"
+
+static unsigned int
+recategorize_combining_class (hb_codepoint_t u,
+			      unsigned int klass)
+{
+  if (klass >= 200)
+    return klass;
+
+  /* Thai / Lao need some per-character work. */
+  if ((u & ~0xFF) == 0x0E00u)
+  {
+    if (unlikely (klass == 0))
+    {
+      switch (u)
+      {
+	case 0x0E31u:
+	case 0x0E34u:
+	case 0x0E35u:
+	case 0x0E36u:
+	case 0x0E37u:
+	case 0x0E47u:
+	case 0x0E4Cu:
+	case 0x0E4Du:
+	case 0x0E4Eu:
+	  klass = HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT;
+	  break;
+
+	case 0x0EB1u:
+	case 0x0EB4u:
+	case 0x0EB5u:
+	case 0x0EB6u:
+	case 0x0EB7u:
+	case 0x0EBBu:
+	case 0x0ECCu:
+	case 0x0ECDu:
+	  klass = HB_UNICODE_COMBINING_CLASS_ABOVE;
+	  break;
+
+	case 0x0EBCu:
+	  klass = HB_UNICODE_COMBINING_CLASS_BELOW;
+	  break;
+      }
+    } else {
+      /* Thai virama is below-right */
+      if (u == 0x0E3Au)
+	klass = HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT;
+    }
+  }
+
+  switch (klass)
+  {
+
+    /* Hebrew */
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC10: /* sheva */
+    case HB_MODIFIED_COMBINING_CLASS_CCC11: /* hataf segol */
+    case HB_MODIFIED_COMBINING_CLASS_CCC12: /* hataf patah */
+    case HB_MODIFIED_COMBINING_CLASS_CCC13: /* hataf qamats */
+    case HB_MODIFIED_COMBINING_CLASS_CCC14: /* hiriq */
+    case HB_MODIFIED_COMBINING_CLASS_CCC15: /* tsere */
+    case HB_MODIFIED_COMBINING_CLASS_CCC16: /* segol */
+    case HB_MODIFIED_COMBINING_CLASS_CCC17: /* patah */
+    case HB_MODIFIED_COMBINING_CLASS_CCC18: /* qamats */
+    case HB_MODIFIED_COMBINING_CLASS_CCC20: /* qubuts */
+    case HB_MODIFIED_COMBINING_CLASS_CCC22: /* meteg */
+      return HB_UNICODE_COMBINING_CLASS_BELOW;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC23: /* rafe */
+      return HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC24: /* shin dot */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC25: /* sin dot */
+    case HB_MODIFIED_COMBINING_CLASS_CCC19: /* holam */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC26: /* point varika */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC21: /* dagesh */
+      break;
+
+
+    /* Arabic and Syriac */
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC27: /* fathatan */
+    case HB_MODIFIED_COMBINING_CLASS_CCC28: /* dammatan */
+    case HB_MODIFIED_COMBINING_CLASS_CCC30: /* fatha */
+    case HB_MODIFIED_COMBINING_CLASS_CCC31: /* damma */
+    case HB_MODIFIED_COMBINING_CLASS_CCC33: /* shadda */
+    case HB_MODIFIED_COMBINING_CLASS_CCC34: /* sukun */
+    case HB_MODIFIED_COMBINING_CLASS_CCC35: /* superscript alef */
+    case HB_MODIFIED_COMBINING_CLASS_CCC36: /* superscript alaph */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC29: /* kasratan */
+    case HB_MODIFIED_COMBINING_CLASS_CCC32: /* kasra */
+      return HB_UNICODE_COMBINING_CLASS_BELOW;
+
+
+    /* Thai */
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC103: /* sara u / sara uu */
+      return HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC107: /* mai */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT;
+
+
+    /* Lao */
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC118: /* sign u / sign uu */
+      return HB_UNICODE_COMBINING_CLASS_BELOW;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC122: /* mai */
+      return HB_UNICODE_COMBINING_CLASS_ABOVE;
+
+
+    /* Tibetan */
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC129: /* sign aa */
+      return HB_UNICODE_COMBINING_CLASS_BELOW;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC130: /* sign i*/
+      return HB_UNICODE_COMBINING_CLASS_ABOVE;
+
+    case HB_MODIFIED_COMBINING_CLASS_CCC132: /* sign u */
+      return HB_UNICODE_COMBINING_CLASS_BELOW;
+
+  }
+
+  return klass;
+}
+
+void
+_hb_ot_shape_fallback_mark_position_recategorize_marks (const hb_ot_shape_plan_t *plan HB_UNUSED,
+							hb_font_t *font HB_UNUSED,
+							hb_buffer_t  *buffer)
+{
+#ifdef HB_NO_OT_SHAPE_FALLBACK
+  return;
+#endif
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if (_hb_glyph_info_get_general_category (&info[i]) == HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK) {
+      unsigned int combining_class = _hb_glyph_info_get_modified_combining_class (&info[i]);
+      combining_class = recategorize_combining_class (info[i].codepoint, combining_class);
+      _hb_glyph_info_set_modified_combining_class (&info[i], combining_class);
+    }
+}
+
+
+static void
+zero_mark_advances (hb_buffer_t *buffer,
+		    unsigned int start,
+		    unsigned int end,
+		    bool adjust_offsets_when_zeroing)
+{
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = start; i < end; i++)
+    if (_hb_glyph_info_get_general_category (&info[i]) == HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)
+    {
+      if (adjust_offsets_when_zeroing)
+      {
+	buffer->pos[i].x_offset -= buffer->pos[i].x_advance;
+	buffer->pos[i].y_offset -= buffer->pos[i].y_advance;
+      }
+      buffer->pos[i].x_advance = 0;
+      buffer->pos[i].y_advance = 0;
+    }
+}
+
+static inline void
+position_mark (const hb_ot_shape_plan_t *plan HB_UNUSED,
+	       hb_font_t *font,
+	       hb_buffer_t  *buffer,
+	       hb_glyph_extents_t &base_extents,
+	       unsigned int i,
+	       unsigned int combining_class)
+{
+  hb_glyph_extents_t mark_extents;
+  if (!font->get_glyph_extents (buffer->info[i].codepoint, &mark_extents))
+    return;
+
+  hb_position_t y_gap = font->y_scale / 16;
+
+  hb_glyph_position_t &pos = buffer->pos[i];
+  pos.x_offset = pos.y_offset = 0;
+
+
+  /* We don't position LEFT and RIGHT marks. */
+
+  /* X positioning */
+  switch (combining_class)
+  {
+    case HB_UNICODE_COMBINING_CLASS_DOUBLE_BELOW:
+    case HB_UNICODE_COMBINING_CLASS_DOUBLE_ABOVE:
+      if (buffer->props.direction == HB_DIRECTION_LTR) {
+	pos.x_offset += base_extents.x_bearing + base_extents.width - mark_extents.width / 2 - mark_extents.x_bearing;
+	break;
+      } else if (buffer->props.direction == HB_DIRECTION_RTL) {
+	pos.x_offset += base_extents.x_bearing - mark_extents.width / 2 - mark_extents.x_bearing;
+	break;
+      }
+      HB_FALLTHROUGH;
+
+    default:
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW:
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE:
+    case HB_UNICODE_COMBINING_CLASS_BELOW:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE:
+      /* Center align. */
+      pos.x_offset += base_extents.x_bearing + (base_extents.width - mark_extents.width) / 2 - mark_extents.x_bearing;
+      break;
+
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW_LEFT:
+    case HB_UNICODE_COMBINING_CLASS_BELOW_LEFT:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT:
+      /* Left align. */
+      pos.x_offset += base_extents.x_bearing - mark_extents.x_bearing;
+      break;
+
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE_RIGHT:
+    case HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT:
+      /* Right align. */
+      pos.x_offset += base_extents.x_bearing + base_extents.width - mark_extents.width - mark_extents.x_bearing;
+      break;
+  }
+
+  /* Y positioning */
+  switch (combining_class)
+  {
+    case HB_UNICODE_COMBINING_CLASS_DOUBLE_BELOW:
+    case HB_UNICODE_COMBINING_CLASS_BELOW_LEFT:
+    case HB_UNICODE_COMBINING_CLASS_BELOW:
+    case HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT:
+      /* Add gap, fall-through. */
+      base_extents.height -= y_gap;
+      HB_FALLTHROUGH;
+
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW_LEFT:
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW:
+      pos.y_offset = base_extents.y_bearing + base_extents.height - mark_extents.y_bearing;
+      /* Never shift up "below" marks. */
+      if ((y_gap > 0) == (pos.y_offset > 0))
+      {
+	base_extents.height -= pos.y_offset;
+	pos.y_offset = 0;
+      }
+      base_extents.height += mark_extents.height;
+      break;
+
+    case HB_UNICODE_COMBINING_CLASS_DOUBLE_ABOVE:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE:
+    case HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT:
+      /* Add gap, fall-through. */
+      base_extents.y_bearing += y_gap;
+      base_extents.height -= y_gap;
+      HB_FALLTHROUGH;
+
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE:
+    case HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE_RIGHT:
+      pos.y_offset = base_extents.y_bearing - (mark_extents.y_bearing + mark_extents.height);
+      /* Don't shift down "above" marks too much. */
+      if ((y_gap > 0) != (pos.y_offset > 0))
+      {
+	unsigned int correction = -pos.y_offset / 2;
+	base_extents.y_bearing += correction;
+	base_extents.height -= correction;
+	pos.y_offset += correction;
+      }
+      base_extents.y_bearing -= mark_extents.height;
+      base_extents.height += mark_extents.height;
+      break;
+  }
+}
+
+static inline void
+position_around_base (const hb_ot_shape_plan_t *plan,
+		      hb_font_t *font,
+		      hb_buffer_t  *buffer,
+		      unsigned int base,
+		      unsigned int end,
+		      bool adjust_offsets_when_zeroing)
+{
+  hb_direction_t horiz_dir = HB_DIRECTION_INVALID;
+
+  buffer->unsafe_to_break (base, end);
+
+  hb_glyph_extents_t base_extents;
+  if (!font->get_glyph_extents (buffer->info[base].codepoint,
+				&base_extents))
+  {
+    /* If extents don't work, zero marks and go home. */
+    zero_mark_advances (buffer, base + 1, end, adjust_offsets_when_zeroing);
+    return;
+  }
+  base_extents.y_bearing += buffer->pos[base].y_offset;
+  /* Use horizontal advance for horizontal positioning.
+   * Generally a better idea.  Also works for zero-ink glyphs.  See:
+   * https://github.com/harfbuzz/harfbuzz/issues/1532 */
+  base_extents.x_bearing = 0;
+  base_extents.width = font->get_glyph_h_advance (buffer->info[base].codepoint);
+
+  unsigned int lig_id = _hb_glyph_info_get_lig_id (&buffer->info[base]);
+  /* Use integer for num_lig_components such that it doesn't convert to unsigned
+   * when we divide or multiply by it. */
+  int num_lig_components = _hb_glyph_info_get_lig_num_comps (&buffer->info[base]);
+
+  hb_position_t x_offset = 0, y_offset = 0;
+  if (HB_DIRECTION_IS_FORWARD (buffer->props.direction)) {
+    x_offset -= buffer->pos[base].x_advance;
+    y_offset -= buffer->pos[base].y_advance;
+  }
+
+  hb_glyph_extents_t component_extents = base_extents;
+  int last_lig_component = -1;
+  unsigned int last_combining_class = 255;
+  hb_glyph_extents_t cluster_extents = base_extents; /* Initialization is just to shut gcc up. */
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = base + 1; i < end; i++)
+    if (_hb_glyph_info_get_modified_combining_class (&info[i]))
+    {
+      if (num_lig_components > 1) {
+	unsigned int this_lig_id = _hb_glyph_info_get_lig_id (&info[i]);
+	int this_lig_component = _hb_glyph_info_get_lig_comp (&info[i]) - 1;
+	/* Conditions for attaching to the last component. */
+	if (!lig_id || lig_id != this_lig_id || this_lig_component >= num_lig_components)
+	  this_lig_component = num_lig_components - 1;
+	if (last_lig_component != this_lig_component)
+	{
+	  last_lig_component = this_lig_component;
+	  last_combining_class = 255;
+	  component_extents = base_extents;
+	  if (unlikely (horiz_dir == HB_DIRECTION_INVALID)) {
+	    if (HB_DIRECTION_IS_HORIZONTAL (plan->props.direction))
+	      horiz_dir = plan->props.direction;
+	    else
+	      horiz_dir = hb_script_get_horizontal_direction (plan->props.script);
+	  }
+	  if (horiz_dir == HB_DIRECTION_LTR)
+	    component_extents.x_bearing += (this_lig_component * component_extents.width) / num_lig_components;
+	  else
+	    component_extents.x_bearing += ((num_lig_components - 1 - this_lig_component) * component_extents.width) / num_lig_components;
+	  component_extents.width /= num_lig_components;
+	}
+      }
+
+      unsigned int this_combining_class = _hb_glyph_info_get_modified_combining_class (&info[i]);
+      if (last_combining_class != this_combining_class)
+      {
+	last_combining_class = this_combining_class;
+	cluster_extents = component_extents;
+      }
+
+      position_mark (plan, font, buffer, cluster_extents, i, this_combining_class);
+
+      buffer->pos[i].x_advance = 0;
+      buffer->pos[i].y_advance = 0;
+      buffer->pos[i].x_offset += x_offset;
+      buffer->pos[i].y_offset += y_offset;
+
+    } else {
+      if (HB_DIRECTION_IS_FORWARD (buffer->props.direction)) {
+	x_offset -= buffer->pos[i].x_advance;
+	y_offset -= buffer->pos[i].y_advance;
+      } else {
+	x_offset += buffer->pos[i].x_advance;
+	y_offset += buffer->pos[i].y_advance;
+      }
+    }
+}
+
+static inline void
+position_cluster (const hb_ot_shape_plan_t *plan,
+		  hb_font_t *font,
+		  hb_buffer_t  *buffer,
+		  unsigned int start,
+		  unsigned int end,
+		  bool adjust_offsets_when_zeroing)
+{
+  if (end - start < 2)
+    return;
+
+  /* Find the base glyph */
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = start; i < end; i++)
+    if (!_hb_glyph_info_is_unicode_mark (&info[i]))
+    {
+      /* Find mark glyphs */
+      unsigned int j;
+      for (j = i + 1; j < end; j++)
+	if (!_hb_glyph_info_is_unicode_mark (&info[j]))
+	  break;
+
+      position_around_base (plan, font, buffer, i, j, adjust_offsets_when_zeroing);
+
+      i = j - 1;
+    }
+}
+
+void
+_hb_ot_shape_fallback_mark_position (const hb_ot_shape_plan_t *plan,
+				     hb_font_t *font,
+				     hb_buffer_t  *buffer,
+				     bool adjust_offsets_when_zeroing)
+{
+#ifdef HB_NO_OT_SHAPE_FALLBACK
+  return;
+#endif
+
+  _hb_buffer_assert_gsubgpos_vars (buffer);
+
+  unsigned int start = 0;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 1; i < count; i++)
+    if (likely (!_hb_glyph_info_is_unicode_mark (&info[i]))) {
+      position_cluster (plan, font, buffer, start, i, adjust_offsets_when_zeroing);
+      start = i;
+    }
+  position_cluster (plan, font, buffer, start, count, adjust_offsets_when_zeroing);
+}
+
+
+#ifndef HB_DISABLE_DEPRECATED
+struct hb_ot_shape_fallback_kern_driver_t
+{
+  hb_ot_shape_fallback_kern_driver_t (hb_font_t   *font_,
+				      hb_buffer_t *buffer) :
+    font (font_), direction (buffer->props.direction) {}
+
+  hb_position_t get_kerning (hb_codepoint_t first, hb_codepoint_t second) const
+  {
+    hb_position_t kern = 0;
+    font->get_glyph_kerning_for_direction (first, second,
+					   direction,
+					   &kern, &kern);
+    return kern;
+  }
+
+  hb_font_t *font;
+  hb_direction_t direction;
+};
+#endif
+
+/* Performs font-assisted kerning. */
+void
+_hb_ot_shape_fallback_kern (const hb_ot_shape_plan_t *plan,
+			    hb_font_t *font,
+			    hb_buffer_t *buffer)
+{
+#ifdef HB_NO_OT_SHAPE_FALLBACK
+  return;
+#endif
+
+#ifndef HB_DISABLE_DEPRECATED
+  if (HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction) ?
+      !font->has_glyph_h_kerning_func () :
+      !font->has_glyph_v_kerning_func ())
+    return;
+
+  bool reverse = HB_DIRECTION_IS_BACKWARD (buffer->props.direction);
+
+  if (reverse)
+    buffer->reverse ();
+
+  hb_ot_shape_fallback_kern_driver_t driver (font, buffer);
+  OT::hb_kern_machine_t<hb_ot_shape_fallback_kern_driver_t> machine (driver);
+  machine.kern (font, buffer, plan->kern_mask, false);
+
+  if (reverse)
+    buffer->reverse ();
+#endif
+}
+
+
+/* Adjusts width of various spaces. */
+void
+_hb_ot_shape_fallback_spaces (const hb_ot_shape_plan_t *plan HB_UNUSED,
+			      hb_font_t *font,
+			      hb_buffer_t  *buffer)
+{
+  hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pos = buffer->pos;
+  bool horizontal = HB_DIRECTION_IS_HORIZONTAL (buffer->props.direction);
+  unsigned int count = buffer->len;
+  for (unsigned int i = 0; i < count; i++)
+    if (_hb_glyph_info_is_unicode_space (&info[i]) && !_hb_glyph_info_ligated (&info[i]))
+    {
+      hb_unicode_funcs_t::space_t space_type = _hb_glyph_info_get_unicode_space_fallback_type (&info[i]);
+      hb_codepoint_t glyph;
+      typedef hb_unicode_funcs_t t;
+      switch (space_type)
+      {
+	case t::NOT_SPACE: /* Shouldn't happen. */
+	case t::SPACE:
+	  break;
+
+	case t::SPACE_EM:
+	case t::SPACE_EM_2:
+	case t::SPACE_EM_3:
+	case t::SPACE_EM_4:
+	case t::SPACE_EM_5:
+	case t::SPACE_EM_6:
+	case t::SPACE_EM_16:
+	  if (horizontal)
+	    pos[i].x_advance = +(font->x_scale + ((int) space_type)/2) / (int) space_type;
+	  else
+	    pos[i].y_advance = -(font->y_scale + ((int) space_type)/2) / (int) space_type;
+	  break;
+
+	case t::SPACE_4_EM_18:
+	  if (horizontal)
+	    pos[i].x_advance = (int64_t) +font->x_scale * 4 / 18;
+	  else
+	    pos[i].y_advance = (int64_t) -font->y_scale * 4 / 18;
+	  break;
+
+	case t::SPACE_FIGURE:
+	  for (char u = '0'; u <= '9'; u++)
+	    if (font->get_nominal_glyph (u, &glyph))
+	    {
+	      if (horizontal)
+		pos[i].x_advance = font->get_glyph_h_advance (glyph);
+	      else
+		pos[i].y_advance = font->get_glyph_v_advance (glyph);
+	      break;
+	    }
+	  break;
+
+	case t::SPACE_PUNCTUATION:
+	  if (font->get_nominal_glyph ('.', &glyph) ||
+	      font->get_nominal_glyph (',', &glyph))
+	  {
+	    if (horizontal)
+	      pos[i].x_advance = font->get_glyph_h_advance (glyph);
+	    else
+	      pos[i].y_advance = font->get_glyph_v_advance (glyph);
+	  }
+	  break;
+
+	case t::SPACE_NARROW:
+	  /* Half-space?
+	   * Unicode doc https://unicode.org/charts/PDF/U2000.pdf says ~1/4 or 1/5 of EM.
+	   * However, in my testing, many fonts have their regular space being about that
+	   * size.  To me, a percentage of the space width makes more sense.  Half is as
+	   * good as any. */
+	  if (horizontal)
+	    pos[i].x_advance /= 2;
+	  else
+	    pos[i].y_advance /= 2;
+	  break;
+      }
+    }
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-fallback.hh b/thirdparty/harfbuzz/src/hb-ot-shape-fallback.hh
new file mode 100644
index 0000000000..5faf5f2dfb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-fallback.hh
@@ -0,0 +1,54 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_FALLBACK_HH
+#define HB_OT_SHAPE_FALLBACK_HH
+
+#include "hb.hh"
+
+#include "hb-ot-shape.hh"
+
+
+HB_INTERNAL void _hb_ot_shape_fallback_mark_position (const hb_ot_shape_plan_t *plan,
+						      hb_font_t *font,
+						      hb_buffer_t  *buffer,
+						      bool adjust_offsets_when_zeroing);
+
+HB_INTERNAL void _hb_ot_shape_fallback_mark_position_recategorize_marks (const hb_ot_shape_plan_t *plan,
+									 hb_font_t *font,
+									 hb_buffer_t  *buffer);
+
+
+HB_INTERNAL void _hb_ot_shape_fallback_kern (const hb_ot_shape_plan_t *plan,
+					     hb_font_t *font,
+					     hb_buffer_t  *buffer);
+
+HB_INTERNAL void _hb_ot_shape_fallback_spaces (const hb_ot_shape_plan_t *plan,
+					       hb_font_t *font,
+					       hb_buffer_t  *buffer);
+
+
+#endif /* HB_OT_SHAPE_FALLBACK_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-normalize.cc b/thirdparty/harfbuzz/src/hb-ot-shape-normalize.cc
new file mode 100644
index 0000000000..50b5829c4a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-normalize.cc
@@ -0,0 +1,478 @@
+/*
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#include "hb-ot-shape-normalize.hh"
+#include "hb-ot-shape-complex.hh"
+#include "hb-ot-shape.hh"
+
+
+/*
+ * HIGHLEVEL DESIGN:
+ *
+ * This file exports one main function: _hb_ot_shape_normalize().
+ *
+ * This function closely reflects the Unicode Normalization Algorithm,
+ * yet it's different.
+ *
+ * Each shaper specifies whether it prefers decomposed (NFD) or composed (NFC).
+ * The logic however tries to use whatever the font can support.
+ *
+ * In general what happens is that: each grapheme is decomposed in a chain
+ * of 1:2 decompositions, marks reordered, and then recomposed if desired,
+ * so far it's like Unicode Normalization.  However, the decomposition and
+ * recomposition only happens if the font supports the resulting characters.
+ *
+ * The goals are:
+ *
+ *   - Try to render all canonically equivalent strings similarly.  To really
+ *     achieve this we have to always do the full decomposition and then
+ *     selectively recompose from there.  It's kinda too expensive though, so
+ *     we skip some cases.  For example, if composed is desired, we simply
+ *     don't touch 1-character clusters that are supported by the font, even
+ *     though their NFC may be different.
+ *
+ *   - When a font has a precomposed character for a sequence but the 'ccmp'
+ *     feature in the font is not adequate, use the precomposed character
+ *     which typically has better mark positioning.
+ *
+ *   - When a font does not support a combining mark, but supports it precomposed
+ *     with previous base, use that.  This needs the itemizer to have this
+ *     knowledge too.  We need to provide assistance to the itemizer.
+ *
+ *   - When a font does not support a character but supports its canonical
+ *     decomposition, well, use the decomposition.
+ *
+ *   - The complex shapers can customize the compose and decompose functions to
+ *     offload some of their requirements to the normalizer.  For example, the
+ *     Indic shaper may want to disallow recomposing of two matras.
+ */
+
+static bool
+decompose_unicode (const hb_ot_shape_normalize_context_t *c,
+		   hb_codepoint_t  ab,
+		   hb_codepoint_t *a,
+		   hb_codepoint_t *b)
+{
+  return (bool) c->unicode->decompose (ab, a, b);
+}
+
+static bool
+compose_unicode (const hb_ot_shape_normalize_context_t *c,
+		 hb_codepoint_t  a,
+		 hb_codepoint_t  b,
+		 hb_codepoint_t *ab)
+{
+  return (bool) c->unicode->compose (a, b, ab);
+}
+
+static inline void
+set_glyph (hb_glyph_info_t &info, hb_font_t *font)
+{
+  (void) font->get_nominal_glyph (info.codepoint, &info.glyph_index());
+}
+
+static inline void
+output_char (hb_buffer_t *buffer, hb_codepoint_t unichar, hb_codepoint_t glyph)
+{
+  buffer->cur().glyph_index() = glyph;
+  buffer->output_glyph (unichar); /* This is very confusing indeed. */
+  _hb_glyph_info_set_unicode_props (&buffer->prev(), buffer);
+}
+
+static inline void
+next_char (hb_buffer_t *buffer, hb_codepoint_t glyph)
+{
+  buffer->cur().glyph_index() = glyph;
+  buffer->next_glyph ();
+}
+
+static inline void
+skip_char (hb_buffer_t *buffer)
+{
+  buffer->skip_glyph ();
+}
+
+/* Returns 0 if didn't decompose, number of resulting characters otherwise. */
+static inline unsigned int
+decompose (const hb_ot_shape_normalize_context_t *c, bool shortest, hb_codepoint_t ab)
+{
+  hb_codepoint_t a = 0, b = 0, a_glyph = 0, b_glyph = 0;
+  hb_buffer_t * const buffer = c->buffer;
+  hb_font_t * const font = c->font;
+
+  if (!c->decompose (c, ab, &a, &b) ||
+      (b && !font->get_nominal_glyph (b, &b_glyph)))
+    return 0;
+
+  bool has_a = (bool) font->get_nominal_glyph (a, &a_glyph);
+  if (shortest && has_a) {
+    /* Output a and b */
+    output_char (buffer, a, a_glyph);
+    if (likely (b)) {
+      output_char (buffer, b, b_glyph);
+      return 2;
+    }
+    return 1;
+  }
+
+  unsigned int ret;
+  if ((ret = decompose (c, shortest, a))) {
+    if (b) {
+      output_char (buffer, b, b_glyph);
+      return ret + 1;
+    }
+    return ret;
+  }
+
+  if (has_a) {
+    output_char (buffer, a, a_glyph);
+    if (likely (b)) {
+      output_char (buffer, b, b_glyph);
+      return 2;
+    }
+    return 1;
+  }
+
+  return 0;
+}
+
+static inline void
+decompose_current_character (const hb_ot_shape_normalize_context_t *c, bool shortest)
+{
+  hb_buffer_t * const buffer = c->buffer;
+  hb_codepoint_t u = buffer->cur().codepoint;
+  hb_codepoint_t glyph = 0;
+
+  if (shortest && c->font->get_nominal_glyph (u, &glyph))
+  {
+    next_char (buffer, glyph);
+    return;
+  }
+
+  if (decompose (c, shortest, u))
+  {
+    skip_char (buffer);
+    return;
+  }
+
+  if (!shortest && c->font->get_nominal_glyph (u, &glyph))
+  {
+    next_char (buffer, glyph);
+    return;
+  }
+
+  if (_hb_glyph_info_is_unicode_space (&buffer->cur()))
+  {
+    hb_codepoint_t space_glyph;
+    hb_unicode_funcs_t::space_t space_type = buffer->unicode->space_fallback_type (u);
+    if (space_type != hb_unicode_funcs_t::NOT_SPACE && c->font->get_nominal_glyph (0x0020u, &space_glyph))
+    {
+      _hb_glyph_info_set_unicode_space_fallback_type (&buffer->cur(), space_type);
+      next_char (buffer, space_glyph);
+      buffer->scratch_flags |= HB_BUFFER_SCRATCH_FLAG_HAS_SPACE_FALLBACK;
+      return;
+    }
+  }
+
+  if (u == 0x2011u)
+  {
+    /* U+2011 is the only sensible character that is a no-break version of another character
+     * and not a space.  The space ones are handled already.  Handle this lone one. */
+    hb_codepoint_t other_glyph;
+    if (c->font->get_nominal_glyph (0x2010u, &other_glyph))
+    {
+      next_char (buffer, other_glyph);
+      return;
+    }
+  }
+
+  next_char (buffer, glyph); /* glyph is initialized in earlier branches. */
+}
+
+static inline void
+handle_variation_selector_cluster (const hb_ot_shape_normalize_context_t *c,
+				   unsigned int end,
+				   bool short_circuit HB_UNUSED)
+{
+  /* TODO Currently if there's a variation-selector we give-up, it's just too hard. */
+  hb_buffer_t * const buffer = c->buffer;
+  hb_font_t * const font = c->font;
+  for (; buffer->idx < end - 1 && buffer->successful;) {
+    if (unlikely (buffer->unicode->is_variation_selector (buffer->cur(+1).codepoint))) {
+      if (font->get_variation_glyph (buffer->cur().codepoint, buffer->cur(+1).codepoint, &buffer->cur().glyph_index()))
+      {
+	hb_codepoint_t unicode = buffer->cur().codepoint;
+	buffer->replace_glyphs (2, 1, &unicode);
+      }
+      else
+      {
+	/* Just pass on the two characters separately, let GSUB do its magic. */
+	set_glyph (buffer->cur(), font);
+	buffer->next_glyph ();
+	set_glyph (buffer->cur(), font);
+	buffer->next_glyph ();
+      }
+      /* Skip any further variation selectors. */
+      while (buffer->idx < end && unlikely (buffer->unicode->is_variation_selector (buffer->cur().codepoint)))
+      {
+	set_glyph (buffer->cur(), font);
+	buffer->next_glyph ();
+      }
+    } else {
+      set_glyph (buffer->cur(), font);
+      buffer->next_glyph ();
+    }
+  }
+  if (likely (buffer->idx < end)) {
+    set_glyph (buffer->cur(), font);
+    buffer->next_glyph ();
+  }
+}
+
+static inline void
+decompose_multi_char_cluster (const hb_ot_shape_normalize_context_t *c, unsigned int end, bool short_circuit)
+{
+  hb_buffer_t * const buffer = c->buffer;
+  for (unsigned int i = buffer->idx; i < end && buffer->successful; i++)
+    if (unlikely (buffer->unicode->is_variation_selector (buffer->info[i].codepoint))) {
+      handle_variation_selector_cluster (c, end, short_circuit);
+      return;
+    }
+
+  while (buffer->idx < end && buffer->successful)
+    decompose_current_character (c, short_circuit);
+}
+
+
+static int
+compare_combining_class (const hb_glyph_info_t *pa, const hb_glyph_info_t *pb)
+{
+  unsigned int a = _hb_glyph_info_get_modified_combining_class (pa);
+  unsigned int b = _hb_glyph_info_get_modified_combining_class (pb);
+
+  return a < b ? -1 : a == b ? 0 : +1;
+}
+
+
+void
+_hb_ot_shape_normalize (const hb_ot_shape_plan_t *plan,
+			hb_buffer_t *buffer,
+			hb_font_t *font)
+{
+  if (unlikely (!buffer->len)) return;
+
+  _hb_buffer_assert_unicode_vars (buffer);
+
+  hb_ot_shape_normalization_mode_t mode = plan->shaper->normalization_preference;
+  if (mode == HB_OT_SHAPE_NORMALIZATION_MODE_AUTO)
+  {
+    if (plan->has_gpos_mark)
+      // https://github.com/harfbuzz/harfbuzz/issues/653#issuecomment-423905920
+      //mode = HB_OT_SHAPE_NORMALIZATION_MODE_DECOMPOSED;
+      mode = HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS;
+    else
+      mode = HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS;
+  }
+
+  const hb_ot_shape_normalize_context_t c = {
+    plan,
+    buffer,
+    font,
+    buffer->unicode,
+    plan->shaper->decompose ? plan->shaper->decompose : decompose_unicode,
+    plan->shaper->compose   ? plan->shaper->compose   : compose_unicode
+  };
+
+  bool always_short_circuit = mode == HB_OT_SHAPE_NORMALIZATION_MODE_NONE;
+  bool might_short_circuit = always_short_circuit ||
+			     (mode != HB_OT_SHAPE_NORMALIZATION_MODE_DECOMPOSED &&
+			      mode != HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT);
+  unsigned int count;
+
+  /* We do a fairly straightforward yet custom normalization process in three
+   * separate rounds: decompose, reorder, recompose (if desired).  Currently
+   * this makes two buffer swaps.  We can make it faster by moving the last
+   * two rounds into the inner loop for the first round, but it's more readable
+   * this way. */
+
+
+  /* First round, decompose */
+
+  bool all_simple = true;
+  {
+    buffer->clear_output ();
+    count = buffer->len;
+    buffer->idx = 0;
+    do
+    {
+      unsigned int end;
+      for (end = buffer->idx + 1; end < count; end++)
+	if (unlikely (_hb_glyph_info_is_unicode_mark (&buffer->info[end])))
+	  break;
+
+      if (end < count)
+	end--; /* Leave one base for the marks to cluster with. */
+
+      /* From idx to end are simple clusters. */
+      if (might_short_circuit)
+      {
+	unsigned int done = font->get_nominal_glyphs (end - buffer->idx,
+						      &buffer->cur().codepoint,
+						      sizeof (buffer->info[0]),
+						      &buffer->cur().glyph_index(),
+						      sizeof (buffer->info[0]));
+	buffer->next_glyphs (done);
+      }
+      while (buffer->idx < end && buffer->successful)
+	decompose_current_character (&c, might_short_circuit);
+
+      if (buffer->idx == count || !buffer->successful)
+	break;
+
+      all_simple = false;
+
+      /* Find all the marks now. */
+      for (end = buffer->idx + 1; end < count; end++)
+	if (!_hb_glyph_info_is_unicode_mark(&buffer->info[end]))
+	  break;
+
+      /* idx to end is one non-simple cluster. */
+      decompose_multi_char_cluster (&c, end, always_short_circuit);
+    }
+    while (buffer->idx < count && buffer->successful);
+    buffer->swap_buffers ();
+  }
+
+
+  /* Second round, reorder (inplace) */
+
+  if (!all_simple)
+  {
+    count = buffer->len;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (_hb_glyph_info_get_modified_combining_class (&buffer->info[i]) == 0)
+	continue;
+
+      unsigned int end;
+      for (end = i + 1; end < count; end++)
+	if (_hb_glyph_info_get_modified_combining_class (&buffer->info[end]) == 0)
+	  break;
+
+      /* We are going to do a O(n^2).  Only do this if the sequence is short. */
+      if (end - i > HB_OT_SHAPE_COMPLEX_MAX_COMBINING_MARKS) {
+	i = end;
+	continue;
+      }
+
+      buffer->sort (i, end, compare_combining_class);
+
+      if (plan->shaper->reorder_marks)
+	plan->shaper->reorder_marks (plan, buffer, i, end);
+
+      i = end;
+    }
+  }
+  if (buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_CGJ)
+  {
+    /* For all CGJ, check if it prevented any reordering at all.
+     * If it did NOT, then make it skippable.
+     * https://github.com/harfbuzz/harfbuzz/issues/554
+     */
+    for (unsigned int i = 1; i + 1 < buffer->len; i++)
+      if (buffer->info[i].codepoint == 0x034Fu/*CGJ*/ &&
+	  info_cc(buffer->info[i-1]) <= info_cc(buffer->info[i+1]))
+      {
+	_hb_glyph_info_unhide (&buffer->info[i]);
+      }
+  }
+
+
+  /* Third round, recompose */
+
+  if (!all_simple &&
+      (mode == HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS ||
+       mode == HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT))
+  {
+    /* As noted in the comment earlier, we don't try to combine
+     * ccc=0 chars with their previous Starter. */
+
+    buffer->clear_output ();
+    count = buffer->len;
+    unsigned int starter = 0;
+    buffer->next_glyph ();
+    while (buffer->idx < count && buffer->successful)
+    {
+      hb_codepoint_t composed, glyph;
+      if (/* We don't try to compose a non-mark character with it's preceding starter.
+	   * This is both an optimization to avoid trying to compose every two neighboring
+	   * glyphs in most scripts AND a desired feature for Hangul.  Apparently Hangul
+	   * fonts are not designed to mix-and-match pre-composed syllables and Jamo. */
+	  _hb_glyph_info_is_unicode_mark(&buffer->cur()))
+      {
+	if (/* If there's anything between the starter and this char, they should have CCC
+	     * smaller than this character's. */
+	    (starter == buffer->out_len - 1 ||
+	     info_cc (buffer->prev()) < info_cc (buffer->cur())) &&
+	    /* And compose. */
+	    c.compose (&c,
+		       buffer->out_info[starter].codepoint,
+		       buffer->cur().codepoint,
+		       &composed) &&
+	    /* And the font has glyph for the composite. */
+	    font->get_nominal_glyph (composed, &glyph))
+	{
+	  /* Composes. */
+	  buffer->next_glyph (); /* Copy to out-buffer. */
+	  if (unlikely (!buffer->successful))
+	    return;
+	  buffer->merge_out_clusters (starter, buffer->out_len);
+	  buffer->out_len--; /* Remove the second composable. */
+	  /* Modify starter and carry on. */
+	  buffer->out_info[starter].codepoint = composed;
+	  buffer->out_info[starter].glyph_index() = glyph;
+	  _hb_glyph_info_set_unicode_props (&buffer->out_info[starter], buffer);
+
+	  continue;
+	}
+      }
+
+      /* Blocked, or doesn't compose. */
+      buffer->next_glyph ();
+
+      if (info_cc (buffer->prev()) == 0)
+	starter = buffer->out_len - 1;
+    }
+    buffer->swap_buffers ();
+  }
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape-normalize.hh b/thirdparty/harfbuzz/src/hb-ot-shape-normalize.hh
new file mode 100644
index 0000000000..04f1a80091
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape-normalize.hh
@@ -0,0 +1,70 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_NORMALIZE_HH
+#define HB_OT_SHAPE_NORMALIZE_HH
+
+#include "hb.hh"
+
+
+/* buffer var allocations, used during the normalization process */
+#define glyph_index()	var1.u32
+
+struct hb_ot_shape_plan_t;
+
+enum hb_ot_shape_normalization_mode_t {
+  HB_OT_SHAPE_NORMALIZATION_MODE_NONE,
+  HB_OT_SHAPE_NORMALIZATION_MODE_DECOMPOSED,
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS, /* Never composes base-to-base */
+  HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT, /* Always fully decomposes and then recompose back */
+
+  HB_OT_SHAPE_NORMALIZATION_MODE_AUTO, /* See hb-ot-shape-normalize.cc for logic. */
+  HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT = HB_OT_SHAPE_NORMALIZATION_MODE_AUTO
+};
+
+HB_INTERNAL void _hb_ot_shape_normalize (const hb_ot_shape_plan_t *shaper,
+					 hb_buffer_t *buffer,
+					 hb_font_t *font);
+
+
+struct hb_ot_shape_normalize_context_t
+{
+  const hb_ot_shape_plan_t *plan;
+  hb_buffer_t *buffer;
+  hb_font_t *font;
+  hb_unicode_funcs_t *unicode;
+  bool (*decompose) (const hb_ot_shape_normalize_context_t *c,
+		     hb_codepoint_t  ab,
+		     hb_codepoint_t *a,
+		     hb_codepoint_t *b);
+  bool (*compose) (const hb_ot_shape_normalize_context_t *c,
+		   hb_codepoint_t  a,
+		   hb_codepoint_t  b,
+		   hb_codepoint_t *ab);
+};
+
+
+#endif /* HB_OT_SHAPE_NORMALIZE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape.cc b/thirdparty/harfbuzz/src/hb-ot-shape.cc
new file mode 100644
index 0000000000..fe0444987a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape.cc
@@ -0,0 +1,1223 @@
+/*
+ * Copyright © 2009,2010  Red Hat, Inc.
+ * Copyright © 2010,2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_SHAPE
+
+#ifdef HB_NO_OT_LAYOUT
+#error "Cannot compile 'ot' shaper with HB_NO_OT_LAYOUT."
+#endif
+
+#include "hb-shaper-impl.hh"
+
+#include "hb-ot-shape.hh"
+#include "hb-ot-shape-complex.hh"
+#include "hb-ot-shape-fallback.hh"
+#include "hb-ot-shape-normalize.hh"
+
+#include "hb-ot-face.hh"
+
+#include "hb-set.hh"
+
+#include "hb-aat-layout.hh"
+
+
+#ifndef HB_NO_AAT_SHAPE
+static inline bool
+_hb_apply_morx (hb_face_t *face, const hb_segment_properties_t *props)
+{
+  /* https://github.com/harfbuzz/harfbuzz/issues/2124 */
+  return hb_aat_layout_has_substitution (face) &&
+	 (HB_DIRECTION_IS_HORIZONTAL (props->direction) || !hb_ot_layout_has_substitution (face));
+}
+#endif
+
+/**
+ * SECTION:hb-ot-shape
+ * @title: hb-ot-shape
+ * @short_description: OpenType shaping support
+ * @include: hb-ot.h
+ *
+ * Support functions for OpenType shaping related queries.
+ **/
+
+
+static void
+hb_ot_shape_collect_features (hb_ot_shape_planner_t          *planner,
+			      const hb_feature_t             *user_features,
+			      unsigned int                    num_user_features);
+
+hb_ot_shape_planner_t::hb_ot_shape_planner_t (hb_face_t                     *face,
+					      const hb_segment_properties_t *props) :
+						face (face),
+						props (*props),
+						map (face, props),
+						aat_map (face, props)
+#ifndef HB_NO_AAT_SHAPE
+						, apply_morx (_hb_apply_morx (face, props))
+#endif
+{
+  shaper = hb_ot_shape_complex_categorize (this);
+
+  script_zero_marks = shaper->zero_width_marks != HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE;
+  script_fallback_mark_positioning = shaper->fallback_position;
+
+  /* https://github.com/harfbuzz/harfbuzz/issues/1528 */
+  if (apply_morx && shaper != &_hb_ot_complex_shaper_default)
+    shaper = &_hb_ot_complex_shaper_dumber;
+}
+
+void
+hb_ot_shape_planner_t::compile (hb_ot_shape_plan_t           &plan,
+				const hb_ot_shape_plan_key_t &key)
+{
+  plan.props = props;
+  plan.shaper = shaper;
+  map.compile (plan.map, key);
+#ifndef HB_NO_AAT_SHAPE
+  if (apply_morx)
+    aat_map.compile (plan.aat_map);
+#endif
+
+#ifndef HB_NO_OT_SHAPE_FRACTIONS
+  plan.frac_mask = plan.map.get_1_mask (HB_TAG ('f','r','a','c'));
+  plan.numr_mask = plan.map.get_1_mask (HB_TAG ('n','u','m','r'));
+  plan.dnom_mask = plan.map.get_1_mask (HB_TAG ('d','n','o','m'));
+  plan.has_frac = plan.frac_mask || (plan.numr_mask && plan.dnom_mask);
+#endif
+
+  plan.rtlm_mask = plan.map.get_1_mask (HB_TAG ('r','t','l','m'));
+  plan.has_vert = !!plan.map.get_1_mask (HB_TAG ('v','e','r','t'));
+
+  hb_tag_t kern_tag = HB_DIRECTION_IS_HORIZONTAL (props.direction) ?
+		      HB_TAG ('k','e','r','n') : HB_TAG ('v','k','r','n');
+#ifndef HB_NO_OT_KERN
+  plan.kern_mask = plan.map.get_mask (kern_tag);
+  plan.requested_kerning = !!plan.kern_mask;
+#endif
+#ifndef HB_NO_AAT_SHAPE
+  plan.trak_mask = plan.map.get_mask (HB_TAG ('t','r','a','k'));
+  plan.requested_tracking = !!plan.trak_mask;
+#endif
+
+  bool has_gpos_kern = plan.map.get_feature_index (1, kern_tag) != HB_OT_LAYOUT_NO_FEATURE_INDEX;
+  bool disable_gpos = plan.shaper->gpos_tag &&
+		      plan.shaper->gpos_tag != plan.map.chosen_script[1];
+
+  /*
+   * Decide who provides glyph classes. GDEF or Unicode.
+   */
+
+  if (!hb_ot_layout_has_glyph_classes (face))
+    plan.fallback_glyph_classes = true;
+
+  /*
+   * Decide who does substitutions. GSUB, morx, or fallback.
+   */
+
+#ifndef HB_NO_AAT_SHAPE
+  plan.apply_morx = apply_morx;
+#endif
+
+  /*
+   * Decide who does positioning. GPOS, kerx, kern, or fallback.
+   */
+
+  if (0)
+    ;
+#ifndef HB_NO_AAT_SHAPE
+  else if (hb_aat_layout_has_positioning (face))
+    plan.apply_kerx = true;
+#endif
+  else if (!apply_morx && !disable_gpos && hb_ot_layout_has_positioning (face))
+    plan.apply_gpos = true;
+
+  if (!plan.apply_kerx && (!has_gpos_kern || !plan.apply_gpos))
+  {
+    /* Apparently Apple applies kerx if GPOS kern was not applied. */
+#ifndef HB_NO_AAT_SHAPE
+    if (hb_aat_layout_has_positioning (face))
+      plan.apply_kerx = true;
+    else
+#endif
+#ifndef HB_NO_OT_KERN
+    if (hb_ot_layout_has_kerning (face))
+      plan.apply_kern = true;
+#endif
+  }
+
+  plan.zero_marks = script_zero_marks &&
+		    !plan.apply_kerx &&
+		    (!plan.apply_kern
+#ifndef HB_NO_OT_KERN
+		     || !hb_ot_layout_has_machine_kerning (face)
+#endif
+		    );
+  plan.has_gpos_mark = !!plan.map.get_1_mask (HB_TAG ('m','a','r','k'));
+
+  plan.adjust_mark_positioning_when_zeroing = !plan.apply_gpos &&
+					      !plan.apply_kerx &&
+					      (!plan.apply_kern
+#ifndef HB_NO_OT_KERN
+					       || !hb_ot_layout_has_cross_kerning (face)
+#endif
+					      );
+
+  plan.fallback_mark_positioning = plan.adjust_mark_positioning_when_zeroing &&
+				   script_fallback_mark_positioning;
+
+#ifndef HB_NO_AAT_SHAPE
+  /* Currently we always apply trak. */
+  plan.apply_trak = plan.requested_tracking && hb_aat_layout_has_tracking (face);
+#endif
+}
+
+bool
+hb_ot_shape_plan_t::init0 (hb_face_t                     *face,
+			   const hb_shape_plan_key_t     *key)
+{
+  map.init ();
+#ifndef HB_NO_AAT_SHAPE
+  aat_map.init ();
+#endif
+
+  hb_ot_shape_planner_t planner (face,
+				 &key->props);
+
+  hb_ot_shape_collect_features (&planner,
+				key->user_features,
+				key->num_user_features);
+
+  planner.compile (*this, key->ot);
+
+  if (shaper->data_create)
+  {
+    data = shaper->data_create (this);
+    if (unlikely (!data))
+    {
+      map.fini ();
+#ifndef HB_NO_AAT_SHAPE
+      aat_map.fini ();
+#endif
+      return false;
+    }
+  }
+
+  return true;
+}
+
+void
+hb_ot_shape_plan_t::fini ()
+{
+  if (shaper->data_destroy)
+    shaper->data_destroy (const_cast<void *> (data));
+
+  map.fini ();
+#ifndef HB_NO_AAT_SHAPE
+  aat_map.fini ();
+#endif
+}
+
+void
+hb_ot_shape_plan_t::substitute (hb_font_t   *font,
+				hb_buffer_t *buffer) const
+{
+#ifndef HB_NO_AAT_SHAPE
+  if (unlikely (apply_morx))
+    hb_aat_layout_substitute (this, font, buffer);
+  else
+#endif
+    map.substitute (this, font, buffer);
+}
+
+void
+hb_ot_shape_plan_t::position (hb_font_t   *font,
+			      hb_buffer_t *buffer) const
+{
+  if (this->apply_gpos)
+    map.position (this, font, buffer);
+#ifndef HB_NO_AAT_SHAPE
+  else if (this->apply_kerx)
+    hb_aat_layout_position (this, font, buffer);
+#endif
+#ifndef HB_NO_OT_KERN
+  else if (this->apply_kern)
+    hb_ot_layout_kern (this, font, buffer);
+#endif
+  else
+    _hb_ot_shape_fallback_kern (this, font, buffer);
+
+#ifndef HB_NO_AAT_SHAPE
+  if (this->apply_trak)
+    hb_aat_layout_track (this, font, buffer);
+#endif
+}
+
+
+static const hb_ot_map_feature_t
+common_features[] =
+{
+  {HB_TAG('a','b','v','m'), F_GLOBAL},
+  {HB_TAG('b','l','w','m'), F_GLOBAL},
+  {HB_TAG('c','c','m','p'), F_GLOBAL},
+  {HB_TAG('l','o','c','l'), F_GLOBAL},
+  {HB_TAG('m','a','r','k'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('m','k','m','k'), F_GLOBAL_MANUAL_JOINERS},
+  {HB_TAG('r','l','i','g'), F_GLOBAL},
+};
+
+
+static const hb_ot_map_feature_t
+horizontal_features[] =
+{
+  {HB_TAG('c','a','l','t'), F_GLOBAL},
+  {HB_TAG('c','l','i','g'), F_GLOBAL},
+  {HB_TAG('c','u','r','s'), F_GLOBAL},
+  {HB_TAG('d','i','s','t'), F_GLOBAL},
+  {HB_TAG('k','e','r','n'), F_GLOBAL_HAS_FALLBACK},
+  {HB_TAG('l','i','g','a'), F_GLOBAL},
+  {HB_TAG('r','c','l','t'), F_GLOBAL},
+};
+
+static void
+hb_ot_shape_collect_features (hb_ot_shape_planner_t          *planner,
+			      const hb_feature_t             *user_features,
+			      unsigned int                    num_user_features)
+{
+  hb_ot_map_builder_t *map = &planner->map;
+
+  map->enable_feature (HB_TAG('r','v','r','n'));
+  map->add_gsub_pause (nullptr);
+
+  switch (planner->props.direction) {
+    case HB_DIRECTION_LTR:
+      map->enable_feature (HB_TAG ('l','t','r','a'));
+      map->enable_feature (HB_TAG ('l','t','r','m'));
+      break;
+    case HB_DIRECTION_RTL:
+      map->enable_feature (HB_TAG ('r','t','l','a'));
+      map->add_feature (HB_TAG ('r','t','l','m'));
+      break;
+    case HB_DIRECTION_TTB:
+    case HB_DIRECTION_BTT:
+    case HB_DIRECTION_INVALID:
+    default:
+      break;
+  }
+
+#ifndef HB_NO_OT_SHAPE_FRACTIONS
+  /* Automatic fractions. */
+  map->add_feature (HB_TAG ('f','r','a','c'));
+  map->add_feature (HB_TAG ('n','u','m','r'));
+  map->add_feature (HB_TAG ('d','n','o','m'));
+#endif
+
+  /* Random! */
+  map->enable_feature (HB_TAG ('r','a','n','d'), F_RANDOM, HB_OT_MAP_MAX_VALUE);
+
+#ifndef HB_NO_AAT_SHAPE
+  /* Tracking.  We enable dummy feature here just to allow disabling
+   * AAT 'trak' table using features.
+   * https://github.com/harfbuzz/harfbuzz/issues/1303 */
+  map->enable_feature (HB_TAG ('t','r','a','k'), F_HAS_FALLBACK);
+#endif
+
+  map->enable_feature (HB_TAG ('H','A','R','F'));
+
+  if (planner->shaper->collect_features)
+    planner->shaper->collect_features (planner);
+
+  map->enable_feature (HB_TAG ('B','U','Z','Z'));
+
+  for (unsigned int i = 0; i < ARRAY_LENGTH (common_features); i++)
+    map->add_feature (common_features[i]);
+
+  if (HB_DIRECTION_IS_HORIZONTAL (planner->props.direction))
+    for (unsigned int i = 0; i < ARRAY_LENGTH (horizontal_features); i++)
+      map->add_feature (horizontal_features[i]);
+  else
+  {
+    /* We really want to find a 'vert' feature if there's any in the font, no
+     * matter which script/langsys it is listed (or not) under.
+     * See various bugs referenced from:
+     * https://github.com/harfbuzz/harfbuzz/issues/63 */
+    map->enable_feature (HB_TAG ('v','e','r','t'), F_GLOBAL_SEARCH);
+  }
+
+  for (unsigned int i = 0; i < num_user_features; i++)
+  {
+    const hb_feature_t *feature = &user_features[i];
+    map->add_feature (feature->tag,
+		      (feature->start == HB_FEATURE_GLOBAL_START &&
+		       feature->end == HB_FEATURE_GLOBAL_END) ?  F_GLOBAL : F_NONE,
+		      feature->value);
+  }
+
+#ifndef HB_NO_AAT_SHAPE
+  if (planner->apply_morx)
+  {
+    hb_aat_map_builder_t *aat_map = &planner->aat_map;
+    for (unsigned int i = 0; i < num_user_features; i++)
+    {
+      const hb_feature_t *feature = &user_features[i];
+      aat_map->add_feature (feature->tag, feature->value);
+    }
+  }
+#endif
+
+  if (planner->shaper->override_features)
+    planner->shaper->override_features (planner);
+}
+
+
+/*
+ * shaper face data
+ */
+
+struct hb_ot_face_data_t {};
+
+hb_ot_face_data_t *
+_hb_ot_shaper_face_data_create (hb_face_t *face)
+{
+  return (hb_ot_face_data_t *) HB_SHAPER_DATA_SUCCEEDED;
+}
+
+void
+_hb_ot_shaper_face_data_destroy (hb_ot_face_data_t *data)
+{
+}
+
+
+/*
+ * shaper font data
+ */
+
+struct hb_ot_font_data_t {};
+
+hb_ot_font_data_t *
+_hb_ot_shaper_font_data_create (hb_font_t *font HB_UNUSED)
+{
+  return (hb_ot_font_data_t *) HB_SHAPER_DATA_SUCCEEDED;
+}
+
+void
+_hb_ot_shaper_font_data_destroy (hb_ot_font_data_t *data HB_UNUSED)
+{
+}
+
+
+/*
+ * shaper
+ */
+
+struct hb_ot_shape_context_t
+{
+  hb_ot_shape_plan_t *plan;
+  hb_font_t *font;
+  hb_face_t *face;
+  hb_buffer_t  *buffer;
+  const hb_feature_t *user_features;
+  unsigned int        num_user_features;
+
+  /* Transient stuff */
+  hb_direction_t target_direction;
+};
+
+
+
+/* Main shaper */
+
+
+/* Prepare */
+
+static void
+hb_set_unicode_props (hb_buffer_t *buffer)
+{
+  /* Implement enough of Unicode Graphemes here that shaping
+   * in reverse-direction wouldn't break graphemes.  Namely,
+   * we mark all marks and ZWJ and ZWJ,Extended_Pictographic
+   * sequences as continuations.  The foreach_grapheme()
+   * macro uses this bit.
+   *
+   * https://www.unicode.org/reports/tr29/#Regex_Definitions
+   */
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    _hb_glyph_info_set_unicode_props (&info[i], buffer);
+
+    /* Marks are already set as continuation by the above line.
+     * Handle Emoji_Modifier and ZWJ-continuation. */
+    if (unlikely (_hb_glyph_info_get_general_category (&info[i]) == HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL &&
+		  hb_in_range<hb_codepoint_t> (info[i].codepoint, 0x1F3FBu, 0x1F3FFu)))
+    {
+	_hb_glyph_info_set_continuation (&info[i]);
+    }
+#ifndef HB_NO_EMOJI_SEQUENCES
+    else if (unlikely (_hb_glyph_info_is_zwj (&info[i])))
+    {
+      _hb_glyph_info_set_continuation (&info[i]);
+      if (i + 1 < count &&
+	  _hb_unicode_is_emoji_Extended_Pictographic (info[i + 1].codepoint))
+      {
+	i++;
+	_hb_glyph_info_set_unicode_props (&info[i], buffer);
+	_hb_glyph_info_set_continuation (&info[i]);
+      }
+    }
+#endif
+    /* Or part of the Other_Grapheme_Extend that is not marks.
+     * As of Unicode 11 that is just:
+     *
+     * 200C          ; Other_Grapheme_Extend # Cf       ZERO WIDTH NON-JOINER
+     * FF9E..FF9F    ; Other_Grapheme_Extend # Lm   [2] HALFWIDTH KATAKANA VOICED SOUND MARK..HALFWIDTH KATAKANA SEMI-VOICED SOUND MARK
+     * E0020..E007F  ; Other_Grapheme_Extend # Cf  [96] TAG SPACE..CANCEL TAG
+     *
+     * ZWNJ is special, we don't want to merge it as there's no need, and keeping
+     * it separate results in more granular clusters.  Ignore Katakana for now.
+     * Tags are used for Emoji sub-region flag sequences:
+     * https://github.com/harfbuzz/harfbuzz/issues/1556
+     */
+    else if (unlikely (hb_in_range<hb_codepoint_t> (info[i].codepoint, 0xE0020u, 0xE007Fu)))
+      _hb_glyph_info_set_continuation (&info[i]);
+  }
+}
+
+static void
+hb_insert_dotted_circle (hb_buffer_t *buffer, hb_font_t *font)
+{
+  if (unlikely (buffer->flags & HB_BUFFER_FLAG_DO_NOT_INSERT_DOTTED_CIRCLE))
+    return;
+
+  if (!(buffer->flags & HB_BUFFER_FLAG_BOT) ||
+      buffer->context_len[0] ||
+      !_hb_glyph_info_is_unicode_mark (&buffer->info[0]))
+    return;
+
+  if (!font->has_glyph (0x25CCu))
+    return;
+
+  hb_glyph_info_t dottedcircle = {0};
+  dottedcircle.codepoint = 0x25CCu;
+  _hb_glyph_info_set_unicode_props (&dottedcircle, buffer);
+
+  buffer->clear_output ();
+
+  buffer->idx = 0;
+  hb_glyph_info_t info = dottedcircle;
+  info.cluster = buffer->cur().cluster;
+  info.mask = buffer->cur().mask;
+  buffer->output_info (info);
+  while (buffer->idx < buffer->len && buffer->successful)
+    buffer->next_glyph ();
+  buffer->swap_buffers ();
+}
+
+static void
+hb_form_clusters (hb_buffer_t *buffer)
+{
+  if (!(buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_NON_ASCII))
+    return;
+
+  if (buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_GRAPHEMES)
+    foreach_grapheme (buffer, start, end)
+      buffer->merge_clusters (start, end);
+  else
+    foreach_grapheme (buffer, start, end)
+      buffer->unsafe_to_break (start, end);
+}
+
+static void
+hb_ensure_native_direction (hb_buffer_t *buffer)
+{
+  hb_direction_t direction = buffer->props.direction;
+  hb_direction_t horiz_dir = hb_script_get_horizontal_direction (buffer->props.script);
+
+  /* TODO vertical:
+   * The only BTT vertical script is Ogham, but it's not clear to me whether OpenType
+   * Ogham fonts are supposed to be implemented BTT or not.  Need to research that
+   * first. */
+  if ((HB_DIRECTION_IS_HORIZONTAL (direction) &&
+       direction != horiz_dir && horiz_dir != HB_DIRECTION_INVALID) ||
+      (HB_DIRECTION_IS_VERTICAL   (direction) &&
+       direction != HB_DIRECTION_TTB))
+  {
+
+    if (buffer->cluster_level == HB_BUFFER_CLUSTER_LEVEL_MONOTONE_CHARACTERS)
+      foreach_grapheme (buffer, start, end)
+      {
+	buffer->merge_clusters (start, end);
+	buffer->reverse_range (start, end);
+      }
+    else
+      foreach_grapheme (buffer, start, end)
+	/* form_clusters() merged clusters already, we don't merge. */
+	buffer->reverse_range (start, end);
+
+    buffer->reverse ();
+
+    buffer->props.direction = HB_DIRECTION_REVERSE (buffer->props.direction);
+  }
+}
+
+
+/*
+ * Substitute
+ */
+
+static hb_codepoint_t
+hb_vert_char_for (hb_codepoint_t u)
+{
+  switch (u >> 8)
+  {
+    case 0x20: switch (u) {
+      case 0x2013u: return 0xfe32u; // EN DASH
+      case 0x2014u: return 0xfe31u; // EM DASH
+      case 0x2025u: return 0xfe30u; // TWO DOT LEADER
+      case 0x2026u: return 0xfe19u; // HORIZONTAL ELLIPSIS
+    } break;
+    case 0x30: switch (u) {
+      case 0x3001u: return 0xfe11u; // IDEOGRAPHIC COMMA
+      case 0x3002u: return 0xfe12u; // IDEOGRAPHIC FULL STOP
+      case 0x3008u: return 0xfe3fu; // LEFT ANGLE BRACKET
+      case 0x3009u: return 0xfe40u; // RIGHT ANGLE BRACKET
+      case 0x300au: return 0xfe3du; // LEFT DOUBLE ANGLE BRACKET
+      case 0x300bu: return 0xfe3eu; // RIGHT DOUBLE ANGLE BRACKET
+      case 0x300cu: return 0xfe41u; // LEFT CORNER BRACKET
+      case 0x300du: return 0xfe42u; // RIGHT CORNER BRACKET
+      case 0x300eu: return 0xfe43u; // LEFT WHITE CORNER BRACKET
+      case 0x300fu: return 0xfe44u; // RIGHT WHITE CORNER BRACKET
+      case 0x3010u: return 0xfe3bu; // LEFT BLACK LENTICULAR BRACKET
+      case 0x3011u: return 0xfe3cu; // RIGHT BLACK LENTICULAR BRACKET
+      case 0x3014u: return 0xfe39u; // LEFT TORTOISE SHELL BRACKET
+      case 0x3015u: return 0xfe3au; // RIGHT TORTOISE SHELL BRACKET
+      case 0x3016u: return 0xfe17u; // LEFT WHITE LENTICULAR BRACKET
+      case 0x3017u: return 0xfe18u; // RIGHT WHITE LENTICULAR BRACKET
+    } break;
+    case 0xfe: switch (u) {
+      case 0xfe4fu: return 0xfe34u; // WAVY LOW LINE
+    } break;
+    case 0xff: switch (u) {
+      case 0xff01u: return 0xfe15u; // FULLWIDTH EXCLAMATION MARK
+      case 0xff08u: return 0xfe35u; // FULLWIDTH LEFT PARENTHESIS
+      case 0xff09u: return 0xfe36u; // FULLWIDTH RIGHT PARENTHESIS
+      case 0xff0cu: return 0xfe10u; // FULLWIDTH COMMA
+      case 0xff1au: return 0xfe13u; // FULLWIDTH COLON
+      case 0xff1bu: return 0xfe14u; // FULLWIDTH SEMICOLON
+      case 0xff1fu: return 0xfe16u; // FULLWIDTH QUESTION MARK
+      case 0xff3bu: return 0xfe47u; // FULLWIDTH LEFT SQUARE BRACKET
+      case 0xff3du: return 0xfe48u; // FULLWIDTH RIGHT SQUARE BRACKET
+      case 0xff3fu: return 0xfe33u; // FULLWIDTH LOW LINE
+      case 0xff5bu: return 0xfe37u; // FULLWIDTH LEFT CURLY BRACKET
+      case 0xff5du: return 0xfe38u; // FULLWIDTH RIGHT CURLY BRACKET
+    } break;
+  }
+
+  return u;
+}
+
+static inline void
+hb_ot_rotate_chars (const hb_ot_shape_context_t *c)
+{
+  hb_buffer_t *buffer = c->buffer;
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+
+  if (HB_DIRECTION_IS_BACKWARD (c->target_direction))
+  {
+    hb_unicode_funcs_t *unicode = buffer->unicode;
+    hb_mask_t rtlm_mask = c->plan->rtlm_mask;
+
+    for (unsigned int i = 0; i < count; i++) {
+      hb_codepoint_t codepoint = unicode->mirroring (info[i].codepoint);
+      if (unlikely (codepoint != info[i].codepoint && c->font->has_glyph (codepoint)))
+	info[i].codepoint = codepoint;
+      else
+	info[i].mask |= rtlm_mask;
+    }
+  }
+
+  if (HB_DIRECTION_IS_VERTICAL (c->target_direction) && !c->plan->has_vert)
+  {
+    for (unsigned int i = 0; i < count; i++) {
+      hb_codepoint_t codepoint = hb_vert_char_for (info[i].codepoint);
+      if (unlikely (codepoint != info[i].codepoint && c->font->has_glyph (codepoint)))
+	info[i].codepoint = codepoint;
+    }
+  }
+}
+
+static inline void
+hb_ot_shape_setup_masks_fraction (const hb_ot_shape_context_t *c)
+{
+#ifdef HB_NO_OT_SHAPE_FRACTIONS
+  return;
+#endif
+
+  if (!(c->buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_NON_ASCII) ||
+      !c->plan->has_frac)
+    return;
+
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_mask_t pre_mask, post_mask;
+  if (HB_DIRECTION_IS_FORWARD (buffer->props.direction))
+  {
+    pre_mask = c->plan->numr_mask | c->plan->frac_mask;
+    post_mask = c->plan->frac_mask | c->plan->dnom_mask;
+  }
+  else
+  {
+    pre_mask = c->plan->frac_mask | c->plan->dnom_mask;
+    post_mask = c->plan->numr_mask | c->plan->frac_mask;
+  }
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    if (info[i].codepoint == 0x2044u) /* FRACTION SLASH */
+    {
+      unsigned int start = i, end = i + 1;
+      while (start &&
+	     _hb_glyph_info_get_general_category (&info[start - 1]) ==
+	     HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER)
+	start--;
+      while (end < count &&
+	     _hb_glyph_info_get_general_category (&info[end]) ==
+	     HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER)
+	end++;
+
+      buffer->unsafe_to_break (start, end);
+
+      for (unsigned int j = start; j < i; j++)
+	info[j].mask |= pre_mask;
+      info[i].mask |= c->plan->frac_mask;
+      for (unsigned int j = i + 1; j < end; j++)
+	info[j].mask |= post_mask;
+
+      i = end - 1;
+    }
+  }
+}
+
+static inline void
+hb_ot_shape_initialize_masks (const hb_ot_shape_context_t *c)
+{
+  hb_ot_map_t *map = &c->plan->map;
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_mask_t global_mask = map->get_global_mask ();
+  buffer->reset_masks (global_mask);
+}
+
+static inline void
+hb_ot_shape_setup_masks (const hb_ot_shape_context_t *c)
+{
+  hb_ot_map_t *map = &c->plan->map;
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_ot_shape_setup_masks_fraction (c);
+
+  if (c->plan->shaper->setup_masks)
+    c->plan->shaper->setup_masks (c->plan, buffer, c->font);
+
+  for (unsigned int i = 0; i < c->num_user_features; i++)
+  {
+    const hb_feature_t *feature = &c->user_features[i];
+    if (!(feature->start == HB_FEATURE_GLOBAL_START && feature->end == HB_FEATURE_GLOBAL_END)) {
+      unsigned int shift;
+      hb_mask_t mask = map->get_mask (feature->tag, &shift);
+      buffer->set_masks (feature->value << shift, mask, feature->start, feature->end);
+    }
+  }
+}
+
+static void
+hb_ot_zero_width_default_ignorables (const hb_buffer_t *buffer)
+{
+  if (!(buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_DEFAULT_IGNORABLES) ||
+      (buffer->flags & HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES) ||
+      (buffer->flags & HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES))
+    return;
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  hb_glyph_position_t *pos = buffer->pos;
+  unsigned int i = 0;
+  for (i = 0; i < count; i++)
+    if (unlikely (_hb_glyph_info_is_default_ignorable (&info[i])))
+      pos[i].x_advance = pos[i].y_advance = pos[i].x_offset = pos[i].y_offset = 0;
+}
+
+static void
+hb_ot_hide_default_ignorables (hb_buffer_t *buffer,
+			       hb_font_t   *font)
+{
+  if (!(buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_DEFAULT_IGNORABLES) ||
+      (buffer->flags & HB_BUFFER_FLAG_PRESERVE_DEFAULT_IGNORABLES))
+    return;
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+
+  hb_codepoint_t invisible = buffer->invisible;
+  if (!(buffer->flags & HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES) &&
+      (invisible || font->get_nominal_glyph (' ', &invisible)))
+  {
+    /* Replace default-ignorables with a zero-advance invisible glyph. */
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (_hb_glyph_info_is_default_ignorable (&info[i]))
+	info[i].codepoint = invisible;
+    }
+  }
+  else
+    hb_ot_layout_delete_glyphs_inplace (buffer, _hb_glyph_info_is_default_ignorable);
+}
+
+
+static inline void
+hb_ot_map_glyphs_fast (hb_buffer_t  *buffer)
+{
+  /* Normalization process sets up glyph_index(), we just copy it. */
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    info[i].codepoint = info[i].glyph_index();
+
+  buffer->content_type = HB_BUFFER_CONTENT_TYPE_GLYPHS;
+}
+
+static inline void
+hb_synthesize_glyph_classes (hb_buffer_t *buffer)
+{
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+  {
+    hb_ot_layout_glyph_props_flags_t klass;
+
+    /* Never mark default-ignorables as marks.
+     * They won't get in the way of lookups anyway,
+     * but having them as mark will cause them to be skipped
+     * over if the lookup-flag says so, but at least for the
+     * Mongolian variation selectors, looks like Uniscribe
+     * marks them as non-mark.  Some Mongolian fonts without
+     * GDEF rely on this.  Another notable character that
+     * this applies to is COMBINING GRAPHEME JOINER. */
+    klass = (_hb_glyph_info_get_general_category (&info[i]) !=
+	     HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK ||
+	     _hb_glyph_info_is_default_ignorable (&info[i])) ?
+	    HB_OT_LAYOUT_GLYPH_PROPS_BASE_GLYPH :
+	    HB_OT_LAYOUT_GLYPH_PROPS_MARK;
+    _hb_glyph_info_set_glyph_props (&info[i], klass);
+  }
+}
+
+static inline void
+hb_ot_substitute_default (const hb_ot_shape_context_t *c)
+{
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_ot_rotate_chars (c);
+
+  HB_BUFFER_ALLOCATE_VAR (buffer, glyph_index);
+
+  _hb_ot_shape_normalize (c->plan, buffer, c->font);
+
+  hb_ot_shape_setup_masks (c);
+
+  /* This is unfortunate to go here, but necessary... */
+  if (c->plan->fallback_mark_positioning)
+    _hb_ot_shape_fallback_mark_position_recategorize_marks (c->plan, c->font, buffer);
+
+  hb_ot_map_glyphs_fast (buffer);
+
+  HB_BUFFER_DEALLOCATE_VAR (buffer, glyph_index);
+}
+
+static inline void
+hb_ot_substitute_complex (const hb_ot_shape_context_t *c)
+{
+  hb_buffer_t *buffer = c->buffer;
+
+  hb_ot_layout_substitute_start (c->font, buffer);
+
+  if (c->plan->fallback_glyph_classes)
+    hb_synthesize_glyph_classes (c->buffer);
+
+  c->plan->substitute (c->font, buffer);
+}
+
+static inline void
+hb_ot_substitute_pre (const hb_ot_shape_context_t *c)
+{
+  hb_ot_substitute_default (c);
+
+  _hb_buffer_allocate_gsubgpos_vars (c->buffer);
+
+  hb_ot_substitute_complex (c);
+}
+
+static inline void
+hb_ot_substitute_post (const hb_ot_shape_context_t *c)
+{
+  hb_ot_hide_default_ignorables (c->buffer, c->font);
+#ifndef HB_NO_AAT_SHAPE
+  if (c->plan->apply_morx)
+    hb_aat_layout_remove_deleted_glyphs (c->buffer);
+#endif
+
+  if (c->plan->shaper->postprocess_glyphs)
+    c->plan->shaper->postprocess_glyphs (c->plan, c->buffer, c->font);
+}
+
+
+/*
+ * Position
+ */
+
+static inline void
+adjust_mark_offsets (hb_glyph_position_t *pos)
+{
+  pos->x_offset -= pos->x_advance;
+  pos->y_offset -= pos->y_advance;
+}
+
+static inline void
+zero_mark_width (hb_glyph_position_t *pos)
+{
+  pos->x_advance = 0;
+  pos->y_advance = 0;
+}
+
+static inline void
+zero_mark_widths_by_gdef (hb_buffer_t *buffer, bool adjust_offsets)
+{
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    if (_hb_glyph_info_is_mark (&info[i]))
+    {
+      if (adjust_offsets)
+	adjust_mark_offsets (&buffer->pos[i]);
+      zero_mark_width (&buffer->pos[i]);
+    }
+}
+
+static inline void
+hb_ot_position_default (const hb_ot_shape_context_t *c)
+{
+  hb_direction_t direction = c->buffer->props.direction;
+  unsigned int count = c->buffer->len;
+  hb_glyph_info_t *info = c->buffer->info;
+  hb_glyph_position_t *pos = c->buffer->pos;
+
+  if (HB_DIRECTION_IS_HORIZONTAL (direction))
+  {
+    c->font->get_glyph_h_advances (count, &info[0].codepoint, sizeof(info[0]),
+				   &pos[0].x_advance, sizeof(pos[0]));
+    /* The nil glyph_h_origin() func returns 0, so no need to apply it. */
+    if (c->font->has_glyph_h_origin_func ())
+      for (unsigned int i = 0; i < count; i++)
+	c->font->subtract_glyph_h_origin (info[i].codepoint,
+					  &pos[i].x_offset,
+					  &pos[i].y_offset);
+  }
+  else
+  {
+    c->font->get_glyph_v_advances (count, &info[0].codepoint, sizeof(info[0]),
+				   &pos[0].y_advance, sizeof(pos[0]));
+    for (unsigned int i = 0; i < count; i++)
+    {
+      c->font->subtract_glyph_v_origin (info[i].codepoint,
+					&pos[i].x_offset,
+					&pos[i].y_offset);
+    }
+  }
+  if (c->buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_SPACE_FALLBACK)
+    _hb_ot_shape_fallback_spaces (c->plan, c->font, c->buffer);
+}
+
+static inline void
+hb_ot_position_complex (const hb_ot_shape_context_t *c)
+{
+  unsigned int count = c->buffer->len;
+  hb_glyph_info_t *info = c->buffer->info;
+  hb_glyph_position_t *pos = c->buffer->pos;
+
+  /* If the font has no GPOS and direction is forward, then when
+   * zeroing mark widths, we shift the mark with it, such that the
+   * mark is positioned hanging over the previous glyph.  When
+   * direction is backward we don't shift and it will end up
+   * hanging over the next glyph after the final reordering.
+   *
+   * Note: If fallback positinoing happens, we don't care about
+   * this as it will be overriden.
+   */
+  bool adjust_offsets_when_zeroing = c->plan->adjust_mark_positioning_when_zeroing &&
+				     HB_DIRECTION_IS_FORWARD (c->buffer->props.direction);
+
+  /* We change glyph origin to what GPOS expects (horizontal), apply GPOS, change it back. */
+
+  /* The nil glyph_h_origin() func returns 0, so no need to apply it. */
+  if (c->font->has_glyph_h_origin_func ())
+    for (unsigned int i = 0; i < count; i++)
+      c->font->add_glyph_h_origin (info[i].codepoint,
+				   &pos[i].x_offset,
+				   &pos[i].y_offset);
+
+  hb_ot_layout_position_start (c->font, c->buffer);
+
+  if (c->plan->zero_marks)
+    switch (c->plan->shaper->zero_width_marks)
+    {
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY:
+	zero_mark_widths_by_gdef (c->buffer, adjust_offsets_when_zeroing);
+	break;
+
+      default:
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE:
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE:
+	break;
+    }
+
+  c->plan->position (c->font, c->buffer);
+
+  if (c->plan->zero_marks)
+    switch (c->plan->shaper->zero_width_marks)
+    {
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE:
+	zero_mark_widths_by_gdef (c->buffer, adjust_offsets_when_zeroing);
+	break;
+
+      default:
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_NONE:
+      case HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY:
+	break;
+    }
+
+  /* Finish off.  Has to follow a certain order. */
+  hb_ot_layout_position_finish_advances (c->font, c->buffer);
+  hb_ot_zero_width_default_ignorables (c->buffer);
+#ifndef HB_NO_AAT_SHAPE
+  if (c->plan->apply_morx)
+    hb_aat_layout_zero_width_deleted_glyphs (c->buffer);
+#endif
+  hb_ot_layout_position_finish_offsets (c->font, c->buffer);
+
+  /* The nil glyph_h_origin() func returns 0, so no need to apply it. */
+  if (c->font->has_glyph_h_origin_func ())
+    for (unsigned int i = 0; i < count; i++)
+      c->font->subtract_glyph_h_origin (info[i].codepoint,
+					&pos[i].x_offset,
+					&pos[i].y_offset);
+
+  if (c->plan->fallback_mark_positioning)
+    _hb_ot_shape_fallback_mark_position (c->plan, c->font, c->buffer,
+					 adjust_offsets_when_zeroing);
+}
+
+static inline void
+hb_ot_position (const hb_ot_shape_context_t *c)
+{
+  c->buffer->clear_positions ();
+
+  hb_ot_position_default (c);
+
+  hb_ot_position_complex (c);
+
+  if (HB_DIRECTION_IS_BACKWARD (c->buffer->props.direction))
+    hb_buffer_reverse (c->buffer);
+
+  _hb_buffer_deallocate_gsubgpos_vars (c->buffer);
+}
+
+static inline void
+hb_propagate_flags (hb_buffer_t *buffer)
+{
+  /* Propagate cluster-level glyph flags to be the same on all cluster glyphs.
+   * Simplifies using them. */
+
+  if (!(buffer->scratch_flags & HB_BUFFER_SCRATCH_FLAG_HAS_UNSAFE_TO_BREAK))
+    return;
+
+  hb_glyph_info_t *info = buffer->info;
+
+  foreach_cluster (buffer, start, end)
+  {
+    unsigned int mask = 0;
+    for (unsigned int i = start; i < end; i++)
+      if (info[i].mask & HB_GLYPH_FLAG_UNSAFE_TO_BREAK)
+      {
+	 mask = HB_GLYPH_FLAG_UNSAFE_TO_BREAK;
+	 break;
+      }
+    if (mask)
+      for (unsigned int i = start; i < end; i++)
+	info[i].mask |= mask;
+  }
+}
+
+/* Pull it all together! */
+
+static void
+hb_ot_shape_internal (hb_ot_shape_context_t *c)
+{
+  c->buffer->deallocate_var_all ();
+  c->buffer->scratch_flags = HB_BUFFER_SCRATCH_FLAG_DEFAULT;
+  if (likely (!hb_unsigned_mul_overflows (c->buffer->len, HB_BUFFER_MAX_LEN_FACTOR)))
+  {
+    c->buffer->max_len = hb_max (c->buffer->len * HB_BUFFER_MAX_LEN_FACTOR,
+				 (unsigned) HB_BUFFER_MAX_LEN_MIN);
+  }
+  if (likely (!hb_unsigned_mul_overflows (c->buffer->len, HB_BUFFER_MAX_OPS_FACTOR)))
+  {
+    c->buffer->max_ops = hb_max (c->buffer->len * HB_BUFFER_MAX_OPS_FACTOR,
+				 (unsigned) HB_BUFFER_MAX_OPS_MIN);
+  }
+
+  /* Save the original direction, we use it later. */
+  c->target_direction = c->buffer->props.direction;
+
+  _hb_buffer_allocate_unicode_vars (c->buffer);
+
+  c->buffer->clear_output ();
+
+  hb_ot_shape_initialize_masks (c);
+  hb_set_unicode_props (c->buffer);
+  hb_insert_dotted_circle (c->buffer, c->font);
+
+  hb_form_clusters (c->buffer);
+
+  hb_ensure_native_direction (c->buffer);
+
+  if (c->plan->shaper->preprocess_text)
+    c->plan->shaper->preprocess_text (c->plan, c->buffer, c->font);
+
+  hb_ot_substitute_pre (c);
+  hb_ot_position (c);
+  hb_ot_substitute_post (c);
+
+  hb_propagate_flags (c->buffer);
+
+  _hb_buffer_deallocate_unicode_vars (c->buffer);
+
+  c->buffer->props.direction = c->target_direction;
+
+  c->buffer->max_len = HB_BUFFER_MAX_LEN_DEFAULT;
+  c->buffer->max_ops = HB_BUFFER_MAX_OPS_DEFAULT;
+  c->buffer->deallocate_var_all ();
+}
+
+
+hb_bool_t
+_hb_ot_shape (hb_shape_plan_t    *shape_plan,
+	      hb_font_t          *font,
+	      hb_buffer_t        *buffer,
+	      const hb_feature_t *features,
+	      unsigned int        num_features)
+{
+  hb_ot_shape_context_t c = {&shape_plan->ot, font, font->face, buffer, features, num_features};
+  hb_ot_shape_internal (&c);
+
+  return true;
+}
+
+
+/**
+ * hb_ot_shape_plan_collect_lookups:
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_ot_shape_plan_collect_lookups (hb_shape_plan_t *shape_plan,
+				  hb_tag_t         table_tag,
+				  hb_set_t        *lookup_indexes /* OUT */)
+{
+  shape_plan->ot.collect_lookups (table_tag, lookup_indexes);
+}
+
+
+/* TODO Move this to hb-ot-shape-normalize, make it do decompose, and make it public. */
+static void
+add_char (hb_font_t          *font,
+	  hb_unicode_funcs_t *unicode,
+	  hb_bool_t           mirror,
+	  hb_codepoint_t      u,
+	  hb_set_t           *glyphs)
+{
+  hb_codepoint_t glyph;
+  if (font->get_nominal_glyph (u, &glyph))
+    glyphs->add (glyph);
+  if (mirror)
+  {
+    hb_codepoint_t m = unicode->mirroring (u);
+    if (m != u && font->get_nominal_glyph (m, &glyph))
+      glyphs->add (glyph);
+  }
+}
+
+
+/**
+ * hb_ot_shape_glyphs_closure:
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_ot_shape_glyphs_closure (hb_font_t          *font,
+			    hb_buffer_t        *buffer,
+			    const hb_feature_t *features,
+			    unsigned int        num_features,
+			    hb_set_t           *glyphs)
+{
+  const char *shapers[] = {"ot", nullptr};
+  hb_shape_plan_t *shape_plan = hb_shape_plan_create_cached (font->face, &buffer->props,
+							     features, num_features, shapers);
+
+  bool mirror = hb_script_get_horizontal_direction (buffer->props.script) == HB_DIRECTION_RTL;
+
+  unsigned int count = buffer->len;
+  hb_glyph_info_t *info = buffer->info;
+  for (unsigned int i = 0; i < count; i++)
+    add_char (font, buffer->unicode, mirror, info[i].codepoint, glyphs);
+
+  hb_set_t *lookups = hb_set_create ();
+  hb_ot_shape_plan_collect_lookups (shape_plan, HB_OT_TAG_GSUB, lookups);
+  hb_ot_layout_lookups_substitute_closure (font->face, lookups, glyphs);
+
+  hb_set_destroy (lookups);
+
+  hb_shape_plan_destroy (shape_plan);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape.h b/thirdparty/harfbuzz/src/hb-ot-shape.h
new file mode 100644
index 0000000000..7b1bcc0637
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape.h
@@ -0,0 +1,53 @@
+/*
+ * Copyright © 2013  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_SHAPE_H
+#define HB_OT_SHAPE_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+/* TODO port to shape-plan / set. */
+HB_EXTERN void
+hb_ot_shape_glyphs_closure (hb_font_t          *font,
+			    hb_buffer_t        *buffer,
+			    const hb_feature_t *features,
+			    unsigned int        num_features,
+			    hb_set_t           *glyphs);
+
+HB_EXTERN void
+hb_ot_shape_plan_collect_lookups (hb_shape_plan_t *shape_plan,
+				  hb_tag_t         table_tag,
+				  hb_set_t        *lookup_indexes /* OUT */);
+
+HB_END_DECLS
+
+#endif /* HB_OT_SHAPE_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-shape.hh b/thirdparty/harfbuzz/src/hb-ot-shape.hh
new file mode 100644
index 0000000000..acc98772a9
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-shape.hh
@@ -0,0 +1,170 @@
+/*
+ * Copyright © 2010  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_SHAPE_HH
+#define HB_OT_SHAPE_HH
+
+#include "hb.hh"
+
+#include "hb-ot-map.hh"
+#include "hb-aat-map.hh"
+
+
+struct hb_ot_shape_plan_key_t
+{
+  unsigned int variations_index[2];
+
+  void init (hb_face_t *face,
+	     const int *coords,
+	     unsigned   num_coords)
+  {
+    for (unsigned int table_index = 0; table_index < 2; table_index++)
+      hb_ot_layout_table_find_feature_variations (face,
+						  table_tags[table_index],
+						  coords,
+						  num_coords,
+						  &variations_index[table_index]);
+  }
+
+  bool equal (const hb_ot_shape_plan_key_t *other)
+  {
+    return 0 == memcmp (this, other, sizeof (*this));
+  }
+};
+
+
+struct hb_shape_plan_key_t;
+
+struct hb_ot_shape_plan_t
+{
+  hb_segment_properties_t props;
+  const struct hb_ot_complex_shaper_t *shaper;
+  hb_ot_map_t map;
+  hb_aat_map_t aat_map;
+  const void *data;
+#ifndef HB_NO_OT_SHAPE_FRACTIONS
+  hb_mask_t frac_mask, numr_mask, dnom_mask;
+#else
+  static constexpr hb_mask_t frac_mask = 0;
+  static constexpr hb_mask_t numr_mask = 0;
+  static constexpr hb_mask_t dnom_mask = 0;
+#endif
+  hb_mask_t rtlm_mask;
+#ifndef HB_NO_OT_KERN
+  hb_mask_t kern_mask;
+#else
+  static constexpr hb_mask_t kern_mask = 0;
+#endif
+#ifndef HB_NO_AAT_SHAPE
+  hb_mask_t trak_mask;
+#else
+  static constexpr hb_mask_t trak_mask = 0;
+#endif
+
+#ifndef HB_NO_OT_KERN
+  bool requested_kerning : 1;
+#else
+  static constexpr bool requested_kerning = false;
+#endif
+#ifndef HB_NO_AAT_SHAPE
+  bool requested_tracking : 1;
+#else
+  static constexpr bool requested_tracking = false;
+#endif
+#ifndef HB_NO_OT_SHAPE_FRACTIONS
+  bool has_frac : 1;
+#else
+  static constexpr bool has_frac = false;
+#endif
+  bool has_vert : 1;
+  bool has_gpos_mark : 1;
+  bool zero_marks : 1;
+  bool fallback_glyph_classes : 1;
+  bool fallback_mark_positioning : 1;
+  bool adjust_mark_positioning_when_zeroing : 1;
+
+  bool apply_gpos : 1;
+#ifndef HB_NO_OT_KERN
+  bool apply_kern : 1;
+#else
+  static constexpr bool apply_kern = false;
+#endif
+#ifndef HB_NO_AAT_SHAPE
+  bool apply_kerx : 1;
+  bool apply_morx : 1;
+  bool apply_trak : 1;
+#else
+  static constexpr bool apply_kerx = false;
+  static constexpr bool apply_morx = false;
+  static constexpr bool apply_trak = false;
+#endif
+
+  void collect_lookups (hb_tag_t table_tag, hb_set_t *lookups) const
+  {
+    unsigned int table_index;
+    switch (table_tag) {
+      case HB_OT_TAG_GSUB: table_index = 0; break;
+      case HB_OT_TAG_GPOS: table_index = 1; break;
+      default: return;
+    }
+    map.collect_lookups (table_index, lookups);
+  }
+
+  HB_INTERNAL bool init0 (hb_face_t                     *face,
+			  const hb_shape_plan_key_t     *key);
+  HB_INTERNAL void fini ();
+
+  HB_INTERNAL void substitute (hb_font_t *font, hb_buffer_t *buffer) const;
+  HB_INTERNAL void position (hb_font_t *font, hb_buffer_t *buffer) const;
+};
+
+struct hb_shape_plan_t;
+
+struct hb_ot_shape_planner_t
+{
+  /* In the order that they are filled in. */
+  hb_face_t *face;
+  hb_segment_properties_t props;
+  hb_ot_map_builder_t map;
+  hb_aat_map_builder_t aat_map;
+#ifndef HB_NO_AAT_SHAPE
+  bool apply_morx : 1;
+#else
+  static constexpr bool apply_morx = false;
+#endif
+  bool script_zero_marks : 1;
+  bool script_fallback_mark_positioning : 1;
+  const struct hb_ot_complex_shaper_t *shaper;
+
+  HB_INTERNAL hb_ot_shape_planner_t (hb_face_t                     *face,
+				     const hb_segment_properties_t *props);
+
+  HB_INTERNAL void compile (hb_ot_shape_plan_t           &plan,
+			    const hb_ot_shape_plan_key_t &key);
+};
+
+
+#endif /* HB_OT_SHAPE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-stat-table.hh b/thirdparty/harfbuzz/src/hb-ot-stat-table.hh
new file mode 100644
index 0000000000..6aa4fa4492
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-stat-table.hh
@@ -0,0 +1,404 @@
+/*
+ * Copyright © 2018  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_OT_STAT_TABLE_HH
+#define HB_OT_STAT_TABLE_HH
+
+#include "hb-open-type.hh"
+#include "hb-ot-layout-common.hh"
+
+/*
+ * STAT -- Style Attributes
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/stat
+ */
+#define HB_OT_TAG_STAT HB_TAG('S','T','A','T')
+
+
+namespace OT {
+
+enum
+{
+  OLDER_SIBLING_FONT_ATTRIBUTE = 0x0001,	/* If set, this axis value table
+						 * provides axis value information
+						 * that is applicable to other fonts
+						 * within the same font family. This
+						 * is used if the other fonts were
+						 * released earlier and did not include
+						 * information about values for some axis.
+						 * If newer versions of the other
+						 * fonts include the information
+						 * themselves and are present,
+						 * then this record is ignored. */
+  ELIDABLE_AXIS_VALUE_NAME = 0x0002		/* If set, it indicates that the axis
+						 * value represents the “normal” value
+						 * for the axis and may be omitted when
+						 * composing name strings. */
+  // Reserved = 0xFFFC				/* Reserved for future use — set to zero. */
+};
+
+struct AxisValueFormat1
+{
+  unsigned int get_axis_index () const { return axisIndex; }
+  float get_value ()             const { return value.to_float (); }
+
+  hb_ot_name_id_t get_value_name_id () const { return valueNameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier — set to 1. */
+  HBUINT16	axisIndex;	/* Zero-base index into the axis record array
+				 * identifying the axis of design variation
+				 * to which the axis value record applies.
+				 * Must be less than designAxisCount. */
+  HBUINT16	flags;		/* Flags — see below for details. */
+  NameID	valueNameID;	/* The name ID for entries in the 'name' table
+				 * that provide a display string for this
+				 * attribute value. */
+  HBFixed	value;		/* A numeric value for this attribute value. */
+  public:
+  DEFINE_SIZE_STATIC (12);
+};
+
+struct AxisValueFormat2
+{
+  unsigned int get_axis_index () const { return axisIndex; }
+  float get_value ()             const { return nominalValue.to_float (); }
+
+  hb_ot_name_id_t get_value_name_id () const { return valueNameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier — set to 2. */
+  HBUINT16	axisIndex;	/* Zero-base index into the axis record array
+				 * identifying the axis of design variation
+				 * to which the axis value record applies.
+				 * Must be less than designAxisCount. */
+  HBUINT16	flags;		/* Flags — see below for details. */
+  NameID	valueNameID;	/* The name ID for entries in the 'name' table
+				 * that provide a display string for this
+				 * attribute value. */
+  HBFixed	nominalValue;	/* A numeric value for this attribute value. */
+  HBFixed	rangeMinValue;	/* The minimum value for a range associated
+				 * with the specified name ID. */
+  HBFixed	rangeMaxValue;	/* The maximum value for a range associated
+				 * with the specified name ID. */
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+struct AxisValueFormat3
+{
+  unsigned int get_axis_index () const { return axisIndex; }
+  float get_value ()             const { return value.to_float (); }
+
+  hb_ot_name_id_t get_value_name_id () const { return valueNameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier — set to 3. */
+  HBUINT16	axisIndex;	/* Zero-base index into the axis record array
+				 * identifying the axis of design variation
+				 * to which the axis value record applies.
+				 * Must be less than designAxisCount. */
+  HBUINT16	flags;		/* Flags — see below for details. */
+  NameID	valueNameID;	/* The name ID for entries in the 'name' table
+				 * that provide a display string for this
+				 * attribute value. */
+  HBFixed	value;		/* A numeric value for this attribute value. */
+  HBFixed	linkedValue;	/* The numeric value for a style-linked mapping
+				 * from this value. */
+  public:
+  DEFINE_SIZE_STATIC (16);
+};
+
+struct AxisValueRecord
+{
+  unsigned int get_axis_index () const { return axisIndex; }
+  float get_value ()             const { return value.to_float (); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	axisIndex;	/* Zero-base index into the axis record array
+				 * identifying the axis to which this value
+				 * applies. Must be less than designAxisCount. */
+  HBFixed	value;		/* A numeric value for this attribute value. */
+  public:
+  DEFINE_SIZE_STATIC (6);
+};
+
+struct AxisValueFormat4
+{
+  const AxisValueRecord &get_axis_record (unsigned int axis_index) const
+  { return axisValues.as_array (axisCount)[axis_index]; }
+
+  hb_ot_name_id_t get_value_name_id () const { return valueNameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  HBUINT16	format;		/* Format identifier — set to 4. */
+  HBUINT16	axisCount;	/* The total number of axes contributing to
+				 * this axis-values combination. */
+  HBUINT16	flags;		/* Flags — see below for details. */
+  NameID	valueNameID;	/* The name ID for entries in the 'name' table
+				 * that provide a display string for this
+				 * attribute value. */
+  UnsizedArrayOf<AxisValueRecord>
+		axisValues;	/* Array of AxisValue records that provide the
+				 * combination of axis values, one for each
+				 * contributing axis. */
+  public:
+  DEFINE_SIZE_ARRAY (8, axisValues);
+};
+
+struct AxisValue
+{
+  bool get_value (unsigned int axis_index) const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.get_value ();
+    case 2: return u.format2.get_value ();
+    case 3: return u.format3.get_value ();
+    case 4: return u.format4.get_axis_record (axis_index).get_value ();
+    default:return 0;
+    }
+  }
+
+  unsigned int get_axis_index () const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.get_axis_index ();
+    case 2: return u.format2.get_axis_index ();
+    case 3: return u.format3.get_axis_index ();
+    /* case 4: Makes more sense for variable fonts which are handled by fvar in hb-style */
+    default:return -1;
+    }
+  }
+
+  hb_ot_name_id_t get_value_name_id () const
+  {
+    switch (u.format)
+    {
+    case 1: return u.format1.get_value_name_id ();
+    case 2: return u.format2.get_value_name_id ();
+    case 3: return u.format3.get_value_name_id ();
+    case 4: return u.format4.get_value_name_id ();
+    default:return HB_OT_NAME_ID_INVALID;
+    }
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!c->check_struct (this)))
+      return_trace (false);
+
+    switch (u.format)
+    {
+    case 1: return_trace (u.format1.sanitize (c));
+    case 2: return_trace (u.format2.sanitize (c));
+    case 3: return_trace (u.format3.sanitize (c));
+    case 4: return_trace (u.format4.sanitize (c));
+    default:return_trace (true);
+    }
+  }
+
+  protected:
+  union
+  {
+  HBUINT16		format;
+  AxisValueFormat1	format1;
+  AxisValueFormat2	format2;
+  AxisValueFormat3	format3;
+  AxisValueFormat4	format4;
+  } u;
+  public:
+  DEFINE_SIZE_UNION (2, format);
+};
+
+struct StatAxisRecord
+{
+  int cmp (hb_tag_t key) const { return tag.cmp (key); }
+
+  hb_ot_name_id_t get_name_id () const { return nameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this)));
+  }
+
+  protected:
+  Tag		tag;		/* A tag identifying the axis of design variation. */
+  NameID	nameID;		/* The name ID for entries in the 'name' table that
+				 * provide a display string for this axis. */
+  HBUINT16	ordering;	/* A value that applications can use to determine
+				 * primary sorting of face names, or for ordering
+				 * of descriptors when composing family or face names. */
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+struct STAT
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_STAT;
+
+  bool has_data () const { return version.to_int (); }
+
+  bool get_value (hb_tag_t tag, float *value) const
+  {
+    unsigned int axis_index;
+    if (!get_design_axes ().lfind (tag, &axis_index)) return false;
+
+    hb_array_t<const OffsetTo<AxisValue>> axis_values = get_axis_value_offsets ();
+    for (unsigned int i = 0; i < axis_values.length; i++)
+    {
+      const AxisValue& axis_value = this+axis_values[i];
+      if (axis_value.get_axis_index () == axis_index)
+      {
+	if (value)
+	  *value = axis_value.get_value (axis_index);
+	return true;
+      }
+    }
+    return false;
+  }
+
+  unsigned get_design_axis_count () const { return designAxisCount; }
+
+  hb_ot_name_id_t get_axis_record_name_id (unsigned axis_record_index) const
+  {
+    if (unlikely (axis_record_index >= designAxisCount)) return HB_OT_NAME_ID_INVALID;
+    const StatAxisRecord &axis_record = get_design_axes ()[axis_record_index];
+    return axis_record.get_name_id ();
+  }
+
+  unsigned get_axis_value_count () const { return axisValueCount; }
+
+  hb_ot_name_id_t get_axis_value_name_id (unsigned axis_value_index) const
+  {
+    if (unlikely (axis_value_index >= axisValueCount)) return HB_OT_NAME_ID_INVALID;
+    const AxisValue &axis_value = (this + get_axis_value_offsets ()[axis_value_index]);
+    return axis_value.get_value_name_id ();
+  }
+
+  void collect_name_ids (hb_set_t *nameids_to_retain) const
+  {
+    if (!has_data ()) return;
+
+    + get_design_axes ()
+    | hb_map (&StatAxisRecord::get_name_id)
+    | hb_sink (nameids_to_retain)
+    ;
+
+    + get_axis_value_offsets ()
+    | hb_map (hb_add (&(this + offsetToAxisValueOffsets)))
+    | hb_map (&AxisValue::get_value_name_id)
+    | hb_sink (nameids_to_retain)
+    ;
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (likely (c->check_struct (this) &&
+			  version.major == 1 &&
+			  version.minor > 0 &&
+			  designAxesOffset.sanitize (c, this, designAxisCount) &&
+			  offsetToAxisValueOffsets.sanitize (c, this, axisValueCount, &(this+offsetToAxisValueOffsets))));
+  }
+
+  protected:
+  hb_array_t<const StatAxisRecord> const get_design_axes () const
+  { return (this+designAxesOffset).as_array (designAxisCount); }
+
+  hb_array_t<const OffsetTo<AxisValue>> const get_axis_value_offsets () const
+  { return (this+offsetToAxisValueOffsets).as_array (axisValueCount); }
+
+
+  protected:
+  FixedVersion<>version;	/* Version of the stat table
+				 * initially set to 0x00010002u */
+  HBUINT16	designAxisSize;	/* The size in bytes of each axis record. */
+  HBUINT16	designAxisCount;/* The number of design axis records. In a
+				 * font with an 'fvar' table, this value must be
+				 * greater than or equal to the axisCount value
+				 * in the 'fvar' table. In all fonts, must
+				 * be greater than zero if axisValueCount
+				 * is greater than zero. */
+  LNNOffsetTo<UnsizedArrayOf<StatAxisRecord>>
+		designAxesOffset;
+				/* Offset in bytes from the beginning of
+				 * the STAT table to the start of the design
+				 * axes array. If designAxisCount is zero,
+				 * set to zero; if designAxisCount is greater
+				 * than zero, must be greater than zero. */
+  HBUINT16	axisValueCount;	/* The number of axis value tables. */
+  LNNOffsetTo<UnsizedArrayOf<OffsetTo<AxisValue>>>
+		offsetToAxisValueOffsets;
+				/* Offset in bytes from the beginning of
+				 * the STAT table to the start of the design
+				 * axes value offsets array. If axisValueCount
+				 * is zero, set to zero; if axisValueCount is
+				 * greater than zero, must be greater than zero. */
+  NameID	elidedFallbackNameID;
+				/* Name ID used as fallback when projection of
+				 * names into a particular font model produces
+				 * a subfamily name containing only elidable
+				 * elements. */
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_STAT_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-tag-table.hh b/thirdparty/harfbuzz/src/hb-ot-tag-table.hh
new file mode 100644
index 0000000000..99937d9f69
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-tag-table.hh
@@ -0,0 +1,2176 @@
+/* == Start of generated table == */
+/*
+ * The following table is generated by running:
+ *
+ *   ./gen-tag-table.py languagetags language-subtag-registry
+ *
+ * on files with these headers:
+ *
+ * <meta name="updated_at" content="2019-05-22 06:05 PM" />
+ * File-Date: 2020-05-12
+ */
+
+#ifndef HB_OT_TAG_TABLE_HH
+#define HB_OT_TAG_TABLE_HH
+
+static const LangTag ot_languages[] = {
+  {"aa",	HB_TAG('A','F','R',' ')},	/* Afar */
+  {"aae",	HB_TAG('S','Q','I',' ')},	/* Arbëreshë Albanian -> Albanian */
+  {"aao",	HB_TAG('A','R','A',' ')},	/* Algerian Saharan Arabic -> Arabic */
+  {"aat",	HB_TAG('S','Q','I',' ')},	/* Arvanitika Albanian -> Albanian */
+  {"ab",	HB_TAG('A','B','K',' ')},	/* Abkhazian */
+  {"abh",	HB_TAG('A','R','A',' ')},	/* Tajiki Arabic -> Arabic */
+  {"abq",	HB_TAG('A','B','A',' ')},	/* Abaza */
+  {"abv",	HB_TAG('A','R','A',' ')},	/* Baharna Arabic -> Arabic */
+  {"acf",	HB_TAG('F','A','N',' ')},	/* Saint Lucian Creole French -> French Antillean */
+/*{"ach",	HB_TAG('A','C','H',' ')},*/	/* Acoli -> Acholi */
+  {"acm",	HB_TAG('A','R','A',' ')},	/* Mesopotamian Arabic -> Arabic */
+  {"acq",	HB_TAG('A','R','A',' ')},	/* Ta'izzi-Adeni Arabic -> Arabic */
+/*{"acr",	HB_TAG('A','C','R',' ')},*/	/* Achi */
+  {"acw",	HB_TAG('A','R','A',' ')},	/* Hijazi Arabic -> Arabic */
+  {"acx",	HB_TAG('A','R','A',' ')},	/* Omani Arabic -> Arabic */
+  {"acy",	HB_TAG('A','R','A',' ')},	/* Cypriot Arabic -> Arabic */
+  {"ada",	HB_TAG('D','N','G',' ')},	/* Adangme -> Dangme */
+  {"adf",	HB_TAG('A','R','A',' ')},	/* Dhofari Arabic -> Arabic */
+  {"adp",	HB_TAG('D','Z','N',' ')},	/* Adap (retired code) -> Dzongkha */
+/*{"ady",	HB_TAG('A','D','Y',' ')},*/	/* Adyghe */
+  {"aeb",	HB_TAG('A','R','A',' ')},	/* Tunisian Arabic -> Arabic */
+  {"aec",	HB_TAG('A','R','A',' ')},	/* Saidi Arabic -> Arabic */
+  {"af",	HB_TAG('A','F','K',' ')},	/* Afrikaans */
+  {"afb",	HB_TAG('A','R','A',' ')},	/* Gulf Arabic -> Arabic */
+  {"ahg",	HB_TAG('A','G','W',' ')},	/* Qimant -> Agaw */
+  {"aht",	HB_TAG('A','T','H',' ')},	/* Ahtena -> Athapaskan */
+  {"aii",	HB_TAG('S','W','A',' ')},	/* Assyrian Neo-Aramaic -> Swadaya Aramaic */
+  {"aii",	HB_TAG('S','Y','R',' ')},	/* Assyrian Neo-Aramaic -> Syriac */
+/*{"aio",	HB_TAG('A','I','O',' ')},*/	/* Aiton */
+  {"aiw",	HB_TAG('A','R','I',' ')},	/* Aari */
+  {"ajp",	HB_TAG('A','R','A',' ')},	/* South Levantine Arabic -> Arabic */
+  {"ak",	HB_TAG('A','K','A',' ')},	/* Akan [macrolanguage] */
+  {"ak",	HB_TAG('T','W','I',' ')},	/* Akan [macrolanguage] -> Twi */
+  {"aln",	HB_TAG('S','Q','I',' ')},	/* Gheg Albanian -> Albanian */
+  {"als",	HB_TAG('S','Q','I',' ')},	/* Tosk Albanian -> Albanian */
+/*{"alt",	HB_TAG('A','L','T',' ')},*/	/* Southern Altai -> Altai */
+  {"am",	HB_TAG('A','M','H',' ')},	/* Amharic */
+  {"amf",	HB_TAG('H','B','N',' ')},	/* Hamer-Banna -> Hammer-Banna */
+  {"amw",	HB_TAG('S','Y','R',' ')},	/* Western Neo-Aramaic -> Syriac */
+  {"an",	HB_TAG('A','R','G',' ')},	/* Aragonese */
+/*{"ang",	HB_TAG('A','N','G',' ')},*/	/* Old English (ca. 450-1100) -> Anglo-Saxon */
+  {"apc",	HB_TAG('A','R','A',' ')},	/* North Levantine Arabic -> Arabic */
+  {"apd",	HB_TAG('A','R','A',' ')},	/* Sudanese Arabic -> Arabic */
+  {"apj",	HB_TAG('A','T','H',' ')},	/* Jicarilla Apache -> Athapaskan */
+  {"apk",	HB_TAG('A','T','H',' ')},	/* Kiowa Apache -> Athapaskan */
+  {"apl",	HB_TAG('A','T','H',' ')},	/* Lipan Apache -> Athapaskan */
+  {"apm",	HB_TAG('A','T','H',' ')},	/* Mescalero-Chiricahua Apache -> Athapaskan */
+  {"apw",	HB_TAG('A','T','H',' ')},	/* Western Apache -> Athapaskan */
+  {"ar",	HB_TAG('A','R','A',' ')},	/* Arabic [macrolanguage] */
+  {"arb",	HB_TAG('A','R','A',' ')},	/* Standard Arabic -> Arabic */
+  {"arn",	HB_TAG('M','A','P',' ')},	/* Mapudungun */
+  {"arq",	HB_TAG('A','R','A',' ')},	/* Algerian Arabic -> Arabic */
+  {"ars",	HB_TAG('A','R','A',' ')},	/* Najdi Arabic -> Arabic */
+  {"ary",	HB_TAG('M','O','R',' ')},	/* Moroccan Arabic -> Moroccan */
+  {"arz",	HB_TAG('A','R','A',' ')},	/* Egyptian Arabic -> Arabic */
+  {"as",	HB_TAG('A','S','M',' ')},	/* Assamese */
+/*{"ast",	HB_TAG('A','S','T',' ')},*/	/* Asturian */
+/*{"ath",	HB_TAG('A','T','H',' ')},*/	/* Athapascan [family] -> Athapaskan */
+  {"atj",	HB_TAG('R','C','R',' ')},	/* Atikamekw -> R-Cree */
+  {"atv",	HB_TAG('A','L','T',' ')},	/* Northern Altai -> Altai */
+  {"auz",	HB_TAG('A','R','A',' ')},	/* Uzbeki Arabic -> Arabic */
+  {"av",	HB_TAG('A','V','R',' ')},	/* Avaric -> Avar */
+  {"avl",	HB_TAG('A','R','A',' ')},	/* Eastern Egyptian Bedawi Arabic -> Arabic */
+/*{"awa",	HB_TAG('A','W','A',' ')},*/	/* Awadhi */
+  {"ay",	HB_TAG('A','Y','M',' ')},	/* Aymara [macrolanguage] */
+  {"ayc",	HB_TAG('A','Y','M',' ')},	/* Southern Aymara -> Aymara */
+  {"ayh",	HB_TAG('A','R','A',' ')},	/* Hadrami Arabic -> Arabic */
+  {"ayl",	HB_TAG('A','R','A',' ')},	/* Libyan Arabic -> Arabic */
+  {"ayn",	HB_TAG('A','R','A',' ')},	/* Sanaani Arabic -> Arabic */
+  {"ayp",	HB_TAG('A','R','A',' ')},	/* North Mesopotamian Arabic -> Arabic */
+  {"ayr",	HB_TAG('A','Y','M',' ')},	/* Central Aymara -> Aymara */
+  {"az",	HB_TAG('A','Z','E',' ')},	/* Azerbaijani [macrolanguage] */
+/*{"azb",	HB_TAG('A','Z','B',' ')},*/	/* South Azerbaijani -> Torki */
+  {"azj",	HB_TAG('A','Z','E',' ')},	/* North Azerbaijani -> Azerbaijani */
+  {"ba",	HB_TAG('B','S','H',' ')},	/* Bashkir */
+  {"bad",	HB_TAG('B','A','D','0')},	/* Banda [family] */
+  {"bai",	HB_TAG('B','M','L',' ')},	/* Bamileke [family] */
+  {"bal",	HB_TAG('B','L','I',' ')},	/* Baluchi [macrolanguage] */
+/*{"ban",	HB_TAG('B','A','N',' ')},*/	/* Balinese */
+/*{"bar",	HB_TAG('B','A','R',' ')},*/	/* Bavarian */
+/*{"bbc",	HB_TAG('B','B','C',' ')},*/	/* Batak Toba */
+  {"bbz",	HB_TAG('A','R','A',' ')},	/* Babalia Creole Arabic (retired code) -> Arabic */
+  {"bcc",	HB_TAG('B','L','I',' ')},	/* Southern Balochi -> Baluchi */
+  {"bci",	HB_TAG('B','A','U',' ')},	/* Baoulé -> Baulé */
+  {"bcl",	HB_TAG('B','I','K',' ')},	/* Central Bikol -> Bikol */
+  {"bcq",	HB_TAG('B','C','H',' ')},	/* Bench */
+  {"bcr",	HB_TAG('A','T','H',' ')},	/* Babine -> Athapaskan */
+/*{"bdy",	HB_TAG('B','D','Y',' ')},*/	/* Bandjalang */
+  {"be",	HB_TAG('B','E','L',' ')},	/* Belarusian -> Belarussian */
+  {"bea",	HB_TAG('A','T','H',' ')},	/* Beaver -> Athapaskan */
+  {"beb",	HB_TAG('B','T','I',' ')},	/* Bebele -> Beti */
+/*{"bem",	HB_TAG('B','E','M',' ')},*/	/* Bemba (Zambia) */
+  {"ber",	HB_TAG('B','B','R',' ')},	/* Berber [family] */
+  {"bfq",	HB_TAG('B','A','D',' ')},	/* Badaga */
+  {"bft",	HB_TAG('B','L','T',' ')},	/* Balti */
+  {"bfu",	HB_TAG('L','A','H',' ')},	/* Gahri -> Lahuli */
+  {"bfy",	HB_TAG('B','A','G',' ')},	/* Bagheli -> Baghelkhandi */
+  {"bg",	HB_TAG('B','G','R',' ')},	/* Bulgarian */
+/*{"bgc",	HB_TAG('B','G','C',' ')},*/	/* Haryanvi */
+  {"bgn",	HB_TAG('B','L','I',' ')},	/* Western Balochi -> Baluchi */
+  {"bgp",	HB_TAG('B','L','I',' ')},	/* Eastern Balochi -> Baluchi */
+/*{"bgq",	HB_TAG('B','G','Q',' ')},*/	/* Bagri */
+  {"bgr",	HB_TAG('Q','I','N',' ')},	/* Bawm Chin -> Chin */
+  {"bhb",	HB_TAG('B','H','I',' ')},	/* Bhili */
+/*{"bhi",	HB_TAG('B','H','I',' ')},*/	/* Bhilali -> Bhili */
+  {"bhk",	HB_TAG('B','I','K',' ')},	/* Albay Bicolano (retired code) -> Bikol */
+/*{"bho",	HB_TAG('B','H','O',' ')},*/	/* Bhojpuri */
+  {"bhr",	HB_TAG('M','L','G',' ')},	/* Bara Malagasy -> Malagasy */
+  {"bi",	HB_TAG('B','I','S',' ')},	/* Bislama */
+/*{"bik",	HB_TAG('B','I','K',' ')},*/	/* Bikol [macrolanguage] */
+  {"bin",	HB_TAG('E','D','O',' ')},	/* Edo */
+/*{"bjj",	HB_TAG('B','J','J',' ')},*/	/* Kanauji */
+  {"bjn",	HB_TAG('M','L','Y',' ')},	/* Banjar -> Malay */
+  {"bjq",	HB_TAG('M','L','G',' ')},	/* Southern Betsimisaraka Malagasy (retired code) -> Malagasy */
+  {"bjt",	HB_TAG('B','L','N',' ')},	/* Balanta-Ganja -> Balante */
+  {"bla",	HB_TAG('B','K','F',' ')},	/* Siksika -> Blackfoot */
+  {"ble",	HB_TAG('B','L','N',' ')},	/* Balanta-Kentohe -> Balante */
+/*{"blk",	HB_TAG('B','L','K',' ')},*/	/* Pa’o Karen */
+  {"bln",	HB_TAG('B','I','K',' ')},	/* Southern Catanduanes Bikol -> Bikol */
+  {"bm",	HB_TAG('B','M','B',' ')},	/* Bambara (Bamanankan) */
+  {"bmm",	HB_TAG('M','L','G',' ')},	/* Northern Betsimisaraka Malagasy -> Malagasy */
+  {"bn",	HB_TAG('B','E','N',' ')},	/* Bengali */
+  {"bo",	HB_TAG('T','I','B',' ')},	/* Tibetan */
+/*{"bpy",	HB_TAG('B','P','Y',' ')},*/	/* Bishnupriya -> Bishnupriya Manipuri */
+  {"bqi",	HB_TAG('L','R','C',' ')},	/* Bakhtiari -> Luri */
+  {"br",	HB_TAG('B','R','E',' ')},	/* Breton */
+  {"bra",	HB_TAG('B','R','I',' ')},	/* Braj -> Braj Bhasha */
+/*{"brh",	HB_TAG('B','R','H',' ')},*/	/* Brahui */
+/*{"brx",	HB_TAG('B','R','X',' ')},*/	/* Bodo (India) */
+  {"bs",	HB_TAG('B','O','S',' ')},	/* Bosnian */
+/*{"bsk",	HB_TAG('B','S','K',' ')},*/	/* Burushaski */
+  {"btb",	HB_TAG('B','T','I',' ')},	/* Beti (Cameroon) (retired code) */
+  {"btj",	HB_TAG('M','L','Y',' ')},	/* Bacanese Malay -> Malay */
+  {"bto",	HB_TAG('B','I','K',' ')},	/* Rinconada Bikol -> Bikol */
+/*{"bts",	HB_TAG('B','T','S',' ')},*/	/* Batak Simalungun */
+/*{"bug",	HB_TAG('B','U','G',' ')},*/	/* Buginese -> Bugis */
+  {"bum",	HB_TAG('B','T','I',' ')},	/* Bulu (Cameroon) -> Beti */
+  {"bve",	HB_TAG('M','L','Y',' ')},	/* Berau Malay -> Malay */
+  {"bvu",	HB_TAG('M','L','Y',' ')},	/* Bukit Malay -> Malay */
+  {"bxk",	HB_TAG('L','U','H',' ')},	/* Bukusu -> Luyia */
+  {"bxp",	HB_TAG('B','T','I',' ')},	/* Bebil -> Beti */
+  {"bxr",	HB_TAG('R','B','U',' ')},	/* Russia Buriat -> Russian Buriat */
+  {"byn",	HB_TAG('B','I','L',' ')},	/* Bilin -> Bilen */
+/*{"byv",	HB_TAG('B','Y','V',' ')},*/	/* Medumba */
+  {"bzc",	HB_TAG('M','L','G',' ')},	/* Southern Betsimisaraka Malagasy -> Malagasy */
+  {"ca",	HB_TAG('C','A','T',' ')},	/* Catalan */
+  {"caf",	HB_TAG('C','R','R',' ')},	/* Southern Carrier -> Carrier */
+  {"caf",	HB_TAG('A','T','H',' ')},	/* Southern Carrier -> Athapaskan */
+/*{"cak",	HB_TAG('C','A','K',' ')},*/	/* Kaqchikel */
+/*{"cbk",	HB_TAG('C','B','K',' ')},*/	/* Chavacano -> Zamboanga Chavacano */
+  {"cbl",	HB_TAG('Q','I','N',' ')},	/* Bualkhaw Chin -> Chin */
+  {"cco",	HB_TAG('C','C','H','N')},	/* Comaltepec Chinantec -> Chinantec */
+  {"ccq",	HB_TAG('A','R','K',' ')},	/* Chaungtha (retired code) -> Rakhine */
+  {"cdo",	HB_TAG('Z','H','S',' ')},	/* Min Dong Chinese -> Chinese Simplified */
+  {"ce",	HB_TAG('C','H','E',' ')},	/* Chechen */
+/*{"ceb",	HB_TAG('C','E','B',' ')},*/	/* Cebuano */
+  {"cfm",	HB_TAG('H','A','L',' ')},	/* Halam (Falam Chin) */
+/*{"cgg",	HB_TAG('C','G','G',' ')},*/	/* Chiga */
+  {"ch",	HB_TAG('C','H','A',' ')},	/* Chamorro */
+  {"chj",	HB_TAG('C','C','H','N')},	/* Ojitlán Chinantec -> Chinantec */
+  {"chk",	HB_TAG('C','H','K','0')},	/* Chuukese */
+/*{"cho",	HB_TAG('C','H','O',' ')},*/	/* Choctaw */
+  {"chp",	HB_TAG('C','H','P',' ')},	/* Chipewyan */
+  {"chp",	HB_TAG('S','A','Y',' ')},	/* Chipewyan -> Sayisi */
+  {"chp",	HB_TAG('A','T','H',' ')},	/* Chipewyan -> Athapaskan */
+  {"chq",	HB_TAG('C','C','H','N')},	/* Quiotepec Chinantec -> Chinantec */
+/*{"chr",	HB_TAG('C','H','R',' ')},*/	/* Cherokee */
+/*{"chy",	HB_TAG('C','H','Y',' ')},*/	/* Cheyenne */
+  {"chz",	HB_TAG('C','C','H','N')},	/* Ozumacín Chinantec -> Chinantec */
+  {"ciw",	HB_TAG('O','J','B',' ')},	/* Chippewa -> Ojibway */
+/*{"cja",	HB_TAG('C','J','A',' ')},*/	/* Western Cham */
+/*{"cjm",	HB_TAG('C','J','M',' ')},*/	/* Eastern Cham */
+  {"cjy",	HB_TAG('Z','H','S',' ')},	/* Jinyu Chinese -> Chinese Simplified */
+  {"cka",	HB_TAG('Q','I','N',' ')},	/* Khumi Awa Chin (retired code) -> Chin */
+  {"ckb",	HB_TAG('K','U','R',' ')},	/* Central Kurdish -> Kurdish */
+  {"ckt",	HB_TAG('C','H','K',' ')},	/* Chukot -> Chukchi */
+  {"clc",	HB_TAG('A','T','H',' ')},	/* Chilcotin -> Athapaskan */
+  {"cld",	HB_TAG('S','Y','R',' ')},	/* Chaldean Neo-Aramaic -> Syriac */
+  {"cle",	HB_TAG('C','C','H','N')},	/* Lealao Chinantec -> Chinantec */
+  {"cmn",	HB_TAG('Z','H','S',' ')},	/* Mandarin Chinese -> Chinese Simplified */
+  {"cmr",	HB_TAG('Q','I','N',' ')},	/* Mro-Khimi Chin -> Chin */
+  {"cnb",	HB_TAG('Q','I','N',' ')},	/* Chinbon Chin -> Chin */
+  {"cnh",	HB_TAG('Q','I','N',' ')},	/* Hakha Chin -> Chin */
+  {"cnk",	HB_TAG('Q','I','N',' ')},	/* Khumi Chin -> Chin */
+  {"cnl",	HB_TAG('C','C','H','N')},	/* Lalana Chinantec -> Chinantec */
+  {"cnp",	HB_TAG('Z','H','S',' ')},	/* Northern Ping Chinese -> Chinese Simplified */
+  {"cnt",	HB_TAG('C','C','H','N')},	/* Tepetotutla Chinantec -> Chinantec */
+  {"cnw",	HB_TAG('Q','I','N',' ')},	/* Ngawn Chin -> Chin */
+  {"co",	HB_TAG('C','O','S',' ')},	/* Corsican */
+  {"coa",	HB_TAG('M','L','Y',' ')},	/* Cocos Islands Malay -> Malay */
+/*{"cop",	HB_TAG('C','O','P',' ')},*/	/* Coptic */
+  {"coq",	HB_TAG('A','T','H',' ')},	/* Coquille -> Athapaskan */
+  {"cpa",	HB_TAG('C','C','H','N')},	/* Palantla Chinantec -> Chinantec */
+  {"cpe",	HB_TAG('C','P','P',' ')},	/* English-based creoles and pidgins [family] -> Creoles */
+  {"cpf",	HB_TAG('C','P','P',' ')},	/* French-based creoles and pidgins [family] -> Creoles */
+/*{"cpp",	HB_TAG('C','P','P',' ')},*/	/* Portuguese-based creoles and pidgins [family] -> Creoles */
+  {"cpx",	HB_TAG('Z','H','S',' ')},	/* Pu-Xian Chinese -> Chinese Simplified */
+  {"cqd",	HB_TAG('H','M','N',' ')},	/* Chuanqiandian Cluster Miao -> Hmong */
+  {"cqu",	HB_TAG('Q','U','H',' ')},	/* Chilean Quechua (retired code) -> Quechua (Bolivia) */
+  {"cr",	HB_TAG('C','R','E',' ')},	/* Cree [macrolanguage] */
+  {"cr",	HB_TAG('Y','C','R',' ')},	/* Cree [macrolanguage] -> Y-Cree */
+  {"crh",	HB_TAG('C','R','T',' ')},	/* Crimean Tatar */
+  {"crj",	HB_TAG('E','C','R',' ')},	/* Southern East Cree -> Eastern Cree */
+  {"crk",	HB_TAG('W','C','R',' ')},	/* Plains Cree -> West-Cree */
+  {"crl",	HB_TAG('E','C','R',' ')},	/* Northern East Cree -> Eastern Cree */
+  {"crm",	HB_TAG('M','C','R',' ')},	/* Moose Cree */
+  {"crm",	HB_TAG('L','C','R',' ')},	/* Moose Cree -> L-Cree */
+  {"crp",	HB_TAG('C','P','P',' ')},	/* Creoles and pidgins [family] -> Creoles */
+  {"crx",	HB_TAG('C','R','R',' ')},	/* Carrier */
+  {"crx",	HB_TAG('A','T','H',' ')},	/* Carrier -> Athapaskan */
+  {"cs",	HB_TAG('C','S','Y',' ')},	/* Czech */
+  {"csa",	HB_TAG('C','C','H','N')},	/* Chiltepec Chinantec -> Chinantec */
+/*{"csb",	HB_TAG('C','S','B',' ')},*/	/* Kashubian */
+  {"csh",	HB_TAG('Q','I','N',' ')},	/* Asho Chin -> Chin */
+  {"cso",	HB_TAG('C','C','H','N')},	/* Sochiapam Chinantec -> Chinantec */
+  {"csp",	HB_TAG('Z','H','S',' ')},	/* Southern Ping Chinese -> Chinese Simplified */
+  {"csw",	HB_TAG('N','C','R',' ')},	/* Swampy Cree -> N-Cree */
+  {"csw",	HB_TAG('N','H','C',' ')},	/* Swampy Cree -> Norway House Cree */
+  {"csy",	HB_TAG('Q','I','N',' ')},	/* Siyin Chin -> Chin */
+  {"ctc",	HB_TAG('A','T','H',' ')},	/* Chetco -> Athapaskan */
+  {"ctd",	HB_TAG('Q','I','N',' ')},	/* Tedim Chin -> Chin */
+  {"cte",	HB_TAG('C','C','H','N')},	/* Tepinapa Chinantec -> Chinantec */
+/*{"ctg",	HB_TAG('C','T','G',' ')},*/	/* Chittagonian */
+  {"ctl",	HB_TAG('C','C','H','N')},	/* Tlacoatzintepec Chinantec -> Chinantec */
+  {"cts",	HB_TAG('B','I','K',' ')},	/* Northern Catanduanes Bikol -> Bikol */
+  {"cu",	HB_TAG('C','S','L',' ')},	/* Church Slavonic */
+  {"cuc",	HB_TAG('C','C','H','N')},	/* Usila Chinantec -> Chinantec */
+/*{"cuk",	HB_TAG('C','U','K',' ')},*/	/* San Blas Kuna */
+  {"cv",	HB_TAG('C','H','U',' ')},	/* Chuvash */
+  {"cvn",	HB_TAG('C','C','H','N')},	/* Valle Nacional Chinantec -> Chinantec */
+  {"cwd",	HB_TAG('D','C','R',' ')},	/* Woods Cree */
+  {"cwd",	HB_TAG('T','C','R',' ')},	/* Woods Cree -> TH-Cree */
+  {"cy",	HB_TAG('W','E','L',' ')},	/* Welsh */
+  {"czh",	HB_TAG('Z','H','S',' ')},	/* Huizhou Chinese -> Chinese Simplified */
+  {"czo",	HB_TAG('Z','H','S',' ')},	/* Min Zhong Chinese -> Chinese Simplified */
+  {"czt",	HB_TAG('Q','I','N',' ')},	/* Zotung Chin -> Chin */
+  {"da",	HB_TAG('D','A','N',' ')},	/* Danish */
+  {"dao",	HB_TAG('Q','I','N',' ')},	/* Daai Chin -> Chin */
+  {"dap",	HB_TAG('N','I','S',' ')},	/* Nisi (India) (retired code) */
+/*{"dar",	HB_TAG('D','A','R',' ')},*/	/* Dargwa */
+/*{"dax",	HB_TAG('D','A','X',' ')},*/	/* Dayi */
+  {"de",	HB_TAG('D','E','U',' ')},	/* German */
+  {"den",	HB_TAG('S','L','A',' ')},	/* Slave (Athapascan) [macrolanguage] -> Slavey */
+  {"den",	HB_TAG('A','T','H',' ')},	/* Slave (Athapascan) [macrolanguage] -> Athapaskan */
+/*{"dgo",	HB_TAG('D','G','O',' ')},*/	/* Dogri */
+  {"dgr",	HB_TAG('A','T','H',' ')},	/* Dogrib -> Athapaskan */
+  {"dhd",	HB_TAG('M','A','W',' ')},	/* Dhundari -> Marwari */
+/*{"dhg",	HB_TAG('D','H','G',' ')},*/	/* Dhangu */
+  {"dib",	HB_TAG('D','N','K',' ')},	/* South Central Dinka -> Dinka */
+  {"dik",	HB_TAG('D','N','K',' ')},	/* Southwestern Dinka -> Dinka */
+  {"din",	HB_TAG('D','N','K',' ')},	/* Dinka [macrolanguage] */
+  {"dip",	HB_TAG('D','N','K',' ')},	/* Northeastern Dinka -> Dinka */
+/*{"diq",	HB_TAG('D','I','Q',' ')},*/	/* Dimli */
+  {"diw",	HB_TAG('D','N','K',' ')},	/* Northwestern Dinka -> Dinka */
+  {"dje",	HB_TAG('D','J','R',' ')},	/* Zarma */
+  {"djr",	HB_TAG('D','J','R','0')},	/* Djambarrpuyngu */
+  {"dks",	HB_TAG('D','N','K',' ')},	/* Southeastern Dinka -> Dinka */
+  {"dng",	HB_TAG('D','U','N',' ')},	/* Dungan */
+/*{"dnj",	HB_TAG('D','N','J',' ')},*/	/* Dan */
+  {"doi",	HB_TAG('D','G','R',' ')},	/* Dogri [macrolanguage] */
+  {"drh",	HB_TAG('M','N','G',' ')},	/* Darkhat (retired code) -> Mongolian */
+  {"drw",	HB_TAG('D','R','I',' ')},	/* Darwazi (retired code) -> Dari */
+  {"dsb",	HB_TAG('L','S','B',' ')},	/* Lower Sorbian */
+  {"dty",	HB_TAG('N','E','P',' ')},	/* Dotyali -> Nepali */
+/*{"duj",	HB_TAG('D','U','J',' ')},*/	/* Dhuwal (retired code) */
+  {"dup",	HB_TAG('M','L','Y',' ')},	/* Duano -> Malay */
+  {"dv",	HB_TAG('D','I','V',' ')},	/* Divehi (Dhivehi, Maldivian) */
+  {"dv",	HB_TAG('D','H','V',' ')},	/* Divehi (Dhivehi, Maldivian) (deprecated) */
+  {"dwk",	HB_TAG('K','U','I',' ')},	/* Dawik Kui -> Kui */
+  {"dwu",	HB_TAG('D','U','J',' ')},	/* Dhuwal */
+  {"dwy",	HB_TAG('D','U','J',' ')},	/* Dhuwaya -> Dhuwal */
+  {"dyu",	HB_TAG('J','U','L',' ')},	/* Dyula -> Jula */
+  {"dz",	HB_TAG('D','Z','N',' ')},	/* Dzongkha */
+  {"ee",	HB_TAG('E','W','E',' ')},	/* Ewe */
+/*{"efi",	HB_TAG('E','F','I',' ')},*/	/* Efik */
+  {"ekk",	HB_TAG('E','T','I',' ')},	/* Standard Estonian -> Estonian */
+  {"el",	HB_TAG('E','L','L',' ')},	/* Modern Greek (1453-) -> Greek */
+  {"emk",	HB_TAG('E','M','K',' ')},	/* Eastern Maninkakan */
+  {"emk",	HB_TAG('M','N','K',' ')},	/* Eastern Maninkakan -> Maninka */
+  {"en",	HB_TAG('E','N','G',' ')},	/* English */
+  {"enb",	HB_TAG('K','A','L',' ')},	/* Markweeta -> Kalenjin */
+  {"enf",	HB_TAG('F','N','E',' ')},	/* Forest Enets -> Forest Nenets */
+  {"enh",	HB_TAG('T','N','E',' ')},	/* Tundra Enets -> Tundra Nenets */
+  {"eo",	HB_TAG('N','T','O',' ')},	/* Esperanto */
+  {"es",	HB_TAG('E','S','P',' ')},	/* Spanish */
+  {"esg",	HB_TAG('G','O','N',' ')},	/* Aheri Gondi -> Gondi */
+  {"esi",	HB_TAG('I','P','K',' ')},	/* North Alaskan Inupiatun -> Inupiat */
+  {"esk",	HB_TAG('I','P','K',' ')},	/* Northwest Alaska Inupiatun -> Inupiat */
+/*{"esu",	HB_TAG('E','S','U',' ')},*/	/* Central Yupik */
+  {"et",	HB_TAG('E','T','I',' ')},	/* Estonian [macrolanguage] */
+  {"eto",	HB_TAG('B','T','I',' ')},	/* Eton (Cameroon) -> Beti */
+  {"eu",	HB_TAG('E','U','Q',' ')},	/* Basque */
+  {"eve",	HB_TAG('E','V','N',' ')},	/* Even */
+  {"evn",	HB_TAG('E','V','K',' ')},	/* Evenki */
+  {"ewo",	HB_TAG('B','T','I',' ')},	/* Ewondo -> Beti */
+  {"eyo",	HB_TAG('K','A','L',' ')},	/* Keiyo -> Kalenjin */
+  {"fa",	HB_TAG('F','A','R',' ')},	/* Persian [macrolanguage] */
+  {"fan",	HB_TAG('F','A','N','0')},	/* Fang (Equatorial Guinea) */
+/*{"fat",	HB_TAG('F','A','T',' ')},*/	/* Fanti */
+  {"fbl",	HB_TAG('B','I','K',' ')},	/* West Albay Bikol -> Bikol */
+  {"ff",	HB_TAG('F','U','L',' ')},	/* Fulah [macrolanguage] */
+  {"ffm",	HB_TAG('F','U','L',' ')},	/* Maasina Fulfulde -> Fulah */
+  {"fi",	HB_TAG('F','I','N',' ')},	/* Finnish */
+  {"fil",	HB_TAG('P','I','L',' ')},	/* Filipino */
+  {"fj",	HB_TAG('F','J','I',' ')},	/* Fijian */
+  {"flm",	HB_TAG('H','A','L',' ')},	/* Halam (Falam Chin) (retired code) */
+  {"flm",	HB_TAG('Q','I','N',' ')},	/* Falam Chin (retired code) -> Chin */
+/*{"fmp",	HB_TAG('F','M','P',' ')},*/	/* Fe’fe’ */
+  {"fo",	HB_TAG('F','O','S',' ')},	/* Faroese */
+/*{"fon",	HB_TAG('F','O','N',' ')},*/	/* Fon */
+  {"fr",	HB_TAG('F','R','A',' ')},	/* French */
+/*{"frc",	HB_TAG('F','R','C',' ')},*/	/* Cajun French */
+/*{"frp",	HB_TAG('F','R','P',' ')},*/	/* Arpitan */
+  {"fub",	HB_TAG('F','U','L',' ')},	/* Adamawa Fulfulde -> Fulah */
+  {"fuc",	HB_TAG('F','U','L',' ')},	/* Pulaar -> Fulah */
+  {"fue",	HB_TAG('F','U','L',' ')},	/* Borgu Fulfulde -> Fulah */
+  {"fuf",	HB_TAG('F','T','A',' ')},	/* Pular -> Futa */
+  {"fuh",	HB_TAG('F','U','L',' ')},	/* Western Niger Fulfulde -> Fulah */
+  {"fui",	HB_TAG('F','U','L',' ')},	/* Bagirmi Fulfulde -> Fulah */
+  {"fuq",	HB_TAG('F','U','L',' ')},	/* Central-Eastern Niger Fulfulde -> Fulah */
+  {"fur",	HB_TAG('F','R','L',' ')},	/* Friulian */
+/*{"fuv",	HB_TAG('F','U','V',' ')},*/	/* Nigerian Fulfulde */
+  {"fy",	HB_TAG('F','R','I',' ')},	/* Western Frisian -> Frisian */
+  {"ga",	HB_TAG('I','R','I',' ')},	/* Irish */
+  {"gaa",	HB_TAG('G','A','D',' ')},	/* Ga */
+/*{"gag",	HB_TAG('G','A','G',' ')},*/	/* Gagauz */
+  {"gan",	HB_TAG('Z','H','S',' ')},	/* Gan Chinese -> Chinese Simplified */
+  {"gax",	HB_TAG('O','R','O',' ')},	/* Borana-Arsi-Guji Oromo -> Oromo */
+  {"gaz",	HB_TAG('O','R','O',' ')},	/* West Central Oromo -> Oromo */
+  {"gbm",	HB_TAG('G','A','W',' ')},	/* Garhwali */
+  {"gce",	HB_TAG('A','T','H',' ')},	/* Galice -> Athapaskan */
+  {"gd",	HB_TAG('G','A','E',' ')},	/* Scottish Gaelic (Gaelic) */
+  {"gda",	HB_TAG('R','A','J',' ')},	/* Gade Lohar -> Rajasthani */
+/*{"gez",	HB_TAG('G','E','Z',' ')},*/	/* Geez */
+  {"ggo",	HB_TAG('G','O','N',' ')},	/* Southern Gondi (retired code) -> Gondi */
+/*{"gih",	HB_TAG('G','I','H',' ')},*/	/* Githabul */
+  {"gil",	HB_TAG('G','I','L','0')},	/* Kiribati (Gilbertese) */
+  {"gju",	HB_TAG('R','A','J',' ')},	/* Gujari -> Rajasthani */
+/*{"gkp",	HB_TAG('G','K','P',' ')},*/	/* Guinea Kpelle -> Kpelle (Guinea) */
+  {"gl",	HB_TAG('G','A','L',' ')},	/* Galician */
+  {"gld",	HB_TAG('N','A','N',' ')},	/* Nanai */
+/*{"glk",	HB_TAG('G','L','K',' ')},*/	/* Gilaki */
+  {"gn",	HB_TAG('G','U','A',' ')},	/* Guarani [macrolanguage] */
+/*{"gnn",	HB_TAG('G','N','N',' ')},*/	/* Gumatj */
+  {"gno",	HB_TAG('G','O','N',' ')},	/* Northern Gondi -> Gondi */
+  {"gnw",	HB_TAG('G','U','A',' ')},	/* Western Bolivian Guaraní -> Guarani */
+/*{"gog",	HB_TAG('G','O','G',' ')},*/	/* Gogo */
+  {"gom",	HB_TAG('K','O','K',' ')},	/* Goan Konkani -> Konkani */
+/*{"gon",	HB_TAG('G','O','N',' ')},*/	/* Gondi [macrolanguage] */
+  {"grt",	HB_TAG('G','R','O',' ')},	/* Garo */
+  {"gru",	HB_TAG('S','O','G',' ')},	/* Kistane -> Sodo Gurage */
+  {"gsw",	HB_TAG('A','L','S',' ')},	/* Alsatian */
+  {"gu",	HB_TAG('G','U','J',' ')},	/* Gujarati */
+/*{"guc",	HB_TAG('G','U','C',' ')},*/	/* Wayuu */
+/*{"guf",	HB_TAG('G','U','F',' ')},*/	/* Gupapuyngu */
+  {"gug",	HB_TAG('G','U','A',' ')},	/* Paraguayan Guaraní -> Guarani */
+  {"gui",	HB_TAG('G','U','A',' ')},	/* Eastern Bolivian Guaraní -> Guarani */
+  {"guk",	HB_TAG('G','M','Z',' ')},	/* Gumuz */
+  {"guk",	HB_TAG('G','U','K',' ')},	/* Gumuz (SIL fonts) */
+  {"gun",	HB_TAG('G','U','A',' ')},	/* Mbyá Guaraní -> Guarani */
+/*{"guz",	HB_TAG('G','U','Z',' ')},*/	/* Gusii */
+  {"gv",	HB_TAG('M','N','X',' ')},	/* Manx */
+  {"gwi",	HB_TAG('A','T','H',' ')},	/* Gwichʼin -> Athapaskan */
+  {"ha",	HB_TAG('H','A','U',' ')},	/* Hausa */
+  {"haa",	HB_TAG('A','T','H',' ')},	/* Han -> Athapaskan */
+  {"hae",	HB_TAG('O','R','O',' ')},	/* Eastern Oromo -> Oromo */
+  {"hak",	HB_TAG('Z','H','S',' ')},	/* Hakka Chinese -> Chinese Simplified */
+  {"har",	HB_TAG('H','R','I',' ')},	/* Harari */
+/*{"haw",	HB_TAG('H','A','W',' ')},*/	/* Hawaiian */
+/*{"hay",	HB_TAG('H','A','Y',' ')},*/	/* Haya */
+/*{"haz",	HB_TAG('H','A','Z',' ')},*/	/* Hazaragi */
+  {"he",	HB_TAG('I','W','R',' ')},	/* Hebrew */
+  {"hea",	HB_TAG('H','M','N',' ')},	/* Northern Qiandong Miao -> Hmong */
+  {"hi",	HB_TAG('H','I','N',' ')},	/* Hindi */
+/*{"hil",	HB_TAG('H','I','L',' ')},*/	/* Hiligaynon */
+  {"hji",	HB_TAG('M','L','Y',' ')},	/* Haji -> Malay */
+  {"hlt",	HB_TAG('Q','I','N',' ')},	/* Matu Chin -> Chin */
+  {"hma",	HB_TAG('H','M','N',' ')},	/* Southern Mashan Hmong -> Hmong */
+  {"hmc",	HB_TAG('H','M','N',' ')},	/* Central Huishui Hmong -> Hmong */
+  {"hmd",	HB_TAG('H','M','N',' ')},	/* Large Flowery Miao -> Hmong */
+  {"hme",	HB_TAG('H','M','N',' ')},	/* Eastern Huishui Hmong -> Hmong */
+  {"hmg",	HB_TAG('H','M','N',' ')},	/* Southwestern Guiyang Hmong -> Hmong */
+  {"hmh",	HB_TAG('H','M','N',' ')},	/* Southwestern Huishui Hmong -> Hmong */
+  {"hmi",	HB_TAG('H','M','N',' ')},	/* Northern Huishui Hmong -> Hmong */
+  {"hmj",	HB_TAG('H','M','N',' ')},	/* Ge -> Hmong */
+  {"hml",	HB_TAG('H','M','N',' ')},	/* Luopohe Hmong -> Hmong */
+  {"hmm",	HB_TAG('H','M','N',' ')},	/* Central Mashan Hmong -> Hmong */
+/*{"hmn",	HB_TAG('H','M','N',' ')},*/	/* Hmong [macrolanguage] */
+  {"hmp",	HB_TAG('H','M','N',' ')},	/* Northern Mashan Hmong -> Hmong */
+  {"hmq",	HB_TAG('H','M','N',' ')},	/* Eastern Qiandong Miao -> Hmong */
+  {"hms",	HB_TAG('H','M','N',' ')},	/* Southern Qiandong Miao -> Hmong */
+  {"hmw",	HB_TAG('H','M','N',' ')},	/* Western Mashan Hmong -> Hmong */
+  {"hmy",	HB_TAG('H','M','N',' ')},	/* Southern Guiyang Hmong -> Hmong */
+  {"hmz",	HB_TAG('H','M','N',' ')},	/* Hmong Shua -> Hmong */
+/*{"hnd",	HB_TAG('H','N','D',' ')},*/	/* Southern Hindko -> Hindko */
+  {"hne",	HB_TAG('C','H','H',' ')},	/* Chhattisgarhi -> Chattisgarhi */
+  {"hnj",	HB_TAG('H','M','N',' ')},	/* Hmong Njua -> Hmong */
+  {"hno",	HB_TAG('H','N','D',' ')},	/* Northern Hindko -> Hindko */
+  {"ho",	HB_TAG('H','M','O',' ')},	/* Hiri Motu */
+  {"hoc",	HB_TAG('H','O',' ',' ')},	/* Ho */
+  {"hoi",	HB_TAG('A','T','H',' ')},	/* Holikachuk -> Athapaskan */
+  {"hoj",	HB_TAG('H','A','R',' ')},	/* Hadothi -> Harauti */
+  {"hr",	HB_TAG('H','R','V',' ')},	/* Croatian */
+  {"hrm",	HB_TAG('H','M','N',' ')},	/* Horned Miao -> Hmong */
+  {"hsb",	HB_TAG('U','S','B',' ')},	/* Upper Sorbian */
+  {"hsn",	HB_TAG('Z','H','S',' ')},	/* Xiang Chinese -> Chinese Simplified */
+  {"ht",	HB_TAG('H','A','I',' ')},	/* Haitian (Haitian Creole) */
+  {"hu",	HB_TAG('H','U','N',' ')},	/* Hungarian */
+  {"huj",	HB_TAG('H','M','N',' ')},	/* Northern Guiyang Hmong -> Hmong */
+  {"hup",	HB_TAG('A','T','H',' ')},	/* Hupa -> Athapaskan */
+  {"hy",	HB_TAG('H','Y','E','0')},	/* Armenian -> Armenian East */
+  {"hy",	HB_TAG('H','Y','E',' ')},	/* Armenian */
+  {"hyw",	HB_TAG('H','Y','E',' ')},	/* Western Armenian -> Armenian */
+  {"hz",	HB_TAG('H','E','R',' ')},	/* Herero */
+  {"ia",	HB_TAG('I','N','A',' ')},	/* Interlingua (International Auxiliary Language Association) */
+/*{"iba",	HB_TAG('I','B','A',' ')},*/	/* Iban */
+/*{"ibb",	HB_TAG('I','B','B',' ')},*/	/* Ibibio */
+  {"id",	HB_TAG('I','N','D',' ')},	/* Indonesian */
+  {"ida",	HB_TAG('L','U','H',' ')},	/* Idakho-Isukha-Tiriki -> Luyia */
+  {"ie",	HB_TAG('I','L','E',' ')},	/* Interlingue */
+  {"ig",	HB_TAG('I','B','O',' ')},	/* Igbo */
+  {"igb",	HB_TAG('E','B','I',' ')},	/* Ebira */
+  {"ii",	HB_TAG('Y','I','M',' ')},	/* Sichuan Yi -> Yi Modern */
+  {"ijc",	HB_TAG('I','J','O',' ')},	/* Izon -> Ijo */
+/*{"ijo",	HB_TAG('I','J','O',' ')},*/	/* Ijo [family] */
+  {"ik",	HB_TAG('I','P','K',' ')},	/* Inupiaq [macrolanguage] -> Inupiat */
+  {"ike",	HB_TAG('I','N','U',' ')},	/* Eastern Canadian Inuktitut -> Inuktitut */
+  {"ikt",	HB_TAG('I','N','U',' ')},	/* Inuinnaqtun -> Inuktitut */
+/*{"ilo",	HB_TAG('I','L','O',' ')},*/	/* Iloko -> Ilokano */
+  {"in",	HB_TAG('I','N','D',' ')},	/* Indonesian (retired code) */
+  {"ing",	HB_TAG('A','T','H',' ')},	/* Degexit'an -> Athapaskan */
+  {"inh",	HB_TAG('I','N','G',' ')},	/* Ingush */
+  {"io",	HB_TAG('I','D','O',' ')},	/* Ido */
+  {"is",	HB_TAG('I','S','L',' ')},	/* Icelandic */
+  {"it",	HB_TAG('I','T','A',' ')},	/* Italian */
+  {"iu",	HB_TAG('I','N','U',' ')},	/* Inuktitut [macrolanguage] */
+  {"iw",	HB_TAG('I','W','R',' ')},	/* Hebrew (retired code) */
+  {"ja",	HB_TAG('J','A','N',' ')},	/* Japanese */
+  {"jak",	HB_TAG('M','L','Y',' ')},	/* Jakun -> Malay */
+/*{"jam",	HB_TAG('J','A','M',' ')},*/	/* Jamaican Creole English -> Jamaican Creole */
+  {"jax",	HB_TAG('M','L','Y',' ')},	/* Jambi Malay -> Malay */
+/*{"jbo",	HB_TAG('J','B','O',' ')},*/	/* Lojban */
+/*{"jct",	HB_TAG('J','C','T',' ')},*/	/* Krymchak */
+  {"ji",	HB_TAG('J','I','I',' ')},	/* Yiddish (retired code) */
+  {"jv",	HB_TAG('J','A','V',' ')},	/* Javanese */
+  {"jw",	HB_TAG('J','A','V',' ')},	/* Javanese (retired code) */
+  {"ka",	HB_TAG('K','A','T',' ')},	/* Georgian */
+  {"kaa",	HB_TAG('K','R','K',' ')},	/* Karakalpak */
+  {"kab",	HB_TAG('K','A','B','0')},	/* Kabyle */
+  {"kam",	HB_TAG('K','M','B',' ')},	/* Kamba (Kenya) */
+  {"kar",	HB_TAG('K','R','N',' ')},	/* Karen [family] */
+  {"kbd",	HB_TAG('K','A','B',' ')},	/* Kabardian */
+  {"kby",	HB_TAG('K','N','R',' ')},	/* Manga Kanuri -> Kanuri */
+  {"kca",	HB_TAG('K','H','K',' ')},	/* Khanty -> Khanty-Kazim */
+  {"kca",	HB_TAG('K','H','S',' ')},	/* Khanty -> Khanty-Shurishkar */
+  {"kca",	HB_TAG('K','H','V',' ')},	/* Khanty -> Khanty-Vakhi */
+/*{"kde",	HB_TAG('K','D','E',' ')},*/	/* Makonde */
+  {"kdr",	HB_TAG('K','R','M',' ')},	/* Karaim */
+  {"kdt",	HB_TAG('K','U','Y',' ')},	/* Kuy */
+/*{"kea",	HB_TAG('K','E','A',' ')},*/	/* Kabuverdianu (Crioulo) */
+/*{"kek",	HB_TAG('K','E','K',' ')},*/	/* Kekchi */
+  {"kex",	HB_TAG('K','K','N',' ')},	/* Kukna -> Kokni */
+  {"kfa",	HB_TAG('K','O','D',' ')},	/* Kodava -> Kodagu */
+  {"kfr",	HB_TAG('K','A','C',' ')},	/* Kachhi -> Kachchi */
+  {"kfx",	HB_TAG('K','U','L',' ')},	/* Kullu Pahari -> Kulvi */
+  {"kfy",	HB_TAG('K','M','N',' ')},	/* Kumaoni */
+  {"kg",	HB_TAG('K','O','N','0')},	/* Kongo [macrolanguage] */
+  {"kha",	HB_TAG('K','S','I',' ')},	/* Khasi */
+  {"khb",	HB_TAG('X','B','D',' ')},	/* Lü */
+  {"khk",	HB_TAG('M','N','G',' ')},	/* Halh Mongolian -> Mongolian */
+  {"kht",	HB_TAG('K','H','N',' ')},	/* Khamti -> Khamti Shan (Microsoft fonts) */
+  {"kht",	HB_TAG('K','H','T',' ')},	/* Khamti -> Khamti Shan (OpenType spec and SIL fonts) */
+/*{"khw",	HB_TAG('K','H','W',' ')},*/	/* Khowar */
+  {"ki",	HB_TAG('K','I','K',' ')},	/* Kikuyu (Gikuyu) */
+/*{"kiu",	HB_TAG('K','I','U',' ')},*/	/* Kirmanjki */
+  {"kj",	HB_TAG('K','U','A',' ')},	/* Kuanyama */
+/*{"kjd",	HB_TAG('K','J','D',' ')},*/	/* Southern Kiwai */
+  {"kjh",	HB_TAG('K','H','A',' ')},	/* Khakas -> Khakass */
+/*{"kjp",	HB_TAG('K','J','P',' ')},*/	/* Pwo Eastern Karen -> Eastern Pwo Karen */
+/*{"kjz",	HB_TAG('K','J','Z',' ')},*/	/* Bumthangkha */
+  {"kk",	HB_TAG('K','A','Z',' ')},	/* Kazakh */
+  {"kkz",	HB_TAG('A','T','H',' ')},	/* Kaska -> Athapaskan */
+  {"kl",	HB_TAG('G','R','N',' ')},	/* Greenlandic */
+  {"kln",	HB_TAG('K','A','L',' ')},	/* Kalenjin [macrolanguage] */
+  {"km",	HB_TAG('K','H','M',' ')},	/* Khmer */
+  {"kmb",	HB_TAG('M','B','N',' ')},	/* Kimbundu -> Mbundu */
+  {"kmr",	HB_TAG('K','U','R',' ')},	/* Northern Kurdish -> Kurdish */
+  {"kmw",	HB_TAG('K','M','O',' ')},	/* Komo (Democratic Republic of Congo) */
+/*{"kmz",	HB_TAG('K','M','Z',' ')},*/	/* Khorasani Turkish -> Khorasani Turkic */
+  {"kn",	HB_TAG('K','A','N',' ')},	/* Kannada */
+  {"knc",	HB_TAG('K','N','R',' ')},	/* Central Kanuri -> Kanuri */
+  {"kng",	HB_TAG('K','O','N','0')},	/* Koongo -> Kongo */
+  {"knn",	HB_TAG('K','O','K',' ')},	/* Konkani */
+  {"ko",	HB_TAG('K','O','R',' ')},	/* Korean */
+  {"koi",	HB_TAG('K','O','P',' ')},	/* Komi-Permyak */
+/*{"kok",	HB_TAG('K','O','K',' ')},*/	/* Konkani [macrolanguage] */
+/*{"kos",	HB_TAG('K','O','S',' ')},*/	/* Kosraean */
+  {"koy",	HB_TAG('A','T','H',' ')},	/* Koyukon -> Athapaskan */
+  {"kpe",	HB_TAG('K','P','L',' ')},	/* Kpelle [macrolanguage] */
+  {"kpv",	HB_TAG('K','O','Z',' ')},	/* Komi-Zyrian */
+  {"kpy",	HB_TAG('K','Y','K',' ')},	/* Koryak */
+  {"kqs",	HB_TAG('K','I','S',' ')},	/* Northern Kissi -> Kisii */
+  {"kqy",	HB_TAG('K','R','T',' ')},	/* Koorete */
+  {"kr",	HB_TAG('K','N','R',' ')},	/* Kanuri [macrolanguage] */
+  {"krc",	HB_TAG('K','A','R',' ')},	/* Karachay-Balkar -> Karachay */
+  {"krc",	HB_TAG('B','A','L',' ')},	/* Karachay-Balkar -> Balkar */
+/*{"kri",	HB_TAG('K','R','I',' ')},*/	/* Krio */
+/*{"krl",	HB_TAG('K','R','L',' ')},*/	/* Karelian */
+  {"krt",	HB_TAG('K','N','R',' ')},	/* Tumari Kanuri -> Kanuri */
+  {"kru",	HB_TAG('K','U','U',' ')},	/* Kurukh */
+  {"ks",	HB_TAG('K','S','H',' ')},	/* Kashmiri */
+  {"ksh",	HB_TAG('K','S','H','0')},	/* Kölsch -> Ripuarian */
+  {"kss",	HB_TAG('K','I','S',' ')},	/* Southern Kisi -> Kisii */
+/*{"ksw",	HB_TAG('K','S','W',' ')},*/	/* S’gaw Karen */
+  {"ktb",	HB_TAG('K','E','B',' ')},	/* Kambaata -> Kebena */
+  {"ktu",	HB_TAG('K','O','N',' ')},	/* Kituba (Democratic Republic of Congo) -> Kikongo */
+  {"ktw",	HB_TAG('A','T','H',' ')},	/* Kato -> Athapaskan */
+  {"ku",	HB_TAG('K','U','R',' ')},	/* Kurdish [macrolanguage] */
+/*{"kum",	HB_TAG('K','U','M',' ')},*/	/* Kumyk */
+  {"kuu",	HB_TAG('A','T','H',' ')},	/* Upper Kuskokwim -> Athapaskan */
+  {"kv",	HB_TAG('K','O','M',' ')},	/* Komi [macrolanguage] */
+  {"kvb",	HB_TAG('M','L','Y',' ')},	/* Kubu -> Malay */
+  {"kvr",	HB_TAG('M','L','Y',' ')},	/* Kerinci -> Malay */
+  {"kw",	HB_TAG('C','O','R',' ')},	/* Cornish */
+  {"kwy",	HB_TAG('K','O','N','0')},	/* San Salvador Kongo -> Kongo */
+  {"kxc",	HB_TAG('K','M','S',' ')},	/* Konso -> Komso */
+  {"kxd",	HB_TAG('M','L','Y',' ')},	/* Brunei -> Malay */
+  {"kxl",	HB_TAG('K','U','U',' ')},	/* Nepali Kurux (retired code) -> Kurukh */
+  {"kxu",	HB_TAG('K','U','I',' ')},	/* Kui (India) (retired code) */
+  {"ky",	HB_TAG('K','I','R',' ')},	/* Kirghiz (Kyrgyz) */
+/*{"kyu",	HB_TAG('K','Y','U',' ')},*/	/* Western Kayah */
+  {"la",	HB_TAG('L','A','T',' ')},	/* Latin */
+  {"lad",	HB_TAG('J','U','D',' ')},	/* Ladino */
+  {"lb",	HB_TAG('L','T','Z',' ')},	/* Luxembourgish */
+  {"lbe",	HB_TAG('L','A','K',' ')},	/* Lak */
+  {"lbj",	HB_TAG('L','D','K',' ')},	/* Ladakhi */
+  {"lbl",	HB_TAG('B','I','K',' ')},	/* Libon Bikol -> Bikol */
+  {"lce",	HB_TAG('M','L','Y',' ')},	/* Loncong -> Malay */
+  {"lcf",	HB_TAG('M','L','Y',' ')},	/* Lubu -> Malay */
+  {"ldi",	HB_TAG('K','O','N','0')},	/* Laari -> Kongo */
+/*{"lez",	HB_TAG('L','E','Z',' ')},*/	/* Lezghian -> Lezgi */
+  {"lg",	HB_TAG('L','U','G',' ')},	/* Ganda */
+  {"li",	HB_TAG('L','I','M',' ')},	/* Limburgish */
+  {"lif",	HB_TAG('L','M','B',' ')},	/* Limbu */
+/*{"lij",	HB_TAG('L','I','J',' ')},*/	/* Ligurian */
+/*{"lis",	HB_TAG('L','I','S',' ')},*/	/* Lisu */
+  {"liw",	HB_TAG('M','L','Y',' ')},	/* Col -> Malay */
+/*{"ljp",	HB_TAG('L','J','P',' ')},*/	/* Lampung Api -> Lampung */
+  {"lkb",	HB_TAG('L','U','H',' ')},	/* Kabras -> Luyia */
+/*{"lki",	HB_TAG('L','K','I',' ')},*/	/* Laki */
+  {"lko",	HB_TAG('L','U','H',' ')},	/* Khayo -> Luyia */
+  {"lks",	HB_TAG('L','U','H',' ')},	/* Kisa -> Luyia */
+  {"lld",	HB_TAG('L','A','D',' ')},	/* Ladin */
+  {"lmn",	HB_TAG('L','A','M',' ')},	/* Lambadi -> Lambani */
+/*{"lmo",	HB_TAG('L','M','O',' ')},*/	/* Lombard */
+  {"ln",	HB_TAG('L','I','N',' ')},	/* Lingala */
+  {"lo",	HB_TAG('L','A','O',' ')},	/* Lao */
+/*{"lom",	HB_TAG('L','O','M',' ')},*/	/* Loma (Liberia) */
+/*{"lrc",	HB_TAG('L','R','C',' ')},*/	/* Northern Luri -> Luri */
+  {"lri",	HB_TAG('L','U','H',' ')},	/* Marachi -> Luyia */
+  {"lrm",	HB_TAG('L','U','H',' ')},	/* Marama -> Luyia */
+  {"lsm",	HB_TAG('L','U','H',' ')},	/* Saamia -> Luyia */
+  {"lt",	HB_TAG('L','T','H',' ')},	/* Lithuanian */
+  {"ltg",	HB_TAG('L','V','I',' ')},	/* Latgalian -> Latvian */
+  {"lto",	HB_TAG('L','U','H',' ')},	/* Tsotso -> Luyia */
+  {"lts",	HB_TAG('L','U','H',' ')},	/* Tachoni -> Luyia */
+  {"lu",	HB_TAG('L','U','B',' ')},	/* Luba-Katanga */
+/*{"lua",	HB_TAG('L','U','A',' ')},*/	/* Luba-Lulua */
+/*{"luo",	HB_TAG('L','U','O',' ')},*/	/* Luo (Kenya and Tanzania) */
+  {"lus",	HB_TAG('M','I','Z',' ')},	/* Lushai -> Mizo */
+  {"luy",	HB_TAG('L','U','H',' ')},	/* Luyia [macrolanguage] */
+  {"luz",	HB_TAG('L','R','C',' ')},	/* Southern Luri -> Luri */
+  {"lv",	HB_TAG('L','V','I',' ')},	/* Latvian [macrolanguage] */
+  {"lvs",	HB_TAG('L','V','I',' ')},	/* Standard Latvian -> Latvian */
+  {"lwg",	HB_TAG('L','U','H',' ')},	/* Wanga -> Luyia */
+  {"lzh",	HB_TAG('Z','H','T',' ')},	/* Literary Chinese -> Chinese Traditional */
+  {"lzz",	HB_TAG('L','A','Z',' ')},	/* Laz */
+/*{"mad",	HB_TAG('M','A','D',' ')},*/	/* Madurese -> Madura */
+/*{"mag",	HB_TAG('M','A','G',' ')},*/	/* Magahi */
+  {"mai",	HB_TAG('M','T','H',' ')},	/* Maithili */
+  {"mak",	HB_TAG('M','K','R',' ')},	/* Makasar */
+/*{"mam",	HB_TAG('M','A','M',' ')},*/	/* Mam */
+  {"man",	HB_TAG('M','N','K',' ')},	/* Mandingo [macrolanguage] -> Maninka */
+  {"max",	HB_TAG('M','L','Y',' ')},	/* North Moluccan Malay -> Malay */
+/*{"mbo",	HB_TAG('M','B','O',' ')},*/	/* Mbo (Cameroon) */
+  {"mct",	HB_TAG('B','T','I',' ')},	/* Mengisa -> Beti */
+  {"mdf",	HB_TAG('M','O','K',' ')},	/* Moksha */
+/*{"mdr",	HB_TAG('M','D','R',' ')},*/	/* Mandar */
+  {"mdy",	HB_TAG('M','L','E',' ')},	/* Male (Ethiopia) */
+  {"men",	HB_TAG('M','D','E',' ')},	/* Mende (Sierra Leone) */
+  {"meo",	HB_TAG('M','L','Y',' ')},	/* Kedah Malay -> Malay */
+/*{"mer",	HB_TAG('M','E','R',' ')},*/	/* Meru */
+/*{"mfa",	HB_TAG('M','F','A',' ')},*/	/* Pattani Malay */
+  {"mfb",	HB_TAG('M','L','Y',' ')},	/* Bangka -> Malay */
+/*{"mfe",	HB_TAG('M','F','E',' ')},*/	/* Morisyen */
+  {"mg",	HB_TAG('M','L','G',' ')},	/* Malagasy [macrolanguage] */
+  {"mh",	HB_TAG('M','A','H',' ')},	/* Marshallese */
+  {"mhr",	HB_TAG('L','M','A',' ')},	/* Eastern Mari -> Low Mari */
+  {"mhv",	HB_TAG('A','R','K',' ')},	/* Arakanese (retired code) -> Rakhine */
+  {"mi",	HB_TAG('M','R','I',' ')},	/* Maori */
+/*{"min",	HB_TAG('M','I','N',' ')},*/	/* Minangkabau */
+  {"mk",	HB_TAG('M','K','D',' ')},	/* Macedonian */
+  {"mku",	HB_TAG('M','N','K',' ')},	/* Konyanka Maninka -> Maninka */
+/*{"mkw",	HB_TAG('M','K','W',' ')},*/	/* Kituba (Congo) */
+  {"ml",	HB_TAG('M','A','L',' ')},	/* Malayalam -> Malayalam Traditional */
+  {"ml",	HB_TAG('M','L','R',' ')},	/* Malayalam -> Malayalam Reformed */
+  {"mlq",	HB_TAG('M','L','N',' ')},	/* Western Maninkakan -> Malinke */
+  {"mlq",	HB_TAG('M','N','K',' ')},	/* Western Maninkakan -> Maninka */
+  {"mmr",	HB_TAG('H','M','N',' ')},	/* Western Xiangxi Miao -> Hmong */
+  {"mn",	HB_TAG('M','N','G',' ')},	/* Mongolian [macrolanguage] */
+  {"mnc",	HB_TAG('M','C','H',' ')},	/* Manchu */
+/*{"mni",	HB_TAG('M','N','I',' ')},*/	/* Manipuri */
+  {"mnk",	HB_TAG('M','N','D',' ')},	/* Mandinka */
+  {"mnk",	HB_TAG('M','N','K',' ')},	/* Mandinka -> Maninka */
+  {"mnp",	HB_TAG('Z','H','S',' ')},	/* Min Bei Chinese -> Chinese Simplified */
+  {"mns",	HB_TAG('M','A','N',' ')},	/* Mansi */
+  {"mnw",	HB_TAG('M','O','N',' ')},	/* Mon */
+  {"mo",	HB_TAG('M','O','L',' ')},	/* Moldavian (retired code) */
+/*{"moh",	HB_TAG('M','O','H',' ')},*/	/* Mohawk */
+/*{"mos",	HB_TAG('M','O','S',' ')},*/	/* Mossi */
+  {"mpe",	HB_TAG('M','A','J',' ')},	/* Majang */
+  {"mqg",	HB_TAG('M','L','Y',' ')},	/* Kota Bangun Kutai Malay -> Malay */
+  {"mr",	HB_TAG('M','A','R',' ')},	/* Marathi */
+  {"mrh",	HB_TAG('Q','I','N',' ')},	/* Mara Chin -> Chin */
+  {"mrj",	HB_TAG('H','M','A',' ')},	/* Western Mari -> High Mari */
+  {"ms",	HB_TAG('M','L','Y',' ')},	/* Malay [macrolanguage] */
+  {"msc",	HB_TAG('M','N','K',' ')},	/* Sankaran Maninka -> Maninka */
+  {"msh",	HB_TAG('M','L','G',' ')},	/* Masikoro Malagasy -> Malagasy */
+  {"msi",	HB_TAG('M','L','Y',' ')},	/* Sabah Malay -> Malay */
+  {"mt",	HB_TAG('M','T','S',' ')},	/* Maltese */
+  {"mtr",	HB_TAG('M','A','W',' ')},	/* Mewari -> Marwari */
+  {"mui",	HB_TAG('M','L','Y',' ')},	/* Musi -> Malay */
+  {"mup",	HB_TAG('R','A','J',' ')},	/* Malvi -> Rajasthani */
+  {"muq",	HB_TAG('H','M','N',' ')},	/* Eastern Xiangxi Miao -> Hmong */
+/*{"mus",	HB_TAG('M','U','S',' ')},*/	/* Creek -> Muscogee */
+  {"mvb",	HB_TAG('A','T','H',' ')},	/* Mattole -> Athapaskan */
+  {"mve",	HB_TAG('M','A','W',' ')},	/* Marwari (Pakistan) */
+  {"mvf",	HB_TAG('M','N','G',' ')},	/* Peripheral Mongolian -> Mongolian */
+  {"mwk",	HB_TAG('M','N','K',' ')},	/* Kita Maninkakan -> Maninka */
+/*{"mwl",	HB_TAG('M','W','L',' ')},*/	/* Mirandese */
+  {"mwr",	HB_TAG('M','A','W',' ')},	/* Marwari [macrolanguage] */
+/*{"mww",	HB_TAG('M','W','W',' ')},*/	/* Hmong Daw */
+  {"my",	HB_TAG('B','R','M',' ')},	/* Burmese */
+  {"mym",	HB_TAG('M','E','N',' ')},	/* Me’en */
+/*{"myn",	HB_TAG('M','Y','N',' ')},*/	/* Mayan [family] */
+  {"myq",	HB_TAG('M','N','K',' ')},	/* Forest Maninka (retired code) -> Maninka */
+  {"myv",	HB_TAG('E','R','Z',' ')},	/* Erzya */
+/*{"mzn",	HB_TAG('M','Z','N',' ')},*/	/* Mazanderani */
+  {"na",	HB_TAG('N','A','U',' ')},	/* Nauru -> Nauruan */
+/*{"nag",	HB_TAG('N','A','G',' ')},*/	/* Naga Pidgin -> Naga-Assamese */
+/*{"nah",	HB_TAG('N','A','H',' ')},*/	/* Nahuatl [family] */
+  {"nan",	HB_TAG('Z','H','S',' ')},	/* Min Nan Chinese -> Chinese Simplified */
+/*{"nap",	HB_TAG('N','A','P',' ')},*/	/* Neapolitan */
+  {"nb",	HB_TAG('N','O','R',' ')},	/* Norwegian Bokmål -> Norwegian */
+  {"nd",	HB_TAG('N','D','B',' ')},	/* North Ndebele -> Ndebele */
+/*{"ndc",	HB_TAG('N','D','C',' ')},*/	/* Ndau */
+/*{"nds",	HB_TAG('N','D','S',' ')},*/	/* Low Saxon */
+  {"ne",	HB_TAG('N','E','P',' ')},	/* Nepali [macrolanguage] */
+/*{"new",	HB_TAG('N','E','W',' ')},*/	/* Newari */
+  {"ng",	HB_TAG('N','D','G',' ')},	/* Ndonga */
+/*{"nga",	HB_TAG('N','G','A',' ')},*/	/* Ngbaka */
+  {"ngl",	HB_TAG('L','M','W',' ')},	/* Lomwe */
+  {"ngo",	HB_TAG('S','X','T',' ')},	/* Ngoni -> Sutu */
+  {"nhd",	HB_TAG('G','U','A',' ')},	/* Chiripá -> Guarani */
+  {"niq",	HB_TAG('K','A','L',' ')},	/* Nandi -> Kalenjin */
+/*{"niu",	HB_TAG('N','I','U',' ')},*/	/* Niuean */
+  {"niv",	HB_TAG('G','I','L',' ')},	/* Gilyak */
+  {"njz",	HB_TAG('N','I','S',' ')},	/* Nyishi -> Nisi */
+  {"nl",	HB_TAG('N','L','D',' ')},	/* Dutch */
+  {"nle",	HB_TAG('L','U','H',' ')},	/* East Nyala -> Luyia */
+  {"nn",	HB_TAG('N','Y','N',' ')},	/* Norwegian Nynorsk (Nynorsk, Norwegian) */
+  {"no",	HB_TAG('N','O','R',' ')},	/* Norwegian [macrolanguage] */
+  {"nod",	HB_TAG('N','T','A',' ')},	/* Northern Thai -> Northern Tai */
+/*{"noe",	HB_TAG('N','O','E',' ')},*/	/* Nimadi */
+/*{"nog",	HB_TAG('N','O','G',' ')},*/	/* Nogai */
+/*{"nov",	HB_TAG('N','O','V',' ')},*/	/* Novial */
+  {"npi",	HB_TAG('N','E','P',' ')},	/* Nepali */
+  {"nqo",	HB_TAG('N','K','O',' ')},	/* N’Ko */
+  {"nr",	HB_TAG('N','D','B',' ')},	/* South Ndebele -> Ndebele */
+  {"nsk",	HB_TAG('N','A','S',' ')},	/* Naskapi */
+/*{"nso",	HB_TAG('N','S','O',' ')},*/	/* Pedi -> Sotho, Northern */
+  {"nv",	HB_TAG('N','A','V',' ')},	/* Navajo */
+  {"nv",	HB_TAG('A','T','H',' ')},	/* Navajo -> Athapaskan */
+  {"ny",	HB_TAG('C','H','I',' ')},	/* Chichewa (Chewa, Nyanja) */
+  {"nyd",	HB_TAG('L','U','H',' ')},	/* Nyore -> Luyia */
+/*{"nym",	HB_TAG('N','Y','M',' ')},*/	/* Nyamwezi */
+  {"nyn",	HB_TAG('N','K','L',' ')},	/* Nyankole */
+/*{"nza",	HB_TAG('N','Z','A',' ')},*/	/* Tigon Mbembe -> Mbembe Tigon */
+  {"oc",	HB_TAG('O','C','I',' ')},	/* Occitan (post 1500) */
+  {"oj",	HB_TAG('O','J','B',' ')},	/* Ojibwa [macrolanguage] -> Ojibway */
+/*{"ojb",	HB_TAG('O','J','B',' ')},*/	/* Northwestern Ojibwa -> Ojibway */
+  {"ojc",	HB_TAG('O','J','B',' ')},	/* Central Ojibwa -> Ojibway */
+  {"ojg",	HB_TAG('O','J','B',' ')},	/* Eastern Ojibwa -> Ojibway */
+  {"ojs",	HB_TAG('O','C','R',' ')},	/* Severn Ojibwa -> Oji-Cree */
+  {"ojw",	HB_TAG('O','J','B',' ')},	/* Western Ojibwa -> Ojibway */
+  {"oki",	HB_TAG('K','A','L',' ')},	/* Okiek -> Kalenjin */
+  {"okm",	HB_TAG('K','O','H',' ')},	/* Middle Korean (10th-16th cent.) -> Korean Old Hangul */
+  {"om",	HB_TAG('O','R','O',' ')},	/* Oromo [macrolanguage] */
+  {"or",	HB_TAG('O','R','I',' ')},	/* Odia (formerly Oriya) [macrolanguage] */
+  {"orc",	HB_TAG('O','R','O',' ')},	/* Orma -> Oromo */
+  {"orn",	HB_TAG('M','L','Y',' ')},	/* Orang Kanaq -> Malay */
+  {"ors",	HB_TAG('M','L','Y',' ')},	/* Orang Seletar -> Malay */
+  {"ory",	HB_TAG('O','R','I',' ')},	/* Odia (formerly Oriya) */
+  {"os",	HB_TAG('O','S','S',' ')},	/* Ossetian */
+  {"otw",	HB_TAG('O','J','B',' ')},	/* Ottawa -> Ojibway */
+  {"pa",	HB_TAG('P','A','N',' ')},	/* Punjabi */
+/*{"pag",	HB_TAG('P','A','G',' ')},*/	/* Pangasinan */
+/*{"pam",	HB_TAG('P','A','M',' ')},*/	/* Pampanga -> Pampangan */
+  {"pap",	HB_TAG('P','A','P','0')},	/* Papiamento -> Papiamentu */
+/*{"pau",	HB_TAG('P','A','U',' ')},*/	/* Palauan */
+  {"pbt",	HB_TAG('P','A','S',' ')},	/* Southern Pashto -> Pashto */
+  {"pbu",	HB_TAG('P','A','S',' ')},	/* Northern Pashto -> Pashto */
+/*{"pcc",	HB_TAG('P','C','C',' ')},*/	/* Bouyei */
+/*{"pcd",	HB_TAG('P','C','D',' ')},*/	/* Picard */
+  {"pce",	HB_TAG('P','L','G',' ')},	/* Ruching Palaung -> Palaung */
+  {"pck",	HB_TAG('Q','I','N',' ')},	/* Paite Chin -> Chin */
+/*{"pdc",	HB_TAG('P','D','C',' ')},*/	/* Pennsylvania German */
+  {"pel",	HB_TAG('M','L','Y',' ')},	/* Pekal -> Malay */
+  {"pes",	HB_TAG('F','A','R',' ')},	/* Iranian Persian -> Persian */
+  {"pga",	HB_TAG('A','R','A',' ')},	/* Sudanese Creole Arabic -> Arabic */
+/*{"phk",	HB_TAG('P','H','K',' ')},*/	/* Phake */
+  {"pi",	HB_TAG('P','A','L',' ')},	/* Pali */
+/*{"pih",	HB_TAG('P','I','H',' ')},*/	/* Pitcairn-Norfolk -> Norfolk */
+  {"pko",	HB_TAG('K','A','L',' ')},	/* Pökoot -> Kalenjin */
+  {"pl",	HB_TAG('P','L','K',' ')},	/* Polish */
+  {"pll",	HB_TAG('P','L','G',' ')},	/* Shwe Palaung -> Palaung */
+  {"plp",	HB_TAG('P','A','P',' ')},	/* Palpa (retired code) */
+  {"plt",	HB_TAG('M','L','G',' ')},	/* Plateau Malagasy -> Malagasy */
+/*{"pms",	HB_TAG('P','M','S',' ')},*/	/* Piemontese */
+/*{"pnb",	HB_TAG('P','N','B',' ')},*/	/* Western Panjabi */
+/*{"poh",	HB_TAG('P','O','H',' ')},*/	/* Poqomchi' -> Pocomchi */
+/*{"pon",	HB_TAG('P','O','N',' ')},*/	/* Pohnpeian */
+  {"ppa",	HB_TAG('B','A','G',' ')},	/* Pao (retired code) -> Baghelkhandi */
+/*{"pro",	HB_TAG('P','R','O',' ')},*/	/* Old Provençal (to 1500) -> Provençal / Old Provençal */
+  {"prs",	HB_TAG('D','R','I',' ')},	/* Dari */
+  {"ps",	HB_TAG('P','A','S',' ')},	/* Pashto [macrolanguage] */
+  {"pse",	HB_TAG('M','L','Y',' ')},	/* Central Malay -> Malay */
+  {"pst",	HB_TAG('P','A','S',' ')},	/* Central Pashto -> Pashto */
+  {"pt",	HB_TAG('P','T','G',' ')},	/* Portuguese */
+/*{"pwo",	HB_TAG('P','W','O',' ')},*/	/* Pwo Western Karen -> Western Pwo Karen */
+  {"qu",	HB_TAG('Q','U','Z',' ')},	/* Quechua [macrolanguage] */
+  {"qub",	HB_TAG('Q','W','H',' ')},	/* Huallaga Huánuco Quechua -> Quechua (Peru) */
+/*{"quc",	HB_TAG('Q','U','C',' ')},*/	/* K’iche’ */
+  {"qud",	HB_TAG('Q','V','I',' ')},	/* Calderón Highland Quichua -> Quechua (Ecuador) */
+  {"quf",	HB_TAG('Q','U','Z',' ')},	/* Lambayeque Quechua -> Quechua */
+  {"qug",	HB_TAG('Q','V','I',' ')},	/* Chimborazo Highland Quichua -> Quechua (Ecuador) */
+/*{"quh",	HB_TAG('Q','U','H',' ')},*/	/* South Bolivian Quechua -> Quechua (Bolivia) */
+  {"quk",	HB_TAG('Q','U','Z',' ')},	/* Chachapoyas Quechua -> Quechua */
+  {"qul",	HB_TAG('Q','U','Z',' ')},	/* North Bolivian Quechua -> Quechua */
+  {"qup",	HB_TAG('Q','V','I',' ')},	/* Southern Pastaza Quechua -> Quechua (Ecuador) */
+  {"qur",	HB_TAG('Q','W','H',' ')},	/* Yanahuanca Pasco Quechua -> Quechua (Peru) */
+  {"qus",	HB_TAG('Q','U','H',' ')},	/* Santiago del Estero Quichua -> Quechua (Bolivia) */
+  {"quw",	HB_TAG('Q','V','I',' ')},	/* Tena Lowland Quichua -> Quechua (Ecuador) */
+  {"qux",	HB_TAG('Q','W','H',' ')},	/* Yauyos Quechua -> Quechua (Peru) */
+  {"quy",	HB_TAG('Q','U','Z',' ')},	/* Ayacucho Quechua -> Quechua */
+/*{"quz",	HB_TAG('Q','U','Z',' ')},*/	/* Cusco Quechua -> Quechua */
+  {"qva",	HB_TAG('Q','W','H',' ')},	/* Ambo-Pasco Quechua -> Quechua (Peru) */
+  {"qvc",	HB_TAG('Q','U','Z',' ')},	/* Cajamarca Quechua -> Quechua */
+  {"qve",	HB_TAG('Q','U','Z',' ')},	/* Eastern Apurímac Quechua -> Quechua */
+  {"qvh",	HB_TAG('Q','W','H',' ')},	/* Huamalíes-Dos de Mayo Huánuco Quechua -> Quechua (Peru) */
+/*{"qvi",	HB_TAG('Q','V','I',' ')},*/	/* Imbabura Highland Quichua -> Quechua (Ecuador) */
+  {"qvj",	HB_TAG('Q','V','I',' ')},	/* Loja Highland Quichua -> Quechua (Ecuador) */
+  {"qvl",	HB_TAG('Q','W','H',' ')},	/* Cajatambo North Lima Quechua -> Quechua (Peru) */
+  {"qvm",	HB_TAG('Q','W','H',' ')},	/* Margos-Yarowilca-Lauricocha Quechua -> Quechua (Peru) */
+  {"qvn",	HB_TAG('Q','W','H',' ')},	/* North Junín Quechua -> Quechua (Peru) */
+  {"qvo",	HB_TAG('Q','V','I',' ')},	/* Napo Lowland Quechua -> Quechua (Ecuador) */
+  {"qvp",	HB_TAG('Q','W','H',' ')},	/* Pacaraos Quechua -> Quechua (Peru) */
+  {"qvs",	HB_TAG('Q','U','Z',' ')},	/* San Martín Quechua -> Quechua */
+  {"qvw",	HB_TAG('Q','W','H',' ')},	/* Huaylla Wanca Quechua -> Quechua (Peru) */
+  {"qvz",	HB_TAG('Q','V','I',' ')},	/* Northern Pastaza Quichua -> Quechua (Ecuador) */
+  {"qwa",	HB_TAG('Q','W','H',' ')},	/* Corongo Ancash Quechua -> Quechua (Peru) */
+  {"qwc",	HB_TAG('Q','U','Z',' ')},	/* Classical Quechua -> Quechua */
+/*{"qwh",	HB_TAG('Q','W','H',' ')},*/	/* Huaylas Ancash Quechua -> Quechua (Peru) */
+  {"qws",	HB_TAG('Q','W','H',' ')},	/* Sihuas Ancash Quechua -> Quechua (Peru) */
+  {"qxa",	HB_TAG('Q','W','H',' ')},	/* Chiquián Ancash Quechua -> Quechua (Peru) */
+  {"qxc",	HB_TAG('Q','W','H',' ')},	/* Chincha Quechua -> Quechua (Peru) */
+  {"qxh",	HB_TAG('Q','W','H',' ')},	/* Panao Huánuco Quechua -> Quechua (Peru) */
+  {"qxl",	HB_TAG('Q','V','I',' ')},	/* Salasaca Highland Quichua -> Quechua (Ecuador) */
+  {"qxn",	HB_TAG('Q','W','H',' ')},	/* Northern Conchucos Ancash Quechua -> Quechua (Peru) */
+  {"qxo",	HB_TAG('Q','W','H',' ')},	/* Southern Conchucos Ancash Quechua -> Quechua (Peru) */
+  {"qxp",	HB_TAG('Q','U','Z',' ')},	/* Puno Quechua -> Quechua */
+  {"qxr",	HB_TAG('Q','V','I',' ')},	/* Cañar Highland Quichua -> Quechua (Ecuador) */
+  {"qxt",	HB_TAG('Q','W','H',' ')},	/* Santa Ana de Tusi Pasco Quechua -> Quechua (Peru) */
+  {"qxu",	HB_TAG('Q','U','Z',' ')},	/* Arequipa-La Unión Quechua -> Quechua */
+  {"qxw",	HB_TAG('Q','W','H',' ')},	/* Jauja Wanca Quechua -> Quechua (Peru) */
+  {"rag",	HB_TAG('L','U','H',' ')},	/* Logooli -> Luyia */
+/*{"raj",	HB_TAG('R','A','J',' ')},*/	/* Rajasthani [macrolanguage] */
+/*{"rar",	HB_TAG('R','A','R',' ')},*/	/* Rarotongan */
+  {"rbb",	HB_TAG('P','L','G',' ')},	/* Rumai Palaung -> Palaung */
+  {"rbl",	HB_TAG('B','I','K',' ')},	/* Miraya Bikol -> Bikol */
+/*{"rej",	HB_TAG('R','E','J',' ')},*/	/* Rejang */
+/*{"ria",	HB_TAG('R','I','A',' ')},*/	/* Riang (India) */
+/*{"rif",	HB_TAG('R','I','F',' ')},*/	/* Tarifit */
+/*{"rit",	HB_TAG('R','I','T',' ')},*/	/* Ritharrngu -> Ritarungo */
+  {"rki",	HB_TAG('A','R','K',' ')},	/* Rakhine */
+/*{"rkw",	HB_TAG('R','K','W',' ')},*/	/* Arakwal */
+  {"rm",	HB_TAG('R','M','S',' ')},	/* Romansh */
+  {"rmc",	HB_TAG('R','O','Y',' ')},	/* Carpathian Romani -> Romany */
+  {"rmf",	HB_TAG('R','O','Y',' ')},	/* Kalo Finnish Romani -> Romany */
+  {"rml",	HB_TAG('R','O','Y',' ')},	/* Baltic Romani -> Romany */
+  {"rmn",	HB_TAG('R','O','Y',' ')},	/* Balkan Romani -> Romany */
+  {"rmo",	HB_TAG('R','O','Y',' ')},	/* Sinte Romani -> Romany */
+  {"rmw",	HB_TAG('R','O','Y',' ')},	/* Welsh Romani -> Romany */
+/*{"rmy",	HB_TAG('R','M','Y',' ')},*/	/* Vlax Romani */
+  {"rmz",	HB_TAG('A','R','K',' ')},	/* Marma -> Rakhine */
+  {"rn",	HB_TAG('R','U','N',' ')},	/* Rundi */
+  {"rnl",	HB_TAG('H','A','L',' ')},	/* Ranglong -> Halam (Falam Chin) */
+  {"ro",	HB_TAG('R','O','M',' ')},	/* Romanian */
+  {"rom",	HB_TAG('R','O','Y',' ')},	/* Romany [macrolanguage] */
+/*{"rtm",	HB_TAG('R','T','M',' ')},*/	/* Rotuman */
+  {"ru",	HB_TAG('R','U','S',' ')},	/* Russian */
+  {"rue",	HB_TAG('R','S','Y',' ')},	/* Rusyn */
+/*{"rup",	HB_TAG('R','U','P',' ')},*/	/* Aromanian */
+  {"rw",	HB_TAG('R','U','A',' ')},	/* Kinyarwanda */
+  {"rwr",	HB_TAG('M','A','W',' ')},	/* Marwari (India) */
+  {"sa",	HB_TAG('S','A','N',' ')},	/* Sanskrit */
+  {"sah",	HB_TAG('Y','A','K',' ')},	/* Yakut -> Sakha */
+  {"sam",	HB_TAG('P','A','A',' ')},	/* Samaritan Aramaic -> Palestinian Aramaic */
+/*{"sas",	HB_TAG('S','A','S',' ')},*/	/* Sasak */
+/*{"sat",	HB_TAG('S','A','T',' ')},*/	/* Santali */
+  {"sc",	HB_TAG('S','R','D',' ')},	/* Sardinian [macrolanguage] */
+  {"sck",	HB_TAG('S','A','D',' ')},	/* Sadri */
+/*{"scn",	HB_TAG('S','C','N',' ')},*/	/* Sicilian */
+/*{"sco",	HB_TAG('S','C','O',' ')},*/	/* Scots */
+  {"scs",	HB_TAG('S','C','S',' ')},	/* North Slavey */
+  {"scs",	HB_TAG('S','L','A',' ')},	/* North Slavey -> Slavey */
+  {"scs",	HB_TAG('A','T','H',' ')},	/* North Slavey -> Athapaskan */
+  {"sd",	HB_TAG('S','N','D',' ')},	/* Sindhi */
+  {"sdc",	HB_TAG('S','R','D',' ')},	/* Sassarese Sardinian -> Sardinian */
+  {"sdh",	HB_TAG('K','U','R',' ')},	/* Southern Kurdish -> Kurdish */
+  {"sdn",	HB_TAG('S','R','D',' ')},	/* Gallurese Sardinian -> Sardinian */
+  {"se",	HB_TAG('N','S','M',' ')},	/* Northern Sami */
+  {"seh",	HB_TAG('S','N','A',' ')},	/* Sena */
+  {"sek",	HB_TAG('A','T','H',' ')},	/* Sekani -> Athapaskan */
+/*{"sel",	HB_TAG('S','E','L',' ')},*/	/* Selkup */
+  {"sez",	HB_TAG('Q','I','N',' ')},	/* Senthang Chin -> Chin */
+  {"sfm",	HB_TAG('H','M','N',' ')},	/* Small Flowery Miao -> Hmong */
+  {"sg",	HB_TAG('S','G','O',' ')},	/* Sango */
+/*{"sga",	HB_TAG('S','G','A',' ')},*/	/* Old Irish (to 900) */
+  {"sgc",	HB_TAG('K','A','L',' ')},	/* Kipsigis -> Kalenjin */
+/*{"sgs",	HB_TAG('S','G','S',' ')},*/	/* Samogitian */
+  {"sgw",	HB_TAG('C','H','G',' ')},	/* Sebat Bet Gurage -> Chaha Gurage */
+  {"sgw",	HB_TAG('S','G','W',' ')},	/* Sebat Bet Gurage -> Chaha Gurage (SIL fonts) */
+/*{"shi",	HB_TAG('S','H','I',' ')},*/	/* Tachelhit */
+/*{"shn",	HB_TAG('S','H','N',' ')},*/	/* Shan */
+  {"shu",	HB_TAG('A','R','A',' ')},	/* Chadian Arabic -> Arabic */
+  {"si",	HB_TAG('S','N','H',' ')},	/* Sinhala (Sinhalese) */
+/*{"sid",	HB_TAG('S','I','D',' ')},*/	/* Sidamo */
+  {"sjd",	HB_TAG('K','S','M',' ')},	/* Kildin Sami */
+  {"sjo",	HB_TAG('S','I','B',' ')},	/* Xibe -> Sibe */
+  {"sk",	HB_TAG('S','K','Y',' ')},	/* Slovak */
+  {"skg",	HB_TAG('M','L','G',' ')},	/* Sakalava Malagasy -> Malagasy */
+  {"skr",	HB_TAG('S','R','K',' ')},	/* Saraiki */
+  {"sl",	HB_TAG('S','L','V',' ')},	/* Slovenian */
+  {"sm",	HB_TAG('S','M','O',' ')},	/* Samoan */
+  {"sma",	HB_TAG('S','S','M',' ')},	/* Southern Sami */
+  {"smj",	HB_TAG('L','S','M',' ')},	/* Lule Sami */
+  {"smn",	HB_TAG('I','S','M',' ')},	/* Inari Sami */
+  {"sms",	HB_TAG('S','K','S',' ')},	/* Skolt Sami */
+  {"sn",	HB_TAG('S','N','A','0')},	/* Shona */
+/*{"snk",	HB_TAG('S','N','K',' ')},*/	/* Soninke */
+  {"so",	HB_TAG('S','M','L',' ')},	/* Somali */
+/*{"sop",	HB_TAG('S','O','P',' ')},*/	/* Songe */
+  {"spv",	HB_TAG('O','R','I',' ')},	/* Sambalpuri -> Odia (formerly Oriya) */
+  {"spy",	HB_TAG('K','A','L',' ')},	/* Sabaot -> Kalenjin */
+  {"sq",	HB_TAG('S','Q','I',' ')},	/* Albanian [macrolanguage] */
+  {"sr",	HB_TAG('S','R','B',' ')},	/* Serbian */
+  {"src",	HB_TAG('S','R','D',' ')},	/* Logudorese Sardinian -> Sardinian */
+  {"sro",	HB_TAG('S','R','D',' ')},	/* Campidanese Sardinian -> Sardinian */
+/*{"srr",	HB_TAG('S','R','R',' ')},*/	/* Serer */
+  {"srs",	HB_TAG('A','T','H',' ')},	/* Sarsi -> Athapaskan */
+  {"ss",	HB_TAG('S','W','Z',' ')},	/* Swati */
+  {"ssh",	HB_TAG('A','R','A',' ')},	/* Shihhi Arabic -> Arabic */
+  {"st",	HB_TAG('S','O','T',' ')},	/* Southern Sotho -> Sotho, Southern */
+/*{"stq",	HB_TAG('S','T','Q',' ')},*/	/* Saterfriesisch -> Saterland Frisian */
+  {"stv",	HB_TAG('S','I','G',' ')},	/* Silt'e -> Silte Gurage */
+  {"su",	HB_TAG('S','U','N',' ')},	/* Sundanese */
+/*{"suk",	HB_TAG('S','U','K',' ')},*/	/* Sukuma */
+  {"suq",	HB_TAG('S','U','R',' ')},	/* Suri */
+  {"sv",	HB_TAG('S','V','E',' ')},	/* Swedish */
+/*{"sva",	HB_TAG('S','V','A',' ')},*/	/* Svan */
+  {"sw",	HB_TAG('S','W','K',' ')},	/* Swahili [macrolanguage] */
+  {"swb",	HB_TAG('C','M','R',' ')},	/* Maore Comorian -> Comorian */
+  {"swc",	HB_TAG('S','W','K',' ')},	/* Congo Swahili -> Swahili */
+  {"swh",	HB_TAG('S','W','K',' ')},	/* Swahili */
+  {"swv",	HB_TAG('M','A','W',' ')},	/* Shekhawati -> Marwari */
+/*{"sxu",	HB_TAG('S','X','U',' ')},*/	/* Upper Saxon */
+  {"syc",	HB_TAG('S','Y','R',' ')},	/* Classical Syriac -> Syriac */
+/*{"syl",	HB_TAG('S','Y','L',' ')},*/	/* Sylheti */
+/*{"syr",	HB_TAG('S','Y','R',' ')},*/	/* Syriac [macrolanguage] */
+/*{"szl",	HB_TAG('S','Z','L',' ')},*/	/* Silesian */
+  {"ta",	HB_TAG('T','A','M',' ')},	/* Tamil */
+  {"taa",	HB_TAG('A','T','H',' ')},	/* Lower Tanana -> Athapaskan */
+/*{"tab",	HB_TAG('T','A','B',' ')},*/	/* Tabassaran -> Tabasaran */
+  {"taq",	HB_TAG('T','M','H',' ')},	/* Tamasheq -> Tamashek */
+  {"tau",	HB_TAG('A','T','H',' ')},	/* Upper Tanana -> Athapaskan */
+  {"tcb",	HB_TAG('A','T','H',' ')},	/* Tanacross -> Athapaskan */
+  {"tce",	HB_TAG('A','T','H',' ')},	/* Southern Tutchone -> Athapaskan */
+  {"tcp",	HB_TAG('Q','I','N',' ')},	/* Tawr Chin -> Chin */
+  {"tcy",	HB_TAG('T','U','L',' ')},	/* Tulu -> Tumbuka */
+  {"tcz",	HB_TAG('Q','I','N',' ')},	/* Thado Chin -> Chin */
+/*{"tdd",	HB_TAG('T','D','D',' ')},*/	/* Tai Nüa -> Dehong Dai */
+  {"tdx",	HB_TAG('M','L','G',' ')},	/* Tandroy-Mahafaly Malagasy -> Malagasy */
+  {"te",	HB_TAG('T','E','L',' ')},	/* Telugu */
+  {"tec",	HB_TAG('K','A','L',' ')},	/* Terik -> Kalenjin */
+  {"tem",	HB_TAG('T','M','N',' ')},	/* Timne -> Temne */
+/*{"tet",	HB_TAG('T','E','T',' ')},*/	/* Tetum */
+  {"tfn",	HB_TAG('A','T','H',' ')},	/* Tanaina -> Athapaskan */
+  {"tg",	HB_TAG('T','A','J',' ')},	/* Tajik -> Tajiki */
+  {"tgj",	HB_TAG('N','I','S',' ')},	/* Tagin -> Nisi */
+  {"tgx",	HB_TAG('A','T','H',' ')},	/* Tagish -> Athapaskan */
+  {"th",	HB_TAG('T','H','A',' ')},	/* Thai */
+  {"tht",	HB_TAG('A','T','H',' ')},	/* Tahltan -> Athapaskan */
+  {"thv",	HB_TAG('T','M','H',' ')},	/* Tahaggart Tamahaq -> Tamashek */
+  {"thz",	HB_TAG('T','M','H',' ')},	/* Tayart Tamajeq -> Tamashek */
+  {"ti",	HB_TAG('T','G','Y',' ')},	/* Tigrinya */
+  {"tig",	HB_TAG('T','G','R',' ')},	/* Tigre */
+/*{"tiv",	HB_TAG('T','I','V',' ')},*/	/* Tiv */
+  {"tk",	HB_TAG('T','K','M',' ')},	/* Turkmen */
+  {"tkg",	HB_TAG('M','L','G',' ')},	/* Tesaka Malagasy -> Malagasy */
+  {"tl",	HB_TAG('T','G','L',' ')},	/* Tagalog */
+/*{"tmh",	HB_TAG('T','M','H',' ')},*/	/* Tamashek [macrolanguage] */
+  {"tmw",	HB_TAG('M','L','Y',' ')},	/* Temuan -> Malay */
+  {"tn",	HB_TAG('T','N','A',' ')},	/* Tswana */
+  {"tnf",	HB_TAG('D','R','I',' ')},	/* Tangshewi (retired code) -> Dari */
+  {"to",	HB_TAG('T','G','N',' ')},	/* Tonga (Tonga Islands) -> Tongan */
+  {"tod",	HB_TAG('T','O','D','0')},	/* Toma */
+  {"toi",	HB_TAG('T','N','G',' ')},	/* Tonga (Zambia) */
+  {"tol",	HB_TAG('A','T','H',' ')},	/* Tolowa -> Athapaskan */
+/*{"tpi",	HB_TAG('T','P','I',' ')},*/	/* Tok Pisin */
+  {"tr",	HB_TAG('T','R','K',' ')},	/* Turkish */
+  {"tru",	HB_TAG('T','U','A',' ')},	/* Turoyo -> Turoyo Aramaic */
+  {"tru",	HB_TAG('S','Y','R',' ')},	/* Turoyo -> Syriac */
+  {"ts",	HB_TAG('T','S','G',' ')},	/* Tsonga */
+/*{"tsj",	HB_TAG('T','S','J',' ')},*/	/* Tshangla */
+  {"tt",	HB_TAG('T','A','T',' ')},	/* Tatar */
+  {"ttm",	HB_TAG('A','T','H',' ')},	/* Northern Tutchone -> Athapaskan */
+  {"ttq",	HB_TAG('T','M','H',' ')},	/* Tawallammat Tamajaq -> Tamashek */
+/*{"tum",	HB_TAG('T','U','M',' ')},*/	/* Tumbuka -> Tulu */
+  {"tuu",	HB_TAG('A','T','H',' ')},	/* Tututni -> Athapaskan */
+  {"tuy",	HB_TAG('K','A','L',' ')},	/* Tugen -> Kalenjin */
+/*{"tvl",	HB_TAG('T','V','L',' ')},*/	/* Tuvalu */
+  {"tw",	HB_TAG('T','W','I',' ')},	/* Twi */
+  {"tw",	HB_TAG('A','K','A',' ')},	/* Twi -> Akan */
+  {"txc",	HB_TAG('A','T','H',' ')},	/* Tsetsaut -> Athapaskan */
+  {"txy",	HB_TAG('M','L','G',' ')},	/* Tanosy Malagasy -> Malagasy */
+  {"ty",	HB_TAG('T','H','T',' ')},	/* Tahitian */
+  {"tyv",	HB_TAG('T','U','V',' ')},	/* Tuvinian -> Tuvin */
+/*{"tyz",	HB_TAG('T','Y','Z',' ')},*/	/* Tày */
+/*{"tzm",	HB_TAG('T','Z','M',' ')},*/	/* Central Atlas Tamazight -> Tamazight */
+/*{"tzo",	HB_TAG('T','Z','O',' ')},*/	/* Tzotzil */
+  {"ubl",	HB_TAG('B','I','K',' ')},	/* Buhi'non Bikol -> Bikol */
+/*{"udm",	HB_TAG('U','D','M',' ')},*/	/* Udmurt */
+  {"ug",	HB_TAG('U','Y','G',' ')},	/* Uyghur */
+  {"uk",	HB_TAG('U','K','R',' ')},	/* Ukrainian */
+  {"uki",	HB_TAG('K','U','I',' ')},	/* Kui (India) */
+/*{"umb",	HB_TAG('U','M','B',' ')},*/	/* Umbundu */
+  {"unr",	HB_TAG('M','U','N',' ')},	/* Mundari */
+  {"ur",	HB_TAG('U','R','D',' ')},	/* Urdu */
+  {"urk",	HB_TAG('M','L','Y',' ')},	/* Urak Lawoi' -> Malay */
+  {"uz",	HB_TAG('U','Z','B',' ')},	/* Uzbek [macrolanguage] */
+  {"uzn",	HB_TAG('U','Z','B',' ')},	/* Northern Uzbek -> Uzbek */
+  {"uzs",	HB_TAG('U','Z','B',' ')},	/* Southern Uzbek -> Uzbek */
+  {"ve",	HB_TAG('V','E','N',' ')},	/* Venda */
+/*{"vec",	HB_TAG('V','E','C',' ')},*/	/* Venetian */
+  {"vi",	HB_TAG('V','I','T',' ')},	/* Vietnamese */
+  {"vkk",	HB_TAG('M','L','Y',' ')},	/* Kaur -> Malay */
+  {"vkt",	HB_TAG('M','L','Y',' ')},	/* Tenggarong Kutai Malay -> Malay */
+  {"vls",	HB_TAG('F','L','E',' ')},	/* Vlaams -> Dutch (Flemish) */
+  {"vmw",	HB_TAG('M','A','K',' ')},	/* Makhuwa */
+  {"vo",	HB_TAG('V','O','L',' ')},	/* Volapük */
+/*{"vro",	HB_TAG('V','R','O',' ')},*/	/* Võro */
+  {"wa",	HB_TAG('W','L','N',' ')},	/* Walloon */
+/*{"war",	HB_TAG('W','A','R',' ')},*/	/* Waray (Philippines) -> Waray-Waray */
+  {"wbm",	HB_TAG('W','A',' ',' ')},	/* Wa */
+  {"wbr",	HB_TAG('W','A','G',' ')},	/* Wagdi */
+  {"wlc",	HB_TAG('C','M','R',' ')},	/* Mwali Comorian -> Comorian */
+  {"wle",	HB_TAG('S','I','G',' ')},	/* Wolane -> Silte Gurage */
+  {"wlk",	HB_TAG('A','T','H',' ')},	/* Wailaki -> Athapaskan */
+  {"wni",	HB_TAG('C','M','R',' ')},	/* Ndzwani Comorian -> Comorian */
+  {"wo",	HB_TAG('W','L','F',' ')},	/* Wolof */
+  {"wry",	HB_TAG('M','A','W',' ')},	/* Merwari -> Marwari */
+  {"wsg",	HB_TAG('G','O','N',' ')},	/* Adilabad Gondi -> Gondi */
+/*{"wtm",	HB_TAG('W','T','M',' ')},*/	/* Mewati */
+  {"wuu",	HB_TAG('Z','H','S',' ')},	/* Wu Chinese -> Chinese Simplified */
+  {"xal",	HB_TAG('K','L','M',' ')},	/* Kalmyk */
+  {"xal",	HB_TAG('T','O','D',' ')},	/* Kalmyk -> Todo */
+  {"xan",	HB_TAG('S','E','K',' ')},	/* Xamtanga -> Sekota */
+  {"xh",	HB_TAG('X','H','S',' ')},	/* Xhosa */
+/*{"xjb",	HB_TAG('X','J','B',' ')},*/	/* Minjungbal -> Minjangbal */
+/*{"xkf",	HB_TAG('X','K','F',' ')},*/	/* Khengkha */
+  {"xmm",	HB_TAG('M','L','Y',' ')},	/* Manado Malay -> Malay */
+  {"xmv",	HB_TAG('M','L','G',' ')},	/* Antankarana Malagasy -> Malagasy */
+  {"xmw",	HB_TAG('M','L','G',' ')},	/* Tsimihety Malagasy -> Malagasy */
+  {"xnr",	HB_TAG('D','G','R',' ')},	/* Kangri -> Dogri */
+/*{"xog",	HB_TAG('X','O','G',' ')},*/	/* Soga */
+/*{"xpe",	HB_TAG('X','P','E',' ')},*/	/* Liberia Kpelle -> Kpelle (Liberia) */
+  {"xsl",	HB_TAG('S','S','L',' ')},	/* South Slavey */
+  {"xsl",	HB_TAG('S','L','A',' ')},	/* South Slavey -> Slavey */
+  {"xsl",	HB_TAG('A','T','H',' ')},	/* South Slavey -> Athapaskan */
+  {"xst",	HB_TAG('S','I','G',' ')},	/* Silt'e (retired code) -> Silte Gurage */
+  {"xwo",	HB_TAG('T','O','D',' ')},	/* Written Oirat -> Todo */
+/*{"yao",	HB_TAG('Y','A','O',' ')},*/	/* Yao */
+/*{"yap",	HB_TAG('Y','A','P',' ')},*/	/* Yapese */
+  {"ybd",	HB_TAG('A','R','K',' ')},	/* Yangbye (retired code) -> Rakhine */
+  {"ydd",	HB_TAG('J','I','I',' ')},	/* Eastern Yiddish -> Yiddish */
+  {"yi",	HB_TAG('J','I','I',' ')},	/* Yiddish [macrolanguage] */
+  {"yih",	HB_TAG('J','I','I',' ')},	/* Western Yiddish -> Yiddish */
+  {"yo",	HB_TAG('Y','B','A',' ')},	/* Yoruba */
+  {"yos",	HB_TAG('Q','I','N',' ')},	/* Yos (retired code) -> Chin */
+  {"yrk",	HB_TAG('T','N','E',' ')},	/* Nenets -> Tundra Nenets */
+  {"yrk",	HB_TAG('F','N','E',' ')},	/* Nenets -> Forest Nenets */
+  {"yue",	HB_TAG('Z','H','H',' ')},	/* Yue Chinese -> Chinese, Hong Kong SAR */
+  {"za",	HB_TAG('Z','H','A',' ')},	/* Zhuang [macrolanguage] */
+  {"zch",	HB_TAG('Z','H','A',' ')},	/* Central Hongshuihe Zhuang -> Zhuang */
+  {"zdj",	HB_TAG('C','M','R',' ')},	/* Ngazidja Comorian -> Comorian */
+/*{"zea",	HB_TAG('Z','E','A',' ')},*/	/* Zeeuws -> Zealandic */
+  {"zeh",	HB_TAG('Z','H','A',' ')},	/* Eastern Hongshuihe Zhuang -> Zhuang */
+  {"zgb",	HB_TAG('Z','H','A',' ')},	/* Guibei Zhuang -> Zhuang */
+/*{"zgh",	HB_TAG('Z','G','H',' ')},*/	/* Standard Moroccan Tamazight */
+  {"zgm",	HB_TAG('Z','H','A',' ')},	/* Minz Zhuang -> Zhuang */
+  {"zgn",	HB_TAG('Z','H','A',' ')},	/* Guibian Zhuang -> Zhuang */
+  {"zh",	HB_TAG('Z','H','S',' ')},	/* Chinese [macrolanguage] -> Chinese Simplified */
+  {"zhd",	HB_TAG('Z','H','A',' ')},	/* Dai Zhuang -> Zhuang */
+  {"zhn",	HB_TAG('Z','H','A',' ')},	/* Nong Zhuang -> Zhuang */
+  {"zlj",	HB_TAG('Z','H','A',' ')},	/* Liujiang Zhuang -> Zhuang */
+  {"zlm",	HB_TAG('M','L','Y',' ')},	/* Malay */
+  {"zln",	HB_TAG('Z','H','A',' ')},	/* Lianshan Zhuang -> Zhuang */
+  {"zlq",	HB_TAG('Z','H','A',' ')},	/* Liuqian Zhuang -> Zhuang */
+  {"zmi",	HB_TAG('M','L','Y',' ')},	/* Negeri Sembilan Malay -> Malay */
+  {"zne",	HB_TAG('Z','N','D',' ')},	/* Zande */
+  {"zom",	HB_TAG('Q','I','N',' ')},	/* Zou -> Chin */
+  {"zqe",	HB_TAG('Z','H','A',' ')},	/* Qiubei Zhuang -> Zhuang */
+  {"zsm",	HB_TAG('M','L','Y',' ')},	/* Standard Malay -> Malay */
+  {"zu",	HB_TAG('Z','U','L',' ')},	/* Zulu */
+  {"zum",	HB_TAG('L','R','C',' ')},	/* Kumzari -> Luri */
+  {"zyb",	HB_TAG('Z','H','A',' ')},	/* Yongbei Zhuang -> Zhuang */
+  {"zyg",	HB_TAG('Z','H','A',' ')},	/* Yang Zhuang -> Zhuang */
+  {"zyj",	HB_TAG('Z','H','A',' ')},	/* Youjiang Zhuang -> Zhuang */
+  {"zyn",	HB_TAG('Z','H','A',' ')},	/* Yongnan Zhuang -> Zhuang */
+/*{"zza",	HB_TAG('Z','Z','A',' ')},*/	/* Zazaki [macrolanguage] */
+  {"zzj",	HB_TAG('Z','H','A',' ')},	/* Zuojiang Zhuang -> Zhuang */
+};
+
+/**
+ * hb_ot_tags_from_complex_language:
+ * @lang_str: a BCP 47 language tag to convert.
+ * @limit: a pointer to the end of the substring of @lang_str to consider for
+ * conversion.
+ * @count: maximum number of language tags to retrieve (IN) and actual number of
+ * language tags retrieved (OUT). If no tags are retrieved, it is not modified.
+ * @tags: array of size at least @language_count to store the language tag
+ * results
+ *
+ * Converts a multi-subtag BCP 47 language tag to language tags.
+ *
+ * Return value: Whether any language systems were retrieved.
+ **/
+static bool
+hb_ot_tags_from_complex_language (const char   *lang_str,
+				  const char   *limit,
+				  unsigned int *count /* IN/OUT */,
+				  hb_tag_t     *tags /* OUT */)
+{
+  if (subtag_matches (lang_str, limit, "-fonnapa"))
+  {
+    /* Undetermined; North American Phonetic Alphabet */
+    tags[0] = HB_TAG('A','P','P','H');  /* Phonetic transcription—Americanist conventions */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-polyton"))
+  {
+    /* Modern Greek (1453-); Polytonic Greek */
+    tags[0] = HB_TAG('P','G','R',' ');  /* Polytonic Greek */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-provenc"))
+  {
+    /* Occitan (post 1500); Provençal */
+    tags[0] = HB_TAG('P','R','O',' ');  /* Provençal / Old Provençal */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-fonipa"))
+  {
+    /* Undetermined; International Phonetic Alphabet */
+    tags[0] = HB_TAG('I','P','P','H');  /* Phonetic transcription—IPA conventions */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-geok"))
+  {
+    /* Undetermined; Khutsuri (Asomtavruli and Nuskhuri) */
+    tags[0] = HB_TAG('K','G','E',' ');  /* Khutsuri Georgian */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-syre"))
+  {
+    /* Undetermined; Syriac (Estrangelo variant) */
+    tags[0] = HB_TAG('S','Y','R','E');  /* Syriac, Estrangela script-variant (equivalent to ISO 15924 'Syre') */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-syrj"))
+  {
+    /* Undetermined; Syriac (Western variant) */
+    tags[0] = HB_TAG('S','Y','R','J');  /* Syriac, Western script-variant (equivalent to ISO 15924 'Syrj') */
+    *count = 1;
+    return true;
+  }
+  if (subtag_matches (lang_str, limit, "-syrn"))
+  {
+    /* Undetermined; Syriac (Eastern variant) */
+    tags[0] = HB_TAG('S','Y','R','N');  /* Syriac, Eastern script-variant (equivalent to ISO 15924 'Syrn') */
+    *count = 1;
+    return true;
+  }
+  switch (lang_str[0])
+  {
+  case 'a':
+    if (0 == strcmp (&lang_str[1], "rt-lojban"))
+    {
+      /* Lojban */
+      tags[0] = HB_TAG('J','B','O',' ');  /* Lojban */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'c':
+    if (lang_matches (&lang_str[1], "do-hant-hk"))
+    {
+      /* Min Dong Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "do-hant-mo"))
+    {
+      /* Min Dong Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "jy-hant-hk"))
+    {
+      /* Jinyu Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "jy-hant-mo"))
+    {
+      /* Jinyu Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "mn-hant-hk"))
+    {
+      /* Mandarin Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "mn-hant-mo"))
+    {
+      /* Mandarin Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hant-hk"))
+    {
+      /* Northern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hant-mo"))
+    {
+      /* Northern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "px-hant-hk"))
+    {
+      /* Pu-Xian Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "px-hant-mo"))
+    {
+      /* Pu-Xian Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sp-hant-hk"))
+    {
+      /* Southern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sp-hant-mo"))
+    {
+      /* Southern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zh-hant-hk"))
+    {
+      /* Huizhou Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zh-hant-mo"))
+    {
+      /* Huizhou Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zo-hant-hk"))
+    {
+      /* Min Zhong Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zo-hant-mo"))
+    {
+      /* Min Zhong Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "do-hans"))
+    {
+      /* Min Dong Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "do-hant"))
+    {
+      /* Min Dong Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "jy-hans"))
+    {
+      /* Jinyu Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "jy-hant"))
+    {
+      /* Jinyu Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "mn-hans"))
+    {
+      /* Mandarin Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "mn-hant"))
+    {
+      /* Mandarin Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hans"))
+    {
+      /* Northern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hant"))
+    {
+      /* Northern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "px-hans"))
+    {
+      /* Pu-Xian Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "px-hant"))
+    {
+      /* Pu-Xian Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sp-hans"))
+    {
+      /* Southern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sp-hant"))
+    {
+      /* Southern Ping Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zh-hans"))
+    {
+      /* Huizhou Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zh-hant"))
+    {
+      /* Huizhou Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zo-hans"))
+    {
+      /* Min Zhong Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "zo-hant"))
+    {
+      /* Min Zhong Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "do-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Min Dong Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "do-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Min Dong Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "do-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Min Dong Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "jy-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Jinyu Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "jy-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Jinyu Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "jy-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Jinyu Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "mn-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Mandarin Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "mn-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Mandarin Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "mn-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Mandarin Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Northern Ping Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Northern Ping Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Northern Ping Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "px-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Pu-Xian Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "px-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Pu-Xian Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "px-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Pu-Xian Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sp-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Southern Ping Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sp-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Southern Ping Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sp-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Southern Ping Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zh-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Huizhou Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zh-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Huizhou Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zh-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Huizhou Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zo-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Min Zhong Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zo-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Min Zhong Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "zo-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Min Zhong Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'g':
+    if (lang_matches (&lang_str[1], "an-hant-hk"))
+    {
+      /* Gan Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hant-mo"))
+    {
+      /* Gan Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hans"))
+    {
+      /* Gan Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hant"))
+    {
+      /* Gan Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "a-latg"))
+    {
+      /* Irish */
+      tags[0] = HB_TAG('I','R','T',' ');  /* Irish Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Gan Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Gan Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Gan Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'h':
+    if (lang_matches (&lang_str[1], "ak-hant-hk"))
+    {
+      /* Hakka Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "ak-hant-mo"))
+    {
+      /* Hakka Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sn-hant-hk"))
+    {
+      /* Xiang Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sn-hant-mo"))
+    {
+      /* Xiang Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "ak-hans"))
+    {
+      /* Hakka Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "ak-hant"))
+    {
+      /* Hakka Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sn-hans"))
+    {
+      /* Xiang Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "sn-hant"))
+    {
+      /* Xiang Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "ak-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Hakka Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "ak-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Hakka Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "ak-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Hakka Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sn-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Xiang Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sn-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Xiang Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "sn-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Xiang Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'i':
+    if (0 == strcmp (&lang_str[1], "-navajo"))
+    {
+      /* Navajo */
+      unsigned int i;
+      hb_tag_t possible_tags[] = {
+	HB_TAG('N','A','V',' '),  /* Navajo */
+	HB_TAG('A','T','H',' '),  /* Athapaskan */
+      };
+      for (i = 0; i < 2 && i < *count; i++)
+	tags[i] = possible_tags[i];
+      *count = i;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "-hak"))
+    {
+      /* Hakka */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "-lux"))
+    {
+      /* Luxembourgish */
+      tags[0] = HB_TAG('L','T','Z',' ');  /* Luxembourgish */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'l':
+    if (lang_matches (&lang_str[1], "zh-hans"))
+    {
+      /* Literary Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'm':
+    if (lang_matches (&lang_str[1], "np-hant-hk"))
+    {
+      /* Min Bei Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hant-mo"))
+    {
+      /* Min Bei Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hans"))
+    {
+      /* Min Bei Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "np-hant"))
+    {
+      /* Min Bei Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Min Bei Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Min Bei Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "np-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Min Bei Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'n':
+    if (lang_matches (&lang_str[1], "an-hant-hk"))
+    {
+      /* Min Nan Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hant-mo"))
+    {
+      /* Min Nan Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hans"))
+    {
+      /* Min Nan Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "an-hant"))
+    {
+      /* Min Nan Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Min Nan Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Min Nan Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "an-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Min Nan Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "o-bok"))
+    {
+      /* Norwegian Bokmal */
+      tags[0] = HB_TAG('N','O','R',' ');  /* Norwegian */
+      *count = 1;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "o-nyn"))
+    {
+      /* Norwegian Nynorsk */
+      tags[0] = HB_TAG('N','Y','N',' ');  /* Norwegian Nynorsk (Nynorsk, Norwegian) */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'r':
+    if (0 == strncmp (&lang_str[1], "o-", 2)
+	&& subtag_matches (lang_str, limit, "-md"))
+    {
+      /* Romanian; Moldova */
+      tags[0] = HB_TAG('M','O','L',' ');  /* Moldavian */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'w':
+    if (lang_matches (&lang_str[1], "uu-hant-hk"))
+    {
+      /* Wu Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "uu-hant-mo"))
+    {
+      /* Wu Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "uu-hans"))
+    {
+      /* Wu Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "uu-hant"))
+    {
+      /* Wu Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "uu-", 3)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Wu Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "uu-", 3)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Wu Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "uu-", 3)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Wu Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'y':
+    if (lang_matches (&lang_str[1], "ue-hans"))
+    {
+      /* Yue Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    break;
+  case 'z':
+    if (lang_matches (&lang_str[1], "h-hant-hk"))
+    {
+      /* Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "h-hant-mo"))
+    {
+      /* Chinese */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "h-min-nan"))
+    {
+      /* Minnan, Hokkien, Amoy, Taiwanese, Southern Min, Southern Fujian, Hoklo, Southern Fukien, Ho-lo */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "h-hans"))
+    {
+      /* Chinese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (lang_matches (&lang_str[1], "h-hant"))
+    {
+      /* Chinese */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    if (0 == strcmp (&lang_str[1], "h-min"))
+    {
+      /* Min, Fuzhou, Hokkien, Amoy, or Taiwanese */
+      tags[0] = HB_TAG('Z','H','S',' ');  /* Chinese Simplified */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "h-", 2)
+	&& subtag_matches (lang_str, limit, "-hk"))
+    {
+      /* Chinese; Hong Kong */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "h-", 2)
+	&& subtag_matches (lang_str, limit, "-mo"))
+    {
+      /* Chinese; Macao */
+      tags[0] = HB_TAG('Z','H','H',' ');  /* Chinese, Hong Kong SAR */
+      *count = 1;
+      return true;
+    }
+    if (0 == strncmp (&lang_str[1], "h-", 2)
+	&& subtag_matches (lang_str, limit, "-tw"))
+    {
+      /* Chinese; Taiwan, Province of China */
+      tags[0] = HB_TAG('Z','H','T',' ');  /* Chinese Traditional */
+      *count = 1;
+      return true;
+    }
+    break;
+  }
+  return false;
+}
+
+/**
+ * hb_ot_ambiguous_tag_to_language
+ * @tag: A language tag.
+ *
+ * Converts @tag to a BCP 47 language tag if it is ambiguous (it corresponds to
+ * many language tags) and the best tag is not the alphabetically first, or if
+ * the best tag consists of multiple subtags, or if the best tag does not appear
+ * in #ot_languages.
+ *
+ * Return value: The #hb_language_t corresponding to the BCP 47 language tag,
+ * or #HB_LANGUAGE_INVALID if @tag is not ambiguous.
+ **/
+static hb_language_t
+hb_ot_ambiguous_tag_to_language (hb_tag_t tag)
+{
+  switch (tag)
+  {
+  case HB_TAG('A','L','T',' '):  /* Altai */
+    return hb_language_from_string ("alt", -1);  /* Southern Altai */
+  case HB_TAG('A','P','P','H'):  /* Phonetic transcription—Americanist conventions */
+    return hb_language_from_string ("und-fonnapa", -1);  /* Undetermined; North American Phonetic Alphabet */
+  case HB_TAG('A','R','A',' '):  /* Arabic */
+    return hb_language_from_string ("ar", -1);  /* Arabic */
+  case HB_TAG('A','R','K',' '):  /* Rakhine */
+    return hb_language_from_string ("rki", -1);  /* Rakhine */
+  case HB_TAG('A','T','H',' '):  /* Athapaskan */
+    return hb_language_from_string ("ath", -1);  /* Athapascan */
+  case HB_TAG('B','I','K',' '):  /* Bikol */
+    return hb_language_from_string ("bik", -1);  /* Bikol */
+  case HB_TAG('C','P','P',' '):  /* Creoles */
+    return hb_language_from_string ("crp", -1);  /* Creoles and pidgins */
+  case HB_TAG('C','R','R',' '):  /* Carrier */
+    return hb_language_from_string ("crx", -1);  /* Carrier */
+  case HB_TAG('D','N','K',' '):  /* Dinka */
+    return hb_language_from_string ("din", -1);  /* Dinka */
+  case HB_TAG('D','R','I',' '):  /* Dari */
+    return hb_language_from_string ("prs", -1);  /* Dari */
+  case HB_TAG('D','Z','N',' '):  /* Dzongkha */
+    return hb_language_from_string ("dz", -1);  /* Dzongkha */
+  case HB_TAG('E','T','I',' '):  /* Estonian */
+    return hb_language_from_string ("et", -1);  /* Estonian */
+  case HB_TAG('G','O','N',' '):  /* Gondi */
+    return hb_language_from_string ("gon", -1);  /* Gondi */
+  case HB_TAG('H','M','N',' '):  /* Hmong */
+    return hb_language_from_string ("hmn", -1);  /* Hmong */
+  case HB_TAG('H','N','D',' '):  /* Hindko */
+    return hb_language_from_string ("hnd", -1);  /* Southern Hindko */
+  case HB_TAG('I','J','O',' '):  /* Ijo */
+    return hb_language_from_string ("ijo", -1);  /* Ijo */
+  case HB_TAG('I','N','U',' '):  /* Inuktitut */
+    return hb_language_from_string ("iu", -1);  /* Inuktitut */
+  case HB_TAG('I','P','K',' '):  /* Inupiat */
+    return hb_language_from_string ("ik", -1);  /* Inupiaq */
+  case HB_TAG('I','P','P','H'):  /* Phonetic transcription—IPA conventions */
+    return hb_language_from_string ("und-fonipa", -1);  /* Undetermined; International Phonetic Alphabet */
+  case HB_TAG('I','R','T',' '):  /* Irish Traditional */
+    return hb_language_from_string ("ga-Latg", -1);  /* Irish; Latin (Gaelic variant) */
+  case HB_TAG('J','I','I',' '):  /* Yiddish */
+    return hb_language_from_string ("yi", -1);  /* Yiddish */
+  case HB_TAG('K','A','L',' '):  /* Kalenjin */
+    return hb_language_from_string ("kln", -1);  /* Kalenjin */
+  case HB_TAG('K','G','E',' '):  /* Khutsuri Georgian */
+    return hb_language_from_string ("und-Geok", -1);  /* Undetermined; Khutsuri (Asomtavruli and Nuskhuri) */
+  case HB_TAG('K','N','R',' '):  /* Kanuri */
+    return hb_language_from_string ("kr", -1);  /* Kanuri */
+  case HB_TAG('K','O','K',' '):  /* Konkani */
+    return hb_language_from_string ("kok", -1);  /* Konkani */
+  case HB_TAG('K','U','I',' '):  /* Kui */
+    return hb_language_from_string ("uki", -1);  /* Kui (India) */
+  case HB_TAG('K','U','R',' '):  /* Kurdish */
+    return hb_language_from_string ("ku", -1);  /* Kurdish */
+  case HB_TAG('L','U','H',' '):  /* Luyia */
+    return hb_language_from_string ("luy", -1);  /* Luyia */
+  case HB_TAG('L','V','I',' '):  /* Latvian */
+    return hb_language_from_string ("lv", -1);  /* Latvian */
+  case HB_TAG('M','A','W',' '):  /* Marwari */
+    return hb_language_from_string ("mwr", -1);  /* Marwari */
+  case HB_TAG('M','L','G',' '):  /* Malagasy */
+    return hb_language_from_string ("mg", -1);  /* Malagasy */
+  case HB_TAG('M','L','Y',' '):  /* Malay */
+    return hb_language_from_string ("ms", -1);  /* Malay */
+  case HB_TAG('M','N','G',' '):  /* Mongolian */
+    return hb_language_from_string ("mn", -1);  /* Mongolian */
+  case HB_TAG('M','O','L',' '):  /* Moldavian */
+    return hb_language_from_string ("ro-MD", -1);  /* Romanian; Moldova */
+  case HB_TAG('N','E','P',' '):  /* Nepali */
+    return hb_language_from_string ("ne", -1);  /* Nepali */
+  case HB_TAG('N','I','S',' '):  /* Nisi */
+    return hb_language_from_string ("njz", -1);  /* Nyishi */
+  case HB_TAG('N','O','R',' '):  /* Norwegian */
+    return hb_language_from_string ("no", -1);  /* Norwegian */
+  case HB_TAG('O','J','B',' '):  /* Ojibway */
+    return hb_language_from_string ("oj", -1);  /* Ojibwa */
+  case HB_TAG('O','R','O',' '):  /* Oromo */
+    return hb_language_from_string ("om", -1);  /* Oromo */
+  case HB_TAG('P','A','S',' '):  /* Pashto */
+    return hb_language_from_string ("ps", -1);  /* Pashto */
+  case HB_TAG('P','G','R',' '):  /* Polytonic Greek */
+    return hb_language_from_string ("el-polyton", -1);  /* Modern Greek (1453-); Polytonic Greek */
+  case HB_TAG('P','R','O',' '):  /* Provençal / Old Provençal */
+    return hb_language_from_string ("pro", -1);  /* Old Provençal (to 1500) */
+  case HB_TAG('Q','U','H',' '):  /* Quechua (Bolivia) */
+    return hb_language_from_string ("quh", -1);  /* South Bolivian Quechua */
+  case HB_TAG('Q','V','I',' '):  /* Quechua (Ecuador) */
+    return hb_language_from_string ("qvi", -1);  /* Imbabura Highland Quichua */
+  case HB_TAG('Q','W','H',' '):  /* Quechua (Peru) */
+    return hb_language_from_string ("qwh", -1);  /* Huaylas Ancash Quechua */
+  case HB_TAG('R','A','J',' '):  /* Rajasthani */
+    return hb_language_from_string ("raj", -1);  /* Rajasthani */
+  case HB_TAG('R','O','Y',' '):  /* Romany */
+    return hb_language_from_string ("rom", -1);  /* Romany */
+  case HB_TAG('S','Q','I',' '):  /* Albanian */
+    return hb_language_from_string ("sq", -1);  /* Albanian */
+  case HB_TAG('S','Y','R',' '):  /* Syriac */
+    return hb_language_from_string ("syr", -1);  /* Syriac */
+  case HB_TAG('S','Y','R','E'):  /* Syriac, Estrangela script-variant (equivalent to ISO 15924 'Syre') */
+    return hb_language_from_string ("und-Syre", -1);  /* Undetermined; Syriac (Estrangelo variant) */
+  case HB_TAG('S','Y','R','J'):  /* Syriac, Western script-variant (equivalent to ISO 15924 'Syrj') */
+    return hb_language_from_string ("und-Syrj", -1);  /* Undetermined; Syriac (Western variant) */
+  case HB_TAG('S','Y','R','N'):  /* Syriac, Eastern script-variant (equivalent to ISO 15924 'Syrn') */
+    return hb_language_from_string ("und-Syrn", -1);  /* Undetermined; Syriac (Eastern variant) */
+  case HB_TAG('T','M','H',' '):  /* Tamashek */
+    return hb_language_from_string ("tmh", -1);  /* Tamashek */
+  case HB_TAG('T','N','E',' '):  /* Tundra Nenets */
+    return hb_language_from_string ("yrk", -1);  /* Nenets */
+  case HB_TAG('Z','H','H',' '):  /* Chinese, Hong Kong SAR */
+    return hb_language_from_string ("zh-HK", -1);  /* Chinese; Hong Kong */
+  case HB_TAG('Z','H','S',' '):  /* Chinese Simplified */
+    return hb_language_from_string ("zh-Hans", -1);  /* Chinese; Han (Simplified variant) */
+  case HB_TAG('Z','H','T',' '):  /* Chinese Traditional */
+    return hb_language_from_string ("zh-Hant", -1);  /* Chinese; Han (Traditional variant) */
+  default:
+    return HB_LANGUAGE_INVALID;
+  }
+}
+
+#endif /* HB_OT_TAG_TABLE_HH */
+
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-ot-tag.cc b/thirdparty/harfbuzz/src/hb-ot-tag.cc
new file mode 100644
index 0000000000..7ec91c5815
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-tag.cc
@@ -0,0 +1,567 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod, Roozbeh Pournader
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_OT_TAG
+
+
+/* hb_script_t */
+
+static hb_tag_t
+hb_ot_old_tag_from_script (hb_script_t script)
+{
+  /* This seems to be accurate as of end of 2012. */
+
+  switch ((hb_tag_t) script)
+  {
+    case HB_SCRIPT_INVALID:		return HB_OT_TAG_DEFAULT_SCRIPT;
+
+    /* KATAKANA and HIRAGANA both map to 'kana' */
+    case HB_SCRIPT_HIRAGANA:		return HB_TAG('k','a','n','a');
+
+    /* Spaces at the end are preserved, unlike ISO 15924 */
+    case HB_SCRIPT_LAO:			return HB_TAG('l','a','o',' ');
+    case HB_SCRIPT_YI:			return HB_TAG('y','i',' ',' ');
+    /* Unicode-5.0 additions */
+    case HB_SCRIPT_NKO:			return HB_TAG('n','k','o',' ');
+    /* Unicode-5.1 additions */
+    case HB_SCRIPT_VAI:			return HB_TAG('v','a','i',' ');
+  }
+
+  /* Else, just change first char to lowercase and return */
+  return ((hb_tag_t) script) | 0x20000000u;
+}
+
+static hb_script_t
+hb_ot_old_tag_to_script (hb_tag_t tag)
+{
+  if (unlikely (tag == HB_OT_TAG_DEFAULT_SCRIPT))
+    return HB_SCRIPT_INVALID;
+
+  /* This side of the conversion is fully algorithmic. */
+
+  /* Any spaces at the end of the tag are replaced by repeating the last
+   * letter.  Eg 'nko ' -> 'Nkoo' */
+  if (unlikely ((tag & 0x0000FF00u) == 0x00002000u))
+    tag |= (tag >> 8) & 0x0000FF00u; /* Copy second letter to third */
+  if (unlikely ((tag & 0x000000FFu) == 0x00000020u))
+    tag |= (tag >> 8) & 0x000000FFu; /* Copy third letter to fourth */
+
+  /* Change first char to uppercase and return */
+  return (hb_script_t) (tag & ~0x20000000u);
+}
+
+static hb_tag_t
+hb_ot_new_tag_from_script (hb_script_t script)
+{
+  switch ((hb_tag_t) script) {
+    case HB_SCRIPT_BENGALI:		return HB_TAG('b','n','g','2');
+    case HB_SCRIPT_DEVANAGARI:		return HB_TAG('d','e','v','2');
+    case HB_SCRIPT_GUJARATI:		return HB_TAG('g','j','r','2');
+    case HB_SCRIPT_GURMUKHI:		return HB_TAG('g','u','r','2');
+    case HB_SCRIPT_KANNADA:		return HB_TAG('k','n','d','2');
+    case HB_SCRIPT_MALAYALAM:		return HB_TAG('m','l','m','2');
+    case HB_SCRIPT_ORIYA:		return HB_TAG('o','r','y','2');
+    case HB_SCRIPT_TAMIL:		return HB_TAG('t','m','l','2');
+    case HB_SCRIPT_TELUGU:		return HB_TAG('t','e','l','2');
+    case HB_SCRIPT_MYANMAR:		return HB_TAG('m','y','m','2');
+  }
+
+  return HB_OT_TAG_DEFAULT_SCRIPT;
+}
+
+static hb_script_t
+hb_ot_new_tag_to_script (hb_tag_t tag)
+{
+  switch (tag) {
+    case HB_TAG('b','n','g','2'):	return HB_SCRIPT_BENGALI;
+    case HB_TAG('d','e','v','2'):	return HB_SCRIPT_DEVANAGARI;
+    case HB_TAG('g','j','r','2'):	return HB_SCRIPT_GUJARATI;
+    case HB_TAG('g','u','r','2'):	return HB_SCRIPT_GURMUKHI;
+    case HB_TAG('k','n','d','2'):	return HB_SCRIPT_KANNADA;
+    case HB_TAG('m','l','m','2'):	return HB_SCRIPT_MALAYALAM;
+    case HB_TAG('o','r','y','2'):	return HB_SCRIPT_ORIYA;
+    case HB_TAG('t','m','l','2'):	return HB_SCRIPT_TAMIL;
+    case HB_TAG('t','e','l','2'):	return HB_SCRIPT_TELUGU;
+    case HB_TAG('m','y','m','2'):	return HB_SCRIPT_MYANMAR;
+  }
+
+  return HB_SCRIPT_UNKNOWN;
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+void
+hb_ot_tags_from_script (hb_script_t  script,
+			hb_tag_t    *script_tag_1,
+			hb_tag_t    *script_tag_2)
+{
+  unsigned int count = 2;
+  hb_tag_t tags[2];
+  hb_ot_tags_from_script_and_language (script, HB_LANGUAGE_INVALID, &count, tags, nullptr, nullptr);
+  *script_tag_1 = count > 0 ? tags[0] : HB_OT_TAG_DEFAULT_SCRIPT;
+  *script_tag_2 = count > 1 ? tags[1] : HB_OT_TAG_DEFAULT_SCRIPT;
+}
+#endif
+
+/*
+ * Complete list at:
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/scripttags
+ *
+ * Most of the script tags are the same as the ISO 15924 tag but lowercased.
+ * So we just do that, and handle the exceptional cases in a switch.
+ */
+
+static void
+hb_ot_all_tags_from_script (hb_script_t   script,
+			    unsigned int *count /* IN/OUT */,
+			    hb_tag_t     *tags /* OUT */)
+{
+  unsigned int i = 0;
+
+  hb_tag_t new_tag = hb_ot_new_tag_from_script (script);
+  if (unlikely (new_tag != HB_OT_TAG_DEFAULT_SCRIPT))
+  {
+    /* HB_SCRIPT_MYANMAR maps to 'mym2', but there is no 'mym3'. */
+    if (new_tag != HB_TAG('m','y','m','2'))
+      tags[i++] = new_tag | '3';
+    if (*count > i)
+      tags[i++] = new_tag;
+  }
+
+  if (*count > i)
+  {
+    hb_tag_t old_tag = hb_ot_old_tag_from_script (script);
+    if (old_tag != HB_OT_TAG_DEFAULT_SCRIPT)
+      tags[i++] = old_tag;
+  }
+
+  *count = i;
+}
+
+hb_script_t
+hb_ot_tag_to_script (hb_tag_t tag)
+{
+  unsigned char digit = tag & 0x000000FFu;
+  if (unlikely (digit == '2' || digit == '3'))
+    return hb_ot_new_tag_to_script (tag & 0xFFFFFF32);
+
+  return hb_ot_old_tag_to_script (tag);
+}
+
+
+/* hb_language_t */
+
+static bool
+subtag_matches (const char *lang_str,
+		const char *limit,
+		const char *subtag)
+{
+  do {
+    const char *s = strstr (lang_str, subtag);
+    if (!s || s >= limit)
+      return false;
+    if (!ISALNUM (s[strlen (subtag)]))
+      return true;
+    lang_str = s + strlen (subtag);
+  } while (true);
+}
+
+static hb_bool_t
+lang_matches (const char *lang_str, const char *spec)
+{
+  unsigned int len = strlen (spec);
+
+  return strncmp (lang_str, spec, len) == 0 &&
+	 (lang_str[len] == '\0' || lang_str[len] == '-');
+}
+
+struct LangTag
+{
+  char language[4];
+  hb_tag_t tag;
+
+  int cmp (const char *a) const
+  {
+    const char *b = this->language;
+    unsigned int da, db;
+    const char *p;
+
+    p = strchr (a, '-');
+    da = p ? (unsigned int) (p - a) : strlen (a);
+
+    p = strchr (b, '-');
+    db = p ? (unsigned int) (p - b) : strlen (b);
+
+    return strncmp (a, b, hb_max (da, db));
+  }
+  int cmp (const LangTag *that) const
+  { return cmp (that->language); }
+};
+
+#include "hb-ot-tag-table.hh"
+
+/* The corresponding languages IDs for the following IDs are unclear,
+ * overlap, or are architecturally weird. Needs more research. */
+
+/*{"??",	{HB_TAG('B','C','R',' ')}},*/	/* Bible Cree */
+/*{"zh?",	{HB_TAG('C','H','N',' ')}},*/	/* Chinese (seen in Microsoft fonts) */
+/*{"ar-Syrc?",	{HB_TAG('G','A','R',' ')}},*/	/* Garshuni */
+/*{"??",	{HB_TAG('N','G','R',' ')}},*/	/* Nagari */
+/*{"??",	{HB_TAG('Y','I','C',' ')}},*/	/* Yi Classic */
+/*{"zh?",	{HB_TAG('Z','H','P',' ')}},*/	/* Chinese Phonetic */
+
+#ifndef HB_DISABLE_DEPRECATED
+hb_tag_t
+hb_ot_tag_from_language (hb_language_t language)
+{
+  unsigned int count = 1;
+  hb_tag_t tags[1];
+  hb_ot_tags_from_script_and_language (HB_SCRIPT_UNKNOWN, language, nullptr, nullptr, &count, tags);
+  return count > 0 ? tags[0] : HB_OT_TAG_DEFAULT_LANGUAGE;
+}
+#endif
+
+static void
+hb_ot_tags_from_language (const char   *lang_str,
+			  const char   *limit,
+			  unsigned int *count,
+			  hb_tag_t     *tags)
+{
+  const char *s;
+  unsigned int tag_idx;
+
+  /* Check for matches of multiple subtags. */
+  if (hb_ot_tags_from_complex_language (lang_str, limit, count, tags))
+    return;
+
+  /* Find a language matching in the first component. */
+  s = strchr (lang_str, '-');
+  {
+    if (s && limit - lang_str >= 6)
+    {
+      const char *extlang_end = strchr (s + 1, '-');
+      /* If there is an extended language tag, use it. */
+      if (3 == (extlang_end ? extlang_end - s - 1 : strlen (s + 1)) &&
+	  ISALPHA (s[1]))
+	lang_str = s + 1;
+    }
+    if (hb_sorted_array (ot_languages).bfind (lang_str, &tag_idx))
+    {
+      unsigned int i;
+      while (tag_idx != 0 &&
+	     0 == strcmp (ot_languages[tag_idx].language, ot_languages[tag_idx - 1].language))
+	tag_idx--;
+      for (i = 0;
+	   i < *count &&
+	   tag_idx + i < ARRAY_LENGTH (ot_languages) &&
+	   0 == strcmp (ot_languages[tag_idx + i].language, ot_languages[tag_idx].language);
+	   i++)
+	tags[i] = ot_languages[tag_idx + i].tag;
+      *count = i;
+      return;
+    }
+  }
+
+  if (!s)
+    s = lang_str + strlen (lang_str);
+  if (s - lang_str == 3) {
+    /* Assume it's ISO-639-3 and upper-case and use it. */
+    tags[0] = hb_tag_from_string (lang_str, s - lang_str) & ~0x20202000u;
+    *count = 1;
+    return;
+  }
+
+  *count = 0;
+}
+
+static bool
+parse_private_use_subtag (const char     *private_use_subtag,
+			  unsigned int   *count,
+			  hb_tag_t       *tags,
+			  const char     *prefix,
+			  unsigned char (*normalize) (unsigned char))
+{
+#ifdef HB_NO_LANGUAGE_PRIVATE_SUBTAG
+  return false;
+#endif
+
+  if (!(private_use_subtag && count && tags && *count)) return false;
+
+  const char *s = strstr (private_use_subtag, prefix);
+  if (!s) return false;
+
+  char tag[4];
+  int i;
+  s += strlen (prefix);
+  if (s[0] == '-') {
+    s += 1;
+    char c;
+    for (i = 0; i < 8 && ISHEX (s[i]); i++)
+    {
+      c = FROMHEX (s[i]);
+      if (i % 2 == 0)
+	tag[i / 2] = c << 4;
+      else
+	tag[i / 2] += c;
+    }
+    if (i != 8) return false;
+  } else {
+    for (i = 0; i < 4 && ISALNUM (s[i]); i++)
+      tag[i] = normalize (s[i]);
+    if (!i) return false;
+
+    for (; i < 4; i++)
+      tag[i] = ' ';
+  }
+  tags[0] = HB_TAG (tag[0], tag[1], tag[2], tag[3]);
+  if ((tags[0] & 0xDFDFDFDF) == HB_OT_TAG_DEFAULT_SCRIPT)
+    tags[0] ^= ~0xDFDFDFDF;
+  *count = 1;
+  return true;
+}
+
+/**
+ * hb_ot_tags_from_script_and_language:
+ * @script: an #hb_script_t to convert.
+ * @language: an #hb_language_t to convert.
+ * @script_count: (allow-none): maximum number of script tags to retrieve (IN)
+ * and actual number of script tags retrieved (OUT)
+ * @script_tags: (out) (allow-none): array of size at least @script_count to store the
+ * script tag results
+ * @language_count: (allow-none): maximum number of language tags to retrieve
+ * (IN) and actual number of language tags retrieved (OUT)
+ * @language_tags: (out) (allow-none): array of size at least @language_count to store
+ * the language tag results
+ *
+ * Converts an #hb_script_t and an #hb_language_t to script and language tags.
+ *
+ * Since: 2.0.0
+ **/
+void
+hb_ot_tags_from_script_and_language (hb_script_t   script,
+				     hb_language_t language,
+				     unsigned int *script_count /* IN/OUT */,
+				     hb_tag_t     *script_tags /* OUT */,
+				     unsigned int *language_count /* IN/OUT */,
+				     hb_tag_t     *language_tags /* OUT */)
+{
+  bool needs_script = true;
+
+  if (language == HB_LANGUAGE_INVALID)
+  {
+    if (language_count && language_tags && *language_count)
+      *language_count = 0;
+  }
+  else
+  {
+    const char *lang_str, *s, *limit, *private_use_subtag;
+    bool needs_language;
+
+    lang_str = hb_language_to_string (language);
+    limit = nullptr;
+    private_use_subtag = nullptr;
+    if (lang_str[0] == 'x' && lang_str[1] == '-')
+    {
+      private_use_subtag = lang_str;
+    } else {
+      for (s = lang_str + 1; *s; s++)
+      {
+	if (s[-1] == '-' && s[1] == '-')
+	{
+	  if (s[0] == 'x')
+	  {
+	    private_use_subtag = s;
+	    if (!limit)
+	      limit = s - 1;
+	    break;
+	  } else if (!limit)
+	  {
+	    limit = s - 1;
+	  }
+	}
+      }
+      if (!limit)
+	limit = s;
+    }
+
+    needs_script = !parse_private_use_subtag (private_use_subtag, script_count, script_tags, "-hbsc", TOLOWER);
+    needs_language = !parse_private_use_subtag (private_use_subtag, language_count, language_tags, "-hbot", TOUPPER);
+
+    if (needs_language && language_count && language_tags && *language_count)
+      hb_ot_tags_from_language (lang_str, limit, language_count, language_tags);
+  }
+
+  if (needs_script && script_count && script_tags && *script_count)
+    hb_ot_all_tags_from_script (script, script_count, script_tags);
+}
+
+/**
+ * hb_ot_tag_to_language:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+hb_language_t
+hb_ot_tag_to_language (hb_tag_t tag)
+{
+  unsigned int i;
+
+  if (tag == HB_OT_TAG_DEFAULT_LANGUAGE)
+    return nullptr;
+
+  {
+    hb_language_t disambiguated_tag = hb_ot_ambiguous_tag_to_language (tag);
+    if (disambiguated_tag != HB_LANGUAGE_INVALID)
+      return disambiguated_tag;
+  }
+
+  for (i = 0; i < ARRAY_LENGTH (ot_languages); i++)
+    if (ot_languages[i].tag == tag)
+      return hb_language_from_string (ot_languages[i].language, -1);
+
+  /* Return a custom language in the form of "x-hbot-AABBCCDD".
+   * If it's three letters long, also guess it's ISO 639-3 and lower-case and
+   * prepend it (if it's not a registered tag, the private use subtags will
+   * ensure that calling hb_ot_tag_from_language on the result will still return
+   * the same tag as the original tag).
+   */
+  {
+    char buf[20];
+    char *str = buf;
+    if (ISALPHA (tag >> 24)
+	&& ISALPHA ((tag >> 16) & 0xFF)
+	&& ISALPHA ((tag >> 8) & 0xFF)
+	&& (tag & 0xFF) == ' ')
+    {
+      buf[0] = TOLOWER (tag >> 24);
+      buf[1] = TOLOWER ((tag >> 16) & 0xFF);
+      buf[2] = TOLOWER ((tag >> 8) & 0xFF);
+      buf[3] = '-';
+      str += 4;
+    }
+    snprintf (str, 16, "x-hbot-%08x", tag);
+    return hb_language_from_string (&*buf, -1);
+  }
+}
+
+/**
+ * hb_ot_tags_to_script_and_language:
+ * @script_tag: a script tag
+ * @language_tag: a language tag
+ * @script: (allow-none): the #hb_script_t corresponding to @script_tag (OUT).
+ * @language: (allow-none): the #hb_language_t corresponding to @script_tag and
+ * @language_tag (OUT).
+ *
+ * Converts a script tag and a language tag to an #hb_script_t and an
+ * #hb_language_t.
+ *
+ * Since: 2.0.0
+ **/
+void
+hb_ot_tags_to_script_and_language (hb_tag_t       script_tag,
+				   hb_tag_t       language_tag,
+				   hb_script_t   *script /* OUT */,
+				   hb_language_t *language /* OUT */)
+{
+  hb_script_t script_out = hb_ot_tag_to_script (script_tag);
+  if (script)
+    *script = script_out;
+  if (language)
+  {
+    unsigned int script_count = 1;
+    hb_tag_t primary_script_tag[1];
+    hb_ot_tags_from_script_and_language (script_out,
+					 HB_LANGUAGE_INVALID,
+					 &script_count,
+					 primary_script_tag,
+					 nullptr, nullptr);
+    *language = hb_ot_tag_to_language (language_tag);
+    if (script_count == 0 || primary_script_tag[0] != script_tag)
+    {
+      unsigned char *buf;
+      const char *lang_str = hb_language_to_string (*language);
+      size_t len = strlen (lang_str);
+      buf = (unsigned char *) malloc (len + 16);
+      if (unlikely (!buf))
+      {
+	*language = nullptr;
+      }
+      else
+      {
+	int shift;
+	memcpy (buf, lang_str, len);
+	if (lang_str[0] != 'x' || lang_str[1] != '-') {
+	  buf[len++] = '-';
+	  buf[len++] = 'x';
+	}
+	buf[len++] = '-';
+	buf[len++] = 'h';
+	buf[len++] = 'b';
+	buf[len++] = 's';
+	buf[len++] = 'c';
+	buf[len++] = '-';
+	for (shift = 28; shift >= 0; shift -= 4)
+	  buf[len++] = TOHEX (script_tag >> shift);
+	*language = hb_language_from_string ((char *) buf, len);
+	free (buf);
+      }
+    }
+  }
+}
+
+#ifdef MAIN
+static inline void
+test_langs_sorted ()
+{
+  for (unsigned int i = 1; i < ARRAY_LENGTH (ot_languages); i++)
+  {
+    int c = ot_languages[i].cmp (&ot_languages[i - 1]);
+    if (c > 0)
+    {
+      fprintf (stderr, "ot_languages not sorted at index %d: %s %d %s\n",
+	       i, ot_languages[i-1].language, c, ot_languages[i].language);
+      abort();
+    }
+  }
+}
+
+int
+main ()
+{
+  test_langs_sorted ();
+  return 0;
+}
+
+#endif
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-var-avar-table.hh b/thirdparty/harfbuzz/src/hb-ot-var-avar-table.hh
new file mode 100644
index 0000000000..29219adb0a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var-avar-table.hh
@@ -0,0 +1,169 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_VAR_AVAR_TABLE_HH
+#define HB_OT_VAR_AVAR_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * avar -- Axis Variations
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/avar
+ */
+
+#define HB_OT_TAG_avar HB_TAG('a','v','a','r')
+
+
+namespace OT {
+
+
+struct AxisValueMap
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  F2DOT14	coords[2];
+//   F2DOT14	fromCoord;	/* A normalized coordinate value obtained using
+//				 * default normalization. */
+//   F2DOT14	toCoord;	/* The modified, normalized coordinate value. */
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct SegmentMaps : ArrayOf<AxisValueMap>
+{
+  int map (int value, unsigned int from_offset = 0, unsigned int to_offset = 1) const
+  {
+#define fromCoord coords[from_offset]
+#define toCoord coords[to_offset]
+    /* The following special-cases are not part of OpenType, which requires
+     * that at least -1, 0, and +1 must be mapped. But we include these as
+     * part of a better error recovery scheme. */
+    if (len < 2)
+    {
+      if (!len)
+	return value;
+      else /* len == 1*/
+	return value - arrayZ[0].fromCoord + arrayZ[0].toCoord;
+    }
+
+    if (value <= arrayZ[0].fromCoord)
+      return value - arrayZ[0].fromCoord + arrayZ[0].toCoord;
+
+    unsigned int i;
+    unsigned int count = len - 1;
+    for (i = 1; i < count && value > arrayZ[i].fromCoord; i++)
+      ;
+
+    if (value >= arrayZ[i].fromCoord)
+      return value - arrayZ[i].fromCoord + arrayZ[i].toCoord;
+
+    if (unlikely (arrayZ[i-1].fromCoord == arrayZ[i].fromCoord))
+      return arrayZ[i-1].toCoord;
+
+    int denom = arrayZ[i].fromCoord - arrayZ[i-1].fromCoord;
+    return roundf (arrayZ[i-1].toCoord + ((float) (arrayZ[i].toCoord - arrayZ[i-1].toCoord) *
+					  (value - arrayZ[i-1].fromCoord)) / denom);
+#undef toCoord
+#undef fromCoord
+  }
+
+  int unmap (int value) const { return map (value, 1, 0); }
+
+  public:
+  DEFINE_SIZE_ARRAY (2, *this);
+};
+
+struct avar
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_avar;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    if (unlikely (!(version.sanitize (c) &&
+		    version.major == 1 &&
+		    c->check_struct (this))))
+      return_trace (false);
+
+    const SegmentMaps *map = &firstAxisSegmentMaps;
+    unsigned int count = axisCount;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      if (unlikely (!map->sanitize (c)))
+	return_trace (false);
+      map = &StructAfter<SegmentMaps> (*map);
+    }
+
+    return_trace (true);
+  }
+
+  void map_coords (int *coords, unsigned int coords_length) const
+  {
+    unsigned int count = hb_min (coords_length, axisCount);
+
+    const SegmentMaps *map = &firstAxisSegmentMaps;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      coords[i] = map->map (coords[i]);
+      map = &StructAfter<SegmentMaps> (*map);
+    }
+  }
+
+  void unmap_coords (int *coords, unsigned int coords_length) const
+  {
+    unsigned int count = hb_min (coords_length, axisCount);
+
+    const SegmentMaps *map = &firstAxisSegmentMaps;
+    for (unsigned int i = 0; i < count; i++)
+    {
+      coords[i] = map->unmap (coords[i]);
+      map = &StructAfter<SegmentMaps> (*map);
+    }
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the avar table
+				 * initially set to 0x00010000u */
+  HBUINT16	reserved;	/* This field is permanently reserved. Set to 0. */
+  HBUINT16	axisCount;	/* The number of variation axes in the font. This
+				 * must be the same number as axisCount in the
+				 * 'fvar' table. */
+  SegmentMaps	firstAxisSegmentMaps;
+
+  public:
+  DEFINE_SIZE_MIN (8);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_VAR_AVAR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-var-fvar-table.hh b/thirdparty/harfbuzz/src/hb-ot-var-fvar-table.hh
new file mode 100644
index 0000000000..f9e933fb2b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var-fvar-table.hh
@@ -0,0 +1,327 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_VAR_FVAR_TABLE_HH
+#define HB_OT_VAR_FVAR_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * fvar -- Font Variations
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/fvar
+ */
+
+#define HB_OT_TAG_fvar HB_TAG('f','v','a','r')
+
+
+namespace OT {
+
+
+struct InstanceRecord
+{
+  friend struct fvar;
+
+  hb_array_t<const HBFixed> get_coordinates (unsigned int axis_count) const
+  { return coordinatesZ.as_array (axis_count); }
+
+  bool sanitize (hb_sanitize_context_t *c, unsigned int axis_count) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  c->check_array (coordinatesZ.arrayZ, axis_count));
+  }
+
+  protected:
+  NameID	subfamilyNameID;/* The name ID for entries in the 'name' table
+				 * that provide subfamily names for this instance. */
+  HBUINT16	flags;		/* Reserved for future use — set to 0. */
+  UnsizedArrayOf<HBFixed>
+		coordinatesZ;	/* The coordinates array for this instance. */
+  //NameID	postScriptNameIDX;/*Optional. The name ID for entries in the 'name'
+  //				  * table that provide PostScript names for this
+  //				  * instance. */
+
+  public:
+  DEFINE_SIZE_UNBOUNDED (4);
+};
+
+struct AxisRecord
+{
+  int cmp (hb_tag_t key) const { return axisTag.cmp (key); }
+
+  enum
+  {
+    AXIS_FLAG_HIDDEN	= 0x0001,
+  };
+
+#ifndef HB_DISABLE_DEPRECATED
+  void get_axis_deprecated (hb_ot_var_axis_t *info) const
+  {
+    info->tag = axisTag;
+    info->name_id = axisNameID;
+    get_coordinates (info->min_value, info->default_value, info->max_value);
+  }
+#endif
+
+  void get_axis_info (unsigned axis_index, hb_ot_var_axis_info_t *info) const
+  {
+    info->axis_index = axis_index;
+    info->tag = axisTag;
+    info->name_id = axisNameID;
+    info->flags = (hb_ot_var_axis_flags_t) (unsigned int) flags;
+    get_coordinates (info->min_value, info->default_value, info->max_value);
+    info->reserved = 0;
+  }
+
+  int normalize_axis_value (float v) const
+  {
+    float min_value, default_value, max_value;
+    get_coordinates (min_value, default_value, max_value);
+
+    v = hb_clamp (v, min_value, max_value);
+
+    if (v == default_value)
+      return 0;
+    else if (v < default_value)
+      v = (v - default_value) / (default_value - min_value);
+    else
+      v = (v - default_value) / (max_value - default_value);
+    return roundf (v * 16384.f);
+  }
+
+  float unnormalize_axis_value (int v) const
+  {
+    float min_value, default_value, max_value;
+    get_coordinates (min_value, default_value, max_value);
+
+    if (v == 0)
+      return default_value;
+    else if (v < 0)
+      return v * (default_value - min_value) / 16384.f + default_value;
+    else
+      return v * (max_value - default_value) / 16384.f + default_value;
+  }
+
+  hb_ot_name_id_t get_name_id () const { return axisNameID; }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  protected:
+  void get_coordinates (float &min, float &default_, float &max) const
+  {
+    default_ = defaultValue / 65536.f;
+    /* Ensure order, to simplify client math. */
+    min = hb_min (default_, minValue / 65536.f);
+    max = hb_max (default_, maxValue / 65536.f);
+  }
+
+  protected:
+  Tag		axisTag;	/* Tag identifying the design variation for the axis. */
+  HBFixed	minValue;	/* The minimum coordinate value for the axis. */
+  HBFixed	defaultValue;	/* The default coordinate value for the axis. */
+  HBFixed	maxValue;	/* The maximum coordinate value for the axis. */
+  HBUINT16	flags;		/* Axis flags. */
+  NameID	axisNameID;	/* The name ID for entries in the 'name' table that
+				 * provide a display name for this axis. */
+
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+struct fvar
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_fvar;
+
+  bool has_data () const { return version.to_int (); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  c->check_struct (this) &&
+		  axisSize == 20 && /* Assumed in our code. */
+		  instanceSize >= axisCount * 4 + 4 &&
+		  get_axes ().sanitize (c) &&
+		  c->check_range (get_instance (0), instanceCount, instanceSize));
+  }
+
+  unsigned int get_axis_count () const { return axisCount; }
+
+#ifndef HB_DISABLE_DEPRECATED
+  unsigned int get_axes_deprecated (unsigned int      start_offset,
+				    unsigned int     *axes_count /* IN/OUT */,
+				    hb_ot_var_axis_t *axes_array /* OUT */) const
+  {
+    if (axes_count)
+    {
+      hb_array_t<const AxisRecord> arr = get_axes ().sub_array (start_offset, axes_count);
+      for (unsigned i = 0; i < arr.length; ++i)
+	arr[i].get_axis_deprecated (&axes_array[i]);
+    }
+    return axisCount;
+  }
+#endif
+
+  unsigned int get_axis_infos (unsigned int           start_offset,
+			       unsigned int          *axes_count /* IN/OUT */,
+			       hb_ot_var_axis_info_t *axes_array /* OUT */) const
+  {
+    if (axes_count)
+    {
+      hb_array_t<const AxisRecord> arr = get_axes ().sub_array (start_offset, axes_count);
+      for (unsigned i = 0; i < arr.length; ++i)
+	arr[i].get_axis_info (start_offset + i, &axes_array[i]);
+    }
+    return axisCount;
+  }
+
+#ifndef HB_DISABLE_DEPRECATED
+  bool
+  find_axis_deprecated (hb_tag_t tag, unsigned *axis_index, hb_ot_var_axis_t *info) const
+  {
+    unsigned i;
+    if (!axis_index) axis_index = &i;
+    *axis_index = HB_OT_VAR_NO_AXIS_INDEX;
+    auto axes = get_axes ();
+    return axes.lfind (tag, axis_index) && (axes[*axis_index].get_axis_deprecated (info), true);
+  }
+#endif
+
+  bool
+  find_axis_info (hb_tag_t tag, hb_ot_var_axis_info_t *info) const
+  {
+    unsigned i;
+    auto axes = get_axes ();
+    return axes.lfind (tag, &i) && (axes[i].get_axis_info (i, info), true);
+  }
+
+  int normalize_axis_value (unsigned int axis_index, float v) const
+  { return get_axes ()[axis_index].normalize_axis_value (v); }
+
+  float unnormalize_axis_value (unsigned int axis_index, int v) const
+  { return get_axes ()[axis_index].unnormalize_axis_value (v); }
+
+  unsigned int get_instance_count () const { return instanceCount; }
+
+  hb_ot_name_id_t get_instance_subfamily_name_id (unsigned int instance_index) const
+  {
+    const InstanceRecord *instance = get_instance (instance_index);
+    if (unlikely (!instance)) return HB_OT_NAME_ID_INVALID;
+    return instance->subfamilyNameID;
+  }
+
+  hb_ot_name_id_t get_instance_postscript_name_id (unsigned int instance_index) const
+  {
+    const InstanceRecord *instance = get_instance (instance_index);
+    if (unlikely (!instance)) return HB_OT_NAME_ID_INVALID;
+    if (instanceSize >= axisCount * 4 + 6)
+      return StructAfter<NameID> (instance->get_coordinates (axisCount));
+    return HB_OT_NAME_ID_INVALID;
+  }
+
+  unsigned int get_instance_coords (unsigned int  instance_index,
+				    unsigned int *coords_length, /* IN/OUT */
+				    float        *coords         /* OUT */) const
+  {
+    const InstanceRecord *instance = get_instance (instance_index);
+    if (unlikely (!instance))
+    {
+      if (coords_length)
+	*coords_length = 0;
+      return 0;
+    }
+
+    if (coords_length && *coords_length)
+    {
+      hb_array_t<const HBFixed> instanceCoords = instance->get_coordinates (axisCount)
+							 .sub_array (0, *coords_length);
+      for (unsigned int i = 0; i < instanceCoords.length; i++)
+	coords[i] = instanceCoords.arrayZ[i].to_float ();
+    }
+    return axisCount;
+  }
+
+  void collect_name_ids (hb_set_t *nameids) const
+  {
+    if (!has_data ()) return;
+
+    + get_axes ()
+    | hb_map (&AxisRecord::get_name_id)
+    | hb_sink (nameids)
+    ;
+
+    + hb_range ((unsigned) instanceCount)
+    | hb_map ([this] (const unsigned _) { return get_instance_subfamily_name_id (_); })
+    | hb_sink (nameids)
+    ;
+
+    + hb_range ((unsigned) instanceCount)
+    | hb_map ([this] (const unsigned _) { return get_instance_postscript_name_id (_); })
+    | hb_sink (nameids)
+    ;
+  }
+
+  protected:
+  hb_array_t<const AxisRecord> get_axes () const
+  { return hb_array (&(this+firstAxis), axisCount); }
+
+  const InstanceRecord *get_instance (unsigned int i) const
+  {
+    if (unlikely (i >= instanceCount)) return nullptr;
+   return &StructAtOffset<InstanceRecord> (&StructAfter<InstanceRecord> (get_axes ()),
+					   i * instanceSize);
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the fvar table
+				 * initially set to 0x00010000u */
+  OffsetTo<AxisRecord>
+		firstAxis;	/* Offset in bytes from the beginning of the table
+				 * to the start of the AxisRecord array. */
+  HBUINT16	reserved;	/* This field is permanently reserved. Set to 2. */
+  HBUINT16	axisCount;	/* The number of variation axes in the font (the
+				 * number of records in the axes array). */
+  HBUINT16	axisSize;	/* The size in bytes of each VariationAxisRecord —
+				 * set to 20 (0x0014) for this version. */
+  HBUINT16	instanceCount;	/* The number of named instances defined in the font
+				 * (the number of records in the instances array). */
+  HBUINT16	instanceSize;	/* The size in bytes of each InstanceRecord — set
+				 * to either axisCount * sizeof(HBFixed) + 4, or to
+				 * axisCount * sizeof(HBFixed) + 6. */
+
+  public:
+  DEFINE_SIZE_STATIC (16);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_VAR_FVAR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-var-gvar-table.hh b/thirdparty/harfbuzz/src/hb-ot-var-gvar-table.hh
new file mode 100644
index 0000000000..4d4e6dcae4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var-gvar-table.hh
@@ -0,0 +1,701 @@
+/*
+ * Copyright © 2019  Adobe Inc.
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_VAR_GVAR_TABLE_HH
+#define HB_OT_VAR_GVAR_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * gvar -- Glyph Variation Table
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/gvar
+ */
+#define HB_OT_TAG_gvar HB_TAG('g','v','a','r')
+
+namespace OT {
+
+struct contour_point_t
+{
+  void init (float x_ = 0.f, float y_ = 0.f, bool is_end_point_ = false)
+  { flag = 0; x = x_; y = y_; is_end_point = is_end_point_; }
+
+  void translate (const contour_point_t &p) { x += p.x; y += p.y; }
+
+  uint8_t flag;
+  float x, y;
+  bool is_end_point;
+};
+
+struct contour_point_vector_t : hb_vector_t<contour_point_t>
+{
+  void extend (const hb_array_t<contour_point_t> &a)
+  {
+    unsigned int old_len = length;
+    resize (old_len + a.length);
+    for (unsigned int i = 0; i < a.length; i++)
+      (*this)[old_len + i] = a[i];
+  }
+
+  void transform (const float (&matrix)[4])
+  {
+    for (unsigned int i = 0; i < length; i++)
+    {
+      contour_point_t &p = (*this)[i];
+      float x_ = p.x * matrix[0] + p.y * matrix[2];
+	   p.y = p.x * matrix[1] + p.y * matrix[3];
+      p.x = x_;
+    }
+  }
+
+  void translate (const contour_point_t& delta)
+  {
+    for (unsigned int i = 0; i < length; i++)
+      (*this)[i].translate (delta);
+  }
+};
+
+/* https://docs.microsoft.com/en-us/typography/opentype/spec/otvarcommonformats#tuplevariationheader */
+struct TupleVariationHeader
+{
+  unsigned get_size (unsigned axis_count) const
+  { return min_size + get_all_tuples (axis_count).get_size (); }
+
+  unsigned get_data_size () const { return varDataSize; }
+
+  const TupleVariationHeader &get_next (unsigned axis_count) const
+  { return StructAtOffset<TupleVariationHeader> (this, get_size (axis_count)); }
+
+  float calculate_scalar (const int *coords, unsigned int coord_count,
+			  const hb_array_t<const F2DOT14> shared_tuples) const
+  {
+    hb_array_t<const F2DOT14> peak_tuple;
+
+    if (has_peak ())
+      peak_tuple = get_peak_tuple (coord_count);
+    else
+    {
+      unsigned int index = get_index ();
+      if (unlikely (index * coord_count >= shared_tuples.length))
+	return 0.f;
+      peak_tuple = shared_tuples.sub_array (coord_count * index, coord_count);
+    }
+
+    hb_array_t<const F2DOT14> start_tuple;
+    hb_array_t<const F2DOT14> end_tuple;
+    if (has_intermediate ())
+    {
+      start_tuple = get_start_tuple (coord_count);
+      end_tuple = get_end_tuple (coord_count);
+    }
+
+    float scalar = 1.f;
+    for (unsigned int i = 0; i < coord_count; i++)
+    {
+      int v = coords[i];
+      int peak = peak_tuple[i];
+      if (!peak || v == peak) continue;
+
+      if (has_intermediate ())
+      {
+	int start = start_tuple[i];
+	int end = end_tuple[i];
+	if (unlikely (start > peak || peak > end ||
+		      (start < 0 && end > 0 && peak))) continue;
+	if (v < start || v > end) return 0.f;
+	if (v < peak)
+	{ if (peak != start) scalar *= (float) (v - start) / (peak - start); }
+	else
+	{ if (peak != end) scalar *= (float) (end - v) / (end - peak); }
+      }
+      else if (!v || v < hb_min (0, peak) || v > hb_max (0, peak)) return 0.f;
+      else
+	scalar *= (float) v / peak;
+    }
+    return scalar;
+  }
+
+  bool           has_peak () const { return tupleIndex & TuppleIndex::EmbeddedPeakTuple; }
+  bool   has_intermediate () const { return tupleIndex & TuppleIndex::IntermediateRegion; }
+  bool has_private_points () const { return tupleIndex & TuppleIndex::PrivatePointNumbers; }
+  unsigned      get_index () const { return tupleIndex & TuppleIndex::TupleIndexMask; }
+
+  protected:
+  struct TuppleIndex : HBUINT16
+  {
+    enum Flags {
+      EmbeddedPeakTuple   = 0x8000u,
+      IntermediateRegion  = 0x4000u,
+      PrivatePointNumbers = 0x2000u,
+      TupleIndexMask      = 0x0FFFu
+    };
+
+    DEFINE_SIZE_STATIC (2);
+  };
+
+  hb_array_t<const F2DOT14> get_all_tuples (unsigned axis_count) const
+  { return StructAfter<UnsizedArrayOf<F2DOT14>> (tupleIndex).as_array ((has_peak () + has_intermediate () * 2) * axis_count); }
+  hb_array_t<const F2DOT14> get_peak_tuple (unsigned axis_count) const
+  { return get_all_tuples (axis_count).sub_array (0, axis_count); }
+  hb_array_t<const F2DOT14> get_start_tuple (unsigned axis_count) const
+  { return get_all_tuples (axis_count).sub_array (has_peak () * axis_count, axis_count); }
+  hb_array_t<const F2DOT14> get_end_tuple (unsigned axis_count) const
+  { return get_all_tuples (axis_count).sub_array (has_peak () * axis_count + axis_count, axis_count); }
+
+  HBUINT16	varDataSize;	/* The size in bytes of the serialized
+				 * data for this tuple variation table. */
+  TuppleIndex	tupleIndex;	/* A packed field. The high 4 bits are flags (see below).
+				   The low 12 bits are an index into a shared tuple
+				   records array. */
+  /* UnsizedArrayOf<F2DOT14> peakTuple - optional */
+				/* Peak tuple record for this tuple variation table — optional,
+				 * determined by flags in the tupleIndex value.
+				 *
+				 * Note that this must always be included in the 'cvar' table. */
+  /* UnsizedArrayOf<F2DOT14> intermediateStartTuple - optional */
+				/* Intermediate start tuple record for this tuple variation table — optional,
+				   determined by flags in the tupleIndex value. */
+  /* UnsizedArrayOf<F2DOT14> intermediateEndTuple - optional */
+				/* Intermediate end tuple record for this tuple variation table — optional,
+				 * determined by flags in the tupleIndex value. */
+  public:
+  DEFINE_SIZE_MIN (4);
+};
+
+struct GlyphVariationData
+{
+  const TupleVariationHeader &get_tuple_var_header (void) const
+  { return StructAfter<TupleVariationHeader> (data); }
+
+  struct tuple_iterator_t
+  {
+    void init (hb_bytes_t var_data_bytes_, unsigned int axis_count_)
+    {
+      var_data_bytes = var_data_bytes_;
+      var_data = var_data_bytes_.as<GlyphVariationData> ();
+      index = 0;
+      axis_count = axis_count_;
+      current_tuple = &var_data->get_tuple_var_header ();
+      data_offset = 0;
+    }
+
+    bool get_shared_indices (hb_vector_t<unsigned int> &shared_indices /* OUT */)
+    {
+      if (var_data->has_shared_point_numbers ())
+      {
+	const HBUINT8 *base = &(var_data+var_data->data);
+	const HBUINT8 *p = base;
+	if (!unpack_points (p, shared_indices, var_data_bytes)) return false;
+	data_offset = p - base;
+      }
+      return true;
+    }
+
+    bool is_valid () const
+    {
+      return (index < var_data->tupleVarCount.get_count ()) &&
+	     var_data_bytes.check_range (current_tuple, TupleVariationHeader::min_size) &&
+	     var_data_bytes.check_range (current_tuple, hb_max (current_tuple->get_data_size (), current_tuple->get_size (axis_count))) &&
+	     current_tuple->get_size (axis_count);
+    }
+
+    bool move_to_next ()
+    {
+      data_offset += current_tuple->get_data_size ();
+      current_tuple = &current_tuple->get_next (axis_count);
+      index++;
+      return is_valid ();
+    }
+
+    const HBUINT8 *get_serialized_data () const
+    { return &(var_data+var_data->data) + data_offset; }
+
+    private:
+    const GlyphVariationData *var_data;
+    unsigned int index;
+    unsigned int axis_count;
+    unsigned int data_offset;
+
+    public:
+    hb_bytes_t var_data_bytes;
+    const TupleVariationHeader *current_tuple;
+  };
+
+  static bool get_tuple_iterator (hb_bytes_t var_data_bytes, unsigned axis_count,
+				  hb_vector_t<unsigned int> &shared_indices /* OUT */,
+				  tuple_iterator_t *iterator /* OUT */)
+  {
+    iterator->init (var_data_bytes, axis_count);
+    if (!iterator->get_shared_indices (shared_indices))
+      return false;
+    return iterator->is_valid ();
+  }
+
+  bool has_shared_point_numbers () const { return tupleVarCount.has_shared_point_numbers (); }
+
+  static bool unpack_points (const HBUINT8 *&p /* IN/OUT */,
+			     hb_vector_t<unsigned int> &points /* OUT */,
+			     const hb_bytes_t &bytes)
+  {
+    enum packed_point_flag_t
+    {
+      POINTS_ARE_WORDS     = 0x80,
+      POINT_RUN_COUNT_MASK = 0x7F
+    };
+
+    if (unlikely (!bytes.check_range (p))) return false;
+
+    uint16_t count = *p++;
+    if (count & POINTS_ARE_WORDS)
+    {
+      if (unlikely (!bytes.check_range (p))) return false;
+      count = ((count & POINT_RUN_COUNT_MASK) << 8) | *p++;
+    }
+    points.resize (count);
+
+    unsigned int n = 0;
+    uint16_t i = 0;
+    while (i < count)
+    {
+      if (unlikely (!bytes.check_range (p))) return false;
+      uint16_t j;
+      uint8_t control = *p++;
+      uint16_t run_count = (control & POINT_RUN_COUNT_MASK) + 1;
+      if (control & POINTS_ARE_WORDS)
+      {
+	for (j = 0; j < run_count && i < count; j++, i++)
+	{
+	  if (unlikely (!bytes.check_range ((const HBUINT16 *) p)))
+	    return false;
+	  n += *(const HBUINT16 *)p;
+	  points[i] = n;
+	  p += HBUINT16::static_size;
+	}
+      }
+      else
+      {
+	for (j = 0; j < run_count && i < count; j++, i++)
+	{
+	  if (unlikely (!bytes.check_range (p))) return false;
+	  n += *p++;
+	  points[i] = n;
+	}
+      }
+      if (j < run_count) return false;
+    }
+    return true;
+  }
+
+  static bool unpack_deltas (const HBUINT8 *&p /* IN/OUT */,
+			     hb_vector_t<int> &deltas /* IN/OUT */,
+			     const hb_bytes_t &bytes)
+  {
+    enum packed_delta_flag_t
+    {
+      DELTAS_ARE_ZERO      = 0x80,
+      DELTAS_ARE_WORDS     = 0x40,
+      DELTA_RUN_COUNT_MASK = 0x3F
+    };
+
+    unsigned int i = 0;
+    unsigned int count = deltas.length;
+    while (i < count)
+    {
+      if (unlikely (!bytes.check_range (p))) return false;
+      uint8_t control = *p++;
+      unsigned int run_count = (control & DELTA_RUN_COUNT_MASK) + 1;
+      unsigned int j;
+      if (control & DELTAS_ARE_ZERO)
+	for (j = 0; j < run_count && i < count; j++, i++)
+	  deltas[i] = 0;
+      else if (control & DELTAS_ARE_WORDS)
+	for (j = 0; j < run_count && i < count; j++, i++)
+	{
+	  if (unlikely (!bytes.check_range ((const HBUINT16 *) p)))
+	    return false;
+	  deltas[i] = *(const HBINT16 *) p;
+	  p += HBUINT16::static_size;
+	}
+      else
+	for (j = 0; j < run_count && i < count; j++, i++)
+	{
+	  if (unlikely (!bytes.check_range (p)))
+	    return false;
+	  deltas[i] = *(const HBINT8 *) p++;
+	}
+      if (j < run_count)
+	return false;
+    }
+    return true;
+  }
+
+  bool has_data () const { return tupleVarCount; }
+
+  protected:
+  struct TupleVarCount : HBUINT16
+  {
+    bool has_shared_point_numbers () const { return ((*this) & SharedPointNumbers); }
+    unsigned int get_count () const { return (*this) & CountMask; }
+
+    protected:
+    enum Flags
+    {
+      SharedPointNumbers= 0x8000u,
+      CountMask		= 0x0FFFu
+    };
+    public:
+    DEFINE_SIZE_STATIC (2);
+  };
+
+  TupleVarCount	tupleVarCount;  /* A packed field. The high 4 bits are flags, and the
+				 * low 12 bits are the number of tuple variation tables
+				 * for this glyph. The number of tuple variation tables
+				 * can be any number between 1 and 4095. */
+  OffsetTo<HBUINT8>
+		data;		/* Offset from the start of the GlyphVariationData table
+				 * to the serialized data. */
+  /* TupleVariationHeader tupleVariationHeaders[] *//* Array of tuple variation headers. */
+  public:
+  DEFINE_SIZE_MIN (4);
+};
+
+struct gvar
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_gvar;
+
+  bool sanitize_shallow (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) && (version.major == 1) &&
+		  (glyphCount == c->get_num_glyphs ()) &&
+		  sharedTuples.sanitize (c, this, axisCount * sharedTupleCount) &&
+		  (is_long_offset () ?
+		     c->check_array (get_long_offset_array (), glyphCount+1) :
+		     c->check_array (get_short_offset_array (), glyphCount+1)) &&
+		  c->check_array (((const HBUINT8*)&(this+dataZ)) + get_offset (0),
+				  get_offset (glyphCount) - get_offset (0)));
+  }
+
+  /* GlyphVariationData not sanitized here; must be checked while accessing each glyph varation data */
+  bool sanitize (hb_sanitize_context_t *c) const
+  { return sanitize_shallow (c); }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+
+    gvar *out = c->serializer->allocate_min<gvar> ();
+    if (unlikely (!out)) return_trace (false);
+
+    out->version.major = 1;
+    out->version.minor = 0;
+    out->axisCount = axisCount;
+    out->sharedTupleCount = sharedTupleCount;
+
+    unsigned int num_glyphs = c->plan->num_output_glyphs ();
+    out->glyphCount = num_glyphs;
+
+    unsigned int subset_data_size = 0;
+    for (hb_codepoint_t gid = 0; gid < num_glyphs; gid++)
+    {
+      hb_codepoint_t old_gid;
+      if (!c->plan->old_gid_for_new_gid (gid, &old_gid)) continue;
+      subset_data_size += get_glyph_var_data_bytes (c->source_blob, old_gid).length;
+    }
+
+    bool long_offset = subset_data_size & ~0xFFFFu;
+    out->flags = long_offset ? 1 : 0;
+
+    HBUINT8 *subset_offsets = c->serializer->allocate_size<HBUINT8> ((long_offset ? 4 : 2) * (num_glyphs + 1));
+    if (!subset_offsets) return_trace (false);
+
+    /* shared tuples */
+    if (!sharedTupleCount || !sharedTuples)
+      out->sharedTuples = 0;
+    else
+    {
+      unsigned int shared_tuple_size = F2DOT14::static_size * axisCount * sharedTupleCount;
+      F2DOT14 *tuples = c->serializer->allocate_size<F2DOT14> (shared_tuple_size);
+      if (!tuples) return_trace (false);
+      out->sharedTuples = (char *) tuples - (char *) out;
+      memcpy (tuples, this+sharedTuples, shared_tuple_size);
+    }
+
+    char *subset_data = c->serializer->allocate_size<char> (subset_data_size);
+    if (!subset_data) return_trace (false);
+    out->dataZ = subset_data - (char *) out;
+
+    unsigned int glyph_offset = 0;
+    for (hb_codepoint_t gid = 0; gid < num_glyphs; gid++)
+    {
+      hb_codepoint_t old_gid;
+      hb_bytes_t var_data_bytes = c->plan->old_gid_for_new_gid (gid, &old_gid)
+				? get_glyph_var_data_bytes (c->source_blob, old_gid)
+				: hb_bytes_t ();
+
+      if (long_offset)
+	((HBUINT32 *) subset_offsets)[gid] = glyph_offset;
+      else
+	((HBUINT16 *) subset_offsets)[gid] = glyph_offset / 2;
+
+      if (var_data_bytes.length > 0)
+	memcpy (subset_data, var_data_bytes.arrayZ, var_data_bytes.length);
+      subset_data += var_data_bytes.length;
+      glyph_offset += var_data_bytes.length;
+    }
+    if (long_offset)
+      ((HBUINT32 *) subset_offsets)[num_glyphs] = glyph_offset;
+    else
+      ((HBUINT16 *) subset_offsets)[num_glyphs] = glyph_offset / 2;
+
+    return_trace (true);
+  }
+
+  protected:
+  const hb_bytes_t get_glyph_var_data_bytes (hb_blob_t *blob, hb_codepoint_t glyph) const
+  {
+    unsigned start_offset = get_offset (glyph);
+    unsigned length = get_offset (glyph+1) - start_offset;
+    hb_bytes_t var_data = blob->as_bytes ().sub_array (((unsigned) dataZ) + start_offset, length);
+    return likely (var_data.length >= GlyphVariationData::min_size) ? var_data : hb_bytes_t ();
+  }
+
+  bool is_long_offset () const { return flags & 1; }
+
+  unsigned get_offset (unsigned i) const
+  { return is_long_offset () ? get_long_offset_array ()[i] : get_short_offset_array ()[i] * 2; }
+
+  const HBUINT32 * get_long_offset_array () const { return (const HBUINT32 *) &offsetZ; }
+  const HBUINT16 *get_short_offset_array () const { return (const HBUINT16 *) &offsetZ; }
+
+  public:
+  struct accelerator_t
+  {
+    void init (hb_face_t *face)
+    { table = hb_sanitize_context_t ().reference_table<gvar> (face); }
+    void fini () { table.destroy (); }
+
+    private:
+    struct x_getter { static float get (const contour_point_t &p) { return p.x; } };
+    struct y_getter { static float get (const contour_point_t &p) { return p.y; } };
+
+    template <typename T>
+    static float infer_delta (const hb_array_t<contour_point_t> points,
+			      const hb_array_t<contour_point_t> deltas,
+			      unsigned int target, unsigned int prev, unsigned int next)
+    {
+      float target_val = T::get (points[target]);
+      float prev_val = T::get (points[prev]);
+      float next_val = T::get (points[next]);
+      float prev_delta = T::get (deltas[prev]);
+      float next_delta = T::get (deltas[next]);
+
+      if (prev_val == next_val)
+	return (prev_delta == next_delta) ? prev_delta : 0.f;
+      else if (target_val <= hb_min (prev_val, next_val))
+	return (prev_val < next_val) ? prev_delta : next_delta;
+      else if (target_val >= hb_max (prev_val, next_val))
+	return (prev_val > next_val) ? prev_delta : next_delta;
+
+      /* linear interpolation */
+      float r = (target_val - prev_val) / (next_val - prev_val);
+      return (1.f - r) * prev_delta + r * next_delta;
+    }
+
+    static unsigned int next_index (unsigned int i, unsigned int start, unsigned int end)
+    { return (i >= end) ? start : (i + 1); }
+
+    public:
+    bool apply_deltas_to_points (hb_codepoint_t glyph, hb_font_t *font,
+				 const hb_array_t<contour_point_t> points) const
+    {
+      /* num_coords should exactly match gvar's axisCount due to how GlyphVariationData tuples are aligned */
+      if (!font->num_coords || font->num_coords != table->axisCount) return true;
+
+      if (unlikely (glyph >= table->glyphCount)) return true;
+
+      hb_bytes_t var_data_bytes = table->get_glyph_var_data_bytes (table.get_blob (), glyph);
+      if (!var_data_bytes.as<GlyphVariationData> ()->has_data ()) return true;
+      hb_vector_t<unsigned int> shared_indices;
+      GlyphVariationData::tuple_iterator_t iterator;
+      if (!GlyphVariationData::get_tuple_iterator (var_data_bytes, table->axisCount,
+						   shared_indices, &iterator))
+	return true; /* so isn't applied at all */
+
+      /* Save original points for inferred delta calculation */
+      contour_point_vector_t orig_points;
+      orig_points.resize (points.length);
+      for (unsigned int i = 0; i < orig_points.length; i++)
+	orig_points[i] = points[i];
+
+      contour_point_vector_t deltas; /* flag is used to indicate referenced point */
+      deltas.resize (points.length);
+
+      hb_vector_t<unsigned> end_points;
+      for (unsigned i = 0; i < points.length; ++i)
+	if (points[i].is_end_point)
+	  end_points.push (i);
+
+      int *coords = font->coords;
+      unsigned num_coords = font->num_coords;
+      hb_array_t<const F2DOT14> shared_tuples = (table+table->sharedTuples).as_array (table->sharedTupleCount * table->axisCount);
+      do
+      {
+	float scalar = iterator.current_tuple->calculate_scalar (coords, num_coords, shared_tuples);
+	if (scalar == 0.f) continue;
+	const HBUINT8 *p = iterator.get_serialized_data ();
+	unsigned int length = iterator.current_tuple->get_data_size ();
+	if (unlikely (!iterator.var_data_bytes.check_range (p, length)))
+	  return false;
+
+	hb_bytes_t bytes ((const char *) p, length);
+	hb_vector_t<unsigned int> private_indices;
+	if (iterator.current_tuple->has_private_points () &&
+	    !GlyphVariationData::unpack_points (p, private_indices, bytes))
+	  return false;
+	const hb_array_t<unsigned int> &indices = private_indices.length ? private_indices : shared_indices;
+
+	bool apply_to_all = (indices.length == 0);
+	unsigned int num_deltas = apply_to_all ? points.length : indices.length;
+	hb_vector_t<int> x_deltas;
+	x_deltas.resize (num_deltas);
+	if (!GlyphVariationData::unpack_deltas (p, x_deltas, bytes))
+	  return false;
+	hb_vector_t<int> y_deltas;
+	y_deltas.resize (num_deltas);
+	if (!GlyphVariationData::unpack_deltas (p, y_deltas, bytes))
+	  return false;
+
+	for (unsigned int i = 0; i < deltas.length; i++)
+	  deltas[i].init ();
+	for (unsigned int i = 0; i < num_deltas; i++)
+	{
+	  unsigned int pt_index = apply_to_all ? i : indices[i];
+	  deltas[pt_index].flag = 1;	/* this point is referenced, i.e., explicit deltas specified */
+	  deltas[pt_index].x += x_deltas[i] * scalar;
+	  deltas[pt_index].y += y_deltas[i] * scalar;
+	}
+
+	/* infer deltas for unreferenced points */
+	unsigned start_point = 0;
+	for (unsigned c = 0; c < end_points.length; c++)
+	{
+	  unsigned end_point = end_points[c];
+
+	  /* Check the number of unreferenced points in a contour. If no unref points or no ref points, nothing to do. */
+	  unsigned unref_count = 0;
+	  for (unsigned i = start_point; i <= end_point; i++)
+	    if (!deltas[i].flag) unref_count++;
+
+	  unsigned j = start_point;
+	  if (unref_count == 0 || unref_count > end_point - start_point)
+	    goto no_more_gaps;
+
+	  for (;;)
+	  {
+	    /* Locate the next gap of unreferenced points between two referenced points prev and next.
+	     * Note that a gap may wrap around at left (start_point) and/or at right (end_point).
+	     */
+	    unsigned int prev, next, i;
+	    for (;;)
+	    {
+	      i = j;
+	      j = next_index (i, start_point, end_point);
+	      if (deltas[i].flag && !deltas[j].flag) break;
+	    }
+	    prev = j = i;
+	    for (;;)
+	    {
+	      i = j;
+	      j = next_index (i, start_point, end_point);
+	      if (!deltas[i].flag && deltas[j].flag) break;
+	    }
+	    next = j;
+	    /* Infer deltas for all unref points in the gap between prev and next */
+	    i = prev;
+	    for (;;)
+	    {
+	      i = next_index (i, start_point, end_point);
+	      if (i == next) break;
+	      deltas[i].x = infer_delta<x_getter> (orig_points.as_array (), deltas.as_array (), i, prev, next);
+	      deltas[i].y = infer_delta<y_getter> (orig_points.as_array (), deltas.as_array (), i, prev, next);
+	      if (--unref_count == 0) goto no_more_gaps;
+	    }
+	  }
+no_more_gaps:
+	  start_point = end_point + 1;
+	}
+
+	/* apply specified / inferred deltas to points */
+	for (unsigned int i = 0; i < points.length; i++)
+	{
+	  points[i].x += roundf (deltas[i].x);
+	  points[i].y += roundf (deltas[i].y);
+	}
+      } while (iterator.move_to_next ());
+
+      return true;
+    }
+
+    unsigned int get_axis_count () const { return table->axisCount; }
+
+    private:
+    hb_blob_ptr_t<gvar> table;
+  };
+
+  protected:
+  FixedVersion<>version;	/* Version number of the glyph variations table
+				 * Set to 0x00010000u. */
+  HBUINT16	axisCount;	/* The number of variation axes for this font. This must be
+				 * the same number as axisCount in the 'fvar' table. */
+  HBUINT16	sharedTupleCount;
+				/* The number of shared tuple records. Shared tuple records
+				 * can be referenced within glyph variation data tables for
+				 * multiple glyphs, as opposed to other tuple records stored
+				 * directly within a glyph variation data table. */
+  LNNOffsetTo<UnsizedArrayOf<F2DOT14>>
+		sharedTuples;	/* Offset from the start of this table to the shared tuple records.
+				 * Array of tuple records shared across all glyph variation data tables. */
+  HBUINT16	glyphCount;	/* The number of glyphs in this font. This must match the number of
+				 * glyphs stored elsewhere in the font. */
+  HBUINT16	flags;		/* Bit-field that gives the format of the offset array that follows.
+				 * If bit 0 is clear, the offsets are uint16; if bit 0 is set, the
+				 * offsets are uint32. */
+  LOffsetTo<GlyphVariationData>
+		dataZ;		/* Offset from the start of this table to the array of
+				 * GlyphVariationData tables. */
+  UnsizedArrayOf<HBUINT8>
+		offsetZ;	/* Offsets from the start of the GlyphVariationData array
+				 * to each GlyphVariationData table. */
+  public:
+  DEFINE_SIZE_MIN (20);
+};
+
+struct gvar_accelerator_t : gvar::accelerator_t {};
+
+} /* namespace OT */
+
+#endif /* HB_OT_VAR_GVAR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-var-hvar-table.hh b/thirdparty/harfbuzz/src/hb-ot-var-hvar-table.hh
new file mode 100644
index 0000000000..fdcc88d674
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var-hvar-table.hh
@@ -0,0 +1,488 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_VAR_HVAR_TABLE_HH
+#define HB_OT_VAR_HVAR_TABLE_HH
+
+#include "hb-ot-layout-common.hh"
+
+
+namespace OT {
+
+
+struct DeltaSetIndexMap
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  c->check_range (mapDataZ.arrayZ,
+				  mapCount,
+				  get_width ()));
+  }
+
+  template <typename T>
+  bool serialize (hb_serialize_context_t *c, const T &plan)
+  {
+    unsigned int width = plan.get_width ();
+    unsigned int inner_bit_count = plan.get_inner_bit_count ();
+    const hb_array_t<const unsigned int> output_map = plan.get_output_map ();
+
+    TRACE_SERIALIZE (this);
+    if (unlikely (output_map.length && ((((inner_bit_count-1)&~0xF)!=0) || (((width-1)&~0x3)!=0))))
+      return_trace (false);
+    if (unlikely (!c->extend_min (*this))) return_trace (false);
+
+    format = ((width-1)<<4)|(inner_bit_count-1);
+    mapCount = output_map.length;
+    HBUINT8 *p = c->allocate_size<HBUINT8> (width * output_map.length);
+    if (unlikely (!p)) return_trace (false);
+    for (unsigned int i = 0; i < output_map.length; i++)
+    {
+      unsigned int v = output_map[i];
+      unsigned int outer = v >> 16;
+      unsigned int inner = v & 0xFFFF;
+      unsigned int u = (outer << inner_bit_count) | inner;
+      for (unsigned int w = width; w > 0;)
+      {
+	p[--w] = u;
+	u >>= 8;
+      }
+      p += width;
+    }
+    return_trace (true);
+  }
+
+  unsigned int map (unsigned int v) const /* Returns 16.16 outer.inner. */
+  {
+    /* If count is zero, pass value unchanged.  This takes
+     * care of direct mapping for advance map. */
+    if (!mapCount)
+      return v;
+
+    if (v >= mapCount)
+      v = mapCount - 1;
+
+    unsigned int u = 0;
+    { /* Fetch it. */
+      unsigned int w = get_width ();
+      const HBUINT8 *p = mapDataZ.arrayZ + w * v;
+      for (; w; w--)
+	u = (u << 8) + *p++;
+    }
+
+    { /* Repack it. */
+      unsigned int n = get_inner_bit_count ();
+      unsigned int outer = u >> n;
+      unsigned int inner = u & ((1 << n) - 1);
+      u = (outer<<16) | inner;
+    }
+
+    return u;
+  }
+
+  unsigned int get_map_count () const	    { return mapCount; }
+  unsigned int get_width () const           { return ((format >> 4) & 3) + 1; }
+  unsigned int get_inner_bit_count () const { return (format & 0xF) + 1; }
+
+  protected:
+  HBUINT16	format;		/* A packed field that describes the compressed
+				 * representation of delta-set indices. */
+  HBUINT16	mapCount;	/* The number of mapping entries. */
+  UnsizedArrayOf<HBUINT8>
+		mapDataZ;	/* The delta-set index mapping data. */
+
+  public:
+  DEFINE_SIZE_ARRAY (4, mapDataZ);
+};
+
+struct index_map_subset_plan_t
+{
+  enum index_map_index_t {
+    ADV_INDEX,
+    LSB_INDEX,	/* dual as TSB */
+    RSB_INDEX,	/* dual as BSB */
+    VORG_INDEX
+  };
+
+  void init (const DeltaSetIndexMap  &index_map,
+	     hb_inc_bimap_t	     &outer_map,
+	     hb_vector_t<hb_set_t *> &inner_sets,
+	     const hb_subset_plan_t  *plan)
+  {
+    map_count = 0;
+    outer_bit_count = 0;
+    inner_bit_count = 1;
+    max_inners.init ();
+    output_map.init ();
+
+    if (&index_map == &Null (DeltaSetIndexMap)) return;
+
+    unsigned int	last_val = (unsigned int)-1;
+    hb_codepoint_t	last_gid = (hb_codepoint_t)-1;
+    hb_codepoint_t	gid = (hb_codepoint_t) hb_min (index_map.get_map_count (), plan->num_output_glyphs ());
+
+    outer_bit_count = (index_map.get_width () * 8) - index_map.get_inner_bit_count ();
+    max_inners.resize (inner_sets.length);
+    for (unsigned i = 0; i < inner_sets.length; i++) max_inners[i] = 0;
+
+    /* Search backwards for a map value different from the last map value */
+    for (; gid > 0; gid--)
+    {
+      hb_codepoint_t	old_gid;
+      if (!plan->old_gid_for_new_gid (gid - 1, &old_gid))
+      {
+	if (last_gid == (hb_codepoint_t) -1)
+	  continue;
+	else
+	  break;
+      }
+
+      unsigned int v = index_map.map (old_gid);
+      if (last_gid == (hb_codepoint_t) -1)
+      {
+	last_val = v;
+	last_gid = gid;
+	continue;
+      }
+      if (v != last_val) break;
+
+      last_gid = gid;
+    }
+
+    if (unlikely (last_gid == (hb_codepoint_t)-1)) return;
+    map_count = last_gid;
+    for (gid = 0; gid < map_count; gid++)
+    {
+      hb_codepoint_t	old_gid;
+      if (plan->old_gid_for_new_gid (gid, &old_gid))
+      {
+	unsigned int v = index_map.map (old_gid);
+	unsigned int outer = v >> 16;
+	unsigned int inner = v & 0xFFFF;
+	outer_map.add (outer);
+	if (inner > max_inners[outer]) max_inners[outer] = inner;
+	if (outer >= inner_sets.length) return;
+	inner_sets[outer]->add (inner);
+      }
+    }
+  }
+
+  void fini ()
+  {
+    max_inners.fini ();
+    output_map.fini ();
+  }
+
+  void remap (const DeltaSetIndexMap *input_map,
+	      const hb_inc_bimap_t &outer_map,
+	      const hb_vector_t<hb_inc_bimap_t> &inner_maps,
+	      const hb_subset_plan_t *plan)
+  {
+    if (input_map == &Null (DeltaSetIndexMap)) return;
+
+    for (unsigned int i = 0; i < max_inners.length; i++)
+    {
+      if (inner_maps[i].get_population () == 0) continue;
+      unsigned int bit_count = (max_inners[i]==0)? 1: hb_bit_storage (inner_maps[i][max_inners[i]]);
+      if (bit_count > inner_bit_count) inner_bit_count = bit_count;
+    }
+
+    output_map.resize (map_count);
+    for (hb_codepoint_t gid = 0; gid < output_map.length; gid++)
+    {
+      hb_codepoint_t	old_gid;
+      if (plan->old_gid_for_new_gid (gid, &old_gid))
+      {
+	unsigned int v = input_map->map (old_gid);
+	unsigned int outer = v >> 16;
+	output_map[gid] = (outer_map[outer] << 16) | (inner_maps[outer][v & 0xFFFF]);
+      }
+      else
+	output_map[gid] = 0;	/* Map unused glyph to outer/inner=0/0 */
+    }
+  }
+
+  unsigned int get_inner_bit_count () const { return inner_bit_count; }
+  unsigned int get_width ()           const { return ((outer_bit_count + inner_bit_count + 7) / 8); }
+  unsigned int get_map_count ()       const { return map_count; }
+
+  unsigned int get_size () const
+  { return (map_count? (DeltaSetIndexMap::min_size + get_width () * map_count): 0); }
+
+  bool is_identity () const { return get_output_map ().length == 0; }
+  hb_array_t<const unsigned int> get_output_map () const { return output_map.as_array (); }
+
+  protected:
+  unsigned int map_count;
+  hb_vector_t<unsigned int> max_inners;
+  unsigned int outer_bit_count;
+  unsigned int inner_bit_count;
+  hb_vector_t<unsigned int> output_map;
+};
+
+struct hvarvvar_subset_plan_t
+{
+  hvarvvar_subset_plan_t() : inner_maps (), index_map_plans () {}
+  ~hvarvvar_subset_plan_t() { fini (); }
+
+  void init (const hb_array_t<const DeltaSetIndexMap *> &index_maps,
+	     const VariationStore &_var_store,
+	     const hb_subset_plan_t *plan)
+  {
+    index_map_plans.resize (index_maps.length);
+
+    var_store = &_var_store;
+    inner_sets.resize (var_store->get_sub_table_count ());
+    for (unsigned int i = 0; i < inner_sets.length; i++)
+      inner_sets[i] = hb_set_create ();
+    adv_set = hb_set_create ();
+
+    inner_maps.resize (var_store->get_sub_table_count ());
+
+    for (unsigned int i = 0; i < inner_maps.length; i++)
+      inner_maps[i].init ();
+
+    if (unlikely (!index_map_plans.length || !inner_sets.length || !inner_maps.length)) return;
+
+    bool retain_adv_map = false;
+    index_map_plans[0].init (*index_maps[0], outer_map, inner_sets, plan);
+    if (index_maps[0] == &Null (DeltaSetIndexMap))
+    {
+      retain_adv_map = plan->retain_gids;
+      outer_map.add (0);
+      for (hb_codepoint_t gid = 0; gid < plan->num_output_glyphs (); gid++)
+      {
+	hb_codepoint_t old_gid;
+	if (plan->old_gid_for_new_gid (gid, &old_gid))
+	  inner_sets[0]->add (old_gid);
+      }
+      hb_set_union (adv_set, inner_sets[0]);
+    }
+
+    for (unsigned int i = 1; i < index_maps.length; i++)
+      index_map_plans[i].init (*index_maps[i], outer_map, inner_sets, plan);
+
+    outer_map.sort ();
+
+    if (retain_adv_map)
+    {
+      for (hb_codepoint_t gid = 0; gid < plan->num_output_glyphs (); gid++)
+	if (inner_sets[0]->has (gid))
+	  inner_maps[0].add (gid);
+	else
+	  inner_maps[0].skip ();
+    }
+    else
+    {
+      inner_maps[0].add_set (adv_set);
+      hb_set_subtract (inner_sets[0], adv_set);
+      inner_maps[0].add_set (inner_sets[0]);
+    }
+
+    for (unsigned int i = 1; i < inner_maps.length; i++)
+      inner_maps[i].add_set (inner_sets[i]);
+
+    for (unsigned int i = 0; i < index_maps.length; i++)
+      index_map_plans[i].remap (index_maps[i], outer_map, inner_maps, plan);
+  }
+
+  void fini ()
+  {
+    for (unsigned int i = 0; i < inner_sets.length; i++)
+      hb_set_destroy (inner_sets[i]);
+    hb_set_destroy (adv_set);
+    inner_maps.fini_deep ();
+    index_map_plans.fini_deep ();
+  }
+
+  hb_inc_bimap_t outer_map;
+  hb_vector_t<hb_inc_bimap_t> inner_maps;
+  hb_vector_t<index_map_subset_plan_t> index_map_plans;
+  const VariationStore *var_store;
+
+  protected:
+  hb_vector_t<hb_set_t *> inner_sets;
+  hb_set_t *adv_set;
+};
+
+/*
+ * HVAR -- Horizontal Metrics Variations
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/hvar
+ * VVAR -- Vertical Metrics Variations
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/vvar
+ */
+#define HB_OT_TAG_HVAR HB_TAG('H','V','A','R')
+#define HB_OT_TAG_VVAR HB_TAG('V','V','A','R')
+
+struct HVARVVAR
+{
+  static constexpr hb_tag_t HVARTag = HB_OT_TAG_HVAR;
+  static constexpr hb_tag_t VVARTag = HB_OT_TAG_VVAR;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  varStore.sanitize (c, this) &&
+		  advMap.sanitize (c, this) &&
+		  lsbMap.sanitize (c, this) &&
+		  rsbMap.sanitize (c, this));
+  }
+
+  void listup_index_maps (hb_vector_t<const DeltaSetIndexMap *> &index_maps) const
+  {
+    index_maps.push (&(this+advMap));
+    index_maps.push (&(this+lsbMap));
+    index_maps.push (&(this+rsbMap));
+  }
+
+  bool serialize_index_maps (hb_serialize_context_t *c,
+			     const hb_array_t<index_map_subset_plan_t> &im_plans)
+  {
+    TRACE_SERIALIZE (this);
+    if (im_plans[index_map_subset_plan_t::ADV_INDEX].is_identity ())
+      advMap = 0;
+    else if (unlikely (!advMap.serialize (c, this).serialize (c, im_plans[index_map_subset_plan_t::ADV_INDEX])))
+      return_trace (false);
+    if (im_plans[index_map_subset_plan_t::LSB_INDEX].is_identity ())
+      lsbMap = 0;
+    else if (unlikely (!lsbMap.serialize (c, this).serialize (c, im_plans[index_map_subset_plan_t::LSB_INDEX])))
+      return_trace (false);
+    if (im_plans[index_map_subset_plan_t::RSB_INDEX].is_identity ())
+      rsbMap = 0;
+    else if (unlikely (!rsbMap.serialize (c, this).serialize (c, im_plans[index_map_subset_plan_t::RSB_INDEX])))
+      return_trace (false);
+
+    return_trace (true);
+  }
+
+  template <typename T>
+  bool _subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    hvarvvar_subset_plan_t	hvar_plan;
+    hb_vector_t<const DeltaSetIndexMap *>
+				index_maps;
+
+    ((T*)this)->listup_index_maps (index_maps);
+    hvar_plan.init (index_maps.as_array (), this+varStore, c->plan);
+
+    T *out = c->serializer->allocate_min<T> ();
+    if (unlikely (!out)) return_trace (false);
+
+    out->version.major = 1;
+    out->version.minor = 0;
+
+    if (unlikely (!out->varStore.serialize (c->serializer, out)
+		     .serialize (c->serializer, hvar_plan.var_store, hvar_plan.inner_maps.as_array ())))
+      return_trace (false);
+
+    return_trace (out->T::serialize_index_maps (c->serializer,
+						hvar_plan.index_map_plans.as_array ()));
+  }
+
+  float get_advance_var (hb_codepoint_t glyph, hb_font_t *font) const
+  {
+    unsigned int varidx = (this+advMap).map (glyph);
+    return (this+varStore).get_delta (varidx, font->coords, font->num_coords);
+  }
+
+  float get_side_bearing_var (hb_codepoint_t glyph,
+			      const int *coords, unsigned int coord_count) const
+  {
+    if (!has_side_bearing_deltas ()) return 0.f;
+    unsigned int varidx = (this+lsbMap).map (glyph);
+    return (this+varStore).get_delta (varidx, coords, coord_count);
+  }
+
+  bool has_side_bearing_deltas () const { return lsbMap && rsbMap; }
+
+  protected:
+  FixedVersion<>version;	/* Version of the metrics variation table
+				 * initially set to 0x00010000u */
+  LOffsetTo<VariationStore>
+		varStore;	/* Offset to item variation store table. */
+  LOffsetTo<DeltaSetIndexMap>
+		advMap;		/* Offset to advance var-idx mapping. */
+  LOffsetTo<DeltaSetIndexMap>
+		lsbMap;		/* Offset to lsb/tsb var-idx mapping. */
+  LOffsetTo<DeltaSetIndexMap>
+		rsbMap;		/* Offset to rsb/bsb var-idx mapping. */
+
+  public:
+  DEFINE_SIZE_STATIC (20);
+};
+
+struct HVAR : HVARVVAR {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_HVAR;
+  bool subset (hb_subset_context_t *c) const { return HVARVVAR::_subset<HVAR> (c); }
+};
+struct VVAR : HVARVVAR {
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_VVAR;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (static_cast<const HVARVVAR *> (this)->sanitize (c) &&
+		  vorgMap.sanitize (c, this));
+  }
+
+  void listup_index_maps (hb_vector_t<const DeltaSetIndexMap *> &index_maps) const
+  {
+    HVARVVAR::listup_index_maps (index_maps);
+    index_maps.push (&(this+vorgMap));
+  }
+
+  bool serialize_index_maps (hb_serialize_context_t *c,
+			     const hb_array_t<index_map_subset_plan_t> &im_plans)
+  {
+    TRACE_SERIALIZE (this);
+    if (unlikely (!HVARVVAR::serialize_index_maps (c, im_plans)))
+      return_trace (false);
+    if (!im_plans[index_map_subset_plan_t::VORG_INDEX].get_map_count ())
+      vorgMap = 0;
+    else if (unlikely (!vorgMap.serialize (c, this).serialize (c, im_plans[index_map_subset_plan_t::VORG_INDEX])))
+      return_trace (false);
+
+    return_trace (true);
+  }
+
+  bool subset (hb_subset_context_t *c) const { return HVARVVAR::_subset<VVAR> (c); }
+
+  protected:
+  LOffsetTo<DeltaSetIndexMap>
+		vorgMap;	/* Offset to vertical-origin var-idx mapping. */
+
+  public:
+  DEFINE_SIZE_STATIC (24);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_VAR_HVAR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-var-mvar-table.hh b/thirdparty/harfbuzz/src/hb-ot-var-mvar-table.hh
new file mode 100644
index 0000000000..1b7fad9cec
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var-mvar-table.hh
@@ -0,0 +1,119 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_VAR_MVAR_TABLE_HH
+#define HB_OT_VAR_MVAR_TABLE_HH
+
+#include "hb-ot-layout-common.hh"
+
+
+namespace OT {
+
+
+struct VariationValueRecord
+{
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  Tag		valueTag;	/* Four-byte tag identifying a font-wide measure. */
+  HBUINT32		varIdx;		/* Outer/inner index into VariationStore item. */
+
+  public:
+  DEFINE_SIZE_STATIC (8);
+};
+
+
+/*
+ * MVAR -- Metrics Variations
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/mvar
+ */
+#define HB_OT_TAG_MVAR HB_TAG('M','V','A','R')
+
+struct MVAR
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_MVAR;
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (version.sanitize (c) &&
+		  likely (version.major == 1) &&
+		  c->check_struct (this) &&
+		  valueRecordSize >= VariationValueRecord::static_size &&
+		  varStore.sanitize (c, this) &&
+		  c->check_range (valuesZ.arrayZ,
+				  valueRecordCount,
+				  valueRecordSize));
+  }
+
+  float get_var (hb_tag_t tag,
+		 const int *coords, unsigned int coord_count) const
+  {
+    const VariationValueRecord *record;
+    record = (VariationValueRecord *) hb_bsearch (tag,
+						  (const VariationValueRecord *)
+						    (const HBUINT8 *) valuesZ,
+						  valueRecordCount, valueRecordSize,
+						  tag_compare);
+    if (!record)
+      return 0.;
+
+    return (this+varStore).get_delta (record->varIdx, coords, coord_count);
+  }
+
+protected:
+  static int tag_compare (const void *pa, const void *pb)
+  {
+    const hb_tag_t *a = (const hb_tag_t *) pa;
+    const Tag *b = (const Tag *) pb;
+    return b->cmp (*a);
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of the metrics variation table
+				 * initially set to 0x00010000u */
+  HBUINT16	reserved;	/* Not used; set to 0. */
+  HBUINT16	valueRecordSize;/* The size in bytes of each value record —
+				 * must be greater than zero. */
+  HBUINT16	valueRecordCount;/* The number of value records — may be zero. */
+  OffsetTo<VariationStore>
+		varStore;	/* Offset to item variation store table. */
+  UnsizedArrayOf<HBUINT8>
+		valuesZ;	/* Array of value records. The records must be
+				 * in binary order of their valueTag field. */
+
+  public:
+  DEFINE_SIZE_ARRAY (12, valuesZ);
+};
+
+} /* namespace OT */
+
+
+#endif /* HB_OT_VAR_MVAR_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot-var.cc b/thirdparty/harfbuzz/src/hb-ot-var.cc
new file mode 100644
index 0000000000..6b8b09b6ba
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var.cc
@@ -0,0 +1,220 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_VAR
+
+#include "hb-ot-var.h"
+
+#include "hb-ot-var-avar-table.hh"
+#include "hb-ot-var-fvar-table.hh"
+#include "hb-ot-var-mvar-table.hh"
+
+
+/**
+ * SECTION:hb-ot-var
+ * @title: hb-ot-var
+ * @short_description: OpenType Font Variations
+ * @include: hb-ot.h
+ *
+ * Functions for fetching information about OpenType Variable Fonts.
+ **/
+
+
+/*
+ * fvar/avar
+ */
+
+
+/**
+ * hb_ot_var_has_data:
+ * @face: #hb_face_t to test
+ *
+ * This function allows to verify the presence of OpenType variation data on the face.
+ *
+ * Return value: true if face has a `fvar' table and false otherwise
+ *
+ * Since: 1.4.2
+ **/
+hb_bool_t
+hb_ot_var_has_data (hb_face_t *face)
+{
+  return face->table.fvar->has_data ();
+}
+
+/**
+ * hb_ot_var_get_axis_count:
+ *
+ * Since: 1.4.2
+ **/
+unsigned int
+hb_ot_var_get_axis_count (hb_face_t *face)
+{
+  return face->table.fvar->get_axis_count ();
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_ot_var_get_axes:
+ *
+ * Since: 1.4.2
+ * Deprecated: 2.2.0
+ **/
+unsigned int
+hb_ot_var_get_axes (hb_face_t        *face,
+		    unsigned int      start_offset,
+		    unsigned int     *axes_count /* IN/OUT */,
+		    hb_ot_var_axis_t *axes_array /* OUT */)
+{
+  return face->table.fvar->get_axes_deprecated (start_offset, axes_count, axes_array);
+}
+
+/**
+ * hb_ot_var_find_axis:
+ *
+ * Since: 1.4.2
+ * Deprecated: 2.2.0
+ **/
+hb_bool_t
+hb_ot_var_find_axis (hb_face_t        *face,
+		     hb_tag_t          axis_tag,
+		     unsigned int     *axis_index,
+		     hb_ot_var_axis_t *axis_info)
+{
+  return face->table.fvar->find_axis_deprecated (axis_tag, axis_index, axis_info);
+}
+#endif
+
+/**
+ * hb_ot_var_get_axis_infos:
+ *
+ * Since: 2.2.0
+ **/
+HB_EXTERN unsigned int
+hb_ot_var_get_axis_infos (hb_face_t             *face,
+			  unsigned int           start_offset,
+			  unsigned int          *axes_count /* IN/OUT */,
+			  hb_ot_var_axis_info_t *axes_array /* OUT */)
+{
+  return face->table.fvar->get_axis_infos (start_offset, axes_count, axes_array);
+}
+
+/**
+ * hb_ot_var_find_axis_info:
+ *
+ * Since: 2.2.0
+ **/
+HB_EXTERN hb_bool_t
+hb_ot_var_find_axis_info (hb_face_t             *face,
+			  hb_tag_t               axis_tag,
+			  hb_ot_var_axis_info_t *axis_info)
+{
+  return face->table.fvar->find_axis_info (axis_tag, axis_info);
+}
+
+
+/*
+ * Named instances.
+ */
+
+unsigned int
+hb_ot_var_get_named_instance_count (hb_face_t *face)
+{
+  return face->table.fvar->get_instance_count ();
+}
+
+hb_ot_name_id_t
+hb_ot_var_named_instance_get_subfamily_name_id (hb_face_t   *face,
+						unsigned int instance_index)
+{
+  return face->table.fvar->get_instance_subfamily_name_id (instance_index);
+}
+
+hb_ot_name_id_t
+hb_ot_var_named_instance_get_postscript_name_id (hb_face_t  *face,
+						unsigned int instance_index)
+{
+  return face->table.fvar->get_instance_postscript_name_id (instance_index);
+}
+
+unsigned int
+hb_ot_var_named_instance_get_design_coords (hb_face_t    *face,
+					    unsigned int  instance_index,
+					    unsigned int *coords_length, /* IN/OUT */
+					    float        *coords         /* OUT */)
+{
+  return face->table.fvar->get_instance_coords (instance_index, coords_length, coords);
+}
+
+
+/**
+ * hb_ot_var_normalize_variations:
+ *
+ * Since: 1.4.2
+ **/
+void
+hb_ot_var_normalize_variations (hb_face_t            *face,
+				const hb_variation_t *variations, /* IN */
+				unsigned int          variations_length,
+				int                  *coords, /* OUT */
+				unsigned int          coords_length)
+{
+  for (unsigned int i = 0; i < coords_length; i++)
+    coords[i] = 0;
+
+  const OT::fvar &fvar = *face->table.fvar;
+  for (unsigned int i = 0; i < variations_length; i++)
+  {
+    hb_ot_var_axis_info_t info;
+    if (hb_ot_var_find_axis_info (face, variations[i].tag, &info) &&
+	info.axis_index < coords_length)
+      coords[info.axis_index] = fvar.normalize_axis_value (info.axis_index, variations[i].value);
+  }
+
+  face->table.avar->map_coords (coords, coords_length);
+}
+
+/**
+ * hb_ot_var_normalize_coords:
+ *
+ * Since: 1.4.2
+ **/
+void
+hb_ot_var_normalize_coords (hb_face_t    *face,
+			    unsigned int coords_length,
+			    const float *design_coords, /* IN */
+			    int *normalized_coords /* OUT */)
+{
+  const OT::fvar &fvar = *face->table.fvar;
+  for (unsigned int i = 0; i < coords_length; i++)
+    normalized_coords[i] = fvar.normalize_axis_value (i, design_coords[i]);
+
+  face->table.avar->map_coords (normalized_coords, coords_length);
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-ot-var.h b/thirdparty/harfbuzz/src/hb-ot-var.h
new file mode 100644
index 0000000000..df89bc5a23
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-var.h
@@ -0,0 +1,146 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H_IN
+#error "Include <hb-ot.h> instead."
+#endif
+
+#ifndef HB_OT_VAR_H
+#define HB_OT_VAR_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+
+#define HB_OT_TAG_VAR_AXIS_ITALIC	HB_TAG('i','t','a','l')
+#define HB_OT_TAG_VAR_AXIS_OPTICAL_SIZE	HB_TAG('o','p','s','z')
+#define HB_OT_TAG_VAR_AXIS_SLANT	HB_TAG('s','l','n','t')
+#define HB_OT_TAG_VAR_AXIS_WIDTH	HB_TAG('w','d','t','h')
+#define HB_OT_TAG_VAR_AXIS_WEIGHT	HB_TAG('w','g','h','t')
+
+
+/*
+ * fvar / avar
+ */
+
+HB_EXTERN hb_bool_t
+hb_ot_var_has_data (hb_face_t *face);
+
+
+/*
+ * Variation axes.
+ */
+
+
+HB_EXTERN unsigned int
+hb_ot_var_get_axis_count (hb_face_t *face);
+
+/**
+ * hb_ot_var_axis_flags_t:
+ * @HB_OT_VAR_AXIS_FLAG_HIDDEN: The axis should not be exposed directly in user interfaces.
+ *
+ * Since: 2.2.0
+ */
+typedef enum { /*< flags >*/
+  HB_OT_VAR_AXIS_FLAG_HIDDEN	= 0x00000001u,
+
+  _HB_OT_VAR_AXIS_FLAG_MAX_VALUE= HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_ot_var_axis_flags_t;
+
+/**
+ * hb_ot_var_axis_info_t:
+ *
+ * Since: 2.2.0
+ */
+typedef struct hb_ot_var_axis_info_t
+{
+  unsigned int			axis_index;
+  hb_tag_t			tag;
+  hb_ot_name_id_t		name_id;
+  hb_ot_var_axis_flags_t	flags;
+  float				min_value;
+  float				default_value;
+  float				max_value;
+  /*< private >*/
+  unsigned int			reserved;
+} hb_ot_var_axis_info_t;
+
+HB_EXTERN unsigned int
+hb_ot_var_get_axis_infos (hb_face_t             *face,
+			  unsigned int           start_offset,
+			  unsigned int          *axes_count /* IN/OUT */,
+			  hb_ot_var_axis_info_t *axes_array /* OUT */);
+
+HB_EXTERN hb_bool_t
+hb_ot_var_find_axis_info (hb_face_t             *face,
+			  hb_tag_t               axis_tag,
+			  hb_ot_var_axis_info_t *axis_info);
+
+
+/*
+ * Named instances.
+ */
+
+HB_EXTERN unsigned int
+hb_ot_var_get_named_instance_count (hb_face_t *face);
+
+HB_EXTERN hb_ot_name_id_t
+hb_ot_var_named_instance_get_subfamily_name_id (hb_face_t   *face,
+						unsigned int instance_index);
+
+HB_EXTERN hb_ot_name_id_t
+hb_ot_var_named_instance_get_postscript_name_id (hb_face_t  *face,
+						unsigned int instance_index);
+
+HB_EXTERN unsigned int
+hb_ot_var_named_instance_get_design_coords (hb_face_t    *face,
+					    unsigned int  instance_index,
+					    unsigned int *coords_length, /* IN/OUT */
+					    float        *coords         /* OUT */);
+
+
+/*
+ * Conversions.
+ */
+
+HB_EXTERN void
+hb_ot_var_normalize_variations (hb_face_t            *face,
+				const hb_variation_t *variations, /* IN */
+				unsigned int          variations_length,
+				int                  *coords, /* OUT */
+				unsigned int          coords_length);
+
+HB_EXTERN void
+hb_ot_var_normalize_coords (hb_face_t    *face,
+			    unsigned int coords_length,
+			    const float *design_coords, /* IN */
+			    int *normalized_coords /* OUT */);
+
+
+HB_END_DECLS
+
+#endif /* HB_OT_VAR_H */
diff --git a/thirdparty/harfbuzz/src/hb-ot-vorg-table.hh b/thirdparty/harfbuzz/src/hb-ot-vorg-table.hh
new file mode 100644
index 0000000000..c6803200f9
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot-vorg-table.hh
@@ -0,0 +1,136 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_OT_VORG_TABLE_HH
+#define HB_OT_VORG_TABLE_HH
+
+#include "hb-open-type.hh"
+
+/*
+ * VORG -- Vertical Origin Table
+ * https://docs.microsoft.com/en-us/typography/opentype/spec/vorg
+ */
+#define HB_OT_TAG_VORG HB_TAG('V','O','R','G')
+
+namespace OT {
+
+struct VertOriginMetric
+{
+  int cmp (hb_codepoint_t g) const { return glyph.cmp (g); }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this));
+  }
+
+  public:
+  HBGlyphID	glyph;
+  FWORD		vertOriginY;
+
+  public:
+  DEFINE_SIZE_STATIC (4);
+};
+
+struct VORG
+{
+  static constexpr hb_tag_t tableTag = HB_OT_TAG_VORG;
+
+  bool has_data () const { return version.to_int (); }
+
+  int get_y_origin (hb_codepoint_t glyph) const
+  {
+    unsigned int i;
+    if (!vertYOrigins.bfind (glyph, &i))
+      return defaultVertOriginY;
+    return vertYOrigins[i].vertOriginY;
+  }
+
+  template <typename Iterator,
+	    hb_requires (hb_is_iterator (Iterator))>
+  void serialize (hb_serialize_context_t *c,
+		  Iterator it,
+		  FWORD defaultVertOriginY)
+  {
+
+    if (unlikely (!c->extend_min ((*this))))  return;
+
+    this->version.major = 1;
+    this->version.minor = 0;
+
+    this->defaultVertOriginY = defaultVertOriginY;
+    this->vertYOrigins.len = it.len ();
+
+    c->copy_all (it);
+  }
+
+  bool subset (hb_subset_context_t *c) const
+  {
+    TRACE_SUBSET (this);
+    VORG *vorg_prime = c->serializer->start_embed<VORG> ();
+    if (unlikely (!c->serializer->check_success (vorg_prime))) return_trace (false);
+
+    auto it =
+    + vertYOrigins.as_array ()
+    | hb_filter (c->plan->glyphset (), &VertOriginMetric::glyph)
+    | hb_map ([&] (const VertOriginMetric& _)
+	      {
+		hb_codepoint_t new_glyph = HB_SET_VALUE_INVALID;
+		c->plan->new_gid_for_old_gid (_.glyph, &new_glyph);
+
+		VertOriginMetric metric;
+		metric.glyph = new_glyph;
+		metric.vertOriginY = _.vertOriginY;
+		return metric;
+	      })
+    ;
+
+    /* serialize the new table */
+    vorg_prime->serialize (c->serializer, it, defaultVertOriginY);
+    return_trace (true);
+  }
+
+  bool sanitize (hb_sanitize_context_t *c) const
+  {
+    TRACE_SANITIZE (this);
+    return_trace (c->check_struct (this) &&
+		  version.major == 1 &&
+		  vertYOrigins.sanitize (c));
+  }
+
+  protected:
+  FixedVersion<>version;	/* Version of VORG table. Set to 0x00010000u. */
+  FWORD		defaultVertOriginY;
+				/* The default vertical origin. */
+  SortedArrayOf<VertOriginMetric>
+		vertYOrigins;	/* The array of vertical origins. */
+
+  public:
+  DEFINE_SIZE_ARRAY(8, vertYOrigins);
+};
+} /* namespace OT */
+
+#endif /* HB_OT_VORG_TABLE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ot.h b/thirdparty/harfbuzz/src/hb-ot.h
new file mode 100644
index 0000000000..f2dbaa1b31
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ot.h
@@ -0,0 +1,49 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_OT_H
+#define HB_OT_H
+#define HB_OT_H_IN
+
+#include "hb.h"
+
+#include "hb-ot-color.h"
+#include "hb-ot-deprecated.h"
+#include "hb-ot-font.h"
+#include "hb-ot-layout.h"
+#include "hb-ot-math.h"
+#include "hb-ot-meta.h"
+#include "hb-ot-metrics.h"
+#include "hb-ot-name.h"
+#include "hb-ot-shape.h"
+#include "hb-ot-var.h"
+
+HB_BEGIN_DECLS
+
+HB_END_DECLS
+
+#undef HB_OT_H_IN
+#endif /* HB_OT_H */
diff --git a/thirdparty/harfbuzz/src/hb-pool.hh b/thirdparty/harfbuzz/src/hb-pool.hh
new file mode 100644
index 0000000000..dcf0faf2a9
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-pool.hh
@@ -0,0 +1,100 @@
+/*
+ * Copyright © 2019  Facebook, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_POOL_HH
+#define HB_POOL_HH
+
+#include "hb.hh"
+
+/* Memory pool for persistent allocation of small objects. */
+
+template <typename T, unsigned ChunkLen = 16>
+struct hb_pool_t
+{
+  hb_pool_t () : next (nullptr) {}
+  ~hb_pool_t () { fini (); }
+
+  void fini ()
+  {
+    next = nullptr;
+
+    for (chunk_t *_ : chunks) ::free (_);
+
+    chunks.fini ();
+  }
+
+  T* alloc ()
+  {
+    if (unlikely (!next))
+    {
+      if (unlikely (!chunks.alloc (chunks.length + 1))) return nullptr;
+      chunk_t *chunk = (chunk_t *) calloc (1, sizeof (chunk_t));
+      if (unlikely (!chunk)) return nullptr;
+      chunks.push (chunk);
+      next = chunk->thread ();
+    }
+
+    T* obj = next;
+    next = * ((T**) next);
+
+    memset (obj, 0, sizeof (T));
+
+    return obj;
+  }
+
+  void free (T* obj)
+  {
+    * (T**) obj = next;
+    next = obj;
+  }
+
+  private:
+
+  static_assert (ChunkLen > 1, "");
+  static_assert (sizeof (T) >= sizeof (void *), "");
+  static_assert (alignof (T) % alignof (void *) == 0, "");
+
+  struct chunk_t
+  {
+    T* thread ()
+    {
+      for (unsigned i = 0; i < ARRAY_LENGTH (arrayZ) - 1; i++)
+	* (T**) &arrayZ[i] = &arrayZ[i + 1];
+
+      * (T**) &arrayZ[ARRAY_LENGTH (arrayZ) - 1] = nullptr;
+
+      return arrayZ;
+    }
+
+    T arrayZ[ChunkLen];
+  };
+
+  T* next;
+  hb_vector_t<chunk_t *> chunks;
+};
+
+
+#endif /* HB_POOL_HH */
diff --git a/thirdparty/harfbuzz/src/hb-sanitize.hh b/thirdparty/harfbuzz/src/hb-sanitize.hh
new file mode 100644
index 0000000000..024b4d1c99
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-sanitize.hh
@@ -0,0 +1,412 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2012,2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SANITIZE_HH
+#define HB_SANITIZE_HH
+
+#include "hb.hh"
+#include "hb-blob.hh"
+#include "hb-dispatch.hh"
+
+
+/*
+ * Sanitize
+ *
+ *
+ * === Introduction ===
+ *
+ * The sanitize machinery is at the core of our zero-cost font loading.  We
+ * mmap() font file into memory and create a blob out of it.  Font subtables
+ * are returned as a readonly sub-blob of the main font blob.  These table
+ * blobs are then sanitized before use, to ensure invalid memory access does
+ * not happen.  The toplevel sanitize API use is like, eg. to load the 'head'
+ * table:
+ *
+ *   hb_blob_t *head_blob = hb_sanitize_context_t ().reference_table<OT::head> (face);
+ *
+ * The blob then can be converted to a head table struct with:
+ *
+ *   const head *head_table = head_blob->as<head> ();
+ *
+ * What the reference_table does is, to call hb_face_reference_table() to load
+ * the table blob, sanitize it and return either the sanitized blob, or empty
+ * blob if sanitization failed.  The blob->as() function returns the null
+ * object of its template type argument if the blob is empty.  Otherwise, it
+ * just casts the blob contents to the desired type.
+ *
+ * Sanitizing a blob of data with a type T works as follows (with minor
+ * simplification):
+ *
+ *   - Cast blob content to T*, call sanitize() method of it,
+ *   - If sanitize succeeded, return blob.
+ *   - Otherwise, if blob is not writable, try making it writable,
+ *     or copy if cannot be made writable in-place,
+ *   - Call sanitize() again.  Return blob if sanitize succeeded.
+ *   - Return empty blob otherwise.
+ *
+ *
+ * === The sanitize() contract ===
+ *
+ * The sanitize() method of each object type shall return true if it's safe to
+ * call other methods of the object, and false otherwise.
+ *
+ * Note that what sanitize() checks for might align with what the specification
+ * describes as valid table data, but does not have to be.  In particular, we
+ * do NOT want to be pedantic and concern ourselves with validity checks that
+ * are irrelevant to our use of the table.  On the contrary, we want to be
+ * lenient with error handling and accept invalid data to the extent that it
+ * does not impose extra burden on us.
+ *
+ * Based on the sanitize contract, one can see that what we check for depends
+ * on how we use the data in other table methods.  Ie. if other table methods
+ * assume that offsets do NOT point out of the table data block, then that's
+ * something sanitize() must check for (GSUB/GPOS/GDEF/etc work this way).  On
+ * the other hand, if other methods do such checks themselves, then sanitize()
+ * does not have to bother with them (glyf/local work this way).  The choice
+ * depends on the table structure and sanitize() performance.  For example, to
+ * check glyf/loca offsets in sanitize() would cost O(num-glyphs).  We try hard
+ * to avoid such costs during font loading.  By postponing such checks to the
+ * actual glyph loading, we reduce the sanitize cost to O(1) and total runtime
+ * cost to O(used-glyphs).  As such, this is preferred.
+ *
+ * The same argument can be made re GSUB/GPOS/GDEF, but there, the table
+ * structure is so complicated that by checking all offsets at sanitize() time,
+ * we make the code much simpler in other methods, as offsets and referenced
+ * objects do not need to be validated at each use site.
+ */
+
+/* This limits sanitizing time on really broken fonts. */
+#ifndef HB_SANITIZE_MAX_EDITS
+#define HB_SANITIZE_MAX_EDITS 32
+#endif
+#ifndef HB_SANITIZE_MAX_OPS_FACTOR
+#define HB_SANITIZE_MAX_OPS_FACTOR 8
+#endif
+#ifndef HB_SANITIZE_MAX_OPS_MIN
+#define HB_SANITIZE_MAX_OPS_MIN 16384
+#endif
+#ifndef HB_SANITIZE_MAX_OPS_MAX
+#define HB_SANITIZE_MAX_OPS_MAX 0x3FFFFFFF
+#endif
+#ifndef HB_SANITIZE_MAX_SUTABLES
+#define HB_SANITIZE_MAX_SUTABLES 0x4000
+#endif
+
+struct hb_sanitize_context_t :
+       hb_dispatch_context_t<hb_sanitize_context_t, bool, HB_DEBUG_SANITIZE>
+{
+  hb_sanitize_context_t () :
+	start (nullptr), end (nullptr),
+	max_ops (0), max_subtables (0),
+	writable (false), edit_count (0),
+	blob (nullptr),
+	num_glyphs (65536),
+	num_glyphs_set (false) {}
+
+  const char *get_name () { return "SANITIZE"; }
+  template <typename T, typename F>
+  bool may_dispatch (const T *obj HB_UNUSED, const F *format)
+  { return format->sanitize (this); }
+  static return_t default_return_value () { return true; }
+  static return_t no_dispatch_return_value () { return false; }
+  bool stop_sublookup_iteration (const return_t r) const { return !r; }
+
+  bool visit_subtables (unsigned count)
+  {
+    max_subtables += count;
+    return max_subtables < HB_SANITIZE_MAX_SUTABLES;
+  }
+
+  private:
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<1>, Ts&&... ds) HB_AUTO_RETURN
+  ( obj.sanitize (this, hb_forward<Ts> (ds)...) )
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<0>, Ts&&... ds) HB_AUTO_RETURN
+  ( obj.dispatch (this, hb_forward<Ts> (ds)...) )
+  public:
+  template <typename T, typename ...Ts> auto
+  dispatch (const T &obj, Ts&&... ds) HB_AUTO_RETURN
+  ( _dispatch (obj, hb_prioritize, hb_forward<Ts> (ds)...) )
+
+
+  void init (hb_blob_t *b)
+  {
+    this->blob = hb_blob_reference (b);
+    this->writable = false;
+  }
+
+  void set_num_glyphs (unsigned int num_glyphs_)
+  {
+    num_glyphs = num_glyphs_;
+    num_glyphs_set = true;
+  }
+  unsigned int get_num_glyphs () { return num_glyphs; }
+
+  void set_max_ops (int max_ops_) { max_ops = max_ops_; }
+
+  template <typename T>
+  void set_object (const T *obj)
+  {
+    reset_object ();
+
+    if (!obj) return;
+
+    const char *obj_start = (const char *) obj;
+    if (unlikely (obj_start < this->start || this->end <= obj_start))
+      this->start = this->end = nullptr;
+    else
+    {
+      this->start = obj_start;
+      this->end   = obj_start + hb_min (size_t (this->end - obj_start), obj->get_size ());
+    }
+  }
+
+  void reset_object ()
+  {
+    this->start = this->blob->data;
+    this->end = this->start + this->blob->length;
+    assert (this->start <= this->end); /* Must not overflow. */
+  }
+
+  void start_processing ()
+  {
+    reset_object ();
+    if (unlikely (hb_unsigned_mul_overflows (this->end - this->start, HB_SANITIZE_MAX_OPS_FACTOR)))
+      this->max_ops = HB_SANITIZE_MAX_OPS_MAX;
+    else
+      this->max_ops = hb_clamp ((unsigned) (this->end - this->start) * HB_SANITIZE_MAX_OPS_FACTOR,
+				(unsigned) HB_SANITIZE_MAX_OPS_MIN,
+				(unsigned) HB_SANITIZE_MAX_OPS_MAX);
+    this->edit_count = 0;
+    this->debug_depth = 0;
+
+    DEBUG_MSG_LEVEL (SANITIZE, start, 0, +1,
+		     "start [%p..%p] (%lu bytes)",
+		     this->start, this->end,
+		     (unsigned long) (this->end - this->start));
+  }
+
+  void end_processing ()
+  {
+    DEBUG_MSG_LEVEL (SANITIZE, this->start, 0, -1,
+		     "end [%p..%p] %u edit requests",
+		     this->start, this->end, this->edit_count);
+
+    hb_blob_destroy (this->blob);
+    this->blob = nullptr;
+    this->start = this->end = nullptr;
+  }
+
+  unsigned get_edit_count () { return edit_count; }
+
+  bool check_range (const void *base,
+		    unsigned int len) const
+  {
+    const char *p = (const char *) base;
+    bool ok = !len ||
+	      (this->start <= p &&
+	       p <= this->end &&
+	       (unsigned int) (this->end - p) >= len &&
+	       this->max_ops-- > 0);
+
+    DEBUG_MSG_LEVEL (SANITIZE, p, this->debug_depth+1, 0,
+		     "check_range [%p..%p]"
+		     " (%d bytes) in [%p..%p] -> %s",
+		     p, p + len, len,
+		     this->start, this->end,
+		     ok ? "OK" : "OUT-OF-RANGE");
+
+    return likely (ok);
+  }
+
+  template <typename T>
+  bool check_range (const T *base,
+		    unsigned int a,
+		    unsigned int b) const
+  {
+    return !hb_unsigned_mul_overflows (a, b) &&
+	   this->check_range (base, a * b);
+  }
+
+  template <typename T>
+  bool check_range (const T *base,
+		    unsigned int a,
+		    unsigned int b,
+		    unsigned int c) const
+  {
+    return !hb_unsigned_mul_overflows (a, b) &&
+	   this->check_range (base, a * b, c);
+  }
+
+  template <typename T>
+  bool check_array (const T *base, unsigned int len) const
+  {
+    return this->check_range (base, len, hb_static_size (T));
+  }
+
+  template <typename T>
+  bool check_array (const T *base,
+		    unsigned int a,
+		    unsigned int b) const
+  {
+    return this->check_range (base, a, b, hb_static_size (T));
+  }
+
+  template <typename Type>
+  bool check_struct (const Type *obj) const
+  { return likely (this->check_range (obj, obj->min_size)); }
+
+  bool may_edit (const void *base, unsigned int len)
+  {
+    if (this->edit_count >= HB_SANITIZE_MAX_EDITS)
+      return false;
+
+    const char *p = (const char *) base;
+    this->edit_count++;
+
+    DEBUG_MSG_LEVEL (SANITIZE, p, this->debug_depth+1, 0,
+       "may_edit(%u) [%p..%p] (%d bytes) in [%p..%p] -> %s",
+       this->edit_count,
+       p, p + len, len,
+       this->start, this->end,
+       this->writable ? "GRANTED" : "DENIED");
+
+    return this->writable;
+  }
+
+  template <typename Type, typename ValueType>
+  bool try_set (const Type *obj, const ValueType &v)
+  {
+    if (this->may_edit (obj, hb_static_size (Type)))
+    {
+      * const_cast<Type *> (obj) = v;
+      return true;
+    }
+    return false;
+  }
+
+  template <typename Type>
+  hb_blob_t *sanitize_blob (hb_blob_t *blob)
+  {
+    bool sane;
+
+    init (blob);
+
+  retry:
+    DEBUG_MSG_FUNC (SANITIZE, start, "start");
+
+    start_processing ();
+
+    if (unlikely (!start))
+    {
+      end_processing ();
+      return blob;
+    }
+
+    Type *t = reinterpret_cast<Type *> (const_cast<char *> (start));
+
+    sane = t->sanitize (this);
+    if (sane)
+    {
+      if (edit_count)
+      {
+	DEBUG_MSG_FUNC (SANITIZE, start, "passed first round with %d edits; going for second round", edit_count);
+
+	/* sanitize again to ensure no toe-stepping */
+	edit_count = 0;
+	sane = t->sanitize (this);
+	if (edit_count) {
+	  DEBUG_MSG_FUNC (SANITIZE, start, "requested %d edits in second round; FAILLING", edit_count);
+	  sane = false;
+	}
+      }
+    }
+    else
+    {
+      if (edit_count && !writable) {
+	start = hb_blob_get_data_writable (blob, nullptr);
+	end = start + blob->length;
+
+	if (start)
+	{
+	  writable = true;
+	  /* ok, we made it writable by relocating.  try again */
+	  DEBUG_MSG_FUNC (SANITIZE, start, "retry");
+	  goto retry;
+	}
+      }
+    }
+
+    end_processing ();
+
+    DEBUG_MSG_FUNC (SANITIZE, start, sane ? "PASSED" : "FAILED");
+    if (sane)
+    {
+      hb_blob_make_immutable (blob);
+      return blob;
+    }
+    else
+    {
+      hb_blob_destroy (blob);
+      return hb_blob_get_empty ();
+    }
+  }
+
+  template <typename Type>
+  hb_blob_t *reference_table (const hb_face_t *face, hb_tag_t tableTag = Type::tableTag)
+  {
+    if (!num_glyphs_set)
+      set_num_glyphs (hb_face_get_glyph_count (face));
+    return sanitize_blob<Type> (hb_face_reference_table (face, tableTag));
+  }
+
+  const char *start, *end;
+  mutable int max_ops, max_subtables;
+  private:
+  bool writable;
+  unsigned int edit_count;
+  hb_blob_t *blob;
+  unsigned int num_glyphs;
+  bool  num_glyphs_set;
+};
+
+struct hb_sanitize_with_object_t
+{
+  template <typename T>
+  hb_sanitize_with_object_t (hb_sanitize_context_t *c, const T& obj) : c (c)
+  { c->set_object (obj); }
+  ~hb_sanitize_with_object_t ()
+  { c->reset_object (); }
+
+  private:
+  hb_sanitize_context_t *c;
+};
+
+
+#endif /* HB_SANITIZE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-serialize.hh b/thirdparty/harfbuzz/src/hb-serialize.hh
new file mode 100644
index 0000000000..4566153a59
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-serialize.hh
@@ -0,0 +1,553 @@
+/*
+ * Copyright © 2007,2008,2009,2010  Red Hat, Inc.
+ * Copyright © 2012,2018  Google, Inc.
+ * Copyright © 2019  Facebook, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ * Facebook Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SERIALIZE_HH
+#define HB_SERIALIZE_HH
+
+#include "hb.hh"
+#include "hb-blob.hh"
+#include "hb-map.hh"
+#include "hb-pool.hh"
+
+
+/*
+ * Serialize
+ */
+
+struct hb_serialize_context_t
+{
+  typedef unsigned objidx_t;
+
+  enum whence_t {
+     Head,	/* Relative to the current object head (default). */
+     Tail,	/* Relative to the current object tail after packed. */
+     Absolute	/* Absolute: from the start of the serialize buffer. */
+   };
+
+  struct object_t
+  {
+    void fini () { links.fini (); }
+
+    bool operator == (const object_t &o) const
+    {
+      return (tail - head == o.tail - o.head)
+	  && (links.length == o.links.length)
+	  && 0 == hb_memcmp (head, o.head, tail - head)
+	  && links.as_bytes () == o.links.as_bytes ();
+    }
+    uint32_t hash () const
+    {
+      return hb_bytes_t (head, tail - head).hash () ^
+	     links.as_bytes ().hash ();
+    }
+
+    struct link_t
+    {
+      bool is_wide: 1;
+      bool is_signed: 1;
+      unsigned whence: 2;
+      unsigned position: 28;
+      unsigned bias;
+      objidx_t objidx;
+    };
+
+    char *head;
+    char *tail;
+    hb_vector_t<link_t> links;
+    object_t *next;
+  };
+
+  struct snapshot_t
+  {
+    char *head;
+    char *tail;
+    object_t *current; // Just for sanity check
+    unsigned num_links;
+  };
+
+  snapshot_t snapshot ()
+  { return snapshot_t { head, tail, current, current->links.length }; }
+
+  hb_serialize_context_t (void *start_, unsigned int size) :
+    start ((char *) start_),
+    end (start + size),
+    current (nullptr)
+  { reset (); }
+  ~hb_serialize_context_t () { fini (); }
+
+  void fini ()
+  {
+    for (object_t *_ : ++hb_iter (packed)) _->fini ();
+    packed.fini ();
+    this->packed_map.fini ();
+
+    while (current)
+    {
+      auto *_ = current;
+      current = current->next;
+      _->fini ();
+    }
+    object_pool.fini ();
+  }
+
+  bool in_error () const { return !this->successful; }
+
+  void reset ()
+  {
+    this->successful = true;
+    this->ran_out_of_room = false;
+    this->head = this->start;
+    this->tail = this->end;
+    this->debug_depth = 0;
+
+    fini ();
+    this->packed.push (nullptr);
+  }
+
+  bool check_success (bool success)
+  { return this->successful && (success || (err_other_error (), false)); }
+
+  template <typename T1, typename T2>
+  bool check_equal (T1 &&v1, T2 &&v2)
+  { return check_success ((long long) v1 == (long long) v2); }
+
+  template <typename T1, typename T2>
+  bool check_assign (T1 &v1, T2 &&v2)
+  { return check_equal (v1 = v2, v2); }
+
+  template <typename T> bool propagate_error (T &&obj)
+  { return check_success (!hb_deref (obj).in_error ()); }
+
+  template <typename T1, typename... Ts> bool propagate_error (T1 &&o1, Ts&&... os)
+  { return propagate_error (hb_forward<T1> (o1)) &&
+	   propagate_error (hb_forward<Ts> (os)...); }
+
+  /* To be called around main operation. */
+  template <typename Type>
+  Type *start_serialize ()
+  {
+    DEBUG_MSG_LEVEL (SERIALIZE, this->start, 0, +1,
+		     "start [%p..%p] (%lu bytes)",
+		     this->start, this->end,
+		     (unsigned long) (this->end - this->start));
+
+    assert (!current);
+    return push<Type> ();
+  }
+  void end_serialize ()
+  {
+    DEBUG_MSG_LEVEL (SERIALIZE, this->start, 0, -1,
+		     "end [%p..%p] serialized %u bytes; %s",
+		     this->start, this->end,
+		     (unsigned) (this->head - this->start),
+		     this->successful ? "successful" : "UNSUCCESSFUL");
+
+    propagate_error (packed, packed_map);
+
+    if (unlikely (!current)) return;
+    if (unlikely (in_error())) return;
+
+    assert (!current->next);
+
+    /* Only "pack" if there exist other objects... Otherwise, don't bother.
+     * Saves a move. */
+    if (packed.length <= 1)
+      return;
+
+    pop_pack (false);
+
+    resolve_links ();
+  }
+
+  template <typename Type = void>
+  Type *push ()
+  {
+    if (unlikely (in_error ())) return start_embed<Type> ();
+
+    object_t *obj = object_pool.alloc ();
+    if (unlikely (!obj))
+      check_success (false);
+    else
+    {
+      obj->head = head;
+      obj->tail = tail;
+      obj->next = current;
+      current = obj;
+    }
+    return start_embed<Type> ();
+  }
+  void pop_discard ()
+  {
+    object_t *obj = current;
+    if (unlikely (!obj)) return;
+    if (unlikely (in_error())) return;
+
+    current = current->next;
+    revert (obj->head, obj->tail);
+    obj->fini ();
+    object_pool.free (obj);
+  }
+
+  /* Set share to false when an object is unlikely sharable with others
+   * so not worth an attempt, or a contiguous table is serialized as
+   * multiple consecutive objects in the reverse order so can't be shared.
+   */
+  objidx_t pop_pack (bool share=true)
+  {
+    object_t *obj = current;
+    if (unlikely (!obj)) return 0;
+    if (unlikely (in_error())) return 0;
+
+    current = current->next;
+    obj->tail = head;
+    obj->next = nullptr;
+    unsigned len = obj->tail - obj->head;
+    head = obj->head; /* Rewind head. */
+
+    if (!len)
+    {
+      assert (!obj->links.length);
+      return 0;
+    }
+
+    objidx_t objidx;
+    if (share)
+    {
+      objidx = packed_map.get (obj);
+      if (objidx)
+      {
+	obj->fini ();
+	return objidx;
+      }
+    }
+
+    tail -= len;
+    memmove (tail, obj->head, len);
+
+    obj->head = tail;
+    obj->tail = tail + len;
+
+    packed.push (obj);
+
+    if (unlikely (packed.in_error ())) {
+      // obj wasn't successfully added to packed, so clean it up otherwise it's
+      // links will be leaked.
+      propagate_error (packed);
+      obj->fini ();
+      return 0;
+    }
+
+    objidx = packed.length - 1;
+
+    if (share) packed_map.set (obj, objidx);
+    propagate_error (packed_map);
+
+    return objidx;
+  }
+
+  void revert (snapshot_t snap)
+  {
+    if (unlikely (in_error ())) return;
+    assert (snap.current == current);
+    current->links.shrink (snap.num_links);
+    revert (snap.head, snap.tail);
+  }
+
+  void revert (char *snap_head,
+	       char *snap_tail)
+  {
+    if (unlikely (in_error ())) return;
+    assert (snap_head <= head);
+    assert (tail <= snap_tail);
+    head = snap_head;
+    tail = snap_tail;
+    discard_stale_objects ();
+  }
+
+  void discard_stale_objects ()
+  {
+    if (unlikely (in_error ())) return;
+    while (packed.length > 1 &&
+	   packed.tail ()->head < tail)
+    {
+      packed_map.del (packed.tail ());
+      assert (!packed.tail ()->next);
+      packed.tail ()->fini ();
+      packed.pop ();
+    }
+    if (packed.length > 1)
+      assert (packed.tail ()->head == tail);
+  }
+
+  template <typename T>
+  void add_link (T &ofs, objidx_t objidx,
+		 whence_t whence = Head,
+		 unsigned bias = 0)
+  {
+    static_assert (sizeof (T) == 2 || sizeof (T) == 4, "");
+    if (unlikely (in_error ())) return;
+
+    if (!objidx)
+      return;
+
+    assert (current);
+    assert (current->head <= (const char *) &ofs);
+
+    auto& link = *current->links.push ();
+
+    link.is_wide = sizeof (T) == 4;
+    link.is_signed = hb_is_signed (hb_unwrap_type (T));
+    link.whence = (unsigned) whence;
+    link.position = (const char *) &ofs - current->head;
+    link.bias = bias;
+    link.objidx = objidx;
+  }
+
+  unsigned to_bias (const void *base) const
+  {
+    if (unlikely (in_error ())) return 0;
+    if (!base) return 0;
+    assert (current);
+    assert (current->head <= (const char *) base);
+    return (const char *) base - current->head;
+  }
+
+  void resolve_links ()
+  {
+    if (unlikely (in_error ())) return;
+
+    assert (!current);
+    assert (packed.length > 1);
+
+    for (const object_t* parent : ++hb_iter (packed))
+      for (const object_t::link_t &link : parent->links)
+      {
+	const object_t* child = packed[link.objidx];
+	if (unlikely (!child)) { err_other_error(); return; }
+	unsigned offset = 0;
+	switch ((whence_t) link.whence) {
+	case Head:     offset = child->head - parent->head; break;
+	case Tail:     offset = child->head - parent->tail; break;
+	case Absolute: offset = (head - start) + (child->head - tail); break;
+	}
+
+	assert (offset >= link.bias);
+	offset -= link.bias;
+	if (link.is_signed)
+	{
+	  if (link.is_wide)
+	    assign_offset<int32_t> (parent, link, offset);
+	  else
+	    assign_offset<int16_t> (parent, link, offset);
+	}
+	else
+	{
+	  if (link.is_wide)
+	    assign_offset<uint32_t> (parent, link, offset);
+	  else
+	    assign_offset<uint16_t> (parent, link, offset);
+	}
+      }
+  }
+
+  unsigned int length () const
+  {
+    if (unlikely (!current)) return 0;
+    return this->head - current->head;
+  }
+
+  void align (unsigned int alignment)
+  {
+    unsigned int l = length () % alignment;
+    if (l)
+      allocate_size<void> (alignment - l);
+  }
+
+  template <typename Type = void>
+  Type *start_embed (const Type *obj HB_UNUSED = nullptr) const
+  { return reinterpret_cast<Type *> (this->head); }
+  template <typename Type>
+  Type *start_embed (const Type &obj) const
+  { return start_embed (hb_addressof (obj)); }
+
+  /* Following two functions exist to allow setting breakpoint on. */
+  void err_ran_out_of_room () { this->ran_out_of_room = true; }
+  void err_other_error () { this->successful = false; }
+
+  template <typename Type>
+  Type *allocate_size (unsigned int size)
+  {
+    if (unlikely (!this->successful)) return nullptr;
+
+    if (this->tail - this->head < ptrdiff_t (size))
+    {
+      err_ran_out_of_room ();
+      this->successful = false;
+      return nullptr;
+    }
+    memset (this->head, 0, size);
+    char *ret = this->head;
+    this->head += size;
+    return reinterpret_cast<Type *> (ret);
+  }
+
+  template <typename Type>
+  Type *allocate_min ()
+  { return this->allocate_size<Type> (Type::min_size); }
+
+  template <typename Type>
+  Type *embed (const Type *obj)
+  {
+    unsigned int size = obj->get_size ();
+    Type *ret = this->allocate_size<Type> (size);
+    if (unlikely (!ret)) return nullptr;
+    memcpy (ret, obj, size);
+    return ret;
+  }
+  template <typename Type>
+  Type *embed (const Type &obj)
+  { return embed (hb_addressof (obj)); }
+
+  template <typename Type, typename ...Ts> auto
+  _copy (const Type &src, hb_priority<1>, Ts&&... ds) HB_RETURN
+  (Type *, src.copy (this, hb_forward<Ts> (ds)...))
+
+  template <typename Type> auto
+  _copy (const Type &src, hb_priority<0>) -> decltype (&(hb_declval<Type> () = src))
+  {
+    Type *ret = this->allocate_size<Type> (sizeof (Type));
+    if (unlikely (!ret)) return nullptr;
+    *ret = src;
+    return ret;
+  }
+
+  /* Like embed, but active: calls obj.operator=() or obj.copy() to transfer data
+   * instead of memcpy(). */
+  template <typename Type, typename ...Ts>
+  Type *copy (const Type &src, Ts&&... ds)
+  { return _copy (src, hb_prioritize, hb_forward<Ts> (ds)...); }
+  template <typename Type, typename ...Ts>
+  Type *copy (const Type *src, Ts&&... ds)
+  { return copy (*src, hb_forward<Ts> (ds)...); }
+
+  template<typename Iterator,
+	   hb_requires (hb_is_iterator (Iterator)),
+	   typename ...Ts>
+  void copy_all (Iterator it, Ts&&... ds)
+  { for (decltype (*it) _ : it) copy (_, hb_forward<Ts> (ds)...); }
+
+  template <typename Type>
+  hb_serialize_context_t& operator << (const Type &obj) & { embed (obj); return *this; }
+
+  template <typename Type>
+  Type *extend_size (Type *obj, unsigned int size)
+  {
+    if (unlikely (in_error ())) return nullptr;
+
+    assert (this->start <= (char *) obj);
+    assert ((char *) obj <= this->head);
+    assert ((char *) obj + size >= this->head);
+    if (unlikely (!this->allocate_size<Type> (((char *) obj) + size - this->head))) return nullptr;
+    return reinterpret_cast<Type *> (obj);
+  }
+  template <typename Type>
+  Type *extend_size (Type &obj, unsigned int size)
+  { return extend_size (hb_addressof (obj), size); }
+
+  template <typename Type>
+  Type *extend_min (Type *obj) { return extend_size (obj, obj->min_size); }
+  template <typename Type>
+  Type *extend_min (Type &obj) { return extend_min (hb_addressof (obj)); }
+
+  template <typename Type, typename ...Ts>
+  Type *extend (Type *obj, Ts&&... ds)
+  { return extend_size (obj, obj->get_size (hb_forward<Ts> (ds)...)); }
+  template <typename Type, typename ...Ts>
+  Type *extend (Type &obj, Ts&&... ds)
+  { return extend (hb_addressof (obj), hb_forward<Ts> (ds)...); }
+
+  /* Output routines. */
+  hb_bytes_t copy_bytes () const
+  {
+    assert (this->successful);
+    /* Copy both items from head side and tail side... */
+    unsigned int len = (this->head - this->start)
+		     + (this->end  - this->tail);
+
+    char *p = (char *) malloc (len);
+    if (unlikely (!p)) return hb_bytes_t ();
+
+    memcpy (p, this->start, this->head - this->start);
+    memcpy (p + (this->head - this->start), this->tail, this->end - this->tail);
+    return hb_bytes_t (p, len);
+  }
+  template <typename Type>
+  Type *copy () const
+  { return reinterpret_cast<Type *> ((char *) copy_bytes ().arrayZ); }
+  hb_blob_t *copy_blob () const
+  {
+    hb_bytes_t b = copy_bytes ();
+    return hb_blob_create (b.arrayZ, b.length,
+			   HB_MEMORY_MODE_WRITABLE,
+			   (char *) b.arrayZ, free);
+  }
+
+  private:
+  template <typename T>
+  void assign_offset (const object_t* parent, const object_t::link_t &link, unsigned offset)
+  {
+    auto &off = * ((BEInt<T, sizeof (T)> *) (parent->head + link.position));
+    assert (0 == off);
+    check_assign (off, offset);
+  }
+
+  public: /* TODO Make private. */
+  char *start, *head, *tail, *end;
+  unsigned int debug_depth;
+  bool successful;
+  bool ran_out_of_room;
+
+  private:
+
+  /* Object memory pool. */
+  hb_pool_t<object_t> object_pool;
+
+  /* Stack of currently under construction objects. */
+  object_t *current;
+
+  /* Stack of packed objects.  Object 0 is always nil object. */
+  hb_vector_t<object_t *> packed;
+
+  /* Map view of packed objects. */
+  hb_hashmap_t<const object_t *, objidx_t, nullptr, 0> packed_map;
+};
+
+
+#endif /* HB_SERIALIZE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-set-digest.hh b/thirdparty/harfbuzz/src/hb-set-digest.hh
new file mode 100644
index 0000000000..b97526f775
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-set-digest.hh
@@ -0,0 +1,174 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SET_DIGEST_HH
+#define HB_SET_DIGEST_HH
+
+#include "hb.hh"
+
+/*
+ * The set digests here implement various "filters" that support
+ * "approximate member query".  Conceptually these are like Bloom
+ * Filter and Quotient Filter, however, much smaller, faster, and
+ * designed to fit the requirements of our uses for glyph coverage
+ * queries.
+ *
+ * Our filters are highly accurate if the lookup covers fairly local
+ * set of glyphs, but fully flooded and ineffective if coverage is
+ * all over the place.
+ *
+ * The frozen-set can be used instead of a digest, to trade more
+ * memory for 100% accuracy, but in practice, that doesn't look like
+ * an attractive trade-off.
+ */
+
+template <typename mask_t, unsigned int shift>
+struct hb_set_digest_lowest_bits_t
+{
+  static constexpr unsigned mask_bytes = sizeof (mask_t);
+  static constexpr unsigned mask_bits = sizeof (mask_t) * 8;
+  static constexpr unsigned num_bits = 0
+				     + (mask_bytes >= 1 ? 3 : 0)
+				     + (mask_bytes >= 2 ? 1 : 0)
+				     + (mask_bytes >= 4 ? 1 : 0)
+				     + (mask_bytes >= 8 ? 1 : 0)
+				     + (mask_bytes >= 16? 1 : 0)
+				     + 0;
+
+  static_assert ((shift < sizeof (hb_codepoint_t) * 8), "");
+  static_assert ((shift + num_bits <= sizeof (hb_codepoint_t) * 8), "");
+
+  void init () { mask = 0; }
+
+  void add (hb_codepoint_t g) { mask |= mask_for (g); }
+
+  bool add_range (hb_codepoint_t a, hb_codepoint_t b)
+  {
+    if ((b >> shift) - (a >> shift) >= mask_bits - 1)
+      mask = (mask_t) -1;
+    else {
+      mask_t ma = mask_for (a);
+      mask_t mb = mask_for (b);
+      mask |= mb + (mb - ma) - (mb < ma);
+    }
+    return true;
+  }
+
+  template <typename T>
+  void add_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    for (unsigned int i = 0; i < count; i++)
+    {
+      add (*array);
+      array = (const T *) (stride + (const char *) array);
+    }
+  }
+  template <typename T>
+  bool add_sorted_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    for (unsigned int i = 0; i < count; i++)
+    {
+      add (*array);
+      array = (const T *) (stride + (const char *) array);
+    }
+    return true;
+  }
+
+  bool may_have (hb_codepoint_t g) const
+  { return !!(mask & mask_for (g)); }
+
+  private:
+
+  static mask_t mask_for (hb_codepoint_t g)
+  { return ((mask_t) 1) << ((g >> shift) & (mask_bits - 1)); }
+  mask_t mask;
+};
+
+template <typename head_t, typename tail_t>
+struct hb_set_digest_combiner_t
+{
+  void init ()
+  {
+    head.init ();
+    tail.init ();
+  }
+
+  void add (hb_codepoint_t g)
+  {
+    head.add (g);
+    tail.add (g);
+  }
+
+  bool add_range (hb_codepoint_t a, hb_codepoint_t b)
+  {
+    head.add_range (a, b);
+    tail.add_range (a, b);
+    return true;
+  }
+  template <typename T>
+  void add_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    head.add_array (array, count, stride);
+    tail.add_array (array, count, stride);
+  }
+  template <typename T>
+  bool add_sorted_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    head.add_sorted_array (array, count, stride);
+    tail.add_sorted_array (array, count, stride);
+    return true;
+  }
+
+  bool may_have (hb_codepoint_t g) const
+  {
+    return head.may_have (g) && tail.may_have (g);
+  }
+
+  private:
+  head_t head;
+  tail_t tail;
+};
+
+
+/*
+ * hb_set_digest_t
+ *
+ * This is a combination of digests that performs "best".
+ * There is not much science to this: it's a result of intuition
+ * and testing.
+ */
+typedef hb_set_digest_combiner_t
+<
+  hb_set_digest_lowest_bits_t<unsigned long, 4>,
+  hb_set_digest_combiner_t
+  <
+    hb_set_digest_lowest_bits_t<unsigned long, 0>,
+    hb_set_digest_lowest_bits_t<unsigned long, 9>
+  >
+> hb_set_digest_t;
+
+
+#endif /* HB_SET_DIGEST_HH */
diff --git a/thirdparty/harfbuzz/src/hb-set.cc b/thirdparty/harfbuzz/src/hb-set.cc
new file mode 100644
index 0000000000..0551ed80f2
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-set.cc
@@ -0,0 +1,541 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb-set.hh"
+
+
+/**
+ * SECTION:hb-set
+ * @title: hb-set
+ * @short_description: Object representing a set of integers
+ * @include: hb.h
+ *
+ * Set objects represent a mathematical set of integer values.  They are
+ * used in non-shaping API to query certain set of characters or glyphs,
+ * or other integer values.
+ **/
+
+
+/**
+ * hb_set_create: (Xconstructor)
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_set_t *
+hb_set_create ()
+{
+  hb_set_t *set;
+
+  if (!(set = hb_object_create<hb_set_t> ()))
+    return hb_set_get_empty ();
+
+  set->init_shallow ();
+
+  return set;
+}
+
+/**
+ * hb_set_get_empty:
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_set_t *
+hb_set_get_empty ()
+{
+  return const_cast<hb_set_t *> (&Null (hb_set_t));
+}
+
+/**
+ * hb_set_reference: (skip)
+ * @set: a set.
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_set_t *
+hb_set_reference (hb_set_t *set)
+{
+  return hb_object_reference (set);
+}
+
+/**
+ * hb_set_destroy: (skip)
+ * @set: a set.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_destroy (hb_set_t *set)
+{
+  if (!hb_object_destroy (set)) return;
+
+  set->fini_shallow ();
+
+  free (set);
+}
+
+/**
+ * hb_set_set_user_data: (skip)
+ * @set: a set.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_set_set_user_data (hb_set_t           *set,
+		      hb_user_data_key_t *key,
+		      void *              data,
+		      hb_destroy_func_t   destroy,
+		      hb_bool_t           replace)
+{
+  return hb_object_set_user_data (set, key, data, destroy, replace);
+}
+
+/**
+ * hb_set_get_user_data: (skip)
+ * @set: a set.
+ * @key:
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_set_get_user_data (hb_set_t           *set,
+		      hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (set, key);
+}
+
+
+/**
+ * hb_set_allocation_successful:
+ * @set: a set.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_set_allocation_successful (const hb_set_t  *set)
+{
+  return set->successful;
+}
+
+/**
+ * hb_set_clear:
+ * @set: a set.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_clear (hb_set_t *set)
+{
+  set->clear ();
+}
+
+/**
+ * hb_set_is_empty:
+ * @set: a set.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_set_is_empty (const hb_set_t *set)
+{
+  return set->is_empty ();
+}
+
+/**
+ * hb_set_has:
+ * @set: a set.
+ * @codepoint:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_set_has (const hb_set_t *set,
+	    hb_codepoint_t  codepoint)
+{
+  return set->has (codepoint);
+}
+
+/**
+ * hb_set_add:
+ * @set: a set.
+ * @codepoint:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_add (hb_set_t       *set,
+	    hb_codepoint_t  codepoint)
+{
+  set->add (codepoint);
+}
+
+/**
+ * hb_set_add_range:
+ * @set: a set.
+ * @first:
+ * @last:
+ *
+ *
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_set_add_range (hb_set_t       *set,
+		  hb_codepoint_t  first,
+		  hb_codepoint_t  last)
+{
+  set->add_range (first, last);
+}
+
+/**
+ * hb_set_del:
+ * @set: a set.
+ * @codepoint:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_del (hb_set_t       *set,
+	    hb_codepoint_t  codepoint)
+{
+  set->del (codepoint);
+}
+
+/**
+ * hb_set_del_range:
+ * @set: a set.
+ * @first:
+ * @last:
+ *
+ *
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_set_del_range (hb_set_t       *set,
+		  hb_codepoint_t  first,
+		  hb_codepoint_t  last)
+{
+  set->del_range (first, last);
+}
+
+/**
+ * hb_set_is_equal:
+ * @set: a set.
+ * @other: other set.
+ *
+ *
+ *
+ * Return value: %TRUE if the two sets are equal, %FALSE otherwise.
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_set_is_equal (const hb_set_t *set,
+		 const hb_set_t *other)
+{
+  return set->is_equal (other);
+}
+
+/**
+ * hb_set_is_subset:
+ * @set: a set.
+ * @larger_set: other set.
+ *
+ *
+ *
+ * Return value: %TRUE if the @set is a subset of (or equal to) @larger_set, %FALSE otherwise.
+ *
+ * Since: 1.8.1
+ **/
+hb_bool_t
+hb_set_is_subset (const hb_set_t *set,
+		  const hb_set_t *larger_set)
+{
+  return set->is_subset (larger_set);
+}
+
+/**
+ * hb_set_set:
+ * @set: a set.
+ * @other:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_set (hb_set_t       *set,
+	    const hb_set_t *other)
+{
+  set->set (other);
+}
+
+/**
+ * hb_set_union:
+ * @set: a set.
+ * @other:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_union (hb_set_t       *set,
+	      const hb_set_t *other)
+{
+  set->union_ (other);
+}
+
+/**
+ * hb_set_intersect:
+ * @set: a set.
+ * @other:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_intersect (hb_set_t       *set,
+		  const hb_set_t *other)
+{
+  set->intersect (other);
+}
+
+/**
+ * hb_set_subtract:
+ * @set: a set.
+ * @other:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_subtract (hb_set_t       *set,
+		 const hb_set_t *other)
+{
+  set->subtract (other);
+}
+
+/**
+ * hb_set_symmetric_difference:
+ * @set: a set.
+ * @other:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_set_symmetric_difference (hb_set_t       *set,
+			     const hb_set_t *other)
+{
+  set->symmetric_difference (other);
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_set_invert:
+ * @set: a set.
+ *
+ *
+ *
+ * Since: 0.9.10
+ *
+ * Deprecated: 1.6.1
+ **/
+void
+hb_set_invert (hb_set_t *set HB_UNUSED)
+{
+}
+#endif
+
+/**
+ * hb_set_get_population:
+ * @set: a set.
+ *
+ * Returns the number of numbers in the set.
+ *
+ * Return value: set population.
+ *
+ * Since: 0.9.7
+ **/
+unsigned int
+hb_set_get_population (const hb_set_t *set)
+{
+  return set->get_population ();
+}
+
+/**
+ * hb_set_get_min:
+ * @set: a set.
+ *
+ * Finds the minimum number in the set.
+ *
+ * Return value: minimum of the set, or %HB_SET_VALUE_INVALID if set is empty.
+ *
+ * Since: 0.9.7
+ **/
+hb_codepoint_t
+hb_set_get_min (const hb_set_t *set)
+{
+  return set->get_min ();
+}
+
+/**
+ * hb_set_get_max:
+ * @set: a set.
+ *
+ * Finds the maximum number in the set.
+ *
+ * Return value: minimum of the set, or %HB_SET_VALUE_INVALID if set is empty.
+ *
+ * Since: 0.9.7
+ **/
+hb_codepoint_t
+hb_set_get_max (const hb_set_t *set)
+{
+  return set->get_max ();
+}
+
+/**
+ * hb_set_next:
+ * @set: a set.
+ * @codepoint: (inout):
+ *
+ * Gets the next number in @set that is greater than current value of @codepoint.
+ *
+ * Set @codepoint to %HB_SET_VALUE_INVALID to get started.
+ *
+ * Return value: whether there was a next value.
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_set_next (const hb_set_t *set,
+	     hb_codepoint_t *codepoint)
+{
+  return set->next (codepoint);
+}
+
+/**
+ * hb_set_previous:
+ * @set: a set.
+ * @codepoint: (inout):
+ *
+ * Gets the previous number in @set that is lower than current value of @codepoint.
+ *
+ * Set @codepoint to %HB_SET_VALUE_INVALID to get started.
+ *
+ * Return value: whether there was a previous value.
+ *
+ * Since: 1.8.0
+ **/
+hb_bool_t
+hb_set_previous (const hb_set_t *set,
+		 hb_codepoint_t *codepoint)
+{
+  return set->previous (codepoint);
+}
+
+/**
+ * hb_set_next_range:
+ * @set: a set.
+ * @first: (out): output first codepoint in the range.
+ * @last: (inout): input current last and output last codepoint in the range.
+ *
+ * Gets the next consecutive range of numbers in @set that
+ * are greater than current value of @last.
+ *
+ * Set @last to %HB_SET_VALUE_INVALID to get started.
+ *
+ * Return value: whether there was a next range.
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_set_next_range (const hb_set_t *set,
+		   hb_codepoint_t *first,
+		   hb_codepoint_t *last)
+{
+  return set->next_range (first, last);
+}
+
+/**
+ * hb_set_previous_range:
+ * @set: a set.
+ * @first: (inout): input current first and output first codepoint in the range.
+ * @last: (out): output last codepoint in the range.
+ *
+ * Gets the previous consecutive range of numbers in @set that
+ * are less than current value of @first.
+ *
+ * Set @first to %HB_SET_VALUE_INVALID to get started.
+ *
+ * Return value: whether there was a previous range.
+ *
+ * Since: 1.8.0
+ **/
+hb_bool_t
+hb_set_previous_range (const hb_set_t *set,
+		       hb_codepoint_t *first,
+		       hb_codepoint_t *last)
+{
+  return set->previous_range (first, last);
+}
diff --git a/thirdparty/harfbuzz/src/hb-set.h b/thirdparty/harfbuzz/src/hb-set.h
new file mode 100644
index 0000000000..ed0e05db2e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-set.h
@@ -0,0 +1,167 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_SET_H
+#define HB_SET_H
+
+#include "hb-common.h"
+
+HB_BEGIN_DECLS
+
+
+/*
+ * Since: 0.9.21
+ */
+#define HB_SET_VALUE_INVALID ((hb_codepoint_t) -1)
+
+typedef struct hb_set_t hb_set_t;
+
+
+HB_EXTERN hb_set_t *
+hb_set_create (void);
+
+HB_EXTERN hb_set_t *
+hb_set_get_empty (void);
+
+HB_EXTERN hb_set_t *
+hb_set_reference (hb_set_t *set);
+
+HB_EXTERN void
+hb_set_destroy (hb_set_t *set);
+
+HB_EXTERN hb_bool_t
+hb_set_set_user_data (hb_set_t           *set,
+		      hb_user_data_key_t *key,
+		      void *              data,
+		      hb_destroy_func_t   destroy,
+		      hb_bool_t           replace);
+
+HB_EXTERN void *
+hb_set_get_user_data (hb_set_t           *set,
+		      hb_user_data_key_t *key);
+
+
+/* Returns false if allocation has failed before */
+HB_EXTERN hb_bool_t
+hb_set_allocation_successful (const hb_set_t *set);
+
+HB_EXTERN void
+hb_set_clear (hb_set_t *set);
+
+HB_EXTERN hb_bool_t
+hb_set_is_empty (const hb_set_t *set);
+
+HB_EXTERN hb_bool_t
+hb_set_has (const hb_set_t *set,
+	    hb_codepoint_t  codepoint);
+
+HB_EXTERN void
+hb_set_add (hb_set_t       *set,
+	    hb_codepoint_t  codepoint);
+
+HB_EXTERN void
+hb_set_add_range (hb_set_t       *set,
+		  hb_codepoint_t  first,
+		  hb_codepoint_t  last);
+
+HB_EXTERN void
+hb_set_del (hb_set_t       *set,
+	    hb_codepoint_t  codepoint);
+
+HB_EXTERN void
+hb_set_del_range (hb_set_t       *set,
+		  hb_codepoint_t  first,
+		  hb_codepoint_t  last);
+
+HB_EXTERN hb_bool_t
+hb_set_is_equal (const hb_set_t *set,
+		 const hb_set_t *other);
+
+HB_EXTERN hb_bool_t
+hb_set_is_subset (const hb_set_t *set,
+		  const hb_set_t *larger_set);
+
+HB_EXTERN void
+hb_set_set (hb_set_t       *set,
+	    const hb_set_t *other);
+
+HB_EXTERN void
+hb_set_union (hb_set_t       *set,
+	      const hb_set_t *other);
+
+HB_EXTERN void
+hb_set_intersect (hb_set_t       *set,
+		  const hb_set_t *other);
+
+HB_EXTERN void
+hb_set_subtract (hb_set_t       *set,
+		 const hb_set_t *other);
+
+HB_EXTERN void
+hb_set_symmetric_difference (hb_set_t       *set,
+			     const hb_set_t *other);
+
+HB_EXTERN unsigned int
+hb_set_get_population (const hb_set_t *set);
+
+/* Returns HB_SET_VALUE_INVALID if set empty. */
+HB_EXTERN hb_codepoint_t
+hb_set_get_min (const hb_set_t *set);
+
+/* Returns HB_SET_VALUE_INVALID if set empty. */
+HB_EXTERN hb_codepoint_t
+hb_set_get_max (const hb_set_t *set);
+
+/* Pass HB_SET_VALUE_INVALID in to get started. */
+HB_EXTERN hb_bool_t
+hb_set_next (const hb_set_t *set,
+	     hb_codepoint_t *codepoint);
+
+/* Pass HB_SET_VALUE_INVALID in to get started. */
+HB_EXTERN hb_bool_t
+hb_set_previous (const hb_set_t *set,
+		 hb_codepoint_t *codepoint);
+
+/* Pass HB_SET_VALUE_INVALID for first and last to get started. */
+HB_EXTERN hb_bool_t
+hb_set_next_range (const hb_set_t *set,
+		   hb_codepoint_t *first,
+		   hb_codepoint_t *last);
+
+/* Pass HB_SET_VALUE_INVALID for first and last to get started. */
+HB_EXTERN hb_bool_t
+hb_set_previous_range (const hb_set_t *set,
+		       hb_codepoint_t *first,
+		       hb_codepoint_t *last);
+
+
+HB_END_DECLS
+
+#endif /* HB_SET_H */
diff --git a/thirdparty/harfbuzz/src/hb-set.hh b/thirdparty/harfbuzz/src/hb-set.hh
new file mode 100644
index 0000000000..b6e2086a2e
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-set.hh
@@ -0,0 +1,884 @@
+/*
+ * Copyright © 2012,2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SET_HH
+#define HB_SET_HH
+
+#include "hb.hh"
+#include "hb-machinery.hh"
+
+
+/*
+ * hb_set_t
+ */
+
+/* TODO Keep a free-list so we can free pages that are completely zeroed.  At that
+ * point maybe also use a sentinel value for "all-1" pages? */
+
+struct hb_set_t
+{
+  HB_DELETE_COPY_ASSIGN (hb_set_t);
+  hb_set_t ()  { init (); }
+  ~hb_set_t () { fini (); }
+
+  struct page_map_t
+  {
+    int cmp (const page_map_t &o) const { return (int) o.major - (int) major; }
+
+    uint32_t major;
+    uint32_t index;
+  };
+
+  struct page_t
+  {
+    void init0 () { v.clear (); }
+    void init1 () { v.clear (0xFF); }
+
+    unsigned int len () const
+    { return ARRAY_LENGTH_CONST (v); }
+
+    bool is_empty () const
+    {
+      for (unsigned int i = 0; i < len (); i++)
+	if (v[i])
+	  return false;
+      return true;
+    }
+
+    void add (hb_codepoint_t g) { elt (g) |= mask (g); }
+    void del (hb_codepoint_t g) { elt (g) &= ~mask (g); }
+    bool get (hb_codepoint_t g) const { return elt (g) & mask (g); }
+
+    void add_range (hb_codepoint_t a, hb_codepoint_t b)
+    {
+      elt_t *la = &elt (a);
+      elt_t *lb = &elt (b);
+      if (la == lb)
+	*la |= (mask (b) << 1) - mask(a);
+      else
+      {
+	*la |= ~(mask (a) - 1);
+	la++;
+
+	memset (la, 0xff, (char *) lb - (char *) la);
+
+	*lb |= ((mask (b) << 1) - 1);
+      }
+    }
+
+    void del_range (hb_codepoint_t a, hb_codepoint_t b)
+    {
+      elt_t *la = &elt (a);
+      elt_t *lb = &elt (b);
+      if (la == lb)
+	*la &= ~((mask (b) << 1) - mask(a));
+      else
+      {
+	*la &= mask (a) - 1;
+	la++;
+
+	memset (la, 0, (char *) lb - (char *) la);
+
+	*lb &= ~((mask (b) << 1) - 1);
+      }
+    }
+
+    bool is_equal (const page_t *other) const
+    {
+      return 0 == hb_memcmp (&v, &other->v, sizeof (v));
+    }
+
+    unsigned int get_population () const
+    {
+      unsigned int pop = 0;
+      for (unsigned int i = 0; i < len (); i++)
+	pop += hb_popcount (v[i]);
+      return pop;
+    }
+
+    bool next (hb_codepoint_t *codepoint) const
+    {
+      unsigned int m = (*codepoint + 1) & MASK;
+      if (!m)
+      {
+	*codepoint = INVALID;
+	return false;
+      }
+      unsigned int i = m / ELT_BITS;
+      unsigned int j = m & ELT_MASK;
+
+      const elt_t vv = v[i] & ~((elt_t (1) << j) - 1);
+      for (const elt_t *p = &vv; i < len (); p = &v[++i])
+	if (*p)
+	{
+	  *codepoint = i * ELT_BITS + elt_get_min (*p);
+	  return true;
+	}
+
+      *codepoint = INVALID;
+      return false;
+    }
+    bool previous (hb_codepoint_t *codepoint) const
+    {
+      unsigned int m = (*codepoint - 1) & MASK;
+      if (m == MASK)
+      {
+	*codepoint = INVALID;
+	return false;
+      }
+      unsigned int i = m / ELT_BITS;
+      unsigned int j = m & ELT_MASK;
+
+      /* Fancy mask to avoid shifting by elt_t bitsize, which is undefined. */
+      const elt_t mask = j < 8 * sizeof (elt_t) - 1 ?
+			 ((elt_t (1) << (j + 1)) - 1) :
+			 (elt_t) -1;
+      const elt_t vv = v[i] & mask;
+      const elt_t *p = &vv;
+      while (true)
+      {
+	if (*p)
+	{
+	  *codepoint = i * ELT_BITS + elt_get_max (*p);
+	  return true;
+	}
+	if ((int) i <= 0) break;
+	p = &v[--i];
+      }
+
+      *codepoint = INVALID;
+      return false;
+    }
+    hb_codepoint_t get_min () const
+    {
+      for (unsigned int i = 0; i < len (); i++)
+	if (v[i])
+	  return i * ELT_BITS + elt_get_min (v[i]);
+      return INVALID;
+    }
+    hb_codepoint_t get_max () const
+    {
+      for (int i = len () - 1; i >= 0; i--)
+	if (v[i])
+	  return i * ELT_BITS + elt_get_max (v[i]);
+      return 0;
+    }
+
+    typedef unsigned long long elt_t;
+    static constexpr unsigned PAGE_BITS = 512;
+    static_assert ((PAGE_BITS & ((PAGE_BITS) - 1)) == 0, "");
+
+    static unsigned int elt_get_min (const elt_t &elt) { return hb_ctz (elt); }
+    static unsigned int elt_get_max (const elt_t &elt) { return hb_bit_storage (elt) - 1; }
+
+    typedef hb_vector_size_t<elt_t, PAGE_BITS / 8> vector_t;
+
+    static constexpr unsigned ELT_BITS = sizeof (elt_t) * 8;
+    static constexpr unsigned ELT_MASK = ELT_BITS - 1;
+    static constexpr unsigned BITS = sizeof (vector_t) * 8;
+    static constexpr unsigned MASK = BITS - 1;
+    static_assert ((unsigned) PAGE_BITS == (unsigned) BITS, "");
+
+    elt_t &elt (hb_codepoint_t g) { return v[(g & MASK) / ELT_BITS]; }
+    elt_t const &elt (hb_codepoint_t g) const { return v[(g & MASK) / ELT_BITS]; }
+    elt_t mask (hb_codepoint_t g) const { return elt_t (1) << (g & ELT_MASK); }
+
+    vector_t v;
+  };
+  static_assert (page_t::PAGE_BITS == sizeof (page_t) * 8, "");
+
+  hb_object_header_t header;
+  bool successful; /* Allocations successful */
+  mutable unsigned int population;
+  hb_sorted_vector_t<page_map_t> page_map;
+  hb_vector_t<page_t> pages;
+
+  void init_shallow ()
+  {
+    successful = true;
+    population = 0;
+    page_map.init ();
+    pages.init ();
+  }
+  void init ()
+  {
+    hb_object_init (this);
+    init_shallow ();
+  }
+  void fini_shallow ()
+  {
+    population = 0;
+    page_map.fini ();
+    pages.fini ();
+  }
+  void fini ()
+  {
+    hb_object_fini (this);
+    fini_shallow ();
+  }
+
+  bool in_error () const { return !successful; }
+
+  bool resize (unsigned int count)
+  {
+    if (unlikely (!successful)) return false;
+    if (!pages.resize (count) || !page_map.resize (count))
+    {
+      pages.resize (page_map.length);
+      successful = false;
+      return false;
+    }
+    return true;
+  }
+
+  void reset ()
+  {
+    if (unlikely (hb_object_is_immutable (this)))
+      return;
+    clear ();
+    successful = true;
+  }
+
+  void clear ()
+  {
+    if (unlikely (hb_object_is_immutable (this)))
+      return;
+    population = 0;
+    page_map.resize (0);
+    pages.resize (0);
+  }
+  bool is_empty () const
+  {
+    unsigned int count = pages.length;
+    for (unsigned int i = 0; i < count; i++)
+      if (!pages[i].is_empty ())
+	return false;
+    return true;
+  }
+
+  void dirty () { population = UINT_MAX; }
+
+  void add (hb_codepoint_t g)
+  {
+    if (unlikely (!successful)) return;
+    if (unlikely (g == INVALID)) return;
+    dirty ();
+    page_t *page = page_for_insert (g); if (unlikely (!page)) return;
+    page->add (g);
+  }
+  bool add_range (hb_codepoint_t a, hb_codepoint_t b)
+  {
+    if (unlikely (!successful)) return true; /* https://github.com/harfbuzz/harfbuzz/issues/657 */
+    if (unlikely (a > b || a == INVALID || b == INVALID)) return false;
+    dirty ();
+    unsigned int ma = get_major (a);
+    unsigned int mb = get_major (b);
+    if (ma == mb)
+    {
+      page_t *page = page_for_insert (a); if (unlikely (!page)) return false;
+      page->add_range (a, b);
+    }
+    else
+    {
+      page_t *page = page_for_insert (a); if (unlikely (!page)) return false;
+      page->add_range (a, major_start (ma + 1) - 1);
+
+      for (unsigned int m = ma + 1; m < mb; m++)
+      {
+	page = page_for_insert (major_start (m)); if (unlikely (!page)) return false;
+	page->init1 ();
+      }
+
+      page = page_for_insert (b); if (unlikely (!page)) return false;
+      page->add_range (major_start (mb), b);
+    }
+    return true;
+  }
+
+  template <typename T>
+  void add_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    if (unlikely (!successful)) return;
+    if (!count) return;
+    dirty ();
+    hb_codepoint_t g = *array;
+    while (count)
+    {
+      unsigned int m = get_major (g);
+      page_t *page = page_for_insert (g); if (unlikely (!page)) return;
+      unsigned int start = major_start (m);
+      unsigned int end = major_start (m + 1);
+      do
+      {
+	page->add (g);
+
+	array = &StructAtOffsetUnaligned<T> (array, stride);
+	count--;
+      }
+      while (count && (g = *array, start <= g && g < end));
+    }
+  }
+
+  /* Might return false if array looks unsorted.
+   * Used for faster rejection of corrupt data. */
+  template <typename T>
+  bool add_sorted_array (const T *array, unsigned int count, unsigned int stride=sizeof(T))
+  {
+    if (unlikely (!successful)) return true; /* https://github.com/harfbuzz/harfbuzz/issues/657 */
+    if (!count) return true;
+    dirty ();
+    hb_codepoint_t g = *array;
+    hb_codepoint_t last_g = g;
+    while (count)
+    {
+      unsigned int m = get_major (g);
+      page_t *page = page_for_insert (g); if (unlikely (!page)) return false;
+      unsigned int end = major_start (m + 1);
+      do
+      {
+	/* If we try harder we can change the following comparison to <=;
+	 * Not sure if it's worth it. */
+	if (g < last_g) return false;
+	last_g = g;
+	page->add (g);
+
+	array = (const T *) ((const char *) array + stride);
+	count--;
+      }
+      while (count && (g = *array, g < end));
+    }
+    return true;
+  }
+
+  void del (hb_codepoint_t g)
+  {
+    /* TODO perform op even if !successful. */
+    if (unlikely (!successful)) return;
+    page_t *page = page_for (g);
+    if (!page)
+      return;
+    dirty ();
+    page->del (g);
+  }
+
+  private:
+  void del_pages (int ds, int de)
+  {
+    if (ds <= de)
+    {
+      unsigned int write_index = 0;
+      for (unsigned int i = 0; i < page_map.length; i++)
+      {
+	int m = (int) page_map[i].major;
+	if (m < ds || de < m)
+	  page_map[write_index++] = page_map[i];
+      }
+      compact (write_index);
+      resize (write_index);
+    }
+  }
+
+  public:
+  void del_range (hb_codepoint_t a, hb_codepoint_t b)
+  {
+    /* TODO perform op even if !successful. */
+    if (unlikely (!successful)) return;
+    if (unlikely (a > b || a == INVALID || b == INVALID)) return;
+    dirty ();
+    unsigned int ma = get_major (a);
+    unsigned int mb = get_major (b);
+    /* Delete pages from ds through de if ds <= de. */
+    int ds = (a == major_start (ma))? (int) ma: (int) (ma + 1);
+    int de = (b + 1 == major_start (mb + 1))? (int) mb: ((int) mb - 1);
+    if (ds > de || (int) ma < ds)
+    {
+      page_t *page = page_for (a);
+      if (page)
+      {
+	if (ma == mb)
+	  page->del_range (a, b);
+	else
+	  page->del_range (a, major_start (ma + 1) - 1);
+      }
+    }
+    if (de < (int) mb && ma != mb)
+    {
+      page_t *page = page_for (b);
+      if (page)
+	page->del_range (major_start (mb), b);
+    }
+    del_pages (ds, de);
+  }
+
+  bool get (hb_codepoint_t g) const
+  {
+    const page_t *page = page_for (g);
+    if (!page)
+      return false;
+    return page->get (g);
+  }
+
+  /* Has interface. */
+  static constexpr bool SENTINEL = false;
+  typedef bool value_t;
+  value_t operator [] (hb_codepoint_t k) const { return get (k); }
+  bool has (hb_codepoint_t k) const { return (*this)[k] != SENTINEL; }
+  /* Predicate. */
+  bool operator () (hb_codepoint_t k) const { return has (k); }
+
+  /* Sink interface. */
+  hb_set_t& operator << (hb_codepoint_t v)
+  { add (v); return *this; }
+  hb_set_t& operator << (const hb_pair_t<hb_codepoint_t, hb_codepoint_t>& range)
+  { add_range (range.first, range.second); return *this; }
+
+  bool intersects (hb_codepoint_t first, hb_codepoint_t last) const
+  {
+    hb_codepoint_t c = first - 1;
+    return next (&c) && c <= last;
+  }
+  void set (const hb_set_t *other)
+  {
+    if (unlikely (!successful)) return;
+    unsigned int count = other->pages.length;
+    if (!resize (count))
+      return;
+    population = other->population;
+    memcpy ((void *) pages, (const void *) other->pages, count * pages.item_size);
+    memcpy ((void *) page_map, (const void *) other->page_map, count * page_map.item_size);
+  }
+
+  bool is_equal (const hb_set_t *other) const
+  {
+    if (get_population () != other->get_population ())
+      return false;
+
+    unsigned int na = pages.length;
+    unsigned int nb = other->pages.length;
+
+    unsigned int a = 0, b = 0;
+    for (; a < na && b < nb; )
+    {
+      if (page_at (a).is_empty ()) { a++; continue; }
+      if (other->page_at (b).is_empty ()) { b++; continue; }
+      if (page_map[a].major != other->page_map[b].major ||
+	  !page_at (a).is_equal (&other->page_at (b)))
+	return false;
+      a++;
+      b++;
+    }
+    for (; a < na; a++)
+      if (!page_at (a).is_empty ()) { return false; }
+    for (; b < nb; b++)
+      if (!other->page_at (b).is_empty ()) { return false; }
+
+    return true;
+  }
+
+  bool is_subset (const hb_set_t *larger_set) const
+  {
+    if (get_population () > larger_set->get_population ())
+      return false;
+
+    /* TODO Optimize to use pages. */
+    hb_codepoint_t c = INVALID;
+    while (next (&c))
+      if (!larger_set->has (c))
+	return false;
+
+    return true;
+  }
+
+  void compact (unsigned int length)
+  {
+    hb_vector_t<uint32_t> old_index_to_page_map_index;
+    old_index_to_page_map_index.resize(pages.length);
+    for (uint32_t i = 0; i < old_index_to_page_map_index.length; i++)
+      old_index_to_page_map_index[i] = 0xFFFFFFFF;
+
+    for (uint32_t i = 0; i < length; i++)
+      old_index_to_page_map_index[page_map[i].index] =  i;
+
+    compact_pages (old_index_to_page_map_index);
+  }
+
+  void compact_pages (const hb_vector_t<uint32_t>& old_index_to_page_map_index)
+  {
+    unsigned int write_index = 0;
+    for (unsigned int i = 0; i < pages.length; i++)
+    {
+      if (old_index_to_page_map_index[i] == 0xFFFFFFFF) continue;
+
+      if (write_index < i)
+	pages[write_index] = pages[i];
+
+      page_map[old_index_to_page_map_index[i]].index = write_index;
+      write_index++;
+    }
+  }
+
+  template <typename Op>
+  void process (const Op& op, const hb_set_t *other)
+  {
+    if (unlikely (!successful)) return;
+
+    dirty ();
+
+    unsigned int na = pages.length;
+    unsigned int nb = other->pages.length;
+    unsigned int next_page = na;
+
+    unsigned int count = 0, newCount = 0;
+    unsigned int a = 0, b = 0;
+    unsigned int write_index = 0;
+    for (; a < na && b < nb; )
+    {
+      if (page_map[a].major == other->page_map[b].major)
+      {
+	if (!Op::passthru_left)
+	{
+	  // Move page_map entries that we're keeping from the left side set
+	  // to the front of the page_map vector. This isn't necessary if
+	  // passthru_left is set since no left side pages will be removed
+	  // in that case.
+	  if (write_index < a)
+	    page_map[write_index] = page_map[a];
+	  write_index++;
+	}
+
+	count++;
+	a++;
+	b++;
+      }
+      else if (page_map[a].major < other->page_map[b].major)
+      {
+	if (Op::passthru_left)
+	  count++;
+	a++;
+      }
+      else
+      {
+	if (Op::passthru_right)
+	  count++;
+	b++;
+      }
+    }
+    if (Op::passthru_left)
+      count += na - a;
+    if (Op::passthru_right)
+      count += nb - b;
+
+    if (!Op::passthru_left)
+    {
+      na  = write_index;
+      next_page = write_index;
+      compact (write_index);
+    }
+
+    if (!resize (count))
+      return;
+
+    newCount = count;
+
+    /* Process in-place backward. */
+    a = na;
+    b = nb;
+    for (; a && b; )
+    {
+      if (page_map[a - 1].major == other->page_map[b - 1].major)
+      {
+	a--;
+	b--;
+	count--;
+	page_map[count] = page_map[a];
+	page_at (count).v = op (page_at (a).v, other->page_at (b).v);
+      }
+      else if (page_map[a - 1].major > other->page_map[b - 1].major)
+      {
+	a--;
+	if (Op::passthru_left)
+	{
+	  count--;
+	  page_map[count] = page_map[a];
+	}
+      }
+      else
+      {
+	b--;
+	if (Op::passthru_right)
+	{
+	  count--;
+	  page_map[count].major = other->page_map[b].major;
+	  page_map[count].index = next_page++;
+	  page_at (count).v = other->page_at (b).v;
+	}
+      }
+    }
+    if (Op::passthru_left)
+      while (a)
+      {
+	a--;
+	count--;
+	page_map[count] = page_map [a];
+      }
+    if (Op::passthru_right)
+      while (b)
+      {
+	b--;
+	count--;
+	page_map[count].major = other->page_map[b].major;
+	page_map[count].index = next_page++;
+	page_at (count).v = other->page_at (b).v;
+      }
+    assert (!count);
+    if (pages.length > newCount)
+      resize (newCount);
+  }
+
+  void union_ (const hb_set_t *other)
+  {
+    process (hb_bitwise_or, other);
+  }
+  void intersect (const hb_set_t *other)
+  {
+    process (hb_bitwise_and, other);
+  }
+  void subtract (const hb_set_t *other)
+  {
+    process (hb_bitwise_sub, other);
+  }
+  void symmetric_difference (const hb_set_t *other)
+  {
+    process (hb_bitwise_xor, other);
+  }
+  bool next (hb_codepoint_t *codepoint) const
+  {
+    if (unlikely (*codepoint == INVALID)) {
+      *codepoint = get_min ();
+      return *codepoint != INVALID;
+    }
+
+    page_map_t map = {get_major (*codepoint), 0};
+    unsigned int i;
+    page_map.bfind (map, &i, HB_BFIND_NOT_FOUND_STORE_CLOSEST);
+    if (i < page_map.length && page_map[i].major == map.major)
+    {
+      if (pages[page_map[i].index].next (codepoint))
+      {
+	*codepoint += page_map[i].major * page_t::PAGE_BITS;
+	return true;
+      }
+      i++;
+    }
+    for (; i < page_map.length; i++)
+    {
+      hb_codepoint_t m = pages[page_map[i].index].get_min ();
+      if (m != INVALID)
+      {
+	*codepoint = page_map[i].major * page_t::PAGE_BITS + m;
+	return true;
+      }
+    }
+    *codepoint = INVALID;
+    return false;
+  }
+  bool previous (hb_codepoint_t *codepoint) const
+  {
+    if (unlikely (*codepoint == INVALID)) {
+      *codepoint = get_max ();
+      return *codepoint != INVALID;
+    }
+
+    page_map_t map = {get_major (*codepoint), 0};
+    unsigned int i;
+    page_map.bfind (map, &i, HB_BFIND_NOT_FOUND_STORE_CLOSEST);
+    if (i < page_map.length && page_map[i].major == map.major)
+    {
+      if (pages[page_map[i].index].previous (codepoint))
+      {
+	*codepoint += page_map[i].major * page_t::PAGE_BITS;
+	return true;
+      }
+    }
+    i--;
+    for (; (int) i >= 0; i--)
+    {
+      hb_codepoint_t m = pages[page_map[i].index].get_max ();
+      if (m != INVALID)
+      {
+	*codepoint = page_map[i].major * page_t::PAGE_BITS + m;
+	return true;
+      }
+    }
+    *codepoint = INVALID;
+    return false;
+  }
+  bool next_range (hb_codepoint_t *first, hb_codepoint_t *last) const
+  {
+    hb_codepoint_t i;
+
+    i = *last;
+    if (!next (&i))
+    {
+      *last = *first = INVALID;
+      return false;
+    }
+
+    /* TODO Speed up. */
+    *last = *first = i;
+    while (next (&i) && i == *last + 1)
+      (*last)++;
+
+    return true;
+  }
+  bool previous_range (hb_codepoint_t *first, hb_codepoint_t *last) const
+  {
+    hb_codepoint_t i;
+
+    i = *first;
+    if (!previous (&i))
+    {
+      *last = *first = INVALID;
+      return false;
+    }
+
+    /* TODO Speed up. */
+    *last = *first = i;
+    while (previous (&i) && i == *first - 1)
+      (*first)--;
+
+    return true;
+  }
+
+  unsigned int get_population () const
+  {
+    if (population != UINT_MAX)
+      return population;
+
+    unsigned int pop = 0;
+    unsigned int count = pages.length;
+    for (unsigned int i = 0; i < count; i++)
+      pop += pages[i].get_population ();
+
+    population = pop;
+    return pop;
+  }
+  hb_codepoint_t get_min () const
+  {
+    unsigned int count = pages.length;
+    for (unsigned int i = 0; i < count; i++)
+      if (!page_at (i).is_empty ())
+	return page_map[i].major * page_t::PAGE_BITS + page_at (i).get_min ();
+    return INVALID;
+  }
+  hb_codepoint_t get_max () const
+  {
+    unsigned int count = pages.length;
+    for (int i = count - 1; i >= 0; i++)
+      if (!page_at (i).is_empty ())
+	return page_map[(unsigned) i].major * page_t::PAGE_BITS + page_at (i).get_max ();
+    return INVALID;
+  }
+
+  static constexpr hb_codepoint_t INVALID = HB_SET_VALUE_INVALID;
+
+  /*
+   * Iterator implementation.
+   */
+  struct iter_t : hb_iter_with_fallback_t<iter_t, hb_codepoint_t>
+  {
+    static constexpr bool is_sorted_iterator = true;
+    iter_t (const hb_set_t &s_ = Null (hb_set_t),
+	    bool init = true) : s (&s_), v (INVALID), l(0)
+    {
+      if (init)
+      {
+	l = s->get_population () + 1;
+	__next__ ();
+      }
+    }
+
+    typedef hb_codepoint_t __item_t__;
+    hb_codepoint_t __item__ () const { return v; }
+    bool __more__ () const { return v != INVALID; }
+    void __next__ () { s->next (&v); if (l) l--; }
+    void __prev__ () { s->previous (&v); }
+    unsigned __len__ () const { return l; }
+    iter_t end () const { return iter_t (*s, false); }
+    bool operator != (const iter_t& o) const
+    { return s != o.s || v != o.v; }
+
+    protected:
+    const hb_set_t *s;
+    hb_codepoint_t v;
+    unsigned l;
+  };
+  iter_t iter () const { return iter_t (*this); }
+  operator iter_t () const { return iter (); }
+
+  protected:
+
+  page_t *page_for_insert (hb_codepoint_t g)
+  {
+    page_map_t map = {get_major (g), pages.length};
+    unsigned int i;
+    if (!page_map.bfind (map, &i, HB_BFIND_NOT_FOUND_STORE_CLOSEST))
+    {
+      if (!resize (pages.length + 1))
+	return nullptr;
+
+      pages[map.index].init0 ();
+      memmove (page_map + i + 1,
+	       page_map + i,
+	       (page_map.length - 1 - i) * page_map.item_size);
+      page_map[i] = map;
+    }
+    return &pages[page_map[i].index];
+  }
+  page_t *page_for (hb_codepoint_t g)
+  {
+    page_map_t key = {get_major (g)};
+    const page_map_t *found = page_map.bsearch (key);
+    if (found)
+      return &pages[found->index];
+    return nullptr;
+  }
+  const page_t *page_for (hb_codepoint_t g) const
+  {
+    page_map_t key = {get_major (g)};
+    const page_map_t *found = page_map.bsearch (key);
+    if (found)
+      return &pages[found->index];
+    return nullptr;
+  }
+  page_t &page_at (unsigned int i) { return pages[page_map[i].index]; }
+  const page_t &page_at (unsigned int i) const { return pages[page_map[i].index]; }
+  unsigned int get_major (hb_codepoint_t g) const { return g / page_t::PAGE_BITS; }
+  hb_codepoint_t major_start (unsigned int major) const { return major * page_t::PAGE_BITS; }
+};
+
+
+#endif /* HB_SET_HH */
diff --git a/thirdparty/harfbuzz/src/hb-shape-plan.cc b/thirdparty/harfbuzz/src/hb-shape-plan.cc
new file mode 100644
index 0000000000..666470b4f1
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shape-plan.cc
@@ -0,0 +1,513 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+#include "hb-shape-plan.hh"
+#include "hb-shaper.hh"
+#include "hb-font.hh"
+#include "hb-buffer.hh"
+
+
+/**
+ * SECTION:hb-shape-plan
+ * @title: hb-shape-plan
+ * @short_description: Object representing a shaping plan
+ * @include: hb.h
+ *
+ * Shape plans are not used for shaping directly, but can be access to query
+ * certain information about how shaping will perform given a set of input
+ * parameters (script, language, direction, features, etc.)
+ * Most client would not need to deal with shape plans directly.
+ **/
+
+
+/*
+ * hb_shape_plan_key_t
+ */
+
+bool
+hb_shape_plan_key_t::init (bool                           copy,
+			   hb_face_t                     *face,
+			   const hb_segment_properties_t *props,
+			   const hb_feature_t            *user_features,
+			   unsigned int                   num_user_features,
+			   const int                     *coords,
+			   unsigned int                   num_coords,
+			   const char * const            *shaper_list)
+{
+  hb_feature_t *features = nullptr;
+  if (copy && num_user_features && !(features = (hb_feature_t *) calloc (num_user_features, sizeof (hb_feature_t))))
+    goto bail;
+
+  this->props = *props;
+  this->num_user_features = num_user_features;
+  this->user_features = copy ? features : user_features;
+  if (copy && num_user_features)
+  {
+    memcpy (features, user_features, num_user_features * sizeof (hb_feature_t));
+    /* Make start/end uniform to easier catch bugs. */
+    for (unsigned int i = 0; i < num_user_features; i++)
+    {
+      if (features[0].start != HB_FEATURE_GLOBAL_START)
+	features[0].start = 1;
+      if (features[0].end   != HB_FEATURE_GLOBAL_END)
+	features[0].end   = 2;
+    }
+  }
+  this->shaper_func = nullptr;
+  this->shaper_name = nullptr;
+#ifndef HB_NO_OT_SHAPE
+  this->ot.init (face, coords, num_coords);
+#endif
+
+  /*
+   * Choose shaper.
+   */
+
+#define HB_SHAPER_PLAN(shaper) \
+	HB_STMT_START { \
+	  if (face->data.shaper) \
+	  { \
+	    this->shaper_func = _hb_##shaper##_shape; \
+	    this->shaper_name = #shaper; \
+	    return true; \
+	  } \
+	} HB_STMT_END
+
+  if (unlikely (shaper_list))
+  {
+    for (; *shaper_list; shaper_list++)
+      if (false)
+	;
+#define HB_SHAPER_IMPLEMENT(shaper) \
+      else if (0 == strcmp (*shaper_list, #shaper)) \
+	HB_SHAPER_PLAN (shaper);
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+  }
+  else
+  {
+    const hb_shaper_entry_t *shapers = _hb_shapers_get ();
+    for (unsigned int i = 0; i < HB_SHAPERS_COUNT; i++)
+      if (false)
+	;
+#define HB_SHAPER_IMPLEMENT(shaper) \
+      else if (shapers[i].func == _hb_##shaper##_shape) \
+	HB_SHAPER_PLAN (shaper);
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+  }
+#undef HB_SHAPER_PLAN
+
+bail:
+  ::free (features);
+  return false;
+}
+
+bool
+hb_shape_plan_key_t::user_features_match (const hb_shape_plan_key_t *other)
+{
+  if (this->num_user_features != other->num_user_features)
+    return false;
+  for (unsigned int i = 0; i < num_user_features; i++)
+  {
+    if (this->user_features[i].tag   != other->user_features[i].tag   ||
+	this->user_features[i].value != other->user_features[i].value ||
+	(this->user_features[i].start == HB_FEATURE_GLOBAL_START &&
+	 this->user_features[i].end   == HB_FEATURE_GLOBAL_END) !=
+	(other->user_features[i].start == HB_FEATURE_GLOBAL_START &&
+	 other->user_features[i].end   == HB_FEATURE_GLOBAL_END))
+      return false;
+  }
+  return true;
+}
+
+bool
+hb_shape_plan_key_t::equal (const hb_shape_plan_key_t *other)
+{
+  return hb_segment_properties_equal (&this->props, &other->props) &&
+	 this->user_features_match (other) &&
+#ifndef HB_NO_OT_SHAPE
+	 this->ot.equal (&other->ot) &&
+#endif
+	 this->shaper_func == other->shaper_func;
+}
+
+
+/*
+ * hb_shape_plan_t
+ */
+
+
+/**
+ * hb_shape_plan_create: (Xconstructor)
+ * @face:
+ * @props:
+ * @user_features: (array length=num_user_features):
+ * @num_user_features:
+ * @shaper_list: (array zero-terminated=1):
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.7
+ **/
+hb_shape_plan_t *
+hb_shape_plan_create (hb_face_t                     *face,
+		      const hb_segment_properties_t *props,
+		      const hb_feature_t            *user_features,
+		      unsigned int                   num_user_features,
+		      const char * const            *shaper_list)
+{
+  return hb_shape_plan_create2 (face, props,
+				user_features, num_user_features,
+				nullptr, 0,
+				shaper_list);
+}
+
+hb_shape_plan_t *
+hb_shape_plan_create2 (hb_face_t                     *face,
+		       const hb_segment_properties_t *props,
+		       const hb_feature_t            *user_features,
+		       unsigned int                   num_user_features,
+		       const int                     *coords,
+		       unsigned int                   num_coords,
+		       const char * const            *shaper_list)
+{
+  DEBUG_MSG_FUNC (SHAPE_PLAN, nullptr,
+		  "face=%p num_features=%d num_coords=%d shaper_list=%p",
+		  face,
+		  num_user_features,
+		  num_coords,
+		  shaper_list);
+
+  assert (props->direction != HB_DIRECTION_INVALID);
+
+  hb_shape_plan_t *shape_plan;
+
+  if (unlikely (!props))
+    goto bail;
+  if (!(shape_plan = hb_object_create<hb_shape_plan_t> ()))
+    goto bail;
+
+  if (unlikely (!face))
+    face = hb_face_get_empty ();
+  hb_face_make_immutable (face);
+  shape_plan->face_unsafe = face;
+
+  if (unlikely (!shape_plan->key.init (true,
+				       face,
+				       props,
+				       user_features,
+				       num_user_features,
+				       coords,
+				       num_coords,
+				       shaper_list)))
+    goto bail2;
+#ifndef HB_NO_OT_SHAPE
+  if (unlikely (!shape_plan->ot.init0 (face, &shape_plan->key)))
+    goto bail3;
+#endif
+
+  return shape_plan;
+
+#ifndef HB_NO_OT_SHAPE
+bail3:
+#endif
+  shape_plan->key.free ();
+bail2:
+  free (shape_plan);
+bail:
+  return hb_shape_plan_get_empty ();
+}
+
+/**
+ * hb_shape_plan_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.7
+ **/
+hb_shape_plan_t *
+hb_shape_plan_get_empty ()
+{
+  return const_cast<hb_shape_plan_t *> (&Null (hb_shape_plan_t));
+}
+
+/**
+ * hb_shape_plan_reference: (skip)
+ * @shape_plan: a shape plan.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.7
+ **/
+hb_shape_plan_t *
+hb_shape_plan_reference (hb_shape_plan_t *shape_plan)
+{
+  return hb_object_reference (shape_plan);
+}
+
+/**
+ * hb_shape_plan_destroy: (skip)
+ * @shape_plan: a shape plan.
+ *
+ *
+ *
+ * Since: 0.9.7
+ **/
+void
+hb_shape_plan_destroy (hb_shape_plan_t *shape_plan)
+{
+  if (!hb_object_destroy (shape_plan)) return;
+
+#ifndef HB_NO_OT_SHAPE
+  shape_plan->ot.fini ();
+#endif
+  shape_plan->key.free ();
+  free (shape_plan);
+}
+
+/**
+ * hb_shape_plan_set_user_data: (skip)
+ * @shape_plan: a shape plan.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_shape_plan_set_user_data (hb_shape_plan_t    *shape_plan,
+			     hb_user_data_key_t *key,
+			     void *              data,
+			     hb_destroy_func_t   destroy,
+			     hb_bool_t           replace)
+{
+  return hb_object_set_user_data (shape_plan, key, data, destroy, replace);
+}
+
+/**
+ * hb_shape_plan_get_user_data: (skip)
+ * @shape_plan: a shape plan.
+ * @key:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.7
+ **/
+void *
+hb_shape_plan_get_user_data (hb_shape_plan_t    *shape_plan,
+			     hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (shape_plan, key);
+}
+
+/**
+ * hb_shape_plan_get_shaper:
+ * @shape_plan: a shape plan.
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.7
+ **/
+const char *
+hb_shape_plan_get_shaper (hb_shape_plan_t *shape_plan)
+{
+  return shape_plan->key.shaper_name;
+}
+
+
+/**
+ * hb_shape_plan_execute:
+ * @shape_plan: a shape plan.
+ * @font: a font.
+ * @buffer: a buffer.
+ * @features: (array length=num_features):
+ * @num_features:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.7
+ **/
+hb_bool_t
+hb_shape_plan_execute (hb_shape_plan_t    *shape_plan,
+		       hb_font_t          *font,
+		       hb_buffer_t        *buffer,
+		       const hb_feature_t *features,
+		       unsigned int        num_features)
+{
+  DEBUG_MSG_FUNC (SHAPE_PLAN, shape_plan,
+		  "num_features=%d shaper_func=%p, shaper_name=%s",
+		  num_features,
+		  shape_plan->key.shaper_func,
+		  shape_plan->key.shaper_name);
+
+  if (unlikely (!buffer->len))
+    return true;
+
+  assert (!hb_object_is_immutable (buffer));
+  assert (buffer->content_type == HB_BUFFER_CONTENT_TYPE_UNICODE);
+
+  if (unlikely (hb_object_is_inert (shape_plan)))
+    return false;
+
+  assert (shape_plan->face_unsafe == font->face);
+  assert (hb_segment_properties_equal (&shape_plan->key.props, &buffer->props));
+
+#define HB_SHAPER_EXECUTE(shaper) \
+	HB_STMT_START { \
+	  return font->data.shaper && \
+		 _hb_##shaper##_shape (shape_plan, font, buffer, features, num_features); \
+	} HB_STMT_END
+
+  if (false)
+    ;
+#define HB_SHAPER_IMPLEMENT(shaper) \
+  else if (shape_plan->key.shaper_func == _hb_##shaper##_shape) \
+    HB_SHAPER_EXECUTE (shaper);
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+
+#undef HB_SHAPER_EXECUTE
+
+  return false;
+}
+
+
+/*
+ * Caching
+ */
+
+/**
+ * hb_shape_plan_create_cached:
+ * @face:
+ * @props:
+ * @user_features: (array length=num_user_features):
+ * @num_user_features:
+ * @shaper_list: (array zero-terminated=1):
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.7
+ **/
+hb_shape_plan_t *
+hb_shape_plan_create_cached (hb_face_t                     *face,
+			     const hb_segment_properties_t *props,
+			     const hb_feature_t            *user_features,
+			     unsigned int                   num_user_features,
+			     const char * const            *shaper_list)
+{
+  return hb_shape_plan_create_cached2 (face, props,
+				       user_features, num_user_features,
+				       nullptr, 0,
+				       shaper_list);
+}
+
+hb_shape_plan_t *
+hb_shape_plan_create_cached2 (hb_face_t                     *face,
+			      const hb_segment_properties_t *props,
+			      const hb_feature_t            *user_features,
+			      unsigned int                   num_user_features,
+			      const int                     *coords,
+			      unsigned int                   num_coords,
+			      const char * const            *shaper_list)
+{
+  DEBUG_MSG_FUNC (SHAPE_PLAN, nullptr,
+		  "face=%p num_features=%d shaper_list=%p",
+		  face,
+		  num_user_features,
+		  shaper_list);
+
+retry:
+  hb_face_t::plan_node_t *cached_plan_nodes = face->shape_plans;
+
+  bool dont_cache = hb_object_is_inert (face);
+
+  if (likely (!dont_cache))
+  {
+    hb_shape_plan_key_t key;
+    if (!key.init (false,
+		   face,
+		   props,
+		   user_features,
+		   num_user_features,
+		   coords,
+		   num_coords,
+		   shaper_list))
+      return hb_shape_plan_get_empty ();
+
+    for (hb_face_t::plan_node_t *node = cached_plan_nodes; node; node = node->next)
+      if (node->shape_plan->key.equal (&key))
+      {
+	DEBUG_MSG_FUNC (SHAPE_PLAN, node->shape_plan, "fulfilled from cache");
+	return hb_shape_plan_reference (node->shape_plan);
+      }
+  }
+
+  hb_shape_plan_t *shape_plan = hb_shape_plan_create2 (face, props,
+						       user_features, num_user_features,
+						       coords, num_coords,
+						       shaper_list);
+
+  if (unlikely (dont_cache))
+    return shape_plan;
+
+  hb_face_t::plan_node_t *node = (hb_face_t::plan_node_t *) calloc (1, sizeof (hb_face_t::plan_node_t));
+  if (unlikely (!node))
+    return shape_plan;
+
+  node->shape_plan = shape_plan;
+  node->next = cached_plan_nodes;
+
+  if (unlikely (!face->shape_plans.cmpexch (cached_plan_nodes, node)))
+  {
+    hb_shape_plan_destroy (shape_plan);
+    free (node);
+    goto retry;
+  }
+  DEBUG_MSG_FUNC (SHAPE_PLAN, shape_plan, "inserted into cache");
+
+  return hb_shape_plan_reference (shape_plan);
+}
diff --git a/thirdparty/harfbuzz/src/hb-shape-plan.h b/thirdparty/harfbuzz/src/hb-shape-plan.h
new file mode 100644
index 0000000000..b62ae7ca35
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shape-plan.h
@@ -0,0 +1,108 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_SHAPE_PLAN_H
+#define HB_SHAPE_PLAN_H
+
+#include "hb-common.h"
+#include "hb-font.h"
+
+HB_BEGIN_DECLS
+
+typedef struct hb_shape_plan_t hb_shape_plan_t;
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_create (hb_face_t                     *face,
+		      const hb_segment_properties_t *props,
+		      const hb_feature_t            *user_features,
+		      unsigned int                   num_user_features,
+		      const char * const            *shaper_list);
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_create_cached (hb_face_t                     *face,
+			     const hb_segment_properties_t *props,
+			     const hb_feature_t            *user_features,
+			     unsigned int                   num_user_features,
+			     const char * const            *shaper_list);
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_create2 (hb_face_t                     *face,
+		       const hb_segment_properties_t *props,
+		       const hb_feature_t            *user_features,
+		       unsigned int                   num_user_features,
+		       const int                     *coords,
+		       unsigned int                   num_coords,
+		       const char * const            *shaper_list);
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_create_cached2 (hb_face_t                     *face,
+			      const hb_segment_properties_t *props,
+			      const hb_feature_t            *user_features,
+			      unsigned int                   num_user_features,
+			      const int                     *coords,
+			      unsigned int                   num_coords,
+			      const char * const            *shaper_list);
+
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_get_empty (void);
+
+HB_EXTERN hb_shape_plan_t *
+hb_shape_plan_reference (hb_shape_plan_t *shape_plan);
+
+HB_EXTERN void
+hb_shape_plan_destroy (hb_shape_plan_t *shape_plan);
+
+HB_EXTERN hb_bool_t
+hb_shape_plan_set_user_data (hb_shape_plan_t    *shape_plan,
+			     hb_user_data_key_t *key,
+			     void *              data,
+			     hb_destroy_func_t   destroy,
+			     hb_bool_t           replace);
+
+HB_EXTERN void *
+hb_shape_plan_get_user_data (hb_shape_plan_t    *shape_plan,
+			     hb_user_data_key_t *key);
+
+
+HB_EXTERN hb_bool_t
+hb_shape_plan_execute (hb_shape_plan_t    *shape_plan,
+		       hb_font_t          *font,
+		       hb_buffer_t        *buffer,
+		       const hb_feature_t *features,
+		       unsigned int        num_features);
+
+HB_EXTERN const char *
+hb_shape_plan_get_shaper (hb_shape_plan_t *shape_plan);
+
+
+HB_END_DECLS
+
+#endif /* HB_SHAPE_PLAN_H */
diff --git a/thirdparty/harfbuzz/src/hb-shape-plan.hh b/thirdparty/harfbuzz/src/hb-shape-plan.hh
new file mode 100644
index 0000000000..6da7edb2f8
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shape-plan.hh
@@ -0,0 +1,76 @@
+/*
+ * Copyright © 2012,2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SHAPE_PLAN_HH
+#define HB_SHAPE_PLAN_HH
+
+#include "hb.hh"
+#include "hb-shaper.hh"
+#include "hb-ot-shape.hh"
+
+
+struct hb_shape_plan_key_t
+{
+  hb_segment_properties_t  props;
+
+  const hb_feature_t      *user_features;
+  unsigned int             num_user_features;
+
+#ifndef HB_NO_OT_SHAPE
+  hb_ot_shape_plan_key_t   ot;
+#endif
+
+  hb_shape_func_t         *shaper_func;
+  const char              *shaper_name;
+
+  HB_INTERNAL bool init (bool                           copy,
+			 hb_face_t                     *face,
+			 const hb_segment_properties_t *props,
+			 const hb_feature_t            *user_features,
+			 unsigned int                   num_user_features,
+			 const int                     *coords,
+			 unsigned int                   num_coords,
+			 const char * const            *shaper_list);
+
+  HB_INTERNAL void free () { ::free ((void *) user_features); }
+
+  HB_INTERNAL bool user_features_match (const hb_shape_plan_key_t *other);
+
+  HB_INTERNAL bool equal (const hb_shape_plan_key_t *other);
+};
+
+struct hb_shape_plan_t
+{
+  hb_object_header_t header;
+  hb_face_t *face_unsafe; /* We don't carry a reference to face. */
+  hb_shape_plan_key_t key;
+#ifndef HB_NO_OT_SHAPE
+  hb_ot_shape_plan_t ot;
+#endif
+};
+
+
+#endif /* HB_SHAPE_PLAN_HH */
diff --git a/thirdparty/harfbuzz/src/hb-shape.cc b/thirdparty/harfbuzz/src/hb-shape.cc
new file mode 100644
index 0000000000..017fb91b69
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shape.cc
@@ -0,0 +1,172 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-shaper.hh"
+#include "hb-shape-plan.hh"
+#include "hb-buffer.hh"
+#include "hb-font.hh"
+#include "hb-machinery.hh"
+
+
+/**
+ * SECTION:hb-shape
+ * @title: hb-shape
+ * @short_description: Conversion of text strings into positioned glyphs
+ * @include: hb.h
+ *
+ * Shaping is the central operation of HarfBuzz. Shaping operates on buffers,
+ * which are sequences of Unicode characters that use the same font and have
+ * the same text direction, script, and language. After shaping the buffer
+ * contains the output glyphs and their positions.
+ **/
+
+
+#if HB_USE_ATEXIT
+static void free_static_shaper_list ();
+#endif
+
+static const char *nil_shaper_list[] = {nullptr};
+
+static struct hb_shaper_list_lazy_loader_t : hb_lazy_loader_t<const char *,
+							      hb_shaper_list_lazy_loader_t>
+{
+  static const char ** create ()
+  {
+    const char **shaper_list = (const char **) calloc (1 + HB_SHAPERS_COUNT, sizeof (const char *));
+    if (unlikely (!shaper_list))
+      return nullptr;
+
+    const hb_shaper_entry_t *shapers = _hb_shapers_get ();
+    unsigned int i;
+    for (i = 0; i < HB_SHAPERS_COUNT; i++)
+      shaper_list[i] = shapers[i].name;
+    shaper_list[i] = nullptr;
+
+#if HB_USE_ATEXIT
+    atexit (free_static_shaper_list);
+#endif
+
+    return shaper_list;
+  }
+  static void destroy (const char **l)
+  { free (l); }
+  static const char ** get_null ()
+  { return nil_shaper_list; }
+} static_shaper_list;
+
+#if HB_USE_ATEXIT
+static
+void free_static_shaper_list ()
+{
+  static_shaper_list.free_instance ();
+}
+#endif
+
+
+/**
+ * hb_shape_list_shapers:
+ *
+ * Retrieves the list of shapers supported by HarfBuzz.
+ *
+ * Return value: (transfer none) (array zero-terminated=1): an array of
+ *    constant strings
+ *
+ * Since: 0.9.2
+ **/
+const char **
+hb_shape_list_shapers ()
+{
+  return static_shaper_list.get_unconst ();
+}
+
+
+/**
+ * hb_shape_full:
+ * @font: an #hb_font_t to use for shaping
+ * @buffer: an #hb_buffer_t to shape
+ * @features: (array length=num_features) (allow-none): an array of user
+ *    specified #hb_feature_t or %NULL
+ * @num_features: the length of @features array
+ * @shaper_list: (array zero-terminated=1) (allow-none): a %NULL-terminated
+ *    array of shapers to use or %NULL
+ *
+ * See hb_shape() for details. If @shaper_list is not %NULL, the specified
+ * shapers will be used in the given order, otherwise the default shapers list
+ * will be used.
+ *
+ * Return value: false if all shapers failed, true otherwise
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_shape_full (hb_font_t          *font,
+	       hb_buffer_t        *buffer,
+	       const hb_feature_t *features,
+	       unsigned int        num_features,
+	       const char * const *shaper_list)
+{
+  if (unlikely (hb_object_is_immutable (buffer))) return false;
+
+  hb_shape_plan_t *shape_plan = hb_shape_plan_create_cached2 (font->face, &buffer->props,
+							      features, num_features,
+							      font->coords, font->num_coords,
+							      shaper_list);
+  hb_bool_t res = hb_shape_plan_execute (shape_plan, font, buffer, features, num_features);
+  hb_shape_plan_destroy (shape_plan);
+
+  if (res)
+    buffer->content_type = HB_BUFFER_CONTENT_TYPE_GLYPHS;
+  return res;
+}
+
+/**
+ * hb_shape:
+ * @font: an #hb_font_t to use for shaping
+ * @buffer: an #hb_buffer_t to shape
+ * @features: (array length=num_features) (allow-none): an array of user
+ *    specified #hb_feature_t or %NULL
+ * @num_features: the length of @features array
+ *
+ * Shapes @buffer using @font turning its Unicode characters content to
+ * positioned glyphs. If @features is not %NULL, it will be used to control the
+ * features applied during shaping. If two @features have the same tag but
+ * overlapping ranges the value of the feature with the higher index takes
+ * precedence.
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_shape (hb_font_t           *font,
+	  hb_buffer_t         *buffer,
+	  const hb_feature_t  *features,
+	  unsigned int         num_features)
+{
+  hb_shape_full (font, buffer, features, num_features, nullptr);
+}
diff --git a/thirdparty/harfbuzz/src/hb-shape.h b/thirdparty/harfbuzz/src/hb-shape.h
new file mode 100644
index 0000000000..39507ff744
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shape.h
@@ -0,0 +1,62 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_SHAPE_H
+#define HB_SHAPE_H
+
+#include "hb-common.h"
+#include "hb-buffer.h"
+#include "hb-font.h"
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN void
+hb_shape (hb_font_t           *font,
+	  hb_buffer_t         *buffer,
+	  const hb_feature_t  *features,
+	  unsigned int         num_features);
+
+HB_EXTERN hb_bool_t
+hb_shape_full (hb_font_t          *font,
+	       hb_buffer_t        *buffer,
+	       const hb_feature_t *features,
+	       unsigned int        num_features,
+	       const char * const *shaper_list);
+
+HB_EXTERN const char **
+hb_shape_list_shapers (void);
+
+
+HB_END_DECLS
+
+#endif /* HB_SHAPE_H */
diff --git a/thirdparty/harfbuzz/src/hb-shaper-impl.hh b/thirdparty/harfbuzz/src/hb-shaper-impl.hh
new file mode 100644
index 0000000000..b674fceb6a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shaper-impl.hh
@@ -0,0 +1,38 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SHAPER_IMPL_HH
+#define HB_SHAPER_IMPL_HH
+
+#include "hb.hh"
+
+#include "hb-shaper.hh"
+#include "hb-face.hh"
+#include "hb-font.hh"
+#include "hb-shape-plan.hh"
+#include "hb-buffer.hh"
+
+#endif /* HB_SHAPER_IMPL_HH */
diff --git a/thirdparty/harfbuzz/src/hb-shaper-list.hh b/thirdparty/harfbuzz/src/hb-shaper-list.hh
new file mode 100644
index 0000000000..0d63933a76
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shaper-list.hh
@@ -0,0 +1,60 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SHAPER_LIST_HH
+#define HB_SHAPER_LIST_HH
+#endif /* HB_SHAPER_LIST_HH */ /* Dummy header guards */
+
+#ifndef HB_NO_SHAPER
+
+
+/* v--- Add new shapers in the right place here. */
+
+#ifdef HAVE_GRAPHITE2
+/* Only picks up fonts that have a "Silf" table. */
+HB_SHAPER_IMPLEMENT (graphite2)
+#endif
+
+#ifndef HB_NO_OT_SHAPE
+HB_SHAPER_IMPLEMENT (ot) /* <--- This is our main OpenType shaper. */
+#endif
+
+#ifdef HAVE_UNISCRIBE
+HB_SHAPER_IMPLEMENT (uniscribe)
+#endif
+#ifdef HAVE_DIRECTWRITE
+HB_SHAPER_IMPLEMENT (directwrite)
+#endif
+#ifdef HAVE_CORETEXT
+HB_SHAPER_IMPLEMENT (coretext)
+#endif
+
+#ifndef HB_NO_FALLBACK_SHAPE
+HB_SHAPER_IMPLEMENT (fallback) /* <--- This should be last. */
+#endif
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-shaper.cc b/thirdparty/harfbuzz/src/hb-shaper.cc
new file mode 100644
index 0000000000..0ea68ad1f5
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shaper.cc
@@ -0,0 +1,108 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+#include "hb-shaper.hh"
+#include "hb-machinery.hh"
+
+
+static const hb_shaper_entry_t all_shapers[] = {
+#define HB_SHAPER_IMPLEMENT(name) {#name, _hb_##name##_shape},
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+};
+#ifndef HB_NO_SHAPER
+static_assert (0 != ARRAY_LENGTH_CONST (all_shapers), "No shaper enabled.");
+#endif
+
+#if HB_USE_ATEXIT
+static void free_static_shapers ();
+#endif
+
+static struct hb_shapers_lazy_loader_t : hb_lazy_loader_t<const hb_shaper_entry_t,
+							  hb_shapers_lazy_loader_t>
+{
+  static hb_shaper_entry_t *create ()
+  {
+    char *env = getenv ("HB_SHAPER_LIST");
+    if (!env || !*env)
+      return nullptr;
+
+    hb_shaper_entry_t *shapers = (hb_shaper_entry_t *) calloc (1, sizeof (all_shapers));
+    if (unlikely (!shapers))
+      return nullptr;
+
+    memcpy (shapers, all_shapers, sizeof (all_shapers));
+
+     /* Reorder shaper list to prefer requested shapers. */
+    unsigned int i = 0;
+    char *end, *p = env;
+    for (;;)
+    {
+      end = strchr (p, ',');
+      if (!end)
+	end = p + strlen (p);
+
+      for (unsigned int j = i; j < ARRAY_LENGTH (all_shapers); j++)
+	if (end - p == (int) strlen (shapers[j].name) &&
+	    0 == strncmp (shapers[j].name, p, end - p))
+	{
+	  /* Reorder this shaper to position i */
+	 struct hb_shaper_entry_t t = shapers[j];
+	 memmove (&shapers[i + 1], &shapers[i], sizeof (shapers[i]) * (j - i));
+	 shapers[i] = t;
+	 i++;
+	}
+
+      if (!*end)
+	break;
+      else
+	p = end + 1;
+    }
+
+#if HB_USE_ATEXIT
+    atexit (free_static_shapers);
+#endif
+
+    return shapers;
+  }
+  static void destroy (const hb_shaper_entry_t *p) { free ((void *) p); }
+  static const hb_shaper_entry_t *get_null ()      { return all_shapers; }
+} static_shapers;
+
+#if HB_USE_ATEXIT
+static
+void free_static_shapers ()
+{
+  static_shapers.free_instance ();
+}
+#endif
+
+const hb_shaper_entry_t *
+_hb_shapers_get ()
+{
+  return static_shapers.get_unconst ();
+}
diff --git a/thirdparty/harfbuzz/src/hb-shaper.hh b/thirdparty/harfbuzz/src/hb-shaper.hh
new file mode 100644
index 0000000000..b4138a324f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-shaper.hh
@@ -0,0 +1,134 @@
+/*
+ * Copyright © 2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_SHAPER_HH
+#define HB_SHAPER_HH
+
+#include "hb.hh"
+#include "hb-machinery.hh"
+
+typedef hb_bool_t hb_shape_func_t (hb_shape_plan_t    *shape_plan,
+				   hb_font_t          *font,
+				   hb_buffer_t        *buffer,
+				   const hb_feature_t *features,
+				   unsigned int        num_features);
+
+#define HB_SHAPER_IMPLEMENT(name) \
+	extern "C" HB_INTERNAL hb_shape_func_t _hb_##name##_shape;
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+
+struct hb_shaper_entry_t {
+  char name[16];
+  hb_shape_func_t *func;
+};
+
+HB_INTERNAL const hb_shaper_entry_t *
+_hb_shapers_get ();
+
+
+template <typename Data, unsigned int WheresData, typename T>
+struct hb_shaper_lazy_loader_t;
+
+#define HB_SHAPER_ORDER(Shaper) \
+  HB_PASTE (HB_SHAPER_ORDER_, Shaper)
+enum hb_shaper_order_t
+{
+  _HB_SHAPER_ORDER_ORDER_ZERO,
+#define HB_SHAPER_IMPLEMENT(Shaper) \
+      HB_SHAPER_ORDER (Shaper),
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+  _HB_SHAPERS_COUNT_PLUS_ONE,
+  HB_SHAPERS_COUNT = _HB_SHAPERS_COUNT_PLUS_ONE - 1,
+};
+
+template <enum hb_shaper_order_t order, typename Object> struct hb_shaper_object_data_type_t;
+
+#define HB_SHAPER_DATA_SUCCEEDED ((void *) +1)
+#define HB_SHAPER_DATA_TYPE(shaper, object)		hb_##shaper##_##object##_data_t
+#define HB_SHAPER_DATA_CREATE_FUNC(shaper, object)	_hb_##shaper##_shaper_##object##_data_create
+#define HB_SHAPER_DATA_DESTROY_FUNC(shaper, object)	_hb_##shaper##_shaper_##object##_data_destroy
+
+#define HB_SHAPER_DATA_INSTANTIATE_SHAPERS(shaper, object) \
+	\
+	struct HB_SHAPER_DATA_TYPE (shaper, object); /* Type forward declaration. */ \
+	extern "C" HB_INTERNAL HB_SHAPER_DATA_TYPE (shaper, object) * \
+	HB_SHAPER_DATA_CREATE_FUNC (shaper, object) (hb_##object##_t *object); \
+	extern "C" HB_INTERNAL void \
+	HB_SHAPER_DATA_DESTROY_FUNC (shaper, object) (HB_SHAPER_DATA_TYPE (shaper, object) *shaper##_##object); \
+	\
+	template <> \
+	struct hb_shaper_object_data_type_t<HB_SHAPER_ORDER (shaper), hb_##object##_t> \
+	{ \
+	  typedef HB_SHAPER_DATA_TYPE(shaper, object) value; \
+	}; \
+	\
+	template <unsigned int WheresData> \
+	struct hb_shaper_lazy_loader_t<hb_##object##_t, WheresData, HB_SHAPER_DATA_TYPE(shaper, object)> \
+		: hb_lazy_loader_t<HB_SHAPER_DATA_TYPE(shaper, object), \
+				   hb_shaper_lazy_loader_t<hb_##object##_t, \
+							   WheresData, \
+							   HB_SHAPER_DATA_TYPE(shaper, object)>, \
+				   hb_##object##_t, WheresData> \
+	{ \
+	  typedef HB_SHAPER_DATA_TYPE(shaper, object) Type; \
+	  static Type* create (hb_##object##_t *data) \
+	  { return HB_SHAPER_DATA_CREATE_FUNC (shaper, object) (data); } \
+	  static Type *get_null () { return nullptr; } \
+	  static void destroy (Type *p) { HB_SHAPER_DATA_DESTROY_FUNC (shaper, object) (p); } \
+	}; \
+	\
+	static_assert (true, "") /* Require semicolon after. */
+
+
+template <typename Object>
+struct hb_shaper_object_dataset_t
+{
+  void init0 (Object *parent_data)
+  {
+    this->parent_data = parent_data;
+#define HB_SHAPER_IMPLEMENT(shaper) shaper.init0 ();
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+  }
+  void fini ()
+  {
+#define HB_SHAPER_IMPLEMENT(shaper) shaper.fini ();
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+  }
+
+  Object *parent_data; /* MUST be JUST before the lazy loaders. */
+#define HB_SHAPER_IMPLEMENT(shaper) \
+	hb_shaper_lazy_loader_t<Object, HB_SHAPER_ORDER(shaper), \
+				typename hb_shaper_object_data_type_t<HB_SHAPER_ORDER(shaper), Object>::value \
+			       > shaper;
+#include "hb-shaper-list.hh"
+#undef HB_SHAPER_IMPLEMENT
+};
+
+#endif /* HB_SHAPER_HH */
diff --git a/thirdparty/harfbuzz/src/hb-static.cc b/thirdparty/harfbuzz/src/hb-static.cc
new file mode 100644
index 0000000000..f5b7fa50a0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-static.cc
@@ -0,0 +1,112 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-open-type.hh"
+#include "hb-face.hh"
+
+#include "hb-aat-layout-common.hh"
+#include "hb-aat-layout-feat-table.hh"
+#include "hb-ot-layout-common.hh"
+#include "hb-ot-cmap-table.hh"
+#include "hb-ot-head-table.hh"
+#include "hb-ot-maxp-table.hh"
+
+#ifndef HB_NO_VISIBILITY
+#include "hb-ot-name-language-static.hh"
+
+uint64_t const _hb_NullPool[(HB_NULL_POOL_SIZE + sizeof (uint64_t) - 1) / sizeof (uint64_t)] = {};
+/*thread_local*/ uint64_t _hb_CrapPool[(HB_NULL_POOL_SIZE + sizeof (uint64_t) - 1) / sizeof (uint64_t)] = {};
+
+DEFINE_NULL_NAMESPACE_BYTES (OT, Index) =  {0xFF,0xFF};
+DEFINE_NULL_NAMESPACE_BYTES (OT, LangSys) = {0x00,0x00, 0xFF,0xFF, 0x00,0x00};
+DEFINE_NULL_NAMESPACE_BYTES (OT, RangeRecord) = {0x00,0x01, 0x00,0x00, 0x00, 0x00};
+DEFINE_NULL_NAMESPACE_BYTES (OT, CmapSubtableLongGroup) = {0x00,0x00,0x00,0x01, 0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00};
+DEFINE_NULL_NAMESPACE_BYTES (AAT, SettingName) = {0xFF,0xFF, 0xFF,0xFF};
+/* Hand-coded because Lookup is a template.  Sad. */
+const unsigned char _hb_Null_AAT_Lookup[2] = {0xFF, 0xFF};
+
+
+
+/* hb_face_t */
+
+unsigned int
+hb_face_t::load_num_glyphs () const
+{
+  hb_sanitize_context_t c = hb_sanitize_context_t ();
+  c.set_num_glyphs (0); /* So we don't recurse ad infinitum. */
+  hb_blob_t *maxp_blob = c.reference_table<OT::maxp> (this);
+  const OT::maxp *maxp_table = maxp_blob->as<OT::maxp> ();
+
+  unsigned int ret = maxp_table->get_num_glyphs ();
+  num_glyphs.set_relaxed (ret);
+  hb_blob_destroy (maxp_blob);
+  return ret;
+}
+
+unsigned int
+hb_face_t::load_upem () const
+{
+  unsigned int ret = table.head->get_upem ();
+  upem.set_relaxed (ret);
+  return ret;
+}
+
+
+/* hb_user_data_array_t */
+
+bool
+hb_user_data_array_t::set (hb_user_data_key_t *key,
+			   void *              data,
+			   hb_destroy_func_t   destroy,
+			   hb_bool_t           replace)
+{
+  if (!key)
+    return false;
+
+  if (replace) {
+    if (!data && !destroy) {
+      items.remove (key, lock);
+      return true;
+    }
+  }
+  hb_user_data_item_t item = {key, data, destroy};
+  bool ret = !!items.replace_or_insert (item, lock, (bool) replace);
+
+  return ret;
+}
+
+void *
+hb_user_data_array_t::get (hb_user_data_key_t *key)
+{
+  hb_user_data_item_t item = {nullptr, nullptr, nullptr};
+
+  return items.find (key, &item, lock) ? item.data : nullptr;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-string-array.hh b/thirdparty/harfbuzz/src/hb-string-array.hh
new file mode 100644
index 0000000000..e7ac119232
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-string-array.hh
@@ -0,0 +1,85 @@
+/*
+ * Copyright © 2017  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_STRING_ARRAY_HH
+#if 0 /* Make checks happy. */
+#define HB_STRING_ARRAY_HH
+#endif
+
+#include "hb.hh"
+
+/* Based on Bruno Haible's code in Appendix B of Ulrich Drepper's dsohowto.pdf:
+ * https://software.intel.com/sites/default/files/m/a/1/e/dsohowto.pdf */
+
+#define HB_STRING_ARRAY_TYPE_NAME	HB_PASTE(HB_STRING_ARRAY_NAME, _msgstr_t)
+#define HB_STRING_ARRAY_POOL_NAME	HB_PASTE(HB_STRING_ARRAY_NAME, _msgstr)
+#define HB_STRING_ARRAY_OFFS_NAME	HB_PASTE(HB_STRING_ARRAY_NAME, _msgidx)
+#define HB_STRING_ARRAY_LENG_NAME	HB_PASTE(HB_STRING_ARRAY_NAME, _length)
+
+static const union HB_STRING_ARRAY_TYPE_NAME {
+  struct {
+/* I like to avoid storing the nul-termination byte since we don't need it,
+ * but C++ does not allow that.
+ * https://stackoverflow.com/q/28433862
+ */
+#define _S(s) char HB_PASTE (str, __LINE__)[sizeof (s)];
+#include HB_STRING_ARRAY_LIST
+#undef _S
+  } st;
+  char str[HB_VAR_ARRAY];
+}
+HB_STRING_ARRAY_POOL_NAME =
+{
+  {
+#define _S(s) s,
+#include HB_STRING_ARRAY_LIST
+#undef _S
+  }
+};
+static const unsigned int HB_STRING_ARRAY_OFFS_NAME[] =
+{
+#define _S(s) offsetof (union HB_STRING_ARRAY_TYPE_NAME, st.HB_PASTE(str, __LINE__)),
+#include HB_STRING_ARRAY_LIST
+#undef _S
+  sizeof (HB_STRING_ARRAY_TYPE_NAME)
+};
+
+static const unsigned int HB_STRING_ARRAY_LENG_NAME = ARRAY_LENGTH_CONST (HB_STRING_ARRAY_OFFS_NAME) - 1;
+
+static inline hb_bytes_t
+HB_STRING_ARRAY_NAME (unsigned int i)
+{
+  assert (i < ARRAY_LENGTH (HB_STRING_ARRAY_OFFS_NAME) - 1);
+  return hb_bytes_t (HB_STRING_ARRAY_POOL_NAME.str + HB_STRING_ARRAY_OFFS_NAME[i],
+		     HB_STRING_ARRAY_OFFS_NAME[i + 1] - HB_STRING_ARRAY_OFFS_NAME[i] - 1);
+}
+
+#undef HB_STRING_ARRAY_TYPE_NAME
+#undef HB_STRING_ARRAY_POOL_NAME
+#undef HB_STRING_ARRAY_OFFS_NAME
+#undef HB_STRING_ARRAY_LENG_NAME
+
+#endif /* HB_STRING_ARRAY_HH */
diff --git a/thirdparty/harfbuzz/src/hb-style.cc b/thirdparty/harfbuzz/src/hb-style.cc
new file mode 100644
index 0000000000..86b9f7da5f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-style.cc
@@ -0,0 +1,135 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_STYLE
+#ifdef HB_EXPERIMENTAL_API
+
+#include "hb-ot-var-avar-table.hh"
+#include "hb-ot-var-fvar-table.hh"
+#include "hb-ot-stat-table.hh"
+#include "hb-ot-os2-table.hh"
+#include "hb-ot-head-table.hh"
+#include "hb-ot-post-table.hh"
+#include "hb-ot-face.hh"
+
+/**
+ * hb_style_tag_t:
+ * @HB_STYLE_TAG_ITALIC: Used to vary between non-italic and italic.
+ * A value of 0 can be interpreted as "Roman" (non-italic); a value of 1 can
+ * be interpreted as (fully) italic.
+ * @HB_STYLE_TAG_OPTICAL_SIZE: Used to vary design to suit different text sizes.
+ * Non-zero. Values can be interpreted as text size, in points.
+ * @HB_STYLE_TAG_SLANT: Used to vary between upright and slanted text. Values
+ * must be greater than -90 and less than +90. Values can be interpreted as
+ * the angle, in counter-clockwise degrees, of oblique slant from whatever the
+ * designer considers to be upright for that font design.
+ * @HB_STYLE_TAG_WIDTH: Used to vary width of text from narrower to wider.
+ * Non-zero. Values can be interpreted as a percentage of whatever the font
+ * designer considers “normal width” for that font design.
+ * @HB_STYLE_TAG_WEIGHT: Used to vary stroke thicknesses or other design details
+ * to give variation from lighter to blacker. Values can be interpreted in direct
+ * comparison to values for usWeightClass in the OS/2 table,
+ * or the CSS font-weight property.
+ *
+ * Defined by https://docs.microsoft.com/en-us/typography/opentype/spec/dvaraxisreg
+ *
+ * Since: EXPERIMENTAL
+ **/
+typedef enum {
+  HB_STYLE_TAG_ITALIC		= HB_TAG ('i','t','a','l'),
+  HB_STYLE_TAG_OPTICAL_SIZE	= HB_TAG ('o','p','s','z'),
+  HB_STYLE_TAG_SLANT		= HB_TAG ('s','l','n','t'),
+  HB_STYLE_TAG_WIDTH		= HB_TAG ('w','d','t','h'),
+  HB_STYLE_TAG_WEIGHT		= HB_TAG ('w','g','h','t'),
+
+  _HB_STYLE_TAG_MAX_VALUE	= HB_TAG_MAX_SIGNED /*< skip >*/
+} hb_style_tag_t;
+
+/**
+ * hb_style_get_value:
+ * @font: a #hb_font_t object.
+ * @style_tag: a style tag.
+ *
+ * Searches variation axes of a hb_font_t object for a specific axis first,
+ * if not set, then tries to get default style values from different
+ * tables of the font.
+ *
+ * Returns: Corresponding axis or default value to a style tag.
+ *
+ * Since: EXPERIMENTAL
+ **/
+float
+hb_style_get_value (hb_font_t *font, hb_tag_t tag)
+{
+  hb_style_tag_t style_tag = (hb_style_tag_t) tag;
+  hb_face_t *face = font->face;
+
+#ifndef HB_NO_VAR
+  hb_ot_var_axis_info_t axis;
+  if (hb_ot_var_find_axis_info (face, style_tag, &axis))
+  {
+    if (axis.axis_index < font->num_coords) return font->design_coords[axis.axis_index];
+    /* If a face is variable, fvar's default_value is better than STAT records */
+    return axis.default_value;
+  }
+#endif
+
+  if (style_tag == HB_STYLE_TAG_OPTICAL_SIZE && font->ptem)
+    return font->ptem;
+
+  /* STAT */
+  float value;
+  if (face->table.STAT->get_value (style_tag, &value))
+    return value;
+
+  switch ((unsigned) style_tag)
+  {
+  case HB_STYLE_TAG_ITALIC:
+    return face->table.OS2->is_italic () || face->table.head->is_italic () ? 1 : 0;
+  case HB_STYLE_TAG_OPTICAL_SIZE:
+  {
+    unsigned int lower, upper;
+    return face->table.OS2->v5 ().get_optical_size (&lower, &upper)
+	   ? (float) (lower + upper) / 2.f
+	   : 12.f;
+  }
+  case HB_STYLE_TAG_SLANT:
+    return face->table.post->table->italicAngle.to_float ();
+  case HB_STYLE_TAG_WIDTH:
+    return face->table.OS2->has_data ()
+	   ? face->table.OS2->get_width ()
+	   : (face->table.head->is_condensed () ? 75 : 100);
+  case HB_STYLE_TAG_WEIGHT:
+    return face->table.OS2->has_data ()
+	   ? face->table.OS2->usWeightClass
+	   : (face->table.head->is_bold () ? 700 : 400);
+  default:
+    return 0;
+  }
+}
+
+#endif
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-style.h b/thirdparty/harfbuzz/src/hb-style.h
new file mode 100644
index 0000000000..1209c79e94
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-style.h
@@ -0,0 +1,43 @@
+/*
+ * Copyright © 2019  Ebrahim Byagowi
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_STYLE_H
+#define HB_STYLE_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+#ifdef HB_EXPERIMENTAL_API
+HB_EXTERN float
+hb_style_get_value (hb_font_t *font, hb_tag_t style_tag);
+#endif
+
+HB_END_DECLS
+
+#endif /* HB_STYLE_H */
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff-common.cc b/thirdparty/harfbuzz/src/hb-subset-cff-common.cc
new file mode 100644
index 0000000000..04e1db24ac
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff-common.cc
@@ -0,0 +1,227 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_SUBSET_CFF
+
+#include "hb-ot-cff-common.hh"
+#include "hb-ot-cff2-table.hh"
+#include "hb-subset-cff-common.hh"
+
+/* Disable FDSelect format 0 for compatibility with fonttools which doesn't seem choose it.
+ * Rarely any/much smaller than format 3 anyway. */
+#define CFF_SERIALIZE_FDSELECT_0  0
+
+using namespace CFF;
+
+/**
+ * hb_plan_subset_cff_fdselect
+ * Determine an optimal FDSelect format according to a provided plan.
+ *
+ * Return value: FDSelect format, size, and ranges for the most compact subset FDSelect
+ * along with a font index remapping table
+ **/
+
+bool
+hb_plan_subset_cff_fdselect (const hb_subset_plan_t *plan,
+			     unsigned int fdCount,
+			     const FDSelect &src, /* IN */
+			     unsigned int &subset_fd_count /* OUT */,
+			     unsigned int &subset_fdselect_size /* OUT */,
+			     unsigned int &subset_fdselect_format /* OUT */,
+			     hb_vector_t<code_pair_t> &fdselect_ranges /* OUT */,
+			     hb_inc_bimap_t &fdmap /* OUT */)
+{
+  subset_fd_count = 0;
+  subset_fdselect_size = 0;
+  subset_fdselect_format = 0;
+  unsigned int num_ranges = 0;
+
+  unsigned int subset_num_glyphs = plan->num_output_glyphs ();
+  if (subset_num_glyphs == 0)
+    return true;
+
+  {
+    /* use hb_set to determine the subset of font dicts */
+    hb_set_t *set = hb_set_create ();
+    if (unlikely (set == &Null (hb_set_t))) return false;
+    hb_codepoint_t prev_fd = CFF_UNDEF_CODE;
+    for (hb_codepoint_t i = 0; i < subset_num_glyphs; i++)
+    {
+      hb_codepoint_t glyph;
+      hb_codepoint_t fd;
+      if (!plan->old_gid_for_new_gid (i, &glyph))
+      {
+	/* fonttools retains FDSelect & font dicts for missing glyphs. do the same */
+	glyph = i;
+      }
+      fd = src.get_fd (glyph);
+      set->add (fd);
+
+      if (fd != prev_fd)
+      {
+	num_ranges++;
+	prev_fd = fd;
+	code_pair_t pair = { fd, i };
+	fdselect_ranges.push (pair);
+      }
+    }
+
+    subset_fd_count = set->get_population ();
+    if (subset_fd_count == fdCount)
+    {
+      /* all font dicts belong to the subset. no need to subset FDSelect & FDArray */
+      fdmap.identity (fdCount);
+      hb_set_destroy (set);
+    }
+    else
+    {
+      /* create a fdmap */
+      fdmap.reset ();
+
+      hb_codepoint_t fd = CFF_UNDEF_CODE;
+      while (set->next (&fd))
+	fdmap.add (fd);
+      hb_set_destroy (set);
+      if (unlikely (fdmap.get_population () != subset_fd_count))
+	return false;
+    }
+
+    /* update each font dict index stored as "code" in fdselect_ranges */
+    for (unsigned int i = 0; i < fdselect_ranges.length; i++)
+      fdselect_ranges[i].code = fdmap[fdselect_ranges[i].code];
+  }
+
+  /* determine which FDSelect format is most compact */
+  if (subset_fd_count > 0xFF)
+  {
+    if (unlikely (src.format != 4))
+      return false;
+    subset_fdselect_format = 4;
+    subset_fdselect_size = FDSelect::min_size + FDSelect4::min_size + FDSelect4_Range::static_size * num_ranges + HBUINT32::static_size;
+  }
+  else
+  {
+#if CFF_SERIALIZE_FDSELECT_0
+    unsigned int format0_size = FDSelect::min_size + FDSelect0::min_size + HBUINT8::static_size * subset_num_glyphs;
+#endif
+    unsigned int format3_size = FDSelect::min_size + FDSelect3::min_size + FDSelect3_Range::static_size * num_ranges + HBUINT16::static_size;
+
+#if CFF_SERIALIZE_FDSELECT_0
+    if (format0_size <= format3_size)
+    {
+      // subset_fdselect_format = 0;
+      subset_fdselect_size = format0_size;
+    }
+    else
+#endif
+    {
+      subset_fdselect_format = 3;
+      subset_fdselect_size = format3_size;
+    }
+  }
+
+  return true;
+}
+
+template <typename FDSELECT3_4>
+static inline bool
+serialize_fdselect_3_4 (hb_serialize_context_t *c,
+			const unsigned int num_glyphs,
+			const FDSelect &src,
+			unsigned int size,
+			const hb_vector_t<code_pair_t> &fdselect_ranges)
+{
+  TRACE_SERIALIZE (this);
+  FDSELECT3_4 *p = c->allocate_size<FDSELECT3_4> (size);
+  if (unlikely (!p)) return_trace (false);
+  p->nRanges () = fdselect_ranges.length;
+  for (unsigned int i = 0; i < fdselect_ranges.length; i++)
+  {
+    p->ranges[i].first = fdselect_ranges[i].glyph;
+    p->ranges[i].fd = fdselect_ranges[i].code;
+  }
+  p->sentinel () = num_glyphs;
+  return_trace (true);
+}
+
+/**
+ * hb_serialize_cff_fdselect
+ * Serialize a subset FDSelect format planned above.
+ **/
+bool
+hb_serialize_cff_fdselect (hb_serialize_context_t *c,
+			   const unsigned int num_glyphs,
+			   const FDSelect &src,
+			   unsigned int fd_count,
+			   unsigned int fdselect_format,
+			   unsigned int size,
+			   const hb_vector_t<code_pair_t> &fdselect_ranges)
+{
+  TRACE_SERIALIZE (this);
+  FDSelect *p = c->allocate_min<FDSelect> ();
+  if (unlikely (!p)) return_trace (false);
+  p->format = fdselect_format;
+  size -= FDSelect::min_size;
+
+  switch (fdselect_format)
+  {
+#if CFF_SERIALIZE_FDSELECT_0
+  case 0:
+  {
+    FDSelect0 *p = c->allocate_size<FDSelect0> (size);
+    if (unlikely (!p)) return_trace (false);
+    unsigned int range_index = 0;
+    unsigned int fd = fdselect_ranges[range_index++].code;
+    for (unsigned int i = 0; i < num_glyphs; i++)
+    {
+      if ((range_index < fdselect_ranges.len) &&
+	  (i >= fdselect_ranges[range_index].glyph))
+      {
+	fd = fdselect_ranges[range_index++].code;
+      }
+      p->fds[i] = fd;
+    }
+    return_trace (true);
+  }
+#endif /* CFF_SERIALIZE_FDSELECT_0 */
+
+  case 3:
+    return serialize_fdselect_3_4<FDSelect3> (c, num_glyphs, src,
+					      size, fdselect_ranges);
+
+  case 4:
+    return serialize_fdselect_3_4<FDSelect4> (c, num_glyphs, src,
+					      size, fdselect_ranges);
+
+  default:
+    return_trace (false);
+  }
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff-common.hh b/thirdparty/harfbuzz/src/hb-subset-cff-common.hh
new file mode 100644
index 0000000000..422b20b8d0
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff-common.hh
@@ -0,0 +1,989 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_SUBSET_CFF_COMMON_HH
+#define HB_SUBSET_CFF_COMMON_HH
+
+#include "hb.hh"
+
+#include "hb-subset-plan.hh"
+#include "hb-cff-interp-cs-common.hh"
+
+namespace CFF {
+
+/* Used for writing a temporary charstring */
+struct str_encoder_t
+{
+  str_encoder_t (str_buff_t &buff_)
+    : buff (buff_), error (false) {}
+
+  void reset () { buff.resize (0); }
+
+  void encode_byte (unsigned char b)
+  {
+    if (unlikely (buff.push (b) == &Crap (unsigned char)))
+      set_error ();
+  }
+
+  void encode_int (int v)
+  {
+    if ((-1131 <= v) && (v <= 1131))
+    {
+      if ((-107 <= v) && (v <= 107))
+	encode_byte (v + 139);
+      else if (v > 0)
+      {
+	v -= 108;
+	encode_byte ((v >> 8) + OpCode_TwoBytePosInt0);
+	encode_byte (v & 0xFF);
+      }
+      else
+      {
+	v = -v - 108;
+	encode_byte ((v >> 8) + OpCode_TwoByteNegInt0);
+	encode_byte (v & 0xFF);
+      }
+    }
+    else
+    {
+      if (unlikely (v < -32768))
+	v = -32768;
+      else if (unlikely (v > 32767))
+	v = 32767;
+      encode_byte (OpCode_shortint);
+      encode_byte ((v >> 8) & 0xFF);
+      encode_byte (v & 0xFF);
+    }
+  }
+
+  void encode_num (const number_t& n)
+  {
+    if (n.in_int_range ())
+    {
+      encode_int (n.to_int ());
+    }
+    else
+    {
+      int32_t v = n.to_fixed ();
+      encode_byte (OpCode_fixedcs);
+      encode_byte ((v >> 24) & 0xFF);
+      encode_byte ((v >> 16) & 0xFF);
+      encode_byte ((v >> 8) & 0xFF);
+      encode_byte (v & 0xFF);
+    }
+  }
+
+  void encode_op (op_code_t op)
+  {
+    if (Is_OpCode_ESC (op))
+    {
+      encode_byte (OpCode_escape);
+      encode_byte (Unmake_OpCode_ESC (op));
+    }
+    else
+      encode_byte (op);
+  }
+
+  void copy_str (const byte_str_t &str)
+  {
+    unsigned int  offset = buff.length;
+    if (unlikely (!buff.resize (offset + str.length)))
+    {
+      set_error ();
+      return;
+    }
+    if (unlikely (buff.length < offset + str.length))
+    {
+      set_error ();
+      return;
+    }
+    memcpy (&buff[offset], &str[0], str.length);
+  }
+
+  bool is_error () const { return error; }
+
+  protected:
+  void set_error () { error = true; }
+
+  str_buff_t &buff;
+  bool    error;
+};
+
+struct cff_sub_table_info_t {
+  cff_sub_table_info_t ()
+    : fd_array_link (0),
+      char_strings_link (0)
+  {
+    fd_select.init ();
+  }
+
+  table_info_t     fd_select;
+  objidx_t     	   fd_array_link;
+  objidx_t     	   char_strings_link;
+};
+
+template <typename OPSTR=op_str_t>
+struct cff_top_dict_op_serializer_t : op_serializer_t
+{
+  bool serialize (hb_serialize_context_t *c,
+		  const OPSTR &opstr,
+		  const cff_sub_table_info_t &info) const
+  {
+    TRACE_SERIALIZE (this);
+
+    switch (opstr.op)
+    {
+      case OpCode_CharStrings:
+	return_trace (FontDict::serialize_link4_op(c, opstr.op, info.char_strings_link, whence_t::Absolute));
+
+      case OpCode_FDArray:
+	return_trace (FontDict::serialize_link4_op(c, opstr.op, info.fd_array_link, whence_t::Absolute));
+
+      case OpCode_FDSelect:
+	return_trace (FontDict::serialize_link4_op(c, opstr.op, info.fd_select.link, whence_t::Absolute));
+
+      default:
+	return_trace (copy_opstr (c, opstr));
+    }
+    return_trace (true);
+  }
+};
+
+struct cff_font_dict_op_serializer_t : op_serializer_t
+{
+  bool serialize (hb_serialize_context_t *c,
+		  const op_str_t &opstr,
+		  const table_info_t &privateDictInfo) const
+  {
+    TRACE_SERIALIZE (this);
+
+    if (opstr.op == OpCode_Private)
+    {
+      /* serialize the private dict size & offset as 2-byte & 4-byte integers */
+      return_trace (UnsizedByteStr::serialize_int2 (c, privateDictInfo.size) &&
+		    Dict::serialize_link4_op (c, opstr.op, privateDictInfo.link, whence_t::Absolute));
+    }
+    else
+    {
+      HBUINT8 *d = c->allocate_size<HBUINT8> (opstr.str.length);
+      if (unlikely (!d)) return_trace (false);
+      memcpy (d, &opstr.str[0], opstr.str.length);
+    }
+    return_trace (true);
+  }
+};
+
+struct cff_private_dict_op_serializer_t : op_serializer_t
+{
+  cff_private_dict_op_serializer_t (bool desubroutinize_, bool drop_hints_)
+    : desubroutinize (desubroutinize_), drop_hints (drop_hints_) {}
+
+  bool serialize (hb_serialize_context_t *c,
+		  const op_str_t &opstr,
+		  objidx_t subrs_link) const
+  {
+    TRACE_SERIALIZE (this);
+
+    if (drop_hints && dict_opset_t::is_hint_op (opstr.op))
+      return true;
+    if (opstr.op == OpCode_Subrs)
+    {
+      if (desubroutinize || !subrs_link)
+	return_trace (true);
+      else
+	return_trace (FontDict::serialize_link2_op (c, opstr.op, subrs_link));
+    }
+    else
+      return_trace (copy_opstr (c, opstr));
+  }
+
+  protected:
+  const bool  desubroutinize;
+  const bool  drop_hints;
+};
+
+struct flatten_param_t
+{
+  str_buff_t     &flatStr;
+  bool	drop_hints;
+};
+
+template <typename ACC, typename ENV, typename OPSET, op_code_t endchar_op=OpCode_Invalid>
+struct subr_flattener_t
+{
+  subr_flattener_t (const ACC &acc_,
+		    const hb_subset_plan_t *plan_)
+		   : acc (acc_), plan (plan_) {}
+
+  bool flatten (str_buff_vec_t &flat_charstrings)
+  {
+    if (!flat_charstrings.resize (plan->num_output_glyphs ()))
+      return false;
+    for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+      flat_charstrings[i].init ();
+    for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+    {
+      hb_codepoint_t  glyph;
+      if (!plan->old_gid_for_new_gid (i, &glyph))
+      {
+	/* add an endchar only charstring for a missing glyph if CFF1 */
+	if (endchar_op != OpCode_Invalid) flat_charstrings[i].push (endchar_op);
+	continue;
+      }
+      const byte_str_t str = (*acc.charStrings)[glyph];
+      unsigned int fd = acc.fdSelect->get_fd (glyph);
+      if (unlikely (fd >= acc.fdCount))
+	return false;
+      cs_interpreter_t<ENV, OPSET, flatten_param_t> interp;
+      interp.env.init (str, acc, fd);
+      flatten_param_t  param = { flat_charstrings[i], plan->drop_hints };
+      if (unlikely (!interp.interpret (param)))
+	return false;
+    }
+    return true;
+  }
+
+  const ACC &acc;
+  const hb_subset_plan_t *plan;
+};
+
+struct subr_closures_t
+{
+  subr_closures_t () : valid (false), global_closure (nullptr)
+  { local_closures.init (); }
+
+  void init (unsigned int fd_count)
+  {
+    valid = true;
+    global_closure = hb_set_create ();
+    if (global_closure == hb_set_get_empty ())
+      valid = false;
+    if (!local_closures.resize (fd_count))
+      valid = false;
+
+    for (unsigned int i = 0; i < local_closures.length; i++)
+    {
+      local_closures[i] = hb_set_create ();
+      if (local_closures[i] == hb_set_get_empty ())
+	valid = false;
+    }
+  }
+
+  void fini ()
+  {
+    hb_set_destroy (global_closure);
+    for (unsigned int i = 0; i < local_closures.length; i++)
+      hb_set_destroy (local_closures[i]);
+    local_closures.fini ();
+  }
+
+  void reset ()
+  {
+    hb_set_clear (global_closure);
+    for (unsigned int i = 0; i < local_closures.length; i++)
+      hb_set_clear (local_closures[i]);
+  }
+
+  bool is_valid () const { return valid; }
+  bool  valid;
+  hb_set_t  *global_closure;
+  hb_vector_t<hb_set_t *> local_closures;
+};
+
+struct parsed_cs_op_t : op_str_t
+{
+  void init (unsigned int subr_num_ = 0)
+  {
+    op_str_t::init ();
+    subr_num = subr_num_;
+    drop_flag = false;
+    keep_flag = false;
+    skip_flag = false;
+  }
+
+  void fini () { op_str_t::fini (); }
+
+  bool for_drop () const { return drop_flag; }
+  void set_drop ()       { if (!for_keep ()) drop_flag = true; }
+
+  bool for_keep () const { return keep_flag; }
+  void set_keep ()       { keep_flag = true; }
+
+  bool for_skip () const { return skip_flag; }
+  void set_skip ()       { skip_flag = true; }
+
+  unsigned int  subr_num;
+
+  protected:
+  bool	  drop_flag : 1;
+  bool	  keep_flag : 1;
+  bool	  skip_flag : 1;
+};
+
+struct parsed_cs_str_t : parsed_values_t<parsed_cs_op_t>
+{
+  void init ()
+  {
+    SUPER::init ();
+    parsed = false;
+    hint_dropped = false;
+    has_prefix_ = false;
+  }
+
+  void add_op (op_code_t op, const byte_str_ref_t& str_ref)
+  {
+    if (!is_parsed ())
+      SUPER::add_op (op, str_ref);
+  }
+
+  void add_call_op (op_code_t op, const byte_str_ref_t& str_ref, unsigned int subr_num)
+  {
+    if (!is_parsed ())
+    {
+      unsigned int parsed_len = get_count ();
+      if (likely (parsed_len > 0))
+	values[parsed_len-1].set_skip ();
+
+      parsed_cs_op_t val;
+      val.init (subr_num);
+      SUPER::add_op (op, str_ref, val);
+    }
+  }
+
+  void set_prefix (const number_t &num, op_code_t op = OpCode_Invalid)
+  {
+    has_prefix_ = true;
+    prefix_op_ = op;
+    prefix_num_ = num;
+  }
+
+  bool at_end (unsigned int pos) const
+  {
+    return ((pos + 1 >= values.length) /* CFF2 */
+	|| (values[pos + 1].op == OpCode_return));
+  }
+
+  bool is_parsed () const { return parsed; }
+  void set_parsed ()      { parsed = true; }
+
+  bool is_hint_dropped () const { return hint_dropped; }
+  void set_hint_dropped ()      { hint_dropped = true; }
+
+  bool is_vsindex_dropped () const { return vsindex_dropped; }
+  void set_vsindex_dropped ()      { vsindex_dropped = true; }
+
+  bool has_prefix () const          { return has_prefix_; }
+  op_code_t prefix_op () const         { return prefix_op_; }
+  const number_t &prefix_num () const { return prefix_num_; }
+
+  protected:
+  bool    parsed;
+  bool    hint_dropped;
+  bool    vsindex_dropped;
+  bool    has_prefix_;
+  op_code_t	prefix_op_;
+  number_t	prefix_num_;
+
+  private:
+  typedef parsed_values_t<parsed_cs_op_t> SUPER;
+};
+
+struct parsed_cs_str_vec_t : hb_vector_t<parsed_cs_str_t>
+{
+  void init (unsigned int len_ = 0)
+  {
+    SUPER::init ();
+    if (unlikely (!resize (len_)))
+      return;
+    for (unsigned int i = 0; i < length; i++)
+      (*this)[i].init ();
+  }
+  void fini () { SUPER::fini_deep (); }
+
+  private:
+  typedef hb_vector_t<parsed_cs_str_t> SUPER;
+};
+
+struct subr_subset_param_t
+{
+  void init (parsed_cs_str_t *parsed_charstring_,
+	     parsed_cs_str_vec_t *parsed_global_subrs_, parsed_cs_str_vec_t *parsed_local_subrs_,
+	     hb_set_t *global_closure_, hb_set_t *local_closure_,
+	     bool drop_hints_)
+  {
+    parsed_charstring = parsed_charstring_;
+    current_parsed_str = parsed_charstring;
+    parsed_global_subrs = parsed_global_subrs_;
+    parsed_local_subrs = parsed_local_subrs_;
+    global_closure = global_closure_;
+    local_closure = local_closure_;
+    drop_hints = drop_hints_;
+  }
+
+  parsed_cs_str_t *get_parsed_str_for_context (call_context_t &context)
+  {
+    switch (context.type)
+    {
+      case CSType_CharString:
+	return parsed_charstring;
+
+      case CSType_LocalSubr:
+	if (likely (context.subr_num < parsed_local_subrs->length))
+	  return &(*parsed_local_subrs)[context.subr_num];
+	break;
+
+      case CSType_GlobalSubr:
+	if (likely (context.subr_num < parsed_global_subrs->length))
+	  return &(*parsed_global_subrs)[context.subr_num];
+	break;
+    }
+    return nullptr;
+  }
+
+  template <typename ENV>
+  void set_current_str (ENV &env, bool calling)
+  {
+    parsed_cs_str_t *parsed_str = get_parsed_str_for_context (env.context);
+    if (unlikely (!parsed_str))
+    {
+      env.set_error ();
+      return;
+    }
+    /* If the called subroutine is parsed partially but not completely yet,
+     * it must be because we are calling it recursively.
+     * Handle it as an error. */
+    if (unlikely (calling && !parsed_str->is_parsed () && (parsed_str->values.length > 0)))
+      env.set_error ();
+    else
+      current_parsed_str = parsed_str;
+  }
+
+  parsed_cs_str_t	*current_parsed_str;
+
+  parsed_cs_str_t	*parsed_charstring;
+  parsed_cs_str_vec_t	*parsed_global_subrs;
+  parsed_cs_str_vec_t	*parsed_local_subrs;
+  hb_set_t      *global_closure;
+  hb_set_t      *local_closure;
+  bool	  drop_hints;
+};
+
+struct subr_remap_t : hb_inc_bimap_t
+{
+  void create (hb_set_t *closure)
+  {
+    /* create a remapping of subroutine numbers from old to new.
+     * no optimization based on usage counts. fonttools doesn't appear doing that either.
+     */
+
+    hb_codepoint_t old_num = HB_SET_VALUE_INVALID;
+    while (hb_set_next (closure, &old_num))
+      add (old_num);
+
+    if (get_population () < 1240)
+      bias = 107;
+    else if (get_population () < 33900)
+      bias = 1131;
+    else
+      bias = 32768;
+  }
+
+  int biased_num (unsigned int old_num) const
+  {
+    hb_codepoint_t new_num = get (old_num);
+    return (int)new_num - bias;
+  }
+
+  protected:
+  int bias;
+};
+
+struct subr_remaps_t
+{
+  subr_remaps_t ()
+  {
+    global_remap.init ();
+    local_remaps.init ();
+  }
+
+  ~subr_remaps_t () { fini (); }
+
+  void init (unsigned int fdCount)
+  {
+    if (unlikely (!local_remaps.resize (fdCount))) return;
+    for (unsigned int i = 0; i < fdCount; i++)
+      local_remaps[i].init ();
+  }
+
+  bool in_error()
+  {
+    return local_remaps.in_error ();
+  }
+
+  void create (subr_closures_t& closures)
+  {
+    global_remap.create (closures.global_closure);
+    for (unsigned int i = 0; i < local_remaps.length; i++)
+      local_remaps[i].create (closures.local_closures[i]);
+  }
+
+  void fini ()
+  {
+    global_remap.fini ();
+    local_remaps.fini_deep ();
+  }
+
+  subr_remap_t	       global_remap;
+  hb_vector_t<subr_remap_t>  local_remaps;
+};
+
+template <typename SUBSETTER, typename SUBRS, typename ACC, typename ENV, typename OPSET, op_code_t endchar_op=OpCode_Invalid>
+struct subr_subsetter_t
+{
+  subr_subsetter_t (ACC &acc_, const hb_subset_plan_t *plan_)
+    : acc (acc_), plan (plan_)
+  {
+    parsed_charstrings.init ();
+    parsed_global_subrs.init ();
+    parsed_local_subrs.init ();
+  }
+
+  ~subr_subsetter_t ()
+  {
+    closures.fini ();
+    remaps.fini ();
+    parsed_charstrings.fini_deep ();
+    parsed_global_subrs.fini_deep ();
+    parsed_local_subrs.fini_deep ();
+  }
+
+  /* Subroutine subsetting with --no-desubroutinize runs in phases:
+   *
+   * 1. execute charstrings/subroutines to determine subroutine closures
+   * 2. parse out all operators and numbers
+   * 3. mark hint operators and operands for removal if --no-hinting
+   * 4. re-encode all charstrings and subroutines with new subroutine numbers
+   *
+   * Phases #1 and #2 are done at the same time in collect_subrs ().
+   * Phase #3 walks charstrings/subroutines forward then backward (hence parsing required),
+   * because we can't tell if a number belongs to a hint op until we see the first moveto.
+   *
+   * Assumption: a callsubr/callgsubr operator must immediately follow a (biased) subroutine number
+   * within the same charstring/subroutine, e.g., not split across a charstring and a subroutine.
+   */
+  bool subset (void)
+  {
+    closures.init (acc.fdCount);
+    remaps.init (acc.fdCount);
+
+    parsed_charstrings.init (plan->num_output_glyphs ());
+    parsed_global_subrs.init (acc.globalSubrs->count);
+
+    if (unlikely (remaps.in_error()
+                  || parsed_charstrings.in_error ()
+                  || parsed_global_subrs.in_error ())) {
+      return false;
+    }
+
+    if (unlikely (!parsed_local_subrs.resize (acc.fdCount))) return false;
+
+    for (unsigned int i = 0; i < acc.fdCount; i++)
+    {
+      parsed_local_subrs[i].init (acc.privateDicts[i].localSubrs->count);
+      if (unlikely (parsed_local_subrs[i].in_error ())) return false;
+    }
+    if (unlikely (!closures.valid))
+      return false;
+
+    /* phase 1 & 2 */
+    for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+    {
+      hb_codepoint_t  glyph;
+      if (!plan->old_gid_for_new_gid (i, &glyph))
+	continue;
+      const byte_str_t str = (*acc.charStrings)[glyph];
+      unsigned int fd = acc.fdSelect->get_fd (glyph);
+      if (unlikely (fd >= acc.fdCount))
+	return false;
+
+      cs_interpreter_t<ENV, OPSET, subr_subset_param_t> interp;
+      interp.env.init (str, acc, fd);
+
+      subr_subset_param_t  param;
+      param.init (&parsed_charstrings[i],
+		  &parsed_global_subrs,  &parsed_local_subrs[fd],
+		  closures.global_closure, closures.local_closures[fd],
+		  plan->drop_hints);
+
+      if (unlikely (!interp.interpret (param)))
+	return false;
+
+      /* complete parsed string esp. copy CFF1 width or CFF2 vsindex to the parsed charstring for encoding */
+      SUBSETTER::complete_parsed_str (interp.env, param, parsed_charstrings[i]);
+    }
+
+    if (plan->drop_hints)
+    {
+      /* mark hint ops and arguments for drop */
+      for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+      {
+	hb_codepoint_t  glyph;
+	if (!plan->old_gid_for_new_gid (i, &glyph))
+	  continue;
+	unsigned int fd = acc.fdSelect->get_fd (glyph);
+	if (unlikely (fd >= acc.fdCount))
+	  return false;
+	subr_subset_param_t  param;
+	param.init (&parsed_charstrings[i],
+		    &parsed_global_subrs,  &parsed_local_subrs[fd],
+		    closures.global_closure, closures.local_closures[fd],
+		    plan->drop_hints);
+
+	drop_hints_param_t  drop;
+	if (drop_hints_in_str (parsed_charstrings[i], param, drop))
+	{
+	  parsed_charstrings[i].set_hint_dropped ();
+	  if (drop.vsindex_dropped)
+	    parsed_charstrings[i].set_vsindex_dropped ();
+	}
+      }
+
+      /* after dropping hints recreate closures of actually used subrs */
+      closures.reset ();
+      for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+      {
+	hb_codepoint_t  glyph;
+	if (!plan->old_gid_for_new_gid (i, &glyph))
+	  continue;
+	unsigned int fd = acc.fdSelect->get_fd (glyph);
+	if (unlikely (fd >= acc.fdCount))
+	  return false;
+	subr_subset_param_t  param;
+	param.init (&parsed_charstrings[i],
+		    &parsed_global_subrs,  &parsed_local_subrs[fd],
+		    closures.global_closure, closures.local_closures[fd],
+		    plan->drop_hints);
+	collect_subr_refs_in_str (parsed_charstrings[i], param);
+      }
+    }
+
+    remaps.create (closures);
+
+    return true;
+  }
+
+  bool encode_charstrings (str_buff_vec_t &buffArray) const
+  {
+    if (unlikely (!buffArray.resize (plan->num_output_glyphs ())))
+      return false;
+    for (unsigned int i = 0; i < plan->num_output_glyphs (); i++)
+    {
+      hb_codepoint_t  glyph;
+      if (!plan->old_gid_for_new_gid (i, &glyph))
+      {
+	/* add an endchar only charstring for a missing glyph if CFF1 */
+	if (endchar_op != OpCode_Invalid) buffArray[i].push (endchar_op);
+	continue;
+      }
+      unsigned int  fd = acc.fdSelect->get_fd (glyph);
+      if (unlikely (fd >= acc.fdCount))
+	return false;
+      if (unlikely (!encode_str (parsed_charstrings[i], fd, buffArray[i])))
+	return false;
+    }
+    return true;
+  }
+
+  bool encode_subrs (const parsed_cs_str_vec_t &subrs, const subr_remap_t& remap, unsigned int fd, str_buff_vec_t &buffArray) const
+  {
+    unsigned int  count = remap.get_population ();
+
+    if (unlikely (!buffArray.resize (count)))
+      return false;
+    for (unsigned int old_num = 0; old_num < subrs.length; old_num++)
+    {
+      hb_codepoint_t new_num = remap[old_num];
+      if (new_num != CFF_UNDEF_CODE)
+      {
+	if (unlikely (!encode_str (subrs[old_num], fd, buffArray[new_num])))
+	  return false;
+      }
+    }
+    return true;
+  }
+
+  bool encode_globalsubrs (str_buff_vec_t &buffArray)
+  {
+    return encode_subrs (parsed_global_subrs, remaps.global_remap, 0, buffArray);
+  }
+
+  bool encode_localsubrs (unsigned int fd, str_buff_vec_t &buffArray) const
+  {
+    return encode_subrs (parsed_local_subrs[fd], remaps.local_remaps[fd], fd, buffArray);
+  }
+
+  protected:
+  struct drop_hints_param_t
+  {
+    drop_hints_param_t ()
+      : seen_moveto (false),
+	ends_in_hint (false),
+	all_dropped (false),
+	vsindex_dropped (false) {}
+
+    bool  seen_moveto;
+    bool  ends_in_hint;
+    bool  all_dropped;
+    bool  vsindex_dropped;
+  };
+
+  bool drop_hints_in_subr (parsed_cs_str_t &str, unsigned int pos,
+			   parsed_cs_str_vec_t &subrs, unsigned int subr_num,
+			   const subr_subset_param_t &param, drop_hints_param_t &drop)
+  {
+    drop.ends_in_hint = false;
+    bool has_hint = drop_hints_in_str (subrs[subr_num], param, drop);
+
+    /* if this subr ends with a stem hint (i.e., not a number; potential argument for moveto),
+     * then this entire subroutine must be a hint. drop its call. */
+    if (drop.ends_in_hint)
+    {
+      str.values[pos].set_drop ();
+      /* if this subr call is at the end of the parent subr, propagate the flag
+       * otherwise reset the flag */
+      if (!str.at_end (pos))
+	drop.ends_in_hint = false;
+    }
+    else if (drop.all_dropped)
+    {
+      str.values[pos].set_drop ();
+    }
+
+    return has_hint;
+  }
+
+  /* returns true if it sees a hint op before the first moveto */
+  bool drop_hints_in_str (parsed_cs_str_t &str, const subr_subset_param_t &param, drop_hints_param_t &drop)
+  {
+    bool  seen_hint = false;
+
+    for (unsigned int pos = 0; pos < str.values.length; pos++)
+    {
+      bool  has_hint = false;
+      switch (str.values[pos].op)
+      {
+	case OpCode_callsubr:
+	  has_hint = drop_hints_in_subr (str, pos,
+					*param.parsed_local_subrs, str.values[pos].subr_num,
+					param, drop);
+	  break;
+
+	case OpCode_callgsubr:
+	  has_hint = drop_hints_in_subr (str, pos,
+					*param.parsed_global_subrs, str.values[pos].subr_num,
+					param, drop);
+	  break;
+
+	case OpCode_rmoveto:
+	case OpCode_hmoveto:
+	case OpCode_vmoveto:
+	  drop.seen_moveto = true;
+	  break;
+
+	case OpCode_hintmask:
+	case OpCode_cntrmask:
+	  if (drop.seen_moveto)
+	  {
+	    str.values[pos].set_drop ();
+	    break;
+	  }
+	  HB_FALLTHROUGH;
+
+	case OpCode_hstemhm:
+	case OpCode_vstemhm:
+	case OpCode_hstem:
+	case OpCode_vstem:
+	  has_hint = true;
+	  str.values[pos].set_drop ();
+	  if (str.at_end (pos))
+	    drop.ends_in_hint = true;
+	  break;
+
+	case OpCode_dotsection:
+	  str.values[pos].set_drop ();
+	  break;
+
+	default:
+	  /* NONE */
+	  break;
+      }
+      if (has_hint)
+      {
+	for (int i = pos - 1; i >= 0; i--)
+	{
+	  parsed_cs_op_t  &csop = str.values[(unsigned)i];
+	  if (csop.for_drop ())
+	    break;
+	  csop.set_drop ();
+	  if (csop.op == OpCode_vsindexcs)
+	    drop.vsindex_dropped = true;
+	}
+	seen_hint |= has_hint;
+      }
+    }
+
+    /* Raise all_dropped flag if all operators except return are dropped from a subr.
+     * It may happen even after seeing the first moveto if a subr contains
+     * only (usually one) hintmask operator, then calls to this subr can be dropped.
+     */
+    drop.all_dropped = true;
+    for (unsigned int pos = 0; pos < str.values.length; pos++)
+    {
+      parsed_cs_op_t  &csop = str.values[pos];
+      if (csop.op == OpCode_return)
+	break;
+      if (!csop.for_drop ())
+      {
+	drop.all_dropped = false;
+	break;
+      }
+    }
+
+    return seen_hint;
+  }
+
+  void collect_subr_refs_in_subr (parsed_cs_str_t &str, unsigned int pos,
+				  unsigned int subr_num, parsed_cs_str_vec_t &subrs,
+				  hb_set_t *closure,
+				  const subr_subset_param_t &param)
+  {
+    closure->add (subr_num);
+    collect_subr_refs_in_str (subrs[subr_num], param);
+  }
+
+  void collect_subr_refs_in_str (parsed_cs_str_t &str, const subr_subset_param_t &param)
+  {
+    for (unsigned int pos = 0; pos < str.values.length; pos++)
+    {
+      if (!str.values[pos].for_drop ())
+      {
+	switch (str.values[pos].op)
+	{
+	  case OpCode_callsubr:
+	    collect_subr_refs_in_subr (str, pos,
+				       str.values[pos].subr_num, *param.parsed_local_subrs,
+				       param.local_closure, param);
+	    break;
+
+	  case OpCode_callgsubr:
+	    collect_subr_refs_in_subr (str, pos,
+				       str.values[pos].subr_num, *param.parsed_global_subrs,
+				       param.global_closure, param);
+	    break;
+
+	  default: break;
+	}
+      }
+    }
+  }
+
+  bool encode_str (const parsed_cs_str_t &str, const unsigned int fd, str_buff_t &buff) const
+  {
+    buff.init ();
+    str_encoder_t  encoder (buff);
+    encoder.reset ();
+    /* if a prefix (CFF1 width or CFF2 vsindex) has been removed along with hints,
+     * re-insert it at the beginning of charstreing */
+    if (str.has_prefix () && str.is_hint_dropped ())
+    {
+      encoder.encode_num (str.prefix_num ());
+      if (str.prefix_op () != OpCode_Invalid)
+	encoder.encode_op (str.prefix_op ());
+    }
+    for (unsigned int i = 0; i < str.get_count(); i++)
+    {
+      const parsed_cs_op_t  &opstr = str.values[i];
+      if (!opstr.for_drop () && !opstr.for_skip ())
+      {
+	switch (opstr.op)
+	{
+	  case OpCode_callsubr:
+	    encoder.encode_int (remaps.local_remaps[fd].biased_num (opstr.subr_num));
+	    encoder.encode_op (OpCode_callsubr);
+	    break;
+
+	  case OpCode_callgsubr:
+	    encoder.encode_int (remaps.global_remap.biased_num (opstr.subr_num));
+	    encoder.encode_op (OpCode_callgsubr);
+	    break;
+
+	  default:
+	    encoder.copy_str (opstr.str);
+	    break;
+	}
+      }
+    }
+    return !encoder.is_error ();
+  }
+
+  protected:
+  const ACC			&acc;
+  const hb_subset_plan_t	*plan;
+
+  subr_closures_t		closures;
+
+  parsed_cs_str_vec_t		parsed_charstrings;
+  parsed_cs_str_vec_t		parsed_global_subrs;
+  hb_vector_t<parsed_cs_str_vec_t>  parsed_local_subrs;
+
+  subr_remaps_t			remaps;
+
+  private:
+  typedef typename SUBRS::count_type subr_count_type;
+};
+
+} /* namespace CFF */
+
+HB_INTERNAL bool
+hb_plan_subset_cff_fdselect (const hb_subset_plan_t *plan,
+			    unsigned int fdCount,
+			    const CFF::FDSelect &src, /* IN */
+			    unsigned int &subset_fd_count /* OUT */,
+			    unsigned int &subset_fdselect_size /* OUT */,
+			    unsigned int &subset_fdselect_format /* OUT */,
+			    hb_vector_t<CFF::code_pair_t> &fdselect_ranges /* OUT */,
+			    hb_inc_bimap_t &fdmap /* OUT */);
+
+HB_INTERNAL bool
+hb_serialize_cff_fdselect (hb_serialize_context_t *c,
+			  unsigned int num_glyphs,
+			  const CFF::FDSelect &src,
+			  unsigned int fd_count,
+			  unsigned int fdselect_format,
+			  unsigned int size,
+			  const hb_vector_t<CFF::code_pair_t> &fdselect_ranges);
+
+#endif /* HB_SUBSET_CFF_COMMON_HH */
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff1.cc b/thirdparty/harfbuzz/src/hb-subset-cff1.cc
new file mode 100644
index 0000000000..df322f8451
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff1.cc
@@ -0,0 +1,940 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_SUBSET_CFF
+
+#include "hb-open-type.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-set.h"
+#include "hb-bimap.hh"
+#include "hb-subset-cff1.hh"
+#include "hb-subset-plan.hh"
+#include "hb-subset-cff-common.hh"
+#include "hb-cff1-interp-cs.hh"
+
+using namespace CFF;
+
+struct remap_sid_t : hb_inc_bimap_t
+{
+  unsigned int add (unsigned int sid)
+  {
+    if ((sid != CFF_UNDEF_SID) && !is_std_std (sid))
+      return offset_sid (hb_inc_bimap_t::add (unoffset_sid (sid)));
+    else
+      return sid;
+  }
+
+  unsigned int operator[] (unsigned int sid) const
+  {
+    if (is_std_std (sid) || (sid == CFF_UNDEF_SID))
+      return sid;
+    else
+      return offset_sid (get (unoffset_sid (sid)));
+  }
+
+  static const unsigned int num_std_strings = 391;
+
+  static bool is_std_std (unsigned int sid) { return sid < num_std_strings; }
+  static unsigned int offset_sid (unsigned int sid) { return sid + num_std_strings; }
+  static unsigned int unoffset_sid (unsigned int sid) { return sid - num_std_strings; }
+};
+
+struct cff1_sub_table_info_t : cff_sub_table_info_t
+{
+  cff1_sub_table_info_t ()
+    : cff_sub_table_info_t (),
+      encoding_link (0),
+      charset_link (0)
+   {
+    privateDictInfo.init ();
+  }
+
+  objidx_t	encoding_link;
+  objidx_t	charset_link;
+  table_info_t	privateDictInfo;
+};
+
+/* a copy of a parsed out cff1_top_dict_values_t augmented with additional operators */
+struct cff1_top_dict_values_mod_t : cff1_top_dict_values_t
+{
+  void init (const cff1_top_dict_values_t *base_= &Null (cff1_top_dict_values_t))
+  {
+    SUPER::init ();
+    base = base_;
+  }
+
+  void fini () { SUPER::fini (); }
+
+  unsigned get_count () const { return base->get_count () + SUPER::get_count (); }
+  const cff1_top_dict_val_t &get_value (unsigned int i) const
+  {
+    if (i < base->get_count ())
+      return (*base)[i];
+    else
+      return SUPER::values[i - base->get_count ()];
+  }
+  const cff1_top_dict_val_t &operator [] (unsigned int i) const { return get_value (i); }
+
+  void reassignSIDs (const remap_sid_t& sidmap)
+  {
+    for (unsigned int i = 0; i < name_dict_values_t::ValCount; i++)
+      nameSIDs[i] = sidmap[base->nameSIDs[i]];
+  }
+
+  protected:
+  typedef cff1_top_dict_values_t SUPER;
+  const cff1_top_dict_values_t *base;
+};
+
+struct top_dict_modifiers_t
+{
+  top_dict_modifiers_t (const cff1_sub_table_info_t &info_,
+			const unsigned int (&nameSIDs_)[name_dict_values_t::ValCount])
+    : info (info_),
+      nameSIDs (nameSIDs_)
+  {}
+
+  const cff1_sub_table_info_t &info;
+  const unsigned int	(&nameSIDs)[name_dict_values_t::ValCount];
+};
+
+struct cff1_top_dict_op_serializer_t : cff_top_dict_op_serializer_t<cff1_top_dict_val_t>
+{
+  bool serialize (hb_serialize_context_t *c,
+		  const cff1_top_dict_val_t &opstr,
+		  const top_dict_modifiers_t &mod) const
+  {
+    TRACE_SERIALIZE (this);
+
+    op_code_t op = opstr.op;
+    switch (op)
+    {
+      case OpCode_charset:
+	if (mod.info.charset_link)
+	  return_trace (FontDict::serialize_link4_op(c, op, mod.info.charset_link, whence_t::Absolute));
+	else
+	  goto fall_back;
+
+      case OpCode_Encoding:
+	if (mod.info.encoding_link)
+	  return_trace (FontDict::serialize_link4_op(c, op, mod.info.encoding_link, whence_t::Absolute));
+	else
+	  goto fall_back;
+
+      case OpCode_Private:
+	return_trace (UnsizedByteStr::serialize_int2 (c, mod.info.privateDictInfo.size) &&
+		      Dict::serialize_link4_op (c, op, mod.info.privateDictInfo.link, whence_t::Absolute));
+
+      case OpCode_version:
+      case OpCode_Notice:
+      case OpCode_Copyright:
+      case OpCode_FullName:
+      case OpCode_FamilyName:
+      case OpCode_Weight:
+      case OpCode_PostScript:
+      case OpCode_BaseFontName:
+      case OpCode_FontName:
+	return_trace (FontDict::serialize_int2_op (c, op, mod.nameSIDs[name_dict_values_t::name_op_to_index (op)]));
+
+      case OpCode_ROS:
+	{
+	  /* for registry & ordering, reassigned SIDs are serialized
+	   * for supplement, the original byte string is copied along with the op code */
+	  op_str_t supp_op;
+	  supp_op.op = op;
+	  if ( unlikely (!(opstr.str.length >= opstr.last_arg_offset + 3)))
+	    return_trace (false);
+	  supp_op.str = byte_str_t (&opstr.str + opstr.last_arg_offset, opstr.str.length - opstr.last_arg_offset);
+	  return_trace (UnsizedByteStr::serialize_int2 (c, mod.nameSIDs[name_dict_values_t::registry]) &&
+			UnsizedByteStr::serialize_int2 (c, mod.nameSIDs[name_dict_values_t::ordering]) &&
+			copy_opstr (c, supp_op));
+	}
+      fall_back:
+      default:
+	return_trace (cff_top_dict_op_serializer_t<cff1_top_dict_val_t>::serialize (c, opstr, mod.info));
+    }
+    return_trace (true);
+  }
+
+};
+
+struct cff1_font_dict_op_serializer_t : cff_font_dict_op_serializer_t
+{
+  bool serialize (hb_serialize_context_t *c,
+		  const op_str_t &opstr,
+		  const cff1_font_dict_values_mod_t &mod) const
+  {
+    TRACE_SERIALIZE (this);
+
+    if (opstr.op == OpCode_FontName)
+      return_trace (FontDict::serialize_int2_op (c, opstr.op, mod.fontName));
+    else
+      return_trace (SUPER::serialize (c, opstr, mod.privateDictInfo));
+  }
+
+  private:
+  typedef cff_font_dict_op_serializer_t SUPER;
+};
+
+struct cff1_cs_opset_flatten_t : cff1_cs_opset_t<cff1_cs_opset_flatten_t, flatten_param_t>
+{
+  static void flush_args_and_op (op_code_t op, cff1_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    if (env.arg_start > 0)
+      flush_width (env, param);
+
+    switch (op)
+    {
+      case OpCode_hstem:
+      case OpCode_hstemhm:
+      case OpCode_vstem:
+      case OpCode_vstemhm:
+      case OpCode_hintmask:
+      case OpCode_cntrmask:
+      case OpCode_dotsection:
+	if (param.drop_hints)
+	{
+	  env.clear_args ();
+	  return;
+	}
+	HB_FALLTHROUGH;
+
+      default:
+	SUPER::flush_args_and_op (op, env, param);
+	break;
+    }
+  }
+  static void flush_args (cff1_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    str_encoder_t  encoder (param.flatStr);
+    for (unsigned int i = env.arg_start; i < env.argStack.get_count (); i++)
+      encoder.encode_num (env.eval_arg (i));
+    SUPER::flush_args (env, param);
+  }
+
+  static void flush_op (op_code_t op, cff1_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    str_encoder_t  encoder (param.flatStr);
+    encoder.encode_op (op);
+  }
+
+  static void flush_width (cff1_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    assert (env.has_width);
+    str_encoder_t  encoder (param.flatStr);
+    encoder.encode_num (env.width);
+  }
+
+  static void flush_hintmask (op_code_t op, cff1_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    SUPER::flush_hintmask (op, env, param);
+    if (!param.drop_hints)
+    {
+      str_encoder_t  encoder (param.flatStr);
+      for (unsigned int i = 0; i < env.hintmask_size; i++)
+	encoder.encode_byte (env.str_ref[i]);
+    }
+  }
+
+  private:
+  typedef cff1_cs_opset_t<cff1_cs_opset_flatten_t, flatten_param_t> SUPER;
+};
+
+struct range_list_t : hb_vector_t<code_pair_t>
+{
+  /* replace the first glyph ID in the "glyph" field each range with a nLeft value */
+  bool complete (unsigned int last_glyph)
+  {
+    bool  two_byte = false;
+    for (unsigned int i = (*this).length; i > 0; i--)
+    {
+      code_pair_t &pair = (*this)[i - 1];
+      unsigned int  nLeft = last_glyph - pair.glyph - 1;
+      if (nLeft >= 0x100)
+	two_byte = true;
+      last_glyph = pair.glyph;
+      pair.glyph = nLeft;
+    }
+    return two_byte;
+  }
+};
+
+struct cff1_cs_opset_subr_subset_t : cff1_cs_opset_t<cff1_cs_opset_subr_subset_t, subr_subset_param_t>
+{
+  static void process_op (op_code_t op, cff1_cs_interp_env_t &env, subr_subset_param_t& param)
+  {
+    switch (op) {
+
+      case OpCode_return:
+	param.current_parsed_str->add_op (op, env.str_ref);
+	param.current_parsed_str->set_parsed ();
+	env.return_from_subr ();
+	param.set_current_str (env, false);
+	break;
+
+      case OpCode_endchar:
+	param.current_parsed_str->add_op (op, env.str_ref);
+	param.current_parsed_str->set_parsed ();
+	SUPER::process_op (op, env, param);
+	break;
+
+      case OpCode_callsubr:
+	process_call_subr (op, CSType_LocalSubr, env, param, env.localSubrs, param.local_closure);
+	break;
+
+      case OpCode_callgsubr:
+	process_call_subr (op, CSType_GlobalSubr, env, param, env.globalSubrs, param.global_closure);
+	break;
+
+      default:
+	SUPER::process_op (op, env, param);
+	param.current_parsed_str->add_op (op, env.str_ref);
+	break;
+    }
+  }
+
+  protected:
+  static void process_call_subr (op_code_t op, cs_type_t type,
+				 cff1_cs_interp_env_t &env, subr_subset_param_t& param,
+				 cff1_biased_subrs_t& subrs, hb_set_t *closure)
+  {
+    byte_str_ref_t    str_ref = env.str_ref;
+    env.call_subr (subrs, type);
+    param.current_parsed_str->add_call_op (op, str_ref, env.context.subr_num);
+    closure->add (env.context.subr_num);
+    param.set_current_str (env, true);
+  }
+
+  private:
+  typedef cff1_cs_opset_t<cff1_cs_opset_subr_subset_t, subr_subset_param_t> SUPER;
+};
+
+struct cff1_subr_subsetter_t : subr_subsetter_t<cff1_subr_subsetter_t, CFF1Subrs, const OT::cff1::accelerator_subset_t, cff1_cs_interp_env_t, cff1_cs_opset_subr_subset_t, OpCode_endchar>
+{
+  cff1_subr_subsetter_t (const OT::cff1::accelerator_subset_t &acc_, const hb_subset_plan_t *plan_)
+    : subr_subsetter_t (acc_, plan_) {}
+
+  static void complete_parsed_str (cff1_cs_interp_env_t &env, subr_subset_param_t& param, parsed_cs_str_t &charstring)
+  {
+    /* insert width at the beginning of the charstring as necessary */
+    if (env.has_width)
+      charstring.set_prefix (env.width);
+
+    /* subroutines/charstring left on the call stack are legally left unmarked
+     * unmarked when a subroutine terminates with endchar. mark them.
+     */
+    param.current_parsed_str->set_parsed ();
+    for (unsigned int i = 0; i < env.callStack.get_count (); i++)
+    {
+      parsed_cs_str_t *parsed_str = param.get_parsed_str_for_context (env.callStack[i]);
+      if (likely (parsed_str))
+	parsed_str->set_parsed ();
+      else
+	env.set_error ();
+    }
+  }
+};
+
+struct cff_subset_plan {
+  cff_subset_plan ()
+    : info (),
+      orig_fdcount (0),
+      subset_fdcount (1),
+      subset_fdselect_format (0),
+      drop_hints (false),
+      desubroutinize(false)
+  {
+    topdict_mod.init ();
+    subset_fdselect_ranges.init ();
+    fdmap.init ();
+    subset_charstrings.init ();
+    subset_globalsubrs.init ();
+    subset_localsubrs.init ();
+    fontdicts_mod.init ();
+    subset_enc_code_ranges.init ();
+    subset_enc_supp_codes.init ();
+    subset_charset_ranges.init ();
+    sidmap.init ();
+    for (unsigned int i = 0; i < name_dict_values_t::ValCount; i++)
+      topDictModSIDs[i] = CFF_UNDEF_SID;
+  }
+
+  ~cff_subset_plan ()
+  {
+    topdict_mod.fini ();
+    subset_fdselect_ranges.fini ();
+    fdmap.fini ();
+    subset_charstrings.fini_deep ();
+    subset_globalsubrs.fini_deep ();
+    subset_localsubrs.fini_deep ();
+    fontdicts_mod.fini ();
+    subset_enc_code_ranges.fini ();
+    subset_enc_supp_codes.fini ();
+    subset_charset_ranges.fini ();
+    sidmap.fini ();
+  }
+
+  void plan_subset_encoding (const OT::cff1::accelerator_subset_t &acc, hb_subset_plan_t *plan)
+  {
+    const Encoding *encoding = acc.encoding;
+    unsigned int  size0, size1, supp_size;
+    hb_codepoint_t  code, last_code = CFF_UNDEF_CODE;
+    hb_vector_t<hb_codepoint_t> supp_codes;
+
+    if (unlikely (!subset_enc_code_ranges.resize (0)))
+    {
+      plan->check_success (false);
+      return;
+    }
+
+    supp_size = 0;
+    supp_codes.init ();
+
+    subset_enc_num_codes = plan->num_output_glyphs () - 1;
+    unsigned int glyph;
+    for (glyph = 1; glyph < plan->num_output_glyphs (); glyph++)
+    {
+      hb_codepoint_t  old_glyph;
+      if (!plan->old_gid_for_new_gid (glyph, &old_glyph))
+      {
+	/* Retain the code for the old missing glyph ID */
+	old_glyph = glyph;
+      }
+      code = acc.glyph_to_code (old_glyph);
+      if (code == CFF_UNDEF_CODE)
+      {
+	subset_enc_num_codes = glyph - 1;
+	break;
+      }
+
+      if ((last_code == CFF_UNDEF_CODE) || (code != last_code + 1))
+      {
+	code_pair_t pair = { code, glyph };
+	subset_enc_code_ranges.push (pair);
+      }
+      last_code = code;
+
+      if (encoding != &Null (Encoding))
+      {
+	hb_codepoint_t  sid = acc.glyph_to_sid (old_glyph);
+	encoding->get_supplement_codes (sid, supp_codes);
+	for (unsigned int i = 0; i < supp_codes.length; i++)
+	{
+	  code_pair_t pair = { supp_codes[i], sid };
+	  subset_enc_supp_codes.push (pair);
+	}
+	supp_size += SuppEncoding::static_size * supp_codes.length;
+      }
+    }
+    supp_codes.fini ();
+
+    subset_enc_code_ranges.complete (glyph);
+
+    assert (subset_enc_num_codes <= 0xFF);
+    size0 = Encoding0::min_size + HBUINT8::static_size * subset_enc_num_codes;
+    size1 = Encoding1::min_size + Encoding1_Range::static_size * subset_enc_code_ranges.length;
+
+    if (size0 < size1)
+      subset_enc_format = 0;
+    else
+      subset_enc_format = 1;
+  }
+
+  void plan_subset_charset (const OT::cff1::accelerator_subset_t &acc, hb_subset_plan_t *plan)
+  {
+    unsigned int  size0, size_ranges;
+    hb_codepoint_t  sid, last_sid = CFF_UNDEF_CODE;
+
+    if (unlikely (!subset_charset_ranges.resize (0)))
+    {
+      plan->check_success (false);
+      return;
+    }
+
+    unsigned int glyph;
+    for (glyph = 1; glyph < plan->num_output_glyphs (); glyph++)
+    {
+      hb_codepoint_t  old_glyph;
+      if (!plan->old_gid_for_new_gid (glyph, &old_glyph))
+      {
+	/* Retain the SID for the old missing glyph ID */
+	old_glyph = glyph;
+      }
+      sid = acc.glyph_to_sid (old_glyph);
+
+      if (!acc.is_CID ())
+	sid = sidmap.add (sid);
+
+      if ((last_sid == CFF_UNDEF_CODE) || (sid != last_sid + 1))
+      {
+	code_pair_t pair = { sid, glyph };
+	subset_charset_ranges.push (pair);
+      }
+      last_sid = sid;
+    }
+
+    bool two_byte = subset_charset_ranges.complete (glyph);
+
+    size0 = Charset0::min_size + HBUINT16::static_size * (plan->num_output_glyphs () - 1);
+    if (!two_byte)
+      size_ranges = Charset1::min_size + Charset1_Range::static_size * subset_charset_ranges.length;
+    else
+      size_ranges = Charset2::min_size + Charset2_Range::static_size * subset_charset_ranges.length;
+
+    if (size0 < size_ranges)
+      subset_charset_format = 0;
+    else if (!two_byte)
+      subset_charset_format = 1;
+    else
+      subset_charset_format = 2;
+  }
+
+  bool collect_sids_in_dicts (const OT::cff1::accelerator_subset_t &acc)
+  {
+    sidmap.reset ();
+
+    for (unsigned int i = 0; i < name_dict_values_t::ValCount; i++)
+    {
+      unsigned int sid = acc.topDict.nameSIDs[i];
+      if (sid != CFF_UNDEF_SID)
+      {
+	(void)sidmap.add (sid);
+	topDictModSIDs[i] = sidmap[sid];
+      }
+    }
+
+    if (acc.fdArray != &Null (CFF1FDArray))
+      for (unsigned int i = 0; i < orig_fdcount; i++)
+	if (fdmap.has (i))
+	  (void)sidmap.add (acc.fontDicts[i].fontName);
+
+    return true;
+  }
+
+  bool create (const OT::cff1::accelerator_subset_t &acc,
+	       hb_subset_plan_t *plan)
+  {
+    /* make sure notdef is first */
+    hb_codepoint_t old_glyph;
+    if (!plan->old_gid_for_new_gid (0, &old_glyph) || (old_glyph != 0)) return false;
+
+    num_glyphs = plan->num_output_glyphs ();
+    orig_fdcount = acc.fdCount;
+    drop_hints = plan->drop_hints;
+    desubroutinize = plan->desubroutinize;
+
+    /* check whether the subset renumbers any glyph IDs */
+    gid_renum = false;
+    for (hb_codepoint_t new_glyph = 0; new_glyph < plan->num_output_glyphs (); new_glyph++)
+    {
+      if (!plan->old_gid_for_new_gid(new_glyph, &old_glyph))
+	continue;
+      if (new_glyph != old_glyph) {
+	gid_renum = true;
+	break;
+      }
+    }
+
+    subset_charset = gid_renum || !acc.is_predef_charset ();
+    subset_encoding = !acc.is_CID() && !acc.is_predef_encoding ();
+
+    /* top dict INDEX */
+    {
+      /* Add encoding/charset to a (copy of) top dict as necessary */
+      topdict_mod.init (&acc.topDict);
+      bool need_to_add_enc = (subset_encoding && !acc.topDict.has_op (OpCode_Encoding));
+      bool need_to_add_set = (subset_charset && !acc.topDict.has_op (OpCode_charset));
+      if (need_to_add_enc || need_to_add_set)
+      {
+	if (need_to_add_enc)
+	  topdict_mod.add_op (OpCode_Encoding);
+	if (need_to_add_set)
+	  topdict_mod.add_op (OpCode_charset);
+      }
+    }
+
+    /* Determine re-mapping of font index as fdmap among other info */
+    if (acc.fdSelect != &Null (CFF1FDSelect))
+    {
+	if (unlikely (!hb_plan_subset_cff_fdselect (plan,
+				  orig_fdcount,
+				  *acc.fdSelect,
+				  subset_fdcount,
+				  info.fd_select.size,
+				  subset_fdselect_format,
+				  subset_fdselect_ranges,
+				  fdmap)))
+	return false;
+    }
+    else
+      fdmap.identity (1);
+
+    /* remove unused SIDs & reassign SIDs */
+    {
+      /* SIDs for name strings in dicts are added before glyph names so they fit in 16-bit int range */
+      if (unlikely (!collect_sids_in_dicts (acc)))
+	return false;
+      if (unlikely (sidmap.get_population () > 0x8000))	/* assumption: a dict won't reference that many strings */
+	return false;
+
+      if (subset_charset) plan_subset_charset (acc, plan);
+
+      topdict_mod.reassignSIDs (sidmap);
+    }
+
+    if (desubroutinize)
+    {
+      /* Flatten global & local subrs */
+      subr_flattener_t<const OT::cff1::accelerator_subset_t, cff1_cs_interp_env_t, cff1_cs_opset_flatten_t, OpCode_endchar>
+		    flattener(acc, plan);
+      if (!flattener.flatten (subset_charstrings))
+	return false;
+    }
+    else
+    {
+      cff1_subr_subsetter_t       subr_subsetter (acc, plan);
+
+      /* Subset subrs: collect used subroutines, leaving all unused ones behind */
+      if (!subr_subsetter.subset ())
+	return false;
+
+      /* encode charstrings, global subrs, local subrs with new subroutine numbers */
+      if (!subr_subsetter.encode_charstrings (subset_charstrings))
+	return false;
+
+      if (!subr_subsetter.encode_globalsubrs (subset_globalsubrs))
+	return false;
+
+      /* local subrs */
+      if (!subset_localsubrs.resize (orig_fdcount))
+	return false;
+      for (unsigned int fd = 0; fd < orig_fdcount; fd++)
+      {
+	subset_localsubrs[fd].init ();
+	if (fdmap.has (fd))
+	{
+	  if (!subr_subsetter.encode_localsubrs (fd, subset_localsubrs[fd]))
+	    return false;
+	}
+      }
+    }
+
+    /* Encoding */
+    if (subset_encoding)
+      plan_subset_encoding (acc, plan);
+
+    /* private dicts & local subrs */
+    if (!acc.is_CID ())
+      fontdicts_mod.push (cff1_font_dict_values_mod_t ());
+    else
+    {
+      + hb_iter (acc.fontDicts)
+      | hb_filter ([&] (const cff1_font_dict_values_t &_)
+	{ return fdmap.has (&_ - &acc.fontDicts[0]); } )
+      | hb_map ([&] (const cff1_font_dict_values_t &_)
+	{
+	  cff1_font_dict_values_mod_t mod;
+	  mod.init (&_, sidmap[_.fontName]);
+	  return mod;
+	})
+      | hb_sink (fontdicts_mod)
+      ;
+    }
+
+    return ((subset_charstrings.length == plan->num_output_glyphs ())
+	   && (fontdicts_mod.length == subset_fdcount));
+  }
+
+  cff1_top_dict_values_mod_t	topdict_mod;
+  cff1_sub_table_info_t		info;
+
+  unsigned int    num_glyphs;
+  unsigned int    orig_fdcount;
+  unsigned int    subset_fdcount;
+  unsigned int    subset_fdselect_format;
+  hb_vector_t<code_pair_t>   subset_fdselect_ranges;
+
+  /* font dict index remap table from fullset FDArray to subset FDArray.
+   * set to CFF_UNDEF_CODE if excluded from subset */
+  hb_inc_bimap_t   fdmap;
+
+  str_buff_vec_t		subset_charstrings;
+  str_buff_vec_t		subset_globalsubrs;
+  hb_vector_t<str_buff_vec_t>	subset_localsubrs;
+  hb_vector_t<cff1_font_dict_values_mod_t>  fontdicts_mod;
+
+  bool		drop_hints;
+
+  bool		gid_renum;
+  bool		subset_encoding;
+  uint8_t	subset_enc_format;
+  unsigned int	subset_enc_num_codes;
+  range_list_t	subset_enc_code_ranges;
+  hb_vector_t<code_pair_t>  subset_enc_supp_codes;
+
+  uint8_t	subset_charset_format;
+  range_list_t	subset_charset_ranges;
+  bool		subset_charset;
+
+  remap_sid_t	sidmap;
+  unsigned int	topDictModSIDs[name_dict_values_t::ValCount];
+
+  bool		desubroutinize;
+};
+
+static bool _serialize_cff1 (hb_serialize_context_t *c,
+			     cff_subset_plan &plan,
+			     const OT::cff1::accelerator_subset_t  &acc,
+			     unsigned int num_glyphs)
+{
+  /* private dicts & local subrs */
+  for (int i = (int)acc.privateDicts.length; --i >= 0 ;)
+  {
+    if (plan.fdmap.has (i))
+    {
+      objidx_t	subrs_link = 0;
+      if (plan.subset_localsubrs[i].length > 0)
+      {
+	CFF1Subrs *dest = c->start_embed <CFF1Subrs> ();
+	if (unlikely (!dest)) return false;
+	c->push ();
+	if (likely (dest && dest->serialize (c, plan.subset_localsubrs[i])))
+	  subrs_link = c->pop_pack ();
+	else
+	{
+	  c->pop_discard ();
+	  return false;
+	}
+      }
+
+      PrivateDict *pd = c->start_embed<PrivateDict> ();
+      if (unlikely (!pd)) return false;
+      c->push ();
+      cff_private_dict_op_serializer_t privSzr (plan.desubroutinize, plan.drop_hints);
+      /* N.B. local subrs immediately follows its corresponding private dict. i.e., subr offset == private dict size */
+      if (likely (pd->serialize (c, acc.privateDicts[i], privSzr, subrs_link)))
+      {
+	unsigned fd = plan.fdmap[i];
+	plan.fontdicts_mod[fd].privateDictInfo.size = c->length ();
+	plan.fontdicts_mod[fd].privateDictInfo.link = c->pop_pack ();
+      }
+      else
+      {
+	c->pop_discard ();
+	return false;
+      }
+    }
+  }
+
+  if (!acc.is_CID ())
+    plan.info.privateDictInfo = plan.fontdicts_mod[0].privateDictInfo;
+
+  /* CharStrings */
+  {
+    CFF1CharStrings  *cs = c->start_embed<CFF1CharStrings> ();
+    if (unlikely (!cs)) return false;
+    c->push ();
+    if (likely (cs->serialize (c, plan.subset_charstrings)))
+      plan.info.char_strings_link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* FDArray (FD Index) */
+  if (acc.fdArray != &Null (CFF1FDArray))
+  {
+    CFF1FDArray *fda = c->start_embed<CFF1FDArray> ();
+    if (unlikely (!fda)) return false;
+    c->push ();
+    cff1_font_dict_op_serializer_t  fontSzr;
+    auto it = + hb_zip (+ hb_iter (plan.fontdicts_mod), + hb_iter (plan.fontdicts_mod));
+    if (likely (fda->serialize (c, it, fontSzr)))
+      plan.info.fd_array_link = c->pop_pack (false);
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* FDSelect */
+  if (acc.fdSelect != &Null (CFF1FDSelect))
+  {
+    c->push ();
+    if (likely (hb_serialize_cff_fdselect (c, num_glyphs, *acc.fdSelect, acc.fdCount,
+					   plan.subset_fdselect_format, plan.info.fd_select.size,
+					   plan.subset_fdselect_ranges)))
+      plan.info.fd_select.link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* Charset */
+  if (plan.subset_charset)
+  {
+    Charset *dest = c->start_embed<Charset> ();
+    if (unlikely (!dest)) return false;
+    c->push ();
+    if (likely (dest->serialize (c,
+				 plan.subset_charset_format,
+				 plan.num_glyphs,
+				 plan.subset_charset_ranges)))
+      plan.info.charset_link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* Encoding */
+  if (plan.subset_encoding)
+  {
+    Encoding *dest = c->start_embed<Encoding> ();
+    if (unlikely (!dest)) return false;
+    c->push ();
+    if (likely (dest->serialize (c,
+				 plan.subset_enc_format,
+				 plan.subset_enc_num_codes,
+				 plan.subset_enc_code_ranges,
+				 plan.subset_enc_supp_codes)))
+      plan.info.encoding_link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* global subrs */
+  {
+    c->push ();
+    CFF1Subrs *dest = c->start_embed <CFF1Subrs> ();
+    if (unlikely (!dest)) return false;
+    if (likely (dest->serialize (c, plan.subset_globalsubrs)))
+      c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* String INDEX */
+  {
+    CFF1StringIndex *dest = c->start_embed<CFF1StringIndex> ();
+    if (unlikely (!dest)) return false;
+    c->push ();
+    if (likely (dest->serialize (c, *acc.stringIndex, plan.sidmap)))
+      c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  OT::cff1 *cff = c->allocate_min<OT::cff1> ();
+  if (unlikely (!cff))
+    return false;
+
+  /* header */
+  cff->version.major = 0x01;
+  cff->version.minor = 0x00;
+  cff->nameIndex = cff->min_size;
+  cff->offSize = 4; /* unused? */
+
+  /* name INDEX */
+  if (unlikely (!(*acc.nameIndex).copy (c))) return false;
+
+  /* top dict INDEX */
+  {
+    /* serialize singleton TopDict */
+    TopDict *top = c->start_embed<TopDict> ();
+    if (!top) return false;
+    c->push ();
+    cff1_top_dict_op_serializer_t topSzr;
+    unsigned top_size = 0;
+    top_dict_modifiers_t  modifier (plan.info, plan.topDictModSIDs);
+    if (likely (top->serialize (c, plan.topdict_mod, topSzr, modifier)))
+    {
+      top_size = c->length ();
+      c->pop_pack (false);
+    }
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+    /* serialize INDEX header for above */
+    CFF1Index *dest = c->start_embed<CFF1Index> ();
+    if (!dest) return false;
+    return dest->serialize_header (c, hb_iter (hb_array_t<unsigned> (&top_size, 1)));
+  }
+}
+
+static bool
+_hb_subset_cff1 (const OT::cff1::accelerator_subset_t  &acc,
+		hb_subset_context_t	*c)
+{
+  cff_subset_plan cff_plan;
+
+  if (unlikely (!cff_plan.create (acc, c->plan)))
+  {
+    DEBUG_MSG(SUBSET, nullptr, "Failed to generate a cff subsetting plan.");
+    return false;
+  }
+
+  return _serialize_cff1 (c->serializer, cff_plan, acc, c->plan->num_output_glyphs ());
+}
+
+/**
+ * hb_subset_cff1:
+ * Subsets the CFF table according to a provided plan.
+ *
+ * Return value: subsetted cff table.
+ **/
+bool
+hb_subset_cff1 (hb_subset_context_t *c)
+{
+  OT::cff1::accelerator_subset_t acc;
+  acc.init (c->plan->source);
+  bool result = likely (acc.is_valid ()) && _hb_subset_cff1 (acc, c);
+  acc.fini ();
+
+  return result;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff1.hh b/thirdparty/harfbuzz/src/hb-subset-cff1.hh
new file mode 100644
index 0000000000..aaf5def1ed
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff1.hh
@@ -0,0 +1,37 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_SUBSET_CFF1_HH
+#define HB_SUBSET_CFF1_HH
+
+#include "hb.hh"
+
+#include "hb-subset-plan.hh"
+
+HB_INTERNAL bool
+hb_subset_cff1 (hb_subset_context_t *c);
+
+#endif /* HB_SUBSET_CFF1_HH */
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff2.cc b/thirdparty/harfbuzz/src/hb-subset-cff2.cc
new file mode 100644
index 0000000000..17ee040deb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff2.cc
@@ -0,0 +1,488 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#include "hb.hh"
+
+#ifndef HB_NO_SUBSET_CFF
+
+#include "hb-open-type.hh"
+#include "hb-ot-cff2-table.hh"
+#include "hb-set.h"
+#include "hb-subset-cff2.hh"
+#include "hb-subset-plan.hh"
+#include "hb-subset-cff-common.hh"
+#include "hb-cff2-interp-cs.hh"
+
+using namespace CFF;
+
+struct cff2_sub_table_info_t : cff_sub_table_info_t
+{
+  cff2_sub_table_info_t ()
+    : cff_sub_table_info_t (),
+      var_store_link (0)
+  {}
+
+  objidx_t  var_store_link;
+};
+
+struct cff2_top_dict_op_serializer_t : cff_top_dict_op_serializer_t<>
+{
+  bool serialize (hb_serialize_context_t *c,
+		  const op_str_t &opstr,
+		  const cff2_sub_table_info_t &info) const
+  {
+    TRACE_SERIALIZE (this);
+
+    switch (opstr.op)
+    {
+      case OpCode_vstore:
+	return_trace (FontDict::serialize_link4_op(c, opstr.op, info.var_store_link));
+
+      default:
+	return_trace (cff_top_dict_op_serializer_t<>::serialize (c, opstr, info));
+    }
+  }
+};
+
+struct cff2_cs_opset_flatten_t : cff2_cs_opset_t<cff2_cs_opset_flatten_t, flatten_param_t>
+{
+  static void flush_args_and_op (op_code_t op, cff2_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    switch (op)
+    {
+      case OpCode_return:
+      case OpCode_endchar:
+	/* dummy opcodes in CFF2. ignore */
+	break;
+
+      case OpCode_hstem:
+      case OpCode_hstemhm:
+      case OpCode_vstem:
+      case OpCode_vstemhm:
+      case OpCode_hintmask:
+      case OpCode_cntrmask:
+	if (param.drop_hints)
+	{
+	  env.clear_args ();
+	  return;
+	}
+	HB_FALLTHROUGH;
+
+      default:
+	SUPER::flush_args_and_op (op, env, param);
+	break;
+    }
+  }
+
+  static void flush_args (cff2_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    for (unsigned int i = 0; i < env.argStack.get_count ();)
+    {
+      const blend_arg_t &arg = env.argStack[i];
+      if (arg.blending ())
+      {
+	if (unlikely (!((arg.numValues > 0) && (env.argStack.get_count () >= arg.numValues))))
+	{
+	  env.set_error ();
+	  return;
+	}
+	flatten_blends (arg, i, env, param);
+	i += arg.numValues;
+      }
+      else
+      {
+	str_encoder_t  encoder (param.flatStr);
+	encoder.encode_num (arg);
+	i++;
+      }
+    }
+    SUPER::flush_args (env, param);
+  }
+
+  static void flatten_blends (const blend_arg_t &arg, unsigned int i, cff2_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    /* flatten the default values */
+    str_encoder_t  encoder (param.flatStr);
+    for (unsigned int j = 0; j < arg.numValues; j++)
+    {
+      const blend_arg_t &arg1 = env.argStack[i + j];
+      if (unlikely (!((arg1.blending () && (arg.numValues == arg1.numValues) && (arg1.valueIndex == j) &&
+	      (arg1.deltas.length == env.get_region_count ())))))
+      {
+	env.set_error ();
+	return;
+      }
+      encoder.encode_num (arg1);
+    }
+    /* flatten deltas for each value */
+    for (unsigned int j = 0; j < arg.numValues; j++)
+    {
+      const blend_arg_t &arg1 = env.argStack[i + j];
+      for (unsigned int k = 0; k < arg1.deltas.length; k++)
+	encoder.encode_num (arg1.deltas[k]);
+    }
+    /* flatten the number of values followed by blend operator */
+    encoder.encode_int (arg.numValues);
+    encoder.encode_op (OpCode_blendcs);
+  }
+
+  static void flush_op (op_code_t op, cff2_cs_interp_env_t &env, flatten_param_t& param)
+  {
+    switch (op)
+    {
+      case OpCode_return:
+      case OpCode_endchar:
+	return;
+      default:
+	str_encoder_t  encoder (param.flatStr);
+	encoder.encode_op (op);
+    }
+  }
+
+  private:
+  typedef cff2_cs_opset_t<cff2_cs_opset_flatten_t, flatten_param_t> SUPER;
+  typedef cs_opset_t<blend_arg_t, cff2_cs_opset_flatten_t, cff2_cs_opset_flatten_t, cff2_cs_interp_env_t, flatten_param_t> CSOPSET;
+};
+
+struct cff2_cs_opset_subr_subset_t : cff2_cs_opset_t<cff2_cs_opset_subr_subset_t, subr_subset_param_t>
+{
+  static void process_op (op_code_t op, cff2_cs_interp_env_t &env, subr_subset_param_t& param)
+  {
+    switch (op) {
+
+      case OpCode_return:
+	param.current_parsed_str->set_parsed ();
+	env.return_from_subr ();
+	param.set_current_str (env, false);
+	break;
+
+      case OpCode_endchar:
+	param.current_parsed_str->set_parsed ();
+	SUPER::process_op (op, env, param);
+	break;
+
+      case OpCode_callsubr:
+	process_call_subr (op, CSType_LocalSubr, env, param, env.localSubrs, param.local_closure);
+	break;
+
+      case OpCode_callgsubr:
+	process_call_subr (op, CSType_GlobalSubr, env, param, env.globalSubrs, param.global_closure);
+	break;
+
+      default:
+	SUPER::process_op (op, env, param);
+	param.current_parsed_str->add_op (op, env.str_ref);
+	break;
+    }
+  }
+
+  protected:
+  static void process_call_subr (op_code_t op, cs_type_t type,
+				 cff2_cs_interp_env_t &env, subr_subset_param_t& param,
+				 cff2_biased_subrs_t& subrs, hb_set_t *closure)
+  {
+    byte_str_ref_t    str_ref = env.str_ref;
+    env.call_subr (subrs, type);
+    param.current_parsed_str->add_call_op (op, str_ref, env.context.subr_num);
+    closure->add (env.context.subr_num);
+    param.set_current_str (env, true);
+  }
+
+  private:
+  typedef cff2_cs_opset_t<cff2_cs_opset_subr_subset_t, subr_subset_param_t> SUPER;
+};
+
+struct cff2_subr_subsetter_t : subr_subsetter_t<cff2_subr_subsetter_t, CFF2Subrs, const OT::cff2::accelerator_subset_t, cff2_cs_interp_env_t, cff2_cs_opset_subr_subset_t>
+{
+  cff2_subr_subsetter_t (const OT::cff2::accelerator_subset_t &acc_, const hb_subset_plan_t *plan_)
+    : subr_subsetter_t (acc_, plan_) {}
+
+  static void complete_parsed_str (cff2_cs_interp_env_t &env, subr_subset_param_t& param, parsed_cs_str_t &charstring)
+  {
+    /* vsindex is inserted at the beginning of the charstring as necessary */
+    if (env.seen_vsindex ())
+    {
+      number_t  ivs;
+      ivs.set_int ((int)env.get_ivs ());
+      charstring.set_prefix (ivs, OpCode_vsindexcs);
+    }
+  }
+};
+
+struct cff2_subset_plan {
+  cff2_subset_plan ()
+    : orig_fdcount (0),
+      subset_fdcount(1),
+      subset_fdselect_size (0),
+      subset_fdselect_format (0),
+      drop_hints (false),
+      desubroutinize (false)
+  {
+    subset_fdselect_ranges.init ();
+    fdmap.init ();
+    subset_charstrings.init ();
+    subset_globalsubrs.init ();
+    subset_localsubrs.init ();
+  }
+
+  ~cff2_subset_plan ()
+  {
+    subset_fdselect_ranges.fini ();
+    fdmap.fini ();
+    subset_charstrings.fini_deep ();
+    subset_globalsubrs.fini_deep ();
+    subset_localsubrs.fini_deep ();
+  }
+
+  bool create (const OT::cff2::accelerator_subset_t &acc,
+	      hb_subset_plan_t *plan)
+  {
+    orig_fdcount = acc.fdArray->count;
+
+    drop_hints = plan->drop_hints;
+    desubroutinize = plan->desubroutinize;
+
+    if (desubroutinize)
+    {
+      /* Flatten global & local subrs */
+      subr_flattener_t<const OT::cff2::accelerator_subset_t, cff2_cs_interp_env_t, cff2_cs_opset_flatten_t>
+		    flattener(acc, plan);
+      if (!flattener.flatten (subset_charstrings))
+	return false;
+    }
+    else
+    {
+      cff2_subr_subsetter_t	subr_subsetter (acc, plan);
+
+      /* Subset subrs: collect used subroutines, leaving all unused ones behind */
+      if (!subr_subsetter.subset ())
+	return false;
+
+      /* encode charstrings, global subrs, local subrs with new subroutine numbers */
+      if (!subr_subsetter.encode_charstrings (subset_charstrings))
+	return false;
+
+      if (!subr_subsetter.encode_globalsubrs (subset_globalsubrs))
+	return false;
+
+      /* local subrs */
+      if (!subset_localsubrs.resize (orig_fdcount))
+	return false;
+      for (unsigned int fd = 0; fd < orig_fdcount; fd++)
+      {
+	subset_localsubrs[fd].init ();
+	if (!subr_subsetter.encode_localsubrs (fd, subset_localsubrs[fd]))
+	  return false;
+      }
+    }
+
+    /* FDSelect */
+    if (acc.fdSelect != &Null (CFF2FDSelect))
+    {
+      if (unlikely (!hb_plan_subset_cff_fdselect (plan,
+						  orig_fdcount,
+						  *(const FDSelect *)acc.fdSelect,
+						  subset_fdcount,
+						  subset_fdselect_size,
+						  subset_fdselect_format,
+						  subset_fdselect_ranges,
+						  fdmap)))
+	return false;
+    }
+    else
+      fdmap.identity (1);
+
+    return true;
+  }
+
+  cff2_sub_table_info_t info;
+
+  unsigned int    orig_fdcount;
+  unsigned int    subset_fdcount;
+  unsigned int	  subset_fdselect_size;
+  unsigned int    subset_fdselect_format;
+  hb_vector_t<code_pair_t>   subset_fdselect_ranges;
+
+  hb_inc_bimap_t   fdmap;
+
+  str_buff_vec_t	    subset_charstrings;
+  str_buff_vec_t	    subset_globalsubrs;
+  hb_vector_t<str_buff_vec_t> subset_localsubrs;
+
+  bool	    drop_hints;
+  bool	    desubroutinize;
+};
+
+static bool _serialize_cff2 (hb_serialize_context_t *c,
+			     cff2_subset_plan &plan,
+			     const OT::cff2::accelerator_subset_t  &acc,
+			     unsigned int num_glyphs)
+{
+  /* private dicts & local subrs */
+  hb_vector_t<table_info_t>  private_dict_infos;
+  if (unlikely (!private_dict_infos.resize (plan.subset_fdcount))) return false;
+
+  for (int i = (int)acc.privateDicts.length; --i >= 0 ;)
+  {
+    if (plan.fdmap.has (i))
+    {
+      objidx_t	subrs_link = 0;
+
+      if (plan.subset_localsubrs[i].length > 0)
+      {
+	CFF2Subrs *dest = c->start_embed <CFF2Subrs> ();
+	if (unlikely (!dest)) return false;
+	c->push ();
+	if (likely (dest->serialize (c, plan.subset_localsubrs[i])))
+	  subrs_link = c->pop_pack ();
+	else
+	{
+	  c->pop_discard ();
+	  return false;
+	}
+      }
+      PrivateDict *pd = c->start_embed<PrivateDict> ();
+      if (unlikely (!pd)) return false;
+      c->push ();
+      cff_private_dict_op_serializer_t privSzr (plan.desubroutinize, plan.drop_hints);
+      if (likely (pd->serialize (c, acc.privateDicts[i], privSzr, subrs_link)))
+      {
+	unsigned fd = plan.fdmap[i];
+	private_dict_infos[fd].size = c->length ();
+	private_dict_infos[fd].link = c->pop_pack ();
+      }
+      else
+      {
+	c->pop_discard ();
+	return false;
+      }
+    }
+  }
+
+  /* CharStrings */
+  {
+    CFF2CharStrings  *cs = c->start_embed<CFF2CharStrings> ();
+    if (unlikely (!cs)) return false;
+    c->push ();
+    if (likely (cs->serialize (c, plan.subset_charstrings)))
+      plan.info.char_strings_link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* FDSelect */
+  if (acc.fdSelect != &Null (CFF2FDSelect))
+  {
+    c->push ();
+    if (likely (hb_serialize_cff_fdselect (c, num_glyphs, *(const FDSelect *)acc.fdSelect, 					      plan.orig_fdcount,
+					    plan.subset_fdselect_format, plan.subset_fdselect_size,
+					    plan.subset_fdselect_ranges)))
+      plan.info.fd_select.link = c->pop_pack ();
+    else
+    {
+      c->pop_discard ();
+      return false;
+    }
+  }
+
+  /* FDArray (FD Index) */
+  {
+    c->push ();
+    CFF2FDArray *fda = c->start_embed<CFF2FDArray> ();
+    if (unlikely (!fda)) return false;
+    cff_font_dict_op_serializer_t fontSzr;
+    auto it =
+    + hb_zip (+ hb_iter (acc.fontDicts)
+	      | hb_filter ([&] (const cff2_font_dict_values_t &_)
+		{ return plan.fdmap.has (&_ - &acc.fontDicts[0]); }),
+	      hb_iter (private_dict_infos))
+    ;
+    if (unlikely (!fda->serialize (c, it, fontSzr))) return false;
+    plan.info.fd_array_link = c->pop_pack ();
+  }
+
+  /* variation store */
+  if (acc.varStore != &Null (CFF2VariationStore))
+  {
+    c->push ();
+    CFF2VariationStore *dest = c->start_embed<CFF2VariationStore> ();
+    if (unlikely (!dest || !dest->serialize (c, acc.varStore))) return false;
+    plan.info.var_store_link = c->pop_pack ();
+  }
+
+  OT::cff2 *cff2 = c->allocate_min<OT::cff2> ();
+  if (unlikely (!cff2)) return false;
+
+  /* header */
+  cff2->version.major = 0x02;
+  cff2->version.minor = 0x00;
+  cff2->topDict = OT::cff2::static_size;
+
+  /* top dict */
+  {
+    TopDict &dict = cff2 + cff2->topDict;
+    cff2_top_dict_op_serializer_t topSzr;
+    if (unlikely (!dict.serialize (c, acc.topDict, topSzr, plan.info))) return false;
+    cff2->topDictSize = c->head - (const char *)&dict;
+  }
+
+  /* global subrs */
+  {
+    CFF2Subrs *dest = c->start_embed <CFF2Subrs> ();
+    if (unlikely (!dest)) return false;
+    return dest->serialize (c, plan.subset_globalsubrs);
+  }
+}
+
+static bool
+_hb_subset_cff2 (const OT::cff2::accelerator_subset_t  &acc,
+		 hb_subset_context_t	*c)
+{
+  cff2_subset_plan cff2_plan;
+
+  if (unlikely (!cff2_plan.create (acc, c->plan))) return false;
+  return _serialize_cff2 (c->serializer, cff2_plan, acc, c->plan->num_output_glyphs ());
+}
+
+/**
+ * hb_subset_cff2:
+ * Subsets the CFF2 table according to a provided subset context.
+ **/
+bool
+hb_subset_cff2 (hb_subset_context_t *c)
+{
+  OT::cff2::accelerator_subset_t acc;
+  acc.init (c->plan->source);
+  bool result = likely (acc.is_valid ()) && _hb_subset_cff2 (acc, c);
+  acc.fini ();
+
+  return result;
+}
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-subset-cff2.hh b/thirdparty/harfbuzz/src/hb-subset-cff2.hh
new file mode 100644
index 0000000000..f10556ddd7
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-cff2.hh
@@ -0,0 +1,37 @@
+/*
+ * Copyright © 2018 Adobe Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Adobe Author(s): Michiharu Ariza
+ */
+
+#ifndef HB_SUBSET_CFF2_HH
+#define HB_SUBSET_CFF2_HH
+
+#include "hb.hh"
+
+#include "hb-subset-plan.hh"
+
+HB_INTERNAL bool
+hb_subset_cff2 (hb_subset_context_t *c);
+
+#endif /* HB_SUBSET_CFF2_HH */
diff --git a/thirdparty/harfbuzz/src/hb-subset-input.cc b/thirdparty/harfbuzz/src/hb-subset-input.cc
new file mode 100644
index 0000000000..fe9be3ce02
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-input.cc
@@ -0,0 +1,229 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Rod Sheeter, Behdad Esfahbod
+ */
+
+#include "hb-subset.hh"
+#include "hb-set.hh"
+
+/**
+ * hb_subset_input_create_or_fail:
+ *
+ * Return value: New subset input.
+ *
+ * Since: 1.8.0
+ **/
+hb_subset_input_t *
+hb_subset_input_create_or_fail ()
+{
+  hb_subset_input_t *input = hb_object_create<hb_subset_input_t>();
+
+  if (unlikely (!input))
+    return nullptr;
+
+  input->unicodes = hb_set_create ();
+  input->glyphs = hb_set_create ();
+  input->name_ids = hb_set_create ();
+  hb_set_add_range (input->name_ids, 0, 6);
+  input->name_languages = hb_set_create ();
+  hb_set_add (input->name_languages, 0x0409);
+  input->drop_tables = hb_set_create ();
+  input->drop_hints = false;
+  input->desubroutinize = false;
+  input->retain_gids = false;
+  input->name_legacy = false;
+
+  hb_tag_t default_drop_tables[] = {
+    // Layout disabled by default
+    HB_TAG ('G', 'S', 'U', 'B'),
+    HB_TAG ('G', 'P', 'O', 'S'),
+    HB_TAG ('G', 'D', 'E', 'F'),
+    HB_TAG ('m', 'o', 'r', 'x'),
+    HB_TAG ('m', 'o', 'r', 't'),
+    HB_TAG ('k', 'e', 'r', 'x'),
+    HB_TAG ('k', 'e', 'r', 'n'),
+
+    // Copied from fontTools:
+    HB_TAG ('B', 'A', 'S', 'E'),
+    HB_TAG ('J', 'S', 'T', 'F'),
+    HB_TAG ('D', 'S', 'I', 'G'),
+    HB_TAG ('E', 'B', 'D', 'T'),
+    HB_TAG ('E', 'B', 'L', 'C'),
+    HB_TAG ('E', 'B', 'S', 'C'),
+    HB_TAG ('S', 'V', 'G', ' '),
+    HB_TAG ('P', 'C', 'L', 'T'),
+    HB_TAG ('L', 'T', 'S', 'H'),
+    // Graphite tables
+    HB_TAG ('F', 'e', 'a', 't'),
+    HB_TAG ('G', 'l', 'a', 't'),
+    HB_TAG ('G', 'l', 'o', 'c'),
+    HB_TAG ('S', 'i', 'l', 'f'),
+    HB_TAG ('S', 'i', 'l', 'l'),
+  };
+
+  input->drop_tables->add_array (default_drop_tables, ARRAY_LENGTH (default_drop_tables));
+
+  return input;
+}
+
+/**
+ * hb_subset_input_reference: (skip)
+ * @subset_input: a subset_input.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 1.8.0
+ **/
+hb_subset_input_t *
+hb_subset_input_reference (hb_subset_input_t *subset_input)
+{
+  return hb_object_reference (subset_input);
+}
+
+/**
+ * hb_subset_input_destroy:
+ * @subset_input: a subset_input.
+ *
+ * Since: 1.8.0
+ **/
+void
+hb_subset_input_destroy (hb_subset_input_t *subset_input)
+{
+  if (!hb_object_destroy (subset_input)) return;
+
+  hb_set_destroy (subset_input->unicodes);
+  hb_set_destroy (subset_input->glyphs);
+  hb_set_destroy (subset_input->name_ids);
+  hb_set_destroy (subset_input->name_languages);
+  hb_set_destroy (subset_input->drop_tables);
+
+  free (subset_input);
+}
+
+/**
+ * hb_subset_input_unicode_set:
+ * @subset_input: a subset_input.
+ *
+ * Since: 1.8.0
+ **/
+HB_EXTERN hb_set_t *
+hb_subset_input_unicode_set (hb_subset_input_t *subset_input)
+{
+  return subset_input->unicodes;
+}
+
+/**
+ * hb_subset_input_glyph_set:
+ * @subset_input: a subset_input.
+ *
+ * Since: 1.8.0
+ **/
+HB_EXTERN hb_set_t *
+hb_subset_input_glyph_set (hb_subset_input_t *subset_input)
+{
+  return subset_input->glyphs;
+}
+
+HB_EXTERN hb_set_t *
+hb_subset_input_nameid_set (hb_subset_input_t *subset_input)
+{
+  return subset_input->name_ids;
+}
+
+HB_EXTERN hb_set_t *
+hb_subset_input_namelangid_set (hb_subset_input_t *subset_input)
+{
+  return subset_input->name_languages;
+}
+
+HB_EXTERN hb_set_t *
+hb_subset_input_drop_tables_set (hb_subset_input_t *subset_input)
+{
+  return subset_input->drop_tables;
+}
+
+HB_EXTERN void
+hb_subset_input_set_drop_hints (hb_subset_input_t *subset_input,
+				hb_bool_t drop_hints)
+{
+  subset_input->drop_hints = drop_hints;
+}
+
+HB_EXTERN hb_bool_t
+hb_subset_input_get_drop_hints (hb_subset_input_t *subset_input)
+{
+  return subset_input->drop_hints;
+}
+
+HB_EXTERN void
+hb_subset_input_set_desubroutinize (hb_subset_input_t *subset_input,
+				    hb_bool_t desubroutinize)
+{
+  subset_input->desubroutinize = desubroutinize;
+}
+
+HB_EXTERN hb_bool_t
+hb_subset_input_get_desubroutinize (hb_subset_input_t *subset_input)
+{
+  return subset_input->desubroutinize;
+}
+
+/**
+ * hb_subset_input_set_retain_gids:
+ * @subset_input: a subset_input.
+ * @retain_gids: If true the subsetter will not renumber glyph ids.
+ * Since: 2.4.0
+ **/
+HB_EXTERN void
+hb_subset_input_set_retain_gids (hb_subset_input_t *subset_input,
+				 hb_bool_t retain_gids)
+{
+  subset_input->retain_gids = retain_gids;
+}
+
+/**
+ * hb_subset_input_get_retain_gids:
+ * Returns: value of retain_gids.
+ * Since: 2.4.0
+ **/
+HB_EXTERN hb_bool_t
+hb_subset_input_get_retain_gids (hb_subset_input_t *subset_input)
+{
+  return subset_input->retain_gids;
+}
+
+HB_EXTERN void
+hb_subset_input_set_name_legacy (hb_subset_input_t *subset_input,
+				 hb_bool_t name_legacy)
+{
+  subset_input->name_legacy = name_legacy;
+}
+
+HB_EXTERN hb_bool_t
+hb_subset_input_get_name_legacy (hb_subset_input_t *subset_input)
+{
+  return subset_input->name_legacy;
+}
diff --git a/thirdparty/harfbuzz/src/hb-subset-input.hh b/thirdparty/harfbuzz/src/hb-subset-input.hh
new file mode 100644
index 0000000000..0aeb96695b
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-input.hh
@@ -0,0 +1,61 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Roderick Sheeter
+ */
+
+#ifndef HB_SUBSET_INPUT_HH
+#define HB_SUBSET_INPUT_HH
+
+
+#include "hb.hh"
+
+#include "hb-subset.h"
+
+#include "hb-font.hh"
+
+struct hb_subset_input_t
+{
+  hb_object_header_t header;
+
+  hb_set_t *unicodes;
+  hb_set_t *glyphs;
+  hb_set_t *name_ids;
+  hb_set_t *name_languages;
+  hb_set_t *drop_tables;
+
+  bool drop_hints;
+  bool desubroutinize;
+  bool retain_gids;
+  bool name_legacy;
+  /* TODO
+   *
+   * features
+   * lookups
+   * name_ids
+   * ...
+   */
+};
+
+
+#endif /* HB_SUBSET_INPUT_HH */
diff --git a/thirdparty/harfbuzz/src/hb-subset-plan.cc b/thirdparty/harfbuzz/src/hb-subset-plan.cc
new file mode 100644
index 0000000000..24beada3e8
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-plan.cc
@@ -0,0 +1,395 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Roderick Sheeter
+ */
+
+#include "hb-subset-plan.hh"
+#include "hb-map.hh"
+#include "hb-set.hh"
+
+#include "hb-ot-cmap-table.hh"
+#include "hb-ot-glyf-table.hh"
+#include "hb-ot-layout-gdef-table.hh"
+#include "hb-ot-layout-gpos-table.hh"
+#include "hb-ot-layout-gsub-table.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-ot-color-colr-table.hh"
+#include "hb-ot-var-fvar-table.hh"
+#include "hb-ot-stat-table.hh"
+
+
+#ifndef HB_NO_SUBSET_CFF
+static inline void
+_add_cff_seac_components (const OT::cff1::accelerator_t &cff,
+			  hb_codepoint_t gid,
+			  hb_set_t *gids_to_retain)
+{
+  hb_codepoint_t base_gid, accent_gid;
+  if (cff.get_seac_components (gid, &base_gid, &accent_gid))
+  {
+    gids_to_retain->add (base_gid);
+    gids_to_retain->add (accent_gid);
+  }
+}
+#endif
+
+#ifndef HB_NO_SUBSET_LAYOUT
+static void
+_remap_indexes (const hb_set_t *indexes,
+		hb_map_t       *mapping /* OUT */)
+{
+  unsigned count = indexes->get_population ();
+
+  for (auto _ : + hb_zip (indexes->iter (), hb_range (count)))
+    mapping->set (_.first, _.second);
+
+}
+
+static inline void
+_gsub_closure_glyphs_lookups_features (hb_face_t *face,
+				       hb_set_t *gids_to_retain,
+				       hb_map_t *gsub_lookups,
+				       hb_map_t *gsub_features)
+{
+  hb_set_t lookup_indices;
+  hb_ot_layout_collect_lookups (face,
+				HB_OT_TAG_GSUB,
+				nullptr,
+				nullptr,
+				nullptr,
+				&lookup_indices);
+  hb_ot_layout_lookups_substitute_closure (face,
+					   &lookup_indices,
+					   gids_to_retain);
+  hb_blob_ptr_t<OT::GSUB> gsub = hb_sanitize_context_t ().reference_table<OT::GSUB> (face);
+  gsub->closure_lookups (face,
+			 gids_to_retain,
+			 &lookup_indices);
+  _remap_indexes (&lookup_indices, gsub_lookups);
+
+  //closure features
+  hb_set_t feature_indices;
+  gsub->closure_features (gsub_lookups, &feature_indices);
+  _remap_indexes (&feature_indices, gsub_features);
+  gsub.destroy ();
+}
+
+static inline void
+_gpos_closure_lookups_features (hb_face_t      *face,
+				const hb_set_t *gids_to_retain,
+				hb_map_t       *gpos_lookups,
+				hb_map_t       *gpos_features)
+{
+  hb_set_t lookup_indices;
+  hb_ot_layout_collect_lookups (face,
+				HB_OT_TAG_GPOS,
+				nullptr,
+				nullptr,
+				nullptr,
+				&lookup_indices);
+  hb_blob_ptr_t<OT::GPOS> gpos = hb_sanitize_context_t ().reference_table<OT::GPOS> (face);
+  gpos->closure_lookups (face,
+			 gids_to_retain,
+			 &lookup_indices);
+  _remap_indexes (&lookup_indices, gpos_lookups);
+
+  //closure features
+  hb_set_t feature_indices;
+  gpos->closure_features (gpos_lookups, &feature_indices);
+  _remap_indexes (&feature_indices, gpos_features);
+  gpos.destroy ();
+}
+#endif
+
+#ifndef HB_NO_VAR
+static inline void
+  _collect_layout_variation_indices (hb_face_t *face,
+				     const hb_set_t *glyphset,
+				     const hb_map_t *gpos_lookups,
+				     hb_set_t  *layout_variation_indices,
+				     hb_map_t  *layout_variation_idx_map)
+{
+  hb_blob_ptr_t<OT::GDEF> gdef = hb_sanitize_context_t ().reference_table<OT::GDEF> (face);
+  hb_blob_ptr_t<OT::GPOS> gpos = hb_sanitize_context_t ().reference_table<OT::GPOS> (face);
+
+  if (!gdef->has_data ())
+  {
+    gdef.destroy ();
+    gpos.destroy ();
+    return;
+  }
+  OT::hb_collect_variation_indices_context_t c (layout_variation_indices, glyphset, gpos_lookups);
+  gdef->collect_variation_indices (&c);
+
+  if (hb_ot_layout_has_positioning (face))
+    gpos->collect_variation_indices (&c);
+
+  gdef->remap_layout_variation_indices (layout_variation_indices, layout_variation_idx_map);
+
+  gdef.destroy ();
+  gpos.destroy ();
+}
+#endif
+
+static inline void
+_cmap_closure (hb_face_t           *face,
+	       const hb_set_t      *unicodes,
+	       hb_set_t            *glyphset)
+{
+  OT::cmap::accelerator_t cmap;
+  cmap.init (face);
+  cmap.table->closure_glyphs (unicodes, glyphset);
+  cmap.fini ();
+}
+
+static inline void
+_remove_invalid_gids (hb_set_t *glyphs,
+		      unsigned int num_glyphs)
+{
+  hb_codepoint_t gid = HB_SET_VALUE_INVALID;
+  while (glyphs->next (&gid))
+  {
+    if (gid >= num_glyphs)
+      glyphs->del (gid);
+  }
+}
+
+static void
+_populate_gids_to_retain (hb_subset_plan_t* plan,
+			  const hb_set_t *unicodes,
+			  const hb_set_t *input_glyphs_to_retain,
+			  bool close_over_gsub,
+			  bool close_over_gpos,
+			  bool close_over_gdef)
+{
+  OT::cmap::accelerator_t cmap;
+  OT::glyf::accelerator_t glyf;
+#ifndef HB_NO_SUBSET_CFF
+  OT::cff1::accelerator_t cff;
+#endif
+  OT::COLR::accelerator_t colr;
+  cmap.init (plan->source);
+  glyf.init (plan->source);
+#ifndef HB_NO_SUBSET_CFF
+  cff.init (plan->source);
+#endif
+  colr.init (plan->source);
+
+  plan->_glyphset_gsub->add (0); // Not-def
+  hb_set_union (plan->_glyphset_gsub, input_glyphs_to_retain);
+
+  hb_codepoint_t cp = HB_SET_VALUE_INVALID;
+  while (unicodes->next (&cp))
+  {
+    hb_codepoint_t gid;
+    if (!cmap.get_nominal_glyph (cp, &gid))
+    {
+      DEBUG_MSG(SUBSET, nullptr, "Drop U+%04X; no gid", cp);
+      continue;
+    }
+    plan->unicodes->add (cp);
+    plan->codepoint_to_glyph->set (cp, gid);
+    plan->_glyphset_gsub->add (gid);
+  }
+
+  _cmap_closure (plan->source, plan->unicodes, plan->_glyphset_gsub);
+
+#ifndef HB_NO_SUBSET_LAYOUT
+  if (close_over_gsub)
+    // closure all glyphs/lookups/features needed for GSUB substitutions.
+    _gsub_closure_glyphs_lookups_features (plan->source, plan->_glyphset_gsub, plan->gsub_lookups, plan->gsub_features);
+
+  if (close_over_gpos)
+    _gpos_closure_lookups_features (plan->source, plan->_glyphset_gsub, plan->gpos_lookups, plan->gpos_features);
+#endif
+  _remove_invalid_gids (plan->_glyphset_gsub, plan->source->get_num_glyphs ());
+
+  // Populate a full set of glyphs to retain by adding all referenced
+  // composite glyphs.
+  hb_codepoint_t gid = HB_SET_VALUE_INVALID;
+  while (plan->_glyphset_gsub->next (&gid))
+  {
+    glyf.add_gid_and_children (gid, plan->_glyphset);
+#ifndef HB_NO_SUBSET_CFF
+    if (cff.is_valid ())
+      _add_cff_seac_components (cff, gid, plan->_glyphset);
+#endif
+    if (colr.is_valid ())
+      colr.closure_glyphs (gid, plan->_glyphset);
+  }
+
+  _remove_invalid_gids (plan->_glyphset, plan->source->get_num_glyphs ());
+
+#ifndef HB_NO_VAR
+  if (close_over_gdef)
+    _collect_layout_variation_indices (plan->source, plan->_glyphset, plan->gpos_lookups, plan->layout_variation_indices, plan->layout_variation_idx_map);
+#endif
+
+#ifndef HB_NO_SUBSET_CFF
+  cff.fini ();
+#endif
+  glyf.fini ();
+  cmap.fini ();
+}
+
+static void
+_create_old_gid_to_new_gid_map (const hb_face_t *face,
+				bool             retain_gids,
+				const hb_set_t  *all_gids_to_retain,
+				hb_map_t        *glyph_map, /* OUT */
+				hb_map_t        *reverse_glyph_map, /* OUT */
+				unsigned int    *num_glyphs /* OUT */)
+{
+  if (!retain_gids)
+  {
+    + hb_enumerate (hb_iter (all_gids_to_retain), (hb_codepoint_t) 0)
+    | hb_sink (reverse_glyph_map)
+    ;
+    *num_glyphs = reverse_glyph_map->get_population ();
+  } else {
+    + hb_iter (all_gids_to_retain)
+    | hb_map ([] (hb_codepoint_t _) {
+		return hb_pair_t<hb_codepoint_t, hb_codepoint_t> (_, _);
+	      })
+    | hb_sink (reverse_glyph_map)
+    ;
+
+    unsigned max_glyph =
+    + hb_iter (all_gids_to_retain)
+    | hb_reduce (hb_max, 0u)
+    ;
+    *num_glyphs = max_glyph + 1;
+  }
+
+  + reverse_glyph_map->iter ()
+  | hb_map (&hb_pair_t<hb_codepoint_t, hb_codepoint_t>::reverse)
+  | hb_sink (glyph_map)
+  ;
+}
+
+static void
+_nameid_closure (hb_face_t *face,
+		 hb_set_t  *nameids)
+{
+#ifndef HB_NO_STYLE
+  face->table.STAT->collect_name_ids (nameids);
+#endif
+#ifndef HB_NO_VAR
+  face->table.fvar->collect_name_ids (nameids);
+#endif
+}
+
+/**
+ * hb_subset_plan_create:
+ * Computes a plan for subsetting the supplied face according
+ * to a provided input. The plan describes
+ * which tables and glyphs should be retained.
+ *
+ * Return value: New subset plan.
+ *
+ * Since: 1.7.5
+ **/
+hb_subset_plan_t *
+hb_subset_plan_create (hb_face_t         *face,
+		       hb_subset_input_t *input)
+{
+  hb_subset_plan_t *plan;
+  if (unlikely (!(plan = hb_object_create<hb_subset_plan_t> ())))
+    return const_cast<hb_subset_plan_t *> (&Null (hb_subset_plan_t));
+
+  plan->successful = true;
+  plan->drop_hints = input->drop_hints;
+  plan->desubroutinize = input->desubroutinize;
+  plan->retain_gids = input->retain_gids;
+  plan->name_legacy = input->name_legacy;
+  plan->unicodes = hb_set_create ();
+  plan->name_ids = hb_set_reference (input->name_ids);
+  _nameid_closure (face, plan->name_ids);
+  plan->name_languages = hb_set_reference (input->name_languages);
+  plan->glyphs_requested = hb_set_reference (input->glyphs);
+  plan->drop_tables = hb_set_reference (input->drop_tables);
+  plan->source = hb_face_reference (face);
+  plan->dest = hb_face_builder_create ();
+
+  plan->_glyphset = hb_set_create ();
+  plan->_glyphset_gsub = hb_set_create ();
+  plan->codepoint_to_glyph = hb_map_create ();
+  plan->glyph_map = hb_map_create ();
+  plan->reverse_glyph_map = hb_map_create ();
+  plan->gsub_lookups = hb_map_create ();
+  plan->gpos_lookups = hb_map_create ();
+  plan->gsub_features = hb_map_create ();
+  plan->gpos_features = hb_map_create ();
+  plan->layout_variation_indices = hb_set_create ();
+  plan->layout_variation_idx_map = hb_map_create ();
+
+  _populate_gids_to_retain (plan,
+			    input->unicodes,
+			    input->glyphs,
+			    !input->drop_tables->has (HB_OT_TAG_GSUB),
+			    !input->drop_tables->has (HB_OT_TAG_GPOS),
+			    !input->drop_tables->has (HB_OT_TAG_GDEF));
+
+  _create_old_gid_to_new_gid_map (face,
+				  input->retain_gids,
+				  plan->_glyphset,
+				  plan->glyph_map,
+				  plan->reverse_glyph_map,
+				  &plan->_num_output_glyphs);
+
+  return plan;
+}
+
+/**
+ * hb_subset_plan_destroy:
+ *
+ * Since: 1.7.5
+ **/
+void
+hb_subset_plan_destroy (hb_subset_plan_t *plan)
+{
+  if (!hb_object_destroy (plan)) return;
+
+  hb_set_destroy (plan->unicodes);
+  hb_set_destroy (plan->name_ids);
+  hb_set_destroy (plan->name_languages);
+  hb_set_destroy (plan->glyphs_requested);
+  hb_set_destroy (plan->drop_tables);
+  hb_face_destroy (plan->source);
+  hb_face_destroy (plan->dest);
+  hb_map_destroy (plan->codepoint_to_glyph);
+  hb_map_destroy (plan->glyph_map);
+  hb_map_destroy (plan->reverse_glyph_map);
+  hb_set_destroy (plan->_glyphset);
+  hb_set_destroy (plan->_glyphset_gsub);
+  hb_map_destroy (plan->gsub_lookups);
+  hb_map_destroy (plan->gpos_lookups);
+  hb_map_destroy (plan->gsub_features);
+  hb_map_destroy (plan->gpos_features);
+  hb_set_destroy (plan->layout_variation_indices);
+  hb_map_destroy (plan->layout_variation_idx_map);
+
+
+  free (plan);
+}
diff --git a/thirdparty/harfbuzz/src/hb-subset-plan.hh b/thirdparty/harfbuzz/src/hb-subset-plan.hh
new file mode 100644
index 0000000000..e9f603dd1d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset-plan.hh
@@ -0,0 +1,194 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Roderick Sheeter
+ */
+
+#ifndef HB_SUBSET_PLAN_HH
+#define HB_SUBSET_PLAN_HH
+
+#include "hb.hh"
+
+#include "hb-subset.h"
+#include "hb-subset-input.hh"
+
+#include "hb-map.hh"
+#include "hb-set.hh"
+
+struct hb_subset_plan_t
+{
+  hb_object_header_t header;
+
+  bool successful : 1;
+  bool drop_hints : 1;
+  bool desubroutinize : 1;
+  bool retain_gids : 1;
+  bool name_legacy : 1;
+
+  // For each cp that we'd like to retain maps to the corresponding gid.
+  hb_set_t *unicodes;
+
+  // name_ids we would like to retain
+  hb_set_t *name_ids;
+
+  // name_languages we would like to retain
+  hb_set_t *name_languages;
+
+  //glyph ids requested to retain
+  hb_set_t *glyphs_requested;
+
+  // Tables which should be dropped.
+  hb_set_t *drop_tables;
+
+  // The glyph subset
+  hb_map_t *codepoint_to_glyph;
+
+  // Old -> New glyph id mapping
+  hb_map_t *glyph_map;
+  hb_map_t *reverse_glyph_map;
+
+  // Plan is only good for a specific source/dest so keep them with it
+  hb_face_t *source;
+  hb_face_t *dest;
+
+  unsigned int _num_output_glyphs;
+  hb_set_t *_glyphset;
+  hb_set_t *_glyphset_gsub;
+
+  //active lookups we'd like to retain
+  hb_map_t *gsub_lookups;
+  hb_map_t *gpos_lookups;
+
+  //active features we'd like to retain
+  hb_map_t *gsub_features;
+  hb_map_t *gpos_features;
+
+  //The set of layout item variation store delta set indices to be retained
+  hb_set_t *layout_variation_indices;
+  //Old -> New layout item variation store delta set index mapping
+  hb_map_t *layout_variation_idx_map;
+
+ public:
+
+  bool in_error () const { return !successful; }
+
+  bool check_success(bool success)
+  {
+    successful = (successful && success);
+    return successful;
+  }
+
+  /*
+   * The set of input glyph ids which will be retained in the subset.
+   * Does NOT include ids kept due to retain_gids. You probably want to use
+   * glyph_map/reverse_glyph_map.
+   */
+  inline const hb_set_t *
+  glyphset () const
+  {
+    return _glyphset;
+  }
+
+  /*
+   * The set of input glyph ids which will be retained in the subset.
+   */
+  inline const hb_set_t *
+  glyphset_gsub () const
+  {
+    return _glyphset_gsub;
+  }
+
+  /*
+   * The total number of output glyphs in the final subset.
+   */
+  inline unsigned int
+  num_output_glyphs () const
+  {
+    return _num_output_glyphs;
+  }
+
+  /*
+   * Given an output gid , returns true if that glyph id is an empty
+   * glyph (ie. it's a gid that we are dropping all data for).
+   */
+  inline bool is_empty_glyph (hb_codepoint_t gid) const
+  {
+    return !_glyphset->has (gid);
+  }
+
+  inline bool new_gid_for_codepoint (hb_codepoint_t codepoint,
+				     hb_codepoint_t *new_gid) const
+  {
+    hb_codepoint_t old_gid = codepoint_to_glyph->get (codepoint);
+    if (old_gid == HB_MAP_VALUE_INVALID)
+      return false;
+
+    return new_gid_for_old_gid (old_gid, new_gid);
+  }
+
+  inline bool new_gid_for_old_gid (hb_codepoint_t old_gid,
+				   hb_codepoint_t *new_gid) const
+  {
+    hb_codepoint_t gid = glyph_map->get (old_gid);
+    if (gid == HB_MAP_VALUE_INVALID)
+      return false;
+
+    *new_gid = gid;
+    return true;
+  }
+
+  inline bool old_gid_for_new_gid (hb_codepoint_t  new_gid,
+				   hb_codepoint_t *old_gid) const
+  {
+    hb_codepoint_t gid = reverse_glyph_map->get (new_gid);
+    if (gid == HB_MAP_VALUE_INVALID)
+      return false;
+
+    *old_gid = gid;
+    return true;
+  }
+
+  inline bool
+  add_table (hb_tag_t tag,
+	     hb_blob_t *contents)
+  {
+    hb_blob_t *source_blob = source->reference_table (tag);
+    DEBUG_MSG(SUBSET, nullptr, "add table %c%c%c%c, dest %d bytes, source %d bytes",
+	      HB_UNTAG(tag),
+	      hb_blob_get_length (contents),
+	      hb_blob_get_length (source_blob));
+    hb_blob_destroy (source_blob);
+    return hb_face_builder_add_table (dest, tag, contents);
+  }
+};
+
+typedef struct hb_subset_plan_t hb_subset_plan_t;
+
+HB_INTERNAL hb_subset_plan_t *
+hb_subset_plan_create (hb_face_t           *face,
+		       hb_subset_input_t   *input);
+
+HB_INTERNAL void
+hb_subset_plan_destroy (hb_subset_plan_t *plan);
+
+#endif /* HB_SUBSET_PLAN_HH */
diff --git a/thirdparty/harfbuzz/src/hb-subset.cc b/thirdparty/harfbuzz/src/hb-subset.cc
new file mode 100644
index 0000000000..8b77ecd45a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset.cc
@@ -0,0 +1,269 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Rod Sheeter, Behdad Esfahbod
+ */
+
+#include "hb.hh"
+#include "hb-open-type.hh"
+
+#include "hb-subset.hh"
+
+#include "hb-open-file.hh"
+#include "hb-ot-cmap-table.hh"
+#include "hb-ot-glyf-table.hh"
+#include "hb-ot-hdmx-table.hh"
+#include "hb-ot-head-table.hh"
+#include "hb-ot-hhea-table.hh"
+#include "hb-ot-hmtx-table.hh"
+#include "hb-ot-maxp-table.hh"
+#include "hb-ot-color-sbix-table.hh"
+#include "hb-ot-color-colr-table.hh"
+#include "hb-ot-os2-table.hh"
+#include "hb-ot-post-table.hh"
+#include "hb-ot-cff1-table.hh"
+#include "hb-ot-cff2-table.hh"
+#include "hb-ot-vorg-table.hh"
+#include "hb-ot-name-table.hh"
+#include "hb-ot-color-cbdt-table.hh"
+#include "hb-ot-layout-gsub-table.hh"
+#include "hb-ot-layout-gpos-table.hh"
+#include "hb-ot-var-gvar-table.hh"
+#include "hb-ot-var-hvar-table.hh"
+
+
+static unsigned
+_plan_estimate_subset_table_size (hb_subset_plan_t *plan, unsigned table_len)
+{
+  unsigned src_glyphs = plan->source->get_num_glyphs ();
+  unsigned dst_glyphs = plan->glyphset ()->get_population ();
+
+  if (unlikely (!src_glyphs))
+    return 512 + table_len;
+
+  return 512 + (unsigned) (table_len * sqrt ((double) dst_glyphs / src_glyphs));
+}
+
+template<typename TableType>
+static bool
+_subset (hb_subset_plan_t *plan)
+{
+  bool result = false;
+  hb_blob_t *source_blob = hb_sanitize_context_t ().reference_table<TableType> (plan->source);
+  const TableType *table = source_blob->as<TableType> ();
+
+  hb_tag_t tag = TableType::tableTag;
+  if (source_blob->data)
+  {
+    hb_vector_t<char> buf;
+    /* TODO Not all tables are glyph-related.  'name' table size for example should not be
+     * affected by number of glyphs.  Accommodate that. */
+    unsigned buf_size = _plan_estimate_subset_table_size (plan, source_blob->length);
+    DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c initial estimated table size: %u bytes.", HB_UNTAG (tag), buf_size);
+    if (unlikely (!buf.alloc (buf_size)))
+    {
+      DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c failed to allocate %u bytes.", HB_UNTAG (tag), buf_size);
+      hb_blob_destroy (source_blob);
+      return false;
+    }
+  retry:
+    hb_serialize_context_t serializer ((void *) buf, buf_size);
+    serializer.start_serialize<TableType> ();
+    hb_subset_context_t c (source_blob, plan, &serializer, tag);
+    bool needed = table->subset (&c);
+    if (serializer.ran_out_of_room)
+    {
+      buf_size += (buf_size >> 1) + 32;
+      DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c ran out of room; reallocating to %u bytes.", HB_UNTAG (tag), buf_size);
+      if (unlikely (!buf.alloc (buf_size)))
+      {
+	DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c failed to reallocate %u bytes.", HB_UNTAG (tag), buf_size);
+	hb_blob_destroy (source_blob);
+	return false;
+      }
+      goto retry;
+    }
+    serializer.end_serialize ();
+
+    result = !serializer.in_error ();
+
+    if (result)
+    {
+      if (needed)
+      {
+	hb_blob_t *dest_blob = serializer.copy_blob ();
+	DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c final subset table size: %u bytes.", HB_UNTAG (tag), dest_blob->length);
+	result = c.plan->add_table (tag, dest_blob);
+	hb_blob_destroy (dest_blob);
+      }
+      else
+      {
+	DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c::subset table subsetted to empty.", HB_UNTAG (tag));
+      }
+    }
+  }
+  else
+    DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c::subset sanitize failed on source table.", HB_UNTAG (tag));
+
+  hb_blob_destroy (source_blob);
+  DEBUG_MSG (SUBSET, nullptr, "OT::%c%c%c%c::subset %s", HB_UNTAG (tag), result ? "success" : "FAILED!");
+  return result;
+}
+
+static bool
+_is_table_present (hb_face_t *source, hb_tag_t tag)
+{
+  hb_tag_t table_tags[32];
+  unsigned offset = 0, num_tables = ARRAY_LENGTH (table_tags);
+  while ((hb_face_get_table_tags (source, offset, &num_tables, table_tags), num_tables))
+  {
+    for (unsigned i = 0; i < num_tables; ++i)
+      if (table_tags[i] == tag)
+	return true;
+    offset += num_tables;
+  }
+  return false;
+}
+
+static bool
+_should_drop_table (hb_subset_plan_t *plan, hb_tag_t tag)
+{
+  if (plan->drop_tables->has (tag))
+    return true;
+
+  switch (tag)
+  {
+  case HB_TAG ('c','v','a','r'): /* hint table, fallthrough */
+  case HB_TAG ('c','v','t',' '): /* hint table, fallthrough */
+  case HB_TAG ('f','p','g','m'): /* hint table, fallthrough */
+  case HB_TAG ('p','r','e','p'): /* hint table, fallthrough */
+  case HB_TAG ('h','d','m','x'): /* hint table, fallthrough */
+  case HB_TAG ('V','D','M','X'): /* hint table, fallthrough */
+    return plan->drop_hints;
+
+#ifdef HB_NO_SUBSET_LAYOUT
+    // Drop Layout Tables if requested.
+  case HB_OT_TAG_GDEF:
+  case HB_OT_TAG_GPOS:
+  case HB_OT_TAG_GSUB:
+  case HB_TAG ('m','o','r','x'):
+  case HB_TAG ('m','o','r','t'):
+  case HB_TAG ('k','e','r','x'):
+  case HB_TAG ('k','e','r','n'):
+    return true;
+#endif
+
+  default:
+    return false;
+  }
+}
+
+static bool
+_subset_table (hb_subset_plan_t *plan, hb_tag_t tag)
+{
+  DEBUG_MSG (SUBSET, nullptr, "subset %c%c%c%c", HB_UNTAG (tag));
+  switch (tag)
+  {
+  case HB_OT_TAG_glyf: return _subset<const OT::glyf> (plan);
+  case HB_OT_TAG_hdmx: return _subset<const OT::hdmx> (plan);
+  case HB_OT_TAG_name: return _subset<const OT::name> (plan);
+  case HB_OT_TAG_head:
+    if (_is_table_present (plan->source, HB_OT_TAG_glyf) && !_should_drop_table (plan, HB_OT_TAG_glyf))
+      return true; /* skip head, handled by glyf */
+    return _subset<const OT::head> (plan);
+  case HB_OT_TAG_hhea: return true; /* skip hhea, handled by hmtx */
+  case HB_OT_TAG_hmtx: return _subset<const OT::hmtx> (plan);
+  case HB_OT_TAG_vhea: return true; /* skip vhea, handled by vmtx */
+  case HB_OT_TAG_vmtx: return _subset<const OT::vmtx> (plan);
+  case HB_OT_TAG_maxp: return _subset<const OT::maxp> (plan);
+  case HB_OT_TAG_sbix: return _subset<const OT::sbix> (plan);
+  case HB_OT_TAG_loca: return true; /* skip loca, handled by glyf */
+  case HB_OT_TAG_cmap: return _subset<const OT::cmap> (plan);
+  case HB_OT_TAG_OS2 : return _subset<const OT::OS2 > (plan);
+  case HB_OT_TAG_post: return _subset<const OT::post> (plan);
+  case HB_OT_TAG_COLR: return _subset<const OT::COLR> (plan);
+  case HB_OT_TAG_CBLC: return _subset<const OT::CBLC> (plan);
+  case HB_OT_TAG_CBDT: return true; /* skip CBDT, handled by CBLC */
+
+#ifndef HB_NO_SUBSET_CFF
+  case HB_OT_TAG_cff1: return _subset<const OT::cff1> (plan);
+  case HB_OT_TAG_cff2: return _subset<const OT::cff2> (plan);
+  case HB_OT_TAG_VORG: return _subset<const OT::VORG> (plan);
+#endif
+
+#ifndef HB_NO_SUBSET_LAYOUT
+  case HB_OT_TAG_GDEF: return _subset<const OT::GDEF> (plan);
+  case HB_OT_TAG_GSUB: return _subset<const OT::GSUB> (plan);
+  case HB_OT_TAG_GPOS: return _subset<const OT::GPOS> (plan);
+  case HB_OT_TAG_gvar: return _subset<const OT::gvar> (plan);
+  case HB_OT_TAG_HVAR: return _subset<const OT::HVAR> (plan);
+  case HB_OT_TAG_VVAR: return _subset<const OT::VVAR> (plan);
+#endif
+
+  default:
+    hb_blob_t *source_table = hb_face_reference_table (plan->source, tag);
+    bool result = plan->add_table (tag, source_table);
+    hb_blob_destroy (source_table);
+    return result;
+  }
+}
+
+/**
+ * hb_subset:
+ * @source: font face data to be subset.
+ * @input: input to use for the subsetting.
+ *
+ * Subsets a font according to provided input.
+ **/
+hb_face_t *
+hb_subset (hb_face_t *source, hb_subset_input_t *input)
+{
+  if (unlikely (!input || !source)) return hb_face_get_empty ();
+
+  hb_subset_plan_t *plan = hb_subset_plan_create (source, input);
+  if (unlikely (plan->in_error ()))
+    return hb_face_get_empty ();
+
+  hb_set_t tags_set;
+  bool success = true;
+  hb_tag_t table_tags[32];
+  unsigned offset = 0, num_tables = ARRAY_LENGTH (table_tags);
+  while ((hb_face_get_table_tags (source, offset, &num_tables, table_tags), num_tables))
+  {
+    for (unsigned i = 0; i < num_tables; ++i)
+    {
+      hb_tag_t tag = table_tags[i];
+      if (_should_drop_table (plan, tag) && !tags_set.has (tag)) continue;
+      tags_set.add (tag);
+      success = _subset_table (plan, tag);
+      if (unlikely (!success)) goto end;
+    }
+    offset += num_tables;
+  }
+end:
+
+  hb_face_t *result = success ? hb_face_reference (plan->dest) : hb_face_get_empty ();
+
+  hb_subset_plan_destroy (plan);
+  return result;
+}
diff --git a/thirdparty/harfbuzz/src/hb-subset.h b/thirdparty/harfbuzz/src/hb-subset.h
new file mode 100644
index 0000000000..ddf4409734
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset.h
@@ -0,0 +1,97 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Rod Sheeter
+ */
+
+#ifndef HB_SUBSET_H
+#define HB_SUBSET_H
+
+#include "hb.h"
+
+HB_BEGIN_DECLS
+
+/*
+ * hb_subset_input_t
+ *
+ * Things that change based on the input. Characters to keep, etc.
+ */
+
+typedef struct hb_subset_input_t hb_subset_input_t;
+
+HB_EXTERN hb_subset_input_t *
+hb_subset_input_create_or_fail (void);
+
+HB_EXTERN hb_subset_input_t *
+hb_subset_input_reference (hb_subset_input_t *subset_input);
+
+HB_EXTERN void
+hb_subset_input_destroy (hb_subset_input_t *subset_input);
+
+HB_EXTERN hb_set_t *
+hb_subset_input_unicode_set (hb_subset_input_t *subset_input);
+
+HB_EXTERN hb_set_t *
+hb_subset_input_glyph_set (hb_subset_input_t *subset_input);
+
+HB_EXTERN hb_set_t *
+hb_subset_input_nameid_set (hb_subset_input_t *subset_input);
+
+HB_EXTERN hb_set_t *
+hb_subset_input_namelangid_set (hb_subset_input_t *subset_input);
+
+HB_EXTERN hb_set_t *
+hb_subset_input_drop_tables_set (hb_subset_input_t *subset_input);
+
+HB_EXTERN void
+hb_subset_input_set_drop_hints (hb_subset_input_t *subset_input,
+				hb_bool_t drop_hints);
+HB_EXTERN hb_bool_t
+hb_subset_input_get_drop_hints (hb_subset_input_t *subset_input);
+
+HB_EXTERN void
+hb_subset_input_set_desubroutinize (hb_subset_input_t *subset_input,
+				    hb_bool_t desubroutinize);
+HB_EXTERN hb_bool_t
+hb_subset_input_get_desubroutinize (hb_subset_input_t *subset_input);
+
+HB_EXTERN void
+hb_subset_input_set_retain_gids (hb_subset_input_t *subset_input,
+				 hb_bool_t retain_gids);
+HB_EXTERN hb_bool_t
+hb_subset_input_get_retain_gids (hb_subset_input_t *subset_input);
+
+HB_EXTERN void
+hb_subset_input_set_name_legacy (hb_subset_input_t *subset_input,
+				 hb_bool_t name_legacy);
+HB_EXTERN hb_bool_t
+hb_subset_input_get_name_legacy (hb_subset_input_t *subset_input);
+
+/* hb_subset () */
+HB_EXTERN hb_face_t *
+hb_subset (hb_face_t *source, hb_subset_input_t *input);
+
+
+HB_END_DECLS
+
+#endif /* HB_SUBSET_H */
diff --git a/thirdparty/harfbuzz/src/hb-subset.hh b/thirdparty/harfbuzz/src/hb-subset.hh
new file mode 100644
index 0000000000..c9b01c67f3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-subset.hh
@@ -0,0 +1,73 @@
+/*
+ * Copyright © 2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Garret Rieger, Roderick Sheeter
+ */
+
+#ifndef HB_SUBSET_HH
+#define HB_SUBSET_HH
+
+
+#include "hb.hh"
+
+#include "hb-subset.h"
+
+#include "hb-machinery.hh"
+#include "hb-subset-input.hh"
+#include "hb-subset-plan.hh"
+
+struct hb_subset_context_t :
+       hb_dispatch_context_t<hb_subset_context_t, bool, HB_DEBUG_SUBSET>
+{
+  const char *get_name () { return "SUBSET"; }
+  static return_t default_return_value () { return true; }
+
+  private:
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<1>, Ts&&... ds) HB_AUTO_RETURN
+  ( obj.subset (this, hb_forward<Ts> (ds)...) )
+  template <typename T, typename ...Ts> auto
+  _dispatch (const T &obj, hb_priority<0>, Ts&&... ds) HB_AUTO_RETURN
+  ( obj.dispatch (this, hb_forward<Ts> (ds)...) )
+  public:
+  template <typename T, typename ...Ts> auto
+  dispatch (const T &obj, Ts&&... ds) HB_AUTO_RETURN
+  ( _dispatch (obj, hb_prioritize, hb_forward<Ts> (ds)...) )
+
+  hb_blob_t *source_blob;
+  hb_subset_plan_t *plan;
+  hb_serialize_context_t *serializer;
+  hb_tag_t table_tag;
+
+  hb_subset_context_t (hb_blob_t *source_blob_,
+		       hb_subset_plan_t *plan_,
+		       hb_serialize_context_t *serializer_,
+		       hb_tag_t table_tag_) :
+		        source_blob (source_blob_),
+			plan (plan_),
+			serializer (serializer_),
+			table_tag (table_tag_) {}
+};
+
+
+#endif /* HB_SUBSET_HH */
diff --git a/thirdparty/harfbuzz/src/hb-ucd-table.hh b/thirdparty/harfbuzz/src/hb-ucd-table.hh
new file mode 100644
index 0000000000..88623db338
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ucd-table.hh
@@ -0,0 +1,6780 @@
+/* == Start of generated table == */
+/*
+ * The following table is generated by running:
+ *
+ *   ./gen-ucd-table.py ucd.nounihan.grouped.xml
+ *
+ * on file with this description: Unicode 13.0.0
+ */
+
+#ifndef HB_UCD_TABLE_HH
+#define HB_UCD_TABLE_HH
+
+#include "hb.hh"
+
+static const hb_script_t
+_hb_ucd_sc_map[157] =
+{
+                   HB_SCRIPT_COMMON,              HB_SCRIPT_INHERITED,
+                  HB_SCRIPT_UNKNOWN,                 HB_SCRIPT_ARABIC,
+                 HB_SCRIPT_ARMENIAN,                HB_SCRIPT_BENGALI,
+                 HB_SCRIPT_CYRILLIC,             HB_SCRIPT_DEVANAGARI,
+                 HB_SCRIPT_GEORGIAN,                  HB_SCRIPT_GREEK,
+                 HB_SCRIPT_GUJARATI,               HB_SCRIPT_GURMUKHI,
+                   HB_SCRIPT_HANGUL,                    HB_SCRIPT_HAN,
+                   HB_SCRIPT_HEBREW,               HB_SCRIPT_HIRAGANA,
+                  HB_SCRIPT_KANNADA,               HB_SCRIPT_KATAKANA,
+                      HB_SCRIPT_LAO,                  HB_SCRIPT_LATIN,
+                HB_SCRIPT_MALAYALAM,                  HB_SCRIPT_ORIYA,
+                    HB_SCRIPT_TAMIL,                 HB_SCRIPT_TELUGU,
+                     HB_SCRIPT_THAI,                HB_SCRIPT_TIBETAN,
+                 HB_SCRIPT_BOPOMOFO,                HB_SCRIPT_BRAILLE,
+       HB_SCRIPT_CANADIAN_SYLLABICS,               HB_SCRIPT_CHEROKEE,
+                 HB_SCRIPT_ETHIOPIC,                  HB_SCRIPT_KHMER,
+                HB_SCRIPT_MONGOLIAN,                HB_SCRIPT_MYANMAR,
+                    HB_SCRIPT_OGHAM,                  HB_SCRIPT_RUNIC,
+                  HB_SCRIPT_SINHALA,                 HB_SCRIPT_SYRIAC,
+                   HB_SCRIPT_THAANA,                     HB_SCRIPT_YI,
+                  HB_SCRIPT_DESERET,                 HB_SCRIPT_GOTHIC,
+               HB_SCRIPT_OLD_ITALIC,                  HB_SCRIPT_BUHID,
+                  HB_SCRIPT_HANUNOO,                HB_SCRIPT_TAGALOG,
+                 HB_SCRIPT_TAGBANWA,                HB_SCRIPT_CYPRIOT,
+                    HB_SCRIPT_LIMBU,               HB_SCRIPT_LINEAR_B,
+                  HB_SCRIPT_OSMANYA,                HB_SCRIPT_SHAVIAN,
+                   HB_SCRIPT_TAI_LE,               HB_SCRIPT_UGARITIC,
+                 HB_SCRIPT_BUGINESE,                 HB_SCRIPT_COPTIC,
+               HB_SCRIPT_GLAGOLITIC,             HB_SCRIPT_KHAROSHTHI,
+              HB_SCRIPT_NEW_TAI_LUE,            HB_SCRIPT_OLD_PERSIAN,
+             HB_SCRIPT_SYLOTI_NAGRI,               HB_SCRIPT_TIFINAGH,
+                 HB_SCRIPT_BALINESE,              HB_SCRIPT_CUNEIFORM,
+                      HB_SCRIPT_NKO,               HB_SCRIPT_PHAGS_PA,
+               HB_SCRIPT_PHOENICIAN,                 HB_SCRIPT_CARIAN,
+                     HB_SCRIPT_CHAM,               HB_SCRIPT_KAYAH_LI,
+                   HB_SCRIPT_LEPCHA,                 HB_SCRIPT_LYCIAN,
+                   HB_SCRIPT_LYDIAN,               HB_SCRIPT_OL_CHIKI,
+                   HB_SCRIPT_REJANG,             HB_SCRIPT_SAURASHTRA,
+                HB_SCRIPT_SUNDANESE,                    HB_SCRIPT_VAI,
+                  HB_SCRIPT_AVESTAN,                  HB_SCRIPT_BAMUM,
+     HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,       HB_SCRIPT_IMPERIAL_ARAMAIC,
+    HB_SCRIPT_INSCRIPTIONAL_PAHLAVI, HB_SCRIPT_INSCRIPTIONAL_PARTHIAN,
+                 HB_SCRIPT_JAVANESE,                 HB_SCRIPT_KAITHI,
+                     HB_SCRIPT_LISU,           HB_SCRIPT_MEETEI_MAYEK,
+        HB_SCRIPT_OLD_SOUTH_ARABIAN,             HB_SCRIPT_OLD_TURKIC,
+                HB_SCRIPT_SAMARITAN,               HB_SCRIPT_TAI_THAM,
+                 HB_SCRIPT_TAI_VIET,                  HB_SCRIPT_BATAK,
+                   HB_SCRIPT_BRAHMI,                HB_SCRIPT_MANDAIC,
+                   HB_SCRIPT_CHAKMA,       HB_SCRIPT_MEROITIC_CURSIVE,
+     HB_SCRIPT_MEROITIC_HIEROGLYPHS,                   HB_SCRIPT_MIAO,
+                  HB_SCRIPT_SHARADA,           HB_SCRIPT_SORA_SOMPENG,
+                    HB_SCRIPT_TAKRI,              HB_SCRIPT_BASSA_VAH,
+       HB_SCRIPT_CAUCASIAN_ALBANIAN,               HB_SCRIPT_DUPLOYAN,
+                  HB_SCRIPT_ELBASAN,                HB_SCRIPT_GRANTHA,
+                   HB_SCRIPT_KHOJKI,              HB_SCRIPT_KHUDAWADI,
+                 HB_SCRIPT_LINEAR_A,               HB_SCRIPT_MAHAJANI,
+               HB_SCRIPT_MANICHAEAN,          HB_SCRIPT_MENDE_KIKAKUI,
+                     HB_SCRIPT_MODI,                    HB_SCRIPT_MRO,
+                HB_SCRIPT_NABATAEAN,      HB_SCRIPT_OLD_NORTH_ARABIAN,
+               HB_SCRIPT_OLD_PERMIC,           HB_SCRIPT_PAHAWH_HMONG,
+                HB_SCRIPT_PALMYRENE,            HB_SCRIPT_PAU_CIN_HAU,
+          HB_SCRIPT_PSALTER_PAHLAVI,                HB_SCRIPT_SIDDHAM,
+                  HB_SCRIPT_TIRHUTA,            HB_SCRIPT_WARANG_CITI,
+                     HB_SCRIPT_AHOM,  HB_SCRIPT_ANATOLIAN_HIEROGLYPHS,
+                   HB_SCRIPT_HATRAN,                HB_SCRIPT_MULTANI,
+            HB_SCRIPT_OLD_HUNGARIAN,            HB_SCRIPT_SIGNWRITING,
+                    HB_SCRIPT_ADLAM,              HB_SCRIPT_BHAIKSUKI,
+                  HB_SCRIPT_MARCHEN,                  HB_SCRIPT_OSAGE,
+                   HB_SCRIPT_TANGUT,                   HB_SCRIPT_NEWA,
+            HB_SCRIPT_MASARAM_GONDI,                  HB_SCRIPT_NUSHU,
+                  HB_SCRIPT_SOYOMBO,       HB_SCRIPT_ZANABAZAR_SQUARE,
+                    HB_SCRIPT_DOGRA,          HB_SCRIPT_GUNJALA_GONDI,
+          HB_SCRIPT_HANIFI_ROHINGYA,                HB_SCRIPT_MAKASAR,
+              HB_SCRIPT_MEDEFAIDRIN,            HB_SCRIPT_OLD_SOGDIAN,
+                  HB_SCRIPT_SOGDIAN,                HB_SCRIPT_ELYMAIC,
+              HB_SCRIPT_NANDINAGARI, HB_SCRIPT_NYIAKENG_PUACHUE_HMONG,
+                   HB_SCRIPT_WANCHO,             HB_SCRIPT_CHORASMIAN,
+              HB_SCRIPT_DIVES_AKURU,    HB_SCRIPT_KHITAN_SMALL_SCRIPT,
+                   HB_SCRIPT_YEZIDI,
+};
+static const uint16_t
+_hb_ucd_dm1_p0_map[825] =
+{
+   0x003Bu, 0x004Bu, 0x0060u, 0x00B4u, 0x00B7u, 0x00C5u, 0x02B9u, 0x0300u,
+   0x0301u, 0x0313u, 0x0385u, 0x0386u, 0x0388u, 0x0389u, 0x038Au, 0x038Cu,
+   0x038Eu, 0x038Fu, 0x0390u, 0x03A9u, 0x03ACu, 0x03ADu, 0x03AEu, 0x03AFu,
+   0x03B0u, 0x03B9u, 0x03CCu, 0x03CDu, 0x03CEu, 0x2002u, 0x2003u, 0x3008u,
+   0x3009u, 0x349Eu, 0x34B9u, 0x34BBu, 0x34DFu, 0x3515u, 0x36EEu, 0x36FCu,
+   0x3781u, 0x382Fu, 0x3862u, 0x387Cu, 0x38C7u, 0x38E3u, 0x391Cu, 0x393Au,
+   0x3A2Eu, 0x3A6Cu, 0x3AE4u, 0x3B08u, 0x3B19u, 0x3B49u, 0x3B9Du, 0x3C18u,
+   0x3C4Eu, 0x3D33u, 0x3D96u, 0x3EACu, 0x3EB8u, 0x3F1Bu, 0x3FFCu, 0x4008u,
+   0x4018u, 0x4039u, 0x4046u, 0x4096u, 0x40E3u, 0x412Fu, 0x4202u, 0x4227u,
+   0x42A0u, 0x4301u, 0x4334u, 0x4359u, 0x43D5u, 0x43D9u, 0x440Bu, 0x446Bu,
+   0x452Bu, 0x455Du, 0x4561u, 0x456Bu, 0x45D7u, 0x45F9u, 0x4635u, 0x46BEu,
+   0x46C7u, 0x4995u, 0x49E6u, 0x4A6Eu, 0x4A76u, 0x4AB2u, 0x4B33u, 0x4BCEu,
+   0x4CCEu, 0x4CEDu, 0x4CF8u, 0x4D56u, 0x4E0Du, 0x4E26u, 0x4E32u, 0x4E38u,
+   0x4E39u, 0x4E3Du, 0x4E41u, 0x4E82u, 0x4E86u, 0x4EAEu, 0x4EC0u, 0x4ECCu,
+   0x4EE4u, 0x4F60u, 0x4F80u, 0x4F86u, 0x4F8Bu, 0x4FAEu, 0x4FBBu, 0x4FBFu,
+   0x5002u, 0x502Bu, 0x507Au, 0x5099u, 0x50CFu, 0x50DAu, 0x50E7u, 0x5140u,
+   0x5145u, 0x514Du, 0x5154u, 0x5164u, 0x5167u, 0x5168u, 0x5169u, 0x516Du,
+   0x5177u, 0x5180u, 0x518Du, 0x5192u, 0x5195u, 0x5197u, 0x51A4u, 0x51ACu,
+   0x51B5u, 0x51B7u, 0x51C9u, 0x51CCu, 0x51DCu, 0x51DEu, 0x51F5u, 0x5203u,
+   0x5207u, 0x5217u, 0x5229u, 0x523Au, 0x523Bu, 0x5246u, 0x5272u, 0x5277u,
+   0x5289u, 0x529Bu, 0x52A3u, 0x52B3u, 0x52C7u, 0x52C9u, 0x52D2u, 0x52DEu,
+   0x52E4u, 0x52F5u, 0x52FAu, 0x5305u, 0x5306u, 0x5317u, 0x533Fu, 0x5349u,
+   0x5351u, 0x535Au, 0x5373u, 0x5375u, 0x537Du, 0x537Fu, 0x53C3u, 0x53CAu,
+   0x53DFu, 0x53E5u, 0x53EBu, 0x53F1u, 0x5406u, 0x540Fu, 0x541Du, 0x5438u,
+   0x5442u, 0x5448u, 0x5468u, 0x549Eu, 0x54A2u, 0x54BDu, 0x54F6u, 0x5510u,
+   0x5553u, 0x5555u, 0x5563u, 0x5584u, 0x5587u, 0x5599u, 0x559Du, 0x55ABu,
+   0x55B3u, 0x55C0u, 0x55C2u, 0x55E2u, 0x5606u, 0x5651u, 0x5668u, 0x5674u,
+   0x56F9u, 0x5716u, 0x5717u, 0x578Bu, 0x57CEu, 0x57F4u, 0x580Du, 0x5831u,
+   0x5832u, 0x5840u, 0x585Au, 0x585Eu, 0x58A8u, 0x58ACu, 0x58B3u, 0x58D8u,
+   0x58DFu, 0x58EEu, 0x58F2u, 0x58F7u, 0x5906u, 0x591Au, 0x5922u, 0x5944u,
+   0x5948u, 0x5951u, 0x5954u, 0x5962u, 0x5973u, 0x59D8u, 0x59ECu, 0x5A1Bu,
+   0x5A27u, 0x5A62u, 0x5A66u, 0x5AB5u, 0x5B08u, 0x5B28u, 0x5B3Eu, 0x5B85u,
+   0x5BC3u, 0x5BD8u, 0x5BE7u, 0x5BEEu, 0x5BF3u, 0x5BFFu, 0x5C06u, 0x5C22u,
+   0x5C3Fu, 0x5C60u, 0x5C62u, 0x5C64u, 0x5C65u, 0x5C6Eu, 0x5C8Du, 0x5CC0u,
+   0x5D19u, 0x5D43u, 0x5D50u, 0x5D6Bu, 0x5D6Eu, 0x5D7Cu, 0x5DB2u, 0x5DBAu,
+   0x5DE1u, 0x5DE2u, 0x5DFDu, 0x5E28u, 0x5E3Du, 0x5E69u, 0x5E74u, 0x5EA6u,
+   0x5EB0u, 0x5EB3u, 0x5EB6u, 0x5EC9u, 0x5ECAu, 0x5ED2u, 0x5ED3u, 0x5ED9u,
+   0x5EECu, 0x5EFEu, 0x5F04u, 0x5F22u, 0x5F53u, 0x5F62u, 0x5F69u, 0x5F6Bu,
+   0x5F8Bu, 0x5F9Au, 0x5FA9u, 0x5FADu, 0x5FCDu, 0x5FD7u, 0x5FF5u, 0x5FF9u,
+   0x6012u, 0x601Cu, 0x6075u, 0x6081u, 0x6094u, 0x60C7u, 0x60D8u, 0x60E1u,
+   0x6108u, 0x6144u, 0x6148u, 0x614Cu, 0x614Eu, 0x6160u, 0x6168u, 0x617Au,
+   0x618Eu, 0x6190u, 0x61A4u, 0x61AFu, 0x61B2u, 0x61DEu, 0x61F2u, 0x61F6u,
+   0x6200u, 0x6210u, 0x621Bu, 0x622Eu, 0x6234u, 0x625Du, 0x62B1u, 0x62C9u,
+   0x62CFu, 0x62D3u, 0x62D4u, 0x62FCu, 0x62FEu, 0x633Du, 0x6350u, 0x6368u,
+   0x637Bu, 0x6383u, 0x63A0u, 0x63A9u, 0x63C4u, 0x63C5u, 0x63E4u, 0x641Cu,
+   0x6422u, 0x6452u, 0x6469u, 0x6477u, 0x647Eu, 0x649Au, 0x649Du, 0x64C4u,
+   0x654Fu, 0x6556u, 0x656Cu, 0x6578u, 0x6599u, 0x65C5u, 0x65E2u, 0x65E3u,
+   0x6613u, 0x6649u, 0x6674u, 0x6688u, 0x6691u, 0x669Cu, 0x66B4u, 0x66C6u,
+   0x66F4u, 0x66F8u, 0x6700u, 0x6717u, 0x671Bu, 0x6721u, 0x674Eu, 0x6753u,
+   0x6756u, 0x675Eu, 0x677Bu, 0x6785u, 0x6797u, 0x67F3u, 0x67FAu, 0x6817u,
+   0x681Fu, 0x6852u, 0x6881u, 0x6885u, 0x688Eu, 0x68A8u, 0x6914u, 0x6942u,
+   0x69A3u, 0x69EAu, 0x6A02u, 0x6A13u, 0x6AA8u, 0x6AD3u, 0x6ADBu, 0x6B04u,
+   0x6B21u, 0x6B54u, 0x6B72u, 0x6B77u, 0x6B79u, 0x6B9Fu, 0x6BAEu, 0x6BBAu,
+   0x6BBBu, 0x6C4Eu, 0x6C67u, 0x6C88u, 0x6CBFu, 0x6CCCu, 0x6CCDu, 0x6CE5u,
+   0x6D16u, 0x6D1Bu, 0x6D1Eu, 0x6D34u, 0x6D3Eu, 0x6D41u, 0x6D69u, 0x6D6Au,
+   0x6D77u, 0x6D78u, 0x6D85u, 0x6DCBu, 0x6DDAu, 0x6DEAu, 0x6DF9u, 0x6E1Au,
+   0x6E2Fu, 0x6E6Eu, 0x6E9Cu, 0x6EBAu, 0x6EC7u, 0x6ECBu, 0x6ED1u, 0x6EDBu,
+   0x6F0Fu, 0x6F22u, 0x6F23u, 0x6F6Eu, 0x6FC6u, 0x6FEBu, 0x6FFEu, 0x701Bu,
+   0x701Eu, 0x7039u, 0x704Au, 0x7070u, 0x7077u, 0x707Du, 0x7099u, 0x70ADu,
+   0x70C8u, 0x70D9u, 0x7145u, 0x7149u, 0x716Eu, 0x719Cu, 0x71CEu, 0x71D0u,
+   0x7210u, 0x721Bu, 0x7228u, 0x722Bu, 0x7235u, 0x7250u, 0x7262u, 0x7280u,
+   0x7295u, 0x72AFu, 0x72C0u, 0x72FCu, 0x732Au, 0x7375u, 0x737Au, 0x7387u,
+   0x738Bu, 0x73A5u, 0x73B2u, 0x73DEu, 0x7406u, 0x7409u, 0x7422u, 0x7447u,
+   0x745Cu, 0x7469u, 0x7471u, 0x7485u, 0x7489u, 0x7498u, 0x74CAu, 0x7506u,
+   0x7524u, 0x753Bu, 0x753Eu, 0x7559u, 0x7565u, 0x7570u, 0x75E2u, 0x7610u,
+   0x761Du, 0x761Fu, 0x7642u, 0x7669u, 0x76CAu, 0x76DBu, 0x76E7u, 0x76F4u,
+   0x7701u, 0x771Eu, 0x771Fu, 0x7740u, 0x774Au, 0x778Bu, 0x77A7u, 0x784Eu,
+   0x786Bu, 0x788Cu, 0x7891u, 0x78CAu, 0x78CCu, 0x78FBu, 0x792Au, 0x793Cu,
+   0x793Eu, 0x7948u, 0x7949u, 0x7950u, 0x7956u, 0x795Du, 0x795Eu, 0x7965u,
+   0x797Fu, 0x798Du, 0x798Eu, 0x798Fu, 0x79AEu, 0x79CAu, 0x79EBu, 0x7A1Cu,
+   0x7A40u, 0x7A4Au, 0x7A4Fu, 0x7A81u, 0x7AB1u, 0x7ACBu, 0x7AEEu, 0x7B20u,
+   0x7BC0u, 0x7BC6u, 0x7BC9u, 0x7C3Eu, 0x7C60u, 0x7C7Bu, 0x7C92u, 0x7CBEu,
+   0x7CD2u, 0x7CD6u, 0x7CE3u, 0x7CE7u, 0x7CE8u, 0x7D00u, 0x7D10u, 0x7D22u,
+   0x7D2Fu, 0x7D5Bu, 0x7D63u, 0x7DA0u, 0x7DBEu, 0x7DC7u, 0x7DF4u, 0x7E02u,
+   0x7E09u, 0x7E37u, 0x7E41u, 0x7E45u, 0x7F3Eu, 0x7F72u, 0x7F79u, 0x7F7Au,
+   0x7F85u, 0x7F95u, 0x7F9Au, 0x7FBDu, 0x7FFAu, 0x8001u, 0x8005u, 0x8046u,
+   0x8060u, 0x806Fu, 0x8070u, 0x807Eu, 0x808Bu, 0x80ADu, 0x80B2u, 0x8103u,
+   0x813Eu, 0x81D8u, 0x81E8u, 0x81EDu, 0x8201u, 0x8204u, 0x8218u, 0x826Fu,
+   0x8279u, 0x828Bu, 0x8291u, 0x829Du, 0x82B1u, 0x82B3u, 0x82BDu, 0x82E5u,
+   0x82E6u, 0x831Du, 0x8323u, 0x8336u, 0x8352u, 0x8353u, 0x8363u, 0x83ADu,
+   0x83BDu, 0x83C9u, 0x83CAu, 0x83CCu, 0x83DCu, 0x83E7u, 0x83EFu, 0x83F1u,
+   0x843Du, 0x8449u, 0x8457u, 0x84EEu, 0x84F1u, 0x84F3u, 0x84FCu, 0x8516u,
+   0x8564u, 0x85CDu, 0x85FAu, 0x8606u, 0x8612u, 0x862Du, 0x863Fu, 0x8650u,
+   0x865Cu, 0x8667u, 0x8669u, 0x8688u, 0x86A9u, 0x86E2u, 0x870Eu, 0x8728u,
+   0x876Bu, 0x8779u, 0x8786u, 0x87BAu, 0x87E1u, 0x8801u, 0x881Fu, 0x884Cu,
+   0x8860u, 0x8863u, 0x88C2u, 0x88CFu, 0x88D7u, 0x88DEu, 0x88E1u, 0x88F8u,
+   0x88FAu, 0x8910u, 0x8941u, 0x8964u, 0x8986u, 0x898Bu, 0x8996u, 0x8AA0u,
+   0x8AAAu, 0x8ABFu, 0x8ACBu, 0x8AD2u, 0x8AD6u, 0x8AEDu, 0x8AF8u, 0x8AFEu,
+   0x8B01u, 0x8B39u, 0x8B58u, 0x8B80u, 0x8B8Au, 0x8C48u, 0x8C55u, 0x8CABu,
+   0x8CC1u, 0x8CC2u, 0x8CC8u, 0x8CD3u, 0x8D08u, 0x8D1Bu, 0x8D77u, 0x8DBCu,
+   0x8DCBu, 0x8DEFu, 0x8DF0u, 0x8ECAu, 0x8ED4u, 0x8F26u, 0x8F2Au, 0x8F38u,
+   0x8F3Bu, 0x8F62u, 0x8F9Eu, 0x8FB0u, 0x8FB6u, 0x9023u, 0x9038u, 0x9072u,
+   0x907Cu, 0x908Fu, 0x9094u, 0x90CEu, 0x90DEu, 0x90F1u, 0x90FDu, 0x9111u,
+   0x911Bu, 0x916Au, 0x9199u, 0x91B4u, 0x91CCu, 0x91CFu, 0x91D1u, 0x9234u,
+   0x9238u, 0x9276u, 0x927Cu, 0x92D7u, 0x92D8u, 0x9304u, 0x934Au, 0x93F9u,
+   0x9415u, 0x958Bu, 0x95ADu, 0x95B7u, 0x962Eu, 0x964Bu, 0x964Du, 0x9675u,
+   0x9678u, 0x967Cu, 0x9686u, 0x96A3u, 0x96B7u, 0x96B8u, 0x96C3u, 0x96E2u,
+   0x96E3u, 0x96F6u, 0x96F7u, 0x9723u, 0x9732u, 0x9748u, 0x9756u, 0x97DBu,
+   0x97E0u, 0x97FFu, 0x980Bu, 0x9818u, 0x9829u, 0x983Bu, 0x985Eu, 0x98E2u,
+   0x98EFu, 0x98FCu, 0x9928u, 0x9929u, 0x99A7u, 0x99C2u, 0x99F1u, 0x99FEu,
+   0x9A6Au, 0x9B12u, 0x9B6Fu, 0x9C40u, 0x9C57u, 0x9CFDu, 0x9D67u, 0x9DB4u,
+   0x9DFAu, 0x9E1Eu, 0x9E7Fu, 0x9E97u, 0x9E9Fu, 0x9EBBu, 0x9ECEu, 0x9EF9u,
+   0x9EFEu, 0x9F05u, 0x9F0Fu, 0x9F16u, 0x9F3Bu, 0x9F43u, 0x9F8Du, 0x9F8Eu,
+   0x9F9Cu,
+};
+static const uint16_t
+_hb_ucd_dm1_p2_map[110] =
+{
+   0x0122u, 0x051Cu, 0x0525u, 0x054Bu, 0x063Au, 0x0804u, 0x08DEu, 0x0A2Cu,
+   0x0B63u, 0x14E4u, 0x16A8u, 0x16EAu, 0x19C8u, 0x1B18u, 0x1D0Bu, 0x1DE4u,
+   0x1DE6u, 0x2183u, 0x219Fu, 0x2331u, 0x26D4u, 0x2844u, 0x284Au, 0x2B0Cu,
+   0x2BF1u, 0x300Au, 0x32B8u, 0x335Fu, 0x3393u, 0x339Cu, 0x33C3u, 0x33D5u,
+   0x346Du, 0x36A3u, 0x38A7u, 0x3A8Du, 0x3AFAu, 0x3CBCu, 0x3D1Eu, 0x3ED1u,
+   0x3F5Eu, 0x3F8Eu, 0x4263u, 0x42EEu, 0x43ABu, 0x4608u, 0x4735u, 0x4814u,
+   0x4C36u, 0x4C92u, 0x4FA1u, 0x4FB8u, 0x5044u, 0x50F2u, 0x50F3u, 0x5119u,
+   0x5133u, 0x5249u, 0x541Du, 0x5626u, 0x569Au, 0x56C5u, 0x597Cu, 0x5AA7u,
+   0x5BABu, 0x5C80u, 0x5CD0u, 0x5F86u, 0x61DAu, 0x6228u, 0x6247u, 0x62D9u,
+   0x633Eu, 0x64DAu, 0x6523u, 0x65A8u, 0x67A7u, 0x67B5u, 0x6B3Cu, 0x6C36u,
+   0x6CD5u, 0x6D6Bu, 0x6F2Cu, 0x6FB1u, 0x70D2u, 0x73CAu, 0x7667u, 0x78AEu,
+   0x7966u, 0x7CA8u, 0x7ED3u, 0x7F2Fu, 0x85D2u, 0x85EDu, 0x872Eu, 0x8BFAu,
+   0x8D77u, 0x9145u, 0x91DFu, 0x921Au, 0x940Au, 0x9496u, 0x95B6u, 0x9B30u,
+   0xA0CEu, 0xA105u, 0xA20Eu, 0xA291u, 0xA392u, 0xA600u,
+};
+static const uint32_t
+_hb_ucd_dm2_u32_map[638] =
+{
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x003Cu, 0x0338u, 0x226Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x003Du, 0x0338u, 0x2260u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x003Eu, 0x0338u, 0x226Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0300u, 0x00C0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0301u, 0x00C1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0302u, 0x00C2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0303u, 0x00C3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0304u, 0x0100u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0306u, 0x0102u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0307u, 0x0226u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0308u, 0x00C4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0309u, 0x1EA2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x030Au, 0x00C5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x030Cu, 0x01CDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x030Fu, 0x0200u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0311u, 0x0202u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0323u, 0x1EA0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0325u, 0x1E00u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0041u, 0x0328u, 0x0104u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0042u, 0x0307u, 0x1E02u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0042u, 0x0323u, 0x1E04u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0042u, 0x0331u, 0x1E06u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0043u, 0x0301u, 0x0106u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0043u, 0x0302u, 0x0108u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0043u, 0x0307u, 0x010Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0043u, 0x030Cu, 0x010Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0043u, 0x0327u, 0x00C7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x0307u, 0x1E0Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x030Cu, 0x010Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x0323u, 0x1E0Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x0327u, 0x1E10u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x032Du, 0x1E12u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0044u, 0x0331u, 0x1E0Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0300u, 0x00C8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0301u, 0x00C9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0302u, 0x00CAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0303u, 0x1EBCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0304u, 0x0112u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0306u, 0x0114u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0307u, 0x0116u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0308u, 0x00CBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0309u, 0x1EBAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x030Cu, 0x011Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x030Fu, 0x0204u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0311u, 0x0206u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0323u, 0x1EB8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0327u, 0x0228u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0328u, 0x0118u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x032Du, 0x1E18u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0045u, 0x0330u, 0x1E1Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0046u, 0x0307u, 0x1E1Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0301u, 0x01F4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0302u, 0x011Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0304u, 0x1E20u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0306u, 0x011Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0307u, 0x0120u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x030Cu, 0x01E6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0047u, 0x0327u, 0x0122u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x0302u, 0x0124u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x0307u, 0x1E22u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x0308u, 0x1E26u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x030Cu, 0x021Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x0323u, 0x1E24u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x0327u, 0x1E28u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0048u, 0x032Eu, 0x1E2Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0300u, 0x00CCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0301u, 0x00CDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0302u, 0x00CEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0303u, 0x0128u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0304u, 0x012Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0306u, 0x012Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0307u, 0x0130u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0308u, 0x00CFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0309u, 0x1EC8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x030Cu, 0x01CFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x030Fu, 0x0208u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0311u, 0x020Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0323u, 0x1ECAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0328u, 0x012Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0049u, 0x0330u, 0x1E2Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Au, 0x0302u, 0x0134u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Bu, 0x0301u, 0x1E30u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Bu, 0x030Cu, 0x01E8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Bu, 0x0323u, 0x1E32u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Bu, 0x0327u, 0x0136u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Bu, 0x0331u, 0x1E34u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x0301u, 0x0139u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x030Cu, 0x013Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x0323u, 0x1E36u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x0327u, 0x013Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x032Du, 0x1E3Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Cu, 0x0331u, 0x1E3Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Du, 0x0301u, 0x1E3Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Du, 0x0307u, 0x1E40u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Du, 0x0323u, 0x1E42u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0300u, 0x01F8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0301u, 0x0143u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0303u, 0x00D1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0307u, 0x1E44u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x030Cu, 0x0147u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0323u, 0x1E46u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0327u, 0x0145u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x032Du, 0x1E4Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Eu, 0x0331u, 0x1E48u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0300u, 0x00D2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0301u, 0x00D3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0302u, 0x00D4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0303u, 0x00D5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0304u, 0x014Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0306u, 0x014Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0307u, 0x022Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0308u, 0x00D6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0309u, 0x1ECEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x030Bu, 0x0150u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x030Cu, 0x01D1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x030Fu, 0x020Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0311u, 0x020Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x031Bu, 0x01A0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0323u, 0x1ECCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x004Fu, 0x0328u, 0x01EAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0050u, 0x0301u, 0x1E54u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0050u, 0x0307u, 0x1E56u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0301u, 0x0154u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0307u, 0x1E58u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x030Cu, 0x0158u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x030Fu, 0x0210u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0311u, 0x0212u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0323u, 0x1E5Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0327u, 0x0156u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0052u, 0x0331u, 0x1E5Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0301u, 0x015Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0302u, 0x015Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0307u, 0x1E60u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x030Cu, 0x0160u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0323u, 0x1E62u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0326u, 0x0218u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0053u, 0x0327u, 0x015Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x0307u, 0x1E6Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x030Cu, 0x0164u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x0323u, 0x1E6Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x0326u, 0x021Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x0327u, 0x0162u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x032Du, 0x1E70u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0054u, 0x0331u, 0x1E6Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0300u, 0x00D9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0301u, 0x00DAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0302u, 0x00DBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0303u, 0x0168u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0304u, 0x016Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0306u, 0x016Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0308u, 0x00DCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0309u, 0x1EE6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x030Au, 0x016Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x030Bu, 0x0170u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x030Cu, 0x01D3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x030Fu, 0x0214u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0311u, 0x0216u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x031Bu, 0x01AFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0323u, 0x1EE4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0324u, 0x1E72u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0328u, 0x0172u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x032Du, 0x1E76u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0055u, 0x0330u, 0x1E74u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0056u, 0x0303u, 0x1E7Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0056u, 0x0323u, 0x1E7Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0300u, 0x1E80u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0301u, 0x1E82u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0302u, 0x0174u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0307u, 0x1E86u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0308u, 0x1E84u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0057u, 0x0323u, 0x1E88u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0058u, 0x0307u, 0x1E8Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0058u, 0x0308u, 0x1E8Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0300u, 0x1EF2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0301u, 0x00DDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0302u, 0x0176u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0303u, 0x1EF8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0304u, 0x0232u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0307u, 0x1E8Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0308u, 0x0178u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0309u, 0x1EF6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0059u, 0x0323u, 0x1EF4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x0301u, 0x0179u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x0302u, 0x1E90u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x0307u, 0x017Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x030Cu, 0x017Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x0323u, 0x1E92u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x005Au, 0x0331u, 0x1E94u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0300u, 0x00E0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0301u, 0x00E1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0302u, 0x00E2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0303u, 0x00E3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0304u, 0x0101u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0306u, 0x0103u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0307u, 0x0227u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0308u, 0x00E4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0309u, 0x1EA3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x030Au, 0x00E5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x030Cu, 0x01CEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x030Fu, 0x0201u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0311u, 0x0203u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0323u, 0x1EA1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0325u, 0x1E01u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0061u, 0x0328u, 0x0105u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0062u, 0x0307u, 0x1E03u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0062u, 0x0323u, 0x1E05u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0062u, 0x0331u, 0x1E07u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0063u, 0x0301u, 0x0107u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0063u, 0x0302u, 0x0109u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0063u, 0x0307u, 0x010Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0063u, 0x030Cu, 0x010Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0063u, 0x0327u, 0x00E7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x0307u, 0x1E0Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x030Cu, 0x010Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x0323u, 0x1E0Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x0327u, 0x1E11u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x032Du, 0x1E13u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0064u, 0x0331u, 0x1E0Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0300u, 0x00E8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0301u, 0x00E9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0302u, 0x00EAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0303u, 0x1EBDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0304u, 0x0113u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0306u, 0x0115u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0307u, 0x0117u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0308u, 0x00EBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0309u, 0x1EBBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x030Cu, 0x011Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x030Fu, 0x0205u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0311u, 0x0207u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0323u, 0x1EB9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0327u, 0x0229u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0328u, 0x0119u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x032Du, 0x1E19u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0065u, 0x0330u, 0x1E1Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0066u, 0x0307u, 0x1E1Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0301u, 0x01F5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0302u, 0x011Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0304u, 0x1E21u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0306u, 0x011Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0307u, 0x0121u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x030Cu, 0x01E7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0067u, 0x0327u, 0x0123u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0302u, 0x0125u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0307u, 0x1E23u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0308u, 0x1E27u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x030Cu, 0x021Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0323u, 0x1E25u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0327u, 0x1E29u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x032Eu, 0x1E2Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0068u, 0x0331u, 0x1E96u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0300u, 0x00ECu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0301u, 0x00EDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0302u, 0x00EEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0303u, 0x0129u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0304u, 0x012Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0306u, 0x012Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0308u, 0x00EFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0309u, 0x1EC9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x030Cu, 0x01D0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x030Fu, 0x0209u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0311u, 0x020Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0323u, 0x1ECBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0328u, 0x012Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0069u, 0x0330u, 0x1E2Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Au, 0x0302u, 0x0135u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Au, 0x030Cu, 0x01F0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Bu, 0x0301u, 0x1E31u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Bu, 0x030Cu, 0x01E9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Bu, 0x0323u, 0x1E33u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Bu, 0x0327u, 0x0137u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Bu, 0x0331u, 0x1E35u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x0301u, 0x013Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x030Cu, 0x013Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x0323u, 0x1E37u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x0327u, 0x013Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x032Du, 0x1E3Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Cu, 0x0331u, 0x1E3Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Du, 0x0301u, 0x1E3Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Du, 0x0307u, 0x1E41u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Du, 0x0323u, 0x1E43u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0300u, 0x01F9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0301u, 0x0144u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0303u, 0x00F1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0307u, 0x1E45u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x030Cu, 0x0148u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0323u, 0x1E47u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0327u, 0x0146u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x032Du, 0x1E4Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Eu, 0x0331u, 0x1E49u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0300u, 0x00F2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0301u, 0x00F3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0302u, 0x00F4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0303u, 0x00F5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0304u, 0x014Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0306u, 0x014Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0307u, 0x022Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0308u, 0x00F6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0309u, 0x1ECFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x030Bu, 0x0151u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x030Cu, 0x01D2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x030Fu, 0x020Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0311u, 0x020Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x031Bu, 0x01A1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0323u, 0x1ECDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x006Fu, 0x0328u, 0x01EBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0070u, 0x0301u, 0x1E55u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0070u, 0x0307u, 0x1E57u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0301u, 0x0155u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0307u, 0x1E59u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x030Cu, 0x0159u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x030Fu, 0x0211u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0311u, 0x0213u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0323u, 0x1E5Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0327u, 0x0157u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0072u, 0x0331u, 0x1E5Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0301u, 0x015Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0302u, 0x015Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0307u, 0x1E61u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x030Cu, 0x0161u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0323u, 0x1E63u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0326u, 0x0219u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0073u, 0x0327u, 0x015Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0307u, 0x1E6Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0308u, 0x1E97u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x030Cu, 0x0165u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0323u, 0x1E6Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0326u, 0x021Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0327u, 0x0163u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x032Du, 0x1E71u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0074u, 0x0331u, 0x1E6Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0300u, 0x00F9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0301u, 0x00FAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0302u, 0x00FBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0303u, 0x0169u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0304u, 0x016Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0306u, 0x016Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0308u, 0x00FCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0309u, 0x1EE7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x030Au, 0x016Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x030Bu, 0x0171u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x030Cu, 0x01D4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x030Fu, 0x0215u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0311u, 0x0217u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x031Bu, 0x01B0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0323u, 0x1EE5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0324u, 0x1E73u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0328u, 0x0173u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x032Du, 0x1E77u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0075u, 0x0330u, 0x1E75u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0076u, 0x0303u, 0x1E7Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0076u, 0x0323u, 0x1E7Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0300u, 0x1E81u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0301u, 0x1E83u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0302u, 0x0175u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0307u, 0x1E87u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0308u, 0x1E85u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x030Au, 0x1E98u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0077u, 0x0323u, 0x1E89u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0078u, 0x0307u, 0x1E8Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0078u, 0x0308u, 0x1E8Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0300u, 0x1EF3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0301u, 0x00FDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0302u, 0x0177u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0303u, 0x1EF9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0304u, 0x0233u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0307u, 0x1E8Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0308u, 0x00FFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0309u, 0x1EF7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x030Au, 0x1E99u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0079u, 0x0323u, 0x1EF5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x0301u, 0x017Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x0302u, 0x1E91u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x0307u, 0x017Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x030Cu, 0x017Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x0323u, 0x1E93u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x007Au, 0x0331u, 0x1E95u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00A8u, 0x0300u, 0x1FEDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00A8u, 0x0301u, 0x0385u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00A8u, 0x0342u, 0x1FC1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C2u, 0x0300u, 0x1EA6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C2u, 0x0301u, 0x1EA4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C2u, 0x0303u, 0x1EAAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C2u, 0x0309u, 0x1EA8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C4u, 0x0304u, 0x01DEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C5u, 0x0301u, 0x01FAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C6u, 0x0301u, 0x01FCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C6u, 0x0304u, 0x01E2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00C7u, 0x0301u, 0x1E08u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00CAu, 0x0300u, 0x1EC0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00CAu, 0x0301u, 0x1EBEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00CAu, 0x0303u, 0x1EC4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00CAu, 0x0309u, 0x1EC2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00CFu, 0x0301u, 0x1E2Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D4u, 0x0300u, 0x1ED2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D4u, 0x0301u, 0x1ED0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D4u, 0x0303u, 0x1ED6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D4u, 0x0309u, 0x1ED4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D5u, 0x0301u, 0x1E4Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D5u, 0x0304u, 0x022Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D5u, 0x0308u, 0x1E4Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D6u, 0x0304u, 0x022Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00D8u, 0x0301u, 0x01FEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00DCu, 0x0300u, 0x01DBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00DCu, 0x0301u, 0x01D7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00DCu, 0x0304u, 0x01D5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00DCu, 0x030Cu, 0x01D9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E2u, 0x0300u, 0x1EA7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E2u, 0x0301u, 0x1EA5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E2u, 0x0303u, 0x1EABu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E2u, 0x0309u, 0x1EA9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E4u, 0x0304u, 0x01DFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E5u, 0x0301u, 0x01FBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E6u, 0x0301u, 0x01FDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E6u, 0x0304u, 0x01E3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00E7u, 0x0301u, 0x1E09u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00EAu, 0x0300u, 0x1EC1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00EAu, 0x0301u, 0x1EBFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00EAu, 0x0303u, 0x1EC5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00EAu, 0x0309u, 0x1EC3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00EFu, 0x0301u, 0x1E2Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F4u, 0x0300u, 0x1ED3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F4u, 0x0301u, 0x1ED1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F4u, 0x0303u, 0x1ED7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F4u, 0x0309u, 0x1ED5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F5u, 0x0301u, 0x1E4Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F5u, 0x0304u, 0x022Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F5u, 0x0308u, 0x1E4Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F6u, 0x0304u, 0x022Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00F8u, 0x0301u, 0x01FFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00FCu, 0x0300u, 0x01DCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00FCu, 0x0301u, 0x01D8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00FCu, 0x0304u, 0x01D6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x00FCu, 0x030Cu, 0x01DAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0102u, 0x0300u, 0x1EB0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0102u, 0x0301u, 0x1EAEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0102u, 0x0303u, 0x1EB4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0102u, 0x0309u, 0x1EB2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0103u, 0x0300u, 0x1EB1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0103u, 0x0301u, 0x1EAFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0103u, 0x0303u, 0x1EB5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0103u, 0x0309u, 0x1EB3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0112u, 0x0300u, 0x1E14u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0112u, 0x0301u, 0x1E16u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0113u, 0x0300u, 0x1E15u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0113u, 0x0301u, 0x1E17u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x014Cu, 0x0300u, 0x1E50u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x014Cu, 0x0301u, 0x1E52u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x014Du, 0x0300u, 0x1E51u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x014Du, 0x0301u, 0x1E53u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x015Au, 0x0307u, 0x1E64u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x015Bu, 0x0307u, 0x1E65u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0160u, 0x0307u, 0x1E66u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0161u, 0x0307u, 0x1E67u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0168u, 0x0301u, 0x1E78u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0169u, 0x0301u, 0x1E79u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x016Au, 0x0308u, 0x1E7Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x016Bu, 0x0308u, 0x1E7Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x017Fu, 0x0307u, 0x1E9Bu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A0u, 0x0300u, 0x1EDCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A0u, 0x0301u, 0x1EDAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A0u, 0x0303u, 0x1EE0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A0u, 0x0309u, 0x1EDEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A0u, 0x0323u, 0x1EE2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A1u, 0x0300u, 0x1EDDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A1u, 0x0301u, 0x1EDBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A1u, 0x0303u, 0x1EE1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A1u, 0x0309u, 0x1EDFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01A1u, 0x0323u, 0x1EE3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01AFu, 0x0300u, 0x1EEAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01AFu, 0x0301u, 0x1EE8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01AFu, 0x0303u, 0x1EEEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01AFu, 0x0309u, 0x1EECu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01AFu, 0x0323u, 0x1EF0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B0u, 0x0300u, 0x1EEBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B0u, 0x0301u, 0x1EE9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B0u, 0x0303u, 0x1EEFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B0u, 0x0309u, 0x1EEDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B0u, 0x0323u, 0x1EF1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01B7u, 0x030Cu, 0x01EEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01EAu, 0x0304u, 0x01ECu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x01EBu, 0x0304u, 0x01EDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0226u, 0x0304u, 0x01E0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0227u, 0x0304u, 0x01E1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0228u, 0x0306u, 0x1E1Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0229u, 0x0306u, 0x1E1Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x022Eu, 0x0304u, 0x0230u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x022Fu, 0x0304u, 0x0231u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0292u, 0x030Cu, 0x01EFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0308u, 0x0301u, 0x0000u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0300u, 0x1FBAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0301u, 0x0386u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0304u, 0x1FB9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0306u, 0x1FB8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0313u, 0x1F08u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0314u, 0x1F09u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0391u, 0x0345u, 0x1FBCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0395u, 0x0300u, 0x1FC8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0395u, 0x0301u, 0x0388u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0395u, 0x0313u, 0x1F18u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0395u, 0x0314u, 0x1F19u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0397u, 0x0300u, 0x1FCAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0397u, 0x0301u, 0x0389u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0397u, 0x0313u, 0x1F28u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0397u, 0x0314u, 0x1F29u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0397u, 0x0345u, 0x1FCCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0300u, 0x1FDAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0301u, 0x038Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0304u, 0x1FD9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0306u, 0x1FD8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0308u, 0x03AAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0313u, 0x1F38u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0399u, 0x0314u, 0x1F39u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x039Fu, 0x0300u, 0x1FF8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x039Fu, 0x0301u, 0x038Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x039Fu, 0x0313u, 0x1F48u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x039Fu, 0x0314u, 0x1F49u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A1u, 0x0314u, 0x1FECu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0300u, 0x1FEAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0301u, 0x038Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0304u, 0x1FE9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0306u, 0x1FE8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0308u, 0x03ABu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A5u, 0x0314u, 0x1F59u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A9u, 0x0300u, 0x1FFAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A9u, 0x0301u, 0x038Fu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A9u, 0x0313u, 0x1F68u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A9u, 0x0314u, 0x1F69u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03A9u, 0x0345u, 0x1FFCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03ACu, 0x0345u, 0x1FB4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03AEu, 0x0345u, 0x1FC4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0300u, 0x1F70u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0301u, 0x03ACu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0304u, 0x1FB1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0306u, 0x1FB0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0313u, 0x1F00u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0314u, 0x1F01u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0342u, 0x1FB6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B1u, 0x0345u, 0x1FB3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B5u, 0x0300u, 0x1F72u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B5u, 0x0301u, 0x03ADu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B5u, 0x0313u, 0x1F10u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B5u, 0x0314u, 0x1F11u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0300u, 0x1F74u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0301u, 0x03AEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0313u, 0x1F20u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0314u, 0x1F21u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0342u, 0x1FC6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B7u, 0x0345u, 0x1FC3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0300u, 0x1F76u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0301u, 0x03AFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0304u, 0x1FD1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0306u, 0x1FD0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0308u, 0x03CAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0313u, 0x1F30u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0314u, 0x1F31u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03B9u, 0x0342u, 0x1FD6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03BFu, 0x0300u, 0x1F78u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03BFu, 0x0301u, 0x03CCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03BFu, 0x0313u, 0x1F40u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03BFu, 0x0314u, 0x1F41u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C1u, 0x0313u, 0x1FE4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C1u, 0x0314u, 0x1FE5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0300u, 0x1F7Au),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0301u, 0x03CDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0304u, 0x1FE1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0306u, 0x1FE0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0308u, 0x03CBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0313u, 0x1F50u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0314u, 0x1F51u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C5u, 0x0342u, 0x1FE6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0300u, 0x1F7Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0301u, 0x03CEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0313u, 0x1F60u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0314u, 0x1F61u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0342u, 0x1FF6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03C9u, 0x0345u, 0x1FF3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CAu, 0x0300u, 0x1FD2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CAu, 0x0301u, 0x0390u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CAu, 0x0342u, 0x1FD7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CBu, 0x0300u, 0x1FE2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CBu, 0x0301u, 0x03B0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CBu, 0x0342u, 0x1FE7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03CEu, 0x0345u, 0x1FF4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03D2u, 0x0301u, 0x03D3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x03D2u, 0x0308u, 0x03D4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0406u, 0x0308u, 0x0407u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0410u, 0x0306u, 0x04D0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0410u, 0x0308u, 0x04D2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0413u, 0x0301u, 0x0403u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0415u, 0x0300u, 0x0400u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0415u, 0x0306u, 0x04D6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0415u, 0x0308u, 0x0401u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0416u, 0x0306u, 0x04C1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0416u, 0x0308u, 0x04DCu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0417u, 0x0308u, 0x04DEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0418u, 0x0300u, 0x040Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0418u, 0x0304u, 0x04E2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0418u, 0x0306u, 0x0419u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0418u, 0x0308u, 0x04E4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x041Au, 0x0301u, 0x040Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x041Eu, 0x0308u, 0x04E6u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0423u, 0x0304u, 0x04EEu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0423u, 0x0306u, 0x040Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0423u, 0x0308u, 0x04F0u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0423u, 0x030Bu, 0x04F2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0427u, 0x0308u, 0x04F4u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x042Bu, 0x0308u, 0x04F8u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x042Du, 0x0308u, 0x04ECu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0430u, 0x0306u, 0x04D1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0430u, 0x0308u, 0x04D3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0433u, 0x0301u, 0x0453u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0435u, 0x0300u, 0x0450u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0435u, 0x0306u, 0x04D7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0435u, 0x0308u, 0x0451u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0436u, 0x0306u, 0x04C2u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0436u, 0x0308u, 0x04DDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0437u, 0x0308u, 0x04DFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0438u, 0x0300u, 0x045Du),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0438u, 0x0304u, 0x04E3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0438u, 0x0306u, 0x0439u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0438u, 0x0308u, 0x04E5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x043Au, 0x0301u, 0x045Cu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x043Eu, 0x0308u, 0x04E7u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0443u, 0x0304u, 0x04EFu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0443u, 0x0306u, 0x045Eu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0443u, 0x0308u, 0x04F1u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0443u, 0x030Bu, 0x04F3u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0447u, 0x0308u, 0x04F5u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x044Bu, 0x0308u, 0x04F9u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x044Du, 0x0308u, 0x04EDu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0456u, 0x0308u, 0x0457u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0474u, 0x030Fu, 0x0476u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x0475u, 0x030Fu, 0x0477u),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x04D8u, 0x0308u, 0x04DAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x04D9u, 0x0308u, 0x04DBu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x04E8u, 0x0308u, 0x04EAu),
+   HB_CODEPOINT_ENCODE3_11_7_14 (0x04E9u, 0x0308u, 0x04EBu),
+};
+static const uint64_t
+_hb_ucd_dm2_u64_map[388] =
+{
+     HB_CODEPOINT_ENCODE3 (0x05D0u, 0x05B7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D0u, 0x05B8u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D0u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D1u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D1u, 0x05BFu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D2u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D3u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D4u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D5u, 0x05B9u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D5u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D6u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D8u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05D9u, 0x05B4u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05D9u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05DAu, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05DBu, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05DBu, 0x05BFu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05DCu, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05DEu, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E0u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05E1u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E3u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05E4u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E4u, 0x05BFu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05E6u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E7u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05E8u, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E9u, 0x05BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05E9u, 0x05C1u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05E9u, 0x05C2u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x05EAu, 0x05BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x05F2u, 0x05B7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0627u, 0x0653u, 0x0622u),   HB_CODEPOINT_ENCODE3 (0x0627u, 0x0654u, 0x0623u),
+     HB_CODEPOINT_ENCODE3 (0x0627u, 0x0655u, 0x0625u),   HB_CODEPOINT_ENCODE3 (0x0648u, 0x0654u, 0x0624u),
+     HB_CODEPOINT_ENCODE3 (0x064Au, 0x0654u, 0x0626u),   HB_CODEPOINT_ENCODE3 (0x06C1u, 0x0654u, 0x06C2u),
+     HB_CODEPOINT_ENCODE3 (0x06D2u, 0x0654u, 0x06D3u),   HB_CODEPOINT_ENCODE3 (0x06D5u, 0x0654u, 0x06C0u),
+     HB_CODEPOINT_ENCODE3 (0x0915u, 0x093Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0916u, 0x093Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0917u, 0x093Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x091Cu, 0x093Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0921u, 0x093Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0922u, 0x093Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0928u, 0x093Cu, 0x0929u),   HB_CODEPOINT_ENCODE3 (0x092Bu, 0x093Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x092Fu, 0x093Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0930u, 0x093Cu, 0x0931u),
+     HB_CODEPOINT_ENCODE3 (0x0933u, 0x093Cu, 0x0934u),   HB_CODEPOINT_ENCODE3 (0x09A1u, 0x09BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x09A2u, 0x09BCu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x09AFu, 0x09BCu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x09C7u, 0x09BEu, 0x09CBu),   HB_CODEPOINT_ENCODE3 (0x09C7u, 0x09D7u, 0x09CCu),
+     HB_CODEPOINT_ENCODE3 (0x0A16u, 0x0A3Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0A17u, 0x0A3Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0A1Cu, 0x0A3Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0A2Bu, 0x0A3Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0A32u, 0x0A3Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0A38u, 0x0A3Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0B21u, 0x0B3Cu, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0B22u, 0x0B3Cu, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0B47u, 0x0B3Eu, 0x0B4Bu),   HB_CODEPOINT_ENCODE3 (0x0B47u, 0x0B56u, 0x0B48u),
+     HB_CODEPOINT_ENCODE3 (0x0B47u, 0x0B57u, 0x0B4Cu),   HB_CODEPOINT_ENCODE3 (0x0B92u, 0x0BD7u, 0x0B94u),
+     HB_CODEPOINT_ENCODE3 (0x0BC6u, 0x0BBEu, 0x0BCAu),   HB_CODEPOINT_ENCODE3 (0x0BC6u, 0x0BD7u, 0x0BCCu),
+     HB_CODEPOINT_ENCODE3 (0x0BC7u, 0x0BBEu, 0x0BCBu),   HB_CODEPOINT_ENCODE3 (0x0C46u, 0x0C56u, 0x0C48u),
+     HB_CODEPOINT_ENCODE3 (0x0CBFu, 0x0CD5u, 0x0CC0u),   HB_CODEPOINT_ENCODE3 (0x0CC6u, 0x0CC2u, 0x0CCAu),
+     HB_CODEPOINT_ENCODE3 (0x0CC6u, 0x0CD5u, 0x0CC7u),   HB_CODEPOINT_ENCODE3 (0x0CC6u, 0x0CD6u, 0x0CC8u),
+     HB_CODEPOINT_ENCODE3 (0x0CCAu, 0x0CD5u, 0x0CCBu),   HB_CODEPOINT_ENCODE3 (0x0D46u, 0x0D3Eu, 0x0D4Au),
+     HB_CODEPOINT_ENCODE3 (0x0D46u, 0x0D57u, 0x0D4Cu),   HB_CODEPOINT_ENCODE3 (0x0D47u, 0x0D3Eu, 0x0D4Bu),
+     HB_CODEPOINT_ENCODE3 (0x0DD9u, 0x0DCAu, 0x0DDAu),   HB_CODEPOINT_ENCODE3 (0x0DD9u, 0x0DCFu, 0x0DDCu),
+     HB_CODEPOINT_ENCODE3 (0x0DD9u, 0x0DDFu, 0x0DDEu),   HB_CODEPOINT_ENCODE3 (0x0DDCu, 0x0DCAu, 0x0DDDu),
+     HB_CODEPOINT_ENCODE3 (0x0F40u, 0x0FB5u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F42u, 0x0FB7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0F4Cu, 0x0FB7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F51u, 0x0FB7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0F56u, 0x0FB7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F5Bu, 0x0FB7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0F71u, 0x0F72u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F71u, 0x0F74u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0F71u, 0x0F80u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F90u, 0x0FB5u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0F92u, 0x0FB7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0F9Cu, 0x0FB7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0FA1u, 0x0FB7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0FA6u, 0x0FB7u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0FABu, 0x0FB7u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x0FB2u, 0x0F80u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x0FB3u, 0x0F80u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0x1025u, 0x102Eu, 0x1026u),
+     HB_CODEPOINT_ENCODE3 (0x1B05u, 0x1B35u, 0x1B06u),   HB_CODEPOINT_ENCODE3 (0x1B07u, 0x1B35u, 0x1B08u),
+     HB_CODEPOINT_ENCODE3 (0x1B09u, 0x1B35u, 0x1B0Au),   HB_CODEPOINT_ENCODE3 (0x1B0Bu, 0x1B35u, 0x1B0Cu),
+     HB_CODEPOINT_ENCODE3 (0x1B0Du, 0x1B35u, 0x1B0Eu),   HB_CODEPOINT_ENCODE3 (0x1B11u, 0x1B35u, 0x1B12u),
+     HB_CODEPOINT_ENCODE3 (0x1B3Au, 0x1B35u, 0x1B3Bu),   HB_CODEPOINT_ENCODE3 (0x1B3Cu, 0x1B35u, 0x1B3Du),
+     HB_CODEPOINT_ENCODE3 (0x1B3Eu, 0x1B35u, 0x1B40u),   HB_CODEPOINT_ENCODE3 (0x1B3Fu, 0x1B35u, 0x1B41u),
+     HB_CODEPOINT_ENCODE3 (0x1B42u, 0x1B35u, 0x1B43u),   HB_CODEPOINT_ENCODE3 (0x1E36u, 0x0304u, 0x1E38u),
+     HB_CODEPOINT_ENCODE3 (0x1E37u, 0x0304u, 0x1E39u),   HB_CODEPOINT_ENCODE3 (0x1E5Au, 0x0304u, 0x1E5Cu),
+     HB_CODEPOINT_ENCODE3 (0x1E5Bu, 0x0304u, 0x1E5Du),   HB_CODEPOINT_ENCODE3 (0x1E62u, 0x0307u, 0x1E68u),
+     HB_CODEPOINT_ENCODE3 (0x1E63u, 0x0307u, 0x1E69u),   HB_CODEPOINT_ENCODE3 (0x1EA0u, 0x0302u, 0x1EACu),
+     HB_CODEPOINT_ENCODE3 (0x1EA0u, 0x0306u, 0x1EB6u),   HB_CODEPOINT_ENCODE3 (0x1EA1u, 0x0302u, 0x1EADu),
+     HB_CODEPOINT_ENCODE3 (0x1EA1u, 0x0306u, 0x1EB7u),   HB_CODEPOINT_ENCODE3 (0x1EB8u, 0x0302u, 0x1EC6u),
+     HB_CODEPOINT_ENCODE3 (0x1EB9u, 0x0302u, 0x1EC7u),   HB_CODEPOINT_ENCODE3 (0x1ECCu, 0x0302u, 0x1ED8u),
+     HB_CODEPOINT_ENCODE3 (0x1ECDu, 0x0302u, 0x1ED9u),   HB_CODEPOINT_ENCODE3 (0x1F00u, 0x0300u, 0x1F02u),
+     HB_CODEPOINT_ENCODE3 (0x1F00u, 0x0301u, 0x1F04u),   HB_CODEPOINT_ENCODE3 (0x1F00u, 0x0342u, 0x1F06u),
+     HB_CODEPOINT_ENCODE3 (0x1F00u, 0x0345u, 0x1F80u),   HB_CODEPOINT_ENCODE3 (0x1F01u, 0x0300u, 0x1F03u),
+     HB_CODEPOINT_ENCODE3 (0x1F01u, 0x0301u, 0x1F05u),   HB_CODEPOINT_ENCODE3 (0x1F01u, 0x0342u, 0x1F07u),
+     HB_CODEPOINT_ENCODE3 (0x1F01u, 0x0345u, 0x1F81u),   HB_CODEPOINT_ENCODE3 (0x1F02u, 0x0345u, 0x1F82u),
+     HB_CODEPOINT_ENCODE3 (0x1F03u, 0x0345u, 0x1F83u),   HB_CODEPOINT_ENCODE3 (0x1F04u, 0x0345u, 0x1F84u),
+     HB_CODEPOINT_ENCODE3 (0x1F05u, 0x0345u, 0x1F85u),   HB_CODEPOINT_ENCODE3 (0x1F06u, 0x0345u, 0x1F86u),
+     HB_CODEPOINT_ENCODE3 (0x1F07u, 0x0345u, 0x1F87u),   HB_CODEPOINT_ENCODE3 (0x1F08u, 0x0300u, 0x1F0Au),
+     HB_CODEPOINT_ENCODE3 (0x1F08u, 0x0301u, 0x1F0Cu),   HB_CODEPOINT_ENCODE3 (0x1F08u, 0x0342u, 0x1F0Eu),
+     HB_CODEPOINT_ENCODE3 (0x1F08u, 0x0345u, 0x1F88u),   HB_CODEPOINT_ENCODE3 (0x1F09u, 0x0300u, 0x1F0Bu),
+     HB_CODEPOINT_ENCODE3 (0x1F09u, 0x0301u, 0x1F0Du),   HB_CODEPOINT_ENCODE3 (0x1F09u, 0x0342u, 0x1F0Fu),
+     HB_CODEPOINT_ENCODE3 (0x1F09u, 0x0345u, 0x1F89u),   HB_CODEPOINT_ENCODE3 (0x1F0Au, 0x0345u, 0x1F8Au),
+     HB_CODEPOINT_ENCODE3 (0x1F0Bu, 0x0345u, 0x1F8Bu),   HB_CODEPOINT_ENCODE3 (0x1F0Cu, 0x0345u, 0x1F8Cu),
+     HB_CODEPOINT_ENCODE3 (0x1F0Du, 0x0345u, 0x1F8Du),   HB_CODEPOINT_ENCODE3 (0x1F0Eu, 0x0345u, 0x1F8Eu),
+     HB_CODEPOINT_ENCODE3 (0x1F0Fu, 0x0345u, 0x1F8Fu),   HB_CODEPOINT_ENCODE3 (0x1F10u, 0x0300u, 0x1F12u),
+     HB_CODEPOINT_ENCODE3 (0x1F10u, 0x0301u, 0x1F14u),   HB_CODEPOINT_ENCODE3 (0x1F11u, 0x0300u, 0x1F13u),
+     HB_CODEPOINT_ENCODE3 (0x1F11u, 0x0301u, 0x1F15u),   HB_CODEPOINT_ENCODE3 (0x1F18u, 0x0300u, 0x1F1Au),
+     HB_CODEPOINT_ENCODE3 (0x1F18u, 0x0301u, 0x1F1Cu),   HB_CODEPOINT_ENCODE3 (0x1F19u, 0x0300u, 0x1F1Bu),
+     HB_CODEPOINT_ENCODE3 (0x1F19u, 0x0301u, 0x1F1Du),   HB_CODEPOINT_ENCODE3 (0x1F20u, 0x0300u, 0x1F22u),
+     HB_CODEPOINT_ENCODE3 (0x1F20u, 0x0301u, 0x1F24u),   HB_CODEPOINT_ENCODE3 (0x1F20u, 0x0342u, 0x1F26u),
+     HB_CODEPOINT_ENCODE3 (0x1F20u, 0x0345u, 0x1F90u),   HB_CODEPOINT_ENCODE3 (0x1F21u, 0x0300u, 0x1F23u),
+     HB_CODEPOINT_ENCODE3 (0x1F21u, 0x0301u, 0x1F25u),   HB_CODEPOINT_ENCODE3 (0x1F21u, 0x0342u, 0x1F27u),
+     HB_CODEPOINT_ENCODE3 (0x1F21u, 0x0345u, 0x1F91u),   HB_CODEPOINT_ENCODE3 (0x1F22u, 0x0345u, 0x1F92u),
+     HB_CODEPOINT_ENCODE3 (0x1F23u, 0x0345u, 0x1F93u),   HB_CODEPOINT_ENCODE3 (0x1F24u, 0x0345u, 0x1F94u),
+     HB_CODEPOINT_ENCODE3 (0x1F25u, 0x0345u, 0x1F95u),   HB_CODEPOINT_ENCODE3 (0x1F26u, 0x0345u, 0x1F96u),
+     HB_CODEPOINT_ENCODE3 (0x1F27u, 0x0345u, 0x1F97u),   HB_CODEPOINT_ENCODE3 (0x1F28u, 0x0300u, 0x1F2Au),
+     HB_CODEPOINT_ENCODE3 (0x1F28u, 0x0301u, 0x1F2Cu),   HB_CODEPOINT_ENCODE3 (0x1F28u, 0x0342u, 0x1F2Eu),
+     HB_CODEPOINT_ENCODE3 (0x1F28u, 0x0345u, 0x1F98u),   HB_CODEPOINT_ENCODE3 (0x1F29u, 0x0300u, 0x1F2Bu),
+     HB_CODEPOINT_ENCODE3 (0x1F29u, 0x0301u, 0x1F2Du),   HB_CODEPOINT_ENCODE3 (0x1F29u, 0x0342u, 0x1F2Fu),
+     HB_CODEPOINT_ENCODE3 (0x1F29u, 0x0345u, 0x1F99u),   HB_CODEPOINT_ENCODE3 (0x1F2Au, 0x0345u, 0x1F9Au),
+     HB_CODEPOINT_ENCODE3 (0x1F2Bu, 0x0345u, 0x1F9Bu),   HB_CODEPOINT_ENCODE3 (0x1F2Cu, 0x0345u, 0x1F9Cu),
+     HB_CODEPOINT_ENCODE3 (0x1F2Du, 0x0345u, 0x1F9Du),   HB_CODEPOINT_ENCODE3 (0x1F2Eu, 0x0345u, 0x1F9Eu),
+     HB_CODEPOINT_ENCODE3 (0x1F2Fu, 0x0345u, 0x1F9Fu),   HB_CODEPOINT_ENCODE3 (0x1F30u, 0x0300u, 0x1F32u),
+     HB_CODEPOINT_ENCODE3 (0x1F30u, 0x0301u, 0x1F34u),   HB_CODEPOINT_ENCODE3 (0x1F30u, 0x0342u, 0x1F36u),
+     HB_CODEPOINT_ENCODE3 (0x1F31u, 0x0300u, 0x1F33u),   HB_CODEPOINT_ENCODE3 (0x1F31u, 0x0301u, 0x1F35u),
+     HB_CODEPOINT_ENCODE3 (0x1F31u, 0x0342u, 0x1F37u),   HB_CODEPOINT_ENCODE3 (0x1F38u, 0x0300u, 0x1F3Au),
+     HB_CODEPOINT_ENCODE3 (0x1F38u, 0x0301u, 0x1F3Cu),   HB_CODEPOINT_ENCODE3 (0x1F38u, 0x0342u, 0x1F3Eu),
+     HB_CODEPOINT_ENCODE3 (0x1F39u, 0x0300u, 0x1F3Bu),   HB_CODEPOINT_ENCODE3 (0x1F39u, 0x0301u, 0x1F3Du),
+     HB_CODEPOINT_ENCODE3 (0x1F39u, 0x0342u, 0x1F3Fu),   HB_CODEPOINT_ENCODE3 (0x1F40u, 0x0300u, 0x1F42u),
+     HB_CODEPOINT_ENCODE3 (0x1F40u, 0x0301u, 0x1F44u),   HB_CODEPOINT_ENCODE3 (0x1F41u, 0x0300u, 0x1F43u),
+     HB_CODEPOINT_ENCODE3 (0x1F41u, 0x0301u, 0x1F45u),   HB_CODEPOINT_ENCODE3 (0x1F48u, 0x0300u, 0x1F4Au),
+     HB_CODEPOINT_ENCODE3 (0x1F48u, 0x0301u, 0x1F4Cu),   HB_CODEPOINT_ENCODE3 (0x1F49u, 0x0300u, 0x1F4Bu),
+     HB_CODEPOINT_ENCODE3 (0x1F49u, 0x0301u, 0x1F4Du),   HB_CODEPOINT_ENCODE3 (0x1F50u, 0x0300u, 0x1F52u),
+     HB_CODEPOINT_ENCODE3 (0x1F50u, 0x0301u, 0x1F54u),   HB_CODEPOINT_ENCODE3 (0x1F50u, 0x0342u, 0x1F56u),
+     HB_CODEPOINT_ENCODE3 (0x1F51u, 0x0300u, 0x1F53u),   HB_CODEPOINT_ENCODE3 (0x1F51u, 0x0301u, 0x1F55u),
+     HB_CODEPOINT_ENCODE3 (0x1F51u, 0x0342u, 0x1F57u),   HB_CODEPOINT_ENCODE3 (0x1F59u, 0x0300u, 0x1F5Bu),
+     HB_CODEPOINT_ENCODE3 (0x1F59u, 0x0301u, 0x1F5Du),   HB_CODEPOINT_ENCODE3 (0x1F59u, 0x0342u, 0x1F5Fu),
+     HB_CODEPOINT_ENCODE3 (0x1F60u, 0x0300u, 0x1F62u),   HB_CODEPOINT_ENCODE3 (0x1F60u, 0x0301u, 0x1F64u),
+     HB_CODEPOINT_ENCODE3 (0x1F60u, 0x0342u, 0x1F66u),   HB_CODEPOINT_ENCODE3 (0x1F60u, 0x0345u, 0x1FA0u),
+     HB_CODEPOINT_ENCODE3 (0x1F61u, 0x0300u, 0x1F63u),   HB_CODEPOINT_ENCODE3 (0x1F61u, 0x0301u, 0x1F65u),
+     HB_CODEPOINT_ENCODE3 (0x1F61u, 0x0342u, 0x1F67u),   HB_CODEPOINT_ENCODE3 (0x1F61u, 0x0345u, 0x1FA1u),
+     HB_CODEPOINT_ENCODE3 (0x1F62u, 0x0345u, 0x1FA2u),   HB_CODEPOINT_ENCODE3 (0x1F63u, 0x0345u, 0x1FA3u),
+     HB_CODEPOINT_ENCODE3 (0x1F64u, 0x0345u, 0x1FA4u),   HB_CODEPOINT_ENCODE3 (0x1F65u, 0x0345u, 0x1FA5u),
+     HB_CODEPOINT_ENCODE3 (0x1F66u, 0x0345u, 0x1FA6u),   HB_CODEPOINT_ENCODE3 (0x1F67u, 0x0345u, 0x1FA7u),
+     HB_CODEPOINT_ENCODE3 (0x1F68u, 0x0300u, 0x1F6Au),   HB_CODEPOINT_ENCODE3 (0x1F68u, 0x0301u, 0x1F6Cu),
+     HB_CODEPOINT_ENCODE3 (0x1F68u, 0x0342u, 0x1F6Eu),   HB_CODEPOINT_ENCODE3 (0x1F68u, 0x0345u, 0x1FA8u),
+     HB_CODEPOINT_ENCODE3 (0x1F69u, 0x0300u, 0x1F6Bu),   HB_CODEPOINT_ENCODE3 (0x1F69u, 0x0301u, 0x1F6Du),
+     HB_CODEPOINT_ENCODE3 (0x1F69u, 0x0342u, 0x1F6Fu),   HB_CODEPOINT_ENCODE3 (0x1F69u, 0x0345u, 0x1FA9u),
+     HB_CODEPOINT_ENCODE3 (0x1F6Au, 0x0345u, 0x1FAAu),   HB_CODEPOINT_ENCODE3 (0x1F6Bu, 0x0345u, 0x1FABu),
+     HB_CODEPOINT_ENCODE3 (0x1F6Cu, 0x0345u, 0x1FACu),   HB_CODEPOINT_ENCODE3 (0x1F6Du, 0x0345u, 0x1FADu),
+     HB_CODEPOINT_ENCODE3 (0x1F6Eu, 0x0345u, 0x1FAEu),   HB_CODEPOINT_ENCODE3 (0x1F6Fu, 0x0345u, 0x1FAFu),
+     HB_CODEPOINT_ENCODE3 (0x1F70u, 0x0345u, 0x1FB2u),   HB_CODEPOINT_ENCODE3 (0x1F74u, 0x0345u, 0x1FC2u),
+     HB_CODEPOINT_ENCODE3 (0x1F7Cu, 0x0345u, 0x1FF2u),   HB_CODEPOINT_ENCODE3 (0x1FB6u, 0x0345u, 0x1FB7u),
+     HB_CODEPOINT_ENCODE3 (0x1FBFu, 0x0300u, 0x1FCDu),   HB_CODEPOINT_ENCODE3 (0x1FBFu, 0x0301u, 0x1FCEu),
+     HB_CODEPOINT_ENCODE3 (0x1FBFu, 0x0342u, 0x1FCFu),   HB_CODEPOINT_ENCODE3 (0x1FC6u, 0x0345u, 0x1FC7u),
+     HB_CODEPOINT_ENCODE3 (0x1FF6u, 0x0345u, 0x1FF7u),   HB_CODEPOINT_ENCODE3 (0x1FFEu, 0x0300u, 0x1FDDu),
+     HB_CODEPOINT_ENCODE3 (0x1FFEu, 0x0301u, 0x1FDEu),   HB_CODEPOINT_ENCODE3 (0x1FFEu, 0x0342u, 0x1FDFu),
+     HB_CODEPOINT_ENCODE3 (0x2190u, 0x0338u, 0x219Au),   HB_CODEPOINT_ENCODE3 (0x2192u, 0x0338u, 0x219Bu),
+     HB_CODEPOINT_ENCODE3 (0x2194u, 0x0338u, 0x21AEu),   HB_CODEPOINT_ENCODE3 (0x21D0u, 0x0338u, 0x21CDu),
+     HB_CODEPOINT_ENCODE3 (0x21D2u, 0x0338u, 0x21CFu),   HB_CODEPOINT_ENCODE3 (0x21D4u, 0x0338u, 0x21CEu),
+     HB_CODEPOINT_ENCODE3 (0x2203u, 0x0338u, 0x2204u),   HB_CODEPOINT_ENCODE3 (0x2208u, 0x0338u, 0x2209u),
+     HB_CODEPOINT_ENCODE3 (0x220Bu, 0x0338u, 0x220Cu),   HB_CODEPOINT_ENCODE3 (0x2223u, 0x0338u, 0x2224u),
+     HB_CODEPOINT_ENCODE3 (0x2225u, 0x0338u, 0x2226u),   HB_CODEPOINT_ENCODE3 (0x223Cu, 0x0338u, 0x2241u),
+     HB_CODEPOINT_ENCODE3 (0x2243u, 0x0338u, 0x2244u),   HB_CODEPOINT_ENCODE3 (0x2245u, 0x0338u, 0x2247u),
+     HB_CODEPOINT_ENCODE3 (0x2248u, 0x0338u, 0x2249u),   HB_CODEPOINT_ENCODE3 (0x224Du, 0x0338u, 0x226Du),
+     HB_CODEPOINT_ENCODE3 (0x2261u, 0x0338u, 0x2262u),   HB_CODEPOINT_ENCODE3 (0x2264u, 0x0338u, 0x2270u),
+     HB_CODEPOINT_ENCODE3 (0x2265u, 0x0338u, 0x2271u),   HB_CODEPOINT_ENCODE3 (0x2272u, 0x0338u, 0x2274u),
+     HB_CODEPOINT_ENCODE3 (0x2273u, 0x0338u, 0x2275u),   HB_CODEPOINT_ENCODE3 (0x2276u, 0x0338u, 0x2278u),
+     HB_CODEPOINT_ENCODE3 (0x2277u, 0x0338u, 0x2279u),   HB_CODEPOINT_ENCODE3 (0x227Au, 0x0338u, 0x2280u),
+     HB_CODEPOINT_ENCODE3 (0x227Bu, 0x0338u, 0x2281u),   HB_CODEPOINT_ENCODE3 (0x227Cu, 0x0338u, 0x22E0u),
+     HB_CODEPOINT_ENCODE3 (0x227Du, 0x0338u, 0x22E1u),   HB_CODEPOINT_ENCODE3 (0x2282u, 0x0338u, 0x2284u),
+     HB_CODEPOINT_ENCODE3 (0x2283u, 0x0338u, 0x2285u),   HB_CODEPOINT_ENCODE3 (0x2286u, 0x0338u, 0x2288u),
+     HB_CODEPOINT_ENCODE3 (0x2287u, 0x0338u, 0x2289u),   HB_CODEPOINT_ENCODE3 (0x2291u, 0x0338u, 0x22E2u),
+     HB_CODEPOINT_ENCODE3 (0x2292u, 0x0338u, 0x22E3u),   HB_CODEPOINT_ENCODE3 (0x22A2u, 0x0338u, 0x22ACu),
+     HB_CODEPOINT_ENCODE3 (0x22A8u, 0x0338u, 0x22ADu),   HB_CODEPOINT_ENCODE3 (0x22A9u, 0x0338u, 0x22AEu),
+     HB_CODEPOINT_ENCODE3 (0x22ABu, 0x0338u, 0x22AFu),   HB_CODEPOINT_ENCODE3 (0x22B2u, 0x0338u, 0x22EAu),
+     HB_CODEPOINT_ENCODE3 (0x22B3u, 0x0338u, 0x22EBu),   HB_CODEPOINT_ENCODE3 (0x22B4u, 0x0338u, 0x22ECu),
+     HB_CODEPOINT_ENCODE3 (0x22B5u, 0x0338u, 0x22EDu),   HB_CODEPOINT_ENCODE3 (0x2ADDu, 0x0338u, 0x0000u),
+     HB_CODEPOINT_ENCODE3 (0x3046u, 0x3099u, 0x3094u),   HB_CODEPOINT_ENCODE3 (0x304Bu, 0x3099u, 0x304Cu),
+     HB_CODEPOINT_ENCODE3 (0x304Du, 0x3099u, 0x304Eu),   HB_CODEPOINT_ENCODE3 (0x304Fu, 0x3099u, 0x3050u),
+     HB_CODEPOINT_ENCODE3 (0x3051u, 0x3099u, 0x3052u),   HB_CODEPOINT_ENCODE3 (0x3053u, 0x3099u, 0x3054u),
+     HB_CODEPOINT_ENCODE3 (0x3055u, 0x3099u, 0x3056u),   HB_CODEPOINT_ENCODE3 (0x3057u, 0x3099u, 0x3058u),
+     HB_CODEPOINT_ENCODE3 (0x3059u, 0x3099u, 0x305Au),   HB_CODEPOINT_ENCODE3 (0x305Bu, 0x3099u, 0x305Cu),
+     HB_CODEPOINT_ENCODE3 (0x305Du, 0x3099u, 0x305Eu),   HB_CODEPOINT_ENCODE3 (0x305Fu, 0x3099u, 0x3060u),
+     HB_CODEPOINT_ENCODE3 (0x3061u, 0x3099u, 0x3062u),   HB_CODEPOINT_ENCODE3 (0x3064u, 0x3099u, 0x3065u),
+     HB_CODEPOINT_ENCODE3 (0x3066u, 0x3099u, 0x3067u),   HB_CODEPOINT_ENCODE3 (0x3068u, 0x3099u, 0x3069u),
+     HB_CODEPOINT_ENCODE3 (0x306Fu, 0x3099u, 0x3070u),   HB_CODEPOINT_ENCODE3 (0x306Fu, 0x309Au, 0x3071u),
+     HB_CODEPOINT_ENCODE3 (0x3072u, 0x3099u, 0x3073u),   HB_CODEPOINT_ENCODE3 (0x3072u, 0x309Au, 0x3074u),
+     HB_CODEPOINT_ENCODE3 (0x3075u, 0x3099u, 0x3076u),   HB_CODEPOINT_ENCODE3 (0x3075u, 0x309Au, 0x3077u),
+     HB_CODEPOINT_ENCODE3 (0x3078u, 0x3099u, 0x3079u),   HB_CODEPOINT_ENCODE3 (0x3078u, 0x309Au, 0x307Au),
+     HB_CODEPOINT_ENCODE3 (0x307Bu, 0x3099u, 0x307Cu),   HB_CODEPOINT_ENCODE3 (0x307Bu, 0x309Au, 0x307Du),
+     HB_CODEPOINT_ENCODE3 (0x309Du, 0x3099u, 0x309Eu),   HB_CODEPOINT_ENCODE3 (0x30A6u, 0x3099u, 0x30F4u),
+     HB_CODEPOINT_ENCODE3 (0x30ABu, 0x3099u, 0x30ACu),   HB_CODEPOINT_ENCODE3 (0x30ADu, 0x3099u, 0x30AEu),
+     HB_CODEPOINT_ENCODE3 (0x30AFu, 0x3099u, 0x30B0u),   HB_CODEPOINT_ENCODE3 (0x30B1u, 0x3099u, 0x30B2u),
+     HB_CODEPOINT_ENCODE3 (0x30B3u, 0x3099u, 0x30B4u),   HB_CODEPOINT_ENCODE3 (0x30B5u, 0x3099u, 0x30B6u),
+     HB_CODEPOINT_ENCODE3 (0x30B7u, 0x3099u, 0x30B8u),   HB_CODEPOINT_ENCODE3 (0x30B9u, 0x3099u, 0x30BAu),
+     HB_CODEPOINT_ENCODE3 (0x30BBu, 0x3099u, 0x30BCu),   HB_CODEPOINT_ENCODE3 (0x30BDu, 0x3099u, 0x30BEu),
+     HB_CODEPOINT_ENCODE3 (0x30BFu, 0x3099u, 0x30C0u),   HB_CODEPOINT_ENCODE3 (0x30C1u, 0x3099u, 0x30C2u),
+     HB_CODEPOINT_ENCODE3 (0x30C4u, 0x3099u, 0x30C5u),   HB_CODEPOINT_ENCODE3 (0x30C6u, 0x3099u, 0x30C7u),
+     HB_CODEPOINT_ENCODE3 (0x30C8u, 0x3099u, 0x30C9u),   HB_CODEPOINT_ENCODE3 (0x30CFu, 0x3099u, 0x30D0u),
+     HB_CODEPOINT_ENCODE3 (0x30CFu, 0x309Au, 0x30D1u),   HB_CODEPOINT_ENCODE3 (0x30D2u, 0x3099u, 0x30D3u),
+     HB_CODEPOINT_ENCODE3 (0x30D2u, 0x309Au, 0x30D4u),   HB_CODEPOINT_ENCODE3 (0x30D5u, 0x3099u, 0x30D6u),
+     HB_CODEPOINT_ENCODE3 (0x30D5u, 0x309Au, 0x30D7u),   HB_CODEPOINT_ENCODE3 (0x30D8u, 0x3099u, 0x30D9u),
+     HB_CODEPOINT_ENCODE3 (0x30D8u, 0x309Au, 0x30DAu),   HB_CODEPOINT_ENCODE3 (0x30DBu, 0x3099u, 0x30DCu),
+     HB_CODEPOINT_ENCODE3 (0x30DBu, 0x309Au, 0x30DDu),   HB_CODEPOINT_ENCODE3 (0x30EFu, 0x3099u, 0x30F7u),
+     HB_CODEPOINT_ENCODE3 (0x30F0u, 0x3099u, 0x30F8u),   HB_CODEPOINT_ENCODE3 (0x30F1u, 0x3099u, 0x30F9u),
+     HB_CODEPOINT_ENCODE3 (0x30F2u, 0x3099u, 0x30FAu),   HB_CODEPOINT_ENCODE3 (0x30FDu, 0x3099u, 0x30FEu),
+     HB_CODEPOINT_ENCODE3 (0xFB49u, 0x05C1u, 0x0000u),   HB_CODEPOINT_ENCODE3 (0xFB49u, 0x05C2u, 0x0000u),
+  HB_CODEPOINT_ENCODE3 (0x11099u, 0x110BAu, 0x1109Au),HB_CODEPOINT_ENCODE3 (0x1109Bu, 0x110BAu, 0x1109Cu),
+  HB_CODEPOINT_ENCODE3 (0x110A5u, 0x110BAu, 0x110ABu),HB_CODEPOINT_ENCODE3 (0x11131u, 0x11127u, 0x1112Eu),
+  HB_CODEPOINT_ENCODE3 (0x11132u, 0x11127u, 0x1112Fu),HB_CODEPOINT_ENCODE3 (0x11347u, 0x1133Eu, 0x1134Bu),
+  HB_CODEPOINT_ENCODE3 (0x11347u, 0x11357u, 0x1134Cu),HB_CODEPOINT_ENCODE3 (0x114B9u, 0x114B0u, 0x114BCu),
+  HB_CODEPOINT_ENCODE3 (0x114B9u, 0x114BAu, 0x114BBu),HB_CODEPOINT_ENCODE3 (0x114B9u, 0x114BDu, 0x114BEu),
+  HB_CODEPOINT_ENCODE3 (0x115B8u, 0x115AFu, 0x115BAu),HB_CODEPOINT_ENCODE3 (0x115B9u, 0x115AFu, 0x115BBu),
+  HB_CODEPOINT_ENCODE3 (0x11935u, 0x11930u, 0x11938u), HB_CODEPOINT_ENCODE3 (0x1D157u, 0x1D165u, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D158u, 0x1D165u, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D15Fu, 0x1D16Eu, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D15Fu, 0x1D16Fu, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D15Fu, 0x1D170u, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D15Fu, 0x1D171u, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D15Fu, 0x1D172u, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D1B9u, 0x1D165u, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D1BAu, 0x1D165u, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D1BBu, 0x1D16Eu, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D1BBu, 0x1D16Fu, 0x0000u),
+   HB_CODEPOINT_ENCODE3 (0x1D1BCu, 0x1D16Eu, 0x0000u), HB_CODEPOINT_ENCODE3 (0x1D1BCu, 0x1D16Fu, 0x0000u),
+};
+
+#ifndef HB_OPTIMIZE_SIZE
+
+static const uint8_t
+_hb_ucd_u8[32480] =
+{
+    0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15,
+   16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 27, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 28,
+   29, 26, 30, 31, 32, 33, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 34, 35, 35, 35, 35,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 37, 38, 39, 40,
+   41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56,
+   26, 57, 58, 59, 59, 59, 59, 59, 26, 26, 60, 59, 59, 59, 59, 59,
+   59, 59, 26, 61, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 26, 62, 59, 63, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 64, 26, 26, 65, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 66, 67, 59, 59, 59, 59, 68, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 69, 70, 71, 72, 73, 74, 59, 59,
+   75, 76, 59, 59, 77, 59, 78, 79, 80, 81, 73, 82, 83, 84, 59, 59,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 85, 26, 26, 26, 26, 26, 26, 26, 86, 87, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 88, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 89, 59, 59, 59, 59, 59, 59, 26, 90, 59, 59,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 91, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   92, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 93,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 94,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   29, 21, 21, 21, 23, 21, 21, 21, 22, 18, 21, 25, 21, 17, 21, 21,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 21, 21, 25, 25, 25, 21,
+   21,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 22, 21, 18, 24, 16,
+   24,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 22, 25, 18, 25,  0,
+   29, 21, 23, 23, 23, 23, 26, 21, 24, 26,  7, 20, 25,  1, 26, 24,
+   26, 25, 15, 15, 24,  5, 21, 21, 24, 15,  7, 19, 15, 15, 15, 21,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9, 25,  9,  9,  9,  9,  9,  9,  9,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5, 25,  5,  5,  5,  5,  5,  5,  5,  5,
+    9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    9,  5,  9,  5,  9,  5,  9,  5,  5,  9,  5,  9,  5,  9,  5,  9,
+    5,  9,  5,  9,  5,  9,  5,  9,  5,  5,  9,  5,  9,  5,  9,  5,
+    9,  5,  9,  5,  9,  5,  9,  5,  9,  9,  5,  9,  5,  9,  5,  5,
+    5,  9,  9,  5,  9,  5,  9,  9,  5,  9,  9,  9,  5,  5,  9,  9,
+    9,  9,  5,  9,  9,  5,  9,  9,  9,  5,  5,  5,  9,  9,  5,  9,
+    9,  5,  9,  5,  9,  5,  9,  9,  5,  9,  5,  5,  9,  5,  9,  9,
+    5,  9,  9,  9,  5,  9,  5,  9,  9,  5,  5,  7,  9,  5,  5,  5,
+    7,  7,  7,  7,  9,  8,  5,  9,  8,  5,  9,  8,  5,  9,  5,  9,
+    5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  5,  9,  5,
+    5,  9,  8,  5,  9,  5,  9,  9,  9,  5,  9,  5,  9,  5,  9,  5,
+    9,  5,  9,  5,  5,  5,  5,  5,  5,  5,  9,  9,  5,  9,  9,  5,
+    5,  9,  5,  9,  9,  9,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    5,  5,  5,  5,  7,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,
+    6,  6, 24, 24, 24, 24,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,
+    6,  6, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+    6,  6,  6,  6,  6, 24, 24, 24, 24, 24, 24, 24,  6, 24,  6, 24,
+   24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+    9,  5,  9,  5,  6, 24,  9,  5,  2,  2,  6,  5,  5,  5, 21,  9,
+    2,  2,  2,  2, 24, 24,  9, 21,  9,  9,  9,  2,  9,  2,  9,  9,
+    5,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  2,  9,  9,  9,  9,  9,  9,  9,  9,  9,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  9,
+    5,  5,  9,  9,  9,  5,  5,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    5,  5,  5,  5,  9,  5, 25,  9,  5,  9,  9,  5,  5,  9,  9,  9,
+    9,  5, 26, 12, 12, 12, 12, 12, 11, 11,  9,  5,  9,  5,  9,  5,
+    9,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  5,
+    2,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  2,  2,  6, 21, 21, 21, 21, 21, 21,
+    5,  5,  5,  5,  5,  5,  5,  5,  5, 21, 17,  2,  2, 26, 26, 23,
+    2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 17, 12,
+   21, 12, 12, 21, 12, 12, 21, 12,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  7,
+    7,  7,  7, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    1,  1,  1,  1,  1,  1, 25, 25, 25, 21, 21, 23, 21, 21, 26, 26,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 21,  1,  2, 21, 21,
+    6,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 21, 21, 21, 21,  7,  7,
+   12,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7, 21,  7, 12, 12, 12, 12, 12, 12, 12,  1, 26, 12,
+   12, 12, 12, 12, 12,  6,  6, 12, 12, 26, 12, 12, 12, 12,  7,  7,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  7,  7,  7, 26, 26,  7,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  1,
+    7, 12,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,  2,  2,  7,  7,  7,
+    7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12,
+   12, 12, 12, 12,  6,  6, 26, 21, 21, 21,  6,  2,  2, 12, 23, 23,
+    7,  7,  7,  7,  7,  7, 12, 12, 12, 12,  6, 12, 12, 12, 12, 12,
+   12, 12, 12, 12,  6, 12, 12, 12,  6, 12, 12, 12, 12, 12,  2,  2,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12,  2,  2, 21,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12,  1, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 10, 12,  7, 10, 10,
+   10, 12, 12, 12, 12, 12, 12, 12, 12, 10, 10, 10, 10, 12, 10, 10,
+    7, 12, 12, 12, 12, 12, 12, 12,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7, 12, 12, 21, 21, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   21,  6,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7, 12, 10, 10,  2,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  7,
+    7,  2,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,
+    7,  2,  7,  2,  2,  2,  7,  7,  7,  7,  2,  2, 12,  7, 10, 10,
+   10, 12, 12, 12, 12,  2,  2, 10, 10,  2,  2, 10, 10, 12,  7,  2,
+    2,  2,  2,  2,  2,  2,  2, 10,  2,  2,  2,  2,  7,  7,  2,  7,
+    7,  7, 12, 12,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    7,  7, 23, 23, 15, 15, 15, 15, 15, 15, 26, 23,  7, 21, 12,  2,
+    2, 12, 12, 10,  2,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  7,
+    7,  2,  7,  7,  2,  7,  7,  2,  7,  7,  2,  2, 12,  2, 10, 10,
+   10, 12, 12,  2,  2,  2,  2, 12, 12,  2,  2, 12, 12, 12,  2,  2,
+    2, 12,  2,  2,  2,  2,  2,  2,  2,  7,  7,  7,  7,  2,  7,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   12, 12,  7,  7,  7, 12, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2, 12, 12, 10,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,
+    7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  2,  2, 12,  7, 10, 10,
+   10, 12, 12, 12, 12, 12,  2, 12, 12, 10,  2, 10, 10, 12,  2,  2,
+    7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   21, 23,  2,  2,  2,  2,  2,  2,  2,  7, 12, 12, 12, 12, 12, 12,
+    2, 12, 10, 10,  2,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  7,
+    7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  2,  2, 12,  7, 10, 12,
+   10, 12, 12, 12, 12,  2,  2, 10, 10,  2,  2, 10, 10, 12,  2,  2,
+    2,  2,  2,  2,  2, 12, 12, 10,  2,  2,  2,  2,  7,  7,  2,  7,
+   26,  7, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2, 12,  7,  2,  7,  7,  7,  7,  7,  7,  2,  2,  2,  7,  7,
+    7,  2,  7,  7,  7,  7,  2,  2,  2,  7,  7,  2,  7,  2,  7,  7,
+    2,  2,  2,  7,  7,  2,  2,  2,  7,  7,  7,  2,  2,  2,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2, 10, 10,
+   12, 10, 10,  2,  2,  2, 10, 10, 10,  2, 10, 10, 10, 12,  2,  2,
+    7,  2,  2,  2,  2,  2,  2, 10,  2,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 26, 26, 26, 26, 26, 26, 23, 26,  2,  2,  2,  2,  2,
+   12, 10, 10, 10, 12,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,
+    7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  7, 12, 12,
+   12, 10, 10, 10, 10,  2, 12, 12, 12,  2, 12, 12, 12, 12,  2,  2,
+    2,  2,  2,  2,  2, 12, 12,  2,  7,  7,  7,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2, 21, 15, 15, 15, 15, 15, 15, 15, 26,
+    7, 12, 10, 10, 21,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,
+    7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  2,  2, 12,  7, 10, 12,
+   10, 10, 10, 10, 10,  2, 12, 10, 10,  2, 10, 10, 12, 12,  2,  2,
+    2,  2,  2,  2,  2, 10, 10,  2,  2,  2,  2,  2,  2,  2,  7,  2,
+    2,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 10, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12,  7, 10, 10,
+   10, 12, 12, 12, 12,  2, 10, 10, 10,  2, 10, 10, 10, 12,  7, 26,
+    2,  2,  2,  2,  7,  7,  7, 10, 15, 15, 15, 15, 15, 15, 15,  7,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 26,  7,  7,  7,  7,  7,  7,
+    2, 12, 10, 10,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  7,  7,  7,  7,  7,  7,
+    7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  2,  2, 12,  2,  2,  2,  2, 10,
+   10, 10, 12, 12, 12,  2, 12,  2, 10, 10, 10, 10, 10, 10, 10, 10,
+    2,  2, 10, 10, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7, 12,  7,  7, 12, 12, 12, 12, 12, 12, 12,  2,  2,  2,  2, 23,
+    7,  7,  7,  7,  7,  7,  6, 12, 12, 12, 12, 12, 12, 12, 12, 21,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 21, 21,  2,  2,  2,  2,
+    2,  7,  7,  2,  7,  2,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,
+    7,  7,  7,  7,  2,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7, 12,  7,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,  7,  2,  2,
+    7,  7,  7,  7,  7,  2,  6,  2, 12, 12, 12, 12, 12, 12,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  7,  7,  7,  7,
+    7, 26, 26, 26, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 26, 21, 26, 26, 26, 12, 12, 26, 26, 26, 26, 26, 26,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 26, 12, 26, 12, 26, 12, 22, 18, 22, 18, 10, 10,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,
+    2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 10,
+   12, 12, 12, 12, 12, 21, 12, 12,  7,  7,  7,  7,  7, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12,  2, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,  2, 26, 26,
+   26, 26, 26, 26, 26, 26, 12, 26, 26, 26, 26, 26, 26,  2, 26, 26,
+   21, 21, 21, 21, 21, 26, 26, 26, 26, 21, 21,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 10, 10, 12, 12, 12,
+   12, 10, 12, 12, 12, 12, 12, 12, 10, 12, 12, 10, 10, 12, 12,  7,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 21, 21, 21, 21, 21, 21,
+    7,  7,  7,  7,  7,  7, 10, 10, 12, 12,  7,  7,  7,  7, 12, 12,
+   12,  7, 10, 10, 10,  7,  7, 10, 10, 10, 10, 10, 10, 10,  7,  7,
+    7, 12, 12, 12, 12,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7, 12, 10, 10, 12, 12, 10, 10, 10, 10, 10, 10, 12,  7, 10,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 10, 10, 10, 12, 26, 26,
+    9,  9,  9,  9,  9,  9,  2,  9,  2,  2,  2,  2,  2,  9,  2,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 21,  6,  5,  5,  5,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  2,  7,  7,  7,  7,  2,  2,
+    7,  2,  7,  7,  7,  7,  2,  2,  7,  7,  7,  7,  7,  7,  7,  2,
+    7,  2,  7,  7,  7,  7,  2,  2,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2, 12, 12, 12,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,  2,
+    9,  9,  9,  9,  9,  9,  2,  2,  5,  5,  5,  5,  5,  5,  2,  2,
+   17,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 26, 21,  7,
+   29,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 22, 18,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 21, 21, 21, 14, 14,
+   14,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,
+    7,  7, 12, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7, 12, 12, 12, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  2, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7, 12, 12, 10, 12, 12, 12, 12, 12, 12, 12, 10, 10,
+   10, 10, 10, 10, 10, 10, 12, 10, 10, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 21, 21, 21,  6, 21, 21, 21, 23,  7, 12,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,
+   21, 21, 21, 21, 21, 21, 17, 21, 21, 21, 21, 12, 12, 12,  1,  2,
+    7,  7,  7,  6,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7, 12, 12,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7, 12,  7,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,
+   12, 12, 12, 10, 10, 10, 10, 12, 12, 10, 10, 10,  2,  2,  2,  2,
+   10, 10, 12, 10, 10, 10, 10, 10, 10, 12, 12, 12,  2,  2,  2,  2,
+   26,  2,  2,  2, 21, 21, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,
+    7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 15,  2,  2,  2, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+    7,  7,  7,  7,  7,  7,  7, 12, 12, 10, 10, 12,  2,  2, 21, 21,
+    7,  7,  7,  7,  7, 10, 12, 10, 12, 12, 12, 12, 12, 12, 12,  2,
+   12, 10, 12, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12, 10, 10, 10,
+   10, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,  2,  2, 12,
+   21, 21, 21, 21, 21, 21, 21,  6, 21, 21, 21, 21, 21, 21,  2,  2,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 11, 12,
+   12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 12, 12, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7, 12, 10, 12, 12, 12, 12, 12, 10, 12, 10, 10, 10,
+   10, 10, 12, 10, 10,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,
+   21, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,
+   12, 12, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7, 10, 12, 12, 12, 12, 10, 10, 12, 12, 10, 12, 12, 12,  7,  7,
+    7,  7,  7,  7,  7,  7, 12, 10, 12, 12, 10, 10, 10, 12, 10, 12,
+   12, 12, 10, 10,  2,  2,  2,  2,  2,  2,  2,  2, 21, 21, 21, 21,
+    7,  7,  7,  7, 10, 10, 10, 10, 10, 10, 10, 10, 12, 12, 12, 12,
+   12, 12, 12, 12, 10, 10, 12, 12,  2,  2,  2, 21, 21, 21, 21, 21,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  6,  6,  6,  6,  6,  6, 21, 21,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2,  2,  2,  2,  2,  2,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  2,  2,  9,  9,  9,
+   21, 21, 21, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 12, 21, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 10, 12, 12, 12, 12, 12, 12, 12,  7,  7,  7,  7, 12,  7,  7,
+    7,  7,  7,  7, 12,  7,  7, 10, 12, 12,  7,  2,  2,  2,  2,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  6,  6,  6,  6,
+    6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  6,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  6,  6,  6,  6,  6,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12,  2, 12, 12, 12, 12, 12,
+    9,  5,  9,  5,  9,  5,  5,  5,  5,  5,  5,  5,  5,  5,  9,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  9,  9,  9,  9,  9,  9,  9,  9,
+    5,  5,  5,  5,  5,  5,  2,  2,  9,  9,  9,  9,  9,  9,  2,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  2,  9,  2,  9,  2,  9,  2,  9,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  8,  8,  8,  8,  8,  8,  8,  8,
+    5,  5,  5,  5,  5,  2,  5,  5,  9,  9,  9,  9,  8, 24,  5, 24,
+   24, 24,  5,  5,  5,  2,  5,  5,  9,  9,  9,  9,  8, 24, 24, 24,
+    5,  5,  5,  5,  2,  2,  5,  5,  9,  9,  9,  9,  2, 24, 24, 24,
+    5,  5,  5,  5,  5,  5,  5,  5,  9,  9,  9,  9,  9, 24, 24, 24,
+    2,  2,  5,  5,  5,  2,  5,  5,  9,  9,  9,  9,  8, 24, 24,  2,
+   29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,  1,  1,  1,  1,  1,
+   17, 17, 17, 17, 17, 17, 21, 21, 20, 19, 22, 20, 20, 19, 22, 20,
+   21, 21, 21, 21, 21, 21, 21, 21, 27, 28,  1,  1,  1,  1,  1, 29,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 20, 19, 21, 21, 21, 21, 16,
+   16, 21, 21, 21, 25, 22, 18, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 25, 21, 16, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 29,
+    1,  1,  1,  1,  1,  2,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+   15,  6,  2,  2, 15, 15, 15, 15, 15, 15, 25, 25, 25, 22, 18,  6,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 25, 25, 25, 22, 18,  2,
+    6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  2,  2,  2,
+   23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 11, 11, 11,
+   11, 12, 11, 11, 11, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   26, 26,  9, 26, 26, 26, 26,  9, 26, 26,  5,  9,  9,  9,  5,  5,
+    9,  9,  9,  5, 26,  9, 26, 26, 25,  9,  9,  9,  9,  9, 26, 26,
+   26, 26, 26, 26,  9, 26,  9, 26,  9, 26,  9,  9,  9,  9, 26,  5,
+    9,  9,  9,  9,  5,  7,  7,  7,  7,  5, 26, 26,  5,  5,  9,  9,
+   25, 25, 25, 25, 25,  9,  5,  5,  5,  5, 26, 25, 26, 26,  5, 26,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14,  9,  5, 14, 14, 14, 14, 15, 26, 26,  2,  2,  2,  2,
+   25, 25, 25, 25, 25, 26, 26, 26, 26, 26, 25, 25, 26, 26, 26, 26,
+   25, 26, 26, 25, 26, 26, 25, 26, 26, 26, 26, 26, 26, 26, 25, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 25, 25,
+   26, 26, 25, 26, 25, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   26, 26, 26, 26, 26, 26, 26, 26, 22, 18, 22, 18, 26, 26, 26, 26,
+   25, 25, 26, 26, 26, 26, 26, 26, 26, 22, 18, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 25, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 25, 25, 25, 25,
+   25, 25, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 15, 15, 15, 15, 15, 15,
+   26, 26, 26, 26, 26, 26, 26, 25, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 25, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 25, 25, 25, 25, 25, 25, 25, 25,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 25,
+   26, 26, 26, 26, 26, 26, 26, 26, 22, 18, 22, 18, 22, 18, 22, 18,
+   22, 18, 22, 18, 22, 18, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   25, 25, 25, 25, 25, 22, 18, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 22, 18, 22, 18, 22, 18, 22, 18, 22, 18,
+   25, 25, 25, 22, 18, 22, 18, 22, 18, 22, 18, 22, 18, 22, 18, 22,
+   18, 22, 18, 22, 18, 22, 18, 22, 18, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 22, 18, 22, 18, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 22, 18, 25, 25,
+   25, 25, 25, 25, 25, 26, 26, 25, 25, 25, 25, 25, 25, 26, 26, 26,
+   26, 26, 26, 26,  2,  2, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26,  2, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,
+    9,  5,  9,  9,  9,  5,  5,  9,  5,  9,  5,  9,  5,  9,  9,  9,
+    9,  5,  9,  5,  5,  9,  5,  5,  5,  5,  5,  5,  6,  6,  9,  9,
+    9,  5,  9,  5,  5, 26, 26, 26, 26, 26, 26,  9,  5,  9,  5, 12,
+   12, 12,  9,  5,  2,  2,  2,  2,  2, 21, 21, 21, 21, 15, 21, 21,
+    5,  5,  5,  5,  5,  5,  2,  5,  2,  2,  2,  2,  2,  5,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  6,
+   21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 12,
+    7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  2,
+   21, 21, 20, 19, 20, 19, 21, 21, 21, 20, 19, 21, 20, 19, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 17, 21, 21, 17, 21, 20, 19, 21, 21,
+   20, 19, 22, 18, 22, 18, 22, 18, 22, 18, 21, 21, 21, 21, 21,  6,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 17, 17, 21, 21, 21, 21,
+   17, 21, 22, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   26, 26, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2, 26, 26, 26, 26, 26,
+   26, 26, 26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,
+   29, 21, 21, 21, 26,  6,  7, 14, 22, 18, 22, 18, 22, 18, 22, 18,
+   22, 18, 26, 26, 22, 18, 22, 18, 22, 18, 22, 18, 17, 22, 18, 18,
+   26, 14, 14, 14, 14, 14, 14, 14, 14, 14, 12, 12, 12, 12, 10, 10,
+   17,  6,  6,  6,  6,  6, 26, 26, 14, 14, 14,  6,  7, 21, 26, 26,
+    7,  7,  7,  7,  7,  7,  7,  2,  2, 12, 12, 24, 24,  6,  6,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 21,  6,  6,  6,  7,
+    2,  2,  2,  2,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+   26, 26, 15, 15, 15, 15, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 15, 15, 15, 15, 15, 15, 15, 15,
+   26, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    7,  7,  7,  7,  7,  6,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  6, 21, 21, 21,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  7,  7,  2,  2,  2,  2,
+    9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  7, 12,
+   11, 11, 11, 21, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 21,  6,
+    9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  6,  6, 12, 12,
+    7,  7,  7,  7,  7,  7, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   12, 12, 21, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,
+   24, 24, 24, 24, 24, 24, 24,  6,  6,  6,  6,  6,  6,  6,  6,  6,
+   24, 24,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    5,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    6,  5,  5,  5,  5,  5,  5,  5,  5,  9,  5,  9,  5,  9,  9,  5,
+    9,  5,  9,  5,  9,  5,  9,  5,  6, 24, 24,  9,  5,  9,  5,  7,
+    9,  5,  9,  5,  5,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  9,  9,  9,  9,  5,
+    9,  9,  9,  9,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,  9,  5,
+    2,  2,  9,  5,  9,  9,  9,  9,  5,  9,  5,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  9,  5,  7,  6,  6,  5,  7,  7,  7,  7,  7,
+    7,  7, 12,  7,  7,  7, 12,  7,  7,  7,  7, 12,  7,  7,  7,  7,
+    7,  7,  7, 10, 10, 12, 12, 10, 26, 26, 26, 26, 12,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 26, 26, 23, 26,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,
+   10, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+   10, 10, 10, 10, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2, 21, 21,
+   12, 12,  7,  7,  7,  7,  7,  7, 21, 21, 21,  7, 21,  7,  7, 12,
+    7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12, 12, 12, 21, 21,
+    7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 10, 10,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 21,
+    7,  7,  7, 12, 10, 10, 12, 12, 12, 12, 10, 10, 12, 12, 10, 10,
+   10, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  6,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2, 21, 21,
+    7,  7,  7,  7,  7, 12,  6,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  7,  7,  7,  7,  7,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12, 10,
+   10, 12, 12, 10, 10, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7, 12,  7,  7,  7,  7,  7,  7,  7,  7, 12, 10,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2, 21, 21, 21, 21,
+    6,  7,  7,  7,  7,  7,  7, 26, 26, 26,  7, 10, 12, 10,  7,  7,
+   12,  7, 12, 12, 12,  7,  7, 12, 12,  7,  7,  7,  7,  7, 12, 12,
+    7, 12,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  7,  7,  6, 21, 21,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 10, 12, 12, 10, 10,
+   21, 21,  7,  6,  6, 10, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  7,  7,  7,  7,  7,  7,  2,  2,  7,  7,  7,  7,  7,  7,  2,
+    2,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 24,  6,  6,  6,  6,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  6, 24, 24,  2,  2,  2,  2,
+    7,  7,  7, 10, 10, 12, 10, 10, 12, 10, 10, 21, 10, 12,  2,  2,
+    7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  7,  7,  7,  7,  7,
+    4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    5,  5,  5,  5,  5,  5,  5,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  5,  5,  5,  5,  5,  2,  2,  2,  2,  2,  7, 12,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7, 25,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  2,  7,  2,
+    7,  7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24, 24,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 18, 22,
+    2,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 23, 26,  2,  2,
+   21, 21, 21, 21, 21, 21, 21, 22, 18, 21,  2,  2,  2,  2,  2,  2,
+   21, 17, 17, 16, 16, 22, 18, 22, 18, 22, 18, 22, 18, 22, 18, 22,
+   18, 22, 18, 22, 18, 21, 21, 22, 18, 21, 21, 21, 21, 16, 16, 16,
+   21, 21, 21,  2, 21, 21, 21, 21, 17, 22, 18, 22, 18, 22, 18, 21,
+   21, 21, 25, 17, 25, 25, 25,  2, 21, 23, 21, 21,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  1,
+    2, 21, 21, 21, 23, 21, 21, 21, 22, 18, 21, 25, 21, 17, 21, 21,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 22, 25, 18, 25, 22,
+   18, 21, 22, 18, 21, 21,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    6,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  6,  6,
+    2,  2,  7,  7,  7,  7,  7,  7,  2,  2,  7,  7,  7,  7,  7,  7,
+    2,  2,  7,  7,  7,  7,  7,  7,  2,  2,  7,  7,  7,  2,  2,  2,
+   23, 23, 25, 24, 26, 23, 23,  2, 26, 25, 25, 25, 25, 26, 26,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  1,  1,  1, 26, 26,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  2,  7,
+   21, 21, 21,  2,  2,  2,  2, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15,  2,  2,  2, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   14, 14, 14, 14, 14, 15, 15, 15, 15, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 15, 15, 26, 26, 26,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,
+   26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 12,  2,  2,
+   12, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,
+   15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,  2,  7,  7,  7,
+    7, 14,  7,  7,  7,  7,  7,  7,  7,  7, 14,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2, 21,
+    7,  7,  7,  7,  2,  2,  2,  2,  7,  7,  7,  7,  7,  7,  7,  7,
+   21, 14, 14, 14, 14, 14,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    9,  9,  9,  9,  9,  9,  9,  9,  5,  5,  5,  5,  5,  5,  5,  5,
+    9,  9,  9,  9,  2,  2,  2,  2,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2,  2,  2,
+    7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 21,
+    7,  7,  7,  7,  7,  7,  2,  2,  7,  2,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  2,  7,  7,  2,  2,  2,  7,  2,  2,  7,
+    7,  7,  7,  7,  7,  7,  2, 21, 15, 15, 15, 15, 15, 15, 15, 15,
+    7,  7,  7,  7,  7,  7,  7, 26, 26, 15, 15, 15, 15, 15, 15, 15,
+    2,  2,  2,  2,  2,  2,  2, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    7,  7,  7,  2,  7,  7,  2,  2,  2,  2,  2, 15, 15, 15, 15, 15,
+    7,  7,  7,  7,  7,  7, 15, 15, 15, 15, 15, 15,  2,  2,  2, 21,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2, 21,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2, 15, 15,  7,  7,
+    2,  2, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    7, 12, 12, 12,  2, 12, 12,  2,  2,  2,  2,  2, 12, 12, 12, 12,
+    7,  7,  7,  7,  2,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  2,  2, 12, 12, 12,  2,  2,  2,  2, 12,
+   15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,
+   21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 15, 15, 21,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 15, 15, 15,
+    7,  7,  7,  7,  7,  7,  7,  7, 26,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7, 12, 12,  2,  2,  2,  2, 15, 15, 15, 15, 15,
+   21, 21, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  2,  2,  2, 21, 21, 21, 21, 21, 21, 21,
+    7,  7,  7,  7,  7,  7,  2,  2, 15, 15, 15, 15, 15, 15, 15, 15,
+    7,  7,  7,  2,  2,  2,  2,  2, 15, 15, 15, 15, 15, 15, 15, 15,
+    7,  7,  2,  2,  2,  2,  2,  2,  2, 21, 21, 21, 21,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2, 15, 15, 15, 15, 15, 15, 15,
+    9,  9,  9,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    5,  5,  5,  2,  2,  2,  2,  2,  2,  2, 15, 15, 15, 15, 15, 15,
+    7,  7,  7,  7, 12, 12, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2, 12, 12, 17,  2,  2,
+    7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15,  7,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 15, 15, 15, 15, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,
+   10, 12, 10,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 21, 21, 21, 21, 21, 21, 21,  2,  2,
+   15, 15, 15, 15, 15, 15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 12,
+   10, 10, 10, 12, 12, 12, 12, 10, 10, 12, 12, 21, 21,  1, 21, 21,
+   21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  1,  2,  2,
+   12, 12, 12,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 10, 12, 12, 12,
+   12, 12, 12, 12, 12,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   21, 21, 21, 21,  7, 10, 10,  7,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7, 12, 21, 21,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7, 10, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12, 12, 10,
+   10,  7,  7,  7,  7, 21, 21, 21, 21, 12, 12, 12, 12, 21, 10, 12,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  7, 21,  7, 21, 21, 21,
+    2, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 10, 10, 10, 12,
+   12, 12, 10, 10, 12, 10, 12, 12, 21, 21, 21, 21, 21, 21, 12,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  2,  7,  7,  7,  7,  2,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7, 21,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12,
+   10, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12,  2,  2,  2,  2,  2,
+   12, 12, 10, 10,  2,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  7,
+    7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  2, 12, 12,  7, 10, 10,
+   12, 10, 10, 10, 10,  2,  2, 10, 10,  2,  2, 10, 10, 10,  2,  2,
+    7,  2,  2,  2,  2,  2,  2, 10,  2,  2,  2,  2,  2,  7,  7,  7,
+    7,  7, 10, 10,  2,  2, 12, 12, 12, 12, 12, 12, 12,  2,  2,  2,
+   12, 12, 12, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7, 10, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12,
+   10, 10, 12, 12, 12, 10, 12,  7,  7,  7,  7, 21, 21, 21, 21, 21,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 21, 21,  2, 21, 12,  7,
+   10, 10, 10, 12, 12, 12, 12, 12, 12, 10, 12, 10, 10, 10, 10, 12,
+   12, 10, 12, 12,  7,  7, 21,  7,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 10,
+   10, 10, 12, 12, 12, 12,  2,  2, 10, 10, 10, 10, 12, 12, 10, 12,
+   12, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21,  7,  7,  7,  7, 12, 12,  2,  2,
+   10, 10, 10, 12, 12, 12, 12, 12, 12, 12, 12, 10, 10, 12, 10, 12,
+   12, 21, 21, 21,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 10, 12, 10, 10,
+   12, 12, 12, 12, 12, 12, 10, 12,  7,  2,  2,  2,  2,  2,  2,  2,
+   10, 10, 12, 12, 12, 12, 10, 12, 12, 12, 12, 12,  2,  2,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 15, 15, 21, 21, 21, 26,
+   12, 12, 12, 12, 12, 12, 12, 12, 10, 12, 12, 21,  2,  2,  2,  2,
+   15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  7,
+    7,  7,  7,  7,  7,  7,  7,  2,  2,  7,  2,  2,  7,  7,  7,  7,
+    7,  7,  7,  7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,
+   10, 10, 10, 10, 10, 10,  2, 10, 10,  2,  2, 12, 12, 10, 12,  7,
+   10,  7, 10, 12, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  7,  7,  7,  7,  7,  7,
+    7, 10, 10, 10, 12, 12, 12, 12,  2,  2, 12, 12, 10, 10, 10, 10,
+   12,  7, 21,  7, 10,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,  7,  7,  7,  7,  7,
+    7,  7,  7, 12, 12, 12, 12, 12, 12, 10,  7, 12, 12, 12, 12, 21,
+   21, 21, 21, 21, 21, 21, 21, 12,  2,  2,  2,  2,  2,  2,  2,  2,
+    7, 12, 12, 12, 12, 12, 12, 10, 10, 12, 12, 12,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 10, 12, 12, 21, 21, 21,  7, 21, 21,
+   21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 12, 12, 12, 12, 12,  2, 12, 12, 12, 12, 12, 12, 10, 12,
+    7, 21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   21, 21,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    2,  2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12,  2, 10, 12, 12, 12, 12, 12, 12,
+   12, 10, 12, 12, 10, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  2,  7,  7,  7,  7,  7,
+    7, 12, 12, 12, 12, 12, 12,  2,  2,  2, 12,  2, 12, 12,  2, 12,
+   12, 12, 12, 12, 12, 12,  7, 12,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  2,  7,  7,  2,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 10, 10, 10, 10, 10,  2,
+   12, 12,  2, 10, 10, 12, 10, 12,  7,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7, 12, 12, 10, 10, 21, 21,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 26, 26, 26, 26, 26, 26, 26, 26, 23, 23, 23,
+   23, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 21,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,  2,
+   21, 21, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    1,  1,  1,  1,  1,  1,  1,  1,  1,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 12, 12, 12, 21,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   12, 12, 12, 12, 12, 12, 12, 21, 21, 21, 21, 21, 26, 26, 26, 26,
+    6,  6,  6,  6, 21, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2, 15, 15, 15, 15, 15,
+   15, 15,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2,  2,  7,  7,  7,
+   15, 15, 15, 15, 15, 15, 15, 21, 21, 21, 21,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2,  2,  2, 12,
+    7, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+   10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
+   10, 10, 10, 10, 10, 10, 10, 10,  2,  2,  2,  2,  2,  2,  2, 12,
+   12, 12, 12,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,
+    6,  6, 21,  6, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   10, 10,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  7,  7,  7,  7,  2,  2,  2,  2,  2,  2,  2,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  2, 26, 12, 12, 21,
+    1,  1,  1,  1,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26,  2,  2, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 10, 10, 12, 12, 12, 26, 26, 26, 10, 10, 10,
+   10, 10, 10,  1,  1,  1,  1,  1,  1,  1,  1, 12, 12, 12, 12, 12,
+   12, 12, 12, 26, 26, 12, 12, 12, 12, 12, 12, 12, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 12, 12, 12, 12, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 12, 12, 12, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  5,  5,
+    5,  5,  5,  5,  5,  2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    9,  9,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  9,  2,  9,  9,
+    2,  2,  9,  2,  2,  9,  9,  2,  2,  9,  9,  9,  9,  2,  9,  9,
+    9,  9,  9,  9,  9,  9,  5,  5,  5,  5,  2,  5,  2,  5,  5,  5,
+    5,  5,  5,  5,  2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  9,  9,  2,  9,  9,  9,  9,  2,  2,  9,  9,  9,
+    9,  9,  9,  9,  9,  2,  9,  9,  9,  9,  9,  9,  9,  2,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  9,  9,  2,  9,  9,  9,  9,  2,
+    9,  9,  9,  9,  9,  2,  9,  2,  2,  2,  9,  9,  9,  9,  9,  9,
+    9,  2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  2,  2,  9,  9,  9,  9,  9,  9,  9,  9,
+    9, 25,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 25,  5,  5,  5,  5,
+    5,  5,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 25,  5,  5,  5,  5,
+    5,  5,  5,  5,  5, 25,  5,  5,  5,  5,  5,  5,  9,  9,  9,  9,
+    9,  9,  9,  9,  9, 25,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 25,
+    5,  5,  5,  5,  5,  5,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 25,
+    5,  5,  5,  5,  5,  5,  5,  5,  5, 25,  5,  5,  5,  5,  5,  5,
+    9,  9,  9,  9,  9,  9,  9,  9,  9, 25,  5,  5,  5,  5,  5,  5,
+    5,  5,  5, 25,  5,  5,  5,  5,  5,  5,  9,  5,  2,  2, 13, 13,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   12, 12, 12, 12, 12, 12, 12, 26, 26, 26, 26, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 26, 26, 26,
+   26, 26, 26, 26, 26, 12, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 12, 26, 26, 21, 21, 21, 21, 21,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12,  2, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12,  2,  2, 12, 12, 12, 12, 12,
+   12, 12,  2, 12, 12,  2, 12, 12, 12, 12, 12,  2,  2,  2,  2,  2,
+   12, 12, 12, 12, 12, 12, 12,  6,  6,  6,  6,  6,  6,  6,  2,  2,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2,  7, 26,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 12, 12, 12, 12,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2,  2, 23,
+    7,  7,  7,  7,  7,  2,  2, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   12, 12, 12, 12, 12, 12, 12,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    5,  5,  5,  5, 12, 12, 12, 12, 12, 12, 12,  6,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 26, 15, 15, 15,
+   23, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 26, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,
+    7,  7,  7,  7,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    2,  7,  7,  2,  7,  2,  2,  7,  2,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  2,  7,  7,  7,  7,  2,  7,  2,  7,  2,  2,  2,  2,
+    2,  2,  7,  2,  2,  2,  2,  7,  2,  7,  2,  7,  2,  7,  7,  7,
+    2,  7,  7,  2,  7,  2,  2,  7,  2,  7,  2,  7,  2,  7,  2,  7,
+    2,  7,  7,  2,  7,  2,  2,  7,  7,  7,  7,  2,  7,  7,  7,  7,
+    7,  7,  7,  2,  7,  7,  7,  7,  2,  7,  7,  7,  7,  2,  7,  2,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,
+    2,  7,  7,  7,  2,  7,  7,  7,  7,  7,  2,  7,  7,  7,  7,  7,
+   25, 25,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2,  2,
+    2,  2,  2,  2,  2,  2, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 24, 24, 24, 24, 24,
+   26, 26, 26, 26, 26, 26, 26, 26,  2,  2,  2,  2,  2,  2,  2,  2,
+   26, 26, 26, 26, 26, 26, 26, 26, 26,  2, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,  2, 26, 26, 26,
+   26, 26, 26, 26, 26,  2,  2,  2, 26, 26, 26,  2,  2,  2,  2,  2,
+   26, 26, 26,  2, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+    2,  1,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  2,  2,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  0,  0,
+    0,  0,  0,  0,  0,  0,  1,  0,  0,  2,  0,  3,  4,  5,  6,  7,
+    8,  9, 10, 11, 12, 12, 12, 13, 14, 12, 15, 16, 17, 18, 19, 20,
+   21, 22,  0,  0,  0,  0, 23,  0,  0,  0,  0,  0,  0,  0, 24, 25,
+    0, 26, 27,  0, 28, 29, 30, 31, 32, 33,  0, 34,  0,  0,  0,  0,
+    0, 35,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0, 36, 37, 38,  0,  0,  0,  0,
+   39, 40,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 41, 42,  0,  0,
+   43, 44, 45, 46,  0, 47,  0, 48,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 49,  0,  0,  0,  0,  0, 50,  0,  0,  0,
+    0,  0,  0, 51,  0, 52, 53,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 54, 55,  0,  0,  0,  0, 56,  0,  0, 57, 58,  0,
+   59, 60, 61, 62, 63, 64, 65,  0, 66, 67,  0, 68, 69, 70, 71,  0,
+   60,  0, 72, 73, 74, 75,  0,  0, 69,  0, 76, 77,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 78, 79,  0,  0,  0,  0,  0,  0,  0,  0, 80,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0, 81,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 82, 83, 84,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   85,  0, 79,  0,  0, 86,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 87, 88,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  1,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11,
+   12,  1,  0,  0, 13,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 14, 15, 16, 17, 18, 19, 20,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  1, 21,  0,  0,  0,  0,  0, 22, 23, 24,
+    0,  0, 25,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 26, 27,
+   28, 29,  0,  0,  0,  0, 30,  0,  0,  0, 31, 32, 33, 34,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 13, 35, 36,  0,  0, 26, 37, 38, 39,  0,  0,  0,  0,  0, 40,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 41,  1,
+   42, 43, 44, 45,  0,  0,  0,  0,  0,  0,  0, 46,  0, 47, 48,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 46,  0, 47,  0,  0,
+    0,  0,  0, 49,  0,  0,  0,  0,  0,  0,  0, 46,  0, 47,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 47,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 47, 50,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 51,  0, 47,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 52,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 53,  0, 54,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 55,  0, 56,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 57,  0,  0, 58, 59,  0,  0,  0,  0,
+    0,  0, 60, 61, 62,  0,  0,  0,  0,  0,  0,  0, 63,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 64, 65,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 36,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 66,
+    0,  0,  0,  0,  0,  0, 67,  0,  0,  0, 67,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 52, 68,
+    0,  0,  0,  0,  0,  0,  0,  0,  0, 69,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 70,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 71, 72,  0,  0,  0,  0,  0,  0,  0,  0,
+   73,  0, 66, 74,  0,  0,  0,  0,  0,  0, 75, 76, 72,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 46,  0, 67,  0,  0,  0,
+    0, 77, 78,  0,  0,  0,  0,  0,  0, 79,  0,  0,  0,  0,  0,  0,
+   80,  0, 79,  0,  0,  0,  0,  0,  0,  0, 64,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 81, 82,
+   83, 84, 85, 86,  0,  0,  0,  0,  0,  0,  0,  0, 87, 88, 89,  1,
+    1,  1, 90, 91,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 92, 93,
+   94, 95, 96,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 71, 86,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 97,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    1,  1,  1,  1,  0,  0,  0,  0,  0, 98,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 99,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 71,100,101,  0,  0,  0, 26,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 86,  0,102,  0,  0,  0,  0, 67,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 67,  0,  0,  0,
+    1,  1, 86,  0,  0,  0,  0,  0,  0,103,  0,  0,  0,  0,104,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,105,  0, 73,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,106,107,108,  0,  0,  0,
+    0,  0,102,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 47,  0,  0,  0,  0,  0,109,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,110,111,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 36,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   72,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 26,112,  0,113,  0,  0,  0,  0,  0,114,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+  115,  0,  0,  0,  0,  0,  0,  0,100,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,116,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,117,118, 72,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,102,  0,  0,  0,
+    0,  0,  0, 97,  0,  0,  0,  0,  0,  0,  0,119,  0,  0,  0,  0,
+    0,  0,  0,  0,112,  0,  0,  0,  0,  0, 51,  0,  0,  0,  0,  0,
+    0,  0,105,  0,  0,  0,  0,  0,  0,  0,  0,  0, 73,120,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,121,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,122,  0,  0,  0,  0,  0,  0,  0,  0,  0,123,  0, 47,  0,  0,
+   26,124,124,  0,  0,  0,  0,  0,  0,  0,  0,  0,125,  0,  0, 49,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,126,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 97,127,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 97,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,128,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,104,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,129,105,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   73,  0,  0,  0,  0,  0,  0,  0,  0,  0, 67,  0, 97,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,130,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,131,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 97,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,132,  0,  0,  0,  0,  0,  0,  0,133,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,134,  0,  0,  0,  0,135,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+  136,137,138,139,140,141,  0,  0,  0,142,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,143,  0,  0,  0,
+    0,  0,  0,  0,133,  1,  1,144,145,112,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,100,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,146,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,100,147,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,230,230,230,230,
+  230,230,230,230,230,230,230,230,230,232,220,220,220,220,232,216,
+  220,220,220,220,220,202,202,220,220,220,220,202,202,220,220,220,
+  220,220,220,220,220,220,220,220,  1,  1,  1,  1,  1,220,220,220,
+  220,230,230,230,230,230,230,230,230,240,230,220,220,220,230,230,
+  230,220,220,  0,230,230,230,220,220,220,220,230,232,220,220,230,
+  233,234,234,233,234,234,233,230,230,230,230,230,  0,  0,  0,230,
+  230,230,230,230,  0,220,230,230,230,230,220,230,230,230,222,220,
+  230,230,230,230,230,230,220,220,220,220,220,220,230,230,220,230,
+  230,222,228,230, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 19, 20,
+   21, 22,  0, 23,  0, 24, 25,  0,230,220,  0, 18, 30, 31, 32,  0,
+    0,  0,  0,  0,  0,  0,  0, 27, 28, 29, 30, 31, 32, 33, 34,230,
+  230,220,220,230,230,230,230,230,220,230,230,220, 35,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,230,230,230,230,230,230,
+  230,  0,  0,230,230,230,230,220,230,  0,  0,230,230,  0,220,230,
+  230,220,  0,  0,  0, 36,  0,  0,  0,  0,  0,  0,230,220,230,230,
+  220,230,230,220,220,220,230,220,220,230,220,230,230,230,220,230,
+  220,230,220,230,220,230,230,  0,  0,  0,  0,  0,230,230,220,230,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,220,  0,  0,230,230,  0,230,
+  230,230,230,230,230,230,230,230,  0,230,230,230,  0,230,230,230,
+  230,230,  0,  0,  0,220,220,220,  0,  0,  0,  0,  0,  0,  0,220,
+  230,230,230,230,230,230,  0,220,230,230,220,230,230,220,230,230,
+  230,220,220,220, 27, 28, 29,230,230,230,220,230,230,220,220,230,
+  230,230,230,230,  0,  0,  0,  0,  7,  0,  0,  0,  0,  0,  0,  0,
+    0,  9,  0,  0,  0,230,220,230,230,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,230,  0,  0,  0,  0,  0,  0, 84, 91,  0,  0,  0,  0,  9,
+    9,  0,  0,  0,  0,  0,  9,  0,  0,  0,  0,  0,103,103,  9,  0,
+    0,  0,  0,  0,107,107,107,107,  0,  0,  0,  0,118,118,  9,  0,
+    0,  0,  0,  0,122,122,122,122,  0,  0,  0,  0,220,220,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,220,  0,220,  0,216,  0,  0,
+    0,  0,  0,  0,  0,129,130,  0,132,  0,  0,  0,  0,  0,130,130,
+  130,130,  0,  0,130,  0,230,230,  9,  0,230,230,  0,  0,  0,  0,
+    0,  0,220,  0,  0,  0,  0,  0,  0,  0,  0,  7,  0,  9,  9,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,230,230,230,  0,  0,  0,  0,
+    9,  0,  0,  0,  0,  0,  0,  0,  0,230,  0,  0,  0,228,  0,  0,
+    0,  0,  0,  0,  0,222,230,220,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,230,220,  0,  0,  0,  0,  0,  0,  0,  9,  0,  0,  0,
+    0,  0,  0,  0,230,230,230,230,230,  0,  0,220,230,230,230,230,
+  230,220,220,220,220,220,220,230,230,220,  0,220,  0,  0,  0,230,
+  220,230,230,230,230,230,230,230,  0,  0,  0,  0,  0,  0,  9,  9,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  7,  0,230,230,230,  0,
+    1,220,220,220,220,220,230,230,220,220,220,220,230,  0,  1,  1,
+    1,  1,  1,  1,  1,  0,  0,  0,  0,220,  0,  0,  0,  0,  0,  0,
+  230,  0,  0,  0,230,230,  0,  0,  0,  0,  0,  0,230,230,220,230,
+  230,230,230,230,230,230,220,230,230,234,214,220,202,230,230,230,
+  230,230,230,230,230,230,230,230,230,230,232,228,228,220,  0,230,
+  233,220,230,220,230,230,  1,  1,230,230,230,230,  1,  1,  1,230,
+  230,  0,  0,  0,  0,230,  0,  0,  0,  1,  1,230,220,230,  1,  1,
+  220,220,220,220,230,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  9,  0,  0,218,228,232,222,224,224,  0,  8,  8,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,230,230,230,230,230,230,230,230,
+  230,230,  0,  0,  0,  0,  0,  0,  0,  0,  9,  0,  0,  0,  0,220,
+  220,220,  0,  0,  0,  0,  0,  9,  0,  0,  0,  0,  0,  0,  0,  7,
+    0,  0,  0,  0,230,  0,230,230,220,  0,  0,230,230,  0,  0,  0,
+    0,  0,230,230,  0,230,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 26,  0,230,230,230,230,230,230,230,220,220,220,220,220,
+  220,220,230,230,230,230,230,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,220,  0,230,230,  1,220,  0,  0,  0,  0,  9,  0,  0,  0,  0,
+    0,230,220,  0,  0,  0,  0,230,230,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,220,220,230,230,230,220,230,220,220,220,  0,  9,  7,  0,
+    0,  0,  0,  0,  0,  0,  7,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  9,  7,  0,  0,  7,  9,  0,  0,  0,  0,  0,  0,  0,  0,  7,
+    7,  0,  0,  0,230,230,230,230,230,  0,  0,  0,  0,  0,  9,  0,
+    0,  0,  7,  0,  0,  0,  9,  7,  0,  0,  0,  0,  7,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  9,  7,  0,  0,  0,  0,
+    0,  9,  9,  0,  0,  9,  0,  0,  0,  0,  0,  0,  0,  0,  7,  0,
+    9,  9,  0,  0,  1,  1,  1,  1,  1,  0,  0,  0,230,230,230,230,
+  230,230,230,  0,  6,  6,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  1,  0,  0,  0,  0,  0,  0,216,216,  1,  1,  1,  0,  0,
+    0,226,216,216,216,216,216,  0,  0,  0,  0,  0,  0,  0,  0,220,
+  220,220,220,220,220,220,220,  0,  0,230,230,230,230,230,220,220,
+    0,  0,  0,  0,  0,  0,230,230,230,230,  0,  0,  0,  0,230,230,
+  230,  0,  0,  0,230,  0,  0,230,230,230,230,230,230,230,  0,230,
+  230,  0,230,230,220,220,220,220,220,220,220,  0,230,230,  7,  0,
+    0,  0,  0,  0, 16, 17, 17, 17, 17, 17, 17, 33, 17, 17, 17, 19,
+   17, 17, 17, 17, 20,101, 17,113,129,169, 17, 27, 28, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17,237,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,
+    0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  3,  4,  0,  0,  0,  0,
+    0,  0,  3,  4,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  5,  0,
+    0,  0,  6,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  7,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  8,  9,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 10,  0,  0, 10,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 10,  0,  0,  0, 10,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 11, 12,  0, 13,  0, 14, 15, 16,  0,  0,
+    0,  0,  0,  1, 17, 18,  0, 19,  7,  1,  0,  0,  0, 20, 20,  7,
+   20, 20, 20, 20, 20, 20, 20,  8, 21,  0, 22,  0,  7, 23, 24,  0,
+   20, 20, 25,  0,  0,  0, 26, 27,  1,  7, 20, 20, 20, 20, 20,  1,
+   28, 29, 30, 31,  0,  0, 20,  0,  0,  0,  0,  0,  0,  0, 10,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 20, 20,
+   20,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  8, 21, 32,  4,  0, 10,  0, 33,  7, 20, 20, 20,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  8, 34, 34, 35, 36, 34, 37,  0, 38,  1, 20, 20,
+    0,  0, 39,  0,  1,  1,  0,  8, 21,  1, 20,  0,  0,  0,  1,  0,
+    0, 40,  1,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  8, 21,
+    0,  1,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0,
+    0,  0, 26, 34, 34, 34, 34, 34, 34, 34, 34, 34, 21,  7, 20, 41,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 21,  0, 42, 43, 44,  0, 45,
+    0,  8, 21,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 46,  7,  1, 10,  1,  0,  0,  0,  1, 20, 20,  1,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 20, 20,  1, 20, 20,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 26, 21,  0,  1,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  0,  2,  0,  0,  0,  0,
+    0,  0,  3,  4,  0,  0,  0,  0,  0,  0,  3, 47, 48,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,
+    0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  3,  4,  0,  0,  0,  0,
+    0,  0,  3,  4,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11,
+   12, 13, 14, 15, 16, 17, 18, 17, 19, 20, 21, 22, 23, 24, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 26, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 27, 28, 28, 29, 30, 31, 32, 33, 33, 33, 33, 33, 33,
+   33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 34,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51,
+   52, 53, 54, 55, 56, 57, 58, 35, 35, 35, 35, 35, 59, 59, 60, 35,
+   35, 35, 35, 35, 35, 35, 61, 62, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 63, 64, 35, 65, 66, 66, 66, 66,
+   66, 66, 66, 66, 66, 66, 66, 67, 66, 68, 69, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 70, 71, 35, 35,
+   35, 35, 72, 35, 35, 35, 35, 35, 35, 35, 35, 35, 73, 74, 75, 76,
+   77, 78, 35, 35, 79, 80, 35, 35, 81, 35, 82, 83, 84, 85, 17, 86,
+   87, 88, 35, 35, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 89, 25, 25, 25, 25, 25, 25, 25, 90,
+   91, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 92, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 93, 35, 35, 35, 35, 35, 35,
+   25, 94, 35, 35, 25, 25, 25, 25, 25, 25, 25, 25, 25, 95, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 96,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   19,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,  0, 19, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,  0,
+    0,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19,  0,  0,  0,  0,  0,
+   26, 26,  0,  0,  0,  0,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+    1,  1,  1,  1,  1,  1,  9,  9,  9,  9,  0,  9,  9,  9,  2,  2,
+    9,  9,  9,  9,  0,  9,  2,  2,  2,  2,  9,  0,  9,  0,  9,  9,
+    9,  2,  9,  2,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  2,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9, 55, 55, 55, 55, 55, 55, 55, 55,
+   55, 55, 55, 55, 55, 55,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,
+    6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  1,  1,  6,  6,  6,
+    6,  6,  6,  6,  6,  6,  2,  4,  4,  4,  4,  4,  4,  4,  4,  4,
+    4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,
+    4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  2,  2,  4,
+    4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,
+    4,  2,  2,  4,  4,  4,  2, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,  2,  2,
+    2,  2,  2,  2,  2,  2, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14,  2,  2,  2,  2, 14, 14, 14, 14, 14, 14,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  0,  3,  3,  3,  3,
+    3,  3,  0,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  0,  3,  2,  3,  0,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  0,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  1,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  0,  3,  3, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37,
+   37, 37, 37, 37,  2, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37,
+   37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37,
+   37,  2,  2, 37, 37, 37, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38,
+   38, 38, 38, 38, 38, 38, 38, 38,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64,
+   64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64, 64,
+   64,  2,  2, 64, 64, 64, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90,
+   90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90,
+   90, 90, 90, 90,  2,  2, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90,
+   90, 90, 90, 90, 90,  2, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95,
+   95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95,
+   95, 95,  2,  2, 95,  2, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37,
+   37,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  2,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  0,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  1,  1,  1,  1,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  0,  0,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  5,  5,  5,  5,  2,  5,  5,  5,  5,  5,
+    5,  5,  5,  2,  2,  5,  5,  2,  2,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,
+    5,  5,  5,  5,  5,  5,  5,  2,  5,  2,  2,  2,  5,  5,  5,  5,
+    2,  2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2,  5,  5,  2,
+    2,  5,  5,  5,  5,  2,  2,  2,  2,  2,  2,  2,  2,  5,  2,  2,
+    2,  2,  5,  5,  2,  5,  5,  5,  5,  5,  2,  2,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,
+    5,  5,  5,  5,  5,  2,  2, 11, 11, 11,  2, 11, 11, 11, 11, 11,
+   11,  2,  2,  2,  2, 11, 11,  2,  2, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,  2,
+   11, 11, 11, 11, 11, 11, 11,  2, 11, 11,  2, 11, 11,  2, 11, 11,
+    2,  2, 11,  2, 11, 11, 11, 11, 11,  2,  2,  2,  2, 11, 11,  2,
+    2, 11, 11, 11,  2,  2,  2, 11,  2,  2,  2,  2,  2,  2,  2, 11,
+   11, 11, 11,  2, 11,  2,  2,  2,  2,  2,  2,  2, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2, 10, 10, 10,  2, 10, 10, 10, 10, 10,
+   10, 10, 10, 10,  2, 10, 10, 10,  2, 10, 10, 10, 10, 10, 10, 10,
+   10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  2,
+   10, 10, 10, 10, 10, 10, 10,  2, 10, 10,  2, 10, 10, 10, 10, 10,
+    2,  2, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  2, 10, 10, 10,
+    2, 10, 10, 10,  2,  2, 10,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 10, 10, 10, 10,  2,  2, 10, 10, 10, 10,
+   10, 10, 10, 10, 10, 10, 10, 10,  2,  2,  2,  2,  2,  2,  2, 10,
+   10, 10, 10, 10, 10, 10,  2, 21, 21, 21,  2, 21, 21, 21, 21, 21,
+   21, 21, 21,  2,  2, 21, 21,  2,  2, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,
+   21, 21, 21, 21, 21, 21, 21,  2, 21, 21,  2, 21, 21, 21, 21, 21,
+    2,  2, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  2, 21, 21,  2,
+    2, 21, 21, 21,  2,  2,  2,  2,  2,  2,  2, 21, 21, 21,  2,  2,
+    2,  2, 21, 21,  2, 21, 21, 21, 21, 21,  2,  2, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2, 22, 22,  2, 22, 22, 22, 22, 22,
+   22,  2,  2,  2, 22, 22, 22,  2, 22, 22, 22, 22,  2,  2,  2, 22,
+   22,  2, 22,  2, 22, 22,  2,  2,  2, 22, 22,  2,  2,  2, 22, 22,
+   22,  2,  2,  2, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22,
+    2,  2,  2,  2, 22, 22, 22, 22, 22,  2,  2,  2, 22, 22, 22,  2,
+   22, 22, 22, 22,  2,  2, 22,  2,  2,  2,  2,  2,  2, 22,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 22, 22, 22, 22,
+   22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22,
+   22,  2,  2,  2,  2,  2, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,
+   23, 23, 23,  2, 23, 23, 23,  2, 23, 23, 23, 23, 23, 23, 23, 23,
+   23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,  2,
+   23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,
+    2,  2,  2, 23, 23, 23, 23, 23, 23, 23, 23,  2, 23, 23, 23,  2,
+   23, 23, 23, 23,  2,  2,  2,  2,  2,  2,  2, 23, 23,  2, 23, 23,
+   23,  2,  2,  2,  2,  2, 23, 23, 23, 23,  2,  2, 23, 23, 23, 23,
+   23, 23, 23, 23, 23, 23,  2,  2,  2,  2,  2,  2,  2, 23, 23, 23,
+   23, 23, 23, 23, 23, 23, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16,  2, 16, 16, 16,  2, 16, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,  2,
+   16, 16, 16, 16, 16, 16, 16, 16, 16, 16,  2, 16, 16, 16, 16, 16,
+    2,  2, 16, 16, 16, 16, 16, 16, 16, 16, 16,  2, 16, 16, 16,  2,
+   16, 16, 16, 16,  2,  2,  2,  2,  2,  2,  2, 16, 16,  2,  2,  2,
+    2,  2,  2,  2, 16,  2, 16, 16, 16, 16,  2,  2, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 16,  2, 16, 16,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,
+   20, 20, 20,  2, 20, 20, 20,  2, 20, 20, 20, 20, 20, 20, 20, 20,
+   20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,
+   20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,  2, 20, 20, 20,  2,
+   20, 20, 20, 20, 20, 20,  2,  2,  2,  2, 20, 20, 20, 20, 20, 20,
+   20, 20, 20, 20, 20, 20, 20, 20, 20, 20,  2,  2, 20, 20, 20, 20,
+   20, 20, 20, 20, 20, 20,  2, 36, 36, 36,  2, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,  2,  2,  2,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36,  2, 36, 36, 36, 36, 36, 36, 36,
+   36, 36,  2, 36,  2,  2, 36, 36, 36, 36, 36, 36, 36,  2,  2,  2,
+   36,  2,  2,  2,  2, 36, 36, 36, 36, 36, 36,  2, 36,  2, 36, 36,
+   36, 36, 36, 36, 36, 36,  2,  2,  2,  2,  2,  2, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36,  2,  2, 36, 36, 36,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24,  2,  2,  2,  2,  0, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24, 24,  2,  2,  2,  2,  2, 18, 18,  2, 18,  2, 18, 18, 18, 18,
+   18,  2, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+   18, 18, 18, 18, 18, 18, 18, 18, 18, 18,  2, 18,  2, 18, 18, 18,
+   18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+   18, 18, 18, 18,  2,  2, 18, 18, 18, 18, 18,  2, 18,  2, 18, 18,
+   18, 18, 18, 18,  2,  2, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+    2,  2, 18, 18, 18, 18, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,  2, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25,  2,  2,  2,  2, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25,  2, 25, 25, 25, 25, 25, 25, 25,  0,  0,  0,  0, 25,
+   25,  2,  2,  2,  2,  2, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33,
+   33, 33, 33, 33, 33, 33,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  2,  8,  2,  2,
+    2,  2,  2,  8,  2,  2,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    8,  0,  8,  8,  8,  8, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,
+   30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,  2,
+   30, 30, 30, 30,  2,  2, 30, 30, 30, 30, 30, 30, 30,  2, 30,  2,
+   30, 30, 30, 30,  2,  2, 30,  2, 30, 30, 30, 30,  2,  2, 30, 30,
+   30, 30, 30, 30, 30,  2, 30,  2, 30, 30, 30, 30,  2,  2, 30, 30,
+   30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,  2, 30, 30,
+   30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,
+   30,  2,  2, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,
+   30, 30, 30,  2,  2,  2, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30,
+    2,  2,  2,  2,  2,  2, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,
+   29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,  2,  2, 29, 29,
+   29, 29, 29, 29,  2,  2, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28,
+   28, 28, 28, 28, 28, 28, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34,  2,  2,  2, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35,  0,  0,  0, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,  2,
+    2,  2,  2,  2,  2,  2, 45, 45, 45, 45, 45, 45, 45, 45, 45, 45,
+   45, 45, 45,  2, 45, 45, 45, 45, 45, 45, 45,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 44, 44, 44, 44, 44, 44, 44, 44, 44, 44,
+   44, 44, 44, 44, 44, 44, 44, 44, 44, 44, 44,  0,  0,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43,
+   43, 43, 43, 43, 43, 43, 43, 43, 43, 43,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46,
+   46, 46, 46,  2, 46, 46, 46,  2, 46, 46,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31,
+   31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31,
+   31, 31, 31, 31,  2,  2, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31,
+    2,  2,  2,  2,  2,  2, 32, 32,  0,  0, 32,  0, 32, 32, 32, 32,
+   32, 32, 32, 32, 32,  2, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+    2,  2,  2,  2,  2,  2, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,  2,
+    2,  2,  2,  2,  2,  2, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32,  2,  2,  2,  2,  2, 28, 28, 28, 28, 28, 28,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48,
+   48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48,
+   48, 48, 48, 48, 48,  2, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48,
+   48, 48,  2,  2,  2,  2, 48,  2,  2,  2, 48, 48, 48, 48, 48, 48,
+   48, 48, 48, 48, 48, 48, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52,
+   52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52,
+   52, 52, 52, 52,  2,  2, 52, 52, 52, 52, 52,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58,
+   58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58,
+   58, 58,  2,  2,  2,  2, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58,
+    2,  2,  2,  2,  2,  2, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58,
+   58,  2,  2,  2, 58, 58, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54,
+   54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54,
+   54, 54,  2,  2, 54, 54, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+   91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+   91, 91, 91, 91, 91,  2, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+   91, 91, 91,  2,  2, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+    2,  2,  2,  2,  2,  2, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+   91, 91, 91, 91,  2,  2,  1,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62,  2,  2,  2,  2, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62, 62,  2,  2,  2, 76, 76, 76, 76, 76, 76, 76, 76, 76, 76,
+   76, 76, 76, 76, 76, 76, 93, 93, 93, 93, 93, 93, 93, 93, 93, 93,
+   93, 93, 93, 93, 93, 93, 93, 93, 93, 93,  2,  2,  2,  2,  2,  2,
+    2,  2, 93, 93, 93, 93, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,  2,  2,
+    2, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,
+    2,  2,  2, 70, 70, 70, 73, 73, 73, 73, 73, 73, 73, 73, 73, 73,
+   73, 73, 73, 73, 73, 73,  6,  6,  6,  6,  6,  6,  6,  6,  6,  2,
+    2,  2,  2,  2,  2,  2,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    8,  2,  2,  8,  8,  8, 76, 76, 76, 76, 76, 76, 76, 76,  2,  2,
+    2,  2,  2,  2,  2,  2,  1,  1,  1,  0,  1,  1,  1,  1,  1,  1,
+    1,  1,  1,  1,  1,  1,  1,  0,  1,  1,  1,  1,  1,  1,  1,  0,
+    0,  0,  0,  1,  0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  1,  1,
+    0,  2,  2,  2,  2,  2, 19, 19, 19, 19, 19, 19,  9,  9,  9,  9,
+    9,  6, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19,  9,  9,  9,  9,  9, 19, 19, 19, 19,  9,  9,  9,  9,
+    9, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,  6, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19,  9,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+    2,  1,  1,  1,  1,  1,  9,  9,  9,  9,  9,  9,  2,  2,  9,  9,
+    9,  9,  9,  9,  2,  2,  9,  9,  9,  9,  9,  9,  9,  9,  2,  9,
+    2,  9,  2,  9,  2,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  2,  2,  9,  9,  9,  9,  9,  2,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  2,  2,  9,  9,  9,  9,
+    9,  9,  2,  9,  9,  9,  2,  2,  9,  9,  9,  2,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  1,  1,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 19,  2,  2,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 19,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  2, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19,  2,  2,  2,  0,  0,  0,  0,  0,  0,  9,  0,  0,  0,
+   19, 19,  0,  0,  0,  0,  0,  0, 19,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 19,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19,  0,
+    0,  0,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  2,  2,  2,  2,  2, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,
+   27, 27, 27, 27, 27, 27,  0,  0,  0,  0,  2,  2,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+   56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+   56, 56, 56, 56, 56,  2, 55, 55, 55, 55, 55, 55, 55, 55, 55, 55,
+   55, 55, 55, 55, 55, 55, 55, 55, 55, 55,  2,  2,  2,  2,  2, 55,
+   55, 55, 55, 55, 55, 55, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61,
+   61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61,  2,  2,
+    2,  2,  2,  2,  2, 61, 61,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 61, 30, 30, 30, 30, 30, 30, 30,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 30, 30, 30, 30, 30, 30, 30,  2, 30, 30,
+   30, 30, 30, 30, 30,  2,  0,  0,  0,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    2, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  2,  2,  2,  2,  0,  0,  0,  0,  0, 13,  0, 13,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    1,  1,  1,  1, 12, 12,  0,  0,  0,  0,  0,  0,  0,  0, 13, 13,
+   13, 13,  0,  0,  0,  0,  2, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  1,
+    1,  0,  0, 15, 15, 15,  0, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17,  0,  0, 17, 17, 17,  2,  2,  2,  2,  2, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26,  2, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12,  2,  0,  0,  0,  0,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12,  0, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17,  0, 17, 17, 17, 17, 17, 17, 17, 17,  0,  0,
+    0,  0,  0,  0,  0,  0, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   13, 13, 13,  2,  2,  2, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39,
+   39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39,
+   39, 39, 39,  2,  2,  2, 39, 39, 39, 39, 39, 39, 39,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 86, 86, 86, 86, 86, 86, 86, 86, 86, 86,
+   86, 86, 86, 86, 86, 86, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77,
+   77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77, 77,
+   77, 77,  2,  2,  2,  2, 79, 79, 79, 79, 79, 79, 79, 79, 79, 79,
+   79, 79, 79, 79, 79, 79, 79, 79, 79, 79, 79, 79, 79, 79,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,  0,  0,
+    0, 19, 19, 19, 19, 19,  2,  2, 19, 19, 19, 19, 19, 19, 19, 19,
+   19,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19, 19, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60,
+   60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60,
+   60, 60, 60,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    2,  2,  2,  2,  2,  2, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65,
+   65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65,  2,  2,
+    2,  2,  2,  2,  2,  2, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75,
+   75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75,  2,  2,  2,  2,
+    2,  2,  2,  2, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75,
+    2,  2,  2,  2,  2,  2, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69,
+   69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69,
+   69, 69, 69, 69,  0, 69, 74, 74, 74, 74, 74, 74, 74, 74, 74, 74,
+   74, 74, 74, 74, 74, 74, 74, 74, 74, 74,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 74, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12,  2,  2,  2, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84,
+   84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84,
+   84, 84, 84, 84,  2,  0, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84,
+    2,  2,  2,  2, 84, 84, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33,
+   33, 33, 33, 33, 33,  2, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68,
+   68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68,
+   68, 68, 68, 68,  2,  2, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68,
+    2,  2, 68, 68, 68, 68, 92, 92, 92, 92, 92, 92, 92, 92, 92, 92,
+   92, 92, 92, 92, 92, 92, 92, 92, 92,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2, 92, 92, 92, 92, 92, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2, 30, 30, 30, 30, 30, 30,  2,  2, 30,
+   30, 30, 30, 30, 30,  2,  2, 30, 30, 30, 30, 30, 30,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19,  9, 19, 19, 19, 19,
+    0,  0,  2,  2,  2,  2, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+   87, 87, 87, 87,  2,  2, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+    2,  2,  2,  2,  2,  2, 12, 12, 12, 12,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 12, 12, 12, 12, 12, 12, 12,  2,  2,  2,
+    2, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   13, 13, 13, 13,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    2,  2,  2,  2,  2,  2, 19, 19, 19, 19, 19, 19, 19,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  4,  4,  4,  4,  4,  2,  2,
+    2,  2,  2, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,  2, 14, 14,
+   14, 14, 14,  2, 14,  2, 14, 14,  2, 14, 14,  2, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14,  3,  3,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  0,  0,  2,  2,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  2,  2,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+    1,  1,  1,  1,  6,  6,  0,  0,  0,  2,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,
+    0,  0,  2,  2,  2,  2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  2,  2,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17,  0,  0,  2,  2, 12, 12, 12, 12, 12, 12,  2,  2,
+   12, 12, 12, 12, 12, 12,  2,  2, 12, 12, 12, 12, 12, 12,  2,  2,
+   12, 12, 12,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,
+    0,  0,  0,  0,  0,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  0,
+    0,  0,  0,  0,  2,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49, 49,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,  2, 49, 49,
+   49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49,  2, 49, 49,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49, 49, 49, 49,  2,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49,  2,  2,  2,  2,  2,  0,  0,  0,  2,  2,  2,  2,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,  2,  2,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  2,  2,  2,  9,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  1,  2,  2, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71,
+   71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71,
+   71, 71, 71,  2,  2,  2, 67, 67, 67, 67, 67, 67, 67, 67, 67, 67,
+   67, 67, 67, 67, 67, 67, 67,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42,
+   42, 42, 42, 42, 42, 42, 42, 42, 42, 42,  2,  2,  2,  2,  2,  2,
+    2,  2,  2, 42, 42, 42, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41,
+   41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41,
+   41,  2,  2,  2,  2,  2,118,118,118,118,118,118,118,118,118,118,
+  118,118,118,118,118,118,118,118,118,118,118,118,118,118,118,118,
+  118,  2,  2,  2,  2,  2, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53,
+   53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53,
+   53, 53, 53, 53,  2, 53, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59,  2,  2,  2,  2, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59, 59,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40,
+   40, 40, 40, 40, 40, 40, 51, 51, 51, 51, 51, 51, 51, 51, 51, 51,
+   51, 51, 51, 51, 51, 51, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
+   50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
+   50, 50, 50, 50,  2,  2, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
+    2,  2,  2,  2,  2,  2,135,135,135,135,135,135,135,135,135,135,
+  135,135,135,135,135,135,135,135,135,135,  2,  2,  2,  2,135,135,
+  135,135,135,135,135,135,135,135,135,135,135,135,135,135,135,135,
+  135,135,  2,  2,  2,  2,106,106,106,106,106,106,106,106,106,106,
+  106,106,106,106,106,106,106,106,106,106,106,106,106,106,  2,  2,
+    2,  2,  2,  2,  2,  2,104,104,104,104,104,104,104,104,104,104,
+  104,104,104,104,104,104,104,104,104,104,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,104,110,110,110,110,110,110,110,110,110,110,
+  110,110,110,110,110,110,110,110,110,110,110,110,110,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,110,110,110,110,110,110,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,110,110,110,110,110,110,110,110,  2,  2,
+    2,  2,  2,  2,  2,  2, 47, 47, 47, 47, 47, 47,  2,  2, 47,  2,
+   47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47,
+   47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47,  2, 47, 47,  2,
+    2,  2, 47,  2,  2, 47, 81, 81, 81, 81, 81, 81, 81, 81, 81, 81,
+   81, 81, 81, 81, 81, 81, 81, 81, 81, 81, 81, 81,  2, 81, 81, 81,
+   81, 81, 81, 81, 81, 81,120,120,120,120,120,120,120,120,120,120,
+  120,120,120,120,120,120,116,116,116,116,116,116,116,116,116,116,
+  116,116,116,116,116,116,116,116,116,116,116,116,116,116,116,116,
+  116,116,116,116,116,  2,  2,  2,  2,  2,  2,  2,  2,116,116,116,
+  116,116,116,116,116,116,128,128,128,128,128,128,128,128,128,128,
+  128,128,128,128,128,128,128,128,128,  2,128,128,  2,  2,  2,  2,
+    2,128,128,128,128,128, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,
+   66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,
+   66, 66,  2,  2,  2, 66, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72,
+   72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72,
+    2,  2,  2,  2,  2, 72, 98, 98, 98, 98, 98, 98, 98, 98, 98, 98,
+   98, 98, 98, 98, 98, 98, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+   97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,  2,  2,
+    2,  2, 97, 97, 97, 97,  2,  2, 97, 97, 97, 97, 97, 97, 97, 97,
+   97, 97, 97, 97, 97, 97, 57, 57, 57, 57,  2, 57, 57,  2,  2,  2,
+    2,  2, 57, 57, 57, 57, 57, 57, 57, 57,  2, 57, 57, 57,  2, 57,
+   57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57,
+   57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57,  2,  2, 57, 57,
+   57,  2,  2,  2,  2, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57,  2,
+    2,  2,  2,  2,  2,  2, 88, 88, 88, 88, 88, 88, 88, 88, 88, 88,
+   88, 88, 88, 88, 88, 88,117,117,117,117,117,117,117,117,117,117,
+  117,117,117,117,117,117,112,112,112,112,112,112,112,112,112,112,
+  112,112,112,112,112,112,112,112,112,112,112,112,112,  2,  2,  2,
+    2,112,112,112,112,112,112,112,112,112,112,112,112,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 78, 78, 78, 78, 78, 78, 78, 78, 78, 78,
+   78, 78, 78, 78, 78, 78, 78, 78, 78, 78, 78, 78,  2,  2,  2, 78,
+   78, 78, 78, 78, 78, 78, 83, 83, 83, 83, 83, 83, 83, 83, 83, 83,
+   83, 83, 83, 83, 83, 83, 83, 83, 83, 83, 83, 83,  2,  2, 83, 83,
+   83, 83, 83, 83, 83, 83, 82, 82, 82, 82, 82, 82, 82, 82, 82, 82,
+   82, 82, 82, 82, 82, 82, 82, 82, 82,  2,  2,  2,  2,  2, 82, 82,
+   82, 82, 82, 82, 82, 82,122,122,122,122,122,122,122,122,122,122,
+  122,122,122,122,122,122,122,122,  2,  2,  2,  2,  2,  2,  2,122,
+  122,122,122,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,122,
+  122,122,122,122,122,122, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89,
+   89, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89, 89,  2,
+    2,  2,  2,  2,  2,  2,130,130,130,130,130,130,130,130,130,130,
+  130,130,130,130,130,130,130,130,130,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,130,130,130,  2,  2,  2,  2,  2,  2,  2,
+  130,130,130,130,130,130,144,144,144,144,144,144,144,144,144,144,
+  144,144,144,144,144,144,144,144,144,144,144,144,144,144,  2,  2,
+    2,  2,  2,  2,  2,  2,144,144,144,144,144,144,144,144,144,144,
+    2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  2,156,156,156,156,156,156,156,156,156,156,
+  156,156,156,156,156,156,156,156,156,156,156,156,156,156,156,156,
+    2,156,156,156,  2,  2,156,156,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,147,147,147,147,147,147,147,147,147,147,
+  147,147,147,147,147,147,147,147,147,147,147,147,147,147,  2,  2,
+    2,  2,  2,  2,  2,  2,148,148,148,148,148,148,148,148,148,148,
+  148,148,148,148,148,148,148,148,148,148,148,148,148,148,148,148,
+    2,  2,  2,  2,  2,  2,153,153,153,153,153,153,153,153,153,153,
+  153,153,153,153,153,153,153,153,153,153,153,153,153,153,153,153,
+  153,153,  2,  2,  2,  2,149,149,149,149,149,149,149,149,149,149,
+  149,149,149,149,149,149,149,149,149,149,149,149,149,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94,
+   94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94,
+   94, 94, 94, 94,  2,  2,  2,  2, 94, 94, 94, 94, 94, 94, 94, 94,
+   94, 94, 94, 94, 94, 94,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 94, 85, 85, 85, 85, 85, 85, 85, 85, 85, 85,
+   85, 85, 85, 85, 85, 85, 85, 85,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2, 85,  2,  2,101,101,101,101,101,101,101,101,101,101,
+  101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,  2,
+    2,  2,  2,  2,  2,  2,101,101,101,101,101,101,101,101,101,101,
+    2,  2,  2,  2,  2,  2, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,
+   96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,  2, 96, 96, 96, 96,
+   96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,  2,  2,
+    2,  2,  2,  2,  2,  2,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,100,100,100,100,100,100,100,100,100,100,
+  100,100,100,100,100,100,  2, 36, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,108,108,108,108,108,108,108,108,108,108,
+  108,108,108,108,108,108,108,108,  2,108,108,108,108,108,108,108,
+  108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,
+  108,108,108,108,108,  2,129,129,129,129,129,129,129,  2,129,  2,
+  129,129,129,129,  2,129,129,129,129,129,129,129,129,129,129,129,
+  129,129,129,129,  2,129,129,129,129,129,129,129,129,129,129,129,
+    2,  2,  2,  2,  2,  2,109,109,109,109,109,109,109,109,109,109,
+  109,109,109,109,109,109,109,109,109,109,109,109,109,109,109,109,
+  109,  2,  2,  2,  2,  2,109,109,109,109,109,109,109,109,109,109,
+    2,  2,  2,  2,  2,  2,107,107,107,107,  2,107,107,107,107,107,
+  107,107,107,  2,  2,107,107,  2,  2,107,107,107,107,107,107,107,
+  107,107,107,107,107,107,107,107,107,107,107,107,107,107,107,  2,
+  107,107,107,107,107,107,107,  2,107,107,  2,107,107,107,107,107,
+    2,  1,107,107,107,107,107,107,107,107,107,  2,  2,107,107,  2,
+    2,107,107,107,  2,  2,107,  2,  2,  2,  2,  2,  2,107,  2,  2,
+    2,  2,  2,107,107,107,107,107,107,107,  2,  2,107,107,107,107,
+  107,107,107,  2,  2,  2,107,107,107,107,107,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,137,137,137,137,137,137,137,137,137,137,
+  137,137,137,137,137,137,137,137,137,137,137,137,137,137,137,137,
+  137,137,  2,137,137,137,137,137,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,124,124,124,124,124,124,124,124,124,124,
+  124,124,124,124,124,124,124,124,124,124,124,124,124,124,  2,  2,
+    2,  2,  2,  2,  2,  2,124,124,124,124,124,124,124,124,124,124,
+    2,  2,  2,  2,  2,  2,123,123,123,123,123,123,123,123,123,123,
+  123,123,123,123,123,123,123,123,123,123,123,123,  2,  2,123,123,
+  123,123,123,123,123,123,123,123,123,123,123,123,123,123,123,123,
+  123,123,123,123,  2,  2,114,114,114,114,114,114,114,114,114,114,
+  114,114,114,114,114,114,114,114,114,114,114,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,114,114,114,114,114,114,114,114,114,114,
+    2,  2,  2,  2,  2,  2, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32,  2,  2,  2,102,102,102,102,102,102,102,102,102,102,
+  102,102,102,102,102,102,102,102,102,102,102,102,102,102,102,  2,
+    2,  2,  2,  2,  2,  2,102,102,102,102,102,102,102,102,102,102,
+    2,  2,  2,  2,  2,  2,126,126,126,126,126,126,126,126,126,126,
+  126,126,126,126,126,126,126,126,126,126,126,126,126,126,126,126,
+  126,  2,  2,126,126,126,126,126,126,126,126,126,126,126,126,126,
+  126,126,  2,  2,  2,  2,142,142,142,142,142,142,142,142,142,142,
+  142,142,142,142,142,142,142,142,142,142,142,142,142,142,142,142,
+  142,142,  2,  2,  2,  2,125,125,125,125,125,125,125,125,125,125,
+  125,125,125,125,125,125,125,125,125,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,125,154,154,154,154,154,154,154,  2,  2,154,
+    2,  2,154,154,154,154,154,154,154,154,  2,154,154,  2,154,154,
+  154,154,154,154,154,154,154,154,154,154,154,154,154,154,154,154,
+  154,154,154,154,154,154,154,154,154,154,154,154,  2,154,154,  2,
+    2,154,154,154,154,154,154,154,154,154,154,154,154,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,154,154,154,154,154,154,154,154,154,154,
+    2,  2,  2,  2,  2,  2,150,150,150,150,150,150,150,150,  2,  2,
+  150,150,150,150,150,150,150,150,150,150,150,150,150,150,150,150,
+  150,150,150,150,150,150,150,150,150,150,150,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,141,141,141,141,141,141,141,141,141,141,
+  141,141,141,141,141,141,141,141,141,141,141,141,141,141,  2,  2,
+    2,  2,  2,  2,  2,  2,140,140,140,140,140,140,140,140,140,140,
+  140,140,140,140,140,140,140,140,140,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,121,121,121,121,121,121,121,121,121,121,
+  121,121,121,121,121,121,121,121,121,121,121,121,121,121,121,  2,
+    2,  2,  2,  2,  2,  2,133,133,133,133,133,133,133,133,133,  2,
+  133,133,133,133,133,133,133,133,133,133,133,133,133,133,133,133,
+  133,133,133,133,133,133,133,133,133,133,133,133,133,  2,133,133,
+  133,133,133,133,133,133,133,133,133,133,133,133,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,133,133,133,133,133,133,133,133,133,133,
+  133,133,133,  2,  2,  2,134,134,134,134,134,134,134,134,134,134,
+  134,134,134,134,134,134,  2,  2,134,134,134,134,134,134,134,134,
+  134,134,134,134,134,134,134,134,134,134,134,134,134,134,  2,134,
+  134,134,134,134,134,134,134,134,134,134,134,134,134,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,138,138,138,138,138,138,138,  2,138,138,
+    2,138,138,138,138,138,138,138,138,138,138,138,138,138,138,138,
+  138,138,138,138,138,138,138,138,138,138,138,138,138,  2,  2,  2,
+  138,  2,138,138,  2,138,138,138,138,138,138,138,138,138,  2,  2,
+    2,  2,  2,  2,  2,  2,138,138,138,138,138,138,138,138,138,138,
+    2,  2,  2,  2,  2,  2,143,143,143,143,143,143,  2,143,143,  2,
+  143,143,143,143,143,143,143,143,143,143,143,143,143,143,143,143,
+  143,143,143,143,143,143,143,143,143,143,143,143,143,143,143,143,
+  143,143,143,143,143,  2,143,143,  2,143,143,143,143,143,143,  2,
+    2,  2,  2,  2,  2,  2,143,143,143,143,143,143,143,143,143,143,
+    2,  2,  2,  2,  2,  2,145,145,145,145,145,145,145,145,145,145,
+  145,145,145,145,145,145,145,145,145,145,145,145,145,145,145,  2,
+    2,  2,  2,  2,  2,  2, 86,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 22, 22, 22, 22, 22, 22, 22, 22, 22, 22,
+   22, 22, 22, 22, 22, 22, 22, 22,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 22, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63,
+   63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63,
+    2,  2,  2,  2,  2,  2, 63, 63, 63, 63, 63, 63, 63, 63, 63, 63,
+   63, 63, 63, 63, 63,  2, 63, 63, 63, 63, 63,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 63, 63, 63, 63,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80,
+   80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80,
+   80, 80, 80, 80, 80,  2, 80, 80, 80, 80, 80, 80, 80, 80, 80,  2,
+    2,  2,  2,  2,  2,  2,127,127,127,127,127,127,127,127,127,127,
+  127,127,127,127,127,127,127,127,127,127,127,127,127,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 79, 79, 79, 79, 79, 79, 79, 79, 79,  2,
+    2,  2,  2,  2,  2,  2,115,115,115,115,115,115,115,115,115,115,
+  115,115,115,115,115,115,115,115,115,115,115,115,115,115,115,115,
+  115,115,115,115,115,  2,115,115,115,115,115,115,115,115,115,115,
+    2,  2,  2,  2,115,115,103,103,103,103,103,103,103,103,103,103,
+  103,103,103,103,103,103,103,103,103,103,103,103,103,103,103,103,
+  103,103,103,103,  2,  2,103,103,103,103,103,103,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,119,119,119,119,119,119,119,119,119,119,
+  119,119,119,119,119,119,119,119,119,119,119,119,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,119,119,119,119,119,119,119,119,119,119,
+    2,119,119,119,119,119,119,119,  2,119,119,119,119,119,119,119,
+  119,119,119,119,119,119,119,119,119,119,119,119,119,119,  2,  2,
+    2,  2,  2,119,119,119,146,146,146,146,146,146,146,146,146,146,
+  146,146,146,146,146,146,146,146,146,146,146,146,146,146,146,146,
+  146,  2,  2,  2,  2,  2, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
+   99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
+   99,  2,  2,  2,  2, 99, 99, 99, 99, 99, 99, 99, 99, 99,  2,  2,
+    2,  2,  2,  2,  2, 99,136,139,  0,  0,155,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 13, 13,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,136,136,136,136,136,136,136,136,136,136,
+  136,136,136,136,136,136,136,136,136,136,136,136,136,136,  2,  2,
+    2,  2,  2,  2,  2,  2,155,155,155,155,155,155,155,155,155,155,
+  155,155,155,155,155,155,155,155,155,155,155,155,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,136,136,136,136,136,136,136,136,136,  2,
+    2,  2,  2,  2,  2,  2, 17, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15,  2, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 17, 17, 17, 17,  2,  2,
+    2,  2,  2,  2,  2,  2,139,139,139,139,139,139,139,139,139,139,
+  139,139,139,139,139,139,139,139,139,139,139,139,139,139,139,139,
+  139,139,  2,  2,  2,  2,105,105,105,105,105,105,105,105,105,105,
+  105,105,105,105,105,105,105,105,105,105,105,105,105,105,105,105,
+  105,  2,  2,  2,  2,  2,105,105,105,105,105,105,105,105,105,105,
+  105,105,105,  2,  2,  2,105,105,105,105,105,105,105,105,105,  2,
+    2,  2,  2,  2,  2,  2,105,105,105,105,105,105,105,105,105,105,
+    2,  2,105,105,105,105,  0,  0,  0,  0,  0,  0,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  2,  2,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  1,  1,  1,  1,  1,  1,  1,  1,  0,  0,  1,  1,  1,  1,  1,
+    1,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    1,  1,  1,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,
+    2,  2,  2,  2,  2,  2,  9,  9,  9,  9,  9,  9,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  2,  0,  0,  2,  2,  0,  2,  2,  0,  0,  2,  2,  0,
+    0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    2,  0,  2,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,
+    0,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,
+    0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    2,  0,  0,  0,  0,  2,  0,  0,  0,  0,  0,  2,  0,  2,  2,  2,
+    0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,  2,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  2,  2,  0,  0,131,131,131,131,131,131,131,131,131,131,
+  131,131,131,131,131,131,131,131,131,131,131,131,131,131,131,131,
+  131,131,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,131,131,131,131,131,  2,131,131,131,131,131,131,131,131,131,
+  131,131,131,131,131,131, 56, 56, 56, 56, 56, 56, 56,  2, 56, 56,
+   56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56,  2,
+    2, 56, 56, 56, 56, 56, 56, 56,  2, 56, 56,  2, 56, 56, 56, 56,
+   56,  2,  2,  2,  2,  2,151,151,151,151,151,151,151,151,151,151,
+  151,151,151,151,151,151,151,151,151,151,151,151,151,151,151,151,
+  151,151,151,  2,  2,  2,151,151,151,151,151,151,151,151,151,151,
+  151,151,151,151,  2,  2,151,151,151,151,151,151,151,151,151,151,
+    2,  2,  2,  2,151,151,152,152,152,152,152,152,152,152,152,152,
+  152,152,152,152,152,152,152,152,152,152,152,152,152,152,152,152,
+    2,  2,  2,  2,  2,152,113,113,113,113,113,113,113,113,113,113,
+  113,113,113,113,113,113,113,113,113,113,113,  2,  2,113,113,113,
+  113,113,113,113,113,113,113,113,113,113,113,113,113,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,132,132,132,132,132,132,132,132,132,132,
+  132,132,132,132,132,132,132,132,132,132,132,132,132,132,132,132,
+  132,132,  2,  2,  2,  2,132,132,132,132,132,132,132,132,132,132,
+    2,  2,  2,  2,132,132,  0,  0,  0,  0,  0,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  2,  2,  3,  3,  3,  3,  2,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  2,  3,  3,  2,  3,  2,  2,  3,  2,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  2,  3,  3,  3,  3,  2,  3,
+    2,  3,  2,  2,  2,  2,  2,  2,  3,  2,  2,  2,  2,  3,  2,  3,
+    2,  3,  2,  3,  3,  3,  2,  3,  3,  2,  3,  2,  2,  3,  2,  3,
+    2,  3,  2,  3,  2,  3,  2,  3,  3,  2,  3,  2,  2,  3,  3,  3,
+    3,  2,  3,  3,  3,  3,  3,  3,  3,  2,  3,  3,  3,  3,  2,  3,
+    3,  3,  3,  2,  3,  2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  2,  2,  2,  2,  2,  3,  3,  3,  2,  3,  3,  3,  3,  3,
+    2,  3,  3,  3,  3,  3,  2,  2,  2,  2,  2,  2,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 15,  0,  0,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  2,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,  2,  2,  2,  0,  0,
+    0,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 13,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+   13,  2,  2,  2,  2,  2,  2,  0,  2,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  1,  2,  3,  4,  5,  6,  0,
+    0,  0,  0,  7,  8,  9, 10, 11,  0, 12,  0,  0,  0,  0, 13,  0,
+    0, 14,  0,  0,  0,  0,  0,  0,  0,  0, 15, 16,  0, 17, 18, 19,
+    0,  0,  0, 20, 21, 22,  0, 23,  0, 24,  0, 25,  0, 26,  0,  0,
+    0,  0,  0, 27, 28,  0, 29,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 30, 31,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 32, 33,
+   34, 35, 36, 37, 38, 39, 40,  0,  0,  0, 41,  0, 42, 43, 44, 45,
+   46, 47, 48,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 49,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 50, 51, 52,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 53, 54, 55, 56, 57, 58,
+   59, 60, 61, 62,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 63,  0, 64,  0,  0,  0,  0,  0,
+    0,  0,  0, 65,  0,  0,  0,  0, 66,  0,  0,  0, 67,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 68,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 69, 70, 71,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 72, 73, 74, 75, 76, 77, 78, 79, 80,  0,
+};
+static const uint16_t
+_hb_ucd_u16[11328] =
+{
+     0,   0,   1,   2,   3,   4,   5,   6,   0,   0,   7,   8,   9,  10,  11,  12,
+    13,  13,  13,  14,  15,  13,  13,  16,  17,  18,  19,  20,  21,  22,  13,  23,
+    13,  13,  13,  24,  25,  11,  11,  11,  11,  26,  11,  27,  28,  29,  30,  31,
+    32,  32,  32,  32,  32,  32,  32,  33,  34,  35,  36,  11,  37,  38,  13,  39,
+     9,   9,   9,  11,  11,  11,  13,  13,  40,  13,  13,  13,  41,  13,  13,  13,
+    13,  13,  13,  42,   9,  43,  11,  11,  44,  45,  32,  46,  47,  48,  49,  50,
+    51,  52,  48,  48,  53,  32,  54,  55,  48,  48,  48,  48,  48,  56,  57,  58,
+    59,  60,  48,  32,  61,  48,  48,  48,  48,  48,  62,  63,  64,  48,  65,  66,
+    48,  67,  68,  69,  48,  70,  71,  72,  72,  72,  48,  73,  74,  75,  76,  32,
+    77,  48,  48,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,
+    91,  84,  85,  92,  93,  94,  95,  96,  97,  98,  85,  99, 100, 101,  89, 102,
+   103,  84,  85, 104, 105, 106,  89, 107, 108, 109, 110, 111, 112, 113,  95, 114,
+   115, 116,  85, 117, 118, 119,  89, 120, 121, 116,  85, 122, 123, 124,  89, 125,
+   126, 116,  48, 127, 128, 129,  89, 130, 131, 132,  48, 133, 134, 135,  95, 136,
+   137,  48,  48, 138, 139, 140,  72,  72, 141,  48, 142, 143, 144, 145,  72,  72,
+   146, 147, 148, 149, 150,  48, 151, 152, 153, 154,  32, 155, 156, 157,  72,  72,
+    48,  48, 158, 159, 160, 161, 162, 163, 164, 165,   9,   9, 166,  11,  11, 167,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48, 168, 169,  48,  48, 168,  48,  48, 170, 171, 172,  48,  48,
+    48, 171,  48,  48,  48, 173, 174, 175,  48, 176,   9,   9,   9,   9,   9, 177,
+   178,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48, 179,  48, 180, 181,  48,  48,  48,  48, 182, 183,
+   184, 185,  48, 186,  48, 187, 184, 188,  48,  48,  48, 189, 190, 191, 192, 193,
+   194, 192,  48,  48, 195,  48,  48, 196, 197,  48, 198,  48,  48,  48,  48, 199,
+    48, 200, 201, 202, 203,  48, 204, 205,  48,  48, 206,  48, 207, 208, 209, 209,
+    48, 210,  48,  48,  48, 211, 212, 213, 192, 192, 214, 215, 216,  72,  72,  72,
+   217,  48,  48, 218, 219, 160, 220, 221, 222,  48, 223,  64,  48,  48, 224, 225,
+    48,  48, 226, 227, 228,  64,  48, 229, 230,   9,   9, 231, 232, 233, 234, 235,
+    11,  11, 236,  27,  27,  27, 237, 238,  11, 239,  27,  27,  32,  32,  32, 240,
+    13,  13,  13,  13,  13,  13,  13,  13,  13, 241,  13,  13,  13,  13,  13,  13,
+   242, 243, 242, 242, 243, 244, 242, 245, 246, 246, 246, 247, 248, 249, 250, 251,
+   252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 262,  72, 263, 264, 216,
+   265, 266, 267, 268, 269, 270, 271, 271, 272, 273, 274, 209, 275, 276, 209, 277,
+   278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278,
+   279, 209, 280, 209, 209, 209, 209, 281, 209, 282, 278, 283, 209, 284, 285, 209,
+   209, 209, 286,  72, 287,  72, 270, 270, 270, 288, 209, 209, 209, 209, 289, 270,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 290, 291, 209, 209, 292,
+   209, 209, 209, 209, 209, 209, 293, 209, 209, 209, 209, 209, 209, 209, 209, 209,
+   209, 209, 209, 209, 209, 209, 294, 295, 270, 296, 209, 209, 297, 278, 298, 278,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209,
+   278, 278, 278, 278, 278, 278, 278, 278, 299, 300, 278, 278, 278, 301, 278, 302,
+   278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278,
+   209, 209, 209, 278, 303, 209, 209, 304, 209, 305, 209, 209, 209, 209, 209, 209,
+     9,   9, 306,  11,  11, 307, 308, 309,  13,  13,  13,  13,  13,  13, 310, 311,
+    11,  11, 312,  48,  48,  48, 313, 314,  48, 315, 316, 316, 316, 316,  32,  32,
+   317, 318, 319, 320, 321, 322,  72,  72, 209, 323, 209, 209, 209, 209, 209, 324,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 325,  72, 326,
+   327, 328, 329, 330, 137,  48,  48,  48,  48, 331, 178,  48,  48,  48,  48, 332,
+   333,  48,  48, 137,  48,  48,  48,  48, 200, 334,  48,  48, 209, 209, 324,  48,
+   209, 335, 336, 209, 337, 338, 209, 209, 336, 209, 209, 338, 209, 209, 209, 209,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 209, 209, 209, 209,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 151,
+    48, 339,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48, 151, 209, 209, 209, 286,  48,  48, 229,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+   340,  48, 341,  72,  13,  13, 342, 343,  13, 344,  48,  48,  48,  48, 345, 346,
+    31, 347, 348, 349,  13,  13,  13, 350, 351, 352, 353, 354, 355,  72,  72, 356,
+   357,  48, 358, 359,  48,  48,  48, 360, 361,  48,  48, 362, 363, 192,  32, 364,
+    64,  48, 365,  48, 366, 367,  48, 151,  77,  48,  48, 368, 369, 370, 371, 372,
+    48,  48, 373, 374, 375, 376,  48, 377,  48,  48,  48, 378, 379, 380, 381, 382,
+   383, 384, 316,  11,  11, 385, 386,  11,  11,  11,  11,  11,  48,  48, 387, 192,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 388,  48, 389,  48,  48, 206,
+   390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390,
+   390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390, 390,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48, 204,  48,  48,  48,  48,  48,  48, 207,  72,  72,
+   392, 393, 394, 395, 396,  48,  48,  48,  48,  48,  48, 397, 398, 399,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48, 400,  72,  48,  48,  48,  48, 401,  48,  48,  74,  72,  72, 402,
+    32, 403,  32, 404, 405, 406, 407,  73,  48,  48,  48,  48,  48,  48,  48, 408,
+   409,   2,   3,   4,   5, 410, 411, 412,  48, 413,  48, 200, 414, 415, 416, 417,
+   418,  48, 172, 419, 204, 204,  72,  72,  48,  48,  48,  48,  48,  48,  48,  71,
+   420, 270, 270, 421, 271, 271, 271, 422, 423, 424, 425,  72,  72, 209, 209, 426,
+    72,  72,  72,  72,  72,  72,  72,  72,  48, 151,  48,  48,  48, 101, 427, 428,
+    48,  48, 429,  48, 430,  48,  48, 431,  48, 432,  48,  48, 433, 434,  72,  72,
+     9,   9, 435,  11,  11,  48,  48,  48,  48, 204, 192,   9,   9, 436,  11, 437,
+    48,  48,  74,  48,  48,  48, 438,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48, 315,  48, 199,  74,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   439,  48,  48, 440,  48, 441,  48, 442,  48, 200, 443,  72,  72,  72,  48, 444,
+    48, 445,  48, 446,  72,  72,  72,  72,  48,  48,  48, 447, 270, 448, 270, 270,
+   449, 450,  48, 451, 452, 453,  48, 454,  48, 455,  72,  72, 456,  48, 457, 458,
+    48,  48,  48, 459,  48, 460,  48, 461,  48, 462, 463,  72,  72,  72,  72,  72,
+    48,  48,  48,  48, 196,  72,  72,  72,   9,   9,   9, 464,  11,  11,  11, 465,
+    48,  48, 466, 192,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72, 270, 467,  48,  48, 468, 469,  72,  72,  72,  72,
+    48, 455, 470,  48,  62, 471,  72,  72,  72,  72,  72,  48, 472,  72,  48, 315,
+   473,  48,  48, 474, 475, 448, 476, 477, 222,  48,  48, 478, 479,  48, 196, 192,
+   480,  48, 481, 482, 483,  48,  48, 484, 222,  48,  48, 485, 486, 487, 488, 489,
+    48,  98, 490, 491,  72,  72,  72,  72, 492, 493, 494,  48,  48, 495, 496, 192,
+   497,  84,  85, 498, 499, 500, 501, 502,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48, 503, 504, 505, 469,  72,  48,  48,  48, 506, 507, 192,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  48,  48, 508, 509, 510, 511,  72,  72,
+    48,  48,  48, 512, 513, 192, 514,  72,  48,  48, 515, 516, 192,  72,  72,  72,
+    48, 173, 517, 518,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48, 490, 519,  72,  72,  72,  72,  72,  72,   9,   9,  11,  11, 148, 520,
+   521, 522,  48, 523, 524, 192,  72,  72,  72,  72, 525,  48,  48, 526, 527,  72,
+   528,  48,  48, 529, 530, 531,  48,  48, 532, 533, 534,  72,  48,  48,  48, 196,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    85,  48, 508, 535, 536, 148, 175, 537,  48, 538, 539, 540,  72,  72,  72,  72,
+   541,  48,  48, 542, 543, 192, 544,  48, 545, 546, 192,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  48, 547,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72, 101, 270, 548, 549, 550,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48, 207,  72,  72,  72,  72,  72,  72,
+   271, 271, 271, 271, 271, 271, 551, 552,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48, 388,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48, 200, 553,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48, 315,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48, 196,  48, 200, 370,  72,  72,  72,  72,  72,  72,  48, 204, 554,
+    48,  48,  48, 555, 556, 557, 558, 559,  48,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,   9,   9,  11,  11, 270, 560,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48, 561, 562, 563, 563, 564, 565,  72,  72,  72,  72, 566, 567,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  74,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 199,  72,  72,
+   196,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48, 200,  72,  72,  72, 568, 569,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 206,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48,  48,  48,  71, 151, 196, 570, 571,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 325,
+   209, 209, 572, 209, 209, 209, 573, 574, 575, 209, 576, 209, 209, 209, 577,  72,
+   209, 209, 209, 209, 578,  72,  72,  72,  72,  72,  72,  72,  72,  72, 270, 579,
+   209, 209, 209, 209, 209, 286, 270, 452,  72,  72,  72,  72,  72,  72,  72,  72,
+     9, 580,  11, 581, 582, 583, 242,   9, 584, 585, 586, 587, 588,   9, 580,  11,
+   589, 590,  11, 591, 592, 593, 594,   9, 595,  11,   9, 580,  11, 581, 582,  11,
+   242,   9, 584, 594,   9, 595,  11,   9, 580,  11, 596,   9, 597, 598, 599, 600,
+    11, 601,   9, 602, 603, 604, 605,  11, 606,   9, 607,  11, 608, 609, 609, 609,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209,
+    32,  32,  32, 610,  32,  32, 611, 612, 613, 614,  45,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   615, 616, 617,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48, 151, 618, 619,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  48,  48, 620, 621,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 622, 623,  72,  72,
+     9,   9, 584,  11, 624, 370,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72, 488, 270, 270, 625, 626,  72,  72,  72,  72,
+   488, 270, 627, 628,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   629,  48, 630, 631, 632, 633, 634, 635, 636, 206, 637, 206,  72,  72,  72, 638,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   209, 209, 326, 209, 209, 209, 209, 209, 209, 324, 335, 639, 639, 639, 209, 325,
+   640, 209, 209, 209, 209, 209, 209, 209, 209, 209, 641,  72,  72,  72, 642, 209,
+   643, 209, 209, 326, 577, 644, 325,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 645,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 209, 646, 424, 424,
+   209, 209, 209, 209, 209, 209, 209, 324, 209, 209, 209, 209, 209, 577, 326,  72,
+   326, 209, 209, 209, 646, 176, 209, 209, 646, 209, 641, 644,  72,  72,  72,  72,
+   209, 209, 209, 209, 209, 209, 209, 647, 209, 209, 209, 209, 648, 209, 209, 209,
+   209, 209, 209, 209, 209, 324, 641, 649, 286, 209, 577, 286, 643, 286,  72,  72,
+   209, 209, 209, 209, 209, 209, 209, 209, 209, 650, 209, 209, 287,  72,  72, 192,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 204,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48, 205,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48, 204,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 469,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 101,  72,
+    48, 204,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  71,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,
+   651,  72, 652, 652, 652, 652, 652, 652,  72,  72,  72,  72,  72,  72,  72,  72,
+    32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  32,  72,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 653,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391,
+   391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 391, 654,
+     0,   0,   0,   0,   1,   2,   1,   2,   0,   0,   3,   3,   4,   5,   4,   5,
+     4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,
+     4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   6,   0,   0,   7,   0,
+     8,   8,   8,   8,   8,   8,   8,   9,  10,  11,  12,  11,  11,  11,  13,  11,
+    14,  14,  14,  14,  14,  14,  14,  14,  15,  14,  14,  14,  14,  14,  14,  14,
+    14,  14,  14,  16,  17,  18,  17,  17,  19,  20,  21,  21,  22,  21,  23,  24,
+    25,  26,  27,  27,  28,  29,  27,  30,  27,  27,  27,  27,  27,  31,  27,  27,
+    32,  33,  33,  33,  34,  27,  27,  27,  35,  35,  35,  36,  37,  37,  37,  38,
+    39,  39,  40,  41,  42,  43,  44,  45,  45,  45,  27,  46,  47,  48,  49,  27,
+    50,  50,  50,  50,  50,  51,  52,  50,  53,  54,  55,  56,  57,  58,  59,  60,
+    61,  62,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,  73,  74,  75,  76,
+    77,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,  91,  92,
+    93,  94,  95,  96,  97,  98,  99, 100, 101, 102, 103, 104, 105, 106, 107, 108,
+   109, 110, 111, 111, 112, 113, 114, 111, 115, 116, 117, 118, 119, 120, 121, 122,
+   123, 124, 124, 125, 124, 126,  45,  45, 127, 128, 129, 130, 131, 132,  45,  45,
+   133, 133, 133, 133, 134, 133, 135, 136, 133, 134, 133, 137, 137, 138,  45,  45,
+   139, 139, 139, 139, 139, 139, 139, 139, 139, 139, 140, 140, 141, 140, 140, 142,
+   143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143,
+   144, 144, 144, 144, 145, 146, 144, 144, 145, 144, 144, 147, 148, 149, 144, 144,
+   144, 148, 144, 144, 144, 150, 144, 151, 144, 152, 153, 153, 153, 153, 153, 154,
+   155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155,
+   155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155, 155,
+   155, 155, 155, 155, 155, 155, 155, 155, 156, 157, 158, 158, 158, 158, 159, 160,
+   161, 162, 163, 164, 165, 166, 167, 168, 169, 169, 169, 169, 169, 170, 171, 171,
+   172, 173, 174, 174, 174, 174, 174, 175, 174, 174, 176, 155, 155, 155, 155, 177,
+   178, 179, 180, 180, 181, 182, 183, 184, 185, 185, 186, 185, 187, 188, 169, 169,
+   189, 190, 191, 191, 191, 192, 191, 193, 194, 194, 195,   8, 196,  45,  45,  45,
+   197, 197, 197, 197, 198, 197, 197, 199, 200, 200, 200, 200, 201, 201, 201, 202,
+   203, 203, 203, 204, 205, 206, 206, 206, 207, 140, 140, 208, 209, 210, 211, 212,
+     4,   4, 213,   4,   4, 214, 215, 216,   4,   4,   4, 217,   8,   8,   8, 218,
+     4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,   4,
+    11, 219,  11,  11, 219, 220,  11, 221,  11,  11,  11, 222, 222, 223,  11, 224,
+   225,   0,   0,   0,   0,   0, 226, 227, 228, 229,   0,   0,  45,   8,   8, 196,
+     0,   0, 230, 231, 232,   0,   4,   4, 233,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 234,  45, 235,  45,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236, 236,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0, 237,   0, 238,   0,   0,   0,   0,   0,   0,
+   239, 239, 240, 239, 239, 240,   4,   4, 241, 241, 241, 241, 241, 241, 241, 242,
+   140, 140, 141, 243, 243, 243, 244, 245, 144, 246, 247, 247, 247, 247,  14,  14,
+     0,   0,   0,   0,   0, 248,  45,  45, 249, 250, 249, 249, 249, 249, 249, 251,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 252,  45, 253,
+   254,   0, 255, 256, 257, 258, 258, 258, 258, 259, 260, 261, 261, 261, 261, 262,
+   263, 264, 264, 265, 143, 143, 143, 143, 266,   0, 264, 264,   0,   0, 267, 261,
+   143, 266,   0,   0,   0,   0, 143, 268,   0,   0,   0,   0,   0, 261, 261, 269,
+   261, 261, 261, 261, 261, 270,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,   0,   0,   0,   0,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 271,
+   272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272,
+   272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272, 272,
+   272, 272, 272, 272, 272, 272, 272, 272, 273, 272, 272, 272, 274, 275, 275, 275,
+   276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276, 276,
+   276, 276, 277,  45,  14,  14,  14,  14,  14,  14, 278, 278, 278, 278, 278, 279,
+     0,   0, 280,   4,   4,   4,   4,   4, 281,   4,   4,   4, 282,  45,  45, 283,
+   284, 284, 285, 286, 287, 287, 287, 288, 289, 289, 289, 289, 290, 291,  50,  50,
+   292, 292, 293, 294, 294, 295, 143, 296, 297, 297, 297, 297, 298, 299, 139, 300,
+   301, 301, 301, 302, 303, 304, 139, 139, 305, 305, 305, 305, 306, 307, 308, 309,
+   310, 311, 247,   4,   4, 312, 313, 153, 153, 153, 153, 153, 308, 308, 314, 315,
+   143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143,
+   143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143,
+   143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 143,
+   143, 143, 143, 143, 143, 143, 143, 143, 143, 143, 316, 143, 317, 143, 143, 318,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 319, 249, 249, 249, 249, 249, 249, 320,  45,  45,
+   321, 322,  21, 323, 324,  27,  27,  27,  27,  27,  27,  27, 325,  48,  27,  27,
+    27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,  27,
+    27,  27,  27, 326,  45,  27,  27,  27,  27, 327,  27,  27,  47,  45,  45, 328,
+     8, 286, 329,   0,   0, 330, 331,  46,  27,  27,  27,  27,  27,  27,  27, 332,
+   333,   0,   1,   2,   1,   2, 334, 260, 261, 335, 143, 266, 336, 337, 338, 339,
+   340, 341, 342, 343, 344, 344,  45,  45, 341, 341, 341, 341, 341, 341, 341, 345,
+   346,   0,   0, 347,  11,  11,  11,  11, 348, 349, 350,  45,  45,   0,   0, 351,
+    45,  45,  45,  45,  45,  45,  45,  45, 352, 353, 354, 354, 354, 355, 356, 253,
+   357, 357, 358, 359, 360, 361, 361, 362, 363, 364, 365, 365, 366, 367,  45,  45,
+   368, 368, 368, 368, 368, 369, 369, 369, 370, 371, 372, 373, 373, 374, 373, 375,
+   376, 376, 377, 378, 378, 378, 379,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380, 380,
+   380, 380, 380, 381, 380, 382, 383,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   384, 385, 385, 386, 387, 388, 389, 389, 390, 391, 392,  45,  45,  45, 393, 394,
+   395, 396, 397, 398,  45,  45,  45,  45, 399, 399, 400, 401, 400, 402, 400, 400,
+   403, 404, 405, 406, 407, 407, 408, 408, 409, 409,  45,  45, 410, 410, 411, 412,
+   413, 413, 413, 414, 415, 416, 417, 418, 419, 420, 421,  45,  45,  45,  45,  45,
+   422, 422, 422, 422, 423,  45,  45,  45, 424, 424, 424, 425, 424, 424, 424, 426,
+   427, 427, 428, 429,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  27, 430, 431, 431, 432, 433,  45,  45,  45,  45,
+   434, 434, 435, 436, 436, 437,  45,  45,  45,  45,  45, 438, 439,  45, 440, 441,
+   442, 442, 442, 442, 443, 444, 442, 445, 446, 446, 446, 446, 447, 448, 449, 450,
+   451, 451, 451, 452, 453, 454, 454, 455, 456, 456, 456, 456, 456, 456, 457, 458,
+   459, 460, 459, 461,  45,  45,  45,  45, 462, 463, 464, 465, 465, 465, 466, 467,
+   468, 469, 470, 471, 472, 473, 474, 475,  45,  45,  45,  45,  45,  45,  45,  45,
+   476, 476, 476, 476, 476, 477, 478,  45, 479, 479, 479, 479, 480, 481,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45, 482, 482, 482, 483, 482, 484,  45,  45,
+   485, 485, 485, 485, 486, 487, 488,  45, 489, 489, 489, 490, 491,  45,  45,  45,
+   492, 493, 494, 492,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   495, 495, 495, 496,  45,  45,  45,  45,  45,  45, 497, 497, 497, 497, 497, 498,
+   499, 500, 501, 502, 503, 504,  45,  45,  45,  45, 505, 506, 506, 505, 507,  45,
+   508, 508, 508, 508, 509, 510, 510, 510, 510, 510, 511,  45, 512, 512, 512, 513,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   514, 515, 515, 516, 517, 515, 518, 519, 519, 520, 521, 522,  45,  45,  45,  45,
+   523, 524, 524, 525, 526, 527, 528, 529, 530, 531, 532,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45, 533, 534,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45, 535, 536, 536, 536, 537,
+   538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538,
+   538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538,
+   538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538, 538,
+   538, 538, 538, 538, 538, 538, 538, 538, 538, 539,  45,  45,  45,  45,  45,  45,
+   538, 538, 538, 538, 538, 538, 540, 541, 538, 538, 538, 538, 538, 538, 538, 538,
+   538, 538, 538, 538, 542,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543,
+   543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543, 543,
+   543, 543, 544, 545,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546,
+   546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546,
+   546, 546, 546, 546, 547,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278,
+   278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278, 278,
+   278, 278, 278, 548, 549, 550, 551,  45,  45,  45,  45,  45,  45, 552, 553, 554,
+   555, 555, 555, 555, 556, 557, 558, 559, 555,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45, 560, 560, 560, 560, 560, 561,  45,  45,  45,  45,  45,  45,
+   562, 562, 562, 562, 563, 562, 562, 562, 564, 562,  45,  45,  45,  45, 565, 566,
+   567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567,
+   567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567,
+   567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567,
+   567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 568,
+   567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567, 567,
+   569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569,
+   569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 569, 570,  45,  45,
+   571,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   572, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258, 258,
+   258, 573,  45,  45,  45, 574, 575, 576, 576, 576, 576, 576, 576, 576, 576, 576,
+   576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 577,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   578, 578, 578, 578, 578, 578, 579, 580, 581, 582, 267,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 583,
+     0,   0, 584,   0,   0,   0, 585, 586, 587,   0, 588,   0,   0,   0, 589,  45,
+    11,  11,  11,  11, 590,  45,  45,  45,  45,  45,  45,  45,  45,  45,   0, 267,
+     0,   0,   0,   0,   0, 234,   0, 589,  45,  45,  45,  45,  45,  45,  45,  45,
+     0,   0,   0,   0,   0, 226,   0,   0,   0, 591, 592, 593, 594,   0,   0,   0,
+   595, 596,   0, 597, 598, 599,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 600,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 601,   0,   0,   0,
+   602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602,
+   602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602,
+   602, 602, 602, 602, 602, 602, 602, 602, 603, 604, 605,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   606, 607, 608,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   609, 609, 610, 611, 612,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45, 613, 613, 613, 614,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   615, 615, 615, 615, 615, 615, 615, 615, 615, 615, 615, 615, 616, 617,  45,  45,
+   618, 618, 618, 618, 619, 620,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45, 333,   0,   0,   0, 621,  45,  45,  45,  45,
+   333,   0,   0, 622,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   623,  27, 624, 625, 626, 627, 628, 629, 630, 631, 632, 631,  45,  45,  45, 325,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+     0,   0, 253,   0,   0,   0,   0,   0,   0, 267, 228, 333, 333, 333,   0, 583,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 622,  45,  45,  45, 633,   0,
+   634,   0,   0, 253, 589, 635, 583,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 636, 349, 349,
+     0,   0,   0,   0,   0,   0,   0, 267,   0,   0,   0,   0,   0, 589, 253,  45,
+   253,   0,   0,   0, 636, 286,   0,   0, 636,   0, 622, 635,  45,  45,  45,  45,
+     0,   0,   0,   0,   0,   0,   0, 637,   0,   0,   0,   0, 638,   0,   0,   0,
+     0,   0,   0,   0,   0, 267, 622, 639, 234,   0, 589, 234, 248, 234,  45,  45,
+     0,   0,   0,   0,   0,   0,   0,   0,   0, 330,   0,   0, 235,  45,  45, 286,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 319,  45,  45,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 640, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 319, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 566, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 641,  45,
+   249, 319,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+    45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249, 249,
+   249, 249, 249, 249, 642,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,  45,
+   643,  45,   0,   0,   0,   0,   0,   0,  45,  45,  45,  45,  45,  45,  45,  45,
+     8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   8,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   939, 940, 941, 942, 946, 948,   0, 962, 969, 970, 971, 976,1001,1002,1003,1008,
+     0,1033,1040,1041,1042,1043,1047,   0,   0,1080,1081,1082,1086,1110,   0,   0,
+  1124,1125,1126,1127,1131,1133,   0,1147,1154,1155,1156,1161,1187,1188,1189,1193,
+     0,1219,1226,1227,1228,1229,1233,   0,   0,1267,1268,1269,1273,1298,   0,1303,
+   943,1128, 944,1129, 954,1139, 958,1143, 959,1144, 960,1145, 961,1146, 964,1149,
+     0,   0, 973,1158, 974,1159, 975,1160, 983,1168, 978,1163, 988,1173, 990,1175,
+   991,1176, 993,1178, 994,1179,   0,   0,1004,1190,1005,1191,1006,1192,1014,1199,
+  1007,   0,   0,   0,1016,1201,1020,1206,   0,1022,1208,1025,1211,1023,1209,   0,
+     0,   0,   0,1032,1218,1037,1223,1035,1221,   0,   0,   0,1044,1230,1045,1231,
+  1049,1235,   0,   0,1058,1244,1064,1250,1060,1246,1066,1252,1067,1253,1072,1258,
+  1069,1255,1077,1264,1074,1261,   0,   0,1083,1270,1084,1271,1085,1272,1088,1275,
+  1089,1276,1096,1283,1103,1290,1111,1299,1115,1118,1307,1120,1309,1121,1310,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1053,1239,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1093,
+  1280,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 949,1134,1010,
+  1195,1050,1236,1090,1277,1341,1368,1340,1367,1342,1369,1339,1366,   0,1320,1347,
+  1418,1419,1323,1350,   0,   0, 992,1177,1018,1204,1055,1241,1416,1417,1415,1424,
+  1202,   0,   0,   0, 987,1172,   0,   0,1031,1217,1321,1348,1322,1349,1338,1365,
+   950,1135, 951,1136, 979,1164, 980,1165,1011,1196,1012,1197,1051,1237,1052,1238,
+  1061,1247,1062,1248,1091,1278,1092,1279,1071,1257,1076,1263,   0,   0, 997,1182,
+     0,   0,   0,   0,   0,   0, 945,1130, 982,1167,1337,1364,1335,1362,1046,1232,
+  1422,1423,1113,1301,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     8,   9,   0,  10,1425,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   7,   0,   0,   0,   0,   0,   0,   0,   0,   0,   1,   0,
+     0,   0,   0,   0,   0,1314,1427,   5,1434,1438,1443,   0,1450,   0,1455,1461,
+  1514,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1446,1458,1468,1476,1480,1486,
+  1517,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1489,1503,1494,1500,1508,   0,
+     0,   0,   0,1520,1521,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1526,1528,   0,1525,   0,   0,   0,1522,   0,   0,   0,   0,1536,1532,1539,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1534,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1556,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1548,1550,   0,1547,   0,   0,   0,1567,   0,   0,   0,   0,1558,1554,1561,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1568,1569,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1529,1551,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1523,1545,1524,1546,   0,   0,1527,1549,   0,   0,1570,1571,1530,1552,1531,1553,
+     0,   0,1533,1555,1535,1557,1537,1559,   0,   0,1572,1573,1544,1566,1538,1560,
+  1540,1562,1541,1563,1542,1564,   0,   0,1543,1565,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1606,1607,1609,1608,1610,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1613,   0,1611,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1612,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1620,   0,   0,   0,   0,   0,   0,
+     0,1623,   0,   0,1624,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1614,1615,1616,1617,1618,1619,1621,1622,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1628,1629,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1625,1626,   0,1627,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1634,   0,   0,1635,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1630,1631,1632,   0,   0,1633,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1639,   0,   0,1638,1640,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1636,1637,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1641,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1642,1644,1643,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1645,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1646,   0,   0,   0,   0,   0,   0,1648,1649,   0,1647,1650,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1651,1653,1652,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1654,   0,1655,1657,1656,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1659,   0,   0,   0,   0,   0,   0,   0,   0,   0,1660,   0,   0,
+     0,   0,1661,   0,   0,   0,   0,1662,   0,   0,   0,   0,1663,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1658,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1664,   0,1665,1673,   0,1674,   0,   0,   0,   0,   0,   0,   0,
+     0,1666,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1668,   0,   0,   0,   0,   0,   0,   0,   0,   0,1669,   0,   0,
+     0,   0,1670,   0,   0,   0,   0,1671,   0,   0,   0,   0,1672,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1667,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1675,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1676,   0,1677,   0,1678,   0,1679,   0,1680,   0,
+     0,   0,1681,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1682,   0,1683,   0,   0,
+  1684,1685,   0,1686,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   953,1138, 955,1140, 956,1141, 957,1142,1324,1351, 963,1148, 965,1150, 968,1153,
+   966,1151, 967,1152,1378,1380,1379,1381, 984,1169, 985,1170,1420,1421, 986,1171,
+   989,1174, 995,1180, 998,1183, 996,1181, 999,1184,1000,1185,1015,1200,1329,1356,
+  1017,1203,1019,1205,1021,1207,1024,1210,1687,1688,1027,1213,1026,1212,1028,1214,
+  1029,1215,1030,1216,1034,1220,1036,1222,1039,1225,1038,1224,1334,1361,1336,1363,
+  1382,1384,1383,1385,1056,1242,1057,1243,1059,1245,1063,1249,1689,1690,1065,1251,
+  1068,1254,1070,1256,1386,1387,1388,1389,1691,1692,1073,1259,1075,1262,1079,1266,
+  1078,1265,1095,1282,1098,1285,1097,1284,1390,1391,1392,1393,1099,1286,1100,1287,
+  1101,1288,1102,1289,1105,1292,1104,1291,1106,1294,1107,1295,1108,1296,1114,1302,
+  1119,1308,1122,1311,1123,1312,1186,1260,1293,1305,   0,1394,   0,   0,   0,   0,
+   952,1137, 947,1132,1317,1344,1316,1343,1319,1346,1318,1345,1693,1695,1371,1375,
+  1370,1374,1373,1377,1372,1376,1694,1696, 981,1166, 977,1162, 972,1157,1326,1353,
+  1325,1352,1328,1355,1327,1354,1697,1698,1009,1194,1013,1198,1054,1240,1048,1234,
+  1331,1358,1330,1357,1333,1360,1332,1359,1699,1700,1396,1401,1395,1400,1398,1403,
+  1397,1402,1399,1404,1094,1281,1087,1274,1406,1411,1405,1410,1408,1413,1407,1412,
+  1409,1414,1109,1297,1117,1306,1116,1304,1112,1300,   0,   0,   0,   0,   0,   0,
+  1471,1472,1701,1705,1702,1706,1703,1707,1430,1431,1715,1719,1716,1720,1717,1721,
+  1477,1478,1729,1731,1730,1732,   0,   0,1435,1436,1733,1735,1734,1736,   0,   0,
+  1481,1482,1737,1741,1738,1742,1739,1743,1439,1440,1751,1755,1752,1756,1753,1757,
+  1490,1491,1765,1768,1766,1769,1767,1770,1447,1448,1771,1774,1772,1775,1773,1776,
+  1495,1496,1777,1779,1778,1780,   0,   0,1451,1452,1781,1783,1782,1784,   0,   0,
+  1504,1505,1785,1788,1786,1789,1787,1790,   0,1459,   0,1791,   0,1792,   0,1793,
+  1509,1510,1794,1798,1795,1799,1796,1800,1462,1463,1808,1812,1809,1813,1810,1814,
+  1467,  21,1475,  22,1479,  23,1485,  24,1493,  27,1499,  28,1507,  29,   0,   0,
+  1704,1708,1709,1710,1711,1712,1713,1714,1718,1722,1723,1724,1725,1726,1727,1728,
+  1740,1744,1745,1746,1747,1748,1749,1750,1754,1758,1759,1760,1761,1762,1763,1764,
+  1797,1801,1802,1803,1804,1805,1806,1807,1811,1815,1816,1817,1818,1819,1820,1821,
+  1470,1469,1822,1474,1465,   0,1473,1825,1429,1428,1426,  12,1432,   0,  26,   0,
+     0,1315,1823,1484,1466,   0,1483,1829,1433,  13,1437,  14,1441,1826,1827,1828,
+  1488,1487,1513,  19,   0,   0,1492,1515,1445,1444,1442,  15,   0,1831,1832,1833,
+  1502,1501,1516,  25,1497,1498,1506,1518,1457,1456,1454,  17,1453,1313,  11,   3,
+     0,   0,1824,1512,1519,   0,1511,1830,1449,  16,1460,  18,1464,   4,   0,   0,
+    30,  31,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,  20,   0,   0,   0,   2,   6,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1834,1835,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1836,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1837,1839,1838,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1840,   0,   0,   0,   0,1841,   0,   0,1842,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1843,   0,1844,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1845,   0,   0,1846,   0,   0,1847,   0,1848,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   937,   0,1850,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1849, 936, 938,
+  1851,1852,   0,   0,1853,1854,   0,   0,1855,1856,   0,   0,   0,   0,   0,   0,
+  1857,1858,   0,   0,1861,1862,   0,   0,1863,1864,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1867,1868,1869,1870,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1859,1860,1865,1866,   0,   0,   0,   0,   0,   0,1871,1872,1873,1874,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,  32,  33,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1875,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1877,   0,1878,   0,
+  1879,   0,1880,   0,1881,   0,1882,   0,1883,   0,1884,   0,1885,   0,1886,   0,
+  1887,   0,1888,   0,   0,1889,   0,1890,   0,1891,   0,   0,   0,   0,   0,   0,
+  1892,1893,   0,1894,1895,   0,1896,1897,   0,1898,1899,   0,1900,1901,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1876,   0,   0,   0,   0,   0,   0,   0,   0,   0,1902,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1904,   0,1905,   0,
+  1906,   0,1907,   0,1908,   0,1909,   0,1910,   0,1911,   0,1912,   0,1913,   0,
+  1914,   0,1915,   0,   0,1916,   0,1917,   0,1918,   0,   0,   0,   0,   0,   0,
+  1919,1920,   0,1921,1922,   0,1923,1924,   0,1925,1926,   0,1927,1928,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1903,   0,   0,1929,1930,1931,1932,   0,   0,   0,1933,   0,
+   710, 385, 724, 715, 455, 103, 186, 825, 825, 242, 751, 205, 241, 336, 524, 601,
+   663, 676, 688, 738, 411, 434, 474, 500, 649, 746, 799, 108, 180, 416, 482, 662,
+   810, 275, 462, 658, 692, 344, 618, 679, 293, 388, 440, 492, 740, 116, 146, 168,
+   368, 414, 481, 527, 606, 660, 665, 722, 781, 803, 809, 538, 553, 588, 642, 758,
+   811, 701, 233, 299, 573, 612, 487, 540, 714, 779, 232, 267, 412, 445, 457, 585,
+   594, 766, 167, 613, 149, 148, 560, 589, 648, 768, 708, 345, 411, 704, 105, 259,
+   313, 496, 518, 174, 542, 120, 307, 101, 430, 372, 584, 183, 228, 529, 650, 697,
+   424, 732, 428, 349, 632, 355, 517, 110, 135, 147, 403, 580, 624, 700, 750, 170,
+   193, 245, 297, 374, 463, 543, 763, 801, 812, 815, 162, 384, 420, 730, 287, 330,
+   337, 366, 459, 476, 509, 558, 591, 610, 726, 652, 734, 759, 154, 163, 198, 473,
+   683, 697, 292, 311, 353, 423, 572, 494, 113, 217, 259, 280, 314, 499, 506, 603,
+   608, 752, 778, 782, 788, 117, 557, 748, 774, 320, 109, 126, 260, 265, 373, 411,
+   479, 523, 655, 737, 823, 380, 765, 161, 395, 398, 438, 451, 502, 516, 537, 583,
+   791, 136, 340, 769, 122, 273, 446, 727, 305, 322, 400, 496, 771, 155, 190, 269,
+   377, 391, 406, 432, 501, 519, 599, 684, 687, 749, 776, 175, 452, 191, 480, 510,
+   659, 772, 805, 813, 397, 444, 619, 566, 568, 575, 491, 471, 707, 111, 636, 156,
+   153, 288, 346, 578, 256, 435, 383, 729, 680, 767, 694, 295, 128, 210,   0,   0,
+   227,   0, 379,   0,   0, 150, 493, 525, 544, 551, 552, 556, 783, 576, 604,   0,
+   661,   0, 703,   0,   0, 735, 743,   0,   0,   0, 793, 794, 795, 808, 741, 773,
+   118, 127, 130, 166, 169, 177, 207, 213, 215, 226, 229, 268, 270, 317, 327, 329,
+   335, 369, 375, 381, 404, 441, 448, 458, 477, 484, 503, 539, 545, 547, 546, 548,
+   549, 550, 554, 555, 561, 564, 569, 591, 593, 595, 598, 607, 620, 625, 625, 651,
+   690, 695, 705, 706, 716, 717, 733, 735, 777, 786, 790, 315, 869, 623,   0,   0,
+   102, 145, 134, 115, 129, 138, 165, 171, 207, 202, 206, 212, 227, 231, 240, 243,
+   250, 254, 294, 296, 303, 308, 319, 325, 321, 329, 326, 335, 341, 357, 360, 362,
+   370, 379, 388, 389, 393, 421, 424, 438, 456, 454, 458, 465, 477, 535, 485, 490,
+   493, 507, 512, 514, 521, 522, 525, 526, 528, 533, 532, 541, 565, 569, 574, 586,
+   591, 597, 607, 637, 647, 674, 691, 693, 695, 698, 703, 699, 705, 704, 702, 706,
+   709, 717, 728, 736, 747, 754, 770, 777, 783, 784, 786, 787, 790, 802, 825, 848,
+   847, 857,  55,  65,  66, 883, 892, 916, 822, 824,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1586,   0,1605,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1602,1603,1934,1935,1574,1575,
+  1576,1577,1579,1580,1581,1583,1584,   0,1585,1587,1588,1589,1591,   0,1592,   0,
+  1593,1594,   0,1595,1596,   0,1598,1599,1600,1601,1604,1582,1578,1590,1597,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1936,   0,1937,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1938,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1939,1940,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1941,1942,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1944,1943,   0,1945,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1946,1947,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1948,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1949,1950,
+  1951,1952,1953,1954,1955,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1956,1957,1958,1960,1959,
+  1961,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   106, 104, 107, 826, 114, 118, 119, 121, 123, 124, 127, 125,  34, 830, 130, 131,
+   132, 137, 827,  35, 133, 139, 829, 142, 143, 112, 144, 145, 924, 151, 152,  37,
+   157, 158, 159, 160,  38, 165, 166, 169, 171, 172, 173, 174, 176, 177, 178, 179,
+   181, 182, 182, 182, 833, 468, 184, 185, 834, 187, 188, 189, 196, 192, 194, 195,
+   197, 199, 200, 201, 203, 204, 204, 206, 208, 209, 211, 218, 213, 219, 214, 216,
+   153, 234, 221, 222, 223, 220, 225, 224, 230, 835, 235, 236, 237, 238, 239, 244,
+   836, 837, 247, 248, 249, 246, 251,  39,  40, 253, 255, 255, 838, 257, 258, 259,
+   261, 839, 262, 263, 301, 264,  41, 266, 270, 272, 271, 841, 274, 842, 277, 276,
+   278, 281, 282,  42, 283, 284, 285, 286,  43, 843,  44, 289, 290, 291, 293, 934,
+   298, 845, 845, 621, 300, 300,  45, 852, 894, 302, 304,  46, 306, 309, 310, 312,
+   316,  48,  47, 317, 846, 318, 323, 324, 325, 324, 328, 329, 333, 331, 332, 334,
+   335, 336, 338, 339, 342, 343, 347, 351, 849, 350, 348, 352, 354, 359, 850, 361,
+   358, 356,  49, 363, 365, 367, 364,  50, 369, 371, 851, 376, 386, 378,  53, 381,
+    52,  51, 140, 141, 387, 382, 614,  78, 388, 389, 390, 394, 392, 856,  54, 399,
+   396, 402, 404, 858, 405, 401, 407,  55, 408, 409, 410, 413, 859, 415,  56, 417,
+   860, 418,  57, 419, 422, 424, 425, 861, 840, 862, 426, 863, 429, 431, 427, 433,
+   437, 441, 438, 439, 442, 443, 864, 436, 449, 450,  58, 454, 453, 865, 447, 460,
+   866, 867, 461, 466, 465, 464,  59, 467, 470, 469, 472, 828, 475, 868, 478, 870,
+   483, 485, 486, 871, 488, 489, 872, 873, 495, 497,  60, 498,  61,  61, 504, 505,
+   507, 508, 511,  62, 513, 874, 515, 875, 518, 844, 520, 876, 877, 878,  63,  64,
+   528, 880, 879, 881, 882, 530, 531, 531, 533,  66, 534,  67,  68, 884, 536, 538,
+   541,  69, 885, 549, 886, 887, 556, 559,  70, 561, 562, 563, 888, 889, 889, 567,
+    71, 890, 570, 571,  72, 891, 577,  73, 581, 579, 582, 893, 587,  74, 590, 592,
+   596,  75, 895, 896,  76, 897, 600, 898, 602, 605, 607, 899, 900, 609, 901, 611,
+   853,  77, 615, 616,  79, 617, 252, 902, 903, 854, 855, 621, 622, 731,  80, 627,
+   626, 628, 164, 629, 630, 631, 633, 904, 632, 634, 639, 640, 635, 641, 646, 651,
+   638, 643, 644, 645, 905, 907, 906,  81, 653, 654, 656, 911, 657, 908,  82,  83,
+   909, 910,  84, 664, 665, 666, 667, 669, 668, 671, 670, 674, 672, 673, 675,  85,
+   677, 678,  86, 681, 682, 912, 685, 686,  87, 689,  36, 913, 914,  88,  89, 696,
+   702, 709, 711, 915, 712, 713, 718, 719, 917, 831, 721, 720, 723, 832, 725, 728,
+   918, 919, 739, 742, 744, 920, 745, 753, 756, 757, 755, 760, 761, 921, 762,  90,
+   764, 922,  91, 775, 279, 780, 923, 925,  92,  93, 785, 926,  94, 927, 787, 787,
+   789, 928, 792,  95, 796, 797, 798, 800,  96, 929, 802, 804, 806,  97,  98, 807,
+   930,  99, 931, 932, 933, 814, 100, 816, 817, 818, 819, 820, 821, 935,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+};
+static const int16_t
+_hb_ucd_i16[196] =
+{
+      0,    0,    0,    0,    1,   -1,    0,    0,    2,    0,   -2,    0,    0,    0,    0,    2,
+      0,   -2,    0,    0,    0,    0,    0,   16,    0,    0,    0,  -16,    0,    0,    1,   -1,
+      0,    0,    0,    1,   -1,    0,    0,    0,    0,    1,   -1,    0,    3,    3,    3,   -3,
+     -3,   -3,    0,    0,    0, 2016,    0,    0,    0,    0,    0, 2527, 1923, 1914, 1918,    0,
+   2250,    0,    0,    0,    0,    0,    0,  138,    0,    7,    0,    0,   -7,    0,    0,    0,
+      1,   -1,    1,   -1,   -1,    1,   -1,    0, 1824,    0,    0,    0,    0,    0, 2104,    0,
+   2108, 2106,    0, 2106, 1316,    0,    0,    0,    0,    1,   -1,    1,   -1, -138,    0,    0,
+      1,   -1,    8,    8,    8,    0,    7,    7,    0,    0,   -8,   -8,   -8,   -7,   -7,    0,
+      1,   -1,    0,    2,-1316,    1,   -1,    0,   -1,    1,   -1,    1,   -1,    3,    1,   -1,
+     -3,    1,   -1,    1,   -1,    0,    0,-1914,-1918,    0,    0,-1923,-1824,    0,    0,    0,
+      0,-2016,    0,    0,    1,   -1,    0,    1,    0,    0,-2104,    0,    0,    0,    0,-2106,
+  -2108,-2106,    0,    0,    1,   -1,-2250,    0,    0,    0,-2527,    0,    0,   -2,    0,    1,
+     -1,    0,    1,   -1,
+};
+
+static inline uint_fast8_t
+_hb_ucd_gc (unsigned u)
+{
+  return u<1114110u?_hb_ucd_u8[2176+(((_hb_ucd_u16[((_hb_ucd_u8[u>>4>>5])<<5)+((u>>4)&31u)])<<4)+((u)&15u))]:2;
+}
+static inline uint_fast8_t
+_hb_ucd_ccc (unsigned u)
+{
+  return u<125259u?_hb_ucd_u8[15060+(((_hb_ucd_u8[13636+(((_hb_ucd_u8[12656+(u>>3>>4)])<<4)+((u>>3)&15u))])<<3)+((u)&7u))]:0;
+}
+static inline unsigned
+_hb_ucd_b4 (const uint8_t* a, unsigned i)
+{
+  return (a[i>>1]>>((i&1u)<<2))&15u;
+}
+static inline int_fast16_t
+_hb_ucd_bmg (unsigned u)
+{
+  return u<65380u?_hb_ucd_i16[((_hb_ucd_u8[16372+(((_hb_ucd_b4(16244+_hb_ucd_u8,u>>2>>6))<<6)+((u>>2)&63u))])<<2)+((u)&3u)]:0;
+}
+static inline uint_fast8_t
+_hb_ucd_sc (unsigned u)
+{
+  return u<918000u?_hb_ucd_u8[19126+(((_hb_ucd_u16[3040+(((_hb_ucd_u8[17332+(u>>4>>5)])<<5)+((u>>4)&31u))])<<4)+((u)&15u))]:2;
+}
+static inline uint_fast16_t
+_hb_ucd_dm (unsigned u)
+{
+  return u<195102u?_hb_ucd_u16[6144+(((_hb_ucd_u8[29430+(u>>6)])<<6)+((u)&63u))]:0;
+}
+
+
+#elif !defined(HB_NO_UCD_UNASSIGNED)
+
+static const uint8_t
+_hb_ucd_u8[17508] =
+{
+    0,  1,  2,  3,  4,  5,  6,  7,  7,  8,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  9, 10, 11,  7,  7,  7,  7, 12, 13, 14, 14, 14, 15,
+   16, 17, 18, 19, 20, 21, 22, 21, 23, 21, 21, 21, 21, 24,  7,  7,
+   25, 26, 21, 21, 21, 21, 27, 28, 21, 21, 29, 30, 31, 32, 33, 34,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7, 35,  7, 36, 37,  7, 38,  7,  7,  7, 39, 21, 40,
+    7,  7, 41, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   42, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 43,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 44,
+    0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15,
+   16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31,
+   32, 33, 34, 34, 35, 36, 37, 38, 39, 34, 34, 34, 40, 41, 42, 43,
+   44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59,
+   60, 61, 62, 63, 64, 64, 65, 66, 67, 68, 69, 70, 71, 69, 72, 73,
+   69, 69, 64, 74, 64, 64, 75, 76, 77, 78, 79, 80, 81, 82, 69, 83,
+   84, 85, 86, 87, 88, 89, 69, 69, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 90, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 91,
+   92, 34, 34, 34, 34, 34, 34, 34, 34, 93, 34, 34, 94, 95, 96, 97,
+   98, 99,100,101,102,103,104,105, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,106,
+  107,107,107,107,107,107,107,107,107,107,107,107,107,107,107,107,
+  108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,
+  108,108, 34, 34,109,110,111,112, 34, 34,113,114,115,116,117,118,
+  119,120,121,122,123,124,125,126,127,128,129,123, 34, 34,130,123,
+  131,132,133,134,135,136,137,138,139,140,141,123,142,143,144,145,
+  146,147,148,149,150,151,152,123,153,154,123,155,156,157,158,123,
+  159,160,161,162,163,164,123,123,165,166,167,168,123,169,123,170,
+   34, 34, 34, 34, 34, 34, 34,171,172, 34,173,123,123,123,123,123,
+  123,123,123,123,123,123,123,123,123,123,123,123,123,123,123,123,
+   34, 34, 34, 34, 34, 34, 34, 34,174,123,123,123,123,123,123,123,
+  123,123,123,123,123,123,123,123, 34, 34, 34, 34,175,123,123,123,
+   34, 34, 34, 34,176,177,178,179,123,123,123,123,180,181,182,183,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,184,
+   34, 34, 34, 34, 34, 34, 34, 34, 34,185,186,123,123,123,123,123,
+   34, 34,187, 34, 34,188,123,123,123,123,123,123,123,123,123,123,
+  123,123,123,123,123,123,123,123,189,190,123,123,123,123,123,123,
+   69,191,192,193,194,195,196,123,197,198,199,200,201,202,203,204,
+   69, 69, 69, 69,205,206,123,123,123,123,123,123,123,123,123,123,
+  207,123,208,123,123,209,123,123,123,123,123,123,123,123,123,123,
+   34,210,211,123,123,123,123,123,212,213,214,123,215,216,123,123,
+  217,218,219,220,221,123, 69,222, 69, 69, 69, 69, 69,223,224,225,
+  226,227,228,229,230,231, 69,232,123,123,123,123,123,123,123,123,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,233, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,234, 34,
+  235, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,236, 34, 34,
+   34, 34, 34, 34, 34, 34, 34,237,123,123,123,123,123,123,123,123,
+   34, 34, 34, 34,238,123,123,123,123,123,123,123,123,123,123,123,
+   34, 34, 34, 34, 34, 34,239,123,123,123,123,123,123,123,123,123,
+  240,123,241,242,123,123,123,123,123,123,123,123,123,123,123,123,
+  108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,243,
+  108,108,108,108,108,108,108,108,108,108,108,108,108,108,108,244,
+    0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,  2,  4,  5,  6,  2,
+    7,  7,  7,  7,  7,  2,  8,  9, 10, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 17, 18, 19,  1, 20, 20, 21, 22, 23, 24, 25,
+   26, 27, 15,  2, 28, 29, 27, 30, 11, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 31, 11, 11, 11, 32, 16, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 33, 16, 16, 16, 16, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 34, 34, 34, 34, 34, 34, 34, 34, 16, 32, 32, 32,
+   32, 32, 32, 32, 11, 34, 34, 16, 34, 32, 32, 11, 34, 11, 16, 11,
+   11, 34, 32, 11, 32, 16, 11, 34, 32, 32, 32, 11, 34, 16, 32, 11,
+   34, 11, 34, 34, 32, 35, 32, 16, 36, 36, 37, 34, 38, 37, 34, 34,
+   34, 34, 34, 34, 34, 34, 16, 32, 34, 38, 32, 11, 32, 32, 32, 32,
+   32, 32, 16, 16, 16, 11, 34, 32, 34, 34, 11, 32, 32, 32, 32, 32,
+   16, 16, 39, 16, 16, 16, 16, 16, 40, 40, 40, 40, 40, 40, 40, 40,
+   40, 41, 41, 40, 40, 40, 40, 40, 40, 41, 41, 41, 41, 41, 41, 41,
+   40, 40, 42, 41, 41, 41, 42, 42, 41, 41, 41, 41, 41, 41, 41, 41,
+   43, 43, 43, 43, 43, 43, 43, 43, 32, 32, 42, 32, 44, 45, 16, 10,
+   44, 44, 41, 46, 11, 47, 47, 11, 34, 11, 11, 11, 11, 11, 11, 11,
+   11, 48, 11, 11, 11, 11, 16, 16, 16, 16, 16, 16, 16, 16, 16, 34,
+   16, 11, 32, 16, 32, 32, 32, 32, 16, 16, 32, 49, 34, 32, 34, 11,
+   32, 50, 43, 43, 51, 32, 32, 32, 11, 34, 34, 34, 34, 34, 34, 16,
+   48, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 47, 52,  2,  2,  2,
+   16, 16, 16, 16, 53, 54, 55, 56, 57, 43, 43, 43, 43, 43, 43, 43,
+   43, 43, 43, 43, 43, 43, 43, 58, 59, 60, 43, 59, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 61, 44, 62,
+   36, 63, 64, 44, 44, 44, 44, 44, 65, 65, 65,  8,  9, 66,  2, 67,
+   43, 43, 43, 43, 43, 60, 68,  2, 69, 36, 36, 36, 36, 70, 43, 43,
+    7,  7,  7,  7,  7,  2,  2, 36, 71, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 72, 43, 43, 43, 73, 50, 43, 43, 74, 75, 76, 43, 43, 36,
+    7,  7,  7,  7,  7, 36, 77, 78,  2,  2,  2,  2,  2,  2,  2, 79,
+   70, 36, 36, 36, 36, 36, 36, 36, 43, 43, 43, 43, 43, 80, 62, 36,
+   36, 36, 36, 43, 43, 43, 43, 43, 71, 44, 44, 44, 44, 44, 44, 44,
+    7,  7,  7,  7,  7, 36, 36, 36, 36, 36, 36, 36, 36, 70, 43, 43,
+   43, 43, 40, 21,  2, 81, 57, 20, 36, 36, 36, 43, 43, 75, 43, 43,
+   43, 43, 75, 43, 75, 43, 43, 44,  2,  2,  2,  2,  2,  2,  2, 64,
+   36, 36, 36, 36, 70, 43, 44, 64, 36, 36, 36, 36, 36, 61, 44, 44,
+   44, 44, 44, 44, 44, 44, 44, 44, 36, 36, 61, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 44, 44, 44, 44, 44, 57, 43, 43, 43, 43, 43, 43,
+   43, 82, 43, 43, 43, 43, 43, 43, 43, 83, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 83, 71, 84, 85, 43, 43, 43, 83, 84, 85, 84,
+   70, 43, 43, 43, 36, 36, 36, 36, 36, 43,  2,  7,  7,  7,  7,  7,
+   86, 36, 36, 36, 36, 36, 36, 36, 70, 84, 62, 36, 36, 36, 61, 62,
+   61, 62, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 61, 36, 36, 36,
+   61, 61, 44, 36, 36, 44, 71, 84, 85, 43, 80, 87, 88, 87, 85, 61,
+   44, 44, 44, 87, 44, 44, 36, 62, 36, 43, 44,  7,  7,  7,  7,  7,
+   36, 20, 27, 27, 27, 56, 63, 80, 57, 83, 62, 36, 36, 61, 44, 62,
+   61, 36, 62, 61, 36, 44, 80, 84, 85, 80, 44, 57, 80, 57, 43, 44,
+   57, 44, 44, 44, 62, 36, 61, 61, 44, 44, 44,  7,  7,  7,  7,  7,
+   43, 36, 70, 64, 44, 44, 44, 44, 57, 83, 62, 36, 36, 36, 36, 62,
+   36, 62, 36, 36, 36, 36, 36, 36, 61, 36, 62, 36, 36, 44, 71, 84,
+   85, 43, 43, 57, 83, 87, 85, 44, 61, 44, 44, 44, 44, 44, 44, 44,
+   66, 44, 44, 44, 62, 43, 43, 43, 57, 84, 62, 36, 36, 36, 61, 62,
+   61, 36, 62, 36, 36, 44, 71, 85, 85, 43, 80, 87, 88, 87, 85, 44,
+   44, 44, 57, 83, 44, 44, 36, 62, 78, 27, 27, 27, 44, 44, 44, 44,
+   44, 71, 62, 36, 36, 61, 44, 36, 61, 36, 36, 44, 62, 61, 61, 36,
+   44, 62, 61, 44, 36, 61, 44, 36, 36, 36, 36, 36, 36, 44, 44, 84,
+   83, 88, 44, 84, 88, 84, 85, 44, 61, 44, 44, 87, 44, 44, 44, 44,
+   27, 89, 67, 67, 56, 90, 44, 44, 83, 84, 71, 36, 36, 36, 61, 36,
+   61, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 44, 62, 43,
+   83, 84, 88, 43, 80, 43, 43, 44, 44, 44, 57, 80, 36, 61, 44, 44,
+   44, 44, 44, 91, 27, 27, 27, 89, 70, 84, 72, 36, 36, 36, 61, 36,
+   36, 36, 62, 36, 36, 44, 71, 85, 84, 84, 88, 83, 88, 84, 43, 44,
+   44, 44, 87, 88, 44, 44, 44, 61, 62, 61, 44, 44, 44, 44, 44, 44,
+   43, 84, 36, 36, 36, 36, 61, 36, 36, 36, 36, 36, 36, 70, 71, 84,
+   85, 43, 80, 84, 88, 84, 85, 77, 44, 44, 36, 92, 27, 27, 27, 93,
+   27, 27, 27, 27, 89, 36, 36, 36, 57, 84, 62, 36, 36, 36, 36, 36,
+   36, 36, 36, 61, 44, 36, 36, 36, 36, 62, 36, 36, 36, 36, 62, 44,
+   36, 36, 36, 61, 44, 80, 44, 87, 84, 43, 80, 80, 84, 84, 84, 84,
+   44, 84, 64, 44, 44, 44, 44, 44, 62, 36, 36, 36, 36, 36, 36, 36,
+   70, 36, 43, 43, 43, 80, 44, 94, 36, 36, 36, 75, 43, 43, 43, 60,
+    7,  7,  7,  7,  7,  2, 44, 44, 62, 61, 61, 36, 36, 61, 36, 36,
+   36, 36, 62, 62, 36, 36, 36, 36, 70, 36, 43, 43, 43, 43, 71, 44,
+   36, 36, 61, 81, 43, 43, 43, 44,  7,  7,  7,  7,  7, 44, 36, 36,
+   77, 67,  2,  2,  2,  2,  2,  2,  2, 95, 95, 67, 43, 67, 67, 67,
+    7,  7,  7,  7,  7, 27, 27, 27, 27, 27, 50, 50, 50,  4,  4, 84,
+   36, 36, 36, 36, 62, 36, 36, 36, 36, 36, 36, 36, 36, 36, 61, 44,
+   57, 43, 43, 43, 43, 43, 43, 83, 43, 43, 60, 43, 36, 36, 70, 43,
+   43, 43, 43, 43, 57, 43, 43, 43, 43, 43, 43, 43, 43, 43, 80, 67,
+   67, 67, 67, 76, 67, 67, 90, 67,  2,  2, 95, 67, 21, 64, 44, 44,
+   36, 36, 36, 36, 36, 92, 85, 43, 83, 43, 43, 43, 85, 83, 85, 71,
+    7,  7,  7,  7,  7,  2,  2,  2, 36, 36, 36, 84, 43, 36, 36, 43,
+   71, 84, 96, 92, 84, 84, 84, 36, 70, 43, 71, 36, 36, 36, 36, 36,
+   36, 83, 85, 83, 84, 84, 85, 92,  7,  7,  7,  7,  7, 84, 85, 67,
+   11, 11, 11, 48, 44, 44, 48, 44, 16, 16, 16, 16, 16, 53, 45, 16,
+   36, 36, 36, 36, 61, 36, 36, 44, 36, 36, 36, 61, 61, 36, 36, 44,
+   61, 36, 36, 44, 36, 36, 36, 61, 61, 36, 36, 44, 36, 36, 36, 36,
+   36, 36, 36, 61, 36, 36, 36, 36, 36, 36, 36, 36, 36, 61, 57, 43,
+    2,  2,  2,  2, 97, 27, 27, 27, 27, 27, 27, 27, 27, 27, 98, 44,
+   67, 67, 67, 67, 67, 44, 44, 44, 11, 11, 11, 44, 16, 16, 16, 44,
+   99, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 77, 72,
+  100, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,101,102, 44,
+   36, 36, 36, 36, 36, 63,  2,103,104, 36, 36, 36, 61, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 61, 36, 36, 43, 80, 44, 44, 44, 44, 44,
+   36, 43, 60, 64, 44, 44, 44, 44, 36, 43, 44, 44, 44, 44, 44, 44,
+   61, 43, 44, 44, 44, 44, 44, 44, 36, 36, 43, 85, 43, 43, 43, 84,
+   84, 84, 84, 83, 85, 43, 43, 43, 43, 43,  2, 86,  2, 66, 70, 44,
+    7,  7,  7,  7,  7, 44, 44, 44, 27, 27, 27, 27, 27, 44, 44, 44,
+    2,  2,  2,105,  2, 59, 43, 68, 36,106, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 61, 44, 44, 44, 36, 36, 70, 71, 36, 36, 36, 36,
+   36, 36, 36, 36, 70, 61, 44, 44, 36, 36, 36, 44, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 36, 61, 43, 83, 84, 85, 83, 84, 44, 44,
+   84, 83, 84, 84, 85, 43, 44, 44, 90, 44,  2,  7,  7,  7,  7,  7,
+   36, 36, 36, 36, 36, 36, 36, 44, 36, 36, 61, 44, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 44, 44, 36, 36, 36, 36, 36, 44, 44, 44,
+    7,  7,  7,  7,  7, 98, 44, 67, 67, 67, 67, 67, 67, 67, 67, 67,
+   36, 36, 36, 70, 83, 85, 44,  2, 36, 36, 92, 83, 43, 43, 43, 80,
+   83, 83, 85, 43, 43, 43, 83, 84, 84, 85, 43, 43, 43, 43, 80, 57,
+    2,  2,  2, 86,  2,  2,  2, 44, 43, 43, 43, 43, 43, 43, 43,107,
+   80, 44, 44, 44, 44, 44, 44, 44, 43, 43, 96, 36, 36, 36, 36, 36,
+   36, 36, 83, 43, 43, 83, 83, 84, 84, 83, 96, 36, 36, 36, 44, 44,
+   95, 67, 67, 67, 67, 50, 43, 43, 43, 43, 67, 67, 67, 67, 90, 44,
+   43, 96, 36, 36, 36, 36, 36, 36, 92, 43, 43, 84, 43, 85, 43, 36,
+   36, 36, 36, 83, 43, 84, 85, 85, 43, 84, 44, 44, 44, 44,  2,  2,
+   36, 36, 84, 84, 84, 84, 43, 43, 43, 43, 84, 43, 44, 91,  2,  2,
+    7,  7,  7,  7,  7, 44, 62, 36, 36, 36, 36, 36, 40, 40, 40,  2,
+   16, 16, 16, 16,108, 44, 44, 44, 11, 11, 11, 11, 11, 47, 48, 11,
+    2,  2,  2,  2, 44, 44, 44, 44, 43, 60, 43, 43, 43, 43, 43, 43,
+   83, 43, 43, 43, 71, 36, 70, 36, 36, 36, 71, 92, 43, 61, 44, 44,
+   16, 16, 16, 16, 16, 16, 40, 40, 40, 40, 40, 40, 40, 45, 16, 16,
+   16, 16, 16, 16, 45, 16, 16, 16, 16, 16, 16, 16, 16,109, 40, 40,
+   43, 43, 43, 43, 43, 57, 43, 43, 32, 32, 32, 16, 16, 16, 16, 32,
+   16, 16, 16, 16, 11, 11, 11, 11, 16, 16, 16, 44, 11, 11, 11, 44,
+   16, 16, 16, 16, 48, 48, 48, 48, 16, 16, 16, 16, 16, 16, 16, 44,
+   16, 16, 16, 16,110,110,110,110, 16, 16,108, 16, 11, 11,111,112,
+   41, 16,108, 16, 11, 11,111, 41, 16, 16, 44, 16, 11, 11,113, 41,
+   16, 16, 16, 16, 11, 11,114, 41, 44, 16,108, 16, 11, 11,111,115,
+  116,116,116,116,116,117, 65, 65,118,118,118,  2,119,120,119,120,
+    2,  2,  2,  2,121, 65, 65,122,  2,  2,  2,  2,123,124,  2,125,
+  126,  2,127,128,  2,  2,  2,  2,  2,  9,126,  2,  2,  2,  2,129,
+   65, 65, 68, 65, 65, 65, 65, 65,130, 44, 27, 27, 27,  8,127,131,
+   27, 27, 27, 27, 27,  8,127,102, 40, 40, 40, 40, 40, 40, 81, 44,
+   20, 20, 20, 20, 20, 20, 20, 20, 43, 43, 43, 43, 43, 43,132, 51,
+  107, 51,107, 43, 43, 43, 43, 43, 67,133, 67,134, 67, 34, 11, 16,
+   11, 32,134, 67, 49, 11, 11, 67, 67, 67,133,133,133, 11, 11,135,
+   11, 11, 35, 36, 39, 67, 16, 11,  8,  8, 49, 16, 16, 26, 67,136,
+   27, 27, 27, 27, 27, 27, 27, 27,103,103,103,103,103,103,103,103,
+  103,137,138,103,139, 67, 44, 44,  8,  8,140, 67, 67,  8, 67, 67,
+  140, 26, 67,140, 67, 67, 67,140, 67, 67, 67, 67, 67, 67, 67,  8,
+   67,140,140, 67, 67, 67, 67, 67, 67, 67,  8,  8,  8,  8,  8,  8,
+    8,  8,  8,  8,  8,  8,  8,  8, 67, 67, 67, 67,  4,  4, 67, 67,
+    8, 67, 67, 67,141,142, 67, 67, 67, 67, 67, 67, 67, 67,140, 67,
+   67, 67, 67, 67, 67, 26,  8,  8,  8,  8, 67, 67, 67, 67, 67, 67,
+   67, 67, 67, 67, 67, 67,  8,  8,  8, 67, 67, 67, 67, 67, 67, 67,
+   67, 67, 67, 90, 44, 44, 44, 44, 67, 67, 67, 67, 67, 90, 44, 44,
+   27, 27, 27, 27, 27, 27, 67, 67, 67, 67, 67, 67, 67, 27, 27, 27,
+   67, 67, 67, 26, 67, 67, 67, 67, 26, 67, 67, 67, 67, 67, 67, 67,
+   67, 67, 67, 67,  8,  8,  8,  8, 67, 67, 67, 67, 67, 67, 67, 26,
+   67, 67, 67, 67,  4,  4,  4,  4,  4,  4,  4, 27, 27, 27, 27, 27,
+   27, 27, 67, 67, 67, 67, 67, 67,  8,  8,127,143,  8,  8,  8,  8,
+    8,  8,  8,  4,  4,  4,  4,  4,  8,127,144,144,144,144,144,144,
+  144,144,144,144,143,  8,  8,  8,  8,  8,  8,  8,  4,  4,  8,  8,
+    8,  8,  8,  8,  8,  8,  4,  8,  8,  8,140, 26,  8,  8,140, 67,
+   67, 67, 44, 67, 67, 67, 67, 67, 67, 67, 67, 55, 67, 67, 67, 67,
+   11, 11, 11, 11, 11, 11, 11, 47, 16, 16, 16, 16, 16, 16, 16,108,
+   32, 11, 32, 34, 34, 34, 34, 11, 32, 32, 34, 16, 16, 16, 40, 11,
+   32, 32,136, 67, 67,134, 34,145, 43, 32, 44, 44, 91,  2, 97,  2,
+   16, 16, 16,146, 44, 44,146, 44, 36, 36, 36, 36, 44, 44, 44, 52,
+   64, 44, 44, 44, 44, 44, 44, 57, 36, 36, 36, 61, 44, 44, 44, 44,
+   36, 36, 36, 61, 36, 36, 36, 61,  2,119,119,  2,123,124,119,  2,
+    2,  2,  2,  6,  2,105,119,  2,119,  4,  4,  4,  4,  2,  2, 86,
+    2,  2,  2,  2,  2,118,  2,  2,105,147,  2,  2,  2,  2,  2,  2,
+   67, 64, 44, 44, 44, 44, 44, 44, 67, 67, 67, 67, 67, 55, 67, 67,
+   67, 67, 44, 44, 44, 44, 44, 44, 67, 67, 67, 44, 44, 44, 44, 44,
+   67, 67, 67, 67, 67, 67, 44, 44,  1,  2,148,149,  4,  4,  4,  4,
+    4, 67,  4,  4,  4,  4,150,151,152,103,103,103,103, 43, 43, 84,
+  153, 40, 40, 67,103,154, 63, 67, 36, 36, 36, 61, 57,155,156, 69,
+   36, 36, 36, 36, 36, 63, 40, 69, 44, 44, 62, 36, 36, 36, 36, 36,
+   67, 27, 27, 67, 67, 67, 67, 67, 67, 67, 67, 67, 67, 67, 67, 90,
+   27, 27, 27, 27, 27, 67, 67, 67, 67, 67, 67, 67, 27, 27, 27, 27,
+  157, 27, 27, 27, 27, 27, 27, 27, 36, 36,106, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36,158,  2,  7,  7,  7,  7,  7, 36, 44, 44,
+   32, 32, 32, 32, 32, 32, 32, 70, 51,159, 43, 43, 43, 43, 43, 86,
+   32, 32, 32, 32, 32, 32, 40, 43, 36, 36, 36,103,103,103,103,103,
+   43,  2,  2,  2, 44, 44, 44, 44, 41, 41, 41,156, 40, 40, 40, 40,
+   41, 32, 32, 32, 32, 32, 32, 32, 16, 32, 32, 32, 32, 32, 32, 32,
+   45, 16, 16, 16, 34, 34, 34, 32, 32, 32, 32, 32, 42,160, 34, 35,
+   32, 32, 16, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 11, 11, 32,
+   11, 11, 32, 32, 32, 32, 32, 32, 44, 32, 11, 11, 34,108, 44, 44,
+   44, 44, 48, 35, 40, 35, 36, 36, 36, 71, 36, 71, 36, 70, 36, 36,
+   36, 92, 85, 83, 67, 67, 80, 44, 27, 27, 27, 67,161, 44, 44, 44,
+   36, 36,  2,  2, 44, 44, 44, 44, 84, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 84, 84, 84, 84, 84, 84, 84, 84, 43, 44, 44, 44, 44,  2,
+   43, 36, 36, 36,  2, 72, 72, 70, 36, 36, 36, 43, 43, 43, 43,  2,
+   36, 36, 36, 70, 43, 43, 43, 43, 43, 84, 44, 44, 44, 44, 44, 91,
+   36, 70, 84, 43, 43, 84, 43, 84,162,  2,  2,  2,  2,  2,  2, 52,
+    7,  7,  7,  7,  7, 44, 44,  2, 36, 36, 70, 69, 36, 36, 36, 36,
+    7,  7,  7,  7,  7, 36, 36, 61, 36, 36, 36, 36, 70, 43, 43, 83,
+   85, 83, 85, 80, 44, 44, 44, 44, 36, 70, 36, 36, 36, 36, 83, 44,
+    7,  7,  7,  7,  7, 44,  2,  2, 69, 36, 36, 77, 67, 92, 83, 36,
+   71, 43, 71, 70, 71, 36, 36, 43, 70, 61, 44, 44, 44, 44, 44, 44,
+   44, 44, 44, 44, 44, 62,106,  2, 36, 36, 36, 36, 36, 92, 43, 84,
+    2,106,163, 80, 44, 44, 44, 44, 62, 36, 36, 61, 62, 36, 36, 61,
+   62, 36, 36, 61, 44, 44, 44, 44, 16, 16, 16, 16, 16,112, 40, 40,
+   16, 16, 16, 16,109, 41, 44, 44, 36, 92, 85, 84, 83,162, 85, 44,
+   36, 36, 44, 44, 44, 44, 44, 44, 36, 36, 36, 61, 44, 62, 36, 36,
+  164,164,164,164,164,164,164,164,165,165,165,165,165,165,165,165,
+   16, 16, 16,108, 44, 44, 44, 44, 44,146, 16, 16, 44, 44, 62, 71,
+   36, 36, 36, 36,166, 36, 36, 36, 36, 36, 36, 61, 36, 36, 61, 61,
+   36, 62, 61, 36, 36, 36, 36, 36, 36, 41, 41, 41, 41, 41, 41, 41,
+   41, 44, 44, 44, 44, 44, 44, 44, 44, 62, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36,144, 44, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36,161, 44,  2,  2,  2,167,128, 44, 44, 44,
+    6,168,169,144,144,144,144,144,144,144,128,167,128,  2,125,170,
+    2, 64,  2,  2,150,144,144,128,  2,171,  8,172, 66,  2, 44, 44,
+   36, 36, 36, 36, 36, 36, 61, 79, 91,  2,  3,  2,  4,  5,  6,  2,
+   16, 16, 16, 16, 16, 17, 18,127,128,  4,  2, 36, 36, 36, 36, 36,
+   69, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 40,
+   44, 36, 36, 36, 44, 36, 36, 36, 44, 36, 36, 36, 44, 36, 61, 44,
+   20,173, 56,174, 26,  8,140, 90, 44, 44, 44, 44, 79, 65, 67, 44,
+   36, 36, 36, 36, 36, 36, 62, 36, 36, 36, 36, 36, 36, 61, 36, 62,
+    2, 64, 44,175, 27, 27, 27, 27, 27, 27, 44, 55, 67, 67, 67, 67,
+  103,103,139, 27, 89, 67, 67, 67, 67, 67, 67, 67, 67, 27, 67, 90,
+   67, 67, 67, 67, 67, 67, 90, 44, 90, 44, 44, 44, 44, 44, 44, 44,
+   67, 67, 67, 67, 67, 67, 50, 44,176, 27, 27, 27, 27, 27, 27, 27,
+   27, 27, 27, 27, 27, 27, 44, 44, 27, 27, 44, 44, 44, 44, 62, 36,
+  149, 36, 36, 36, 36,177, 44, 44, 36, 36, 36, 43, 43, 80, 44, 44,
+   36, 36, 36, 36, 36, 36, 36, 91, 36, 36, 44, 44, 36, 36, 36, 36,
+  178,103,103, 44, 44, 44, 44, 44, 11, 11, 11, 11, 16, 16, 16, 16,
+   11, 11, 44, 44, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 44, 44,
+   36, 36, 44, 44, 44, 44, 44, 91, 36, 36, 36, 44, 61, 36, 36, 36,
+   36, 36, 36, 62, 61, 44, 61, 62, 36, 36, 36, 91, 27, 27, 27, 27,
+   36, 36, 36, 77,157, 27, 27, 27, 44, 44, 44,175, 27, 27, 27, 27,
+   36, 61, 36, 44, 44,175, 27, 27, 36, 36, 36, 27, 27, 27, 44, 91,
+   36, 36, 36, 36, 36, 44, 44, 91, 36, 36, 36, 36, 44, 44, 27, 36,
+   44, 27, 27, 27, 27, 27, 27, 27, 70, 43, 57, 80, 44, 44, 43, 43,
+   36, 36, 62, 36, 62, 36, 36, 36, 36, 36, 36, 44, 43, 80, 44, 57,
+   27, 27, 27, 27, 98, 44, 44, 44,  2,  2,  2,  2, 64, 44, 44, 44,
+   36, 36, 36, 36, 36, 36,179, 30, 36, 36, 36, 36, 36, 36,179, 27,
+   36, 36, 36, 36, 78, 36, 36, 36, 36, 36, 70, 80, 44,175, 27, 27,
+    2,  2,  2, 64, 44, 44, 44, 44, 36, 36, 36, 44, 91,  2,  2,  2,
+   36, 36, 36, 44, 27, 27, 27, 27, 36, 61, 44, 44, 27, 27, 27, 27,
+   36, 44, 44, 44, 91,  2, 64, 44, 44, 44, 44, 44,175, 27, 27, 27,
+   11, 47, 44, 44, 44, 44, 44, 44, 16,108, 44, 44, 44, 27, 27, 27,
+   36, 36, 43, 43, 44, 44, 44, 44, 27, 27, 27, 27, 27, 27, 27, 98,
+   36, 36, 36, 36, 36, 57,180, 44, 36, 44, 44, 44, 44, 44, 44, 44,
+   27, 27, 27, 93, 44, 44, 44, 44,176, 27, 30,  2,  2, 44, 44, 44,
+   36, 36,179, 27, 27, 27, 44, 44, 85, 96, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 43, 43, 43, 43, 43, 43, 43, 60,  2,  2,  2, 44,
+   27, 27, 27,  7,  7,  7,  7,  7, 44, 44, 44, 44, 44, 44, 44, 57,
+   84, 85, 43, 83, 85, 60,181,  2,  2, 44, 44, 44, 44, 44, 79, 44,
+   43, 71, 36, 36, 36, 36, 36, 36, 36, 36, 36, 70, 43, 43, 85, 43,
+   43, 43, 80,  7,  7,  7,  7,  7,  2,  2, 92, 96, 44, 44, 44, 44,
+   36, 70,  2, 61, 44, 44, 44, 44, 36, 92, 84, 43, 43, 43, 43, 83,
+   96, 36, 63,  2, 59, 43, 60, 85,  7,  7,  7,  7,  7, 63, 63,  2,
+  175, 27, 27, 27, 27, 27, 27, 27, 27, 27, 98, 44, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 84, 85, 43, 84, 83, 43,  2,  2,  2, 80,
+   36, 36, 36, 61, 61, 36, 36, 62, 36, 36, 36, 36, 36, 36, 36, 62,
+   36, 36, 36, 36, 63, 44, 44, 44, 36, 36, 36, 36, 36, 36, 36, 70,
+   84, 85, 43, 43, 43, 80, 44, 44, 43, 84, 62, 36, 36, 36, 61, 62,
+   61, 36, 62, 36, 36, 57, 71, 84, 83, 84, 88, 87, 88, 87, 84, 44,
+   61, 44, 44, 87, 44, 44, 62, 36, 36, 84, 44, 43, 43, 43, 80, 44,
+   43, 43, 80, 44, 44, 44, 44, 44, 36, 36, 92, 84, 43, 43, 43, 43,
+   84, 43, 83, 71, 36, 63,  2,  2,  7,  7,  7,  7,  7,  2, 91, 71,
+   84, 85, 43, 43, 83, 83, 84, 85, 83, 43, 36, 72, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 36, 36, 92, 84, 43, 43, 44, 84, 84, 43, 85,
+   60,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 36, 36, 43, 44,
+   84, 85, 43, 43, 43, 83, 85, 85, 60,  2, 61, 44, 44, 44, 44, 44,
+    2,  2,  2,  2,  2,  2, 64, 44, 36, 36, 36, 36, 36, 70, 85, 84,
+   43, 43, 43, 85, 61, 44, 44, 44, 84, 43, 43, 85, 43, 43, 44, 44,
+    7,  7,  7,  7,  7, 27,  2, 95, 43, 43, 43, 43, 85, 60, 44, 44,
+   27, 98, 44, 44, 44, 44, 44, 62, 36, 36, 36, 61, 62, 44, 36, 36,
+   36, 36, 62, 61, 36, 36, 36, 36, 84, 84, 84, 87, 88, 57, 83, 71,
+   96, 85,  2, 64, 44, 44, 44, 44, 36, 36, 36, 36, 44, 36, 36, 36,
+   92, 84, 43, 43, 44, 43, 84, 84, 71, 72, 88, 44, 44, 44, 44, 44,
+   70, 43, 43, 43, 43, 71, 36, 36, 36, 70, 43, 43, 83, 70, 43, 60,
+    2,  2,  2, 59, 44, 44, 44, 44, 70, 43, 43, 83, 85, 43, 36, 36,
+   36, 36, 36, 36, 36, 43, 43, 43, 43, 43, 43, 83, 43,  2, 72,  2,
+    2, 64, 44, 44, 44, 44, 44, 44, 43, 43, 43, 80, 43, 43, 43, 85,
+   63,  2,  2, 44, 44, 44, 44, 44,  2, 36, 36, 36, 36, 36, 36, 36,
+   44, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 87, 43, 43, 43,
+   83, 43, 85, 80, 44, 44, 44, 44, 36, 36, 36, 61, 36, 62, 36, 36,
+   70, 43, 43, 80, 44, 80, 43, 57, 43, 43, 43, 70, 44, 44, 44, 44,
+   36, 36, 36, 62, 61, 36, 36, 36, 36, 36, 36, 36, 36, 84, 84, 88,
+   43, 87, 85, 85, 61, 44, 44, 44, 36, 70, 83,162, 64, 44, 44, 44,
+   27, 27, 89, 67, 67, 67, 56, 20,161, 67, 67, 67, 67, 67, 67, 67,
+   67, 44, 44, 44, 44, 44, 44, 91,103,103,103,103,103,103,103,177,
+    2,  2, 64, 44, 44, 44, 44, 44, 65, 65, 65, 65, 68, 44, 44, 44,
+   43, 43, 60, 44, 44, 44, 44, 44, 43, 43, 43, 60,  2,  2, 67, 67,
+   40, 40, 95, 44, 44, 44, 44, 44,  7,  7,  7,  7,  7,175, 27, 27,
+   27, 62, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 44, 44, 62, 36,
+   27, 27, 27, 30,  2, 64, 44, 44, 36, 36, 36, 36, 36, 61, 44, 57,
+   92, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84,
+   84, 84, 84, 84, 44, 44, 44, 57, 43, 74, 40, 40, 40, 40, 40, 40,
+   40, 86, 80, 44, 44, 44, 44, 44, 84, 44, 44, 44, 44, 44, 44, 44,
+   36, 61, 44, 44, 44, 44, 44, 44, 44, 44, 36, 36, 44, 44, 44, 44,
+   36, 36, 36, 36, 36, 44, 50, 60, 65, 65, 44, 44, 44, 44, 44, 44,
+   67, 67, 67, 90, 55, 67, 67, 67, 67, 67,182, 85, 43, 67,182, 84,
+   84,183, 65, 65, 65, 82, 43, 43, 43, 76, 50, 43, 43, 43, 67, 67,
+   67, 67, 67, 67, 67, 43, 43, 67, 67, 67, 67, 67, 90, 44, 44, 44,
+   67, 43, 76, 44, 44, 44, 44, 44, 27, 27, 44, 44, 44, 44, 44, 44,
+   11, 11, 11, 11, 11, 16, 16, 16, 16, 16, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 11, 16, 16, 16,108, 16, 16, 16, 16, 16,
+   11, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 47, 11,
+   44, 47, 48, 47, 48, 11, 47, 11, 11, 11, 11, 16, 16,146,146, 16,
+   16, 16,146, 16, 16, 16, 16, 16, 16, 16, 11, 48, 11, 47, 48, 11,
+   11, 11, 47, 11, 11, 11, 47, 16, 16, 16, 16, 16, 11, 48, 11, 47,
+   11, 11, 47, 47, 44, 11, 11, 11, 47, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 16, 11, 11, 11, 11, 11, 16, 16, 16, 16, 16,
+   16, 16, 16, 44, 11, 11, 11, 11, 31, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 33, 16, 16, 16, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 31, 16, 16, 16, 16, 33, 16, 16, 16, 11, 11,
+   11, 11, 31, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 33,
+   16, 16, 16, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 31,
+   16, 16, 16, 16, 33, 16, 16, 16, 11, 11, 11, 11, 31, 16, 16, 16,
+   16, 33, 16, 16, 16, 32, 44,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+   43, 43, 43, 76, 67, 50, 43, 43, 43, 43, 43, 43, 43, 43, 76, 67,
+   67, 67, 50, 67, 67, 67, 67, 67, 67, 67, 76, 21,  2,  2, 44, 44,
+   44, 44, 44, 44, 44, 57, 43, 43, 43, 43, 43, 80, 43, 43, 43, 43,
+   43, 43, 43, 43, 80, 57, 43, 43, 43, 57, 80, 43, 43, 80, 44, 44,
+   43, 43, 43, 74, 40, 40, 40, 44,  7,  7,  7,  7,  7, 44, 44, 77,
+   36, 36, 36, 36, 36, 36, 43, 43,  7,  7,  7,  7,  7, 44, 44, 94,
+   36, 36, 61,175, 27, 27, 27, 27, 43, 43, 43, 80, 44, 44, 44, 44,
+   16, 16, 43, 43, 43, 74, 44, 44, 27, 27, 27, 27, 27, 27,157, 27,
+  184, 27, 98, 44, 44, 44, 44, 44, 27, 27, 27, 27, 27, 27, 27,157,
+   27, 27, 27, 27, 27, 27, 27, 44, 36, 36, 62, 36, 36, 36, 36, 36,
+   62, 61, 61, 62, 62, 36, 36, 36, 36, 61, 36, 36, 62, 62, 44, 44,
+   44, 61, 44, 62, 62, 62, 62, 36, 62, 61, 61, 62, 62, 62, 62, 62,
+   62, 61, 61, 62, 36, 61, 36, 36, 36, 61, 36, 36, 62, 36, 61, 61,
+   36, 36, 36, 36, 36, 62, 36, 36, 62, 36, 62, 36, 36, 62, 36, 36,
+    8, 44, 44, 44, 44, 44, 44, 44, 55, 67, 67, 67, 67, 67, 67, 67,
+   27, 27, 27, 27, 27, 27, 89, 67, 67, 67, 67, 67, 67, 67, 67, 44,
+   44, 44, 44, 67, 67, 67, 67, 67, 67, 90, 44, 44, 44, 44, 44, 44,
+   67, 44, 44, 44, 44, 44, 44, 44, 67, 67, 67, 67, 67, 25, 41, 41,
+   67, 67, 67, 67, 44, 44, 44, 44, 67, 67, 67, 67, 90, 67, 67, 67,
+   67, 67, 67, 67, 67, 67, 55, 67, 67, 67, 90, 44, 67, 90, 44, 44,
+   67, 90, 67, 67, 67, 67, 67, 67, 79, 44, 44, 44, 44, 44, 44, 44,
+   65, 65, 65, 65, 65, 65, 65, 65,165,165,165,165,165,165,165, 44,
+  165,165,165,165,165,165,165,  0,  0,  0, 29, 21, 21, 21, 23, 21,
+   22, 18, 21, 25, 21, 17, 13, 13, 25, 25, 25, 21, 21,  9,  9,  9,
+    9, 22, 21, 18, 24, 16, 24,  5,  5,  5,  5, 22, 25, 18, 25,  0,
+   23, 23, 26, 21, 24, 26,  7, 20, 25,  1, 26, 24, 26, 25, 15, 15,
+   24, 15,  7, 19, 15, 21,  9, 25,  9,  5,  5, 25,  5,  9,  5,  7,
+    7,  7,  9,  8,  8,  5,  7,  5,  6,  6, 24, 24,  6, 24, 12, 12,
+    2,  2,  6,  5,  9, 21,  9,  2,  2,  9, 25,  9, 26, 12, 11, 11,
+    2,  6,  5, 21, 17,  2,  2, 26, 26, 23,  2, 12, 17, 12, 21, 12,
+   12, 21,  7,  2,  2,  7,  7, 21, 21,  2,  1,  1, 21, 23, 26, 26,
+    1,  2,  6,  7,  7, 12, 12,  7, 21,  7, 12,  1, 12,  6,  6, 12,
+   12, 26,  7, 26, 26,  7,  2,  1, 12,  2,  6,  2,  1, 12, 12, 10,
+   10, 10, 10, 12, 21,  6,  2, 10, 10,  2, 15, 26, 26,  2,  2, 21,
+    7, 10, 15,  7,  2, 23, 21, 26, 10,  7, 21, 15, 15,  2, 17,  7,
+   29,  7,  7, 22, 18,  2, 14, 14, 14,  7, 17, 21,  7,  6, 11, 12,
+    5,  2,  5,  6,  8,  8,  8, 24,  5, 24,  2, 24,  9, 24, 24,  2,
+   29, 29, 29,  1, 17, 17, 20, 19, 22, 20, 27, 28,  1, 29, 21, 20,
+   19, 21, 21, 16, 16, 21, 25, 22, 18, 21, 21, 29, 15,  6, 18,  6,
+   12, 11,  9, 26, 26,  9, 26,  5,  5, 26, 14,  9,  5, 14, 14, 15,
+   25, 26, 26, 22, 18, 26, 18, 25, 18, 22,  5, 12,  2,  5, 22, 21,
+   26,  6,  7, 14, 17, 22, 18, 18, 26, 14, 17,  6, 14,  6, 12, 24,
+   24,  6, 26, 15,  6, 21, 11, 21, 24,  9, 23, 26, 10, 21,  6, 10,
+    4,  4,  3,  3,  7, 25, 21, 22, 17, 16, 16, 22, 16, 16, 25, 17,
+   25,  2, 25, 24, 23,  2,  2, 15, 12, 15, 14,  2, 21, 14,  7, 15,
+   12, 17, 21,  1, 26, 10, 10,  1, 23, 15,  0,  1,  2,  3,  4,  5,
+    6,  7,  8,  9,  0, 10, 11, 12, 13,  0, 14,  0,  0,  0,  0,  0,
+   15,  0, 16,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 17, 18, 19,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 20,  0, 21, 22, 23,  0,  0,  0, 24,
+   25, 26, 27, 28, 29, 30, 31, 32, 33,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 34,
+    0, 35,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 36,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 37, 38,  0,  0,  0,  0,  0,  0, 39, 40,  0,  0, 41,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,  4,  0,  0,  0,  0,
+    0,  0,  0,  0,  5,  0,  0,  0,  0,  0,  0,  0,  6,  7,  8,  0,
+    9,  0, 10, 11,  0,  0, 12, 13, 14, 15, 16,  0,  0,  0,  0, 17,
+   18, 19, 20,  0,  0,  0, 21, 22,  0, 23, 24,  0,  0, 23, 25, 26,
+    0, 23, 25,  0,  0, 23, 25,  0,  0, 23, 25,  0,  0,  0, 25,  0,
+    0,  0, 27,  0,  0, 23, 25,  0,  0, 28, 25,  0,  0,  0, 29,  0,
+    0, 30, 31,  0,  0, 32, 33,  0, 34, 35,  0, 36, 37,  0, 38,  0,
+    0, 39,  0,  0, 40,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 41,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 42, 42,  0,  0,  0,  0, 43,  0,
+    0,  0,  0,  0,  0, 44,  0,  0,  0, 45,  0,  0,  0,  0,  0,  0,
+   46,  0,  0, 47,  0, 48, 49,  0,  0, 50, 51, 52,  0, 53,  0, 54,
+    0, 55,  0,  0,  0,  0, 56, 57,  0,  0,  0,  0,  0,  0, 58, 59,
+    0,  0,  0,  0,  0,  0, 60, 61,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 62,  0,  0,  0, 63,  0,  0,  0, 64,
+    0, 65,  0,  0, 66,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 67, 68,  0,  0, 69,  0,  0,  0,  0,  0,  0,  0,  0,
+   70, 71,  0,  0,  0,  0, 51, 72,  0, 73, 74,  0,  0, 75, 76,  0,
+    0,  0,  0,  0,  0, 77, 78, 79,  0,  0,  0,  0,  0,  0,  0, 25,
+    0,  0,  0,  0,  0,  0,  0,  0, 80,  0,  0,  0,  0,  0,  0,  0,
+    0, 81,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 82,
+    0,  0,  0,  0,  0,  0,  0, 49,  0,  0,  0, 83,  0,  0,  0,  0,
+   84, 85,  0,  0,  0,  0,  0, 86,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0, 87,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 88,  0,  0,  0,  0, 89,  0,  0,  0,  0,  0,
+    0,  0, 70, 63,  0, 90,  0,  0, 91, 92,  0, 75,  0,  0, 93,  0,
+    0, 94,  0,  0,  0,  0,  0, 95,  0, 96, 25, 97,  0,  0,  0,  0,
+    0,  0, 98,  0,  0,  0, 99,  0,  0,  0,  0,  0,  0, 63,100,  0,
+    0, 63,  0,  0,  0,101,  0,  0,  0,102,  0,  0,  0,  0,  0,  0,
+    0, 90,  0,  0,  0,  0,  0,  0,  0,103,104,  0,  0,  0,  0, 76,
+    0, 42,105,  0,106,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 63,  0,  0,  0,  0,  0,  0,  0,  0,107,  0,108,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,109,  0,110,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,111,
+    0,  0,  0,  0,112,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,113,114,115,  0,  0,
+    0,  0,116,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+  117,118,  0,  0,  0,  0,  0,  0,  0,110,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,119,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,120,  0,  0,  0,121,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,  1,  1,  2,  3,  4,
+    5,  6,  7,  4,  4,  8,  9, 10,  1, 11, 12, 13, 14, 15, 16, 17,
+   18,  1,  1,  1,  0,  0,  0,  0, 19,  1,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 20, 21, 22,  1, 23,  4, 21, 24, 25, 26, 27, 28,
+   29, 30,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  1, 31,  0,
+    0,  0, 32, 33, 34, 35,  1, 36,  0,  0,  0,  0, 37,  0,  0,  0,
+    0,  0,  0,  0,  0, 38,  1, 39, 14, 39, 40, 41,  0,  0,  0,  0,
+    0,  0,  0,  0, 42,  0,  0,  0,  0,  0,  0,  0, 43, 36, 44, 45,
+   21, 45, 46,  0,  0,  0,  0,  0,  0,  0, 19,  1, 21,  0,  0, 47,
+    0,  0,  0,  0,  0, 38, 48,  1,  1, 49, 49, 50,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 51,  0,  0,  0,  0,  0, 52,  1,  1,  1,
+   53, 21, 43, 54, 55, 21, 35,  1,  0,  0,  0,  0,  0,  0,  0, 56,
+    0,  0,  0, 57, 58, 59,  0,  0,  0,  0,  0, 57,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 60,  0,  0,  0, 57,  0, 61,  0,  0,
+    0,  0,  0,  0,  0,  0, 62, 63,  0,  0, 64,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 65,  0,  0,  0, 66,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 67,  0,  0,  0, 68,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 69,  0,  0,  0,  0,  0,  0, 70, 71,  0,
+    0,  0,  0,  0, 72, 73, 74, 75, 76, 77,  0,  0,  0,  0,  0,  0,
+    0, 78,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 79, 80,  0,
+    0,  0,  0, 47,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 49,
+    0,  0,  0,  0,  0, 63,  0,  0,  0,  0,  0,  0, 64,  0,  0, 81,
+    0,  0, 82,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 83,  0,
+    0,  0,  0,  0,  0, 19, 84,  0, 63,  0,  0,  0,  0, 49,  1, 85,
+    0,  0,  0,  0,  1, 54, 15, 86, 84,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 56,  0,  0,  0, 63,  0,  0,  0,  0,  0,  0,
+    0,  0, 19, 10,  1,  0,  0,  0,  0,  0, 87,  0,  0,  0,  0,  0,
+    0, 88,  0,  0, 87,  0,  0,  0,  0,  0,  0,  0,  0, 79,  0,  0,
+    0,  0,  0,  0, 89,  9, 12,  4, 90,  8, 91, 47,  0, 59, 50,  0,
+   21,  1, 21, 92, 93,  1,  1,  1,  1,  1,  1,  1,  1, 94, 95, 96,
+    0,  0,  0,  0, 97,  1, 98, 59, 81, 99,100,  4, 59,  0,  0,  0,
+    0,  0,  0, 19, 50,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 62,
+    1,  1,  1,  1,  1,  1,  1,  1,  0,  0,101,102,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,103,  0,  0,  0,  0, 19,  0,  1,  1, 50,
+    0,  0,  0,  0,  0,  0,  0, 38,  0,  0,  0,  0, 50,  0,  0,  0,
+    0, 64,  0,  0,  0,  0,  0,  0,  0,  0,  0, 63,  0,  0,  0,  0,
+    1,  1,  1,  1, 50,  0,  0,  0,  0,  0, 52, 69,  0,  0,  0,  0,
+    0,  0,  0,  0, 62,  0,  0,  0,  0,  0,  0,  0, 79,  0,  0,  0,
+   63,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,104,105, 59, 38,
+   81,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 64,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,106,  1, 14,  4, 12,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 47,  0,  0,  0,  0,  0, 38, 89,  0,
+    0,  0,  0,107,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,108, 62,
+    0,109,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  0,  0,  0,
+    0,  0, 19, 59,  0,  0,  0,  0,  0,110, 14, 54, 84,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,111,  0, 89,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 62, 63,  0,  0, 63,  0, 88,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,111,  0,  0,  0,  0,112,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 79, 56,  0, 38,  1, 59,  1, 59,  0,  0,
+   64, 88,  0,  0,  0,  0,  0, 60,113,  0,  0,  0,  0,  0,  0,  0,
+   56,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,113,  0,  0,
+    0,  0, 62,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 80,
+   79,  0,  0,  0,  0,  0,  0,  0,  0, 62,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 57,  0, 88,114,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0, 62,  0,  0,  0,  0,  0,  0,  8, 91,  0,  0,
+    0,  0,  0,  0,  1, 89,  0,  0,  0,  0,  0,  0,115,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,116,  0,117,118,119,120,  0, 52,  4,
+  121, 49, 23,  0,  0,  0,  0,  0,  0,  0, 38, 50,  0,  0,  0,  0,
+   38, 59,  0,  0,  0,  0,  0,  0,  1, 89,  1,  1,  1,  1, 39,  1,
+   48,104, 89,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  0,
+    0,  0,  0,  0,  4,121,  0,  0,  0,  1,122,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,230,230,230,230,230,232,220,220,220,220,232,216,
+  220,220,220,220,220,202,202,220,220,220,220,202,202,220,220,220,
+    1,  1,  1,  1,  1,220,220,220,220,230,230,230,230,240,230,220,
+  220,220,230,230,230,220,220,  0,230,230,230,220,220,220,220,230,
+  232,220,220,230,233,234,234,233,234,234,233,230,  0,  0,  0,230,
+    0,220,230,230,230,230,220,230,230,230,222,220,230,230,220,220,
+  230,222,228,230, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 19, 20,
+   21, 22,  0, 23,  0, 24, 25,  0,230,220,  0, 18, 30, 31, 32,  0,
+    0,  0,  0, 27, 28, 29, 30, 31, 32, 33, 34,230,230,220,220,230,
+  220,230,230,220, 35,  0,  0,  0,  0,  0,230,230,230,  0,  0,230,
+  230,  0,220,230,230,220,  0,  0,  0, 36,  0,  0,230,220,230,230,
+  220,220,230,220,220,230,220,230,220,230,230,  0,  0,220,  0,  0,
+  230,230,  0,230,  0,230,230,230,230,230,  0,  0,  0,220,220,220,
+    0,  0,  0,220,230,230,  0,220,230,220,220,220, 27, 28, 29,230,
+    7,  0,  0,  0,  0,  9,  0,  0,  0,230,220,230,230,  0,  0,  0,
+    0,  0,230,  0,  0, 84, 91,  0,  0,  0,  0,  9,  9,  0,  0,  0,
+    0,  0,  9,  0,103,103,  9,  0,107,107,107,107,118,118,  9,  0,
+  122,122,122,122,220,220,  0,  0,  0,220,  0,220,  0,216,  0,  0,
+    0,129,130,  0,132,  0,  0,  0,  0,  0,130,130,130,130,  0,  0,
+  130,  0,230,230,  9,  0,230,230,  0,  0,220,  0,  0,  0,  0,  7,
+    0,  9,  9,  0,  0,230,  0,  0,  0,228,  0,  0,  0,222,230,220,
+  220,  0,  0,  0,230,  0,  0,220,230,220,  0,220,  0,  0,  9,  9,
+    0,  0,  7,  0,230,230,230,  0,230,  0,  1,  1,  1,  0,  0,  0,
+  230,234,214,220,202,230,230,230,230,230,232,228,228,220,  0,230,
+  233,220,230,220,230,230,  1,  1,  1,  1,  1,230,  0,  1,  1,230,
+  220,230,  1,  1,  0,  0,218,228,232,222,224,224,  0,  8,  8,  0,
+  230,  0,230,230,220,  0,  0,230,  0,  0, 26,  0,  0,220,  0,230,
+  230,  1,220,  0,  0,230,220,  0,  0,  0,220,220,  0,  9,  7,  0,
+    0,  7,  9,  0,  0,  0,  9,  7,  9,  9,  0,  0,  6,  6,  0,  0,
+    0,  0,  1,  0,  0,216,216,  1,  1,  1,  0,  0,  0,226,216,216,
+  216,216,216,  0,220,220,220,  0,230,230,  7,  0, 16, 17, 17, 17,
+   17, 17, 17, 33, 17, 17, 17, 19, 17, 17, 17, 17, 20,101, 17,113,
+  129,169, 17, 27, 28, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,237,  0,  1,  2,  2,
+    0,  3,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  4,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  5,  0,  0,  0,  0,  6,  7,  8,
+    9,  0,  0,  0, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 20,  0,  0, 21, 22,  0,  0,  0,  0,
+   23, 24, 25, 26,  0, 27,  0, 28, 29, 30, 31, 32,  0,  0,  0,  0,
+    0,  0,  0, 33, 34, 35,  0,  0,  0,  0,  0,  0, 36,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 37, 38,  0,  0,  0,  0,  1,  2, 39, 40,
+    0,  1,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,
+    0,  0,  0,  2,  0,  0,  0,  0,  0,  0,  3,  4,  0,  0,  5,  0,
+    0,  0,  6,  0,  0,  0,  0,  0,  0,  0,  7,  1,  0,  0,  0,  0,
+    0,  0,  8,  9,  0,  0,  0,  0,  0,  0, 10,  0,  0, 10,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 10,  0,  0,  0, 10,
+    0,  0,  0,  0,  0,  0, 11, 12,  0, 13,  0, 14, 15, 16,  0,  0,
+    0,  0,  0,  1, 17, 18,  0, 19,  7,  1,  0,  0,  0, 20, 20,  7,
+   20, 20, 20, 20, 20, 20, 20,  8, 21,  0, 22,  0,  7, 23, 24,  0,
+   20, 20, 25,  0,  0,  0, 26, 27,  1,  7, 20, 20, 20, 20, 20,  1,
+   28, 29, 30, 31,  0,  0, 20,  0,  0,  0,  0,  0,  0,  0, 10,  0,
+    0,  0,  0,  0,  0,  0, 20, 20, 20,  1,  0,  0,  8, 21, 32,  4,
+    0, 10,  0, 33,  7, 20, 20, 20,  0,  0,  0,  0,  8, 34, 34, 35,
+   36, 34, 37,  0, 38,  1, 20, 20,  0,  0, 39,  0,  1,  1,  0,  8,
+   21,  1, 20,  0,  0,  0,  1,  0,  0, 40,  1,  1,  0,  0,  8, 21,
+    0,  1,  0,  1,  0,  1,  0,  0,  0,  0, 26, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 21,  7, 20, 41, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 21,  0, 42, 43, 44,  0, 45,  0,  8, 21,  0,  0,  0,  0,  0,
+    0,  0,  0, 46,  7,  1, 10,  1,  0,  0,  0,  1, 20, 20,  1,  0,
+    0,  0,  0,  0,  0,  0, 20, 20,  1, 20, 20,  0,  0,  0,  0,  0,
+    0,  0, 26, 21,  0,  1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  3, 47, 48,  0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,
+    4,  5,  6,  7,  7,  8,  7,  7,  7,  7,  7,  7,  7,  7,  7,  9,
+   10, 11, 12, 12, 12, 12, 13, 14, 14, 14, 14, 15, 16, 17, 18, 19,
+   20, 14, 21, 14, 22, 14, 14, 14, 14, 23, 24, 24, 25, 26, 14, 14,
+   14, 14, 27, 28, 14, 14, 29, 30, 31, 32, 33, 34,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+   35,  7, 36, 37,  7, 38,  7,  7,  7, 39, 14, 40,  7,  7, 41, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 42,  0,  0,  1,
+    2,  2,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15,
+   16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31,
+   32, 32, 33, 34, 35, 36, 37, 37, 37, 37, 37, 38, 39, 40, 41, 42,
+   43, 44, 45, 46, 47, 48, 49, 50, 51, 52,  2,  2, 53, 54, 55, 56,
+   57, 58, 59, 59, 59, 59, 60, 59, 59, 59, 59, 59, 59, 59, 61, 61,
+   59, 59, 59, 59, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73,
+   74, 75, 76, 77, 78, 59, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 79, 70, 70, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 80, 81, 81,
+   81, 81, 81, 81, 81, 81, 81, 82, 83, 83, 84, 85, 86, 87, 88, 89,
+   90, 91, 92, 93, 94, 95, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 96, 97, 97,
+   97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+   70, 70, 98, 99,100,101,102,102,103,104,105,106,107,108,109,110,
+  111,112, 97,113,114,115,116,117,118, 97,119,119,120, 97,121,122,
+  123,124,125,126,127,128,129,130,131, 97,132,133,134,135,136,137,
+  138,139,140,141,142, 97,143,144, 97,145,146,147,148, 97,149,150,
+  151,152,153,154, 97, 97,155,156,157,158, 97,159, 97,160,161,161,
+  161,161,161,161,161,162,163,161,164, 97, 97, 97, 97, 97,165,165,
+  165,165,165,165,165,165,166, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+   97, 97, 97, 97, 97, 97,167,167,167,167,168, 97, 97, 97,169,169,
+  169,169,170,171,172,173, 97, 97, 97, 97,174,175,176,177,178,178,
+  178,178,178,178,178,178,178,178,178,178,178,178,178,178,178,178,
+  178,178,178,178,178,178,178,178,178,178,178,178,178,179,178,178,
+  178,178,178,178,180,180,180,181,182, 97, 97, 97, 97, 97,183,184,
+  185,186,186,187, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,
+   97, 97, 97, 97, 97, 97,188,189, 97, 97, 97, 97, 97, 97, 59,190,
+  191,192,193,194,195, 97,196,197,198, 59, 59,199, 59,200,201,201,
+  201,201,201,202, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,203, 97,
+  204, 97, 97,205, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97,206,207,
+  208, 97, 97, 97, 97, 97,209,210,211, 97,212,213, 97, 97,214,215,
+   59,216,217, 97, 59, 59, 59, 59, 59, 59, 59,218,219,220,221,222,
+  223,224,225,226, 59,227, 97, 97, 97, 97, 97, 97, 97, 97, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,228, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,229, 70,230, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,
+   70, 70, 70, 70, 70, 70, 70, 70, 70, 70, 70,231, 70, 70, 70, 70,
+   70, 70, 70, 70, 70,232, 97, 97, 97, 97, 97, 97, 97, 97, 70, 70,
+   70, 70,233, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 97, 70, 70,
+   70, 70, 70, 70,234, 97, 97, 97, 97, 97, 97, 97, 97, 97,235, 97,
+  236,237,  0,  1,  2,  2,  0,  1,  2,  2,  2,  3,  4,  5,  0,  0,
+    0,  0,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19,  0,  0,  0,  0,  0,  0,  0,
+   19,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19, 19, 19,  0, 19,  0,
+    0,  0,  0,  0,  0,  0, 19, 19, 19, 19, 19,  0,  0,  0,  0,  0,
+   26, 26,  0,  0,  0,  0,  1,  1,  1,  1,  1,  1,  1,  1,  9,  9,
+    9,  9,  0,  9,  9,  9,  2,  2,  9,  9,  9,  9,  0,  9,  2,  2,
+    2,  2,  9,  0,  9,  0,  9,  9,  9,  2,  9,  2,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  2,  9,  9,  9,  9,  9,  9,  9,
+   55, 55, 55, 55, 55, 55, 55, 55, 55, 55, 55, 55, 55, 55,  6,  6,
+    6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  6,  1,  1,  6,  2,  4,
+    4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,  4,
+    4,  4,  4,  4,  4,  2,  4,  4,  4,  2,  2,  4,  4,  4,  2, 14,
+   14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,  2,  2,
+    2,  2,  2,  2,  2,  2, 14, 14, 14,  2,  2,  2,  2, 14, 14, 14,
+   14, 14, 14,  2,  2,  2,  3,  3,  3,  3,  3,  0,  3,  3,  3,  3,
+    3,  3,  0,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  0,  3,  2,  3,  0,  0,  3,  3,  3,  3,  3,  3,  3,  3,  3,
+    3,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  3,  3,  1,  3,
+    3,  3,  3,  3,  3,  3, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37,
+   37, 37, 37, 37,  2, 37, 37, 37, 37,  2,  2, 37, 37, 37, 38, 38,
+   38, 38, 38, 38, 38, 38, 38, 38,  2,  2,  2,  2,  2,  2, 64, 64,
+   64, 64, 64, 64, 64, 64, 64, 64, 64,  2,  2, 64, 64, 64, 90, 90,
+   90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90, 90,  2,  2, 90, 90,
+   90, 90, 90, 90, 90,  2, 95, 95, 95, 95, 95, 95, 95, 95, 95, 95,
+   95, 95,  2,  2, 95,  2, 37, 37, 37,  2,  2,  2,  2,  2,  3,  3,
+    3,  3,  3,  2,  3,  3,  2,  2,  2,  3,  3,  3,  3,  3,  3,  3,
+    0,  3,  3,  3,  3,  3,  7,  7,  7,  7,  7,  7,  7,  7,  7,  1,
+    1,  1,  1,  7,  7,  7,  7,  7,  7,  7,  0,  0,  7,  7,  5,  5,
+    5,  5,  2,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2,  5,  5,  2,
+    2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,
+    5,  5,  5,  5,  5,  5,  5,  2,  5,  2,  2,  2,  5,  5,  5,  5,
+    2,  2,  5,  5,  5,  5,  5,  2,  2,  5,  5,  5,  5,  2,  2,  2,
+    2,  2,  2,  2,  2,  5,  2,  2,  2,  2,  5,  5,  2,  5,  5,  5,
+    5,  5,  2,  2,  5,  5,  5,  5,  5,  5,  5,  5,  5,  2,  2, 11,
+   11, 11,  2, 11, 11, 11, 11, 11, 11,  2,  2,  2,  2, 11, 11,  2,
+    2, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,  2,
+   11, 11, 11, 11, 11, 11, 11,  2, 11, 11,  2, 11, 11,  2, 11, 11,
+    2,  2, 11,  2, 11, 11, 11,  2,  2, 11, 11, 11,  2,  2,  2, 11,
+    2,  2,  2,  2,  2,  2,  2, 11, 11, 11, 11,  2, 11,  2,  2,  2,
+    2,  2,  2,  2, 11, 11, 11, 11, 11, 11, 11, 11, 11,  2,  2, 10,
+   10, 10,  2, 10, 10, 10, 10, 10, 10, 10, 10, 10,  2, 10, 10, 10,
+    2, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  2,
+   10, 10, 10, 10, 10, 10, 10,  2, 10, 10,  2, 10, 10, 10, 10, 10,
+    2,  2, 10, 10, 10, 10, 10, 10,  2, 10, 10, 10,  2,  2, 10,  2,
+    2,  2,  2,  2,  2,  2, 10, 10, 10, 10,  2,  2, 10, 10, 10, 10,
+    2,  2,  2,  2,  2,  2,  2, 10, 10, 10, 10, 10, 10, 10,  2, 21,
+   21, 21,  2, 21, 21, 21, 21, 21, 21, 21, 21,  2,  2, 21, 21,  2,
+    2, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,  2,
+   21, 21, 21, 21, 21, 21, 21,  2, 21, 21,  2, 21, 21, 21, 21, 21,
+    2,  2, 21, 21, 21, 21, 21,  2,  2, 21, 21, 21,  2,  2,  2,  2,
+    2,  2,  2, 21, 21, 21,  2,  2,  2,  2, 21, 21,  2, 21, 21, 21,
+   21, 21,  2,  2, 21, 21,  2,  2, 22, 22,  2, 22, 22, 22, 22, 22,
+   22,  2,  2,  2, 22, 22, 22,  2, 22, 22, 22, 22,  2,  2,  2, 22,
+   22,  2, 22,  2, 22, 22,  2,  2,  2, 22, 22,  2,  2,  2, 22, 22,
+   22, 22, 22, 22, 22, 22, 22, 22,  2,  2,  2,  2, 22, 22, 22,  2,
+    2,  2,  2,  2,  2, 22,  2,  2,  2,  2,  2,  2, 22, 22, 22, 22,
+   22,  2,  2,  2,  2,  2, 23, 23, 23, 23, 23, 23, 23, 23, 23, 23,
+   23, 23, 23,  2, 23, 23, 23,  2, 23, 23, 23, 23, 23, 23, 23, 23,
+    2,  2,  2, 23, 23, 23, 23,  2, 23, 23, 23, 23,  2,  2,  2,  2,
+    2,  2,  2, 23, 23,  2, 23, 23, 23,  2,  2,  2,  2,  2, 23, 23,
+   23, 23,  2,  2, 23, 23,  2,  2,  2,  2,  2,  2,  2, 23, 16, 16,
+   16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,  2, 16, 16, 16,  2,
+   16, 16, 16, 16, 16, 16, 16, 16, 16, 16,  2, 16, 16, 16, 16, 16,
+    2,  2, 16, 16, 16, 16, 16,  2, 16, 16, 16, 16,  2,  2,  2,  2,
+    2,  2,  2, 16, 16,  2,  2,  2,  2,  2,  2,  2, 16,  2, 16, 16,
+   16, 16,  2,  2, 16, 16,  2, 16, 16,  2,  2,  2,  2,  2, 20, 20,
+   20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20,  2, 20, 20, 20,  2,
+   20, 20, 20, 20, 20, 20,  2,  2,  2,  2, 20, 20, 20, 20, 20, 20,
+   20, 20,  2,  2, 20, 20,  2, 36, 36, 36,  2, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,  2,  2,  2,
+   36, 36, 36, 36, 36, 36, 36, 36,  2, 36, 36, 36, 36, 36, 36, 36,
+   36, 36,  2, 36,  2,  2,  2,  2, 36,  2,  2,  2,  2, 36, 36, 36,
+   36, 36, 36,  2, 36,  2,  2,  2,  2,  2,  2,  2, 36, 36,  2,  2,
+   36, 36, 36,  2,  2,  2,  2, 24, 24, 24, 24, 24, 24, 24, 24, 24,
+   24, 24, 24, 24, 24, 24, 24, 24, 24,  2,  2,  2,  2,  0, 24, 24,
+   24, 24,  2,  2,  2,  2,  2, 18, 18,  2, 18,  2, 18, 18, 18, 18,
+   18,  2, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18,
+   18, 18,  2, 18,  2, 18, 18, 18, 18, 18, 18, 18,  2,  2, 18, 18,
+   18, 18, 18,  2, 18,  2, 18, 18,  2,  2, 18, 18, 18, 18, 25, 25,
+   25, 25, 25, 25, 25, 25,  2, 25, 25, 25, 25, 25, 25, 25, 25, 25,
+   25, 25, 25,  2,  2,  2, 25, 25, 25, 25, 25,  2, 25, 25, 25, 25,
+   25, 25, 25,  0,  0,  0,  0, 25, 25,  2,  2,  2,  2,  2, 33, 33,
+   33, 33, 33, 33, 33, 33,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    8,  8,  8,  8,  2,  8,  2,  2,  2,  2,  2,  8,  2,  2,  8,  8,
+    8,  0,  8,  8,  8,  8, 12, 12, 12, 12, 12, 12, 12, 12, 30, 30,
+   30, 30, 30, 30, 30, 30, 30,  2, 30, 30, 30, 30,  2,  2, 30, 30,
+   30, 30, 30, 30, 30,  2, 30, 30, 30,  2,  2, 30, 30, 30, 30, 30,
+   30, 30, 30,  2,  2,  2, 30, 30,  2,  2,  2,  2,  2,  2, 29, 29,
+   29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29, 29,  2,  2, 28, 28,
+   28, 28, 28, 28, 28, 28, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34,  2,  2,  2, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35,
+   35,  0,  0,  0, 35, 35, 35,  2,  2,  2,  2,  2,  2,  2, 45, 45,
+   45, 45, 45, 45, 45, 45, 45, 45, 45, 45, 45,  2, 45, 45, 45, 45,
+   45, 45, 45,  2,  2,  2, 44, 44, 44, 44, 44, 44, 44, 44, 44, 44,
+   44, 44, 44,  0,  0,  2, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43,
+   43, 43,  2,  2,  2,  2, 46, 46, 46, 46, 46, 46, 46, 46, 46, 46,
+   46, 46, 46,  2, 46, 46, 46,  2, 46, 46,  2,  2,  2,  2, 31, 31,
+   31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31, 31,  2,  2, 31, 31,
+    2,  2,  2,  2,  2,  2, 32, 32,  0,  0, 32,  0, 32, 32, 32, 32,
+   32, 32, 32, 32, 32,  2, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+    2,  2,  2,  2,  2,  2, 32,  2,  2,  2,  2,  2,  2,  2, 32, 32,
+   32,  2,  2,  2,  2,  2, 28, 28, 28, 28, 28, 28,  2,  2, 48, 48,
+   48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48,  2, 48, 48,
+   48, 48,  2,  2,  2,  2, 48,  2,  2,  2, 48, 48, 48, 48, 52, 52,
+   52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52, 52,  2,  2, 52, 52,
+   52, 52, 52,  2,  2,  2, 58, 58, 58, 58, 58, 58, 58, 58, 58, 58,
+   58, 58,  2,  2,  2,  2, 58, 58,  2,  2,  2,  2,  2,  2, 58, 58,
+   58,  2,  2,  2, 58, 58, 54, 54, 54, 54, 54, 54, 54, 54, 54, 54,
+   54, 54,  2,  2, 54, 54, 91, 91, 91, 91, 91, 91, 91, 91, 91, 91,
+   91, 91, 91, 91, 91,  2, 91, 91, 91, 91, 91,  2,  2, 91, 91, 91,
+    2,  2,  2,  2,  2,  2, 91, 91, 91, 91, 91, 91,  2,  2,  1,  2,
+    2,  2,  2,  2,  2,  2, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62,  2,  2,  2,  2, 62, 62, 62, 62, 62,  2,  2,  2, 76, 76,
+   76, 76, 76, 76, 76, 76, 93, 93, 93, 93, 93, 93, 93, 93, 93, 93,
+   93, 93,  2,  2,  2,  2,  2,  2,  2,  2, 93, 93, 93, 93, 70, 70,
+   70, 70, 70, 70, 70, 70,  2,  2,  2, 70, 70, 70, 70, 70, 70, 70,
+    2,  2,  2, 70, 70, 70, 73, 73, 73, 73, 73, 73, 73, 73,  6,  2,
+    2,  2,  2,  2,  2,  2,  8,  8,  8,  2,  2,  8,  8,  8,  1,  1,
+    1,  0,  1,  1,  1,  1,  1,  0,  1,  1,  1,  1,  1,  1,  1,  0,
+    0,  0,  0,  1,  0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  1,  1,
+    0,  2,  2,  2,  2,  2, 19, 19, 19, 19, 19, 19,  9,  9,  9,  9,
+    9,  6, 19, 19, 19, 19, 19, 19, 19, 19, 19,  9,  9,  9,  9,  9,
+   19, 19, 19, 19,  9,  9,  9,  9,  9, 19, 19, 19, 19, 19,  6, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19,  9,  1,  1,
+    2,  1,  1,  1,  1,  1,  9,  9,  9,  9,  9,  9,  2,  2,  2,  9,
+    2,  9,  2,  9,  2,  9,  9,  9,  9,  9,  9,  2,  9,  9,  9,  9,
+    9,  9,  2,  2,  9,  9,  9,  9,  9,  9,  2,  9,  9,  9,  2,  2,
+    9,  9,  9,  2,  9,  9,  9,  9,  9,  9,  9,  9,  9,  2,  0,  0,
+    0,  0,  1,  1,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0, 19,
+    2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 19,  0,  0,
+    0,  0,  0,  0,  0,  2, 19, 19, 19, 19, 19,  2,  2,  2,  0,  0,
+    0,  0,  0,  0,  9,  0,  0,  0, 19, 19,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 19,  0, 19,  0,  0,  0,  2,  2,  2,  2,  0,  0,
+    0,  2,  2,  2,  2,  2, 27, 27, 27, 27, 27, 27, 27, 27,  0,  0,
+    0,  0,  2,  2,  0,  0,  0,  0,  0,  0,  0,  0,  2,  0, 56, 56,
+   56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56,  2, 55, 55,
+   55, 55,  2,  2,  2,  2,  2, 55, 55, 55, 55, 55, 55, 55, 61, 61,
+   61, 61, 61, 61, 61, 61,  2,  2,  2,  2,  2,  2,  2, 61, 61,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    2, 13, 13, 13, 13, 13, 13, 13, 13, 13,  2,  2,  2,  2, 13, 13,
+   13, 13, 13, 13,  2,  2,  0,  0,  0,  0,  2,  2,  2,  2,  0,  0,
+    0,  0,  0, 13,  0, 13,  0, 13, 13, 13, 13, 13, 13, 13, 13, 13,
+    1,  1,  1,  1, 12, 12, 13, 13, 13, 13,  0,  0,  0,  0,  2, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15,  2,  2,  1,  1,  0,  0, 15, 15, 15,  0, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17,  0,  0, 17, 17, 17,  2,  2,  2,  2,  2, 26, 26, 26, 26, 26,
+   26, 26, 26, 26, 26, 26,  2, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12,  2, 12, 12, 12, 12, 12, 12, 12,  0, 17, 17,
+   17, 17, 17, 17, 17,  0, 13, 13, 13, 13, 13,  2,  2,  2, 39, 39,
+   39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39,  2,  2,  2, 39, 39,
+   39, 39, 39, 39, 39,  2, 86, 86, 86, 86, 86, 86, 86, 86, 77, 77,
+   77, 77, 77, 77, 77, 77, 77, 77, 77, 77,  2,  2,  2,  2, 79, 79,
+   79, 79, 79, 79, 79, 79,  0,  0, 19, 19, 19, 19, 19, 19,  0,  0,
+    0, 19, 19, 19, 19, 19,  2,  2, 19, 19, 19, 19, 19, 19, 19, 19,
+   19,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 19, 19, 19, 60, 60,
+   60, 60, 60, 60, 60, 60, 60, 60, 60, 60, 60,  2,  2,  2,  0,  0,
+    2,  2,  2,  2,  2,  2, 65, 65, 65, 65, 65, 65, 65, 65, 75, 75,
+   75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75,  2,  2,  2,  2,
+    2,  2,  2,  2, 75, 75, 75, 75,  2,  2,  2,  2,  2,  2, 69, 69,
+   69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69, 69,  0, 69, 74, 74,
+   74, 74, 74, 74, 74, 74, 74, 74, 74, 74,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 74, 12, 12, 12, 12, 12,  2,  2,  2, 84, 84,
+   84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84, 84,  2,  0, 84, 84,
+    2,  2,  2,  2, 84, 84, 33, 33, 33, 33, 33, 33, 33,  2, 68, 68,
+   68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68, 68,  2, 68, 68,
+   68, 68, 68, 68,  2,  2, 68, 68,  2,  2, 68, 68, 68, 68, 92, 92,
+   92, 92, 92, 92, 92, 92, 92, 92, 92,  2,  2,  2,  2,  2,  2,  2,
+    2, 92, 92, 92, 92, 92, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+   87, 87, 87, 87, 87,  2,  2, 30, 30, 30, 30, 30, 30,  2, 19, 19,
+   19,  0, 19, 19, 19, 19, 19, 19, 19, 19, 19,  9, 19, 19, 19, 19,
+    0,  0,  2,  2,  2,  2, 87, 87, 87, 87, 87, 87,  2,  2, 87, 87,
+    2,  2,  2,  2,  2,  2, 12, 12, 12, 12,  2,  2,  2,  2,  2,  2,
+    2, 12, 12, 12, 12, 12, 13, 13,  2,  2,  2,  2,  2,  2, 19, 19,
+   19, 19, 19, 19, 19,  2,  2,  2,  2,  4,  4,  4,  4,  4,  2,  2,
+    2,  2,  2, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14,  2, 14, 14,
+   14, 14, 14,  2, 14,  2, 14, 14,  2, 14, 14,  2, 14, 14,  3,  3,
+    2,  2,  2,  2,  2,  2,  3,  3,  3,  3,  3,  3,  0,  0,  2,  2,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  2,  2,  1,  1,
+    1,  1,  1,  1,  6,  6,  0,  0,  0,  2,  0,  0,  0,  0,  3,  3,
+    3,  3,  3,  2,  2,  0,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 17, 17, 17, 17, 17, 17, 17, 17,  0,  0,  2,  2,
+   12, 12, 12, 12, 12, 12,  2,  2, 12, 12, 12,  2,  2,  2,  2,  0,
+    0,  0,  0,  0,  2,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,
+   49, 49,  2, 49, 49, 49, 49, 49, 49, 49, 49, 49, 49,  2, 49, 49,
+   49,  2, 49, 49,  2, 49, 49, 49, 49, 49, 49, 49,  2,  2, 49, 49,
+   49,  2,  2,  2,  2,  2,  0,  0,  0,  2,  2,  2,  2,  0,  0,  0,
+    0,  0,  2,  2,  2,  0,  0,  0,  0,  0,  0,  2,  2,  2,  9,  2,
+    2,  2,  2,  2,  2,  2,  0,  0,  0,  0,  0,  1,  2,  2, 71, 71,
+   71, 71, 71, 71, 71, 71, 71, 71, 71, 71, 71,  2,  2,  2, 67, 67,
+   67, 67, 67, 67, 67, 67, 67,  2,  2,  2,  2,  2,  2,  2,  1,  0,
+    0,  0,  0,  0,  0,  0, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42,
+   42, 42,  2,  2,  2,  2,  2,  2,  2,  2,  2, 42, 42, 42, 41, 41,
+   41, 41, 41, 41, 41, 41, 41, 41, 41,  2,  2,  2,  2,  2,118,118,
+  118,118,118,118,118,118,118,118,118,  2,  2,  2,  2,  2, 53, 53,
+   53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53,  2, 53, 59, 59,
+   59, 59, 59, 59, 59, 59, 59, 59, 59, 59,  2,  2,  2,  2, 59, 59,
+   59, 59, 59, 59,  2,  2, 40, 40, 40, 40, 40, 40, 40, 40, 51, 51,
+   51, 51, 51, 51, 51, 51, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50,
+   50, 50, 50, 50,  2,  2, 50, 50,  2,  2,  2,  2,  2,  2,135,135,
+  135,135,135,135,135,135,135,135,135,135,  2,  2,  2,  2,106,106,
+  106,106,106,106,106,106,104,104,104,104,104,104,104,104,104,104,
+  104,104,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,104,110,110,
+  110,110,110,110,110,110,110,110,110,110,110,110,110,  2,110,110,
+  110,110,110,110,  2,  2, 47, 47, 47, 47, 47, 47,  2,  2, 47,  2,
+   47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47, 47,
+   47, 47, 47, 47,  2, 47, 47,  2,  2,  2, 47,  2,  2, 47, 81, 81,
+   81, 81, 81, 81, 81, 81, 81, 81, 81, 81, 81, 81,  2, 81,120,120,
+  120,120,120,120,120,120,116,116,116,116,116,116,116,116,116,116,
+  116,116,116,116,116,  2,  2,  2,  2,  2,  2,  2,  2,116,128,128,
+  128,128,128,128,128,128,128,128,128,  2,128,128,  2,  2,  2,  2,
+    2,128,128,128,128,128, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,
+   66, 66,  2,  2,  2, 66, 72, 72, 72, 72, 72, 72, 72, 72, 72, 72,
+    2,  2,  2,  2,  2, 72, 98, 98, 98, 98, 98, 98, 98, 98, 97, 97,
+   97, 97, 97, 97, 97, 97,  2,  2,  2,  2, 97, 97, 97, 97,  2,  2,
+   97, 97, 97, 97, 97, 97, 57, 57, 57, 57,  2, 57, 57,  2,  2,  2,
+    2,  2, 57, 57, 57, 57, 57, 57, 57, 57,  2, 57, 57, 57,  2, 57,
+   57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57, 57,
+   57, 57, 57, 57,  2,  2, 57, 57, 57,  2,  2,  2,  2, 57, 57,  2,
+    2,  2,  2,  2,  2,  2, 88, 88, 88, 88, 88, 88, 88, 88,117,117,
+  117,117,117,117,117,117,112,112,112,112,112,112,112,112,112,112,
+  112,112,112,112,112,  2,  2,  2,  2,112,112,112,112,112, 78, 78,
+   78, 78, 78, 78, 78, 78, 78, 78, 78, 78, 78, 78,  2,  2,  2, 78,
+   78, 78, 78, 78, 78, 78, 83, 83, 83, 83, 83, 83, 83, 83, 83, 83,
+   83, 83, 83, 83,  2,  2, 82, 82, 82, 82, 82, 82, 82, 82, 82, 82,
+   82,  2,  2,  2,  2,  2,122,122,122,122,122,122,122,122,122,122,
+    2,  2,  2,  2,  2,  2,  2,122,122,122,122,  2,  2,  2,  2,122,
+  122,122,122,122,122,122, 89, 89, 89, 89, 89, 89, 89, 89, 89,  2,
+    2,  2,  2,  2,  2,  2,130,130,130,130,130,130,130,130,130,130,
+  130,  2,  2,  2,  2,  2,  2,  2,130,130,130,130,130,130,144,144,
+  144,144,144,144,144,144,144,144,  2,  2,  2,  2,  2,  2,  3,  3,
+    3,  3,  3,  3,  3,  2,156,156,156,156,156,156,156,156,156,156,
+    2,156,156,156,  2,  2,156,156,  2,  2,  2,  2,  2,  2,147,147,
+  147,147,147,147,147,147,148,148,148,148,148,148,148,148,148,148,
+    2,  2,  2,  2,  2,  2,153,153,153,153,153,153,153,153,153,153,
+  153,153,  2,  2,  2,  2,149,149,149,149,149,149,149,149,149,149,
+  149,149,149,149,149,  2, 94, 94, 94, 94, 94, 94, 94, 94, 94, 94,
+   94, 94, 94, 94,  2,  2,  2,  2, 94, 94, 94, 94, 94, 94,  2,  2,
+    2,  2,  2,  2,  2, 94, 85, 85, 85, 85, 85, 85, 85, 85, 85, 85,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 85,  2,  2,101,101,
+  101,101,101,101,101,101,101,  2,  2,  2,  2,  2,  2,  2,101,101,
+    2,  2,  2,  2,  2,  2, 96, 96, 96, 96, 96, 96, 96, 96, 96, 96,
+   96, 96, 96,  2, 96, 96,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,  2,100,100,100,100,100,100,100,100,  2, 36,
+   36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,  2,  2,  2,108,108,
+  108,108,108,108,108,108,108,108,  2,108,108,108,108,108,108,108,
+  108,108,108,108,108,  2,129,129,129,129,129,129,129,  2,129,  2,
+  129,129,129,129,  2,129,129,129,129,129,129,129,129,129,129,129,
+  129,129,129,129,  2,129,129,129,  2,  2,  2,  2,  2,  2,109,109,
+  109,109,109,109,109,109,109,109,109,  2,  2,  2,  2,  2,109,109,
+    2,  2,  2,  2,  2,  2,107,107,107,107,  2,107,107,107,107,107,
+  107,107,107,  2,  2,107,107,  2,  2,107,107,107,107,107,107,107,
+  107,107,107,107,107,107,107,  2,107,107,107,107,107,107,107,  2,
+  107,107,  2,107,107,107,107,107,  2,  1,107,107,107,107,107,  2,
+    2,107,107,107,  2,  2,107,  2,  2,  2,  2,  2,  2,107,  2,  2,
+    2,  2,  2,107,107,107,107,107,107,107,  2,  2,107,107,107,107,
+  107,107,107,  2,  2,  2,137,137,137,137,137,137,137,137,137,137,
+  137,137,  2,137,137,137,137,137,  2,  2,  2,  2,  2,  2,124,124,
+  124,124,124,124,124,124,124,124,  2,  2,  2,  2,  2,  2,123,123,
+  123,123,123,123,123,123,123,123,123,123,123,123,  2,  2,114,114,
+  114,114,114,114,114,114,114,114,114,114,114,  2,  2,  2,114,114,
+    2,  2,  2,  2,  2,  2, 32, 32, 32, 32, 32,  2,  2,  2,102,102,
+  102,102,102,102,102,102,102,  2,  2,  2,  2,  2,  2,  2,102,102,
+    2,  2,  2,  2,  2,  2,126,126,126,126,126,126,126,126,126,126,
+  126,  2,  2,126,126,126,126,126,126,126,  2,  2,  2,  2,142,142,
+  142,142,142,142,142,142,142,142,142,142,  2,  2,  2,  2,125,125,
+  125,125,125,125,125,125,125,125,125,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2,  2,125,154,154,154,154,154,154,154,  2,  2,154,
+    2,  2,154,154,154,154,154,154,154,154,  2,154,154,  2,154,154,
+  154,154,154,154,154,154,154,154,154,154,154,154,  2,154,154,  2,
+    2,154,154,154,154,154,154,154,  2,  2,  2,  2,  2,  2,150,150,
+  150,150,150,150,150,150,  2,  2,150,150,150,150,150,150,150,150,
+  150,150,150,  2,  2,  2,141,141,141,141,141,141,141,141,140,140,
+  140,140,140,140,140,140,140,140,140,  2,  2,  2,  2,  2,121,121,
+  121,121,121,121,121,121,121,  2,  2,  2,  2,  2,  2,  2,133,133,
+  133,133,133,133,133,133,133,  2,133,133,133,133,133,133,133,133,
+  133,133,133,133,133,  2,133,133,133,133,133,133,  2,  2,133,133,
+  133,133,133,  2,  2,  2,134,134,134,134,134,134,134,134,  2,  2,
+  134,134,134,134,134,134,  2,134,134,134,134,134,134,134,134,134,
+  134,134,134,134,134,  2,138,138,138,138,138,138,138,  2,138,138,
+    2,138,138,138,138,138,138,138,138,138,138,138,138,138,  2,  2,
+  138,  2,138,138,  2,138,138,138,  2,  2,  2,  2,  2,  2,143,143,
+  143,143,143,143,  2,143,143,  2,143,143,143,143,143,143,143,143,
+  143,143,143,143,143,143,143,143,143,143,143,143,143,  2,143,143,
+    2,143,143,143,143,143,143,  2,  2,  2,  2,  2,  2,  2,143,143,
+    2,  2,  2,  2,  2,  2,145,145,145,145,145,145,145,145,145,  2,
+    2,  2,  2,  2,  2,  2, 86,  2,  2,  2,  2,  2,  2,  2, 22, 22,
+    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2, 22, 63, 63,
+   63, 63, 63, 63, 63, 63, 63, 63,  2,  2,  2,  2,  2,  2, 63, 63,
+   63, 63, 63, 63, 63,  2, 63, 63, 63, 63, 63,  2,  2,  2, 63, 63,
+   63, 63,  2,  2,  2,  2, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80,
+   80, 80, 80, 80, 80,  2, 80,  2,  2,  2,  2,  2,  2,  2,127,127,
+  127,127,127,127,127,127,127,127,127,127,127,127,127,  2, 79,  2,
+    2,  2,  2,  2,  2,  2,115,115,115,115,115,115,115,115,115,115,
+  115,115,115,115,115,  2,115,115,  2,  2,  2,  2,115,115,103,103,
+  103,103,103,103,103,103,103,103,103,103,103,103,  2,  2,119,119,
+  119,119,119,119,119,119,119,119,119,119,119,119,  2,  2,119,119,
+    2,119,119,119,119,119,  2,  2,  2,  2,  2,119,119,119,146,146,
+  146,146,146,146,146,146,146,146,146,  2,  2,  2,  2,  2, 99, 99,
+   99, 99, 99, 99, 99, 99, 99, 99, 99,  2,  2,  2,  2, 99,  2,  2,
+    2,  2,  2,  2,  2, 99,136,139,  0,  0,155,  2,  2,  2,136,136,
+  136,136,136,136,136,136,155,155,155,155,155,155,155,155,155,155,
+  155,155,155,155,  2,  2,136,  2,  2,  2,  2,  2,  2,  2, 17, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15,  2,  2,  2,  2,  2,  2,  2,
+    2,  2, 17, 17, 17, 17,139,139,139,139,139,139,139,139,139,139,
+  139,139,  2,  2,  2,  2,105,105,105,105,105,105,105,105,105,105,
+  105,  2,  2,  2,  2,  2,105,105,105,105,105,  2,  2,  2,105,  2,
+    2,  2,  2,  2,  2,  2,105,105,  2,  2,105,105,105,105,  0,  0,
+    0,  0,  0,  0,  2,  2,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,  1,  1,  1,  1,
+    1,  0,  0,  1,  1,  1,  1,  1,  1,  1,  0,  0,  0,  0,  0,  0,
+    1,  1,  1,  1,  0,  0,  0,  2,  2,  2,  2,  2,  2,  2,  2,  2,
+    0,  2,  2,  0,  0,  2,  2,  0,  0,  0,  0,  2,  0,  0,  0,  0,
+    2,  0,  2,  0,  0,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,  0,
+    0,  2,  2,  0,  0,  0,  0,  0,  2,  0,  0,  0,  0,  2,  0,  0,
+    0,  0,  0,  2,  0,  2,  2,  2,  0,  0,  0,  0,  0,  0,  0,  2,
+    0,  0,  0,  0,  0,  0,131,131,131,131,131,131,131,131,131,131,
+  131,131,  2,  2,  2,  2,  2,  2,  2,131,131,131,131,131,  2,131,
+  131,131,131,131,131,131, 56,  2,  2, 56, 56, 56, 56, 56, 56, 56,
+    2, 56, 56,  2, 56, 56, 56, 56, 56,  2,  2,  2,  2,  2,151,151,
+  151,151,151,151,151,151,151,151,151,151,151,  2,  2,  2,151,151,
+  151,151,151,151,  2,  2,151,151,  2,  2,  2,  2,151,151,152,152,
+  152,152,152,152,152,152,152,152,  2,  2,  2,  2,  2,152,113,113,
+  113,113,113,113,113,113,113,113,113,113,113,  2,  2,113,113,113,
+  113,113,113,113,113,  2,132,132,132,132,132,132,132,132,132,132,
+  132,132,  2,  2,  2,  2,132,132,  2,  2,  2,  2,132,132,  3,  3,
+    3,  3,  2,  3,  3,  3,  2,  3,  3,  2,  3,  2,  2,  3,  2,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  2,  3,  3,  3,  3,  2,  3,
+    2,  3,  2,  2,  2,  2,  2,  2,  3,  2,  2,  2,  2,  3,  2,  3,
+    2,  3,  2,  3,  3,  3,  2,  3,  2,  3,  2,  3,  2,  3,  2,  3,
+    3,  3,  3,  2,  3,  2,  3,  3,  2,  3,  3,  3,  3,  3,  3,  3,
+    3,  3,  2,  2,  2,  2,  2,  3,  3,  3,  2,  3,  3,  3,  2,  2,
+    2,  2,  2,  2,  0,  0, 15,  0,  0,  2,  2,  2,  2,  2, 13,  2,
+    2,  2,  2,  2,  2,  2, 13, 13, 13,  2,  2,  2,  2,  2,  2,  0,
+    2,  2,  2,  2,  2,  2,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9,
+    9,  9,  9, 10,  9, 11, 12, 13,  9,  9,  9, 14,  9,  9, 15,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9, 16, 17,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 18, 19,
+   20,  9, 21,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 22,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9, 23, 24,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12,  0,  0, 13, 14,
+   15, 16, 17, 18, 19, 20, 21, 22,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 23,  0,  0, 24, 25, 26, 27, 28,
+   29, 30,  0,  0, 31, 32,  0, 33,  0, 34,  0, 35,  0,  0,  0,  0,
+   36, 37, 38, 39,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 40,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   41, 42,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 43, 44,  0, 45,  0,  0,  0,  0,  0,  0,
+   46, 47,  0,  0,  0,  0,  0, 48,  0, 49,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 50, 51,  0,  0,  0, 52,  0,  0,
+   53,  0,  0,  0,  0,  0,  0,  0, 54,  0,  0,  0,  0,  0,  0,  0,
+   55,  0,  0,  0,  0,  0,  0,  0, 56,  0,  0,  0,  0,  0,  0,  0,
+    0, 57,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0, 58, 59, 60, 61, 62, 63, 64, 65,
+    0,  0,  0,  0,  0,  0, 66,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 67, 68,  0, 69, 70,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82,
+   83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98,
+   99,100,101,102,103,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,104,  0,  0,  0,  0,  0,  0,105,106,  0,
+  107,  0,  0,  0,108,  0,109,  0,110,  0,111,112,113,  0,114,  0,
+    0,  0,115,  0,  0,  0,116,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,117,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,118,119,120,121,  0,122,123,124,
+  125,126,  0,127,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,128,129,130,131,132,133,134,135,136,137,138,139,
+  140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,
+  156,157,  0,  0,  0,158,159,160,161,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,162,163,  0,
+    0,  0,  0,  0,  0,  0,164,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,165,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,166,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,167,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,168,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,169,170,  0,  0,  0,  0,171,
+  172,  0,  0,  0,173,174,175,176,177,178,179,180,181,182,183,184,
+  185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,
+  201,202,203,204,205,206,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    1,  2,  3,  4,
+};
+static const uint16_t
+_hb_ucd_u16[9080] =
+{
+     0,   0,   1,   2,   3,   4,   5,   6,   0,   0,   7,   8,   9,  10,  11,  12,
+    13,  13,  13,  14,  15,  13,  13,  16,  17,  18,  19,  20,  21,  22,  13,  23,
+    13,  13,  13,  24,  25,  11,  11,  11,  11,  26,  11,  27,  28,  29,  30,  31,
+    32,  32,  32,  32,  32,  32,  32,  33,  34,  35,  36,  11,  37,  38,  13,  39,
+     9,   9,   9,  11,  11,  11,  13,  13,  40,  13,  13,  13,  41,  13,  13,  13,
+    13,  13,  13,  42,   9,  43,  11,  11,  44,  45,  32,  46,  47,  48,  49,  50,
+    51,  52,  48,  48,  53,  32,  54,  55,  48,  48,  48,  48,  48,  56,  57,  58,
+    59,  60,  48,  32,  61,  48,  48,  48,  48,  48,  62,  63,  64,  48,  65,  66,
+    48,  67,  68,  69,  48,  70,  71,  72,  72,  72,  48,  73,  74,  75,  76,  32,
+    77,  48,  48,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,
+    91,  84,  85,  92,  93,  94,  95,  96,  97,  98,  85,  99, 100, 101,  89, 102,
+   103,  84,  85, 104, 105, 106,  89, 107, 108, 109, 110, 111, 112, 113,  95, 114,
+   115, 116,  85, 117, 118, 119,  89, 120, 121, 116,  85, 122, 123, 124,  89, 125,
+   126, 116,  48, 127, 128, 129,  89, 130, 131, 132,  48, 133, 134, 135,  95, 136,
+   137,  48,  48, 138, 139, 140,  72,  72, 141,  48, 142, 143, 144, 145,  72,  72,
+   146, 147, 148, 149, 150,  48, 151, 152, 153, 154,  32, 155, 156, 157,  72,  72,
+    48,  48, 158, 159, 160, 161, 162, 163, 164, 165,   9,   9, 166,  11,  11, 167,
+    48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 168, 169,  48,  48,
+   168,  48,  48, 170, 171, 172,  48,  48,  48, 171,  48,  48,  48, 173, 174, 175,
+    48, 176,   9,   9,   9,   9,   9, 177, 178,  48,  48,  48,  48,  48,  48,  48,
+    48,  48,  48,  48,  48,  48, 179,  48, 180, 181,  48,  48,  48,  48, 182, 183,
+   184, 185,  48, 186,  48, 187, 184, 188,  48,  48,  48, 189, 190, 191, 192, 193,
+   194, 192,  48,  48, 195,  48,  48, 196, 197,  48, 198,  48,  48,  48,  48, 199,
+    48, 200, 201, 202, 203,  48, 204, 205,  48,  48, 206,  48, 207, 208, 209, 209,
+    48, 210,  48,  48,  48, 211, 212, 213, 192, 192, 214, 215, 216,  72,  72,  72,
+   217,  48,  48, 218, 219, 160, 220, 221, 222,  48, 223,  64,  48,  48, 224, 225,
+    48,  48, 226, 227, 228,  64,  48, 229, 230,   9,   9, 231, 232, 233, 234, 235,
+    11,  11, 236,  27,  27,  27, 237, 238,  11, 239,  27,  27,  32,  32,  32, 240,
+    13,  13,  13,  13,  13,  13,  13,  13,  13, 241,  13,  13,  13,  13,  13,  13,
+   242, 243, 242, 242, 243, 244, 242, 245, 246, 246, 246, 247, 248, 249, 250, 251,
+   252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 262,  72, 263, 264, 216,
+   265, 266, 267, 268, 269, 270, 271, 271, 272, 273, 274, 209, 275, 276, 209, 277,
+   278, 278, 278, 278, 278, 278, 278, 278, 279, 209, 280, 209, 209, 209, 209, 281,
+   209, 282, 278, 283, 209, 284, 285, 209, 209, 209, 286,  72, 287,  72, 270, 270,
+   270, 288, 209, 209, 209, 209, 289, 270, 209, 209, 209, 209, 209, 209, 209, 209,
+   209, 209, 209, 290, 291, 209, 209, 292, 209, 209, 209, 209, 209, 209, 293, 209,
+   209, 209, 209, 209, 209, 209, 294, 295, 270, 296, 209, 209, 297, 278, 298, 278,
+   299, 300, 278, 278, 278, 301, 278, 302, 209, 209, 209, 278, 303, 209, 209, 304,
+   209, 305, 209, 209, 209, 209, 209, 209,   9,   9, 306,  11,  11, 307, 308, 309,
+    13,  13,  13,  13,  13,  13, 310, 311,  11,  11, 312,  48,  48,  48, 313, 314,
+    48, 315, 316, 316, 316, 316,  32,  32, 317, 318, 319, 320, 321, 322,  72,  72,
+   209, 323, 209, 209, 209, 209, 209, 324, 209, 209, 209, 209, 209, 325,  72, 326,
+   327, 328, 329, 330, 137,  48,  48,  48,  48, 331, 178,  48,  48,  48,  48, 332,
+   333,  48,  48, 137,  48,  48,  48,  48, 200, 334,  48,  48, 209, 209, 324,  48,
+   209, 335, 336, 209, 337, 338, 209, 209, 336, 209, 209, 338, 209, 209, 209, 209,
+    48,  48,  48,  48, 209, 209, 209, 209,  48,  48,  48,  48,  48,  48,  48, 151,
+    48, 339,  48,  48,  48,  48,  48,  48, 151, 209, 209, 209, 286,  48,  48, 229,
+   340,  48, 341,  72,  13,  13, 342, 343,  13, 344,  48,  48,  48,  48, 345, 346,
+    31, 347, 348, 349,  13,  13,  13, 350, 351, 352, 353, 354, 355,  72,  72, 356,
+   357,  48, 358, 359,  48,  48,  48, 360, 361,  48,  48, 362, 363, 192,  32, 364,
+    64,  48, 365,  48, 366, 367,  48, 151,  77,  48,  48, 368, 369, 370, 371, 372,
+    48,  48, 373, 374, 375, 376,  48, 377,  48,  48,  48, 378, 379, 380, 381, 382,
+   383, 384, 316,  11,  11, 385, 386,  11,  11,  11,  11,  11,  48,  48, 387, 192,
+    48,  48, 388,  48, 389,  48,  48, 206, 390, 390, 390, 390, 390, 390, 390, 390,
+   391, 391, 391, 391, 391, 391, 391, 391,  48,  48,  48,  48,  48,  48, 204,  48,
+    48,  48,  48,  48,  48, 207,  72,  72, 392, 393, 394, 395, 396,  48,  48,  48,
+    48,  48,  48, 397, 398, 399,  48,  48,  48,  48,  48, 400,  72,  48,  48,  48,
+    48, 401,  48,  48,  74,  72,  72, 402,  32, 403,  32, 404, 405, 406, 407,  73,
+    48,  48,  48,  48,  48,  48,  48, 408, 409,   2,   3,   4,   5, 410, 411, 412,
+    48, 413,  48, 200, 414, 415, 416, 417, 418,  48, 172, 419, 204, 204,  72,  72,
+    48,  48,  48,  48,  48,  48,  48,  71, 420, 270, 270, 421, 271, 271, 271, 422,
+   423, 424, 425,  72,  72, 209, 209, 426,  72,  72,  72,  72,  72,  72,  72,  72,
+    48, 151,  48,  48,  48, 101, 427, 428,  48,  48, 429,  48, 430,  48,  48, 431,
+    48, 432,  48,  48, 433, 434,  72,  72,   9,   9, 435,  11,  11,  48,  48,  48,
+    48, 204, 192,   9,   9, 436,  11, 437,  48,  48,  74,  48,  48,  48, 438,  72,
+    48,  48,  48, 315,  48, 199,  74,  72, 439,  48,  48, 440,  48, 441,  48, 442,
+    48, 200, 443,  72,  72,  72,  48, 444,  48, 445,  48, 446,  72,  72,  72,  72,
+    48,  48,  48, 447, 270, 448, 270, 270, 449, 450,  48, 451, 452, 453,  48, 454,
+    48, 455,  72,  72, 456,  48, 457, 458,  48,  48,  48, 459,  48, 460,  48, 461,
+    48, 462, 463,  72,  72,  72,  72,  72,  48,  48,  48,  48, 196,  72,  72,  72,
+     9,   9,   9, 464,  11,  11,  11, 465,  48,  48, 466, 192,  72,  72,  72,  72,
+    72,  72,  72,  72,  72,  72, 270, 467,  48,  48, 468, 469,  72,  72,  72,  72,
+    48, 455, 470,  48,  62, 471,  72,  72,  72,  72,  72,  48, 472,  72,  48, 315,
+   473,  48,  48, 474, 475, 448, 476, 477, 222,  48,  48, 478, 479,  48, 196, 192,
+   480,  48, 481, 482, 483,  48,  48, 484, 222,  48,  48, 485, 486, 487, 488, 489,
+    48,  98, 490, 491,  72,  72,  72,  72, 492, 493, 494,  48,  48, 495, 496, 192,
+   497,  84,  85, 498, 499, 500, 501, 502,  48,  48,  48, 503, 504, 505, 469,  72,
+    48,  48,  48, 506, 507, 192,  72,  72,  48,  48, 508, 509, 510, 511,  72,  72,
+    48,  48,  48, 512, 513, 192, 514,  72,  48,  48, 515, 516, 192,  72,  72,  72,
+    48, 173, 517, 518,  72,  72,  72,  72,  48,  48, 490, 519,  72,  72,  72,  72,
+    72,  72,   9,   9,  11,  11, 148, 520, 521, 522,  48, 523, 524, 192,  72,  72,
+    72,  72, 525,  48,  48, 526, 527,  72, 528,  48,  48, 529, 530, 531,  48,  48,
+   532, 533, 534,  72,  48,  48,  48, 196,  85,  48, 508, 535, 536, 148, 175, 537,
+    48, 538, 539, 540,  72,  72,  72,  72, 541,  48,  48, 542, 543, 192, 544,  48,
+   545, 546, 192,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  72,  48, 547,
+    72,  72,  72, 101, 270, 548, 549, 550,  48, 207,  72,  72,  72,  72,  72,  72,
+   271, 271, 271, 271, 271, 271, 551, 552,  48,  48,  48,  48, 388,  72,  72,  72,
+    48,  48, 200, 553,  72,  72,  72,  72,  48,  48,  48,  48, 315,  72,  72,  72,
+    48,  48,  48, 196,  48, 200, 370,  72,  72,  72,  72,  72,  72,  48, 204, 554,
+    48,  48,  48, 555, 556, 557, 558, 559,  48,  72,  72,  72,  72,  72,  72,  72,
+    72,  72,  72,  72,   9,   9,  11,  11, 270, 560,  72,  72,  72,  72,  72,  72,
+    48,  48,  48,  48, 561, 562, 563, 563, 564, 565,  72,  72,  72,  72, 566, 567,
+    48,  48,  48,  48,  48,  48,  48,  74,  48,  48,  48,  48,  48, 199,  72,  72,
+   196,  72,  72,  72,  72,  72,  72,  72,  48, 200,  72,  72,  72, 568, 569,  48,
+    48,  48,  48,  48,  48,  48,  48, 206,  48,  48,  48,  48,  48,  48,  71, 151,
+   196, 570, 571,  72,  72,  72,  72,  72, 209, 209, 209, 209, 209, 209, 209, 325,
+   209, 209, 572, 209, 209, 209, 573, 574, 575, 209, 576, 209, 209, 209, 577,  72,
+   209, 209, 209, 209, 578,  72,  72,  72,  72,  72,  72,  72,  72,  72, 270, 579,
+   209, 209, 209, 209, 209, 286, 270, 452,   9, 580,  11, 581, 582, 583, 242,   9,
+   584, 585, 586, 587, 588,   9, 580,  11, 589, 590,  11, 591, 592, 593, 594,   9,
+   595,  11,   9, 580,  11, 581, 582,  11, 242,   9, 584, 594,   9, 595,  11,   9,
+   580,  11, 596,   9, 597, 598, 599, 600,  11, 601,   9, 602, 603, 604, 605,  11,
+   606,   9, 607,  11, 608, 609, 609, 609,  32,  32,  32, 610,  32,  32, 611, 612,
+   613, 614,  45,  72,  72,  72,  72,  72, 615, 616, 617,  72,  72,  72,  72,  72,
+    48,  48, 151, 618, 619,  72,  72,  72,  72,  72,  72,  72,  48,  48, 620, 621,
+    48,  48,  48,  48, 622, 623,  72,  72,   9,   9, 584,  11, 624, 370,  72,  72,
+    72,  72,  72,  72,  72,  72,  72, 488, 270, 270, 625, 626,  72,  72,  72,  72,
+   488, 270, 627, 628,  72,  72,  72,  72, 629,  48, 630, 631, 632, 633, 634, 635,
+   636, 206, 637, 206,  72,  72,  72, 638, 209, 209, 326, 209, 209, 209, 209, 209,
+   209, 324, 335, 639, 639, 639, 209, 325, 640, 209, 209, 209, 209, 209, 209, 209,
+   209, 209, 641,  72,  72,  72, 642, 209, 643, 209, 209, 326, 577, 644, 325,  72,
+   209, 209, 209, 209, 209, 209, 209, 645, 209, 209, 209, 209, 209, 646, 424, 424,
+   209, 209, 209, 209, 209, 209, 209, 324, 209, 209, 209, 209, 209, 577, 326,  72,
+   326, 209, 209, 209, 646, 176, 209, 209, 646, 209, 641, 644,  72,  72,  72,  72,
+   209, 209, 209, 209, 209, 209, 209, 647, 209, 209, 209, 209, 648, 209, 209, 209,
+   209, 209, 209, 209, 209, 324, 641, 649, 286, 209, 577, 286, 643, 286,  72,  72,
+   209, 650, 209, 209, 287,  72,  72, 192,  48,  48,  48,  48,  48, 204,  72,  72,
+    48,  48,  48, 205,  48,  48,  48,  48,  48, 204,  48,  48,  48,  48,  48,  48,
+    48,  48, 469,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48,  48, 101,  72,
+    48, 204,  72,  72,  72,  72,  72,  72,  48,  48,  48,  48,  71,  72,  72,  72,
+   651,  72, 652, 652, 652, 652, 652, 652,  32,  32,  32,  32,  32,  32,  32,  32,
+    32,  32,  32,  32,  32,  32,  32,  72, 391, 391, 391, 391, 391, 391, 391, 653,
+   391, 391, 391, 391, 391, 391, 391, 654,   0,   0,   0,   0,   0,   0,   0,   0,
+     1,   2,   2,   3,   1,   2,   2,   3,   0,   0,   0,   0,   0,   4,   0,   4,
+     2,   2,   5,   2,   2,   2,   5,   2,   2,   2,   2,   2,   2,   2,   2,   2,
+     2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   2,   6,
+     0,   0,   0,   0,   7,   8,   0,   0,   9,   9,   9,   9,   9,   9,   9,   9,
+     9,   9,   9,   9,   9,   9,  10,  11,  12,  13,  14,  14,  15,  14,  14,  14,
+    14,  14,  14,  14,  16,  17,  14,  14,  18,  18,  18,  18,  18,  18,  18,  18,
+    18,  18,  18,  18,  18,  18,  18,  18,  19,  18,  18,  18,  18,  18,  18,  18,
+    18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  20,  21,
+    21,  21,  22,  20,  21,  21,  21,  21,  21,  23,  24,  25,  25,  25,  25,  25,
+    25,  26,  25,  25,  25,  27,  28,  26,  29,  30,  31,  32,  31,  31,  31,  31,
+    33,  34,  35,  31,  31,  31,  36,  31,  31,  31,  31,  31,  31,  31,  31,  31,
+    31,  31,  31,  29,  31,  31,  31,  31,  37,  38,  37,  37,  37,  37,  37,  37,
+    37,  39,  31,  31,  31,  31,  31,  31,  40,  40,  40,  40,  40,  40,  41,  26,
+    42,  42,  42,  42,  42,  42,  42,  43,  44,  44,  44,  44,  44,  45,  44,  46,
+    47,  47,  47,  48,  37,  49,  26,  26,  26,  26,  26,  26,  31,  31,  50,  31,
+    31,  26,  51,  31,  52,  31,  31,  31,  53,  53,  53,  53,  53,  53,  53,  53,
+    53,  53,  54,  53,  55,  53,  53,  53,  56,  57,  58,  59,  59,  60,  61,  62,
+    57,  63,  64,  65,  66,  59,  59,  67,  68,  69,  70,  71,  71,  72,  73,  74,
+    69,  75,  76,  77,  78,  71,  79,  26,  80,  81,  82,  83,  83,  84,  85,  86,
+    81,  87,  88,  26,  89,  83,  90,  91,  92,  93,  94,  95,  95,  96,  97,  98,
+    93,  99, 100, 101, 102,  95,  95,  26, 103, 104, 105, 106, 107, 104, 108, 109,
+   104, 105, 110,  26, 111, 108, 108, 112, 113, 114, 115, 113, 113, 115, 113, 116,
+   114, 117, 118, 119, 120, 113, 121, 113, 122, 123, 124, 122, 122, 124, 125, 126,
+   123, 127, 128, 129, 130, 122, 131,  26, 132, 133, 134, 132, 132, 132, 132, 132,
+   133, 134, 135, 132, 136, 132, 132, 132, 137, 138, 139, 140, 138, 138, 141, 142,
+   139, 143, 144, 138, 145, 138, 146,  26, 147, 148, 148, 148, 148, 148, 148, 149,
+   148, 148, 148, 150,  26,  26,  26,  26, 151, 152, 153, 153, 154, 153, 153, 155,
+   156, 155, 153, 157,  26,  26,  26,  26, 158, 158, 158, 158, 158, 158, 158, 158,
+   158, 159, 158, 158, 158, 160, 159, 158, 158, 158, 158, 159, 158, 158, 158, 161,
+   158, 161, 162, 163,  26,  26,  26,  26, 164, 164, 164, 164, 164, 164, 164, 164,
+   164, 164, 164, 164, 164, 164, 164, 164, 164, 164, 164, 164, 165, 165, 165, 165,
+   166, 167, 165, 165, 165, 165, 165, 168, 169, 169, 169, 169, 169, 169, 169, 169,
+   169, 169, 169, 169, 169, 169, 169, 169, 170, 170, 170, 170, 170, 170, 170, 170,
+   170, 171, 172, 171, 170, 170, 170, 170, 170, 171, 170, 170, 170, 170, 171, 172,
+   171, 170, 172, 170, 170, 170, 170, 170, 170, 170, 171, 170, 170, 170, 170, 170,
+   170, 170, 170, 173, 170, 170, 170, 174, 170, 170, 170, 175, 176, 176, 176, 176,
+   176, 176, 176, 176, 176, 176, 177, 177, 178, 178, 178, 178, 178, 178, 178, 178,
+   178, 178, 178, 178, 178, 178, 178, 178, 179, 179, 179, 180, 181, 181, 181, 181,
+   181, 181, 181, 181, 181, 182, 181, 183, 184, 185, 186,  26, 187, 187, 188,  26,
+   189, 189, 190,  26, 191, 192, 193,  26, 194, 194, 194, 194, 194, 194, 194, 194,
+   194, 194, 194, 195, 194, 196, 194, 196, 197, 198, 199, 200, 199, 199, 199, 199,
+   199, 199, 199, 199, 199, 199, 199, 201, 199, 199, 199, 199, 199, 202, 178, 178,
+   178, 178, 178, 178, 178, 178, 203,  26, 204, 204, 204, 205, 204, 206, 204, 206,
+   207, 204, 208, 208, 208, 209, 210,  26, 211, 211, 211, 211, 211, 212, 211, 211,
+   211, 213, 211, 214, 194, 194, 194, 194, 215, 215, 215, 216, 217, 217, 217, 217,
+   217, 217, 217, 218, 217, 217, 217, 219, 217, 220, 217, 220, 217, 221,   9,   9,
+   222,  26,  26,  26,  26,  26,  26,  26, 223, 223, 223, 223, 223, 223, 223, 223,
+   223, 224, 223, 223, 223, 223, 223, 225, 226, 226, 226, 226, 226, 226, 226, 226,
+   227, 227, 227, 227, 227, 227, 228, 229, 230, 230, 230, 230, 230, 230, 230, 231,
+   230, 232, 233, 233, 233, 233, 233, 233,  18, 234, 165, 165, 165, 165, 165, 235,
+   226,  26, 236,   9, 237, 238, 239, 240,   2,   2,   2,   2, 241, 242,   2,   2,
+     2,   2,   2, 243, 244, 245,   2, 246,   2,   2,   2,   2,   2,   2,   2, 247,
+     9,   9,   9,   9,   9,   9,   9, 248,  14,  14, 249, 249,  14,  14,  14,  14,
+   249, 249,  14, 250,  14,  14,  14, 249,  14,  14,  14,  14,  14,  14, 251,  14,
+   251,  14, 252, 253,  14,  14, 254, 255,   0, 256,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0, 257,   0, 258, 259,   0, 260,   2, 261,   0,   0,   0,   0,
+    26,  26,   9,   9,   9,   9, 222,  26,   0,   0,   0,   0, 262, 263,   4,   0,
+     0, 264,   0,   0,   2,   2,   2,   2,   2, 265,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 260,  26,  26,  26,
+     0, 266,  26,  26,   0,   0,   0,   0, 267, 267, 267, 267, 267, 267, 267, 267,
+   267, 267, 267, 267, 267, 267, 267, 267,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0, 268,   0,   0,   0, 269,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0, 270, 270, 270, 270, 270, 271, 270, 270,
+   270, 270, 270, 271,   2,   2,   2,   2,  17,  17,  17,  17,  17,  17,  17,  17,
+    17,  17,  17,  17,  17,  17, 272, 273, 165, 165, 165, 165, 166, 167, 274, 274,
+   274, 274, 274, 274, 274, 275, 276, 275, 170, 170, 172,  26, 172, 172, 172, 172,
+   172, 172, 172, 172,  18,  18,  18,  18,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 266,  26,  26,  26,  26,  26, 277, 277, 277, 278, 277, 277, 277, 277,
+   277, 277, 277, 277, 277, 277, 279,  26, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 280,  26,  26,  26,   0, 281, 282,   0,   0,   0, 283, 284,   0, 285,
+   286, 287, 287, 287, 287, 287, 287, 287, 287, 287, 288, 289, 290, 291, 291, 291,
+   291, 291, 291, 291, 291, 291, 291, 292, 293, 294, 294, 294, 294, 294, 295, 169,
+   169, 169, 169, 169, 169, 169, 169, 169, 169, 296,   0,   0, 294, 294, 294, 294,
+     0,   0,   0,   0, 281,  26, 291, 291, 169, 169, 169, 296,   0,   0,   0,   0,
+     0,   0,   0,   0, 169, 169, 169, 297,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 291, 291, 291, 291, 291, 298, 291, 291, 291, 291, 291, 291, 291, 291,
+   291, 291, 291,   0,   0,   0,   0,   0, 277, 277, 277, 277, 277, 277, 277, 277,
+     0,   0,   0,   0,   0,   0,   0,   0, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 277, 277, 277, 277, 299, 300, 300, 300, 300, 300, 300, 300, 300,
+   300, 300, 300, 300, 300, 300, 300, 300, 300, 301, 300, 300, 300, 300, 300, 300,
+   302,  26, 303, 303, 303, 303, 303, 303, 304, 304, 304, 304, 304, 304, 304, 304,
+   304, 304, 304, 304, 304, 304, 304, 304, 304, 304, 304, 304, 304, 305,  26,  26,
+    18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18,  18, 306, 306, 306, 306,
+   306, 306, 306, 306, 306, 306, 306,  26,   0,   0,   0,   0, 307,   2,   2,   2,
+     2,   2,   2,   2,   2,   2,   2,   2,   2, 308,   2,   2,   2,   2,   2,   2,
+   309, 310,  26,  26,  26,  26, 311,   2, 312, 312, 312, 312, 312, 313,   0, 314,
+   315, 315, 315, 315, 315, 315, 315,  26, 316, 316, 316, 316, 316, 316, 316, 316,
+   317, 318, 316, 319,  53,  53,  53,  53, 320, 320, 320, 320, 320, 321, 322, 322,
+   322, 322, 323, 324, 169, 169, 169, 325, 326, 326, 326, 326, 326, 326, 326, 326,
+   326, 327, 326, 328, 164, 164, 164, 329, 330, 330, 330, 330, 330, 330, 331,  26,
+   330, 332, 330, 333, 164, 164, 164, 164, 334, 334, 334, 334, 334, 334, 334, 334,
+   335,  26,  26, 336, 337, 337, 338,  26, 339, 339, 339,  26, 172, 172,   2,   2,
+     2,   2,   2, 340, 341, 342, 176, 176, 176, 176, 176, 176, 176, 176, 176, 176,
+   337, 337, 337, 337, 337, 343, 337, 344, 169, 169, 169, 169, 345,  26, 169, 169,
+   296, 346, 169, 169, 169, 169, 169, 345,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 277, 277, 280, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 347,  26,  26,  26,  26, 348,  26, 349, 350,  25,  25, 351, 352,
+   353,  25,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,
+   354,  26,  51,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,
+    31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31,  31, 355,
+    26,  26,  31,  31,  31,  31,  31,  31,  31,  31, 356,  31,  31,  31,  31,  31,
+    31,  26,  26,  26,  26,  26,  31, 357,   9,   9,   0, 314,   9, 358,   0,   0,
+     0,   0, 359,   0, 260, 281,  50,  31,  31,  31,  31,  31,  31,  31,  31,  31,
+    31,  31,  31,  31,  31,  31,  31, 360, 361,   0,   0,   0,   1,   2,   2,   3,
+     1,   2,   2,   3, 362, 291, 290, 291, 291, 291, 291, 363, 169, 169, 169, 296,
+   364, 364, 364, 365, 260, 260,  26, 366, 367, 368, 367, 367, 369, 367, 367, 370,
+   367, 371, 367, 371,  26,  26,  26,  26, 367, 367, 367, 367, 367, 367, 367, 367,
+   367, 367, 367, 367, 367, 367, 367, 372, 373,   0,   0,   0,   0,   0, 374,   0,
+    14,  14,  14,  14,  14,  14,  14,  14,  14, 255,   0, 375, 376,  26,  26,  26,
+    26,  26,   0,   0,   0,   0,   0, 377, 378, 378, 378, 379, 380, 380, 380, 380,
+   380, 380, 381,  26, 382,   0,   0, 281, 383, 383, 383, 383, 384, 385, 386, 386,
+   386, 387, 388, 388, 388, 388, 388, 389, 390, 390, 390, 391, 392, 392, 392, 392,
+   393, 392, 394,  26,  26,  26,  26,  26, 395, 395, 395, 395, 395, 395, 395, 395,
+   395, 395, 396, 396, 396, 396, 396, 396, 397, 397, 397, 398, 397, 399, 400, 400,
+   400, 400, 401, 400, 400, 400, 400, 401, 402, 402, 402, 402, 402,  26, 403, 403,
+   403, 403, 403, 403, 404, 405,  26,  26, 406, 406, 406, 406, 406, 406, 406, 406,
+   406, 406, 406, 406, 406, 406, 406, 406, 406, 406, 406, 406, 406, 406, 407,  26,
+   406, 406, 408,  26, 406,  26,  26,  26, 409, 410, 411, 411, 411, 411, 412, 413,
+   414, 414, 415, 414, 416, 416, 416, 416, 417, 417, 417, 418, 419, 417,  26,  26,
+    26,  26,  26,  26, 420, 420, 421, 422, 423, 423, 423, 424, 425, 425, 425, 426,
+    26,  26,  26,  26,  26,  26,  26,  26, 427, 427, 427, 427, 428, 428, 428, 429,
+   428, 428, 430, 428, 428, 428, 428, 428, 431, 432, 433, 434, 435, 435, 436, 437,
+   435, 438, 435, 438, 439, 439, 439, 439, 440, 440, 440, 440,  26,  26,  26,  26,
+   441, 441, 441, 441, 442, 443, 442,  26, 444, 444, 444, 444, 444, 444, 445, 446,
+   447, 447, 448, 447, 449, 449, 450, 449, 451, 451, 452, 453,  26, 454,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 455, 455, 455, 455, 455, 455, 455, 455,
+   455, 456,  26,  26,  26,  26,  26,  26, 457, 457, 457, 457, 457, 457, 458,  26,
+   457, 457, 457, 457, 457, 457, 458, 459, 460, 460, 460, 460, 460,  26, 460, 461,
+    26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,  31,  31,  31, 462, 463, 463, 463, 463, 463, 464, 465,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 466, 466, 466, 466, 466,  26, 467, 467,
+   467, 467, 467, 468,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26, 469, 469,
+   469, 470,  26,  26, 471, 471, 472,  26, 473, 473, 473, 473, 473, 473, 473, 473,
+   473, 474, 475, 473, 473, 473,  26, 476, 477, 477, 477, 477, 477, 477, 477, 477,
+   478, 479, 480, 480, 480, 481, 480, 482, 483, 483, 483, 483, 483, 483, 484, 483,
+   483,  26, 485, 485, 485, 485, 486,  26, 487, 487, 487, 487, 487, 487, 487, 487,
+   487, 487, 487, 487, 488, 138, 489,  26, 490, 490, 491, 490, 490, 490, 490, 492,
+    26,  26,  26,  26,  26,  26,  26,  26, 493, 494, 495, 496, 495, 497, 498, 498,
+   498, 498, 498, 498, 498, 499, 498, 500, 501, 502, 503, 504, 504, 505, 506, 507,
+   502, 508, 509, 510, 511, 512, 512,  26, 513, 513, 513, 513, 513, 513, 513, 513,
+   513, 513, 513, 514, 515,  26,  26,  26, 516, 516, 516, 516, 516, 516, 516, 516,
+   516,  26, 516, 517,  26,  26,  26,  26, 518, 518, 518, 518, 518, 518, 519, 518,
+   518, 518, 518, 519,  26,  26,  26,  26, 520, 520, 520, 520, 520, 520, 520, 520,
+   521,  26, 520, 522, 199, 523,  26,  26, 524, 524, 524, 524, 524, 524, 524, 525,
+   524, 526,  26,  26,  26,  26,  26,  26, 527, 527, 527, 528, 527, 529, 527, 527,
+    26,  26,  26,  26,  26,  26,  26,  26, 530, 530, 530, 530, 530, 530, 530, 531,
+    26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26, 532, 532, 532, 532,
+   532, 532, 532, 532, 532, 532, 533, 534, 535, 536, 537, 538, 538, 538, 539, 540,
+   535,  26, 538, 541,  26,  26,  26,  26,  26,  26,  26,  26, 542, 543, 542, 542,
+   542, 542, 542, 543, 544,  26,  26,  26, 545, 545, 545, 545, 545, 545, 545, 545,
+   545,  26, 546, 546, 546, 546, 546, 546, 546, 546, 546, 546, 547,  26,  26,  26,
+   548, 548, 548, 548, 548, 548, 548, 549, 550, 551, 550, 550, 550, 550, 552, 550,
+   553,  26, 550, 550, 550, 554, 555, 555, 555, 555, 556, 555, 555, 557, 558,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 559, 560, 561, 561, 561, 561, 559, 562,
+   561,  26, 561, 563, 564, 565, 566, 566, 566, 567, 568, 569, 566, 570,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26, 571, 571, 571, 572,  26,  26,  26,  26,  26,  26, 573,  26,
+   108, 108, 108, 108, 108, 108, 574, 575, 576, 576, 576, 576, 576, 576, 576, 576,
+   576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 576, 577,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 576, 576, 576, 576, 576, 576, 576, 576,
+   576, 576, 576, 576, 576, 578, 579,  26, 576, 576, 576, 576, 576, 576, 576, 576,
+   580,  26,  26,  26,  26,  26,  26,  26, 581, 581, 581, 581, 581, 581, 581, 581,
+   581, 581, 581, 581, 581, 581, 581, 581, 581, 581, 581, 581, 581, 582, 581, 583,
+    26,  26,  26,  26,  26,  26,  26,  26, 584, 584, 584, 584, 584, 584, 584, 584,
+   584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584, 584,
+   585,  26,  26,  26,  26,  26,  26,  26, 306, 306, 306, 306, 306, 306, 306, 306,
+   306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 306, 586,
+   587, 587, 587, 588, 587, 589,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26, 590, 590, 590, 591, 591,  26, 592, 592, 592, 592, 592, 592, 592, 592,
+   593,  26, 592, 594, 594, 592, 592, 595, 592, 592,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+   596, 596, 596, 596, 596, 596, 596, 596, 596, 596, 596, 597,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 598, 598, 598, 598, 598, 598, 598, 598,
+   598, 599, 598, 598, 598, 598, 598, 598, 598, 600, 598, 598,  26,  26,  26,  26,
+    26,  26,  26,  26, 601,  26, 347,  26, 602, 602, 602, 602, 602, 602, 602, 602,
+   602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602, 602,
+   602, 602, 602, 602, 602, 602, 602,  26, 603, 603, 603, 603, 603, 603, 603, 603,
+   603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603, 603,
+   603, 603, 604,  26,  26,  26,  26,  26, 602, 605,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 606, 287, 287, 287, 287, 287, 287, 287,
+   287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287,
+   287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 287, 288,  26,  26,  26,  26,
+    26,  26, 607,  26, 608,  26, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609,
+   609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609, 609,
+   609, 609, 609, 609, 609, 609, 609, 610, 611, 611, 611, 611, 611, 611, 611, 611,
+   611, 611, 611, 611, 611, 612, 611, 613, 611, 614, 611, 615, 281,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0, 616,  26,   0,   0,   0,   0, 260, 361,   0,   0,
+     0,   0,   0,   0, 617, 618,   0, 619, 620, 621,   0,   0,   0, 622,   0,   0,
+     0,   0,   0,   0,   0, 623,  26,  26,  14,  14,  14,  14,  14,  14,  14,  14,
+   249,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,   0,   0, 281,  26,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 260,  26,   0,   0,   0, 623,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 257,   0,   0,   0,   0,   0,   0,   0,   0, 257, 624, 625,   0, 626,
+   627,   0,   0,   0,   0,   0,   0,   0, 269, 628, 257, 257,   0,   0,   0, 629,
+   630, 631, 632,   0,   0,   0,   0,   0,   0,   0,   0,   0, 616,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0, 268,   0,   0,   0,   0,   0,   0, 633, 633, 633, 633, 633, 633, 633, 633,
+   633, 633, 633, 633, 633, 633, 633, 633, 633, 634,  26, 635, 636, 633,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 271, 270, 270, 637, 638, 639,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 640, 640, 640, 640, 640, 641, 640, 642,
+   640, 643,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+   644, 644, 644, 644, 644, 644, 644, 645, 646, 646, 646, 646, 646, 646, 646, 646,
+   646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646, 646,
+   647, 646, 648,  26,  26,  26,  26,  26, 649, 649, 649, 649, 649, 649, 649, 649,
+   649, 650, 649, 651,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26, 361,   0,   0,   0,   0,   0,   0,   0, 375,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 361,   0,   0,   0,   0,   0,   0, 616,
+    26,  26,  26,  26,  26,  26,  26,  26, 652,  31,  31,  31, 653, 654, 655, 656,
+   657, 658, 653, 659, 653, 655, 655, 660,  31, 661,  31, 662, 663, 661,  31, 662,
+    26,  26,  26,  26,  26,  26, 354,  26,   0,   0,   0,   0,   0, 281,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 281,  26,   0, 260, 361,   0,
+   361,   0, 361,   0,   0,   0, 616,  26,   0,   0,   0,   0,   0, 616,  26,  26,
+    26,  26,  26,  26, 664,   0,   0,   0, 665,  26,   0,   0,   0,   0,   0, 281,
+     0, 623, 314,  26, 616,  26,  26,  26,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,  26,   0, 375,   0, 375,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0, 281,  26,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0, 623,   0, 281,  26,  26,   0, 281,   0,   0,   0,   0,   0,   0,
+     0,  26,   0, 314,   0,   0,   0,   0,   0,  26,   0,   0,   0, 616, 314,  26,
+    26,  26,  26,  26,  26,  26,  26,  26,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0, 632,   0,   0,   0,   0,   0,   0,   0,   0,
+     0, 627,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0, 281,  26,   0, 616, 375, 266, 260,  26,   0,   0,   0, 623, 260,  26,
+   266,  26, 260,  26,  26,  26,  26,  26,   0,   0, 359,   0,   0,   0,   0,   0,
+     0, 266,  26,  26,  26,  26,   0, 314, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 280,  26,  26,  26,  26, 277, 277, 277, 277, 277, 277, 299,  26,
+   277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 280, 277, 277, 277, 277,
+   277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 347,  26, 277, 277,
+   277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 277, 277, 277, 666,  26,  26,  26, 277, 277, 277, 280,  26,  26,  26,  26,
+    26,  26,  26,  26,  26,  26,  26,  26, 277, 277, 277, 277, 277, 277, 277, 277,
+   277, 667,  26,  26,  26,  26,  26,  26, 668,  26,  26,  26,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   9,   9,   9,   9,   9,   9,   9,   9,
+     9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,   9,
+     9,   9,   9,   9,   9,   9,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0, 939, 940, 941, 942, 946, 948,   0, 962,
+   969, 970, 971, 976,1001,1002,1003,1008,   0,1033,1040,1041,1042,1043,1047,   0,
+     0,1080,1081,1082,1086,1110,   0,   0,1124,1125,1126,1127,1131,1133,   0,1147,
+  1154,1155,1156,1161,1187,1188,1189,1193,   0,1219,1226,1227,1228,1229,1233,   0,
+     0,1267,1268,1269,1273,1298,   0,1303, 943,1128, 944,1129, 954,1139, 958,1143,
+   959,1144, 960,1145, 961,1146, 964,1149,   0,   0, 973,1158, 974,1159, 975,1160,
+   983,1168, 978,1163, 988,1173, 990,1175, 991,1176, 993,1178, 994,1179,   0,   0,
+  1004,1190,1005,1191,1006,1192,1014,1199,1007,   0,   0,   0,1016,1201,1020,1206,
+     0,1022,1208,1025,1211,1023,1209,   0,   0,   0,   0,1032,1218,1037,1223,1035,
+  1221,   0,   0,   0,1044,1230,1045,1231,1049,1235,   0,   0,1058,1244,1064,1250,
+  1060,1246,1066,1252,1067,1253,1072,1258,1069,1255,1077,1264,1074,1261,   0,   0,
+  1083,1270,1084,1271,1085,1272,1088,1275,1089,1276,1096,1283,1103,1290,1111,1299,
+  1115,1118,1307,1120,1309,1121,1310,   0,1053,1239,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,1093,1280,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0, 949,1134,1010,1195,1050,1236,1090,1277,1341,1368,1340,
+  1367,1342,1369,1339,1366,   0,1320,1347,1418,1419,1323,1350,   0,   0, 992,1177,
+  1018,1204,1055,1241,1416,1417,1415,1424,1202,   0,   0,   0, 987,1172,   0,   0,
+  1031,1217,1321,1348,1322,1349,1338,1365, 950,1135, 951,1136, 979,1164, 980,1165,
+  1011,1196,1012,1197,1051,1237,1052,1238,1061,1247,1062,1248,1091,1278,1092,1279,
+  1071,1257,1076,1263,   0,   0, 997,1182,   0,   0,   0,   0,   0,   0, 945,1130,
+   982,1167,1337,1364,1335,1362,1046,1232,1422,1423,1113,1301,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   8,   9,   0,  10,1425,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   7,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   1,   0,   0,   0,   0,   0,   0,1314,1427,   5,
+  1434,1438,1443,   0,1450,   0,1455,1461,1514,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1446,1458,1468,1476,1480,1486,1517,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1489,1503,1494,1500,1508,   0,   0,   0,   0,1520,1521,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1526,1528,   0,1525,   0,   0,   0,1522,
+     0,   0,   0,   0,1536,1532,1539,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1534,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1556,   0,   0,   0,   0,   0,   0,1548,1550,   0,1547,   0,   0,   0,1567,
+     0,   0,   0,   0,1558,1554,1561,   0,   0,   0,   0,   0,   0,   0,1568,1569,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1529,1551,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1523,1545,1524,1546,   0,   0,1527,1549,
+     0,   0,1570,1571,1530,1552,1531,1553,   0,   0,1533,1555,1535,1557,1537,1559,
+     0,   0,1572,1573,1544,1566,1538,1560,1540,1562,1541,1563,1542,1564,   0,   0,
+  1543,1565,   0,   0,   0,   0,   0,   0,   0,   0,1606,1607,1609,1608,1610,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1613,   0,1611,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1612,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1620,   0,   0,   0,   0,   0,   0,   0,1623,   0,   0,1624,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1614,1615,1616,1617,1618,1619,1621,1622,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1628,1629,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1625,1626,   0,1627,   0,   0,   0,1634,   0,   0,1635,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1630,1631,1632,   0,   0,1633,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1639,   0,   0,1638,1640,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1636,1637,   0,   0,   0,   0,   0,   0,1641,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1642,1644,1643,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1645,   0,   0,   0,   0,   0,   0,   0,1646,   0,   0,   0,   0,   0,   0,1648,
+  1649,   0,1647,1650,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1651,1653,1652,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1654,   0,1655,1657,1656,   0,   0,   0,   0,1659,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,1660,   0,   0,   0,   0,1661,   0,   0,   0,   0,1662,
+     0,   0,   0,   0,1663,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1658,   0,   0,   0,   0,   0,   0,   0,   0,   0,1664,   0,1665,1673,   0,
+  1674,   0,   0,   0,   0,   0,   0,   0,   0,1666,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1668,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,1669,   0,   0,   0,   0,1670,   0,   0,   0,   0,1671,
+     0,   0,   0,   0,1672,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1667,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1675,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1676,   0,
+  1677,   0,1678,   0,1679,   0,1680,   0,   0,   0,1681,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1682,   0,1683,   0,   0,1684,1685,   0,1686,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0, 953,1138, 955,1140, 956,1141, 957,1142,
+  1324,1351, 963,1148, 965,1150, 968,1153, 966,1151, 967,1152,1378,1380,1379,1381,
+   984,1169, 985,1170,1420,1421, 986,1171, 989,1174, 995,1180, 998,1183, 996,1181,
+   999,1184,1000,1185,1015,1200,1329,1356,1017,1203,1019,1205,1021,1207,1024,1210,
+  1687,1688,1027,1213,1026,1212,1028,1214,1029,1215,1030,1216,1034,1220,1036,1222,
+  1039,1225,1038,1224,1334,1361,1336,1363,1382,1384,1383,1385,1056,1242,1057,1243,
+  1059,1245,1063,1249,1689,1690,1065,1251,1068,1254,1070,1256,1386,1387,1388,1389,
+  1691,1692,1073,1259,1075,1262,1079,1266,1078,1265,1095,1282,1098,1285,1097,1284,
+  1390,1391,1392,1393,1099,1286,1100,1287,1101,1288,1102,1289,1105,1292,1104,1291,
+  1106,1294,1107,1295,1108,1296,1114,1302,1119,1308,1122,1311,1123,1312,1186,1260,
+  1293,1305,   0,1394,   0,   0,   0,   0, 952,1137, 947,1132,1317,1344,1316,1343,
+  1319,1346,1318,1345,1693,1695,1371,1375,1370,1374,1373,1377,1372,1376,1694,1696,
+   981,1166, 977,1162, 972,1157,1326,1353,1325,1352,1328,1355,1327,1354,1697,1698,
+  1009,1194,1013,1198,1054,1240,1048,1234,1331,1358,1330,1357,1333,1360,1332,1359,
+  1699,1700,1396,1401,1395,1400,1398,1403,1397,1402,1399,1404,1094,1281,1087,1274,
+  1406,1411,1405,1410,1408,1413,1407,1412,1409,1414,1109,1297,1117,1306,1116,1304,
+  1112,1300,   0,   0,   0,   0,   0,   0,1471,1472,1701,1705,1702,1706,1703,1707,
+  1430,1431,1715,1719,1716,1720,1717,1721,1477,1478,1729,1731,1730,1732,   0,   0,
+  1435,1436,1733,1735,1734,1736,   0,   0,1481,1482,1737,1741,1738,1742,1739,1743,
+  1439,1440,1751,1755,1752,1756,1753,1757,1490,1491,1765,1768,1766,1769,1767,1770,
+  1447,1448,1771,1774,1772,1775,1773,1776,1495,1496,1777,1779,1778,1780,   0,   0,
+  1451,1452,1781,1783,1782,1784,   0,   0,1504,1505,1785,1788,1786,1789,1787,1790,
+     0,1459,   0,1791,   0,1792,   0,1793,1509,1510,1794,1798,1795,1799,1796,1800,
+  1462,1463,1808,1812,1809,1813,1810,1814,1467,  21,1475,  22,1479,  23,1485,  24,
+  1493,  27,1499,  28,1507,  29,   0,   0,1704,1708,1709,1710,1711,1712,1713,1714,
+  1718,1722,1723,1724,1725,1726,1727,1728,1740,1744,1745,1746,1747,1748,1749,1750,
+  1754,1758,1759,1760,1761,1762,1763,1764,1797,1801,1802,1803,1804,1805,1806,1807,
+  1811,1815,1816,1817,1818,1819,1820,1821,1470,1469,1822,1474,1465,   0,1473,1825,
+  1429,1428,1426,  12,1432,   0,  26,   0,   0,1315,1823,1484,1466,   0,1483,1829,
+  1433,  13,1437,  14,1441,1826,1827,1828,1488,1487,1513,  19,   0,   0,1492,1515,
+  1445,1444,1442,  15,   0,1831,1832,1833,1502,1501,1516,  25,1497,1498,1506,1518,
+  1457,1456,1454,  17,1453,1313,  11,   3,   0,   0,1824,1512,1519,   0,1511,1830,
+  1449,  16,1460,  18,1464,   4,   0,   0,  30,  31,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,  20,   0,
+     0,   0,   2,   6,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1834,1835,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1836,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,1837,1839,1838,   0,   0,   0,   0,1840,   0,   0,   0,
+     0,1841,   0,   0,1842,   0,   0,   0,   0,   0,   0,   0,1843,   0,1844,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1845,   0,   0,1846,   0,   0,1847,
+     0,1848,   0,   0,   0,   0,   0,   0, 937,   0,1850,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,1849, 936, 938,1851,1852,   0,   0,1853,1854,   0,   0,
+  1855,1856,   0,   0,   0,   0,   0,   0,1857,1858,   0,   0,1861,1862,   0,   0,
+  1863,1864,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1867,1868,1869,1870,1859,1860,1865,1866,   0,   0,   0,   0,
+     0,   0,1871,1872,1873,1874,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,  32,  33,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1875,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1877,   0,1878,   0,1879,   0,1880,   0,1881,   0,1882,   0,
+  1883,   0,1884,   0,1885,   0,1886,   0,1887,   0,1888,   0,   0,1889,   0,1890,
+     0,1891,   0,   0,   0,   0,   0,   0,1892,1893,   0,1894,1895,   0,1896,1897,
+     0,1898,1899,   0,1900,1901,   0,   0,   0,   0,   0,   0,1876,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1902,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,1904,   0,1905,   0,1906,   0,1907,   0,1908,   0,1909,   0,
+  1910,   0,1911,   0,1912,   0,1913,   0,1914,   0,1915,   0,   0,1916,   0,1917,
+     0,1918,   0,   0,   0,   0,   0,   0,1919,1920,   0,1921,1922,   0,1923,1924,
+     0,1925,1926,   0,1927,1928,   0,   0,   0,   0,   0,   0,1903,   0,   0,1929,
+  1930,1931,1932,   0,   0,   0,1933,   0, 710, 385, 724, 715, 455, 103, 186, 825,
+   825, 242, 751, 205, 241, 336, 524, 601, 663, 676, 688, 738, 411, 434, 474, 500,
+   649, 746, 799, 108, 180, 416, 482, 662, 810, 275, 462, 658, 692, 344, 618, 679,
+   293, 388, 440, 492, 740, 116, 146, 168, 368, 414, 481, 527, 606, 660, 665, 722,
+   781, 803, 809, 538, 553, 588, 642, 758, 811, 701, 233, 299, 573, 612, 487, 540,
+   714, 779, 232, 267, 412, 445, 457, 585, 594, 766, 167, 613, 149, 148, 560, 589,
+   648, 768, 708, 345, 411, 704, 105, 259, 313, 496, 518, 174, 542, 120, 307, 101,
+   430, 372, 584, 183, 228, 529, 650, 697, 424, 732, 428, 349, 632, 355, 517, 110,
+   135, 147, 403, 580, 624, 700, 750, 170, 193, 245, 297, 374, 463, 543, 763, 801,
+   812, 815, 162, 384, 420, 730, 287, 330, 337, 366, 459, 476, 509, 558, 591, 610,
+   726, 652, 734, 759, 154, 163, 198, 473, 683, 697, 292, 311, 353, 423, 572, 494,
+   113, 217, 259, 280, 314, 499, 506, 603, 608, 752, 778, 782, 788, 117, 557, 748,
+   774, 320, 109, 126, 260, 265, 373, 411, 479, 523, 655, 737, 823, 380, 765, 161,
+   395, 398, 438, 451, 502, 516, 537, 583, 791, 136, 340, 769, 122, 273, 446, 727,
+   305, 322, 400, 496, 771, 155, 190, 269, 377, 391, 406, 432, 501, 519, 599, 684,
+   687, 749, 776, 175, 452, 191, 480, 510, 659, 772, 805, 813, 397, 444, 619, 566,
+   568, 575, 491, 471, 707, 111, 636, 156, 153, 288, 346, 578, 256, 435, 383, 729,
+   680, 767, 694, 295, 128, 210,   0,   0, 227,   0, 379,   0,   0, 150, 493, 525,
+   544, 551, 552, 556, 783, 576, 604,   0, 661,   0, 703,   0,   0, 735, 743,   0,
+     0,   0, 793, 794, 795, 808, 741, 773, 118, 127, 130, 166, 169, 177, 207, 213,
+   215, 226, 229, 268, 270, 317, 327, 329, 335, 369, 375, 381, 404, 441, 448, 458,
+   477, 484, 503, 539, 545, 547, 546, 548, 549, 550, 554, 555, 561, 564, 569, 591,
+   593, 595, 598, 607, 620, 625, 625, 651, 690, 695, 705, 706, 716, 717, 733, 735,
+   777, 786, 790, 315, 869, 623,   0,   0, 102, 145, 134, 115, 129, 138, 165, 171,
+   207, 202, 206, 212, 227, 231, 240, 243, 250, 254, 294, 296, 303, 308, 319, 325,
+   321, 329, 326, 335, 341, 357, 360, 362, 370, 379, 388, 389, 393, 421, 424, 438,
+   456, 454, 458, 465, 477, 535, 485, 490, 493, 507, 512, 514, 521, 522, 525, 526,
+   528, 533, 532, 541, 565, 569, 574, 586, 591, 597, 607, 637, 647, 674, 691, 693,
+   695, 698, 703, 699, 705, 704, 702, 706, 709, 717, 728, 736, 747, 754, 770, 777,
+   783, 784, 786, 787, 790, 802, 825, 848, 847, 857,  55,  65,  66, 883, 892, 916,
+   822, 824,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,1586,   0,1605,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1602,1603,1934,1935,1574,1575,1576,1577,1579,1580,1581,1583,1584,   0,
+  1585,1587,1588,1589,1591,   0,1592,   0,1593,1594,   0,1595,1596,   0,1598,1599,
+  1600,1601,1604,1582,1578,1590,1597,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1936,   0,1937,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1938,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1939,1940,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1941,1942,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1944,1943,   0,1945,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,1946,1947,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1948,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1949,1950,1951,1952,1953,1954,1955,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1956,1957,1958,1960,1959,1961,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0, 106, 104, 107, 826, 114, 118, 119, 121,
+   123, 124, 127, 125,  34, 830, 130, 131, 132, 137, 827,  35, 133, 139, 829, 142,
+   143, 112, 144, 145, 924, 151, 152,  37, 157, 158, 159, 160,  38, 165, 166, 169,
+   171, 172, 173, 174, 176, 177, 178, 179, 181, 182, 182, 182, 833, 468, 184, 185,
+   834, 187, 188, 189, 196, 192, 194, 195, 197, 199, 200, 201, 203, 204, 204, 206,
+   208, 209, 211, 218, 213, 219, 214, 216, 153, 234, 221, 222, 223, 220, 225, 224,
+   230, 835, 235, 236, 237, 238, 239, 244, 836, 837, 247, 248, 249, 246, 251,  39,
+    40, 253, 255, 255, 838, 257, 258, 259, 261, 839, 262, 263, 301, 264,  41, 266,
+   270, 272, 271, 841, 274, 842, 277, 276, 278, 281, 282,  42, 283, 284, 285, 286,
+    43, 843,  44, 289, 290, 291, 293, 934, 298, 845, 845, 621, 300, 300,  45, 852,
+   894, 302, 304,  46, 306, 309, 310, 312, 316,  48,  47, 317, 846, 318, 323, 324,
+   325, 324, 328, 329, 333, 331, 332, 334, 335, 336, 338, 339, 342, 343, 347, 351,
+   849, 350, 348, 352, 354, 359, 850, 361, 358, 356,  49, 363, 365, 367, 364,  50,
+   369, 371, 851, 376, 386, 378,  53, 381,  52,  51, 140, 141, 387, 382, 614,  78,
+   388, 389, 390, 394, 392, 856,  54, 399, 396, 402, 404, 858, 405, 401, 407,  55,
+   408, 409, 410, 413, 859, 415,  56, 417, 860, 418,  57, 419, 422, 424, 425, 861,
+   840, 862, 426, 863, 429, 431, 427, 433, 437, 441, 438, 439, 442, 443, 864, 436,
+   449, 450,  58, 454, 453, 865, 447, 460, 866, 867, 461, 466, 465, 464,  59, 467,
+   470, 469, 472, 828, 475, 868, 478, 870, 483, 485, 486, 871, 488, 489, 872, 873,
+   495, 497,  60, 498,  61,  61, 504, 505, 507, 508, 511,  62, 513, 874, 515, 875,
+   518, 844, 520, 876, 877, 878,  63,  64, 528, 880, 879, 881, 882, 530, 531, 531,
+   533,  66, 534,  67,  68, 884, 536, 538, 541,  69, 885, 549, 886, 887, 556, 559,
+    70, 561, 562, 563, 888, 889, 889, 567,  71, 890, 570, 571,  72, 891, 577,  73,
+   581, 579, 582, 893, 587,  74, 590, 592, 596,  75, 895, 896,  76, 897, 600, 898,
+   602, 605, 607, 899, 900, 609, 901, 611, 853,  77, 615, 616,  79, 617, 252, 902,
+   903, 854, 855, 621, 622, 731,  80, 627, 626, 628, 164, 629, 630, 631, 633, 904,
+   632, 634, 639, 640, 635, 641, 646, 651, 638, 643, 644, 645, 905, 907, 906,  81,
+   653, 654, 656, 911, 657, 908,  82,  83, 909, 910,  84, 664, 665, 666, 667, 669,
+   668, 671, 670, 674, 672, 673, 675,  85, 677, 678,  86, 681, 682, 912, 685, 686,
+    87, 689,  36, 913, 914,  88,  89, 696, 702, 709, 711, 915, 712, 713, 718, 719,
+   917, 831, 721, 720, 723, 832, 725, 728, 918, 919, 739, 742, 744, 920, 745, 753,
+   756, 757, 755, 760, 761, 921, 762,  90, 764, 922,  91, 775, 279, 780, 923, 925,
+    92,  93, 785, 926,  94, 927, 787, 787, 789, 928, 792,  95, 796, 797, 798, 800,
+    96, 929, 802, 804, 806,  97,  98, 807, 930,  99, 931, 932, 933, 814, 100, 816,
+   817, 818, 819, 820, 821, 935,   0,   0,
+};
+static const int16_t
+_hb_ucd_i16[196] =
+{
+      0,    0,    0,    0,    1,   -1,    0,    0,    2,    0,   -2,    0,    0,    0,    0,    2,
+      0,   -2,    0,    0,    0,    0,    0,   16,    0,    0,    0,  -16,    0,    0,    1,   -1,
+      0,    0,    0,    1,   -1,    0,    0,    0,    0,    1,   -1,    0,    3,    3,    3,   -3,
+     -3,   -3,    0,    0,    0, 2016,    0,    0,    0,    0,    0, 2527, 1923, 1914, 1918,    0,
+   2250,    0,    0,    0,    0,    0,    0,  138,    0,    7,    0,    0,   -7,    0,    0,    0,
+      1,   -1,    1,   -1,   -1,    1,   -1,    0, 1824,    0,    0,    0,    0,    0, 2104,    0,
+   2108, 2106,    0, 2106, 1316,    0,    0,    0,    0,    1,   -1,    1,   -1, -138,    0,    0,
+      1,   -1,    8,    8,    8,    0,    7,    7,    0,    0,   -8,   -8,   -8,   -7,   -7,    0,
+      1,   -1,    0,    2,-1316,    1,   -1,    0,   -1,    1,   -1,    1,   -1,    3,    1,   -1,
+     -3,    1,   -1,    1,   -1,    0,    0,-1914,-1918,    0,    0,-1923,-1824,    0,    0,    0,
+      0,-2016,    0,    0,    1,   -1,    0,    1,    0,    0,-2104,    0,    0,    0,    0,-2106,
+  -2108,-2106,    0,    0,    1,   -1,-2250,    0,    0,    0,-2527,    0,    0,   -2,    0,    1,
+     -1,    0,    1,   -1,
+};
+
+static inline uint_fast8_t
+_hb_ucd_gc (unsigned u)
+{
+  return u<1114110u?_hb_ucd_u8[6504+(((_hb_ucd_u8[1264+(((_hb_ucd_u16[((_hb_ucd_u8[544+(((_hb_ucd_u8[u>>1>>3>>3>>4])<<4)+((u>>1>>3>>3)&15u))])<<3)+((u>>1>>3)&7u)])<<3)+((u>>1)&7u))])<<1)+((u)&1u))]:2;
+}
+static inline uint_fast8_t
+_hb_ucd_ccc (unsigned u)
+{
+  return u<125259u?_hb_ucd_u8[8768+(((_hb_ucd_u8[7792+(((_hb_ucd_u8[7120+(((_hb_ucd_u8[6874+(u>>2>>3>>4)])<<4)+((u>>2>>3)&15u))])<<3)+((u>>2)&7u))])<<2)+((u)&3u))]:0;
+}
+static inline unsigned
+_hb_ucd_b4 (const uint8_t* a, unsigned i)
+{
+  return (a[i>>1]>>((i&1u)<<2))&15u;
+}
+static inline int_fast16_t
+_hb_ucd_bmg (unsigned u)
+{
+  return u<65380u?_hb_ucd_i16[((_hb_ucd_u8[9508+(((_hb_ucd_u8[9388+(((_hb_ucd_b4(9260+_hb_ucd_u8,u>>2>>3>>3))<<3)+((u>>2>>3)&7u))])<<3)+((u>>2)&7u))])<<2)+((u)&3u)]:0;
+}
+static inline uint_fast8_t
+_hb_ucd_sc (unsigned u)
+{
+  return u<918000u?_hb_ucd_u8[10974+(((_hb_ucd_u16[1960+(((_hb_ucd_u8[10286+(((_hb_ucd_u8[9836+(u>>3>>4>>4)])<<4)+((u>>3>>4)&15u))])<<4)+((u>>3)&15u))])<<3)+((u)&7u))]:2;
+}
+static inline uint_fast16_t
+_hb_ucd_dm (unsigned u)
+{
+  return u<195102u?_hb_ucd_u16[5768+(((_hb_ucd_u8[16708+(((_hb_ucd_u8[16326+(u>>4>>5)])<<5)+((u>>4)&31u))])<<4)+((u)&15u))]:0;
+}
+
+
+#else
+
+static const uint8_t
+_hb_ucd_u8[13344] =
+{
+    0,  1,  2,  3,  4,  5,  5,  5,  5,  5,  6,  5,  5,  7,  8,  9,
+   10, 11, 12, 13, 14, 15, 16,  5, 17, 15, 15, 18, 15, 19, 20, 21,
+    5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5,  5, 22, 23,
+    5, 24, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+   25, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15,
+    8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+    0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15,
+   16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31,
+   32, 33, 34, 34, 34, 34, 35, 36, 37, 34, 34, 34, 38, 39, 40, 41,
+   42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57,
+   58, 59, 60, 61, 62, 62, 63, 64, 65, 66, 67, 68, 69, 67, 70, 71,
+   67, 67, 62, 72, 62, 62, 73, 67, 74, 75, 76, 77, 78, 67, 67, 67,
+   79, 80, 34, 81, 82, 83, 67, 67, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 84, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   85, 34, 34, 34, 34, 34, 34, 34, 34, 86, 34, 34, 87, 88, 89, 90,
+   91, 92, 93, 94, 95, 96, 97, 98, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
+  100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,
+  100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,
+  100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,100,
+  100,100, 34, 34, 34, 34,101,102, 34, 34,103,104,105,106,107,108,
+   34, 34,109,110,111,112,113,114,115,116,117,111, 34, 34, 34,111,
+  118,119,120,121,122,123,124,125, 34,126,127,111,128,129,130,131,
+  132,133,134,135,136,137,138,111,139,140,111,141,142,143,144,111,
+  145,146,147,148,149,150,111,111,151,152,153,154,111,155,111,156,
+   34, 34, 34, 34, 34, 34, 34, 34,157, 34, 34,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+   34, 34, 34, 34, 34, 34, 34, 34,158,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111, 34, 34, 34, 34, 34,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+   34, 34, 34, 34,159,160,161, 34,111,111,111,111,162,163,164,165,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,111,111,111,111,111,
+   34, 34, 34, 34, 34, 34,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111, 34,166,111,111,111,111,111,111,
+   67, 67,167,168,169,128, 65,111,170,171,172,173,174,175,176,177,
+   67, 67, 67, 67,178,179,111,111,111,111,111,111,111,111,111,111,
+  180,111,181,111,111,182,111,111,111,111,111,111,111,111,111,111,
+   34,183,184,111,111,111,111,111,128,185,186,111, 34,187,111,111,
+   67, 67,188, 67, 67,111, 67,189, 67, 67, 67, 67, 67, 67, 67, 67,
+   67, 67, 67, 67, 67, 67, 67,190,111,111,111,111,111,111,111,111,
+   34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34,
+   34, 34, 34, 34, 34, 34, 34, 34,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+   34, 34, 34, 34, 34,111,111,111,111,111,111,111,111,111,111,111,
+   34, 34, 34, 34, 34, 34, 34,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+  191,111,180,180,111,111,111,111,111,111,111,111,111,111,111,111,
+  111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,111,
+    0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,  2,  4,  5,  6,  2,
+    7,  7,  7,  7,  7,  2,  8,  9, 10, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 17, 18, 19,  1, 20, 20, 21, 22, 23, 24, 25,
+   26, 27, 15,  2, 28, 29, 27, 30, 11, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 31, 11, 11, 11, 32, 16, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 33, 16, 16, 16, 16, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 34, 34, 34, 34, 34, 34, 34, 34, 16, 32, 32, 32,
+   32, 32, 32, 32, 11, 34, 34, 16, 34, 32, 32, 11, 34, 11, 16, 11,
+   11, 34, 32, 11, 32, 16, 11, 34, 32, 32, 32, 11, 34, 16, 32, 11,
+   34, 11, 34, 34, 32, 35, 32, 16, 36, 36, 37, 34, 38, 37, 34, 34,
+   34, 34, 34, 34, 34, 34, 16, 32, 34, 38, 32, 11, 32, 32, 32, 32,
+   32, 32, 16, 16, 16, 11, 34, 32, 34, 34, 11, 32, 32, 32, 32, 32,
+   16, 16, 39, 16, 16, 16, 16, 16, 40, 40, 40, 40, 40, 40, 40, 40,
+   40, 41, 41, 40, 40, 40, 40, 40, 40, 41, 41, 41, 41, 41, 41, 41,
+   40, 40, 42, 41, 41, 41, 42, 42, 41, 41, 41, 41, 41, 41, 41, 41,
+   43, 43, 43, 43, 43, 43, 43, 43, 32, 32, 42, 32, 16, 44, 16, 10,
+   41, 41, 41, 45, 11, 11, 11, 11, 34, 11, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 16, 16, 16, 16, 16, 16, 16, 16, 16, 34,
+   16, 11, 32, 16, 32, 32, 32, 32, 16, 16, 32, 46, 34, 32, 34, 11,
+   32, 47, 43, 43, 48, 32, 32, 32, 11, 34, 34, 34, 34, 34, 34, 16,
+   11, 11, 11, 11, 49,  2,  2,  2, 16, 16, 16, 16, 50, 51, 52, 53,
+   54, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 55,
+   56, 57, 43, 56, 43, 43, 43, 43, 36, 36, 36, 36, 36, 36, 36, 36,
+   36, 58,  2,  2,  2,  2,  2,  2, 59, 59, 59,  8,  9, 60,  2, 61,
+   43, 43, 43, 43, 43, 57, 59,  2, 62, 36, 36, 36, 36, 63, 43, 43,
+    7,  7,  7,  7,  7,  2,  2, 36, 64, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 65, 43, 43, 43, 66, 47, 43, 43, 67, 68, 69, 43, 43, 36,
+    7,  7,  7,  7,  7, 36, 70, 71,  2,  2,  2,  2,  2,  2,  2, 72,
+   63, 36, 36, 36, 36, 36, 36, 36, 43, 43, 43, 43, 43, 43, 64, 36,
+   36, 36, 36, 43, 43, 43, 43, 43,  7,  7,  7,  7,  7, 36, 36, 36,
+   36, 36, 36, 36, 36, 63, 43, 43, 43, 43, 40, 21,  2, 40, 68, 20,
+   36, 36, 36, 43, 43, 68, 43, 43, 43, 43, 68, 43, 68, 43, 43, 43,
+    2,  2,  2,  2,  2,  2,  2,  2, 36, 36, 36, 36, 63, 43, 43,  2,
+   36, 63, 43, 43, 43, 43, 43, 43, 43, 73, 43, 43, 43, 43, 43, 43,
+   43, 74, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 74, 64, 75,
+   76, 43, 43, 43, 74, 75, 76, 75, 63, 43, 43, 43, 36, 36, 36, 36,
+   36, 43,  2,  7,  7,  7,  7,  7, 77, 36, 36, 36, 36, 36, 36, 36,
+   63, 75, 78, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 64, 75,
+   76, 43, 43, 74, 75, 75, 76, 36, 36, 36, 36, 79, 75, 75, 36, 36,
+   36, 43, 43,  7,  7,  7,  7,  7, 36, 20, 27, 27, 27, 53, 58, 43,
+   43, 74, 78, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 43, 75,
+   76, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 43, 64, 36, 36, 36,
+   36, 36, 36,  7,  7,  7,  7,  7, 43, 36, 63,  2,  2,  2,  2,  2,
+   76, 43, 43, 43, 74, 75, 76, 43, 60, 20, 20, 20, 80, 43, 43, 43,
+   43, 75, 78, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 64, 76,
+   76, 43, 43, 74, 75, 75, 76, 43, 43, 43, 43, 74, 75, 75, 36, 36,
+   71, 27, 27, 27, 27, 27, 27, 27, 43, 64, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 75, 74, 75, 75, 75, 75, 75, 76, 43,
+   36, 36, 36, 79, 75, 75, 75, 75, 75, 75, 75,  7,  7,  7,  7,  7,
+   27, 81, 61, 61, 53, 61, 61, 61, 74, 75, 64, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36, 43, 74, 75, 75, 43, 43, 43, 43, 43,
+   43, 43, 43, 43, 36, 36, 36, 36,  7,  7,  7, 82, 27, 27, 27, 81,
+   63, 75, 65, 36, 36, 36, 36, 36, 75, 75, 75, 74, 75, 75, 43, 43,
+   43, 43, 74, 75, 75, 75, 75, 36, 83, 36, 36, 36, 36, 36, 36, 36,
+   43, 75, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 63, 64, 75,
+   76, 43, 43, 75, 75, 75, 76, 70, 61, 61, 36, 79, 27, 27, 27, 84,
+   27, 27, 27, 27, 81, 36, 36, 36, 36, 36, 36, 36, 36, 43, 43, 74,
+   75, 43, 43, 43, 75, 75, 75, 75,  7, 75,  2,  2,  2,  2,  2,  2,
+   63, 36, 43, 43, 43, 43, 43, 85, 36, 36, 36, 68, 43, 43, 43, 57,
+    7,  7,  7,  7,  7,  2,  2,  2, 63, 36, 43, 43, 43, 43, 64, 36,
+   36, 36, 36, 40, 43, 43, 43, 43,  7,  7,  7,  7,  7,  7, 36, 36,
+   70, 61,  2,  2,  2,  2,  2,  2,  2, 86, 86, 61, 43, 61, 61, 61,
+    7,  7,  7,  7,  7, 27, 27, 27, 27, 27, 47, 47, 47,  4,  4, 75,
+   63, 43, 43, 43, 43, 43, 43, 74, 43, 43, 57, 43, 36, 36, 63, 43,
+   43, 43, 43, 43, 43, 43, 43, 61, 61, 61, 61, 69, 61, 61, 61, 61,
+    2,  2, 86, 61, 21,  2,  2,  2, 36, 36, 36, 36, 36, 79, 76, 43,
+   74, 43, 43, 43, 76, 74, 76, 64, 36, 36, 36, 75, 43, 36, 36, 43,
+   64, 75, 78, 79, 75, 75, 75, 36, 63, 43, 64, 36, 36, 36, 36, 36,
+   36, 74, 76, 74, 75, 75, 76, 79,  7,  7,  7,  7,  7, 75, 76, 61,
+   16, 16, 16, 16, 16, 50, 44, 16, 36, 36, 36, 36, 36, 36, 63, 43,
+    2,  2,  2,  2, 87, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27,
+   61, 61, 61, 61, 61, 61, 61, 61, 11, 11, 11, 11, 16, 16, 16, 16,
+   88, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 70, 65,
+   89, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 90, 91, 91,
+   36, 36, 36, 36, 36, 58,  2, 92, 93, 36, 36, 36, 36, 36, 36, 36,
+   36, 43, 43, 43, 43, 43, 43, 43, 36, 43, 57,  2,  2,  2,  2,  2,
+   36, 36, 43, 76, 43, 43, 43, 75, 75, 75, 75, 74, 76, 43, 43, 43,
+   43, 43,  2, 77,  2, 60, 63, 43,  7,  7,  7,  7,  7,  7,  7,  7,
+    2,  2,  2, 94,  2, 56, 43, 59, 36, 95, 36, 36, 36, 36, 36, 36,
+   36, 36, 63, 64, 36, 36, 36, 36, 36, 36, 36, 36, 63, 36, 36, 36,
+   43, 74, 75, 76, 74, 75, 75, 75, 75, 74, 75, 75, 76, 43, 43, 43,
+   61, 61,  2,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 27, 27, 61,
+   36, 36, 36, 63, 74, 76, 43,  2, 36, 36, 79, 74, 43, 43, 43, 43,
+   74, 74, 76, 43, 43, 43, 74, 75, 75, 76, 43, 43, 43, 43, 43, 43,
+    2,  2,  2, 77,  2,  2,  2,  2, 43, 43, 43, 43, 43, 43, 43, 96,
+   43, 43, 78, 36, 36, 36, 36, 36, 36, 36, 74, 43, 43, 74, 74, 75,
+   75, 74, 78, 36, 36, 36, 36, 36, 86, 61, 61, 61, 61, 47, 43, 43,
+   43, 43, 61, 61, 61, 61, 61, 61, 43, 78, 36, 36, 36, 36, 36, 36,
+   79, 43, 43, 75, 43, 76, 43, 36, 36, 36, 36, 74, 43, 75, 76, 76,
+   43, 75, 75, 75, 75, 75,  2,  2, 36, 36, 75, 75, 75, 75, 43, 43,
+   43, 43, 75, 43, 43, 57,  2,  2,  7,  7,  7,  7,  7,  7, 83, 36,
+   36, 36, 36, 36, 40, 40, 40,  2, 43, 57, 43, 43, 43, 43, 43, 43,
+   74, 43, 43, 43, 64, 36, 63, 36, 36, 36, 64, 79, 43, 36, 36, 36,
+   16, 16, 16, 16, 16, 16, 40, 40, 40, 40, 40, 40, 40, 44, 16, 16,
+   16, 16, 16, 16, 44, 16, 16, 16, 16, 16, 16, 16, 16, 97, 40, 40,
+   32, 32, 32, 16, 16, 16, 16, 32, 16, 16, 16, 16, 11, 11, 11, 11,
+   16, 16, 16, 16, 34, 11, 11, 11, 16, 16, 16, 16, 98, 98, 98, 98,
+   16, 16, 16, 16, 11, 11, 99,100, 41, 16, 16, 16, 11, 11, 99, 41,
+   16, 16, 16, 16, 11, 11,101, 41,102,102,102,102,102,103, 59, 59,
+   51, 51, 51,  2,104,105,104,105,  2,  2,  2,  2,106, 59, 59,107,
+    2,  2,  2,  2,108,109,  2,110,111,  2,112,113,  2,  2,  2,  2,
+    2,  9,111,  2,  2,  2,  2,114, 59, 59, 59, 59, 59, 59, 59, 59,
+  115, 40, 27, 27, 27,  8,112,116, 27, 27, 27, 27, 27,  8,112, 91,
+   20, 20, 20, 20, 20, 20, 20, 20, 43, 43, 43, 43, 43, 43,117, 48,
+   96, 48, 96, 43, 43, 43, 43, 43, 61,118, 61,119, 61, 34, 11, 16,
+   11, 32,119, 61, 46, 11, 11, 61, 61, 61,118,118,118, 11, 11,120,
+   11, 11, 35, 36, 39, 61, 16, 11,  8,  8, 46, 16, 16, 26, 61,121,
+   92, 92, 92, 92, 92, 92, 92, 92, 92,122,123, 92,124, 61, 61, 61,
+    8,  8,125, 61, 61,  8, 61, 61,125, 26, 61,125, 61, 61, 61,125,
+   61, 61, 61, 61, 61, 61, 61,  8, 61,125,125, 61, 61, 61, 61, 61,
+   61, 61,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,  8,
+   61, 61, 61, 61,  4,  4, 61, 61,  8, 61, 61, 61,126,127, 61, 61,
+   61, 61, 61, 61, 61, 61,125, 61, 61, 61, 61, 61, 61, 26,  8,  8,
+    8,  8, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61,  8,  8,
+    8, 61, 61, 61, 61, 61, 61, 61, 27, 27, 27, 27, 27, 27, 61, 61,
+   61, 61, 61, 61, 61, 27, 27, 27, 61, 61, 61, 26, 61, 61, 61, 61,
+   26, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61,  8,  8,  8,  8,
+   61, 61, 61, 61, 61, 61, 61, 26, 61, 61, 61, 61,  4,  4,  4,  4,
+    4,  4,  4, 27, 27, 27, 27, 27, 27, 27, 61, 61, 61, 61, 61, 61,
+    8,  8,112,128,  8,  8,  8,  8,  8,  8,  8,  4,  4,  4,  4,  4,
+    8,112,129,129,129,129,129,129,129,129,129,129,128,  8,  8,  8,
+    8,  8,  8,  8,  4,  4,  8,  8,  8,  8,  8,  8,  8,  8,  4,  8,
+    8,  8,125, 26,  8,  8,125, 61, 32, 11, 32, 34, 34, 34, 34, 11,
+   32, 32, 34, 16, 16, 16, 40, 11, 32, 32,121, 61, 61,119, 34,130,
+   43, 32, 16, 16, 50,  2, 87,  2, 36, 36, 36, 36, 36, 36, 36, 95,
+    2,  2,  2,  2,  2,  2,  2, 56,  2,104,104,  2,108,109,104,  2,
+    2,  2,  2,  6,  2, 94,104,  2,104,  4,  4,  4,  4,  2,  2, 77,
+    2,  2,  2,  2,  2, 51,  2,  2, 94,131,  2,  2,  2,  2,  2,  2,
+   61,  2,  2,  2,  2,  2,  2,  2,  1,  2,132,133,  4,  4,  4,  4,
+    4, 61,  4,  4,  4,  4,134, 91,135, 92, 92, 92, 92, 43, 43, 75,
+  136, 40, 40, 61, 92,137, 58, 61, 71, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 63,138,139, 62, 36, 36, 36, 36, 36, 58, 40, 62,
+   61, 27, 27, 61, 61, 61, 61, 61, 27, 27, 27, 27, 27, 61, 61, 61,
+   61, 61, 61, 61, 27, 27, 27, 27,140, 27, 27, 27, 27, 27, 27, 27,
+   36, 36, 95, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36,141,  2,
+   32, 32, 32, 32, 32, 32, 32, 63, 48,142, 43, 43, 43, 43, 43, 77,
+   32, 32, 32, 32, 32, 32, 40, 43, 36, 36, 36, 92, 92, 92, 92, 92,
+   43,  2,  2,  2,  2,  2,  2,  2, 41, 41, 41,139, 40, 40, 40, 40,
+   41, 32, 32, 32, 32, 32, 32, 32, 16, 32, 32, 32, 32, 32, 32, 32,
+   44, 16, 16, 16, 34, 34, 34, 32, 32, 32, 32, 32, 42,143, 34, 35,
+   32, 32, 16, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 11, 11, 32,
+   11, 11, 32, 32, 32, 32, 32, 32, 16, 32, 11, 11, 34, 16, 16, 16,
+   16, 16, 34, 35, 40, 35, 36, 36, 36, 64, 36, 64, 36, 63, 36, 36,
+   36, 79, 76, 74, 61, 61, 43, 43, 27, 27, 27, 61,144, 61, 61, 61,
+   36, 36,  2,  2,  2,  2,  2,  2, 75, 36, 36, 36, 36, 36, 36, 36,
+   36, 36, 75, 75, 75, 75, 75, 75, 75, 75, 43, 43, 43, 43, 43,  2,
+   43, 36, 36, 36,  2, 65, 65, 63, 36, 36, 36, 43, 43, 43, 43,  2,
+   36, 36, 36, 63, 43, 43, 43, 43, 43, 75, 75, 75, 75, 75, 75,145,
+   36, 63, 75, 43, 43, 75, 43, 75,145,  2,  2,  2,  2,  2,  2, 77,
+    7,  7,  7,  7,  7,  7,  7,  2, 36, 36, 63, 62, 36, 36, 36, 36,
+   36, 36, 36, 36, 63, 43, 43, 74, 76, 74, 76, 43, 43, 43, 43, 43,
+   36, 63, 36, 36, 36, 36, 74, 75,  7,  7,  7,  7,  7,  7,  2,  2,
+   62, 36, 36, 70, 61, 79, 74, 36, 64, 43, 64, 63, 64, 36, 36, 43,
+   36, 36, 36, 36, 36, 36, 95,  2, 36, 36, 36, 36, 36, 79, 43, 75,
+    2, 95,146, 43, 43, 43, 43, 43, 16, 16, 16, 16, 16,100, 40, 40,
+   16, 16, 16, 16, 97, 41, 41, 41, 36, 79, 76, 75, 74,145, 76, 43,
+  147,147,147,147,147,147,147,147,148,148,148,148,148,148,148,148,
+   16, 16, 16, 16, 16, 16, 35, 64, 36, 36, 36, 36,149, 36, 36, 36,
+   36, 41, 41, 41, 41, 41, 41, 41, 41,150, 36, 36, 36, 36, 36, 36,
+   36, 36, 36, 36, 36, 36, 36,129,151,151,151,151,151,151,151,151,
+   36, 36, 36, 36, 36, 36,144, 61,  2,  2,  2,152,113,  2,  2,  2,
+    6,153,154,129,129,129,129,129,129,129,113,152,113,  2,110,155,
+    2,  2,  2,  2,134,129,129,113,  2,156,  8,  8, 60,  2,  2,  2,
+   36, 36, 36, 36, 36, 36, 36,157,  2,  2,  3,  2,  4,  5,  6,  2,
+   16, 16, 16, 16, 16, 17, 18,112,113,  4,  2, 36, 36, 36, 36, 36,
+   62, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 40,
+   20,158, 53, 20, 26,  8,125, 61, 61, 61, 61, 61,159, 59, 61, 61,
+    2,  2,  2, 87, 27, 27, 27, 27, 27, 27, 27, 81, 61, 61, 61, 61,
+   92, 92,124, 27, 81, 61, 61, 61, 61, 61, 61, 61, 61, 27, 61, 61,
+   61, 61, 61, 61, 61, 61, 47, 43,160,160,160,160,160,160,160,160,
+  161, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 27, 84, 36,
+  133, 36, 36, 36, 36, 92, 92, 92, 36, 36, 36, 36, 36, 36, 36, 58,
+  162, 92, 92, 92, 92, 92, 92, 92, 36, 36, 36, 58, 27, 27, 27, 27,
+   36, 36, 36, 70,140, 27, 27, 27, 36, 36, 36,163, 27, 27, 27, 27,
+   36, 36, 36, 36, 36,163, 27, 27, 36, 36, 36, 27, 27, 27, 27, 30,
+   36, 36, 36, 36, 36, 36, 27, 36, 63, 43, 43, 43, 43, 43, 43, 43,
+   36, 36, 36, 36, 43, 43, 43, 43, 36, 36, 36, 36, 36, 36,163, 30,
+   36, 36, 36, 36, 36, 36,163, 27, 36, 36, 36, 36, 71, 36, 36, 36,
+   36, 36, 63, 43, 43,161, 27, 27, 36, 36, 36, 36, 58,  2,  2,  2,
+   36, 36, 36, 36, 27, 27, 27, 27, 16, 16, 16, 16, 16, 27, 27, 27,
+   36, 36, 43, 43, 43, 43, 43, 43, 36, 36, 36, 36, 36, 63,164, 51,
+   27, 27, 27, 84, 36, 36, 36, 36,161, 27, 30,  2,  2,  2,  2,  2,
+   36, 36,163, 27, 27, 27, 27, 27, 76, 78, 36, 36, 36, 36, 36, 36,
+   43, 43, 43, 57,  2,  2,  2,  2,  2, 27, 27, 27, 27, 27, 27, 27,
+   27, 27, 27,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,165,
+   75, 76, 43, 74, 76, 57, 72,  2,  2,  2,  2,  2,  2,  2, 72, 59,
+   36, 36, 36, 63, 43, 43, 76, 43, 43, 43, 43,  7,  7,  7,  7,  7,
+    2,  2, 79, 78, 36, 36, 36, 36, 36, 63,  2, 36, 36, 36, 36, 36,
+   36, 79, 75, 43, 43, 43, 43, 74, 78, 36, 58,  2, 56, 43, 57, 76,
+    7,  7,  7,  7,  7, 58, 58,  2, 87, 27, 27, 27, 27, 27, 27, 27,
+   36, 36, 36, 36, 36, 36, 75, 76, 43, 75, 74, 43,  2,  2,  2, 43,
+   36, 36, 36, 36, 36, 36, 36, 63, 74, 75, 75, 75, 75, 75, 75, 75,
+   36, 36, 36, 79, 75, 75, 78, 36, 36, 75, 75, 43, 43, 43, 43, 43,
+   36, 36, 79, 75, 43, 43, 43, 43, 75, 43, 74, 64, 36, 58,  2,  2,
+    7,  7,  7,  7,  7,  2,  2, 64, 75, 76, 43, 43, 74, 74, 75, 76,
+   74, 43, 36, 65, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 79,
+   75, 43, 43, 43, 75, 75, 43, 76, 57,  2,  2,  2,  2,  2,  2,  2,
+    2,  2,  2,  2, 36, 36, 43, 43, 75, 76, 43, 43, 43, 74, 76, 76,
+   57,  2, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 63, 76, 75,
+   43, 43, 43, 76, 36, 36, 36, 36, 75, 43, 43, 76, 43, 43, 43, 43,
+    7,  7,  7,  7,  7, 27,  2, 86, 43, 43, 43, 43, 76, 57,  2,  2,
+   27, 27, 27, 27, 27, 27, 27, 84, 75, 75, 75, 75, 75, 76, 74, 64,
+   78, 76,  2,  2,  2,  2,  2,  2, 79, 75, 43, 43, 43, 43, 75, 75,
+   64, 65, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75,
+   63, 43, 43, 43, 43, 64, 36, 36, 36, 63, 43, 43, 74, 63, 43, 57,
+    2,  2,  2, 56, 43, 43, 43, 43, 63, 43, 43, 74, 76, 43, 36, 36,
+   36, 36, 36, 36, 36, 43, 43, 43, 43, 43, 43, 74, 43,  2, 65,  2,
+   43, 43, 43, 43, 43, 43, 43, 76, 58,  2,  2,  2,  2,  2,  2,  2,
+    2, 36, 36, 36, 36, 36, 36, 36, 43, 43, 43, 43, 74, 43, 43, 43,
+   74, 43, 76, 43, 43, 43, 43, 43, 43, 43, 43, 63, 43, 43, 43, 43,
+   36, 36, 36, 36, 36, 75, 75, 75, 43, 74, 76, 76, 36, 36, 36, 36,
+   36, 63, 74,145,  2,  2,  2,  2, 27, 27, 81, 61, 61, 61, 53, 20,
+  144, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 61, 21,
+   43, 43, 57,  2,  2,  2,  2,  2, 43, 43, 43, 57,  2,  2, 61, 61,
+   40, 40, 86, 61, 61, 61, 61, 61,  7,  7,  7,  7,  7,166, 27, 27,
+   27, 84, 36, 36, 36, 36, 36, 36, 27, 27, 27, 30,  2,  2,  2,  2,
+   79, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 75, 76,
+   43, 67, 40, 40, 40, 40, 40, 40, 40, 77, 43, 43, 43, 43, 43, 43,
+   36, 36, 36, 36, 36, 36, 47, 57, 61, 61,167, 76, 43, 61,167, 75,
+   75,168, 59, 59, 59, 73, 43, 43, 43, 69, 47, 43, 43, 43, 61, 61,
+   61, 61, 61, 61, 61, 43, 43, 61, 61, 43, 69, 61, 61, 61, 61, 61,
+   11, 11, 11, 11, 11, 16, 16, 16, 16, 16, 11, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 11, 16, 11, 16, 16, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 16, 11, 11, 11, 11, 11, 16, 16, 16, 16, 16,
+   31, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 33, 16, 16,
+   16, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 31, 16, 16,
+   16, 16, 33, 16, 16, 16, 11, 11, 11, 11, 31, 16, 16, 16, 16, 16,
+   16, 16, 16, 16, 16, 16, 16, 33, 16, 16, 16, 11, 11, 11, 11, 11,
+   11, 11, 11, 11, 11, 11, 11, 31, 16, 16, 16, 16, 33, 16, 16, 16,
+   11, 11, 11, 11, 31, 16, 16, 16, 16, 33, 16, 16, 16, 32, 16,  7,
+   43, 43, 43, 69, 61, 47, 43, 43, 43, 43, 43, 43, 43, 43, 69, 61,
+   61, 61, 47, 61, 61, 61, 61, 61, 61, 61, 69, 21,  2,  2,  2,  2,
+    2,  2,  2,  2,  2, 56, 43, 43, 43, 43, 43, 67, 40, 40, 40, 40,
+    7,  7,  7,  7,  7,  7,  7, 70, 36, 36, 36, 36, 36, 36, 43, 43,
+    7,  7,  7,  7,  7,  7,  7,169, 16, 16, 43, 43, 43, 67, 40, 40,
+   27, 27, 27, 27, 27, 27,140, 27,170, 27, 27, 27, 27, 27, 27, 27,
+   27, 27, 27, 27, 27, 27, 27,140, 27, 27, 27, 27, 27, 27, 81, 61,
+   61, 61, 61, 61, 61, 25, 41, 41,  0,  0, 29, 21, 21, 21, 23, 21,
+   22, 18, 21, 25, 21, 17, 13, 13, 25, 25, 25, 21, 21,  9,  9,  9,
+    9, 22, 21, 18, 24, 16, 24,  5,  5,  5,  5, 22, 25, 18, 25,  0,
+   23, 23, 26, 21, 24, 26,  7, 20, 25,  1, 26, 24, 26, 25, 15, 15,
+   24, 15,  7, 19, 15, 21,  9, 25,  9,  5,  5, 25,  5,  9,  5,  7,
+    7,  7,  9,  8,  8,  5,  7,  5,  6,  6, 24, 24,  6, 24, 12, 12,
+    6,  5,  9, 21, 25,  9, 26, 12, 11, 11,  9,  6,  5, 21, 17, 17,
+   17, 26, 26, 23, 23, 12, 17, 12, 21, 12, 12, 21,  7, 21,  1,  1,
+   21, 23, 26, 26,  6,  7,  7, 12, 12,  7, 21,  7, 12,  1, 12,  6,
+    6, 12, 12, 26,  7, 26, 26,  7, 21,  1,  1, 12, 12, 10, 10, 10,
+   10, 12, 21,  6, 10,  7,  7, 10, 23,  7, 15, 26, 13, 21, 13,  7,
+   15,  7, 12, 23, 21, 26, 21, 15, 17,  7, 29,  7,  7, 22, 18, 18,
+   14, 14, 14,  7, 17, 21,  7,  6, 11, 12,  5,  6,  8,  8,  8, 24,
+    5, 24,  9, 24, 29, 29, 29,  1, 20, 19, 22, 20, 27, 28,  1, 29,
+   21, 20, 19, 21, 21, 16, 16, 21, 25, 22, 18, 21, 21, 29, 15,  6,
+   18,  6, 12, 11,  9, 26, 26,  9, 26,  5,  5, 26, 14,  9,  5, 14,
+   14, 15, 25, 26, 26, 22, 18, 26, 18, 25, 18, 22,  5, 12, 22, 21,
+   26,  6,  7, 14, 17, 22, 26, 14, 17,  6, 14,  6, 12, 24, 24,  6,
+   26, 15,  6, 21, 11, 21, 24,  9, 23, 26, 10, 21,  6, 10,  4,  4,
+    3,  3,  7, 25, 24,  7, 22, 22, 21, 22, 17, 16, 16, 22, 16, 16,
+   25, 17,  7,  1, 25, 24, 26,  1,  2,  2, 12, 15, 21, 14,  7, 15,
+   12, 17, 13, 12, 13, 15, 26, 10, 10,  1, 13, 23, 23, 15,  0,  1,
+    2,  3,  4,  5,  6,  7,  8,  9,  0, 10, 11, 12, 13,  0, 14,  0,
+    0,  0,  0,  0, 15,  0, 16,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 17, 18, 19,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 20,  0, 21, 22, 23,
+    0,  0,  0, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 34,  0, 35,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 36,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 37, 38,  0,  0,  0,  0,  0,  0, 39, 40,
+    0,  0, 41,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    1,  2,  0,  0,  0,  0,  3,  0,  0,  0,  4,  5,  6,  7,  0,  8,
+    9, 10,  0, 11, 12, 13,  0, 14, 15, 16, 15, 17, 15, 18, 15, 18,
+   15, 18,  0, 18,  0, 19, 15, 18, 20, 18,  0, 21, 22, 23, 24, 25,
+   26, 27, 28, 29, 30,  0, 31,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0, 32,  0,  0,  0,  0,  0,  0, 33,  0,  0, 34,  0,  0, 35,  0,
+   36,  0,  0,  0, 37, 38, 39, 40, 41, 42, 43, 44, 45,  0,  0, 46,
+    0,  0,  0, 47,  0,  0,  0, 48,  0,  0,  0,  0,  0,  0,  0, 49,
+    0, 50,  0, 51, 52,  0, 53,  0,  0,  0,  0,  0,  0, 54, 55, 56,
+    0,  0,  0,  0, 57,  0,  0, 58, 59, 60, 61, 62,  0,  0, 63, 64,
+    0,  0,  0, 65,  0,  0,  0,  0, 66,  0,  0,  0, 67,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 68,  0,  0,  0, 69,
+    0, 70,  0,  0, 71,  0,  0, 72,  0,  0,  0,  0,  0,  0,  0,  0,
+   73,  0,  0,  0,  0,  0, 74,  0,  0, 75,  0,  0,  0, 76, 77,  0,
+   78, 61,  0, 79, 80,  0,  0, 81, 82, 83,  0,  0,  0, 84,  0, 85,
+    0,  0, 50, 86, 50,  0, 87,  0, 88,  0,  0,  0, 77,  0,  0,  0,
+   89, 90,  0, 91, 92, 93, 94,  0,  0,  0,  0,  0, 50,  0,  0,  0,
+    0, 95, 96,  0,  0,  0,  0, 97, 98,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0, 99,  0,  0,100,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,101,102,  0,  0,103,  0,  0,  0,  0,  0,  0,104,  0,  0,  0,
+   98,  0,  0,  0,  0,  0,  0,105,  0,  0,  0,  0,  0,  0,  0,106,
+    0,107,  0,  0,  0,  0,  0,  0,  1,  2,  3,  4,  5,  6,  7,  0,
+    8,  0,  0,  0,  0,  9, 10, 11, 12,  0,  0,  0,  0, 13,  0,  0,
+   14, 15,  0, 16,  0, 17, 18,  0,  0, 19,  0, 20, 21,  0,  0,  0,
+    0,  0, 22, 23,  0, 24, 25,  0,  0, 26,  0,  0,  0, 27, 28, 29,
+    0,  0,  0, 30, 31, 32,  0,  0, 31,  0,  0, 33, 31,  0,  0,  0,
+   31, 34,  0,  0,  0,  0,  0, 35, 36,  0,  0,  0,  0,  0,  0, 37,
+   38,  0,  0,  0,  0,  0,  0, 39, 40,  0,  0,  0,  0, 41,  0, 42,
+    0,  0,  0, 43, 44,  0,  0,  0, 45,  0,  0,  0,  0,  0,  0, 46,
+   47,  0,  0,  0,  0, 48,  0,  0,  0, 49,  0, 49,  0, 50,  0,  0,
+    0,  0, 51,  0,  0,  0,  0, 52,  0, 53,  0,  0,  0,  0, 54, 55,
+    0,  0,  0, 56, 57,  0,  0,  0,  0,  0,  0, 58, 49,  0, 59, 60,
+    0,  0, 61,  0,  0,  0, 62, 63,  0,  0,  0, 64,  0, 65, 66, 67,
+   68, 69,  1, 70,  0, 71, 72, 73,  0,  0, 74, 75,  0,  0,  0, 76,
+    0,  0,  1,  1,  0,  0, 77,  0,  0, 78,  0,  0,  0,  0, 74, 79,
+    0, 80,  0,  0,  0,  0,  0, 75, 81,  0, 82,  0, 49,  0,  1, 75,
+    0,  0, 83,  0,  0, 84,  0,  0,  0,  0,  0, 85, 54,  0,  0,  0,
+    0,  0,  0, 86, 87,  0,  0, 81,  0,  0, 31,  0,  0, 88,  0,  0,
+    0,  0, 89,  0,  0,  0,  0, 47,  0,  0, 57,  0,  0,  0,  0, 90,
+   91,  0,  0, 92,  0,  0, 93,  0,  0,  0, 94,  0,  0,  0, 95,  0,
+   96, 57,  0,  0, 81,  0,  0, 76,  0,  0,  0, 97, 98,  0,  0, 99,
+  100,  0,  0,  0,  0,  0,  0,101,  0,  0,102,  0,  0,  0,  0,103,
+   31,  0,104,105,106, 33,  0,  0,107,  0,  0,  0,108,  0,  0,  0,
+    0,  0,  0,109,  0,  0,110,  0,  0,  0,  0,111, 85,  0,  0,  0,
+    0,  0, 54,  0,  0,  0,  0, 49,112,  0,  0,  0,  0,113,  0,  0,
+  114,  0,  0,  0,  0,112,  0,  0,  0,  0,  0,115,  0,  0,  0,116,
+    0,  0,  0,117,  0,118,  0,  0,  0,  0,119,120,121,  0,122,  0,
+  123,  0,  0,  0,124,125,126,  0,  0,  0,127,  0,  0,128,  0,  0,
+  129,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,  1,  1,  2,  3,  4,
+    5,  6,  7,  4,  4,  8,  9, 10,  1, 11, 12, 13, 14, 15, 16, 17,
+   18,  1,  1,  1, 19,  1,  0,  0, 20, 21, 22,  1, 23,  4, 21, 24,
+   25, 26, 27, 28, 29, 30,  0,  0,  1,  1, 31,  0,  0,  0, 32, 33,
+   34, 35,  1, 36, 37,  0,  0,  0,  0, 38,  1, 39, 14, 39, 40, 41,
+   42,  0,  0,  0, 43, 36, 44, 45, 21, 45, 46,  0,  0,  0, 19,  1,
+   21,  0,  0, 47,  0, 38, 48,  1,  1, 49, 49, 50,  0,  0, 51,  0,
+   52,  1,  1,  1, 53, 21, 43, 54, 55, 21, 35,  1,  0,  0,  0, 56,
+    0,  0,  0, 57, 58, 59,  0,  0,  0,  0,  0, 60,  0, 61,  0,  0,
+    0,  0, 62, 63,  0,  0, 64,  0,  0,  0, 65,  0,  0,  0, 66,  0,
+    0,  0, 67,  0,  0,  0, 68,  0,  0,  0, 69,  0,  0, 70, 71,  0,
+   72, 73, 74, 75, 76, 77,  0,  0,  0, 78,  0,  0,  0, 79, 80,  0,
+    0,  0,  0, 47,  0,  0,  0, 49,  0, 63,  0,  0, 64,  0,  0, 81,
+    0,  0, 82,  0,  0,  0, 83,  0,  0, 19, 84,  0, 63,  0,  0,  0,
+    0, 49,  1, 85,  1, 54, 15, 86, 84,  0,  0,  0,  0, 56,  0,  0,
+    0,  0, 19, 10,  1,  0,  0,  0,  0,  0, 87,  0,  0, 88,  0,  0,
+   87,  0,  0,  0,  0, 79,  0,  0, 89,  9, 12,  4, 90,  8, 91, 47,
+    0, 59, 50,  0, 21,  1, 21, 92, 93,  1,  1,  1,  1, 94, 95, 96,
+   97,  1, 98, 59, 81, 99,100,  4, 59,  0,  0,  0,  0,  0,  0, 19,
+   50,  0,  0,  0,  0,  0,  0, 62,  0,  0,101,102,  0,  0,103,  0,
+    0,  1,  1, 50,  0,  0,  0, 38,  0, 64,  0,  0,  0,  0,  0, 63,
+    0,  0, 52, 69, 62,  0,  0,  0, 79,  0,  0,  0,104,105, 59, 38,
+   81,  0,  0,  0,  0,  0,  0,106,  1, 14,  4, 12,  0, 38, 89,  0,
+    0,  0,  0,107,  0,  0,108, 62,  0,109,  0,  0,  0,  1,  0,  0,
+    0,  0, 19, 59,  0,110, 14, 54,  0,  0,111,  0, 89,  0,  0,  0,
+   62, 63,  0,  0, 63,  0, 88,  0,  0,111,  0,  0,  0,  0,112,  0,
+    0,  0, 79, 56,  0, 38,  1, 59,  1, 59,  0,  0, 64, 88,  0,  0,
+  113,  0,  0,  0, 56,  0,  0,  0,  0,113,  0,  0,  0,  0, 62,  0,
+    0,  0,  0, 80,  0, 62,  0,  0,  0,  0, 57,  0, 88,114,  0,  0,
+    8, 91,  0,  0,  1, 89,  0,  0,115,  0,  0,  0,  0,  0,  0,116,
+    0,117,118,119,120,  0, 52,  4,121, 49, 23,  0,  0,  0, 38, 50,
+   38, 59,  0,  0,  1, 89,  1,  1,  1,  1, 39,  1, 48,104, 89,  0,
+    0,  0,  0,  1,  4,121,  0,  0,  0,  1,122,  0,  0,  0,  0,  0,
+  230,230,230,230,230,232,220,220,220,220,232,216,220,220,220,220,
+  220,202,202,220,220,220,220,202,202,220,220,220,  1,  1,  1,  1,
+    1,220,220,220,220,230,230,230,230,240,230,220,220,220,230,230,
+  230,220,220,  0,230,230,230,220,220,220,220,230,232,220,220,230,
+  233,234,234,233,234,234,233,230,  0,  0,  0,230,  0,220,230,230,
+  230,230,220,230,230,230,222,220,230,230,220,220,230,222,228,230,
+   10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 19, 20, 21, 22,  0, 23,
+    0, 24, 25,  0,230,220,  0, 18, 30, 31, 32,  0,  0,  0,  0, 27,
+   28, 29, 30, 31, 32, 33, 34,230,230,220,220,230,220,230,230,220,
+   35,  0,  0,  0,  0,  0,230,230,230,  0,  0,230,230,  0,220,230,
+  230,220,  0,  0,  0, 36,  0,  0,230,220,230,230,220,220,230,220,
+  220,230,220,230,220,230,230,  0,  0,220,  0,  0,230,230,  0,230,
+    0,230,230,230,230,230,  0,  0,  0,220,220,220,  0,  0,  0,220,
+  230,230,  0,220,230,220,220,220, 27, 28, 29,230,  7,  0,  0,  0,
+    0,  9,  0,  0,  0,230,220,230,230,  0,  0,  0,  0,  0,230,  0,
+    0, 84, 91,  0,  0,  0,  0,  9,  9,  0,  0,  0,  0,  0,  9,  0,
+  103,103,  9,  0,107,107,107,107,118,118,  9,  0,122,122,122,122,
+  220,220,  0,  0,  0,220,  0,220,  0,216,  0,  0,  0,129,130,  0,
+  132,  0,  0,  0,  0,  0,130,130,130,130,  0,  0,130,  0,230,230,
+    9,  0,230,230,  0,  0,220,  0,  0,  0,  0,  7,  0,  9,  9,  0,
+    0,230,  0,  0,  0,228,  0,  0,  0,222,230,220,220,  0,  0,  0,
+  230,  0,  0,220,230,220,  0,220,  0,  0,  9,  9,  0,  0,  7,  0,
+  230,230,230,  0,230,  0,  1,  1,  1,  0,  0,  0,230,234,214,220,
+  202,230,230,230,230,230,232,228,228,220,  0,230,233,220,230,220,
+  230,230,  1,  1,  1,  1,  1,230,  0,  1,  1,230,220,230,  1,  1,
+    0,  0,218,228,232,222,224,224,  0,  8,  8,  0,230,  0,230,230,
+  220,  0,  0,230,  0,  0, 26,  0,  0,220,  0,230,230,  1,220,  0,
+    0,230,220,  0,  0,  0,220,220,  0,  9,  7,  0,  0,  7,  9,  0,
+    0,  0,  9,  7,  9,  9,  0,  0,  6,  6,  0,  0,  0,  0,  1,  0,
+    0,216,216,  1,  1,  1,  0,  0,  0,226,216,216,216,216,216,  0,
+  220,220,220,  0,230,230,  7,  0, 16, 17, 17, 33, 17, 49, 17, 17,
+   84, 97,135,145, 26, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17, 17, 17,177,  0,  1,  2,  3,  3,  3,  3,  3,
+    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  4,  3,  3,  3,
+    3,  3,  5,  3,  3,  3,  3,  3,  6,  7,  8,  3,  3,  3,  3,  3,
+    9, 10, 11, 12, 13,  3,  3,  3,  3,  3,  3,  3,  3, 14,  3, 15,
+    3,  3,  3,  3,  3,  3, 16, 17, 18, 19, 20, 21,  3,  3,  3, 22,
+   23,  3,  3,  3,  3,  3,  3,  3, 24,  3,  3,  3,  3,  3,  3,  3,
+    3, 25,  3,  3, 26, 27,  0,  1,  0,  0,  0,  0,  0,  1,  0,  2,
+    0,  0,  0,  3,  0,  0,  0,  3,  0,  0,  0,  0,  0,  4,  0,  5,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  6,
+    0,  0,  0,  7,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  8,
+    9,  0,  0,  0,  0,  0,  0,  9,  0,  9,  0,  0,  0,  0,  0,  0,
+    0, 10, 11, 12, 13,  0,  0, 14, 15, 16,  6,  0, 17, 18, 19, 19,
+   19, 20, 21, 22, 23, 24, 19, 25,  0, 26, 27, 19, 19, 28, 29, 30,
+    0, 31,  0,  0,  0,  8,  0,  0,  0,  0,  0,  0,  0, 19, 28,  0,
+   32, 33,  9, 34, 35, 19,  0,  0, 36, 37, 38, 39, 40, 19,  0, 41,
+   42, 43, 44, 31,  0,  1, 45, 42,  0,  0,  0,  0,  0, 32, 14, 14,
+    0,  0,  0,  0, 14,  0,  0, 46, 47, 47, 47, 47, 48, 49, 47, 47,
+   47, 47, 50, 51, 52, 53, 43, 21,  0,  0,  0,  0,  0,  0,  0, 54,
+    6, 55,  0, 14, 19,  1,  0,  0,  0, 19, 56, 31,  0,  0,  0,  0,
+    0,  0,  0, 57, 14,  0,  0,  0,  0,  1,  0,  2,  0,  0,  0,  3,
+    0,  0,  0, 58, 59,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0,  0,
+    0,  0,  2,  3,  0,  4,  5,  0,  0,  6,  0,  0,  0,  7,  0,  0,
+    0,  1,  1,  0,  0,  8,  9,  0,  8,  9,  0,  0,  0,  0,  8,  9,
+   10, 11, 12,  0,  0,  0, 13,  0,  0,  0,  0, 14, 15, 16, 17,  0,
+    0,  0,  1,  0,  0, 18, 19,  0,  0,  0, 20,  0,  0,  0,  1,  1,
+    1,  1,  0,  1,  1,  1,  1,  1,  1,  1,  0,  8, 21,  9,  0,  0,
+   22,  0,  0,  0,  0,  1,  0, 23, 24, 25,  0,  0, 26,  0,  0,  0,
+    8, 21, 27,  0,  1,  0,  0,  1,  1,  1,  1,  0,  1, 28, 29, 30,
+    0, 31, 32, 20,  1,  1,  0,  0,  0,  8, 21,  9,  1,  4,  5,  0,
+    0,  0, 33,  9,  0,  1,  1,  1,  0,  8, 21, 21, 21, 21, 34,  1,
+   35, 21, 21, 21,  9, 36,  0,  0, 37, 38,  1,  0, 39,  0,  0,  0,
+    1,  0,  1,  0,  0,  0,  0,  8, 21,  9,  1,  0,  0,  0, 40,  0,
+    8, 21, 21, 21, 21, 21, 21, 21, 21,  9,  0,  1,  1,  1,  1,  8,
+   21, 21, 21,  9,  0,  0,  0, 41,  0, 42, 43,  0,  0,  0,  1, 44,
+    0,  0,  0, 45,  8,  9,  1,  0,  1,  0,  1,  1,  8, 21, 21,  9,
+    0,  4,  5,  8,  9,  1,  0,  0,  0,  1,  2,  3,  4,  5,  6,  7,
+    7,  8,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  9, 10, 11, 11,
+   11, 11, 11, 12, 12, 12, 12, 13, 14, 15, 16, 17, 18, 12, 19, 12,
+   20, 12, 12, 12, 12, 21, 22, 22, 22, 23, 12, 12, 12, 12, 24, 25,
+   12, 12, 26, 27, 28, 29, 30, 31,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,
+    7,  7,  7,  7,  7, 32, 12, 33,  7,  7, 34, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12,
+   12, 12, 12, 12, 12, 12, 12, 12, 35,  0,  0,  1,  2,  2,  2,  3,
+    4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
+   20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 32, 33, 33,
+   33, 34, 35, 35, 35, 35, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44,
+   45, 46, 47, 48, 49, 50,  2,  2, 51, 51, 52, 53, 54, 55, 56, 56,
+   56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 57, 57, 56, 56, 56, 56,
+   56, 56, 58, 59, 60, 61, 56, 62, 62, 63, 64, 65, 66, 67, 68, 69,
+   70, 56, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62, 62,
+   62, 62, 62, 62, 62, 71, 62, 62, 62, 62, 72, 72, 72, 72, 72, 72,
+   72, 72, 72, 73, 74, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84,
+   85, 86, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+   32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 87, 87, 87, 87, 87, 87,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 62, 62, 62, 62,
+   88, 89, 89, 89, 90, 89, 91, 92, 93, 94, 95, 95, 96, 97, 87, 98,
+   99,100,101,102,103, 87,104,104,104, 87,105,106,107,108,109,110,
+  111,112,113,114,115, 87, 89,116,117,118,119,120,121,122,123,124,
+  125, 87,126,127, 87,128,129,130,131, 87,132,133,134,135,136,137,
+   87, 87,138,139,140,141, 87,142, 87,143,144,144,144,144,144,144,
+  144,144,144,144,144, 87, 87, 87, 87, 87,145,145,145,145,145,145,
+  145,145,145, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+   87, 87,146,146,146,146,146, 87, 87, 87,147,147,147,147,148,149,
+  150,150, 87, 87, 87, 87,151,151,152,153,154,154,154,154,154,154,
+  154,154,154,154,154,154,154,154,154,154,154,154,154,154,154,154,
+  155,155,155,155,154, 87, 87, 87, 87, 87,156,157,158,159,159,159,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 87,
+   87, 87,160,161, 87, 87, 87, 87, 87, 87, 56, 56,162,163, 51, 56,
+   56, 87, 56, 56, 56, 56, 56, 56, 56, 56,164,164,164,164,164,164,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87,165, 87,166, 87, 87,167,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87,168,168,169, 87, 87, 87,
+   87, 87, 56, 56, 56, 87, 89, 89, 87, 87, 56, 56, 56, 56,170, 87,
+   56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+   56, 56, 87, 87, 87, 87, 87, 87, 87, 87, 62, 62, 62, 62, 62, 62,
+   62, 62, 87, 87, 87, 87, 87, 87, 87, 87, 62, 62, 62, 62, 62, 87,
+   87, 87, 87, 87, 87, 87, 87, 87, 87, 87, 62, 62, 62, 62, 62, 62,
+   62, 87, 87, 87, 87, 87, 87, 87, 87, 87, 56, 87,171,171,  0,  1,
+    2,  2,  0,  1,  2,  2,  2,  3,  4,  5,  0,  0,  0,  0,  1,  2,
+    1,  2,  0,  0,  3,  3,  4,  5,  4,  5,  4,  4,  4,  4,  4,  4,
+    4,  4,  4,  4,  4,  6,  0,  0,  7,  0,  8,  8,  8,  8,  8,  8,
+    8,  9, 10, 11, 11, 11, 11, 11, 12, 11, 13, 13, 13, 13, 13, 13,
+   13, 13, 14, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 15, 16, 16,
+   16, 16, 16, 17, 18, 18, 18, 18, 18, 18, 19, 20, 21, 21, 22, 23,
+   21, 24, 21, 21, 21, 21, 21, 25, 21, 21, 26, 26, 26, 26, 26, 21,
+   21, 21, 27, 27, 27, 27, 28, 28, 28, 28, 29, 29, 29, 29, 30, 30,
+   26, 26, 21, 21, 21, 21, 21, 21, 31, 21, 32, 32, 32, 32, 32, 33,
+   34, 32, 35, 35, 35, 35, 35, 35, 35, 35, 36, 36, 36, 36, 36, 36,
+   36, 36, 37, 37, 37, 37, 37, 37, 37, 37, 38, 38, 38, 38, 38, 38,
+   38, 38, 39, 39, 39, 39, 39, 39, 39, 39, 40, 40, 40, 40, 40, 40,
+   40, 40, 41, 41, 41, 41, 41, 41, 41, 41, 42, 42, 42, 42, 42, 42,
+   42, 42, 43, 43, 43, 43, 43, 43, 43, 43, 44, 44, 44, 45, 44, 44,
+   44, 44, 46, 46, 46, 46, 46, 46, 46, 46, 47, 47, 47, 47, 47, 47,
+   47, 47, 47, 47, 47, 47, 47, 48, 47, 47, 49, 49, 49, 49, 49, 49,
+   49, 49, 49, 49, 50, 50, 50, 50, 50, 51, 52, 52, 52, 52, 52, 52,
+   52, 52, 53, 53, 53, 53, 53, 53, 53, 53, 53, 53, 54, 54, 54, 54,
+   54, 54, 55, 55, 55, 55, 55, 55, 55, 55, 56, 56, 57, 57, 57, 57,
+   58, 57, 59, 59, 60, 61, 62, 62, 63, 63, 64, 64, 64, 64, 64, 64,
+   64, 64, 65, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 55, 55, 55,
+   55, 55, 67, 67, 67, 67, 67, 68, 68, 68, 69, 69, 69, 69, 69, 69,
+   64, 64, 70, 70, 71, 71, 71, 71, 71, 71, 71, 71, 71,  8,  8,  8,
+    8,  8, 72, 72, 72, 72, 72, 72, 72, 72, 73, 73, 73, 73, 74, 74,
+   74, 74, 75, 75, 75, 75, 75, 76, 76, 76, 13, 50, 50, 50, 73, 77,
+   78, 79,  4,  4, 80,  4,  4, 81, 82, 83,  4,  4,  4, 84,  8,  8,
+    8,  8, 11, 11, 11, 11, 11, 11, 11, 11, 85,  0,  0,  0,  0,  0,
+    0, 86,  0,  4,  0,  0,  0,  8,  8,  8,  0,  0, 87, 88, 89,  0,
+    4,  4,  6,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 90, 90, 90, 90, 90, 90, 90, 90, 91, 91, 91, 91, 91, 91,
+    4,  4, 92, 92, 92, 92, 92, 92, 92, 92, 50, 50, 50, 93, 93, 93,
+   93, 93, 53, 53, 53, 53, 53, 53, 13, 13, 94, 94, 94, 94, 94, 94,
+   94, 94, 94, 94, 94, 94, 94, 94, 94,  0, 95,  0, 96, 97, 98, 99,
+   99, 99, 99,100,101,102,102,102,102,103,104,104,104,105, 52, 52,
+   52, 52, 52,  0,104,104,  0,  0,  0,102, 52, 52,  0,  0,  0,  0,
+   52,106,  0,  0,  0,  0,  0,102,102,107,102,102,102,102,102,108,
+    0,  0, 94, 94, 94, 94,  0,  0,  0,  0,109,109,109,109,109,109,
+  109,109,109,109,109,109,109,110,110,110,111,111,111,111,111,111,
+  111,111,111,111,111,111, 13, 13, 13, 13, 13, 13,112,112,112,112,
+  112,112,  0,  0,113,  4,  4,  4,  4,  4,114,  4,  4,  4,  4,  4,
+    4,  4,115,115,115,  0,116,116,116,116,117,117,117,117,117,117,
+   32, 32,118,118,119,120,120,120, 52, 52,121,121,121,121,122,121,
+   49, 49,123,123,123,123,123,123, 49, 49,124,124,124,124,124,124,
+  125,125, 53, 53, 53,  4,  4,126,127, 54, 54, 54, 54, 54,125,125,
+  125,125,128,128,128,128,128,128,128,128,  4,129, 18, 18, 18, 21,
+   21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21, 21,130,  0, 21,
+   21, 21,  8,  0,131,  0,  0,  0,  0, 21, 21, 21, 21, 21, 21, 21,
+   21,132,  0,  0,  1,  2,  1,  2,133,101,102,134, 52, 52, 52, 52,
+    0,  0,135,135,135,135,135,135,135,135,  0,  0,  0,  0, 11, 11,
+   11, 11, 11,  0, 11, 11, 11,  0,  0,136,137,137,138,138,138,138,
+  139,  0,140,140,140,141,141,142,142,142,143,143,144,144,144,144,
+  144,144,145,145,145,145,145,146,146,146,147,147,147,148,148,148,
+  148,148,149,149,149,150,150,150,150,150,151,151,151,151,151,151,
+  151,151,152,152,152,152,153,153,154,154,155,155,155,155,155,155,
+  156,156,157,157,158,158,158,158,158,158,159,159,160,160,160,160,
+  160,160,161,161,161,161,161,161,162,162,163,163,163,163,164,164,
+  164,164,165,165,165,165,166,166,167,167,168,168,168,168,168,168,
+  168,168,169,169,169,169,169,169,169,169,170,170,170,170,170,170,
+  170,170,171,171,171,171,171,171,171,171,172,172,172,172,172,172,
+  172,172,173,173,173,174,174,174,174,174,175,175,175,175,175,175,
+  176,176,177,177,177,177,177,177,177,177,178,178,178,178,178,179,
+  179,179,180,180,180,180,180,181,181,181,182,182,182,182,182,182,
+  183, 43,184,184,184,184,184,184,184,184,185,185,185,186,186,186,
+  186,186,187,187,187,188,187,187,187,187,189,189,189,189,189,189,
+  189,189,190,190,190,190,190,190,190,190,191,191,191,191,191,191,
+  191,191,192,192,192,192,192,192, 66, 66,193,193,193,193,193,193,
+  193,193,194,194,194,194,194,194,194,194,195,195,195,195,195,195,
+  195,195,196,196,196,196,196,196,196,196,197,197,197,197,197,197,
+  197,197,198,198,198,198,198,198,198,198,199,199,199,199,199,200,
+  200,200,200,200,200,200,201,201,201,201,202,202,202,202,202,202,
+  202,203,203,203,203,203,203,203,203,203,204,204,204,204,204,204,
+  205,205,205,205,205,205,205,205,205,205,206,206,206,206,206,206,
+  206,206,110,110,110,110, 39, 39, 39, 39,207,207,207,207,207,207,
+  207,207,208,208,208,208,208,208,208,208,209,209,209,209,209,209,
+  209,209,112,112,112,112,112,112,112,112,112,112,112,112,210,210,
+  210,210,211,211,211,211,211,211,211,211,212,212,212,212,212,212,
+  212,212,213,213,213,213,213,213,213,213,214,214,214,214,214,214,
+  214,214,214,214,214,214,214,214,215, 94,216,216,216,216,216,216,
+  216,216,217,217,217,217,217,217,217,217,218, 99, 99, 99, 99, 99,
+   99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
+  219,220,220,220,220,220,220,220,220,220,221,221,221,221,221,221,
+  221,221,221,221,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+  222,223,224,  0,225,  0,  0,  0,  0,  0,226,226,226,226,226,226,
+  226,226, 91, 91, 91, 91, 91, 91, 91, 91,227,227,227,227,227,227,
+  227,227,228,228,228,228,228,228,228,228,229,229,229,229,229,229,
+  229,229,230,230,230,230,230,230,230,230,231,  0,  0,  0,  0,  0,
+    0,  0,  8,  8,  8,  8,  8,  8,  8,  8,  0,  0,  0,  0,  1,  2,
+    2,  2,  2,  2,  3,  0,  0,  0,  4,  0,  2,  2,  2,  2,  2,  3,
+    2,  2,  2,  2,  5,  0,  2,  5,  6,  0,  7,  7,  7,  7,  8,  9,
+    8, 10,  8, 11,  8,  8,  8,  8,  8,  8, 12, 13, 13, 13, 14, 14,
+   14, 14, 14, 15, 14, 14, 16, 17, 17, 17, 17, 17, 17, 17, 18, 19,
+   19, 19, 19, 19, 19, 19, 20, 21, 20, 22, 20, 20, 23, 23, 20, 20,
+   20, 20, 22, 20, 24,  7,  7, 25, 20, 20, 26, 20, 20, 20, 20, 20,
+   20, 21, 27, 27, 27, 27, 28, 28, 28, 28, 29, 29, 29, 29, 30, 30,
+   30, 30, 31, 31, 31, 31, 32, 20, 20, 20, 33, 33, 33, 33, 34, 35,
+   33, 33, 33, 36, 33, 33, 37, 37, 37, 37, 38, 38, 38, 38, 39, 39,
+   39, 39, 40, 40, 40, 40, 41, 41, 41, 41, 42, 42, 42, 42, 43, 43,
+   43, 43, 44, 44, 44, 44, 45, 45, 45, 45, 46, 46, 46, 46, 46, 46,
+   46, 47, 48, 48, 48, 48, 49, 49, 49, 49, 49, 50, 51, 49, 52, 52,
+   52, 52, 53, 53, 53, 53, 53, 53, 54, 53, 55, 55, 55, 55, 56, 56,
+   56, 56, 57, 57, 57, 57, 58, 58, 58, 58, 59, 59, 59, 59, 60, 60,
+   60, 60, 60, 60, 61, 62, 63, 63, 63, 63, 64, 64, 64, 64, 64, 65,
+    0,  0, 66, 66, 66, 66, 67, 67, 67, 67, 68, 68, 68, 68, 69, 70,
+   71, 71, 71, 71, 71, 71, 72, 72, 72, 72, 73, 73, 73, 73, 74, 74,
+   74, 74, 75, 75, 75, 75, 76, 76, 76, 76, 77, 77, 77, 77, 78, 78,
+   78, 78, 79, 79, 79, 79, 80, 80, 80, 80, 81, 81, 81, 81, 82,  7,
+    7,  7, 83,  7, 84, 85,  0, 84, 86,  0,  2, 87, 88,  2,  2,  2,
+    2, 89, 90, 87, 91,  2,  2,  2, 92,  2,  2,  2,  2, 93,  0,  0,
+    0, 86,  1,  0,  0, 94,  0, 95, 96,  0,  4,  0,  0,  0,  0,  0,
+    0,  4, 97, 97, 97, 97, 98, 98, 98, 98, 13, 13, 13, 13, 99, 99,
+   99, 99,100,100,100,100,  0,101,  0,  0,102,100,103,104,  0,  0,
+  100,  0,105,106,106,106,106,106,106,106,106,106,107,105,108,109,
+  109,109,109,109,109,109,109,109,110,108,111,111,111,111,112, 55,
+   55, 55, 55, 55, 55,113,109,109,109,110,109,109,  0,  0,114,114,
+  114,114,115,115,115,115,116,116,116,116,117,117,117,117, 96,  2,
+    2,  2,  2,  2, 94,  2,118,118,118,118,119,119,119,119,120,120,
+  120,120,121,121,121,121,121,121,121,122,123,123,123,123,124,124,
+  124,124,124,124,124,125,126,126,126,126,127,127,127,127,128,128,
+  128,128,  2,  2,  3,  2,  2,129,130,  0,131,131,131,131,132, 17,
+   17, 18, 20, 20, 20,133,  7,  7,  7,134, 20, 20, 20, 23,  0,135,
+  109,109,109,109,109,136,137,137,137,137,  0,  0,  0,138,139,139,
+  139,139,140,140,140,140, 84,  0,  0,  0,141,141,141,141,142,142,
+  142,142,143,143,143,143,144,144,144,144,145,145,145,145,146,146,
+  146,146,147,147,147,147,148,148,148,148,149,149,149,149,150,150,
+  150,150,151,151,151,151,152,152,152,152,153,153,153,153,154,154,
+  154,154,155,155,155,155,156,156,156,156,157,157,157,157,158,158,
+  158,158,159,159,159,159,160,160,160,160,161,161,161,161,162,162,
+  162,162,163,163,163,163,164,164,164,164,165,165,165,165,166,166,
+  166,166,167,167,167,167,168,168,168,168,169,169,169,169,170,170,
+  170,170,171,171,171,171,172,172,172,172,173,173,173,173,174,174,
+  174,174,175,175,175,175,176,176,176,176,177,177,177,177,178,178,
+  178,178,179,179,179,179,180,180,180,180,181,181,181,181,182,182,
+  182,182,183,183,183,183,184, 45, 45, 45,185,185,185,185,186,186,
+  186,186,187,187,187,187,188,188,188,188,188,188,189,188,190,190,
+  190,190,191,191,191,191,192,192,192,192,193,193,193,193,194,194,
+  194,194,195,195,195,195,196,196,196,196,197,197,197,197,198,198,
+  198,198,199,199,199,199,200,200,200,200,201,201,201,201,202,202,
+  202,202,203,203,203,203,204,204,204,204,205,205,205,205,206,206,
+  206,206,207,207,207,207,208,208,208,208,209,209,209,209,210,210,
+  210,210,211,211,211,211,212,212,212,212,213,213,213,213,214,214,
+  214,214,215,215,215,215,216,217,217,217,218,218,218,218,217,217,
+  217,217,219,106,106,106,106,109,109,109,220,220,220,220,221,221,
+  221,221,  0,222, 86,  0,  0,  0,222,  7, 82,138,  7,  0,  0,  0,
+  223, 86,224,224,224,224,225,225,225,225,226,226,226,226,227,227,
+  227,227,228,228,228,228,229,  0,  0,  0,  0,  0,  0,  0,  0, 19,
+   19, 19, 19, 19, 19, 19, 19, 19, 19,  0,  0,  0, 19,  0, 19,  0,
+    0,  0,  0,  0, 26, 26,  1,  1,  1,  1,  9,  9,  9,  9,  0,  9,
+    9,  9,  9,  9,  0,  9,  9,  0,  9,  0,  9,  9, 55, 55, 55, 55,
+   55, 55,  6,  6,  6,  6,  6,  1,  1,  6,  6,  4,  4,  4,  4,  4,
+    4,  4,  4, 14, 14, 14, 14, 14, 14, 14,  3,  3,  3,  3,  3,  0,
+    3,  3,  0,  3,  3,  3,  3,  3,  3,  0,  3,  3,  3,  1,  1,  1,
+    3,  3,  1,  3,  3,  3, 37, 37, 37, 37, 38, 38, 38, 38, 64, 64,
+   64, 64, 90, 90, 90, 90, 95, 95, 95, 95,  3,  3,  0,  3,  7,  7,
+    7,  7,  7,  1,  1,  1,  1,  7,  7,  7,  0,  0,  7,  7,  5,  5,
+    5,  5, 11, 11, 11, 11, 10, 10, 10, 10, 21, 21, 21, 21, 22, 22,
+   22, 22, 23, 23, 23, 23, 16, 16, 16, 16, 20, 20, 20, 20, 36, 36,
+   36, 36, 24, 24, 24, 24, 24, 24, 24,  0, 18, 18, 18, 18, 25, 25,
+   25, 25, 25,  0,  0,  0,  0, 25, 25, 25, 33, 33, 33, 33,  8,  8,
+    8,  8,  8,  8,  8,  0, 12, 12, 12, 12, 30, 30, 30, 30, 29, 29,
+   29, 29, 28, 28, 28, 28, 34, 34, 34, 34, 35, 35, 35, 35, 35, 35,
+   35,  0,  0,  0, 35, 35, 45, 45, 45, 45, 44, 44, 44, 44, 44,  0,
+    0,  0, 43, 43, 43, 43, 46, 46, 46, 46, 31, 31, 31, 31, 32, 32,
+    0,  0, 32,  0, 32, 32, 32, 32, 32, 32, 48, 48, 48, 48, 52, 52,
+   52, 52, 58, 58, 58, 58, 54, 54, 54, 54, 91, 91, 91, 91, 62, 62,
+   62, 62, 76, 76, 76, 76, 93, 93, 93, 93, 70, 70, 70, 70, 73, 73,
+   73, 73,  1,  1,  1,  0,  1,  0,  1,  1,  1,  0,  0,  0,  0,  1,
+    0,  0,  1,  1,  0,  0, 19, 19,  9,  9,  9,  9,  9,  6, 19,  9,
+    9,  9,  9,  9, 19, 19,  9,  9,  9, 19,  6, 19, 19, 19, 19, 19,
+   19,  9,  0,  0,  0, 19,  0,  0,  9,  0,  0,  0, 19, 19, 27, 27,
+   27, 27, 56, 56, 56, 56, 61, 61, 61, 61, 13, 13, 13, 13,  0, 13,
+    0, 13,  0, 13, 13, 13, 13, 13,  1,  1,  1,  1, 12, 12,  0, 15,
+   15, 15, 15, 15, 15, 15, 15,  1,  1,  0,  0, 17, 17, 17, 17, 17,
+   17, 17, 17, 17, 17,  0, 26, 26, 26, 26, 26, 12, 12, 12, 12, 12,
+   12,  0, 39, 39, 39, 39, 86, 86, 86, 86, 77, 77, 77, 77, 79, 79,
+   79, 79, 60, 60, 60, 60, 65, 65, 65, 65, 75, 75, 75, 75, 69, 69,
+   69, 69, 69, 69,  0, 69, 74, 74, 74, 74, 84, 84, 84, 84, 84, 84,
+   84,  0, 68, 68, 68, 68, 92, 92, 92, 92, 87, 87, 87, 87, 19,  9,
+   19, 19, 19, 19,  0,  0,  2,  2,  2,  2, 19, 19, 19,  4,  3,  3,
+    0,  0,  1,  1,  6,  6,  0,  0, 17, 17, 17, 17,  0,  0, 49, 49,
+   49, 49,  0,  1,  1,  1, 71, 71, 71, 71, 67, 67, 67, 67, 42, 42,
+   42, 42, 41, 41, 41, 41,118,118,118,118, 53, 53, 53, 53, 59, 59,
+   59, 59, 40, 40, 40, 40, 51, 51, 51, 51, 50, 50, 50, 50,135,135,
+  135,135,106,106,106,106,104,104,104,104,110,110,110,110, 47, 47,
+   47, 47, 81, 81, 81, 81,120,120,120,120,116,116,116,116,128,128,
+  128,128, 66, 66, 66, 66, 72, 72, 72, 72, 98, 98, 98, 98, 97, 97,
+   97, 97, 57, 57, 57, 57, 88, 88, 88, 88,117,117,117,117,112,112,
+  112,112, 78, 78, 78, 78, 83, 83, 83, 83, 82, 82, 82, 82,122,122,
+  122,122, 89, 89, 89, 89,130,130,130,130,144,144,144,144,156,156,
+  156,156,147,147,147,147,148,148,148,148,153,153,153,153,149,149,
+  149,149, 94, 94, 94, 94, 85, 85, 85, 85,101,101,101,101, 96, 96,
+   96, 96,111,111,111,111,100,100,100,100,100, 36, 36, 36,108,108,
+  108,108,129,129,129,129,109,109,109,109,107,107,107,107,107,107,
+  107,  1,137,137,137,137,124,124,124,124,123,123,123,123,114,114,
+  114,114,102,102,102,102,126,126,126,126,142,142,142,142,125,125,
+  125,125,154,154,154,154,150,150,150,150,141,141,141,141,140,140,
+  140,140,121,121,121,121,133,133,133,133,134,134,134,134,138,138,
+  138,138,143,143,143,143,145,145,145,145, 63, 63, 63, 63, 80, 80,
+   80, 80,127,127,127,127,115,115,115,115,103,103,103,103,119,119,
+  119,119,146,146,146,146, 99, 99, 99, 99,136,139,  0,  0,155,155,
+  155,155,136,136,136,136, 17, 15, 15, 15,139,139,139,139,105,105,
+  105,105,  0,  0,  0,  1,  0,  0,  1,  1,131,131,131,131,151,151,
+  151,151,152,152,152,152,113,113,113,113,132,132,132,132, 15,  0,
+    0,  0,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  9,  9,  9, 10,
+    9, 11, 12, 13,  9,  9,  9, 14,  9,  9, 15,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 16, 17,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 18, 19, 20,  9, 21,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 22,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,
+    9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9,  9, 23, 24,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,  4,
+    5,  6,  7,  8,  9, 10, 11, 12,  0,  0, 13, 14, 15, 16, 17, 18,
+   19, 20, 21, 22,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 23,  0,  0, 24, 25, 26, 27, 28, 29, 30,  0,  0,
+   31, 32,  0, 33,  0, 34,  0, 35,  0,  0,  0,  0, 36, 37, 38, 39,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 40,  0,  0,  0,  0,  0,  0,  0,  0,  0, 41, 42,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0, 43, 44,  0, 45,  0,  0,  0,  0,  0,  0, 46, 47,  0,  0,
+    0,  0,  0, 48,  0, 49,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 50, 51,  0,  0,  0, 52,  0,  0, 53,  0,  0,  0,
+    0,  0,  0,  0, 54,  0,  0,  0,  0,  0,  0,  0, 55,  0,  0,  0,
+    0,  0,  0,  0, 56,  0,  0,  0,  0,  0,  0,  0,  0, 57,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0, 58, 59, 60, 61, 62, 63, 64, 65,  0,  0,  0,  0,
+    0,  0, 66,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   67, 68,  0, 69, 70,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+   71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86,
+   87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99,100,101,102,
+  103,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,104,  0,  0,  0,  0,  0,  0,105,106,  0,107,  0,  0,  0,
+  108,  0,109,  0,110,  0,111,112,113,  0,114,  0,  0,  0,115,  0,
+    0,  0,116,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,117,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,118,119,120,121,  0,122,123,124,125,126,  0,127,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+  128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,
+  144,145,146,147,148,149,150,151,152,153,154,155,156,157,  0,  0,
+    0,158,159,160,161,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,162,163,  0,  0,  0,  0,  0,
+    0,  0,164,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,165,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,166,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,167,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,168,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,169,170,  0,  0,  0,  0,171,172,  0,  0,  0,
+  173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,
+  189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,
+  205,206,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  2,  3,  4,
+};
+static const uint16_t
+_hb_ucd_u16[4848] =
+{
+     0,   0,   1,   2,   3,   4,   5,   6,   0,   0,   7,   8,   9,  10,  11,  12,
+    13,  13,  13,  14,  15,  13,  13,  16,  17,  18,  19,  20,  21,  22,  13,  23,
+    13,  13,  13,  24,  25,  11,  11,  11,  11,  26,  11,  27,  28,  29,  30,  31,
+    32,  32,  32,  32,  32,  32,  32,  33,  34,  35,  36,  11,  37,  38,  13,  39,
+     9,   9,   9,  11,  11,  11,  13,  13,  40,  13,  13,  13,  41,  13,  13,  13,
+    13,  13,  13,  35,   9,  42,  11,  11,  43,  44,  32,  45,  46,  47,  47,  48,
+    49,  50,  47,  47,  51,  32,  52,  53,  47,  47,  47,  47,  47,  54,  55,  56,
+    57,  58,  47,  32,  59,  47,  47,  47,  47,  47,  60,  53,  61,  47,  62,  63,
+    47,  64,  65,  66,  47,  67,  47,  47,  47,  47,  47,  47,  47,  68,  69,  32,
+    70,  47,  47,  71,  72,  73,  74,  75,  76,  47,  47,  77,  78,  79,  80,  81,
+    82,  47,  47,  83,  84,  85,  86,  87,  82,  47,  47,  77,  88,  47,  80,  89,
+    90,  47,  47,  91,  92,  93,  80,  94,  95,  47,  47,  96,  97,  98,  99, 100,
+   101,  47,  47, 102, 103, 104,  80, 105, 106,  47,  47,  91, 107, 108,  80, 109,
+   110,  47,  47, 111, 112, 113,  80, 114,  90,  47,  47,  47, 115, 116,  99, 117,
+    47,  47,  47, 118, 119, 120,  66,  66,  47,  47,  47, 121, 122, 123,  47,  47,
+   124, 125, 126, 127,  47,  47,  47, 128, 129,  32,  32, 130, 131, 132,  66,  66,
+    47,  47, 133, 134, 120, 135, 136, 137, 138, 139,   9,   9,   9,  11,  11, 140,
+    47,  47,  47,  47,  47,  47,  47,  47,  47,  47,  47,  47,  47, 141, 142, 143,
+    47, 144,   9,   9,   9,   9,   9, 145, 146,  47,  47,  47,  47,  47,  47,  47,
+    47,  47,  47,  47,  47,  47, 147,  47, 148, 149,  47,  47,  47,  47, 150, 151,
+    47, 152,  47, 153,  47, 152,  47, 152,  47,  47,  47, 154, 155, 156, 157, 143,
+   158, 157,  47,  47, 159,  47,  47,  47, 160,  47, 161,  47,  47,  47,  47,  47,
+    47,  47, 162, 163, 164,  47,  47,  47,  47,  47,  47,  47,  47, 165, 144, 144,
+    47, 166,  47,  47,  47, 167, 168, 169, 157, 157, 170, 171,  32,  32,  32,  32,
+   172,  47,  47, 173, 174, 120, 175, 176, 177,  47, 178,  61,  47,  47, 179, 180,
+    47,  47, 181, 182, 183,  61,  47, 184,  11,   9,   9,   9,  66, 185, 186, 187,
+    11,  11, 188,  27,  27,  27, 189, 190,  11, 191,  27,  27,  32,  32,  32,  32,
+    13,  13,  13,  13,  13,  13,  13,  13,  13, 192,  13,  13,  13,  13,  13,  13,
+   193, 193, 193, 193, 193, 194, 193,  11, 195, 195, 195, 196, 197, 198, 198, 197,
+   199, 200, 201, 202, 203, 204, 205, 206, 207,  27, 208, 208, 208, 209, 210,  32,
+   211, 212, 213, 214, 215, 143, 216, 216, 217, 218, 219, 144, 220, 221, 144, 222,
+   223, 223, 223, 223, 223, 223, 223, 223, 224, 144, 225, 144, 144, 144, 144, 226,
+   144, 227, 223, 228, 144, 229, 230, 144, 144, 144, 144, 144, 144, 144, 143, 143,
+   143, 231, 144, 144, 144, 144, 232, 143, 144, 144, 144, 144, 144, 144, 144, 144,
+   144, 144, 144, 233, 234, 144, 144, 235, 144, 144, 144, 144, 144, 144, 236, 144,
+   144, 144, 144, 144, 144, 144, 237, 238, 143, 239, 144, 144, 240, 223, 241, 223,
+   242, 243, 223, 223, 223, 244, 223, 245, 144, 144, 144, 223, 246, 144, 144, 144,
+     9,   9,   9,  11,  11,  11, 247, 248,  13,  13,  13,  13,  13,  13, 249, 250,
+    11,  11,  11,  47,  47,  47, 251, 252,  47,  47,  47,  47,  47,  47,  32,  32,
+   253, 254, 255, 256, 257, 258,  66,  66, 259, 260, 261, 262, 263,  47,  47,  47,
+    47, 264, 146,  47,  47,  47,  47, 265,  47, 266,  47,  47, 144, 144, 144,  47,
+   144, 144, 267, 144, 268, 269, 144, 144, 267, 144, 144, 269, 144, 144, 144, 144,
+    47,  47,  47,  47, 144, 144, 144, 144,  47, 270,  47,  47,  47,  47,  47,  47,
+    47, 144, 144, 144, 144,  47,  47, 184, 271,  47,  61,  47,  13,  13, 272, 273,
+    13, 274,  47,  47,  47,  47, 275, 276,  31, 277, 278, 279,  13,  13,  13, 280,
+   281, 282, 283, 284, 285,  11,  11, 286, 287,  47, 288, 289,  47,  47,  47, 290,
+   291,  47,  47, 292, 293, 157,  32, 294,  61,  47, 295,  47, 296, 297,  47,  47,
+    70,  47,  47, 298, 299, 300, 301,  61,  47,  47, 302, 303, 304, 305,  47, 306,
+    47,  47,  47, 307,  58, 308, 309, 310,  47,  47,  47,  11,  11, 311, 312,  11,
+    11,  11,  11,  11,  47,  47, 313, 157, 314, 314, 314, 314, 314, 314, 314, 314,
+   315, 315, 315, 315, 315, 315, 315, 315,  11, 316, 317,  47,  47,  47,  47,  47,
+    47,  47,  47, 318,  31, 319,  47,  47,  47,  47,  47, 320, 321,  47,  47,  47,
+    47,  47,  47,  47,  47,  47,  47, 322,  32, 323,  32, 324, 325, 326, 327,  47,
+    47,  47,  47,  47,  47,  47,  47, 328, 329,   2,   3,   4,   5, 330, 331, 332,
+    47, 333,  47,  47,  47,  47, 334, 335, 336, 143, 143, 337, 216, 216, 216, 338,
+   339, 144, 144, 144, 144, 144, 144, 340, 341, 341, 341, 341, 341, 341, 341, 341,
+    47,  47,  47,  47,  47,  47, 342, 143,  47,  47, 343,  47, 344,  47,  47,  60,
+    47, 345,  47,  47,  47, 346, 216, 216,   9,   9, 145,  11,  11,  47,  47,  47,
+    47,  47, 157,   9,   9, 145,  11,  11,  47,  47,  47,  47,  47,  47, 345,  66,
+    47,  47,  47,  47,  47, 347,  47, 348,  47,  47, 349, 143, 143, 143,  47, 350,
+    47, 351,  47, 345,  66,  66,  66,  66,  47,  47,  47, 352, 143, 143, 143, 143,
+   353,  47,  47, 354, 143,  66,  47, 355,  47, 356, 143, 143, 357,  47, 358,  66,
+    47,  47,  47, 359,  47, 360,  47, 360,  47, 359, 142, 143, 143, 143, 143, 143,
+     9,   9,   9,   9,  11,  11,  11, 361,  47,  47, 362, 157, 157, 157, 157, 157,
+   143, 143, 143, 143, 143, 143, 143, 143,  47,  47, 363,  47,  47,  47,  47,  47,
+    47, 356, 364,  47,  60, 365,  66,  66,  47,  47,  47,  47, 366, 143,  47,  47,
+   367,  47,  47, 354, 368, 369, 370, 371, 177,  47,  47, 372, 373,  47,  47, 157,
+    95,  47, 374, 375, 376,  47,  47, 377, 177,  47,  47, 378, 379, 380, 381, 143,
+    47,  47, 382, 383,  32,  32,  32,  32,  47,  47, 359,  47,  47, 384, 169, 157,
+    90,  47,  47, 111, 385, 386, 387,  32,  47,  47,  47, 388, 389, 390,  47,  47,
+    47,  47,  47, 391, 392, 157, 157, 157,  47,  47, 393, 394, 395, 396,  32,  32,
+    47,  47,  47, 397, 398, 157,  66,  66,  47,  47, 399, 400, 157, 157, 157, 157,
+    47, 141, 401, 402, 144, 144, 144, 144,  47,  47, 382, 403,  66,  66,  66,  66,
+     9,   9,   9,   9,  11,  11, 126, 404,  47,  47,  47, 405, 406, 157, 157, 157,
+    47,  47,  47,  47,  47, 407, 408, 409, 410,  47,  47, 411, 412, 413,  47,  47,
+   414, 415,  66,  66,  47,  47,  47,  47,  47,  47, 393, 416, 417, 126, 143, 418,
+    47, 152, 419, 420,  32,  32,  32,  32,  47,  47,  47, 353, 421, 157,  47,  47,
+   422, 423, 157, 157, 157, 157, 157, 157,  47,  47,  47,  47,  47,  47,  47, 424,
+    47,  47,  47,  47, 143, 425, 426, 427, 216, 216, 216, 216, 216, 216, 216,  66,
+    47,  47,  47, 205, 205, 205, 205, 205,  47,  47,  47,  47,  47,  47, 300,  66,
+    47,  47,  47,  47,  47,  47,  47, 428,  47,  47,  47, 429, 430, 431, 432,  47,
+     9,   9,   9,   9,   9,   9,  11,  11, 143, 433,  66,  66,  66,  66,  66,  66,
+    47,  47,  47,  47, 384, 434, 409, 409, 435, 436,  27,  27,  27,  27, 437, 409,
+    47, 438, 205, 205, 205, 205, 205, 205, 144, 144, 144, 144, 144, 144, 439, 440,
+   441, 144, 442, 144, 144, 144, 144, 144, 144, 144, 144, 144, 443, 144, 144, 144,
+     9, 444,  11, 445, 446,  11, 193,   9, 447, 448,   9, 449,  11,   9, 444,  11,
+   445, 446,  11, 193,   9, 447, 448,   9, 449,  11,   9, 444,  11, 445, 446,  11,
+   193,   9, 447, 448,   9, 449,  11,   9, 444,  11, 193,   9, 450, 451, 452, 453,
+    11, 454,   9, 455, 456, 457, 458,  11, 459,   9, 460,  11, 461, 157, 157, 157,
+    32,  32,  32, 462,  32,  32, 463, 464, 465, 466,  32,  32,  32,  32,  32,  32,
+    32,  32,  32,  32,  32,  32,  32,  32,  47,  47,  47, 467, 468, 144, 144, 144,
+    47,  47,  47,  47,  47,  47, 469, 470,  47,  47,  47,  47, 349,  32,  32,  32,
+     9,   9, 447,  11, 471, 300,  66,  66, 143, 143, 472, 473, 143, 143, 143, 143,
+   143, 143, 474, 143, 143, 143, 143, 143,  47,  47,  47,  47,  47,  47,  47, 223,
+   475, 144, 144, 144, 144, 144, 144, 144, 144, 144, 144, 144, 144, 144, 144, 476,
+   144, 144, 144, 144, 144, 144, 144, 157, 205, 205, 205, 205, 205, 205, 205, 205,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   939, 940, 941, 942, 946, 948,   0, 962, 969, 970, 971, 976,1001,1002,1003,1008,
+     0,1033,1040,1041,1042,1043,1047,   0,   0,1080,1081,1082,1086,1110,   0,   0,
+  1124,1125,1126,1127,1131,1133,   0,1147,1154,1155,1156,1161,1187,1188,1189,1193,
+     0,1219,1226,1227,1228,1229,1233,   0,   0,1267,1268,1269,1273,1298,   0,1303,
+   943,1128, 944,1129, 954,1139, 958,1143, 959,1144, 960,1145, 961,1146, 964,1149,
+     0,   0, 973,1158, 974,1159, 975,1160, 983,1168, 978,1163, 988,1173, 990,1175,
+   991,1176, 993,1178, 994,1179,   0,   0,1004,1190,1005,1191,1006,1192,1014,1199,
+  1007,   0,   0,   0,1016,1201,1020,1206,   0,1022,1208,1025,1211,1023,1209,   0,
+     0,   0,   0,1032,1218,1037,1223,1035,1221,   0,   0,   0,1044,1230,1045,1231,
+  1049,1235,   0,   0,1058,1244,1064,1250,1060,1246,1066,1252,1067,1253,1072,1258,
+  1069,1255,1077,1264,1074,1261,   0,   0,1083,1270,1084,1271,1085,1272,1088,1275,
+  1089,1276,1096,1283,1103,1290,1111,1299,1115,1118,1307,1120,1309,1121,1310,   0,
+  1053,1239,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1093,
+  1280,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0, 949,1134,1010,
+  1195,1050,1236,1090,1277,1341,1368,1340,1367,1342,1369,1339,1366,   0,1320,1347,
+  1418,1419,1323,1350,   0,   0, 992,1177,1018,1204,1055,1241,1416,1417,1415,1424,
+  1202,   0,   0,   0, 987,1172,   0,   0,1031,1217,1321,1348,1322,1349,1338,1365,
+   950,1135, 951,1136, 979,1164, 980,1165,1011,1196,1012,1197,1051,1237,1052,1238,
+  1061,1247,1062,1248,1091,1278,1092,1279,1071,1257,1076,1263,   0,   0, 997,1182,
+     0,   0,   0,   0,   0,   0, 945,1130, 982,1167,1337,1364,1335,1362,1046,1232,
+  1422,1423,1113,1301,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     8,   9,   0,  10,1425,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   7,   0,   0,   0,   0,   0,   0,   0,   0,   0,   1,   0,
+     0,   0,   0,   0,   0,1314,1427,   5,1434,1438,1443,   0,1450,   0,1455,1461,
+  1514,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1446,1458,1468,1476,1480,1486,
+  1517,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1489,1503,1494,1500,1508,   0,
+     0,   0,   0,1520,1521,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1526,1528,   0,1525,   0,   0,   0,1522,   0,   0,   0,   0,1536,1532,1539,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1534,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1556,   0,   0,   0,   0,   0,   0,
+  1548,1550,   0,1547,   0,   0,   0,1567,   0,   0,   0,   0,1558,1554,1561,   0,
+     0,   0,   0,   0,   0,   0,1568,1569,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1529,1551,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1523,1545,1524,1546,   0,   0,1527,1549,   0,   0,1570,1571,1530,1552,1531,1553,
+     0,   0,1533,1555,1535,1557,1537,1559,   0,   0,1572,1573,1544,1566,1538,1560,
+  1540,1562,1541,1563,1542,1564,   0,   0,1543,1565,   0,   0,   0,   0,   0,   0,
+     0,   0,1606,1607,1609,1608,1610,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+  1613,   0,1611,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1612,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1620,   0,   0,   0,   0,   0,   0,
+     0,1623,   0,   0,1624,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1614,1615,1616,1617,1618,1619,1621,1622,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1628,1629,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1625,1626,   0,1627,
+     0,   0,   0,1634,   0,   0,1635,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1630,1631,1632,   0,   0,1633,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1639,   0,   0,1638,1640,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1636,1637,   0,   0,
+     0,   0,   0,   0,1641,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1642,1644,1643,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1645,   0,   0,   0,   0,   0,   0,   0,
+  1646,   0,   0,   0,   0,   0,   0,1648,1649,   0,1647,1650,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1651,1653,1652,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1654,   0,1655,1657,1656,   0,
+     0,   0,   0,1659,   0,   0,   0,   0,   0,   0,   0,   0,   0,1660,   0,   0,
+     0,   0,1661,   0,   0,   0,   0,1662,   0,   0,   0,   0,1663,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1658,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1664,   0,1665,1673,   0,1674,   0,   0,   0,   0,   0,   0,   0,
+     0,1666,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,1668,   0,   0,   0,   0,   0,   0,   0,   0,   0,1669,   0,   0,
+     0,   0,1670,   0,   0,   0,   0,1671,   0,   0,   0,   0,1672,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,1667,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1675,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,1676,   0,1677,   0,1678,   0,1679,   0,1680,   0,
+     0,   0,1681,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1682,   0,1683,   0,   0,
+  1684,1685,   0,1686,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   953,1138, 955,1140, 956,1141, 957,1142,1324,1351, 963,1148, 965,1150, 968,1153,
+   966,1151, 967,1152,1378,1380,1379,1381, 984,1169, 985,1170,1420,1421, 986,1171,
+   989,1174, 995,1180, 998,1183, 996,1181, 999,1184,1000,1185,1015,1200,1329,1356,
+  1017,1203,1019,1205,1021,1207,1024,1210,1687,1688,1027,1213,1026,1212,1028,1214,
+  1029,1215,1030,1216,1034,1220,1036,1222,1039,1225,1038,1224,1334,1361,1336,1363,
+  1382,1384,1383,1385,1056,1242,1057,1243,1059,1245,1063,1249,1689,1690,1065,1251,
+  1068,1254,1070,1256,1386,1387,1388,1389,1691,1692,1073,1259,1075,1262,1079,1266,
+  1078,1265,1095,1282,1098,1285,1097,1284,1390,1391,1392,1393,1099,1286,1100,1287,
+  1101,1288,1102,1289,1105,1292,1104,1291,1106,1294,1107,1295,1108,1296,1114,1302,
+  1119,1308,1122,1311,1123,1312,1186,1260,1293,1305,   0,1394,   0,   0,   0,   0,
+   952,1137, 947,1132,1317,1344,1316,1343,1319,1346,1318,1345,1693,1695,1371,1375,
+  1370,1374,1373,1377,1372,1376,1694,1696, 981,1166, 977,1162, 972,1157,1326,1353,
+  1325,1352,1328,1355,1327,1354,1697,1698,1009,1194,1013,1198,1054,1240,1048,1234,
+  1331,1358,1330,1357,1333,1360,1332,1359,1699,1700,1396,1401,1395,1400,1398,1403,
+  1397,1402,1399,1404,1094,1281,1087,1274,1406,1411,1405,1410,1408,1413,1407,1412,
+  1409,1414,1109,1297,1117,1306,1116,1304,1112,1300,   0,   0,   0,   0,   0,   0,
+  1471,1472,1701,1705,1702,1706,1703,1707,1430,1431,1715,1719,1716,1720,1717,1721,
+  1477,1478,1729,1731,1730,1732,   0,   0,1435,1436,1733,1735,1734,1736,   0,   0,
+  1481,1482,1737,1741,1738,1742,1739,1743,1439,1440,1751,1755,1752,1756,1753,1757,
+  1490,1491,1765,1768,1766,1769,1767,1770,1447,1448,1771,1774,1772,1775,1773,1776,
+  1495,1496,1777,1779,1778,1780,   0,   0,1451,1452,1781,1783,1782,1784,   0,   0,
+  1504,1505,1785,1788,1786,1789,1787,1790,   0,1459,   0,1791,   0,1792,   0,1793,
+  1509,1510,1794,1798,1795,1799,1796,1800,1462,1463,1808,1812,1809,1813,1810,1814,
+  1467,  21,1475,  22,1479,  23,1485,  24,1493,  27,1499,  28,1507,  29,   0,   0,
+  1704,1708,1709,1710,1711,1712,1713,1714,1718,1722,1723,1724,1725,1726,1727,1728,
+  1740,1744,1745,1746,1747,1748,1749,1750,1754,1758,1759,1760,1761,1762,1763,1764,
+  1797,1801,1802,1803,1804,1805,1806,1807,1811,1815,1816,1817,1818,1819,1820,1821,
+  1470,1469,1822,1474,1465,   0,1473,1825,1429,1428,1426,  12,1432,   0,  26,   0,
+     0,1315,1823,1484,1466,   0,1483,1829,1433,  13,1437,  14,1441,1826,1827,1828,
+  1488,1487,1513,  19,   0,   0,1492,1515,1445,1444,1442,  15,   0,1831,1832,1833,
+  1502,1501,1516,  25,1497,1498,1506,1518,1457,1456,1454,  17,1453,1313,  11,   3,
+     0,   0,1824,1512,1519,   0,1511,1830,1449,  16,1460,  18,1464,   4,   0,   0,
+    30,  31,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,  20,   0,   0,   0,   2,   6,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1834,1835,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1836,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1837,1839,1838,
+     0,   0,   0,   0,1840,   0,   0,   0,   0,1841,   0,   0,1842,   0,   0,   0,
+     0,   0,   0,   0,1843,   0,1844,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,1845,   0,   0,1846,   0,   0,1847,   0,1848,   0,   0,   0,   0,   0,   0,
+   937,   0,1850,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1849, 936, 938,
+  1851,1852,   0,   0,1853,1854,   0,   0,1855,1856,   0,   0,   0,   0,   0,   0,
+  1857,1858,   0,   0,1861,1862,   0,   0,1863,1864,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1867,1868,1869,1870,
+  1859,1860,1865,1866,   0,   0,   0,   0,   0,   0,1871,1872,1873,1874,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,  32,  33,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1875,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1877,   0,1878,   0,
+  1879,   0,1880,   0,1881,   0,1882,   0,1883,   0,1884,   0,1885,   0,1886,   0,
+  1887,   0,1888,   0,   0,1889,   0,1890,   0,1891,   0,   0,   0,   0,   0,   0,
+  1892,1893,   0,1894,1895,   0,1896,1897,   0,1898,1899,   0,1900,1901,   0,   0,
+     0,   0,   0,   0,1876,   0,   0,   0,   0,   0,   0,   0,   0,   0,1902,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1904,   0,1905,   0,
+  1906,   0,1907,   0,1908,   0,1909,   0,1910,   0,1911,   0,1912,   0,1913,   0,
+  1914,   0,1915,   0,   0,1916,   0,1917,   0,1918,   0,   0,   0,   0,   0,   0,
+  1919,1920,   0,1921,1922,   0,1923,1924,   0,1925,1926,   0,1927,1928,   0,   0,
+     0,   0,   0,   0,1903,   0,   0,1929,1930,1931,1932,   0,   0,   0,1933,   0,
+   710, 385, 724, 715, 455, 103, 186, 825, 825, 242, 751, 205, 241, 336, 524, 601,
+   663, 676, 688, 738, 411, 434, 474, 500, 649, 746, 799, 108, 180, 416, 482, 662,
+   810, 275, 462, 658, 692, 344, 618, 679, 293, 388, 440, 492, 740, 116, 146, 168,
+   368, 414, 481, 527, 606, 660, 665, 722, 781, 803, 809, 538, 553, 588, 642, 758,
+   811, 701, 233, 299, 573, 612, 487, 540, 714, 779, 232, 267, 412, 445, 457, 585,
+   594, 766, 167, 613, 149, 148, 560, 589, 648, 768, 708, 345, 411, 704, 105, 259,
+   313, 496, 518, 174, 542, 120, 307, 101, 430, 372, 584, 183, 228, 529, 650, 697,
+   424, 732, 428, 349, 632, 355, 517, 110, 135, 147, 403, 580, 624, 700, 750, 170,
+   193, 245, 297, 374, 463, 543, 763, 801, 812, 815, 162, 384, 420, 730, 287, 330,
+   337, 366, 459, 476, 509, 558, 591, 610, 726, 652, 734, 759, 154, 163, 198, 473,
+   683, 697, 292, 311, 353, 423, 572, 494, 113, 217, 259, 280, 314, 499, 506, 603,
+   608, 752, 778, 782, 788, 117, 557, 748, 774, 320, 109, 126, 260, 265, 373, 411,
+   479, 523, 655, 737, 823, 380, 765, 161, 395, 398, 438, 451, 502, 516, 537, 583,
+   791, 136, 340, 769, 122, 273, 446, 727, 305, 322, 400, 496, 771, 155, 190, 269,
+   377, 391, 406, 432, 501, 519, 599, 684, 687, 749, 776, 175, 452, 191, 480, 510,
+   659, 772, 805, 813, 397, 444, 619, 566, 568, 575, 491, 471, 707, 111, 636, 156,
+   153, 288, 346, 578, 256, 435, 383, 729, 680, 767, 694, 295, 128, 210,   0,   0,
+   227,   0, 379,   0,   0, 150, 493, 525, 544, 551, 552, 556, 783, 576, 604,   0,
+   661,   0, 703,   0,   0, 735, 743,   0,   0,   0, 793, 794, 795, 808, 741, 773,
+   118, 127, 130, 166, 169, 177, 207, 213, 215, 226, 229, 268, 270, 317, 327, 329,
+   335, 369, 375, 381, 404, 441, 448, 458, 477, 484, 503, 539, 545, 547, 546, 548,
+   549, 550, 554, 555, 561, 564, 569, 591, 593, 595, 598, 607, 620, 625, 625, 651,
+   690, 695, 705, 706, 716, 717, 733, 735, 777, 786, 790, 315, 869, 623,   0,   0,
+   102, 145, 134, 115, 129, 138, 165, 171, 207, 202, 206, 212, 227, 231, 240, 243,
+   250, 254, 294, 296, 303, 308, 319, 325, 321, 329, 326, 335, 341, 357, 360, 362,
+   370, 379, 388, 389, 393, 421, 424, 438, 456, 454, 458, 465, 477, 535, 485, 490,
+   493, 507, 512, 514, 521, 522, 525, 526, 528, 533, 532, 541, 565, 569, 574, 586,
+   591, 597, 607, 637, 647, 674, 691, 693, 695, 698, 703, 699, 705, 704, 702, 706,
+   709, 717, 728, 736, 747, 754, 770, 777, 783, 784, 786, 787, 790, 802, 825, 848,
+   847, 857,  55,  65,  66, 883, 892, 916, 822, 824,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1586,   0,1605,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1602,1603,1934,1935,1574,1575,
+  1576,1577,1579,1580,1581,1583,1584,   0,1585,1587,1588,1589,1591,   0,1592,   0,
+  1593,1594,   0,1595,1596,   0,1598,1599,1600,1601,1604,1582,1578,1590,1597,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1936,   0,1937,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1938,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1939,1940,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1941,1942,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1944,1943,   0,1945,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1946,1947,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,1948,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1949,1950,
+  1951,1952,1953,1954,1955,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+     0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,1956,1957,1958,1960,1959,
+  1961,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,
+   106, 104, 107, 826, 114, 118, 119, 121, 123, 124, 127, 125,  34, 830, 130, 131,
+   132, 137, 827,  35, 133, 139, 829, 142, 143, 112, 144, 145, 924, 151, 152,  37,
+   157, 158, 159, 160,  38, 165, 166, 169, 171, 172, 173, 174, 176, 177, 178, 179,
+   181, 182, 182, 182, 833, 468, 184, 185, 834, 187, 188, 189, 196, 192, 194, 195,
+   197, 199, 200, 201, 203, 204, 204, 206, 208, 209, 211, 218, 213, 219, 214, 216,
+   153, 234, 221, 222, 223, 220, 225, 224, 230, 835, 235, 236, 237, 238, 239, 244,
+   836, 837, 247, 248, 249, 246, 251,  39,  40, 253, 255, 255, 838, 257, 258, 259,
+   261, 839, 262, 263, 301, 264,  41, 266, 270, 272, 271, 841, 274, 842, 277, 276,
+   278, 281, 282,  42, 283, 284, 285, 286,  43, 843,  44, 289, 290, 291, 293, 934,
+   298, 845, 845, 621, 300, 300,  45, 852, 894, 302, 304,  46, 306, 309, 310, 312,
+   316,  48,  47, 317, 846, 318, 323, 324, 325, 324, 328, 329, 333, 331, 332, 334,
+   335, 336, 338, 339, 342, 343, 347, 351, 849, 350, 348, 352, 354, 359, 850, 361,
+   358, 356,  49, 363, 365, 367, 364,  50, 369, 371, 851, 376, 386, 378,  53, 381,
+    52,  51, 140, 141, 387, 382, 614,  78, 388, 389, 390, 394, 392, 856,  54, 399,
+   396, 402, 404, 858, 405, 401, 407,  55, 408, 409, 410, 413, 859, 415,  56, 417,
+   860, 418,  57, 419, 422, 424, 425, 861, 840, 862, 426, 863, 429, 431, 427, 433,
+   437, 441, 438, 439, 442, 443, 864, 436, 449, 450,  58, 454, 453, 865, 447, 460,
+   866, 867, 461, 466, 465, 464,  59, 467, 470, 469, 472, 828, 475, 868, 478, 870,
+   483, 485, 486, 871, 488, 489, 872, 873, 495, 497,  60, 498,  61,  61, 504, 505,
+   507, 508, 511,  62, 513, 874, 515, 875, 518, 844, 520, 876, 877, 878,  63,  64,
+   528, 880, 879, 881, 882, 530, 531, 531, 533,  66, 534,  67,  68, 884, 536, 538,
+   541,  69, 885, 549, 886, 887, 556, 559,  70, 561, 562, 563, 888, 889, 889, 567,
+    71, 890, 570, 571,  72, 891, 577,  73, 581, 579, 582, 893, 587,  74, 590, 592,
+   596,  75, 895, 896,  76, 897, 600, 898, 602, 605, 607, 899, 900, 609, 901, 611,
+   853,  77, 615, 616,  79, 617, 252, 902, 903, 854, 855, 621, 622, 731,  80, 627,
+   626, 628, 164, 629, 630, 631, 633, 904, 632, 634, 639, 640, 635, 641, 646, 651,
+   638, 643, 644, 645, 905, 907, 906,  81, 653, 654, 656, 911, 657, 908,  82,  83,
+   909, 910,  84, 664, 665, 666, 667, 669, 668, 671, 670, 674, 672, 673, 675,  85,
+   677, 678,  86, 681, 682, 912, 685, 686,  87, 689,  36, 913, 914,  88,  89, 696,
+   702, 709, 711, 915, 712, 713, 718, 719, 917, 831, 721, 720, 723, 832, 725, 728,
+   918, 919, 739, 742, 744, 920, 745, 753, 756, 757, 755, 760, 761, 921, 762,  90,
+   764, 922,  91, 775, 279, 780, 923, 925,  92,  93, 785, 926,  94, 927, 787, 787,
+   789, 928, 792,  95, 796, 797, 798, 800,  96, 929, 802, 804, 806,  97,  98, 807,
+   930,  99, 931, 932, 933, 814, 100, 816, 817, 818, 819, 820, 821, 935,   0,   0,
+};
+static const int16_t
+_hb_ucd_i16[92] =
+{
+      0,    0,    1,   -1,    2,    0,   -2,    0,    0,    2,    0,   -2,    0,   16,    0,  -16,
+      0,    1,   -1,    0,    3,    3,    3,   -3,   -3,   -3,    0, 2016,    0, 2527, 1923, 1914,
+   1918,    0, 2250,    0,    0,  138,    0,    7,   -7,    0,   -1,    1, 1824,    0, 2104,    0,
+   2108, 2106,    0, 2106, 1316,    0,   -1, -138,    8,    8,    8,    0,    7,    7,   -8,   -8,
+     -8,   -7,-1316,    1,   -1,    3,   -3,    1,    0,-1914,-1918,    0,    0,-1923,-1824,    0,
+      0,-2016,-2104,    0,    0,-2106,-2108,-2106,-2250,    0,-2527,    0,
+};
+
+static inline uint_fast8_t
+_hb_ucd_gc (unsigned u)
+{
+  return u<1114112u?_hb_ucd_u8[4920+(((_hb_ucd_u8[1104+(((_hb_ucd_u16[((_hb_ucd_u8[272+(((_hb_ucd_u8[u>>1>>3>>3>>5])<<5)+((u>>1>>3>>3)&31u))])<<3)+((u>>1>>3)&7u)])<<3)+((u>>1)&7u))])<<1)+((u)&1u))]:2;
+}
+static inline uint_fast8_t
+_hb_ucd_ccc (unsigned u)
+{
+  return u<125259u?_hb_ucd_u8[6796+(((_hb_ucd_u8[6276+(((_hb_ucd_u8[5844+(((_hb_ucd_u8[5508+(((_hb_ucd_u8[5262+(u>>2>>2>>2>>3)])<<3)+((u>>2>>2>>2)&7u))])<<2)+((u>>2>>2)&3u))])<<2)+((u>>2)&3u))])<<2)+((u)&3u))]:0;
+}
+static inline unsigned
+_hb_ucd_b4 (const uint8_t* a, unsigned i)
+{
+  return (a[i>>1]>>((i&1u)<<2))&15u;
+}
+static inline int_fast16_t
+_hb_ucd_bmg (unsigned u)
+{
+  return u<65380u?_hb_ucd_i16[((_hb_ucd_u8[7672+(((_hb_ucd_u8[7448+(((_hb_ucd_u8[7352+(((_hb_ucd_b4(7288+_hb_ucd_u8,u>>1>>2>>3>>3))<<3)+((u>>1>>2>>3)&7u))])<<3)+((u>>1>>2)&7u))])<<2)+((u>>1)&3u))])<<1)+((u)&1u)]:0;
+}
+static inline uint_fast8_t
+_hb_ucd_sc (unsigned u)
+{
+  return u<918016u?_hb_ucd_u8[11242+(((_hb_ucd_u8[10314+(((_hb_ucd_u8[8938+(((_hb_ucd_u8[8362+(((_hb_ucd_u8[7912+(u>>2>>2>>3>>4)])<<4)+((u>>2>>2>>3)&15u))])<<3)+((u>>2>>2)&7u))])<<2)+((u>>2)&3u))])<<2)+((u)&3u))]:2;
+}
+static inline uint_fast16_t
+_hb_ucd_dm (unsigned u)
+{
+  return u<195102u?_hb_ucd_u16[1536+(((_hb_ucd_u8[12544+(((_hb_ucd_u8[12162+(u>>4>>5)])<<5)+((u>>4)&31u))])<<4)+((u)&15u))]:0;
+}
+
+#endif
+
+
+#endif /* HB_UCD_TABLE_HH */
+
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-ucd.cc b/thirdparty/harfbuzz/src/hb-ucd.cc
new file mode 100644
index 0000000000..ad72a26c04
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-ucd.cc
@@ -0,0 +1,248 @@
+/*
+ * Copyright (C) 2012 Grigori Goronzy <greg@kinoho.net>
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#include "hb.hh"
+#include "hb-unicode.hh"
+#include "hb-machinery.hh"
+
+#include "hb-ucd-table.hh"
+
+static hb_unicode_combining_class_t
+hb_ucd_combining_class (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			hb_codepoint_t unicode,
+			void *user_data HB_UNUSED)
+{
+  return (hb_unicode_combining_class_t) _hb_ucd_ccc (unicode);
+}
+
+static hb_unicode_general_category_t
+hb_ucd_general_category (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+			 hb_codepoint_t unicode,
+			 void *user_data HB_UNUSED)
+{
+  return (hb_unicode_general_category_t) _hb_ucd_gc (unicode);
+}
+
+static hb_codepoint_t
+hb_ucd_mirroring (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+		  hb_codepoint_t unicode,
+		  void *user_data HB_UNUSED)
+{
+  return unicode + _hb_ucd_bmg (unicode);
+}
+
+static hb_script_t
+hb_ucd_script (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+	       hb_codepoint_t unicode,
+	       void *user_data HB_UNUSED)
+{
+  return _hb_ucd_sc_map[_hb_ucd_sc (unicode)];
+}
+
+
+#define SBASE 0xAC00u
+#define LBASE 0x1100u
+#define VBASE 0x1161u
+#define TBASE 0x11A7u
+#define SCOUNT 11172u
+#define LCOUNT 19u
+#define VCOUNT 21u
+#define TCOUNT 28u
+#define NCOUNT (VCOUNT * TCOUNT)
+
+static inline bool
+_hb_ucd_decompose_hangul (hb_codepoint_t ab, hb_codepoint_t *a, hb_codepoint_t *b)
+{
+  unsigned si = ab - SBASE;
+
+  if (si >= SCOUNT)
+    return false;
+
+  if (si % TCOUNT)
+  {
+    /* LV,T */
+    *a = SBASE + (si / TCOUNT) * TCOUNT;
+    *b = TBASE + (si % TCOUNT);
+    return true;
+  } else {
+    /* L,V */
+    *a = LBASE + (si / NCOUNT);
+    *b = VBASE + (si % NCOUNT) / TCOUNT;
+    return true;
+  }
+}
+
+static inline bool
+_hb_ucd_compose_hangul (hb_codepoint_t a, hb_codepoint_t b, hb_codepoint_t *ab)
+{
+  if (a >= SBASE && a < (SBASE + SCOUNT) && b > TBASE && b < (TBASE + TCOUNT) &&
+    !((a - SBASE) % TCOUNT))
+  {
+    /* LV,T */
+    *ab = a + (b - TBASE);
+    return true;
+  }
+  else if (a >= LBASE && a < (LBASE + LCOUNT) && b >= VBASE && b < (VBASE + VCOUNT))
+  {
+    /* L,V */
+    int li = a - LBASE;
+    int vi = b - VBASE;
+    *ab = SBASE + li * NCOUNT + vi * TCOUNT;
+    return true;
+  }
+  else
+    return false;
+}
+
+static int
+_cmp_pair (const void *_key, const void *_item)
+{
+  uint64_t& a = * (uint64_t*) _key;
+  uint64_t b = (* (uint64_t*) _item) & HB_CODEPOINT_ENCODE3(0x1FFFFFu, 0x1FFFFFu, 0);
+
+  return a < b ? -1 : a > b ? +1 : 0;
+}
+static int
+_cmp_pair_11_7_14 (const void *_key, const void *_item)
+{
+  uint32_t& a = * (uint32_t*) _key;
+  uint32_t b = (* (uint32_t*) _item) & HB_CODEPOINT_ENCODE3_11_7_14(0x1FFFFFu, 0x1FFFFFu, 0);
+
+  return a < b ? -1 : a > b ? +1 : 0;
+}
+
+static hb_bool_t
+hb_ucd_compose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+		hb_codepoint_t a, hb_codepoint_t b, hb_codepoint_t *ab,
+		void *user_data HB_UNUSED)
+{
+  if (_hb_ucd_compose_hangul (a, b, ab)) return true;
+
+  hb_codepoint_t u = 0;
+
+  if ((a & 0xFFFFF800u) == 0x0000u && (b & 0xFFFFFF80) == 0x0300u)
+  {
+    uint32_t k = HB_CODEPOINT_ENCODE3_11_7_14 (a, b, 0);
+    const uint32_t *v = hb_bsearch (k,
+				    _hb_ucd_dm2_u32_map,
+				    ARRAY_LENGTH (_hb_ucd_dm2_u32_map),
+				    sizeof (*_hb_ucd_dm2_u32_map),
+				    _cmp_pair_11_7_14);
+    if (likely (!v)) return false;
+    u = HB_CODEPOINT_DECODE3_11_7_14_3 (*v);
+  }
+  else
+  {
+    uint64_t k = HB_CODEPOINT_ENCODE3 (a, b, 0);
+    const uint64_t *v = hb_bsearch (k,
+				    _hb_ucd_dm2_u64_map,
+				    ARRAY_LENGTH (_hb_ucd_dm2_u64_map),
+				    sizeof (*_hb_ucd_dm2_u64_map),
+				    _cmp_pair);
+    if (likely (!v)) return false;
+    u = HB_CODEPOINT_DECODE3_3 (*v);
+  }
+
+  if (unlikely (!u)) return false;
+  *ab = u;
+  return true;
+}
+
+static hb_bool_t
+hb_ucd_decompose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
+		  hb_codepoint_t ab, hb_codepoint_t *a, hb_codepoint_t *b,
+		  void *user_data HB_UNUSED)
+{
+  if (_hb_ucd_decompose_hangul (ab, a, b)) return true;
+
+  unsigned i = _hb_ucd_dm (ab);
+
+  if (likely (!i)) return false;
+  i--;
+
+  if (i < ARRAY_LENGTH (_hb_ucd_dm1_p0_map) + ARRAY_LENGTH (_hb_ucd_dm1_p2_map))
+  {
+    if (i < ARRAY_LENGTH (_hb_ucd_dm1_p0_map))
+      *a = _hb_ucd_dm1_p0_map[i];
+    else
+    {
+      i -= ARRAY_LENGTH (_hb_ucd_dm1_p0_map);
+      *a = 0x20000 | _hb_ucd_dm1_p2_map[i];
+    }
+    *b = 0;
+    return true;
+  }
+  i -= ARRAY_LENGTH (_hb_ucd_dm1_p0_map) + ARRAY_LENGTH (_hb_ucd_dm1_p2_map);
+
+  if (i < ARRAY_LENGTH (_hb_ucd_dm2_u32_map))
+  {
+    uint32_t v = _hb_ucd_dm2_u32_map[i];
+    *a = HB_CODEPOINT_DECODE3_11_7_14_1 (v);
+    *b = HB_CODEPOINT_DECODE3_11_7_14_2 (v);
+    return true;
+  }
+  i -= ARRAY_LENGTH (_hb_ucd_dm2_u32_map);
+
+  uint64_t v = _hb_ucd_dm2_u64_map[i];
+  *a = HB_CODEPOINT_DECODE3_1 (v);
+  *b = HB_CODEPOINT_DECODE3_2 (v);
+  return true;
+}
+
+
+#if HB_USE_ATEXIT
+static void free_static_ucd_funcs ();
+#endif
+
+static struct hb_ucd_unicode_funcs_lazy_loader_t : hb_unicode_funcs_lazy_loader_t<hb_ucd_unicode_funcs_lazy_loader_t>
+{
+  static hb_unicode_funcs_t *create ()
+  {
+    hb_unicode_funcs_t *funcs = hb_unicode_funcs_create (nullptr);
+
+    hb_unicode_funcs_set_combining_class_func (funcs, hb_ucd_combining_class, nullptr, nullptr);
+    hb_unicode_funcs_set_general_category_func (funcs, hb_ucd_general_category, nullptr, nullptr);
+    hb_unicode_funcs_set_mirroring_func (funcs, hb_ucd_mirroring, nullptr, nullptr);
+    hb_unicode_funcs_set_script_func (funcs, hb_ucd_script, nullptr, nullptr);
+    hb_unicode_funcs_set_compose_func (funcs, hb_ucd_compose, nullptr, nullptr);
+    hb_unicode_funcs_set_decompose_func (funcs, hb_ucd_decompose, nullptr, nullptr);
+
+    hb_unicode_funcs_make_immutable (funcs);
+
+#if HB_USE_ATEXIT
+    atexit (free_static_ucd_funcs);
+#endif
+
+    return funcs;
+  }
+} static_ucd_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_ucd_funcs ()
+{
+  static_ucd_funcs.free_instance ();
+}
+#endif
+
+hb_unicode_funcs_t *
+hb_ucd_get_unicode_funcs ()
+{
+#ifdef HB_NO_UCD
+  return hb_unicode_funcs_get_empty ();
+#endif
+  return static_ucd_funcs.get_unconst ();
+}
diff --git a/thirdparty/harfbuzz/src/hb-unicode-emoji-table.hh b/thirdparty/harfbuzz/src/hb-unicode-emoji-table.hh
new file mode 100644
index 0000000000..eb7776eecb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-unicode-emoji-table.hh
@@ -0,0 +1,78 @@
+/* == Start of generated table == */
+/*
+ * The following tables are generated by running:
+ *
+ *   ./gen-emoji-table.py emoji-data.txt
+ *
+ * on file with this header:
+ *
+ * # emoji-data.txt
+ * # Date: 2020-01-28, 20:52:38 GMT
+ * # © 2020 Unicode®, Inc.
+ * # Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
+ * # For terms of use, see http://www.unicode.org/terms_of_use.html
+ * #
+ * # Emoji Data for UTS #51
+ * # Version: 13.0
+ * #
+ * # For documentation and usage, see http://www.unicode.org/reports/tr51
+ */
+
+#ifndef HB_UNICODE_EMOJI_TABLE_HH
+#define HB_UNICODE_EMOJI_TABLE_HH
+
+#include "hb-unicode.hh"
+
+static const uint8_t
+_hb_emoji_u8[448] =
+{
+    0,  0,  0,  0, 33,  3,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, 84,118,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0,  1,  0,  0,  0,  0,  0,  2,  0,  0,  3,
+    0,  0,  0,  0,  0,  0,  4,  5,  6,  7,  8,  7,  9, 10, 11,  0,
+    0,  0,  0,  0, 12,  0,  0,  0,  0,  0,  0,  0, 13,  0,  0,  0,
+    7,  7,  7, 14, 15, 16, 17, 18, 19, 20,  7,  7,  7,  7,  7, 21,
+    7,  7,  7,  7, 22, 23,  7,  7,  7, 24,  7, 14,  0, 25,  0, 26,
+   27, 28, 29, 14, 30, 31,  7,  7,  7,  7,  7, 14,  0,  0,  0,  0,
+    7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7,  7, 22,
+    0,  0,  0,  0,  0,  0,  0,  0,  0,  0,240,  1,  0,  2,  0,  0,
+    0,  0,  0,  4,  0,  0,  0,  0,  0,  0,  0,  0,  0,254,  7,  3,
+    0,  0,  0,  0,  0,  4,  0,  0,  0,  0,  0,  0,  0,  0,  0, 56,
+  159,255,243,255,255,255,255,255,255,255,255,255,255,255,255,255,
+   31,  0,255,255,255,255,255,255, 31,255,  3,  0,  0,  0,  8,  0,
+    0,  0, 24,  0,120,  0,  0,  0,  0,  0, 96,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,  0, 16,  0, 96,  0,  0,  8,  0,  0,  0,  0,
+  255,255,255,255,255,255,255,127,  0, 96,  0,  0,  0,  0,  0,  0,
+    0,  0,  0,  0,  0,240,  1, 64,  0,  0,254,  3,  0,224,255,255,
+  255,255,255,255, 31,  0,  0,  0,254,127,  0,  0,  0,  0,252,115,
+    0,254,255,255,255,255,255,255,255,255,255,255,255,255,255,  3,
+  255,255,255,255,255,255,255, 31,192,255,255,255,255,255,255,255,
+  255,127,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,240,127,
+    0,  0,224,255,255,255,255,127,  0,112,  0,  0,  0,  0,  0,  0,
+    0,127,  0,124,  0,  0,  0,  0,  0,127,  0,  0,  0,192,255,255,
+    0,240,255,255,255,255,255,243,159,255,255,255,255,255,255,255,
+};
+
+static inline unsigned
+_hb_emoji_b4 (const uint8_t* a, unsigned i)
+{
+  return (a[i>>1]>>((i&1u)<<2))&15u;
+}
+static inline unsigned
+_hb_emoji_b1 (const uint8_t* a, unsigned i)
+{
+  return (a[i>>3]>>((i&7u)<<0))&1u;
+}
+static inline uint_fast8_t
+_hb_emoji_is_Extended_Pictographic (unsigned u)
+{
+  return u<131069u?_hb_emoji_b1(192+_hb_emoji_u8,((_hb_emoji_u8[64+(((_hb_emoji_b4(_hb_emoji_u8,u>>6>>4))<<4)+((u>>6)&15u))])<<6)+((u)&63u)):0;
+}
+
+
+#endif /* HB_UNICODE_EMOJI_TABLE_HH */
+
+/* == End of generated table == */
diff --git a/thirdparty/harfbuzz/src/hb-unicode.cc b/thirdparty/harfbuzz/src/hb-unicode.cc
new file mode 100644
index 0000000000..36070a7f18
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-unicode.cc
@@ -0,0 +1,586 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Codethink Limited
+ * Copyright © 2010,2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Codethink Author(s): Ryan Lortie
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#include "hb-unicode.hh"
+
+
+/**
+ * SECTION: hb-unicode
+ * @title: hb-unicode
+ * @short_description: Unicode character property access
+ * @include: hb.h
+ *
+ * Unicode functions are used to access Unicode character properties.
+ * Client can pass its own Unicode functions to HarfBuzz, or access
+ * the built-in Unicode functions that come with HarfBuzz.
+ *
+ * With the Unicode functions, one can query variour Unicode character
+ * properties, such as General Category, Script, Combining Class, etc.
+ **/
+
+
+/*
+ * hb_unicode_funcs_t
+ */
+
+static hb_unicode_combining_class_t
+hb_unicode_combining_class_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				hb_codepoint_t      unicode   HB_UNUSED,
+				void               *user_data HB_UNUSED)
+{
+  return HB_UNICODE_COMBINING_CLASS_NOT_REORDERED;
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+static unsigned int
+hb_unicode_eastasian_width_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				hb_codepoint_t      unicode   HB_UNUSED,
+				void               *user_data HB_UNUSED)
+{
+  return 1;
+}
+#endif
+
+static hb_unicode_general_category_t
+hb_unicode_general_category_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				 hb_codepoint_t      unicode   HB_UNUSED,
+				 void               *user_data HB_UNUSED)
+{
+  return HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER;
+}
+
+static hb_codepoint_t
+hb_unicode_mirroring_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+			  hb_codepoint_t      unicode,
+			  void               *user_data HB_UNUSED)
+{
+  return unicode;
+}
+
+static hb_script_t
+hb_unicode_script_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+		       hb_codepoint_t      unicode   HB_UNUSED,
+		       void               *user_data HB_UNUSED)
+{
+  return HB_SCRIPT_UNKNOWN;
+}
+
+static hb_bool_t
+hb_unicode_compose_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+			hb_codepoint_t      a         HB_UNUSED,
+			hb_codepoint_t      b         HB_UNUSED,
+			hb_codepoint_t     *ab        HB_UNUSED,
+			void               *user_data HB_UNUSED)
+{
+  return false;
+}
+
+static hb_bool_t
+hb_unicode_decompose_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+			  hb_codepoint_t      ab        HB_UNUSED,
+			  hb_codepoint_t     *a         HB_UNUSED,
+			  hb_codepoint_t     *b         HB_UNUSED,
+			  void               *user_data HB_UNUSED)
+{
+  return false;
+}
+
+
+#ifndef HB_DISABLE_DEPRECATED
+static unsigned int
+hb_unicode_decompose_compatibility_nil (hb_unicode_funcs_t *ufuncs     HB_UNUSED,
+					hb_codepoint_t      u          HB_UNUSED,
+					hb_codepoint_t     *decomposed HB_UNUSED,
+					void               *user_data  HB_UNUSED)
+{
+  return 0;
+}
+#endif
+
+#if !defined(HB_NO_UNICODE_FUNCS) && defined(HAVE_GLIB)
+#include "hb-glib.h"
+#endif
+#if !defined(HB_NO_UNICODE_FUNCS) && defined(HAVE_ICU) && defined(HAVE_ICU_BUILTIN)
+#include "hb-icu.h"
+#endif
+
+hb_unicode_funcs_t *
+hb_unicode_funcs_get_default ()
+{
+#if !defined(HB_NO_UNICODE_FUNCS) && !defined(HB_NO_UCD)
+  return hb_ucd_get_unicode_funcs ();
+#elif !defined(HB_NO_UNICODE_FUNCS) && defined(HAVE_GLIB)
+  return hb_glib_get_unicode_funcs ();
+#elif !defined(HB_NO_UNICODE_FUNCS) && defined(HAVE_ICU) && defined(HAVE_ICU_BUILTIN)
+  return hb_icu_get_unicode_funcs ();
+#else
+#define HB_UNICODE_FUNCS_NIL 1
+  return hb_unicode_funcs_get_empty ();
+#endif
+}
+
+#if !defined(HB_NO_UNICODE_FUNCS) && defined(HB_UNICODE_FUNCS_NIL)
+#error "Could not find any Unicode functions implementation, you have to provide your own"
+#error "Consider building hb-ucd.cc.  If you absolutely want to build without any, check the code."
+#endif
+
+/**
+ * hb_unicode_funcs_create: (Xconstructor)
+ * @parent: (nullable):
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_unicode_funcs_t *
+hb_unicode_funcs_create (hb_unicode_funcs_t *parent)
+{
+  hb_unicode_funcs_t *ufuncs;
+
+  if (!(ufuncs = hb_object_create<hb_unicode_funcs_t> ()))
+    return hb_unicode_funcs_get_empty ();
+
+  if (!parent)
+    parent = hb_unicode_funcs_get_empty ();
+
+  hb_unicode_funcs_make_immutable (parent);
+  ufuncs->parent = hb_unicode_funcs_reference (parent);
+
+  ufuncs->func = parent->func;
+
+  /* We can safely copy user_data from parent since we hold a reference
+   * onto it and it's immutable.  We should not copy the destroy notifiers
+   * though. */
+  ufuncs->user_data = parent->user_data;
+
+  return ufuncs;
+}
+
+
+DEFINE_NULL_INSTANCE (hb_unicode_funcs_t) =
+{
+  HB_OBJECT_HEADER_STATIC,
+
+  nullptr, /* parent */
+  {
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_##name##_nil,
+    HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+  }
+};
+
+/**
+ * hb_unicode_funcs_get_empty:
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_unicode_funcs_t *
+hb_unicode_funcs_get_empty ()
+{
+  return const_cast<hb_unicode_funcs_t *> (&Null (hb_unicode_funcs_t));
+}
+
+/**
+ * hb_unicode_funcs_reference: (skip)
+ * @ufuncs: Unicode functions.
+ *
+ *
+ *
+ * Return value: (transfer full):
+ *
+ * Since: 0.9.2
+ **/
+hb_unicode_funcs_t *
+hb_unicode_funcs_reference (hb_unicode_funcs_t *ufuncs)
+{
+  return hb_object_reference (ufuncs);
+}
+
+/**
+ * hb_unicode_funcs_destroy: (skip)
+ * @ufuncs: Unicode functions.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_unicode_funcs_destroy (hb_unicode_funcs_t *ufuncs)
+{
+  if (!hb_object_destroy (ufuncs)) return;
+
+#define HB_UNICODE_FUNC_IMPLEMENT(name) \
+  if (ufuncs->destroy.name) ufuncs->destroy.name (ufuncs->user_data.name);
+    HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+
+  hb_unicode_funcs_destroy (ufuncs->parent);
+
+  free (ufuncs);
+}
+
+/**
+ * hb_unicode_funcs_set_user_data: (skip)
+ * @ufuncs: Unicode functions.
+ * @key:
+ * @data:
+ * @destroy:
+ * @replace:
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_unicode_funcs_set_user_data (hb_unicode_funcs_t *ufuncs,
+				hb_user_data_key_t *key,
+				void *              data,
+				hb_destroy_func_t   destroy,
+				hb_bool_t           replace)
+{
+  return hb_object_set_user_data (ufuncs, key, data, destroy, replace);
+}
+
+/**
+ * hb_unicode_funcs_get_user_data: (skip)
+ * @ufuncs: Unicode functions.
+ * @key:
+ *
+ *
+ *
+ * Return value: (transfer none):
+ *
+ * Since: 0.9.2
+ **/
+void *
+hb_unicode_funcs_get_user_data (hb_unicode_funcs_t *ufuncs,
+				hb_user_data_key_t *key)
+{
+  return hb_object_get_user_data (ufuncs, key);
+}
+
+
+/**
+ * hb_unicode_funcs_make_immutable:
+ * @ufuncs: Unicode functions.
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+void
+hb_unicode_funcs_make_immutable (hb_unicode_funcs_t *ufuncs)
+{
+  if (hb_object_is_immutable (ufuncs))
+    return;
+
+  hb_object_make_immutable (ufuncs);
+}
+
+/**
+ * hb_unicode_funcs_is_immutable:
+ * @ufuncs: Unicode functions.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_unicode_funcs_is_immutable (hb_unicode_funcs_t *ufuncs)
+{
+  return hb_object_is_immutable (ufuncs);
+}
+
+/**
+ * hb_unicode_funcs_get_parent:
+ * @ufuncs: Unicode functions.
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_unicode_funcs_t *
+hb_unicode_funcs_get_parent (hb_unicode_funcs_t *ufuncs)
+{
+  return ufuncs->parent ? ufuncs->parent : hb_unicode_funcs_get_empty ();
+}
+
+
+#define HB_UNICODE_FUNC_IMPLEMENT(name)						\
+										\
+void										\
+hb_unicode_funcs_set_##name##_func (hb_unicode_funcs_t		   *ufuncs,	\
+				    hb_unicode_##name##_func_t	    func,	\
+				    void			   *user_data,	\
+				    hb_destroy_func_t		    destroy)	\
+{										\
+  if (hb_object_is_immutable (ufuncs))						\
+    return;									\
+										\
+  if (ufuncs->destroy.name)							\
+    ufuncs->destroy.name (ufuncs->user_data.name);				\
+										\
+  if (func) {									\
+    ufuncs->func.name = func;							\
+    ufuncs->user_data.name = user_data;						\
+    ufuncs->destroy.name = destroy;						\
+  } else {									\
+    ufuncs->func.name = ufuncs->parent->func.name;				\
+    ufuncs->user_data.name = ufuncs->parent->user_data.name;			\
+    ufuncs->destroy.name = nullptr;						\
+  }										\
+}
+
+HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+
+
+#define HB_UNICODE_FUNC_IMPLEMENT(return_type, name)				\
+										\
+return_type									\
+hb_unicode_##name (hb_unicode_funcs_t *ufuncs,					\
+		   hb_codepoint_t      unicode)					\
+{										\
+  return ufuncs->name (unicode);						\
+}
+HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE
+#undef HB_UNICODE_FUNC_IMPLEMENT
+
+/**
+ * hb_unicode_compose:
+ * @ufuncs: Unicode functions.
+ * @a:
+ * @b:
+ * @ab: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_unicode_compose (hb_unicode_funcs_t *ufuncs,
+		    hb_codepoint_t      a,
+		    hb_codepoint_t      b,
+		    hb_codepoint_t     *ab)
+{
+  return ufuncs->compose (a, b, ab);
+}
+
+/**
+ * hb_unicode_decompose:
+ * @ufuncs: Unicode functions.
+ * @ab:
+ * @a: (out):
+ * @b: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ **/
+hb_bool_t
+hb_unicode_decompose (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t      ab,
+		      hb_codepoint_t     *a,
+		      hb_codepoint_t     *b)
+{
+  return ufuncs->decompose (ab, a, b);
+}
+
+#ifndef HB_DISABLE_DEPRECATED
+/**
+ * hb_unicode_decompose_compatibility:
+ * @ufuncs: Unicode functions.
+ * @u:
+ * @decomposed: (out):
+ *
+ *
+ *
+ * Return value:
+ *
+ * Since: 0.9.2
+ * Deprecated: 2.0.0
+ **/
+unsigned int
+hb_unicode_decompose_compatibility (hb_unicode_funcs_t *ufuncs,
+				    hb_codepoint_t      u,
+				    hb_codepoint_t     *decomposed)
+{
+  return ufuncs->decompose_compatibility (u, decomposed);
+}
+#endif
+
+
+#ifndef HB_NO_OT_SHAPE
+/* See hb-unicode.hh for details. */
+const uint8_t
+_hb_modified_combining_class[256] =
+{
+  0, /* HB_UNICODE_COMBINING_CLASS_NOT_REORDERED */
+  1, /* HB_UNICODE_COMBINING_CLASS_OVERLAY */
+  2, 3, 4, 5, 6,
+  7, /* HB_UNICODE_COMBINING_CLASS_NUKTA */
+  8, /* HB_UNICODE_COMBINING_CLASS_KANA_VOICING */
+  9, /* HB_UNICODE_COMBINING_CLASS_VIRAMA */
+
+  /* Hebrew */
+  HB_MODIFIED_COMBINING_CLASS_CCC10,
+  HB_MODIFIED_COMBINING_CLASS_CCC11,
+  HB_MODIFIED_COMBINING_CLASS_CCC12,
+  HB_MODIFIED_COMBINING_CLASS_CCC13,
+  HB_MODIFIED_COMBINING_CLASS_CCC14,
+  HB_MODIFIED_COMBINING_CLASS_CCC15,
+  HB_MODIFIED_COMBINING_CLASS_CCC16,
+  HB_MODIFIED_COMBINING_CLASS_CCC17,
+  HB_MODIFIED_COMBINING_CLASS_CCC18,
+  HB_MODIFIED_COMBINING_CLASS_CCC19,
+  HB_MODIFIED_COMBINING_CLASS_CCC20,
+  HB_MODIFIED_COMBINING_CLASS_CCC21,
+  HB_MODIFIED_COMBINING_CLASS_CCC22,
+  HB_MODIFIED_COMBINING_CLASS_CCC23,
+  HB_MODIFIED_COMBINING_CLASS_CCC24,
+  HB_MODIFIED_COMBINING_CLASS_CCC25,
+  HB_MODIFIED_COMBINING_CLASS_CCC26,
+
+  /* Arabic */
+  HB_MODIFIED_COMBINING_CLASS_CCC27,
+  HB_MODIFIED_COMBINING_CLASS_CCC28,
+  HB_MODIFIED_COMBINING_CLASS_CCC29,
+  HB_MODIFIED_COMBINING_CLASS_CCC30,
+  HB_MODIFIED_COMBINING_CLASS_CCC31,
+  HB_MODIFIED_COMBINING_CLASS_CCC32,
+  HB_MODIFIED_COMBINING_CLASS_CCC33,
+  HB_MODIFIED_COMBINING_CLASS_CCC34,
+  HB_MODIFIED_COMBINING_CLASS_CCC35,
+
+  /* Syriac */
+  HB_MODIFIED_COMBINING_CLASS_CCC36,
+
+  37, 38, 39,
+  40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59,
+  60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79,
+  80, 81, 82, 83,
+
+  /* Telugu */
+  HB_MODIFIED_COMBINING_CLASS_CCC84,
+  85, 86, 87, 88, 89, 90,
+  HB_MODIFIED_COMBINING_CLASS_CCC91,
+  92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102,
+
+  /* Thai */
+  HB_MODIFIED_COMBINING_CLASS_CCC103,
+  104, 105, 106,
+  HB_MODIFIED_COMBINING_CLASS_CCC107,
+  108, 109, 110, 111, 112, 113, 114, 115, 116, 117,
+
+  /* Lao */
+  HB_MODIFIED_COMBINING_CLASS_CCC118,
+  119, 120, 121,
+  HB_MODIFIED_COMBINING_CLASS_CCC122,
+  123, 124, 125, 126, 127, 128,
+
+  /* Tibetan */
+  HB_MODIFIED_COMBINING_CLASS_CCC129,
+  HB_MODIFIED_COMBINING_CLASS_CCC130,
+  131,
+  HB_MODIFIED_COMBINING_CLASS_CCC132,
+  133, 134, 135, 136, 137, 138, 139,
+
+
+  140, 141, 142, 143, 144, 145, 146, 147, 148, 149,
+  150, 151, 152, 153, 154, 155, 156, 157, 158, 159,
+  160, 161, 162, 163, 164, 165, 166, 167, 168, 169,
+  170, 171, 172, 173, 174, 175, 176, 177, 178, 179,
+  180, 181, 182, 183, 184, 185, 186, 187, 188, 189,
+  190, 191, 192, 193, 194, 195, 196, 197, 198, 199,
+
+  200, /* HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW_LEFT */
+  201,
+  202, /* HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW */
+  203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213,
+  214, /* HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE */
+  215,
+  216, /* HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE_RIGHT */
+  217,
+  218, /* HB_UNICODE_COMBINING_CLASS_BELOW_LEFT */
+  219,
+  220, /* HB_UNICODE_COMBINING_CLASS_BELOW */
+  221,
+  222, /* HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT */
+  223,
+  224, /* HB_UNICODE_COMBINING_CLASS_LEFT */
+  225,
+  226, /* HB_UNICODE_COMBINING_CLASS_RIGHT */
+  227,
+  228, /* HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT */
+  229,
+  230, /* HB_UNICODE_COMBINING_CLASS_ABOVE */
+  231,
+  232, /* HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT */
+  233, /* HB_UNICODE_COMBINING_CLASS_DOUBLE_BELOW */
+  234, /* HB_UNICODE_COMBINING_CLASS_DOUBLE_ABOVE */
+  235, 236, 237, 238, 239,
+  240, /* HB_UNICODE_COMBINING_CLASS_IOTA_SUBSCRIPT */
+  241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254,
+  255, /* HB_UNICODE_COMBINING_CLASS_INVALID */
+};
+#endif
+
+
+/*
+ * Emoji
+ */
+#ifndef HB_NO_EMOJI_SEQUENCES
+
+#include "hb-unicode-emoji-table.hh"
+
+bool
+_hb_unicode_is_emoji_Extended_Pictographic (hb_codepoint_t cp)
+{
+  return _hb_emoji_is_Extended_Pictographic (cp);
+}
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-unicode.h b/thirdparty/harfbuzz/src/hb-unicode.h
new file mode 100644
index 0000000000..61b1b0ba1f
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-unicode.h
@@ -0,0 +1,404 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Codethink Limited
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Codethink Author(s): Ryan Lortie
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_UNICODE_H
+#define HB_UNICODE_H
+
+#include "hb-common.h"
+
+HB_BEGIN_DECLS
+
+
+/**
+ * HB_UNICODE_MAX
+ *
+ * Since: 1.9.0
+ **/
+#define HB_UNICODE_MAX 0x10FFFFu
+
+
+/* hb_unicode_general_category_t */
+
+/* Unicode Character Database property: General_Category (gc) */
+typedef enum
+{
+  HB_UNICODE_GENERAL_CATEGORY_CONTROL,			/* Cc */
+  HB_UNICODE_GENERAL_CATEGORY_FORMAT,			/* Cf */
+  HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED,		/* Cn */
+  HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE,		/* Co */
+  HB_UNICODE_GENERAL_CATEGORY_SURROGATE,		/* Cs */
+  HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER,		/* Ll */
+  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER,		/* Lm */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER,		/* Lo */
+  HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER,		/* Lt */
+  HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER,		/* Lu */
+  HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK,		/* Mc */
+  HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK,		/* Me */
+  HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK,		/* Mn */
+  HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER,		/* Nd */
+  HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER,		/* Nl */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER,		/* No */
+  HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION,	/* Pc */
+  HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION,		/* Pd */
+  HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION,	/* Pe */
+  HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION,	/* Pf */
+  HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION,	/* Pi */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION,	/* Po */
+  HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION,		/* Ps */
+  HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL,		/* Sc */
+  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL,		/* Sk */
+  HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL,		/* Sm */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL,		/* So */
+  HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR,		/* Zl */
+  HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR,	/* Zp */
+  HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR		/* Zs */
+} hb_unicode_general_category_t;
+
+/* hb_unicode_combining_class_t */
+
+/* Note: newer versions of Unicode may add new values.  Clients should be ready to handle
+ * any value in the 0..254 range being returned from hb_unicode_combining_class().
+ */
+
+/* Unicode Character Database property: Canonical_Combining_Class (ccc) */
+typedef enum
+{
+  HB_UNICODE_COMBINING_CLASS_NOT_REORDERED	= 0,
+  HB_UNICODE_COMBINING_CLASS_OVERLAY		= 1,
+  HB_UNICODE_COMBINING_CLASS_NUKTA		= 7,
+  HB_UNICODE_COMBINING_CLASS_KANA_VOICING	= 8,
+  HB_UNICODE_COMBINING_CLASS_VIRAMA		= 9,
+
+  /* Hebrew */
+  HB_UNICODE_COMBINING_CLASS_CCC10	=  10,
+  HB_UNICODE_COMBINING_CLASS_CCC11	=  11,
+  HB_UNICODE_COMBINING_CLASS_CCC12	=  12,
+  HB_UNICODE_COMBINING_CLASS_CCC13	=  13,
+  HB_UNICODE_COMBINING_CLASS_CCC14	=  14,
+  HB_UNICODE_COMBINING_CLASS_CCC15	=  15,
+  HB_UNICODE_COMBINING_CLASS_CCC16	=  16,
+  HB_UNICODE_COMBINING_CLASS_CCC17	=  17,
+  HB_UNICODE_COMBINING_CLASS_CCC18	=  18,
+  HB_UNICODE_COMBINING_CLASS_CCC19	=  19,
+  HB_UNICODE_COMBINING_CLASS_CCC20	=  20,
+  HB_UNICODE_COMBINING_CLASS_CCC21	=  21,
+  HB_UNICODE_COMBINING_CLASS_CCC22	=  22,
+  HB_UNICODE_COMBINING_CLASS_CCC23	=  23,
+  HB_UNICODE_COMBINING_CLASS_CCC24	=  24,
+  HB_UNICODE_COMBINING_CLASS_CCC25	=  25,
+  HB_UNICODE_COMBINING_CLASS_CCC26	=  26,
+
+  /* Arabic */
+  HB_UNICODE_COMBINING_CLASS_CCC27	=  27,
+  HB_UNICODE_COMBINING_CLASS_CCC28	=  28,
+  HB_UNICODE_COMBINING_CLASS_CCC29	=  29,
+  HB_UNICODE_COMBINING_CLASS_CCC30	=  30,
+  HB_UNICODE_COMBINING_CLASS_CCC31	=  31,
+  HB_UNICODE_COMBINING_CLASS_CCC32	=  32,
+  HB_UNICODE_COMBINING_CLASS_CCC33	=  33,
+  HB_UNICODE_COMBINING_CLASS_CCC34	=  34,
+  HB_UNICODE_COMBINING_CLASS_CCC35	=  35,
+
+  /* Syriac */
+  HB_UNICODE_COMBINING_CLASS_CCC36	=  36,
+
+  /* Telugu */
+  HB_UNICODE_COMBINING_CLASS_CCC84	=  84,
+  HB_UNICODE_COMBINING_CLASS_CCC91	=  91,
+
+  /* Thai */
+  HB_UNICODE_COMBINING_CLASS_CCC103	= 103,
+  HB_UNICODE_COMBINING_CLASS_CCC107	= 107,
+
+  /* Lao */
+  HB_UNICODE_COMBINING_CLASS_CCC118	= 118,
+  HB_UNICODE_COMBINING_CLASS_CCC122	= 122,
+
+  /* Tibetan */
+  HB_UNICODE_COMBINING_CLASS_CCC129	= 129,
+  HB_UNICODE_COMBINING_CLASS_CCC130	= 130,
+  HB_UNICODE_COMBINING_CLASS_CCC133	= 132,
+
+
+  HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW_LEFT	= 200,
+  HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW		= 202,
+  HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE		= 214,
+  HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE_RIGHT	= 216,
+  HB_UNICODE_COMBINING_CLASS_BELOW_LEFT			= 218,
+  HB_UNICODE_COMBINING_CLASS_BELOW			= 220,
+  HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT		= 222,
+  HB_UNICODE_COMBINING_CLASS_LEFT			= 224,
+  HB_UNICODE_COMBINING_CLASS_RIGHT			= 226,
+  HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT			= 228,
+  HB_UNICODE_COMBINING_CLASS_ABOVE			= 230,
+  HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT		= 232,
+  HB_UNICODE_COMBINING_CLASS_DOUBLE_BELOW		= 233,
+  HB_UNICODE_COMBINING_CLASS_DOUBLE_ABOVE		= 234,
+
+  HB_UNICODE_COMBINING_CLASS_IOTA_SUBSCRIPT		= 240,
+
+  HB_UNICODE_COMBINING_CLASS_INVALID	= 255
+} hb_unicode_combining_class_t;
+
+
+/*
+ * hb_unicode_funcs_t
+ */
+
+typedef struct hb_unicode_funcs_t hb_unicode_funcs_t;
+
+
+/*
+ * just give me the best implementation you've got there.
+ */
+HB_EXTERN hb_unicode_funcs_t *
+hb_unicode_funcs_get_default (void);
+
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_unicode_funcs_create (hb_unicode_funcs_t *parent);
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_unicode_funcs_get_empty (void);
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_unicode_funcs_reference (hb_unicode_funcs_t *ufuncs);
+
+HB_EXTERN void
+hb_unicode_funcs_destroy (hb_unicode_funcs_t *ufuncs);
+
+HB_EXTERN hb_bool_t
+hb_unicode_funcs_set_user_data (hb_unicode_funcs_t *ufuncs,
+				hb_user_data_key_t *key,
+				void *              data,
+				hb_destroy_func_t   destroy,
+				hb_bool_t           replace);
+
+
+HB_EXTERN void *
+hb_unicode_funcs_get_user_data (hb_unicode_funcs_t *ufuncs,
+				hb_user_data_key_t *key);
+
+
+HB_EXTERN void
+hb_unicode_funcs_make_immutable (hb_unicode_funcs_t *ufuncs);
+
+HB_EXTERN hb_bool_t
+hb_unicode_funcs_is_immutable (hb_unicode_funcs_t *ufuncs);
+
+HB_EXTERN hb_unicode_funcs_t *
+hb_unicode_funcs_get_parent (hb_unicode_funcs_t *ufuncs);
+
+
+/*
+ * funcs
+ */
+
+/* typedefs */
+
+typedef hb_unicode_combining_class_t	(*hb_unicode_combining_class_func_t)	(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      unicode,
+										 void               *user_data);
+typedef hb_unicode_general_category_t	(*hb_unicode_general_category_func_t)	(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      unicode,
+										 void               *user_data);
+typedef hb_codepoint_t			(*hb_unicode_mirroring_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      unicode,
+										 void               *user_data);
+typedef hb_script_t			(*hb_unicode_script_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      unicode,
+										 void               *user_data);
+
+typedef hb_bool_t			(*hb_unicode_compose_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      a,
+										 hb_codepoint_t      b,
+										 hb_codepoint_t     *ab,
+										 void               *user_data);
+typedef hb_bool_t			(*hb_unicode_decompose_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      ab,
+										 hb_codepoint_t     *a,
+										 hb_codepoint_t     *b,
+										 void               *user_data);
+
+/* setters */
+
+/**
+ * hb_unicode_funcs_set_combining_class_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_combining_class_func (hb_unicode_funcs_t *ufuncs,
+					   hb_unicode_combining_class_func_t func,
+					   void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_funcs_set_general_category_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_general_category_func (hb_unicode_funcs_t *ufuncs,
+					    hb_unicode_general_category_func_t func,
+					    void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_funcs_set_mirroring_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_mirroring_func (hb_unicode_funcs_t *ufuncs,
+				     hb_unicode_mirroring_func_t func,
+				     void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_funcs_set_script_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_script_func (hb_unicode_funcs_t *ufuncs,
+				  hb_unicode_script_func_t func,
+				  void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_funcs_set_compose_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_compose_func (hb_unicode_funcs_t *ufuncs,
+				   hb_unicode_compose_func_t func,
+				   void *user_data, hb_destroy_func_t destroy);
+
+/**
+ * hb_unicode_funcs_set_decompose_func:
+ * @ufuncs: a Unicode function structure
+ * @func: (closure user_data) (destroy destroy) (scope notified):
+ * @user_data:
+ * @destroy:
+ *
+ *
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN void
+hb_unicode_funcs_set_decompose_func (hb_unicode_funcs_t *ufuncs,
+				     hb_unicode_decompose_func_t func,
+				     void *user_data, hb_destroy_func_t destroy);
+
+/* accessors */
+
+/**
+ * hb_unicode_combining_class:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN hb_unicode_combining_class_t
+hb_unicode_combining_class (hb_unicode_funcs_t *ufuncs,
+			    hb_codepoint_t unicode);
+
+/**
+ * hb_unicode_general_category:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN hb_unicode_general_category_t
+hb_unicode_general_category (hb_unicode_funcs_t *ufuncs,
+			     hb_codepoint_t unicode);
+
+/**
+ * hb_unicode_mirroring:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN hb_codepoint_t
+hb_unicode_mirroring (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t unicode);
+
+/**
+ * hb_unicode_script:
+ *
+ * Since: 0.9.2
+ **/
+HB_EXTERN hb_script_t
+hb_unicode_script (hb_unicode_funcs_t *ufuncs,
+		   hb_codepoint_t unicode);
+
+HB_EXTERN hb_bool_t
+hb_unicode_compose (hb_unicode_funcs_t *ufuncs,
+		    hb_codepoint_t      a,
+		    hb_codepoint_t      b,
+		    hb_codepoint_t     *ab);
+
+HB_EXTERN hb_bool_t
+hb_unicode_decompose (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t      ab,
+		      hb_codepoint_t     *a,
+		      hb_codepoint_t     *b);
+
+HB_END_DECLS
+
+#endif /* HB_UNICODE_H */
diff --git a/thirdparty/harfbuzz/src/hb-unicode.hh b/thirdparty/harfbuzz/src/hb-unicode.hh
new file mode 100644
index 0000000000..34d66d7aa3
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-unicode.hh
@@ -0,0 +1,398 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ * Copyright © 2011  Codethink Limited
+ * Copyright © 2010,2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Codethink Author(s): Ryan Lortie
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_UNICODE_HH
+#define HB_UNICODE_HH
+
+#include "hb.hh"
+
+
+extern HB_INTERNAL const uint8_t _hb_modified_combining_class[256];
+
+/*
+ * hb_unicode_funcs_t
+ */
+
+#define HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS \
+  HB_UNICODE_FUNC_IMPLEMENT (combining_class) \
+  HB_IF_NOT_DEPRECATED (HB_UNICODE_FUNC_IMPLEMENT (eastasian_width)) \
+  HB_UNICODE_FUNC_IMPLEMENT (general_category) \
+  HB_UNICODE_FUNC_IMPLEMENT (mirroring) \
+  HB_UNICODE_FUNC_IMPLEMENT (script) \
+  HB_UNICODE_FUNC_IMPLEMENT (compose) \
+  HB_UNICODE_FUNC_IMPLEMENT (decompose) \
+  HB_IF_NOT_DEPRECATED (HB_UNICODE_FUNC_IMPLEMENT (decompose_compatibility)) \
+  /* ^--- Add new callbacks here */
+
+/* Simple callbacks are those taking a hb_codepoint_t and returning a hb_codepoint_t */
+#define HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE \
+  HB_UNICODE_FUNC_IMPLEMENT (hb_unicode_combining_class_t, combining_class) \
+  HB_IF_NOT_DEPRECATED (HB_UNICODE_FUNC_IMPLEMENT (unsigned int, eastasian_width)) \
+  HB_UNICODE_FUNC_IMPLEMENT (hb_unicode_general_category_t, general_category) \
+  HB_UNICODE_FUNC_IMPLEMENT (hb_codepoint_t, mirroring) \
+  HB_UNICODE_FUNC_IMPLEMENT (hb_script_t, script) \
+  /* ^--- Add new simple callbacks here */
+
+struct hb_unicode_funcs_t
+{
+  hb_object_header_t header;
+
+  hb_unicode_funcs_t *parent;
+
+#define HB_UNICODE_FUNC_IMPLEMENT(return_type, name) \
+  return_type name (hb_codepoint_t unicode) { return func.name (this, unicode, user_data.name); }
+HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE
+#undef HB_UNICODE_FUNC_IMPLEMENT
+
+  hb_bool_t compose (hb_codepoint_t a, hb_codepoint_t b,
+		     hb_codepoint_t *ab)
+  {
+    *ab = 0;
+    if (unlikely (!a || !b)) return false;
+    return func.compose (this, a, b, ab, user_data.compose);
+  }
+
+  hb_bool_t decompose (hb_codepoint_t ab,
+		       hb_codepoint_t *a, hb_codepoint_t *b)
+  {
+    *a = ab; *b = 0;
+    return func.decompose (this, ab, a, b, user_data.decompose);
+  }
+
+  unsigned int decompose_compatibility (hb_codepoint_t  u,
+					hb_codepoint_t *decomposed)
+  {
+#ifdef HB_DISABLE_DEPRECATED
+    unsigned int ret  = 0;
+#else
+    unsigned int ret = func.decompose_compatibility (this, u, decomposed, user_data.decompose_compatibility);
+#endif
+    if (ret == 1 && u == decomposed[0]) {
+      decomposed[0] = 0;
+      return 0;
+    }
+    decomposed[ret] = 0;
+    return ret;
+  }
+
+  unsigned int
+  modified_combining_class (hb_codepoint_t u)
+  {
+    /* XXX This hack belongs to the USE shaper (for Tai Tham):
+     * Reorder SAKOT to ensure it comes after any tone marks. */
+    if (unlikely (u == 0x1A60u)) return 254;
+
+    /* XXX This hack belongs to the Tibetan shaper:
+     * Reorder PADMA to ensure it comes after any vowel marks. */
+    if (unlikely (u == 0x0FC6u)) return 254;
+    /* Reorder TSA -PHRU to reorder before U+0F74 */
+    if (unlikely (u == 0x0F39u)) return 127;
+
+    return _hb_modified_combining_class[combining_class (u)];
+  }
+
+  static hb_bool_t
+  is_variation_selector (hb_codepoint_t unicode)
+  {
+    /* U+180B..180D MONGOLIAN FREE VARIATION SELECTORs are handled in the
+     * Arabic shaper.  No need to match them here. */
+    return unlikely (hb_in_ranges<hb_codepoint_t> (unicode,
+						   0xFE00u, 0xFE0Fu, /* VARIATION SELECTOR-1..16 */
+						   0xE0100u, 0xE01EFu));  /* VARIATION SELECTOR-17..256 */
+  }
+
+  /* Default_Ignorable codepoints:
+   *
+   * Note: While U+115F, U+1160, U+3164 and U+FFA0 are Default_Ignorable,
+   * we do NOT want to hide them, as the way Uniscribe has implemented them
+   * is with regular spacing glyphs, and that's the way fonts are made to work.
+   * As such, we make exceptions for those four.
+   * Also ignoring U+1BCA0..1BCA3. https://github.com/harfbuzz/harfbuzz/issues/503
+   *
+   * Unicode 7.0:
+   * $ grep '; Default_Ignorable_Code_Point ' DerivedCoreProperties.txt | sed 's/;.*#/#/'
+   * 00AD          # Cf       SOFT HYPHEN
+   * 034F          # Mn       COMBINING GRAPHEME JOINER
+   * 061C          # Cf       ARABIC LETTER MARK
+   * 115F..1160    # Lo   [2] HANGUL CHOSEONG FILLER..HANGUL JUNGSEONG FILLER
+   * 17B4..17B5    # Mn   [2] KHMER VOWEL INHERENT AQ..KHMER VOWEL INHERENT AA
+   * 180B..180D    # Mn   [3] MONGOLIAN FREE VARIATION SELECTOR ONE..MONGOLIAN FREE VARIATION SELECTOR THREE
+   * 180E          # Cf       MONGOLIAN VOWEL SEPARATOR
+   * 200B..200F    # Cf   [5] ZERO WIDTH SPACE..RIGHT-TO-LEFT MARK
+   * 202A..202E    # Cf   [5] LEFT-TO-RIGHT EMBEDDING..RIGHT-TO-LEFT OVERRIDE
+   * 2060..2064    # Cf   [5] WORD JOINER..INVISIBLE PLUS
+   * 2065          # Cn       <reserved-2065>
+   * 2066..206F    # Cf  [10] LEFT-TO-RIGHT ISOLATE..NOMINAL DIGIT SHAPES
+   * 3164          # Lo       HANGUL FILLER
+   * FE00..FE0F    # Mn  [16] VARIATION SELECTOR-1..VARIATION SELECTOR-16
+   * FEFF          # Cf       ZERO WIDTH NO-BREAK SPACE
+   * FFA0          # Lo       HALFWIDTH HANGUL FILLER
+   * FFF0..FFF8    # Cn   [9] <reserved-FFF0>..<reserved-FFF8>
+   * 1BCA0..1BCA3  # Cf   [4] SHORTHAND FORMAT LETTER OVERLAP..SHORTHAND FORMAT UP STEP
+   * 1D173..1D17A  # Cf   [8] MUSICAL SYMBOL BEGIN BEAM..MUSICAL SYMBOL END PHRASE
+   * E0000         # Cn       <reserved-E0000>
+   * E0001         # Cf       LANGUAGE TAG
+   * E0002..E001F  # Cn  [30] <reserved-E0002>..<reserved-E001F>
+   * E0020..E007F  # Cf  [96] TAG SPACE..CANCEL TAG
+   * E0080..E00FF  # Cn [128] <reserved-E0080>..<reserved-E00FF>
+   * E0100..E01EF  # Mn [240] VARIATION SELECTOR-17..VARIATION SELECTOR-256
+   * E01F0..E0FFF  # Cn [3600] <reserved-E01F0>..<reserved-E0FFF>
+   */
+  static hb_bool_t
+  is_default_ignorable (hb_codepoint_t ch)
+  {
+    hb_codepoint_t plane = ch >> 16;
+    if (likely (plane == 0))
+    {
+      /* BMP */
+      hb_codepoint_t page = ch >> 8;
+      switch (page) {
+	case 0x00: return unlikely (ch == 0x00ADu);
+	case 0x03: return unlikely (ch == 0x034Fu);
+	case 0x06: return unlikely (ch == 0x061Cu);
+	case 0x17: return hb_in_range<hb_codepoint_t> (ch, 0x17B4u, 0x17B5u);
+	case 0x18: return hb_in_range<hb_codepoint_t> (ch, 0x180Bu, 0x180Eu);
+	case 0x20: return hb_in_ranges<hb_codepoint_t> (ch, 0x200Bu, 0x200Fu,
+					    0x202Au, 0x202Eu,
+					    0x2060u, 0x206Fu);
+	case 0xFE: return hb_in_range<hb_codepoint_t> (ch, 0xFE00u, 0xFE0Fu) || ch == 0xFEFFu;
+	case 0xFF: return hb_in_range<hb_codepoint_t> (ch, 0xFFF0u, 0xFFF8u);
+	default: return false;
+      }
+    }
+    else
+    {
+      /* Other planes */
+      switch (plane) {
+	case 0x01: return hb_in_range<hb_codepoint_t> (ch, 0x1D173u, 0x1D17Au);
+	case 0x0E: return hb_in_range<hb_codepoint_t> (ch, 0xE0000u, 0xE0FFFu);
+	default: return false;
+      }
+    }
+  }
+
+  /* Space estimates based on:
+   * https://unicode.org/charts/PDF/U2000.pdf
+   * https://docs.microsoft.com/en-us/typography/develop/character-design-standards/whitespace
+   */
+  enum space_t {
+    NOT_SPACE = 0,
+    SPACE_EM   = 1,
+    SPACE_EM_2 = 2,
+    SPACE_EM_3 = 3,
+    SPACE_EM_4 = 4,
+    SPACE_EM_5 = 5,
+    SPACE_EM_6 = 6,
+    SPACE_EM_16 = 16,
+    SPACE_4_EM_18,	/* 4/18th of an EM! */
+    SPACE,
+    SPACE_FIGURE,
+    SPACE_PUNCTUATION,
+    SPACE_NARROW,
+  };
+  static space_t
+  space_fallback_type (hb_codepoint_t u)
+  {
+    switch (u)
+    {
+      /* All GC=Zs chars that can use a fallback. */
+      default:	    return NOT_SPACE;	/* U+1680 OGHAM SPACE MARK */
+      case 0x0020u: return SPACE;	/* U+0020 SPACE */
+      case 0x00A0u: return SPACE;	/* U+00A0 NO-BREAK SPACE */
+      case 0x2000u: return SPACE_EM_2;	/* U+2000 EN QUAD */
+      case 0x2001u: return SPACE_EM;	/* U+2001 EM QUAD */
+      case 0x2002u: return SPACE_EM_2;	/* U+2002 EN SPACE */
+      case 0x2003u: return SPACE_EM;	/* U+2003 EM SPACE */
+      case 0x2004u: return SPACE_EM_3;	/* U+2004 THREE-PER-EM SPACE */
+      case 0x2005u: return SPACE_EM_4;	/* U+2005 FOUR-PER-EM SPACE */
+      case 0x2006u: return SPACE_EM_6;	/* U+2006 SIX-PER-EM SPACE */
+      case 0x2007u: return SPACE_FIGURE;	/* U+2007 FIGURE SPACE */
+      case 0x2008u: return SPACE_PUNCTUATION;	/* U+2008 PUNCTUATION SPACE */
+      case 0x2009u: return SPACE_EM_5;		/* U+2009 THIN SPACE */
+      case 0x200Au: return SPACE_EM_16;		/* U+200A HAIR SPACE */
+      case 0x202Fu: return SPACE_NARROW;	/* U+202F NARROW NO-BREAK SPACE */
+      case 0x205Fu: return SPACE_4_EM_18;	/* U+205F MEDIUM MATHEMATICAL SPACE */
+      case 0x3000u: return SPACE_EM;		/* U+3000 IDEOGRAPHIC SPACE */
+    }
+  }
+
+  struct {
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_##name##_func_t name;
+    HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+  } func;
+
+  struct {
+#define HB_UNICODE_FUNC_IMPLEMENT(name) void *name;
+    HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+  } user_data;
+
+  struct {
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_destroy_func_t name;
+    HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+  } destroy;
+};
+DECLARE_NULL_INSTANCE (hb_unicode_funcs_t);
+
+
+/*
+ * Modified combining marks
+ */
+
+/* Hebrew
+ *
+ * We permute the "fixed-position" classes 10-26 into the order
+ * described in the SBL Hebrew manual:
+ *
+ * https://www.sbl-site.org/Fonts/SBLHebrewUserManual1.5x.pdf
+ *
+ * (as recommended by:
+ *  https://forum.fontlab.com/archive-old-microsoft-volt-group/vista-and-diacritic-ordering/msg22823/)
+ *
+ * More details here:
+ * https://bugzilla.mozilla.org/show_bug.cgi?id=662055
+ */
+#define HB_MODIFIED_COMBINING_CLASS_CCC10 22 /* sheva */
+#define HB_MODIFIED_COMBINING_CLASS_CCC11 15 /* hataf segol */
+#define HB_MODIFIED_COMBINING_CLASS_CCC12 16 /* hataf patah */
+#define HB_MODIFIED_COMBINING_CLASS_CCC13 17 /* hataf qamats */
+#define HB_MODIFIED_COMBINING_CLASS_CCC14 23 /* hiriq */
+#define HB_MODIFIED_COMBINING_CLASS_CCC15 18 /* tsere */
+#define HB_MODIFIED_COMBINING_CLASS_CCC16 19 /* segol */
+#define HB_MODIFIED_COMBINING_CLASS_CCC17 20 /* patah */
+#define HB_MODIFIED_COMBINING_CLASS_CCC18 21 /* qamats */
+#define HB_MODIFIED_COMBINING_CLASS_CCC19 14 /* holam */
+#define HB_MODIFIED_COMBINING_CLASS_CCC20 24 /* qubuts */
+#define HB_MODIFIED_COMBINING_CLASS_CCC21 12 /* dagesh */
+#define HB_MODIFIED_COMBINING_CLASS_CCC22 25 /* meteg */
+#define HB_MODIFIED_COMBINING_CLASS_CCC23 13 /* rafe */
+#define HB_MODIFIED_COMBINING_CLASS_CCC24 10 /* shin dot */
+#define HB_MODIFIED_COMBINING_CLASS_CCC25 11 /* sin dot */
+#define HB_MODIFIED_COMBINING_CLASS_CCC26 26 /* point varika */
+
+/*
+ * Arabic
+ *
+ * Modify to move Shadda (ccc=33) before other marks.  See:
+ * https://unicode.org/faq/normalization.html#8
+ * https://unicode.org/faq/normalization.html#9
+ */
+#define HB_MODIFIED_COMBINING_CLASS_CCC27 28 /* fathatan */
+#define HB_MODIFIED_COMBINING_CLASS_CCC28 29 /* dammatan */
+#define HB_MODIFIED_COMBINING_CLASS_CCC29 30 /* kasratan */
+#define HB_MODIFIED_COMBINING_CLASS_CCC30 31 /* fatha */
+#define HB_MODIFIED_COMBINING_CLASS_CCC31 32 /* damma */
+#define HB_MODIFIED_COMBINING_CLASS_CCC32 33 /* kasra */
+#define HB_MODIFIED_COMBINING_CLASS_CCC33 27 /* shadda */
+#define HB_MODIFIED_COMBINING_CLASS_CCC34 34 /* sukun */
+#define HB_MODIFIED_COMBINING_CLASS_CCC35 35 /* superscript alef */
+
+/* Syriac */
+#define HB_MODIFIED_COMBINING_CLASS_CCC36 36 /* superscript alaph */
+
+/* Telugu
+ *
+ * Modify Telugu length marks (ccc=84, ccc=91).
+ * These are the only matras in the main Indic scripts range that have
+ * a non-zero ccc.  That makes them reorder with the Halant (ccc=9).
+ * Assign 4 and 5, which are otherwise unassigned.
+ */
+#define HB_MODIFIED_COMBINING_CLASS_CCC84 4 /* length mark */
+#define HB_MODIFIED_COMBINING_CLASS_CCC91 5 /* ai length mark */
+
+/* Thai
+ *
+ * Modify U+0E38 and U+0E39 (ccc=103) to be reordered before U+0E3A (ccc=9).
+ * Assign 3, which is unassigned otherwise.
+ * Uniscribe does this reordering too.
+ */
+#define HB_MODIFIED_COMBINING_CLASS_CCC103 3 /* sara u / sara uu */
+#define HB_MODIFIED_COMBINING_CLASS_CCC107 107 /* mai * */
+
+/* Lao */
+#define HB_MODIFIED_COMBINING_CLASS_CCC118 118 /* sign u / sign uu */
+#define HB_MODIFIED_COMBINING_CLASS_CCC122 122 /* mai * */
+
+/* Tibetan
+ *
+ * In case of multiple vowel-signs, use u first (but after achung)
+ * this allows Dzongkha multi-vowel shortcuts to render correctly
+ */
+#define HB_MODIFIED_COMBINING_CLASS_CCC129 129 /* sign aa */
+#define HB_MODIFIED_COMBINING_CLASS_CCC130 132 /* sign i */
+#define HB_MODIFIED_COMBINING_CLASS_CCC132 131 /* sign u */
+
+/* Misc */
+
+#define HB_UNICODE_GENERAL_CATEGORY_IS_MARK(gen_cat) \
+	(FLAG_UNSAFE (gen_cat) & \
+	 (FLAG (HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK) | \
+	  FLAG (HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)))
+
+
+/*
+ * Ranges, used for bsearch tables.
+ */
+
+struct hb_unicode_range_t
+{
+  static int
+  cmp (const void *_key, const void *_item)
+  {
+    hb_codepoint_t cp = *((hb_codepoint_t *) _key);
+    const hb_unicode_range_t *range = (hb_unicode_range_t *) _item;
+
+    if (cp < range->start)
+      return -1;
+    else if (cp <= range->end)
+      return 0;
+    else
+      return +1;
+  }
+
+  hb_codepoint_t start;
+  hb_codepoint_t end;
+};
+
+/*
+ * Emoji.
+ */
+
+HB_INTERNAL bool
+_hb_unicode_is_emoji_Extended_Pictographic (hb_codepoint_t cp);
+
+
+extern "C" HB_INTERNAL hb_unicode_funcs_t *hb_ucd_get_unicode_funcs ();
+
+
+#endif /* HB_UNICODE_HH */
diff --git a/thirdparty/harfbuzz/src/hb-uniscribe.cc b/thirdparty/harfbuzz/src/hb-uniscribe.cc
new file mode 100644
index 0000000000..48a5dc50ad
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-uniscribe.cc
@@ -0,0 +1,1047 @@
+/*
+ * Copyright © 2011,2012,2013  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#include "hb.hh"
+
+#ifdef HAVE_UNISCRIBE
+
+#ifdef HB_NO_OT_TAG
+#error "Cannot compile 'uniscribe' shaper with HB_NO_OT_TAG."
+#endif
+
+#include "hb-shaper-impl.hh"
+
+#include <windows.h>
+#include <usp10.h>
+#include <rpc.h>
+
+#ifndef E_NOT_SUFFICIENT_BUFFER
+#define E_NOT_SUFFICIENT_BUFFER HRESULT_FROM_WIN32 (ERROR_INSUFFICIENT_BUFFER)
+#endif
+
+#include "hb-uniscribe.h"
+
+#include "hb-open-file.hh"
+#include "hb-ot-name-table.hh"
+#include "hb-ot-layout.h"
+
+
+/**
+ * SECTION:hb-uniscribe
+ * @title: hb-uniscribe
+ * @short_description: Windows integration
+ * @include: hb-uniscribe.h
+ *
+ * Functions for using HarfBuzz with Windows fonts.
+ **/
+
+typedef HRESULT (WINAPI *SIOT) /*ScriptItemizeOpenType*/(
+  const WCHAR *pwcInChars,
+  int cInChars,
+  int cMaxItems,
+  const SCRIPT_CONTROL *psControl,
+  const SCRIPT_STATE *psState,
+  SCRIPT_ITEM *pItems,
+  OPENTYPE_TAG *pScriptTags,
+  int *pcItems
+);
+
+typedef HRESULT (WINAPI *SSOT) /*ScriptShapeOpenType*/(
+  HDC hdc,
+  SCRIPT_CACHE *psc,
+  SCRIPT_ANALYSIS *psa,
+  OPENTYPE_TAG tagScript,
+  OPENTYPE_TAG tagLangSys,
+  int *rcRangeChars,
+  TEXTRANGE_PROPERTIES **rpRangeProperties,
+  int cRanges,
+  const WCHAR *pwcChars,
+  int cChars,
+  int cMaxGlyphs,
+  WORD *pwLogClust,
+  SCRIPT_CHARPROP *pCharProps,
+  WORD *pwOutGlyphs,
+  SCRIPT_GLYPHPROP *pOutGlyphProps,
+  int *pcGlyphs
+);
+
+typedef HRESULT (WINAPI *SPOT) /*ScriptPlaceOpenType*/(
+  HDC hdc,
+  SCRIPT_CACHE *psc,
+  SCRIPT_ANALYSIS *psa,
+  OPENTYPE_TAG tagScript,
+  OPENTYPE_TAG tagLangSys,
+  int *rcRangeChars,
+  TEXTRANGE_PROPERTIES **rpRangeProperties,
+  int cRanges,
+  const WCHAR *pwcChars,
+  WORD *pwLogClust,
+  SCRIPT_CHARPROP *pCharProps,
+  int cChars,
+  const WORD *pwGlyphs,
+  const SCRIPT_GLYPHPROP *pGlyphProps,
+  int cGlyphs,
+  int *piAdvance,
+  GOFFSET *pGoffset,
+  ABC *pABC
+);
+
+
+/* Fallback implementations. */
+
+static HRESULT WINAPI
+hb_ScriptItemizeOpenType(
+  const WCHAR *pwcInChars,
+  int cInChars,
+  int cMaxItems,
+  const SCRIPT_CONTROL *psControl,
+  const SCRIPT_STATE *psState,
+  SCRIPT_ITEM *pItems,
+  OPENTYPE_TAG *pScriptTags,
+  int *pcItems
+)
+{
+{
+  return ScriptItemize (pwcInChars,
+			cInChars,
+			cMaxItems,
+			psControl,
+			psState,
+			pItems,
+			pcItems);
+}
+}
+
+static HRESULT WINAPI
+hb_ScriptShapeOpenType(
+  HDC hdc,
+  SCRIPT_CACHE *psc,
+  SCRIPT_ANALYSIS *psa,
+  OPENTYPE_TAG tagScript,
+  OPENTYPE_TAG tagLangSys,
+  int *rcRangeChars,
+  TEXTRANGE_PROPERTIES **rpRangeProperties,
+  int cRanges,
+  const WCHAR *pwcChars,
+  int cChars,
+  int cMaxGlyphs,
+  WORD *pwLogClust,
+  SCRIPT_CHARPROP *pCharProps,
+  WORD *pwOutGlyphs,
+  SCRIPT_GLYPHPROP *pOutGlyphProps,
+  int *pcGlyphs
+)
+{
+  SCRIPT_VISATTR *psva = (SCRIPT_VISATTR *) pOutGlyphProps;
+  return ScriptShape (hdc,
+		      psc,
+		      pwcChars,
+		      cChars,
+		      cMaxGlyphs,
+		      psa,
+		      pwOutGlyphs,
+		      pwLogClust,
+		      psva,
+		      pcGlyphs);
+}
+
+static HRESULT WINAPI
+hb_ScriptPlaceOpenType(
+  HDC hdc,
+  SCRIPT_CACHE *psc,
+  SCRIPT_ANALYSIS *psa,
+  OPENTYPE_TAG tagScript,
+  OPENTYPE_TAG tagLangSys,
+  int *rcRangeChars,
+  TEXTRANGE_PROPERTIES **rpRangeProperties,
+  int cRanges,
+  const WCHAR *pwcChars,
+  WORD *pwLogClust,
+  SCRIPT_CHARPROP *pCharProps,
+  int cChars,
+  const WORD *pwGlyphs,
+  const SCRIPT_GLYPHPROP *pGlyphProps,
+  int cGlyphs,
+  int *piAdvance,
+  GOFFSET *pGoffset,
+  ABC *pABC
+)
+{
+  SCRIPT_VISATTR *psva = (SCRIPT_VISATTR *) pGlyphProps;
+  return ScriptPlace (hdc,
+		      psc,
+		      pwGlyphs,
+		      cGlyphs,
+		      psva,
+		      psa,
+		      piAdvance,
+		      pGoffset,
+		      pABC);
+}
+
+
+struct hb_uniscribe_shaper_funcs_t
+{
+  SIOT ScriptItemizeOpenType;
+  SSOT ScriptShapeOpenType;
+  SPOT ScriptPlaceOpenType;
+
+  void init ()
+  {
+    HMODULE hinstLib;
+    this->ScriptItemizeOpenType = nullptr;
+    this->ScriptShapeOpenType   = nullptr;
+    this->ScriptPlaceOpenType   = nullptr;
+
+    hinstLib = GetModuleHandle (TEXT ("usp10.dll"));
+    if (hinstLib)
+    {
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wcast-function-type"
+      this->ScriptItemizeOpenType = (SIOT) GetProcAddress (hinstLib, "ScriptItemizeOpenType");
+      this->ScriptShapeOpenType   = (SSOT) GetProcAddress (hinstLib, "ScriptShapeOpenType");
+      this->ScriptPlaceOpenType   = (SPOT) GetProcAddress (hinstLib, "ScriptPlaceOpenType");
+#pragma GCC diagnostic pop
+    }
+    if (!this->ScriptItemizeOpenType ||
+	!this->ScriptShapeOpenType   ||
+	!this->ScriptPlaceOpenType)
+    {
+      DEBUG_MSG (UNISCRIBE, nullptr, "OpenType versions of functions not found; falling back.");
+      this->ScriptItemizeOpenType = hb_ScriptItemizeOpenType;
+      this->ScriptShapeOpenType   = hb_ScriptShapeOpenType;
+      this->ScriptPlaceOpenType   = hb_ScriptPlaceOpenType;
+    }
+  }
+};
+
+#if HB_USE_ATEXIT
+static void free_static_uniscribe_shaper_funcs ();
+#endif
+
+static struct hb_uniscribe_shaper_funcs_lazy_loader_t : hb_lazy_loader_t<hb_uniscribe_shaper_funcs_t,
+									 hb_uniscribe_shaper_funcs_lazy_loader_t>
+{
+  static hb_uniscribe_shaper_funcs_t *create ()
+  {
+    hb_uniscribe_shaper_funcs_t *funcs = (hb_uniscribe_shaper_funcs_t *) calloc (1, sizeof (hb_uniscribe_shaper_funcs_t));
+    if (unlikely (!funcs))
+      return nullptr;
+
+    funcs->init ();
+
+#if HB_USE_ATEXIT
+    atexit (free_static_uniscribe_shaper_funcs);
+#endif
+
+    return funcs;
+  }
+  static void destroy (hb_uniscribe_shaper_funcs_t *p)
+  {
+    free ((void *) p);
+  }
+  static hb_uniscribe_shaper_funcs_t *get_null ()
+  {
+    return nullptr;
+  }
+} static_uniscribe_shaper_funcs;
+
+#if HB_USE_ATEXIT
+static
+void free_static_uniscribe_shaper_funcs ()
+{
+  static_uniscribe_shaper_funcs.free_instance ();
+}
+#endif
+
+static hb_uniscribe_shaper_funcs_t *
+hb_uniscribe_shaper_get_funcs ()
+{
+  return static_uniscribe_shaper_funcs.get_unconst ();
+}
+
+
+struct active_feature_t {
+  OPENTYPE_FEATURE_RECORD rec;
+  unsigned int order;
+
+  HB_INTERNAL static int cmp (const void *pa, const void *pb) {
+    const active_feature_t *a = (const active_feature_t *) pa;
+    const active_feature_t *b = (const active_feature_t *) pb;
+    return a->rec.tagFeature < b->rec.tagFeature ? -1 : a->rec.tagFeature > b->rec.tagFeature ? 1 :
+	   a->order < b->order ? -1 : a->order > b->order ? 1 :
+	   a->rec.lParameter < b->rec.lParameter ? -1 : a->rec.lParameter > b->rec.lParameter ? 1 :
+	   0;
+  }
+  bool operator== (const active_feature_t *f)
+  { return cmp (this, f) == 0; }
+};
+
+struct feature_event_t {
+  unsigned int index;
+  bool start;
+  active_feature_t feature;
+
+  HB_INTERNAL static int cmp (const void *pa, const void *pb)
+  {
+    const feature_event_t *a = (const feature_event_t *) pa;
+    const feature_event_t *b = (const feature_event_t *) pb;
+    return a->index < b->index ? -1 : a->index > b->index ? 1 :
+	   a->start < b->start ? -1 : a->start > b->start ? 1 :
+	   active_feature_t::cmp (&a->feature, &b->feature);
+  }
+};
+
+struct range_record_t {
+  TEXTRANGE_PROPERTIES props;
+  unsigned int index_first; /* == start */
+  unsigned int index_last;  /* == end - 1 */
+};
+
+
+/*
+ * shaper face data
+ */
+
+struct hb_uniscribe_face_data_t {
+  HANDLE fh;
+  hb_uniscribe_shaper_funcs_t *funcs;
+  wchar_t face_name[LF_FACESIZE];
+};
+
+/* face_name should point to a wchar_t[LF_FACESIZE] object. */
+static void
+_hb_generate_unique_face_name (wchar_t *face_name, unsigned int *plen)
+{
+  /* We'll create a private name for the font from a UUID using a simple,
+   * somewhat base64-like encoding scheme */
+  const char *enc = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+-";
+  UUID id;
+  UuidCreate ((UUID*) &id);
+  static_assert ((2 + 3 * (16/2) < LF_FACESIZE), "");
+  unsigned int name_str_len = 0;
+  face_name[name_str_len++] = 'F';
+  face_name[name_str_len++] = '_';
+  unsigned char *p = (unsigned char *) &id;
+  for (unsigned int i = 0; i < 16; i += 2)
+  {
+    /* Spread the 16 bits from two bytes of the UUID across three chars of face_name,
+     * using the bits in groups of 5,5,6 to select chars from enc.
+     * This will generate 24 characters; with the 'F_' prefix we already provided,
+     * the name will be 26 chars (plus the NUL terminator), so will always fit within
+     * face_name (LF_FACESIZE = 32). */
+    face_name[name_str_len++] = enc[p[i] >> 3];
+    face_name[name_str_len++] = enc[((p[i] << 2) | (p[i + 1] >> 6)) & 0x1f];
+    face_name[name_str_len++] = enc[p[i + 1] & 0x3f];
+  }
+  face_name[name_str_len] = 0;
+  if (plen)
+    *plen = name_str_len;
+}
+
+/* Destroys blob. */
+static hb_blob_t *
+_hb_rename_font (hb_blob_t *blob, wchar_t *new_name)
+{
+  /* Create a copy of the font data, with the 'name' table replaced by a
+   * table that names the font with our private F_* name created above.
+   * For simplicity, we just append a new 'name' table and update the
+   * sfnt directory; the original table is left in place, but unused.
+   *
+   * The new table will contain just 5 name IDs: family, style, unique,
+   * full, PS. All of them point to the same name data with our unique name.
+   */
+
+  blob = hb_sanitize_context_t ().sanitize_blob<OT::OpenTypeFontFile> (blob);
+
+  unsigned int length, new_length, name_str_len;
+  const char *orig_sfnt_data = hb_blob_get_data (blob, &length);
+
+  _hb_generate_unique_face_name (new_name, &name_str_len);
+
+  static const uint16_t name_IDs[] = { 1, 2, 3, 4, 6 };
+
+  unsigned int name_table_length = OT::name::min_size +
+				   ARRAY_LENGTH (name_IDs) * OT::NameRecord::static_size +
+				   name_str_len * 2; /* for name data in UTF16BE form */
+  unsigned int padded_name_table_length = ((name_table_length + 3) & ~3);
+  unsigned int name_table_offset = (length + 3) & ~3;
+
+  new_length = name_table_offset + padded_name_table_length;
+  void *new_sfnt_data = calloc (1, new_length);
+  if (!new_sfnt_data)
+  {
+    hb_blob_destroy (blob);
+    return nullptr;
+  }
+
+  memcpy(new_sfnt_data, orig_sfnt_data, length);
+
+  OT::name &name = StructAtOffset<OT::name> (new_sfnt_data, name_table_offset);
+  name.format = 0;
+  name.count = ARRAY_LENGTH (name_IDs);
+  name.stringOffset = name.get_size ();
+  for (unsigned int i = 0; i < ARRAY_LENGTH (name_IDs); i++)
+  {
+    OT::NameRecord &record = name.nameRecordZ[i];
+    record.platformID = 3;
+    record.encodingID = 1;
+    record.languageID = 0x0409u; /* English */
+    record.nameID = name_IDs[i];
+    record.length = name_str_len * 2;
+    record.offset = 0;
+  }
+
+  /* Copy string data from new_name, converting wchar_t to UTF16BE. */
+  unsigned char *p = &StructAfter<unsigned char> (name);
+  for (unsigned int i = 0; i < name_str_len; i++)
+  {
+    *p++ = new_name[i] >> 8;
+    *p++ = new_name[i] & 0xff;
+  }
+
+  /* Adjust name table entry to point to new name table */
+  const OT::OpenTypeFontFile &file = * (OT::OpenTypeFontFile *) (new_sfnt_data);
+  unsigned int face_count = file.get_face_count ();
+  for (unsigned int face_index = 0; face_index < face_count; face_index++)
+  {
+    /* Note: doing multiple edits (ie. TTC) can be unsafe.  There may be
+     * toe-stepping.  But we don't really care. */
+    const OT::OpenTypeFontFace &face = file.get_face (face_index);
+    unsigned int index;
+    if (face.find_table_index (HB_OT_TAG_name, &index))
+    {
+      OT::TableRecord &record = const_cast<OT::TableRecord &> (face.get_table (index));
+      record.checkSum.set_for_data (&name, padded_name_table_length);
+      record.offset = name_table_offset;
+      record.length = name_table_length;
+    }
+    else if (face_index == 0) /* Fail if first face doesn't have 'name' table. */
+    {
+      free (new_sfnt_data);
+      hb_blob_destroy (blob);
+      return nullptr;
+    }
+  }
+
+  /* The checkSumAdjustment field in the 'head' table is now wrong,
+   * but that doesn't actually seem to cause any problems so we don't
+   * bother. */
+
+  hb_blob_destroy (blob);
+  return hb_blob_create ((const char *) new_sfnt_data, new_length,
+			 HB_MEMORY_MODE_WRITABLE, new_sfnt_data, free);
+}
+
+hb_uniscribe_face_data_t *
+_hb_uniscribe_shaper_face_data_create (hb_face_t *face)
+{
+  hb_uniscribe_face_data_t *data = (hb_uniscribe_face_data_t *) calloc (1, sizeof (hb_uniscribe_face_data_t));
+  if (unlikely (!data))
+    return nullptr;
+
+  data->funcs = hb_uniscribe_shaper_get_funcs ();
+  if (unlikely (!data->funcs))
+  {
+    free (data);
+    return nullptr;
+  }
+
+  hb_blob_t *blob = hb_face_reference_blob (face);
+  if (unlikely (!hb_blob_get_length (blob)))
+    DEBUG_MSG (UNISCRIBE, face, "Face has empty blob");
+
+  blob = _hb_rename_font (blob, data->face_name);
+  if (unlikely (!blob))
+  {
+    free (data);
+    return nullptr;
+  }
+
+  DWORD num_fonts_installed;
+  data->fh = AddFontMemResourceEx ((void *) hb_blob_get_data (blob, nullptr),
+				   hb_blob_get_length (blob),
+				   0, &num_fonts_installed);
+  if (unlikely (!data->fh))
+  {
+    DEBUG_MSG (UNISCRIBE, face, "Face AddFontMemResourceEx() failed");
+    free (data);
+    return nullptr;
+  }
+
+  return data;
+}
+
+void
+_hb_uniscribe_shaper_face_data_destroy (hb_uniscribe_face_data_t *data)
+{
+  RemoveFontMemResourceEx (data->fh);
+  free (data);
+}
+
+
+/*
+ * shaper font data
+ */
+
+struct hb_uniscribe_font_data_t
+{
+  HDC hdc;
+  mutable LOGFONTW log_font;
+  HFONT hfont;
+  mutable SCRIPT_CACHE script_cache;
+  double x_mult, y_mult; /* From LOGFONT space to HB space. */
+};
+
+static bool
+populate_log_font (LOGFONTW  *lf,
+		   hb_font_t *font,
+		   unsigned int font_size)
+{
+  memset (lf, 0, sizeof (*lf));
+  lf->lfHeight = - (int) font_size;
+  lf->lfCharSet = DEFAULT_CHARSET;
+
+  memcpy (lf->lfFaceName, font->face->data.uniscribe->face_name, sizeof (lf->lfFaceName));
+
+  return true;
+}
+
+hb_uniscribe_font_data_t *
+_hb_uniscribe_shaper_font_data_create (hb_font_t *font)
+{
+  hb_uniscribe_font_data_t *data = (hb_uniscribe_font_data_t *) calloc (1, sizeof (hb_uniscribe_font_data_t));
+  if (unlikely (!data))
+    return nullptr;
+
+  int font_size = font->face->get_upem (); /* Default... */
+  /* No idea if the following is even a good idea. */
+  if (font->y_ppem)
+    font_size = font->y_ppem;
+
+  if (font_size < 0)
+    font_size = -font_size;
+  data->x_mult = (double) font->x_scale / font_size;
+  data->y_mult = (double) font->y_scale / font_size;
+
+  data->hdc = GetDC (nullptr);
+
+  if (unlikely (!populate_log_font (&data->log_font, font, font_size))) {
+    DEBUG_MSG (UNISCRIBE, font, "Font populate_log_font() failed");
+    _hb_uniscribe_shaper_font_data_destroy (data);
+    return nullptr;
+  }
+
+  data->hfont = CreateFontIndirectW (&data->log_font);
+  if (unlikely (!data->hfont)) {
+    DEBUG_MSG (UNISCRIBE, font, "Font CreateFontIndirectW() failed");
+    _hb_uniscribe_shaper_font_data_destroy (data);
+     return nullptr;
+  }
+
+  if (!SelectObject (data->hdc, data->hfont)) {
+    DEBUG_MSG (UNISCRIBE, font, "Font SelectObject() failed");
+    _hb_uniscribe_shaper_font_data_destroy (data);
+     return nullptr;
+  }
+
+  return data;
+}
+
+void
+_hb_uniscribe_shaper_font_data_destroy (hb_uniscribe_font_data_t *data)
+{
+  if (data->hdc)
+    ReleaseDC (nullptr, data->hdc);
+  if (data->hfont)
+    DeleteObject (data->hfont);
+  if (data->script_cache)
+    ScriptFreeCache (&data->script_cache);
+  free (data);
+}
+
+/**
+ * hb_uniscribe_font_get_logfontw:
+ * @font: The #hb_font_t to work upon
+ *
+ * Fetches the LOGFONTW structure that corresponds to the
+ * specified #hb_font_t font.
+ *
+ * Return value: a pointer to the LOGFONTW retrieved
+ *
+ **/
+LOGFONTW *
+hb_uniscribe_font_get_logfontw (hb_font_t *font)
+{
+  const hb_uniscribe_font_data_t *data =  font->data.uniscribe;
+  return data ? &data->log_font : nullptr;
+}
+
+/**
+ * hb_uniscribe_font_get_hfont:
+ * @font: The #hb_font_t to work upon
+ *
+ * Fetches the HFONT handle that corresponds to the
+ * specified #hb_font_t font.
+ *
+ * Return value: the HFONT retreieved
+ *
+ **/
+HFONT
+hb_uniscribe_font_get_hfont (hb_font_t *font)
+{
+  const hb_uniscribe_font_data_t *data =  font->data.uniscribe;
+  return data ? data->hfont : nullptr;
+}
+
+
+/*
+ * shaper
+ */
+
+
+hb_bool_t
+_hb_uniscribe_shape (hb_shape_plan_t    *shape_plan,
+		     hb_font_t          *font,
+		     hb_buffer_t        *buffer,
+		     const hb_feature_t *features,
+		     unsigned int        num_features)
+{
+  hb_face_t *face = font->face;
+  const hb_uniscribe_face_data_t *face_data = face->data.uniscribe;
+  const hb_uniscribe_font_data_t *font_data = font->data.uniscribe;
+  hb_uniscribe_shaper_funcs_t *funcs = face_data->funcs;
+
+  /*
+   * Set up features.
+   */
+  hb_vector_t<OPENTYPE_FEATURE_RECORD> feature_records;
+  hb_vector_t<range_record_t> range_records;
+  if (num_features)
+  {
+    /* Sort features by start/end events. */
+    hb_vector_t<feature_event_t> feature_events;
+    for (unsigned int i = 0; i < num_features; i++)
+    {
+      active_feature_t feature;
+      feature.rec.tagFeature = hb_uint32_swap (features[i].tag);
+      feature.rec.lParameter = features[i].value;
+      feature.order = i;
+
+      feature_event_t *event;
+
+      event = feature_events.push ();
+      event->index = features[i].start;
+      event->start = true;
+      event->feature = feature;
+
+      event = feature_events.push ();
+      event->index = features[i].end;
+      event->start = false;
+      event->feature = feature;
+    }
+    feature_events.qsort ();
+    /* Add a strategic final event. */
+    {
+      active_feature_t feature;
+      feature.rec.tagFeature = 0;
+      feature.rec.lParameter = 0;
+      feature.order = num_features + 1;
+
+      feature_event_t *event = feature_events.push ();
+      event->index = 0; /* This value does magic. */
+      event->start = false;
+      event->feature = feature;
+    }
+
+    /* Scan events and save features for each range. */
+    hb_vector_t<active_feature_t> active_features;
+    unsigned int last_index = 0;
+    for (unsigned int i = 0; i < feature_events.length; i++)
+    {
+      feature_event_t *event = &feature_events[i];
+
+      if (event->index != last_index)
+      {
+	/* Save a snapshot of active features and the range. */
+	range_record_t *range = range_records.push ();
+
+	unsigned int offset = feature_records.length;
+
+	active_features.qsort ();
+	for (unsigned int j = 0; j < active_features.length; j++)
+	{
+	  if (!j || active_features[j].rec.tagFeature != feature_records[feature_records.length - 1].tagFeature)
+	  {
+	    feature_records.push (active_features[j].rec);
+	  }
+	  else
+	  {
+	    /* Overrides value for existing feature. */
+	    feature_records[feature_records.length - 1].lParameter = active_features[j].rec.lParameter;
+	  }
+	}
+
+	/* Will convert to pointer after all is ready, since feature_records.array
+	 * may move as we grow it. */
+	range->props.potfRecords = reinterpret_cast<OPENTYPE_FEATURE_RECORD *> (offset);
+	range->props.cotfRecords = feature_records.length - offset;
+	range->index_first = last_index;
+	range->index_last  = event->index - 1;
+
+	last_index = event->index;
+      }
+
+      if (event->start)
+      {
+	active_features.push (event->feature);
+      }
+      else
+      {
+	active_feature_t *feature = active_features.find (&event->feature);
+	if (feature)
+	  active_features.remove (feature - active_features.arrayZ);
+      }
+    }
+
+    if (!range_records.length) /* No active feature found. */
+      num_features = 0;
+
+    /* Fixup the pointers. */
+    for (unsigned int i = 0; i < range_records.length; i++)
+    {
+      range_record_t *range = &range_records[i];
+      range->props.potfRecords = (OPENTYPE_FEATURE_RECORD *) feature_records + reinterpret_cast<uintptr_t> (range->props.potfRecords);
+    }
+  }
+
+#define FAIL(...) \
+  HB_STMT_START { \
+    DEBUG_MSG (UNISCRIBE, nullptr, __VA_ARGS__); \
+    return false; \
+  } HB_STMT_END
+
+  HRESULT hr;
+
+retry:
+
+  unsigned int scratch_size;
+  hb_buffer_t::scratch_buffer_t *scratch = buffer->get_scratch_buffer (&scratch_size);
+
+#define ALLOCATE_ARRAY(Type, name, len) \
+  Type *name = (Type *) scratch; \
+  do { \
+    unsigned int _consumed = DIV_CEIL ((len) * sizeof (Type), sizeof (*scratch)); \
+    assert (_consumed <= scratch_size); \
+    scratch += _consumed; \
+    scratch_size -= _consumed; \
+  } while (0)
+
+#define utf16_index() var1.u32
+
+  ALLOCATE_ARRAY (WCHAR, pchars, buffer->len * 2);
+
+  unsigned int chars_len = 0;
+  for (unsigned int i = 0; i < buffer->len; i++)
+  {
+    hb_codepoint_t c = buffer->info[i].codepoint;
+    buffer->info[i].utf16_index() = chars_len;
+    if (likely (c <= 0xFFFFu))
+      pchars[chars_len++] = c;
+    else if (unlikely (c > 0x10FFFFu))
+      pchars[chars_len++] = 0xFFFDu;
+    else {
+      pchars[chars_len++] = 0xD800u + ((c - 0x10000u) >> 10);
+      pchars[chars_len++] = 0xDC00u + ((c - 0x10000u) & ((1u << 10) - 1));
+    }
+  }
+
+  ALLOCATE_ARRAY (WORD, log_clusters, chars_len);
+  ALLOCATE_ARRAY (SCRIPT_CHARPROP, char_props, chars_len);
+
+  if (num_features)
+  {
+    /* Need log_clusters to assign features. */
+    chars_len = 0;
+    for (unsigned int i = 0; i < buffer->len; i++)
+    {
+      hb_codepoint_t c = buffer->info[i].codepoint;
+      unsigned int cluster = buffer->info[i].cluster;
+      log_clusters[chars_len++] = cluster;
+      if (hb_in_range (c, 0x10000u, 0x10FFFFu))
+	log_clusters[chars_len++] = cluster; /* Surrogates. */
+    }
+  }
+
+  /* The -2 in the following is to compensate for possible
+   * alignment needed after the WORD array.  sizeof(WORD) == 2. */
+  unsigned int glyphs_size = (scratch_size * sizeof (int) - 2)
+			   / (sizeof (WORD) +
+			      sizeof (SCRIPT_GLYPHPROP) +
+			      sizeof (int) +
+			      sizeof (GOFFSET) +
+			      sizeof (uint32_t));
+
+  ALLOCATE_ARRAY (WORD, glyphs, glyphs_size);
+  ALLOCATE_ARRAY (SCRIPT_GLYPHPROP, glyph_props, glyphs_size);
+  ALLOCATE_ARRAY (int, advances, glyphs_size);
+  ALLOCATE_ARRAY (GOFFSET, offsets, glyphs_size);
+  ALLOCATE_ARRAY (uint32_t, vis_clusters, glyphs_size);
+
+  /* Note:
+   * We can't touch the contents of glyph_props.  Our fallback
+   * implementations of Shape and Place functions use that buffer
+   * by casting it to a different type.  It works because they
+   * both agree about it, but if we want to access it here we
+   * need address that issue first.
+   */
+
+#undef ALLOCATE_ARRAY
+
+#define MAX_ITEMS 256
+
+  SCRIPT_ITEM items[MAX_ITEMS + 1];
+  SCRIPT_CONTROL bidi_control = {0};
+  SCRIPT_STATE bidi_state = {0};
+  ULONG script_tags[MAX_ITEMS];
+  int item_count;
+
+  /* MinGW32 doesn't define fMergeNeutralItems, so we bruteforce */
+  //bidi_control.fMergeNeutralItems = true;
+  *(uint32_t*)&bidi_control |= 1u<<24;
+
+  bidi_state.uBidiLevel = HB_DIRECTION_IS_FORWARD (buffer->props.direction) ? 0 : 1;
+  bidi_state.fOverrideDirection = 1;
+
+  hr = funcs->ScriptItemizeOpenType (pchars,
+				     chars_len,
+				     MAX_ITEMS,
+				     &bidi_control,
+				     &bidi_state,
+				     items,
+				     script_tags,
+				     &item_count);
+  if (unlikely (FAILED (hr)))
+    FAIL ("ScriptItemizeOpenType() failed: 0x%08lx", hr);
+
+#undef MAX_ITEMS
+
+  hb_tag_t lang_tag;
+  unsigned int lang_count = 1;
+  hb_ot_tags_from_script_and_language (buffer->props.script,
+				       buffer->props.language,
+				       nullptr, nullptr,
+				       &lang_count, &lang_tag);
+  OPENTYPE_TAG language_tag = hb_uint32_swap (lang_count ? lang_tag : HB_TAG_NONE);
+  hb_vector_t<TEXTRANGE_PROPERTIES*> range_properties;
+  hb_vector_t<int> range_char_counts;
+
+  unsigned int glyphs_offset = 0;
+  unsigned int glyphs_len;
+  bool backward = HB_DIRECTION_IS_BACKWARD (buffer->props.direction);
+  for (int i = 0; i < item_count; i++)
+  {
+    unsigned int chars_offset = items[i].iCharPos;
+    unsigned int item_chars_len = items[i + 1].iCharPos - chars_offset;
+
+    if (num_features)
+    {
+      range_properties.shrink (0);
+      range_char_counts.shrink (0);
+
+      range_record_t *last_range = &range_records[0];
+
+      for (unsigned int k = chars_offset; k < chars_offset + item_chars_len; k++)
+      {
+	range_record_t *range = last_range;
+	while (log_clusters[k] < range->index_first)
+	  range--;
+	while (log_clusters[k] > range->index_last)
+	  range++;
+	if (!range_properties.length ||
+	    &range->props != range_properties[range_properties.length - 1])
+	{
+	  TEXTRANGE_PROPERTIES **props = range_properties.push ();
+	  int *c = range_char_counts.push ();
+	  if (unlikely (!props || !c))
+	  {
+	    range_properties.shrink (0);
+	    range_char_counts.shrink (0);
+	    break;
+	  }
+	  *props = &range->props;
+	  *c = 1;
+	}
+	else
+	{
+	  range_char_counts[range_char_counts.length - 1]++;
+	}
+
+	last_range = range;
+      }
+    }
+
+    /* Asking for glyphs in logical order circumvents at least
+     * one bug in Uniscribe. */
+    items[i].a.fLogicalOrder = true;
+
+  retry_shape:
+    hr = funcs->ScriptShapeOpenType (font_data->hdc,
+				     &font_data->script_cache,
+				     &items[i].a,
+				     script_tags[i],
+				     language_tag,
+				     range_char_counts.arrayZ,
+				     range_properties.arrayZ,
+				     range_properties.length,
+				     pchars + chars_offset,
+				     item_chars_len,
+				     glyphs_size - glyphs_offset,
+				     /* out */
+				     log_clusters + chars_offset,
+				     char_props + chars_offset,
+				     glyphs + glyphs_offset,
+				     glyph_props + glyphs_offset,
+				     (int *) &glyphs_len);
+
+    if (unlikely (items[i].a.fNoGlyphIndex))
+      FAIL ("ScriptShapeOpenType() set fNoGlyphIndex");
+    if (unlikely (hr == E_OUTOFMEMORY || hr == E_NOT_SUFFICIENT_BUFFER))
+    {
+      if (unlikely (!buffer->ensure (buffer->allocated * 2)))
+	FAIL ("Buffer resize failed");
+      goto retry;
+    }
+    if (unlikely (hr == USP_E_SCRIPT_NOT_IN_FONT))
+    {
+      if (items[i].a.eScript == SCRIPT_UNDEFINED)
+	FAIL ("ScriptShapeOpenType() failed: Font doesn't support script");
+      items[i].a.eScript = SCRIPT_UNDEFINED;
+      goto retry_shape;
+    }
+    if (unlikely (FAILED (hr)))
+    {
+      FAIL ("ScriptShapeOpenType() failed: 0x%08lx", hr);
+    }
+
+    for (unsigned int j = chars_offset; j < chars_offset + item_chars_len; j++)
+      log_clusters[j] += glyphs_offset;
+
+    hr = funcs->ScriptPlaceOpenType (font_data->hdc,
+				     &font_data->script_cache,
+				     &items[i].a,
+				     script_tags[i],
+				     language_tag,
+				     range_char_counts.arrayZ,
+				     range_properties.arrayZ,
+				     range_properties.length,
+				     pchars + chars_offset,
+				     log_clusters + chars_offset,
+				     char_props + chars_offset,
+				     item_chars_len,
+				     glyphs + glyphs_offset,
+				     glyph_props + glyphs_offset,
+				     glyphs_len,
+				     /* out */
+				     advances + glyphs_offset,
+				     offsets + glyphs_offset,
+				     nullptr);
+    if (unlikely (FAILED (hr)))
+      FAIL ("ScriptPlaceOpenType() failed: 0x%08lx", hr);
+
+    if (DEBUG_ENABLED (UNISCRIBE))
+      fprintf (stderr, "Item %d RTL %d LayoutRTL %d LogicalOrder %d ScriptTag %c%c%c%c\n",
+	       i,
+	       items[i].a.fRTL,
+	       items[i].a.fLayoutRTL,
+	       items[i].a.fLogicalOrder,
+	       HB_UNTAG (hb_uint32_swap (script_tags[i])));
+
+    glyphs_offset += glyphs_len;
+  }
+  glyphs_len = glyphs_offset;
+
+  /* Ok, we've got everything we need, now compose output buffer,
+   * very, *very*, carefully! */
+
+  /* Calculate visual-clusters.  That's what we ship. */
+  for (unsigned int i = 0; i < glyphs_len; i++)
+    vis_clusters[i] = (uint32_t) -1;
+  for (unsigned int i = 0; i < buffer->len; i++) {
+    uint32_t *p = &vis_clusters[log_clusters[buffer->info[i].utf16_index()]];
+    *p = hb_min (*p, buffer->info[i].cluster);
+  }
+  for (unsigned int i = 1; i < glyphs_len; i++)
+    if (vis_clusters[i] == (uint32_t) -1)
+      vis_clusters[i] = vis_clusters[i - 1];
+
+#undef utf16_index
+
+  if (unlikely (!buffer->ensure (glyphs_len)))
+    FAIL ("Buffer in error");
+
+#undef FAIL
+
+  /* Set glyph infos */
+  buffer->len = 0;
+  for (unsigned int i = 0; i < glyphs_len; i++)
+  {
+    hb_glyph_info_t *info = &buffer->info[buffer->len++];
+
+    info->codepoint = glyphs[i];
+    info->cluster = vis_clusters[i];
+
+    /* The rest is crap.  Let's store position info there for now. */
+    info->mask = advances[i];
+    info->var1.i32 = offsets[i].du;
+    info->var2.i32 = offsets[i].dv;
+  }
+
+  /* Set glyph positions */
+  buffer->clear_positions ();
+  double x_mult = font_data->x_mult, y_mult = font_data->y_mult;
+  for (unsigned int i = 0; i < glyphs_len; i++)
+  {
+    hb_glyph_info_t *info = &buffer->info[i];
+    hb_glyph_position_t *pos = &buffer->pos[i];
+
+    /* TODO vertical */
+    pos->x_advance = x_mult * (int32_t) info->mask;
+    pos->x_offset = x_mult * (backward ? -info->var1.i32 : info->var1.i32);
+    pos->y_offset = y_mult * info->var2.i32;
+  }
+
+  if (backward)
+    hb_buffer_reverse (buffer);
+
+  buffer->unsafe_to_break_all ();
+
+  /* Wow, done! */
+  return true;
+}
+
+
+#endif
diff --git a/thirdparty/harfbuzz/src/hb-uniscribe.h b/thirdparty/harfbuzz/src/hb-uniscribe.h
new file mode 100644
index 0000000000..4e4ef9986a
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-uniscribe.h
@@ -0,0 +1,46 @@
+/*
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_UNISCRIBE_H
+#define HB_UNISCRIBE_H
+
+#include "hb.h"
+
+#include <windows.h>
+
+HB_BEGIN_DECLS
+
+
+HB_EXTERN LOGFONTW *
+hb_uniscribe_font_get_logfontw (hb_font_t *font);
+
+HB_EXTERN HFONT
+hb_uniscribe_font_get_hfont (hb_font_t *font);
+
+
+HB_END_DECLS
+
+#endif /* HB_UNISCRIBE_H */
diff --git a/thirdparty/harfbuzz/src/hb-utf.hh b/thirdparty/harfbuzz/src/hb-utf.hh
new file mode 100644
index 0000000000..ff5712d16d
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-utf.hh
@@ -0,0 +1,453 @@
+/*
+ * Copyright © 2011,2012,2014  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_UTF_HH
+#define HB_UTF_HH
+
+#include "hb.hh"
+
+#include "hb-open-type.hh"
+
+
+struct hb_utf8_t
+{
+  typedef uint8_t codepoint_t;
+
+  static const codepoint_t *
+  next (const codepoint_t *text,
+	const codepoint_t *end,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    /* Written to only accept well-formed sequences.
+     * Based on ideas from ICU's U8_NEXT.
+     * Generates one "replacement" for each ill-formed byte. */
+
+    hb_codepoint_t c = *text++;
+
+    if (c > 0x7Fu)
+    {
+      if (hb_in_range<hb_codepoint_t> (c, 0xC2u, 0xDFu)) /* Two-byte */
+      {
+	unsigned int t1;
+	if (likely (text < end &&
+		    (t1 = text[0] - 0x80u) <= 0x3Fu))
+	{
+	  c = ((c&0x1Fu)<<6) | t1;
+	  text++;
+	}
+	else
+	  goto error;
+      }
+      else if (hb_in_range<hb_codepoint_t> (c, 0xE0u, 0xEFu)) /* Three-byte */
+      {
+	unsigned int t1, t2;
+	if (likely (1 < end - text &&
+		    (t1 = text[0] - 0x80u) <= 0x3Fu &&
+		    (t2 = text[1] - 0x80u) <= 0x3Fu))
+	{
+	  c = ((c&0xFu)<<12) | (t1<<6) | t2;
+	  if (unlikely (c < 0x0800u || hb_in_range<hb_codepoint_t> (c, 0xD800u, 0xDFFFu)))
+	    goto error;
+	  text += 2;
+	}
+	else
+	  goto error;
+      }
+      else if (hb_in_range<hb_codepoint_t> (c, 0xF0u, 0xF4u)) /* Four-byte */
+      {
+	unsigned int t1, t2, t3;
+	if (likely (2 < end - text &&
+		    (t1 = text[0] - 0x80u) <= 0x3Fu &&
+		    (t2 = text[1] - 0x80u) <= 0x3Fu &&
+		    (t3 = text[2] - 0x80u) <= 0x3Fu))
+	{
+	  c = ((c&0x7u)<<18) | (t1<<12) | (t2<<6) | t3;
+	  if (unlikely (!hb_in_range<hb_codepoint_t> (c, 0x10000u, 0x10FFFFu)))
+	    goto error;
+	  text += 3;
+	}
+	else
+	  goto error;
+      }
+      else
+	goto error;
+    }
+
+    *unicode = c;
+    return text;
+
+  error:
+    *unicode = replacement;
+    return text;
+  }
+
+  static const codepoint_t *
+  prev (const codepoint_t *text,
+	const codepoint_t *start,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    const codepoint_t *end = text--;
+    while (start < text && (*text & 0xc0) == 0x80 && end - text < 4)
+      text--;
+
+    if (likely (next (text, end, unicode, replacement) == end))
+      return text;
+
+    *unicode = replacement;
+    return end - 1;
+  }
+
+  static unsigned int
+  strlen (const codepoint_t *text)
+  { return ::strlen ((const char *) text); }
+
+  static unsigned int
+  encode_len (hb_codepoint_t unicode)
+  {
+    if (unicode <   0x0080u) return 1;
+    if (unicode <   0x0800u) return 2;
+    if (unicode <  0x10000u) return 3;
+    if (unicode < 0x110000u) return 4;
+    return 3;
+  }
+
+  static codepoint_t *
+  encode (codepoint_t *text,
+	  const codepoint_t *end,
+	  hb_codepoint_t unicode)
+  {
+    if (unlikely (unicode >= 0xD800u && (unicode <= 0xDFFFu || unicode > 0x10FFFFu)))
+      unicode = 0xFFFDu;
+    if (unicode < 0x0080u)
+     *text++ = unicode;
+    else if (unicode < 0x0800u)
+    {
+      if (end - text >= 2)
+      {
+	*text++ =  0xC0u + (0x1Fu & (unicode >>  6));
+	*text++ =  0x80u + (0x3Fu & (unicode      ));
+      }
+    }
+    else if (unicode < 0x10000u)
+    {
+      if (end - text >= 3)
+      {
+	*text++ =  0xE0u + (0x0Fu & (unicode >> 12));
+	*text++ =  0x80u + (0x3Fu & (unicode >>  6));
+	*text++ =  0x80u + (0x3Fu & (unicode      ));
+      }
+    }
+    else
+    {
+      if (end - text >= 4)
+      {
+	*text++ =  0xF0u + (0x07u & (unicode >> 18));
+	*text++ =  0x80u + (0x3Fu & (unicode >> 12));
+	*text++ =  0x80u + (0x3Fu & (unicode >>  6));
+	*text++ =  0x80u + (0x3Fu & (unicode      ));
+      }
+    }
+    return text;
+  }
+};
+
+
+template <typename TCodepoint>
+struct hb_utf16_xe_t
+{
+  static_assert (sizeof (TCodepoint) == 2, "");
+  typedef TCodepoint codepoint_t;
+
+  static const codepoint_t *
+  next (const codepoint_t *text,
+	const codepoint_t *end,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    hb_codepoint_t c = *text++;
+
+    if (likely (!hb_in_range<hb_codepoint_t> (c, 0xD800u, 0xDFFFu)))
+    {
+      *unicode = c;
+      return text;
+    }
+
+    if (likely (c <= 0xDBFFu && text < end))
+    {
+      /* High-surrogate in c */
+      hb_codepoint_t l = *text;
+      if (likely (hb_in_range<hb_codepoint_t> (l, 0xDC00u, 0xDFFFu)))
+      {
+	/* Low-surrogate in l */
+	*unicode = (c << 10) + l - ((0xD800u << 10) - 0x10000u + 0xDC00u);
+	 text++;
+	 return text;
+      }
+    }
+
+    /* Lonely / out-of-order surrogate. */
+    *unicode = replacement;
+    return text;
+  }
+
+  static const codepoint_t *
+  prev (const codepoint_t *text,
+	const codepoint_t *start,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    hb_codepoint_t c = *--text;
+
+    if (likely (!hb_in_range<hb_codepoint_t> (c, 0xD800u, 0xDFFFu)))
+    {
+      *unicode = c;
+      return text;
+    }
+
+    if (likely (c >= 0xDC00u && start < text))
+    {
+      /* Low-surrogate in c */
+      hb_codepoint_t h = text[-1];
+      if (likely (hb_in_range<hb_codepoint_t> (h, 0xD800u, 0xDBFFu)))
+      {
+	/* High-surrogate in h */
+	*unicode = (h << 10) + c - ((0xD800u << 10) - 0x10000u + 0xDC00u);
+	text--;
+	return text;
+      }
+    }
+
+    /* Lonely / out-of-order surrogate. */
+    *unicode = replacement;
+    return text;
+  }
+
+
+  static unsigned int
+  strlen (const codepoint_t *text)
+  {
+    unsigned int l = 0;
+    while (*text++) l++;
+    return l;
+  }
+
+  static unsigned int
+  encode_len (hb_codepoint_t unicode)
+  {
+    return unicode < 0x10000 ? 1 : 2;
+  }
+
+  static codepoint_t *
+  encode (codepoint_t *text,
+	  const codepoint_t *end,
+	  hb_codepoint_t unicode)
+  {
+    if (unlikely (unicode >= 0xD800u && (unicode <= 0xDFFFu || unicode > 0x10FFFFu)))
+      unicode = 0xFFFDu;
+    if (unicode < 0x10000u)
+     *text++ = unicode;
+    else if (end - text >= 2)
+    {
+      unicode -= 0x10000u;
+      *text++ =  0xD800u + (unicode >> 10);
+      *text++ =  0xDC00u + (unicode & 0x03FFu);
+    }
+    return text;
+  }
+};
+
+typedef hb_utf16_xe_t<uint16_t> hb_utf16_t;
+typedef hb_utf16_xe_t<OT::HBUINT16> hb_utf16_be_t;
+
+
+template <typename TCodepoint, bool validate=true>
+struct hb_utf32_xe_t
+{
+  static_assert (sizeof (TCodepoint) == 4, "");
+  typedef TCodepoint codepoint_t;
+
+  static const TCodepoint *
+  next (const TCodepoint *text,
+	const TCodepoint *end HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    hb_codepoint_t c = *unicode = *text++;
+    if (validate && unlikely (c >= 0xD800u && (c <= 0xDFFFu || c > 0x10FFFFu)))
+      *unicode = replacement;
+    return text;
+  }
+
+  static const TCodepoint *
+  prev (const TCodepoint *text,
+	const TCodepoint *start HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    hb_codepoint_t c = *unicode = *--text;
+    if (validate && unlikely (c >= 0xD800u && (c <= 0xDFFFu || c > 0x10FFFFu)))
+      *unicode = replacement;
+    return text;
+  }
+
+  static unsigned int
+  strlen (const TCodepoint *text)
+  {
+    unsigned int l = 0;
+    while (*text++) l++;
+    return l;
+  }
+
+  static unsigned int
+  encode_len (hb_codepoint_t unicode HB_UNUSED)
+  {
+    return 1;
+  }
+
+  static codepoint_t *
+  encode (codepoint_t *text,
+	  const codepoint_t *end HB_UNUSED,
+	  hb_codepoint_t unicode)
+  {
+    if (validate && unlikely (unicode >= 0xD800u && (unicode <= 0xDFFFu || unicode > 0x10FFFFu)))
+      unicode = 0xFFFDu;
+    *text++ = unicode;
+    return text;
+  }
+};
+
+typedef hb_utf32_xe_t<uint32_t> hb_utf32_t;
+typedef hb_utf32_xe_t<uint32_t, false> hb_utf32_novalidate_t;
+
+
+struct hb_latin1_t
+{
+  typedef uint8_t codepoint_t;
+
+  static const codepoint_t *
+  next (const codepoint_t *text,
+	const codepoint_t *end HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement HB_UNUSED)
+  {
+    *unicode = *text++;
+    return text;
+  }
+
+  static const codepoint_t *
+  prev (const codepoint_t *text,
+	const codepoint_t *start HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement HB_UNUSED)
+  {
+    *unicode = *--text;
+    return text;
+  }
+
+  static unsigned int
+  strlen (const codepoint_t *text)
+  {
+    unsigned int l = 0;
+    while (*text++) l++;
+    return l;
+  }
+
+  static unsigned int
+  encode_len (hb_codepoint_t unicode HB_UNUSED)
+  {
+    return 1;
+  }
+
+  static codepoint_t *
+  encode (codepoint_t *text,
+	  const codepoint_t *end HB_UNUSED,
+	  hb_codepoint_t unicode)
+  {
+    if (unlikely (unicode >= 0x0100u))
+      unicode = '?';
+    *text++ = unicode;
+    return text;
+  }
+};
+
+
+struct hb_ascii_t
+{
+  typedef uint8_t codepoint_t;
+
+  static const codepoint_t *
+  next (const codepoint_t *text,
+	const codepoint_t *end HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement HB_UNUSED)
+  {
+    *unicode = *text++;
+    if (*unicode >= 0x0080u)
+      *unicode = replacement;
+    return text;
+  }
+
+  static const codepoint_t *
+  prev (const codepoint_t *text,
+	const codepoint_t *start HB_UNUSED,
+	hb_codepoint_t *unicode,
+	hb_codepoint_t replacement)
+  {
+    *unicode = *--text;
+    if (*unicode >= 0x0080u)
+      *unicode = replacement;
+    return text;
+  }
+
+  static unsigned int
+  strlen (const codepoint_t *text)
+  {
+    unsigned int l = 0;
+    while (*text++) l++;
+    return l;
+  }
+
+  static unsigned int
+  encode_len (hb_codepoint_t unicode HB_UNUSED)
+  {
+    return 1;
+  }
+
+  static codepoint_t *
+  encode (codepoint_t *text,
+	  const codepoint_t *end HB_UNUSED,
+	  hb_codepoint_t unicode)
+  {
+    if (unlikely (unicode >= 0x0080u))
+      unicode = '?';
+    *text++ = unicode;
+    return text;
+  }
+};
+
+#endif /* HB_UTF_HH */
diff --git a/thirdparty/harfbuzz/src/hb-vector.hh b/thirdparty/harfbuzz/src/hb-vector.hh
new file mode 100644
index 0000000000..079b94a6b4
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-vector.hh
@@ -0,0 +1,313 @@
+/*
+ * Copyright © 2017,2018  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_VECTOR_HH
+#define HB_VECTOR_HH
+
+#include "hb.hh"
+#include "hb-array.hh"
+#include "hb-null.hh"
+
+
+template <typename Type>
+struct hb_vector_t
+{
+  typedef Type item_t;
+  static constexpr unsigned item_size = hb_static_size (Type);
+
+  hb_vector_t ()  { init (); }
+  hb_vector_t (const hb_vector_t &o)
+  {
+    init ();
+    alloc (o.length);
+    hb_copy (o, *this);
+  }
+  hb_vector_t (hb_vector_t &&o)
+  {
+    allocated = o.allocated;
+    length = o.length;
+    arrayZ = o.arrayZ;
+    o.init ();
+  }
+  ~hb_vector_t () { fini (); }
+
+  private:
+  int allocated; /* == -1 means allocation failed. */
+  public:
+  unsigned int length;
+  public:
+  Type *arrayZ;
+
+  void init ()
+  {
+    allocated = length = 0;
+    arrayZ = nullptr;
+  }
+
+  void fini ()
+  {
+    free (arrayZ);
+    init ();
+  }
+  void fini_deep ()
+  {
+    unsigned int count = length;
+    for (unsigned int i = 0; i < count; i++)
+      arrayZ[i].fini ();
+    fini ();
+  }
+
+  void reset () { resize (0); }
+
+  hb_vector_t& operator = (const hb_vector_t &o)
+  {
+    reset ();
+    alloc (o.length);
+    hb_copy (o, *this);
+    return *this;
+  }
+  hb_vector_t& operator = (hb_vector_t &&o)
+  {
+    fini ();
+    allocated = o.allocated;
+    length = o.length;
+    arrayZ = o.arrayZ;
+    o.init ();
+    return *this;
+  }
+
+  hb_bytes_t as_bytes () const
+  { return hb_bytes_t ((const char *) arrayZ, length * item_size); }
+
+  bool operator == (const hb_vector_t &o) const { return as_array () == o.as_array (); }
+  bool operator != (const hb_vector_t &o) const { return !(*this == o); }
+  uint32_t hash () const { return as_array ().hash (); }
+
+  Type& operator [] (int i_)
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= length))
+      return Crap (Type);
+    return arrayZ[i];
+  }
+  const Type& operator [] (int i_) const
+  {
+    unsigned int i = (unsigned int) i_;
+    if (unlikely (i >= length))
+      return Null (Type);
+    return arrayZ[i];
+  }
+
+  Type& tail () { return (*this)[length - 1]; }
+  const Type& tail () const { return (*this)[length - 1]; }
+
+  explicit operator bool () const { return length; }
+  unsigned get_size () const { return length * item_size; }
+
+  /* Sink interface. */
+  template <typename T>
+  hb_vector_t& operator << (T&& v) { push (hb_forward<T> (v)); return *this; }
+
+  hb_array_t<      Type> as_array ()       { return hb_array (arrayZ, length); }
+  hb_array_t<const Type> as_array () const { return hb_array (arrayZ, length); }
+
+  /* Iterator. */
+  typedef hb_array_t<const Type>   iter_t;
+  typedef hb_array_t<      Type> writer_t;
+    iter_t   iter () const { return as_array (); }
+  writer_t writer ()       { return as_array (); }
+  operator   iter_t () const { return   iter (); }
+  operator writer_t ()       { return writer (); }
+
+  hb_array_t<const Type> sub_array (unsigned int start_offset, unsigned int count) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<const Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */) const
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<Type> sub_array (unsigned int start_offset, unsigned int count)
+  { return as_array ().sub_array (start_offset, count); }
+  hb_array_t<Type> sub_array (unsigned int start_offset, unsigned int *count = nullptr /* IN/OUT */)
+  { return as_array ().sub_array (start_offset, count); }
+
+  hb_sorted_array_t<Type> as_sorted_array ()
+  { return hb_sorted_array (arrayZ, length); }
+  hb_sorted_array_t<const Type> as_sorted_array () const
+  { return hb_sorted_array (arrayZ, length); }
+
+  template <typename T> explicit operator T * () { return arrayZ; }
+  template <typename T> explicit operator const T * () const { return arrayZ; }
+
+  Type * operator  + (unsigned int i) { return arrayZ + i; }
+  const Type * operator  + (unsigned int i) const { return arrayZ + i; }
+
+  Type *push ()
+  {
+    if (unlikely (!resize (length + 1)))
+      return &Crap (Type);
+    return &arrayZ[length - 1];
+  }
+  template <typename T>
+  Type *push (T&& v)
+  {
+    Type *p = push ();
+    *p = hb_forward<T> (v);
+    return p;
+  }
+
+  bool in_error () const { return allocated < 0; }
+
+  /* Allocate for size but don't adjust length. */
+  bool alloc (unsigned int size)
+  {
+    if (unlikely (allocated < 0))
+      return false;
+
+    if (likely (size <= (unsigned) allocated))
+      return true;
+
+    /* Reallocate */
+
+    unsigned int new_allocated = allocated;
+    while (size >= new_allocated)
+      new_allocated += (new_allocated >> 1) + 8;
+
+    Type *new_array = nullptr;
+    bool overflows =
+      (int) new_allocated < 0 ||
+      (new_allocated < (unsigned) allocated) ||
+      hb_unsigned_mul_overflows (new_allocated, sizeof (Type));
+    if (likely (!overflows))
+      new_array = (Type *) realloc (arrayZ, new_allocated * sizeof (Type));
+
+    if (unlikely (!new_array))
+    {
+      allocated = -1;
+      return false;
+    }
+
+    arrayZ = new_array;
+    allocated = new_allocated;
+
+    return true;
+  }
+
+  bool resize (int size_)
+  {
+    unsigned int size = size_ < 0 ? 0u : (unsigned int) size_;
+    if (!alloc (size))
+      return false;
+
+    if (size > length)
+      memset (arrayZ + length, 0, (size - length) * sizeof (*arrayZ));
+
+    length = size;
+    return true;
+  }
+
+  Type pop ()
+  {
+    if (!length) return Null (Type);
+    return hb_move (arrayZ[--length]); /* Does this move actually work? */
+  }
+
+  void remove (unsigned int i)
+  {
+    if (unlikely (i >= length))
+      return;
+    memmove (static_cast<void *> (&arrayZ[i]),
+	     static_cast<void *> (&arrayZ[i + 1]),
+	     (length - i - 1) * sizeof (Type));
+    length--;
+  }
+
+  void shrink (int size_)
+  {
+    unsigned int size = size_ < 0 ? 0u : (unsigned int) size_;
+     if (size < length)
+       length = size;
+  }
+
+  template <typename T>
+  Type *find (T v)
+  {
+    for (unsigned int i = 0; i < length; i++)
+      if (arrayZ[i] == v)
+	return &arrayZ[i];
+    return nullptr;
+  }
+  template <typename T>
+  const Type *find (T v) const
+  {
+    for (unsigned int i = 0; i < length; i++)
+      if (arrayZ[i] == v)
+	return &arrayZ[i];
+    return nullptr;
+  }
+
+  void qsort (int (*cmp)(const void*, const void*))
+  { as_array ().qsort (cmp); }
+  void qsort (unsigned int start = 0, unsigned int end = (unsigned int) -1)
+  { as_array ().qsort (start, end); }
+
+  template <typename T>
+  Type *lsearch (const T &x, Type *not_found = nullptr)
+  { return as_array ().lsearch (x, not_found); }
+  template <typename T>
+  const Type *lsearch (const T &x, const Type *not_found = nullptr) const
+  { return as_array ().lsearch (x, not_found); }
+  template <typename T>
+  bool lfind (const T &x, unsigned *pos = nullptr) const
+  { return as_array ().lfind (x, pos); }
+};
+
+template <typename Type>
+struct hb_sorted_vector_t : hb_vector_t<Type>
+{
+  hb_sorted_array_t<      Type> as_array ()       { return hb_sorted_array (this->arrayZ, this->length); }
+  hb_sorted_array_t<const Type> as_array () const { return hb_sorted_array (this->arrayZ, this->length); }
+
+  /* Iterator. */
+  typedef hb_sorted_array_t<const Type> const_iter_t;
+  typedef hb_sorted_array_t<      Type>       iter_t;
+  const_iter_t  iter () const { return as_array (); }
+  const_iter_t citer () const { return as_array (); }
+	iter_t  iter ()       { return as_array (); }
+  operator       iter_t ()       { return iter (); }
+  operator const_iter_t () const { return iter (); }
+
+  template <typename T>
+  Type *bsearch (const T &x, Type *not_found = nullptr)
+  { return as_array ().bsearch (x, not_found); }
+  template <typename T>
+  const Type *bsearch (const T &x, const Type *not_found = nullptr) const
+  { return as_array ().bsearch (x, not_found); }
+  template <typename T>
+  bool bfind (const T &x, unsigned int *i = nullptr,
+	      hb_bfind_not_found_t not_found = HB_BFIND_NOT_FOUND_DONT_STORE,
+	      unsigned int to_store = (unsigned int) -1) const
+  { return as_array ().bfind (x, i, not_found, to_store); }
+};
+
+#endif /* HB_VECTOR_HH */
diff --git a/thirdparty/harfbuzz/src/hb-version.h b/thirdparty/harfbuzz/src/hb-version.h
new file mode 100644
index 0000000000..92d61b8cdb
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb-version.h
@@ -0,0 +1,66 @@
+/*
+ * Copyright © 2011  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H_IN
+#error "Include <hb.h> instead."
+#endif
+
+#ifndef HB_VERSION_H
+#define HB_VERSION_H
+
+#include "hb-common.h"
+
+HB_BEGIN_DECLS
+
+
+#define HB_VERSION_MAJOR 2
+#define HB_VERSION_MINOR 7
+#define HB_VERSION_MICRO 2
+
+#define HB_VERSION_STRING "2.7.2"
+
+#define HB_VERSION_ATLEAST(major,minor,micro) \
+	((major)*10000+(minor)*100+(micro) <= \
+	 HB_VERSION_MAJOR*10000+HB_VERSION_MINOR*100+HB_VERSION_MICRO)
+
+
+HB_EXTERN void
+hb_version (unsigned int *major,
+	    unsigned int *minor,
+	    unsigned int *micro);
+
+HB_EXTERN const char *
+hb_version_string (void);
+
+HB_EXTERN hb_bool_t
+hb_version_atleast (unsigned int major,
+		    unsigned int minor,
+		    unsigned int micro);
+
+
+HB_END_DECLS
+
+#endif /* HB_VERSION_H */
diff --git a/thirdparty/harfbuzz/src/hb.h b/thirdparty/harfbuzz/src/hb.h
new file mode 100644
index 0000000000..360686ca68
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb.h
@@ -0,0 +1,50 @@
+/*
+ * Copyright © 2009  Red Hat, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_H
+#define HB_H
+#define HB_H_IN
+
+#include "hb-blob.h"
+#include "hb-buffer.h"
+#include "hb-common.h"
+#include "hb-deprecated.h"
+#include "hb-draw.h"
+#include "hb-face.h"
+#include "hb-font.h"
+#include "hb-map.h"
+#include "hb-set.h"
+#include "hb-shape.h"
+#include "hb-shape-plan.h"
+#include "hb-style.h"
+#include "hb-unicode.h"
+#include "hb-version.h"
+
+HB_BEGIN_DECLS
+HB_END_DECLS
+
+#undef HB_H_IN
+#endif /* HB_H */
diff --git a/thirdparty/harfbuzz/src/hb.hh b/thirdparty/harfbuzz/src/hb.hh
new file mode 100644
index 0000000000..274a0e98db
--- /dev/null
+++ b/thirdparty/harfbuzz/src/hb.hh
@@ -0,0 +1,634 @@
+/*
+ * Copyright © 2007,2008,2009  Red Hat, Inc.
+ * Copyright © 2011,2012  Google, Inc.
+ *
+ *  This is part of HarfBuzz, a text shaping library.
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and its documentation for any purpose, provided that the
+ * above copyright notice and the following two paragraphs appear in
+ * all copies of this software.
+ *
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
+ * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
+ * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
+ * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
+ * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
+ * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
+ * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
+ * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
+ *
+ * Red Hat Author(s): Behdad Esfahbod
+ * Google Author(s): Behdad Esfahbod
+ */
+
+#ifndef HB_HH
+#define HB_HH
+
+
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC
+#ifdef _MSC_VER
+#pragma warning( disable: 4068 ) /* Unknown pragma */
+#endif
+#if defined(__GNUC__) || defined(__clang__)
+/* Rules:
+ *
+ * - All pragmas are declared GCC even if they are clang ones.  Otherwise GCC
+ *   nags, even though we instruct it to ignore -Wunknown-pragmas. ¯\_(ツ)_/¯
+ *
+ * - Within each category, keep sorted.
+ *
+ * - Warnings whose scope can be expanded in future compiler versions shall
+ *   be declared as "warning".  Otherwise, either ignored or error.
+ */
+
+/* Setup.  Don't sort order within this category. */
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_WARNING
+#pragma GCC diagnostic warning "-Wall"
+#pragma GCC diagnostic warning "-Wextra"
+#endif
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_IGNORED
+#pragma GCC diagnostic ignored "-Wpragmas"
+#pragma GCC diagnostic ignored "-Wunknown-pragmas"
+#pragma GCC diagnostic ignored "-Wunknown-warning-option"
+#endif
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_WARNING
+//#pragma GCC diagnostic warning "-Weverything"
+#endif
+
+/* Error.  Should never happen. */
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_ERROR
+#pragma GCC diagnostic error   "-Wc++11-narrowing"
+#pragma GCC diagnostic error   "-Wcast-align"
+#pragma GCC diagnostic error   "-Wcast-function-type"
+#pragma GCC diagnostic error   "-Wdelete-non-virtual-dtor"
+#pragma GCC diagnostic error   "-Wembedded-directive"
+#pragma GCC diagnostic error   "-Wextra-semi-stmt"
+#pragma GCC diagnostic error   "-Wformat-security"
+#pragma GCC diagnostic error   "-Wimplicit-function-declaration"
+#pragma GCC diagnostic error   "-Winit-self"
+#pragma GCC diagnostic error   "-Winjected-class-name"
+#pragma GCC diagnostic error   "-Wmissing-braces"
+#pragma GCC diagnostic error   "-Wmissing-declarations"
+#pragma GCC diagnostic error   "-Wmissing-prototypes"
+#pragma GCC diagnostic error   "-Wnested-externs"
+#pragma GCC diagnostic error   "-Wold-style-definition"
+#pragma GCC diagnostic error   "-Wpointer-arith"
+#pragma GCC diagnostic error   "-Wredundant-decls"
+#pragma GCC diagnostic error   "-Wreorder"
+#pragma GCC diagnostic error   "-Wsign-compare"
+#pragma GCC diagnostic error   "-Wstrict-prototypes"
+#pragma GCC diagnostic error   "-Wstring-conversion"
+#pragma GCC diagnostic error   "-Wswitch-enum"
+#pragma GCC diagnostic error   "-Wtautological-overlap-compare"
+#pragma GCC diagnostic error   "-Wunneeded-internal-declaration"
+#pragma GCC diagnostic error   "-Wunused"
+#pragma GCC diagnostic error   "-Wunused-local-typedefs"
+#pragma GCC diagnostic error   "-Wunused-value"
+#pragma GCC diagnostic error   "-Wunused-variable"
+#pragma GCC diagnostic error   "-Wvla"
+#pragma GCC diagnostic error   "-Wwrite-strings"
+#endif
+
+/* Warning.  To be investigated if happens. */
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_WARNING
+#pragma GCC diagnostic warning "-Wbuiltin-macro-redefined"
+#pragma GCC diagnostic warning "-Wdeprecated"
+#pragma GCC diagnostic warning "-Wdeprecated-declarations"
+#pragma GCC diagnostic warning "-Wdisabled-optimization"
+#pragma GCC diagnostic warning "-Wdouble-promotion"
+#pragma GCC diagnostic warning "-Wformat=2"
+#pragma GCC diagnostic warning "-Wignored-pragma-optimize"
+#pragma GCC diagnostic warning "-Wlogical-op"
+#pragma GCC diagnostic warning "-Wmaybe-uninitialized"
+#pragma GCC diagnostic warning "-Wmissing-format-attribute"
+#pragma GCC diagnostic warning "-Wundef"
+#pragma GCC diagnostic warning "-Wunused-but-set-variable"
+#endif
+
+/* Ignored currently, but should be fixed at some point. */
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_IGNORED
+#pragma GCC diagnostic ignored "-Wconversion"			// TODO fix
+#pragma GCC diagnostic ignored "-Wformat-signedness"		// TODO fix
+#pragma GCC diagnostic ignored "-Wshadow"			// TODO fix
+#pragma GCC diagnostic ignored "-Wunsafe-loop-optimizations"	// TODO fix
+#pragma GCC diagnostic ignored "-Wunused-parameter"		// TODO fix
+#endif
+
+/* Ignored intentionally. */
+#ifndef HB_NO_PRAGMA_GCC_DIAGNOSTIC_IGNORED
+#pragma GCC diagnostic ignored "-Wclass-memaccess"
+#pragma GCC diagnostic ignored "-Wformat-nonliteral"
+#pragma GCC diagnostic ignored "-Wformat-zero-length"
+#pragma GCC diagnostic ignored "-Wmissing-field-initializers"
+#pragma GCC diagnostic ignored "-Wpacked" // Erratic impl in clang
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wtype-limits"
+#pragma GCC diagnostic ignored "-Wc++11-compat" // only gcc raises it
+#endif
+
+#endif
+#endif
+
+
+#include "hb-config.hh"
+
+
+/*
+ * Following added based on what AC_USE_SYSTEM_EXTENSIONS adds to
+ * config.h.in.  Copied here for the convenience of those embedding
+ * HarfBuzz and not using our build system.
+ */
+/* Enable extensions on AIX 3, Interix.  */
+#ifndef _ALL_SOURCE
+# define _ALL_SOURCE 1
+#endif
+/* Enable GNU extensions on systems that have them.  */
+#ifndef _GNU_SOURCE
+# define _GNU_SOURCE 1
+#endif
+/* Enable threading extensions on Solaris.  */
+#ifndef _POSIX_PTHREAD_SEMANTICS
+# define _POSIX_PTHREAD_SEMANTICS 1
+#endif
+/* Enable extensions on HP NonStop.  */
+#ifndef _TANDEM_SOURCE
+# define _TANDEM_SOURCE 1
+#endif
+/* Enable general extensions on Solaris.  */
+#ifndef __EXTENSIONS__
+# define __EXTENSIONS__ 1
+#endif
+
+#if defined (_MSC_VER) && defined (HB_DLL_EXPORT)
+#define HB_EXTERN __declspec (dllexport) extern
+#endif
+
+#include "hb.h"
+#define HB_H_IN
+#include "hb-ot.h"
+#define HB_OT_H_IN
+#include "hb-aat.h"
+#define HB_AAT_H_IN
+
+#include <limits.h>
+#include <math.h>
+#include <float.h>
+#include <stdlib.h>
+#include <stddef.h>
+#include <string.h>
+#include <assert.h>
+#include <stdio.h>
+#include <stdarg.h>
+
+#if (defined(_MSC_VER) && _MSC_VER >= 1500) || defined(__MINGW32__)
+#ifdef __MINGW32_VERSION
+#ifndef WIN32_LEAN_AND_MEAN
+#define WIN32_LEAN_AND_MEAN 1
+#endif
+#else
+#include <intrin.h>
+#endif
+#endif
+
+#ifdef _WIN32
+#include <windows.h>
+#include <winapifamily.h>
+#endif
+
+#define HB_PASTE1(a,b) a##b
+#define HB_PASTE(a,b) HB_PASTE1(a,b)
+
+
+/* Compile-time custom allocator support. */
+
+#if !defined(HB_CUSTOM_MALLOC) \
+  && defined(hb_malloc_impl) \
+  && defined(hb_calloc_impl) \
+  && defined(hb_realloc_impl) \
+  && defined(hb_free_impl)
+#define HB_CUSTOM_MALLOC
+#endif
+
+#ifdef HB_CUSTOM_MALLOC
+extern "C" void* hb_malloc_impl(size_t size);
+extern "C" void* hb_calloc_impl(size_t nmemb, size_t size);
+extern "C" void* hb_realloc_impl(void *ptr, size_t size);
+extern "C" void  hb_free_impl(void *ptr);
+#define malloc hb_malloc_impl
+#define calloc hb_calloc_impl
+#define realloc hb_realloc_impl
+#define free hb_free_impl
+#endif
+
+
+/*
+ * Compiler attributes
+ */
+
+#if (defined(__GNUC__) || defined(__clang__)) && defined(__OPTIMIZE__)
+#define likely(expr) (__builtin_expect (!!(expr), 1))
+#define unlikely(expr) (__builtin_expect (!!(expr), 0))
+#else
+#define likely(expr) (expr)
+#define unlikely(expr) (expr)
+#endif
+
+#if !defined(__GNUC__) && !defined(__clang__)
+#undef __attribute__
+#define __attribute__(x)
+#endif
+
+#if defined(__GNUC__) && (__GNUC__ >= 3)
+#define HB_PURE_FUNC	__attribute__((pure))
+#define HB_CONST_FUNC	__attribute__((const))
+#define HB_PRINTF_FUNC(format_idx, arg_idx) __attribute__((__format__ (__printf__, format_idx, arg_idx)))
+#else
+#define HB_PURE_FUNC
+#define HB_CONST_FUNC
+#define HB_PRINTF_FUNC(format_idx, arg_idx)
+#endif
+#if defined(__GNUC__) && (__GNUC__ >= 4) || (__clang__)
+#define HB_UNUSED	__attribute__((unused))
+#elif defined(_MSC_VER) /* https://github.com/harfbuzz/harfbuzz/issues/635 */
+#define HB_UNUSED __pragma(warning(suppress: 4100 4101))
+#else
+#define HB_UNUSED
+#endif
+
+#ifndef HB_INTERNAL
+# if !defined(HB_NO_VISIBILITY) && !defined(__MINGW32__) && !defined(__CYGWIN__) && !defined(_MSC_VER) && !defined(__SUNPRO_CC)
+#  define HB_INTERNAL __attribute__((__visibility__("hidden")))
+# elif defined(__MINGW32__)
+   /* We use -export-symbols on mingw32, since it does not support visibility attributes. */
+#  define HB_INTERNAL
+# elif defined (_MSC_VER) && defined (HB_DLL_EXPORT)
+   /* We do not try to export internal symbols on Visual Studio */
+#  define HB_INTERNAL
+#else
+#  define HB_INTERNAL
+#  define HB_NO_VISIBILITY 1
+# endif
+#endif
+
+/* https://github.com/harfbuzz/harfbuzz/issues/1651 */
+#if defined(__clang__) && __clang_major__ < 10
+#define static_const static
+#else
+#define static_const static const
+#endif
+
+#if defined(__GNUC__) && (__GNUC__ >= 3)
+#define HB_FUNC __PRETTY_FUNCTION__
+#elif defined(_MSC_VER)
+#define HB_FUNC __FUNCSIG__
+#else
+#define HB_FUNC __func__
+#endif
+
+#if defined(__SUNPRO_CC) && (__SUNPRO_CC < 0x5140)
+/* https://github.com/harfbuzz/harfbuzz/issues/630 */
+#define __restrict
+#endif
+
+/*
+ * Borrowed from https://bugzilla.mozilla.org/show_bug.cgi?id=1215411
+ * HB_FALLTHROUGH is an annotation to suppress compiler warnings about switch
+ * cases that fall through without a break or return statement. HB_FALLTHROUGH
+ * is only needed on cases that have code:
+ *
+ * switch (foo) {
+ *   case 1: // These cases have no code. No fallthrough annotations are needed.
+ *   case 2:
+ *   case 3:
+ *     foo = 4; // This case has code, so a fallthrough annotation is needed:
+ *     HB_FALLTHROUGH;
+ *   default:
+ *     return foo;
+ * }
+ */
+#if defined(__clang__) && __cplusplus >= 201103L
+   /* clang's fallthrough annotations are only available starting in C++11. */
+#  define HB_FALLTHROUGH [[clang::fallthrough]]
+#elif defined(__GNUC__) && (__GNUC__ >= 7)
+   /* GNU fallthrough attribute is available from GCC7 */
+#  define HB_FALLTHROUGH __attribute__((fallthrough))
+#elif defined(_MSC_VER)
+   /*
+    * MSVC's __fallthrough annotations are checked by /analyze (Code Analysis):
+    * https://msdn.microsoft.com/en-us/library/ms235402%28VS.80%29.aspx
+    */
+#  include <sal.h>
+#  define HB_FALLTHROUGH __fallthrough
+#else
+#  define HB_FALLTHROUGH /* FALLTHROUGH */
+#endif
+
+/* A tag to enforce use of return value for a function */
+#if __cplusplus >= 201703L
+#  define HB_NODISCARD [[nodiscard]]
+#elif defined(__GNUC__) || defined(__clang__)
+#  define HB_NODISCARD __attribute__((warn_unused_result))
+#elif defined(_MSC_VER)
+#  define HB_NODISCARD _Check_return_
+#else
+#  define HB_NODISCARD
+#endif
+#define hb_success_t HB_NODISCARD bool
+
+/* https://github.com/harfbuzz/harfbuzz/issues/1852 */
+#if defined(__clang__) && !(defined(_AIX) && (defined(__IBMCPP__) || defined(__ibmxl__)))
+/* Disable certain sanitizer errors. */
+/* https://github.com/harfbuzz/harfbuzz/issues/1247 */
+#define HB_NO_SANITIZE_SIGNED_INTEGER_OVERFLOW __attribute__((no_sanitize("signed-integer-overflow")))
+#else
+#define HB_NO_SANITIZE_SIGNED_INTEGER_OVERFLOW
+#endif
+
+
+#ifdef _WIN32
+   /* We need Windows Vista for both Uniscribe backend and for
+    * MemoryBarrier.  We don't support compiling on Windows XP,
+    * though we run on it fine. */
+#  if defined(_WIN32_WINNT) && _WIN32_WINNT < 0x0600
+#    undef _WIN32_WINNT
+#  endif
+#  ifndef _WIN32_WINNT
+#    if WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+#      define _WIN32_WINNT 0x0600
+#    endif
+#  endif
+#  ifndef WIN32_LEAN_AND_MEAN
+#    define WIN32_LEAN_AND_MEAN 1
+#  endif
+#  ifndef STRICT
+#    define STRICT 1
+#  endif
+
+#  if defined(_WIN32_WCE)
+     /* Some things not defined on Windows CE. */
+#    define vsnprintf _vsnprintf
+#    ifndef HB_NO_GETENV
+#      define HB_NO_GETENV
+#    endif
+#    if _WIN32_WCE < 0x800
+#      define HB_NO_SETLOCALE
+#      define HB_NO_ERRNO
+#    endif
+#  elif WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)
+#    ifndef HB_NO_GETENV
+#      define HB_NO_GETENV
+#    endif
+#  endif
+#  if defined(_MSC_VER) && _MSC_VER < 1900
+#    define snprintf _snprintf
+#  endif
+#endif
+
+#ifdef HB_NO_GETENV
+#define getenv(Name) nullptr
+#endif
+
+#ifndef HB_NO_ERRNO
+#  include <errno.h>
+#else
+static int HB_UNUSED _hb_errno = 0;
+#  undef errno
+#  define errno _hb_errno
+#endif
+
+#if defined(HAVE_ATEXIT) && !defined(HB_USE_ATEXIT)
+/* atexit() is only safe to be called from shared libraries on certain
+ * platforms.  Whitelist.
+ * https://bugs.freedesktop.org/show_bug.cgi?id=82246 */
+#  if defined(__linux) && defined(__GLIBC_PREREQ)
+#    if __GLIBC_PREREQ(2,3)
+/* From atexit() manpage, it's safe with glibc 2.2.3 on Linux. */
+#      define HB_USE_ATEXIT 1
+#    endif
+#  elif defined(_MSC_VER) || defined(__MINGW32__)
+/* For MSVC:
+ * https://msdn.microsoft.com/en-us/library/tze57ck3.aspx
+ * https://msdn.microsoft.com/en-us/library/zk17ww08.aspx
+ * mingw32 headers say atexit is safe to use in shared libraries.
+ */
+#    define HB_USE_ATEXIT 1
+#  elif defined(__ANDROID__)
+/* This is available since Android NKD r8 or r8b:
+ * https://issuetracker.google.com/code/p/android/issues/detail?id=6455
+ */
+#    define HB_USE_ATEXIT 1
+#  elif defined(__APPLE__)
+/* For macOS and related platforms, the atexit man page indicates
+ * that it will be invoked when the library is unloaded, not only
+ * at application exit.
+ */
+#    define HB_USE_ATEXIT 1
+#  endif
+#endif
+#ifdef HB_NO_ATEXIT
+#  undef HB_USE_ATEXIT
+#endif
+#ifndef HB_USE_ATEXIT
+#  define HB_USE_ATEXIT 0
+#endif
+
+#define HB_STMT_START do
+#define HB_STMT_END   while (0)
+
+/* Static-assert as expression. */
+template <unsigned int cond> class hb_assert_constant_t;
+template <> class hb_assert_constant_t<1> {};
+#define ASSERT_STATIC_EXPR_ZERO(_cond) (0 * (unsigned int) sizeof (hb_assert_constant_t<_cond>))
+
+/* Lets assert int types.  Saves trouble down the road. */
+static_assert ((sizeof (int8_t) == 1), "");
+static_assert ((sizeof (uint8_t) == 1), "");
+static_assert ((sizeof (int16_t) == 2), "");
+static_assert ((sizeof (uint16_t) == 2), "");
+static_assert ((sizeof (int32_t) == 4), "");
+static_assert ((sizeof (uint32_t) == 4), "");
+static_assert ((sizeof (int64_t) == 8), "");
+static_assert ((sizeof (uint64_t) == 8), "");
+static_assert ((sizeof (hb_codepoint_t) == 4), "");
+static_assert ((sizeof (hb_position_t) == 4), "");
+static_assert ((sizeof (hb_mask_t) == 4), "");
+static_assert ((sizeof (hb_var_int_t) == 4), "");
+
+#define HB_DELETE_COPY_ASSIGN(TypeName) \
+  TypeName(const TypeName&) = delete; \
+  void operator=(const TypeName&) = delete
+#define HB_DELETE_CREATE_COPY_ASSIGN(TypeName) \
+  TypeName() = delete; \
+  TypeName(const TypeName&) = delete; \
+  void operator=(const TypeName&) = delete
+
+
+/* Flags */
+
+/* Enable bitwise ops on enums marked as flags_t */
+/* To my surprise, looks like the function resolver is happy to silently cast
+ * one enum to another...  So this doesn't provide the type-checking that I
+ * originally had in mind... :(.
+ *
+ * For MSVC warnings, see: https://github.com/harfbuzz/harfbuzz/pull/163
+ */
+#ifdef _MSC_VER
+# pragma warning(disable:4200)
+# pragma warning(disable:4800)
+#endif
+#define HB_MARK_AS_FLAG_T(T) \
+	extern "C++" { \
+	  static inline T operator | (T l, T r) { return T ((unsigned) l | (unsigned) r); } \
+	  static inline T operator & (T l, T r) { return T ((unsigned) l & (unsigned) r); } \
+	  static inline T operator ^ (T l, T r) { return T ((unsigned) l ^ (unsigned) r); } \
+	  static inline T operator ~ (T r) { return T (~(unsigned int) r); } \
+	  static inline T& operator |= (T &l, T r) { l = l | r; return l; } \
+	  static inline T& operator &= (T& l, T r) { l = l & r; return l; } \
+	  static inline T& operator ^= (T& l, T r) { l = l ^ r; return l; } \
+	} \
+	static_assert (true, "")
+
+/* Useful for set-operations on small enums.
+ * For example, for testing "x ∈ {x1, x2, x3}" use:
+ * (FLAG_UNSAFE(x) & (FLAG(x1) | FLAG(x2) | FLAG(x3)))
+ */
+#define FLAG(x) (ASSERT_STATIC_EXPR_ZERO ((unsigned)(x) < 32) + (((uint32_t) 1U) << (unsigned)(x)))
+#define FLAG_UNSAFE(x) ((unsigned)(x) < 32 ? (((uint32_t) 1U) << (unsigned)(x)) : 0)
+#define FLAG_RANGE(x,y) (ASSERT_STATIC_EXPR_ZERO ((x) < (y)) + FLAG(y+1) - FLAG(x))
+#define FLAG64(x) (ASSERT_STATIC_EXPR_ZERO ((unsigned)(x) < 64) + (((uint64_t) 1ULL) << (unsigned)(x)))
+#define FLAG64_UNSAFE(x) ((unsigned)(x) < 64 ? (((uint64_t) 1ULL) << (unsigned)(x)) : 0)
+
+
+/* Size signifying variable-sized array */
+#ifndef HB_VAR_ARRAY
+#define HB_VAR_ARRAY 1
+#endif
+
+static inline float
+_hb_roundf (float x) { return floorf (x + .5f); }
+#define roundf(x) _hb_roundf(x)
+
+/* Endian swap, used in Windows related backends */
+static inline uint16_t hb_uint16_swap (const uint16_t v)
+{ return (v >> 8) | (v << 8); }
+static inline uint32_t hb_uint32_swap (const uint32_t v)
+{ return (hb_uint16_swap (v) << 16) | hb_uint16_swap (v >> 16); }
+
+/*
+ * Big-endian integers.  Here because fundamental.
+ */
+
+template <typename Type, int Bytes> struct BEInt;
+
+template <typename Type>
+struct BEInt<Type, 1>
+{
+  public:
+  BEInt<Type, 1>& operator = (Type V)
+  {
+    v = V;
+    return *this;
+  }
+  operator Type () const { return v; }
+  private: uint8_t v;
+};
+template <typename Type>
+struct BEInt<Type, 2>
+{
+  public:
+  BEInt<Type, 2>& operator = (Type V)
+  {
+    v[0] = (V >>  8) & 0xFF;
+    v[1] = (V      ) & 0xFF;
+    return *this;
+  }
+  operator Type () const
+  {
+#if ((defined(__GNUC__) && __GNUC__ >= 5) || defined(__clang__)) && \
+    defined(__BYTE_ORDER) && \
+    (__BYTE_ORDER == __LITTLE_ENDIAN || __BYTE_ORDER == __BIG_ENDIAN)
+    /* Spoon-feed the compiler a big-endian integer with alignment 1.
+     * https://github.com/harfbuzz/harfbuzz/pull/1398 */
+    struct __attribute__((packed)) packed_uint16_t { uint16_t v; };
+#if __BYTE_ORDER == __LITTLE_ENDIAN
+    return __builtin_bswap16 (((packed_uint16_t *) this)->v);
+#else /* __BYTE_ORDER == __BIG_ENDIAN */
+    return ((packed_uint16_t *) this)->v;
+#endif
+#endif
+    return (v[0] <<  8)
+	 + (v[1]      );
+  }
+  private: uint8_t v[2];
+};
+template <typename Type>
+struct BEInt<Type, 3>
+{
+  public:
+  BEInt<Type, 3>& operator = (Type V)
+  {
+    v[0] = (V >> 16) & 0xFF;
+    v[1] = (V >>  8) & 0xFF;
+    v[2] = (V      ) & 0xFF;
+    return *this;
+  }
+  operator Type () const
+  {
+    return (v[0] << 16)
+	 + (v[1] <<  8)
+	 + (v[2]      );
+  }
+  private: uint8_t v[3];
+};
+template <typename Type>
+struct BEInt<Type, 4>
+{
+  public:
+  BEInt<Type, 4>& operator = (Type V)
+  {
+    v[0] = (V >> 24) & 0xFF;
+    v[1] = (V >> 16) & 0xFF;
+    v[2] = (V >>  8) & 0xFF;
+    v[3] = (V      ) & 0xFF;
+    return *this;
+  }
+  operator Type () const
+  {
+    return (v[0] << 24)
+	 + (v[1] << 16)
+	 + (v[2] <<  8)
+	 + (v[3]      );
+  }
+  private: uint8_t v[4];
+};
+
+
+/*
+ * For lack of a better place, put Zawgyi script hack here.
+ * https://github.com/harfbuzz/harfbuzz/issues/1162
+ */
+
+#define HB_SCRIPT_MYANMAR_ZAWGYI	((hb_script_t) HB_TAG ('Q','a','a','g'))
+
+
+/* Headers we include for everyone.  Keep topologically sorted by dependency.
+ * They express dependency amongst themselves, but no other file should include
+ * them directly.*/
+#include "hb-meta.hh"
+#include "hb-mutex.hh"
+#include "hb-number.hh"
+#include "hb-atomic.hh"	// Requires: hb-meta
+#include "hb-null.hh"	// Requires: hb-meta
+#include "hb-algs.hh"	// Requires: hb-meta hb-null hb-number
+#include "hb-iter.hh"	// Requires: hb-algs hb-meta
+#include "hb-debug.hh"	// Requires: hb-algs hb-atomic
+#include "hb-array.hh"	// Requires: hb-algs hb-iter hb-null
+#include "hb-vector.hh"	// Requires: hb-array hb-null
+#include "hb-object.hh"	// Requires: hb-atomic hb-mutex hb-vector
+
+#endif /* HB_HH */
diff --git a/thirdparty/icu4c/APIChangeReport.md b/thirdparty/icu4c/APIChangeReport.md
new file mode 100644
index 0000000000..0cf9ed5bfc
--- /dev/null
+++ b/thirdparty/icu4c/APIChangeReport.md
@@ -0,0 +1,380 @@
+
+  
+<!--
+ Copyright © 2019 and later: Unicode, Inc. and others.
+ License & terms of use: http://www.unicode.org/copyright.html
+-->
+
+# ICU4C API Comparison: ICU 67 with ICU 68
+
+> _Note_ Markdown format of this document is new for ICU 65.
+
+- [Removed from ICU 67](#removed)
+- [Deprecated or Obsoleted in ICU 68](#deprecated)
+- [Changed in  ICU 68](#changed)
+- [Promoted to stable in ICU 68](#promoted)
+- [Added in ICU 68](#added)
+- [Other existing drafts in ICU 68](#other)
+- [Signature Simplifications](#simplifications)
+
+## Removed
+
+Removed from ICU 67
+  
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| measunit.h | LocalArray&lt;MeasureUnit&gt; icu::MeasureUnit::splitToSingleUnits(int32_t&amp;, UErrorCode&amp;) const |  InternalICU 67 | (missing)
+| measunit.h | int32_t icu::MeasureUnit::getIndex() const |  Internal | (missing)
+| measunit.h | <tt>static</tt> int32_t icu::MeasureUnit::getIndexCount() |  Internal | (missing)
+| measunit.h | <tt>static</tt> int32_t icu::MeasureUnit::internalGetIndexForTypeAndSubtype(const char*, const char*) |  Internal | (missing)
+| nounit.h | UClassID icu::NoUnit::getDynamicClassID() const |  DraftICU 60 | (missing)
+| nounit.h | icu::NoUnit::NoUnit(const NoUnit&amp;) |  DraftICU 60 | (missing)
+| nounit.h | icu::NoUnit::~NoUnit() |  DraftICU 60 | (missing)
+| nounit.h | <tt>static</tt> NoUnit icu::NoUnit::base() |  DraftICU 60 | (missing)
+| nounit.h | <tt>static</tt> NoUnit icu::NoUnit::percent() |  DraftICU 60 | (missing)
+| nounit.h | <tt>static</tt> NoUnit icu::NoUnit::permille() |  DraftICU 60 | (missing)
+| nounit.h | <tt>static</tt> UClassID icu::NoUnit::getStaticClassID() |  DraftICU 60 | (missing)
+| nounit.h | void* icu::NoUnit::clone() const |  DraftICU 60 | (missing)
+
+## Deprecated
+
+Deprecated or Obsoleted in ICU 68
+  
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| numberrangeformatter.h | UnicodeString icu::number::FormattedNumberRange::getFirstDecimal(UErrorCode&amp;) const |  DraftICU 63 | DeprecatedICU 68
+| numberrangeformatter.h | UnicodeString icu::number::FormattedNumberRange::getSecondDecimal(UErrorCode&amp;) const |  DraftICU 63 | DeprecatedICU 68
+| umachine.h | <tt>#define</tt> FALSE |  StableICU 2.0 | DeprecatedICU 68
+| umachine.h | <tt>#define</tt> TRUE |  StableICU 2.0 | DeprecatedICU 68
+
+## Changed
+
+Changed in  ICU 68 (old, new)
+
+
+  
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| bytestrie.h | BytesTrie&amp; icu::BytesTrie::resetToState64(uint64_t) |  Draft→StableICU 65
+| bytestrie.h | uint64_t icu::BytesTrie::getState64() const |  Draft→StableICU 65
+| localebuilder.h | UBool icu::LocaleBuilder::copyErrorTo(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::addSupportedLocale(const Locale&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::operator=(Builder&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setDefaultLocale(const Locale*) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setDemotionPerDesiredLocale(ULocMatchDemotion) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setFavorSubtag(ULocMatchFavorSubtag) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocales(Iter, Iter) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocales(Locale::Iterator&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocalesFromListString(StringPiece) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocalesViaConverter(Iter, Iter, Conv) |  Draft→StableICU 65
+| localematcher.h | Locale icu::LocaleMatcher::Result::makeResolvedLocale(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | LocaleMatcher icu::LocaleMatcher::Builder::build(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | LocaleMatcher&amp; icu::LocaleMatcher::operator=(LocaleMatcher&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | Result icu::LocaleMatcher::getBestMatchResult(Locale::Iterator&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Result icu::LocaleMatcher::getBestMatchResult(const Locale&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Result&amp; icu::LocaleMatcher::Result::operator=(Result&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | UBool icu::LocaleMatcher::Builder::copyErrorTo(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::Result::getDesiredLocale() const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::Result::getSupportedLocale() const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatch(Locale::Iterator&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatch(const Locale&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatchForListString(StringPiece, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchDemotion::ULOCMATCH_DEMOTION_NONE |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchDemotion::ULOCMATCH_DEMOTION_REGION |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchFavorSubtag::ULOCMATCH_FAVOR_LANGUAGE |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchFavorSubtag::ULOCMATCH_FAVOR_SCRIPT |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::Builder() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::Builder(Builder&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::~Builder() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::LocaleMatcher(LocaleMatcher&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Result::Result(Result&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Result::~Result() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::~LocaleMatcher() |  Draft→StableICU 65
+| localematcher.h | int32_t icu::LocaleMatcher::Result::getDesiredIndex() const |  Draft→StableICU 65
+| localematcher.h | int32_t icu::LocaleMatcher::Result::getSupportedIndex() const |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::hasNext() const override |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::Iterator::hasNext() const |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::RangeIterator&lt; Iter &gt;::hasNext() const override |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::next() override |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::Iterator::next() |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::RangeIterator&lt; Iter &gt;::next() override |  Draft→StableICU 65
+| locid.h | icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::ConvertingIterator(Iter, Iter, Conv) |  Draft→StableICU 65
+| locid.h | icu::Locale::Iterator::~Iterator() |  Draft→StableICU 65
+| locid.h | icu::Locale::RangeIterator&lt; Iter &gt;::RangeIterator(Iter, Iter) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getBar() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDecade() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDotPerCentimeter() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDotPerInch() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getEm() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getMegapixel() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPascal() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixel() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixelPerCentimeter() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixelPerInch() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getThermUs() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createBar(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDecade(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDotPerCentimeter(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDotPerInch(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createEm(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createMegapixel(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPascal(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixel(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixelPerCentimeter(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixelPerInch(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createThermUs(UErrorCode&amp;) |  Draft→StableICU 65
+| numberformatter.h | StringClass icu::number::FormattedNumber::toDecimalNumber(UErrorCode&amp;) const |  Draft→StableICU 65
+| numberrangeformatter.h | UnicodeString icu::number::FormattedNumberRange::getFirstDecimal(UErrorCode&amp;) const |  DraftICU 63 | DeprecatedICU 68
+| numberrangeformatter.h | UnicodeString icu::number::FormattedNumberRange::getSecondDecimal(UErrorCode&amp;) const |  DraftICU 63 | DeprecatedICU 68
+| reldatefmt.h | <tt>enum</tt> UDateAbsoluteUnit::UDAT_ABSOLUTE_HOUR |  Draft→StableICU 65
+| reldatefmt.h | <tt>enum</tt> UDateAbsoluteUnit::UDAT_ABSOLUTE_MINUTE |  Draft→StableICU 65
+| stringpiece.h | icu::StringPiece::StringPiece(T) |  Draft→StableICU 65
+| ucal.h | int32_t ucal_getHostTimeZone(UChar*, int32_t, UErrorCode*) |  Draft→StableICU 65
+| ucharstrie.h | UCharsTrie&amp; icu::UCharsTrie::resetToState64(uint64_t) |  Draft→StableICU 65
+| ucharstrie.h | uint64_t icu::UCharsTrie::getState64() const |  Draft→StableICU 65
+| uloc.h | UEnumeration* uloc_openAvailableByType(ULocAvailableType, UErrorCode*) |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_DEFAULT |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_ONLY_LEGACY_ALIASES |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_WITH_LEGACY_ALIASES |  Draft→StableICU 65
+| umachine.h | <tt>#define</tt> FALSE |  StableICU 2.0 | DeprecatedICU 68
+| umachine.h | <tt>#define</tt> TRUE |  StableICU 2.0 | DeprecatedICU 68
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_BUNDLE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_DATA_FILE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_RES_FILE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_START |  Draft→StableICU 65
+
+## Promoted
+
+Promoted to stable in ICU 68
+  
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| bytestrie.h | BytesTrie&amp; icu::BytesTrie::resetToState64(uint64_t) |  Draft→StableICU 65
+| bytestrie.h | uint64_t icu::BytesTrie::getState64() const |  Draft→StableICU 65
+| localebuilder.h | UBool icu::LocaleBuilder::copyErrorTo(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::addSupportedLocale(const Locale&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::operator=(Builder&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setDefaultLocale(const Locale*) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setDemotionPerDesiredLocale(ULocMatchDemotion) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setFavorSubtag(ULocMatchFavorSubtag) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocales(Iter, Iter) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocales(Locale::Iterator&amp;) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocalesFromListString(StringPiece) |  Draft→StableICU 65
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setSupportedLocalesViaConverter(Iter, Iter, Conv) |  Draft→StableICU 65
+| localematcher.h | Locale icu::LocaleMatcher::Result::makeResolvedLocale(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | LocaleMatcher icu::LocaleMatcher::Builder::build(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | LocaleMatcher&amp; icu::LocaleMatcher::operator=(LocaleMatcher&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | Result icu::LocaleMatcher::getBestMatchResult(Locale::Iterator&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Result icu::LocaleMatcher::getBestMatchResult(const Locale&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | Result&amp; icu::LocaleMatcher::Result::operator=(Result&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | UBool icu::LocaleMatcher::Builder::copyErrorTo(UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::Result::getDesiredLocale() const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::Result::getSupportedLocale() const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatch(Locale::Iterator&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatch(const Locale&amp;, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | const Locale* icu::LocaleMatcher::getBestMatchForListString(StringPiece, UErrorCode&amp;) const |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchDemotion::ULOCMATCH_DEMOTION_NONE |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchDemotion::ULOCMATCH_DEMOTION_REGION |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchFavorSubtag::ULOCMATCH_FAVOR_LANGUAGE |  Draft→StableICU 65
+| localematcher.h | <tt>enum</tt> ULocMatchFavorSubtag::ULOCMATCH_FAVOR_SCRIPT |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::Builder() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::Builder(Builder&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Builder::~Builder() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::LocaleMatcher(LocaleMatcher&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Result::Result(Result&amp;&amp;) |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::Result::~Result() |  Draft→StableICU 65
+| localematcher.h | icu::LocaleMatcher::~LocaleMatcher() |  Draft→StableICU 65
+| localematcher.h | int32_t icu::LocaleMatcher::Result::getDesiredIndex() const |  Draft→StableICU 65
+| localematcher.h | int32_t icu::LocaleMatcher::Result::getSupportedIndex() const |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::hasNext() const override |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::Iterator::hasNext() const |  Draft→StableICU 65
+| locid.h | UBool icu::Locale::RangeIterator&lt; Iter &gt;::hasNext() const override |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::next() override |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::Iterator::next() |  Draft→StableICU 65
+| locid.h | const Locale&amp; icu::Locale::RangeIterator&lt; Iter &gt;::next() override |  Draft→StableICU 65
+| locid.h | icu::Locale::ConvertingIterator&lt; Iter, Conv &gt;::ConvertingIterator(Iter, Iter, Conv) |  Draft→StableICU 65
+| locid.h | icu::Locale::Iterator::~Iterator() |  Draft→StableICU 65
+| locid.h | icu::Locale::RangeIterator&lt; Iter &gt;::RangeIterator(Iter, Iter) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getBar() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDecade() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDotPerCentimeter() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDotPerInch() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getEm() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getMegapixel() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPascal() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixel() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixelPerCentimeter() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPixelPerInch() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getThermUs() |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createBar(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDecade(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDotPerCentimeter(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDotPerInch(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createEm(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createMegapixel(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPascal(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixel(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixelPerCentimeter(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPixelPerInch(UErrorCode&amp;) |  Draft→StableICU 65
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createThermUs(UErrorCode&amp;) |  Draft→StableICU 65
+| numberformatter.h | StringClass icu::number::FormattedNumber::toDecimalNumber(UErrorCode&amp;) const |  Draft→StableICU 65
+| reldatefmt.h | <tt>enum</tt> UDateAbsoluteUnit::UDAT_ABSOLUTE_HOUR |  Draft→StableICU 65
+| reldatefmt.h | <tt>enum</tt> UDateAbsoluteUnit::UDAT_ABSOLUTE_MINUTE |  Draft→StableICU 65
+| stringpiece.h | icu::StringPiece::StringPiece(T) |  Draft→StableICU 65
+| ucal.h | int32_t ucal_getHostTimeZone(UChar*, int32_t, UErrorCode*) |  Draft→StableICU 65
+| ucharstrie.h | UCharsTrie&amp; icu::UCharsTrie::resetToState64(uint64_t) |  Draft→StableICU 65
+| ucharstrie.h | uint64_t icu::UCharsTrie::getState64() const |  Draft→StableICU 65
+| uloc.h | UEnumeration* uloc_openAvailableByType(ULocAvailableType, UErrorCode*) |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_DEFAULT |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_ONLY_LEGACY_ALIASES |  Draft→StableICU 65
+| uloc.h | <tt>enum</tt> ULocAvailableType::ULOC_AVAILABLE_WITH_LEGACY_ALIASES |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_BUNDLE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_DATA_FILE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_RES_FILE |  Draft→StableICU 65
+| utrace.h | <tt>enum</tt> UTraceFunctionNumber::UTRACE_UDATA_START |  Draft→StableICU 65
+
+## Added
+
+Added in ICU 68
+  
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| dtitvfmt.h | UDisplayContext icu::DateIntervalFormat::getContext(UDisplayContextType, UErrorCode&amp;) const |  (missing) | DraftICU 68
+| dtitvfmt.h | void icu::DateIntervalFormat::setContext(UDisplayContext, UErrorCode&amp;) |  (missing) | DraftICU 68
+| dtptngen.h | <tt>static</tt> DateTimePatternGenerator* icu::DateTimePatternGenerator::createInstanceNoStdPat(const Locale&amp;, UErrorCode&amp;) |  (missing) | Internal
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setMaxDistance(const Locale&amp;, const Locale&amp;) |  (missing) | DraftICU 68
+| localematcher.h | Builder&amp; icu::LocaleMatcher::Builder::setNoDefaultLocale() |  (missing) | DraftICU 68
+| localematcher.h | UBool icu::LocaleMatcher::isMatch(const Locale&amp;, const Locale&amp;, UErrorCode&amp;) const |  (missing) | DraftICU 68
+| measunit.h | int32_t icu::MeasureUnit::getOffset() const |  (missing) | Internal
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getCandela() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDessertSpoon() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDessertSpoonImperial() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDot() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDram() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getDrop() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getEarthRadius() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getGrain() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getJigger() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getLumen() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getPinch() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit icu::MeasureUnit::getQuartImperial() |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createCandela(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDessertSpoon(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDessertSpoonImperial(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDot(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDram(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createDrop(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createEarthRadius(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createGrain(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createJigger(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createLumen(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createPinch(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | <tt>static</tt> MeasureUnit* icu::MeasureUnit::createQuartImperial(UErrorCode&amp;) |  (missing) | DraftICU 68
+| measunit.h | std::pair&lt; LocalArray&lt; MeasureUnit &gt;, int32_t &gt; icu::MeasureUnit::splitToSingleUnits(UErrorCode&amp;) const |  (missing) | DraftICU 68
+| numberformatter.h | Derived icu::number::NumberFormatterSettings&lt; Derived &gt;::usage(StringPiece) const&amp; |  (missing) | DraftICU 68
+| numberformatter.h | Derived icu::number::NumberFormatterSettings&lt; Derived &gt;::usage(StringPiece)&amp;&amp; |  (missing) | DraftICU 68
+| numberformatter.h | MeasureUnit icu::number::FormattedNumber::getOutputUnit(UErrorCode&amp;) const |  (missing) | DraftICU 68
+| numberformatter.h | Usage&amp; icu::number::impl::Usage::operator=(Usage&amp;&amp;) |  (missing) | Internal
+| numberformatter.h | Usage&amp; icu::number::impl::Usage::operator=(const Usage&amp;) |  (missing) | Internal
+| numberformatter.h | bool icu::number::impl::Usage::isSet() const |  (missing) | Internal
+| numberformatter.h | icu::number::impl::Usage::Usage(Usage&amp;&amp;) |  (missing) | Internal
+| numberformatter.h | icu::number::impl::Usage::Usage(const Usage&amp;) |  (missing) | Internal
+| numberformatter.h | icu::number::impl::Usage::~Usage() |  (missing) | Internal
+| numberformatter.h | int16_t icu::number::impl::Usage::length() const |  (missing) | Internal
+| numberformatter.h | void icu::number::impl::Usage::set(StringPiece) |  (missing) | Internal
+| numberrangeformatter.h | std::pair&lt; StringClass, StringClass &gt; icu::number::FormattedNumberRange::getDecimalNumbers(UErrorCode&amp;) const |  (missing) | DraftICU 68
+| plurrule.h | UnicodeString icu::PluralRules::select(const number::FormattedNumberRange&amp;, UErrorCode&amp;) const |  (missing) | DraftICU 68
+| plurrule.h | UnicodeString icu::PluralRules::select(const number::impl::UFormattedNumberRangeData*, UErrorCode&amp;) const |  (missing) | Internal
+| timezone.h | <tt>static</tt> TimeZone* icu::TimeZone::forLocaleOrDefault(const Locale&amp;) |  (missing) | Internal
+| ucurr.h | <tt>enum</tt> UCurrNameStyle::UCURR_FORMAL_SYMBOL_NAME |  (missing) | DraftICU 68
+| ucurr.h | <tt>enum</tt> UCurrNameStyle::UCURR_VARIANT_SYMBOL_NAME |  (missing) | DraftICU 68
+| udateintervalformat.h | UDisplayContext udtitvfmt_getContext(const UDateIntervalFormat*, UDisplayContextType, UErrorCode*) |  (missing) | DraftICU 68
+| udateintervalformat.h | void udtitvfmt_setContext(UDateIntervalFormat*, UDisplayContext, UErrorCode*) |  (missing) | DraftICU 68
+| umachine.h | <tt>#define</tt> U_DEFINE_FALSE_AND_TRUE |  (missing) | InternalICU 68
+| unum.h | <tt>enum</tt> UNumberFormatMinimumGroupingDigits::UNUM_MINIMUM_GROUPING_DIGITS_AUTO |  (missing) | DraftICU 68
+| unum.h | <tt>enum</tt> UNumberFormatMinimumGroupingDigits::UNUM_MINIMUM_GROUPING_DIGITS_MIN2 |  (missing) | DraftICU 68
+| unumberformatter.h | <tt>enum</tt> UNumberUnitWidth::UNUM_UNIT_WIDTH_FORMAL |  (missing) | DraftICU 68
+| unumberformatter.h | <tt>enum</tt> UNumberUnitWidth::UNUM_UNIT_WIDTH_VARIANT |  (missing) | DraftICU 68
+| unumberformatter.h | int32_t unumf_resultToDecimalNumber(const UFormattedNumber*, char*, int32_t, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | UFormattedNumberRange* unumrf_openResult(UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | UNumberRangeFormatter* unumrf_openForSkeletonWithCollapseAndIdentityFallback(const UChar*, int32_t, UNumberRangeCollapse, UNumberRangeIdentityFallback, const char*, UParseError*, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | UNumberRangeIdentityResult unumrf_resultGetIdentityResult(const UFormattedNumberRange*, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | const UFormattedValue* unumrf_resultAsValue(const UFormattedNumberRange*, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | int32_t unumrf_resultGetFirstDecimalNumber(const UFormattedNumberRange*, char*, int32_t, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | int32_t unumrf_resultGetSecondDecimalNumber(const UFormattedNumberRange*, char*, int32_t, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | void unumrf_close(UNumberRangeFormatter*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | void unumrf_closeResult(UFormattedNumberRange*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | void unumrf_formatDecimalRange(const UNumberRangeFormatter*, const char*, int32_t, const char*, int32_t, UFormattedNumberRange*, UErrorCode*) |  (missing) | DraftICU 68
+| unumberrangeformatter.h | void unumrf_formatDoubleRange(const UNumberRangeFormatter*, double, double, UFormattedNumberRange*, UErrorCode*) |  (missing) | DraftICU 68
+| upluralrules.h | int32_t uplrules_selectForRange(const UPluralRules*, const struct UFormattedNumberRange*, UChar*, int32_t, UErrorCode*) |  (missing) | DraftICU 68
+
+## Other
+
+Other existing drafts in ICU 68
+
+| File | API | ICU 67 | ICU 68 |
+|---|---|---|---|
+| bytestream.h |  void icu::ByteSink::AppendU8(const char*, int32_t) | DraftICU 67 | 
+| bytestream.h |  void icu::ByteSink::AppendU8(const char8_t*, int32_t) | DraftICU 67 | 
+| dtptngen.h |  UDateFormatHourCycle icu::DateTimePatternGenerator::getDefaultHourCycle(UErrorCode&amp;) const | DraftICU 67 | 
+| listformatter.h |  <tt>static</tt> ListFormatter* icu::ListFormatter::createInstance(const Locale&amp;, UListFormatterType, UListFormatterWidth, UErrorCode&amp;) | DraftICU 67 | 
+| localematcher.h |  Builder&amp; icu::LocaleMatcher::Builder::setDirection(ULocMatchDirection) | DraftICU 67 | 
+| localematcher.h |  <tt>enum</tt> ULocMatchDirection::ULOCMATCH_DIRECTION_ONLY_TWO_WAY | DraftICU 67 | 
+| localematcher.h |  <tt>enum</tt> ULocMatchDirection::ULOCMATCH_DIRECTION_WITH_ONE_WAY | DraftICU 67 | 
+| locid.h |  void icu::Locale::canonicalize(UErrorCode&amp;) | DraftICU 67 | 
+| measfmt.h |  void icu::MeasureFormat::parseObject(const UnicodeString&amp;, Formattable&amp;, ParsePosition&amp;) const | DraftICU 53 | 
+| measunit.h |  MeasureUnit icu::MeasureUnit::product(const MeasureUnit&amp;, UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  MeasureUnit icu::MeasureUnit::reciprocal(UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  MeasureUnit icu::MeasureUnit::withDimensionality(int32_t, UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  MeasureUnit icu::MeasureUnit::withSIPrefix(UMeasureSIPrefix, UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  MeasureUnit&amp; icu::MeasureUnit::operator=(MeasureUnit&amp;&amp;) noexcept | DraftICU 67 | 
+| measunit.h |  UMeasureSIPrefix icu::MeasureUnit::getSIPrefix(UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  UMeasureUnitComplexity icu::MeasureUnit::getComplexity(UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  const char* icu::MeasureUnit::getIdentifier() const | DraftICU 67 | 
+| measunit.h |  icu::MeasureUnit::MeasureUnit(MeasureUnit&amp;&amp;) noexcept | DraftICU 67 | 
+| measunit.h |  int32_t icu::MeasureUnit::getDimensionality(UErrorCode&amp;) const | DraftICU 67 | 
+| measunit.h |  <tt>static</tt> MeasureUnit icu::MeasureUnit::forIdentifier(StringPiece, UErrorCode&amp;) | DraftICU 67 | 
+| stringpiece.h |  icu::StringPiece::StringPiece(const char8_t*) | DraftICU 67 | 
+| stringpiece.h |  icu::StringPiece::StringPiece(const char8_t*, int32_t) | DraftICU 67 | 
+| stringpiece.h |  icu::StringPiece::StringPiece(const std::u8string&amp;) | DraftICU 67 | 
+| stringpiece.h |  icu::StringPiece::StringPiece(std::nullptr_t) | DraftICU 67 | 
+| stringpiece.h |  int32_t icu::StringPiece::compare(StringPiece) | DraftICU 67 | 
+| stringpiece.h |  int32_t icu::StringPiece::find(StringPiece, int32_t) | DraftICU 67 | 
+| stringpiece.h |  void icu::StringPiece::set(const char8_t*) | DraftICU 67 | 
+| stringpiece.h |  void icu::StringPiece::set(const char8_t*, int32_t) | DraftICU 67 | 
+| udat.h |  <tt>enum</tt> UDateFormatHourCycle::UDAT_HOUR_CYCLE_11 | DraftICU 67 | 
+| udat.h |  <tt>enum</tt> UDateFormatHourCycle::UDAT_HOUR_CYCLE_12 | DraftICU 67 | 
+| udat.h |  <tt>enum</tt> UDateFormatHourCycle::UDAT_HOUR_CYCLE_23 | DraftICU 67 | 
+| udat.h |  <tt>enum</tt> UDateFormatHourCycle::UDAT_HOUR_CYCLE_24 | DraftICU 67 | 
+| udateintervalformat.h |  void udtitvfmt_formatCalendarToResult(const UDateIntervalFormat*, UCalendar*, UCalendar*, UFormattedDateInterval*, UErrorCode*) | DraftICU 67 | 
+| udateintervalformat.h |  void udtitvfmt_formatToResult(const UDateIntervalFormat*, UDate, UDate, UFormattedDateInterval*, UErrorCode*) | DraftICU 67 | 
+| udatpg.h |  UDateFormatHourCycle udatpg_getDefaultHourCycle(const UDateTimePatternGenerator*, UErrorCode*) | DraftICU 67 | 
+| ulistformatter.h |  UListFormatter* ulistfmt_openForType(const char*, UListFormatterType, UListFormatterWidth, UErrorCode*) | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterType::ULISTFMT_TYPE_AND | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterType::ULISTFMT_TYPE_OR | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterType::ULISTFMT_TYPE_UNITS | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterWidth::ULISTFMT_WIDTH_NARROW | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterWidth::ULISTFMT_WIDTH_SHORT | DraftICU 67 | 
+| ulistformatter.h |  <tt>enum</tt> UListFormatterWidth::ULISTFMT_WIDTH_WIDE | DraftICU 67 | 
+| uregex.h |  <tt>enum</tt> URegexpFlag::UREGEX_CANON_EQ | DraftICU 2.4 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_BREAK_ENGINE | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_CHARACTER | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_LINE | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_SENTENCE | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_TITLE | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_CREATE_WORD | DraftICU 67 | 
+| utrace.h |  <tt>enum</tt> UTraceFunctionNumber::UTRACE_UBRK_START | DraftICU 67 | 
+
+## Simplifications
+
+This section shows cases where the signature was "simplified" for the sake of comparison. The simplified form is in bold, followed by
+    all possible variations in "original" form.
+
+
+## Colophon
+
+Contents generated by StableAPI tool on Wed Sep 30 17:44:26 PDT 2020
+
+Copyright © 2019 and later: Unicode, Inc. and others.
+License & terms of use: http://www.unicode.org/copyright.html
+  
\ No newline at end of file
diff --git a/thirdparty/icu4c/LICENSE b/thirdparty/icu4c/LICENSE
new file mode 100644
index 0000000000..5d664a083b
--- /dev/null
+++ b/thirdparty/icu4c/LICENSE
@@ -0,0 +1,414 @@
+COPYRIGHT AND PERMISSION NOTICE (ICU 58 and later)
+
+Copyright © 1991-2020 Unicode, Inc. All rights reserved.
+Distributed under the Terms of Use in https://www.unicode.org/copyright.html.
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of the Unicode data files and any associated documentation
+(the "Data Files") or Unicode software and any associated documentation
+(the "Software") to deal in the Data Files or Software
+without restriction, including without limitation the rights to use,
+copy, modify, merge, publish, distribute, and/or sell copies of
+the Data Files or Software, and to permit persons to whom the Data Files
+or Software are furnished to do so, provided that either
+(a) this copyright and permission notice appear with all copies
+of the Data Files or Software, or
+(b) this copyright and permission notice appear in associated
+Documentation.
+
+THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF
+ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
+WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT OF THIRD PARTY RIGHTS.
+IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS
+NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL
+DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
+PERFORMANCE OF THE DATA FILES OR SOFTWARE.
+
+Except as contained in this notice, the name of a copyright holder
+shall not be used in advertising or otherwise to promote the sale,
+use or other dealings in these Data Files or Software without prior
+written authorization of the copyright holder.
+
+---------------------
+
+Third-Party Software Licenses
+
+This section contains third-party software notices and/or additional
+terms for licensed third-party software components included within ICU
+libraries.
+
+1. ICU License - ICU 1.8.1 to ICU 57.1
+
+COPYRIGHT AND PERMISSION NOTICE
+
+Copyright (c) 1995-2016 International Business Machines Corporation and others
+All rights reserved.
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, and/or sell copies of the Software, and to permit persons
+to whom the Software is furnished to do so, provided that the above
+copyright notice(s) and this permission notice appear in all copies of
+the Software and that both the above copyright notice(s) and this
+permission notice appear in supporting documentation.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT
+OF THIRD PARTY RIGHTS. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+HOLDERS INCLUDED IN THIS NOTICE BE LIABLE FOR ANY CLAIM, OR ANY
+SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES, OR ANY DAMAGES WHATSOEVER
+RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF
+CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN
+CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+
+Except as contained in this notice, the name of a copyright holder
+shall not be used in advertising or otherwise to promote the sale, use
+or other dealings in this Software without prior written authorization
+of the copyright holder.
+
+All trademarks and registered trademarks mentioned herein are the
+property of their respective owners.
+
+2. Chinese/Japanese Word Break Dictionary Data (cjdict.txt)
+
+ #     The Google Chrome software developed by Google is licensed under
+ # the BSD license. Other software included in this distribution is
+ # provided under other licenses, as set forth below.
+ #
+ #  The BSD License
+ #  http://opensource.org/licenses/bsd-license.php
+ #  Copyright (C) 2006-2008, Google Inc.
+ #
+ #  All rights reserved.
+ #
+ #  Redistribution and use in source and binary forms, with or without
+ # modification, are permitted provided that the following conditions are met:
+ #
+ #  Redistributions of source code must retain the above copyright notice,
+ # this list of conditions and the following disclaimer.
+ #  Redistributions in binary form must reproduce the above
+ # copyright notice, this list of conditions and the following
+ # disclaimer in the documentation and/or other materials provided with
+ # the distribution.
+ #  Neither the name of  Google Inc. nor the names of its
+ # contributors may be used to endorse or promote products derived from
+ # this software without specific prior written permission.
+ #
+ #
+ #  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
+ # CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ # INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ # MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ # DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ # LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ # CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ # SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ # BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ # LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ # NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ # SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ #
+ #
+ #  The word list in cjdict.txt are generated by combining three word lists
+ # listed below with further processing for compound word breaking. The
+ # frequency is generated with an iterative training against Google web
+ # corpora.
+ #
+ #  * Libtabe (Chinese)
+ #    - https://sourceforge.net/project/?group_id=1519
+ #    - Its license terms and conditions are shown below.
+ #
+ #  * IPADIC (Japanese)
+ #    - http://chasen.aist-nara.ac.jp/chasen/distribution.html
+ #    - Its license terms and conditions are shown below.
+ #
+ #  ---------COPYING.libtabe ---- BEGIN--------------------
+ #
+ #  /*
+ #   * Copyright (c) 1999 TaBE Project.
+ #   * Copyright (c) 1999 Pai-Hsiang Hsiao.
+ #   * All rights reserved.
+ #   *
+ #   * Redistribution and use in source and binary forms, with or without
+ #   * modification, are permitted provided that the following conditions
+ #   * are met:
+ #   *
+ #   * . Redistributions of source code must retain the above copyright
+ #   *   notice, this list of conditions and the following disclaimer.
+ #   * . Redistributions in binary form must reproduce the above copyright
+ #   *   notice, this list of conditions and the following disclaimer in
+ #   *   the documentation and/or other materials provided with the
+ #   *   distribution.
+ #   * . Neither the name of the TaBE Project nor the names of its
+ #   *   contributors may be used to endorse or promote products derived
+ #   *   from this software without specific prior written permission.
+ #   *
+ #   * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ #   * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ #   * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ #   * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ #   * REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+ #   * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ #   * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ #   * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ #   * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ #   * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ #   * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ #   * OF THE POSSIBILITY OF SUCH DAMAGE.
+ #   */
+ #
+ #  /*
+ #   * Copyright (c) 1999 Computer Systems and Communication Lab,
+ #   *                    Institute of Information Science, Academia
+ #       *                    Sinica. All rights reserved.
+ #   *
+ #   * Redistribution and use in source and binary forms, with or without
+ #   * modification, are permitted provided that the following conditions
+ #   * are met:
+ #   *
+ #   * . Redistributions of source code must retain the above copyright
+ #   *   notice, this list of conditions and the following disclaimer.
+ #   * . Redistributions in binary form must reproduce the above copyright
+ #   *   notice, this list of conditions and the following disclaimer in
+ #   *   the documentation and/or other materials provided with the
+ #   *   distribution.
+ #   * . Neither the name of the Computer Systems and Communication Lab
+ #   *   nor the names of its contributors may be used to endorse or
+ #   *   promote products derived from this software without specific
+ #   *   prior written permission.
+ #   *
+ #   * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ #   * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ #   * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ #   * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ #   * REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+ #   * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ #   * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ #   * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ #   * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ #   * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ #   * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ #   * OF THE POSSIBILITY OF SUCH DAMAGE.
+ #   */
+ #
+ #  Copyright 1996 Chih-Hao Tsai @ Beckman Institute,
+ #      University of Illinois
+ #  c-tsai4@uiuc.edu  http://casper.beckman.uiuc.edu/~c-tsai4
+ #
+ #  ---------------COPYING.libtabe-----END--------------------------------
+ #
+ #
+ #  ---------------COPYING.ipadic-----BEGIN-------------------------------
+ #
+ #  Copyright 2000, 2001, 2002, 2003 Nara Institute of Science
+ #  and Technology.  All Rights Reserved.
+ #
+ #  Use, reproduction, and distribution of this software is permitted.
+ #  Any copy of this software, whether in its original form or modified,
+ #  must include both the above copyright notice and the following
+ #  paragraphs.
+ #
+ #  Nara Institute of Science and Technology (NAIST),
+ #  the copyright holders, disclaims all warranties with regard to this
+ #  software, including all implied warranties of merchantability and
+ #  fitness, in no event shall NAIST be liable for
+ #  any special, indirect or consequential damages or any damages
+ #  whatsoever resulting from loss of use, data or profits, whether in an
+ #  action of contract, negligence or other tortuous action, arising out
+ #  of or in connection with the use or performance of this software.
+ #
+ #  A large portion of the dictionary entries
+ #  originate from ICOT Free Software.  The following conditions for ICOT
+ #  Free Software applies to the current dictionary as well.
+ #
+ #  Each User may also freely distribute the Program, whether in its
+ #  original form or modified, to any third party or parties, PROVIDED
+ #  that the provisions of Section 3 ("NO WARRANTY") will ALWAYS appear
+ #  on, or be attached to, the Program, which is distributed substantially
+ #  in the same form as set out herein and that such intended
+ #  distribution, if actually made, will neither violate or otherwise
+ #  contravene any of the laws and regulations of the countries having
+ #  jurisdiction over the User or the intended distribution itself.
+ #
+ #  NO WARRANTY
+ #
+ #  The program was produced on an experimental basis in the course of the
+ #  research and development conducted during the project and is provided
+ #  to users as so produced on an experimental basis.  Accordingly, the
+ #  program is provided without any warranty whatsoever, whether express,
+ #  implied, statutory or otherwise.  The term "warranty" used herein
+ #  includes, but is not limited to, any warranty of the quality,
+ #  performance, merchantability and fitness for a particular purpose of
+ #  the program and the nonexistence of any infringement or violation of
+ #  any right of any third party.
+ #
+ #  Each user of the program will agree and understand, and be deemed to
+ #  have agreed and understood, that there is no warranty whatsoever for
+ #  the program and, accordingly, the entire risk arising from or
+ #  otherwise connected with the program is assumed by the user.
+ #
+ #  Therefore, neither ICOT, the copyright holder, or any other
+ #  organization that participated in or was otherwise related to the
+ #  development of the program and their respective officials, directors,
+ #  officers and other employees shall be held liable for any and all
+ #  damages, including, without limitation, general, special, incidental
+ #  and consequential damages, arising out of or otherwise in connection
+ #  with the use or inability to use the program or any product, material
+ #  or result produced or otherwise obtained by using the program,
+ #  regardless of whether they have been advised of, or otherwise had
+ #  knowledge of, the possibility of such damages at any time during the
+ #  project or thereafter.  Each user will be deemed to have agreed to the
+ #  foregoing by his or her commencement of use of the program.  The term
+ #  "use" as used herein includes, but is not limited to, the use,
+ #  modification, copying and distribution of the program and the
+ #  production of secondary products from the program.
+ #
+ #  In the case where the program, whether in its original form or
+ #  modified, was distributed or delivered to or received by a user from
+ #  any person, organization or entity other than ICOT, unless it makes or
+ #  grants independently of ICOT any specific warranty to the user in
+ #  writing, such person, organization or entity, will also be exempted
+ #  from and not be held liable to the user for any such damages as noted
+ #  above as far as the program is concerned.
+ #
+ #  ---------------COPYING.ipadic-----END----------------------------------
+
+3. Lao Word Break Dictionary Data (laodict.txt)
+
+ #  Copyright (c) 2013 International Business Machines Corporation
+ #  and others. All Rights Reserved.
+ #
+ # Project: https://github.com/veer66/lao-dictionary
+ # Dictionary: https://github.com/veer66/lao-dictionary/blob/master/Lao-Dictionary.txt
+ # License: https://github.com/veer66/lao-dictionary/blob/master/Lao-Dictionary-LICENSE.txt
+ #              (copied below)
+ #
+ #  This file is derived from the above dictionary, with slight
+ #  modifications.
+ #  ----------------------------------------------------------------------
+ #  Copyright (C) 2013 Brian Eugene Wilson, Robert Martin Campbell.
+ #  All rights reserved.
+ #
+ #  Redistribution and use in source and binary forms, with or without
+ #  modification,
+ #  are permitted provided that the following conditions are met:
+ #
+ #
+ # Redistributions of source code must retain the above copyright notice, this
+ #  list of conditions and the following disclaimer. Redistributions in
+ #  binary form must reproduce the above copyright notice, this list of
+ #  conditions and the following disclaimer in the documentation and/or
+ #  other materials provided with the distribution.
+ #
+ #
+ # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ # INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ # (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ # SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ # HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ # STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ # OF THE POSSIBILITY OF SUCH DAMAGE.
+ #  --------------------------------------------------------------------------
+
+4. Burmese Word Break Dictionary Data (burmesedict.txt)
+
+ #  Copyright (c) 2014 International Business Machines Corporation
+ #  and others. All Rights Reserved.
+ #
+ #  This list is part of a project hosted at:
+ #    github.com/kanyawtech/myanmar-karen-word-lists
+ #
+ #  --------------------------------------------------------------------------
+ #  Copyright (c) 2013, LeRoy Benjamin Sharon
+ #  All rights reserved.
+ #
+ #  Redistribution and use in source and binary forms, with or without
+ #  modification, are permitted provided that the following conditions
+ #  are met: Redistributions of source code must retain the above
+ #  copyright notice, this list of conditions and the following
+ #  disclaimer.  Redistributions in binary form must reproduce the
+ #  above copyright notice, this list of conditions and the following
+ #  disclaimer in the documentation and/or other materials provided
+ #  with the distribution.
+ #
+ #    Neither the name Myanmar Karen Word Lists, nor the names of its
+ #    contributors may be used to endorse or promote products derived
+ #    from this software without specific prior written permission.
+ #
+ #  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
+ #  CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
+ #  INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ #  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ #  DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS
+ #  BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ #  EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
+ #  TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ #  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ #  ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
+ #  TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF
+ #  THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ #  SUCH DAMAGE.
+ #  --------------------------------------------------------------------------
+
+5. Time Zone Database
+
+  ICU uses the public domain data and code derived from Time Zone
+Database for its time zone support. The ownership of the TZ database
+is explained in BCP 175: Procedure for Maintaining the Time Zone
+Database section 7.
+
+ # 7.  Database Ownership
+ #
+ #    The TZ database itself is not an IETF Contribution or an IETF
+ #    document.  Rather it is a pre-existing and regularly updated work
+ #    that is in the public domain, and is intended to remain in the
+ #    public domain.  Therefore, BCPs 78 [RFC5378] and 79 [RFC3979] do
+ #    not apply to the TZ Database or contributions that individuals make
+ #    to it.  Should any claims be made and substantiated against the TZ
+ #    Database, the organization that is providing the IANA
+ #    Considerations defined in this RFC, under the memorandum of
+ #    understanding with the IETF, currently ICANN, may act in accordance
+ #    with all competent court orders.  No ownership claims will be made
+ #    by ICANN or the IETF Trust on the database or the code.  Any person
+ #    making a contribution to the database or code waives all rights to
+ #    future claims in that contribution or in the TZ Database.
+
+6. Google double-conversion
+
+Copyright 2006-2011, the V8 project authors. All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+    * Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+    * Redistributions in binary form must reproduce the above
+      copyright notice, this list of conditions and the following
+      disclaimer in the documentation and/or other materials provided
+      with the distribution.
+    * Neither the name of Google Inc. nor the names of its
+      contributors may be used to endorse or promote products derived
+      from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
diff --git a/thirdparty/icu4c/common/appendable.cpp b/thirdparty/icu4c/common/appendable.cpp
new file mode 100644
index 0000000000..fca3c1e413
--- /dev/null
+++ b/thirdparty/icu4c/common/appendable.cpp
@@ -0,0 +1,74 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  appendable.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010dec07
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/appendable.h"
+#include "unicode/utf16.h"
+
+U_NAMESPACE_BEGIN
+
+Appendable::~Appendable() {}
+
+UBool
+Appendable::appendCodePoint(UChar32 c) {
+    if(c<=0xffff) {
+        return appendCodeUnit((UChar)c);
+    } else {
+        return appendCodeUnit(U16_LEAD(c)) && appendCodeUnit(U16_TRAIL(c));
+    }
+}
+
+UBool
+Appendable::appendString(const UChar *s, int32_t length) {
+    if(length<0) {
+        UChar c;
+        while((c=*s++)!=0) {
+            if(!appendCodeUnit(c)) {
+                return FALSE;
+            }
+        }
+    } else if(length>0) {
+        const UChar *limit=s+length;
+        do {
+            if(!appendCodeUnit(*s++)) {
+                return FALSE;
+            }
+        } while(s<limit);
+    }
+    return TRUE;
+}
+
+UBool
+Appendable::reserveAppendCapacity(int32_t /*appendCapacity*/) {
+    return TRUE;
+}
+
+UChar *
+Appendable::getAppendBuffer(int32_t minCapacity,
+                            int32_t /*desiredCapacityHint*/,
+                            UChar *scratch, int32_t scratchCapacity,
+                            int32_t *resultCapacity) {
+    if(minCapacity<1 || scratchCapacity<minCapacity) {
+        *resultCapacity=0;
+        return NULL;
+    }
+    *resultCapacity=scratchCapacity;
+    return scratch;
+}
+
+// UnicodeStringAppendable is implemented in unistr.cpp.
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bmpset.cpp b/thirdparty/icu4c/common/bmpset.cpp
new file mode 100644
index 0000000000..bc79f5e5a6
--- /dev/null
+++ b/thirdparty/icu4c/common/bmpset.cpp
@@ -0,0 +1,741 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2007-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  bmpset.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2007jan29
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "bmpset.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+BMPSet::BMPSet(const int32_t *parentList, int32_t parentListLength) :
+        list(parentList), listLength(parentListLength) {
+    uprv_memset(latin1Contains, 0, sizeof(latin1Contains));
+    uprv_memset(table7FF, 0, sizeof(table7FF));
+    uprv_memset(bmpBlockBits, 0, sizeof(bmpBlockBits));
+
+    /*
+     * Set the list indexes for binary searches for
+     * U+0800, U+1000, U+2000, .., U+F000, U+10000.
+     * U+0800 is the first 3-byte-UTF-8 code point. Lower code points are
+     * looked up in the bit tables.
+     * The last pair of indexes is for finding supplementary code points.
+     */
+    list4kStarts[0]=findCodePoint(0x800, 0, listLength-1);
+    int32_t i;
+    for(i=1; i<=0x10; ++i) {
+        list4kStarts[i]=findCodePoint(i<<12, list4kStarts[i-1], listLength-1);
+    }
+    list4kStarts[0x11]=listLength-1;
+    containsFFFD=containsSlow(0xfffd, list4kStarts[0xf], list4kStarts[0x10]);
+
+    initBits();
+    overrideIllegal();
+}
+
+BMPSet::BMPSet(const BMPSet &otherBMPSet, const int32_t *newParentList, int32_t newParentListLength) :
+        containsFFFD(otherBMPSet.containsFFFD),
+        list(newParentList), listLength(newParentListLength) {
+    uprv_memcpy(latin1Contains, otherBMPSet.latin1Contains, sizeof(latin1Contains));
+    uprv_memcpy(table7FF, otherBMPSet.table7FF, sizeof(table7FF));
+    uprv_memcpy(bmpBlockBits, otherBMPSet.bmpBlockBits, sizeof(bmpBlockBits));
+    uprv_memcpy(list4kStarts, otherBMPSet.list4kStarts, sizeof(list4kStarts));
+}
+
+BMPSet::~BMPSet() {
+}
+
+/*
+ * Set bits in a bit rectangle in "vertical" bit organization.
+ * start<limit<=0x800
+ */
+static void set32x64Bits(uint32_t table[64], int32_t start, int32_t limit) {
+    U_ASSERT(start<limit);
+    U_ASSERT(limit<=0x800);
+
+    int32_t lead=start>>6;  // Named for UTF-8 2-byte lead byte with upper 5 bits.
+    int32_t trail=start&0x3f;  // Named for UTF-8 2-byte trail byte with lower 6 bits.
+
+    // Set one bit indicating an all-one block.
+    uint32_t bits=(uint32_t)1<<lead;
+    if((start+1)==limit) {  // Single-character shortcut.
+        table[trail]|=bits;
+        return;
+    }
+
+    int32_t limitLead=limit>>6;
+    int32_t limitTrail=limit&0x3f;
+
+    if(lead==limitLead) {
+        // Partial vertical bit column.
+        while(trail<limitTrail) {
+            table[trail++]|=bits;
+        }
+    } else {
+        // Partial vertical bit column,
+        // followed by a bit rectangle,
+        // followed by another partial vertical bit column.
+        if(trail>0) {
+            do {
+                table[trail++]|=bits;
+            } while(trail<64);
+            ++lead;
+        }
+        if(lead<limitLead) {
+            bits=~(((unsigned)1<<lead)-1);
+            if(limitLead<0x20) {
+                bits&=((unsigned)1<<limitLead)-1;
+            }
+            for(trail=0; trail<64; ++trail) {
+                table[trail]|=bits;
+            }
+        }
+        // limit<=0x800. If limit==0x800 then limitLead=32 and limitTrail=0.
+        // In that case, bits=1<<limitLead is undefined but the bits value
+        // is not used because trail<limitTrail is already false.
+        bits=(uint32_t)1<<((limitLead == 0x20) ? (limitLead - 1) : limitLead);
+        for(trail=0; trail<limitTrail; ++trail) {
+            table[trail]|=bits;
+        }
+    }
+}
+
+void BMPSet::initBits() {
+    UChar32 start, limit;
+    int32_t listIndex=0;
+
+    // Set latin1Contains[].
+    do {
+        start=list[listIndex++];
+        if(listIndex<listLength) {
+            limit=list[listIndex++];
+        } else {
+            limit=0x110000;
+        }
+        if(start>=0x100) {
+            break;
+        }
+        do {
+            latin1Contains[start++]=1;
+        } while(start<limit && start<0x100);
+    } while(limit<=0x100);
+
+    // Find the first range overlapping with (or after) 80..FF again,
+    // to include them in table7FF as well.
+    for(listIndex=0;;) {
+        start=list[listIndex++];
+        if(listIndex<listLength) {
+            limit=list[listIndex++];
+        } else {
+            limit=0x110000;
+        }
+        if(limit>0x80) {
+            if(start<0x80) {
+                start=0x80;
+            }
+            break;
+        }
+    }
+
+    // Set table7FF[].
+    while(start<0x800) {
+        set32x64Bits(table7FF, start, limit<=0x800 ? limit : 0x800);
+        if(limit>0x800) {
+            start=0x800;
+            break;
+        }
+
+        start=list[listIndex++];
+        if(listIndex<listLength) {
+            limit=list[listIndex++];
+        } else {
+            limit=0x110000;
+        }
+    }
+
+    // Set bmpBlockBits[].
+    int32_t minStart=0x800;
+    while(start<0x10000) {
+        if(limit>0x10000) {
+            limit=0x10000;
+        }
+
+        if(start<minStart) {
+            start=minStart;
+        }
+        if(start<limit) {  // Else: Another range entirely in a known mixed-value block.
+            if(start&0x3f) {
+                // Mixed-value block of 64 code points.
+                start>>=6;
+                bmpBlockBits[start&0x3f]|=0x10001<<(start>>6);
+                start=(start+1)<<6;  // Round up to the next block boundary.
+                minStart=start;      // Ignore further ranges in this block.
+            }
+            if(start<limit) {
+                if(start<(limit&~0x3f)) {
+                    // Multiple all-ones blocks of 64 code points each.
+                    set32x64Bits(bmpBlockBits, start>>6, limit>>6);
+                }
+
+                if(limit&0x3f) {
+                    // Mixed-value block of 64 code points.
+                    limit>>=6;
+                    bmpBlockBits[limit&0x3f]|=0x10001<<(limit>>6);
+                    limit=(limit+1)<<6;  // Round up to the next block boundary.
+                    minStart=limit;      // Ignore further ranges in this block.
+                }
+            }
+        }
+
+        if(limit==0x10000) {
+            break;
+        }
+
+        start=list[listIndex++];
+        if(listIndex<listLength) {
+            limit=list[listIndex++];
+        } else {
+            limit=0x110000;
+        }
+    }
+}
+
+/*
+ * Override some bits and bytes to the result of contains(FFFD)
+ * for faster validity checking at runtime.
+ * No need to set 0 values where they were reset to 0 in the constructor
+ * and not modified by initBits().
+ * (table7FF[] 0..7F, bmpBlockBits[] 0..7FF)
+ * Need to set 0 values for surrogates D800..DFFF.
+ */
+void BMPSet::overrideIllegal() {
+    uint32_t bits, mask;
+    int32_t i;
+
+    if(containsFFFD) {
+        bits=3;                 // Lead bytes 0xC0 and 0xC1.
+        for(i=0; i<64; ++i) {
+            table7FF[i]|=bits;
+        }
+
+        bits=1;                 // Lead byte 0xE0.
+        for(i=0; i<32; ++i) {   // First half of 4k block.
+            bmpBlockBits[i]|=bits;
+        }
+
+        mask= static_cast<uint32_t>(~(0x10001<<0xd));   // Lead byte 0xED.
+        bits=1<<0xd;
+        for(i=32; i<64; ++i) {  // Second half of 4k block.
+            bmpBlockBits[i]=(bmpBlockBits[i]&mask)|bits;
+        }
+    } else {
+        mask= static_cast<uint32_t>(~(0x10001<<0xd));   // Lead byte 0xED.
+        for(i=32; i<64; ++i) {  // Second half of 4k block.
+            bmpBlockBits[i]&=mask;
+        }
+    }
+}
+
+int32_t BMPSet::findCodePoint(UChar32 c, int32_t lo, int32_t hi) const {
+    /* Examples:
+                                       findCodePoint(c)
+       set              list[]         c=0 1 3 4 7 8
+       ===              ==============   ===========
+       []               [110000]         0 0 0 0 0 0
+       [\u0000-\u0003]  [0, 4, 110000]   1 1 1 2 2 2
+       [\u0004-\u0007]  [4, 8, 110000]   0 0 0 1 1 2
+       [:Any:]          [0, 110000]      1 1 1 1 1 1
+     */
+
+    // Return the smallest i such that c < list[i].  Assume
+    // list[len - 1] == HIGH and that c is legal (0..HIGH-1).
+    if (c < list[lo])
+        return lo;
+    // High runner test.  c is often after the last range, so an
+    // initial check for this condition pays off.
+    if (lo >= hi || c >= list[hi-1])
+        return hi;
+    // invariant: c >= list[lo]
+    // invariant: c < list[hi]
+    for (;;) {
+        int32_t i = (lo + hi) >> 1;
+        if (i == lo) {
+            break; // Found!
+        } else if (c < list[i]) {
+            hi = i;
+        } else {
+            lo = i;
+        }
+    }
+    return hi;
+}
+
+UBool
+BMPSet::contains(UChar32 c) const {
+    if((uint32_t)c<=0xff) {
+        return (UBool)latin1Contains[c];
+    } else if((uint32_t)c<=0x7ff) {
+        return (UBool)((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))!=0);
+    } else if((uint32_t)c<0xd800 || (c>=0xe000 && c<=0xffff)) {
+        int lead=c>>12;
+        uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+        if(twoBits<=1) {
+            // All 64 code points with the same bits 15..6
+            // are either in the set or not.
+            return (UBool)twoBits;
+        } else {
+            // Look up the code point in its 4k block of code points.
+            return containsSlow(c, list4kStarts[lead], list4kStarts[lead+1]);
+        }
+    } else if((uint32_t)c<=0x10ffff) {
+        // surrogate or supplementary code point
+        return containsSlow(c, list4kStarts[0xd], list4kStarts[0x11]);
+    } else {
+        // Out-of-range code points get FALSE, consistent with long-standing
+        // behavior of UnicodeSet::contains(c).
+        return FALSE;
+    }
+}
+
+/*
+ * Check for sufficient length for trail unit for each surrogate pair.
+ * Handle single surrogates as surrogate code points as usual in ICU.
+ */
+const UChar *
+BMPSet::span(const UChar *s, const UChar *limit, USetSpanCondition spanCondition) const {
+    UChar c, c2;
+
+    if(spanCondition) {
+        // span
+        do {
+            c=*s;
+            if(c<=0xff) {
+                if(!latin1Contains[c]) {
+                    break;
+                }
+            } else if(c<=0x7ff) {
+                if((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))==0) {
+                    break;
+                }
+            } else if(c<0xd800 || c>=0xe000) {
+                int lead=c>>12;
+                uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+                if(twoBits<=1) {
+                    // All 64 code points with the same bits 15..6
+                    // are either in the set or not.
+                    if(twoBits==0) {
+                        break;
+                    }
+                } else {
+                    // Look up the code point in its 4k block of code points.
+                    if(!containsSlow(c, list4kStarts[lead], list4kStarts[lead+1])) {
+                        break;
+                    }
+                }
+            } else if(c>=0xdc00 || (s+1)==limit || (c2=s[1])<0xdc00 || c2>=0xe000) {
+                // surrogate code point
+                if(!containsSlow(c, list4kStarts[0xd], list4kStarts[0xe])) {
+                    break;
+                }
+            } else {
+                // surrogate pair
+                if(!containsSlow(U16_GET_SUPPLEMENTARY(c, c2), list4kStarts[0x10], list4kStarts[0x11])) {
+                    break;
+                }
+                ++s;
+            }
+        } while(++s<limit);
+    } else {
+        // span not
+        do {
+            c=*s;
+            if(c<=0xff) {
+                if(latin1Contains[c]) {
+                    break;
+                }
+            } else if(c<=0x7ff) {
+                if((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))!=0) {
+                    break;
+                }
+            } else if(c<0xd800 || c>=0xe000) {
+                int lead=c>>12;
+                uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+                if(twoBits<=1) {
+                    // All 64 code points with the same bits 15..6
+                    // are either in the set or not.
+                    if(twoBits!=0) {
+                        break;
+                    }
+                } else {
+                    // Look up the code point in its 4k block of code points.
+                    if(containsSlow(c, list4kStarts[lead], list4kStarts[lead+1])) {
+                        break;
+                    }
+                }
+            } else if(c>=0xdc00 || (s+1)==limit || (c2=s[1])<0xdc00 || c2>=0xe000) {
+                // surrogate code point
+                if(containsSlow(c, list4kStarts[0xd], list4kStarts[0xe])) {
+                    break;
+                }
+            } else {
+                // surrogate pair
+                if(containsSlow(U16_GET_SUPPLEMENTARY(c, c2), list4kStarts[0x10], list4kStarts[0x11])) {
+                    break;
+                }
+                ++s;
+            }
+        } while(++s<limit);
+    }
+    return s;
+}
+
+/* Symmetrical with span(). */
+const UChar *
+BMPSet::spanBack(const UChar *s, const UChar *limit, USetSpanCondition spanCondition) const {
+    UChar c, c2;
+
+    if(spanCondition) {
+        // span
+        for(;;) {
+            c=*(--limit);
+            if(c<=0xff) {
+                if(!latin1Contains[c]) {
+                    break;
+                }
+            } else if(c<=0x7ff) {
+                if((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))==0) {
+                    break;
+                }
+            } else if(c<0xd800 || c>=0xe000) {
+                int lead=c>>12;
+                uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+                if(twoBits<=1) {
+                    // All 64 code points with the same bits 15..6
+                    // are either in the set or not.
+                    if(twoBits==0) {
+                        break;
+                    }
+                } else {
+                    // Look up the code point in its 4k block of code points.
+                    if(!containsSlow(c, list4kStarts[lead], list4kStarts[lead+1])) {
+                        break;
+                    }
+                }
+            } else if(c<0xdc00 || s==limit || (c2=*(limit-1))<0xd800 || c2>=0xdc00) {
+                // surrogate code point
+                if(!containsSlow(c, list4kStarts[0xd], list4kStarts[0xe])) {
+                    break;
+                }
+            } else {
+                // surrogate pair
+                if(!containsSlow(U16_GET_SUPPLEMENTARY(c2, c), list4kStarts[0x10], list4kStarts[0x11])) {
+                    break;
+                }
+                --limit;
+            }
+            if(s==limit) {
+                return s;
+            }
+        }
+    } else {
+        // span not
+        for(;;) {
+            c=*(--limit);
+            if(c<=0xff) {
+                if(latin1Contains[c]) {
+                    break;
+                }
+            } else if(c<=0x7ff) {
+                if((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))!=0) {
+                    break;
+                }
+            } else if(c<0xd800 || c>=0xe000) {
+                int lead=c>>12;
+                uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+                if(twoBits<=1) {
+                    // All 64 code points with the same bits 15..6
+                    // are either in the set or not.
+                    if(twoBits!=0) {
+                        break;
+                    }
+                } else {
+                    // Look up the code point in its 4k block of code points.
+                    if(containsSlow(c, list4kStarts[lead], list4kStarts[lead+1])) {
+                        break;
+                    }
+                }
+            } else if(c<0xdc00 || s==limit || (c2=*(limit-1))<0xd800 || c2>=0xdc00) {
+                // surrogate code point
+                if(containsSlow(c, list4kStarts[0xd], list4kStarts[0xe])) {
+                    break;
+                }
+            } else {
+                // surrogate pair
+                if(containsSlow(U16_GET_SUPPLEMENTARY(c2, c), list4kStarts[0x10], list4kStarts[0x11])) {
+                    break;
+                }
+                --limit;
+            }
+            if(s==limit) {
+                return s;
+            }
+        }
+    }
+    return limit+1;
+}
+
+/*
+ * Precheck for sufficient trail bytes at end of string only once per span.
+ * Check validity.
+ */
+const uint8_t *
+BMPSet::spanUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const {
+    const uint8_t *limit=s+length;
+    uint8_t b=*s;
+    if(U8_IS_SINGLE(b)) {
+        // Initial all-ASCII span.
+        if(spanCondition) {
+            do {
+                if(!latin1Contains[b] || ++s==limit) {
+                    return s;
+                }
+                b=*s;
+            } while(U8_IS_SINGLE(b));
+        } else {
+            do {
+                if(latin1Contains[b] || ++s==limit) {
+                    return s;
+                }
+                b=*s;
+            } while(U8_IS_SINGLE(b));
+        }
+        length=(int32_t)(limit-s);
+    }
+
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    const uint8_t *limit0=limit;
+
+    /*
+     * Make sure that the last 1/2/3/4-byte sequence before limit is complete
+     * or runs into a lead byte.
+     * In the span loop compare s with limit only once
+     * per multi-byte character.
+     *
+     * Give a trailing illegal sequence the same value as the result of contains(FFFD),
+     * including it if that is part of the span, otherwise set limit0 to before
+     * the truncated sequence.
+     */
+    b=*(limit-1);
+    if((int8_t)b<0) {
+        // b>=0x80: lead or trail byte
+        if(b<0xc0) {
+            // single trail byte, check for preceding 3- or 4-byte lead byte
+            if(length>=2 && (b=*(limit-2))>=0xe0) {
+                limit-=2;
+                if(containsFFFD!=spanCondition) {
+                    limit0=limit;
+                }
+            } else if(b<0xc0 && b>=0x80 && length>=3 && (b=*(limit-3))>=0xf0) {
+                // 4-byte lead byte with only two trail bytes
+                limit-=3;
+                if(containsFFFD!=spanCondition) {
+                    limit0=limit;
+                }
+            }
+        } else {
+            // lead byte with no trail bytes
+            --limit;
+            if(containsFFFD!=spanCondition) {
+                limit0=limit;
+            }
+        }
+    }
+
+    uint8_t t1, t2, t3;
+
+    while(s<limit) {
+        b=*s;
+        if(U8_IS_SINGLE(b)) {
+            // ASCII
+            if(spanCondition) {
+                do {
+                    if(!latin1Contains[b]) {
+                        return s;
+                    } else if(++s==limit) {
+                        return limit0;
+                    }
+                    b=*s;
+                } while(U8_IS_SINGLE(b));
+            } else {
+                do {
+                    if(latin1Contains[b]) {
+                        return s;
+                    } else if(++s==limit) {
+                        return limit0;
+                    }
+                    b=*s;
+                } while(U8_IS_SINGLE(b));
+            }
+        }
+        ++s;  // Advance past the lead byte.
+        if(b>=0xe0) {
+            if(b<0xf0) {
+                if( /* handle U+0000..U+FFFF inline */
+                    (t1=(uint8_t)(s[0]-0x80)) <= 0x3f &&
+                    (t2=(uint8_t)(s[1]-0x80)) <= 0x3f
+                ) {
+                    b&=0xf;
+                    uint32_t twoBits=(bmpBlockBits[t1]>>b)&0x10001;
+                    if(twoBits<=1) {
+                        // All 64 code points with this lead byte and middle trail byte
+                        // are either in the set or not.
+                        if(twoBits!=(uint32_t)spanCondition) {
+                            return s-1;
+                        }
+                    } else {
+                        // Look up the code point in its 4k block of code points.
+                        UChar32 c=(b<<12)|(t1<<6)|t2;
+                        if(containsSlow(c, list4kStarts[b], list4kStarts[b+1]) != spanCondition) {
+                            return s-1;
+                        }
+                    }
+                    s+=2;
+                    continue;
+                }
+            } else if( /* handle U+10000..U+10FFFF inline */
+                (t1=(uint8_t)(s[0]-0x80)) <= 0x3f &&
+                (t2=(uint8_t)(s[1]-0x80)) <= 0x3f &&
+                (t3=(uint8_t)(s[2]-0x80)) <= 0x3f
+            ) {
+                // Give an illegal sequence the same value as the result of contains(FFFD).
+                UChar32 c=((UChar32)(b-0xf0)<<18)|((UChar32)t1<<12)|(t2<<6)|t3;
+                if( (   (0x10000<=c && c<=0x10ffff) ?
+                            containsSlow(c, list4kStarts[0x10], list4kStarts[0x11]) :
+                            containsFFFD
+                    ) != spanCondition
+                ) {
+                    return s-1;
+                }
+                s+=3;
+                continue;
+            }
+        } else {
+            if( /* handle U+0000..U+07FF inline */
+                b>=0xc0 &&
+                (t1=(uint8_t)(*s-0x80)) <= 0x3f
+            ) {
+                if((USetSpanCondition)((table7FF[t1]&((uint32_t)1<<(b&0x1f)))!=0) != spanCondition) {
+                    return s-1;
+                }
+                ++s;
+                continue;
+            }
+        }
+
+        // Give an illegal sequence the same value as the result of contains(FFFD).
+        // Handle each byte of an illegal sequence separately to simplify the code;
+        // no need to optimize error handling.
+        if(containsFFFD!=spanCondition) {
+            return s-1;
+        }
+    }
+
+    return limit0;
+}
+
+/*
+ * While going backwards through UTF-8 optimize only for ASCII.
+ * Unlike UTF-16, UTF-8 is not forward-backward symmetrical, that is, it is not
+ * possible to tell from the last byte in a multi-byte sequence how many
+ * preceding bytes there should be. Therefore, going backwards through UTF-8
+ * is much harder than going forward.
+ */
+int32_t
+BMPSet::spanBackUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    uint8_t b;
+
+    do {
+        b=s[--length];
+        if(U8_IS_SINGLE(b)) {
+            // ASCII sub-span
+            if(spanCondition) {
+                do {
+                    if(!latin1Contains[b]) {
+                        return length+1;
+                    } else if(length==0) {
+                        return 0;
+                    }
+                    b=s[--length];
+                } while(U8_IS_SINGLE(b));
+            } else {
+                do {
+                    if(latin1Contains[b]) {
+                        return length+1;
+                    } else if(length==0) {
+                        return 0;
+                    }
+                    b=s[--length];
+                } while(U8_IS_SINGLE(b));
+            }
+        }
+
+        int32_t prev=length;
+        UChar32 c;
+        // trail byte: collect a multi-byte character
+        // (or  lead byte in last-trail position)
+        c=utf8_prevCharSafeBody(s, 0, &length, b, -3);
+        // c is a valid code point, not ASCII, not a surrogate
+        if(c<=0x7ff) {
+            if((USetSpanCondition)((table7FF[c&0x3f]&((uint32_t)1<<(c>>6)))!=0) != spanCondition) {
+                return prev+1;
+            }
+        } else if(c<=0xffff) {
+            int lead=c>>12;
+            uint32_t twoBits=(bmpBlockBits[(c>>6)&0x3f]>>lead)&0x10001;
+            if(twoBits<=1) {
+                // All 64 code points with the same bits 15..6
+                // are either in the set or not.
+                if(twoBits!=(uint32_t)spanCondition) {
+                    return prev+1;
+                }
+            } else {
+                // Look up the code point in its 4k block of code points.
+                if(containsSlow(c, list4kStarts[lead], list4kStarts[lead+1]) != spanCondition) {
+                    return prev+1;
+                }
+            }
+        } else {
+            if(containsSlow(c, list4kStarts[0x10], list4kStarts[0x11]) != spanCondition) {
+                return prev+1;
+            }
+        }
+    } while(length>0);
+    return 0;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bmpset.h b/thirdparty/icu4c/common/bmpset.h
new file mode 100644
index 0000000000..e1982ac669
--- /dev/null
+++ b/thirdparty/icu4c/common/bmpset.h
@@ -0,0 +1,164 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2007, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  bmpset.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2007jan29
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __BMPSET_H__
+#define __BMPSET_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Helper class for frozen UnicodeSets, implements contains() and span()
+ * optimized for BMP code points. Structured to be UTF-8-friendly.
+ *
+ * Latin-1: Look up bytes.
+ * 2-byte characters: Bits organized vertically.
+ * 3-byte characters: Use zero/one/mixed data per 64-block in U+0000..U+FFFF,
+ *                    with mixed for illegal ranges.
+ * Supplementary characters: Binary search over
+ * the supplementary part of the parent set's inversion list.
+ */
+class BMPSet : public UMemory {
+public:
+    BMPSet(const int32_t *parentList, int32_t parentListLength);
+    BMPSet(const BMPSet &otherBMPSet, const int32_t *newParentList, int32_t newParentListLength);
+    virtual ~BMPSet();
+
+    virtual UBool contains(UChar32 c) const;
+
+    /*
+     * Span the initial substring for which each character c has spanCondition==contains(c).
+     * It must be s<limit and spanCondition==0 or 1.
+     * @return The string pointer which limits the span.
+     */
+    const UChar *span(const UChar *s, const UChar *limit, USetSpanCondition spanCondition) const;
+
+    /*
+     * Span the trailing substring for which each character c has spanCondition==contains(c).
+     * It must be s<limit and spanCondition==0 or 1.
+     * @return The string pointer which starts the span.
+     */
+    const UChar *spanBack(const UChar *s, const UChar *limit, USetSpanCondition spanCondition) const;
+
+    /*
+     * Span the initial substring for which each character c has spanCondition==contains(c).
+     * It must be length>0 and spanCondition==0 or 1.
+     * @return The string pointer which limits the span.
+     */
+    const uint8_t *spanUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    /*
+     * Span the trailing substring for which each character c has spanCondition==contains(c).
+     * It must be length>0 and spanCondition==0 or 1.
+     * @return The start of the span.
+     */
+    int32_t spanBackUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+private:
+    void initBits();
+    void overrideIllegal();
+
+    /**
+     * Same as UnicodeSet::findCodePoint(UChar32 c) const except that the
+     * binary search is restricted for finding code points in a certain range.
+     *
+     * For restricting the search for finding in the range start..end,
+     * pass in
+     *   lo=findCodePoint(start) and
+     *   hi=findCodePoint(end)
+     * with 0<=lo<=hi<len.
+     * findCodePoint(c) defaults to lo=0 and hi=len-1.
+     *
+     * @param c a character in a subrange of MIN_VALUE..MAX_VALUE
+     * @param lo The lowest index to be returned.
+     * @param hi The highest index to be returned.
+     * @return the smallest integer i in the range lo..hi,
+     *         inclusive, such that c < list[i]
+     */
+    int32_t findCodePoint(UChar32 c, int32_t lo, int32_t hi) const;
+
+    inline UBool containsSlow(UChar32 c, int32_t lo, int32_t hi) const;
+
+    /*
+     * One byte 0 or 1 per Latin-1 character.
+     */
+    UBool latin1Contains[0x100];
+
+    /* true if contains(U+FFFD). */
+    UBool containsFFFD;
+
+    /*
+     * One bit per code point from U+0000..U+07FF.
+     * The bits are organized vertically; consecutive code points
+     * correspond to the same bit positions in consecutive table words.
+     * With code point parts
+     *   lead=c{10..6}
+     *   trail=c{5..0}
+     * it is set.contains(c)==(table7FF[trail] bit lead)
+     *
+     * Bits for 0..7F (non-shortest forms) are set to the result of contains(FFFD)
+     * for faster validity checking at runtime.
+     */
+    uint32_t table7FF[64];
+
+    /*
+     * One bit per 64 BMP code points.
+     * The bits are organized vertically; consecutive 64-code point blocks
+     * correspond to the same bit position in consecutive table words.
+     * With code point parts
+     *   lead=c{15..12}
+     *   t1=c{11..6}
+     * test bits (lead+16) and lead in bmpBlockBits[t1].
+     * If the upper bit is 0, then the lower bit indicates if contains(c)
+     * for all code points in the 64-block.
+     * If the upper bit is 1, then the block is mixed and set.contains(c)
+     * must be called.
+     *
+     * Bits for 0..7FF (non-shortest forms) and D800..DFFF are set to
+     * the result of contains(FFFD) for faster validity checking at runtime.
+     */
+    uint32_t bmpBlockBits[64];
+
+    /*
+     * Inversion list indexes for restricted binary searches in
+     * findCodePoint(), from
+     * findCodePoint(U+0800, U+1000, U+2000, .., U+F000, U+10000).
+     * U+0800 is the first 3-byte-UTF-8 code point. Code points below U+0800 are
+     * always looked up in the bit tables.
+     * The last pair of indexes is for finding supplementary code points.
+     */
+    int32_t list4kStarts[18];
+
+    /*
+     * The inversion list of the parent set, for the slower contains() implementation
+     * for mixed BMP blocks and for supplementary code points.
+     * The list is terminated with list[listLength-1]=0x110000.
+     */
+    const int32_t *list;
+    int32_t listLength;
+};
+
+inline UBool BMPSet::containsSlow(UChar32 c, int32_t lo, int32_t hi) const {
+    return (UBool)(findCodePoint(c, lo, hi) & 1);
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/brkeng.cpp b/thirdparty/icu4c/common/brkeng.cpp
new file mode 100644
index 0000000000..78492db662
--- /dev/null
+++ b/thirdparty/icu4c/common/brkeng.cpp
@@ -0,0 +1,284 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ ************************************************************************************
+ * Copyright (C) 2006-2016, International Business Machines Corporation
+ * and others. All Rights Reserved.
+ ************************************************************************************
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/uchar.h"
+#include "unicode/uniset.h"
+#include "unicode/chariter.h"
+#include "unicode/ures.h"
+#include "unicode/udata.h"
+#include "unicode/putil.h"
+#include "unicode/ustring.h"
+#include "unicode/uscript.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/bytestrie.h"
+
+#include "brkeng.h"
+#include "cmemory.h"
+#include "dictbe.h"
+#include "charstr.h"
+#include "dictionarydata.h"
+#include "mutex.h"
+#include "uvector.h"
+#include "umutex.h"
+#include "uresimp.h"
+#include "ubrkimpl.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ ******************************************************************
+ */
+
+LanguageBreakEngine::LanguageBreakEngine() {
+}
+
+LanguageBreakEngine::~LanguageBreakEngine() {
+}
+
+/*
+ ******************************************************************
+ */
+
+LanguageBreakFactory::LanguageBreakFactory() {
+}
+
+LanguageBreakFactory::~LanguageBreakFactory() {
+}
+
+/*
+ ******************************************************************
+ */
+
+UnhandledEngine::UnhandledEngine(UErrorCode &status) : fHandled(nullptr) {
+    (void)status;
+}
+
+UnhandledEngine::~UnhandledEngine() {
+    delete fHandled;
+    fHandled = nullptr;
+}
+
+UBool
+UnhandledEngine::handles(UChar32 c) const {
+    return fHandled && fHandled->contains(c);
+}
+
+int32_t
+UnhandledEngine::findBreaks( UText *text,
+                             int32_t /* startPos */,
+                             int32_t endPos,
+                             UVector32 &/*foundBreaks*/ ) const {
+    UChar32 c = utext_current32(text); 
+    while((int32_t)utext_getNativeIndex(text) < endPos && fHandled->contains(c)) {
+        utext_next32(text);            // TODO:  recast loop to work with post-increment operations.
+        c = utext_current32(text);
+    }
+    return 0;
+}
+
+void
+UnhandledEngine::handleCharacter(UChar32 c) {
+    if (fHandled == nullptr) {
+        fHandled = new UnicodeSet();
+        if (fHandled == nullptr) {
+            return;
+        }
+    }
+    if (!fHandled->contains(c)) {
+        UErrorCode status = U_ZERO_ERROR;
+        // Apply the entire script of the character.
+        int32_t script = u_getIntPropertyValue(c, UCHAR_SCRIPT);
+        fHandled->applyIntPropertyValue(UCHAR_SCRIPT, script, status);
+    }
+}
+
+/*
+ ******************************************************************
+ */
+
+ICULanguageBreakFactory::ICULanguageBreakFactory(UErrorCode &/*status*/) {
+    fEngines = 0;
+}
+
+ICULanguageBreakFactory::~ICULanguageBreakFactory() {
+    if (fEngines != 0) {
+        delete fEngines;
+    }
+}
+
+U_NAMESPACE_END
+U_CDECL_BEGIN
+static void U_CALLCONV _deleteEngine(void *obj) {
+    delete (const icu::LanguageBreakEngine *) obj;
+}
+U_CDECL_END
+U_NAMESPACE_BEGIN
+
+const LanguageBreakEngine *
+ICULanguageBreakFactory::getEngineFor(UChar32 c) {
+    const LanguageBreakEngine *lbe = NULL;
+    UErrorCode  status = U_ZERO_ERROR;
+
+    static UMutex gBreakEngineMutex;
+    Mutex m(&gBreakEngineMutex);
+
+    if (fEngines == NULL) {
+        UStack  *engines = new UStack(_deleteEngine, NULL, status);
+        if (U_FAILURE(status) || engines == NULL) {
+            // Note: no way to return error code to caller.
+            delete engines;
+            return NULL;
+        }
+        fEngines = engines;
+    } else {
+        int32_t i = fEngines->size();
+        while (--i >= 0) {
+            lbe = (const LanguageBreakEngine *)(fEngines->elementAt(i));
+            if (lbe != NULL && lbe->handles(c)) {
+                return lbe;
+            }
+        }
+    }
+    
+    // We didn't find an engine. Create one.
+    lbe = loadEngineFor(c);
+    if (lbe != NULL) {
+        fEngines->push((void *)lbe, status);
+    }
+    return lbe;
+}
+
+const LanguageBreakEngine *
+ICULanguageBreakFactory::loadEngineFor(UChar32 c) {
+    UErrorCode status = U_ZERO_ERROR;
+    UScriptCode code = uscript_getScript(c, &status);
+    if (U_SUCCESS(status)) {
+        DictionaryMatcher *m = loadDictionaryMatcherFor(code);
+        if (m != NULL) {
+            const LanguageBreakEngine *engine = NULL;
+            switch(code) {
+            case USCRIPT_THAI:
+                engine = new ThaiBreakEngine(m, status);
+                break;
+            case USCRIPT_LAO:
+                engine = new LaoBreakEngine(m, status);
+                break;
+            case USCRIPT_MYANMAR:
+                engine = new BurmeseBreakEngine(m, status);
+                break;
+            case USCRIPT_KHMER:
+                engine = new KhmerBreakEngine(m, status);
+                break;
+
+#if !UCONFIG_NO_NORMALIZATION
+                // CJK not available w/o normalization
+            case USCRIPT_HANGUL:
+                engine = new CjkBreakEngine(m, kKorean, status);
+                break;
+
+            // use same BreakEngine and dictionary for both Chinese and Japanese
+            case USCRIPT_HIRAGANA:
+            case USCRIPT_KATAKANA:
+            case USCRIPT_HAN:
+                engine = new CjkBreakEngine(m, kChineseJapanese, status);
+                break;
+#if 0
+            // TODO: Have to get some characters with script=common handled
+            // by CjkBreakEngine (e.g. U+309B). Simply subjecting
+            // them to CjkBreakEngine does not work. The engine has to
+            // special-case them.
+            case USCRIPT_COMMON:
+            {
+                UBlockCode block = ublock_getCode(code);
+                if (block == UBLOCK_HIRAGANA || block == UBLOCK_KATAKANA)
+                   engine = new CjkBreakEngine(dict, kChineseJapanese, status);
+                break;
+            }
+#endif
+#endif
+
+            default:
+                break;
+            }
+            if (engine == NULL) {
+                delete m;
+            }
+            else if (U_FAILURE(status)) {
+                delete engine;
+                engine = NULL;
+            }
+            return engine;
+        }
+    }
+    return NULL;
+}
+
+DictionaryMatcher *
+ICULanguageBreakFactory::loadDictionaryMatcherFor(UScriptCode script) { 
+    UErrorCode status = U_ZERO_ERROR;
+    // open root from brkitr tree.
+    UResourceBundle *b = ures_open(U_ICUDATA_BRKITR, "", &status);
+    b = ures_getByKeyWithFallback(b, "dictionaries", b, &status);
+    int32_t dictnlength = 0;
+    const UChar *dictfname =
+        ures_getStringByKeyWithFallback(b, uscript_getShortName(script), &dictnlength, &status);
+    if (U_FAILURE(status)) {
+        ures_close(b);
+        return NULL;
+    }
+    CharString dictnbuf;
+    CharString ext;
+    const UChar *extStart = u_memrchr(dictfname, 0x002e, dictnlength);  // last dot
+    if (extStart != NULL) {
+        int32_t len = (int32_t)(extStart - dictfname);
+        ext.appendInvariantChars(UnicodeString(FALSE, extStart + 1, dictnlength - len - 1), status);
+        dictnlength = len;
+    }
+    dictnbuf.appendInvariantChars(UnicodeString(FALSE, dictfname, dictnlength), status);
+    ures_close(b);
+
+    UDataMemory *file = udata_open(U_ICUDATA_BRKITR, ext.data(), dictnbuf.data(), &status);
+    if (U_SUCCESS(status)) {
+        // build trie
+        const uint8_t *data = (const uint8_t *)udata_getMemory(file);
+        const int32_t *indexes = (const int32_t *)data;
+        const int32_t offset = indexes[DictionaryData::IX_STRING_TRIE_OFFSET];
+        const int32_t trieType = indexes[DictionaryData::IX_TRIE_TYPE] & DictionaryData::TRIE_TYPE_MASK;
+        DictionaryMatcher *m = NULL;
+        if (trieType == DictionaryData::TRIE_TYPE_BYTES) {
+            const int32_t transform = indexes[DictionaryData::IX_TRANSFORM];
+            const char *characters = (const char *)(data + offset);
+            m = new BytesDictionaryMatcher(characters, transform, file);
+        }
+        else if (trieType == DictionaryData::TRIE_TYPE_UCHARS) {
+            const UChar *characters = (const UChar *)(data + offset);
+            m = new UCharsDictionaryMatcher(characters, file);
+        }
+        if (m == NULL) {
+            // no matcher exists to take ownership - either we are an invalid 
+            // type or memory allocation failed
+            udata_close(file);
+        }
+        return m;
+    } else if (dictfname != NULL) {
+        // we don't have a dictionary matcher.
+        // returning NULL here will cause us to fail to find a dictionary break engine, as expected
+        status = U_ZERO_ERROR;
+        return NULL;
+    }
+    return NULL;
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/brkeng.h b/thirdparty/icu4c/common/brkeng.h
new file mode 100644
index 0000000000..155433b89a
--- /dev/null
+++ b/thirdparty/icu4c/common/brkeng.h
@@ -0,0 +1,271 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ ************************************************************************************
+ * Copyright (C) 2006-2012, International Business Machines Corporation and others. *
+ * All Rights Reserved.                                                             *
+ ************************************************************************************
+ */
+
+#ifndef BRKENG_H
+#define BRKENG_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "unicode/utext.h"
+#include "unicode/uscript.h"
+
+U_NAMESPACE_BEGIN
+
+class UnicodeSet;
+class UStack;
+class UVector32;
+class DictionaryMatcher;
+
+/*******************************************************************
+ * LanguageBreakEngine
+ */
+
+/**
+ * <p>LanguageBreakEngines implement language-specific knowledge for
+ * finding text boundaries within a run of characters belonging to a
+ * specific set. The boundaries will be of a specific kind, e.g. word,
+ * line, etc.</p>
+ *
+ * <p>LanguageBreakEngines should normally be implemented so as to
+ * be shared between threads without locking.</p>
+ */
+class LanguageBreakEngine : public UMemory {
+ public:
+
+  /**
+   * <p>Default constructor.</p>
+   *
+   */
+  LanguageBreakEngine();
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~LanguageBreakEngine();
+
+ /**
+  * <p>Indicate whether this engine handles a particular character for
+  * a particular kind of break.</p>
+  *
+  * @param c A character which begins a run that the engine might handle
+  * @return true if this engine handles the particular character and break
+  * type.
+  */
+  virtual UBool handles(UChar32 c) const = 0;
+
+ /**
+  * <p>Find any breaks within a run in the supplied text.</p>
+  *
+  * @param text A UText representing the text. The
+  * iterator is left at the end of the run of characters which the engine
+  * is capable of handling.
+  * @param startPos The start of the run within the supplied text.
+  * @param endPos The end of the run within the supplied text.
+  * @param foundBreaks A Vector of int32_t to receive the breaks.
+  * @return The number of breaks found.
+  */
+  virtual int32_t findBreaks( UText *text,
+                              int32_t startPos,
+                              int32_t endPos,
+                              UVector32 &foundBreaks ) const = 0;
+
+};
+
+/*******************************************************************
+ * LanguageBreakFactory
+ */
+
+/**
+ * <p>LanguageBreakFactorys find and return a LanguageBreakEngine
+ * that can determine breaks for characters in a specific set, if
+ * such an object can be found.</p>
+ *
+ * <p>If a LanguageBreakFactory is to be shared between threads,
+ * appropriate synchronization must be used; there is none internal
+ * to the factory.</p>
+ *
+ * <p>A LanguageBreakEngine returned by a LanguageBreakFactory can
+ * normally be shared between threads without synchronization, unless
+ * the specific subclass of LanguageBreakFactory indicates otherwise.</p>
+ *
+ * <p>A LanguageBreakFactory is responsible for deleting any LanguageBreakEngine
+ * it returns when it itself is deleted, unless the specific subclass of
+ * LanguageBreakFactory indicates otherwise. Naturally, the factory should
+ * not be deleted until the LanguageBreakEngines it has returned are no
+ * longer needed.</p>
+ */
+class LanguageBreakFactory : public UMemory {
+ public:
+
+  /**
+   * <p>Default constructor.</p>
+   *
+   */
+  LanguageBreakFactory();
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~LanguageBreakFactory();
+
+ /**
+  * <p>Find and return a LanguageBreakEngine that can find the desired
+  * kind of break for the set of characters to which the supplied
+  * character belongs. It is up to the set of available engines to
+  * determine what the sets of characters are.</p>
+  *
+  * @param c A character that begins a run for which a LanguageBreakEngine is
+  * sought.
+  * @return A LanguageBreakEngine with the desired characteristics, or 0.
+  */
+  virtual const LanguageBreakEngine *getEngineFor(UChar32 c) = 0;
+
+};
+
+/*******************************************************************
+ * UnhandledEngine
+ */
+
+/**
+ * <p>UnhandledEngine is a special subclass of LanguageBreakEngine that
+ * handles characters that no other LanguageBreakEngine is available to
+ * handle. It is told the character and the type of break; at its
+ * discretion it may handle more than the specified character (e.g.,
+ * the entire script to which that character belongs.</p>
+ *
+ * <p>UnhandledEngines may not be shared between threads without
+ * external synchronization.</p>
+ */
+
+class UnhandledEngine : public LanguageBreakEngine {
+ private:
+
+    /**
+     * The sets of characters handled.
+     * @internal
+     */
+
+  UnicodeSet    *fHandled;
+
+ public:
+
+  /**
+   * <p>Default constructor.</p>
+   *
+   */
+  UnhandledEngine(UErrorCode &status);
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~UnhandledEngine();
+
+ /**
+  * <p>Indicate whether this engine handles a particular character for
+  * a particular kind of break.</p>
+  *
+  * @param c A character which begins a run that the engine might handle
+  * @return true if this engine handles the particular character and break
+  * type.
+  */
+  virtual UBool handles(UChar32 c) const;
+
+ /**
+  * <p>Find any breaks within a run in the supplied text.</p>
+  *
+  * @param text A UText representing the text (TODO: UText). The
+  * iterator is left at the end of the run of characters which the engine
+  * is capable of handling.
+  * @param startPos The start of the run within the supplied text.
+  * @param endPos The end of the run within the supplied text.
+  * @param foundBreaks An allocated C array of the breaks found, if any
+  * @return The number of breaks found.
+  */
+  virtual int32_t findBreaks( UText *text,
+                              int32_t startPos,
+                              int32_t endPos,
+                              UVector32 &foundBreaks ) const;
+
+ /**
+  * <p>Tell the engine to handle a particular character and break type.</p>
+  *
+  * @param c A character which the engine should handle
+  */
+  virtual void handleCharacter(UChar32 c);
+
+};
+
+/*******************************************************************
+ * ICULanguageBreakFactory
+ */
+
+/**
+ * <p>ICULanguageBreakFactory is the default LanguageBreakFactory for
+ * ICU. It creates dictionary-based LanguageBreakEngines from dictionary
+ * data in the ICU data file.</p>
+ */
+class ICULanguageBreakFactory : public LanguageBreakFactory {
+ private:
+
+    /**
+     * The stack of break engines created by this factory
+     * @internal
+     */
+
+  UStack    *fEngines;
+
+ public:
+
+  /**
+   * <p>Standard constructor.</p>
+   *
+   */
+  ICULanguageBreakFactory(UErrorCode &status);
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~ICULanguageBreakFactory();
+
+ /**
+  * <p>Find and return a LanguageBreakEngine that can find the desired
+  * kind of break for the set of characters to which the supplied
+  * character belongs. It is up to the set of available engines to
+  * determine what the sets of characters are.</p>
+  *
+  * @param c A character that begins a run for which a LanguageBreakEngine is
+  * sought.
+  * @return A LanguageBreakEngine with the desired characteristics, or 0.
+  */
+  virtual const LanguageBreakEngine *getEngineFor(UChar32 c);
+
+protected:
+ /**
+  * <p>Create a LanguageBreakEngine for the set of characters to which
+  * the supplied character belongs, for the specified break type.</p>
+  *
+  * @param c A character that begins a run for which a LanguageBreakEngine is
+  * sought.
+  * @return A LanguageBreakEngine with the desired characteristics, or 0.
+  */
+  virtual const LanguageBreakEngine *loadEngineFor(UChar32 c);
+
+  /**
+   * <p>Create a DictionaryMatcher for the specified script and break type.</p>
+   * @param script An ISO 15924 script code that identifies the dictionary to be
+   * created.
+   * @return A DictionaryMatcher with the desired characteristics, or NULL.
+   */
+  virtual DictionaryMatcher *loadDictionaryMatcherFor(UScriptCode script);
+};
+
+U_NAMESPACE_END
+
+    /* BRKENG_H */
+#endif
diff --git a/thirdparty/icu4c/common/brkiter.cpp b/thirdparty/icu4c/common/brkiter.cpp
new file mode 100644
index 0000000000..b9b6ca65cd
--- /dev/null
+++ b/thirdparty/icu4c/common/brkiter.cpp
@@ -0,0 +1,527 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 1997-2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+*******************************************************************************
+*
+* File brkiter.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   02/18/97    aliu        Converted from OpenClass.  Added DONE.
+*   01/13/2000  helena      Added UErrorCode parameter to createXXXInstance methods.
+*****************************************************************************************
+*/
+
+// *****************************************************************************
+// This file was generated from the java source file BreakIterator.java
+// *****************************************************************************
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/rbbi.h"
+#include "unicode/brkiter.h"
+#include "unicode/udata.h"
+#include "unicode/ures.h"
+#include "unicode/ustring.h"
+#include "unicode/filteredbrk.h"
+#include "ucln_cmn.h"
+#include "cstring.h"
+#include "umutex.h"
+#include "servloc.h"
+#include "locbased.h"
+#include "uresimp.h"
+#include "uassert.h"
+#include "ubrkimpl.h"
+#include "utracimp.h"
+#include "charstr.h"
+
+// *****************************************************************************
+// class BreakIterator
+// This class implements methods for finding the location of boundaries in text.
+// Instances of BreakIterator maintain a current position and scan over text
+// returning the index of characters where boundaries occur.
+// *****************************************************************************
+
+U_NAMESPACE_BEGIN
+
+// -------------------------------------
+
+BreakIterator*
+BreakIterator::buildInstance(const Locale& loc, const char *type, UErrorCode &status)
+{
+    char fnbuff[256];
+    char ext[4]={'\0'};
+    CharString actualLocale;
+    int32_t size;
+    const UChar* brkfname = NULL;
+    UResourceBundle brkRulesStack;
+    UResourceBundle brkNameStack;
+    UResourceBundle *brkRules = &brkRulesStack;
+    UResourceBundle *brkName  = &brkNameStack;
+    RuleBasedBreakIterator *result = NULL;
+
+    if (U_FAILURE(status))
+        return NULL;
+
+    ures_initStackObject(brkRules);
+    ures_initStackObject(brkName);
+
+    // Get the locale
+    UResourceBundle *b = ures_openNoDefault(U_ICUDATA_BRKITR, loc.getName(), &status);
+
+    // Get the "boundaries" array.
+    if (U_SUCCESS(status)) {
+        brkRules = ures_getByKeyWithFallback(b, "boundaries", brkRules, &status);
+        // Get the string object naming the rules file
+        brkName = ures_getByKeyWithFallback(brkRules, type, brkName, &status);
+        // Get the actual string
+        brkfname = ures_getString(brkName, &size, &status);
+        U_ASSERT((size_t)size<sizeof(fnbuff));
+        if ((size_t)size>=sizeof(fnbuff)) {
+            size=0;
+            if (U_SUCCESS(status)) {
+                status = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+
+        // Use the string if we found it
+        if (U_SUCCESS(status) && brkfname) {
+            actualLocale.append(ures_getLocaleInternal(brkName, &status), -1, status);
+
+            UChar* extStart=u_strchr(brkfname, 0x002e);
+            int len = 0;
+            if(extStart!=NULL){
+                len = (int)(extStart-brkfname);
+                u_UCharsToChars(extStart+1, ext, sizeof(ext)); // nul terminates the buff
+                u_UCharsToChars(brkfname, fnbuff, len);
+            }
+            fnbuff[len]=0; // nul terminate
+        }
+    }
+
+    ures_close(brkRules);
+    ures_close(brkName);
+
+    UDataMemory* file = udata_open(U_ICUDATA_BRKITR, ext, fnbuff, &status);
+    if (U_FAILURE(status)) {
+        ures_close(b);
+        return NULL;
+    }
+
+    // Create a RuleBasedBreakIterator
+    result = new RuleBasedBreakIterator(file, status);
+
+    // If there is a result, set the valid locale and actual locale, and the kind
+    if (U_SUCCESS(status) && result != NULL) {
+        U_LOCALE_BASED(locBased, *(BreakIterator*)result);
+        locBased.setLocaleIDs(ures_getLocaleByType(b, ULOC_VALID_LOCALE, &status), 
+                              actualLocale.data());
+    }
+
+    ures_close(b);
+
+    if (U_FAILURE(status) && result != NULL) {  // Sometimes redundant check, but simple
+        delete result;
+        return NULL;
+    }
+
+    if (result == NULL) {
+        udata_close(file);
+        if (U_SUCCESS(status)) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+
+    return result;
+}
+
+// Creates a break iterator for word breaks.
+BreakIterator* U_EXPORT2
+BreakIterator::createWordInstance(const Locale& key, UErrorCode& status)
+{
+    return createInstance(key, UBRK_WORD, status);
+}
+
+// -------------------------------------
+
+// Creates a break iterator  for line breaks.
+BreakIterator* U_EXPORT2
+BreakIterator::createLineInstance(const Locale& key, UErrorCode& status)
+{
+    return createInstance(key, UBRK_LINE, status);
+}
+
+// -------------------------------------
+
+// Creates a break iterator  for character breaks.
+BreakIterator* U_EXPORT2
+BreakIterator::createCharacterInstance(const Locale& key, UErrorCode& status)
+{
+    return createInstance(key, UBRK_CHARACTER, status);
+}
+
+// -------------------------------------
+
+// Creates a break iterator  for sentence breaks.
+BreakIterator* U_EXPORT2
+BreakIterator::createSentenceInstance(const Locale& key, UErrorCode& status)
+{
+    return createInstance(key, UBRK_SENTENCE, status);
+}
+
+// -------------------------------------
+
+// Creates a break iterator for title casing breaks.
+BreakIterator* U_EXPORT2
+BreakIterator::createTitleInstance(const Locale& key, UErrorCode& status)
+{
+    return createInstance(key, UBRK_TITLE, status);
+}
+
+// -------------------------------------
+
+// Gets all the available locales that has localized text boundary data.
+const Locale* U_EXPORT2
+BreakIterator::getAvailableLocales(int32_t& count)
+{
+    return Locale::getAvailableLocales(count);
+}
+
+// ------------------------------------------
+//
+// Constructors, destructor and assignment operator
+//
+//-------------------------------------------
+
+BreakIterator::BreakIterator()
+{
+    *validLocale = *actualLocale = 0;
+}
+
+BreakIterator::BreakIterator(const BreakIterator &other) : UObject(other) {
+    uprv_strncpy(actualLocale, other.actualLocale, sizeof(actualLocale));
+    uprv_strncpy(validLocale, other.validLocale, sizeof(validLocale));
+}
+
+BreakIterator &BreakIterator::operator =(const BreakIterator &other) {
+    if (this != &other) {
+        uprv_strncpy(actualLocale, other.actualLocale, sizeof(actualLocale));
+        uprv_strncpy(validLocale, other.validLocale, sizeof(validLocale));
+    }
+    return *this;
+}
+
+BreakIterator::~BreakIterator()
+{
+}
+
+// ------------------------------------------
+//
+// Registration
+//
+//-------------------------------------------
+#if !UCONFIG_NO_SERVICE
+
+// -------------------------------------
+
+class ICUBreakIteratorFactory : public ICUResourceBundleFactory {
+public:
+    virtual ~ICUBreakIteratorFactory();
+protected:
+    virtual UObject* handleCreate(const Locale& loc, int32_t kind, const ICUService* /*service*/, UErrorCode& status) const {
+        return BreakIterator::makeInstance(loc, kind, status);
+    }
+};
+
+ICUBreakIteratorFactory::~ICUBreakIteratorFactory() {}
+
+// -------------------------------------
+
+class ICUBreakIteratorService : public ICULocaleService {
+public:
+    ICUBreakIteratorService()
+        : ICULocaleService(UNICODE_STRING("Break Iterator", 14))
+    {
+        UErrorCode status = U_ZERO_ERROR;
+        registerFactory(new ICUBreakIteratorFactory(), status);
+    }
+
+    virtual ~ICUBreakIteratorService();
+
+    virtual UObject* cloneInstance(UObject* instance) const {
+        return ((BreakIterator*)instance)->clone();
+    }
+
+    virtual UObject* handleDefault(const ICUServiceKey& key, UnicodeString* /*actualID*/, UErrorCode& status) const {
+        LocaleKey& lkey = (LocaleKey&)key;
+        int32_t kind = lkey.kind();
+        Locale loc;
+        lkey.currentLocale(loc);
+        return BreakIterator::makeInstance(loc, kind, status);
+    }
+
+    virtual UBool isDefault() const {
+        return countFactories() == 1;
+    }
+};
+
+ICUBreakIteratorService::~ICUBreakIteratorService() {}
+
+// -------------------------------------
+
+// defined in ucln_cmn.h
+U_NAMESPACE_END
+
+static icu::UInitOnce gInitOnceBrkiter = U_INITONCE_INITIALIZER;
+static icu::ICULocaleService* gService = NULL;
+
+
+
+/**
+ * Release all static memory held by breakiterator.
+ */
+U_CDECL_BEGIN
+static UBool U_CALLCONV breakiterator_cleanup(void) {
+#if !UCONFIG_NO_SERVICE
+    if (gService) {
+        delete gService;
+        gService = NULL;
+    }
+    gInitOnceBrkiter.reset();
+#endif
+    return TRUE;
+}
+U_CDECL_END
+U_NAMESPACE_BEGIN
+
+static void U_CALLCONV 
+initService(void) {
+    gService = new ICUBreakIteratorService();
+    ucln_common_registerCleanup(UCLN_COMMON_BREAKITERATOR, breakiterator_cleanup);
+}
+
+static ICULocaleService*
+getService(void)
+{
+    umtx_initOnce(gInitOnceBrkiter, &initService);
+    return gService;
+}
+
+
+// -------------------------------------
+
+static inline UBool
+hasService(void)
+{
+    return !gInitOnceBrkiter.isReset() && getService() != NULL;
+}
+
+// -------------------------------------
+
+URegistryKey U_EXPORT2
+BreakIterator::registerInstance(BreakIterator* toAdopt, const Locale& locale, UBreakIteratorType kind, UErrorCode& status)
+{
+    ICULocaleService *service = getService();
+    if (service == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    return service->registerInstance(toAdopt, locale, kind, status);
+}
+
+// -------------------------------------
+
+UBool U_EXPORT2
+BreakIterator::unregister(URegistryKey key, UErrorCode& status)
+{
+    if (U_SUCCESS(status)) {
+        if (hasService()) {
+            return gService->unregister(key, status);
+        }
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return FALSE;
+}
+
+// -------------------------------------
+
+StringEnumeration* U_EXPORT2
+BreakIterator::getAvailableLocales(void)
+{
+    ICULocaleService *service = getService();
+    if (service == NULL) {
+        return NULL;
+    }
+    return service->getAvailableLocales();
+}
+#endif /* UCONFIG_NO_SERVICE */
+
+// -------------------------------------
+
+BreakIterator*
+BreakIterator::createInstance(const Locale& loc, int32_t kind, UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+#if !UCONFIG_NO_SERVICE
+    if (hasService()) {
+        Locale actualLoc("");
+        BreakIterator *result = (BreakIterator*)gService->get(loc, kind, &actualLoc, status);
+        // TODO: The way the service code works in ICU 2.8 is that if
+        // there is a real registered break iterator, the actualLoc
+        // will be populated, but if the handleDefault path is taken
+        // (because nothing is registered that can handle the
+        // requested locale) then the actualLoc comes back empty.  In
+        // that case, the returned object already has its actual/valid
+        // locale data populated (by makeInstance, which is what
+        // handleDefault calls), so we don't touch it.  YES, A COMMENT
+        // THIS LONG is a sign of bad code -- so the action item is to
+        // revisit this in ICU 3.0 and clean it up/fix it/remove it.
+        if (U_SUCCESS(status) && (result != NULL) && *actualLoc.getName() != 0) {
+            U_LOCALE_BASED(locBased, *result);
+            locBased.setLocaleIDs(actualLoc.getName(), actualLoc.getName());
+        }
+        return result;
+    }
+    else
+#endif
+    {
+        return makeInstance(loc, kind, status);
+    }
+}
+
+// -------------------------------------
+enum { kKeyValueLenMax = 32 };
+
+BreakIterator*
+BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
+{
+
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+    char lbType[kKeyValueLenMax];
+
+    BreakIterator *result = NULL;
+    switch (kind) {
+    case UBRK_CHARACTER:
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_CHARACTER);
+            result = BreakIterator::buildInstance(loc, "grapheme", status);
+            UTRACE_EXIT_STATUS(status);
+        }
+        break;
+    case UBRK_WORD:
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_WORD);
+            result = BreakIterator::buildInstance(loc, "word", status);
+            UTRACE_EXIT_STATUS(status);
+        }
+        break;
+    case UBRK_LINE:
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_LINE);
+            uprv_strcpy(lbType, "line");
+            char lbKeyValue[kKeyValueLenMax] = {0};
+            UErrorCode kvStatus = U_ZERO_ERROR;
+            int32_t kLen = loc.getKeywordValue("lb", lbKeyValue, kKeyValueLenMax, kvStatus);
+            if (U_SUCCESS(kvStatus) && kLen > 0 && (uprv_strcmp(lbKeyValue,"strict")==0 || uprv_strcmp(lbKeyValue,"normal")==0 || uprv_strcmp(lbKeyValue,"loose")==0)) {
+                uprv_strcat(lbType, "_");
+                uprv_strcat(lbType, lbKeyValue);
+            }
+            result = BreakIterator::buildInstance(loc, lbType, status);
+
+            UTRACE_DATA1(UTRACE_INFO, "lb=%s", lbKeyValue);
+            UTRACE_EXIT_STATUS(status);
+        }
+        break;
+    case UBRK_SENTENCE:
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_SENTENCE);
+            result = BreakIterator::buildInstance(loc, "sentence", status);
+#if !UCONFIG_NO_FILTERED_BREAK_ITERATION
+            char ssKeyValue[kKeyValueLenMax] = {0};
+            UErrorCode kvStatus = U_ZERO_ERROR;
+            int32_t kLen = loc.getKeywordValue("ss", ssKeyValue, kKeyValueLenMax, kvStatus);
+            if (U_SUCCESS(kvStatus) && kLen > 0 && uprv_strcmp(ssKeyValue,"standard")==0) {
+                FilteredBreakIteratorBuilder* fbiBuilder = FilteredBreakIteratorBuilder::createInstance(loc, kvStatus);
+                if (U_SUCCESS(kvStatus)) {
+                    result = fbiBuilder->build(result, status);
+                    delete fbiBuilder;
+                }
+            }
+#endif
+            UTRACE_EXIT_STATUS(status);
+        }
+        break;
+    case UBRK_TITLE:
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_TITLE);
+            result = BreakIterator::buildInstance(loc, "title", status);
+            UTRACE_EXIT_STATUS(status);
+        }
+        break;
+    default:
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+    return result;
+}
+
+Locale
+BreakIterator::getLocale(ULocDataLocaleType type, UErrorCode& status) const {
+    U_LOCALE_BASED(locBased, *this);
+    return locBased.getLocale(type, status);
+}
+
+const char *
+BreakIterator::getLocaleID(ULocDataLocaleType type, UErrorCode& status) const {
+    U_LOCALE_BASED(locBased, *this);
+    return locBased.getLocaleID(type, status);
+}
+
+
+// This implementation of getRuleStatus is a do-nothing stub, here to
+// provide a default implementation for any derived BreakIterator classes that
+// do not implement it themselves.
+int32_t BreakIterator::getRuleStatus() const {
+    return 0;
+}
+
+// This implementation of getRuleStatusVec is a do-nothing stub, here to
+// provide a default implementation for any derived BreakIterator classes that
+// do not implement it themselves.
+int32_t BreakIterator::getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return 0;
+    }
+    if (capacity < 1) {
+        status = U_BUFFER_OVERFLOW_ERROR;
+        return 1;
+    }
+    *fillInVec = 0;
+    return 1;
+}
+
+BreakIterator::BreakIterator (const Locale& valid, const Locale& actual) {
+  U_LOCALE_BASED(locBased, (*this));
+  locBased.setLocaleIDs(valid, actual);
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+//eof
diff --git a/thirdparty/icu4c/common/bytesinkutil.cpp b/thirdparty/icu4c/common/bytesinkutil.cpp
new file mode 100644
index 0000000000..c64a845f87
--- /dev/null
+++ b/thirdparty/icu4c/common/bytesinkutil.cpp
@@ -0,0 +1,161 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// bytesinkutil.cpp
+// created: 2017sep14 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/bytestream.h"
+#include "unicode/edits.h"
+#include "unicode/stringoptions.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+UBool
+ByteSinkUtil::appendChange(int32_t length, const char16_t *s16, int32_t s16Length,
+                           ByteSink &sink, Edits *edits, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    char scratch[200];
+    int32_t s8Length = 0;
+    for (int32_t i = 0; i < s16Length;) {
+        int32_t capacity;
+        int32_t desiredCapacity = s16Length - i;
+        if (desiredCapacity < (INT32_MAX / 3)) {
+            desiredCapacity *= 3;  // max 3 UTF-8 bytes per UTF-16 code unit
+        } else if (desiredCapacity < (INT32_MAX / 2)) {
+            desiredCapacity *= 2;
+        } else {
+            desiredCapacity = INT32_MAX;
+        }
+        char *buffer = sink.GetAppendBuffer(U8_MAX_LENGTH, desiredCapacity,
+                                            scratch, UPRV_LENGTHOF(scratch), &capacity);
+        capacity -= U8_MAX_LENGTH - 1;
+        int32_t j = 0;
+        for (; i < s16Length && j < capacity;) {
+            UChar32 c;
+            U16_NEXT_UNSAFE(s16, i, c);
+            U8_APPEND_UNSAFE(buffer, j, c);
+        }
+        if (j > (INT32_MAX - s8Length)) {
+            errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            return FALSE;
+        }
+        sink.Append(buffer, j);
+        s8Length += j;
+    }
+    if (edits != nullptr) {
+        edits->addReplace(length, s8Length);
+    }
+    return TRUE;
+}
+
+UBool
+ByteSinkUtil::appendChange(const uint8_t *s, const uint8_t *limit,
+                           const char16_t *s16, int32_t s16Length,
+                           ByteSink &sink, Edits *edits, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    if ((limit - s) > INT32_MAX) {
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return FALSE;
+    }
+    return appendChange((int32_t)(limit - s), s16, s16Length, sink, edits, errorCode);
+}
+
+void
+ByteSinkUtil::appendCodePoint(int32_t length, UChar32 c, ByteSink &sink, Edits *edits) {
+    char s8[U8_MAX_LENGTH];
+    int32_t s8Length = 0;
+    U8_APPEND_UNSAFE(s8, s8Length, c);
+    if (edits != nullptr) {
+        edits->addReplace(length, s8Length);
+    }
+    sink.Append(s8, s8Length);
+}
+
+namespace {
+
+// See unicode/utf8.h U8_APPEND_UNSAFE().
+inline uint8_t getTwoByteLead(UChar32 c) { return (uint8_t)((c >> 6) | 0xc0); }
+inline uint8_t getTwoByteTrail(UChar32 c) { return (uint8_t)((c & 0x3f) | 0x80); }
+
+}  // namespace
+
+void
+ByteSinkUtil::appendTwoBytes(UChar32 c, ByteSink &sink) {
+    U_ASSERT(0x80 <= c && c <= 0x7ff);  // 2-byte UTF-8
+    char s8[2] = { (char)getTwoByteLead(c), (char)getTwoByteTrail(c) };
+    sink.Append(s8, 2);
+}
+
+void
+ByteSinkUtil::appendNonEmptyUnchanged(const uint8_t *s, int32_t length,
+                                      ByteSink &sink, uint32_t options, Edits *edits) {
+    U_ASSERT(length > 0);
+    if (edits != nullptr) {
+        edits->addUnchanged(length);
+    }
+    if ((options & U_OMIT_UNCHANGED_TEXT) == 0) {
+        sink.Append(reinterpret_cast<const char *>(s), length);
+    }
+}
+
+UBool
+ByteSinkUtil::appendUnchanged(const uint8_t *s, const uint8_t *limit,
+                              ByteSink &sink, uint32_t options, Edits *edits,
+                              UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    if ((limit - s) > INT32_MAX) {
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return FALSE;
+    }
+    int32_t length = (int32_t)(limit - s);
+    if (length > 0) {
+        appendNonEmptyUnchanged(s, length, sink, options, edits);
+    }
+    return TRUE;
+}
+
+CharStringByteSink::CharStringByteSink(CharString* dest) : dest_(*dest) {
+}
+
+CharStringByteSink::~CharStringByteSink() = default;
+
+void
+CharStringByteSink::Append(const char* bytes, int32_t n) {
+    UErrorCode status = U_ZERO_ERROR;
+    dest_.append(bytes, n, status);
+    // Any errors are silently ignored.
+}
+
+char*
+CharStringByteSink::GetAppendBuffer(int32_t min_capacity,
+                                    int32_t desired_capacity_hint,
+                                    char* scratch,
+                                    int32_t scratch_capacity,
+                                    int32_t* result_capacity) {
+    if (min_capacity < 1 || scratch_capacity < min_capacity) {
+        *result_capacity = 0;
+        return nullptr;
+    }
+
+    UErrorCode status = U_ZERO_ERROR;
+    char* result = dest_.getAppendBuffer(
+            min_capacity,
+            desired_capacity_hint,
+            *result_capacity,
+            status);
+    if (U_SUCCESS(status)) {
+        return result;
+    }
+
+    *result_capacity = scratch_capacity;
+    return scratch;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bytesinkutil.h b/thirdparty/icu4c/common/bytesinkutil.h
new file mode 100644
index 0000000000..ab2516432d
--- /dev/null
+++ b/thirdparty/icu4c/common/bytesinkutil.h
@@ -0,0 +1,83 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// bytesinkutil.h
+// created: 2017sep14 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/bytestream.h"
+#include "unicode/edits.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+class ByteSink;
+class CharString;
+class Edits;
+
+class U_COMMON_API ByteSinkUtil {
+public:
+    ByteSinkUtil() = delete;  // all static
+
+    /** (length) bytes were mapped to valid (s16, s16Length). */
+    static UBool appendChange(int32_t length,
+                              const char16_t *s16, int32_t s16Length,
+                              ByteSink &sink, Edits *edits, UErrorCode &errorCode);
+
+    /** The bytes at [s, limit[ were mapped to valid (s16, s16Length). */
+    static UBool appendChange(const uint8_t *s, const uint8_t *limit,
+                              const char16_t *s16, int32_t s16Length,
+                              ByteSink &sink, Edits *edits, UErrorCode &errorCode);
+
+    /** (length) bytes were mapped/changed to valid code point c. */
+    static void appendCodePoint(int32_t length, UChar32 c, ByteSink &sink, Edits *edits = nullptr);
+
+    /** The few bytes at [src, nextSrc[ were mapped/changed to valid code point c. */
+    static inline void appendCodePoint(const uint8_t *src, const uint8_t *nextSrc, UChar32 c,
+                                       ByteSink &sink, Edits *edits = nullptr) {
+        appendCodePoint((int32_t)(nextSrc - src), c, sink, edits);
+    }
+
+    /** Append the two-byte character (U+0080..U+07FF). */
+    static void appendTwoBytes(UChar32 c, ByteSink &sink);
+
+    static UBool appendUnchanged(const uint8_t *s, int32_t length,
+                                 ByteSink &sink, uint32_t options, Edits *edits,
+                                 UErrorCode &errorCode) {
+        if (U_FAILURE(errorCode)) { return false; }
+        if (length > 0) { appendNonEmptyUnchanged(s, length, sink, options, edits); }
+        return true;
+    }
+
+    static UBool appendUnchanged(const uint8_t *s, const uint8_t *limit,
+                                 ByteSink &sink, uint32_t options, Edits *edits,
+                                 UErrorCode &errorCode);
+
+private:
+    static void appendNonEmptyUnchanged(const uint8_t *s, int32_t length,
+                                        ByteSink &sink, uint32_t options, Edits *edits);
+};
+
+class U_COMMON_API CharStringByteSink : public ByteSink {
+public:
+    CharStringByteSink(CharString* dest);
+    ~CharStringByteSink() override;
+
+    CharStringByteSink() = delete;
+    CharStringByteSink(const CharStringByteSink&) = delete;
+    CharStringByteSink& operator=(const CharStringByteSink&) = delete;
+
+    void Append(const char* bytes, int32_t n) override;
+
+    char* GetAppendBuffer(int32_t min_capacity,
+                          int32_t desired_capacity_hint,
+                          char* scratch,
+                          int32_t scratch_capacity,
+                          int32_t* result_capacity) override;
+
+private:
+    CharString& dest_;
+};
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bytestream.cpp b/thirdparty/icu4c/common/bytestream.cpp
new file mode 100644
index 0000000000..0d0e4dda39
--- /dev/null
+++ b/thirdparty/icu4c/common/bytestream.cpp
@@ -0,0 +1,85 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+// Copyright (C) 2009-2011, International Business Machines
+// Corporation and others. All Rights Reserved.
+//
+// Copyright 2007 Google Inc. All Rights Reserved.
+// Author: sanjay@google.com (Sanjay Ghemawat)
+
+#include "unicode/utypes.h"
+#include "unicode/bytestream.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+ByteSink::~ByteSink() {}
+
+char* ByteSink::GetAppendBuffer(int32_t min_capacity,
+                                int32_t /*desired_capacity_hint*/,
+                                char* scratch, int32_t scratch_capacity,
+                                int32_t* result_capacity) {
+  if (min_capacity < 1 || scratch_capacity < min_capacity) {
+    *result_capacity = 0;
+    return NULL;
+  }
+  *result_capacity = scratch_capacity;
+  return scratch;
+}
+
+void ByteSink::Flush() {}
+
+CheckedArrayByteSink::CheckedArrayByteSink(char* outbuf, int32_t capacity)
+    : outbuf_(outbuf), capacity_(capacity < 0 ? 0 : capacity),
+      size_(0), appended_(0), overflowed_(FALSE) {
+}
+
+CheckedArrayByteSink::~CheckedArrayByteSink() {}
+
+CheckedArrayByteSink& CheckedArrayByteSink::Reset() {
+  size_ = appended_ = 0;
+  overflowed_ = FALSE;
+  return *this;
+}
+
+void CheckedArrayByteSink::Append(const char* bytes, int32_t n) {
+  if (n <= 0) {
+    return;
+  }
+  if (n > (INT32_MAX - appended_)) {
+    // TODO: Report as integer overflow, not merely buffer overflow.
+    appended_ = INT32_MAX;
+    overflowed_ = TRUE;
+    return;
+  }
+  appended_ += n;
+  int32_t available = capacity_ - size_;
+  if (n > available) {
+    n = available;
+    overflowed_ = TRUE;
+  }
+  if (n > 0 && bytes != (outbuf_ + size_)) {
+    uprv_memcpy(outbuf_ + size_, bytes, n);
+  }
+  size_ += n;
+}
+
+char* CheckedArrayByteSink::GetAppendBuffer(int32_t min_capacity,
+                                            int32_t /*desired_capacity_hint*/,
+                                            char* scratch,
+                                            int32_t scratch_capacity,
+                                            int32_t* result_capacity) {
+  if (min_capacity < 1 || scratch_capacity < min_capacity) {
+    *result_capacity = 0;
+    return NULL;
+  }
+  int32_t available = capacity_ - size_;
+  if (available >= min_capacity) {
+    *result_capacity = available;
+    return outbuf_ + size_;
+  } else {
+    *result_capacity = scratch_capacity;
+    return scratch;
+  }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bytestrie.cpp b/thirdparty/icu4c/common/bytestrie.cpp
new file mode 100644
index 0000000000..c4d498c4bf
--- /dev/null
+++ b/thirdparty/icu4c/common/bytestrie.cpp
@@ -0,0 +1,441 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  bytestrie.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010sep25
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/bytestream.h"
+#include "unicode/bytestrie.h"
+#include "unicode/uobject.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+BytesTrie::~BytesTrie() {
+    uprv_free(ownedArray_);
+}
+
+// lead byte already shifted right by 1.
+int32_t
+BytesTrie::readValue(const uint8_t *pos, int32_t leadByte) {
+    int32_t value;
+    if(leadByte<kMinTwoByteValueLead) {
+        value=leadByte-kMinOneByteValueLead;
+    } else if(leadByte<kMinThreeByteValueLead) {
+        value=((leadByte-kMinTwoByteValueLead)<<8)|*pos;
+    } else if(leadByte<kFourByteValueLead) {
+        value=((leadByte-kMinThreeByteValueLead)<<16)|(pos[0]<<8)|pos[1];
+    } else if(leadByte==kFourByteValueLead) {
+        value=(pos[0]<<16)|(pos[1]<<8)|pos[2];
+    } else {
+        value=(pos[0]<<24)|(pos[1]<<16)|(pos[2]<<8)|pos[3];
+    }
+    return value;
+}
+
+const uint8_t *
+BytesTrie::jumpByDelta(const uint8_t *pos) {
+    int32_t delta=*pos++;
+    if(delta<kMinTwoByteDeltaLead) {
+        // nothing to do
+    } else if(delta<kMinThreeByteDeltaLead) {
+        delta=((delta-kMinTwoByteDeltaLead)<<8)|*pos++;
+    } else if(delta<kFourByteDeltaLead) {
+        delta=((delta-kMinThreeByteDeltaLead)<<16)|(pos[0]<<8)|pos[1];
+        pos+=2;
+    } else if(delta==kFourByteDeltaLead) {
+        delta=(pos[0]<<16)|(pos[1]<<8)|pos[2];
+        pos+=3;
+    } else {
+        delta=(pos[0]<<24)|(pos[1]<<16)|(pos[2]<<8)|pos[3];
+        pos+=4;
+    }
+    return pos+delta;
+}
+
+UStringTrieResult
+BytesTrie::current() const {
+    const uint8_t *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    } else {
+        int32_t node;
+        return (remainingMatchLength_<0 && (node=*pos)>=kMinValueLead) ?
+                valueResult(node) : USTRINGTRIE_NO_VALUE;
+    }
+}
+
+UStringTrieResult
+BytesTrie::branchNext(const uint8_t *pos, int32_t length, int32_t inByte) {
+    // Branch according to the current byte.
+    if(length==0) {
+        length=*pos++;
+    }
+    ++length;
+    // The length of the branch is the number of bytes to select from.
+    // The data structure encodes a binary search.
+    while(length>kMaxBranchLinearSubNodeLength) {
+        if(inByte<*pos++) {
+            length>>=1;
+            pos=jumpByDelta(pos);
+        } else {
+            length=length-(length>>1);
+            pos=skipDelta(pos);
+        }
+    }
+    // Drop down to linear search for the last few bytes.
+    // length>=2 because the loop body above sees length>kMaxBranchLinearSubNodeLength>=3
+    // and divides length by 2.
+    do {
+        if(inByte==*pos++) {
+            UStringTrieResult result;
+            int32_t node=*pos;
+            U_ASSERT(node>=kMinValueLead);
+            if(node&kValueIsFinal) {
+                // Leave the final value for getValue() to read.
+                result=USTRINGTRIE_FINAL_VALUE;
+            } else {
+                // Use the non-final value as the jump delta.
+                ++pos;
+                // int32_t delta=readValue(pos, node>>1);
+                node>>=1;
+                int32_t delta;
+                if(node<kMinTwoByteValueLead) {
+                    delta=node-kMinOneByteValueLead;
+                } else if(node<kMinThreeByteValueLead) {
+                    delta=((node-kMinTwoByteValueLead)<<8)|*pos++;
+                } else if(node<kFourByteValueLead) {
+                    delta=((node-kMinThreeByteValueLead)<<16)|(pos[0]<<8)|pos[1];
+                    pos+=2;
+                } else if(node==kFourByteValueLead) {
+                    delta=(pos[0]<<16)|(pos[1]<<8)|pos[2];
+                    pos+=3;
+                } else {
+                    delta=(pos[0]<<24)|(pos[1]<<16)|(pos[2]<<8)|pos[3];
+                    pos+=4;
+                }
+                // end readValue()
+                pos+=delta;
+                node=*pos;
+                result= node>=kMinValueLead ? valueResult(node) : USTRINGTRIE_NO_VALUE;
+            }
+            pos_=pos;
+            return result;
+        }
+        --length;
+        pos=skipValue(pos);
+    } while(length>1);
+    if(inByte==*pos++) {
+        pos_=pos;
+        int32_t node=*pos;
+        return node>=kMinValueLead ? valueResult(node) : USTRINGTRIE_NO_VALUE;
+    } else {
+        stop();
+        return USTRINGTRIE_NO_MATCH;
+    }
+}
+
+UStringTrieResult
+BytesTrie::nextImpl(const uint8_t *pos, int32_t inByte) {
+    for(;;) {
+        int32_t node=*pos++;
+        if(node<kMinLinearMatch) {
+            return branchNext(pos, node, inByte);
+        } else if(node<kMinValueLead) {
+            // Match the first of length+1 bytes.
+            int32_t length=node-kMinLinearMatch;  // Actual match length minus 1.
+            if(inByte==*pos++) {
+                remainingMatchLength_=--length;
+                pos_=pos;
+                return (length<0 && (node=*pos)>=kMinValueLead) ?
+                        valueResult(node) : USTRINGTRIE_NO_VALUE;
+            } else {
+                // No match.
+                break;
+            }
+        } else if(node&kValueIsFinal) {
+            // No further matching bytes.
+            break;
+        } else {
+            // Skip intermediate value.
+            pos=skipValue(pos, node);
+            // The next node must not also be a value node.
+            U_ASSERT(*pos<kMinValueLead);
+        }
+    }
+    stop();
+    return USTRINGTRIE_NO_MATCH;
+}
+
+UStringTrieResult
+BytesTrie::next(int32_t inByte) {
+    const uint8_t *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    }
+    if(inByte<0) {
+        inByte+=0x100;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    if(length>=0) {
+        // Remaining part of a linear-match node.
+        if(inByte==*pos++) {
+            remainingMatchLength_=--length;
+            pos_=pos;
+            int32_t node;
+            return (length<0 && (node=*pos)>=kMinValueLead) ?
+                    valueResult(node) : USTRINGTRIE_NO_VALUE;
+        } else {
+            stop();
+            return USTRINGTRIE_NO_MATCH;
+        }
+    }
+    return nextImpl(pos, inByte);
+}
+
+UStringTrieResult
+BytesTrie::next(const char *s, int32_t sLength) {
+    if(sLength<0 ? *s==0 : sLength==0) {
+        // Empty input.
+        return current();
+    }
+    const uint8_t *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    for(;;) {
+        // Fetch the next input byte, if there is one.
+        // Continue a linear-match node without rechecking sLength<0.
+        int32_t inByte;
+        if(sLength<0) {
+            for(;;) {
+                if((inByte=*s++)==0) {
+                    remainingMatchLength_=length;
+                    pos_=pos;
+                    int32_t node;
+                    return (length<0 && (node=*pos)>=kMinValueLead) ?
+                            valueResult(node) : USTRINGTRIE_NO_VALUE;
+                }
+                if(length<0) {
+                    remainingMatchLength_=length;
+                    break;
+                }
+                if(inByte!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+            }
+        } else {
+            for(;;) {
+                if(sLength==0) {
+                    remainingMatchLength_=length;
+                    pos_=pos;
+                    int32_t node;
+                    return (length<0 && (node=*pos)>=kMinValueLead) ?
+                            valueResult(node) : USTRINGTRIE_NO_VALUE;
+                }
+                inByte=*s++;
+                --sLength;
+                if(length<0) {
+                    remainingMatchLength_=length;
+                    break;
+                }
+                if(inByte!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+            }
+        }
+        for(;;) {
+            int32_t node=*pos++;
+            if(node<kMinLinearMatch) {
+                UStringTrieResult result=branchNext(pos, node, inByte);
+                if(result==USTRINGTRIE_NO_MATCH) {
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                // Fetch the next input byte, if there is one.
+                if(sLength<0) {
+                    if((inByte=*s++)==0) {
+                        return result;
+                    }
+                } else {
+                    if(sLength==0) {
+                        return result;
+                    }
+                    inByte=*s++;
+                    --sLength;
+                }
+                if(result==USTRINGTRIE_FINAL_VALUE) {
+                    // No further matching bytes.
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                pos=pos_;  // branchNext() advanced pos and wrote it to pos_ .
+            } else if(node<kMinValueLead) {
+                // Match length+1 bytes.
+                length=node-kMinLinearMatch;  // Actual match length minus 1.
+                if(inByte!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+                break;
+            } else if(node&kValueIsFinal) {
+                // No further matching bytes.
+                stop();
+                return USTRINGTRIE_NO_MATCH;
+            } else {
+                // Skip intermediate value.
+                pos=skipValue(pos, node);
+                // The next node must not also be a value node.
+                U_ASSERT(*pos<kMinValueLead);
+            }
+        }
+    }
+}
+
+const uint8_t *
+BytesTrie::findUniqueValueFromBranch(const uint8_t *pos, int32_t length,
+                                     UBool haveUniqueValue, int32_t &uniqueValue) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison byte
+        if(NULL==findUniqueValueFromBranch(jumpByDelta(pos), length>>1, haveUniqueValue, uniqueValue)) {
+            return NULL;
+        }
+        length=length-(length>>1);
+        pos=skipDelta(pos);
+    }
+    do {
+        ++pos;  // ignore a comparison byte
+        // handle its value
+        int32_t node=*pos++;
+        UBool isFinal=(UBool)(node&kValueIsFinal);
+        int32_t value=readValue(pos, node>>1);
+        pos=skipValue(pos, node);
+        if(isFinal) {
+            if(haveUniqueValue) {
+                if(value!=uniqueValue) {
+                    return NULL;
+                }
+            } else {
+                uniqueValue=value;
+                haveUniqueValue=TRUE;
+            }
+        } else {
+            if(!findUniqueValue(pos+value, haveUniqueValue, uniqueValue)) {
+                return NULL;
+            }
+            haveUniqueValue=TRUE;
+        }
+    } while(--length>1);
+    return pos+1;  // ignore the last comparison byte
+}
+
+UBool
+BytesTrie::findUniqueValue(const uint8_t *pos, UBool haveUniqueValue, int32_t &uniqueValue) {
+    for(;;) {
+        int32_t node=*pos++;
+        if(node<kMinLinearMatch) {
+            if(node==0) {
+                node=*pos++;
+            }
+            pos=findUniqueValueFromBranch(pos, node+1, haveUniqueValue, uniqueValue);
+            if(pos==NULL) {
+                return FALSE;
+            }
+            haveUniqueValue=TRUE;
+        } else if(node<kMinValueLead) {
+            // linear-match node
+            pos+=node-kMinLinearMatch+1;  // Ignore the match bytes.
+        } else {
+            UBool isFinal=(UBool)(node&kValueIsFinal);
+            int32_t value=readValue(pos, node>>1);
+            if(haveUniqueValue) {
+                if(value!=uniqueValue) {
+                    return FALSE;
+                }
+            } else {
+                uniqueValue=value;
+                haveUniqueValue=TRUE;
+            }
+            if(isFinal) {
+                return TRUE;
+            }
+            pos=skipValue(pos, node);
+        }
+    }
+}
+
+int32_t
+BytesTrie::getNextBytes(ByteSink &out) const {
+    const uint8_t *pos=pos_;
+    if(pos==NULL) {
+        return 0;
+    }
+    if(remainingMatchLength_>=0) {
+        append(out, *pos);  // Next byte of a pending linear-match node.
+        return 1;
+    }
+    int32_t node=*pos++;
+    if(node>=kMinValueLead) {
+        if(node&kValueIsFinal) {
+            return 0;
+        } else {
+            pos=skipValue(pos, node);
+            node=*pos++;
+            U_ASSERT(node<kMinValueLead);
+        }
+    }
+    if(node<kMinLinearMatch) {
+        if(node==0) {
+            node=*pos++;
+        }
+        getNextBranchBytes(pos, ++node, out);
+        return node;
+    } else {
+        // First byte of the linear-match node.
+        append(out, *pos);
+        return 1;
+    }
+}
+
+void
+BytesTrie::getNextBranchBytes(const uint8_t *pos, int32_t length, ByteSink &out) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison byte
+        getNextBranchBytes(jumpByDelta(pos), length>>1, out);
+        length=length-(length>>1);
+        pos=skipDelta(pos);
+    }
+    do {
+        append(out, *pos++);
+        pos=skipValue(pos);
+    } while(--length>1);
+    append(out, *pos);
+}
+
+void
+BytesTrie::append(ByteSink &out, int c) {
+    char ch=(char)c;
+    out.Append(&ch, 1);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bytestriebuilder.cpp b/thirdparty/icu4c/common/bytestriebuilder.cpp
new file mode 100644
index 0000000000..ec1ab7d8f5
--- /dev/null
+++ b/thirdparty/icu4c/common/bytestriebuilder.cpp
@@ -0,0 +1,504 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  bytestriebuilder.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010sep25
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/bytestriebuilder.h"
+#include "unicode/stringpiece.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "uhash.h"
+#include "uarrsort.h"
+#include "uassert.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Note: This builder implementation stores (bytes, value) pairs with full copies
+ * of the byte sequences, until the BytesTrie is built.
+ * It might(!) take less memory if we collected the data in a temporary, dynamic trie.
+ */
+
+class BytesTrieElement : public UMemory {
+public:
+    // Use compiler's default constructor, initializes nothing.
+
+    void setTo(StringPiece s, int32_t val, CharString &strings, UErrorCode &errorCode);
+
+    StringPiece getString(const CharString &strings) const {
+        int32_t offset=stringOffset;
+        int32_t length;
+        if(offset>=0) {
+            length=(uint8_t)strings[offset++];
+        } else {
+            offset=~offset;
+            length=((int32_t)(uint8_t)strings[offset]<<8)|(uint8_t)strings[offset+1];
+            offset+=2;
+        }
+        return StringPiece(strings.data()+offset, length);
+    }
+    int32_t getStringLength(const CharString &strings) const {
+        int32_t offset=stringOffset;
+        if(offset>=0) {
+            return (uint8_t)strings[offset];
+        } else {
+            offset=~offset;
+            return ((int32_t)(uint8_t)strings[offset]<<8)|(uint8_t)strings[offset+1];
+        }
+    }
+
+    char charAt(int32_t index, const CharString &strings) const { return data(strings)[index]; }
+
+    int32_t getValue() const { return value; }
+
+    int32_t compareStringTo(const BytesTrieElement &o, const CharString &strings) const;
+
+private:
+    const char *data(const CharString &strings) const {
+        int32_t offset=stringOffset;
+        if(offset>=0) {
+            ++offset;
+        } else {
+            offset=~offset+2;
+        }
+        return strings.data()+offset;
+    }
+
+    // If the stringOffset is non-negative, then the first strings byte contains
+    // the string length.
+    // If the stringOffset is negative, then the first two strings bytes contain
+    // the string length (big-endian), and the offset needs to be bit-inverted.
+    // (Compared with a stringLength field here, this saves 3 bytes per string for most strings.)
+    int32_t stringOffset;
+    int32_t value;
+};
+
+void
+BytesTrieElement::setTo(StringPiece s, int32_t val,
+                        CharString &strings, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    int32_t length=s.length();
+    if(length>0xffff) {
+        // Too long: We store the length in 1 or 2 bytes.
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+    int32_t offset=strings.length();
+    if(length>0xff) {
+        offset=~offset;
+        strings.append((char)(length>>8), errorCode);
+    }
+    strings.append((char)length, errorCode);
+    stringOffset=offset;
+    value=val;
+    strings.append(s, errorCode);
+}
+
+int32_t
+BytesTrieElement::compareStringTo(const BytesTrieElement &other, const CharString &strings) const {
+    // TODO: add StringPiece::compare(), see ticket #8187
+    StringPiece thisString=getString(strings);
+    StringPiece otherString=other.getString(strings);
+    int32_t lengthDiff=thisString.length()-otherString.length();
+    int32_t commonLength;
+    if(lengthDiff<=0) {
+        commonLength=thisString.length();
+    } else {
+        commonLength=otherString.length();
+    }
+    int32_t diff=uprv_memcmp(thisString.data(), otherString.data(), commonLength);
+    return diff!=0 ? diff : lengthDiff;
+}
+
+BytesTrieBuilder::BytesTrieBuilder(UErrorCode &errorCode)
+        : strings(NULL), elements(NULL), elementsCapacity(0), elementsLength(0),
+          bytes(NULL), bytesCapacity(0), bytesLength(0) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    strings=new CharString();
+    if(strings==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+BytesTrieBuilder::~BytesTrieBuilder() {
+    delete strings;
+    delete[] elements;
+    uprv_free(bytes);
+}
+
+BytesTrieBuilder &
+BytesTrieBuilder::add(StringPiece s, int32_t value, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return *this;
+    }
+    if(bytesLength>0) {
+        // Cannot add elements after building.
+        errorCode=U_NO_WRITE_PERMISSION;
+        return *this;
+    }
+    if(elementsLength==elementsCapacity) {
+        int32_t newCapacity;
+        if(elementsCapacity==0) {
+            newCapacity=1024;
+        } else {
+            newCapacity=4*elementsCapacity;
+        }
+        BytesTrieElement *newElements=new BytesTrieElement[newCapacity];
+        if(newElements==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return *this; // error instead of dereferencing null
+        }
+        if(elementsLength>0) {
+            uprv_memcpy(newElements, elements, (size_t)elementsLength*sizeof(BytesTrieElement));
+        }
+        delete[] elements;
+        elements=newElements;
+        elementsCapacity=newCapacity;
+    }
+    elements[elementsLength++].setTo(s, value, *strings, errorCode);
+    return *this;
+}
+
+U_CDECL_BEGIN
+
+static int32_t U_CALLCONV
+compareElementStrings(const void *context, const void *left, const void *right) {
+    const CharString *strings=static_cast<const CharString *>(context);
+    const BytesTrieElement *leftElement=static_cast<const BytesTrieElement *>(left);
+    const BytesTrieElement *rightElement=static_cast<const BytesTrieElement *>(right);
+    return leftElement->compareStringTo(*rightElement, *strings);
+}
+
+U_CDECL_END
+
+BytesTrie *
+BytesTrieBuilder::build(UStringTrieBuildOption buildOption, UErrorCode &errorCode) {
+    buildBytes(buildOption, errorCode);
+    BytesTrie *newTrie=NULL;
+    if(U_SUCCESS(errorCode)) {
+        newTrie=new BytesTrie(bytes, bytes+(bytesCapacity-bytesLength));
+        if(newTrie==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            bytes=NULL;  // The new trie now owns the array.
+            bytesCapacity=0;
+        }
+    }
+    return newTrie;
+}
+
+StringPiece
+BytesTrieBuilder::buildStringPiece(UStringTrieBuildOption buildOption, UErrorCode &errorCode) {
+    buildBytes(buildOption, errorCode);
+    StringPiece result;
+    if(U_SUCCESS(errorCode)) {
+        result.set(bytes+(bytesCapacity-bytesLength), bytesLength);
+    }
+    return result;
+}
+
+void
+BytesTrieBuilder::buildBytes(UStringTrieBuildOption buildOption, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    if(bytes!=NULL && bytesLength>0) {
+        // Already built.
+        return;
+    }
+    if(bytesLength==0) {
+        if(elementsLength==0) {
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return;
+        }
+        uprv_sortArray(elements, elementsLength, (int32_t)sizeof(BytesTrieElement),
+                      compareElementStrings, strings,
+                      FALSE,  // need not be a stable sort
+                      &errorCode);
+        if(U_FAILURE(errorCode)) {
+            return;
+        }
+        // Duplicate strings are not allowed.
+        StringPiece prev=elements[0].getString(*strings);
+        for(int32_t i=1; i<elementsLength; ++i) {
+            StringPiece current=elements[i].getString(*strings);
+            if(prev==current) {
+                errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+                return;
+            }
+            prev=current;
+        }
+    }
+    // Create and byte-serialize the trie for the elements.
+    bytesLength=0;
+    int32_t capacity=strings->length();
+    if(capacity<1024) {
+        capacity=1024;
+    }
+    if(bytesCapacity<capacity) {
+        uprv_free(bytes);
+        bytes=static_cast<char *>(uprv_malloc(capacity));
+        if(bytes==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            bytesCapacity=0;
+            return;
+        }
+        bytesCapacity=capacity;
+    }
+    StringTrieBuilder::build(buildOption, elementsLength, errorCode);
+    if(bytes==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+BytesTrieBuilder &
+BytesTrieBuilder::clear() {
+    strings->clear();
+    elementsLength=0;
+    bytesLength=0;
+    return *this;
+}
+
+int32_t
+BytesTrieBuilder::getElementStringLength(int32_t i) const {
+    return elements[i].getStringLength(*strings);
+}
+
+UChar
+BytesTrieBuilder::getElementUnit(int32_t i, int32_t byteIndex) const {
+    return (uint8_t)elements[i].charAt(byteIndex, *strings);
+}
+
+int32_t
+BytesTrieBuilder::getElementValue(int32_t i) const {
+    return elements[i].getValue();
+}
+
+int32_t
+BytesTrieBuilder::getLimitOfLinearMatch(int32_t first, int32_t last, int32_t byteIndex) const {
+    const BytesTrieElement &firstElement=elements[first];
+    const BytesTrieElement &lastElement=elements[last];
+    int32_t minStringLength=firstElement.getStringLength(*strings);
+    while(++byteIndex<minStringLength &&
+            firstElement.charAt(byteIndex, *strings)==
+            lastElement.charAt(byteIndex, *strings)) {}
+    return byteIndex;
+}
+
+int32_t
+BytesTrieBuilder::countElementUnits(int32_t start, int32_t limit, int32_t byteIndex) const {
+    int32_t length=0;  // Number of different bytes at byteIndex.
+    int32_t i=start;
+    do {
+        char byte=elements[i++].charAt(byteIndex, *strings);
+        while(i<limit && byte==elements[i].charAt(byteIndex, *strings)) {
+            ++i;
+        }
+        ++length;
+    } while(i<limit);
+    return length;
+}
+
+int32_t
+BytesTrieBuilder::skipElementsBySomeUnits(int32_t i, int32_t byteIndex, int32_t count) const {
+    do {
+        char byte=elements[i++].charAt(byteIndex, *strings);
+        while(byte==elements[i].charAt(byteIndex, *strings)) {
+            ++i;
+        }
+    } while(--count>0);
+    return i;
+}
+
+int32_t
+BytesTrieBuilder::indexOfElementWithNextUnit(int32_t i, int32_t byteIndex, UChar byte) const {
+    char b=(char)byte;
+    while(b==elements[i].charAt(byteIndex, *strings)) {
+        ++i;
+    }
+    return i;
+}
+
+BytesTrieBuilder::BTLinearMatchNode::BTLinearMatchNode(const char *bytes, int32_t len, Node *nextNode)
+        : LinearMatchNode(len, nextNode), s(bytes) {
+    hash=static_cast<int32_t>(
+        static_cast<uint32_t>(hash)*37u + static_cast<uint32_t>(ustr_hashCharsN(bytes, len)));
+}
+
+UBool
+BytesTrieBuilder::BTLinearMatchNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!LinearMatchNode::operator==(other)) {
+        return FALSE;
+    }
+    const BTLinearMatchNode &o=(const BTLinearMatchNode &)other;
+    return 0==uprv_memcmp(s, o.s, length);
+}
+
+void
+BytesTrieBuilder::BTLinearMatchNode::write(StringTrieBuilder &builder) {
+    BytesTrieBuilder &b=(BytesTrieBuilder &)builder;
+    next->write(builder);
+    b.write(s, length);
+    offset=b.write(b.getMinLinearMatch()+length-1);
+}
+
+StringTrieBuilder::Node *
+BytesTrieBuilder::createLinearMatchNode(int32_t i, int32_t byteIndex, int32_t length,
+                                        Node *nextNode) const {
+    return new BTLinearMatchNode(
+            elements[i].getString(*strings).data()+byteIndex,
+            length,
+            nextNode);
+}
+
+UBool
+BytesTrieBuilder::ensureCapacity(int32_t length) {
+    if(bytes==NULL) {
+        return FALSE;  // previous memory allocation had failed
+    }
+    if(length>bytesCapacity) {
+        int32_t newCapacity=bytesCapacity;
+        do {
+            newCapacity*=2;
+        } while(newCapacity<=length);
+        char *newBytes=static_cast<char *>(uprv_malloc(newCapacity));
+        if(newBytes==NULL) {
+            // unable to allocate memory
+            uprv_free(bytes);
+            bytes=NULL;
+            bytesCapacity=0;
+            return FALSE;
+        }
+        uprv_memcpy(newBytes+(newCapacity-bytesLength),
+                    bytes+(bytesCapacity-bytesLength), bytesLength);
+        uprv_free(bytes);
+        bytes=newBytes;
+        bytesCapacity=newCapacity;
+    }
+    return TRUE;
+}
+
+int32_t
+BytesTrieBuilder::write(int32_t byte) {
+    int32_t newLength=bytesLength+1;
+    if(ensureCapacity(newLength)) {
+        bytesLength=newLength;
+        bytes[bytesCapacity-bytesLength]=(char)byte;
+    }
+    return bytesLength;
+}
+
+int32_t
+BytesTrieBuilder::write(const char *b, int32_t length) {
+    int32_t newLength=bytesLength+length;
+    if(ensureCapacity(newLength)) {
+        bytesLength=newLength;
+        uprv_memcpy(bytes+(bytesCapacity-bytesLength), b, length);
+    }
+    return bytesLength;
+}
+
+int32_t
+BytesTrieBuilder::writeElementUnits(int32_t i, int32_t byteIndex, int32_t length) {
+    return write(elements[i].getString(*strings).data()+byteIndex, length);
+}
+
+int32_t
+BytesTrieBuilder::writeValueAndFinal(int32_t i, UBool isFinal) {
+    if(0<=i && i<=BytesTrie::kMaxOneByteValue) {
+        return write(((BytesTrie::kMinOneByteValueLead+i)<<1)|isFinal);
+    }
+    char intBytes[5];
+    int32_t length=1;
+    if(i<0 || i>0xffffff) {
+        intBytes[0]=(char)BytesTrie::kFiveByteValueLead;
+        intBytes[1]=(char)((uint32_t)i>>24);
+        intBytes[2]=(char)((uint32_t)i>>16);
+        intBytes[3]=(char)((uint32_t)i>>8);
+        intBytes[4]=(char)i;
+        length=5;
+    // } else if(i<=BytesTrie::kMaxOneByteValue) {
+    //     intBytes[0]=(char)(BytesTrie::kMinOneByteValueLead+i);
+    } else {
+        if(i<=BytesTrie::kMaxTwoByteValue) {
+            intBytes[0]=(char)(BytesTrie::kMinTwoByteValueLead+(i>>8));
+        } else {
+            if(i<=BytesTrie::kMaxThreeByteValue) {
+                intBytes[0]=(char)(BytesTrie::kMinThreeByteValueLead+(i>>16));
+            } else {
+                intBytes[0]=(char)BytesTrie::kFourByteValueLead;
+                intBytes[1]=(char)(i>>16);
+                length=2;
+            }
+            intBytes[length++]=(char)(i>>8);
+        }
+        intBytes[length++]=(char)i;
+    }
+    intBytes[0]=(char)((intBytes[0]<<1)|isFinal);
+    return write(intBytes, length);
+}
+
+int32_t
+BytesTrieBuilder::writeValueAndType(UBool hasValue, int32_t value, int32_t node) {
+    int32_t offset=write(node);
+    if(hasValue) {
+        offset=writeValueAndFinal(value, FALSE);
+    }
+    return offset;
+}
+
+int32_t
+BytesTrieBuilder::writeDeltaTo(int32_t jumpTarget) {
+    int32_t i=bytesLength-jumpTarget;
+    U_ASSERT(i>=0);
+    if(i<=BytesTrie::kMaxOneByteDelta) {
+        return write(i);
+    }
+    char intBytes[5];
+    int32_t length;
+    if(i<=BytesTrie::kMaxTwoByteDelta) {
+        intBytes[0]=(char)(BytesTrie::kMinTwoByteDeltaLead+(i>>8));
+        length=1;
+    } else {
+        if(i<=BytesTrie::kMaxThreeByteDelta) {
+            intBytes[0]=(char)(BytesTrie::kMinThreeByteDeltaLead+(i>>16));
+            length=2;
+        } else {
+            if(i<=0xffffff) {
+                intBytes[0]=(char)BytesTrie::kFourByteDeltaLead;
+                length=3;
+            } else {
+                intBytes[0]=(char)BytesTrie::kFiveByteDeltaLead;
+                intBytes[1]=(char)(i>>24);
+                length=4;
+            }
+            intBytes[1]=(char)(i>>16);
+        }
+        intBytes[1]=(char)(i>>8);
+    }
+    intBytes[length++]=(char)i;
+    return write(intBytes, length);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/bytestrieiterator.cpp b/thirdparty/icu4c/common/bytestrieiterator.cpp
new file mode 100644
index 0000000000..e64961a1f1
--- /dev/null
+++ b/thirdparty/icu4c/common/bytestrieiterator.cpp
@@ -0,0 +1,214 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  bytestrieiterator.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov03
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/stringpiece.h"
+#include "charstr.h"
+#include "uvectr32.h"
+
+U_NAMESPACE_BEGIN
+
+BytesTrie::Iterator::Iterator(const void *trieBytes, int32_t maxStringLength,
+                              UErrorCode &errorCode)
+        : bytes_(static_cast<const uint8_t *>(trieBytes)),
+          pos_(bytes_), initialPos_(bytes_),
+          remainingMatchLength_(-1), initialRemainingMatchLength_(-1),
+          str_(NULL), maxLength_(maxStringLength), value_(0), stack_(NULL) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    // str_ and stack_ are pointers so that it's easy to turn bytestrie.h into
+    // a public API header for which we would want it to depend only on
+    // other public headers.
+    // Unlike BytesTrie itself, its Iterator performs memory allocations anyway
+    // via the CharString and UVector32 implementations, so this additional
+    // cost is minimal.
+    str_=new CharString();
+    stack_=new UVector32(errorCode);
+    if(U_SUCCESS(errorCode) && (str_==NULL || stack_==NULL)) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+BytesTrie::Iterator::Iterator(const BytesTrie &trie, int32_t maxStringLength,
+                              UErrorCode &errorCode)
+        : bytes_(trie.bytes_), pos_(trie.pos_), initialPos_(trie.pos_),
+          remainingMatchLength_(trie.remainingMatchLength_),
+          initialRemainingMatchLength_(trie.remainingMatchLength_),
+          str_(NULL), maxLength_(maxStringLength), value_(0), stack_(NULL) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    str_=new CharString();
+    stack_=new UVector32(errorCode);
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    if(str_==NULL || stack_==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    if(length>=0) {
+        // Pending linear-match node, append remaining bytes to str_.
+        ++length;
+        if(maxLength_>0 && length>maxLength_) {
+            length=maxLength_;  // This will leave remainingMatchLength>=0 as a signal.
+        }
+        str_->append(reinterpret_cast<const char *>(pos_), length, errorCode);
+        pos_+=length;
+        remainingMatchLength_-=length;
+    }
+}
+
+BytesTrie::Iterator::~Iterator() {
+    delete str_;
+    delete stack_;
+}
+
+BytesTrie::Iterator &
+BytesTrie::Iterator::reset() {
+    pos_=initialPos_;
+    remainingMatchLength_=initialRemainingMatchLength_;
+    int32_t length=remainingMatchLength_+1;  // Remaining match length.
+    if(maxLength_>0 && length>maxLength_) {
+        length=maxLength_;
+    }
+    str_->truncate(length);
+    pos_+=length;
+    remainingMatchLength_-=length;
+    stack_->setSize(0);
+    return *this;
+}
+
+UBool
+BytesTrie::Iterator::hasNext() const { return pos_!=NULL || !stack_->isEmpty(); }
+
+UBool
+BytesTrie::Iterator::next(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    const uint8_t *pos=pos_;
+    if(pos==NULL) {
+        if(stack_->isEmpty()) {
+            return FALSE;
+        }
+        // Pop the state off the stack and continue with the next outbound edge of
+        // the branch node.
+        int32_t stackSize=stack_->size();
+        int32_t length=stack_->elementAti(stackSize-1);
+        pos=bytes_+stack_->elementAti(stackSize-2);
+        stack_->setSize(stackSize-2);
+        str_->truncate(length&0xffff);
+        length=(int32_t)((uint32_t)length>>16);
+        if(length>1) {
+            pos=branchNext(pos, length, errorCode);
+            if(pos==NULL) {
+                return TRUE;  // Reached a final value.
+            }
+        } else {
+            str_->append((char)*pos++, errorCode);
+        }
+    }
+    if(remainingMatchLength_>=0) {
+        // We only get here if we started in a pending linear-match node
+        // with more than maxLength remaining bytes.
+        return truncateAndStop();
+    }
+    for(;;) {
+        int32_t node=*pos++;
+        if(node>=kMinValueLead) {
+            // Deliver value for the byte sequence so far.
+            UBool isFinal=(UBool)(node&kValueIsFinal);
+            value_=readValue(pos, node>>1);
+            if(isFinal || (maxLength_>0 && str_->length()==maxLength_)) {
+                pos_=NULL;
+            } else {
+                pos_=skipValue(pos, node);
+            }
+            return TRUE;
+        }
+        if(maxLength_>0 && str_->length()==maxLength_) {
+            return truncateAndStop();
+        }
+        if(node<kMinLinearMatch) {
+            if(node==0) {
+                node=*pos++;
+            }
+            pos=branchNext(pos, node+1, errorCode);
+            if(pos==NULL) {
+                return TRUE;  // Reached a final value.
+            }
+        } else {
+            // Linear-match node, append length bytes to str_.
+            int32_t length=node-kMinLinearMatch+1;
+            if(maxLength_>0 && str_->length()+length>maxLength_) {
+                str_->append(reinterpret_cast<const char *>(pos),
+                            maxLength_-str_->length(), errorCode);
+                return truncateAndStop();
+            }
+            str_->append(reinterpret_cast<const char *>(pos), length, errorCode);
+            pos+=length;
+        }
+    }
+}
+
+StringPiece
+BytesTrie::Iterator::getString() const {
+    return str_ == NULL ? StringPiece() : str_->toStringPiece();
+}
+
+UBool
+BytesTrie::Iterator::truncateAndStop() {
+    pos_=NULL;
+    value_=-1;  // no real value for str
+    return TRUE;
+}
+
+// Branch node, needs to take the first outbound edge and push state for the rest.
+const uint8_t *
+BytesTrie::Iterator::branchNext(const uint8_t *pos, int32_t length, UErrorCode &errorCode) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison byte
+        // Push state for the greater-or-equal edge.
+        stack_->addElement((int32_t)(skipDelta(pos)-bytes_), errorCode);
+        stack_->addElement(((length-(length>>1))<<16)|str_->length(), errorCode);
+        // Follow the less-than edge.
+        length>>=1;
+        pos=jumpByDelta(pos);
+    }
+    // List of key-value pairs where values are either final values or jump deltas.
+    // Read the first (key, value) pair.
+    uint8_t trieByte=*pos++;
+    int32_t node=*pos++;
+    UBool isFinal=(UBool)(node&kValueIsFinal);
+    int32_t value=readValue(pos, node>>1);
+    pos=skipValue(pos, node);
+    stack_->addElement((int32_t)(pos-bytes_), errorCode);
+    stack_->addElement(((length-1)<<16)|str_->length(), errorCode);
+    str_->append((char)trieByte, errorCode);
+    if(isFinal) {
+        pos_=NULL;
+        value_=value;
+        return NULL;
+    } else {
+        return pos+value;
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/caniter.cpp b/thirdparty/icu4c/common/caniter.cpp
new file mode 100644
index 0000000000..b28acfc84e
--- /dev/null
+++ b/thirdparty/icu4c/common/caniter.cpp
@@ -0,0 +1,586 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *****************************************************************************
+ * Copyright (C) 1996-2015, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ *****************************************************************************
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/caniter.h"
+#include "unicode/normalizer2.h"
+#include "unicode/uchar.h"
+#include "unicode/uniset.h"
+#include "unicode/usetiter.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "hash.h"
+#include "normalizer2impl.h"
+
+/**
+ * This class allows one to iterate through all the strings that are canonically equivalent to a given
+ * string. For example, here are some sample results:
+Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+1: \u0041\u030A\u0064\u0307\u0327
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+2: \u0041\u030A\u0064\u0327\u0307
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+3: \u0041\u030A\u1E0B\u0327
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+4: \u0041\u030A\u1E11\u0307
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+5: \u00C5\u0064\u0307\u0327
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+6: \u00C5\u0064\u0327\u0307
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+7: \u00C5\u1E0B\u0327
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+8: \u00C5\u1E11\u0307
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+9: \u212B\u0064\u0307\u0327
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+10: \u212B\u0064\u0327\u0307
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+11: \u212B\u1E0B\u0327
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+12: \u212B\u1E11\u0307
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+ *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
+ * since it has not been optimized for that situation.
+ *@author M. Davis
+ *@draft
+ */
+
+// public
+
+U_NAMESPACE_BEGIN
+
+// TODO: add boilerplate methods.
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(CanonicalIterator)
+
+/**
+ *@param source string to get results for
+ */
+CanonicalIterator::CanonicalIterator(const UnicodeString &sourceStr, UErrorCode &status) :
+    pieces(NULL),
+    pieces_length(0),
+    pieces_lengths(NULL),
+    current(NULL),
+    current_length(0),
+    nfd(*Normalizer2::getNFDInstance(status)),
+    nfcImpl(*Normalizer2Factory::getNFCImpl(status))
+{
+    if(U_SUCCESS(status) && nfcImpl.ensureCanonIterData(status)) {
+      setSource(sourceStr, status);
+    }
+}
+
+CanonicalIterator::~CanonicalIterator() {
+  cleanPieces();
+}
+
+void CanonicalIterator::cleanPieces() {
+    int32_t i = 0;
+    if(pieces != NULL) {
+        for(i = 0; i < pieces_length; i++) {
+            if(pieces[i] != NULL) {
+                delete[] pieces[i];
+            }
+        }
+        uprv_free(pieces);
+        pieces = NULL;
+        pieces_length = 0;
+    }
+    if(pieces_lengths != NULL) {
+        uprv_free(pieces_lengths);
+        pieces_lengths = NULL;
+    }
+    if(current != NULL) {
+        uprv_free(current);
+        current = NULL;
+        current_length = 0;
+    }
+}
+
+/**
+ *@return gets the source: NOTE: it is the NFD form of source
+ */
+UnicodeString CanonicalIterator::getSource() {
+  return source;
+}
+
+/**
+ * Resets the iterator so that one can start again from the beginning.
+ */
+void CanonicalIterator::reset() {
+    done = FALSE;
+    for (int i = 0; i < current_length; ++i) {
+        current[i] = 0;
+    }
+}
+
+/**
+ *@return the next string that is canonically equivalent. The value null is returned when
+ * the iteration is done.
+ */
+UnicodeString CanonicalIterator::next() {
+    int32_t i = 0;
+
+    if (done) {
+      buffer.setToBogus();
+      return buffer;
+    }
+
+    // delete old contents
+    buffer.remove();
+
+    // construct return value
+
+    for (i = 0; i < pieces_length; ++i) {
+        buffer.append(pieces[i][current[i]]);
+    }
+    //String result = buffer.toString(); // not needed
+
+    // find next value for next time
+
+    for (i = current_length - 1; ; --i) {
+        if (i < 0) {
+            done = TRUE;
+            break;
+        }
+        current[i]++;
+        if (current[i] < pieces_lengths[i]) break; // got sequence
+        current[i] = 0;
+    }
+    return buffer;
+}
+
+/**
+ *@param set the source string to iterate against. This allows the same iterator to be used
+ * while changing the source string, saving object creation.
+ */
+void CanonicalIterator::setSource(const UnicodeString &newSource, UErrorCode &status) {
+    int32_t list_length = 0;
+    UChar32 cp = 0;
+    int32_t start = 0;
+    int32_t i = 0;
+    UnicodeString *list = NULL;
+
+    nfd.normalize(newSource, source, status);
+    if(U_FAILURE(status)) {
+      return;
+    }
+    done = FALSE;
+
+    cleanPieces();
+
+    // catch degenerate case
+    if (newSource.length() == 0) {
+        pieces = (UnicodeString **)uprv_malloc(sizeof(UnicodeString *));
+        pieces_lengths = (int32_t*)uprv_malloc(1 * sizeof(int32_t));
+        pieces_length = 1;
+        current = (int32_t*)uprv_malloc(1 * sizeof(int32_t));
+        current_length = 1;
+        if (pieces == NULL || pieces_lengths == NULL || current == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            goto CleanPartialInitialization;
+        }
+        current[0] = 0;
+        pieces[0] = new UnicodeString[1];
+        pieces_lengths[0] = 1;
+        if (pieces[0] == 0) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            goto CleanPartialInitialization;
+        }
+        return;
+    }
+
+
+    list = new UnicodeString[source.length()];
+    if (list == 0) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        goto CleanPartialInitialization;
+    }
+
+    // i should initialy be the number of code units at the 
+    // start of the string
+    i = U16_LENGTH(source.char32At(0));
+    //int32_t i = 1;
+    // find the segments
+    // This code iterates through the source string and 
+    // extracts segments that end up on a codepoint that
+    // doesn't start any decompositions. (Analysis is done
+    // on the NFD form - see above).
+    for (; i < source.length(); i += U16_LENGTH(cp)) {
+        cp = source.char32At(i);
+        if (nfcImpl.isCanonSegmentStarter(cp)) {
+            source.extract(start, i-start, list[list_length++]); // add up to i
+            start = i;
+        }
+    }
+    source.extract(start, i-start, list[list_length++]); // add last one
+
+
+    // allocate the arrays, and find the strings that are CE to each segment
+    pieces = (UnicodeString **)uprv_malloc(list_length * sizeof(UnicodeString *));
+    pieces_length = list_length;
+    pieces_lengths = (int32_t*)uprv_malloc(list_length * sizeof(int32_t));
+    current = (int32_t*)uprv_malloc(list_length * sizeof(int32_t));
+    current_length = list_length;
+    if (pieces == NULL || pieces_lengths == NULL || current == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        goto CleanPartialInitialization;
+    }
+
+    for (i = 0; i < current_length; i++) {
+        current[i] = 0;
+    }
+    // for each segment, get all the combinations that can produce 
+    // it after NFD normalization
+    for (i = 0; i < pieces_length; ++i) {
+        //if (PROGRESS) printf("SEGMENT\n");
+        pieces[i] = getEquivalents(list[i], pieces_lengths[i], status);
+    }
+
+    delete[] list;
+    return;
+// Common section to cleanup all local variables and reset object variables.
+CleanPartialInitialization:
+    if (list != NULL) {
+        delete[] list;
+    }
+    cleanPieces();
+}
+
+/**
+ * Dumb recursive implementation of permutation.
+ * TODO: optimize
+ * @param source the string to find permutations for
+ * @return the results in a set.
+ */
+void U_EXPORT2 CanonicalIterator::permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status) {
+    if(U_FAILURE(status)) {
+        return;
+    }
+    //if (PROGRESS) printf("Permute: %s\n", UToS(Tr(source)));
+    int32_t i = 0;
+
+    // optimization:
+    // if zero or one character, just return a set with it
+    // we check for length < 2 to keep from counting code points all the time
+    if (source.length() <= 2 && source.countChar32() <= 1) {
+        UnicodeString *toPut = new UnicodeString(source);
+        /* test for NULL */
+        if (toPut == 0) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        result->put(source, toPut, status);
+        return;
+    }
+
+    // otherwise iterate through the string, and recursively permute all the other characters
+    UChar32 cp;
+    Hashtable subpermute(status);
+    if(U_FAILURE(status)) {
+        return;
+    }
+    subpermute.setValueDeleter(uprv_deleteUObject);
+
+    for (i = 0; i < source.length(); i += U16_LENGTH(cp)) {
+        cp = source.char32At(i);
+        const UHashElement *ne = NULL;
+        int32_t el = UHASH_FIRST;
+        UnicodeString subPermuteString = source;
+
+        // optimization:
+        // if the character is canonical combining class zero,
+        // don't permute it
+        if (skipZeros && i != 0 && u_getCombiningClass(cp) == 0) {
+            //System.out.println("Skipping " + Utility.hex(UTF16.valueOf(source, i)));
+            continue;
+        }
+
+        subpermute.removeAll();
+
+        // see what the permutations of the characters before and after this one are
+        //Hashtable *subpermute = permute(source.substring(0,i) + source.substring(i + UTF16.getCharCount(cp)));
+        permute(subPermuteString.remove(i, U16_LENGTH(cp)), skipZeros, &subpermute, status);
+        /* Test for buffer overflows */
+        if(U_FAILURE(status)) {
+            return;
+        }
+        // The upper remove is destructive. The question is do we have to make a copy, or we don't care about the contents 
+        // of source at this point.
+
+        // prefix this character to all of them
+        ne = subpermute.nextElement(el);
+        while (ne != NULL) {
+            UnicodeString *permRes = (UnicodeString *)(ne->value.pointer);
+            UnicodeString *chStr = new UnicodeString(cp);
+            //test for  NULL
+            if (chStr == NULL) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            chStr->append(*permRes); //*((UnicodeString *)(ne->value.pointer));
+            //if (PROGRESS) printf("  Piece: %s\n", UToS(*chStr));
+            result->put(*chStr, chStr, status);
+            ne = subpermute.nextElement(el);
+        }
+    }
+    //return result;
+}
+
+// privates
+
+// we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
+UnicodeString* CanonicalIterator::getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status) {
+    Hashtable result(status);
+    Hashtable permutations(status);
+    Hashtable basic(status);
+    if (U_FAILURE(status)) {
+        return 0;
+    }
+    result.setValueDeleter(uprv_deleteUObject);
+    permutations.setValueDeleter(uprv_deleteUObject);
+    basic.setValueDeleter(uprv_deleteUObject);
+
+    UChar USeg[256];
+    int32_t segLen = segment.extract(USeg, 256, status);
+    getEquivalents2(&basic, USeg, segLen, status);
+
+    // now get all the permutations
+    // add only the ones that are canonically equivalent
+    // TODO: optimize by not permuting any class zero.
+
+    const UHashElement *ne = NULL;
+    int32_t el = UHASH_FIRST;
+    //Iterator it = basic.iterator();
+    ne = basic.nextElement(el);
+    //while (it.hasNext())
+    while (ne != NULL) {
+        //String item = (String) it.next();
+        UnicodeString item = *((UnicodeString *)(ne->value.pointer));
+
+        permutations.removeAll();
+        permute(item, CANITER_SKIP_ZEROES, &permutations, status);
+        const UHashElement *ne2 = NULL;
+        int32_t el2 = UHASH_FIRST;
+        //Iterator it2 = permutations.iterator();
+        ne2 = permutations.nextElement(el2);
+        //while (it2.hasNext())
+        while (ne2 != NULL) {
+            //String possible = (String) it2.next();
+            //UnicodeString *possible = new UnicodeString(*((UnicodeString *)(ne2->value.pointer)));
+            UnicodeString possible(*((UnicodeString *)(ne2->value.pointer)));
+            UnicodeString attempt;
+            nfd.normalize(possible, attempt, status);
+
+            // TODO: check if operator == is semanticaly the same as attempt.equals(segment)
+            if (attempt==segment) {
+                //if (PROGRESS) printf("Adding Permutation: %s\n", UToS(Tr(*possible)));
+                // TODO: use the hashtable just to catch duplicates - store strings directly (somehow).
+                result.put(possible, new UnicodeString(possible), status); //add(possible);
+            } else {
+                //if (PROGRESS) printf("-Skipping Permutation: %s\n", UToS(Tr(*possible)));
+            }
+
+            ne2 = permutations.nextElement(el2);
+        }
+        ne = basic.nextElement(el);
+    }
+
+    /* Test for buffer overflows */
+    if(U_FAILURE(status)) {
+        return 0;
+    }
+    // convert into a String[] to clean up storage
+    //String[] finalResult = new String[result.size()];
+    UnicodeString *finalResult = NULL;
+    int32_t resultCount;
+    if((resultCount = result.count()) != 0) {
+        finalResult = new UnicodeString[resultCount];
+        if (finalResult == 0) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+    }
+    else {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    //result.toArray(finalResult);
+    result_len = 0;
+    el = UHASH_FIRST;
+    ne = result.nextElement(el);
+    while(ne != NULL) {
+        finalResult[result_len++] = *((UnicodeString *)(ne->value.pointer));
+        ne = result.nextElement(el);
+    }
+
+
+    return finalResult;
+}
+
+Hashtable *CanonicalIterator::getEquivalents2(Hashtable *fillinResult, const UChar *segment, int32_t segLen, UErrorCode &status) {
+
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+    //if (PROGRESS) printf("Adding: %s\n", UToS(Tr(segment)));
+
+    UnicodeString toPut(segment, segLen);
+
+    fillinResult->put(toPut, new UnicodeString(toPut), status);
+
+    UnicodeSet starts;
+
+    // cycle through all the characters
+    UChar32 cp;
+    for (int32_t i = 0; i < segLen; i += U16_LENGTH(cp)) {
+        // see if any character is at the start of some decomposition
+        U16_GET(segment, 0, i, segLen, cp);
+        if (!nfcImpl.getCanonStartSet(cp, starts)) {
+            continue;
+        }
+        // if so, see which decompositions match
+        UnicodeSetIterator iter(starts);
+        while (iter.next()) {
+            UChar32 cp2 = iter.getCodepoint();
+            Hashtable remainder(status);
+            remainder.setValueDeleter(uprv_deleteUObject);
+            if (extract(&remainder, cp2, segment, segLen, i, status) == NULL) {
+                continue;
+            }
+
+            // there were some matches, so add all the possibilities to the set.
+            UnicodeString prefix(segment, i);
+            prefix += cp2;
+
+            int32_t el = UHASH_FIRST;
+            const UHashElement *ne = remainder.nextElement(el);
+            while (ne != NULL) {
+                UnicodeString item = *((UnicodeString *)(ne->value.pointer));
+                UnicodeString *toAdd = new UnicodeString(prefix);
+                /* test for NULL */
+                if (toAdd == 0) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                    return NULL;
+                }
+                *toAdd += item;
+                fillinResult->put(*toAdd, toAdd, status);
+
+                //if (PROGRESS) printf("Adding: %s\n", UToS(Tr(*toAdd)));
+
+                ne = remainder.nextElement(el);
+            }
+        }
+    }
+
+    /* Test for buffer overflows */
+    if(U_FAILURE(status)) {
+        return NULL;
+    }
+    return fillinResult;
+}
+
+/**
+ * See if the decomposition of cp2 is at segment starting at segmentPos 
+ * (with canonical rearrangment!)
+ * If so, take the remainder, and return the equivalents 
+ */
+Hashtable *CanonicalIterator::extract(Hashtable *fillinResult, UChar32 comp, const UChar *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status) {
+//Hashtable *CanonicalIterator::extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status) {
+    //if (PROGRESS) printf(" extract: %s, ", UToS(Tr(UnicodeString(comp))));
+    //if (PROGRESS) printf("%s, %i\n", UToS(Tr(segment)), segmentPos);
+
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+    UnicodeString temp(comp);
+    int32_t inputLen=temp.length();
+    UnicodeString decompString;
+    nfd.normalize(temp, decompString, status);
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+    if (decompString.isBogus()) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    const UChar *decomp=decompString.getBuffer();
+    int32_t decompLen=decompString.length();
+
+    // See if it matches the start of segment (at segmentPos)
+    UBool ok = FALSE;
+    UChar32 cp;
+    int32_t decompPos = 0;
+    UChar32 decompCp;
+    U16_NEXT(decomp, decompPos, decompLen, decompCp);
+
+    int32_t i = segmentPos;
+    while(i < segLen) {
+        U16_NEXT(segment, i, segLen, cp);
+
+        if (cp == decompCp) { // if equal, eat another cp from decomp
+
+            //if (PROGRESS) printf("  matches: %s\n", UToS(Tr(UnicodeString(cp))));
+
+            if (decompPos == decompLen) { // done, have all decomp characters!
+                temp.append(segment+i, segLen-i);
+                ok = TRUE;
+                break;
+            }
+            U16_NEXT(decomp, decompPos, decompLen, decompCp);
+        } else {
+            //if (PROGRESS) printf("  buffer: %s\n", UToS(Tr(UnicodeString(cp))));
+
+            // brute force approach
+            temp.append(cp);
+
+            /* TODO: optimize
+            // since we know that the classes are monotonically increasing, after zero
+            // e.g. 0 5 7 9 0 3
+            // we can do an optimization
+            // there are only a few cases that work: zero, less, same, greater
+            // if both classes are the same, we fail
+            // if the decomp class < the segment class, we fail
+
+            segClass = getClass(cp);
+            if (decompClass <= segClass) return null;
+            */
+        }
+    }
+    if (!ok)
+        return NULL; // we failed, characters left over
+
+    //if (PROGRESS) printf("Matches\n");
+
+    if (inputLen == temp.length()) {
+        fillinResult->put(UnicodeString(), new UnicodeString(), status);
+        return fillinResult; // succeed, but no remainder
+    }
+
+    // brute force approach
+    // check to make sure result is canonically equivalent
+    UnicodeString trial;
+    nfd.normalize(temp, trial, status);
+    if(U_FAILURE(status) || trial.compare(segment+segmentPos, segLen - segmentPos) != 0) {
+        return NULL;
+    }
+
+    return getEquivalents2(fillinResult, temp.getBuffer()+inputLen, temp.length()-inputLen, status);
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
diff --git a/thirdparty/icu4c/common/capi_helper.h b/thirdparty/icu4c/common/capi_helper.h
new file mode 100644
index 0000000000..54b1db9e33
--- /dev/null
+++ b/thirdparty/icu4c/common/capi_helper.h
@@ -0,0 +1,97 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+#ifndef __CAPI_HELPER_H__
+#define __CAPI_HELPER_H__
+
+#include "unicode/utypes.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * An internal helper class to help convert between C and C++ APIs.
+ */
+template<typename CType, typename CPPType, int32_t kMagic>
+class IcuCApiHelper {
+  public:
+    /**
+     * Convert from the C type to the C++ type (const version).
+     */
+    static const CPPType* validate(const CType* input, UErrorCode& status);
+
+    /**
+     * Convert from the C type to the C++ type (non-const version).
+     */
+    static CPPType* validate(CType* input, UErrorCode& status);
+
+    /**
+     * Convert from the C++ type to the C type (const version).
+     */
+    const CType* exportConstForC() const;
+
+    /**
+     * Convert from the C++ type to the C type (non-const version).
+     */
+    CType* exportForC();
+
+    /**
+     * Invalidates the object.
+     */
+    ~IcuCApiHelper();
+
+  private:
+    /**
+     * While the object is valid, fMagic equals kMagic.
+     */
+    int32_t fMagic = kMagic;
+};
+
+
+template<typename CType, typename CPPType, int32_t kMagic>
+const CPPType*
+IcuCApiHelper<CType, CPPType, kMagic>::validate(const CType* input, UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+    if (input == nullptr) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    auto* impl = reinterpret_cast<const CPPType*>(input);
+    if (static_cast<const IcuCApiHelper<CType, CPPType, kMagic>*>(impl)->fMagic != kMagic) {
+        status = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+    return impl;
+}
+
+template<typename CType, typename CPPType, int32_t kMagic>
+CPPType*
+IcuCApiHelper<CType, CPPType, kMagic>::validate(CType* input, UErrorCode& status) {
+    auto* constInput = static_cast<const CType*>(input);
+    auto* validated = validate(constInput, status);
+    return const_cast<CPPType*>(validated);
+}
+
+template<typename CType, typename CPPType, int32_t kMagic>
+const CType*
+IcuCApiHelper<CType, CPPType, kMagic>::exportConstForC() const {
+    return reinterpret_cast<const CType*>(static_cast<const CPPType*>(this));
+}
+
+template<typename CType, typename CPPType, int32_t kMagic>
+CType*
+IcuCApiHelper<CType, CPPType, kMagic>::exportForC() {
+    return reinterpret_cast<CType*>(static_cast<CPPType*>(this));
+}
+
+template<typename CType, typename CPPType, int32_t kMagic>
+IcuCApiHelper<CType, CPPType, kMagic>::~IcuCApiHelper() {
+    // head off application errors by preventing use of of deleted objects.
+    fMagic = 0;
+}
+
+
+U_NAMESPACE_END
+
+#endif // __CAPI_HELPER_H__
diff --git a/thirdparty/icu4c/common/characterproperties.cpp b/thirdparty/icu4c/common/characterproperties.cpp
new file mode 100644
index 0000000000..7b50a4e205
--- /dev/null
+++ b/thirdparty/icu4c/common/characterproperties.cpp
@@ -0,0 +1,383 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// characterproperties.cpp
+// created: 2018sep03 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/localpointer.h"
+#include "unicode/uchar.h"
+#include "unicode/ucpmap.h"
+#include "unicode/ucptrie.h"
+#include "unicode/umutablecptrie.h"
+#include "unicode/uniset.h"
+#include "unicode/uscript.h"
+#include "unicode/uset.h"
+#include "cmemory.h"
+#include "mutex.h"
+#include "normalizer2impl.h"
+#include "uassert.h"
+#include "ubidi_props.h"
+#include "ucase.h"
+#include "ucln_cmn.h"
+#include "umutex.h"
+#include "uprops.h"
+
+using icu::LocalPointer;
+#if !UCONFIG_NO_NORMALIZATION
+using icu::Normalizer2Factory;
+using icu::Normalizer2Impl;
+#endif
+using icu::UInitOnce;
+using icu::UnicodeSet;
+
+namespace {
+
+UBool U_CALLCONV characterproperties_cleanup();
+
+constexpr int32_t NUM_INCLUSIONS = UPROPS_SRC_COUNT + UCHAR_INT_LIMIT - UCHAR_INT_START;
+
+struct Inclusion {
+    UnicodeSet  *fSet = nullptr;
+    UInitOnce    fInitOnce = U_INITONCE_INITIALIZER;
+};
+Inclusion gInclusions[NUM_INCLUSIONS]; // cached getInclusions()
+
+UnicodeSet *sets[UCHAR_BINARY_LIMIT] = {};
+
+UCPMap *maps[UCHAR_INT_LIMIT - UCHAR_INT_START] = {};
+
+icu::UMutex cpMutex;
+
+//----------------------------------------------------------------
+// Inclusions list
+//----------------------------------------------------------------
+
+// USetAdder implementation
+// Does not use uset.h to reduce code dependencies
+void U_CALLCONV
+_set_add(USet *set, UChar32 c) {
+    ((UnicodeSet *)set)->add(c);
+}
+
+void U_CALLCONV
+_set_addRange(USet *set, UChar32 start, UChar32 end) {
+    ((UnicodeSet *)set)->add(start, end);
+}
+
+void U_CALLCONV
+_set_addString(USet *set, const UChar *str, int32_t length) {
+    ((UnicodeSet *)set)->add(icu::UnicodeString((UBool)(length<0), str, length));
+}
+
+UBool U_CALLCONV characterproperties_cleanup() {
+    for (Inclusion &in: gInclusions) {
+        delete in.fSet;
+        in.fSet = nullptr;
+        in.fInitOnce.reset();
+    }
+    for (int32_t i = 0; i < UPRV_LENGTHOF(sets); ++i) {
+        delete sets[i];
+        sets[i] = nullptr;
+    }
+    for (int32_t i = 0; i < UPRV_LENGTHOF(maps); ++i) {
+        ucptrie_close(reinterpret_cast<UCPTrie *>(maps[i]));
+        maps[i] = nullptr;
+    }
+    return TRUE;
+}
+
+void U_CALLCONV initInclusion(UPropertySource src, UErrorCode &errorCode) {
+    // This function is invoked only via umtx_initOnce().
+    U_ASSERT(0 <= src && src < UPROPS_SRC_COUNT);
+    if (src == UPROPS_SRC_NONE) {
+        errorCode = U_INTERNAL_PROGRAM_ERROR;
+        return;
+    }
+    U_ASSERT(gInclusions[src].fSet == nullptr);
+
+    LocalPointer<UnicodeSet> incl(new UnicodeSet());
+    if (incl.isNull()) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    USetAdder sa = {
+        (USet *)incl.getAlias(),
+        _set_add,
+        _set_addRange,
+        _set_addString,
+        nullptr, // don't need remove()
+        nullptr // don't need removeRange()
+    };
+
+    switch(src) {
+    case UPROPS_SRC_CHAR:
+        uchar_addPropertyStarts(&sa, &errorCode);
+        break;
+    case UPROPS_SRC_PROPSVEC:
+        upropsvec_addPropertyStarts(&sa, &errorCode);
+        break;
+    case UPROPS_SRC_CHAR_AND_PROPSVEC:
+        uchar_addPropertyStarts(&sa, &errorCode);
+        upropsvec_addPropertyStarts(&sa, &errorCode);
+        break;
+#if !UCONFIG_NO_NORMALIZATION
+    case UPROPS_SRC_CASE_AND_NORM: {
+        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+        if(U_SUCCESS(errorCode)) {
+            impl->addPropertyStarts(&sa, errorCode);
+        }
+        ucase_addPropertyStarts(&sa, &errorCode);
+        break;
+    }
+    case UPROPS_SRC_NFC: {
+        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+        if(U_SUCCESS(errorCode)) {
+            impl->addPropertyStarts(&sa, errorCode);
+        }
+        break;
+    }
+    case UPROPS_SRC_NFKC: {
+        const Normalizer2Impl *impl=Normalizer2Factory::getNFKCImpl(errorCode);
+        if(U_SUCCESS(errorCode)) {
+            impl->addPropertyStarts(&sa, errorCode);
+        }
+        break;
+    }
+    case UPROPS_SRC_NFKC_CF: {
+        const Normalizer2Impl *impl=Normalizer2Factory::getNFKC_CFImpl(errorCode);
+        if(U_SUCCESS(errorCode)) {
+            impl->addPropertyStarts(&sa, errorCode);
+        }
+        break;
+    }
+    case UPROPS_SRC_NFC_CANON_ITER: {
+        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+        if(U_SUCCESS(errorCode)) {
+            impl->addCanonIterPropertyStarts(&sa, errorCode);
+        }
+        break;
+    }
+#endif
+    case UPROPS_SRC_CASE:
+        ucase_addPropertyStarts(&sa, &errorCode);
+        break;
+    case UPROPS_SRC_BIDI:
+        ubidi_addPropertyStarts(&sa, &errorCode);
+        break;
+    case UPROPS_SRC_INPC:
+    case UPROPS_SRC_INSC:
+    case UPROPS_SRC_VO:
+        uprops_addPropertyStarts((UPropertySource)src, &sa, &errorCode);
+        break;
+    default:
+        errorCode = U_INTERNAL_PROGRAM_ERROR;
+        break;
+    }
+
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if (incl->isBogus()) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    // Compact for caching.
+    incl->compact();
+    gInclusions[src].fSet = incl.orphan();
+    ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
+}
+
+const UnicodeSet *getInclusionsForSource(UPropertySource src, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    if (src < 0 || UPROPS_SRC_COUNT <= src) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    Inclusion &i = gInclusions[src];
+    umtx_initOnce(i.fInitOnce, &initInclusion, src, errorCode);
+    return i.fSet;
+}
+
+void U_CALLCONV initIntPropInclusion(UProperty prop, UErrorCode &errorCode) {
+    // This function is invoked only via umtx_initOnce().
+    U_ASSERT(UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT);
+    int32_t inclIndex = UPROPS_SRC_COUNT + prop - UCHAR_INT_START;
+    U_ASSERT(gInclusions[inclIndex].fSet == nullptr);
+    UPropertySource src = uprops_getSource(prop);
+    const UnicodeSet *incl = getInclusionsForSource(src, errorCode);
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+
+    LocalPointer<UnicodeSet> intPropIncl(new UnicodeSet(0, 0));
+    if (intPropIncl.isNull()) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    int32_t numRanges = incl->getRangeCount();
+    int32_t prevValue = 0;
+    for (int32_t i = 0; i < numRanges; ++i) {
+        UChar32 rangeEnd = incl->getRangeEnd(i);
+        for (UChar32 c = incl->getRangeStart(i); c <= rangeEnd; ++c) {
+            // TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
+            int32_t value = u_getIntPropertyValue(c, prop);
+            if (value != prevValue) {
+                intPropIncl->add(c);
+                prevValue = value;
+            }
+        }
+    }
+
+    if (intPropIncl->isBogus()) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    // Compact for caching.
+    intPropIncl->compact();
+    gInclusions[inclIndex].fSet = intPropIncl.orphan();
+    ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
+}
+
+}  // namespace
+
+U_NAMESPACE_BEGIN
+
+const UnicodeSet *CharacterProperties::getInclusionsForProperty(
+        UProperty prop, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    if (UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT) {
+        int32_t inclIndex = UPROPS_SRC_COUNT + prop - UCHAR_INT_START;
+        Inclusion &i = gInclusions[inclIndex];
+        umtx_initOnce(i.fInitOnce, &initIntPropInclusion, prop, errorCode);
+        return i.fSet;
+    } else {
+        UPropertySource src = uprops_getSource(prop);
+        return getInclusionsForSource(src, errorCode);
+    }
+}
+
+U_NAMESPACE_END
+
+namespace {
+
+UnicodeSet *makeSet(UProperty property, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    LocalPointer<UnicodeSet> set(new UnicodeSet());
+    if (set.isNull()) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+    const UnicodeSet *inclusions =
+        icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    int32_t numRanges = inclusions->getRangeCount();
+    UChar32 startHasProperty = -1;
+
+    for (int32_t i = 0; i < numRanges; ++i) {
+        UChar32 rangeEnd = inclusions->getRangeEnd(i);
+        for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
+            // TODO: Get a UCharacterProperty.BinaryProperty to avoid the property dispatch.
+            if (u_hasBinaryProperty(c, property)) {
+                if (startHasProperty < 0) {
+                    // Transition from false to true.
+                    startHasProperty = c;
+                }
+            } else if (startHasProperty >= 0) {
+                // Transition from true to false.
+                set->add(startHasProperty, c - 1);
+                startHasProperty = -1;
+            }
+        }
+    }
+    if (startHasProperty >= 0) {
+        set->add(startHasProperty, 0x10FFFF);
+    }
+    set->freeze();
+    return set.orphan();
+}
+
+UCPMap *makeMap(UProperty property, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    uint32_t nullValue = property == UCHAR_SCRIPT ? USCRIPT_UNKNOWN : 0;
+    icu::LocalUMutableCPTriePointer mutableTrie(
+        umutablecptrie_open(nullValue, nullValue, &errorCode));
+    const UnicodeSet *inclusions =
+        icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    int32_t numRanges = inclusions->getRangeCount();
+    UChar32 start = 0;
+    uint32_t value = nullValue;
+
+    for (int32_t i = 0; i < numRanges; ++i) {
+        UChar32 rangeEnd = inclusions->getRangeEnd(i);
+        for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
+            // TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
+            uint32_t nextValue = u_getIntPropertyValue(c, property);
+            if (value != nextValue) {
+                if (value != nullValue) {
+                    umutablecptrie_setRange(mutableTrie.getAlias(), start, c - 1, value, &errorCode);
+                }
+                start = c;
+                value = nextValue;
+            }
+        }
+    }
+    if (value != 0) {
+        umutablecptrie_setRange(mutableTrie.getAlias(), start, 0x10FFFF, value, &errorCode);
+    }
+
+    UCPTrieType type;
+    if (property == UCHAR_BIDI_CLASS || property == UCHAR_GENERAL_CATEGORY) {
+        type = UCPTRIE_TYPE_FAST;
+    } else {
+        type = UCPTRIE_TYPE_SMALL;
+    }
+    UCPTrieValueWidth valueWidth;
+    // TODO: UCharacterProperty.IntProperty
+    int32_t max = u_getIntPropertyMaxValue(property);
+    if (max <= 0xff) {
+        valueWidth = UCPTRIE_VALUE_BITS_8;
+    } else if (max <= 0xffff) {
+        valueWidth = UCPTRIE_VALUE_BITS_16;
+    } else {
+        valueWidth = UCPTRIE_VALUE_BITS_32;
+    }
+    return reinterpret_cast<UCPMap *>(
+        umutablecptrie_buildImmutable(mutableTrie.getAlias(), type, valueWidth, &errorCode));
+}
+
+}  // namespace
+
+U_NAMESPACE_USE
+
+U_CAPI const USet * U_EXPORT2
+u_getBinaryPropertySet(UProperty property, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) { return nullptr; }
+    if (property < 0 || UCHAR_BINARY_LIMIT <= property) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    Mutex m(&cpMutex);
+    UnicodeSet *set = sets[property];
+    if (set == nullptr) {
+        sets[property] = set = makeSet(property, *pErrorCode);
+    }
+    if (U_FAILURE(*pErrorCode)) { return nullptr; }
+    return set->toUSet();
+}
+
+U_CAPI const UCPMap * U_EXPORT2
+u_getIntPropertyMap(UProperty property, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) { return nullptr; }
+    if (property < UCHAR_INT_START || UCHAR_INT_LIMIT <= property) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    Mutex m(&cpMutex);
+    UCPMap *map = maps[property - UCHAR_INT_START];
+    if (map == nullptr) {
+        maps[property - UCHAR_INT_START] = map = makeMap(property, *pErrorCode);
+    }
+    return map;
+}
diff --git a/thirdparty/icu4c/common/chariter.cpp b/thirdparty/icu4c/common/chariter.cpp
new file mode 100644
index 0000000000..887119a0eb
--- /dev/null
+++ b/thirdparty/icu4c/common/chariter.cpp
@@ -0,0 +1,100 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "unicode/chariter.h"
+
+U_NAMESPACE_BEGIN
+
+ForwardCharacterIterator::~ForwardCharacterIterator() {}
+ForwardCharacterIterator::ForwardCharacterIterator()
+: UObject()
+{}
+ForwardCharacterIterator::ForwardCharacterIterator(const ForwardCharacterIterator &other)
+: UObject(other)
+{}
+
+
+CharacterIterator::CharacterIterator()
+: textLength(0), pos(0), begin(0), end(0) {
+}
+
+CharacterIterator::CharacterIterator(int32_t length)
+: textLength(length), pos(0), begin(0), end(length) {
+    if(textLength < 0) {
+        textLength = end = 0;
+    }
+}
+
+CharacterIterator::CharacterIterator(int32_t length, int32_t position)
+: textLength(length), pos(position), begin(0), end(length) {
+    if(textLength < 0) {
+        textLength = end = 0;
+    }
+    if(pos < 0) {
+        pos = 0;
+    } else if(pos > end) {
+        pos = end;
+    }
+}
+
+CharacterIterator::CharacterIterator(int32_t length, int32_t textBegin, int32_t textEnd, int32_t position)
+: textLength(length), pos(position), begin(textBegin), end(textEnd) {
+    if(textLength < 0) {
+        textLength = 0;
+    }
+    if(begin < 0) {
+        begin = 0;
+    } else if(begin > textLength) {
+        begin = textLength;
+    }
+    if(end < begin) {
+        end = begin;
+    } else if(end > textLength) {
+        end = textLength;
+    }
+    if(pos < begin) {
+        pos = begin;
+    } else if(pos > end) {
+        pos = end;
+    }
+}
+
+CharacterIterator::~CharacterIterator() {}
+
+CharacterIterator::CharacterIterator(const CharacterIterator &that) :
+ForwardCharacterIterator(that),
+textLength(that.textLength), pos(that.pos), begin(that.begin), end(that.end)
+{
+}
+
+CharacterIterator &
+CharacterIterator::operator=(const CharacterIterator &that) {
+    ForwardCharacterIterator::operator=(that);
+    textLength = that.textLength;
+    pos = that.pos;
+    begin = that.begin;
+    end = that.end;
+    return *this;
+}
+
+// implementing first[32]PostInc() directly in a subclass should be faster
+// but these implementations make subclassing a little easier
+UChar
+CharacterIterator::firstPostInc(void) {
+    setToStart();
+    return nextPostInc();
+}
+
+UChar32
+CharacterIterator::first32PostInc(void) {
+    setToStart();
+    return next32PostInc();
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/charstr.cpp b/thirdparty/icu4c/common/charstr.cpp
new file mode 100644
index 0000000000..318a185b3f
--- /dev/null
+++ b/thirdparty/icu4c/common/charstr.cpp
@@ -0,0 +1,239 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  charstr.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010may19
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uinvchar.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+CharString::CharString(CharString&& src) U_NOEXCEPT
+        : buffer(std::move(src.buffer)), len(src.len) {
+    src.len = 0;  // not strictly necessary because we make no guarantees on the source string
+}
+
+CharString& CharString::operator=(CharString&& src) U_NOEXCEPT {
+    buffer = std::move(src.buffer);
+    len = src.len;
+    src.len = 0;  // not strictly necessary because we make no guarantees on the source string
+    return *this;
+}
+
+char *CharString::cloneData(UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    char *p = static_cast<char *>(uprv_malloc(len + 1));
+    if (p == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+    uprv_memcpy(p, buffer.getAlias(), len + 1);
+    return p;
+}
+
+int32_t CharString::extract(char *dest, int32_t capacity, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) { return len; }
+    if (capacity < 0 || (capacity > 0 && dest == nullptr)) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return len;
+    }
+    const char *src = buffer.getAlias();
+    if (0 < len && len <= capacity && src != dest) {
+        uprv_memcpy(dest, src, len);
+    }
+    return u_terminateChars(dest, capacity, len, &errorCode);
+}
+
+CharString &CharString::copyFrom(const CharString &s, UErrorCode &errorCode) {
+    if(U_SUCCESS(errorCode) && this!=&s && ensureCapacity(s.len+1, 0, errorCode)) {
+        len=s.len;
+        uprv_memcpy(buffer.getAlias(), s.buffer.getAlias(), len+1);
+    }
+    return *this;
+}
+
+int32_t CharString::lastIndexOf(char c) const {
+    for(int32_t i=len; i>0;) {
+        if(buffer[--i]==c) {
+            return i;
+        }
+    }
+    return -1;
+}
+
+bool CharString::contains(StringPiece s) const {
+    if (s.empty()) { return false; }
+    const char *p = buffer.getAlias();
+    int32_t lastStart = len - s.length();
+    for (int32_t i = 0; i <= lastStart; ++i) {
+        if (uprv_memcmp(p + i, s.data(), s.length()) == 0) {
+            return true;
+        }
+    }
+    return false;
+}
+
+CharString &CharString::truncate(int32_t newLength) {
+    if(newLength<0) {
+        newLength=0;
+    }
+    if(newLength<len) {
+        buffer[len=newLength]=0;
+    }
+    return *this;
+}
+
+CharString &CharString::append(char c, UErrorCode &errorCode) {
+    if(ensureCapacity(len+2, 0, errorCode)) {
+        buffer[len++]=c;
+        buffer[len]=0;
+    }
+    return *this;
+}
+
+CharString &CharString::append(const char *s, int32_t sLength, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return *this;
+    }
+    if(sLength<-1 || (s==NULL && sLength!=0)) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    if(sLength<0) {
+        sLength= static_cast<int32_t>(uprv_strlen(s));
+    }
+    if(sLength>0) {
+        if(s==(buffer.getAlias()+len)) {
+            // The caller wrote into the getAppendBuffer().
+            if(sLength>=(buffer.getCapacity()-len)) {
+                // The caller wrote too much.
+                errorCode=U_INTERNAL_PROGRAM_ERROR;
+            } else {
+                buffer[len+=sLength]=0;
+            }
+        } else if(buffer.getAlias()<=s && s<(buffer.getAlias()+len) &&
+                  sLength>=(buffer.getCapacity()-len)
+        ) {
+            // (Part of) this string is appended to itself which requires reallocation,
+            // so we have to make a copy of the substring and append that.
+            return append(CharString(s, sLength, errorCode), errorCode);
+        } else if(ensureCapacity(len+sLength+1, 0, errorCode)) {
+            uprv_memcpy(buffer.getAlias()+len, s, sLength);
+            buffer[len+=sLength]=0;
+        }
+    }
+    return *this;
+}
+
+char *CharString::getAppendBuffer(int32_t minCapacity,
+                                  int32_t desiredCapacityHint,
+                                  int32_t &resultCapacity,
+                                  UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        resultCapacity=0;
+        return NULL;
+    }
+    int32_t appendCapacity=buffer.getCapacity()-len-1;  // -1 for NUL
+    if(appendCapacity>=minCapacity) {
+        resultCapacity=appendCapacity;
+        return buffer.getAlias()+len;
+    }
+    if(ensureCapacity(len+minCapacity+1, len+desiredCapacityHint+1, errorCode)) {
+        resultCapacity=buffer.getCapacity()-len-1;
+        return buffer.getAlias()+len;
+    }
+    resultCapacity=0;
+    return NULL;
+}
+
+CharString &CharString::appendInvariantChars(const UnicodeString &s, UErrorCode &errorCode) {
+    return appendInvariantChars(s.getBuffer(), s.length(), errorCode);
+}
+
+CharString &CharString::appendInvariantChars(const UChar* uchars, int32_t ucharsLen, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return *this;
+    }
+    if (!uprv_isInvariantUString(uchars, ucharsLen)) {
+        errorCode = U_INVARIANT_CONVERSION_ERROR;
+        return *this;
+    }
+    if(ensureCapacity(len+ucharsLen+1, 0, errorCode)) {
+        u_UCharsToChars(uchars, buffer.getAlias()+len, ucharsLen);
+        len += ucharsLen;
+        buffer[len] = 0;
+    }
+    return *this;
+}
+
+UBool CharString::ensureCapacity(int32_t capacity,
+                                 int32_t desiredCapacityHint,
+                                 UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    if(capacity>buffer.getCapacity()) {
+        if(desiredCapacityHint==0) {
+            desiredCapacityHint=capacity+buffer.getCapacity();
+        }
+        if( (desiredCapacityHint<=capacity || buffer.resize(desiredCapacityHint, len+1)==NULL) &&
+            buffer.resize(capacity, len+1)==NULL
+        ) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+CharString &CharString::appendPathPart(StringPiece s, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return *this;
+    }
+    if(s.length()==0) {
+        return *this;
+    }
+    char c;
+    if(len>0 && (c=buffer[len-1])!=U_FILE_SEP_CHAR && c!=U_FILE_ALT_SEP_CHAR) {
+        append(getDirSepChar(), errorCode);
+    }
+    append(s, errorCode);
+    return *this;
+}
+
+CharString &CharString::ensureEndsWithFileSeparator(UErrorCode &errorCode) {
+    char c;
+    if(U_SUCCESS(errorCode) && len>0 &&
+            (c=buffer[len-1])!=U_FILE_SEP_CHAR && c!=U_FILE_ALT_SEP_CHAR) {
+        append(getDirSepChar(), errorCode);
+    }
+    return *this;
+}
+
+char CharString::getDirSepChar() const {
+    char dirSepChar = U_FILE_SEP_CHAR;
+#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)
+    // We may need to return a different directory separator when building for Cygwin or MSYS2.
+    if(len>0 && !uprv_strchr(data(), U_FILE_SEP_CHAR) && uprv_strchr(data(), U_FILE_ALT_SEP_CHAR))
+        dirSepChar = U_FILE_ALT_SEP_CHAR;
+#endif
+    return dirSepChar;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/charstr.h b/thirdparty/icu4c/common/charstr.h
new file mode 100644
index 0000000000..6619faac61
--- /dev/null
+++ b/thirdparty/icu4c/common/charstr.h
@@ -0,0 +1,190 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2001-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   11/19/2001  aliu        Creation.
+*   05/19/2010  markus      Rewritten from scratch
+**********************************************************************
+*/
+
+#ifndef CHARSTRING_H
+#define CHARSTRING_H
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+// Windows needs us to DLL-export the MaybeStackArray template specialization,
+// but MacOS X cannot handle it. Same as in digitlst.h.
+#if !U_PLATFORM_IS_DARWIN_BASED
+template class U_COMMON_API MaybeStackArray<char, 40>;
+#endif
+
+/**
+ * ICU-internal char * string class.
+ * This class does not assume or enforce any particular character encoding.
+ * Raw bytes can be stored. The string object owns its characters.
+ * A terminating NUL is stored, but the class does not prevent embedded NUL characters.
+ *
+ * This class wants to be convenient but is also deliberately minimalist.
+ * Please do not add methods if they only add minor convenience.
+ * For example:
+ *   cs.data()[5]='a';  // no need for setCharAt(5, 'a')
+ */
+class U_COMMON_API CharString : public UMemory {
+public:
+    CharString() : len(0) { buffer[0]=0; }
+    CharString(StringPiece s, UErrorCode &errorCode) : len(0) {
+        buffer[0]=0;
+        append(s, errorCode);
+    }
+    CharString(const CharString &s, UErrorCode &errorCode) : len(0) {
+        buffer[0]=0;
+        append(s, errorCode);
+    }
+    CharString(const char *s, int32_t sLength, UErrorCode &errorCode) : len(0) {
+        buffer[0]=0;
+        append(s, sLength, errorCode);
+    }
+    ~CharString() {}
+
+    /**
+     * Move constructor; might leave src in an undefined state.
+     * This string will have the same contents and state that the source string had.
+     */
+    CharString(CharString &&src) U_NOEXCEPT;
+    /**
+     * Move assignment operator; might leave src in an undefined state.
+     * This string will have the same contents and state that the source string had.
+     * The behavior is undefined if *this and src are the same object.
+     */
+    CharString &operator=(CharString &&src) U_NOEXCEPT;
+
+    /**
+     * Replaces this string's contents with the other string's contents.
+     * CharString does not support the standard copy constructor nor
+     * the assignment operator, to make copies explicit and to
+     * use a UErrorCode where memory allocations might be needed.
+     */
+    CharString &copyFrom(const CharString &other, UErrorCode &errorCode);
+
+    UBool isEmpty() const { return len==0; }
+    int32_t length() const { return len; }
+    char operator[](int32_t index) const { return buffer[index]; }
+    StringPiece toStringPiece() const { return StringPiece(buffer.getAlias(), len); }
+
+    const char *data() const { return buffer.getAlias(); }
+    char *data() { return buffer.getAlias(); }
+    /**
+     * Allocates length()+1 chars and copies the NUL-terminated data().
+     * The caller must uprv_free() the result.
+     */
+    char *cloneData(UErrorCode &errorCode) const;
+    /**
+     * Copies the contents of the string into dest.
+     * Checks if there is enough space in dest, extracts the entire string if possible,
+     * and NUL-terminates dest if possible.
+     *
+     * If the string fits into dest but cannot be NUL-terminated (length()==capacity),
+     * then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
+     * If the string itself does not fit into dest (length()>capacity),
+     * then the error code is set to U_BUFFER_OVERFLOW_ERROR.
+     *
+     * @param dest Destination string buffer.
+     * @param capacity Size of the dest buffer (number of chars).
+     * @param errorCode ICU error code.
+     * @return length()
+     */
+    int32_t extract(char *dest, int32_t capacity, UErrorCode &errorCode) const;
+
+    bool operator==(StringPiece other) const {
+        return len == other.length() && (len == 0 || uprv_memcmp(data(), other.data(), len) == 0);
+    }
+    bool operator!=(StringPiece other) const {
+        return !operator==(other);
+    }
+
+    /** @return last index of c, or -1 if c is not in this string */
+    int32_t lastIndexOf(char c) const;
+
+    bool contains(StringPiece s) const;
+
+    CharString &clear() { len=0; buffer[0]=0; return *this; }
+    CharString &truncate(int32_t newLength);
+
+    CharString &append(char c, UErrorCode &errorCode);
+    CharString &append(StringPiece s, UErrorCode &errorCode) {
+        return append(s.data(), s.length(), errorCode);
+    }
+    CharString &append(const CharString &s, UErrorCode &errorCode) {
+        return append(s.data(), s.length(), errorCode);
+    }
+    CharString &append(const char *s, int32_t sLength, UErrorCode &status);
+    /**
+     * Returns a writable buffer for appending and writes the buffer's capacity to
+     * resultCapacity. Guarantees resultCapacity>=minCapacity if U_SUCCESS().
+     * There will additionally be space for a terminating NUL right at resultCapacity.
+     * (This function is similar to ByteSink.GetAppendBuffer().)
+     *
+     * The returned buffer is only valid until the next write operation
+     * on this string.
+     *
+     * After writing at most resultCapacity bytes, call append() with the
+     * pointer returned from this function and the number of bytes written.
+     *
+     * @param minCapacity required minimum capacity of the returned buffer;
+     *                    must be non-negative
+     * @param desiredCapacityHint desired capacity of the returned buffer;
+     *                            must be non-negative
+     * @param resultCapacity will be set to the capacity of the returned buffer
+     * @param errorCode in/out error code
+     * @return a buffer with resultCapacity>=min_capacity
+     */
+    char *getAppendBuffer(int32_t minCapacity,
+                          int32_t desiredCapacityHint,
+                          int32_t &resultCapacity,
+                          UErrorCode &errorCode);
+
+    CharString &appendInvariantChars(const UnicodeString &s, UErrorCode &errorCode);
+    CharString &appendInvariantChars(const UChar* uchars, int32_t ucharsLen, UErrorCode& errorCode);
+
+    /**
+     * Appends a filename/path part, e.g., a directory name.
+     * First appends a U_FILE_SEP_CHAR or U_FILE_ALT_SEP_CHAR if necessary.
+     * Does nothing if s is empty.
+     */
+    CharString &appendPathPart(StringPiece s, UErrorCode &errorCode);
+
+    /**
+     * Appends a U_FILE_SEP_CHAR or U_FILE_ALT_SEP_CHAR if this string is not empty
+     * and does not already end with a U_FILE_SEP_CHAR or U_FILE_ALT_SEP_CHAR.
+     */
+    CharString &ensureEndsWithFileSeparator(UErrorCode &errorCode);
+
+private:
+    MaybeStackArray<char, 40> buffer;
+    int32_t len;
+
+    UBool ensureCapacity(int32_t capacity, int32_t desiredCapacityHint, UErrorCode &errorCode);
+
+    CharString(const CharString &other); // forbid copying of this class
+    CharString &operator=(const CharString &other); // forbid copying of this class
+
+    /**
+     * Returns U_FILE_ALT_SEP_CHAR if found in string, and U_FILE_SEP_CHAR is not found.
+     * Otherwise returns U_FILE_SEP_CHAR.
+     */
+    char getDirSepChar() const;
+};
+
+U_NAMESPACE_END
+
+#endif
+//eof
diff --git a/thirdparty/icu4c/common/charstrmap.h b/thirdparty/icu4c/common/charstrmap.h
new file mode 100644
index 0000000000..3320a46208
--- /dev/null
+++ b/thirdparty/icu4c/common/charstrmap.h
@@ -0,0 +1,55 @@
+// © 2020 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// charstrmap.h
+// created: 2020sep01 Frank Yung-Fong Tang
+
+#ifndef __CHARSTRMAP_H__
+#define __CHARSTRMAP_H__
+
+#include <utility>
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "uhash.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Map of const char * keys & values.
+ * Stores pointers as is: Does not own/copy/adopt/release strings.
+ */
+class CharStringMap final : public UMemory {
+public:
+    /** Constructs an unusable non-map. */
+    CharStringMap() : map(nullptr) {}
+    CharStringMap(int32_t size, UErrorCode &errorCode) {
+        map = uhash_openSize(uhash_hashChars, uhash_compareChars, uhash_compareChars,
+                             size, &errorCode);
+    }
+    CharStringMap(CharStringMap &&other) U_NOEXCEPT : map(other.map) {
+        other.map = nullptr;
+    }
+    CharStringMap(const CharStringMap &other) = delete;
+    ~CharStringMap() {
+        uhash_close(map);
+    }
+
+    CharStringMap &operator=(CharStringMap &&other) U_NOEXCEPT {
+        map = other.map;
+        other.map = nullptr;
+        return *this;
+    }
+    CharStringMap &operator=(const CharStringMap &other) = delete;
+
+    const char *get(const char *key) const { return static_cast<const char *>(uhash_get(map, key)); }
+    void put(const char *key, const char *value, UErrorCode &errorCode) {
+        uhash_put(map, const_cast<char *>(key), const_cast<char *>(value), &errorCode);
+    }
+
+private:
+    UHashtable *map;
+};
+
+U_NAMESPACE_END
+
+#endif  //  __CHARSTRMAP_H__
diff --git a/thirdparty/icu4c/common/cmemory.cpp b/thirdparty/icu4c/common/cmemory.cpp
new file mode 100644
index 0000000000..663c1411e4
--- /dev/null
+++ b/thirdparty/icu4c/common/cmemory.cpp
@@ -0,0 +1,138 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2002-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File cmemory.c      ICU Heap allocation.
+*                     All ICU heap allocation, both for C and C++ new of ICU
+*                     class types, comes through these functions.
+*
+*                     If you have a need to replace ICU allocation, this is the
+*                     place to do it.
+*
+*                     Note that uprv_malloc(0) returns a non-NULL pointer, and
+*                     that a subsequent free of that pointer value is a NOP.
+*
+******************************************************************************
+*/
+#include "unicode/uclean.h"
+#include "cmemory.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include <stdlib.h>
+
+/* uprv_malloc(0) returns a pointer to this read-only data. */
+static const int32_t zeroMem[] = {0, 0, 0, 0, 0, 0};
+
+/* Function Pointers for user-supplied heap functions  */
+static const void     *pContext;
+static UMemAllocFn    *pAlloc;
+static UMemReallocFn  *pRealloc;
+static UMemFreeFn     *pFree;
+
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+#include <stdio.h>
+static int n=0;
+static long b=0; 
+#endif
+
+U_CAPI void * U_EXPORT2
+uprv_malloc(size_t s) {
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+#if 1
+  putchar('>');
+  fflush(stdout);
+#else
+  fprintf(stderr,"MALLOC\t#%d\t%ul bytes\t%ul total\n", ++n,s,(b+=s)); fflush(stderr);
+#endif
+#endif
+    if (s > 0) {
+        if (pAlloc) {
+            return (*pAlloc)(pContext, s);
+        } else {
+            return uprv_default_malloc(s);
+        }
+    } else {
+        return (void *)zeroMem;
+    }
+}
+
+U_CAPI void * U_EXPORT2
+uprv_realloc(void * buffer, size_t size) {
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+  putchar('~');
+  fflush(stdout);
+#endif
+    if (buffer == zeroMem) {
+        return uprv_malloc(size);
+    } else if (size == 0) {
+        if (pFree) {
+            (*pFree)(pContext, buffer);
+        } else {
+            uprv_default_free(buffer);
+        }
+        return (void *)zeroMem;
+    } else {
+        if (pRealloc) {
+            return (*pRealloc)(pContext, buffer, size);
+        } else {
+            return uprv_default_realloc(buffer, size);
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+uprv_free(void *buffer) {
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+  putchar('<');
+  fflush(stdout);
+#endif
+    if (buffer != zeroMem) {
+        if (pFree) {
+            (*pFree)(pContext, buffer);
+        } else {
+            uprv_default_free(buffer);
+        }
+    }
+}
+
+U_CAPI void * U_EXPORT2
+uprv_calloc(size_t num, size_t size) {
+    void *mem = NULL;
+    size *= num;
+    mem = uprv_malloc(size);
+    if (mem) {
+        uprv_memset(mem, 0, size);
+    }
+    return mem;
+}
+
+U_CAPI void U_EXPORT2
+u_setMemoryFunctions(const void *context, UMemAllocFn *a, UMemReallocFn *r, UMemFreeFn *f,  UErrorCode *status)
+{
+    if (U_FAILURE(*status)) {
+        return;
+    }
+    if (a==NULL || r==NULL || f==NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    pContext  = context;
+    pAlloc    = a;
+    pRealloc  = r;
+    pFree     = f;
+}
+
+
+U_CFUNC UBool cmemory_cleanup(void) {
+    pContext   = NULL;
+    pAlloc     = NULL;
+    pRealloc   = NULL;
+    pFree      = NULL;
+    return TRUE;
+}
diff --git a/thirdparty/icu4c/common/cmemory.h b/thirdparty/icu4c/common/cmemory.h
new file mode 100644
index 0000000000..210bc7645e
--- /dev/null
+++ b/thirdparty/icu4c/common/cmemory.h
@@ -0,0 +1,849 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File CMEMORY.H
+*
+*  Contains stdlib.h/string.h memory functions
+*
+* @author       Bertrand A. Damiba
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   6/20/98     Bertrand    Created.
+*  05/03/99     stephen     Changed from functions to macros.
+*
+******************************************************************************
+*/
+
+#ifndef CMEMORY_H
+#define CMEMORY_H
+
+#include "unicode/utypes.h"
+
+#include <stddef.h>
+#include <string.h>
+#include "unicode/localpointer.h"
+
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+#include <stdio.h>
+#endif
+
+
+#define uprv_memcpy(dst, src, size) U_STANDARD_CPP_NAMESPACE memcpy(dst, src, size)
+#define uprv_memmove(dst, src, size) U_STANDARD_CPP_NAMESPACE memmove(dst, src, size)
+
+/**
+ * \def UPRV_LENGTHOF
+ * Convenience macro to determine the length of a fixed array at compile-time.
+ * @param array A fixed length array
+ * @return The length of the array, in elements
+ * @internal
+ */
+#define UPRV_LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
+#define uprv_memset(buffer, mark, size) U_STANDARD_CPP_NAMESPACE memset(buffer, mark, size)
+#define uprv_memcmp(buffer1, buffer2, size) U_STANDARD_CPP_NAMESPACE memcmp(buffer1, buffer2,size)
+#define uprv_memchr(ptr, value, num) U_STANDARD_CPP_NAMESPACE memchr(ptr, value, num)
+
+U_CAPI void * U_EXPORT2
+uprv_malloc(size_t s) U_MALLOC_ATTR U_ALLOC_SIZE_ATTR(1);
+
+U_CAPI void * U_EXPORT2
+uprv_realloc(void *mem, size_t size) U_ALLOC_SIZE_ATTR(2);
+
+U_CAPI void U_EXPORT2
+uprv_free(void *mem);
+
+U_CAPI void * U_EXPORT2
+uprv_calloc(size_t num, size_t size) U_MALLOC_ATTR U_ALLOC_SIZE_ATTR2(1,2);
+
+/**
+ * Get the least significant bits of a pointer (a memory address).
+ * For example, with a mask of 3, the macro gets the 2 least significant bits,
+ * which will be 0 if the pointer is 32-bit (4-byte) aligned.
+ *
+ * uintptr_t is the most appropriate integer type to cast to.
+ */
+#define U_POINTER_MASK_LSB(ptr, mask) ((uintptr_t)(ptr) & (mask))
+
+/**
+ * Create & return an instance of "type" in statically allocated storage.
+ * e.g.
+ *    static std::mutex *myMutex = STATIC_NEW(std::mutex);
+ * To destroy an object created in this way, invoke the destructor explicitly, e.g.
+ *    myMutex->~mutex();
+ * DO NOT use delete.
+ * DO NOT use with class UMutex, which has specific support for static instances.
+ *
+ * STATIC_NEW is intended for use when
+ *   - We want a static (or global) object.
+ *   - We don't want it to ever be destructed, or to explicitly control destruction,
+ *     to avoid use-after-destruction problems.
+ *   - We want to avoid an ordinary heap allocated object,
+ *     to avoid the possibility of memory allocation failures, and
+ *     to avoid memory leak reports, from valgrind, for example.
+ * This is defined as a macro rather than a template function because each invocation
+ * must define distinct static storage for the object being returned.
+ */
+#define STATIC_NEW(type) [] () { \
+    alignas(type) static char storage[sizeof(type)]; \
+    return new(storage) type();} ()
+
+/**
+  *  Heap clean up function, called from u_cleanup()
+  *    Clears any user heap functions from u_setMemoryFunctions()
+  *    Does NOT deallocate any remaining allocated memory.
+  */
+U_CFUNC UBool 
+cmemory_cleanup(void);
+
+/**
+ * A function called by <TT>uhash_remove</TT>,
+ * <TT>uhash_close</TT>, or <TT>uhash_put</TT> to delete
+ * an existing key or value.
+ * @param obj A key or value stored in a hashtable
+ * @see uprv_deleteUObject
+ */
+typedef void U_CALLCONV UObjectDeleter(void* obj);
+
+/**
+ * Deleter for UObject instances.
+ * Works for all subclasses of UObject because it has a virtual destructor.
+ */
+U_CAPI void U_EXPORT2
+uprv_deleteUObject(void *obj);
+
+#ifdef __cplusplus
+
+#include <utility>
+#include "unicode/uobject.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * "Smart pointer" class, deletes memory via uprv_free().
+ * For most methods see the LocalPointerBase base class.
+ * Adds operator[] for array item access.
+ *
+ * @see LocalPointerBase
+ */
+template<typename T>
+class LocalMemory : public LocalPointerBase<T> {
+public:
+    using LocalPointerBase<T>::operator*;
+    using LocalPointerBase<T>::operator->;
+    /**
+     * Constructor takes ownership.
+     * @param p simple pointer to an array of T items that is adopted
+     */
+    explicit LocalMemory(T *p=NULL) : LocalPointerBase<T>(p) {}
+    /**
+     * Move constructor, leaves src with isNull().
+     * @param src source smart pointer
+     */
+    LocalMemory(LocalMemory<T> &&src) U_NOEXCEPT : LocalPointerBase<T>(src.ptr) {
+        src.ptr=NULL;
+    }
+    /**
+     * Destructor deletes the memory it owns.
+     */
+    ~LocalMemory() {
+        uprv_free(LocalPointerBase<T>::ptr);
+    }
+    /**
+     * Move assignment operator, leaves src with isNull().
+     * The behavior is undefined if *this and src are the same object.
+     * @param src source smart pointer
+     * @return *this
+     */
+    LocalMemory<T> &operator=(LocalMemory<T> &&src) U_NOEXCEPT {
+        uprv_free(LocalPointerBase<T>::ptr);
+        LocalPointerBase<T>::ptr=src.ptr;
+        src.ptr=NULL;
+        return *this;
+    }
+    /**
+     * Swap pointers.
+     * @param other other smart pointer
+     */
+    void swap(LocalMemory<T> &other) U_NOEXCEPT {
+        T *temp=LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=other.ptr;
+        other.ptr=temp;
+    }
+    /**
+     * Non-member LocalMemory swap function.
+     * @param p1 will get p2's pointer
+     * @param p2 will get p1's pointer
+     */
+    friend inline void swap(LocalMemory<T> &p1, LocalMemory<T> &p2) U_NOEXCEPT {
+        p1.swap(p2);
+    }
+    /**
+     * Deletes the array it owns,
+     * and adopts (takes ownership of) the one passed in.
+     * @param p simple pointer to an array of T items that is adopted
+     */
+    void adoptInstead(T *p) {
+        uprv_free(LocalPointerBase<T>::ptr);
+        LocalPointerBase<T>::ptr=p;
+    }
+    /**
+     * Deletes the array it owns, allocates a new one and reset its bytes to 0.
+     * Returns the new array pointer.
+     * If the allocation fails, then the current array is unchanged and
+     * this method returns NULL.
+     * @param newCapacity must be >0
+     * @return the allocated array pointer, or NULL if the allocation failed
+     */
+    inline T *allocateInsteadAndReset(int32_t newCapacity=1);
+    /**
+     * Deletes the array it owns and allocates a new one, copying length T items.
+     * Returns the new array pointer.
+     * If the allocation fails, then the current array is unchanged and
+     * this method returns NULL.
+     * @param newCapacity must be >0
+     * @param length number of T items to be copied from the old array to the new one;
+     *               must be no more than the capacity of the old array,
+     *               which the caller must track because the LocalMemory does not track it
+     * @return the allocated array pointer, or NULL if the allocation failed
+     */
+    inline T *allocateInsteadAndCopy(int32_t newCapacity=1, int32_t length=0);
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    T &operator[](ptrdiff_t i) const { return LocalPointerBase<T>::ptr[i]; }
+};
+
+template<typename T>
+inline T *LocalMemory<T>::allocateInsteadAndReset(int32_t newCapacity) {
+    if(newCapacity>0) {
+        T *p=(T *)uprv_malloc(newCapacity*sizeof(T));
+        if(p!=NULL) {
+            uprv_memset(p, 0, newCapacity*sizeof(T));
+            uprv_free(LocalPointerBase<T>::ptr);
+            LocalPointerBase<T>::ptr=p;
+        }
+        return p;
+    } else {
+        return NULL;
+    }
+}
+
+
+template<typename T>
+inline T *LocalMemory<T>::allocateInsteadAndCopy(int32_t newCapacity, int32_t length) {
+    if(newCapacity>0) {
+        T *p=(T *)uprv_malloc(newCapacity*sizeof(T));
+        if(p!=NULL) {
+            if(length>0) {
+                if(length>newCapacity) {
+                    length=newCapacity;
+                }
+                uprv_memcpy(p, LocalPointerBase<T>::ptr, (size_t)length*sizeof(T));
+            }
+            uprv_free(LocalPointerBase<T>::ptr);
+            LocalPointerBase<T>::ptr=p;
+        }
+        return p;
+    } else {
+        return NULL;
+    }
+}
+
+/**
+ * Simple array/buffer management class using uprv_malloc() and uprv_free().
+ * Provides an internal array with fixed capacity. Can alias another array
+ * or allocate one.
+ *
+ * The array address is properly aligned for type T. It might not be properly
+ * aligned for types larger than T (or larger than the largest subtype of T).
+ *
+ * Unlike LocalMemory and LocalArray, this class never adopts
+ * (takes ownership of) another array.
+ *
+ * WARNING: MaybeStackArray only works with primitive (plain-old data) types.
+ * It does NOT know how to call a destructor! If you work with classes with
+ * destructors, consider:
+ *
+ * - LocalArray in localpointer.h if you know the length ahead of time
+ * - MaybeStackVector if you know the length at runtime
+ */
+template<typename T, int32_t stackCapacity>
+class MaybeStackArray {
+public:
+    // No heap allocation. Use only on the stack.
+    static void* U_EXPORT2 operator new(size_t) U_NOEXCEPT = delete;
+    static void* U_EXPORT2 operator new[](size_t) U_NOEXCEPT = delete;
+#if U_HAVE_PLACEMENT_NEW
+    static void* U_EXPORT2 operator new(size_t, void*) U_NOEXCEPT = delete;
+#endif
+
+    /**
+     * Default constructor initializes with internal T[stackCapacity] buffer.
+     */
+    MaybeStackArray() : ptr(stackArray), capacity(stackCapacity), needToRelease(false) {}
+    /**
+     * Automatically allocates the heap array if the argument is larger than the stack capacity.
+     * Intended for use when an approximate capacity is known at compile time but the true
+     * capacity is not known until runtime.
+     */
+    MaybeStackArray(int32_t newCapacity, UErrorCode status) : MaybeStackArray() {
+        if (U_FAILURE(status)) {
+            return;
+        }
+        if (capacity < newCapacity) {
+            if (resize(newCapacity) == nullptr) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+            }
+        }
+    }
+    /**
+     * Destructor deletes the array (if owned).
+     */
+    ~MaybeStackArray() { releaseArray(); }
+    /**
+     * Move constructor: transfers ownership or copies the stack array.
+     */
+    MaybeStackArray(MaybeStackArray<T, stackCapacity> &&src) U_NOEXCEPT;
+    /**
+     * Move assignment: transfers ownership or copies the stack array.
+     */
+    MaybeStackArray<T, stackCapacity> &operator=(MaybeStackArray<T, stackCapacity> &&src) U_NOEXCEPT;
+    /**
+     * Returns the array capacity (number of T items).
+     * @return array capacity
+     */
+    int32_t getCapacity() const { return capacity; }
+    /**
+     * Access without ownership change.
+     * @return the array pointer
+     */
+    T *getAlias() const { return ptr; }
+    /**
+     * Returns the array limit. Simple convenience method.
+     * @return getAlias()+getCapacity()
+     */
+    T *getArrayLimit() const { return getAlias()+capacity; }
+    // No "operator T *() const" because that can make
+    // expressions like mbs[index] ambiguous for some compilers.
+    /**
+     * Array item access (const).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    const T &operator[](ptrdiff_t i) const { return ptr[i]; }
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    T &operator[](ptrdiff_t i) { return ptr[i]; }
+    /**
+     * Deletes the array (if owned) and aliases another one, no transfer of ownership.
+     * If the arguments are illegal, then the current array is unchanged.
+     * @param otherArray must not be NULL
+     * @param otherCapacity must be >0
+     */
+    void aliasInstead(T *otherArray, int32_t otherCapacity) {
+        if(otherArray!=NULL && otherCapacity>0) {
+            releaseArray();
+            ptr=otherArray;
+            capacity=otherCapacity;
+            needToRelease=false;
+        }
+    }
+    /**
+     * Deletes the array (if owned) and allocates a new one, copying length T items.
+     * Returns the new array pointer.
+     * If the allocation fails, then the current array is unchanged and
+     * this method returns NULL.
+     * @param newCapacity can be less than or greater than the current capacity;
+     *                    must be >0
+     * @param length number of T items to be copied from the old array to the new one
+     * @return the allocated array pointer, or NULL if the allocation failed
+     */
+    inline T *resize(int32_t newCapacity, int32_t length=0);
+    /**
+     * Gives up ownership of the array if owned, or else clones it,
+     * copying length T items; resets itself to the internal stack array.
+     * Returns NULL if the allocation failed.
+     * @param length number of T items to copy when cloning,
+     *        and capacity of the clone when cloning
+     * @param resultCapacity will be set to the returned array's capacity (output-only)
+     * @return the array pointer;
+     *         caller becomes responsible for deleting the array
+     */
+    inline T *orphanOrClone(int32_t length, int32_t &resultCapacity);
+
+protected:
+    // Resizes the array to the size of src, then copies the contents of src.
+    void copyFrom(const MaybeStackArray &src, UErrorCode &status) {
+        if (U_FAILURE(status)) {
+            return;
+        }
+        if (this->resize(src.capacity, 0) == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        uprv_memcpy(this->ptr, src.ptr, (size_t)capacity * sizeof(T));
+    }
+
+private:
+    T *ptr;
+    int32_t capacity;
+    UBool needToRelease;
+    T stackArray[stackCapacity];
+    void releaseArray() {
+        if(needToRelease) {
+            uprv_free(ptr);
+        }
+    }
+    void resetToStackArray() {
+        ptr=stackArray;
+        capacity=stackCapacity;
+        needToRelease=false;
+    }
+    /* No comparison operators with other MaybeStackArray's. */
+    bool operator==(const MaybeStackArray & /*other*/) = delete;
+    bool operator!=(const MaybeStackArray & /*other*/) = delete;
+    /* No ownership transfer: No copy constructor, no assignment operator. */
+    MaybeStackArray(const MaybeStackArray & /*other*/) = delete;
+    void operator=(const MaybeStackArray & /*other*/) = delete;
+};
+
+template<typename T, int32_t stackCapacity>
+icu::MaybeStackArray<T, stackCapacity>::MaybeStackArray(
+        MaybeStackArray <T, stackCapacity>&& src) U_NOEXCEPT
+        : ptr(src.ptr), capacity(src.capacity), needToRelease(src.needToRelease) {
+    if (src.ptr == src.stackArray) {
+        ptr = stackArray;
+        uprv_memcpy(stackArray, src.stackArray, sizeof(T) * src.capacity);
+    } else {
+        src.resetToStackArray();  // take ownership away from src
+    }
+}
+
+template<typename T, int32_t stackCapacity>
+inline MaybeStackArray <T, stackCapacity>&
+MaybeStackArray<T, stackCapacity>::operator=(MaybeStackArray <T, stackCapacity>&& src) U_NOEXCEPT {
+    releaseArray();  // in case this instance had its own memory allocated
+    capacity = src.capacity;
+    needToRelease = src.needToRelease;
+    if (src.ptr == src.stackArray) {
+        ptr = stackArray;
+        uprv_memcpy(stackArray, src.stackArray, sizeof(T) * src.capacity);
+    } else {
+        ptr = src.ptr;
+        src.resetToStackArray();  // take ownership away from src
+    }
+    return *this;
+}
+
+template<typename T, int32_t stackCapacity>
+inline T *MaybeStackArray<T, stackCapacity>::resize(int32_t newCapacity, int32_t length) {
+    if(newCapacity>0) {
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+        ::fprintf(::stderr, "MaybeStackArray (resize) alloc %d * %lu\n", newCapacity, sizeof(T));
+#endif
+        T *p=(T *)uprv_malloc(newCapacity*sizeof(T));
+        if(p!=NULL) {
+            if(length>0) {
+                if(length>capacity) {
+                    length=capacity;
+                }
+                if(length>newCapacity) {
+                    length=newCapacity;
+                }
+                uprv_memcpy(p, ptr, (size_t)length*sizeof(T));
+            }
+            releaseArray();
+            ptr=p;
+            capacity=newCapacity;
+            needToRelease=true;
+        }
+        return p;
+    } else {
+        return NULL;
+    }
+}
+
+template<typename T, int32_t stackCapacity>
+inline T *MaybeStackArray<T, stackCapacity>::orphanOrClone(int32_t length, int32_t &resultCapacity) {
+    T *p;
+    if(needToRelease) {
+        p=ptr;
+    } else if(length<=0) {
+        return NULL;
+    } else {
+        if(length>capacity) {
+            length=capacity;
+        }
+        p=(T *)uprv_malloc(length*sizeof(T));
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+      ::fprintf(::stderr,"MaybeStacArray (orphan) alloc %d * %lu\n", length,sizeof(T));
+#endif
+        if(p==NULL) {
+            return NULL;
+        }
+        uprv_memcpy(p, ptr, (size_t)length*sizeof(T));
+    }
+    resultCapacity=length;
+    resetToStackArray();
+    return p;
+}
+
+/**
+ * Variant of MaybeStackArray that allocates a header struct and an array
+ * in one contiguous memory block, using uprv_malloc() and uprv_free().
+ * Provides internal memory with fixed array capacity. Can alias another memory
+ * block or allocate one.
+ * The stackCapacity is the number of T items in the internal memory,
+ * not counting the H header.
+ * Unlike LocalMemory and LocalArray, this class never adopts
+ * (takes ownership of) another memory block.
+ */
+template<typename H, typename T, int32_t stackCapacity>
+class MaybeStackHeaderAndArray {
+public:
+    // No heap allocation. Use only on the stack.
+    static void* U_EXPORT2 operator new(size_t) U_NOEXCEPT = delete;
+    static void* U_EXPORT2 operator new[](size_t) U_NOEXCEPT = delete;
+#if U_HAVE_PLACEMENT_NEW
+    static void* U_EXPORT2 operator new(size_t, void*) U_NOEXCEPT = delete;
+#endif
+
+    /**
+     * Default constructor initializes with internal H+T[stackCapacity] buffer.
+     */
+    MaybeStackHeaderAndArray() : ptr(&stackHeader), capacity(stackCapacity), needToRelease(false) {}
+    /**
+     * Destructor deletes the memory (if owned).
+     */
+    ~MaybeStackHeaderAndArray() { releaseMemory(); }
+    /**
+     * Returns the array capacity (number of T items).
+     * @return array capacity
+     */
+    int32_t getCapacity() const { return capacity; }
+    /**
+     * Access without ownership change.
+     * @return the header pointer
+     */
+    H *getAlias() const { return ptr; }
+    /**
+     * Returns the array start.
+     * @return array start, same address as getAlias()+1
+     */
+    T *getArrayStart() const { return reinterpret_cast<T *>(getAlias()+1); }
+    /**
+     * Returns the array limit.
+     * @return array limit
+     */
+    T *getArrayLimit() const { return getArrayStart()+capacity; }
+    /**
+     * Access without ownership change. Same as getAlias().
+     * A class instance can be used directly in expressions that take a T *.
+     * @return the header pointer
+     */
+    operator H *() const { return ptr; }
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    T &operator[](ptrdiff_t i) { return getArrayStart()[i]; }
+    /**
+     * Deletes the memory block (if owned) and aliases another one, no transfer of ownership.
+     * If the arguments are illegal, then the current memory is unchanged.
+     * @param otherArray must not be NULL
+     * @param otherCapacity must be >0
+     */
+    void aliasInstead(H *otherMemory, int32_t otherCapacity) {
+        if(otherMemory!=NULL && otherCapacity>0) {
+            releaseMemory();
+            ptr=otherMemory;
+            capacity=otherCapacity;
+            needToRelease=false;
+        }
+    }
+    /**
+     * Deletes the memory block (if owned) and allocates a new one,
+     * copying the header and length T array items.
+     * Returns the new header pointer.
+     * If the allocation fails, then the current memory is unchanged and
+     * this method returns NULL.
+     * @param newCapacity can be less than or greater than the current capacity;
+     *                    must be >0
+     * @param length number of T items to be copied from the old array to the new one
+     * @return the allocated pointer, or NULL if the allocation failed
+     */
+    inline H *resize(int32_t newCapacity, int32_t length=0);
+    /**
+     * Gives up ownership of the memory if owned, or else clones it,
+     * copying the header and length T array items; resets itself to the internal memory.
+     * Returns NULL if the allocation failed.
+     * @param length number of T items to copy when cloning,
+     *        and array capacity of the clone when cloning
+     * @param resultCapacity will be set to the returned array's capacity (output-only)
+     * @return the header pointer;
+     *         caller becomes responsible for deleting the array
+     */
+    inline H *orphanOrClone(int32_t length, int32_t &resultCapacity);
+private:
+    H *ptr;
+    int32_t capacity;
+    UBool needToRelease;
+    // stackHeader must precede stackArray immediately.
+    H stackHeader;
+    T stackArray[stackCapacity];
+    void releaseMemory() {
+        if(needToRelease) {
+            uprv_free(ptr);
+        }
+    }
+    /* No comparison operators with other MaybeStackHeaderAndArray's. */
+    bool operator==(const MaybeStackHeaderAndArray & /*other*/) {return false;}
+    bool operator!=(const MaybeStackHeaderAndArray & /*other*/) {return true;}
+    /* No ownership transfer: No copy constructor, no assignment operator. */
+    MaybeStackHeaderAndArray(const MaybeStackHeaderAndArray & /*other*/) {}
+    void operator=(const MaybeStackHeaderAndArray & /*other*/) {}
+};
+
+template<typename H, typename T, int32_t stackCapacity>
+inline H *MaybeStackHeaderAndArray<H, T, stackCapacity>::resize(int32_t newCapacity,
+                                                                int32_t length) {
+    if(newCapacity>=0) {
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+      ::fprintf(::stderr,"MaybeStackHeaderAndArray alloc %d + %d * %ul\n", sizeof(H),newCapacity,sizeof(T));
+#endif
+        H *p=(H *)uprv_malloc(sizeof(H)+newCapacity*sizeof(T));
+        if(p!=NULL) {
+            if(length<0) {
+                length=0;
+            } else if(length>0) {
+                if(length>capacity) {
+                    length=capacity;
+                }
+                if(length>newCapacity) {
+                    length=newCapacity;
+                }
+            }
+            uprv_memcpy(p, ptr, sizeof(H)+(size_t)length*sizeof(T));
+            releaseMemory();
+            ptr=p;
+            capacity=newCapacity;
+            needToRelease=true;
+        }
+        return p;
+    } else {
+        return NULL;
+    }
+}
+
+template<typename H, typename T, int32_t stackCapacity>
+inline H *MaybeStackHeaderAndArray<H, T, stackCapacity>::orphanOrClone(int32_t length,
+                                                                       int32_t &resultCapacity) {
+    H *p;
+    if(needToRelease) {
+        p=ptr;
+    } else {
+        if(length<0) {
+            length=0;
+        } else if(length>capacity) {
+            length=capacity;
+        }
+#if U_DEBUG && defined(UPRV_MALLOC_COUNT)
+      ::fprintf(::stderr,"MaybeStackHeaderAndArray (orphan) alloc %ul + %d * %lu\n", sizeof(H),length,sizeof(T));
+#endif
+        p=(H *)uprv_malloc(sizeof(H)+length*sizeof(T));
+        if(p==NULL) {
+            return NULL;
+        }
+        uprv_memcpy(p, ptr, sizeof(H)+(size_t)length*sizeof(T));
+    }
+    resultCapacity=length;
+    ptr=&stackHeader;
+    capacity=stackCapacity;
+    needToRelease=false;
+    return p;
+}
+
+/**
+ * A simple memory management class that creates new heap allocated objects (of
+ * any class that has a public constructor), keeps track of them and eventually
+ * deletes them all in its own destructor.
+ *
+ * A typical use-case would be code like this:
+ *
+ *     MemoryPool<MyType> pool;
+ *
+ *     MyType* o1 = pool.create();
+ *     if (o1 != nullptr) {
+ *         foo(o1);
+ *     }
+ *
+ *     MyType* o2 = pool.create(1, 2, 3);
+ *     if (o2 != nullptr) {
+ *         bar(o2);
+ *     }
+ *
+ *     // MemoryPool will take care of deleting the MyType objects.
+ *
+ * It doesn't do anything more than that, and is intentionally kept minimalist.
+ */
+template<typename T, int32_t stackCapacity = 8>
+class MemoryPool : public UMemory {
+public:
+    MemoryPool() : fCount(0), fPool() {}
+
+    ~MemoryPool() {
+        for (int32_t i = 0; i < fCount; ++i) {
+            delete fPool[i];
+        }
+    }
+
+    MemoryPool(const MemoryPool&) = delete;
+    MemoryPool& operator=(const MemoryPool&) = delete;
+
+    MemoryPool(MemoryPool&& other) U_NOEXCEPT : fCount(other.fCount),
+                                                fPool(std::move(other.fPool)) {
+        other.fCount = 0;
+    }
+
+    MemoryPool& operator=(MemoryPool&& other) U_NOEXCEPT {
+        fCount = other.fCount;
+        fPool = std::move(other.fPool);
+        other.fCount = 0;
+        return *this;
+    }
+
+    /**
+     * Creates a new object of typename T, by forwarding any and all arguments
+     * to the typename T constructor.
+     *
+     * @param args Arguments to be forwarded to the typename T constructor.
+     * @return A pointer to the newly created object, or nullptr on error.
+     */
+    template<typename... Args>
+    T* create(Args&&... args) {
+        int32_t capacity = fPool.getCapacity();
+        if (fCount == capacity &&
+            fPool.resize(capacity == stackCapacity ? 4 * capacity : 2 * capacity,
+                         capacity) == nullptr) {
+            return nullptr;
+        }
+        return fPool[fCount++] = new T(std::forward<Args>(args)...);
+    }
+
+    template <typename... Args>
+    T* createAndCheckErrorCode(UErrorCode &status, Args &&... args) {
+        if (U_FAILURE(status)) {
+            return nullptr;
+        }
+        T *pointer = this->create(args...);
+        if (U_SUCCESS(status) && pointer == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+        }
+        return pointer;
+    }
+
+    /**
+     * @return Number of elements that have been allocated.
+     */
+    int32_t count() const {
+        return fCount;
+    }
+
+protected:
+    int32_t fCount;
+    MaybeStackArray<T*, stackCapacity> fPool;
+};
+
+/**
+ * An internal Vector-like implementation based on MemoryPool.
+ *
+ * Heap-allocates each element and stores pointers.
+ *
+ * To append an item to the vector, use emplaceBack.
+ *
+ *     MaybeStackVector<MyType> vector;
+ *     MyType* element = vector.emplaceBack();
+ *     if (!element) {
+ *         status = U_MEMORY_ALLOCATION_ERROR;
+ *     }
+ *     // do stuff with element
+ *
+ * To loop over the vector, use a for loop with indices:
+ *
+ *     for (int32_t i = 0; i < vector.length(); i++) {
+ *         MyType* element = vector[i];
+ *     }
+ */
+template<typename T, int32_t stackCapacity = 8>
+class MaybeStackVector : protected MemoryPool<T, stackCapacity> {
+public:
+    using MemoryPool<T, stackCapacity>::MemoryPool;
+    using MemoryPool<T, stackCapacity>::operator=;
+
+    template<typename... Args>
+    T* emplaceBack(Args&&... args) {
+        return this->create(args...);
+    }
+
+    template <typename... Args>
+    T *emplaceBackAndCheckErrorCode(UErrorCode &status, Args &&... args) {
+        return this->createAndCheckErrorCode(status, args...);
+    }
+
+    int32_t length() const {
+        return this->fCount;
+    }
+
+    T** getAlias() {
+        return this->fPool.getAlias();
+    }
+
+    const T *const *getAlias() const {
+        return this->fPool.getAlias();
+    }
+
+    /**
+     * Array item access (read-only).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    const T* operator[](ptrdiff_t i) const {
+        return this->fPool[i];
+    }
+
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    T* operator[](ptrdiff_t i) {
+        return this->fPool[i];
+    }
+};
+
+
+U_NAMESPACE_END
+
+#endif  /* __cplusplus */
+#endif  /* CMEMORY_H */
diff --git a/thirdparty/icu4c/common/cpputils.h b/thirdparty/icu4c/common/cpputils.h
new file mode 100644
index 0000000000..307e570486
--- /dev/null
+++ b/thirdparty/icu4c/common/cpputils.h
@@ -0,0 +1,97 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  cpputils.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*/
+
+#ifndef CPPUTILS_H
+#define CPPUTILS_H
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "cmemory.h"
+
+/*==========================================================================*/
+/* Array copy utility functions */
+/*==========================================================================*/
+
+static
+inline void uprv_arrayCopy(const double* src, double* dst, int32_t count)
+{ uprv_memcpy(dst, src, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const double* src, int32_t srcStart,
+              double* dst, int32_t dstStart, int32_t count)
+{ uprv_memcpy(dst+dstStart, src+srcStart, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int8_t* src, int8_t* dst, int32_t count)
+    { uprv_memcpy(dst, src, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int8_t* src, int32_t srcStart,
+              int8_t* dst, int32_t dstStart, int32_t count)
+{ uprv_memcpy(dst+dstStart, src+srcStart, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int16_t* src, int16_t* dst, int32_t count)
+{ uprv_memcpy(dst, src, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int16_t* src, int32_t srcStart,
+              int16_t* dst, int32_t dstStart, int32_t count)
+{ uprv_memcpy(dst+dstStart, src+srcStart, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int32_t* src, int32_t* dst, int32_t count)
+{ uprv_memcpy(dst, src, (size_t)count * sizeof(*src)); }
+
+static
+inline void uprv_arrayCopy(const int32_t* src, int32_t srcStart,
+              int32_t* dst, int32_t dstStart, int32_t count)
+{ uprv_memcpy(dst+dstStart, src+srcStart, (size_t)count * sizeof(*src)); }
+
+static
+inline void
+uprv_arrayCopy(const UChar *src, int32_t srcStart,
+        UChar *dst, int32_t dstStart, int32_t count)
+{ uprv_memcpy(dst+dstStart, src+srcStart, (size_t)count * sizeof(*src)); }
+
+/**
+ * Copy an array of UnicodeString OBJECTS (not pointers).
+ * @internal
+ */
+static inline void
+uprv_arrayCopy(const icu::UnicodeString *src, icu::UnicodeString *dst, int32_t count)
+{ while(count-- > 0) *dst++ = *src++; }
+
+/**
+ * Copy an array of UnicodeString OBJECTS (not pointers).
+ * @internal
+ */
+static inline void
+uprv_arrayCopy(const icu::UnicodeString *src, int32_t srcStart,
+               icu::UnicodeString *dst, int32_t dstStart, int32_t count)
+{ uprv_arrayCopy(src+srcStart, dst+dstStart, count); }
+
+/**
+ * Checks that the string is readable and writable.
+ * Sets U_ILLEGAL_ARGUMENT_ERROR if the string isBogus() or has an open getBuffer().
+ */
+inline void
+uprv_checkCanGetBuffer(const icu::UnicodeString &s, UErrorCode &errorCode) {
+    if(U_SUCCESS(errorCode) && s.isBogus()) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+#endif /* _CPPUTILS */
diff --git a/thirdparty/icu4c/common/cstr.cpp b/thirdparty/icu4c/common/cstr.cpp
new file mode 100644
index 0000000000..24654f8fc2
--- /dev/null
+++ b/thirdparty/icu4c/common/cstr.cpp
@@ -0,0 +1,54 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2015-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  charstr.cpp
+*/
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/unistr.h"
+
+#include "cstr.h"
+
+#include "charstr.h"
+#include "uinvchar.h"
+
+U_NAMESPACE_BEGIN
+
+CStr::CStr(const UnicodeString &in) {
+    UErrorCode status = U_ZERO_ERROR;
+#if !UCONFIG_NO_CONVERSION || U_CHARSET_IS_UTF8
+    int32_t length = in.extract(0, in.length(), static_cast<char *>(NULL), static_cast<uint32_t>(0));
+    int32_t resultCapacity = 0;
+    char *buf = s.getAppendBuffer(length, length, resultCapacity, status);
+    if (U_SUCCESS(status)) {
+        in.extract(0, in.length(), buf, resultCapacity);
+        s.append(buf, length, status);
+    }
+#else
+    // No conversion available. Convert any invariant characters; substitute '?' for the rest.
+    // Note: can't just call u_UCharsToChars() or CharString.appendInvariantChars() on the 
+    //       whole string because they require that the entire input be invariant.
+    char buf[2];
+    for (int i=0; i<in.length(); i = in.moveIndex32(i, 1)) {
+        if (uprv_isInvariantUString(in.getBuffer()+i, 1)) {
+            u_UCharsToChars(in.getBuffer()+i, buf, 1);
+        } else {
+            buf[0] = '?';
+        }
+        s.append(buf, 1, status);
+    }
+#endif
+}
+
+CStr::~CStr() {
+}
+
+const char * CStr::operator ()() const {
+    return s.data();
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/cstr.h b/thirdparty/icu4c/common/cstr.h
new file mode 100644
index 0000000000..c33f487ea1
--- /dev/null
+++ b/thirdparty/icu4c/common/cstr.h
@@ -0,0 +1,60 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File: cstr.h
+*/
+
+#ifndef CSTR_H
+#define CSTR_H
+
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+#include "unicode/utypes.h"
+
+#include "charstr.h"
+
+/**
+ * ICU-internal class CStr, a small helper class to facilitate passing UnicodeStrings
+ * to functions needing (const char *) strings, such as printf().
+ *
+ * It is intended primarily for use in debugging or in tests. Uses platform 
+ * default code page conversion, which will do the best job possible,
+ * but may be lossy, depending on the platform.
+ *
+ * If no other conversion is available, use invariant conversion and substitue
+ * '?' for non-invariant characters.
+ *
+ * Example Usage:
+ *   UnicodeString s = whatever;
+ *   printf("%s", CStr(s)());
+ *
+ *   The explicit call to the CStr() constructor creates a temporary object.
+ *   Operator () on the temporary object returns a (const char *) pointer.
+ *   The lifetime of the (const char *) data is that of the temporary object,
+ *   which works well when passing it as a parameter to another function, such as printf.
+ */
+
+U_NAMESPACE_BEGIN
+
+class U_COMMON_API CStr : public UMemory {
+  public:
+    CStr(const UnicodeString &in);
+    ~CStr();
+    const char * operator ()() const;
+
+  private:
+    CharString s;
+    CStr(const CStr &other);               //  Forbid copying of this class.
+    CStr &operator =(const CStr &other);   //  Forbid assignment.
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/cstring.cpp b/thirdparty/icu4c/common/cstring.cpp
new file mode 100644
index 0000000000..06275c4b56
--- /dev/null
+++ b/thirdparty/icu4c/common/cstring.cpp
@@ -0,0 +1,341 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File CSTRING.C
+*
+* @author       Helena Shih
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   6/18/98     hshih       Created
+*   09/08/98    stephen     Added include for ctype, for Mac Port
+*   11/15/99    helena      Integrated S/390 IEEE changes. 
+******************************************************************************
+*/
+
+
+
+#include <stdlib.h>
+#include <stdio.h>
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+
+/*
+ * We hardcode case conversion for invariant characters to match our expectation
+ * and the compiler execution charset.
+ * This prevents problems on systems
+ * - with non-default casing behavior, like Turkish system locales where
+ *   tolower('I') maps to dotless i and toupper('i') maps to dotted I
+ * - where there are no lowercase Latin characters at all, or using different
+ *   codes (some old EBCDIC codepages)
+ *
+ * This works because the compiler usually runs on a platform where the execution
+ * charset includes all of the invariant characters at their expected
+ * code positions, so that the char * string literals in ICU code match
+ * the char literals here.
+ *
+ * Note that the set of lowercase Latin letters is discontiguous in EBCDIC
+ * and the set of uppercase Latin letters is discontiguous as well.
+ */
+
+U_CAPI UBool U_EXPORT2
+uprv_isASCIILetter(char c) {
+#if U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    return
+        ('a'<=c && c<='i') || ('j'<=c && c<='r') || ('s'<=c && c<='z') ||
+        ('A'<=c && c<='I') || ('J'<=c && c<='R') || ('S'<=c && c<='Z');
+#else
+    return ('a'<=c && c<='z') || ('A'<=c && c<='Z');
+#endif
+}
+
+U_CAPI char U_EXPORT2
+uprv_toupper(char c) {
+#if U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    if(('a'<=c && c<='i') || ('j'<=c && c<='r') || ('s'<=c && c<='z')) {
+        c=(char)(c+('A'-'a'));
+    }
+#else
+    if('a'<=c && c<='z') {
+        c=(char)(c+('A'-'a'));
+    }
+#endif
+    return c;
+}
+
+
+#if 0
+/*
+ * Commented out because cstring.h defines uprv_tolower() to be
+ * the same as either uprv_asciitolower() or uprv_ebcdictolower()
+ * to reduce the amount of code to cover with tests.
+ *
+ * Note that this uprv_tolower() definition is likely to work for most
+ * charset families, not just ASCII and EBCDIC, because its #else branch
+ * is written generically.
+ */
+U_CAPI char U_EXPORT2
+uprv_tolower(char c) {
+#if U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    if(('A'<=c && c<='I') || ('J'<=c && c<='R') || ('S'<=c && c<='Z')) {
+        c=(char)(c+('a'-'A'));
+    }
+#else
+    if('A'<=c && c<='Z') {
+        c=(char)(c+('a'-'A'));
+    }
+#endif
+    return c;
+}
+#endif
+
+U_CAPI char U_EXPORT2
+uprv_asciitolower(char c) {
+    if(0x41<=c && c<=0x5a) {
+        c=(char)(c+0x20);
+    }
+    return c;
+}
+
+U_CAPI char U_EXPORT2
+uprv_ebcdictolower(char c) {
+    if( (0xc1<=(uint8_t)c && (uint8_t)c<=0xc9) ||
+        (0xd1<=(uint8_t)c && (uint8_t)c<=0xd9) ||
+        (0xe2<=(uint8_t)c && (uint8_t)c<=0xe9)
+    ) {
+        c=(char)(c-0x40);
+    }
+    return c;
+}
+
+
+U_CAPI char* U_EXPORT2
+T_CString_toLowerCase(char* str)
+{
+    char* origPtr = str;
+
+    if (str) {
+        do
+            *str = (char)uprv_tolower(*str);
+        while (*(str++));
+    }
+
+    return origPtr;
+}
+
+U_CAPI char* U_EXPORT2
+T_CString_toUpperCase(char* str)
+{
+    char* origPtr = str;
+
+    if (str) {
+        do
+            *str = (char)uprv_toupper(*str);
+        while (*(str++));
+    }
+
+    return origPtr;
+}
+
+/*
+ * Takes a int32_t and fills in  a char* string with that number "radix"-based.
+ * Does not handle negative values (makes an empty string for them).
+ * Writes at most 12 chars ("-2147483647" plus NUL).
+ * Returns the length of the string (not including the NUL).
+ */
+U_CAPI int32_t U_EXPORT2
+T_CString_integerToString(char* buffer, int32_t v, int32_t radix)
+{
+    char      tbuf[30];
+    int32_t   tbx    = sizeof(tbuf);
+    uint8_t   digit;
+    int32_t   length = 0;
+    uint32_t  uval;
+    
+    U_ASSERT(radix>=2 && radix<=16);
+    uval = (uint32_t) v;
+    if(v<0 && radix == 10) {
+        /* Only in base 10 do we conside numbers to be signed. */
+        uval = (uint32_t)(-v); 
+        buffer[length++] = '-';
+    }
+    
+    tbx = sizeof(tbuf)-1;
+    tbuf[tbx] = 0;   /* We are generating the digits backwards.  Null term the end. */
+    do {
+        digit = (uint8_t)(uval % radix);
+        tbuf[--tbx] = (char)(T_CString_itosOffset(digit));
+        uval  = uval / radix;
+    } while (uval != 0);
+    
+    /* copy converted number into user buffer  */
+    uprv_strcpy(buffer+length, tbuf+tbx);
+    length += sizeof(tbuf) - tbx -1;
+    return length;
+}
+
+
+
+/*
+ * Takes a int64_t and fills in  a char* string with that number "radix"-based.
+ * Writes at most 21: chars ("-9223372036854775807" plus NUL).
+ * Returns the length of the string, not including the terminating NULL.
+ */
+U_CAPI int32_t U_EXPORT2
+T_CString_int64ToString(char* buffer, int64_t v, uint32_t radix)
+{
+    char      tbuf[30];
+    int32_t   tbx    = sizeof(tbuf);
+    uint8_t   digit;
+    int32_t   length = 0;
+    uint64_t  uval;
+    
+    U_ASSERT(radix>=2 && radix<=16);
+    uval = (uint64_t) v;
+    if(v<0 && radix == 10) {
+        /* Only in base 10 do we conside numbers to be signed. */
+        uval = (uint64_t)(-v); 
+        buffer[length++] = '-';
+    }
+    
+    tbx = sizeof(tbuf)-1;
+    tbuf[tbx] = 0;   /* We are generating the digits backwards.  Null term the end. */
+    do {
+        digit = (uint8_t)(uval % radix);
+        tbuf[--tbx] = (char)(T_CString_itosOffset(digit));
+        uval  = uval / radix;
+    } while (uval != 0);
+    
+    /* copy converted number into user buffer  */
+    uprv_strcpy(buffer+length, tbuf+tbx);
+    length += sizeof(tbuf) - tbx -1;
+    return length;
+}
+
+
+U_CAPI int32_t U_EXPORT2
+T_CString_stringToInteger(const char *integerString, int32_t radix)
+{
+    char *end;
+    return uprv_strtoul(integerString, &end, radix);
+
+}
+
+U_CAPI int U_EXPORT2
+uprv_stricmp(const char *str1, const char *str2) {
+    if(str1==NULL) {
+        if(str2==NULL) {
+            return 0;
+        } else {
+            return -1;
+        }
+    } else if(str2==NULL) {
+        return 1;
+    } else {
+        /* compare non-NULL strings lexically with lowercase */
+        int rc;
+        unsigned char c1, c2;
+
+        for(;;) {
+            c1=(unsigned char)*str1;
+            c2=(unsigned char)*str2;
+            if(c1==0) {
+                if(c2==0) {
+                    return 0;
+                } else {
+                    return -1;
+                }
+            } else if(c2==0) {
+                return 1;
+            } else {
+                /* compare non-zero characters with lowercase */
+                rc=(int)(unsigned char)uprv_tolower(c1)-(int)(unsigned char)uprv_tolower(c2);
+                if(rc!=0) {
+                    return rc;
+                }
+            }
+            ++str1;
+            ++str2;
+        }
+    }
+}
+
+U_CAPI int U_EXPORT2
+uprv_strnicmp(const char *str1, const char *str2, uint32_t n) {
+    if(str1==NULL) {
+        if(str2==NULL) {
+            return 0;
+        } else {
+            return -1;
+        }
+    } else if(str2==NULL) {
+        return 1;
+    } else {
+        /* compare non-NULL strings lexically with lowercase */
+        int rc;
+        unsigned char c1, c2;
+
+        for(; n--;) {
+            c1=(unsigned char)*str1;
+            c2=(unsigned char)*str2;
+            if(c1==0) {
+                if(c2==0) {
+                    return 0;
+                } else {
+                    return -1;
+                }
+            } else if(c2==0) {
+                return 1;
+            } else {
+                /* compare non-zero characters with lowercase */
+                rc=(int)(unsigned char)uprv_tolower(c1)-(int)(unsigned char)uprv_tolower(c2);
+                if(rc!=0) {
+                    return rc;
+                }
+            }
+            ++str1;
+            ++str2;
+        }
+    }
+
+    return 0;
+}
+
+U_CAPI char* U_EXPORT2
+uprv_strdup(const char *src) {
+    size_t len = uprv_strlen(src) + 1;
+    char *dup = (char *) uprv_malloc(len);
+
+    if (dup) {
+        uprv_memcpy(dup, src, len);
+    }
+
+    return dup;
+}
+
+U_CAPI char* U_EXPORT2
+uprv_strndup(const char *src, int32_t n) {
+    char *dup;
+
+    if(n < 0) {
+        dup = uprv_strdup(src);
+    } else {
+        dup = (char*)uprv_malloc(n+1);
+        if (dup) { 
+            uprv_memcpy(dup, src, n);
+            dup[n] = 0;
+        }
+    }
+
+    return dup;
+}
diff --git a/thirdparty/icu4c/common/cstring.h b/thirdparty/icu4c/common/cstring.h
new file mode 100644
index 0000000000..3a14e4216c
--- /dev/null
+++ b/thirdparty/icu4c/common/cstring.h
@@ -0,0 +1,126 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File CSTRING.H
+*
+* Contains CString interface
+*
+* @author       Helena Shih
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   6/17/98     hshih       Created.
+*  05/03/99     stephen     Changed from functions to macros.
+*  06/14/99     stephen     Added icu_strncat, icu_strncmp, icu_tolower
+*
+******************************************************************************
+*/
+
+#ifndef CSTRING_H
+#define CSTRING_H 1
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include <string.h>
+#include <stdlib.h>
+#include <ctype.h>
+
+#define uprv_strcpy(dst, src) U_STANDARD_CPP_NAMESPACE  strcpy(dst, src)
+#define uprv_strlen(str) U_STANDARD_CPP_NAMESPACE strlen(str)
+#define uprv_strcmp(s1, s2) U_STANDARD_CPP_NAMESPACE strcmp(s1, s2)
+#define uprv_strcat(dst, src) U_STANDARD_CPP_NAMESPACE strcat(dst, src)
+#define uprv_strchr(s, c) U_STANDARD_CPP_NAMESPACE strchr(s, c)
+#define uprv_strstr(s, c) U_STANDARD_CPP_NAMESPACE strstr(s, c)
+#define uprv_strrchr(s, c) U_STANDARD_CPP_NAMESPACE strrchr(s, c)
+#define uprv_strncpy(dst, src, size) U_STANDARD_CPP_NAMESPACE strncpy(dst, src, size)
+#define uprv_strncmp(s1, s2, n) U_STANDARD_CPP_NAMESPACE strncmp(s1, s2, n)
+#define uprv_strncat(dst, src, n) U_STANDARD_CPP_NAMESPACE strncat(dst, src, n)
+
+/**
+ * Is c an ASCII-repertoire letter a-z or A-Z?
+ * Note: The implementation is specific to whether ICU is compiled for
+ * an ASCII-based or EBCDIC-based machine. There just does not seem to be a better name for this.
+ */
+U_CAPI UBool U_EXPORT2
+uprv_isASCIILetter(char c);
+
+// NOTE: For u_asciiToUpper that takes a UChar, see ustr_imp.h
+
+U_CAPI char U_EXPORT2
+uprv_toupper(char c);
+
+
+U_CAPI char U_EXPORT2
+uprv_asciitolower(char c);
+
+U_CAPI char U_EXPORT2
+uprv_ebcdictolower(char c);
+
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_tolower uprv_asciitolower
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_tolower uprv_ebcdictolower
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+
+#define uprv_strtod(source, end) U_STANDARD_CPP_NAMESPACE strtod(source, end)
+#define uprv_strtoul(str, end, base) U_STANDARD_CPP_NAMESPACE strtoul(str, end, base)
+#define uprv_strtol(str, end, base) U_STANDARD_CPP_NAMESPACE strtol(str, end, base)
+
+/* Conversion from a digit to the character with radix base from 2-19 */
+/* May need to use U_UPPER_ORDINAL*/
+#define T_CString_itosOffset(a) ((a)<=9?('0'+(a)):('A'+(a)-10))
+
+U_CAPI char* U_EXPORT2
+uprv_strdup(const char *src);
+
+/**
+ * uprv_malloc n+1 bytes, and copy n bytes from src into the new string.
+ * Terminate with a null at offset n.   If n is -1, works like uprv_strdup
+ * @param src
+ * @param n length of the input string, not including null.
+ * @return new string (owned by caller, use uprv_free to free).
+ * @internal
+ */
+U_CAPI char* U_EXPORT2
+uprv_strndup(const char *src, int32_t n);
+
+U_CAPI char* U_EXPORT2
+T_CString_toLowerCase(char* str);
+
+U_CAPI char* U_EXPORT2
+T_CString_toUpperCase(char* str);
+
+U_CAPI int32_t U_EXPORT2
+T_CString_integerToString(char *buffer, int32_t n, int32_t radix);
+
+U_CAPI int32_t U_EXPORT2
+T_CString_int64ToString(char *buffer, int64_t n, uint32_t radix);
+
+U_CAPI int32_t U_EXPORT2
+T_CString_stringToInteger(const char *integerString, int32_t radix);
+
+/**
+ * Case-insensitive, language-independent string comparison
+ * limited to the ASCII character repertoire.
+ */
+U_CAPI int U_EXPORT2
+uprv_stricmp(const char *str1, const char *str2);
+
+/**
+ * Case-insensitive, language-independent string comparison
+ * limited to the ASCII character repertoire.
+ */
+U_CAPI int U_EXPORT2
+uprv_strnicmp(const char *str1, const char *str2, uint32_t n);
+
+#endif /* ! CSTRING_H */
diff --git a/thirdparty/icu4c/common/cwchar.cpp b/thirdparty/icu4c/common/cwchar.cpp
new file mode 100644
index 0000000000..20c7d71e0f
--- /dev/null
+++ b/thirdparty/icu4c/common/cwchar.cpp
@@ -0,0 +1,55 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+******************************************************************************
+*
+*   Copyright (C) 2001, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  cwchar.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001may25
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !U_HAVE_WCSCPY
+
+#include "cwchar.h"
+
+U_CAPI wchar_t *uprv_wcscat(wchar_t *dst, const wchar_t *src) {
+    wchar_t *start=dst;
+    while(*dst!=0) {
+        ++dst;
+    }
+    while((*dst=*src)!=0) {
+        ++dst;
+        ++src;
+    }
+    return start;
+}
+
+U_CAPI wchar_t *uprv_wcscpy(wchar_t *dst, const wchar_t *src) {
+    wchar_t *start=dst;
+    while((*dst=*src)!=0) {
+        ++dst;
+        ++src;
+    }
+    return start;
+}
+
+U_CAPI size_t uprv_wcslen(const wchar_t *src) {
+    const wchar_t *start=src;
+    while(*src!=0) {
+        ++src;
+    }
+    return src-start;
+}
+
+#endif
+
diff --git a/thirdparty/icu4c/common/cwchar.h b/thirdparty/icu4c/common/cwchar.h
new file mode 100644
index 0000000000..8fd041a1b9
--- /dev/null
+++ b/thirdparty/icu4c/common/cwchar.h
@@ -0,0 +1,58 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+******************************************************************************
+*
+*   Copyright (C) 2001, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  cwchar.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001may25
+*   created by: Markus W. Scherer
+*
+*   This file contains ICU-internal definitions of wchar_t operations.
+*   These definitions were moved here from cstring.h so that fewer
+*   ICU implementation files include wchar.h.
+*/
+
+#ifndef __CWCHAR_H__
+#define __CWCHAR_H__
+
+#include <string.h>
+#include <stdlib.h>
+#include "unicode/utypes.h"
+
+/* Do this after utypes.h so that we have U_HAVE_WCHAR_H . */
+#if U_HAVE_WCHAR_H
+#   include <wchar.h>
+#endif
+
+/*===========================================================================*/
+/* Wide-character functions                                                  */
+/*===========================================================================*/
+
+/* The following are not available on all systems, defined in wchar.h or string.h. */
+#if U_HAVE_WCSCPY
+#   define uprv_wcscpy wcscpy
+#   define uprv_wcscat wcscat
+#   define uprv_wcslen wcslen
+#else
+U_CAPI wchar_t* U_EXPORT2 
+uprv_wcscpy(wchar_t *dst, const wchar_t *src);
+U_CAPI wchar_t* U_EXPORT2 
+uprv_wcscat(wchar_t *dst, const wchar_t *src);
+U_CAPI size_t U_EXPORT2 
+uprv_wcslen(const wchar_t *src);
+#endif
+
+/* The following are part of the ANSI C standard, defined in stdlib.h . */
+#define uprv_wcstombs(mbstr, wcstr, count) U_STANDARD_CPP_NAMESPACE wcstombs(mbstr, wcstr, count)
+#define uprv_mbstowcs(wcstr, mbstr, count) U_STANDARD_CPP_NAMESPACE mbstowcs(wcstr, mbstr, count)
+
+
+#endif
diff --git a/thirdparty/icu4c/common/dictbe.cpp b/thirdparty/icu4c/common/dictbe.cpp
new file mode 100644
index 0000000000..b42cdf03fa
--- /dev/null
+++ b/thirdparty/icu4c/common/dictbe.cpp
@@ -0,0 +1,1410 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2006-2016, International Business Machines Corporation
+ * and others. All Rights Reserved.
+ *******************************************************************************
+ */
+
+#include <utility>
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "brkeng.h"
+#include "dictbe.h"
+#include "unicode/uniset.h"
+#include "unicode/chariter.h"
+#include "unicode/ubrk.h"
+#include "utracimp.h"
+#include "uvectr32.h"
+#include "uvector.h"
+#include "uassert.h"
+#include "unicode/normlzr.h"
+#include "cmemory.h"
+#include "dictionarydata.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ ******************************************************************
+ */
+
+DictionaryBreakEngine::DictionaryBreakEngine() {
+}
+
+DictionaryBreakEngine::~DictionaryBreakEngine() {
+}
+
+UBool
+DictionaryBreakEngine::handles(UChar32 c) const {
+    return fSet.contains(c);
+}
+
+int32_t
+DictionaryBreakEngine::findBreaks( UText *text,
+                                 int32_t startPos,
+                                 int32_t endPos,
+                                 UVector32 &foundBreaks ) const {
+    (void)startPos;            // TODO: remove this param?
+    int32_t result = 0;
+
+    // Find the span of characters included in the set.
+    //   The span to break begins at the current position in the text, and
+    //   extends towards the start or end of the text, depending on 'reverse'.
+
+    int32_t start = (int32_t)utext_getNativeIndex(text);
+    int32_t current;
+    int32_t rangeStart;
+    int32_t rangeEnd;
+    UChar32 c = utext_current32(text);
+    while((current = (int32_t)utext_getNativeIndex(text)) < endPos && fSet.contains(c)) {
+        utext_next32(text);         // TODO:  recast loop for postincrement
+        c = utext_current32(text);
+    }
+    rangeStart = start;
+    rangeEnd = current;
+    result = divideUpDictionaryRange(text, rangeStart, rangeEnd, foundBreaks);
+    utext_setNativeIndex(text, current);
+    
+    return result;
+}
+
+void
+DictionaryBreakEngine::setCharacters( const UnicodeSet &set ) {
+    fSet = set;
+    // Compact for caching
+    fSet.compact();
+}
+
+/*
+ ******************************************************************
+ * PossibleWord
+ */
+
+// Helper class for improving readability of the Thai/Lao/Khmer word break
+// algorithm. The implementation is completely inline.
+
+// List size, limited by the maximum number of words in the dictionary
+// that form a nested sequence.
+static const int32_t POSSIBLE_WORD_LIST_MAX = 20;
+
+class PossibleWord {
+private:
+    // list of word candidate lengths, in increasing length order
+    // TODO: bytes would be sufficient for word lengths.
+    int32_t   count;      // Count of candidates
+    int32_t   prefix;     // The longest match with a dictionary word
+    int32_t   offset;     // Offset in the text of these candidates
+    int32_t   mark;       // The preferred candidate's offset
+    int32_t   current;    // The candidate we're currently looking at
+    int32_t   cuLengths[POSSIBLE_WORD_LIST_MAX];   // Word Lengths, in code units.
+    int32_t   cpLengths[POSSIBLE_WORD_LIST_MAX];   // Word Lengths, in code points.
+
+public:
+    PossibleWord() : count(0), prefix(0), offset(-1), mark(0), current(0) {}
+    ~PossibleWord() {}
+  
+    // Fill the list of candidates if needed, select the longest, and return the number found
+    int32_t   candidates( UText *text, DictionaryMatcher *dict, int32_t rangeEnd );
+  
+    // Select the currently marked candidate, point after it in the text, and invalidate self
+    int32_t   acceptMarked( UText *text );
+  
+    // Back up from the current candidate to the next shorter one; return TRUE if that exists
+    // and point the text after it
+    UBool     backUp( UText *text );
+  
+    // Return the longest prefix this candidate location shares with a dictionary word
+    // Return value is in code points.
+    int32_t   longestPrefix() { return prefix; }
+  
+    // Mark the current candidate as the one we like
+    void      markCurrent() { mark = current; }
+    
+    // Get length in code points of the marked word.
+    int32_t   markedCPLength() { return cpLengths[mark]; }
+};
+
+
+int32_t PossibleWord::candidates( UText *text, DictionaryMatcher *dict, int32_t rangeEnd ) {
+    // TODO: If getIndex is too slow, use offset < 0 and add discardAll()
+    int32_t start = (int32_t)utext_getNativeIndex(text);
+    if (start != offset) {
+        offset = start;
+        count = dict->matches(text, rangeEnd-start, UPRV_LENGTHOF(cuLengths), cuLengths, cpLengths, NULL, &prefix);
+        // Dictionary leaves text after longest prefix, not longest word. Back up.
+        if (count <= 0) {
+            utext_setNativeIndex(text, start);
+        }
+    }
+    if (count > 0) {
+        utext_setNativeIndex(text, start+cuLengths[count-1]);
+    }
+    current = count-1;
+    mark = current;
+    return count;
+}
+
+int32_t
+PossibleWord::acceptMarked( UText *text ) {
+    utext_setNativeIndex(text, offset + cuLengths[mark]);
+    return cuLengths[mark];
+}
+
+
+UBool
+PossibleWord::backUp( UText *text ) {
+    if (current > 0) {
+        utext_setNativeIndex(text, offset + cuLengths[--current]);
+        return TRUE;
+    }
+    return FALSE;
+}
+
+/*
+ ******************************************************************
+ * ThaiBreakEngine
+ */
+
+// How many words in a row are "good enough"?
+static const int32_t THAI_LOOKAHEAD = 3;
+
+// Will not combine a non-word with a preceding dictionary word longer than this
+static const int32_t THAI_ROOT_COMBINE_THRESHOLD = 3;
+
+// Will not combine a non-word that shares at least this much prefix with a
+// dictionary word, with a preceding word
+static const int32_t THAI_PREFIX_COMBINE_THRESHOLD = 3;
+
+// Ellision character
+static const int32_t THAI_PAIYANNOI = 0x0E2F;
+
+// Repeat character
+static const int32_t THAI_MAIYAMOK = 0x0E46;
+
+// Minimum word size
+static const int32_t THAI_MIN_WORD = 2;
+
+// Minimum number of characters for two words
+static const int32_t THAI_MIN_WORD_SPAN = THAI_MIN_WORD * 2;
+
+ThaiBreakEngine::ThaiBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status)
+    : DictionaryBreakEngine(),
+      fDictionary(adoptDictionary)
+{
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Thai");
+    fThaiWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Thai:]&[:LineBreak=SA:]]"), status);
+    if (U_SUCCESS(status)) {
+        setCharacters(fThaiWordSet);
+    }
+    fMarkSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Thai:]&[:LineBreak=SA:]&[:M:]]"), status);
+    fMarkSet.add(0x0020);
+    fEndWordSet = fThaiWordSet;
+    fEndWordSet.remove(0x0E31);             // MAI HAN-AKAT
+    fEndWordSet.remove(0x0E40, 0x0E44);     // SARA E through SARA AI MAIMALAI
+    fBeginWordSet.add(0x0E01, 0x0E2E);      // KO KAI through HO NOKHUK
+    fBeginWordSet.add(0x0E40, 0x0E44);      // SARA E through SARA AI MAIMALAI
+    fSuffixSet.add(THAI_PAIYANNOI);
+    fSuffixSet.add(THAI_MAIYAMOK);
+
+    // Compact for caching.
+    fMarkSet.compact();
+    fEndWordSet.compact();
+    fBeginWordSet.compact();
+    fSuffixSet.compact();
+    UTRACE_EXIT_STATUS(status);
+}
+
+ThaiBreakEngine::~ThaiBreakEngine() {
+    delete fDictionary;
+}
+
+int32_t
+ThaiBreakEngine::divideUpDictionaryRange( UText *text,
+                                                int32_t rangeStart,
+                                                int32_t rangeEnd,
+                                                UVector32 &foundBreaks ) const {
+    utext_setNativeIndex(text, rangeStart);
+    utext_moveIndex32(text, THAI_MIN_WORD_SPAN);
+    if (utext_getNativeIndex(text) >= rangeEnd) {
+        return 0;       // Not enough characters for two words
+    }
+    utext_setNativeIndex(text, rangeStart);
+
+
+    uint32_t wordsFound = 0;
+    int32_t cpWordLength = 0;    // Word Length in Code Points.
+    int32_t cuWordLength = 0;    // Word length in code units (UText native indexing)
+    int32_t current;
+    UErrorCode status = U_ZERO_ERROR;
+    PossibleWord words[THAI_LOOKAHEAD];
+    
+    utext_setNativeIndex(text, rangeStart);
+    
+    while (U_SUCCESS(status) && (current = (int32_t)utext_getNativeIndex(text)) < rangeEnd) {
+        cpWordLength = 0;
+        cuWordLength = 0;
+
+        // Look for candidate words at the current position
+        int32_t candidates = words[wordsFound%THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+        
+        // If we found exactly one, use that
+        if (candidates == 1) {
+            cuWordLength = words[wordsFound % THAI_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % THAI_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        // If there was more than one, see which one can take us forward the most words
+        else if (candidates > 1) {
+            // If we're already at the end of the range, we're done
+            if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                goto foundBest;
+            }
+            do {
+                int32_t wordsMatched = 1;
+                if (words[(wordsFound + 1) % THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) > 0) {
+                    if (wordsMatched < 2) {
+                        // Followed by another dictionary word; mark first word as a good candidate
+                        words[wordsFound%THAI_LOOKAHEAD].markCurrent();
+                        wordsMatched = 2;
+                    }
+                    
+                    // If we're already at the end of the range, we're done
+                    if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                        goto foundBest;
+                    }
+                    
+                    // See if any of the possible second words is followed by a third word
+                    do {
+                        // If we find a third word, stop right away
+                        if (words[(wordsFound + 2) % THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd)) {
+                            words[wordsFound % THAI_LOOKAHEAD].markCurrent();
+                            goto foundBest;
+                        }
+                    }
+                    while (words[(wordsFound + 1) % THAI_LOOKAHEAD].backUp(text));
+                }
+            }
+            while (words[wordsFound % THAI_LOOKAHEAD].backUp(text));
+foundBest:
+            // Set UText position to after the accepted word.
+            cuWordLength = words[wordsFound % THAI_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % THAI_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        
+        // We come here after having either found a word or not. We look ahead to the
+        // next word. If it's not a dictionary word, we will combine it with the word we
+        // just found (if there is one), but only if the preceding word does not exceed
+        // the threshold.
+        // The text iterator should now be positioned at the end of the word we found.
+        
+        UChar32 uc = 0;
+        if ((int32_t)utext_getNativeIndex(text) < rangeEnd &&  cpWordLength < THAI_ROOT_COMBINE_THRESHOLD) {
+            // if it is a dictionary word, do nothing. If it isn't, then if there is
+            // no preceding word, or the non-word shares less than the minimum threshold
+            // of characters with a dictionary word, then scan to resynchronize
+            if (words[wordsFound % THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+                  && (cuWordLength == 0
+                      || words[wordsFound%THAI_LOOKAHEAD].longestPrefix() < THAI_PREFIX_COMBINE_THRESHOLD)) {
+                // Look for a plausible word boundary
+                int32_t remaining = rangeEnd - (current+cuWordLength);
+                UChar32 pc;
+                int32_t chars = 0;
+                for (;;) {
+                    int32_t pcIndex = (int32_t)utext_getNativeIndex(text);
+                    pc = utext_next32(text);
+                    int32_t pcSize = (int32_t)utext_getNativeIndex(text) - pcIndex;
+                    chars += pcSize;
+                    remaining -= pcSize;
+                    if (remaining <= 0) {
+                        break;
+                    }
+                    uc = utext_current32(text);
+                    if (fEndWordSet.contains(pc) && fBeginWordSet.contains(uc)) {
+                        // Maybe. See if it's in the dictionary.
+                        // NOTE: In the original Apple code, checked that the next
+                        // two characters after uc were not 0x0E4C THANTHAKHAT before
+                        // checking the dictionary. That is just a performance filter,
+                        // but it's not clear it's faster than checking the trie.
+                        int32_t num_candidates = words[(wordsFound + 1) % THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+                        utext_setNativeIndex(text, current + cuWordLength + chars);
+                        if (num_candidates > 0) {
+                            break;
+                        }
+                    }
+                }
+                
+                // Bump the word count if there wasn't already one
+                if (cuWordLength <= 0) {
+                    wordsFound += 1;
+                }
+                
+                // Update the length with the passed-over characters
+                cuWordLength += chars;
+            }
+            else {
+                // Back up to where we were for next iteration
+                utext_setNativeIndex(text, current+cuWordLength);
+            }
+        }
+        
+        // Never stop before a combining mark.
+        int32_t currPos;
+        while ((currPos = (int32_t)utext_getNativeIndex(text)) < rangeEnd && fMarkSet.contains(utext_current32(text))) {
+            utext_next32(text);
+            cuWordLength += (int32_t)utext_getNativeIndex(text) - currPos;
+        }
+        
+        // Look ahead for possible suffixes if a dictionary word does not follow.
+        // We do this in code rather than using a rule so that the heuristic
+        // resynch continues to function. For example, one of the suffix characters
+        // could be a typo in the middle of a word.
+        if ((int32_t)utext_getNativeIndex(text) < rangeEnd && cuWordLength > 0) {
+            if (words[wordsFound%THAI_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+                && fSuffixSet.contains(uc = utext_current32(text))) {
+                if (uc == THAI_PAIYANNOI) {
+                    if (!fSuffixSet.contains(utext_previous32(text))) {
+                        // Skip over previous end and PAIYANNOI
+                        utext_next32(text);
+                        int32_t paiyannoiIndex = (int32_t)utext_getNativeIndex(text);
+                        utext_next32(text);
+                        cuWordLength += (int32_t)utext_getNativeIndex(text) - paiyannoiIndex;    // Add PAIYANNOI to word
+                        uc = utext_current32(text);     // Fetch next character
+                    }
+                    else {
+                        // Restore prior position
+                        utext_next32(text);
+                    }
+                }
+                if (uc == THAI_MAIYAMOK) {
+                    if (utext_previous32(text) != THAI_MAIYAMOK) {
+                        // Skip over previous end and MAIYAMOK
+                        utext_next32(text);
+                        int32_t maiyamokIndex = (int32_t)utext_getNativeIndex(text);
+                        utext_next32(text);
+                        cuWordLength += (int32_t)utext_getNativeIndex(text) - maiyamokIndex;    // Add MAIYAMOK to word
+                    }
+                    else {
+                        // Restore prior position
+                        utext_next32(text);
+                    }
+                }
+            }
+            else {
+                utext_setNativeIndex(text, current+cuWordLength);
+            }
+        }
+
+        // Did we find a word on this iteration? If so, push it on the break stack
+        if (cuWordLength > 0) {
+            foundBreaks.push((current+cuWordLength), status);
+        }
+    }
+
+    // Don't return a break for the end of the dictionary range if there is one there.
+    if (foundBreaks.peeki() >= rangeEnd) {
+        (void) foundBreaks.popi();
+        wordsFound -= 1;
+    }
+
+    return wordsFound;
+}
+
+/*
+ ******************************************************************
+ * LaoBreakEngine
+ */
+
+// How many words in a row are "good enough"?
+static const int32_t LAO_LOOKAHEAD = 3;
+
+// Will not combine a non-word with a preceding dictionary word longer than this
+static const int32_t LAO_ROOT_COMBINE_THRESHOLD = 3;
+
+// Will not combine a non-word that shares at least this much prefix with a
+// dictionary word, with a preceding word
+static const int32_t LAO_PREFIX_COMBINE_THRESHOLD = 3;
+
+// Minimum word size
+static const int32_t LAO_MIN_WORD = 2;
+
+// Minimum number of characters for two words
+static const int32_t LAO_MIN_WORD_SPAN = LAO_MIN_WORD * 2;
+
+LaoBreakEngine::LaoBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status)
+    : DictionaryBreakEngine(),
+      fDictionary(adoptDictionary)
+{
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Laoo");
+    fLaoWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Laoo:]&[:LineBreak=SA:]]"), status);
+    if (U_SUCCESS(status)) {
+        setCharacters(fLaoWordSet);
+    }
+    fMarkSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Laoo:]&[:LineBreak=SA:]&[:M:]]"), status);
+    fMarkSet.add(0x0020);
+    fEndWordSet = fLaoWordSet;
+    fEndWordSet.remove(0x0EC0, 0x0EC4);     // prefix vowels
+    fBeginWordSet.add(0x0E81, 0x0EAE);      // basic consonants (including holes for corresponding Thai characters)
+    fBeginWordSet.add(0x0EDC, 0x0EDD);      // digraph consonants (no Thai equivalent)
+    fBeginWordSet.add(0x0EC0, 0x0EC4);      // prefix vowels
+
+    // Compact for caching.
+    fMarkSet.compact();
+    fEndWordSet.compact();
+    fBeginWordSet.compact();
+    UTRACE_EXIT_STATUS(status);
+}
+
+LaoBreakEngine::~LaoBreakEngine() {
+    delete fDictionary;
+}
+
+int32_t
+LaoBreakEngine::divideUpDictionaryRange( UText *text,
+                                                int32_t rangeStart,
+                                                int32_t rangeEnd,
+                                                UVector32 &foundBreaks ) const {
+    if ((rangeEnd - rangeStart) < LAO_MIN_WORD_SPAN) {
+        return 0;       // Not enough characters for two words
+    }
+
+    uint32_t wordsFound = 0;
+    int32_t cpWordLength = 0;
+    int32_t cuWordLength = 0;
+    int32_t current;
+    UErrorCode status = U_ZERO_ERROR;
+    PossibleWord words[LAO_LOOKAHEAD];
+    
+    utext_setNativeIndex(text, rangeStart);
+    
+    while (U_SUCCESS(status) && (current = (int32_t)utext_getNativeIndex(text)) < rangeEnd) {
+        cuWordLength = 0;
+        cpWordLength = 0;
+
+        // Look for candidate words at the current position
+        int32_t candidates = words[wordsFound%LAO_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+        
+        // If we found exactly one, use that
+        if (candidates == 1) {
+            cuWordLength = words[wordsFound % LAO_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % LAO_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        // If there was more than one, see which one can take us forward the most words
+        else if (candidates > 1) {
+            // If we're already at the end of the range, we're done
+            if (utext_getNativeIndex(text) >= rangeEnd) {
+                goto foundBest;
+            }
+            do {
+                int32_t wordsMatched = 1;
+                if (words[(wordsFound + 1) % LAO_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) > 0) {
+                    if (wordsMatched < 2) {
+                        // Followed by another dictionary word; mark first word as a good candidate
+                        words[wordsFound%LAO_LOOKAHEAD].markCurrent();
+                        wordsMatched = 2;
+                    }
+                    
+                    // If we're already at the end of the range, we're done
+                    if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                        goto foundBest;
+                    }
+                    
+                    // See if any of the possible second words is followed by a third word
+                    do {
+                        // If we find a third word, stop right away
+                        if (words[(wordsFound + 2) % LAO_LOOKAHEAD].candidates(text, fDictionary, rangeEnd)) {
+                            words[wordsFound % LAO_LOOKAHEAD].markCurrent();
+                            goto foundBest;
+                        }
+                    }
+                    while (words[(wordsFound + 1) % LAO_LOOKAHEAD].backUp(text));
+                }
+            }
+            while (words[wordsFound % LAO_LOOKAHEAD].backUp(text));
+foundBest:
+            cuWordLength = words[wordsFound % LAO_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % LAO_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        
+        // We come here after having either found a word or not. We look ahead to the
+        // next word. If it's not a dictionary word, we will combine it withe the word we
+        // just found (if there is one), but only if the preceding word does not exceed
+        // the threshold.
+        // The text iterator should now be positioned at the end of the word we found.
+        if ((int32_t)utext_getNativeIndex(text) < rangeEnd && cpWordLength < LAO_ROOT_COMBINE_THRESHOLD) {
+            // if it is a dictionary word, do nothing. If it isn't, then if there is
+            // no preceding word, or the non-word shares less than the minimum threshold
+            // of characters with a dictionary word, then scan to resynchronize
+            if (words[wordsFound % LAO_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+                  && (cuWordLength == 0
+                      || words[wordsFound%LAO_LOOKAHEAD].longestPrefix() < LAO_PREFIX_COMBINE_THRESHOLD)) {
+                // Look for a plausible word boundary
+                int32_t remaining = rangeEnd - (current + cuWordLength);
+                UChar32 pc;
+                UChar32 uc;
+                int32_t chars = 0;
+                for (;;) {
+                    int32_t pcIndex = (int32_t)utext_getNativeIndex(text);
+                    pc = utext_next32(text);
+                    int32_t pcSize = (int32_t)utext_getNativeIndex(text) - pcIndex;
+                    chars += pcSize;
+                    remaining -= pcSize;
+                    if (remaining <= 0) {
+                        break;
+                    }
+                    uc = utext_current32(text);
+                    if (fEndWordSet.contains(pc) && fBeginWordSet.contains(uc)) {
+                        // Maybe. See if it's in the dictionary.
+                        // TODO: this looks iffy; compare with old code.
+                        int32_t num_candidates = words[(wordsFound + 1) % LAO_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+                        utext_setNativeIndex(text, current + cuWordLength + chars);
+                        if (num_candidates > 0) {
+                            break;
+                        }
+                    }
+                }
+                
+                // Bump the word count if there wasn't already one
+                if (cuWordLength <= 0) {
+                    wordsFound += 1;
+                }
+                
+                // Update the length with the passed-over characters
+                cuWordLength += chars;
+            }
+            else {
+                // Back up to where we were for next iteration
+                utext_setNativeIndex(text, current + cuWordLength);
+            }
+        }
+        
+        // Never stop before a combining mark.
+        int32_t currPos;
+        while ((currPos = (int32_t)utext_getNativeIndex(text)) < rangeEnd && fMarkSet.contains(utext_current32(text))) {
+            utext_next32(text);
+            cuWordLength += (int32_t)utext_getNativeIndex(text) - currPos;
+        }
+        
+        // Look ahead for possible suffixes if a dictionary word does not follow.
+        // We do this in code rather than using a rule so that the heuristic
+        // resynch continues to function. For example, one of the suffix characters
+        // could be a typo in the middle of a word.
+        // NOT CURRENTLY APPLICABLE TO LAO
+
+        // Did we find a word on this iteration? If so, push it on the break stack
+        if (cuWordLength > 0) {
+            foundBreaks.push((current+cuWordLength), status);
+        }
+    }
+
+    // Don't return a break for the end of the dictionary range if there is one there.
+    if (foundBreaks.peeki() >= rangeEnd) {
+        (void) foundBreaks.popi();
+        wordsFound -= 1;
+    }
+
+    return wordsFound;
+}
+
+/*
+ ******************************************************************
+ * BurmeseBreakEngine
+ */
+
+// How many words in a row are "good enough"?
+static const int32_t BURMESE_LOOKAHEAD = 3;
+
+// Will not combine a non-word with a preceding dictionary word longer than this
+static const int32_t BURMESE_ROOT_COMBINE_THRESHOLD = 3;
+
+// Will not combine a non-word that shares at least this much prefix with a
+// dictionary word, with a preceding word
+static const int32_t BURMESE_PREFIX_COMBINE_THRESHOLD = 3;
+
+// Minimum word size
+static const int32_t BURMESE_MIN_WORD = 2;
+
+// Minimum number of characters for two words
+static const int32_t BURMESE_MIN_WORD_SPAN = BURMESE_MIN_WORD * 2;
+
+BurmeseBreakEngine::BurmeseBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status)
+    : DictionaryBreakEngine(),
+      fDictionary(adoptDictionary)
+{
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Mymr");
+    fBurmeseWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Mymr:]&[:LineBreak=SA:]]"), status);
+    if (U_SUCCESS(status)) {
+        setCharacters(fBurmeseWordSet);
+    }
+    fMarkSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Mymr:]&[:LineBreak=SA:]&[:M:]]"), status);
+    fMarkSet.add(0x0020);
+    fEndWordSet = fBurmeseWordSet;
+    fBeginWordSet.add(0x1000, 0x102A);      // basic consonants and independent vowels
+
+    // Compact for caching.
+    fMarkSet.compact();
+    fEndWordSet.compact();
+    fBeginWordSet.compact();
+    UTRACE_EXIT_STATUS(status);
+}
+
+BurmeseBreakEngine::~BurmeseBreakEngine() {
+    delete fDictionary;
+}
+
+int32_t
+BurmeseBreakEngine::divideUpDictionaryRange( UText *text,
+                                                int32_t rangeStart,
+                                                int32_t rangeEnd,
+                                                UVector32 &foundBreaks ) const {
+    if ((rangeEnd - rangeStart) < BURMESE_MIN_WORD_SPAN) {
+        return 0;       // Not enough characters for two words
+    }
+
+    uint32_t wordsFound = 0;
+    int32_t cpWordLength = 0;
+    int32_t cuWordLength = 0;
+    int32_t current;
+    UErrorCode status = U_ZERO_ERROR;
+    PossibleWord words[BURMESE_LOOKAHEAD];
+    
+    utext_setNativeIndex(text, rangeStart);
+    
+    while (U_SUCCESS(status) && (current = (int32_t)utext_getNativeIndex(text)) < rangeEnd) {
+        cuWordLength = 0;
+        cpWordLength = 0;
+
+        // Look for candidate words at the current position
+        int32_t candidates = words[wordsFound%BURMESE_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+        
+        // If we found exactly one, use that
+        if (candidates == 1) {
+            cuWordLength = words[wordsFound % BURMESE_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % BURMESE_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        // If there was more than one, see which one can take us forward the most words
+        else if (candidates > 1) {
+            // If we're already at the end of the range, we're done
+            if (utext_getNativeIndex(text) >= rangeEnd) {
+                goto foundBest;
+            }
+            do {
+                int32_t wordsMatched = 1;
+                if (words[(wordsFound + 1) % BURMESE_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) > 0) {
+                    if (wordsMatched < 2) {
+                        // Followed by another dictionary word; mark first word as a good candidate
+                        words[wordsFound%BURMESE_LOOKAHEAD].markCurrent();
+                        wordsMatched = 2;
+                    }
+                    
+                    // If we're already at the end of the range, we're done
+                    if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                        goto foundBest;
+                    }
+                    
+                    // See if any of the possible second words is followed by a third word
+                    do {
+                        // If we find a third word, stop right away
+                        if (words[(wordsFound + 2) % BURMESE_LOOKAHEAD].candidates(text, fDictionary, rangeEnd)) {
+                            words[wordsFound % BURMESE_LOOKAHEAD].markCurrent();
+                            goto foundBest;
+                        }
+                    }
+                    while (words[(wordsFound + 1) % BURMESE_LOOKAHEAD].backUp(text));
+                }
+            }
+            while (words[wordsFound % BURMESE_LOOKAHEAD].backUp(text));
+foundBest:
+            cuWordLength = words[wordsFound % BURMESE_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % BURMESE_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+        
+        // We come here after having either found a word or not. We look ahead to the
+        // next word. If it's not a dictionary word, we will combine it withe the word we
+        // just found (if there is one), but only if the preceding word does not exceed
+        // the threshold.
+        // The text iterator should now be positioned at the end of the word we found.
+        if ((int32_t)utext_getNativeIndex(text) < rangeEnd && cpWordLength < BURMESE_ROOT_COMBINE_THRESHOLD) {
+            // if it is a dictionary word, do nothing. If it isn't, then if there is
+            // no preceding word, or the non-word shares less than the minimum threshold
+            // of characters with a dictionary word, then scan to resynchronize
+            if (words[wordsFound % BURMESE_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+                  && (cuWordLength == 0
+                      || words[wordsFound%BURMESE_LOOKAHEAD].longestPrefix() < BURMESE_PREFIX_COMBINE_THRESHOLD)) {
+                // Look for a plausible word boundary
+                int32_t remaining = rangeEnd - (current + cuWordLength);
+                UChar32 pc;
+                UChar32 uc;
+                int32_t chars = 0;
+                for (;;) {
+                    int32_t pcIndex = (int32_t)utext_getNativeIndex(text);
+                    pc = utext_next32(text);
+                    int32_t pcSize = (int32_t)utext_getNativeIndex(text) - pcIndex;
+                    chars += pcSize;
+                    remaining -= pcSize;
+                    if (remaining <= 0) {
+                        break;
+                    }
+                    uc = utext_current32(text);
+                    if (fEndWordSet.contains(pc) && fBeginWordSet.contains(uc)) {
+                        // Maybe. See if it's in the dictionary.
+                        // TODO: this looks iffy; compare with old code.
+                        int32_t num_candidates = words[(wordsFound + 1) % BURMESE_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+                        utext_setNativeIndex(text, current + cuWordLength + chars);
+                        if (num_candidates > 0) {
+                            break;
+                        }
+                    }
+                }
+                
+                // Bump the word count if there wasn't already one
+                if (cuWordLength <= 0) {
+                    wordsFound += 1;
+                }
+                
+                // Update the length with the passed-over characters
+                cuWordLength += chars;
+            }
+            else {
+                // Back up to where we were for next iteration
+                utext_setNativeIndex(text, current + cuWordLength);
+            }
+        }
+        
+        // Never stop before a combining mark.
+        int32_t currPos;
+        while ((currPos = (int32_t)utext_getNativeIndex(text)) < rangeEnd && fMarkSet.contains(utext_current32(text))) {
+            utext_next32(text);
+            cuWordLength += (int32_t)utext_getNativeIndex(text) - currPos;
+        }
+        
+        // Look ahead for possible suffixes if a dictionary word does not follow.
+        // We do this in code rather than using a rule so that the heuristic
+        // resynch continues to function. For example, one of the suffix characters
+        // could be a typo in the middle of a word.
+        // NOT CURRENTLY APPLICABLE TO BURMESE
+
+        // Did we find a word on this iteration? If so, push it on the break stack
+        if (cuWordLength > 0) {
+            foundBreaks.push((current+cuWordLength), status);
+        }
+    }
+
+    // Don't return a break for the end of the dictionary range if there is one there.
+    if (foundBreaks.peeki() >= rangeEnd) {
+        (void) foundBreaks.popi();
+        wordsFound -= 1;
+    }
+
+    return wordsFound;
+}
+
+/*
+ ******************************************************************
+ * KhmerBreakEngine
+ */
+
+// How many words in a row are "good enough"?
+static const int32_t KHMER_LOOKAHEAD = 3;
+
+// Will not combine a non-word with a preceding dictionary word longer than this
+static const int32_t KHMER_ROOT_COMBINE_THRESHOLD = 3;
+
+// Will not combine a non-word that shares at least this much prefix with a
+// dictionary word, with a preceding word
+static const int32_t KHMER_PREFIX_COMBINE_THRESHOLD = 3;
+
+// Minimum word size
+static const int32_t KHMER_MIN_WORD = 2;
+
+// Minimum number of characters for two words
+static const int32_t KHMER_MIN_WORD_SPAN = KHMER_MIN_WORD * 2;
+
+KhmerBreakEngine::KhmerBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status)
+    : DictionaryBreakEngine(),
+      fDictionary(adoptDictionary)
+{
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Khmr");
+    fKhmerWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Khmr:]&[:LineBreak=SA:]]"), status);
+    if (U_SUCCESS(status)) {
+        setCharacters(fKhmerWordSet);
+    }
+    fMarkSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Khmr:]&[:LineBreak=SA:]&[:M:]]"), status);
+    fMarkSet.add(0x0020);
+    fEndWordSet = fKhmerWordSet;
+    fBeginWordSet.add(0x1780, 0x17B3);
+    //fBeginWordSet.add(0x17A3, 0x17A4);      // deprecated vowels
+    //fEndWordSet.remove(0x17A5, 0x17A9);     // Khmer independent vowels that can't end a word
+    //fEndWordSet.remove(0x17B2);             // Khmer independent vowel that can't end a word
+    fEndWordSet.remove(0x17D2);             // KHMER SIGN COENG that combines some following characters
+    //fEndWordSet.remove(0x17B6, 0x17C5);     // Remove dependent vowels
+//    fEndWordSet.remove(0x0E31);             // MAI HAN-AKAT
+//    fEndWordSet.remove(0x0E40, 0x0E44);     // SARA E through SARA AI MAIMALAI
+//    fBeginWordSet.add(0x0E01, 0x0E2E);      // KO KAI through HO NOKHUK
+//    fBeginWordSet.add(0x0E40, 0x0E44);      // SARA E through SARA AI MAIMALAI
+//    fSuffixSet.add(THAI_PAIYANNOI);
+//    fSuffixSet.add(THAI_MAIYAMOK);
+
+    // Compact for caching.
+    fMarkSet.compact();
+    fEndWordSet.compact();
+    fBeginWordSet.compact();
+//    fSuffixSet.compact();
+    UTRACE_EXIT_STATUS(status);
+}
+
+KhmerBreakEngine::~KhmerBreakEngine() {
+    delete fDictionary;
+}
+
+int32_t
+KhmerBreakEngine::divideUpDictionaryRange( UText *text,
+                                                int32_t rangeStart,
+                                                int32_t rangeEnd,
+                                                UVector32 &foundBreaks ) const {
+    if ((rangeEnd - rangeStart) < KHMER_MIN_WORD_SPAN) {
+        return 0;       // Not enough characters for two words
+    }
+
+    uint32_t wordsFound = 0;
+    int32_t cpWordLength = 0;
+    int32_t cuWordLength = 0;
+    int32_t current;
+    UErrorCode status = U_ZERO_ERROR;
+    PossibleWord words[KHMER_LOOKAHEAD];
+
+    utext_setNativeIndex(text, rangeStart);
+
+    while (U_SUCCESS(status) && (current = (int32_t)utext_getNativeIndex(text)) < rangeEnd) {
+        cuWordLength = 0;
+        cpWordLength = 0;
+
+        // Look for candidate words at the current position
+        int32_t candidates = words[wordsFound%KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+
+        // If we found exactly one, use that
+        if (candidates == 1) {
+            cuWordLength = words[wordsFound % KHMER_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % KHMER_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+
+        // If there was more than one, see which one can take us forward the most words
+        else if (candidates > 1) {
+            // If we're already at the end of the range, we're done
+            if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                goto foundBest;
+            }
+            do {
+                int32_t wordsMatched = 1;
+                if (words[(wordsFound + 1) % KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) > 0) {
+                    if (wordsMatched < 2) {
+                        // Followed by another dictionary word; mark first word as a good candidate
+                        words[wordsFound % KHMER_LOOKAHEAD].markCurrent();
+                        wordsMatched = 2;
+                    }
+
+                    // If we're already at the end of the range, we're done
+                    if ((int32_t)utext_getNativeIndex(text) >= rangeEnd) {
+                        goto foundBest;
+                    }
+
+                    // See if any of the possible second words is followed by a third word
+                    do {
+                        // If we find a third word, stop right away
+                        if (words[(wordsFound + 2) % KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd)) {
+                            words[wordsFound % KHMER_LOOKAHEAD].markCurrent();
+                            goto foundBest;
+                        }
+                    }
+                    while (words[(wordsFound + 1) % KHMER_LOOKAHEAD].backUp(text));
+                }
+            }
+            while (words[wordsFound % KHMER_LOOKAHEAD].backUp(text));
+foundBest:
+            cuWordLength = words[wordsFound % KHMER_LOOKAHEAD].acceptMarked(text);
+            cpWordLength = words[wordsFound % KHMER_LOOKAHEAD].markedCPLength();
+            wordsFound += 1;
+        }
+
+        // We come here after having either found a word or not. We look ahead to the
+        // next word. If it's not a dictionary word, we will combine it with the word we
+        // just found (if there is one), but only if the preceding word does not exceed
+        // the threshold.
+        // The text iterator should now be positioned at the end of the word we found.
+        if ((int32_t)utext_getNativeIndex(text) < rangeEnd && cpWordLength < KHMER_ROOT_COMBINE_THRESHOLD) {
+            // if it is a dictionary word, do nothing. If it isn't, then if there is
+            // no preceding word, or the non-word shares less than the minimum threshold
+            // of characters with a dictionary word, then scan to resynchronize
+            if (words[wordsFound % KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+                  && (cuWordLength == 0
+                      || words[wordsFound % KHMER_LOOKAHEAD].longestPrefix() < KHMER_PREFIX_COMBINE_THRESHOLD)) {
+                // Look for a plausible word boundary
+                int32_t remaining = rangeEnd - (current+cuWordLength);
+                UChar32 pc;
+                UChar32 uc;
+                int32_t chars = 0;
+                for (;;) {
+                    int32_t pcIndex = (int32_t)utext_getNativeIndex(text);
+                    pc = utext_next32(text);
+                    int32_t pcSize = (int32_t)utext_getNativeIndex(text) - pcIndex;
+                    chars += pcSize;
+                    remaining -= pcSize;
+                    if (remaining <= 0) {
+                        break;
+                    }
+                    uc = utext_current32(text);
+                    if (fEndWordSet.contains(pc) && fBeginWordSet.contains(uc)) {
+                        // Maybe. See if it's in the dictionary.
+                        int32_t num_candidates = words[(wordsFound + 1) % KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd);
+                        utext_setNativeIndex(text, current+cuWordLength+chars);
+                        if (num_candidates > 0) {
+                            break;
+                        }
+                    }
+                }
+
+                // Bump the word count if there wasn't already one
+                if (cuWordLength <= 0) {
+                    wordsFound += 1;
+                }
+
+                // Update the length with the passed-over characters
+                cuWordLength += chars;
+            }
+            else {
+                // Back up to where we were for next iteration
+                utext_setNativeIndex(text, current+cuWordLength);
+            }
+        }
+
+        // Never stop before a combining mark.
+        int32_t currPos;
+        while ((currPos = (int32_t)utext_getNativeIndex(text)) < rangeEnd && fMarkSet.contains(utext_current32(text))) {
+            utext_next32(text);
+            cuWordLength += (int32_t)utext_getNativeIndex(text) - currPos;
+        }
+
+        // Look ahead for possible suffixes if a dictionary word does not follow.
+        // We do this in code rather than using a rule so that the heuristic
+        // resynch continues to function. For example, one of the suffix characters
+        // could be a typo in the middle of a word.
+//        if ((int32_t)utext_getNativeIndex(text) < rangeEnd && wordLength > 0) {
+//            if (words[wordsFound%KHMER_LOOKAHEAD].candidates(text, fDictionary, rangeEnd) <= 0
+//                && fSuffixSet.contains(uc = utext_current32(text))) {
+//                if (uc == KHMER_PAIYANNOI) {
+//                    if (!fSuffixSet.contains(utext_previous32(text))) {
+//                        // Skip over previous end and PAIYANNOI
+//                        utext_next32(text);
+//                        utext_next32(text);
+//                        wordLength += 1;            // Add PAIYANNOI to word
+//                        uc = utext_current32(text);     // Fetch next character
+//                    }
+//                    else {
+//                        // Restore prior position
+//                        utext_next32(text);
+//                    }
+//                }
+//                if (uc == KHMER_MAIYAMOK) {
+//                    if (utext_previous32(text) != KHMER_MAIYAMOK) {
+//                        // Skip over previous end and MAIYAMOK
+//                        utext_next32(text);
+//                        utext_next32(text);
+//                        wordLength += 1;            // Add MAIYAMOK to word
+//                    }
+//                    else {
+//                        // Restore prior position
+//                        utext_next32(text);
+//                    }
+//                }
+//            }
+//            else {
+//                utext_setNativeIndex(text, current+wordLength);
+//            }
+//        }
+
+        // Did we find a word on this iteration? If so, push it on the break stack
+        if (cuWordLength > 0) {
+            foundBreaks.push((current+cuWordLength), status);
+        }
+    }
+    
+    // Don't return a break for the end of the dictionary range if there is one there.
+    if (foundBreaks.peeki() >= rangeEnd) {
+        (void) foundBreaks.popi();
+        wordsFound -= 1;
+    }
+
+    return wordsFound;
+}
+
+#if !UCONFIG_NO_NORMALIZATION
+/*
+ ******************************************************************
+ * CjkBreakEngine
+ */
+static const uint32_t kuint32max = 0xFFFFFFFF;
+CjkBreakEngine::CjkBreakEngine(DictionaryMatcher *adoptDictionary, LanguageType type, UErrorCode &status)
+: DictionaryBreakEngine(), fDictionary(adoptDictionary) {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Hani");
+    // Korean dictionary only includes Hangul syllables
+    fHangulWordSet.applyPattern(UNICODE_STRING_SIMPLE("[\\uac00-\\ud7a3]"), status);
+    fHanWordSet.applyPattern(UNICODE_STRING_SIMPLE("[:Han:]"), status);
+    fKatakanaWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Katakana:]\\uff9e\\uff9f]"), status);
+    fHiraganaWordSet.applyPattern(UNICODE_STRING_SIMPLE("[:Hiragana:]"), status);
+    nfkcNorm2 = Normalizer2::getNFKCInstance(status);
+
+    if (U_SUCCESS(status)) {
+        // handle Korean and Japanese/Chinese using different dictionaries
+        if (type == kKorean) {
+            setCharacters(fHangulWordSet);
+        } else { //Chinese and Japanese
+            UnicodeSet cjSet;
+            cjSet.addAll(fHanWordSet);
+            cjSet.addAll(fKatakanaWordSet);
+            cjSet.addAll(fHiraganaWordSet);
+            cjSet.add(0xFF70); // HALFWIDTH KATAKANA-HIRAGANA PROLONGED SOUND MARK
+            cjSet.add(0x30FC); // KATAKANA-HIRAGANA PROLONGED SOUND MARK
+            setCharacters(cjSet);
+        }
+    }
+    UTRACE_EXIT_STATUS(status);
+}
+
+CjkBreakEngine::~CjkBreakEngine(){
+    delete fDictionary;
+}
+
+// The katakanaCost values below are based on the length frequencies of all
+// katakana phrases in the dictionary
+static const int32_t kMaxKatakanaLength = 8;
+static const int32_t kMaxKatakanaGroupLength = 20;
+static const uint32_t maxSnlp = 255;
+
+static inline uint32_t getKatakanaCost(int32_t wordLength){
+    //TODO: fill array with actual values from dictionary!
+    static const uint32_t katakanaCost[kMaxKatakanaLength + 1]
+                                       = {8192, 984, 408, 240, 204, 252, 300, 372, 480};
+    return (wordLength > kMaxKatakanaLength) ? 8192 : katakanaCost[wordLength];
+}
+
+static inline bool isKatakana(UChar32 value) {
+    return (value >= 0x30A1 && value <= 0x30FE && value != 0x30FB) ||
+            (value >= 0xFF66 && value <= 0xFF9f);
+}
+
+
+// Function for accessing internal utext flags.
+//   Replicates an internal UText function.
+
+static inline int32_t utext_i32_flag(int32_t bitIndex) {
+    return (int32_t)1 << bitIndex;
+}
+
+       
+/*
+ * @param text A UText representing the text
+ * @param rangeStart The start of the range of dictionary characters
+ * @param rangeEnd The end of the range of dictionary characters
+ * @param foundBreaks vector<int32> to receive the break positions
+ * @return The number of breaks found
+ */
+int32_t 
+CjkBreakEngine::divideUpDictionaryRange( UText *inText,
+        int32_t rangeStart,
+        int32_t rangeEnd,
+        UVector32 &foundBreaks ) const {
+    if (rangeStart >= rangeEnd) {
+        return 0;
+    }
+
+    // UnicodeString version of input UText, NFKC normalized if necessary.
+    UnicodeString inString;
+
+    // inputMap[inStringIndex] = corresponding native index from UText inText.
+    // If NULL then mapping is 1:1
+    LocalPointer<UVector32>     inputMap;
+
+    UErrorCode     status      = U_ZERO_ERROR;
+
+
+    // if UText has the input string as one contiguous UTF-16 chunk
+    if ((inText->providerProperties & utext_i32_flag(UTEXT_PROVIDER_STABLE_CHUNKS)) &&
+         inText->chunkNativeStart <= rangeStart &&
+         inText->chunkNativeLimit >= rangeEnd   &&
+         inText->nativeIndexingLimit >= rangeEnd - inText->chunkNativeStart) {
+
+        // Input UText is in one contiguous UTF-16 chunk.
+        // Use Read-only aliasing UnicodeString.
+        inString.setTo(FALSE,
+                       inText->chunkContents + rangeStart - inText->chunkNativeStart,
+                       rangeEnd - rangeStart);
+    } else {
+        // Copy the text from the original inText (UText) to inString (UnicodeString).
+        // Create a map from UnicodeString indices -> UText offsets.
+        utext_setNativeIndex(inText, rangeStart);
+        int32_t limit = rangeEnd;
+        U_ASSERT(limit <= utext_nativeLength(inText));
+        if (limit > utext_nativeLength(inText)) {
+            limit = (int32_t)utext_nativeLength(inText);
+        }
+        inputMap.adoptInsteadAndCheckErrorCode(new UVector32(status), status);
+        if (U_FAILURE(status)) {
+            return 0;
+        }
+        while (utext_getNativeIndex(inText) < limit) {
+            int32_t nativePosition = (int32_t)utext_getNativeIndex(inText);
+            UChar32 c = utext_next32(inText);
+            U_ASSERT(c != U_SENTINEL);
+            inString.append(c);
+            while (inputMap->size() < inString.length()) {
+                inputMap->addElement(nativePosition, status);
+            }
+        }
+        inputMap->addElement(limit, status);
+    }
+
+
+    if (!nfkcNorm2->isNormalized(inString, status)) {
+        UnicodeString normalizedInput;
+        //  normalizedMap[normalizedInput position] ==  original UText position.
+        LocalPointer<UVector32> normalizedMap(new UVector32(status), status);
+        if (U_FAILURE(status)) {
+            return 0;
+        }
+        
+        UnicodeString fragment;
+        UnicodeString normalizedFragment;
+        for (int32_t srcI = 0; srcI < inString.length();) {  // Once per normalization chunk
+            fragment.remove();
+            int32_t fragmentStartI = srcI;
+            UChar32 c = inString.char32At(srcI);
+            for (;;) {
+                fragment.append(c);
+                srcI = inString.moveIndex32(srcI, 1);
+                if (srcI == inString.length()) {
+                    break;
+                }
+                c = inString.char32At(srcI);
+                if (nfkcNorm2->hasBoundaryBefore(c)) {
+                    break;
+                }
+            }
+            nfkcNorm2->normalize(fragment, normalizedFragment, status);
+            normalizedInput.append(normalizedFragment);
+
+            // Map every position in the normalized chunk to the start of the chunk
+            //   in the original input.
+            int32_t fragmentOriginalStart = inputMap.isValid() ?
+                    inputMap->elementAti(fragmentStartI) : fragmentStartI+rangeStart;
+            while (normalizedMap->size() < normalizedInput.length()) {
+                normalizedMap->addElement(fragmentOriginalStart, status);
+                if (U_FAILURE(status)) {
+                    break;
+                }
+            }
+        }
+        U_ASSERT(normalizedMap->size() == normalizedInput.length());
+        int32_t nativeEnd = inputMap.isValid() ?
+                inputMap->elementAti(inString.length()) : inString.length()+rangeStart;
+        normalizedMap->addElement(nativeEnd, status);
+
+        inputMap = std::move(normalizedMap);
+        inString = std::move(normalizedInput);
+    }
+
+    int32_t numCodePts = inString.countChar32();
+    if (numCodePts != inString.length()) {
+        // There are supplementary characters in the input.
+        // The dictionary will produce boundary positions in terms of code point indexes,
+        //   not in terms of code unit string indexes.
+        // Use the inputMap mechanism to take care of this in addition to indexing differences
+        //    from normalization and/or UTF-8 input.
+        UBool hadExistingMap = inputMap.isValid();
+        if (!hadExistingMap) {
+            inputMap.adoptInsteadAndCheckErrorCode(new UVector32(status), status);
+            if (U_FAILURE(status)) {
+                return 0;
+            }
+        }
+        int32_t cpIdx = 0;
+        for (int32_t cuIdx = 0; ; cuIdx = inString.moveIndex32(cuIdx, 1)) {
+            U_ASSERT(cuIdx >= cpIdx);
+            if (hadExistingMap) {
+                inputMap->setElementAt(inputMap->elementAti(cuIdx), cpIdx);
+            } else {
+                inputMap->addElement(cuIdx+rangeStart, status);
+            }
+            cpIdx++;
+            if (cuIdx == inString.length()) {
+               break;
+            }
+        }
+    }
+                
+    // bestSnlp[i] is the snlp of the best segmentation of the first i
+    // code points in the range to be matched.
+    UVector32 bestSnlp(numCodePts + 1, status);
+    bestSnlp.addElement(0, status);
+    for(int32_t i = 1; i <= numCodePts; i++) {
+        bestSnlp.addElement(kuint32max, status);
+    }
+
+
+    // prev[i] is the index of the last CJK code point in the previous word in 
+    // the best segmentation of the first i characters.
+    UVector32 prev(numCodePts + 1, status);
+    for(int32_t i = 0; i <= numCodePts; i++){
+        prev.addElement(-1, status);
+    }
+
+    const int32_t maxWordSize = 20;
+    UVector32 values(numCodePts, status);
+    values.setSize(numCodePts);
+    UVector32 lengths(numCodePts, status);
+    lengths.setSize(numCodePts);
+
+    UText fu = UTEXT_INITIALIZER;
+    utext_openUnicodeString(&fu, &inString, &status);
+
+    // Dynamic programming to find the best segmentation.
+
+    // In outer loop, i  is the code point index,
+    //                ix is the corresponding string (code unit) index.
+    //    They differ when the string contains supplementary characters.
+    int32_t ix = 0;
+    bool is_prev_katakana = false;
+    for (int32_t i = 0;  i < numCodePts;  ++i, ix = inString.moveIndex32(ix, 1)) {
+        if ((uint32_t)bestSnlp.elementAti(i) == kuint32max) {
+            continue;
+        }
+
+        int32_t count;
+        utext_setNativeIndex(&fu, ix);
+        count = fDictionary->matches(&fu, maxWordSize, numCodePts,
+                             NULL, lengths.getBuffer(), values.getBuffer(), NULL);
+                             // Note: lengths is filled with code point lengths
+                             //       The NULL parameter is the ignored code unit lengths.
+
+        // if there are no single character matches found in the dictionary 
+        // starting with this character, treat character as a 1-character word 
+        // with the highest value possible, i.e. the least likely to occur.
+        // Exclude Korean characters from this treatment, as they should be left
+        // together by default.
+        if ((count == 0 || lengths.elementAti(0) != 1) &&
+                !fHangulWordSet.contains(inString.char32At(ix))) {
+            values.setElementAt(maxSnlp, count);   // 255
+            lengths.setElementAt(1, count++);
+        }
+
+        for (int32_t j = 0; j < count; j++) {
+            uint32_t newSnlp = (uint32_t)bestSnlp.elementAti(i) + (uint32_t)values.elementAti(j);
+            int32_t ln_j_i = lengths.elementAti(j) + i;
+            if (newSnlp < (uint32_t)bestSnlp.elementAti(ln_j_i)) {
+                bestSnlp.setElementAt(newSnlp, ln_j_i);
+                prev.setElementAt(i, ln_j_i);
+            }
+        }
+
+        // In Japanese,
+        // Katakana word in single character is pretty rare. So we apply
+        // the following heuristic to Katakana: any continuous run of Katakana
+        // characters is considered a candidate word with a default cost
+        // specified in the katakanaCost table according to its length.
+
+        bool is_katakana = isKatakana(inString.char32At(ix));
+        int32_t katakanaRunLength = 1;
+        if (!is_prev_katakana && is_katakana) {
+            int32_t j = inString.moveIndex32(ix, 1);
+            // Find the end of the continuous run of Katakana characters
+            while (j < inString.length() && katakanaRunLength < kMaxKatakanaGroupLength &&
+                    isKatakana(inString.char32At(j))) {
+                j = inString.moveIndex32(j, 1);
+                katakanaRunLength++;
+            }
+            if (katakanaRunLength < kMaxKatakanaGroupLength) {
+                uint32_t newSnlp = bestSnlp.elementAti(i) + getKatakanaCost(katakanaRunLength);
+                if (newSnlp < (uint32_t)bestSnlp.elementAti(i+katakanaRunLength)) {
+                    bestSnlp.setElementAt(newSnlp, i+katakanaRunLength);
+                    prev.setElementAt(i, i+katakanaRunLength);  // prev[j] = i;
+                }
+            }
+        }
+        is_prev_katakana = is_katakana;
+    }
+    utext_close(&fu);
+
+    // Start pushing the optimal offset index into t_boundary (t for tentative).
+    // prev[numCodePts] is guaranteed to be meaningful.
+    // We'll first push in the reverse order, i.e.,
+    // t_boundary[0] = numCodePts, and afterwards do a swap.
+    UVector32 t_boundary(numCodePts+1, status);
+
+    int32_t numBreaks = 0;
+    // No segmentation found, set boundary to end of range
+    if ((uint32_t)bestSnlp.elementAti(numCodePts) == kuint32max) {
+        t_boundary.addElement(numCodePts, status);
+        numBreaks++;
+    } else {
+        for (int32_t i = numCodePts; i > 0; i = prev.elementAti(i)) {
+            t_boundary.addElement(i, status);
+            numBreaks++;
+        }
+        U_ASSERT(prev.elementAti(t_boundary.elementAti(numBreaks - 1)) == 0);
+    }
+
+    // Add a break for the start of the dictionary range if there is not one
+    // there already.
+    if (foundBreaks.size() == 0 || foundBreaks.peeki() < rangeStart) {
+        t_boundary.addElement(0, status);
+        numBreaks++;
+    }
+
+    // Now that we're done, convert positions in t_boundary[] (indices in 
+    // the normalized input string) back to indices in the original input UText
+    // while reversing t_boundary and pushing values to foundBreaks.
+    int32_t prevCPPos = -1;
+    int32_t prevUTextPos = -1;
+    for (int32_t i = numBreaks-1; i >= 0; i--) {
+        int32_t cpPos = t_boundary.elementAti(i);
+        U_ASSERT(cpPos > prevCPPos);
+        int32_t utextPos =  inputMap.isValid() ? inputMap->elementAti(cpPos) : cpPos + rangeStart;
+        U_ASSERT(utextPos >= prevUTextPos);
+        if (utextPos > prevUTextPos) {
+            // Boundaries are added to foundBreaks output in ascending order.
+            U_ASSERT(foundBreaks.size() == 0 || foundBreaks.peeki() < utextPos);
+            foundBreaks.push(utextPos, status);
+        } else {
+            // Normalization expanded the input text, the dictionary found a boundary
+            // within the expansion, giving two boundaries with the same index in the
+            // original text. Ignore the second. See ticket #12918.
+            --numBreaks;
+        }
+        prevCPPos = cpPos;
+        prevUTextPos = utextPos;
+    }
+    (void)prevCPPos; // suppress compiler warnings about unused variable
+
+    // inString goes out of scope
+    // inputMap goes out of scope
+    return numBreaks;
+}
+#endif
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
diff --git a/thirdparty/icu4c/common/dictbe.h b/thirdparty/icu4c/common/dictbe.h
new file mode 100644
index 0000000000..4ea676fc71
--- /dev/null
+++ b/thirdparty/icu4c/common/dictbe.h
@@ -0,0 +1,402 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2006-2014, International Business Machines Corporation   *
+ * and others. All Rights Reserved.                                            *
+ *******************************************************************************
+ */
+
+#ifndef DICTBE_H
+#define DICTBE_H
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+#include "unicode/utext.h"
+
+#include "brkeng.h"
+#include "uvectr32.h"
+
+U_NAMESPACE_BEGIN
+
+class DictionaryMatcher;
+class Normalizer2;
+
+/*******************************************************************
+ * DictionaryBreakEngine
+ */
+
+/**
+ * <p>DictionaryBreakEngine is a kind of LanguageBreakEngine that uses a
+ * dictionary to determine language-specific breaks.</p>
+ *
+ * <p>After it is constructed a DictionaryBreakEngine may be shared between
+ * threads without synchronization.</p>
+ */
+class DictionaryBreakEngine : public LanguageBreakEngine {
+ private:
+    /**
+     * The set of characters handled by this engine
+     * @internal
+     */
+
+  UnicodeSet    fSet;
+
+ public:
+
+  /**
+   * <p>Constructor </p>
+   */
+  DictionaryBreakEngine();
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~DictionaryBreakEngine();
+
+  /**
+   * <p>Indicate whether this engine handles a particular character for
+   * a particular kind of break.</p>
+   *
+   * @param c A character which begins a run that the engine might handle
+   * @return true if this engine handles the particular character and break
+   * type.
+   */
+  virtual UBool handles(UChar32 c) const;
+
+  /**
+   * <p>Find any breaks within a run in the supplied text.</p>
+   *
+   * @param text A UText representing the text. The iterator is left at
+   * the end of the run of characters which the engine is capable of handling 
+   * that starts from the first character in the range.
+   * @param startPos The start of the run within the supplied text.
+   * @param endPos The end of the run within the supplied text.
+   * @param foundBreaks vector of int32_t to receive the break positions
+   * @return The number of breaks found.
+   */
+  virtual int32_t findBreaks( UText *text,
+                              int32_t startPos,
+                              int32_t endPos,
+                              UVector32 &foundBreaks ) const;
+
+ protected:
+
+ /**
+  * <p>Set the character set handled by this engine.</p>
+  *
+  * @param set A UnicodeSet of the set of characters handled by the engine
+  */
+  virtual void setCharacters( const UnicodeSet &set );
+
+ /**
+  * <p>Divide up a range of known dictionary characters handled by this break engine.</p>
+  *
+  * @param text A UText representing the text
+  * @param rangeStart The start of the range of dictionary characters
+  * @param rangeEnd The end of the range of dictionary characters
+  * @param foundBreaks Output of C array of int32_t break positions, or 0
+  * @return The number of breaks found
+  */
+  virtual int32_t divideUpDictionaryRange( UText *text,
+                                           int32_t rangeStart,
+                                           int32_t rangeEnd,
+                                           UVector32 &foundBreaks ) const = 0;
+
+};
+
+/*******************************************************************
+ * ThaiBreakEngine
+ */
+
+/**
+ * <p>ThaiBreakEngine is a kind of DictionaryBreakEngine that uses a
+ * dictionary and heuristics to determine Thai-specific breaks.</p>
+ *
+ * <p>After it is constructed a ThaiBreakEngine may be shared between
+ * threads without synchronization.</p>
+ */
+class ThaiBreakEngine : public DictionaryBreakEngine {
+ private:
+    /**
+     * The set of characters handled by this engine
+     * @internal
+     */
+
+  UnicodeSet                fThaiWordSet;
+  UnicodeSet                fEndWordSet;
+  UnicodeSet                fBeginWordSet;
+  UnicodeSet                fSuffixSet;
+  UnicodeSet                fMarkSet;
+  DictionaryMatcher  *fDictionary;
+
+ public:
+
+  /**
+   * <p>Default constructor.</p>
+   *
+   * @param adoptDictionary A DictionaryMatcher to adopt. Deleted when the
+   * engine is deleted.
+   */
+  ThaiBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status);
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~ThaiBreakEngine();
+
+ protected:
+ /**
+  * <p>Divide up a range of known dictionary characters handled by this break engine.</p>
+  *
+  * @param text A UText representing the text
+  * @param rangeStart The start of the range of dictionary characters
+  * @param rangeEnd The end of the range of dictionary characters
+  * @param foundBreaks Output of C array of int32_t break positions, or 0
+  * @return The number of breaks found
+  */
+  virtual int32_t divideUpDictionaryRange( UText *text,
+                                           int32_t rangeStart,
+                                           int32_t rangeEnd,
+                                           UVector32 &foundBreaks ) const;
+
+};
+
+/*******************************************************************
+ * LaoBreakEngine
+ */
+
+/**
+ * <p>LaoBreakEngine is a kind of DictionaryBreakEngine that uses a
+ * dictionary and heuristics to determine Lao-specific breaks.</p>
+ *
+ * <p>After it is constructed a LaoBreakEngine may be shared between
+ * threads without synchronization.</p>
+ */
+class LaoBreakEngine : public DictionaryBreakEngine {
+ private:
+    /**
+     * The set of characters handled by this engine
+     * @internal
+     */
+
+  UnicodeSet                fLaoWordSet;
+  UnicodeSet                fEndWordSet;
+  UnicodeSet                fBeginWordSet;
+  UnicodeSet                fMarkSet;
+  DictionaryMatcher  *fDictionary;
+
+ public:
+
+  /**
+   * <p>Default constructor.</p>
+   *
+   * @param adoptDictionary A DictionaryMatcher to adopt. Deleted when the
+   * engine is deleted.
+   */
+  LaoBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status);
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~LaoBreakEngine();
+
+ protected:
+ /**
+  * <p>Divide up a range of known dictionary characters handled by this break engine.</p>
+  *
+  * @param text A UText representing the text
+  * @param rangeStart The start of the range of dictionary characters
+  * @param rangeEnd The end of the range of dictionary characters
+  * @param foundBreaks Output of C array of int32_t break positions, or 0
+  * @return The number of breaks found
+  */
+  virtual int32_t divideUpDictionaryRange( UText *text,
+                                           int32_t rangeStart,
+                                           int32_t rangeEnd,
+                                           UVector32 &foundBreaks ) const;
+
+};
+
+/******************************************************************* 
+ * BurmeseBreakEngine 
+ */ 
+ 
+/** 
+ * <p>BurmeseBreakEngine is a kind of DictionaryBreakEngine that uses a 
+ * DictionaryMatcher and heuristics to determine Burmese-specific breaks.</p> 
+ * 
+ * <p>After it is constructed a BurmeseBreakEngine may be shared between 
+ * threads without synchronization.</p> 
+ */ 
+class BurmeseBreakEngine : public DictionaryBreakEngine { 
+ private: 
+    /** 
+     * The set of characters handled by this engine 
+     * @internal 
+     */ 
+ 
+  UnicodeSet                fBurmeseWordSet; 
+  UnicodeSet                fEndWordSet; 
+  UnicodeSet                fBeginWordSet; 
+  UnicodeSet                fMarkSet; 
+  DictionaryMatcher  *fDictionary; 
+ 
+ public: 
+ 
+  /** 
+   * <p>Default constructor.</p> 
+   * 
+   * @param adoptDictionary A DictionaryMatcher to adopt. Deleted when the 
+   * engine is deleted. 
+   */ 
+  BurmeseBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status); 
+ 
+  /** 
+   * <p>Virtual destructor.</p> 
+   */ 
+  virtual ~BurmeseBreakEngine(); 
+ 
+ protected: 
+ /** 
+  * <p>Divide up a range of known dictionary characters.</p> 
+  * 
+  * @param text A UText representing the text 
+  * @param rangeStart The start of the range of dictionary characters 
+  * @param rangeEnd The end of the range of dictionary characters 
+  * @param foundBreaks Output of C array of int32_t break positions, or 0 
+  * @return The number of breaks found 
+  */ 
+  virtual int32_t divideUpDictionaryRange( UText *text, 
+                                           int32_t rangeStart, 
+                                           int32_t rangeEnd, 
+                                           UVector32 &foundBreaks ) const; 
+ 
+}; 
+ 
+/******************************************************************* 
+ * KhmerBreakEngine 
+ */ 
+ 
+/** 
+ * <p>KhmerBreakEngine is a kind of DictionaryBreakEngine that uses a 
+ * DictionaryMatcher and heuristics to determine Khmer-specific breaks.</p> 
+ * 
+ * <p>After it is constructed a KhmerBreakEngine may be shared between 
+ * threads without synchronization.</p> 
+ */ 
+class KhmerBreakEngine : public DictionaryBreakEngine { 
+ private: 
+    /** 
+     * The set of characters handled by this engine 
+     * @internal 
+     */ 
+ 
+  UnicodeSet                fKhmerWordSet; 
+  UnicodeSet                fEndWordSet; 
+  UnicodeSet                fBeginWordSet; 
+  UnicodeSet                fMarkSet; 
+  DictionaryMatcher  *fDictionary; 
+ 
+ public: 
+ 
+  /** 
+   * <p>Default constructor.</p> 
+   * 
+   * @param adoptDictionary A DictionaryMatcher to adopt. Deleted when the 
+   * engine is deleted. 
+   */ 
+  KhmerBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &status); 
+ 
+  /** 
+   * <p>Virtual destructor.</p> 
+   */ 
+  virtual ~KhmerBreakEngine(); 
+ 
+ protected: 
+ /** 
+  * <p>Divide up a range of known dictionary characters.</p> 
+  * 
+  * @param text A UText representing the text 
+  * @param rangeStart The start of the range of dictionary characters 
+  * @param rangeEnd The end of the range of dictionary characters 
+  * @param foundBreaks Output of C array of int32_t break positions, or 0 
+  * @return The number of breaks found 
+  */ 
+  virtual int32_t divideUpDictionaryRange( UText *text, 
+                                           int32_t rangeStart, 
+                                           int32_t rangeEnd, 
+                                           UVector32 &foundBreaks ) const; 
+ 
+}; 
+ 
+#if !UCONFIG_NO_NORMALIZATION
+
+/*******************************************************************
+ * CjkBreakEngine
+ */
+
+//indicates language/script that the CjkBreakEngine will handle
+enum LanguageType {
+    kKorean,
+    kChineseJapanese
+};
+
+/**
+ * <p>CjkBreakEngine is a kind of DictionaryBreakEngine that uses a
+ * dictionary with costs associated with each word and
+ * Viterbi decoding to determine CJK-specific breaks.</p>
+ */
+class CjkBreakEngine : public DictionaryBreakEngine {
+ protected:
+    /**
+     * The set of characters handled by this engine
+     * @internal
+     */
+  UnicodeSet                fHangulWordSet;
+  UnicodeSet                fHanWordSet;
+  UnicodeSet                fKatakanaWordSet;
+  UnicodeSet                fHiraganaWordSet;
+
+  DictionaryMatcher        *fDictionary;
+  const Normalizer2        *nfkcNorm2;
+
+ public:
+
+    /**
+     * <p>Default constructor.</p>
+     *
+     * @param adoptDictionary A DictionaryMatcher to adopt. Deleted when the
+     * engine is deleted. The DictionaryMatcher must contain costs for each word
+     * in order for the dictionary to work properly.
+     */
+  CjkBreakEngine(DictionaryMatcher *adoptDictionary, LanguageType type, UErrorCode &status);
+
+    /**
+     * <p>Virtual destructor.</p>
+     */
+  virtual ~CjkBreakEngine();
+
+ protected:
+    /**
+     * <p>Divide up a range of known dictionary characters handled by this break engine.</p>
+     *
+     * @param text A UText representing the text
+     * @param rangeStart The start of the range of dictionary characters
+     * @param rangeEnd The end of the range of dictionary characters
+     * @param foundBreaks Output of C array of int32_t break positions, or 0
+     * @return The number of breaks found
+     */
+  virtual int32_t divideUpDictionaryRange( UText *text,
+          int32_t rangeStart,
+          int32_t rangeEnd,
+          UVector32 &foundBreaks ) const;
+
+};
+
+#endif
+
+U_NAMESPACE_END
+
+    /* DICTBE_H */
+#endif
diff --git a/thirdparty/icu4c/common/dictionarydata.cpp b/thirdparty/icu4c/common/dictionarydata.cpp
new file mode 100644
index 0000000000..6e2dbee5b6
--- /dev/null
+++ b/thirdparty/icu4c/common/dictionarydata.cpp
@@ -0,0 +1,242 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2014-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* dictionarydata.h
+*
+* created on: 2012may31
+* created by: Markus W. Scherer & Maxime Serrano
+*/
+
+#include "dictionarydata.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/bytestrie.h"
+#include "unicode/udata.h"
+#include "cmemory.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+U_NAMESPACE_BEGIN
+
+const int32_t  DictionaryData::TRIE_TYPE_BYTES = 0;
+const int32_t  DictionaryData::TRIE_TYPE_UCHARS = 1;
+const int32_t  DictionaryData::TRIE_TYPE_MASK = 7;
+const int32_t  DictionaryData::TRIE_HAS_VALUES = 8;
+
+const int32_t  DictionaryData::TRANSFORM_NONE = 0;
+const int32_t  DictionaryData::TRANSFORM_TYPE_OFFSET = 0x1000000;
+const int32_t  DictionaryData::TRANSFORM_TYPE_MASK = 0x7f000000;
+const int32_t  DictionaryData::TRANSFORM_OFFSET_MASK = 0x1fffff;
+    
+DictionaryMatcher::~DictionaryMatcher() {
+}
+
+UCharsDictionaryMatcher::~UCharsDictionaryMatcher() {
+    udata_close(file);
+}
+
+int32_t UCharsDictionaryMatcher::getType() const {
+    return DictionaryData::TRIE_TYPE_UCHARS;
+}
+
+int32_t UCharsDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const {
+
+    UCharsTrie uct(characters);
+    int32_t startingTextIndex = (int32_t)utext_getNativeIndex(text);
+    int32_t wordCount = 0;
+    int32_t codePointsMatched = 0;
+
+    for (UChar32 c = utext_next32(text); c >= 0; c=utext_next32(text)) {
+        UStringTrieResult result = (codePointsMatched == 0) ? uct.first(c) : uct.next(c);
+        int32_t lengthMatched = (int32_t)utext_getNativeIndex(text) - startingTextIndex;
+        codePointsMatched += 1;
+        if (USTRINGTRIE_HAS_VALUE(result)) {
+            if (wordCount < limit) {
+                if (values != NULL) {
+                    values[wordCount] = uct.getValue();
+                }
+                if (lengths != NULL) {
+                    lengths[wordCount] = lengthMatched;
+                }
+                if (cpLengths != NULL) {
+                    cpLengths[wordCount] = codePointsMatched;
+                }
+                ++wordCount;
+            }
+            if (result == USTRINGTRIE_FINAL_VALUE) {
+                break;
+            }
+        }
+        else if (result == USTRINGTRIE_NO_MATCH) {
+            break;
+        }
+        if (lengthMatched >= maxLength) {
+            break;
+        }
+    }
+
+    if (prefix != NULL) {
+        *prefix = codePointsMatched;
+    }
+    return wordCount;
+}
+
+BytesDictionaryMatcher::~BytesDictionaryMatcher() {
+    udata_close(file);
+}
+
+UChar32 BytesDictionaryMatcher::transform(UChar32 c) const {
+    if ((transformConstant & DictionaryData::TRANSFORM_TYPE_MASK) == DictionaryData::TRANSFORM_TYPE_OFFSET) {
+        if (c == 0x200D) {
+            return 0xFF;
+        } else if (c == 0x200C) {
+            return 0xFE;
+        }
+        int32_t delta = c - (transformConstant & DictionaryData::TRANSFORM_OFFSET_MASK);
+        if (delta < 0 || 0xFD < delta) {
+            return U_SENTINEL;
+        }
+        return (UChar32)delta;
+    }
+    return c;
+}
+
+int32_t BytesDictionaryMatcher::getType() const {
+    return DictionaryData::TRIE_TYPE_BYTES;
+}
+
+int32_t BytesDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const {
+    BytesTrie bt(characters);
+    int32_t startingTextIndex = (int32_t)utext_getNativeIndex(text);
+    int32_t wordCount = 0;
+    int32_t codePointsMatched = 0;
+
+    for (UChar32 c = utext_next32(text); c >= 0; c=utext_next32(text)) {
+        UStringTrieResult result = (codePointsMatched == 0) ? bt.first(transform(c)) : bt.next(transform(c));
+        int32_t lengthMatched = (int32_t)utext_getNativeIndex(text) - startingTextIndex;
+        codePointsMatched += 1;
+        if (USTRINGTRIE_HAS_VALUE(result)) {
+            if (wordCount < limit) {
+                if (values != NULL) {
+                    values[wordCount] = bt.getValue();
+                }
+                if (lengths != NULL) {
+                    lengths[wordCount] = lengthMatched;
+                }
+                if (cpLengths != NULL) {
+                    cpLengths[wordCount] = codePointsMatched;
+                }
+                ++wordCount;
+            }
+            if (result == USTRINGTRIE_FINAL_VALUE) {
+                break;
+            }
+        }
+        else if (result == USTRINGTRIE_NO_MATCH) {
+            break;
+        }
+        if (lengthMatched >= maxLength) {
+            break;
+        }
+    }
+
+    if (prefix != NULL) {
+        *prefix = codePointsMatched;
+    }
+    return wordCount;
+}
+
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI int32_t U_EXPORT2
+udict_swap(const UDataSwapper *ds, const void *inData, int32_t length,
+           void *outData, UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+    const int32_t *inIndexes;
+    int32_t indexes[DictionaryData::IX_COUNT];
+    int32_t i, offset, size;
+
+    headerSize = udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if (pErrorCode == NULL || U_FAILURE(*pErrorCode)) return 0;
+    pInfo = (const UDataInfo *)((const char *)inData + 4);
+    if (!(pInfo->dataFormat[0] == 0x44 && 
+          pInfo->dataFormat[1] == 0x69 && 
+          pInfo->dataFormat[2] == 0x63 && 
+          pInfo->dataFormat[3] == 0x74 && 
+          pInfo->formatVersion[0] == 1)) {
+        udata_printError(ds, "udict_swap(): data format %02x.%02x.%02x.%02x (format version %02x) is not recognized as dictionary data\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1], pInfo->dataFormat[2], pInfo->dataFormat[3], pInfo->formatVersion[0]);
+        *pErrorCode = U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes = (const uint8_t *)inData + headerSize;
+    outBytes = (uint8_t *)outData + headerSize;
+
+    inIndexes = (const int32_t *)inBytes;
+    if (length >= 0) {
+        length -= headerSize;
+        if (length < (int32_t)(sizeof(indexes))) {
+            udata_printError(ds, "udict_swap(): too few bytes (%d after header) for dictionary data\n", length);
+            *pErrorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    for (i = 0; i < DictionaryData::IX_COUNT; i++) {
+        indexes[i] = udata_readInt32(ds, inIndexes[i]);
+    }
+
+    size = indexes[DictionaryData::IX_TOTAL_SIZE];
+
+    if (length >= 0) {
+        if (length < size) {
+            udata_printError(ds, "udict_swap(): too few bytes (%d after header) for all of dictionary data\n", length);
+            *pErrorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        if (inBytes != outBytes) {
+            uprv_memcpy(outBytes, inBytes, size);
+        }
+
+        offset = 0;
+        ds->swapArray32(ds, inBytes, sizeof(indexes), outBytes, pErrorCode);
+        offset = (int32_t)sizeof(indexes);
+        int32_t trieType = indexes[DictionaryData::IX_TRIE_TYPE] & DictionaryData::TRIE_TYPE_MASK;
+        int32_t nextOffset = indexes[DictionaryData::IX_RESERVED1_OFFSET];
+
+        if (trieType == DictionaryData::TRIE_TYPE_UCHARS) {
+            ds->swapArray16(ds, inBytes + offset, nextOffset - offset, outBytes + offset, pErrorCode);
+        } else if (trieType == DictionaryData::TRIE_TYPE_BYTES) {
+            // nothing to do
+        } else {
+            udata_printError(ds, "udict_swap(): unknown trie type!\n");
+            *pErrorCode = U_UNSUPPORTED_ERROR;
+            return 0;
+        }
+
+        // these next two sections are empty in the current format,
+        // but may be used later.
+        offset = nextOffset;
+        nextOffset = indexes[DictionaryData::IX_RESERVED2_OFFSET];
+        offset = nextOffset;
+        nextOffset = indexes[DictionaryData::IX_TOTAL_SIZE];
+        offset = nextOffset;
+    }
+    return headerSize + size;
+}
+#endif
diff --git a/thirdparty/icu4c/common/dictionarydata.h b/thirdparty/icu4c/common/dictionarydata.h
new file mode 100644
index 0000000000..0d303d9a8d
--- /dev/null
+++ b/thirdparty/icu4c/common/dictionarydata.h
@@ -0,0 +1,191 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* dictionarydata.h
+*
+* created on: 2012may31
+* created by: Markus W. Scherer & Maxime Serrano
+*/
+
+#ifndef __DICTIONARYDATA_H__
+#define __DICTIONARYDATA_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/utext.h"
+#include "unicode/udata.h"
+#include "udataswp.h"
+#include "unicode/uobject.h"
+#include "unicode/ustringtrie.h"
+
+U_NAMESPACE_BEGIN
+
+class UCharsTrie;
+class BytesTrie;
+
+class U_COMMON_API DictionaryData : public UMemory {
+public:
+    static const int32_t TRIE_TYPE_BYTES; // = 0;
+    static const int32_t TRIE_TYPE_UCHARS; // = 1;
+    static const int32_t TRIE_TYPE_MASK; // = 7;
+    static const int32_t TRIE_HAS_VALUES; // = 8;
+
+    static const int32_t TRANSFORM_NONE; // = 0;
+    static const int32_t TRANSFORM_TYPE_OFFSET; // = 0x1000000;
+    static const int32_t TRANSFORM_TYPE_MASK; // = 0x7f000000;
+    static const int32_t TRANSFORM_OFFSET_MASK; // = 0x1fffff;
+
+    enum {
+        // Byte offsets from the start of the data, after the generic header.
+        IX_STRING_TRIE_OFFSET,
+        IX_RESERVED1_OFFSET,
+        IX_RESERVED2_OFFSET,
+        IX_TOTAL_SIZE,
+
+        // Trie type: TRIE_HAS_VALUES | TRIE_TYPE_BYTES etc.
+        IX_TRIE_TYPE,
+        // Transform specification: TRANSFORM_TYPE_OFFSET | 0xe00 etc.
+        IX_TRANSFORM,
+
+        IX_RESERVED6,
+        IX_RESERVED7,
+        IX_COUNT
+    };
+};
+
+/**
+ * Wrapper class around generic dictionaries, implementing matches().
+ * getType() should return a TRIE_TYPE_??? constant from DictionaryData.
+ * 
+ * All implementations of this interface must be thread-safe if they are to be used inside of the
+ * dictionary-based break iteration code.
+ */
+class U_COMMON_API DictionaryMatcher : public UMemory {
+public:
+    DictionaryMatcher() {}
+    virtual ~DictionaryMatcher();
+    // this should emulate CompactTrieDictionary::matches()
+    /*  @param text      The text in which to look for matching words. Matching begins
+     *                   at the current position of the UText.
+     *  @param maxLength The max length of match to consider. Units are the native indexing
+     *                   units of the UText.
+     *  @param limit     Capacity of output arrays, which is also the maximum number of
+     *                   matching words to be found.
+     *  @param lengths   output array, filled with the lengths of the matches, in order,
+     *                   from shortest to longest. Lengths are in native indexing units
+     *                   of the UText. May be NULL.
+     *  @param cpLengths output array, filled with the lengths of the matches, in order,
+     *                   from shortest to longest. Lengths are the number of Unicode code points.
+     *                   May be NULL.
+     *  @param values    Output array, filled with the values associated with the words found.
+     *                   May be NULL.
+     *  @param prefix    Output parameter, the code point length of the prefix match, even if that
+     *                   prefix didn't lead to a complete word. Will always be >= the cpLength
+     *                   of the longest complete word matched. May be NULL.
+     *  @return          Number of matching words found.
+     */
+    virtual int32_t matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const = 0;
+
+    /** @return DictionaryData::TRIE_TYPE_XYZ */
+    virtual int32_t getType() const = 0;
+};
+
+// Implementation of the DictionaryMatcher interface for a UCharsTrie dictionary
+class U_COMMON_API UCharsDictionaryMatcher : public DictionaryMatcher {
+public:
+    // constructs a new UCharsDictionaryMatcher.
+    // The UDataMemory * will be closed on this object's destruction.
+    UCharsDictionaryMatcher(const UChar *c, UDataMemory *f) : characters(c), file(f) { }
+    virtual ~UCharsDictionaryMatcher();
+    virtual int32_t matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const;
+    virtual int32_t getType() const;
+private:
+    const UChar *characters;
+    UDataMemory *file;
+};
+
+// Implementation of the DictionaryMatcher interface for a BytesTrie dictionary
+class U_COMMON_API BytesDictionaryMatcher : public DictionaryMatcher {
+public:
+    // constructs a new BytesTrieDictionaryMatcher
+    // the transform constant should be the constant read from the file, not a masked version!
+    // the UDataMemory * fed in here will be closed on this object's destruction
+    BytesDictionaryMatcher(const char *c, int32_t t, UDataMemory *f)
+            : characters(c), transformConstant(t), file(f) { }
+    virtual ~BytesDictionaryMatcher();
+    virtual int32_t matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const;
+    virtual int32_t getType() const;
+private:
+    UChar32 transform(UChar32 c) const;
+
+    const char *characters;
+    int32_t transformConstant;
+    UDataMemory *file;
+};
+
+U_NAMESPACE_END
+
+U_CAPI int32_t U_EXPORT2
+udict_swap(const UDataSwapper *ds, const void *inData, int32_t length, void *outData, UErrorCode *pErrorCode);
+
+/**
+ * Format of dictionary .dict data files.
+ * Format version 1.0.
+ *
+ * A dictionary .dict data file contains a byte-serialized BytesTrie or
+ * a UChars-serialized UCharsTrie.
+ * Such files are used in dictionary-based break iteration (DBBI).
+ *
+ * For a BytesTrie, a transformation type is specified for
+ * transforming Unicode strings into byte sequences.
+ *
+ * A .dict file begins with a standard ICU data file header
+ * (DataHeader, see ucmndata.h and unicode/udata.h).
+ * The UDataInfo.dataVersion field is currently unused (set to 0.0.0.0).
+ *
+ * After the header, the file contains the following parts.
+ * Constants are defined in the DictionaryData class.
+ *
+ * For the data structure of BytesTrie & UCharsTrie see
+ * http://site.icu-project.org/design/struct/tries
+ * and the bytestrie.h and ucharstrie.h header files.
+ *
+ * int32_t indexes[indexesLength]; -- indexesLength=indexes[IX_STRING_TRIE_OFFSET]/4;
+ *
+ *      The first four indexes are byte offsets in ascending order.
+ *      Each byte offset marks the start of the next part in the data file,
+ *      and the end of the previous one.
+ *      When two consecutive byte offsets are the same, then the corresponding part is empty.
+ *      Byte offsets are offsets from after the header,
+ *      that is, from the beginning of the indexes[].
+ *      Each part starts at an offset with proper alignment for its data.
+ *      If necessary, the previous part may include padding bytes to achieve this alignment.
+ *
+ *      trieType=indexes[IX_TRIE_TYPE] defines the trie type.
+ *      transform=indexes[IX_TRANSFORM] defines the Unicode-to-bytes transformation.
+ *          If the transformation type is TRANSFORM_TYPE_OFFSET,
+ *          then the lower 21 bits contain the offset code point.
+ *          Each code point c is mapped to byte b = (c - offset).
+ *          Code points outside the range offset..(offset+0xff) cannot be mapped
+ *          and do not occur in the dictionary.
+ *
+ * stringTrie; -- a serialized BytesTrie or UCharsTrie
+ *
+ *      The dictionary maps strings to specific values (TRIE_HAS_VALUES bit set in trieType),
+ *      or it maps all strings to 0 (TRIE_HAS_VALUES bit not set).
+ */
+
+#endif  /* !UCONFIG_NO_BREAK_ITERATION */
+#endif  /* __DICTIONARYDATA_H__ */
diff --git a/thirdparty/icu4c/common/dtintrv.cpp b/thirdparty/icu4c/common/dtintrv.cpp
new file mode 100644
index 0000000000..80bb5d6dbd
--- /dev/null
+++ b/thirdparty/icu4c/common/dtintrv.cpp
@@ -0,0 +1,63 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*******************************************************************************
+* Copyright (C) 2008, International Business Machines Corporation and
+* others. All Rights Reserved.
+*******************************************************************************
+*
+* File DTINTRV.CPP 
+*
+*******************************************************************************
+*/
+
+
+
+#include "unicode/dtintrv.h"
+
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(DateInterval)
+
+//DateInterval::DateInterval(){}
+
+
+DateInterval::DateInterval(UDate from, UDate to)
+:   fromDate(from),
+    toDate(to)
+{}
+
+
+DateInterval::~DateInterval(){}
+
+
+DateInterval::DateInterval(const DateInterval& other)
+: UObject(other) {
+    *this = other;
+}   
+
+
+DateInterval&
+DateInterval::operator=(const DateInterval& other) {
+    if ( this != &other ) {
+        fromDate = other.fromDate;
+        toDate = other.toDate;
+    }
+    return *this;
+}
+
+
+DateInterval* 
+DateInterval::clone() const {
+    return new DateInterval(*this);
+}
+
+
+UBool 
+DateInterval::operator==(const DateInterval& other) const { 
+    return ( fromDate == other.fromDate && toDate == other.toDate );
+}
+
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/edits.cpp b/thirdparty/icu4c/common/edits.cpp
new file mode 100644
index 0000000000..95f0c19a72
--- /dev/null
+++ b/thirdparty/icu4c/common/edits.cpp
@@ -0,0 +1,803 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// edits.cpp
+// created: 2017feb08 Markus W. Scherer
+
+#include "unicode/edits.h"
+#include "unicode/unistr.h"
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "util.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+// 0000uuuuuuuuuuuu records u+1 unchanged text units.
+const int32_t MAX_UNCHANGED_LENGTH = 0x1000;
+const int32_t MAX_UNCHANGED = MAX_UNCHANGED_LENGTH - 1;
+
+// 0mmmnnnccccccccc with m=1..6 records ccc+1 replacements of m:n text units.
+const int32_t MAX_SHORT_CHANGE_OLD_LENGTH = 6;
+const int32_t MAX_SHORT_CHANGE_NEW_LENGTH = 7;
+const int32_t SHORT_CHANGE_NUM_MASK = 0x1ff;
+const int32_t MAX_SHORT_CHANGE = 0x6fff;
+
+// 0111mmmmmmnnnnnn records a replacement of m text units with n.
+// m or n = 61: actual length follows in the next edits array unit.
+// m or n = 62..63: actual length follows in the next two edits array units.
+// Bit 30 of the actual length is in the head unit.
+// Trailing units have bit 15 set.
+const int32_t LENGTH_IN_1TRAIL = 61;
+const int32_t LENGTH_IN_2TRAIL = 62;
+
+}  // namespace
+
+void Edits::releaseArray() U_NOEXCEPT {
+    if (array != stackArray) {
+        uprv_free(array);
+    }
+}
+
+Edits &Edits::copyArray(const Edits &other) {
+    if (U_FAILURE(errorCode_)) {
+        length = delta = numChanges = 0;
+        return *this;
+    }
+    if (length > capacity) {
+        uint16_t *newArray = (uint16_t *)uprv_malloc((size_t)length * 2);
+        if (newArray == nullptr) {
+            length = delta = numChanges = 0;
+            errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+            return *this;
+        }
+        releaseArray();
+        array = newArray;
+        capacity = length;
+    }
+    if (length > 0) {
+        uprv_memcpy(array, other.array, (size_t)length * 2);
+    }
+    return *this;
+}
+
+Edits &Edits::moveArray(Edits &src) U_NOEXCEPT {
+    if (U_FAILURE(errorCode_)) {
+        length = delta = numChanges = 0;
+        return *this;
+    }
+    releaseArray();
+    if (length > STACK_CAPACITY) {
+        array = src.array;
+        capacity = src.capacity;
+        src.array = src.stackArray;
+        src.capacity = STACK_CAPACITY;
+        src.reset();
+        return *this;
+    }
+    array = stackArray;
+    capacity = STACK_CAPACITY;
+    if (length > 0) {
+        uprv_memcpy(array, src.array, (size_t)length * 2);
+    }
+    return *this;
+}
+
+Edits &Edits::operator=(const Edits &other) {
+    length = other.length;
+    delta = other.delta;
+    numChanges = other.numChanges;
+    errorCode_ = other.errorCode_;
+    return copyArray(other);
+}
+
+Edits &Edits::operator=(Edits &&src) U_NOEXCEPT {
+    length = src.length;
+    delta = src.delta;
+    numChanges = src.numChanges;
+    errorCode_ = src.errorCode_;
+    return moveArray(src);
+}
+
+Edits::~Edits() {
+    releaseArray();
+}
+
+void Edits::reset() U_NOEXCEPT {
+    length = delta = numChanges = 0;
+    errorCode_ = U_ZERO_ERROR;
+}
+
+void Edits::addUnchanged(int32_t unchangedLength) {
+    if(U_FAILURE(errorCode_) || unchangedLength == 0) { return; }
+    if(unchangedLength < 0) {
+        errorCode_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    // Merge into previous unchanged-text record, if any.
+    int32_t last = lastUnit();
+    if(last < MAX_UNCHANGED) {
+        int32_t remaining = MAX_UNCHANGED - last;
+        if (remaining >= unchangedLength) {
+            setLastUnit(last + unchangedLength);
+            return;
+        }
+        setLastUnit(MAX_UNCHANGED);
+        unchangedLength -= remaining;
+    }
+    // Split large lengths into multiple units.
+    while(unchangedLength >= MAX_UNCHANGED_LENGTH) {
+        append(MAX_UNCHANGED);
+        unchangedLength -= MAX_UNCHANGED_LENGTH;
+    }
+    // Write a small (remaining) length.
+    if(unchangedLength > 0) {
+        append(unchangedLength - 1);
+    }
+}
+
+void Edits::addReplace(int32_t oldLength, int32_t newLength) {
+    if(U_FAILURE(errorCode_)) { return; }
+    if(oldLength < 0 || newLength < 0) {
+        errorCode_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if (oldLength == 0 && newLength == 0) {
+        return;
+    }
+    ++numChanges;
+    int32_t newDelta = newLength - oldLength;
+    if (newDelta != 0) {
+        if ((newDelta > 0 && delta >= 0 && newDelta > (INT32_MAX - delta)) ||
+                (newDelta < 0 && delta < 0 && newDelta < (INT32_MIN - delta))) {
+            // Integer overflow or underflow.
+            errorCode_ = U_INDEX_OUTOFBOUNDS_ERROR;
+            return;
+        }
+        delta += newDelta;
+    }
+
+    if(0 < oldLength && oldLength <= MAX_SHORT_CHANGE_OLD_LENGTH &&
+            newLength <= MAX_SHORT_CHANGE_NEW_LENGTH) {
+        // Merge into previous same-lengths short-replacement record, if any.
+        int32_t u = (oldLength << 12) | (newLength << 9);
+        int32_t last = lastUnit();
+        if(MAX_UNCHANGED < last && last < MAX_SHORT_CHANGE &&
+                (last & ~SHORT_CHANGE_NUM_MASK) == u &&
+                (last & SHORT_CHANGE_NUM_MASK) < SHORT_CHANGE_NUM_MASK) {
+            setLastUnit(last + 1);
+            return;
+        }
+        append(u);
+        return;
+    }
+
+    int32_t head = 0x7000;
+    if (oldLength < LENGTH_IN_1TRAIL && newLength < LENGTH_IN_1TRAIL) {
+        head |= oldLength << 6;
+        head |= newLength;
+        append(head);
+    } else if ((capacity - length) >= 5 || growArray()) {
+        int32_t limit = length + 1;
+        if(oldLength < LENGTH_IN_1TRAIL) {
+            head |= oldLength << 6;
+        } else if(oldLength <= 0x7fff) {
+            head |= LENGTH_IN_1TRAIL << 6;
+            array[limit++] = (uint16_t)(0x8000 | oldLength);
+        } else {
+            head |= (LENGTH_IN_2TRAIL + (oldLength >> 30)) << 6;
+            array[limit++] = (uint16_t)(0x8000 | (oldLength >> 15));
+            array[limit++] = (uint16_t)(0x8000 | oldLength);
+        }
+        if(newLength < LENGTH_IN_1TRAIL) {
+            head |= newLength;
+        } else if(newLength <= 0x7fff) {
+            head |= LENGTH_IN_1TRAIL;
+            array[limit++] = (uint16_t)(0x8000 | newLength);
+        } else {
+            head |= LENGTH_IN_2TRAIL + (newLength >> 30);
+            array[limit++] = (uint16_t)(0x8000 | (newLength >> 15));
+            array[limit++] = (uint16_t)(0x8000 | newLength);
+        }
+        array[length] = (uint16_t)head;
+        length = limit;
+    }
+}
+
+void Edits::append(int32_t r) {
+    if(length < capacity || growArray()) {
+        array[length++] = (uint16_t)r;
+    }
+}
+
+UBool Edits::growArray() {
+    int32_t newCapacity;
+    if (array == stackArray) {
+        newCapacity = 2000;
+    } else if (capacity == INT32_MAX) {
+        // Not U_BUFFER_OVERFLOW_ERROR because that could be confused on a string transform API
+        // with a result-string-buffer overflow.
+        errorCode_ = U_INDEX_OUTOFBOUNDS_ERROR;
+        return FALSE;
+    } else if (capacity >= (INT32_MAX / 2)) {
+        newCapacity = INT32_MAX;
+    } else {
+        newCapacity = 2 * capacity;
+    }
+    // Grow by at least 5 units so that a maximal change record will fit.
+    if ((newCapacity - capacity) < 5) {
+        errorCode_ = U_INDEX_OUTOFBOUNDS_ERROR;
+        return FALSE;
+    }
+    uint16_t *newArray = (uint16_t *)uprv_malloc((size_t)newCapacity * 2);
+    if (newArray == NULL) {
+        errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    uprv_memcpy(newArray, array, (size_t)length * 2);
+    releaseArray();
+    array = newArray;
+    capacity = newCapacity;
+    return TRUE;
+}
+
+UBool Edits::copyErrorTo(UErrorCode &outErrorCode) const {
+    if (U_FAILURE(outErrorCode)) { return TRUE; }
+    if (U_SUCCESS(errorCode_)) { return FALSE; }
+    outErrorCode = errorCode_;
+    return TRUE;
+}
+
+Edits &Edits::mergeAndAppend(const Edits &ab, const Edits &bc, UErrorCode &errorCode) {
+    if (copyErrorTo(errorCode)) { return *this; }
+    // Picture string a --(Edits ab)--> string b --(Edits bc)--> string c.
+    // Parallel iteration over both Edits.
+    Iterator abIter = ab.getFineIterator();
+    Iterator bcIter = bc.getFineIterator();
+    UBool abHasNext = TRUE, bcHasNext = TRUE;
+    // Copy iterator state into local variables, so that we can modify and subdivide spans.
+    // ab old & new length, bc old & new length
+    int32_t aLength = 0, ab_bLength = 0, bc_bLength = 0, cLength = 0;
+    // When we have different-intermediate-length changes, we accumulate a larger change.
+    int32_t pending_aLength = 0, pending_cLength = 0;
+    for (;;) {
+        // At this point, for each of the two iterators:
+        // Either we are done with the locally cached current edit,
+        // and its intermediate-string length has been reset,
+        // or we will continue to work with a truncated remainder of this edit.
+        //
+        // If the current edit is done, and the iterator has not yet reached the end,
+        // then we fetch the next edit. This is true for at least one of the iterators.
+        //
+        // Normally it does not matter whether we fetch from ab and then bc or vice versa.
+        // However, the result is observably different when
+        // ab deletions meet bc insertions at the same intermediate-string index.
+        // Some users expect the bc insertions to come first, so we fetch from bc first.
+        if (bc_bLength == 0) {
+            if (bcHasNext && (bcHasNext = bcIter.next(errorCode)) != 0) {
+                bc_bLength = bcIter.oldLength();
+                cLength = bcIter.newLength();
+                if (bc_bLength == 0) {
+                    // insertion
+                    if (ab_bLength == 0 || !abIter.hasChange()) {
+                        addReplace(pending_aLength, pending_cLength + cLength);
+                        pending_aLength = pending_cLength = 0;
+                    } else {
+                        pending_cLength += cLength;
+                    }
+                    continue;
+                }
+            }
+            // else see if the other iterator is done, too.
+        }
+        if (ab_bLength == 0) {
+            if (abHasNext && (abHasNext = abIter.next(errorCode)) != 0) {
+                aLength = abIter.oldLength();
+                ab_bLength = abIter.newLength();
+                if (ab_bLength == 0) {
+                    // deletion
+                    if (bc_bLength == bcIter.oldLength() || !bcIter.hasChange()) {
+                        addReplace(pending_aLength + aLength, pending_cLength);
+                        pending_aLength = pending_cLength = 0;
+                    } else {
+                        pending_aLength += aLength;
+                    }
+                    continue;
+                }
+            } else if (bc_bLength == 0) {
+                // Both iterators are done at the same time:
+                // The intermediate-string lengths match.
+                break;
+            } else {
+                // The ab output string is shorter than the bc input string.
+                if (!copyErrorTo(errorCode)) {
+                    errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+                }
+                return *this;
+            }
+        }
+        if (bc_bLength == 0) {
+            // The bc input string is shorter than the ab output string.
+            if (!copyErrorTo(errorCode)) {
+                errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            }
+            return *this;
+        }
+        //  Done fetching: ab_bLength > 0 && bc_bLength > 0
+
+        // The current state has two parts:
+        // - Past: We accumulate a longer ac edit in the "pending" variables.
+        // - Current: We have copies of the current ab/bc edits in local variables.
+        //   At least one side is newly fetched.
+        //   One side might be a truncated remainder of an edit we fetched earlier.
+
+        if (!abIter.hasChange() && !bcIter.hasChange()) {
+            // An unchanged span all the way from string a to string c.
+            if (pending_aLength != 0 || pending_cLength != 0) {
+                addReplace(pending_aLength, pending_cLength);
+                pending_aLength = pending_cLength = 0;
+            }
+            int32_t unchangedLength = aLength <= cLength ? aLength : cLength;
+            addUnchanged(unchangedLength);
+            ab_bLength = aLength -= unchangedLength;
+            bc_bLength = cLength -= unchangedLength;
+            // At least one of the unchanged spans is now empty.
+            continue;
+        }
+        if (!abIter.hasChange() && bcIter.hasChange()) {
+            // Unchanged a->b but changed b->c.
+            if (ab_bLength >= bc_bLength) {
+                // Split the longer unchanged span into change + remainder.
+                addReplace(pending_aLength + bc_bLength, pending_cLength + cLength);
+                pending_aLength = pending_cLength = 0;
+                aLength = ab_bLength -= bc_bLength;
+                bc_bLength = 0;
+                continue;
+            }
+            // Handle the shorter unchanged span below like a change.
+        } else if (abIter.hasChange() && !bcIter.hasChange()) {
+            // Changed a->b and then unchanged b->c.
+            if (ab_bLength <= bc_bLength) {
+                // Split the longer unchanged span into change + remainder.
+                addReplace(pending_aLength + aLength, pending_cLength + ab_bLength);
+                pending_aLength = pending_cLength = 0;
+                cLength = bc_bLength -= ab_bLength;
+                ab_bLength = 0;
+                continue;
+            }
+            // Handle the shorter unchanged span below like a change.
+        } else {  // both abIter.hasChange() && bcIter.hasChange()
+            if (ab_bLength == bc_bLength) {
+                // Changes on both sides up to the same position. Emit & reset.
+                addReplace(pending_aLength + aLength, pending_cLength + cLength);
+                pending_aLength = pending_cLength = 0;
+                ab_bLength = bc_bLength = 0;
+                continue;
+            }
+        }
+        // Accumulate the a->c change, reset the shorter side,
+        // keep a remainder of the longer one.
+        pending_aLength += aLength;
+        pending_cLength += cLength;
+        if (ab_bLength < bc_bLength) {
+            bc_bLength -= ab_bLength;
+            cLength = ab_bLength = 0;
+        } else {  // ab_bLength > bc_bLength
+            ab_bLength -= bc_bLength;
+            aLength = bc_bLength = 0;
+        }
+    }
+    if (pending_aLength != 0 || pending_cLength != 0) {
+        addReplace(pending_aLength, pending_cLength);
+    }
+    copyErrorTo(errorCode);
+    return *this;
+}
+
+Edits::Iterator::Iterator(const uint16_t *a, int32_t len, UBool oc, UBool crs) :
+        array(a), index(0), length(len), remaining(0),
+        onlyChanges_(oc), coarse(crs),
+        dir(0), changed(FALSE), oldLength_(0), newLength_(0),
+        srcIndex(0), replIndex(0), destIndex(0) {}
+
+int32_t Edits::Iterator::readLength(int32_t head) {
+    if (head < LENGTH_IN_1TRAIL) {
+        return head;
+    } else if (head < LENGTH_IN_2TRAIL) {
+        U_ASSERT(index < length);
+        U_ASSERT(array[index] >= 0x8000);
+        return array[index++] & 0x7fff;
+    } else {
+        U_ASSERT((index + 2) <= length);
+        U_ASSERT(array[index] >= 0x8000);
+        U_ASSERT(array[index + 1] >= 0x8000);
+        int32_t len = ((head & 1) << 30) |
+                ((int32_t)(array[index] & 0x7fff) << 15) |
+                (array[index + 1] & 0x7fff);
+        index += 2;
+        return len;
+    }
+}
+
+void Edits::Iterator::updateNextIndexes() {
+    srcIndex += oldLength_;
+    if (changed) {
+        replIndex += newLength_;
+    }
+    destIndex += newLength_;
+}
+
+void Edits::Iterator::updatePreviousIndexes() {
+    srcIndex -= oldLength_;
+    if (changed) {
+        replIndex -= newLength_;
+    }
+    destIndex -= newLength_;
+}
+
+UBool Edits::Iterator::noNext() {
+    // No change before or beyond the string.
+    dir = 0;
+    changed = FALSE;
+    oldLength_ = newLength_ = 0;
+    return FALSE;
+}
+
+UBool Edits::Iterator::next(UBool onlyChanges, UErrorCode &errorCode) {
+    // Forward iteration: Update the string indexes to the limit of the current span,
+    // and post-increment-read array units to assemble a new span.
+    // Leaves the array index one after the last unit of that span.
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    // We have an errorCode in case we need to start guarding against integer overflows.
+    // It is also convenient for caller loops if we bail out when an error was set elsewhere.
+    if (dir > 0) {
+        updateNextIndexes();
+    } else {
+        if (dir < 0) {
+            // Turn around from previous() to next().
+            // Post-increment-read the same span again.
+            if (remaining > 0) {
+                // Fine-grained iterator:
+                // Stay on the current one of a sequence of compressed changes.
+                ++index;  // next() rests on the index after the sequence unit.
+                dir = 1;
+                return TRUE;
+            }
+        }
+        dir = 1;
+    }
+    if (remaining >= 1) {
+        // Fine-grained iterator: Continue a sequence of compressed changes.
+        if (remaining > 1) {
+            --remaining;
+            return TRUE;
+        }
+        remaining = 0;
+    }
+    if (index >= length) {
+        return noNext();
+    }
+    int32_t u = array[index++];
+    if (u <= MAX_UNCHANGED) {
+        // Combine adjacent unchanged ranges.
+        changed = FALSE;
+        oldLength_ = u + 1;
+        while (index < length && (u = array[index]) <= MAX_UNCHANGED) {
+            ++index;
+            oldLength_ += u + 1;
+        }
+        newLength_ = oldLength_;
+        if (onlyChanges) {
+            updateNextIndexes();
+            if (index >= length) {
+                return noNext();
+            }
+            // already fetched u > MAX_UNCHANGED at index
+            ++index;
+        } else {
+            return TRUE;
+        }
+    }
+    changed = TRUE;
+    if (u <= MAX_SHORT_CHANGE) {
+        int32_t oldLen = u >> 12;
+        int32_t newLen = (u >> 9) & MAX_SHORT_CHANGE_NEW_LENGTH;
+        int32_t num = (u & SHORT_CHANGE_NUM_MASK) + 1;
+        if (coarse) {
+            oldLength_ = num * oldLen;
+            newLength_ = num * newLen;
+        } else {
+            // Split a sequence of changes that was compressed into one unit.
+            oldLength_ = oldLen;
+            newLength_ = newLen;
+            if (num > 1) {
+                remaining = num;  // This is the first of two or more changes.
+            }
+            return TRUE;
+        }
+    } else {
+        U_ASSERT(u <= 0x7fff);
+        oldLength_ = readLength((u >> 6) & 0x3f);
+        newLength_ = readLength(u & 0x3f);
+        if (!coarse) {
+            return TRUE;
+        }
+    }
+    // Combine adjacent changes.
+    while (index < length && (u = array[index]) > MAX_UNCHANGED) {
+        ++index;
+        if (u <= MAX_SHORT_CHANGE) {
+            int32_t num = (u & SHORT_CHANGE_NUM_MASK) + 1;
+            oldLength_ += (u >> 12) * num;
+            newLength_ += ((u >> 9) & MAX_SHORT_CHANGE_NEW_LENGTH) * num;
+        } else {
+            U_ASSERT(u <= 0x7fff);
+            oldLength_ += readLength((u >> 6) & 0x3f);
+            newLength_ += readLength(u & 0x3f);
+        }
+    }
+    return TRUE;
+}
+
+UBool Edits::Iterator::previous(UErrorCode &errorCode) {
+    // Backward iteration: Pre-decrement-read array units to assemble a new span,
+    // then update the string indexes to the start of that span.
+    // Leaves the array index on the head unit of that span.
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    // We have an errorCode in case we need to start guarding against integer overflows.
+    // It is also convenient for caller loops if we bail out when an error was set elsewhere.
+    if (dir >= 0) {
+        if (dir > 0) {
+            // Turn around from next() to previous().
+            // Set the string indexes to the span limit and
+            // pre-decrement-read the same span again.
+            if (remaining > 0) {
+                // Fine-grained iterator:
+                // Stay on the current one of a sequence of compressed changes.
+                --index;  // previous() rests on the sequence unit.
+                dir = -1;
+                return TRUE;
+            }
+            updateNextIndexes();
+        }
+        dir = -1;
+    }
+    if (remaining > 0) {
+        // Fine-grained iterator: Continue a sequence of compressed changes.
+        int32_t u = array[index];
+        U_ASSERT(MAX_UNCHANGED < u && u <= MAX_SHORT_CHANGE);
+        if (remaining <= (u & SHORT_CHANGE_NUM_MASK)) {
+            ++remaining;
+            updatePreviousIndexes();
+            return TRUE;
+        }
+        remaining = 0;
+    }
+    if (index <= 0) {
+        return noNext();
+    }
+    int32_t u = array[--index];
+    if (u <= MAX_UNCHANGED) {
+        // Combine adjacent unchanged ranges.
+        changed = FALSE;
+        oldLength_ = u + 1;
+        while (index > 0 && (u = array[index - 1]) <= MAX_UNCHANGED) {
+            --index;
+            oldLength_ += u + 1;
+        }
+        newLength_ = oldLength_;
+        // No need to handle onlyChanges as long as previous() is called only from findIndex().
+        updatePreviousIndexes();
+        return TRUE;
+    }
+    changed = TRUE;
+    if (u <= MAX_SHORT_CHANGE) {
+        int32_t oldLen = u >> 12;
+        int32_t newLen = (u >> 9) & MAX_SHORT_CHANGE_NEW_LENGTH;
+        int32_t num = (u & SHORT_CHANGE_NUM_MASK) + 1;
+        if (coarse) {
+            oldLength_ = num * oldLen;
+            newLength_ = num * newLen;
+        } else {
+            // Split a sequence of changes that was compressed into one unit.
+            oldLength_ = oldLen;
+            newLength_ = newLen;
+            if (num > 1) {
+                remaining = 1;  // This is the last of two or more changes.
+            }
+            updatePreviousIndexes();
+            return TRUE;
+        }
+    } else {
+        if (u <= 0x7fff) {
+            // The change is encoded in u alone.
+            oldLength_ = readLength((u >> 6) & 0x3f);
+            newLength_ = readLength(u & 0x3f);
+        } else {
+            // Back up to the head of the change, read the lengths,
+            // and reset the index to the head again.
+            U_ASSERT(index > 0);
+            while ((u = array[--index]) > 0x7fff) {}
+            U_ASSERT(u > MAX_SHORT_CHANGE);
+            int32_t headIndex = index++;
+            oldLength_ = readLength((u >> 6) & 0x3f);
+            newLength_ = readLength(u & 0x3f);
+            index = headIndex;
+        }
+        if (!coarse) {
+            updatePreviousIndexes();
+            return TRUE;
+        }
+    }
+    // Combine adjacent changes.
+    while (index > 0 && (u = array[index - 1]) > MAX_UNCHANGED) {
+        --index;
+        if (u <= MAX_SHORT_CHANGE) {
+            int32_t num = (u & SHORT_CHANGE_NUM_MASK) + 1;
+            oldLength_ += (u >> 12) * num;
+            newLength_ += ((u >> 9) & MAX_SHORT_CHANGE_NEW_LENGTH) * num;
+        } else if (u <= 0x7fff) {
+            // Read the lengths, and reset the index to the head again.
+            int32_t headIndex = index++;
+            oldLength_ += readLength((u >> 6) & 0x3f);
+            newLength_ += readLength(u & 0x3f);
+            index = headIndex;
+        }
+    }
+    updatePreviousIndexes();
+    return TRUE;
+}
+
+int32_t Edits::Iterator::findIndex(int32_t i, UBool findSource, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode) || i < 0) { return -1; }
+    int32_t spanStart, spanLength;
+    if (findSource) {  // find source index
+        spanStart = srcIndex;
+        spanLength = oldLength_;
+    } else {  // find destination index
+        spanStart = destIndex;
+        spanLength = newLength_;
+    }
+    if (i < spanStart) {
+        if (i >= (spanStart / 2)) {
+            // Search backwards.
+            for (;;) {
+                UBool hasPrevious = previous(errorCode);
+                U_ASSERT(hasPrevious);  // because i>=0 and the first span starts at 0
+                (void)hasPrevious;  // avoid unused-variable warning
+                spanStart = findSource ? srcIndex : destIndex;
+                if (i >= spanStart) {
+                    // The index is in the current span.
+                    return 0;
+                }
+                if (remaining > 0) {
+                    // Is the index in one of the remaining compressed edits?
+                    // spanStart is the start of the current span, first of the remaining ones.
+                    spanLength = findSource ? oldLength_ : newLength_;
+                    int32_t u = array[index];
+                    U_ASSERT(MAX_UNCHANGED < u && u <= MAX_SHORT_CHANGE);
+                    int32_t num = (u & SHORT_CHANGE_NUM_MASK) + 1 - remaining;
+                    int32_t len = num * spanLength;
+                    if (i >= (spanStart - len)) {
+                        int32_t n = ((spanStart - i - 1) / spanLength) + 1;
+                        // 1 <= n <= num
+                        srcIndex -= n * oldLength_;
+                        replIndex -= n * newLength_;
+                        destIndex -= n * newLength_;
+                        remaining += n;
+                        return 0;
+                    }
+                    // Skip all of these edits at once.
+                    srcIndex -= num * oldLength_;
+                    replIndex -= num * newLength_;
+                    destIndex -= num * newLength_;
+                    remaining = 0;
+                }
+            }
+        }
+        // Reset the iterator to the start.
+        dir = 0;
+        index = remaining = oldLength_ = newLength_ = srcIndex = replIndex = destIndex = 0;
+    } else if (i < (spanStart + spanLength)) {
+        // The index is in the current span.
+        return 0;
+    }
+    while (next(FALSE, errorCode)) {
+        if (findSource) {
+            spanStart = srcIndex;
+            spanLength = oldLength_;
+        } else {
+            spanStart = destIndex;
+            spanLength = newLength_;
+        }
+        if (i < (spanStart + spanLength)) {
+            // The index is in the current span.
+            return 0;
+        }
+        if (remaining > 1) {
+            // Is the index in one of the remaining compressed edits?
+            // spanStart is the start of the current span, first of the remaining ones.
+            int32_t len = remaining * spanLength;
+            if (i < (spanStart + len)) {
+                int32_t n = (i - spanStart) / spanLength;  // 1 <= n <= remaining - 1
+                srcIndex += n * oldLength_;
+                replIndex += n * newLength_;
+                destIndex += n * newLength_;
+                remaining -= n;
+                return 0;
+            }
+            // Make next() skip all of these edits at once.
+            oldLength_ *= remaining;
+            newLength_ *= remaining;
+            remaining = 0;
+        }
+    }
+    return 1;
+}
+
+int32_t Edits::Iterator::destinationIndexFromSourceIndex(int32_t i, UErrorCode &errorCode) {
+    int32_t where = findIndex(i, TRUE, errorCode);
+    if (where < 0) {
+        // Error or before the string.
+        return 0;
+    }
+    if (where > 0 || i == srcIndex) {
+        // At or after string length, or at start of the found span.
+        return destIndex;
+    }
+    if (changed) {
+        // In a change span, map to its end.
+        return destIndex + newLength_;
+    } else {
+        // In an unchanged span, offset 1:1 within it.
+        return destIndex + (i - srcIndex);
+    }
+}
+
+int32_t Edits::Iterator::sourceIndexFromDestinationIndex(int32_t i, UErrorCode &errorCode) {
+    int32_t where = findIndex(i, FALSE, errorCode);
+    if (where < 0) {
+        // Error or before the string.
+        return 0;
+    }
+    if (where > 0 || i == destIndex) {
+        // At or after string length, or at start of the found span.
+        return srcIndex;
+    }
+    if (changed) {
+        // In a change span, map to its end.
+        return srcIndex + oldLength_;
+    } else {
+        // In an unchanged span, offset within it.
+        return srcIndex + (i - destIndex);
+    }
+}
+
+UnicodeString& Edits::Iterator::toString(UnicodeString& sb) const {
+    sb.append(u"{ src[", -1);
+    ICU_Utility::appendNumber(sb, srcIndex);
+    sb.append(u"..", -1);
+    ICU_Utility::appendNumber(sb, srcIndex + oldLength_);
+    if (changed) {
+        sb.append(u"] ⇝ dest[", -1);
+    } else {
+        sb.append(u"] ≡ dest[", -1);
+    }
+    ICU_Utility::appendNumber(sb, destIndex);
+    sb.append(u"..", -1);
+    ICU_Utility::appendNumber(sb, destIndex + newLength_);
+    if (changed) {
+        sb.append(u"], repl[", -1);
+        ICU_Utility::appendNumber(sb, replIndex);
+        sb.append(u"..", -1);
+        ICU_Utility::appendNumber(sb, replIndex + newLength_);
+        sb.append(u"] }", -1);
+    } else {
+        sb.append(u"] (no-change) }", -1);
+    }
+    return sb;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/errorcode.cpp b/thirdparty/icu4c/common/errorcode.cpp
new file mode 100644
index 0000000000..e7ac43b527
--- /dev/null
+++ b/thirdparty/icu4c/common/errorcode.cpp
@@ -0,0 +1,42 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  errorcode.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009mar10
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/errorcode.h"
+
+U_NAMESPACE_BEGIN
+
+ErrorCode::~ErrorCode() {}
+
+UErrorCode ErrorCode::reset() {
+    UErrorCode code = errorCode;
+    errorCode = U_ZERO_ERROR;
+    return code;
+}
+
+void ErrorCode::assertSuccess() const {
+    if(isFailure()) {
+        handleFailure();
+    }
+}
+
+const char* ErrorCode::errorName() const {
+  return u_errorName(errorCode);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/filteredbrk.cpp b/thirdparty/icu4c/common/filteredbrk.cpp
new file mode 100644
index 0000000000..c07128cbce
--- /dev/null
+++ b/thirdparty/icu4c/common/filteredbrk.cpp
@@ -0,0 +1,710 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2014-2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+*******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#if !UCONFIG_NO_BREAK_ITERATION && !UCONFIG_NO_FILTERED_BREAK_ITERATION
+
+#include "cmemory.h"
+
+#include "unicode/filteredbrk.h"
+#include "unicode/ucharstriebuilder.h"
+#include "unicode/ures.h"
+
+#include "uresimp.h" // ures_getByKeyWithFallback
+#include "ubrkimpl.h" // U_ICUDATA_BRKITR
+#include "uvector.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+#ifndef FB_DEBUG
+#define FB_DEBUG 0
+#endif
+
+#if FB_DEBUG
+#include <stdio.h>
+static void _fb_trace(const char *m, const UnicodeString *s, UBool b, int32_t d, const char *f, int l) {
+  char buf[2048];
+  if(s) {
+    s->extract(0,s->length(),buf,2048);
+  } else {
+    strcpy(buf,"NULL");
+  }
+  fprintf(stderr,"%s:%d: %s. s='%s'(%p), b=%c, d=%d\n",
+          f, l, m, buf, (const void*)s, b?'T':'F',(int)d);
+}
+
+#define FB_TRACE(m,s,b,d) _fb_trace(m,s,b,d,__FILE__,__LINE__)
+#else
+#define FB_TRACE(m,s,b,d)
+#endif
+
+/**
+ * Used with sortedInsert()
+ */
+static int8_t U_CALLCONV compareUnicodeString(UElement t1, UElement t2) {
+    const UnicodeString &a = *(const UnicodeString*)t1.pointer;
+    const UnicodeString &b = *(const UnicodeString*)t2.pointer;
+    return a.compare(b);
+}
+
+/**
+ * A UVector which implements a set of strings.
+ */
+class U_COMMON_API UStringSet : public UVector {
+ public:
+  UStringSet(UErrorCode &status) : UVector(uprv_deleteUObject,
+                                           uhash_compareUnicodeString,
+                                           1,
+                                           status) {}
+  virtual ~UStringSet();
+  /**
+   * Is this UnicodeSet contained?
+   */
+  inline UBool contains(const UnicodeString& s) {
+    return contains((void*) &s);
+  }
+  using UVector::contains;
+  /**
+   * Return the ith UnicodeString alias
+   */
+  inline const UnicodeString* getStringAt(int32_t i) const {
+    return (const UnicodeString*)elementAt(i);
+  }
+  /**
+   * Adopt the UnicodeString if not already contained.
+   * Caller no longer owns the pointer in any case.
+   * @return true if adopted successfully, false otherwise (error, or else duplicate)
+   */
+  inline UBool adopt(UnicodeString *str, UErrorCode &status) {
+    if(U_FAILURE(status) || contains(*str)) {
+      delete str;
+      return false;
+    } else {
+      sortedInsert(str, compareUnicodeString, status);
+      if(U_FAILURE(status)) {
+        delete str;
+        return false;
+      }
+      return true;
+    }
+  }
+  /**
+   * Add by value.
+   * @return true if successfully adopted.
+   */
+  inline UBool add(const UnicodeString& str, UErrorCode &status) {
+    if(U_FAILURE(status)) return false;
+    UnicodeString *t = new UnicodeString(str);
+    if(t==NULL) {
+      status = U_MEMORY_ALLOCATION_ERROR; return false;
+    }
+    return adopt(t, status);
+  }
+  /**
+   * Remove this string.
+   * @return true if successfully removed, false otherwise (error, or else it wasn't there)
+   */
+  inline UBool remove(const UnicodeString &s, UErrorCode &status) {
+    if(U_FAILURE(status)) return false;
+    return removeElement((void*) &s);
+  }
+};
+
+/**
+ * Virtual, won't be inlined
+ */
+UStringSet::~UStringSet() {}
+
+/* ----------------------------------------------------------- */
+
+
+/* Filtered Break constants */
+static const int32_t kPARTIAL = (1<<0); //< partial - need to run through forward trie
+static const int32_t kMATCH   = (1<<1); //< exact match - skip this one.
+static const int32_t kSuppressInReverse = (1<<0);
+static const int32_t kAddToForward = (1<<1);
+static const UChar   kFULLSTOP = 0x002E; // '.'
+
+/**
+ * Shared data for SimpleFilteredSentenceBreakIterator
+ */
+class SimpleFilteredSentenceBreakData : public UMemory {
+public:
+  SimpleFilteredSentenceBreakData(UCharsTrie *forwards, UCharsTrie *backwards ) 
+      : fForwardsPartialTrie(forwards), fBackwardsTrie(backwards), refcount(1) { }
+  SimpleFilteredSentenceBreakData *incr() { refcount++;  return this; }
+  SimpleFilteredSentenceBreakData *decr() { if((--refcount) <= 0) delete this; return 0; }
+  virtual ~SimpleFilteredSentenceBreakData();
+
+  LocalPointer<UCharsTrie>    fForwardsPartialTrie; //  Has ".a" for "a.M."
+  LocalPointer<UCharsTrie>    fBackwardsTrie; //  i.e. ".srM" for Mrs.
+  int32_t                     refcount;
+};
+
+SimpleFilteredSentenceBreakData::~SimpleFilteredSentenceBreakData() {}
+
+/**
+ * Concrete implementation
+ */
+class SimpleFilteredSentenceBreakIterator : public BreakIterator {
+public:
+  SimpleFilteredSentenceBreakIterator(BreakIterator *adopt, UCharsTrie *forwards, UCharsTrie *backwards, UErrorCode &status);
+  SimpleFilteredSentenceBreakIterator(const SimpleFilteredSentenceBreakIterator& other);
+  virtual ~SimpleFilteredSentenceBreakIterator();
+private:
+  SimpleFilteredSentenceBreakData *fData;
+  LocalPointer<BreakIterator> fDelegate;
+  LocalUTextPointer           fText;
+
+  /* -- subclass interface -- */
+public:
+  /* -- cloning and other subclass stuff -- */
+  virtual BreakIterator *  createBufferClone(void * /*stackBuffer*/,
+                                             int32_t &/*BufferSize*/,
+                                             UErrorCode &status) {
+    // for now - always deep clone
+    status = U_SAFECLONE_ALLOCATED_WARNING;
+    return clone();
+  }
+  virtual SimpleFilteredSentenceBreakIterator* clone() const { return new SimpleFilteredSentenceBreakIterator(*this); }
+  virtual UClassID getDynamicClassID(void) const { return NULL; }
+  virtual UBool operator==(const BreakIterator& o) const { if(this==&o) return true; return false; }
+
+  /* -- text modifying -- */
+  virtual void setText(UText *text, UErrorCode &status) { fDelegate->setText(text,status); }
+  virtual BreakIterator &refreshInputText(UText *input, UErrorCode &status) { fDelegate->refreshInputText(input,status); return *this; }
+  virtual void adoptText(CharacterIterator* it) { fDelegate->adoptText(it); }
+  virtual void setText(const UnicodeString &text) { fDelegate->setText(text); }
+
+  /* -- other functions that are just delegated -- */
+  virtual UText *getUText(UText *fillIn, UErrorCode &status) const { return fDelegate->getUText(fillIn,status); }
+  virtual CharacterIterator& getText(void) const { return fDelegate->getText(); }
+
+  /* -- ITERATION -- */
+  virtual int32_t first(void);
+  virtual int32_t preceding(int32_t offset);
+  virtual int32_t previous(void);
+  virtual UBool isBoundary(int32_t offset);
+  virtual int32_t current(void) const { return fDelegate->current(); } // we keep the delegate current, so this should be correct.
+
+  virtual int32_t next(void);
+
+  virtual int32_t next(int32_t n);
+  virtual int32_t following(int32_t offset);
+  virtual int32_t last(void);
+
+private:
+    /**
+     * Given that the fDelegate has already given its "initial" answer,
+     * find the NEXT actual (non-excepted) break.
+     * @param n initial position from delegate
+     * @return new break position or UBRK_DONE
+     */
+    int32_t internalNext(int32_t n);
+    /**
+     * Given that the fDelegate has already given its "initial" answer,
+     * find the PREV actual (non-excepted) break.
+     * @param n initial position from delegate
+     * @return new break position or UBRK_DONE
+     */
+    int32_t internalPrev(int32_t n);
+    /**
+     * set up the UText with the value of the fDelegate.
+     * Call this before calling breakExceptionAt. 
+     * May be able to avoid excess calls
+     */
+    void resetState(UErrorCode &status);
+    /**
+     * Is there a match  (exception) at this spot?
+     */
+    enum EFBMatchResult { kNoExceptionHere, kExceptionHere };
+    /**
+     * Determine if there is an exception at this spot
+     * @param n spot to check
+     * @return kNoExceptionHere or kExceptionHere
+     **/
+    enum EFBMatchResult breakExceptionAt(int32_t n);
+};
+
+SimpleFilteredSentenceBreakIterator::SimpleFilteredSentenceBreakIterator(const SimpleFilteredSentenceBreakIterator& other)
+  : BreakIterator(other), fData(other.fData->incr()), fDelegate(other.fDelegate->clone())
+{
+}
+
+
+SimpleFilteredSentenceBreakIterator::SimpleFilteredSentenceBreakIterator(BreakIterator *adopt, UCharsTrie *forwards, UCharsTrie *backwards, UErrorCode &status) :
+  BreakIterator(adopt->getLocale(ULOC_VALID_LOCALE,status),adopt->getLocale(ULOC_ACTUAL_LOCALE,status)),
+  fData(new SimpleFilteredSentenceBreakData(forwards, backwards)),
+  fDelegate(adopt)
+{
+  // all set..
+}
+
+SimpleFilteredSentenceBreakIterator::~SimpleFilteredSentenceBreakIterator() {
+    fData = fData->decr();
+}
+
+void SimpleFilteredSentenceBreakIterator::resetState(UErrorCode &status) {
+  fText.adoptInstead(fDelegate->getUText(fText.orphan(), status));
+}
+
+SimpleFilteredSentenceBreakIterator::EFBMatchResult
+SimpleFilteredSentenceBreakIterator::breakExceptionAt(int32_t n) {
+    int64_t bestPosn = -1;
+    int32_t bestValue = -1;
+    // loops while 'n' points to an exception.
+    utext_setNativeIndex(fText.getAlias(), n); // from n..
+    fData->fBackwardsTrie->reset();
+    UChar32 uch;
+
+    //if(debug2) u_printf(" n@ %d\n", n);
+    // Assume a space is following the '.'  (so we handle the case:  "Mr. /Brown")
+    if((uch=utext_previous32(fText.getAlias()))==(UChar32)0x0020) {  // TODO: skip a class of chars here??
+      // TODO only do this the 1st time?
+      //if(debug2) u_printf("skipping prev: |%C| \n", (UChar)uch);
+    } else {
+      //if(debug2) u_printf("not skipping prev: |%C| \n", (UChar)uch);
+      uch = utext_next32(fText.getAlias());
+      //if(debug2) u_printf(" -> : |%C| \n", (UChar)uch);
+    }
+
+    UStringTrieResult r = USTRINGTRIE_INTERMEDIATE_VALUE;
+
+    while((uch=utext_previous32(fText.getAlias()))!=U_SENTINEL  &&   // more to consume backwards and..
+          USTRINGTRIE_HAS_NEXT(r=fData->fBackwardsTrie->nextForCodePoint(uch))) {// more in the trie
+      if(USTRINGTRIE_HAS_VALUE(r)) { // remember the best match so far
+        bestPosn = utext_getNativeIndex(fText.getAlias());
+        bestValue = fData->fBackwardsTrie->getValue();
+      }
+      //if(debug2) u_printf("rev< /%C/ cont?%d @%d\n", (UChar)uch, r, utext_getNativeIndex(fText.getAlias()));
+    }
+
+    if(USTRINGTRIE_MATCHES(r)) { // exact match?
+      //if(debug2) u_printf("rev<?/%C/?end of seq.. r=%d, bestPosn=%d, bestValue=%d\n", (UChar)uch, r, bestPosn, bestValue);
+      bestValue = fData->fBackwardsTrie->getValue();
+      bestPosn = utext_getNativeIndex(fText.getAlias());
+      //if(debug2) u_printf("rev<+/%C/+end of seq.. r=%d, bestPosn=%d, bestValue=%d\n", (UChar)uch, r, bestPosn, bestValue);
+    }
+
+    if(bestPosn>=0) {
+      //if(debug2) u_printf("rev< /%C/ end of seq.. r=%d, bestPosn=%d, bestValue=%d\n", (UChar)uch, r, bestPosn, bestValue);
+
+      //if(USTRINGTRIE_MATCHES(r)) {  // matched - so, now what?
+      //int32_t bestValue = fBackwardsTrie->getValue();
+      ////if(debug2) u_printf("rev< /%C/ matched, skip..%d  bestValue=%d\n", (UChar)uch, r, bestValue);
+
+      if(bestValue == kMATCH) { // exact match!
+        //if(debug2) u_printf(" exact backward match\n");
+        return kExceptionHere; // See if the next is another exception.
+      } else if(bestValue == kPARTIAL
+                && fData->fForwardsPartialTrie.isValid()) { // make sure there's a forward trie
+        //if(debug2) u_printf(" partial backward match\n");
+        // We matched the "Ph." in "Ph.D." - now we need to run everything through the forwards trie
+        // to see if it matches something going forward.
+        fData->fForwardsPartialTrie->reset();
+        UStringTrieResult rfwd = USTRINGTRIE_INTERMEDIATE_VALUE;
+        utext_setNativeIndex(fText.getAlias(), bestPosn); // hope that's close ..
+        //if(debug2) u_printf("Retrying at %d\n", bestPosn);
+        while((uch=utext_next32(fText.getAlias()))!=U_SENTINEL &&
+              USTRINGTRIE_HAS_NEXT(rfwd=fData->fForwardsPartialTrie->nextForCodePoint(uch))) {
+          //if(debug2) u_printf("fwd> /%C/ cont?%d @%d\n", (UChar)uch, rfwd, utext_getNativeIndex(fText.getAlias()));
+        }
+        if(USTRINGTRIE_MATCHES(rfwd)) {
+          //if(debug2) u_printf("fwd> /%C/ == forward match!\n", (UChar)uch);
+          // only full matches here, nothing to check
+          // skip the next:
+            return kExceptionHere;
+        } else {
+          //if(debug2) u_printf("fwd> /%C/ no match.\n", (UChar)uch);
+          // no match (no exception) -return the 'underlying' break
+          return kNoExceptionHere;
+        }
+      } else {
+        return kNoExceptionHere; // internal error and/or no forwards trie
+      }
+    } else {
+      //if(debug2) u_printf("rev< /%C/ .. no match..%d\n", (UChar)uch, r);  // no best match
+      return kNoExceptionHere; // No match - so exit. Not an exception.
+    }
+}
+
+// the workhorse single next.
+int32_t
+SimpleFilteredSentenceBreakIterator::internalNext(int32_t n) {
+  if(n == UBRK_DONE || // at end  or
+    fData->fBackwardsTrie.isNull()) { // .. no backwards table loaded == no exceptions
+      return n;
+  }
+  // OK, do we need to break here?
+  UErrorCode status = U_ZERO_ERROR;
+  // refresh text
+  resetState(status);
+  if(U_FAILURE(status)) return UBRK_DONE; // bail out
+  int64_t utextLen = utext_nativeLength(fText.getAlias());
+
+  //if(debug2) u_printf("str, native len=%d\n", utext_nativeLength(fText.getAlias()));
+  while (n != UBRK_DONE && n != utextLen) { // outer loop runs once per underlying break (from fDelegate).
+    SimpleFilteredSentenceBreakIterator::EFBMatchResult m = breakExceptionAt(n);
+
+    switch(m) {
+    case kExceptionHere:
+      n = fDelegate->next(); // skip this one. Find the next lowerlevel break.
+      continue;
+
+    default:
+    case kNoExceptionHere:
+      return n;
+    }    
+  }
+  return n;
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::internalPrev(int32_t n) {
+  if(n == 0 || n == UBRK_DONE || // at end  or
+    fData->fBackwardsTrie.isNull()) { // .. no backwards table loaded == no exceptions
+      return n;
+  }
+  // OK, do we need to break here?
+  UErrorCode status = U_ZERO_ERROR;
+  // refresh text
+  resetState(status);
+  if(U_FAILURE(status)) return UBRK_DONE; // bail out
+
+  //if(debug2) u_printf("str, native len=%d\n", utext_nativeLength(fText.getAlias()));
+  while (n != UBRK_DONE && n != 0) { // outer loop runs once per underlying break (from fDelegate).
+    SimpleFilteredSentenceBreakIterator::EFBMatchResult m = breakExceptionAt(n);
+
+    switch(m) {
+    case kExceptionHere:
+      n = fDelegate->previous(); // skip this one. Find the next lowerlevel break.
+      continue;
+
+    default:
+    case kNoExceptionHere:
+      return n;
+    }    
+  }
+  return n;
+}
+
+
+int32_t
+SimpleFilteredSentenceBreakIterator::next() {
+  return internalNext(fDelegate->next());
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::first(void) {
+  // Don't suppress a break opportunity at the beginning of text.
+  return fDelegate->first();
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::preceding(int32_t offset) {
+  return internalPrev(fDelegate->preceding(offset));
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::previous(void) {
+  return internalPrev(fDelegate->previous());
+}
+
+UBool SimpleFilteredSentenceBreakIterator::isBoundary(int32_t offset) {
+  if (!fDelegate->isBoundary(offset)) return false; // no break to suppress
+
+  if (fData->fBackwardsTrie.isNull()) return true; // no data = no suppressions
+
+  UErrorCode status = U_ZERO_ERROR;
+  resetState(status);
+
+  SimpleFilteredSentenceBreakIterator::EFBMatchResult m = breakExceptionAt(offset);
+
+  switch(m) {
+  case kExceptionHere:
+    return false;
+  default:
+  case kNoExceptionHere:
+    return true;
+  }    
+}
+ 
+int32_t
+SimpleFilteredSentenceBreakIterator::next(int32_t offset) {
+  return internalNext(fDelegate->next(offset));
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::following(int32_t offset) {
+  return internalNext(fDelegate->following(offset));
+}
+
+int32_t
+SimpleFilteredSentenceBreakIterator::last(void) {
+  // Don't suppress a break opportunity at the end of text.
+  return fDelegate->last();
+}
+
+
+/**
+ * Concrete implementation of builder class.
+ */
+class U_COMMON_API SimpleFilteredBreakIteratorBuilder : public FilteredBreakIteratorBuilder {
+public:
+  virtual ~SimpleFilteredBreakIteratorBuilder();
+  SimpleFilteredBreakIteratorBuilder(const Locale &fromLocale, UErrorCode &status);
+  SimpleFilteredBreakIteratorBuilder(UErrorCode &status);
+  virtual UBool suppressBreakAfter(const UnicodeString& exception, UErrorCode& status);
+  virtual UBool unsuppressBreakAfter(const UnicodeString& exception, UErrorCode& status);
+  virtual BreakIterator *build(BreakIterator* adoptBreakIterator, UErrorCode& status);
+private:
+  UStringSet fSet;
+};
+
+SimpleFilteredBreakIteratorBuilder::~SimpleFilteredBreakIteratorBuilder()
+{
+}
+
+SimpleFilteredBreakIteratorBuilder::SimpleFilteredBreakIteratorBuilder(UErrorCode &status) 
+  : fSet(status)
+{
+}
+
+SimpleFilteredBreakIteratorBuilder::SimpleFilteredBreakIteratorBuilder(const Locale &fromLocale, UErrorCode &status)
+  : fSet(status)
+{
+  if(U_SUCCESS(status)) {
+    UErrorCode subStatus = U_ZERO_ERROR;
+    LocalUResourceBundlePointer b(ures_open(U_ICUDATA_BRKITR, fromLocale.getBaseName(), &subStatus));
+    if (U_FAILURE(subStatus) || (subStatus == U_USING_DEFAULT_WARNING) ) {    
+      status = subStatus; // copy the failing status 
+#if FB_DEBUG
+      fprintf(stderr, "open BUNDLE %s : %s, %s\n", fromLocale.getBaseName(), "[exit]", u_errorName(status));
+#endif
+      return;  // leaves the builder empty, if you try to use it.
+    }
+    LocalUResourceBundlePointer exceptions(ures_getByKeyWithFallback(b.getAlias(), "exceptions", NULL, &subStatus));
+    if (U_FAILURE(subStatus) || (subStatus == U_USING_DEFAULT_WARNING) ) {    
+      status = subStatus; // copy the failing status 
+#if FB_DEBUG
+      fprintf(stderr, "open EXCEPTIONS %s : %s, %s\n", fromLocale.getBaseName(), "[exit]", u_errorName(status));
+#endif
+      return;  // leaves the builder empty, if you try to use it.
+    }
+    LocalUResourceBundlePointer breaks(ures_getByKeyWithFallback(exceptions.getAlias(), "SentenceBreak", NULL, &subStatus));
+
+#if FB_DEBUG
+    {
+      UErrorCode subsub = subStatus;
+      fprintf(stderr, "open SentenceBreak %s => %s, %s\n", fromLocale.getBaseName(), ures_getLocale(breaks.getAlias(), &subsub), u_errorName(subStatus));
+    }
+#endif
+    
+    if (U_FAILURE(subStatus) || (subStatus == U_USING_DEFAULT_WARNING) ) {    
+      status = subStatus; // copy the failing status 
+#if FB_DEBUG
+      fprintf(stderr, "open %s : %s, %s\n", fromLocale.getBaseName(), "[exit]", u_errorName(status));
+#endif
+      return;  // leaves the builder empty, if you try to use it.
+    }
+
+    LocalUResourceBundlePointer strs;
+    subStatus = status; // Pick up inherited warning status now 
+    do {
+      strs.adoptInstead(ures_getNextResource(breaks.getAlias(), strs.orphan(), &subStatus));
+      if(strs.isValid() && U_SUCCESS(subStatus)) {
+        UnicodeString str(ures_getUnicodeString(strs.getAlias(), &status));
+        suppressBreakAfter(str, status); // load the string
+      }
+    } while (strs.isValid() && U_SUCCESS(subStatus));
+    if(U_FAILURE(subStatus)&&subStatus!=U_INDEX_OUTOFBOUNDS_ERROR&&U_SUCCESS(status)) {
+      status = subStatus;
+    }
+  }
+}
+
+UBool
+SimpleFilteredBreakIteratorBuilder::suppressBreakAfter(const UnicodeString& exception, UErrorCode& status)
+{
+  UBool r = fSet.add(exception, status);
+  FB_TRACE("suppressBreakAfter",&exception,r,0);
+  return r;
+}
+
+UBool
+SimpleFilteredBreakIteratorBuilder::unsuppressBreakAfter(const UnicodeString& exception, UErrorCode& status)
+{
+  UBool r = fSet.remove(exception, status);
+  FB_TRACE("unsuppressBreakAfter",&exception,r,0);
+  return r;
+}
+
+/**
+ * Jitterbug 2974: MSVC has a bug whereby new X[0] behaves badly.
+ * Work around this.
+ *
+ * Note: "new UnicodeString[subCount]" ends up calling global operator new
+ * on MSVC2012 for some reason.
+ */
+static inline UnicodeString* newUnicodeStringArray(size_t count) {
+    return new UnicodeString[count ? count : 1];
+}
+
+BreakIterator *
+SimpleFilteredBreakIteratorBuilder::build(BreakIterator* adoptBreakIterator, UErrorCode& status) {
+  LocalPointer<BreakIterator> adopt(adoptBreakIterator);
+
+  LocalPointer<UCharsTrieBuilder> builder(new UCharsTrieBuilder(status), status);
+  LocalPointer<UCharsTrieBuilder> builder2(new UCharsTrieBuilder(status), status);
+  if(U_FAILURE(status)) {
+    return NULL;
+  }
+
+  int32_t revCount = 0;
+  int32_t fwdCount = 0;
+
+  int32_t subCount = fSet.size();
+
+  UnicodeString *ustrs_ptr = newUnicodeStringArray(subCount);
+  
+  LocalArray<UnicodeString> ustrs(ustrs_ptr);
+
+  LocalMemory<int> partials;
+  partials.allocateInsteadAndReset(subCount);
+
+  LocalPointer<UCharsTrie>    backwardsTrie; //  i.e. ".srM" for Mrs.
+  LocalPointer<UCharsTrie>    forwardsPartialTrie; //  Has ".a" for "a.M."
+
+  int n=0;
+  for ( int32_t i = 0;
+        i<fSet.size();
+        i++) {
+    const UnicodeString *abbr = fSet.getStringAt(i);
+    if(abbr) {
+      FB_TRACE("build",abbr,TRUE,i);
+      ustrs[n] = *abbr; // copy by value
+      FB_TRACE("ustrs[n]",&ustrs[n],TRUE,i);
+    } else {
+      FB_TRACE("build",abbr,FALSE,i);
+      status = U_MEMORY_ALLOCATION_ERROR;
+      return NULL;
+    }
+    partials[n] = 0; // default: not partial
+    n++;
+  }
+  // first pass - find partials.
+  for(int i=0;i<subCount;i++) {
+    int nn = ustrs[i].indexOf(kFULLSTOP); // TODO: non-'.' abbreviations
+    if(nn>-1 && (nn+1)!=ustrs[i].length()) {
+      FB_TRACE("partial",&ustrs[i],FALSE,i);
+      // is partial.
+      // is it unique?
+      int sameAs = -1;
+      for(int j=0;j<subCount;j++) {
+        if(j==i) continue;
+        if(ustrs[i].compare(0,nn+1,ustrs[j],0,nn+1)==0) {
+          FB_TRACE("prefix",&ustrs[j],FALSE,nn+1);
+          //UBool otherIsPartial = ((nn+1)!=ustrs[j].length());  // true if ustrs[j] doesn't end at nn
+          if(partials[j]==0) { // hasn't been processed yet
+            partials[j] = kSuppressInReverse | kAddToForward;
+            FB_TRACE("suppressing",&ustrs[j],FALSE,j);
+          } else if(partials[j] & kSuppressInReverse) {
+            sameAs = j; // the other entry is already in the reverse table.
+          }
+        }
+      }
+      FB_TRACE("for partial same-",&ustrs[i],FALSE,sameAs);
+      FB_TRACE(" == partial #",&ustrs[i],FALSE,partials[i]);
+      UnicodeString prefix(ustrs[i], 0, nn+1);
+      if(sameAs == -1 && partials[i] == 0) {
+        // first one - add the prefix to the reverse table.
+        prefix.reverse();
+        builder->add(prefix, kPARTIAL, status);
+        revCount++;
+        FB_TRACE("Added partial",&prefix,FALSE, i);
+        FB_TRACE(u_errorName(status),&ustrs[i],FALSE,i);
+        partials[i] = kSuppressInReverse | kAddToForward;
+      } else {
+        FB_TRACE("NOT adding partial",&prefix,FALSE, i);
+        FB_TRACE(u_errorName(status),&ustrs[i],FALSE,i);
+      }
+    }
+  }
+  for(int i=0;i<subCount;i++) {
+    if(partials[i]==0) {
+      ustrs[i].reverse();
+      builder->add(ustrs[i], kMATCH, status);
+      revCount++;
+      FB_TRACE(u_errorName(status), &ustrs[i], FALSE, i);
+    } else {
+      FB_TRACE("Adding fwd",&ustrs[i], FALSE, i);
+
+      // an optimization would be to only add the portion after the '.'
+      // for example, for "Ph.D." we store ".hP" in the reverse table. We could just store "D." in the forward,
+      // instead of "Ph.D." since we already know the "Ph." part is a match.
+      // would need the trie to be able to hold 0-length strings, though.
+      builder2->add(ustrs[i], kMATCH, status); // forward
+      fwdCount++;
+      //ustrs[i].reverse();
+      ////if(debug2) u_printf("SUPPRESS- not Added(%d):  /%S/ status=%s\n",partials[i], ustrs[i].getTerminatedBuffer(), u_errorName(status));
+    }
+  }
+  FB_TRACE("AbbrCount",NULL,FALSE, subCount);
+
+  if(revCount>0) {
+    backwardsTrie.adoptInstead(builder->build(USTRINGTRIE_BUILD_FAST, status));
+    if(U_FAILURE(status)) {
+      FB_TRACE(u_errorName(status),NULL,FALSE, -1);
+      return NULL;
+    }
+  }
+
+  if(fwdCount>0) {
+    forwardsPartialTrie.adoptInstead(builder2->build(USTRINGTRIE_BUILD_FAST, status));
+    if(U_FAILURE(status)) {
+      FB_TRACE(u_errorName(status),NULL,FALSE, -1);
+      return NULL;
+    }
+  }
+
+  return new SimpleFilteredSentenceBreakIterator(adopt.orphan(), forwardsPartialTrie.orphan(), backwardsTrie.orphan(), status);
+}
+
+
+// ----------- Base class implementation
+
+FilteredBreakIteratorBuilder::FilteredBreakIteratorBuilder() {
+}
+
+FilteredBreakIteratorBuilder::~FilteredBreakIteratorBuilder() {
+}
+
+FilteredBreakIteratorBuilder *
+FilteredBreakIteratorBuilder::createInstance(const Locale& where, UErrorCode& status) {
+  if(U_FAILURE(status)) return NULL;
+  LocalPointer<FilteredBreakIteratorBuilder> ret(new SimpleFilteredBreakIteratorBuilder(where, status), status);
+  return (U_SUCCESS(status))? ret.orphan(): NULL;
+}
+
+FilteredBreakIteratorBuilder *
+FilteredBreakIteratorBuilder::createInstance(UErrorCode &status) {
+  return createEmptyInstance(status);
+}
+
+FilteredBreakIteratorBuilder *
+FilteredBreakIteratorBuilder::createEmptyInstance(UErrorCode& status) {
+  if(U_FAILURE(status)) return NULL;
+  LocalPointer<FilteredBreakIteratorBuilder> ret(new SimpleFilteredBreakIteratorBuilder(status), status);
+  return (U_SUCCESS(status))? ret.orphan(): NULL;
+}
+
+U_NAMESPACE_END
+
+#endif //#if !UCONFIG_NO_BREAK_ITERATION && !UCONFIG_NO_FILTERED_BREAK_ITERATION
diff --git a/thirdparty/icu4c/common/filterednormalizer2.cpp b/thirdparty/icu4c/common/filterednormalizer2.cpp
new file mode 100644
index 0000000000..1a0914d3f7
--- /dev/null
+++ b/thirdparty/icu4c/common/filterednormalizer2.cpp
@@ -0,0 +1,363 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  filterednormalizer2.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009dec10
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/edits.h"
+#include "unicode/normalizer2.h"
+#include "unicode/stringoptions.h"
+#include "unicode/uniset.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm.h"
+#include "cpputils.h"
+
+U_NAMESPACE_BEGIN
+
+FilteredNormalizer2::~FilteredNormalizer2() {}
+
+UnicodeString &
+FilteredNormalizer2::normalize(const UnicodeString &src,
+                               UnicodeString &dest,
+                               UErrorCode &errorCode) const {
+    uprv_checkCanGetBuffer(src, errorCode);
+    if(U_FAILURE(errorCode)) {
+        dest.setToBogus();
+        return dest;
+    }
+    if(&dest==&src) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return dest;
+    }
+    dest.remove();
+    return normalize(src, dest, USET_SPAN_SIMPLE, errorCode);
+}
+
+// Internal: No argument checking, and appends to dest.
+// Pass as input spanCondition the one that is likely to yield a non-zero
+// span length at the start of src.
+// For set=[:age=3.2:], since almost all common characters were in Unicode 3.2,
+// USET_SPAN_SIMPLE should be passed in for the start of src
+// and USET_SPAN_NOT_CONTAINED should be passed in if we continue after
+// an in-filter prefix.
+UnicodeString &
+FilteredNormalizer2::normalize(const UnicodeString &src,
+                               UnicodeString &dest,
+                               USetSpanCondition spanCondition,
+                               UErrorCode &errorCode) const {
+    UnicodeString tempDest;  // Don't throw away destination buffer between iterations.
+    for(int32_t prevSpanLimit=0; prevSpanLimit<src.length();) {
+        int32_t spanLimit=set.span(src, prevSpanLimit, spanCondition);
+        int32_t spanLength=spanLimit-prevSpanLimit;
+        if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+            if(spanLength!=0) {
+                dest.append(src, prevSpanLimit, spanLength);
+            }
+            spanCondition=USET_SPAN_SIMPLE;
+        } else {
+            if(spanLength!=0) {
+                // Not norm2.normalizeSecondAndAppend() because we do not want
+                // to modify the non-filter part of dest.
+                dest.append(norm2.normalize(src.tempSubStringBetween(prevSpanLimit, spanLimit),
+                                            tempDest, errorCode));
+                if(U_FAILURE(errorCode)) {
+                    break;
+                }
+            }
+            spanCondition=USET_SPAN_NOT_CONTAINED;
+        }
+        prevSpanLimit=spanLimit;
+    }
+    return dest;
+}
+
+void
+FilteredNormalizer2::normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink,
+                                   Edits *edits, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if (edits != nullptr && (options & U_EDITS_NO_RESET) == 0) {
+        edits->reset();
+    }
+    options |= U_EDITS_NO_RESET;  // Do not reset for each span.
+    normalizeUTF8(options, src.data(), src.length(), sink, edits, USET_SPAN_SIMPLE, errorCode);
+}
+
+void
+FilteredNormalizer2::normalizeUTF8(uint32_t options, const char *src, int32_t length,
+                                   ByteSink &sink, Edits *edits,
+                                   USetSpanCondition spanCondition,
+                                   UErrorCode &errorCode) const {
+    while (length > 0) {
+        int32_t spanLength = set.spanUTF8(src, length, spanCondition);
+        if (spanCondition == USET_SPAN_NOT_CONTAINED) {
+            if (spanLength != 0) {
+                if (edits != nullptr) {
+                    edits->addUnchanged(spanLength);
+                }
+                if ((options & U_OMIT_UNCHANGED_TEXT) == 0) {
+                    sink.Append(src, spanLength);
+                }
+            }
+            spanCondition = USET_SPAN_SIMPLE;
+        } else {
+            if (spanLength != 0) {
+                // Not norm2.normalizeSecondAndAppend() because we do not want
+                // to modify the non-filter part of dest.
+                norm2.normalizeUTF8(options, StringPiece(src, spanLength), sink, edits, errorCode);
+                if (U_FAILURE(errorCode)) {
+                    break;
+                }
+            }
+            spanCondition = USET_SPAN_NOT_CONTAINED;
+        }
+        src += spanLength;
+        length -= spanLength;
+    }
+}
+
+UnicodeString &
+FilteredNormalizer2::normalizeSecondAndAppend(UnicodeString &first,
+                                              const UnicodeString &second,
+                                              UErrorCode &errorCode) const {
+    return normalizeSecondAndAppend(first, second, TRUE, errorCode);
+}
+
+UnicodeString &
+FilteredNormalizer2::append(UnicodeString &first,
+                            const UnicodeString &second,
+                            UErrorCode &errorCode) const {
+    return normalizeSecondAndAppend(first, second, FALSE, errorCode);
+}
+
+UnicodeString &
+FilteredNormalizer2::normalizeSecondAndAppend(UnicodeString &first,
+                                              const UnicodeString &second,
+                                              UBool doNormalize,
+                                              UErrorCode &errorCode) const {
+    uprv_checkCanGetBuffer(first, errorCode);
+    uprv_checkCanGetBuffer(second, errorCode);
+    if(U_FAILURE(errorCode)) {
+        return first;
+    }
+    if(&first==&second) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return first;
+    }
+    if(first.isEmpty()) {
+        if(doNormalize) {
+            return normalize(second, first, errorCode);
+        } else {
+            return first=second;
+        }
+    }
+    // merge the in-filter suffix of the first string with the in-filter prefix of the second
+    int32_t prefixLimit=set.span(second, 0, USET_SPAN_SIMPLE);
+    if(prefixLimit!=0) {
+        UnicodeString prefix(second.tempSubString(0, prefixLimit));
+        int32_t suffixStart=set.spanBack(first, INT32_MAX, USET_SPAN_SIMPLE);
+        if(suffixStart==0) {
+            if(doNormalize) {
+                norm2.normalizeSecondAndAppend(first, prefix, errorCode);
+            } else {
+                norm2.append(first, prefix, errorCode);
+            }
+        } else {
+            UnicodeString middle(first, suffixStart, INT32_MAX);
+            if(doNormalize) {
+                norm2.normalizeSecondAndAppend(middle, prefix, errorCode);
+            } else {
+                norm2.append(middle, prefix, errorCode);
+            }
+            first.replace(suffixStart, INT32_MAX, middle);
+        }
+    }
+    if(prefixLimit<second.length()) {
+        UnicodeString rest(second.tempSubString(prefixLimit, INT32_MAX));
+        if(doNormalize) {
+            normalize(rest, first, USET_SPAN_NOT_CONTAINED, errorCode);
+        } else {
+            first.append(rest);
+        }
+    }
+    return first;
+}
+
+UBool
+FilteredNormalizer2::getDecomposition(UChar32 c, UnicodeString &decomposition) const {
+    return set.contains(c) && norm2.getDecomposition(c, decomposition);
+}
+
+UBool
+FilteredNormalizer2::getRawDecomposition(UChar32 c, UnicodeString &decomposition) const {
+    return set.contains(c) && norm2.getRawDecomposition(c, decomposition);
+}
+
+UChar32
+FilteredNormalizer2::composePair(UChar32 a, UChar32 b) const {
+    return (set.contains(a) && set.contains(b)) ? norm2.composePair(a, b) : U_SENTINEL;
+}
+
+uint8_t
+FilteredNormalizer2::getCombiningClass(UChar32 c) const {
+    return set.contains(c) ? norm2.getCombiningClass(c) : 0;
+}
+
+UBool
+FilteredNormalizer2::isNormalized(const UnicodeString &s, UErrorCode &errorCode) const {
+    uprv_checkCanGetBuffer(s, errorCode);
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    USetSpanCondition spanCondition=USET_SPAN_SIMPLE;
+    for(int32_t prevSpanLimit=0; prevSpanLimit<s.length();) {
+        int32_t spanLimit=set.span(s, prevSpanLimit, spanCondition);
+        if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+            spanCondition=USET_SPAN_SIMPLE;
+        } else {
+            if( !norm2.isNormalized(s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode) ||
+                U_FAILURE(errorCode)
+            ) {
+                return FALSE;
+            }
+            spanCondition=USET_SPAN_NOT_CONTAINED;
+        }
+        prevSpanLimit=spanLimit;
+    }
+    return TRUE;
+}
+
+UBool
+FilteredNormalizer2::isNormalizedUTF8(StringPiece sp, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    const char *s = sp.data();
+    int32_t length = sp.length();
+    USetSpanCondition spanCondition = USET_SPAN_SIMPLE;
+    while (length > 0) {
+        int32_t spanLength = set.spanUTF8(s, length, spanCondition);
+        if (spanCondition == USET_SPAN_NOT_CONTAINED) {
+            spanCondition = USET_SPAN_SIMPLE;
+        } else {
+            if (!norm2.isNormalizedUTF8(StringPiece(s, spanLength), errorCode) ||
+                    U_FAILURE(errorCode)) {
+                return FALSE;
+            }
+            spanCondition = USET_SPAN_NOT_CONTAINED;
+        }
+        s += spanLength;
+        length -= spanLength;
+    }
+    return TRUE;
+}
+
+UNormalizationCheckResult
+FilteredNormalizer2::quickCheck(const UnicodeString &s, UErrorCode &errorCode) const {
+    uprv_checkCanGetBuffer(s, errorCode);
+    if(U_FAILURE(errorCode)) {
+        return UNORM_MAYBE;
+    }
+    UNormalizationCheckResult result=UNORM_YES;
+    USetSpanCondition spanCondition=USET_SPAN_SIMPLE;
+    for(int32_t prevSpanLimit=0; prevSpanLimit<s.length();) {
+        int32_t spanLimit=set.span(s, prevSpanLimit, spanCondition);
+        if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+            spanCondition=USET_SPAN_SIMPLE;
+        } else {
+            UNormalizationCheckResult qcResult=
+                norm2.quickCheck(s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode);
+            if(U_FAILURE(errorCode) || qcResult==UNORM_NO) {
+                return qcResult;
+            } else if(qcResult==UNORM_MAYBE) {
+                result=qcResult;
+            }
+            spanCondition=USET_SPAN_NOT_CONTAINED;
+        }
+        prevSpanLimit=spanLimit;
+    }
+    return result;
+}
+
+int32_t
+FilteredNormalizer2::spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const {
+    uprv_checkCanGetBuffer(s, errorCode);
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    USetSpanCondition spanCondition=USET_SPAN_SIMPLE;
+    for(int32_t prevSpanLimit=0; prevSpanLimit<s.length();) {
+        int32_t spanLimit=set.span(s, prevSpanLimit, spanCondition);
+        if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+            spanCondition=USET_SPAN_SIMPLE;
+        } else {
+            int32_t yesLimit=
+                prevSpanLimit+
+                norm2.spanQuickCheckYes(
+                    s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode);
+            if(U_FAILURE(errorCode) || yesLimit<spanLimit) {
+                return yesLimit;
+            }
+            spanCondition=USET_SPAN_NOT_CONTAINED;
+        }
+        prevSpanLimit=spanLimit;
+    }
+    return s.length();
+}
+
+UBool
+FilteredNormalizer2::hasBoundaryBefore(UChar32 c) const {
+    return !set.contains(c) || norm2.hasBoundaryBefore(c);
+}
+
+UBool
+FilteredNormalizer2::hasBoundaryAfter(UChar32 c) const {
+    return !set.contains(c) || norm2.hasBoundaryAfter(c);
+}
+
+UBool
+FilteredNormalizer2::isInert(UChar32 c) const {
+    return !set.contains(c) || norm2.isInert(c);
+}
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+U_CAPI UNormalizer2 * U_EXPORT2
+unorm2_openFiltered(const UNormalizer2 *norm2, const USet *filterSet, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(filterSet==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    Normalizer2 *fn2=new FilteredNormalizer2(*(Normalizer2 *)norm2,
+                                             *UnicodeSet::fromUSet(filterSet));
+    if(fn2==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+    return (UNormalizer2 *)fn2;
+}
+
+#endif  // !UCONFIG_NO_NORMALIZATION
diff --git a/thirdparty/icu4c/common/hash.h b/thirdparty/icu4c/common/hash.h
new file mode 100644
index 0000000000..f02cb7087a
--- /dev/null
+++ b/thirdparty/icu4c/common/hash.h
@@ -0,0 +1,248 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 1997-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   03/28/00    aliu        Creation.
+******************************************************************************
+*/
+
+#ifndef HASH_H
+#define HASH_H
+
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+#include "cmemory.h"
+#include "uhash.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Hashtable is a thin C++ wrapper around UHashtable, a general-purpose void*
+ * hashtable implemented in C.  Hashtable is designed to be idiomatic and
+ * easy-to-use in C++.
+ *
+ * Hashtable is an INTERNAL CLASS.
+ */
+class U_COMMON_API Hashtable : public UMemory {
+    UHashtable* hash;
+    UHashtable hashObj;
+
+    inline void init(UHashFunction *keyHash, UKeyComparator *keyComp, UValueComparator *valueComp, UErrorCode& status);
+
+    inline void initSize(UHashFunction *keyHash, UKeyComparator *keyComp, UValueComparator *valueComp, int32_t size, UErrorCode& status);
+
+public:
+    /**
+     * Construct a hashtable
+     * @param ignoreKeyCase If true, keys are case insensitive.
+     * @param status Error code
+    */
+    inline Hashtable(UBool ignoreKeyCase, UErrorCode& status);
+
+    /**
+     * Construct a hashtable
+     * @param ignoreKeyCase If true, keys are case insensitive.
+     * @param size initial size allocation
+     * @param status Error code
+    */
+    inline Hashtable(UBool ignoreKeyCase, int32_t size, UErrorCode& status);
+
+    /**
+     * Construct a hashtable
+     * @param keyComp Comparator for comparing the keys
+     * @param valueComp Comparator for comparing the values
+     * @param status Error code
+    */
+    inline Hashtable(UKeyComparator *keyComp, UValueComparator *valueComp, UErrorCode& status);
+
+    /**
+     * Construct a hashtable
+     * @param status Error code
+    */
+    inline Hashtable(UErrorCode& status);
+
+    /**
+     * Construct a hashtable, _disregarding any error_.  Use this constructor
+     * with caution.
+     */
+    inline Hashtable();
+
+    /**
+     * Non-virtual destructor; make this virtual if Hashtable is subclassed
+     * in the future.
+     */
+    inline ~Hashtable();
+
+    inline UObjectDeleter *setValueDeleter(UObjectDeleter *fn);
+
+    inline int32_t count() const;
+
+    inline void* put(const UnicodeString& key, void* value, UErrorCode& status);
+
+    inline int32_t puti(const UnicodeString& key, int32_t value, UErrorCode& status);
+
+    inline void* get(const UnicodeString& key) const;
+
+    inline int32_t geti(const UnicodeString& key) const;
+
+    inline void* remove(const UnicodeString& key);
+
+    inline int32_t removei(const UnicodeString& key);
+
+    inline void removeAll(void);
+
+    inline const UHashElement* find(const UnicodeString& key) const;
+
+    /**
+     * @param pos - must be UHASH_FIRST on first call, and untouched afterwards.
+     * @see uhash_nextElement
+     */
+    inline const UHashElement* nextElement(int32_t& pos) const;
+
+    inline UKeyComparator* setKeyComparator(UKeyComparator*keyComp);
+
+    inline UValueComparator* setValueComparator(UValueComparator* valueComp);
+
+    inline UBool equals(const Hashtable& that) const;
+private:
+    Hashtable(const Hashtable &other); // forbid copying of this class
+    Hashtable &operator=(const Hashtable &other); // forbid copying of this class
+};
+
+/*********************************************************************
+ * Implementation
+ ********************************************************************/
+
+inline void Hashtable::init(UHashFunction *keyHash, UKeyComparator *keyComp,
+                            UValueComparator *valueComp, UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    uhash_init(&hashObj, keyHash, keyComp, valueComp, &status);
+    if (U_SUCCESS(status)) {
+        hash = &hashObj;
+        uhash_setKeyDeleter(hash, uprv_deleteUObject);
+    }
+}
+
+inline void Hashtable::initSize(UHashFunction *keyHash, UKeyComparator *keyComp,
+                                UValueComparator *valueComp, int32_t size, UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    uhash_initSize(&hashObj, keyHash, keyComp, valueComp, size, &status);
+    if (U_SUCCESS(status)) {
+        hash = &hashObj;
+        uhash_setKeyDeleter(hash, uprv_deleteUObject);
+    }
+}
+
+inline Hashtable::Hashtable(UKeyComparator *keyComp, UValueComparator *valueComp,
+                 UErrorCode& status) : hash(0) {
+    init( uhash_hashUnicodeString, keyComp, valueComp, status);
+}
+
+inline Hashtable::Hashtable(UBool ignoreKeyCase, UErrorCode& status)
+ : hash(0)
+{
+    init(ignoreKeyCase ? uhash_hashCaselessUnicodeString
+                        : uhash_hashUnicodeString,
+            ignoreKeyCase ? uhash_compareCaselessUnicodeString
+                        : uhash_compareUnicodeString,
+            NULL,
+            status);
+}
+
+inline Hashtable::Hashtable(UBool ignoreKeyCase, int32_t size, UErrorCode& status)
+ : hash(0)
+{
+    initSize(ignoreKeyCase ? uhash_hashCaselessUnicodeString
+                        : uhash_hashUnicodeString,
+            ignoreKeyCase ? uhash_compareCaselessUnicodeString
+                        : uhash_compareUnicodeString,
+            NULL, size,
+            status);
+}
+
+inline Hashtable::Hashtable(UErrorCode& status)
+ : hash(0)
+{
+    init(uhash_hashUnicodeString, uhash_compareUnicodeString, NULL, status);
+}
+
+inline Hashtable::Hashtable()
+ : hash(0)
+{
+    UErrorCode status = U_ZERO_ERROR;
+    init(uhash_hashUnicodeString, uhash_compareUnicodeString, NULL, status);
+}
+
+inline Hashtable::~Hashtable() {
+    if (hash != NULL) {
+        uhash_close(hash);
+    }
+}
+
+inline UObjectDeleter *Hashtable::setValueDeleter(UObjectDeleter *fn) {
+    return uhash_setValueDeleter(hash, fn);
+}
+
+inline int32_t Hashtable::count() const {
+    return uhash_count(hash);
+}
+
+inline void* Hashtable::put(const UnicodeString& key, void* value, UErrorCode& status) {
+    return uhash_put(hash, new UnicodeString(key), value, &status);
+}
+
+inline int32_t Hashtable::puti(const UnicodeString& key, int32_t value, UErrorCode& status) {
+    return uhash_puti(hash, new UnicodeString(key), value, &status);
+}
+
+inline void* Hashtable::get(const UnicodeString& key) const {
+    return uhash_get(hash, &key);
+}
+
+inline int32_t Hashtable::geti(const UnicodeString& key) const {
+    return uhash_geti(hash, &key);
+}
+
+inline void* Hashtable::remove(const UnicodeString& key) {
+    return uhash_remove(hash, &key);
+}
+
+inline int32_t Hashtable::removei(const UnicodeString& key) {
+    return uhash_removei(hash, &key);
+}
+
+inline const UHashElement* Hashtable::find(const UnicodeString& key) const {
+    return uhash_find(hash, &key);
+}
+
+inline const UHashElement* Hashtable::nextElement(int32_t& pos) const {
+    return uhash_nextElement(hash, &pos);
+}
+
+inline void Hashtable::removeAll(void) {
+    uhash_removeAll(hash);
+}
+
+inline UKeyComparator* Hashtable::setKeyComparator(UKeyComparator*keyComp){
+    return uhash_setKeyComparator(hash, keyComp);
+}
+
+inline UValueComparator* Hashtable::setValueComparator(UValueComparator* valueComp){
+    return uhash_setValueComparator(hash, valueComp);
+}
+
+inline UBool Hashtable::equals(const Hashtable& that)const{
+   return uhash_equals(hash, that.hash);
+}
+U_NAMESPACE_END
+
+#endif
+
diff --git a/thirdparty/icu4c/common/icudataver.cpp b/thirdparty/icu4c/common/icudataver.cpp
new file mode 100644
index 0000000000..d314411374
--- /dev/null
+++ b/thirdparty/icu4c/common/icudataver.cpp
@@ -0,0 +1,31 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2009-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/icudataver.h"
+#include "unicode/ures.h"
+#include "uresimp.h" /* for ures_getVersionByKey */
+
+U_CAPI void U_EXPORT2 u_getDataVersion(UVersionInfo dataVersionFillin, UErrorCode *status) {
+    UResourceBundle *icudatares = NULL;
+    
+    if (U_FAILURE(*status)) {
+        return;
+    }
+    
+    if (dataVersionFillin != NULL) {
+        icudatares = ures_openDirect(NULL, U_ICU_VERSION_BUNDLE , status);
+        if (U_SUCCESS(*status)) {
+            ures_getVersionByKey(icudatares, U_ICU_DATA_KEY, dataVersionFillin, status);
+        }
+        ures_close(icudatares);
+    }
+}
diff --git a/thirdparty/icu4c/common/icuplug.cpp b/thirdparty/icu4c/common/icuplug.cpp
new file mode 100644
index 0000000000..4ab8c66ebe
--- /dev/null
+++ b/thirdparty/icu4c/common/icuplug.cpp
@@ -0,0 +1,884 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2009-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : icuplug.c
+*
+*   Date         Name        Description
+*   10/29/2009   sl          New.
+******************************************************************************
+*/
+
+#include "unicode/icuplug.h"
+
+
+#if UCONFIG_ENABLE_PLUGINS
+
+
+#include "icuplugimp.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "putilimp.h"
+#include "ucln.h"
+#include <stdio.h>
+#ifdef __MVS__  /* defined by z/OS compiler */
+#define _POSIX_SOURCE
+#include <cics.h> /* 12 Nov 2011 JAM iscics() function */
+#endif
+#include "charstr.h"
+
+using namespace icu;
+
+#ifndef UPLUG_TRACE
+#define UPLUG_TRACE 0
+#endif
+
+#if UPLUG_TRACE
+#include <stdio.h>
+#define DBG(x) fprintf(stderr, "%s:%d: ",__FILE__,__LINE__); fprintf x
+#endif
+
+/**
+ * Internal structure of an ICU plugin. 
+ */
+
+struct UPlugData {
+  UPlugEntrypoint  *entrypoint; /**< plugin entrypoint */
+  uint32_t structSize;    /**< initialized to the size of this structure */
+  uint32_t token;         /**< must be U_PLUG_TOKEN */
+  void *lib;              /**< plugin library, or NULL */
+  char libName[UPLUG_NAME_MAX];   /**< library name */
+  char sym[UPLUG_NAME_MAX];        /**< plugin symbol, or NULL */
+  char config[UPLUG_NAME_MAX];     /**< configuration data */
+  void *context;          /**< user context data */
+  char name[UPLUG_NAME_MAX];   /**< name of plugin */
+  UPlugLevel  level; /**< level of plugin */
+  UBool   awaitingLoad; /**< TRUE if the plugin is awaiting a load call */
+  UBool   dontUnload; /**< TRUE if plugin must stay resident (leak plugin and lib) */
+  UErrorCode pluginStatus; /**< status code of plugin */
+};
+
+
+
+#define UPLUG_LIBRARY_INITIAL_COUNT 8
+#define UPLUG_PLUGIN_INITIAL_COUNT 12
+
+/**
+ * Remove an item
+ * @param list the full list
+ * @param listSize the number of entries in the list
+ * @param memberSize the size of one member
+ * @param itemToRemove the item number of the member
+ * @return the new listsize 
+ */
+static int32_t uplug_removeEntryAt(void *list, int32_t listSize, int32_t memberSize, int32_t itemToRemove) {
+  uint8_t *bytePtr = (uint8_t *)list;
+    
+  /* get rid of some bad cases first */
+  if(listSize<1) {
+    return listSize;
+  }
+    
+  /* is there anything to move? */
+  if(listSize > itemToRemove+1) {
+    memmove(bytePtr+(itemToRemove*memberSize), bytePtr+((itemToRemove+1)*memberSize), memberSize);
+  }
+    
+  return listSize-1;
+}
+
+
+
+
+#if U_ENABLE_DYLOAD
+/**
+ * Library management. Internal. 
+ * @internal
+ */
+struct UPlugLibrary;
+
+/**
+ * Library management. Internal. 
+ * @internal
+ */
+typedef struct UPlugLibrary {
+  void *lib;                           /**< library ptr */
+  char name[UPLUG_NAME_MAX]; /**< library name */
+  uint32_t ref;                        /**< reference count */
+} UPlugLibrary;
+
+static UPlugLibrary   staticLibraryList[UPLUG_LIBRARY_INITIAL_COUNT];
+static UPlugLibrary * libraryList = staticLibraryList;
+static int32_t libraryCount = 0;
+static int32_t libraryMax = UPLUG_LIBRARY_INITIAL_COUNT;
+
+/**
+ * Search for a library. Doesn't lock
+ * @param libName libname to search for
+ * @return the library's struct
+ */
+static int32_t searchForLibraryName(const char *libName) {
+  int32_t i;
+    
+  for(i=0;i<libraryCount;i++) {
+    if(!uprv_strcmp(libName, libraryList[i].name)) {
+      return i;
+    }
+  }
+  return -1;
+}
+
+static int32_t searchForLibrary(void *lib) {
+  int32_t i;
+    
+  for(i=0;i<libraryCount;i++) {
+    if(lib==libraryList[i].lib) {
+      return i;
+    }
+  }
+  return -1;
+}
+
+U_CAPI char * U_EXPORT2
+uplug_findLibrary(void *lib, UErrorCode *status) {
+  int32_t libEnt;
+  char *ret = NULL;
+  if(U_FAILURE(*status)) {
+    return NULL;
+  }
+  libEnt = searchForLibrary(lib);
+  if(libEnt!=-1) { 
+    ret = libraryList[libEnt].name;
+  } else {
+    *status = U_MISSING_RESOURCE_ERROR;
+  }
+  return ret;
+}
+
+U_CAPI void * U_EXPORT2
+uplug_openLibrary(const char *libName, UErrorCode *status) {
+  int32_t libEntry = -1;
+  void *lib = NULL;
+    
+  if(U_FAILURE(*status)) return NULL;
+
+  libEntry = searchForLibraryName(libName);
+  if(libEntry == -1) {
+    libEntry = libraryCount++;
+    if(libraryCount >= libraryMax) {
+      /* Ran out of library slots. Statically allocated because we can't depend on allocating memory.. */
+      *status = U_MEMORY_ALLOCATION_ERROR;
+#if UPLUG_TRACE
+      DBG((stderr, "uplug_openLibrary() - out of library slots (max %d)\n", libraryMax));
+#endif
+      return NULL;
+    }
+    /* Some operating systems don't want 
+       DL operations from multiple threads. */
+    libraryList[libEntry].lib = uprv_dl_open(libName, status);
+#if UPLUG_TRACE
+    DBG((stderr, "uplug_openLibrary(%s,%s) libEntry %d, lib %p\n", libName, u_errorName(*status), libEntry, lib));
+#endif
+        
+    if(libraryList[libEntry].lib == NULL || U_FAILURE(*status)) {
+      /* cleanup. */
+      libraryList[libEntry].lib = NULL; /* failure with open */
+      libraryList[libEntry].name[0] = 0;
+#if UPLUG_TRACE
+      DBG((stderr, "uplug_openLibrary(%s,%s) libEntry %d, lib %p\n", libName, u_errorName(*status), libEntry, lib));
+#endif
+      /* no need to free - just won't increase the count. */
+      libraryCount--;
+    } else { /* is it still there? */
+      /* link it in */
+      uprv_strncpy(libraryList[libEntry].name,libName,UPLUG_NAME_MAX);
+      libraryList[libEntry].ref=1;
+      lib = libraryList[libEntry].lib;
+    }
+
+  } else {
+    lib = libraryList[libEntry].lib;
+    libraryList[libEntry].ref++;
+  }
+  return lib;
+}
+
+U_CAPI void U_EXPORT2
+uplug_closeLibrary(void *lib, UErrorCode *status) {
+  int32_t i;
+    
+#if UPLUG_TRACE
+  DBG((stderr, "uplug_closeLibrary(%p,%s) list %p\n", lib, u_errorName(*status), (void*)libraryList));
+#endif
+  if(U_FAILURE(*status)) return;
+    
+  for(i=0;i<libraryCount;i++) {
+    if(lib==libraryList[i].lib) {
+      if(--(libraryList[i].ref) == 0) {
+        uprv_dl_close(libraryList[i].lib, status);
+        libraryCount = uplug_removeEntryAt(libraryList, libraryCount, sizeof(*libraryList), i);
+      }
+      return;
+    }
+  }
+  *status = U_INTERNAL_PROGRAM_ERROR; /* could not find the entry! */
+}
+
+#endif
+
+static UPlugData pluginList[UPLUG_PLUGIN_INITIAL_COUNT];
+static int32_t pluginCount = 0;
+
+
+
+  
+static int32_t uplug_pluginNumber(UPlugData* d) {
+  UPlugData *pastPlug = &pluginList[pluginCount];
+  if(d<=pluginList) {
+    return 0;
+  } else if(d>=pastPlug) {
+    return pluginCount;
+  } else {
+    return (d-pluginList)/sizeof(pluginList[0]);
+  }
+}
+
+
+U_CAPI UPlugData * U_EXPORT2
+uplug_nextPlug(UPlugData *prior) {
+  if(prior==NULL) {
+    return pluginList;
+  } else {
+    UPlugData *nextPlug = &prior[1];
+    UPlugData *pastPlug = &pluginList[pluginCount];
+    
+    if(nextPlug>=pastPlug) {
+      return NULL;
+    } else {
+      return nextPlug;
+    }
+  }
+}
+
+
+
+/**
+ * Call the plugin with some params
+ */
+static void uplug_callPlug(UPlugData *plug, UPlugReason reason, UErrorCode *status) {
+  UPlugTokenReturn token;
+  if(plug==NULL||U_FAILURE(*status)) {
+    return;
+  }
+  token = (*(plug->entrypoint))(plug, reason, status);
+  if(token!=UPLUG_TOKEN) {
+    *status = U_INTERNAL_PROGRAM_ERROR;
+  }
+}
+
+
+static void uplug_unloadPlug(UPlugData *plug, UErrorCode *status) {
+  if(plug->awaitingLoad) {  /* shouldn't happen. Plugin hasn'tbeen loaded yet.*/
+    *status = U_INTERNAL_PROGRAM_ERROR;
+    return; 
+  }
+  if(U_SUCCESS(plug->pluginStatus)) {
+    /* Don't unload a plug which has a failing load status - means it didn't actually load. */
+    uplug_callPlug(plug, UPLUG_REASON_UNLOAD, status);
+  }
+}
+
+static void uplug_queryPlug(UPlugData *plug, UErrorCode *status) {
+  if(!plug->awaitingLoad || !(plug->level == UPLUG_LEVEL_UNKNOWN) ) {  /* shouldn't happen. Plugin hasn'tbeen loaded yet.*/
+    *status = U_INTERNAL_PROGRAM_ERROR;
+    return; 
+  }
+  plug->level = UPLUG_LEVEL_INVALID;
+  uplug_callPlug(plug, UPLUG_REASON_QUERY, status);
+  if(U_SUCCESS(*status)) { 
+    if(plug->level == UPLUG_LEVEL_INVALID) {
+      plug->pluginStatus = U_PLUGIN_DIDNT_SET_LEVEL;
+      plug->awaitingLoad = FALSE;
+    }
+  } else {
+    plug->pluginStatus = U_INTERNAL_PROGRAM_ERROR;
+    plug->awaitingLoad = FALSE;
+  }
+}
+
+
+static void uplug_loadPlug(UPlugData *plug, UErrorCode *status) {
+  if(U_FAILURE(*status)) {
+    return;
+  }
+  if(!plug->awaitingLoad || (plug->level < UPLUG_LEVEL_LOW) ) {  /* shouldn't happen. Plugin hasn'tbeen loaded yet.*/
+    *status = U_INTERNAL_PROGRAM_ERROR;
+    return;
+  }
+  uplug_callPlug(plug, UPLUG_REASON_LOAD, status);
+  plug->awaitingLoad = FALSE;
+  if(!U_SUCCESS(*status)) {
+    plug->pluginStatus = U_INTERNAL_PROGRAM_ERROR;
+  }
+}
+
+static UPlugData *uplug_allocateEmptyPlug(UErrorCode *status)
+{
+  UPlugData *plug = NULL;
+
+  if(U_FAILURE(*status)) {
+    return NULL;
+  }
+
+  if(pluginCount == UPLUG_PLUGIN_INITIAL_COUNT) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+
+  plug = &pluginList[pluginCount++];
+
+  plug->token = UPLUG_TOKEN;
+  plug->structSize = sizeof(UPlugData);
+  plug->name[0]=0;
+  plug->level = UPLUG_LEVEL_UNKNOWN; /* initialize to null state */
+  plug->awaitingLoad = TRUE;
+  plug->dontUnload = FALSE;
+  plug->pluginStatus = U_ZERO_ERROR;
+  plug->libName[0] = 0;
+  plug->config[0]=0;
+  plug->sym[0]=0;
+  plug->lib=NULL;
+  plug->entrypoint=NULL;
+
+
+  return plug;
+}
+
+static UPlugData *uplug_allocatePlug(UPlugEntrypoint *entrypoint, const char *config, void *lib, const char *symName,
+                                     UErrorCode *status) {
+  UPlugData *plug = uplug_allocateEmptyPlug(status);
+  if(U_FAILURE(*status)) {
+    return NULL;
+  }
+
+  if(config!=NULL) {
+    uprv_strncpy(plug->config, config, UPLUG_NAME_MAX);
+  } else {
+    plug->config[0] = 0;
+  }
+    
+  if(symName!=NULL) {
+    uprv_strncpy(plug->sym, symName, UPLUG_NAME_MAX);
+  } else {
+    plug->sym[0] = 0;
+  }
+    
+  plug->entrypoint = entrypoint;
+  plug->lib = lib;
+  uplug_queryPlug(plug, status);
+    
+  return plug;
+}
+
+static void uplug_deallocatePlug(UPlugData *plug, UErrorCode *status) {
+  UErrorCode subStatus = U_ZERO_ERROR;
+  if(!plug->dontUnload) {
+#if U_ENABLE_DYLOAD
+    uplug_closeLibrary(plug->lib, &subStatus);
+#endif
+  }
+  plug->lib = NULL;
+  if(U_SUCCESS(*status) && U_FAILURE(subStatus)) {
+    *status = subStatus;
+  }
+  /* shift plugins up and decrement count. */
+  if(U_SUCCESS(*status)) {
+    /* all ok- remove. */
+    pluginCount = uplug_removeEntryAt(pluginList, pluginCount, sizeof(plug[0]), uplug_pluginNumber(plug));
+  } else {
+    /* not ok- leave as a message. */
+    plug->awaitingLoad=FALSE;
+    plug->entrypoint=0;
+    plug->dontUnload=TRUE;
+  }
+}
+
+static void uplug_doUnloadPlug(UPlugData *plugToRemove, UErrorCode *status) {
+  if(plugToRemove != NULL) {
+    uplug_unloadPlug(plugToRemove, status);
+    uplug_deallocatePlug(plugToRemove, status);
+  }
+}
+
+U_CAPI void U_EXPORT2
+uplug_removePlug(UPlugData *plug, UErrorCode *status)  {
+  UPlugData *cursor = NULL;
+  UPlugData *plugToRemove = NULL;
+  if(U_FAILURE(*status)) return;
+    
+  for(cursor=pluginList;cursor!=NULL;) {
+    if(cursor==plug) {
+      plugToRemove = plug;
+      cursor=NULL;
+    } else {
+      cursor = uplug_nextPlug(cursor);
+    }
+  }
+    
+  uplug_doUnloadPlug(plugToRemove, status);
+}
+
+
+
+
+U_CAPI void U_EXPORT2 
+uplug_setPlugNoUnload(UPlugData *data, UBool dontUnload)
+{
+  data->dontUnload = dontUnload;
+}
+
+
+U_CAPI void U_EXPORT2
+uplug_setPlugLevel(UPlugData *data, UPlugLevel level) {
+  data->level = level;
+}
+
+
+U_CAPI UPlugLevel U_EXPORT2
+uplug_getPlugLevel(UPlugData *data) {
+  return data->level;
+}
+
+
+U_CAPI void U_EXPORT2
+uplug_setPlugName(UPlugData *data, const char *name) {
+  uprv_strncpy(data->name, name, UPLUG_NAME_MAX);
+}
+
+
+U_CAPI const char * U_EXPORT2
+uplug_getPlugName(UPlugData *data) {
+  return data->name;
+}
+
+
+U_CAPI const char * U_EXPORT2
+uplug_getSymbolName(UPlugData *data) {
+  return data->sym;
+}
+
+U_CAPI const char * U_EXPORT2
+uplug_getLibraryName(UPlugData *data, UErrorCode *status) {
+  if(data->libName[0]) {
+    return data->libName;
+  } else {
+#if U_ENABLE_DYLOAD
+    return uplug_findLibrary(data->lib, status);
+#else
+    return NULL;
+#endif
+  }
+}
+
+U_CAPI void * U_EXPORT2
+uplug_getLibrary(UPlugData *data) {
+  return data->lib;
+}
+
+U_CAPI void * U_EXPORT2
+uplug_getContext(UPlugData *data) {
+  return data->context;
+}
+
+
+U_CAPI void U_EXPORT2
+uplug_setContext(UPlugData *data, void *context) {
+  data->context = context;
+}
+
+U_CAPI const char* U_EXPORT2
+uplug_getConfiguration(UPlugData *data) {
+  return data->config;
+}
+
+U_CAPI UPlugData* U_EXPORT2
+uplug_getPlugInternal(int32_t n) { 
+  if(n <0 || n >= pluginCount) {
+    return NULL;
+  } else { 
+    return &(pluginList[n]);
+  }
+}
+
+
+U_CAPI UErrorCode U_EXPORT2
+uplug_getPlugLoadStatus(UPlugData *plug) {
+  return plug->pluginStatus;
+}
+
+
+
+
+/**
+ * Initialize a plugin fron an entrypoint and library - but don't load it.
+ */
+static UPlugData* uplug_initPlugFromEntrypointAndLibrary(UPlugEntrypoint *entrypoint, const char *config, void *lib, const char *sym,
+                                                         UErrorCode *status) {
+  UPlugData *plug = NULL;
+
+  plug = uplug_allocatePlug(entrypoint, config, lib, sym, status);
+
+  if(U_SUCCESS(*status)) {
+    return plug;
+  } else {
+    uplug_deallocatePlug(plug, status);
+    return NULL;
+  }
+}
+
+U_CAPI UPlugData* U_EXPORT2
+uplug_loadPlugFromEntrypoint(UPlugEntrypoint *entrypoint, const char *config, UErrorCode *status) {
+  UPlugData* plug = uplug_initPlugFromEntrypointAndLibrary(entrypoint, config, NULL, NULL, status);
+  uplug_loadPlug(plug, status);
+  return plug;
+}
+
+#if U_ENABLE_DYLOAD
+
+static UPlugData* 
+uplug_initErrorPlug(const char *libName, const char *sym, const char *config, const char *nameOrError, UErrorCode loadStatus, UErrorCode *status)
+{
+  UPlugData *plug = uplug_allocateEmptyPlug(status);
+  if(U_FAILURE(*status)) return NULL;
+
+  plug->pluginStatus = loadStatus;
+  plug->awaitingLoad = FALSE; /* Won't load. */
+  plug->dontUnload = TRUE; /* cannot unload. */
+
+  if(sym!=NULL) {
+    uprv_strncpy(plug->sym, sym, UPLUG_NAME_MAX);
+  }
+
+  if(libName!=NULL) {
+    uprv_strncpy(plug->libName, libName, UPLUG_NAME_MAX);
+  }
+
+  if(nameOrError!=NULL) {
+    uprv_strncpy(plug->name, nameOrError, UPLUG_NAME_MAX);
+  }
+
+  if(config!=NULL) {
+    uprv_strncpy(plug->config, config, UPLUG_NAME_MAX);
+  }
+
+  return plug;
+}
+
+/**
+ * Fetch a plugin from DLL, and then initialize it from a library- but don't load it.
+ */
+static UPlugData* 
+uplug_initPlugFromLibrary(const char *libName, const char *sym, const char *config, UErrorCode *status) {
+  void *lib = NULL;
+  UPlugData *plug = NULL;
+  if(U_FAILURE(*status)) { return NULL; }
+  lib = uplug_openLibrary(libName, status);
+  if(lib!=NULL && U_SUCCESS(*status)) {
+    UPlugEntrypoint *entrypoint = NULL;
+    entrypoint = (UPlugEntrypoint*)uprv_dlsym_func(lib, sym, status);
+
+    if(entrypoint!=NULL&&U_SUCCESS(*status)) {
+      plug = uplug_initPlugFromEntrypointAndLibrary(entrypoint, config, lib, sym, status);
+      if(plug!=NULL&&U_SUCCESS(*status)) {
+        plug->lib = lib; /* plug takes ownership of library */
+        lib = NULL; /* library is now owned by plugin. */
+      }
+    } else {
+      UErrorCode subStatus = U_ZERO_ERROR;
+      plug = uplug_initErrorPlug(libName,sym,config,"ERROR: Could not load entrypoint",(lib==NULL)?U_MISSING_RESOURCE_ERROR:*status,&subStatus);
+    }
+    if(lib!=NULL) { /* still need to close the lib */
+      UErrorCode subStatus = U_ZERO_ERROR;
+      uplug_closeLibrary(lib, &subStatus); /* don't care here */
+    }
+  } else {
+    UErrorCode subStatus = U_ZERO_ERROR;
+    plug = uplug_initErrorPlug(libName,sym,config,"ERROR: could not load library",(lib==NULL)?U_MISSING_RESOURCE_ERROR:*status,&subStatus);
+  }
+  return plug;
+}
+
+U_CAPI UPlugData* U_EXPORT2
+uplug_loadPlugFromLibrary(const char *libName, const char *sym, const char *config, UErrorCode *status) { 
+  UPlugData *plug = NULL;
+  if(U_FAILURE(*status)) { return NULL; }
+  plug = uplug_initPlugFromLibrary(libName, sym, config, status);
+  uplug_loadPlug(plug, status);
+
+  return plug;
+}
+
+#endif
+
+static UPlugLevel gCurrentLevel = UPLUG_LEVEL_LOW;
+
+U_CAPI UPlugLevel U_EXPORT2 uplug_getCurrentLevel() {
+  return gCurrentLevel;
+}
+
+static UBool U_CALLCONV uplug_cleanup(void)
+{
+  int32_t i;
+    
+  UPlugData *pluginToRemove;
+  /* cleanup plugs */
+  for(i=0;i<pluginCount;i++) {
+    UErrorCode subStatus = U_ZERO_ERROR;
+    pluginToRemove = &pluginList[i];
+    /* unload and deallocate */
+    uplug_doUnloadPlug(pluginToRemove, &subStatus);
+  }
+  /* close other held libs? */
+  gCurrentLevel = UPLUG_LEVEL_LOW;
+  return TRUE;
+}
+
+#if U_ENABLE_DYLOAD
+
+static void uplug_loadWaitingPlugs(UErrorCode *status) {
+  int32_t i;
+  UPlugLevel currentLevel = uplug_getCurrentLevel();
+    
+  if(U_FAILURE(*status)) {
+    return;
+  }
+#if UPLUG_TRACE
+  DBG((stderr,  "uplug_loadWaitingPlugs() Level: %d\n", currentLevel));
+#endif
+  /* pass #1: low level plugs */
+  for(i=0;i<pluginCount;i++) {
+    UErrorCode subStatus = U_ZERO_ERROR;
+    UPlugData *pluginToLoad = &pluginList[i];
+    if(pluginToLoad->awaitingLoad) {
+      if(pluginToLoad->level == UPLUG_LEVEL_LOW) {
+        if(currentLevel > UPLUG_LEVEL_LOW) {
+          pluginToLoad->pluginStatus = U_PLUGIN_TOO_HIGH;
+        } else {
+          UPlugLevel newLevel;
+          uplug_loadPlug(pluginToLoad, &subStatus);
+          newLevel = uplug_getCurrentLevel();
+          if(newLevel > currentLevel) {
+            pluginToLoad->pluginStatus = U_PLUGIN_CHANGED_LEVEL_WARNING;
+            currentLevel = newLevel;
+          }
+        }
+        pluginToLoad->awaitingLoad = FALSE;
+      } 
+    }
+  }    
+  for(i=0;i<pluginCount;i++) {
+    UErrorCode subStatus = U_ZERO_ERROR;
+    UPlugData *pluginToLoad = &pluginList[i];
+        
+    if(pluginToLoad->awaitingLoad) {
+      if(pluginToLoad->level == UPLUG_LEVEL_INVALID) { 
+        pluginToLoad->pluginStatus = U_PLUGIN_DIDNT_SET_LEVEL;
+      } else if(pluginToLoad->level == UPLUG_LEVEL_UNKNOWN) {
+        pluginToLoad->pluginStatus = U_INTERNAL_PROGRAM_ERROR;
+      } else {
+        uplug_loadPlug(pluginToLoad, &subStatus);
+      }
+      pluginToLoad->awaitingLoad = FALSE;
+    }
+  }
+    
+#if UPLUG_TRACE
+  DBG((stderr,  " Done Loading Plugs. Level: %d\n", (int32_t)uplug_getCurrentLevel()));
+#endif
+}
+
+/* Name of the plugin config file */
+static char plugin_file[2048] = "";
+#endif
+
+U_CAPI const char* U_EXPORT2
+uplug_getPluginFile() {
+#if U_ENABLE_DYLOAD && !UCONFIG_NO_FILE_IO
+  return plugin_file;
+#else
+  return NULL;
+#endif
+}
+
+
+//  uplug_init()  is called first thing from u_init().
+
+U_CAPI void U_EXPORT2
+uplug_init(UErrorCode *status) {
+#if !U_ENABLE_DYLOAD
+  (void)status; /* unused */
+#elif !UCONFIG_NO_FILE_IO
+  CharString plugin_dir;
+  const char *env = getenv("ICU_PLUGINS");
+
+  if(U_FAILURE(*status)) return;
+  if(env != NULL) {
+    plugin_dir.append(env, -1, *status);
+  }
+  if(U_FAILURE(*status)) return;
+
+#if defined(DEFAULT_ICU_PLUGINS) 
+  if(plugin_dir.isEmpty()) {
+    plugin_dir.append(DEFAULT_ICU_PLUGINS, -1, *status);
+  }
+#endif
+
+#if UPLUG_TRACE
+  DBG((stderr, "ICU_PLUGINS=%s\n", plugin_dir.data()));
+#endif
+
+  if(!plugin_dir.isEmpty()) {
+    FILE *f;
+        
+    CharString pluginFile;
+#ifdef OS390BATCH
+/* There are potentially a lot of ways to implement a plugin directory on OS390/zOS  */
+/* Keeping in mind that unauthorized file access is logged, monitored, and enforced  */
+/* I've chosen to open a DDNAME if BATCH and leave it alone for (presumably) UNIX    */
+/* System Services.  Alternative techniques might be allocating a member in          */
+/* SYS1.PARMLIB or setting an environment variable "ICU_PLUGIN_PATH" (?).  The       */
+/* DDNAME can be connected to a file in the HFS if need be.                          */
+
+    pluginFile.append("//DD:ICUPLUG", -1, *status);        /* JAM 20 Oct 2011 */
+#else
+    pluginFile.append(plugin_dir, *status);
+    pluginFile.append(U_FILE_SEP_STRING, -1, *status);
+    pluginFile.append("icuplugins", -1, *status);
+    pluginFile.append(U_ICU_VERSION_SHORT, -1, *status);
+    pluginFile.append(".txt", -1, *status);
+#endif
+
+#if UPLUG_TRACE
+    DBG((stderr, "status=%s\n", u_errorName(*status)));
+#endif
+
+    if(U_FAILURE(*status)) {
+      return;
+    }
+    if((size_t)pluginFile.length() > (sizeof(plugin_file)-1)) {
+      *status = U_BUFFER_OVERFLOW_ERROR;
+#if UPLUG_TRACE
+      DBG((stderr, "status=%s\n", u_errorName(*status)));
+#endif
+      return;
+    }
+    
+    /* plugin_file is not used for processing - it is only used 
+       so that uplug_getPluginFile() works (i.e. icuinfo)
+    */
+    pluginFile.extract(plugin_file, sizeof(plugin_file), *status);
+
+#if UPLUG_TRACE
+    DBG((stderr, "pluginfile= %s len %d/%d\n", plugin_file, (int)strlen(plugin_file), (int)sizeof(plugin_file)));
+#endif
+        
+#ifdef __MVS__
+    if (iscics()) /* 12 Nov 2011 JAM */
+    {
+        f = NULL;
+    }
+    else
+#endif
+    {
+        f = fopen(pluginFile.data(), "r");
+    }
+
+    if(f != NULL) {
+      char linebuf[1024];
+      char *p, *libName=NULL, *symName=NULL, *config=NULL;
+      int32_t line = 0;
+            
+            
+      while(fgets(linebuf,1023,f)) {
+        line++;
+
+        if(!*linebuf || *linebuf=='#') {
+          continue;
+        } else {
+          p = linebuf;
+          while(*p&&isspace((int)*p))
+            p++;
+          if(!*p || *p=='#') continue;
+          libName = p;
+          while(*p&&!isspace((int)*p)) {
+            p++;
+          }
+          if(!*p || *p=='#') continue; /* no tab after libname */
+          *p=0; /* end of libname */
+          p++;
+          while(*p&&isspace((int)*p)) {
+            p++;
+          }
+          if(!*p||*p=='#') continue; /* no symname after libname +tab */
+          symName = p;
+          while(*p&&!isspace((int)*p)) {
+            p++;
+          }
+                    
+          if(*p) { /* has config */
+            *p=0;
+            ++p;
+            while(*p&&isspace((int)*p)) {
+              p++;
+            }
+            if(*p) {
+              config = p;
+            }
+          }
+                    
+          /* chop whitespace at the end of the config */
+          if(config!=NULL&&*config!=0) {
+            p = config+strlen(config);
+            while(p>config&&isspace((int)*(--p))) {
+              *p=0;
+            }
+          }
+                
+          /* OK, we're good. */
+          { 
+            UErrorCode subStatus = U_ZERO_ERROR;
+            UPlugData *plug = uplug_initPlugFromLibrary(libName, symName, config, &subStatus);
+            if(U_FAILURE(subStatus) && U_SUCCESS(*status)) {
+              *status = subStatus;
+            }
+#if UPLUG_TRACE
+            DBG((stderr, "PLUGIN libName=[%s], sym=[%s], config=[%s]\n", libName, symName, config));
+            DBG((stderr, " -> %p, %s\n", (void*)plug, u_errorName(subStatus)));
+#else
+            (void)plug; /* unused */
+#endif
+          }
+        }
+      }
+      fclose(f);
+    } else {
+#if UPLUG_TRACE
+      DBG((stderr, "Can't open plugin file %s\n", plugin_file));
+#endif
+    }
+  }
+  uplug_loadWaitingPlugs(status);
+#endif /* U_ENABLE_DYLOAD */
+  gCurrentLevel = UPLUG_LEVEL_HIGH;
+  ucln_registerCleanup(UCLN_UPLUG, uplug_cleanup);
+}
+
+#endif
+
+
diff --git a/thirdparty/icu4c/common/icuplugimp.h b/thirdparty/icu4c/common/icuplugimp.h
new file mode 100644
index 0000000000..9df309204e
--- /dev/null
+++ b/thirdparty/icu4c/common/icuplugimp.h
@@ -0,0 +1,93 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2009-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : icuplugimp.h
+* 
+*  Internal functions for the ICU plugin system
+*
+*   Date         Name        Description
+*   10/29/2009   sl          New.
+******************************************************************************
+*/
+
+
+#ifndef ICUPLUGIMP_H
+#define ICUPLUGIMP_H
+
+#include "unicode/icuplug.h"
+
+#if UCONFIG_ENABLE_PLUGINS
+
+/*========================*/
+/** @{ Library Manipulation  
+ */
+
+/**
+ * Open a library, adding a reference count if needed.
+ * @param libName library name to load
+ * @param status error code
+ * @return the library pointer, or NULL
+ * @internal internal use only
+ */
+U_CAPI void * U_EXPORT2
+uplug_openLibrary(const char *libName, UErrorCode *status);
+
+/**
+ * Close a library, if its reference count is 0
+ * @param lib the library to close
+ * @param status error code
+ * @internal internal use only
+ */
+U_CAPI void U_EXPORT2
+uplug_closeLibrary(void *lib, UErrorCode *status);
+
+/**
+ * Get a library's name, or NULL if not found.
+ * @param lib the library's name
+ * @param status error code
+ * @return the library name, or NULL if not found.
+ * @internal internal use only
+ */
+U_CAPI  char * U_EXPORT2
+uplug_findLibrary(void *lib, UErrorCode *status);
+
+/** @} */
+
+/*========================*/
+/** {@ ICU Plugin internal interfaces
+ */
+
+/**
+ * Initialize the plugins 
+ * @param status error result
+ * @internal - Internal use only.
+ */
+U_CAPI void U_EXPORT2
+uplug_init(UErrorCode *status);
+
+/**
+ * Get raw plug N
+ * @internal - Internal use only
+ */ 
+U_CAPI UPlugData* U_EXPORT2
+uplug_getPlugInternal(int32_t n);
+
+/**
+ * Get the name of the plugin file. 
+ * @internal - Internal use only.
+ */
+U_CAPI const char* U_EXPORT2
+uplug_getPluginFile(void);
+
+/** @} */
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/loadednormalizer2impl.cpp b/thirdparty/icu4c/common/loadednormalizer2impl.cpp
new file mode 100644
index 0000000000..e4b36f1055
--- /dev/null
+++ b/thirdparty/icu4c/common/loadednormalizer2impl.cpp
@@ -0,0 +1,418 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* loadednormalizer2impl.cpp
+*
+* created on: 2014sep03
+* created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/udata.h"
+#include "unicode/localpointer.h"
+#include "unicode/normalizer2.h"
+#include "unicode/ucptrie.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "norm2allmodes.h"
+#include "normalizer2impl.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "uhash.h"
+
+U_NAMESPACE_BEGIN
+
+class LoadedNormalizer2Impl : public Normalizer2Impl {
+public:
+    LoadedNormalizer2Impl() : memory(NULL), ownedTrie(NULL) {}
+    virtual ~LoadedNormalizer2Impl();
+
+    void load(const char *packageName, const char *name, UErrorCode &errorCode);
+
+private:
+    static UBool U_CALLCONV
+    isAcceptable(void *context, const char *type, const char *name, const UDataInfo *pInfo);
+
+    UDataMemory *memory;
+    UCPTrie *ownedTrie;
+};
+
+LoadedNormalizer2Impl::~LoadedNormalizer2Impl() {
+    udata_close(memory);
+    ucptrie_close(ownedTrie);
+}
+
+UBool U_CALLCONV
+LoadedNormalizer2Impl::isAcceptable(void * /*context*/,
+                                    const char * /* type */, const char * /*name*/,
+                                    const UDataInfo *pInfo) {
+    if(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->dataFormat[0]==0x4e &&    /* dataFormat="Nrm2" */
+        pInfo->dataFormat[1]==0x72 &&
+        pInfo->dataFormat[2]==0x6d &&
+        pInfo->dataFormat[3]==0x32 &&
+        pInfo->formatVersion[0]==4
+    ) {
+        // Normalizer2Impl *me=(Normalizer2Impl *)context;
+        // uprv_memcpy(me->dataVersion, pInfo->dataVersion, 4);
+        return TRUE;
+    } else {
+        return FALSE;
+    }
+}
+
+void
+LoadedNormalizer2Impl::load(const char *packageName, const char *name, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    memory=udata_openChoice(packageName, "nrm", name, isAcceptable, this, &errorCode);
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    const uint8_t *inBytes=(const uint8_t *)udata_getMemory(memory);
+    const int32_t *inIndexes=(const int32_t *)inBytes;
+    int32_t indexesLength=inIndexes[IX_NORM_TRIE_OFFSET]/4;
+    if(indexesLength<=IX_MIN_LCCC_CP) {
+        errorCode=U_INVALID_FORMAT_ERROR;  // Not enough indexes.
+        return;
+    }
+
+    int32_t offset=inIndexes[IX_NORM_TRIE_OFFSET];
+    int32_t nextOffset=inIndexes[IX_EXTRA_DATA_OFFSET];
+    ownedTrie=ucptrie_openFromBinary(UCPTRIE_TYPE_FAST, UCPTRIE_VALUE_BITS_16,
+                                     inBytes+offset, nextOffset-offset, NULL,
+                                     &errorCode);
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+
+    offset=nextOffset;
+    nextOffset=inIndexes[IX_SMALL_FCD_OFFSET];
+    const uint16_t *inExtraData=(const uint16_t *)(inBytes+offset);
+
+    // smallFCD: new in formatVersion 2
+    offset=nextOffset;
+    const uint8_t *inSmallFCD=inBytes+offset;
+
+    init(inIndexes, ownedTrie, inExtraData, inSmallFCD);
+}
+
+// instance cache ---------------------------------------------------------- ***
+
+Norm2AllModes *
+Norm2AllModes::createInstance(const char *packageName,
+                              const char *name,
+                              UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    LoadedNormalizer2Impl *impl=new LoadedNormalizer2Impl;
+    if(impl==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    impl->load(packageName, name, errorCode);
+    return createInstance(impl, errorCode);
+}
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV uprv_loaded_normalizer2_cleanup();
+U_CDECL_END
+
+#if !NORM2_HARDCODE_NFC_DATA
+static Norm2AllModes *nfcSingleton;
+static icu::UInitOnce nfcInitOnce = U_INITONCE_INITIALIZER;
+#endif
+
+static Norm2AllModes *nfkcSingleton;
+static icu::UInitOnce nfkcInitOnce = U_INITONCE_INITIALIZER;
+
+static Norm2AllModes *nfkc_cfSingleton;
+static icu::UInitOnce nfkc_cfInitOnce = U_INITONCE_INITIALIZER;
+
+static UHashtable    *cache=NULL;
+
+// UInitOnce singleton initialization function
+static void U_CALLCONV initSingletons(const char *what, UErrorCode &errorCode) {
+#if !NORM2_HARDCODE_NFC_DATA
+    if (uprv_strcmp(what, "nfc") == 0) {
+        nfcSingleton    = Norm2AllModes::createInstance(NULL, "nfc", errorCode);
+    } else
+#endif
+    if (uprv_strcmp(what, "nfkc") == 0) {
+        nfkcSingleton    = Norm2AllModes::createInstance(NULL, "nfkc", errorCode);
+    } else if (uprv_strcmp(what, "nfkc_cf") == 0) {
+        nfkc_cfSingleton = Norm2AllModes::createInstance(NULL, "nfkc_cf", errorCode);
+    } else {
+        UPRV_UNREACHABLE;   // Unknown singleton
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_LOADED_NORMALIZER2, uprv_loaded_normalizer2_cleanup);
+}
+
+U_CDECL_BEGIN
+
+static void U_CALLCONV deleteNorm2AllModes(void *allModes) {
+    delete (Norm2AllModes *)allModes;
+}
+
+static UBool U_CALLCONV uprv_loaded_normalizer2_cleanup() {
+#if !NORM2_HARDCODE_NFC_DATA
+    delete nfcSingleton;
+    nfcSingleton = NULL;
+    nfcInitOnce.reset();
+#endif
+
+    delete nfkcSingleton;
+    nfkcSingleton = NULL;
+    nfkcInitOnce.reset();
+
+    delete nfkc_cfSingleton;
+    nfkc_cfSingleton = NULL;
+    nfkc_cfInitOnce.reset();
+
+    uhash_close(cache);
+    cache=NULL;
+    return TRUE;
+}
+
+U_CDECL_END
+
+#if !NORM2_HARDCODE_NFC_DATA
+const Norm2AllModes *
+Norm2AllModes::getNFCInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return NULL; }
+    umtx_initOnce(nfcInitOnce, &initSingletons, "nfc", errorCode);
+    return nfcSingleton;
+}
+#endif
+
+const Norm2AllModes *
+Norm2AllModes::getNFKCInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return NULL; }
+    umtx_initOnce(nfkcInitOnce, &initSingletons, "nfkc", errorCode);
+    return nfkcSingleton;
+}
+
+const Norm2AllModes *
+Norm2AllModes::getNFKC_CFInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return NULL; }
+    umtx_initOnce(nfkc_cfInitOnce, &initSingletons, "nfkc_cf", errorCode);
+    return nfkc_cfSingleton;
+}
+
+#if !NORM2_HARDCODE_NFC_DATA
+const Normalizer2 *
+Normalizer2::getNFCInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->comp : NULL;
+}
+
+const Normalizer2 *
+Normalizer2::getNFDInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->decomp : NULL;
+}
+
+const Normalizer2 *Normalizer2Factory::getFCDInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->fcd : NULL;
+}
+
+const Normalizer2 *Normalizer2Factory::getFCCInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->fcc : NULL;
+}
+
+const Normalizer2Impl *
+Normalizer2Factory::getNFCImpl(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? allModes->impl : NULL;
+}
+#endif
+
+const Normalizer2 *
+Normalizer2::getNFKCInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFKCInstance(errorCode);
+    return allModes!=NULL ? &allModes->comp : NULL;
+}
+
+const Normalizer2 *
+Normalizer2::getNFKDInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFKCInstance(errorCode);
+    return allModes!=NULL ? &allModes->decomp : NULL;
+}
+
+const Normalizer2 *
+Normalizer2::getNFKCCasefoldInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFKC_CFInstance(errorCode);
+    return allModes!=NULL ? &allModes->comp : NULL;
+}
+
+const Normalizer2 *
+Normalizer2::getInstance(const char *packageName,
+                         const char *name,
+                         UNormalization2Mode mode,
+                         UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    if(name==NULL || *name==0) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    const Norm2AllModes *allModes=NULL;
+    if(packageName==NULL) {
+        if(0==uprv_strcmp(name, "nfc")) {
+            allModes=Norm2AllModes::getNFCInstance(errorCode);
+        } else if(0==uprv_strcmp(name, "nfkc")) {
+            allModes=Norm2AllModes::getNFKCInstance(errorCode);
+        } else if(0==uprv_strcmp(name, "nfkc_cf")) {
+            allModes=Norm2AllModes::getNFKC_CFInstance(errorCode);
+        }
+    }
+    if(allModes==NULL && U_SUCCESS(errorCode)) {
+        {
+            Mutex lock;
+            if(cache!=NULL) {
+                allModes=(Norm2AllModes *)uhash_get(cache, name);
+            }
+        }
+        if(allModes==NULL) {
+            ucln_common_registerCleanup(UCLN_COMMON_LOADED_NORMALIZER2, uprv_loaded_normalizer2_cleanup);
+            LocalPointer<Norm2AllModes> localAllModes(
+                Norm2AllModes::createInstance(packageName, name, errorCode));
+            if(U_SUCCESS(errorCode)) {
+                Mutex lock;
+                if(cache==NULL) {
+                    cache=uhash_open(uhash_hashChars, uhash_compareChars, NULL, &errorCode);
+                    if(U_FAILURE(errorCode)) {
+                        return NULL;
+                    }
+                    uhash_setKeyDeleter(cache, uprv_free);
+                    uhash_setValueDeleter(cache, deleteNorm2AllModes);
+                }
+                void *temp=uhash_get(cache, name);
+                if(temp==NULL) {
+                    int32_t keyLength= static_cast<int32_t>(uprv_strlen(name)+1);
+                    char *nameCopy=(char *)uprv_malloc(keyLength);
+                    if(nameCopy==NULL) {
+                        errorCode=U_MEMORY_ALLOCATION_ERROR;
+                        return NULL;
+                    }
+                    uprv_memcpy(nameCopy, name, keyLength);
+                    allModes=localAllModes.getAlias();
+                    uhash_put(cache, nameCopy, localAllModes.orphan(), &errorCode);
+                } else {
+                    // race condition
+                    allModes=(Norm2AllModes *)temp;
+                }
+            }
+        }
+    }
+    if(allModes!=NULL && U_SUCCESS(errorCode)) {
+        switch(mode) {
+        case UNORM2_COMPOSE:
+            return &allModes->comp;
+        case UNORM2_DECOMPOSE:
+            return &allModes->decomp;
+        case UNORM2_FCD:
+            return &allModes->fcd;
+        case UNORM2_COMPOSE_CONTIGUOUS:
+            return &allModes->fcc;
+        default:
+            break;  // do nothing
+        }
+    }
+    return NULL;
+}
+
+const Normalizer2 *
+Normalizer2Factory::getInstance(UNormalizationMode mode, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    switch(mode) {
+    case UNORM_NFD:
+        return Normalizer2::getNFDInstance(errorCode);
+    case UNORM_NFKD:
+        return Normalizer2::getNFKDInstance(errorCode);
+    case UNORM_NFC:
+        return Normalizer2::getNFCInstance(errorCode);
+    case UNORM_NFKC:
+        return Normalizer2::getNFKCInstance(errorCode);
+    case UNORM_FCD:
+        return getFCDInstance(errorCode);
+    default:  // UNORM_NONE
+        return getNoopInstance(errorCode);
+    }
+}
+
+const Normalizer2Impl *
+Normalizer2Factory::getNFKCImpl(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFKCInstance(errorCode);
+    return allModes!=NULL ? allModes->impl : NULL;
+}
+
+const Normalizer2Impl *
+Normalizer2Factory::getNFKC_CFImpl(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFKC_CFInstance(errorCode);
+    return allModes!=NULL ? allModes->impl : NULL;
+}
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKCInstance(UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getNFKCInstance(*pErrorCode);
+}
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKDInstance(UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getNFKDInstance(*pErrorCode);
+}
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKCCasefoldInstance(UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getNFKCCasefoldInstance(*pErrorCode);
+}
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getInstance(const char *packageName,
+                   const char *name,
+                   UNormalization2Mode mode,
+                   UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getInstance(packageName, name, mode, *pErrorCode);
+}
+
+U_CFUNC UNormalizationCheckResult
+unorm_getQuickCheck(UChar32 c, UNormalizationMode mode) {
+    if(mode<=UNORM_NONE || UNORM_FCD<=mode) {
+        return UNORM_YES;
+    }
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2 *norm2=Normalizer2Factory::getInstance(mode, errorCode);
+    if(U_SUCCESS(errorCode)) {
+        return ((const Normalizer2WithImpl *)norm2)->getQuickCheck(c);
+    } else {
+        return UNORM_MAYBE;
+    }
+}
+
+#endif  // !UCONFIG_NO_NORMALIZATION
diff --git a/thirdparty/icu4c/common/localebuilder.cpp b/thirdparty/icu4c/common/localebuilder.cpp
new file mode 100644
index 0000000000..1dd8131e58
--- /dev/null
+++ b/thirdparty/icu4c/common/localebuilder.cpp
@@ -0,0 +1,468 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+#include <utility>
+
+#include "bytesinkutil.h"  // CharStringByteSink
+#include "charstr.h"
+#include "cstring.h"
+#include "ulocimp.h"
+#include "unicode/localebuilder.h"
+#include "unicode/locid.h"
+
+U_NAMESPACE_BEGIN
+
+#define UPRV_ISDIGIT(c) (((c) >= '0') && ((c) <= '9'))
+#define UPRV_ISALPHANUM(c) (uprv_isASCIILetter(c) || UPRV_ISDIGIT(c) )
+
+const char* kAttributeKey = "attribute";
+
+static bool _isExtensionSubtags(char key, const char* s, int32_t len) {
+    switch (uprv_tolower(key)) {
+        case 'u':
+            return ultag_isUnicodeExtensionSubtags(s, len);
+        case 't':
+            return ultag_isTransformedExtensionSubtags(s, len);
+        case 'x':
+            return ultag_isPrivateuseValueSubtags(s, len);
+        default:
+            return ultag_isExtensionSubtags(s, len);
+    }
+}
+
+LocaleBuilder::LocaleBuilder() : UObject(), status_(U_ZERO_ERROR), language_(),
+    script_(), region_(), variant_(nullptr), extensions_(nullptr)
+{
+    language_[0] = 0;
+    script_[0] = 0;
+    region_[0] = 0;
+}
+
+LocaleBuilder::~LocaleBuilder()
+{
+    delete variant_;
+    delete extensions_;
+}
+
+LocaleBuilder& LocaleBuilder::setLocale(const Locale& locale)
+{
+    clear();
+    setLanguage(locale.getLanguage());
+    setScript(locale.getScript());
+    setRegion(locale.getCountry());
+    setVariant(locale.getVariant());
+    extensions_ = locale.clone();
+    if (extensions_ == nullptr) {
+        status_ = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::setLanguageTag(StringPiece tag)
+{
+    Locale l = Locale::forLanguageTag(tag, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    // Because setLocale will reset status_ we need to return
+    // first if we have error in forLanguageTag.
+    setLocale(l);
+    return *this;
+}
+
+static void setField(StringPiece input, char* dest, UErrorCode& errorCode,
+                     UBool (*test)(const char*, int32_t)) {
+    if (U_FAILURE(errorCode)) { return; }
+    if (input.empty()) {
+        dest[0] = '\0';
+    } else if (test(input.data(), input.length())) {
+        uprv_memcpy(dest, input.data(), input.length());
+        dest[input.length()] = '\0';
+    } else {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+LocaleBuilder& LocaleBuilder::setLanguage(StringPiece language)
+{
+    setField(language, language_, status_, &ultag_isLanguageSubtag);
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::setScript(StringPiece script)
+{
+    setField(script, script_, status_, &ultag_isScriptSubtag);
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::setRegion(StringPiece region)
+{
+    setField(region, region_, status_, &ultag_isRegionSubtag);
+    return *this;
+}
+
+static void transform(char* data, int32_t len) {
+    for (int32_t i = 0; i < len; i++, data++) {
+        if (*data == '_') {
+            *data = '-';
+        } else {
+            *data = uprv_tolower(*data);
+        }
+    }
+}
+
+LocaleBuilder& LocaleBuilder::setVariant(StringPiece variant)
+{
+    if (U_FAILURE(status_)) { return *this; }
+    if (variant.empty()) {
+        delete variant_;
+        variant_ = nullptr;
+        return *this;
+    }
+    CharString* new_variant = new CharString(variant, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    if (new_variant == nullptr) {
+        status_ = U_MEMORY_ALLOCATION_ERROR;
+        return *this;
+    }
+    transform(new_variant->data(), new_variant->length());
+    if (!ultag_isVariantSubtags(new_variant->data(), new_variant->length())) {
+        delete new_variant;
+        status_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    delete variant_;
+    variant_ = new_variant;
+    return *this;
+}
+
+static bool
+_isKeywordValue(const char* key, const char* value, int32_t value_len)
+{
+    if (key[1] == '\0') {
+        // one char key
+        return (UPRV_ISALPHANUM(uprv_tolower(key[0])) &&
+                _isExtensionSubtags(key[0], value, value_len));
+    } else if (uprv_strcmp(key, kAttributeKey) == 0) {
+        // unicode attributes
+        return ultag_isUnicodeLocaleAttributes(value, value_len);
+    }
+    // otherwise: unicode extension value
+    // We need to convert from legacy key/value to unicode
+    // key/value
+    const char* unicode_locale_key = uloc_toUnicodeLocaleKey(key);
+    const char* unicode_locale_type = uloc_toUnicodeLocaleType(key, value);
+
+    return unicode_locale_key && unicode_locale_type &&
+           ultag_isUnicodeLocaleKey(unicode_locale_key, -1) &&
+           ultag_isUnicodeLocaleType(unicode_locale_type, -1);
+}
+
+static void
+_copyExtensions(const Locale& from, icu::StringEnumeration *keywords,
+                Locale& to, bool validate, UErrorCode& errorCode)
+{
+    if (U_FAILURE(errorCode)) { return; }
+    LocalPointer<icu::StringEnumeration> ownedKeywords;
+    if (keywords == nullptr) {
+        ownedKeywords.adoptInstead(from.createKeywords(errorCode));
+        if (U_FAILURE(errorCode) || ownedKeywords.isNull()) { return; }
+        keywords = ownedKeywords.getAlias();
+    }
+    const char* key;
+    while ((key = keywords->next(nullptr, errorCode)) != nullptr) {
+        CharString value;
+        CharStringByteSink sink(&value);
+        from.getKeywordValue(key, sink, errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        if (uprv_strcmp(key, kAttributeKey) == 0) {
+            transform(value.data(), value.length());
+        }
+        if (validate &&
+            !_isKeywordValue(key, value.data(), value.length())) {
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        to.setKeywordValue(key, value.data(), errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+    }
+}
+
+void static
+_clearUAttributesAndKeyType(Locale& locale, UErrorCode& errorCode)
+{
+    // Clear Unicode attributes
+    locale.setKeywordValue(kAttributeKey, "", errorCode);
+
+    // Clear all Unicode keyword values
+    LocalPointer<icu::StringEnumeration> iter(locale.createUnicodeKeywords(errorCode));
+    if (U_FAILURE(errorCode) || iter.isNull()) { return; }
+    const char* key;
+    while ((key = iter->next(nullptr, errorCode)) != nullptr) {
+        locale.setUnicodeKeywordValue(key, nullptr, errorCode);
+    }
+}
+
+static void
+_setUnicodeExtensions(Locale& locale, const CharString& value, UErrorCode& errorCode)
+{
+    // Add the unicode extensions to extensions_
+    CharString locale_str("und-u-", errorCode);
+    locale_str.append(value, errorCode);
+    _copyExtensions(
+        Locale::forLanguageTag(locale_str.data(), errorCode), nullptr,
+        locale, false, errorCode);
+}
+
+LocaleBuilder& LocaleBuilder::setExtension(char key, StringPiece value)
+{
+    if (U_FAILURE(status_)) { return *this; }
+    if (!UPRV_ISALPHANUM(key)) {
+        status_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    CharString value_str(value, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    transform(value_str.data(), value_str.length());
+    if (!value_str.isEmpty() &&
+            !_isExtensionSubtags(key, value_str.data(), value_str.length())) {
+        status_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    if (extensions_ == nullptr) {
+        extensions_ = new Locale();
+        if (extensions_ == nullptr) {
+            status_ = U_MEMORY_ALLOCATION_ERROR;
+            return *this;
+        }
+    }
+    if (uprv_tolower(key) != 'u') {
+        // for t, x and others extension.
+        extensions_->setKeywordValue(StringPiece(&key, 1), value_str.data(),
+                                     status_);
+        return *this;
+    }
+    _clearUAttributesAndKeyType(*extensions_, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    if (!value.empty()) {
+        _setUnicodeExtensions(*extensions_, value_str, status_);
+    }
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::setUnicodeLocaleKeyword(
+      StringPiece key, StringPiece type)
+{
+    if (U_FAILURE(status_)) { return *this; }
+    if (!ultag_isUnicodeLocaleKey(key.data(), key.length()) ||
+            (!type.empty() &&
+                 !ultag_isUnicodeLocaleType(type.data(), type.length()))) {
+      status_ = U_ILLEGAL_ARGUMENT_ERROR;
+      return *this;
+    }
+    if (extensions_ == nullptr) {
+        extensions_ = new Locale();
+    }
+    if (extensions_ == nullptr) {
+        status_ = U_MEMORY_ALLOCATION_ERROR;
+        return *this;
+    }
+    extensions_->setUnicodeKeywordValue(key, type, status_);
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::addUnicodeLocaleAttribute(
+    StringPiece value)
+{
+    CharString value_str(value, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    transform(value_str.data(), value_str.length());
+    if (!ultag_isUnicodeLocaleAttribute(value_str.data(), value_str.length())) {
+        status_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    if (extensions_ == nullptr) {
+        extensions_ = new Locale();
+        if (extensions_ == nullptr) {
+            status_ = U_MEMORY_ALLOCATION_ERROR;
+            return *this;
+        }
+        extensions_->setKeywordValue(kAttributeKey, value_str.data(), status_);
+        return *this;
+    }
+
+    CharString attributes;
+    CharStringByteSink sink(&attributes);
+    UErrorCode localErrorCode = U_ZERO_ERROR;
+    extensions_->getKeywordValue(kAttributeKey, sink, localErrorCode);
+    if (U_FAILURE(localErrorCode)) {
+        CharString new_attributes(value_str.data(), status_);
+        // No attributes, set the attribute.
+        extensions_->setKeywordValue(kAttributeKey, new_attributes.data(), status_);
+        return *this;
+    }
+
+    transform(attributes.data(),attributes.length());
+    const char* start = attributes.data();
+    const char* limit = attributes.data() + attributes.length();
+    CharString new_attributes;
+    bool inserted = false;
+    while (start < limit) {
+        if (!inserted) {
+            int cmp = uprv_strcmp(start, value_str.data());
+            if (cmp == 0) { return *this; }  // Found it in attributes: Just return
+            if (cmp > 0) {
+                if (!new_attributes.isEmpty()) new_attributes.append('_', status_);
+                new_attributes.append(value_str.data(), status_);
+                inserted = true;
+            }
+        }
+        if (!new_attributes.isEmpty()) {
+            new_attributes.append('_', status_);
+        }
+        new_attributes.append(start, status_);
+        start += uprv_strlen(start) + 1;
+    }
+    if (!inserted) {
+        if (!new_attributes.isEmpty()) {
+            new_attributes.append('_', status_);
+        }
+        new_attributes.append(value_str.data(), status_);
+    }
+    // Not yet in the attributes, set the attribute.
+    extensions_->setKeywordValue(kAttributeKey, new_attributes.data(), status_);
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::removeUnicodeLocaleAttribute(
+    StringPiece value)
+{
+    CharString value_str(value, status_);
+    if (U_FAILURE(status_)) { return *this; }
+    transform(value_str.data(), value_str.length());
+    if (!ultag_isUnicodeLocaleAttribute(value_str.data(), value_str.length())) {
+        status_ = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    if (extensions_ == nullptr) { return *this; }
+    UErrorCode localErrorCode = U_ZERO_ERROR;
+    CharString attributes;
+    CharStringByteSink sink(&attributes);
+    extensions_->getKeywordValue(kAttributeKey, sink, localErrorCode);
+    // get failure, just return
+    if (U_FAILURE(localErrorCode)) { return *this; }
+    // Do not have any attributes, just return.
+    if (attributes.isEmpty()) { return *this; }
+
+    char* p = attributes.data();
+    // Replace null terminiator in place for _ and - so later
+    // we can use uprv_strcmp to compare.
+    for (int32_t i = 0; i < attributes.length(); i++, p++) {
+        *p = (*p == '_' || *p == '-') ? '\0' : uprv_tolower(*p);
+    }
+
+    const char* start = attributes.data();
+    const char* limit = attributes.data() + attributes.length();
+    CharString new_attributes;
+    bool found = false;
+    while (start < limit) {
+        if (uprv_strcmp(start, value_str.data()) == 0) {
+            found = true;
+        } else {
+            if (!new_attributes.isEmpty()) {
+                new_attributes.append('_', status_);
+            }
+            new_attributes.append(start, status_);
+        }
+        start += uprv_strlen(start) + 1;
+    }
+    // Found the value in attributes, set the attribute.
+    if (found) {
+        extensions_->setKeywordValue(kAttributeKey, new_attributes.data(), status_);
+    }
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::clear()
+{
+    status_ = U_ZERO_ERROR;
+    language_[0] = 0;
+    script_[0] = 0;
+    region_[0] = 0;
+    delete variant_;
+    variant_ = nullptr;
+    clearExtensions();
+    return *this;
+}
+
+LocaleBuilder& LocaleBuilder::clearExtensions()
+{
+    delete extensions_;
+    extensions_ = nullptr;
+    return *this;
+}
+
+Locale makeBogusLocale() {
+  Locale bogus;
+  bogus.setToBogus();
+  return bogus;
+}
+
+void LocaleBuilder::copyExtensionsFrom(const Locale& src, UErrorCode& errorCode)
+{
+    if (U_FAILURE(errorCode)) { return; }
+    LocalPointer<icu::StringEnumeration> keywords(src.createKeywords(errorCode));
+    if (U_FAILURE(errorCode) || keywords.isNull() || keywords->count(errorCode) == 0) {
+        // Error, or no extensions to copy.
+        return;
+    }
+    if (extensions_ == nullptr) {
+        extensions_ = new Locale();
+        if (extensions_ == nullptr) {
+            status_ = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    }
+    _copyExtensions(src, keywords.getAlias(), *extensions_, false, errorCode);
+}
+
+Locale LocaleBuilder::build(UErrorCode& errorCode)
+{
+    if (U_FAILURE(errorCode)) {
+        return makeBogusLocale();
+    }
+    if (U_FAILURE(status_)) {
+        errorCode = status_;
+        return makeBogusLocale();
+    }
+    CharString locale_str(language_, errorCode);
+    if (uprv_strlen(script_) > 0) {
+        locale_str.append('-', errorCode).append(StringPiece(script_), errorCode);
+    }
+    if (uprv_strlen(region_) > 0) {
+        locale_str.append('-', errorCode).append(StringPiece(region_), errorCode);
+    }
+    if (variant_ != nullptr) {
+        locale_str.append('-', errorCode).append(StringPiece(variant_->data()), errorCode);
+    }
+    if (U_FAILURE(errorCode)) {
+        return makeBogusLocale();
+    }
+    Locale product(locale_str.data());
+    if (extensions_ != nullptr) {
+        _copyExtensions(*extensions_, nullptr, product, true, errorCode);
+    }
+    if (U_FAILURE(errorCode)) {
+        return makeBogusLocale();
+    }
+    return product;
+}
+
+UBool LocaleBuilder::copyErrorTo(UErrorCode &outErrorCode) const {
+    if (U_FAILURE(outErrorCode)) {
+        // Do not overwrite the older error code
+        return TRUE;
+    }
+    outErrorCode = status_;
+    return U_FAILURE(outErrorCode);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/localematcher.cpp b/thirdparty/icu4c/common/localematcher.cpp
new file mode 100644
index 0000000000..5795cbf87e
--- /dev/null
+++ b/thirdparty/icu4c/common/localematcher.cpp
@@ -0,0 +1,846 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// localematcher.cpp
+// created: 2019may08 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/localebuilder.h"
+#include "unicode/localematcher.h"
+#include "unicode/locid.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uloc.h"
+#include "unicode/uobject.h"
+#include "cstring.h"
+#include "localeprioritylist.h"
+#include "loclikelysubtags.h"
+#include "locdistance.h"
+#include "lsr.h"
+#include "uassert.h"
+#include "uhash.h"
+#include "ustr_imp.h"
+#include "uvector.h"
+
+#define UND_LSR LSR("und", "", "", LSR::EXPLICIT_LSR)
+
+/**
+ * Indicator for the lifetime of desired-locale objects passed into the LocaleMatcher.
+ *
+ * @draft ICU 65
+ */
+enum ULocMatchLifetime {
+    /**
+     * Locale objects are temporary.
+     * The matcher will make a copy of a locale that will be used beyond one function call.
+     *
+     * @draft ICU 65
+     */
+    ULOCMATCH_TEMPORARY_LOCALES,
+    /**
+     * Locale objects are stored at least as long as the matcher is used.
+     * The matcher will keep only a pointer to a locale that will be used beyond one function call,
+     * avoiding a copy.
+     *
+     * @draft ICU 65
+     */
+    ULOCMATCH_STORED_LOCALES  // TODO: permanent? cached? clone?
+};
+#ifndef U_IN_DOXYGEN
+typedef enum ULocMatchLifetime ULocMatchLifetime;
+#endif
+
+U_NAMESPACE_BEGIN
+
+LocaleMatcher::Result::Result(LocaleMatcher::Result &&src) U_NOEXCEPT :
+        desiredLocale(src.desiredLocale),
+        supportedLocale(src.supportedLocale),
+        desiredIndex(src.desiredIndex),
+        supportedIndex(src.supportedIndex),
+        desiredIsOwned(src.desiredIsOwned) {
+    if (desiredIsOwned) {
+        src.desiredLocale = nullptr;
+        src.desiredIndex = -1;
+        src.desiredIsOwned = FALSE;
+    }
+}
+
+LocaleMatcher::Result::~Result() {
+    if (desiredIsOwned) {
+        delete desiredLocale;
+    }
+}
+
+LocaleMatcher::Result &LocaleMatcher::Result::operator=(LocaleMatcher::Result &&src) U_NOEXCEPT {
+    this->~Result();
+
+    desiredLocale = src.desiredLocale;
+    supportedLocale = src.supportedLocale;
+    desiredIndex = src.desiredIndex;
+    supportedIndex = src.supportedIndex;
+    desiredIsOwned = src.desiredIsOwned;
+
+    if (desiredIsOwned) {
+        src.desiredLocale = nullptr;
+        src.desiredIndex = -1;
+        src.desiredIsOwned = FALSE;
+    }
+    return *this;
+}
+
+Locale LocaleMatcher::Result::makeResolvedLocale(UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode) || supportedLocale == nullptr) {
+        return Locale::getRoot();
+    }
+    const Locale *bestDesired = getDesiredLocale();
+    if (bestDesired == nullptr || *supportedLocale == *bestDesired) {
+        return *supportedLocale;
+    }
+    LocaleBuilder b;
+    b.setLocale(*supportedLocale);
+
+    // Copy the region from bestDesired, if there is one.
+    const char *region = bestDesired->getCountry();
+    if (*region != 0) {
+        b.setRegion(region);
+    }
+
+    // Copy the variants from bestDesired, if there are any.
+    // Note that this will override any supportedLocale variants.
+    // For example, "sco-ulster-fonipa" + "...-fonupa" => "sco-fonupa" (replacing ulster).
+    const char *variants = bestDesired->getVariant();
+    if (*variants != 0) {
+        b.setVariant(variants);
+    }
+
+    // Copy the extensions from bestDesired, if there are any.
+    // C++ note: The following note, copied from Java, may not be true,
+    // as long as C++ copies by legacy ICU keyword, not by extension singleton.
+    // Note that this will override any supportedLocale extensions.
+    // For example, "th-u-nu-latn-ca-buddhist" + "...-u-nu-native" => "th-u-nu-native"
+    // (replacing calendar).
+    b.copyExtensionsFrom(*bestDesired, errorCode);
+    return b.build(errorCode);
+}
+
+LocaleMatcher::Builder::Builder(LocaleMatcher::Builder &&src) U_NOEXCEPT :
+        errorCode_(src.errorCode_),
+        supportedLocales_(src.supportedLocales_),
+        thresholdDistance_(src.thresholdDistance_),
+        demotion_(src.demotion_),
+        defaultLocale_(src.defaultLocale_),
+        withDefault_(src.withDefault_),
+        favor_(src.favor_),
+        direction_(src.direction_) {
+    src.supportedLocales_ = nullptr;
+    src.defaultLocale_ = nullptr;
+}
+
+LocaleMatcher::Builder::~Builder() {
+    delete supportedLocales_;
+    delete defaultLocale_;
+    delete maxDistanceDesired_;
+    delete maxDistanceSupported_;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::operator=(LocaleMatcher::Builder &&src) U_NOEXCEPT {
+    this->~Builder();
+
+    errorCode_ = src.errorCode_;
+    supportedLocales_ = src.supportedLocales_;
+    thresholdDistance_ = src.thresholdDistance_;
+    demotion_ = src.demotion_;
+    defaultLocale_ = src.defaultLocale_;
+    withDefault_ = src.withDefault_,
+    favor_ = src.favor_;
+    direction_ = src.direction_;
+
+    src.supportedLocales_ = nullptr;
+    src.defaultLocale_ = nullptr;
+    return *this;
+}
+
+void LocaleMatcher::Builder::clearSupportedLocales() {
+    if (supportedLocales_ != nullptr) {
+        supportedLocales_->removeAllElements();
+    }
+}
+
+bool LocaleMatcher::Builder::ensureSupportedLocaleVector() {
+    if (U_FAILURE(errorCode_)) { return false; }
+    if (supportedLocales_ != nullptr) { return true; }
+    supportedLocales_ = new UVector(uprv_deleteUObject, nullptr, errorCode_);
+    if (U_FAILURE(errorCode_)) { return false; }
+    if (supportedLocales_ == nullptr) {
+        errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+        return false;
+    }
+    return true;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setSupportedLocalesFromListString(
+        StringPiece locales) {
+    LocalePriorityList list(locales, errorCode_);
+    if (U_FAILURE(errorCode_)) { return *this; }
+    clearSupportedLocales();
+    if (!ensureSupportedLocaleVector()) { return *this; }
+    int32_t length = list.getLengthIncludingRemoved();
+    for (int32_t i = 0; i < length; ++i) {
+        Locale *locale = list.orphanLocaleAt(i);
+        if (locale == nullptr) { continue; }
+        supportedLocales_->addElement(locale, errorCode_);
+        if (U_FAILURE(errorCode_)) {
+            delete locale;
+            break;
+        }
+    }
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setSupportedLocales(Locale::Iterator &locales) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    clearSupportedLocales();
+    if (!ensureSupportedLocaleVector()) { return *this; }
+    while (locales.hasNext()) {
+        const Locale &locale = locales.next();
+        Locale *clone = locale.clone();
+        if (clone == nullptr) {
+            errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+            break;
+        }
+        supportedLocales_->addElement(clone, errorCode_);
+        if (U_FAILURE(errorCode_)) {
+            delete clone;
+            break;
+        }
+    }
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::addSupportedLocale(const Locale &locale) {
+    if (!ensureSupportedLocaleVector()) { return *this; }
+    Locale *clone = locale.clone();
+    if (clone == nullptr) {
+        errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+        return *this;
+    }
+    supportedLocales_->addElement(clone, errorCode_);
+    if (U_FAILURE(errorCode_)) {
+        delete clone;
+    }
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setNoDefaultLocale() {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    delete defaultLocale_;
+    defaultLocale_ = nullptr;
+    withDefault_ = false;
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setDefaultLocale(const Locale *defaultLocale) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    Locale *clone = nullptr;
+    if (defaultLocale != nullptr) {
+        clone = defaultLocale->clone();
+        if (clone == nullptr) {
+            errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+            return *this;
+        }
+    }
+    delete defaultLocale_;
+    defaultLocale_ = clone;
+    withDefault_ = true;
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setFavorSubtag(ULocMatchFavorSubtag subtag) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    favor_ = subtag;
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setDemotionPerDesiredLocale(ULocMatchDemotion demotion) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    demotion_ = demotion;
+    return *this;
+}
+
+LocaleMatcher::Builder &LocaleMatcher::Builder::setMaxDistance(const Locale &desired,
+                                                               const Locale &supported) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    Locale *desiredClone = desired.clone();
+    Locale *supportedClone = supported.clone();
+    if (desiredClone == nullptr || supportedClone == nullptr) {
+        delete desiredClone;  // in case only one could not be allocated
+        delete supportedClone;
+        errorCode_ = U_MEMORY_ALLOCATION_ERROR;
+        return *this;
+    }
+    delete maxDistanceDesired_;
+    delete maxDistanceSupported_;
+    maxDistanceDesired_ = desiredClone;
+    maxDistanceSupported_ = supportedClone;
+    return *this;
+}
+
+#if 0
+/**
+ * <i>Internal only!</i>
+ *
+ * @param thresholdDistance the thresholdDistance to set, with -1 = default
+ * @return this Builder object
+ * @internal
+ * @deprecated This API is ICU internal only.
+ */
+@Deprecated
+LocaleMatcher::Builder &LocaleMatcher::Builder::internalSetThresholdDistance(int32_t thresholdDistance) {
+    if (U_FAILURE(errorCode_)) { return *this; }
+    if (thresholdDistance > 100) {
+        thresholdDistance = 100;
+    }
+    thresholdDistance_ = thresholdDistance;
+    return *this;
+}
+#endif
+
+UBool LocaleMatcher::Builder::copyErrorTo(UErrorCode &outErrorCode) const {
+    if (U_FAILURE(outErrorCode)) { return TRUE; }
+    if (U_SUCCESS(errorCode_)) { return FALSE; }
+    outErrorCode = errorCode_;
+    return TRUE;
+}
+
+LocaleMatcher LocaleMatcher::Builder::build(UErrorCode &errorCode) const {
+    if (U_SUCCESS(errorCode) && U_FAILURE(errorCode_)) {
+        errorCode = errorCode_;
+    }
+    return LocaleMatcher(*this, errorCode);
+}
+
+namespace {
+
+LSR getMaximalLsrOrUnd(const XLikelySubtags &likelySubtags, const Locale &locale,
+                       UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode) || locale.isBogus() || *locale.getName() == 0 /* "und" */) {
+        return UND_LSR;
+    } else {
+        return likelySubtags.makeMaximizedLsrFrom(locale, errorCode);
+    }
+}
+
+int32_t hashLSR(const UHashTok token) {
+    const LSR *lsr = static_cast<const LSR *>(token.pointer);
+    return lsr->hashCode;
+}
+
+UBool compareLSRs(const UHashTok t1, const UHashTok t2) {
+    const LSR *lsr1 = static_cast<const LSR *>(t1.pointer);
+    const LSR *lsr2 = static_cast<const LSR *>(t2.pointer);
+    return *lsr1 == *lsr2;
+}
+
+}  // namespace
+
+int32_t LocaleMatcher::putIfAbsent(const LSR &lsr, int32_t i, int32_t suppLength,
+                                   UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return suppLength; }
+    int32_t index = uhash_geti(supportedLsrToIndex, &lsr);
+    if (index == 0) {
+        uhash_puti(supportedLsrToIndex, const_cast<LSR *>(&lsr), i + 1, &errorCode);
+        if (U_SUCCESS(errorCode)) {
+            supportedLSRs[suppLength] = &lsr;
+            supportedIndexes[suppLength++] = i;
+        }
+    }
+    return suppLength;
+}
+
+LocaleMatcher::LocaleMatcher(const Builder &builder, UErrorCode &errorCode) :
+        likelySubtags(*XLikelySubtags::getSingleton(errorCode)),
+        localeDistance(*LocaleDistance::getSingleton(errorCode)),
+        thresholdDistance(builder.thresholdDistance_),
+        demotionPerDesiredLocale(0),
+        favorSubtag(builder.favor_),
+        direction(builder.direction_),
+        supportedLocales(nullptr), lsrs(nullptr), supportedLocalesLength(0),
+        supportedLsrToIndex(nullptr),
+        supportedLSRs(nullptr), supportedIndexes(nullptr), supportedLSRsLength(0),
+        ownedDefaultLocale(nullptr), defaultLocale(nullptr) {
+    if (U_FAILURE(errorCode)) { return; }
+    const Locale *def = builder.defaultLocale_;
+    LSR builderDefaultLSR;
+    const LSR *defLSR = nullptr;
+    if (def != nullptr) {
+        ownedDefaultLocale = def->clone();
+        if (ownedDefaultLocale == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        def = ownedDefaultLocale;
+        builderDefaultLSR = getMaximalLsrOrUnd(likelySubtags, *def, errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        defLSR = &builderDefaultLSR;
+    }
+    supportedLocalesLength = builder.supportedLocales_ != nullptr ?
+        builder.supportedLocales_->size() : 0;
+    if (supportedLocalesLength > 0) {
+        // Store the supported locales in input order,
+        // so that when different types are used (e.g., language tag strings)
+        // we can return those by parallel index.
+        supportedLocales = static_cast<const Locale **>(
+            uprv_malloc(supportedLocalesLength * sizeof(const Locale *)));
+        // Supported LRSs in input order.
+        // In C++, we store these permanently to simplify ownership management
+        // in the hash tables. Duplicate LSRs (if any) are unused overhead.
+        lsrs = new LSR[supportedLocalesLength];
+        if (supportedLocales == nullptr || lsrs == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        // If the constructor fails partway, we need null pointers for destructibility.
+        uprv_memset(supportedLocales, 0, supportedLocalesLength * sizeof(const Locale *));
+        for (int32_t i = 0; i < supportedLocalesLength; ++i) {
+            const Locale &locale = *static_cast<Locale *>(builder.supportedLocales_->elementAt(i));
+            supportedLocales[i] = locale.clone();
+            if (supportedLocales[i] == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            const Locale &supportedLocale = *supportedLocales[i];
+            LSR &lsr = lsrs[i] = getMaximalLsrOrUnd(likelySubtags, supportedLocale, errorCode);
+            lsr.setHashCode();
+            if (U_FAILURE(errorCode)) { return; }
+        }
+
+        // We need an unordered map from LSR to first supported locale with that LSR,
+        // and an ordered list of (LSR, supported index) for
+        // the supported locales in the following order:
+        // 1. Default locale, if it is supported.
+        // 2. Priority locales (aka "paradigm locales") in builder order.
+        // 3. Remaining locales in builder order.
+        supportedLsrToIndex = uhash_openSize(hashLSR, compareLSRs, uhash_compareLong,
+                                             supportedLocalesLength, &errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        supportedLSRs = static_cast<const LSR **>(
+            uprv_malloc(supportedLocalesLength * sizeof(const LSR *)));
+        supportedIndexes = static_cast<int32_t *>(
+            uprv_malloc(supportedLocalesLength * sizeof(int32_t)));
+        if (supportedLSRs == nullptr || supportedIndexes == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        int32_t suppLength = 0;
+        // Determine insertion order.
+        // Add locales immediately that are equivalent to the default.
+        MaybeStackArray<int8_t, 100> order(supportedLocalesLength, errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        int32_t numParadigms = 0;
+        for (int32_t i = 0; i < supportedLocalesLength; ++i) {
+            const Locale &locale = *supportedLocales[i];
+            const LSR &lsr = lsrs[i];
+            if (defLSR == nullptr && builder.withDefault_) {
+                // Implicit default locale = first supported locale, if not turned off.
+                U_ASSERT(i == 0);
+                def = &locale;
+                defLSR = &lsr;
+                order[i] = 1;
+                suppLength = putIfAbsent(lsr, 0, suppLength, errorCode);
+            } else if (defLSR != nullptr && lsr.isEquivalentTo(*defLSR)) {
+                order[i] = 1;
+                suppLength = putIfAbsent(lsr, i, suppLength, errorCode);
+            } else if (localeDistance.isParadigmLSR(lsr)) {
+                order[i] = 2;
+                ++numParadigms;
+            } else {
+                order[i] = 3;
+            }
+            if (U_FAILURE(errorCode)) { return; }
+        }
+        // Add supported paradigm locales.
+        int32_t paradigmLimit = suppLength + numParadigms;
+        for (int32_t i = 0; i < supportedLocalesLength && suppLength < paradigmLimit; ++i) {
+            if (order[i] == 2) {
+                suppLength = putIfAbsent(lsrs[i], i, suppLength, errorCode);
+            }
+        }
+        // Add remaining supported locales.
+        for (int32_t i = 0; i < supportedLocalesLength; ++i) {
+            if (order[i] == 3) {
+                suppLength = putIfAbsent(lsrs[i], i, suppLength, errorCode);
+            }
+        }
+        supportedLSRsLength = suppLength;
+        // If supportedLSRsLength < supportedLocalesLength then
+        // we waste as many array slots as there are duplicate supported LSRs,
+        // but the amount of wasted space is small as long as there are few duplicates.
+    }
+
+    defaultLocale = def;
+
+    if (builder.demotion_ == ULOCMATCH_DEMOTION_REGION) {
+        demotionPerDesiredLocale = localeDistance.getDefaultDemotionPerDesiredLocale();
+    }
+
+    if (thresholdDistance >= 0) {
+        // already copied
+    } else if (builder.maxDistanceDesired_ != nullptr) {
+        LSR suppLSR = getMaximalLsrOrUnd(likelySubtags, *builder.maxDistanceSupported_, errorCode);
+        const LSR *pSuppLSR = &suppLSR;
+        int32_t indexAndDistance = localeDistance.getBestIndexAndDistance(
+                getMaximalLsrOrUnd(likelySubtags, *builder.maxDistanceDesired_, errorCode),
+                &pSuppLSR, 1,
+                LocaleDistance::shiftDistance(100), favorSubtag, direction);
+        if (U_SUCCESS(errorCode)) {
+            // +1 for an exclusive threshold from an inclusive max.
+            thresholdDistance = LocaleDistance::getDistanceFloor(indexAndDistance) + 1;
+        } else {
+            thresholdDistance = 0;
+        }
+    } else {
+        thresholdDistance = localeDistance.getDefaultScriptDistance();
+    }
+}
+
+LocaleMatcher::LocaleMatcher(LocaleMatcher &&src) U_NOEXCEPT :
+        likelySubtags(src.likelySubtags),
+        localeDistance(src.localeDistance),
+        thresholdDistance(src.thresholdDistance),
+        demotionPerDesiredLocale(src.demotionPerDesiredLocale),
+        favorSubtag(src.favorSubtag),
+        direction(src.direction),
+        supportedLocales(src.supportedLocales), lsrs(src.lsrs),
+        supportedLocalesLength(src.supportedLocalesLength),
+        supportedLsrToIndex(src.supportedLsrToIndex),
+        supportedLSRs(src.supportedLSRs),
+        supportedIndexes(src.supportedIndexes),
+        supportedLSRsLength(src.supportedLSRsLength),
+        ownedDefaultLocale(src.ownedDefaultLocale), defaultLocale(src.defaultLocale) {
+    src.supportedLocales = nullptr;
+    src.lsrs = nullptr;
+    src.supportedLocalesLength = 0;
+    src.supportedLsrToIndex = nullptr;
+    src.supportedLSRs = nullptr;
+    src.supportedIndexes = nullptr;
+    src.supportedLSRsLength = 0;
+    src.ownedDefaultLocale = nullptr;
+    src.defaultLocale = nullptr;
+}
+
+LocaleMatcher::~LocaleMatcher() {
+    for (int32_t i = 0; i < supportedLocalesLength; ++i) {
+        delete supportedLocales[i];
+    }
+    uprv_free(supportedLocales);
+    delete[] lsrs;
+    uhash_close(supportedLsrToIndex);
+    uprv_free(supportedLSRs);
+    uprv_free(supportedIndexes);
+    delete ownedDefaultLocale;
+}
+
+LocaleMatcher &LocaleMatcher::operator=(LocaleMatcher &&src) U_NOEXCEPT {
+    this->~LocaleMatcher();
+
+    thresholdDistance = src.thresholdDistance;
+    demotionPerDesiredLocale = src.demotionPerDesiredLocale;
+    favorSubtag = src.favorSubtag;
+    direction = src.direction;
+    supportedLocales = src.supportedLocales;
+    lsrs = src.lsrs;
+    supportedLocalesLength = src.supportedLocalesLength;
+    supportedLsrToIndex = src.supportedLsrToIndex;
+    supportedLSRs = src.supportedLSRs;
+    supportedIndexes = src.supportedIndexes;
+    supportedLSRsLength = src.supportedLSRsLength;
+    ownedDefaultLocale = src.ownedDefaultLocale;
+    defaultLocale = src.defaultLocale;
+
+    src.supportedLocales = nullptr;
+    src.lsrs = nullptr;
+    src.supportedLocalesLength = 0;
+    src.supportedLsrToIndex = nullptr;
+    src.supportedLSRs = nullptr;
+    src.supportedIndexes = nullptr;
+    src.supportedLSRsLength = 0;
+    src.ownedDefaultLocale = nullptr;
+    src.defaultLocale = nullptr;
+    return *this;
+}
+
+class LocaleLsrIterator {
+public:
+    LocaleLsrIterator(const XLikelySubtags &likelySubtags, Locale::Iterator &locales,
+                      ULocMatchLifetime lifetime) :
+            likelySubtags(likelySubtags), locales(locales), lifetime(lifetime) {}
+
+    ~LocaleLsrIterator() {
+        if (lifetime == ULOCMATCH_TEMPORARY_LOCALES) {
+            delete remembered;
+        }
+    }
+
+    bool hasNext() const {
+        return locales.hasNext();
+    }
+
+    LSR next(UErrorCode &errorCode) {
+        current = &locales.next();
+        return getMaximalLsrOrUnd(likelySubtags, *current, errorCode);
+    }
+
+    void rememberCurrent(int32_t desiredIndex, UErrorCode &errorCode) {
+        if (U_FAILURE(errorCode)) { return; }
+        bestDesiredIndex = desiredIndex;
+        if (lifetime == ULOCMATCH_STORED_LOCALES) {
+            remembered = current;
+        } else {
+            // ULOCMATCH_TEMPORARY_LOCALES
+            delete remembered;
+            remembered = new Locale(*current);
+            if (remembered == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+            }
+        }
+    }
+
+    const Locale *orphanRemembered() {
+        const Locale *rem = remembered;
+        remembered = nullptr;
+        return rem;
+    }
+
+    int32_t getBestDesiredIndex() const {
+        return bestDesiredIndex;
+    }
+
+private:
+    const XLikelySubtags &likelySubtags;
+    Locale::Iterator &locales;
+    ULocMatchLifetime lifetime;
+    const Locale *current = nullptr, *remembered = nullptr;
+    int32_t bestDesiredIndex = -1;
+};
+
+const Locale *LocaleMatcher::getBestMatch(const Locale &desiredLocale, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    int32_t suppIndex = getBestSuppIndex(
+        getMaximalLsrOrUnd(likelySubtags, desiredLocale, errorCode),
+        nullptr, errorCode);
+    return U_SUCCESS(errorCode) && suppIndex >= 0 ? supportedLocales[suppIndex] : defaultLocale;
+}
+
+const Locale *LocaleMatcher::getBestMatch(Locale::Iterator &desiredLocales,
+                                          UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    if (!desiredLocales.hasNext()) {
+        return defaultLocale;
+    }
+    LocaleLsrIterator lsrIter(likelySubtags, desiredLocales, ULOCMATCH_TEMPORARY_LOCALES);
+    int32_t suppIndex = getBestSuppIndex(lsrIter.next(errorCode), &lsrIter, errorCode);
+    return U_SUCCESS(errorCode) && suppIndex >= 0 ? supportedLocales[suppIndex] : defaultLocale;
+}
+
+const Locale *LocaleMatcher::getBestMatchForListString(
+        StringPiece desiredLocaleList, UErrorCode &errorCode) const {
+    LocalePriorityList list(desiredLocaleList, errorCode);
+    LocalePriorityList::Iterator iter = list.iterator();
+    return getBestMatch(iter, errorCode);
+}
+
+LocaleMatcher::Result LocaleMatcher::getBestMatchResult(
+        const Locale &desiredLocale, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return Result(nullptr, defaultLocale, -1, -1, FALSE);
+    }
+    int32_t suppIndex = getBestSuppIndex(
+        getMaximalLsrOrUnd(likelySubtags, desiredLocale, errorCode),
+        nullptr, errorCode);
+    if (U_FAILURE(errorCode) || suppIndex < 0) {
+        return Result(nullptr, defaultLocale, -1, -1, FALSE);
+    } else {
+        return Result(&desiredLocale, supportedLocales[suppIndex], 0, suppIndex, FALSE);
+    }
+}
+
+LocaleMatcher::Result LocaleMatcher::getBestMatchResult(
+        Locale::Iterator &desiredLocales, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode) || !desiredLocales.hasNext()) {
+        return Result(nullptr, defaultLocale, -1, -1, FALSE);
+    }
+    LocaleLsrIterator lsrIter(likelySubtags, desiredLocales, ULOCMATCH_TEMPORARY_LOCALES);
+    int32_t suppIndex = getBestSuppIndex(lsrIter.next(errorCode), &lsrIter, errorCode);
+    if (U_FAILURE(errorCode) || suppIndex < 0) {
+        return Result(nullptr, defaultLocale, -1, -1, FALSE);
+    } else {
+        return Result(lsrIter.orphanRemembered(), supportedLocales[suppIndex],
+                      lsrIter.getBestDesiredIndex(), suppIndex, TRUE);
+    }
+}
+
+int32_t LocaleMatcher::getBestSuppIndex(LSR desiredLSR, LocaleLsrIterator *remainingIter,
+                                        UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) { return -1; }
+    int32_t desiredIndex = 0;
+    int32_t bestSupportedLsrIndex = -1;
+    for (int32_t bestShiftedDistance = LocaleDistance::shiftDistance(thresholdDistance);;) {
+        // Quick check for exact maximized LSR.
+        // Returns suppIndex+1 where 0 means not found.
+        if (supportedLsrToIndex != nullptr) {
+            desiredLSR.setHashCode();
+            int32_t index = uhash_geti(supportedLsrToIndex, &desiredLSR);
+            if (index != 0) {
+                int32_t suppIndex = index - 1;
+                if (remainingIter != nullptr) {
+                    remainingIter->rememberCurrent(desiredIndex, errorCode);
+                }
+                return suppIndex;
+            }
+        }
+        int32_t bestIndexAndDistance = localeDistance.getBestIndexAndDistance(
+                desiredLSR, supportedLSRs, supportedLSRsLength,
+                bestShiftedDistance, favorSubtag, direction);
+        if (bestIndexAndDistance >= 0) {
+            bestShiftedDistance = LocaleDistance::getShiftedDistance(bestIndexAndDistance);
+            if (remainingIter != nullptr) {
+                remainingIter->rememberCurrent(desiredIndex, errorCode);
+                if (U_FAILURE(errorCode)) { return -1; }
+            }
+            bestSupportedLsrIndex = LocaleDistance::getIndex(bestIndexAndDistance);
+        }
+        if ((bestShiftedDistance -= LocaleDistance::shiftDistance(demotionPerDesiredLocale)) <= 0) {
+            break;
+        }
+        if (remainingIter == nullptr || !remainingIter->hasNext()) {
+            break;
+        }
+        desiredLSR = remainingIter->next(errorCode);
+        if (U_FAILURE(errorCode)) { return -1; }
+        ++desiredIndex;
+    }
+    if (bestSupportedLsrIndex < 0) {
+        // no good match
+        return -1;
+    }
+    return supportedIndexes[bestSupportedLsrIndex];
+}
+
+UBool LocaleMatcher::isMatch(const Locale &desired, const Locale &supported,
+                             UErrorCode &errorCode) const {
+    LSR suppLSR = getMaximalLsrOrUnd(likelySubtags, supported, errorCode);
+    if (U_FAILURE(errorCode)) { return 0; }
+    const LSR *pSuppLSR = &suppLSR;
+    int32_t indexAndDistance = localeDistance.getBestIndexAndDistance(
+            getMaximalLsrOrUnd(likelySubtags, desired, errorCode),
+            &pSuppLSR, 1,
+            LocaleDistance::shiftDistance(thresholdDistance), favorSubtag, direction);
+    return indexAndDistance >= 0;
+}
+
+double LocaleMatcher::internalMatch(const Locale &desired, const Locale &supported, UErrorCode &errorCode) const {
+    // Returns the inverse of the distance: That is, 1-distance(desired, supported).
+    LSR suppLSR = getMaximalLsrOrUnd(likelySubtags, supported, errorCode);
+    if (U_FAILURE(errorCode)) { return 0; }
+    const LSR *pSuppLSR = &suppLSR;
+    int32_t indexAndDistance = localeDistance.getBestIndexAndDistance(
+            getMaximalLsrOrUnd(likelySubtags, desired, errorCode),
+            &pSuppLSR, 1,
+            LocaleDistance::shiftDistance(thresholdDistance), favorSubtag, direction);
+    double distance = LocaleDistance::getDistanceDouble(indexAndDistance);
+    return (100.0 - distance) / 100.0;
+}
+
+U_NAMESPACE_END
+
+// uloc_acceptLanguage() --------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+namespace {
+
+class LocaleFromTag {
+public:
+    LocaleFromTag() : locale(Locale::getRoot()) {}
+    const Locale &operator()(const char *tag) { return locale = Locale(tag); }
+
+private:
+    // Store the locale in the converter, rather than return a reference to a temporary,
+    // or a value which could go out of scope with the caller's reference to it.
+    Locale locale;
+};
+
+int32_t acceptLanguage(UEnumeration &supportedLocales, Locale::Iterator &desiredLocales,
+                       char *dest, int32_t capacity, UAcceptResult *acceptResult,
+                       UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return 0; }
+    LocaleMatcher::Builder builder;
+    const char *locString;
+    while ((locString = uenum_next(&supportedLocales, nullptr, &errorCode)) != nullptr) {
+        Locale loc(locString);
+        if (loc.isBogus()) {
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            return 0;
+        }
+        builder.addSupportedLocale(loc);
+    }
+    LocaleMatcher matcher = builder.build(errorCode);
+    LocaleMatcher::Result result = matcher.getBestMatchResult(desiredLocales, errorCode);
+    if (U_FAILURE(errorCode)) { return 0; }
+    if (result.getDesiredIndex() >= 0) {
+        if (acceptResult != nullptr) {
+            *acceptResult = *result.getDesiredLocale() == *result.getSupportedLocale() ?
+                ULOC_ACCEPT_VALID : ULOC_ACCEPT_FALLBACK;
+        }
+        const char *bestStr = result.getSupportedLocale()->getName();
+        int32_t bestLength = (int32_t)uprv_strlen(bestStr);
+        if (bestLength <= capacity) {
+            uprv_memcpy(dest, bestStr, bestLength);
+        }
+        return u_terminateChars(dest, capacity, bestLength, &errorCode);
+    } else {
+        if (acceptResult != nullptr) {
+            *acceptResult = ULOC_ACCEPT_FAILED;
+        }
+        return u_terminateChars(dest, capacity, 0, &errorCode);
+    }
+}
+
+}  // namespace
+
+U_CAPI int32_t U_EXPORT2
+uloc_acceptLanguage(char *result, int32_t resultAvailable,
+                    UAcceptResult *outResult,
+                    const char **acceptList, int32_t acceptListCount,
+                    UEnumeration *availableLocales,
+                    UErrorCode *status) {
+    if (U_FAILURE(*status)) { return 0; }
+    if ((result == nullptr ? resultAvailable != 0 : resultAvailable < 0) ||
+            (acceptList == nullptr ? acceptListCount != 0 : acceptListCount < 0) ||
+            availableLocales == nullptr) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    LocaleFromTag converter;
+    Locale::ConvertingIterator<const char **, LocaleFromTag> desiredLocales(
+        acceptList, acceptList + acceptListCount, converter);
+    return acceptLanguage(*availableLocales, desiredLocales,
+                          result, resultAvailable, outResult, *status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_acceptLanguageFromHTTP(char *result, int32_t resultAvailable,
+                            UAcceptResult *outResult,
+                            const char *httpAcceptLanguage,
+                            UEnumeration *availableLocales,
+                            UErrorCode *status) {
+    if (U_FAILURE(*status)) { return 0; }
+    if ((result == nullptr ? resultAvailable != 0 : resultAvailable < 0) ||
+            httpAcceptLanguage == nullptr || availableLocales == nullptr) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    LocalePriorityList list(httpAcceptLanguage, *status);
+    LocalePriorityList::Iterator desiredLocales = list.iterator();
+    return acceptLanguage(*availableLocales, desiredLocales,
+                          result, resultAvailable, outResult, *status);
+}
diff --git a/thirdparty/icu4c/common/localeprioritylist.cpp b/thirdparty/icu4c/common/localeprioritylist.cpp
new file mode 100644
index 0000000000..8916b121be
--- /dev/null
+++ b/thirdparty/icu4c/common/localeprioritylist.cpp
@@ -0,0 +1,239 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// localeprioritylist.cpp
+// created: 2019jul11 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/localpointer.h"
+#include "unicode/locid.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "localeprioritylist.h"
+#include "uarrsort.h"
+#include "uassert.h"
+#include "uhash.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+int32_t hashLocale(const UHashTok token) {
+    auto *locale = static_cast<const Locale *>(token.pointer);
+    return locale->hashCode();
+}
+
+UBool compareLocales(const UHashTok t1, const UHashTok t2) {
+    auto *l1 = static_cast<const Locale *>(t1.pointer);
+    auto *l2 = static_cast<const Locale *>(t2.pointer);
+    return *l1 == *l2;
+}
+
+constexpr int32_t WEIGHT_ONE = 1000;
+
+struct LocaleAndWeight {
+    Locale *locale;
+    int32_t weight;  // 0..1000 = 0.0..1.0
+    int32_t index;  // force stable sort
+
+    int32_t compare(const LocaleAndWeight &other) const {
+        int32_t diff = other.weight - weight;  // descending: other-this
+        if (diff != 0) { return diff; }
+        return index - other.index;
+    }
+};
+
+int32_t U_CALLCONV
+compareLocaleAndWeight(const void * /*context*/, const void *left, const void *right) {
+    return static_cast<const LocaleAndWeight *>(left)->
+        compare(*static_cast<const LocaleAndWeight *>(right));
+}
+
+const char *skipSpaces(const char *p, const char *limit) {
+    while (p < limit && *p == ' ') { ++p; }
+    return p;
+}
+
+int32_t findTagLength(const char *p, const char *limit) {
+    // Look for accept-language delimiters.
+    // Leave other validation up to the Locale constructor.
+    const char *q;
+    for (q = p; q < limit; ++q) {
+        char c = *q;
+        if (c == ' ' || c == ',' || c == ';') { break; }
+    }
+    return static_cast<int32_t>(q - p);
+}
+
+/**
+ * Parses and returns a qvalue weight in millis.
+ * Advances p to after the parsed substring.
+ * Returns a negative value if parsing fails.
+ */
+int32_t parseWeight(const char *&p, const char *limit) {
+    p = skipSpaces(p, limit);
+    char c;
+    if (p == limit || ((c = *p) != '0' && c != '1')) { return -1; }
+    int32_t weight = (c - '0') * 1000;
+    if (++p == limit || *p != '.') { return weight; }
+    int32_t multiplier = 100;
+    while (++p != limit && '0' <= (c = *p) && c <= '9') {
+        c -= '0';
+        if (multiplier > 0) {
+            weight += c * multiplier;
+            multiplier /= 10;
+        } else if (multiplier == 0) {
+            // round up
+            if (c >= 5) { ++weight; }
+            multiplier = -1;
+        }  // else ignore further fraction digits
+    }
+    return weight <= WEIGHT_ONE ? weight : -1;  // bad if > 1.0
+}
+
+}  // namespace
+
+/**
+ * Nothing but a wrapper over a MaybeStackArray of LocaleAndWeight.
+ *
+ * This wrapper exists (and is not in an anonymous namespace)
+ * so that we can forward-declare it in the header file and
+ * don't have to expose the MaybeStackArray specialization and
+ * the LocaleAndWeight to code (like the test) that #includes localeprioritylist.h.
+ * Also, otherwise we would have to do a platform-specific
+ * template export declaration of some kind for the MaybeStackArray specialization
+ * to be properly exported from the common DLL.
+ */
+struct LocaleAndWeightArray : public UMemory {
+    MaybeStackArray<LocaleAndWeight, 20> array;
+};
+
+LocalePriorityList::LocalePriorityList(StringPiece s, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return; }
+    list = new LocaleAndWeightArray();
+    if (list == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    const char *p = s.data();
+    const char *limit = p + s.length();
+    while ((p = skipSpaces(p, limit)) != limit) {
+        if (*p == ',') {  // empty range field
+            ++p;
+            continue;
+        }
+        int32_t tagLength = findTagLength(p, limit);
+        if (tagLength == 0) {
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        CharString tag(p, tagLength, errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        Locale locale = Locale(tag.data());
+        if (locale.isBogus()) {
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        int32_t weight = WEIGHT_ONE;
+        if ((p = skipSpaces(p + tagLength, limit)) != limit && *p == ';') {
+            if ((p = skipSpaces(p + 1, limit)) == limit || *p != 'q' ||
+                    (p = skipSpaces(p + 1, limit)) == limit || *p != '=' ||
+                    (++p, (weight = parseWeight(p, limit)) < 0)) {
+                errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+                return;
+            }
+            p = skipSpaces(p, limit);
+        }
+        if (p != limit && *p != ',') {  // trailing junk
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        add(locale, weight, errorCode);
+        if (p == limit) { break; }
+        ++p;
+    }
+    sort(errorCode);
+}
+
+LocalePriorityList::~LocalePriorityList() {
+    if (list != nullptr) {
+        for (int32_t i = 0; i < listLength; ++i) {
+            delete list->array[i].locale;
+        }
+        delete list;
+    }
+    uhash_close(map);
+}
+
+const Locale *LocalePriorityList::localeAt(int32_t i) const {
+    return list->array[i].locale;
+}
+
+Locale *LocalePriorityList::orphanLocaleAt(int32_t i) {
+    if (list == nullptr) { return nullptr; }
+    LocaleAndWeight &lw = list->array[i];
+    Locale *l = lw.locale;
+    lw.locale = nullptr;
+    return l;
+}
+
+bool LocalePriorityList::add(const Locale &locale, int32_t weight, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return false; }
+    if (map == nullptr) {
+        if (weight <= 0) { return true; }  // do not add q=0
+        map = uhash_open(hashLocale, compareLocales, uhash_compareLong, &errorCode);
+        if (U_FAILURE(errorCode)) { return false; }
+    }
+    LocalPointer<Locale> clone;
+    int32_t index = uhash_geti(map, &locale);
+    if (index != 0) {
+        // Duplicate: Remove the old item and append it anew.
+        LocaleAndWeight &lw = list->array[index - 1];
+        clone.adoptInstead(lw.locale);
+        lw.locale = nullptr;
+        lw.weight = 0;
+        ++numRemoved;
+    }
+    if (weight <= 0) {  // do not add q=0
+        if (index != 0) {
+            // Not strictly necessary but cleaner.
+            uhash_removei(map, &locale);
+        }
+        return true;
+    }
+    if (clone.isNull()) {
+        clone.adoptInstead(locale.clone());
+        if (clone.isNull() || (clone->isBogus() && !locale.isBogus())) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return false;
+        }
+    }
+    if (listLength == list->array.getCapacity()) {
+        int32_t newCapacity = listLength < 50 ? 100 : 4 * listLength;
+        if (list->array.resize(newCapacity, listLength) == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return false;
+        }
+    }
+    uhash_puti(map, clone.getAlias(), listLength + 1, &errorCode);
+    if (U_FAILURE(errorCode)) { return false; }
+    LocaleAndWeight &lw = list->array[listLength];
+    lw.locale = clone.orphan();
+    lw.weight = weight;
+    lw.index = listLength++;
+    if (weight < WEIGHT_ONE) { hasWeights = true; }
+    U_ASSERT(uhash_count(map) == getLength());
+    return true;
+}
+
+void LocalePriorityList::sort(UErrorCode &errorCode) {
+    // Sort by descending weights if there is a mix of weights.
+    // The comparator forces a stable sort via the item index.
+    if (U_FAILURE(errorCode) || getLength() <= 1 || !hasWeights) { return; }
+    uprv_sortArray(list->array.getAlias(), listLength, sizeof(LocaleAndWeight),
+                   compareLocaleAndWeight, nullptr, FALSE, &errorCode);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/localeprioritylist.h b/thirdparty/icu4c/common/localeprioritylist.h
new file mode 100644
index 0000000000..41e9d3ea08
--- /dev/null
+++ b/thirdparty/icu4c/common/localeprioritylist.h
@@ -0,0 +1,115 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// localeprioritylist.h
+// created: 2019jul11 Markus W. Scherer
+
+#ifndef __LOCALEPRIORITYLIST_H__
+#define __LOCALEPRIORITYLIST_H__
+
+#include "unicode/utypes.h"
+#include "unicode/locid.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+
+struct UHashtable;
+
+U_NAMESPACE_BEGIN
+
+struct LocaleAndWeightArray;
+
+/**
+ * Parses a list of locales from an accept-language string.
+ * We are a bit more lenient than the spec:
+ * We accept extra whitespace in more places, empty range fields,
+ * and any number of qvalue fraction digits.
+ *
+ * https://tools.ietf.org/html/rfc2616#section-14.4
+ * 14.4 Accept-Language
+ *
+ *        Accept-Language = "Accept-Language" ":"
+ *                          1#( language-range [ ";" "q" "=" qvalue ] )
+ *        language-range  = ( ( 1*8ALPHA *( "-" 1*8ALPHA ) ) | "*" )
+ *
+ *    Each language-range MAY be given an associated quality value which
+ *    represents an estimate of the user's preference for the languages
+ *    specified by that range. The quality value defaults to "q=1". For
+ *    example,
+ *
+ *        Accept-Language: da, en-gb;q=0.8, en;q=0.7
+ *
+ * https://tools.ietf.org/html/rfc2616#section-3.9
+ * 3.9 Quality Values
+ *
+ *    HTTP content negotiation (section 12) uses short "floating point"
+ *    numbers to indicate the relative importance ("weight") of various
+ *    negotiable parameters.  A weight is normalized to a real number in
+ *    the range 0 through 1, where 0 is the minimum and 1 the maximum
+ *    value. If a parameter has a quality value of 0, then content with
+ *    this parameter is `not acceptable' for the client. HTTP/1.1
+ *    applications MUST NOT generate more than three digits after the
+ *    decimal point. User configuration of these values SHOULD also be
+ *    limited in this fashion.
+ *
+ *        qvalue         = ( "0" [ "." 0*3DIGIT ] )
+ *                       | ( "1" [ "." 0*3("0") ] )
+ */
+class U_COMMON_API LocalePriorityList : public UMemory {
+public:
+    class Iterator : public Locale::Iterator {
+    public:
+        UBool hasNext() const override { return count < length; }
+
+        const Locale &next() override {
+            for(;;) {
+                const Locale *locale = list.localeAt(index++);
+                if (locale != nullptr) {
+                    ++count;
+                    return *locale;
+                }
+            }
+        }
+
+    private:
+        friend class LocalePriorityList;
+
+        Iterator(const LocalePriorityList &list) : list(list), length(list.getLength()) {}
+
+        const LocalePriorityList &list;
+        int32_t index = 0;
+        int32_t count = 0;
+        const int32_t length;
+    };
+
+    LocalePriorityList(StringPiece s, UErrorCode &errorCode);
+
+    ~LocalePriorityList();
+
+    int32_t getLength() const { return listLength - numRemoved; }
+
+    int32_t getLengthIncludingRemoved() const { return listLength; }
+
+    Iterator iterator() const { return Iterator(*this); }
+
+    const Locale *localeAt(int32_t i) const;
+
+    Locale *orphanLocaleAt(int32_t i);
+
+private:
+    LocalePriorityList(const LocalePriorityList &) = delete;
+    LocalePriorityList &operator=(const LocalePriorityList &) = delete;
+
+    bool add(const Locale &locale, int32_t weight, UErrorCode &errorCode);
+
+    void sort(UErrorCode &errorCode);
+
+    LocaleAndWeightArray *list = nullptr;
+    int32_t listLength = 0;
+    int32_t numRemoved = 0;
+    bool hasWeights = false;  // other than 1.0
+    UHashtable *map = nullptr;
+};
+
+U_NAMESPACE_END
+
+#endif  // __LOCALEPRIORITYLIST_H__
diff --git a/thirdparty/icu4c/common/localsvc.h b/thirdparty/icu4c/common/localsvc.h
new file mode 100644
index 0000000000..3364019513
--- /dev/null
+++ b/thirdparty/icu4c/common/localsvc.h
@@ -0,0 +1,27 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+*   Copyright (C) 2006 International Business Machines Corporation        *
+*   and others. All rights reserved.                                      *
+***************************************************************************
+*/
+
+#ifndef LOCALSVC_H
+#define LOCALSVC_H
+
+#include "unicode/utypes.h"
+
+#if defined(U_LOCAL_SERVICE_HOOK) && U_LOCAL_SERVICE_HOOK
+/**
+ * Prototype for user-supplied service hook. This function is expected to return
+ * a type of factory object specific to the requested service.
+ * 
+ * @param what service-specific string identifying the specific user hook
+ * @param status error status
+ * @return a service-specific hook, or NULL on failure.
+ */
+U_CAPI void* uprv_svc_hook(const char *what, UErrorCode *status);
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/locavailable.cpp b/thirdparty/icu4c/common/locavailable.cpp
new file mode 100644
index 0000000000..e8ec512e37
--- /dev/null
+++ b/thirdparty/icu4c/common/locavailable.cpp
@@ -0,0 +1,270 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  locavailable.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010feb25
+*   created by: Markus W. Scherer
+*
+*   Code for available locales, separated out from other .cpp files
+*   that then do not depend on resource bundle code and res_index bundles.
+*/
+
+#include "unicode/errorcode.h"
+#include "unicode/utypes.h"
+#include "unicode/locid.h"
+#include "unicode/uloc.h"
+#include "unicode/ures.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "ucln_cmn.h"
+#include "uassert.h"
+#include "umutex.h"
+#include "uresimp.h"
+
+// C++ API ----------------------------------------------------------------- ***
+
+U_NAMESPACE_BEGIN
+
+static icu::Locale*  availableLocaleList = NULL;
+static int32_t  availableLocaleListCount;
+static icu::UInitOnce gInitOnceLocale = U_INITONCE_INITIALIZER;
+
+U_NAMESPACE_END
+
+U_CDECL_BEGIN
+
+static UBool U_CALLCONV locale_available_cleanup(void)
+{
+    U_NAMESPACE_USE
+
+    if (availableLocaleList) {
+        delete []availableLocaleList;
+        availableLocaleList = NULL;
+    }
+    availableLocaleListCount = 0;
+    gInitOnceLocale.reset();
+
+    return TRUE;
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+void U_CALLCONV locale_available_init() {
+    // This function is a friend of class Locale.
+    // This function is only invoked via umtx_initOnce().
+    
+    // for now, there is a hardcoded list, so just walk through that list and set it up.
+    //  Note: this function is a friend of class Locale.
+    availableLocaleListCount = uloc_countAvailable();
+    if(availableLocaleListCount) {
+       availableLocaleList = new Locale[availableLocaleListCount];
+    }
+    if (availableLocaleList == NULL) {
+        availableLocaleListCount= 0;
+    }
+    for (int32_t locCount=availableLocaleListCount-1; locCount>=0; --locCount) {
+        availableLocaleList[locCount].setFromPOSIXID(uloc_getAvailable(locCount));
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE_AVAILABLE, locale_available_cleanup);
+}
+
+const Locale* U_EXPORT2
+Locale::getAvailableLocales(int32_t& count)
+{
+    umtx_initOnce(gInitOnceLocale, &locale_available_init);
+    count = availableLocaleListCount;
+    return availableLocaleList;
+}
+
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+/* ### Constants **************************************************/
+
+namespace {
+
+// Enough capacity for the two lists in the res_index.res file
+const char** gAvailableLocaleNames[2] = {};
+int32_t gAvailableLocaleCounts[2] = {};
+icu::UInitOnce ginstalledLocalesInitOnce = U_INITONCE_INITIALIZER;
+
+class AvailableLocalesSink : public ResourceSink {
+  public:
+    void put(const char *key, ResourceValue &value, UBool /*noFallback*/, UErrorCode &status) U_OVERRIDE {
+        ResourceTable resIndexTable = value.getTable(status);
+        if (U_FAILURE(status)) {
+            return;
+        }
+        for (int32_t i = 0; resIndexTable.getKeyAndValue(i, key, value); ++i) {
+            ULocAvailableType type;
+            if (uprv_strcmp(key, "InstalledLocales") == 0) {
+                type = ULOC_AVAILABLE_DEFAULT;
+            } else if (uprv_strcmp(key, "AliasLocales") == 0) {
+                type = ULOC_AVAILABLE_ONLY_LEGACY_ALIASES;
+            } else {
+                // CLDRVersion, etc.
+                continue;
+            }
+            ResourceTable availableLocalesTable = value.getTable(status);
+            if (U_FAILURE(status)) {
+                return;
+            }
+            gAvailableLocaleCounts[type] = availableLocalesTable.getSize();
+            gAvailableLocaleNames[type] = static_cast<const char**>(
+                uprv_malloc(gAvailableLocaleCounts[type] * sizeof(const char*)));
+            if (gAvailableLocaleNames[type] == nullptr) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            for (int32_t j = 0; availableLocalesTable.getKeyAndValue(j, key, value); ++j) {
+                gAvailableLocaleNames[type][j] = key;
+            }
+        }
+    }
+};
+
+class AvailableLocalesStringEnumeration : public StringEnumeration {
+  public:
+    AvailableLocalesStringEnumeration(ULocAvailableType type) : fType(type) {
+    }
+
+    const char* next(int32_t *resultLength, UErrorCode&) override {
+        ULocAvailableType actualType = fType;
+        int32_t actualIndex = fIndex++;
+
+        // If the "combined" list was requested, resolve that now
+        if (fType == ULOC_AVAILABLE_WITH_LEGACY_ALIASES) {
+            int32_t defaultLocalesCount = gAvailableLocaleCounts[ULOC_AVAILABLE_DEFAULT];
+            if (actualIndex < defaultLocalesCount) {
+                actualType = ULOC_AVAILABLE_DEFAULT;
+            } else {
+                actualIndex -= defaultLocalesCount;
+                actualType = ULOC_AVAILABLE_ONLY_LEGACY_ALIASES;
+            }
+        }
+
+        // Return the requested string
+        int32_t count = gAvailableLocaleCounts[actualType];
+        const char* result;
+        if (actualIndex < count) {
+            result = gAvailableLocaleNames[actualType][actualIndex];
+            if (resultLength != nullptr) {
+                *resultLength = static_cast<int32_t>(uprv_strlen(result));
+            }
+        } else {
+            result = nullptr;
+            if (resultLength != nullptr) {
+                *resultLength = 0;
+            }
+        }
+        return result;
+    }
+
+    void reset(UErrorCode&) override {
+        fIndex = 0;
+    }
+
+    int32_t count(UErrorCode&) const override {
+        if (fType == ULOC_AVAILABLE_WITH_LEGACY_ALIASES) {
+            return gAvailableLocaleCounts[ULOC_AVAILABLE_DEFAULT]
+                + gAvailableLocaleCounts[ULOC_AVAILABLE_ONLY_LEGACY_ALIASES];
+        } else {
+            return gAvailableLocaleCounts[fType];
+        }
+    }
+
+  private:
+    ULocAvailableType fType;
+    int32_t fIndex = 0;
+};
+
+/* ### Get available **************************************************/
+
+static UBool U_CALLCONV uloc_cleanup(void) {
+    for (int32_t i = 0; i < UPRV_LENGTHOF(gAvailableLocaleNames); i++) {
+        uprv_free(gAvailableLocaleNames[i]);
+        gAvailableLocaleNames[i] = nullptr;
+        gAvailableLocaleCounts[i] = 0;
+    }
+    ginstalledLocalesInitOnce.reset();
+    return TRUE;
+}
+
+// Load Installed Locales. This function will be called exactly once
+//   via the initOnce mechanism.
+
+static void U_CALLCONV loadInstalledLocales(UErrorCode& status) {
+    ucln_common_registerCleanup(UCLN_COMMON_ULOC, uloc_cleanup);
+
+    icu::LocalUResourceBundlePointer rb(ures_openDirect(NULL, "res_index", &status));
+    AvailableLocalesSink sink;
+    ures_getAllItemsWithFallback(rb.getAlias(), "", sink, status);
+}
+
+void _load_installedLocales(UErrorCode& status) {
+    umtx_initOnce(ginstalledLocalesInitOnce, &loadInstalledLocales, status);
+}
+
+} // namespace
+
+U_CAPI const char* U_EXPORT2
+uloc_getAvailable(int32_t offset) {
+    icu::ErrorCode status;
+    _load_installedLocales(status);
+    if (status.isFailure()) {
+        return nullptr;
+    }
+    if (offset > gAvailableLocaleCounts[0]) {
+        // *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    return gAvailableLocaleNames[0][offset];
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_countAvailable() {
+    icu::ErrorCode status;
+    _load_installedLocales(status);
+    if (status.isFailure()) {
+        return 0;
+    }
+    return gAvailableLocaleCounts[0];
+}
+
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openAvailableByType(ULocAvailableType type, UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return nullptr;
+    }
+    if (type < 0 || type >= ULOC_AVAILABLE_COUNT) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    _load_installedLocales(*status);
+    if (U_FAILURE(*status)) {
+        return nullptr;
+    }
+    LocalPointer<AvailableLocalesStringEnumeration> result(
+        new AvailableLocalesStringEnumeration(type), *status);
+    if (U_FAILURE(*status)) {
+        return nullptr;
+    }
+    return uenum_openFromStringEnumeration(result.orphan(), status);
+}
+
diff --git a/thirdparty/icu4c/common/locbased.cpp b/thirdparty/icu4c/common/locbased.cpp
new file mode 100644
index 0000000000..ff378b4cc7
--- /dev/null
+++ b/thirdparty/icu4c/common/locbased.cpp
@@ -0,0 +1,55 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2004-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: January 16 2004
+* Since: ICU 2.8
+**********************************************************************
+*/
+#include "locbased.h"
+#include "cstring.h"
+
+U_NAMESPACE_BEGIN
+
+Locale LocaleBased::getLocale(ULocDataLocaleType type, UErrorCode& status) const {
+    const char* id = getLocaleID(type, status);
+    return Locale((id != 0) ? id : "");
+}
+
+const char* LocaleBased::getLocaleID(ULocDataLocaleType type, UErrorCode& status) const {
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+    switch(type) {
+    case ULOC_VALID_LOCALE:
+        return valid;
+    case ULOC_ACTUAL_LOCALE:
+        return actual;
+    default:
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+}
+
+void LocaleBased::setLocaleIDs(const char* validID, const char* actualID) {
+    if (validID != 0) {
+      uprv_strncpy(valid, validID, ULOC_FULLNAME_CAPACITY);
+      valid[ULOC_FULLNAME_CAPACITY-1] = 0; // always terminate
+    }
+    if (actualID != 0) {
+      uprv_strncpy(actual, actualID, ULOC_FULLNAME_CAPACITY);
+      actual[ULOC_FULLNAME_CAPACITY-1] = 0; // always terminate
+    }
+}
+
+void LocaleBased::setLocaleIDs(const Locale& validID, const Locale& actualID) {
+  uprv_strcpy(valid, validID.getName());
+  uprv_strcpy(actual, actualID.getName());
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/locbased.h b/thirdparty/icu4c/common/locbased.h
new file mode 100644
index 0000000000..45738863b5
--- /dev/null
+++ b/thirdparty/icu4c/common/locbased.h
@@ -0,0 +1,107 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2004-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: January 16 2004
+* Since: ICU 2.8
+**********************************************************************
+*/
+#ifndef LOCBASED_H
+#define LOCBASED_H
+
+#include "unicode/locid.h"
+#include "unicode/uobject.h"
+
+/**
+ * Macro to declare a locale LocaleBased wrapper object for the given
+ * object, which must have two members named `validLocale' and
+ * `actualLocale' of size ULOC_FULLNAME_CAPACITY
+ */
+#define U_LOCALE_BASED(varname, objname) \
+  LocaleBased varname((objname).validLocale, (objname).actualLocale)
+
+U_NAMESPACE_BEGIN
+
+/**
+ * A utility class that unifies the implementation of getLocale() by
+ * various ICU services.  This class is likely to be removed in the
+ * ICU 3.0 time frame in favor of an integrated approach with the
+ * services framework.
+ * @since ICU 2.8
+ */
+class U_COMMON_API LocaleBased : public UMemory {
+
+ public:
+
+    /**
+     * Construct a LocaleBased wrapper around the two pointers.  These
+     * will be aliased for the lifetime of this object.
+     */
+    inline LocaleBased(char* validAlias, char* actualAlias);
+
+    /**
+     * Construct a LocaleBased wrapper around the two const pointers.
+     * These will be aliased for the lifetime of this object.
+     */
+    inline LocaleBased(const char* validAlias, const char* actualAlias);
+
+    /**
+     * Return locale meta-data for the service object wrapped by this
+     * object.  Either the valid or the actual locale may be
+     * retrieved.
+     * @param type either ULOC_VALID_LOCALE or ULOC_ACTUAL_LOCALE
+     * @param status input-output error code
+     * @return the indicated locale
+     */
+    Locale getLocale(ULocDataLocaleType type, UErrorCode& status) const;
+
+    /**
+     * Return the locale ID for the service object wrapped by this
+     * object.  Either the valid or the actual locale may be
+     * retrieved.
+     * @param type either ULOC_VALID_LOCALE or ULOC_ACTUAL_LOCALE
+     * @param status input-output error code
+     * @return the indicated locale ID
+     */
+    const char* getLocaleID(ULocDataLocaleType type, UErrorCode& status) const;
+
+    /**
+     * Set the locale meta-data for the service object wrapped by this
+     * object.  If either parameter is zero, it is ignored.
+     * @param valid the ID of the valid locale
+     * @param actual the ID of the actual locale
+     */
+    void setLocaleIDs(const char* valid, const char* actual);
+
+    /**
+     * Set the locale meta-data for the service object wrapped by this
+     * object.
+     * @param valid the ID of the valid locale
+     * @param actual the ID of the actual locale
+     */
+    void setLocaleIDs(const Locale& valid, const Locale& actual);
+
+ private:
+
+    char* valid;
+    
+    char* actual;
+};
+
+inline LocaleBased::LocaleBased(char* validAlias, char* actualAlias) :
+    valid(validAlias), actual(actualAlias) {
+}
+
+inline LocaleBased::LocaleBased(const char* validAlias,
+                                const char* actualAlias) :
+    // ugh: cast away const
+    valid((char*)validAlias), actual((char*)actualAlias) {
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/locdispnames.cpp b/thirdparty/icu4c/common/locdispnames.cpp
new file mode 100644
index 0000000000..47c0667417
--- /dev/null
+++ b/thirdparty/icu4c/common/locdispnames.cpp
@@ -0,0 +1,890 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  locdispnames.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010feb25
+*   created by: Markus W. Scherer
+*
+*   Code for locale display names, separated out from other .cpp files
+*   that then do not depend on resource bundle code and display name data.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/brkiter.h"
+#include "unicode/locid.h"
+#include "unicode/uenum.h"
+#include "unicode/uloc.h"
+#include "unicode/ures.h"
+#include "unicode/ustring.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "putilimp.h"
+#include "ulocimp.h"
+#include "uresimp.h"
+#include "ureslocs.h"
+#include "ustr_imp.h"
+
+// C++ API ----------------------------------------------------------------- ***
+
+U_NAMESPACE_BEGIN
+
+UnicodeString&
+Locale::getDisplayLanguage(UnicodeString& dispLang) const
+{
+    return this->getDisplayLanguage(getDefault(), dispLang);
+}
+
+/*We cannot make any assumptions on the size of the output display strings
+* Yet, since we are calling through to a C API, we need to set limits on
+* buffer size. For all the following getDisplay functions we first attempt
+* to fill up a stack allocated buffer. If it is to small we heap allocated
+* the exact buffer we need copy it to the UnicodeString and delete it*/
+
+UnicodeString&
+Locale::getDisplayLanguage(const Locale &displayLocale,
+                           UnicodeString &result) const {
+    UChar *buffer;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length;
+
+    buffer=result.getBuffer(ULOC_FULLNAME_CAPACITY);
+    if(buffer==0) {
+        result.truncate(0);
+        return result;
+    }
+
+    length=uloc_getDisplayLanguage(fullName, displayLocale.fullName,
+                                   buffer, result.getCapacity(),
+                                   &errorCode);
+    result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        buffer=result.getBuffer(length);
+        if(buffer==0) {
+            result.truncate(0);
+            return result;
+        }
+        errorCode=U_ZERO_ERROR;
+        length=uloc_getDisplayLanguage(fullName, displayLocale.fullName,
+                                       buffer, result.getCapacity(),
+                                       &errorCode);
+        result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+    }
+
+    return result;
+}
+
+UnicodeString&
+Locale::getDisplayScript(UnicodeString& dispScript) const
+{
+    return this->getDisplayScript(getDefault(), dispScript);
+}
+
+UnicodeString&
+Locale::getDisplayScript(const Locale &displayLocale,
+                          UnicodeString &result) const {
+    UChar *buffer;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length;
+
+    buffer=result.getBuffer(ULOC_FULLNAME_CAPACITY);
+    if(buffer==0) {
+        result.truncate(0);
+        return result;
+    }
+
+    length=uloc_getDisplayScript(fullName, displayLocale.fullName,
+                                  buffer, result.getCapacity(),
+                                  &errorCode);
+    result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        buffer=result.getBuffer(length);
+        if(buffer==0) {
+            result.truncate(0);
+            return result;
+        }
+        errorCode=U_ZERO_ERROR;
+        length=uloc_getDisplayScript(fullName, displayLocale.fullName,
+                                      buffer, result.getCapacity(),
+                                      &errorCode);
+        result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+    }
+
+    return result;
+}
+
+UnicodeString&
+Locale::getDisplayCountry(UnicodeString& dispCntry) const
+{
+    return this->getDisplayCountry(getDefault(), dispCntry);
+}
+
+UnicodeString&
+Locale::getDisplayCountry(const Locale &displayLocale,
+                          UnicodeString &result) const {
+    UChar *buffer;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length;
+
+    buffer=result.getBuffer(ULOC_FULLNAME_CAPACITY);
+    if(buffer==0) {
+        result.truncate(0);
+        return result;
+    }
+
+    length=uloc_getDisplayCountry(fullName, displayLocale.fullName,
+                                  buffer, result.getCapacity(),
+                                  &errorCode);
+    result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        buffer=result.getBuffer(length);
+        if(buffer==0) {
+            result.truncate(0);
+            return result;
+        }
+        errorCode=U_ZERO_ERROR;
+        length=uloc_getDisplayCountry(fullName, displayLocale.fullName,
+                                      buffer, result.getCapacity(),
+                                      &errorCode);
+        result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+    }
+
+    return result;
+}
+
+UnicodeString&
+Locale::getDisplayVariant(UnicodeString& dispVar) const
+{
+    return this->getDisplayVariant(getDefault(), dispVar);
+}
+
+UnicodeString&
+Locale::getDisplayVariant(const Locale &displayLocale,
+                          UnicodeString &result) const {
+    UChar *buffer;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length;
+
+    buffer=result.getBuffer(ULOC_FULLNAME_CAPACITY);
+    if(buffer==0) {
+        result.truncate(0);
+        return result;
+    }
+
+    length=uloc_getDisplayVariant(fullName, displayLocale.fullName,
+                                  buffer, result.getCapacity(),
+                                  &errorCode);
+    result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        buffer=result.getBuffer(length);
+        if(buffer==0) {
+            result.truncate(0);
+            return result;
+        }
+        errorCode=U_ZERO_ERROR;
+        length=uloc_getDisplayVariant(fullName, displayLocale.fullName,
+                                      buffer, result.getCapacity(),
+                                      &errorCode);
+        result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+    }
+
+    return result;
+}
+
+UnicodeString&
+Locale::getDisplayName( UnicodeString& name ) const
+{
+    return this->getDisplayName(getDefault(), name);
+}
+
+UnicodeString&
+Locale::getDisplayName(const Locale &displayLocale,
+                       UnicodeString &result) const {
+    UChar *buffer;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length;
+
+    buffer=result.getBuffer(ULOC_FULLNAME_CAPACITY);
+    if(buffer==0) {
+        result.truncate(0);
+        return result;
+    }
+
+    length=uloc_getDisplayName(fullName, displayLocale.fullName,
+                               buffer, result.getCapacity(),
+                               &errorCode);
+    result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        buffer=result.getBuffer(length);
+        if(buffer==0) {
+            result.truncate(0);
+            return result;
+        }
+        errorCode=U_ZERO_ERROR;
+        length=uloc_getDisplayName(fullName, displayLocale.fullName,
+                                   buffer, result.getCapacity(),
+                                   &errorCode);
+        result.releaseBuffer(U_SUCCESS(errorCode) ? length : 0);
+    }
+
+    return result;
+}
+
+#if ! UCONFIG_NO_BREAK_ITERATION
+
+// -------------------------------------
+// Gets the objectLocale display name in the default locale language.
+UnicodeString& U_EXPORT2
+BreakIterator::getDisplayName(const Locale& objectLocale,
+                             UnicodeString& name)
+{
+    return objectLocale.getDisplayName(name);
+}
+
+// -------------------------------------
+// Gets the objectLocale display name in the displayLocale language.
+UnicodeString& U_EXPORT2
+BreakIterator::getDisplayName(const Locale& objectLocale,
+                             const Locale& displayLocale,
+                             UnicodeString& name)
+{
+    return objectLocale.getDisplayName(displayLocale, name);
+}
+
+#endif
+
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+/* ### Constants **************************************************/
+
+/* These strings describe the resources we attempt to load from
+ the locale ResourceBundle data file.*/
+static const char _kLanguages[]       = "Languages";
+static const char _kScripts[]         = "Scripts";
+static const char _kScriptsStandAlone[] = "Scripts%stand-alone";
+static const char _kCountries[]       = "Countries";
+static const char _kVariants[]        = "Variants";
+static const char _kKeys[]            = "Keys";
+static const char _kTypes[]           = "Types";
+//static const char _kRootName[]        = "root";
+static const char _kCurrency[]        = "currency";
+static const char _kCurrencies[]      = "Currencies";
+static const char _kLocaleDisplayPattern[] = "localeDisplayPattern";
+static const char _kPattern[]         = "pattern";
+static const char _kSeparator[]       = "separator";
+
+/* ### Display name **************************************************/
+
+static int32_t
+_getStringOrCopyKey(const char *path, const char *locale,
+                    const char *tableKey, 
+                    const char* subTableKey,
+                    const char *itemKey,
+                    const char *substitute,
+                    UChar *dest, int32_t destCapacity,
+                    UErrorCode *pErrorCode) {
+    const UChar *s = NULL;
+    int32_t length = 0;
+
+    if(itemKey==NULL) {
+        /* top-level item: normal resource bundle access */
+        icu::LocalUResourceBundlePointer rb(ures_open(path, locale, pErrorCode));
+
+        if(U_SUCCESS(*pErrorCode)) {
+            s=ures_getStringByKey(rb.getAlias(), tableKey, &length, pErrorCode);
+            /* see comment about closing rb near "return item;" in _res_getTableStringWithFallback() */
+        }
+    } else {
+        /* Language code should not be a number. If it is, set the error code. */
+        if (!uprv_strncmp(tableKey, "Languages", 9) && uprv_strtol(itemKey, NULL, 10)) {
+            *pErrorCode = U_MISSING_RESOURCE_ERROR;
+        } else {
+            /* second-level item, use special fallback */
+            s=uloc_getTableStringWithFallback(path, locale,
+                                               tableKey, 
+                                               subTableKey,
+                                               itemKey,
+                                               &length,
+                                               pErrorCode);
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode)) {
+        int32_t copyLength=uprv_min(length, destCapacity);
+        if(copyLength>0 && s != NULL) {
+            u_memcpy(dest, s, copyLength);
+        }
+    } else {
+        /* no string from a resource bundle: convert the substitute */
+        length=(int32_t)uprv_strlen(substitute);
+        u_charsToUChars(substitute, dest, uprv_min(length, destCapacity));
+        *pErrorCode=U_USING_DEFAULT_WARNING;
+    }
+
+    return u_terminateUChars(dest, destCapacity, length, pErrorCode);
+}
+
+typedef  int32_t U_CALLCONV UDisplayNameGetter(const char *, char *, int32_t, UErrorCode *);
+
+static int32_t
+_getDisplayNameForComponent(const char *locale,
+                            const char *displayLocale,
+                            UChar *dest, int32_t destCapacity,
+                            UDisplayNameGetter *getter,
+                            const char *tag,
+                            UErrorCode *pErrorCode) {
+    char localeBuffer[ULOC_FULLNAME_CAPACITY*4];
+    int32_t length;
+    UErrorCode localStatus;
+    const char* root = NULL;
+
+    /* argument checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    localStatus = U_ZERO_ERROR;
+    length=(*getter)(locale, localeBuffer, sizeof(localeBuffer), &localStatus);
+    if(U_FAILURE(localStatus) || localStatus==U_STRING_NOT_TERMINATED_WARNING) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    if(length==0) {
+        // For the display name, we treat this as unknown language (ICU-20273).
+        if (getter == uloc_getLanguage) {
+            uprv_strcpy(localeBuffer, "und");
+        } else {
+            return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
+        }
+    }
+
+    root = tag == _kCountries ? U_ICUDATA_REGION : U_ICUDATA_LANG;
+
+    return _getStringOrCopyKey(root, displayLocale,
+                               tag, NULL, localeBuffer,
+                               localeBuffer,
+                               dest, destCapacity,
+                               pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayLanguage(const char *locale,
+                        const char *displayLocale,
+                        UChar *dest, int32_t destCapacity,
+                        UErrorCode *pErrorCode) {
+    return _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                uloc_getLanguage, _kLanguages, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayScript(const char* locale,
+                      const char* displayLocale,
+                      UChar *dest, int32_t destCapacity,
+                      UErrorCode *pErrorCode)
+{
+    UErrorCode err = U_ZERO_ERROR;
+    int32_t res = _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                uloc_getScript, _kScriptsStandAlone, &err);
+
+    if (destCapacity == 0 && err == U_BUFFER_OVERFLOW_ERROR) {
+        // For preflight, return the max of the value and the fallback.
+        int32_t fallback_res = _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                                                           uloc_getScript, _kScripts, pErrorCode);
+        return (fallback_res > res) ? fallback_res : res;
+    }
+    if ( err == U_USING_DEFAULT_WARNING ) {
+        return _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                                           uloc_getScript, _kScripts, pErrorCode);
+    } else {
+        *pErrorCode = err;
+        return res;
+    }
+}
+
+static int32_t
+uloc_getDisplayScriptInContext(const char* locale,
+                      const char* displayLocale,
+                      UChar *dest, int32_t destCapacity,
+                      UErrorCode *pErrorCode)
+{
+    return _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                    uloc_getScript, _kScripts, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayCountry(const char *locale,
+                       const char *displayLocale,
+                       UChar *dest, int32_t destCapacity,
+                       UErrorCode *pErrorCode) {
+    return _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                uloc_getCountry, _kCountries, pErrorCode);
+}
+
+/*
+ * TODO separate variant1_variant2_variant3...
+ * by getting each tag's display string and concatenating them with ", "
+ * in between - similar to uloc_getDisplayName()
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayVariant(const char *locale,
+                       const char *displayLocale,
+                       UChar *dest, int32_t destCapacity,
+                       UErrorCode *pErrorCode) {
+    return _getDisplayNameForComponent(locale, displayLocale, dest, destCapacity,
+                uloc_getVariant, _kVariants, pErrorCode);
+}
+
+/* Instead of having a separate pass for 'special' patterns, reintegrate the two
+ * so we don't get bitten by preflight bugs again.  We can be reasonably efficient
+ * without two separate code paths, this code isn't that performance-critical.
+ *
+ * This code is general enough to deal with patterns that have a prefix or swap the
+ * language and remainder components, since we gave developers enough rope to do such
+ * things if they futz with the pattern data.  But since we don't give them a way to
+ * specify a pattern for arbitrary combinations of components, there's not much use in
+ * that.  I don't think our data includes such patterns, the only variable I know if is
+ * whether there is a space before the open paren, or not.  Oh, and zh uses different
+ * chars than the standard open/close paren (which ja and ko use, btw).
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayName(const char *locale,
+                    const char *displayLocale,
+                    UChar *dest, int32_t destCapacity,
+                    UErrorCode *pErrorCode)
+{
+    static const UChar defaultSeparator[9] = { 0x007b, 0x0030, 0x007d, 0x002c, 0x0020, 0x007b, 0x0031, 0x007d, 0x0000 }; /* "{0}, {1}" */
+    static const UChar sub0[4] = { 0x007b, 0x0030, 0x007d , 0x0000 } ; /* {0} */
+    static const UChar sub1[4] = { 0x007b, 0x0031, 0x007d , 0x0000 } ; /* {1} */
+    static const int32_t subLen = 3;
+    static const UChar defaultPattern[10] = {
+        0x007b, 0x0030, 0x007d, 0x0020, 0x0028, 0x007b, 0x0031, 0x007d, 0x0029, 0x0000
+    }; /* {0} ({1}) */
+    static const int32_t defaultPatLen = 9;
+    static const int32_t defaultSub0Pos = 0;
+    static const int32_t defaultSub1Pos = 5;
+
+    int32_t length; /* of formatted result */
+
+    const UChar *separator;
+    int32_t sepLen = 0;
+    const UChar *pattern;
+    int32_t patLen = 0;
+    int32_t sub0Pos, sub1Pos;
+    
+    UChar formatOpenParen         = 0x0028; // (
+    UChar formatReplaceOpenParen  = 0x005B; // [
+    UChar formatCloseParen        = 0x0029; // )
+    UChar formatReplaceCloseParen = 0x005D; // ]
+
+    UBool haveLang = TRUE; /* assume true, set false if we find we don't have
+                              a lang component in the locale */
+    UBool haveRest = TRUE; /* assume true, set false if we find we don't have
+                              any other component in the locale */
+    UBool retry = FALSE; /* set true if we need to retry, see below */
+
+    int32_t langi = 0; /* index of the language substitution (0 or 1), virtually always 0 */
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    {
+        UErrorCode status = U_ZERO_ERROR;
+
+        icu::LocalUResourceBundlePointer locbundle(
+                ures_open(U_ICUDATA_LANG, displayLocale, &status));
+        icu::LocalUResourceBundlePointer dspbundle(
+                ures_getByKeyWithFallback(locbundle.getAlias(), _kLocaleDisplayPattern, NULL, &status));
+
+        separator=ures_getStringByKeyWithFallback(dspbundle.getAlias(), _kSeparator, &sepLen, &status);
+        pattern=ures_getStringByKeyWithFallback(dspbundle.getAlias(), _kPattern, &patLen, &status);
+    }
+
+    /* If we couldn't find any data, then use the defaults */
+    if(sepLen == 0) {
+       separator = defaultSeparator;
+    }
+    /* #10244: Even though separator is now a pattern, it is awkward to handle it as such
+     * here since we are trying to build the display string in place in the dest buffer,
+     * and to handle it as a pattern would entail having separate storage for the
+     * substrings that need to be combined (the first of which may be the result of
+     * previous such combinations). So for now we continue to treat the portion between
+     * {0} and {1} as a string to be appended when joining substrings, ignoring anything
+     * that is before {0} or after {1} (no existing separator pattern has any such thing).
+     * This is similar to how pattern is handled below.
+     */
+    {
+        UChar *p0=u_strstr(separator, sub0);
+        UChar *p1=u_strstr(separator, sub1);
+        if (p0==NULL || p1==NULL || p1<p0) {
+            *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return 0;
+        }
+        separator = (const UChar *)p0 + subLen;
+        sepLen = static_cast<int32_t>(p1 - separator);
+    }
+
+    if(patLen==0 || (patLen==defaultPatLen && !u_strncmp(pattern, defaultPattern, patLen))) {
+        pattern=defaultPattern;
+        patLen=defaultPatLen;
+        sub0Pos=defaultSub0Pos;
+        sub1Pos=defaultSub1Pos;
+        // use default formatOpenParen etc. set above
+    } else { /* non-default pattern */
+        UChar *p0=u_strstr(pattern, sub0);
+        UChar *p1=u_strstr(pattern, sub1);
+        if (p0==NULL || p1==NULL) {
+            *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return 0;
+        }
+        sub0Pos = static_cast<int32_t>(p0-pattern);
+        sub1Pos = static_cast<int32_t>(p1-pattern);
+        if (sub1Pos < sub0Pos) { /* a very odd pattern */
+            int32_t t=sub0Pos; sub0Pos=sub1Pos; sub1Pos=t;
+            langi=1;
+        }
+        if (u_strchr(pattern, 0xFF08) != NULL) {
+            formatOpenParen         = 0xFF08; // fullwidth (
+            formatReplaceOpenParen  = 0xFF3B; // fullwidth [
+            formatCloseParen        = 0xFF09; // fullwidth )
+            formatReplaceCloseParen = 0xFF3D; // fullwidth ]
+        }
+    }
+
+    /* We loop here because there is one case in which after the first pass we could need to
+     * reextract the data.  If there's initial padding before the first element, we put in
+     * the padding and then write that element.  If it turns out there's no second element,
+     * we didn't need the padding.  If we do need the data (no preflight), and the first element
+     * would have fit but for the padding, we need to reextract.  In this case (only) we
+     * adjust the parameters so padding is not added, and repeat.
+     */
+    do {
+        UChar* p=dest;
+        int32_t patPos=0; /* position in the pattern, used for non-substitution portions */
+        int32_t langLen=0; /* length of language substitution */
+        int32_t langPos=0; /* position in output of language substitution */
+        int32_t restLen=0; /* length of 'everything else' substitution */
+        int32_t restPos=0; /* position in output of 'everything else' substitution */
+        icu::LocalUEnumerationPointer kenum; /* keyword enumeration */
+
+        /* prefix of pattern, extremely likely to be empty */
+        if(sub0Pos) {
+            if(destCapacity >= sub0Pos) {
+                while (patPos < sub0Pos) {
+                    *p++ = pattern[patPos++];
+                }
+            } else {
+                patPos=sub0Pos;
+            }
+            length=sub0Pos;
+        } else {
+            length=0;
+        }
+
+        for(int32_t subi=0,resti=0;subi<2;) { /* iterate through patterns 0 and 1*/
+            UBool subdone = FALSE; /* set true when ready to move to next substitution */
+
+            /* prep p and cap for calls to get display components, pin cap to 0 since
+               they complain if cap is negative */
+            int32_t cap=destCapacity-length;
+            if (cap <= 0) {
+                cap=0;
+            } else {
+                p=dest+length;
+            }
+
+            if (subi == langi) { /* {0}*/
+                if(haveLang) {
+                    langPos=length;
+                    langLen=uloc_getDisplayLanguage(locale, displayLocale, p, cap, pErrorCode);
+                    length+=langLen;
+                    haveLang=langLen>0;
+                }
+                subdone=TRUE;
+            } else { /* {1} */
+                if(!haveRest) {
+                    subdone=TRUE;
+                } else {
+                    int32_t len; /* length of component (plus other stuff) we just fetched */
+                    switch(resti++) {
+                        case 0:
+                            restPos=length;
+                            len=uloc_getDisplayScriptInContext(locale, displayLocale, p, cap, pErrorCode);
+                            break;
+                        case 1:
+                            len=uloc_getDisplayCountry(locale, displayLocale, p, cap, pErrorCode);
+                            break;
+                        case 2:
+                            len=uloc_getDisplayVariant(locale, displayLocale, p, cap, pErrorCode);
+                            break;
+                        case 3:
+                            kenum.adoptInstead(uloc_openKeywords(locale, pErrorCode));
+                            U_FALLTHROUGH;
+                        default: {
+                            const char* kw=uenum_next(kenum.getAlias(), &len, pErrorCode);
+                            if (kw == NULL) {
+                                len=0; /* mark that we didn't add a component */
+                                subdone=TRUE;
+                            } else {
+                                /* incorporating this behavior into the loop made it even more complex,
+                                   so just special case it here */
+                                len = uloc_getDisplayKeyword(kw, displayLocale, p, cap, pErrorCode);
+                                if(len) {
+                                    if(len < cap) {
+                                        p[len]=0x3d; /* '=', assume we'll need it */
+                                    }
+                                    len+=1;
+
+                                    /* adjust for call to get keyword */
+                                    cap-=len;
+                                    if(cap <= 0) {
+                                        cap=0;
+                                    } else {
+                                        p+=len;
+                                    }
+                                }
+                                /* reset for call below */
+                                if(*pErrorCode == U_BUFFER_OVERFLOW_ERROR) {
+                                    *pErrorCode=U_ZERO_ERROR;
+                                }
+                                int32_t vlen = uloc_getDisplayKeywordValue(locale, kw, displayLocale,
+                                                                           p, cap, pErrorCode);
+                                if(len) {
+                                    if(vlen==0) {
+                                        --len; /* remove unneeded '=' */
+                                    }
+                                    /* restore cap and p to what they were at start */
+                                    cap=destCapacity-length;
+                                    if(cap <= 0) {
+                                        cap=0;
+                                    } else {
+                                        p=dest+length;
+                                    }
+                                }
+                                len+=vlen; /* total we added for key + '=' + value */
+                            }
+                        } break;
+                    } /* end switch */
+
+                    if (len>0) {
+                        /* we addeed a component, so add separator and write it if there's room. */
+                        if(len+sepLen<=cap) {
+                            const UChar * plimit = p + len;
+                            for (; p < plimit; p++) {
+                                if (*p == formatOpenParen) {
+                                    *p = formatReplaceOpenParen;
+                                } else if (*p == formatCloseParen) {
+                                    *p = formatReplaceCloseParen;
+                                }
+                            }
+                            for(int32_t i=0;i<sepLen;++i) {
+                                *p++=separator[i];
+                            }
+                        }
+                        length+=len+sepLen;
+                    } else if(subdone) {
+                        /* remove separator if we added it */
+                        if (length!=restPos) {
+                            length-=sepLen;
+                        }
+                        restLen=length-restPos;
+                        haveRest=restLen>0;
+                    }
+                }
+            }
+
+            if(*pErrorCode == U_BUFFER_OVERFLOW_ERROR) {
+                *pErrorCode=U_ZERO_ERROR;
+            }
+
+            if(subdone) {
+                if(haveLang && haveRest) {
+                    /* append internal portion of pattern, the first time,
+                       or last portion of pattern the second time */
+                    int32_t padLen;
+                    patPos+=subLen;
+                    padLen=(subi==0 ? sub1Pos : patLen)-patPos;
+                    if(length+padLen <= destCapacity) {
+                        p=dest+length;
+                        for(int32_t i=0;i<padLen;++i) {
+                            *p++=pattern[patPos++];
+                        }
+                    } else {
+                        patPos+=padLen;
+                    }
+                    length+=padLen;
+                } else if(subi==0) {
+                    /* don't have first component, reset for second component */
+                    sub0Pos=0;
+                    length=0;
+                } else if(length>0) {
+                    /* true length is the length of just the component we got. */
+                    length=haveLang?langLen:restLen;
+                    if(dest && sub0Pos!=0) {
+                        if (sub0Pos+length<=destCapacity) {
+                            /* first component not at start of result,
+                               but we have full component in buffer. */
+                            u_memmove(dest, dest+(haveLang?langPos:restPos), length);
+                        } else {
+                            /* would have fit, but didn't because of pattern prefix. */
+                            sub0Pos=0; /* stops initial padding (and a second retry,
+                                          so we won't end up here again) */
+                            retry=TRUE;
+                        }
+                    }
+                }
+
+                ++subi; /* move on to next substitution */
+            }
+        }
+    } while(retry);
+
+    return u_terminateUChars(dest, destCapacity, length, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayKeyword(const char* keyword,
+                       const char* displayLocale,
+                       UChar* dest,
+                       int32_t destCapacity,
+                       UErrorCode* status){
+
+    /* argument checking */
+    if(status==NULL || U_FAILURE(*status)) {
+        return 0;
+    }
+
+    if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+
+    /* pass itemKey=NULL to look for a top-level item */
+    return _getStringOrCopyKey(U_ICUDATA_LANG, displayLocale,
+                               _kKeys, NULL, 
+                               keyword, 
+                               keyword,      
+                               dest, destCapacity,
+                               status);
+
+}
+
+
+#define UCURRENCY_DISPLAY_NAME_INDEX 1
+
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayKeywordValue(   const char* locale,
+                               const char* keyword,
+                               const char* displayLocale,
+                               UChar* dest,
+                               int32_t destCapacity,
+                               UErrorCode* status){
+
+
+    /* argument checking */
+    if(status==NULL || U_FAILURE(*status)) {
+        return 0;
+    }
+
+    if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* get the keyword value */
+    CharString keywordValue;
+    {
+        CharStringByteSink sink(&keywordValue);
+        ulocimp_getKeywordValue(locale, keyword, sink, status);
+    }
+
+    /* 
+     * if the keyword is equal to currency .. then to get the display name 
+     * we need to do the fallback ourselves
+     */
+    if(uprv_stricmp(keyword, _kCurrency)==0){
+
+        int32_t dispNameLen = 0;
+        const UChar *dispName = NULL;
+
+        icu::LocalUResourceBundlePointer bundle(
+                ures_open(U_ICUDATA_CURR, displayLocale, status));
+        icu::LocalUResourceBundlePointer currencies(
+                ures_getByKey(bundle.getAlias(), _kCurrencies, NULL, status));
+        icu::LocalUResourceBundlePointer currency(
+                ures_getByKeyWithFallback(currencies.getAlias(), keywordValue.data(), NULL, status));
+
+        dispName = ures_getStringByIndex(currency.getAlias(), UCURRENCY_DISPLAY_NAME_INDEX, &dispNameLen, status);
+
+        if(U_FAILURE(*status)){
+            if(*status == U_MISSING_RESOURCE_ERROR){
+                /* we just want to write the value over if nothing is available */
+                *status = U_USING_DEFAULT_WARNING;
+            }else{
+                return 0;
+            }
+        }
+
+        /* now copy the dispName over if not NULL */
+        if(dispName != NULL){
+            if(dispNameLen <= destCapacity){
+                u_memcpy(dest, dispName, dispNameLen);
+                return u_terminateUChars(dest, destCapacity, dispNameLen, status);
+            }else{
+                *status = U_BUFFER_OVERFLOW_ERROR;
+                return dispNameLen;
+            }
+        }else{
+            /* we have not found the display name for the value .. just copy over */
+            if(keywordValue.length() <= destCapacity){
+                u_charsToUChars(keywordValue.data(), dest, keywordValue.length());
+                return u_terminateUChars(dest, destCapacity, keywordValue.length(), status);
+            }else{
+                 *status = U_BUFFER_OVERFLOW_ERROR;
+                return keywordValue.length();
+            }
+        }
+
+        
+    }else{
+
+        return _getStringOrCopyKey(U_ICUDATA_LANG, displayLocale,
+                                   _kTypes, keyword, 
+                                   keywordValue.data(),
+                                   keywordValue.data(),
+                                   dest, destCapacity,
+                                   status);
+    }
+}
diff --git a/thirdparty/icu4c/common/locdistance.cpp b/thirdparty/icu4c/common/locdistance.cpp
new file mode 100644
index 0000000000..ff8892791b
--- /dev/null
+++ b/thirdparty/icu4c/common/locdistance.cpp
@@ -0,0 +1,415 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// locdistance.cpp
+// created: 2019may08 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/localematcher.h"
+#include "unicode/locid.h"
+#include "unicode/uobject.h"
+#include "unicode/ures.h"
+#include "cstring.h"
+#include "locdistance.h"
+#include "loclikelysubtags.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "uinvchar.h"
+#include "umutex.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+/**
+ * Bit flag used on the last character of a subtag in the trie.
+ * Must be set consistently by the builder and the lookup code.
+ */
+constexpr int32_t END_OF_SUBTAG = 0x80;
+/** Distance value bit flag, set by the builder. */
+constexpr int32_t DISTANCE_SKIP_SCRIPT = 0x80;
+/** Distance value bit flag, set by trieNext(). */
+constexpr int32_t DISTANCE_IS_FINAL = 0x100;
+constexpr int32_t DISTANCE_IS_FINAL_OR_SKIP_SCRIPT = DISTANCE_IS_FINAL | DISTANCE_SKIP_SCRIPT;
+
+constexpr int32_t ABOVE_THRESHOLD = 100;
+
+// Indexes into array of distances.
+enum {
+    IX_DEF_LANG_DISTANCE,
+    IX_DEF_SCRIPT_DISTANCE,
+    IX_DEF_REGION_DISTANCE,
+    IX_MIN_REGION_DISTANCE,
+    IX_LIMIT
+};
+
+LocaleDistance *gLocaleDistance = nullptr;
+UInitOnce gInitOnce = U_INITONCE_INITIALIZER;
+
+UBool U_CALLCONV cleanup() {
+    delete gLocaleDistance;
+    gLocaleDistance = nullptr;
+    gInitOnce.reset();
+    return TRUE;
+}
+
+}  // namespace
+
+void U_CALLCONV LocaleDistance::initLocaleDistance(UErrorCode &errorCode) {
+    // This function is invoked only via umtx_initOnce().
+    U_ASSERT(gLocaleDistance == nullptr);
+    const XLikelySubtags &likely = *XLikelySubtags::getSingleton(errorCode);
+    if (U_FAILURE(errorCode)) { return; }
+    const LocaleDistanceData &data = likely.getDistanceData();
+    if (data.distanceTrieBytes == nullptr ||
+            data.regionToPartitions == nullptr || data.partitions == nullptr ||
+            // ok if no paradigms
+            data.distances == nullptr) {
+        errorCode = U_MISSING_RESOURCE_ERROR;
+        return;
+    }
+    gLocaleDistance = new LocaleDistance(data, likely);
+    if (gLocaleDistance == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE_DISTANCE, cleanup);
+}
+
+const LocaleDistance *LocaleDistance::getSingleton(UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    umtx_initOnce(gInitOnce, &LocaleDistance::initLocaleDistance, errorCode);
+    return gLocaleDistance;
+}
+
+LocaleDistance::LocaleDistance(const LocaleDistanceData &data, const XLikelySubtags &likely) :
+        likelySubtags(likely),
+        trie(data.distanceTrieBytes),
+        regionToPartitionsIndex(data.regionToPartitions), partitionArrays(data.partitions),
+        paradigmLSRs(data.paradigms), paradigmLSRsLength(data.paradigmsLength),
+        defaultLanguageDistance(data.distances[IX_DEF_LANG_DISTANCE]),
+        defaultScriptDistance(data.distances[IX_DEF_SCRIPT_DISTANCE]),
+        defaultRegionDistance(data.distances[IX_DEF_REGION_DISTANCE]),
+        minRegionDistance(data.distances[IX_MIN_REGION_DISTANCE]) {
+    // For the default demotion value, use the
+    // default region distance between unrelated Englishes.
+    // Thus, unless demotion is turned off,
+    // a mere region difference for one desired locale
+    // is as good as a perfect match for the next following desired locale.
+    // As of CLDR 36, we have <languageMatch desired="en_*_*" supported="en_*_*" distance="5"/>.
+    LSR en("en", "Latn", "US", LSR::EXPLICIT_LSR);
+    LSR enGB("en", "Latn", "GB", LSR::EXPLICIT_LSR);
+    const LSR *p_enGB = &enGB;
+    int32_t indexAndDistance = getBestIndexAndDistance(en, &p_enGB, 1,
+            shiftDistance(50), ULOCMATCH_FAVOR_LANGUAGE, ULOCMATCH_DIRECTION_WITH_ONE_WAY);
+    defaultDemotionPerDesiredLocale  = getDistanceFloor(indexAndDistance);
+}
+
+int32_t LocaleDistance::getBestIndexAndDistance(
+        const LSR &desired,
+        const LSR **supportedLSRs, int32_t supportedLSRsLength,
+        int32_t shiftedThreshold,
+        ULocMatchFavorSubtag favorSubtag, ULocMatchDirection direction) const {
+    BytesTrie iter(trie);
+    // Look up the desired language only once for all supported LSRs.
+    // Its "distance" is either a match point value of 0, or a non-match negative value.
+    // Note: The data builder verifies that there are no <*, supported> or <desired, *> rules.
+    int32_t desLangDistance = trieNext(iter, desired.language, false);
+    uint64_t desLangState = desLangDistance >= 0 && supportedLSRsLength > 1 ? iter.getState64() : 0;
+    // Index of the supported LSR with the lowest distance.
+    int32_t bestIndex = -1;
+    // Cached lookup info from XLikelySubtags.compareLikely().
+    int32_t bestLikelyInfo = -1;
+    for (int32_t slIndex = 0; slIndex < supportedLSRsLength; ++slIndex) {
+        const LSR &supported = *supportedLSRs[slIndex];
+        bool star = false;
+        int32_t distance = desLangDistance;
+        if (distance >= 0) {
+            U_ASSERT((distance & DISTANCE_IS_FINAL) == 0);
+            if (slIndex != 0) {
+                iter.resetToState64(desLangState);
+            }
+            distance = trieNext(iter, supported.language, true);
+        }
+        // Note: The data builder verifies that there are no rules with "any" (*) language and
+        // real (non *) script or region subtags.
+        // This means that if the lookup for either language fails we can use
+        // the default distances without further lookups.
+        int32_t flags;
+        if (distance >= 0) {
+            flags = distance & DISTANCE_IS_FINAL_OR_SKIP_SCRIPT;
+            distance &= ~DISTANCE_IS_FINAL_OR_SKIP_SCRIPT;
+        } else {  // <*, *>
+            if (uprv_strcmp(desired.language, supported.language) == 0) {
+                distance = 0;
+            } else {
+                distance = defaultLanguageDistance;
+            }
+            flags = 0;
+            star = true;
+        }
+        U_ASSERT(0 <= distance && distance <= 100);
+        // Round up the shifted threshold (if fraction bits are not 0)
+        // for comparison with un-shifted distances until we need fraction bits.
+        // (If we simply shifted non-zero fraction bits away, then we might ignore a language
+        // when it's really still a micro distance below the threshold.)
+        int32_t roundedThreshold = (shiftedThreshold + DISTANCE_FRACTION_MASK) >> DISTANCE_SHIFT;
+        // We implement "favor subtag" by reducing the language subtag distance
+        // (unscientifically reducing it to a quarter of the normal value),
+        // so that the script distance is relatively more important.
+        // For example, given a default language distance of 80, we reduce it to 20,
+        // which is below the default threshold of 50, which is the default script distance.
+        if (favorSubtag == ULOCMATCH_FAVOR_SCRIPT) {
+            distance >>= 2;
+        }
+        // Let distance == roundedThreshold pass until the tie-breaker logic
+        // at the end of the loop.
+        if (distance > roundedThreshold) {
+            continue;
+        }
+
+        int32_t scriptDistance;
+        if (star || flags != 0) {
+            if (uprv_strcmp(desired.script, supported.script) == 0) {
+                scriptDistance = 0;
+            } else {
+                scriptDistance = defaultScriptDistance;
+            }
+        } else {
+            scriptDistance = getDesSuppScriptDistance(iter, iter.getState64(),
+                    desired.script, supported.script);
+            flags = scriptDistance & DISTANCE_IS_FINAL;
+            scriptDistance &= ~DISTANCE_IS_FINAL;
+        }
+        distance += scriptDistance;
+        if (distance > roundedThreshold) {
+            continue;
+        }
+
+        if (uprv_strcmp(desired.region, supported.region) == 0) {
+            // regionDistance = 0
+        } else if (star || (flags & DISTANCE_IS_FINAL) != 0) {
+            distance += defaultRegionDistance;
+        } else {
+            int32_t remainingThreshold = roundedThreshold - distance;
+            if (minRegionDistance > remainingThreshold) {
+                continue;
+            }
+
+            // From here on we know the regions are not equal.
+            // Map each region to zero or more partitions. (zero = one non-matching string)
+            // (Each array of single-character partition strings is encoded as one string.)
+            // If either side has more than one, then we find the maximum distance.
+            // This could be optimized by adding some more structure, but probably not worth it.
+            distance += getRegionPartitionsDistance(
+                    iter, iter.getState64(),
+                    partitionsForRegion(desired),
+                    partitionsForRegion(supported),
+                    remainingThreshold);
+        }
+        int32_t shiftedDistance = shiftDistance(distance);
+        if (shiftedDistance == 0) {
+            // Distinguish between equivalent but originally unequal locales via an
+            // additional micro distance.
+            shiftedDistance |= (desired.flags ^ supported.flags);
+            if (shiftedDistance < shiftedThreshold) {
+                if (direction != ULOCMATCH_DIRECTION_ONLY_TWO_WAY ||
+                        // Is there also a match when we swap desired/supported?
+                        isMatch(supported, desired, shiftedThreshold, favorSubtag)) {
+                    if (shiftedDistance == 0) {
+                        return slIndex << INDEX_SHIFT;
+                    }
+                    bestIndex = slIndex;
+                    shiftedThreshold = shiftedDistance;
+                    bestLikelyInfo = -1;
+                }
+            }
+        } else {
+            if (shiftedDistance < shiftedThreshold) {
+                if (direction != ULOCMATCH_DIRECTION_ONLY_TWO_WAY ||
+                        // Is there also a match when we swap desired/supported?
+                        isMatch(supported, desired, shiftedThreshold, favorSubtag)) {
+                    bestIndex = slIndex;
+                    shiftedThreshold = shiftedDistance;
+                    bestLikelyInfo = -1;
+                }
+            } else if (shiftedDistance == shiftedThreshold && bestIndex >= 0) {
+                if (direction != ULOCMATCH_DIRECTION_ONLY_TWO_WAY ||
+                        // Is there also a match when we swap desired/supported?
+                        isMatch(supported, desired, shiftedThreshold, favorSubtag)) {
+                    bestLikelyInfo = likelySubtags.compareLikely(
+                            supported, *supportedLSRs[bestIndex], bestLikelyInfo);
+                    if ((bestLikelyInfo & 1) != 0) {
+                        // This supported locale matches as well as the previous best match,
+                        // and neither matches perfectly,
+                        // but this one is "more likely" (has more-default subtags).
+                        bestIndex = slIndex;
+                    }
+                }
+            }
+        }
+    }
+    return bestIndex >= 0 ?
+            (bestIndex << INDEX_SHIFT) | shiftedThreshold :
+            INDEX_NEG_1 | shiftDistance(ABOVE_THRESHOLD);
+}
+
+int32_t LocaleDistance::getDesSuppScriptDistance(
+        BytesTrie &iter, uint64_t startState, const char *desired, const char *supported) {
+    // Note: The data builder verifies that there are no <*, supported> or <desired, *> rules.
+    int32_t distance = trieNext(iter, desired, false);
+    if (distance >= 0) {
+        distance = trieNext(iter, supported, true);
+    }
+    if (distance < 0) {
+        UStringTrieResult result = iter.resetToState64(startState).next(u'*');  // <*, *>
+        U_ASSERT(USTRINGTRIE_HAS_VALUE(result));
+        if (uprv_strcmp(desired, supported) == 0) {
+            distance = 0;  // same script
+        } else {
+            distance = iter.getValue();
+            U_ASSERT(distance >= 0);
+        }
+        if (result == USTRINGTRIE_FINAL_VALUE) {
+            distance |= DISTANCE_IS_FINAL;
+        }
+    }
+    return distance;
+}
+
+int32_t LocaleDistance::getRegionPartitionsDistance(
+        BytesTrie &iter, uint64_t startState,
+        const char *desiredPartitions, const char *supportedPartitions, int32_t threshold) {
+    char desired = *desiredPartitions++;
+    char supported = *supportedPartitions++;
+    U_ASSERT(desired != 0 && supported != 0);
+    // See if we have single desired/supported partitions, from NUL-terminated
+    // partition strings without explicit length.
+    bool suppLengthGt1 = *supportedPartitions != 0;  // gt1: more than 1 character
+    // equivalent to: if (desLength == 1 && suppLength == 1)
+    if (*desiredPartitions == 0 && !suppLengthGt1) {
+        // Fastpath for single desired/supported partitions.
+        UStringTrieResult result = iter.next(uprv_invCharToAscii(desired) | END_OF_SUBTAG);
+        if (USTRINGTRIE_HAS_NEXT(result)) {
+            result = iter.next(uprv_invCharToAscii(supported) | END_OF_SUBTAG);
+            if (USTRINGTRIE_HAS_VALUE(result)) {
+                return iter.getValue();
+            }
+        }
+        return getFallbackRegionDistance(iter, startState);
+    }
+
+    const char *supportedStart = supportedPartitions - 1;  // for restart of inner loop
+    int32_t regionDistance = 0;
+    // Fall back to * only once, not for each pair of partition strings.
+    bool star = false;
+    for (;;) {
+        // Look up each desired-partition string only once,
+        // not for each (desired, supported) pair.
+        UStringTrieResult result = iter.next(uprv_invCharToAscii(desired) | END_OF_SUBTAG);
+        if (USTRINGTRIE_HAS_NEXT(result)) {
+            uint64_t desState = suppLengthGt1 ? iter.getState64() : 0;
+            for (;;) {
+                result = iter.next(uprv_invCharToAscii(supported) | END_OF_SUBTAG);
+                int32_t d;
+                if (USTRINGTRIE_HAS_VALUE(result)) {
+                    d = iter.getValue();
+                } else if (star) {
+                    d = 0;
+                } else {
+                    d = getFallbackRegionDistance(iter, startState);
+                    star = true;
+                }
+                if (d > threshold) {
+                    return d;
+                } else if (regionDistance < d) {
+                    regionDistance = d;
+                }
+                if ((supported = *supportedPartitions++) != 0) {
+                    iter.resetToState64(desState);
+                } else {
+                    break;
+                }
+            }
+        } else if (!star) {
+            int32_t d = getFallbackRegionDistance(iter, startState);
+            if (d > threshold) {
+                return d;
+            } else if (regionDistance < d) {
+                regionDistance = d;
+            }
+            star = true;
+        }
+        if ((desired = *desiredPartitions++) != 0) {
+            iter.resetToState64(startState);
+            supportedPartitions = supportedStart;
+            supported = *supportedPartitions++;
+        } else {
+            break;
+        }
+    }
+    return regionDistance;
+}
+
+int32_t LocaleDistance::getFallbackRegionDistance(BytesTrie &iter, uint64_t startState) {
+#if U_DEBUG
+    UStringTrieResult result =
+#endif
+    iter.resetToState64(startState).next(u'*');  // <*, *>
+    U_ASSERT(USTRINGTRIE_HAS_VALUE(result));
+    int32_t distance = iter.getValue();
+    U_ASSERT(distance >= 0);
+    return distance;
+}
+
+int32_t LocaleDistance::trieNext(BytesTrie &iter, const char *s, bool wantValue) {
+    uint8_t c;
+    if ((c = *s) == 0) {
+        return -1;  // no empty subtags in the distance data
+    }
+    for (;;) {
+        c = uprv_invCharToAscii(c);
+        // EBCDIC: If *s is not an invariant character,
+        // then c is now 0 and will simply not match anything, which is harmless.
+        uint8_t next = *++s;
+        if (next != 0) {
+            if (!USTRINGTRIE_HAS_NEXT(iter.next(c))) {
+                return -1;
+            }
+        } else {
+            // last character of this subtag
+            UStringTrieResult result = iter.next(c | END_OF_SUBTAG);
+            if (wantValue) {
+                if (USTRINGTRIE_HAS_VALUE(result)) {
+                    int32_t value = iter.getValue();
+                    if (result == USTRINGTRIE_FINAL_VALUE) {
+                        value |= DISTANCE_IS_FINAL;
+                    }
+                    return value;
+                }
+            } else {
+                if (USTRINGTRIE_HAS_NEXT(result)) {
+                    return 0;
+                }
+            }
+            return -1;
+        }
+        c = next;
+    }
+}
+
+UBool LocaleDistance::isParadigmLSR(const LSR &lsr) const {
+    // Linear search for a very short list (length 6 as of 2019),
+    // because we look for equivalence not equality, and
+    // because it's easy.
+    // If there are many paradigm LSRs we should use a hash set
+    // with custom comparator and hasher.
+    U_ASSERT(paradigmLSRsLength <= 15);
+    for (int32_t i = 0; i < paradigmLSRsLength; ++i) {
+        if (lsr.isEquivalentTo(paradigmLSRs[i])) { return true; }
+    }
+    return false;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/locdistance.h b/thirdparty/icu4c/common/locdistance.h
new file mode 100644
index 0000000000..51b777e627
--- /dev/null
+++ b/thirdparty/icu4c/common/locdistance.h
@@ -0,0 +1,151 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// locdistance.h
+// created: 2019may08 Markus W. Scherer
+
+#ifndef __LOCDISTANCE_H__
+#define __LOCDISTANCE_H__
+
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/localematcher.h"
+#include "unicode/locid.h"
+#include "unicode/uobject.h"
+#include "lsr.h"
+
+U_NAMESPACE_BEGIN
+
+struct LocaleDistanceData;
+
+/**
+ * Offline-built data for LocaleMatcher.
+ * Mostly but not only the data for mapping locales to their maximized forms.
+ */
+class LocaleDistance final : public UMemory {
+public:
+    static const LocaleDistance *getSingleton(UErrorCode &errorCode);
+
+    static int32_t shiftDistance(int32_t distance) {
+        return distance << DISTANCE_SHIFT;
+    }
+
+    static int32_t getShiftedDistance(int32_t indexAndDistance) {
+        return indexAndDistance & DISTANCE_MASK;
+    }
+
+    static double getDistanceDouble(int32_t indexAndDistance) {
+        double shiftedDistance = getShiftedDistance(indexAndDistance);
+        return shiftedDistance / (1 << DISTANCE_SHIFT);
+    }
+
+    static int32_t getDistanceFloor(int32_t indexAndDistance) {
+        return (indexAndDistance & DISTANCE_MASK) >> DISTANCE_SHIFT;
+    }
+
+    static int32_t getIndex(int32_t indexAndDistance) {
+        // assert indexAndDistance >= 0;
+        return indexAndDistance >> INDEX_SHIFT;
+    }
+
+    /**
+     * Finds the supported LSR with the smallest distance from the desired one.
+     * Equivalent LSR subtags must be normalized into a canonical form.
+     *
+     * <p>Returns the index of the lowest-distance supported LSR in the high bits
+     * (negative if none has a distance below the threshold),
+     * and its distance (0..ABOVE_THRESHOLD) in the low bits.
+     */
+    int32_t getBestIndexAndDistance(const LSR &desired,
+                                    const LSR **supportedLSRs, int32_t supportedLSRsLength,
+                                    int32_t shiftedThreshold,
+                                    ULocMatchFavorSubtag favorSubtag,
+                                    ULocMatchDirection direction) const;
+
+    UBool isParadigmLSR(const LSR &lsr) const;
+
+    int32_t getDefaultScriptDistance() const {
+        return defaultScriptDistance;
+    }
+
+    int32_t getDefaultDemotionPerDesiredLocale() const {
+        return defaultDemotionPerDesiredLocale;
+    }
+
+private:
+    // The distance is shifted left to gain some fraction bits.
+    static constexpr int32_t DISTANCE_SHIFT = 3;
+    static constexpr int32_t DISTANCE_FRACTION_MASK = 7;
+    // 7 bits for 0..100
+    static constexpr int32_t DISTANCE_INT_SHIFT = 7;
+    static constexpr int32_t INDEX_SHIFT = DISTANCE_INT_SHIFT + DISTANCE_SHIFT;
+    static constexpr int32_t DISTANCE_MASK = 0x3ff;
+    // tic constexpr int32_t MAX_INDEX = 0x1fffff;  // avoids sign bit
+    static constexpr int32_t INDEX_NEG_1 = 0xfffffc00;
+
+    LocaleDistance(const LocaleDistanceData &data, const XLikelySubtags &likely);
+    LocaleDistance(const LocaleDistance &other) = delete;
+    LocaleDistance &operator=(const LocaleDistance &other) = delete;
+
+    static void initLocaleDistance(UErrorCode &errorCode);
+
+    UBool isMatch(const LSR &desired, const LSR &supported,
+                  int32_t shiftedThreshold, ULocMatchFavorSubtag favorSubtag) const {
+        const LSR *pSupp = &supported;
+        return getBestIndexAndDistance(
+            desired, &pSupp, 1,
+            shiftedThreshold, favorSubtag, ULOCMATCH_DIRECTION_WITH_ONE_WAY) >= 0;
+    }
+
+    static int32_t getDesSuppScriptDistance(BytesTrie &iter, uint64_t startState,
+                                            const char *desired, const char *supported);
+
+    static int32_t getRegionPartitionsDistance(
+        BytesTrie &iter, uint64_t startState,
+        const char *desiredPartitions, const char *supportedPartitions,
+        int32_t threshold);
+
+    static int32_t getFallbackRegionDistance(BytesTrie &iter, uint64_t startState);
+
+    static int32_t trieNext(BytesTrie &iter, const char *s, bool wantValue);
+
+    const char *partitionsForRegion(const LSR &lsr) const {
+        // ill-formed region -> one non-matching string
+        int32_t pIndex = regionToPartitionsIndex[lsr.regionIndex];
+        return partitionArrays[pIndex];
+    }
+
+    int32_t getDefaultRegionDistance() const {
+        return defaultRegionDistance;
+    }
+
+    const XLikelySubtags &likelySubtags;
+
+    // The trie maps each dlang+slang+dscript+sscript+dregion+sregion
+    // (encoded in ASCII with bit 7 set on the last character of each subtag) to a distance.
+    // There is also a trie value for each subsequence of whole subtags.
+    // One '*' is used for a (desired, supported) pair of "und", "Zzzz"/"", or "ZZ"/"".
+    BytesTrie trie;
+
+    /**
+     * Maps each region to zero or more single-character partitions.
+     */
+    const uint8_t *regionToPartitionsIndex;
+    const char **partitionArrays;
+
+    /**
+     * Used to get the paradigm region for a cluster, if there is one.
+     */
+    const LSR *paradigmLSRs;
+    int32_t paradigmLSRsLength;
+
+    int32_t defaultLanguageDistance;
+    int32_t defaultScriptDistance;
+    int32_t defaultRegionDistance;
+    int32_t minRegionDistance;
+    int32_t defaultDemotionPerDesiredLocale;
+};
+
+U_NAMESPACE_END
+
+#endif  // __LOCDISTANCE_H__
diff --git a/thirdparty/icu4c/common/locdspnm.cpp b/thirdparty/icu4c/common/locdspnm.cpp
new file mode 100644
index 0000000000..43334f5196
--- /dev/null
+++ b/thirdparty/icu4c/common/locdspnm.cpp
@@ -0,0 +1,1110 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2010-2016, International Business Machines Corporation and
+* others. All Rights Reserved.
+*******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/locdspnm.h"
+#include "unicode/simpleformatter.h"
+#include "unicode/ucasemap.h"
+#include "unicode/ures.h"
+#include "unicode/udisplaycontext.h"
+#include "unicode/brkiter.h"
+#include "unicode/ucurr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "ulocimp.h"
+#include "umutex.h"
+#include "ureslocs.h"
+#include "uresimp.h"
+
+#include <stdarg.h>
+
+/**
+ * Concatenate a number of null-terminated strings to buffer, leaving a
+ * null-terminated string.  The last argument should be the null pointer.
+ * Return the length of the string in the buffer, not counting the trailing
+ * null.  Return -1 if there is an error (buffer is null, or buflen < 1).
+ */
+static int32_t ncat(char *buffer, uint32_t buflen, ...) {
+  va_list args;
+  char *str;
+  char *p = buffer;
+  const char* e = buffer + buflen - 1;
+
+  if (buffer == NULL || buflen < 1) {
+    return -1;
+  }
+
+  va_start(args, buflen);
+  while ((str = va_arg(args, char *)) != 0) {
+    char c;
+    while (p != e && (c = *str++) != 0) {
+      *p++ = c;
+    }
+  }
+  *p = 0;
+  va_end(args);
+
+  return static_cast<int32_t>(p - buffer);
+}
+
+U_NAMESPACE_BEGIN
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+// Access resource data for locale components.
+// Wrap code in uloc.c for now.
+class ICUDataTable {
+    const char* path;
+    Locale locale;
+
+public:
+    ICUDataTable(const char* path, const Locale& locale);
+    ~ICUDataTable();
+
+    const Locale& getLocale();
+
+    UnicodeString& get(const char* tableKey, const char* itemKey,
+                        UnicodeString& result) const;
+    UnicodeString& get(const char* tableKey, const char* subTableKey, const char* itemKey,
+                        UnicodeString& result) const;
+
+    UnicodeString& getNoFallback(const char* tableKey, const char* itemKey,
+                                UnicodeString &result) const;
+    UnicodeString& getNoFallback(const char* tableKey, const char* subTableKey, const char* itemKey,
+                                UnicodeString &result) const;
+};
+
+inline UnicodeString &
+ICUDataTable::get(const char* tableKey, const char* itemKey, UnicodeString& result) const {
+    return get(tableKey, NULL, itemKey, result);
+}
+
+inline UnicodeString &
+ICUDataTable::getNoFallback(const char* tableKey, const char* itemKey, UnicodeString& result) const {
+    return getNoFallback(tableKey, NULL, itemKey, result);
+}
+
+ICUDataTable::ICUDataTable(const char* path, const Locale& locale)
+    : path(NULL), locale(Locale::getRoot())
+{
+  if (path) {
+    int32_t len = static_cast<int32_t>(uprv_strlen(path));
+    this->path = (const char*) uprv_malloc(len + 1);
+    if (this->path) {
+      uprv_strcpy((char *)this->path, path);
+      this->locale = locale;
+    }
+  }
+}
+
+ICUDataTable::~ICUDataTable() {
+  if (path) {
+    uprv_free((void*) path);
+    path = NULL;
+  }
+}
+
+const Locale&
+ICUDataTable::getLocale() {
+  return locale;
+}
+
+UnicodeString &
+ICUDataTable::get(const char* tableKey, const char* subTableKey, const char* itemKey,
+                  UnicodeString &result) const {
+  UErrorCode status = U_ZERO_ERROR;
+  int32_t len = 0;
+
+  const UChar *s = uloc_getTableStringWithFallback(path, locale.getName(),
+                                                   tableKey, subTableKey, itemKey,
+                                                   &len, &status);
+  if (U_SUCCESS(status) && len > 0) {
+    return result.setTo(s, len);
+  }
+  return result.setTo(UnicodeString(itemKey, -1, US_INV));
+}
+
+UnicodeString &
+ICUDataTable::getNoFallback(const char* tableKey, const char* subTableKey, const char* itemKey,
+                            UnicodeString& result) const {
+  UErrorCode status = U_ZERO_ERROR;
+  int32_t len = 0;
+
+  const UChar *s = uloc_getTableStringWithFallback(path, locale.getName(),
+                                                   tableKey, subTableKey, itemKey,
+                                                   &len, &status);
+  if (U_SUCCESS(status)) {
+    return result.setTo(s, len);
+  }
+
+  result.setToBogus();
+  return result;
+}
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+LocaleDisplayNames::~LocaleDisplayNames() {}
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+#if 0  // currently unused
+
+class DefaultLocaleDisplayNames : public LocaleDisplayNames {
+  UDialectHandling dialectHandling;
+
+public:
+  // constructor
+  DefaultLocaleDisplayNames(UDialectHandling dialectHandling);
+
+  virtual ~DefaultLocaleDisplayNames();
+
+  virtual const Locale& getLocale() const;
+  virtual UDialectHandling getDialectHandling() const;
+
+  virtual UnicodeString& localeDisplayName(const Locale& locale,
+                                           UnicodeString& result) const;
+  virtual UnicodeString& localeDisplayName(const char* localeId,
+                                           UnicodeString& result) const;
+  virtual UnicodeString& languageDisplayName(const char* lang,
+                                             UnicodeString& result) const;
+  virtual UnicodeString& scriptDisplayName(const char* script,
+                                           UnicodeString& result) const;
+  virtual UnicodeString& scriptDisplayName(UScriptCode scriptCode,
+                                           UnicodeString& result) const;
+  virtual UnicodeString& regionDisplayName(const char* region,
+                                           UnicodeString& result) const;
+  virtual UnicodeString& variantDisplayName(const char* variant,
+                                            UnicodeString& result) const;
+  virtual UnicodeString& keyDisplayName(const char* key,
+                                        UnicodeString& result) const;
+  virtual UnicodeString& keyValueDisplayName(const char* key,
+                                             const char* value,
+                                             UnicodeString& result) const;
+};
+
+DefaultLocaleDisplayNames::DefaultLocaleDisplayNames(UDialectHandling dialectHandling)
+    : dialectHandling(dialectHandling) {
+}
+
+DefaultLocaleDisplayNames::~DefaultLocaleDisplayNames() {
+}
+
+const Locale&
+DefaultLocaleDisplayNames::getLocale() const {
+  return Locale::getRoot();
+}
+
+UDialectHandling
+DefaultLocaleDisplayNames::getDialectHandling() const {
+  return dialectHandling;
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::localeDisplayName(const Locale& locale,
+                                             UnicodeString& result) const {
+  return result = UnicodeString(locale.getName(), -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::localeDisplayName(const char* localeId,
+                                             UnicodeString& result) const {
+  return result = UnicodeString(localeId, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::languageDisplayName(const char* lang,
+                                               UnicodeString& result) const {
+  return result = UnicodeString(lang, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::scriptDisplayName(const char* script,
+                                             UnicodeString& result) const {
+  return result = UnicodeString(script, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::scriptDisplayName(UScriptCode scriptCode,
+                                             UnicodeString& result) const {
+  const char* name = uscript_getName(scriptCode);
+  if (name) {
+    return result = UnicodeString(name, -1, US_INV);
+  }
+  return result.remove();
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::regionDisplayName(const char* region,
+                                             UnicodeString& result) const {
+  return result = UnicodeString(region, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::variantDisplayName(const char* variant,
+                                              UnicodeString& result) const {
+  return result = UnicodeString(variant, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::keyDisplayName(const char* key,
+                                          UnicodeString& result) const {
+  return result = UnicodeString(key, -1, US_INV);
+}
+
+UnicodeString&
+DefaultLocaleDisplayNames::keyValueDisplayName(const char* /* key */,
+                                               const char* value,
+                                               UnicodeString& result) const {
+  return result = UnicodeString(value, -1, US_INV);
+}
+
+#endif  // currently unused class DefaultLocaleDisplayNames
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+class LocaleDisplayNamesImpl : public LocaleDisplayNames {
+    Locale locale;
+    UDialectHandling dialectHandling;
+    ICUDataTable langData;
+    ICUDataTable regionData;
+    SimpleFormatter separatorFormat;
+    SimpleFormatter format;
+    SimpleFormatter keyTypeFormat;
+    UDisplayContext capitalizationContext;
+#if !UCONFIG_NO_BREAK_ITERATION
+    BreakIterator* capitalizationBrkIter;
+#else
+    UObject* capitalizationBrkIter;
+#endif
+    UnicodeString formatOpenParen;
+    UnicodeString formatReplaceOpenParen;
+    UnicodeString formatCloseParen;
+    UnicodeString formatReplaceCloseParen;
+    UDisplayContext nameLength;
+    UDisplayContext substitute;
+
+    // Constants for capitalization context usage types.
+    enum CapContextUsage {
+        kCapContextUsageLanguage,
+        kCapContextUsageScript,
+        kCapContextUsageTerritory,
+        kCapContextUsageVariant,
+        kCapContextUsageKey,
+        kCapContextUsageKeyValue,
+        kCapContextUsageCount
+    };
+    // Capitalization transforms. For each usage type, indicates whether to titlecase for
+    // the context specified in capitalizationContext (which we know at construction time)
+     UBool fCapitalization[kCapContextUsageCount];
+
+public:
+    // constructor
+    LocaleDisplayNamesImpl(const Locale& locale, UDialectHandling dialectHandling);
+    LocaleDisplayNamesImpl(const Locale& locale, UDisplayContext *contexts, int32_t length);
+    virtual ~LocaleDisplayNamesImpl();
+
+    virtual const Locale& getLocale() const;
+    virtual UDialectHandling getDialectHandling() const;
+    virtual UDisplayContext getContext(UDisplayContextType type) const;
+
+    virtual UnicodeString& localeDisplayName(const Locale& locale,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& localeDisplayName(const char* localeId,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& languageDisplayName(const char* lang,
+                                               UnicodeString& result) const;
+    virtual UnicodeString& scriptDisplayName(const char* script,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& scriptDisplayName(UScriptCode scriptCode,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& regionDisplayName(const char* region,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& variantDisplayName(const char* variant,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& keyDisplayName(const char* key,
+                                                UnicodeString& result) const;
+    virtual UnicodeString& keyValueDisplayName(const char* key,
+                                                const char* value,
+                                                UnicodeString& result) const;
+private:
+    UnicodeString& localeIdName(const char* localeId,
+                                UnicodeString& result, bool substitute) const;
+    UnicodeString& appendWithSep(UnicodeString& buffer, const UnicodeString& src) const;
+    UnicodeString& adjustForUsageAndContext(CapContextUsage usage, UnicodeString& result) const;
+    UnicodeString& scriptDisplayName(const char* script, UnicodeString& result, UBool skipAdjust) const;
+    UnicodeString& regionDisplayName(const char* region, UnicodeString& result, UBool skipAdjust) const;
+    UnicodeString& variantDisplayName(const char* variant, UnicodeString& result, UBool skipAdjust) const;
+    UnicodeString& keyDisplayName(const char* key, UnicodeString& result, UBool skipAdjust) const;
+    UnicodeString& keyValueDisplayName(const char* key, const char* value,
+                                        UnicodeString& result, UBool skipAdjust) const;
+    void initialize(void);
+
+    struct CapitalizationContextSink;
+};
+
+LocaleDisplayNamesImpl::LocaleDisplayNamesImpl(const Locale& locale,
+                                               UDialectHandling dialectHandling)
+    : dialectHandling(dialectHandling)
+    , langData(U_ICUDATA_LANG, locale)
+    , regionData(U_ICUDATA_REGION, locale)
+    , capitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
+    , capitalizationBrkIter(NULL)
+    , nameLength(UDISPCTX_LENGTH_FULL)
+    , substitute(UDISPCTX_SUBSTITUTE)
+{
+    initialize();
+}
+
+LocaleDisplayNamesImpl::LocaleDisplayNamesImpl(const Locale& locale,
+                                               UDisplayContext *contexts, int32_t length)
+    : dialectHandling(ULDN_STANDARD_NAMES)
+    , langData(U_ICUDATA_LANG, locale)
+    , regionData(U_ICUDATA_REGION, locale)
+    , capitalizationContext(UDISPCTX_CAPITALIZATION_NONE)
+    , capitalizationBrkIter(NULL)
+    , nameLength(UDISPCTX_LENGTH_FULL)
+    , substitute(UDISPCTX_SUBSTITUTE)
+{
+    while (length-- > 0) {
+        UDisplayContext value = *contexts++;
+        UDisplayContextType selector = (UDisplayContextType)((uint32_t)value >> 8);
+        switch (selector) {
+            case UDISPCTX_TYPE_DIALECT_HANDLING:
+                dialectHandling = (UDialectHandling)value;
+                break;
+            case UDISPCTX_TYPE_CAPITALIZATION:
+                capitalizationContext = value;
+                break;
+            case UDISPCTX_TYPE_DISPLAY_LENGTH:
+                nameLength = value;
+                break;
+            case UDISPCTX_TYPE_SUBSTITUTE_HANDLING:
+                substitute = value;
+                break;
+            default:
+                break;
+        }
+    }
+    initialize();
+}
+
+struct LocaleDisplayNamesImpl::CapitalizationContextSink : public ResourceSink {
+    UBool hasCapitalizationUsage;
+    LocaleDisplayNamesImpl& parent;
+
+    CapitalizationContextSink(LocaleDisplayNamesImpl& _parent)
+      : hasCapitalizationUsage(FALSE), parent(_parent) {}
+    virtual ~CapitalizationContextSink();
+
+    virtual void put(const char *key, ResourceValue &value, UBool /*noFallback*/,
+            UErrorCode &errorCode) {
+        ResourceTable contexts = value.getTable(errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        for (int i = 0; contexts.getKeyAndValue(i, key, value); ++i) {
+
+            CapContextUsage usageEnum;
+            if (uprv_strcmp(key, "key") == 0) {
+                usageEnum = kCapContextUsageKey;
+            } else if (uprv_strcmp(key, "keyValue") == 0) {
+                usageEnum = kCapContextUsageKeyValue;
+            } else if (uprv_strcmp(key, "languages") == 0) {
+                usageEnum = kCapContextUsageLanguage;
+            } else if (uprv_strcmp(key, "script") == 0) {
+                usageEnum = kCapContextUsageScript;
+            } else if (uprv_strcmp(key, "territory") == 0) {
+                usageEnum = kCapContextUsageTerritory;
+            } else if (uprv_strcmp(key, "variant") == 0) {
+                usageEnum = kCapContextUsageVariant;
+            } else {
+                continue;
+            }
+
+            int32_t len = 0;
+            const int32_t* intVector = value.getIntVector(len, errorCode);
+            if (U_FAILURE(errorCode)) { return; }
+            if (len < 2) { continue; }
+
+            int32_t titlecaseInt = (parent.capitalizationContext == UDISPCTX_CAPITALIZATION_FOR_UI_LIST_OR_MENU) ? intVector[0] : intVector[1];
+            if (titlecaseInt == 0) { continue; }
+
+            parent.fCapitalization[usageEnum] = TRUE;
+            hasCapitalizationUsage = TRUE;
+        }
+    }
+};
+
+// Virtual destructors must be defined out of line.
+LocaleDisplayNamesImpl::CapitalizationContextSink::~CapitalizationContextSink() {}
+
+void
+LocaleDisplayNamesImpl::initialize(void) {
+    LocaleDisplayNamesImpl *nonConstThis = (LocaleDisplayNamesImpl *)this;
+    nonConstThis->locale = langData.getLocale() == Locale::getRoot()
+        ? regionData.getLocale()
+        : langData.getLocale();
+
+    UnicodeString sep;
+    langData.getNoFallback("localeDisplayPattern", "separator", sep);
+    if (sep.isBogus()) {
+        sep = UnicodeString("{0}, {1}", -1, US_INV);
+    }
+    UErrorCode status = U_ZERO_ERROR;
+    separatorFormat.applyPatternMinMaxArguments(sep, 2, 2, status);
+
+    UnicodeString pattern;
+    langData.getNoFallback("localeDisplayPattern", "pattern", pattern);
+    if (pattern.isBogus()) {
+        pattern = UnicodeString("{0} ({1})", -1, US_INV);
+    }
+    format.applyPatternMinMaxArguments(pattern, 2, 2, status);
+    if (pattern.indexOf((UChar)0xFF08) >= 0) {
+        formatOpenParen.setTo((UChar)0xFF08);         // fullwidth (
+        formatReplaceOpenParen.setTo((UChar)0xFF3B);  // fullwidth [
+        formatCloseParen.setTo((UChar)0xFF09);        // fullwidth )
+        formatReplaceCloseParen.setTo((UChar)0xFF3D); // fullwidth ]
+    } else {
+        formatOpenParen.setTo((UChar)0x0028);         // (
+        formatReplaceOpenParen.setTo((UChar)0x005B);  // [
+        formatCloseParen.setTo((UChar)0x0029);        // )
+        formatReplaceCloseParen.setTo((UChar)0x005D); // ]
+    }
+
+    UnicodeString ktPattern;
+    langData.get("localeDisplayPattern", "keyTypePattern", ktPattern);
+    if (ktPattern.isBogus()) {
+        ktPattern = UnicodeString("{0}={1}", -1, US_INV);
+    }
+    keyTypeFormat.applyPatternMinMaxArguments(ktPattern, 2, 2, status);
+
+    uprv_memset(fCapitalization, 0, sizeof(fCapitalization));
+#if !UCONFIG_NO_BREAK_ITERATION
+    // Only get the context data if we need it! This is a const object so we know now...
+    // Also check whether we will need a break iterator (depends on the data)
+    UBool needBrkIter = FALSE;
+    if (capitalizationContext == UDISPCTX_CAPITALIZATION_FOR_UI_LIST_OR_MENU || capitalizationContext == UDISPCTX_CAPITALIZATION_FOR_STANDALONE) {
+        LocalUResourceBundlePointer resource(ures_open(NULL, locale.getName(), &status));
+        if (U_FAILURE(status)) { return; }
+        CapitalizationContextSink sink(*this);
+        ures_getAllItemsWithFallback(resource.getAlias(), "contextTransforms", sink, status);
+        if (status == U_MISSING_RESOURCE_ERROR) {
+            // Silently ignore.  Not every locale has contextTransforms.
+            status = U_ZERO_ERROR;
+        } else if (U_FAILURE(status)) {
+            return;
+        }
+        needBrkIter = sink.hasCapitalizationUsage;
+    }
+    // Get a sentence break iterator if we will need it
+    if (needBrkIter || capitalizationContext == UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE) {
+        status = U_ZERO_ERROR;
+        capitalizationBrkIter = BreakIterator::createSentenceInstance(locale, status);
+        if (U_FAILURE(status)) {
+            delete capitalizationBrkIter;
+            capitalizationBrkIter = NULL;
+        }
+    }
+#endif
+}
+
+LocaleDisplayNamesImpl::~LocaleDisplayNamesImpl() {
+#if !UCONFIG_NO_BREAK_ITERATION
+    delete capitalizationBrkIter;
+#endif
+}
+
+const Locale&
+LocaleDisplayNamesImpl::getLocale() const {
+    return locale;
+}
+
+UDialectHandling
+LocaleDisplayNamesImpl::getDialectHandling() const {
+    return dialectHandling;
+}
+
+UDisplayContext
+LocaleDisplayNamesImpl::getContext(UDisplayContextType type) const {
+    switch (type) {
+        case UDISPCTX_TYPE_DIALECT_HANDLING:
+            return (UDisplayContext)dialectHandling;
+        case UDISPCTX_TYPE_CAPITALIZATION:
+            return capitalizationContext;
+        case UDISPCTX_TYPE_DISPLAY_LENGTH:
+            return nameLength;
+        case UDISPCTX_TYPE_SUBSTITUTE_HANDLING:
+            return substitute;
+        default:
+            break;
+    }
+    return (UDisplayContext)0;
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::adjustForUsageAndContext(CapContextUsage usage,
+                                                UnicodeString& result) const {
+#if !UCONFIG_NO_BREAK_ITERATION
+    // check to see whether we need to titlecase result
+    if ( result.length() > 0 && u_islower(result.char32At(0)) && capitalizationBrkIter!= NULL &&
+          ( capitalizationContext==UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE || fCapitalization[usage] ) ) {
+        // note fCapitalization[usage] won't be set unless capitalizationContext is UI_LIST_OR_MENU or STANDALONE
+        static UMutex capitalizationBrkIterLock;
+        Mutex lock(&capitalizationBrkIterLock);
+        result.toTitle(capitalizationBrkIter, locale, U_TITLECASE_NO_LOWERCASE | U_TITLECASE_NO_BREAK_ADJUSTMENT);
+    }
+#endif
+    return result;
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::localeDisplayName(const Locale& loc,
+                                          UnicodeString& result) const {
+  if (loc.isBogus()) {
+    result.setToBogus();
+    return result;
+  }
+  UnicodeString resultName;
+
+  const char* lang = loc.getLanguage();
+  if (uprv_strlen(lang) == 0) {
+    lang = "root";
+  }
+  const char* script = loc.getScript();
+  const char* country = loc.getCountry();
+  const char* variant = loc.getVariant();
+
+  UBool hasScript = uprv_strlen(script) > 0;
+  UBool hasCountry = uprv_strlen(country) > 0;
+  UBool hasVariant = uprv_strlen(variant) > 0;
+
+  if (dialectHandling == ULDN_DIALECT_NAMES) {
+    char buffer[ULOC_FULLNAME_CAPACITY];
+    do { // loop construct is so we can break early out of search
+      if (hasScript && hasCountry) {
+        ncat(buffer, ULOC_FULLNAME_CAPACITY, lang, "_", script, "_", country, (char *)0);
+        localeIdName(buffer, resultName, false);
+        if (!resultName.isBogus()) {
+          hasScript = FALSE;
+          hasCountry = FALSE;
+          break;
+        }
+      }
+      if (hasScript) {
+        ncat(buffer, ULOC_FULLNAME_CAPACITY, lang, "_", script, (char *)0);
+        localeIdName(buffer, resultName, false);
+        if (!resultName.isBogus()) {
+          hasScript = FALSE;
+          break;
+        }
+      }
+      if (hasCountry) {
+        ncat(buffer, ULOC_FULLNAME_CAPACITY, lang, "_", country, (char*)0);
+        localeIdName(buffer, resultName, false);
+        if (!resultName.isBogus()) {
+          hasCountry = FALSE;
+          break;
+        }
+      }
+    } while (FALSE);
+  }
+  if (resultName.isBogus() || resultName.isEmpty()) {
+    localeIdName(lang, resultName, substitute == UDISPCTX_SUBSTITUTE);
+    if (resultName.isBogus()) {
+      result.setToBogus();
+      return result;
+    }
+  }
+
+  UnicodeString resultRemainder;
+  UnicodeString temp;
+  UErrorCode status = U_ZERO_ERROR;
+
+  if (hasScript) {
+    UnicodeString script_str = scriptDisplayName(script, temp, TRUE);
+    if (script_str.isBogus()) {
+      result.setToBogus();
+      return result;
+    }
+    resultRemainder.append(script_str);
+  }
+  if (hasCountry) {
+    UnicodeString region_str = regionDisplayName(country, temp, TRUE);
+    if (region_str.isBogus()) {
+      result.setToBogus();
+      return result;
+    }
+    appendWithSep(resultRemainder, region_str);
+  }
+  if (hasVariant) {
+    UnicodeString variant_str = variantDisplayName(variant, temp, TRUE);
+    if (variant_str.isBogus()) {
+      result.setToBogus();
+      return result;
+    }
+    appendWithSep(resultRemainder, variant_str);
+  }
+  resultRemainder.findAndReplace(formatOpenParen, formatReplaceOpenParen);
+  resultRemainder.findAndReplace(formatCloseParen, formatReplaceCloseParen);
+
+  LocalPointer<StringEnumeration> e(loc.createKeywords(status));
+  if (e.isValid() && U_SUCCESS(status)) {
+    UnicodeString temp2;
+    char value[ULOC_KEYWORD_AND_VALUES_CAPACITY]; // sigh, no ULOC_VALUE_CAPACITY
+    const char* key;
+    while ((key = e->next((int32_t *)0, status)) != NULL) {
+      value[0] = 0;
+      loc.getKeywordValue(key, value, ULOC_KEYWORD_AND_VALUES_CAPACITY, status);
+      if (U_FAILURE(status) || status == U_STRING_NOT_TERMINATED_WARNING) {
+        return result;
+      }
+      keyDisplayName(key, temp, TRUE);
+      temp.findAndReplace(formatOpenParen, formatReplaceOpenParen);
+      temp.findAndReplace(formatCloseParen, formatReplaceCloseParen);
+      keyValueDisplayName(key, value, temp2, TRUE);
+      temp2.findAndReplace(formatOpenParen, formatReplaceOpenParen);
+      temp2.findAndReplace(formatCloseParen, formatReplaceCloseParen);
+      if (temp2 != UnicodeString(value, -1, US_INV)) {
+        appendWithSep(resultRemainder, temp2);
+      } else if (temp != UnicodeString(key, -1, US_INV)) {
+        UnicodeString temp3;
+        keyTypeFormat.format(temp, temp2, temp3, status);
+        appendWithSep(resultRemainder, temp3);
+      } else {
+        appendWithSep(resultRemainder, temp)
+          .append((UChar)0x3d /* = */)
+          .append(temp2);
+      }
+    }
+  }
+
+  if (!resultRemainder.isEmpty()) {
+    format.format(resultName, resultRemainder, result.remove(), status);
+    return adjustForUsageAndContext(kCapContextUsageLanguage, result);
+  }
+
+  result = resultName;
+  return adjustForUsageAndContext(kCapContextUsageLanguage, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::appendWithSep(UnicodeString& buffer, const UnicodeString& src) const {
+    if (buffer.isEmpty()) {
+        buffer.setTo(src);
+    } else {
+        const UnicodeString *values[2] = { &buffer, &src };
+        UErrorCode status = U_ZERO_ERROR;
+        separatorFormat.formatAndReplace(values, 2, buffer, NULL, 0, status);
+    }
+    return buffer;
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::localeDisplayName(const char* localeId,
+                                          UnicodeString& result) const {
+    return localeDisplayName(Locale(localeId), result);
+}
+
+// private
+UnicodeString&
+LocaleDisplayNamesImpl::localeIdName(const char* localeId,
+                                     UnicodeString& result, bool substitute) const {
+    if (nameLength == UDISPCTX_LENGTH_SHORT) {
+        langData.getNoFallback("Languages%short", localeId, result);
+        if (!result.isBogus()) {
+            return result;
+        }
+    }
+    if (substitute) {
+        return langData.get("Languages", localeId, result);
+    } else {
+        return langData.getNoFallback("Languages", localeId, result);
+    }
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::languageDisplayName(const char* lang,
+                                            UnicodeString& result) const {
+    if (uprv_strcmp("root", lang) == 0 || uprv_strchr(lang, '_') != NULL) {
+        return result = UnicodeString(lang, -1, US_INV);
+    }
+    if (nameLength == UDISPCTX_LENGTH_SHORT) {
+        langData.getNoFallback("Languages%short", lang, result);
+        if (!result.isBogus()) {
+            return adjustForUsageAndContext(kCapContextUsageLanguage, result);
+        }
+    }
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        langData.get("Languages", lang, result);
+    } else {
+        langData.getNoFallback("Languages", lang, result);
+    }
+    return adjustForUsageAndContext(kCapContextUsageLanguage, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::scriptDisplayName(const char* script,
+                                          UnicodeString& result,
+                                          UBool skipAdjust) const {
+    if (nameLength == UDISPCTX_LENGTH_SHORT) {
+        langData.getNoFallback("Scripts%short", script, result);
+        if (!result.isBogus()) {
+            return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageScript, result);
+        }
+    }
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        langData.get("Scripts", script, result);
+    } else {
+        langData.getNoFallback("Scripts", script, result);
+    }
+    return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageScript, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::scriptDisplayName(const char* script,
+                                          UnicodeString& result) const {
+    return scriptDisplayName(script, result, FALSE);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::scriptDisplayName(UScriptCode scriptCode,
+                                          UnicodeString& result) const {
+    return scriptDisplayName(uscript_getName(scriptCode), result, FALSE);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::regionDisplayName(const char* region,
+                                          UnicodeString& result,
+                                          UBool skipAdjust) const {
+    if (nameLength == UDISPCTX_LENGTH_SHORT) {
+         regionData.getNoFallback("Countries%short", region, result);
+        if (!result.isBogus()) {
+            return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageTerritory, result);
+        }
+    }
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        regionData.get("Countries", region, result);
+    } else {
+        regionData.getNoFallback("Countries", region, result);
+    }
+    return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageTerritory, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::regionDisplayName(const char* region,
+                                          UnicodeString& result) const {
+    return regionDisplayName(region, result, FALSE);
+}
+
+
+UnicodeString&
+LocaleDisplayNamesImpl::variantDisplayName(const char* variant,
+                                           UnicodeString& result,
+                                           UBool skipAdjust) const {
+    // don't have a resource for short variant names
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        langData.get("Variants", variant, result);
+    } else {
+        langData.getNoFallback("Variants", variant, result);
+    }
+    return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageVariant, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::variantDisplayName(const char* variant,
+                                           UnicodeString& result) const {
+    return variantDisplayName(variant, result, FALSE);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::keyDisplayName(const char* key,
+                                       UnicodeString& result,
+                                       UBool skipAdjust) const {
+    // don't have a resource for short key names
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        langData.get("Keys", key, result);
+    } else {
+        langData.getNoFallback("Keys", key, result);
+    }
+    return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageKey, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::keyDisplayName(const char* key,
+                                       UnicodeString& result) const {
+    return keyDisplayName(key, result, FALSE);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::keyValueDisplayName(const char* key,
+                                            const char* value,
+                                            UnicodeString& result,
+                                            UBool skipAdjust) const {
+    if (uprv_strcmp(key, "currency") == 0) {
+        // ICU4C does not have ICU4J CurrencyDisplayInfo equivalent for now.
+        UErrorCode sts = U_ZERO_ERROR;
+        UnicodeString ustrValue(value, -1, US_INV);
+        int32_t len;
+        const UChar *currencyName = ucurr_getName(ustrValue.getTerminatedBuffer(),
+            locale.getBaseName(), UCURR_LONG_NAME, nullptr /* isChoiceFormat */, &len, &sts);
+        if (U_FAILURE(sts)) {
+            // Return the value as is on failure
+            result = ustrValue;
+            return result;
+        }
+        result.setTo(currencyName, len);
+        return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageKeyValue, result);
+    }
+
+    if (nameLength == UDISPCTX_LENGTH_SHORT) {
+        langData.getNoFallback("Types%short", key, value, result);
+        if (!result.isBogus()) {
+            return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageKeyValue, result);
+        }
+    }
+    if (substitute == UDISPCTX_SUBSTITUTE) {
+        langData.get("Types", key, value, result);
+    } else {
+        langData.getNoFallback("Types", key, value, result);
+    }
+    return skipAdjust? result: adjustForUsageAndContext(kCapContextUsageKeyValue, result);
+}
+
+UnicodeString&
+LocaleDisplayNamesImpl::keyValueDisplayName(const char* key,
+                                            const char* value,
+                                            UnicodeString& result) const {
+    return keyValueDisplayName(key, value, result, FALSE);
+}
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+LocaleDisplayNames*
+LocaleDisplayNames::createInstance(const Locale& locale,
+                                   UDialectHandling dialectHandling) {
+    return new LocaleDisplayNamesImpl(locale, dialectHandling);
+}
+
+LocaleDisplayNames*
+LocaleDisplayNames::createInstance(const Locale& locale,
+                                   UDisplayContext *contexts, int32_t length) {
+    if (contexts == NULL) {
+        length = 0;
+    }
+    return new LocaleDisplayNamesImpl(locale, contexts, length);
+}
+
+U_NAMESPACE_END
+
+////////////////////////////////////////////////////////////////////////////////////////////////////
+
+U_NAMESPACE_USE
+
+U_CAPI ULocaleDisplayNames * U_EXPORT2
+uldn_open(const char * locale,
+          UDialectHandling dialectHandling,
+          UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (locale == NULL) {
+    locale = uloc_getDefault();
+  }
+  return (ULocaleDisplayNames *)LocaleDisplayNames::createInstance(Locale(locale), dialectHandling);
+}
+
+U_CAPI ULocaleDisplayNames * U_EXPORT2
+uldn_openForContext(const char * locale,
+                    UDisplayContext *contexts, int32_t length,
+                    UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (locale == NULL) {
+    locale = uloc_getDefault();
+  }
+  return (ULocaleDisplayNames *)LocaleDisplayNames::createInstance(Locale(locale), contexts, length);
+}
+
+
+U_CAPI void U_EXPORT2
+uldn_close(ULocaleDisplayNames *ldn) {
+  delete (LocaleDisplayNames *)ldn;
+}
+
+U_CAPI const char * U_EXPORT2
+uldn_getLocale(const ULocaleDisplayNames *ldn) {
+  if (ldn) {
+    return ((const LocaleDisplayNames *)ldn)->getLocale().getName();
+  }
+  return NULL;
+}
+
+U_CAPI UDialectHandling U_EXPORT2
+uldn_getDialectHandling(const ULocaleDisplayNames *ldn) {
+  if (ldn) {
+    return ((const LocaleDisplayNames *)ldn)->getDialectHandling();
+  }
+  return ULDN_STANDARD_NAMES;
+}
+
+U_CAPI UDisplayContext U_EXPORT2
+uldn_getContext(const ULocaleDisplayNames *ldn,
+              UDisplayContextType type,
+              UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return (UDisplayContext)0;
+  }
+  return ((const LocaleDisplayNames *)ldn)->getContext(type);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_localeDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *locale,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || locale == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->localeDisplayName(locale, temp);
+  if (temp.isBogus()) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_languageDisplayName(const ULocaleDisplayNames *ldn,
+                         const char *lang,
+                         UChar *result,
+                         int32_t maxResultSize,
+                         UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || lang == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->languageDisplayName(lang, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_scriptDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *script,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || script == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->scriptDisplayName(script, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_scriptCodeDisplayName(const ULocaleDisplayNames *ldn,
+                           UScriptCode scriptCode,
+                           UChar *result,
+                           int32_t maxResultSize,
+                           UErrorCode *pErrorCode) {
+  return uldn_scriptDisplayName(ldn, uscript_getName(scriptCode), result, maxResultSize, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_regionDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *region,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || region == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->regionDisplayName(region, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_variantDisplayName(const ULocaleDisplayNames *ldn,
+                        const char *variant,
+                        UChar *result,
+                        int32_t maxResultSize,
+                        UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || variant == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->variantDisplayName(variant, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_keyDisplayName(const ULocaleDisplayNames *ldn,
+                    const char *key,
+                    UChar *result,
+                    int32_t maxResultSize,
+                    UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || key == NULL || (result == NULL && maxResultSize > 0) || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->keyDisplayName(key, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uldn_keyValueDisplayName(const ULocaleDisplayNames *ldn,
+                         const char *key,
+                         const char *value,
+                         UChar *result,
+                         int32_t maxResultSize,
+                         UErrorCode *pErrorCode) {
+  if (U_FAILURE(*pErrorCode)) {
+    return 0;
+  }
+  if (ldn == NULL || key == NULL || value == NULL || (result == NULL && maxResultSize > 0)
+      || maxResultSize < 0) {
+    *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  UnicodeString temp(result, 0, maxResultSize);
+  ((const LocaleDisplayNames *)ldn)->keyValueDisplayName(key, value, temp);
+  return temp.extract(result, maxResultSize, *pErrorCode);
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/locid.cpp b/thirdparty/icu4c/common/locid.cpp
new file mode 100644
index 0000000000..2804e36bf6
--- /dev/null
+++ b/thirdparty/icu4c/common/locid.cpp
@@ -0,0 +1,2536 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ **********************************************************************
+ *   Copyright (C) 1997-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ **********************************************************************
+*
+* File locid.cpp
+*
+* Created by: Richard Gillam
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   02/11/97    aliu        Changed gLocPath to fgDataDirectory and added
+*                           methods to get and set it.
+*   04/02/97    aliu        Made operator!= inline; fixed return value
+*                           of getName().
+*   04/15/97    aliu        Cleanup for AIX/Win32.
+*   04/24/97    aliu        Numerous changes per code review.
+*   08/18/98    stephen     Changed getDisplayName()
+*                           Added SIMPLIFIED_CHINESE, TRADITIONAL_CHINESE
+*                           Added getISOCountries(), getISOLanguages(),
+*                           getLanguagesForCountry()
+*   03/16/99    bertrand    rehaul.
+*   07/21/99    stephen     Added U_CFUNC setDefault
+*   11/09/99    weiv        Added const char * getName() const;
+*   04/12/00    srl         removing unicodestring api's and cached hash code
+*   08/10/01    grhoten     Change the static Locales to accessor functions
+******************************************************************************
+*/
+
+#include <utility>
+
+#include "unicode/bytestream.h"
+#include "unicode/locid.h"
+#include "unicode/strenum.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uloc.h"
+#include "unicode/ures.h"
+
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "charstrmap.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "uhash.h"
+#include "ulocimp.h"
+#include "umutex.h"
+#include "uniquecharstr.h"
+#include "ustr_imp.h"
+#include "uvector.h"
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV locale_cleanup(void);
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+static Locale   *gLocaleCache = NULL;
+static UInitOnce gLocaleCacheInitOnce = U_INITONCE_INITIALIZER;
+
+// gDefaultLocaleMutex protects all access to gDefaultLocalesHashT and gDefaultLocale.
+static UMutex gDefaultLocaleMutex;
+static UHashtable *gDefaultLocalesHashT = NULL;
+static Locale *gDefaultLocale = NULL;
+
+/**
+ * \def ULOC_STRING_LIMIT
+ * strings beyond this value crash in CharString
+ */
+#define ULOC_STRING_LIMIT 357913941
+
+U_NAMESPACE_END
+
+typedef enum ELocalePos {
+    eENGLISH,
+    eFRENCH,
+    eGERMAN,
+    eITALIAN,
+    eJAPANESE,
+    eKOREAN,
+    eCHINESE,
+
+    eFRANCE,
+    eGERMANY,
+    eITALY,
+    eJAPAN,
+    eKOREA,
+    eCHINA,      /* Alias for PRC */
+    eTAIWAN,
+    eUK,
+    eUS,
+    eCANADA,
+    eCANADA_FRENCH,
+    eROOT,
+
+
+    //eDEFAULT,
+    eMAX_LOCALES
+} ELocalePos;
+
+U_CDECL_BEGIN
+//
+// Deleter function for Locales owned by the default Locale hash table/
+//
+static void U_CALLCONV
+deleteLocale(void *obj) {
+    delete (icu::Locale *) obj;
+}
+
+static UBool U_CALLCONV locale_cleanup(void)
+{
+    U_NAMESPACE_USE
+
+    delete [] gLocaleCache;
+    gLocaleCache = NULL;
+    gLocaleCacheInitOnce.reset();
+
+    if (gDefaultLocalesHashT) {
+        uhash_close(gDefaultLocalesHashT);   // Automatically deletes all elements, using deleter func.
+        gDefaultLocalesHashT = NULL;
+    }
+    gDefaultLocale = NULL;
+    return TRUE;
+}
+
+
+static void U_CALLCONV locale_init(UErrorCode &status) {
+    U_NAMESPACE_USE
+
+    U_ASSERT(gLocaleCache == NULL);
+    gLocaleCache = new Locale[(int)eMAX_LOCALES];
+    if (gLocaleCache == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE, locale_cleanup);
+    gLocaleCache[eROOT]          = Locale("");
+    gLocaleCache[eENGLISH]       = Locale("en");
+    gLocaleCache[eFRENCH]        = Locale("fr");
+    gLocaleCache[eGERMAN]        = Locale("de");
+    gLocaleCache[eITALIAN]       = Locale("it");
+    gLocaleCache[eJAPANESE]      = Locale("ja");
+    gLocaleCache[eKOREAN]        = Locale("ko");
+    gLocaleCache[eCHINESE]       = Locale("zh");
+    gLocaleCache[eFRANCE]        = Locale("fr", "FR");
+    gLocaleCache[eGERMANY]       = Locale("de", "DE");
+    gLocaleCache[eITALY]         = Locale("it", "IT");
+    gLocaleCache[eJAPAN]         = Locale("ja", "JP");
+    gLocaleCache[eKOREA]         = Locale("ko", "KR");
+    gLocaleCache[eCHINA]         = Locale("zh", "CN");
+    gLocaleCache[eTAIWAN]        = Locale("zh", "TW");
+    gLocaleCache[eUK]            = Locale("en", "GB");
+    gLocaleCache[eUS]            = Locale("en", "US");
+    gLocaleCache[eCANADA]        = Locale("en", "CA");
+    gLocaleCache[eCANADA_FRENCH] = Locale("fr", "CA");
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+Locale *locale_set_default_internal(const char *id, UErrorCode& status) {
+    // Synchronize this entire function.
+    Mutex lock(&gDefaultLocaleMutex);
+
+    UBool canonicalize = FALSE;
+
+    // If given a NULL string for the locale id, grab the default
+    //   name from the system.
+    //   (Different from most other locale APIs, where a null name means use
+    //    the current ICU default locale.)
+    if (id == NULL) {
+        id = uprv_getDefaultLocaleID();   // This function not thread safe? TODO: verify.
+        canonicalize = TRUE; // always canonicalize host ID
+    }
+
+    CharString localeNameBuf;
+    {
+        CharStringByteSink sink(&localeNameBuf);
+        if (canonicalize) {
+            ulocimp_canonicalize(id, sink, &status);
+        } else {
+            ulocimp_getName(id, sink, &status);
+        }
+    }
+
+    if (U_FAILURE(status)) {
+        return gDefaultLocale;
+    }
+
+    if (gDefaultLocalesHashT == NULL) {
+        gDefaultLocalesHashT = uhash_open(uhash_hashChars, uhash_compareChars, NULL, &status);
+        if (U_FAILURE(status)) {
+            return gDefaultLocale;
+        }
+        uhash_setValueDeleter(gDefaultLocalesHashT, deleteLocale);
+        ucln_common_registerCleanup(UCLN_COMMON_LOCALE, locale_cleanup);
+    }
+
+    Locale *newDefault = (Locale *)uhash_get(gDefaultLocalesHashT, localeNameBuf.data());
+    if (newDefault == NULL) {
+        newDefault = new Locale(Locale::eBOGUS);
+        if (newDefault == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return gDefaultLocale;
+        }
+        newDefault->init(localeNameBuf.data(), FALSE);
+        uhash_put(gDefaultLocalesHashT, (char*) newDefault->getName(), newDefault, &status);
+        if (U_FAILURE(status)) {
+            return gDefaultLocale;
+        }
+    }
+    gDefaultLocale = newDefault;
+    return gDefaultLocale;
+}
+
+U_NAMESPACE_END
+
+/* sfb 07/21/99 */
+U_CFUNC void
+locale_set_default(const char *id)
+{
+    U_NAMESPACE_USE
+    UErrorCode status = U_ZERO_ERROR;
+    locale_set_default_internal(id, status);
+}
+/* end */
+
+U_CFUNC const char *
+locale_get_default(void)
+{
+    U_NAMESPACE_USE
+    return Locale::getDefault().getName();
+}
+
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(Locale)
+
+/*Character separating the posix id fields*/
+// '_'
+// In the platform codepage.
+#define SEP_CHAR '_'
+#define NULL_CHAR '\0'
+
+Locale::~Locale()
+{
+    if (baseName != fullName) {
+        uprv_free(baseName);
+    }
+    baseName = NULL;
+    /*if fullName is on the heap, we free it*/
+    if (fullName != fullNameBuffer)
+    {
+        uprv_free(fullName);
+        fullName = NULL;
+    }
+}
+
+Locale::Locale()
+    : UObject(), fullName(fullNameBuffer), baseName(NULL)
+{
+    init(NULL, FALSE);
+}
+
+/*
+ * Internal constructor to allow construction of a locale object with
+ *   NO side effects.   (Default constructor tries to get
+ *   the default locale.)
+ */
+Locale::Locale(Locale::ELocaleType)
+    : UObject(), fullName(fullNameBuffer), baseName(NULL)
+{
+    setToBogus();
+}
+
+
+Locale::Locale( const   char * newLanguage,
+                const   char * newCountry,
+                const   char * newVariant,
+                const   char * newKeywords)
+    : UObject(), fullName(fullNameBuffer), baseName(NULL)
+{
+    if( (newLanguage==NULL) && (newCountry == NULL) && (newVariant == NULL) )
+    {
+        init(NULL, FALSE); /* shortcut */
+    }
+    else
+    {
+        UErrorCode status = U_ZERO_ERROR;
+        int32_t size = 0;
+        int32_t lsize = 0;
+        int32_t csize = 0;
+        int32_t vsize = 0;
+        int32_t ksize = 0;
+
+        // Calculate the size of the resulting string.
+
+        // Language
+        if ( newLanguage != NULL )
+        {
+            lsize = (int32_t)uprv_strlen(newLanguage);
+            if ( lsize < 0 || lsize > ULOC_STRING_LIMIT ) { // int32 wrap
+                setToBogus();
+                return;
+            }
+            size = lsize;
+        }
+
+        CharString togo(newLanguage, lsize, status); // start with newLanguage
+
+        // _Country
+        if ( newCountry != NULL )
+        {
+            csize = (int32_t)uprv_strlen(newCountry);
+            if ( csize < 0 || csize > ULOC_STRING_LIMIT ) { // int32 wrap
+                setToBogus();
+                return;
+            }
+            size += csize;
+        }
+
+        // _Variant
+        if ( newVariant != NULL )
+        {
+            // remove leading _'s
+            while(newVariant[0] == SEP_CHAR)
+            {
+                newVariant++;
+            }
+
+            // remove trailing _'s
+            vsize = (int32_t)uprv_strlen(newVariant);
+            if ( vsize < 0 || vsize > ULOC_STRING_LIMIT ) { // int32 wrap
+                setToBogus();
+                return;
+            }
+            while( (vsize>1) && (newVariant[vsize-1] == SEP_CHAR) )
+            {
+                vsize--;
+            }
+        }
+
+        if( vsize > 0 )
+        {
+            size += vsize;
+        }
+
+        // Separator rules:
+        if ( vsize > 0 )
+        {
+            size += 2;  // at least: __v
+        }
+        else if ( csize > 0 )
+        {
+            size += 1;  // at least: _v
+        }
+
+        if ( newKeywords != NULL)
+        {
+            ksize = (int32_t)uprv_strlen(newKeywords);
+            if ( ksize < 0 || ksize > ULOC_STRING_LIMIT ) {
+              setToBogus();
+              return;
+            }
+            size += ksize + 1;
+        }
+
+        //  NOW we have the full locale string..
+        // Now, copy it back.
+
+        // newLanguage is already copied
+
+        if ( ( vsize != 0 ) || (csize != 0) )  // at least:  __v
+        {                                      //            ^
+            togo.append(SEP_CHAR, status);
+        }
+
+        if ( csize != 0 )
+        {
+            togo.append(newCountry, status);
+        }
+
+        if ( vsize != 0)
+        {
+            togo.append(SEP_CHAR, status)
+                .append(newVariant, vsize, status);
+        }
+
+        if ( ksize != 0)
+        {
+            if (uprv_strchr(newKeywords, '=')) {
+                togo.append('@', status); /* keyword parsing */
+            }
+            else {
+                togo.append('_', status); /* Variant parsing with a script */
+                if ( vsize == 0) {
+                    togo.append('_', status); /* No country found */
+                }
+            }
+            togo.append(newKeywords, status);
+        }
+
+        if (U_FAILURE(status)) {
+            // Something went wrong with appending, etc.
+            setToBogus();
+            return;
+        }
+        // Parse it, because for example 'language' might really be a complete
+        // string.
+        init(togo.data(), FALSE);
+    }
+}
+
+Locale::Locale(const Locale &other)
+    : UObject(other), fullName(fullNameBuffer), baseName(NULL)
+{
+    *this = other;
+}
+
+Locale::Locale(Locale&& other) U_NOEXCEPT
+    : UObject(other), fullName(fullNameBuffer), baseName(fullName) {
+  *this = std::move(other);
+}
+
+Locale& Locale::operator=(const Locale& other) {
+    if (this == &other) {
+        return *this;
+    }
+
+    setToBogus();
+
+    if (other.fullName == other.fullNameBuffer) {
+        uprv_strcpy(fullNameBuffer, other.fullNameBuffer);
+    } else if (other.fullName == nullptr) {
+        fullName = nullptr;
+    } else {
+        fullName = uprv_strdup(other.fullName);
+        if (fullName == nullptr) return *this;
+    }
+
+    if (other.baseName == other.fullName) {
+        baseName = fullName;
+    } else if (other.baseName != nullptr) {
+        baseName = uprv_strdup(other.baseName);
+        if (baseName == nullptr) return *this;
+    }
+
+    uprv_strcpy(language, other.language);
+    uprv_strcpy(script, other.script);
+    uprv_strcpy(country, other.country);
+
+    variantBegin = other.variantBegin;
+    fIsBogus = other.fIsBogus;
+
+    return *this;
+}
+
+Locale& Locale::operator=(Locale&& other) U_NOEXCEPT {
+    if (baseName != fullName) uprv_free(baseName);
+    if (fullName != fullNameBuffer) uprv_free(fullName);
+
+    if (other.fullName == other.fullNameBuffer) {
+        uprv_strcpy(fullNameBuffer, other.fullNameBuffer);
+        fullName = fullNameBuffer;
+    } else {
+        fullName = other.fullName;
+    }
+
+    if (other.baseName == other.fullName) {
+        baseName = fullName;
+    } else {
+        baseName = other.baseName;
+    }
+
+    uprv_strcpy(language, other.language);
+    uprv_strcpy(script, other.script);
+    uprv_strcpy(country, other.country);
+
+    variantBegin = other.variantBegin;
+    fIsBogus = other.fIsBogus;
+
+    other.baseName = other.fullName = other.fullNameBuffer;
+
+    return *this;
+}
+
+Locale *
+Locale::clone() const {
+    return new Locale(*this);
+}
+
+UBool
+Locale::operator==( const   Locale& other) const
+{
+    return (uprv_strcmp(other.fullName, fullName) == 0);
+}
+
+namespace {
+
+UInitOnce gKnownCanonicalizedInitOnce = U_INITONCE_INITIALIZER;
+UHashtable *gKnownCanonicalized = nullptr;
+
+static const char* const KNOWN_CANONICALIZED[] = {
+    "c",
+    // Commonly used locales known are already canonicalized
+    "af", "af_ZA", "am", "am_ET", "ar", "ar_001", "as", "as_IN", "az", "az_AZ",
+    "be", "be_BY", "bg", "bg_BG", "bn", "bn_IN", "bs", "bs_BA", "ca", "ca_ES",
+    "cs", "cs_CZ", "cy", "cy_GB", "da", "da_DK", "de", "de_DE", "el", "el_GR",
+    "en", "en_GB", "en_US", "es", "es_419", "es_ES", "et", "et_EE", "eu",
+    "eu_ES", "fa", "fa_IR", "fi", "fi_FI", "fil", "fil_PH", "fr", "fr_FR",
+    "ga", "ga_IE", "gl", "gl_ES", "gu", "gu_IN", "he", "he_IL", "hi", "hi_IN",
+    "hr", "hr_HR", "hu", "hu_HU", "hy", "hy_AM", "id", "id_ID", "is", "is_IS",
+    "it", "it_IT", "ja", "ja_JP", "jv", "jv_ID", "ka", "ka_GE", "kk", "kk_KZ",
+    "km", "km_KH", "kn", "kn_IN", "ko", "ko_KR", "ky", "ky_KG", "lo", "lo_LA",
+    "lt", "lt_LT", "lv", "lv_LV", "mk", "mk_MK", "ml", "ml_IN", "mn", "mn_MN",
+    "mr", "mr_IN", "ms", "ms_MY", "my", "my_MM", "nb", "nb_NO", "ne", "ne_NP",
+    "nl", "nl_NL", "or", "or_IN", "pa", "pa_IN", "pl", "pl_PL", "ps", "ps_AF",
+    "pt", "pt_BR", "pt_PT", "ro", "ro_RO", "ru", "ru_RU", "sd", "sd_IN", "si",
+    "si_LK", "sk", "sk_SK", "sl", "sl_SI", "so", "so_SO", "sq", "sq_AL", "sr",
+    "sr_Cyrl_RS", "sr_Latn", "sr_RS", "sv", "sv_SE", "sw", "sw_TZ", "ta",
+    "ta_IN", "te", "te_IN", "th", "th_TH", "tk", "tk_TM", "tr", "tr_TR", "uk",
+    "uk_UA", "ur", "ur_PK", "uz", "uz_UZ", "vi", "vi_VN", "yue", "yue_Hant",
+    "yue_Hant_HK", "yue_HK", "zh", "zh_CN", "zh_Hans", "zh_Hans_CN", "zh_Hant",
+    "zh_Hant_TW", "zh_TW", "zu", "zu_ZA"
+};
+
+static UBool U_CALLCONV cleanupKnownCanonicalized() {
+    gKnownCanonicalizedInitOnce.reset();
+    if (gKnownCanonicalized) { uhash_close(gKnownCanonicalized); }
+    return TRUE;
+}
+
+static void U_CALLCONV loadKnownCanonicalized(UErrorCode &status) {
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE_KNOWN_CANONICALIZED,
+                                cleanupKnownCanonicalized);
+    LocalUHashtablePointer newKnownCanonicalizedMap(
+        uhash_open(uhash_hashChars, uhash_compareChars, nullptr, &status));
+    for (int32_t i = 0;
+            U_SUCCESS(status) && i < UPRV_LENGTHOF(KNOWN_CANONICALIZED);
+            i++) {
+        uhash_puti(newKnownCanonicalizedMap.getAlias(),
+                   (void*)KNOWN_CANONICALIZED[i],
+                   1, &status);
+    }
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    gKnownCanonicalized = newKnownCanonicalizedMap.orphan();
+}
+
+class AliasData;
+
+/**
+ * A Builder class to build the alias data.
+ */
+class AliasDataBuilder {
+public:
+    AliasDataBuilder() {
+    }
+
+    // Build the AliasData from resource.
+    AliasData* build(UErrorCode &status);
+
+private:
+    void readAlias(UResourceBundle* alias,
+                   UniqueCharStrings* strings,
+                   LocalMemory<const char*>& types,
+                   LocalMemory<int32_t>& replacementIndexes,
+                   int32_t &length,
+                   void (*checkType)(const char* type),
+                   void (*checkReplacement)(const UnicodeString& replacement),
+                   UErrorCode &status);
+
+    // Read the languageAlias data from alias to
+    // strings+types+replacementIndexes
+    // The number of record will be stored into length.
+    // Allocate length items for types, to store the type field.
+    // Allocate length items for replacementIndexes,
+    // to store the index in the strings for the replacement script.
+    void readLanguageAlias(UResourceBundle* alias,
+                           UniqueCharStrings* strings,
+                           LocalMemory<const char*>& types,
+                           LocalMemory<int32_t>& replacementIndexes,
+                           int32_t &length,
+                           UErrorCode &status);
+
+    // Read the scriptAlias data from alias to
+    // strings+types+replacementIndexes
+    // Allocate length items for types, to store the type field.
+    // Allocate length items for replacementIndexes,
+    // to store the index in the strings for the replacement script.
+    void readScriptAlias(UResourceBundle* alias,
+                         UniqueCharStrings* strings,
+                         LocalMemory<const char*>& types,
+                         LocalMemory<int32_t>& replacementIndexes,
+                         int32_t &length, UErrorCode &status);
+
+    // Read the territoryAlias data from alias to
+    // strings+types+replacementIndexes
+    // Allocate length items for types, to store the type field.
+    // Allocate length items for replacementIndexes,
+    // to store the index in the strings for the replacement script.
+    void readTerritoryAlias(UResourceBundle* alias,
+                            UniqueCharStrings* strings,
+                            LocalMemory<const char*>& types,
+                            LocalMemory<int32_t>& replacementIndexes,
+                            int32_t &length, UErrorCode &status);
+
+    // Read the variantAlias data from alias to
+    // strings+types+replacementIndexes
+    // Allocate length items for types, to store the type field.
+    // Allocate length items for replacementIndexes,
+    // to store the index in the strings for the replacement variant.
+    void readVariantAlias(UResourceBundle* alias,
+                          UniqueCharStrings* strings,
+                          LocalMemory<const char*>& types,
+                          LocalMemory<int32_t>& replacementIndexes,
+                          int32_t &length, UErrorCode &status);
+};
+
+/**
+ * A class to hold the Alias Data.
+ */
+class AliasData : public UMemory {
+public:
+    static const AliasData* singleton(UErrorCode& status) {
+        if (U_FAILURE(status)) {
+            // Do not get into loadData if the status already has error.
+            return nullptr;
+        }
+        umtx_initOnce(AliasData::gInitOnce, &AliasData::loadData, status);
+        return gSingleton;
+    }
+
+    const CharStringMap& languageMap() const { return language; }
+    const CharStringMap& scriptMap() const { return script; }
+    const CharStringMap& territoryMap() const { return territory; }
+    const CharStringMap& variantMap() const { return variant; }
+
+    static void U_CALLCONV loadData(UErrorCode &status);
+    static UBool U_CALLCONV cleanup();
+
+    static UInitOnce gInitOnce;
+
+private:
+    AliasData(CharStringMap languageMap,
+              CharStringMap scriptMap,
+              CharStringMap territoryMap,
+              CharStringMap variantMap,
+              CharString* strings)
+        : language(std::move(languageMap)),
+          script(std::move(scriptMap)),
+          territory(std::move(territoryMap)),
+          variant(std::move(variantMap)),
+          strings(strings) {
+    }
+
+    ~AliasData() {
+        delete strings;
+    }
+
+    static const AliasData* gSingleton;
+
+    CharStringMap language;
+    CharStringMap script;
+    CharStringMap territory;
+    CharStringMap variant;
+    CharString* strings;
+
+    friend class AliasDataBuilder;
+};
+
+
+const AliasData* AliasData::gSingleton = nullptr;
+UInitOnce AliasData::gInitOnce = U_INITONCE_INITIALIZER;
+
+UBool U_CALLCONV
+AliasData::cleanup()
+{
+    gInitOnce.reset();
+    delete gSingleton;
+    return TRUE;
+}
+
+void
+AliasDataBuilder::readAlias(
+        UResourceBundle* alias,
+        UniqueCharStrings* strings,
+        LocalMemory<const char*>& types,
+        LocalMemory<int32_t>& replacementIndexes,
+        int32_t &length,
+        void (*checkType)(const char* type),
+        void (*checkReplacement)(const UnicodeString& replacement),
+        UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    length = ures_getSize(alias);
+    const char** rawTypes = types.allocateInsteadAndCopy(length);
+    if (rawTypes == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    int32_t* rawIndexes = replacementIndexes.allocateInsteadAndCopy(length);
+    if (rawIndexes == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    int i = 0;
+    while (ures_hasNext(alias)) {
+        LocalUResourceBundlePointer res(
+            ures_getNextResource(alias, nullptr, &status));
+        const char* aliasFrom = ures_getKey(res.getAlias());
+        UnicodeString aliasTo =
+            ures_getUnicodeStringByKey(res.getAlias(), "replacement", &status);
+
+        checkType(aliasFrom);
+        checkReplacement(aliasTo);
+
+        rawTypes[i] = aliasFrom;
+        rawIndexes[i] = strings->add(aliasTo, status);
+        i++;
+    }
+}
+
+/**
+ * Read the languageAlias data from alias to strings+types+replacementIndexes.
+ * Allocate length items for types, to store the type field. Allocate length
+ * items for replacementIndexes, to store the index in the strings for the
+ * replacement language.
+ */
+void
+AliasDataBuilder::readLanguageAlias(
+        UResourceBundle* alias,
+        UniqueCharStrings* strings,
+        LocalMemory<const char*>& types,
+        LocalMemory<int32_t>& replacementIndexes,
+        int32_t &length,
+        UErrorCode &status)
+{
+    return readAlias(
+        alias, strings, types, replacementIndexes, length,
+#if U_DEBUG
+        [](const char* type) {
+            // Assert the aliasFrom only contains the following possibilties
+            // language_REGION_variant
+            // language_REGION
+            // language_variant
+            // language
+            // und_variant
+            Locale test(type);
+            // Assert no script in aliasFrom
+            U_ASSERT(test.getScript()[0] == '\0');
+            // Assert when language is und, no REGION in aliasFrom.
+            U_ASSERT(test.getLanguage()[0] != '\0' || test.getCountry()[0] == '\0');
+        },
+#else
+        [](const char*) {},
+#endif
+        [](const UnicodeString&) {}, status);
+}
+
+/**
+ * Read the scriptAlias data from alias to strings+types+replacementIndexes.
+ * Allocate length items for types, to store the type field. Allocate length
+ * items for replacementIndexes, to store the index in the strings for the
+ * replacement script.
+ */
+void
+AliasDataBuilder::readScriptAlias(
+        UResourceBundle* alias,
+        UniqueCharStrings* strings,
+        LocalMemory<const char*>& types,
+        LocalMemory<int32_t>& replacementIndexes,
+        int32_t &length,
+        UErrorCode &status)
+{
+    return readAlias(
+        alias, strings, types, replacementIndexes, length,
+#if U_DEBUG
+        [](const char* type) {
+            U_ASSERT(uprv_strlen(type) == 4);
+        },
+        [](const UnicodeString& replacement) {
+            U_ASSERT(replacement.length() == 4);
+        },
+#else
+        [](const char*) {},
+        [](const UnicodeString&) { },
+#endif
+        status);
+}
+
+/**
+ * Read the territoryAlias data from alias to strings+types+replacementIndexes.
+ * Allocate length items for types, to store the type field. Allocate length
+ * items for replacementIndexes, to store the index in the strings for the
+ * replacement regions.
+ */
+void
+AliasDataBuilder::readTerritoryAlias(
+        UResourceBundle* alias,
+        UniqueCharStrings* strings,
+        LocalMemory<const char*>& types,
+        LocalMemory<int32_t>& replacementIndexes,
+        int32_t &length,
+        UErrorCode &status)
+{
+    return readAlias(
+        alias, strings, types, replacementIndexes, length,
+#if U_DEBUG
+        [](const char* type) {
+            U_ASSERT(uprv_strlen(type) == 2 || uprv_strlen(type) == 3);
+        },
+#else
+        [](const char*) {},
+#endif
+        [](const UnicodeString&) { },
+        status);
+}
+
+/**
+ * Read the variantAlias data from alias to strings+types+replacementIndexes.
+ * Allocate length items for types, to store the type field. Allocate length
+ * items for replacementIndexes, to store the index in the strings for the
+ * replacement variant.
+ */
+void
+AliasDataBuilder::readVariantAlias(
+        UResourceBundle* alias,
+        UniqueCharStrings* strings,
+        LocalMemory<const char*>& types,
+        LocalMemory<int32_t>& replacementIndexes,
+        int32_t &length,
+        UErrorCode &status)
+{
+    return readAlias(
+        alias, strings, types, replacementIndexes, length,
+#if U_DEBUG
+        [](const char* type) {
+            U_ASSERT(uprv_strlen(type) >= 4 && uprv_strlen(type) <= 8);
+            U_ASSERT(uprv_strlen(type) != 4 ||
+                     (type[0] >= '0' && type[0] <= '9'));
+        },
+        [](const UnicodeString& replacement) {
+            U_ASSERT(replacement.length() >= 4 && replacement.length() <= 8);
+            U_ASSERT(replacement.length() != 4 ||
+                     (replacement.charAt(0) >= u'0' &&
+                      replacement.charAt(0) <= u'9'));
+        },
+#else
+        [](const char*) {},
+        [](const UnicodeString&) { },
+#endif
+        status);
+}
+
+/**
+ * Initializes the alias data from the ICU resource bundles. The alias data
+ * contains alias of language, country, script and variants.
+ *
+ * If the alias data has already loaded, then this method simply returns without
+ * doing anything meaningful.
+ */
+void U_CALLCONV
+AliasData::loadData(UErrorCode &status)
+{
+#ifdef LOCALE_CANONICALIZATION_DEBUG
+    UDate start = uprv_getRawUTCtime();
+#endif  // LOCALE_CANONICALIZATION_DEBUG
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE_ALIAS, cleanup);
+    AliasDataBuilder builder;
+    gSingleton = builder.build(status);
+#ifdef LOCALE_CANONICALIZATION_DEBUG
+    UDate end = uprv_getRawUTCtime();
+    printf("AliasData::loadData took total %f ms\n", end - start);
+#endif  // LOCALE_CANONICALIZATION_DEBUG
+}
+
+/**
+ * Build the alias data from resources.
+ */
+AliasData*
+AliasDataBuilder::build(UErrorCode &status) {
+    LocalUResourceBundlePointer metadata(
+        ures_openDirect(nullptr, "metadata", &status));
+    LocalUResourceBundlePointer metadataAlias(
+        ures_getByKey(metadata.getAlias(), "alias", nullptr, &status));
+    LocalUResourceBundlePointer languageAlias(
+        ures_getByKey(metadataAlias.getAlias(), "language", nullptr, &status));
+    LocalUResourceBundlePointer scriptAlias(
+        ures_getByKey(metadataAlias.getAlias(), "script", nullptr, &status));
+    LocalUResourceBundlePointer territoryAlias(
+        ures_getByKey(metadataAlias.getAlias(), "territory", nullptr, &status));
+    LocalUResourceBundlePointer variantAlias(
+        ures_getByKey(metadataAlias.getAlias(), "variant", nullptr, &status));
+
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+    int32_t languagesLength = 0, scriptLength = 0, territoryLength = 0,
+            variantLength = 0;
+
+    // Read the languageAlias into languageTypes, languageReplacementIndexes
+    // and strings
+    UniqueCharStrings strings(status);
+    LocalMemory<const char*> languageTypes;
+    LocalMemory<int32_t> languageReplacementIndexes;
+    readLanguageAlias(languageAlias.getAlias(),
+                      &strings,
+                      languageTypes,
+                      languageReplacementIndexes,
+                      languagesLength,
+                      status);
+
+    // Read the scriptAlias into scriptTypes, scriptReplacementIndexes
+    // and strings
+    LocalMemory<const char*> scriptTypes;
+    LocalMemory<int32_t> scriptReplacementIndexes;
+    readScriptAlias(scriptAlias.getAlias(),
+                    &strings,
+                    scriptTypes,
+                    scriptReplacementIndexes,
+                    scriptLength,
+                    status);
+
+    // Read the territoryAlias into territoryTypes, territoryReplacementIndexes
+    // and strings
+    LocalMemory<const char*> territoryTypes;
+    LocalMemory<int32_t> territoryReplacementIndexes;
+    readTerritoryAlias(territoryAlias.getAlias(),
+                       &strings,
+                       territoryTypes,
+                       territoryReplacementIndexes,
+                       territoryLength, status);
+
+    // Read the variantAlias into variantTypes, variantReplacementIndexes
+    // and strings
+    LocalMemory<const char*> variantTypes;
+    LocalMemory<int32_t> variantReplacementIndexes;
+    readVariantAlias(variantAlias.getAlias(),
+                     &strings,
+                     variantTypes,
+                     variantReplacementIndexes,
+                     variantLength, status);
+
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+
+    // We can only use strings after freeze it.
+    strings.freeze();
+
+    // Build the languageMap from languageTypes & languageReplacementIndexes
+    CharStringMap languageMap(490, status);
+    for (int32_t i = 0; U_SUCCESS(status) && i < languagesLength; i++) {
+        languageMap.put(languageTypes[i],
+                        strings.get(languageReplacementIndexes[i]),
+                        status);
+    }
+
+    // Build the scriptMap from scriptTypes & scriptReplacementIndexes
+    CharStringMap scriptMap(1, status);
+    for (int32_t i = 0; U_SUCCESS(status) && i < scriptLength; i++) {
+        scriptMap.put(scriptTypes[i],
+                      strings.get(scriptReplacementIndexes[i]),
+                      status);
+    }
+
+    // Build the territoryMap from territoryTypes & territoryReplacementIndexes
+    CharStringMap territoryMap(650, status);
+    for (int32_t i = 0; U_SUCCESS(status) && i < territoryLength; i++) {
+        territoryMap.put(territoryTypes[i],
+                         strings.get(territoryReplacementIndexes[i]),
+                         status);
+    }
+
+    // Build the variantMap from variantTypes & variantReplacementIndexes.
+    CharStringMap variantMap(2, status);
+    for (int32_t i = 0; U_SUCCESS(status) && i < variantLength; i++) {
+        variantMap.put(variantTypes[i],
+                       strings.get(variantReplacementIndexes[i]),
+                       status);
+    }
+
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+
+    // copy hashtables
+    auto *data = new AliasData(
+        std::move(languageMap),
+        std::move(scriptMap),
+        std::move(territoryMap),
+        std::move(variantMap),
+        strings.orphanCharStrings());
+
+    if (data == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return data;
+}
+
+/**
+ * A class that find the replacement values of locale fields by using AliasData.
+ */
+class AliasReplacer {
+public:
+    AliasReplacer(UErrorCode status) :
+            language(nullptr), script(nullptr), region(nullptr),
+            extensions(nullptr), variants(status),
+            data(nullptr) {
+    }
+    ~AliasReplacer() {
+    }
+
+    // Check the fields inside locale, if need to replace fields,
+    // place the the replaced locale ID in out and return true.
+    // Otherwise return false for no replacement or error.
+    bool replace(
+        const Locale& locale, CharString& out, UErrorCode status);
+
+private:
+    const char* language;
+    const char* script;
+    const char* region;
+    const char* extensions;
+    UVector variants;
+
+    const AliasData* data;
+
+    inline bool notEmpty(const char* str) {
+        return str && str[0] != NULL_CHAR;
+    }
+
+    /**
+     * If replacement is neither null nor empty and input is either null or empty,
+     * return replacement.
+     * If replacement is neither null nor empty but input is not empty, return input.
+     * If replacement is either null or empty and type is either null or empty,
+     * return input.
+     * Otherwise return null.
+     *   replacement     input      type        return
+     *    AAA             nullptr    *           AAA
+     *    AAA             BBB        *           BBB
+     *    nullptr || ""   CCC        nullptr     CCC
+     *    nullptr || ""   *          DDD         nullptr
+     */
+    inline const char* deleteOrReplace(
+            const char* input, const char* type, const char* replacement) {
+        return notEmpty(replacement) ?
+            ((input == nullptr) ?  replacement : input) :
+            ((type == nullptr) ? input  : nullptr);
+    }
+
+    inline bool same(const char* a, const char* b) {
+        if (a == nullptr && b == nullptr) {
+            return true;
+        }
+        if ((a == nullptr && b != nullptr) ||
+            (a != nullptr && b == nullptr)) {
+          return false;
+        }
+        return uprv_strcmp(a, b) == 0;
+    }
+
+    // Gather fields and generate locale ID into out.
+    CharString& outputToString(CharString& out, UErrorCode status);
+
+    // Generate the lookup key.
+    CharString& generateKey(const char* language, const char* region,
+                            const char* variant, CharString& out,
+                            UErrorCode status);
+
+    void parseLanguageReplacement(const char* replacement,
+                                  const char*& replaceLanguage,
+                                  const char*& replaceScript,
+                                  const char*& replaceRegion,
+                                  const char*& replaceVariant,
+                                  const char*& replaceExtensions,
+                                  UVector& toBeFreed,
+                                  UErrorCode& status);
+
+    // Replace by using languageAlias.
+    bool replaceLanguage(bool checkLanguage, bool checkRegion,
+                         bool checkVariants, UVector& toBeFreed,
+                         UErrorCode& status);
+
+    // Replace by using territoryAlias.
+    bool replaceTerritory(UVector& toBeFreed, UErrorCode& status);
+
+    // Replace by using scriptAlias.
+    bool replaceScript(UErrorCode& status);
+
+    // Replace by using variantAlias.
+    bool replaceVariant(UErrorCode& status);
+};
+
+CharString&
+AliasReplacer::generateKey(
+        const char* language, const char* region, const char* variant,
+        CharString& out, UErrorCode status)
+{
+    out.append(language, status);
+    if (notEmpty(region)) {
+        out.append(SEP_CHAR, status)
+            .append(region, status);
+    }
+    if (notEmpty(variant)) {
+       out.append(SEP_CHAR, status)
+           .append(variant, status);
+    }
+    return out;
+}
+
+void
+AliasReplacer::parseLanguageReplacement(
+    const char* replacement,
+    const char*& replacedLanguage,
+    const char*& replacedScript,
+    const char*& replacedRegion,
+    const char*& replacedVariant,
+    const char*& replacedExtensions,
+    UVector& toBeFreed,
+    UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+    replacedScript = replacedRegion = replacedVariant
+        = replacedExtensions = nullptr;
+    if (uprv_strchr(replacement, '_') == nullptr) {
+        replacedLanguage = replacement;
+        // reach the end, just return it.
+        return;
+    }
+    // We have multiple field so we have to allocate and parse
+    CharString* str = new CharString(
+        replacement, (int32_t)uprv_strlen(replacement), status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (str == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    toBeFreed.addElement(str, status);
+    char* data = str->data();
+    replacedLanguage = (const char*) data;
+    char* endOfField = uprv_strchr(data, '_');
+    *endOfField = '\0'; // null terminiate it.
+    endOfField++;
+    const char* start = endOfField;
+    endOfField = (char*) uprv_strchr(start, '_');
+    size_t len = 0;
+    if (endOfField == nullptr) {
+        len = uprv_strlen(start);
+    } else {
+        len = endOfField - start;
+        *endOfField = '\0'; // null terminiate it.
+    }
+    if (len == 4 && uprv_isASCIILetter(*start)) {
+        // Got a script
+        replacedScript = start;
+        if (endOfField == nullptr) {
+            return;
+        }
+        start = endOfField++;
+        endOfField = (char*)uprv_strchr(start, '_');
+        if (endOfField == nullptr) {
+            len = uprv_strlen(start);
+        } else {
+            len = endOfField - start;
+            *endOfField = '\0'; // null terminiate it.
+        }
+    }
+    if (len >= 2 && len <= 3) {
+        // Got a region
+        replacedRegion = start;
+        if (endOfField == nullptr) {
+            return;
+        }
+        start = endOfField++;
+        endOfField = (char*)uprv_strchr(start, '_');
+        if (endOfField == nullptr) {
+            len = uprv_strlen(start);
+        } else {
+            len = endOfField - start;
+            *endOfField = '\0'; // null terminiate it.
+        }
+    }
+    if (len >= 4) {
+        // Got a variant
+        replacedVariant = start;
+        if (endOfField == nullptr) {
+            return;
+        }
+        start = endOfField++;
+    }
+    replacedExtensions = start;
+}
+
+bool
+AliasReplacer::replaceLanguage(
+        bool checkLanguage, bool checkRegion,
+        bool checkVariants, UVector& toBeFreed, UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    if (    (checkRegion && region == nullptr) ||
+            (checkVariants && variants.size() == 0)) {
+        // Nothing to search.
+        return false;
+    }
+    int32_t variant_size = checkVariants ? variants.size() : 1;
+    // Since we may have more than one variant, we need to loop through them.
+    const char* searchLanguage = checkLanguage ? language : "und";
+    const char* searchRegion = checkRegion ? region : nullptr;
+    const char* searchVariant = nullptr;
+    for (int32_t variant_index = 0;
+            variant_index < variant_size;
+            variant_index++) {
+        if (checkVariants) {
+            U_ASSERT(variant_index < variant_size);
+            searchVariant = (const char*)(variants.elementAt(variant_index));
+        }
+
+        if (searchVariant != nullptr && uprv_strlen(searchVariant) < 4) {
+            // Do not consider  ill-formed variant subtag.
+            searchVariant = nullptr;
+        }
+        CharString typeKey;
+        generateKey(searchLanguage, searchRegion, searchVariant, typeKey,
+                    status);
+        if (U_FAILURE(status)) {
+            return false;
+        }
+        const char *replacement = data->languageMap().get(typeKey.data());
+        if (replacement == nullptr) {
+            // Found no replacement data.
+            continue;
+        }
+
+        const char* replacedLanguage = nullptr;
+        const char* replacedScript = nullptr;
+        const char* replacedRegion = nullptr;
+        const char* replacedVariant = nullptr;
+        const char* replacedExtensions = nullptr;
+        parseLanguageReplacement(replacement,
+                                 replacedLanguage,
+                                 replacedScript,
+                                 replacedRegion,
+                                 replacedVariant,
+                                 replacedExtensions,
+                                 toBeFreed,
+                                 status);
+        replacedLanguage =
+            (replacedLanguage != nullptr && uprv_strcmp(replacedLanguage, "und") == 0) ?
+            language : replacedLanguage;
+        replacedScript = deleteOrReplace(script, nullptr, replacedScript);
+        replacedRegion = deleteOrReplace(region, searchRegion, replacedRegion);
+        replacedVariant = deleteOrReplace(
+            searchVariant, searchVariant, replacedVariant);
+
+        if (    same(language, replacedLanguage) &&
+                same(script, replacedScript) &&
+                same(region, replacedRegion) &&
+                same(searchVariant, replacedVariant) &&
+                replacedExtensions == nullptr) {
+            // Replacement produce no changes.
+            continue;
+        }
+
+        language = replacedLanguage;
+        region = replacedRegion;
+        script = replacedScript;
+        if (searchVariant != nullptr) {
+            if (notEmpty(replacedVariant)) {
+                variants.setElementAt((void*)replacedVariant, variant_index);
+            } else {
+                variants.removeElementAt(variant_index);
+            }
+        }
+        if (replacedExtensions != nullptr) {
+            // TODO(ICU-21292)
+            // DO NOTHING
+            // UTS35 does not specifiy what should we do if we have extensions in the
+            // replacement. Currently we know only the following 4 "BCP47 LegacyRules" have
+            // extensions in them languageAlias:
+            //  i_default => en_x_i_default
+            //  i_enochian => und_x_i_enochian
+            //  i_mingo => see_x_i_mingo
+            //  zh_min => nan_x_zh_min
+            // But all of them are already changed by code inside ultag_parse() before
+            // hitting this code.
+        }
+
+        // Something changed by language alias data.
+        return true;
+    }
+    // Nothing changed by language alias data.
+    return false;
+}
+
+bool
+AliasReplacer::replaceTerritory(UVector& toBeFreed, UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    if (region == nullptr) {
+        // No region to search.
+        return false;
+    }
+    const char *replacement = data->territoryMap().get(region);
+    if (replacement == nullptr) {
+        // Found no replacement data for this region.
+        return false;
+    }
+    const char* replacedRegion = replacement;
+    const char* firstSpace = uprv_strchr(replacement, ' ');
+    if (firstSpace != nullptr) {
+        // If there are are more than one region in the replacement.
+        // We need to check which one match based on the language.
+        // Cannot use nullptr for language because that will construct
+        // the default locale, in that case, use "und" to get the correct
+        // locale.
+        Locale l(language == nullptr ? "und" : language, nullptr, script);
+        l.addLikelySubtags(status);
+        const char* likelyRegion = l.getCountry();
+        CharString* item = nullptr;
+        if (likelyRegion != nullptr && uprv_strlen(likelyRegion) > 0) {
+            size_t len = uprv_strlen(likelyRegion);
+            const char* foundInReplacement = uprv_strstr(replacement,
+                                                         likelyRegion);
+            if (foundInReplacement != nullptr) {
+                // Assuming the case there are no three letter region code in
+                // the replacement of territoryAlias
+                U_ASSERT(foundInReplacement == replacement ||
+                         *(foundInReplacement-1) == ' ');
+                U_ASSERT(foundInReplacement[len] == ' ' ||
+                         foundInReplacement[len] == '\0');
+                item = new CharString(foundInReplacement, (int32_t)len, status);
+            }
+        }
+        if (item == nullptr) {
+            item = new CharString(replacement,
+                                  (int32_t)(firstSpace - replacement), status);
+        }
+        if (U_FAILURE(status)) { return false; }
+        if (item == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return false;
+        }
+        replacedRegion = item->data();
+        toBeFreed.addElement(item, status);
+    }
+    U_ASSERT(!same(region, replacedRegion));
+    region = replacedRegion;
+    // The region is changed by data in territory alias.
+    return true;
+}
+
+bool
+AliasReplacer::replaceScript(UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    if (script == nullptr) {
+        // No script to search.
+        return false;
+    }
+    const char *replacement = data->scriptMap().get(script);
+    if (replacement == nullptr) {
+        // Found no replacement data for this script.
+        return false;
+    }
+    U_ASSERT(!same(script, replacement));
+    script = replacement;
+    // The script is changed by data in script alias.
+    return true;
+}
+
+bool
+AliasReplacer::replaceVariant(UErrorCode& status)
+{
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    // Since we may have more than one variant, we need to loop through them.
+    for (int32_t i = 0; i < variants.size(); i++) {
+        const char *variant = (const char*)(variants.elementAt(i));
+        const char *replacement = data->variantMap().get(variant);
+        if (replacement == nullptr) {
+            // Found no replacement data for this variant.
+            continue;
+        }
+        U_ASSERT((uprv_strlen(replacement) >= 5  &&
+                  uprv_strlen(replacement) <= 8) ||
+                 (uprv_strlen(replacement) == 4 &&
+                  replacement[0] >= '0' &&
+                  replacement[0] <= '9'));
+        if (!same(variant, replacement)) {
+            variants.setElementAt((void*)replacement, i);
+            // Special hack to handle hepburn-heploc => alalc97
+            if (uprv_strcmp(variant, "heploc") == 0) {
+                for (int32_t j = 0; j < variants.size(); j++) {
+                     if (uprv_strcmp((const char*)(variants.elementAt(j)),
+                                     "hepburn") == 0) {
+                         variants.removeElementAt(j);
+                     }
+                }
+            }
+            return true;
+        }
+    }
+    return false;
+}
+
+CharString&
+AliasReplacer::outputToString(
+    CharString& out, UErrorCode status)
+{
+    out.append(language, status);
+    if (notEmpty(script)) {
+        out.append(SEP_CHAR, status)
+            .append(script, status);
+    }
+    if (notEmpty(region)) {
+        out.append(SEP_CHAR, status)
+            .append(region, status);
+    }
+    if (variants.size() > 0) {
+        if (!notEmpty(script) && !notEmpty(region)) {
+          out.append(SEP_CHAR, status);
+        }
+        variants.sort([](UElement e1, UElement e2) -> int8_t {
+            return uprv_strcmp(
+                (const char*)e1.pointer, (const char*)e2.pointer);
+        }, status);
+        int32_t variantsStart = out.length();
+        for (int32_t i = 0; i < variants.size(); i++) {
+             out.append(SEP_CHAR, status)
+                 .append((const char*)((UVector*)variants.elementAt(i)),
+                         status);
+        }
+        T_CString_toUpperCase(out.data() + variantsStart);
+    }
+    if (notEmpty(extensions)) {
+        CharString tmp("und_", status);
+        tmp.append(extensions, status);
+        Locale tmpLocale(tmp.data());
+        // only support x extension inside CLDR for now.
+        U_ASSERT(extensions[0] == 'x');
+        out.append(tmpLocale.getName() + 1, status);
+    }
+    return out;
+}
+
+bool
+AliasReplacer::replace(const Locale& locale, CharString& out, UErrorCode status)
+{
+    data = AliasData::singleton(status);
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    U_ASSERT(data != nullptr);
+    out.clear();
+    language = locale.getLanguage();
+    if (!notEmpty(language)) {
+        language = nullptr;
+    }
+    script = locale.getScript();
+    if (!notEmpty(script)) {
+        script = nullptr;
+    }
+    region = locale.getCountry();
+    if (!notEmpty(region)) {
+        region = nullptr;
+    }
+    const char* variantsStr = locale.getVariant();
+    const char* extensionsStr = locale_getKeywordsStart(locale.getName());
+    CharString variantsBuff(variantsStr, -1, status);
+    if (!variantsBuff.isEmpty()) {
+        if (U_FAILURE(status)) { return false; }
+        char* start = variantsBuff.data();
+        T_CString_toLowerCase(start);
+        char* end;
+        while ((end = uprv_strchr(start, SEP_CHAR)) != nullptr &&
+               U_SUCCESS(status)) {
+            *end = NULL_CHAR;  // null terminate inside variantsBuff
+            variants.addElement(start, status);
+            start = end + 1;
+        }
+        variants.addElement(start, status);
+    }
+    if (U_FAILURE(status)) { return false; }
+
+    // Sort the variants
+    variants.sort([](UElement e1, UElement e2) -> int8_t {
+        return uprv_strcmp(
+            (const char*)e1.pointer, (const char*)e2.pointer);
+    }, status);
+
+    // A changed count to assert when loop too many times.
+    int changed = 0;
+    // A UVector to to hold CharString allocated by the replace* method
+    // and freed when out of scope from his function.
+    UVector stringsToBeFreed([](void *obj){ delete ((CharString*) obj); },
+                             nullptr, 10, status);
+    while (U_SUCCESS(status)) {
+        // Something wrong with the data cause looping here more than 10 times
+        // already.
+        U_ASSERT(changed < 5);
+        // From observation of key in data/misc/metadata.txt
+        // we know currently we only need to search in the following combination
+        // of fields for type in languageAlias:
+        // * lang_region_variant
+        // * lang_region
+        // * lang_variant
+        // * lang
+        // * und_variant
+        // This assumption is ensured by the U_ASSERT in readLanguageAlias
+        //
+        //                      lang  REGION variant
+        if (    replaceLanguage(true, true,  true,  stringsToBeFreed, status) ||
+                replaceLanguage(true, true,  false, stringsToBeFreed, status) ||
+                replaceLanguage(true, false, true,  stringsToBeFreed, status) ||
+                replaceLanguage(true, false, false, stringsToBeFreed, status) ||
+                replaceLanguage(false,false, true,  stringsToBeFreed, status) ||
+                replaceTerritory(stringsToBeFreed, status) ||
+                replaceScript(status) ||
+                replaceVariant(status)) {
+            // Some values in data is changed, try to match from the beginning
+            // again.
+            changed++;
+            continue;
+        }
+        // Nothing changed. Break out.
+        break;
+    }  // while(1)
+
+    if (U_FAILURE(status)) { return false; }
+    // Nothing changed and we know the order of the vaiants are not change
+    // because we have no variant or only one.
+    if (changed == 0 && variants.size() <= 1) {
+        return false;
+    }
+    outputToString(out, status);
+    if (extensionsStr != nullptr) {
+        out.append(extensionsStr, status);
+    }
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    // If the tag is not changed, return.
+    if (uprv_strcmp(out.data(), locale.getName()) == 0) {
+        U_ASSERT(changed == 0);
+        U_ASSERT(variants.size() > 1);
+        out.clear();
+        return false;
+    }
+    return true;
+}
+
+// Return true if the locale is changed during canonicalization.
+// The replaced value then will be put into out.
+bool
+canonicalizeLocale(const Locale& locale, CharString& out, UErrorCode& status)
+{
+    AliasReplacer replacer(status);
+    return replacer.replace(locale, out, status);
+}
+
+// Function to optimize for known cases without so we can skip the loading
+// of resources in the startup time until we really need it.
+bool
+isKnownCanonicalizedLocale(const char* locale, UErrorCode& status)
+{
+    if (    uprv_strcmp(locale, "c") == 0 ||
+            uprv_strcmp(locale, "en") == 0 ||
+            uprv_strcmp(locale, "en_US") == 0) {
+        return true;
+    }
+
+    // common well-known Canonicalized.
+    umtx_initOnce(gKnownCanonicalizedInitOnce,
+                  &loadKnownCanonicalized, status);
+    if (U_FAILURE(status)) {
+        return false;
+    }
+    U_ASSERT(gKnownCanonicalized != nullptr);
+    return uhash_geti(gKnownCanonicalized, locale) != 0;
+}
+
+}  // namespace
+
+// Function for testing.
+U_CAPI const char* const*
+ulocimp_getKnownCanonicalizedLocaleForTest(int32_t* length)
+{
+    *length = UPRV_LENGTHOF(KNOWN_CANONICALIZED);
+    return KNOWN_CANONICALIZED;
+}
+
+// Function for testing.
+U_CAPI bool
+ulocimp_isCanonicalizedLocaleForTest(const char* localeName)
+{
+    Locale l(localeName);
+    UErrorCode status = U_ZERO_ERROR;
+    CharString temp;
+    return !canonicalizeLocale(l, temp, status) && U_SUCCESS(status);
+}
+
+/*This function initializes a Locale from a C locale ID*/
+Locale& Locale::init(const char* localeID, UBool canonicalize)
+{
+    fIsBogus = FALSE;
+    /* Free our current storage */
+    if (baseName != fullName) {
+        uprv_free(baseName);
+    }
+    baseName = NULL;
+    if(fullName != fullNameBuffer) {
+        uprv_free(fullName);
+        fullName = fullNameBuffer;
+    }
+
+    // not a loop:
+    // just an easy way to have a common error-exit
+    // without goto and without another function
+    do {
+        char *separator;
+        char *field[5] = {0};
+        int32_t fieldLen[5] = {0};
+        int32_t fieldIdx;
+        int32_t variantField;
+        int32_t length;
+        UErrorCode err;
+
+        if(localeID == NULL) {
+            // not an error, just set the default locale
+            return *this = getDefault();
+        }
+
+        /* preset all fields to empty */
+        language[0] = script[0] = country[0] = 0;
+
+        // "canonicalize" the locale ID to ICU/Java format
+        err = U_ZERO_ERROR;
+        length = canonicalize ?
+            uloc_canonicalize(localeID, fullName, sizeof(fullNameBuffer), &err) :
+            uloc_getName(localeID, fullName, sizeof(fullNameBuffer), &err);
+
+        if(err == U_BUFFER_OVERFLOW_ERROR || length >= (int32_t)sizeof(fullNameBuffer)) {
+            /*Go to heap for the fullName if necessary*/
+            fullName = (char *)uprv_malloc(sizeof(char)*(length + 1));
+            if(fullName == 0) {
+                fullName = fullNameBuffer;
+                break; // error: out of memory
+            }
+            err = U_ZERO_ERROR;
+            length = canonicalize ?
+                uloc_canonicalize(localeID, fullName, length+1, &err) :
+                uloc_getName(localeID, fullName, length+1, &err);
+        }
+        if(U_FAILURE(err) || err == U_STRING_NOT_TERMINATED_WARNING) {
+            /* should never occur */
+            break;
+        }
+
+        variantBegin = length;
+
+        /* after uloc_getName/canonicalize() we know that only '_' are separators */
+        /* But _ could also appeared in timezone such as "en@timezone=America/Los_Angeles" */
+        separator = field[0] = fullName;
+        fieldIdx = 1;
+        char* at = uprv_strchr(fullName, '@');
+        while ((separator = uprv_strchr(field[fieldIdx-1], SEP_CHAR)) != 0 &&
+               fieldIdx < UPRV_LENGTHOF(field)-1 &&
+               (at == nullptr || separator < at)) {
+            field[fieldIdx] = separator + 1;
+            fieldLen[fieldIdx-1] = (int32_t)(separator - field[fieldIdx-1]);
+            fieldIdx++;
+        }
+        // variant may contain @foo or .foo POSIX cruft; remove it
+        separator = uprv_strchr(field[fieldIdx-1], '@');
+        char* sep2 = uprv_strchr(field[fieldIdx-1], '.');
+        if (separator!=NULL || sep2!=NULL) {
+            if (separator==NULL || (sep2!=NULL && separator > sep2)) {
+                separator = sep2;
+            }
+            fieldLen[fieldIdx-1] = (int32_t)(separator - field[fieldIdx-1]);
+        } else {
+            fieldLen[fieldIdx-1] = length - (int32_t)(field[fieldIdx-1] - fullName);
+        }
+
+        if (fieldLen[0] >= (int32_t)(sizeof(language)))
+        {
+            break; // error: the language field is too long
+        }
+
+        variantField = 1; /* Usually the 2nd one, except when a script or country is also used. */
+        if (fieldLen[0] > 0) {
+            /* We have a language */
+            uprv_memcpy(language, fullName, fieldLen[0]);
+            language[fieldLen[0]] = 0;
+        }
+        if (fieldLen[1] == 4 && uprv_isASCIILetter(field[1][0]) &&
+                uprv_isASCIILetter(field[1][1]) && uprv_isASCIILetter(field[1][2]) &&
+                uprv_isASCIILetter(field[1][3])) {
+            /* We have at least a script */
+            uprv_memcpy(script, field[1], fieldLen[1]);
+            script[fieldLen[1]] = 0;
+            variantField++;
+        }
+
+        if (fieldLen[variantField] == 2 || fieldLen[variantField] == 3) {
+            /* We have a country */
+            uprv_memcpy(country, field[variantField], fieldLen[variantField]);
+            country[fieldLen[variantField]] = 0;
+            variantField++;
+        } else if (fieldLen[variantField] == 0) {
+            variantField++; /* script or country empty but variant in next field (i.e. en__POSIX) */
+        }
+
+        if (fieldLen[variantField] > 0) {
+            /* We have a variant */
+            variantBegin = (int32_t)(field[variantField] - fullName);
+        }
+
+        err = U_ZERO_ERROR;
+        initBaseName(err);
+        if (U_FAILURE(err)) {
+            break;
+        }
+
+        if (canonicalize) {
+            if (!isKnownCanonicalizedLocale(fullName, err)) {
+                CharString replaced;
+                // Not sure it is already canonicalized
+                if (canonicalizeLocale(*this, replaced, err)) {
+                    U_ASSERT(U_SUCCESS(err));
+                    // If need replacement, call init again.
+                    init(replaced.data(), false);
+                }
+                if (U_FAILURE(err)) {
+                    break;
+                }
+            }
+        }   // if (canonicalize) {
+
+        // successful end of init()
+        return *this;
+    } while(0); /*loop doesn't iterate*/
+
+    // when an error occurs, then set this object to "bogus" (there is no UErrorCode here)
+    setToBogus();
+
+    return *this;
+}
+
+/*
+ * Set up the base name.
+ * If there are no key words, it's exactly the full name.
+ * If key words exist, it's the full name truncated at the '@' character.
+ * Need to set up both at init() and after setting a keyword.
+ */
+void
+Locale::initBaseName(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    U_ASSERT(baseName==NULL || baseName==fullName);
+    const char *atPtr = uprv_strchr(fullName, '@');
+    const char *eqPtr = uprv_strchr(fullName, '=');
+    if (atPtr && eqPtr && atPtr < eqPtr) {
+        // Key words exist.
+        int32_t baseNameLength = (int32_t)(atPtr - fullName);
+        baseName = (char *)uprv_malloc(baseNameLength + 1);
+        if (baseName == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        uprv_strncpy(baseName, fullName, baseNameLength);
+        baseName[baseNameLength] = 0;
+
+        // The original computation of variantBegin leaves it equal to the length
+        // of fullName if there is no variant.  It should instead be
+        // the length of the baseName.
+        if (variantBegin > baseNameLength) {
+            variantBegin = baseNameLength;
+        }
+    } else {
+        baseName = fullName;
+    }
+}
+
+
+int32_t
+Locale::hashCode() const
+{
+    return ustr_hashCharsN(fullName, static_cast<int32_t>(uprv_strlen(fullName)));
+}
+
+void
+Locale::setToBogus() {
+    /* Free our current storage */
+    if(baseName != fullName) {
+        uprv_free(baseName);
+    }
+    baseName = NULL;
+    if(fullName != fullNameBuffer) {
+        uprv_free(fullName);
+        fullName = fullNameBuffer;
+    }
+    *fullNameBuffer = 0;
+    *language = 0;
+    *script = 0;
+    *country = 0;
+    fIsBogus = TRUE;
+    variantBegin = 0;
+}
+
+const Locale& U_EXPORT2
+Locale::getDefault()
+{
+    {
+        Mutex lock(&gDefaultLocaleMutex);
+        if (gDefaultLocale != NULL) {
+            return *gDefaultLocale;
+        }
+    }
+    UErrorCode status = U_ZERO_ERROR;
+    return *locale_set_default_internal(NULL, status);
+}
+
+
+
+void U_EXPORT2
+Locale::setDefault( const   Locale&     newLocale,
+                            UErrorCode&  status)
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    /* Set the default from the full name string of the supplied locale.
+     * This is a convenient way to access the default locale caching mechanisms.
+     */
+    const char *localeID = newLocale.getName();
+    locale_set_default_internal(localeID, status);
+}
+
+void
+Locale::addLikelySubtags(UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    CharString maximizedLocaleID;
+    {
+        CharStringByteSink sink(&maximizedLocaleID);
+        ulocimp_addLikelySubtags(fullName, sink, &status);
+    }
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    init(maximizedLocaleID.data(), /*canonicalize=*/FALSE);
+    if (isBogus()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+void
+Locale::minimizeSubtags(UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    CharString minimizedLocaleID;
+    {
+        CharStringByteSink sink(&minimizedLocaleID);
+        ulocimp_minimizeSubtags(fullName, sink, &status);
+    }
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    init(minimizedLocaleID.data(), /*canonicalize=*/FALSE);
+    if (isBogus()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+void
+Locale::canonicalize(UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (isBogus()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    CharString uncanonicalized(fullName, status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    init(uncanonicalized.data(), /*canonicalize=*/TRUE);
+    if (isBogus()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+Locale U_EXPORT2
+Locale::forLanguageTag(StringPiece tag, UErrorCode& status)
+{
+    Locale result(Locale::eBOGUS);
+
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    // If a BCP 47 language tag is passed as the language parameter to the
+    // normal Locale constructor, it will actually fall back to invoking
+    // uloc_forLanguageTag() to parse it if it somehow is able to detect that
+    // the string actually is BCP 47. This works well for things like strings
+    // using BCP 47 extensions, but it does not at all work for things like
+    // legacy language tags (marked as “Type: grandfathered” in BCP 47,
+    // e.g., "en-GB-oed") which are possible to also
+    // interpret as ICU locale IDs and because of that won't trigger the BCP 47
+    // parsing. Therefore the code here explicitly calls uloc_forLanguageTag()
+    // and then Locale::init(), instead of just calling the normal constructor.
+
+    CharString localeID;
+    int32_t parsedLength;
+    {
+        CharStringByteSink sink(&localeID);
+        ulocimp_forLanguageTag(
+                tag.data(),
+                tag.length(),
+                sink,
+                &parsedLength,
+                &status);
+    }
+
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    if (parsedLength != tag.size()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return result;
+    }
+
+    result.init(localeID.data(), /*canonicalize=*/FALSE);
+    if (result.isBogus()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return result;
+}
+
+void
+Locale::toLanguageTag(ByteSink& sink, UErrorCode& status) const
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    if (fIsBogus) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    ulocimp_toLanguageTag(fullName, sink, /*strict=*/FALSE, &status);
+}
+
+Locale U_EXPORT2
+Locale::createFromName (const char *name)
+{
+    if (name) {
+        Locale l("");
+        l.init(name, FALSE);
+        return l;
+    }
+    else {
+        return getDefault();
+    }
+}
+
+Locale U_EXPORT2
+Locale::createCanonical(const char* name) {
+    Locale loc("");
+    loc.init(name, TRUE);
+    return loc;
+}
+
+const char *
+Locale::getISO3Language() const
+{
+    return uloc_getISO3Language(fullName);
+}
+
+
+const char *
+Locale::getISO3Country() const
+{
+    return uloc_getISO3Country(fullName);
+}
+
+/**
+ * Return the LCID value as specified in the "LocaleID" resource for this
+ * locale.  The LocaleID must be expressed as a hexadecimal number, from
+ * one to four digits.  If the LocaleID resource is not present, or is
+ * in an incorrect format, 0 is returned.  The LocaleID is for use in
+ * Windows (it is an LCID), but is available on all platforms.
+ */
+uint32_t
+Locale::getLCID() const
+{
+    return uloc_getLCID(fullName);
+}
+
+const char* const* U_EXPORT2 Locale::getISOCountries()
+{
+    return uloc_getISOCountries();
+}
+
+const char* const* U_EXPORT2 Locale::getISOLanguages()
+{
+    return uloc_getISOLanguages();
+}
+
+// Set the locale's data based on a posix id.
+void Locale::setFromPOSIXID(const char *posixID)
+{
+    init(posixID, TRUE);
+}
+
+const Locale & U_EXPORT2
+Locale::getRoot(void)
+{
+    return getLocale(eROOT);
+}
+
+const Locale & U_EXPORT2
+Locale::getEnglish(void)
+{
+    return getLocale(eENGLISH);
+}
+
+const Locale & U_EXPORT2
+Locale::getFrench(void)
+{
+    return getLocale(eFRENCH);
+}
+
+const Locale & U_EXPORT2
+Locale::getGerman(void)
+{
+    return getLocale(eGERMAN);
+}
+
+const Locale & U_EXPORT2
+Locale::getItalian(void)
+{
+    return getLocale(eITALIAN);
+}
+
+const Locale & U_EXPORT2
+Locale::getJapanese(void)
+{
+    return getLocale(eJAPANESE);
+}
+
+const Locale & U_EXPORT2
+Locale::getKorean(void)
+{
+    return getLocale(eKOREAN);
+}
+
+const Locale & U_EXPORT2
+Locale::getChinese(void)
+{
+    return getLocale(eCHINESE);
+}
+
+const Locale & U_EXPORT2
+Locale::getSimplifiedChinese(void)
+{
+    return getLocale(eCHINA);
+}
+
+const Locale & U_EXPORT2
+Locale::getTraditionalChinese(void)
+{
+    return getLocale(eTAIWAN);
+}
+
+
+const Locale & U_EXPORT2
+Locale::getFrance(void)
+{
+    return getLocale(eFRANCE);
+}
+
+const Locale & U_EXPORT2
+Locale::getGermany(void)
+{
+    return getLocale(eGERMANY);
+}
+
+const Locale & U_EXPORT2
+Locale::getItaly(void)
+{
+    return getLocale(eITALY);
+}
+
+const Locale & U_EXPORT2
+Locale::getJapan(void)
+{
+    return getLocale(eJAPAN);
+}
+
+const Locale & U_EXPORT2
+Locale::getKorea(void)
+{
+    return getLocale(eKOREA);
+}
+
+const Locale & U_EXPORT2
+Locale::getChina(void)
+{
+    return getLocale(eCHINA);
+}
+
+const Locale & U_EXPORT2
+Locale::getPRC(void)
+{
+    return getLocale(eCHINA);
+}
+
+const Locale & U_EXPORT2
+Locale::getTaiwan(void)
+{
+    return getLocale(eTAIWAN);
+}
+
+const Locale & U_EXPORT2
+Locale::getUK(void)
+{
+    return getLocale(eUK);
+}
+
+const Locale & U_EXPORT2
+Locale::getUS(void)
+{
+    return getLocale(eUS);
+}
+
+const Locale & U_EXPORT2
+Locale::getCanada(void)
+{
+    return getLocale(eCANADA);
+}
+
+const Locale & U_EXPORT2
+Locale::getCanadaFrench(void)
+{
+    return getLocale(eCANADA_FRENCH);
+}
+
+const Locale &
+Locale::getLocale(int locid)
+{
+    Locale *localeCache = getLocaleCache();
+    U_ASSERT((locid < eMAX_LOCALES)&&(locid>=0));
+    if (localeCache == NULL) {
+        // Failure allocating the locale cache.
+        //   The best we can do is return a NULL reference.
+        locid = 0;
+    }
+    return localeCache[locid]; /*operating on NULL*/
+}
+
+/*
+This function is defined this way in order to get around static
+initialization and static destruction.
+ */
+Locale *
+Locale::getLocaleCache(void)
+{
+    UErrorCode status = U_ZERO_ERROR;
+    umtx_initOnce(gLocaleCacheInitOnce, locale_init, status);
+    return gLocaleCache;
+}
+
+class KeywordEnumeration : public StringEnumeration {
+private:
+    char *keywords;
+    char *current;
+    int32_t length;
+    UnicodeString currUSKey;
+    static const char fgClassID;/* Warning this is used beyond the typical RTTI usage. */
+
+public:
+    static UClassID U_EXPORT2 getStaticClassID(void) { return (UClassID)&fgClassID; }
+    virtual UClassID getDynamicClassID(void) const { return getStaticClassID(); }
+public:
+    KeywordEnumeration(const char *keys, int32_t keywordLen, int32_t currentIndex, UErrorCode &status)
+        : keywords((char *)&fgClassID), current((char *)&fgClassID), length(0) {
+        if(U_SUCCESS(status) && keywordLen != 0) {
+            if(keys == NULL || keywordLen < 0) {
+                status = U_ILLEGAL_ARGUMENT_ERROR;
+            } else {
+                keywords = (char *)uprv_malloc(keywordLen+1);
+                if (keywords == NULL) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                }
+                else {
+                    uprv_memcpy(keywords, keys, keywordLen);
+                    keywords[keywordLen] = 0;
+                    current = keywords + currentIndex;
+                    length = keywordLen;
+                }
+            }
+        }
+    }
+
+    virtual ~KeywordEnumeration();
+
+    virtual StringEnumeration * clone() const
+    {
+        UErrorCode status = U_ZERO_ERROR;
+        return new KeywordEnumeration(keywords, length, (int32_t)(current - keywords), status);
+    }
+
+    virtual int32_t count(UErrorCode &/*status*/) const {
+        char *kw = keywords;
+        int32_t result = 0;
+        while(*kw) {
+            result++;
+            kw += uprv_strlen(kw)+1;
+        }
+        return result;
+    }
+
+    virtual const char* next(int32_t* resultLength, UErrorCode& status) {
+        const char* result;
+        int32_t len;
+        if(U_SUCCESS(status) && *current != 0) {
+            result = current;
+            len = (int32_t)uprv_strlen(current);
+            current += len+1;
+            if(resultLength != NULL) {
+                *resultLength = len;
+            }
+        } else {
+            if(resultLength != NULL) {
+                *resultLength = 0;
+            }
+            result = NULL;
+        }
+        return result;
+    }
+
+    virtual const UnicodeString* snext(UErrorCode& status) {
+        int32_t resultLength = 0;
+        const char *s = next(&resultLength, status);
+        return setChars(s, resultLength, status);
+    }
+
+    virtual void reset(UErrorCode& /*status*/) {
+        current = keywords;
+    }
+};
+
+const char KeywordEnumeration::fgClassID = '\0';
+
+KeywordEnumeration::~KeywordEnumeration() {
+    uprv_free(keywords);
+}
+
+// A wrapper around KeywordEnumeration that calls uloc_toUnicodeLocaleKey() in
+// the next() method for each keyword before returning it.
+class UnicodeKeywordEnumeration : public KeywordEnumeration {
+public:
+    using KeywordEnumeration::KeywordEnumeration;
+    virtual ~UnicodeKeywordEnumeration();
+
+    virtual const char* next(int32_t* resultLength, UErrorCode& status) {
+        const char* legacy_key = KeywordEnumeration::next(nullptr, status);
+        if (U_SUCCESS(status) && legacy_key != nullptr) {
+            const char* key = uloc_toUnicodeLocaleKey(legacy_key);
+            if (key == nullptr) {
+                status = U_ILLEGAL_ARGUMENT_ERROR;
+            } else {
+                if (resultLength != nullptr) {
+                    *resultLength = static_cast<int32_t>(uprv_strlen(key));
+                }
+                return key;
+            }
+        }
+        if (resultLength != nullptr) *resultLength = 0;
+        return nullptr;
+    }
+};
+
+// Out-of-line virtual destructor to serve as the "key function".
+UnicodeKeywordEnumeration::~UnicodeKeywordEnumeration() = default;
+
+StringEnumeration *
+Locale::createKeywords(UErrorCode &status) const
+{
+    StringEnumeration *result = NULL;
+
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    const char* variantStart = uprv_strchr(fullName, '@');
+    const char* assignment = uprv_strchr(fullName, '=');
+    if(variantStart) {
+        if(assignment > variantStart) {
+            CharString keywords;
+            CharStringByteSink sink(&keywords);
+            ulocimp_getKeywords(variantStart+1, '@', sink, FALSE, &status);
+            if (U_SUCCESS(status) && !keywords.isEmpty()) {
+                result = new KeywordEnumeration(keywords.data(), keywords.length(), 0, status);
+                if (!result) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                }
+            }
+        } else {
+            status = U_INVALID_FORMAT_ERROR;
+        }
+    }
+    return result;
+}
+
+StringEnumeration *
+Locale::createUnicodeKeywords(UErrorCode &status) const
+{
+    StringEnumeration *result = NULL;
+
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    const char* variantStart = uprv_strchr(fullName, '@');
+    const char* assignment = uprv_strchr(fullName, '=');
+    if(variantStart) {
+        if(assignment > variantStart) {
+            CharString keywords;
+            CharStringByteSink sink(&keywords);
+            ulocimp_getKeywords(variantStart+1, '@', sink, FALSE, &status);
+            if (U_SUCCESS(status) && !keywords.isEmpty()) {
+                result = new UnicodeKeywordEnumeration(keywords.data(), keywords.length(), 0, status);
+                if (!result) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                }
+            }
+        } else {
+            status = U_INVALID_FORMAT_ERROR;
+        }
+    }
+    return result;
+}
+
+int32_t
+Locale::getKeywordValue(const char* keywordName, char *buffer, int32_t bufLen, UErrorCode &status) const
+{
+    return uloc_getKeywordValue(fullName, keywordName, buffer, bufLen, &status);
+}
+
+void
+Locale::getKeywordValue(StringPiece keywordName, ByteSink& sink, UErrorCode& status) const {
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    if (fIsBogus) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    // TODO: Remove the need for a const char* to a NUL terminated buffer.
+    const CharString keywordName_nul(keywordName, status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    ulocimp_getKeywordValue(fullName, keywordName_nul.data(), sink, &status);
+}
+
+void
+Locale::getUnicodeKeywordValue(StringPiece keywordName,
+                               ByteSink& sink,
+                               UErrorCode& status) const {
+    // TODO: Remove the need for a const char* to a NUL terminated buffer.
+    const CharString keywordName_nul(keywordName, status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    const char* legacy_key = uloc_toLegacyKey(keywordName_nul.data());
+
+    if (legacy_key == nullptr) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    CharString legacy_value;
+    {
+        CharStringByteSink sink(&legacy_value);
+        getKeywordValue(legacy_key, sink, status);
+    }
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    const char* unicode_value = uloc_toUnicodeLocaleType(
+            keywordName_nul.data(), legacy_value.data());
+
+    if (unicode_value == nullptr) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    sink.Append(unicode_value, static_cast<int32_t>(uprv_strlen(unicode_value)));
+}
+
+void
+Locale::setKeywordValue(const char* keywordName, const char* keywordValue, UErrorCode &status)
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+    int32_t bufferLength = uprv_max((int32_t)(uprv_strlen(fullName) + 1), ULOC_FULLNAME_CAPACITY);
+    int32_t newLength = uloc_setKeywordValue(keywordName, keywordValue, fullName,
+                                             bufferLength, &status) + 1;
+    /* Handle the case the current buffer is not enough to hold the new id */
+    if (status == U_BUFFER_OVERFLOW_ERROR) {
+        U_ASSERT(newLength > bufferLength);
+        char* newFullName = (char *)uprv_malloc(newLength);
+        if (newFullName == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        uprv_strcpy(newFullName, fullName);
+        if (fullName != fullNameBuffer) {
+            // if full Name is already on the heap, need to free it.
+            uprv_free(fullName);
+        }
+        fullName = newFullName;
+        status = U_ZERO_ERROR;
+        uloc_setKeywordValue(keywordName, keywordValue, fullName, newLength, &status);
+    } else {
+        U_ASSERT(newLength <= bufferLength);
+    }
+    if (U_SUCCESS(status) && baseName == fullName) {
+        // May have added the first keyword, meaning that the fullName is no longer also the baseName.
+        initBaseName(status);
+    }
+}
+
+void
+Locale::setKeywordValue(StringPiece keywordName,
+                        StringPiece keywordValue,
+                        UErrorCode& status) {
+    // TODO: Remove the need for a const char* to a NUL terminated buffer.
+    const CharString keywordName_nul(keywordName, status);
+    const CharString keywordValue_nul(keywordValue, status);
+    setKeywordValue(keywordName_nul.data(), keywordValue_nul.data(), status);
+}
+
+void
+Locale::setUnicodeKeywordValue(StringPiece keywordName,
+                               StringPiece keywordValue,
+                               UErrorCode& status) {
+    // TODO: Remove the need for a const char* to a NUL terminated buffer.
+    const CharString keywordName_nul(keywordName, status);
+    const CharString keywordValue_nul(keywordValue, status);
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    const char* legacy_key = uloc_toLegacyKey(keywordName_nul.data());
+
+    if (legacy_key == nullptr) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    const char* legacy_value = nullptr;
+
+    if (!keywordValue_nul.isEmpty()) {
+        legacy_value =
+            uloc_toLegacyType(keywordName_nul.data(), keywordValue_nul.data());
+
+        if (legacy_value == nullptr) {
+            status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+    }
+
+    setKeywordValue(legacy_key, legacy_value, status);
+}
+
+const char *
+Locale::getBaseName() const {
+    return baseName;
+}
+
+Locale::Iterator::~Iterator() = default;
+
+//eof
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/loclikely.cpp b/thirdparty/icu4c/common/loclikely.cpp
new file mode 100644
index 0000000000..94a60aba3e
--- /dev/null
+++ b/thirdparty/icu4c/common/loclikely.cpp
@@ -0,0 +1,1410 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  loclikely.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010feb25
+*   created by: Markus W. Scherer
+*
+*   Code for likely and minimized locale subtags, separated out from other .cpp files
+*   that then do not depend on resource bundle code and likely-subtags data.
+*/
+
+#include "unicode/bytestream.h"
+#include "unicode/utypes.h"
+#include "unicode/locid.h"
+#include "unicode/putil.h"
+#include "unicode/uchar.h"
+#include "unicode/uloc.h"
+#include "unicode/ures.h"
+#include "unicode/uscript.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "ulocimp.h"
+#include "ustr_imp.h"
+
+/**
+ * These are the canonical strings for unknown languages, scripts and regions.
+ **/
+static const char* const unknownLanguage = "und";
+static const char* const unknownScript = "Zzzz";
+static const char* const unknownRegion = "ZZ";
+
+/**
+ * This function looks for the localeID in the likelySubtags resource.
+ *
+ * @param localeID The tag to find.
+ * @param buffer A buffer to hold the matching entry
+ * @param bufferLength The length of the output buffer
+ * @return A pointer to "buffer" if found, or a null pointer if not.
+ */
+static const char*  U_CALLCONV
+findLikelySubtags(const char* localeID,
+                  char* buffer,
+                  int32_t bufferLength,
+                  UErrorCode* err) {
+    const char* result = NULL;
+
+    if (!U_FAILURE(*err)) {
+        int32_t resLen = 0;
+        const UChar* s = NULL;
+        UErrorCode tmpErr = U_ZERO_ERROR;
+        icu::LocalUResourceBundlePointer subtags(ures_openDirect(NULL, "likelySubtags", &tmpErr));
+        if (U_SUCCESS(tmpErr)) {
+            icu::CharString und;
+            if (localeID != NULL) {
+                if (*localeID == '\0') {
+                    localeID = unknownLanguage;
+                } else if (*localeID == '_') {
+                    und.append(unknownLanguage, *err);
+                    und.append(localeID, *err);
+                    if (U_FAILURE(*err)) {
+                        return NULL;
+                    }
+                    localeID = und.data();
+                }
+            }
+            s = ures_getStringByKey(subtags.getAlias(), localeID, &resLen, &tmpErr);
+
+            if (U_FAILURE(tmpErr)) {
+                /*
+                 * If a resource is missing, it's not really an error, it's
+                 * just that we don't have any data for that particular locale ID.
+                 */
+                if (tmpErr != U_MISSING_RESOURCE_ERROR) {
+                    *err = tmpErr;
+                }
+            }
+            else if (resLen >= bufferLength) {
+                /* The buffer should never overflow. */
+                *err = U_INTERNAL_PROGRAM_ERROR;
+            }
+            else {
+                u_UCharsToChars(s, buffer, resLen + 1);
+                if (resLen >= 3 &&
+                    uprv_strnicmp(buffer, unknownLanguage, 3) == 0 &&
+                    (resLen == 3 || buffer[3] == '_')) {
+                    uprv_memmove(buffer, buffer + 3, resLen - 3 + 1);
+                }
+                result = buffer;
+            }
+        } else {
+            *err = tmpErr;
+        }
+    }
+
+    return result;
+}
+
+/**
+ * Append a tag to a buffer, adding the separator if necessary.  The buffer
+ * must be large enough to contain the resulting tag plus any separator
+ * necessary. The tag must not be a zero-length string.
+ *
+ * @param tag The tag to add.
+ * @param tagLength The length of the tag.
+ * @param buffer The output buffer.
+ * @param bufferLength The length of the output buffer.  This is an input/ouput parameter.
+ **/
+static void U_CALLCONV
+appendTag(
+    const char* tag,
+    int32_t tagLength,
+    char* buffer,
+    int32_t* bufferLength,
+    UBool withSeparator) {
+
+    if (withSeparator) {
+        buffer[*bufferLength] = '_';
+        ++(*bufferLength);
+    }
+
+    uprv_memmove(
+        &buffer[*bufferLength],
+        tag,
+        tagLength);
+
+    *bufferLength += tagLength;
+}
+
+/**
+ * Create a tag string from the supplied parameters.  The lang, script and region
+ * parameters may be NULL pointers. If they are, their corresponding length parameters
+ * must be less than or equal to 0.
+ *
+ * If any of the language, script or region parameters are empty, and the alternateTags
+ * parameter is not NULL, it will be parsed for potential language, script and region tags
+ * to be used when constructing the new tag.  If the alternateTags parameter is NULL, or
+ * it contains no language tag, the default tag for the unknown language is used.
+ *
+ * If the length of the new string exceeds the capacity of the output buffer, 
+ * the function copies as many bytes to the output buffer as it can, and returns
+ * the error U_BUFFER_OVERFLOW_ERROR.
+ *
+ * If an illegal argument is provided, the function returns the error
+ * U_ILLEGAL_ARGUMENT_ERROR.
+ *
+ * Note that this function can return the warning U_STRING_NOT_TERMINATED_WARNING if
+ * the tag string fits in the output buffer, but the null terminator doesn't.
+ *
+ * @param lang The language tag to use.
+ * @param langLength The length of the language tag.
+ * @param script The script tag to use.
+ * @param scriptLength The length of the script tag.
+ * @param region The region tag to use.
+ * @param regionLength The length of the region tag.
+ * @param trailing Any trailing data to append to the new tag.
+ * @param trailingLength The length of the trailing data.
+ * @param alternateTags A string containing any alternate tags.
+ * @param sink The output sink receiving the tag string.
+ * @param err A pointer to a UErrorCode for error reporting.
+ **/
+static void U_CALLCONV
+createTagStringWithAlternates(
+    const char* lang,
+    int32_t langLength,
+    const char* script,
+    int32_t scriptLength,
+    const char* region,
+    int32_t regionLength,
+    const char* trailing,
+    int32_t trailingLength,
+    const char* alternateTags,
+    icu::ByteSink& sink,
+    UErrorCode* err) {
+
+    if (U_FAILURE(*err)) {
+        goto error;
+    }
+    else if (langLength >= ULOC_LANG_CAPACITY ||
+             scriptLength >= ULOC_SCRIPT_CAPACITY ||
+             regionLength >= ULOC_COUNTRY_CAPACITY) {
+        goto error;
+    }
+    else {
+        /**
+         * ULOC_FULLNAME_CAPACITY will provide enough capacity
+         * that we can build a string that contains the language,
+         * script and region code without worrying about overrunning
+         * the user-supplied buffer.
+         **/
+        char tagBuffer[ULOC_FULLNAME_CAPACITY];
+        int32_t tagLength = 0;
+        UBool regionAppended = FALSE;
+
+        if (langLength > 0) {
+            appendTag(
+                lang,
+                langLength,
+                tagBuffer,
+                &tagLength,
+                /*withSeparator=*/FALSE);
+        }
+        else if (alternateTags == NULL) {
+            /*
+             * Use the empty string for an unknown language, if
+             * we found no language.
+             */
+        }
+        else {
+            /*
+             * Parse the alternateTags string for the language.
+             */
+            char alternateLang[ULOC_LANG_CAPACITY];
+            int32_t alternateLangLength = sizeof(alternateLang);
+
+            alternateLangLength =
+                uloc_getLanguage(
+                    alternateTags,
+                    alternateLang,
+                    alternateLangLength,
+                    err);
+            if(U_FAILURE(*err) ||
+                alternateLangLength >= ULOC_LANG_CAPACITY) {
+                goto error;
+            }
+            else if (alternateLangLength == 0) {
+                /*
+                 * Use the empty string for an unknown language, if
+                 * we found no language.
+                 */
+            }
+            else {
+                appendTag(
+                    alternateLang,
+                    alternateLangLength,
+                    tagBuffer,
+                    &tagLength,
+                    /*withSeparator=*/FALSE);
+            }
+        }
+
+        if (scriptLength > 0) {
+            appendTag(
+                script,
+                scriptLength,
+                tagBuffer,
+                &tagLength,
+                /*withSeparator=*/TRUE);
+        }
+        else if (alternateTags != NULL) {
+            /*
+             * Parse the alternateTags string for the script.
+             */
+            char alternateScript[ULOC_SCRIPT_CAPACITY];
+
+            const int32_t alternateScriptLength =
+                uloc_getScript(
+                    alternateTags,
+                    alternateScript,
+                    sizeof(alternateScript),
+                    err);
+
+            if (U_FAILURE(*err) ||
+                alternateScriptLength >= ULOC_SCRIPT_CAPACITY) {
+                goto error;
+            }
+            else if (alternateScriptLength > 0) {
+                appendTag(
+                    alternateScript,
+                    alternateScriptLength,
+                    tagBuffer,
+                    &tagLength,
+                    /*withSeparator=*/TRUE);
+            }
+        }
+
+        if (regionLength > 0) {
+            appendTag(
+                region,
+                regionLength,
+                tagBuffer,
+                &tagLength,
+                /*withSeparator=*/TRUE);
+
+            regionAppended = TRUE;
+        }
+        else if (alternateTags != NULL) {
+            /*
+             * Parse the alternateTags string for the region.
+             */
+            char alternateRegion[ULOC_COUNTRY_CAPACITY];
+
+            const int32_t alternateRegionLength =
+                uloc_getCountry(
+                    alternateTags,
+                    alternateRegion,
+                    sizeof(alternateRegion),
+                    err);
+            if (U_FAILURE(*err) ||
+                alternateRegionLength >= ULOC_COUNTRY_CAPACITY) {
+                goto error;
+            }
+            else if (alternateRegionLength > 0) {
+                appendTag(
+                    alternateRegion,
+                    alternateRegionLength,
+                    tagBuffer,
+                    &tagLength,
+                    /*withSeparator=*/TRUE);
+
+                regionAppended = TRUE;
+            }
+        }
+
+        /**
+         * Copy the partial tag from our internal buffer to the supplied
+         * target.
+         **/
+        sink.Append(tagBuffer, tagLength);
+
+        if (trailingLength > 0) {
+            if (*trailing != '@') {
+                sink.Append("_", 1);
+                if (!regionAppended) {
+                    /* extra separator is required */
+                    sink.Append("_", 1);
+                }
+            }
+
+            /*
+             * Copy the trailing data into the supplied buffer.
+             */
+            sink.Append(trailing, trailingLength);
+        }
+
+        return;
+    }
+
+error:
+
+    /**
+     * An overflow indicates the locale ID passed in
+     * is ill-formed.  If we got here, and there was
+     * no previous error, it's an implicit overflow.
+     **/
+    if (*err ==  U_BUFFER_OVERFLOW_ERROR ||
+        U_SUCCESS(*err)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+/**
+ * Create a tag string from the supplied parameters.  The lang, script and region
+ * parameters may be NULL pointers. If they are, their corresponding length parameters
+ * must be less than or equal to 0.  If the lang parameter is an empty string, the
+ * default value for an unknown language is written to the output buffer.
+ *
+ * If the length of the new string exceeds the capacity of the output buffer, 
+ * the function copies as many bytes to the output buffer as it can, and returns
+ * the error U_BUFFER_OVERFLOW_ERROR.
+ *
+ * If an illegal argument is provided, the function returns the error
+ * U_ILLEGAL_ARGUMENT_ERROR.
+ *
+ * @param lang The language tag to use.
+ * @param langLength The length of the language tag.
+ * @param script The script tag to use.
+ * @param scriptLength The length of the script tag.
+ * @param region The region tag to use.
+ * @param regionLength The length of the region tag.
+ * @param trailing Any trailing data to append to the new tag.
+ * @param trailingLength The length of the trailing data.
+ * @param sink The output sink receiving the tag string.
+ * @param err A pointer to a UErrorCode for error reporting.
+ **/
+static void U_CALLCONV
+createTagString(
+    const char* lang,
+    int32_t langLength,
+    const char* script,
+    int32_t scriptLength,
+    const char* region,
+    int32_t regionLength,
+    const char* trailing,
+    int32_t trailingLength,
+    icu::ByteSink& sink,
+    UErrorCode* err)
+{
+    createTagStringWithAlternates(
+                lang,
+                langLength,
+                script,
+                scriptLength,
+                region,
+                regionLength,
+                trailing,
+                trailingLength,
+                NULL,
+                sink,
+                err);
+}
+
+/**
+ * Parse the language, script, and region subtags from a tag string, and copy the
+ * results into the corresponding output parameters. The buffers are null-terminated,
+ * unless overflow occurs.
+ *
+ * The langLength, scriptLength, and regionLength parameters are input/output
+ * parameters, and must contain the capacity of their corresponding buffers on
+ * input.  On output, they will contain the actual length of the buffers, not
+ * including the null terminator.
+ *
+ * If the length of any of the output subtags exceeds the capacity of the corresponding
+ * buffer, the function copies as many bytes to the output buffer as it can, and returns
+ * the error U_BUFFER_OVERFLOW_ERROR.  It will not parse any more subtags once overflow
+ * occurs.
+ *
+ * If an illegal argument is provided, the function returns the error
+ * U_ILLEGAL_ARGUMENT_ERROR.
+ *
+ * @param localeID The locale ID to parse.
+ * @param lang The language tag buffer.
+ * @param langLength The length of the language tag.
+ * @param script The script tag buffer.
+ * @param scriptLength The length of the script tag.
+ * @param region The region tag buffer.
+ * @param regionLength The length of the region tag.
+ * @param err A pointer to a UErrorCode for error reporting.
+ * @return The number of chars of the localeID parameter consumed.
+ **/
+static int32_t U_CALLCONV
+parseTagString(
+    const char* localeID,
+    char* lang,
+    int32_t* langLength,
+    char* script,
+    int32_t* scriptLength,
+    char* region,
+    int32_t* regionLength,
+    UErrorCode* err)
+{
+    const char* position = localeID;
+    int32_t subtagLength = 0;
+
+    if(U_FAILURE(*err) ||
+       localeID == NULL ||
+       lang == NULL ||
+       langLength == NULL ||
+       script == NULL ||
+       scriptLength == NULL ||
+       region == NULL ||
+       regionLength == NULL) {
+        goto error;
+    }
+
+    subtagLength = ulocimp_getLanguage(position, &position, *err).extract(lang, *langLength, *err);
+
+    /*
+     * Note that we explicit consider U_STRING_NOT_TERMINATED_WARNING
+     * to be an error, because it indicates the user-supplied tag is
+     * not well-formed.
+     */
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    *langLength = subtagLength;
+
+    /*
+     * If no language was present, use the empty string instead.
+     * Otherwise, move past any separator.
+     */
+    if (_isIDSeparator(*position)) {
+        ++position;
+    }
+
+    subtagLength = ulocimp_getScript(position, &position, *err).extract(script, *scriptLength, *err);
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    *scriptLength = subtagLength;
+
+    if (*scriptLength > 0) {
+        if (uprv_strnicmp(script, unknownScript, *scriptLength) == 0) {
+            /**
+             * If the script part is the "unknown" script, then don't return it.
+             **/
+            *scriptLength = 0;
+        }
+
+        /*
+         * Move past any separator.
+         */
+        if (_isIDSeparator(*position)) {
+            ++position;
+        }    
+    }
+
+    subtagLength = ulocimp_getCountry(position, &position, *err).extract(region, *regionLength, *err);
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    *regionLength = subtagLength;
+
+    if (*regionLength > 0) {
+        if (uprv_strnicmp(region, unknownRegion, *regionLength) == 0) {
+            /**
+             * If the region part is the "unknown" region, then don't return it.
+             **/
+            *regionLength = 0;
+        }
+    } else if (*position != 0 && *position != '@') {
+        /* back up over consumed trailing separator */
+        --position;
+    }
+
+exit:
+
+    return (int32_t)(position - localeID);
+
+error:
+
+    /**
+     * If we get here, we have no explicit error, it's the result of an
+     * illegal argument.
+     **/
+    if (!U_FAILURE(*err)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+
+    goto exit;
+}
+
+static UBool U_CALLCONV
+createLikelySubtagsString(
+    const char* lang,
+    int32_t langLength,
+    const char* script,
+    int32_t scriptLength,
+    const char* region,
+    int32_t regionLength,
+    const char* variants,
+    int32_t variantsLength,
+    icu::ByteSink& sink,
+    UErrorCode* err) {
+    /**
+     * ULOC_FULLNAME_CAPACITY will provide enough capacity
+     * that we can build a string that contains the language,
+     * script and region code without worrying about overrunning
+     * the user-supplied buffer.
+     **/
+    char likelySubtagsBuffer[ULOC_FULLNAME_CAPACITY];
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    /**
+     * Try the language with the script and region first.
+     **/
+    if (scriptLength > 0 && regionLength > 0) {
+
+        const char* likelySubtags = NULL;
+
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink sink(&tagBuffer);
+            createTagString(
+                lang,
+                langLength,
+                script,
+                scriptLength,
+                region,
+                regionLength,
+                NULL,
+                0,
+                sink,
+                err);
+        }
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        likelySubtags =
+            findLikelySubtags(
+                tagBuffer.data(),
+                likelySubtagsBuffer,
+                sizeof(likelySubtagsBuffer),
+                err);
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        if (likelySubtags != NULL) {
+            /* Always use the language tag from the
+               maximal string, since it may be more
+               specific than the one provided. */
+            createTagStringWithAlternates(
+                        NULL,
+                        0,
+                        NULL,
+                        0,
+                        NULL,
+                        0,
+                        variants,
+                        variantsLength,
+                        likelySubtags,
+                        sink,
+                        err);
+            return TRUE;
+        }
+    }
+
+    /**
+     * Try the language with just the script.
+     **/
+    if (scriptLength > 0) {
+
+        const char* likelySubtags = NULL;
+
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink sink(&tagBuffer);
+            createTagString(
+                lang,
+                langLength,
+                script,
+                scriptLength,
+                NULL,
+                0,
+                NULL,
+                0,
+                sink,
+                err);
+        }
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        likelySubtags =
+            findLikelySubtags(
+                tagBuffer.data(),
+                likelySubtagsBuffer,
+                sizeof(likelySubtagsBuffer),
+                err);
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        if (likelySubtags != NULL) {
+            /* Always use the language tag from the
+               maximal string, since it may be more
+               specific than the one provided. */
+            createTagStringWithAlternates(
+                        NULL,
+                        0,
+                        NULL,
+                        0,
+                        region,
+                        regionLength,
+                        variants,
+                        variantsLength,
+                        likelySubtags,
+                        sink,
+                        err);
+            return TRUE;
+        }
+    }
+
+    /**
+     * Try the language with just the region.
+     **/
+    if (regionLength > 0) {
+
+        const char* likelySubtags = NULL;
+
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink sink(&tagBuffer);
+            createTagString(
+                lang,
+                langLength,
+                NULL,
+                0,
+                region,
+                regionLength,
+                NULL,
+                0,
+                sink,
+                err);
+        }
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        likelySubtags =
+            findLikelySubtags(
+                tagBuffer.data(),
+                likelySubtagsBuffer,
+                sizeof(likelySubtagsBuffer),
+                err);
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        if (likelySubtags != NULL) {
+            /* Always use the language tag from the
+               maximal string, since it may be more
+               specific than the one provided. */
+            createTagStringWithAlternates(
+                        NULL,
+                        0,
+                        script,
+                        scriptLength,
+                        NULL,
+                        0,
+                        variants,
+                        variantsLength,
+                        likelySubtags,
+                        sink,
+                        err);
+            return TRUE;
+        }
+    }
+
+    /**
+     * Finally, try just the language.
+     **/
+    {
+        const char* likelySubtags = NULL;
+
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink sink(&tagBuffer);
+            createTagString(
+                lang,
+                langLength,
+                NULL,
+                0,
+                NULL,
+                0,
+                NULL,
+                0,
+                sink,
+                err);
+        }
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        likelySubtags =
+            findLikelySubtags(
+                tagBuffer.data(),
+                likelySubtagsBuffer,
+                sizeof(likelySubtagsBuffer),
+                err);
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+
+        if (likelySubtags != NULL) {
+            /* Always use the language tag from the
+               maximal string, since it may be more
+               specific than the one provided. */
+            createTagStringWithAlternates(
+                        NULL,
+                        0,
+                        script,
+                        scriptLength,
+                        region,
+                        regionLength,
+                        variants,
+                        variantsLength,
+                        likelySubtags,
+                        sink,
+                        err);
+            return TRUE;
+        }
+    }
+
+    return FALSE;
+
+error:
+
+    if (!U_FAILURE(*err)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+
+    return FALSE;
+}
+
+#define CHECK_TRAILING_VARIANT_SIZE(trailing, trailingLength) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t count = 0; \
+    int32_t i; \
+    for (i = 0; i < trailingLength; i++) { \
+        if (trailing[i] == '-' || trailing[i] == '_') { \
+            count = 0; \
+            if (count > 8) { \
+                goto error; \
+            } \
+        } else if (trailing[i] == '@') { \
+            break; \
+        } else if (count > 8) { \
+            goto error; \
+        } else { \
+            count++; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+static UBool
+_uloc_addLikelySubtags(const char* localeID,
+                       icu::ByteSink& sink,
+                       UErrorCode* err) {
+    char lang[ULOC_LANG_CAPACITY];
+    int32_t langLength = sizeof(lang);
+    char script[ULOC_SCRIPT_CAPACITY];
+    int32_t scriptLength = sizeof(script);
+    char region[ULOC_COUNTRY_CAPACITY];
+    int32_t regionLength = sizeof(region);
+    const char* trailing = "";
+    int32_t trailingLength = 0;
+    int32_t trailingIndex = 0;
+    UBool success = FALSE;
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+    if (localeID == NULL) {
+        goto error;
+    }
+
+    trailingIndex = parseTagString(
+        localeID,
+        lang,
+        &langLength,
+        script,
+        &scriptLength,
+        region,
+        &regionLength,
+        err);
+    if(U_FAILURE(*err)) {
+        /* Overflow indicates an illegal argument error */
+        if (*err == U_BUFFER_OVERFLOW_ERROR) {
+            *err = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+
+        goto error;
+    }
+
+    /* Find the length of the trailing portion. */
+    while (_isIDSeparator(localeID[trailingIndex])) {
+        trailingIndex++;
+    }
+    trailing = &localeID[trailingIndex];
+    trailingLength = (int32_t)uprv_strlen(trailing);
+
+    CHECK_TRAILING_VARIANT_SIZE(trailing, trailingLength);
+
+    success =
+        createLikelySubtagsString(
+            lang,
+            langLength,
+            script,
+            scriptLength,
+            region,
+            regionLength,
+            trailing,
+            trailingLength,
+            sink,
+            err);
+
+    if (!success) {
+        const int32_t localIDLength = (int32_t)uprv_strlen(localeID);
+
+        /*
+         * If we get here, we need to return localeID.
+         */
+        sink.Append(localeID, localIDLength);
+    }
+
+    return success;
+
+error:
+
+    if (!U_FAILURE(*err)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return FALSE;
+}
+
+// Add likely subtags to the sink
+// return true if the value in the sink is produced by a match during the lookup
+// return false if the value in the sink is the same as input because there are
+// no match after the lookup.
+static UBool _ulocimp_addLikelySubtags(const char*, icu::ByteSink&, UErrorCode*);
+
+static void
+_uloc_minimizeSubtags(const char* localeID,
+                      icu::ByteSink& sink,
+                      UErrorCode* err) {
+    icu::CharString maximizedTagBuffer;
+
+    char lang[ULOC_LANG_CAPACITY];
+    int32_t langLength = sizeof(lang);
+    char script[ULOC_SCRIPT_CAPACITY];
+    int32_t scriptLength = sizeof(script);
+    char region[ULOC_COUNTRY_CAPACITY];
+    int32_t regionLength = sizeof(region);
+    const char* trailing = "";
+    int32_t trailingLength = 0;
+    int32_t trailingIndex = 0;
+    UBool successGetMax = FALSE;
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+    else if (localeID == NULL) {
+        goto error;
+    }
+
+    trailingIndex =
+        parseTagString(
+            localeID,
+            lang,
+            &langLength,
+            script,
+            &scriptLength,
+            region,
+            &regionLength,
+            err);
+    if(U_FAILURE(*err)) {
+
+        /* Overflow indicates an illegal argument error */
+        if (*err == U_BUFFER_OVERFLOW_ERROR) {
+            *err = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+
+        goto error;
+    }
+
+    /* Find the spot where the variants or the keywords begin, if any. */
+    while (_isIDSeparator(localeID[trailingIndex])) {
+        trailingIndex++;
+    }
+    trailing = &localeID[trailingIndex];
+    trailingLength = (int32_t)uprv_strlen(trailing);
+
+    CHECK_TRAILING_VARIANT_SIZE(trailing, trailingLength);
+
+    {
+        icu::CharString base;
+        {
+            icu::CharStringByteSink baseSink(&base);
+            createTagString(
+                lang,
+                langLength,
+                script,
+                scriptLength,
+                region,
+                regionLength,
+                NULL,
+                0,
+                baseSink,
+                err);
+        }
+
+        /**
+         * First, we need to first get the maximization
+         * from AddLikelySubtags.
+         **/
+        {
+            icu::CharStringByteSink maxSink(&maximizedTagBuffer);
+            successGetMax = _ulocimp_addLikelySubtags(base.data(), maxSink, err);
+        }
+    }
+
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    if (!successGetMax) {
+        /**
+         * If we got here, return the locale ID parameter unchanged.
+         **/
+        const int32_t localeIDLength = (int32_t)uprv_strlen(localeID);
+        sink.Append(localeID, localeIDLength);
+        return;
+    }
+
+    // In the following, the lang, script, region are referring to those in
+    // the maximizedTagBuffer, not the one in the localeID.
+    langLength = sizeof(lang);
+    scriptLength = sizeof(script);
+    regionLength = sizeof(region);
+    parseTagString(
+        maximizedTagBuffer.data(),
+        lang,
+        &langLength,
+        script,
+        &scriptLength,
+        region,
+        &regionLength,
+        err);
+    if(U_FAILURE(*err)) {
+        goto error;
+    }
+
+    /**
+     * Start first with just the language.
+     **/
+    {
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink tagSink(&tagBuffer);
+            createLikelySubtagsString(
+                lang,
+                langLength,
+                NULL,
+                0,
+                NULL,
+                0,
+                NULL,
+                0,
+                tagSink,
+                err);
+        }
+
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+        else if (!tagBuffer.isEmpty() &&
+                 uprv_strnicmp(
+                    maximizedTagBuffer.data(),
+                    tagBuffer.data(),
+                    tagBuffer.length()) == 0) {
+
+            createTagString(
+                        lang,
+                        langLength,
+                        NULL,
+                        0,
+                        NULL,
+                        0,
+                        trailing,
+                        trailingLength,
+                        sink,
+                        err);
+            return;
+        }
+    }
+
+    /**
+     * Next, try the language and region.
+     **/
+    if (regionLength > 0) {
+
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink tagSink(&tagBuffer);
+            createLikelySubtagsString(
+                lang,
+                langLength,
+                NULL,
+                0,
+                region,
+                regionLength,
+                NULL,
+                0,
+                tagSink,
+                err);
+        }
+
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+        else if (!tagBuffer.isEmpty() &&
+                 uprv_strnicmp(
+                    maximizedTagBuffer.data(),
+                    tagBuffer.data(),
+                    tagBuffer.length()) == 0) {
+
+            createTagString(
+                        lang,
+                        langLength,
+                        NULL,
+                        0,
+                        region,
+                        regionLength,
+                        trailing,
+                        trailingLength,
+                        sink,
+                        err);
+            return;
+        }
+    }
+
+    /**
+     * Finally, try the language and script.  This is our last chance,
+     * since trying with all three subtags would only yield the
+     * maximal version that we already have.
+     **/
+    if (scriptLength > 0) {
+        icu::CharString tagBuffer;
+        {
+            icu::CharStringByteSink tagSink(&tagBuffer);
+            createLikelySubtagsString(
+                lang,
+                langLength,
+                script,
+                scriptLength,
+                NULL,
+                0,
+                NULL,
+                0,
+                tagSink,
+                err);
+        }
+
+        if(U_FAILURE(*err)) {
+            goto error;
+        }
+        else if (!tagBuffer.isEmpty() &&
+                 uprv_strnicmp(
+                    maximizedTagBuffer.data(),
+                    tagBuffer.data(),
+                    tagBuffer.length()) == 0) {
+
+            createTagString(
+                        lang,
+                        langLength,
+                        script,
+                        scriptLength,
+                        NULL,
+                        0,
+                        trailing,
+                        trailingLength,
+                        sink,
+                        err);
+            return;
+        }
+    }
+
+    {
+        /**
+         * If we got here, return the max + trail.
+         **/
+        createTagString(
+                    lang,
+                    langLength,
+                    script,
+                    scriptLength,
+                    region,
+                    regionLength,
+                    trailing,
+                    trailingLength,
+                    sink,
+                    err);
+        return;
+    }
+
+error:
+
+    if (!U_FAILURE(*err)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static UBool
+do_canonicalize(const char*    localeID,
+         char* buffer,
+         int32_t bufferCapacity,
+         UErrorCode* err)
+{
+    uloc_canonicalize(
+        localeID,
+        buffer,
+        bufferCapacity,
+        err);
+
+    if (*err == U_STRING_NOT_TERMINATED_WARNING ||
+        *err == U_BUFFER_OVERFLOW_ERROR) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+
+        return FALSE;
+    }
+    else if (U_FAILURE(*err)) {
+
+        return FALSE;
+    }
+    else {
+        return TRUE;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_addLikelySubtags(const char* localeID,
+                      char* maximizedLocaleID,
+                      int32_t maximizedLocaleIDCapacity,
+                      UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    icu::CheckedArrayByteSink sink(
+            maximizedLocaleID, maximizedLocaleIDCapacity);
+
+    ulocimp_addLikelySubtags(localeID, sink, status);
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*status)) {
+        return sink.Overflowed() ? reslen : -1;
+    }
+
+    if (sink.Overflowed()) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(
+                maximizedLocaleID, maximizedLocaleIDCapacity, reslen, status);
+    }
+
+    return reslen;
+}
+
+static UBool
+_ulocimp_addLikelySubtags(const char* localeID,
+                          icu::ByteSink& sink,
+                          UErrorCode* status) {
+    char localeBuffer[ULOC_FULLNAME_CAPACITY];
+
+    if (do_canonicalize(localeID, localeBuffer, sizeof localeBuffer, status)) {
+        return _uloc_addLikelySubtags(localeBuffer, sink, status);
+    }
+    return FALSE;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_addLikelySubtags(const char* localeID,
+                         icu::ByteSink& sink,
+                         UErrorCode* status) {
+    _ulocimp_addLikelySubtags(localeID, sink, status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_minimizeSubtags(const char* localeID,
+                     char* minimizedLocaleID,
+                     int32_t minimizedLocaleIDCapacity,
+                     UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    icu::CheckedArrayByteSink sink(
+            minimizedLocaleID, minimizedLocaleIDCapacity);
+
+    ulocimp_minimizeSubtags(localeID, sink, status);
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*status)) {
+        return sink.Overflowed() ? reslen : -1;
+    }
+
+    if (sink.Overflowed()) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(
+                minimizedLocaleID, minimizedLocaleIDCapacity, reslen, status);
+    }
+
+    return reslen;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_minimizeSubtags(const char* localeID,
+                        icu::ByteSink& sink,
+                        UErrorCode* status) {
+    char localeBuffer[ULOC_FULLNAME_CAPACITY];
+
+    if (do_canonicalize(localeID, localeBuffer, sizeof localeBuffer, status)) {
+        _uloc_minimizeSubtags(localeBuffer, sink, status);
+    }
+}
+
+// Pairs of (language subtag, + or -) for finding out fast if common languages
+// are LTR (minus) or RTL (plus).
+static const char LANG_DIR_STRING[] =
+        "root-en-es-pt-zh-ja-ko-de-fr-it-ar+he+fa+ru-nl-pl-th-tr-";
+
+// Implemented here because this calls ulocimp_addLikelySubtags().
+U_CAPI UBool U_EXPORT2
+uloc_isRightToLeft(const char *locale) {
+    UErrorCode errorCode = U_ZERO_ERROR;
+    char script[8];
+    int32_t scriptLength = uloc_getScript(locale, script, UPRV_LENGTHOF(script), &errorCode);
+    if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING ||
+            scriptLength == 0) {
+        // Fastpath: We know the likely scripts and their writing direction
+        // for some common languages.
+        errorCode = U_ZERO_ERROR;
+        char lang[8];
+        int32_t langLength = uloc_getLanguage(locale, lang, UPRV_LENGTHOF(lang), &errorCode);
+        if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING) {
+            return FALSE;
+        }
+        if (langLength > 0) {
+            const char* langPtr = uprv_strstr(LANG_DIR_STRING, lang);
+            if (langPtr != NULL) {
+                switch (langPtr[langLength]) {
+                case '-': return FALSE;
+                case '+': return TRUE;
+                default: break;  // partial match of a longer code
+                }
+            }
+        }
+        // Otherwise, find the likely script.
+        errorCode = U_ZERO_ERROR;
+        icu::CharString likely;
+        {
+            icu::CharStringByteSink sink(&likely);
+            ulocimp_addLikelySubtags(locale, sink, &errorCode);
+        }
+        if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING) {
+            return FALSE;
+        }
+        scriptLength = uloc_getScript(likely.data(), script, UPRV_LENGTHOF(script), &errorCode);
+        if (U_FAILURE(errorCode) || errorCode == U_STRING_NOT_TERMINATED_WARNING ||
+                scriptLength == 0) {
+            return FALSE;
+        }
+    }
+    UScriptCode scriptCode = (UScriptCode)u_getPropertyValueEnum(UCHAR_SCRIPT, script);
+    return uscript_isRightToLeft(scriptCode);
+}
+
+U_NAMESPACE_BEGIN
+
+UBool
+Locale::isRightToLeft() const {
+    return uloc_isRightToLeft(getBaseName());
+}
+
+U_NAMESPACE_END
+
+// The following must at least allow for rg key value (6) plus terminator (1).
+#define ULOC_RG_BUFLEN 8
+
+U_CAPI int32_t U_EXPORT2
+ulocimp_getRegionForSupplementalData(const char *localeID, UBool inferRegion,
+                                     char *region, int32_t regionCapacity, UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+    char rgBuf[ULOC_RG_BUFLEN];
+    UErrorCode rgStatus = U_ZERO_ERROR;
+
+    // First check for rg keyword value
+    int32_t rgLen = uloc_getKeywordValue(localeID, "rg", rgBuf, ULOC_RG_BUFLEN, &rgStatus);
+    if (U_FAILURE(rgStatus) || rgLen != 6) {
+        rgLen = 0;
+    } else {
+        // rgBuf guaranteed to be zero terminated here, with text len 6
+        char *rgPtr = rgBuf;
+        for (; *rgPtr!= 0; rgPtr++) {
+            *rgPtr = uprv_toupper(*rgPtr);
+        }
+        rgLen = (uprv_strcmp(rgBuf+2, "ZZZZ") == 0)? 2: 0;
+    }
+
+    if (rgLen == 0) {
+        // No valid rg keyword value, try for unicode_region_subtag
+        rgLen = uloc_getCountry(localeID, rgBuf, ULOC_RG_BUFLEN, status);
+        if (U_FAILURE(*status)) {
+            rgLen = 0;
+        } else if (rgLen == 0 && inferRegion) {
+            // no unicode_region_subtag but inferRegion TRUE, try likely subtags
+            rgStatus = U_ZERO_ERROR;
+            icu::CharString locBuf;
+            {
+                icu::CharStringByteSink sink(&locBuf);
+                ulocimp_addLikelySubtags(localeID, sink, &rgStatus);
+            }
+            if (U_SUCCESS(rgStatus)) {
+                rgLen = uloc_getCountry(locBuf.data(), rgBuf, ULOC_RG_BUFLEN, status);
+                if (U_FAILURE(*status)) {
+                    rgLen = 0;
+                }
+            }
+        }
+    }
+
+    rgBuf[rgLen] = 0;
+    uprv_strncpy(region, rgBuf, regionCapacity);
+    return u_terminateChars(region, regionCapacity, rgLen, status);
+}
+
diff --git a/thirdparty/icu4c/common/loclikelysubtags.cpp b/thirdparty/icu4c/common/loclikelysubtags.cpp
new file mode 100644
index 0000000000..a031bfa587
--- /dev/null
+++ b/thirdparty/icu4c/common/loclikelysubtags.cpp
@@ -0,0 +1,682 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// loclikelysubtags.cpp
+// created: 2019may08 Markus W. Scherer
+
+#include <utility>
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/localpointer.h"
+#include "unicode/locid.h"
+#include "unicode/uobject.h"
+#include "unicode/ures.h"
+#include "charstr.h"
+#include "cstring.h"
+#include "loclikelysubtags.h"
+#include "lsr.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "uhash.h"
+#include "uinvchar.h"
+#include "umutex.h"
+#include "uniquecharstr.h"
+#include "uresdata.h"
+#include "uresimp.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+constexpr char PSEUDO_ACCENTS_PREFIX = '\'';  // -XA, -PSACCENT
+constexpr char PSEUDO_BIDI_PREFIX = '+';  // -XB, -PSBIDI
+constexpr char PSEUDO_CRACKED_PREFIX = ',';  // -XC, -PSCRACK
+
+}  // namespace
+
+LocaleDistanceData::LocaleDistanceData(LocaleDistanceData &&data) :
+        distanceTrieBytes(data.distanceTrieBytes),
+        regionToPartitions(data.regionToPartitions),
+        partitions(data.partitions),
+        paradigms(data.paradigms), paradigmsLength(data.paradigmsLength),
+        distances(data.distances) {
+    data.partitions = nullptr;
+    data.paradigms = nullptr;
+}
+
+LocaleDistanceData::~LocaleDistanceData() {
+    uprv_free(partitions);
+    delete[] paradigms;
+}
+
+// TODO(ICU-20777): Rename to just LikelySubtagsData.
+struct XLikelySubtagsData {
+    UResourceBundle *langInfoBundle = nullptr;
+    UniqueCharStrings strings;
+    CharStringMap languageAliases;
+    CharStringMap regionAliases;
+    const uint8_t *trieBytes = nullptr;
+    LSR *lsrs = nullptr;
+    int32_t lsrsLength = 0;
+
+    LocaleDistanceData distanceData;
+
+    XLikelySubtagsData(UErrorCode &errorCode) : strings(errorCode) {}
+
+    ~XLikelySubtagsData() {
+        ures_close(langInfoBundle);
+        delete[] lsrs;
+    }
+
+    void load(UErrorCode &errorCode) {
+        langInfoBundle = ures_openDirect(nullptr, "langInfo", &errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        StackUResourceBundle stackTempBundle;
+        ResourceDataValue value;
+        ures_getValueWithFallback(langInfoBundle, "likely", stackTempBundle.getAlias(),
+                                  value, errorCode);
+        ResourceTable likelyTable = value.getTable(errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+
+        // Read all strings in the resource bundle and convert them to invariant char *.
+        LocalMemory<int32_t> languageIndexes, regionIndexes, lsrSubtagIndexes;
+        int32_t languagesLength = 0, regionsLength = 0, lsrSubtagsLength = 0;
+        if (!readStrings(likelyTable, "languageAliases", value,
+                         languageIndexes, languagesLength, errorCode) ||
+                !readStrings(likelyTable, "regionAliases", value,
+                             regionIndexes, regionsLength, errorCode) ||
+                !readStrings(likelyTable, "lsrs", value,
+                             lsrSubtagIndexes,lsrSubtagsLength, errorCode)) {
+            return;
+        }
+        if ((languagesLength & 1) != 0 ||
+                (regionsLength & 1) != 0 ||
+                (lsrSubtagsLength % 3) != 0) {
+            errorCode = U_INVALID_FORMAT_ERROR;
+            return;
+        }
+        if (lsrSubtagsLength == 0) {
+            errorCode = U_MISSING_RESOURCE_ERROR;
+            return;
+        }
+
+        if (!likelyTable.findValue("trie", value)) {
+            errorCode = U_MISSING_RESOURCE_ERROR;
+            return;
+        }
+        int32_t length;
+        trieBytes = value.getBinary(length, errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+
+        // Also read distance/matcher data if available,
+        // to open & keep only one resource bundle pointer
+        // and to use one single UniqueCharStrings.
+        UErrorCode matchErrorCode = U_ZERO_ERROR;
+        ures_getValueWithFallback(langInfoBundle, "match", stackTempBundle.getAlias(),
+                                  value, matchErrorCode);
+        LocalMemory<int32_t> partitionIndexes, paradigmSubtagIndexes;
+        int32_t partitionsLength = 0, paradigmSubtagsLength = 0;
+        if (U_SUCCESS(matchErrorCode)) {
+            ResourceTable matchTable = value.getTable(errorCode);
+            if (U_FAILURE(errorCode)) { return; }
+
+            if (matchTable.findValue("trie", value)) {
+                distanceData.distanceTrieBytes = value.getBinary(length, errorCode);
+                if (U_FAILURE(errorCode)) { return; }
+            }
+
+            if (matchTable.findValue("regionToPartitions", value)) {
+                distanceData.regionToPartitions = value.getBinary(length, errorCode);
+                if (U_FAILURE(errorCode)) { return; }
+                if (length < LSR::REGION_INDEX_LIMIT) {
+                    errorCode = U_INVALID_FORMAT_ERROR;
+                    return;
+                }
+            }
+
+            if (!readStrings(matchTable, "partitions", value,
+                             partitionIndexes, partitionsLength, errorCode) ||
+                    !readStrings(matchTable, "paradigms", value,
+                                 paradigmSubtagIndexes, paradigmSubtagsLength, errorCode)) {
+                return;
+            }
+            if ((paradigmSubtagsLength % 3) != 0) {
+                errorCode = U_INVALID_FORMAT_ERROR;
+                return;
+            }
+
+            if (matchTable.findValue("distances", value)) {
+                distanceData.distances = value.getIntVector(length, errorCode);
+                if (U_FAILURE(errorCode)) { return; }
+                if (length < 4) {  // LocaleDistance IX_LIMIT
+                    errorCode = U_INVALID_FORMAT_ERROR;
+                    return;
+                }
+            }
+        } else if (matchErrorCode == U_MISSING_RESOURCE_ERROR) {
+            // ok for likely subtags
+        } else {  // error other than missing resource
+            errorCode = matchErrorCode;
+            return;
+        }
+
+        // Fetch & store invariant-character versions of strings
+        // only after we have collected and de-duplicated all of them.
+        strings.freeze();
+
+        languageAliases = CharStringMap(languagesLength / 2, errorCode);
+        for (int32_t i = 0; i < languagesLength; i += 2) {
+            languageAliases.put(strings.get(languageIndexes[i]),
+                                strings.get(languageIndexes[i + 1]), errorCode);
+        }
+
+        regionAliases = CharStringMap(regionsLength / 2, errorCode);
+        for (int32_t i = 0; i < regionsLength; i += 2) {
+            regionAliases.put(strings.get(regionIndexes[i]),
+                              strings.get(regionIndexes[i + 1]), errorCode);
+        }
+        if (U_FAILURE(errorCode)) { return; }
+
+        lsrsLength = lsrSubtagsLength / 3;
+        lsrs = new LSR[lsrsLength];
+        if (lsrs == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        for (int32_t i = 0, j = 0; i < lsrSubtagsLength; i += 3, ++j) {
+            lsrs[j] = LSR(strings.get(lsrSubtagIndexes[i]),
+                          strings.get(lsrSubtagIndexes[i + 1]),
+                          strings.get(lsrSubtagIndexes[i + 2]),
+                          LSR::IMPLICIT_LSR);
+        }
+
+        if (partitionsLength > 0) {
+            distanceData.partitions = static_cast<const char **>(
+                uprv_malloc(partitionsLength * sizeof(const char *)));
+            if (distanceData.partitions == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            for (int32_t i = 0; i < partitionsLength; ++i) {
+                distanceData.partitions[i] = strings.get(partitionIndexes[i]);
+            }
+        }
+
+        if (paradigmSubtagsLength > 0) {
+            distanceData.paradigmsLength = paradigmSubtagsLength / 3;
+            LSR *paradigms = new LSR[distanceData.paradigmsLength];
+            if (paradigms == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            for (int32_t i = 0, j = 0; i < paradigmSubtagsLength; i += 3, ++j) {
+                paradigms[j] = LSR(strings.get(paradigmSubtagIndexes[i]),
+                                   strings.get(paradigmSubtagIndexes[i + 1]),
+                                   strings.get(paradigmSubtagIndexes[i + 2]),
+                                   LSR::DONT_CARE_FLAGS);
+            }
+            distanceData.paradigms = paradigms;
+        }
+    }
+
+private:
+    bool readStrings(const ResourceTable &table, const char *key, ResourceValue &value,
+                     LocalMemory<int32_t> &indexes, int32_t &length, UErrorCode &errorCode) {
+        if (table.findValue(key, value)) {
+            ResourceArray stringArray = value.getArray(errorCode);
+            if (U_FAILURE(errorCode)) { return false; }
+            length = stringArray.getSize();
+            if (length == 0) { return true; }
+            int32_t *rawIndexes = indexes.allocateInsteadAndCopy(length);
+            if (rawIndexes == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+                return false;
+            }
+            for (int i = 0; i < length; ++i) {
+                stringArray.getValue(i, value);  // returns TRUE because i < length
+                rawIndexes[i] = strings.add(value.getUnicodeString(errorCode), errorCode);
+                if (U_FAILURE(errorCode)) { return false; }
+            }
+        }
+        return true;
+    }
+};
+
+namespace {
+
+XLikelySubtags *gLikelySubtags = nullptr;
+UInitOnce gInitOnce = U_INITONCE_INITIALIZER;
+
+UBool U_CALLCONV cleanup() {
+    delete gLikelySubtags;
+    gLikelySubtags = nullptr;
+    gInitOnce.reset();
+    return TRUE;
+}
+
+}  // namespace
+
+void U_CALLCONV XLikelySubtags::initLikelySubtags(UErrorCode &errorCode) {
+    // This function is invoked only via umtx_initOnce().
+    U_ASSERT(gLikelySubtags == nullptr);
+    XLikelySubtagsData data(errorCode);
+    data.load(errorCode);
+    if (U_FAILURE(errorCode)) { return; }
+    gLikelySubtags = new XLikelySubtags(data);
+    if (gLikelySubtags == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_LIKELY_SUBTAGS, cleanup);
+}
+
+const XLikelySubtags *XLikelySubtags::getSingleton(UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    umtx_initOnce(gInitOnce, &XLikelySubtags::initLikelySubtags, errorCode);
+    return gLikelySubtags;
+}
+
+XLikelySubtags::XLikelySubtags(XLikelySubtagsData &data) :
+        langInfoBundle(data.langInfoBundle),
+        strings(data.strings.orphanCharStrings()),
+        languageAliases(std::move(data.languageAliases)),
+        regionAliases(std::move(data.regionAliases)),
+        trie(data.trieBytes),
+        lsrs(data.lsrs),
+#if U_DEBUG
+        lsrsLength(data.lsrsLength),
+#endif
+        distanceData(std::move(data.distanceData)) {
+    data.langInfoBundle = nullptr;
+    data.lsrs = nullptr;
+
+    // Cache the result of looking up language="und" encoded as "*", and "und-Zzzz" ("**").
+    UStringTrieResult result = trie.next(u'*');
+    U_ASSERT(USTRINGTRIE_HAS_NEXT(result));
+    trieUndState = trie.getState64();
+    result = trie.next(u'*');
+    U_ASSERT(USTRINGTRIE_HAS_NEXT(result));
+    trieUndZzzzState = trie.getState64();
+    result = trie.next(u'*');
+    U_ASSERT(USTRINGTRIE_HAS_VALUE(result));
+    defaultLsrIndex = trie.getValue();
+    trie.reset();
+
+    for (char16_t c = u'a'; c <= u'z'; ++c) {
+        result = trie.next(c);
+        if (result == USTRINGTRIE_NO_VALUE) {
+            trieFirstLetterStates[c - u'a'] = trie.getState64();
+        }
+        trie.reset();
+    }
+}
+
+XLikelySubtags::~XLikelySubtags() {
+    ures_close(langInfoBundle);
+    delete strings;
+    delete[] lsrs;
+}
+
+LSR XLikelySubtags::makeMaximizedLsrFrom(const Locale &locale, UErrorCode &errorCode) const {
+    const char *name = locale.getName();
+    if (uprv_isAtSign(name[0]) && name[1] == 'x' && name[2] == '=') {  // name.startsWith("@x=")
+        // Private use language tag x-subtag-subtag...
+        return LSR(name, "", "", LSR::EXPLICIT_LSR);
+    }
+    return makeMaximizedLsr(locale.getLanguage(), locale.getScript(), locale.getCountry(),
+                            locale.getVariant(), errorCode);
+}
+
+namespace {
+
+const char *getCanonical(const CharStringMap &aliases, const char *alias) {
+    const char *canonical = aliases.get(alias);
+    return canonical == nullptr ? alias : canonical;
+}
+
+}  // namespace
+
+LSR XLikelySubtags::makeMaximizedLsr(const char *language, const char *script, const char *region,
+                                     const char *variant, UErrorCode &errorCode) const {
+    // Handle pseudolocales like en-XA, ar-XB, fr-PSCRACK.
+    // They should match only themselves,
+    // not other locales with what looks like the same language and script subtags.
+    char c1;
+    if (region[0] == 'X' && (c1 = region[1]) != 0 && region[2] == 0) {
+        switch (c1) {
+        case 'A':
+            return LSR(PSEUDO_ACCENTS_PREFIX, language, script, region,
+                       LSR::EXPLICIT_LSR, errorCode);
+        case 'B':
+            return LSR(PSEUDO_BIDI_PREFIX, language, script, region,
+                       LSR::EXPLICIT_LSR, errorCode);
+        case 'C':
+            return LSR(PSEUDO_CRACKED_PREFIX, language, script, region,
+                       LSR::EXPLICIT_LSR, errorCode);
+        default:  // normal locale
+            break;
+        }
+    }
+
+    if (variant[0] == 'P' && variant[1] == 'S') {
+        int32_t lsrFlags = *region == 0 ?
+            LSR::EXPLICIT_LANGUAGE | LSR::EXPLICIT_SCRIPT : LSR::EXPLICIT_LSR;
+        if (uprv_strcmp(variant, "PSACCENT") == 0) {
+            return LSR(PSEUDO_ACCENTS_PREFIX, language, script,
+                       *region == 0 ? "XA" : region, lsrFlags, errorCode);
+        } else if (uprv_strcmp(variant, "PSBIDI") == 0) {
+            return LSR(PSEUDO_BIDI_PREFIX, language, script,
+                       *region == 0 ? "XB" : region, lsrFlags, errorCode);
+        } else if (uprv_strcmp(variant, "PSCRACK") == 0) {
+            return LSR(PSEUDO_CRACKED_PREFIX, language, script,
+                       *region == 0 ? "XC" : region, lsrFlags, errorCode);
+        }
+        // else normal locale
+    }
+
+    language = getCanonical(languageAliases, language);
+    // (We have no script mappings.)
+    region = getCanonical(regionAliases, region);
+    return maximize(language, script, region);
+}
+
+LSR XLikelySubtags::maximize(const char *language, const char *script, const char *region) const {
+    if (uprv_strcmp(language, "und") == 0) {
+        language = "";
+    }
+    if (uprv_strcmp(script, "Zzzz") == 0) {
+        script = "";
+    }
+    if (uprv_strcmp(region, "ZZ") == 0) {
+        region = "";
+    }
+    if (*script != 0 && *region != 0 && *language != 0) {
+        return LSR(language, script, region, LSR::EXPLICIT_LSR);  // already maximized
+    }
+
+    uint32_t retainOldMask = 0;
+    BytesTrie iter(trie);
+    uint64_t state;
+    int32_t value;
+    // Small optimization: Array lookup for first language letter.
+    int32_t c0;
+    if (0 <= (c0 = uprv_lowerOrdinal(language[0])) && c0 <= 25 &&
+            language[1] != 0 &&  // language.length() >= 2
+            (state = trieFirstLetterStates[c0]) != 0) {
+        value = trieNext(iter.resetToState64(state), language, 1);
+    } else {
+        value = trieNext(iter, language, 0);
+    }
+    if (value >= 0) {
+        if (*language != 0) {
+            retainOldMask |= 4;
+        }
+        state = iter.getState64();
+    } else {
+        retainOldMask |= 4;
+        iter.resetToState64(trieUndState);  // "und" ("*")
+        state = 0;
+    }
+
+    if (value > 0) {
+        // Intermediate or final value from just language.
+        if (value == SKIP_SCRIPT) {
+            value = 0;
+        }
+        if (*script != 0) {
+            retainOldMask |= 2;
+        }
+    } else {
+        value = trieNext(iter, script, 0);
+        if (value >= 0) {
+            if (*script != 0) {
+                retainOldMask |= 2;
+            }
+            state = iter.getState64();
+        } else {
+            retainOldMask |= 2;
+            if (state == 0) {
+                iter.resetToState64(trieUndZzzzState);  // "und-Zzzz" ("**")
+            } else {
+                iter.resetToState64(state);
+                value = trieNext(iter, "", 0);
+                U_ASSERT(value >= 0);
+                state = iter.getState64();
+            }
+        }
+    }
+
+    if (value > 0) {
+        // Final value from just language or language+script.
+        if (*region != 0) {
+            retainOldMask |= 1;
+        }
+    } else {
+        value = trieNext(iter, region, 0);
+        if (value >= 0) {
+            if (*region != 0) {
+                retainOldMask |= 1;
+            }
+        } else {
+            retainOldMask |= 1;
+            if (state == 0) {
+                value = defaultLsrIndex;
+            } else {
+                iter.resetToState64(state);
+                value = trieNext(iter, "", 0);
+                U_ASSERT(value > 0);
+            }
+        }
+    }
+    U_ASSERT(value < lsrsLength);
+    const LSR &result = lsrs[value];
+
+    if (*language == 0) {
+        language = "und";
+    }
+
+    if (retainOldMask == 0) {
+        // Quickly return a copy of the lookup-result LSR
+        // without new allocation of the subtags.
+        return LSR(result.language, result.script, result.region, result.flags);
+    }
+    if ((retainOldMask & 4) == 0) {
+        language = result.language;
+    }
+    if ((retainOldMask & 2) == 0) {
+        script = result.script;
+    }
+    if ((retainOldMask & 1) == 0) {
+        region = result.region;
+    }
+    // retainOldMask flags = LSR explicit-subtag flags
+    return LSR(language, script, region, retainOldMask);
+}
+
+int32_t XLikelySubtags::compareLikely(const LSR &lsr, const LSR &other, int32_t likelyInfo) const {
+    // If likelyInfo >= 0:
+    // likelyInfo bit 1 is set if the previous comparison with lsr
+    // was for equal language and script.
+    // Otherwise the scripts differed.
+    if (uprv_strcmp(lsr.language, other.language) != 0) {
+        return 0xfffffffc;  // negative, lsr not better than other
+    }
+    if (uprv_strcmp(lsr.script, other.script) != 0) {
+        int32_t index;
+        if (likelyInfo >= 0 && (likelyInfo & 2) == 0) {
+            index = likelyInfo >> 2;
+        } else {
+            index = getLikelyIndex(lsr.language, "");
+            likelyInfo = index << 2;
+        }
+        const LSR &likely = lsrs[index];
+        if (uprv_strcmp(lsr.script, likely.script) == 0) {
+            return likelyInfo | 1;
+        } else {
+            return likelyInfo & ~1;
+        }
+    }
+    if (uprv_strcmp(lsr.region, other.region) != 0) {
+        int32_t index;
+        if (likelyInfo >= 0 && (likelyInfo & 2) != 0) {
+            index = likelyInfo >> 2;
+        } else {
+            index = getLikelyIndex(lsr.language, lsr.region);
+            likelyInfo = (index << 2) | 2;
+        }
+        const LSR &likely = lsrs[index];
+        if (uprv_strcmp(lsr.region, likely.region) == 0) {
+            return likelyInfo | 1;
+        } else {
+            return likelyInfo & ~1;
+        }
+    }
+    return likelyInfo & ~1;  // lsr not better than other
+}
+
+// Subset of maximize().
+int32_t XLikelySubtags::getLikelyIndex(const char *language, const char *script) const {
+    if (uprv_strcmp(language, "und") == 0) {
+        language = "";
+    }
+    if (uprv_strcmp(script, "Zzzz") == 0) {
+        script = "";
+    }
+
+    BytesTrie iter(trie);
+    uint64_t state;
+    int32_t value;
+    // Small optimization: Array lookup for first language letter.
+    int32_t c0;
+    if (0 <= (c0 = uprv_lowerOrdinal(language[0])) && c0 <= 25 &&
+            language[1] != 0 &&  // language.length() >= 2
+            (state = trieFirstLetterStates[c0]) != 0) {
+        value = trieNext(iter.resetToState64(state), language, 1);
+    } else {
+        value = trieNext(iter, language, 0);
+    }
+    if (value >= 0) {
+        state = iter.getState64();
+    } else {
+        iter.resetToState64(trieUndState);  // "und" ("*")
+        state = 0;
+    }
+
+    if (value > 0) {
+        // Intermediate or final value from just language.
+        if (value == SKIP_SCRIPT) {
+            value = 0;
+        }
+    } else {
+        value = trieNext(iter, script, 0);
+        if (value >= 0) {
+            state = iter.getState64();
+        } else {
+            if (state == 0) {
+                iter.resetToState64(trieUndZzzzState);  // "und-Zzzz" ("**")
+            } else {
+                iter.resetToState64(state);
+                value = trieNext(iter, "", 0);
+                U_ASSERT(value >= 0);
+                state = iter.getState64();
+            }
+        }
+    }
+
+    if (value > 0) {
+        // Final value from just language or language+script.
+    } else {
+        value = trieNext(iter, "", 0);
+        U_ASSERT(value > 0);
+    }
+    U_ASSERT(value < lsrsLength);
+    return value;
+}
+
+int32_t XLikelySubtags::trieNext(BytesTrie &iter, const char *s, int32_t i) {
+    UStringTrieResult result;
+    uint8_t c;
+    if ((c = s[i]) == 0) {
+        result = iter.next(u'*');
+    } else {
+        for (;;) {
+            c = uprv_invCharToAscii(c);
+            // EBCDIC: If s[i] is not an invariant character,
+            // then c is now 0 and will simply not match anything, which is harmless.
+            uint8_t next = s[++i];
+            if (next != 0) {
+                if (!USTRINGTRIE_HAS_NEXT(iter.next(c))) {
+                    return -1;
+                }
+            } else {
+                // last character of this subtag
+                result = iter.next(c | 0x80);
+                break;
+            }
+            c = next;
+        }
+    }
+    switch (result) {
+    case USTRINGTRIE_NO_MATCH: return -1;
+    case USTRINGTRIE_NO_VALUE: return 0;
+    case USTRINGTRIE_INTERMEDIATE_VALUE:
+        U_ASSERT(iter.getValue() == SKIP_SCRIPT);
+        return SKIP_SCRIPT;
+    case USTRINGTRIE_FINAL_VALUE: return iter.getValue();
+    default: return -1;
+    }
+}
+
+// TODO(ICU-20777): Switch Locale/uloc_ likely-subtags API from the old code
+// in loclikely.cpp to this new code, including activating this
+// minimizeSubtags() function. The LocaleMatcher does not minimize.
+#if 0
+LSR XLikelySubtags::minimizeSubtags(const char *languageIn, const char *scriptIn,
+                                    const char *regionIn, ULocale.Minimize fieldToFavor,
+                                    UErrorCode &errorCode) const {
+    LSR result = maximize(languageIn, scriptIn, regionIn);
+
+    // We could try just a series of checks, like:
+    // LSR result2 = addLikelySubtags(languageIn, "", "");
+    // if result.equals(result2) return result2;
+    // However, we can optimize 2 of the cases:
+    //   (languageIn, "", "")
+    //   (languageIn, "", regionIn)
+
+    // value00 = lookup(result.language, "", "")
+    BytesTrie iter = new BytesTrie(trie);
+    int value = trieNext(iter, result.language, 0);
+    U_ASSERT(value >= 0);
+    if (value == 0) {
+        value = trieNext(iter, "", 0);
+        U_ASSERT(value >= 0);
+        if (value == 0) {
+            value = trieNext(iter, "", 0);
+        }
+    }
+    U_ASSERT(value > 0);
+    LSR value00 = lsrs[value];
+    boolean favorRegionOk = false;
+    if (result.script.equals(value00.script)) { //script is default
+        if (result.region.equals(value00.region)) {
+            return new LSR(result.language, "", "", LSR.DONT_CARE_FLAGS);
+        } else if (fieldToFavor == ULocale.Minimize.FAVOR_REGION) {
+            return new LSR(result.language, "", result.region, LSR.DONT_CARE_FLAGS);
+        } else {
+            favorRegionOk = true;
+        }
+    }
+
+    // The last case is not as easy to optimize.
+    // Maybe do later, but for now use the straightforward code.
+    LSR result2 = maximize(languageIn, scriptIn, "");
+    if (result2.equals(result)) {
+        return new LSR(result.language, result.script, "", LSR.DONT_CARE_FLAGS);
+    } else if (favorRegionOk) {
+        return new LSR(result.language, "", result.region, LSR.DONT_CARE_FLAGS);
+    }
+    return result;
+}
+#endif
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/loclikelysubtags.h b/thirdparty/icu4c/common/loclikelysubtags.h
new file mode 100644
index 0000000000..14a01a5eac
--- /dev/null
+++ b/thirdparty/icu4c/common/loclikelysubtags.h
@@ -0,0 +1,121 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// loclikelysubtags.h
+// created: 2019may08 Markus W. Scherer
+
+#ifndef __LOCLIKELYSUBTAGS_H__
+#define __LOCLIKELYSUBTAGS_H__
+
+#include <utility>
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/locid.h"
+#include "unicode/uobject.h"
+#include "unicode/ures.h"
+#include "charstrmap.h"
+#include "lsr.h"
+
+U_NAMESPACE_BEGIN
+
+struct XLikelySubtagsData;
+
+struct LocaleDistanceData {
+    LocaleDistanceData() = default;
+    LocaleDistanceData(LocaleDistanceData &&data);
+    ~LocaleDistanceData();
+
+    const uint8_t *distanceTrieBytes = nullptr;
+    const uint8_t *regionToPartitions = nullptr;
+    const char **partitions = nullptr;
+    const LSR *paradigms = nullptr;
+    int32_t paradigmsLength = 0;
+    const int32_t *distances = nullptr;
+
+private:
+    LocaleDistanceData &operator=(const LocaleDistanceData &) = delete;
+};
+
+// TODO(ICU-20777): Rename to just LikelySubtags.
+class XLikelySubtags final : public UMemory {
+public:
+    ~XLikelySubtags();
+
+    static constexpr int32_t SKIP_SCRIPT = 1;
+
+    // VisibleForTesting
+    static const XLikelySubtags *getSingleton(UErrorCode &errorCode);
+
+    // VisibleForTesting
+    LSR makeMaximizedLsrFrom(const Locale &locale, UErrorCode &errorCode) const;
+
+    /**
+     * Tests whether lsr is "more likely" than other.
+     * For example, fr-Latn-FR is more likely than fr-Latn-CH because
+     * FR is the default region for fr-Latn.
+     *
+     * The likelyInfo caches lookup information between calls.
+     * The return value is an updated likelyInfo value,
+     * with bit 0 set if lsr is "more likely".
+     * The initial value of likelyInfo must be negative.
+     */
+    int32_t compareLikely(const LSR &lsr, const LSR &other, int32_t likelyInfo) const;
+
+    // TODO(ICU-20777): Switch Locale/uloc_ likely-subtags API from the old code
+    // in loclikely.cpp to this new code, including activating this
+    // minimizeSubtags() function. The LocaleMatcher does not minimize.
+#if 0
+    LSR minimizeSubtags(const char *languageIn, const char *scriptIn, const char *regionIn,
+                        ULocale.Minimize fieldToFavor, UErrorCode &errorCode) const;
+#endif
+
+    // visible for LocaleDistance
+    const LocaleDistanceData &getDistanceData() const { return distanceData; }
+
+private:
+    XLikelySubtags(XLikelySubtagsData &data);
+    XLikelySubtags(const XLikelySubtags &other) = delete;
+    XLikelySubtags &operator=(const XLikelySubtags &other) = delete;
+
+    static void initLikelySubtags(UErrorCode &errorCode);
+
+    LSR makeMaximizedLsr(const char *language, const char *script, const char *region,
+                         const char *variant, UErrorCode &errorCode) const;
+
+    /**
+     * Raw access to addLikelySubtags. Input must be in canonical format, eg "en", not "eng" or "EN".
+     */
+    LSR maximize(const char *language, const char *script, const char *region) const;
+
+    int32_t getLikelyIndex(const char *language, const char *script) const;
+
+    static int32_t trieNext(BytesTrie &iter, const char *s, int32_t i);
+
+    UResourceBundle *langInfoBundle;
+    // We could store the strings by value, except that if there were few enough strings,
+    // moving the contents could copy it to a different array,
+    // invalidating the pointers stored in the maps.
+    CharString *strings;
+    CharStringMap languageAliases;
+    CharStringMap regionAliases;
+
+    // The trie maps each lang+script+region (encoded in ASCII) to an index into lsrs.
+    // There is also a trie value for each intermediate lang and lang+script.
+    // '*' is used instead of "und", "Zzzz"/"" and "ZZ"/"".
+    BytesTrie trie;
+    uint64_t trieUndState;
+    uint64_t trieUndZzzzState;
+    int32_t defaultLsrIndex;
+    uint64_t trieFirstLetterStates[26];
+    const LSR *lsrs;
+#if U_DEBUG
+    int32_t lsrsLength;
+#endif
+
+    // distance/matcher data: see comment in XLikelySubtagsData::load()
+    LocaleDistanceData distanceData;
+};
+
+U_NAMESPACE_END
+
+#endif  // __LOCLIKELYSUBTAGS_H__
diff --git a/thirdparty/icu4c/common/locmap.cpp b/thirdparty/icu4c/common/locmap.cpp
new file mode 100644
index 0000000000..29a5646385
--- /dev/null
+++ b/thirdparty/icu4c/common/locmap.cpp
@@ -0,0 +1,1315 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ **********************************************************************
+ *   Copyright (C) 1996-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ **********************************************************************
+ *
+ * Provides functionality for mapping between
+ * LCID and Posix IDs or ICU locale to codepage
+ *
+ * Note: All classes and code in this file are
+ *       intended for internal use only.
+ *
+ * Methods of interest:
+ *   unsigned long convertToLCID(const char*);
+ *   const char* convertToPosix(unsigned long);
+ *
+ * Kathleen Wilson, 4/30/96
+ *
+ *  Date        Name        Description
+ *  3/11/97     aliu        Fixed off-by-one bug in assignment operator. Added
+ *                          setId() method and safety check against 
+ *                          MAX_ID_LENGTH.
+ * 04/23/99     stephen     Added C wrapper for convertToPosix.
+ * 09/18/00     george      Removed the memory leaks.
+ * 08/23/01     george      Convert to C
+ */
+
+#include "locmap.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "ulocimp.h"
+#include "unicode/uloc.h"
+
+#if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+#include <windows.h>
+#include <winnls.h> // LCIDToLocaleName and LocaleNameToLCID
+#endif
+
+/*
+ * Note:
+ * The mapping from Win32 locale ID numbers to POSIX locale strings should
+ * be the faster one.
+ *
+ * Windows LCIDs are defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx
+ * [MS-LCID] Windows Language Code Identifier (LCID) Reference
+ */
+
+/*
+////////////////////////////////////////////////
+//
+// Internal Classes for LCID <--> POSIX Mapping
+//
+/////////////////////////////////////////////////
+*/
+
+typedef struct ILcidPosixElement
+{
+    const uint32_t hostID;
+    const char * const posixID;
+} ILcidPosixElement;
+
+typedef struct ILcidPosixMap
+{
+    const uint32_t numRegions;
+    const struct ILcidPosixElement* const regionMaps;
+} ILcidPosixMap;
+
+
+/*
+/////////////////////////////////////////////////
+//
+// Easy macros to make the LCID <--> POSIX Mapping
+//
+/////////////////////////////////////////////////
+*/
+
+/**
+ * The standard one language/one country mapping for LCID.
+ * The first element must be the language, and the following
+ * elements are the language with the country.
+ * @param hostID LCID in host format such as 0x044d
+ * @param languageID posix ID of just the language such as 'de'
+ * @param posixID posix ID of the language_TERRITORY such as 'de_CH'
+ */
+#define ILCID_POSIX_ELEMENT_ARRAY(hostID, languageID, posixID) \
+static const ILcidPosixElement locmap_ ## languageID [] = { \
+    {LANGUAGE_LCID(hostID), #languageID},     /* parent locale */ \
+    {hostID, #posixID}, \
+};
+
+/**
+ * Define a subtable by ID
+ * @param id the POSIX ID, either a language or language_TERRITORY
+ */
+#define ILCID_POSIX_SUBTABLE(id) \
+static const ILcidPosixElement locmap_ ## id [] =
+
+
+/**
+ * Create the map for the posixID. This macro supposes that the language string
+ * name is the same as the global variable name, and that the first element
+ * in the ILcidPosixElement is just the language.
+ * @param _posixID the full POSIX ID for this entry.
+ */
+#define ILCID_POSIX_MAP(_posixID) \
+    {UPRV_LENGTHOF(locmap_ ## _posixID), locmap_ ## _posixID}
+
+/*
+////////////////////////////////////////////
+//
+// Create the table of LCID to POSIX Mapping
+// None of it should be dynamically created.
+//
+// Keep static locale variables inside the function so that
+// it can be created properly during static init.
+//
+// Note: This table should be updated periodically. Check the [MS-LCID] Windows Language Code Identifier 
+//       (LCID) Reference defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx
+//
+//       Microsoft is moving away from LCID in favor of locale name as of Vista.  This table needs to be
+//       maintained for support of older Windows version.
+//       Update: Windows 7 (091130)
+//
+// Note: Microsoft assign a different LCID if a locale has a sorting variant. POSIX IDs below may contain
+//       @collation=XXX, but no other keywords are allowed (at least for now). When uprv_convertToLCID() is
+//       called from uloc_getLCID(), keywords other than collation are already removed. If we really need
+//       to support other keywords in this mapping data, we must update the implementation.
+////////////////////////////////////////////
+*/
+
+// TODO: For Windows ideally this table would be a list of exceptions rather than a complete list as 
+// LocaleNameToLCID and LCIDToLocaleName provide 90% of these.
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0436, af, af_ZA)
+
+ILCID_POSIX_SUBTABLE(ar) {
+    {0x01,   "ar"},
+    {0x3801, "ar_AE"},
+    {0x3c01, "ar_BH"},
+    {0x1401, "ar_DZ"},
+    {0x0c01, "ar_EG"},
+    {0x0801, "ar_IQ"},
+    {0x2c01, "ar_JO"},
+    {0x3401, "ar_KW"},
+    {0x3001, "ar_LB"},
+    {0x1001, "ar_LY"},
+    {0x1801, "ar_MA"},
+    {0x1801, "ar_MO"},
+    {0x2001, "ar_OM"},
+    {0x4001, "ar_QA"},
+    {0x0401, "ar_SA"},
+    {0x2801, "ar_SY"},
+    {0x1c01, "ar_TN"},
+    {0x2401, "ar_YE"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x044d, as, as_IN)
+ILCID_POSIX_ELEMENT_ARRAY(0x045e, am, am_ET)
+ILCID_POSIX_ELEMENT_ARRAY(0x047a, arn,arn_CL)
+
+ILCID_POSIX_SUBTABLE(az) {
+    {0x2c,   "az"},
+    {0x082c, "az_Cyrl_AZ"},  /* Cyrillic based */
+    {0x742c, "az_Cyrl"},  /* Cyrillic based */
+    {0x042c, "az_Latn_AZ"}, /* Latin based */
+    {0x782c, "az_Latn"}, /* Latin based */
+    {0x042c, "az_AZ"} /* Latin based */
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x046d, ba, ba_RU)
+ILCID_POSIX_ELEMENT_ARRAY(0x0423, be, be_BY)
+
+/*ILCID_POSIX_SUBTABLE(ber) {
+    {0x5f,   "ber"},
+    {0x045f, "ber_Arab_DZ"},
+    {0x045f, "ber_Arab"},
+    {0x085f, "ber_Latn_DZ"},
+    {0x085f, "ber_Latn"}
+};*/
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0402, bg, bg_BG)
+
+ILCID_POSIX_SUBTABLE(bin) {
+    {0x66, "bin"},
+    {0x0466, "bin_NG"}
+};
+
+ILCID_POSIX_SUBTABLE(bn) {
+    {0x45,   "bn"},
+    {0x0845, "bn_BD"},
+    {0x0445, "bn_IN"}
+};
+
+ILCID_POSIX_SUBTABLE(bo) {
+    {0x51,   "bo"},
+    {0x0851, "bo_BT"},
+    {0x0451, "bo_CN"},
+    {0x0c51, "dz_BT"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x047e, br, br_FR)
+
+ILCID_POSIX_SUBTABLE(ca) {
+    {0x03,   "ca"},
+    {0x0403, "ca_ES"},
+    {0x0803, "ca_ES_VALENCIA"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0483, co, co_FR)
+
+ILCID_POSIX_SUBTABLE(chr) {
+    {0x05c,  "chr"},
+    {0x7c5c, "chr_Cher"},
+    {0x045c, "chr_Cher_US"},
+    {0x045c, "chr_US"}
+};
+
+// ICU has chosen different names for these.
+ILCID_POSIX_SUBTABLE(ckb) {
+    {0x92,   "ckb"},
+    {0x7c92, "ckb_Arab"},
+    {0x0492, "ckb_Arab_IQ"}
+};
+
+/* Declared as cs_CZ to get around compiler errors on z/OS, which defines cs as a function */
+ILCID_POSIX_ELEMENT_ARRAY(0x0405, cs, cs_CZ)
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0452, cy, cy_GB)
+ILCID_POSIX_ELEMENT_ARRAY(0x0406, da, da_DK)
+
+// Windows doesn't know POSIX or BCP47 Unicode phonebook sort names
+ILCID_POSIX_SUBTABLE(de) {
+    {0x07,   "de"},
+    {0x0c07, "de_AT"},
+    {0x0807, "de_CH"},
+    {0x0407, "de_DE"},
+    {0x1407, "de_LI"},
+    {0x1007, "de_LU"},
+    {0x10407,"de_DE@collation=phonebook"},  /*This is really de_DE_PHONEBOOK on Windows*/
+    {0x10407,"de@collation=phonebook"}  /*This is really de_DE_PHONEBOOK on Windows*/
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0465, dv, dv_MV)
+ILCID_POSIX_ELEMENT_ARRAY(0x0408, el, el_GR)
+
+// Windows uses an empty string for 'invariant'
+ILCID_POSIX_SUBTABLE(en) {
+    {0x09,   "en"},
+    {0x0c09, "en_AU"},
+    {0x2809, "en_BZ"},
+    {0x1009, "en_CA"},
+    {0x0809, "en_GB"},
+    {0x3c09, "en_HK"},
+    {0x3809, "en_ID"},
+    {0x1809, "en_IE"},
+    {0x4009, "en_IN"},
+    {0x2009, "en_JM"},
+    {0x4409, "en_MY"},
+    {0x1409, "en_NZ"},
+    {0x3409, "en_PH"},
+    {0x4809, "en_SG"},
+    {0x2C09, "en_TT"},
+    {0x0409, "en_US"},
+    {0x007f, "en_US_POSIX"}, /* duplicate for round-tripping */
+    {0x2409, "en_029"},
+    {0x1c09, "en_ZA"},
+    {0x3009, "en_ZW"},
+    {0x2409, "en_VI"},  /* Virgin Islands AKA Caribbean Islands (en_CB). On Windows8+ This is 0x1000 or dynamically assigned */
+    {0x0409, "en_AS"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
+    {0x0409, "en_GU"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
+    {0x0409, "en_MH"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
+    {0x0409, "en_MP"},  /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
+    {0x0409, "en_UM"}   /* Alias for en_US. Leave last.  On Windows8+ This is 0x1000 or dynamically assigned */
+};
+
+ILCID_POSIX_SUBTABLE(en_US_POSIX) {
+    {0x007f, "en_US_POSIX"} /* duplicate for roundtripping */
+};
+
+// Windows doesn't know POSIX or BCP47 Unicode traditional sort names
+ILCID_POSIX_SUBTABLE(es) {
+    {0x0a,   "es"},
+    {0x2c0a, "es_AR"},
+    {0x400a, "es_BO"},
+    {0x340a, "es_CL"},
+    {0x240a, "es_CO"},
+    {0x140a, "es_CR"},
+    {0x5c0a, "es_CU"},
+    {0x1c0a, "es_DO"},
+    {0x300a, "es_EC"},
+    {0x0c0a, "es_ES"},      /*Modern sort.*/
+    {0x100a, "es_GT"},
+    {0x480a, "es_HN"},
+    {0x080a, "es_MX"},
+    {0x4c0a, "es_NI"},
+    {0x180a, "es_PA"},
+    {0x280a, "es_PE"},
+    {0x500a, "es_PR"},
+    {0x3c0a, "es_PY"},
+    {0x440a, "es_SV"},
+    {0x540a, "es_US"},
+    {0x380a, "es_UY"},
+    {0x200a, "es_VE"},
+    {0x580a, "es_419"},
+    {0x040a, "es_ES@collation=traditional"},
+    {0x040a, "es@collation=traditional"}        // Windows will treat this as es-ES@collation=traditional
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0425, et, et_EE)
+ILCID_POSIX_ELEMENT_ARRAY(0x042d, eu, eu_ES)
+
+/* ISO-639 doesn't distinguish between Persian and Dari.*/
+ILCID_POSIX_SUBTABLE(fa) {
+    {0x29,   "fa"},
+    {0x0429, "fa_IR"},  /* Persian/Farsi (Iran) */
+    {0x048c, "fa_AF"}   /* Persian/Dari (Afghanistan) */
+};
+
+
+/* duplicate for roundtripping */
+ILCID_POSIX_SUBTABLE(fa_AF) {
+    {0x8c,   "fa_AF"},  /* Persian/Dari (Afghanistan) */
+    {0x048c, "fa_AF"}   /* Persian/Dari (Afghanistan) */
+};
+
+ILCID_POSIX_SUBTABLE(ff) {
+    {0x67,   "ff"},
+    {0x7c67, "ff_Latn"},
+    {0x0867, "ff_Latn_SN"},
+    {0x0467, "ff_NG"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x040b, fi, fi_FI)
+ILCID_POSIX_ELEMENT_ARRAY(0x0464, fil,fil_PH)
+ILCID_POSIX_ELEMENT_ARRAY(0x0438, fo, fo_FO)
+
+ILCID_POSIX_SUBTABLE(fr) {
+    {0x0c,   "fr"},
+    {0x080c, "fr_BE"},
+    {0x0c0c, "fr_CA"},
+    {0x240c, "fr_CD"},
+    {0x240c, "fr_CG"},
+    {0x100c, "fr_CH"},
+    {0x300c, "fr_CI"},
+    {0x2c0c, "fr_CM"},
+    {0x040c, "fr_FR"},
+    {0x3c0c, "fr_HT"},
+    {0x140c, "fr_LU"},
+    {0x380c, "fr_MA"},
+    {0x180c, "fr_MC"},
+    {0x340c, "fr_ML"},
+    {0x200c, "fr_RE"},
+    {0x280c, "fr_SN"},
+    {0xe40c, "fr_015"},
+    {0x1c0c, "fr_029"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0467, fuv, fuv_NG)
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0462, fy, fy_NL)
+
+ILCID_POSIX_SUBTABLE(ga) { /* Gaelic (Ireland) */
+    {0x3c,   "ga"},
+    {0x083c, "ga_IE"},
+    {0x043c, "gd_GB"}
+};
+
+ILCID_POSIX_SUBTABLE(gd) { /* Gaelic (Scotland) */
+    {0x91,   "gd"},
+    {0x0491, "gd_GB"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0456, gl, gl_ES)
+ILCID_POSIX_ELEMENT_ARRAY(0x0447, gu, gu_IN)
+ILCID_POSIX_ELEMENT_ARRAY(0x0474, gn, gn_PY)
+ILCID_POSIX_ELEMENT_ARRAY(0x0484, gsw,gsw_FR)
+
+ILCID_POSIX_SUBTABLE(ha) {
+    {0x68,   "ha"},
+    {0x7c68, "ha_Latn"},
+    {0x0468, "ha_Latn_NG"},
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0475, haw,haw_US)
+ILCID_POSIX_ELEMENT_ARRAY(0x040d, he, he_IL)
+ILCID_POSIX_ELEMENT_ARRAY(0x0439, hi, hi_IN)
+
+/* This LCID is really four different locales.*/
+ILCID_POSIX_SUBTABLE(hr) {
+    {0x1a,   "hr"},
+    {0x141a, "bs_Latn_BA"},  /* Bosnian, Bosnia and Herzegovina */
+    {0x681a, "bs_Latn"},  /* Bosnian, Bosnia and Herzegovina */
+    {0x141a, "bs_BA"},  /* Bosnian, Bosnia and Herzegovina */
+    {0x781a, "bs"},     /* Bosnian */
+    {0x201a, "bs_Cyrl_BA"},  /* Bosnian, Bosnia and Herzegovina */
+    {0x641a, "bs_Cyrl"},  /* Bosnian, Bosnia and Herzegovina */
+    {0x101a, "hr_BA"},  /* Croatian in Bosnia */
+    {0x041a, "hr_HR"},  /* Croatian*/
+    {0x2c1a, "sr_Latn_ME"},
+    {0x241a, "sr_Latn_RS"},
+    {0x181a, "sr_Latn_BA"}, /* Serbo-Croatian in Bosnia */
+    {0x081a, "sr_Latn_CS"}, /* Serbo-Croatian*/
+    {0x701a, "sr_Latn"},    /* It's 0x1a or 0x081a, pick one to make the test program happy. */
+    {0x1c1a, "sr_Cyrl_BA"}, /* Serbo-Croatian in Bosnia */
+    {0x0c1a, "sr_Cyrl_CS"}, /* Serbian*/
+    {0x301a, "sr_Cyrl_ME"},
+    {0x281a, "sr_Cyrl_RS"},
+    {0x6c1a, "sr_Cyrl"},    /* It's 0x1a or 0x0c1a, pick one to make the test program happy. */
+    {0x7c1a, "sr"}          /* In CLDR sr is sr_Cyrl. */
+};
+
+ILCID_POSIX_SUBTABLE(hsb) {
+    {0x2E,   "hsb"},
+    {0x042E, "hsb_DE"},
+    {0x082E, "dsb_DE"},
+    {0x7C2E, "dsb"},
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x040e, hu, hu_HU)
+ILCID_POSIX_ELEMENT_ARRAY(0x042b, hy, hy_AM)
+
+ILCID_POSIX_SUBTABLE(ibb) {
+    {0x69, "ibb"},
+    {0x0469, "ibb_NG"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0421, id, id_ID)
+ILCID_POSIX_ELEMENT_ARRAY(0x0470, ig, ig_NG)
+ILCID_POSIX_ELEMENT_ARRAY(0x0478, ii, ii_CN)
+ILCID_POSIX_ELEMENT_ARRAY(0x040f, is, is_IS)
+
+ILCID_POSIX_SUBTABLE(it) {
+    {0x10,   "it"},
+    {0x0810, "it_CH"},
+    {0x0410, "it_IT"}
+};
+
+ILCID_POSIX_SUBTABLE(iu) {
+    {0x5d,   "iu"},
+    {0x045d, "iu_Cans_CA"},
+    {0x785d, "iu_Cans"},
+    {0x085d, "iu_Latn_CA"},
+    {0x7c5d, "iu_Latn"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x040d, iw, iw_IL)    /*Left in for compatibility*/
+ILCID_POSIX_ELEMENT_ARRAY(0x0411, ja, ja_JP)
+ILCID_POSIX_ELEMENT_ARRAY(0x0437, ka, ka_GE)
+ILCID_POSIX_ELEMENT_ARRAY(0x043f, kk, kk_KZ)
+ILCID_POSIX_ELEMENT_ARRAY(0x046f, kl, kl_GL)
+ILCID_POSIX_ELEMENT_ARRAY(0x0453, km, km_KH)
+ILCID_POSIX_ELEMENT_ARRAY(0x044b, kn, kn_IN)
+
+ILCID_POSIX_SUBTABLE(ko) {
+    {0x12,   "ko"},
+    {0x0812, "ko_KP"},
+    {0x0412, "ko_KR"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0457, kok, kok_IN)
+ILCID_POSIX_ELEMENT_ARRAY(0x0471, kr,  kr_NG)
+
+ILCID_POSIX_SUBTABLE(ks) {         /* We could add PK and CN too */
+    {0x60,   "ks"},
+    {0x0460, "ks_Arab_IN"},
+    {0x0860, "ks_Deva_IN"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0440, ky, ky_KG)   /* Kyrgyz is spoken in Kyrgyzstan */
+
+ILCID_POSIX_SUBTABLE(la) {
+    {0x76,   "la"},
+    {0x0476, "la_001"},
+    {0x0476, "la_IT"}       /*Left in for compatibility*/
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x046e, lb, lb_LU)
+ILCID_POSIX_ELEMENT_ARRAY(0x0454, lo, lo_LA)
+ILCID_POSIX_ELEMENT_ARRAY(0x0427, lt, lt_LT)
+ILCID_POSIX_ELEMENT_ARRAY(0x0426, lv, lv_LV)
+ILCID_POSIX_ELEMENT_ARRAY(0x0481, mi, mi_NZ)
+ILCID_POSIX_ELEMENT_ARRAY(0x042f, mk, mk_MK)
+ILCID_POSIX_ELEMENT_ARRAY(0x044c, ml, ml_IN)
+
+ILCID_POSIX_SUBTABLE(mn) {
+    {0x50,   "mn"},
+    {0x0450, "mn_MN"},
+    {0x7c50, "mn_Mong"},
+    {0x0850, "mn_Mong_CN"},
+    {0x0850, "mn_CN"},
+    {0x7850, "mn_Cyrl"},
+    {0x0c50, "mn_Mong_MN"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0458, mni,mni_IN)
+ILCID_POSIX_ELEMENT_ARRAY(0x047c, moh,moh_CA)
+ILCID_POSIX_ELEMENT_ARRAY(0x044e, mr, mr_IN)
+
+ILCID_POSIX_SUBTABLE(ms) {
+    {0x3e,   "ms"},
+    {0x083e, "ms_BN"},   /* Brunei Darussalam*/
+    {0x043e, "ms_MY"}    /* Malaysia*/
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x043a, mt, mt_MT)
+ILCID_POSIX_ELEMENT_ARRAY(0x0455, my, my_MM)
+
+ILCID_POSIX_SUBTABLE(ne) {
+    {0x61,   "ne"},
+    {0x0861, "ne_IN"},   /* India*/
+    {0x0461, "ne_NP"}    /* Nepal*/
+};
+
+ILCID_POSIX_SUBTABLE(nl) {
+    {0x13,   "nl"},
+    {0x0813, "nl_BE"},
+    {0x0413, "nl_NL"}
+};
+
+/* The "no" locale split into nb and nn.  By default in ICU, "no" is nb.*/
+// TODO: Not all of these are needed on Windows, but I don't know how ICU treats preferred ones here.
+ILCID_POSIX_SUBTABLE(no) {
+    {0x14,   "no"},     /* really nb_NO - actually Windows differentiates between neutral (no region) and specific (with region) */ 
+    {0x7c14, "nb"},     /* really nb */
+    {0x0414, "nb_NO"},  /* really nb_NO. Keep first in the 414 list. */
+    {0x0414, "no_NO"},  /* really nb_NO */
+    {0x0814, "nn_NO"},  /* really nn_NO. Keep first in the 814 list.  */
+    {0x7814, "nn"},     /* It's 0x14 or 0x814, pick one to make the test program happy. */
+    {0x0814, "no_NO_NY"}/* really nn_NO */
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x046c, nso,nso_ZA)   /* TODO: Verify the ISO-639 code */
+ILCID_POSIX_ELEMENT_ARRAY(0x0482, oc, oc_FR)
+
+ILCID_POSIX_SUBTABLE(om) { /* TODO: Verify the country */
+    {0x72,   "om"},
+    {0x0472, "om_ET"},
+    {0x0472, "gaz_ET"}
+};
+
+/* Declared as or_IN to get around compiler errors*/
+ILCID_POSIX_SUBTABLE(or_IN) {
+    {0x48,   "or"},
+    {0x0448, "or_IN"},
+};
+
+ILCID_POSIX_SUBTABLE(pa) {
+    {0x46,   "pa"},
+    {0x0446, "pa_IN"},
+    {0x0846, "pa_Arab_PK"},
+    {0x0846, "pa_PK"}
+};
+
+ILCID_POSIX_SUBTABLE(pap) {
+    {0x79, "pap"},
+    {0x0479, "pap_029"},
+    {0x0479, "pap_AN"}     /*Left in for compatibility*/
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0415, pl, pl_PL)
+ILCID_POSIX_ELEMENT_ARRAY(0x0463, ps, ps_AF)
+
+ILCID_POSIX_SUBTABLE(pt) {
+    {0x16,   "pt"},
+    {0x0416, "pt_BR"},
+    {0x0816, "pt_PT"}
+};
+
+ILCID_POSIX_SUBTABLE(qu) {
+    {0x6b,   "qu"},
+    {0x046b, "qu_BO"},
+    {0x086b, "qu_EC"},
+    {0x0C6b, "qu_PE"},
+    {0x046b, "quz_BO"},
+    {0x086b, "quz_EC"},
+    {0x0C6b, "quz_PE"}
+};
+
+ILCID_POSIX_SUBTABLE(quc) {
+    {0x93,   "quc"},
+    {0x0493, "quc_CO"},
+    /*
+        "quc_Latn_GT" is an exceptional case. Language ID of "quc"
+        is 0x93, but LCID of "quc_Latn_GT" is 0x486, which should be
+        under the group of "qut". "qut" is a retired ISO 639-3 language
+        code for West Central Quiche, and merged to "quc".
+        It looks Windows previously reserved "qut" for K'iche', but,
+        decided to use "quc" when adding a locale for K'iche' (Guatemala).
+
+        This data structure used here assumes language ID bits in
+        LCID is unique for alphabetic language code. But this is not true
+        for "quc_Latn_GT". If we don't have the data below, LCID look up
+        by alphabetic locale ID (POSIX) will fail. The same entry is found
+        under "qut" below, which is required for reverse look up.
+    */
+    {0x0486, "quc_Latn_GT"}
+};
+
+ILCID_POSIX_SUBTABLE(qut) {
+    {0x86,   "qut"},
+    {0x0486, "qut_GT"},
+    /*
+        See the note in "quc" above.
+    */
+    {0x0486, "quc_Latn_GT"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0417, rm, rm_CH)
+
+ILCID_POSIX_SUBTABLE(ro) {
+    {0x18,   "ro"},
+    {0x0418, "ro_RO"},
+    {0x0818, "ro_MD"}
+};
+
+// TODO: This is almost certainly 'wrong'.  0 in Windows is a synonym for LOCALE_USER_DEFAULT.
+// More likely this is a similar concept to the Windows 0x7f Invariant locale ""
+// (Except that it's not invariant in ICU)
+ILCID_POSIX_SUBTABLE(root) {
+    {0x00,   "root"}
+};
+
+ILCID_POSIX_SUBTABLE(ru) {
+    {0x19,   "ru"},
+    {0x0419, "ru_RU"},
+    {0x0819, "ru_MD"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0487, rw, rw_RW)
+ILCID_POSIX_ELEMENT_ARRAY(0x044f, sa, sa_IN)
+ILCID_POSIX_ELEMENT_ARRAY(0x0485, sah,sah_RU)
+
+ILCID_POSIX_SUBTABLE(sd) {
+    {0x59,   "sd"},
+    {0x0459, "sd_Deva_IN"},
+    {0x0459, "sd_IN"},
+    {0x0859, "sd_Arab_PK"},
+    {0x0859, "sd_PK"},
+    {0x7c59, "sd_Arab"}
+};
+
+ILCID_POSIX_SUBTABLE(se) {
+    {0x3b,   "se"},
+    {0x0c3b, "se_FI"},
+    {0x043b, "se_NO"},
+    {0x083b, "se_SE"},
+    {0x783b, "sma"},
+    {0x183b, "sma_NO"},
+    {0x1c3b, "sma_SE"},
+    {0x7c3b, "smj"},
+    {0x703b, "smn"},
+    {0x743b, "sms"},
+    {0x103b, "smj_NO"},
+    {0x143b, "smj_SE"},
+    {0x243b, "smn_FI"},
+    {0x203b, "sms_FI"},
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x045b, si, si_LK)
+ILCID_POSIX_ELEMENT_ARRAY(0x041b, sk, sk_SK)
+ILCID_POSIX_ELEMENT_ARRAY(0x0424, sl, sl_SI)
+
+ILCID_POSIX_SUBTABLE(so) {
+    {0x77,   "so"},
+    {0x0477, "so_SO"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x041c, sq, sq_AL)
+ILCID_POSIX_ELEMENT_ARRAY(0x0430, st, st_ZA)
+
+ILCID_POSIX_SUBTABLE(sv) {
+    {0x1d,   "sv"},
+    {0x081d, "sv_FI"},
+    {0x041d, "sv_SE"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0441, sw, sw_KE)
+ILCID_POSIX_ELEMENT_ARRAY(0x045A, syr, syr_SY)
+
+ILCID_POSIX_SUBTABLE(ta) {
+    {0x49,   "ta"},
+    {0x0449, "ta_IN"},
+    {0x0849, "ta_LK"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x044a, te, te_IN)
+
+/* Cyrillic based by default */
+ILCID_POSIX_SUBTABLE(tg) {
+    {0x28,   "tg"},
+    {0x7c28, "tg_Cyrl"},
+    {0x0428, "tg_Cyrl_TJ"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x041e, th, th_TH)
+
+ILCID_POSIX_SUBTABLE(ti) {
+    {0x73,   "ti"},
+    {0x0873, "ti_ER"},
+    {0x0473, "ti_ET"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0442, tk, tk_TM)
+
+ILCID_POSIX_SUBTABLE(tn) {
+    {0x32,   "tn"},
+    {0x0832, "tn_BW"},
+    {0x0432, "tn_ZA"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x041f, tr, tr_TR)
+ILCID_POSIX_ELEMENT_ARRAY(0x0431, ts, ts_ZA)
+ILCID_POSIX_ELEMENT_ARRAY(0x0444, tt, tt_RU)
+
+ILCID_POSIX_SUBTABLE(tzm) {
+    {0x5f,   "tzm"},
+    {0x7c5f, "tzm_Latn"},
+    {0x085f, "tzm_Latn_DZ"},
+    {0x105f, "tzm_Tfng_MA"},
+    {0x045f, "tzm_Arab_MA"},
+    {0x045f, "tmz"}
+};
+
+ILCID_POSIX_SUBTABLE(ug) {
+    {0x80,   "ug"},
+    {0x0480, "ug_CN"},
+    {0x0480, "ug_Arab_CN"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0422, uk, uk_UA)
+
+ILCID_POSIX_SUBTABLE(ur) {
+    {0x20,   "ur"},
+    {0x0820, "ur_IN"},
+    {0x0420, "ur_PK"}
+};
+
+ILCID_POSIX_SUBTABLE(uz) {
+    {0x43,   "uz"},
+    {0x0843, "uz_Cyrl_UZ"},  /* Cyrillic based */
+    {0x7843, "uz_Cyrl"},  /* Cyrillic based */
+    {0x0843, "uz_UZ"},  /* Cyrillic based */
+    {0x0443, "uz_Latn_UZ"}, /* Latin based */
+    {0x7c43, "uz_Latn"} /* Latin based */
+};
+
+ILCID_POSIX_SUBTABLE(ve) { /* TODO: Verify the country */
+    {0x33,   "ve"},
+    {0x0433, "ve_ZA"},
+    {0x0433, "ven_ZA"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x042a, vi, vi_VN)
+ILCID_POSIX_ELEMENT_ARRAY(0x0488, wo, wo_SN)
+ILCID_POSIX_ELEMENT_ARRAY(0x0434, xh, xh_ZA)
+
+ILCID_POSIX_SUBTABLE(yi) {
+    {0x003d, "yi"},
+    {0x043d, "yi_001"}
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x046a, yo, yo_NG)
+
+// Windows & ICU tend to different names for some of these
+// TODO: Windows probably does not need all of these entries, but I don't know how the precedence works.
+ILCID_POSIX_SUBTABLE(zh) {
+    {0x0004, "zh_Hans"},
+    {0x7804, "zh"},
+    {0x0804, "zh_CN"},
+    {0x0804, "zh_Hans_CN"},
+    {0x0c04, "zh_Hant_HK"},
+    {0x0c04, "zh_HK"},
+    {0x1404, "zh_Hant_MO"},
+    {0x1404, "zh_MO"},
+    {0x1004, "zh_Hans_SG"},
+    {0x1004, "zh_SG"},
+    {0x0404, "zh_Hant_TW"},
+    {0x7c04, "zh_Hant"},
+    {0x0404, "zh_TW"},
+    {0x30404,"zh_Hant_TW"},     /* Bopomofo order */
+    {0x30404,"zh_TW"},          /* Bopomofo order */
+    {0x20004,"zh@collation=stroke"},
+    {0x20404,"zh_Hant@collation=stroke"},
+    {0x20404,"zh_Hant_TW@collation=stroke"},
+    {0x20404,"zh_TW@collation=stroke"},
+    {0x20804,"zh_Hans@collation=stroke"},
+    {0x20804,"zh_Hans_CN@collation=stroke"},
+    {0x20804,"zh_CN@collation=stroke"}
+    // TODO: Alternate collations for other LCIDs are missing, eg: 0x50804
+};
+
+ILCID_POSIX_ELEMENT_ARRAY(0x0435, zu, zu_ZA)
+
+/* This must be static and grouped by LCID. */
+static const ILcidPosixMap gPosixIDmap[] = {
+    ILCID_POSIX_MAP(af),    /*  af  Afrikaans                 0x36 */
+    ILCID_POSIX_MAP(am),    /*  am  Amharic                   0x5e */
+    ILCID_POSIX_MAP(ar),    /*  ar  Arabic                    0x01 */
+    ILCID_POSIX_MAP(arn),   /*  arn Araucanian/Mapudungun     0x7a */
+    ILCID_POSIX_MAP(as),    /*  as  Assamese                  0x4d */
+    ILCID_POSIX_MAP(az),    /*  az  Azerbaijani               0x2c */
+    ILCID_POSIX_MAP(ba),    /*  ba  Bashkir                   0x6d */
+    ILCID_POSIX_MAP(be),    /*  be  Belarusian                0x23 */
+/*    ILCID_POSIX_MAP(ber),     ber Berber/Tamazight          0x5f */
+    ILCID_POSIX_MAP(bg),    /*  bg  Bulgarian                 0x02 */
+    ILCID_POSIX_MAP(bin),   /*  bin Edo                       0x66 */
+    ILCID_POSIX_MAP(bn),    /*  bn  Bengali; Bangla           0x45 */
+    ILCID_POSIX_MAP(bo),    /*  bo  Tibetan                   0x51 */
+    ILCID_POSIX_MAP(br),    /*  br  Breton                    0x7e */
+    ILCID_POSIX_MAP(ca),    /*  ca  Catalan                   0x03 */
+    ILCID_POSIX_MAP(chr),   /*  chr Cherokee                  0x5c */
+    ILCID_POSIX_MAP(ckb),   /*  ckb Sorani (Central Kurdish)  0x92 */
+    ILCID_POSIX_MAP(co),    /*  co  Corsican                  0x83 */
+    ILCID_POSIX_MAP(cs),    /*  cs  Czech                     0x05 */
+    ILCID_POSIX_MAP(cy),    /*  cy  Welsh                     0x52 */
+    ILCID_POSIX_MAP(da),    /*  da  Danish                    0x06 */
+    ILCID_POSIX_MAP(de),    /*  de  German                    0x07 */
+    ILCID_POSIX_MAP(dv),    /*  dv  Divehi                    0x65 */
+    ILCID_POSIX_MAP(el),    /*  el  Greek                     0x08 */
+    ILCID_POSIX_MAP(en),    /*  en  English                   0x09 */
+    ILCID_POSIX_MAP(en_US_POSIX), /*    invariant             0x7f */
+    ILCID_POSIX_MAP(es),    /*  es  Spanish                   0x0a */
+    ILCID_POSIX_MAP(et),    /*  et  Estonian                  0x25 */
+    ILCID_POSIX_MAP(eu),    /*  eu  Basque                    0x2d */
+    ILCID_POSIX_MAP(fa),    /*  fa  Persian/Farsi             0x29 */
+    ILCID_POSIX_MAP(fa_AF), /*  fa  Persian/Dari              0x8c */
+    ILCID_POSIX_MAP(ff),    /*  ff  Fula                      0x67 */
+    ILCID_POSIX_MAP(fi),    /*  fi  Finnish                   0x0b */
+    ILCID_POSIX_MAP(fil),   /*  fil Filipino                  0x64 */
+    ILCID_POSIX_MAP(fo),    /*  fo  Faroese                   0x38 */
+    ILCID_POSIX_MAP(fr),    /*  fr  French                    0x0c */
+    ILCID_POSIX_MAP(fuv),   /*  fuv Fulfulde - Nigeria        0x67 */
+    ILCID_POSIX_MAP(fy),    /*  fy  Frisian                   0x62 */
+    ILCID_POSIX_MAP(ga),    /*  *   Gaelic (Ireland,Scotland) 0x3c */
+    ILCID_POSIX_MAP(gd),    /*  gd  Gaelic (United Kingdom)   0x91 */
+    ILCID_POSIX_MAP(gl),    /*  gl  Galician                  0x56 */
+    ILCID_POSIX_MAP(gn),    /*  gn  Guarani                   0x74 */
+    ILCID_POSIX_MAP(gsw),   /*  gsw Alemanic/Alsatian/Swiss German 0x84 */
+    ILCID_POSIX_MAP(gu),    /*  gu  Gujarati                  0x47 */
+    ILCID_POSIX_MAP(ha),    /*  ha  Hausa                     0x68 */
+    ILCID_POSIX_MAP(haw),   /*  haw Hawaiian                  0x75 */
+    ILCID_POSIX_MAP(he),    /*  he  Hebrew (formerly iw)      0x0d */
+    ILCID_POSIX_MAP(hi),    /*  hi  Hindi                     0x39 */
+    ILCID_POSIX_MAP(hr),    /*  *   Croatian and others       0x1a */
+    ILCID_POSIX_MAP(hsb),   /*  hsb Upper Sorbian             0x2e */
+    ILCID_POSIX_MAP(hu),    /*  hu  Hungarian                 0x0e */
+    ILCID_POSIX_MAP(hy),    /*  hy  Armenian                  0x2b */
+    ILCID_POSIX_MAP(ibb),   /*  ibb Ibibio - Nigeria          0x69 */
+    ILCID_POSIX_MAP(id),    /*  id  Indonesian (formerly in)  0x21 */
+    ILCID_POSIX_MAP(ig),    /*  ig  Igbo                      0x70 */
+    ILCID_POSIX_MAP(ii),    /*  ii  Sichuan Yi                0x78 */
+    ILCID_POSIX_MAP(is),    /*  is  Icelandic                 0x0f */
+    ILCID_POSIX_MAP(it),    /*  it  Italian                   0x10 */
+    ILCID_POSIX_MAP(iu),    /*  iu  Inuktitut                 0x5d */
+    ILCID_POSIX_MAP(iw),    /*  iw  Hebrew                    0x0d */
+    ILCID_POSIX_MAP(ja),    /*  ja  Japanese                  0x11 */
+    ILCID_POSIX_MAP(ka),    /*  ka  Georgian                  0x37 */
+    ILCID_POSIX_MAP(kk),    /*  kk  Kazakh                    0x3f */
+    ILCID_POSIX_MAP(kl),    /*  kl  Kalaallisut               0x6f */
+    ILCID_POSIX_MAP(km),    /*  km  Khmer                     0x53 */
+    ILCID_POSIX_MAP(kn),    /*  kn  Kannada                   0x4b */
+    ILCID_POSIX_MAP(ko),    /*  ko  Korean                    0x12 */
+    ILCID_POSIX_MAP(kok),   /*  kok Konkani                   0x57 */
+    ILCID_POSIX_MAP(kr),    /*  kr  Kanuri                    0x71 */
+    ILCID_POSIX_MAP(ks),    /*  ks  Kashmiri                  0x60 */
+    ILCID_POSIX_MAP(ky),    /*  ky  Kyrgyz                    0x40 */
+    ILCID_POSIX_MAP(lb),    /*  lb  Luxembourgish             0x6e */
+    ILCID_POSIX_MAP(la),    /*  la  Latin                     0x76 */
+    ILCID_POSIX_MAP(lo),    /*  lo  Lao                       0x54 */
+    ILCID_POSIX_MAP(lt),    /*  lt  Lithuanian                0x27 */
+    ILCID_POSIX_MAP(lv),    /*  lv  Latvian, Lettish          0x26 */
+    ILCID_POSIX_MAP(mi),    /*  mi  Maori                     0x81 */
+    ILCID_POSIX_MAP(mk),    /*  mk  Macedonian                0x2f */
+    ILCID_POSIX_MAP(ml),    /*  ml  Malayalam                 0x4c */
+    ILCID_POSIX_MAP(mn),    /*  mn  Mongolian                 0x50 */
+    ILCID_POSIX_MAP(mni),   /*  mni Manipuri                  0x58 */
+    ILCID_POSIX_MAP(moh),   /*  moh Mohawk                    0x7c */
+    ILCID_POSIX_MAP(mr),    /*  mr  Marathi                   0x4e */
+    ILCID_POSIX_MAP(ms),    /*  ms  Malay                     0x3e */
+    ILCID_POSIX_MAP(mt),    /*  mt  Maltese                   0x3a */
+    ILCID_POSIX_MAP(my),    /*  my  Burmese                   0x55 */
+/*    ILCID_POSIX_MAP(nb),    //  no  Norwegian                 0x14 */
+    ILCID_POSIX_MAP(ne),    /*  ne  Nepali                    0x61 */
+    ILCID_POSIX_MAP(nl),    /*  nl  Dutch                     0x13 */
+/*    ILCID_POSIX_MAP(nn),    //  no  Norwegian                 0x14 */
+    ILCID_POSIX_MAP(no),    /*  *   Norwegian                 0x14 */
+    ILCID_POSIX_MAP(nso),   /*  nso Sotho, Northern (Sepedi dialect) 0x6c */
+    ILCID_POSIX_MAP(oc),    /*  oc  Occitan                   0x82 */
+    ILCID_POSIX_MAP(om),    /*  om  Oromo                     0x72 */
+    ILCID_POSIX_MAP(or_IN), /*  or  Oriya                     0x48 */
+    ILCID_POSIX_MAP(pa),    /*  pa  Punjabi                   0x46 */
+    ILCID_POSIX_MAP(pap),   /*  pap Papiamentu                0x79 */
+    ILCID_POSIX_MAP(pl),    /*  pl  Polish                    0x15 */
+    ILCID_POSIX_MAP(ps),    /*  ps  Pashto                    0x63 */
+    ILCID_POSIX_MAP(pt),    /*  pt  Portuguese                0x16 */
+    ILCID_POSIX_MAP(qu),    /*  qu  Quechua                   0x6B */
+    ILCID_POSIX_MAP(quc),   /*  quc K'iche                    0x93 */
+    ILCID_POSIX_MAP(qut),   /*  qut K'iche                    0x86 */
+    ILCID_POSIX_MAP(rm),    /*  rm  Raeto-Romance/Romansh     0x17 */
+    ILCID_POSIX_MAP(ro),    /*  ro  Romanian                  0x18 */
+    ILCID_POSIX_MAP(root),  /*  root                          0x00 */
+    ILCID_POSIX_MAP(ru),    /*  ru  Russian                   0x19 */
+    ILCID_POSIX_MAP(rw),    /*  rw  Kinyarwanda               0x87 */
+    ILCID_POSIX_MAP(sa),    /*  sa  Sanskrit                  0x4f */
+    ILCID_POSIX_MAP(sah),   /*  sah Yakut                     0x85 */
+    ILCID_POSIX_MAP(sd),    /*  sd  Sindhi                    0x59 */
+    ILCID_POSIX_MAP(se),    /*  se  Sami                      0x3b */
+/*    ILCID_POSIX_MAP(sh),    //  sh  Serbo-Croatian            0x1a */
+    ILCID_POSIX_MAP(si),    /*  si  Sinhalese                 0x5b */
+    ILCID_POSIX_MAP(sk),    /*  sk  Slovak                    0x1b */
+    ILCID_POSIX_MAP(sl),    /*  sl  Slovenian                 0x24 */
+    ILCID_POSIX_MAP(so),    /*  so  Somali                    0x77 */
+    ILCID_POSIX_MAP(sq),    /*  sq  Albanian                  0x1c */
+/*    ILCID_POSIX_MAP(sr),    //  sr  Serbian                   0x1a */
+    ILCID_POSIX_MAP(st),    /*  st  Sutu                      0x30 */
+    ILCID_POSIX_MAP(sv),    /*  sv  Swedish                   0x1d */
+    ILCID_POSIX_MAP(sw),    /*  sw  Swahili                   0x41 */
+    ILCID_POSIX_MAP(syr),   /*  syr Syriac                    0x5A */
+    ILCID_POSIX_MAP(ta),    /*  ta  Tamil                     0x49 */
+    ILCID_POSIX_MAP(te),    /*  te  Telugu                    0x4a */
+    ILCID_POSIX_MAP(tg),    /*  tg  Tajik                     0x28 */
+    ILCID_POSIX_MAP(th),    /*  th  Thai                      0x1e */
+    ILCID_POSIX_MAP(ti),    /*  ti  Tigrigna                  0x73 */
+    ILCID_POSIX_MAP(tk),    /*  tk  Turkmen                   0x42 */
+    ILCID_POSIX_MAP(tn),    /*  tn  Tswana                    0x32 */
+    ILCID_POSIX_MAP(tr),    /*  tr  Turkish                   0x1f */
+    ILCID_POSIX_MAP(ts),    /*  ts  Tsonga                    0x31 */
+    ILCID_POSIX_MAP(tt),    /*  tt  Tatar                     0x44 */
+    ILCID_POSIX_MAP(tzm),   /*  tzm Tamazight                 0x5f */
+    ILCID_POSIX_MAP(ug),    /*  ug  Uighur                    0x80 */
+    ILCID_POSIX_MAP(uk),    /*  uk  Ukrainian                 0x22 */
+    ILCID_POSIX_MAP(ur),    /*  ur  Urdu                      0x20 */
+    ILCID_POSIX_MAP(uz),    /*  uz  Uzbek                     0x43 */
+    ILCID_POSIX_MAP(ve),    /*  ve  Venda                     0x33 */
+    ILCID_POSIX_MAP(vi),    /*  vi  Vietnamese                0x2a */
+    ILCID_POSIX_MAP(wo),    /*  wo  Wolof                     0x88 */
+    ILCID_POSIX_MAP(xh),    /*  xh  Xhosa                     0x34 */
+    ILCID_POSIX_MAP(yi),    /*  yi  Yiddish                   0x3d */
+    ILCID_POSIX_MAP(yo),    /*  yo  Yoruba                    0x6a */
+    ILCID_POSIX_MAP(zh),    /*  zh  Chinese                   0x04 */
+    ILCID_POSIX_MAP(zu),    /*  zu  Zulu                      0x35 */
+};
+
+static const uint32_t gLocaleCount = UPRV_LENGTHOF(gPosixIDmap);
+
+/**
+ * Do not call this function. It is called by hostID.
+ * The function is not private because this struct must stay as a C struct,
+ * and this is an internal class.
+ */
+static int32_t
+idCmp(const char* id1, const char* id2)
+{
+    int32_t diffIdx = 0;
+    while (*id1 == *id2 && *id1 != 0) {
+        diffIdx++;
+        id1++;
+        id2++;
+    }
+    return diffIdx;
+}
+
+/**
+ * Searches for a Windows LCID
+ *
+ * @param posixID the Posix style locale id.
+ * @param status gets set to U_ILLEGAL_ARGUMENT_ERROR when the Posix ID has
+ *               no equivalent Windows LCID.
+ * @return the LCID
+ */
+static uint32_t
+getHostID(const ILcidPosixMap *this_0, const char* posixID, UErrorCode* status)
+{
+    int32_t bestIdx = 0;
+    int32_t bestIdxDiff = 0;
+    int32_t posixIDlen = (int32_t)uprv_strlen(posixID);
+    uint32_t idx;
+
+    for (idx = 0; idx < this_0->numRegions; idx++ ) {
+        int32_t sameChars = idCmp(posixID, this_0->regionMaps[idx].posixID);
+        if (sameChars > bestIdxDiff && this_0->regionMaps[idx].posixID[sameChars] == 0) {
+            if (posixIDlen == sameChars) {
+                /* Exact match */
+                return this_0->regionMaps[idx].hostID;
+            }
+            bestIdxDiff = sameChars;
+            bestIdx = idx;
+        }
+    }
+    /* We asked for something unusual, like en_ZZ, and we try to return the number for the same language. */
+    /* We also have to make sure that sid and si and similar string subsets don't match. */
+    if ((posixID[bestIdxDiff] == '_' || posixID[bestIdxDiff] == '@')
+        && this_0->regionMaps[bestIdx].posixID[bestIdxDiff] == 0)
+    {
+        *status = U_USING_FALLBACK_WARNING;
+        return this_0->regionMaps[bestIdx].hostID;
+    }
+
+    /*no match found */
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return this_0->regionMaps->hostID;
+}
+
+static const char*
+getPosixID(const ILcidPosixMap *this_0, uint32_t hostID)
+{
+    uint32_t i;
+    for (i = 0; i < this_0->numRegions; i++)
+    {
+        if (this_0->regionMaps[i].hostID == hostID)
+        {
+            return this_0->regionMaps[i].posixID;
+        }
+    }
+
+    /* If you get here, then no matching region was found,
+       so return the language id with the wild card region. */
+    return this_0->regionMaps[0].posixID;
+}
+
+/*
+//////////////////////////////////////
+//
+// LCID --> POSIX
+//
+/////////////////////////////////////
+*/
+#if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+/*
+ * Various language tags needs to be changed:
+ * quz -> qu
+ * prs -> fa
+ */
+#define FIX_LANGUAGE_ID_TAG(buffer, len) \
+    if (len >= 3) { \
+        if (buffer[0] == 'q' && buffer[1] == 'u' && buffer[2] == 'z') {\
+            buffer[2] = 0; \
+            uprv_strcat(buffer, buffer+3); \
+        } else if (buffer[0] == 'p' && buffer[1] == 'r' && buffer[2] == 's') {\
+            buffer[0] = 'f'; buffer[1] = 'a'; buffer[2] = 0; \
+            uprv_strcat(buffer, buffer+3); \
+        } \
+    }
+
+#endif
+
+U_CAPI int32_t
+uprv_convertToPosix(uint32_t hostid, char *posixID, int32_t posixIDCapacity, UErrorCode* status)
+{
+    uint16_t langID;
+    uint32_t localeIndex;
+    UBool bLookup = TRUE;
+    const char *pPosixID = NULL;
+
+#if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+    static_assert(ULOC_FULLNAME_CAPACITY > LOCALE_NAME_MAX_LENGTH, "Windows locale names have smaller length than ICU locale names.");
+
+    char locName[LOCALE_NAME_MAX_LENGTH] = {};
+
+    // Note: Windows primary lang ID 0x92 in LCID is used for Central Kurdish and
+    // GetLocaleInfo() maps such LCID to "ku". However, CLDR uses "ku" for
+    // Northern Kurdish and "ckb" for Central Kurdish. For this reason, we cannot
+    // use the Windows API to resolve locale ID for this specific case.
+    if ((hostid & 0x3FF) != 0x92) {
+        int32_t tmpLen = 0;
+        char16_t windowsLocaleName[LOCALE_NAME_MAX_LENGTH] = {};
+
+        // Note: LOCALE_ALLOW_NEUTRAL_NAMES was enabled in Windows7+, prior versions did not handle neutral (no-region) locale names.
+        tmpLen = LCIDToLocaleName(hostid, (PWSTR)windowsLocaleName, UPRV_LENGTHOF(windowsLocaleName), LOCALE_ALLOW_NEUTRAL_NAMES);
+        if (tmpLen > 1) {
+            int32_t i = 0;
+            // Only need to look up in table if have _, eg for de-de_phoneb type alternate sort.
+            bLookup = FALSE;
+            for (i = 0; i < UPRV_LENGTHOF(locName); i++)
+            {
+                locName[i] = (char)(windowsLocaleName[i]);
+
+                // Windows locale name may contain sorting variant, such as "es-ES_tradnl".
+                // In such cases, we need special mapping data found in the hardcoded table
+                // in this source file.
+                if (windowsLocaleName[i] == L'_')
+                {
+                    // Keep the base locale, without variant
+                    // TODO: Should these be mapped from _phoneb to @collation=phonebook, etc.?
+                    locName[i] = '\0';
+                    tmpLen = i;
+                    bLookup = TRUE;
+                    break;
+                }
+                else if (windowsLocaleName[i] == L'-')
+                {
+                    // Windows names use -, ICU uses _
+                    locName[i] = '_';
+                }
+                else if (windowsLocaleName[i] == L'\0')
+                {
+                    // No point in doing more work than necessary
+                    break;
+                }
+            }
+            // TODO: Need to understand this better, why isn't it an alias?
+            FIX_LANGUAGE_ID_TAG(locName, tmpLen);
+            pPosixID = locName;
+        }
+    }
+#endif
+
+    if (bLookup) {
+        const char *pCandidate = NULL;
+        langID = LANGUAGE_LCID(hostid);
+
+        for (localeIndex = 0; localeIndex < gLocaleCount; localeIndex++) {
+            if (langID == gPosixIDmap[localeIndex].regionMaps->hostID) {
+                pCandidate = getPosixID(&gPosixIDmap[localeIndex], hostid);
+                break;
+            }
+        }
+
+        /* On Windows, when locale name has a variant, we still look up the hardcoded table.
+           If a match in the hardcoded table is longer than the Windows locale name without
+           variant, we use the one as the result */
+        if (pCandidate && (pPosixID == NULL || uprv_strlen(pCandidate) > uprv_strlen(pPosixID))) {
+            pPosixID = pCandidate;
+        }
+    }
+
+    if (pPosixID) {
+        int32_t resLen = static_cast<int32_t>(uprv_strlen(pPosixID));
+        int32_t copyLen = resLen <= posixIDCapacity ? resLen : posixIDCapacity;
+        uprv_memcpy(posixID, pPosixID, copyLen);
+        if (resLen < posixIDCapacity) {
+            posixID[resLen] = 0;
+            if (*status == U_STRING_NOT_TERMINATED_WARNING) {
+                *status = U_ZERO_ERROR;
+            }
+        } else if (resLen == posixIDCapacity) {
+            *status = U_STRING_NOT_TERMINATED_WARNING;
+        } else {
+            *status = U_BUFFER_OVERFLOW_ERROR;
+        }
+        return resLen;
+    }
+
+    /* no match found */
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return -1;
+}
+
+/*
+//////////////////////////////////////
+//
+// POSIX --> LCID
+// This should only be called from uloc_getLCID.
+// The locale ID must be in canonical form.
+//
+/////////////////////////////////////
+*/
+U_CAPI uint32_t
+uprv_convertToLCIDPlatform(const char* localeID, UErrorCode* status)
+{
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    // The purpose of this function is to leverage the Windows platform name->lcid
+    // conversion functionality when available.
+#if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+    int32_t len;
+    char baseName[ULOC_FULLNAME_CAPACITY] = {};
+    const char * mylocaleID = localeID;
+
+    // Check any for keywords.
+    if (uprv_strchr(localeID, '@'))
+    {
+        icu::CharString collVal;
+        {
+            icu::CharStringByteSink sink(&collVal);
+            ulocimp_getKeywordValue(localeID, "collation", sink, status);
+        }
+        if (U_SUCCESS(*status) && !collVal.isEmpty())
+        {
+            // If it contains the keyword collation, return 0 so that the LCID lookup table will be used.
+            return 0;
+        }
+        else
+        {
+            // If the locale ID contains keywords other than collation, just use the base name.
+            len = uloc_getBaseName(localeID, baseName, UPRV_LENGTHOF(baseName) - 1, status);
+
+            if (U_SUCCESS(*status) && len > 0)
+            {
+                baseName[len] = 0;
+                mylocaleID = baseName;
+            }
+        }
+    }
+
+    char asciiBCP47Tag[LOCALE_NAME_MAX_LENGTH] = {};
+    // this will change it from de_DE@collation=phonebook to de-DE-u-co-phonebk form
+    (void)uloc_toLanguageTag(mylocaleID, asciiBCP47Tag, UPRV_LENGTHOF(asciiBCP47Tag), FALSE, status);
+
+    if (U_SUCCESS(*status))
+    {
+        // Need it to be UTF-16, not 8-bit
+        wchar_t bcp47Tag[LOCALE_NAME_MAX_LENGTH] = {};
+        int32_t i;
+        for (i = 0; i < UPRV_LENGTHOF(bcp47Tag); i++)
+        {
+            if (asciiBCP47Tag[i] == '\0')
+            {
+                break;
+            }
+            else
+            {
+                // Copy the character
+                bcp47Tag[i] = static_cast<wchar_t>(asciiBCP47Tag[i]);
+            }
+        }
+
+        if (i < (UPRV_LENGTHOF(bcp47Tag) - 1))
+        {
+            // Ensure it's null terminated
+            bcp47Tag[i] = L'\0';
+            LCID lcid = LocaleNameToLCID(bcp47Tag, LOCALE_ALLOW_NEUTRAL_NAMES);
+            if (lcid > 0)
+            {
+                // Found LCID from windows, return that one, unless its completely ambiguous
+                // LOCALE_USER_DEFAULT and transients are OK because they will round trip
+                // for this process.
+                if (lcid != LOCALE_CUSTOM_UNSPECIFIED)
+                {
+                    return lcid;
+                }
+            }
+        }
+    }
+#else
+    (void) localeID; // Suppress unused variable warning.
+#endif
+
+    // Nothing found, or not implemented.
+    return 0;
+}
+
+U_CAPI uint32_t
+uprv_convertToLCID(const char *langID, const char* posixID, UErrorCode* status)
+{
+    // This function does the table lookup when native platform name->lcid conversion isn't available,
+    // or for locales that don't follow patterns the platform expects.
+    uint32_t   low    = 0;
+    uint32_t   high   = gLocaleCount;
+    uint32_t   mid;
+    uint32_t   oldmid = 0;
+    int32_t    compVal;
+
+    uint32_t   value         = 0;
+    uint32_t   fallbackValue = (uint32_t)-1;
+    UErrorCode myStatus;
+    uint32_t   idx;
+
+    /* Check for incomplete id. */
+    if (!langID || !posixID || uprv_strlen(langID) < 2 || uprv_strlen(posixID) < 2) {
+        return 0;
+    }
+
+    /*Binary search for the map entry for normal cases */
+
+    while (high > low)  /*binary search*/{
+
+        mid = (high+low) >> 1; /*Finds median*/
+
+        if (mid == oldmid) 
+            break;
+
+        compVal = uprv_strcmp(langID, gPosixIDmap[mid].regionMaps->posixID);
+        if (compVal < 0){
+            high = mid;
+        }
+        else if (compVal > 0){
+            low = mid;
+        }
+        else /*we found it*/{
+            return getHostID(&gPosixIDmap[mid], posixID, status);
+        }
+        oldmid = mid;
+    }
+
+    /*
+     * Sometimes we can't do a binary search on posixID because some LCIDs
+     * go to different locales.  We hit one of those special cases.
+     */
+    for (idx = 0; idx < gLocaleCount; idx++ ) {
+        myStatus = U_ZERO_ERROR;
+        value = getHostID(&gPosixIDmap[idx], posixID, &myStatus);
+        if (myStatus == U_ZERO_ERROR) {
+            return value;
+        }
+        else if (myStatus == U_USING_FALLBACK_WARNING) {
+            fallbackValue = value;
+        }
+    }
+
+    if (fallbackValue != (uint32_t)-1) {
+        *status = U_USING_FALLBACK_WARNING;
+        return fallbackValue;
+    }
+
+    /* no match found */
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;   /* return international (root) */
+}
diff --git a/thirdparty/icu4c/common/locmap.h b/thirdparty/icu4c/common/locmap.h
new file mode 100644
index 0000000000..e669873a14
--- /dev/null
+++ b/thirdparty/icu4c/common/locmap.h
@@ -0,0 +1,40 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1996-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File locmap.h      : Locale Mapping Classes
+* 
+*
+* Created by: Helena Shih
+*
+* Modification History:
+*
+*  Date        Name        Description
+*  3/11/97     aliu        Added setId().
+*  4/20/99     Madhu       Added T_convertToPosix()
+* 09/18/00     george      Removed the memory leaks.
+* 08/23/01     george      Convert to C
+*============================================================================
+*/
+
+#ifndef LOCMAP_H
+#define LOCMAP_H
+
+#include "unicode/utypes.h"
+
+#define LANGUAGE_LCID(hostID) (uint16_t)(0x03FF & hostID)
+
+U_CAPI int32_t uprv_convertToPosix(uint32_t hostid, char* posixID, int32_t posixIDCapacity, UErrorCode* status);
+
+/* Don't call these functions directly. Use uloc_getLCID instead. */
+U_CAPI uint32_t uprv_convertToLCIDPlatform(const char* localeID, UErrorCode* status); // Leverage platform conversion if possible
+U_CAPI uint32_t uprv_convertToLCID(const char* langID, const char* posixID, UErrorCode* status);
+
+#endif /* LOCMAP_H */
+
diff --git a/thirdparty/icu4c/common/locresdata.cpp b/thirdparty/icu4c/common/locresdata.cpp
new file mode 100644
index 0000000000..d1d9a4729f
--- /dev/null
+++ b/thirdparty/icu4c/common/locresdata.cpp
@@ -0,0 +1,220 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  loclikely.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010feb25
+*   created by: Markus W. Scherer
+*
+*   Code for miscellaneous locale-related resource bundle data access,
+*   separated out from other .cpp files
+*   that then do not depend on resource bundle code and this data.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/uloc.h"
+#include "unicode/ures.h"
+#include "cstring.h"
+#include "ulocimp.h"
+#include "uresimp.h"
+
+/*
+ * Lookup a resource bundle table item with fallback on the table level.
+ * Regular resource bundle lookups perform fallback to parent locale bundles
+ * and eventually the root bundle, but only for top-level items.
+ * This function takes the name of a top-level table and of an item in that table
+ * and performs a lookup of both, falling back until a bundle contains a table
+ * with this item.
+ *
+ * Note: Only the opening of entire bundles falls back through the default locale
+ * before root. Once a bundle is open, item lookups do not go through the
+ * default locale because that would result in a mix of languages that is
+ * unpredictable to the programmer and most likely useless.
+ */
+U_CAPI const UChar * U_EXPORT2
+uloc_getTableStringWithFallback(const char *path, const char *locale,
+                              const char *tableKey, const char *subTableKey,
+                              const char *itemKey,
+                              int32_t *pLength,
+                              UErrorCode *pErrorCode)
+{
+/*    char localeBuffer[ULOC_FULLNAME_CAPACITY*4];*/
+    const UChar *item=NULL;
+    UErrorCode errorCode;
+    char explicitFallbackName[ULOC_FULLNAME_CAPACITY] = {0};
+
+    /*
+     * open the bundle for the current locale
+     * this falls back through the locale's chain to root
+     */
+    errorCode=U_ZERO_ERROR;
+    icu::LocalUResourceBundlePointer rb(ures_open(path, locale, &errorCode));
+
+    if(U_FAILURE(errorCode)) {
+        /* total failure, not even root could be opened */
+        *pErrorCode=errorCode;
+        return NULL;
+    } else if(errorCode==U_USING_DEFAULT_WARNING ||
+                (errorCode==U_USING_FALLBACK_WARNING && *pErrorCode!=U_USING_DEFAULT_WARNING)
+    ) {
+        /* set the "strongest" error code (success->fallback->default->failure) */
+        *pErrorCode=errorCode;
+    }
+
+    for(;;){
+        icu::StackUResourceBundle table;
+        icu::StackUResourceBundle subTable;
+        ures_getByKeyWithFallback(rb.getAlias(), tableKey, table.getAlias(), &errorCode);
+
+        if (subTableKey != NULL) {
+            /*
+            ures_getByKeyWithFallback(table.getAlias(), subTableKey, subTable.getAlias(), &errorCode);
+            item = ures_getStringByKeyWithFallback(subTable.getAlias(), itemKey, pLength, &errorCode);
+            if(U_FAILURE(errorCode)){
+                *pErrorCode = errorCode;
+            }
+            
+            break;*/
+            
+            ures_getByKeyWithFallback(table.getAlias(), subTableKey, table.getAlias(), &errorCode);
+        }
+        if(U_SUCCESS(errorCode)){
+            item = ures_getStringByKeyWithFallback(table.getAlias(), itemKey, pLength, &errorCode);
+            if(U_FAILURE(errorCode)){
+                const char* replacement = NULL;
+                *pErrorCode = errorCode; /*save the errorCode*/
+                errorCode = U_ZERO_ERROR;
+                /* may be a deprecated code */
+                if(uprv_strcmp(tableKey, "Countries")==0){
+                    replacement =  uloc_getCurrentCountryID(itemKey);
+                }else if(uprv_strcmp(tableKey, "Languages")==0){
+                    replacement =  uloc_getCurrentLanguageID(itemKey);
+                }
+                /*pointer comparison is ok since uloc_getCurrentCountryID & uloc_getCurrentLanguageID return the key itself is replacement is not found*/
+                if(replacement!=NULL && itemKey != replacement){
+                    item = ures_getStringByKeyWithFallback(table.getAlias(), replacement, pLength, &errorCode);
+                    if(U_SUCCESS(errorCode)){
+                        *pErrorCode = errorCode;
+                        break;
+                    }
+                }
+            }else{
+                break;
+            }
+        }
+        
+        if(U_FAILURE(errorCode)){    
+
+            /* still can't figure out ?.. try the fallback mechanism */
+            int32_t len = 0;
+            const UChar* fallbackLocale =  NULL;
+            *pErrorCode = errorCode;
+            errorCode = U_ZERO_ERROR;
+
+            fallbackLocale = ures_getStringByKeyWithFallback(table.getAlias(), "Fallback", &len, &errorCode);
+            if(U_FAILURE(errorCode)){
+               *pErrorCode = errorCode;
+                break;
+            }
+            
+            u_UCharsToChars(fallbackLocale, explicitFallbackName, len);
+            
+            /* guard against recursive fallback */
+            if(uprv_strcmp(explicitFallbackName, locale)==0){
+                *pErrorCode = U_INTERNAL_PROGRAM_ERROR;
+                break;
+            }
+            rb.adoptInstead(ures_open(path, explicitFallbackName, &errorCode));
+            if(U_FAILURE(errorCode)){
+                *pErrorCode = errorCode;
+                break;
+            }
+            /* succeeded in opening the fallback bundle .. continue and try to fetch the item */
+        }else{
+            break;
+        }
+    }
+
+    return item;
+}
+
+static ULayoutType
+_uloc_getOrientationHelper(const char* localeId,
+                           const char* key,
+                           UErrorCode *status)
+{
+    ULayoutType result = ULOC_LAYOUT_UNKNOWN;
+
+    if (!U_FAILURE(*status)) {
+        int32_t length = 0;
+        char localeBuffer[ULOC_FULLNAME_CAPACITY];
+
+        uloc_canonicalize(localeId, localeBuffer, sizeof(localeBuffer), status);
+
+        if (!U_FAILURE(*status)) {
+            const UChar* const value =
+                uloc_getTableStringWithFallback(
+                    NULL,
+                    localeBuffer,
+                    "layout",
+                    NULL,
+                    key,
+                    &length,
+                    status);
+
+            if (!U_FAILURE(*status) && length != 0) {
+                switch(value[0])
+                {
+                case 0x0062: /* 'b' */
+                    result = ULOC_LAYOUT_BTT;
+                    break;
+                case 0x006C: /* 'l' */
+                    result = ULOC_LAYOUT_LTR;
+                    break;
+                case 0x0072: /* 'r' */
+                    result = ULOC_LAYOUT_RTL;
+                    break;
+                case 0x0074: /* 't' */
+                    result = ULOC_LAYOUT_TTB;
+                    break;
+                default:
+                    *status = U_INTERNAL_PROGRAM_ERROR;
+                    break;
+                }
+            }
+        }
+    }
+
+    return result;
+}
+
+U_CAPI ULayoutType U_EXPORT2
+uloc_getCharacterOrientation(const char* localeId,
+                             UErrorCode *status)
+{
+    return _uloc_getOrientationHelper(localeId, "characters", status);
+}
+
+/**
+ * Get the layout line orientation for the specified locale.
+ * 
+ * @param localeID locale name
+ * @param status Error status
+ * @return an enum indicating the layout orientation for lines.
+ */
+U_CAPI ULayoutType U_EXPORT2
+uloc_getLineOrientation(const char* localeId,
+                        UErrorCode *status)
+{
+    return _uloc_getOrientationHelper(localeId, "lines", status);
+}
diff --git a/thirdparty/icu4c/common/locutil.cpp b/thirdparty/icu4c/common/locutil.cpp
new file mode 100644
index 0000000000..3d9d69ff7e
--- /dev/null
+++ b/thirdparty/icu4c/common/locutil.cpp
@@ -0,0 +1,275 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ * Copyright (C) 2002-2014, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE || !UCONFIG_NO_TRANSLITERATION
+
+#include "unicode/resbund.h"
+#include "unicode/uenum.h"
+#include "cmemory.h"
+#include "ustrfmt.h"
+#include "locutil.h"
+#include "charstr.h"
+#include "ucln_cmn.h"
+#include "uassert.h"
+#include "umutex.h"
+
+// see LocaleUtility::getAvailableLocaleNames
+static icu::UInitOnce   LocaleUtilityInitOnce = U_INITONCE_INITIALIZER;
+static icu::Hashtable * LocaleUtility_cache = NULL;
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * Release all static memory held by Locale Utility.  
+ */
+U_CDECL_BEGIN
+static UBool U_CALLCONV service_cleanup(void) {
+    if (LocaleUtility_cache) {
+        delete LocaleUtility_cache;
+        LocaleUtility_cache = NULL;
+    }
+    return TRUE;
+}
+
+
+static void U_CALLCONV locale_utility_init(UErrorCode &status) {
+    using namespace icu;
+    U_ASSERT(LocaleUtility_cache == NULL);
+    ucln_common_registerCleanup(UCLN_COMMON_SERVICE, service_cleanup);
+    LocaleUtility_cache = new Hashtable(status);
+    if (U_FAILURE(status)) {
+        delete LocaleUtility_cache;
+        LocaleUtility_cache = NULL;
+        return;
+    }
+    if (LocaleUtility_cache == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    LocaleUtility_cache->setValueDeleter(uhash_deleteHashtable);
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+UnicodeString&
+LocaleUtility::canonicalLocaleString(const UnicodeString* id, UnicodeString& result)
+{
+  if (id == NULL) {
+    result.setToBogus();
+  } else {
+    // Fix case only (no other changes) up to the first '@' or '.' or
+    // end of string, whichever comes first.  In 3.0 I changed this to
+    // stop at first '@' or '.'.  It used to run out to the end of
+    // string.  My fix makes the tests pass but is probably
+    // structurally incorrect.  See below.  [alan 3.0]
+
+    // TODO: Doug, you might want to revise this...
+    result = *id;
+    int32_t i = 0;
+    int32_t end = result.indexOf(AT_SIGN_CHAR);
+    int32_t n = result.indexOf(PERIOD_CHAR);
+    if (n >= 0 && n < end) {
+        end = n;
+    }
+    if (end < 0) {
+        end = result.length();
+    }
+    n = result.indexOf(UNDERSCORE_CHAR);
+    if (n < 0) {
+      n = end;
+    }
+    for (; i < n; ++i) {
+      UChar c = result.charAt(i);
+      if (c >= 0x0041 && c <= 0x005a) {
+        c += 0x20;
+        result.setCharAt(i, c);
+      }
+    }
+    for (n = end; i < n; ++i) {
+      UChar c = result.charAt(i);
+      if (c >= 0x0061 && c <= 0x007a) {
+        c -= 0x20;
+        result.setCharAt(i, c);
+      }
+    }
+  }
+  return result;
+
+#if 0
+    // This code does a proper full level 2 canonicalization of id.
+    // It's nasty to go from UChar to char to char to UChar -- but
+    // that's what you have to do to use the uloc_canonicalize
+    // function on UnicodeStrings.
+
+    // I ended up doing the alternate fix (see above) not for
+    // performance reasons, although performance will certainly be
+    // better, but because doing a full level 2 canonicalization
+    // causes some tests to fail.  [alan 3.0]
+
+    // TODO: Doug, you might want to revisit this...
+    result.setToBogus();
+    if (id != 0) {
+        int32_t buflen = id->length() + 8; // space for NUL
+        char* buf = (char*) uprv_malloc(buflen);
+        char* canon = (buf == 0) ? 0 : (char*) uprv_malloc(buflen);
+        if (buf != 0 && canon != 0) {
+            U_ASSERT(id->extract(0, INT32_MAX, buf, buflen) < buflen);
+            UErrorCode ec = U_ZERO_ERROR;
+            uloc_canonicalize(buf, canon, buflen, &ec);
+            if (U_SUCCESS(ec)) {
+                result = UnicodeString(canon);
+            }
+        }
+        uprv_free(buf);
+        uprv_free(canon);
+    }
+    return result;
+#endif
+}
+
+Locale&
+LocaleUtility::initLocaleFromName(const UnicodeString& id, Locale& result)
+{
+    enum { BUFLEN = 128 }; // larger than ever needed
+
+    if (id.isBogus() || id.length() >= BUFLEN) {
+        result.setToBogus();
+    } else {
+        /*
+         * We need to convert from a UnicodeString to char * in order to
+         * create a Locale.
+         *
+         * Problem: Locale ID strings may contain '@' which is a variant
+         * character and cannot be handled by invariant-character conversion.
+         *
+         * Hack: Since ICU code can handle locale IDs with multiple encodings
+         * of '@' (at least for EBCDIC; it's not known to be a problem for
+         * ASCII-based systems),
+         * we use regular invariant-character conversion for everything else
+         * and manually convert U+0040 into a compiler-char-constant '@'.
+         * While this compilation-time constant may not match the runtime
+         * encoding of '@', it should be one of the encodings which ICU
+         * recognizes.
+         *
+         * There should be only at most one '@' in a locale ID.
+         */
+        char buffer[BUFLEN];
+        int32_t prev, i;
+        prev = 0;
+        for(;;) {
+            i = id.indexOf((UChar)0x40, prev);
+            if(i < 0) {
+                // no @ between prev and the rest of the string
+                id.extract(prev, INT32_MAX, buffer + prev, BUFLEN - prev, US_INV);
+                break; // done
+            } else {
+                // normal invariant-character conversion for text between @s
+                id.extract(prev, i - prev, buffer + prev, BUFLEN - prev, US_INV);
+                // manually "convert" U+0040 at id[i] into '@' at buffer[i]
+                buffer[i] = '@';
+                prev = i + 1;
+            }
+        }
+        result = Locale::createFromName(buffer);
+    }
+    return result;
+}
+
+UnicodeString&
+LocaleUtility::initNameFromLocale(const Locale& locale, UnicodeString& result)
+{
+    if (locale.isBogus()) {
+        result.setToBogus();
+    } else {
+        result.append(UnicodeString(locale.getName(), -1, US_INV));
+    }
+    return result;
+}
+
+const Hashtable*
+LocaleUtility::getAvailableLocaleNames(const UnicodeString& bundleID)
+{
+    // LocaleUtility_cache is a hash-of-hashes.  The top-level keys
+    // are path strings ('bundleID') passed to
+    // ures_openAvailableLocales.  The top-level values are
+    // second-level hashes.  The second-level keys are result strings
+    // from ures_openAvailableLocales.  The second-level values are
+    // garbage ((void*)1 or other random pointer).
+
+    UErrorCode status = U_ZERO_ERROR;
+    umtx_initOnce(LocaleUtilityInitOnce, locale_utility_init, status);
+    Hashtable *cache = LocaleUtility_cache;
+    if (cache == NULL) {
+        // Catastrophic failure.
+        return NULL;
+    }
+
+    Hashtable* htp;
+    umtx_lock(NULL);
+    htp = (Hashtable*) cache->get(bundleID);
+    umtx_unlock(NULL);
+
+    if (htp == NULL) {
+        htp = new Hashtable(status);
+        if (htp && U_SUCCESS(status)) {
+            CharString cbundleID;
+            cbundleID.appendInvariantChars(bundleID, status);
+            const char* path = cbundleID.isEmpty() ? NULL : cbundleID.data();
+            icu::LocalUEnumerationPointer uenum(ures_openAvailableLocales(path, &status));
+            for (;;) {
+                const UChar* id = uenum_unext(uenum.getAlias(), NULL, &status);
+                if (id == NULL) {
+                    break;
+                }
+                htp->put(UnicodeString(id), (void*)htp, status);
+            }
+            if (U_FAILURE(status)) {
+                delete htp;
+                return NULL;
+            }
+            umtx_lock(NULL);
+            Hashtable *t = static_cast<Hashtable *>(cache->get(bundleID));
+            if (t != NULL) {
+                // Another thread raced through this code, creating the cache entry first.
+                // Discard ours and return theirs.
+                umtx_unlock(NULL);
+                delete htp;
+                htp = t;
+            } else {
+                cache->put(bundleID, (void*)htp, status);
+                umtx_unlock(NULL);
+            }
+        }
+    }
+    return htp;
+}
+
+UBool
+LocaleUtility::isFallbackOf(const UnicodeString& root, const UnicodeString& child)
+{
+    return child.indexOf(root) == 0 &&
+      (child.length() == root.length() ||
+       child.charAt(root.length()) == UNDERSCORE_CHAR);
+}
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/locutil.h b/thirdparty/icu4c/common/locutil.h
new file mode 100644
index 0000000000..31bfffd7a5
--- /dev/null
+++ b/thirdparty/icu4c/common/locutil.h
@@ -0,0 +1,39 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2002-2005, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#ifndef LOCUTIL_H
+#define LOCUTIL_H
+
+#include "unicode/utypes.h"
+#include "hash.h"
+
+#if !UCONFIG_NO_SERVICE || !UCONFIG_NO_TRANSLITERATION
+
+
+U_NAMESPACE_BEGIN
+
+// temporary utility functions, till I know where to find them
+// in header so tests can also access them
+
+class U_COMMON_API LocaleUtility {
+public:
+  static UnicodeString& canonicalLocaleString(const UnicodeString* id, UnicodeString& result);
+  static Locale& initLocaleFromName(const UnicodeString& id, Locale& result);
+  static UnicodeString& initNameFromLocale(const Locale& locale, UnicodeString& result);
+  static const Hashtable* getAvailableLocaleNames(const UnicodeString& bundleID);
+  static UBool isFallbackOf(const UnicodeString& root, const UnicodeString& child);
+};
+
+U_NAMESPACE_END
+
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/lsr.cpp b/thirdparty/icu4c/common/lsr.cpp
new file mode 100644
index 0000000000..b81808f2c4
--- /dev/null
+++ b/thirdparty/icu4c/common/lsr.cpp
@@ -0,0 +1,114 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// lsr.cpp
+// created: 2019may08 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "lsr.h"
+#include "uinvchar.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+LSR::LSR(char prefix, const char *lang, const char *scr, const char *r, int32_t f,
+         UErrorCode &errorCode) :
+        language(nullptr), script(nullptr), region(r),
+        regionIndex(indexForRegion(region)), flags(f) {
+    if (U_SUCCESS(errorCode)) {
+        CharString langScript;
+        langScript.append(prefix, errorCode).append(lang, errorCode).append('\0', errorCode);
+        int32_t scriptOffset = langScript.length();
+        langScript.append(prefix, errorCode).append(scr, errorCode);
+        owned = langScript.cloneData(errorCode);
+        if (U_SUCCESS(errorCode)) {
+            language = owned;
+            script = owned + scriptOffset;
+        }
+    }
+}
+
+LSR::LSR(LSR &&other) U_NOEXCEPT :
+        language(other.language), script(other.script), region(other.region), owned(other.owned),
+        regionIndex(other.regionIndex), flags(other.flags),
+        hashCode(other.hashCode) {
+    if (owned != nullptr) {
+        other.language = other.script = "";
+        other.owned = nullptr;
+        other.hashCode = 0;
+    }
+}
+
+void LSR::deleteOwned() {
+    uprv_free(owned);
+}
+
+LSR &LSR::operator=(LSR &&other) U_NOEXCEPT {
+    this->~LSR();
+    language = other.language;
+    script = other.script;
+    region = other.region;
+    regionIndex = other.regionIndex;
+    flags = other.flags;
+    owned = other.owned;
+    hashCode = other.hashCode;
+    if (owned != nullptr) {
+        other.language = other.script = "";
+        other.owned = nullptr;
+        other.hashCode = 0;
+    }
+    return *this;
+}
+
+UBool LSR::isEquivalentTo(const LSR &other) const {
+    return
+        uprv_strcmp(language, other.language) == 0 &&
+        uprv_strcmp(script, other.script) == 0 &&
+        regionIndex == other.regionIndex &&
+        // Compare regions if both are ill-formed (and their indexes are 0).
+        (regionIndex > 0 || uprv_strcmp(region, other.region) == 0);
+}
+
+UBool LSR::operator==(const LSR &other) const {
+    return
+        uprv_strcmp(language, other.language) == 0 &&
+        uprv_strcmp(script, other.script) == 0 &&
+        regionIndex == other.regionIndex &&
+        // Compare regions if both are ill-formed (and their indexes are 0).
+        (regionIndex > 0 || uprv_strcmp(region, other.region) == 0) &&
+        flags == other.flags;
+}
+
+int32_t LSR::indexForRegion(const char *region) {
+    int32_t c = region[0];
+    int32_t a = c - '0';
+    if (0 <= a && a <= 9) {  // digits: "419"
+        int32_t b = region[1] - '0';
+        if (b < 0 || 9 < b) { return 0; }
+        c = region[2] - '0';
+        if (c < 0 || 9 < c || region[3] != 0) { return 0; }
+        return (10 * a + b) * 10 + c + 1;
+    } else {  // letters: "DE"
+        a = uprv_upperOrdinal(c);
+        if (a < 0 || 25 < a) { return 0; }
+        int32_t b = uprv_upperOrdinal(region[1]);
+        if (b < 0 || 25 < b || region[2] != 0) { return 0; }
+        return 26 * a + b + 1001;
+    }
+    return 0;
+}
+
+LSR &LSR::setHashCode() {
+    if (hashCode == 0) {
+        uint32_t h = ustr_hashCharsN(language, static_cast<int32_t>(uprv_strlen(language)));
+        h = h * 37 + ustr_hashCharsN(script, static_cast<int32_t>(uprv_strlen(script)));
+        h = h * 37 + regionIndex;
+        hashCode = h * 37 + flags;
+    }
+    return *this;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/lsr.h b/thirdparty/icu4c/common/lsr.h
new file mode 100644
index 0000000000..a33f855245
--- /dev/null
+++ b/thirdparty/icu4c/common/lsr.h
@@ -0,0 +1,82 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// lsr.h
+// created: 2019may08 Markus W. Scherer
+
+#ifndef __LSR_H__
+#define __LSR_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "cstring.h"
+
+U_NAMESPACE_BEGIN
+
+struct LSR final : public UMemory {
+    static constexpr int32_t REGION_INDEX_LIMIT = 1001 + 26 * 26;
+
+    static constexpr int32_t EXPLICIT_LSR = 7;
+    static constexpr int32_t EXPLICIT_LANGUAGE = 4;
+    static constexpr int32_t EXPLICIT_SCRIPT = 2;
+    static constexpr int32_t EXPLICIT_REGION = 1;
+    static constexpr int32_t IMPLICIT_LSR = 0;
+    static constexpr int32_t DONT_CARE_FLAGS = 0;
+
+    const char *language;
+    const char *script;
+    const char *region;
+    char *owned = nullptr;
+    /** Index for region, 0 if ill-formed. @see indexForRegion */
+    int32_t regionIndex = 0;
+    int32_t flags = 0;
+    /** Only set for LSRs that will be used in a hash table. */
+    int32_t hashCode = 0;
+
+    LSR() : language("und"), script(""), region("") {}
+
+    /** Constructor which aliases all subtag pointers. */
+    LSR(const char *lang, const char *scr, const char *r, int32_t f) :
+            language(lang),  script(scr), region(r),
+            regionIndex(indexForRegion(region)), flags(f) {}
+    /**
+     * Constructor which prepends the prefix to the language and script,
+     * copies those into owned memory, and aliases the region.
+     */
+    LSR(char prefix, const char *lang, const char *scr, const char *r, int32_t f,
+        UErrorCode &errorCode);
+    LSR(LSR &&other) U_NOEXCEPT;
+    LSR(const LSR &other) = delete;
+    inline ~LSR() {
+        // Pure inline code for almost all instances.
+        if (owned != nullptr) {
+            deleteOwned();
+        }
+    }
+
+    LSR &operator=(LSR &&other) U_NOEXCEPT;
+    LSR &operator=(const LSR &other) = delete;
+
+    /**
+     * Returns a positive index (>0) for a well-formed region code.
+     * Do not rely on a particular region->index mapping; it may change.
+     * Returns 0 for ill-formed strings.
+     */
+    static int32_t indexForRegion(const char *region);
+
+    UBool isEquivalentTo(const LSR &other) const;
+    UBool operator==(const LSR &other) const;
+
+    inline UBool operator!=(const LSR &other) const {
+        return !operator==(other);
+    }
+
+    LSR &setHashCode();
+
+private:
+    void deleteOwned();
+};
+
+U_NAMESPACE_END
+
+#endif  // __LSR_H__
diff --git a/thirdparty/icu4c/common/messageimpl.h b/thirdparty/icu4c/common/messageimpl.h
new file mode 100644
index 0000000000..a56479066b
--- /dev/null
+++ b/thirdparty/icu4c/common/messageimpl.h
@@ -0,0 +1,65 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  messageimpl.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011apr04
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __MESSAGEIMPL_H__
+#define __MESSAGEIMPL_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/messagepattern.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Helper functions for use of MessagePattern.
+ * In Java, these are package-private methods in MessagePattern itself.
+ * In C++, they are declared here and implemented in messagepattern.cpp.
+ */
+class U_COMMON_API MessageImpl {
+public:
+    /**
+     * @return true if getApostropheMode()==UMSGPAT_APOS_DOUBLE_REQUIRED
+     */
+    static UBool jdkAposMode(const MessagePattern &msgPattern) {
+        return msgPattern.getApostropheMode()==UMSGPAT_APOS_DOUBLE_REQUIRED;
+    }
+
+    /**
+     * Appends the s[start, limit[ substring to sb, but with only half of the apostrophes
+     * according to JDK pattern behavior.
+     */
+    static void appendReducedApostrophes(const UnicodeString &s, int32_t start, int32_t limit,
+                                         UnicodeString &sb);
+
+    /**
+     * Appends the sub-message to the result string.
+     * Omits SKIP_SYNTAX and appends whole arguments using appendReducedApostrophes().
+     */
+    static UnicodeString &appendSubMessageWithoutSkipSyntax(const MessagePattern &msgPattern,
+                                                            int32_t msgStart,
+                                                            UnicodeString &result);
+
+private:
+    MessageImpl();  // no constructor: all static methods
+};
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_FORMATTING
+
+#endif  // __MESSAGEIMPL_H__
diff --git a/thirdparty/icu4c/common/messagepattern.cpp b/thirdparty/icu4c/common/messagepattern.cpp
new file mode 100644
index 0000000000..f223d06711
--- /dev/null
+++ b/thirdparty/icu4c/common/messagepattern.cpp
@@ -0,0 +1,1233 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  messagepattern.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011mar14
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/messagepattern.h"
+#include "unicode/unistr.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "messageimpl.h"
+#include "patternprops.h"
+#include "putilimp.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+// Unicode character/code point constants ---------------------------------- ***
+
+static const UChar u_pound=0x23;
+static const UChar u_apos=0x27;
+static const UChar u_plus=0x2B;
+static const UChar u_comma=0x2C;
+static const UChar u_minus=0x2D;
+static const UChar u_dot=0x2E;
+static const UChar u_colon=0x3A;
+static const UChar u_lessThan=0x3C;
+static const UChar u_equal=0x3D;
+static const UChar u_A=0x41;
+static const UChar u_C=0x43;
+static const UChar u_D=0x44;
+static const UChar u_E=0x45;
+static const UChar u_H=0x48;
+static const UChar u_I=0x49;
+static const UChar u_L=0x4C;
+static const UChar u_N=0x4E;
+static const UChar u_O=0x4F;
+static const UChar u_P=0x50;
+static const UChar u_R=0x52;
+static const UChar u_S=0x53;
+static const UChar u_T=0x54;
+static const UChar u_U=0x55;
+static const UChar u_Z=0x5A;
+static const UChar u_a=0x61;
+static const UChar u_c=0x63;
+static const UChar u_d=0x64;
+static const UChar u_e=0x65;
+static const UChar u_f=0x66;
+static const UChar u_h=0x68;
+static const UChar u_i=0x69;
+static const UChar u_l=0x6C;
+static const UChar u_n=0x6E;
+static const UChar u_o=0x6F;
+static const UChar u_p=0x70;
+static const UChar u_r=0x72;
+static const UChar u_s=0x73;
+static const UChar u_t=0x74;
+static const UChar u_u=0x75;
+static const UChar u_z=0x7A;
+static const UChar u_leftCurlyBrace=0x7B;
+static const UChar u_pipe=0x7C;
+static const UChar u_rightCurlyBrace=0x7D;
+static const UChar u_lessOrEqual=0x2264;  // U+2264 is <=
+
+static const UChar kOffsetColon[]={  // "offset:"
+    u_o, u_f, u_f, u_s, u_e, u_t, u_colon
+};
+
+static const UChar kOther[]={  // "other"
+    u_o, u_t, u_h, u_e, u_r
+};
+
+// MessagePatternList ------------------------------------------------------ ***
+
+template<typename T, int32_t stackCapacity>
+class MessagePatternList : public UMemory {
+public:
+    MessagePatternList() {}
+    void copyFrom(const MessagePatternList<T, stackCapacity> &other,
+                  int32_t length,
+                  UErrorCode &errorCode);
+    UBool ensureCapacityForOneMore(int32_t oldLength, UErrorCode &errorCode);
+    UBool equals(const MessagePatternList<T, stackCapacity> &other, int32_t length) const {
+        for(int32_t i=0; i<length; ++i) {
+            if(a[i]!=other.a[i]) { return FALSE; }
+        }
+        return TRUE;
+    }
+
+    MaybeStackArray<T, stackCapacity> a;
+};
+
+template<typename T, int32_t stackCapacity>
+void
+MessagePatternList<T, stackCapacity>::copyFrom(
+        const MessagePatternList<T, stackCapacity> &other,
+        int32_t length,
+        UErrorCode &errorCode) {
+    if(U_SUCCESS(errorCode) && length>0) {
+        if(length>a.getCapacity() && NULL==a.resize(length)) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        uprv_memcpy(a.getAlias(), other.a.getAlias(), (size_t)length*sizeof(T));
+    }
+}
+
+template<typename T, int32_t stackCapacity>
+UBool
+MessagePatternList<T, stackCapacity>::ensureCapacityForOneMore(int32_t oldLength, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    if(a.getCapacity()>oldLength || a.resize(2*oldLength, oldLength)!=NULL) {
+        return TRUE;
+    }
+    errorCode=U_MEMORY_ALLOCATION_ERROR;
+    return FALSE;
+}
+
+// MessagePatternList specializations -------------------------------------- ***
+
+class MessagePatternDoubleList : public MessagePatternList<double, 8> {
+};
+
+class MessagePatternPartsList : public MessagePatternList<MessagePattern::Part, 32> {
+};
+
+// MessagePattern constructors etc. ---------------------------------------- ***
+
+MessagePattern::MessagePattern(UErrorCode &errorCode)
+        : aposMode(UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE),
+          partsList(NULL), parts(NULL), partsLength(0),
+          numericValuesList(NULL), numericValues(NULL), numericValuesLength(0),
+          hasArgNames(FALSE), hasArgNumbers(FALSE), needsAutoQuoting(FALSE) {
+    init(errorCode);
+}
+
+MessagePattern::MessagePattern(UMessagePatternApostropheMode mode, UErrorCode &errorCode)
+        : aposMode(mode),
+          partsList(NULL), parts(NULL), partsLength(0),
+          numericValuesList(NULL), numericValues(NULL), numericValuesLength(0),
+          hasArgNames(FALSE), hasArgNumbers(FALSE), needsAutoQuoting(FALSE) {
+    init(errorCode);
+}
+
+MessagePattern::MessagePattern(const UnicodeString &pattern, UParseError *parseError, UErrorCode &errorCode)
+        : aposMode(UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE),
+          partsList(NULL), parts(NULL), partsLength(0),
+          numericValuesList(NULL), numericValues(NULL), numericValuesLength(0),
+          hasArgNames(FALSE), hasArgNumbers(FALSE), needsAutoQuoting(FALSE) {
+    if(init(errorCode)) {
+        parse(pattern, parseError, errorCode);
+    }
+}
+
+UBool
+MessagePattern::init(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    partsList=new MessagePatternPartsList();
+    if(partsList==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    parts=partsList->a.getAlias();
+    return TRUE;
+}
+
+MessagePattern::MessagePattern(const MessagePattern &other)
+        : UObject(other), aposMode(other.aposMode), msg(other.msg),
+          partsList(NULL), parts(NULL), partsLength(0),
+          numericValuesList(NULL), numericValues(NULL), numericValuesLength(0),
+          hasArgNames(other.hasArgNames), hasArgNumbers(other.hasArgNumbers),
+          needsAutoQuoting(other.needsAutoQuoting) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    if(!copyStorage(other, errorCode)) {
+        clear();
+    }
+}
+
+MessagePattern &
+MessagePattern::operator=(const MessagePattern &other) {
+    if(this==&other) {
+        return *this;
+    }
+    aposMode=other.aposMode;
+    msg=other.msg;
+    hasArgNames=other.hasArgNames;
+    hasArgNumbers=other.hasArgNumbers;
+    needsAutoQuoting=other.needsAutoQuoting;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    if(!copyStorage(other, errorCode)) {
+        clear();
+    }
+    return *this;
+}
+
+UBool
+MessagePattern::copyStorage(const MessagePattern &other, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    parts=NULL;
+    partsLength=0;
+    numericValues=NULL;
+    numericValuesLength=0;
+    if(partsList==NULL) {
+        partsList=new MessagePatternPartsList();
+        if(partsList==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return FALSE;
+        }
+        parts=partsList->a.getAlias();
+    }
+    if(other.partsLength>0) {
+        partsList->copyFrom(*other.partsList, other.partsLength, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return FALSE;
+        }
+        parts=partsList->a.getAlias();
+        partsLength=other.partsLength;
+    }
+    if(other.numericValuesLength>0) {
+        if(numericValuesList==NULL) {
+            numericValuesList=new MessagePatternDoubleList();
+            if(numericValuesList==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+                return FALSE;
+            }
+            numericValues=numericValuesList->a.getAlias();
+        }
+        numericValuesList->copyFrom(
+            *other.numericValuesList, other.numericValuesLength, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return FALSE;
+        }
+        numericValues=numericValuesList->a.getAlias();
+        numericValuesLength=other.numericValuesLength;
+    }
+    return TRUE;
+}
+
+MessagePattern::~MessagePattern() {
+    delete partsList;
+    delete numericValuesList;
+}
+
+// MessagePattern API ------------------------------------------------------ ***
+
+MessagePattern &
+MessagePattern::parse(const UnicodeString &pattern, UParseError *parseError, UErrorCode &errorCode) {
+    preParse(pattern, parseError, errorCode);
+    parseMessage(0, 0, 0, UMSGPAT_ARG_TYPE_NONE, parseError, errorCode);
+    postParse();
+    return *this;
+}
+
+MessagePattern &
+MessagePattern::parseChoiceStyle(const UnicodeString &pattern,
+                                 UParseError *parseError, UErrorCode &errorCode) {
+    preParse(pattern, parseError, errorCode);
+    parseChoiceStyle(0, 0, parseError, errorCode);
+    postParse();
+    return *this;
+}
+
+MessagePattern &
+MessagePattern::parsePluralStyle(const UnicodeString &pattern,
+                                 UParseError *parseError, UErrorCode &errorCode) {
+    preParse(pattern, parseError, errorCode);
+    parsePluralOrSelectStyle(UMSGPAT_ARG_TYPE_PLURAL, 0, 0, parseError, errorCode);
+    postParse();
+    return *this;
+}
+
+MessagePattern &
+MessagePattern::parseSelectStyle(const UnicodeString &pattern,
+                                 UParseError *parseError, UErrorCode &errorCode) {
+    preParse(pattern, parseError, errorCode);
+    parsePluralOrSelectStyle(UMSGPAT_ARG_TYPE_SELECT, 0, 0, parseError, errorCode);
+    postParse();
+    return *this;
+}
+
+void
+MessagePattern::clear() {
+    // Mostly the same as preParse().
+    msg.remove();
+    hasArgNames=hasArgNumbers=FALSE;
+    needsAutoQuoting=FALSE;
+    partsLength=0;
+    numericValuesLength=0;
+}
+
+UBool
+MessagePattern::operator==(const MessagePattern &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    return
+        aposMode==other.aposMode &&
+        msg==other.msg &&
+        // parts.equals(o.parts)
+        partsLength==other.partsLength &&
+        (partsLength==0 || partsList->equals(*other.partsList, partsLength));
+    // No need to compare numericValues if msg and parts are the same.
+}
+
+int32_t
+MessagePattern::hashCode() const {
+    int32_t hash=(aposMode*37+msg.hashCode())*37+partsLength;
+    for(int32_t i=0; i<partsLength; ++i) {
+        hash=hash*37+parts[i].hashCode();
+    }
+    return hash;
+}
+
+int32_t
+MessagePattern::validateArgumentName(const UnicodeString &name) {
+    if(!PatternProps::isIdentifier(name.getBuffer(), name.length())) {
+        return UMSGPAT_ARG_NAME_NOT_VALID;
+    }
+    return parseArgNumber(name, 0, name.length());
+}
+
+UnicodeString
+MessagePattern::autoQuoteApostropheDeep() const {
+    if(!needsAutoQuoting) {
+        return msg;
+    }
+    UnicodeString modified(msg);
+    // Iterate backward so that the insertion indexes do not change.
+    int32_t count=countParts();
+    for(int32_t i=count; i>0;) {
+        const Part &part=getPart(--i);
+        if(part.getType()==UMSGPAT_PART_TYPE_INSERT_CHAR) {
+           modified.insert(part.index, (UChar)part.value);
+        }
+    }
+    return modified;
+}
+
+double
+MessagePattern::getNumericValue(const Part &part) const {
+    UMessagePatternPartType type=part.type;
+    if(type==UMSGPAT_PART_TYPE_ARG_INT) {
+        return part.value;
+    } else if(type==UMSGPAT_PART_TYPE_ARG_DOUBLE) {
+        return numericValues[part.value];
+    } else {
+        return UMSGPAT_NO_NUMERIC_VALUE;
+    }
+}
+
+/**
+  * Returns the "offset:" value of a PluralFormat argument, or 0 if none is specified.
+  * @param pluralStart the index of the first PluralFormat argument style part. (0..countParts()-1)
+  * @return the "offset:" value.
+  * @draft ICU 4.8
+  */
+double
+MessagePattern::getPluralOffset(int32_t pluralStart) const {
+    const Part &part=getPart(pluralStart);
+    if(Part::hasNumericValue(part.type)) {
+        return getNumericValue(part);
+    } else {
+        return 0;
+    }
+}
+
+// MessagePattern::Part ---------------------------------------------------- ***
+
+UBool
+MessagePattern::Part::operator==(const Part &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    return
+        type==other.type &&
+        index==other.index &&
+        length==other.length &&
+        value==other.value &&
+        limitPartIndex==other.limitPartIndex;
+}
+
+// MessagePattern parser --------------------------------------------------- ***
+
+void
+MessagePattern::preParse(const UnicodeString &pattern, UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    if(parseError!=NULL) {
+        parseError->line=0;
+        parseError->offset=0;
+        parseError->preContext[0]=0;
+        parseError->postContext[0]=0;
+    }
+    msg=pattern;
+    hasArgNames=hasArgNumbers=FALSE;
+    needsAutoQuoting=FALSE;
+    partsLength=0;
+    numericValuesLength=0;
+}
+
+void
+MessagePattern::postParse() {
+    if(partsList!=NULL) {
+        parts=partsList->a.getAlias();
+    }
+    if(numericValuesList!=NULL) {
+        numericValues=numericValuesList->a.getAlias();
+    }
+}
+
+int32_t
+MessagePattern::parseMessage(int32_t index, int32_t msgStartLength,
+                             int32_t nestingLevel, UMessagePatternArgType parentType,
+                             UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(nestingLevel>Part::MAX_VALUE) {
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    int32_t msgStart=partsLength;
+    addPart(UMSGPAT_PART_TYPE_MSG_START, index, msgStartLength, nestingLevel, errorCode);
+    index+=msgStartLength;
+    for(;;) {  // while(index<msg.length()) with U_FAILURE(errorCode) check
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+        if(index>=msg.length()) {
+            break;
+        }
+        UChar c=msg.charAt(index++);
+        if(c==u_apos) {
+            if(index==msg.length()) {
+                // The apostrophe is the last character in the pattern. 
+                // Add a Part for auto-quoting.
+                addPart(UMSGPAT_PART_TYPE_INSERT_CHAR, index, 0,
+                        u_apos, errorCode);  // value=char to be inserted
+                needsAutoQuoting=TRUE;
+            } else {
+                c=msg.charAt(index);
+                if(c==u_apos) {
+                    // double apostrophe, skip the second one
+                    addPart(UMSGPAT_PART_TYPE_SKIP_SYNTAX, index++, 1, 0, errorCode);
+                } else if(
+                    aposMode==UMSGPAT_APOS_DOUBLE_REQUIRED ||
+                    c==u_leftCurlyBrace || c==u_rightCurlyBrace ||
+                    (parentType==UMSGPAT_ARG_TYPE_CHOICE && c==u_pipe) ||
+                    (UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(parentType) && c==u_pound)
+                ) {
+                    // skip the quote-starting apostrophe
+                    addPart(UMSGPAT_PART_TYPE_SKIP_SYNTAX, index-1, 1, 0, errorCode);
+                    // find the end of the quoted literal text
+                    for(;;) {
+                        index=msg.indexOf(u_apos, index+1);
+                        if(index>=0) {
+                            if(/*(index+1)<msg.length() &&*/ msg.charAt(index+1)==u_apos) {
+                                // double apostrophe inside quoted literal text
+                                // still encodes a single apostrophe, skip the second one
+                                addPart(UMSGPAT_PART_TYPE_SKIP_SYNTAX, ++index, 1, 0, errorCode);
+                            } else {
+                                // skip the quote-ending apostrophe
+                                addPart(UMSGPAT_PART_TYPE_SKIP_SYNTAX, index++, 1, 0, errorCode);
+                                break;
+                            }
+                        } else {
+                            // The quoted text reaches to the end of the of the message.
+                            index=msg.length();
+                            // Add a Part for auto-quoting.
+                            addPart(UMSGPAT_PART_TYPE_INSERT_CHAR, index, 0,
+                                    u_apos, errorCode);  // value=char to be inserted
+                            needsAutoQuoting=TRUE;
+                            break;
+                        }
+                    }
+                } else {
+                    // Interpret the apostrophe as literal text.
+                    // Add a Part for auto-quoting.
+                    addPart(UMSGPAT_PART_TYPE_INSERT_CHAR, index, 0,
+                            u_apos, errorCode);  // value=char to be inserted
+                    needsAutoQuoting=TRUE;
+                }
+            }
+        } else if(UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(parentType) && c==u_pound) {
+            // The unquoted # in a plural message fragment will be replaced
+            // with the (number-offset).
+            addPart(UMSGPAT_PART_TYPE_REPLACE_NUMBER, index-1, 1, 0, errorCode);
+        } else if(c==u_leftCurlyBrace) {
+            index=parseArg(index-1, 1, nestingLevel, parseError, errorCode);
+        } else if((nestingLevel>0 && c==u_rightCurlyBrace) ||
+                  (parentType==UMSGPAT_ARG_TYPE_CHOICE && c==u_pipe)) {
+            // Finish the message before the terminator.
+            // In a choice style, report the "}" substring only for the following ARG_LIMIT,
+            // not for this MSG_LIMIT.
+            int32_t limitLength=(parentType==UMSGPAT_ARG_TYPE_CHOICE && c==u_rightCurlyBrace) ? 0 : 1;
+            addLimitPart(msgStart, UMSGPAT_PART_TYPE_MSG_LIMIT, index-1, limitLength,
+                         nestingLevel, errorCode);
+            if(parentType==UMSGPAT_ARG_TYPE_CHOICE) {
+                // Let the choice style parser see the '}' or '|'.
+                return index-1;
+            } else {
+                // continue parsing after the '}'
+                return index;
+            }
+        }  // else: c is part of literal text
+    }
+    if(nestingLevel>0 && !inTopLevelChoiceMessage(nestingLevel, parentType)) {
+        setParseError(parseError, 0);  // Unmatched '{' braces in message.
+        errorCode=U_UNMATCHED_BRACES;
+        return 0;
+    }
+    addLimitPart(msgStart, UMSGPAT_PART_TYPE_MSG_LIMIT, index, 0, nestingLevel, errorCode);
+    return index;
+}
+
+int32_t
+MessagePattern::parseArg(int32_t index, int32_t argStartLength, int32_t nestingLevel,
+                         UParseError *parseError, UErrorCode &errorCode) {
+    int32_t argStart=partsLength;
+    UMessagePatternArgType argType=UMSGPAT_ARG_TYPE_NONE;
+    addPart(UMSGPAT_PART_TYPE_ARG_START, index, argStartLength, argType, errorCode);
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    int32_t nameIndex=index=skipWhiteSpace(index+argStartLength);
+    if(index==msg.length()) {
+        setParseError(parseError, 0);  // Unmatched '{' braces in message.
+        errorCode=U_UNMATCHED_BRACES;
+        return 0;
+    }
+    // parse argument name or number
+    index=skipIdentifier(index);
+    int32_t number=parseArgNumber(nameIndex, index);
+    if(number>=0) {
+        int32_t length=index-nameIndex;
+        if(length>Part::MAX_LENGTH || number>Part::MAX_VALUE) {
+            setParseError(parseError, nameIndex);  // Argument number too large.
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+        hasArgNumbers=TRUE;
+        addPart(UMSGPAT_PART_TYPE_ARG_NUMBER, nameIndex, length, number, errorCode);
+    } else if(number==UMSGPAT_ARG_NAME_NOT_NUMBER) {
+        int32_t length=index-nameIndex;
+        if(length>Part::MAX_LENGTH) {
+            setParseError(parseError, nameIndex);  // Argument name too long.
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+        hasArgNames=TRUE;
+        addPart(UMSGPAT_PART_TYPE_ARG_NAME, nameIndex, length, 0, errorCode);
+    } else {  // number<-1 (ARG_NAME_NOT_VALID)
+        setParseError(parseError, nameIndex);  // Bad argument syntax.
+        errorCode=U_PATTERN_SYNTAX_ERROR;
+        return 0;
+    }
+    index=skipWhiteSpace(index);
+    if(index==msg.length()) {
+        setParseError(parseError, 0);  // Unmatched '{' braces in message.
+        errorCode=U_UNMATCHED_BRACES;
+        return 0;
+    }
+    UChar c=msg.charAt(index);
+    if(c==u_rightCurlyBrace) {
+        // all done
+    } else if(c!=u_comma) {
+        setParseError(parseError, nameIndex);  // Bad argument syntax.
+        errorCode=U_PATTERN_SYNTAX_ERROR;
+        return 0;
+    } else /* ',' */ {
+        // parse argument type: case-sensitive a-zA-Z
+        int32_t typeIndex=index=skipWhiteSpace(index+1);
+        while(index<msg.length() && isArgTypeChar(msg.charAt(index))) {
+            ++index;
+        }
+        int32_t length=index-typeIndex;
+        index=skipWhiteSpace(index);
+        if(index==msg.length()) {
+            setParseError(parseError, 0);  // Unmatched '{' braces in message.
+            errorCode=U_UNMATCHED_BRACES;
+            return 0;
+        }
+        if(length==0 || ((c=msg.charAt(index))!=u_comma && c!=u_rightCurlyBrace)) {
+            setParseError(parseError, nameIndex);  // Bad argument syntax.
+            errorCode=U_PATTERN_SYNTAX_ERROR;
+            return 0;
+        }
+        if(length>Part::MAX_LENGTH) {
+            setParseError(parseError, nameIndex);  // Argument type name too long.
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+        argType=UMSGPAT_ARG_TYPE_SIMPLE;
+        if(length==6) {
+            // case-insensitive comparisons for complex-type names
+            if(isChoice(typeIndex)) {
+                argType=UMSGPAT_ARG_TYPE_CHOICE;
+            } else if(isPlural(typeIndex)) {
+                argType=UMSGPAT_ARG_TYPE_PLURAL;
+            } else if(isSelect(typeIndex)) {
+                argType=UMSGPAT_ARG_TYPE_SELECT;
+            }
+        } else if(length==13) {
+            if(isSelect(typeIndex) && isOrdinal(typeIndex+6)) {
+                argType=UMSGPAT_ARG_TYPE_SELECTORDINAL;
+            }
+        }
+        // change the ARG_START type from NONE to argType
+        partsList->a[argStart].value=(int16_t)argType;
+        if(argType==UMSGPAT_ARG_TYPE_SIMPLE) {
+            addPart(UMSGPAT_PART_TYPE_ARG_TYPE, typeIndex, length, 0, errorCode);
+        }
+        // look for an argument style (pattern)
+        if(c==u_rightCurlyBrace) {
+            if(argType!=UMSGPAT_ARG_TYPE_SIMPLE) {
+                setParseError(parseError, nameIndex);  // No style field for complex argument.
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+        } else /* ',' */ {
+            ++index;
+            if(argType==UMSGPAT_ARG_TYPE_SIMPLE) {
+                index=parseSimpleStyle(index, parseError, errorCode);
+            } else if(argType==UMSGPAT_ARG_TYPE_CHOICE) {
+                index=parseChoiceStyle(index, nestingLevel, parseError, errorCode);
+            } else {
+                index=parsePluralOrSelectStyle(argType, index, nestingLevel, parseError, errorCode);
+            }
+        }
+    }
+    // Argument parsing stopped on the '}'.
+    addLimitPart(argStart, UMSGPAT_PART_TYPE_ARG_LIMIT, index, 1, argType, errorCode);
+    return index+1;
+}
+
+int32_t
+MessagePattern::parseSimpleStyle(int32_t index, UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    int32_t start=index;
+    int32_t nestedBraces=0;
+    while(index<msg.length()) {
+        UChar c=msg.charAt(index++);
+        if(c==u_apos) {
+            // Treat apostrophe as quoting but include it in the style part.
+            // Find the end of the quoted literal text.
+            index=msg.indexOf(u_apos, index);
+            if(index<0) {
+                // Quoted literal argument style text reaches to the end of the message.
+                setParseError(parseError, start);
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+            // skip the quote-ending apostrophe
+            ++index;
+        } else if(c==u_leftCurlyBrace) {
+            ++nestedBraces;
+        } else if(c==u_rightCurlyBrace) {
+            if(nestedBraces>0) {
+                --nestedBraces;
+            } else {
+                int32_t length=--index-start;
+                if(length>Part::MAX_LENGTH) {
+                    setParseError(parseError, start);  // Argument style text too long.
+                    errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                    return 0;
+                }
+                addPart(UMSGPAT_PART_TYPE_ARG_STYLE, start, length, 0, errorCode);
+                return index;
+            }
+        }  // c is part of literal text
+    }
+    setParseError(parseError, 0);  // Unmatched '{' braces in message.
+    errorCode=U_UNMATCHED_BRACES;
+    return 0;
+}
+
+int32_t
+MessagePattern::parseChoiceStyle(int32_t index, int32_t nestingLevel,
+                                 UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    int32_t start=index;
+    index=skipWhiteSpace(index);
+    if(index==msg.length() || msg.charAt(index)==u_rightCurlyBrace) {
+        setParseError(parseError, 0);  // Missing choice argument pattern.
+        errorCode=U_PATTERN_SYNTAX_ERROR;
+        return 0;
+    }
+    for(;;) {
+        // The choice argument style contains |-separated (number, separator, message) triples.
+        // Parse the number.
+        int32_t numberIndex=index;
+        index=skipDouble(index);
+        int32_t length=index-numberIndex;
+        if(length==0) {
+            setParseError(parseError, start);  // Bad choice pattern syntax.
+            errorCode=U_PATTERN_SYNTAX_ERROR;
+            return 0;
+        }
+        if(length>Part::MAX_LENGTH) {
+            setParseError(parseError, numberIndex);  // Choice number too long.
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+        parseDouble(numberIndex, index, TRUE, parseError, errorCode);  // adds ARG_INT or ARG_DOUBLE
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+        // Parse the separator.
+        index=skipWhiteSpace(index);
+        if(index==msg.length()) {
+            setParseError(parseError, start);  // Bad choice pattern syntax.
+            errorCode=U_PATTERN_SYNTAX_ERROR;
+            return 0;
+        }
+        UChar c=msg.charAt(index);
+        if(!(c==u_pound || c==u_lessThan || c==u_lessOrEqual)) {  // U+2264 is <=
+            setParseError(parseError, start);  // Expected choice separator (#<\u2264) instead of c.
+            errorCode=U_PATTERN_SYNTAX_ERROR;
+            return 0;
+        }
+        addPart(UMSGPAT_PART_TYPE_ARG_SELECTOR, index, 1, 0, errorCode);
+        // Parse the message fragment.
+        index=parseMessage(++index, 0, nestingLevel+1, UMSGPAT_ARG_TYPE_CHOICE, parseError, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+        // parseMessage(..., CHOICE) returns the index of the terminator, or msg.length().
+        if(index==msg.length()) {
+            return index;
+        }
+        if(msg.charAt(index)==u_rightCurlyBrace) {
+            if(!inMessageFormatPattern(nestingLevel)) {
+                setParseError(parseError, start);  // Bad choice pattern syntax.
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+            return index;
+        }  // else the terminator is '|'
+        index=skipWhiteSpace(index+1);
+    }
+}
+
+int32_t
+MessagePattern::parsePluralOrSelectStyle(UMessagePatternArgType argType,
+                                         int32_t index, int32_t nestingLevel,
+                                         UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    int32_t start=index;
+    UBool isEmpty=TRUE;
+    UBool hasOther=FALSE;
+    for(;;) {
+        // First, collect the selector looking for a small set of terminators.
+        // It would be a little faster to consider the syntax of each possible
+        // token right here, but that makes the code too complicated.
+        index=skipWhiteSpace(index);
+        UBool eos=index==msg.length();
+        if(eos || msg.charAt(index)==u_rightCurlyBrace) {
+            if(eos==inMessageFormatPattern(nestingLevel)) {
+                setParseError(parseError, start);  // Bad plural/select pattern syntax.
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+            if(!hasOther) {
+                setParseError(parseError, 0);  // Missing 'other' keyword in plural/select pattern.
+                errorCode=U_DEFAULT_KEYWORD_MISSING;
+                return 0;
+            }
+            return index;
+        }
+        int32_t selectorIndex=index;
+        if(UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(argType) && msg.charAt(selectorIndex)==u_equal) {
+            // explicit-value plural selector: =double
+            index=skipDouble(index+1);
+            int32_t length=index-selectorIndex;
+            if(length==1) {
+                setParseError(parseError, start);  // Bad plural/select pattern syntax.
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+            if(length>Part::MAX_LENGTH) {
+                setParseError(parseError, selectorIndex);  // Argument selector too long.
+                errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+            addPart(UMSGPAT_PART_TYPE_ARG_SELECTOR, selectorIndex, length, 0, errorCode);
+            parseDouble(selectorIndex+1, index, FALSE,
+                        parseError, errorCode);  // adds ARG_INT or ARG_DOUBLE
+        } else {
+            index=skipIdentifier(index);
+            int32_t length=index-selectorIndex;
+            if(length==0) {
+                setParseError(parseError, start);  // Bad plural/select pattern syntax.
+                errorCode=U_PATTERN_SYNTAX_ERROR;
+                return 0;
+            }
+            // Note: The ':' in "offset:" is just beyond the skipIdentifier() range.
+            if( UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(argType) && length==6 && index<msg.length() &&
+                0==msg.compare(selectorIndex, 7, kOffsetColon, 0, 7)
+            ) {
+                // plural offset, not a selector
+                if(!isEmpty) {
+                    // Plural argument 'offset:' (if present) must precede key-message pairs.
+                    setParseError(parseError, start);
+                    errorCode=U_PATTERN_SYNTAX_ERROR;
+                    return 0;
+                }
+                // allow whitespace between offset: and its value
+                int32_t valueIndex=skipWhiteSpace(index+1);  // The ':' is at index.
+                index=skipDouble(valueIndex);
+                if(index==valueIndex) {
+                    setParseError(parseError, start);  // Missing value for plural 'offset:'.
+                    errorCode=U_PATTERN_SYNTAX_ERROR;
+                    return 0;
+                }
+                if((index-valueIndex)>Part::MAX_LENGTH) {
+                    setParseError(parseError, valueIndex);  // Plural offset value too long.
+                    errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                    return 0;
+                }
+                parseDouble(valueIndex, index, FALSE,
+                            parseError, errorCode);  // adds ARG_INT or ARG_DOUBLE
+                if(U_FAILURE(errorCode)) {
+                    return 0;
+                }
+                isEmpty=FALSE;
+                continue;  // no message fragment after the offset
+            } else {
+                // normal selector word
+                if(length>Part::MAX_LENGTH) {
+                    setParseError(parseError, selectorIndex);  // Argument selector too long.
+                    errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                    return 0;
+                }
+                addPart(UMSGPAT_PART_TYPE_ARG_SELECTOR, selectorIndex, length, 0, errorCode);
+                if(0==msg.compare(selectorIndex, length, kOther, 0, 5)) {
+                    hasOther=TRUE;
+                }
+            }
+        }
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+
+        // parse the message fragment following the selector
+        index=skipWhiteSpace(index);
+        if(index==msg.length() || msg.charAt(index)!=u_leftCurlyBrace) {
+            setParseError(parseError, selectorIndex);  // No message fragment after plural/select selector.
+            errorCode=U_PATTERN_SYNTAX_ERROR;
+            return 0;
+        }
+        index=parseMessage(index, 1, nestingLevel+1, argType, parseError, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+        isEmpty=FALSE;
+    }
+}
+
+int32_t
+MessagePattern::parseArgNumber(const UnicodeString &s, int32_t start, int32_t limit) {
+    // If the identifier contains only ASCII digits, then it is an argument _number_
+    // and must not have leading zeros (except "0" itself).
+    // Otherwise it is an argument _name_.
+    if(start>=limit) {
+        return UMSGPAT_ARG_NAME_NOT_VALID;
+    }
+    int32_t number;
+    // Defer numeric errors until we know there are only digits.
+    UBool badNumber;
+    UChar c=s.charAt(start++);
+    if(c==0x30) {
+        if(start==limit) {
+            return 0;
+        } else {
+            number=0;
+            badNumber=TRUE;  // leading zero
+        }
+    } else if(0x31<=c && c<=0x39) {
+        number=c-0x30;
+        badNumber=FALSE;
+    } else {
+        return UMSGPAT_ARG_NAME_NOT_NUMBER;
+    }
+    while(start<limit) {
+        c=s.charAt(start++);
+        if(0x30<=c && c<=0x39) {
+            if(number>=INT32_MAX/10) {
+                badNumber=TRUE;  // overflow
+            }
+            number=number*10+(c-0x30);
+        } else {
+            return UMSGPAT_ARG_NAME_NOT_NUMBER;
+        }
+    }
+    // There are only ASCII digits.
+    if(badNumber) {
+        return UMSGPAT_ARG_NAME_NOT_VALID;
+    } else {
+        return number;
+    }
+}
+
+void
+MessagePattern::parseDouble(int32_t start, int32_t limit, UBool allowInfinity,
+                            UParseError *parseError, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    U_ASSERT(start<limit);
+    // fake loop for easy exit and single throw statement
+    for(;;) { /*loop doesn't iterate*/
+        // fast path for small integers and infinity
+        int32_t value=0;
+        int32_t isNegative=0;  // not boolean so that we can easily add it to value
+        int32_t index=start;
+        UChar c=msg.charAt(index++);
+        if(c==u_minus) {
+            isNegative=1;
+            if(index==limit) {
+                break;  // no number
+            }
+            c=msg.charAt(index++);
+        } else if(c==u_plus) {
+            if(index==limit) {
+                break;  // no number
+            }
+            c=msg.charAt(index++);
+        }
+        if(c==0x221e) {  // infinity
+            if(allowInfinity && index==limit) {
+                double infinity=uprv_getInfinity();
+                addArgDoublePart(
+                    isNegative!=0 ? -infinity : infinity,
+                    start, limit-start, errorCode);
+                return;
+            } else {
+                break;
+            }
+        }
+        // try to parse the number as a small integer but fall back to a double
+        while('0'<=c && c<='9') {
+            value=value*10+(c-'0');
+            if(value>(Part::MAX_VALUE+isNegative)) {
+                break;  // not a small-enough integer
+            }
+            if(index==limit) {
+                addPart(UMSGPAT_PART_TYPE_ARG_INT, start, limit-start,
+                        isNegative!=0 ? -value : value, errorCode);
+                return;
+            }
+            c=msg.charAt(index++);
+        }
+        // Let Double.parseDouble() throw a NumberFormatException.
+        char numberChars[128];
+        int32_t capacity=(int32_t)sizeof(numberChars);
+        int32_t length=limit-start;
+        if(length>=capacity) {
+            break;  // number too long
+        }
+        msg.extract(start, length, numberChars, capacity, US_INV);
+        if((int32_t)uprv_strlen(numberChars)<length) {
+            break;  // contains non-invariant character that was turned into NUL
+        }
+        char *end;
+        double numericValue=uprv_strtod(numberChars, &end);
+        if(end!=(numberChars+length)) {
+            break;  // parsing error
+        }
+        addArgDoublePart(numericValue, start, length, errorCode);
+        return;
+    }
+    setParseError(parseError, start /*, limit*/);  // Bad syntax for numeric value.
+    errorCode=U_PATTERN_SYNTAX_ERROR;
+    return;
+}
+
+int32_t
+MessagePattern::skipWhiteSpace(int32_t index) {
+    const UChar *s=msg.getBuffer();
+    int32_t msgLength=msg.length();
+    const UChar *t=PatternProps::skipWhiteSpace(s+index, msgLength-index);
+    return (int32_t)(t-s);
+}
+
+int32_t
+MessagePattern::skipIdentifier(int32_t index) {
+    const UChar *s=msg.getBuffer();
+    int32_t msgLength=msg.length();
+    const UChar *t=PatternProps::skipIdentifier(s+index, msgLength-index);
+    return (int32_t)(t-s);
+}
+
+int32_t
+MessagePattern::skipDouble(int32_t index) {
+    int32_t msgLength=msg.length();
+    while(index<msgLength) {
+        UChar c=msg.charAt(index);
+        // U+221E: Allow the infinity symbol, for ChoiceFormat patterns.
+        if((c<0x30 && c!=u_plus && c!=u_minus && c!=u_dot) || (c>0x39 && c!=u_e && c!=u_E && c!=0x221e)) {
+            break;
+        }
+        ++index;
+    }
+    return index;
+}
+
+UBool
+MessagePattern::isArgTypeChar(UChar32 c) {
+    return (u_a<=c && c<=u_z) || (u_A<=c && c<=u_Z);
+}
+
+UBool
+MessagePattern::isChoice(int32_t index) {
+    UChar c;
+    return
+        ((c=msg.charAt(index++))==u_c || c==u_C) &&
+        ((c=msg.charAt(index++))==u_h || c==u_H) &&
+        ((c=msg.charAt(index++))==u_o || c==u_O) &&
+        ((c=msg.charAt(index++))==u_i || c==u_I) &&
+        ((c=msg.charAt(index++))==u_c || c==u_C) &&
+        ((c=msg.charAt(index))==u_e || c==u_E);
+}
+
+UBool
+MessagePattern::isPlural(int32_t index) {
+    UChar c;
+    return
+        ((c=msg.charAt(index++))==u_p || c==u_P) &&
+        ((c=msg.charAt(index++))==u_l || c==u_L) &&
+        ((c=msg.charAt(index++))==u_u || c==u_U) &&
+        ((c=msg.charAt(index++))==u_r || c==u_R) &&
+        ((c=msg.charAt(index++))==u_a || c==u_A) &&
+        ((c=msg.charAt(index))==u_l || c==u_L);
+}
+
+UBool
+MessagePattern::isSelect(int32_t index) {
+    UChar c;
+    return
+        ((c=msg.charAt(index++))==u_s || c==u_S) &&
+        ((c=msg.charAt(index++))==u_e || c==u_E) &&
+        ((c=msg.charAt(index++))==u_l || c==u_L) &&
+        ((c=msg.charAt(index++))==u_e || c==u_E) &&
+        ((c=msg.charAt(index++))==u_c || c==u_C) &&
+        ((c=msg.charAt(index))==u_t || c==u_T);
+}
+
+UBool
+MessagePattern::isOrdinal(int32_t index) {
+    UChar c;
+    return
+        ((c=msg.charAt(index++))==u_o || c==u_O) &&
+        ((c=msg.charAt(index++))==u_r || c==u_R) &&
+        ((c=msg.charAt(index++))==u_d || c==u_D) &&
+        ((c=msg.charAt(index++))==u_i || c==u_I) &&
+        ((c=msg.charAt(index++))==u_n || c==u_N) &&
+        ((c=msg.charAt(index++))==u_a || c==u_A) &&
+        ((c=msg.charAt(index))==u_l || c==u_L);
+}
+
+UBool
+MessagePattern::inMessageFormatPattern(int32_t nestingLevel) {
+    return nestingLevel>0 || partsList->a[0].type==UMSGPAT_PART_TYPE_MSG_START;
+}
+
+UBool
+MessagePattern::inTopLevelChoiceMessage(int32_t nestingLevel, UMessagePatternArgType parentType) {
+    return
+        nestingLevel==1 &&
+        parentType==UMSGPAT_ARG_TYPE_CHOICE &&
+        partsList->a[0].type!=UMSGPAT_PART_TYPE_MSG_START;
+}
+
+void
+MessagePattern::addPart(UMessagePatternPartType type, int32_t index, int32_t length,
+                        int32_t value, UErrorCode &errorCode) {
+    if(partsList->ensureCapacityForOneMore(partsLength, errorCode)) {
+        Part &part=partsList->a[partsLength++];
+        part.type=type;
+        part.index=index;
+        part.length=(uint16_t)length;
+        part.value=(int16_t)value;
+        part.limitPartIndex=0;
+    }
+}
+
+void
+MessagePattern::addLimitPart(int32_t start,
+                             UMessagePatternPartType type, int32_t index, int32_t length,
+                             int32_t value, UErrorCode &errorCode) {
+    partsList->a[start].limitPartIndex=partsLength;
+    addPart(type, index, length, value, errorCode);
+}
+
+void
+MessagePattern::addArgDoublePart(double numericValue, int32_t start, int32_t length,
+                                 UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    int32_t numericIndex=numericValuesLength;
+    if(numericValuesList==NULL) {
+        numericValuesList=new MessagePatternDoubleList();
+        if(numericValuesList==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    } else if(!numericValuesList->ensureCapacityForOneMore(numericValuesLength, errorCode)) {
+        return;
+    } else {
+        if(numericIndex>Part::MAX_VALUE) {
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return;
+        }
+    }
+    numericValuesList->a[numericValuesLength++]=numericValue;
+    addPart(UMSGPAT_PART_TYPE_ARG_DOUBLE, start, length, numericIndex, errorCode);
+}
+
+void
+MessagePattern::setParseError(UParseError *parseError, int32_t index) {
+    if(parseError==NULL) {
+        return;
+    }
+    parseError->offset=index;
+
+    // Set preContext to some of msg before index.
+    // Avoid splitting a surrogate pair.
+    int32_t length=index;
+    if(length>=U_PARSE_CONTEXT_LEN) {
+        length=U_PARSE_CONTEXT_LEN-1;
+        if(length>0 && U16_IS_TRAIL(msg[index-length])) {
+            --length;
+        }
+    }
+    msg.extract(index-length, length, parseError->preContext);
+    parseError->preContext[length]=0;
+
+    // Set postContext to some of msg starting at index.
+    length=msg.length()-index;
+    if(length>=U_PARSE_CONTEXT_LEN) {
+        length=U_PARSE_CONTEXT_LEN-1;
+        if(length>0 && U16_IS_LEAD(msg[index+length-1])) {
+            --length;
+        }
+    }
+    msg.extract(index, length, parseError->postContext);
+    parseError->postContext[length]=0;
+}
+
+// MessageImpl ------------------------------------------------------------- ***
+
+void
+MessageImpl::appendReducedApostrophes(const UnicodeString &s, int32_t start, int32_t limit,
+                                      UnicodeString &sb) {
+    int32_t doubleApos=-1;
+    for(;;) {
+        int32_t i=s.indexOf(u_apos, start);
+        if(i<0 || i>=limit) {
+            sb.append(s, start, limit-start);
+            break;
+        }
+        if(i==doubleApos) {
+            // Double apostrophe at start-1 and start==i, append one.
+            sb.append(u_apos);
+            ++start;
+            doubleApos=-1;
+        } else {
+            // Append text between apostrophes and skip this one.
+            sb.append(s, start, i-start);
+            doubleApos=start=i+1;
+        }
+    }
+}
+
+// Ported from second half of ICU4J SelectFormat.format(String).
+UnicodeString &
+MessageImpl::appendSubMessageWithoutSkipSyntax(const MessagePattern &msgPattern,
+                                               int32_t msgStart,
+                                               UnicodeString &result) {
+    const UnicodeString &msgString=msgPattern.getPatternString();
+    int32_t prevIndex=msgPattern.getPart(msgStart).getLimit();
+    for(int32_t i=msgStart;;) {
+        const MessagePattern::Part &part=msgPattern.getPart(++i);
+        UMessagePatternPartType type=part.getType();
+        int32_t index=part.getIndex();
+        if(type==UMSGPAT_PART_TYPE_MSG_LIMIT) {
+            return result.append(msgString, prevIndex, index-prevIndex);
+        } else if(type==UMSGPAT_PART_TYPE_SKIP_SYNTAX) {
+            result.append(msgString, prevIndex, index-prevIndex);
+            prevIndex=part.getLimit();
+        } else if(type==UMSGPAT_PART_TYPE_ARG_START) {
+            result.append(msgString, prevIndex, index-prevIndex);
+            prevIndex=index;
+            i=msgPattern.getLimitPartIndex(i);
+            index=msgPattern.getPart(i).getLimit();
+            appendReducedApostrophes(msgString, prevIndex, index, result);
+            prevIndex=index;
+        }
+    }
+}
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_FORMATTING
diff --git a/thirdparty/icu4c/common/msvcres.h b/thirdparty/icu4c/common/msvcres.h
new file mode 100644
index 0000000000..0cace85e74
--- /dev/null
+++ b/thirdparty/icu4c/common/msvcres.h
@@ -0,0 +1,25 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//{{NO_DEPENDENCIES}}
+// Copyright (c) 2003-2010 International Business Machines
+// Corporation and others. All Rights Reserved.
+//
+// Used by common.rc and other .rc files.
+//Do not edit with Microsoft Developer Studio because it will modify this
+//header the wrong way. This is here to prevent Visual Studio .NET from
+//unnessarily building the resource files when it's not needed.
+//
+
+/*
+These are defined before unicode/uversion.h in order to prevent 
+STLPort's broken stddef.h from being used when rc.exe parses this file. 
+*/
+#define _STLP_OUTERMOST_HEADER_ID 0
+#define _STLP_WINCE 1
+
+#include "unicode/uversion.h"
+
+#define ICU_WEBSITE "http://icu-project.org"
+#define ICU_COMPANY "The ICU Project"
+#define ICU_PRODUCT_PREFIX "ICU"
+#define ICU_PRODUCT "International Components for Unicode"
diff --git a/thirdparty/icu4c/common/mutex.h b/thirdparty/icu4c/common/mutex.h
new file mode 100644
index 0000000000..44b1f90ba0
--- /dev/null
+++ b/thirdparty/icu4c/common/mutex.h
@@ -0,0 +1,77 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*/
+//----------------------------------------------------------------------------
+// File:     mutex.h
+//
+// Lightweight C++ wrapper for umtx_ C mutex functions
+//
+// Author:   Alan Liu  1/31/97
+// History:
+// 06/04/97   helena         Updated setImplementation as per feedback from 5/21 drop.
+// 04/07/1999  srl               refocused as a thin wrapper
+//
+//----------------------------------------------------------------------------
+#ifndef MUTEX_H
+#define MUTEX_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "umutex.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+  * Mutex is a helper class for convenient locking and unlocking of a UMutex.
+  *
+  * Creating a local scope Mutex will lock a UMutex, holding the lock until the Mutex
+  * goes out of scope.
+  *
+  *  If no UMutex is specified, the ICU global mutex is implied.
+  *
+  *  For example:
+  *
+  *  static UMutex myMutex;
+  *
+  *  void Function(int arg1, int arg2)
+  *  {
+  *     static Object* foo;      // Shared read-write object
+  *     Mutex mutex(&myMutex);   // or no args for the global lock
+  *     foo->Method();
+  *     // When 'mutex' goes out of scope and gets destroyed here, the lock is released
+  *  }
+  *
+  *  Note:  Do NOT use the form 'Mutex mutex();' as that merely forward-declares a function
+  *         returning a Mutex. This is a common mistake which silently slips through the
+  *         compiler!!
+  */
+
+class U_COMMON_API Mutex : public UMemory {
+public:
+    Mutex(UMutex *mutex = nullptr) : fMutex(mutex) {
+        umtx_lock(fMutex);
+    }
+    ~Mutex() {
+        umtx_unlock(fMutex);
+    }
+
+    Mutex(const Mutex &other) = delete; // forbid assigning of this class
+    Mutex &operator=(const Mutex &other) = delete; // forbid copying of this class
+    void *operator new(size_t s) = delete;  // forbid heap allocation. Locals only.
+
+private:
+    UMutex   *fMutex;
+};
+
+
+U_NAMESPACE_END
+
+#endif //_MUTEX_
+//eof
diff --git a/thirdparty/icu4c/common/norm2_nfc_data.h b/thirdparty/icu4c/common/norm2_nfc_data.h
new file mode 100644
index 0000000000..455cc0c428
--- /dev/null
+++ b/thirdparty/icu4c/common/norm2_nfc_data.h
@@ -0,0 +1,1149 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+// Copyright (C) 1999-2016, International Business Machines
+// Corporation and others.  All Rights Reserved.
+//
+// file name: norm2_nfc_data.h
+//
+// machine-generated by: icu/source/tools/gennorm2/n2builder.cpp
+
+
+#ifdef INCLUDED_FROM_NORMALIZER2_CPP
+
+static const UVersionInfo norm2_nfc_data_formatVersion={4,0,0,0};
+static const UVersionInfo norm2_nfc_data_dataVersion={0xd,0,0,0};
+
+static const int32_t norm2_nfc_data_indexes[Normalizer2Impl::IX_COUNT]={
+0x50,0x4bac,0x8814,0x8914,0x8914,0x8914,0x8914,0x8914,0xc0,0x300,0xae2,0x29e0,0x3c66,0xfc00,0x1288,0x3b9c,
+0x3c34,0x3c66,0x300,0
+};
+
+static const uint16_t norm2_nfc_data_trieIndex[1746]={
+0,0x40,0x7b,0xbb,0xfb,0x13a,0x17a,0x1b2,0x1f2,0x226,0x254,0x226,0x294,0x2d4,0x313,0x353,
+0x393,0x3d2,0x40f,0x44e,0x226,0x226,0x488,0x4c8,0x4f8,0x530,0x226,0x570,0x59f,0x5de,0x226,0x5f3,
+0x631,0x65f,0x226,0x68c,0x6cc,0x709,0x729,0x768,0x7a7,0x7e4,0x803,0x840,0x729,0x879,0x8a7,0x8e6,
+0x226,0x920,0x937,0x977,0x98e,0x9cd,0x226,0xa03,0xa23,0xa5e,0xa6a,0xaa5,0xacd,0xb0a,0xb4a,0xb84,
+0xb9f,0x226,0xbda,0x226,0xc1a,0xc39,0xc6f,0xcac,0x226,0x226,0x226,0x226,0x226,0xccf,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0xcfb,0x226,0x226,0xd30,
+0x226,0x226,0xd4e,0x226,0xd78,0x226,0x226,0x226,0xdb4,0xdd4,0xe14,0xe53,0xe8e,0xece,0xf02,0xf2e,
+0x808,0x226,0x226,0xf62,0x226,0x226,0x226,0xfa2,0xfe2,0x1022,0x1062,0x10a2,0x10e2,0x1122,0x1162,0x11a2,
+0x11e2,0x226,0x226,0x1212,0x1243,0x226,0x1273,0x12a6,0x12e3,0x1322,0x1362,0x1398,0x13c6,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x13f1,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0xcbd,0x226,0x140e,0x226,0x144e,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x148e,0x14c8,0x1506,0x1546,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1585,0x15c3,0x15e3,0x226,0x226,0x226,0x226,
+0x161d,0x226,0x226,0x1645,0x1677,0x16a5,0x80c,0x16b8,0x226,0x226,0x16c8,0x1708,0x226,0x226,0x226,0x1420,
+0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,
+0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,
+0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,
+0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,
+0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,
+0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,
+0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,
+0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,
+0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,
+0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,
+0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x175c,0x1748,0x1750,0x1758,0x1760,0x174c,0x1754,0x1794,0x226,
+0x17d4,0x180f,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x184f,0x188f,0x18cf,0x190f,0x194f,0x198f,0x19cf,0x1a0f,0x1a32,0x1a72,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1a92,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x655,0x664,0x67c,0x69b,0x6b0,0x6b0,0x6b0,0x6b4,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0xbda,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x54f,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x40c,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1ac5,0x226,0x226,0x1ad5,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0xdc6,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1ae5,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x15d6,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x1aef,0x54f,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x7eb,0x226,0x226,
+0x9ba,0x226,0x1aff,0x1b0c,0x1b18,0x226,0x226,0x226,0x226,0x414,0x226,0x1b23,0x1b33,0x226,0x226,0x226,
+0x7e0,0x226,0x226,0x226,0x226,0x1b43,0x226,0x226,0x226,0x1b4e,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x1b55,0x226,0x226,0x226,0x226,0x1b60,0x1b6f,0x8f6,0x1b7d,0x412,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x1b8b,0x798,0x226,0x226,0x226,0x226,0x226,0x1b9b,0x1baa,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x8d6,0x1bb2,0x1bc2,0x226,
+0x226,0x226,0x9ba,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1bcc,0x226,0x226,0x226,0x226,0x226,
+0x226,0x7e6,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1bc9,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1bdc,
+0x7e0,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x84d,0x226,0x226,0x226,0x7ed,0x7ea,
+0x226,0x226,0x226,0x226,0x7e8,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x9ba,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0xbd4,0x226,0x226,0x226,
+0x226,0x7ea,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x1bec,0x226,0x226,0x226,0xefb,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x1bfc,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1bfe,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x1c0d,0x1c1d,0x1c2b,0x1c38,0x226,0x1c44,0x1c52,0x1c62,0x226,0x226,
+0x226,0x226,0xcea,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x1c72,0x1c7a,
+0x1c88,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0xefb,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x4fc,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x1c98,0x226,0x226,0x226,0x226,0x226,0x226,0x1ca4,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x1cb4,0x1cc4,0x1cd4,0x1ce4,0x1cf4,0x1d04,0x1d14,0x1d24,0x1d34,0x1d44,0x1d54,
+0x1d64,0x1d74,0x1d84,0x1d94,0x1da4,0x1db4,0x1dc4,0x1dd4,0x1de4,0x1df4,0x1e04,0x1e14,0x1e24,0x1e34,0x1e44,0x1e54,
+0x1e64,0x1e74,0x1e84,0x1e94,0x1ea4,0x1eb4,0x1ec4,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,0x226,
+0x226,0x226,0x226,0x226,0x226,0x408,0x428,0xc4,0xc4,0xc4,0x448,0x457,0x46d,0x489,0x4a6,0x4c2,
+0x4df,0x4fc,0x51b,0x538,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,
+0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0x552,0xc4,0x566,0xc4,0xc4,0xc4,0xc4,
+0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,
+0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0x586,0xc4,0xc4,0xc4,0xc4,0xc4,
+0xc4,0xc4,0xc4,0x591,0x5ae,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0x5ce,0x5e2,0xc4,0xc4,0x5f5,
+0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,
+0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,0xc4,
+0x615,0x635
+};
+
+static const uint16_t norm2_nfc_data_trieData[7892]={
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,4,8,0xc,1,
+1,0x10,0x50,0x5c,0x70,0x88,0xcc,0xd0,0xec,0x108,0x144,0x148,0x15c,0x174,0x180,0x1a4,
+0x1e4,1,0x1ec,0x20c,0x228,0x244,0x290,0x298,0x2b0,0x2b8,0x2dc,1,1,1,1,1,
+1,0x2f4,0x334,0x340,0x354,0x36c,0x3b0,0x3b4,0x3d0,0x3f0,0x428,0x430,0x444,0x45c,0x468,0x48c,
+0x4cc,1,0x4d4,0x4f4,0x510,0x530,0x57c,0x584,0x5a0,0x5a8,0x5d0,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x5e8,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0x128a,0x1290,0xae4,0x1296,0xafa,
+0xb04,0x5f4,0xb0e,0x129c,0x12a2,0xb18,0x12a8,0x12ae,0x12b4,0x12ba,0xb2e,1,0x12c0,0x12c6,0x12cc,0xb38,
+0xb4e,0xb60,1,0x5fc,0x12d2,0x12d8,0x12de,0xb6a,0x12e4,1,1,0x12ea,0x12f0,0xb80,0x12f6,0xb96,
+0xba0,0x600,0xbaa,0x12fc,0x1302,0xbb4,0x1308,0x130e,0x1314,0x131a,0xbca,1,0x1320,0x1326,0x132c,0xbd4,
+0xbea,0xbfc,1,0x608,0x1332,0x1338,0x133e,0xc06,0x1344,1,0x134a,0x1350,0x1356,0xc1c,0xc32,0x135d,
+0x1363,0x1368,0x136e,0x1374,0x137a,0x1380,0x1386,0x138c,0x1392,0x1398,0x139e,1,1,0xc48,0xc56,0x13a4,
+0x13aa,0x13b0,0x13b6,0x13bd,0x13c3,0x13c8,0x13ce,0x13d4,0x13da,0x13e0,0x13e6,0x13ec,0x13f2,0x13f9,0x13ff,0x1404,
+0x140a,1,1,0x1410,0x1416,0x141c,0x1422,0x1428,0x142e,0x1435,0x143b,0x1440,1,1,1,0x1447,
+0x144d,0x1453,0x1459,1,0x145e,0x1464,0x146b,0x1471,0x1476,0x147c,1,1,1,0x1482,0x1488,0x148f,
+0x1495,0x149a,0x14a0,1,1,1,0xc64,0xc72,0x14a6,0x14ac,0x14b2,0x14b8,1,1,0x14be,0x14c4,
+0x14cb,0x14d1,0x14d6,0x14dc,0xc80,0xc8a,0x14e2,0x14e8,0x14ef,0x14f5,0xc94,0xc9e,0x14fb,0x1501,0x1506,0x150c,
+1,1,0xca8,0xcb2,0xcbc,0xcc6,0x1512,0x1518,0x151e,0x1524,0x152a,0x1530,0x1537,0x153d,0x1542,0x1548,
+0x154e,0x1554,0x155a,0x1560,0x1566,0x156c,0x1572,0x1578,0x157e,0x60c,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,0xcd0,0xcea,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xd04,0xd1e,1,1,1,1,1,
+1,0x610,1,1,1,1,1,1,1,1,1,1,1,1,1,0x1584,
+0x158a,0x1590,0x1596,0x159c,0x15a2,0x15a8,0x15ae,0x15b6,0x15c0,0x15ca,0x15d4,0x15de,0x15e8,0x15f2,0x15fc,1,
+0x1606,0x1610,0x161a,0x1624,0x162d,0x1633,1,1,0x1638,0x163e,0x1644,0x164a,0xd38,0xd42,0x1653,0x165d,
+0x1665,0x166b,0x1671,1,1,1,0x1676,0x167c,1,1,0x1682,0x1688,0x1690,0x169a,0x16a3,0x16a9,
+0x16af,0x16b5,0x16ba,0x16c0,0x16c6,0x16cc,0x16d2,0x16d8,0x16de,0x16e4,0x16ea,0x16f0,0x16f6,0x16fc,0x1702,0x1708,
+0x170e,0x1714,0x171a,0x1720,0x1726,0x172c,0x1732,0x1738,0x173e,0x1744,0x174a,0x1750,0x1756,0x175c,1,1,
+0x1762,0x1768,1,1,1,1,1,1,0xd4c,0xd56,0xd60,0xd6a,0x1770,0x177a,0x1784,0x178e,
+0xd74,0xd7e,0x1798,0x17a2,0x17aa,0x17b0,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0x614,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xfdcc,0xfdcc,0xfdcc,0xfdcc,0xfdcc,0xffcc,0xfdcc,0xfdcc,0xfdcc,0xfdcc,0xfdcc,0xfdcc,
+0xfdcc,0xffcc,0xffcc,0xfdcc,0xffcc,0xfdcc,0xffcc,0xfdcc,0xfdcc,0xffd0,0xffb8,0xffb8,0xffb8,0xffb8,0xffd0,0xfdb0,
+0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xff94,0xff94,0xfdb8,0xfdb8,0xfdb8,0xfdb8,0xfd94,0xfd94,0xffb8,0xffb8,0xffb8,
+0xffb8,0xfdb8,0xfdb8,0xffb8,0xfdb8,0xfdb8,0xffb8,0xffb8,0xfe02,0xfe02,0xfe02,0xfe02,0xfc02,0xffb8,0xffb8,0xffb8,
+0xffb8,0xffcc,0xffcc,0xffcc,0x3c36,0x3c3c,0xfdcc,0x3c42,0x3c48,0xfde0,0xffcc,0xffb8,0xffb8,0xffb8,0xffcc,0xffcc,
+0xffcc,0xffb8,0xffb8,1,0xffcc,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xffb8,0xffcc,0xffd0,0xffb8,0xffb8,0xffcc,
+0xffd2,0xffd4,0xffd4,0xffd2,0xffd4,0xffd4,0xffd2,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,1,0x29e1,1,1,1,1,1,1,1,
+1,1,0x29e5,1,1,1,1,1,0x17b7,0x17bd,0x29e9,0x17c3,0x17c9,0x17cf,1,0x17d5,
+1,0x17db,0x17e1,0x17e9,0x618,1,1,1,0x634,1,0x644,1,0x658,1,1,1,
+1,1,0x674,1,0x684,1,1,1,0x688,1,1,1,0x6a0,0x17f1,0x17f7,0xd88,
+0x17fd,0xd92,0x1803,0x180b,0x6b4,1,1,1,0x6d4,1,0x6e4,1,0x6fc,1,1,1,
+1,1,0x71c,1,0x72c,1,1,1,0x734,1,1,1,0x754,0xd9c,0xdae,0x1813,
+0x1819,0xdc0,1,1,1,0x76c,0x181f,0x1825,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x182b,0x1831,1,0x1837,1,1,0x774,0x183d,1,1,1,1,0x1843,
+0x1849,0x184f,1,0x778,1,1,0x780,1,0x784,0x790,0x798,0x79c,0x1855,0x7ac,1,1,
+1,0x7b0,1,1,1,1,0x7b4,1,1,1,0x7c4,1,1,1,0x7c8,1,
+0x7cc,1,1,0x7d0,1,1,0x7d8,1,0x7dc,0x7e8,0x7f0,0x7f4,0x185b,0x804,1,1,
+1,0x808,1,1,1,0x80c,1,1,1,0x81c,1,1,1,0x820,1,0x824,
+1,1,0x1861,0x1867,1,0x186d,1,1,0x828,0x1873,1,1,1,1,0x1879,0x187f,
+0x1885,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0x82c,0x830,0x188b,0x1891,1,1,1,1,1,1,
+1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x1897,
+0x189d,1,1,1,1,1,1,1,1,1,1,1,1,1,0x18a3,0x18a9,
+0x18af,0x18b5,1,1,0x18bb,0x18c1,0x834,0x838,0x18c7,0x18cd,0x18d3,0x18d9,0x18df,0x18e5,1,1,
+0x18eb,0x18f1,0x18f7,0x18fd,0x1903,0x1909,0x83c,0x840,0x190f,0x1915,0x191b,0x1921,0x1927,0x192d,0x1933,0x1939,
+0x193f,0x1945,0x194b,0x1951,1,1,0x1957,0x195d,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,
+0xffcc,0xffcc,0xffbc,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,
+0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffbc,0xffc8,0xffcc,0xfe14,0xfe16,0xfe18,0xfe1a,0xfe1c,0xfe1e,0xfe20,0xfe22,
+0xfe24,0xfe26,0xfe26,0xfe28,0xfe2a,0xfe2c,1,0xfe2e,1,0xfe30,0xfe32,1,0xffcc,0xffb8,1,0xfe24,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xfe3c,0xfe3e,0xfe40,1,1,1,1,1,1,1,0x1962,0x1968,0x196f,0x1975,0x197b,0x844,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0x850,1,0x854,0xfe36,0xfe38,0xfe3a,0xfe3c,0xfe3e,
+0xfe40,0xfe42,0xfe44,0xfdcc,0xfdcc,0xfdb8,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffb8,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+0xfe46,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+0x1981,0x858,0x1987,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0x85c,0x198d,1,0x860,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,1,1,0xffcc,0xffcc,1,0xffb8,0xffcc,0xffcc,0xffb8,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+0xfe48,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xffcc,
+0xffb8,0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xffcc,0xffb8,0xffb8,0xffcc,0xffb8,0xffcc,0xffcc,
+0xffb8,0xffcc,0xffb8,0xffcc,0xffb8,0xffcc,0xffb8,0xffcc,0xffcc,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,1,1,1,1,1,1,1,1,1,
+0xffb8,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,1,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,0xffcc,0xffcc,0xffcc,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xffb8,0xffb8,0xffb8,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xffb8,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,0xffb8,
+0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xfe36,0xfe38,0xfe3a,0xffcc,
+0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffb8,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0x864,0x1993,1,1,1,1,1,1,0x868,0x1999,1,0x86c,
+0x199f,1,1,1,1,1,1,1,0xfc0e,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0xfe12,1,1,1,0xffcc,0xffb8,0xffcc,0xffcc,1,1,
+1,0x29ec,0x29f2,0x29f8,0x29fe,0x2a04,0x2a0a,0x2a10,0x2a16,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,0xfe0e,1,0xfc00,1,1,1,1,1,1,1,0x870,
+1,1,1,0x19a5,0x19ab,0xfe12,1,1,1,1,1,1,1,1,1,0xfc00,
+1,1,1,1,0x2a1c,0x2a22,1,0x2a28,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0xffcc,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,0x2a2e,1,1,0x2a34,1,1,
+1,1,1,0xfe0e,1,1,1,1,1,1,1,1,1,1,1,1,
+1,0xfe12,1,1,1,1,1,1,1,1,1,1,1,0x2a3a,0x2a40,0x2a46,
+1,1,0x2a4c,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe0e,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+0x878,0x19b1,1,1,0x19b7,0x19bd,0xfe12,1,1,1,1,1,1,1,1,0xfc00,
+0xfc00,1,1,1,1,0x2a52,0x2a58,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0x884,1,0x19c3,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,0xfc00,1,1,1,1,1,1,0x888,0x890,1,1,
+0x19c9,0x19cf,0x19d5,0xfe12,1,1,1,1,1,1,1,1,1,0xfc00,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0x894,1,0x19db,1,1,1,1,0xfe12,1,1,
+1,1,1,1,1,0xfea8,0xfcb6,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0xfe0e,1,1,0x898,0x19e1,1,0xfc00,1,1,1,0x89c,0x19e7,0x19ed,
+1,0xdca,0x19f5,1,0xfe12,1,1,1,1,1,1,1,0xfc00,0xfc00,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xfe12,0xfe12,1,0xfc00,1,1,1,
+1,1,1,0x8a8,0x8b0,1,1,0x19fd,0x1a03,0x1a09,0xfe12,1,1,1,1,1,
+1,1,1,1,0xfc00,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0xfc12,1,1,
+1,1,0xfc00,1,1,1,1,1,1,1,1,1,0x8b4,0x1a0f,1,0xdd4,
+0x1a17,0x1a1f,0xfc00,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0xfece,0xfece,0xfe12,1,1,
+1,1,1,1,1,1,0xfed6,0xfed6,0xfed6,0xfed6,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xfeec,0xfeec,0xfe12,1,1,1,1,1,1,1,1,0xfef4,0xfef4,0xfef4,
+0xfef4,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,0xffb8,0xffb8,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffb8,1,0xffb8,1,0xffb0,1,1,1,1,1,1,0x2a5f,1,1,
+1,1,1,1,1,1,1,0x2a65,1,1,1,1,0x2a6b,1,1,1,
+1,0x2a71,1,1,1,1,0x2a77,1,1,1,1,1,1,1,1,1,
+1,1,1,0x2a7d,1,1,1,1,1,1,1,0xff02,0xff04,0x3c50,0xff08,0x3c58,
+0x2a82,1,0x2a88,1,0xff04,0xff04,0xff04,0xff04,1,1,0xff04,0x3c60,0xffcc,0xffcc,0xfe12,1,
+0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,0x2a8f,1,1,
+1,1,1,1,1,1,1,0x2a95,1,1,1,1,0x2a9b,1,1,1,
+1,0x2aa1,1,1,1,1,0x2aa7,1,1,1,1,1,1,1,1,1,
+1,1,1,0x2aad,1,1,1,1,1,1,0xffb8,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0x8c0,0x1a25,1,1,1,1,1,1,1,0xfc00,1,1,
+1,1,1,1,1,1,0xfe0e,1,0xfe12,0xfe12,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0xffb8,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,
+0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,
+0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,0xfe00,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0xffcc,0xffcc,0xffcc,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xfe12,1,1,1,1,1,1,1,1,1,1,0xffcc,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0xffc8,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,0xffbc,0xffcc,0xffb8,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0xffcc,0xffb8,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xfe12,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,1,1,0xffb8,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffcc,
+0xffcc,0xffb8,1,0xffb8,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x8c4,0x1a2b,0x8c8,0x1a31,0x8cc,0x1a37,0x8d0,0x1a3d,0x8d4,0x1a43,1,1,0x8d8,
+0x1a49,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xfe0e,0xfc00,1,1,1,1,0x8dc,0x1a4f,0x8e0,0x1a55,0x8e4,0x8e8,0x1a5b,0x1a61,
+0x8ec,0x1a67,0xfe12,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xffcc,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,0xfe12,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xfe0e,1,1,1,1,1,1,1,1,1,1,1,
+0xfe12,0xfe12,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffcc,0xffcc,0xffcc,1,0xfe02,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffcc,0xffcc,0xffb8,0xffb8,
+0xffb8,0xffb8,0xffcc,1,0xfe02,0xfe02,0xfe02,0xfe02,0xfe02,0xfe02,0xfe02,1,1,1,1,0xffb8,
+1,1,1,1,1,1,0xffcc,1,1,1,0xffcc,0xffcc,1,1,1,1,
+1,1,0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,0xffcc,0xffd4,
+0xffac,0xffb8,0xff94,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffd0,0xffc8,0xffc8,0xffb8,1,0xffcc,0xffd2,0xffb8,
+0xffcc,0xffb8,0x1a6c,0x1a72,0x1a78,0x1a7e,0x1a85,0x1a8b,0x1a91,0x1a97,0x1a9f,0x1aa9,0x1ab0,0x1ab6,0x1abc,0x1ac2,
+0x1ac8,0x1ace,0x1ad5,0x1adb,0x1ae0,0x1ae6,0x1aee,0x1af8,0x1b02,0x1b0c,0x1b14,0x1b1a,0x1b20,0x1b26,0x1b2f,0x1b39,
+0x1b41,0x1b47,0x1b4c,0x1b52,0x1b58,0x1b5e,0x1b64,0x1b6a,0x1b70,0x1b76,0x1b7d,0x1b83,0x1b88,0x1b8e,0x1b94,0x1b9a,
+0x1ba2,0x1bac,0x1bb4,0x1bba,0x1bc0,0x1bc6,0x1bcc,0x1bd2,0xdde,0xde8,0x1bda,0x1be4,0x1bec,0x1bf2,0x1bf8,0x1bfe,
+0x1c04,0x1c0a,0x1c10,0x1c16,0x1c1d,0x1c23,0x1c28,0x1c2e,0x1c34,0x1c3a,0x1c40,0x1c46,0x1c4c,0x1c52,0x1c5a,0x1c64,
+0x1c6e,0x1c78,0x1c82,0x1c8c,0x1c96,0x1ca0,0x1ca9,0x1caf,0x1cb5,0x1cbb,0x1cc0,0x1cc6,0xdf2,0xdfc,0x1cce,0x1cd8,
+0x1ce0,0x1ce6,0x1cec,0x1cf2,0xe06,0xe10,0x1cfa,0x1d04,0x1d0e,0x1d18,0x1d22,0x1d2c,0x1d34,0x1d3a,0x1d40,0x1d46,
+0x1d4c,0x1d52,0x1d58,0x1d5e,0x1d64,0x1d6a,0x1d70,0x1d76,0x1d7c,0x1d82,0x1d8a,0x1d94,0x1d9e,0x1da8,0x1db0,0x1db6,
+0x1dbd,0x1dc3,0x1dc8,0x1dce,0x1dd4,0x1dda,0x1de0,0x1de6,0x1dec,0x1df2,0x1df9,0x1dff,0x1e05,0x1e0b,0x1e11,0x1e17,
+0x1e1c,0x1e22,0x1e28,0x1e2e,0x1e35,0x1e3b,0x1e41,0x1e47,0x1e4c,0x1e52,0x1e58,0x1e5e,1,0x1e65,1,1,
+1,1,0xe1a,0xe28,0x1e6a,0x1e70,0x1e78,0x1e82,0x1e8c,0x1e96,0x1ea0,0x1eaa,0x1eb4,0x1ebe,0x1ec8,0x1ed2,
+0x1edc,0x1ee6,0x1ef0,0x1efa,0x1f04,0x1f0e,0x1f18,0x1f22,0x1f2c,0x1f36,0xe36,0xe40,0x1f3e,0x1f44,0x1f4a,0x1f50,
+0x1f58,0x1f62,0x1f6c,0x1f76,0x1f80,0x1f8a,0x1f94,0x1f9e,0x1fa8,0x1fb2,0x1fba,0x1fc0,0x1fc6,0x1fcc,0xe4a,0xe54,
+0x1fd2,0x1fd8,0x1fe0,0x1fea,0x1ff4,0x1ffe,0x2008,0x2012,0x201c,0x2026,0x2030,0x203a,0x2044,0x204e,0x2058,0x2062,
+0x206c,0x2076,0x2080,0x208a,0x2094,0x209e,0x20a6,0x20ac,0x20b2,0x20b8,0x20c0,0x20ca,0x20d4,0x20de,0x20e8,0x20f2,
+0x20fc,0x2106,0x2110,0x211a,0x2122,0x2128,0x212f,0x2135,0x213a,0x2140,0x2146,0x214c,1,1,1,1,
+1,1,0xe5e,0xe74,0xe8c,0xe9a,0xea8,0xeb6,0xec4,0xed2,0xede,0xef4,0xf0c,0xf1a,0xf28,0xf36,
+0xf44,0xf52,0xf5e,0xf6c,0x2155,0x215f,0x2169,0x2173,1,1,0xf7a,0xf88,0x217d,0x2187,0x2191,0x219b,
+1,1,0xf96,0xfac,0xfc4,0xfd2,0xfe0,0xfee,0xffc,0x100a,0x1016,0x102c,0x1044,0x1052,0x1060,0x106e,
+0x107c,0x108a,0x1096,0x10a8,0x21a5,0x21af,0x21b9,0x21c3,0x21cd,0x21d7,0x10ba,0x10cc,0x21e1,0x21eb,0x21f5,0x21ff,
+0x2209,0x2213,0x10de,0x10ec,0x221d,0x2227,0x2231,0x223b,1,1,0x10fa,0x1108,0x2245,0x224f,0x2259,0x2263,
+1,1,0x1116,0x1128,0x226d,0x2277,0x2281,0x228b,0x2295,0x229f,1,0x113a,1,0x22a9,1,0x22b3,
+1,0x22bd,0x114c,0x1162,0x117a,0x1188,0x1196,0x11a4,0x11b2,0x11c0,0x11cc,0x11e2,0x11fa,0x1208,0x1216,0x1224,
+0x1232,0x1240,0x124c,0x3b9e,0x22c5,0x3ba6,0x1256,0x3bae,0x22cb,0x3bb6,0x22d1,0x3bbe,0x22d7,0x3bc6,0x1260,0x3bce,
+1,1,0x22de,0x22e8,0x22f7,0x2307,0x2317,0x2327,0x2337,0x2347,0x2352,0x235c,0x236b,0x237b,0x238b,0x239b,
+0x23ab,0x23bb,0x23c6,0x23d0,0x23df,0x23ef,0x23ff,0x240f,0x241f,0x242f,0x243a,0x2444,0x2453,0x2463,0x2473,0x2483,
+0x2493,0x24a3,0x24ae,0x24b8,0x24c7,0x24d7,0x24e7,0x24f7,0x2507,0x2517,0x2522,0x252c,0x253b,0x254b,0x255b,0x256b,
+0x257b,0x258b,0x2595,0x259b,0x25a3,0x25aa,0x25b3,1,0x126a,0x25bd,0x25c5,0x25cb,0x25d1,0x3bd6,0x25d6,1,
+0x2ab2,0x8f0,1,0x25dd,0x25e5,0x25ec,0x25f5,1,0x1274,0x25ff,0x2607,0x3bde,0x260d,0x3be6,0x2612,0x2619,
+0x261f,0x2625,0x262b,0x2631,0x2639,0x3bf0,1,1,0x2641,0x2649,0x2651,0x2657,0x265d,0x3bfa,1,0x2663,
+0x2669,0x266f,0x2675,0x267b,0x2683,0x3c04,0x268b,0x2691,0x2697,0x269f,0x26a7,0x26ad,0x26b3,0x3c0e,0x26b9,0x26bf,
+0x3c16,0x2ab7,1,1,0x26c7,0x26ce,0x26d7,1,0x127e,0x26e1,0x26e9,0x3c1e,0x26ef,0x3c26,0x26f4,0x2abb,
+0x8fc,1,0xfa09,0xfa09,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffcc,0xffcc,0xfe02,0xfe02,0xffcc,0xffcc,0xffcc,0xffcc,0xfe02,0xfe02,0xfe02,0xffcc,0xffcc,1,
+1,1,1,0xffcc,1,1,1,0xfe02,0xfe02,0xffcc,0xffb8,0xffcc,0xfe02,0xfe02,0xffb8,0xffb8,
+0xffb8,0xffb8,0xffcc,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0x2abe,1,1,1,0x2ac2,0x3c2e,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x908,1,0x90c,1,0x910,1,1,1,1,1,0x26fb,0x2701,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,0x2707,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x270d,0x2713,0x2719,0x914,1,0x918,1,0x91c,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0x920,0x271f,1,1,1,0x924,0x2725,1,0x928,0x272b,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0x92c,0x2731,0x930,0x2737,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x934,
+1,1,1,0x273d,1,0x938,0x2743,0x93c,1,0x2749,0x940,0x274f,1,1,1,0x944,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0x2755,0x948,0x275b,1,0x94c,0x950,1,1,1,1,1,1,1,0x2761,
+0x2767,0x276d,0x2773,0x2779,0x954,0x958,0x277f,0x2785,0x95c,0x960,0x278b,0x2791,0x964,0x968,0x96c,0x970,
+1,1,0x2797,0x279d,0x974,0x978,0x27a3,0x27a9,0x97c,0x980,0x27af,0x27b5,1,1,1,1,
+1,1,1,0x984,0x988,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0x98c,1,1,1,1,1,0x990,0x994,1,0x998,0x27bb,0x27c1,
+0x27c7,0x27cd,1,1,0x99c,0x9a0,0x9a4,0x9a8,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0x27d3,0x27d9,0x27df,0x27e5,1,1,1,1,
+1,1,0x27eb,0x27f1,0x27f7,0x27fd,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x2ac7,
+0x2acb,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0x2acf,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xffb4,0xffc8,0xffd0,0xffbc,0xffc0,0xffc0,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x9ac,1,
+1,1,1,0x9b0,0x2803,0x9b4,0x2809,0x9b8,0x280f,0x9bc,0x2815,0x9c0,0x281b,0x9c4,0x2821,0x9c8,
+0x2827,0x9cc,0x282d,0x9d0,0x2833,0x9d4,0x2839,0x9d8,0x283f,0x9dc,0x2845,1,0x9e0,0x284b,0x9e4,0x2851,
+0x9e8,0x2857,1,1,1,1,1,0x9ec,0x285d,0x2863,0x9f4,0x2869,0x286f,0x9fc,0x2875,0x287b,
+0xa04,0x2881,0x2887,0xa0c,0x288d,0x2893,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,0x2899,1,1,1,1,0xfc10,
+0xfc10,1,1,0xa14,0x289f,1,1,1,1,1,1,1,0xa18,1,1,1,
+1,0xa1c,0x28a5,0xa20,0x28ab,0xa24,0x28b1,0xa28,0x28b7,0xa2c,0x28bd,0xa30,0x28c3,0xa34,0x28c9,0xa38,
+0x28cf,0xa3c,0x28d5,0xa40,0x28db,0xa44,0x28e1,0xa48,0x28e7,1,0xa4c,0x28ed,0xa50,0x28f3,0xa54,0x28f9,
+1,1,1,1,1,0xa58,0x28ff,0x2905,0xa60,0x290b,0x2911,0xa68,0x2917,0x291d,0xa70,0x2923,
+0x2929,0xa78,0x292f,0x2935,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,0xa80,0xa84,0xa88,0xa8c,1,0x293b,1,1,0x2941,0x2947,0x294d,
+0x2953,1,1,0xa90,0x2959,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xffcc,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0xfe12,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xfe12,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffb8,0xffb8,0xffb8,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xfe12,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xffcc,1,0xffcc,0xffcc,0xffb8,1,1,0xffcc,
+0xffcc,1,1,1,1,1,0xffcc,0xffcc,1,0xffcc,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xfe12,1,
+1,1,1,1,1,1,1,1,0xae2,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0xae2,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0xae2,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0xae2,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,
+0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,0x1289,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0x3c66,1,0x3c66,0x3c66,0x3c66,0x3c66,0x3c66,0x3c66,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x3c66,0x3c66,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x3c66,1,1,1,1,0x3c66,1,1,1,0x3c66,1,0x3c66,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0x3b97,1,0x2ad5,
+0x2ad9,0x2add,0x2ae1,0x2ae5,0x2ae9,0x2aed,0x2af1,0x2af1,0x2af5,0x2af9,0x2afd,0x2b01,0x2b05,0x2b09,0x2b0d,0x2b11,
+0x2b15,0x2b19,0x2b1d,0x2b21,0x2b25,0x2b29,0x2b2d,0x2b31,0x2b35,0x2b39,0x2b3d,0x2b41,0x2b45,0x2b49,0x2b4d,0x2b51,
+0x2b55,0x2b59,0x2b5d,0x2b61,0x2b65,0x2b69,0x2b6d,0x2b71,0x2b75,0x2b79,0x2b7d,0x2b81,0x2b85,0x2b89,0x2b8d,0x2b91,
+0x2b95,0x2b99,0x2b9d,0x2ba1,0x2ba5,0x2ba9,0x2bad,0x2bb1,0x2bb5,0x2bb9,0x2bbd,0x2bc1,0x2bc5,0x2bc9,0x2bcd,0x2bd1,
+0x2bd5,0x2bd9,0x2bdd,0x2be1,0x2be5,0x2be9,0x2bed,0x2bf1,0x2bf5,0x2bf9,0x2bfd,0x2c01,0x2c05,0x2c09,0x2c0d,0x2c11,
+0x2c15,0x2c19,0x2c1d,0x2c21,0x2c25,0x2c29,0x2c2d,0x2c31,0x2c35,0x2c39,0x2c3d,0x2b21,0x2c41,0x2c45,0x2c49,0x2c4d,
+0x2c51,0x2c55,0x2c59,0x2c5d,0x2c61,0x2c65,0x2c69,0x2c6d,0x2c71,0x2c75,0x2c79,0x2c7d,0x2c81,0x2c85,0x2c89,0x2c8d,
+0x2c91,0x2c95,0x2c99,0x2c9d,0x2ca1,0x2ca5,0x2ca9,0x2cad,0x2cb1,0x2cb5,0x2cb9,0x2cbd,0x2cc1,0x2cc5,0x2cc9,0x2ccd,
+0x2cd1,0x2cd5,0x2cd9,0x2cdd,0x2ce1,0x2ce5,0x2ce9,0x2ced,0x2cf1,0x2cf5,0x2cf9,0x2cfd,0x2d01,0x2d05,0x2d09,0x2d0d,
+0x2d11,0x2d15,0x2d19,0x2d1d,0x2d21,0x2d25,0x2d29,0x2d2d,0x2d31,0x2d35,0x2d39,0x2d3d,0x2d41,0x2d45,0x2d49,0x2d4d,
+0x2c89,0x2d51,0x2d55,0x2d59,0x2d5d,0x2d61,0x2d65,0x2d69,0x2d6d,0x2c49,0x2d71,0x2d75,0x2d79,0x2d7d,0x2d81,0x2d85,
+0x2d89,0x2d8d,0x2d91,0x2d95,0x2d99,0x2d9d,0x2da1,0x2da5,0x2da9,0x2dad,0x2db1,0x2db5,0x2db9,0x2dbd,0x2b21,0x2dc1,
+0x2dc5,0x2dc9,0x2dcd,0x2dd1,0x2dd5,0x2dd9,0x2ddd,0x2de1,0x2de5,0x2de9,0x2ded,0x2df1,0x2df5,0x2df9,0x2dfd,0x2e01,
+0x2e05,0x2e09,0x2e0d,0x2e11,0x2e15,0x2e19,0x2e1d,0x2e21,0x2e25,0x2e29,0x2c51,0x2e2d,0x2e31,0x2e35,0x2e39,0x2e3d,
+0x2e41,0x2e45,0x2e49,0x2e4d,0x2e51,0x2e55,0x2e59,0x2e5d,0x2e61,0x2e65,0x2e69,0x2e6d,0x2e71,0x2e75,0x2e79,0x2e7d,
+0x2e81,0x2e85,0x2e89,0x2e8d,0x2e91,0x2e95,0x2e99,0x2e9d,0x2ea1,0x2ea5,0x2ea9,0x2ead,0x2eb1,0x2eb5,0x2eb9,0x2ebd,
+0x2ec1,0x2ec5,0x2ec9,0x2ecd,0x2ed1,0x2ed5,0x2ed9,0x2edd,0x2ee1,0x2ee5,0x2ee9,0x2eed,0x2ef1,1,1,0x2ef5,
+1,0x2ef9,1,1,0x2efd,0x2f01,0x2f05,0x2f09,0x2f0d,0x2f11,0x2f15,0x2f19,0x2f1d,0x2f21,1,0x2f25,
+1,0x2f29,1,1,0x2f2d,0x2f31,1,1,1,0x2f35,0x2f39,0x2f3d,0x2f41,0x2f45,0x2f49,0x2f4d,
+0x2f51,0x2f55,0x2f59,0x2f5d,0x2f61,0x2f65,0x2f69,0x2f6d,0x2f71,0x2f75,0x2f79,0x2f7d,0x2f81,0x2f85,0x2f89,0x2f8d,
+0x2f91,0x2f95,0x2f99,0x2f9d,0x2fa1,0x2fa5,0x2fa9,0x2fad,0x2fb1,0x2fb5,0x2fb9,0x2fbd,0x2fc1,0x2fc5,0x2fc9,0x2fcd,
+0x2fd1,0x2fd5,0x2fd9,0x2fdd,0x2fe1,0x2fe5,0x2d25,0x2fe9,0x2fed,0x2ff1,0x2ff5,0x2ff9,0x2ffd,0x2ffd,0x3001,0x3005,
+0x3009,0x300d,0x3011,0x3015,0x3019,0x301d,0x2f2d,0x3021,0x3025,0x3029,0x302d,0x3031,0x3037,1,1,0x303b,
+0x303f,0x3043,0x3047,0x304b,0x304f,0x3053,0x3057,0x2f65,0x305b,0x305f,0x3063,0x2ef5,0x3067,0x306b,0x306f,0x3073,
+0x3077,0x307b,0x307f,0x3083,0x3087,0x308b,0x308f,0x3093,0x2f89,0x3097,0x2f8d,0x309b,0x309f,0x30a3,0x30a7,0x30ab,
+0x2ef9,0x2b75,0x30af,0x30b3,0x30b7,0x2c8d,0x2de9,0x30bb,0x30bf,0x2fa9,0x30c3,0x2fad,0x30c7,0x30cb,0x30cf,0x2f01,
+0x30d3,0x30d7,0x30db,0x30df,0x30e3,0x2f05,0x30e7,0x30eb,0x30ef,0x30f3,0x30f7,0x30fb,0x2fe5,0x30ff,0x3103,0x2d25,
+0x3107,0x2ff5,0x310b,0x310f,0x3113,0x3117,0x311b,0x3009,0x311f,0x2f29,0x3123,0x300d,0x2c41,0x3127,0x3011,0x312b,
+0x3019,0x312f,0x3133,0x3137,0x313b,0x313f,0x3021,0x2f19,0x3143,0x3025,0x3147,0x3029,0x314b,0x2af1,0x314f,0x3155,
+0x315b,0x3161,0x3165,0x3169,0x316d,0x3173,0x3179,0x317f,0x3183,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x3186,
+0xfe34,0x318c,1,1,1,1,1,1,1,1,1,1,0x3192,0x3198,0x31a0,0x31aa,
+0x31b2,0x31b8,0x31be,0x31c4,0x31ca,0x31d0,0x31d6,0x31dc,0x31e2,1,0x31e8,0x31ee,0x31f4,0x31fa,0x3200,1,
+0x3206,1,0x320c,0x3212,1,0x3218,0x321e,1,0x3224,0x322a,0x3230,0x3236,0x323c,0x3242,0x3248,0x324e,
+0x3254,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,
+0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffb8,1,0xffcc,1,1,1,1,1,1,1,1,0xffcc,0xfe02,0xffb8,
+1,1,1,1,0xfe12,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,
+1,1,1,1,1,0xffb8,0xffb8,0xffcc,0xffcc,0xffcc,0xffb8,0xffcc,0xffb8,0xffb8,0xffb8,1,
+1,1,1,1,1,1,1,1,0xa94,0x295f,0xa9a,0x2969,1,1,1,1,
+1,0xaa0,1,1,1,1,1,0x2973,1,1,1,1,1,1,1,1,
+1,0xfe12,0xfc0e,1,1,1,1,1,1,1,0xfc00,1,1,1,1,1,
+1,0x297d,0x2987,1,0xaa6,0xaac,0xfe12,0xfe12,1,1,1,1,1,1,1,1,
+1,1,1,0xfe12,1,1,1,1,1,1,1,1,1,0xfe0e,1,1,
+1,1,1,0xfe12,0xfe0e,1,1,1,1,1,1,1,1,1,0xfe0e,0xfe12,
+1,1,1,1,1,1,1,1,1,1,1,0xfe0e,0xfe0e,1,0xfc00,1,
+1,1,1,1,1,1,0xab2,1,1,1,0x2991,0x299b,0xfe12,1,1,1,
+1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,1,1,0xfe12,1,1,
+1,0xfe0e,1,1,1,1,1,1,1,1,1,0xfc00,1,1,1,1,
+1,1,1,1,0xabe,0xfc00,0x29a5,0x29af,0xfc00,0x29b9,1,1,0xfe12,0xfe0e,1,1,
+1,1,1,1,1,1,1,1,1,1,0xad0,0xad6,0x29c3,0x29cd,1,1,
+1,0xfe12,0xfe0e,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xfe12,0xfe0e,1,1,1,1,1,1,1,1,0xfc00,1,1,1,
+1,0xadc,1,1,0x29d7,1,1,1,1,0xfe12,0xfe12,1,0xfe02,0xfe02,0xfe02,0xfe02,
+0xfe02,1,1,1,1,1,1,1,1,1,1,1,0xfe0c,0xfe0c,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0xfe02,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0x325a,0x3264,0x3278,0x3290,0x32a8,
+0x32c0,0x32d8,0xffb0,0xffb0,0xfe02,0xfe02,0xfe02,1,1,1,0xffc4,0xffb0,0xffb0,0xffb0,1,1,
+1,1,1,1,1,1,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,1,1,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffb8,0xffb8,1,1,1,1,1,1,1,1,1,1,0xffcc,0xffcc,
+0xffcc,0xffcc,1,1,1,1,1,1,1,1,1,1,1,0x32e6,0x32f0,0x3304,
+0x331c,0x3334,0x334c,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,1,0xffcc,0xffcc,1,0xffcc,0xffcc,
+0xffcc,0xffcc,0xffcc,1,1,1,1,1,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,0xffb8,1,
+1,1,1,1,1,1,1,1,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xffcc,0xfe0e,1,
+1,1,1,1,0x335b,0x335f,0x3363,0x3367,0x336d,0x2f4d,0x3371,0x3375,0x3379,0x337d,0x2f51,0x3381,
+0x3385,0x3389,0x2f55,0x338f,0x3393,0x3397,0x339b,0x33a1,0x33a5,0x33a9,0x33ad,0x33b3,0x33b7,0x33bb,0x33bf,0x303f,
+0x33c3,0x33c9,0x33cd,0x33d1,0x33d5,0x33d9,0x33dd,0x33e1,0x33e5,0x3053,0x2f59,0x2f5d,0x3057,0x33e9,0x33ed,0x2c59,
+0x33f1,0x2f61,0x33f5,0x33f9,0x33fd,0x3401,0x3401,0x3401,0x3405,0x340b,0x340f,0x3413,0x3417,0x341d,0x3421,0x3425,
+0x3429,0x342d,0x3431,0x3435,0x3439,0x343d,0x3441,0x3445,0x3449,0x344d,0x344d,0x305f,0x3451,0x3455,0x3459,0x345d,
+0x2f69,0x3461,0x3465,0x3469,0x2ebd,0x346d,0x3471,0x3475,0x3479,0x347d,0x3481,0x3485,0x3489,0x348d,0x3493,0x3497,
+0x349b,0x349f,0x34a3,0x34a7,0x34ab,0x34b1,0x34b7,0x34bb,0x34bf,0x34c3,0x34c7,0x34cb,0x34cf,0x34d3,0x34d7,0x34d7,
+0x34db,0x34e1,0x34e5,0x2c49,0x34e9,0x34ed,0x34f3,0x34f7,0x34fb,0x34ff,0x3503,0x3507,0x2f7d,0x350b,0x350f,0x3513,
+0x3519,0x351d,0x3523,0x3527,0x352b,0x352f,0x3533,0x3537,0x353b,0x353f,0x3543,0x3547,0x354b,0x354f,0x3555,0x3559,
+0x355d,0x3561,0x2b71,0x3565,0x356b,0x356f,0x356f,0x3575,0x3579,0x3579,0x357d,0x3581,0x3587,0x358d,0x3591,0x3595,
+0x3599,0x359d,0x35a1,0x35a5,0x35a9,0x35ad,0x35b1,0x2f81,0x35b5,0x35bb,0x35bf,0x35c3,0x308f,0x35c3,0x35c7,0x2f89,
+0x35cb,0x35cf,0x35d3,0x35d7,0x2f8d,0x2b05,0x35db,0x35df,0x35e3,0x35e7,0x35eb,0x35ef,0x35f3,0x35f9,0x35fd,0x3601,
+0x3605,0x3609,0x360d,0x3613,0x3617,0x361b,0x361f,0x3623,0x3627,0x362b,0x362f,0x3633,0x2f91,0x3637,0x363b,0x3641,
+0x3645,0x3649,0x364d,0x2f99,0x3651,0x3655,0x3659,0x365d,0x3661,0x3665,0x3669,0x366d,0x2b75,0x30af,0x3671,0x3675,
+0x3679,0x367d,0x3683,0x3687,0x368b,0x368f,0x2f9d,0x3693,0x3699,0x369d,0x36a1,0x3161,0x36a5,0x36a9,0x36ad,0x36b1,
+0x36b5,0x36bb,0x36bf,0x36c3,0x36c7,0x36cd,0x36d1,0x36d5,0x36d9,0x2c8d,0x36dd,0x36e1,0x36e7,0x36ed,0x36f3,0x36f7,
+0x36fd,0x3701,0x3705,0x3709,0x370d,0x2fa1,0x2de9,0x3711,0x3715,0x3719,0x371d,0x3723,0x3727,0x372b,0x372f,0x30bf,
+0x3733,0x3737,0x373d,0x3741,0x3745,0x374b,0x3751,0x3755,0x30c3,0x3759,0x375d,0x3761,0x3765,0x3769,0x376d,0x3771,
+0x3777,0x377b,0x3781,0x3785,0x378b,0x30cb,0x378f,0x3793,0x3799,0x379d,0x37a1,0x37a7,0x37ad,0x37b1,0x37b5,0x37b9,
+0x37bd,0x37bd,0x37c1,0x37c5,0x30d3,0x37c9,0x37cd,0x37d1,0x37d5,0x37d9,0x37df,0x37e3,0x2c55,0x37e9,0x37ef,0x37f3,
+0x37f9,0x37ff,0x3805,0x3809,0x30eb,0x380d,0x3813,0x3819,0x381f,0x3825,0x3829,0x3829,0x30ef,0x3169,0x382d,0x3831,
+0x3835,0x3839,0x383f,0x2bbd,0x30f7,0x3843,0x3847,0x2fcd,0x384d,0x3853,0x2f15,0x3859,0x385d,0x2fdd,0x3861,0x3865,
+0x3869,0x386f,0x386f,0x3875,0x3879,0x387d,0x3883,0x3887,0x388b,0x388f,0x3895,0x3899,0x389d,0x38a1,0x38a5,0x38a9,
+0x38af,0x38b3,0x38b7,0x38bb,0x38bf,0x38c3,0x38c7,0x38cd,0x38d3,0x38d7,0x38dd,0x38e1,0x38e7,0x38eb,0x2ff5,0x38ef,
+0x38f5,0x38fb,0x38ff,0x3905,0x3909,0x390f,0x3913,0x3917,0x391b,0x391f,0x3923,0x3927,0x392d,0x3933,0x3939,0x3575,
+0x393f,0x3943,0x3947,0x394b,0x394f,0x3953,0x3957,0x395b,0x395f,0x3963,0x3967,0x396b,0x2c9d,0x3971,0x3975,0x3979,
+0x397d,0x3981,0x3985,0x3001,0x3989,0x398d,0x3991,0x3995,0x3999,0x399f,0x39a5,0x39ab,0x39af,0x39b3,0x39b7,0x39bb,
+0x39c1,0x39c5,0x39cb,0x39cf,0x39d3,0x39d9,0x39df,0x39e3,0x2ba9,0x39e7,0x39eb,0x39ef,0x39f3,0x39f7,0x39fb,0x3113,
+0x39ff,0x3a03,0x3a07,0x3a0b,0x3a0f,0x3a13,0x3a17,0x3a1b,0x3a1f,0x3a23,0x3a29,0x3a2d,0x3a31,0x3a35,0x3a39,0x3a3d,
+0x3a43,0x3a49,0x3a4d,0x3a51,0x3127,0x312b,0x3a55,0x3a59,0x3a5f,0x3a63,0x3a67,0x3a6b,0x3a6f,0x3a75,0x3a7b,0x3a7f,
+0x3a83,0x3a87,0x3a8d,0x312f,0x3a91,0x3a97,0x3a9d,0x3aa1,0x3aa5,0x3aa9,0x3aaf,0x3ab3,0x3ab7,0x3abb,0x3abf,0x3ac3,
+0x3ac7,0x3acb,0x3ad1,0x3ad5,0x3ad9,0x3add,0x3ae3,0x3ae7,0x3aeb,0x3aef,0x3af3,0x3af9,0x3aff,0x3b03,0x3b07,0x3b0b,
+0x3b11,0x3b15,0x3147,0x3147,0x3b1b,0x3b1f,0x3b25,0x3b29,0x3b2d,0x3b31,0x3b35,0x3b39,0x3b3d,0x3b41,0x314b,0x3b47,
+0x3b4b,0x3b4f,0x3b53,0x3b57,0x3b5b,0x3b61,0x3b65,0x3b6b,0x3b71,0x3b77,0x3b7b,0x3b7f,0x3b83,0x3b87,0x3b8b,0x3b8f,
+0x3b93,0x3b97,1,1
+};
+
+static const UCPTrie norm2_nfc_data_trie={
+    norm2_nfc_data_trieIndex,
+    { norm2_nfc_data_trieData },
+    1746, 7892,
+    0x2fc00, 0x30,
+    0, 0,
+    0, 0,
+    0xc4, 0x226,
+    0x1,
+};
+
+static const uint16_t norm2_nfc_data_extraData[7732]={
+0xffff,0xffff,0x8670,0x44dc,0x8670,0x44c0,0x8670,0x44de,0x600,0x180,0x602,0x182,0x604,0x185,0x606,0x186,
+0x608,0x200,0x60c,0x205,0x60e,0x44d,0x610,0x189,0x612,0x3d44,0x614,0x18b,0x618,0x39a,0x61e,0x400,
+0x622,0x404,0x646,0x3d41,0x64a,0x3c00,0x8650,0x208,0x60e,0x3c04,0x646,0x3c08,0x8662,0x3c0c,0x602,0x20c,
+0x604,0x210,0x60e,0x214,0x618,0x218,0x864e,0x18f,0x60e,0x3c14,0x618,0x21c,0x646,0x3c18,0x64e,0x3c20,
+0x65a,0x3c24,0x8662,0x3c1c,0x600,0x190,0x602,0x192,0x604,0x195,0x606,0x3d78,0x608,0x225,0x60c,0x228,
+0x60e,0x22c,0x610,0x196,0x612,0x3d74,0x618,0x234,0x61e,0x408,0x622,0x40c,0x646,0x3d71,0x64e,0x451,
+0x650,0x230,0x65a,0x3c30,0x8660,0x3c34,0x860e,0x3c3c,0x602,0x3e8,0x604,0x238,0x608,0x3c40,0x60c,0x23c,
+0x60e,0x240,0x618,0x3cc,0x864e,0x244,0x604,0x248,0x60e,0x3c44,0x610,0x3c4c,0x618,0x43c,0x646,0x3c48,
+0x64e,0x3c50,0x865c,0x3c54,0x600,0x198,0x602,0x19a,0x604,0x19c,0x606,0x250,0x608,0x254,0x60c,0x258,
+0x60e,0x260,0x610,0x19f,0x612,0x3d90,0x618,0x39e,0x61e,0x410,0x622,0x414,0x646,0x3d94,0x650,0x25c,
+0x8660,0x3c58,0x8604,0x268,0x602,0x3c60,0x618,0x3d0,0x646,0x3c64,0x64e,0x26c,0x8662,0x3c68,0x602,0x272,
+0x618,0x27a,0x646,0x3c6d,0x64e,0x276,0x65a,0x3c78,0x8662,0x3c74,0x602,0x3c7c,0x60e,0x3c80,0x8646,0x3c84,
+0x600,0x3f0,0x602,0x286,0x606,0x1a2,0x60e,0x3c88,0x618,0x28e,0x646,0x3c8c,0x64e,0x28a,0x65a,0x3c94,
+0x8662,0x3c90,0x600,0x1a4,0x602,0x1a6,0x604,0x1a9,0x606,0x1ab,0x608,0x299,0x60c,0x29c,0x60e,0x45d,
+0x610,0x1ad,0x612,0x3d9c,0x616,0x2a0,0x618,0x3a2,0x61e,0x418,0x622,0x41c,0x636,0x341,0x646,0x3d99,
+0x8650,0x3d5,0x602,0x3ca8,0x860e,0x3cac,0x602,0x2a8,0x60e,0x3cb0,0x618,0x2b0,0x61e,0x420,0x622,0x424,
+0x646,0x3cb5,0x64e,0x2ac,0x8662,0x3cbc,0x602,0x2b5,0x604,0x2b8,0x60e,0x3cc0,0x618,0x2c1,0x646,0x3cc5,
+0x64c,0x430,0x864e,0x2bc,0x60e,0x3cd4,0x618,0x2c8,0x646,0x3cd8,0x64c,0x434,0x64e,0x2c4,0x65a,0x3ce0,
+0x8662,0x3cdc,0x600,0x1b2,0x602,0x1b4,0x604,0x1b6,0x606,0x2d1,0x608,0x2d5,0x60c,0x2d8,0x610,0x1b9,
+0x612,0x3dcc,0x614,0x2dc,0x616,0x2e0,0x618,0x3a6,0x61e,0x428,0x622,0x42c,0x636,0x35f,0x646,0x3dc8,
+0x648,0x3ce4,0x650,0x2e4,0x65a,0x3cec,0x8660,0x3ce8,0x606,0x3cf8,0x8646,0x3cfc,0x600,0x3d00,0x602,0x3d04,
+0x604,0x2e8,0x60e,0x3d0c,0x610,0x3d08,0x8646,0x3d10,0x60e,0x3d14,0x8610,0x3d18,0x600,0x3de4,0x602,0x1ba,
+0x604,0x2ec,0x606,0x3df0,0x608,0x464,0x60e,0x3d1c,0x610,0x2f0,0x612,0x3dec,0x8646,0x3de8,0x602,0x2f2,
+0x604,0x3d20,0x60e,0x2f6,0x618,0x2fa,0x646,0x3d24,0x8662,0x3d28,0x600,0x1c0,0x602,0x1c2,0x604,0x1c5,
+0x606,0x1c6,0x608,0x202,0x60c,0x207,0x60e,0x44f,0x610,0x1c9,0x612,0x3d46,0x614,0x1cb,0x618,0x39c,
+0x61e,0x402,0x622,0x406,0x646,0x3d43,0x64a,0x3c02,0x8650,0x20a,0x60e,0x3c06,0x646,0x3c0a,0x8662,0x3c0e,
+0x602,0x20e,0x604,0x212,0x60e,0x216,0x618,0x21a,0x864e,0x1cf,0x60e,0x3c16,0x618,0x21e,0x646,0x3c1a,
+0x64e,0x3c22,0x65a,0x3c26,0x8662,0x3c1e,0x600,0x1d0,0x602,0x1d2,0x604,0x1d5,0x606,0x3d7a,0x608,0x227,
+0x60c,0x22a,0x60e,0x22e,0x610,0x1d6,0x612,0x3d76,0x618,0x236,0x61e,0x40a,0x622,0x40e,0x646,0x3d73,
+0x64e,0x453,0x650,0x232,0x65a,0x3c32,0x8660,0x3c36,0x860e,0x3c3e,0x602,0x3ea,0x604,0x23a,0x608,0x3c42,
+0x60c,0x23e,0x60e,0x242,0x618,0x3ce,0x864e,0x246,0x604,0x24a,0x60e,0x3c46,0x610,0x3c4e,0x618,0x43e,
+0x646,0x3c4a,0x64e,0x3c52,0x65c,0x3c56,0x8662,0x3d2c,0x600,0x1d8,0x602,0x1da,0x604,0x1dc,0x606,0x252,
+0x608,0x256,0x60c,0x25a,0x610,0x1df,0x612,0x3d92,0x618,0x3a0,0x61e,0x412,0x622,0x416,0x646,0x3d96,
+0x650,0x25e,0x8660,0x3c5a,0x604,0x26a,0x8618,0x3e0,0x602,0x3c62,0x618,0x3d2,0x646,0x3c66,0x64e,0x26e,
+0x8662,0x3c6a,0x602,0x274,0x618,0x27c,0x646,0x3c6f,0x64e,0x278,0x65a,0x3c7a,0x8662,0x3c76,0x602,0x3c7e,
+0x60e,0x3c82,0x8646,0x3c86,0x600,0x3f2,0x602,0x288,0x606,0x1e2,0x60e,0x3c8a,0x618,0x290,0x646,0x3c8e,
+0x64e,0x28c,0x65a,0x3c96,0x8662,0x3c92,0x600,0x1e4,0x602,0x1e6,0x604,0x1e9,0x606,0x1eb,0x608,0x29b,
+0x60c,0x29e,0x60e,0x45f,0x610,0x1ed,0x612,0x3d9e,0x616,0x2a2,0x618,0x3a4,0x61e,0x41a,0x622,0x41e,
+0x636,0x343,0x646,0x3d9b,0x8650,0x3d7,0x602,0x3caa,0x860e,0x3cae,0x602,0x2aa,0x60e,0x3cb2,0x618,0x2b2,
+0x61e,0x422,0x622,0x426,0x646,0x3cb7,0x64e,0x2ae,0x8662,0x3cbe,0x602,0x2b7,0x604,0x2ba,0x60e,0x3cc2,
+0x618,0x2c3,0x646,0x3cc7,0x64c,0x432,0x864e,0x2be,0x60e,0x3cd6,0x610,0x3d2e,0x618,0x2ca,0x646,0x3cda,
+0x64c,0x436,0x64e,0x2c6,0x65a,0x3ce2,0x8662,0x3cde,0x600,0x1f2,0x602,0x1f4,0x604,0x1f6,0x606,0x2d3,
+0x608,0x2d7,0x60c,0x2da,0x610,0x1f9,0x612,0x3dce,0x614,0x2de,0x616,0x2e2,0x618,0x3a8,0x61e,0x42a,
+0x622,0x42e,0x636,0x361,0x646,0x3dca,0x648,0x3ce6,0x650,0x2e6,0x65a,0x3cee,0x8660,0x3cea,0x606,0x3cfa,
+0x8646,0x3cfe,0x600,0x3d02,0x602,0x3d06,0x604,0x2ea,0x60e,0x3d0e,0x610,0x3d0a,0x614,0x3d30,0x8646,0x3d12,
+0x60e,0x3d16,0x8610,0x3d1a,0x600,0x3de6,0x602,0x1fa,0x604,0x2ee,0x606,0x3df2,0x608,0x466,0x60e,0x3d1e,
+0x610,0x1fe,0x612,0x3dee,0x614,0x3d32,0x8646,0x3dea,0x602,0x2f4,0x604,0x3d22,0x60e,0x2f8,0x618,0x2fc,
+0x646,0x3d26,0x8662,0x3d2a,0x600,0x3fda,0x602,0x70a,0x8684,0x3f82,0x602,0x3f8,0x8608,0x3c4,0x8602,0x3fc,
+0x602,0x3fa,0x8608,0x3c6,0x8602,0x3fe,0x860e,0x3d36,0x8618,0x3dc,0x8618,0x3de,0x600,0x3f74,0x602,0x70c,
+0x608,0x3f72,0x60c,0x3f70,0x626,0x3e11,0x628,0x3e13,0x868a,0x3f78,0x600,0x3f90,0x602,0x710,0x626,0x3e31,
+0x8628,0x3e33,0x600,0x3f94,0x602,0x712,0x626,0x3e51,0x628,0x3e53,0x868a,0x3f98,0x600,0x3fb4,0x602,0x714,
+0x608,0x3fb2,0x60c,0x3fb0,0x610,0x754,0x626,0x3e71,0x8628,0x3e73,0x600,0x3ff0,0x602,0x718,0x626,0x3e91,
+0x8628,0x3e93,0x8628,0x3fd8,0x600,0x3fd4,0x602,0x71c,0x608,0x3fd2,0x60c,0x3fd0,0x610,0x756,0x8628,0x3eb3,
+0x600,0x3ff4,0x602,0x71e,0x626,0x3ed1,0x628,0x3ed3,0x868a,0x3ff8,0x600,0x3ee1,0x602,0x759,0x608,0x3f62,
+0x60c,0x3f60,0x626,0x3e01,0x628,0x3e03,0x684,0x3f6d,0x868a,0x3f66,0x600,0x3ee4,0x602,0x75a,0x626,0x3e21,
+0x8628,0x3e23,0x600,0x3ee9,0x602,0x75d,0x626,0x3e41,0x628,0x3e43,0x684,0x3f8d,0x868a,0x3f86,0x600,0x3eec,
+0x602,0x75e,0x608,0x3fa2,0x60c,0x3fa0,0x610,0x795,0x626,0x3e61,0x628,0x3e63,0x8684,0x3fac,0x600,0x3ef0,
+0x602,0x798,0x626,0x3e81,0x8628,0x3e83,0x626,0x3fc8,0x8628,0x3fca,0x600,0x3ef4,0x602,0x79a,0x608,0x3fc2,
+0x60c,0x3fc0,0x610,0x797,0x626,0x3ea1,0x628,0x3ea3,0x8684,0x3fcc,0x600,0x3ef9,0x602,0x79d,0x626,0x3ec1,
+0x628,0x3ec3,0x684,0x3fed,0x868a,0x3fe6,0x602,0x7a6,0x8610,0x7a8,0x8610,0x80e,0x60c,0x9a0,0x8610,0x9a4,
+0x8602,0x806,0x600,0x800,0x60c,0x9ac,0x8610,0x802,0x60c,0x982,0x8610,0x9b8,0x8610,0x9bc,0x600,0x81a,
+0x608,0x9c4,0x60c,0x832,0x8610,0x9c8,0x8602,0x818,0x8610,0x9cc,0x608,0x9dc,0x60c,0x81c,0x610,0x9e0,
+0x8616,0x9e4,0x8610,0x9e8,0x8610,0x9f0,0x8610,0x9d8,0x60c,0x9a2,0x8610,0x9a6,0x8602,0x8a6,0x600,0x8a0,
+0x60c,0x9ae,0x8610,0x8a2,0x60c,0x984,0x8610,0x9ba,0x8610,0x9be,0x600,0x8ba,0x608,0x9c6,0x60c,0x872,
+0x8610,0x9ca,0x8602,0x8b8,0x8610,0x9ce,0x608,0x9de,0x60c,0x8bc,0x610,0x9e2,0x8616,0x9e6,0x8610,0x9ea,
+0x8610,0x9f2,0x8610,0x9da,0x8610,0x8ae,0x861e,0x8ec,0x861e,0x8ee,0x8610,0x9b4,0x8610,0x9b6,0x8610,0x9d4,
+0x8610,0x9d6,0xca6,0xc44,0xca8,0xc46,0x8caa,0xc4a,0x8ca8,0xc48,0x8ca8,0xc4c,0x8ca8,0xd84,0x8ca8,0xda6,
+0x8ca8,0xd80,0x9278,0x1252,0x9278,0x1262,0x9278,0x1268,0x137c,0x1396,0x93ae,0x1398,0x167c,0x1696,0x16ac,0x1690,
+0x96ae,0x1698,0x97ae,0x1728,0x177c,0x1794,0x97ae,0x1798,0x977c,0x1796,0x98ac,0x1890,0x99aa,0x1980,0x1984,0x1995,
+0x19aa,0x198e,0x99ac,0x1990,0x1a7c,0x1a94,0x9aae,0x1a98,0x9a7c,0x1a96,0x1b94,0x1bb4,0x1b9e,0x1bb9,0x9bbe,0x1bbc,
+0xa05c,0x204c,0xb66a,0x360c,0xb66a,0x3610,0xb66a,0x3614,0xb66a,0x3618,0xb66a,0x361c,0xb66a,0x3624,0xb66a,0x3676,
+0xb66a,0x367a,0xb66a,0x3680,0xb66a,0x3682,0xb66a,0x3686,0x600,0x3f9a,0x602,0x3f9c,0x8684,0x3f9e,0x600,0x3fba,
+0x602,0x3fbc,0x8684,0x3fbe,0x8670,0x4334,0x8670,0x4336,0x8670,0x435c,0x8670,0x439a,0x8670,0x439e,0x8670,0x439c,
+0x8670,0x4408,0x8670,0x4412,0x8670,0x4418,0x8670,0x4448,0x8670,0x444c,0x8670,0x4482,0x8670,0x4488,0x8670,0x448e,
+0x8670,0x4492,0x8670,0x44da,0x8670,0x44c4,0x8670,0x44e0,0x8670,0x44e2,0x8670,0x44e8,0x8670,0x44ea,0x8670,0x44f0,
+0x8670,0x44f2,0x8670,0x4500,0x8670,0x4502,0x8670,0x45c0,0x8670,0x45c2,0x8670,0x4508,0x8670,0x450a,0x8670,0x4510,
+0x8670,0x4512,0x8670,0x45c4,0x8670,0x45c6,0x8670,0x4558,0x8670,0x455a,0x8670,0x455c,0x8670,0x455e,0x8670,0x45d4,
+0x8670,0x45d6,0x8670,0x45d8,0x8670,0x45da,0xe132,0x6128,0xe132,0x6098,0xe132,0x609c,0xe132,0x60a0,0xe132,0x60a4,
+0xe132,0x60a8,0xe132,0x60ac,0xe132,0x60b0,0xe132,0x60b4,0xe132,0x60b8,0xe132,0x60bc,0xe132,0x60c0,0xe132,0x60c4,
+0xe132,0x60ca,0xe132,0x60ce,0xe132,0x60d2,0x6132,0x60e0,0xe134,0x60e2,0x6132,0x60e6,0xe134,0x60e8,0x6132,0x60ec,
+0xe134,0x60ee,0x6132,0x60f2,0xe134,0x60f4,0x6132,0x60f8,0xe134,0x60fa,0xe132,0x613c,0xe132,0x61e8,0xe132,0x6158,
+0xe132,0x615c,0xe132,0x6160,0xe132,0x6164,0xe132,0x6168,0xe132,0x616c,0xe132,0x6170,0xe132,0x6174,0xe132,0x6178,
+0xe132,0x617c,0xe132,0x6180,0xe132,0x6184,0xe132,0x618a,0xe132,0x618e,0xe132,0x6192,0x6132,0x61a0,0xe134,0x61a2,
+0x6132,0x61a6,0xe134,0x61a8,0x6132,0x61ac,0xe134,0x61ae,0x6132,0x61b2,0xe134,0x61b4,0x6132,0x61b8,0xe134,0x61ba,
+0xe132,0x61ee,0xe132,0x61f0,0xe132,0x61f2,0xe132,0x61f4,0xe132,0x61fc,0xb489,0x2e82,0x2134,0xb489,0x2e82,0x2138,
+0xb489,0x2e82,0x2156,0xb489,0x49c2,0x225c,0xb489,0x49c2,0x225e,0x3489,0xcf82,0x2696,0xb489,0xd5c2,0x2698,0x348b,
+0x2c02,0x2978,0x348b,0x2e82,0x2976,0xb48b,0x2f42,0x297c,0xb48b,0x6bc2,0x2b74,0xb48b,0x6bc2,0x2b76,0xb48d,0x4c02,
+0x3270,2,0xe602,0x41,0x302,0x600,0x3d4c,0x602,0x3d48,0x606,0x3d54,0x8612,0x3d50,0xe602,0x41,0x308,
+0x8608,0x3bc,0xe602,0x41,0x30a,0x8602,0x3f4,0xca02,0x43,0x327,0x8602,0x3c10,0xe602,0x45,0x302,0x600,
+0x3d80,0x602,0x3d7c,0x606,0x3d88,0x8612,0x3d84,0xe602,0x49,0x308,0x8602,0x3c5c,0xe602,0x4f,0x302,0x600,
+0x3da4,0x602,0x3da0,0x606,0x3dac,0x8612,0x3da8,0xe602,0x4f,0x303,0x602,0x3c98,0x608,0x458,0x8610,0x3c9c,
+0xe602,0x4f,0x308,0x8608,0x454,0xe602,0x55,0x308,0x600,0x3b6,0x602,0x3ae,0x608,0x3aa,0x8618,0x3b2,
+0xe602,0x61,0x302,0x600,0x3d4e,0x602,0x3d4a,0x606,0x3d56,0x8612,0x3d52,0xe602,0x61,0x308,0x8608,0x3be,
+0xe602,0x61,0x30a,0x8602,0x3f6,0xca02,0x63,0x327,0x8602,0x3c12,0xe602,0x65,0x302,0x600,0x3d82,0x602,
+0x3d7e,0x606,0x3d8a,0x8612,0x3d86,0xe602,0x69,0x308,0x8602,0x3c5e,0xe602,0x6f,0x302,0x600,0x3da6,0x602,
+0x3da2,0x606,0x3dae,0x8612,0x3daa,0xe602,0x6f,0x303,0x602,0x3c9a,0x608,0x45a,0x8610,0x3c9e,0xe602,0x6f,
+0x308,0x8608,0x456,0xe602,0x75,0x308,0x600,0x3b8,0x602,0x3b0,0x608,0x3ac,0x8618,0x3b4,0xe602,0x41,
+0x306,0x600,0x3d60,0x602,0x3d5c,0x606,0x3d68,0x8612,0x3d64,0xe602,0x61,0x306,0x600,0x3d62,0x602,0x3d5e,
+0x606,0x3d6a,0x8612,0x3d66,0xe602,0x45,0x304,0x600,0x3c28,0x8602,0x3c2c,0xe602,0x65,0x304,0x600,0x3c2a,
+0x8602,0x3c2e,0xe602,0x4f,0x304,0x600,0x3ca0,0x8602,0x3ca4,0xe602,0x6f,0x304,0x600,0x3ca2,0x8602,0x3ca6,
+0xe602,0x53,0x301,0x860e,0x3cc8,0xe602,0x73,0x301,0x860e,0x3cca,0xe602,0x53,0x30c,0x860e,0x3ccc,0xe602,
+0x73,0x30c,0x860e,0x3cce,0xe602,0x55,0x303,0x8602,0x3cf0,0xe602,0x75,0x303,0x8602,0x3cf2,0xe602,0x55,
+0x304,0x8610,0x3cf4,0xe602,0x75,0x304,0x8610,0x3cf6,0xd802,0x4f,0x31b,0x600,0x3db8,0x602,0x3db4,0x606,
+0x3dc0,0x612,0x3dbc,0x8646,0x3dc4,0xd802,0x6f,0x31b,0x600,0x3dba,0x602,0x3db6,0x606,0x3dc2,0x612,0x3dbe,
+0x8646,0x3dc6,0xd802,0x55,0x31b,0x600,0x3dd4,0x602,0x3dd0,0x606,0x3ddc,0x612,0x3dd8,0x8646,0x3de0,0xd802,
+0x75,0x31b,0x600,0x3dd6,0x602,0x3dd2,0x606,0x3dde,0x612,0x3dda,0x8646,0x3de2,0xca02,0x4f,0x328,0x8608,
+0x3d8,0xca02,0x6f,0x328,0x8608,0x3da,0xe602,0x41,0x307,0x8608,0x3c0,0xe602,0x61,0x307,0x8608,0x3c2,
+0xca02,0x45,0x327,0x860c,0x3c38,0xca02,0x65,0x327,0x860c,0x3c3a,0xe602,0x4f,0x307,0x8608,0x460,0xe602,
+0x6f,0x307,0x8608,0x462,0xe602,0x3b1,0x301,0x868a,0x3f68,0xe602,0x3b7,0x301,0x868a,0x3f88,0xe602,0x3b9,
+0x308,0x600,0x3fa4,0x602,0x720,0x8684,0x3fae,0xe602,0x3c5,0x308,0x600,0x3fc4,0x602,0x760,0x8684,0x3fce,
+0xe602,0x3c9,0x301,0x868a,0x3fe8,2,0xcc6,0xcc2,0x99aa,0x1996,2,0xdd9,0xdcf,0x9b94,0x1bba,0xdc02,
+0x4c,0x323,0x8608,0x3c70,0xdc02,0x6c,0x323,0x8608,0x3c72,0xdc02,0x52,0x323,0x8608,0x3cb8,0xdc02,0x72,
+0x323,0x8608,0x3cba,0xdc02,0x53,0x323,0x860e,0x3cd0,0xdc02,0x73,0x323,0x860e,0x3cd2,0xdc02,0x41,0x323,
+0x604,0x3d58,0x860c,0x3d6c,0xdc02,0x61,0x323,0x604,0x3d5a,0x860c,0x3d6e,0xdc02,0x45,0x323,0x8604,0x3d8c,
+0xdc02,0x65,0x323,0x8604,0x3d8e,0xdc02,0x4f,0x323,0x8604,0x3db0,0xdc02,0x6f,0x323,0x8604,0x3db2,0xe602,
+0x3b1,0x313,0x600,0x3e05,0x602,0x3e09,0x684,0x3e0d,0x868a,0x3f00,0xe602,0x3b1,0x314,0x600,0x3e07,0x602,
+0x3e0b,0x684,0x3e0f,0x868a,0x3f02,0x1f00,0xe643,0x3b1,0x313,0x300,0x868a,0x3f04,0x1f01,0xe643,0x3b1,0x314,
+0x300,0x868a,0x3f06,0x1f00,0xe643,0x3b1,0x313,0x301,0x868a,0x3f08,0x1f01,0xe643,0x3b1,0x314,0x301,0x868a,
+0x3f0a,0x1f00,0xe643,0x3b1,0x313,0x342,0x868a,0x3f0c,0x1f01,0xe643,0x3b1,0x314,0x342,0x868a,0x3f0e,0xe602,
+0x391,0x313,0x600,0x3e15,0x602,0x3e19,0x684,0x3e1d,0x868a,0x3f10,0xe602,0x391,0x314,0x600,0x3e17,0x602,
+0x3e1b,0x684,0x3e1f,0x868a,0x3f12,0x1f08,0xe643,0x391,0x313,0x300,0x868a,0x3f14,0x1f09,0xe643,0x391,0x314,
+0x300,0x868a,0x3f16,0x1f08,0xe643,0x391,0x313,0x301,0x868a,0x3f18,0x1f09,0xe643,0x391,0x314,0x301,0x868a,
+0x3f1a,0x1f08,0xe643,0x391,0x313,0x342,0x868a,0x3f1c,0x1f09,0xe643,0x391,0x314,0x342,0x868a,0x3f1e,0xe602,
+0x3b5,0x313,0x600,0x3e24,0x8602,0x3e28,0xe602,0x3b5,0x314,0x600,0x3e26,0x8602,0x3e2a,0xe602,0x395,0x313,
+0x600,0x3e34,0x8602,0x3e38,0xe602,0x395,0x314,0x600,0x3e36,0x8602,0x3e3a,0xe602,0x3b7,0x313,0x600,0x3e45,
+0x602,0x3e49,0x684,0x3e4d,0x868a,0x3f20,0xe602,0x3b7,0x314,0x600,0x3e47,0x602,0x3e4b,0x684,0x3e4f,0x868a,
+0x3f22,0x1f20,0xe643,0x3b7,0x313,0x300,0x868a,0x3f24,0x1f21,0xe643,0x3b7,0x314,0x300,0x868a,0x3f26,0x1f20,
+0xe643,0x3b7,0x313,0x301,0x868a,0x3f28,0x1f21,0xe643,0x3b7,0x314,0x301,0x868a,0x3f2a,0x1f20,0xe643,0x3b7,
+0x313,0x342,0x868a,0x3f2c,0x1f21,0xe643,0x3b7,0x314,0x342,0x868a,0x3f2e,0xe602,0x397,0x313,0x600,0x3e55,
+0x602,0x3e59,0x684,0x3e5d,0x868a,0x3f30,0xe602,0x397,0x314,0x600,0x3e57,0x602,0x3e5b,0x684,0x3e5f,0x868a,
+0x3f32,0x1f28,0xe643,0x397,0x313,0x300,0x868a,0x3f34,0x1f29,0xe643,0x397,0x314,0x300,0x868a,0x3f36,0x1f28,
+0xe643,0x397,0x313,0x301,0x868a,0x3f38,0x1f29,0xe643,0x397,0x314,0x301,0x868a,0x3f3a,0x1f28,0xe643,0x397,
+0x313,0x342,0x868a,0x3f3c,0x1f29,0xe643,0x397,0x314,0x342,0x868a,0x3f3e,0xe602,0x3b9,0x313,0x600,0x3e64,
+0x602,0x3e68,0x8684,0x3e6c,0xe602,0x3b9,0x314,0x600,0x3e66,0x602,0x3e6a,0x8684,0x3e6e,0xe602,0x399,0x313,
+0x600,0x3e74,0x602,0x3e78,0x8684,0x3e7c,0xe602,0x399,0x314,0x600,0x3e76,0x602,0x3e7a,0x8684,0x3e7e,0xe602,
+0x3bf,0x313,0x600,0x3e84,0x8602,0x3e88,0xe602,0x3bf,0x314,0x600,0x3e86,0x8602,0x3e8a,0xe602,0x39f,0x313,
+0x600,0x3e94,0x8602,0x3e98,0xe602,0x39f,0x314,0x600,0x3e96,0x8602,0x3e9a,0xe602,0x3c5,0x313,0x600,0x3ea4,
+0x602,0x3ea8,0x8684,0x3eac,0xe602,0x3c5,0x314,0x600,0x3ea6,0x602,0x3eaa,0x8684,0x3eae,0xe602,0x3a5,0x314,
+0x600,0x3eb6,0x602,0x3eba,0x8684,0x3ebe,0xe602,0x3c9,0x313,0x600,0x3ec5,0x602,0x3ec9,0x684,0x3ecd,0x868a,
+0x3f40,0xe602,0x3c9,0x314,0x600,0x3ec7,0x602,0x3ecb,0x684,0x3ecf,0x868a,0x3f42,0x1f60,0xe643,0x3c9,0x313,
+0x300,0x868a,0x3f44,0x1f61,0xe643,0x3c9,0x314,0x300,0x868a,0x3f46,0x1f60,0xe643,0x3c9,0x313,0x301,0x868a,
+0x3f48,0x1f61,0xe643,0x3c9,0x314,0x301,0x868a,0x3f4a,0x1f60,0xe643,0x3c9,0x313,0x342,0x868a,0x3f4c,0x1f61,
+0xe643,0x3c9,0x314,0x342,0x868a,0x3f4e,0xe602,0x3a9,0x313,0x600,0x3ed5,0x602,0x3ed9,0x684,0x3edd,0x868a,
+0x3f50,0xe602,0x3a9,0x314,0x600,0x3ed7,0x602,0x3edb,0x684,0x3edf,0x868a,0x3f52,0x1f68,0xe643,0x3a9,0x313,
+0x300,0x868a,0x3f54,0x1f69,0xe643,0x3a9,0x314,0x300,0x868a,0x3f56,0x1f68,0xe643,0x3a9,0x313,0x301,0x868a,
+0x3f58,0x1f69,0xe643,0x3a9,0x314,0x301,0x868a,0x3f5a,0x1f68,0xe643,0x3a9,0x313,0x342,0x868a,0x3f5c,0x1f69,
+0xe643,0x3a9,0x314,0x342,0x868a,0x3f5e,0xe602,0x3b1,0x300,0x868a,0x3f64,0xe602,0x3b7,0x300,0x868a,0x3f84,
+0xe602,0x3c9,0x300,0x868a,0x3fe4,0xe602,0x3b1,0x342,0x868a,0x3f6e,0xe602,0x3b7,0x342,0x868a,0x3f8e,0xe602,
+0x3c9,0x342,0x868a,0x3fee,3,0xe602,0x41,0x300,0xe602,0x41,0x301,0xe602,0x41,0x303,0xe602,0x45,
+0x300,0xe602,0x45,0x301,0xe602,0x45,0x308,0xe602,0x49,0x300,0xe602,0x49,0x301,0xe602,0x49,0x302,
+0xe602,0x4e,0x303,0xe602,0x4f,0x300,0xe602,0x4f,0x301,0xe602,0x55,0x300,0xe602,0x55,0x301,0xe602,
+0x55,0x302,0xe602,0x59,0x301,0xe602,0x61,0x300,0xe602,0x61,0x301,0xe602,0x61,0x303,0xe602,0x65,
+0x300,0xe602,0x65,0x301,0xe602,0x65,0x308,0xe602,0x69,0x300,0xe602,0x69,0x301,0xe602,0x69,0x302,
+0xe602,0x6e,0x303,0xe602,0x6f,0x300,0xe602,0x6f,0x301,0xe602,0x75,0x300,0xe602,0x75,0x301,0xe602,
+0x75,0x302,0xe602,0x79,0x301,0xe602,0x79,0x308,0xe602,0x41,0x304,0xe602,0x61,0x304,0xca02,0x41,
+0x328,0xca02,0x61,0x328,0xe602,0x43,0x301,0xe602,0x63,0x301,0xe602,0x43,0x302,0xe602,0x63,0x302,
+0xe602,0x43,0x307,0xe602,0x63,0x307,0xe602,0x43,0x30c,0xe602,0x63,0x30c,0xe602,0x44,0x30c,0xe602,
+0x64,0x30c,0xe602,0x45,0x306,0xe602,0x65,0x306,0xe602,0x45,0x307,0xe602,0x65,0x307,0xca02,0x45,
+0x328,0xca02,0x65,0x328,0xe602,0x45,0x30c,0xe602,0x65,0x30c,0xe602,0x47,0x302,0xe602,0x67,0x302,
+0xe602,0x47,0x306,0xe602,0x67,0x306,0xe602,0x47,0x307,0xe602,0x67,0x307,0xca02,0x47,0x327,0xca02,
+0x67,0x327,0xe602,0x48,0x302,0xe602,0x68,0x302,0xe602,0x49,0x303,0xe602,0x69,0x303,0xe602,0x49,
+0x304,0xe602,0x69,0x304,0xe602,0x49,0x306,0xe602,0x69,0x306,0xca02,0x49,0x328,0xca02,0x69,0x328,
+0xe602,0x49,0x307,0xe602,0x4a,0x302,0xe602,0x6a,0x302,0xca02,0x4b,0x327,0xca02,0x6b,0x327,0xe602,
+0x4c,0x301,0xe602,0x6c,0x301,0xca02,0x4c,0x327,0xca02,0x6c,0x327,0xe602,0x4c,0x30c,0xe602,0x6c,
+0x30c,0xe602,0x4e,0x301,0xe602,0x6e,0x301,0xca02,0x4e,0x327,0xca02,0x6e,0x327,0xe602,0x4e,0x30c,
+0xe602,0x6e,0x30c,0xe602,0x4f,0x306,0xe602,0x6f,0x306,0xe602,0x4f,0x30b,0xe602,0x6f,0x30b,0xe602,
+0x52,0x301,0xe602,0x72,0x301,0xca02,0x52,0x327,0xca02,0x72,0x327,0xe602,0x52,0x30c,0xe602,0x72,
+0x30c,0xe602,0x53,0x302,0xe602,0x73,0x302,0xca02,0x53,0x327,0xca02,0x73,0x327,0xca02,0x54,0x327,
+0xca02,0x74,0x327,0xe602,0x54,0x30c,0xe602,0x74,0x30c,0xe602,0x55,0x306,0xe602,0x75,0x306,0xe602,
+0x55,0x30a,0xe602,0x75,0x30a,0xe602,0x55,0x30b,0xe602,0x75,0x30b,0xca02,0x55,0x328,0xca02,0x75,
+0x328,0xe602,0x57,0x302,0xe602,0x77,0x302,0xe602,0x59,0x302,0xe602,0x79,0x302,0xe602,0x59,0x308,
+0xe602,0x5a,0x301,0xe602,0x7a,0x301,0xe602,0x5a,0x307,0xe602,0x7a,0x307,0xe602,0x5a,0x30c,0xe602,
+0x7a,0x30c,0xe602,0x41,0x30c,0xe602,0x61,0x30c,0xe602,0x49,0x30c,0xe602,0x69,0x30c,0xe602,0x4f,
+0x30c,0xe602,0x6f,0x30c,0xe602,0x55,0x30c,0xe602,0x75,0x30c,0xdc,0xe643,0x55,0x308,0x304,0xfc,
+0xe643,0x75,0x308,0x304,0xdc,0xe643,0x55,0x308,0x301,0xfc,0xe643,0x75,0x308,0x301,0xdc,0xe643,
+0x55,0x308,0x30c,0xfc,0xe643,0x75,0x308,0x30c,0xdc,0xe643,0x55,0x308,0x300,0xfc,0xe643,0x75,
+0x308,0x300,0xc4,0xe643,0x41,0x308,0x304,0xe4,0xe643,0x61,0x308,0x304,0x226,0xe643,0x41,0x307,
+0x304,0x227,0xe643,0x61,0x307,0x304,0xe602,0xc6,0x304,0xe602,0xe6,0x304,0xe602,0x47,0x30c,0xe602,
+0x67,0x30c,0xe602,0x4b,0x30c,0xe602,0x6b,0x30c,0x1ea,0xe643,0x4f,0x328,0x304,0x1eb,0xe643,0x6f,
+0x328,0x304,0xe602,0x1b7,0x30c,0xe602,0x292,0x30c,0xe602,0x6a,0x30c,0xe602,0x47,0x301,0xe602,0x67,
+0x301,0xe602,0x4e,0x300,0xe602,0x6e,0x300,0xc5,0xe643,0x41,0x30a,0x301,0xe5,0xe643,0x61,0x30a,
+0x301,0xe602,0xc6,0x301,0xe602,0xe6,0x301,0xe602,0xd8,0x301,0xe602,0xf8,0x301,0xe602,0x41,0x30f,
+0xe602,0x61,0x30f,0xe602,0x41,0x311,0xe602,0x61,0x311,0xe602,0x45,0x30f,0xe602,0x65,0x30f,0xe602,
+0x45,0x311,0xe602,0x65,0x311,0xe602,0x49,0x30f,0xe602,0x69,0x30f,0xe602,0x49,0x311,0xe602,0x69,
+0x311,0xe602,0x4f,0x30f,0xe602,0x6f,0x30f,0xe602,0x4f,0x311,0xe602,0x6f,0x311,0xe602,0x52,0x30f,
+0xe602,0x72,0x30f,0xe602,0x52,0x311,0xe602,0x72,0x311,0xe602,0x55,0x30f,0xe602,0x75,0x30f,0xe602,
+0x55,0x311,0xe602,0x75,0x311,0xdc02,0x53,0x326,0xdc02,0x73,0x326,0xdc02,0x54,0x326,0xdc02,0x74,
+0x326,0xe602,0x48,0x30c,0xe602,0x68,0x30c,0xd6,0xe643,0x4f,0x308,0x304,0xf6,0xe643,0x6f,0x308,
+0x304,0xd5,0xe643,0x4f,0x303,0x304,0xf5,0xe643,0x6f,0x303,0x304,0x22e,0xe643,0x4f,0x307,0x304,
+0x22f,0xe643,0x6f,0x307,0x304,0xe602,0x59,0x304,0xe602,0x79,0x304,0xe602,0xa8,0x301,0xe602,0x391,
+0x301,0xe602,0x395,0x301,0xe602,0x397,0x301,0xe602,0x399,0x301,0xe602,0x39f,0x301,0xe602,0x3a5,0x301,
+0xe602,0x3a9,0x301,0x3ca,0xe643,0x3b9,0x308,0x301,0xe602,0x399,0x308,0xe602,0x3a5,0x308,0xe602,0x3b5,
+0x301,0xe602,0x3b9,0x301,0x3cb,0xe643,0x3c5,0x308,0x301,0xe602,0x3bf,0x301,0xe602,0x3c5,0x301,0xe602,
+0x3d2,0x301,0xe602,0x3d2,0x308,0xe602,0x415,0x300,0xe602,0x415,0x308,0xe602,0x413,0x301,0xe602,0x406,
+0x308,0xe602,0x41a,0x301,0xe602,0x418,0x300,0xe602,0x423,0x306,0xe602,0x418,0x306,0xe602,0x438,0x306,
+0xe602,0x435,0x300,0xe602,0x435,0x308,0xe602,0x433,0x301,0xe602,0x456,0x308,0xe602,0x43a,0x301,0xe602,
+0x438,0x300,0xe602,0x443,0x306,0xe602,0x474,0x30f,0xe602,0x475,0x30f,0xe602,0x416,0x306,0xe602,0x436,
+0x306,0xe602,0x410,0x306,0xe602,0x430,0x306,0xe602,0x410,0x308,0xe602,0x430,0x308,0xe602,0x415,0x306,
+0xe602,0x435,0x306,0xe602,0x4d8,0x308,0xe602,0x4d9,0x308,0xe602,0x416,0x308,0xe602,0x436,0x308,0xe602,
+0x417,0x308,0xe602,0x437,0x308,0xe602,0x418,0x304,0xe602,0x438,0x304,0xe602,0x418,0x308,0xe602,0x438,
+0x308,0xe602,0x41e,0x308,0xe602,0x43e,0x308,0xe602,0x4e8,0x308,0xe602,0x4e9,0x308,0xe602,0x42d,0x308,
+0xe602,0x44d,0x308,0xe602,0x423,0x304,0xe602,0x443,0x304,0xe602,0x423,0x308,0xe602,0x443,0x308,0xe602,
+0x423,0x30b,0xe602,0x443,0x30b,0xe602,0x427,0x308,0xe602,0x447,0x308,0xe602,0x42b,0x308,0xe602,0x44b,
+0x308,0xe602,0x627,0x653,0xe602,0x627,0x654,0xe602,0x648,0x654,0xdc02,0x627,0x655,0xe602,0x64a,0x654,
+0xe602,0x6d5,0x654,0xe602,0x6c1,0x654,0xe602,0x6d2,0x654,0x702,0x928,0x93c,0x702,0x930,0x93c,0x702,
+0x933,0x93c,2,0x9c7,0x9be,2,0x9c7,0x9d7,2,0xb47,0xb56,2,0xb47,0xb3e,2,0xb47,
+0xb57,2,0xb92,0xbd7,2,0xbc6,0xbbe,2,0xbc7,0xbbe,2,0xbc6,0xbd7,0x5b02,0xc46,0xc56,
+2,0xcbf,0xcd5,2,0xcc6,0xcd5,2,0xcc6,0xcd6,0xcca,0x43,0xcc6,0xcc2,0xcd5,2,0xd46,
+0xd3e,2,0xd47,0xd3e,2,0xd46,0xd57,0x902,0xdd9,0xdca,0xddc,0x943,0xdd9,0xdcf,0xdca,2,
+0xdd9,0xddf,2,0x1025,0x102e,2,0x1b05,0x1b35,2,0x1b07,0x1b35,2,0x1b09,0x1b35,2,0x1b0b,
+0x1b35,2,0x1b0d,0x1b35,2,0x1b11,0x1b35,2,0x1b3a,0x1b35,2,0x1b3c,0x1b35,2,0x1b3e,0x1b35,
+2,0x1b3f,0x1b35,2,0x1b42,0x1b35,0xdc02,0x41,0x325,0xdc02,0x61,0x325,0xe602,0x42,0x307,0xe602,
+0x62,0x307,0xdc02,0x42,0x323,0xdc02,0x62,0x323,0xdc02,0x42,0x331,0xdc02,0x62,0x331,0xc7,0xe643,
+0x43,0x327,0x301,0xe7,0xe643,0x63,0x327,0x301,0xe602,0x44,0x307,0xe602,0x64,0x307,0xdc02,0x44,
+0x323,0xdc02,0x64,0x323,0xdc02,0x44,0x331,0xdc02,0x64,0x331,0xca02,0x44,0x327,0xca02,0x64,0x327,
+0xdc02,0x44,0x32d,0xdc02,0x64,0x32d,0x112,0xe643,0x45,0x304,0x300,0x113,0xe643,0x65,0x304,0x300,
+0x112,0xe643,0x45,0x304,0x301,0x113,0xe643,0x65,0x304,0x301,0xdc02,0x45,0x32d,0xdc02,0x65,0x32d,
+0xdc02,0x45,0x330,0xdc02,0x65,0x330,0x228,0xe643,0x45,0x327,0x306,0x229,0xe643,0x65,0x327,0x306,
+0xe602,0x46,0x307,0xe602,0x66,0x307,0xe602,0x47,0x304,0xe602,0x67,0x304,0xe602,0x48,0x307,0xe602,
+0x68,0x307,0xdc02,0x48,0x323,0xdc02,0x68,0x323,0xe602,0x48,0x308,0xe602,0x68,0x308,0xca02,0x48,
+0x327,0xca02,0x68,0x327,0xdc02,0x48,0x32e,0xdc02,0x68,0x32e,0xdc02,0x49,0x330,0xdc02,0x69,0x330,
+0xcf,0xe643,0x49,0x308,0x301,0xef,0xe643,0x69,0x308,0x301,0xe602,0x4b,0x301,0xe602,0x6b,0x301,
+0xdc02,0x4b,0x323,0xdc02,0x6b,0x323,0xdc02,0x4b,0x331,0xdc02,0x6b,0x331,0x1e36,0xe643,0x4c,0x323,
+0x304,0x1e37,0xe643,0x6c,0x323,0x304,0xdc02,0x4c,0x331,0xdc02,0x6c,0x331,0xdc02,0x4c,0x32d,0xdc02,
+0x6c,0x32d,0xe602,0x4d,0x301,0xe602,0x6d,0x301,0xe602,0x4d,0x307,0xe602,0x6d,0x307,0xdc02,0x4d,
+0x323,0xdc02,0x6d,0x323,0xe602,0x4e,0x307,0xe602,0x6e,0x307,0xdc02,0x4e,0x323,0xdc02,0x6e,0x323,
+0xdc02,0x4e,0x331,0xdc02,0x6e,0x331,0xdc02,0x4e,0x32d,0xdc02,0x6e,0x32d,0xd5,0xe643,0x4f,0x303,
+0x301,0xf5,0xe643,0x6f,0x303,0x301,0xd5,0xe643,0x4f,0x303,0x308,0xf5,0xe643,0x6f,0x303,0x308,
+0x14c,0xe643,0x4f,0x304,0x300,0x14d,0xe643,0x6f,0x304,0x300,0x14c,0xe643,0x4f,0x304,0x301,0x14d,
+0xe643,0x6f,0x304,0x301,0xe602,0x50,0x301,0xe602,0x70,0x301,0xe602,0x50,0x307,0xe602,0x70,0x307,
+0xe602,0x52,0x307,0xe602,0x72,0x307,0x1e5a,0xe643,0x52,0x323,0x304,0x1e5b,0xe643,0x72,0x323,0x304,
+0xdc02,0x52,0x331,0xdc02,0x72,0x331,0xe602,0x53,0x307,0xe602,0x73,0x307,0x15a,0xe643,0x53,0x301,
+0x307,0x15b,0xe643,0x73,0x301,0x307,0x160,0xe643,0x53,0x30c,0x307,0x161,0xe643,0x73,0x30c,0x307,
+0x1e62,0xe643,0x53,0x323,0x307,0x1e63,0xe643,0x73,0x323,0x307,0xe602,0x54,0x307,0xe602,0x74,0x307,
+0xdc02,0x54,0x323,0xdc02,0x74,0x323,0xdc02,0x54,0x331,0xdc02,0x74,0x331,0xdc02,0x54,0x32d,0xdc02,
+0x74,0x32d,0xdc02,0x55,0x324,0xdc02,0x75,0x324,0xdc02,0x55,0x330,0xdc02,0x75,0x330,0xdc02,0x55,
+0x32d,0xdc02,0x75,0x32d,0x168,0xe643,0x55,0x303,0x301,0x169,0xe643,0x75,0x303,0x301,0x16a,0xe643,
+0x55,0x304,0x308,0x16b,0xe643,0x75,0x304,0x308,0xe602,0x56,0x303,0xe602,0x76,0x303,0xdc02,0x56,
+0x323,0xdc02,0x76,0x323,0xe602,0x57,0x300,0xe602,0x77,0x300,0xe602,0x57,0x301,0xe602,0x77,0x301,
+0xe602,0x57,0x308,0xe602,0x77,0x308,0xe602,0x57,0x307,0xe602,0x77,0x307,0xdc02,0x57,0x323,0xdc02,
+0x77,0x323,0xe602,0x58,0x307,0xe602,0x78,0x307,0xe602,0x58,0x308,0xe602,0x78,0x308,0xe602,0x59,
+0x307,0xe602,0x79,0x307,0xe602,0x5a,0x302,0xe602,0x7a,0x302,0xdc02,0x5a,0x323,0xdc02,0x7a,0x323,
+0xdc02,0x5a,0x331,0xdc02,0x7a,0x331,0xdc02,0x68,0x331,0xe602,0x74,0x308,0xe602,0x77,0x30a,0xe602,
+0x79,0x30a,0xe602,0x17f,0x307,0xe602,0x41,0x309,0xe602,0x61,0x309,0xc2,0xe643,0x41,0x302,0x301,
+0xe2,0xe643,0x61,0x302,0x301,0xc2,0xe643,0x41,0x302,0x300,0xe2,0xe643,0x61,0x302,0x300,0xc2,
+0xe643,0x41,0x302,0x309,0xe2,0xe643,0x61,0x302,0x309,0xc2,0xe643,0x41,0x302,0x303,0xe2,0xe643,
+0x61,0x302,0x303,0x1ea0,0xe643,0x41,0x323,0x302,0x1ea1,0xe643,0x61,0x323,0x302,0x102,0xe643,0x41,
+0x306,0x301,0x103,0xe643,0x61,0x306,0x301,0x102,0xe643,0x41,0x306,0x300,0x103,0xe643,0x61,0x306,
+0x300,0x102,0xe643,0x41,0x306,0x309,0x103,0xe643,0x61,0x306,0x309,0x102,0xe643,0x41,0x306,0x303,
+0x103,0xe643,0x61,0x306,0x303,0x1ea0,0xe643,0x41,0x323,0x306,0x1ea1,0xe643,0x61,0x323,0x306,0xe602,
+0x45,0x309,0xe602,0x65,0x309,0xe602,0x45,0x303,0xe602,0x65,0x303,0xca,0xe643,0x45,0x302,0x301,
+0xea,0xe643,0x65,0x302,0x301,0xca,0xe643,0x45,0x302,0x300,0xea,0xe643,0x65,0x302,0x300,0xca,
+0xe643,0x45,0x302,0x309,0xea,0xe643,0x65,0x302,0x309,0xca,0xe643,0x45,0x302,0x303,0xea,0xe643,
+0x65,0x302,0x303,0x1eb8,0xe643,0x45,0x323,0x302,0x1eb9,0xe643,0x65,0x323,0x302,0xe602,0x49,0x309,
+0xe602,0x69,0x309,0xdc02,0x49,0x323,0xdc02,0x69,0x323,0xe602,0x4f,0x309,0xe602,0x6f,0x309,0xd4,
+0xe643,0x4f,0x302,0x301,0xf4,0xe643,0x6f,0x302,0x301,0xd4,0xe643,0x4f,0x302,0x300,0xf4,0xe643,
+0x6f,0x302,0x300,0xd4,0xe643,0x4f,0x302,0x309,0xf4,0xe643,0x6f,0x302,0x309,0xd4,0xe643,0x4f,
+0x302,0x303,0xf4,0xe643,0x6f,0x302,0x303,0x1ecc,0xe643,0x4f,0x323,0x302,0x1ecd,0xe643,0x6f,0x323,
+0x302,0x1a0,0xe643,0x4f,0x31b,0x301,0x1a1,0xe643,0x6f,0x31b,0x301,0x1a0,0xe643,0x4f,0x31b,0x300,
+0x1a1,0xe643,0x6f,0x31b,0x300,0x1a0,0xe643,0x4f,0x31b,0x309,0x1a1,0xe643,0x6f,0x31b,0x309,0x1a0,
+0xe643,0x4f,0x31b,0x303,0x1a1,0xe643,0x6f,0x31b,0x303,0x1a0,0xdc43,0x4f,0x31b,0x323,0x1a1,0xdc43,
+0x6f,0x31b,0x323,0xdc02,0x55,0x323,0xdc02,0x75,0x323,0xe602,0x55,0x309,0xe602,0x75,0x309,0x1af,
+0xe643,0x55,0x31b,0x301,0x1b0,0xe643,0x75,0x31b,0x301,0x1af,0xe643,0x55,0x31b,0x300,0x1b0,0xe643,
+0x75,0x31b,0x300,0x1af,0xe643,0x55,0x31b,0x309,0x1b0,0xe643,0x75,0x31b,0x309,0x1af,0xe643,0x55,
+0x31b,0x303,0x1b0,0xe643,0x75,0x31b,0x303,0x1af,0xdc43,0x55,0x31b,0x323,0x1b0,0xdc43,0x75,0x31b,
+0x323,0xe602,0x59,0x300,0xe602,0x79,0x300,0xdc02,0x59,0x323,0xdc02,0x79,0x323,0xe602,0x59,0x309,
+0xe602,0x79,0x309,0xe602,0x59,0x303,0xe602,0x79,0x303,0x1f10,0xe643,0x3b5,0x313,0x300,0x1f11,0xe643,
+0x3b5,0x314,0x300,0x1f10,0xe643,0x3b5,0x313,0x301,0x1f11,0xe643,0x3b5,0x314,0x301,0x1f18,0xe643,0x395,
+0x313,0x300,0x1f19,0xe643,0x395,0x314,0x300,0x1f18,0xe643,0x395,0x313,0x301,0x1f19,0xe643,0x395,0x314,
+0x301,0x1f30,0xe643,0x3b9,0x313,0x300,0x1f31,0xe643,0x3b9,0x314,0x300,0x1f30,0xe643,0x3b9,0x313,0x301,
+0x1f31,0xe643,0x3b9,0x314,0x301,0x1f30,0xe643,0x3b9,0x313,0x342,0x1f31,0xe643,0x3b9,0x314,0x342,0x1f38,
+0xe643,0x399,0x313,0x300,0x1f39,0xe643,0x399,0x314,0x300,0x1f38,0xe643,0x399,0x313,0x301,0x1f39,0xe643,
+0x399,0x314,0x301,0x1f38,0xe643,0x399,0x313,0x342,0x1f39,0xe643,0x399,0x314,0x342,0x1f40,0xe643,0x3bf,
+0x313,0x300,0x1f41,0xe643,0x3bf,0x314,0x300,0x1f40,0xe643,0x3bf,0x313,0x301,0x1f41,0xe643,0x3bf,0x314,
+0x301,0x1f48,0xe643,0x39f,0x313,0x300,0x1f49,0xe643,0x39f,0x314,0x300,0x1f48,0xe643,0x39f,0x313,0x301,
+0x1f49,0xe643,0x39f,0x314,0x301,0x1f50,0xe643,0x3c5,0x313,0x300,0x1f51,0xe643,0x3c5,0x314,0x300,0x1f50,
+0xe643,0x3c5,0x313,0x301,0x1f51,0xe643,0x3c5,0x314,0x301,0x1f50,0xe643,0x3c5,0x313,0x342,0x1f51,0xe643,
+0x3c5,0x314,0x342,0x1f59,0xe643,0x3a5,0x314,0x300,0x1f59,0xe643,0x3a5,0x314,0x301,0x1f59,0xe643,0x3a5,
+0x314,0x342,0xe602,0x3b5,0x300,0xe602,0x3b9,0x300,0xe602,0x3bf,0x300,0xe602,0x3c5,0x300,0x1f00,0xf043,
+0x3b1,0x313,0x345,0x1f01,0xf043,0x3b1,0x314,0x345,0x1f02,0x345,2,0xf044,0x3b1,0x313,0x300,0x345,
+0x1f03,0x345,2,0xf044,0x3b1,0x314,0x300,0x345,0x1f04,0x345,2,0xf044,0x3b1,0x313,0x301,0x345,
+0x1f05,0x345,2,0xf044,0x3b1,0x314,0x301,0x345,0x1f06,0x345,2,0xf044,0x3b1,0x313,0x342,0x345,
+0x1f07,0x345,2,0xf044,0x3b1,0x314,0x342,0x345,0x1f08,0xf043,0x391,0x313,0x345,0x1f09,0xf043,0x391,
+0x314,0x345,0x1f0a,0x345,2,0xf044,0x391,0x313,0x300,0x345,0x1f0b,0x345,2,0xf044,0x391,0x314,
+0x300,0x345,0x1f0c,0x345,2,0xf044,0x391,0x313,0x301,0x345,0x1f0d,0x345,2,0xf044,0x391,0x314,
+0x301,0x345,0x1f0e,0x345,2,0xf044,0x391,0x313,0x342,0x345,0x1f0f,0x345,2,0xf044,0x391,0x314,
+0x342,0x345,0x1f20,0xf043,0x3b7,0x313,0x345,0x1f21,0xf043,0x3b7,0x314,0x345,0x1f22,0x345,2,0xf044,
+0x3b7,0x313,0x300,0x345,0x1f23,0x345,2,0xf044,0x3b7,0x314,0x300,0x345,0x1f24,0x345,2,0xf044,
+0x3b7,0x313,0x301,0x345,0x1f25,0x345,2,0xf044,0x3b7,0x314,0x301,0x345,0x1f26,0x345,2,0xf044,
+0x3b7,0x313,0x342,0x345,0x1f27,0x345,2,0xf044,0x3b7,0x314,0x342,0x345,0x1f28,0xf043,0x397,0x313,
+0x345,0x1f29,0xf043,0x397,0x314,0x345,0x1f2a,0x345,2,0xf044,0x397,0x313,0x300,0x345,0x1f2b,0x345,
+2,0xf044,0x397,0x314,0x300,0x345,0x1f2c,0x345,2,0xf044,0x397,0x313,0x301,0x345,0x1f2d,0x345,
+2,0xf044,0x397,0x314,0x301,0x345,0x1f2e,0x345,2,0xf044,0x397,0x313,0x342,0x345,0x1f2f,0x345,
+2,0xf044,0x397,0x314,0x342,0x345,0x1f60,0xf043,0x3c9,0x313,0x345,0x1f61,0xf043,0x3c9,0x314,0x345,
+0x1f62,0x345,2,0xf044,0x3c9,0x313,0x300,0x345,0x1f63,0x345,2,0xf044,0x3c9,0x314,0x300,0x345,
+0x1f64,0x345,2,0xf044,0x3c9,0x313,0x301,0x345,0x1f65,0x345,2,0xf044,0x3c9,0x314,0x301,0x345,
+0x1f66,0x345,2,0xf044,0x3c9,0x313,0x342,0x345,0x1f67,0x345,2,0xf044,0x3c9,0x314,0x342,0x345,
+0x1f68,0xf043,0x3a9,0x313,0x345,0x1f69,0xf043,0x3a9,0x314,0x345,0x1f6a,0x345,2,0xf044,0x3a9,0x313,
+0x300,0x345,0x1f6b,0x345,2,0xf044,0x3a9,0x314,0x300,0x345,0x1f6c,0x345,2,0xf044,0x3a9,0x313,
+0x301,0x345,0x1f6d,0x345,2,0xf044,0x3a9,0x314,0x301,0x345,0x1f6e,0x345,2,0xf044,0x3a9,0x313,
+0x342,0x345,0x1f6f,0x345,2,0xf044,0x3a9,0x314,0x342,0x345,0xe602,0x3b1,0x306,0xe602,0x3b1,0x304,
+0x1f70,0xf043,0x3b1,0x300,0x345,0xf002,0x3b1,0x345,0x3ac,0xf043,0x3b1,0x301,0x345,0x1fb6,0xf043,0x3b1,
+0x342,0x345,0xe602,0x391,0x306,0xe602,0x391,0x304,0xe602,0x391,0x300,0xf002,0x391,0x345,0xe602,0xa8,
+0x342,0x1f74,0xf043,0x3b7,0x300,0x345,0xf002,0x3b7,0x345,0x3ae,0xf043,0x3b7,0x301,0x345,0x1fc6,0xf043,
+0x3b7,0x342,0x345,0xe602,0x395,0x300,0xe602,0x397,0x300,0xf002,0x397,0x345,0xe602,0x1fbf,0x300,0xe602,
+0x1fbf,0x301,0xe602,0x1fbf,0x342,0xe602,0x3b9,0x306,0xe602,0x3b9,0x304,0x3ca,0xe643,0x3b9,0x308,0x300,
+0xe602,0x3b9,0x342,0x3ca,0xe643,0x3b9,0x308,0x342,0xe602,0x399,0x306,0xe602,0x399,0x304,0xe602,0x399,
+0x300,0xe602,0x1ffe,0x300,0xe602,0x1ffe,0x301,0xe602,0x1ffe,0x342,0xe602,0x3c5,0x306,0xe602,0x3c5,0x304,
+0x3cb,0xe643,0x3c5,0x308,0x300,0xe602,0x3c1,0x313,0xe602,0x3c1,0x314,0xe602,0x3c5,0x342,0x3cb,0xe643,
+0x3c5,0x308,0x342,0xe602,0x3a5,0x306,0xe602,0x3a5,0x304,0xe602,0x3a5,0x300,0xe602,0x3a1,0x314,0xe602,
+0xa8,0x300,0x1f7c,0xf043,0x3c9,0x300,0x345,0xf002,0x3c9,0x345,0x3ce,0xf043,0x3c9,0x301,0x345,0x1ff6,
+0xf043,0x3c9,0x342,0x345,0xe602,0x39f,0x300,0xe602,0x3a9,0x300,0xf002,0x3a9,0x345,0x102,0x2190,0x338,
+0x102,0x2192,0x338,0x102,0x2194,0x338,0x102,0x21d0,0x338,0x102,0x21d4,0x338,0x102,0x21d2,0x338,0x102,
+0x2203,0x338,0x102,0x2208,0x338,0x102,0x220b,0x338,0x102,0x2223,0x338,0x102,0x2225,0x338,0x102,0x223c,
+0x338,0x102,0x2243,0x338,0x102,0x2245,0x338,0x102,0x2248,0x338,0x102,0x3d,0x338,0x102,0x2261,0x338,
+0x102,0x224d,0x338,0x102,0x3c,0x338,0x102,0x3e,0x338,0x102,0x2264,0x338,0x102,0x2265,0x338,0x102,
+0x2272,0x338,0x102,0x2273,0x338,0x102,0x2276,0x338,0x102,0x2277,0x338,0x102,0x227a,0x338,0x102,0x227b,
+0x338,0x102,0x2282,0x338,0x102,0x2283,0x338,0x102,0x2286,0x338,0x102,0x2287,0x338,0x102,0x22a2,0x338,
+0x102,0x22a8,0x338,0x102,0x22a9,0x338,0x102,0x22ab,0x338,0x102,0x227c,0x338,0x102,0x227d,0x338,0x102,
+0x2291,0x338,0x102,0x2292,0x338,0x102,0x22b2,0x338,0x102,0x22b3,0x338,0x102,0x22b4,0x338,0x102,0x22b5,
+0x338,0x802,0x304b,0x3099,0x802,0x304d,0x3099,0x802,0x304f,0x3099,0x802,0x3051,0x3099,0x802,0x3053,0x3099,
+0x802,0x3055,0x3099,0x802,0x3057,0x3099,0x802,0x3059,0x3099,0x802,0x305b,0x3099,0x802,0x305d,0x3099,0x802,
+0x305f,0x3099,0x802,0x3061,0x3099,0x802,0x3064,0x3099,0x802,0x3066,0x3099,0x802,0x3068,0x3099,0x802,0x306f,
+0x3099,0x802,0x306f,0x309a,0x802,0x3072,0x3099,0x802,0x3072,0x309a,0x802,0x3075,0x3099,0x802,0x3075,0x309a,
+0x802,0x3078,0x3099,0x802,0x3078,0x309a,0x802,0x307b,0x3099,0x802,0x307b,0x309a,0x802,0x3046,0x3099,0x802,
+0x309d,0x3099,0x802,0x30ab,0x3099,0x802,0x30ad,0x3099,0x802,0x30af,0x3099,0x802,0x30b1,0x3099,0x802,0x30b3,
+0x3099,0x802,0x30b5,0x3099,0x802,0x30b7,0x3099,0x802,0x30b9,0x3099,0x802,0x30bb,0x3099,0x802,0x30bd,0x3099,
+0x802,0x30bf,0x3099,0x802,0x30c1,0x3099,0x802,0x30c4,0x3099,0x802,0x30c6,0x3099,0x802,0x30c8,0x3099,0x802,
+0x30cf,0x3099,0x802,0x30cf,0x309a,0x802,0x30d2,0x3099,0x802,0x30d2,0x309a,0x802,0x30d5,0x3099,0x802,0x30d5,
+0x309a,0x802,0x30d8,0x3099,0x802,0x30d8,0x309a,0x802,0x30db,0x3099,0x802,0x30db,0x309a,0x802,0x30a6,0x3099,
+0x802,0x30ef,0x3099,0x802,0x30f0,0x3099,0x802,0x30f1,0x3099,0x802,0x30f2,0x3099,0x802,0x30fd,0x3099,0x704,
+0xd804,0xdc99,0xd804,0xdcba,0x704,0xd804,0xdc9b,0xd804,0xdcba,0x704,0xd804,0xdca5,0xd804,0xdcba,4,0xd804,
+0xdd31,0xd804,0xdd27,4,0xd804,0xdd32,0xd804,0xdd27,4,0xd804,0xdf47,0xd804,0xdf3e,4,0xd804,0xdf47,
+0xd804,0xdf57,4,0xd805,0xdcb9,0xd805,0xdcba,4,0xd805,0xdcb9,0xd805,0xdcb0,4,0xd805,0xdcb9,0xd805,
+0xdcbd,4,0xd805,0xddb8,0xd805,0xddaf,4,0xd805,0xddb9,0xd805,0xddaf,4,0xd806,0xdd35,0xd806,0xdd30,
+1,0x2b9,1,0x3b,1,0xb7,0x702,0x915,0x93c,0x702,0x916,0x93c,0x702,0x917,0x93c,0x702,
+0x91c,0x93c,0x702,0x921,0x93c,0x702,0x922,0x93c,0x702,0x92b,0x93c,0x702,0x92f,0x93c,0x702,0x9a1,
+0x9bc,0x702,0x9a2,0x9bc,0x702,0x9af,0x9bc,0x702,0xa32,0xa3c,0x702,0xa38,0xa3c,0x702,0xa16,0xa3c,
+0x702,0xa17,0xa3c,0x702,0xa1c,0xa3c,0x702,0xa2b,0xa3c,0x702,0xb21,0xb3c,0x702,0xb22,0xb3c,2,
+0xf42,0xfb7,2,0xf4c,0xfb7,2,0xf51,0xfb7,2,0xf56,0xfb7,2,0xf5b,0xfb7,2,0xf40,
+0xfb5,0x8202,0xfb2,0xf80,0x8202,0xfb3,0xf80,2,0xf92,0xfb7,2,0xf9c,0xfb7,2,0xfa1,0xfb7,
+2,0xfa6,0xfb7,2,0xfab,0xfb7,2,0xf90,0xfb5,1,0x3b9,1,0x60,1,0xb4,1,
+0x3a9,1,0x4b,1,0x3008,1,0x3009,0x102,0x2add,0x338,1,0x8c48,1,0x66f4,1,0x8eca,
+1,0x8cc8,1,0x6ed1,1,0x4e32,1,0x53e5,1,0x9f9c,1,0x5951,1,0x91d1,1,0x5587,
+1,0x5948,1,0x61f6,1,0x7669,1,0x7f85,1,0x863f,1,0x87ba,1,0x88f8,1,0x908f,
+1,0x6a02,1,0x6d1b,1,0x70d9,1,0x73de,1,0x843d,1,0x916a,1,0x99f1,1,0x4e82,
+1,0x5375,1,0x6b04,1,0x721b,1,0x862d,1,0x9e1e,1,0x5d50,1,0x6feb,1,0x85cd,
+1,0x8964,1,0x62c9,1,0x81d8,1,0x881f,1,0x5eca,1,0x6717,1,0x6d6a,1,0x72fc,
+1,0x90ce,1,0x4f86,1,0x51b7,1,0x52de,1,0x64c4,1,0x6ad3,1,0x7210,1,0x76e7,
+1,0x8001,1,0x8606,1,0x865c,1,0x8def,1,0x9732,1,0x9b6f,1,0x9dfa,1,0x788c,
+1,0x797f,1,0x7da0,1,0x83c9,1,0x9304,1,0x9e7f,1,0x8ad6,1,0x58df,1,0x5f04,
+1,0x7c60,1,0x807e,1,0x7262,1,0x78ca,1,0x8cc2,1,0x96f7,1,0x58d8,1,0x5c62,
+1,0x6a13,1,0x6dda,1,0x6f0f,1,0x7d2f,1,0x7e37,1,0x964b,1,0x52d2,1,0x808b,
+1,0x51dc,1,0x51cc,1,0x7a1c,1,0x7dbe,1,0x83f1,1,0x9675,1,0x8b80,1,0x62cf,
+1,0x8afe,1,0x4e39,1,0x5be7,1,0x6012,1,0x7387,1,0x7570,1,0x5317,1,0x78fb,
+1,0x4fbf,1,0x5fa9,1,0x4e0d,1,0x6ccc,1,0x6578,1,0x7d22,1,0x53c3,1,0x585e,
+1,0x7701,1,0x8449,1,0x8aaa,1,0x6bba,1,0x8fb0,1,0x6c88,1,0x62fe,1,0x82e5,
+1,0x63a0,1,0x7565,1,0x4eae,1,0x5169,1,0x51c9,1,0x6881,1,0x7ce7,1,0x826f,
+1,0x8ad2,1,0x91cf,1,0x52f5,1,0x5442,1,0x5973,1,0x5eec,1,0x65c5,1,0x6ffe,
+1,0x792a,1,0x95ad,1,0x9a6a,1,0x9e97,1,0x9ece,1,0x529b,1,0x66c6,1,0x6b77,
+1,0x8f62,1,0x5e74,1,0x6190,1,0x6200,1,0x649a,1,0x6f23,1,0x7149,1,0x7489,
+1,0x79ca,1,0x7df4,1,0x806f,1,0x8f26,1,0x84ee,1,0x9023,1,0x934a,1,0x5217,
+1,0x52a3,1,0x54bd,1,0x70c8,1,0x88c2,1,0x5ec9,1,0x5ff5,1,0x637b,1,0x6bae,
+1,0x7c3e,1,0x7375,1,0x4ee4,1,0x56f9,1,0x5dba,1,0x601c,1,0x73b2,1,0x7469,
+1,0x7f9a,1,0x8046,1,0x9234,1,0x96f6,1,0x9748,1,0x9818,1,0x4f8b,1,0x79ae,
+1,0x91b4,1,0x96b8,1,0x60e1,1,0x4e86,1,0x50da,1,0x5bee,1,0x5c3f,1,0x6599,
+1,0x71ce,1,0x7642,1,0x84fc,1,0x907c,1,0x9f8d,1,0x6688,1,0x962e,1,0x5289,
+1,0x677b,1,0x67f3,1,0x6d41,1,0x6e9c,1,0x7409,1,0x7559,1,0x786b,1,0x7d10,
+1,0x985e,1,0x516d,1,0x622e,1,0x9678,1,0x502b,1,0x5d19,1,0x6dea,1,0x8f2a,
+1,0x5f8b,1,0x6144,1,0x6817,1,0x9686,1,0x5229,1,0x540f,1,0x5c65,1,0x6613,
+1,0x674e,1,0x68a8,1,0x6ce5,1,0x7406,1,0x75e2,1,0x7f79,1,0x88cf,1,0x88e1,
+1,0x91cc,1,0x96e2,1,0x533f,1,0x6eba,1,0x541d,1,0x71d0,1,0x7498,1,0x85fa,
+1,0x96a3,1,0x9c57,1,0x9e9f,1,0x6797,1,0x6dcb,1,0x81e8,1,0x7acb,1,0x7b20,
+1,0x7c92,1,0x72c0,1,0x7099,1,0x8b58,1,0x4ec0,1,0x8336,1,0x523a,1,0x5207,
+1,0x5ea6,1,0x62d3,1,0x7cd6,1,0x5b85,1,0x6d1e,1,0x66b4,1,0x8f3b,1,0x884c,
+1,0x964d,1,0x898b,1,0x5ed3,1,0x5140,1,0x55c0,1,0x585a,1,0x6674,1,0x51de,
+1,0x732a,1,0x76ca,1,0x793c,1,0x795e,1,0x7965,1,0x798f,1,0x9756,1,0x7cbe,
+1,0x7fbd,1,0x8612,1,0x8af8,1,0x9038,1,0x90fd,1,0x98ef,1,0x98fc,1,0x9928,
+1,0x9db4,1,0x90de,1,0x96b7,1,0x4fae,1,0x50e7,1,0x514d,1,0x52c9,1,0x52e4,
+1,0x5351,1,0x559d,1,0x5606,1,0x5668,1,0x5840,1,0x58a8,1,0x5c64,1,0x5c6e,
+1,0x6094,1,0x6168,1,0x618e,1,0x61f2,1,0x654f,1,0x65e2,1,0x6691,1,0x6885,
+1,0x6d77,1,0x6e1a,1,0x6f22,1,0x716e,1,0x722b,1,0x7422,1,0x7891,1,0x793e,
+1,0x7949,1,0x7948,1,0x7950,1,0x7956,1,0x795d,1,0x798d,1,0x798e,1,0x7a40,
+1,0x7a81,1,0x7bc0,1,0x7e09,1,0x7e41,1,0x7f72,1,0x8005,1,0x81ed,1,0x8279,
+1,0x8457,1,0x8910,1,0x8996,1,0x8b01,1,0x8b39,1,0x8cd3,1,0x8d08,1,0x8fb6,
+1,0x96e3,1,0x97ff,1,0x983b,1,0x6075,2,0xd850,0xdeee,1,0x8218,1,0x4e26,1,
+0x51b5,1,0x5168,1,0x4f80,1,0x5145,1,0x5180,1,0x52c7,1,0x52fa,1,0x5555,1,
+0x5599,1,0x55e2,1,0x58b3,1,0x5944,1,0x5954,1,0x5a62,1,0x5b28,1,0x5ed2,1,
+0x5ed9,1,0x5f69,1,0x5fad,1,0x60d8,1,0x614e,1,0x6108,1,0x6160,1,0x6234,1,
+0x63c4,1,0x641c,1,0x6452,1,0x6556,1,0x671b,1,0x6756,1,0x6b79,1,0x6edb,1,
+0x6ecb,1,0x701e,1,0x77a7,1,0x7235,1,0x72af,1,0x7471,1,0x7506,1,0x753b,1,
+0x761d,1,0x761f,1,0x76db,1,0x76f4,1,0x774a,1,0x7740,1,0x78cc,1,0x7ab1,1,
+0x7c7b,1,0x7d5b,1,0x7f3e,1,0x8352,1,0x83ef,1,0x8779,1,0x8941,1,0x8986,1,
+0x8abf,1,0x8acb,1,0x8aed,1,0x8b8a,1,0x8f38,1,0x9072,1,0x9199,1,0x9276,1,
+0x967c,1,0x97db,1,0x980b,1,0x9b12,2,0xd84a,0xdc4a,2,0xd84a,0xdc44,2,0xd84c,0xdfd5,
+1,0x3b9d,1,0x4018,1,0x4039,2,0xd854,0xde49,2,0xd857,0xdcd0,2,0xd85f,0xded3,1,
+0x9f43,1,0x9f8e,0xe02,0x5d9,0x5b4,0x1102,0x5f2,0x5b7,0x1802,0x5e9,0x5c1,0x1902,0x5e9,0x5c2,0xfb49,
+0x1843,0x5e9,0x5bc,0x5c1,0xfb49,0x1943,0x5e9,0x5bc,0x5c2,0x1102,0x5d0,0x5b7,0x1202,0x5d0,0x5b8,0x1502,
+0x5d0,0x5bc,0x1502,0x5d1,0x5bc,0x1502,0x5d2,0x5bc,0x1502,0x5d3,0x5bc,0x1502,0x5d4,0x5bc,0x1502,0x5d5,
+0x5bc,0x1502,0x5d6,0x5bc,0x1502,0x5d8,0x5bc,0x1502,0x5d9,0x5bc,0x1502,0x5da,0x5bc,0x1502,0x5db,0x5bc,
+0x1502,0x5dc,0x5bc,0x1502,0x5de,0x5bc,0x1502,0x5e0,0x5bc,0x1502,0x5e1,0x5bc,0x1502,0x5e3,0x5bc,0x1502,
+0x5e4,0x5bc,0x1502,0x5e6,0x5bc,0x1502,0x5e7,0x5bc,0x1502,0x5e8,0x5bc,0x1502,0x5e9,0x5bc,0x1502,0x5ea,
+0x5bc,0x1302,0x5d5,0x5b9,0x1702,0x5d1,0x5bf,0x1702,0x5db,0x5bf,0x1702,0x5e4,0x5bf,0xd804,0xd834,0xdd57,
+0xd834,0xdd65,0xd804,0xd834,0xdd58,0xd834,0xdd65,0xd834,0xdd5f,0xd834,0xdd6e,4,0xd846,0xd834,0xdd58,0xd834,
+0xdd65,0xd834,0xdd6e,0xd834,0xdd5f,0xd834,0xdd6f,4,0xd846,0xd834,0xdd58,0xd834,0xdd65,0xd834,0xdd6f,0xd834,
+0xdd5f,0xd834,0xdd70,4,0xd846,0xd834,0xdd58,0xd834,0xdd65,0xd834,0xdd70,0xd834,0xdd5f,0xd834,0xdd71,4,
+0xd846,0xd834,0xdd58,0xd834,0xdd65,0xd834,0xdd71,0xd834,0xdd5f,0xd834,0xdd72,4,0xd846,0xd834,0xdd58,0xd834,
+0xdd65,0xd834,0xdd72,0xd804,0xd834,0xddb9,0xd834,0xdd65,0xd804,0xd834,0xddba,0xd834,0xdd65,0xd834,0xddbb,0xd834,
+0xdd6e,4,0xd846,0xd834,0xddb9,0xd834,0xdd65,0xd834,0xdd6e,0xd834,0xddbc,0xd834,0xdd6e,4,0xd846,0xd834,
+0xddba,0xd834,0xdd65,0xd834,0xdd6e,0xd834,0xddbb,0xd834,0xdd6f,4,0xd846,0xd834,0xddb9,0xd834,0xdd65,0xd834,
+0xdd6f,0xd834,0xddbc,0xd834,0xdd6f,4,0xd846,0xd834,0xddba,0xd834,0xdd65,0xd834,0xdd6f,1,0x4e3d,1,
+0x4e38,1,0x4e41,2,0xd840,0xdd22,1,0x4f60,1,0x4fbb,1,0x5002,1,0x507a,1,0x5099,
+1,0x50cf,1,0x349e,2,0xd841,0xde3a,1,0x5154,1,0x5164,1,0x5177,2,0xd841,0xdd1c,
+1,0x34b9,1,0x5167,1,0x518d,2,0xd841,0xdd4b,1,0x5197,1,0x51a4,1,0x4ecc,1,
+0x51ac,2,0xd864,0xdddf,1,0x51f5,1,0x5203,1,0x34df,1,0x523b,1,0x5246,1,0x5272,
+1,0x5277,1,0x3515,1,0x5305,1,0x5306,1,0x5349,1,0x535a,1,0x5373,1,0x537d,
+1,0x537f,2,0xd842,0xde2c,1,0x7070,1,0x53ca,1,0x53df,2,0xd842,0xdf63,1,0x53eb,
+1,0x53f1,1,0x5406,1,0x549e,1,0x5438,1,0x5448,1,0x5468,1,0x54a2,1,0x54f6,
+1,0x5510,1,0x5553,1,0x5563,1,0x5584,1,0x55ab,1,0x55b3,1,0x55c2,1,0x5716,
+1,0x5717,1,0x5651,1,0x5674,1,0x58ee,1,0x57ce,1,0x57f4,1,0x580d,1,0x578b,
+1,0x5832,1,0x5831,1,0x58ac,2,0xd845,0xdce4,1,0x58f2,1,0x58f7,1,0x5906,1,
+0x591a,1,0x5922,1,0x5962,2,0xd845,0xdea8,2,0xd845,0xdeea,1,0x59ec,1,0x5a1b,1,
+0x5a27,1,0x59d8,1,0x5a66,1,0x36ee,1,0x36fc,1,0x5b08,1,0x5b3e,2,0xd846,0xddc8,
+1,0x5bc3,1,0x5bd8,1,0x5bf3,2,0xd846,0xdf18,1,0x5bff,1,0x5c06,1,0x5f53,1,
+0x5c22,1,0x3781,1,0x5c60,1,0x5cc0,1,0x5c8d,2,0xd847,0xdde4,1,0x5d43,2,0xd847,
+0xdde6,1,0x5d6e,1,0x5d6b,1,0x5d7c,1,0x5de1,1,0x5de2,1,0x382f,1,0x5dfd,1,
+0x5e28,1,0x5e3d,1,0x5e69,1,0x3862,2,0xd848,0xdd83,1,0x387c,1,0x5eb0,1,0x5eb3,
+1,0x5eb6,2,0xd868,0xdf92,1,0x5efe,2,0xd848,0xdf31,1,0x8201,1,0x5f22,1,0x38c7,
+2,0xd84c,0xdeb8,2,0xd858,0xddda,1,0x5f62,1,0x5f6b,1,0x38e3,1,0x5f9a,1,0x5fcd,
+1,0x5fd7,1,0x5ff9,1,0x6081,1,0x393a,1,0x391c,2,0xd849,0xded4,1,0x60c7,1,
+0x6148,1,0x614c,1,0x617a,1,0x61b2,1,0x61a4,1,0x61af,1,0x61de,1,0x6210,1,
+0x621b,1,0x625d,1,0x62b1,1,0x62d4,1,0x6350,2,0xd84a,0xdf0c,1,0x633d,1,0x62fc,
+1,0x6368,1,0x6383,1,0x63e4,2,0xd84a,0xdff1,1,0x6422,1,0x63c5,1,0x63a9,1,
+0x3a2e,1,0x6469,1,0x647e,1,0x649d,1,0x6477,1,0x3a6c,1,0x656c,2,0xd84c,0xdc0a,
+1,0x65e3,1,0x66f8,1,0x6649,1,0x3b19,1,0x3b08,1,0x3ae4,1,0x5192,1,0x5195,
+1,0x6700,1,0x669c,1,0x80ad,1,0x43d9,1,0x6721,1,0x675e,1,0x6753,2,0xd84c,
+0xdfc3,1,0x3b49,1,0x67fa,1,0x6785,1,0x6852,2,0xd84d,0xdc6d,1,0x688e,1,0x681f,
+1,0x6914,1,0x6942,1,0x69a3,1,0x69ea,1,0x6aa8,2,0xd84d,0xdea3,1,0x6adb,1,
+0x3c18,1,0x6b21,2,0xd84e,0xdca7,1,0x6b54,1,0x3c4e,1,0x6b72,1,0x6b9f,1,0x6bbb,
+2,0xd84e,0xde8d,2,0xd847,0xdd0b,2,0xd84e,0xdefa,1,0x6c4e,2,0xd84f,0xdcbc,1,0x6cbf,
+1,0x6ccd,1,0x6c67,1,0x6d16,1,0x6d3e,1,0x6d69,1,0x6d78,1,0x6d85,2,0xd84f,
+0xdd1e,1,0x6d34,1,0x6e2f,1,0x6e6e,1,0x3d33,1,0x6ec7,2,0xd84f,0xded1,1,0x6df9,
+1,0x6f6e,2,0xd84f,0xdf5e,2,0xd84f,0xdf8e,1,0x6fc6,1,0x7039,1,0x701b,1,0x3d96,
+1,0x704a,1,0x707d,1,0x7077,1,0x70ad,2,0xd841,0xdd25,1,0x7145,2,0xd850,0xde63,
+1,0x719c,2,0xd850,0xdfab,1,0x7228,1,0x7250,2,0xd851,0xde08,1,0x7280,1,0x7295,
+2,0xd851,0xdf35,2,0xd852,0xdc14,1,0x737a,1,0x738b,1,0x3eac,1,0x73a5,1,0x3eb8,
+1,0x7447,1,0x745c,1,0x7485,1,0x74ca,1,0x3f1b,1,0x7524,2,0xd853,0xdc36,1,
+0x753e,2,0xd853,0xdc92,2,0xd848,0xdd9f,1,0x7610,2,0xd853,0xdfa1,2,0xd853,0xdfb8,2,
+0xd854,0xdc44,1,0x3ffc,1,0x4008,2,0xd854,0xdcf3,2,0xd854,0xdcf2,2,0xd854,0xdd19,2,
+0xd854,0xdd33,1,0x771e,1,0x771f,1,0x778b,1,0x4046,1,0x4096,2,0xd855,0xdc1d,1,
+0x784e,1,0x40e3,2,0xd855,0xde26,2,0xd855,0xde9a,2,0xd855,0xdec5,1,0x79eb,1,0x412f,
+1,0x7a4a,1,0x7a4f,2,0xd856,0xdd7c,2,0xd856,0xdea7,1,0x7aee,1,0x4202,2,0xd856,
+0xdfab,1,0x7bc6,1,0x7bc9,1,0x4227,2,0xd857,0xdc80,1,0x7cd2,1,0x42a0,1,0x7ce8,
+1,0x7ce3,1,0x7d00,2,0xd857,0xdf86,1,0x7d63,1,0x4301,1,0x7dc7,1,0x7e02,1,
+0x7e45,1,0x4334,2,0xd858,0xde28,2,0xd858,0xde47,1,0x4359,2,0xd858,0xded9,1,0x7f7a,
+2,0xd858,0xdf3e,1,0x7f95,1,0x7ffa,2,0xd859,0xdcda,2,0xd859,0xdd23,1,0x8060,2,
+0xd859,0xdda8,1,0x8070,2,0xd84c,0xdf5f,1,0x43d5,1,0x80b2,1,0x8103,1,0x440b,1,
+0x813e,1,0x5ab5,2,0xd859,0xdfa7,2,0xd859,0xdfb5,2,0xd84c,0xdf93,2,0xd84c,0xdf9c,1,
+0x8204,1,0x8f9e,1,0x446b,1,0x8291,1,0x828b,1,0x829d,1,0x52b3,1,0x82b1,1,
+0x82b3,1,0x82bd,1,0x82e6,2,0xd85a,0xdf3c,1,0x831d,1,0x8363,1,0x83ad,1,0x8323,
+1,0x83bd,1,0x83e7,1,0x8353,1,0x83ca,1,0x83cc,1,0x83dc,2,0xd85b,0xdc36,2,
+0xd85b,0xdd6b,2,0xd85b,0xdcd5,1,0x452b,1,0x84f1,1,0x84f3,1,0x8516,2,0xd85c,0xdfca,
+1,0x8564,2,0xd85b,0xdf2c,1,0x455d,1,0x4561,2,0xd85b,0xdfb1,2,0xd85c,0xdcd2,1,
+0x456b,1,0x8650,1,0x8667,1,0x8669,1,0x86a9,1,0x8688,1,0x870e,1,0x86e2,1,
+0x8728,1,0x876b,1,0x8786,1,0x45d7,1,0x87e1,1,0x8801,1,0x45f9,1,0x8860,1,
+0x8863,2,0xd85d,0xde67,1,0x88d7,1,0x88de,1,0x4635,1,0x88fa,1,0x34bb,2,0xd85e,
+0xdcae,2,0xd85e,0xdd66,1,0x46be,1,0x46c7,1,0x8aa0,1,0x8c55,2,0xd85f,0xdca8,1,
+0x8cab,1,0x8cc1,1,0x8d1b,1,0x8d77,2,0xd85f,0xdf2f,2,0xd842,0xdc04,1,0x8dcb,1,
+0x8dbc,1,0x8df0,2,0xd842,0xdcde,1,0x8ed4,2,0xd861,0xddd2,2,0xd861,0xdded,1,0x9094,
+1,0x90f1,1,0x9111,2,0xd861,0xdf2e,1,0x911b,1,0x9238,1,0x92d7,1,0x92d8,1,
+0x927c,1,0x93f9,1,0x9415,2,0xd862,0xdffa,1,0x958b,1,0x4995,1,0x95b7,2,0xd863,
+0xdd77,1,0x49e6,1,0x96c3,1,0x5db2,1,0x9723,2,0xd864,0xdd45,2,0xd864,0xde1a,1,
+0x4a6e,1,0x4a76,1,0x97e0,2,0xd865,0xdc0a,1,0x4ab2,2,0xd865,0xdc96,1,0x9829,2,
+0xd865,0xddb6,1,0x98e2,1,0x4b33,1,0x9929,1,0x99a7,1,0x99c2,1,0x99fe,1,0x4bce,
+2,0xd866,0xdf30,1,0x9c40,1,0x9cfd,1,0x4cce,1,0x4ced,1,0x9d67,2,0xd868,0xdcce,
+1,0x4cf8,2,0xd868,0xdd05,2,0xd868,0xde0e,2,0xd868,0xde91,1,0x9ebb,1,0x4d56,1,
+0x9ef9,1,0x9efe,1,0x9f05,1,0x9f0f,1,0x9f16,1,0x9f3b,2,0xd869,0xde00,0x3ac,0xe642,
+0x3b1,0x301,0x3ad,0xe642,0x3b5,0x301,0x3ae,0xe642,0x3b7,0x301,0x3af,0xe642,0x3b9,0x301,0x3cc,0xe642,
+0x3bf,0x301,0x3cd,0xe642,0x3c5,0x301,0x3ce,0xe642,0x3c9,0x301,0x386,0xe642,0x391,0x301,0x388,0xe642,
+0x395,0x301,0x389,0xe642,0x397,0x301,0x390,1,0xe643,0x3b9,0x308,0x301,0x38a,0xe642,0x399,0x301,
+0x3b0,1,0xe643,0x3c5,0x308,0x301,0x38e,0xe642,0x3a5,0x301,0x385,0xe642,0xa8,0x301,0x38c,0xe642,
+0x39f,0x301,0x38f,0xe642,0x3a9,0x301,0xc5,0xe642,0x41,0x30a,0xe6e6,0xe681,0x300,0xe6e6,0xe681,0x301,
+0xe6e6,0xe681,0x313,0xe6e6,0xe682,0x308,0x301,0x8100,0x8282,0xf71,0xf72,0x8100,0x8482,0xf71,0xf74,0x8100,
+0x8282,0xf71,0xf80,0
+};
+
+static const uint8_t norm2_nfc_data_smallFCD[256]={
+0xc0,0xef,3,0x7f,0xdf,0x70,0xcf,0x87,0xc7,0xe6,0x66,0x46,0x64,0x46,0x66,0x5b,
+0x12,0,0,4,0,0,0,0x43,0x20,2,0x69,0xae,0xc2,0xc0,0xff,0xff,
+0xc0,0x72,0xbf,0,0,0,0,0,0,0,0x40,0,0x80,0x88,0,0,
+0xfe,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0x98,0,0xc3,0x66,0xe0,0x80,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,3,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,7,0,0,2,0
+};
+
+#endif  // INCLUDED_FROM_NORMALIZER2_CPP
diff --git a/thirdparty/icu4c/common/norm2allmodes.h b/thirdparty/icu4c/common/norm2allmodes.h
new file mode 100644
index 0000000000..e8bd52c6ae
--- /dev/null
+++ b/thirdparty/icu4c/common/norm2allmodes.h
@@ -0,0 +1,369 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* norm2allmodes.h
+*
+* created on: 2014sep07
+* created by: Markus W. Scherer
+*/
+
+#ifndef __NORM2ALLMODES_H__
+#define __NORM2ALLMODES_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/edits.h"
+#include "unicode/normalizer2.h"
+#include "unicode/stringoptions.h"
+#include "unicode/unistr.h"
+#include "cpputils.h"
+#include "normalizer2impl.h"
+
+U_NAMESPACE_BEGIN
+
+// Intermediate class:
+// Has Normalizer2Impl and does boilerplate argument checking and setup.
+class Normalizer2WithImpl : public Normalizer2 {
+public:
+    Normalizer2WithImpl(const Normalizer2Impl &ni) : impl(ni) {}
+    virtual ~Normalizer2WithImpl();
+
+    // normalize
+    virtual UnicodeString &
+    normalize(const UnicodeString &src,
+              UnicodeString &dest,
+              UErrorCode &errorCode) const {
+        if(U_FAILURE(errorCode)) {
+            dest.setToBogus();
+            return dest;
+        }
+        const UChar *sArray=src.getBuffer();
+        if(&dest==&src || sArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            dest.setToBogus();
+            return dest;
+        }
+        dest.remove();
+        ReorderingBuffer buffer(impl, dest);
+        if(buffer.init(src.length(), errorCode)) {
+            normalize(sArray, sArray+src.length(), buffer, errorCode);
+        }
+        return dest;
+    }
+    virtual void
+    normalize(const UChar *src, const UChar *limit,
+              ReorderingBuffer &buffer, UErrorCode &errorCode) const = 0;
+
+    // normalize and append
+    virtual UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UErrorCode &errorCode) const {
+        return normalizeSecondAndAppend(first, second, true, errorCode);
+    }
+    virtual UnicodeString &
+    append(UnicodeString &first,
+           const UnicodeString &second,
+           UErrorCode &errorCode) const {
+        return normalizeSecondAndAppend(first, second, false, errorCode);
+    }
+    UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UBool doNormalize,
+                             UErrorCode &errorCode) const {
+        uprv_checkCanGetBuffer(first, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return first;
+        }
+        const UChar *secondArray=second.getBuffer();
+        if(&first==&second || secondArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return first;
+        }
+        int32_t firstLength=first.length();
+        UnicodeString safeMiddle;
+        {
+            ReorderingBuffer buffer(impl, first);
+            if(buffer.init(firstLength+second.length(), errorCode)) {
+                normalizeAndAppend(secondArray, secondArray+second.length(), doNormalize,
+                                   safeMiddle, buffer, errorCode);
+            }
+        }  // The ReorderingBuffer destructor finalizes the first string.
+        if(U_FAILURE(errorCode)) {
+            // Restore the modified suffix of the first string.
+            first.replace(firstLength-safeMiddle.length(), 0x7fffffff, safeMiddle);
+        }
+        return first;
+    }
+    virtual void
+    normalizeAndAppend(const UChar *src, const UChar *limit, UBool doNormalize,
+                       UnicodeString &safeMiddle,
+                       ReorderingBuffer &buffer, UErrorCode &errorCode) const = 0;
+    virtual UBool
+    getDecomposition(UChar32 c, UnicodeString &decomposition) const {
+        UChar buffer[4];
+        int32_t length;
+        const UChar *d=impl.getDecomposition(c, buffer, length);
+        if(d==NULL) {
+            return false;
+        }
+        if(d==buffer) {
+            decomposition.setTo(buffer, length);  // copy the string (Jamos from Hangul syllable c)
+        } else {
+            decomposition.setTo(false, d, length);  // read-only alias
+        }
+        return true;
+    }
+    virtual UBool
+    getRawDecomposition(UChar32 c, UnicodeString &decomposition) const {
+        UChar buffer[30];
+        int32_t length;
+        const UChar *d=impl.getRawDecomposition(c, buffer, length);
+        if(d==NULL) {
+            return false;
+        }
+        if(d==buffer) {
+            decomposition.setTo(buffer, length);  // copy the string (algorithmic decomposition)
+        } else {
+            decomposition.setTo(false, d, length);  // read-only alias
+        }
+        return true;
+    }
+    virtual UChar32
+    composePair(UChar32 a, UChar32 b) const {
+        return impl.composePair(a, b);
+    }
+
+    virtual uint8_t
+    getCombiningClass(UChar32 c) const {
+        return impl.getCC(impl.getNorm16(c));
+    }
+
+    // quick checks
+    virtual UBool
+    isNormalized(const UnicodeString &s, UErrorCode &errorCode) const {
+        if(U_FAILURE(errorCode)) {
+            return false;
+        }
+        const UChar *sArray=s.getBuffer();
+        if(sArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return false;
+        }
+        const UChar *sLimit=sArray+s.length();
+        return sLimit==spanQuickCheckYes(sArray, sLimit, errorCode);
+    }
+    virtual UNormalizationCheckResult
+    quickCheck(const UnicodeString &s, UErrorCode &errorCode) const {
+        return Normalizer2WithImpl::isNormalized(s, errorCode) ? UNORM_YES : UNORM_NO;
+    }
+    virtual int32_t
+    spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const {
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+        const UChar *sArray=s.getBuffer();
+        if(sArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return 0;
+        }
+        return (int32_t)(spanQuickCheckYes(sArray, sArray+s.length(), errorCode)-sArray);
+    }
+    virtual const UChar *
+    spanQuickCheckYes(const UChar *src, const UChar *limit, UErrorCode &errorCode) const = 0;
+
+    virtual UNormalizationCheckResult getQuickCheck(UChar32) const {
+        return UNORM_YES;
+    }
+
+    const Normalizer2Impl &impl;
+};
+
+class DecomposeNormalizer2 : public Normalizer2WithImpl {
+public:
+    DecomposeNormalizer2(const Normalizer2Impl &ni) : Normalizer2WithImpl(ni) {}
+    virtual ~DecomposeNormalizer2();
+
+private:
+    virtual void
+    normalize(const UChar *src, const UChar *limit,
+              ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+        impl.decompose(src, limit, &buffer, errorCode);
+    }
+    using Normalizer2WithImpl::normalize;  // Avoid warning about hiding base class function.
+    virtual void
+    normalizeAndAppend(const UChar *src, const UChar *limit, UBool doNormalize,
+                       UnicodeString &safeMiddle,
+                       ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+        impl.decomposeAndAppend(src, limit, doNormalize, safeMiddle, buffer, errorCode);
+    }
+    virtual const UChar *
+    spanQuickCheckYes(const UChar *src, const UChar *limit, UErrorCode &errorCode) const {
+        return impl.decompose(src, limit, NULL, errorCode);
+    }
+    using Normalizer2WithImpl::spanQuickCheckYes;  // Avoid warning about hiding base class function.
+    virtual UNormalizationCheckResult getQuickCheck(UChar32 c) const {
+        return impl.isDecompYes(impl.getNorm16(c)) ? UNORM_YES : UNORM_NO;
+    }
+    virtual UBool hasBoundaryBefore(UChar32 c) const { return impl.hasDecompBoundaryBefore(c); }
+    virtual UBool hasBoundaryAfter(UChar32 c) const { return impl.hasDecompBoundaryAfter(c); }
+    virtual UBool isInert(UChar32 c) const { return impl.isDecompInert(c); }
+};
+
+class ComposeNormalizer2 : public Normalizer2WithImpl {
+public:
+    ComposeNormalizer2(const Normalizer2Impl &ni, UBool fcc) :
+        Normalizer2WithImpl(ni), onlyContiguous(fcc) {}
+    virtual ~ComposeNormalizer2();
+
+private:
+    virtual void
+    normalize(const UChar *src, const UChar *limit,
+              ReorderingBuffer &buffer, UErrorCode &errorCode) const U_OVERRIDE {
+        impl.compose(src, limit, onlyContiguous, true, buffer, errorCode);
+    }
+    using Normalizer2WithImpl::normalize;  // Avoid warning about hiding base class function.
+
+    void
+    normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink,
+                  Edits *edits, UErrorCode &errorCode) const U_OVERRIDE {
+        if (U_FAILURE(errorCode)) {
+            return;
+        }
+        if (edits != nullptr && (options & U_EDITS_NO_RESET) == 0) {
+            edits->reset();
+        }
+        const uint8_t *s = reinterpret_cast<const uint8_t *>(src.data());
+        impl.composeUTF8(options, onlyContiguous, s, s + src.length(),
+                         &sink, edits, errorCode);
+        sink.Flush();
+    }
+
+    virtual void
+    normalizeAndAppend(const UChar *src, const UChar *limit, UBool doNormalize,
+                       UnicodeString &safeMiddle,
+                       ReorderingBuffer &buffer, UErrorCode &errorCode) const U_OVERRIDE {
+        impl.composeAndAppend(src, limit, doNormalize, onlyContiguous, safeMiddle, buffer, errorCode);
+    }
+
+    virtual UBool
+    isNormalized(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_FAILURE(errorCode)) {
+            return false;
+        }
+        const UChar *sArray=s.getBuffer();
+        if(sArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return false;
+        }
+        UnicodeString temp;
+        ReorderingBuffer buffer(impl, temp);
+        if(!buffer.init(5, errorCode)) {  // small destCapacity for substring normalization
+            return false;
+        }
+        return impl.compose(sArray, sArray+s.length(), onlyContiguous, false, buffer, errorCode);
+    }
+    virtual UBool
+    isNormalizedUTF8(StringPiece sp, UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_FAILURE(errorCode)) {
+            return false;
+        }
+        const uint8_t *s = reinterpret_cast<const uint8_t *>(sp.data());
+        return impl.composeUTF8(0, onlyContiguous, s, s + sp.length(), nullptr, nullptr, errorCode);
+    }
+    virtual UNormalizationCheckResult
+    quickCheck(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_FAILURE(errorCode)) {
+            return UNORM_MAYBE;
+        }
+        const UChar *sArray=s.getBuffer();
+        if(sArray==NULL) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return UNORM_MAYBE;
+        }
+        UNormalizationCheckResult qcResult=UNORM_YES;
+        impl.composeQuickCheck(sArray, sArray+s.length(), onlyContiguous, &qcResult);
+        return qcResult;
+    }
+    virtual const UChar *
+    spanQuickCheckYes(const UChar *src, const UChar *limit, UErrorCode &) const U_OVERRIDE {
+        return impl.composeQuickCheck(src, limit, onlyContiguous, NULL);
+    }
+    using Normalizer2WithImpl::spanQuickCheckYes;  // Avoid warning about hiding base class function.
+    virtual UNormalizationCheckResult getQuickCheck(UChar32 c) const U_OVERRIDE {
+        return impl.getCompQuickCheck(impl.getNorm16(c));
+    }
+    virtual UBool hasBoundaryBefore(UChar32 c) const U_OVERRIDE {
+        return impl.hasCompBoundaryBefore(c);
+    }
+    virtual UBool hasBoundaryAfter(UChar32 c) const U_OVERRIDE {
+        return impl.hasCompBoundaryAfter(c, onlyContiguous);
+    }
+    virtual UBool isInert(UChar32 c) const U_OVERRIDE {
+        return impl.isCompInert(c, onlyContiguous);
+    }
+
+    const UBool onlyContiguous;
+};
+
+class FCDNormalizer2 : public Normalizer2WithImpl {
+public:
+    FCDNormalizer2(const Normalizer2Impl &ni) : Normalizer2WithImpl(ni) {}
+    virtual ~FCDNormalizer2();
+
+private:
+    virtual void
+    normalize(const UChar *src, const UChar *limit,
+              ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+        impl.makeFCD(src, limit, &buffer, errorCode);
+    }
+    using Normalizer2WithImpl::normalize;  // Avoid warning about hiding base class function.
+    virtual void
+    normalizeAndAppend(const UChar *src, const UChar *limit, UBool doNormalize,
+                       UnicodeString &safeMiddle,
+                       ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+        impl.makeFCDAndAppend(src, limit, doNormalize, safeMiddle, buffer, errorCode);
+    }
+    virtual const UChar *
+    spanQuickCheckYes(const UChar *src, const UChar *limit, UErrorCode &errorCode) const {
+        return impl.makeFCD(src, limit, NULL, errorCode);
+    }
+    using Normalizer2WithImpl::spanQuickCheckYes;  // Avoid warning about hiding base class function.
+    virtual UBool hasBoundaryBefore(UChar32 c) const { return impl.hasFCDBoundaryBefore(c); }
+    virtual UBool hasBoundaryAfter(UChar32 c) const { return impl.hasFCDBoundaryAfter(c); }
+    virtual UBool isInert(UChar32 c) const { return impl.isFCDInert(c); }
+};
+
+struct Norm2AllModes : public UMemory {
+    Norm2AllModes(Normalizer2Impl *i)
+            : impl(i), comp(*i, false), decomp(*i), fcd(*i), fcc(*i, true) {}
+    ~Norm2AllModes();
+
+    static Norm2AllModes *createInstance(Normalizer2Impl *impl, UErrorCode &errorCode);
+    static Norm2AllModes *createNFCInstance(UErrorCode &errorCode);
+    static Norm2AllModes *createInstance(const char *packageName,
+                                         const char *name,
+                                         UErrorCode &errorCode);
+
+    static const Norm2AllModes *getNFCInstance(UErrorCode &errorCode);
+    static const Norm2AllModes *getNFKCInstance(UErrorCode &errorCode);
+    static const Norm2AllModes *getNFKC_CFInstance(UErrorCode &errorCode);
+
+    Normalizer2Impl *impl;
+    ComposeNormalizer2 comp;
+    DecomposeNormalizer2 decomp;
+    FCDNormalizer2 fcd;
+    ComposeNormalizer2 fcc;
+};
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_NORMALIZATION
+#endif  // __NORM2ALLMODES_H__
diff --git a/thirdparty/icu4c/common/normalizer2.cpp b/thirdparty/icu4c/common/normalizer2.cpp
new file mode 100644
index 0000000000..6be7e0b21a
--- /dev/null
+++ b/thirdparty/icu4c/common/normalizer2.cpp
@@ -0,0 +1,572 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  normalizer2.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009nov22
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/edits.h"
+#include "unicode/normalizer2.h"
+#include "unicode/stringoptions.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "norm2allmodes.h"
+#include "normalizer2impl.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+
+using icu::Normalizer2Impl;
+
+#if NORM2_HARDCODE_NFC_DATA
+// NFC/NFD data machine-generated by gennorm2 --csource
+#define INCLUDED_FROM_NORMALIZER2_CPP
+#include "norm2_nfc_data.h"
+#endif
+
+U_NAMESPACE_BEGIN
+
+// Public API dispatch via Normalizer2 subclasses -------------------------- ***
+
+Normalizer2::~Normalizer2() {}
+
+void
+Normalizer2::normalizeUTF8(uint32_t /*options*/, StringPiece src, ByteSink &sink,
+                           Edits *edits, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if (edits != nullptr) {
+        errorCode = U_UNSUPPORTED_ERROR;
+        return;
+    }
+    UnicodeString src16 = UnicodeString::fromUTF8(src);
+    normalize(src16, errorCode).toUTF8(sink);
+}
+
+UBool
+Normalizer2::getRawDecomposition(UChar32, UnicodeString &) const {
+    return FALSE;
+}
+
+UChar32
+Normalizer2::composePair(UChar32, UChar32) const {
+    return U_SENTINEL;
+}
+
+uint8_t
+Normalizer2::getCombiningClass(UChar32 /*c*/) const {
+    return 0;
+}
+
+UBool
+Normalizer2::isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const {
+    return U_SUCCESS(errorCode) && isNormalized(UnicodeString::fromUTF8(s), errorCode);
+}
+
+// Normalizer2 implementation for the old UNORM_NONE.
+class NoopNormalizer2 : public Normalizer2 {
+    virtual ~NoopNormalizer2();
+
+    virtual UnicodeString &
+    normalize(const UnicodeString &src,
+              UnicodeString &dest,
+              UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_SUCCESS(errorCode)) {
+            if(&dest!=&src) {
+                dest=src;
+            } else {
+                errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            }
+        }
+        return dest;
+    }
+    virtual void
+    normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink,
+                  Edits *edits, UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_SUCCESS(errorCode)) {
+            if (edits != nullptr) {
+                if ((options & U_EDITS_NO_RESET) == 0) {
+                    edits->reset();
+                }
+                edits->addUnchanged(src.length());
+            }
+            if ((options & U_OMIT_UNCHANGED_TEXT) == 0) {
+                sink.Append(src.data(), src.length());
+            }
+            sink.Flush();
+        }
+    }
+
+    virtual UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_SUCCESS(errorCode)) {
+            if(&first!=&second) {
+                first.append(second);
+            } else {
+                errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            }
+        }
+        return first;
+    }
+    virtual UnicodeString &
+    append(UnicodeString &first,
+           const UnicodeString &second,
+           UErrorCode &errorCode) const U_OVERRIDE {
+        if(U_SUCCESS(errorCode)) {
+            if(&first!=&second) {
+                first.append(second);
+            } else {
+                errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            }
+        }
+        return first;
+    }
+    virtual UBool
+    getDecomposition(UChar32, UnicodeString &) const U_OVERRIDE {
+        return FALSE;
+    }
+    // No need to U_OVERRIDE the default getRawDecomposition().
+    virtual UBool
+    isNormalized(const UnicodeString &, UErrorCode &errorCode) const U_OVERRIDE {
+        return U_SUCCESS(errorCode);
+    }
+    virtual UBool
+    isNormalizedUTF8(StringPiece, UErrorCode &errorCode) const U_OVERRIDE {
+        return U_SUCCESS(errorCode);
+    }
+    virtual UNormalizationCheckResult
+    quickCheck(const UnicodeString &, UErrorCode &) const U_OVERRIDE {
+        return UNORM_YES;
+    }
+    virtual int32_t
+    spanQuickCheckYes(const UnicodeString &s, UErrorCode &) const U_OVERRIDE {
+        return s.length();
+    }
+    virtual UBool hasBoundaryBefore(UChar32) const U_OVERRIDE { return TRUE; }
+    virtual UBool hasBoundaryAfter(UChar32) const U_OVERRIDE { return TRUE; }
+    virtual UBool isInert(UChar32) const U_OVERRIDE { return TRUE; }
+};
+
+NoopNormalizer2::~NoopNormalizer2() {}
+
+Normalizer2WithImpl::~Normalizer2WithImpl() {}
+
+DecomposeNormalizer2::~DecomposeNormalizer2() {}
+
+ComposeNormalizer2::~ComposeNormalizer2() {}
+
+FCDNormalizer2::~FCDNormalizer2() {}
+
+// instance cache ---------------------------------------------------------- ***
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV uprv_normalizer2_cleanup();
+U_CDECL_END
+
+static Normalizer2   *noopSingleton;
+static icu::UInitOnce noopInitOnce = U_INITONCE_INITIALIZER;
+
+static void U_CALLCONV initNoopSingleton(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    noopSingleton=new NoopNormalizer2;
+    if(noopSingleton==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_NORMALIZER2, uprv_normalizer2_cleanup);
+}
+
+const Normalizer2 *Normalizer2Factory::getNoopInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return NULL; }
+    umtx_initOnce(noopInitOnce, &initNoopSingleton, errorCode);
+    return noopSingleton;
+}
+
+const Normalizer2Impl *
+Normalizer2Factory::getImpl(const Normalizer2 *norm2) {
+    return &((Normalizer2WithImpl *)norm2)->impl;
+}
+
+Norm2AllModes::~Norm2AllModes() {
+    delete impl;
+}
+
+Norm2AllModes *
+Norm2AllModes::createInstance(Normalizer2Impl *impl, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        delete impl;
+        return NULL;
+    }
+    Norm2AllModes *allModes=new Norm2AllModes(impl);
+    if(allModes==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        delete impl;
+        return NULL;
+    }
+    return allModes;
+}
+
+#if NORM2_HARDCODE_NFC_DATA
+Norm2AllModes *
+Norm2AllModes::createNFCInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    Normalizer2Impl *impl=new Normalizer2Impl;
+    if(impl==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    impl->init(norm2_nfc_data_indexes, &norm2_nfc_data_trie,
+               norm2_nfc_data_extraData, norm2_nfc_data_smallFCD);
+    return createInstance(impl, errorCode);
+}
+
+static Norm2AllModes *nfcSingleton;
+
+static icu::UInitOnce nfcInitOnce = U_INITONCE_INITIALIZER;
+
+static void U_CALLCONV initNFCSingleton(UErrorCode &errorCode) {
+    nfcSingleton=Norm2AllModes::createNFCInstance(errorCode);
+    ucln_common_registerCleanup(UCLN_COMMON_NORMALIZER2, uprv_normalizer2_cleanup);
+}
+
+const Norm2AllModes *
+Norm2AllModes::getNFCInstance(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return NULL; }
+    umtx_initOnce(nfcInitOnce, &initNFCSingleton, errorCode);
+    return nfcSingleton;
+}
+
+const Normalizer2 *
+Normalizer2::getNFCInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->comp : NULL;
+}
+
+const Normalizer2 *
+Normalizer2::getNFDInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->decomp : NULL;
+}
+
+const Normalizer2 *Normalizer2Factory::getFCDInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->fcd : NULL;
+}
+
+const Normalizer2 *Normalizer2Factory::getFCCInstance(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? &allModes->fcc : NULL;
+}
+
+const Normalizer2Impl *
+Normalizer2Factory::getNFCImpl(UErrorCode &errorCode) {
+    const Norm2AllModes *allModes=Norm2AllModes::getNFCInstance(errorCode);
+    return allModes!=NULL ? allModes->impl : NULL;
+}
+#endif  // NORM2_HARDCODE_NFC_DATA
+
+U_CDECL_BEGIN
+
+static UBool U_CALLCONV uprv_normalizer2_cleanup() {
+    delete noopSingleton;
+    noopSingleton = NULL;
+    noopInitOnce.reset(); 
+#if NORM2_HARDCODE_NFC_DATA
+    delete nfcSingleton;
+    nfcSingleton = NULL;
+    nfcInitOnce.reset(); 
+#endif
+    return TRUE;
+}
+
+U_CDECL_END
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFCInstance(UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getNFCInstance(*pErrorCode);
+}
+
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFDInstance(UErrorCode *pErrorCode) {
+    return (const UNormalizer2 *)Normalizer2::getNFDInstance(*pErrorCode);
+}
+
+U_CAPI void U_EXPORT2
+unorm2_close(UNormalizer2 *norm2) {
+    delete (Normalizer2 *)norm2;
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_normalize(const UNormalizer2 *norm2,
+                 const UChar *src, int32_t length,
+                 UChar *dest, int32_t capacity,
+                 UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if( (src==NULL ? length!=0 : length<-1) ||
+        (dest==NULL ? capacity!=0 : capacity<0) ||
+        (src==dest && src!=NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString destString(dest, 0, capacity);
+    // length==0: Nothing to do, and n2wi->normalize(NULL, NULL, buffer, ...) would crash.
+    if(length!=0) {
+        const Normalizer2 *n2=(const Normalizer2 *)norm2;
+        const Normalizer2WithImpl *n2wi=dynamic_cast<const Normalizer2WithImpl *>(n2);
+        if(n2wi!=NULL) {
+            // Avoid duplicate argument checking and support NUL-terminated src.
+            ReorderingBuffer buffer(n2wi->impl, destString);
+            if(buffer.init(length, *pErrorCode)) {
+                n2wi->normalize(src, length>=0 ? src+length : NULL, buffer, *pErrorCode);
+            }
+        } else {
+            UnicodeString srcString(length<0, src, length);
+            n2->normalize(srcString, destString, *pErrorCode);
+        }
+    }
+    return destString.extract(dest, capacity, *pErrorCode);
+}
+
+static int32_t
+normalizeSecondAndAppend(const UNormalizer2 *norm2,
+                         UChar *first, int32_t firstLength, int32_t firstCapacity,
+                         const UChar *second, int32_t secondLength,
+                         UBool doNormalize,
+                         UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if( (second==NULL ? secondLength!=0 : secondLength<-1) ||
+        (first==NULL ? (firstCapacity!=0 || firstLength!=0) :
+                       (firstCapacity<0 || firstLength<-1)) ||
+        (first==second && first!=NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString firstString(first, firstLength, firstCapacity);
+    firstLength=firstString.length();  // In case it was -1.
+    // secondLength==0: Nothing to do, and n2wi->normalizeAndAppend(NULL, NULL, buffer, ...) would crash.
+    if(secondLength!=0) {
+        const Normalizer2 *n2=(const Normalizer2 *)norm2;
+        const Normalizer2WithImpl *n2wi=dynamic_cast<const Normalizer2WithImpl *>(n2);
+        if(n2wi!=NULL) {
+            // Avoid duplicate argument checking and support NUL-terminated src.
+            UnicodeString safeMiddle;
+            {
+                ReorderingBuffer buffer(n2wi->impl, firstString);
+                if(buffer.init(firstLength+secondLength+1, *pErrorCode)) {  // destCapacity>=-1
+                    n2wi->normalizeAndAppend(second, secondLength>=0 ? second+secondLength : NULL,
+                                             doNormalize, safeMiddle, buffer, *pErrorCode);
+                }
+            }  // The ReorderingBuffer destructor finalizes firstString.
+            if(U_FAILURE(*pErrorCode) || firstString.length()>firstCapacity) {
+                // Restore the modified suffix of the first string.
+                // This does not restore first[] array contents between firstLength and firstCapacity.
+                // (That might be uninitialized memory, as far as we know.)
+                if(first!=NULL) { /* don't dereference NULL */
+                  safeMiddle.extract(0, 0x7fffffff, first+firstLength-safeMiddle.length());
+                  if(firstLength<firstCapacity) {
+                    first[firstLength]=0;  // NUL-terminate in case it was originally.
+                  }
+                }
+            }
+        } else {
+            UnicodeString secondString(secondLength<0, second, secondLength);
+            if(doNormalize) {
+                n2->normalizeSecondAndAppend(firstString, secondString, *pErrorCode);
+            } else {
+                n2->append(firstString, secondString, *pErrorCode);
+            }
+        }
+    }
+    return firstString.extract(first, firstCapacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_normalizeSecondAndAppend(const UNormalizer2 *norm2,
+                                UChar *first, int32_t firstLength, int32_t firstCapacity,
+                                const UChar *second, int32_t secondLength,
+                                UErrorCode *pErrorCode) {
+    return normalizeSecondAndAppend(norm2,
+                                    first, firstLength, firstCapacity,
+                                    second, secondLength,
+                                    TRUE, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_append(const UNormalizer2 *norm2,
+              UChar *first, int32_t firstLength, int32_t firstCapacity,
+              const UChar *second, int32_t secondLength,
+              UErrorCode *pErrorCode) {
+    return normalizeSecondAndAppend(norm2,
+                                    first, firstLength, firstCapacity,
+                                    second, secondLength,
+                                    FALSE, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_getDecomposition(const UNormalizer2 *norm2,
+                        UChar32 c, UChar *decomposition, int32_t capacity,
+                        UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(decomposition==NULL ? capacity!=0 : capacity<0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString destString(decomposition, 0, capacity);
+    if(reinterpret_cast<const Normalizer2 *>(norm2)->getDecomposition(c, destString)) {
+        return destString.extract(decomposition, capacity, *pErrorCode);
+    } else {
+        return -1;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_getRawDecomposition(const UNormalizer2 *norm2,
+                           UChar32 c, UChar *decomposition, int32_t capacity,
+                           UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(decomposition==NULL ? capacity!=0 : capacity<0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString destString(decomposition, 0, capacity);
+    if(reinterpret_cast<const Normalizer2 *>(norm2)->getRawDecomposition(c, destString)) {
+        return destString.extract(decomposition, capacity, *pErrorCode);
+    } else {
+        return -1;
+    }
+}
+
+U_CAPI UChar32 U_EXPORT2
+unorm2_composePair(const UNormalizer2 *norm2, UChar32 a, UChar32 b) {
+    return reinterpret_cast<const Normalizer2 *>(norm2)->composePair(a, b);
+}
+
+U_CAPI uint8_t U_EXPORT2
+unorm2_getCombiningClass(const UNormalizer2 *norm2, UChar32 c) {
+    return reinterpret_cast<const Normalizer2 *>(norm2)->getCombiningClass(c);
+}
+
+U_CAPI UBool U_EXPORT2
+unorm2_isNormalized(const UNormalizer2 *norm2,
+                    const UChar *s, int32_t length,
+                    UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if((s==NULL && length!=0) || length<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString sString(length<0, s, length);
+    return ((const Normalizer2 *)norm2)->isNormalized(sString, *pErrorCode);
+}
+
+U_CAPI UNormalizationCheckResult U_EXPORT2
+unorm2_quickCheck(const UNormalizer2 *norm2,
+                  const UChar *s, int32_t length,
+                  UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return UNORM_NO;
+    }
+    if((s==NULL && length!=0) || length<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return UNORM_NO;
+    }
+    UnicodeString sString(length<0, s, length);
+    return ((const Normalizer2 *)norm2)->quickCheck(sString, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm2_spanQuickCheckYes(const UNormalizer2 *norm2,
+                         const UChar *s, int32_t length,
+                         UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if((s==NULL && length!=0) || length<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    UnicodeString sString(length<0, s, length);
+    return ((const Normalizer2 *)norm2)->spanQuickCheckYes(sString, *pErrorCode);
+}
+
+U_CAPI UBool U_EXPORT2
+unorm2_hasBoundaryBefore(const UNormalizer2 *norm2, UChar32 c) {
+    return ((const Normalizer2 *)norm2)->hasBoundaryBefore(c);
+}
+
+U_CAPI UBool U_EXPORT2
+unorm2_hasBoundaryAfter(const UNormalizer2 *norm2, UChar32 c) {
+    return ((const Normalizer2 *)norm2)->hasBoundaryAfter(c);
+}
+
+U_CAPI UBool U_EXPORT2
+unorm2_isInert(const UNormalizer2 *norm2, UChar32 c) {
+    return ((const Normalizer2 *)norm2)->isInert(c);
+}
+
+// Some properties APIs ---------------------------------------------------- ***
+
+U_CAPI uint8_t U_EXPORT2
+u_getCombiningClass(UChar32 c) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2 *nfd=Normalizer2::getNFDInstance(errorCode);
+    if(U_SUCCESS(errorCode)) {
+        return nfd->getCombiningClass(c);
+    } else {
+        return 0;
+    }
+}
+
+U_CFUNC uint16_t
+unorm_getFCD16(UChar32 c) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+    if(U_SUCCESS(errorCode)) {
+        return impl->getFCD16(c);
+    } else {
+        return 0;
+    }
+}
+
+#endif  // !UCONFIG_NO_NORMALIZATION
diff --git a/thirdparty/icu4c/common/normalizer2impl.cpp b/thirdparty/icu4c/common/normalizer2impl.cpp
new file mode 100644
index 0000000000..cbf6b4d980
--- /dev/null
+++ b/thirdparty/icu4c/common/normalizer2impl.cpp
@@ -0,0 +1,2669 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  normalizer2impl.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009nov22
+*   created by: Markus W. Scherer
+*/
+
+// #define UCPTRIE_DEBUG
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/bytestream.h"
+#include "unicode/edits.h"
+#include "unicode/normalizer2.h"
+#include "unicode/stringoptions.h"
+#include "unicode/ucptrie.h"
+#include "unicode/udata.h"
+#include "unicode/umutablecptrie.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "unicode/utf8.h"
+#include "bytesinkutil.h"
+#include "cmemory.h"
+#include "mutex.h"
+#include "normalizer2impl.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include "ucptrie_impl.h"
+#include "uset_imp.h"
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+/**
+ * UTF-8 lead byte for minNoMaybeCP.
+ * Can be lower than the actual lead byte for c.
+ * Typically U+0300 for NFC/NFD, U+00A0 for NFKC/NFKD, U+0041 for NFKC_Casefold.
+ */
+inline uint8_t leadByteForCP(UChar32 c) {
+    if (c <= 0x7f) {
+        return (uint8_t)c;
+    } else if (c <= 0x7ff) {
+        return (uint8_t)(0xc0+(c>>6));
+    } else {
+        // Should not occur because ccc(U+0300)!=0.
+        return 0xe0;
+    }
+}
+
+/**
+ * Returns the code point from one single well-formed UTF-8 byte sequence
+ * between cpStart and cpLimit.
+ *
+ * Trie UTF-8 macros do not assemble whole code points (for efficiency).
+ * When we do need the code point, we call this function.
+ * We should not need it for normalization-inert data (norm16==0).
+ * Illegal sequences yield the error value norm16==0 just like real normalization-inert code points.
+ */
+UChar32 codePointFromValidUTF8(const uint8_t *cpStart, const uint8_t *cpLimit) {
+    // Similar to U8_NEXT_UNSAFE(s, i, c).
+    U_ASSERT(cpStart < cpLimit);
+    uint8_t c = *cpStart;
+    switch(cpLimit-cpStart) {
+    case 1:
+        return c;
+    case 2:
+        return ((c&0x1f)<<6) | (cpStart[1]&0x3f);
+    case 3:
+        // no need for (c&0xf) because the upper bits are truncated after <<12 in the cast to (UChar)
+        return (UChar)((c<<12) | ((cpStart[1]&0x3f)<<6) | (cpStart[2]&0x3f));
+    case 4:
+        return ((c&7)<<18) | ((cpStart[1]&0x3f)<<12) | ((cpStart[2]&0x3f)<<6) | (cpStart[3]&0x3f);
+    default:
+        UPRV_UNREACHABLE;  // Should not occur.
+    }
+}
+
+/**
+ * Returns the last code point in [start, p[ if it is valid and in U+1000..U+D7FF.
+ * Otherwise returns a negative value.
+ */
+UChar32 previousHangulOrJamo(const uint8_t *start, const uint8_t *p) {
+    if ((p - start) >= 3) {
+        p -= 3;
+        uint8_t l = *p;
+        uint8_t t1, t2;
+        if (0xe1 <= l && l <= 0xed &&
+                (t1 = (uint8_t)(p[1] - 0x80)) <= 0x3f &&
+                (t2 = (uint8_t)(p[2] - 0x80)) <= 0x3f &&
+                (l < 0xed || t1 <= 0x1f)) {
+            return ((l & 0xf) << 12) | (t1 << 6) | t2;
+        }
+    }
+    return U_SENTINEL;
+}
+
+/**
+ * Returns the offset from the Jamo T base if [src, limit[ starts with a single Jamo T code point.
+ * Otherwise returns a negative value.
+ */
+int32_t getJamoTMinusBase(const uint8_t *src, const uint8_t *limit) {
+    // Jamo T: E1 86 A8..E1 87 82
+    if ((limit - src) >= 3 && *src == 0xe1) {
+        if (src[1] == 0x86) {
+            uint8_t t = src[2];
+            // The first Jamo T is U+11A8 but JAMO_T_BASE is 11A7.
+            // Offset 0 does not correspond to any conjoining Jamo.
+            if (0xa8 <= t && t <= 0xbf) {
+                return t - 0xa7;
+            }
+        } else if (src[1] == 0x87) {
+            uint8_t t = src[2];
+            if ((int8_t)t <= (int8_t)0x82u) {
+                return t - (0xa7 - 0x40);
+            }
+        }
+    }
+    return -1;
+}
+
+void
+appendCodePointDelta(const uint8_t *cpStart, const uint8_t *cpLimit, int32_t delta,
+                     ByteSink &sink, Edits *edits) {
+    char buffer[U8_MAX_LENGTH];
+    int32_t length;
+    int32_t cpLength = (int32_t)(cpLimit - cpStart);
+    if (cpLength == 1) {
+        // The builder makes ASCII map to ASCII.
+        buffer[0] = (uint8_t)(*cpStart + delta);
+        length = 1;
+    } else {
+        int32_t trail = *(cpLimit-1) + delta;
+        if (0x80 <= trail && trail <= 0xbf) {
+            // The delta only changes the last trail byte.
+            --cpLimit;
+            length = 0;
+            do { buffer[length++] = *cpStart++; } while (cpStart < cpLimit);
+            buffer[length++] = (uint8_t)trail;
+        } else {
+            // Decode the code point, add the delta, re-encode.
+            UChar32 c = codePointFromValidUTF8(cpStart, cpLimit) + delta;
+            length = 0;
+            U8_APPEND_UNSAFE(buffer, length, c);
+        }
+    }
+    if (edits != nullptr) {
+        edits->addReplace(cpLength, length);
+    }
+    sink.Append(buffer, length);
+}
+
+}  // namespace
+
+// ReorderingBuffer -------------------------------------------------------- ***
+
+ReorderingBuffer::ReorderingBuffer(const Normalizer2Impl &ni, UnicodeString &dest,
+                                   UErrorCode &errorCode) :
+        impl(ni), str(dest),
+        start(str.getBuffer(8)), reorderStart(start), limit(start),
+        remainingCapacity(str.getCapacity()), lastCC(0) {
+    if (start == nullptr && U_SUCCESS(errorCode)) {
+        // getBuffer() already did str.setToBogus()
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+UBool ReorderingBuffer::init(int32_t destCapacity, UErrorCode &errorCode) {
+    int32_t length=str.length();
+    start=str.getBuffer(destCapacity);
+    if(start==NULL) {
+        // getBuffer() already did str.setToBogus()
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    limit=start+length;
+    remainingCapacity=str.getCapacity()-length;
+    reorderStart=start;
+    if(start==limit) {
+        lastCC=0;
+    } else {
+        setIterator();
+        lastCC=previousCC();
+        // Set reorderStart after the last code point with cc<=1 if there is one.
+        if(lastCC>1) {
+            while(previousCC()>1) {}
+        }
+        reorderStart=codePointLimit;
+    }
+    return TRUE;
+}
+
+UBool ReorderingBuffer::equals(const UChar *otherStart, const UChar *otherLimit) const {
+    int32_t length=(int32_t)(limit-start);
+    return
+        length==(int32_t)(otherLimit-otherStart) &&
+        0==u_memcmp(start, otherStart, length);
+}
+
+UBool ReorderingBuffer::equals(const uint8_t *otherStart, const uint8_t *otherLimit) const {
+    U_ASSERT((otherLimit - otherStart) <= INT32_MAX);  // ensured by caller
+    int32_t length = (int32_t)(limit - start);
+    int32_t otherLength = (int32_t)(otherLimit - otherStart);
+    // For equal strings, UTF-8 is at least as long as UTF-16, and at most three times as long.
+    if (otherLength < length || (otherLength / 3) > length) {
+        return FALSE;
+    }
+    // Compare valid strings from between normalization boundaries.
+    // (Invalid sequences are normalization-inert.)
+    for (int32_t i = 0, j = 0;;) {
+        if (i >= length) {
+            return j >= otherLength;
+        } else if (j >= otherLength) {
+            return FALSE;
+        }
+        // Not at the end of either string yet.
+        UChar32 c, other;
+        U16_NEXT_UNSAFE(start, i, c);
+        U8_NEXT_UNSAFE(otherStart, j, other);
+        if (c != other) {
+            return FALSE;
+        }
+    }
+}
+
+UBool ReorderingBuffer::appendSupplementary(UChar32 c, uint8_t cc, UErrorCode &errorCode) {
+    if(remainingCapacity<2 && !resize(2, errorCode)) {
+        return FALSE;
+    }
+    if(lastCC<=cc || cc==0) {
+        limit[0]=U16_LEAD(c);
+        limit[1]=U16_TRAIL(c);
+        limit+=2;
+        lastCC=cc;
+        if(cc<=1) {
+            reorderStart=limit;
+        }
+    } else {
+        insert(c, cc);
+    }
+    remainingCapacity-=2;
+    return TRUE;
+}
+
+UBool ReorderingBuffer::append(const UChar *s, int32_t length, UBool isNFD,
+                               uint8_t leadCC, uint8_t trailCC,
+                               UErrorCode &errorCode) {
+    if(length==0) {
+        return TRUE;
+    }
+    if(remainingCapacity<length && !resize(length, errorCode)) {
+        return FALSE;
+    }
+    remainingCapacity-=length;
+    if(lastCC<=leadCC || leadCC==0) {
+        if(trailCC<=1) {
+            reorderStart=limit+length;
+        } else if(leadCC<=1) {
+            reorderStart=limit+1;  // Ok if not a code point boundary.
+        }
+        const UChar *sLimit=s+length;
+        do { *limit++=*s++; } while(s!=sLimit);
+        lastCC=trailCC;
+    } else {
+        int32_t i=0;
+        UChar32 c;
+        U16_NEXT(s, i, length, c);
+        insert(c, leadCC);  // insert first code point
+        while(i<length) {
+            U16_NEXT(s, i, length, c);
+            if(i<length) {
+                if (isNFD) {
+                    leadCC = Normalizer2Impl::getCCFromYesOrMaybe(impl.getRawNorm16(c));
+                } else {
+                    leadCC = impl.getCC(impl.getNorm16(c));
+                }
+            } else {
+                leadCC=trailCC;
+            }
+            append(c, leadCC, errorCode);
+        }
+    }
+    return TRUE;
+}
+
+UBool ReorderingBuffer::appendZeroCC(UChar32 c, UErrorCode &errorCode) {
+    int32_t cpLength=U16_LENGTH(c);
+    if(remainingCapacity<cpLength && !resize(cpLength, errorCode)) {
+        return FALSE;
+    }
+    remainingCapacity-=cpLength;
+    if(cpLength==1) {
+        *limit++=(UChar)c;
+    } else {
+        limit[0]=U16_LEAD(c);
+        limit[1]=U16_TRAIL(c);
+        limit+=2;
+    }
+    lastCC=0;
+    reorderStart=limit;
+    return TRUE;
+}
+
+UBool ReorderingBuffer::appendZeroCC(const UChar *s, const UChar *sLimit, UErrorCode &errorCode) {
+    if(s==sLimit) {
+        return TRUE;
+    }
+    int32_t length=(int32_t)(sLimit-s);
+    if(remainingCapacity<length && !resize(length, errorCode)) {
+        return FALSE;
+    }
+    u_memcpy(limit, s, length);
+    limit+=length;
+    remainingCapacity-=length;
+    lastCC=0;
+    reorderStart=limit;
+    return TRUE;
+}
+
+void ReorderingBuffer::remove() {
+    reorderStart=limit=start;
+    remainingCapacity=str.getCapacity();
+    lastCC=0;
+}
+
+void ReorderingBuffer::removeSuffix(int32_t suffixLength) {
+    if(suffixLength<(limit-start)) {
+        limit-=suffixLength;
+        remainingCapacity+=suffixLength;
+    } else {
+        limit=start;
+        remainingCapacity=str.getCapacity();
+    }
+    lastCC=0;
+    reorderStart=limit;
+}
+
+UBool ReorderingBuffer::resize(int32_t appendLength, UErrorCode &errorCode) {
+    int32_t reorderStartIndex=(int32_t)(reorderStart-start);
+    int32_t length=(int32_t)(limit-start);
+    str.releaseBuffer(length);
+    int32_t newCapacity=length+appendLength;
+    int32_t doubleCapacity=2*str.getCapacity();
+    if(newCapacity<doubleCapacity) {
+        newCapacity=doubleCapacity;
+    }
+    if(newCapacity<256) {
+        newCapacity=256;
+    }
+    start=str.getBuffer(newCapacity);
+    if(start==NULL) {
+        // getBuffer() already did str.setToBogus()
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    reorderStart=start+reorderStartIndex;
+    limit=start+length;
+    remainingCapacity=str.getCapacity()-length;
+    return TRUE;
+}
+
+void ReorderingBuffer::skipPrevious() {
+    codePointLimit=codePointStart;
+    UChar c=*--codePointStart;
+    if(U16_IS_TRAIL(c) && start<codePointStart && U16_IS_LEAD(*(codePointStart-1))) {
+        --codePointStart;
+    }
+}
+
+uint8_t ReorderingBuffer::previousCC() {
+    codePointLimit=codePointStart;
+    if(reorderStart>=codePointStart) {
+        return 0;
+    }
+    UChar32 c=*--codePointStart;
+    UChar c2;
+    if(U16_IS_TRAIL(c) && start<codePointStart && U16_IS_LEAD(c2=*(codePointStart-1))) {
+        --codePointStart;
+        c=U16_GET_SUPPLEMENTARY(c2, c);
+    }
+    return impl.getCCFromYesOrMaybeCP(c);
+}
+
+// Inserts c somewhere before the last character.
+// Requires 0<cc<lastCC which implies reorderStart<limit.
+void ReorderingBuffer::insert(UChar32 c, uint8_t cc) {
+    for(setIterator(), skipPrevious(); previousCC()>cc;) {}
+    // insert c at codePointLimit, after the character with prevCC<=cc
+    UChar *q=limit;
+    UChar *r=limit+=U16_LENGTH(c);
+    do {
+        *--r=*--q;
+    } while(codePointLimit!=q);
+    writeCodePoint(q, c);
+    if(cc<=1) {
+        reorderStart=r;
+    }
+}
+
+// Normalizer2Impl --------------------------------------------------------- ***
+
+struct CanonIterData : public UMemory {
+    CanonIterData(UErrorCode &errorCode);
+    ~CanonIterData();
+    void addToStartSet(UChar32 origin, UChar32 decompLead, UErrorCode &errorCode);
+    UMutableCPTrie *mutableTrie;
+    UCPTrie *trie;
+    UVector canonStartSets;  // contains UnicodeSet *
+};
+
+Normalizer2Impl::~Normalizer2Impl() {
+    delete fCanonIterData;
+}
+
+void
+Normalizer2Impl::init(const int32_t *inIndexes, const UCPTrie *inTrie,
+                      const uint16_t *inExtraData, const uint8_t *inSmallFCD) {
+    minDecompNoCP = static_cast<UChar>(inIndexes[IX_MIN_DECOMP_NO_CP]);
+    minCompNoMaybeCP = static_cast<UChar>(inIndexes[IX_MIN_COMP_NO_MAYBE_CP]);
+    minLcccCP = static_cast<UChar>(inIndexes[IX_MIN_LCCC_CP]);
+
+    minYesNo = static_cast<uint16_t>(inIndexes[IX_MIN_YES_NO]);
+    minYesNoMappingsOnly = static_cast<uint16_t>(inIndexes[IX_MIN_YES_NO_MAPPINGS_ONLY]);
+    minNoNo = static_cast<uint16_t>(inIndexes[IX_MIN_NO_NO]);
+    minNoNoCompBoundaryBefore = static_cast<uint16_t>(inIndexes[IX_MIN_NO_NO_COMP_BOUNDARY_BEFORE]);
+    minNoNoCompNoMaybeCC = static_cast<uint16_t>(inIndexes[IX_MIN_NO_NO_COMP_NO_MAYBE_CC]);
+    minNoNoEmpty = static_cast<uint16_t>(inIndexes[IX_MIN_NO_NO_EMPTY]);
+    limitNoNo = static_cast<uint16_t>(inIndexes[IX_LIMIT_NO_NO]);
+    minMaybeYes = static_cast<uint16_t>(inIndexes[IX_MIN_MAYBE_YES]);
+    U_ASSERT((minMaybeYes & 7) == 0);  // 8-aligned for noNoDelta bit fields
+    centerNoNoDelta = (minMaybeYes >> DELTA_SHIFT) - MAX_DELTA - 1;
+
+    normTrie=inTrie;
+
+    maybeYesCompositions=inExtraData;
+    extraData=maybeYesCompositions+((MIN_NORMAL_MAYBE_YES-minMaybeYes)>>OFFSET_SHIFT);
+
+    smallFCD=inSmallFCD;
+}
+
+U_CDECL_BEGIN
+
+static uint32_t U_CALLCONV
+segmentStarterMapper(const void * /*context*/, uint32_t value) {
+    return value&CANON_NOT_SEGMENT_STARTER;
+}
+
+U_CDECL_END
+
+void
+Normalizer2Impl::addLcccChars(UnicodeSet &set) const {
+    UChar32 start = 0, end;
+    uint32_t norm16;
+    while ((end = ucptrie_getRange(normTrie, start, UCPMAP_RANGE_FIXED_LEAD_SURROGATES, INERT,
+                                   nullptr, nullptr, &norm16)) >= 0) {
+        if (norm16 > Normalizer2Impl::MIN_NORMAL_MAYBE_YES &&
+                norm16 != Normalizer2Impl::JAMO_VT) {
+            set.add(start, end);
+        } else if (minNoNoCompNoMaybeCC <= norm16 && norm16 < limitNoNo) {
+            uint16_t fcd16 = getFCD16(start);
+            if (fcd16 > 0xff) { set.add(start, end); }
+        }
+        start = end + 1;
+    }
+}
+
+void
+Normalizer2Impl::addPropertyStarts(const USetAdder *sa, UErrorCode & /*errorCode*/) const {
+    // Add the start code point of each same-value range of the trie.
+    UChar32 start = 0, end;
+    uint32_t value;
+    while ((end = ucptrie_getRange(normTrie, start, UCPMAP_RANGE_FIXED_LEAD_SURROGATES, INERT,
+                                   nullptr, nullptr, &value)) >= 0) {
+        sa->add(sa->set, start);
+        if (start != end && isAlgorithmicNoNo((uint16_t)value) &&
+                (value & Normalizer2Impl::DELTA_TCCC_MASK) > Normalizer2Impl::DELTA_TCCC_1) {
+            // Range of code points with same-norm16-value algorithmic decompositions.
+            // They might have different non-zero FCD16 values.
+            uint16_t prevFCD16 = getFCD16(start);
+            while (++start <= end) {
+                uint16_t fcd16 = getFCD16(start);
+                if (fcd16 != prevFCD16) {
+                    sa->add(sa->set, start);
+                    prevFCD16 = fcd16;
+                }
+            }
+        }
+        start = end + 1;
+    }
+
+    /* add Hangul LV syllables and LV+1 because of skippables */
+    for(UChar c=Hangul::HANGUL_BASE; c<Hangul::HANGUL_LIMIT; c+=Hangul::JAMO_T_COUNT) {
+        sa->add(sa->set, c);
+        sa->add(sa->set, c+1);
+    }
+    sa->add(sa->set, Hangul::HANGUL_LIMIT); /* add Hangul+1 to continue with other properties */
+}
+
+void
+Normalizer2Impl::addCanonIterPropertyStarts(const USetAdder *sa, UErrorCode &errorCode) const {
+    // Add the start code point of each same-value range of the canonical iterator data trie.
+    if (!ensureCanonIterData(errorCode)) { return; }
+    // Currently only used for the SEGMENT_STARTER property.
+    UChar32 start = 0, end;
+    uint32_t value;
+    while ((end = ucptrie_getRange(fCanonIterData->trie, start, UCPMAP_RANGE_NORMAL, 0,
+                                   segmentStarterMapper, nullptr, &value)) >= 0) {
+        sa->add(sa->set, start);
+        start = end + 1;
+    }
+}
+
+const UChar *
+Normalizer2Impl::copyLowPrefixFromNulTerminated(const UChar *src,
+                                                UChar32 minNeedDataCP,
+                                                ReorderingBuffer *buffer,
+                                                UErrorCode &errorCode) const {
+    // Make some effort to support NUL-terminated strings reasonably.
+    // Take the part of the fast quick check loop that does not look up
+    // data and check the first part of the string.
+    // After this prefix, determine the string length to simplify the rest
+    // of the code.
+    const UChar *prevSrc=src;
+    UChar c;
+    while((c=*src++)<minNeedDataCP && c!=0) {}
+    // Back out the last character for full processing.
+    // Copy this prefix.
+    if(--src!=prevSrc) {
+        if(buffer!=NULL) {
+            buffer->appendZeroCC(prevSrc, src, errorCode);
+        }
+    }
+    return src;
+}
+
+UnicodeString &
+Normalizer2Impl::decompose(const UnicodeString &src, UnicodeString &dest,
+                           UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        dest.setToBogus();
+        return dest;
+    }
+    const UChar *sArray=src.getBuffer();
+    if(&dest==&src || sArray==NULL) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        dest.setToBogus();
+        return dest;
+    }
+    decompose(sArray, sArray+src.length(), dest, src.length(), errorCode);
+    return dest;
+}
+
+void
+Normalizer2Impl::decompose(const UChar *src, const UChar *limit,
+                           UnicodeString &dest,
+                           int32_t destLengthEstimate,
+                           UErrorCode &errorCode) const {
+    if(destLengthEstimate<0 && limit!=NULL) {
+        destLengthEstimate=(int32_t)(limit-src);
+    }
+    dest.remove();
+    ReorderingBuffer buffer(*this, dest);
+    if(buffer.init(destLengthEstimate, errorCode)) {
+        decompose(src, limit, &buffer, errorCode);
+    }
+}
+
+// Dual functionality:
+// buffer!=NULL: normalize
+// buffer==NULL: isNormalized/spanQuickCheckYes
+const UChar *
+Normalizer2Impl::decompose(const UChar *src, const UChar *limit,
+                           ReorderingBuffer *buffer,
+                           UErrorCode &errorCode) const {
+    UChar32 minNoCP=minDecompNoCP;
+    if(limit==NULL) {
+        src=copyLowPrefixFromNulTerminated(src, minNoCP, buffer, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return src;
+        }
+        limit=u_strchr(src, 0);
+    }
+
+    const UChar *prevSrc;
+    UChar32 c=0;
+    uint16_t norm16=0;
+
+    // only for quick check
+    const UChar *prevBoundary=src;
+    uint8_t prevCC=0;
+
+    for(;;) {
+        // count code units below the minimum or with irrelevant data for the quick check
+        for(prevSrc=src; src!=limit;) {
+            if( (c=*src)<minNoCP ||
+                isMostDecompYesAndZeroCC(norm16=UCPTRIE_FAST_BMP_GET(normTrie, UCPTRIE_16, c))
+            ) {
+                ++src;
+            } else if(!U16_IS_LEAD(c)) {
+                break;
+            } else {
+                UChar c2;
+                if((src+1)!=limit && U16_IS_TRAIL(c2=src[1])) {
+                    c=U16_GET_SUPPLEMENTARY(c, c2);
+                    norm16=UCPTRIE_FAST_SUPP_GET(normTrie, UCPTRIE_16, c);
+                    if(isMostDecompYesAndZeroCC(norm16)) {
+                        src+=2;
+                    } else {
+                        break;
+                    }
+                } else {
+                    ++src;  // unpaired lead surrogate: inert
+                }
+            }
+        }
+        // copy these code units all at once
+        if(src!=prevSrc) {
+            if(buffer!=NULL) {
+                if(!buffer->appendZeroCC(prevSrc, src, errorCode)) {
+                    break;
+                }
+            } else {
+                prevCC=0;
+                prevBoundary=src;
+            }
+        }
+        if(src==limit) {
+            break;
+        }
+
+        // Check one above-minimum, relevant code point.
+        src+=U16_LENGTH(c);
+        if(buffer!=NULL) {
+            if(!decompose(c, norm16, *buffer, errorCode)) {
+                break;
+            }
+        } else {
+            if(isDecompYes(norm16)) {
+                uint8_t cc=getCCFromYesOrMaybe(norm16);
+                if(prevCC<=cc || cc==0) {
+                    prevCC=cc;
+                    if(cc<=1) {
+                        prevBoundary=src;
+                    }
+                    continue;
+                }
+            }
+            return prevBoundary;  // "no" or cc out of order
+        }
+    }
+    return src;
+}
+
+// Decompose a short piece of text which is likely to contain characters that
+// fail the quick check loop and/or where the quick check loop's overhead
+// is unlikely to be amortized.
+// Called by the compose() and makeFCD() implementations.
+const UChar *
+Normalizer2Impl::decomposeShort(const UChar *src, const UChar *limit,
+                                UBool stopAtCompBoundary, UBool onlyContiguous,
+                                ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return nullptr;
+    }
+    while(src<limit) {
+        if (stopAtCompBoundary && *src < minCompNoMaybeCP) {
+            return src;
+        }
+        const UChar *prevSrc = src;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, src, limit, c, norm16);
+        if (stopAtCompBoundary && norm16HasCompBoundaryBefore(norm16)) {
+            return prevSrc;
+        }
+        if(!decompose(c, norm16, buffer, errorCode)) {
+            return nullptr;
+        }
+        if (stopAtCompBoundary && norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+            return src;
+        }
+    }
+    return src;
+}
+
+UBool Normalizer2Impl::decompose(UChar32 c, uint16_t norm16,
+                                 ReorderingBuffer &buffer,
+                                 UErrorCode &errorCode) const {
+    // get the decomposition and the lead and trail cc's
+    if (norm16 >= limitNoNo) {
+        if (isMaybeOrNonZeroCC(norm16)) {
+            return buffer.append(c, getCCFromYesOrMaybe(norm16), errorCode);
+        }
+        // Maps to an isCompYesAndZeroCC.
+        c=mapAlgorithmic(c, norm16);
+        norm16=getRawNorm16(c);
+    }
+    if (norm16 < minYesNo) {
+        // c does not decompose
+        return buffer.append(c, 0, errorCode);
+    } else if(isHangulLV(norm16) || isHangulLVT(norm16)) {
+        // Hangul syllable: decompose algorithmically
+        UChar jamos[3];
+        return buffer.appendZeroCC(jamos, jamos+Hangul::decompose(c, jamos), errorCode);
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    uint16_t firstUnit=*mapping;
+    int32_t length=firstUnit&MAPPING_LENGTH_MASK;
+    uint8_t leadCC, trailCC;
+    trailCC=(uint8_t)(firstUnit>>8);
+    if(firstUnit&MAPPING_HAS_CCC_LCCC_WORD) {
+        leadCC=(uint8_t)(*(mapping-1)>>8);
+    } else {
+        leadCC=0;
+    }
+    return buffer.append((const UChar *)mapping+1, length, TRUE, leadCC, trailCC, errorCode);
+}
+
+const uint8_t *
+Normalizer2Impl::decomposeShort(const uint8_t *src, const uint8_t *limit,
+                                UBool stopAtCompBoundary, UBool onlyContiguous,
+                                ReorderingBuffer &buffer, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return nullptr;
+    }
+    while (src < limit) {
+        const uint8_t *prevSrc = src;
+        uint16_t norm16;
+        UCPTRIE_FAST_U8_NEXT(normTrie, UCPTRIE_16, src, limit, norm16);
+        // Get the decomposition and the lead and trail cc's.
+        UChar32 c = U_SENTINEL;
+        if (norm16 >= limitNoNo) {
+            if (isMaybeOrNonZeroCC(norm16)) {
+                // No boundaries around this character.
+                c = codePointFromValidUTF8(prevSrc, src);
+                if (!buffer.append(c, getCCFromYesOrMaybe(norm16), errorCode)) {
+                    return nullptr;
+                }
+                continue;
+            }
+            // Maps to an isCompYesAndZeroCC.
+            if (stopAtCompBoundary) {
+                return prevSrc;
+            }
+            c = codePointFromValidUTF8(prevSrc, src);
+            c = mapAlgorithmic(c, norm16);
+            norm16 = getRawNorm16(c);
+        } else if (stopAtCompBoundary && norm16 < minNoNoCompNoMaybeCC) {
+            return prevSrc;
+        }
+        // norm16!=INERT guarantees that [prevSrc, src[ is valid UTF-8.
+        // We do not see invalid UTF-8 here because
+        // its norm16==INERT is normalization-inert,
+        // so it gets copied unchanged in the fast path,
+        // and we stop the slow path where invalid UTF-8 begins.
+        U_ASSERT(norm16 != INERT);
+        if (norm16 < minYesNo) {
+            if (c < 0) {
+                c = codePointFromValidUTF8(prevSrc, src);
+            }
+            // does not decompose
+            if (!buffer.append(c, 0, errorCode)) {
+                return nullptr;
+            }
+        } else if (isHangulLV(norm16) || isHangulLVT(norm16)) {
+            // Hangul syllable: decompose algorithmically
+            if (c < 0) {
+                c = codePointFromValidUTF8(prevSrc, src);
+            }
+            char16_t jamos[3];
+            if (!buffer.appendZeroCC(jamos, jamos+Hangul::decompose(c, jamos), errorCode)) {
+                return nullptr;
+            }
+        } else {
+            // The character decomposes, get everything from the variable-length extra data.
+            const uint16_t *mapping = getMapping(norm16);
+            uint16_t firstUnit = *mapping;
+            int32_t length = firstUnit & MAPPING_LENGTH_MASK;
+            uint8_t trailCC = (uint8_t)(firstUnit >> 8);
+            uint8_t leadCC;
+            if (firstUnit & MAPPING_HAS_CCC_LCCC_WORD) {
+                leadCC = (uint8_t)(*(mapping-1) >> 8);
+            } else {
+                leadCC = 0;
+            }
+            if (!buffer.append((const char16_t *)mapping+1, length, TRUE, leadCC, trailCC, errorCode)) {
+                return nullptr;
+            }
+        }
+        if (stopAtCompBoundary && norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+            return src;
+        }
+    }
+    return src;
+}
+
+const UChar *
+Normalizer2Impl::getDecomposition(UChar32 c, UChar buffer[4], int32_t &length) const {
+    uint16_t norm16;
+    if(c<minDecompNoCP || isMaybeOrNonZeroCC(norm16=getNorm16(c))) {
+        // c does not decompose
+        return nullptr;
+    }
+    const UChar *decomp = nullptr;
+    if(isDecompNoAlgorithmic(norm16)) {
+        // Maps to an isCompYesAndZeroCC.
+        c=mapAlgorithmic(c, norm16);
+        decomp=buffer;
+        length=0;
+        U16_APPEND_UNSAFE(buffer, length, c);
+        // The mapping might decompose further.
+        norm16 = getRawNorm16(c);
+    }
+    if (norm16 < minYesNo) {
+        return decomp;
+    } else if(isHangulLV(norm16) || isHangulLVT(norm16)) {
+        // Hangul syllable: decompose algorithmically
+        length=Hangul::decompose(c, buffer);
+        return buffer;
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    length=*mapping&MAPPING_LENGTH_MASK;
+    return (const UChar *)mapping+1;
+}
+
+// The capacity of the buffer must be 30=MAPPING_LENGTH_MASK-1
+// so that a raw mapping fits that consists of one unit ("rm0")
+// plus all but the first two code units of the normal mapping.
+// The maximum length of a normal mapping is 31=MAPPING_LENGTH_MASK.
+const UChar *
+Normalizer2Impl::getRawDecomposition(UChar32 c, UChar buffer[30], int32_t &length) const {
+    uint16_t norm16;
+    if(c<minDecompNoCP || isDecompYes(norm16=getNorm16(c))) {
+        // c does not decompose
+        return NULL;
+    } else if(isHangulLV(norm16) || isHangulLVT(norm16)) {
+        // Hangul syllable: decompose algorithmically
+        Hangul::getRawDecomposition(c, buffer);
+        length=2;
+        return buffer;
+    } else if(isDecompNoAlgorithmic(norm16)) {
+        c=mapAlgorithmic(c, norm16);
+        length=0;
+        U16_APPEND_UNSAFE(buffer, length, c);
+        return buffer;
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    uint16_t firstUnit=*mapping;
+    int32_t mLength=firstUnit&MAPPING_LENGTH_MASK;  // length of normal mapping
+    if(firstUnit&MAPPING_HAS_RAW_MAPPING) {
+        // Read the raw mapping from before the firstUnit and before the optional ccc/lccc word.
+        // Bit 7=MAPPING_HAS_CCC_LCCC_WORD
+        const uint16_t *rawMapping=mapping-((firstUnit>>7)&1)-1;
+        uint16_t rm0=*rawMapping;
+        if(rm0<=MAPPING_LENGTH_MASK) {
+            length=rm0;
+            return (const UChar *)rawMapping-rm0;
+        } else {
+            // Copy the normal mapping and replace its first two code units with rm0.
+            buffer[0]=(UChar)rm0;
+            u_memcpy(buffer+1, (const UChar *)mapping+1+2, mLength-2);
+            length=mLength-1;
+            return buffer;
+        }
+    } else {
+        length=mLength;
+        return (const UChar *)mapping+1;
+    }
+}
+
+void Normalizer2Impl::decomposeAndAppend(const UChar *src, const UChar *limit,
+                                         UBool doDecompose,
+                                         UnicodeString &safeMiddle,
+                                         ReorderingBuffer &buffer,
+                                         UErrorCode &errorCode) const {
+    buffer.copyReorderableSuffixTo(safeMiddle);
+    if(doDecompose) {
+        decompose(src, limit, &buffer, errorCode);
+        return;
+    }
+    // Just merge the strings at the boundary.
+    bool isFirst = true;
+    uint8_t firstCC = 0, prevCC = 0, cc;
+    const UChar *p = src;
+    while (p != limit) {
+        const UChar *codePointStart = p;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, p, limit, c, norm16);
+        if ((cc = getCC(norm16)) == 0) {
+            p = codePointStart;
+            break;
+        }
+        if (isFirst) {
+            firstCC = cc;
+            isFirst = false;
+        }
+        prevCC = cc;
+    }
+    if(limit==NULL) {  // appendZeroCC() needs limit!=NULL
+        limit=u_strchr(p, 0);
+    }
+
+    if (buffer.append(src, (int32_t)(p - src), FALSE, firstCC, prevCC, errorCode)) {
+        buffer.appendZeroCC(p, limit, errorCode);
+    }
+}
+
+UBool Normalizer2Impl::hasDecompBoundaryBefore(UChar32 c) const {
+    return c < minLcccCP || (c <= 0xffff && !singleLeadMightHaveNonZeroFCD16(c)) ||
+        norm16HasDecompBoundaryBefore(getNorm16(c));
+}
+
+UBool Normalizer2Impl::norm16HasDecompBoundaryBefore(uint16_t norm16) const {
+    if (norm16 < minNoNoCompNoMaybeCC) {
+        return TRUE;
+    }
+    if (norm16 >= limitNoNo) {
+        return norm16 <= MIN_NORMAL_MAYBE_YES || norm16 == JAMO_VT;
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    uint16_t firstUnit=*mapping;
+    // TRUE if leadCC==0 (hasFCDBoundaryBefore())
+    return (firstUnit&MAPPING_HAS_CCC_LCCC_WORD)==0 || (*(mapping-1)&0xff00)==0;
+}
+
+UBool Normalizer2Impl::hasDecompBoundaryAfter(UChar32 c) const {
+    if (c < minDecompNoCP) {
+        return TRUE;
+    }
+    if (c <= 0xffff && !singleLeadMightHaveNonZeroFCD16(c)) {
+        return TRUE;
+    }
+    return norm16HasDecompBoundaryAfter(getNorm16(c));
+}
+
+UBool Normalizer2Impl::norm16HasDecompBoundaryAfter(uint16_t norm16) const {
+    if(norm16 <= minYesNo || isHangulLVT(norm16)) {
+        return TRUE;
+    }
+    if (norm16 >= limitNoNo) {
+        if (isMaybeOrNonZeroCC(norm16)) {
+            return norm16 <= MIN_NORMAL_MAYBE_YES || norm16 == JAMO_VT;
+        }
+        // Maps to an isCompYesAndZeroCC.
+        return (norm16 & DELTA_TCCC_MASK) <= DELTA_TCCC_1;
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    uint16_t firstUnit=*mapping;
+    // decomp after-boundary: same as hasFCDBoundaryAfter(),
+    // fcd16<=1 || trailCC==0
+    if(firstUnit>0x1ff) {
+        return FALSE;  // trailCC>1
+    }
+    if(firstUnit<=0xff) {
+        return TRUE;  // trailCC==0
+    }
+    // if(trailCC==1) test leadCC==0, same as checking for before-boundary
+    // TRUE if leadCC==0 (hasFCDBoundaryBefore())
+    return (firstUnit&MAPPING_HAS_CCC_LCCC_WORD)==0 || (*(mapping-1)&0xff00)==0;
+}
+
+/*
+ * Finds the recomposition result for
+ * a forward-combining "lead" character,
+ * specified with a pointer to its compositions list,
+ * and a backward-combining "trail" character.
+ *
+ * If the lead and trail characters combine, then this function returns
+ * the following "compositeAndFwd" value:
+ * Bits 21..1  composite character
+ * Bit      0  set if the composite is a forward-combining starter
+ * otherwise it returns -1.
+ *
+ * The compositions list has (trail, compositeAndFwd) pair entries,
+ * encoded as either pairs or triples of 16-bit units.
+ * The last entry has the high bit of its first unit set.
+ *
+ * The list is sorted by ascending trail characters (there are no duplicates).
+ * A linear search is used.
+ *
+ * See normalizer2impl.h for a more detailed description
+ * of the compositions list format.
+ */
+int32_t Normalizer2Impl::combine(const uint16_t *list, UChar32 trail) {
+    uint16_t key1, firstUnit;
+    if(trail<COMP_1_TRAIL_LIMIT) {
+        // trail character is 0..33FF
+        // result entry may have 2 or 3 units
+        key1=(uint16_t)(trail<<1);
+        while(key1>(firstUnit=*list)) {
+            list+=2+(firstUnit&COMP_1_TRIPLE);
+        }
+        if(key1==(firstUnit&COMP_1_TRAIL_MASK)) {
+            if(firstUnit&COMP_1_TRIPLE) {
+                return ((int32_t)list[1]<<16)|list[2];
+            } else {
+                return list[1];
+            }
+        }
+    } else {
+        // trail character is 3400..10FFFF
+        // result entry has 3 units
+        key1=(uint16_t)(COMP_1_TRAIL_LIMIT+
+                        (((trail>>COMP_1_TRAIL_SHIFT))&
+                          ~COMP_1_TRIPLE));
+        uint16_t key2=(uint16_t)(trail<<COMP_2_TRAIL_SHIFT);
+        uint16_t secondUnit;
+        for(;;) {
+            if(key1>(firstUnit=*list)) {
+                list+=2+(firstUnit&COMP_1_TRIPLE);
+            } else if(key1==(firstUnit&COMP_1_TRAIL_MASK)) {
+                if(key2>(secondUnit=list[1])) {
+                    if(firstUnit&COMP_1_LAST_TUPLE) {
+                        break;
+                    } else {
+                        list+=3;
+                    }
+                } else if(key2==(secondUnit&COMP_2_TRAIL_MASK)) {
+                    return ((int32_t)(secondUnit&~COMP_2_TRAIL_MASK)<<16)|list[2];
+                } else {
+                    break;
+                }
+            } else {
+                break;
+            }
+        }
+    }
+    return -1;
+}
+
+/**
+  * @param list some character's compositions list
+  * @param set recursively receives the composites from these compositions
+  */
+void Normalizer2Impl::addComposites(const uint16_t *list, UnicodeSet &set) const {
+    uint16_t firstUnit;
+    int32_t compositeAndFwd;
+    do {
+        firstUnit=*list;
+        if((firstUnit&COMP_1_TRIPLE)==0) {
+            compositeAndFwd=list[1];
+            list+=2;
+        } else {
+            compositeAndFwd=(((int32_t)list[1]&~COMP_2_TRAIL_MASK)<<16)|list[2];
+            list+=3;
+        }
+        UChar32 composite=compositeAndFwd>>1;
+        if((compositeAndFwd&1)!=0) {
+            addComposites(getCompositionsListForComposite(getRawNorm16(composite)), set);
+        }
+        set.add(composite);
+    } while((firstUnit&COMP_1_LAST_TUPLE)==0);
+}
+
+/*
+ * Recomposes the buffer text starting at recomposeStartIndex
+ * (which is in NFD - decomposed and canonically ordered),
+ * and truncates the buffer contents.
+ *
+ * Note that recomposition never lengthens the text:
+ * Any character consists of either one or two code units;
+ * a composition may contain at most one more code unit than the original starter,
+ * while the combining mark that is removed has at least one code unit.
+ */
+void Normalizer2Impl::recompose(ReorderingBuffer &buffer, int32_t recomposeStartIndex,
+                                UBool onlyContiguous) const {
+    UChar *p=buffer.getStart()+recomposeStartIndex;
+    UChar *limit=buffer.getLimit();
+    if(p==limit) {
+        return;
+    }
+
+    UChar *starter, *pRemove, *q, *r;
+    const uint16_t *compositionsList;
+    UChar32 c, compositeAndFwd;
+    uint16_t norm16;
+    uint8_t cc, prevCC;
+    UBool starterIsSupplementary;
+
+    // Some of the following variables are not used until we have a forward-combining starter
+    // and are only initialized now to avoid compiler warnings.
+    compositionsList=NULL;  // used as indicator for whether we have a forward-combining starter
+    starter=NULL;
+    starterIsSupplementary=FALSE;
+    prevCC=0;
+
+    for(;;) {
+        UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, p, limit, c, norm16);
+        cc=getCCFromYesOrMaybe(norm16);
+        if( // this character combines backward and
+            isMaybe(norm16) &&
+            // we have seen a starter that combines forward and
+            compositionsList!=NULL &&
+            // the backward-combining character is not blocked
+            (prevCC<cc || prevCC==0)
+        ) {
+            if(isJamoVT(norm16)) {
+                // c is a Jamo V/T, see if we can compose it with the previous character.
+                if(c<Hangul::JAMO_T_BASE) {
+                    // c is a Jamo Vowel, compose with previous Jamo L and following Jamo T.
+                    UChar prev=(UChar)(*starter-Hangul::JAMO_L_BASE);
+                    if(prev<Hangul::JAMO_L_COUNT) {
+                        pRemove=p-1;
+                        UChar syllable=(UChar)
+                            (Hangul::HANGUL_BASE+
+                             (prev*Hangul::JAMO_V_COUNT+(c-Hangul::JAMO_V_BASE))*
+                             Hangul::JAMO_T_COUNT);
+                        UChar t;
+                        if(p!=limit && (t=(UChar)(*p-Hangul::JAMO_T_BASE))<Hangul::JAMO_T_COUNT) {
+                            ++p;
+                            syllable+=t;  // The next character was a Jamo T.
+                        }
+                        *starter=syllable;
+                        // remove the Jamo V/T
+                        q=pRemove;
+                        r=p;
+                        while(r<limit) {
+                            *q++=*r++;
+                        }
+                        limit=q;
+                        p=pRemove;
+                    }
+                }
+                /*
+                 * No "else" for Jamo T:
+                 * Since the input is in NFD, there are no Hangul LV syllables that
+                 * a Jamo T could combine with.
+                 * All Jamo Ts are combined above when handling Jamo Vs.
+                 */
+                if(p==limit) {
+                    break;
+                }
+                compositionsList=NULL;
+                continue;
+            } else if((compositeAndFwd=combine(compositionsList, c))>=0) {
+                // The starter and the combining mark (c) do combine.
+                UChar32 composite=compositeAndFwd>>1;
+
+                // Replace the starter with the composite, remove the combining mark.
+                pRemove=p-U16_LENGTH(c);  // pRemove & p: start & limit of the combining mark
+                if(starterIsSupplementary) {
+                    if(U_IS_SUPPLEMENTARY(composite)) {
+                        // both are supplementary
+                        starter[0]=U16_LEAD(composite);
+                        starter[1]=U16_TRAIL(composite);
+                    } else {
+                        *starter=(UChar)composite;
+                        // The composite is shorter than the starter,
+                        // move the intermediate characters forward one.
+                        starterIsSupplementary=FALSE;
+                        q=starter+1;
+                        r=q+1;
+                        while(r<pRemove) {
+                            *q++=*r++;
+                        }
+                        --pRemove;
+                    }
+                } else if(U_IS_SUPPLEMENTARY(composite)) {
+                    // The composite is longer than the starter,
+                    // move the intermediate characters back one.
+                    starterIsSupplementary=TRUE;
+                    ++starter;  // temporarily increment for the loop boundary
+                    q=pRemove;
+                    r=++pRemove;
+                    while(starter<q) {
+                        *--r=*--q;
+                    }
+                    *starter=U16_TRAIL(composite);
+                    *--starter=U16_LEAD(composite);  // undo the temporary increment
+                } else {
+                    // both are on the BMP
+                    *starter=(UChar)composite;
+                }
+
+                /* remove the combining mark by moving the following text over it */
+                if(pRemove<p) {
+                    q=pRemove;
+                    r=p;
+                    while(r<limit) {
+                        *q++=*r++;
+                    }
+                    limit=q;
+                    p=pRemove;
+                }
+                // Keep prevCC because we removed the combining mark.
+
+                if(p==limit) {
+                    break;
+                }
+                // Is the composite a starter that combines forward?
+                if(compositeAndFwd&1) {
+                    compositionsList=
+                        getCompositionsListForComposite(getRawNorm16(composite));
+                } else {
+                    compositionsList=NULL;
+                }
+
+                // We combined; continue with looking for compositions.
+                continue;
+            }
+        }
+
+        // no combination this time
+        prevCC=cc;
+        if(p==limit) {
+            break;
+        }
+
+        // If c did not combine, then check if it is a starter.
+        if(cc==0) {
+            // Found a new starter.
+            if((compositionsList=getCompositionsListForDecompYes(norm16))!=NULL) {
+                // It may combine with something, prepare for it.
+                if(U_IS_BMP(c)) {
+                    starterIsSupplementary=FALSE;
+                    starter=p-1;
+                } else {
+                    starterIsSupplementary=TRUE;
+                    starter=p-2;
+                }
+            }
+        } else if(onlyContiguous) {
+            // FCC: no discontiguous compositions; any intervening character blocks.
+            compositionsList=NULL;
+        }
+    }
+    buffer.setReorderingLimit(limit);
+}
+
+UChar32
+Normalizer2Impl::composePair(UChar32 a, UChar32 b) const {
+    uint16_t norm16=getNorm16(a);  // maps an out-of-range 'a' to inert norm16
+    const uint16_t *list;
+    if(isInert(norm16)) {
+        return U_SENTINEL;
+    } else if(norm16<minYesNoMappingsOnly) {
+        // a combines forward.
+        if(isJamoL(norm16)) {
+            b-=Hangul::JAMO_V_BASE;
+            if(0<=b && b<Hangul::JAMO_V_COUNT) {
+                return
+                    (Hangul::HANGUL_BASE+
+                     ((a-Hangul::JAMO_L_BASE)*Hangul::JAMO_V_COUNT+b)*
+                     Hangul::JAMO_T_COUNT);
+            } else {
+                return U_SENTINEL;
+            }
+        } else if(isHangulLV(norm16)) {
+            b-=Hangul::JAMO_T_BASE;
+            if(0<b && b<Hangul::JAMO_T_COUNT) {  // not b==0!
+                return a+b;
+            } else {
+                return U_SENTINEL;
+            }
+        } else {
+            // 'a' has a compositions list in extraData
+            list=getMapping(norm16);
+            if(norm16>minYesNo) {  // composite 'a' has both mapping & compositions list
+                list+=  // mapping pointer
+                    1+  // +1 to skip the first unit with the mapping length
+                    (*list&MAPPING_LENGTH_MASK);  // + mapping length
+            }
+        }
+    } else if(norm16<minMaybeYes || MIN_NORMAL_MAYBE_YES<=norm16) {
+        return U_SENTINEL;
+    } else {
+        list=getCompositionsListForMaybe(norm16);
+    }
+    if(b<0 || 0x10ffff<b) {  // combine(list, b) requires a valid code point b
+        return U_SENTINEL;
+    }
+#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+    return combine(list, b)>>1;
+#else
+    int32_t compositeAndFwd=combine(list, b);
+    return compositeAndFwd>=0 ? compositeAndFwd>>1 : U_SENTINEL;
+#endif
+}
+
+// Very similar to composeQuickCheck(): Make the same changes in both places if relevant.
+// doCompose: normalize
+// !doCompose: isNormalized (buffer must be empty and initialized)
+UBool
+Normalizer2Impl::compose(const UChar *src, const UChar *limit,
+                         UBool onlyContiguous,
+                         UBool doCompose,
+                         ReorderingBuffer &buffer,
+                         UErrorCode &errorCode) const {
+    const UChar *prevBoundary=src;
+    UChar32 minNoMaybeCP=minCompNoMaybeCP;
+    if(limit==NULL) {
+        src=copyLowPrefixFromNulTerminated(src, minNoMaybeCP,
+                                           doCompose ? &buffer : NULL,
+                                           errorCode);
+        if(U_FAILURE(errorCode)) {
+            return FALSE;
+        }
+        limit=u_strchr(src, 0);
+        if (prevBoundary != src) {
+            if (hasCompBoundaryAfter(*(src-1), onlyContiguous)) {
+                prevBoundary = src;
+            } else {
+                buffer.removeSuffix(1);
+                prevBoundary = --src;
+            }
+        }
+    }
+
+    for (;;) {
+        // Fast path: Scan over a sequence of characters below the minimum "no or maybe" code point,
+        // or with (compYes && ccc==0) properties.
+        const UChar *prevSrc;
+        UChar32 c = 0;
+        uint16_t norm16 = 0;
+        for (;;) {
+            if (src == limit) {
+                if (prevBoundary != limit && doCompose) {
+                    buffer.appendZeroCC(prevBoundary, limit, errorCode);
+                }
+                return TRUE;
+            }
+            if( (c=*src)<minNoMaybeCP ||
+                isCompYesAndZeroCC(norm16=UCPTRIE_FAST_BMP_GET(normTrie, UCPTRIE_16, c))
+            ) {
+                ++src;
+            } else {
+                prevSrc = src++;
+                if(!U16_IS_LEAD(c)) {
+                    break;
+                } else {
+                    UChar c2;
+                    if(src!=limit && U16_IS_TRAIL(c2=*src)) {
+                        ++src;
+                        c=U16_GET_SUPPLEMENTARY(c, c2);
+                        norm16=UCPTRIE_FAST_SUPP_GET(normTrie, UCPTRIE_16, c);
+                        if(!isCompYesAndZeroCC(norm16)) {
+                            break;
+                        }
+                    }
+                }
+            }
+        }
+        // isCompYesAndZeroCC(norm16) is false, that is, norm16>=minNoNo.
+        // The current character is either a "noNo" (has a mapping)
+        // or a "maybeYes" (combines backward)
+        // or a "yesYes" with ccc!=0.
+        // It is not a Hangul syllable or Jamo L because those have "yes" properties.
+
+        // Medium-fast path: Handle cases that do not require full decomposition and recomposition.
+        if (!isMaybeOrNonZeroCC(norm16)) {  // minNoNo <= norm16 < minMaybeYes
+            if (!doCompose) {
+                return FALSE;
+            }
+            // Fast path for mapping a character that is immediately surrounded by boundaries.
+            // In this case, we need not decompose around the current character.
+            if (isDecompNoAlgorithmic(norm16)) {
+                // Maps to a single isCompYesAndZeroCC character
+                // which also implies hasCompBoundaryBefore.
+                if (norm16HasCompBoundaryAfter(norm16, onlyContiguous) ||
+                        hasCompBoundaryBefore(src, limit)) {
+                    if (prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+                        break;
+                    }
+                    if(!buffer.append(mapAlgorithmic(c, norm16), 0, errorCode)) {
+                        break;
+                    }
+                    prevBoundary = src;
+                    continue;
+                }
+            } else if (norm16 < minNoNoCompBoundaryBefore) {
+                // The mapping is comp-normalized which also implies hasCompBoundaryBefore.
+                if (norm16HasCompBoundaryAfter(norm16, onlyContiguous) ||
+                        hasCompBoundaryBefore(src, limit)) {
+                    if (prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+                        break;
+                    }
+                    const UChar *mapping = reinterpret_cast<const UChar *>(getMapping(norm16));
+                    int32_t length = *mapping++ & MAPPING_LENGTH_MASK;
+                    if(!buffer.appendZeroCC(mapping, mapping + length, errorCode)) {
+                        break;
+                    }
+                    prevBoundary = src;
+                    continue;
+                }
+            } else if (norm16 >= minNoNoEmpty) {
+                // The current character maps to nothing.
+                // Simply omit it from the output if there is a boundary before _or_ after it.
+                // The character itself implies no boundaries.
+                if (hasCompBoundaryBefore(src, limit) ||
+                        hasCompBoundaryAfter(prevBoundary, prevSrc, onlyContiguous)) {
+                    if (prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+                        break;
+                    }
+                    prevBoundary = src;
+                    continue;
+                }
+            }
+            // Other "noNo" type, or need to examine more text around this character:
+            // Fall through to the slow path.
+        } else if (isJamoVT(norm16) && prevBoundary != prevSrc) {
+            UChar prev=*(prevSrc-1);
+            if(c<Hangul::JAMO_T_BASE) {
+                // The current character is a Jamo Vowel,
+                // compose with previous Jamo L and following Jamo T.
+                UChar l = (UChar)(prev-Hangul::JAMO_L_BASE);
+                if(l<Hangul::JAMO_L_COUNT) {
+                    if (!doCompose) {
+                        return FALSE;
+                    }
+                    int32_t t;
+                    if (src != limit &&
+                            0 < (t = ((int32_t)*src - Hangul::JAMO_T_BASE)) &&
+                            t < Hangul::JAMO_T_COUNT) {
+                        // The next character is a Jamo T.
+                        ++src;
+                    } else if (hasCompBoundaryBefore(src, limit)) {
+                        // No Jamo T follows, not even via decomposition.
+                        t = 0;
+                    } else {
+                        t = -1;
+                    }
+                    if (t >= 0) {
+                        UChar32 syllable = Hangul::HANGUL_BASE +
+                            (l*Hangul::JAMO_V_COUNT + (c-Hangul::JAMO_V_BASE)) *
+                            Hangul::JAMO_T_COUNT + t;
+                        --prevSrc;  // Replace the Jamo L as well.
+                        if (prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+                            break;
+                        }
+                        if(!buffer.appendBMP((UChar)syllable, 0, errorCode)) {
+                            break;
+                        }
+                        prevBoundary = src;
+                        continue;
+                    }
+                    // If we see L+V+x where x!=T then we drop to the slow path,
+                    // decompose and recompose.
+                    // This is to deal with NFKC finding normal L and V but a
+                    // compatibility variant of a T.
+                    // We need to either fully compose that combination here
+                    // (which would complicate the code and may not work with strange custom data)
+                    // or use the slow path.
+                }
+            } else if (Hangul::isHangulLV(prev)) {
+                // The current character is a Jamo Trailing consonant,
+                // compose with previous Hangul LV that does not contain a Jamo T.
+                if (!doCompose) {
+                    return FALSE;
+                }
+                UChar32 syllable = prev + c - Hangul::JAMO_T_BASE;
+                --prevSrc;  // Replace the Hangul LV as well.
+                if (prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+                    break;
+                }
+                if(!buffer.appendBMP((UChar)syllable, 0, errorCode)) {
+                    break;
+                }
+                prevBoundary = src;
+                continue;
+            }
+            // No matching context, or may need to decompose surrounding text first:
+            // Fall through to the slow path.
+        } else if (norm16 > JAMO_VT) {  // norm16 >= MIN_YES_YES_WITH_CC
+            // One or more combining marks that do not combine-back:
+            // Check for canonical order, copy unchanged if ok and
+            // if followed by a character with a boundary-before.
+            uint8_t cc = getCCFromNormalYesOrMaybe(norm16);  // cc!=0
+            if (onlyContiguous /* FCC */ && getPreviousTrailCC(prevBoundary, prevSrc) > cc) {
+                // Fails FCD test, need to decompose and contiguously recompose.
+                if (!doCompose) {
+                    return FALSE;
+                }
+            } else {
+                // If !onlyContiguous (not FCC), then we ignore the tccc of
+                // the previous character which passed the quick check "yes && ccc==0" test.
+                const UChar *nextSrc;
+                uint16_t n16;
+                for (;;) {
+                    if (src == limit) {
+                        if (doCompose) {
+                            buffer.appendZeroCC(prevBoundary, limit, errorCode);
+                        }
+                        return TRUE;
+                    }
+                    uint8_t prevCC = cc;
+                    nextSrc = src;
+                    UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, nextSrc, limit, c, n16);
+                    if (n16 >= MIN_YES_YES_WITH_CC) {
+                        cc = getCCFromNormalYesOrMaybe(n16);
+                        if (prevCC > cc) {
+                            if (!doCompose) {
+                                return FALSE;
+                            }
+                            break;
+                        }
+                    } else {
+                        break;
+                    }
+                    src = nextSrc;
+                }
+                // src is after the last in-order combining mark.
+                // If there is a boundary here, then we continue with no change.
+                if (norm16HasCompBoundaryBefore(n16)) {
+                    if (isCompYesAndZeroCC(n16)) {
+                        src = nextSrc;
+                    }
+                    continue;
+                }
+                // Use the slow path. There is no boundary in [prevSrc, src[.
+            }
+        }
+
+        // Slow path: Find the nearest boundaries around the current character,
+        // decompose and recompose.
+        if (prevBoundary != prevSrc && !norm16HasCompBoundaryBefore(norm16)) {
+            const UChar *p = prevSrc;
+            UCPTRIE_FAST_U16_PREV(normTrie, UCPTRIE_16, prevBoundary, p, c, norm16);
+            if (!norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+                prevSrc = p;
+            }
+        }
+        if (doCompose && prevBoundary != prevSrc && !buffer.appendZeroCC(prevBoundary, prevSrc, errorCode)) {
+            break;
+        }
+        int32_t recomposeStartIndex=buffer.length();
+        // We know there is not a boundary here.
+        decomposeShort(prevSrc, src, FALSE /* !stopAtCompBoundary */, onlyContiguous,
+                       buffer, errorCode);
+        // Decompose until the next boundary.
+        src = decomposeShort(src, limit, TRUE /* stopAtCompBoundary */, onlyContiguous,
+                             buffer, errorCode);
+        if (U_FAILURE(errorCode)) {
+            break;
+        }
+        if ((src - prevSrc) > INT32_MAX) {  // guard before buffer.equals()
+            errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            return TRUE;
+        }
+        recompose(buffer, recomposeStartIndex, onlyContiguous);
+        if(!doCompose) {
+            if(!buffer.equals(prevSrc, src)) {
+                return FALSE;
+            }
+            buffer.remove();
+        }
+        prevBoundary=src;
+    }
+    return TRUE;
+}
+
+// Very similar to compose(): Make the same changes in both places if relevant.
+// pQCResult==NULL: spanQuickCheckYes
+// pQCResult!=NULL: quickCheck (*pQCResult must be UNORM_YES)
+const UChar *
+Normalizer2Impl::composeQuickCheck(const UChar *src, const UChar *limit,
+                                   UBool onlyContiguous,
+                                   UNormalizationCheckResult *pQCResult) const {
+    const UChar *prevBoundary=src;
+    UChar32 minNoMaybeCP=minCompNoMaybeCP;
+    if(limit==NULL) {
+        UErrorCode errorCode=U_ZERO_ERROR;
+        src=copyLowPrefixFromNulTerminated(src, minNoMaybeCP, NULL, errorCode);
+        limit=u_strchr(src, 0);
+        if (prevBoundary != src) {
+            if (hasCompBoundaryAfter(*(src-1), onlyContiguous)) {
+                prevBoundary = src;
+            } else {
+                prevBoundary = --src;
+            }
+        }
+    }
+
+    for(;;) {
+        // Fast path: Scan over a sequence of characters below the minimum "no or maybe" code point,
+        // or with (compYes && ccc==0) properties.
+        const UChar *prevSrc;
+        UChar32 c = 0;
+        uint16_t norm16 = 0;
+        for (;;) {
+            if(src==limit) {
+                return src;
+            }
+            if( (c=*src)<minNoMaybeCP ||
+                isCompYesAndZeroCC(norm16=UCPTRIE_FAST_BMP_GET(normTrie, UCPTRIE_16, c))
+            ) {
+                ++src;
+            } else {
+                prevSrc = src++;
+                if(!U16_IS_LEAD(c)) {
+                    break;
+                } else {
+                    UChar c2;
+                    if(src!=limit && U16_IS_TRAIL(c2=*src)) {
+                        ++src;
+                        c=U16_GET_SUPPLEMENTARY(c, c2);
+                        norm16=UCPTRIE_FAST_SUPP_GET(normTrie, UCPTRIE_16, c);
+                        if(!isCompYesAndZeroCC(norm16)) {
+                            break;
+                        }
+                    }
+                }
+            }
+        }
+        // isCompYesAndZeroCC(norm16) is false, that is, norm16>=minNoNo.
+        // The current character is either a "noNo" (has a mapping)
+        // or a "maybeYes" (combines backward)
+        // or a "yesYes" with ccc!=0.
+        // It is not a Hangul syllable or Jamo L because those have "yes" properties.
+
+        uint16_t prevNorm16 = INERT;
+        if (prevBoundary != prevSrc) {
+            if (norm16HasCompBoundaryBefore(norm16)) {
+                prevBoundary = prevSrc;
+            } else {
+                const UChar *p = prevSrc;
+                uint16_t n16;
+                UCPTRIE_FAST_U16_PREV(normTrie, UCPTRIE_16, prevBoundary, p, c, n16);
+                if (norm16HasCompBoundaryAfter(n16, onlyContiguous)) {
+                    prevBoundary = prevSrc;
+                } else {
+                    prevBoundary = p;
+                    prevNorm16 = n16;
+                }
+            }
+        }
+
+        if(isMaybeOrNonZeroCC(norm16)) {
+            uint8_t cc=getCCFromYesOrMaybe(norm16);
+            if (onlyContiguous /* FCC */ && cc != 0 &&
+                    getTrailCCFromCompYesAndZeroCC(prevNorm16) > cc) {
+                // The [prevBoundary..prevSrc[ character
+                // passed the quick check "yes && ccc==0" test
+                // but is out of canonical order with the current combining mark.
+            } else {
+                // If !onlyContiguous (not FCC), then we ignore the tccc of
+                // the previous character which passed the quick check "yes && ccc==0" test.
+                const UChar *nextSrc;
+                for (;;) {
+                    if (norm16 < MIN_YES_YES_WITH_CC) {
+                        if (pQCResult != nullptr) {
+                            *pQCResult = UNORM_MAYBE;
+                        } else {
+                            return prevBoundary;
+                        }
+                    }
+                    if (src == limit) {
+                        return src;
+                    }
+                    uint8_t prevCC = cc;
+                    nextSrc = src;
+                    UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, nextSrc, limit, c, norm16);
+                    if (isMaybeOrNonZeroCC(norm16)) {
+                        cc = getCCFromYesOrMaybe(norm16);
+                        if (!(prevCC <= cc || cc == 0)) {
+                            break;
+                        }
+                    } else {
+                        break;
+                    }
+                    src = nextSrc;
+                }
+                // src is after the last in-order combining mark.
+                if (isCompYesAndZeroCC(norm16)) {
+                    prevBoundary = src;
+                    src = nextSrc;
+                    continue;
+                }
+            }
+        }
+        if(pQCResult!=NULL) {
+            *pQCResult=UNORM_NO;
+        }
+        return prevBoundary;
+    }
+}
+
+void Normalizer2Impl::composeAndAppend(const UChar *src, const UChar *limit,
+                                       UBool doCompose,
+                                       UBool onlyContiguous,
+                                       UnicodeString &safeMiddle,
+                                       ReorderingBuffer &buffer,
+                                       UErrorCode &errorCode) const {
+    if(!buffer.isEmpty()) {
+        const UChar *firstStarterInSrc=findNextCompBoundary(src, limit, onlyContiguous);
+        if(src!=firstStarterInSrc) {
+            const UChar *lastStarterInDest=findPreviousCompBoundary(buffer.getStart(),
+                                                                    buffer.getLimit(), onlyContiguous);
+            int32_t destSuffixLength=(int32_t)(buffer.getLimit()-lastStarterInDest);
+            UnicodeString middle(lastStarterInDest, destSuffixLength);
+            buffer.removeSuffix(destSuffixLength);
+            safeMiddle=middle;
+            middle.append(src, (int32_t)(firstStarterInSrc-src));
+            const UChar *middleStart=middle.getBuffer();
+            compose(middleStart, middleStart+middle.length(), onlyContiguous,
+                    TRUE, buffer, errorCode);
+            if(U_FAILURE(errorCode)) {
+                return;
+            }
+            src=firstStarterInSrc;
+        }
+    }
+    if(doCompose) {
+        compose(src, limit, onlyContiguous, TRUE, buffer, errorCode);
+    } else {
+        if(limit==NULL) {  // appendZeroCC() needs limit!=NULL
+            limit=u_strchr(src, 0);
+        }
+        buffer.appendZeroCC(src, limit, errorCode);
+    }
+}
+
+UBool
+Normalizer2Impl::composeUTF8(uint32_t options, UBool onlyContiguous,
+                             const uint8_t *src, const uint8_t *limit,
+                             ByteSink *sink, Edits *edits, UErrorCode &errorCode) const {
+    U_ASSERT(limit != nullptr);
+    UnicodeString s16;
+    uint8_t minNoMaybeLead = leadByteForCP(minCompNoMaybeCP);
+    const uint8_t *prevBoundary = src;
+
+    for (;;) {
+        // Fast path: Scan over a sequence of characters below the minimum "no or maybe" code point,
+        // or with (compYes && ccc==0) properties.
+        const uint8_t *prevSrc;
+        uint16_t norm16 = 0;
+        for (;;) {
+            if (src == limit) {
+                if (prevBoundary != limit && sink != nullptr) {
+                    ByteSinkUtil::appendUnchanged(prevBoundary, limit,
+                                                  *sink, options, edits, errorCode);
+                }
+                return TRUE;
+            }
+            if (*src < minNoMaybeLead) {
+                ++src;
+            } else {
+                prevSrc = src;
+                UCPTRIE_FAST_U8_NEXT(normTrie, UCPTRIE_16, src, limit, norm16);
+                if (!isCompYesAndZeroCC(norm16)) {
+                    break;
+                }
+            }
+        }
+        // isCompYesAndZeroCC(norm16) is false, that is, norm16>=minNoNo.
+        // The current character is either a "noNo" (has a mapping)
+        // or a "maybeYes" (combines backward)
+        // or a "yesYes" with ccc!=0.
+        // It is not a Hangul syllable or Jamo L because those have "yes" properties.
+
+        // Medium-fast path: Handle cases that do not require full decomposition and recomposition.
+        if (!isMaybeOrNonZeroCC(norm16)) {  // minNoNo <= norm16 < minMaybeYes
+            if (sink == nullptr) {
+                return FALSE;
+            }
+            // Fast path for mapping a character that is immediately surrounded by boundaries.
+            // In this case, we need not decompose around the current character.
+            if (isDecompNoAlgorithmic(norm16)) {
+                // Maps to a single isCompYesAndZeroCC character
+                // which also implies hasCompBoundaryBefore.
+                if (norm16HasCompBoundaryAfter(norm16, onlyContiguous) ||
+                        hasCompBoundaryBefore(src, limit)) {
+                    if (prevBoundary != prevSrc &&
+                            !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                           *sink, options, edits, errorCode)) {
+                        break;
+                    }
+                    appendCodePointDelta(prevSrc, src, getAlgorithmicDelta(norm16), *sink, edits);
+                    prevBoundary = src;
+                    continue;
+                }
+            } else if (norm16 < minNoNoCompBoundaryBefore) {
+                // The mapping is comp-normalized which also implies hasCompBoundaryBefore.
+                if (norm16HasCompBoundaryAfter(norm16, onlyContiguous) ||
+                        hasCompBoundaryBefore(src, limit)) {
+                    if (prevBoundary != prevSrc &&
+                            !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                           *sink, options, edits, errorCode)) {
+                        break;
+                    }
+                    const uint16_t *mapping = getMapping(norm16);
+                    int32_t length = *mapping++ & MAPPING_LENGTH_MASK;
+                    if (!ByteSinkUtil::appendChange(prevSrc, src, (const UChar *)mapping, length,
+                                                    *sink, edits, errorCode)) {
+                        break;
+                    }
+                    prevBoundary = src;
+                    continue;
+                }
+            } else if (norm16 >= minNoNoEmpty) {
+                // The current character maps to nothing.
+                // Simply omit it from the output if there is a boundary before _or_ after it.
+                // The character itself implies no boundaries.
+                if (hasCompBoundaryBefore(src, limit) ||
+                        hasCompBoundaryAfter(prevBoundary, prevSrc, onlyContiguous)) {
+                    if (prevBoundary != prevSrc &&
+                            !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                           *sink, options, edits, errorCode)) {
+                        break;
+                    }
+                    if (edits != nullptr) {
+                        edits->addReplace((int32_t)(src - prevSrc), 0);
+                    }
+                    prevBoundary = src;
+                    continue;
+                }
+            }
+            // Other "noNo" type, or need to examine more text around this character:
+            // Fall through to the slow path.
+        } else if (isJamoVT(norm16)) {
+            // Jamo L: E1 84 80..92
+            // Jamo V: E1 85 A1..B5
+            // Jamo T: E1 86 A8..E1 87 82
+            U_ASSERT((src - prevSrc) == 3 && *prevSrc == 0xe1);
+            UChar32 prev = previousHangulOrJamo(prevBoundary, prevSrc);
+            if (prevSrc[1] == 0x85) {
+                // The current character is a Jamo Vowel,
+                // compose with previous Jamo L and following Jamo T.
+                UChar32 l = prev - Hangul::JAMO_L_BASE;
+                if ((uint32_t)l < Hangul::JAMO_L_COUNT) {
+                    if (sink == nullptr) {
+                        return FALSE;
+                    }
+                    int32_t t = getJamoTMinusBase(src, limit);
+                    if (t >= 0) {
+                        // The next character is a Jamo T.
+                        src += 3;
+                    } else if (hasCompBoundaryBefore(src, limit)) {
+                        // No Jamo T follows, not even via decomposition.
+                        t = 0;
+                    }
+                    if (t >= 0) {
+                        UChar32 syllable = Hangul::HANGUL_BASE +
+                            (l*Hangul::JAMO_V_COUNT + (prevSrc[2]-0xa1)) *
+                            Hangul::JAMO_T_COUNT + t;
+                        prevSrc -= 3;  // Replace the Jamo L as well.
+                        if (prevBoundary != prevSrc &&
+                                !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                               *sink, options, edits, errorCode)) {
+                            break;
+                        }
+                        ByteSinkUtil::appendCodePoint(prevSrc, src, syllable, *sink, edits);
+                        prevBoundary = src;
+                        continue;
+                    }
+                    // If we see L+V+x where x!=T then we drop to the slow path,
+                    // decompose and recompose.
+                    // This is to deal with NFKC finding normal L and V but a
+                    // compatibility variant of a T.
+                    // We need to either fully compose that combination here
+                    // (which would complicate the code and may not work with strange custom data)
+                    // or use the slow path.
+                }
+            } else if (Hangul::isHangulLV(prev)) {
+                // The current character is a Jamo Trailing consonant,
+                // compose with previous Hangul LV that does not contain a Jamo T.
+                if (sink == nullptr) {
+                    return FALSE;
+                }
+                UChar32 syllable = prev + getJamoTMinusBase(prevSrc, src);
+                prevSrc -= 3;  // Replace the Hangul LV as well.
+                if (prevBoundary != prevSrc &&
+                        !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                       *sink, options, edits, errorCode)) {
+                    break;
+                }
+                ByteSinkUtil::appendCodePoint(prevSrc, src, syllable, *sink, edits);
+                prevBoundary = src;
+                continue;
+            }
+            // No matching context, or may need to decompose surrounding text first:
+            // Fall through to the slow path.
+        } else if (norm16 > JAMO_VT) {  // norm16 >= MIN_YES_YES_WITH_CC
+            // One or more combining marks that do not combine-back:
+            // Check for canonical order, copy unchanged if ok and
+            // if followed by a character with a boundary-before.
+            uint8_t cc = getCCFromNormalYesOrMaybe(norm16);  // cc!=0
+            if (onlyContiguous /* FCC */ && getPreviousTrailCC(prevBoundary, prevSrc) > cc) {
+                // Fails FCD test, need to decompose and contiguously recompose.
+                if (sink == nullptr) {
+                    return FALSE;
+                }
+            } else {
+                // If !onlyContiguous (not FCC), then we ignore the tccc of
+                // the previous character which passed the quick check "yes && ccc==0" test.
+                const uint8_t *nextSrc;
+                uint16_t n16;
+                for (;;) {
+                    if (src == limit) {
+                        if (sink != nullptr) {
+                            ByteSinkUtil::appendUnchanged(prevBoundary, limit,
+                                                          *sink, options, edits, errorCode);
+                        }
+                        return TRUE;
+                    }
+                    uint8_t prevCC = cc;
+                    nextSrc = src;
+                    UCPTRIE_FAST_U8_NEXT(normTrie, UCPTRIE_16, nextSrc, limit, n16);
+                    if (n16 >= MIN_YES_YES_WITH_CC) {
+                        cc = getCCFromNormalYesOrMaybe(n16);
+                        if (prevCC > cc) {
+                            if (sink == nullptr) {
+                                return FALSE;
+                            }
+                            break;
+                        }
+                    } else {
+                        break;
+                    }
+                    src = nextSrc;
+                }
+                // src is after the last in-order combining mark.
+                // If there is a boundary here, then we continue with no change.
+                if (norm16HasCompBoundaryBefore(n16)) {
+                    if (isCompYesAndZeroCC(n16)) {
+                        src = nextSrc;
+                    }
+                    continue;
+                }
+                // Use the slow path. There is no boundary in [prevSrc, src[.
+            }
+        }
+
+        // Slow path: Find the nearest boundaries around the current character,
+        // decompose and recompose.
+        if (prevBoundary != prevSrc && !norm16HasCompBoundaryBefore(norm16)) {
+            const uint8_t *p = prevSrc;
+            UCPTRIE_FAST_U8_PREV(normTrie, UCPTRIE_16, prevBoundary, p, norm16);
+            if (!norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+                prevSrc = p;
+            }
+        }
+        ReorderingBuffer buffer(*this, s16, errorCode);
+        if (U_FAILURE(errorCode)) {
+            break;
+        }
+        // We know there is not a boundary here.
+        decomposeShort(prevSrc, src, FALSE /* !stopAtCompBoundary */, onlyContiguous,
+                       buffer, errorCode);
+        // Decompose until the next boundary.
+        src = decomposeShort(src, limit, TRUE /* stopAtCompBoundary */, onlyContiguous,
+                             buffer, errorCode);
+        if (U_FAILURE(errorCode)) {
+            break;
+        }
+        if ((src - prevSrc) > INT32_MAX) {  // guard before buffer.equals()
+            errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            return TRUE;
+        }
+        recompose(buffer, 0, onlyContiguous);
+        if (!buffer.equals(prevSrc, src)) {
+            if (sink == nullptr) {
+                return FALSE;
+            }
+            if (prevBoundary != prevSrc &&
+                    !ByteSinkUtil::appendUnchanged(prevBoundary, prevSrc,
+                                                   *sink, options, edits, errorCode)) {
+                break;
+            }
+            if (!ByteSinkUtil::appendChange(prevSrc, src, buffer.getStart(), buffer.length(),
+                                            *sink, edits, errorCode)) {
+                break;
+            }
+            prevBoundary = src;
+        }
+    }
+    return TRUE;
+}
+
+UBool Normalizer2Impl::hasCompBoundaryBefore(const UChar *src, const UChar *limit) const {
+    if (src == limit || *src < minCompNoMaybeCP) {
+        return TRUE;
+    }
+    UChar32 c;
+    uint16_t norm16;
+    UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, src, limit, c, norm16);
+    return norm16HasCompBoundaryBefore(norm16);
+}
+
+UBool Normalizer2Impl::hasCompBoundaryBefore(const uint8_t *src, const uint8_t *limit) const {
+    if (src == limit) {
+        return TRUE;
+    }
+    uint16_t norm16;
+    UCPTRIE_FAST_U8_NEXT(normTrie, UCPTRIE_16, src, limit, norm16);
+    return norm16HasCompBoundaryBefore(norm16);
+}
+
+UBool Normalizer2Impl::hasCompBoundaryAfter(const UChar *start, const UChar *p,
+                                            UBool onlyContiguous) const {
+    if (start == p) {
+        return TRUE;
+    }
+    UChar32 c;
+    uint16_t norm16;
+    UCPTRIE_FAST_U16_PREV(normTrie, UCPTRIE_16, start, p, c, norm16);
+    return norm16HasCompBoundaryAfter(norm16, onlyContiguous);
+}
+
+UBool Normalizer2Impl::hasCompBoundaryAfter(const uint8_t *start, const uint8_t *p,
+                                            UBool onlyContiguous) const {
+    if (start == p) {
+        return TRUE;
+    }
+    uint16_t norm16;
+    UCPTRIE_FAST_U8_PREV(normTrie, UCPTRIE_16, start, p, norm16);
+    return norm16HasCompBoundaryAfter(norm16, onlyContiguous);
+}
+
+const UChar *Normalizer2Impl::findPreviousCompBoundary(const UChar *start, const UChar *p,
+                                                       UBool onlyContiguous) const {
+    while (p != start) {
+        const UChar *codePointLimit = p;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_PREV(normTrie, UCPTRIE_16, start, p, c, norm16);
+        if (norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+            return codePointLimit;
+        }
+        if (hasCompBoundaryBefore(c, norm16)) {
+            return p;
+        }
+    }
+    return p;
+}
+
+const UChar *Normalizer2Impl::findNextCompBoundary(const UChar *p, const UChar *limit,
+                                                   UBool onlyContiguous) const {
+    while (p != limit) {
+        const UChar *codePointStart = p;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, p, limit, c, norm16);
+        if (hasCompBoundaryBefore(c, norm16)) {
+            return codePointStart;
+        }
+        if (norm16HasCompBoundaryAfter(norm16, onlyContiguous)) {
+            return p;
+        }
+    }
+    return p;
+}
+
+uint8_t Normalizer2Impl::getPreviousTrailCC(const UChar *start, const UChar *p) const {
+    if (start == p) {
+        return 0;
+    }
+    int32_t i = (int32_t)(p - start);
+    UChar32 c;
+    U16_PREV(start, 0, i, c);
+    return (uint8_t)getFCD16(c);
+}
+
+uint8_t Normalizer2Impl::getPreviousTrailCC(const uint8_t *start, const uint8_t *p) const {
+    if (start == p) {
+        return 0;
+    }
+    int32_t i = (int32_t)(p - start);
+    UChar32 c;
+    U8_PREV(start, 0, i, c);
+    return (uint8_t)getFCD16(c);
+}
+
+// Note: normalizer2impl.cpp r30982 (2011-nov-27)
+// still had getFCDTrie() which built and cached an FCD trie.
+// That provided faster access to FCD data than getFCD16FromNormData()
+// but required synchronization and consumed some 10kB of heap memory
+// in any process that uses FCD (e.g., via collation).
+// minDecompNoCP etc. and smallFCD[] are intended to help with any loss of performance,
+// at least for ASCII & CJK.
+
+// Ticket 20907 - The optimizer in MSVC/Visual Studio versions below 16.4 has trouble with this
+// function on Windows ARM64. As a work-around, we disable optimizations for this function.
+// This work-around could/should be removed once the following versions of Visual Studio are no
+// longer supported: All versions of VS2017, and versions of VS2019 below 16.4.
+#if (defined(_MSC_VER) && (defined(_M_ARM64)) && (_MSC_VER < 1924))
+#pragma optimize( "", off )
+#endif
+// Gets the FCD value from the regular normalization data.
+uint16_t Normalizer2Impl::getFCD16FromNormData(UChar32 c) const {
+    uint16_t norm16=getNorm16(c);
+    if (norm16 >= limitNoNo) {
+        if(norm16>=MIN_NORMAL_MAYBE_YES) {
+            // combining mark
+            norm16=getCCFromNormalYesOrMaybe(norm16);
+            return norm16|(norm16<<8);
+        } else if(norm16>=minMaybeYes) {
+            return 0;
+        } else {  // isDecompNoAlgorithmic(norm16)
+            uint16_t deltaTrailCC = norm16 & DELTA_TCCC_MASK;
+            if (deltaTrailCC <= DELTA_TCCC_1) {
+                return deltaTrailCC >> OFFSET_SHIFT;
+            }
+            // Maps to an isCompYesAndZeroCC.
+            c=mapAlgorithmic(c, norm16);
+            norm16=getRawNorm16(c);
+        }
+    }
+    if(norm16<=minYesNo || isHangulLVT(norm16)) {
+        // no decomposition or Hangul syllable, all zeros
+        return 0;
+    }
+    // c decomposes, get everything from the variable-length extra data
+    const uint16_t *mapping=getMapping(norm16);
+    uint16_t firstUnit=*mapping;
+    norm16=firstUnit>>8;  // tccc
+    if(firstUnit&MAPPING_HAS_CCC_LCCC_WORD) {
+        norm16|=*(mapping-1)&0xff00;  // lccc
+    }
+    return norm16;
+}
+#if (defined(_MSC_VER) && (defined(_M_ARM64)) && (_MSC_VER < 1924))
+#pragma optimize( "", on )
+#endif
+
+// Dual functionality:
+// buffer!=NULL: normalize
+// buffer==NULL: isNormalized/quickCheck/spanQuickCheckYes
+const UChar *
+Normalizer2Impl::makeFCD(const UChar *src, const UChar *limit,
+                         ReorderingBuffer *buffer,
+                         UErrorCode &errorCode) const {
+    // Tracks the last FCD-safe boundary, before lccc=0 or after properly-ordered tccc<=1.
+    // Similar to the prevBoundary in the compose() implementation.
+    const UChar *prevBoundary=src;
+    int32_t prevFCD16=0;
+    if(limit==NULL) {
+        src=copyLowPrefixFromNulTerminated(src, minLcccCP, buffer, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return src;
+        }
+        if(prevBoundary<src) {
+            prevBoundary=src;
+            // We know that the previous character's lccc==0.
+            // Fetching the fcd16 value was deferred for this below-U+0300 code point.
+            prevFCD16=getFCD16(*(src-1));
+            if(prevFCD16>1) {
+                --prevBoundary;
+            }
+        }
+        limit=u_strchr(src, 0);
+    }
+
+    // Note: In this function we use buffer->appendZeroCC() because we track
+    // the lead and trail combining classes here, rather than leaving it to
+    // the ReorderingBuffer.
+    // The exception is the call to decomposeShort() which uses the buffer
+    // in the normal way.
+
+    const UChar *prevSrc;
+    UChar32 c=0;
+    uint16_t fcd16=0;
+
+    for(;;) {
+        // count code units with lccc==0
+        for(prevSrc=src; src!=limit;) {
+            if((c=*src)<minLcccCP) {
+                prevFCD16=~c;
+                ++src;
+            } else if(!singleLeadMightHaveNonZeroFCD16(c)) {
+                prevFCD16=0;
+                ++src;
+            } else {
+                if(U16_IS_LEAD(c)) {
+                    UChar c2;
+                    if((src+1)!=limit && U16_IS_TRAIL(c2=src[1])) {
+                        c=U16_GET_SUPPLEMENTARY(c, c2);
+                    }
+                }
+                if((fcd16=getFCD16FromNormData(c))<=0xff) {
+                    prevFCD16=fcd16;
+                    src+=U16_LENGTH(c);
+                } else {
+                    break;
+                }
+            }
+        }
+        // copy these code units all at once
+        if(src!=prevSrc) {
+            if(buffer!=NULL && !buffer->appendZeroCC(prevSrc, src, errorCode)) {
+                break;
+            }
+            if(src==limit) {
+                break;
+            }
+            prevBoundary=src;
+            // We know that the previous character's lccc==0.
+            if(prevFCD16<0) {
+                // Fetching the fcd16 value was deferred for this below-minLcccCP code point.
+                UChar32 prev=~prevFCD16;
+                if(prev<minDecompNoCP) {
+                    prevFCD16=0;
+                } else {
+                    prevFCD16=getFCD16FromNormData(prev);
+                    if(prevFCD16>1) {
+                        --prevBoundary;
+                    }
+                }
+            } else {
+                const UChar *p=src-1;
+                if(U16_IS_TRAIL(*p) && prevSrc<p && U16_IS_LEAD(*(p-1))) {
+                    --p;
+                    // Need to fetch the previous character's FCD value because
+                    // prevFCD16 was just for the trail surrogate code point.
+                    prevFCD16=getFCD16FromNormData(U16_GET_SUPPLEMENTARY(p[0], p[1]));
+                    // Still known to have lccc==0 because its lead surrogate unit had lccc==0.
+                }
+                if(prevFCD16>1) {
+                    prevBoundary=p;
+                }
+            }
+            // The start of the current character (c).
+            prevSrc=src;
+        } else if(src==limit) {
+            break;
+        }
+
+        src+=U16_LENGTH(c);
+        // The current character (c) at [prevSrc..src[ has a non-zero lead combining class.
+        // Check for proper order, and decompose locally if necessary.
+        if((prevFCD16&0xff)<=(fcd16>>8)) {
+            // proper order: prev tccc <= current lccc
+            if((fcd16&0xff)<=1) {
+                prevBoundary=src;
+            }
+            if(buffer!=NULL && !buffer->appendZeroCC(c, errorCode)) {
+                break;
+            }
+            prevFCD16=fcd16;
+            continue;
+        } else if(buffer==NULL) {
+            return prevBoundary;  // quick check "no"
+        } else {
+            /*
+             * Back out the part of the source that we copied or appended
+             * already but is now going to be decomposed.
+             * prevSrc is set to after what was copied/appended.
+             */
+            buffer->removeSuffix((int32_t)(prevSrc-prevBoundary));
+            /*
+             * Find the part of the source that needs to be decomposed,
+             * up to the next safe boundary.
+             */
+            src=findNextFCDBoundary(src, limit);
+            /*
+             * The source text does not fulfill the conditions for FCD.
+             * Decompose and reorder a limited piece of the text.
+             */
+            decomposeShort(prevBoundary, src, FALSE, FALSE, *buffer, errorCode);
+            if (U_FAILURE(errorCode)) {
+                break;
+            }
+            prevBoundary=src;
+            prevFCD16=0;
+        }
+    }
+    return src;
+}
+
+void Normalizer2Impl::makeFCDAndAppend(const UChar *src, const UChar *limit,
+                                       UBool doMakeFCD,
+                                       UnicodeString &safeMiddle,
+                                       ReorderingBuffer &buffer,
+                                       UErrorCode &errorCode) const {
+    if(!buffer.isEmpty()) {
+        const UChar *firstBoundaryInSrc=findNextFCDBoundary(src, limit);
+        if(src!=firstBoundaryInSrc) {
+            const UChar *lastBoundaryInDest=findPreviousFCDBoundary(buffer.getStart(),
+                                                                    buffer.getLimit());
+            int32_t destSuffixLength=(int32_t)(buffer.getLimit()-lastBoundaryInDest);
+            UnicodeString middle(lastBoundaryInDest, destSuffixLength);
+            buffer.removeSuffix(destSuffixLength);
+            safeMiddle=middle;
+            middle.append(src, (int32_t)(firstBoundaryInSrc-src));
+            const UChar *middleStart=middle.getBuffer();
+            makeFCD(middleStart, middleStart+middle.length(), &buffer, errorCode);
+            if(U_FAILURE(errorCode)) {
+                return;
+            }
+            src=firstBoundaryInSrc;
+        }
+    }
+    if(doMakeFCD) {
+        makeFCD(src, limit, &buffer, errorCode);
+    } else {
+        if(limit==NULL) {  // appendZeroCC() needs limit!=NULL
+            limit=u_strchr(src, 0);
+        }
+        buffer.appendZeroCC(src, limit, errorCode);
+    }
+}
+
+const UChar *Normalizer2Impl::findPreviousFCDBoundary(const UChar *start, const UChar *p) const {
+    while(start<p) {
+        const UChar *codePointLimit = p;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_PREV(normTrie, UCPTRIE_16, start, p, c, norm16);
+        if (c < minDecompNoCP || norm16HasDecompBoundaryAfter(norm16)) {
+            return codePointLimit;
+        }
+        if (norm16HasDecompBoundaryBefore(norm16)) {
+            return p;
+        }
+    }
+    return p;
+}
+
+const UChar *Normalizer2Impl::findNextFCDBoundary(const UChar *p, const UChar *limit) const {
+    while(p<limit) {
+        const UChar *codePointStart=p;
+        UChar32 c;
+        uint16_t norm16;
+        UCPTRIE_FAST_U16_NEXT(normTrie, UCPTRIE_16, p, limit, c, norm16);
+        if (c < minLcccCP || norm16HasDecompBoundaryBefore(norm16)) {
+            return codePointStart;
+        }
+        if (norm16HasDecompBoundaryAfter(norm16)) {
+            return p;
+        }
+    }
+    return p;
+}
+
+// CanonicalIterator data -------------------------------------------------- ***
+
+CanonIterData::CanonIterData(UErrorCode &errorCode) :
+        mutableTrie(umutablecptrie_open(0, 0, &errorCode)), trie(nullptr),
+        canonStartSets(uprv_deleteUObject, NULL, errorCode) {}
+
+CanonIterData::~CanonIterData() {
+    umutablecptrie_close(mutableTrie);
+    ucptrie_close(trie);
+}
+
+void CanonIterData::addToStartSet(UChar32 origin, UChar32 decompLead, UErrorCode &errorCode) {
+    uint32_t canonValue = umutablecptrie_get(mutableTrie, decompLead);
+    if((canonValue&(CANON_HAS_SET|CANON_VALUE_MASK))==0 && origin!=0) {
+        // origin is the first character whose decomposition starts with
+        // the character for which we are setting the value.
+        umutablecptrie_set(mutableTrie, decompLead, canonValue|origin, &errorCode);
+    } else {
+        // origin is not the first character, or it is U+0000.
+        UnicodeSet *set;
+        if((canonValue&CANON_HAS_SET)==0) {
+            set=new UnicodeSet;
+            if(set==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            UChar32 firstOrigin=(UChar32)(canonValue&CANON_VALUE_MASK);
+            canonValue=(canonValue&~CANON_VALUE_MASK)|CANON_HAS_SET|(uint32_t)canonStartSets.size();
+            umutablecptrie_set(mutableTrie, decompLead, canonValue, &errorCode);
+            canonStartSets.addElement(set, errorCode);
+            if(firstOrigin!=0) {
+                set->add(firstOrigin);
+            }
+        } else {
+            set=(UnicodeSet *)canonStartSets[(int32_t)(canonValue&CANON_VALUE_MASK)];
+        }
+        set->add(origin);
+    }
+}
+
+// C++ class for friend access to private Normalizer2Impl members.
+class InitCanonIterData {
+public:
+    static void doInit(Normalizer2Impl *impl, UErrorCode &errorCode);
+};
+
+U_CDECL_BEGIN
+
+// UInitOnce instantiation function for CanonIterData
+static void U_CALLCONV
+initCanonIterData(Normalizer2Impl *impl, UErrorCode &errorCode) {
+    InitCanonIterData::doInit(impl, errorCode);
+}
+
+U_CDECL_END
+
+void InitCanonIterData::doInit(Normalizer2Impl *impl, UErrorCode &errorCode) {
+    U_ASSERT(impl->fCanonIterData == NULL);
+    impl->fCanonIterData = new CanonIterData(errorCode);
+    if (impl->fCanonIterData == NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_SUCCESS(errorCode)) {
+        UChar32 start = 0, end;
+        uint32_t value;
+        while ((end = ucptrie_getRange(impl->normTrie, start,
+                                       UCPMAP_RANGE_FIXED_LEAD_SURROGATES, Normalizer2Impl::INERT,
+                                       nullptr, nullptr, &value)) >= 0) {
+            // Call Normalizer2Impl::makeCanonIterDataFromNorm16() for a range of same-norm16 characters.
+            if (value != Normalizer2Impl::INERT) {
+                impl->makeCanonIterDataFromNorm16(start, end, value, *impl->fCanonIterData, errorCode);
+            }
+            start = end + 1;
+        }
+#ifdef UCPTRIE_DEBUG
+        umutablecptrie_setName(impl->fCanonIterData->mutableTrie, "CanonIterData");
+#endif
+        impl->fCanonIterData->trie = umutablecptrie_buildImmutable(
+            impl->fCanonIterData->mutableTrie, UCPTRIE_TYPE_SMALL, UCPTRIE_VALUE_BITS_32, &errorCode);
+        umutablecptrie_close(impl->fCanonIterData->mutableTrie);
+        impl->fCanonIterData->mutableTrie = nullptr;
+    }
+    if (U_FAILURE(errorCode)) {
+        delete impl->fCanonIterData;
+        impl->fCanonIterData = NULL;
+    }
+}
+
+void Normalizer2Impl::makeCanonIterDataFromNorm16(UChar32 start, UChar32 end, const uint16_t norm16,
+                                                  CanonIterData &newData,
+                                                  UErrorCode &errorCode) const {
+    if(isInert(norm16) || (minYesNo<=norm16 && norm16<minNoNo)) {
+        // Inert, or 2-way mapping (including Hangul syllable).
+        // We do not write a canonStartSet for any yesNo character.
+        // Composites from 2-way mappings are added at runtime from the
+        // starter's compositions list, and the other characters in
+        // 2-way mappings get CANON_NOT_SEGMENT_STARTER set because they are
+        // "maybe" characters.
+        return;
+    }
+    for(UChar32 c=start; c<=end; ++c) {
+        uint32_t oldValue = umutablecptrie_get(newData.mutableTrie, c);
+        uint32_t newValue=oldValue;
+        if(isMaybeOrNonZeroCC(norm16)) {
+            // not a segment starter if it occurs in a decomposition or has cc!=0
+            newValue|=CANON_NOT_SEGMENT_STARTER;
+            if(norm16<MIN_NORMAL_MAYBE_YES) {
+                newValue|=CANON_HAS_COMPOSITIONS;
+            }
+        } else if(norm16<minYesNo) {
+            newValue|=CANON_HAS_COMPOSITIONS;
+        } else {
+            // c has a one-way decomposition
+            UChar32 c2=c;
+            // Do not modify the whole-range norm16 value.
+            uint16_t norm16_2=norm16;
+            if (isDecompNoAlgorithmic(norm16_2)) {
+                // Maps to an isCompYesAndZeroCC.
+                c2 = mapAlgorithmic(c2, norm16_2);
+                norm16_2 = getRawNorm16(c2);
+                // No compatibility mappings for the CanonicalIterator.
+                U_ASSERT(!(isHangulLV(norm16_2) || isHangulLVT(norm16_2)));
+            }
+            if (norm16_2 > minYesNo) {
+                // c decomposes, get everything from the variable-length extra data
+                const uint16_t *mapping=getMapping(norm16_2);
+                uint16_t firstUnit=*mapping;
+                int32_t length=firstUnit&MAPPING_LENGTH_MASK;
+                if((firstUnit&MAPPING_HAS_CCC_LCCC_WORD)!=0) {
+                    if(c==c2 && (*(mapping-1)&0xff)!=0) {
+                        newValue|=CANON_NOT_SEGMENT_STARTER;  // original c has cc!=0
+                    }
+                }
+                // Skip empty mappings (no characters in the decomposition).
+                if(length!=0) {
+                    ++mapping;  // skip over the firstUnit
+                    // add c to first code point's start set
+                    int32_t i=0;
+                    U16_NEXT_UNSAFE(mapping, i, c2);
+                    newData.addToStartSet(c, c2, errorCode);
+                    // Set CANON_NOT_SEGMENT_STARTER for each remaining code point of a
+                    // one-way mapping. A 2-way mapping is possible here after
+                    // intermediate algorithmic mapping.
+                    if(norm16_2>=minNoNo) {
+                        while(i<length) {
+                            U16_NEXT_UNSAFE(mapping, i, c2);
+                            uint32_t c2Value = umutablecptrie_get(newData.mutableTrie, c2);
+                            if((c2Value&CANON_NOT_SEGMENT_STARTER)==0) {
+                                umutablecptrie_set(newData.mutableTrie, c2,
+                                                   c2Value|CANON_NOT_SEGMENT_STARTER, &errorCode);
+                            }
+                        }
+                    }
+                }
+            } else {
+                // c decomposed to c2 algorithmically; c has cc==0
+                newData.addToStartSet(c, c2, errorCode);
+            }
+        }
+        if(newValue!=oldValue) {
+            umutablecptrie_set(newData.mutableTrie, c, newValue, &errorCode);
+        }
+    }
+}
+
+UBool Normalizer2Impl::ensureCanonIterData(UErrorCode &errorCode) const {
+    // Logically const: Synchronized instantiation.
+    Normalizer2Impl *me=const_cast<Normalizer2Impl *>(this);
+    umtx_initOnce(me->fCanonIterDataInitOnce, &initCanonIterData, me, errorCode);
+    return U_SUCCESS(errorCode);
+}
+
+int32_t Normalizer2Impl::getCanonValue(UChar32 c) const {
+    return (int32_t)ucptrie_get(fCanonIterData->trie, c);
+}
+
+const UnicodeSet &Normalizer2Impl::getCanonStartSet(int32_t n) const {
+    return *(const UnicodeSet *)fCanonIterData->canonStartSets[n];
+}
+
+UBool Normalizer2Impl::isCanonSegmentStarter(UChar32 c) const {
+    return getCanonValue(c)>=0;
+}
+
+UBool Normalizer2Impl::getCanonStartSet(UChar32 c, UnicodeSet &set) const {
+    int32_t canonValue=getCanonValue(c)&~CANON_NOT_SEGMENT_STARTER;
+    if(canonValue==0) {
+        return FALSE;
+    }
+    set.clear();
+    int32_t value=canonValue&CANON_VALUE_MASK;
+    if((canonValue&CANON_HAS_SET)!=0) {
+        set.addAll(getCanonStartSet(value));
+    } else if(value!=0) {
+        set.add(value);
+    }
+    if((canonValue&CANON_HAS_COMPOSITIONS)!=0) {
+        uint16_t norm16=getRawNorm16(c);
+        if(norm16==JAMO_L) {
+            UChar32 syllable=
+                (UChar32)(Hangul::HANGUL_BASE+(c-Hangul::JAMO_L_BASE)*Hangul::JAMO_VT_COUNT);
+            set.add(syllable, syllable+Hangul::JAMO_VT_COUNT-1);
+        } else {
+            addComposites(getCompositionsList(norm16), set);
+        }
+    }
+    return TRUE;
+}
+
+U_NAMESPACE_END
+
+// Normalizer2 data swapping ----------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+U_CAPI int32_t U_EXPORT2
+unorm2_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    const int32_t *inIndexes;
+    int32_t indexes[Normalizer2Impl::IX_TOTAL_SIZE+1];
+
+    int32_t i, offset, nextOffset, size;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    uint8_t formatVersion0=pInfo->formatVersion[0];
+    if(!(
+        pInfo->dataFormat[0]==0x4e &&   /* dataFormat="Nrm2" */
+        pInfo->dataFormat[1]==0x72 &&
+        pInfo->dataFormat[2]==0x6d &&
+        pInfo->dataFormat[3]==0x32 &&
+        (1<=formatVersion0 && formatVersion0<=4)
+    )) {
+        udata_printError(ds, "unorm2_swap(): data format %02x.%02x.%02x.%02x (format version %02x) is not recognized as Normalizer2 data\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData+headerSize;
+    outBytes=(uint8_t *)outData+headerSize;
+
+    inIndexes=(const int32_t *)inBytes;
+    int32_t minIndexesLength;
+    if(formatVersion0==1) {
+        minIndexesLength=Normalizer2Impl::IX_MIN_MAYBE_YES+1;
+    } else if(formatVersion0==2) {
+        minIndexesLength=Normalizer2Impl::IX_MIN_YES_NO_MAPPINGS_ONLY+1;
+    } else {
+        minIndexesLength=Normalizer2Impl::IX_MIN_LCCC_CP+1;
+    }
+
+    if(length>=0) {
+        length-=headerSize;
+        if(length<minIndexesLength*4) {
+            udata_printError(ds, "unorm2_swap(): too few bytes (%d after header) for Normalizer2 data\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    /* read the first few indexes */
+    for(i=0; i<UPRV_LENGTHOF(indexes); ++i) {
+        indexes[i]=udata_readInt32(ds, inIndexes[i]);
+    }
+
+    /* get the total length of the data */
+    size=indexes[Normalizer2Impl::IX_TOTAL_SIZE];
+
+    if(length>=0) {
+        if(length<size) {
+            udata_printError(ds, "unorm2_swap(): too few bytes (%d after header) for all of Normalizer2 data\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        /* copy the data for inaccessible bytes */
+        if(inBytes!=outBytes) {
+            uprv_memcpy(outBytes, inBytes, size);
+        }
+
+        offset=0;
+
+        /* swap the int32_t indexes[] */
+        nextOffset=indexes[Normalizer2Impl::IX_NORM_TRIE_OFFSET];
+        ds->swapArray32(ds, inBytes, nextOffset-offset, outBytes, pErrorCode);
+        offset=nextOffset;
+
+        /* swap the trie */
+        nextOffset=indexes[Normalizer2Impl::IX_EXTRA_DATA_OFFSET];
+        utrie_swapAnyVersion(ds, inBytes+offset, nextOffset-offset, outBytes+offset, pErrorCode);
+        offset=nextOffset;
+
+        /* swap the uint16_t extraData[] */
+        nextOffset=indexes[Normalizer2Impl::IX_SMALL_FCD_OFFSET];
+        ds->swapArray16(ds, inBytes+offset, nextOffset-offset, outBytes+offset, pErrorCode);
+        offset=nextOffset;
+
+        /* no need to swap the uint8_t smallFCD[] (new in formatVersion 2) */
+        nextOffset=indexes[Normalizer2Impl::IX_SMALL_FCD_OFFSET+1];
+        offset=nextOffset;
+
+        U_ASSERT(offset==size);
+    }
+
+    return headerSize+size;
+}
+
+#endif  // !UCONFIG_NO_NORMALIZATION
diff --git a/thirdparty/icu4c/common/normalizer2impl.h b/thirdparty/icu4c/common/normalizer2impl.h
new file mode 100644
index 0000000000..4218a30a34
--- /dev/null
+++ b/thirdparty/icu4c/common/normalizer2impl.h
@@ -0,0 +1,978 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  normalizer2impl.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009nov22
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __NORMALIZER2IMPL_H__
+#define __NORMALIZER2IMPL_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/normalizer2.h"
+#include "unicode/ucptrie.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "mutex.h"
+#include "udataswp.h"
+#include "uset_imp.h"
+
+// When the nfc.nrm data is *not* hardcoded into the common library
+// (with this constant set to 0),
+// then it needs to be built into the data package:
+// Add nfc.nrm to icu4c/source/data/Makefile.in DAT_FILES_SHORT
+#define NORM2_HARDCODE_NFC_DATA 1
+
+U_NAMESPACE_BEGIN
+
+struct CanonIterData;
+
+class ByteSink;
+class Edits;
+class InitCanonIterData;
+class LcccContext;
+
+class U_COMMON_API Hangul {
+public:
+    /* Korean Hangul and Jamo constants */
+    enum {
+        JAMO_L_BASE=0x1100,     /* "lead" jamo */
+        JAMO_L_END=0x1112,
+        JAMO_V_BASE=0x1161,     /* "vowel" jamo */
+        JAMO_V_END=0x1175,
+        JAMO_T_BASE=0x11a7,     /* "trail" jamo */
+        JAMO_T_END=0x11c2,
+
+        HANGUL_BASE=0xac00,
+        HANGUL_END=0xd7a3,
+
+        JAMO_L_COUNT=19,
+        JAMO_V_COUNT=21,
+        JAMO_T_COUNT=28,
+
+        JAMO_VT_COUNT=JAMO_V_COUNT*JAMO_T_COUNT,
+
+        HANGUL_COUNT=JAMO_L_COUNT*JAMO_V_COUNT*JAMO_T_COUNT,
+        HANGUL_LIMIT=HANGUL_BASE+HANGUL_COUNT
+    };
+
+    static inline UBool isHangul(UChar32 c) {
+        return HANGUL_BASE<=c && c<HANGUL_LIMIT;
+    }
+    static inline UBool
+    isHangulLV(UChar32 c) {
+        c-=HANGUL_BASE;
+        return 0<=c && c<HANGUL_COUNT && c%JAMO_T_COUNT==0;
+    }
+    static inline UBool isJamoL(UChar32 c) {
+        return (uint32_t)(c-JAMO_L_BASE)<JAMO_L_COUNT;
+    }
+    static inline UBool isJamoV(UChar32 c) {
+        return (uint32_t)(c-JAMO_V_BASE)<JAMO_V_COUNT;
+    }
+    static inline UBool isJamoT(UChar32 c) {
+        int32_t t=c-JAMO_T_BASE;
+        return 0<t && t<JAMO_T_COUNT;  // not JAMO_T_BASE itself
+    }
+    static UBool isJamo(UChar32 c) {
+        return JAMO_L_BASE<=c && c<=JAMO_T_END &&
+            (c<=JAMO_L_END || (JAMO_V_BASE<=c && c<=JAMO_V_END) || JAMO_T_BASE<c);
+    }
+
+    /**
+     * Decomposes c, which must be a Hangul syllable, into buffer
+     * and returns the length of the decomposition (2 or 3).
+     */
+    static inline int32_t decompose(UChar32 c, UChar buffer[3]) {
+        c-=HANGUL_BASE;
+        UChar32 c2=c%JAMO_T_COUNT;
+        c/=JAMO_T_COUNT;
+        buffer[0]=(UChar)(JAMO_L_BASE+c/JAMO_V_COUNT);
+        buffer[1]=(UChar)(JAMO_V_BASE+c%JAMO_V_COUNT);
+        if(c2==0) {
+            return 2;
+        } else {
+            buffer[2]=(UChar)(JAMO_T_BASE+c2);
+            return 3;
+        }
+    }
+
+    /**
+     * Decomposes c, which must be a Hangul syllable, into buffer.
+     * This is the raw, not recursive, decomposition. Its length is always 2.
+     */
+    static inline void getRawDecomposition(UChar32 c, UChar buffer[2]) {
+        UChar32 orig=c;
+        c-=HANGUL_BASE;
+        UChar32 c2=c%JAMO_T_COUNT;
+        if(c2==0) {
+            c/=JAMO_T_COUNT;
+            buffer[0]=(UChar)(JAMO_L_BASE+c/JAMO_V_COUNT);
+            buffer[1]=(UChar)(JAMO_V_BASE+c%JAMO_V_COUNT);
+        } else {
+            buffer[0]=(UChar)(orig-c2);  // LV syllable
+            buffer[1]=(UChar)(JAMO_T_BASE+c2);
+        }
+    }
+private:
+    Hangul();  // no instantiation
+};
+
+class Normalizer2Impl;
+
+class U_COMMON_API ReorderingBuffer : public UMemory {
+public:
+    /** Constructs only; init() should be called. */
+    ReorderingBuffer(const Normalizer2Impl &ni, UnicodeString &dest) :
+        impl(ni), str(dest),
+        start(NULL), reorderStart(NULL), limit(NULL),
+        remainingCapacity(0), lastCC(0) {}
+    /** Constructs, removes the string contents, and initializes for a small initial capacity. */
+    ReorderingBuffer(const Normalizer2Impl &ni, UnicodeString &dest, UErrorCode &errorCode);
+    ~ReorderingBuffer() {
+        if(start!=NULL) {
+            str.releaseBuffer((int32_t)(limit-start));
+        }
+    }
+    UBool init(int32_t destCapacity, UErrorCode &errorCode);
+
+    UBool isEmpty() const { return start==limit; }
+    int32_t length() const { return (int32_t)(limit-start); }
+    UChar *getStart() { return start; }
+    UChar *getLimit() { return limit; }
+    uint8_t getLastCC() const { return lastCC; }
+
+    UBool equals(const UChar *start, const UChar *limit) const;
+    UBool equals(const uint8_t *otherStart, const uint8_t *otherLimit) const;
+
+    UBool append(UChar32 c, uint8_t cc, UErrorCode &errorCode) {
+        return (c<=0xffff) ?
+            appendBMP((UChar)c, cc, errorCode) :
+            appendSupplementary(c, cc, errorCode);
+    }
+    UBool append(const UChar *s, int32_t length, UBool isNFD,
+                 uint8_t leadCC, uint8_t trailCC,
+                 UErrorCode &errorCode);
+    UBool appendBMP(UChar c, uint8_t cc, UErrorCode &errorCode) {
+        if(remainingCapacity==0 && !resize(1, errorCode)) {
+            return false;
+        }
+        if(lastCC<=cc || cc==0) {
+            *limit++=c;
+            lastCC=cc;
+            if(cc<=1) {
+                reorderStart=limit;
+            }
+        } else {
+            insert(c, cc);
+        }
+        --remainingCapacity;
+        return true;
+    }
+    UBool appendZeroCC(UChar32 c, UErrorCode &errorCode);
+    UBool appendZeroCC(const UChar *s, const UChar *sLimit, UErrorCode &errorCode);
+    void remove();
+    void removeSuffix(int32_t suffixLength);
+    void setReorderingLimit(UChar *newLimit) {
+        remainingCapacity+=(int32_t)(limit-newLimit);
+        reorderStart=limit=newLimit;
+        lastCC=0;
+    }
+    void copyReorderableSuffixTo(UnicodeString &s) const {
+        s.setTo(ConstChar16Ptr(reorderStart), (int32_t)(limit-reorderStart));
+    }
+private:
+    /*
+     * TODO: Revisit whether it makes sense to track reorderStart.
+     * It is set to after the last known character with cc<=1,
+     * which stops previousCC() before it reads that character and looks up its cc.
+     * previousCC() is normally only called from insert().
+     * In other words, reorderStart speeds up the insertion of a combining mark
+     * into a multi-combining mark sequence where it does not belong at the end.
+     * This might not be worth the trouble.
+     * On the other hand, it's not a huge amount of trouble.
+     *
+     * We probably need it for UNORM_SIMPLE_APPEND.
+     */
+
+    UBool appendSupplementary(UChar32 c, uint8_t cc, UErrorCode &errorCode);
+    void insert(UChar32 c, uint8_t cc);
+    static void writeCodePoint(UChar *p, UChar32 c) {
+        if(c<=0xffff) {
+            *p=(UChar)c;
+        } else {
+            p[0]=U16_LEAD(c);
+            p[1]=U16_TRAIL(c);
+        }
+    }
+    UBool resize(int32_t appendLength, UErrorCode &errorCode);
+
+    const Normalizer2Impl &impl;
+    UnicodeString &str;
+    UChar *start, *reorderStart, *limit;
+    int32_t remainingCapacity;
+    uint8_t lastCC;
+
+    // private backward iterator
+    void setIterator() { codePointStart=limit; }
+    void skipPrevious();  // Requires start<codePointStart.
+    uint8_t previousCC();  // Returns 0 if there is no previous character.
+
+    UChar *codePointStart, *codePointLimit;
+};
+
+/**
+ * Low-level implementation of the Unicode Normalization Algorithm.
+ * For the data structure and details see the documentation at the end of
+ * this normalizer2impl.h and in the design doc at
+ * http://site.icu-project.org/design/normalization/custom
+ */
+class U_COMMON_API Normalizer2Impl : public UObject {
+public:
+    Normalizer2Impl() : normTrie(NULL), fCanonIterData(NULL) { }
+    virtual ~Normalizer2Impl();
+
+    void init(const int32_t *inIndexes, const UCPTrie *inTrie,
+              const uint16_t *inExtraData, const uint8_t *inSmallFCD);
+
+    void addLcccChars(UnicodeSet &set) const;
+    void addPropertyStarts(const USetAdder *sa, UErrorCode &errorCode) const;
+    void addCanonIterPropertyStarts(const USetAdder *sa, UErrorCode &errorCode) const;
+
+    // low-level properties ------------------------------------------------ ***
+
+    UBool ensureCanonIterData(UErrorCode &errorCode) const;
+
+    // The trie stores values for lead surrogate code *units*.
+    // Surrogate code *points* are inert.
+    uint16_t getNorm16(UChar32 c) const {
+        return U_IS_LEAD(c) ?
+            static_cast<uint16_t>(INERT) :
+            UCPTRIE_FAST_GET(normTrie, UCPTRIE_16, c);
+    }
+    uint16_t getRawNorm16(UChar32 c) const { return UCPTRIE_FAST_GET(normTrie, UCPTRIE_16, c); }
+
+    UNormalizationCheckResult getCompQuickCheck(uint16_t norm16) const {
+        if(norm16<minNoNo || MIN_YES_YES_WITH_CC<=norm16) {
+            return UNORM_YES;
+        } else if(minMaybeYes<=norm16) {
+            return UNORM_MAYBE;
+        } else {
+            return UNORM_NO;
+        }
+    }
+    UBool isAlgorithmicNoNo(uint16_t norm16) const { return limitNoNo<=norm16 && norm16<minMaybeYes; }
+    UBool isCompNo(uint16_t norm16) const { return minNoNo<=norm16 && norm16<minMaybeYes; }
+    UBool isDecompYes(uint16_t norm16) const { return norm16<minYesNo || minMaybeYes<=norm16; }
+
+    uint8_t getCC(uint16_t norm16) const {
+        if(norm16>=MIN_NORMAL_MAYBE_YES) {
+            return getCCFromNormalYesOrMaybe(norm16);
+        }
+        if(norm16<minNoNo || limitNoNo<=norm16) {
+            return 0;
+        }
+        return getCCFromNoNo(norm16);
+    }
+    static uint8_t getCCFromNormalYesOrMaybe(uint16_t norm16) {
+        return (uint8_t)(norm16 >> OFFSET_SHIFT);
+    }
+    static uint8_t getCCFromYesOrMaybe(uint16_t norm16) {
+        return norm16>=MIN_NORMAL_MAYBE_YES ? getCCFromNormalYesOrMaybe(norm16) : 0;
+    }
+    uint8_t getCCFromYesOrMaybeCP(UChar32 c) const {
+        if (c < minCompNoMaybeCP) { return 0; }
+        return getCCFromYesOrMaybe(getNorm16(c));
+    }
+
+    /**
+     * Returns the FCD data for code point c.
+     * @param c A Unicode code point.
+     * @return The lccc(c) in bits 15..8 and tccc(c) in bits 7..0.
+     */
+    uint16_t getFCD16(UChar32 c) const {
+        if(c<minDecompNoCP) {
+            return 0;
+        } else if(c<=0xffff) {
+            if(!singleLeadMightHaveNonZeroFCD16(c)) { return 0; }
+        }
+        return getFCD16FromNormData(c);
+    }
+    /**
+     * Returns the FCD data for the next code point (post-increment).
+     * Might skip only a lead surrogate rather than the whole surrogate pair if none of
+     * the supplementary code points associated with the lead surrogate have non-zero FCD data.
+     * @param s A valid pointer into a string. Requires s!=limit.
+     * @param limit The end of the string, or NULL.
+     * @return The lccc(c) in bits 15..8 and tccc(c) in bits 7..0.
+     */
+    uint16_t nextFCD16(const UChar *&s, const UChar *limit) const {
+        UChar32 c=*s++;
+        if(c<minDecompNoCP || !singleLeadMightHaveNonZeroFCD16(c)) {
+            return 0;
+        }
+        UChar c2;
+        if(U16_IS_LEAD(c) && s!=limit && U16_IS_TRAIL(c2=*s)) {
+            c=U16_GET_SUPPLEMENTARY(c, c2);
+            ++s;
+        }
+        return getFCD16FromNormData(c);
+    }
+    /**
+     * Returns the FCD data for the previous code point (pre-decrement).
+     * @param start The start of the string.
+     * @param s A valid pointer into a string. Requires start<s.
+     * @return The lccc(c) in bits 15..8 and tccc(c) in bits 7..0.
+     */
+    uint16_t previousFCD16(const UChar *start, const UChar *&s) const {
+        UChar32 c=*--s;
+        if(c<minDecompNoCP) {
+            return 0;
+        }
+        if(!U16_IS_TRAIL(c)) {
+            if(!singleLeadMightHaveNonZeroFCD16(c)) {
+                return 0;
+            }
+        } else {
+            UChar c2;
+            if(start<s && U16_IS_LEAD(c2=*(s-1))) {
+                c=U16_GET_SUPPLEMENTARY(c2, c);
+                --s;
+            }
+        }
+        return getFCD16FromNormData(c);
+    }
+
+    /** Returns true if the single-or-lead code unit c might have non-zero FCD data. */
+    UBool singleLeadMightHaveNonZeroFCD16(UChar32 lead) const {
+        // 0<=lead<=0xffff
+        uint8_t bits=smallFCD[lead>>8];
+        if(bits==0) { return false; }
+        return (UBool)((bits>>((lead>>5)&7))&1);
+    }
+    /** Returns the FCD value from the regular normalization data. */
+    uint16_t getFCD16FromNormData(UChar32 c) const;
+
+    /**
+     * Gets the decomposition for one code point.
+     * @param c code point
+     * @param buffer out-only buffer for algorithmic decompositions
+     * @param length out-only, takes the length of the decomposition, if any
+     * @return pointer to the decomposition, or NULL if none
+     */
+    const UChar *getDecomposition(UChar32 c, UChar buffer[4], int32_t &length) const;
+
+    /**
+     * Gets the raw decomposition for one code point.
+     * @param c code point
+     * @param buffer out-only buffer for algorithmic decompositions
+     * @param length out-only, takes the length of the decomposition, if any
+     * @return pointer to the decomposition, or NULL if none
+     */
+    const UChar *getRawDecomposition(UChar32 c, UChar buffer[30], int32_t &length) const;
+
+    UChar32 composePair(UChar32 a, UChar32 b) const;
+
+    UBool isCanonSegmentStarter(UChar32 c) const;
+    UBool getCanonStartSet(UChar32 c, UnicodeSet &set) const;
+
+    enum {
+        // Fixed norm16 values.
+        MIN_YES_YES_WITH_CC=0xfe02,
+        JAMO_VT=0xfe00,
+        MIN_NORMAL_MAYBE_YES=0xfc00,
+        JAMO_L=2,  // offset=1 hasCompBoundaryAfter=false
+        INERT=1,  // offset=0 hasCompBoundaryAfter=true
+
+        // norm16 bit 0 is comp-boundary-after.
+        HAS_COMP_BOUNDARY_AFTER=1,
+        OFFSET_SHIFT=1,
+
+        // For algorithmic one-way mappings, norm16 bits 2..1 indicate the
+        // tccc (0, 1, >1) for quick FCC boundary-after tests.
+        DELTA_TCCC_0=0,
+        DELTA_TCCC_1=2,
+        DELTA_TCCC_GT_1=4,
+        DELTA_TCCC_MASK=6,
+        DELTA_SHIFT=3,
+
+        MAX_DELTA=0x40
+    };
+
+    enum {
+        // Byte offsets from the start of the data, after the generic header.
+        IX_NORM_TRIE_OFFSET,
+        IX_EXTRA_DATA_OFFSET,
+        IX_SMALL_FCD_OFFSET,
+        IX_RESERVED3_OFFSET,
+        IX_RESERVED4_OFFSET,
+        IX_RESERVED5_OFFSET,
+        IX_RESERVED6_OFFSET,
+        IX_TOTAL_SIZE,
+
+        // Code point thresholds for quick check codes.
+        IX_MIN_DECOMP_NO_CP,
+        IX_MIN_COMP_NO_MAYBE_CP,
+
+        // Norm16 value thresholds for quick check combinations and types of extra data.
+
+        /** Mappings & compositions in [minYesNo..minYesNoMappingsOnly[. */
+        IX_MIN_YES_NO,
+        /** Mappings are comp-normalized. */
+        IX_MIN_NO_NO,
+        IX_LIMIT_NO_NO,
+        IX_MIN_MAYBE_YES,
+
+        /** Mappings only in [minYesNoMappingsOnly..minNoNo[. */
+        IX_MIN_YES_NO_MAPPINGS_ONLY,
+        /** Mappings are not comp-normalized but have a comp boundary before. */
+        IX_MIN_NO_NO_COMP_BOUNDARY_BEFORE,
+        /** Mappings do not have a comp boundary before. */
+        IX_MIN_NO_NO_COMP_NO_MAYBE_CC,
+        /** Mappings to the empty string. */
+        IX_MIN_NO_NO_EMPTY,
+
+        IX_MIN_LCCC_CP,
+        IX_RESERVED19,
+        IX_COUNT
+    };
+
+    enum {
+        MAPPING_HAS_CCC_LCCC_WORD=0x80,
+        MAPPING_HAS_RAW_MAPPING=0x40,
+        // unused bit 0x20,
+        MAPPING_LENGTH_MASK=0x1f
+    };
+
+    enum {
+        COMP_1_LAST_TUPLE=0x8000,
+        COMP_1_TRIPLE=1,
+        COMP_1_TRAIL_LIMIT=0x3400,
+        COMP_1_TRAIL_MASK=0x7ffe,
+        COMP_1_TRAIL_SHIFT=9,  // 10-1 for the "triple" bit
+        COMP_2_TRAIL_SHIFT=6,
+        COMP_2_TRAIL_MASK=0xffc0
+    };
+
+    // higher-level functionality ------------------------------------------ ***
+
+    // NFD without an NFD Normalizer2 instance.
+    UnicodeString &decompose(const UnicodeString &src, UnicodeString &dest,
+                             UErrorCode &errorCode) const;
+    /**
+     * Decomposes [src, limit[ and writes the result to dest.
+     * limit can be NULL if src is NUL-terminated.
+     * destLengthEstimate is the initial dest buffer capacity and can be -1.
+     */
+    void decompose(const UChar *src, const UChar *limit,
+                   UnicodeString &dest, int32_t destLengthEstimate,
+                   UErrorCode &errorCode) const;
+
+    const UChar *decompose(const UChar *src, const UChar *limit,
+                           ReorderingBuffer *buffer, UErrorCode &errorCode) const;
+    void decomposeAndAppend(const UChar *src, const UChar *limit,
+                            UBool doDecompose,
+                            UnicodeString &safeMiddle,
+                            ReorderingBuffer &buffer,
+                            UErrorCode &errorCode) const;
+    UBool compose(const UChar *src, const UChar *limit,
+                  UBool onlyContiguous,
+                  UBool doCompose,
+                  ReorderingBuffer &buffer,
+                  UErrorCode &errorCode) const;
+    const UChar *composeQuickCheck(const UChar *src, const UChar *limit,
+                                   UBool onlyContiguous,
+                                   UNormalizationCheckResult *pQCResult) const;
+    void composeAndAppend(const UChar *src, const UChar *limit,
+                          UBool doCompose,
+                          UBool onlyContiguous,
+                          UnicodeString &safeMiddle,
+                          ReorderingBuffer &buffer,
+                          UErrorCode &errorCode) const;
+
+    /** sink==nullptr: isNormalized() */
+    UBool composeUTF8(uint32_t options, UBool onlyContiguous,
+                      const uint8_t *src, const uint8_t *limit,
+                      ByteSink *sink, icu::Edits *edits, UErrorCode &errorCode) const;
+
+    const UChar *makeFCD(const UChar *src, const UChar *limit,
+                         ReorderingBuffer *buffer, UErrorCode &errorCode) const;
+    void makeFCDAndAppend(const UChar *src, const UChar *limit,
+                          UBool doMakeFCD,
+                          UnicodeString &safeMiddle,
+                          ReorderingBuffer &buffer,
+                          UErrorCode &errorCode) const;
+
+    UBool hasDecompBoundaryBefore(UChar32 c) const;
+    UBool norm16HasDecompBoundaryBefore(uint16_t norm16) const;
+    UBool hasDecompBoundaryAfter(UChar32 c) const;
+    UBool norm16HasDecompBoundaryAfter(uint16_t norm16) const;
+    UBool isDecompInert(UChar32 c) const { return isDecompYesAndZeroCC(getNorm16(c)); }
+
+    UBool hasCompBoundaryBefore(UChar32 c) const {
+        return c<minCompNoMaybeCP || norm16HasCompBoundaryBefore(getNorm16(c));
+    }
+    UBool hasCompBoundaryAfter(UChar32 c, UBool onlyContiguous) const {
+        return norm16HasCompBoundaryAfter(getNorm16(c), onlyContiguous);
+    }
+    UBool isCompInert(UChar32 c, UBool onlyContiguous) const {
+        uint16_t norm16=getNorm16(c);
+        return isCompYesAndZeroCC(norm16) &&
+            (norm16 & HAS_COMP_BOUNDARY_AFTER) != 0 &&
+            (!onlyContiguous || isInert(norm16) || *getMapping(norm16) <= 0x1ff);
+    }
+
+    UBool hasFCDBoundaryBefore(UChar32 c) const { return hasDecompBoundaryBefore(c); }
+    UBool hasFCDBoundaryAfter(UChar32 c) const { return hasDecompBoundaryAfter(c); }
+    UBool isFCDInert(UChar32 c) const { return getFCD16(c)<=1; }
+private:
+    friend class InitCanonIterData;
+    friend class LcccContext;
+
+    UBool isMaybe(uint16_t norm16) const { return minMaybeYes<=norm16 && norm16<=JAMO_VT; }
+    UBool isMaybeOrNonZeroCC(uint16_t norm16) const { return norm16>=minMaybeYes; }
+    static UBool isInert(uint16_t norm16) { return norm16==INERT; }
+    static UBool isJamoL(uint16_t norm16) { return norm16==JAMO_L; }
+    static UBool isJamoVT(uint16_t norm16) { return norm16==JAMO_VT; }
+    uint16_t hangulLVT() const { return minYesNoMappingsOnly|HAS_COMP_BOUNDARY_AFTER; }
+    UBool isHangulLV(uint16_t norm16) const { return norm16==minYesNo; }
+    UBool isHangulLVT(uint16_t norm16) const {
+        return norm16==hangulLVT();
+    }
+    UBool isCompYesAndZeroCC(uint16_t norm16) const { return norm16<minNoNo; }
+    // UBool isCompYes(uint16_t norm16) const {
+    //     return norm16>=MIN_YES_YES_WITH_CC || norm16<minNoNo;
+    // }
+    // UBool isCompYesOrMaybe(uint16_t norm16) const {
+    //     return norm16<minNoNo || minMaybeYes<=norm16;
+    // }
+    // UBool hasZeroCCFromDecompYes(uint16_t norm16) const {
+    //     return norm16<=MIN_NORMAL_MAYBE_YES || norm16==JAMO_VT;
+    // }
+    UBool isDecompYesAndZeroCC(uint16_t norm16) const {
+        return norm16<minYesNo ||
+               norm16==JAMO_VT ||
+               (minMaybeYes<=norm16 && norm16<=MIN_NORMAL_MAYBE_YES);
+    }
+    /**
+     * A little faster and simpler than isDecompYesAndZeroCC() but does not include
+     * the MaybeYes which combine-forward and have ccc=0.
+     * (Standard Unicode 10 normalization does not have such characters.)
+     */
+    UBool isMostDecompYesAndZeroCC(uint16_t norm16) const {
+        return norm16<minYesNo || norm16==MIN_NORMAL_MAYBE_YES || norm16==JAMO_VT;
+    }
+    UBool isDecompNoAlgorithmic(uint16_t norm16) const { return norm16>=limitNoNo; }
+
+    // For use with isCompYes().
+    // Perhaps the compiler can combine the two tests for MIN_YES_YES_WITH_CC.
+    // static uint8_t getCCFromYes(uint16_t norm16) {
+    //     return norm16>=MIN_YES_YES_WITH_CC ? getCCFromNormalYesOrMaybe(norm16) : 0;
+    // }
+    uint8_t getCCFromNoNo(uint16_t norm16) const {
+        const uint16_t *mapping=getMapping(norm16);
+        if(*mapping&MAPPING_HAS_CCC_LCCC_WORD) {
+            return (uint8_t)*(mapping-1);
+        } else {
+            return 0;
+        }
+    }
+    // requires that the [cpStart..cpLimit[ character passes isCompYesAndZeroCC()
+    uint8_t getTrailCCFromCompYesAndZeroCC(uint16_t norm16) const {
+        if(norm16<=minYesNo) {
+            return 0;  // yesYes and Hangul LV have ccc=tccc=0
+        } else {
+            // For Hangul LVT we harmlessly fetch a firstUnit with tccc=0 here.
+            return (uint8_t)(*getMapping(norm16)>>8);  // tccc from yesNo
+        }
+    }
+    uint8_t getPreviousTrailCC(const UChar *start, const UChar *p) const;
+    uint8_t getPreviousTrailCC(const uint8_t *start, const uint8_t *p) const;
+
+    // Requires algorithmic-NoNo.
+    UChar32 mapAlgorithmic(UChar32 c, uint16_t norm16) const {
+        return c+(norm16>>DELTA_SHIFT)-centerNoNoDelta;
+    }
+    UChar32 getAlgorithmicDelta(uint16_t norm16) const {
+        return (norm16>>DELTA_SHIFT)-centerNoNoDelta;
+    }
+
+    // Requires minYesNo<norm16<limitNoNo.
+    const uint16_t *getMapping(uint16_t norm16) const { return extraData+(norm16>>OFFSET_SHIFT); }
+    const uint16_t *getCompositionsListForDecompYes(uint16_t norm16) const {
+        if(norm16<JAMO_L || MIN_NORMAL_MAYBE_YES<=norm16) {
+            return NULL;
+        } else if(norm16<minMaybeYes) {
+            return getMapping(norm16);  // for yesYes; if Jamo L: harmless empty list
+        } else {
+            return maybeYesCompositions+norm16-minMaybeYes;
+        }
+    }
+    const uint16_t *getCompositionsListForComposite(uint16_t norm16) const {
+        // A composite has both mapping & compositions list.
+        const uint16_t *list=getMapping(norm16);
+        return list+  // mapping pointer
+            1+  // +1 to skip the first unit with the mapping length
+            (*list&MAPPING_LENGTH_MASK);  // + mapping length
+    }
+    const uint16_t *getCompositionsListForMaybe(uint16_t norm16) const {
+        // minMaybeYes<=norm16<MIN_NORMAL_MAYBE_YES
+        return maybeYesCompositions+((norm16-minMaybeYes)>>OFFSET_SHIFT);
+    }
+    /**
+     * @param c code point must have compositions
+     * @return compositions list pointer
+     */
+    const uint16_t *getCompositionsList(uint16_t norm16) const {
+        return isDecompYes(norm16) ?
+                getCompositionsListForDecompYes(norm16) :
+                getCompositionsListForComposite(norm16);
+    }
+
+    const UChar *copyLowPrefixFromNulTerminated(const UChar *src,
+                                                UChar32 minNeedDataCP,
+                                                ReorderingBuffer *buffer,
+                                                UErrorCode &errorCode) const;
+    const UChar *decomposeShort(const UChar *src, const UChar *limit,
+                                UBool stopAtCompBoundary, UBool onlyContiguous,
+                                ReorderingBuffer &buffer, UErrorCode &errorCode) const;
+    UBool decompose(UChar32 c, uint16_t norm16,
+                    ReorderingBuffer &buffer, UErrorCode &errorCode) const;
+
+    const uint8_t *decomposeShort(const uint8_t *src, const uint8_t *limit,
+                                  UBool stopAtCompBoundary, UBool onlyContiguous,
+                                  ReorderingBuffer &buffer, UErrorCode &errorCode) const;
+
+    static int32_t combine(const uint16_t *list, UChar32 trail);
+    void addComposites(const uint16_t *list, UnicodeSet &set) const;
+    void recompose(ReorderingBuffer &buffer, int32_t recomposeStartIndex,
+                   UBool onlyContiguous) const;
+
+    UBool hasCompBoundaryBefore(UChar32 c, uint16_t norm16) const {
+        return c<minCompNoMaybeCP || norm16HasCompBoundaryBefore(norm16);
+    }
+    UBool norm16HasCompBoundaryBefore(uint16_t norm16) const  {
+        return norm16 < minNoNoCompNoMaybeCC || isAlgorithmicNoNo(norm16);
+    }
+    UBool hasCompBoundaryBefore(const UChar *src, const UChar *limit) const;
+    UBool hasCompBoundaryBefore(const uint8_t *src, const uint8_t *limit) const;
+    UBool hasCompBoundaryAfter(const UChar *start, const UChar *p,
+                               UBool onlyContiguous) const;
+    UBool hasCompBoundaryAfter(const uint8_t *start, const uint8_t *p,
+                               UBool onlyContiguous) const;
+    UBool norm16HasCompBoundaryAfter(uint16_t norm16, UBool onlyContiguous) const {
+        return (norm16 & HAS_COMP_BOUNDARY_AFTER) != 0 &&
+            (!onlyContiguous || isTrailCC01ForCompBoundaryAfter(norm16));
+    }
+    /** For FCC: Given norm16 HAS_COMP_BOUNDARY_AFTER, does it have tccc<=1? */
+    UBool isTrailCC01ForCompBoundaryAfter(uint16_t norm16) const {
+        return isInert(norm16) || (isDecompNoAlgorithmic(norm16) ?
+            (norm16 & DELTA_TCCC_MASK) <= DELTA_TCCC_1 : *getMapping(norm16) <= 0x1ff);
+    }
+
+    const UChar *findPreviousCompBoundary(const UChar *start, const UChar *p, UBool onlyContiguous) const;
+    const UChar *findNextCompBoundary(const UChar *p, const UChar *limit, UBool onlyContiguous) const;
+
+    const UChar *findPreviousFCDBoundary(const UChar *start, const UChar *p) const;
+    const UChar *findNextFCDBoundary(const UChar *p, const UChar *limit) const;
+
+    void makeCanonIterDataFromNorm16(UChar32 start, UChar32 end, const uint16_t norm16,
+                                     CanonIterData &newData, UErrorCode &errorCode) const;
+
+    int32_t getCanonValue(UChar32 c) const;
+    const UnicodeSet &getCanonStartSet(int32_t n) const;
+
+    // UVersionInfo dataVersion;
+
+    // BMP code point thresholds for quick check loops looking at single UTF-16 code units.
+    UChar minDecompNoCP;
+    UChar minCompNoMaybeCP;
+    UChar minLcccCP;
+
+    // Norm16 value thresholds for quick check combinations and types of extra data.
+    uint16_t minYesNo;
+    uint16_t minYesNoMappingsOnly;
+    uint16_t minNoNo;
+    uint16_t minNoNoCompBoundaryBefore;
+    uint16_t minNoNoCompNoMaybeCC;
+    uint16_t minNoNoEmpty;
+    uint16_t limitNoNo;
+    uint16_t centerNoNoDelta;
+    uint16_t minMaybeYes;
+
+    const UCPTrie *normTrie;
+    const uint16_t *maybeYesCompositions;
+    const uint16_t *extraData;  // mappings and/or compositions for yesYes, yesNo & noNo characters
+    const uint8_t *smallFCD;  // [0x100] one bit per 32 BMP code points, set if any FCD!=0
+
+    UInitOnce       fCanonIterDataInitOnce = U_INITONCE_INITIALIZER;
+    CanonIterData  *fCanonIterData;
+};
+
+// bits in canonIterData
+#define CANON_NOT_SEGMENT_STARTER 0x80000000
+#define CANON_HAS_COMPOSITIONS 0x40000000
+#define CANON_HAS_SET 0x200000
+#define CANON_VALUE_MASK 0x1fffff
+
+/**
+ * ICU-internal shortcut for quick access to standard Unicode normalization.
+ */
+class U_COMMON_API Normalizer2Factory {
+public:
+    static const Normalizer2 *getFCDInstance(UErrorCode &errorCode);
+    static const Normalizer2 *getFCCInstance(UErrorCode &errorCode);
+    static const Normalizer2 *getNoopInstance(UErrorCode &errorCode);
+
+    static const Normalizer2 *getInstance(UNormalizationMode mode, UErrorCode &errorCode);
+
+    static const Normalizer2Impl *getNFCImpl(UErrorCode &errorCode);
+    static const Normalizer2Impl *getNFKCImpl(UErrorCode &errorCode);
+    static const Normalizer2Impl *getNFKC_CFImpl(UErrorCode &errorCode);
+
+    // Get the Impl instance of the Normalizer2.
+    // Must be used only when it is known that norm2 is a Normalizer2WithImpl instance.
+    static const Normalizer2Impl *getImpl(const Normalizer2 *norm2);
+private:
+    Normalizer2Factory();  // No instantiation.
+};
+
+U_NAMESPACE_END
+
+U_CAPI int32_t U_EXPORT2
+unorm2_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode);
+
+/**
+ * Get the NF*_QC property for a code point, for u_getIntPropertyValue().
+ * @internal
+ */
+U_CFUNC UNormalizationCheckResult
+unorm_getQuickCheck(UChar32 c, UNormalizationMode mode);
+
+/**
+ * Gets the 16-bit FCD value (lead & trail CCs) for a code point, for u_getIntPropertyValue().
+ * @internal
+ */
+U_CFUNC uint16_t
+unorm_getFCD16(UChar32 c);
+
+/**
+ * Format of Normalizer2 .nrm data files.
+ * Format version 4.0.
+ *
+ * Normalizer2 .nrm data files provide data for the Unicode Normalization algorithms.
+ * ICU ships with data files for standard Unicode Normalization Forms
+ * NFC and NFD (nfc.nrm), NFKC and NFKD (nfkc.nrm) and NFKC_Casefold (nfkc_cf.nrm).
+ * Custom (application-specific) data can be built into additional .nrm files
+ * with the gennorm2 build tool.
+ * ICU ships with one such file, uts46.nrm, for the implementation of UTS #46.
+ *
+ * Normalizer2.getInstance() causes a .nrm file to be loaded, unless it has been
+ * cached already. Internally, Normalizer2Impl.load() reads the .nrm file.
+ *
+ * A .nrm file begins with a standard ICU data file header
+ * (DataHeader, see ucmndata.h and unicode/udata.h).
+ * The UDataInfo.dataVersion field usually contains the Unicode version
+ * for which the data was generated.
+ *
+ * After the header, the file contains the following parts.
+ * Constants are defined as enum values of the Normalizer2Impl class.
+ *
+ * Many details of the data structures are described in the design doc
+ * which is at http://site.icu-project.org/design/normalization/custom
+ *
+ * int32_t indexes[indexesLength]; -- indexesLength=indexes[IX_NORM_TRIE_OFFSET]/4;
+ *
+ *      The first eight indexes are byte offsets in ascending order.
+ *      Each byte offset marks the start of the next part in the data file,
+ *      and the end of the previous one.
+ *      When two consecutive byte offsets are the same, then the corresponding part is empty.
+ *      Byte offsets are offsets from after the header,
+ *      that is, from the beginning of the indexes[].
+ *      Each part starts at an offset with proper alignment for its data.
+ *      If necessary, the previous part may include padding bytes to achieve this alignment.
+ *
+ *      minDecompNoCP=indexes[IX_MIN_DECOMP_NO_CP] is the lowest code point
+ *      with a decomposition mapping, that is, with NF*D_QC=No.
+ *      minCompNoMaybeCP=indexes[IX_MIN_COMP_NO_MAYBE_CP] is the lowest code point
+ *      with NF*C_QC=No (has a one-way mapping) or Maybe (combines backward).
+ *      minLcccCP=indexes[IX_MIN_LCCC_CP] (index 18, new in formatVersion 3)
+ *      is the lowest code point with lccc!=0.
+ *
+ *      The next eight indexes are thresholds of 16-bit trie values for ranges of
+ *      values indicating multiple normalization properties.
+ *      They are listed here in threshold order, not in the order they are stored in the indexes.
+ *          minYesNo=indexes[IX_MIN_YES_NO];
+ *          minYesNoMappingsOnly=indexes[IX_MIN_YES_NO_MAPPINGS_ONLY];
+ *          minNoNo=indexes[IX_MIN_NO_NO];
+ *          minNoNoCompBoundaryBefore=indexes[IX_MIN_NO_NO_COMP_BOUNDARY_BEFORE];
+ *          minNoNoCompNoMaybeCC=indexes[IX_MIN_NO_NO_COMP_NO_MAYBE_CC];
+ *          minNoNoEmpty=indexes[IX_MIN_NO_NO_EMPTY];
+ *          limitNoNo=indexes[IX_LIMIT_NO_NO];
+ *          minMaybeYes=indexes[IX_MIN_MAYBE_YES];
+ *      See the normTrie description below and the design doc for details.
+ *
+ * UCPTrie normTrie; -- see ucptrie_impl.h and ucptrie.h, same as Java CodePointTrie
+ *
+ *      The trie holds the main normalization data. Each code point is mapped to a 16-bit value.
+ *      Rather than using independent bits in the value (which would require more than 16 bits),
+ *      information is extracted primarily via range checks.
+ *      Except, format version 3 uses bit 0 for hasCompBoundaryAfter().
+ *      For example, a 16-bit value norm16 in the range minYesNo<=norm16<minNoNo
+ *      means that the character has NF*C_QC=Yes and NF*D_QC=No properties,
+ *      which means it has a two-way (round-trip) decomposition mapping.
+ *      Values in the range 2<=norm16<limitNoNo are also directly indexes into the extraData
+ *      pointing to mappings, compositions lists, or both.
+ *      Value norm16==INERT (0 in versions 1 & 2, 1 in version 3)
+ *      means that the character is normalization-inert, that is,
+ *      it does not have a mapping, does not participate in composition, has a zero
+ *      canonical combining class, and forms a boundary where text before it and after it
+ *      can be normalized independently.
+ *      For details about how multiple properties are encoded in 16-bit values
+ *      see the design doc.
+ *      Note that the encoding cannot express all combinations of the properties involved;
+ *      it only supports those combinations that are allowed by
+ *      the Unicode Normalization algorithms. Details are in the design doc as well.
+ *      The gennorm2 tool only builds .nrm files for data that conforms to the limitations.
+ *
+ *      The trie has a value for each lead surrogate code unit representing the "worst case"
+ *      properties of the 1024 supplementary characters whose UTF-16 form starts with
+ *      the lead surrogate. If all of the 1024 supplementary characters are normalization-inert,
+ *      then their lead surrogate code unit has the trie value INERT.
+ *      When the lead surrogate unit's value exceeds the quick check minimum during processing,
+ *      the properties for the full supplementary code point need to be looked up.
+ *
+ * uint16_t maybeYesCompositions[MIN_NORMAL_MAYBE_YES-minMaybeYes];
+ * uint16_t extraData[];
+ *
+ *      There is only one byte offset for the end of these two arrays.
+ *      The split between them is given by the constant and variable mentioned above.
+ *      In version 3, the difference must be shifted right by OFFSET_SHIFT.
+ *
+ *      The maybeYesCompositions array contains compositions lists for characters that
+ *      combine both forward (as starters in composition pairs)
+ *      and backward (as trailing characters in composition pairs).
+ *      Such characters do not occur in Unicode 5.2 but are allowed by
+ *      the Unicode Normalization algorithms.
+ *      If there are no such characters, then minMaybeYes==MIN_NORMAL_MAYBE_YES
+ *      and the maybeYesCompositions array is empty.
+ *      If there are such characters, then minMaybeYes is subtracted from their norm16 values
+ *      to get the index into this array.
+ *
+ *      The extraData array contains compositions lists for "YesYes" characters,
+ *      followed by mappings and optional compositions lists for "YesNo" characters,
+ *      followed by only mappings for "NoNo" characters.
+ *      (Referring to pairs of NFC/NFD quick check values.)
+ *      The norm16 values of those characters are directly indexes into the extraData array.
+ *      In version 3, the norm16 values must be shifted right by OFFSET_SHIFT
+ *      for accessing extraData.
+ *
+ *      The data structures for compositions lists and mappings are described in the design doc.
+ *
+ * uint8_t smallFCD[0x100]; -- new in format version 2
+ *
+ *      This is a bit set to help speed up FCD value lookups in the absence of a full
+ *      UTrie2 or other large data structure with the full FCD value mapping.
+ *
+ *      Each smallFCD bit is set if any of the corresponding 32 BMP code points
+ *      has a non-zero FCD value (lccc!=0 or tccc!=0).
+ *      Bit 0 of smallFCD[0] is for U+0000..U+001F. Bit 7 of smallFCD[0xff] is for U+FFE0..U+FFFF.
+ *      A bit for 32 lead surrogates is set if any of the 32k corresponding
+ *      _supplementary_ code points has a non-zero FCD value.
+ *
+ *      This bit set is most useful for the large blocks of CJK characters with FCD=0.
+ *
+ * Changes from format version 1 to format version 2 ---------------------------
+ *
+ * - Addition of data for raw (not recursively decomposed) mappings.
+ *   + The MAPPING_NO_COMP_BOUNDARY_AFTER bit in the extraData is now also set when
+ *     the mapping is to an empty string or when the character combines-forward.
+ *     This subsumes the one actual use of the MAPPING_PLUS_COMPOSITION_LIST bit which
+ *     is then repurposed for the MAPPING_HAS_RAW_MAPPING bit.
+ *   + For details see the design doc.
+ * - Addition of indexes[IX_MIN_YES_NO_MAPPINGS_ONLY] and separation of the yesNo extraData into
+ *   distinct ranges (combines-forward vs. not)
+ *   so that a range check can be used to find out if there is a compositions list.
+ *   This is fully equivalent with formatVersion 1's MAPPING_PLUS_COMPOSITION_LIST flag.
+ *   It is needed for the new (in ICU 49) composePair(), not for other normalization.
+ * - Addition of the smallFCD[] bit set.
+ *
+ * Changes from format version 2 to format version 3 (ICU 60) ------------------
+ *
+ * - norm16 bit 0 indicates hasCompBoundaryAfter(),
+ *   except that for contiguous composition (FCC) the tccc must be checked as well.
+ *   Data indexes and ccc values are shifted left by one (OFFSET_SHIFT).
+ *   Thresholds like minNoNo are tested before shifting.
+ *
+ * - Algorithmic mapping deltas are shifted left by two more bits (total DELTA_SHIFT),
+ *   to make room for two bits (three values) indicating whether the tccc is 0, 1, or greater.
+ *   See DELTA_TCCC_MASK etc.
+ *   This helps with fetching tccc/FCD values and FCC hasCompBoundaryAfter().
+ *   minMaybeYes is 8-aligned so that the DELTA_TCCC_MASK bits can be tested directly.
+ *
+ * - Algorithmic mappings are only used for mapping to "comp yes and ccc=0" characters,
+ *   and ASCII characters are mapped algorithmically only to other ASCII characters.
+ *   This helps with hasCompBoundaryBefore() and compose() fast paths.
+ *   It is never necessary any more to loop for algorithmic mappings.
+ *
+ * - Addition of indexes[IX_MIN_NO_NO_COMP_BOUNDARY_BEFORE],
+ *   indexes[IX_MIN_NO_NO_COMP_NO_MAYBE_CC], and indexes[IX_MIN_NO_NO_EMPTY],
+ *   and separation of the noNo extraData into distinct ranges.
+ *   With this, the noNo norm16 value indicates whether the mapping is
+ *   compose-normalized, not normalized but hasCompBoundaryBefore(),
+ *   not even that, or maps to an empty string.
+ *   hasCompBoundaryBefore() can be determined solely from the norm16 value.
+ *
+ * - The norm16 value for Hangul LVT is now different from that for Hangul LV,
+ *   so that hasCompBoundaryAfter() need not check for the syllable type.
+ *   For Hangul LV, minYesNo continues to be used (no comp-boundary-after).
+ *   For Hangul LVT, minYesNoMappingsOnly|HAS_COMP_BOUNDARY_AFTER is used.
+ *   The extraData units at these indexes are set to firstUnit=2 and firstUnit=3, respectively,
+ *   to simplify some code.
+ *
+ * - The extraData firstUnit bit 5 is no longer necessary
+ *   (norm16 bit 0 used instead of firstUnit MAPPING_NO_COMP_BOUNDARY_AFTER),
+ *   is reserved again, and always set to 0.
+ *
+ * - Addition of indexes[IX_MIN_LCCC_CP], the first code point where lccc!=0.
+ *   This used to be hardcoded to U+0300, but in data like NFKC_Casefold it is lower:
+ *   U+00AD Soft Hyphen maps to an empty string,
+ *   which is artificially assigned "worst case" values lccc=1 and tccc=255.
+ *
+ * - A mapping to an empty string has explicit lccc=1 and tccc=255 values.
+ *
+ * Changes from format version 3 to format version 4 (ICU 63) ------------------
+ *
+ * Switched from UTrie2 to UCPTrie/CodePointTrie.
+ *
+ * The new trie no longer stores different values for surrogate code *units* vs.
+ * surrogate code *points*.
+ * Lead surrogates still have values for optimized UTF-16 string processing.
+ * When looking up code point properties, the code now checks for lead surrogates and
+ * treats them as inert.
+ *
+ * gennorm2 now has to reject mappings for surrogate code points.
+ * UTS #46 maps unpaired surrogates to U+FFFD in code rather than via its
+ * custom normalization data file.
+ */
+
+#endif  /* !UCONFIG_NO_NORMALIZATION */
+#endif  /* __NORMALIZER2IMPL_H__ */
diff --git a/thirdparty/icu4c/common/normlzr.cpp b/thirdparty/icu4c/common/normlzr.cpp
new file mode 100644
index 0000000000..2dea0ffc33
--- /dev/null
+++ b/thirdparty/icu4c/common/normlzr.cpp
@@ -0,0 +1,529 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *************************************************************************
+ * COPYRIGHT: 
+ * Copyright (c) 1996-2012, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ *************************************************************************
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/uniset.h"
+#include "unicode/unistr.h"
+#include "unicode/chariter.h"
+#include "unicode/schriter.h"
+#include "unicode/uchriter.h"
+#include "unicode/normlzr.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "normalizer2impl.h"
+#include "uprops.h"  // for uniset_getUnicode32Instance()
+
+#if defined(move32)
+ // System can define move32 intrinsics, but the char iters define move32 method
+ // using same undef trick in headers, so undef here to re-enable the method.
+#undef move32
+#endif
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(Normalizer)
+
+//-------------------------------------------------------------------------
+// Constructors and other boilerplate
+//-------------------------------------------------------------------------
+
+Normalizer::Normalizer(const UnicodeString& str, UNormalizationMode mode) :
+    UObject(), fFilteredNorm2(NULL), fNorm2(NULL), fUMode(mode), fOptions(0),
+    text(new StringCharacterIterator(str)),
+    currentIndex(0), nextIndex(0),
+    buffer(), bufferPos(0)
+{
+    init();
+}
+
+Normalizer::Normalizer(ConstChar16Ptr str, int32_t length, UNormalizationMode mode) :
+    UObject(), fFilteredNorm2(NULL), fNorm2(NULL), fUMode(mode), fOptions(0),
+    text(new UCharCharacterIterator(str, length)),
+    currentIndex(0), nextIndex(0),
+    buffer(), bufferPos(0)
+{
+    init();
+}
+
+Normalizer::Normalizer(const CharacterIterator& iter, UNormalizationMode mode) :
+    UObject(), fFilteredNorm2(NULL), fNorm2(NULL), fUMode(mode), fOptions(0),
+    text(iter.clone()),
+    currentIndex(0), nextIndex(0),
+    buffer(), bufferPos(0)
+{
+    init();
+}
+
+Normalizer::Normalizer(const Normalizer &copy) :
+    UObject(copy), fFilteredNorm2(NULL), fNorm2(NULL), fUMode(copy.fUMode), fOptions(copy.fOptions),
+    text(copy.text->clone()),
+    currentIndex(copy.currentIndex), nextIndex(copy.nextIndex),
+    buffer(copy.buffer), bufferPos(copy.bufferPos)
+{
+    init();
+}
+
+void
+Normalizer::init() {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    fNorm2=Normalizer2Factory::getInstance(fUMode, errorCode);
+    if(fOptions&UNORM_UNICODE_3_2) {
+        delete fFilteredNorm2;
+        fNorm2=fFilteredNorm2=
+            new FilteredNormalizer2(*fNorm2, *uniset_getUnicode32Instance(errorCode));
+    }
+    if(U_FAILURE(errorCode)) {
+        errorCode=U_ZERO_ERROR;
+        fNorm2=Normalizer2Factory::getNoopInstance(errorCode);
+    }
+}
+
+Normalizer::~Normalizer()
+{
+    delete fFilteredNorm2;
+    delete text;
+}
+
+Normalizer* 
+Normalizer::clone() const
+{
+    return new Normalizer(*this);
+}
+
+/**
+ * Generates a hash code for this iterator.
+ */
+int32_t Normalizer::hashCode() const
+{
+    return text->hashCode() + fUMode + fOptions + buffer.hashCode() + bufferPos + currentIndex + nextIndex;
+}
+    
+UBool Normalizer::operator==(const Normalizer& that) const
+{
+    return
+        this==&that ||
+        (fUMode==that.fUMode &&
+        fOptions==that.fOptions &&
+        *text==*that.text &&
+        buffer==that.buffer &&
+        bufferPos==that.bufferPos &&
+        nextIndex==that.nextIndex);
+}
+
+//-------------------------------------------------------------------------
+// Static utility methods
+//-------------------------------------------------------------------------
+
+void U_EXPORT2
+Normalizer::normalize(const UnicodeString& source, 
+                      UNormalizationMode mode, int32_t options,
+                      UnicodeString& result, 
+                      UErrorCode &status) {
+    if(source.isBogus() || U_FAILURE(status)) {
+        result.setToBogus();
+        if(U_SUCCESS(status)) {
+            status=U_ILLEGAL_ARGUMENT_ERROR;
+        }
+    } else {
+        UnicodeString localDest;
+        UnicodeString *dest;
+
+        if(&source!=&result) {
+            dest=&result;
+        } else {
+            // the source and result strings are the same object, use a temporary one
+            dest=&localDest;
+        }
+        const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, status);
+        if(U_SUCCESS(status)) {
+            if(options&UNORM_UNICODE_3_2) {
+                FilteredNormalizer2(*n2, *uniset_getUnicode32Instance(status)).
+                    normalize(source, *dest, status);
+            } else {
+                n2->normalize(source, *dest, status);
+            }
+        }
+        if(dest==&localDest && U_SUCCESS(status)) {
+            result=*dest;
+        }
+    }
+}
+
+void U_EXPORT2
+Normalizer::compose(const UnicodeString& source, 
+                    UBool compat, int32_t options,
+                    UnicodeString& result, 
+                    UErrorCode &status) {
+    normalize(source, compat ? UNORM_NFKC : UNORM_NFC, options, result, status);
+}
+
+void U_EXPORT2
+Normalizer::decompose(const UnicodeString& source, 
+                      UBool compat, int32_t options,
+                      UnicodeString& result, 
+                      UErrorCode &status) {
+    normalize(source, compat ? UNORM_NFKD : UNORM_NFD, options, result, status);
+}
+
+UNormalizationCheckResult
+Normalizer::quickCheck(const UnicodeString& source,
+                       UNormalizationMode mode, int32_t options,
+                       UErrorCode &status) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, status);
+    if(U_SUCCESS(status)) {
+        if(options&UNORM_UNICODE_3_2) {
+            return FilteredNormalizer2(*n2, *uniset_getUnicode32Instance(status)).
+                quickCheck(source, status);
+        } else {
+            return n2->quickCheck(source, status);
+        }
+    } else {
+        return UNORM_MAYBE;
+    }
+}
+
+UBool
+Normalizer::isNormalized(const UnicodeString& source,
+                         UNormalizationMode mode, int32_t options,
+                         UErrorCode &status) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, status);
+    if(U_SUCCESS(status)) {
+        if(options&UNORM_UNICODE_3_2) {
+            return FilteredNormalizer2(*n2, *uniset_getUnicode32Instance(status)).
+                isNormalized(source, status);
+        } else {
+            return n2->isNormalized(source, status);
+        }
+    } else {
+        return FALSE;
+    }
+}
+
+UnicodeString & U_EXPORT2
+Normalizer::concatenate(const UnicodeString &left, const UnicodeString &right,
+                        UnicodeString &result,
+                        UNormalizationMode mode, int32_t options,
+                        UErrorCode &errorCode) {
+    if(left.isBogus() || right.isBogus() || U_FAILURE(errorCode)) {
+        result.setToBogus();
+        if(U_SUCCESS(errorCode)) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        }
+    } else {
+        UnicodeString localDest;
+        UnicodeString *dest;
+
+        if(&right!=&result) {
+            dest=&result;
+        } else {
+            // the right and result strings are the same object, use a temporary one
+            dest=&localDest;
+        }
+        *dest=left;
+        const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, errorCode);
+        if(U_SUCCESS(errorCode)) {
+            if(options&UNORM_UNICODE_3_2) {
+                FilteredNormalizer2(*n2, *uniset_getUnicode32Instance(errorCode)).
+                    append(*dest, right, errorCode);
+            } else {
+                n2->append(*dest, right, errorCode);
+            }
+        }
+        if(dest==&localDest && U_SUCCESS(errorCode)) {
+            result=*dest;
+        }
+    }
+    return result;
+}
+
+//-------------------------------------------------------------------------
+// Iteration API
+//-------------------------------------------------------------------------
+
+/**
+ * Return the current character in the normalized text.
+ */
+UChar32 Normalizer::current() {
+    if(bufferPos<buffer.length() || nextNormalize()) {
+        return buffer.char32At(bufferPos);
+    } else {
+        return DONE;
+    }
+}
+
+/**
+ * Return the next character in the normalized text and advance
+ * the iteration position by one.  If the end
+ * of the text has already been reached, {@link #DONE} is returned.
+ */
+UChar32 Normalizer::next() {
+    if(bufferPos<buffer.length() ||  nextNormalize()) {
+        UChar32 c=buffer.char32At(bufferPos);
+        bufferPos+=U16_LENGTH(c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+/**
+ * Return the previous character in the normalized text and decrement
+ * the iteration position by one.  If the beginning
+ * of the text has already been reached, {@link #DONE} is returned.
+ */
+UChar32 Normalizer::previous() {
+    if(bufferPos>0 || previousNormalize()) {
+        UChar32 c=buffer.char32At(bufferPos-1);
+        bufferPos-=U16_LENGTH(c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+void Normalizer::reset() {
+    currentIndex=nextIndex=text->setToStart();
+    clearBuffer();
+}
+
+void
+Normalizer::setIndexOnly(int32_t index) {
+    text->setIndex(index);  // pins index
+    currentIndex=nextIndex=text->getIndex();
+    clearBuffer();
+}
+
+/**
+ * Return the first character in the normalized text.  This resets
+ * the <tt>Normalizer's</tt> position to the beginning of the text.
+ */
+UChar32 Normalizer::first() {
+    reset();
+    return next();
+}
+
+/**
+ * Return the last character in the normalized text.  This resets
+ * the <tt>Normalizer's</tt> position to be just before the
+ * the input text corresponding to that normalized character.
+ */
+UChar32 Normalizer::last() {
+    currentIndex=nextIndex=text->setToEnd();
+    clearBuffer();
+    return previous();
+}
+
+/**
+ * Retrieve the current iteration position in the input text that is
+ * being normalized.  This method is useful in applications such as
+ * searching, where you need to be able to determine the position in
+ * the input text that corresponds to a given normalized output character.
+ * <p>
+ * <b>Note:</b> This method sets the position in the <em>input</em>, while
+ * {@link #next} and {@link #previous} iterate through characters in the
+ * <em>output</em>.  This means that there is not necessarily a one-to-one
+ * correspondence between characters returned by <tt>next</tt> and
+ * <tt>previous</tt> and the indices passed to and returned from
+ * <tt>setIndex</tt> and {@link #getIndex}.
+ *
+ */
+int32_t Normalizer::getIndex() const {
+    if(bufferPos<buffer.length()) {
+        return currentIndex;
+    } else {
+        return nextIndex;
+    }
+}
+
+/**
+ * Retrieve the index of the start of the input text.  This is the begin index
+ * of the <tt>CharacterIterator</tt> or the start (i.e. 0) of the <tt>String</tt>
+ * over which this <tt>Normalizer</tt> is iterating
+ */
+int32_t Normalizer::startIndex() const {
+    return text->startIndex();
+}
+
+/**
+ * Retrieve the index of the end of the input text.  This is the end index
+ * of the <tt>CharacterIterator</tt> or the length of the <tt>String</tt>
+ * over which this <tt>Normalizer</tt> is iterating
+ */
+int32_t Normalizer::endIndex() const {
+    return text->endIndex();
+}
+
+//-------------------------------------------------------------------------
+// Property access methods
+//-------------------------------------------------------------------------
+
+void
+Normalizer::setMode(UNormalizationMode newMode) 
+{
+    fUMode = newMode;
+    init();
+}
+
+UNormalizationMode
+Normalizer::getUMode() const
+{
+    return fUMode;
+}
+
+void
+Normalizer::setOption(int32_t option, 
+                      UBool value) 
+{
+    if (value) {
+        fOptions |= option;
+    } else {
+        fOptions &= (~option);
+    }
+    init();
+}
+
+UBool
+Normalizer::getOption(int32_t option) const
+{
+    return (fOptions & option) != 0;
+}
+
+/**
+ * Set the input text over which this <tt>Normalizer</tt> will iterate.
+ * The iteration position is set to the beginning of the input text.
+ */
+void
+Normalizer::setText(const UnicodeString& newText, 
+                    UErrorCode &status)
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+    CharacterIterator *newIter = new StringCharacterIterator(newText);
+    if (newIter == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    delete text;
+    text = newIter;
+    reset();
+}
+
+/**
+ * Set the input text over which this <tt>Normalizer</tt> will iterate.
+ * The iteration position is set to the beginning of the string.
+ */
+void
+Normalizer::setText(const CharacterIterator& newText, 
+                    UErrorCode &status) 
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+    CharacterIterator *newIter = newText.clone();
+    if (newIter == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    delete text;
+    text = newIter;
+    reset();
+}
+
+void
+Normalizer::setText(ConstChar16Ptr newText,
+                    int32_t length,
+                    UErrorCode &status)
+{
+    if (U_FAILURE(status)) {
+        return;
+    }
+    CharacterIterator *newIter = new UCharCharacterIterator(newText, length);
+    if (newIter == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    delete text;
+    text = newIter;
+    reset();
+}
+
+/**
+ * Copies the text under iteration into the UnicodeString referred to by "result".
+ * @param result Receives a copy of the text under iteration.
+ */
+void
+Normalizer::getText(UnicodeString&  result) 
+{
+    text->getText(result);
+}
+
+//-------------------------------------------------------------------------
+// Private utility methods
+//-------------------------------------------------------------------------
+
+void Normalizer::clearBuffer() {
+    buffer.remove();
+    bufferPos=0;
+}
+
+UBool
+Normalizer::nextNormalize() {
+    clearBuffer();
+    currentIndex=nextIndex;
+    text->setIndex(nextIndex);
+    if(!text->hasNext()) {
+        return FALSE;
+    }
+    // Skip at least one character so we make progress.
+    UnicodeString segment(text->next32PostInc());
+    while(text->hasNext()) {
+        UChar32 c;
+        if(fNorm2->hasBoundaryBefore(c=text->next32PostInc())) {
+            text->move32(-1, CharacterIterator::kCurrent);
+            break;
+        }
+        segment.append(c);
+    }
+    nextIndex=text->getIndex();
+    UErrorCode errorCode=U_ZERO_ERROR;
+    fNorm2->normalize(segment, buffer, errorCode);
+    return U_SUCCESS(errorCode) && !buffer.isEmpty();
+}
+
+UBool
+Normalizer::previousNormalize() {
+    clearBuffer();
+    nextIndex=currentIndex;
+    text->setIndex(currentIndex);
+    if(!text->hasPrevious()) {
+        return FALSE;
+    }
+    UnicodeString segment;
+    while(text->hasPrevious()) {
+        UChar32 c=text->previous32();
+        segment.insert(0, c);
+        if(fNorm2->hasBoundaryBefore(c)) {
+            break;
+        }
+    }
+    currentIndex=text->getIndex();
+    UErrorCode errorCode=U_ZERO_ERROR;
+    fNorm2->normalize(segment, buffer, errorCode);
+    bufferPos=buffer.length();
+    return U_SUCCESS(errorCode) && !buffer.isEmpty();
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
diff --git a/thirdparty/icu4c/common/parsepos.cpp b/thirdparty/icu4c/common/parsepos.cpp
new file mode 100644
index 0000000000..56c6c78813
--- /dev/null
+++ b/thirdparty/icu4c/common/parsepos.cpp
@@ -0,0 +1,23 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2003-2003, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "unicode/parsepos.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ParsePosition)
+
+ParsePosition::~ParsePosition() {}
+
+ParsePosition *
+ParsePosition::clone() const {
+    return new ParsePosition(*this);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/patternprops.cpp b/thirdparty/icu4c/common/patternprops.cpp
new file mode 100644
index 0000000000..c38a7e276d
--- /dev/null
+++ b/thirdparty/icu4c/common/patternprops.cpp
@@ -0,0 +1,230 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  patternprops.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011mar13
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "patternprops.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * One byte per Latin-1 character.
+ * Bit 0 is set if either Pattern property is true,
+ * bit 1 if Pattern_Syntax is true,
+ * bit 2 if Pattern_White_Space is true.
+ * That is, Pattern_Syntax is encoded as 3 and Pattern_White_Space as 5.
+ */
+static const uint8_t latin1[256]={
+    // WS: 9..D
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 5, 5, 5, 5, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // WS: 20  Syntax: 21..2F
+    5, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3,
+    // Syntax: 3A..40
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 3, 3,
+    3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // Syntax: 5B..5E
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 0,
+    // Syntax: 60
+    3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // Syntax: 7B..7E
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3, 3, 3, 0,
+    // WS: 85
+    0, 0, 0, 0, 0, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // Syntax: A1..A7, A9, AB, AC, AE
+    0, 3, 3, 3, 3, 3, 3, 3, 0, 3, 0, 3, 3, 0, 3, 0,
+    // Syntax: B0, B1, B6, BB, BF
+    3, 3, 0, 0, 0, 0, 3, 0, 0, 0, 0, 3, 0, 0, 0, 3,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // Syntax: D7
+    0, 0, 0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    // Syntax: F7
+    0, 0, 0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 0
+};
+
+/*
+ * One byte per 32 characters from U+2000..U+303F indexing into
+ * a small table of 32-bit data words.
+ * The first two data words are all-zeros and all-ones.
+ */
+static const uint8_t index2000[130]={
+    2, 3, 4, 0, 0, 0, 0, 0,  // 20xx
+    0, 0, 0, 0, 5, 1, 1, 1,  // 21xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 22xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 23xx
+    1, 1, 1, 0, 0, 0, 0, 0,  // 24xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 25xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 26xx
+    1, 1, 1, 6, 7, 1, 1, 1,  // 27xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 28xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 29xx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 2Axx
+    1, 1, 1, 1, 1, 1, 1, 1,  // 2Bxx
+    0, 0, 0, 0, 0, 0, 0, 0,  // 2Cxx
+    0, 0, 0, 0, 0, 0, 0, 0,  // 2Dxx
+    1, 1, 1, 1, 0, 0, 0, 0,  // 2Exx
+    0, 0, 0, 0, 0, 0, 0, 0,  // 2Fxx
+    8, 9  // 3000..303F
+};
+
+/*
+ * One 32-bit integer per 32 characters. Ranges of all-false and all-true
+ * are mapped to the first two values, other ranges map to appropriate bit patterns.
+ */
+static const uint32_t syntax2000[]={
+    0,
+    0xffffffff,
+    0xffff0000,  // 2: 2010..201F
+    0x7fff00ff,  // 3: 2020..2027, 2030..203E
+    0x7feffffe,  // 4: 2041..2053, 2055..205E
+    0xffff0000,  // 5: 2190..219F
+    0x003fffff,  // 6: 2760..2775
+    0xfff00000,  // 7: 2794..279F
+    0xffffff0e,  // 8: 3001..3003, 3008..301F
+    0x00010001   // 9: 3020, 3030
+};
+
+/*
+ * Same as syntax2000, but with additional bits set for the
+ * Pattern_White_Space characters 200E 200F 2028 2029.
+ */
+static const uint32_t syntaxOrWhiteSpace2000[]={
+    0,
+    0xffffffff,
+    0xffffc000,  // 2: 200E..201F
+    0x7fff03ff,  // 3: 2020..2029, 2030..203E
+    0x7feffffe,  // 4: 2041..2053, 2055..205E
+    0xffff0000,  // 5: 2190..219F
+    0x003fffff,  // 6: 2760..2775
+    0xfff00000,  // 7: 2794..279F
+    0xffffff0e,  // 8: 3001..3003, 3008..301F
+    0x00010001   // 9: 3020, 3030
+};
+
+UBool
+PatternProps::isSyntax(UChar32 c) {
+    if(c<0) {
+        return FALSE;
+    } else if(c<=0xff) {
+        return (UBool)(latin1[c]>>1)&1;
+    } else if(c<0x2010) {
+        return FALSE;
+    } else if(c<=0x3030) {
+        uint32_t bits=syntax2000[index2000[(c-0x2000)>>5]];
+        return (UBool)((bits>>(c&0x1f))&1);
+    } else if(0xfd3e<=c && c<=0xfe46) {
+        return c<=0xfd3f || 0xfe45<=c;
+    } else {
+        return FALSE;
+    }
+}
+
+UBool
+PatternProps::isSyntaxOrWhiteSpace(UChar32 c) {
+    if(c<0) {
+        return FALSE;
+    } else if(c<=0xff) {
+        return (UBool)(latin1[c]&1);
+    } else if(c<0x200e) {
+        return FALSE;
+    } else if(c<=0x3030) {
+        uint32_t bits=syntaxOrWhiteSpace2000[index2000[(c-0x2000)>>5]];
+        return (UBool)((bits>>(c&0x1f))&1);
+    } else if(0xfd3e<=c && c<=0xfe46) {
+        return c<=0xfd3f || 0xfe45<=c;
+    } else {
+        return FALSE;
+    }
+}
+
+UBool
+PatternProps::isWhiteSpace(UChar32 c) {
+    if(c<0) {
+        return FALSE;
+    } else if(c<=0xff) {
+        return (UBool)(latin1[c]>>2)&1;
+    } else if(0x200e<=c && c<=0x2029) {
+        return c<=0x200f || 0x2028<=c;
+    } else {
+        return FALSE;
+    }
+}
+
+const UChar *
+PatternProps::skipWhiteSpace(const UChar *s, int32_t length) {
+    while(length>0 && isWhiteSpace(*s)) {
+        ++s;
+        --length;
+    }
+    return s;
+}
+
+int32_t
+PatternProps::skipWhiteSpace(const UnicodeString& s, int32_t start) {
+    int32_t i = start;
+    int32_t length = s.length();
+    while(i<length && isWhiteSpace(s.charAt(i))) {
+        ++i;
+    }
+    return i;
+}
+
+const UChar *
+PatternProps::trimWhiteSpace(const UChar *s, int32_t &length) {
+    if(length<=0 || (!isWhiteSpace(s[0]) && !isWhiteSpace(s[length-1]))) {
+        return s;
+    }
+    int32_t start=0;
+    int32_t limit=length;
+    while(start<limit && isWhiteSpace(s[start])) {
+        ++start;
+    }
+    if(start<limit) {
+        // There is non-white space at start; we will not move limit below that,
+        // so we need not test start<limit in the loop.
+        while(isWhiteSpace(s[limit-1])) {
+            --limit;
+        }
+    }
+    length=limit-start;
+    return s+start;
+}
+
+UBool
+PatternProps::isIdentifier(const UChar *s, int32_t length) {
+    if(length<=0) {
+        return FALSE;
+    }
+    const UChar *limit=s+length;
+    do {
+        if(isSyntaxOrWhiteSpace(*s++)) {
+            return FALSE;
+        }
+    } while(s<limit);
+    return TRUE;
+}
+
+const UChar *
+PatternProps::skipIdentifier(const UChar *s, int32_t length) {
+    while(length>0 && !isSyntaxOrWhiteSpace(*s)) {
+        ++s;
+        --length;
+    }
+    return s;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/patternprops.h b/thirdparty/icu4c/common/patternprops.h
new file mode 100644
index 0000000000..95898d580c
--- /dev/null
+++ b/thirdparty/icu4c/common/patternprops.h
@@ -0,0 +1,98 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  patternprops.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011mar13
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __PATTERNPROPS_H__
+#define __PATTERNPROPS_H__
+
+#include "unicode/unistr.h"
+#include "unicode/utypes.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Implements the immutable Unicode properties Pattern_Syntax and Pattern_White_Space.
+ * Hardcodes these properties, does not load data, does not depend on other ICU classes.
+ * <p>
+ * Note: Both properties include ASCII as well as non-ASCII, non-Latin-1 code points,
+ * and both properties only include BMP code points (no supplementary ones).
+ * Pattern_Syntax includes some unassigned code points.
+ * <p>
+ * [:Pattern_White_Space:] =
+ *   [\u0009-\u000D\ \u0085\u200E\u200F\u2028\u2029]
+ * <p>
+ * [:Pattern_Syntax:] =
+ *   [!-/\:-@\[-\^`\{-~\u00A1-\u00A7\u00A9\u00AB\u00AC\u00AE
+ *    \u00B0\u00B1\u00B6\u00BB\u00BF\u00D7\u00F7
+ *    \u2010-\u2027\u2030-\u203E\u2041-\u2053\u2055-\u205E
+ *    \u2190-\u245F\u2500-\u2775\u2794-\u2BFF\u2E00-\u2E7F
+ *    \u3001-\u3003\u3008-\u3020\u3030\uFD3E\uFD3F\uFE45\uFE46]
+ * @author mscherer
+ */
+class U_COMMON_API PatternProps {
+public:
+    /**
+     * @return true if c is a Pattern_Syntax code point.
+     */
+    static UBool isSyntax(UChar32 c);
+
+    /**
+     * @return true if c is a Pattern_Syntax or Pattern_White_Space code point.
+     */
+    static UBool isSyntaxOrWhiteSpace(UChar32 c);
+
+    /**
+     * @return true if c is a Pattern_White_Space character.
+     */
+    static UBool isWhiteSpace(UChar32 c);
+
+    /**
+     * Skips over Pattern_White_Space starting at s.
+     * @return The smallest pointer at or after s with a non-white space character.
+     */
+    static const UChar *skipWhiteSpace(const UChar *s, int32_t length);
+
+    /**
+     * Skips over Pattern_White_Space starting at index start in s.
+     * @return The smallest index at or after start with a non-white space character.
+     */
+    static int32_t skipWhiteSpace(const UnicodeString &s, int32_t start);
+
+    /**
+     * @return s except with leading and trailing Pattern_White_Space removed and length adjusted.
+     */
+    static const UChar *trimWhiteSpace(const UChar *s, int32_t &length);
+
+    /**
+     * Tests whether the string contains a "pattern identifier", that is,
+     * whether it contains only non-Pattern_White_Space, non-Pattern_Syntax characters.
+     * @return true if there are no Pattern_White_Space or Pattern_Syntax characters in s.
+     */
+    static UBool isIdentifier(const UChar *s, int32_t length);
+
+    /**
+     * Skips over a "pattern identifier" starting at index s.
+     * @return The smallest pointer at or after s with
+     *         a Pattern_White_Space or Pattern_Syntax character.
+     */
+    static const UChar *skipIdentifier(const UChar *s, int32_t length);
+
+private:
+    PatternProps();  // no constructor: all static methods
+};
+
+U_NAMESPACE_END
+
+#endif  // __PATTERNPROPS_H__
diff --git a/thirdparty/icu4c/common/pluralmap.cpp b/thirdparty/icu4c/common/pluralmap.cpp
new file mode 100644
index 0000000000..ec87f0198e
--- /dev/null
+++ b/thirdparty/icu4c/common/pluralmap.cpp
@@ -0,0 +1,44 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ * Copyright (C) 2015, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ */
+
+#include "unicode/unistr.h"
+#include "charstr.h"
+#include "cstring.h"
+#include "pluralmap.h"
+
+U_NAMESPACE_BEGIN
+
+static const char * const gPluralForms[] = {
+        "other", "zero", "one", "two", "few", "many"};
+
+PluralMapBase::Category
+PluralMapBase::toCategory(const char *pluralForm) {
+    for (int32_t i = 0; i < UPRV_LENGTHOF(gPluralForms); ++i) {
+        if (uprv_strcmp(pluralForm, gPluralForms[i]) == 0) {
+            return static_cast<Category>(i);
+        }
+    }
+    return NONE;
+}
+
+PluralMapBase::Category
+PluralMapBase::toCategory(const UnicodeString &pluralForm) {
+    CharString cCategory;
+    UErrorCode status = U_ZERO_ERROR;
+    cCategory.appendInvariantChars(pluralForm, status);    
+    return U_FAILURE(status) ? NONE : toCategory(cCategory.data());
+}
+
+const char *PluralMapBase::getCategoryName(Category c) {
+    int32_t index = c;
+    return (index < 0 || index >= UPRV_LENGTHOF(gPluralForms)) ?
+            NULL : gPluralForms[index];
+}
+
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/pluralmap.h b/thirdparty/icu4c/common/pluralmap.h
new file mode 100644
index 0000000000..d898ac4671
--- /dev/null
+++ b/thirdparty/icu4c/common/pluralmap.h
@@ -0,0 +1,292 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File pluralmap.h - PluralMap class that maps plural categories to values.
+******************************************************************************
+*/
+
+#ifndef __PLURAL_MAP_H__
+#define __PLURAL_MAP_H__
+
+#include "unicode/uobject.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+
+class U_COMMON_API PluralMapBase : public UMemory {
+public:
+    /**
+     * The names of all the plural categories. NONE is not an actual plural
+     * category, but rather represents the absense of a plural category.
+     */
+    enum Category {
+        NONE = -1,
+        OTHER,
+        ZERO,
+        ONE,
+        TWO,
+        FEW,
+        MANY,
+        CATEGORY_COUNT
+    };
+
+    /**
+     * Converts a category name such as "zero", "one", "two", "few", "many"
+     * or "other" to a category enum. Returns NONE for an unrecognized
+     * category name.
+     */
+    static Category toCategory(const char *categoryName);
+
+    /**
+     * Converts a category name such as "zero", "one", "two", "few", "many"
+     * or "other" to a category enum.  Returns NONE for urecongized
+     * category name.
+     */
+    static Category toCategory(const UnicodeString &categoryName);
+
+    /**
+     * Converts a category to a name.
+     * Passing NONE or CATEGORY_COUNT for category returns NULL.
+     */
+    static const char *getCategoryName(Category category);
+};
+
+/**
+ * A Map of plural categories to values. It maintains ownership of the
+ * values.
+ *
+ * Type T is the value type. T must provide the followng:
+ * 1) Default constructor
+ * 2) Copy constructor
+ * 3) Assignment operator
+ * 4) Must extend UMemory
+ */
+template<typename T>
+class PluralMap : public PluralMapBase {
+public:
+    /**
+     * Other category is maps to a copy of the default value.
+     */
+    PluralMap() : fOtherVariant() {
+        initializeNew();
+    }
+
+    /**
+     * Other category is mapped to otherVariant.
+     */
+    PluralMap(const T &otherVariant) : fOtherVariant(otherVariant) {
+        initializeNew();
+    }
+
+    PluralMap(const PluralMap<T> &other) : fOtherVariant(other.fOtherVariant) {
+        fVariants[0] = &fOtherVariant;
+        for (int32_t i = 1; i < UPRV_LENGTHOF(fVariants); ++i) {
+            fVariants[i] = other.fVariants[i] ?
+                    new T(*other.fVariants[i]) : NULL;
+        }
+    }
+
+    PluralMap<T> &operator=(const PluralMap<T> &other) {
+        if (this == &other) {
+            return *this;
+        }
+        for (int32_t i = 0; i < UPRV_LENGTHOF(fVariants); ++i) {
+            if (fVariants[i] != NULL && other.fVariants[i] != NULL) {
+                *fVariants[i] = *other.fVariants[i];
+            } else if (fVariants[i] != NULL) {
+                delete fVariants[i];
+                fVariants[i] = NULL;
+            } else if (other.fVariants[i] != NULL) {
+                fVariants[i] = new T(*other.fVariants[i]);
+            } else {
+                // do nothing
+            }
+        }
+        return *this;
+    }
+
+    ~PluralMap() {
+        for (int32_t i = 1; i < UPRV_LENGTHOF(fVariants); ++i) {
+            delete fVariants[i];
+        }
+    }
+
+    /**
+     * Removes all mappings and makes 'other' point to the default value.
+     */
+    void clear() {
+        *fVariants[0] = T();
+        for (int32_t i = 1; i < UPRV_LENGTHOF(fVariants); ++i) {
+            delete fVariants[i];
+            fVariants[i] = NULL;
+        }
+    }
+
+    /**
+     * Iterates through the mappings in this instance, set index to NONE
+     * prior to using. Call next repeatedly to get the values until it
+     * returns NULL. Each time next returns, caller may pass index
+     * to getCategoryName() to get the name of the plural category.
+     * When this function returns NULL, index is CATEGORY_COUNT
+     */
+    const T *next(Category &index) const {
+        int32_t idx = index;
+        ++idx;
+        for (; idx < UPRV_LENGTHOF(fVariants); ++idx) {
+            if (fVariants[idx] != NULL) {
+                index = static_cast<Category>(idx);
+                return fVariants[idx];
+            }
+        }
+        index = static_cast<Category>(idx);
+        return NULL;
+    }
+
+    /**
+     * non const version of next.
+     */
+    T *nextMutable(Category &index) {
+        const T *result = next(index);
+        return const_cast<T *>(result);
+    }
+
+    /**
+     * Returns the 'other' variant.
+     * Same as calling get(OTHER).
+     */
+    const T &getOther() const {
+        return get(OTHER);
+    }
+
+    /**
+     * Returns the value associated with a category.
+     * If no value found, or v is NONE or CATEGORY_COUNT, falls
+     * back to returning the value for the 'other' category.
+     */
+    const T &get(Category v) const {
+        int32_t index = v;
+        if (index < 0 || index >= UPRV_LENGTHOF(fVariants) || fVariants[index] == NULL) {
+            return *fVariants[0];
+        }
+        return *fVariants[index];
+    }
+
+    /**
+     * Convenience routine to get the value by category name. Otherwise
+     * works just like get(Category).
+     */
+    const T &get(const char *category) const {
+        return get(toCategory(category));
+    }
+
+    /**
+     * Convenience routine to get the value by category name as a
+     * UnicodeString. Otherwise works just like get(category).
+     */
+    const T &get(const UnicodeString &category) const {
+        return get(toCategory(category));
+    }
+
+    /**
+     * Returns a pointer to the value associated with a category
+     * that caller can safely modify. If the value was defaulting to the 'other'
+     * variant because no explicit value was stored, this method creates a
+     * new value using the default constructor at the returned pointer.
+     *
+     * @param category the category with the value to change.
+     * @param status error returned here if index is NONE or CATEGORY_COUNT
+     *  or memory could not be allocated, or any other error happens.
+     */
+    T *getMutable(
+            Category category,
+            UErrorCode &status) {
+        return getMutable(category, NULL, status);
+    }
+
+    /**
+     * Convenience routine to get a mutable pointer to a value by category name.
+     * Otherwise works just like getMutable(Category, UErrorCode &).
+     * reports an error if the category name is invalid.
+     */
+    T *getMutable(
+            const char *category,
+            UErrorCode &status) {
+        return getMutable(toCategory(category), NULL, status);
+    }
+
+    /**
+     * Just like getMutable(Category, UErrorCode &) but copies defaultValue to
+     * returned pointer if it was defaulting to the 'other' variant
+     * because no explicit value was stored.
+     */
+    T *getMutableWithDefault(
+            Category category,
+            const T &defaultValue,
+            UErrorCode &status) {
+        return getMutable(category, &defaultValue, status);
+    }
+
+    /**
+     * Returns true if this object equals rhs.
+     */
+    UBool equals(
+            const PluralMap<T> &rhs,
+            UBool (*eqFunc)(const T &, const T &)) const {
+        for (int32_t i = 0; i < UPRV_LENGTHOF(fVariants); ++i) {
+            if (fVariants[i] == rhs.fVariants[i]) {
+                continue;
+            }
+            if (fVariants[i] == NULL || rhs.fVariants[i] == NULL) {
+                return false;
+            }
+            if (!eqFunc(*fVariants[i], *rhs.fVariants[i])) {
+                return false;
+            }
+        }
+        return true;
+    }
+
+private:
+    T fOtherVariant;
+    T* fVariants[6];
+
+    T *getMutable(
+            Category category,
+            const T *defaultValue,
+            UErrorCode &status) {
+        if (U_FAILURE(status)) {
+            return NULL;
+        }
+        int32_t index = category;
+        if (index < 0 || index >= UPRV_LENGTHOF(fVariants)) {
+            status = U_ILLEGAL_ARGUMENT_ERROR;
+            return NULL;
+        }
+        if (fVariants[index] == NULL) {
+            fVariants[index] = defaultValue == NULL ?
+                    new T() : new T(*defaultValue);
+        }
+        if (!fVariants[index]) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+        }
+        return fVariants[index];
+    }
+
+    void initializeNew() {
+        fVariants[0] = &fOtherVariant;
+        for (int32_t i = 1; i < UPRV_LENGTHOF(fVariants); ++i) {
+            fVariants[i] = NULL;
+        }
+    }
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/propname.cpp b/thirdparty/icu4c/common/propname.cpp
new file mode 100644
index 0000000000..a12eb7d913
--- /dev/null
+++ b/thirdparty/icu4c/common/propname.cpp
@@ -0,0 +1,328 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: October 30 2002
+* Since: ICU 2.4
+* 2010nov19 Markus Scherer  Rewrite for formatVersion 2.
+**********************************************************************
+*/
+#include "propname.h"
+#include "unicode/uchar.h"
+#include "unicode/udata.h"
+#include "unicode/uscript.h"
+#include "umutex.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uarrsort.h"
+#include "uinvchar.h"
+
+#define INCLUDED_FROM_PROPNAME_CPP
+#include "propname_data.h"
+
+U_CDECL_BEGIN
+
+/**
+ * Get the next non-ignorable ASCII character from a property name
+ * and lowercases it.
+ * @return ((advance count for the name)<<8)|character
+ */
+static inline int32_t
+getASCIIPropertyNameChar(const char *name) {
+    int32_t i;
+    char c;
+
+    /* Ignore delimiters '-', '_', and ASCII White_Space */
+    for(i=0;
+        (c=name[i++])==0x2d || c==0x5f ||
+        c==0x20 || (0x09<=c && c<=0x0d);
+    ) {}
+
+    if(c!=0) {
+        return (i<<8)|(uint8_t)uprv_asciitolower((char)c);
+    } else {
+        return i<<8;
+    }
+}
+
+/**
+ * Get the next non-ignorable EBCDIC character from a property name
+ * and lowercases it.
+ * @return ((advance count for the name)<<8)|character
+ */
+static inline int32_t
+getEBCDICPropertyNameChar(const char *name) {
+    int32_t i;
+    char c;
+
+    /* Ignore delimiters '-', '_', and EBCDIC White_Space */
+    for(i=0;
+        (c=name[i++])==0x60 || c==0x6d ||
+        c==0x40 || c==0x05 || c==0x15 || c==0x25 || c==0x0b || c==0x0c || c==0x0d;
+    ) {}
+
+    if(c!=0) {
+        return (i<<8)|(uint8_t)uprv_ebcdictolower((char)c);
+    } else {
+        return i<<8;
+    }
+}
+
+/**
+ * Unicode property names and property value names are compared "loosely".
+ *
+ * UCD.html 4.0.1 says:
+ *   For all property names, property value names, and for property values for
+ *   Enumerated, Binary, or Catalog properties, use the following
+ *   loose matching rule:
+ *
+ *   LM3. Ignore case, whitespace, underscore ('_'), and hyphens.
+ *
+ * This function does just that, for (char *) name strings.
+ * It is almost identical to ucnv_compareNames() but also ignores
+ * C0 White_Space characters (U+0009..U+000d, and U+0085 on EBCDIC).
+ *
+ * @internal
+ */
+
+U_CAPI int32_t U_EXPORT2
+uprv_compareASCIIPropertyNames(const char *name1, const char *name2) {
+    int32_t rc, r1, r2;
+
+    for(;;) {
+        r1=getASCIIPropertyNameChar(name1);
+        r2=getASCIIPropertyNameChar(name2);
+
+        /* If we reach the ends of both strings then they match */
+        if(((r1|r2)&0xff)==0) {
+            return 0;
+        }
+
+        /* Compare the lowercased characters */
+        if(r1!=r2) {
+            rc=(r1&0xff)-(r2&0xff);
+            if(rc!=0) {
+                return rc;
+            }
+        }
+
+        name1+=r1>>8;
+        name2+=r2>>8;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_compareEBCDICPropertyNames(const char *name1, const char *name2) {
+    int32_t rc, r1, r2;
+
+    for(;;) {
+        r1=getEBCDICPropertyNameChar(name1);
+        r2=getEBCDICPropertyNameChar(name2);
+
+        /* If we reach the ends of both strings then they match */
+        if(((r1|r2)&0xff)==0) {
+            return 0;
+        }
+
+        /* Compare the lowercased characters */
+        if(r1!=r2) {
+            rc=(r1&0xff)-(r2&0xff);
+            if(rc!=0) {
+                return rc;
+            }
+        }
+
+        name1+=r1>>8;
+        name2+=r2>>8;
+    }
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+int32_t PropNameData::findProperty(int32_t property) {
+    int32_t i=1;  // valueMaps index, initially after numRanges
+    for(int32_t numRanges=valueMaps[0]; numRanges>0; --numRanges) {
+        // Read and skip the start and limit of this range.
+        int32_t start=valueMaps[i];
+        int32_t limit=valueMaps[i+1];
+        i+=2;
+        if(property<start) {
+            break;
+        }
+        if(property<limit) {
+            return i+(property-start)*2;
+        }
+        i+=(limit-start)*2;  // Skip all entries for this range.
+    }
+    return 0;
+}
+
+int32_t PropNameData::findPropertyValueNameGroup(int32_t valueMapIndex, int32_t value) {
+    if(valueMapIndex==0) {
+        return 0;  // The property does not have named values.
+    }
+    ++valueMapIndex;  // Skip the BytesTrie offset.
+    int32_t numRanges=valueMaps[valueMapIndex++];
+    if(numRanges<0x10) {
+        // Ranges of values.
+        for(; numRanges>0; --numRanges) {
+            // Read and skip the start and limit of this range.
+            int32_t start=valueMaps[valueMapIndex];
+            int32_t limit=valueMaps[valueMapIndex+1];
+            valueMapIndex+=2;
+            if(value<start) {
+                break;
+            }
+            if(value<limit) {
+                return valueMaps[valueMapIndex+value-start];
+            }
+            valueMapIndex+=limit-start;  // Skip all entries for this range.
+        }
+    } else {
+        // List of values.
+        int32_t valuesStart=valueMapIndex;
+        int32_t nameGroupOffsetsStart=valueMapIndex+numRanges-0x10;
+        do {
+            int32_t v=valueMaps[valueMapIndex];
+            if(value<v) {
+                break;
+            }
+            if(value==v) {
+                return valueMaps[nameGroupOffsetsStart+valueMapIndex-valuesStart];
+            }
+        } while(++valueMapIndex<nameGroupOffsetsStart);
+    }
+    return 0;
+}
+
+const char *PropNameData::getName(const char *nameGroup, int32_t nameIndex) {
+    int32_t numNames=*nameGroup++;
+    if(nameIndex<0 || numNames<=nameIndex) {
+        return NULL;
+    }
+    // Skip nameIndex names.
+    for(; nameIndex>0; --nameIndex) {
+        nameGroup=uprv_strchr(nameGroup, 0)+1;
+    }
+    if(*nameGroup==0) {
+        return NULL;  // no name (Property[Value]Aliases.txt has "n/a")
+    }
+    return nameGroup;
+}
+
+UBool PropNameData::containsName(BytesTrie &trie, const char *name) {
+    if(name==NULL) {
+        return FALSE;
+    }
+    UStringTrieResult result=USTRINGTRIE_NO_VALUE;
+    char c;
+    while((c=*name++)!=0) {
+        c=uprv_invCharToLowercaseAscii(c);
+        // Ignore delimiters '-', '_', and ASCII White_Space.
+        if(c==0x2d || c==0x5f || c==0x20 || (0x09<=c && c<=0x0d)) {
+            continue;
+        }
+        if(!USTRINGTRIE_HAS_NEXT(result)) {
+            return FALSE;
+        }
+        result=trie.next((uint8_t)c);
+    }
+    return USTRINGTRIE_HAS_VALUE(result);
+}
+
+const char *PropNameData::getPropertyName(int32_t property, int32_t nameChoice) {
+    int32_t valueMapIndex=findProperty(property);
+    if(valueMapIndex==0) {
+        return NULL;  // Not a known property.
+    }
+    return getName(nameGroups+valueMaps[valueMapIndex], nameChoice);
+}
+
+const char *PropNameData::getPropertyValueName(int32_t property, int32_t value, int32_t nameChoice) {
+    int32_t valueMapIndex=findProperty(property);
+    if(valueMapIndex==0) {
+        return NULL;  // Not a known property.
+    }
+    int32_t nameGroupOffset=findPropertyValueNameGroup(valueMaps[valueMapIndex+1], value);
+    if(nameGroupOffset==0) {
+        return NULL;
+    }
+    return getName(nameGroups+nameGroupOffset, nameChoice);
+}
+
+int32_t PropNameData::getPropertyOrValueEnum(int32_t bytesTrieOffset, const char *alias) {
+    BytesTrie trie(bytesTries+bytesTrieOffset);
+    if(containsName(trie, alias)) {
+        return trie.getValue();
+    } else {
+        return UCHAR_INVALID_CODE;
+    }
+}
+
+int32_t PropNameData::getPropertyEnum(const char *alias) {
+    return getPropertyOrValueEnum(0, alias);
+}
+
+int32_t PropNameData::getPropertyValueEnum(int32_t property, const char *alias) {
+    int32_t valueMapIndex=findProperty(property);
+    if(valueMapIndex==0) {
+        return UCHAR_INVALID_CODE;  // Not a known property.
+    }
+    valueMapIndex=valueMaps[valueMapIndex+1];
+    if(valueMapIndex==0) {
+        return UCHAR_INVALID_CODE;  // The property does not have named values.
+    }
+    // valueMapIndex is the start of the property's valueMap,
+    // where the first word is the BytesTrie offset.
+    return getPropertyOrValueEnum(valueMaps[valueMapIndex], alias);
+}
+U_NAMESPACE_END
+
+//----------------------------------------------------------------------
+// Public API implementation
+
+U_CAPI const char* U_EXPORT2
+u_getPropertyName(UProperty property,
+                  UPropertyNameChoice nameChoice) {
+    U_NAMESPACE_USE
+    return PropNameData::getPropertyName(property, nameChoice);
+}
+
+U_CAPI UProperty U_EXPORT2
+u_getPropertyEnum(const char* alias) {
+    U_NAMESPACE_USE
+    return (UProperty)PropNameData::getPropertyEnum(alias);
+}
+
+U_CAPI const char* U_EXPORT2
+u_getPropertyValueName(UProperty property,
+                       int32_t value,
+                       UPropertyNameChoice nameChoice) {
+    U_NAMESPACE_USE
+    return PropNameData::getPropertyValueName(property, value, nameChoice);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_getPropertyValueEnum(UProperty property,
+                       const char* alias) {
+    U_NAMESPACE_USE
+    return PropNameData::getPropertyValueEnum(property, alias);
+}
+
+U_CAPI const char*  U_EXPORT2
+uscript_getName(UScriptCode scriptCode){
+    return u_getPropertyValueName(UCHAR_SCRIPT, scriptCode,
+                                  U_LONG_PROPERTY_NAME);
+}
+
+U_CAPI const char*  U_EXPORT2
+uscript_getShortName(UScriptCode scriptCode){
+    return u_getPropertyValueName(UCHAR_SCRIPT, scriptCode,
+                                  U_SHORT_PROPERTY_NAME);
+}
diff --git a/thirdparty/icu4c/common/propname.h b/thirdparty/icu4c/common/propname.h
new file mode 100644
index 0000000000..1a8ced5b87
--- /dev/null
+++ b/thirdparty/icu4c/common/propname.h
@@ -0,0 +1,212 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2011, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: October 30 2002
+* Since: ICU 2.4
+* 2010nov19 Markus Scherer  Rewrite for formatVersion 2.
+**********************************************************************
+*/
+#ifndef PROPNAME_H
+#define PROPNAME_H
+
+#include "unicode/utypes.h"
+#include "unicode/bytestrie.h"
+#include "unicode/uchar.h"
+#include "udataswp.h"
+#include "uprops.h"
+
+/*
+ * This header defines the in-memory layout of the property names data
+ * structure representing the UCD data files PropertyAliases.txt and
+ * PropertyValueAliases.txt.  It is used by:
+ *   propname.cpp - reads data
+ *   genpname     - creates data
+ */
+
+/* low-level char * property name comparison -------------------------------- */
+
+U_CDECL_BEGIN
+
+/**
+ * \var uprv_comparePropertyNames
+ * Unicode property names and property value names are compared "loosely".
+ *
+ * UCD.html 4.0.1 says:
+ *   For all property names, property value names, and for property values for
+ *   Enumerated, Binary, or Catalog properties, use the following
+ *   loose matching rule:
+ *
+ *   LM3. Ignore case, whitespace, underscore ('_'), and hyphens.
+ *
+ * This function does just that, for (char *) name strings.
+ * It is almost identical to ucnv_compareNames() but also ignores
+ * C0 White_Space characters (U+0009..U+000d, and U+0085 on EBCDIC).
+ *
+ * @internal
+ */
+
+U_CAPI int32_t U_EXPORT2
+uprv_compareASCIIPropertyNames(const char *name1, const char *name2);
+
+U_CAPI int32_t U_EXPORT2
+uprv_compareEBCDICPropertyNames(const char *name1, const char *name2);
+
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_comparePropertyNames uprv_compareASCIIPropertyNames
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_comparePropertyNames uprv_compareEBCDICPropertyNames
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+
+U_CDECL_END
+
+/* UDataMemory structure and signatures ------------------------------------- */
+
+#define PNAME_DATA_NAME "pnames"
+#define PNAME_DATA_TYPE "icu"
+
+/* Fields in UDataInfo: */
+
+/* PNAME_SIG[] is encoded as numeric literals for compatibility with the HP compiler */
+#define PNAME_SIG_0 ((uint8_t)0x70) /* p */
+#define PNAME_SIG_1 ((uint8_t)0x6E) /* n */
+#define PNAME_SIG_2 ((uint8_t)0x61) /* a */
+#define PNAME_SIG_3 ((uint8_t)0x6D) /* m */
+
+U_NAMESPACE_BEGIN
+
+class PropNameData {
+public:
+    enum {
+        // Byte offsets from the start of the data, after the generic header.
+        IX_VALUE_MAPS_OFFSET,
+        IX_BYTE_TRIES_OFFSET,
+        IX_NAME_GROUPS_OFFSET,
+        IX_RESERVED3_OFFSET,
+        IX_RESERVED4_OFFSET,
+        IX_TOTAL_SIZE,
+
+        // Other values.
+        IX_MAX_NAME_LENGTH,
+        IX_RESERVED7,
+        IX_COUNT
+    };
+
+    static const char *getPropertyName(int32_t property, int32_t nameChoice);
+    static const char *getPropertyValueName(int32_t property, int32_t value, int32_t nameChoice);
+
+    static int32_t getPropertyEnum(const char *alias);
+    static int32_t getPropertyValueEnum(int32_t property, const char *alias);
+
+private:
+    static int32_t findProperty(int32_t property);
+    static int32_t findPropertyValueNameGroup(int32_t valueMapIndex, int32_t value);
+    static const char *getName(const char *nameGroup, int32_t nameIndex);
+    static UBool containsName(BytesTrie &trie, const char *name);
+
+    static int32_t getPropertyOrValueEnum(int32_t bytesTrieOffset, const char *alias);
+
+    static const int32_t indexes[];
+    static const int32_t valueMaps[];
+    static const uint8_t bytesTries[];
+    static const char nameGroups[];
+};
+
+/*
+ * pnames.icu formatVersion 2
+ *
+ * formatVersion 2 is new in ICU 4.8.
+ * In ICU 4.8, the pnames.icu data file is used only in ICU4J.
+ * ICU4C 4.8 has the same data structures hardcoded in source/common/propname_data.h.
+ *
+ * For documentation of pnames.icu formatVersion 1 see ICU4C 4.6 (2010-dec-01)
+ * or earlier versions of this header file (source/common/propname.h).
+ *
+ * The pnames.icu begins with the standard ICU DataHeader/UDataInfo.
+ * After that:
+ *
+ * int32_t indexes[8];
+ *
+ *      (See the PropNameData::IX_... constants.)
+ *
+ *      The first 6 indexes are byte offsets from the beginning of the data
+ *      (beginning of indexes[]) to following structures.
+ *      The length of each structure is the difference between its offset
+ *      and the next one.
+ *      All offsets are filled in: Where there is no data between two offsets,
+ *      those two offsets are the same.
+ *      The last offset (indexes[PropNameData::IX_TOTAL_SIZE]) indicates the
+ *      total number of bytes in the file. (Not counting the standard headers.)
+ *
+ *      The sixth index (indexes[PropNameData::IX_MAX_NAME_LENGTH]) has the
+ *      maximum length of any Unicode property (or property value) alias.
+ *      (Without normalization, that is, including underscores etc.)
+ *
+ * int32_t valueMaps[];
+ *
+ *      The valueMaps[] begins with a map from UProperty enums to properties,
+ *      followed by the per-property value maps from property values to names,
+ *      for those properties that have named values.
+ *      (Binary & enumerated, plus General_Category_Mask.)
+ *
+ *      valueMaps[0] contains the number of UProperty enum ranges.
+ *      For each range:
+ *        int32_t start, limit -- first and last+1 UProperty enum of a dense range
+ *        Followed by (limit-start) pairs of
+ *          int32_t nameGroupOffset;
+ *            Offset into nameGroups[] for the property's names/aliases.
+ *          int32_t valueMapIndex;
+ *            Offset of the property's value map in the valueMaps[] array.
+ *            If the valueMapIndex is 0, then the property does not have named values.
+ *
+ *      For each property's value map:
+ *      int32_t bytesTrieOffset; -- Offset into bytesTries[] for name->value mapping.
+ *      int32_t numRanges;
+ *        If numRanges is in the range 1..15, then that many ranges of values follow.
+ *        Per range:
+ *          int32_t start, limit -- first and last+1 UProperty enum of a range
+ *          Followed by (limit-start) entries of
+ *            int32_t nameGroupOffset;
+ *              Offset into nameGroups[] for the property value's names/aliases.
+ *              If the nameGroupOffset is 0, then this is not a named value for this property.
+ *              (That is, the ranges need not be dense.)
+ *        If numRanges is >=0x10, then (numRanges-0x10) sorted values
+ *        and then (numRanges-0x10) corresponding nameGroupOffsets follow.
+ *        Values are sorted as signed integers.
+ *        In this case, the set of values is dense; no nameGroupOffset will be 0.
+ *
+ *      For both properties and property values, ranges are sorted by their start/limit values.
+ *
+ * uint8_t bytesTries[];
+ *
+ *      This is a sequence of BytesTrie structures, byte-serialized tries for
+ *      mapping from names/aliases to values.
+ *      The first one maps from property names/aliases to UProperty enum constants.
+ *      The following ones are indexed by property value map bytesTrieOffsets
+ *      for mapping each property's names/aliases to their property values.
+ *
+ * char nameGroups[];
+ *
+ *      This is a sequence of property name groups.
+ *      Each group is a list of names/aliases (invariant-character strings) for
+ *      one property or property value, in the order of UCharNameChoice.
+ *      The first byte of each group is the number of names in the group.
+ *      It is followed by that many NUL-terminated strings.
+ *      The first string is for the short name; if there is no short name,
+ *      then the first string is empty.
+ *      The second string is the long name. Further strings are additional aliases.
+ *
+ *      The first name group is for a property rather than a property value,
+ *      so that a nameGroupOffset of 0 can be used to indicate "no value"
+ *      in a property's sparse value ranges.
+ */
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/propname_data.h b/thirdparty/icu4c/common/propname_data.h
new file mode 100644
index 0000000000..6f63e9cdd4
--- /dev/null
+++ b/thirdparty/icu4c/common/propname_data.h
@@ -0,0 +1,1919 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+// Copyright (C) 1999-2016, International Business Machines
+// Corporation and others.  All Rights Reserved.
+//
+// file name: propname_data.h
+//
+// machine-generated by: icu/tools/unicode/c/genprops/pnamesbuilder.cpp
+
+
+#ifdef INCLUDED_FROM_PROPNAME_CPP
+
+U_NAMESPACE_BEGIN
+
+const int32_t PropNameData::indexes[8]={0x20,0x15b8,0x5048,0xa69a,0xa69a,0xa69a,0x2f,0};
+
+const int32_t PropNameData::valueMaps[1382]={
+6,0,0x41,0,0xe3,0x368,0xe3,0x37e,0xe3,0x393,0xe3,0x3a9,0xe3,0x3b4,0xe3,0x3d5,
+0xe3,0x3e5,0xe3,0x3f4,0xe3,0x402,0xe3,0x426,0xe3,0x43d,0xe3,0x455,0xe3,0x46c,0xe3,0x47b,
+0xe3,0x48a,0xe3,0x49b,0xe3,0x4a9,0xe3,0x4bb,0xe3,0x4d5,0xe3,0x4f0,0xe3,0x505,0xe3,0x522,
+0xe3,0x533,0xe3,0x53e,0xe3,0x55d,0xe3,0x573,0xe3,0x584,0xe3,0x594,0xe3,0x5af,0xe3,0x5c8,
+0xe3,0x5d9,0xe3,0x5f3,0xe3,0x606,0xe3,0x616,0xe3,0x630,0xe3,0x649,0xe3,0x660,0xe3,0x674,
+0xe3,0x68a,0xe3,0x69e,0xe3,0x6b4,0xe3,0x6ce,0xe3,0x6e6,0xe3,0x702,0xe3,0x70a,0xe3,0x712,
+0xe3,0x71a,0xe3,0x722,0xe3,0x72b,0xe3,0x738,0xe3,0x74b,0xe3,0x768,0xe3,0x785,0xe3,0x7a2,
+0xe3,0x7c0,0xe3,0x7de,0xe3,0x802,0xe3,0x80f,0xe3,0x829,0xe3,0x83e,0xe3,0x859,0xe3,0x870,
+0xe3,0x887,0xe3,0x8a9,0xe3,0x1000,0x1019,0x8c8,0x15f,0xae8,0x17a,0x2f11,0xe9,0x2f30,0x2b3,0x306e,
+0x2c9,0x30c8,0x2d3,0x3325,0x2f5,0x3c20,0x35f,0x3c90,0x369,0x3f2a,0x398,0x3f68,0x3a0,0x4a5b,0x465,0x4ad9,
+0x46f,0x4afe,0x475,0x4b18,0x47b,0x4b39,0x482,0x4b53,0xe9,0x4b78,0xe9,0x4b9e,0x489,0x4c48,0x49f,0x4cc1,
+0x4b2,0x4d73,0x4cd,0x4daa,0x4d4,0x4f8a,0x4e8,0x540a,0x510,0x2000,0x2001,0x5469,0x518,0x3000,0x3001,0x54f5,
+0,0x4000,0x400e,0x5507,0,0x5510,0,0x552a,0,0x553b,0,0x554c,0,0x5562,0,0x556b,
+0,0x5588,0,0x55a6,0,0x55c4,0,0x55e2,0,0x55f8,0,0x560c,0,0x5622,0,0x7000,
+0x7001,0x563b,0,0x7d6,0x12,0,1,0x12,0x20,0x7f4,0x4a,0,1,6,7,8,
+9,0xa,0xb,0xc,0xd,0xe,0xf,0x10,0x11,0x12,0x13,0x14,0x15,0x16,0x17,0x18,
+0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,0x54,0x5b,0x67,0x6b,
+0x76,0x7a,0x81,0x82,0x84,0x85,0xc8,0xca,0xd6,0xd8,0xda,0xdc,0xde,0xe0,0xe2,0xe4,
+0xe6,0xe8,0xe9,0xea,0xf0,0x2e,0x40,0x4c,0x5e,0x68,0x79,0x84,0x91,0x9e,0xab,0xb8,
+0xc5,0xd2,0xdf,0xec,0xf9,0x106,0x113,0x120,0x12d,0x13a,0x147,0x154,0x161,0x16e,0x17b,0x188,
+0x195,0x1a2,0x1af,0x1bc,0x1c9,0x1d6,0x1e3,0x1f0,0x1fd,0x20c,0x21b,0x22a,0x239,0x248,0x257,0x266,
+0x275,0x28f,0x2a3,0x2b7,0x2d2,0x2e1,0x2ea,0x2fa,0x302,0x30b,0x31a,0x323,0x333,0x344,0x355,0x995,
+1,0,0x17,0x8d7,0x8e8,0x8f9,0x90d,0x924,0x93c,0x94e,0x963,0x97a,0x98f,0x99f,0x9b1,0x9ce,
+0x9ea,0x9fc,0xa19,0xa35,0xa51,0xa66,0xa7b,0xa95,0xab0,0xacb,0xb37,1,0,0x135,0xaf3,0xb00,
+0xb13,0xb3b,0xb59,0xb77,0xb8f,0xbba,0xbe4,0xbfc,0xc0f,0xc22,0xc31,0xc40,0xc4f,0xc5e,0xc75,0xc86,
+0xc99,0xcac,0xcb9,0xcc6,0xcd5,0xce6,0xcfb,0xd0c,0xd17,0xd20,0xd31,0xd42,0xd55,0xd67,0xd7a,0xd8d,
+0xdcc,0xdd9,0xde6,0xdf3,0xe08,0xe38,0xe52,0xe73,0xe9e,0xec1,0xf1f,0xf46,0xf61,0xf70,0xf97,0xfbf,
+0xfe2,0x1005,0x102f,0x1048,0x1067,0x108a,0x10ae,0x10c1,0x10db,0x1105,0x111d,0x1145,0x116e,0x1181,0x1194,0x11a7,
+0x11ce,0x11dd,0x11fd,0x122b,0x1249,0x1277,0x1293,0x12ae,0x12c7,0x12e0,0x1301,0x1331,0x1350,0x1372,0x13a6,0x13d3,
+0x1418,0x1439,0x1463,0x1484,0x14ad,0x14c0,0x14f3,0x150a,0x1519,0x152a,0x1555,0x156c,0x159d,0x15cb,0x160e,0x1619,
+0x1652,0x1663,0x1674,0x1681,0x1694,0x16ce,0x16f2,0x1716,0x1750,0x1788,0x17b3,0x17cb,0x17f7,0x1823,0x1830,0x183f,
+0x185c,0x187e,0x18ac,0x18cc,0x18f3,0x191a,0x1939,0x194c,0x195d,0x196e,0x1993,0x19b8,0x19df,0x1a13,0x1a40,0x1a5e,
+0x1a71,0x1a8a,0x1ac3,0x1ad2,0x1af2,0x1b14,0x1b36,0x1b4d,0x1b64,0x1b91,0x1baa,0x1bc3,0x1bf4,0x1c1e,0x1c39,0x1c4c,
+0x1c6b,0x1c74,0x1c87,0x1ca5,0x1cc3,0x1cd6,0x1ced,0x1d02,0x1d37,0x1d5b,0x1d70,0x1d7f,0x1d92,0x1db6,0x1dbf,0x1de3,
+0x1dfa,0x1e0d,0x1e1c,0x1e27,0x1e48,0x1e60,0x1e6f,0x1e7e,0x1e8d,0x1ea4,0x1eb9,0x1ece,0x1f07,0x1f1a,0x1f36,0x1f41,
+0x1f4e,0x1f7c,0x1fa0,0x1fc3,0x1fd6,0x1ff8,0x200b,0x2026,0x2049,0x206c,0x2091,0x20a2,0x20d1,0x20fe,0x2115,0x2130,
+0x213f,0x216a,0x21a2,0x21dc,0x220a,0x221b,0x2228,0x224c,0x225b,0x2277,0x2291,0x22ae,0x22e6,0x22fb,0x2328,0x2347,
+0x2375,0x2395,0x23c9,0x23d8,0x2402,0x2425,0x2450,0x245b,0x246c,0x2487,0x24ab,0x24b8,0x24cd,0x24f4,0x251f,0x2556,
+0x2569,0x257a,0x25aa,0x25bb,0x25ca,0x25df,0x25fd,0x2610,0x2623,0x263a,0x2657,0x2662,0x266b,0x268d,0x26a2,0x26c7,
+0x26de,0x2707,0x2722,0x2737,0x2750,0x2771,0x27a6,0x27b7,0x27e8,0x280c,0x281d,0x2836,0x2841,0x286e,0x2890,0x28be,
+0x28f1,0x2900,0x2911,0x292e,0x2970,0x2997,0x29a4,0x29b9,0x29dd,0x2a03,0x2a3c,0x2a4d,0x2a71,0x2a7c,0x2a89,0x2a98,
+0x2abd,0x2aeb,0x2b07,0x2b24,0x2b31,0x2b42,0x2b60,0x2b83,0x2ba0,0x2bad,0x2bcd,0x2bea,0x2c0b,0x2c34,0x2c45,0x2c64,
+0x2c7d,0x2c96,0x2ca7,0x2cf0,0x2d01,0x2d1a,0x2d49,0x2d76,0x2d9b,0x2ddd,0x2df9,0x2e08,0x2e1f,0x2e4d,0x2e66,0x2e8f,
+0x2ea9,0x2ee4,0x2f02,0x1e85,1,0,0x12,0x2f47,0x2f57,0x2f6a,0x2f7a,0x2f8a,0x2f99,0x2fa9,0x2fbb,0x2fce,
+0x2fe0,0x2ff0,0x3000,0x300f,0x301e,0x302e,0x303b,0x304a,0x305e,0x1f43,1,0,6,0x3083,0x308e,0x309b,
+0x30a8,0x30b5,0x30c0,0x1f87,1,0,0x1e,0x30dd,0x30ec,0x3101,0x3116,0x312b,0x313f,0x3150,0x3164,0x3177,
+0x3188,0x31a1,0x31b3,0x31c4,0x31d8,0x31eb,0x3203,0x3215,0x3220,0x3230,0x323e,0x3253,0x3268,0x327e,0x3298,0x32ae,
+0x32be,0x32d2,0x32e6,0x32f7,0x330f,0x21b2,1,0,0x66,0x3337,0x335a,0x3363,0x3370,0x337b,0x3384,0x338f,
+0x3398,0x33b1,0x33b6,0x33bf,0x33dc,0x33e5,0x33f2,0x33fb,0x341f,0x3426,0x342f,0x3442,0x344d,0x3456,0x3461,0x347a,
+0x3483,0x3492,0x349d,0x34a6,0x34b1,0x34ba,0x34c1,0x34ca,0x34d5,0x34de,0x34f7,0x3500,0x350d,0x3518,0x3529,0x3534,
+0x3549,0x3560,0x3569,0x3572,0x358b,0x3596,0x359f,0x35a8,0x35bf,0x35dc,0x35e7,0x35f8,0x3603,0x360a,0x3617,0x3624,
+0x3651,0x3666,0x366f,0x368a,0x36ad,0x36ce,0x36ef,0x3714,0x373b,0x375c,0x377f,0x37a0,0x37c7,0x37e8,0x380d,0x382c,
+0x384b,0x386a,0x3887,0x38a8,0x38c9,0x38ec,0x3911,0x3930,0x394f,0x3970,0x3997,0x39bc,0x39db,0x39fc,0x3a1f,0x3a3a,
+0x3a53,0x3a6e,0x3a87,0x3aa4,0x3abf,0x3adc,0x3afb,0x3b18,0x3b35,0x3b54,0x3b71,0x3b8c,0x3ba9,0x3bc6,0x3bf9,0x24f7,
+1,0,6,0x3c31,0x3c40,0x3c50,0x3c60,0x3c70,0x3c81,0x2555,1,0,0x2b,0x3c9f,0x3cab,0x3cb9,
+0x3cc8,0x3cd7,0x3ce7,0x3cf8,0x3d0c,0x3d21,0x3d37,0x3d4a,0x3d5e,0x3d6e,0x3d77,0x3d82,0x3d92,0x3dae,0x3dc0,0x3dce,
+0x3ddd,0x3de9,0x3dfe,0x3e12,0x3e25,0x3e33,0x3e47,0x3e55,0x3e5f,0x3e71,0x3e7d,0x3e8b,0x3e9b,0x3ea2,0x3ea9,0x3eb0,
+0x3eb7,0x3ebe,0x3ed4,0x3ef5,0x870,0x3f07,0x3f12,0x3f21,0x27ae,1,0,4,0x3f3b,0x3f46,0x3f52,0x3f5c,
+0x27d4,1,0,0xc1,0x3f73,0x3f80,0x3f95,0x3fa2,0x3fb1,0x3fbf,0x3fce,0x3fdd,0x3fef,0x3ffe,0x400c,0x401d,
+0x402c,0x403b,0x4048,0x4054,0x4063,0x4072,0x407c,0x4089,0x4096,0x40a5,0x40b3,0x40c2,0x40ce,0x40d8,0x40e4,0x40f4,
+0x4104,0x4112,0x411e,0x412f,0x413b,0x4147,0x4155,0x4162,0x416e,0x417b,0xd0c,0x4188,0x4196,0x41b0,0x41b9,0x41c7,
+0x41d5,0x41e1,0x41f0,0x41fe,0x420c,0x4218,0x4227,0x4235,0x4243,0x4250,0x425f,0x427a,0x4289,0x429a,0x42ab,0x42be,
+0x42d0,0x42df,0x42f1,0x4300,0x430c,0x4317,0x1e1c,0x4324,0x432f,0x433a,0x4345,0x4350,0x436b,0x4376,0x4381,0x438c,
+0x439f,0x43b3,0x43be,0x43cd,0x43dc,0x43e7,0x43f2,0x43ff,0x440e,0x441c,0x4427,0x4442,0x444c,0x445d,0x446e,0x447d,
+0x448e,0x4499,0x44a4,0x44af,0x44ba,0x44c5,0x44d0,0x44db,0x44e5,0x44f0,0x4500,0x450b,0x4519,0x4526,0x4531,0x4540,
+0x454d,0x455a,0x4569,0x4576,0x4587,0x4599,0x45a9,0x45b4,0x45c7,0x45de,0x45ec,0x45f9,0x4604,0x4611,0x4622,0x463e,
+0x4654,0x465f,0x467c,0x468c,0x469b,0x46a6,0x46b1,0x1f36,0x46bd,0x46c8,0x46e0,0x46f0,0x46ff,0x470d,0x471b,0x4726,
+0x4731,0x4745,0x475c,0x4774,0x4784,0x4794,0x47a4,0x47b6,0x47c1,0x47cc,0x47d6,0x47e2,0x47f0,0x4803,0x480f,0x481c,
+0x4827,0x4843,0x4850,0x485e,0x4877,0x2836,0x4886,0x2657,0x4893,0x48a1,0x48b3,0x48c1,0x48cd,0x48dd,0x2a71,0x48eb,
+0x48f7,0x4902,0x490d,0x4918,0x492c,0x493a,0x4951,0x495d,0x4971,0x497f,0x4991,0x49a7,0x49b5,0x49c7,0x49d5,0x49f2,
+0x4a04,0x4a11,0x4a22,0x4a34,0x4a4e,0x31cc,1,0,6,0x4a75,0x4a88,0x4a98,0x4aa6,0x4ab7,0x4ac7,0x3228,
+0x12,0,1,0x4af1,0x4af7,0x3235,0x12,0,1,0x4af1,0x4af7,0x3242,1,0,3,0x4af1,
+0x4af7,0x4b30,0x3258,1,0,3,0x4af1,0x4af7,0x4b30,0x326e,1,0,0x12,0x4bba,0x4bc4,0x4bd0,
+0x4bd7,0x4be2,0x4be7,0x4bee,0x4bf5,0x4bfe,0x4c03,0x4c08,0x4c18,0x870,0x3f07,0x4c24,0x3f12,0x4c34,0x3f21,0x3317,
+1,0,0xf,0x4bba,0x4c5b,0x4c65,0x4c6f,0x4c7a,0x3ddd,0x4c84,0x4c90,0x4c98,0x4c9f,0x4ca9,0x4bd0,0x4bd7,
+0x4be7,0x4cb3,0x339e,1,0,0x17,0x4bba,0x4cd0,0x4c6f,0x4cdc,0x4ce9,0x4cf7,0x3ddd,0x4d02,0x4bd0,0x4d13,
+0x4be7,0x4d22,0x4d30,0x870,0x3ef5,0x4d3c,0x4d4d,0x3f07,0x4c24,0x3f12,0x4c34,0x3f21,0x4d5e,0x34bb,1,0,
+3,0x4d91,0x4d99,0x4da1,0x34d4,1,0,0x10,0x4dca,0x4dd1,0x4de0,0x4e01,0x4e24,0x4e2f,0x4e4e,0x4e65,
+0x4e72,0x4e7b,0x4e9a,0x4ecd,0x4ee8,0x4f17,0x4f34,0x4f59,0x356d,1,0,0x24,0x4fa8,0x4fb5,0x4fc8,0x4fd5,
+0x5002,0x5027,0x503c,0x505b,0x507c,0x50a9,0x50e2,0x5105,0x5128,0x5155,0x518a,0x51b1,0x51da,0x5211,0x5240,0x5261,
+0x5286,0x5295,0x52b8,0x52cf,0x52dc,0x52eb,0x5308,0x5321,0x5344,0x5369,0x5382,0x5397,0x53a6,0x53b7,0x53c4,0x53e5,
+0x373d,1,0,4,0x5423,0x542e,0x5446,0x545e,0x3779,0x36,1,2,4,8,0xe,0x10,
+0x20,0x3e,0x40,0x80,0x100,0x1c0,0x200,0x400,0x800,0xe00,0x1000,0x2000,0x4000,0x7000,0x8000,0x10000,
+0x20000,0x40000,0x78001,0x80000,0x100000,0x200000,0x400000,0x800000,0x1000000,0x2000000,0x4000000,0x8000000,0xf000000,0x10000000,0x20000000,0x30f80000,
+0x30dd,0x30ec,0x3101,0x3116,0x5497,0x312b,0x313f,0x548d,0x3150,0x3164,0x3177,0x54a8,0x3188,0x31a1,0x31b3,0x54bf,
+0x31c4,0x31d8,0x31eb,0x54e8,0x3203,0x3215,0x3220,0x3230,0x5484,0x323e,0x3253,0x3268,0x327e,0x3298,0x32ae,0x32be,
+0x32d2,0x32e6,0x54de,0x32f7,0x330f,0x54c9
+};
+
+const uint8_t PropNameData::bytesTries[14992]={
+0,0x15,0x6d,0xc3,0x78,0x73,0xc2,0x12,0x76,0x7a,0x76,0x6a,0x77,0xa2,0x52,0x78,
+1,0x64,0x50,0x69,0x10,0x64,1,0x63,0x30,0x73,0x62,0x13,0x74,0x61,0x72,0x74,
+0x63,0x60,0x16,0x6f,0x6e,0x74,0x69,0x6e,0x75,0x65,0x61,0x13,0x69,0x67,0x69,0x74,
+0x81,3,0x61,0x2e,0x65,0x4c,0x6f,0xc3,0x18,0x73,0x69,0x1e,0x72,0x69,0x61,0x74,
+0x69,0x6f,0x6e,0x73,0x65,0x6c,0x65,0x63,0x74,0x6f,0x72,0x69,0x10,0x72,0x1f,0x74,
+0x69,0x63,0x61,0x6c,0x6f,0x72,0x69,0x65,0x6e,0x74,0x61,0x74,0x69,0x6f,0x6e,0xc3,
+0x18,3,0x62,0xc3,0x14,0x68,0x32,0x6f,0x42,0x73,0x13,0x70,0x61,0x63,0x65,0x5f,
+0x17,0x69,0x74,0x65,0x73,0x70,0x61,0x63,0x65,0x5f,0x16,0x72,0x64,0x62,0x72,0x65,
+0x61,0x6b,0xc3,0x14,0x73,0xa2,0x49,0x74,0xa4,0x3b,0x75,3,0x63,0xd9,0x40,0xc,
+0x69,0x52,0x6e,0x58,0x70,0x12,0x70,0x65,0x72,0x5c,0x13,0x63,0x61,0x73,0x65,0x5c,
+0x16,0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,0x40,0xc,0x12,0x64,0x65,0x6f,0x5b,
+0x10,0x69,1,0x63,0x3e,0x66,0x1b,0x69,0x65,0x64,0x69,0x64,0x65,0x6f,0x67,0x72,
+0x61,0x70,0x68,0x5b,0x17,0x6f,0x64,0x65,0x31,0x6e,0x61,0x6d,0x65,0xd9,0x40,0xb,
+0xa,0x69,0x84,0x70,0x19,0x70,0x30,0x74,0x36,0x75,0x10,0x63,0xd9,0x40,9,0x12,
+0x61,0x63,0x65,0x5f,1,0x63,0xd9,0x40,8,0x65,0x11,0x72,0x6d,0x67,0x69,0x3c,
+0x6c,0xa2,0x5f,0x6f,0x17,0x66,0x74,0x64,0x6f,0x74,0x74,0x65,0x64,0x57,0x13,0x6d,
+0x70,0x6c,0x65,3,0x63,0x50,0x6c,0x68,0x74,0x8a,0x75,0x1e,0x70,0x70,0x65,0x72,
+0x63,0x61,0x73,0x65,0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,0x40,9,0x19,0x61,
+0x73,0x65,0x66,0x6f,0x6c,0x64,0x69,0x6e,0x67,0xd9,0x40,6,0x1e,0x6f,0x77,0x65,
+0x72,0x63,0x61,0x73,0x65,0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,0x40,7,0x1e,
+0x69,0x74,0x6c,0x65,0x63,0x61,0x73,0x65,0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,
+0x40,8,0x10,0x63,0xd9,0x40,7,0x62,0xc3,0x13,0x63,0x34,0x64,0x57,0x65,0x6e,
+0x66,0x10,0x63,0xd9,0x40,6,0xc2,0xa,2,0x66,0xd9,0x40,6,0x72,0x28,0x78,
+0xd9,0x70,0,0x12,0x69,0x70,0x74,0xc2,0xa,0x19,0x65,0x78,0x74,0x65,0x6e,0x73,
+0x69,0x6f,0x6e,0x73,0xd9,0x70,0,1,0x67,0x6a,0x6e,1,0x73,0x54,0x74,0x13,
+0x65,0x6e,0x63,0x65,1,0x62,0x34,0x74,0x16,0x65,0x72,0x6d,0x69,0x6e,0x61,0x6c,
+0x67,0x13,0x72,0x65,0x61,0x6b,0xc3,0x13,0x14,0x69,0x74,0x69,0x76,0x65,0x65,1,
+0x6d,0x2e,0x73,0x13,0x74,0x61,0x72,0x74,0x73,0x19,0x65,0x6e,0x74,0x73,0x74,0x61,
+0x72,0x74,0x65,0x72,0x73,3,0x63,0x66,0x65,0x72,0x69,0x98,0x72,0x19,0x61,0x69,
+0x6c,0x63,0x61,0x6e,0x6f,0x6e,0x69,0x63,0x1f,0x61,0x6c,0x63,0x6f,0x6d,0x62,0x69,
+0x6e,0x69,0x6e,0x67,0x63,0x6c,0x61,0x73,0x73,0xc3,0x11,0xd8,0x40,0xa,0x11,0x63,
+0x63,0xc3,0x11,0x11,0x72,0x6d,0x58,0x1e,0x69,0x6e,0x61,0x6c,0x70,0x75,0x6e,0x63,
+0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x59,0x1d,0x74,0x6c,0x65,0x63,0x61,0x73,0x65,
+0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,0x40,0xa,0x6d,0x70,0x6e,0x76,0x70,0xa2,
+0xf1,0x71,0xa4,0x43,0x72,2,0x61,0x28,0x65,0x32,0x69,0x9d,0x14,0x64,0x69,0x63,
+0x61,0x6c,0x55,0x1e,0x67,0x69,0x6f,0x6e,0x61,0x6c,0x69,0x6e,0x64,0x69,0x63,0x61,
+0x74,0x6f,0x72,0x9d,0x12,0x61,0x74,0x68,0x4f,6,0x6f,0x39,0x6f,0x32,0x74,0xc3,
+9,0x75,0x54,0x76,0xd9,0x30,0,0x12,0x6e,0x63,0x68,0x1f,0x61,0x72,0x61,0x63,
+0x74,0x65,0x72,0x63,0x6f,0x64,0x65,0x70,0x6f,0x69,0x6e,0x74,0x51,0x14,0x6d,0x65,
+0x72,0x69,0x63,1,0x74,0x32,0x76,0x13,0x61,0x6c,0x75,0x65,0xd9,0x30,0,0x12,
+0x79,0x70,0x65,0xc3,9,0x61,0xa2,0x77,0x63,0xa2,0x82,0x66,2,0x63,0x98,0x64,
+0xa2,0x53,0x6b,1,0x63,0x56,0x64,1,0x69,0x42,0x71,1,0x63,0xc3,0xd,0x75,
+0x17,0x69,0x63,0x6b,0x63,0x68,0x65,0x63,0x6b,0xc3,0xd,0x13,0x6e,0x65,0x72,0x74,
+0x6d,1,0x69,0x42,0x71,1,0x63,0xc3,0xf,0x75,0x17,0x69,0x63,0x6b,0x63,0x68,
+0x65,0x63,0x6b,0xc3,0xf,0x13,0x6e,0x65,0x72,0x74,0x71,1,0x69,0x42,0x71,1,
+0x63,0xc3,0xe,0x75,0x17,0x69,0x63,0x6b,0x63,0x68,0x65,0x63,0x6b,0xc3,0xe,0x13,
+0x6e,0x65,0x72,0x74,0x6f,1,0x69,0x42,0x71,1,0x63,0xc3,0xc,0x75,0x17,0x69,
+0x63,0x6b,0x63,0x68,0x65,0x63,0x6b,0xc3,0xc,0x13,0x6e,0x65,0x72,0x74,0x6b,0xd8,
+0x40,5,1,0x31,0xd9,0x40,0xb,0x6d,0x10,0x65,0xd9,0x40,5,0x12,0x68,0x61,
+0x72,0x51,2,0x61,0x6c,0x63,0xa2,0x4c,0x72,1,0x65,0x2a,0x69,0x11,0x6e,0x74,
+0x7f,0x16,0x70,0x65,0x6e,0x64,0x65,0x64,0x63,0x1f,0x6f,0x6e,0x63,0x61,0x74,0x65,
+0x6e,0x61,0x74,0x69,0x6f,0x6e,0x6d,0x61,0x72,0x6b,0x9f,0x10,0x74,2,0x73,0x2c,
+0x74,0x30,0x77,0x10,0x73,0x77,0x11,0x79,0x6e,0x75,0x12,0x65,0x72,0x6e,1,0x73,
+0x38,0x77,0x18,0x68,0x69,0x74,0x65,0x73,0x70,0x61,0x63,0x65,0x77,0x14,0x79,0x6e,
+0x74,0x61,0x78,0x75,0x10,0x6d,0x9f,1,0x6d,0x3c,0x75,0x1a,0x6f,0x74,0x61,0x74,
+0x69,0x6f,0x6e,0x6d,0x61,0x72,0x6b,0x53,0x12,0x61,0x72,0x6b,0x53,0x66,0xc1,0xf8,
+0x69,0xc1,0x3c,0x69,0xa2,0x6f,0x6a,0xa4,9,0x6c,4,0x62,0xc3,8,0x63,0x8c,
+0x65,0x98,0x69,0xa2,0x56,0x6f,2,0x65,0x4b,0x67,0x4c,0x77,0x11,0x65,0x72,0x4c,
+0x13,0x63,0x61,0x73,0x65,0x4c,0x16,0x6d,0x61,0x70,0x70,0x69,0x6e,0x67,0xd9,0x40,
+4,0x11,0x69,0x63,0x1f,0x61,0x6c,0x6f,0x72,0x64,0x65,0x72,0x65,0x78,0x63,0x65,
+0x70,0x74,0x69,0x6f,0x6e,0x4b,0xd8,0x40,4,0x11,0x63,0x63,0xc3,0x10,0x18,0x61,
+0x64,0x63,0x61,0x6e,0x6f,0x6e,0x69,0x63,0x1f,0x61,0x6c,0x63,0x6f,0x6d,0x62,0x69,
+0x6e,0x69,0x6e,0x67,0x63,0x6c,0x61,0x73,0x73,0xc3,0x10,0x16,0x6e,0x65,0x62,0x72,
+0x65,0x61,0x6b,0xc3,8,2,0x64,0x4a,0x6e,0xa2,0x5b,0x73,1,0x63,0xd9,0x40,
+3,0x6f,0x16,0x63,0x6f,0x6d,0x6d,0x65,0x6e,0x74,0xd9,0x40,3,2,0x63,0x80,
+0x65,0x90,0x73,0x40,1,0x62,0x52,0x74,0x46,1,0x61,0x40,0x72,0x1c,0x69,0x6e,
+0x61,0x72,0x79,0x6f,0x70,0x65,0x72,0x61,0x74,0x6f,0x72,0x47,0x11,0x72,0x74,0x41,
+0x44,0x1c,0x69,0x6e,0x61,0x72,0x79,0x6f,0x70,0x65,0x72,0x61,0x74,0x6f,0x72,0x45,
+0x3e,0x16,0x6f,0x6e,0x74,0x69,0x6e,0x75,0x65,0x3f,0x10,0x6f,0x42,0x16,0x67,0x72,
+0x61,0x70,0x68,0x69,0x63,0x43,2,0x64,0x2e,0x70,0x86,0x73,0x10,0x63,0xc3,0x17,
+0x11,0x69,0x63,1,0x70,0x46,0x73,0x1e,0x79,0x6c,0x6c,0x61,0x62,0x69,0x63,0x63,
+0x61,0x74,0x65,0x67,0x6f,0x72,0x79,0xc3,0x17,0x10,0x6f,0x1f,0x73,0x69,0x74,0x69,
+0x6f,0x6e,0x61,0x6c,0x63,0x61,0x74,0x65,0x67,0x6f,0x72,0x79,0xc3,0x16,0x10,0x63,
+0xc3,0x16,2,0x67,0xc3,6,0x6f,0x26,0x74,0xc3,7,0x11,0x69,0x6e,1,0x63,
+0x4a,0x69,0x11,0x6e,0x67,1,0x67,0x2e,0x74,0x12,0x79,0x70,0x65,0xc3,7,0x13,
+0x72,0x6f,0x75,0x70,0xc3,6,0x48,0x15,0x6f,0x6e,0x74,0x72,0x6f,0x6c,0x49,0x66,
+0x86,0x67,0xa2,0x4a,0x68,3,0x61,0x36,0x65,0x58,0x73,0x68,0x79,0x13,0x70,0x68,
+0x65,0x6e,0x3d,0x1f,0x6e,0x67,0x75,0x6c,0x73,0x79,0x6c,0x6c,0x61,0x62,0x6c,0x65,
+0x74,0x79,0x70,0x65,0xc3,0xb,0x10,0x78,0x3a,0x14,0x64,0x69,0x67,0x69,0x74,0x3b,
+0x10,0x74,0xc3,0xb,0x16,0x75,0x6c,0x6c,0x63,0x6f,0x6d,0x70,0x1f,0x6f,0x73,0x69,
+0x74,0x69,0x6f,0x6e,0x65,0x78,0x63,0x6c,0x75,0x73,0x69,0x6f,0x6e,0x33,2,0x63,
+0xa2,0x44,0x65,0xa2,0x4b,0x72,3,0x61,0x34,0x62,0x84,0x65,0x8a,0x6c,0x12,0x69,
+0x6e,0x6b,0x39,0x11,0x70,0x68,0x7c,0x12,0x65,0x6d,0x65,3,0x62,0x5e,0x63,0x30,
+0x65,0x48,0x6c,0x12,0x69,0x6e,0x6b,0x39,0x1a,0x6c,0x75,0x73,0x74,0x65,0x72,0x62,
+0x72,0x65,0x61,0x6b,0xc3,0x12,0x14,0x78,0x74,0x65,0x6e,0x64,0x37,0x12,0x61,0x73,
+0x65,0x35,0x11,0x78,0x74,0x37,0xc2,5,1,0x62,0xc3,0x12,0x6d,0xd9,0x20,0,
+0x1c,0x6e,0x65,0x72,0x61,0x6c,0x63,0x61,0x74,0x65,0x67,0x6f,0x72,0x79,0xc2,5,
+0x13,0x6d,0x61,0x73,0x6b,0xd9,0x20,0,0x61,0xa2,0x90,0x62,0xa2,0xbe,0x63,0xa4,
+0x30,0x64,0xa4,0xfd,0x65,5,0x6d,0x63,0x6d,0x6e,0x70,0xa2,0x59,0x78,0x10,0x74,
+0x30,1,0x65,0x2c,0x70,0x12,0x69,0x63,0x74,0xa1,0x12,0x6e,0x64,0x65,1,0x64,
+0x24,0x72,0x31,0x1b,0x70,0x69,0x63,0x74,0x6f,0x67,0x72,0x61,0x70,0x68,0x69,0x63,
+0xa1,0x10,0x6f,1,0x64,0x97,0x6a,0x10,0x69,0x92,2,0x63,0x40,0x6d,0x50,0x70,
+0x1a,0x72,0x65,0x73,0x65,0x6e,0x74,0x61,0x74,0x69,0x6f,0x6e,0x95,0x17,0x6f,0x6d,
+0x70,0x6f,0x6e,0x65,0x6e,0x74,0x9b,0x16,0x6f,0x64,0x69,0x66,0x69,0x65,0x72,0x96,
+0x13,0x62,0x61,0x73,0x65,0x99,0x12,0x72,0x65,0x73,0x95,0x61,0x30,0x62,0x4e,0x63,
+0x12,0x6f,0x6d,0x70,0x9b,0xc2,4,0x1b,0x73,0x74,0x61,0x73,0x69,0x61,0x6e,0x77,
+0x69,0x64,0x74,0x68,0xc3,4,0x12,0x61,0x73,0x65,0x99,3,0x67,0x44,0x68,0x4a,
+0x6c,0x4e,0x73,0x1a,0x63,0x69,0x69,0x68,0x65,0x78,0x64,0x69,0x67,0x69,0x74,0x23,
+0x10,0x65,0xd9,0x40,0,0x11,0x65,0x78,0x23,1,0x6e,0x38,0x70,0x11,0x68,0x61,
+0x20,0x14,0x62,0x65,0x74,0x69,0x63,0x21,0x11,0x75,0x6d,0x79,4,0x63,0xc3,0,
+0x69,0x3e,0x6c,0xa2,0x57,0x6d,0xa2,0x64,0x70,1,0x62,0xd9,0x40,0xd,0x74,0xc3,
+0x15,0x11,0x64,0x69,2,0x63,0x54,0x6d,0x74,0x70,0x1b,0x61,0x69,0x72,0x65,0x64,
+0x62,0x72,0x61,0x63,0x6b,0x65,0x74,0xd8,0x40,0xd,0x13,0x74,0x79,0x70,0x65,0xc3,
+0x15,0x24,1,0x6c,0x30,0x6f,0x14,0x6e,0x74,0x72,0x6f,0x6c,0x25,0x12,0x61,0x73,
+0x73,0xc3,0,0x26,0x14,0x69,0x72,0x72,0x6f,0x72,1,0x65,0x38,0x69,0x16,0x6e,
+0x67,0x67,0x6c,0x79,0x70,0x68,0xd9,0x40,1,0x10,0x64,0x27,2,0x61,0x32,0x6b,
+0xc3,1,0x6f,0x11,0x63,0x6b,0xc3,1,0x11,0x6e,0x6b,0x7b,0x10,0x67,0xd9,0x40,
+1,6,0x68,0x7c,0x68,0x54,0x69,0x85,0x6f,0xa2,0x6f,0x77,4,0x63,0x30,0x6b,
+0x36,0x6c,0x87,0x74,0x8b,0x75,0x89,1,0x66,0x8d,0x6d,0x8f,0x11,0x63,0x66,0x91,
+0x18,0x61,0x6e,0x67,0x65,0x73,0x77,0x68,0x65,0x6e,4,0x63,0x44,0x6c,0x6c,0x6e,
+0x7e,0x74,0x98,0x75,0x18,0x70,0x70,0x65,0x72,0x63,0x61,0x73,0x65,0x64,0x89,0x12,
+0x61,0x73,0x65,1,0x66,0x30,0x6d,0x14,0x61,0x70,0x70,0x65,0x64,0x8f,0x14,0x6f,
+0x6c,0x64,0x65,0x64,0x8d,0x18,0x6f,0x77,0x65,0x72,0x63,0x61,0x73,0x65,0x64,0x87,
+0x1c,0x66,0x6b,0x63,0x63,0x61,0x73,0x65,0x66,0x6f,0x6c,0x64,0x65,0x64,0x91,0x18,
+0x69,0x74,0x6c,0x65,0x63,0x61,0x73,0x65,0x64,0x8b,0x13,0x6d,0x70,0x65,0x78,0x33,
+0x61,0x2e,0x63,0xa2,0x48,0x66,0xd9,0x40,2,1,0x6e,0x72,0x73,0x10,0x65,3,
+0x64,0x83,0x66,0x3a,0x69,0x4a,0x73,0x17,0x65,0x6e,0x73,0x69,0x74,0x69,0x76,0x65,
+0x65,0x15,0x6f,0x6c,0x64,0x69,0x6e,0x67,0xd9,0x40,2,0x17,0x67,0x6e,0x6f,0x72,
+0x61,0x62,0x6c,0x65,0x85,0x13,0x6f,0x6e,0x69,0x63,0x1f,0x61,0x6c,0x63,0x6f,0x6d,
+0x62,0x69,0x6e,0x69,0x6e,0x67,0x63,0x6c,0x61,0x73,0x73,0xc3,2,0x10,0x63,0xc3,
+2,3,0x61,0x30,0x65,0x34,0x69,0xa2,0x41,0x74,0xc3,3,0x11,0x73,0x68,0x29,
+2,0x63,0x3a,0x66,0x58,0x70,0x2c,0x16,0x72,0x65,0x63,0x61,0x74,0x65,0x64,0x2d,
+0x1d,0x6f,0x6d,0x70,0x6f,0x73,0x69,0x74,0x69,0x6f,0x6e,0x74,0x79,0x70,0x65,0xc3,
+3,0x15,0x61,0x75,0x6c,0x74,0x69,0x67,0x1f,0x6e,0x6f,0x72,0x61,0x62,0x6c,0x65,
+0x63,0x6f,0x64,0x65,0x70,0x6f,0x69,0x6e,0x74,0x2b,0x2a,0x10,0x61,0x2e,0x15,0x63,
+0x72,0x69,0x74,0x69,0x63,0x2f,3,0x66,0x34,0x6e,0x3e,0x74,0x42,0x79,0x22,0x11,
+0x65,0x73,0x23,0x20,0x13,0x61,0x6c,0x73,0x65,0x21,0x20,0x10,0x6f,0x21,0x22,0x12,
+0x72,0x75,0x65,0x23,0xb,0x6b,0x5b,0x6f,0x23,0x6f,0x3c,0x72,0x4c,0x76,1,0x69,
+0x24,0x72,0x33,0x13,0x72,0x61,0x6d,0x61,0x33,0x10,0x76,0x22,0x14,0x65,0x72,0x6c,
+0x61,0x79,0x23,0xa2,0xe2,0x13,0x69,0x67,0x68,0x74,0xa3,0xe2,0x6b,0x58,0x6c,0x74,
+0x6e,3,0x6b,0x2f,0x6f,0x30,0x72,0x21,0x75,0x12,0x6b,0x74,0x61,0x2f,0x19,0x74,
+0x72,0x65,0x6f,0x72,0x64,0x65,0x72,0x65,0x64,0x21,1,0x61,0x24,0x76,0x31,0x18,
+0x6e,0x61,0x76,0x6f,0x69,0x63,0x69,0x6e,0x67,0x31,0xa2,0xe0,0x12,0x65,0x66,0x74,
+0xa3,0xe0,0x64,0x45,0x64,0x4e,0x68,0x88,0x69,1,0x6f,0x26,0x73,0xa3,0xf0,0x1a,
+0x74,0x61,0x73,0x75,0x62,0x73,0x63,0x72,0x69,0x70,0x74,0xa3,0xf0,2,0x61,0xa3,
+0xea,0x62,0xa3,0xe9,0x6f,0x13,0x75,0x62,0x6c,0x65,1,0x61,0x30,0x62,0x13,0x65,
+0x6c,0x6f,0x77,0xa3,0xe9,0x13,0x62,0x6f,0x76,0x65,0xa3,0xea,0x12,0x61,0x6e,0x72,
+0x2c,0x15,0x65,0x61,0x64,0x69,0x6e,0x67,0x2d,0x61,0xa2,0x7b,0x62,0xa2,0xd4,0x63,
+0x11,0x63,0x63,4,0x31,0x3c,0x32,0xa2,0x42,0x33,0xa2,0x56,0x38,0xa2,0x64,0x39,
+0x10,0x31,0xa3,0x5b,9,0x35,0xa,0x35,0x3f,0x36,0x41,0x37,0x43,0x38,0x45,0x39,
+0x47,0x30,0x30,0x31,0x3c,0x32,0x42,0x33,0x4e,0x34,0x3d,0x34,1,0x33,0xa3,0x67,
+0x37,0xa3,0x6b,0x36,0x10,0x38,0xa3,0x76,0x38,1,0x32,0xa3,0x7a,0x39,0xa3,0x81,
+0x3a,2,0x30,0xa3,0x82,0x32,0xa3,0x84,0x33,0xa3,0x85,9,0x35,0xa,0x35,0x53,
+0x36,0x55,0x37,0x57,0x38,0x59,0x39,0x5b,0x30,0x49,0x31,0x4b,0x32,0x4d,0x33,0x4f,
+0x34,0x51,6,0x33,8,0x33,0x63,0x34,0x65,0x35,0x67,0x36,0x69,0x30,0x5d,0x31,
+0x5f,0x32,0x61,0x10,0x34,0xa3,0x54,0xa2,0xe6,3,0x62,0xa0,0x6c,0xa3,0xe4,0x72,
+0xa3,0xe8,0x74,2,0x61,0x74,0x62,0x7c,0x74,0x14,0x61,0x63,0x68,0x65,0x64,1,
+0x61,0x3e,0x62,0x13,0x65,0x6c,0x6f,0x77,0xa2,0xca,0x13,0x6c,0x65,0x66,0x74,0xa3,
+0xc8,0x13,0x62,0x6f,0x76,0x65,0xa2,0xd6,0x14,0x72,0x69,0x67,0x68,0x74,0xa3,0xd8,
+0xa2,0xd6,0x10,0x72,0xa3,0xd8,0xa2,0xca,0x10,0x6c,0xa3,0xc8,0x12,0x6f,0x76,0x65,
+0xa2,0xe6,1,0x6c,0x30,0x72,0x13,0x69,0x67,0x68,0x74,0xa3,0xe8,0x12,0x65,0x66,
+0x74,0xa3,0xe4,0xa2,0xdc,2,0x65,0x2c,0x6c,0xa3,0xda,0x72,0xa3,0xde,0x12,0x6c,
+0x6f,0x77,0xa2,0xdc,1,0x6c,0x30,0x72,0x13,0x69,0x67,0x68,0x74,0xa3,0xde,0x12,
+0x65,0x66,0x74,0xa3,0xda,0xb,0x6e,0xc0,0xca,0x72,0x5f,0x72,0x46,0x73,0xa2,0x48,
+0x77,1,0x68,0x24,0x73,0x33,0x17,0x69,0x74,0x65,0x73,0x70,0x61,0x63,0x65,0x33,
+0x22,1,0x69,0x30,0x6c,2,0x65,0x3d,0x69,0x4b,0x6f,0x3f,0x18,0x67,0x68,0x74,
+0x74,0x6f,0x6c,0x65,0x66,0x74,0x22,2,0x65,0x38,0x69,0x48,0x6f,0x16,0x76,0x65,
+0x72,0x72,0x69,0x64,0x65,0x3f,0x17,0x6d,0x62,0x65,0x64,0x64,0x69,0x6e,0x67,0x3d,
+0x15,0x73,0x6f,0x6c,0x61,0x74,0x65,0x4b,0x30,0x1e,0x65,0x67,0x6d,0x65,0x6e,0x74,
+0x73,0x65,0x70,0x61,0x72,0x61,0x74,0x6f,0x72,0x31,0x6e,0xa2,0x41,0x6f,0xa2,0x53,
+0x70,2,0x61,0x66,0x64,0x86,0x6f,0x1b,0x70,0x64,0x69,0x72,0x65,0x63,0x74,0x69,
+0x6f,0x6e,0x61,0x6c,1,0x66,0x32,0x69,0x15,0x73,0x6f,0x6c,0x61,0x74,0x65,0x4d,
+0x14,0x6f,0x72,0x6d,0x61,0x74,0x41,0x1f,0x72,0x61,0x67,0x72,0x61,0x70,0x68,0x73,
+0x65,0x70,0x61,0x72,0x61,0x74,0x6f,0x72,0x2f,1,0x66,0x41,0x69,0x4d,1,0x6f,
+0x28,0x73,0x10,0x6d,0x43,0x1b,0x6e,0x73,0x70,0x61,0x63,0x69,0x6e,0x67,0x6d,0x61,
+0x72,0x6b,0x43,1,0x6e,0x35,0x74,0x19,0x68,0x65,0x72,0x6e,0x65,0x75,0x74,0x72,
+0x61,0x6c,0x35,0x65,0x88,0x65,0x98,0x66,0xa2,0x6a,0x6c,0x20,1,0x65,0x30,0x72,
+2,0x65,0x37,0x69,0x49,0x6f,0x39,0x18,0x66,0x74,0x74,0x6f,0x72,0x69,0x67,0x68,
+0x74,0x20,2,0x65,0x38,0x69,0x48,0x6f,0x16,0x76,0x65,0x72,0x72,0x69,0x64,0x65,
+0x39,0x17,0x6d,0x62,0x65,0x64,0x64,0x69,0x6e,0x67,0x37,0x15,0x73,0x6f,0x6c,0x61,
+0x74,0x65,0x49,3,0x6e,0x25,0x73,0x27,0x74,0x29,0x75,0x15,0x72,0x6f,0x70,0x65,
+0x61,0x6e,2,0x6e,0x3c,0x73,0x46,0x74,0x18,0x65,0x72,0x6d,0x69,0x6e,0x61,0x74,
+0x6f,0x72,0x29,0x14,0x75,0x6d,0x62,0x65,0x72,0x25,0x17,0x65,0x70,0x61,0x72,0x61,
+0x74,0x6f,0x72,0x27,1,0x69,0x28,0x73,0x10,0x69,0x47,0x1f,0x72,0x73,0x74,0x73,
+0x74,0x72,0x6f,0x6e,0x67,0x69,0x73,0x6f,0x6c,0x61,0x74,0x65,0x47,0x61,0x4e,0x62,
+0x84,0x63,1,0x6f,0x24,0x73,0x2d,0x1c,0x6d,0x6d,0x6f,0x6e,0x73,0x65,0x70,0x61,
+0x72,0x61,0x74,0x6f,0x72,0x2d,2,0x6c,0x3b,0x6e,0x2b,0x72,0x13,0x61,0x62,0x69,
+0x63,1,0x6c,0x30,0x6e,0x14,0x75,0x6d,0x62,0x65,0x72,0x2b,0x14,0x65,0x74,0x74,
+0x65,0x72,0x3b,0x2e,1,0x6e,0x45,0x6f,0x1c,0x75,0x6e,0x64,0x61,0x72,0x79,0x6e,
+0x65,0x75,0x74,0x72,0x61,0x6c,0x45,0,0x16,0x6d,0xc8,0xc8,0x74,0xc1,0xee,0x77,
+0x6a,0x77,0x48,0x79,0x70,0x7a,0x1d,0x61,0x6e,0x61,0x62,0x61,0x7a,0x61,0x72,0x73,
+0x71,0x75,0x61,0x72,0x65,0xa5,0x18,0x10,0x61,1,0x6e,0x36,0x72,0x16,0x61,0x6e,
+0x67,0x63,0x69,0x74,0x69,0xa3,0xfc,0x12,0x63,0x68,0x6f,0xa5,0x2c,1,0x65,0x88,
+0x69,2,0x6a,0x3c,0x72,0x68,0x73,0x17,0x79,0x6c,0x6c,0x61,0x62,0x6c,0x65,0x73,
+0xa3,0x48,0x12,0x69,0x6e,0x67,0xa2,0x74,0x1e,0x68,0x65,0x78,0x61,0x67,0x72,0x61,
+0x6d,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0x74,0x16,0x61,0x64,0x69,0x63,0x61,
+0x6c,0x73,0xa3,0x49,0x13,0x7a,0x69,0x64,0x69,0xa5,0x34,0x74,0xa2,0x59,0x75,0xa4,
+0x35,0x76,2,0x61,0x36,0x65,0x7a,0x73,0xa2,0x6c,0x12,0x73,0x75,0x70,0xa3,0x7d,
+1,0x69,0xa3,0x9f,0x72,0x1e,0x69,0x61,0x74,0x69,0x6f,0x6e,0x73,0x65,0x6c,0x65,
+0x63,0x74,0x6f,0x72,0x73,0xa2,0x6c,0x19,0x73,0x75,0x70,0x70,0x6c,0x65,0x6d,0x65,
+0x6e,0x74,0xa3,0x7d,1,0x64,0x3c,0x72,0x19,0x74,0x69,0x63,0x61,0x6c,0x66,0x6f,
+0x72,0x6d,0x73,0xa3,0x91,0x14,0x69,0x63,0x65,0x78,0x74,0xa2,0xaf,0x16,0x65,0x6e,
+0x73,0x69,0x6f,0x6e,0x73,0xa3,0xaf,4,0x61,0x68,0x65,0xa2,0xad,0x68,0xa2,0xb0,
+0x69,0xa2,0xb8,0x72,0x1c,0x61,0x6e,0x73,0x70,0x6f,0x72,0x74,0x61,0x6e,0x64,0x6d,
+0x61,0x70,0xa2,0xcf,0x16,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0xcf,4,0x67,
+0x7e,0x69,0xa2,0x41,0x6b,0xa2,0x6a,0x6d,0xa2,0x6c,0x6e,0x12,0x67,0x75,0x74,0xa4,
+0x10,1,0x63,0x40,0x73,0x11,0x75,0x70,0xa4,0x33,0x16,0x70,0x6c,0x65,0x6d,0x65,
+0x6e,0x74,0xa5,0x33,0x18,0x6f,0x6d,0x70,0x6f,0x6e,0x65,0x6e,0x74,0x73,0xa5,0x11,
+2,0x61,0x2a,0x62,0x32,0x73,0xa3,0x60,0x12,0x6c,0x6f,0x67,0xa3,0x62,0x13,0x61,
+0x6e,0x77,0x61,0xa3,0x65,3,0x6c,0x52,0x74,0x56,0x76,0x5e,0x78,0x16,0x75,0x61,
+0x6e,0x6a,0x69,0x6e,0x67,0xa2,0x7c,0x16,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,
+0x7c,0x10,0x65,0xa3,0x70,0x12,0x68,0x61,0x6d,0xa3,0xae,0x12,0x69,0x65,0x74,0xa3,
+0xb7,0x11,0x72,0x69,0xa3,0xdc,0x11,0x69,0x6c,0x48,0x12,0x73,0x75,0x70,0xa4,0x2b,
+0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa5,0x2b,0x13,0x6c,0x75,0x67,0x75,0x4b,
+0x10,0x61,1,0x61,0x24,0x69,0x53,0x11,0x6e,0x61,0x3d,2,0x62,0x34,0x66,0x3c,
+0x72,0x13,0x68,0x75,0x74,0x61,0xa3,0xfb,0x13,0x65,0x74,0x61,0x6e,0x57,0x14,0x69,
+0x6e,0x61,0x67,0x68,0xa3,0x90,2,0x63,0x82,0x67,0x92,0x6e,0x1f,0x69,0x66,0x69,
+0x65,0x64,0x63,0x61,0x6e,0x61,0x64,0x69,0x61,0x6e,0x61,0x62,0x6f,0x1f,0x72,0x69,
+0x67,0x69,0x6e,0x61,0x6c,0x73,0x79,0x6c,0x6c,0x61,0x62,0x69,0x63,0x73,0x62,0x17,
+0x65,0x78,0x74,0x65,0x6e,0x64,0x65,0x64,0xa3,0xad,0x11,0x61,0x73,0x62,0x12,0x65,
+0x78,0x74,0xa3,0xad,0x15,0x61,0x72,0x69,0x74,0x69,0x63,0xa3,0x78,0x70,0xc3,0x4b,
+0x70,0xa6,0x61,0x72,0xa8,0x1d,0x73,7,0x6f,0xc1,0xbe,0x6f,0xa2,0x69,0x70,0xa2,
+0x85,0x75,0xa2,0xa4,0x79,2,0x6c,0x50,0x6d,0x62,0x72,0x12,0x69,0x61,0x63,0x3a,
+0x12,0x73,0x75,0x70,0xa4,0x17,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa5,0x17,
+0x17,0x6f,0x74,0x69,0x6e,0x61,0x67,0x72,0x69,0xa3,0x8f,0x13,0x62,0x6f,0x6c,0x73,
+1,0x61,0x4c,0x66,0x10,0x6f,0x1f,0x72,0x6c,0x65,0x67,0x61,0x63,0x79,0x63,0x6f,
+0x6d,0x70,0x75,0x74,0x69,0x6e,0x67,0xa5,0x32,0x1f,0x6e,0x64,0x70,0x69,0x63,0x74,
+0x6f,0x67,0x72,0x61,0x70,0x68,0x73,0x65,0x78,0x74,1,0x61,0xa5,0x2a,0x65,0x14,
+0x6e,0x64,0x65,0x64,0x61,0xa5,0x2a,2,0x67,0x34,0x72,0x3e,0x79,0x13,0x6f,0x6d,
+0x62,0x6f,0xa5,0x16,0x13,0x64,0x69,0x61,0x6e,0xa5,0x23,0x17,0x61,0x73,0x6f,0x6d,
+0x70,0x65,0x6e,0x67,0xa3,0xda,1,0x61,0x32,0x65,0x14,0x63,0x69,0x61,0x6c,0x73,
+0xa3,0x56,0x12,0x63,0x69,0x6e,0x1f,0x67,0x6d,0x6f,0x64,0x69,0x66,0x69,0x65,0x72,
+0x6c,0x65,0x74,0x74,0x65,0x72,0x73,0x2d,2,0x6e,0x48,0x70,0x76,0x74,0x1d,0x74,
+0x6f,0x6e,0x73,0x69,0x67,0x6e,0x77,0x72,0x69,0x74,0x69,0x6e,0x67,0xa5,6,0x15,
+0x64,0x61,0x6e,0x65,0x73,0x65,0xa2,0x9b,0x12,0x73,0x75,0x70,0xa2,0xdb,0x16,0x70,
+0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0xdb,4,0x61,0xa2,0xa8,0x65,0x5c,0x6d,0x9e,
+0x70,0xa2,0x4b,0x73,0x13,0x79,0x6d,0x62,0x6f,0x1f,0x6c,0x73,0x61,0x6e,0x64,0x70,
+0x69,0x63,0x74,0x6f,0x67,0x72,0x61,0x70,0x68,0x73,0xa5,5,0x10,0x72,1,0x61,
+0x4e,0x73,0x12,0x63,0x72,0x69,0x1f,0x70,0x74,0x73,0x61,0x6e,0x64,0x73,0x75,0x62,
+0x73,0x63,0x72,0x69,0x70,0x74,0x73,0x73,0x14,0x6e,0x64,0x73,0x75,0x62,0x73,0x1b,
+0x61,0x74,0x68,0x6f,0x70,0x65,0x72,0x61,0x74,0x6f,0x72,0x73,0xa3,0x6a,1,0x6c,
+0x40,0x75,1,0x61,0x6e,0x6e,0x17,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xa3,
+0x8e,0x15,0x65,0x6d,0x65,0x6e,0x74,0x61,1,0x6c,0x50,0x72,0x1e,0x79,0x70,0x72,
+0x69,0x76,0x61,0x74,0x65,0x75,0x73,0x65,0x61,0x72,0x65,0x61,1,0x61,0xa3,0x6d,
+0x62,0xa3,0x6e,3,0x61,0x5c,0x6d,0x78,0x70,0xa2,0x41,0x73,0x13,0x79,0x6d,0x62,
+0x6f,0x1f,0x6c,0x73,0x61,0x6e,0x64,0x70,0x69,0x63,0x74,0x6f,0x67,0x72,0x61,0x70,
+0x68,0x73,0xa5,5,0x14,0x72,0x72,0x6f,0x77,0x73,2,0x61,0xa3,0x67,0x62,0xa3,
+0x68,0x63,0xa3,0xfa,0x13,0x61,0x74,0x68,0x65,0x1f,0x6d,0x61,0x74,0x69,0x63,0x61,
+0x6c,0x6f,0x70,0x65,0x72,0x61,0x74,0x6f,0x72,0x73,0xa3,0x6a,0x19,0x75,0x6e,0x63,
+0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xa3,0x8e,0x61,0x88,0x68,0xa2,0x48,0x69,0xa2,
+0x71,0x6d,0x12,0x61,0x6c,0x6c,1,0x66,0x46,0x6b,0x15,0x61,0x6e,0x61,0x65,0x78,
+0x74,0xa4,0x29,0x15,0x65,0x6e,0x73,0x69,0x6f,0x6e,0xa5,0x29,0x12,0x6f,0x72,0x6d,
+1,0x73,0xa3,0x54,0x76,0x16,0x61,0x72,0x69,0x61,0x6e,0x74,0x73,0xa3,0x54,1,
+0x6d,0x36,0x75,0x16,0x72,0x61,0x73,0x68,0x74,0x72,0x61,0xa3,0xa1,0x15,0x61,0x72,
+0x69,0x74,0x61,0x6e,0xa3,0xac,1,0x61,0x52,0x6f,0x13,0x72,0x74,0x68,0x61,0x1f,
+0x6e,0x64,0x66,0x6f,0x72,0x6d,0x61,0x74,0x63,0x6f,0x6e,0x74,0x72,0x6f,0x6c,0x73,
+0xa3,0xf7,1,0x72,0x2e,0x76,0x12,0x69,0x61,0x6e,0xa3,0x79,0x12,0x61,0x64,0x61,
+0xa3,0xd9,1,0x64,0x50,0x6e,0x13,0x68,0x61,0x6c,0x61,0x50,0x1d,0x61,0x72,0x63,
+0x68,0x61,0x69,0x63,0x6e,0x75,0x6d,0x62,0x65,0x72,0x73,0xa3,0xf9,0x13,0x64,0x68,
+0x61,0x6d,0xa3,0xf8,5,0x72,0x35,0x72,0x44,0x73,0x64,0x75,1,0x61,0xa3,0x4e,
+0x6e,0x17,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x71,0x17,0x69,0x76,0x61,0x74,
+0x65,0x75,0x73,0x65,0xa2,0x4e,0x13,0x61,0x72,0x65,0x61,0xa3,0x4e,0x1b,0x61,0x6c,
+0x74,0x65,0x72,0x70,0x61,0x68,0x6c,0x61,0x76,0x69,0xa3,0xf6,0x61,0x40,0x68,0x82,
+0x6c,0x19,0x61,0x79,0x69,0x6e,0x67,0x63,0x61,0x72,0x64,0x73,0xa3,0xcc,2,0x68,
+0x38,0x6c,0x4a,0x75,0x15,0x63,0x69,0x6e,0x68,0x61,0x75,0xa3,0xf5,0x17,0x61,0x77,
+0x68,0x68,0x6d,0x6f,0x6e,0x67,0xa3,0xf3,0x15,0x6d,0x79,0x72,0x65,0x6e,0x65,0xa3,
+0xf4,1,0x61,0x8e,0x6f,1,0x65,0x74,0x6e,0x16,0x65,0x74,0x69,0x63,0x65,0x78,
+0x74,0xa2,0x72,1,0x65,0x2c,0x73,0x11,0x75,0x70,0xa3,0x8d,0x15,0x6e,0x73,0x69,
+0x6f,0x6e,0x73,0xa2,0x72,0x19,0x73,0x75,0x70,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,
+0xa3,0x8d,0x15,0x6e,0x69,0x63,0x69,0x61,0x6e,0xa3,0x97,1,0x67,0x3e,0x69,0x13,
+0x73,0x74,0x6f,0x73,0xa2,0xa6,0x13,0x64,0x69,0x73,0x63,0xa3,0xa6,0x12,0x73,0x70,
+0x61,0xa3,0x96,1,0x65,0x5c,0x75,1,0x6d,0x2a,0x6e,0x11,0x69,0x63,0x67,0x10,
+0x69,0xa2,0xc0,0x1d,0x6e,0x75,0x6d,0x65,0x72,0x61,0x6c,0x73,0x79,0x6d,0x62,0x6f,
+0x6c,0x73,0xa3,0xc0,0x13,0x6a,0x61,0x6e,0x67,0xa3,0xa3,0x6d,0xa2,0xe6,0x6e,0xa8,
+0x19,0x6f,6,0x70,0x63,0x70,0x56,0x72,0x8a,0x73,0xa2,0x4c,0x74,0x10,0x74,0x1f,
+0x6f,0x6d,0x61,0x6e,0x73,0x69,0x79,0x61,0x71,0x6e,0x75,0x6d,0x62,0x65,0x72,0x73,
+0xa5,0x28,0x18,0x74,0x69,0x63,0x61,0x6c,0x63,0x68,0x61,0x72,0x1f,0x61,0x63,0x74,
+0x65,0x72,0x72,0x65,0x63,0x6f,0x67,0x6e,0x69,0x74,0x69,0x6f,0x6e,0x85,1,0x69,
+0x46,0x6e,0x1e,0x61,0x6d,0x65,0x6e,0x74,0x61,0x6c,0x64,0x69,0x6e,0x67,0x62,0x61,
+0x74,0x73,0xa3,0xf2,0x11,0x79,0x61,0x47,1,0x61,0x30,0x6d,0x13,0x61,0x6e,0x79,
+0x61,0xa3,0x7a,0x11,0x67,0x65,0xa5,0xf,0x63,0xa2,0x71,0x67,0xa2,0x71,0x6c,1,
+0x63,0xa2,0x62,0x64,5,0x70,0x38,0x70,0x36,0x73,0x56,0x74,0x14,0x75,0x72,0x6b,
+0x69,0x63,0xa3,0xbf,0x11,0x65,0x72,1,0x6d,0x2e,0x73,0x12,0x69,0x61,0x6e,0xa3,
+0x8c,0x11,0x69,0x63,0xa3,0xf1,0x10,0x6f,1,0x67,0x3a,0x75,0x18,0x74,0x68,0x61,
+0x72,0x61,0x62,0x69,0x61,0x6e,0xa3,0xbb,0x13,0x64,0x69,0x61,0x6e,0xa5,0x22,0x68,
+0x42,0x69,0x54,0x6e,0x1a,0x6f,0x72,0x74,0x68,0x61,0x72,0x61,0x62,0x69,0x61,0x6e,
+0xa3,0xf0,0x17,0x75,0x6e,0x67,0x61,0x72,0x69,0x61,0x6e,0xa5,4,0x14,0x74,0x61,
+0x6c,0x69,0x63,0xa3,0x58,0x13,0x68,0x69,0x6b,0x69,0xa3,0x9d,0x10,0x72,0x85,0x12,
+0x68,0x61,0x6d,0x65,6,0x6f,0x86,0x6f,0x6c,0x72,0xa2,0x61,0x75,0xa2,0x62,0x79,
+0x14,0x61,0x6e,0x6d,0x61,0x72,0x58,0x12,0x65,0x78,0x74,2,0x61,0xa3,0xb6,0x62,
+0xa3,0xee,0x65,0x13,0x6e,0x64,0x65,0x64,1,0x61,0xa3,0xb6,0x62,0xa3,0xee,1,
+0x64,0x52,0x6e,0x15,0x67,0x6f,0x6c,0x69,0x61,0x6e,0x6a,0x12,0x73,0x75,0x70,0xa4,
+0xd,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa5,0xd,0x10,0x69,0xa2,0xec,0x13,
+0x66,0x69,0x65,0x72,1,0x6c,0x3c,0x74,0x19,0x6f,0x6e,0x65,0x6c,0x65,0x74,0x74,
+0x65,0x72,0x73,0xa3,0x8a,0x15,0x65,0x74,0x74,0x65,0x72,0x73,0x2d,0x10,0x6f,0xa3,
+0xed,1,0x6c,0x44,0x73,0x11,0x69,0x63,0xa2,0x5c,0x18,0x61,0x6c,0x73,0x79,0x6d,
+0x62,0x6f,0x6c,0x73,0xa3,0x5c,0x13,0x74,0x61,0x6e,0x69,0xa5,3,0x61,0xa2,0x9b,
+0x65,0xa4,0x4c,0x69,1,0x61,0xa2,0x8f,0x73,0x10,0x63,5,0x70,0x18,0x70,0xa2,
+0x71,0x73,0x36,0x74,0x17,0x65,0x63,0x68,0x6e,0x69,0x63,0x61,0x6c,0x81,0x15,0x79,
+0x6d,0x62,0x6f,0x6c,0x73,0x8f,0x61,0xa2,0x66,0x65,0x46,0x6d,0x19,0x61,0x74,0x68,
+0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,1,0x61,0xa3,0x66,0x62,0xa3,0x69,0x17,0x6c,
+0x6c,0x61,0x6e,0x65,0x6f,0x75,0x73,2,0x6d,0x3a,0x73,0x6c,0x74,0x17,0x65,0x63,
+0x68,0x6e,0x69,0x63,0x61,0x6c,0x81,0x11,0x61,0x74,0x1f,0x68,0x65,0x6d,0x61,0x74,
+0x69,0x63,0x61,0x6c,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,1,0x61,0xa3,0x66,0x62,
+0xa3,0x69,0x15,0x79,0x6d,0x62,0x6f,0x6c,0x73,0x8e,0x12,0x61,0x6e,0x64,1,0x61,
+0x3c,0x70,0x19,0x69,0x63,0x74,0x6f,0x67,0x72,0x61,0x70,0x68,0x73,0xa3,0xcd,0x14,
+0x72,0x72,0x6f,0x77,0x73,0xa3,0x73,0x10,0x6f,0xa3,0xd8,7,0x72,0x6f,0x72,0x44,
+0x73,0x4e,0x74,0x62,0x79,0x19,0x61,0x6e,0x6e,0x75,0x6d,0x65,0x72,0x61,0x6c,0x73,
+0xa5,0x20,0x13,0x63,0x68,0x65,0x6e,0xa5,0xc,0x18,0x61,0x72,0x61,0x6d,0x67,0x6f,
+0x6e,0x64,0x69,0xa5,0x14,0x10,0x68,2,0x61,0x3a,0x65,0x4a,0x6f,0x17,0x70,0x65,
+0x72,0x61,0x74,0x6f,0x72,0x73,0x7f,0x16,0x6c,0x70,0x68,0x61,0x6e,0x75,0x6d,0xa3,
+0x5d,0x16,0x6d,0x61,0x74,0x69,0x63,0x61,0x6c,1,0x61,0x36,0x6f,0x17,0x70,0x65,
+0x72,0x61,0x74,0x6f,0x72,0x73,0x7f,0x11,0x6c,0x70,0x1f,0x68,0x61,0x6e,0x75,0x6d,
+0x65,0x72,0x69,0x63,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0x5d,0x68,0x50,0x6b,
+0x7e,0x6c,0x88,0x6e,1,0x64,0x34,0x69,0x15,0x63,0x68,0x61,0x65,0x61,0x6e,0xa3,
+0xea,0x12,0x61,0x69,0x63,0xa3,0xc6,1,0x61,0x3e,0x6a,0x12,0x6f,0x6e,0x67,0xa2,
+0xaa,0x14,0x74,0x69,0x6c,0x65,0x73,0xa3,0xaa,0x13,0x6a,0x61,0x6e,0x69,0xa3,0xe9,
+0x13,0x61,0x73,0x61,0x72,0xa5,0x1f,0x15,0x61,0x79,0x61,0x6c,0x61,0x6d,0x4f,3,
+0x64,0x6c,0x65,0x7e,0x6e,0xa2,0x47,0x72,0x14,0x6f,0x69,0x74,0x69,0x63,1,0x63,
+0x3c,0x68,0x19,0x69,0x65,0x72,0x6f,0x67,0x6c,0x79,0x70,0x68,0x73,0xa3,0xd7,0x15,
+0x75,0x72,0x73,0x69,0x76,0x65,0xa3,0xd6,0x17,0x65,0x66,0x61,0x69,0x64,0x72,0x69,
+0x6e,0xa5,0x21,0x17,0x74,0x65,0x69,0x6d,0x61,0x79,0x65,0x6b,0xa2,0xb8,0x12,0x65,
+0x78,0x74,0xa2,0xd5,0x16,0x65,0x6e,0x73,0x69,0x6f,0x6e,0x73,0xa3,0xd5,0x18,0x64,
+0x65,0x6b,0x69,0x6b,0x61,0x6b,0x75,0x69,0xa3,0xeb,6,0x6b,0x3b,0x6b,0x56,0x6f,
+0x5a,0x75,0x64,0x79,0x11,0x69,0x61,0x1f,0x6b,0x65,0x6e,0x67,0x70,0x75,0x61,0x63,
+0x68,0x75,0x65,0x68,0x6d,0x6f,0x6e,0x67,0xa5,0x27,0x10,0x6f,0xa3,0x92,0x14,0x62,
+0x6c,0x6f,0x63,0x6b,0x21,1,0x6d,0x2c,0x73,0x11,0x68,0x75,0xa5,0x15,0x17,0x62,
+0x65,0x72,0x66,0x6f,0x72,0x6d,0x73,0x7b,0x61,0x44,0x62,0x21,0x65,0x10,0x77,1,
+0x61,0xa5,0xe,0x74,0x14,0x61,0x69,0x6c,0x75,0x65,0xa3,0x8b,1,0x62,0x38,0x6e,
+0x17,0x64,0x69,0x6e,0x61,0x67,0x61,0x72,0x69,0xa5,0x26,0x15,0x61,0x74,0x61,0x65,
+0x61,0x6e,0xa3,0xef,0x67,0xc4,0x32,0x6a,0xc1,0xb9,0x6a,0xa2,0xd5,0x6b,0xa2,0xee,
+0x6c,4,0x61,0x54,0x65,0xa2,0x61,0x69,0xa2,0x78,0x6f,0xa2,0xb7,0x79,1,0x63,
+0x2e,0x64,0x12,0x69,0x61,0x6e,0xa3,0xa9,0x12,0x69,0x61,0x6e,0xa3,0xa7,1,0x6f,
+0x55,0x74,0x11,0x69,0x6e,1,0x31,0x82,0x65,0x11,0x78,0x74,4,0x61,0x5c,0x62,
+0x29,0x63,0xa3,0x94,0x64,0xa3,0x95,0x65,0xa2,0xe7,0x13,0x6e,0x64,0x65,0x64,4,
+0x61,0x36,0x62,0x29,0x63,0xa3,0x94,0x64,0xa3,0x95,0x65,0xa3,0xe7,0x26,0x18,0x64,
+0x64,0x69,0x74,0x69,0x6f,0x6e,0x61,0x6c,0x6d,0x24,0x12,0x73,0x75,0x70,0x24,0x16,
+0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0x25,1,0x70,0x42,0x74,0x1d,0x74,0x65,0x72,
+0x6c,0x69,0x6b,0x65,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0x79,0x12,0x63,0x68,0x61,
+0xa3,0x9c,2,0x6d,0x4e,0x6e,0x54,0x73,0x10,0x75,0xa2,0xb0,0x12,0x73,0x75,0x70,
+0xa4,0x31,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa5,0x31,0x11,0x62,0x75,0xa3,
+0x6f,0x12,0x65,0x61,0x72,1,0x61,0xa3,0xe8,0x62,1,0x69,0x38,0x73,0x17,0x79,
+0x6c,0x6c,0x61,0x62,0x61,0x72,0x79,0xa3,0x75,0x17,0x64,0x65,0x6f,0x67,0x72,0x61,
+0x6d,0x73,0xa3,0x76,0x1a,0x77,0x73,0x75,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,0x73,
+0xa3,0x4d,0x10,0x61,1,0x6d,0x32,0x76,0x14,0x61,0x6e,0x65,0x73,0x65,0xa3,0xb5,
+0x10,0x6f,0x5c,0x12,0x65,0x78,0x74,1,0x61,0xa3,0xb4,0x62,0xa3,0xb9,1,0x61,
+0xa2,0x43,0x68,4,0x61,0x40,0x69,0x50,0x6d,0x6e,0x6f,0x86,0x75,0x15,0x64,0x61,
+0x77,0x61,0x64,0x69,0xa3,0xe6,0x16,0x72,0x6f,0x73,0x68,0x74,0x68,0x69,0xa3,0x89,
+0x1d,0x74,0x61,0x6e,0x73,0x6d,0x61,0x6c,0x6c,0x73,0x63,0x72,0x69,0x70,0x74,0xa5,
+0x30,0x11,0x65,0x72,0x68,0x16,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0x71,0x12,
+0x6a,0x6b,0x69,0xa3,0xe5,3,0x69,0x3a,0x6e,0x42,0x74,0xa2,0x51,0x79,0x13,0x61,
+0x68,0x6c,0x69,0xa3,0xa2,0x12,0x74,0x68,0x69,0xa3,0xc1,3,0x61,0x34,0x62,0x76,
+0x67,0x7c,0x6e,0x12,0x61,0x64,0x61,0x4d,1,0x65,0x40,0x73,0x11,0x75,0x70,0xa2,
+0xcb,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0xcb,0x11,0x78,0x74,1,0x61,
+0xa5,0x13,0x65,0x14,0x6e,0x64,0x65,0x64,0x61,0xa5,0x13,0x11,0x75,0x6e,0xa3,0x42,
+0x11,0x78,0x69,0x96,0x17,0x72,0x61,0x64,0x69,0x63,0x61,0x6c,0x73,0x97,0x14,0x61,
+0x6b,0x61,0x6e,0x61,0x9e,1,0x65,0x4c,0x70,0x10,0x68,0x1f,0x6f,0x6e,0x65,0x74,
+0x69,0x63,0x65,0x78,0x74,0x65,0x6e,0x73,0x69,0x6f,0x6e,0x73,0xa3,0x6b,0x11,0x78,
+0x74,0xa3,0x6b,0x67,0xa2,0xb5,0x68,0xa4,0x84,0x69,3,0x64,0x4c,0x6d,0xa2,0x55,
+0x6e,0xa2,0x62,0x70,0x13,0x61,0x65,0x78,0x74,0x2a,0x16,0x65,0x6e,0x73,0x69,0x6f,
+0x6e,0x73,0x2b,1,0x63,0x99,0x65,0x17,0x6f,0x67,0x72,0x61,0x70,0x68,0x69,0x63,
+1,0x64,0x56,0x73,0x15,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa4,0xb,0x1d,0x61,0x6e,
+0x64,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xa5,0xb,0x13,0x65,
+0x73,0x63,0x72,0x1f,0x69,0x70,0x74,0x69,0x6f,0x6e,0x63,0x68,0x61,0x72,0x61,0x63,
+0x74,0x65,0x72,0x73,0x99,0x1c,0x70,0x65,0x72,0x69,0x61,0x6c,0x61,0x72,0x61,0x6d,
+0x61,0x69,0x63,0xa3,0xba,1,0x64,0x62,0x73,0x1b,0x63,0x72,0x69,0x70,0x74,0x69,
+0x6f,0x6e,0x61,0x6c,0x70,0x61,1,0x68,0x32,0x72,0x14,0x74,0x68,0x69,0x61,0x6e,
+0xa3,0xbd,0x13,0x6c,0x61,0x76,0x69,0xa3,0xbe,0x11,0x69,0x63,1,0x6e,0x3e,0x73,
+0x1a,0x69,0x79,0x61,0x71,0x6e,0x75,0x6d,0x62,0x65,0x72,0x73,0xa5,0x1e,0x19,0x75,
+0x6d,0x62,0x65,0x72,0x66,0x6f,0x72,0x6d,0x73,0xa3,0xb2,4,0x65,0x74,0x6c,0xa2,
+0x82,0x6f,0xa2,0x9a,0x72,0xa2,0x9e,0x75,2,0x6a,0x34,0x6e,0x3e,0x72,0x14,0x6d,
+0x75,0x6b,0x68,0x69,0x43,0x14,0x61,0x72,0x61,0x74,0x69,0x45,0x18,0x6a,0x61,0x6c,
+0x61,0x67,0x6f,0x6e,0x64,0x69,0xa5,0x1c,1,0x6e,0xa2,0x46,0x6f,1,0x6d,0x6e,
+0x72,0x13,0x67,0x69,0x61,0x6e,0x5a,1,0x65,0x40,0x73,0x11,0x75,0x70,0xa2,0x87,
+0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0x87,0x11,0x78,0x74,0xa4,0x1b,0x14,
+0x65,0x6e,0x64,0x65,0x64,0xa5,0x1b,0x1a,0x65,0x74,0x72,0x69,0x63,0x73,0x68,0x61,
+0x70,0x65,0x73,0x8c,0x12,0x65,0x78,0x74,0xa2,0xe3,0x14,0x65,0x6e,0x64,0x65,0x64,
+0xa3,0xe3,0x1e,0x65,0x72,0x61,0x6c,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,
+0x6f,0x6e,0x71,0x17,0x61,0x67,0x6f,0x6c,0x69,0x74,0x69,0x63,0xa2,0x88,0x12,0x73,
+0x75,0x70,0xa4,0xa,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa5,0xa,0x13,0x74,
+0x68,0x69,0x63,0xa3,0x59,1,0x61,0x5c,0x65,0x11,0x65,0x6b,0x30,1,0x61,0x38,
+0x65,0x11,0x78,0x74,0x6e,0x14,0x65,0x6e,0x64,0x65,0x64,0x6f,0x17,0x6e,0x64,0x63,
+0x6f,0x70,0x74,0x69,0x63,0x31,0x13,0x6e,0x74,0x68,0x61,0xa3,0xe4,2,0x61,0xa2,
+0x48,0x65,0xa2,0xdf,0x69,1,0x67,0x30,0x72,0x14,0x61,0x67,0x61,0x6e,0x61,0x9d,
+0x10,0x68,1,0x70,0x3a,0x73,0x18,0x75,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,0x73,
+0xa3,0x4b,1,0x72,0x3c,0x75,0x19,0x73,0x75,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,
+0x73,0xa3,0x4c,0x11,0x69,0x76,0x1f,0x61,0x74,0x65,0x75,0x73,0x65,0x73,0x75,0x72,
+0x72,0x6f,0x67,0x61,0x74,0x65,0x73,0xa3,0x4c,2,0x6c,0x32,0x6e,0x9a,0x74,0x12,
+0x72,0x61,0x6e,0xa5,2,0x10,0x66,2,0x61,0x58,0x6d,0x70,0x77,0x14,0x69,0x64,
+0x74,0x68,0x61,0x1f,0x6e,0x64,0x66,0x75,0x6c,0x6c,0x77,0x69,0x64,0x74,0x68,0x66,
+0x6f,0x72,0x6d,0x73,0xa3,0x57,0x1a,0x6e,0x64,0x66,0x75,0x6c,0x6c,0x66,0x6f,0x72,
+0x6d,0x73,0xa3,0x57,0x13,0x61,0x72,0x6b,0x73,0xa3,0x52,2,0x67,0x34,0x69,0xa2,
+0x45,0x75,0x12,0x6e,0x6f,0x6f,0xa3,0x63,0x11,0x75,0x6c,0xa2,0x4a,2,0x63,0x3c,
+0x6a,0x5e,0x73,0x17,0x79,0x6c,0x6c,0x61,0x62,0x6c,0x65,0x73,0xa3,0x4a,0x1f,0x6f,
+0x6d,0x70,0x61,0x74,0x69,0x62,0x69,0x6c,0x69,0x74,0x79,0x6a,0x61,0x6d,0x6f,0xa3,
+0x41,0x12,0x61,0x6d,0x6f,0x5c,0x17,0x65,0x78,0x74,0x65,0x6e,0x64,0x65,0x64,1,
+0x61,0xa3,0xb4,0x62,0xa3,0xb9,0x19,0x66,0x69,0x72,0x6f,0x68,0x69,0x6e,0x67,0x79,
+0x61,0xa5,0x1d,0x13,0x62,0x72,0x65,0x77,0x37,0x61,0xa4,5,0x62,0xa6,0x45,0x63,
+0xa8,0x1a,0x64,0xac,0xb8,0x65,5,0x6d,0xa2,0x6d,0x86,0x6e,0x96,0x74,0x15,0x68,
+0x69,0x6f,0x70,0x69,0x63,0x5e,1,0x65,0x40,0x73,0x11,0x75,0x70,0xa2,0x86,0x16,
+0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0x86,0x11,0x78,0x74,0xa2,0x85,1,0x61,
+0xa3,0xc8,0x65,0x13,0x6e,0x64,0x65,0x64,0xa2,0x85,0x10,0x61,0xa3,0xc8,0x16,0x6f,
+0x74,0x69,0x63,0x6f,0x6e,0x73,0xa3,0xce,0x15,0x63,0x6c,0x6f,0x73,0x65,0x64,2,
+0x61,0x5a,0x63,0x9e,0x69,0x1c,0x64,0x65,0x6f,0x67,0x72,0x61,0x70,0x68,0x69,0x63,
+0x73,0x75,0x70,0xa2,0xc4,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0xc4,0x16,
+0x6c,0x70,0x68,0x61,0x6e,0x75,0x6d,0x86,1,0x65,0x2c,0x73,0x11,0x75,0x70,0xa3,
+0xc3,0x13,0x72,0x69,0x63,0x73,0x86,0x18,0x75,0x70,0x70,0x6c,0x65,0x6d,0x65,0x6e,
+0x74,0xa3,0xc3,0x11,0x6a,0x6b,0xa2,0x44,0x1f,0x6c,0x65,0x74,0x74,0x65,0x72,0x73,
+0x61,0x6e,0x64,0x6d,0x6f,0x6e,0x74,0x68,0x73,0xa3,0x44,0x61,0x4a,0x67,0x76,0x6c,
+1,0x62,0x30,0x79,0x13,0x6d,0x61,0x69,0x63,0xa5,0x25,0x13,0x61,0x73,0x61,0x6e,
+0xa3,0xe2,0x13,0x72,0x6c,0x79,0x64,0x1f,0x79,0x6e,0x61,0x73,0x74,0x69,0x63,0x63,
+0x75,0x6e,0x65,0x69,0x66,0x6f,0x72,0x6d,0xa5,1,0x1f,0x79,0x70,0x74,0x69,0x61,
+0x6e,0x68,0x69,0x65,0x72,0x6f,0x67,0x6c,0x79,0x70,0x68,1,0x66,0x26,0x73,0xa3,
+0xc2,0x1c,0x6f,0x72,0x6d,0x61,0x74,0x63,0x6f,0x6e,0x74,0x72,0x6f,0x6c,0x73,0xa5,
+0x24,7,0x6e,0xc0,0xe5,0x6e,0x3e,0x72,0xa2,0x5d,0x73,0xa2,0xd8,0x76,0x14,0x65,
+0x73,0x74,0x61,0x6e,0xa3,0xbc,1,0x61,0x92,0x63,0x13,0x69,0x65,0x6e,0x74,1,
+0x67,0x34,0x73,0x15,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0xa5,0x13,0x72,0x65,0x65,
+0x6b,1,0x6d,0x34,0x6e,0x15,0x75,0x6d,0x62,0x65,0x72,0x73,0xa3,0x7f,0x13,0x75,
+0x73,0x69,0x63,0xa2,0x7e,0x19,0x61,0x6c,0x6e,0x6f,0x74,0x61,0x74,0x69,0x6f,0x6e,
+0xa3,0x7e,0x10,0x74,0x1f,0x6f,0x6c,0x69,0x61,0x6e,0x68,0x69,0x65,0x72,0x6f,0x67,
+0x6c,0x79,0x70,0x68,0x73,0xa3,0xfe,2,0x61,0x32,0x6d,0xa2,0x71,0x72,0x12,0x6f,
+0x77,0x73,0x7d,0x12,0x62,0x69,0x63,0x38,3,0x65,0x4a,0x6d,0x66,0x70,0xa2,0x43,
+0x73,0x11,0x75,0x70,0xa2,0x80,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0x80,
+0x11,0x78,0x74,1,0x61,0xa3,0xd2,0x65,0x14,0x6e,0x64,0x65,0x64,0x61,0xa3,0xd2,
+0x12,0x61,0x74,0x68,0xa2,0xd3,0x18,0x65,0x6d,0x61,0x74,0x69,0x63,0x61,0x6c,0x61,
+0x1f,0x6c,0x70,0x68,0x61,0x62,0x65,0x74,0x69,0x63,0x73,0x79,0x6d,0x62,0x6f,0x6c,
+0x73,0xa3,0xd3,1,0x66,0x42,0x72,0x1e,0x65,0x73,0x65,0x6e,0x74,0x61,0x74,0x69,
+0x6f,0x6e,0x66,0x6f,0x72,0x6d,0x73,1,0x61,0xa3,0x51,0x62,0xa3,0x55,0x14,0x65,
+0x6e,0x69,0x61,0x6e,0x35,0x12,0x63,0x69,0x69,0x23,0x64,0x9e,0x65,0xa2,0x42,0x68,
+0xa2,0x4d,0x6c,1,0x63,0x62,0x70,0x17,0x68,0x61,0x62,0x65,0x74,0x69,0x63,0x70,
+1,0x66,0xa3,0x50,0x72,0x1e,0x65,0x73,0x65,0x6e,0x74,0x61,0x74,0x69,0x6f,0x6e,
+0x66,0x6f,0x72,0x6d,0x73,0xa3,0x50,0x16,0x68,0x65,0x6d,0x69,0x63,0x61,0x6c,0xa2,
+0xd0,0x16,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0xd0,0x12,0x6c,0x61,0x6d,0xa5,
+7,0x1a,0x67,0x65,0x61,0x6e,0x6e,0x75,0x6d,0x62,0x65,0x72,0x73,0xa3,0x77,0x11,
+0x6f,0x6d,0xa3,0xfd,7,0x6f,0x71,0x6f,0x64,0x72,0xa2,0x41,0x75,0xa2,0x58,0x79,
+0x1b,0x7a,0x61,0x6e,0x74,0x69,0x6e,0x65,0x6d,0x75,0x73,0x69,0x63,0xa2,0x5b,0x18,
+0x61,0x6c,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0xa3,0x5b,1,0x70,0x34,0x78,0x16,
+0x64,0x72,0x61,0x77,0x69,0x6e,0x67,0x89,0x14,0x6f,0x6d,0x6f,0x66,0x6f,0xa0,0x12,
+0x65,0x78,0x74,0xa2,0x43,0x14,0x65,0x6e,0x64,0x65,0x64,0xa3,0x43,0x10,0x61,1,
+0x68,0x40,0x69,0x12,0x6c,0x6c,0x65,0x92,0x17,0x70,0x61,0x74,0x74,0x65,0x72,0x6e,
+0x73,0x93,0x11,0x6d,0x69,0xa3,0xc9,1,0x67,0x2c,0x68,0x11,0x69,0x64,0xa3,0x64,
+0x14,0x69,0x6e,0x65,0x73,0x65,0xa3,0x81,0x61,0x48,0x65,0xa2,0x4e,0x68,0xa2,0x52,
+0x6c,0x1a,0x6f,0x63,0x6b,0x65,0x6c,0x65,0x6d,0x65,0x6e,0x74,0x73,0x8b,3,0x6c,
+0x34,0x6d,0x40,0x73,0x66,0x74,0x11,0x61,0x6b,0xa3,0xc7,0x14,0x69,0x6e,0x65,0x73,
+0x65,0xa3,0x93,0x11,0x75,0x6d,0xa2,0xb1,0x12,0x73,0x75,0x70,0xa2,0xca,0x16,0x70,
+0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0xca,1,0x69,0x30,0x73,0x13,0x61,0x76,0x61,
+0x68,0xa3,0xdd,0x15,0x63,0x6c,0x61,0x74,0x69,0x6e,0x23,0x14,0x6e,0x67,0x61,0x6c,
+0x69,0x41,0x16,0x61,0x69,0x6b,0x73,0x75,0x6b,0x69,0xa5,8,5,0x6f,0xc1,0x4c,
+0x6f,0xa2,0x55,0x75,0xa4,0x10,0x79,1,0x70,0x9c,0x72,0x14,0x69,0x6c,0x6c,0x69,
+0x63,0x32,1,0x65,0x4c,0x73,0x11,0x75,0x70,0xa2,0x61,0x16,0x70,0x6c,0x65,0x6d,
+0x65,0x6e,0x74,0xa2,0x61,0x12,0x61,0x72,0x79,0xa3,0x61,0x11,0x78,0x74,3,0x61,
+0xa3,0x9e,0x62,0xa3,0xa0,0x63,0xa5,9,0x65,0x13,0x6e,0x64,0x65,0x64,2,0x61,
+0xa3,0x9e,0x62,0xa3,0xa0,0x63,0xa5,9,0x1c,0x72,0x69,0x6f,0x74,0x73,0x79,0x6c,
+0x6c,0x61,0x62,0x61,0x72,0x79,0xa3,0x7b,3,0x6d,0x5a,0x6e,0xa2,0x95,0x70,0xa2,
+0xa0,0x75,0x17,0x6e,0x74,0x69,0x6e,0x67,0x72,0x6f,0x64,0xa2,0x9a,0x17,0x6e,0x75,
+0x6d,0x65,0x72,0x61,0x6c,0x73,0xa3,0x9a,2,0x62,0x3a,0x6d,0xa2,0x5f,0x70,0x15,
+0x61,0x74,0x6a,0x61,0x6d,0x6f,0xa3,0x41,0x14,0x69,0x6e,0x69,0x6e,0x67,2,0x64,
+0x46,0x68,0x9e,0x6d,0x1d,0x61,0x72,0x6b,0x73,0x66,0x6f,0x72,0x73,0x79,0x6d,0x62,
+0x6f,0x6c,0x73,0x77,0x1e,0x69,0x61,0x63,0x72,0x69,0x74,0x69,0x63,0x61,0x6c,0x6d,
+0x61,0x72,0x6b,0x73,0x2e,2,0x65,0x40,0x66,0xa6,0x41,0x73,0x18,0x75,0x70,0x70,
+0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0x83,0x16,0x78,0x74,0x65,0x6e,0x64,0x65,0x64,
+0xa3,0xe0,0x17,0x61,0x6c,0x66,0x6d,0x61,0x72,0x6b,0x73,0xa3,0x52,0x11,0x6f,0x6e,
+0x1f,0x69,0x6e,0x64,0x69,0x63,0x6e,0x75,0x6d,0x62,0x65,0x72,0x66,0x6f,0x72,0x6d,
+0x73,0xa3,0xb2,0x1b,0x74,0x72,0x6f,0x6c,0x70,0x69,0x63,0x74,0x75,0x72,0x65,0x73,
+0x83,0x12,0x74,0x69,0x63,0xa2,0x84,0x1b,0x65,0x70,0x61,0x63,0x74,0x6e,0x75,0x6d,
+0x62,0x65,0x72,0x73,0xa3,0xdf,1,0x6e,0x3e,0x72,0x1b,0x72,0x65,0x6e,0x63,0x79,
+0x73,0x79,0x6d,0x62,0x6f,0x6c,0x73,0x75,0x15,0x65,0x69,0x66,0x6f,0x72,0x6d,0xa2,
+0x98,0x16,0x6e,0x75,0x6d,0x62,0x65,0x72,0x73,0xa2,0x99,0x1d,0x61,0x6e,0x64,0x70,
+0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xa3,0x99,0x61,0xa2,0xe1,0x68,
+0xa4,0xb,0x6a,0x10,0x6b,0xa2,0x47,4,0x63,0x8c,0x65,0xa2,0x80,0x72,0xa2,0x98,
+0x73,0xa2,0xaa,0x75,0x1f,0x6e,0x69,0x66,0x69,0x65,0x64,0x69,0x64,0x65,0x6f,0x67,
+0x72,0x61,0x70,0x68,0x73,0xa2,0x47,0x18,0x65,0x78,0x74,0x65,0x6e,0x73,0x69,0x6f,
+0x6e,6,0x64,0x6b,0x64,0xa3,0xd1,0x65,0xa5,0,0x66,0xa5,0x12,0x67,0xa5,0x2e,
+0x14,0x6f,0x6d,0x70,0x61,0x74,0xa2,0x45,1,0x66,0x96,0x69,1,0x62,0x44,0x64,
+0x17,0x65,0x6f,0x67,0x72,0x61,0x70,0x68,0x73,0xa2,0x4f,0x12,0x73,0x75,0x70,0xa3,
+0x5f,0x14,0x69,0x6c,0x69,0x74,0x79,0xa2,0x45,1,0x66,0x54,0x69,0x18,0x64,0x65,
+0x6f,0x67,0x72,0x61,0x70,0x68,0x73,0xa2,0x4f,0x19,0x73,0x75,0x70,0x70,0x6c,0x65,
+0x6d,0x65,0x6e,0x74,0xa3,0x5f,0x13,0x6f,0x72,0x6d,0x73,0xa3,0x53,0x11,0x78,0x74,
+6,0x64,0xc,0x64,0xa3,0xd1,0x65,0xa5,0,0x66,0xa5,0x12,0x67,0xa5,0x2e,0x61,
+0xa3,0x46,0x62,0xa3,0x5e,0x63,0xa3,0xc5,0x19,0x61,0x64,0x69,0x63,0x61,0x6c,0x73,
+0x73,0x75,0x70,0x94,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0x95,1,0x74,0x50,
+0x79,0x14,0x6d,0x62,0x6f,0x6c,0x73,0x9a,0x1d,0x61,0x6e,0x64,0x70,0x75,0x6e,0x63,
+0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x9b,0x14,0x72,0x6f,0x6b,0x65,0x73,0xa3,0x82,
+2,0x6e,0x48,0x72,0x64,0x75,0x1d,0x63,0x61,0x73,0x69,0x61,0x6e,0x61,0x6c,0x62,
+0x61,0x6e,0x69,0x61,0x6e,0xa3,0xde,0x1d,0x61,0x64,0x69,0x61,0x6e,0x73,0x79,0x6c,
+0x6c,0x61,0x62,0x69,0x63,0x73,0x63,0x12,0x69,0x61,0x6e,0xa3,0xa8,2,0x61,0x3a,
+0x65,0x4c,0x6f,0x16,0x72,0x61,0x73,0x6d,0x69,0x61,0x6e,0xa5,0x2d,1,0x6b,0x26,
+0x6d,0xa3,0xa4,0x11,0x6d,0x61,0xa3,0xd4,1,0x72,0x38,0x73,0x17,0x73,0x73,0x79,
+0x6d,0x62,0x6f,0x6c,0x73,0xa5,0x19,0x13,0x6f,0x6b,0x65,0x65,0x60,0x12,0x73,0x75,
+0x70,0xa2,0xff,0x16,0x70,0x6c,0x65,0x6d,0x65,0x6e,0x74,0xa3,0xff,3,0x65,0x3e,
+0x69,0x7e,0x6f,0xa2,0x69,0x75,0x15,0x70,0x6c,0x6f,0x79,0x61,0x6e,0xa3,0xe1,1,
+0x73,0x50,0x76,0x16,0x61,0x6e,0x61,0x67,0x61,0x72,0x69,0x3e,0x12,0x65,0x78,0x74,
+0xa2,0xb3,0x14,0x65,0x6e,0x64,0x65,0x64,0xa3,0xb3,0x13,0x65,0x72,0x65,0x74,0xa3,
+0x5a,2,0x61,0x3a,0x6e,0x82,0x76,0x16,0x65,0x73,0x61,0x6b,0x75,0x72,0x75,0xa5,
+0x2f,0x18,0x63,0x72,0x69,0x74,0x69,0x63,0x61,0x6c,0x73,0x2e,2,0x65,0x30,0x66,
+0x36,0x73,0x11,0x75,0x70,0xa3,0x83,0x11,0x78,0x74,0xa3,0xe0,0x18,0x6f,0x72,0x73,
+0x79,0x6d,0x62,0x6f,0x6c,0x73,0x77,0x14,0x67,0x62,0x61,0x74,0x73,0x91,1,0x67,
+0x3e,0x6d,0x12,0x69,0x6e,0x6f,0xa2,0xab,0x14,0x74,0x69,0x6c,0x65,0x73,0xa3,0xab,
+0x11,0x72,0x61,0xa5,0x1a,8,0x6d,0x5f,0x6d,0x3a,0x6e,0x48,0x73,0x7a,0x76,0xa2,
+0x4b,0x77,0x12,0x69,0x64,0x65,0x43,0x11,0x65,0x64,0x32,0x12,0x69,0x61,0x6c,0x33,
+2,0x61,0x40,0x62,0x37,0x6f,1,0x62,0x28,0x6e,0x10,0x65,0x21,0x13,0x72,0x65,
+0x61,0x6b,0x37,0x10,0x72,0x34,0x12,0x72,0x6f,0x77,0x35,2,0x6d,0x38,0x71,0x46,
+0x75,1,0x62,0x3d,0x70,0x3e,0x11,0x65,0x72,0x3f,1,0x61,0x24,0x6c,0x39,0x11,
+0x6c,0x6c,0x39,1,0x72,0x3b,0x75,0x12,0x61,0x72,0x65,0x3b,0x12,0x65,0x72,0x74,
+0x40,0x13,0x69,0x63,0x61,0x6c,0x41,0x63,0x58,0x65,0x92,0x66,0x96,0x69,1,0x6e,
+0x36,0x73,0x10,0x6f,0x30,0x14,0x6c,0x61,0x74,0x65,0x64,0x31,0x11,0x69,0x74,0x2e,
+0x12,0x69,0x61,0x6c,0x2f,2,0x61,0x36,0x69,0x48,0x6f,0x10,0x6d,0x24,0x12,0x70,
+0x61,0x74,0x25,0x10,0x6e,0x22,0x15,0x6f,0x6e,0x69,0x63,0x61,0x6c,0x23,0x13,0x72,
+0x63,0x6c,0x65,0x27,0x11,0x6e,0x63,0x27,2,0x69,0x3a,0x6f,0x44,0x72,0x10,0x61,
+0x2c,0x14,0x63,0x74,0x69,0x6f,0x6e,0x2d,0x10,0x6e,0x28,0x11,0x61,0x6c,0x29,0x11,
+0x6e,0x74,0x2b,4,0x61,0x3a,0x66,0x4c,0x68,0x5e,0x6e,0x70,0x77,0x2a,0x12,0x69,
+0x64,0x65,0x2b,0x22,0x17,0x6d,0x62,0x69,0x67,0x75,0x6f,0x75,0x73,0x23,0x26,0x17,
+0x75,0x6c,0x6c,0x77,0x69,0x64,0x74,0x68,0x27,0x24,0x17,0x61,0x6c,0x66,0x77,0x69,
+0x64,0x74,0x68,0x25,0x20,1,0x61,0x30,0x65,0x14,0x75,0x74,0x72,0x61,0x6c,0x21,
+0x28,0x13,0x72,0x72,0x6f,0x77,0x29,0xd,0x6e,0xc0,0xfb,0x73,0x6d,0x73,0x3a,0x74,
+0x98,0x75,0xa2,0x49,0x7a,2,0x6c,0x3b,0x70,0x3d,0x73,0x39,5,0x6f,0x28,0x6f,
+0x57,0x70,0x34,0x75,0x16,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,0x45,0x11,0x61,0x63,
+1,0x65,0x32,0x69,0x15,0x6e,0x67,0x6d,0x61,0x72,0x6b,0x31,0x18,0x73,0x65,0x70,
+0x61,0x72,0x61,0x74,0x6f,0x72,0x39,0x63,0x53,0x6b,0x55,0x6d,0x51,0x1d,0x69,0x74,
+0x6c,0x65,0x63,0x61,0x73,0x65,0x6c,0x65,0x74,0x74,0x65,0x72,0x27,1,0x6e,0x40,
+0x70,0x1c,0x70,0x65,0x72,0x63,0x61,0x73,0x65,0x6c,0x65,0x74,0x74,0x65,0x72,0x23,
+0x17,0x61,0x73,0x73,0x69,0x67,0x6e,0x65,0x64,0x21,0x6e,0x8a,0x6f,0xa2,0x47,0x70,
+8,0x66,0x14,0x66,0x5b,0x69,0x59,0x6f,0x4f,0x72,0x24,0x73,0x49,0x17,0x69,0x76,
+0x61,0x74,0x65,0x75,0x73,0x65,0x43,0x61,0x2c,0x63,0x4d,0x64,0x47,0x65,0x4b,0x1f,
+0x72,0x61,0x67,0x72,0x61,0x70,0x68,0x73,0x65,0x70,0x61,0x72,0x61,0x74,0x6f,0x72,
+0x3d,2,0x64,0x33,0x6c,0x35,0x6f,0x36,0x1b,0x6e,0x73,0x70,0x61,0x63,0x69,0x6e,
+0x67,0x6d,0x61,0x72,0x6b,0x2d,1,0x70,0x7c,0x74,0x12,0x68,0x65,0x72,3,0x6c,
+0x38,0x6e,0x42,0x70,0x4c,0x73,0x14,0x79,0x6d,0x62,0x6f,0x6c,0x57,0x14,0x65,0x74,
+0x74,0x65,0x72,0x2b,0x14,0x75,0x6d,0x62,0x65,0x72,0x37,0x19,0x75,0x6e,0x63,0x74,
+0x75,0x61,0x74,0x69,0x6f,0x6e,0x4f,0x1c,0x65,0x6e,0x70,0x75,0x6e,0x63,0x74,0x75,
+0x61,0x74,0x69,0x6f,0x6e,0x49,0x66,0x9e,0x66,0x88,0x69,0xa2,0x4b,0x6c,0xa2,0x5c,
+0x6d,4,0x61,0x60,0x63,0x31,0x65,0x2f,0x6e,0x2d,0x6f,0x15,0x64,0x69,0x66,0x69,
+0x65,0x72,1,0x6c,0x30,0x73,0x14,0x79,0x6d,0x62,0x6f,0x6c,0x55,0x14,0x65,0x74,
+0x74,0x65,0x72,0x29,0x17,0x74,0x68,0x73,0x79,0x6d,0x62,0x6f,0x6c,0x51,1,0x69,
+0x2e,0x6f,0x13,0x72,0x6d,0x61,0x74,0x41,0x1d,0x6e,0x61,0x6c,0x70,0x75,0x6e,0x63,
+0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x5b,0x10,0x6e,0x1f,0x69,0x74,0x69,0x61,0x6c,
+0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x59,6,0x6d,0x18,0x6d,
+0x29,0x6f,0x28,0x74,0x27,0x75,0x23,0x2a,0x1c,0x77,0x65,0x72,0x63,0x61,0x73,0x65,
+0x6c,0x65,0x74,0x74,0x65,0x72,0x25,0x65,0x28,0x69,0x3c,0x6c,0x25,0x19,0x74,0x74,
+0x65,0x72,0x6e,0x75,0x6d,0x62,0x65,0x72,0x35,0x1a,0x6e,0x65,0x73,0x65,0x70,0x61,
+0x72,0x61,0x74,0x6f,0x72,0x3b,0x63,0x44,0x64,0xa2,0x60,0x65,0x1b,0x6e,0x63,0x6c,
+0x6f,0x73,0x69,0x6e,0x67,0x6d,0x61,0x72,0x6b,0x2f,6,0x6e,0x39,0x6e,0x46,0x6f,
+0x4e,0x73,0x45,0x75,0x1b,0x72,0x72,0x65,0x6e,0x63,0x79,0x73,0x79,0x6d,0x62,0x6f,
+0x6c,0x53,0x20,0x12,0x74,0x72,0x6c,0x3f,0x42,0x10,0x6e,1,0x6e,0x2c,0x74,0x12,
+0x72,0x6f,0x6c,0x3f,0x1f,0x65,0x63,0x74,0x6f,0x72,0x70,0x75,0x6e,0x63,0x74,0x75,
+0x61,0x74,0x69,0x6f,0x6e,0x4d,0x63,0x3f,0x66,0x41,0x6c,0x1d,0x6f,0x73,0x65,0x70,
+0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x4b,2,0x61,0x30,0x65,0x4a,
+0x69,0x12,0x67,0x69,0x74,0x33,0x1c,0x73,0x68,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,
+0x74,0x69,0x6f,0x6e,0x47,0x1a,0x63,0x69,0x6d,0x61,0x6c,0x6e,0x75,0x6d,0x62,0x65,
+0x72,0x33,0,0x12,0x6d,0xc2,0x3f,0x73,0xa1,0x73,0x4e,0x74,0xa2,0x56,0x77,0xa2,
+0x72,0x79,0xa2,0x73,0x7a,1,0x61,0x2c,0x68,0x12,0x61,0x69,0x6e,0x8b,0x11,0x69,
+0x6e,0x85,5,0x74,0x22,0x74,0x38,0x77,0x4c,0x79,0x16,0x72,0x69,0x61,0x63,0x77,
+0x61,0x77,0x6f,0x18,0x72,0x61,0x69,0x67,0x68,0x74,0x77,0x61,0x77,0xa3,0x55,0x15,
+0x61,0x73,0x68,0x6b,0x61,0x66,0x6d,0x61,0x2e,0x65,0x38,0x68,0x11,0x69,0x6e,0x6b,
+0x10,0x64,0x62,0x11,0x68,0x65,0x65,1,0x65,0x2e,0x6d,0x13,0x6b,0x61,0x74,0x68,
+0x69,0x10,0x6e,0x67,1,0x61,0x4e,0x65,1,0x68,0x28,0x74,0x10,0x68,0x77,0x16,
+0x6d,0x61,0x72,0x62,0x75,0x74,0x61,0x74,0x13,0x67,0x6f,0x61,0x6c,0x3d,1,0x68,
+0x71,0x77,0x73,0x11,0x61,0x77,0x79,1,0x65,0x32,0x75,0x11,0x64,0x68,0x80,0x11,
+0x68,0x65,0x83,0x10,0x68,0x7a,1,0x62,0x34,0x77,0x16,0x69,0x74,0x68,0x74,0x61,
+0x69,0x6c,0x7f,0x14,0x61,0x72,0x72,0x65,0x65,0x7d,0x6d,0x6c,0x6e,0xa4,0x6b,0x70,
+0xa4,0x88,0x71,0xa4,0x88,0x72,1,0x65,0x38,0x6f,0x18,0x68,0x69,0x6e,0x67,0x79,
+0x61,0x79,0x65,0x68,0x93,1,0x68,0x5f,0x76,0x16,0x65,0x72,0x73,0x65,0x64,0x70,
+0x65,0x61,2,0x61,0x2e,0x65,0xa4,0x3e,0x69,0x10,0x6d,0x53,1,0x6c,0xa2,0xe7,
+0x6e,0x16,0x69,0x63,0x68,0x61,0x65,0x61,0x6e,0,0x12,0x6e,0x76,0x73,0x51,0x73,
+0x3e,0x74,0x5c,0x77,0xa0,0x79,0xa2,0x42,0x7a,0x13,0x61,0x79,0x69,0x6e,0xa3,0x54,
+0x10,0x61,1,0x64,0x2e,0x6d,0x12,0x65,0x6b,0x68,0xa3,0x4c,0x11,0x68,0x65,0xa3,
+0x4b,3,0x61,0x38,0x65,0x3c,0x68,0x4a,0x77,0x13,0x65,0x6e,0x74,0x79,0xa3,0x51,
+0x10,0x77,0xa3,0x4d,1,0x6e,0xa3,0x4e,0x74,0x10,0x68,0xa3,0x4f,0x14,0x61,0x6d,
+0x65,0x64,0x68,0xa3,0x50,0x11,0x61,0x77,0xa3,0x52,0x12,0x6f,0x64,0x68,0xa3,0x53,
+0x6e,0x3a,0x6f,0x40,0x70,0x46,0x71,0x4a,0x72,0x12,0x65,0x73,0x68,0xa3,0x4a,0x11,
+0x75,0x6e,0xa3,0x46,0x11,0x6e,0x65,0xa3,0x47,0x10,0x65,0xa3,0x48,0x12,0x6f,0x70,
+0x68,0xa3,0x49,0x67,0x33,0x67,0x38,0x68,0x40,0x6b,0x5e,0x6c,0x66,0x6d,0x11,0x65,
+0x6d,0xa3,0x45,0x13,0x69,0x6d,0x65,0x6c,0xa1,1,0x65,0x32,0x75,0x14,0x6e,0x64,
+0x72,0x65,0x64,0xa3,0x42,0x11,0x74,0x68,0xa3,0x41,0x12,0x61,0x70,0x68,0xa3,0x43,
+0x14,0x61,0x6d,0x65,0x64,0x68,0xa3,0x44,0x61,0x34,0x62,0x4a,0x64,0x50,0x66,0x12,
+0x69,0x76,0x65,0x9f,1,0x6c,0x2a,0x79,0x11,0x69,0x6e,0x97,0x12,0x65,0x70,0x68,
+0x95,0x12,0x65,0x74,0x68,0x99,1,0x61,0x30,0x68,0x14,0x61,0x6d,0x65,0x64,0x68,
+0x9d,0x13,0x6c,0x65,0x74,0x68,0x9b,0x15,0x61,0x79,0x61,0x6c,0x61,0x6d,6,0x6e,
+0x2c,0x6e,0x34,0x72,0x5e,0x73,0x62,0x74,0x11,0x74,0x61,0xa3,0x63,2,0x67,0x2e,
+0x6e,0x32,0x79,0x10,0x61,0xa3,0x60,0x10,0x61,0xa3,0x5d,1,0x61,0xa3,0x5e,0x6e,
+0x10,0x61,0xa3,0x5f,0x10,0x61,0xa3,0x61,0x11,0x73,0x61,0xa3,0x62,0x62,0x3c,0x6a,
+0x42,0x6c,0x10,0x6c,1,0x61,0xa3,0x5b,0x6c,0x10,0x61,0xa3,0x5c,0x11,0x68,0x61,
+0xa3,0x59,0x10,0x61,0xa3,0x5a,0x11,0x65,0x6d,0x51,2,0x6f,0x2c,0x75,0x50,0x79,
+0x10,0x61,0x91,1,0x6a,0x28,0x6f,0x10,0x6e,0x55,0x1a,0x6f,0x69,0x6e,0x69,0x6e,
+0x67,0x67,0x72,0x6f,0x75,0x70,0x21,0x10,0x6e,0x57,0x10,0x65,0x59,0x10,0x61,1,
+0x66,0x5b,0x70,0x10,0x68,0x5d,0x66,0x9a,0x66,0x42,0x67,0x7a,0x68,0x8a,0x6b,0xa2,
+0x75,0x6c,0x11,0x61,0x6d,0x4c,0x12,0x61,0x64,0x68,0x4f,2,0x61,0x3e,0x65,0x4a,
+0x69,0x19,0x6e,0x61,0x6c,0x73,0x65,0x6d,0x6b,0x61,0x74,0x68,0x35,0x15,0x72,0x73,
+0x69,0x79,0x65,0x68,0x8f,0x86,0x10,0x68,0x33,0x10,0x61,1,0x66,0x37,0x6d,0x11,
+0x61,0x6c,0x39,1,0x61,0x40,0x65,0x3e,1,0x68,0x28,0x74,0x10,0x68,0x45,0x40,
+0x13,0x67,0x6f,0x61,0x6c,0x43,2,0x68,0x3b,0x6d,0x5c,0x6e,0x1a,0x69,0x66,0x69,
+0x72,0x6f,0x68,0x69,0x6e,0x67,0x79,0x61,1,0x6b,0x2a,0x70,0x10,0x61,0xa3,0x65,
+0x15,0x69,0x6e,0x6e,0x61,0x79,0x61,0xa3,0x64,0x1a,0x7a,0x61,0x6f,0x6e,0x68,0x65,
+0x68,0x67,0x6f,0x61,0x6c,0x3d,2,0x61,0x3a,0x68,0x44,0x6e,0x17,0x6f,0x74,0x74,
+0x65,0x64,0x68,0x65,0x68,0x4b,1,0x66,0x47,0x70,0x10,0x68,0x49,0x12,0x61,0x70,
+0x68,0x89,0x61,0x2e,0x62,0x8a,0x64,0xa2,0x51,0x65,0x31,2,0x66,0x3c,0x69,0x70,
+0x6c,1,0x61,0x28,0x65,0x10,0x66,0x27,0x11,0x70,0x68,0x25,0x14,0x72,0x69,0x63,
+0x61,0x6e,2,0x66,0x30,0x6e,0x36,0x71,0x11,0x61,0x66,0xa3,0x58,0x11,0x65,0x68,
+0xa3,0x56,0x12,0x6f,0x6f,0x6e,0xa3,0x57,0x10,0x6e,0x23,1,0x65,0x4a,0x75,0x10,
+0x72,0x1f,0x75,0x73,0x68,0x61,0x73,0x6b,0x69,0x79,0x65,0x68,0x62,0x61,0x72,0x72,
+0x65,0x65,0x8d,1,0x68,0x29,0x74,0x10,0x68,0x2b,0x11,0x61,0x6c,0x2c,0x16,0x61,
+0x74,0x68,0x72,0x69,0x73,0x68,0x2f,7,0x6e,0x2e,0x6e,0x2c,0x72,0x3e,0x74,0x56,
+0x75,0x21,0x18,0x6f,0x6e,0x6a,0x6f,0x69,0x6e,0x69,0x6e,0x67,0x21,0x28,0x1a,0x69,
+0x67,0x68,0x74,0x6a,0x6f,0x69,0x6e,0x69,0x6e,0x67,0x29,0x2a,0x19,0x72,0x61,0x6e,
+0x73,0x70,0x61,0x72,0x65,0x6e,0x74,0x2b,0x63,0x23,0x64,0x40,0x6a,0x56,0x6c,0x26,
+0x19,0x65,0x66,0x74,0x6a,0x6f,0x69,0x6e,0x69,0x6e,0x67,0x27,0x24,0x19,0x75,0x61,
+0x6c,0x6a,0x6f,0x69,0x6e,0x69,0x6e,0x67,0x25,0x19,0x6f,0x69,0x6e,0x63,0x61,0x75,
+0x73,0x69,0x6e,0x67,0x23,0,0x13,0x6e,0xc0,0xd0,0x73,0x49,0x73,0x48,0x75,0x78,
+0x77,0x84,0x78,0x9c,0x7a,0x10,0x77,0x58,1,0x6a,0x75,0x73,0x13,0x70,0x61,0x63,
+0x65,0x59,4,0x61,0x51,0x67,0x53,0x70,0x28,0x75,0x30,0x79,0x57,0x54,0x12,0x61,
+0x63,0x65,0x55,0x16,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,0x53,0x15,0x6e,0x6b,0x6e,
+0x6f,0x77,0x6e,0x21,1,0x6a,0x5d,0x6f,0x17,0x72,0x64,0x6a,0x6f,0x69,0x6e,0x65,
+0x72,0x5d,0x10,0x78,0x21,0x6e,0x60,0x6f,0xa2,0x41,0x70,0xa2,0x50,0x71,0xa2,0x6e,
+0x72,1,0x65,0x24,0x69,0x6f,0x1e,0x67,0x69,0x6f,0x6e,0x61,0x6c,0x69,0x6e,0x64,
+0x69,0x63,0x61,0x74,0x6f,0x72,0x6f,4,0x65,0x3e,0x6c,0x5b,0x6f,0x46,0x73,0x45,
+0x75,0x46,0x14,0x6d,0x65,0x72,0x69,0x63,0x47,0x15,0x78,0x74,0x6c,0x69,0x6e,0x65,
+0x5b,0x17,0x6e,0x73,0x74,0x61,0x72,0x74,0x65,0x72,0x45,0x10,0x70,0x48,0x1c,0x65,
+0x6e,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0x49,1,0x6f,0x3e,
+0x72,0x4c,0x1a,0x65,0x66,0x69,0x78,0x6e,0x75,0x6d,0x65,0x72,0x69,0x63,0x4d,0x4a,
+0x1b,0x73,0x74,0x66,0x69,0x78,0x6e,0x75,0x6d,0x65,0x72,0x69,0x63,0x4b,0x10,0x75,
+0x4e,0x16,0x6f,0x74,0x61,0x74,0x69,0x6f,0x6e,0x4f,0x68,0x7b,0x68,0x50,0x69,0x86,
+0x6a,0xa2,0x61,0x6c,0xa2,0x65,0x6d,0x1c,0x61,0x6e,0x64,0x61,0x74,0x6f,0x72,0x79,
+0x62,0x72,0x65,0x61,0x6b,0x2d,4,0x32,0x5f,0x33,0x61,0x65,0x34,0x6c,0x6d,0x79,
+0x3a,0x13,0x70,0x68,0x65,0x6e,0x3b,0x19,0x62,0x72,0x65,0x77,0x6c,0x65,0x74,0x74,
+0x65,0x72,0x6d,2,0x64,0x28,0x6e,0x3c,0x73,0x41,0x3c,0x18,0x65,0x6f,0x67,0x72,
+0x61,0x70,0x68,0x69,0x63,0x3d,0x3e,1,0x66,0x3e,0x73,0x11,0x65,0x70,1,0x61,
+0x22,0x65,0x14,0x72,0x61,0x62,0x6c,0x65,0x3f,0x18,0x69,0x78,0x6e,0x75,0x6d,0x65,
+0x72,0x69,0x63,0x41,2,0x6c,0x63,0x74,0x65,0x76,0x67,1,0x66,0x43,0x69,0x15,
+0x6e,0x65,0x66,0x65,0x65,0x64,0x43,0x61,0x40,0x62,0x70,0x63,0xa2,0x55,0x65,0xa2,
+0xdb,0x67,0x10,0x6c,0x38,0x11,0x75,0x65,0x39,2,0x69,0x23,0x6c,0x34,0x6d,0x16,
+0x62,0x69,0x67,0x75,0x6f,0x75,0x73,0x23,0x24,0x17,0x70,0x68,0x61,0x62,0x65,0x74,
+0x69,0x63,0x25,4,0x32,0x27,0x61,0x29,0x62,0x2b,0x6b,0x2d,0x72,0x12,0x65,0x61,
+0x6b,2,0x61,0x36,0x62,0x3e,0x73,0x15,0x79,0x6d,0x62,0x6f,0x6c,0x73,0x57,0x13,
+0x66,0x74,0x65,0x72,0x29,1,0x65,0x2a,0x6f,0x11,0x74,0x68,0x27,0x13,0x66,0x6f,
+0x72,0x65,0x2b,7,0x6d,0x51,0x6d,0x33,0x6f,0x28,0x70,0x69,0x72,0x35,1,0x6d,
+0x76,0x6e,1,0x64,0x3c,0x74,0x1a,0x69,0x6e,0x67,0x65,0x6e,0x74,0x62,0x72,0x65,
+0x61,0x6b,0x2f,0x15,0x69,0x74,0x69,0x6f,0x6e,0x61,0x1f,0x6c,0x6a,0x61,0x70,0x61,
+0x6e,0x65,0x73,0x65,0x73,0x74,0x61,0x72,0x74,0x65,0x72,0x6b,1,0x62,0x3a,0x70,
+0x19,0x6c,0x65,0x78,0x63,0x6f,0x6e,0x74,0x65,0x78,0x74,0x51,0x18,0x69,0x6e,0x69,
+0x6e,0x67,0x6d,0x61,0x72,0x6b,0x33,0x61,0x6a,0x62,0x2f,0x6a,0x6b,0x6c,0x30,0x13,
+0x6f,0x73,0x65,0x70,1,0x61,0x38,0x75,0x18,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,
+0x6f,0x6e,0x31,0x18,0x72,0x65,0x6e,0x74,0x68,0x65,0x73,0x69,0x73,0x69,0x1b,0x72,
+0x72,0x69,0x61,0x67,0x65,0x72,0x65,0x74,0x75,0x72,0x6e,0x35,2,0x62,0x3e,0x6d,
+0x46,0x78,0x36,0x18,0x63,0x6c,0x61,0x6d,0x61,0x74,0x69,0x6f,0x6e,0x37,0x70,0x12,
+0x61,0x73,0x65,0x71,0x72,0x16,0x6f,0x64,0x69,0x66,0x69,0x65,0x72,0x73,1,0x64,
+0x42,0x6e,1,0x6f,0x32,0x75,0x26,0x14,0x6d,0x65,0x72,0x69,0x63,0x27,0x11,0x6e,
+0x65,0x21,1,0x65,0x2e,0x69,0x24,0x12,0x67,0x69,0x74,0x25,0x22,0x14,0x63,0x69,
+0x6d,0x61,0x6c,0x23,0,0x18,0x6e,0xc4,0x2a,0x74,0xc1,0x6d,0x77,0x96,0x77,0xa2,
+0x4c,0x78,0xa2,0x70,0x79,0xa2,0x7a,0x7a,6,0x73,0x1e,0x73,0x34,0x78,0x42,0x79,
+0x48,0x7a,0x11,0x7a,0x7a,0xa3,0x67,0x10,0x79,1,0x65,0xa3,0xae,0x6d,0xa3,0x81,
+0x11,0x78,0x78,0xa3,0x66,0x11,0x79,0x79,0x21,0x61,0x30,0x69,0x58,0x6d,0x11,0x74,
+0x68,0xa3,0x80,0x10,0x6e,1,0x61,0x26,0x62,0xa3,0xb1,0x1a,0x62,0x61,0x7a,0x61,
+0x72,0x73,0x71,0x75,0x61,0x72,0x65,0xa3,0xb1,0x11,0x6e,0x68,0x23,2,0x61,0x30,
+0x63,0x5a,0x6f,0x11,0x6c,0x65,0xa3,0x9b,1,0x6e,0x3c,0x72,0x10,0x61,0xa2,0x92,
+0x15,0x6e,0x67,0x63,0x69,0x74,0x69,0xa3,0x92,0x12,0x63,0x68,0x6f,0xa3,0xbc,0x11,
+0x68,0x6f,0xa3,0xbc,1,0x70,0x2c,0x73,0x11,0x75,0x78,0xa3,0x65,0x11,0x65,0x6f,
+0x9b,1,0x65,0x2c,0x69,0x72,0x11,0x69,0x69,0x73,0x11,0x7a,0x69,0xa2,0xc0,0x11,
+0x64,0x69,0xa3,0xc0,0x74,0x4a,0x75,0xa2,0xba,0x76,1,0x61,0x2c,0x69,0x11,0x73,
+0x70,0xa3,0x64,0x10,0x69,0xa2,0x63,0x10,0x69,0xa3,0x63,5,0x67,0x36,0x67,0x68,
+0x68,0x6c,0x69,2,0x62,0x3a,0x66,0x4a,0x72,0x10,0x68,0xa2,0x9e,0x12,0x75,0x74,
+0x61,0xa3,0x9e,1,0x65,0x24,0x74,0x6f,0x12,0x74,0x61,0x6e,0x6f,0x14,0x69,0x6e,
+0x61,0x67,0x68,0x99,0x11,0x6c,0x67,0x75,0x10,0x61,1,0x61,0x24,0x69,0x6d,0x6a,
+0x11,0x6e,0x61,0x6b,0x61,0x30,0x65,0xa2,0x5b,0x66,0x11,0x6e,0x67,0x99,6,0x6c,
+0x21,0x6c,0x32,0x6d,0x38,0x6e,0x44,0x76,0x10,0x74,0xa3,0x7f,1,0x65,0x89,0x75,
+0x97,1,0x69,0x24,0x6c,0x67,0x10,0x6c,0x67,0x10,0x67,0xa2,0x9a,0x11,0x75,0x74,
+0xa3,0x9a,0x67,0x36,0x69,0x52,0x6b,0x10,0x72,0xa2,0x99,0x10,0x69,0xa3,0x99,1,
+0x61,0x30,0x62,0x7a,0x13,0x61,0x6e,0x77,0x61,0x7b,0x12,0x6c,0x6f,0x67,0x75,2,
+0x6c,0x32,0x74,0x34,0x76,0x12,0x69,0x65,0x74,0xa3,0x7f,0x10,0x65,0x89,0x12,0x68,
+0x61,0x6d,0xa3,0x6a,1,0x6c,0x2a,0x6e,0x10,0x67,0xa3,0x62,0x10,0x75,0x68,0x11,
+0x67,0x75,0x69,1,0x67,0x32,0x6e,0x14,0x6b,0x6e,0x6f,0x77,0x6e,0xa3,0x67,0x11,
+0x61,0x72,0x8a,0x13,0x69,0x74,0x69,0x63,0x8b,0x71,0xc1,0x13,0x71,0xa2,0xde,0x72,
+0xa2,0xe3,0x73,6,0x69,0x8a,0x69,0x72,0x6f,0xa2,0x4c,0x75,0xa2,0x75,0x79,1,
+0x6c,0x46,0x72,4,0x63,0x65,0x65,0xa3,0x5f,0x69,0x2c,0x6a,0xa3,0x60,0x6e,0xa3,
+0x61,0x11,0x61,0x63,0x65,0x10,0x6f,0x94,0x16,0x74,0x69,0x6e,0x61,0x67,0x72,0x69,
+0x95,2,0x64,0x3c,0x67,0x4c,0x6e,1,0x64,0xa3,0x91,0x68,0x62,0x12,0x61,0x6c,
+0x61,0x63,0x10,0x64,0xa2,0xa6,0x12,0x68,0x61,0x6d,0xa3,0xa6,0x17,0x6e,0x77,0x72,
+0x69,0x74,0x69,0x6e,0x67,0xa3,0x70,2,0x67,0x3a,0x72,0x52,0x79,0x10,0x6f,0xa2,
+0xb0,0x12,0x6d,0x62,0x6f,0xa3,0xb0,1,0x64,0x26,0x6f,0xa3,0xb8,0xa2,0xb7,0x12,
+0x69,0x61,0x6e,0xa3,0xb7,0x10,0x61,0xa2,0x98,0x16,0x73,0x6f,0x6d,0x70,0x65,0x6e,
+0x67,0xa3,0x98,0x11,0x6e,0x64,0xa2,0x71,0x14,0x61,0x6e,0x65,0x73,0x65,0xa3,0x71,
+0x61,0x5c,0x67,0xa2,0x43,0x68,1,0x61,0x2a,0x72,0x10,0x64,0xa3,0x97,2,0x72,
+0x28,0x76,0x30,0x77,0x87,0x12,0x61,0x64,0x61,0xa3,0x97,0x12,0x69,0x61,0x6e,0x87,
+2,0x6d,0x40,0x72,0x58,0x75,0x10,0x72,0xa2,0x6f,0x15,0x61,0x73,0x68,0x74,0x72,
+0x61,0xa3,0x6f,1,0x61,0x26,0x72,0xa3,0x7e,0x14,0x72,0x69,0x74,0x61,0x6e,0xa3,
+0x7e,1,0x61,0xa3,0x5e,0x62,0xa3,0x85,0x11,0x6e,0x77,0xa3,0x70,0x11,0x61,0x61,
+1,0x63,0x2f,0x69,0x23,3,0x65,0x3e,0x6a,0x48,0x6f,0x4e,0x75,0x10,0x6e,1,
+0x69,0x24,0x72,0x61,0x10,0x63,0x61,0x13,0x6a,0x61,0x6e,0x67,0xa3,0x6e,0x11,0x6e,
+0x67,0xa3,0x6e,1,0x68,0x2a,0x72,0x10,0x6f,0xa3,0x5d,0x10,0x67,0xa3,0xb6,0x6e,
+0xa2,0x83,0x6f,0xa2,0xf2,0x70,5,0x6c,0x1e,0x6c,0x44,0x72,0x4a,0x73,0x1b,0x61,
+0x6c,0x74,0x65,0x72,0x70,0x61,0x68,0x6c,0x61,0x76,0x69,0xa3,0x7b,0x11,0x72,0x64,
+0xa3,0x5c,0x11,0x74,0x69,0xa3,0x7d,0x61,0x7c,0x65,0xa2,0x54,0x68,3,0x61,0x3e,
+0x6c,0x4e,0x6e,0x5e,0x6f,0x16,0x65,0x6e,0x69,0x63,0x69,0x61,0x6e,0xa3,0x5b,0x10,
+0x67,0xa2,0x5a,0x12,0x73,0x70,0x61,0xa3,0x5a,2,0x69,0xa3,0x7a,0x70,0xa3,0x7b,
+0x76,0xa3,0x7c,0x10,0x78,0xa3,0x5b,2,0x68,0x3e,0x6c,0x50,0x75,0x10,0x63,0xa2,
+0xa5,0x14,0x69,0x6e,0x68,0x61,0x75,0xa3,0xa5,0x17,0x61,0x77,0x68,0x68,0x6d,0x6f,
+0x6e,0x67,0xa3,0x4b,0x10,0x6d,0xa2,0x90,0x14,0x79,0x72,0x65,0x6e,0x65,0xa3,0x90,
+0x11,0x72,0x6d,0xa3,0x59,6,0x6b,0x36,0x6b,0x56,0x73,0x6e,0x75,0x74,0x79,0x11,
+0x69,0x61,0x1f,0x6b,0x65,0x6e,0x67,0x70,0x75,0x61,0x63,0x68,0x75,0x65,0x68,0x6d,
+0x6f,0x6e,0x67,0xa3,0xba,1,0x67,0x2e,0x6f,0xa2,0x57,0x10,0x6f,0xa3,0x57,0x10,
+0x62,0xa3,0x84,0x11,0x68,0x75,0xa3,0x96,0x12,0x73,0x68,0x75,0xa3,0x96,0x61,0x42,
+0x62,0x80,0x65,0x10,0x77,1,0x61,0xa3,0xaa,0x74,0x14,0x61,0x69,0x6c,0x75,0x65,
+0x97,2,0x62,0x2e,0x6e,0x3c,0x72,0x10,0x62,0xa3,0x8e,0x15,0x61,0x74,0x61,0x65,
+0x61,0x6e,0xa3,0x8f,0x10,0x64,0xa2,0xbb,0x16,0x69,0x6e,0x61,0x67,0x61,0x72,0x69,
+0xa3,0xbb,0x11,0x61,0x74,0xa3,0x8f,3,0x67,0x5a,0x6c,0x6c,0x72,0xa2,0x93,0x73,
+2,0x61,0x36,0x67,0x3c,0x6d,0x10,0x61,0x84,0x12,0x6e,0x79,0x61,0x85,0x11,0x67,
+0x65,0xa3,0xab,0x10,0x65,0xa3,0xab,1,0x61,0x2a,0x68,0x11,0x61,0x6d,0x5b,0x10,
+0x6d,0x5b,1,0x63,0xa2,0x60,0x64,5,0x70,0x37,0x70,0x36,0x73,0x54,0x74,0x14,
+0x75,0x72,0x6b,0x69,0x63,0xa3,0x58,0x11,0x65,0x72,1,0x6d,0x2c,0x73,0x12,0x69,
+0x61,0x6e,0x9b,0x11,0x69,0x63,0xa3,0x59,0x10,0x6f,1,0x67,0x3a,0x75,0x18,0x74,
+0x68,0x61,0x72,0x61,0x62,0x69,0x61,0x6e,0xa3,0x85,0x13,0x64,0x69,0x61,0x6e,0xa3,
+0xb8,0x68,0x42,0x69,0x54,0x6e,0x1a,0x6f,0x72,0x74,0x68,0x61,0x72,0x61,0x62,0x69,
+0x61,0x6e,0xa3,0x8e,0x17,0x75,0x6e,0x67,0x61,0x72,0x69,0x61,0x6e,0xa3,0x4c,0x14,
+0x74,0x61,0x6c,0x69,0x63,0x5d,1,0x68,0x26,0x6b,0xa3,0x6d,0x12,0x69,0x6b,0x69,
+0xa3,0x6d,2,0x69,0x2c,0x6b,0x30,0x79,0x10,0x61,0x5f,0x11,0x79,0x61,0x5f,0x10,
+0x68,0xa3,0x58,0x68,0xc3,0xd,0x6b,0xc2,0x24,0x6b,0xa4,0x17,0x6c,0xa4,0xb2,0x6d,
+8,0x6f,0x46,0x6f,0x48,0x72,0x74,0x74,0x80,0x75,0x86,0x79,1,0x61,0x28,0x6d,
+0x10,0x72,0x59,0x13,0x6e,0x6d,0x61,0x72,0x59,2,0x64,0x2e,0x6e,0x32,0x6f,0x10,
+0x6e,0xa3,0x72,0x10,0x69,0xa3,0xa3,0x10,0x67,0x56,0x14,0x6f,0x6c,0x69,0x61,0x6e,
+0x57,0x10,0x6f,0xa2,0x95,0x10,0x6f,0xa3,0x95,0x11,0x65,0x69,0xa3,0x73,0x11,0x6c,
+0x74,0xa2,0xa4,0x12,0x61,0x6e,0x69,0xa3,0xa4,0x61,0x36,0x65,0xa2,0x67,0x69,0xa2,
+0xbd,0x6c,0x11,0x79,0x6d,0x55,6,0x6e,0x38,0x6e,0x32,0x72,0x5c,0x73,0x6c,0x79,
+0x10,0x61,0xa3,0x55,1,0x64,0x38,0x69,0xa2,0x79,0x15,0x63,0x68,0x61,0x65,0x61,
+0x6e,0xa3,0x79,0xa2,0x54,0x12,0x61,0x69,0x63,0xa3,0x54,0x10,0x63,0xa2,0xa9,0x12,
+0x68,0x65,0x6e,0xa3,0xa9,0x18,0x61,0x72,0x61,0x6d,0x67,0x6f,0x6e,0x64,0x69,0xa3,
+0xaf,0x68,0x36,0x6b,0x4c,0x6c,0x15,0x61,0x79,0x61,0x6c,0x61,0x6d,0x55,1,0x61,
+0x26,0x6a,0xa3,0xa0,0x13,0x6a,0x61,0x6e,0x69,0xa3,0xa0,0x10,0x61,0xa2,0xb4,0x12,
+0x73,0x61,0x72,0xa3,0xb4,3,0x64,0x78,0x65,0x94,0x6e,0xa2,0x42,0x72,1,0x63,
+0xa3,0x8d,0x6f,0xa2,0x56,0x13,0x69,0x74,0x69,0x63,1,0x63,0x3c,0x68,0x19,0x69,
+0x65,0x72,0x6f,0x67,0x6c,0x79,0x70,0x68,0x73,0xa3,0x56,0x15,0x75,0x72,0x73,0x69,
+0x76,0x65,0xa3,0x8d,1,0x65,0x26,0x66,0xa3,0xb5,0x16,0x66,0x61,0x69,0x64,0x72,
+0x69,0x6e,0xa3,0xb5,0x17,0x74,0x65,0x69,0x6d,0x61,0x79,0x65,0x6b,0xa3,0x73,0x10,
+0x64,0xa2,0x8c,0x17,0x65,0x6b,0x69,0x6b,0x61,0x6b,0x75,0x69,0xa3,0x8c,0x11,0x61,
+0x6f,0xa3,0x5c,6,0x6e,0x1a,0x6e,0x34,0x6f,0x38,0x70,0x3e,0x74,0x11,0x68,0x69,
+0xa3,0x78,0x11,0x64,0x61,0x4b,0x11,0x72,0x65,0xa3,0x77,0x11,0x65,0x6c,0xa3,0x8a,
+0x61,0x30,0x68,0x9a,0x69,0x11,0x74,0x73,0xa3,0xbf,4,0x69,0x3c,0x6c,0x44,0x6e,
+0x48,0x74,0x56,0x79,0x13,0x61,0x68,0x6c,0x69,0xa3,0x4f,0x12,0x74,0x68,0x69,0xa3,
+0x78,0x10,0x69,0xa3,0x4f,1,0x61,0x4d,0x6e,0x12,0x61,0x64,0x61,0x4b,0x14,0x61,
+0x6b,0x61,0x6e,0x61,0x4c,0x19,0x6f,0x72,0x68,0x69,0x72,0x61,0x67,0x61,0x6e,0x61,
+0x8d,4,0x61,0x40,0x69,0x52,0x6d,0x70,0x6f,0x7c,0x75,0x15,0x64,0x61,0x77,0x61,
+0x64,0x69,0xa3,0x91,0x10,0x72,0x92,0x15,0x6f,0x73,0x68,0x74,0x68,0x69,0x93,0x1d,
+0x74,0x61,0x6e,0x73,0x6d,0x61,0x6c,0x6c,0x73,0x63,0x72,0x69,0x70,0x74,0xa3,0xbf,
+1,0x65,0x24,0x72,0x4f,0x10,0x72,0x4f,0x10,0x6a,0xa2,0x9d,0x11,0x6b,0x69,0xa3,
+0x9d,4,0x61,0x5c,0x65,0x90,0x69,0xa0,0x6f,0xa2,0x5d,0x79,1,0x63,0x34,0x64,
+0x10,0x69,0xa2,0x6c,0x11,0x61,0x6e,0xa3,0x6c,0x10,0x69,0xa2,0x6b,0x11,0x61,0x6e,
+0xa3,0x6b,2,0x6e,0x42,0x6f,0x46,0x74,3,0x66,0xa3,0x50,0x67,0xa3,0x51,0x69,
+0x24,0x6e,0x53,0x10,0x6e,0x53,0x10,0x61,0xa3,0x6a,0x50,0x10,0x6f,0x51,0x11,0x70,
+0x63,0xa2,0x52,0x11,0x68,0x61,0xa3,0x52,2,0x6d,0x2e,0x6e,0x36,0x73,0x10,0x75,
+0xa3,0x83,0x10,0x62,0x80,0x10,0x75,0x81,2,0x61,0xa3,0x53,0x62,0x83,0x65,0x11,
+0x61,0x72,1,0x61,0xa3,0x53,0x62,0x83,0x11,0x6d,0x61,0xa3,0x8b,0x68,0x6e,0x69,
+0xa2,0x95,0x6a,2,0x61,0x30,0x70,0x52,0x75,0x11,0x72,0x63,0xa3,0x94,1,0x6d,
+0x38,0x76,0x10,0x61,0xa2,0x4e,0x13,0x6e,0x65,0x73,0x65,0xa3,0x4e,0x10,0x6f,0xa3,
+0xad,0x11,0x61,0x6e,0xa3,0x69,6,0x6c,0x1e,0x6c,0x34,0x6d,0x3a,0x72,0x48,0x75,
+0x11,0x6e,0x67,0xa3,0x4c,0x11,0x75,0x77,0xa3,0x9c,0x10,0x6e,1,0x67,0xa3,0x4b,
+0x70,0xa3,0xba,0x11,0x6b,0x74,0x8d,0x61,0x3c,0x65,0xa2,0x43,0x69,0x11,0x72,0x61,
+0x48,0x13,0x67,0x61,0x6e,0x61,0x49,1,0x6e,0x34,0x74,0x10,0x72,0xa2,0xa2,0x11,
+0x61,0x6e,0xa3,0xa2,0x42,6,0x6f,0xe,0x6f,0x77,0x73,0xa3,0x49,0x74,0xa3,0x4a,
+0x75,0x12,0x6e,0x6f,0x6f,0x77,0x62,0xa3,0xac,0x67,0x3e,0x69,0x42,0x19,0x66,0x69,
+0x72,0x6f,0x68,0x69,0x6e,0x67,0x79,0x61,0xa3,0xb6,0x44,0x11,0x75,0x6c,0x45,0x11,
+0x62,0x72,0x46,0x11,0x65,0x77,0x47,2,0x6d,0x2e,0x6e,0x4a,0x74,0x11,0x61,0x6c,
+0x5d,0x1c,0x70,0x65,0x72,0x69,0x61,0x6c,0x61,0x72,0x61,0x6d,0x61,0x69,0x63,0xa3,
+0x74,2,0x64,0x66,0x68,0x6a,0x73,0x1b,0x63,0x72,0x69,0x70,0x74,0x69,0x6f,0x6e,
+0x61,0x6c,0x70,0x61,1,0x68,0x32,0x72,0x14,0x74,0x68,0x69,0x61,0x6e,0xa3,0x7d,
+0x13,0x6c,0x61,0x76,0x69,0xa3,0x7a,0x10,0x73,0xa3,0x4d,0x15,0x65,0x72,0x69,0x74,
+0x65,0x64,0x23,0x64,0xc1,0xd,0x64,0xa2,0x7a,0x65,0xa2,0xc1,0x67,4,0x65,0x82,
+0x6c,0x9a,0x6f,0xa2,0x46,0x72,0xa2,0x55,0x75,2,0x6a,0x3c,0x6e,0x4e,0x72,1,
+0x6d,0x24,0x75,0x41,0x13,0x75,0x6b,0x68,0x69,0x41,1,0x61,0x24,0x72,0x3f,0x13,
+0x72,0x61,0x74,0x69,0x3f,0x18,0x6a,0x61,0x6c,0x61,0x67,0x6f,0x6e,0x64,0x69,0xa3,
+0xb3,0x10,0x6f,1,0x6b,0xa3,0x48,0x72,0x38,0x13,0x67,0x69,0x61,0x6e,0x39,0x11,
+0x61,0x67,0x90,0x15,0x6f,0x6c,0x69,0x74,0x69,0x63,0x91,1,0x6e,0x30,0x74,0x10,
+0x68,0x3a,0x11,0x69,0x63,0x3b,1,0x67,0xa3,0xb3,0x6d,0xa3,0xaf,1,0x61,0x32,
+0x65,1,0x65,0x24,0x6b,0x3d,0x10,0x6b,0x3d,0x10,0x6e,0xa2,0x89,0x12,0x74,0x68,
+0x61,0xa3,0x89,4,0x65,0x46,0x69,0x6c,0x6f,0x8c,0x73,0x9a,0x75,0x11,0x70,0x6c,
+0xa2,0x87,0x13,0x6f,0x79,0x61,0x6e,0xa3,0x87,1,0x73,0x38,0x76,0x10,0x61,0x34,
+0x15,0x6e,0x61,0x67,0x61,0x72,0x69,0x35,0x13,0x65,0x72,0x65,0x74,0x33,1,0x61,
+0x36,0x76,0x16,0x65,0x73,0x61,0x6b,0x75,0x72,0x75,0xa3,0xbe,0x10,0x6b,0xa3,0xbe,
+0x11,0x67,0x72,0xa2,0xb2,0x10,0x61,0xa3,0xb2,0x11,0x72,0x74,0x33,2,0x67,0x3a,
+0x6c,0x72,0x74,0x11,0x68,0x69,0x36,0x13,0x6f,0x70,0x69,0x63,0x37,0x10,0x79,2,
+0x64,0xa3,0x45,0x68,0xa3,0x46,0x70,0xa2,0x47,0x1e,0x74,0x69,0x61,0x6e,0x68,0x69,
+0x65,0x72,0x6f,0x67,0x6c,0x79,0x70,0x68,0x73,0xa3,0x47,1,0x62,0x36,0x79,0x10,
+0x6d,0xa2,0xb9,0x12,0x61,0x69,0x63,0xa3,0xb9,0x10,0x61,0xa2,0x88,0x12,0x73,0x61,
+0x6e,0xa3,0x88,0x61,0xa2,0xb4,0x62,0xa4,0x19,0x63,6,0x6f,0x3d,0x6f,0x5a,0x70,
+0x76,0x75,0x7a,0x79,1,0x70,0x3e,0x72,2,0x69,0x2a,0x6c,0x31,0x73,0xa3,0x44,
+0x13,0x6c,0x6c,0x69,0x63,0x31,0x13,0x72,0x69,0x6f,0x74,0x7f,1,0x6d,0x30,0x70,
+0x10,0x74,0x2e,0x11,0x69,0x63,0x2f,0x12,0x6d,0x6f,0x6e,0x21,0x11,0x72,0x74,0x7f,
+0x16,0x6e,0x65,0x69,0x66,0x6f,0x72,0x6d,0xa3,0x65,0x61,0x32,0x68,0xa2,0x41,0x69,
+0x11,0x72,0x74,0xa3,0x43,3,0x6b,0x4c,0x6e,0x50,0x72,0x76,0x75,0x1d,0x63,0x61,
+0x73,0x69,0x61,0x6e,0x61,0x6c,0x62,0x61,0x6e,0x69,0x61,0x6e,0xa3,0x9f,0x10,0x6d,
+0xa3,0x76,1,0x61,0x24,0x73,0x71,0x1d,0x64,0x69,0x61,0x6e,0x61,0x62,0x6f,0x72,
+0x69,0x67,0x69,0x6e,0x61,0x6c,0x71,0x10,0x69,0xa2,0x68,0x11,0x61,0x6e,0xa3,0x68,
+3,0x61,0x32,0x65,0x44,0x6f,0x52,0x72,0x10,0x73,0xa3,0xbd,1,0x6b,0x26,0x6d,
+0xa3,0x42,0x11,0x6d,0x61,0xa3,0x76,0x10,0x72,0x2c,0x13,0x6f,0x6b,0x65,0x65,0x2d,
+0x16,0x72,0x61,0x73,0x6d,0x69,0x61,0x6e,0xa3,0xbd,6,0x68,0x4a,0x68,0x48,0x6e,
+0x4e,0x72,0x76,0x76,1,0x65,0x2a,0x73,0x10,0x74,0xa3,0x75,0x13,0x73,0x74,0x61,
+0x6e,0xa3,0x75,0x11,0x6f,0x6d,0xa3,0xa1,0x11,0x61,0x74,0x1f,0x6f,0x6c,0x69,0x61,
+0x6e,0x68,0x69,0x65,0x72,0x6f,0x67,0x6c,0x79,0x70,0x68,0x73,0xa3,0x9c,1,0x61,
+0x3e,0x6d,2,0x65,0x2a,0x69,0xa3,0x74,0x6e,0x27,0x13,0x6e,0x69,0x61,0x6e,0x27,
+0x10,0x62,0x24,0x11,0x69,0x63,0x25,0x64,0x30,0x66,0x44,0x67,0x11,0x68,0x62,0xa3,
+0x9f,0x10,0x6c,1,0x61,0x26,0x6d,0xa3,0xa7,0x10,0x6d,0xa3,0xa7,0x11,0x61,0x6b,
+0xa3,0x93,6,0x6c,0x3c,0x6c,0x52,0x6f,0x56,0x72,0x66,0x75,1,0x67,0x30,0x68,
+1,0x64,0x79,0x69,0x10,0x64,0x79,0x10,0x69,0x8e,0x13,0x6e,0x65,0x73,0x65,0x8f,
+0x11,0x69,0x73,0xa1,0x11,0x70,0x6f,0x2a,0x13,0x6d,0x6f,0x66,0x6f,0x2b,0x10,0x61,
+1,0x68,0x2e,0x69,0x7c,0x12,0x6c,0x6c,0x65,0x7d,0xa2,0x41,0x11,0x6d,0x69,0xa3,
+0x41,0x61,0x48,0x65,0x9c,0x68,1,0x61,0x2a,0x6b,0x10,0x73,0xa3,0xa8,0x15,0x69,
+0x6b,0x73,0x75,0x6b,0x69,0xa3,0xa8,3,0x6c,0x3a,0x6d,0x48,0x73,0x54,0x74,1,
+0x61,0x24,0x6b,0x9f,0x10,0x6b,0x9f,0x10,0x69,0x9c,0x13,0x6e,0x65,0x73,0x65,0x9d,
+0x10,0x75,0xa2,0x82,0x10,0x6d,0xa3,0x82,0x10,0x73,0xa2,0x86,0x13,0x61,0x76,0x61,
+0x68,0xa3,0x86,0x11,0x6e,0x67,0x28,0x12,0x61,0x6c,0x69,0x29,3,0x6c,0x42,0x6e,
+0x90,0x74,0xa2,0x46,0x76,0x24,0x17,0x6f,0x77,0x65,0x6c,0x6a,0x61,0x6d,0x6f,0x25,
+0x22,1,0x65,0x54,0x76,0x28,1,0x73,0x38,0x74,0x2a,0x17,0x73,0x79,0x6c,0x6c,
+0x61,0x62,0x6c,0x65,0x2b,0x16,0x79,0x6c,0x6c,0x61,0x62,0x6c,0x65,0x29,0x18,0x61,
+0x64,0x69,0x6e,0x67,0x6a,0x61,0x6d,0x6f,0x23,1,0x61,0x21,0x6f,0x1a,0x74,0x61,
+0x70,0x70,0x6c,0x69,0x63,0x61,0x62,0x6c,0x65,0x21,0x26,0x1a,0x72,0x61,0x69,0x6c,
+0x69,0x6e,0x67,0x6a,0x61,0x6d,0x6f,0x27,1,0x6e,0x2c,0x79,0x22,0x11,0x65,0x73,
+0x23,0x20,0x10,0x6f,0x21,1,0x6e,0x2c,0x79,0x22,0x11,0x65,0x73,0x23,0x20,0x10,
+0x6f,0x21,2,0x6d,0x30,0x6e,0x3a,0x79,0x22,0x11,0x65,0x73,0x23,0x24,0x13,0x61,
+0x79,0x62,0x65,0x25,0x20,0x10,0x6f,0x21,2,0x6d,0x30,0x6e,0x3a,0x79,0x22,0x11,
+0x65,0x73,0x23,0x24,0x13,0x61,0x79,0x62,0x65,0x25,0x20,0x10,0x6f,0x21,0xb,0x72,
+0x39,0x76,0xc,0x76,0x33,0x78,0x2a,0x7a,0x11,0x77,0x6a,0x43,0x10,0x78,0x21,0x72,
+0x28,0x73,0x50,0x74,0x31,1,0x65,0x24,0x69,0x39,0x1e,0x67,0x69,0x6f,0x6e,0x61,
+0x6c,0x69,0x6e,0x64,0x69,0x63,0x61,0x74,0x6f,0x72,0x39,1,0x6d,0x35,0x70,0x18,
+0x61,0x63,0x69,0x6e,0x67,0x6d,0x61,0x72,0x6b,0x35,0x6c,0x1f,0x6c,0x3c,0x6f,0x4a,
+0x70,1,0x70,0x37,0x72,0x14,0x65,0x70,0x65,0x6e,0x64,0x37,0x28,1,0x66,0x2b,
+0x76,0x2c,0x10,0x74,0x2f,0x13,0x74,0x68,0x65,0x72,0x21,0x63,0x4c,0x65,0x64,0x67,
+1,0x61,0x3a,0x6c,0x19,0x75,0x65,0x61,0x66,0x74,0x65,0x72,0x7a,0x77,0x6a,0x41,
+0x10,0x7a,0x41,2,0x6e,0x23,0x6f,0x24,0x72,0x25,0x14,0x6e,0x74,0x72,0x6f,0x6c,
+0x23,2,0x62,0x34,0x6d,0x4e,0x78,0x26,0x13,0x74,0x65,0x6e,0x64,0x27,0x3a,1,
+0x61,0x24,0x67,0x3d,0x11,0x73,0x65,0x3a,0x12,0x67,0x61,0x7a,0x3d,0x3e,0x16,0x6f,
+0x64,0x69,0x66,0x69,0x65,0x72,0x3f,9,0x6e,0x4a,0x6e,0x34,0x6f,0x44,0x73,0x60,
+0x75,0x94,0x78,0x10,0x78,0x21,0x10,0x75,0x2a,0x14,0x6d,0x65,0x72,0x69,0x63,0x2b,
+1,0x6c,0x2c,0x74,0x12,0x68,0x65,0x72,0x21,0x14,0x65,0x74,0x74,0x65,0x72,0x2d,
+3,0x63,0x36,0x65,0x46,0x70,0x31,0x74,0x32,0x12,0x65,0x72,0x6d,0x33,0x3c,0x16,
+0x6f,0x6e,0x74,0x69,0x6e,0x75,0x65,0x3d,0x2e,0x10,0x70,0x2f,0x10,0x70,0x34,0x12,
+0x70,0x65,0x72,0x35,0x61,0x46,0x63,0x52,0x65,0x64,0x66,0x72,0x6c,2,0x65,0x2d,
+0x66,0x3b,0x6f,0x28,0x12,0x77,0x65,0x72,0x29,0x10,0x74,0x22,0x12,0x65,0x72,0x6d,
+0x23,1,0x6c,0x24,0x72,0x37,0x24,0x12,0x6f,0x73,0x65,0x25,0x10,0x78,0x38,0x13,
+0x74,0x65,0x6e,0x64,0x39,0x10,0x6f,0x26,0x13,0x72,0x6d,0x61,0x74,0x27,0,0x10,
+0x6c,0x88,0x72,0x40,0x72,0x36,0x73,0x5e,0x77,0x7a,0x78,0x8a,0x7a,0x11,0x77,0x6a,
+0x4b,1,0x65,0x24,0x69,0x3b,0x1e,0x67,0x69,0x6f,0x6e,0x61,0x6c,0x69,0x6e,0x64,
+0x69,0x63,0x61,0x74,0x6f,0x72,0x3b,1,0x69,0x24,0x71,0x3f,0x18,0x6e,0x67,0x6c,
+0x65,0x71,0x75,0x6f,0x74,0x65,0x3f,0x17,0x73,0x65,0x67,0x73,0x70,0x61,0x63,0x65,
+0x4d,0x10,0x78,0x21,0x6c,0x36,0x6d,0x3c,0x6e,0x76,0x6f,0x13,0x74,0x68,0x65,0x72,
+0x21,1,0x65,0x23,0x66,0x35,3,0x62,0x37,0x69,0x28,0x6c,0x29,0x6e,0x2b,0x10,
+0x64,1,0x6c,0x34,0x6e,0x11,0x75,0x6d,0x2a,0x12,0x6c,0x65,0x74,0x37,0x14,0x65,
+0x74,0x74,0x65,0x72,0x29,2,0x65,0x36,0x6c,0x39,0x75,0x2c,0x14,0x6d,0x65,0x72,
+0x69,0x63,0x2d,0x14,0x77,0x6c,0x69,0x6e,0x65,0x39,0x66,0x3f,0x66,0x40,0x67,0x4e,
+0x68,0x70,0x6b,0x10,0x61,0x26,0x15,0x74,0x61,0x6b,0x61,0x6e,0x61,0x27,0x10,0x6f,
+0x24,0x13,0x72,0x6d,0x61,0x74,0x25,1,0x61,0x3a,0x6c,0x19,0x75,0x65,0x61,0x66,
+0x74,0x65,0x72,0x7a,0x77,0x6a,0x49,0x10,0x7a,0x49,1,0x65,0x24,0x6c,0x3d,0x19,
+0x62,0x72,0x65,0x77,0x6c,0x65,0x74,0x74,0x65,0x72,0x3d,0x61,0x86,0x63,0x92,0x64,
+0x94,0x65,2,0x62,0x44,0x6d,0x5e,0x78,0x2e,0x13,0x74,0x65,0x6e,0x64,0x32,0x15,
+0x6e,0x75,0x6d,0x6c,0x65,0x74,0x2f,0x42,1,0x61,0x24,0x67,0x45,0x11,0x73,0x65,
+0x42,0x12,0x67,0x61,0x7a,0x45,0x46,0x16,0x6f,0x64,0x69,0x66,0x69,0x65,0x72,0x47,
+0x15,0x6c,0x65,0x74,0x74,0x65,0x72,0x23,0x10,0x72,0x31,1,0x6f,0x24,0x71,0x41,
+0x18,0x75,0x62,0x6c,0x65,0x71,0x75,0x6f,0x74,0x65,0x41,2,0x63,0x32,0x6e,0x3c,
+0x6f,0x22,0x12,0x70,0x65,0x6e,0x23,0x24,0x13,0x6c,0x6f,0x73,0x65,0x25,0x20,0x12,
+0x6f,0x6e,0x65,0x21,6,0x6f,0x65,0x6f,0x4a,0x72,0x5c,0x74,0x64,0x76,0x1d,0x69,
+0x73,0x75,0x61,0x6c,0x6f,0x72,0x64,0x65,0x72,0x6c,0x65,0x66,0x74,0x3d,0x18,0x76,
+0x65,0x72,0x73,0x74,0x72,0x75,0x63,0x6b,0x2d,0x13,0x69,0x67,0x68,0x74,0x2f,0x11,
+0x6f,0x70,0x30,0x12,0x61,0x6e,0x64,2,0x62,0x32,0x6c,0x62,0x72,0x13,0x69,0x67,
+0x68,0x74,0x3b,0x14,0x6f,0x74,0x74,0x6f,0x6d,0x32,0x12,0x61,0x6e,0x64,1,0x6c,
+0x2e,0x72,0x13,0x69,0x67,0x68,0x74,0x35,0x12,0x65,0x66,0x74,0x3f,0x12,0x65,0x66,
+0x74,0x36,0x17,0x61,0x6e,0x64,0x72,0x69,0x67,0x68,0x74,0x39,0x62,0x2c,0x6c,0x5c,
+0x6e,0x10,0x61,0x21,0x14,0x6f,0x74,0x74,0x6f,0x6d,0x22,0x12,0x61,0x6e,0x64,1,
+0x6c,0x2e,0x72,0x13,0x69,0x67,0x68,0x74,0x27,0x12,0x65,0x66,0x74,0x25,0x12,0x65,
+0x66,0x74,0x28,0x17,0x61,0x6e,0x64,0x72,0x69,0x67,0x68,0x74,0x2b,0xd,0x6e,0xaa,
+0x72,0x70,0x72,0x92,0x73,0xa2,0x46,0x74,0xa2,0x54,0x76,1,0x69,0x60,0x6f,0x12,
+0x77,0x65,0x6c,0x62,1,0x64,0x3a,0x69,0x19,0x6e,0x64,0x65,0x70,0x65,0x6e,0x64,
+0x65,0x6e,0x74,0x67,0x17,0x65,0x70,0x65,0x6e,0x64,0x65,0x6e,0x74,0x65,1,0x72,
+0x2e,0x73,0x13,0x61,0x72,0x67,0x61,0x61,0x12,0x61,0x6d,0x61,0x5f,0x1d,0x65,0x67,
+0x69,0x73,0x74,0x65,0x72,0x73,0x68,0x69,0x66,0x74,0x65,0x72,0x57,0x1e,0x79,0x6c,
+0x6c,0x61,0x62,0x6c,0x65,0x6d,0x6f,0x64,0x69,0x66,0x69,0x65,0x72,0x59,0x12,0x6f,
+0x6e,0x65,1,0x6c,0x2c,0x6d,0x12,0x61,0x72,0x6b,0x5d,0x14,0x65,0x74,0x74,0x65,
+0x72,0x5b,0x6e,0x3c,0x6f,0x7c,0x70,0x18,0x75,0x72,0x65,0x6b,0x69,0x6c,0x6c,0x65,
+0x72,0x55,1,0x6f,0x4c,0x75,1,0x6b,0x3c,0x6d,0x12,0x62,0x65,0x72,0x50,0x15,
+0x6a,0x6f,0x69,0x6e,0x65,0x72,0x53,0x11,0x74,0x61,0x4f,0x16,0x6e,0x6a,0x6f,0x69,
+0x6e,0x65,0x72,0x4d,0x13,0x74,0x68,0x65,0x72,0x21,0x67,0x3e,0x67,0x4a,0x69,0x64,
+0x6a,0x82,0x6d,0x1d,0x6f,0x64,0x69,0x66,0x79,0x69,0x6e,0x67,0x6c,0x65,0x74,0x74,
+0x65,0x72,0x4b,0x1c,0x65,0x6d,0x69,0x6e,0x61,0x74,0x69,0x6f,0x6e,0x6d,0x61,0x72,
+0x6b,0x45,0x1e,0x6e,0x76,0x69,0x73,0x69,0x62,0x6c,0x65,0x73,0x74,0x61,0x63,0x6b,
+0x65,0x72,0x47,0x14,0x6f,0x69,0x6e,0x65,0x72,0x49,0x61,0xa2,0xba,0x62,0xa2,0xc0,
+0x63,1,0x61,0xa2,0xa2,0x6f,0x16,0x6e,0x73,0x6f,0x6e,0x61,0x6e,0x74,0x2a,8,
+0x6b,0x67,0x6b,0x48,0x6d,0x52,0x70,0x5c,0x73,0xa2,0x42,0x77,0x19,0x69,0x74,0x68,
+0x73,0x74,0x61,0x63,0x6b,0x65,0x72,0x43,0x14,0x69,0x6c,0x6c,0x65,0x72,0x35,0x14,
+0x65,0x64,0x69,0x61,0x6c,0x37,1,0x6c,0x52,0x72,0x10,0x65,1,0x63,0x2e,0x66,
+0x13,0x69,0x78,0x65,0x64,0x3d,0x19,0x65,0x64,0x69,0x6e,0x67,0x72,0x65,0x70,0x68,
+0x61,0x3b,0x18,0x61,0x63,0x65,0x68,0x6f,0x6c,0x64,0x65,0x72,0x39,0x10,0x75,1,
+0x62,0x3e,0x63,0x1b,0x63,0x65,0x65,0x64,0x69,0x6e,0x67,0x72,0x65,0x70,0x68,0x61,
+0x41,0x15,0x6a,0x6f,0x69,0x6e,0x65,0x64,0x3f,0x64,0x4c,0x66,0x52,0x68,0x5a,0x69,
+0x1e,0x6e,0x69,0x74,0x69,0x61,0x6c,0x70,0x6f,0x73,0x74,0x66,0x69,0x78,0x65,0x64,
+0x33,0x12,0x65,0x61,0x64,0x2d,0x13,0x69,0x6e,0x61,0x6c,0x2f,0x18,0x65,0x61,0x64,
+0x6c,0x65,0x74,0x74,0x65,0x72,0x31,0x1d,0x6e,0x74,0x69,0x6c,0x6c,0x61,0x74,0x69,
+0x6f,0x6e,0x6d,0x61,0x72,0x6b,0x29,0x16,0x76,0x61,0x67,0x72,0x61,0x68,0x61,0x23,
+1,0x69,0x4a,0x72,0x10,0x61,0x1f,0x68,0x6d,0x69,0x6a,0x6f,0x69,0x6e,0x69,0x6e,
+0x67,0x6e,0x75,0x6d,0x62,0x65,0x72,0x27,0x12,0x6e,0x64,0x75,0x25,2,0x72,0x38,
+0x74,0x46,0x75,0x26,0x15,0x70,0x72,0x69,0x67,0x68,0x74,0x27,0x20,0x15,0x6f,0x74,
+0x61,0x74,0x65,0x64,0x21,1,0x72,0x24,0x75,0x25,0x22,0x18,0x61,0x6e,0x73,0x66,
+0x6f,0x72,0x6d,0x65,0x64,1,0x72,0x32,0x75,0x15,0x70,0x72,0x69,0x67,0x68,0x74,
+0x25,0x15,0x6f,0x74,0x61,0x74,0x65,0x64,0x23,0xd,0x6e,0xc1,0x86,0x73,0xa8,0x73,
+0x4c,0x74,0xa2,0x76,0x75,0xa2,0x83,0x7a,0xd8,0x70,0,2,0x6c,0xd9,0x20,0,
+0x70,0xd9,0x40,0,0x73,0xc3,0,0xfe,0xf,0,0,0,7,0x6f,0x3c,0x6f,
+0xff,8,0,0,0,0x70,0x3a,0x75,0x6e,0x79,0x13,0x6d,0x62,0x6f,0x6c,0xff,
+0xf,0,0,0,0x11,0x61,0x63,1,0x65,0x34,0x69,0x15,0x6e,0x67,0x6d,0x61,
+0x72,0x6b,0xa5,0,0x18,0x73,0x65,0x70,0x61,0x72,0x61,0x74,0x6f,0x72,0xc3,0,
+0x16,0x72,0x72,0x6f,0x67,0x61,0x74,0x65,0xe1,0,0,0x63,0xff,2,0,0,
+0,0x65,0x38,0x6b,0xff,4,0,0,0,0x6d,0xff,1,0,0,0,0x16,
+0x70,0x61,0x72,0x61,0x74,0x6f,0x72,0xd9,0x70,0,0x1d,0x69,0x74,0x6c,0x65,0x63,
+0x61,0x73,0x65,0x6c,0x65,0x74,0x74,0x65,0x72,0x31,1,0x6e,0x40,0x70,0x1c,0x70,
+0x65,0x72,0x63,0x61,0x73,0x65,0x6c,0x65,0x74,0x74,0x65,0x72,0x25,0x17,0x61,0x73,
+0x73,0x69,0x67,0x6e,0x65,0x64,0x23,0x6e,0xa2,0x69,0x6f,0xa2,0x89,0x70,0xfe,0x30,
+0xf8,0,0,9,0x69,0x33,0x69,0xff,0x10,0,0,0,0x6f,0xfd,0x80,0,
+0,0x72,0x54,0x73,0xf9,0,0,0x75,0x12,0x6e,0x63,0x74,0xfe,0x30,0xf8,0,
+0,0x15,0x75,0x61,0x74,0x69,0x6f,0x6e,0xff,0x30,0xf8,0,0,0x17,0x69,0x76,
+0x61,0x74,0x65,0x75,0x73,0x65,0xdd,0,0,0x61,0x48,0x63,0xfd,0x40,0,0,
+0x64,0xe9,0,0,0x65,0xfd,0x20,0,0,0x66,0xff,0x20,0,0,0,0x1f,
+0x72,0x61,0x67,0x72,0x61,0x70,0x68,0x73,0x65,0x70,0x61,0x72,0x61,0x74,0x6f,0x72,
+0xd9,0x40,0,0xbe,0,3,0x64,0xa7,0,0x6c,0xab,0,0x6f,0x30,0x75,0x13,
+0x6d,0x62,0x65,0x72,0xbf,0,0xb2,0,0x1b,0x6e,0x73,0x70,0x61,0x63,0x69,0x6e,
+0x67,0x6d,0x61,0x72,0x6b,0xa1,1,0x70,0x92,0x74,0x12,0x68,0x65,0x72,0xe6,0x80,
+1,3,0x6c,0x40,0x6e,0x4a,0x70,0x56,0x73,0x14,0x79,0x6d,0x62,0x6f,0x6c,0xff,
+8,0,0,0,0x14,0x65,0x74,0x74,0x65,0x72,0x61,0x14,0x75,0x6d,0x62,0x65,
+0x72,0xb3,0,0x19,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xfd,0x80,
+0,0,0x1c,0x65,0x6e,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,
+0xf9,0,0,0x66,0xc0,0xc4,0x66,0xa2,0x47,0x69,0xa2,0x64,0x6c,0xa2,0x79,0x6d,
+0xa4,0xc0,4,0x61,0x6c,0x63,0xa5,0,0x65,0xa3,0x80,0x6e,0xa1,0x6f,0x15,0x64,
+0x69,0x66,0x69,0x65,0x72,1,0x6c,0x38,0x73,0x14,0x79,0x6d,0x62,0x6f,0x6c,0xff,
+4,0,0,0,0x14,0x65,0x74,0x74,0x65,0x72,0x41,1,0x72,0x3c,0x74,0x16,
+0x68,0x73,0x79,0x6d,0x62,0x6f,0x6c,0xff,1,0,0,0,0x10,0x6b,0xa5,0xc0,
+1,0x69,0x32,0x6f,0x13,0x72,0x6d,0x61,0x74,0xdb,0,0,0x1d,0x6e,0x61,0x6c,
+0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xff,0x20,0,0,0,
+0x10,0x6e,0x1f,0x69,0x74,0x69,0x61,0x6c,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,
+0x69,0x6f,0x6e,0xff,0x10,0,0,0,0x9c,7,0x6d,0x18,0x6d,0x41,0x6f,0x28,
+0x74,0x31,0x75,0x25,0x60,0x1c,0x77,0x65,0x72,0x63,0x61,0x73,0x65,0x6c,0x65,0x74,
+0x74,0x65,0x72,0x29,0x63,0x3d,0x65,0x28,0x69,0x42,0x6c,0x29,0x13,0x74,0x74,0x65,
+0x72,0x9c,0x15,0x6e,0x75,0x6d,0x62,0x65,0x72,0xab,0,0x1a,0x6e,0x65,0x73,0x65,
+0x70,0x61,0x72,0x61,0x74,0x6f,0x72,0xd9,0x20,0,0x63,0x46,0x64,0xa2,0x96,0x65,
+0x1b,0x6e,0x63,0x6c,0x6f,0x73,0x69,0x6e,0x67,0x6d,0x61,0x72,0x6b,0xa3,0x80,0xe6,
+0x80,1,7,0x6e,0x57,0x6e,0x52,0x6f,0x5e,0x73,0xe1,0,0,0x75,0x1b,0x72,
+0x72,0x65,0x6e,0x63,0x79,0x73,0x79,0x6d,0x62,0x6f,0x6c,0xff,2,0,0,0,
+0x22,0x12,0x74,0x72,0x6c,0xd9,0x80,0,0xdc,0,0,1,0x6d,0x62,0x6e,1,
+0x6e,0x30,0x74,0x12,0x72,0x6f,0x6c,0xd9,0x80,0,0x1f,0x65,0x63,0x74,0x6f,0x72,
+0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xfd,0x40,0,0,0x19,
+0x62,0x69,0x6e,0x69,0x6e,0x67,0x6d,0x61,0x72,0x6b,0xa5,0xc0,0x61,0x58,0x63,0xd9,
+0x80,0,0x66,0xdb,0,0,0x6c,0x1d,0x6f,0x73,0x65,0x70,0x75,0x6e,0x63,0x74,
+0x75,0x61,0x74,0x69,0x6f,0x6e,0xfd,0x20,0,0,0x18,0x73,0x65,0x64,0x6c,0x65,
+0x74,0x74,0x65,0x72,0x3d,2,0x61,0x32,0x65,0x50,0x69,0x12,0x67,0x69,0x74,0xa7,
+0,0x1c,0x73,0x68,0x70,0x75,0x6e,0x63,0x74,0x75,0x61,0x74,0x69,0x6f,0x6e,0xe9,
+0,0,0x1a,0x63,0x69,0x6d,0x61,0x6c,0x6e,0x75,0x6d,0x62,0x65,0x72,0xa7,0
+};
+
+const char PropNameData::nameGroups[22098]={
+2,'A','l','p','h','a',0,'A','l','p','h','a','b','e','t','i','c',0,
+4,'N',0,'N','o',0,'F',0,'F','a','l','s','e',0,4,'Y',0,'Y','e','s',0,'T',0,'T','r','u','e',0,
+2,'N','R',0,'N','o','t','_','R','e','o','r','d','e','r','e','d',0,
+2,'O','V',0,'O','v','e','r','l','a','y',0,2,'H','A','N','R',0,'H','a','n','_','R','e','a','d','i','n','g',0,
+2,'N','K',0,'N','u','k','t','a',0,2,'K','V',0,'K','a','n','a','_','V','o','i','c','i','n','g',0,
+2,'V','R',0,'V','i','r','a','m','a',0,2,'C','C','C','1','0',0,'C','C','C','1','0',0,
+2,'C','C','C','1','1',0,'C','C','C','1','1',0,2,'C','C','C','1','2',0,'C','C','C','1','2',0,
+2,'C','C','C','1','3',0,'C','C','C','1','3',0,2,'C','C','C','1','4',0,'C','C','C','1','4',0,
+2,'C','C','C','1','5',0,'C','C','C','1','5',0,2,'C','C','C','1','6',0,'C','C','C','1','6',0,
+2,'C','C','C','1','7',0,'C','C','C','1','7',0,2,'C','C','C','1','8',0,'C','C','C','1','8',0,
+2,'C','C','C','1','9',0,'C','C','C','1','9',0,2,'C','C','C','2','0',0,'C','C','C','2','0',0,
+2,'C','C','C','2','1',0,'C','C','C','2','1',0,2,'C','C','C','2','2',0,'C','C','C','2','2',0,
+2,'C','C','C','2','3',0,'C','C','C','2','3',0,2,'C','C','C','2','4',0,'C','C','C','2','4',0,
+2,'C','C','C','2','5',0,'C','C','C','2','5',0,2,'C','C','C','2','6',0,'C','C','C','2','6',0,
+2,'C','C','C','2','7',0,'C','C','C','2','7',0,2,'C','C','C','2','8',0,'C','C','C','2','8',0,
+2,'C','C','C','2','9',0,'C','C','C','2','9',0,2,'C','C','C','3','0',0,'C','C','C','3','0',0,
+2,'C','C','C','3','1',0,'C','C','C','3','1',0,2,'C','C','C','3','2',0,'C','C','C','3','2',0,
+2,'C','C','C','3','3',0,'C','C','C','3','3',0,2,'C','C','C','3','4',0,'C','C','C','3','4',0,
+2,'C','C','C','3','5',0,'C','C','C','3','5',0,2,'C','C','C','3','6',0,'C','C','C','3','6',0,
+2,'C','C','C','8','4',0,'C','C','C','8','4',0,2,'C','C','C','9','1',0,'C','C','C','9','1',0,
+2,'C','C','C','1','0','3',0,'C','C','C','1','0','3',0,2,'C','C','C','1','0','7',0,'C','C','C','1','0','7',0,
+2,'C','C','C','1','1','8',0,'C','C','C','1','1','8',0,2,'C','C','C','1','2','2',0,'C','C','C','1','2','2',0,
+2,'C','C','C','1','2','9',0,'C','C','C','1','2','9',0,2,'C','C','C','1','3','0',0,'C','C','C','1','3','0',0,
+2,'C','C','C','1','3','2',0,'C','C','C','1','3','2',0,2,'C','C','C','1','3','3',0,'C','C','C','1','3','3',0,
+2,'A','T','B','L',0,'A','t','t','a','c','h','e','d','_','B','e','l','o','w','_','L','e','f','t',0,
+2,'A','T','B',0,'A','t','t','a','c','h','e','d','_','B','e','l','o','w',0,
+2,'A','T','A',0,'A','t','t','a','c','h','e','d','_','A','b','o','v','e',0,
+2,'A','T','A','R',0,'A','t','t','a','c','h','e','d','_','A','b','o','v','e','_','R','i','g','h','t',0,
+2,'B','L',0,'B','e','l','o','w','_','L','e','f','t',0,2,'B',0,'B','e','l','o','w',0,
+2,'B','R',0,'B','e','l','o','w','_','R','i','g','h','t',0,
+2,'L',0,'L','e','f','t',0,2,'R',0,'R','i','g','h','t',0,
+2,'A','L',0,'A','b','o','v','e','_','L','e','f','t',0,2,'A',0,'A','b','o','v','e',0,
+2,'A','R',0,'A','b','o','v','e','_','R','i','g','h','t',0,
+2,'D','B',0,'D','o','u','b','l','e','_','B','e','l','o','w',0,
+2,'D','A',0,'D','o','u','b','l','e','_','A','b','o','v','e',0,
+2,'I','S',0,'I','o','t','a','_','S','u','b','s','c','r','i','p','t',0,
+2,'A','H','e','x',0,'A','S','C','I','I','_','H','e','x','_','D','i','g','i','t',0,
+2,'B','i','d','i','_','C',0,'B','i','d','i','_','C','o','n','t','r','o','l',0,
+2,'B','i','d','i','_','M',0,'B','i','d','i','_','M','i','r','r','o','r','e','d',0,
+2,'D','a','s','h',0,'D','a','s','h',0,2,'D','I',0,'D','e','f','a','u','l','t','_','I','g','n','o','r','a','b','l','e',
+'_','C','o','d','e','_','P','o','i','n','t',0,2,'D','e','p',0,'D','e','p','r','e','c','a','t','e','d',0,
+2,'D','i','a',0,'D','i','a','c','r','i','t','i','c',0,2,'E','x','t',0,'E','x','t','e','n','d','e','r',0,
+2,'C','o','m','p','_','E','x',0,'F','u','l','l','_','C','o','m','p','o','s','i','t','i','o','n','_','E','x','c','l','u','s',
+'i','o','n',0,2,'G','r','_','B','a','s','e',0,'G','r','a','p','h','e','m','e','_','B','a','s','e',0,
+2,'G','r','_','E','x','t',0,'G','r','a','p','h','e','m','e','_','E','x','t','e','n','d',0,
+2,'G','r','_','L','i','n','k',0,'G','r','a','p','h','e','m','e','_','L','i','n','k',0,
+2,'H','e','x',0,'H','e','x','_','D','i','g','i','t',0,2,'H','y','p','h','e','n',0,'H','y','p','h','e','n',0,
+2,'I','D','C',0,'I','D','_','C','o','n','t','i','n','u','e',0,
+2,'I','D','S',0,'I','D','_','S','t','a','r','t',0,2,'I','d','e','o',0,'I','d','e','o','g','r','a','p','h','i','c',0,
+2,'I','D','S','B',0,'I','D','S','_','B','i','n','a','r','y','_','O','p','e','r','a','t','o','r',0,
+2,'I','D','S','T',0,'I','D','S','_','T','r','i','n','a','r','y','_','O','p','e','r','a','t','o','r',0,
+2,'J','o','i','n','_','C',0,'J','o','i','n','_','C','o','n','t','r','o','l',0,
+2,'L','O','E',0,'L','o','g','i','c','a','l','_','O','r','d','e','r','_','E','x','c','e','p','t','i','o','n',0,
+2,'L','o','w','e','r',0,'L','o','w','e','r','c','a','s','e',0,
+2,'M','a','t','h',0,'M','a','t','h',0,2,'N','C','h','a','r',0,'N','o','n','c','h','a','r','a','c','t','e','r','_','C',
+'o','d','e','_','P','o','i','n','t',0,2,'Q','M','a','r','k',0,'Q','u','o','t','a','t','i','o','n','_','M','a','r','k',0,
+2,'R','a','d','i','c','a','l',0,'R','a','d','i','c','a','l',0,
+2,'S','D',0,'S','o','f','t','_','D','o','t','t','e','d',0,
+2,'T','e','r','m',0,'T','e','r','m','i','n','a','l','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'U','I','d','e','o',0,'U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',0,
+2,'U','p','p','e','r',0,'U','p','p','e','r','c','a','s','e',0,
+3,'W','S','p','a','c','e',0,'W','h','i','t','e','_','S','p','a','c','e',0,'s','p','a','c','e',0,
+2,'X','I','D','C',0,'X','I','D','_','C','o','n','t','i','n','u','e',0,
+2,'X','I','D','S',0,'X','I','D','_','S','t','a','r','t',0,
+2,'S','e','n','s','i','t','i','v','e',0,'C','a','s','e','_','S','e','n','s','i','t','i','v','e',0,
+2,'S','T','e','r','m',0,'S','e','n','t','e','n','c','e','_','T','e','r','m','i','n','a','l',0,
+2,'V','S',0,'V','a','r','i','a','t','i','o','n','_','S','e','l','e','c','t','o','r',0,
+2,'n','f','d','i','n','e','r','t',0,'N','F','D','_','I','n','e','r','t',0,
+2,'n','f','k','d','i','n','e','r','t',0,'N','F','K','D','_','I','n','e','r','t',0,
+2,'n','f','c','i','n','e','r','t',0,'N','F','C','_','I','n','e','r','t',0,
+2,'n','f','k','c','i','n','e','r','t',0,'N','F','K','C','_','I','n','e','r','t',0,
+2,'s','e','g','s','t','a','r','t',0,'S','e','g','m','e','n','t','_','S','t','a','r','t','e','r',0,
+2,'P','a','t','_','S','y','n',0,'P','a','t','t','e','r','n','_','S','y','n','t','a','x',0,
+2,'P','a','t','_','W','S',0,'P','a','t','t','e','r','n','_','W','h','i','t','e','_','S','p','a','c','e',0,
+2,0,'a','l','n','u','m',0,2,0,'b','l','a','n','k',0,
+2,0,'g','r','a','p','h',0,2,0,'p','r','i','n','t',0,
+2,0,'x','d','i','g','i','t',0,2,'C','a','s','e','d',0,'C','a','s','e','d',0,
+2,'C','I',0,'C','a','s','e','_','I','g','n','o','r','a','b','l','e',0,
+2,'C','W','L',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','L','o','w','e','r','c','a','s','e','d',0,
+2,'C','W','U',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','U','p','p','e','r','c','a','s','e','d',0,
+2,'C','W','T',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','T','i','t','l','e','c','a','s','e','d',0,
+2,'C','W','C','F',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','C','a','s','e','f','o','l','d','e','d',0,
+2,'C','W','C','M',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','C','a','s','e','m','a','p','p','e','d',0,
+2,'C','W','K','C','F',0,'C','h','a','n','g','e','s','_','W','h','e','n','_','N','F','K','C','_','C','a','s','e','f','o','l',
+'d','e','d',0,2,'E','m','o','j','i',0,'E','m','o','j','i',0,
+2,'E','P','r','e','s',0,'E','m','o','j','i','_','P','r','e','s','e','n','t','a','t','i','o','n',0,
+2,'E','M','o','d',0,'E','m','o','j','i','_','M','o','d','i','f','i','e','r',0,
+2,'E','B','a','s','e',0,'E','m','o','j','i','_','M','o','d','i','f','i','e','r','_','B','a','s','e',0,
+2,'E','C','o','m','p',0,'E','m','o','j','i','_','C','o','m','p','o','n','e','n','t',0,
+2,'R','I',0,'R','e','g','i','o','n','a','l','_','I','n','d','i','c','a','t','o','r',0,
+2,'P','C','M',0,'P','r','e','p','e','n','d','e','d','_','C','o','n','c','a','t','e','n','a','t','i','o','n','_','M','a','r',
+'k',0,2,'E','x','t','P','i','c','t',0,'E','x','t','e','n','d','e','d','_','P','i','c','t','o','g','r','a','p','h','i','c',
+0,2,'b','c',0,'B','i','d','i','_','C','l','a','s','s',0,
+2,'L',0,'L','e','f','t','_','T','o','_','R','i','g','h','t',0,
+2,'R',0,'R','i','g','h','t','_','T','o','_','L','e','f','t',0,
+2,'E','N',0,'E','u','r','o','p','e','a','n','_','N','u','m','b','e','r',0,
+2,'E','S',0,'E','u','r','o','p','e','a','n','_','S','e','p','a','r','a','t','o','r',0,
+2,'E','T',0,'E','u','r','o','p','e','a','n','_','T','e','r','m','i','n','a','t','o','r',0,
+2,'A','N',0,'A','r','a','b','i','c','_','N','u','m','b','e','r',0,
+2,'C','S',0,'C','o','m','m','o','n','_','S','e','p','a','r','a','t','o','r',0,
+2,'B',0,'P','a','r','a','g','r','a','p','h','_','S','e','p','a','r','a','t','o','r',0,
+2,'S',0,'S','e','g','m','e','n','t','_','S','e','p','a','r','a','t','o','r',0,
+2,'W','S',0,'W','h','i','t','e','_','S','p','a','c','e',0,
+2,'O','N',0,'O','t','h','e','r','_','N','e','u','t','r','a','l',0,
+2,'L','R','E',0,'L','e','f','t','_','T','o','_','R','i','g','h','t','_','E','m','b','e','d','d','i','n','g',0,
+2,'L','R','O',0,'L','e','f','t','_','T','o','_','R','i','g','h','t','_','O','v','e','r','r','i','d','e',0,
+2,'A','L',0,'A','r','a','b','i','c','_','L','e','t','t','e','r',0,
+2,'R','L','E',0,'R','i','g','h','t','_','T','o','_','L','e','f','t','_','E','m','b','e','d','d','i','n','g',0,
+2,'R','L','O',0,'R','i','g','h','t','_','T','o','_','L','e','f','t','_','O','v','e','r','r','i','d','e',0,
+2,'P','D','F',0,'P','o','p','_','D','i','r','e','c','t','i','o','n','a','l','_','F','o','r','m','a','t',0,
+2,'N','S','M',0,'N','o','n','s','p','a','c','i','n','g','_','M','a','r','k',0,
+2,'B','N',0,'B','o','u','n','d','a','r','y','_','N','e','u','t','r','a','l',0,
+2,'F','S','I',0,'F','i','r','s','t','_','S','t','r','o','n','g','_','I','s','o','l','a','t','e',0,
+2,'L','R','I',0,'L','e','f','t','_','T','o','_','R','i','g','h','t','_','I','s','o','l','a','t','e',0,
+2,'R','L','I',0,'R','i','g','h','t','_','T','o','_','L','e','f','t','_','I','s','o','l','a','t','e',0,
+2,'P','D','I',0,'P','o','p','_','D','i','r','e','c','t','i','o','n','a','l','_','I','s','o','l','a','t','e',0,
+2,'b','l','k',0,'B','l','o','c','k',0,2,'N','B',0,'N','o','_','B','l','o','c','k',0,
+2,'A','S','C','I','I',0,'B','a','s','i','c','_','L','a','t','i','n',0,
+3,'L','a','t','i','n','_','1','_','S','u','p',0,'L','a','t','i','n','_','1','_','S','u','p','p','l','e','m','e','n','t',0,
+'L','a','t','i','n','_','1',0,2,'L','a','t','i','n','_','E','x','t','_','A',0,'L','a','t','i','n','_','E','x','t','e','n',
+'d','e','d','_','A',0,2,'L','a','t','i','n','_','E','x','t','_','B',0,'L','a','t','i','n','_','E','x','t','e','n','d','e',
+'d','_','B',0,2,'I','P','A','_','E','x','t',0,'I','P','A','_','E','x','t','e','n','s','i','o','n','s',0,
+2,'M','o','d','i','f','i','e','r','_','L','e','t','t','e','r','s',0,'S','p','a','c','i','n','g','_','M','o','d','i','f','i',
+'e','r','_','L','e','t','t','e','r','s',0,2,'D','i','a','c','r','i','t','i','c','a','l','s',0,
+'C','o','m','b','i','n','i','n','g','_','D','i','a','c','r','i','t','i','c','a','l','_','M','a','r','k','s',0,
+2,'G','r','e','e','k',0,'G','r','e','e','k','_','A','n','d','_','C','o','p','t','i','c',0,
+2,'C','y','r','i','l','l','i','c',0,'C','y','r','i','l','l','i','c',0,
+2,'A','r','m','e','n','i','a','n',0,'A','r','m','e','n','i','a','n',0,
+2,'H','e','b','r','e','w',0,'H','e','b','r','e','w',0,2,'A','r','a','b','i','c',0,'A','r','a','b','i','c',0,
+2,'S','y','r','i','a','c',0,'S','y','r','i','a','c',0,2,'T','h','a','a','n','a',0,'T','h','a','a','n','a',0,
+2,'D','e','v','a','n','a','g','a','r','i',0,'D','e','v','a','n','a','g','a','r','i',0,
+2,'B','e','n','g','a','l','i',0,'B','e','n','g','a','l','i',0,
+2,'G','u','r','m','u','k','h','i',0,'G','u','r','m','u','k','h','i',0,
+2,'G','u','j','a','r','a','t','i',0,'G','u','j','a','r','a','t','i',0,
+2,'O','r','i','y','a',0,'O','r','i','y','a',0,2,'T','a','m','i','l',0,'T','a','m','i','l',0,
+2,'T','e','l','u','g','u',0,'T','e','l','u','g','u',0,2,'K','a','n','n','a','d','a',0,
+'K','a','n','n','a','d','a',0,2,'M','a','l','a','y','a','l','a','m',0,'M','a','l','a','y','a','l','a','m',0,
+2,'S','i','n','h','a','l','a',0,'S','i','n','h','a','l','a',0,
+2,'T','h','a','i',0,'T','h','a','i',0,2,'L','a','o',0,'L','a','o',0,
+2,'T','i','b','e','t','a','n',0,'T','i','b','e','t','a','n',0,
+2,'M','y','a','n','m','a','r',0,'M','y','a','n','m','a','r',0,
+2,'G','e','o','r','g','i','a','n',0,'G','e','o','r','g','i','a','n',0,
+2,'J','a','m','o',0,'H','a','n','g','u','l','_','J','a','m','o',0,
+2,'E','t','h','i','o','p','i','c',0,'E','t','h','i','o','p','i','c',0,
+2,'C','h','e','r','o','k','e','e',0,'C','h','e','r','o','k','e','e',0,
+3,'U','C','A','S',0,'U','n','i','f','i','e','d','_','C','a','n','a','d','i','a','n','_','A','b','o','r','i','g','i','n','a',
+'l','_','S','y','l','l','a','b','i','c','s',0,'C','a','n','a','d','i','a','n','_','S','y','l','l','a','b','i','c','s',0,
+2,'O','g','h','a','m',0,'O','g','h','a','m',0,2,'R','u','n','i','c',0,'R','u','n','i','c',0,
+2,'K','h','m','e','r',0,'K','h','m','e','r',0,2,'M','o','n','g','o','l','i','a','n',0,
+'M','o','n','g','o','l','i','a','n',0,2,'L','a','t','i','n','_','E','x','t','_','A','d','d','i','t','i','o','n','a','l',0,
+'L','a','t','i','n','_','E','x','t','e','n','d','e','d','_','A','d','d','i','t','i','o','n','a','l',0,
+2,'G','r','e','e','k','_','E','x','t',0,'G','r','e','e','k','_','E','x','t','e','n','d','e','d',0,
+2,'P','u','n','c','t','u','a','t','i','o','n',0,'G','e','n','e','r','a','l','_','P','u','n','c','t','u','a','t','i','o','n',
+0,2,'S','u','p','e','r','_','A','n','d','_','S','u','b',0,'S','u','p','e','r','s','c','r','i','p','t','s','_','A','n','d',
+'_','S','u','b','s','c','r','i','p','t','s',0,2,'C','u','r','r','e','n','c','y','_','S','y','m','b','o','l','s',0,
+'C','u','r','r','e','n','c','y','_','S','y','m','b','o','l','s',0,
+3,'D','i','a','c','r','i','t','i','c','a','l','s','_','F','o','r','_','S','y','m','b','o','l','s',0,
+'C','o','m','b','i','n','i','n','g','_','D','i','a','c','r','i','t','i','c','a','l','_','M','a','r','k','s','_','F','o','r','_',
+'S','y','m','b','o','l','s',0,'C','o','m','b','i','n','i','n','g','_','M','a','r','k','s','_','F','o','r','_','S','y','m','b',
+'o','l','s',0,2,'L','e','t','t','e','r','l','i','k','e','_','S','y','m','b','o','l','s',0,
+'L','e','t','t','e','r','l','i','k','e','_','S','y','m','b','o','l','s',0,
+2,'N','u','m','b','e','r','_','F','o','r','m','s',0,'N','u','m','b','e','r','_','F','o','r','m','s',0,
+2,'A','r','r','o','w','s',0,'A','r','r','o','w','s',0,2,'M','a','t','h','_','O','p','e','r','a','t','o','r','s',0,
+'M','a','t','h','e','m','a','t','i','c','a','l','_','O','p','e','r','a','t','o','r','s',0,
+2,'M','i','s','c','_','T','e','c','h','n','i','c','a','l',0,'M','i','s','c','e','l','l','a','n','e','o','u','s','_','T','e',
+'c','h','n','i','c','a','l',0,2,'C','o','n','t','r','o','l','_','P','i','c','t','u','r','e','s',0,
+'C','o','n','t','r','o','l','_','P','i','c','t','u','r','e','s',0,
+2,'O','C','R',0,'O','p','t','i','c','a','l','_','C','h','a','r','a','c','t','e','r','_','R','e','c','o','g','n','i','t','i',
+'o','n',0,2,'E','n','c','l','o','s','e','d','_','A','l','p','h','a','n','u','m',0,'E','n','c','l','o','s','e','d','_','A',
+'l','p','h','a','n','u','m','e','r','i','c','s',0,2,'B','o','x','_','D','r','a','w','i','n','g',0,
+'B','o','x','_','D','r','a','w','i','n','g',0,2,'B','l','o','c','k','_','E','l','e','m','e','n','t','s',0,
+'B','l','o','c','k','_','E','l','e','m','e','n','t','s',0,2,'G','e','o','m','e','t','r','i','c','_','S','h','a','p','e','s',
+0,'G','e','o','m','e','t','r','i','c','_','S','h','a','p','e','s',0,
+2,'M','i','s','c','_','S','y','m','b','o','l','s',0,'M','i','s','c','e','l','l','a','n','e','o','u','s','_','S','y','m','b',
+'o','l','s',0,2,'D','i','n','g','b','a','t','s',0,'D','i','n','g','b','a','t','s',0,
+2,'B','r','a','i','l','l','e',0,'B','r','a','i','l','l','e','_','P','a','t','t','e','r','n','s',0,
+2,'C','J','K','_','R','a','d','i','c','a','l','s','_','S','u','p',0,'C','J','K','_','R','a','d','i','c','a','l','s','_','S',
+'u','p','p','l','e','m','e','n','t',0,2,'K','a','n','g','x','i',0,'K','a','n','g','x','i','_','R','a','d','i','c','a','l',
+'s',0,2,'I','D','C',0,'I','d','e','o','g','r','a','p','h','i','c','_','D','e','s','c','r','i','p','t','i','o','n','_','C',
+'h','a','r','a','c','t','e','r','s',0,2,'C','J','K','_','S','y','m','b','o','l','s',0,'C','J','K','_','S','y','m','b','o',
+'l','s','_','A','n','d','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'H','i','r','a','g','a','n','a',0,'H','i','r','a','g','a','n','a',0,
+2,'K','a','t','a','k','a','n','a',0,'K','a','t','a','k','a','n','a',0,
+2,'B','o','p','o','m','o','f','o',0,'B','o','p','o','m','o','f','o',0,
+2,'C','o','m','p','a','t','_','J','a','m','o',0,'H','a','n','g','u','l','_','C','o','m','p','a','t','i','b','i','l','i','t',
+'y','_','J','a','m','o',0,2,'K','a','n','b','u','n',0,'K','a','n','b','u','n',0,
+2,'B','o','p','o','m','o','f','o','_','E','x','t',0,'B','o','p','o','m','o','f','o','_','E','x','t','e','n','d','e','d',0,
+2,'E','n','c','l','o','s','e','d','_','C','J','K',0,'E','n','c','l','o','s','e','d','_','C','J','K','_','L','e','t','t','e',
+'r','s','_','A','n','d','_','M','o','n','t','h','s',0,2,'C','J','K','_','C','o','m','p','a','t',0,
+'C','J','K','_','C','o','m','p','a','t','i','b','i','l','i','t','y',0,
+2,'C','J','K','_','E','x','t','_','A',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',
+'s','_','E','x','t','e','n','s','i','o','n','_','A',0,2,'C','J','K',0,'C','J','K','_','U','n','i','f','i','e','d','_','I',
+'d','e','o','g','r','a','p','h','s',0,2,'Y','i','_','S','y','l','l','a','b','l','e','s',0,
+'Y','i','_','S','y','l','l','a','b','l','e','s',0,2,'Y','i','_','R','a','d','i','c','a','l','s',0,
+'Y','i','_','R','a','d','i','c','a','l','s',0,2,'H','a','n','g','u','l',0,'H','a','n','g','u','l','_','S','y','l','l','a',
+'b','l','e','s',0,2,'H','i','g','h','_','S','u','r','r','o','g','a','t','e','s',0,'H','i','g','h','_','S','u','r','r','o',
+'g','a','t','e','s',0,2,'H','i','g','h','_','P','U','_','S','u','r','r','o','g','a','t','e','s',0,
+'H','i','g','h','_','P','r','i','v','a','t','e','_','U','s','e','_','S','u','r','r','o','g','a','t','e','s',0,
+2,'L','o','w','_','S','u','r','r','o','g','a','t','e','s',0,'L','o','w','_','S','u','r','r','o','g','a','t','e','s',0,
+3,'P','U','A',0,'P','r','i','v','a','t','e','_','U','s','e','_','A','r','e','a',0,'P','r','i','v','a','t','e','_','U','s',
+'e',0,2,'C','J','K','_','C','o','m','p','a','t','_','I','d','e','o','g','r','a','p','h','s',0,
+'C','J','K','_','C','o','m','p','a','t','i','b','i','l','i','t','y','_','I','d','e','o','g','r','a','p','h','s',0,
+2,'A','l','p','h','a','b','e','t','i','c','_','P','F',0,'A','l','p','h','a','b','e','t','i','c','_','P','r','e','s','e','n',
+'t','a','t','i','o','n','_','F','o','r','m','s',0,3,'A','r','a','b','i','c','_','P','F','_','A',0,
+'A','r','a','b','i','c','_','P','r','e','s','e','n','t','a','t','i','o','n','_','F','o','r','m','s','_','A',0,
+'A','r','a','b','i','c','_','P','r','e','s','e','n','t','a','t','i','o','n','_','F','o','r','m','s','-','A',0,
+2,'H','a','l','f','_','M','a','r','k','s',0,'C','o','m','b','i','n','i','n','g','_','H','a','l','f','_','M','a','r','k','s',
+0,2,'C','J','K','_','C','o','m','p','a','t','_','F','o','r','m','s',0,'C','J','K','_','C','o','m','p','a','t','i','b','i',
+'l','i','t','y','_','F','o','r','m','s',0,2,'S','m','a','l','l','_','F','o','r','m','s',0,
+'S','m','a','l','l','_','F','o','r','m','_','V','a','r','i','a','n','t','s',0,
+2,'A','r','a','b','i','c','_','P','F','_','B',0,'A','r','a','b','i','c','_','P','r','e','s','e','n','t','a','t','i','o','n',
+'_','F','o','r','m','s','_','B',0,2,'S','p','e','c','i','a','l','s',0,'S','p','e','c','i','a','l','s',0,
+2,'H','a','l','f','_','A','n','d','_','F','u','l','l','_','F','o','r','m','s',0,'H','a','l','f','w','i','d','t','h','_','A',
+'n','d','_','F','u','l','l','w','i','d','t','h','_','F','o','r','m','s',0,
+2,'O','l','d','_','I','t','a','l','i','c',0,'O','l','d','_','I','t','a','l','i','c',0,
+2,'G','o','t','h','i','c',0,'G','o','t','h','i','c',0,2,'D','e','s','e','r','e','t',0,
+'D','e','s','e','r','e','t',0,2,'B','y','z','a','n','t','i','n','e','_','M','u','s','i','c',0,
+'B','y','z','a','n','t','i','n','e','_','M','u','s','i','c','a','l','_','S','y','m','b','o','l','s',0,
+2,'M','u','s','i','c',0,'M','u','s','i','c','a','l','_','S','y','m','b','o','l','s',0,
+2,'M','a','t','h','_','A','l','p','h','a','n','u','m',0,'M','a','t','h','e','m','a','t','i','c','a','l','_','A','l','p','h',
+'a','n','u','m','e','r','i','c','_','S','y','m','b','o','l','s',0,
+2,'C','J','K','_','E','x','t','_','B',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',
+'s','_','E','x','t','e','n','s','i','o','n','_','B',0,2,'C','J','K','_','C','o','m','p','a','t','_','I','d','e','o','g','r',
+'a','p','h','s','_','S','u','p',0,'C','J','K','_','C','o','m','p','a','t','i','b','i','l','i','t','y','_','I','d','e','o','g',
+'r','a','p','h','s','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'T','a','g','s',0,'T','a','g','s',0,3,'C','y','r','i','l','l','i','c','_','S','u','p',0,
+'C','y','r','i','l','l','i','c','_','S','u','p','p','l','e','m','e','n','t',0,'C','y','r','i','l','l','i','c','_','S','u','p',
+'p','l','e','m','e','n','t','a','r','y',0,2,'T','a','g','a','l','o','g',0,'T','a','g','a','l','o','g',0,
+2,'H','a','n','u','n','o','o',0,'H','a','n','u','n','o','o',0,
+2,'B','u','h','i','d',0,'B','u','h','i','d',0,2,'T','a','g','b','a','n','w','a',0,'T','a','g','b','a','n','w','a',0,
+2,'M','i','s','c','_','M','a','t','h','_','S','y','m','b','o','l','s','_','A',0,'M','i','s','c','e','l','l','a','n','e','o',
+'u','s','_','M','a','t','h','e','m','a','t','i','c','a','l','_','S','y','m','b','o','l','s','_','A',0,
+2,'S','u','p','_','A','r','r','o','w','s','_','A',0,'S','u','p','p','l','e','m','e','n','t','a','l','_','A','r','r','o','w',
+'s','_','A',0,2,'S','u','p','_','A','r','r','o','w','s','_','B',0,'S','u','p','p','l','e','m','e','n','t','a','l','_','A',
+'r','r','o','w','s','_','B',0,2,'M','i','s','c','_','M','a','t','h','_','S','y','m','b','o','l','s','_','B',0,
+'M','i','s','c','e','l','l','a','n','e','o','u','s','_','M','a','t','h','e','m','a','t','i','c','a','l','_','S','y','m','b','o',
+'l','s','_','B',0,2,'S','u','p','_','M','a','t','h','_','O','p','e','r','a','t','o','r','s',0,
+'S','u','p','p','l','e','m','e','n','t','a','l','_','M','a','t','h','e','m','a','t','i','c','a','l','_','O','p','e','r','a','t',
+'o','r','s',0,2,'K','a','t','a','k','a','n','a','_','E','x','t',0,'K','a','t','a','k','a','n','a','_','P','h','o','n','e',
+'t','i','c','_','E','x','t','e','n','s','i','o','n','s',0,2,'V','S',0,'V','a','r','i','a','t','i','o','n','_','S','e','l',
+'e','c','t','o','r','s',0,2,'S','u','p','_','P','U','A','_','A',0,'S','u','p','p','l','e','m','e','n','t','a','r','y','_',
+'P','r','i','v','a','t','e','_','U','s','e','_','A','r','e','a','_','A',0,
+2,'S','u','p','_','P','U','A','_','B',0,'S','u','p','p','l','e','m','e','n','t','a','r','y','_','P','r','i','v','a','t','e',
+'_','U','s','e','_','A','r','e','a','_','B',0,2,'L','i','m','b','u',0,'L','i','m','b','u',0,
+2,'T','a','i','_','L','e',0,'T','a','i','_','L','e',0,2,'K','h','m','e','r','_','S','y','m','b','o','l','s',0,
+'K','h','m','e','r','_','S','y','m','b','o','l','s',0,2,'P','h','o','n','e','t','i','c','_','E','x','t',0,
+'P','h','o','n','e','t','i','c','_','E','x','t','e','n','s','i','o','n','s',0,
+2,'M','i','s','c','_','A','r','r','o','w','s',0,'M','i','s','c','e','l','l','a','n','e','o','u','s','_','S','y','m','b','o',
+'l','s','_','A','n','d','_','A','r','r','o','w','s',0,2,'Y','i','j','i','n','g',0,'Y','i','j','i','n','g','_','H','e','x',
+'a','g','r','a','m','_','S','y','m','b','o','l','s',0,2,'L','i','n','e','a','r','_','B','_','S','y','l','l','a','b','a','r',
+'y',0,'L','i','n','e','a','r','_','B','_','S','y','l','l','a','b','a','r','y',0,
+2,'L','i','n','e','a','r','_','B','_','I','d','e','o','g','r','a','m','s',0,'L','i','n','e','a','r','_','B','_','I','d','e',
+'o','g','r','a','m','s',0,2,'A','e','g','e','a','n','_','N','u','m','b','e','r','s',0,'A','e','g','e','a','n','_','N','u',
+'m','b','e','r','s',0,2,'U','g','a','r','i','t','i','c',0,'U','g','a','r','i','t','i','c',0,
+2,'S','h','a','v','i','a','n',0,'S','h','a','v','i','a','n',0,
+2,'O','s','m','a','n','y','a',0,'O','s','m','a','n','y','a',0,
+2,'C','y','p','r','i','o','t','_','S','y','l','l','a','b','a','r','y',0,'C','y','p','r','i','o','t','_','S','y','l','l','a',
+'b','a','r','y',0,2,'T','a','i','_','X','u','a','n','_','J','i','n','g',0,'T','a','i','_','X','u','a','n','_','J','i','n',
+'g','_','S','y','m','b','o','l','s',0,2,'V','S','_','S','u','p',0,'V','a','r','i','a','t','i','o','n','_','S','e','l','e',
+'c','t','o','r','s','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'A','n','c','i','e','n','t','_','G','r','e','e','k','_','M','u','s','i','c',0,'A','n','c','i','e','n','t','_','G','r','e',
+'e','k','_','M','u','s','i','c','a','l','_','N','o','t','a','t','i','o','n',0,
+2,'A','n','c','i','e','n','t','_','G','r','e','e','k','_','N','u','m','b','e','r','s',0,'A','n','c','i','e','n','t','_','G',
+'r','e','e','k','_','N','u','m','b','e','r','s',0,2,'A','r','a','b','i','c','_','S','u','p',0,
+'A','r','a','b','i','c','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'B','u','g','i','n','e','s','e',0,'B','u','g','i','n','e','s','e',0,
+2,'C','J','K','_','S','t','r','o','k','e','s',0,'C','J','K','_','S','t','r','o','k','e','s',0,
+2,'D','i','a','c','r','i','t','i','c','a','l','s','_','S','u','p',0,'C','o','m','b','i','n','i','n','g','_','D','i','a','c',
+'r','i','t','i','c','a','l','_','M','a','r','k','s','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'C','o','p','t','i','c',0,'C','o','p','t','i','c',0,2,'E','t','h','i','o','p','i','c','_','E','x','t',0,
+'E','t','h','i','o','p','i','c','_','E','x','t','e','n','d','e','d',0,
+2,'E','t','h','i','o','p','i','c','_','S','u','p',0,'E','t','h','i','o','p','i','c','_','S','u','p','p','l','e','m','e','n',
+'t',0,2,'G','e','o','r','g','i','a','n','_','S','u','p',0,'G','e','o','r','g','i','a','n','_','S','u','p','p','l','e','m',
+'e','n','t',0,2,'G','l','a','g','o','l','i','t','i','c',0,'G','l','a','g','o','l','i','t','i','c',0,
+2,'K','h','a','r','o','s','h','t','h','i',0,'K','h','a','r','o','s','h','t','h','i',0,
+2,'M','o','d','i','f','i','e','r','_','T','o','n','e','_','L','e','t','t','e','r','s',0,'M','o','d','i','f','i','e','r','_',
+'T','o','n','e','_','L','e','t','t','e','r','s',0,2,'N','e','w','_','T','a','i','_','L','u','e',0,
+'N','e','w','_','T','a','i','_','L','u','e',0,2,'O','l','d','_','P','e','r','s','i','a','n',0,
+'O','l','d','_','P','e','r','s','i','a','n',0,2,'P','h','o','n','e','t','i','c','_','E','x','t','_','S','u','p',0,
+'P','h','o','n','e','t','i','c','_','E','x','t','e','n','s','i','o','n','s','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'S','u','p','_','P','u','n','c','t','u','a','t','i','o','n',0,'S','u','p','p','l','e','m','e','n','t','a','l','_','P','u',
+'n','c','t','u','a','t','i','o','n',0,2,'S','y','l','o','t','i','_','N','a','g','r','i',0,
+'S','y','l','o','t','i','_','N','a','g','r','i',0,2,'T','i','f','i','n','a','g','h',0,'T','i','f','i','n','a','g','h',0,
+2,'V','e','r','t','i','c','a','l','_','F','o','r','m','s',0,'V','e','r','t','i','c','a','l','_','F','o','r','m','s',0,
+2,'N','K','o',0,'N','K','o',0,2,'B','a','l','i','n','e','s','e',0,'B','a','l','i','n','e','s','e',0,
+2,'L','a','t','i','n','_','E','x','t','_','C',0,'L','a','t','i','n','_','E','x','t','e','n','d','e','d','_','C',0,
+2,'L','a','t','i','n','_','E','x','t','_','D',0,'L','a','t','i','n','_','E','x','t','e','n','d','e','d','_','D',0,
+2,'P','h','a','g','s','_','P','a',0,'P','h','a','g','s','_','P','a',0,
+2,'P','h','o','e','n','i','c','i','a','n',0,'P','h','o','e','n','i','c','i','a','n',0,
+2,'C','u','n','e','i','f','o','r','m',0,'C','u','n','e','i','f','o','r','m',0,
+2,'C','u','n','e','i','f','o','r','m','_','N','u','m','b','e','r','s',0,'C','u','n','e','i','f','o','r','m','_','N','u','m',
+'b','e','r','s','_','A','n','d','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'C','o','u','n','t','i','n','g','_','R','o','d',0,'C','o','u','n','t','i','n','g','_','R','o','d','_','N','u','m','e','r',
+'a','l','s',0,2,'S','u','n','d','a','n','e','s','e',0,'S','u','n','d','a','n','e','s','e',0,
+2,'L','e','p','c','h','a',0,'L','e','p','c','h','a',0,2,'O','l','_','C','h','i','k','i',0,
+'O','l','_','C','h','i','k','i',0,2,'C','y','r','i','l','l','i','c','_','E','x','t','_','A',0,
+'C','y','r','i','l','l','i','c','_','E','x','t','e','n','d','e','d','_','A',0,
+2,'V','a','i',0,'V','a','i',0,2,'C','y','r','i','l','l','i','c','_','E','x','t','_','B',0,
+'C','y','r','i','l','l','i','c','_','E','x','t','e','n','d','e','d','_','B',0,
+2,'S','a','u','r','a','s','h','t','r','a',0,'S','a','u','r','a','s','h','t','r','a',0,
+2,'K','a','y','a','h','_','L','i',0,'K','a','y','a','h','_','L','i',0,
+2,'R','e','j','a','n','g',0,'R','e','j','a','n','g',0,2,'C','h','a','m',0,'C','h','a','m',0,
+2,'A','n','c','i','e','n','t','_','S','y','m','b','o','l','s',0,'A','n','c','i','e','n','t','_','S','y','m','b','o','l','s',
+0,2,'P','h','a','i','s','t','o','s',0,'P','h','a','i','s','t','o','s','_','D','i','s','c',0,
+2,'L','y','c','i','a','n',0,'L','y','c','i','a','n',0,2,'C','a','r','i','a','n',0,'C','a','r','i','a','n',0,
+2,'L','y','d','i','a','n',0,'L','y','d','i','a','n',0,2,'M','a','h','j','o','n','g',0,
+'M','a','h','j','o','n','g','_','T','i','l','e','s',0,2,'D','o','m','i','n','o',0,'D','o','m','i','n','o','_','T','i','l',
+'e','s',0,2,'S','a','m','a','r','i','t','a','n',0,'S','a','m','a','r','i','t','a','n',0,
+2,'U','C','A','S','_','E','x','t',0,'U','n','i','f','i','e','d','_','C','a','n','a','d','i','a','n','_','A','b','o','r','i',
+'g','i','n','a','l','_','S','y','l','l','a','b','i','c','s','_','E','x','t','e','n','d','e','d',0,
+2,'T','a','i','_','T','h','a','m',0,'T','a','i','_','T','h','a','m',0,
+2,'V','e','d','i','c','_','E','x','t',0,'V','e','d','i','c','_','E','x','t','e','n','s','i','o','n','s',0,
+2,'L','i','s','u',0,'L','i','s','u',0,2,'B','a','m','u','m',0,'B','a','m','u','m',0,
+2,'I','n','d','i','c','_','N','u','m','b','e','r','_','F','o','r','m','s',0,'C','o','m','m','o','n','_','I','n','d','i','c',
+'_','N','u','m','b','e','r','_','F','o','r','m','s',0,2,'D','e','v','a','n','a','g','a','r','i','_','E','x','t',0,
+'D','e','v','a','n','a','g','a','r','i','_','E','x','t','e','n','d','e','d',0,
+2,'J','a','m','o','_','E','x','t','_','A',0,'H','a','n','g','u','l','_','J','a','m','o','_','E','x','t','e','n','d','e','d',
+'_','A',0,2,'J','a','v','a','n','e','s','e',0,'J','a','v','a','n','e','s','e',0,
+2,'M','y','a','n','m','a','r','_','E','x','t','_','A',0,'M','y','a','n','m','a','r','_','E','x','t','e','n','d','e','d','_',
+'A',0,2,'T','a','i','_','V','i','e','t',0,'T','a','i','_','V','i','e','t',0,
+2,'M','e','e','t','e','i','_','M','a','y','e','k',0,'M','e','e','t','e','i','_','M','a','y','e','k',0,
+2,'J','a','m','o','_','E','x','t','_','B',0,'H','a','n','g','u','l','_','J','a','m','o','_','E','x','t','e','n','d','e','d',
+'_','B',0,2,'I','m','p','e','r','i','a','l','_','A','r','a','m','a','i','c',0,'I','m','p','e','r','i','a','l','_','A','r',
+'a','m','a','i','c',0,2,'O','l','d','_','S','o','u','t','h','_','A','r','a','b','i','a','n',0,
+'O','l','d','_','S','o','u','t','h','_','A','r','a','b','i','a','n',0,
+2,'A','v','e','s','t','a','n',0,'A','v','e','s','t','a','n',0,
+2,'I','n','s','c','r','i','p','t','i','o','n','a','l','_','P','a','r','t','h','i','a','n',0,
+'I','n','s','c','r','i','p','t','i','o','n','a','l','_','P','a','r','t','h','i','a','n',0,
+2,'I','n','s','c','r','i','p','t','i','o','n','a','l','_','P','a','h','l','a','v','i',0,'I','n','s','c','r','i','p','t','i',
+'o','n','a','l','_','P','a','h','l','a','v','i',0,2,'O','l','d','_','T','u','r','k','i','c',0,
+'O','l','d','_','T','u','r','k','i','c',0,2,'R','u','m','i',0,'R','u','m','i','_','N','u','m','e','r','a','l','_','S','y',
+'m','b','o','l','s',0,2,'K','a','i','t','h','i',0,'K','a','i','t','h','i',0,
+2,'E','g','y','p','t','i','a','n','_','H','i','e','r','o','g','l','y','p','h','s',0,'E','g','y','p','t','i','a','n','_','H',
+'i','e','r','o','g','l','y','p','h','s',0,2,'E','n','c','l','o','s','e','d','_','A','l','p','h','a','n','u','m','_','S','u',
+'p',0,'E','n','c','l','o','s','e','d','_','A','l','p','h','a','n','u','m','e','r','i','c','_','S','u','p','p','l','e','m','e',
+'n','t',0,2,'E','n','c','l','o','s','e','d','_','I','d','e','o','g','r','a','p','h','i','c','_','S','u','p',0,
+'E','n','c','l','o','s','e','d','_','I','d','e','o','g','r','a','p','h','i','c','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'C','J','K','_','E','x','t','_','C',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',
+'s','_','E','x','t','e','n','s','i','o','n','_','C',0,2,'M','a','n','d','a','i','c',0,'M','a','n','d','a','i','c',0,
+2,'B','a','t','a','k',0,'B','a','t','a','k',0,2,'E','t','h','i','o','p','i','c','_','E','x','t','_','A',0,
+'E','t','h','i','o','p','i','c','_','E','x','t','e','n','d','e','d','_','A',0,
+2,'B','r','a','h','m','i',0,'B','r','a','h','m','i',0,2,'B','a','m','u','m','_','S','u','p',0,
+'B','a','m','u','m','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'K','a','n','a','_','S','u','p',0,'K','a','n','a','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'P','l','a','y','i','n','g','_','C','a','r','d','s',0,'P','l','a','y','i','n','g','_','C','a','r','d','s',0,
+2,'M','i','s','c','_','P','i','c','t','o','g','r','a','p','h','s',0,'M','i','s','c','e','l','l','a','n','e','o','u','s','_',
+'S','y','m','b','o','l','s','_','A','n','d','_','P','i','c','t','o','g','r','a','p','h','s',0,
+2,'E','m','o','t','i','c','o','n','s',0,'E','m','o','t','i','c','o','n','s',0,
+2,'T','r','a','n','s','p','o','r','t','_','A','n','d','_','M','a','p',0,'T','r','a','n','s','p','o','r','t','_','A','n','d',
+'_','M','a','p','_','S','y','m','b','o','l','s',0,2,'A','l','c','h','e','m','i','c','a','l',0,
+'A','l','c','h','e','m','i','c','a','l','_','S','y','m','b','o','l','s',0,
+2,'C','J','K','_','E','x','t','_','D',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',
+'s','_','E','x','t','e','n','s','i','o','n','_','D',0,2,'A','r','a','b','i','c','_','E','x','t','_','A',0,
+'A','r','a','b','i','c','_','E','x','t','e','n','d','e','d','_','A',0,
+2,'A','r','a','b','i','c','_','M','a','t','h',0,'A','r','a','b','i','c','_','M','a','t','h','e','m','a','t','i','c','a','l',
+'_','A','l','p','h','a','b','e','t','i','c','_','S','y','m','b','o','l','s',0,
+2,'C','h','a','k','m','a',0,'C','h','a','k','m','a',0,2,'M','e','e','t','e','i','_','M','a','y','e','k','_','E','x','t',
+0,'M','e','e','t','e','i','_','M','a','y','e','k','_','E','x','t','e','n','s','i','o','n','s',0,
+2,'M','e','r','o','i','t','i','c','_','C','u','r','s','i','v','e',0,'M','e','r','o','i','t','i','c','_','C','u','r','s','i',
+'v','e',0,2,'M','e','r','o','i','t','i','c','_','H','i','e','r','o','g','l','y','p','h','s',0,
+'M','e','r','o','i','t','i','c','_','H','i','e','r','o','g','l','y','p','h','s',0,
+2,'M','i','a','o',0,'M','i','a','o',0,2,'S','h','a','r','a','d','a',0,'S','h','a','r','a','d','a',0,
+2,'S','o','r','a','_','S','o','m','p','e','n','g',0,'S','o','r','a','_','S','o','m','p','e','n','g',0,
+2,'S','u','n','d','a','n','e','s','e','_','S','u','p',0,'S','u','n','d','a','n','e','s','e','_','S','u','p','p','l','e','m',
+'e','n','t',0,2,'T','a','k','r','i',0,'T','a','k','r','i',0,
+2,'B','a','s','s','a','_','V','a','h',0,'B','a','s','s','a','_','V','a','h',0,
+2,'C','a','u','c','a','s','i','a','n','_','A','l','b','a','n','i','a','n',0,'C','a','u','c','a','s','i','a','n','_','A','l',
+'b','a','n','i','a','n',0,2,'C','o','p','t','i','c','_','E','p','a','c','t','_','N','u','m','b','e','r','s',0,
+'C','o','p','t','i','c','_','E','p','a','c','t','_','N','u','m','b','e','r','s',0,
+2,'D','i','a','c','r','i','t','i','c','a','l','s','_','E','x','t',0,'C','o','m','b','i','n','i','n','g','_','D','i','a','c',
+'r','i','t','i','c','a','l','_','M','a','r','k','s','_','E','x','t','e','n','d','e','d',0,
+2,'D','u','p','l','o','y','a','n',0,'D','u','p','l','o','y','a','n',0,
+2,'E','l','b','a','s','a','n',0,'E','l','b','a','s','a','n',0,
+2,'G','e','o','m','e','t','r','i','c','_','S','h','a','p','e','s','_','E','x','t',0,'G','e','o','m','e','t','r','i','c','_',
+'S','h','a','p','e','s','_','E','x','t','e','n','d','e','d',0,
+2,'G','r','a','n','t','h','a',0,'G','r','a','n','t','h','a',0,
+2,'K','h','o','j','k','i',0,'K','h','o','j','k','i',0,2,'K','h','u','d','a','w','a','d','i',0,
+'K','h','u','d','a','w','a','d','i',0,2,'L','a','t','i','n','_','E','x','t','_','E',0,'L','a','t','i','n','_','E','x','t',
+'e','n','d','e','d','_','E',0,2,'L','i','n','e','a','r','_','A',0,'L','i','n','e','a','r','_','A',0,
+2,'M','a','h','a','j','a','n','i',0,'M','a','h','a','j','a','n','i',0,
+2,'M','a','n','i','c','h','a','e','a','n',0,'M','a','n','i','c','h','a','e','a','n',0,
+2,'M','e','n','d','e','_','K','i','k','a','k','u','i',0,'M','e','n','d','e','_','K','i','k','a','k','u','i',0,
+2,'M','o','d','i',0,'M','o','d','i',0,2,'M','r','o',0,'M','r','o',0,
+2,'M','y','a','n','m','a','r','_','E','x','t','_','B',0,'M','y','a','n','m','a','r','_','E','x','t','e','n','d','e','d','_',
+'B',0,2,'N','a','b','a','t','a','e','a','n',0,'N','a','b','a','t','a','e','a','n',0,
+2,'O','l','d','_','N','o','r','t','h','_','A','r','a','b','i','a','n',0,'O','l','d','_','N','o','r','t','h','_','A','r','a',
+'b','i','a','n',0,2,'O','l','d','_','P','e','r','m','i','c',0,'O','l','d','_','P','e','r','m','i','c',0,
+2,'O','r','n','a','m','e','n','t','a','l','_','D','i','n','g','b','a','t','s',0,'O','r','n','a','m','e','n','t','a','l','_',
+'D','i','n','g','b','a','t','s',0,2,'P','a','h','a','w','h','_','H','m','o','n','g',0,'P','a','h','a','w','h','_','H','m',
+'o','n','g',0,2,'P','a','l','m','y','r','e','n','e',0,'P','a','l','m','y','r','e','n','e',0,
+2,'P','a','u','_','C','i','n','_','H','a','u',0,'P','a','u','_','C','i','n','_','H','a','u',0,
+2,'P','s','a','l','t','e','r','_','P','a','h','l','a','v','i',0,'P','s','a','l','t','e','r','_','P','a','h','l','a','v','i',
+0,2,'S','h','o','r','t','h','a','n','d','_','F','o','r','m','a','t','_','C','o','n','t','r','o','l','s',0,
+'S','h','o','r','t','h','a','n','d','_','F','o','r','m','a','t','_','C','o','n','t','r','o','l','s',0,
+2,'S','i','d','d','h','a','m',0,'S','i','d','d','h','a','m',0,
+2,'S','i','n','h','a','l','a','_','A','r','c','h','a','i','c','_','N','u','m','b','e','r','s',0,
+'S','i','n','h','a','l','a','_','A','r','c','h','a','i','c','_','N','u','m','b','e','r','s',0,
+2,'S','u','p','_','A','r','r','o','w','s','_','C',0,'S','u','p','p','l','e','m','e','n','t','a','l','_','A','r','r','o','w',
+'s','_','C',0,2,'T','i','r','h','u','t','a',0,'T','i','r','h','u','t','a',0,
+2,'W','a','r','a','n','g','_','C','i','t','i',0,'W','a','r','a','n','g','_','C','i','t','i',0,
+2,'A','h','o','m',0,'A','h','o','m',0,2,'A','n','a','t','o','l','i','a','n','_','H','i','e','r','o','g','l','y','p','h',
+'s',0,'A','n','a','t','o','l','i','a','n','_','H','i','e','r','o','g','l','y','p','h','s',0,
+2,'C','h','e','r','o','k','e','e','_','S','u','p',0,'C','h','e','r','o','k','e','e','_','S','u','p','p','l','e','m','e','n',
+'t',0,2,'C','J','K','_','E','x','t','_','E',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a',
+'p','h','s','_','E','x','t','e','n','s','i','o','n','_','E',0,
+2,'E','a','r','l','y','_','D','y','n','a','s','t','i','c','_','C','u','n','e','i','f','o','r','m',0,
+'E','a','r','l','y','_','D','y','n','a','s','t','i','c','_','C','u','n','e','i','f','o','r','m',0,
+2,'H','a','t','r','a','n',0,'H','a','t','r','a','n',0,2,'M','u','l','t','a','n','i',0,
+'M','u','l','t','a','n','i',0,2,'O','l','d','_','H','u','n','g','a','r','i','a','n',0,'O','l','d','_','H','u','n','g','a',
+'r','i','a','n',0,2,'S','u','p','_','S','y','m','b','o','l','s','_','A','n','d','_','P','i','c','t','o','g','r','a','p','h',
+'s',0,'S','u','p','p','l','e','m','e','n','t','a','l','_','S','y','m','b','o','l','s','_','A','n','d','_','P','i','c','t','o',
+'g','r','a','p','h','s',0,2,'S','u','t','t','o','n','_','S','i','g','n','W','r','i','t','i','n','g',0,
+'S','u','t','t','o','n','_','S','i','g','n','W','r','i','t','i','n','g',0,
+2,'A','d','l','a','m',0,'A','d','l','a','m',0,2,'B','h','a','i','k','s','u','k','i',0,
+'B','h','a','i','k','s','u','k','i',0,2,'C','y','r','i','l','l','i','c','_','E','x','t','_','C',0,
+'C','y','r','i','l','l','i','c','_','E','x','t','e','n','d','e','d','_','C',0,
+2,'G','l','a','g','o','l','i','t','i','c','_','S','u','p',0,'G','l','a','g','o','l','i','t','i','c','_','S','u','p','p','l',
+'e','m','e','n','t',0,2,'I','d','e','o','g','r','a','p','h','i','c','_','S','y','m','b','o','l','s',0,
+'I','d','e','o','g','r','a','p','h','i','c','_','S','y','m','b','o','l','s','_','A','n','d','_','P','u','n','c','t','u','a','t',
+'i','o','n',0,2,'M','a','r','c','h','e','n',0,'M','a','r','c','h','e','n',0,
+2,'M','o','n','g','o','l','i','a','n','_','S','u','p',0,'M','o','n','g','o','l','i','a','n','_','S','u','p','p','l','e','m',
+'e','n','t',0,2,'N','e','w','a',0,'N','e','w','a',0,2,'O','s','a','g','e',0,'O','s','a','g','e',0,
+2,'T','a','n','g','u','t',0,'T','a','n','g','u','t',0,2,'T','a','n','g','u','t','_','C','o','m','p','o','n','e','n','t',
+'s',0,'T','a','n','g','u','t','_','C','o','m','p','o','n','e','n','t','s',0,
+2,'C','J','K','_','E','x','t','_','F',0,'C','J','K','_','U','n','i','f','i','e','d','_','I','d','e','o','g','r','a','p','h',
+'s','_','E','x','t','e','n','s','i','o','n','_','F',0,2,'K','a','n','a','_','E','x','t','_','A',0,
+'K','a','n','a','_','E','x','t','e','n','d','e','d','_','A',0,
+2,'M','a','s','a','r','a','m','_','G','o','n','d','i',0,'M','a','s','a','r','a','m','_','G','o','n','d','i',0,
+2,'N','u','s','h','u',0,'N','u','s','h','u',0,2,'S','o','y','o','m','b','o',0,'S','o','y','o','m','b','o',0,
+2,'S','y','r','i','a','c','_','S','u','p',0,'S','y','r','i','a','c','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'Z','a','n','a','b','a','z','a','r','_','S','q','u','a','r','e',0,'Z','a','n','a','b','a','z','a','r','_','S','q','u','a',
+'r','e',0,2,'C','h','e','s','s','_','S','y','m','b','o','l','s',0,'C','h','e','s','s','_','S','y','m','b','o','l','s',0,
+2,'D','o','g','r','a',0,'D','o','g','r','a',0,2,'G','e','o','r','g','i','a','n','_','E','x','t',0,
+'G','e','o','r','g','i','a','n','_','E','x','t','e','n','d','e','d',0,
+2,'G','u','n','j','a','l','a','_','G','o','n','d','i',0,'G','u','n','j','a','l','a','_','G','o','n','d','i',0,
+2,'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a',0,'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a',
+0,2,'I','n','d','i','c','_','S','i','y','a','q','_','N','u','m','b','e','r','s',0,'I','n','d','i','c','_','S','i','y','a',
+'q','_','N','u','m','b','e','r','s',0,2,'M','a','k','a','s','a','r',0,'M','a','k','a','s','a','r',0,
+2,'M','a','y','a','n','_','N','u','m','e','r','a','l','s',0,'M','a','y','a','n','_','N','u','m','e','r','a','l','s',0,
+2,'M','e','d','e','f','a','i','d','r','i','n',0,'M','e','d','e','f','a','i','d','r','i','n',0,
+2,'O','l','d','_','S','o','g','d','i','a','n',0,'O','l','d','_','S','o','g','d','i','a','n',0,
+2,'S','o','g','d','i','a','n',0,'S','o','g','d','i','a','n',0,
+2,'E','g','y','p','t','i','a','n','_','H','i','e','r','o','g','l','y','p','h','_','F','o','r','m','a','t','_','C','o','n','t',
+'r','o','l','s',0,'E','g','y','p','t','i','a','n','_','H','i','e','r','o','g','l','y','p','h','_','F','o','r','m','a','t','_',
+'C','o','n','t','r','o','l','s',0,2,'E','l','y','m','a','i','c',0,'E','l','y','m','a','i','c',0,
+2,'N','a','n','d','i','n','a','g','a','r','i',0,'N','a','n','d','i','n','a','g','a','r','i',0,
+2,'N','y','i','a','k','e','n','g','_','P','u','a','c','h','u','e','_','H','m','o','n','g',0,
+'N','y','i','a','k','e','n','g','_','P','u','a','c','h','u','e','_','H','m','o','n','g',0,
+2,'O','t','t','o','m','a','n','_','S','i','y','a','q','_','N','u','m','b','e','r','s',0,'O','t','t','o','m','a','n','_','S',
+'i','y','a','q','_','N','u','m','b','e','r','s',0,2,'S','m','a','l','l','_','K','a','n','a','_','E','x','t',0,
+'S','m','a','l','l','_','K','a','n','a','_','E','x','t','e','n','s','i','o','n',0,
+2,'S','y','m','b','o','l','s','_','A','n','d','_','P','i','c','t','o','g','r','a','p','h','s','_','E','x','t','_','A',0,
+'S','y','m','b','o','l','s','_','A','n','d','_','P','i','c','t','o','g','r','a','p','h','s','_','E','x','t','e','n','d','e','d',
+'_','A',0,2,'T','a','m','i','l','_','S','u','p',0,'T','a','m','i','l','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'W','a','n','c','h','o',0,'W','a','n','c','h','o',0,2,'C','h','o','r','a','s','m','i','a','n',0,
+'C','h','o','r','a','s','m','i','a','n',0,2,'C','J','K','_','E','x','t','_','G',0,'C','J','K','_','U','n','i','f','i','e',
+'d','_','I','d','e','o','g','r','a','p','h','s','_','E','x','t','e','n','s','i','o','n','_','G',0,
+2,'D','i','v','e','s','_','A','k','u','r','u',0,'D','i','v','e','s','_','A','k','u','r','u',0,
+2,'K','h','i','t','a','n','_','S','m','a','l','l','_','S','c','r','i','p','t',0,'K','h','i','t','a','n','_','S','m','a','l',
+'l','_','S','c','r','i','p','t',0,2,'L','i','s','u','_','S','u','p',0,'L','i','s','u','_','S','u','p','p','l','e','m','e',
+'n','t',0,2,'S','y','m','b','o','l','s','_','F','o','r','_','L','e','g','a','c','y','_','C','o','m','p','u','t','i','n','g',
+0,'S','y','m','b','o','l','s','_','F','o','r','_','L','e','g','a','c','y','_','C','o','m','p','u','t','i','n','g',0,
+2,'T','a','n','g','u','t','_','S','u','p',0,'T','a','n','g','u','t','_','S','u','p','p','l','e','m','e','n','t',0,
+2,'Y','e','z','i','d','i',0,'Y','e','z','i','d','i',0,2,'c','c','c',0,'C','a','n','o','n','i','c','a','l','_','C','o',
+'m','b','i','n','i','n','g','_','C','l','a','s','s',0,2,'d','t',0,'D','e','c','o','m','p','o','s','i','t','i','o','n','_',
+'T','y','p','e',0,3,'N','o','n','e',0,'N','o','n','e',0,'n','o','n','e',0,
+3,'C','a','n',0,'C','a','n','o','n','i','c','a','l',0,'c','a','n',0,
+3,'C','o','m',0,'C','o','m','p','a','t',0,'c','o','m',0,
+3,'E','n','c',0,'C','i','r','c','l','e',0,'e','n','c',0,
+3,'F','i','n',0,'F','i','n','a','l',0,'f','i','n',0,3,'F','o','n','t',0,'F','o','n','t',0,
+'f','o','n','t',0,3,'F','r','a',0,'F','r','a','c','t','i','o','n',0,'f','r','a',0,
+3,'I','n','i','t',0,'I','n','i','t','i','a','l',0,'i','n','i','t',0,
+3,'I','s','o',0,'I','s','o','l','a','t','e','d',0,'i','s','o',0,
+3,'M','e','d',0,'M','e','d','i','a','l',0,'m','e','d',0,
+3,'N','a','r',0,'N','a','r','r','o','w',0,'n','a','r',0,
+3,'N','b',0,'N','o','b','r','e','a','k',0,'n','b',0,3,'S','m','l',0,'S','m','a','l','l',0,
+'s','m','l',0,3,'S','q','r',0,'S','q','u','a','r','e',0,'s','q','r',0,
+3,'S','u','b',0,'S','u','b',0,'s','u','b',0,3,'S','u','p',0,'S','u','p','e','r',0,
+'s','u','p',0,3,'V','e','r','t',0,'V','e','r','t','i','c','a','l',0,'v','e','r','t',0,
+3,'W','i','d','e',0,'W','i','d','e',0,'w','i','d','e',0,
+2,'e','a',0,'E','a','s','t','_','A','s','i','a','n','_','W','i','d','t','h',0,
+2,'N',0,'N','e','u','t','r','a','l',0,2,'A',0,'A','m','b','i','g','u','o','u','s',0,
+2,'H',0,'H','a','l','f','w','i','d','t','h',0,2,'F',0,'F','u','l','l','w','i','d','t','h',0,
+2,'N','a',0,'N','a','r','r','o','w',0,2,'W',0,'W','i','d','e',0,
+2,'g','c',0,'G','e','n','e','r','a','l','_','C','a','t','e','g','o','r','y',0,
+2,'C','n',0,'U','n','a','s','s','i','g','n','e','d',0,2,'L','u',0,'U','p','p','e','r','c','a','s','e','_','L','e','t',
+'t','e','r',0,2,'L','l',0,'L','o','w','e','r','c','a','s','e','_','L','e','t','t','e','r',0,
+2,'L','t',0,'T','i','t','l','e','c','a','s','e','_','L','e','t','t','e','r',0,
+2,'L','m',0,'M','o','d','i','f','i','e','r','_','L','e','t','t','e','r',0,
+2,'L','o',0,'O','t','h','e','r','_','L','e','t','t','e','r',0,
+2,'M','n',0,'N','o','n','s','p','a','c','i','n','g','_','M','a','r','k',0,
+2,'M','e',0,'E','n','c','l','o','s','i','n','g','_','M','a','r','k',0,
+2,'M','c',0,'S','p','a','c','i','n','g','_','M','a','r','k',0,
+3,'N','d',0,'D','e','c','i','m','a','l','_','N','u','m','b','e','r',0,'d','i','g','i','t',0,
+2,'N','l',0,'L','e','t','t','e','r','_','N','u','m','b','e','r',0,
+2,'N','o',0,'O','t','h','e','r','_','N','u','m','b','e','r',0,
+2,'Z','s',0,'S','p','a','c','e','_','S','e','p','a','r','a','t','o','r',0,
+2,'Z','l',0,'L','i','n','e','_','S','e','p','a','r','a','t','o','r',0,
+2,'Z','p',0,'P','a','r','a','g','r','a','p','h','_','S','e','p','a','r','a','t','o','r',0,
+3,'C','c',0,'C','o','n','t','r','o','l',0,'c','n','t','r','l',0,
+2,'C','f',0,'F','o','r','m','a','t',0,2,'C','o',0,'P','r','i','v','a','t','e','_','U','s','e',0,
+2,'C','s',0,'S','u','r','r','o','g','a','t','e',0,2,'P','d',0,'D','a','s','h','_','P','u','n','c','t','u','a','t','i',
+'o','n',0,2,'P','s',0,'O','p','e','n','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'P','e',0,'C','l','o','s','e','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'P','c',0,'C','o','n','n','e','c','t','o','r','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'P','o',0,'O','t','h','e','r','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'S','m',0,'M','a','t','h','_','S','y','m','b','o','l',0,
+2,'S','c',0,'C','u','r','r','e','n','c','y','_','S','y','m','b','o','l',0,
+2,'S','k',0,'M','o','d','i','f','i','e','r','_','S','y','m','b','o','l',0,
+2,'S','o',0,'O','t','h','e','r','_','S','y','m','b','o','l',0,
+2,'P','i',0,'I','n','i','t','i','a','l','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'P','f',0,'F','i','n','a','l','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'j','g',0,'J','o','i','n','i','n','g','_','G','r','o','u','p',0,
+2,'N','o','_','J','o','i','n','i','n','g','_','G','r','o','u','p',0,'N','o','_','J','o','i','n','i','n','g','_','G','r','o',
+'u','p',0,2,'A','i','n',0,'A','i','n',0,2,'A','l','a','p','h',0,'A','l','a','p','h',0,
+2,'A','l','e','f',0,'A','l','e','f',0,2,'B','e','h',0,'B','e','h',0,
+2,'B','e','t','h',0,'B','e','t','h',0,2,'D','a','l',0,'D','a','l',0,
+2,'D','a','l','a','t','h','_','R','i','s','h',0,'D','a','l','a','t','h','_','R','i','s','h',0,
+2,'E',0,'E',0,2,'F','e','h',0,'F','e','h',0,2,'F','i','n','a','l','_','S','e','m','k','a','t','h',0,
+'F','i','n','a','l','_','S','e','m','k','a','t','h',0,2,'G','a','f',0,'G','a','f',0,
+2,'G','a','m','a','l',0,'G','a','m','a','l',0,2,'H','a','h',0,'H','a','h',0,
+2,'T','e','h','_','M','a','r','b','u','t','a','_','G','o','a','l',0,'H','a','m','z','a','_','O','n','_','H','e','h','_','G',
+'o','a','l',0,2,'H','e',0,'H','e',0,2,'H','e','h',0,'H','e','h',0,
+2,'H','e','h','_','G','o','a','l',0,'H','e','h','_','G','o','a','l',0,
+2,'H','e','t','h',0,'H','e','t','h',0,2,'K','a','f',0,'K','a','f',0,
+2,'K','a','p','h',0,'K','a','p','h',0,2,'K','n','o','t','t','e','d','_','H','e','h',0,
+'K','n','o','t','t','e','d','_','H','e','h',0,2,'L','a','m',0,'L','a','m',0,
+2,'L','a','m','a','d','h',0,'L','a','m','a','d','h',0,2,'M','e','e','m',0,'M','e','e','m',0,
+2,'M','i','m',0,'M','i','m',0,2,'N','o','o','n',0,'N','o','o','n',0,
+2,'N','u','n',0,'N','u','n',0,2,'P','e',0,'P','e',0,
+2,'Q','a','f',0,'Q','a','f',0,2,'Q','a','p','h',0,'Q','a','p','h',0,
+2,'R','e','h',0,'R','e','h',0,2,'R','e','v','e','r','s','e','d','_','P','e',0,'R','e','v','e','r','s','e','d','_','P',
+'e',0,2,'S','a','d',0,'S','a','d',0,2,'S','a','d','h','e',0,'S','a','d','h','e',0,
+2,'S','e','e','n',0,'S','e','e','n',0,2,'S','e','m','k','a','t','h',0,'S','e','m','k','a','t','h',0,
+2,'S','h','i','n',0,'S','h','i','n',0,2,'S','w','a','s','h','_','K','a','f',0,'S','w','a','s','h','_','K','a','f',0,
+2,'S','y','r','i','a','c','_','W','a','w',0,'S','y','r','i','a','c','_','W','a','w',0,
+2,'T','a','h',0,'T','a','h',0,2,'T','a','w',0,'T','a','w',0,
+2,'T','e','h','_','M','a','r','b','u','t','a',0,'T','e','h','_','M','a','r','b','u','t','a',0,
+2,'T','e','t','h',0,'T','e','t','h',0,2,'W','a','w',0,'W','a','w',0,
+2,'Y','e','h',0,'Y','e','h',0,2,'Y','e','h','_','B','a','r','r','e','e',0,'Y','e','h','_','B','a','r','r','e','e',0,
+2,'Y','e','h','_','W','i','t','h','_','T','a','i','l',0,'Y','e','h','_','W','i','t','h','_','T','a','i','l',0,
+2,'Y','u','d','h',0,'Y','u','d','h',0,2,'Y','u','d','h','_','H','e',0,'Y','u','d','h','_','H','e',0,
+2,'Z','a','i','n',0,'Z','a','i','n',0,2,'F','e',0,'F','e',0,
+2,'K','h','a','p','h',0,'K','h','a','p','h',0,2,'Z','h','a','i','n',0,'Z','h','a','i','n',0,
+2,'B','u','r','u','s','h','a','s','k','i','_','Y','e','h','_','B','a','r','r','e','e',0,'B','u','r','u','s','h','a','s','k',
+'i','_','Y','e','h','_','B','a','r','r','e','e',0,2,'F','a','r','s','i','_','Y','e','h',0,
+'F','a','r','s','i','_','Y','e','h',0,2,'N','y','a',0,'N','y','a',0,
+2,'R','o','h','i','n','g','y','a','_','Y','e','h',0,'R','o','h','i','n','g','y','a','_','Y','e','h',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','A','l','e','p','h',0,'M','a','n','i','c','h','a','e','a','n','_','A','l','e',
+'p','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','A','y','i','n',0,'M','a','n','i','c','h','a','e','a','n','_','A',
+'y','i','n',0,2,'M','a','n','i','c','h','a','e','a','n','_','B','e','t','h',0,'M','a','n','i','c','h','a','e','a','n','_',
+'B','e','t','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','D','a','l','e','t','h',0,
+'M','a','n','i','c','h','a','e','a','n','_','D','a','l','e','t','h',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','D','h','a','m','e','d','h',0,'M','a','n','i','c','h','a','e','a','n','_','D',
+'h','a','m','e','d','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','F','i','v','e',0,
+'M','a','n','i','c','h','a','e','a','n','_','F','i','v','e',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','G','i','m','e','l',0,'M','a','n','i','c','h','a','e','a','n','_','G','i','m',
+'e','l',0,2,'M','a','n','i','c','h','a','e','a','n','_','H','e','t','h',0,'M','a','n','i','c','h','a','e','a','n','_','H',
+'e','t','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','H','u','n','d','r','e','d',0,
+'M','a','n','i','c','h','a','e','a','n','_','H','u','n','d','r','e','d',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','K','a','p','h',0,'M','a','n','i','c','h','a','e','a','n','_','K','a','p','h',
+0,2,'M','a','n','i','c','h','a','e','a','n','_','L','a','m','e','d','h',0,'M','a','n','i','c','h','a','e','a','n','_','L',
+'a','m','e','d','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','M','e','m',0,'M','a','n','i','c','h','a','e','a','n',
+'_','M','e','m',0,2,'M','a','n','i','c','h','a','e','a','n','_','N','u','n',0,'M','a','n','i','c','h','a','e','a','n','_',
+'N','u','n',0,2,'M','a','n','i','c','h','a','e','a','n','_','O','n','e',0,'M','a','n','i','c','h','a','e','a','n','_','O',
+'n','e',0,2,'M','a','n','i','c','h','a','e','a','n','_','P','e',0,'M','a','n','i','c','h','a','e','a','n','_','P','e',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','Q','o','p','h',0,'M','a','n','i','c','h','a','e','a','n','_','Q','o','p','h',
+0,2,'M','a','n','i','c','h','a','e','a','n','_','R','e','s','h',0,'M','a','n','i','c','h','a','e','a','n','_','R','e','s',
+'h',0,2,'M','a','n','i','c','h','a','e','a','n','_','S','a','d','h','e',0,'M','a','n','i','c','h','a','e','a','n','_','S',
+'a','d','h','e',0,2,'M','a','n','i','c','h','a','e','a','n','_','S','a','m','e','k','h',0,
+'M','a','n','i','c','h','a','e','a','n','_','S','a','m','e','k','h',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','T','a','w',0,'M','a','n','i','c','h','a','e','a','n','_','T','a','w',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','T','e','n',0,'M','a','n','i','c','h','a','e','a','n','_','T','e','n',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','T','e','t','h',0,'M','a','n','i','c','h','a','e','a','n','_','T','e','t','h',
+0,2,'M','a','n','i','c','h','a','e','a','n','_','T','h','a','m','e','d','h',0,'M','a','n','i','c','h','a','e','a','n','_',
+'T','h','a','m','e','d','h',0,2,'M','a','n','i','c','h','a','e','a','n','_','T','w','e','n','t','y',0,
+'M','a','n','i','c','h','a','e','a','n','_','T','w','e','n','t','y',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','W','a','w',0,'M','a','n','i','c','h','a','e','a','n','_','W','a','w',0,
+2,'M','a','n','i','c','h','a','e','a','n','_','Y','o','d','h',0,'M','a','n','i','c','h','a','e','a','n','_','Y','o','d','h',
+0,2,'M','a','n','i','c','h','a','e','a','n','_','Z','a','y','i','n',0,'M','a','n','i','c','h','a','e','a','n','_','Z','a',
+'y','i','n',0,2,'S','t','r','a','i','g','h','t','_','W','a','w',0,'S','t','r','a','i','g','h','t','_','W','a','w',0,
+2,'A','f','r','i','c','a','n','_','F','e','h',0,'A','f','r','i','c','a','n','_','F','e','h',0,
+2,'A','f','r','i','c','a','n','_','N','o','o','n',0,'A','f','r','i','c','a','n','_','N','o','o','n',0,
+2,'A','f','r','i','c','a','n','_','Q','a','f',0,'A','f','r','i','c','a','n','_','Q','a','f',0,
+2,'M','a','l','a','y','a','l','a','m','_','B','h','a',0,'M','a','l','a','y','a','l','a','m','_','B','h','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','J','a',0,'M','a','l','a','y','a','l','a','m','_','J','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','L','l','a',0,'M','a','l','a','y','a','l','a','m','_','L','l','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','L','l','l','a',0,'M','a','l','a','y','a','l','a','m','_','L','l','l','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','N','g','a',0,'M','a','l','a','y','a','l','a','m','_','N','g','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','N','n','a',0,'M','a','l','a','y','a','l','a','m','_','N','n','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','N','n','n','a',0,'M','a','l','a','y','a','l','a','m','_','N','n','n','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','N','y','a',0,'M','a','l','a','y','a','l','a','m','_','N','y','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','R','a',0,'M','a','l','a','y','a','l','a','m','_','R','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','S','s','a',0,'M','a','l','a','y','a','l','a','m','_','S','s','a',0,
+2,'M','a','l','a','y','a','l','a','m','_','T','t','a',0,'M','a','l','a','y','a','l','a','m','_','T','t','a',0,
+2,'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a','_','K','i','n','n','a','_','Y','a',0,
+'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a','_','K','i','n','n','a','_','Y','a',0,
+2,'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a','_','P','a',0,'H','a','n','i','f','i','_','R','o','h','i','n',
+'g','y','a','_','P','a',0,2,'j','t',0,'J','o','i','n','i','n','g','_','T','y','p','e',0,
+2,'U',0,'N','o','n','_','J','o','i','n','i','n','g',0,2,'C',0,'J','o','i','n','_','C','a','u','s','i','n','g',0,
+2,'D',0,'D','u','a','l','_','J','o','i','n','i','n','g',0,
+2,'L',0,'L','e','f','t','_','J','o','i','n','i','n','g',0,
+2,'R',0,'R','i','g','h','t','_','J','o','i','n','i','n','g',0,
+2,'T',0,'T','r','a','n','s','p','a','r','e','n','t',0,2,'l','b',0,'L','i','n','e','_','B','r','e','a','k',0,
+2,'X','X',0,'U','n','k','n','o','w','n',0,2,'A','I',0,'A','m','b','i','g','u','o','u','s',0,
+2,'A','L',0,'A','l','p','h','a','b','e','t','i','c',0,2,'B','2',0,'B','r','e','a','k','_','B','o','t','h',0,
+2,'B','A',0,'B','r','e','a','k','_','A','f','t','e','r',0,
+2,'B','B',0,'B','r','e','a','k','_','B','e','f','o','r','e',0,
+2,'B','K',0,'M','a','n','d','a','t','o','r','y','_','B','r','e','a','k',0,
+2,'C','B',0,'C','o','n','t','i','n','g','e','n','t','_','B','r','e','a','k',0,
+2,'C','L',0,'C','l','o','s','e','_','P','u','n','c','t','u','a','t','i','o','n',0,
+2,'C','M',0,'C','o','m','b','i','n','i','n','g','_','M','a','r','k',0,
+2,'C','R',0,'C','a','r','r','i','a','g','e','_','R','e','t','u','r','n',0,
+2,'E','X',0,'E','x','c','l','a','m','a','t','i','o','n',0,
+2,'G','L',0,'G','l','u','e',0,2,'H','Y',0,'H','y','p','h','e','n',0,
+2,'I','D',0,'I','d','e','o','g','r','a','p','h','i','c',0,
+3,'I','N',0,'I','n','s','e','p','a','r','a','b','l','e',0,'I','n','s','e','p','e','r','a','b','l','e',0,
+2,'I','S',0,'I','n','f','i','x','_','N','u','m','e','r','i','c',0,
+2,'L','F',0,'L','i','n','e','_','F','e','e','d',0,2,'N','S',0,'N','o','n','s','t','a','r','t','e','r',0,
+2,'N','U',0,'N','u','m','e','r','i','c',0,2,'O','P',0,'O','p','e','n','_','P','u','n','c','t','u','a','t','i','o','n',
+0,2,'P','O',0,'P','o','s','t','f','i','x','_','N','u','m','e','r','i','c',0,
+2,'P','R',0,'P','r','e','f','i','x','_','N','u','m','e','r','i','c',0,
+2,'Q','U',0,'Q','u','o','t','a','t','i','o','n',0,2,'S','A',0,'C','o','m','p','l','e','x','_','C','o','n','t','e','x',
+'t',0,2,'S','G',0,'S','u','r','r','o','g','a','t','e',0,
+2,'S','P',0,'S','p','a','c','e',0,2,'S','Y',0,'B','r','e','a','k','_','S','y','m','b','o','l','s',0,
+2,'Z','W',0,'Z','W','S','p','a','c','e',0,2,'N','L',0,'N','e','x','t','_','L','i','n','e',0,
+2,'W','J',0,'W','o','r','d','_','J','o','i','n','e','r',0,
+2,'H','2',0,'H','2',0,2,'H','3',0,'H','3',0,2,'J','L',0,'J','L',0,
+2,'J','T',0,'J','T',0,2,'J','V',0,'J','V',0,2,'C','P',0,'C','l','o','s','e','_','P','a','r','e','n','t','h','e',
+'s','i','s',0,2,'C','J',0,'C','o','n','d','i','t','i','o','n','a','l','_','J','a','p','a','n','e','s','e','_','S','t','a',
+'r','t','e','r',0,2,'H','L',0,'H','e','b','r','e','w','_','L','e','t','t','e','r',0,
+2,'E','B',0,'E','_','B','a','s','e',0,2,'E','M',0,'E','_','M','o','d','i','f','i','e','r',0,
+2,'Z','W','J',0,'Z','W','J',0,2,'n','t',0,'N','u','m','e','r','i','c','_','T','y','p','e',0,
+2,'N','o','n','e',0,'N','o','n','e',0,2,'D','e',0,'D','e','c','i','m','a','l',0,
+2,'D','i',0,'D','i','g','i','t',0,2,'N','u',0,'N','u','m','e','r','i','c',0,
+2,'s','c',0,'S','c','r','i','p','t',0,2,'Z','y','y','y',0,'C','o','m','m','o','n',0,
+3,'Z','i','n','h',0,'I','n','h','e','r','i','t','e','d',0,'Q','a','a','i',0,
+2,'A','r','a','b',0,'A','r','a','b','i','c',0,2,'A','r','m','n',0,'A','r','m','e','n','i','a','n',0,
+2,'B','e','n','g',0,'B','e','n','g','a','l','i',0,2,'B','o','p','o',0,'B','o','p','o','m','o','f','o',0,
+2,'C','h','e','r',0,'C','h','e','r','o','k','e','e',0,3,'C','o','p','t',0,'C','o','p','t','i','c',0,
+'Q','a','a','c',0,2,'C','y','r','l',0,'C','y','r','i','l','l','i','c',0,
+2,'D','s','r','t',0,'D','e','s','e','r','e','t',0,2,'D','e','v','a',0,'D','e','v','a','n','a','g','a','r','i',0,
+2,'E','t','h','i',0,'E','t','h','i','o','p','i','c',0,2,'G','e','o','r',0,'G','e','o','r','g','i','a','n',0,
+2,'G','o','t','h',0,'G','o','t','h','i','c',0,2,'G','r','e','k',0,'G','r','e','e','k',0,
+2,'G','u','j','r',0,'G','u','j','a','r','a','t','i',0,2,'G','u','r','u',0,'G','u','r','m','u','k','h','i',0,
+2,'H','a','n','i',0,'H','a','n',0,2,'H','a','n','g',0,'H','a','n','g','u','l',0,
+2,'H','e','b','r',0,'H','e','b','r','e','w',0,2,'H','i','r','a',0,'H','i','r','a','g','a','n','a',0,
+2,'K','n','d','a',0,'K','a','n','n','a','d','a',0,2,'K','a','n','a',0,'K','a','t','a','k','a','n','a',0,
+2,'K','h','m','r',0,'K','h','m','e','r',0,2,'L','a','o','o',0,'L','a','o',0,
+2,'L','a','t','n',0,'L','a','t','i','n',0,2,'M','l','y','m',0,'M','a','l','a','y','a','l','a','m',0,
+2,'M','o','n','g',0,'M','o','n','g','o','l','i','a','n',0,
+2,'M','y','m','r',0,'M','y','a','n','m','a','r',0,2,'O','g','a','m',0,'O','g','h','a','m',0,
+2,'I','t','a','l',0,'O','l','d','_','I','t','a','l','i','c',0,
+2,'O','r','y','a',0,'O','r','i','y','a',0,2,'R','u','n','r',0,'R','u','n','i','c',0,
+2,'S','i','n','h',0,'S','i','n','h','a','l','a',0,2,'S','y','r','c',0,'S','y','r','i','a','c',0,
+2,'T','a','m','l',0,'T','a','m','i','l',0,2,'T','e','l','u',0,'T','e','l','u','g','u',0,
+2,'T','h','a','a',0,'T','h','a','a','n','a',0,2,'T','i','b','t',0,'T','i','b','e','t','a','n',0,
+2,'C','a','n','s',0,'C','a','n','a','d','i','a','n','_','A','b','o','r','i','g','i','n','a','l',0,
+2,'Y','i','i','i',0,'Y','i',0,2,'T','g','l','g',0,'T','a','g','a','l','o','g',0,
+2,'H','a','n','o',0,'H','a','n','u','n','o','o',0,2,'B','u','h','d',0,'B','u','h','i','d',0,
+2,'T','a','g','b',0,'T','a','g','b','a','n','w','a',0,2,'B','r','a','i',0,'B','r','a','i','l','l','e',0,
+2,'C','p','r','t',0,'C','y','p','r','i','o','t',0,2,'L','i','m','b',0,'L','i','m','b','u',0,
+2,'L','i','n','b',0,'L','i','n','e','a','r','_','B',0,2,'O','s','m','a',0,'O','s','m','a','n','y','a',0,
+2,'S','h','a','w',0,'S','h','a','v','i','a','n',0,2,'T','a','l','e',0,'T','a','i','_','L','e',0,
+2,'U','g','a','r',0,'U','g','a','r','i','t','i','c',0,2,'H','r','k','t',0,'K','a','t','a','k','a','n','a','_','O','r',
+'_','H','i','r','a','g','a','n','a',0,2,'B','u','g','i',0,'B','u','g','i','n','e','s','e',0,
+2,'G','l','a','g',0,'G','l','a','g','o','l','i','t','i','c',0,
+2,'K','h','a','r',0,'K','h','a','r','o','s','h','t','h','i',0,
+2,'S','y','l','o',0,'S','y','l','o','t','i','_','N','a','g','r','i',0,
+2,'T','a','l','u',0,'N','e','w','_','T','a','i','_','L','u','e',0,
+2,'T','f','n','g',0,'T','i','f','i','n','a','g','h',0,2,'X','p','e','o',0,'O','l','d','_','P','e','r','s','i','a','n',
+0,2,'B','a','l','i',0,'B','a','l','i','n','e','s','e',0,
+2,'B','a','t','k',0,'B','a','t','a','k',0,2,'B','l','i','s',0,'B','l','i','s',0,
+2,'B','r','a','h',0,'B','r','a','h','m','i',0,2,'C','i','r','t',0,'C','i','r','t',0,
+2,'C','y','r','s',0,'C','y','r','s',0,2,'E','g','y','d',0,'E','g','y','d',0,
+2,'E','g','y','h',0,'E','g','y','h',0,2,'E','g','y','p',0,'E','g','y','p','t','i','a','n','_','H','i','e','r','o','g',
+'l','y','p','h','s',0,2,'G','e','o','k',0,'G','e','o','k',0,
+2,'H','a','n','s',0,'H','a','n','s',0,2,'H','a','n','t',0,'H','a','n','t',0,
+2,'H','m','n','g',0,'P','a','h','a','w','h','_','H','m','o','n','g',0,
+2,'H','u','n','g',0,'O','l','d','_','H','u','n','g','a','r','i','a','n',0,
+2,'I','n','d','s',0,'I','n','d','s',0,2,'J','a','v','a',0,'J','a','v','a','n','e','s','e',0,
+2,'K','a','l','i',0,'K','a','y','a','h','_','L','i',0,2,'L','a','t','f',0,'L','a','t','f',0,
+2,'L','a','t','g',0,'L','a','t','g',0,2,'L','e','p','c',0,'L','e','p','c','h','a',0,
+2,'L','i','n','a',0,'L','i','n','e','a','r','_','A',0,2,'M','a','n','d',0,'M','a','n','d','a','i','c',0,
+2,'M','a','y','a',0,'M','a','y','a',0,2,'M','e','r','o',0,'M','e','r','o','i','t','i','c','_','H','i','e','r','o','g',
+'l','y','p','h','s',0,2,'N','k','o','o',0,'N','k','o',0,
+2,'O','r','k','h',0,'O','l','d','_','T','u','r','k','i','c',0,
+2,'P','e','r','m',0,'O','l','d','_','P','e','r','m','i','c',0,
+2,'P','h','a','g',0,'P','h','a','g','s','_','P','a',0,2,'P','h','n','x',0,'P','h','o','e','n','i','c','i','a','n',0,
+2,'P','l','r','d',0,'M','i','a','o',0,2,'R','o','r','o',0,'R','o','r','o',0,
+2,'S','a','r','a',0,'S','a','r','a',0,2,'S','y','r','e',0,'S','y','r','e',0,
+2,'S','y','r','j',0,'S','y','r','j',0,2,'S','y','r','n',0,'S','y','r','n',0,
+2,'T','e','n','g',0,'T','e','n','g',0,2,'V','a','i','i',0,'V','a','i',0,
+2,'V','i','s','p',0,'V','i','s','p',0,2,'X','s','u','x',0,'C','u','n','e','i','f','o','r','m',0,
+2,'Z','x','x','x',0,'Z','x','x','x',0,2,'Z','z','z','z',0,'U','n','k','n','o','w','n',0,
+2,'C','a','r','i',0,'C','a','r','i','a','n',0,2,'J','p','a','n',0,'J','p','a','n',0,
+2,'L','a','n','a',0,'T','a','i','_','T','h','a','m',0,2,'L','y','c','i',0,'L','y','c','i','a','n',0,
+2,'L','y','d','i',0,'L','y','d','i','a','n',0,2,'O','l','c','k',0,'O','l','_','C','h','i','k','i',0,
+2,'R','j','n','g',0,'R','e','j','a','n','g',0,2,'S','a','u','r',0,'S','a','u','r','a','s','h','t','r','a',0,
+2,'S','g','n','w',0,'S','i','g','n','W','r','i','t','i','n','g',0,
+2,'S','u','n','d',0,'S','u','n','d','a','n','e','s','e',0,
+2,'M','o','o','n',0,'M','o','o','n',0,2,'M','t','e','i',0,'M','e','e','t','e','i','_','M','a','y','e','k',0,
+2,'A','r','m','i',0,'I','m','p','e','r','i','a','l','_','A','r','a','m','a','i','c',0,
+2,'A','v','s','t',0,'A','v','e','s','t','a','n',0,2,'C','a','k','m',0,'C','h','a','k','m','a',0,
+2,'K','o','r','e',0,'K','o','r','e',0,2,'K','t','h','i',0,'K','a','i','t','h','i',0,
+2,'M','a','n','i',0,'M','a','n','i','c','h','a','e','a','n',0,
+2,'P','h','l','i',0,'I','n','s','c','r','i','p','t','i','o','n','a','l','_','P','a','h','l','a','v','i',0,
+2,'P','h','l','p',0,'P','s','a','l','t','e','r','_','P','a','h','l','a','v','i',0,
+2,'P','h','l','v',0,'P','h','l','v',0,2,'P','r','t','i',0,'I','n','s','c','r','i','p','t','i','o','n','a','l','_','P',
+'a','r','t','h','i','a','n',0,2,'S','a','m','r',0,'S','a','m','a','r','i','t','a','n',0,
+2,'T','a','v','t',0,'T','a','i','_','V','i','e','t',0,2,'Z','m','t','h',0,'Z','m','t','h',0,
+2,'Z','s','y','m',0,'Z','s','y','m',0,2,'B','a','m','u',0,'B','a','m','u','m',0,
+2,'N','k','g','b',0,'N','k','g','b',0,2,'S','a','r','b',0,'O','l','d','_','S','o','u','t','h','_','A','r','a','b','i',
+'a','n',0,2,'B','a','s','s',0,'B','a','s','s','a','_','V','a','h',0,
+2,'D','u','p','l',0,'D','u','p','l','o','y','a','n',0,2,'E','l','b','a',0,'E','l','b','a','s','a','n',0,
+2,'G','r','a','n',0,'G','r','a','n','t','h','a',0,2,'K','p','e','l',0,'K','p','e','l',0,
+2,'L','o','m','a',0,'L','o','m','a',0,2,'M','e','n','d',0,'M','e','n','d','e','_','K','i','k','a','k','u','i',0,
+2,'M','e','r','c',0,'M','e','r','o','i','t','i','c','_','C','u','r','s','i','v','e',0,
+2,'N','a','r','b',0,'O','l','d','_','N','o','r','t','h','_','A','r','a','b','i','a','n',0,
+2,'N','b','a','t',0,'N','a','b','a','t','a','e','a','n',0,
+2,'P','a','l','m',0,'P','a','l','m','y','r','e','n','e',0,
+2,'S','i','n','d',0,'K','h','u','d','a','w','a','d','i',0,
+2,'W','a','r','a',0,'W','a','r','a','n','g','_','C','i','t','i',0,
+2,'A','f','a','k',0,'A','f','a','k',0,2,'J','u','r','c',0,'J','u','r','c',0,
+2,'M','r','o','o',0,'M','r','o',0,2,'N','s','h','u',0,'N','u','s','h','u',0,
+2,'S','h','r','d',0,'S','h','a','r','a','d','a',0,2,'S','o','r','a',0,'S','o','r','a','_','S','o','m','p','e','n','g',
+0,2,'T','a','k','r',0,'T','a','k','r','i',0,2,'T','a','n','g',0,'T','a','n','g','u','t',0,
+2,'W','o','l','e',0,'W','o','l','e',0,2,'H','l','u','w',0,'A','n','a','t','o','l','i','a','n','_','H','i','e','r','o',
+'g','l','y','p','h','s',0,2,'K','h','o','j',0,'K','h','o','j','k','i',0,
+2,'T','i','r','h',0,'T','i','r','h','u','t','a',0,2,'A','g','h','b',0,'C','a','u','c','a','s','i','a','n','_','A','l',
+'b','a','n','i','a','n',0,2,'M','a','h','j',0,'M','a','h','a','j','a','n','i',0,
+2,'H','a','t','r',0,'H','a','t','r','a','n',0,2,'M','u','l','t',0,'M','u','l','t','a','n','i',0,
+2,'P','a','u','c',0,'P','a','u','_','C','i','n','_','H','a','u',0,
+2,'S','i','d','d',0,'S','i','d','d','h','a','m',0,2,'A','d','l','m',0,'A','d','l','a','m',0,
+2,'B','h','k','s',0,'B','h','a','i','k','s','u','k','i',0,
+2,'M','a','r','c',0,'M','a','r','c','h','e','n',0,2,'O','s','g','e',0,'O','s','a','g','e',0,
+2,'H','a','n','b',0,'H','a','n','b',0,2,'J','a','m','o',0,'J','a','m','o',0,
+2,'Z','s','y','e',0,'Z','s','y','e',0,2,'G','o','n','m',0,'M','a','s','a','r','a','m','_','G','o','n','d','i',0,
+2,'S','o','y','o',0,'S','o','y','o','m','b','o',0,2,'Z','a','n','b',0,'Z','a','n','a','b','a','z','a','r','_','S','q',
+'u','a','r','e',0,2,'D','o','g','r',0,'D','o','g','r','a',0,
+2,'G','o','n','g',0,'G','u','n','j','a','l','a','_','G','o','n','d','i',0,
+2,'M','a','k','a',0,'M','a','k','a','s','a','r',0,2,'M','e','d','f',0,'M','e','d','e','f','a','i','d','r','i','n',0,
+2,'R','o','h','g',0,'H','a','n','i','f','i','_','R','o','h','i','n','g','y','a',0,
+2,'S','o','g','d',0,'S','o','g','d','i','a','n',0,2,'S','o','g','o',0,'O','l','d','_','S','o','g','d','i','a','n',0,
+2,'E','l','y','m',0,'E','l','y','m','a','i','c',0,2,'H','m','n','p',0,'N','y','i','a','k','e','n','g','_','P','u','a',
+'c','h','u','e','_','H','m','o','n','g',0,2,'N','a','n','d',0,'N','a','n','d','i','n','a','g','a','r','i',0,
+2,'W','c','h','o',0,'W','a','n','c','h','o',0,2,'C','h','r','s',0,'C','h','o','r','a','s','m','i','a','n',0,
+2,'D','i','a','k',0,'D','i','v','e','s','_','A','k','u','r','u',0,
+2,'K','i','t','s',0,'K','h','i','t','a','n','_','S','m','a','l','l','_','S','c','r','i','p','t',0,
+2,'Y','e','z','i',0,'Y','e','z','i','d','i',0,2,'h','s','t',0,'H','a','n','g','u','l','_','S','y','l','l','a','b','l',
+'e','_','T','y','p','e',0,2,'N','A',0,'N','o','t','_','A','p','p','l','i','c','a','b','l','e',0,
+2,'L',0,'L','e','a','d','i','n','g','_','J','a','m','o',0,
+2,'V',0,'V','o','w','e','l','_','J','a','m','o',0,2,'T',0,'T','r','a','i','l','i','n','g','_','J','a','m','o',0,
+2,'L','V',0,'L','V','_','S','y','l','l','a','b','l','e',0,
+2,'L','V','T',0,'L','V','T','_','S','y','l','l','a','b','l','e',0,
+2,'N','F','D','_','Q','C',0,'N','F','D','_','Q','u','i','c','k','_','C','h','e','c','k',0,
+2,'N',0,'N','o',0,2,'Y',0,'Y','e','s',0,2,'N','F','K','D','_','Q','C',0,'N','F','K','D','_','Q','u','i','c','k',
+'_','C','h','e','c','k',0,2,'N','F','C','_','Q','C',0,'N','F','C','_','Q','u','i','c','k','_','C','h','e','c','k',0,
+2,'M',0,'M','a','y','b','e',0,2,'N','F','K','C','_','Q','C',0,'N','F','K','C','_','Q','u','i','c','k','_','C','h','e',
+'c','k',0,2,'l','c','c','c',0,'L','e','a','d','_','C','a','n','o','n','i','c','a','l','_','C','o','m','b','i','n','i','n',
+'g','_','C','l','a','s','s',0,2,'t','c','c','c',0,'T','r','a','i','l','_','C','a','n','o','n','i','c','a','l','_','C','o',
+'m','b','i','n','i','n','g','_','C','l','a','s','s',0,2,'G','C','B',0,'G','r','a','p','h','e','m','e','_','C','l','u','s',
+'t','e','r','_','B','r','e','a','k',0,2,'X','X',0,'O','t','h','e','r',0,
+2,'C','N',0,'C','o','n','t','r','o','l',0,2,'C','R',0,'C','R',0,
+2,'E','X',0,'E','x','t','e','n','d',0,2,'L',0,'L',0,
+2,'L','F',0,'L','F',0,2,'L','V',0,'L','V',0,2,'L','V','T',0,'L','V','T',0,
+2,'T',0,'T',0,2,'V',0,'V',0,2,'S','M',0,'S','p','a','c','i','n','g','M','a','r','k',0,
+2,'P','P',0,'P','r','e','p','e','n','d',0,2,'E','B','G',0,'E','_','B','a','s','e','_','G','A','Z',0,
+2,'G','A','Z',0,'G','l','u','e','_','A','f','t','e','r','_','Z','w','j',0,
+2,'S','B',0,'S','e','n','t','e','n','c','e','_','B','r','e','a','k',0,
+2,'A','T',0,'A','T','e','r','m',0,2,'C','L',0,'C','l','o','s','e',0,
+2,'F','O',0,'F','o','r','m','a','t',0,2,'L','O',0,'L','o','w','e','r',0,
+2,'L','E',0,'O','L','e','t','t','e','r',0,2,'S','E',0,'S','e','p',0,
+2,'S','P',0,'S','p',0,2,'S','T',0,'S','T','e','r','m',0,
+2,'U','P',0,'U','p','p','e','r',0,2,'S','C',0,'S','C','o','n','t','i','n','u','e',0,
+2,'W','B',0,'W','o','r','d','_','B','r','e','a','k',0,2,'L','E',0,'A','L','e','t','t','e','r',0,
+2,'K','A',0,'K','a','t','a','k','a','n','a',0,2,'M','L',0,'M','i','d','L','e','t','t','e','r',0,
+2,'M','N',0,'M','i','d','N','u','m',0,2,'E','X',0,'E','x','t','e','n','d','N','u','m','L','e','t',0,
+2,'E','x','t','e','n','d',0,'E','x','t','e','n','d',0,2,'M','B',0,'M','i','d','N','u','m','L','e','t',0,
+2,'N','L',0,'N','e','w','l','i','n','e',0,2,'S','Q',0,'S','i','n','g','l','e','_','Q','u','o','t','e',0,
+2,'D','Q',0,'D','o','u','b','l','e','_','Q','u','o','t','e',0,
+2,'W','S','e','g','S','p','a','c','e',0,'W','S','e','g','S','p','a','c','e',0,
+2,'b','p','t',0,'B','i','d','i','_','P','a','i','r','e','d','_','B','r','a','c','k','e','t','_','T','y','p','e',0,
+2,'n',0,'N','o','n','e',0,2,'o',0,'O','p','e','n',0,
+2,'c',0,'C','l','o','s','e',0,2,'I','n','P','C',0,'I','n','d','i','c','_','P','o','s','i','t','i','o','n','a','l','_',
+'C','a','t','e','g','o','r','y',0,2,'N','A',0,'N','A',0,
+2,'B','o','t','t','o','m',0,'B','o','t','t','o','m',0,2,'B','o','t','t','o','m','_','A','n','d','_','L','e','f','t',0,
+'B','o','t','t','o','m','_','A','n','d','_','L','e','f','t',0,
+2,'B','o','t','t','o','m','_','A','n','d','_','R','i','g','h','t',0,'B','o','t','t','o','m','_','A','n','d','_','R','i','g',
+'h','t',0,2,'L','e','f','t',0,'L','e','f','t',0,2,'L','e','f','t','_','A','n','d','_','R','i','g','h','t',0,
+'L','e','f','t','_','A','n','d','_','R','i','g','h','t',0,2,'O','v','e','r','s','t','r','u','c','k',0,
+'O','v','e','r','s','t','r','u','c','k',0,2,'R','i','g','h','t',0,'R','i','g','h','t',0,
+2,'T','o','p',0,'T','o','p',0,2,'T','o','p','_','A','n','d','_','B','o','t','t','o','m',0,
+'T','o','p','_','A','n','d','_','B','o','t','t','o','m',0,2,'T','o','p','_','A','n','d','_','B','o','t','t','o','m','_','A',
+'n','d','_','R','i','g','h','t',0,'T','o','p','_','A','n','d','_','B','o','t','t','o','m','_','A','n','d','_','R','i','g','h',
+'t',0,2,'T','o','p','_','A','n','d','_','L','e','f','t',0,'T','o','p','_','A','n','d','_','L','e','f','t',0,
+2,'T','o','p','_','A','n','d','_','L','e','f','t','_','A','n','d','_','R','i','g','h','t',0,
+'T','o','p','_','A','n','d','_','L','e','f','t','_','A','n','d','_','R','i','g','h','t',0,
+2,'T','o','p','_','A','n','d','_','R','i','g','h','t',0,'T','o','p','_','A','n','d','_','R','i','g','h','t',0,
+2,'V','i','s','u','a','l','_','O','r','d','e','r','_','L','e','f','t',0,'V','i','s','u','a','l','_','O','r','d','e','r','_',
+'L','e','f','t',0,2,'T','o','p','_','A','n','d','_','B','o','t','t','o','m','_','A','n','d','_','L','e','f','t',0,
+'T','o','p','_','A','n','d','_','B','o','t','t','o','m','_','A','n','d','_','L','e','f','t',0,
+2,'I','n','S','C',0,'I','n','d','i','c','_','S','y','l','l','a','b','i','c','_','C','a','t','e','g','o','r','y',0,
+2,'O','t','h','e','r',0,'O','t','h','e','r',0,2,'A','v','a','g','r','a','h','a',0,'A','v','a','g','r','a','h','a',0,
+2,'B','i','n','d','u',0,'B','i','n','d','u',0,2,'B','r','a','h','m','i','_','J','o','i','n','i','n','g','_','N','u','m',
+'b','e','r',0,'B','r','a','h','m','i','_','J','o','i','n','i','n','g','_','N','u','m','b','e','r',0,
+2,'C','a','n','t','i','l','l','a','t','i','o','n','_','M','a','r','k',0,'C','a','n','t','i','l','l','a','t','i','o','n','_',
+'M','a','r','k',0,2,'C','o','n','s','o','n','a','n','t',0,'C','o','n','s','o','n','a','n','t',0,
+2,'C','o','n','s','o','n','a','n','t','_','D','e','a','d',0,'C','o','n','s','o','n','a','n','t','_','D','e','a','d',0,
+2,'C','o','n','s','o','n','a','n','t','_','F','i','n','a','l',0,'C','o','n','s','o','n','a','n','t','_','F','i','n','a','l',
+0,2,'C','o','n','s','o','n','a','n','t','_','H','e','a','d','_','L','e','t','t','e','r',0,
+'C','o','n','s','o','n','a','n','t','_','H','e','a','d','_','L','e','t','t','e','r',0,
+2,'C','o','n','s','o','n','a','n','t','_','I','n','i','t','i','a','l','_','P','o','s','t','f','i','x','e','d',0,
+'C','o','n','s','o','n','a','n','t','_','I','n','i','t','i','a','l','_','P','o','s','t','f','i','x','e','d',0,
+2,'C','o','n','s','o','n','a','n','t','_','K','i','l','l','e','r',0,'C','o','n','s','o','n','a','n','t','_','K','i','l','l',
+'e','r',0,2,'C','o','n','s','o','n','a','n','t','_','M','e','d','i','a','l',0,'C','o','n','s','o','n','a','n','t','_','M',
+'e','d','i','a','l',0,2,'C','o','n','s','o','n','a','n','t','_','P','l','a','c','e','h','o','l','d','e','r',0,
+'C','o','n','s','o','n','a','n','t','_','P','l','a','c','e','h','o','l','d','e','r',0,
+2,'C','o','n','s','o','n','a','n','t','_','P','r','e','c','e','d','i','n','g','_','R','e','p','h','a',0,
+'C','o','n','s','o','n','a','n','t','_','P','r','e','c','e','d','i','n','g','_','R','e','p','h','a',0,
+2,'C','o','n','s','o','n','a','n','t','_','P','r','e','f','i','x','e','d',0,'C','o','n','s','o','n','a','n','t','_','P','r',
+'e','f','i','x','e','d',0,2,'C','o','n','s','o','n','a','n','t','_','S','u','b','j','o','i','n','e','d',0,
+'C','o','n','s','o','n','a','n','t','_','S','u','b','j','o','i','n','e','d',0,
+2,'C','o','n','s','o','n','a','n','t','_','S','u','c','c','e','e','d','i','n','g','_','R','e','p','h','a',0,
+'C','o','n','s','o','n','a','n','t','_','S','u','c','c','e','e','d','i','n','g','_','R','e','p','h','a',0,
+2,'C','o','n','s','o','n','a','n','t','_','W','i','t','h','_','S','t','a','c','k','e','r',0,
+'C','o','n','s','o','n','a','n','t','_','W','i','t','h','_','S','t','a','c','k','e','r',0,
+2,'G','e','m','i','n','a','t','i','o','n','_','M','a','r','k',0,'G','e','m','i','n','a','t','i','o','n','_','M','a','r','k',
+0,2,'I','n','v','i','s','i','b','l','e','_','S','t','a','c','k','e','r',0,'I','n','v','i','s','i','b','l','e','_','S','t',
+'a','c','k','e','r',0,2,'J','o','i','n','e','r',0,'J','o','i','n','e','r',0,
+2,'M','o','d','i','f','y','i','n','g','_','L','e','t','t','e','r',0,'M','o','d','i','f','y','i','n','g','_','L','e','t','t',
+'e','r',0,2,'N','o','n','_','J','o','i','n','e','r',0,'N','o','n','_','J','o','i','n','e','r',0,
+2,'N','u','k','t','a',0,'N','u','k','t','a',0,2,'N','u','m','b','e','r',0,'N','u','m','b','e','r',0,
+2,'N','u','m','b','e','r','_','J','o','i','n','e','r',0,'N','u','m','b','e','r','_','J','o','i','n','e','r',0,
+2,'P','u','r','e','_','K','i','l','l','e','r',0,'P','u','r','e','_','K','i','l','l','e','r',0,
+2,'R','e','g','i','s','t','e','r','_','S','h','i','f','t','e','r',0,'R','e','g','i','s','t','e','r','_','S','h','i','f','t',
+'e','r',0,2,'S','y','l','l','a','b','l','e','_','M','o','d','i','f','i','e','r',0,'S','y','l','l','a','b','l','e','_','M',
+'o','d','i','f','i','e','r',0,2,'T','o','n','e','_','L','e','t','t','e','r',0,'T','o','n','e','_','L','e','t','t','e','r',
+0,2,'T','o','n','e','_','M','a','r','k',0,'T','o','n','e','_','M','a','r','k',0,
+2,'V','i','r','a','m','a',0,'V','i','r','a','m','a',0,2,'V','i','s','a','r','g','a',0,
+'V','i','s','a','r','g','a',0,2,'V','o','w','e','l',0,'V','o','w','e','l',0,
+2,'V','o','w','e','l','_','D','e','p','e','n','d','e','n','t',0,'V','o','w','e','l','_','D','e','p','e','n','d','e','n','t',
+0,2,'V','o','w','e','l','_','I','n','d','e','p','e','n','d','e','n','t',0,'V','o','w','e','l','_','I','n','d','e','p','e',
+'n','d','e','n','t',0,2,'v','o',0,'V','e','r','t','i','c','a','l','_','O','r','i','e','n','t','a','t','i','o','n',0,
+2,'R',0,'R','o','t','a','t','e','d',0,2,'T','r',0,'T','r','a','n','s','f','o','r','m','e','d','_','R','o','t','a','t',
+'e','d',0,2,'T','u',0,'T','r','a','n','s','f','o','r','m','e','d','_','U','p','r','i','g','h','t',0,
+2,'U',0,'U','p','r','i','g','h','t',0,2,'g','c','m',0,'G','e','n','e','r','a','l','_','C','a','t','e','g','o','r','y',
+'_','M','a','s','k',0,2,'C',0,'O','t','h','e','r',0,2,'L',0,'L','e','t','t','e','r',0,
+2,'L','C',0,'C','a','s','e','d','_','L','e','t','t','e','r',0,
+3,'M',0,'M','a','r','k',0,'C','o','m','b','i','n','i','n','g','_','M','a','r','k',0,
+2,'N',0,'N','u','m','b','e','r',0,3,'P',0,'P','u','n','c','t','u','a','t','i','o','n',0,
+'p','u','n','c','t',0,2,'S',0,'S','y','m','b','o','l',0,
+2,'Z',0,'S','e','p','a','r','a','t','o','r',0,2,'n','v',0,'N','u','m','e','r','i','c','_','V','a','l','u','e',0,
+2,'a','g','e',0,'A','g','e',0,2,'b','m','g',0,'B','i','d','i','_','M','i','r','r','o','r','i','n','g','_','G','l','y',
+'p','h',0,2,'c','f',0,'C','a','s','e','_','F','o','l','d','i','n','g',0,
+2,'i','s','c',0,'I','S','O','_','C','o','m','m','e','n','t',0,
+2,'l','c',0,'L','o','w','e','r','c','a','s','e','_','M','a','p','p','i','n','g',0,
+2,'n','a',0,'N','a','m','e',0,3,'s','c','f',0,'S','i','m','p','l','e','_','C','a','s','e','_','F','o','l','d','i','n',
+'g',0,'s','f','c',0,2,'s','l','c',0,'S','i','m','p','l','e','_','L','o','w','e','r','c','a','s','e','_','M','a','p','p',
+'i','n','g',0,2,'s','t','c',0,'S','i','m','p','l','e','_','T','i','t','l','e','c','a','s','e','_','M','a','p','p','i','n',
+'g',0,2,'s','u','c',0,'S','i','m','p','l','e','_','U','p','p','e','r','c','a','s','e','_','M','a','p','p','i','n','g',0,
+2,'t','c',0,'T','i','t','l','e','c','a','s','e','_','M','a','p','p','i','n','g',0,
+2,'n','a','1',0,'U','n','i','c','o','d','e','_','1','_','N','a','m','e',0,
+2,'u','c',0,'U','p','p','e','r','c','a','s','e','_','M','a','p','p','i','n','g',0,
+2,'b','p','b',0,'B','i','d','i','_','P','a','i','r','e','d','_','B','r','a','c','k','e','t',0,
+2,'s','c','x',0,'S','c','r','i','p','t','_','E','x','t','e','n','s','i','o','n','s',0
+};
+
+U_NAMESPACE_END
+
+#endif  // INCLUDED_FROM_PROPNAME_CPP
diff --git a/thirdparty/icu4c/common/propsvec.cpp b/thirdparty/icu4c/common/propsvec.cpp
new file mode 100644
index 0000000000..056fcda9cf
--- /dev/null
+++ b/thirdparty/icu4c/common/propsvec.cpp
@@ -0,0 +1,529 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  propsvec.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002feb22
+*   created by: Markus W. Scherer
+*
+*   Store bits (Unicode character properties) in bit set vectors.
+*/
+
+#include <stdlib.h>
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "utrie.h"
+#include "utrie2.h"
+#include "uarrsort.h"
+#include "propsvec.h"
+#include "uassert.h"
+
+struct UPropsVectors {
+    uint32_t *v;
+    int32_t columns;  /* number of columns, plus two for start & limit values */
+    int32_t maxRows;
+    int32_t rows;
+    int32_t prevRow;  /* search optimization: remember last row seen */
+    UBool isCompacted;
+};
+
+#define UPVEC_INITIAL_ROWS (1<<12)
+#define UPVEC_MEDIUM_ROWS ((int32_t)1<<16)
+#define UPVEC_MAX_ROWS (UPVEC_MAX_CP+1)
+
+U_CAPI UPropsVectors * U_EXPORT2
+upvec_open(int32_t columns, UErrorCode *pErrorCode) {
+    UPropsVectors *pv;
+    uint32_t *v, *row;
+    uint32_t cp;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(columns<1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    columns+=2; /* count range start and limit columns */
+
+    pv=(UPropsVectors *)uprv_malloc(sizeof(UPropsVectors));
+    v=(uint32_t *)uprv_malloc(UPVEC_INITIAL_ROWS*columns*4);
+    if(pv==NULL || v==NULL) {
+        uprv_free(pv);
+        uprv_free(v);
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memset(pv, 0, sizeof(UPropsVectors));
+    pv->v=v;
+    pv->columns=columns;
+    pv->maxRows=UPVEC_INITIAL_ROWS;
+    pv->rows=2+(UPVEC_MAX_CP-UPVEC_FIRST_SPECIAL_CP);
+
+    /* set the all-Unicode row and the special-value rows */
+    row=pv->v;
+    uprv_memset(row, 0, pv->rows*columns*4);
+    row[0]=0;
+    row[1]=0x110000;
+    row+=columns;
+    for(cp=UPVEC_FIRST_SPECIAL_CP; cp<=UPVEC_MAX_CP; ++cp) {
+        row[0]=cp;
+        row[1]=cp+1;
+        row+=columns;
+    }
+    return pv;
+}
+
+U_CAPI void U_EXPORT2
+upvec_close(UPropsVectors *pv) {
+    if(pv!=NULL) {
+        uprv_free(pv->v);
+        uprv_free(pv);
+    }
+}
+
+static uint32_t *
+_findRow(UPropsVectors *pv, UChar32 rangeStart) {
+    uint32_t *row;
+    int32_t columns, i, start, limit, prevRow;
+
+    columns=pv->columns;
+    limit=pv->rows;
+    prevRow=pv->prevRow;
+
+    /* check the vicinity of the last-seen row (start searching with an unrolled loop) */
+    row=pv->v+prevRow*columns;
+    if(rangeStart>=(UChar32)row[0]) {
+        if(rangeStart<(UChar32)row[1]) {
+            /* same row as last seen */
+            return row;
+        } else if(rangeStart<(UChar32)(row+=columns)[1]) {
+            /* next row after the last one */
+            pv->prevRow=prevRow+1;
+            return row;
+        } else if(rangeStart<(UChar32)(row+=columns)[1]) {
+            /* second row after the last one */
+            pv->prevRow=prevRow+2;
+            return row;
+        } else if((rangeStart-(UChar32)row[1])<10) {
+            /* we are close, continue looping */
+            prevRow+=2;
+            do {
+                ++prevRow;
+                row+=columns;
+            } while(rangeStart>=(UChar32)row[1]);
+            pv->prevRow=prevRow;
+            return row;
+        }
+    } else if(rangeStart<(UChar32)pv->v[1]) {
+        /* the very first row */
+        pv->prevRow=0;
+        return pv->v;
+    }
+
+    /* do a binary search for the start of the range */
+    start=0;
+    while(start<limit-1) {
+        i=(start+limit)/2;
+        row=pv->v+i*columns;
+        if(rangeStart<(UChar32)row[0]) {
+            limit=i;
+        } else if(rangeStart<(UChar32)row[1]) {
+            pv->prevRow=i;
+            return row;
+        } else {
+            start=i;
+        }
+    }
+
+    /* must be found because all ranges together always cover all of Unicode */
+    pv->prevRow=start;
+    return pv->v+start*columns;
+}
+
+U_CAPI void U_EXPORT2
+upvec_setValue(UPropsVectors *pv,
+               UChar32 start, UChar32 end,
+               int32_t column,
+               uint32_t value, uint32_t mask,
+               UErrorCode *pErrorCode) {
+    uint32_t *firstRow, *lastRow;
+    int32_t columns;
+    UChar32 limit;
+    UBool splitFirstRow, splitLastRow;
+
+    /* argument checking */
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if( pv==NULL ||
+        start<0 || start>end || end>UPVEC_MAX_CP ||
+        column<0 || column>=(pv->columns-2)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if(pv->isCompacted) {
+        *pErrorCode=U_NO_WRITE_PERMISSION;
+        return;
+    }
+    limit=end+1;
+
+    /* initialize */
+    columns=pv->columns;
+    column+=2; /* skip range start and limit columns */
+    value&=mask;
+
+    /* find the rows whose ranges overlap with the input range */
+
+    /* find the first and last rows, always successful */
+    firstRow=_findRow(pv, start);
+    lastRow=_findRow(pv, end);
+
+    /*
+     * Rows need to be split if they partially overlap with the
+     * input range (only possible for the first and last rows)
+     * and if their value differs from the input value.
+     */
+    splitFirstRow= (UBool)(start!=(UChar32)firstRow[0] && value!=(firstRow[column]&mask));
+    splitLastRow= (UBool)(limit!=(UChar32)lastRow[1] && value!=(lastRow[column]&mask));
+
+    /* split first/last rows if necessary */
+    if(splitFirstRow || splitLastRow) {
+        int32_t count, rows;
+
+        rows=pv->rows;
+        if((rows+splitFirstRow+splitLastRow)>pv->maxRows) {
+            uint32_t *newVectors;
+            int32_t newMaxRows;
+
+            if(pv->maxRows<UPVEC_MEDIUM_ROWS) {
+                newMaxRows=UPVEC_MEDIUM_ROWS;
+            } else if(pv->maxRows<UPVEC_MAX_ROWS) {
+                newMaxRows=UPVEC_MAX_ROWS;
+            } else {
+                /* Implementation bug, or UPVEC_MAX_ROWS too low. */
+                *pErrorCode=U_INTERNAL_PROGRAM_ERROR;
+                return;
+            }
+            newVectors=(uint32_t *)uprv_malloc(newMaxRows*columns*4);
+            if(newVectors==NULL) {
+                *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            uprv_memcpy(newVectors, pv->v, (size_t)rows*columns*4);
+            firstRow=newVectors+(firstRow-pv->v);
+            lastRow=newVectors+(lastRow-pv->v);
+            uprv_free(pv->v);
+            pv->v=newVectors;
+            pv->maxRows=newMaxRows;
+        }
+
+        /* count the number of row cells to move after the last row, and move them */
+        count = (int32_t)((pv->v+rows*columns)-(lastRow+columns));
+        if(count>0) {
+            uprv_memmove(
+                lastRow+(1+splitFirstRow+splitLastRow)*columns,
+                lastRow+columns,
+                count*4);
+        }
+        pv->rows=rows+splitFirstRow+splitLastRow;
+
+        /* split the first row, and move the firstRow pointer to the second part */
+        if(splitFirstRow) {
+            /* copy all affected rows up one and move the lastRow pointer */
+            count = (int32_t)((lastRow-firstRow)+columns);
+            uprv_memmove(firstRow+columns, firstRow, (size_t)count*4);
+            lastRow+=columns;
+
+            /* split the range and move the firstRow pointer */
+            firstRow[1]=firstRow[columns]=(uint32_t)start;
+            firstRow+=columns;
+        }
+
+        /* split the last row */
+        if(splitLastRow) {
+            /* copy the last row data */
+            uprv_memcpy(lastRow+columns, lastRow, (size_t)columns*4);
+
+            /* split the range and move the firstRow pointer */
+            lastRow[1]=lastRow[columns]=(uint32_t)limit;
+        }
+    }
+
+    /* set the "row last seen" to the last row for the range */
+    pv->prevRow=(int32_t)((lastRow-(pv->v))/columns);
+
+    /* set the input value in all remaining rows */
+    firstRow+=column;
+    lastRow+=column;
+    mask=~mask;
+    for(;;) {
+        *firstRow=(*firstRow&mask)|value;
+        if(firstRow==lastRow) {
+            break;
+        }
+        firstRow+=columns;
+    }
+}
+
+U_CAPI uint32_t U_EXPORT2
+upvec_getValue(const UPropsVectors *pv, UChar32 c, int32_t column) {
+    uint32_t *row;
+    UPropsVectors *ncpv;
+
+    if(pv->isCompacted || c<0 || c>UPVEC_MAX_CP || column<0 || column>=(pv->columns-2)) {
+        return 0;
+    }
+    ncpv=(UPropsVectors *)pv;
+    row=_findRow(ncpv, c);
+    return row[2+column];
+}
+
+U_CAPI uint32_t * U_EXPORT2
+upvec_getRow(const UPropsVectors *pv, int32_t rowIndex,
+             UChar32 *pRangeStart, UChar32 *pRangeEnd) {
+    uint32_t *row;
+    int32_t columns;
+
+    if(pv->isCompacted || rowIndex<0 || rowIndex>=pv->rows) {
+        return NULL;
+    }
+
+    columns=pv->columns;
+    row=pv->v+rowIndex*columns;
+    if(pRangeStart!=NULL) {
+        *pRangeStart=(UChar32)row[0];
+    }
+    if(pRangeEnd!=NULL) {
+        *pRangeEnd=(UChar32)row[1]-1;
+    }
+    return row+2;
+}
+
+static int32_t U_CALLCONV
+upvec_compareRows(const void *context, const void *l, const void *r) {
+    const uint32_t *left=(const uint32_t *)l, *right=(const uint32_t *)r;
+    const UPropsVectors *pv=(const UPropsVectors *)context;
+    int32_t i, count, columns;
+
+    count=columns=pv->columns; /* includes start/limit columns */
+
+    /* start comparing after start/limit but wrap around to them */
+    i=2;
+    do {
+        if(left[i]!=right[i]) {
+            return left[i]<right[i] ? -1 : 1;
+        }
+        if(++i==columns) {
+            i=0;
+        }
+    } while(--count>0);
+
+    return 0;
+}
+
+U_CAPI void U_EXPORT2
+upvec_compact(UPropsVectors *pv, UPVecCompactHandler *handler, void *context, UErrorCode *pErrorCode) {
+    uint32_t *row;
+    int32_t i, columns, valueColumns, rows, count;
+    UChar32 start, limit;
+
+    /* argument checking */
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if(handler==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if(pv->isCompacted) {
+        return;
+    }
+
+    /* Set the flag now: Sorting and compacting destroys the builder data structure. */
+    pv->isCompacted=TRUE;
+
+    rows=pv->rows;
+    columns=pv->columns;
+    U_ASSERT(columns>=3); /* upvec_open asserts this */
+    valueColumns=columns-2; /* not counting start & limit */
+
+    /* sort the properties vectors to find unique vector values */
+    uprv_sortArray(pv->v, rows, columns*4,
+                   upvec_compareRows, pv, FALSE, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /*
+     * Find and set the special values.
+     * This has to do almost the same work as the compaction below,
+     * to find the indexes where the special-value rows will move.
+     */
+    row=pv->v;
+    count=-valueColumns;
+    for(i=0; i<rows; ++i) {
+        start=(UChar32)row[0];
+
+        /* count a new values vector if it is different from the current one */
+        if(count<0 || 0!=uprv_memcmp(row+2, row-valueColumns, valueColumns*4)) {
+            count+=valueColumns;
+        }
+
+        if(start>=UPVEC_FIRST_SPECIAL_CP) {
+            handler(context, start, start, count, row+2, valueColumns, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                return;
+            }
+        }
+
+        row+=columns;
+    }
+
+    /* count is at the beginning of the last vector, add valueColumns to include that last vector */
+    count+=valueColumns;
+
+    /* Call the handler once more to signal the start of delivering real values. */
+    handler(context, UPVEC_START_REAL_VALUES_CP, UPVEC_START_REAL_VALUES_CP,
+            count, row-valueColumns, valueColumns, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /*
+     * Move vector contents up to a contiguous array with only unique
+     * vector values, and call the handler function for each vector.
+     *
+     * This destroys the Properties Vector structure and replaces it
+     * with an array of just vector values.
+     */
+    row=pv->v;
+    count=-valueColumns;
+    for(i=0; i<rows; ++i) {
+        /* fetch these first before memmove() may overwrite them */
+        start=(UChar32)row[0];
+        limit=(UChar32)row[1];
+
+        /* add a new values vector if it is different from the current one */
+        if(count<0 || 0!=uprv_memcmp(row+2, pv->v+count, valueColumns*4)) {
+            count+=valueColumns;
+            uprv_memmove(pv->v+count, row+2, (size_t)valueColumns*4);
+        }
+
+        if(start<UPVEC_FIRST_SPECIAL_CP) {
+            handler(context, start, limit-1, count, pv->v+count, valueColumns, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                return;
+            }
+        }
+
+        row+=columns;
+    }
+
+    /* count is at the beginning of the last vector, add one to include that last vector */
+    pv->rows=count/valueColumns+1;
+}
+
+U_CAPI const uint32_t * U_EXPORT2
+upvec_getArray(const UPropsVectors *pv, int32_t *pRows, int32_t *pColumns) {
+    if(!pv->isCompacted) {
+        return NULL;
+    }
+    if(pRows!=NULL) {
+        *pRows=pv->rows;
+    }
+    if(pColumns!=NULL) {
+        *pColumns=pv->columns-2;
+    }
+    return pv->v;
+}
+
+U_CAPI uint32_t * U_EXPORT2
+upvec_cloneArray(const UPropsVectors *pv,
+                 int32_t *pRows, int32_t *pColumns, UErrorCode *pErrorCode) {
+    uint32_t *clonedArray;
+    int32_t byteLength;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(!pv->isCompacted) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    byteLength=pv->rows*(pv->columns-2)*4;
+    clonedArray=(uint32_t *)uprv_malloc(byteLength);
+    if(clonedArray==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memcpy(clonedArray, pv->v, byteLength);
+    if(pRows!=NULL) {
+        *pRows=pv->rows;
+    }
+    if(pColumns!=NULL) {
+        *pColumns=pv->columns-2;
+    }
+    return clonedArray;
+}
+
+U_CAPI UTrie2 * U_EXPORT2
+upvec_compactToUTrie2WithRowIndexes(UPropsVectors *pv, UErrorCode *pErrorCode) {
+    UPVecToUTrie2Context toUTrie2={ NULL, 0, 0, 0 };
+    upvec_compact(pv, upvec_compactToUTrie2Handler, &toUTrie2, pErrorCode);
+    utrie2_freeze(toUTrie2.trie, UTRIE2_16_VALUE_BITS, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        utrie2_close(toUTrie2.trie);
+        toUTrie2.trie=NULL;
+    }
+    return toUTrie2.trie;
+}
+
+/*
+ * TODO(markus): Add upvec_16BitsToUTrie2() function that enumerates all rows, extracts
+ * some 16-bit field and builds and returns a UTrie2.
+ */
+
+U_CAPI void U_CALLCONV
+upvec_compactToUTrie2Handler(void *context,
+                             UChar32 start, UChar32 end,
+                             int32_t rowIndex, uint32_t *row, int32_t columns,
+                             UErrorCode *pErrorCode) {
+    (void)row;
+    (void)columns;
+    UPVecToUTrie2Context *toUTrie2=(UPVecToUTrie2Context *)context;
+    if(start<UPVEC_FIRST_SPECIAL_CP) {
+        utrie2_setRange32(toUTrie2->trie, start, end, (uint32_t)rowIndex, TRUE, pErrorCode);
+    } else {
+        switch(start) {
+        case UPVEC_INITIAL_VALUE_CP:
+            toUTrie2->initialValue=rowIndex;
+            break;
+        case UPVEC_ERROR_VALUE_CP:
+            toUTrie2->errorValue=rowIndex;
+            break;
+        case UPVEC_START_REAL_VALUES_CP:
+            toUTrie2->maxValue=rowIndex;
+            if(rowIndex>0xffff) {
+                /* too many rows for a 16-bit trie */
+                *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            } else {
+                toUTrie2->trie=utrie2_open(toUTrie2->initialValue,
+                                           toUTrie2->errorValue, pErrorCode);
+            }
+            break;
+        default:
+            break;
+        }
+    }
+}
diff --git a/thirdparty/icu4c/common/propsvec.h b/thirdparty/icu4c/common/propsvec.h
new file mode 100644
index 0000000000..39080615ea
--- /dev/null
+++ b/thirdparty/icu4c/common/propsvec.h
@@ -0,0 +1,178 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2010, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  propsvec.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002feb22
+*   created by: Markus W. Scherer
+*
+*   Store bits (Unicode character properties) in bit set vectors.
+*/
+
+#ifndef __UPROPSVEC_H__
+#define __UPROPSVEC_H__
+
+#include "unicode/utypes.h"
+#include "utrie.h"
+#include "utrie2.h"
+
+U_CDECL_BEGIN
+
+/**
+ * Unicode Properties Vectors associated with code point ranges.
+ *
+ * Rows of uint32_t integers in a contiguous array store
+ * the range limits and the properties vectors.
+ *
+ * Logically, each row has a certain number of uint32_t values,
+ * which is set via the upvec_open() "columns" parameter.
+ *
+ * Internally, two additional columns are stored.
+ * In each internal row,
+ * row[0] contains the start code point and
+ * row[1] contains the limit code point,
+ * which is the start of the next range.
+ *
+ * Initially, there is only one "normal" row for
+ * range [0..0x110000[ with values 0.
+ * There are additional rows for special purposes, see UPVEC_FIRST_SPECIAL_CP.
+ *
+ * It would be possible to store only one range boundary per row,
+ * but self-contained rows allow to later sort them by contents.
+ */
+struct UPropsVectors;
+typedef struct UPropsVectors UPropsVectors;
+
+/*
+ * Special pseudo code points for storing the initialValue and the errorValue,
+ * which are used to initialize a UTrie2 or similar.
+ */
+#define UPVEC_FIRST_SPECIAL_CP 0x110000
+#define UPVEC_INITIAL_VALUE_CP 0x110000
+#define UPVEC_ERROR_VALUE_CP 0x110001
+#define UPVEC_MAX_CP 0x110001
+
+/*
+ * Special pseudo code point used in upvec_compact() signalling the end of
+ * delivering special values and the beginning of delivering real ones.
+ * Stable value, unlike UPVEC_MAX_CP which might grow over time.
+ */
+#define UPVEC_START_REAL_VALUES_CP 0x200000
+
+/*
+ * Open a UPropsVectors object.
+ * @param columns Number of value integers (uint32_t) per row.
+ */
+U_CAPI UPropsVectors * U_EXPORT2
+upvec_open(int32_t columns, UErrorCode *pErrorCode);
+
+U_CAPI void U_EXPORT2
+upvec_close(UPropsVectors *pv);
+
+/*
+ * In rows for code points [start..end], select the column,
+ * reset the mask bits and set the value bits (ANDed with the mask).
+ *
+ * Will set U_NO_WRITE_PERMISSION if called after upvec_compact().
+ */
+U_CAPI void U_EXPORT2
+upvec_setValue(UPropsVectors *pv,
+               UChar32 start, UChar32 end,
+               int32_t column,
+               uint32_t value, uint32_t mask,
+               UErrorCode *pErrorCode);
+
+/*
+ * Logically const but must not be used on the same pv concurrently!
+ * Always returns 0 if called after upvec_compact().
+ */
+U_CAPI uint32_t U_EXPORT2
+upvec_getValue(const UPropsVectors *pv, UChar32 c, int32_t column);
+
+/*
+ * pRangeStart and pRangeEnd can be NULL.
+ * @return NULL if rowIndex out of range and for illegal arguments,
+ *         or if called after upvec_compact()
+ */
+U_CAPI uint32_t * U_EXPORT2
+upvec_getRow(const UPropsVectors *pv, int32_t rowIndex,
+             UChar32 *pRangeStart, UChar32 *pRangeEnd);
+
+/*
+ * Compact the vectors:
+ * - modify the memory
+ * - keep only unique vectors
+ * - store them contiguously from the beginning of the memory
+ * - for each (non-unique) row, call the handler function
+ *
+ * The handler's rowIndex is the index of the row in the compacted
+ * memory block.
+ * (Therefore, it starts at 0 increases in increments of the columns value.)
+ *
+ * In a first phase, only special values are delivered (each exactly once),
+ * with start==end both equalling a special pseudo code point.
+ * Then the handler is called once more with start==end==UPVEC_START_REAL_VALUES_CP
+ * where rowIndex is the length of the compacted array,
+ * and the row is arbitrary (but not NULL).
+ * Then, in the second phase, the handler is called for each row of real values.
+ */
+typedef void U_CALLCONV
+UPVecCompactHandler(void *context,
+                    UChar32 start, UChar32 end,
+                    int32_t rowIndex, uint32_t *row, int32_t columns,
+                    UErrorCode *pErrorCode);
+
+U_CAPI void U_EXPORT2
+upvec_compact(UPropsVectors *pv, UPVecCompactHandler *handler, void *context, UErrorCode *pErrorCode);
+
+/*
+ * Get the vectors array after calling upvec_compact().
+ * The caller must not modify nor release the returned array.
+ * Returns NULL if called before upvec_compact().
+ */
+U_CAPI const uint32_t * U_EXPORT2
+upvec_getArray(const UPropsVectors *pv, int32_t *pRows, int32_t *pColumns);
+
+/*
+ * Get a clone of the vectors array after calling upvec_compact().
+ * The caller owns the returned array and must uprv_free() it.
+ * Returns NULL if called before upvec_compact().
+ */
+U_CAPI uint32_t * U_EXPORT2
+upvec_cloneArray(const UPropsVectors *pv,
+                 int32_t *pRows, int32_t *pColumns, UErrorCode *pErrorCode);
+
+/*
+ * Call upvec_compact(), create a 16-bit UTrie2 with indexes into the compacted
+ * vectors array, and freeze the trie.
+ */
+U_CAPI UTrie2 * U_EXPORT2
+upvec_compactToUTrie2WithRowIndexes(UPropsVectors *pv, UErrorCode *pErrorCode);
+
+struct UPVecToUTrie2Context {
+    UTrie2 *trie;
+    int32_t initialValue;
+    int32_t errorValue;
+    int32_t maxValue;
+};
+typedef struct UPVecToUTrie2Context UPVecToUTrie2Context;
+
+/* context=UPVecToUTrie2Context, creates the trie and stores the rowIndex values */
+U_CAPI void U_CALLCONV
+upvec_compactToUTrie2Handler(void *context,
+                             UChar32 start, UChar32 end,
+                             int32_t rowIndex, uint32_t *row, int32_t columns,
+                             UErrorCode *pErrorCode);
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/punycode.cpp b/thirdparty/icu4c/common/punycode.cpp
new file mode 100644
index 0000000000..4832938ff7
--- /dev/null
+++ b/thirdparty/icu4c/common/punycode.cpp
@@ -0,0 +1,590 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  punycode.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jan31
+*   created by: Markus W. Scherer
+*/
+
+
+/* This ICU code derived from: */
+/*
+punycode.c 0.4.0 (2001-Nov-17-Sat)
+http://www.cs.berkeley.edu/~amc/idn/
+Adam M. Costello
+http://www.nicemice.net/amc/
+
+Disclaimer and license
+
+    Regarding this entire document or any portion of it (including
+    the pseudocode and C code), the author makes no guarantees and
+    is not responsible for any damage resulting from its use.  The
+    author grants irrevocable permission to anyone to use, modify,
+    and distribute it in any way that does not diminish the rights
+    of anyone else to use, modify, and distribute it, provided that
+    redistributed derivative works do not contain misleading author or
+    version information.  Derivative works need not be licensed under
+    similar terms.
+*/
+/*
+ * ICU modifications:
+ * - ICU data types and coding conventions
+ * - ICU string buffer handling with implicit source lengths
+ *   and destination preflighting
+ * - UTF-16 handling
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/ustring.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "punycode.h"
+#include "uassert.h"
+
+
+/* Punycode ----------------------------------------------------------------- */
+
+/* Punycode parameters for Bootstring */
+#define BASE            36
+#define TMIN            1
+#define TMAX            26
+#define SKEW            38
+#define DAMP            700
+#define INITIAL_BIAS    72
+#define INITIAL_N       0x80
+
+/* "Basic" Unicode/ASCII code points */
+#define _HYPHEN         0X2d
+#define DELIMITER       _HYPHEN
+
+#define _ZERO_          0X30
+#define _NINE           0x39
+
+#define _SMALL_A        0X61
+#define _SMALL_Z        0X7a
+
+#define _CAPITAL_A      0X41
+#define _CAPITAL_Z      0X5a
+
+#define IS_BASIC(c) ((c)<0x80)
+#define IS_BASIC_UPPERCASE(c) (_CAPITAL_A<=(c) && (c)<=_CAPITAL_Z)
+
+/**
+ * digitToBasic() returns the basic code point whose value
+ * (when used for representing integers) is d, which must be in the
+ * range 0 to BASE-1. The lowercase form is used unless the uppercase flag is
+ * nonzero, in which case the uppercase form is used.
+ */
+static inline char
+digitToBasic(int32_t digit, UBool uppercase) {
+    /*  0..25 map to ASCII a..z or A..Z */
+    /* 26..35 map to ASCII 0..9         */
+    if(digit<26) {
+        if(uppercase) {
+            return (char)(_CAPITAL_A+digit);
+        } else {
+            return (char)(_SMALL_A+digit);
+        }
+    } else {
+        return (char)((_ZERO_-26)+digit);
+    }
+}
+
+/**
+ * @return the numeric value of a basic code point (for use in representing integers)
+ *         in the range 0 to BASE-1, or a negative value if cp is invalid.
+ */
+static int32_t decodeDigit(int32_t cp) {
+    if(cp<=u'Z') {
+        if(cp<=u'9') {
+            if(cp<u'0') {
+                return -1;
+            } else {
+                return cp-u'0'+26;  // 0..9 -> 26..35
+            }
+        } else {
+            return cp-u'A';  // A-Z -> 0..25
+        }
+    } else if(cp<=u'z') {
+        return cp-'a';  // a..z -> 0..25
+    } else {
+        return -1;
+    }
+}
+
+static inline char
+asciiCaseMap(char b, UBool uppercase) {
+    if(uppercase) {
+        if(_SMALL_A<=b && b<=_SMALL_Z) {
+            b-=(_SMALL_A-_CAPITAL_A);
+        }
+    } else {
+        if(_CAPITAL_A<=b && b<=_CAPITAL_Z) {
+            b+=(_SMALL_A-_CAPITAL_A);
+        }
+    }
+    return b;
+}
+
+/* Punycode-specific Bootstring code ---------------------------------------- */
+
+/*
+ * The following code omits the {parts} of the pseudo-algorithm in the spec
+ * that are not used with the Punycode parameter set.
+ */
+
+/* Bias adaptation function. */
+static int32_t
+adaptBias(int32_t delta, int32_t length, UBool firstTime) {
+    int32_t count;
+
+    if(firstTime) {
+        delta/=DAMP;
+    } else {
+        delta/=2;
+    }
+
+    delta+=delta/length;
+    for(count=0; delta>((BASE-TMIN)*TMAX)/2; count+=BASE) {
+        delta/=(BASE-TMIN);
+    }
+
+    return count+(((BASE-TMIN+1)*delta)/(delta+SKEW));
+}
+
+namespace {
+
+// ICU-13727: Limit input length for n^2 algorithm
+// where well-formed strings are at most 59 characters long.
+constexpr int32_t ENCODE_MAX_CODE_UNITS=1000;
+constexpr int32_t DECODE_MAX_CHARS=2000;
+
+}  // namespace
+
+// encode
+U_CAPI int32_t
+u_strToPunycode(const UChar *src, int32_t srcLength,
+                UChar *dest, int32_t destCapacity,
+                const UBool *caseFlags,
+                UErrorCode *pErrorCode) {
+
+    int32_t cpBuffer[ENCODE_MAX_CODE_UNITS];
+    int32_t n, delta, handledCPCount, basicLength, destLength, bias, j, m, q, k, t, srcCPCount;
+    UChar c, c2;
+
+    /* argument checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    if (srcLength>ENCODE_MAX_CODE_UNITS) {
+        *pErrorCode=U_INPUT_TOO_LONG_ERROR;
+        return 0;
+    }
+
+    /*
+     * Handle the basic code points and
+     * convert extended ones to UTF-32 in cpBuffer (caseFlag in sign bit):
+     */
+    srcCPCount=destLength=0;
+    if(srcLength==-1) {
+        /* NUL-terminated input */
+        for(j=0; /* no condition */; ++j) {
+            if((c=src[j])==0) {
+                break;
+            }
+            if(j>=ENCODE_MAX_CODE_UNITS) {
+                *pErrorCode=U_INPUT_TOO_LONG_ERROR;
+                return 0;
+            }
+            if(IS_BASIC(c)) {
+                cpBuffer[srcCPCount++]=0;
+                if(destLength<destCapacity) {
+                    dest[destLength]=
+                        caseFlags!=NULL ?
+                            asciiCaseMap((char)c, caseFlags[j]) :
+                            (char)c;
+                }
+                ++destLength;
+            } else {
+                n=(caseFlags!=NULL && caseFlags[j])<<31L;
+                if(U16_IS_SINGLE(c)) {
+                    n|=c;
+                } else if(U16_IS_LEAD(c) && U16_IS_TRAIL(c2=src[j+1])) {
+                    ++j;
+                    n|=(int32_t)U16_GET_SUPPLEMENTARY(c, c2);
+                } else {
+                    /* error: unmatched surrogate */
+                    *pErrorCode=U_INVALID_CHAR_FOUND;
+                    return 0;
+                }
+                cpBuffer[srcCPCount++]=n;
+            }
+        }
+    } else {
+        /* length-specified input */
+        for(j=0; j<srcLength; ++j) {
+            c=src[j];
+            if(IS_BASIC(c)) {
+                cpBuffer[srcCPCount++]=0;
+                if(destLength<destCapacity) {
+                    dest[destLength]=
+                        caseFlags!=NULL ?
+                            asciiCaseMap((char)c, caseFlags[j]) :
+                            (char)c;
+                }
+                ++destLength;
+            } else {
+                n=(caseFlags!=NULL && caseFlags[j])<<31L;
+                if(U16_IS_SINGLE(c)) {
+                    n|=c;
+                } else if(U16_IS_LEAD(c) && (j+1)<srcLength && U16_IS_TRAIL(c2=src[j+1])) {
+                    ++j;
+                    n|=(int32_t)U16_GET_SUPPLEMENTARY(c, c2);
+                } else {
+                    /* error: unmatched surrogate */
+                    *pErrorCode=U_INVALID_CHAR_FOUND;
+                    return 0;
+                }
+                cpBuffer[srcCPCount++]=n;
+            }
+        }
+    }
+
+    /* Finish the basic string - if it is not empty - with a delimiter. */
+    basicLength=destLength;
+    if(basicLength>0) {
+        if(destLength<destCapacity) {
+            dest[destLength]=DELIMITER;
+        }
+        ++destLength;
+    }
+
+    /*
+     * handledCPCount is the number of code points that have been handled
+     * basicLength is the number of basic code points
+     * destLength is the number of chars that have been output
+     */
+
+    /* Initialize the state: */
+    n=INITIAL_N;
+    delta=0;
+    bias=INITIAL_BIAS;
+
+    /* Main encoding loop: */
+    for(handledCPCount=basicLength; handledCPCount<srcCPCount; /* no op */) {
+        /*
+         * All non-basic code points < n have been handled already.
+         * Find the next larger one:
+         */
+        for(m=0x7fffffff, j=0; j<srcCPCount; ++j) {
+            q=cpBuffer[j]&0x7fffffff; /* remove case flag from the sign bit */
+            if(n<=q && q<m) {
+                m=q;
+            }
+        }
+
+        /*
+         * Increase delta enough to advance the decoder's
+         * <n,i> state to <m,0>, but guard against overflow:
+         */
+        if(m-n>(0x7fffffff-handledCPCount-delta)/(handledCPCount+1)) {
+            *pErrorCode=U_INTERNAL_PROGRAM_ERROR;
+            return 0;
+        }
+        delta+=(m-n)*(handledCPCount+1);
+        n=m;
+
+        /* Encode a sequence of same code points n */
+        for(j=0; j<srcCPCount; ++j) {
+            q=cpBuffer[j]&0x7fffffff; /* remove case flag from the sign bit */
+            if(q<n) {
+                ++delta;
+            } else if(q==n) {
+                /* Represent delta as a generalized variable-length integer: */
+                for(q=delta, k=BASE; /* no condition */; k+=BASE) {
+
+                    /** RAM: comment out the old code for conformance with draft-ietf-idn-punycode-03.txt
+
+                    t=k-bias;
+                    if(t<TMIN) {
+                        t=TMIN;
+                    } else if(t>TMAX) {
+                        t=TMAX;
+                    }
+                    */
+
+                    t=k-bias;
+                    if(t<TMIN) {
+                        t=TMIN;
+                    } else if(k>=(bias+TMAX)) {
+                        t=TMAX;
+                    }
+
+                    if(q<t) {
+                        break;
+                    }
+
+                    if(destLength<destCapacity) {
+                        dest[destLength]=digitToBasic(t+(q-t)%(BASE-t), 0);
+                    }
+                    ++destLength;
+                    q=(q-t)/(BASE-t);
+                }
+
+                if(destLength<destCapacity) {
+                    dest[destLength]=digitToBasic(q, (UBool)(cpBuffer[j]<0));
+                }
+                ++destLength;
+                bias=adaptBias(delta, handledCPCount+1, (UBool)(handledCPCount==basicLength));
+                delta=0;
+                ++handledCPCount;
+            }
+        }
+
+        ++delta;
+        ++n;
+    }
+
+    return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
+}
+
+// decode
+U_CAPI int32_t
+u_strFromPunycode(const UChar *src, int32_t srcLength,
+                  UChar *dest, int32_t destCapacity,
+                  UBool *caseFlags,
+                  UErrorCode *pErrorCode) {
+    int32_t n, destLength, i, bias, basicLength, j, in, oldi, w, k, digit, t,
+            destCPCount, firstSupplementaryIndex, cpLength;
+    UChar b;
+
+    /* argument checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(src==NULL || srcLength<-1 || (dest==NULL && destCapacity!=0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(srcLength==-1) {
+        srcLength=u_strlen(src);
+    }
+    if (srcLength>DECODE_MAX_CHARS) {
+        *pErrorCode=U_INPUT_TOO_LONG_ERROR;
+        return 0;
+    }
+
+    /*
+     * Handle the basic code points:
+     * Let basicLength be the number of input code points
+     * before the last delimiter, or 0 if there is none,
+     * then copy the first basicLength code points to the output.
+     *
+     * The two following loops iterate backward.
+     */
+    for(j=srcLength; j>0;) {
+        if(src[--j]==DELIMITER) {
+            break;
+        }
+    }
+    destLength=basicLength=destCPCount=j;
+    U_ASSERT(destLength>=0);
+
+    while(j>0) {
+        b=src[--j];
+        if(!IS_BASIC(b)) {
+            *pErrorCode=U_INVALID_CHAR_FOUND;
+            return 0;
+        }
+
+        if(j<destCapacity) {
+            dest[j]=(UChar)b;
+
+            if(caseFlags!=NULL) {
+                caseFlags[j]=IS_BASIC_UPPERCASE(b);
+            }
+        }
+    }
+
+    /* Initialize the state: */
+    n=INITIAL_N;
+    i=0;
+    bias=INITIAL_BIAS;
+    firstSupplementaryIndex=1000000000;
+
+    /*
+     * Main decoding loop:
+     * Start just after the last delimiter if any
+     * basic code points were copied; start at the beginning otherwise.
+     */
+    for(in=basicLength>0 ? basicLength+1 : 0; in<srcLength; /* no op */) {
+        /*
+         * in is the index of the next character to be consumed, and
+         * destCPCount is the number of code points in the output array.
+         *
+         * Decode a generalized variable-length integer into delta,
+         * which gets added to i.  The overflow checking is easier
+         * if we increase i as we go, then subtract off its starting
+         * value at the end to obtain delta.
+         */
+        for(oldi=i, w=1, k=BASE; /* no condition */; k+=BASE) {
+            if(in>=srcLength) {
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                return 0;
+            }
+
+            digit=decodeDigit(src[in++]);
+            if(digit<0) {
+                *pErrorCode=U_INVALID_CHAR_FOUND;
+                return 0;
+            }
+            if(digit>(0x7fffffff-i)/w) {
+                /* integer overflow */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                return 0;
+            }
+
+            i+=digit*w;
+            /** RAM: comment out the old code for conformance with draft-ietf-idn-punycode-03.txt  
+            t=k-bias;
+            if(t<TMIN) {
+                t=TMIN;
+            } else if(t>TMAX) {
+                t=TMAX;
+            }
+            */
+            t=k-bias;
+            if(t<TMIN) {
+                t=TMIN;
+            } else if(k>=(bias+TMAX)) {
+                t=TMAX;
+            }
+            if(digit<t) {
+                break;
+            }
+
+            if(w>0x7fffffff/(BASE-t)) {
+                /* integer overflow */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                return 0;
+            }
+            w*=BASE-t;
+        }
+
+        /*
+         * Modification from sample code:
+         * Increments destCPCount here,
+         * where needed instead of in for() loop tail.
+         */
+        ++destCPCount;
+        bias=adaptBias(i-oldi, destCPCount, (UBool)(oldi==0));
+
+        /*
+         * i was supposed to wrap around from (incremented) destCPCount to 0,
+         * incrementing n each time, so we'll fix that now:
+         */
+        if(i/destCPCount>(0x7fffffff-n)) {
+            /* integer overflow */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            return 0;
+        }
+
+        n+=i/destCPCount;
+        i%=destCPCount;
+        /* not needed for Punycode: */
+        /* if (decode_digit(n) <= BASE) return punycode_invalid_input; */
+
+        if(n>0x10ffff || U_IS_SURROGATE(n)) {
+            /* Unicode code point overflow */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            return 0;
+        }
+
+        /* Insert n at position i of the output: */
+        cpLength=U16_LENGTH(n);
+        if(dest!=NULL && ((destLength+cpLength)<=destCapacity)) {
+            int32_t codeUnitIndex;
+
+            /*
+             * Handle indexes when supplementary code points are present.
+             *
+             * In almost all cases, there will be only BMP code points before i
+             * and even in the entire string.
+             * This is handled with the same efficiency as with UTF-32.
+             *
+             * Only the rare cases with supplementary code points are handled
+             * more slowly - but not too bad since this is an insertion anyway.
+             */
+            if(i<=firstSupplementaryIndex) {
+                codeUnitIndex=i;
+                if(cpLength>1) {
+                    firstSupplementaryIndex=codeUnitIndex;
+                } else {
+                    ++firstSupplementaryIndex;
+                }
+            } else {
+                codeUnitIndex=firstSupplementaryIndex;
+                U16_FWD_N(dest, codeUnitIndex, destLength, i-codeUnitIndex);
+            }
+
+            /* use the UChar index codeUnitIndex instead of the code point index i */
+            if(codeUnitIndex<destLength) {
+                uprv_memmove(dest+codeUnitIndex+cpLength,
+                             dest+codeUnitIndex,
+                             (destLength-codeUnitIndex)*U_SIZEOF_UCHAR);
+                if(caseFlags!=NULL) {
+                    uprv_memmove(caseFlags+codeUnitIndex+cpLength,
+                                 caseFlags+codeUnitIndex,
+                                 destLength-codeUnitIndex);
+                }
+            }
+            if(cpLength==1) {
+                /* BMP, insert one code unit */
+                dest[codeUnitIndex]=(UChar)n;
+            } else {
+                /* supplementary character, insert two code units */
+                dest[codeUnitIndex]=U16_LEAD(n);
+                dest[codeUnitIndex+1]=U16_TRAIL(n);
+            }
+            if(caseFlags!=NULL) {
+                /* Case of last character determines uppercase flag: */
+                caseFlags[codeUnitIndex]=IS_BASIC_UPPERCASE(src[in-1]);
+                if(cpLength==2) {
+                    caseFlags[codeUnitIndex+1]=FALSE;
+                }
+            }
+        }
+        destLength+=cpLength;
+        U_ASSERT(destLength>=0);
+        ++i;
+    }
+
+    return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
+}
+
+/* ### check notes on overflow handling - only necessary if not IDNA? are these Punycode functions to be public? */
+
+#endif /* #if !UCONFIG_NO_IDNA */
diff --git a/thirdparty/icu4c/common/punycode.h b/thirdparty/icu4c/common/punycode.h
new file mode 100644
index 0000000000..9e28f770c4
--- /dev/null
+++ b/thirdparty/icu4c/common/punycode.h
@@ -0,0 +1,120 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2003, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  punycode.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jan31
+*   created by: Markus W. Scherer
+*/
+
+/* This ICU code derived from: */
+/*
+punycode.c 0.4.0 (2001-Nov-17-Sat)
+http://www.cs.berkeley.edu/~amc/idn/
+Adam M. Costello
+http://www.nicemice.net/amc/
+*/
+
+#ifndef __PUNYCODE_H__
+#define __PUNYCODE_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+/**
+ * u_strToPunycode() converts Unicode to Punycode.
+ *
+ * The input string must not contain single, unpaired surrogates.
+ * The output will be represented as an array of ASCII code points.
+ *
+ * The output string is NUL-terminated according to normal ICU
+ * string output rules.
+ *
+ * @param src Input Unicode string.
+ *            This function handles a limited amount of code points
+ *            (the limit is >=64).
+ *            U_INDEX_OUTOFBOUNDS_ERROR is set if the limit is exceeded.
+ * @param srcLength Number of UChars in src, or -1 if NUL-terminated.
+ * @param dest Output Punycode array.
+ * @param destCapacity Size of dest.
+ * @param caseFlags Vector of boolean values, one per input UChar,
+ *                  indicating that the corresponding character is to be
+ *                  marked for the decoder optionally
+ *                  uppercasing (true) or lowercasing (false)
+ *                  the character.
+ *                  ASCII characters are output directly in the case as marked.
+ *                  Flags corresponding to trail surrogates are ignored.
+ *                  If caseFlags==NULL then input characters are not
+ *                  case-mapped.
+ * @param pErrorCode ICU in/out error code parameter.
+ *                   U_INVALID_CHAR_FOUND if src contains
+ *                   unmatched single surrogates.
+ *                   U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                   too many code points.
+ * @return Number of ASCII characters in puny.
+ *
+ * @see u_strFromPunycode
+ */
+U_CAPI int32_t
+u_strToPunycode(const UChar *src, int32_t srcLength,
+                UChar *dest, int32_t destCapacity,
+                const UBool *caseFlags,
+                UErrorCode *pErrorCode);
+
+/**
+ * u_strFromPunycode() converts Punycode to Unicode.
+ * The Unicode string will be at most as long (in UChars)
+ * than the Punycode string (in chars).
+ *
+ * @param src Input Punycode string.
+ * @param srcLength Length of puny, or -1 if NUL-terminated
+ * @param dest Output Unicode string buffer.
+ * @param destCapacity Size of dest in number of UChars,
+ *                     and of caseFlags in numbers of UBools.
+ * @param caseFlags Output array for case flags as
+ *                  defined by the Punycode string.
+ *                  The caller should uppercase (true) or lowercase (FASLE)
+ *                  the corresponding character in dest.
+ *                  For supplementary characters, only the lead surrogate
+ *                  is marked, and false is stored for the trail surrogate.
+ *                  This is redundant and not necessary for ASCII characters
+ *                  because they are already in the case indicated.
+ *                  Can be NULL if the case flags are not needed.
+ * @param pErrorCode ICU in/out error code parameter.
+ *                   U_INVALID_CHAR_FOUND if a non-ASCII character
+ *                   precedes the last delimiter ('-'),
+ *                   or if an invalid character (not a-zA-Z0-9) is found
+ *                   after the last delimiter.
+ *                   U_ILLEGAL_CHAR_FOUND if the delta sequence is ill-formed.
+ * @return Number of UChars written to dest.
+ *
+ * @see u_strToPunycode
+ */
+U_CAPI int32_t
+u_strFromPunycode(const UChar *src, int32_t srcLength,
+                  UChar *dest, int32_t destCapacity,
+                  UBool *caseFlags,
+                  UErrorCode *pErrorCode);
+
+#endif /* #if !UCONFIG_NO_IDNA */
+
+#endif
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/putil.cpp b/thirdparty/icu4c/common/putil.cpp
new file mode 100644
index 0000000000..3ed6a05d22
--- /dev/null
+++ b/thirdparty/icu4c/common/putil.cpp
@@ -0,0 +1,2482 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : putil.c (previously putil.cpp and ptypes.cpp)
+*
+*   Date        Name        Description
+*   04/14/97    aliu        Creation.
+*   04/24/97    aliu        Added getDefaultDataDirectory() and
+*                            getDefaultLocaleID().
+*   04/28/97    aliu        Rewritten to assume Unix and apply general methods
+*                            for assumed case.  Non-UNIX platforms must be
+*                            special-cased.  Rewrote numeric methods dealing
+*                            with NaN and Infinity to be platform independent
+*                             over all IEEE 754 platforms.
+*   05/13/97    aliu        Restored sign of timezone
+*                            (semantics are hours West of GMT)
+*   06/16/98    erm         Added IEEE_754 stuff, cleaned up isInfinite, isNan,
+*                             nextDouble..
+*   07/22/98    stephen     Added remainder, max, min, trunc
+*   08/13/98    stephen     Added isNegativeInfinity, isPositiveInfinity
+*   08/24/98    stephen     Added longBitsFromDouble
+*   09/08/98    stephen     Minor changes for Mac Port
+*   03/02/99    stephen     Removed openFile().  Added AS400 support.
+*                            Fixed EBCDIC tables
+*   04/15/99    stephen     Converted to C.
+*   06/28/99    stephen     Removed mutex locking in u_isBigEndian().
+*   08/04/99    jeffrey R.  Added OS/2 changes
+*   11/15/99    helena      Integrated S/390 IEEE support.
+*   04/26/01    Barry N.    OS/400 support for uprv_getDefaultLocaleID
+*   08/15/01    Steven H.   OS/400 support for uprv_getDefaultCodepage
+*   01/03/08    Steven L.   Fake Time Support
+******************************************************************************
+*/
+
+// Defines _XOPEN_SOURCE for access to POSIX functions.
+// Must be before any other #includes.
+#include "uposixdefs.h"
+
+// First, the platform type. Need this for U_PLATFORM.
+#include "unicode/platform.h"
+
+#if U_PLATFORM == U_PF_MINGW && defined __STRICT_ANSI__
+/* tzset isn't defined in strict ANSI on MinGW. */
+#undef __STRICT_ANSI__
+#endif
+
+/*
+ * Cygwin with GCC requires inclusion of time.h after the above disabling strict asci mode statement.
+ */
+#include <time.h>
+
+#if !U_PLATFORM_USES_ONLY_WIN32_API
+#include <sys/time.h>
+#endif
+
+/* include the rest of the ICU headers */
+#include "unicode/putil.h"
+#include "unicode/ustring.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include "umutex.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "locmap.h"
+#include "ucln_cmn.h"
+#include "charstr.h"
+
+/* Include standard headers. */
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <math.h>
+#include <locale.h>
+#include <float.h>
+
+#ifndef U_COMMON_IMPLEMENTATION
+#error U_COMMON_IMPLEMENTATION not set - must be set for all ICU source files in common/ - see https://unicode-org.github.io/icu/userguide/howtouseicu
+#endif
+
+
+/* include system headers */
+#if U_PLATFORM_USES_ONLY_WIN32_API
+    /*
+     * TODO: U_PLATFORM_USES_ONLY_WIN32_API includes MinGW.
+     * Should Cygwin be included as well (U_PLATFORM_HAS_WIN32_API)
+     * to use native APIs as much as possible?
+     */
+#ifndef WIN32_LEAN_AND_MEAN
+#   define WIN32_LEAN_AND_MEAN
+#endif
+#   define VC_EXTRALEAN
+#   define NOUSER
+#   define NOSERVICE
+#   define NOIME
+#   define NOMCX
+#   include <windows.h>
+#   include "unicode/uloc.h"
+#   include "wintz.h"
+#elif U_PLATFORM == U_PF_OS400
+#   include <float.h>
+#   include <qusec.h>       /* error code structure */
+#   include <qusrjobi.h>
+#   include <qliept.h>      /* EPT_CALL macro  - this include must be after all other "QSYSINCs" */
+#   include <mih/testptr.h> /* For uprv_maximumPtr */
+#elif U_PLATFORM == U_PF_OS390
+#   include "unicode/ucnv.h"   /* Needed for UCNV_SWAP_LFNL_OPTION_STRING */
+#elif U_PLATFORM_IS_DARWIN_BASED || U_PLATFORM_IS_LINUX_BASED || U_PLATFORM == U_PF_BSD || U_PLATFORM == U_PF_SOLARIS
+#   include <limits.h>
+#   include <unistd.h>
+#   if U_PLATFORM == U_PF_SOLARIS
+#       ifndef _XPG4_2
+#           define _XPG4_2
+#       endif
+#   elif U_PLATFORM == U_PF_ANDROID
+#       include <sys/system_properties.h>
+#       include <dlfcn.h>
+#   endif
+#elif U_PLATFORM == U_PF_QNX
+#   include <sys/neutrino.h>
+#endif
+
+
+/*
+ * Only include langinfo.h if we have a way to get the codeset. If we later
+ * depend on more feature, we can test on U_HAVE_NL_LANGINFO.
+ *
+ */
+
+#if U_HAVE_NL_LANGINFO_CODESET
+#include <langinfo.h>
+#endif
+
+/**
+ * Simple things (presence of functions, etc) should just go in configure.in and be added to
+ * icucfg.h via autoheader.
+ */
+#if U_PLATFORM_IMPLEMENTS_POSIX
+#   if U_PLATFORM == U_PF_OS400
+#    define HAVE_DLFCN_H 0
+#    define HAVE_DLOPEN 0
+#   else
+#   ifndef HAVE_DLFCN_H
+#    define HAVE_DLFCN_H 1
+#   endif
+#   ifndef HAVE_DLOPEN
+#    define HAVE_DLOPEN 1
+#   endif
+#   endif
+#   ifndef HAVE_GETTIMEOFDAY
+#    define HAVE_GETTIMEOFDAY 1
+#   endif
+#else
+#   define HAVE_DLFCN_H 0
+#   define HAVE_DLOPEN 0
+#   define HAVE_GETTIMEOFDAY 0
+#endif
+
+U_NAMESPACE_USE
+
+/* Define the extension for data files, again... */
+#define DATA_TYPE "dat"
+
+/* Leave this copyright notice here! */
+static const char copyright[] = U_COPYRIGHT_STRING;
+
+/* floating point implementations ------------------------------------------- */
+
+/* We return QNAN rather than SNAN*/
+#define SIGN 0x80000000U
+
+/* Make it easy to define certain types of constants */
+typedef union {
+    int64_t i64; /* This must be defined first in order to allow the initialization to work. This is a C89 feature. */
+    double d64;
+} BitPatternConversion;
+static const BitPatternConversion gNan = { (int64_t) INT64_C(0x7FF8000000000000) };
+static const BitPatternConversion gInf = { (int64_t) INT64_C(0x7FF0000000000000) };
+
+/*---------------------------------------------------------------------------
+  Platform utilities
+  Our general strategy is to assume we're on a POSIX platform.  Platforms which
+  are non-POSIX must declare themselves so.  The default POSIX implementation
+  will sometimes work for non-POSIX platforms as well (e.g., the NaN-related
+  functions).
+  ---------------------------------------------------------------------------*/
+
+#if U_PLATFORM_USES_ONLY_WIN32_API || U_PLATFORM == U_PF_OS400
+#   undef U_POSIX_LOCALE
+#else
+#   define U_POSIX_LOCALE    1
+#endif
+
+/*
+    WARNING! u_topNBytesOfDouble and u_bottomNBytesOfDouble
+    can't be properly optimized by the gcc compiler sometimes (i.e. gcc 3.2).
+*/
+#if !IEEE_754
+static char*
+u_topNBytesOfDouble(double* d, int n)
+{
+#if U_IS_BIG_ENDIAN
+    return (char*)d;
+#else
+    return (char*)(d + 1) - n;
+#endif
+}
+
+static char*
+u_bottomNBytesOfDouble(double* d, int n)
+{
+#if U_IS_BIG_ENDIAN
+    return (char*)(d + 1) - n;
+#else
+    return (char*)d;
+#endif
+}
+#endif   /* !IEEE_754 */
+
+#if IEEE_754
+static UBool
+u_signBit(double d) {
+    uint8_t hiByte;
+#if U_IS_BIG_ENDIAN
+    hiByte = *(uint8_t *)&d;
+#else
+    hiByte = *(((uint8_t *)&d) + sizeof(double) - 1);
+#endif
+    return (hiByte & 0x80) != 0;
+}
+#endif
+
+
+
+#if defined (U_DEBUG_FAKETIME)
+/* Override the clock to test things without having to move the system clock.
+ * Assumes POSIX gettimeofday() will function
+ */
+UDate fakeClock_t0 = 0; /** Time to start the clock from **/
+UDate fakeClock_dt = 0; /** Offset (fake time - real time) **/
+UBool fakeClock_set = FALSE; /** True if fake clock has spun up **/
+
+static UDate getUTCtime_real() {
+    struct timeval posixTime;
+    gettimeofday(&posixTime, NULL);
+    return (UDate)(((int64_t)posixTime.tv_sec * U_MILLIS_PER_SECOND) + (posixTime.tv_usec/1000));
+}
+
+static UDate getUTCtime_fake() {
+    static UMutex fakeClockMutex;
+    umtx_lock(&fakeClockMutex);
+    if(!fakeClock_set) {
+        UDate real = getUTCtime_real();
+        const char *fake_start = getenv("U_FAKETIME_START");
+        if((fake_start!=NULL) && (fake_start[0]!=0)) {
+            sscanf(fake_start,"%lf",&fakeClock_t0);
+            fakeClock_dt = fakeClock_t0 - real;
+            fprintf(stderr,"U_DEBUG_FAKETIME was set at compile time, so the ICU clock will start at a preset value\n"
+                    "env variable U_FAKETIME_START=%.0f (%s) for an offset of %.0f ms from the current time %.0f\n",
+                    fakeClock_t0, fake_start, fakeClock_dt, real);
+        } else {
+          fakeClock_dt = 0;
+            fprintf(stderr,"U_DEBUG_FAKETIME was set at compile time, but U_FAKETIME_START was not set.\n"
+                    "Set U_FAKETIME_START to the number of milliseconds since 1/1/1970 to set the ICU clock.\n");
+        }
+        fakeClock_set = TRUE;
+    }
+    umtx_unlock(&fakeClockMutex);
+
+    return getUTCtime_real() + fakeClock_dt;
+}
+#endif
+
+#if U_PLATFORM_USES_ONLY_WIN32_API
+typedef union {
+    int64_t int64;
+    FILETIME fileTime;
+} FileTimeConversion;   /* This is like a ULARGE_INTEGER */
+
+/* Number of 100 nanoseconds from 1/1/1601 to 1/1/1970 */
+#define EPOCH_BIAS  INT64_C(116444736000000000)
+#define HECTONANOSECOND_PER_MILLISECOND   10000
+
+#endif
+
+/*---------------------------------------------------------------------------
+  Universal Implementations
+  These are designed to work on all platforms.  Try these, and if they
+  don't work on your platform, then special case your platform with new
+  implementations.
+---------------------------------------------------------------------------*/
+
+U_CAPI UDate U_EXPORT2
+uprv_getUTCtime()
+{
+#if defined(U_DEBUG_FAKETIME)
+    return getUTCtime_fake(); /* Hook for overriding the clock */
+#else
+    return uprv_getRawUTCtime();
+#endif
+}
+
+/* Return UTC (GMT) time measured in milliseconds since 0:00 on 1/1/70.*/
+U_CAPI UDate U_EXPORT2
+uprv_getRawUTCtime()
+{
+#if U_PLATFORM_USES_ONLY_WIN32_API
+
+    FileTimeConversion winTime;
+    GetSystemTimeAsFileTime(&winTime.fileTime);
+    return (UDate)((winTime.int64 - EPOCH_BIAS) / HECTONANOSECOND_PER_MILLISECOND);
+#else
+
+#if HAVE_GETTIMEOFDAY
+    struct timeval posixTime;
+    gettimeofday(&posixTime, NULL);
+    return (UDate)(((int64_t)posixTime.tv_sec * U_MILLIS_PER_SECOND) + (posixTime.tv_usec/1000));
+#else
+    time_t epochtime;
+    time(&epochtime);
+    return (UDate)epochtime * U_MILLIS_PER_SECOND;
+#endif
+
+#endif
+}
+
+/*-----------------------------------------------------------------------------
+  IEEE 754
+  These methods detect and return NaN and infinity values for doubles
+  conforming to IEEE 754.  Platforms which support this standard include X86,
+  Mac 680x0, Mac PowerPC, AIX RS/6000, and most others.
+  If this doesn't work on your platform, you have non-IEEE floating-point, and
+  will need to code your own versions.  A naive implementation is to return 0.0
+  for getNaN and getInfinity, and false for isNaN and isInfinite.
+  ---------------------------------------------------------------------------*/
+
+U_CAPI UBool U_EXPORT2
+uprv_isNaN(double number)
+{
+#if IEEE_754
+    BitPatternConversion convertedNumber;
+    convertedNumber.d64 = number;
+    /* Infinity is 0x7FF0000000000000U. Anything greater than that is a NaN */
+    return (UBool)((convertedNumber.i64 & U_INT64_MAX) > gInf.i64);
+
+#elif U_PLATFORM == U_PF_OS390
+    uint32_t highBits = *(uint32_t*)u_topNBytesOfDouble(&number,
+                        sizeof(uint32_t));
+    uint32_t lowBits  = *(uint32_t*)u_bottomNBytesOfDouble(&number,
+                        sizeof(uint32_t));
+
+    return ((highBits & 0x7F080000L) == 0x7F080000L) &&
+      (lowBits == 0x00000000L);
+
+#else
+    /* If your platform doesn't support IEEE 754 but *does* have an NaN value,*/
+    /* you'll need to replace this default implementation with what's correct*/
+    /* for your platform.*/
+    return number != number;
+#endif
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_isInfinite(double number)
+{
+#if IEEE_754
+    BitPatternConversion convertedNumber;
+    convertedNumber.d64 = number;
+    /* Infinity is exactly 0x7FF0000000000000U. */
+    return (UBool)((convertedNumber.i64 & U_INT64_MAX) == gInf.i64);
+#elif U_PLATFORM == U_PF_OS390
+    uint32_t highBits = *(uint32_t*)u_topNBytesOfDouble(&number,
+                        sizeof(uint32_t));
+    uint32_t lowBits  = *(uint32_t*)u_bottomNBytesOfDouble(&number,
+                        sizeof(uint32_t));
+
+    return ((highBits  & ~SIGN) == 0x70FF0000L) && (lowBits == 0x00000000L);
+
+#else
+    /* If your platform doesn't support IEEE 754 but *does* have an infinity*/
+    /* value, you'll need to replace this default implementation with what's*/
+    /* correct for your platform.*/
+    return number == (2.0 * number);
+#endif
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_isPositiveInfinity(double number)
+{
+#if IEEE_754 || U_PLATFORM == U_PF_OS390
+    return (UBool)(number > 0 && uprv_isInfinite(number));
+#else
+    return uprv_isInfinite(number);
+#endif
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_isNegativeInfinity(double number)
+{
+#if IEEE_754 || U_PLATFORM == U_PF_OS390
+    return (UBool)(number < 0 && uprv_isInfinite(number));
+
+#else
+    uint32_t highBits = *(uint32_t*)u_topNBytesOfDouble(&number,
+                        sizeof(uint32_t));
+    return((highBits & SIGN) && uprv_isInfinite(number));
+
+#endif
+}
+
+U_CAPI double U_EXPORT2
+uprv_getNaN()
+{
+#if IEEE_754 || U_PLATFORM == U_PF_OS390
+    return gNan.d64;
+#else
+    /* If your platform doesn't support IEEE 754 but *does* have an NaN value,*/
+    /* you'll need to replace this default implementation with what's correct*/
+    /* for your platform.*/
+    return 0.0;
+#endif
+}
+
+U_CAPI double U_EXPORT2
+uprv_getInfinity()
+{
+#if IEEE_754 || U_PLATFORM == U_PF_OS390
+    return gInf.d64;
+#else
+    /* If your platform doesn't support IEEE 754 but *does* have an infinity*/
+    /* value, you'll need to replace this default implementation with what's*/
+    /* correct for your platform.*/
+    return 0.0;
+#endif
+}
+
+U_CAPI double U_EXPORT2
+uprv_floor(double x)
+{
+    return floor(x);
+}
+
+U_CAPI double U_EXPORT2
+uprv_ceil(double x)
+{
+    return ceil(x);
+}
+
+U_CAPI double U_EXPORT2
+uprv_round(double x)
+{
+    return uprv_floor(x + 0.5);
+}
+
+U_CAPI double U_EXPORT2
+uprv_fabs(double x)
+{
+    return fabs(x);
+}
+
+U_CAPI double U_EXPORT2
+uprv_modf(double x, double* y)
+{
+    return modf(x, y);
+}
+
+U_CAPI double U_EXPORT2
+uprv_fmod(double x, double y)
+{
+    return fmod(x, y);
+}
+
+U_CAPI double U_EXPORT2
+uprv_pow(double x, double y)
+{
+    /* This is declared as "double pow(double x, double y)" */
+    return pow(x, y);
+}
+
+U_CAPI double U_EXPORT2
+uprv_pow10(int32_t x)
+{
+    return pow(10.0, (double)x);
+}
+
+U_CAPI double U_EXPORT2
+uprv_fmax(double x, double y)
+{
+#if IEEE_754
+    /* first handle NaN*/
+    if(uprv_isNaN(x) || uprv_isNaN(y))
+        return uprv_getNaN();
+
+    /* check for -0 and 0*/
+    if(x == 0.0 && y == 0.0 && u_signBit(x))
+        return y;
+
+#endif
+
+    /* this should work for all flt point w/o NaN and Inf special cases */
+    return (x > y ? x : y);
+}
+
+U_CAPI double U_EXPORT2
+uprv_fmin(double x, double y)
+{
+#if IEEE_754
+    /* first handle NaN*/
+    if(uprv_isNaN(x) || uprv_isNaN(y))
+        return uprv_getNaN();
+
+    /* check for -0 and 0*/
+    if(x == 0.0 && y == 0.0 && u_signBit(y))
+        return y;
+
+#endif
+
+    /* this should work for all flt point w/o NaN and Inf special cases */
+    return (x > y ? y : x);
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_add32_overflow(int32_t a, int32_t b, int32_t* res) {
+    // NOTE: Some compilers (GCC, Clang) have primitives available, like __builtin_add_overflow.
+    // This function could be optimized by calling one of those primitives.
+    auto a64 = static_cast<int64_t>(a);
+    auto b64 = static_cast<int64_t>(b);
+    int64_t res64 = a64 + b64;
+    *res = static_cast<int32_t>(res64);
+    return res64 != *res;
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_mul32_overflow(int32_t a, int32_t b, int32_t* res) {
+    // NOTE: Some compilers (GCC, Clang) have primitives available, like __builtin_mul_overflow.
+    // This function could be optimized by calling one of those primitives.
+    auto a64 = static_cast<int64_t>(a);
+    auto b64 = static_cast<int64_t>(b);
+    int64_t res64 = a64 * b64;
+    *res = static_cast<int32_t>(res64);
+    return res64 != *res;
+}
+
+/**
+ * Truncates the given double.
+ * trunc(3.3) = 3.0, trunc (-3.3) = -3.0
+ * This is different than calling floor() or ceil():
+ * floor(3.3) = 3, floor(-3.3) = -4
+ * ceil(3.3) = 4, ceil(-3.3) = -3
+ */
+U_CAPI double U_EXPORT2
+uprv_trunc(double d)
+{
+#if IEEE_754
+    /* handle error cases*/
+    if(uprv_isNaN(d))
+        return uprv_getNaN();
+    if(uprv_isInfinite(d))
+        return uprv_getInfinity();
+
+    if(u_signBit(d))    /* Signbit() picks up -0.0;  d<0 does not. */
+        return ceil(d);
+    else
+        return floor(d);
+
+#else
+    return d >= 0 ? floor(d) : ceil(d);
+
+#endif
+}
+
+/**
+ * Return the largest positive number that can be represented by an integer
+ * type of arbitrary bit length.
+ */
+U_CAPI double U_EXPORT2
+uprv_maxMantissa(void)
+{
+    return pow(2.0, DBL_MANT_DIG + 1.0) - 1.0;
+}
+
+U_CAPI double U_EXPORT2
+uprv_log(double d)
+{
+    return log(d);
+}
+
+U_CAPI void * U_EXPORT2
+uprv_maximumPtr(void * base)
+{
+#if U_PLATFORM == U_PF_OS400
+    /*
+     * With the provided function we should never be out of range of a given segment
+     * (a traditional/typical segment that is).  Our segments have 5 bytes for the
+     * id and 3 bytes for the offset.  The key is that the casting takes care of
+     * only retrieving the offset portion minus x1000.  Hence, the smallest offset
+     * seen in a program is x001000 and when casted to an int would be 0.
+     * That's why we can only add 0xffefff.  Otherwise, we would exceed the segment.
+     *
+     * Currently, 16MB is the current addressing limitation on i5/OS if the activation is
+     * non-TERASPACE.  If it is TERASPACE it is 2GB - 4k(header information).
+     * This function determines the activation based on the pointer that is passed in and
+     * calculates the appropriate maximum available size for
+     * each pointer type (TERASPACE and non-TERASPACE)
+     *
+     * Unlike other operating systems, the pointer model isn't determined at
+     * compile time on i5/OS.
+     */
+    if ((base != NULL) && (_TESTPTR(base, _C_TERASPACE_CHECK))) {
+        /* if it is a TERASPACE pointer the max is 2GB - 4k */
+        return ((void *)(((char *)base)-((uint32_t)(base))+((uint32_t)0x7fffefff)));
+    }
+    /* otherwise 16MB since NULL ptr is not checkable or the ptr is not TERASPACE */
+    return ((void *)(((char *)base)-((uint32_t)(base))+((uint32_t)0xffefff)));
+
+#else
+    return U_MAX_PTR(base);
+#endif
+}
+
+/*---------------------------------------------------------------------------
+  Platform-specific Implementations
+  Try these, and if they don't work on your platform, then special case your
+  platform with new implementations.
+  ---------------------------------------------------------------------------*/
+
+/* Generic time zone layer -------------------------------------------------- */
+
+/* Time zone utilities */
+U_CAPI void U_EXPORT2
+uprv_tzset()
+{
+#if defined(U_TZSET)
+    U_TZSET();
+#else
+    /* no initialization*/
+#endif
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_timezone()
+{
+#ifdef U_TIMEZONE
+    return U_TIMEZONE;
+#else
+    time_t t, t1, t2;
+    struct tm tmrec;
+    int32_t tdiff = 0;
+
+    time(&t);
+    uprv_memcpy( &tmrec, localtime(&t), sizeof(tmrec) );
+#if U_PLATFORM != U_PF_IPHONE
+    UBool dst_checked = (tmrec.tm_isdst != 0); /* daylight savings time is checked*/
+#endif
+    t1 = mktime(&tmrec);                 /* local time in seconds*/
+    uprv_memcpy( &tmrec, gmtime(&t), sizeof(tmrec) );
+    t2 = mktime(&tmrec);                 /* GMT (or UTC) in seconds*/
+    tdiff = t2 - t1;
+
+#if U_PLATFORM != U_PF_IPHONE
+    /* imitate NT behaviour, which returns same timezone offset to GMT for
+       winter and summer.
+       This does not work on all platforms. For instance, on glibc on Linux
+       and on Mac OS 10.5, tdiff calculated above remains the same
+       regardless of whether DST is in effect or not. iOS is another
+       platform where this does not work. Linux + glibc and Mac OS 10.5
+       have U_TIMEZONE defined so that this code is not reached.
+    */
+    if (dst_checked)
+        tdiff += 3600;
+#endif
+    return tdiff;
+#endif
+}
+
+/* Note that U_TZNAME does *not* have to be tzname, but if it is,
+   some platforms need to have it declared here. */
+
+#if defined(U_TZNAME) && (U_PLATFORM == U_PF_IRIX || U_PLATFORM_IS_DARWIN_BASED)
+/* RS6000 and others reject char **tzname.  */
+extern U_IMPORT char *U_TZNAME[];
+#endif
+
+#if !UCONFIG_NO_FILE_IO && ((U_PLATFORM_IS_DARWIN_BASED && (U_PLATFORM != U_PF_IPHONE || defined(U_TIMEZONE))) || U_PLATFORM_IS_LINUX_BASED || U_PLATFORM == U_PF_BSD || U_PLATFORM == U_PF_SOLARIS)
+/* These platforms are likely to use Olson timezone IDs. */
+/* common targets of the symbolic link at TZDEFAULT are:
+ * "/usr/share/zoneinfo/<olsonID>" default, older Linux distros, macOS to 10.12
+ * "../usr/share/zoneinfo/<olsonID>" newer Linux distros: Red Hat Enterprise Linux 7, Ubuntu 16, SuSe Linux 12
+ * "/usr/share/lib/zoneinfo/<olsonID>" Solaris
+ * "../usr/share/lib/zoneinfo/<olsonID>" Solaris
+ * "/var/db/timezone/zoneinfo/<olsonID>" macOS 10.13
+ * To avoid checking lots of paths, just check that the target path
+ * before the <olsonID> ends with "/zoneinfo/", and the <olsonID> is valid.
+ */
+
+#define CHECK_LOCALTIME_LINK 1
+#if U_PLATFORM_IS_DARWIN_BASED
+#include <tzfile.h>
+#define TZZONEINFO      (TZDIR "/")
+#elif U_PLATFORM == U_PF_SOLARIS
+#define TZDEFAULT       "/etc/localtime"
+#define TZZONEINFO      "/usr/share/lib/zoneinfo/"
+#define TZ_ENV_CHECK    "localtime"
+#else
+#define TZDEFAULT       "/etc/localtime"
+#define TZZONEINFO      "/usr/share/zoneinfo/"
+#endif
+#define TZZONEINFOTAIL  "/zoneinfo/"
+#if U_HAVE_DIRENT_H
+#define TZFILE_SKIP     "posixrules" /* tz file to skip when searching. */
+/* Some Linux distributions have 'localtime' in /usr/share/zoneinfo
+   symlinked to /etc/localtime, which makes searchForTZFile return
+   'localtime' when it's the first match. */
+#define TZFILE_SKIP2    "localtime"
+#define SEARCH_TZFILE
+#include <dirent.h>  /* Needed to search through system timezone files */
+#endif
+static char gTimeZoneBuffer[PATH_MAX];
+static char *gTimeZoneBufferPtr = NULL;
+#endif
+
+#if !U_PLATFORM_USES_ONLY_WIN32_API
+#define isNonDigit(ch) (ch < '0' || '9' < ch)
+static UBool isValidOlsonID(const char *id) {
+    int32_t idx = 0;
+
+    /* Determine if this is something like Iceland (Olson ID)
+    or AST4ADT (non-Olson ID) */
+    while (id[idx] && isNonDigit(id[idx]) && id[idx] != ',') {
+        idx++;
+    }
+
+    /* If we went through the whole string, then it might be okay.
+    The timezone is sometimes set to "CST-7CDT", "CST6CDT5,J129,J131/19:30",
+    "GRNLNDST3GRNLNDDT" or similar, so we cannot use it.
+    The rest of the time it could be an Olson ID. George */
+    return (UBool)(id[idx] == 0
+        || uprv_strcmp(id, "PST8PDT") == 0
+        || uprv_strcmp(id, "MST7MDT") == 0
+        || uprv_strcmp(id, "CST6CDT") == 0
+        || uprv_strcmp(id, "EST5EDT") == 0);
+}
+
+/* On some Unix-like OS, 'posix' subdirectory in
+   /usr/share/zoneinfo replicates the top-level contents. 'right'
+   subdirectory has the same set of files, but individual files
+   are different from those in the top-level directory or 'posix'
+   because 'right' has files for TAI (Int'l Atomic Time) while 'posix'
+   has files for UTC.
+   When the first match for /etc/localtime is in either of them
+   (usually in posix because 'right' has different file contents),
+   or TZ environment variable points to one of them, createTimeZone
+   fails because, say, 'posix/America/New_York' is not an Olson
+   timezone id ('America/New_York' is). So, we have to skip
+   'posix/' and 'right/' at the beginning. */
+static void skipZoneIDPrefix(const char** id) {
+    if (uprv_strncmp(*id, "posix/", 6) == 0
+        || uprv_strncmp(*id, "right/", 6) == 0)
+    {
+        *id += 6;
+    }
+}
+#endif
+
+#if defined(U_TZNAME) && !U_PLATFORM_USES_ONLY_WIN32_API
+
+#define CONVERT_HOURS_TO_SECONDS(offset) (int32_t)(offset*3600)
+typedef struct OffsetZoneMapping {
+    int32_t offsetSeconds;
+    int32_t daylightType; /* 0=U_DAYLIGHT_NONE, 1=daylight in June-U_DAYLIGHT_JUNE, 2=daylight in December=U_DAYLIGHT_DECEMBER*/
+    const char *stdID;
+    const char *dstID;
+    const char *olsonID;
+} OffsetZoneMapping;
+
+enum { U_DAYLIGHT_NONE=0,U_DAYLIGHT_JUNE=1,U_DAYLIGHT_DECEMBER=2 };
+
+/*
+This list tries to disambiguate a set of abbreviated timezone IDs and offsets
+and maps it to an Olson ID.
+Before adding anything to this list, take a look at
+icu/source/tools/tzcode/tz.alias
+Sometimes no daylight savings (0) is important to define due to aliases.
+This list can be tested with icu/source/test/compat/tzone.pl
+More values could be added to daylightType to increase precision.
+*/
+static const struct OffsetZoneMapping OFFSET_ZONE_MAPPINGS[] = {
+    {-45900, 2, "CHAST", "CHADT", "Pacific/Chatham"},
+    {-43200, 1, "PETT", "PETST", "Asia/Kamchatka"},
+    {-43200, 2, "NZST", "NZDT", "Pacific/Auckland"},
+    {-43200, 1, "ANAT", "ANAST", "Asia/Anadyr"},
+    {-39600, 1, "MAGT", "MAGST", "Asia/Magadan"},
+    {-37800, 2, "LHST", "LHST", "Australia/Lord_Howe"},
+    {-36000, 2, "EST", "EST", "Australia/Sydney"},
+    {-36000, 1, "SAKT", "SAKST", "Asia/Sakhalin"},
+    {-36000, 1, "VLAT", "VLAST", "Asia/Vladivostok"},
+    {-34200, 2, "CST", "CST", "Australia/South"},
+    {-32400, 1, "YAKT", "YAKST", "Asia/Yakutsk"},
+    {-32400, 1, "CHOT", "CHOST", "Asia/Choibalsan"},
+    {-31500, 2, "CWST", "CWST", "Australia/Eucla"},
+    {-28800, 1, "IRKT", "IRKST", "Asia/Irkutsk"},
+    {-28800, 1, "ULAT", "ULAST", "Asia/Ulaanbaatar"},
+    {-28800, 2, "WST", "WST", "Australia/West"},
+    {-25200, 1, "HOVT", "HOVST", "Asia/Hovd"},
+    {-25200, 1, "KRAT", "KRAST", "Asia/Krasnoyarsk"},
+    {-21600, 1, "NOVT", "NOVST", "Asia/Novosibirsk"},
+    {-21600, 1, "OMST", "OMSST", "Asia/Omsk"},
+    {-18000, 1, "YEKT", "YEKST", "Asia/Yekaterinburg"},
+    {-14400, 1, "SAMT", "SAMST", "Europe/Samara"},
+    {-14400, 1, "AMT", "AMST", "Asia/Yerevan"},
+    {-14400, 1, "AZT", "AZST", "Asia/Baku"},
+    {-10800, 1, "AST", "ADT", "Asia/Baghdad"},
+    {-10800, 1, "MSK", "MSD", "Europe/Moscow"},
+    {-10800, 1, "VOLT", "VOLST", "Europe/Volgograd"},
+    {-7200, 0, "EET", "CEST", "Africa/Tripoli"},
+    {-7200, 1, "EET", "EEST", "Europe/Athens"}, /* Conflicts with Africa/Cairo */
+    {-7200, 1, "IST", "IDT", "Asia/Jerusalem"},
+    {-3600, 0, "CET", "WEST", "Africa/Algiers"},
+    {-3600, 2, "WAT", "WAST", "Africa/Windhoek"},
+    {0, 1, "GMT", "IST", "Europe/Dublin"},
+    {0, 1, "GMT", "BST", "Europe/London"},
+    {0, 0, "WET", "WEST", "Africa/Casablanca"},
+    {0, 0, "WET", "WET", "Africa/El_Aaiun"},
+    {3600, 1, "AZOT", "AZOST", "Atlantic/Azores"},
+    {3600, 1, "EGT", "EGST", "America/Scoresbysund"},
+    {10800, 1, "PMST", "PMDT", "America/Miquelon"},
+    {10800, 2, "UYT", "UYST", "America/Montevideo"},
+    {10800, 1, "WGT", "WGST", "America/Godthab"},
+    {10800, 2, "BRT", "BRST", "Brazil/East"},
+    {12600, 1, "NST", "NDT", "America/St_Johns"},
+    {14400, 1, "AST", "ADT", "Canada/Atlantic"},
+    {14400, 2, "AMT", "AMST", "America/Cuiaba"},
+    {14400, 2, "CLT", "CLST", "Chile/Continental"},
+    {14400, 2, "FKT", "FKST", "Atlantic/Stanley"},
+    {14400, 2, "PYT", "PYST", "America/Asuncion"},
+    {18000, 1, "CST", "CDT", "America/Havana"},
+    {18000, 1, "EST", "EDT", "US/Eastern"}, /* Conflicts with America/Grand_Turk */
+    {21600, 2, "EAST", "EASST", "Chile/EasterIsland"},
+    {21600, 0, "CST", "MDT", "Canada/Saskatchewan"},
+    {21600, 0, "CST", "CDT", "America/Guatemala"},
+    {21600, 1, "CST", "CDT", "US/Central"}, /* Conflicts with Mexico/General */
+    {25200, 1, "MST", "MDT", "US/Mountain"}, /* Conflicts with Mexico/BajaSur */
+    {28800, 0, "PST", "PST", "Pacific/Pitcairn"},
+    {28800, 1, "PST", "PDT", "US/Pacific"}, /* Conflicts with Mexico/BajaNorte */
+    {32400, 1, "AKST", "AKDT", "US/Alaska"},
+    {36000, 1, "HAST", "HADT", "US/Aleutian"}
+};
+
+/*#define DEBUG_TZNAME*/
+
+static const char* remapShortTimeZone(const char *stdID, const char *dstID, int32_t daylightType, int32_t offset)
+{
+    int32_t idx;
+#ifdef DEBUG_TZNAME
+    fprintf(stderr, "TZ=%s std=%s dst=%s daylight=%d offset=%d\n", getenv("TZ"), stdID, dstID, daylightType, offset);
+#endif
+    for (idx = 0; idx < UPRV_LENGTHOF(OFFSET_ZONE_MAPPINGS); idx++)
+    {
+        if (offset == OFFSET_ZONE_MAPPINGS[idx].offsetSeconds
+            && daylightType == OFFSET_ZONE_MAPPINGS[idx].daylightType
+            && strcmp(OFFSET_ZONE_MAPPINGS[idx].stdID, stdID) == 0
+            && strcmp(OFFSET_ZONE_MAPPINGS[idx].dstID, dstID) == 0)
+        {
+            return OFFSET_ZONE_MAPPINGS[idx].olsonID;
+        }
+    }
+    return NULL;
+}
+#endif
+
+#ifdef SEARCH_TZFILE
+#define MAX_READ_SIZE 512
+
+typedef struct DefaultTZInfo {
+    char* defaultTZBuffer;
+    int64_t defaultTZFileSize;
+    FILE* defaultTZFilePtr;
+    UBool defaultTZstatus;
+    int32_t defaultTZPosition;
+} DefaultTZInfo;
+
+/*
+ * This method compares the two files given to see if they are a match.
+ * It is currently use to compare two TZ files.
+ */
+static UBool compareBinaryFiles(const char* defaultTZFileName, const char* TZFileName, DefaultTZInfo* tzInfo) {
+    FILE* file;
+    int64_t sizeFile;
+    int64_t sizeFileLeft;
+    int32_t sizeFileRead;
+    int32_t sizeFileToRead;
+    char bufferFile[MAX_READ_SIZE];
+    UBool result = TRUE;
+
+    if (tzInfo->defaultTZFilePtr == NULL) {
+        tzInfo->defaultTZFilePtr = fopen(defaultTZFileName, "r");
+    }
+    file = fopen(TZFileName, "r");
+
+    tzInfo->defaultTZPosition = 0; /* reset position to begin search */
+
+    if (file != NULL && tzInfo->defaultTZFilePtr != NULL) {
+        /* First check that the file size are equal. */
+        if (tzInfo->defaultTZFileSize == 0) {
+            fseek(tzInfo->defaultTZFilePtr, 0, SEEK_END);
+            tzInfo->defaultTZFileSize = ftell(tzInfo->defaultTZFilePtr);
+        }
+        fseek(file, 0, SEEK_END);
+        sizeFile = ftell(file);
+        sizeFileLeft = sizeFile;
+
+        if (sizeFile != tzInfo->defaultTZFileSize) {
+            result = FALSE;
+        } else {
+            /* Store the data from the files in seperate buffers and
+             * compare each byte to determine equality.
+             */
+            if (tzInfo->defaultTZBuffer == NULL) {
+                rewind(tzInfo->defaultTZFilePtr);
+                tzInfo->defaultTZBuffer = (char*)uprv_malloc(sizeof(char) * tzInfo->defaultTZFileSize);
+                sizeFileRead = fread(tzInfo->defaultTZBuffer, 1, tzInfo->defaultTZFileSize, tzInfo->defaultTZFilePtr);
+            }
+            rewind(file);
+            while(sizeFileLeft > 0) {
+                uprv_memset(bufferFile, 0, MAX_READ_SIZE);
+                sizeFileToRead = sizeFileLeft < MAX_READ_SIZE ? sizeFileLeft : MAX_READ_SIZE;
+
+                sizeFileRead = fread(bufferFile, 1, sizeFileToRead, file);
+                if (memcmp(tzInfo->defaultTZBuffer + tzInfo->defaultTZPosition, bufferFile, sizeFileRead) != 0) {
+                    result = FALSE;
+                    break;
+                }
+                sizeFileLeft -= sizeFileRead;
+                tzInfo->defaultTZPosition += sizeFileRead;
+            }
+        }
+    } else {
+        result = FALSE;
+    }
+
+    if (file != NULL) {
+        fclose(file);
+    }
+
+    return result;
+}
+
+
+/* dirent also lists two entries: "." and ".." that we can safely ignore. */
+#define SKIP1 "."
+#define SKIP2 ".."
+static UBool U_CALLCONV putil_cleanup(void);
+static CharString *gSearchTZFileResult = NULL;
+
+/*
+ * This method recursively traverses the directory given for a matching TZ file and returns the first match.
+ * This function is not thread safe - it uses a global, gSearchTZFileResult, to hold its results.
+ */
+static char* searchForTZFile(const char* path, DefaultTZInfo* tzInfo) {
+    DIR* dirp = NULL;
+    struct dirent* dirEntry = NULL;
+    char* result = NULL;
+    UErrorCode status = U_ZERO_ERROR;
+
+    /* Save the current path */
+    CharString curpath(path, -1, status);
+    if (U_FAILURE(status)) {
+        goto cleanupAndReturn;
+    }
+
+    dirp = opendir(path);
+    if (dirp == NULL) {
+        goto cleanupAndReturn;
+    }
+
+    if (gSearchTZFileResult == NULL) {
+        gSearchTZFileResult = new CharString;
+        if (gSearchTZFileResult == NULL) {
+            goto cleanupAndReturn;
+        }
+        ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
+    }
+
+    /* Check each entry in the directory. */
+    while((dirEntry = readdir(dirp)) != NULL) {
+        const char* dirName = dirEntry->d_name;
+        if (uprv_strcmp(dirName, SKIP1) != 0 && uprv_strcmp(dirName, SKIP2) != 0
+            && uprv_strcmp(TZFILE_SKIP, dirName) != 0 && uprv_strcmp(TZFILE_SKIP2, dirName) != 0) {
+            /* Create a newpath with the new entry to test each entry in the directory. */
+            CharString newpath(curpath, status);
+            newpath.append(dirName, -1, status);
+            if (U_FAILURE(status)) {
+                break;
+            }
+
+            DIR* subDirp = NULL;
+            if ((subDirp = opendir(newpath.data())) != NULL) {
+                /* If this new path is a directory, make a recursive call with the newpath. */
+                closedir(subDirp);
+                newpath.append('/', status);
+                if (U_FAILURE(status)) {
+                    break;
+                }
+                result = searchForTZFile(newpath.data(), tzInfo);
+                /*
+                 Have to get out here. Otherwise, we'd keep looking
+                 and return the first match in the top-level directory
+                 if there's a match in the top-level. If not, this function
+                 would return NULL and set gTimeZoneBufferPtr to NULL in initDefault().
+                 It worked without this in most cases because we have a fallback of calling
+                 localtime_r to figure out the default timezone.
+                */
+                if (result != NULL)
+                    break;
+            } else {
+                if(compareBinaryFiles(TZDEFAULT, newpath.data(), tzInfo)) {
+                    int32_t amountToSkip = sizeof(TZZONEINFO) - 1;
+                    if (amountToSkip > newpath.length()) {
+                        amountToSkip = newpath.length();
+                    }
+                    const char* zoneid = newpath.data() + amountToSkip;
+                    skipZoneIDPrefix(&zoneid);
+                    gSearchTZFileResult->clear();
+                    gSearchTZFileResult->append(zoneid, -1, status);
+                    if (U_FAILURE(status)) {
+                        break;
+                    }
+                    result = gSearchTZFileResult->data();
+                    /* Get out after the first one found. */
+                    break;
+                }
+            }
+        }
+    }
+
+  cleanupAndReturn:
+    if (dirp) {
+        closedir(dirp);
+    }
+    return result;
+}
+#endif
+
+#if U_PLATFORM == U_PF_ANDROID
+typedef int(system_property_read_callback)(const prop_info* info,
+                                           void (*callback)(void* cookie,
+                                                            const char* name,
+                                                            const char* value,
+                                                            uint32_t serial),
+                                           void* cookie);
+typedef int(system_property_get)(const char*, char*);
+
+static char gAndroidTimeZone[PROP_VALUE_MAX] = { '\0' };
+
+static void u_property_read(void* cookie, const char* name, const char* value,
+                            uint32_t serial) {
+    uprv_strcpy((char* )cookie, value);
+}
+#endif
+
+U_CAPI void U_EXPORT2
+uprv_tzname_clear_cache(void)
+{
+#if U_PLATFORM == U_PF_ANDROID
+    /* Android's timezone is stored in system property. */
+    gAndroidTimeZone[0] = '\0';
+    void* libc = dlopen("libc.so", RTLD_NOLOAD);
+    if (libc) {
+        /* Android API 26+ has new API to get system property and old API
+         * (__system_property_get) is deprecated */
+        system_property_read_callback* property_read_callback =
+            (system_property_read_callback*)dlsym(
+                libc, "__system_property_read_callback");
+        if (property_read_callback) {
+            const prop_info* info =
+                __system_property_find("persist.sys.timezone");
+            if (info) {
+                property_read_callback(info, &u_property_read, gAndroidTimeZone);
+            }
+        } else {
+            system_property_get* property_get =
+                (system_property_get*)dlsym(libc, "__system_property_get");
+            if (property_get) {
+                property_get("persist.sys.timezone", gAndroidTimeZone);
+            }
+        }
+        dlclose(libc);
+    }
+#endif
+
+#if defined(CHECK_LOCALTIME_LINK) && !defined(DEBUG_SKIP_LOCALTIME_LINK)
+    gTimeZoneBufferPtr = NULL;
+#endif
+}
+
+U_CAPI const char* U_EXPORT2
+uprv_tzname(int n)
+{
+    (void)n; // Avoid unreferenced parameter warning.
+    const char *tzid = NULL;
+#if U_PLATFORM_USES_ONLY_WIN32_API
+    tzid = uprv_detectWindowsTimeZone();
+
+    if (tzid != NULL) {
+        return tzid;
+    }
+
+#ifndef U_TZNAME
+    // The return value is free'd in timezone.cpp on Windows because
+    // the other code path returns a pointer to a heap location.
+    // If we don't have a name already, then tzname wouldn't be any
+    // better, so just fall back.
+    return uprv_strdup("");
+#endif // !U_TZNAME
+
+#else
+
+/*#if U_PLATFORM_IS_DARWIN_BASED
+    int ret;
+
+    tzid = getenv("TZFILE");
+    if (tzid != NULL) {
+        return tzid;
+    }
+#endif*/
+
+/* This code can be temporarily disabled to test tzname resolution later on. */
+#ifndef DEBUG_TZNAME
+#if U_PLATFORM == U_PF_ANDROID
+    tzid = gAndroidTimeZone;
+#else
+    tzid = getenv("TZ");
+#endif
+    if (tzid != NULL && isValidOlsonID(tzid)
+#if U_PLATFORM == U_PF_SOLARIS
+    /* When TZ equals localtime on Solaris, check the /etc/localtime file. */
+        && uprv_strcmp(tzid, TZ_ENV_CHECK) != 0
+#endif
+    ) {
+        /* The colon forces tzset() to treat the remainder as zoneinfo path */
+        if (tzid[0] == ':') {
+            tzid++;
+        }
+        /* This might be a good Olson ID. */
+        skipZoneIDPrefix(&tzid);
+        return tzid;
+    }
+    /* else U_TZNAME will give a better result. */
+#endif
+
+#if defined(CHECK_LOCALTIME_LINK) && !defined(DEBUG_SKIP_LOCALTIME_LINK)
+    /* Caller must handle threading issues */
+    if (gTimeZoneBufferPtr == NULL) {
+        /*
+        This is a trick to look at the name of the link to get the Olson ID
+        because the tzfile contents is underspecified.
+        This isn't guaranteed to work because it may not be a symlink.
+        */
+        int32_t ret = (int32_t)readlink(TZDEFAULT, gTimeZoneBuffer, sizeof(gTimeZoneBuffer)-1);
+        if (0 < ret) {
+            int32_t tzZoneInfoTailLen = uprv_strlen(TZZONEINFOTAIL);
+            gTimeZoneBuffer[ret] = 0;
+            char *  tzZoneInfoTailPtr = uprv_strstr(gTimeZoneBuffer, TZZONEINFOTAIL);
+
+            if (tzZoneInfoTailPtr != NULL
+                && isValidOlsonID(tzZoneInfoTailPtr + tzZoneInfoTailLen))
+            {
+                return (gTimeZoneBufferPtr = tzZoneInfoTailPtr + tzZoneInfoTailLen);
+            }
+        } else {
+#if defined(SEARCH_TZFILE)
+            DefaultTZInfo* tzInfo = (DefaultTZInfo*)uprv_malloc(sizeof(DefaultTZInfo));
+            if (tzInfo != NULL) {
+                tzInfo->defaultTZBuffer = NULL;
+                tzInfo->defaultTZFileSize = 0;
+                tzInfo->defaultTZFilePtr = NULL;
+                tzInfo->defaultTZstatus = FALSE;
+                tzInfo->defaultTZPosition = 0;
+
+                gTimeZoneBufferPtr = searchForTZFile(TZZONEINFO, tzInfo);
+
+                /* Free previously allocated memory */
+                if (tzInfo->defaultTZBuffer != NULL) {
+                    uprv_free(tzInfo->defaultTZBuffer);
+                }
+                if (tzInfo->defaultTZFilePtr != NULL) {
+                    fclose(tzInfo->defaultTZFilePtr);
+                }
+                uprv_free(tzInfo);
+            }
+
+            if (gTimeZoneBufferPtr != NULL && isValidOlsonID(gTimeZoneBufferPtr)) {
+                return gTimeZoneBufferPtr;
+            }
+#endif
+        }
+    }
+    else {
+        return gTimeZoneBufferPtr;
+    }
+#endif
+#endif
+
+#ifdef U_TZNAME
+#if U_PLATFORM_USES_ONLY_WIN32_API
+    /* The return value is free'd in timezone.cpp on Windows because
+     * the other code path returns a pointer to a heap location. */
+    return uprv_strdup(U_TZNAME[n]);
+#else
+    /*
+    U_TZNAME is usually a non-unique abbreviation, which isn't normally usable.
+    So we remap the abbreviation to an olson ID.
+
+    Since Windows exposes a little more timezone information,
+    we normally don't use this code on Windows because
+    uprv_detectWindowsTimeZone should have already given the correct answer.
+    */
+    {
+        struct tm juneSol, decemberSol;
+        int daylightType;
+        static const time_t juneSolstice=1182478260; /*2007-06-21 18:11 UT*/
+        static const time_t decemberSolstice=1198332540; /*2007-12-22 06:09 UT*/
+
+        /* This probing will tell us when daylight savings occurs.  */
+        localtime_r(&juneSolstice, &juneSol);
+        localtime_r(&decemberSolstice, &decemberSol);
+        if(decemberSol.tm_isdst > 0) {
+          daylightType = U_DAYLIGHT_DECEMBER;
+        } else if(juneSol.tm_isdst > 0) {
+          daylightType = U_DAYLIGHT_JUNE;
+        } else {
+          daylightType = U_DAYLIGHT_NONE;
+        }
+        tzid = remapShortTimeZone(U_TZNAME[0], U_TZNAME[1], daylightType, uprv_timezone());
+        if (tzid != NULL) {
+            return tzid;
+        }
+    }
+    return U_TZNAME[n];
+#endif
+#else
+    return "";
+#endif
+}
+
+/* Get and set the ICU data directory --------------------------------------- */
+
+static icu::UInitOnce gDataDirInitOnce = U_INITONCE_INITIALIZER;
+static char *gDataDirectory = NULL;
+
+UInitOnce gTimeZoneFilesInitOnce = U_INITONCE_INITIALIZER;
+static CharString *gTimeZoneFilesDirectory = NULL;
+
+#if U_POSIX_LOCALE || U_PLATFORM_USES_ONLY_WIN32_API
+ static const char *gCorrectedPOSIXLocale = NULL; /* Sometimes heap allocated */
+ static bool gCorrectedPOSIXLocaleHeapAllocated = false;
+#endif
+
+static UBool U_CALLCONV putil_cleanup(void)
+{
+    if (gDataDirectory && *gDataDirectory) {
+        uprv_free(gDataDirectory);
+    }
+    gDataDirectory = NULL;
+    gDataDirInitOnce.reset();
+
+    delete gTimeZoneFilesDirectory;
+    gTimeZoneFilesDirectory = NULL;
+    gTimeZoneFilesInitOnce.reset();
+
+#ifdef SEARCH_TZFILE
+    delete gSearchTZFileResult;
+    gSearchTZFileResult = NULL;
+#endif
+
+#if U_POSIX_LOCALE || U_PLATFORM_USES_ONLY_WIN32_API
+    if (gCorrectedPOSIXLocale && gCorrectedPOSIXLocaleHeapAllocated) {
+        uprv_free(const_cast<char *>(gCorrectedPOSIXLocale));
+        gCorrectedPOSIXLocale = NULL;
+        gCorrectedPOSIXLocaleHeapAllocated = false;
+    }
+#endif
+    return TRUE;
+}
+
+/*
+ * Set the data directory.
+ *    Make a copy of the passed string, and set the global data dir to point to it.
+ */
+U_CAPI void U_EXPORT2
+u_setDataDirectory(const char *directory) {
+    char *newDataDir;
+    int32_t length;
+
+    if(directory==NULL || *directory==0) {
+        /* A small optimization to prevent the malloc and copy when the
+        shared library is used, and this is a way to make sure that NULL
+        is never returned.
+        */
+        newDataDir = (char *)"";
+    }
+    else {
+        length=(int32_t)uprv_strlen(directory);
+        newDataDir = (char *)uprv_malloc(length + 2);
+        /* Exit out if newDataDir could not be created. */
+        if (newDataDir == NULL) {
+            return;
+        }
+        uprv_strcpy(newDataDir, directory);
+
+#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)
+        {
+            char *p;
+            while((p = uprv_strchr(newDataDir, U_FILE_ALT_SEP_CHAR)) != NULL) {
+                *p = U_FILE_SEP_CHAR;
+            }
+        }
+#endif
+    }
+
+    if (gDataDirectory && *gDataDirectory) {
+        uprv_free(gDataDirectory);
+    }
+    gDataDirectory = newDataDir;
+    ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_pathIsAbsolute(const char *path)
+{
+  if(!path || !*path) {
+    return FALSE;
+  }
+
+  if(*path == U_FILE_SEP_CHAR) {
+    return TRUE;
+  }
+
+#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)
+  if(*path == U_FILE_ALT_SEP_CHAR) {
+    return TRUE;
+  }
+#endif
+
+#if U_PLATFORM_USES_ONLY_WIN32_API
+  if( (((path[0] >= 'A') && (path[0] <= 'Z')) ||
+       ((path[0] >= 'a') && (path[0] <= 'z'))) &&
+      path[1] == ':' ) {
+    return TRUE;
+  }
+#endif
+
+  return FALSE;
+}
+
+/* Backup setting of ICU_DATA_DIR_PREFIX_ENV_VAR
+   (needed for some Darwin ICU build environments) */
+#if U_PLATFORM_IS_DARWIN_BASED && TARGET_OS_SIMULATOR
+# if !defined(ICU_DATA_DIR_PREFIX_ENV_VAR)
+#  define ICU_DATA_DIR_PREFIX_ENV_VAR "IPHONE_SIMULATOR_ROOT"
+# endif
+#endif
+
+#if defined(ICU_DATA_DIR_WINDOWS)
+// Helper function to get the ICU Data Directory under the Windows directory location.
+static BOOL U_CALLCONV getIcuDataDirectoryUnderWindowsDirectory(char* directoryBuffer, UINT bufferLength)
+{
+    wchar_t windowsPath[MAX_PATH];
+    char windowsPathUtf8[MAX_PATH];
+
+    UINT length = GetSystemWindowsDirectoryW(windowsPath, UPRV_LENGTHOF(windowsPath));
+    if ((length > 0) && (length < (UPRV_LENGTHOF(windowsPath) - 1))) {
+        // Convert UTF-16 to a UTF-8 string.
+        UErrorCode status = U_ZERO_ERROR;
+        int32_t windowsPathUtf8Len = 0;
+        u_strToUTF8(windowsPathUtf8, static_cast<int32_t>(UPRV_LENGTHOF(windowsPathUtf8)),
+            &windowsPathUtf8Len, reinterpret_cast<const UChar*>(windowsPath), -1, &status);
+
+        if (U_SUCCESS(status) && (status != U_STRING_NOT_TERMINATED_WARNING) &&
+            (windowsPathUtf8Len < (UPRV_LENGTHOF(windowsPathUtf8) - 1))) {
+            // Ensure it always has a separator, so we can append the ICU data path.
+            if (windowsPathUtf8[windowsPathUtf8Len - 1] != U_FILE_SEP_CHAR) {
+                windowsPathUtf8[windowsPathUtf8Len++] = U_FILE_SEP_CHAR;
+                windowsPathUtf8[windowsPathUtf8Len] = '\0';
+            }
+            // Check if the concatenated string will fit.
+            if ((windowsPathUtf8Len + UPRV_LENGTHOF(ICU_DATA_DIR_WINDOWS)) < bufferLength) {
+                uprv_strcpy(directoryBuffer, windowsPathUtf8);
+                uprv_strcat(directoryBuffer, ICU_DATA_DIR_WINDOWS);
+                return TRUE;
+            }
+        }
+    }
+
+    return FALSE;
+}
+#endif
+
+static void U_CALLCONV dataDirectoryInitFn() {
+    /* If we already have the directory, then return immediately. Will happen if user called
+     * u_setDataDirectory().
+     */
+    if (gDataDirectory) {
+        return;
+    }
+
+    const char *path = NULL;
+#if defined(ICU_DATA_DIR_PREFIX_ENV_VAR)
+    char datadir_path_buffer[PATH_MAX];
+#endif
+
+    /*
+    When ICU_NO_USER_DATA_OVERRIDE is defined, users aren't allowed to
+    override ICU's data with the ICU_DATA environment variable. This prevents
+    problems where multiple custom copies of ICU's specific version of data
+    are installed on a system. Either the application must define the data
+    directory with u_setDataDirectory, define ICU_DATA_DIR when compiling
+    ICU, set the data with udata_setCommonData or trust that all of the
+    required data is contained in ICU's data library that contains
+    the entry point defined by U_ICUDATA_ENTRY_POINT.
+
+    There may also be some platforms where environment variables
+    are not allowed.
+    */
+#   if !defined(ICU_NO_USER_DATA_OVERRIDE) && !UCONFIG_NO_FILE_IO
+    /* First try to get the environment variable */
+#     if U_PLATFORM_HAS_WINUWP_API == 0  // Windows UWP does not support getenv
+        path=getenv("ICU_DATA");
+#     endif
+#   endif
+
+    /* ICU_DATA_DIR may be set as a compile option.
+     * U_ICU_DATA_DEFAULT_DIR is provided and is set by ICU at compile time
+     * and is used only when data is built in archive mode eliminating the need
+     * for ICU_DATA_DIR to be set. U_ICU_DATA_DEFAULT_DIR is set to the installation
+     * directory of the data dat file. Users should use ICU_DATA_DIR if they want to
+     * set their own path.
+     */
+#if defined(ICU_DATA_DIR) || defined(U_ICU_DATA_DEFAULT_DIR)
+    if(path==NULL || *path==0) {
+# if defined(ICU_DATA_DIR_PREFIX_ENV_VAR)
+        const char *prefix = getenv(ICU_DATA_DIR_PREFIX_ENV_VAR);
+# endif
+# ifdef ICU_DATA_DIR
+        path=ICU_DATA_DIR;
+# else
+        path=U_ICU_DATA_DEFAULT_DIR;
+# endif
+# if defined(ICU_DATA_DIR_PREFIX_ENV_VAR)
+        if (prefix != NULL) {
+            snprintf(datadir_path_buffer, PATH_MAX, "%s%s", prefix, path);
+            path=datadir_path_buffer;
+        }
+# endif
+    }
+#endif
+
+#if defined(ICU_DATA_DIR_WINDOWS)
+    char datadir_path_buffer[MAX_PATH];
+    if (getIcuDataDirectoryUnderWindowsDirectory(datadir_path_buffer, UPRV_LENGTHOF(datadir_path_buffer))) {
+        path = datadir_path_buffer;
+    }
+#endif
+
+    if(path==NULL) {
+        /* It looks really bad, set it to something. */
+        path = "";
+    }
+
+    u_setDataDirectory(path);
+    return;
+}
+
+U_CAPI const char * U_EXPORT2
+u_getDataDirectory(void) {
+    umtx_initOnce(gDataDirInitOnce, &dataDirectoryInitFn);
+    return gDataDirectory;
+}
+
+static void setTimeZoneFilesDir(const char *path, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    gTimeZoneFilesDirectory->clear();
+    gTimeZoneFilesDirectory->append(path, status);
+#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)
+    char *p = gTimeZoneFilesDirectory->data();
+    while ((p = uprv_strchr(p, U_FILE_ALT_SEP_CHAR)) != NULL) {
+        *p = U_FILE_SEP_CHAR;
+    }
+#endif
+}
+
+#define TO_STRING(x) TO_STRING_2(x)
+#define TO_STRING_2(x) #x
+
+static void U_CALLCONV TimeZoneDataDirInitFn(UErrorCode &status) {
+    U_ASSERT(gTimeZoneFilesDirectory == NULL);
+    ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
+    gTimeZoneFilesDirectory = new CharString();
+    if (gTimeZoneFilesDirectory == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    const char *dir = "";
+
+#if defined(ICU_TIMEZONE_FILES_DIR_PREFIX_ENV_VAR)
+    char timezonefilesdir_path_buffer[PATH_MAX];
+    const char *prefix = getenv(ICU_TIMEZONE_FILES_DIR_PREFIX_ENV_VAR);
+#endif
+
+#if U_PLATFORM_HAS_WINUWP_API == 1
+// The UWP version does not support the environment variable setting.
+
+# if defined(ICU_DATA_DIR_WINDOWS)
+    // When using the Windows system data, we can possibly pick up time zone data from the Windows directory.
+    char datadir_path_buffer[MAX_PATH];
+    if (getIcuDataDirectoryUnderWindowsDirectory(datadir_path_buffer, UPRV_LENGTHOF(datadir_path_buffer))) {
+        dir = datadir_path_buffer;
+    }
+# endif
+
+#else
+    dir = getenv("ICU_TIMEZONE_FILES_DIR");
+#endif // U_PLATFORM_HAS_WINUWP_API
+
+#if defined(U_TIMEZONE_FILES_DIR)
+    if (dir == NULL) {
+        // Build time configuration setting.
+        dir = TO_STRING(U_TIMEZONE_FILES_DIR);
+    }
+#endif
+
+    if (dir == NULL) {
+        dir = "";
+    }
+
+#if defined(ICU_TIMEZONE_FILES_DIR_PREFIX_ENV_VAR)
+    if (prefix != NULL) {
+        snprintf(timezonefilesdir_path_buffer, PATH_MAX, "%s%s", prefix, dir);
+        dir = timezonefilesdir_path_buffer;
+    }
+#endif
+
+    setTimeZoneFilesDir(dir, status);
+}
+
+
+U_CAPI const char * U_EXPORT2
+u_getTimeZoneFilesDirectory(UErrorCode *status) {
+    umtx_initOnce(gTimeZoneFilesInitOnce, &TimeZoneDataDirInitFn, *status);
+    return U_SUCCESS(*status) ? gTimeZoneFilesDirectory->data() : "";
+}
+
+U_CAPI void U_EXPORT2
+u_setTimeZoneFilesDirectory(const char *path, UErrorCode *status) {
+    umtx_initOnce(gTimeZoneFilesInitOnce, &TimeZoneDataDirInitFn, *status);
+    setTimeZoneFilesDir(path, *status);
+
+    // Note: this function does some extra churn, first setting based on the
+    //       environment, then immediately replacing with the value passed in.
+    //       The logic is simpler that way, and performance shouldn't be an issue.
+}
+
+
+#if U_POSIX_LOCALE
+/* A helper function used by uprv_getPOSIXIDForDefaultLocale and
+ * uprv_getPOSIXIDForDefaultCodepage. Returns the posix locale id for
+ * LC_CTYPE and LC_MESSAGES. It doesn't support other locale categories.
+ */
+static const char *uprv_getPOSIXIDForCategory(int category)
+{
+    const char* posixID = NULL;
+    if (category == LC_MESSAGES || category == LC_CTYPE) {
+        /*
+        * On Solaris two different calls to setlocale can result in
+        * different values. Only get this value once.
+        *
+        * We must check this first because an application can set this.
+        *
+        * LC_ALL can't be used because it's platform dependent. The LANG
+        * environment variable seems to affect LC_CTYPE variable by default.
+        * Here is what setlocale(LC_ALL, NULL) can return.
+        * HPUX can return 'C C C C C C C'
+        * Solaris can return /en_US/C/C/C/C/C on the second try.
+        * Linux can return LC_CTYPE=C;LC_NUMERIC=C;...
+        *
+        * The default codepage detection also needs to use LC_CTYPE.
+        *
+        * Do not call setlocale(LC_*, "")! Using an empty string instead
+        * of NULL, will modify the libc behavior.
+        */
+        posixID = setlocale(category, NULL);
+        if ((posixID == 0)
+            || (uprv_strcmp("C", posixID) == 0)
+            || (uprv_strcmp("POSIX", posixID) == 0))
+        {
+            /* Maybe we got some garbage.  Try something more reasonable */
+            posixID = getenv("LC_ALL");
+            /* Solaris speaks POSIX -  See IEEE Std 1003.1-2008
+             * This is needed to properly handle empty env. variables
+             */
+#if U_PLATFORM == U_PF_SOLARIS
+            if ((posixID == 0) || (posixID[0] == '\0')) {
+                posixID = getenv(category == LC_MESSAGES ? "LC_MESSAGES" : "LC_CTYPE");
+                if ((posixID == 0) || (posixID[0] == '\0')) {
+#else
+            if (posixID == 0) {
+                posixID = getenv(category == LC_MESSAGES ? "LC_MESSAGES" : "LC_CTYPE");
+                if (posixID == 0) {
+#endif
+                    posixID = getenv("LANG");
+                }
+            }
+        }
+    }
+    if ((posixID==0)
+        || (uprv_strcmp("C", posixID) == 0)
+        || (uprv_strcmp("POSIX", posixID) == 0))
+    {
+        /* Nothing worked.  Give it a nice POSIX default value. */
+        posixID = "en_US_POSIX";
+        // Note: this test will not catch 'C.UTF-8',
+        // that will be handled in uprv_getDefaultLocaleID().
+        // Leave this mapping here for the uprv_getPOSIXIDForDefaultCodepage()
+        // caller which expects to see "en_US_POSIX" in many branches.
+    }
+    return posixID;
+}
+
+/* Return just the POSIX id for the default locale, whatever happens to be in
+ * it. It gets the value from LC_MESSAGES and indirectly from LC_ALL and LANG.
+ */
+static const char *uprv_getPOSIXIDForDefaultLocale(void)
+{
+    static const char* posixID = NULL;
+    if (posixID == 0) {
+        posixID = uprv_getPOSIXIDForCategory(LC_MESSAGES);
+    }
+    return posixID;
+}
+
+#if !U_CHARSET_IS_UTF8
+/* Return just the POSIX id for the default codepage, whatever happens to be in
+ * it. It gets the value from LC_CTYPE and indirectly from LC_ALL and LANG.
+ */
+static const char *uprv_getPOSIXIDForDefaultCodepage(void)
+{
+    static const char* posixID = NULL;
+    if (posixID == 0) {
+        posixID = uprv_getPOSIXIDForCategory(LC_CTYPE);
+    }
+    return posixID;
+}
+#endif
+#endif
+
+/* NOTE: The caller should handle thread safety */
+U_CAPI const char* U_EXPORT2
+uprv_getDefaultLocaleID()
+{
+#if U_POSIX_LOCALE
+/*
+  Note that:  (a '!' means the ID is improper somehow)
+     LC_ALL  ---->     default_loc          codepage
+--------------------------------------------------------
+     ab.CD             ab                   CD
+     ab@CD             ab__CD               -
+     ab@CD.EF          ab__CD               EF
+
+     ab_CD.EF@GH       ab_CD_GH             EF
+
+Some 'improper' ways to do the same as above:
+  !  ab_CD@GH.EF       ab_CD_GH             EF
+  !  ab_CD.EF@GH.IJ    ab_CD_GH             EF
+  !  ab_CD@ZZ.EF@GH.IJ ab_CD_GH             EF
+
+     _CD@GH            _CD_GH               -
+     _CD.EF@GH         _CD_GH               EF
+
+The variant cannot have dots in it.
+The 'rightmost' variant (@xxx) wins.
+The leftmost codepage (.xxx) wins.
+*/
+    const char* posixID = uprv_getPOSIXIDForDefaultLocale();
+
+    /* Format: (no spaces)
+    ll [ _CC ] [ . MM ] [ @ VV]
+
+      l = lang, C = ctry, M = charmap, V = variant
+    */
+
+    if (gCorrectedPOSIXLocale != nullptr) {
+        return gCorrectedPOSIXLocale;
+    }
+
+    // Copy the ID into owned memory.
+    // Over-allocate in case we replace "C" with "en_US_POSIX" (+10), + null termination
+    char *correctedPOSIXLocale = static_cast<char *>(uprv_malloc(uprv_strlen(posixID) + 10 + 1));
+    if (correctedPOSIXLocale == nullptr) {
+        return nullptr;
+    }
+    uprv_strcpy(correctedPOSIXLocale, posixID);
+
+    char *limit;
+    if ((limit = uprv_strchr(correctedPOSIXLocale, '.')) != nullptr) {
+        *limit = 0;
+    }
+    if ((limit = uprv_strchr(correctedPOSIXLocale, '@')) != nullptr) {
+        *limit = 0;
+    }
+
+    if ((uprv_strcmp("C", correctedPOSIXLocale) == 0) // no @ variant
+        || (uprv_strcmp("POSIX", correctedPOSIXLocale) == 0)) {
+      // Raw input was C.* or POSIX.*, Give it a nice POSIX default value.
+      // (The "C"/"POSIX" case is handled in uprv_getPOSIXIDForCategory())
+      uprv_strcpy(correctedPOSIXLocale, "en_US_POSIX");
+    }
+
+    /* Note that we scan the *uncorrected* ID. */
+    const char *p;
+    if ((p = uprv_strrchr(posixID, '@')) != nullptr) {
+        p++;
+
+        /* Take care of any special cases here.. */
+        if (!uprv_strcmp(p, "nynorsk")) {
+            p = "NY";
+            /* Don't worry about no__NY. In practice, it won't appear. */
+        }
+
+        if (uprv_strchr(correctedPOSIXLocale,'_') == nullptr) {
+            uprv_strcat(correctedPOSIXLocale, "__"); /* aa@b -> aa__b (note this can make the new locale 1 char longer) */
+        }
+        else {
+            uprv_strcat(correctedPOSIXLocale, "_"); /* aa_CC@b -> aa_CC_b */
+        }
+
+        const char *q;
+        if ((q = uprv_strchr(p, '.')) != nullptr) {
+            /* How big will the resulting string be? */
+            int32_t len = (int32_t)(uprv_strlen(correctedPOSIXLocale) + (q-p));
+            uprv_strncat(correctedPOSIXLocale, p, q-p); // do not include charset
+            correctedPOSIXLocale[len] = 0;
+        }
+        else {
+            /* Anything following the @ sign */
+            uprv_strcat(correctedPOSIXLocale, p);
+        }
+
+        /* Should there be a map from 'no@nynorsk' -> no_NO_NY here?
+         * How about 'russian' -> 'ru'?
+         * Many of the other locales using ISO codes will be handled by the
+         * canonicalization functions in uloc_getDefault.
+         */
+    }
+
+    if (gCorrectedPOSIXLocale == nullptr) {
+        gCorrectedPOSIXLocale = correctedPOSIXLocale;
+        gCorrectedPOSIXLocaleHeapAllocated = true;
+        ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
+        correctedPOSIXLocale = nullptr;
+    }
+    posixID = gCorrectedPOSIXLocale;
+
+    if (correctedPOSIXLocale != nullptr) {  /* Was already set - clean up. */
+        uprv_free(correctedPOSIXLocale);
+    }
+
+    return posixID;
+
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#define POSIX_LOCALE_CAPACITY 64
+    UErrorCode status = U_ZERO_ERROR;
+    char *correctedPOSIXLocale = nullptr;
+
+    // If we have already figured this out just use the cached value
+    if (gCorrectedPOSIXLocale != nullptr) {
+        return gCorrectedPOSIXLocale;
+    }
+
+    // No cached value, need to determine the current value
+    static WCHAR windowsLocale[LOCALE_NAME_MAX_LENGTH] = {};
+    int length = GetLocaleInfoEx(LOCALE_NAME_USER_DEFAULT, LOCALE_SNAME, windowsLocale, LOCALE_NAME_MAX_LENGTH);
+
+    // Now we should have a Windows locale name that needs converted to the POSIX style.
+    if (length > 0) // If length is 0, then the GetLocaleInfoEx failed.
+    {
+        // First we need to go from UTF-16 to char (and also convert from _ to - while we're at it.)
+        char modifiedWindowsLocale[LOCALE_NAME_MAX_LENGTH] = {};
+
+        int32_t i;
+        for (i = 0; i < UPRV_LENGTHOF(modifiedWindowsLocale); i++)
+        {
+            if (windowsLocale[i] == '_')
+            {
+                modifiedWindowsLocale[i] = '-';
+            }
+            else
+            {
+                modifiedWindowsLocale[i] = static_cast<char>(windowsLocale[i]);
+            }
+
+            if (modifiedWindowsLocale[i] == '\0')
+            {
+                break;
+            }
+        }
+
+        if (i >= UPRV_LENGTHOF(modifiedWindowsLocale))
+        {
+            // Ran out of room, can't really happen, maybe we'll be lucky about a matching
+            // locale when tags are dropped
+            modifiedWindowsLocale[UPRV_LENGTHOF(modifiedWindowsLocale) - 1] = '\0';
+        }
+
+        // Now normalize the resulting name
+        correctedPOSIXLocale = static_cast<char *>(uprv_malloc(POSIX_LOCALE_CAPACITY + 1));
+        /* TODO: Should we just exit on memory allocation failure? */
+        if (correctedPOSIXLocale)
+        {
+            int32_t posixLen = uloc_canonicalize(modifiedWindowsLocale, correctedPOSIXLocale, POSIX_LOCALE_CAPACITY, &status);
+            if (U_SUCCESS(status))
+            {
+                *(correctedPOSIXLocale + posixLen) = 0;
+                gCorrectedPOSIXLocale = correctedPOSIXLocale;
+                gCorrectedPOSIXLocaleHeapAllocated = true;
+                ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
+            }
+            else
+            {
+                uprv_free(correctedPOSIXLocale);
+            }
+        }
+    }
+
+    // If unable to find a locale we can agree upon, use en-US by default
+    if (gCorrectedPOSIXLocale == nullptr) {
+        gCorrectedPOSIXLocale = "en_US";
+    }
+    return gCorrectedPOSIXLocale;
+
+#elif U_PLATFORM == U_PF_OS400
+    /* locales are process scoped and are by definition thread safe */
+    static char correctedLocale[64];
+    const  char *localeID = getenv("LC_ALL");
+           char *p;
+
+    if (localeID == NULL)
+        localeID = getenv("LANG");
+    if (localeID == NULL)
+        localeID = setlocale(LC_ALL, NULL);
+    /* Make sure we have something... */
+    if (localeID == NULL)
+        return "en_US_POSIX";
+
+    /* Extract the locale name from the path. */
+    if((p = uprv_strrchr(localeID, '/')) != NULL)
+    {
+        /* Increment p to start of locale name. */
+        p++;
+        localeID = p;
+    }
+
+    /* Copy to work location. */
+    uprv_strcpy(correctedLocale, localeID);
+
+    /* Strip off the '.locale' extension. */
+    if((p = uprv_strchr(correctedLocale, '.')) != NULL) {
+        *p = 0;
+    }
+
+    /* Upper case the locale name. */
+    T_CString_toUpperCase(correctedLocale);
+
+    /* See if we are using the POSIX locale.  Any of the
+    * following are equivalent and use the same QLGPGCMA
+    * (POSIX) locale.
+    * QLGPGCMA2 means UCS2
+    * QLGPGCMA_4 means UTF-32
+    * QLGPGCMA_8 means UTF-8
+    */
+    if ((uprv_strcmp("C", correctedLocale) == 0) ||
+        (uprv_strcmp("POSIX", correctedLocale) == 0) ||
+        (uprv_strncmp("QLGPGCMA", correctedLocale, 8) == 0))
+    {
+        uprv_strcpy(correctedLocale, "en_US_POSIX");
+    }
+    else
+    {
+        int16_t LocaleLen;
+
+        /* Lower case the lang portion. */
+        for(p = correctedLocale; *p != 0 && *p != '_'; p++)
+        {
+            *p = uprv_tolower(*p);
+        }
+
+        /* Adjust for Euro.  After '_E' add 'URO'. */
+        LocaleLen = uprv_strlen(correctedLocale);
+        if (correctedLocale[LocaleLen - 2] == '_' &&
+            correctedLocale[LocaleLen - 1] == 'E')
+        {
+            uprv_strcat(correctedLocale, "URO");
+        }
+
+        /* If using Lotus-based locale then convert to
+         * equivalent non Lotus.
+         */
+        else if (correctedLocale[LocaleLen - 2] == '_' &&
+            correctedLocale[LocaleLen - 1] == 'L')
+        {
+            correctedLocale[LocaleLen - 2] = 0;
+        }
+
+        /* There are separate simplified and traditional
+         * locales called zh_HK_S and zh_HK_T.
+         */
+        else if (uprv_strncmp(correctedLocale, "zh_HK", 5) == 0)
+        {
+            uprv_strcpy(correctedLocale, "zh_HK");
+        }
+
+        /* A special zh_CN_GBK locale...
+        */
+        else if (uprv_strcmp(correctedLocale, "zh_CN_GBK") == 0)
+        {
+            uprv_strcpy(correctedLocale, "zh_CN");
+        }
+
+    }
+
+    return correctedLocale;
+#endif
+
+}
+
+#if !U_CHARSET_IS_UTF8
+#if U_POSIX_LOCALE
+/*
+Due to various platform differences, one platform may specify a charset,
+when they really mean a different charset. Remap the names so that they are
+compatible with ICU. Only conflicting/ambiguous aliases should be resolved
+here. Before adding anything to this function, please consider adding unique
+names to the ICU alias table in the data directory.
+*/
+static const char*
+remapPlatformDependentCodepage(const char *locale, const char *name) {
+    if (locale != NULL && *locale == 0) {
+        /* Make sure that an empty locale is handled the same way. */
+        locale = NULL;
+    }
+    if (name == NULL) {
+        return NULL;
+    }
+#if U_PLATFORM == U_PF_AIX
+    if (uprv_strcmp(name, "IBM-943") == 0) {
+        /* Use the ASCII compatible ibm-943 */
+        name = "Shift-JIS";
+    }
+    else if (uprv_strcmp(name, "IBM-1252") == 0) {
+        /* Use the windows-1252 that contains the Euro */
+        name = "IBM-5348";
+    }
+#elif U_PLATFORM == U_PF_SOLARIS
+    if (locale != NULL && uprv_strcmp(name, "EUC") == 0) {
+        /* Solaris underspecifies the "EUC" name. */
+        if (uprv_strcmp(locale, "zh_CN") == 0) {
+            name = "EUC-CN";
+        }
+        else if (uprv_strcmp(locale, "zh_TW") == 0) {
+            name = "EUC-TW";
+        }
+        else if (uprv_strcmp(locale, "ko_KR") == 0) {
+            name = "EUC-KR";
+        }
+    }
+    else if (uprv_strcmp(name, "eucJP") == 0) {
+        /*
+        ibm-954 is the best match.
+        ibm-33722 is the default for eucJP (similar to Windows).
+        */
+        name = "eucjis";
+    }
+    else if (uprv_strcmp(name, "646") == 0) {
+        /*
+         * The default codepage given by Solaris is 646 but the C library routines treat it as if it was
+         * ISO-8859-1 instead of US-ASCII(646).
+         */
+        name = "ISO-8859-1";
+    }
+#elif U_PLATFORM_IS_DARWIN_BASED
+    if (locale == NULL && *name == 0) {
+        /*
+        No locale was specified, and an empty name was passed in.
+        This usually indicates that nl_langinfo didn't return valid information.
+        Mac OS X uses UTF-8 by default (especially the locale data and console).
+        */
+        name = "UTF-8";
+    }
+    else if (uprv_strcmp(name, "CP949") == 0) {
+        /* Remap CP949 to a similar codepage to avoid issues with backslash and won symbol. */
+        name = "EUC-KR";
+    }
+    else if (locale != NULL && uprv_strcmp(locale, "en_US_POSIX") != 0 && uprv_strcmp(name, "US-ASCII") == 0) {
+        /*
+         * For non C/POSIX locale, default the code page to UTF-8 instead of US-ASCII.
+         */
+        name = "UTF-8";
+    }
+#elif U_PLATFORM == U_PF_BSD
+    if (uprv_strcmp(name, "CP949") == 0) {
+        /* Remap CP949 to a similar codepage to avoid issues with backslash and won symbol. */
+        name = "EUC-KR";
+    }
+#elif U_PLATFORM == U_PF_HPUX
+    if (locale != NULL && uprv_strcmp(locale, "zh_HK") == 0 && uprv_strcmp(name, "big5") == 0) {
+        /* HP decided to extend big5 as hkbig5 even though it's not compatible :-( */
+        /* zh_TW.big5 is not the same charset as zh_HK.big5! */
+        name = "hkbig5";
+    }
+    else if (uprv_strcmp(name, "eucJP") == 0) {
+        /*
+        ibm-1350 is the best match, but unavailable.
+        ibm-954 is mostly a superset of ibm-1350.
+        ibm-33722 is the default for eucJP (similar to Windows).
+        */
+        name = "eucjis";
+    }
+#elif U_PLATFORM == U_PF_LINUX
+    if (locale != NULL && uprv_strcmp(name, "euc") == 0) {
+        /* Linux underspecifies the "EUC" name. */
+        if (uprv_strcmp(locale, "korean") == 0) {
+            name = "EUC-KR";
+        }
+        else if (uprv_strcmp(locale, "japanese") == 0) {
+            /* See comment below about eucJP */
+            name = "eucjis";
+        }
+    }
+    else if (uprv_strcmp(name, "eucjp") == 0) {
+        /*
+        ibm-1350 is the best match, but unavailable.
+        ibm-954 is mostly a superset of ibm-1350.
+        ibm-33722 is the default for eucJP (similar to Windows).
+        */
+        name = "eucjis";
+    }
+    else if (locale != NULL && uprv_strcmp(locale, "en_US_POSIX") != 0 &&
+            (uprv_strcmp(name, "ANSI_X3.4-1968") == 0 || uprv_strcmp(name, "US-ASCII") == 0)) {
+        /*
+         * For non C/POSIX locale, default the code page to UTF-8 instead of US-ASCII.
+         */
+        name = "UTF-8";
+    }
+    /*
+     * Linux returns ANSI_X3.4-1968 for C/POSIX, but the call site takes care of
+     * it by falling back to 'US-ASCII' when NULL is returned from this
+     * function. So, we don't have to worry about it here.
+     */
+#endif
+    /* return NULL when "" is passed in */
+    if (*name == 0) {
+        name = NULL;
+    }
+    return name;
+}
+
+static const char*
+getCodepageFromPOSIXID(const char *localeName, char * buffer, int32_t buffCapacity)
+{
+    char localeBuf[100];
+    const char *name = NULL;
+    char *variant = NULL;
+
+    if (localeName != NULL && (name = (uprv_strchr(localeName, '.'))) != NULL) {
+        size_t localeCapacity = uprv_min(sizeof(localeBuf), (name-localeName)+1);
+        uprv_strncpy(localeBuf, localeName, localeCapacity);
+        localeBuf[localeCapacity-1] = 0; /* ensure NULL termination */
+        name = uprv_strncpy(buffer, name+1, buffCapacity);
+        buffer[buffCapacity-1] = 0; /* ensure NULL termination */
+        if ((variant = const_cast<char *>(uprv_strchr(name, '@'))) != NULL) {
+            *variant = 0;
+        }
+        name = remapPlatformDependentCodepage(localeBuf, name);
+    }
+    return name;
+}
+#endif
+
+static const char*
+int_getDefaultCodepage()
+{
+#if U_PLATFORM == U_PF_OS400
+    uint32_t ccsid = 37; /* Default to ibm-37 */
+    static char codepage[64];
+    Qwc_JOBI0400_t jobinfo;
+    Qus_EC_t error = { sizeof(Qus_EC_t) }; /* SPI error code */
+
+    EPT_CALL(QUSRJOBI)(&jobinfo, sizeof(jobinfo), "JOBI0400",
+        "*                         ", "                ", &error);
+
+    if (error.Bytes_Available == 0) {
+        if (jobinfo.Coded_Char_Set_ID != 0xFFFF) {
+            ccsid = (uint32_t)jobinfo.Coded_Char_Set_ID;
+        }
+        else if (jobinfo.Default_Coded_Char_Set_Id != 0xFFFF) {
+            ccsid = (uint32_t)jobinfo.Default_Coded_Char_Set_Id;
+        }
+        /* else use the default */
+    }
+    sprintf(codepage,"ibm-%d", ccsid);
+    return codepage;
+
+#elif U_PLATFORM == U_PF_OS390
+    static char codepage[64];
+
+    strncpy(codepage, nl_langinfo(CODESET),63-strlen(UCNV_SWAP_LFNL_OPTION_STRING));
+    strcat(codepage,UCNV_SWAP_LFNL_OPTION_STRING);
+    codepage[63] = 0; /* NULL terminate */
+
+    return codepage;
+
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+    static char codepage[64];
+    DWORD codepageNumber = 0;
+
+#if U_PLATFORM_HAS_WINUWP_API == 1
+    // UWP doesn't have a direct API to get the default ACP as Microsoft would rather
+    // have folks use Unicode than a "system" code page, however this is the same
+    // codepage as the system default locale codepage.  (FWIW, the system locale is
+    // ONLY used for codepage, it should never be used for anything else)
+    GetLocaleInfoEx(LOCALE_NAME_SYSTEM_DEFAULT, LOCALE_IDEFAULTANSICODEPAGE | LOCALE_RETURN_NUMBER,
+        (LPWSTR)&codepageNumber, sizeof(codepageNumber) / sizeof(WCHAR));
+#else
+    // Win32 apps can call GetACP
+    codepageNumber = GetACP();
+#endif
+    // Special case for UTF-8
+    if (codepageNumber == 65001)
+    {
+        return "UTF-8";
+    }
+    // Windows codepages can look like windows-1252, so format the found number
+    // the numbers are eclectic, however all valid system code pages, besides UTF-8
+    // are between 3 and 19999
+    if (codepageNumber > 0 && codepageNumber < 20000)
+    {
+        sprintf(codepage, "windows-%ld", codepageNumber);
+        return codepage;
+    }
+    // If the codepage number call failed then return UTF-8
+    return "UTF-8";
+
+#elif U_POSIX_LOCALE
+    static char codesetName[100];
+    const char *localeName = NULL;
+    const char *name = NULL;
+
+    localeName = uprv_getPOSIXIDForDefaultCodepage();
+    uprv_memset(codesetName, 0, sizeof(codesetName));
+    /* On Solaris nl_langinfo returns C locale values unless setlocale
+     * was called earlier.
+     */
+#if (U_HAVE_NL_LANGINFO_CODESET && U_PLATFORM != U_PF_SOLARIS)
+    /* When available, check nl_langinfo first because it usually gives more
+       useful names. It depends on LC_CTYPE.
+       nl_langinfo may use the same buffer as setlocale. */
+    {
+        const char *codeset = nl_langinfo(U_NL_LANGINFO_CODESET);
+#if U_PLATFORM_IS_DARWIN_BASED || U_PLATFORM_IS_LINUX_BASED
+        /*
+         * On Linux and MacOSX, ensure that default codepage for non C/POSIX locale is UTF-8
+         * instead of ASCII.
+         */
+        if (uprv_strcmp(localeName, "en_US_POSIX") != 0) {
+            codeset = remapPlatformDependentCodepage(localeName, codeset);
+        } else
+#endif
+        {
+            codeset = remapPlatformDependentCodepage(NULL, codeset);
+        }
+
+        if (codeset != NULL) {
+            uprv_strncpy(codesetName, codeset, sizeof(codesetName));
+            codesetName[sizeof(codesetName)-1] = 0;
+            return codesetName;
+        }
+    }
+#endif
+
+    /* Use setlocale in a nice way, and then check some environment variables.
+       Maybe the application used setlocale already.
+    */
+    uprv_memset(codesetName, 0, sizeof(codesetName));
+    name = getCodepageFromPOSIXID(localeName, codesetName, sizeof(codesetName));
+    if (name) {
+        /* if we can find the codeset name from setlocale, return that. */
+        return name;
+    }
+
+    if (*codesetName == 0)
+    {
+        /* Everything failed. Return US ASCII (ISO 646). */
+        (void)uprv_strcpy(codesetName, "US-ASCII");
+    }
+    return codesetName;
+#else
+    return "US-ASCII";
+#endif
+}
+
+
+U_CAPI const char*  U_EXPORT2
+uprv_getDefaultCodepage()
+{
+    static char const  *name = NULL;
+    umtx_lock(NULL);
+    if (name == NULL) {
+        name = int_getDefaultCodepage();
+    }
+    umtx_unlock(NULL);
+    return name;
+}
+#endif  /* !U_CHARSET_IS_UTF8 */
+
+
+/* end of platform-specific implementation -------------- */
+
+/* version handling --------------------------------------------------------- */
+
+U_CAPI void U_EXPORT2
+u_versionFromString(UVersionInfo versionArray, const char *versionString) {
+    char *end;
+    uint16_t part=0;
+
+    if(versionArray==NULL) {
+        return;
+    }
+
+    if(versionString!=NULL) {
+        for(;;) {
+            versionArray[part]=(uint8_t)uprv_strtoul(versionString, &end, 10);
+            if(end==versionString || ++part==U_MAX_VERSION_LENGTH || *end!=U_VERSION_DELIMITER) {
+                break;
+            }
+            versionString=end+1;
+        }
+    }
+
+    while(part<U_MAX_VERSION_LENGTH) {
+        versionArray[part++]=0;
+    }
+}
+
+U_CAPI void U_EXPORT2
+u_versionFromUString(UVersionInfo versionArray, const UChar *versionString) {
+    if(versionArray!=NULL && versionString!=NULL) {
+        char versionChars[U_MAX_VERSION_STRING_LENGTH+1];
+        int32_t len = u_strlen(versionString);
+        if(len>U_MAX_VERSION_STRING_LENGTH) {
+            len = U_MAX_VERSION_STRING_LENGTH;
+        }
+        u_UCharsToChars(versionString, versionChars, len);
+        versionChars[len]=0;
+        u_versionFromString(versionArray, versionChars);
+    }
+}
+
+U_CAPI void U_EXPORT2
+u_versionToString(const UVersionInfo versionArray, char *versionString) {
+    uint16_t count, part;
+    uint8_t field;
+
+    if(versionString==NULL) {
+        return;
+    }
+
+    if(versionArray==NULL) {
+        versionString[0]=0;
+        return;
+    }
+
+    /* count how many fields need to be written */
+    for(count=4; count>0 && versionArray[count-1]==0; --count) {
+    }
+
+    if(count <= 1) {
+        count = 2;
+    }
+
+    /* write the first part */
+    /* write the decimal field value */
+    field=versionArray[0];
+    if(field>=100) {
+        *versionString++=(char)('0'+field/100);
+        field%=100;
+    }
+    if(field>=10) {
+        *versionString++=(char)('0'+field/10);
+        field%=10;
+    }
+    *versionString++=(char)('0'+field);
+
+    /* write the following parts */
+    for(part=1; part<count; ++part) {
+        /* write a dot first */
+        *versionString++=U_VERSION_DELIMITER;
+
+        /* write the decimal field value */
+        field=versionArray[part];
+        if(field>=100) {
+            *versionString++=(char)('0'+field/100);
+            field%=100;
+        }
+        if(field>=10) {
+            *versionString++=(char)('0'+field/10);
+            field%=10;
+        }
+        *versionString++=(char)('0'+field);
+    }
+
+    /* NUL-terminate */
+    *versionString=0;
+}
+
+U_CAPI void U_EXPORT2
+u_getVersion(UVersionInfo versionArray) {
+    (void)copyright;   // Suppress unused variable warning from clang.
+    u_versionFromString(versionArray, U_ICU_VERSION);
+}
+
+/**
+ * icucfg.h dependent code
+ */
+
+#if U_ENABLE_DYLOAD && HAVE_DLOPEN && !U_PLATFORM_USES_ONLY_WIN32_API
+
+#if HAVE_DLFCN_H
+#ifdef __MVS__
+#ifndef __SUSV3
+#define __SUSV3 1
+#endif
+#endif
+#include <dlfcn.h>
+#endif /* HAVE_DLFCN_H */
+
+U_CAPI void * U_EXPORT2
+uprv_dl_open(const char *libName, UErrorCode *status) {
+  void *ret = NULL;
+  if(U_FAILURE(*status)) return ret;
+  ret =  dlopen(libName, RTLD_NOW|RTLD_GLOBAL);
+  if(ret==NULL) {
+#ifdef U_TRACE_DYLOAD
+    printf("dlerror on dlopen(%s): %s\n", libName, dlerror());
+#endif
+    *status = U_MISSING_RESOURCE_ERROR;
+  }
+  return ret;
+}
+
+U_CAPI void U_EXPORT2
+uprv_dl_close(void *lib, UErrorCode *status) {
+  if(U_FAILURE(*status)) return;
+  dlclose(lib);
+}
+
+U_CAPI UVoidFunction* U_EXPORT2
+uprv_dlsym_func(void *lib, const char* sym, UErrorCode *status) {
+  union {
+      UVoidFunction *fp;
+      void *vp;
+  } uret;
+  uret.fp = NULL;
+  if(U_FAILURE(*status)) return uret.fp;
+  uret.vp = dlsym(lib, sym);
+  if(uret.vp == NULL) {
+#ifdef U_TRACE_DYLOAD
+    printf("dlerror on dlsym(%p,%s): %s\n", lib,sym, dlerror());
+#endif
+    *status = U_MISSING_RESOURCE_ERROR;
+  }
+  return uret.fp;
+}
+
+#elif U_ENABLE_DYLOAD && U_PLATFORM_USES_ONLY_WIN32_API && !U_PLATFORM_HAS_WINUWP_API
+
+/* Windows API implementation. */
+// Note: UWP does not expose/allow these APIs, so the UWP version gets the null implementation. */
+
+U_CAPI void * U_EXPORT2
+uprv_dl_open(const char *libName, UErrorCode *status) {
+  HMODULE lib = NULL;
+
+  if(U_FAILURE(*status)) return NULL;
+
+  lib = LoadLibraryA(libName);
+
+  if(lib==NULL) {
+    *status = U_MISSING_RESOURCE_ERROR;
+  }
+
+  return (void*)lib;
+}
+
+U_CAPI void U_EXPORT2
+uprv_dl_close(void *lib, UErrorCode *status) {
+  HMODULE handle = (HMODULE)lib;
+  if(U_FAILURE(*status)) return;
+
+  FreeLibrary(handle);
+
+  return;
+}
+
+U_CAPI UVoidFunction* U_EXPORT2
+uprv_dlsym_func(void *lib, const char* sym, UErrorCode *status) {
+  HMODULE handle = (HMODULE)lib;
+  UVoidFunction* addr = NULL;
+
+  if(U_FAILURE(*status) || lib==NULL) return NULL;
+
+  addr = (UVoidFunction*)GetProcAddress(handle, sym);
+
+  if(addr==NULL) {
+    DWORD lastError = GetLastError();
+    if(lastError == ERROR_PROC_NOT_FOUND) {
+      *status = U_MISSING_RESOURCE_ERROR;
+    } else {
+      *status = U_UNSUPPORTED_ERROR; /* other unknown error. */
+    }
+  }
+
+  return addr;
+}
+
+#else
+
+/* No dynamic loading, null (nonexistent) implementation. */
+
+U_CAPI void * U_EXPORT2
+uprv_dl_open(const char *libName, UErrorCode *status) {
+    (void)libName;
+    if(U_FAILURE(*status)) return NULL;
+    *status = U_UNSUPPORTED_ERROR;
+    return NULL;
+}
+
+U_CAPI void U_EXPORT2
+uprv_dl_close(void *lib, UErrorCode *status) {
+    (void)lib;
+    if(U_FAILURE(*status)) return;
+    *status = U_UNSUPPORTED_ERROR;
+    return;
+}
+
+U_CAPI UVoidFunction* U_EXPORT2
+uprv_dlsym_func(void *lib, const char* sym, UErrorCode *status) {
+  (void)lib;
+  (void)sym;
+  if(U_SUCCESS(*status)) {
+    *status = U_UNSUPPORTED_ERROR;
+  }
+  return (UVoidFunction*)NULL;
+}
+
+#endif
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/putilimp.h b/thirdparty/icu4c/common/putilimp.h
new file mode 100644
index 0000000000..a325c6c359
--- /dev/null
+++ b/thirdparty/icu4c/common/putilimp.h
@@ -0,0 +1,615 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : putilimp.h
+*
+*   Date        Name        Description
+*   10/17/04    grhoten     Move internal functions from putil.h to this file.
+******************************************************************************
+*/
+
+#ifndef PUTILIMP_H
+#define PUTILIMP_H
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+
+/**
+ * \def U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+ * Nearly all CPUs and compilers implement a right-shift of a signed integer
+ * as an Arithmetic Shift Right which copies the sign bit (the Most Significant Bit (MSB))
+ * into the vacated bits (sign extension).
+ * For example, (int32_t)0xfff5fff3>>4 becomes 0xffff5fff and -1>>1=-1.
+ *
+ * This can be useful for storing a signed value in the upper bits
+ * and another bit field in the lower bits.
+ * The signed value can be retrieved by simple right-shifting.
+ *
+ * This is consistent with the Java language.
+ *
+ * However, the C standard allows compilers to implement a right-shift of a signed integer
+ * as a Logical Shift Right which copies a 0 into the vacated bits.
+ * For example, (int32_t)0xfff5fff3>>4 becomes 0x0fff5fff and -1>>1=0x7fffffff.
+ *
+ * Code that depends on the natural behavior should be guarded with this macro,
+ * with an alternate path for unusual platforms.
+ * @internal
+ */
+#ifdef U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+    /* Use the predefined value. */
+#else
+    /*
+     * Nearly all CPUs & compilers implement a right-shift of a signed integer
+     * as an Arithmetic Shift Right (with sign extension).
+     */
+#   define U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC 1
+#endif
+
+/** Define this to 1 if your platform supports IEEE 754 floating point,
+   to 0 if it does not. */
+#ifndef IEEE_754
+#   define IEEE_754 1
+#endif
+
+/**
+ * uintptr_t is an optional part of the standard definitions in stdint.h.
+ * The opengroup.org documentation for stdint.h says
+ * "On XSI-conformant systems, the intptr_t and uintptr_t types are required;
+ * otherwise, they are optional."
+ * We assume that when uintptr_t is defined, UINTPTR_MAX is defined as well.
+ *
+ * Do not use ptrdiff_t since it is signed. size_t is unsigned.
+ */
+/* TODO: This check fails on some z environments. Filed a ticket #9357 for this. */
+#if !defined(__intptr_t_defined) && !defined(UINTPTR_MAX) && (U_PLATFORM != U_PF_OS390)
+typedef size_t uintptr_t;
+#endif
+
+/*===========================================================================*/
+/** @{ Information about POSIX support                                       */
+/*===========================================================================*/
+
+#ifdef U_HAVE_NL_LANGINFO_CODESET
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API || U_PLATFORM == U_PF_ANDROID || U_PLATFORM == U_PF_QNX
+#   define U_HAVE_NL_LANGINFO_CODESET 0
+#else
+#   define U_HAVE_NL_LANGINFO_CODESET 1
+#endif
+
+#ifdef U_NL_LANGINFO_CODESET
+    /* Use the predefined value. */
+#elif !U_HAVE_NL_LANGINFO_CODESET
+#   define U_NL_LANGINFO_CODESET -1
+#elif U_PLATFORM == U_PF_OS400
+   /* not defined */
+#else
+#   define U_NL_LANGINFO_CODESET CODESET
+#endif
+
+#if defined(U_TZSET) || defined(U_HAVE_TZSET)
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+    // UWP doesn't support tzset or environment variables for tz
+#if U_PLATFORM_HAS_WINUWP_API == 0
+#   define U_TZSET _tzset
+#endif
+#elif U_PLATFORM == U_PF_OS400
+   /* not defined */
+#else
+#   define U_TZSET tzset
+#endif
+
+#if defined(U_TIMEZONE) || defined(U_HAVE_TIMEZONE)
+    /* Use the predefined value. */
+#elif U_PLATFORM == U_PF_ANDROID
+#   define U_TIMEZONE timezone
+#elif defined(__UCLIBC__)
+    // uClibc does not have __timezone or _timezone.
+#elif defined(_NEWLIB_VERSION)
+#   define U_TIMEZONE _timezone
+#elif defined(__GLIBC__)
+    // glibc
+#   define U_TIMEZONE __timezone
+#elif U_PLATFORM_IS_LINUX_BASED
+    // not defined
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_TIMEZONE _timezone
+#elif U_PLATFORM == U_PF_BSD && !defined(__NetBSD__)
+   /* not defined */
+#elif U_PLATFORM == U_PF_OS400
+   /* not defined */
+#elif U_PLATFORM == U_PF_IPHONE
+   /* not defined */
+#else
+#   define U_TIMEZONE timezone
+#endif
+
+#if defined(U_TZNAME) || defined(U_HAVE_TZNAME)
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+    /* not usable on all windows platforms */
+#if U_PLATFORM_HAS_WINUWP_API == 0
+#   define U_TZNAME _tzname
+#endif
+#elif U_PLATFORM == U_PF_OS400
+   /* not defined */
+#else
+#   define U_TZNAME tzname
+#endif
+
+#ifdef U_HAVE_MMAP
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_HAVE_MMAP 0
+#else
+#   define U_HAVE_MMAP 1
+#endif
+
+#ifdef U_HAVE_POPEN
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_HAVE_POPEN 0
+#elif U_PLATFORM == U_PF_OS400
+#   define U_HAVE_POPEN 0
+#else
+#   define U_HAVE_POPEN 1
+#endif
+
+/**
+ * \def U_HAVE_DIRENT_H
+ * Defines whether dirent.h is available.
+ * @internal
+ */
+#ifdef U_HAVE_DIRENT_H
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_HAVE_DIRENT_H 0
+#else
+#   define U_HAVE_DIRENT_H 1
+#endif
+
+/** @} */
+
+/*===========================================================================*/
+/** @{ Programs used by ICU code                                             */
+/*===========================================================================*/
+
+/**
+ * \def U_MAKE_IS_NMAKE
+ * Defines whether the "make" program is Windows nmake.
+ */
+#ifdef U_MAKE_IS_NMAKE
+    /* Use the predefined value. */
+#elif U_PLATFORM == U_PF_WINDOWS
+#   define U_MAKE_IS_NMAKE 1
+#else
+#   define U_MAKE_IS_NMAKE 0
+#endif
+
+/** @} */
+
+/*==========================================================================*/
+/* Platform utilities                                                       */
+/*==========================================================================*/
+
+/**
+ * Platform utilities isolates the platform dependencies of the
+ * library.  For each platform which this code is ported to, these
+ * functions may have to be re-implemented.
+ */
+
+/**
+ * Floating point utility to determine if a double is Not a Number (NaN).
+ * @internal
+ */
+U_CAPI UBool   U_EXPORT2 uprv_isNaN(double d);
+/**
+ * Floating point utility to determine if a double has an infinite value.
+ * @internal
+ */
+U_CAPI UBool   U_EXPORT2 uprv_isInfinite(double d);
+/**
+ * Floating point utility to determine if a double has a positive infinite value.
+ * @internal
+ */
+U_CAPI UBool   U_EXPORT2 uprv_isPositiveInfinity(double d);
+/**
+ * Floating point utility to determine if a double has a negative infinite value.
+ * @internal
+ */
+U_CAPI UBool   U_EXPORT2 uprv_isNegativeInfinity(double d);
+/**
+ * Floating point utility that returns a Not a Number (NaN) value.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_getNaN(void);
+/**
+ * Floating point utility that returns an infinite value.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_getInfinity(void);
+
+/**
+ * Floating point utility to truncate a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_trunc(double d);
+/**
+ * Floating point utility to calculate the floor of a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_floor(double d);
+/**
+ * Floating point utility to calculate the ceiling of a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_ceil(double d);
+/**
+ * Floating point utility to calculate the absolute value of a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_fabs(double d);
+/**
+ * Floating point utility to calculate the fractional and integer parts of a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_modf(double d, double* pinteger);
+/**
+ * Floating point utility to calculate the remainder of a double divided by another double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_fmod(double d, double y);
+/**
+ * Floating point utility to calculate d to the power of exponent (d^exponent).
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_pow(double d, double exponent);
+/**
+ * Floating point utility to calculate 10 to the power of exponent (10^exponent).
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_pow10(int32_t exponent);
+/**
+ * Floating point utility to calculate the maximum value of two doubles.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_fmax(double d, double y);
+/**
+ * Floating point utility to calculate the minimum value of two doubles.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_fmin(double d, double y);
+/**
+ * Private utility to calculate the maximum value of two integers.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2 uprv_max(int32_t d, int32_t y);
+/**
+ * Private utility to calculate the minimum value of two integers.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2 uprv_min(int32_t d, int32_t y);
+
+#if U_IS_BIG_ENDIAN
+#   define uprv_isNegative(number) (*((signed char *)&(number))<0)
+#else
+#   define uprv_isNegative(number) (*((signed char *)&(number)+sizeof(number)-1)<0)
+#endif
+
+/**
+ * Return the largest positive number that can be represented by an integer
+ * type of arbitrary bit length.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_maxMantissa(void);
+
+/**
+ * Floating point utility to calculate the logarithm of a double.
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_log(double d);
+
+/**
+ * Does common notion of rounding e.g. uprv_floor(x + 0.5);
+ * @param x the double number
+ * @return the rounded double
+ * @internal
+ */
+U_CAPI double  U_EXPORT2 uprv_round(double x);
+
+/**
+ * Adds the signed integers a and b, storing the result in res.
+ * Checks for signed integer overflow.
+ * Similar to the GCC/Clang extension __builtin_add_overflow
+ *
+ * @param a The first operand.
+ * @param b The second operand.
+ * @param res a + b
+ * @return true if overflow occurred; false if no overflow occurred.
+ * @internal
+ */
+U_CAPI UBool U_EXPORT2 uprv_add32_overflow(int32_t a, int32_t b, int32_t* res);
+
+/**
+ * Multiplies the signed integers a and b, storing the result in res.
+ * Checks for signed integer overflow.
+ * Similar to the GCC/Clang extension __builtin_mul_overflow
+ *
+ * @param a The first multiplicand.
+ * @param b The second multiplicand.
+ * @param res a * b
+ * @return true if overflow occurred; false if no overflow occurred.
+ * @internal
+ */
+U_CAPI UBool U_EXPORT2 uprv_mul32_overflow(int32_t a, int32_t b, int32_t* res);
+
+#if 0
+/**
+ * Returns the number of digits after the decimal point in a double number x.
+ *
+ * @param x the double number
+ * @return the number of digits after the decimal point in a double number x.
+ * @internal
+ */
+/*U_CAPI int32_t  U_EXPORT2 uprv_digitsAfterDecimal(double x);*/
+#endif
+
+#if !U_CHARSET_IS_UTF8
+/**
+ * Please use ucnv_getDefaultName() instead.
+ * Return the default codepage for this platform and locale.
+ * This function can call setlocale() on Unix platforms. Please read the
+ * platform documentation on setlocale() before calling this function.
+ * @return the default codepage for this platform
+ * @internal
+ */
+U_CAPI const char*  U_EXPORT2 uprv_getDefaultCodepage(void);
+#endif
+
+/**
+ * Please use uloc_getDefault() instead.
+ * Return the default locale ID string by querying the system, or
+ *     zero if one cannot be found.
+ * This function can call setlocale() on Unix platforms. Please read the
+ * platform documentation on setlocale() before calling this function.
+ * @return the default locale ID string
+ * @internal
+ */
+U_CAPI const char*  U_EXPORT2 uprv_getDefaultLocaleID(void);
+
+/**
+ * Time zone utilities
+ *
+ * Wrappers for C runtime library functions relating to timezones.
+ * The t_tzset() function (similar to tzset) uses the current setting
+ * of the environment variable TZ to assign values to three global
+ * variables: daylight, timezone, and tzname. These variables have the
+ * following meanings, and are declared in &lt;time.h&gt;.
+ *
+ *   daylight   Nonzero if daylight-saving-time zone (DST) is specified
+ *              in TZ; otherwise, 0. Default value is 1.
+ *   timezone   Difference in seconds between coordinated universal
+ *              time and local time. E.g., -28,800 for PST (GMT-8hrs)
+ *   tzname(0)  Three-letter time-zone name derived from TZ environment
+ *              variable. E.g., "PST".
+ *   tzname(1)  Three-letter DST zone name derived from TZ environment
+ *              variable.  E.g., "PDT". If DST zone is omitted from TZ,
+ *              tzname(1) is an empty string.
+ *
+ * Notes: For example, to set the TZ environment variable to correspond
+ * to the current time zone in Germany, you can use one of the
+ * following statements:
+ *
+ *   set TZ=GST1GDT
+ *   set TZ=GST+1GDT
+ *
+ * If the TZ value is not set, t_tzset() attempts to use the time zone
+ * information specified by the operating system. Under Windows NT
+ * and Windows 95, this information is specified in the Control Panel's
+ * Date/Time application.
+ * @internal
+ */
+U_CAPI void     U_EXPORT2 uprv_tzset(void);
+
+/**
+ * Difference in seconds between coordinated universal
+ * time and local time. E.g., -28,800 for PST (GMT-8hrs)
+ * @return the difference in seconds between coordinated universal time and local time.
+ * @internal
+ */
+U_CAPI int32_t  U_EXPORT2 uprv_timezone(void);
+
+/**
+ *   tzname(0)  Three-letter time-zone name derived from TZ environment
+ *              variable. E.g., "PST".
+ *   tzname(1)  Three-letter DST zone name derived from TZ environment
+ *              variable.  E.g., "PDT". If DST zone is omitted from TZ,
+ *              tzname(1) is an empty string.
+ * @internal
+ */
+U_CAPI const char* U_EXPORT2 uprv_tzname(int n);
+
+/**
+ * Reset the global tzname cache.
+ * @internal
+ */
+U_CAPI void uprv_tzname_clear_cache(void);
+
+/**
+ * Get UTC (GMT) time measured in milliseconds since 0:00 on 1/1/1970.
+ * This function is affected by 'faketime' and should be the bottleneck for all user-visible ICU time functions.
+ * @return the UTC time measured in milliseconds
+ * @internal
+ */
+U_CAPI UDate U_EXPORT2 uprv_getUTCtime(void);
+
+/**
+ * Get UTC (GMT) time measured in milliseconds since 0:00 on 1/1/1970.
+ * This function is not affected by 'faketime', so it should only be used by low level test functions- not by anything that
+ * exposes time to the end user.
+ * @return the UTC time measured in milliseconds
+ * @internal
+ */
+U_CAPI UDate U_EXPORT2 uprv_getRawUTCtime(void);
+
+/**
+ * Determine whether a pathname is absolute or not, as defined by the platform.
+ * @param path Pathname to test
+ * @return true if the path is absolute
+ * @internal (ICU 3.0)
+ */
+U_CAPI UBool U_EXPORT2 uprv_pathIsAbsolute(const char *path);
+
+/**
+ * Use U_MAX_PTR instead of this function.
+ * @param void pointer to test
+ * @return the largest possible pointer greater than the base
+ * @internal (ICU 3.8)
+ */
+U_CAPI void * U_EXPORT2 uprv_maximumPtr(void *base);
+
+/**
+ * Maximum value of a (void*) - use to indicate the limit of an 'infinite' buffer.
+ * In fact, buffer sizes must not exceed 2GB so that the difference between
+ * the buffer limit and the buffer start can be expressed in an int32_t.
+ *
+ * The definition of U_MAX_PTR must fulfill the following conditions:
+ * - return the largest possible pointer greater than base
+ * - return a valid pointer according to the machine architecture (AS/400, 64-bit, etc.)
+ * - avoid wrapping around at high addresses
+ * - make sure that the returned pointer is not farther from base than 0x7fffffff bytes
+ *
+ * @param base The beginning of a buffer to find the maximum offset from
+ * @internal
+ */
+#ifndef U_MAX_PTR
+#  if U_PLATFORM == U_PF_OS390 && !defined(_LP64)
+    /* We have 31-bit pointers. */
+#    define U_MAX_PTR(base) ((void *)0x7fffffff)
+#  elif U_PLATFORM == U_PF_OS400
+#    define U_MAX_PTR(base) uprv_maximumPtr((void *)base)
+#  elif 0
+    /*
+     * For platforms where pointers are scalar values (which is normal, but unlike i5/OS)
+     * but that do not define uintptr_t.
+     *
+     * However, this does not work on modern compilers:
+     * The C++ standard does not define pointer overflow, and allows compilers to
+     * assume that p+u>p for any pointer p and any integer u>0.
+     * Thus, modern compilers optimize away the ">" comparison.
+     * (See ICU tickets #7187 and #8096.)
+     */
+#    define U_MAX_PTR(base) \
+    ((void *)(((char *)(base)+0x7fffffffu) > (char *)(base) \
+        ? ((char *)(base)+0x7fffffffu) \
+        : (char *)-1))
+#  else
+    /* Default version. C++ standard compliant for scalar pointers. */
+#    define U_MAX_PTR(base) \
+    ((void *)(((uintptr_t)(base)+0x7fffffffu) > (uintptr_t)(base) \
+        ? ((uintptr_t)(base)+0x7fffffffu) \
+        : (uintptr_t)-1))
+#  endif
+#endif
+
+
+#ifdef __cplusplus
+/**
+ * Pin a buffer capacity such that doing pointer arithmetic
+ * on the destination pointer and capacity cannot overflow.
+ *
+ * The pinned capacity must fulfill the following conditions (for positive capacities):
+ *   - dest + capacity is a valid pointer according to the machine arcitecture (AS/400, 64-bit, etc.)
+ *   - (dest + capacity) >= dest
+ *   - The size (in bytes) of T[capacity] does not exceed 0x7fffffff
+ *
+ * @param dest the destination buffer pointer.
+ * @param capacity the requested buffer capacity, in units of type T.
+ * @return the pinned capacity.
+ * @internal
+ */
+template <typename T>
+inline int32_t pinCapacity(T *dest, int32_t capacity) {
+    if (capacity <= 0) { return capacity; }
+
+    uintptr_t destInt = (uintptr_t)dest;
+    uintptr_t maxInt;
+
+#  if U_PLATFORM == U_PF_OS390 && !defined(_LP64)
+    // We have 31-bit pointers.
+    maxInt = 0x7fffffff;
+#  elif U_PLATFORM == U_PF_OS400
+    maxInt = (uintptr_t)uprv_maximumPtr((void *)dest);
+#  else
+    maxInt = destInt + 0x7fffffffu;
+    if (maxInt < destInt) {
+        // Less than 2GB to the end of the address space.
+        // Pin to that to prevent address overflow.
+        maxInt = (uintptr_t)-1;
+    }
+#  endif
+
+    uintptr_t maxBytes = maxInt - destInt;  // max. 2GB
+    int32_t maxCapacity = (int32_t)(maxBytes / sizeof(T));
+    return capacity <= maxCapacity ? capacity : maxCapacity;
+}
+#endif   // __cplusplus
+
+/*  Dynamic Library Functions */
+
+typedef void (UVoidFunction)(void);
+
+#if U_ENABLE_DYLOAD
+/**
+ * Load a library
+ * @internal (ICU 4.4)
+ */
+U_CAPI void * U_EXPORT2 uprv_dl_open(const char *libName, UErrorCode *status);
+
+/**
+ * Close a library
+ * @internal (ICU 4.4)
+ */
+U_CAPI void U_EXPORT2 uprv_dl_close( void *lib, UErrorCode *status);
+
+/**
+ * Extract a symbol from a library (function)
+ * @internal (ICU 4.8)
+ */
+U_CAPI UVoidFunction* U_EXPORT2 uprv_dlsym_func( void *lib, const char *symbolName, UErrorCode *status);
+
+/**
+ * Extract a symbol from a library (function)
+ * Not implemented, no clients.
+ * @internal
+ */
+/* U_CAPI void * U_EXPORT2 uprv_dlsym_data( void *lib, const char *symbolName, UErrorCode *status); */
+
+#endif
+
+/**
+ * Define malloc and related functions
+ * @internal
+ */
+#if U_PLATFORM == U_PF_OS400
+# define uprv_default_malloc(x) _C_TS_malloc(x)
+# define uprv_default_realloc(x,y) _C_TS_realloc(x,y)
+# define uprv_default_free(x) _C_TS_free(x)
+/* also _C_TS_calloc(x) */
+#else
+/* C defaults */
+# define uprv_default_malloc(x) malloc(x)
+# define uprv_default_realloc(x,y) realloc(x,y)
+# define uprv_default_free(x) free(x)
+#endif
+
+
+#endif
diff --git a/thirdparty/icu4c/common/rbbi.cpp b/thirdparty/icu4c/common/rbbi.cpp
new file mode 100644
index 0000000000..9b7e70c3cf
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbi.cpp
@@ -0,0 +1,1301 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+*   Copyright (C) 1999-2016 International Business Machines Corporation
+*   and others. All rights reserved.
+***************************************************************************
+*/
+//
+//  file:  rbbi.cpp  Contains the implementation of the rule based break iterator
+//                   runtime engine and the API implementation for
+//                   class RuleBasedBreakIterator
+//
+
+#include "utypeinfo.h"  // for 'typeid' to work
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include <cinttypes>
+
+#include "unicode/rbbi.h"
+#include "unicode/schriter.h"
+#include "unicode/uchriter.h"
+#include "unicode/uclean.h"
+#include "unicode/udata.h"
+
+#include "brkeng.h"
+#include "ucln_cmn.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "localsvc.h"
+#include "rbbidata.h"
+#include "rbbi_cache.h"
+#include "rbbirb.h"
+#include "uassert.h"
+#include "umutex.h"
+#include "uvectr32.h"
+
+#ifdef RBBI_DEBUG
+static UBool gTrace = FALSE;
+#endif
+
+U_NAMESPACE_BEGIN
+
+// The state number of the starting state
+constexpr int32_t START_STATE = 1;
+
+// The state-transition value indicating "stop"
+constexpr int32_t STOP_STATE = 0;
+
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(RuleBasedBreakIterator)
+
+
+//=======================================================================
+// constructors
+//=======================================================================
+
+/**
+ * Constructs a RuleBasedBreakIterator that uses the already-created
+ * tables object that is passed in as a parameter.
+ */
+RuleBasedBreakIterator::RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode &status)
+ : fSCharIter(UnicodeString())
+{
+    init(status);
+    fData = new RBBIDataWrapper(data, status); // status checked in constructor
+    if (U_FAILURE(status)) {return;}
+    if(fData == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    if (fData->fForwardTable->fLookAheadResultsSize > 0) {
+        fLookAheadMatches = static_cast<int32_t *>(
+            uprv_malloc(fData->fForwardTable->fLookAheadResultsSize * sizeof(int32_t)));
+        if (fLookAheadMatches == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    }
+}
+
+//
+//  Construct from precompiled binary rules (tables).  This constructor is public API,
+//  taking the rules as a (const uint8_t *) to match the type produced by getBinaryRules().
+//
+RuleBasedBreakIterator::RuleBasedBreakIterator(const uint8_t *compiledRules,
+                       uint32_t       ruleLength,
+                       UErrorCode     &status)
+ : fSCharIter(UnicodeString())
+{
+    init(status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (compiledRules == NULL || ruleLength < sizeof(RBBIDataHeader)) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    const RBBIDataHeader *data = (const RBBIDataHeader *)compiledRules;
+    if (data->fLength > ruleLength) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    fData = new RBBIDataWrapper(data, RBBIDataWrapper::kDontAdopt, status);
+    if (U_FAILURE(status)) {return;}
+    if(fData == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    if (fData->fForwardTable->fLookAheadResultsSize > 0) {
+        fLookAheadMatches = static_cast<int32_t *>(
+            uprv_malloc(fData->fForwardTable->fLookAheadResultsSize * sizeof(int32_t)));
+        if (fLookAheadMatches == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    }
+}
+
+
+//-------------------------------------------------------------------------------
+//
+//   Constructor   from a UDataMemory handle to precompiled break rules
+//                 stored in an ICU data file.
+//
+//-------------------------------------------------------------------------------
+RuleBasedBreakIterator::RuleBasedBreakIterator(UDataMemory* udm, UErrorCode &status)
+ : fSCharIter(UnicodeString())
+{
+    init(status);
+    fData = new RBBIDataWrapper(udm, status); // status checked in constructor
+    if (U_FAILURE(status)) {return;}
+    if(fData == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    if (fData->fForwardTable->fLookAheadResultsSize > 0) {
+        fLookAheadMatches = static_cast<int32_t *>(
+            uprv_malloc(fData->fForwardTable->fLookAheadResultsSize * sizeof(int32_t)));
+        if (fLookAheadMatches == nullptr) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    }
+}
+
+
+
+//-------------------------------------------------------------------------------
+//
+//   Constructor       from a set of rules supplied as a string.
+//
+//-------------------------------------------------------------------------------
+RuleBasedBreakIterator::RuleBasedBreakIterator( const UnicodeString  &rules,
+                                                UParseError          &parseError,
+                                                UErrorCode           &status)
+ : fSCharIter(UnicodeString())
+{
+    init(status);
+    if (U_FAILURE(status)) {return;}
+    RuleBasedBreakIterator *bi = (RuleBasedBreakIterator *)
+        RBBIRuleBuilder::createRuleBasedBreakIterator(rules, &parseError, status);
+    // Note:  This is a bit awkward.  The RBBI ruleBuilder has a factory method that
+    //        creates and returns a complete RBBI.  From here, in a constructor, we
+    //        can't just return the object created by the builder factory, hence
+    //        the assignment of the factory created object to "this".
+    if (U_SUCCESS(status)) {
+        *this = *bi;
+        delete bi;
+    }
+}
+
+
+//-------------------------------------------------------------------------------
+//
+// Default Constructor.      Create an empty shell that can be set up later.
+//                           Used when creating a RuleBasedBreakIterator from a set
+//                           of rules.
+//-------------------------------------------------------------------------------
+RuleBasedBreakIterator::RuleBasedBreakIterator()
+ : fSCharIter(UnicodeString())
+{
+    UErrorCode status = U_ZERO_ERROR;
+    init(status);
+}
+
+
+//-------------------------------------------------------------------------------
+//
+//   Copy constructor.  Will produce a break iterator with the same behavior,
+//                      and which iterates over the same text, as the one passed in.
+//
+//-------------------------------------------------------------------------------
+RuleBasedBreakIterator::RuleBasedBreakIterator(const RuleBasedBreakIterator& other)
+: BreakIterator(other),
+  fSCharIter(UnicodeString())
+{
+    UErrorCode status = U_ZERO_ERROR;
+    this->init(status);
+    *this = other;
+}
+
+
+/**
+ * Destructor
+ */
+RuleBasedBreakIterator::~RuleBasedBreakIterator() {
+    if (fCharIter != &fSCharIter) {
+        // fCharIter was adopted from the outside.
+        delete fCharIter;
+    }
+    fCharIter = nullptr;
+
+    utext_close(&fText);
+
+    if (fData != nullptr) {
+        fData->removeReference();
+        fData = nullptr;
+    }
+    delete fBreakCache;
+    fBreakCache = nullptr;
+
+    delete fDictionaryCache;
+    fDictionaryCache = nullptr;
+
+    delete fLanguageBreakEngines;
+    fLanguageBreakEngines = nullptr;
+
+    delete fUnhandledBreakEngine;
+    fUnhandledBreakEngine = nullptr;
+
+    uprv_free(fLookAheadMatches);
+    fLookAheadMatches = nullptr;
+}
+
+/**
+ * Assignment operator.  Sets this iterator to have the same behavior,
+ * and iterate over the same text, as the one passed in.
+ * TODO: needs better handling of memory allocation errors.
+ */
+RuleBasedBreakIterator&
+RuleBasedBreakIterator::operator=(const RuleBasedBreakIterator& that) {
+    if (this == &that) {
+        return *this;
+    }
+    BreakIterator::operator=(that);
+
+    if (fLanguageBreakEngines != NULL) {
+        delete fLanguageBreakEngines;
+        fLanguageBreakEngines = NULL;   // Just rebuild for now
+    }
+    // TODO: clone fLanguageBreakEngines from "that"
+    UErrorCode status = U_ZERO_ERROR;
+    utext_clone(&fText, &that.fText, FALSE, TRUE, &status);
+
+    if (fCharIter != &fSCharIter) {
+        delete fCharIter;
+    }
+    fCharIter = &fSCharIter;
+
+    if (that.fCharIter != NULL && that.fCharIter != &that.fSCharIter) {
+        // This is a little bit tricky - it will intially appear that
+        //  this->fCharIter is adopted, even if that->fCharIter was
+        //  not adopted.  That's ok.
+        fCharIter = that.fCharIter->clone();
+    }
+    fSCharIter = that.fSCharIter;
+    if (fCharIter == NULL) {
+        fCharIter = &fSCharIter;
+    }
+
+    if (fData != NULL) {
+        fData->removeReference();
+        fData = NULL;
+    }
+    if (that.fData != NULL) {
+        fData = that.fData->addReference();
+    }
+
+    uprv_free(fLookAheadMatches);
+    fLookAheadMatches = nullptr;
+    if (fData && fData->fForwardTable->fLookAheadResultsSize > 0) {
+        fLookAheadMatches = static_cast<int32_t *>(
+            uprv_malloc(fData->fForwardTable->fLookAheadResultsSize * sizeof(int32_t)));
+    }
+
+
+    fPosition = that.fPosition;
+    fRuleStatusIndex = that.fRuleStatusIndex;
+    fDone = that.fDone;
+
+    // TODO: both the dictionary and the main cache need to be copied.
+    //       Current position could be within a dictionary range. Trying to continue
+    //       the iteration without the caches present would go to the rules, with
+    //       the assumption that the current position is on a rule boundary.
+    fBreakCache->reset(fPosition, fRuleStatusIndex);
+    fDictionaryCache->reset();
+
+    return *this;
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//    init()      Shared initialization routine.   Used by all the constructors.
+//                Initializes all fields, leaving the object in a consistent state.
+//
+//-----------------------------------------------------------------------------
+void RuleBasedBreakIterator::init(UErrorCode &status) {
+    fCharIter             = nullptr;
+    fData                 = nullptr;
+    fPosition             = 0;
+    fRuleStatusIndex      = 0;
+    fDone                 = false;
+    fDictionaryCharCount  = 0;
+    fLanguageBreakEngines = nullptr;
+    fUnhandledBreakEngine = nullptr;
+    fBreakCache           = nullptr;
+    fDictionaryCache      = nullptr;
+    fLookAheadMatches     = nullptr;
+
+    // Note: IBM xlC is unable to assign or initialize member fText from UTEXT_INITIALIZER.
+    // fText                 = UTEXT_INITIALIZER;
+    static const UText initializedUText = UTEXT_INITIALIZER;
+    uprv_memcpy(&fText, &initializedUText, sizeof(UText));
+
+   if (U_FAILURE(status)) {
+        return;
+    }
+
+    utext_openUChars(&fText, NULL, 0, &status);
+    fDictionaryCache = new DictionaryCache(this, status);
+    fBreakCache      = new BreakCache(this, status);
+    if (U_SUCCESS(status) && (fDictionaryCache == NULL || fBreakCache == NULL)) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+
+#ifdef RBBI_DEBUG
+    static UBool debugInitDone = FALSE;
+    if (debugInitDone == FALSE) {
+        char *debugEnv = getenv("U_RBBIDEBUG");
+        if (debugEnv && uprv_strstr(debugEnv, "trace")) {
+            gTrace = TRUE;
+        }
+        debugInitDone = TRUE;
+    }
+#endif
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//    clone - Returns a newly-constructed RuleBasedBreakIterator with the same
+//            behavior, and iterating over the same text, as this one.
+//            Virtual function: does the right thing with subclasses.
+//
+//-----------------------------------------------------------------------------
+RuleBasedBreakIterator*
+RuleBasedBreakIterator::clone() const {
+    return new RuleBasedBreakIterator(*this);
+}
+
+/**
+ * Equality operator.  Returns TRUE if both BreakIterators are of the
+ * same class, have the same behavior, and iterate over the same text.
+ */
+UBool
+RuleBasedBreakIterator::operator==(const BreakIterator& that) const {
+    if (typeid(*this) != typeid(that)) {
+        return FALSE;
+    }
+    if (this == &that) {
+        return TRUE;
+    }
+
+    // The base class BreakIterator carries no state that participates in equality,
+    // and does not implement an equality function that would otherwise be
+    // checked at this point.
+
+    const RuleBasedBreakIterator& that2 = (const RuleBasedBreakIterator&) that;
+
+    if (!utext_equals(&fText, &that2.fText)) {
+        // The two break iterators are operating on different text,
+        //   or have a different iteration position.
+        //   Note that fText's position is always the same as the break iterator's position.
+        return FALSE;
+    }
+
+    if (!(fPosition == that2.fPosition &&
+            fRuleStatusIndex == that2.fRuleStatusIndex &&
+            fDone == that2.fDone)) {
+        return FALSE;
+    }
+
+    if (that2.fData == fData ||
+        (fData != NULL && that2.fData != NULL && *that2.fData == *fData)) {
+            // The two break iterators are using the same rules.
+            return TRUE;
+        }
+    return FALSE;
+}
+
+/**
+ * Compute a hash code for this BreakIterator
+ * @return A hash code
+ */
+int32_t
+RuleBasedBreakIterator::hashCode(void) const {
+    int32_t   hash = 0;
+    if (fData != NULL) {
+        hash = fData->hashCode();
+    }
+    return hash;
+}
+
+
+void RuleBasedBreakIterator::setText(UText *ut, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    fBreakCache->reset();
+    fDictionaryCache->reset();
+    utext_clone(&fText, ut, FALSE, TRUE, &status);
+
+    // Set up a dummy CharacterIterator to be returned if anyone
+    //   calls getText().  With input from UText, there is no reasonable
+    //   way to return a characterIterator over the actual input text.
+    //   Return one over an empty string instead - this is the closest
+    //   we can come to signaling a failure.
+    //   (GetText() is obsolete, this failure is sort of OK)
+    fSCharIter.setText(UnicodeString());
+
+    if (fCharIter != &fSCharIter) {
+        // existing fCharIter was adopted from the outside.  Delete it now.
+        delete fCharIter;
+    }
+    fCharIter = &fSCharIter;
+
+    this->first();
+}
+
+
+UText *RuleBasedBreakIterator::getUText(UText *fillIn, UErrorCode &status) const {
+    UText *result = utext_clone(fillIn, &fText, FALSE, TRUE, &status);
+    return result;
+}
+
+
+//=======================================================================
+// BreakIterator overrides
+//=======================================================================
+
+/**
+ * Return a CharacterIterator over the text being analyzed.
+ */
+CharacterIterator&
+RuleBasedBreakIterator::getText() const {
+    return *fCharIter;
+}
+
+/**
+ * Set the iterator to analyze a new piece of text.  This function resets
+ * the current iteration position to the beginning of the text.
+ * @param newText An iterator over the text to analyze.
+ */
+void
+RuleBasedBreakIterator::adoptText(CharacterIterator* newText) {
+    // If we are holding a CharacterIterator adopted from a
+    //   previous call to this function, delete it now.
+    if (fCharIter != &fSCharIter) {
+        delete fCharIter;
+    }
+
+    fCharIter = newText;
+    UErrorCode status = U_ZERO_ERROR;
+    fBreakCache->reset();
+    fDictionaryCache->reset();
+    if (newText==NULL || newText->startIndex() != 0) {
+        // startIndex !=0 wants to be an error, but there's no way to report it.
+        // Make the iterator text be an empty string.
+        utext_openUChars(&fText, NULL, 0, &status);
+    } else {
+        utext_openCharacterIterator(&fText, newText, &status);
+    }
+    this->first();
+}
+
+/**
+ * Set the iterator to analyze a new piece of text.  This function resets
+ * the current iteration position to the beginning of the text.
+ * @param newText An iterator over the text to analyze.
+ */
+void
+RuleBasedBreakIterator::setText(const UnicodeString& newText) {
+    UErrorCode status = U_ZERO_ERROR;
+    fBreakCache->reset();
+    fDictionaryCache->reset();
+    utext_openConstUnicodeString(&fText, &newText, &status);
+
+    // Set up a character iterator on the string.
+    //   Needed in case someone calls getText().
+    //  Can not, unfortunately, do this lazily on the (probably never)
+    //  call to getText(), because getText is const.
+    fSCharIter.setText(newText);
+
+    if (fCharIter != &fSCharIter) {
+        // old fCharIter was adopted from the outside.  Delete it.
+        delete fCharIter;
+    }
+    fCharIter = &fSCharIter;
+
+    this->first();
+}
+
+
+/**
+ *  Provide a new UText for the input text.  Must reference text with contents identical
+ *  to the original.
+ *  Intended for use with text data originating in Java (garbage collected) environments
+ *  where the data may be moved in memory at arbitrary times.
+ */
+RuleBasedBreakIterator &RuleBasedBreakIterator::refreshInputText(UText *input, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return *this;
+    }
+    if (input == NULL) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return *this;
+    }
+    int64_t pos = utext_getNativeIndex(&fText);
+    //  Shallow read-only clone of the new UText into the existing input UText
+    utext_clone(&fText, input, FALSE, TRUE, &status);
+    if (U_FAILURE(status)) {
+        return *this;
+    }
+    utext_setNativeIndex(&fText, pos);
+    if (utext_getNativeIndex(&fText) != pos) {
+        // Sanity check.  The new input utext is supposed to have the exact same
+        // contents as the old.  If we can't set to the same position, it doesn't.
+        // The contents underlying the old utext might be invalid at this point,
+        // so it's not safe to check directly.
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return *this;
+}
+
+
+/**
+ * Sets the current iteration position to the beginning of the text, position zero.
+ * @return The new iterator position, which is zero.
+ */
+int32_t RuleBasedBreakIterator::first(void) {
+    UErrorCode status = U_ZERO_ERROR;
+    if (!fBreakCache->seek(0)) {
+        fBreakCache->populateNear(0, status);
+    }
+    fBreakCache->current();
+    U_ASSERT(fPosition == 0);
+    return 0;
+}
+
+/**
+ * Sets the current iteration position to the end of the text.
+ * @return The text's past-the-end offset.
+ */
+int32_t RuleBasedBreakIterator::last(void) {
+    int32_t endPos = (int32_t)utext_nativeLength(&fText);
+    UBool endShouldBeBoundary = isBoundary(endPos);      // Has side effect of setting iterator position.
+    (void)endShouldBeBoundary;
+    U_ASSERT(endShouldBeBoundary);
+    U_ASSERT(fPosition == endPos);
+    return endPos;
+}
+
+/**
+ * Advances the iterator either forward or backward the specified number of steps.
+ * Negative values move backward, and positive values move forward.  This is
+ * equivalent to repeatedly calling next() or previous().
+ * @param n The number of steps to move.  The sign indicates the direction
+ * (negative is backwards, and positive is forwards).
+ * @return The character offset of the boundary position n boundaries away from
+ * the current one.
+ */
+int32_t RuleBasedBreakIterator::next(int32_t n) {
+    int32_t result = 0;
+    if (n > 0) {
+        for (; n > 0 && result != UBRK_DONE; --n) {
+            result = next();
+        }
+    } else if (n < 0) {
+        for (; n < 0 && result != UBRK_DONE; ++n) {
+            result = previous();
+        }
+    } else {
+        result = current();
+    }
+    return result;
+}
+
+/**
+ * Advances the iterator to the next boundary position.
+ * @return The position of the first boundary after this one.
+ */
+int32_t RuleBasedBreakIterator::next(void) {
+    fBreakCache->next();
+    return fDone ? UBRK_DONE : fPosition;
+}
+
+/**
+ * Move the iterator backwards, to the boundary preceding the current one.
+ *
+ *         Starts from the current position within fText.
+ *         Starting position need not be on a boundary.
+ *
+ * @return The position of the boundary position immediately preceding the starting position.
+ */
+int32_t RuleBasedBreakIterator::previous(void) {
+    UErrorCode status = U_ZERO_ERROR;
+    fBreakCache->previous(status);
+    return fDone ? UBRK_DONE : fPosition;
+}
+
+/**
+ * Sets the iterator to refer to the first boundary position following
+ * the specified position.
+ * @param startPos The position from which to begin searching for a break position.
+ * @return The position of the first break after the current position.
+ */
+int32_t RuleBasedBreakIterator::following(int32_t startPos) {
+    // if the supplied position is before the beginning, return the
+    // text's starting offset
+    if (startPos < 0) {
+        return first();
+    }
+
+    // Move requested offset to a code point start. It might be on a trail surrogate,
+    // or on a trail byte if the input is UTF-8. Or it may be beyond the end of the text.
+    utext_setNativeIndex(&fText, startPos);
+    startPos = (int32_t)utext_getNativeIndex(&fText);
+
+    UErrorCode status = U_ZERO_ERROR;
+    fBreakCache->following(startPos, status);
+    return fDone ? UBRK_DONE : fPosition;
+}
+
+/**
+ * Sets the iterator to refer to the last boundary position before the
+ * specified position.
+ * @param offset The position to begin searching for a break from.
+ * @return The position of the last boundary before the starting position.
+ */
+int32_t RuleBasedBreakIterator::preceding(int32_t offset) {
+    if (offset > utext_nativeLength(&fText)) {
+        return last();
+    }
+
+    // Move requested offset to a code point start. It might be on a trail surrogate,
+    // or on a trail byte if the input is UTF-8.
+
+    utext_setNativeIndex(&fText, offset);
+    int32_t adjustedOffset = static_cast<int32_t>(utext_getNativeIndex(&fText));
+
+    UErrorCode status = U_ZERO_ERROR;
+    fBreakCache->preceding(adjustedOffset, status);
+    return fDone ? UBRK_DONE : fPosition;
+}
+
+/**
+ * Returns true if the specfied position is a boundary position.  As a side
+ * effect, leaves the iterator pointing to the first boundary position at
+ * or after "offset".
+ *
+ * @param offset the offset to check.
+ * @return True if "offset" is a boundary position.
+ */
+UBool RuleBasedBreakIterator::isBoundary(int32_t offset) {
+    // out-of-range indexes are never boundary positions
+    if (offset < 0) {
+        first();       // For side effects on current position, tag values.
+        return FALSE;
+    }
+
+    // Adjust offset to be on a code point boundary and not beyond the end of the text.
+    // Note that isBoundary() is always false for offsets that are not on code point boundaries.
+    // But we still need the side effect of leaving iteration at the following boundary.
+
+    utext_setNativeIndex(&fText, offset);
+    int32_t adjustedOffset = static_cast<int32_t>(utext_getNativeIndex(&fText));
+
+    bool result = false;
+    UErrorCode status = U_ZERO_ERROR;
+    if (fBreakCache->seek(adjustedOffset) || fBreakCache->populateNear(adjustedOffset, status)) {
+        result = (fBreakCache->current() == offset);
+    }
+
+    if (result && adjustedOffset < offset && utext_char32At(&fText, offset) == U_SENTINEL) {
+        // Original offset is beyond the end of the text. Return FALSE, it's not a boundary,
+        // but the iteration position remains set to the end of the text, which is a boundary.
+        return FALSE;
+    }
+    if (!result) {
+        // Not on a boundary. isBoundary() must leave iterator on the following boundary.
+        // Cache->seek(), above, left us on the preceding boundary, so advance one.
+        next();
+    }
+    return result;
+}
+
+
+/**
+ * Returns the current iteration position.
+ * @return The current iteration position.
+ */
+int32_t RuleBasedBreakIterator::current(void) const {
+    return fPosition;
+}
+
+
+//=======================================================================
+// implementation
+//=======================================================================
+
+//
+// RBBIRunMode  -  the state machine runs an extra iteration at the beginning and end
+//                 of user text.  A variable with this enum type keeps track of where we
+//                 are.  The state machine only fetches user input while in the RUN mode.
+//
+enum RBBIRunMode {
+    RBBI_START,     // state machine processing is before first char of input
+    RBBI_RUN,       // state machine processing is in the user text
+    RBBI_END        // state machine processing is after end of user text.
+};
+
+
+// Wrapper functions to select the appropriate handleNext() or handleSafePrevious()
+// instantiation, based on whether an 8 or 16 bit table is required.
+//
+// These Trie access functions will be inlined within the handleNext()/Previous() instantions.
+static inline uint16_t TrieFunc8(const UCPTrie *trie, UChar32 c) {
+    return UCPTRIE_FAST_GET(trie, UCPTRIE_8, c);
+}
+
+static inline uint16_t TrieFunc16(const UCPTrie *trie, UChar32 c) {
+    return UCPTRIE_FAST_GET(trie, UCPTRIE_16, c);
+}
+
+int32_t RuleBasedBreakIterator::handleNext() {
+    const RBBIStateTable *statetable = fData->fForwardTable;
+    bool use8BitsTrie = ucptrie_getValueWidth(fData->fTrie) == UCPTRIE_VALUE_BITS_8;
+    if (statetable->fFlags & RBBI_8BITS_ROWS) {
+        if (use8BitsTrie) {
+            return handleNext<RBBIStateTableRow8, TrieFunc8>();
+        } else {
+            return handleNext<RBBIStateTableRow8, TrieFunc16>();
+        }
+    } else {
+        if (use8BitsTrie) {
+            return handleNext<RBBIStateTableRow16, TrieFunc8>();
+        } else {
+            return handleNext<RBBIStateTableRow16, TrieFunc16>();
+        }
+    }
+}
+
+int32_t RuleBasedBreakIterator::handleSafePrevious(int32_t fromPosition) {
+    const RBBIStateTable *statetable = fData->fReverseTable;
+    bool use8BitsTrie = ucptrie_getValueWidth(fData->fTrie) == UCPTRIE_VALUE_BITS_8;
+    if (statetable->fFlags & RBBI_8BITS_ROWS) {
+        if (use8BitsTrie) {
+            return handleSafePrevious<RBBIStateTableRow8, TrieFunc8>(fromPosition);
+        } else {
+            return handleSafePrevious<RBBIStateTableRow8, TrieFunc16>(fromPosition);
+        }
+    } else {
+        if (use8BitsTrie) {
+            return handleSafePrevious<RBBIStateTableRow16, TrieFunc8>(fromPosition);
+        } else {
+            return handleSafePrevious<RBBIStateTableRow16, TrieFunc16>(fromPosition);
+        }
+    }
+}
+
+
+//-----------------------------------------------------------------------------------
+//
+//  handleNext()
+//     Run the state machine to find a boundary
+//
+//-----------------------------------------------------------------------------------
+template <typename RowType, RuleBasedBreakIterator::PTrieFunc trieFunc>
+int32_t RuleBasedBreakIterator::handleNext() {
+    int32_t             state;
+    uint16_t            category        = 0;
+    RBBIRunMode         mode;
+
+    RowType             *row;
+    UChar32             c;
+    int32_t             result             = 0;
+    int32_t             initialPosition    = 0;
+    const RBBIStateTable *statetable       = fData->fForwardTable;
+    const char         *tableData          = statetable->fTableData;
+    uint32_t            tableRowLen        = statetable->fRowLen;
+    uint32_t            dictStart          = statetable->fDictCategoriesStart;
+    #ifdef RBBI_DEBUG
+        if (gTrace) {
+            RBBIDebugPuts("Handle Next   pos   char  state category");
+        }
+    #endif
+
+    // handleNext alway sets the break tag value.
+    // Set the default for it.
+    fRuleStatusIndex = 0;
+
+    fDictionaryCharCount = 0;
+
+    // if we're already at the end of the text, return DONE.
+    initialPosition = fPosition;
+    UTEXT_SETNATIVEINDEX(&fText, initialPosition);
+    result          = initialPosition;
+    c               = UTEXT_NEXT32(&fText);
+    if (c==U_SENTINEL) {
+        fDone = TRUE;
+        return UBRK_DONE;
+    }
+
+    //  Set the initial state for the state machine
+    state = START_STATE;
+    row = (RowType *)
+            //(statetable->fTableData + (statetable->fRowLen * state));
+            (tableData + tableRowLen * state);
+
+
+    mode     = RBBI_RUN;
+    if (statetable->fFlags & RBBI_BOF_REQUIRED) {
+        category = 2;
+        mode     = RBBI_START;
+    }
+
+
+    // loop until we reach the end of the text or transition to state 0
+    //
+    for (;;) {
+        if (c == U_SENTINEL) {
+            // Reached end of input string.
+            if (mode == RBBI_END) {
+                // We have already run the loop one last time with the
+                //   character set to the psueudo {eof} value.  Now it is time
+                //   to unconditionally bail out.
+                break;
+            }
+            // Run the loop one last time with the fake end-of-input character category.
+            mode = RBBI_END;
+            category = 1;
+        }
+
+        //
+        // Get the char category.  An incoming category of 1 or 2 means that
+        //      we are preset for doing the beginning or end of input, and
+        //      that we shouldn't get a category from an actual text input character.
+        //
+        if (mode == RBBI_RUN) {
+            // look up the current character's character category, which tells us
+            // which column in the state table to look at.
+            category = trieFunc(fData->fTrie, c);
+            fDictionaryCharCount += (category >= dictStart);
+        }
+
+       #ifdef RBBI_DEBUG
+            if (gTrace) {
+                RBBIDebugPrintf("             %4" PRId64 "   ", utext_getNativeIndex(&fText));
+                if (0x20<=c && c<0x7f) {
+                    RBBIDebugPrintf("\"%c\"  ", c);
+                } else {
+                    RBBIDebugPrintf("%5x  ", c);
+                }
+                RBBIDebugPrintf("%3d  %3d\n", state, category);
+            }
+        #endif
+
+        // State Transition - move machine to its next state
+        //
+
+        // fNextState is a variable-length array.
+        U_ASSERT(category<fData->fHeader->fCatCount);
+        state = row->fNextState[category];  /*Not accessing beyond memory*/
+        row = (RowType *)
+            // (statetable->fTableData + (statetable->fRowLen * state));
+            (tableData + tableRowLen * state);
+
+
+        uint16_t accepting = row->fAccepting;
+        if (accepting == ACCEPTING_UNCONDITIONAL) {
+            // Match found, common case.
+            if (mode != RBBI_START) {
+                result = (int32_t)UTEXT_GETNATIVEINDEX(&fText);
+            }
+            fRuleStatusIndex = row->fTagsIdx;   // Remember the break status (tag) values.
+        } else if (accepting > ACCEPTING_UNCONDITIONAL) {
+            // Lookahead match is completed.
+            U_ASSERT(accepting < fData->fForwardTable->fLookAheadResultsSize);
+            int32_t lookaheadResult = fLookAheadMatches[accepting];
+            if (lookaheadResult >= 0) {
+                fRuleStatusIndex = row->fTagsIdx;
+                fPosition = lookaheadResult;
+                return lookaheadResult;
+            }
+        }
+
+        // If we are at the position of the '/' in a look-ahead (hard break) rule;
+        // record the current position, to be returned later, if the full rule matches.
+        // TODO: Move this check before the previous check of fAccepting.
+        //       This would enable hard-break rules with no following context.
+        //       But there are line break test failures when trying this. Investigate.
+        //       Issue ICU-20837
+        uint16_t rule = row->fLookAhead;
+        U_ASSERT(rule == 0 || rule > ACCEPTING_UNCONDITIONAL);
+        U_ASSERT(rule == 0 || rule < fData->fForwardTable->fLookAheadResultsSize);
+        if (rule > ACCEPTING_UNCONDITIONAL) {
+            int32_t  pos = (int32_t)UTEXT_GETNATIVEINDEX(&fText);
+            fLookAheadMatches[rule] = pos;
+        }
+
+        if (state == STOP_STATE) {
+            // This is the normal exit from the lookup state machine.
+            // We have advanced through the string until it is certain that no
+            //   longer match is possible, no matter what characters follow.
+            break;
+        }
+
+        // Advance to the next character.
+        // If this is a beginning-of-input loop iteration, don't advance
+        //    the input position.  The next iteration will be processing the
+        //    first real input character.
+        if (mode == RBBI_RUN) {
+            c = UTEXT_NEXT32(&fText);
+        } else {
+            if (mode == RBBI_START) {
+                mode = RBBI_RUN;
+            }
+        }
+    }
+
+    // The state machine is done.  Check whether it found a match...
+
+    // If the iterator failed to advance in the match engine, force it ahead by one.
+    //   (This really indicates a defect in the break rules.  They should always match
+    //    at least one character.)
+    if (result == initialPosition) {
+        utext_setNativeIndex(&fText, initialPosition);
+        utext_next32(&fText);
+        result = (int32_t)utext_getNativeIndex(&fText);
+        fRuleStatusIndex = 0;
+    }
+
+    // Leave the iterator at our result position.
+    fPosition = result;
+    #ifdef RBBI_DEBUG
+        if (gTrace) {
+            RBBIDebugPrintf("result = %d\n\n", result);
+        }
+    #endif
+    return result;
+}
+
+
+//-----------------------------------------------------------------------------------
+//
+//  handleSafePrevious()
+//
+//      Iterate backwards using the safe reverse rules.
+//      The logic of this function is similar to handleNext(), but simpler
+//      because the safe table does not require as many options.
+//
+//-----------------------------------------------------------------------------------
+template <typename RowType, RuleBasedBreakIterator::PTrieFunc trieFunc>
+int32_t RuleBasedBreakIterator::handleSafePrevious(int32_t fromPosition) {
+
+    int32_t             state;
+    uint16_t            category        = 0;
+    RowType            *row;
+    UChar32             c;
+    int32_t             result          = 0;
+
+    const RBBIStateTable *stateTable = fData->fReverseTable;
+    UTEXT_SETNATIVEINDEX(&fText, fromPosition);
+    #ifdef RBBI_DEBUG
+        if (gTrace) {
+            RBBIDebugPuts("Handle Previous   pos   char  state category");
+        }
+    #endif
+
+    // if we're already at the start of the text, return DONE.
+    if (fData == NULL || UTEXT_GETNATIVEINDEX(&fText)==0) {
+        return BreakIterator::DONE;
+    }
+
+    //  Set the initial state for the state machine
+    c = UTEXT_PREVIOUS32(&fText);
+    state = START_STATE;
+    row = (RowType *)
+            (stateTable->fTableData + (stateTable->fRowLen * state));
+
+    // loop until we reach the start of the text or transition to state 0
+    //
+    for (; c != U_SENTINEL; c = UTEXT_PREVIOUS32(&fText)) {
+
+        // look up the current character's character category, which tells us
+        // which column in the state table to look at.
+        //
+        //  Off the dictionary flag bit. For reverse iteration it is not used.
+        category = trieFunc(fData->fTrie, c);
+
+        #ifdef RBBI_DEBUG
+            if (gTrace) {
+                RBBIDebugPrintf("             %4d   ", (int32_t)utext_getNativeIndex(&fText));
+                if (0x20<=c && c<0x7f) {
+                    RBBIDebugPrintf("\"%c\"  ", c);
+                } else {
+                    RBBIDebugPrintf("%5x  ", c);
+                }
+                RBBIDebugPrintf("%3d  %3d\n", state, category);
+            }
+        #endif
+
+        // State Transition - move machine to its next state
+        //
+        // fNextState is a variable-length array.
+        U_ASSERT(category<fData->fHeader->fCatCount);
+        state = row->fNextState[category];  /*Not accessing beyond memory*/
+        row = (RowType *)
+            (stateTable->fTableData + (stateTable->fRowLen * state));
+
+        if (state == STOP_STATE) {
+            // This is the normal exit from the lookup state machine.
+            // Transistion to state zero means we have found a safe point.
+            break;
+        }
+    }
+
+    // The state machine is done.  Check whether it found a match...
+    result = (int32_t)UTEXT_GETNATIVEINDEX(&fText);
+    #ifdef RBBI_DEBUG
+        if (gTrace) {
+            RBBIDebugPrintf("result = %d\n\n", result);
+        }
+    #endif
+    return result;
+}
+
+
+//-------------------------------------------------------------------------------
+//
+//   getRuleStatus()   Return the break rule tag associated with the current
+//                     iterator position.  If the iterator arrived at its current
+//                     position by iterating forwards, the value will have been
+//                     cached by the handleNext() function.
+//
+//-------------------------------------------------------------------------------
+
+int32_t  RuleBasedBreakIterator::getRuleStatus() const {
+
+    // fLastRuleStatusIndex indexes to the start of the appropriate status record
+    //                                                 (the number of status values.)
+    //   This function returns the last (largest) of the array of status values.
+    int32_t  idx = fRuleStatusIndex + fData->fRuleStatusTable[fRuleStatusIndex];
+    int32_t  tagVal = fData->fRuleStatusTable[idx];
+
+    return tagVal;
+}
+
+
+int32_t RuleBasedBreakIterator::getRuleStatusVec(
+             int32_t *fillInVec, int32_t capacity, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return 0;
+    }
+
+    int32_t  numVals = fData->fRuleStatusTable[fRuleStatusIndex];
+    int32_t  numValsToCopy = numVals;
+    if (numVals > capacity) {
+        status = U_BUFFER_OVERFLOW_ERROR;
+        numValsToCopy = capacity;
+    }
+    int i;
+    for (i=0; i<numValsToCopy; i++) {
+        fillInVec[i] = fData->fRuleStatusTable[fRuleStatusIndex + i + 1];
+    }
+    return numVals;
+}
+
+
+
+//-------------------------------------------------------------------------------
+//
+//   getBinaryRules        Access to the compiled form of the rules,
+//                         for use by build system tools that save the data
+//                         for standard iterator types.
+//
+//-------------------------------------------------------------------------------
+const uint8_t  *RuleBasedBreakIterator::getBinaryRules(uint32_t &length) {
+    const uint8_t  *retPtr = NULL;
+    length = 0;
+
+    if (fData != NULL) {
+        retPtr = (const uint8_t *)fData->fHeader;
+        length = fData->fHeader->fLength;
+    }
+    return retPtr;
+}
+
+
+RuleBasedBreakIterator *RuleBasedBreakIterator::createBufferClone(
+        void * /*stackBuffer*/, int32_t &bufferSize, UErrorCode &status) {
+    if (U_FAILURE(status)){
+        return NULL;
+    }
+
+    if (bufferSize == 0) {
+        bufferSize = 1;  // preflighting for deprecated functionality
+        return NULL;
+    }
+
+    BreakIterator *clonedBI = clone();
+    if (clonedBI == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        status = U_SAFECLONE_ALLOCATED_WARNING;
+    }
+    return (RuleBasedBreakIterator *)clonedBI;
+}
+
+U_NAMESPACE_END
+
+
+static icu::UStack *gLanguageBreakFactories = nullptr;
+static const icu::UnicodeString *gEmptyString = nullptr;
+static icu::UInitOnce gLanguageBreakFactoriesInitOnce = U_INITONCE_INITIALIZER;
+static icu::UInitOnce gRBBIInitOnce = U_INITONCE_INITIALIZER;
+
+/**
+ * Release all static memory held by breakiterator.
+ */
+U_CDECL_BEGIN
+UBool U_CALLCONV rbbi_cleanup(void) {
+    delete gLanguageBreakFactories;
+    gLanguageBreakFactories = nullptr;
+    delete gEmptyString;
+    gEmptyString = nullptr;
+    gLanguageBreakFactoriesInitOnce.reset();
+    gRBBIInitOnce.reset();
+    return TRUE;
+}
+U_CDECL_END
+
+U_CDECL_BEGIN
+static void U_CALLCONV _deleteFactory(void *obj) {
+    delete (icu::LanguageBreakFactory *) obj;
+}
+U_CDECL_END
+U_NAMESPACE_BEGIN
+
+static void U_CALLCONV rbbiInit() {
+    gEmptyString = new UnicodeString();
+    ucln_common_registerCleanup(UCLN_COMMON_RBBI, rbbi_cleanup);
+}
+
+static void U_CALLCONV initLanguageFactories() {
+    UErrorCode status = U_ZERO_ERROR;
+    U_ASSERT(gLanguageBreakFactories == NULL);
+    gLanguageBreakFactories = new UStack(_deleteFactory, NULL, status);
+    if (gLanguageBreakFactories != NULL && U_SUCCESS(status)) {
+        ICULanguageBreakFactory *builtIn = new ICULanguageBreakFactory(status);
+        gLanguageBreakFactories->push(builtIn, status);
+#ifdef U_LOCAL_SERVICE_HOOK
+        LanguageBreakFactory *extra = (LanguageBreakFactory *)uprv_svc_hook("languageBreakFactory", &status);
+        if (extra != NULL) {
+            gLanguageBreakFactories->push(extra, status);
+        }
+#endif
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_RBBI, rbbi_cleanup);
+}
+
+
+static const LanguageBreakEngine*
+getLanguageBreakEngineFromFactory(UChar32 c)
+{
+    umtx_initOnce(gLanguageBreakFactoriesInitOnce, &initLanguageFactories);
+    if (gLanguageBreakFactories == NULL) {
+        return NULL;
+    }
+
+    int32_t i = gLanguageBreakFactories->size();
+    const LanguageBreakEngine *lbe = NULL;
+    while (--i >= 0) {
+        LanguageBreakFactory *factory = (LanguageBreakFactory *)(gLanguageBreakFactories->elementAt(i));
+        lbe = factory->getEngineFor(c);
+        if (lbe != NULL) {
+            break;
+        }
+    }
+    return lbe;
+}
+
+
+//-------------------------------------------------------------------------------
+//
+//  getLanguageBreakEngine  Find an appropriate LanguageBreakEngine for the
+//                          the character c.
+//
+//-------------------------------------------------------------------------------
+const LanguageBreakEngine *
+RuleBasedBreakIterator::getLanguageBreakEngine(UChar32 c) {
+    const LanguageBreakEngine *lbe = NULL;
+    UErrorCode status = U_ZERO_ERROR;
+
+    if (fLanguageBreakEngines == NULL) {
+        fLanguageBreakEngines = new UStack(status);
+        if (fLanguageBreakEngines == NULL || U_FAILURE(status)) {
+            delete fLanguageBreakEngines;
+            fLanguageBreakEngines = 0;
+            return NULL;
+        }
+    }
+
+    int32_t i = fLanguageBreakEngines->size();
+    while (--i >= 0) {
+        lbe = (const LanguageBreakEngine *)(fLanguageBreakEngines->elementAt(i));
+        if (lbe->handles(c)) {
+            return lbe;
+        }
+    }
+
+    // No existing dictionary took the character. See if a factory wants to
+    // give us a new LanguageBreakEngine for this character.
+    lbe = getLanguageBreakEngineFromFactory(c);
+
+    // If we got one, use it and push it on our stack.
+    if (lbe != NULL) {
+        fLanguageBreakEngines->push((void *)lbe, status);
+        // Even if we can't remember it, we can keep looking it up, so
+        // return it even if the push fails.
+        return lbe;
+    }
+
+    // No engine is forthcoming for this character. Add it to the
+    // reject set. Create the reject break engine if needed.
+    if (fUnhandledBreakEngine == NULL) {
+        fUnhandledBreakEngine = new UnhandledEngine(status);
+        if (U_SUCCESS(status) && fUnhandledBreakEngine == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return nullptr;
+        }
+        // Put it last so that scripts for which we have an engine get tried
+        // first.
+        fLanguageBreakEngines->insertElementAt(fUnhandledBreakEngine, 0, status);
+        // If we can't insert it, or creation failed, get rid of it
+        if (U_FAILURE(status)) {
+            delete fUnhandledBreakEngine;
+            fUnhandledBreakEngine = 0;
+            return NULL;
+        }
+    }
+
+    // Tell the reject engine about the character; at its discretion, it may
+    // add more than just the one character.
+    fUnhandledBreakEngine->handleCharacter(c);
+
+    return fUnhandledBreakEngine;
+}
+
+void RuleBasedBreakIterator::dumpCache() {
+    fBreakCache->dumpCache();
+}
+
+void RuleBasedBreakIterator::dumpTables() {
+    fData->printData();
+}
+
+/**
+ * Returns the description used to create this iterator
+ */
+
+const UnicodeString&
+RuleBasedBreakIterator::getRules() const {
+    if (fData != NULL) {
+        return fData->getRuleSourceString();
+    } else {
+        umtx_initOnce(gRBBIInitOnce, &rbbiInit);
+        return *gEmptyString;
+    }
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbi_cache.cpp b/thirdparty/icu4c/common/rbbi_cache.cpp
new file mode 100644
index 0000000000..63ff3001c7
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbi_cache.cpp
@@ -0,0 +1,655 @@
+// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// file: rbbi_cache.cpp
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/ubrk.h"
+#include "unicode/rbbi.h"
+
+#include "rbbi_cache.h"
+
+#include "brkeng.h"
+#include "cmemory.h"
+#include "rbbidata.h"
+#include "rbbirb.h"
+#include "uassert.h"
+#include "uvectr32.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * DictionaryCache implementation
+ */
+
+RuleBasedBreakIterator::DictionaryCache::DictionaryCache(RuleBasedBreakIterator *bi, UErrorCode &status) :
+        fBI(bi), fBreaks(status), fPositionInCache(-1),
+        fStart(0), fLimit(0), fFirstRuleStatusIndex(0), fOtherRuleStatusIndex(0) {
+}
+
+RuleBasedBreakIterator::DictionaryCache::~DictionaryCache() {
+}
+
+void RuleBasedBreakIterator::DictionaryCache::reset() {
+    fPositionInCache = -1;
+    fStart = 0;
+    fLimit = 0;
+    fFirstRuleStatusIndex = 0;
+    fOtherRuleStatusIndex = 0;
+    fBreaks.removeAllElements();
+}
+
+UBool RuleBasedBreakIterator::DictionaryCache::following(int32_t fromPos, int32_t *result, int32_t *statusIndex) {
+    if (fromPos >= fLimit || fromPos < fStart) {
+        fPositionInCache = -1;
+        return FALSE;
+    }
+
+    // Sequential iteration, move from previous boundary to the following
+
+    int32_t r = 0;
+    if (fPositionInCache >= 0 && fPositionInCache < fBreaks.size() && fBreaks.elementAti(fPositionInCache) == fromPos) {
+        ++fPositionInCache;
+        if (fPositionInCache >= fBreaks.size()) {
+            fPositionInCache = -1;
+            return FALSE;
+        }
+        r = fBreaks.elementAti(fPositionInCache);
+        U_ASSERT(r > fromPos);
+        *result = r;
+        *statusIndex = fOtherRuleStatusIndex;
+        return TRUE;
+    }
+
+    // Random indexing. Linear search for the boundary following the given position.
+
+    for (fPositionInCache = 0; fPositionInCache < fBreaks.size(); ++fPositionInCache) {
+        r= fBreaks.elementAti(fPositionInCache);
+        if (r > fromPos) {
+            *result = r;
+            *statusIndex = fOtherRuleStatusIndex;
+            return TRUE;
+        }
+    }
+    UPRV_UNREACHABLE;
+}
+
+
+UBool RuleBasedBreakIterator::DictionaryCache::preceding(int32_t fromPos, int32_t *result, int32_t *statusIndex) {
+    if (fromPos <= fStart || fromPos > fLimit) {
+        fPositionInCache = -1;
+        return FALSE;
+    }
+
+    if (fromPos == fLimit) {
+        fPositionInCache = fBreaks.size() - 1;
+        if (fPositionInCache >= 0) {
+            U_ASSERT(fBreaks.elementAti(fPositionInCache) == fromPos);
+        }
+    }
+
+    int32_t r;
+    if (fPositionInCache > 0 && fPositionInCache < fBreaks.size() && fBreaks.elementAti(fPositionInCache) == fromPos) {
+        --fPositionInCache;
+        r = fBreaks.elementAti(fPositionInCache);
+        U_ASSERT(r < fromPos);
+        *result = r;
+        *statusIndex = ( r== fStart) ? fFirstRuleStatusIndex : fOtherRuleStatusIndex;
+        return TRUE;
+    }
+
+    if (fPositionInCache == 0) {
+        fPositionInCache = -1;
+        return FALSE;
+    }
+
+    for (fPositionInCache = fBreaks.size()-1; fPositionInCache >= 0; --fPositionInCache) {
+        r = fBreaks.elementAti(fPositionInCache);
+        if (r < fromPos) {
+            *result = r;
+            *statusIndex = ( r == fStart) ? fFirstRuleStatusIndex : fOtherRuleStatusIndex;
+            return TRUE;
+        }
+    }
+    UPRV_UNREACHABLE;
+}
+
+void RuleBasedBreakIterator::DictionaryCache::populateDictionary(int32_t startPos, int32_t endPos,
+                                       int32_t firstRuleStatus, int32_t otherRuleStatus) {
+    if ((endPos - startPos) <= 1) {
+        return;
+    }
+
+    reset();
+    fFirstRuleStatusIndex = firstRuleStatus;
+    fOtherRuleStatusIndex = otherRuleStatus;
+
+    int32_t rangeStart = startPos;
+    int32_t rangeEnd = endPos;
+
+    uint16_t    category;
+    int32_t     current;
+    UErrorCode  status = U_ZERO_ERROR;
+    int32_t     foundBreakCount = 0;
+    UText      *text = &fBI->fText;
+
+    // Loop through the text, looking for ranges of dictionary characters.
+    // For each span, find the appropriate break engine, and ask it to find
+    // any breaks within the span.
+
+    utext_setNativeIndex(text, rangeStart);
+    UChar32     c = utext_current32(text);
+    category = ucptrie_get(fBI->fData->fTrie, c);
+    uint32_t dictStart = fBI->fData->fForwardTable->fDictCategoriesStart;
+
+    while(U_SUCCESS(status)) {
+        while((current = (int32_t)UTEXT_GETNATIVEINDEX(text)) < rangeEnd
+                && (category < dictStart)) {
+            utext_next32(text);           // TODO: cleaner loop structure.
+            c = utext_current32(text);
+            category = ucptrie_get(fBI->fData->fTrie, c);
+        }
+        if (current >= rangeEnd) {
+            break;
+        }
+
+        // We now have a dictionary character. Get the appropriate language object
+        // to deal with it.
+        const LanguageBreakEngine *lbe = fBI->getLanguageBreakEngine(c);
+
+        // Ask the language object if there are any breaks. It will add them to the cache and
+        // leave the text pointer on the other side of its range, ready to search for the next one.
+        if (lbe != NULL) {
+            foundBreakCount += lbe->findBreaks(text, rangeStart, rangeEnd, fBreaks);
+        }
+
+        // Reload the loop variables for the next go-round
+        c = utext_current32(text);
+        category = ucptrie_get(fBI->fData->fTrie, c);
+    }
+
+    // If we found breaks, ensure that the first and last entries are
+    // the original starting and ending position. And initialize the
+    // cache iteration position to the first entry.
+
+    // printf("foundBreakCount = %d\n", foundBreakCount);
+    if (foundBreakCount > 0) {
+        U_ASSERT(foundBreakCount == fBreaks.size());
+        if (startPos < fBreaks.elementAti(0)) {
+            // The dictionary did not place a boundary at the start of the segment of text.
+            // Add one now. This should not commonly happen, but it would be easy for interactions
+            // of the rules for dictionary segments and the break engine implementations to
+            // inadvertently cause it. Cover it here, just in case.
+            fBreaks.insertElementAt(startPos, 0, status);
+        }
+        if (endPos > fBreaks.peeki()) {
+            fBreaks.push(endPos, status);
+        }
+        fPositionInCache = 0;
+        // Note: Dictionary matching may extend beyond the original limit.
+        fStart = fBreaks.elementAti(0);
+        fLimit = fBreaks.peeki();
+    } else {
+        // there were no language-based breaks, even though the segment contained
+        // dictionary characters. Subsequent attempts to fetch boundaries from the dictionary cache
+        // for this range will fail, and the calling code will fall back to the rule based boundaries.
+    }
+}
+
+
+/*
+ *   BreakCache implemetation
+ */
+
+RuleBasedBreakIterator::BreakCache::BreakCache(RuleBasedBreakIterator *bi, UErrorCode &status) :
+        fBI(bi), fSideBuffer(status) {
+    reset();
+}
+
+
+RuleBasedBreakIterator::BreakCache::~BreakCache() {
+}
+
+
+void RuleBasedBreakIterator::BreakCache::reset(int32_t pos, int32_t ruleStatus) {
+    fStartBufIdx = 0;
+    fEndBufIdx = 0;
+    fTextIdx = pos;
+    fBufIdx = 0;
+    fBoundaries[0] = pos;
+    fStatuses[0] = (uint16_t)ruleStatus;
+}
+
+
+int32_t  RuleBasedBreakIterator::BreakCache::current() {
+    fBI->fPosition = fTextIdx;
+    fBI->fRuleStatusIndex = fStatuses[fBufIdx];
+    fBI->fDone = FALSE;
+    return fTextIdx;
+}
+
+
+void RuleBasedBreakIterator::BreakCache::following(int32_t startPos, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (startPos == fTextIdx || seek(startPos) || populateNear(startPos, status)) {
+        // startPos is in the cache. Do a next() from that position.
+        // TODO: an awkward set of interactions with bi->fDone
+        //       seek() does not clear it; it can't because of interactions with populateNear().
+        //       next() does not clear it in the fast-path case, where everything matters. Maybe it should.
+        //       So clear it here, for the case where seek() succeeded on an iterator that had previously run off the end.
+        fBI->fDone = false;
+        next();
+    }
+    return;
+}
+
+
+void RuleBasedBreakIterator::BreakCache::preceding(int32_t startPos, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (startPos == fTextIdx || seek(startPos) || populateNear(startPos, status)) {
+        if (startPos == fTextIdx) {
+            previous(status);
+        } else {
+            // seek() leaves the BreakCache positioned at the preceding boundary
+            //        if the requested position is between two bounaries.
+            // current() pushes the BreakCache position out to the BreakIterator itself.
+            U_ASSERT(startPos > fTextIdx);
+            current();
+        }
+    }
+    return;
+}
+
+
+/*
+ * Out-of-line code for BreakCache::next().
+ * Cache does not already contain the boundary
+ */
+void RuleBasedBreakIterator::BreakCache::nextOL() {
+    fBI->fDone = !populateFollowing();
+    fBI->fPosition = fTextIdx;
+    fBI->fRuleStatusIndex = fStatuses[fBufIdx];
+    return;
+}
+
+
+void RuleBasedBreakIterator::BreakCache::previous(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    int32_t initialBufIdx = fBufIdx;
+    if (fBufIdx == fStartBufIdx) {
+        // At start of cache. Prepend to it.
+        populatePreceding(status);
+    } else {
+        // Cache already holds the next boundary
+        fBufIdx = modChunkSize(fBufIdx - 1);
+        fTextIdx = fBoundaries[fBufIdx];
+    }
+    fBI->fDone = (fBufIdx == initialBufIdx);
+    fBI->fPosition = fTextIdx;
+    fBI->fRuleStatusIndex = fStatuses[fBufIdx];
+    return;
+}
+
+
+UBool RuleBasedBreakIterator::BreakCache::seek(int32_t pos) {
+    if (pos < fBoundaries[fStartBufIdx] || pos > fBoundaries[fEndBufIdx]) {
+        return FALSE;
+    }
+    if (pos == fBoundaries[fStartBufIdx]) {
+        // Common case: seek(0), from BreakIterator::first()
+        fBufIdx = fStartBufIdx;
+        fTextIdx = fBoundaries[fBufIdx];
+        return TRUE;
+    }
+    if (pos == fBoundaries[fEndBufIdx]) {
+        fBufIdx = fEndBufIdx;
+        fTextIdx = fBoundaries[fBufIdx];
+        return TRUE;
+    }
+
+    int32_t min = fStartBufIdx;
+    int32_t max = fEndBufIdx;
+    while (min != max) {
+        int32_t probe = (min + max + (min>max ? CACHE_SIZE : 0)) / 2;
+        probe = modChunkSize(probe);
+        if (fBoundaries[probe] > pos) {
+            max = probe;
+        } else {
+            min = modChunkSize(probe + 1);
+        }
+    }
+    U_ASSERT(fBoundaries[max] > pos);
+    fBufIdx = modChunkSize(max - 1);
+    fTextIdx = fBoundaries[fBufIdx];
+    U_ASSERT(fTextIdx <= pos);
+    return TRUE;
+}
+
+
+UBool RuleBasedBreakIterator::BreakCache::populateNear(int32_t position, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return FALSE;
+    }
+    U_ASSERT(position < fBoundaries[fStartBufIdx] || position > fBoundaries[fEndBufIdx]);
+
+    // Find a boundary somewhere in the vicinity of the requested position.
+    // Depending on the safe rules and the text data, it could be either before, at, or after
+    // the requested position.
+
+
+    // If the requested position is not near already cached positions, clear the existing cache,
+    // find a near-by boundary and begin new cache contents there.
+
+    if ((position < fBoundaries[fStartBufIdx] - 15) || position > (fBoundaries[fEndBufIdx] + 15)) {
+        int32_t aBoundary = 0;
+        int32_t ruleStatusIndex = 0;
+        if (position > 20) {
+            int32_t backupPos = fBI->handleSafePrevious(position);
+
+            if (backupPos > 0) {
+                // Advance to the boundary following the backup position.
+                // There is a complication: the safe reverse rules identify pairs of code points
+                // that are safe. If advancing from the safe point moves forwards by less than
+                // two code points, we need to advance one more time to ensure that the boundary
+                // is good, including a correct rules status value.
+                //
+                fBI->fPosition = backupPos;
+                aBoundary = fBI->handleNext();
+                if (aBoundary <= backupPos + 4) {
+                    // +4 is a quick test for possibly having advanced only one codepoint.
+                    // Four being the length of the longest potential code point, a supplementary in UTF-8
+                    utext_setNativeIndex(&fBI->fText, aBoundary);
+                    if (backupPos == utext_getPreviousNativeIndex(&fBI->fText)) {
+                        // The initial handleNext() only advanced by a single code point. Go again.
+                        aBoundary = fBI->handleNext();   // Safe rules identify safe pairs.
+                    }
+                }
+                ruleStatusIndex = fBI->fRuleStatusIndex;
+            }
+        }
+        reset(aBoundary, ruleStatusIndex);        // Reset cache to hold aBoundary as a single starting point.
+    }
+
+    // Fill in boundaries between existing cache content and the new requested position.
+
+    if (fBoundaries[fEndBufIdx] < position) {
+        // The last position in the cache precedes the requested position.
+        // Add following position(s) to the cache.
+        while (fBoundaries[fEndBufIdx] < position) {
+            if (!populateFollowing()) {
+                UPRV_UNREACHABLE;
+            }
+        }
+        fBufIdx = fEndBufIdx;                      // Set iterator position to the end of the buffer.
+        fTextIdx = fBoundaries[fBufIdx];           // Required because populateFollowing may add extra boundaries.
+        while (fTextIdx > position) {              // Move backwards to a position at or preceding the requested pos.
+            previous(status);
+        }
+        return true;
+    }
+
+    if (fBoundaries[fStartBufIdx] > position) {
+        // The first position in the cache is beyond the requested position.
+        // back up more until we get a boundary <= the requested position.
+        while (fBoundaries[fStartBufIdx] > position) {
+            populatePreceding(status);
+        }
+        fBufIdx = fStartBufIdx;                    // Set iterator position to the start of the buffer.
+        fTextIdx = fBoundaries[fBufIdx];           // Required because populatePreceding may add extra boundaries.
+        while (fTextIdx < position) {              // Move forwards to a position at or following the requested pos.
+            next();
+        }
+        if (fTextIdx > position) {
+            // If position is not itself a boundary, the next() loop above will overshoot.
+            // Back up one, leaving cache position at the boundary preceding the requested position.
+            previous(status);
+        }
+        return true;
+    }
+
+    U_ASSERT(fTextIdx == position);
+    return true;
+}
+
+
+
+UBool RuleBasedBreakIterator::BreakCache::populateFollowing() {
+    int32_t fromPosition = fBoundaries[fEndBufIdx];
+    int32_t fromRuleStatusIdx = fStatuses[fEndBufIdx];
+    int32_t pos = 0;
+    int32_t ruleStatusIdx = 0;
+
+    if (fBI->fDictionaryCache->following(fromPosition, &pos, &ruleStatusIdx)) {
+        addFollowing(pos, ruleStatusIdx, UpdateCachePosition);
+        return TRUE;
+    }
+
+    fBI->fPosition = fromPosition;
+    pos = fBI->handleNext();
+    if (pos == UBRK_DONE) {
+        return FALSE;
+    }
+
+    ruleStatusIdx = fBI->fRuleStatusIndex;
+    if (fBI->fDictionaryCharCount > 0) {
+        // The text segment obtained from the rules includes dictionary characters.
+        // Subdivide it, with subdivided results going into the dictionary cache.
+        fBI->fDictionaryCache->populateDictionary(fromPosition, pos, fromRuleStatusIdx, ruleStatusIdx);
+        if (fBI->fDictionaryCache->following(fromPosition, &pos, &ruleStatusIdx)) {
+            addFollowing(pos, ruleStatusIdx, UpdateCachePosition);
+            return TRUE;
+            // TODO: may want to move a sizable chunk of dictionary cache to break cache at this point.
+            //       But be careful with interactions with populateNear().
+        }
+    }
+
+    // Rule based segment did not include dictionary characters.
+    // Or, it did contain dictionary chars, but the dictionary segmenter didn't handle them,
+    //    meaning that we didn't take the return, above.
+    // Add its end point to the cache.
+    addFollowing(pos, ruleStatusIdx, UpdateCachePosition);
+
+    // Add several non-dictionary boundaries at this point, to optimize straight forward iteration.
+    //    (subsequent calls to BreakIterator::next() will take the fast path, getting cached results.
+    //
+    for (int count=0; count<6; ++count) {
+        pos = fBI->handleNext();
+        if (pos == UBRK_DONE || fBI->fDictionaryCharCount > 0) {
+            break;
+        }
+        addFollowing(pos, fBI->fRuleStatusIndex, RetainCachePosition);
+    }
+
+    return TRUE;
+}
+
+
+UBool RuleBasedBreakIterator::BreakCache::populatePreceding(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return FALSE;
+    }
+
+    int32_t fromPosition = fBoundaries[fStartBufIdx];
+    if (fromPosition == 0) {
+        return FALSE;
+    }
+
+    int32_t position = 0;
+    int32_t positionStatusIdx = 0;
+
+    if (fBI->fDictionaryCache->preceding(fromPosition, &position, &positionStatusIdx)) {
+        addPreceding(position, positionStatusIdx, UpdateCachePosition);
+        return TRUE;
+    }
+
+    int32_t backupPosition = fromPosition;
+
+    // Find a boundary somewhere preceding the first already-cached boundary
+    do {
+        backupPosition = backupPosition - 30;
+        if (backupPosition <= 0) {
+            backupPosition = 0;
+        } else {
+            backupPosition = fBI->handleSafePrevious(backupPosition);
+        }
+        if (backupPosition == UBRK_DONE || backupPosition == 0) {
+            position = 0;
+            positionStatusIdx = 0;
+        } else {
+            // Advance to the boundary following the backup position.
+            // There is a complication: the safe reverse rules identify pairs of code points
+            // that are safe. If advancing from the safe point moves forwards by less than
+            // two code points, we need to advance one more time to ensure that the boundary
+            // is good, including a correct rules status value.
+            //
+            fBI->fPosition = backupPosition;
+            position = fBI->handleNext();
+            if (position <= backupPosition + 4) {
+                // +4 is a quick test for possibly having advanced only one codepoint.
+                // Four being the length of the longest potential code point, a supplementary in UTF-8
+                utext_setNativeIndex(&fBI->fText, position);
+                if (backupPosition == utext_getPreviousNativeIndex(&fBI->fText)) {
+                    // The initial handleNext() only advanced by a single code point. Go again.
+                    position = fBI->handleNext();   // Safe rules identify safe pairs.
+                }
+            }
+            positionStatusIdx = fBI->fRuleStatusIndex;
+        }
+    } while (position >= fromPosition);
+
+    // Find boundaries between the one we just located and the first already-cached boundary
+    // Put them in a side buffer, because we don't yet know where they will fall in the circular cache buffer..
+
+    fSideBuffer.removeAllElements();
+    fSideBuffer.addElement(position, status);
+    fSideBuffer.addElement(positionStatusIdx, status);
+
+    do {
+        int32_t prevPosition = fBI->fPosition = position;
+        int32_t prevStatusIdx = positionStatusIdx;
+        position = fBI->handleNext();
+        positionStatusIdx = fBI->fRuleStatusIndex;
+        if (position == UBRK_DONE) {
+            break;
+        }
+
+        UBool segmentHandledByDictionary = FALSE;
+        if (fBI->fDictionaryCharCount != 0) {
+            // Segment from the rules includes dictionary characters.
+            // Subdivide it, with subdivided results going into the dictionary cache.
+            int32_t dictSegEndPosition = position;
+            fBI->fDictionaryCache->populateDictionary(prevPosition, dictSegEndPosition, prevStatusIdx, positionStatusIdx);
+            while (fBI->fDictionaryCache->following(prevPosition, &position, &positionStatusIdx)) {
+                segmentHandledByDictionary = true;
+                U_ASSERT(position > prevPosition);
+                if (position >= fromPosition) {
+                    break;
+                }
+                U_ASSERT(position <= dictSegEndPosition);
+                fSideBuffer.addElement(position, status);
+                fSideBuffer.addElement(positionStatusIdx, status);
+                prevPosition = position;
+            }
+            U_ASSERT(position==dictSegEndPosition || position>=fromPosition);
+        }
+
+        if (!segmentHandledByDictionary && position < fromPosition) {
+            fSideBuffer.addElement(position, status);
+            fSideBuffer.addElement(positionStatusIdx, status);
+        }
+    } while (position < fromPosition);
+
+    // Move boundaries from the side buffer to the main circular buffer.
+    UBool success = FALSE;
+    if (!fSideBuffer.isEmpty()) {
+        positionStatusIdx = fSideBuffer.popi();
+        position = fSideBuffer.popi();
+        addPreceding(position, positionStatusIdx, UpdateCachePosition);
+        success = TRUE;
+    }
+
+    while (!fSideBuffer.isEmpty()) {
+        positionStatusIdx = fSideBuffer.popi();
+        position = fSideBuffer.popi();
+        if (!addPreceding(position, positionStatusIdx, RetainCachePosition)) {
+            // No space in circular buffer to hold a new preceding result while
+            // also retaining the current cache (iteration) position.
+            // Bailing out is safe; the cache will refill again if needed.
+            break;
+        }
+    }
+
+    return success;
+}
+
+
+void RuleBasedBreakIterator::BreakCache::addFollowing(int32_t position, int32_t ruleStatusIdx, UpdatePositionValues update) {
+    U_ASSERT(position > fBoundaries[fEndBufIdx]);
+    U_ASSERT(ruleStatusIdx <= UINT16_MAX);
+    int32_t nextIdx = modChunkSize(fEndBufIdx + 1);
+    if (nextIdx == fStartBufIdx) {
+        fStartBufIdx = modChunkSize(fStartBufIdx + 6);    // TODO: experiment. Probably revert to 1.
+    }
+    fBoundaries[nextIdx] = position;
+    fStatuses[nextIdx] = static_cast<uint16_t>(ruleStatusIdx);
+    fEndBufIdx = nextIdx;
+    if (update == UpdateCachePosition) {
+        // Set current position to the newly added boundary.
+        fBufIdx = nextIdx;
+        fTextIdx = position;
+    } else {
+        // Retaining the original cache position.
+        // Check if the added boundary wraps around the buffer, and would over-write the original position.
+        // It's the responsibility of callers of this function to not add too many.
+        U_ASSERT(nextIdx != fBufIdx);
+    }
+}
+
+bool RuleBasedBreakIterator::BreakCache::addPreceding(int32_t position, int32_t ruleStatusIdx, UpdatePositionValues update) {
+    U_ASSERT(position < fBoundaries[fStartBufIdx]);
+    U_ASSERT(ruleStatusIdx <= UINT16_MAX);
+    int32_t nextIdx = modChunkSize(fStartBufIdx - 1);
+    if (nextIdx == fEndBufIdx) {
+        if (fBufIdx == fEndBufIdx && update == RetainCachePosition) {
+            // Failure. The insertion of the new boundary would claim the buffer position that is the
+            // current iteration position. And we also want to retain the current iteration position.
+            // (The buffer is already completely full of entries that precede the iteration position.)
+            return false;
+        }
+        fEndBufIdx = modChunkSize(fEndBufIdx - 1);
+    }
+    fBoundaries[nextIdx] = position;
+    fStatuses[nextIdx] = static_cast<uint16_t>(ruleStatusIdx);
+    fStartBufIdx = nextIdx;
+    if (update == UpdateCachePosition) {
+        fBufIdx = nextIdx;
+        fTextIdx = position;
+    }
+    return true;
+}
+
+
+void RuleBasedBreakIterator::BreakCache::dumpCache() {
+#ifdef RBBI_DEBUG
+    RBBIDebugPrintf("fTextIdx:%d   fBufIdx:%d\n", fTextIdx, fBufIdx);
+    for (int32_t i=fStartBufIdx; ; i=modChunkSize(i+1)) {
+        RBBIDebugPrintf("%d  %d\n", i, fBoundaries[i]);
+        if (i == fEndBufIdx) {
+            break;
+        }
+    }
+#endif
+}
+
+U_NAMESPACE_END
+
+#endif // #if !UCONFIG_NO_BREAK_ITERATION
diff --git a/thirdparty/icu4c/common/rbbi_cache.h b/thirdparty/icu4c/common/rbbi_cache.h
new file mode 100644
index 0000000000..597312e85c
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbi_cache.h
@@ -0,0 +1,203 @@
+// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// file: rbbi_cache.h
+//
+#ifndef RBBI_CACHE_H
+#define RBBI_CACHE_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/rbbi.h"
+#include "unicode/uobject.h"
+
+#include "uvectr32.h"
+
+U_NAMESPACE_BEGIN
+
+/* DictionaryCache  stores the boundaries obtained from a run of dictionary characters.
+ *                 Dictionary boundaries are moved first to this cache, then from here
+ *                 to the main BreakCache, where they may inter-leave with non-dictionary
+ *                 boundaries. The public BreakIterator API always fetches directly
+ *                 from the main BreakCache, not from here.
+ *
+ *                 In common situations, the number of boundaries in a single dictionary run
+ *                 should be quite small, it will be terminated by punctuation, spaces,
+ *                 or any other non-dictionary characters. The main BreakCache may end
+ *                 up with boundaries from multiple dictionary based runs.
+ *
+ *                 The boundaries are stored in a simple ArrayList (vector), with the
+ *                 assumption that they will be accessed sequentially.
+ */                 
+class RuleBasedBreakIterator::DictionaryCache: public UMemory {
+  public:
+     DictionaryCache(RuleBasedBreakIterator *bi, UErrorCode &status);
+     ~DictionaryCache();
+
+     void reset();
+
+     UBool following(int32_t fromPos, int32_t *pos, int32_t *statusIndex);
+     UBool preceding(int32_t fromPos, int32_t *pos, int32_t *statusIndex);
+
+    /**
+     * Populate the cache with the dictionary based boundaries within a region of text.
+     * @param startPos  The start position of a range of text
+     * @param endPos    The end position of a range of text
+     * @param firstRuleStatus The rule status index that applies to the break at startPos
+     * @param otherRuleStatus The rule status index that applies to boundaries other than startPos
+     * @internal
+     */
+    void populateDictionary(int32_t startPos, int32_t endPos,
+                         int32_t firstRuleStatus, int32_t otherRuleStatus);
+
+
+
+    RuleBasedBreakIterator *fBI;
+    
+    UVector32           fBreaks;                // A vector containing the boundaries.
+    int32_t             fPositionInCache;       // Index in fBreaks of last boundary returned by following()
+                                                //    or preceding(). Optimizes sequential access.
+    int32_t             fStart;                 // Text position of first boundary in cache.
+    int32_t             fLimit;                 // Last boundary in cache. Which is the limit of the
+                                                //    text segment being handled by the dictionary.
+    int32_t             fFirstRuleStatusIndex;  // Rule status info for first boundary.
+    int32_t             fOtherRuleStatusIndex;  // Rule status info for 2nd through last boundaries.
+};
+
+
+/*
+ * class BreakCache
+ *
+ * Cache of break boundary positions and rule status values.
+ * Break iterator API functions, next(), previous(), etc., will use cached results
+ * when possible, and otherwise cache new results as they are obtained.
+ *
+ * Uniformly caches both dictionary and rule based (non-dictionary) boundaries.
+ *
+ * The cache is implemented as a single circular buffer.
+ */
+
+/*
+ * size of the circular cache buffer.
+ */
+
+class RuleBasedBreakIterator::BreakCache: public UMemory {
+  public:
+                BreakCache(RuleBasedBreakIterator *bi, UErrorCode &status);
+    virtual     ~BreakCache();
+    void        reset(int32_t pos = 0, int32_t ruleStatus = 0);
+    void        next() {    if (fBufIdx == fEndBufIdx) {
+                                nextOL();
+                            } else {
+                                fBufIdx = modChunkSize(fBufIdx + 1);
+                                fTextIdx = fBI->fPosition = fBoundaries[fBufIdx];
+                                fBI->fRuleStatusIndex = fStatuses[fBufIdx];
+                            }
+                }
+
+
+    void        nextOL();
+    void        previous(UErrorCode &status);
+
+    // Move the iteration state to the position following the startPosition.
+    // Input position must be pinned to the input length.
+    void        following(int32_t startPosition, UErrorCode &status);
+
+    void        preceding(int32_t startPosition, UErrorCode &status);
+
+    /*
+     * Update the state of the public BreakIterator (fBI) to reflect the
+     * current state of the break iterator cache (this).
+     */
+    int32_t     current();
+
+    /**
+     * Add boundaries to the cache near the specified position.
+     * The given position need not be a boundary itself.
+     * The input position must be within the range of the text, and
+     * on a code point boundary.
+     * If the requested position is a break boundary, leave the iteration
+     * position on it.
+     * If the requested position is not a boundary, leave the iteration
+     * position on the preceding boundary and include both the
+     * preceding and following boundaries in the cache.
+     * Additional boundaries, either preceding or following, may be added
+     * to the cache as a side effect.
+     *
+     * Return false if the operation failed.
+     */
+    UBool populateNear(int32_t position, UErrorCode &status);
+
+    /**
+     *  Add boundary(s) to the cache following the current last boundary.
+     *  Return false if at the end of the text, and no more boundaries can be added.
+     *  Leave iteration position at the first newly added boundary, or unchanged if no boundary was added.
+     */
+    UBool populateFollowing();
+
+    /**
+     *  Add one or more boundaries to the cache preceding the first currently cached boundary.
+     *  Leave the iteration position on the first added boundary.
+     *  Return false if no boundaries could be added (if at the start of the text.)
+     */
+    UBool populatePreceding(UErrorCode &status);
+
+    enum UpdatePositionValues {
+        RetainCachePosition = 0,
+        UpdateCachePosition = 1
+    };
+
+    /*
+     * Add the boundary following the current position.
+     * The current position can be left as it was, or changed to the newly added boundary,
+     * as specified by the update parameter.
+     */
+    void addFollowing(int32_t position, int32_t ruleStatusIdx, UpdatePositionValues update);
+
+
+    /*
+     * Add the boundary preceding the current position.
+     * The current position can be left as it was, or changed to the newly added boundary,
+     * as specified by the update parameter.
+     */
+    bool addPreceding(int32_t position, int32_t ruleStatusIdx, UpdatePositionValues update);
+
+    /**
+     *  Set the cache position to the specified position, or, if the position
+     *  falls between to cached boundaries, to the preceding boundary.
+     *  Fails if the requested position is outside of the range of boundaries currently held by the cache.
+     *  The startPosition must be on a code point boundary.
+     *
+     *  Return true if successful, false if the specified position is after
+     *  the last cached boundary or before the first.
+     */
+    UBool                   seek(int32_t startPosition);
+
+    void dumpCache();
+
+  private:
+    static inline int32_t   modChunkSize(int index) { return index & (CACHE_SIZE - 1); }
+
+    static constexpr int32_t CACHE_SIZE = 128;
+    static_assert((CACHE_SIZE & (CACHE_SIZE-1)) == 0, "CACHE_SIZE must be power of two.");
+
+    RuleBasedBreakIterator *fBI;
+    int32_t                 fStartBufIdx;
+    int32_t                 fEndBufIdx;    // inclusive
+
+    int32_t                 fTextIdx;
+    int32_t                 fBufIdx;
+
+    int32_t                 fBoundaries[CACHE_SIZE];
+    uint16_t                fStatuses[CACHE_SIZE];
+
+    UVector32               fSideBuffer;
+};
+
+U_NAMESPACE_END
+
+#endif // #if !UCONFIG_NO_BREAK_ITERATION
+
+#endif // RBBI_CACHE_H
diff --git a/thirdparty/icu4c/common/rbbidata.cpp b/thirdparty/icu4c/common/rbbidata.cpp
new file mode 100644
index 0000000000..193acafc44
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbidata.cpp
@@ -0,0 +1,476 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+*   Copyright (C) 1999-2014 International Business Machines Corporation   *
+*   and others. All rights reserved.                                      *
+***************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/ucptrie.h"
+#include "unicode/utypes.h"
+#include "rbbidata.h"
+#include "rbbirb.h"
+#include "udatamem.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "umutex.h"
+
+#include "uassert.h"
+
+
+U_NAMESPACE_BEGIN
+
+//-----------------------------------------------------------------------------
+//
+//    Constructors.
+//
+//-----------------------------------------------------------------------------
+RBBIDataWrapper::RBBIDataWrapper(const RBBIDataHeader *data, UErrorCode &status) {
+    init0();
+    init(data, status);
+}
+
+RBBIDataWrapper::RBBIDataWrapper(const RBBIDataHeader *data, enum EDontAdopt, UErrorCode &status) {
+    init0();
+    init(data, status);
+    fDontFreeData = TRUE;
+}
+
+RBBIDataWrapper::RBBIDataWrapper(UDataMemory* udm, UErrorCode &status) {
+    init0();
+    if (U_FAILURE(status)) {
+        return;
+    }
+    const DataHeader *dh = udm->pHeader;
+    int32_t headerSize = dh->dataHeader.headerSize;
+    if (  !(headerSize >= 20 &&
+            dh->info.isBigEndian == U_IS_BIG_ENDIAN &&
+            dh->info.charsetFamily == U_CHARSET_FAMILY &&
+            dh->info.dataFormat[0] == 0x42 &&  // dataFormat="Brk "
+            dh->info.dataFormat[1] == 0x72 &&
+            dh->info.dataFormat[2] == 0x6b &&
+            dh->info.dataFormat[3] == 0x20 &&
+            isDataVersionAcceptable(dh->info.formatVersion))
+        ) {
+        status = U_INVALID_FORMAT_ERROR;
+        return;
+    }
+    const char *dataAsBytes = reinterpret_cast<const char *>(dh);
+    const RBBIDataHeader *rbbidh = reinterpret_cast<const RBBIDataHeader *>(dataAsBytes + headerSize);
+    init(rbbidh, status);
+    fUDataMem = udm;
+}
+
+UBool RBBIDataWrapper::isDataVersionAcceptable(const UVersionInfo version) {
+    return RBBI_DATA_FORMAT_VERSION[0] == version[0];
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//    init().   Does most of the work of construction, shared between the
+//              constructors.
+//
+//-----------------------------------------------------------------------------
+void RBBIDataWrapper::init0() {
+    fHeader = NULL;
+    fForwardTable = NULL;
+    fReverseTable = NULL;
+    fRuleSource   = NULL;
+    fRuleStatusTable = NULL;
+    fTrie         = NULL;
+    fUDataMem     = NULL;
+    fRefCount     = 0;
+    fDontFreeData = TRUE;
+}
+
+void RBBIDataWrapper::init(const RBBIDataHeader *data, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    fHeader = data;
+    if (fHeader->fMagic != 0xb1a0 || !isDataVersionAcceptable(fHeader->fFormatVersion)) {
+        status = U_INVALID_FORMAT_ERROR;
+        return;
+    }
+    // Note: in ICU version 3.2 and earlier, there was a formatVersion 1
+    //       that is no longer supported.  At that time fFormatVersion was
+    //       an int32_t field, rather than an array of 4 bytes.
+
+    fDontFreeData = FALSE;
+    if (data->fFTableLen != 0) {
+        fForwardTable = (RBBIStateTable *)((char *)data + fHeader->fFTable);
+    }
+    if (data->fRTableLen != 0) {
+        fReverseTable = (RBBIStateTable *)((char *)data + fHeader->fRTable);
+    }
+
+    fTrie = ucptrie_openFromBinary(UCPTRIE_TYPE_FAST,
+                                   UCPTRIE_VALUE_BITS_ANY,
+                                   (uint8_t *)data + fHeader->fTrie,
+                                   fHeader->fTrieLen,
+                                   nullptr,           // *actual length
+                                   &status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    UCPTrieValueWidth width = ucptrie_getValueWidth(fTrie);
+    if (!(width == UCPTRIE_VALUE_BITS_8 || width == UCPTRIE_VALUE_BITS_16)) {
+        status = U_INVALID_FORMAT_ERROR;
+        return;
+    }
+
+    fRuleSource   = ((char *)data + fHeader->fRuleSource);
+    fRuleString = UnicodeString::fromUTF8(StringPiece(fRuleSource, fHeader->fRuleSourceLen));
+    U_ASSERT(data->fRuleSourceLen > 0);
+
+    fRuleStatusTable = (int32_t *)((char *)data + fHeader->fStatusTable);
+    fStatusMaxIdx    = data->fStatusTableLen / sizeof(int32_t);
+
+    fRefCount = 1;
+
+#ifdef RBBI_DEBUG
+    char *debugEnv = getenv("U_RBBIDEBUG");
+    if (debugEnv && uprv_strstr(debugEnv, "data")) {this->printData();}
+#endif
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//    Destructor.     Don't call this - use removeReference() instead.
+//
+//-----------------------------------------------------------------------------
+RBBIDataWrapper::~RBBIDataWrapper() {
+    U_ASSERT(fRefCount == 0);
+    ucptrie_close(fTrie);
+    fTrie = nullptr;
+    if (fUDataMem) {
+        udata_close(fUDataMem);
+    } else if (!fDontFreeData) {
+        uprv_free((void *)fHeader);
+    }
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   Operator ==    Consider two RBBIDataWrappers to be equal if they
+//                  refer to the same underlying data.  Although
+//                  the data wrappers are normally shared between
+//                  iterator instances, it's possible to independently
+//                  open the same data twice, and get two instances, which
+//                  should still be ==.
+//
+//-----------------------------------------------------------------------------
+UBool RBBIDataWrapper::operator ==(const RBBIDataWrapper &other) const {
+    if (fHeader == other.fHeader) {
+        return TRUE;
+    }
+    if (fHeader->fLength != other.fHeader->fLength) {
+        return FALSE;
+    }
+    if (uprv_memcmp(fHeader, other.fHeader, fHeader->fLength) == 0) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+int32_t  RBBIDataWrapper::hashCode() {
+    return fHeader->fFTableLen;
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//    Reference Counting.   A single RBBIDataWrapper object is shared among
+//                          however many RulesBasedBreakIterator instances are
+//                          referencing the same data.
+//
+//-----------------------------------------------------------------------------
+void RBBIDataWrapper::removeReference() {
+    if (umtx_atomic_dec(&fRefCount) == 0) {
+        delete this;
+    }
+}
+
+
+RBBIDataWrapper *RBBIDataWrapper::addReference() {
+   umtx_atomic_inc(&fRefCount);
+   return this;
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//  getRuleSourceString
+//
+//-----------------------------------------------------------------------------
+const UnicodeString &RBBIDataWrapper::getRuleSourceString() const {
+    return fRuleString;
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//  print   -  debugging function to dump the runtime data tables.
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void  RBBIDataWrapper::printTable(const char *heading, const RBBIStateTable *table) {
+    uint32_t   c;
+    uint32_t   s;
+
+    RBBIDebugPrintf("%s\n", heading);
+
+    RBBIDebugPrintf("   fDictCategoriesStart: %d\n", table->fDictCategoriesStart);
+    RBBIDebugPrintf("   fLookAheadResultsSize: %d\n", table->fLookAheadResultsSize);
+    RBBIDebugPrintf("   Flags: %4x RBBI_LOOKAHEAD_HARD_BREAK=%s RBBI_BOF_REQUIRED=%s  RBBI_8BITS_ROWS=%s\n",
+                    table->fFlags,
+                    table->fFlags & RBBI_LOOKAHEAD_HARD_BREAK ? "T" : "F",
+                    table->fFlags & RBBI_BOF_REQUIRED ? "T" : "F",
+                    table->fFlags & RBBI_8BITS_ROWS ? "T" : "F");
+    RBBIDebugPrintf("\nState |  Acc  LA TagIx");
+    for (c=0; c<fHeader->fCatCount; c++) {RBBIDebugPrintf("%3d ", c);}
+    RBBIDebugPrintf("\n------|---------------"); for (c=0;c<fHeader->fCatCount; c++) {
+        RBBIDebugPrintf("----");
+    }
+    RBBIDebugPrintf("\n");
+
+    if (table == NULL) {
+        RBBIDebugPrintf("         N U L L   T A B L E\n\n");
+        return;
+    }
+    UBool use8Bits = table->fFlags & RBBI_8BITS_ROWS;
+    for (s=0; s<table->fNumStates; s++) {
+        RBBIStateTableRow *row = (RBBIStateTableRow *)
+                                  (table->fTableData + (table->fRowLen * s));
+        if (use8Bits) {
+            RBBIDebugPrintf("%4d  |  %3d %3d %3d ", s, row->r8.fAccepting, row->r8.fLookAhead, row->r8.fTagsIdx);
+            for (c=0; c<fHeader->fCatCount; c++)  {
+                RBBIDebugPrintf("%3d ", row->r8.fNextState[c]);
+            }
+        } else {
+            RBBIDebugPrintf("%4d  |  %3d %3d %3d ", s, row->r16.fAccepting, row->r16.fLookAhead, row->r16.fTagsIdx);
+            for (c=0; c<fHeader->fCatCount; c++)  {
+                RBBIDebugPrintf("%3d ", row->r16.fNextState[c]);
+            }
+        }
+        RBBIDebugPrintf("\n");
+    }
+    RBBIDebugPrintf("\n");
+}
+#endif
+
+
+void  RBBIDataWrapper::printData() {
+#ifdef RBBI_DEBUG
+    RBBIDebugPrintf("RBBI Data at %p\n", (void *)fHeader);
+    RBBIDebugPrintf("   Version = {%d %d %d %d}\n", fHeader->fFormatVersion[0], fHeader->fFormatVersion[1],
+                                                    fHeader->fFormatVersion[2], fHeader->fFormatVersion[3]);
+    RBBIDebugPrintf("   total length of data  = %d\n", fHeader->fLength);
+    RBBIDebugPrintf("   number of character categories = %d\n\n", fHeader->fCatCount);
+
+    printTable("Forward State Transition Table", fForwardTable);
+    printTable("Reverse State Transition Table", fReverseTable);
+
+    RBBIDebugPrintf("\nOrignal Rules source:\n");
+    for (int32_t c=0; fRuleSource[c] != 0; c++) {
+        RBBIDebugPrintf("%c", fRuleSource[c]);
+    }
+    RBBIDebugPrintf("\n\n");
+#endif
+}
+
+
+U_NAMESPACE_END
+U_NAMESPACE_USE
+
+//-----------------------------------------------------------------------------
+//
+//  ubrk_swap   -  byte swap and char encoding swap of RBBI data
+//
+//-----------------------------------------------------------------------------
+
+U_CAPI int32_t U_EXPORT2
+ubrk_swap(const UDataSwapper *ds, const void *inData, int32_t length, void *outData,
+           UErrorCode *status) {
+
+    if (status == NULL || U_FAILURE(*status)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<-1 || (length>0 && outData==NULL)) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    //
+    //  Check that the data header is for for break data.
+    //    (Header contents are defined in genbrk.cpp)
+    //
+    const UDataInfo *pInfo = (const UDataInfo *)((const char *)inData+4);
+    if(!(  pInfo->dataFormat[0]==0x42 &&   /* dataFormat="Brk " */
+           pInfo->dataFormat[1]==0x72 &&
+           pInfo->dataFormat[2]==0x6b &&
+           pInfo->dataFormat[3]==0x20 &&
+           RBBIDataWrapper::isDataVersionAcceptable(pInfo->formatVersion) )) {
+        udata_printError(ds, "ubrk_swap(): data format %02x.%02x.%02x.%02x (format version %02x) is not recognized\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0]);
+        *status=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    //
+    // Swap the data header.  (This is the generic ICU Data Header, not the RBBI Specific
+    //                         RBBIDataHeader).  This swap also conveniently gets us
+    //                         the size of the ICU d.h., which lets us locate the start
+    //                         of the RBBI specific data.
+    //
+    int32_t headerSize=udata_swapDataHeader(ds, inData, length, outData, status);
+
+
+    //
+    // Get the RRBI Data Header, and check that it appears to be OK.
+    //
+    const uint8_t  *inBytes =(const uint8_t *)inData+headerSize;
+    RBBIDataHeader *rbbiDH = (RBBIDataHeader *)inBytes;
+    if (ds->readUInt32(rbbiDH->fMagic) != 0xb1a0 || 
+            !RBBIDataWrapper::isDataVersionAcceptable(rbbiDH->fFormatVersion) ||
+            ds->readUInt32(rbbiDH->fLength)  <  sizeof(RBBIDataHeader)) {
+        udata_printError(ds, "ubrk_swap(): RBBI Data header is invalid.\n");
+        *status=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    //
+    // Prefight operation?  Just return the size
+    //
+    int32_t breakDataLength = ds->readUInt32(rbbiDH->fLength);
+    int32_t totalSize = headerSize + breakDataLength;
+    if (length < 0) {
+        return totalSize;
+    }
+
+    //
+    // Check that length passed in is consistent with length from RBBI data header.
+    //
+    if (length < totalSize) {
+        udata_printError(ds, "ubrk_swap(): too few bytes (%d after ICU Data header) for break data.\n",
+                            breakDataLength);
+        *status=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+        }
+
+
+    //
+    // Swap the Data.  Do the data itself first, then the RBBI Data Header, because
+    //                 we need to reference the header to locate the data, and an
+    //                 inplace swap of the header leaves it unusable.
+    //
+    uint8_t         *outBytes = (uint8_t *)outData + headerSize;
+    RBBIDataHeader  *outputDH = (RBBIDataHeader *)outBytes;
+
+    int32_t   tableStartOffset;
+    int32_t   tableLength;
+
+    //
+    // If not swapping in place, zero out the output buffer before starting.
+    //    Individual tables and other data items within are aligned to 8 byte boundaries
+    //    when originally created.  Any unused space between items needs to be zero.
+    //
+    if (inBytes != outBytes) {
+        uprv_memset(outBytes, 0, breakDataLength);
+    }
+
+    //
+    // Each state table begins with several 32 bit fields.  Calculate the size
+    //   in bytes of these.
+    //
+    int32_t         topSize = offsetof(RBBIStateTable, fTableData);
+
+    // Forward state table.
+    tableStartOffset = ds->readUInt32(rbbiDH->fFTable);
+    tableLength      = ds->readUInt32(rbbiDH->fFTableLen);
+
+    if (tableLength > 0) {
+        RBBIStateTable *rbbiST = (RBBIStateTable *)(inBytes+tableStartOffset);
+        UBool use8Bits = ds->readUInt32(rbbiST->fFlags) & RBBI_8BITS_ROWS;
+
+        ds->swapArray32(ds, inBytes+tableStartOffset, topSize,
+                            outBytes+tableStartOffset, status);
+
+        // Swap the state table if the table is in 16 bits.
+        if (use8Bits) {
+            if (outBytes != inBytes) {
+                uprv_memmove(outBytes+tableStartOffset+topSize,
+                             inBytes+tableStartOffset+topSize,
+                             tableLength-topSize);
+            }
+        } else {
+            ds->swapArray16(ds, inBytes+tableStartOffset+topSize, tableLength-topSize,
+                                outBytes+tableStartOffset+topSize, status);
+        }
+    }
+
+    // Reverse state table.  Same layout as forward table, above.
+    tableStartOffset = ds->readUInt32(rbbiDH->fRTable);
+    tableLength      = ds->readUInt32(rbbiDH->fRTableLen);
+
+    if (tableLength > 0) {
+        RBBIStateTable *rbbiST = (RBBIStateTable *)(inBytes+tableStartOffset);
+        UBool use8Bits = ds->readUInt32(rbbiST->fFlags) & RBBI_8BITS_ROWS;
+
+        ds->swapArray32(ds, inBytes+tableStartOffset, topSize,
+                            outBytes+tableStartOffset, status);
+
+        // Swap the state table if the table is in 16 bits.
+        if (use8Bits) {
+            if (outBytes != inBytes) {
+                uprv_memmove(outBytes+tableStartOffset+topSize,
+                             inBytes+tableStartOffset+topSize,
+                             tableLength-topSize);
+            }
+        } else {
+            ds->swapArray16(ds, inBytes+tableStartOffset+topSize, tableLength-topSize,
+                                outBytes+tableStartOffset+topSize, status);
+        }
+    }
+
+    // Trie table for character categories
+    ucptrie_swap(ds, inBytes+ds->readUInt32(rbbiDH->fTrie), ds->readUInt32(rbbiDH->fTrieLen),
+                     outBytes+ds->readUInt32(rbbiDH->fTrie), status);
+
+    // Source Rules Text.  It's UTF8 data
+    if (outBytes != inBytes) {
+        uprv_memmove(outBytes+ds->readUInt32(rbbiDH->fRuleSource),
+                     inBytes+ds->readUInt32(rbbiDH->fRuleSource),
+                     ds->readUInt32(rbbiDH->fRuleSourceLen));
+    }
+
+    // Table of rule status values.  It's all int_32 values
+    ds->swapArray32(ds, inBytes+ds->readUInt32(rbbiDH->fStatusTable), ds->readUInt32(rbbiDH->fStatusTableLen),
+                        outBytes+ds->readUInt32(rbbiDH->fStatusTable), status);
+
+    // And, last, the header.
+    //   It is all int32_t values except for fFormataVersion, which is an array of four bytes.
+    //   Swap the whole thing as int32_t, then re-swap the one field.
+    //
+    ds->swapArray32(ds, inBytes, sizeof(RBBIDataHeader), outBytes, status);
+    ds->swapArray32(ds, outputDH->fFormatVersion, 4, outputDH->fFormatVersion, status);
+
+    return totalSize;
+}
+
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbidata.h b/thirdparty/icu4c/common/rbbidata.h
new file mode 100644
index 0000000000..3749f16799
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbidata.h
@@ -0,0 +1,212 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2014 International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  rbbidata.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   RBBI data formats  Includes
+*
+*                          Structs that describes the format of the Binary RBBI data,
+*                          as it is stored in ICU's data file.
+*
+*      RBBIDataWrapper  -  Instances of this class sit between the
+*                          raw data structs and the RulesBasedBreakIterator objects
+*                          that are created by applications.  The wrapper class
+*                          provides reference counting for the underlying data,
+*                          and direct pointers to data that would not otherwise
+*                          be accessible without ugly pointer arithmetic.  The
+*                          wrapper does not attempt to provide any higher level
+*                          abstractions for the data itself.
+*
+*                          There will be only one instance of RBBIDataWrapper for any
+*                          set of RBBI run time data being shared by instances
+*                          (clones) of RulesBasedBreakIterator.
+*/
+
+#ifndef __RBBIDATA_H__
+#define __RBBIDATA_H__
+
+#include "unicode/utypes.h"
+#include "unicode/udata.h"
+#include "udataswp.h"
+
+/**
+ * Swap RBBI data. See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode);
+
+#ifdef __cplusplus
+
+#include "unicode/ucptrie.h"
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+#include "unicode/uversion.h"
+#include "umutex.h"
+
+
+U_NAMESPACE_BEGIN
+
+// The current RBBI data format version.
+static const uint8_t RBBI_DATA_FORMAT_VERSION[] = {6, 0, 0, 0};
+
+/*  
+ *   The following structs map exactly onto the raw data from ICU common data file. 
+ */
+struct RBBIDataHeader {
+    uint32_t         fMagic;           /*  == 0xbla0                                               */
+    UVersionInfo     fFormatVersion;   /* Data Format.  Same as the value in struct UDataInfo      */
+                                       /*   if there is one associated with this data.             */
+                                       /*     (version originates in rbbi, is copied to UDataInfo) */
+    uint32_t         fLength;          /*  Total length in bytes of this RBBI Data,                */
+                                       /*      including all sections, not just the header.        */
+    uint32_t         fCatCount;        /*  Number of character categories.                         */
+
+    /*                                                                        */
+    /*  Offsets and sizes of each of the subsections within the RBBI data.    */
+    /*  All offsets are bytes from the start of the RBBIDataHeader.           */
+    /*  All sizes are in bytes.                                               */
+    /*                                                                        */
+    uint32_t         fFTable;         /*  forward state transition table. */
+    uint32_t         fFTableLen;
+    uint32_t         fRTable;         /*  Offset to the reverse state transition table. */
+    uint32_t         fRTableLen;
+    uint32_t         fTrie;           /*  Offset to Trie data for character categories */
+    uint32_t         fTrieLen;
+    uint32_t         fRuleSource;     /*  Offset to the source for for the break */
+    uint32_t         fRuleSourceLen;  /*    rules.  Stored UChar *. */
+    uint32_t         fStatusTable;    /* Offset to the table of rule status values */
+    uint32_t         fStatusTableLen;
+
+    uint32_t         fReserved[6];    /*  Reserved for expansion */
+
+};
+
+
+
+template <typename T>
+struct RBBIStateTableRowT {
+    T               fAccepting;    //  Non-zero if this row is for an accepting state.
+                                   //  Value 0: not an accepting state.
+                                   //        1: (ACCEPTING_UNCONDITIONAL) Unconditional Accepting state.
+                                   //       >1: Look-ahead match has completed.
+                                   //           Actual boundary position happened earlier.
+                                   //           Value here == fLookAhead in earlier
+                                   //           state, at actual boundary pos.
+    T               fLookAhead;    //  Non-zero if this row is for a state that
+                                   //    corresponds to a '/' in the rule source.
+                                   //    Value is the same as the fAccepting
+                                   //    value for the rule (which will appear
+                                   //    in a different state.
+    T               fTagsIdx;      //  Non-zero if this row covers a {tagged} position
+                                   //    from a rule.  Value is the index in the
+                                   //    StatusTable of the set of matching
+                                   //    tags (rule status values)
+    T               fNextState[1]; //  Next State, indexed by char category.
+                                   //    Variable-length array declared with length 1
+                                   //    to disable bounds checkers.
+                                   //    Array Size is actually fData->fHeader->fCatCount
+                                   //    CAUTION:  see RBBITableBuilder::getTableSize()
+                                   //              before changing anything here.
+};
+
+typedef RBBIStateTableRowT<uint8_t> RBBIStateTableRow8;
+typedef RBBIStateTableRowT<uint16_t> RBBIStateTableRow16;
+
+constexpr uint16_t ACCEPTING_UNCONDITIONAL = 1;   // Value constant for RBBIStateTableRow::fAccepting
+
+union RBBIStateTableRow {
+  RBBIStateTableRow16 r16;
+  RBBIStateTableRow8 r8;
+};
+
+struct RBBIStateTable {
+    uint32_t         fNumStates;            // Number of states.
+    uint32_t         fRowLen;               // Length of a state table row, in bytes.
+    uint32_t         fDictCategoriesStart;  // Char category number of the first dictionary
+                                            //   char class, or the the largest category number + 1
+                                            //   if there are no dictionary categories.
+    uint32_t         fLookAheadResultsSize; // Size of run-time array required for holding
+                                            //   look-ahead results. Indexed by row.fLookAhead.
+    uint32_t         fFlags;                // Option Flags for this state table.
+    char             fTableData[1];         // First RBBIStateTableRow begins here.
+                                            //   Variable-length array declared with length 1
+                                            //   to disable bounds checkers.
+                                            //   (making it char[] simplifies ugly address
+                                            //   arithmetic for indexing variable length rows.)
+};
+
+constexpr uint32_t RBBI_LOOKAHEAD_HARD_BREAK = 1;
+constexpr uint32_t RBBI_BOF_REQUIRED = 2;
+constexpr uint32_t RBBI_8BITS_ROWS = 4;
+
+
+/*                                        */
+/*   The reference counting wrapper class */
+/*                                        */
+class RBBIDataWrapper : public UMemory {
+public:
+    enum EDontAdopt {
+        kDontAdopt
+    };
+    RBBIDataWrapper(const RBBIDataHeader *data, UErrorCode &status);
+    RBBIDataWrapper(const RBBIDataHeader *data, enum EDontAdopt dontAdopt, UErrorCode &status);
+    RBBIDataWrapper(UDataMemory* udm, UErrorCode &status);
+    ~RBBIDataWrapper();
+
+    static UBool          isDataVersionAcceptable(const UVersionInfo version);
+
+    void                  init0();
+    void                  init(const RBBIDataHeader *data, UErrorCode &status);
+    RBBIDataWrapper      *addReference();
+    void                  removeReference();
+    UBool                 operator ==(const RBBIDataWrapper &other) const;
+    int32_t               hashCode();
+    const UnicodeString  &getRuleSourceString() const;
+    void                  printData();
+    void                  printTable(const char *heading, const RBBIStateTable *table);
+
+    /*                                     */
+    /*   Pointers to items within the data */
+    /*                                     */
+    const RBBIDataHeader     *fHeader;
+    const RBBIStateTable     *fForwardTable;
+    const RBBIStateTable     *fReverseTable;
+    const char               *fRuleSource;
+    const int32_t            *fRuleStatusTable; 
+
+    /* number of int32_t values in the rule status table.   Used to sanity check indexing */
+    int32_t             fStatusMaxIdx;
+
+    UCPTrie             *fTrie;
+
+private:
+    u_atomic_int32_t    fRefCount;
+    UDataMemory        *fUDataMem;
+    UnicodeString       fRuleString;
+    UBool               fDontFreeData;
+
+    RBBIDataWrapper(const RBBIDataWrapper &other) = delete; /*  forbid copying of this class */
+    RBBIDataWrapper &operator=(const RBBIDataWrapper &other) = delete; /*  forbid copying of this class */
+};
+
+
+
+U_NAMESPACE_END
+
+U_CFUNC UBool rbbi_cleanup(void);
+
+#endif /* C++ */
+
+#endif
diff --git a/thirdparty/icu4c/common/rbbinode.cpp b/thirdparty/icu4c/common/rbbinode.cpp
new file mode 100644
index 0000000000..69d84151fe
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbinode.cpp
@@ -0,0 +1,372 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+*   Copyright (C) 2002-2016 International Business Machines Corporation   *
+*   and others. All rights reserved.                                      *
+***************************************************************************
+*/
+
+//
+//  File:  rbbinode.cpp
+//
+//         Implementation of class RBBINode, which represents a node in the
+//         tree generated when parsing the Rules Based Break Iterator rules.
+//
+//         This "Class" is actually closer to a struct.
+//         Code using it is expected to directly access fields much of the time.
+//
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/unistr.h"
+#include "unicode/uniset.h"
+#include "unicode/uchar.h"
+#include "unicode/parsepos.h"
+
+#include "cstr.h"
+#include "uvector.h"
+
+#include "rbbirb.h"
+#include "rbbinode.h"
+
+#include "uassert.h"
+
+
+U_NAMESPACE_BEGIN
+
+#ifdef RBBI_DEBUG
+static int  gLastSerial = 0;
+#endif
+
+
+//-------------------------------------------------------------------------
+//
+//    Constructor.   Just set the fields to reasonable default values.
+//
+//-------------------------------------------------------------------------
+RBBINode::RBBINode(NodeType t) : UMemory() {
+#ifdef RBBI_DEBUG
+    fSerialNum    = ++gLastSerial;
+#endif
+    fType         = t;
+    fParent       = NULL;
+    fLeftChild    = NULL;
+    fRightChild   = NULL;
+    fInputSet     = NULL;
+    fFirstPos     = 0;
+    fLastPos      = 0;
+    fNullable     = FALSE;
+    fLookAheadEnd = FALSE;
+    fRuleRoot     = FALSE;
+    fChainIn      = FALSE;
+    fVal          = 0;
+    fPrecedence   = precZero;
+
+    UErrorCode     status = U_ZERO_ERROR;
+    fFirstPosSet  = new UVector(status);  // TODO - get a real status from somewhere
+    fLastPosSet   = new UVector(status);
+    fFollowPos    = new UVector(status);
+    if      (t==opCat)    {fPrecedence = precOpCat;}
+    else if (t==opOr)     {fPrecedence = precOpOr;}
+    else if (t==opStart)  {fPrecedence = precStart;}
+    else if (t==opLParen) {fPrecedence = precLParen;}
+
+}
+
+
+RBBINode::RBBINode(const RBBINode &other) : UMemory(other) {
+#ifdef RBBI_DEBUG
+    fSerialNum   = ++gLastSerial;
+#endif
+    fType        = other.fType;
+    fParent      = NULL;
+    fLeftChild   = NULL;
+    fRightChild  = NULL;
+    fInputSet    = other.fInputSet;
+    fPrecedence  = other.fPrecedence;
+    fText        = other.fText;
+    fFirstPos    = other.fFirstPos;
+    fLastPos     = other.fLastPos;
+    fNullable    = other.fNullable;
+    fVal         = other.fVal;
+    fRuleRoot    = FALSE;
+    fChainIn     = other.fChainIn;
+    UErrorCode     status = U_ZERO_ERROR;
+    fFirstPosSet = new UVector(status);   // TODO - get a real status from somewhere
+    fLastPosSet  = new UVector(status);
+    fFollowPos   = new UVector(status);
+}
+
+
+//-------------------------------------------------------------------------
+//
+//    Destructor.   Deletes both this node AND any child nodes,
+//                  except in the case of variable reference nodes.  For
+//                  these, the l. child points back to the definition, which
+//                  is common for all references to the variable, meaning
+//                  it can't be deleted here.
+//
+//-------------------------------------------------------------------------
+RBBINode::~RBBINode() {
+    // printf("deleting node %8x   serial %4d\n", this, this->fSerialNum);
+    delete fInputSet;
+    fInputSet = NULL;
+
+    switch (this->fType) {
+    case varRef:
+    case setRef:
+        // for these node types, multiple instances point to the same "children"
+        // Storage ownership of children handled elsewhere.  Don't delete here.
+        break;
+
+    default:
+        delete        fLeftChild;
+        fLeftChild =   NULL;
+        delete        fRightChild;
+        fRightChild = NULL;
+    }
+
+
+    delete fFirstPosSet;
+    delete fLastPosSet;
+    delete fFollowPos;
+
+}
+
+
+//-------------------------------------------------------------------------
+//
+//    cloneTree     Make a copy of the subtree rooted at this node.
+//                  Discard any variable references encountered along the way,
+//                  and replace with copies of the variable's definitions.
+//                  Used to replicate the expression underneath variable
+//                  references in preparation for generating the DFA tables.
+//
+//-------------------------------------------------------------------------
+RBBINode *RBBINode::cloneTree() {
+    RBBINode    *n;
+
+    if (fType == RBBINode::varRef) {
+        // If the current node is a variable reference, skip over it
+        //   and clone the definition of the variable instead.
+        n = fLeftChild->cloneTree();
+    } else if (fType == RBBINode::uset) {
+        n = this;
+    } else {
+        n = new RBBINode(*this);
+        // Check for null pointer.
+        if (n != NULL) {
+            if (fLeftChild != NULL) {
+                n->fLeftChild          = fLeftChild->cloneTree();
+                n->fLeftChild->fParent = n;
+            }
+            if (fRightChild != NULL) {
+                n->fRightChild          = fRightChild->cloneTree();
+                n->fRightChild->fParent = n;
+            }
+        }
+    }
+    return n;
+}
+
+
+
+//-------------------------------------------------------------------------
+//
+//   flattenVariables   Walk a parse tree, replacing any variable
+//                      references with a copy of the variable's definition.
+//                      Aside from variables, the tree is not changed.
+//
+//                      Return the root of the tree.  If the root was not a variable
+//                      reference, it remains unchanged - the root we started with
+//                      is the root we return.  If, however, the root was a variable
+//                      reference, the root of the newly cloned replacement tree will
+//                      be returned, and the original tree deleted.
+//
+//                      This function works by recursively walking the tree
+//                      without doing anything until a variable reference is
+//                      found, then calling cloneTree() at that point.  Any
+//                      nested references are handled by cloneTree(), not here.
+//
+//-------------------------------------------------------------------------
+RBBINode *RBBINode::flattenVariables() {
+    if (fType == varRef) {
+        RBBINode *retNode  = fLeftChild->cloneTree();
+        if (retNode != NULL) {
+            retNode->fRuleRoot = this->fRuleRoot;
+            retNode->fChainIn  = this->fChainIn;
+        }
+        delete this;   // TODO: undefined behavior. Fix.
+        return retNode;
+    }
+
+    if (fLeftChild != NULL) {
+        fLeftChild = fLeftChild->flattenVariables();
+        fLeftChild->fParent  = this;
+    }
+    if (fRightChild != NULL) {
+        fRightChild = fRightChild->flattenVariables();
+        fRightChild->fParent = this;
+    }
+    return this;
+}
+
+
+//-------------------------------------------------------------------------
+//
+//  flattenSets    Walk the parse tree, replacing any nodes of type setRef
+//                 with a copy of the expression tree for the set.  A set's
+//                 equivalent expression tree is precomputed and saved as
+//                 the left child of the uset node.
+//
+//-------------------------------------------------------------------------
+void RBBINode::flattenSets() {
+    U_ASSERT(fType != setRef);
+
+    if (fLeftChild != NULL) {
+        if (fLeftChild->fType==setRef) {
+            RBBINode *setRefNode = fLeftChild;
+            RBBINode *usetNode   = setRefNode->fLeftChild;
+            RBBINode *replTree   = usetNode->fLeftChild;
+            fLeftChild           = replTree->cloneTree();
+            fLeftChild->fParent  = this;
+            delete setRefNode;
+        } else {
+            fLeftChild->flattenSets();
+        }
+    }
+
+    if (fRightChild != NULL) {
+        if (fRightChild->fType==setRef) {
+            RBBINode *setRefNode = fRightChild;
+            RBBINode *usetNode   = setRefNode->fLeftChild;
+            RBBINode *replTree   = usetNode->fLeftChild;
+            fRightChild           = replTree->cloneTree();
+            fRightChild->fParent  = this;
+            delete setRefNode;
+        } else {
+            fRightChild->flattenSets();
+        }
+    }
+}
+
+
+
+//-------------------------------------------------------------------------
+//
+//   findNodes()     Locate all the nodes of the specified type, starting
+//                   at the specified root.
+//
+//-------------------------------------------------------------------------
+void   RBBINode::findNodes(UVector *dest, RBBINode::NodeType kind, UErrorCode &status) {
+    /* test for buffer overflows */
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (fType == kind) {
+        dest->addElement(this, status);
+    }
+    if (fLeftChild != NULL) {
+        fLeftChild->findNodes(dest, kind, status);
+    }
+    if (fRightChild != NULL) {
+        fRightChild->findNodes(dest, kind, status);
+    }
+}
+
+
+//-------------------------------------------------------------------------
+//
+//    print.         Print out a single node, for debugging.
+//
+//-------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+
+static int32_t serial(const RBBINode *node) {
+    return (node == NULL? -1 : node->fSerialNum);
+}
+
+
+void RBBINode::printNode(const RBBINode *node) {
+    static const char * const nodeTypeNames[] = {
+                "setRef",
+                "uset",
+                "varRef",
+                "leafChar",
+                "lookAhead",
+                "tag",
+                "endMark",
+                "opStart",
+                "opCat",
+                "opOr",
+                "opStar",
+                "opPlus",
+                "opQuestion",
+                "opBreak",
+                "opReverse",
+                "opLParen"
+    };
+
+    if (node==NULL) {
+        RBBIDebugPrintf("%10p", (void *)node);
+    } else {
+        RBBIDebugPrintf("%10p %5d %12s %c%c  %5d       %5d     %5d       %6d     %d ",
+            (void *)node, node->fSerialNum, nodeTypeNames[node->fType],
+            node->fRuleRoot?'R':' ', node->fChainIn?'C':' ',
+            serial(node->fLeftChild), serial(node->fRightChild), serial(node->fParent),
+            node->fFirstPos, node->fVal);
+        if (node->fType == varRef) {
+            RBBI_DEBUG_printUnicodeString(node->fText);
+        }
+    }
+    RBBIDebugPrintf("\n");
+}
+#endif
+
+
+#ifdef RBBI_DEBUG
+U_CFUNC void RBBI_DEBUG_printUnicodeString(const UnicodeString &s, int minWidth) {
+    RBBIDebugPrintf("%*s", minWidth, CStr(s)());
+}
+#endif
+
+
+//-------------------------------------------------------------------------
+//
+//    print.         Print out the tree of nodes rooted at "this"
+//
+//-------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBINode::printNodeHeader() {
+    RBBIDebugPrintf(" Address   serial        type     LeftChild  RightChild   Parent   position value\n");
+}
+    
+void RBBINode::printTree(const RBBINode *node, UBool printHeading) {
+    if (printHeading) {
+        printNodeHeader();
+    }
+    printNode(node);
+    if (node != NULL) {
+        // Only dump the definition under a variable reference if asked to.
+        // Unconditinally dump children of all other node types.
+        if (node->fType != varRef) {
+            if (node->fLeftChild != NULL) {
+                printTree(node->fLeftChild, FALSE);
+            }
+            
+            if (node->fRightChild != NULL) {
+                printTree(node->fRightChild, FALSE);
+            }
+        }
+    }
+}
+#endif
+
+
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbinode.h b/thirdparty/icu4c/common/rbbinode.h
new file mode 100644
index 0000000000..cff3ba70c4
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbinode.h
@@ -0,0 +1,127 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/********************************************************************
+ * COPYRIGHT:
+ * Copyright (c) 2001-2016, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ ********************************************************************/
+
+#ifndef RBBINODE_H
+#define RBBINODE_H
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+
+//
+//  class RBBINode
+//
+//                    Represents a node in the parse tree generated when reading
+//                    a rule file.
+//
+
+U_NAMESPACE_BEGIN
+
+class    UnicodeSet;
+class    UVector;
+
+class RBBINode : public UMemory {
+    public:
+        enum NodeType {
+            setRef,
+            uset,
+            varRef,
+            leafChar,
+            lookAhead,
+            tag,
+            endMark,
+            opStart,
+            opCat,
+            opOr,
+            opStar,
+            opPlus,
+            opQuestion,
+            opBreak,
+            opReverse,
+            opLParen
+        };
+
+        enum OpPrecedence {      
+            precZero,
+            precStart,
+            precLParen,
+            precOpOr,
+            precOpCat
+        };
+            
+        NodeType      fType;
+        RBBINode      *fParent;
+        RBBINode      *fLeftChild;
+        RBBINode      *fRightChild;
+        UnicodeSet    *fInputSet;           // For uset nodes only.
+        OpPrecedence  fPrecedence;          // For binary ops only.
+        
+        UnicodeString fText;                // Text corresponding to this node.
+                                            //   May be lazily evaluated when (if) needed
+                                            //   for some node types.
+        int           fFirstPos;            // Position in the rule source string of the
+                                            //   first text associated with the node.
+                                            //   If there's a left child, this will be the same
+                                            //   as that child's left pos.
+        int           fLastPos;             //  Last position in the rule source string
+                                            //    of any text associated with this node.
+                                            //    If there's a right child, this will be the same
+                                            //    as that child's last postion.
+
+        UBool         fNullable;            // See Aho.
+        int32_t       fVal;                 // For leafChar nodes, the value.
+                                            //   Values are the character category,
+                                            //   corresponds to columns in the final
+                                            //   state transition table.
+
+        UBool         fLookAheadEnd;        // For endMark nodes, set true if
+                                            //   marking the end of a look-ahead rule.
+
+        UBool         fRuleRoot;            // True if this node is the root of a rule.
+        UBool         fChainIn;             // True if chaining into this rule is allowed
+                                            //     (no '^' present).
+
+        UVector       *fFirstPosSet;
+        UVector       *fLastPosSet;         // TODO: rename fFirstPos & fLastPos to avoid confusion.
+        UVector       *fFollowPos;
+
+
+        RBBINode(NodeType t);
+        RBBINode(const RBBINode &other);
+        ~RBBINode();
+        
+        RBBINode    *cloneTree();
+        RBBINode    *flattenVariables();
+        void         flattenSets();
+        void         findNodes(UVector *dest, RBBINode::NodeType kind, UErrorCode &status);
+
+#ifdef RBBI_DEBUG
+        static void printNodeHeader();
+        static void printNode(const RBBINode *n);
+        static void printTree(const RBBINode *n, UBool withHeading);
+#endif
+
+    private:
+        RBBINode &operator = (const RBBINode &other); // No defs.
+        UBool operator == (const RBBINode &other);    // Private, so these functions won't accidently be used.
+
+#ifdef RBBI_DEBUG
+    public:
+        int           fSerialNum;           //  Debugging aids.
+#endif
+};
+
+#ifdef RBBI_DEBUG
+U_CFUNC void 
+RBBI_DEBUG_printUnicodeString(const UnicodeString &s, int minWidth=0);
+#endif
+
+U_NAMESPACE_END
+
+#endif
+
diff --git a/thirdparty/icu4c/common/rbbirb.cpp b/thirdparty/icu4c/common/rbbirb.cpp
new file mode 100644
index 0000000000..e5c250dfe4
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbirb.cpp
@@ -0,0 +1,361 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  file:  rbbirb.cpp
+//
+//  Copyright (C) 2002-2011, International Business Machines Corporation and others.
+//  All Rights Reserved.
+//
+//  This file contains the RBBIRuleBuilder class implementation.  This is the main class for
+//    building (compiling) break rules into the tables required by the runtime
+//    RBBI engine.
+//
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/brkiter.h"
+#include "unicode/rbbi.h"
+#include "unicode/ubrk.h"
+#include "unicode/unistr.h"
+#include "unicode/uniset.h"
+#include "unicode/uchar.h"
+#include "unicode/uchriter.h"
+#include "unicode/ustring.h"
+#include "unicode/parsepos.h"
+#include "unicode/parseerr.h"
+
+#include "cmemory.h"
+#include "cstring.h"
+#include "rbbirb.h"
+#include "rbbinode.h"
+#include "rbbiscan.h"
+#include "rbbisetb.h"
+#include "rbbitblb.h"
+#include "rbbidata.h"
+#include "uassert.h"
+
+
+U_NAMESPACE_BEGIN
+
+
+//----------------------------------------------------------------------------------------
+//
+//  Constructor.
+//
+//----------------------------------------------------------------------------------------
+RBBIRuleBuilder::RBBIRuleBuilder(const UnicodeString   &rules,
+                                       UParseError     *parseErr,
+                                       UErrorCode      &status)
+ : fRules(rules), fStrippedRules(rules)
+{
+    fStatus = &status; // status is checked below
+    fParseError = parseErr;
+    fDebugEnv   = NULL;
+#ifdef RBBI_DEBUG
+    fDebugEnv   = getenv("U_RBBIDEBUG");
+#endif
+
+
+    fForwardTree        = NULL;
+    fReverseTree        = NULL;
+    fSafeFwdTree        = NULL;
+    fSafeRevTree        = NULL;
+    fDefaultTree        = &fForwardTree;
+    fForwardTable       = NULL;
+    fRuleStatusVals     = NULL;
+    fChainRules         = FALSE;
+    fLBCMNoChain        = FALSE;
+    fLookAheadHardBreak = FALSE;
+    fUSetNodes          = NULL;
+    fRuleStatusVals     = NULL;
+    fScanner            = NULL;
+    fSetBuilder         = NULL;
+    if (parseErr) {
+        uprv_memset(parseErr, 0, sizeof(UParseError));
+    }
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    fUSetNodes          = new UVector(status); // bcos status gets overwritten here
+    fRuleStatusVals     = new UVector(status);
+    fScanner            = new RBBIRuleScanner(this);
+    fSetBuilder         = new RBBISetBuilder(this);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if(fSetBuilder == 0 || fScanner == 0 || fUSetNodes == 0 || fRuleStatusVals == 0) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+
+
+//----------------------------------------------------------------------------------------
+//
+//  Destructor
+//
+//----------------------------------------------------------------------------------------
+RBBIRuleBuilder::~RBBIRuleBuilder() {
+
+    int        i;
+    for (i=0; ; i++) {
+        RBBINode *n = (RBBINode *)fUSetNodes->elementAt(i);
+        if (n==NULL) {
+            break;
+        }
+        delete n;
+    }
+
+    delete fUSetNodes;
+    delete fSetBuilder;
+    delete fForwardTable;
+    delete fForwardTree;
+    delete fReverseTree;
+    delete fSafeFwdTree;
+    delete fSafeRevTree;
+    delete fScanner;
+    delete fRuleStatusVals;
+}
+
+
+
+
+
+//----------------------------------------------------------------------------------------
+//
+//   flattenData() -  Collect up the compiled RBBI rule data and put it into
+//                    the format for saving in ICU data files,
+//                    which is also the format needed by the RBBI runtime engine.
+//
+//----------------------------------------------------------------------------------------
+static int32_t align8(int32_t i) {return (i+7) & 0xfffffff8;}
+
+RBBIDataHeader *RBBIRuleBuilder::flattenData() {
+    int32_t    i;
+
+    if (U_FAILURE(*fStatus)) {
+        return NULL;
+    }
+
+    // Remove whitespace from the rules to make it smaller.
+    // The rule parser has already removed comments.
+    fStrippedRules = fScanner->stripRules(fStrippedRules);
+
+    // Calculate the size of each section in the data.
+    //   Sizes here are padded up to a multiple of 8 for better memory alignment.
+    //   Sections sizes actually stored in the header are for the actual data
+    //     without the padding.
+    //
+    int32_t headerSize        = align8(sizeof(RBBIDataHeader));
+    int32_t forwardTableSize  = align8(fForwardTable->getTableSize());
+    int32_t reverseTableSize  = align8(fForwardTable->getSafeTableSize());
+    int32_t trieSize          = align8(fSetBuilder->getTrieSize());
+    int32_t statusTableSize   = align8(fRuleStatusVals->size() * sizeof(int32_t));
+
+    int32_t rulesLengthInUTF8 = 0;
+    u_strToUTF8WithSub(0, 0, &rulesLengthInUTF8,
+                       fStrippedRules.getBuffer(), fStrippedRules.length(),
+                       0xfffd, nullptr, fStatus);
+    *fStatus = U_ZERO_ERROR;
+
+    int32_t rulesSize         = align8((rulesLengthInUTF8+1));
+
+    int32_t         totalSize = headerSize
+                                + forwardTableSize
+                                + reverseTableSize
+                                + statusTableSize + trieSize + rulesSize;
+
+#ifdef RBBI_DEBUG
+    if (fDebugEnv && uprv_strstr(fDebugEnv, "size")) {
+        RBBIDebugPrintf("Header Size:        %8d\n", headerSize);
+        RBBIDebugPrintf("Forward Table Size: %8d\n", forwardTableSize);
+        RBBIDebugPrintf("Reverse Table Size: %8d\n", reverseTableSize);
+        RBBIDebugPrintf("Trie Size:          %8d\n", trieSize);
+        RBBIDebugPrintf("Status Table Size:  %8d\n", statusTableSize);
+        RBBIDebugPrintf("Rules Size:         %8d\n", rulesSize);
+        RBBIDebugPrintf("-----------------------------\n");
+        RBBIDebugPrintf("Total Size:         %8d\n", totalSize);
+    }
+#endif
+
+    RBBIDataHeader  *data     = (RBBIDataHeader *)uprv_malloc(totalSize);
+    if (data == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memset(data, 0, totalSize);
+
+
+    data->fMagic            = 0xb1a0;
+    data->fFormatVersion[0] = RBBI_DATA_FORMAT_VERSION[0];
+    data->fFormatVersion[1] = RBBI_DATA_FORMAT_VERSION[1];
+    data->fFormatVersion[2] = RBBI_DATA_FORMAT_VERSION[2];
+    data->fFormatVersion[3] = RBBI_DATA_FORMAT_VERSION[3];
+    data->fLength           = totalSize;
+    data->fCatCount         = fSetBuilder->getNumCharCategories();
+
+    data->fFTable        = headerSize;
+    data->fFTableLen     = forwardTableSize;
+
+    data->fRTable        = data->fFTable  + data->fFTableLen;
+    data->fRTableLen     = reverseTableSize;
+
+    data->fTrie          = data->fRTable + data->fRTableLen;
+    data->fTrieLen       = trieSize;
+    data->fStatusTable   = data->fTrie    + data->fTrieLen;
+    data->fStatusTableLen= statusTableSize;
+    data->fRuleSource    = data->fStatusTable + statusTableSize;
+    data->fRuleSourceLen = rulesLengthInUTF8;
+
+    uprv_memset(data->fReserved, 0, sizeof(data->fReserved));
+
+    fForwardTable->exportTable((uint8_t *)data + data->fFTable);
+    fForwardTable->exportSafeTable((uint8_t *)data + data->fRTable);
+    fSetBuilder->serializeTrie ((uint8_t *)data + data->fTrie);
+
+    int32_t *ruleStatusTable = (int32_t *)((uint8_t *)data + data->fStatusTable);
+    for (i=0; i<fRuleStatusVals->size(); i++) {
+        ruleStatusTable[i] = fRuleStatusVals->elementAti(i);
+    }
+
+    u_strToUTF8WithSub((char *)data+data->fRuleSource, rulesSize, &rulesLengthInUTF8,
+                       fStrippedRules.getBuffer(), fStrippedRules.length(),
+                       0xfffd, nullptr, fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return NULL;
+    }
+
+    return data;
+}
+
+
+//----------------------------------------------------------------------------------------
+//
+//  createRuleBasedBreakIterator    construct from source rules that are passed in
+//                                  in a UnicodeString
+//
+//----------------------------------------------------------------------------------------
+BreakIterator *
+RBBIRuleBuilder::createRuleBasedBreakIterator( const UnicodeString    &rules,
+                                    UParseError      *parseError,
+                                    UErrorCode       &status)
+{
+    //
+    // Read the input rules, generate a parse tree, symbol table,
+    // and list of all Unicode Sets referenced by the rules.
+    //
+    RBBIRuleBuilder  builder(rules, parseError, status);
+    if (U_FAILURE(status)) { // status checked here bcos build below doesn't
+        return NULL;
+    }
+
+    RBBIDataHeader *data = builder.build(status);
+
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+
+    //
+    //  Create a break iterator from the compiled rules.
+    //     (Identical to creation from stored pre-compiled rules)
+    //
+    // status is checked after init in construction.
+    RuleBasedBreakIterator *This = new RuleBasedBreakIterator(data, status);
+    if (U_FAILURE(status)) {
+        delete This;
+        This = NULL;
+    } 
+    else if(This == NULL) { // test for NULL
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return This;
+}
+
+RBBIDataHeader *RBBIRuleBuilder::build(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+
+    fScanner->parse();
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+
+    //
+    // UnicodeSet processing.
+    //    Munge the Unicode Sets to create an initial set of character categories.
+    //
+    fSetBuilder->buildRanges();
+
+    //
+    //   Generate the DFA state transition table.
+    //
+    fForwardTable = new RBBITableBuilder(this, &fForwardTree, status);
+    if (fForwardTable == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+
+    fForwardTable->buildForwardTable();
+
+    // State table and character category optimization.
+    // Merge equivalent rows and columns.
+    // Note that this process alters the initial set of character categories,
+    // causing the representation of UnicodeSets in the parse tree to become invalid.
+
+    optimizeTables();
+    fForwardTable->buildSafeReverseTable(status);
+
+
+#ifdef RBBI_DEBUG
+    if (fDebugEnv && uprv_strstr(fDebugEnv, "states")) {
+        fForwardTable->printStates();
+        fForwardTable->printRuleStatusTable();
+        fForwardTable->printReverseTable();
+    }
+#endif
+
+    //    Generate the mapping tables (TRIE) from input code points to
+    //    the character categories.
+    //
+    fSetBuilder->buildTrie();
+
+    //
+    //   Package up the compiled data into a memory image
+    //      in the run-time format.
+    //
+    RBBIDataHeader *data = flattenData(); // returns NULL if error
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+    return data;
+}
+
+void RBBIRuleBuilder::optimizeTables() {
+    bool didSomething;
+    do {
+        didSomething = false;
+
+        // Begin looking for duplicates with char class 3.
+        // Classes 0, 1 and 2 are special; they are unused, {bof} and {eof} respectively,
+        // and should not have other categories merged into them.
+        IntPair duplPair = {3, 0};
+        while (fForwardTable->findDuplCharClassFrom(&duplPair)) {
+            fSetBuilder->mergeCategories(duplPair);
+            fForwardTable->removeColumn(duplPair.second);
+            didSomething = true;
+        }
+
+        while (fForwardTable->removeDuplicateStates() > 0) {
+            didSomething = true;
+        }
+    } while (didSomething);
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbirb.h b/thirdparty/icu4c/common/rbbirb.h
new file mode 100644
index 0000000000..037c1dc2ce
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbirb.h
@@ -0,0 +1,237 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  rbbirb.h
+//
+//  Copyright (C) 2002-2008, International Business Machines Corporation and others.
+//  All Rights Reserved.
+//
+//  This file contains declarations for several classes from the
+//    Rule Based Break Iterator rule builder.
+//
+
+
+#ifndef RBBIRB_H
+#define RBBIRB_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include <utility>
+
+#include "unicode/uobject.h"
+#include "unicode/rbbi.h"
+#include "unicode/uniset.h"
+#include "unicode/parseerr.h"
+#include "uhash.h"
+#include "uvector.h"
+#include "unicode/symtable.h"// For UnicodeSet parsing, is the interface that
+                             //    looks up references to $variables within a set.
+
+
+U_NAMESPACE_BEGIN
+
+class               RBBIRuleScanner;
+struct              RBBIRuleTableEl;
+class               RBBISetBuilder;
+class               RBBINode;
+class               RBBITableBuilder;
+
+
+
+//--------------------------------------------------------------------------------
+//
+//   RBBISymbolTable.    Implements SymbolTable interface that is used by the
+//                       UnicodeSet parser to resolve references to $variables.
+//
+//--------------------------------------------------------------------------------
+class RBBISymbolTableEntry : public UMemory { // The symbol table hash table contains one
+public:                                       //   of these structs for each entry.
+    RBBISymbolTableEntry();
+    UnicodeString          key;
+    RBBINode               *val;
+    ~RBBISymbolTableEntry();
+
+private:
+    RBBISymbolTableEntry(const RBBISymbolTableEntry &other); // forbid copying of this class
+    RBBISymbolTableEntry &operator=(const RBBISymbolTableEntry &other); // forbid copying of this class
+};
+
+
+class RBBISymbolTable : public UMemory, public SymbolTable {
+private:
+    const UnicodeString      &fRules;
+    UHashtable               *fHashTable;
+    RBBIRuleScanner          *fRuleScanner;
+
+    // These next two fields are part of the mechanism for passing references to
+    //   already-constructed UnicodeSets back to the UnicodeSet constructor
+    //   when the pattern includes $variable references.
+    const UnicodeString      ffffString;      // = "/uffff"
+    UnicodeSet              *fCachedSetLookup;
+
+public:
+    //  API inherited from class SymbolTable
+    virtual const UnicodeString*  lookup(const UnicodeString& s) const;
+    virtual const UnicodeFunctor* lookupMatcher(UChar32 ch) const;
+    virtual UnicodeString parseReference(const UnicodeString& text,
+                                         ParsePosition& pos, int32_t limit) const;
+
+    //  Additional Functions
+    RBBISymbolTable(RBBIRuleScanner *, const UnicodeString &fRules, UErrorCode &status);
+    virtual ~RBBISymbolTable();
+
+    virtual RBBINode *lookupNode(const UnicodeString &key) const;
+    virtual void      addEntry  (const UnicodeString &key, RBBINode *val, UErrorCode &err);
+
+#ifdef RBBI_DEBUG
+    virtual void      rbbiSymtablePrint() const;
+#else
+    // A do-nothing inline function for non-debug builds.  Member funcs can't be empty
+    //  or the call sites won't compile.
+    int32_t fFakeField;
+    #define rbbiSymtablePrint() fFakeField=0; 
+#endif
+
+private:
+    RBBISymbolTable(const RBBISymbolTable &other); // forbid copying of this class
+    RBBISymbolTable &operator=(const RBBISymbolTable &other); // forbid copying of this class
+};
+
+
+//--------------------------------------------------------------------------------
+//
+//  class RBBIRuleBuilder       The top-level class handling RBBI rule compiling.
+//
+//--------------------------------------------------------------------------------
+class RBBIRuleBuilder : public UMemory {
+public:
+
+    //  Create a rule based break iterator from a set of rules.
+    //  This function is the main entry point into the rule builder.  The
+    //   public ICU API for creating RBBIs uses this function to do the actual work.
+    //
+    static BreakIterator * createRuleBasedBreakIterator( const UnicodeString    &rules,
+                                    UParseError      *parseError,
+                                    UErrorCode       &status);
+
+public:
+    // The "public" functions and data members that appear below are accessed
+    //  (and shared) by the various parts that make up the rule builder.  They
+    //  are NOT intended to be accessed by anything outside of the
+    //  rule builder implementation.
+    RBBIRuleBuilder(const UnicodeString  &rules,
+                    UParseError          *parseErr,
+                    UErrorCode           &status
+    );
+
+    virtual    ~RBBIRuleBuilder();
+
+    /**
+     *  Build the state tables and char class Trie from the source rules.
+     */
+    RBBIDataHeader  *build(UErrorCode &status);
+
+
+    /**
+     * Fold together redundant character classes (table columns) and
+     * redundant states (table rows). Done after initial table generation,
+     * before serializing the result.
+     */
+    void optimizeTables();
+
+    char                          *fDebugEnv;        // controls debug trace output
+    UErrorCode                    *fStatus;          // Error reporting.  Keeping status
+    UParseError                   *fParseError;      //   here avoids passing it everywhere.
+    const UnicodeString           &fRules;           // The rule string that we are compiling
+    UnicodeString                 fStrippedRules;    // The rule string, with comments stripped.
+
+    RBBIRuleScanner               *fScanner;         // The scanner.
+    RBBINode                      *fForwardTree;     // The parse trees, generated by the scanner,
+    RBBINode                      *fReverseTree;     //   then manipulated by subsequent steps.
+    RBBINode                      *fSafeFwdTree;
+    RBBINode                      *fSafeRevTree;
+
+    RBBINode                      **fDefaultTree;    // For rules not qualified with a !
+                                                     //   the tree to which they belong to.
+
+    UBool                         fChainRules;       // True for chained Unicode TR style rules.
+                                                     // False for traditional regexp rules.
+
+    UBool                         fLBCMNoChain;      // True:  suppress chaining of rules on
+                                                     //   chars with LineBreak property == CM.
+
+    UBool                         fLookAheadHardBreak;  // True:  Look ahead matches cause an
+                                                     // immediate break, no continuing for the
+                                                     // longest match.
+
+    RBBISetBuilder                *fSetBuilder;      // Set and Character Category builder.
+    UVector                       *fUSetNodes;       // Vector of all uset nodes.
+
+    RBBITableBuilder              *fForwardTable;    // State transition table, build time form.
+
+    UVector                       *fRuleStatusVals;  // The values that can be returned
+                                                     //   from getRuleStatus().
+
+    RBBIDataHeader                *flattenData();    // Create the flattened (runtime format)
+                                                     // data tables..
+private:
+    RBBIRuleBuilder(const RBBIRuleBuilder &other); // forbid copying of this class
+    RBBIRuleBuilder &operator=(const RBBIRuleBuilder &other); // forbid copying of this class
+};
+
+
+
+
+//----------------------------------------------------------------------------
+//
+//   RBBISetTableEl   is an entry in the hash table of UnicodeSets that have
+//                    been encountered.  The val Node will be of nodetype uset
+//                    and contain pointers to the actual UnicodeSets.
+//                    The Key is the source string for initializing the set.
+//
+//                    The hash table is used to avoid creating duplicate
+//                    unnamed (not $var references) UnicodeSets.
+//
+//                    Memory Management:
+//                       The Hash Table owns these RBBISetTableEl structs and
+//                            the key strings.  It does NOT own the val nodes.
+//
+//----------------------------------------------------------------------------
+struct RBBISetTableEl {
+    UnicodeString *key;
+    RBBINode      *val;
+};
+
+/**
+ *   A pair of ints, used to bundle pairs of states or pairs of character classes.
+ */
+typedef std::pair<int32_t, int32_t> IntPair;
+
+
+//----------------------------------------------------------------------------
+//
+//   RBBIDebugPrintf    Printf equivalent, for debugging output.
+//                      Conditional compilation of the implementation lets us
+//                      get rid of the stdio dependency in environments where it
+//                      is unavailable.
+//
+//----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+#include <stdio.h>
+#define RBBIDebugPrintf printf
+#define RBBIDebugPuts puts
+#else
+#undef RBBIDebugPrintf 
+#define RBBIDebugPuts(arg)
+#endif
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif
+
+
+
diff --git a/thirdparty/icu4c/common/rbbirpt.h b/thirdparty/icu4c/common/rbbirpt.h
new file mode 100644
index 0000000000..586953c90c
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbirpt.h
@@ -0,0 +1,296 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//---------------------------------------------------------------------------------
+//
+// Generated Header File.  Do not edit by hand.
+//    This file contains the state table for the ICU Rule Based Break Iterator
+//    rule parser.
+//    It is generated by the Perl script "rbbicst.pl" from
+//    the rule parser state definitions file "rbbirpt.txt".
+//
+//   Copyright (C) 2002-2016 International Business Machines Corporation 
+//   and others. All rights reserved.  
+//
+//---------------------------------------------------------------------------------
+#ifndef RBBIRPT_H
+#define RBBIRPT_H
+
+#include "unicode/utypes.h"
+
+U_NAMESPACE_BEGIN
+//
+// Character classes for RBBI rule scanning.
+//
+    static const uint8_t kRuleSet_digit_char = 128;
+    static const uint8_t kRuleSet_name_char = 129;
+    static const uint8_t kRuleSet_name_start_char = 130;
+    static const uint8_t kRuleSet_rule_char = 131;
+    static const uint8_t kRuleSet_white_space = 132;
+
+
+enum RBBI_RuleParseAction {
+    doCheckVarDef,
+    doDotAny,
+    doEndAssign,
+    doEndOfRule,
+    doEndVariableName,
+    doExit,
+    doExprCatOperator,
+    doExprFinished,
+    doExprOrOperator,
+    doExprRParen,
+    doExprStart,
+    doLParen,
+    doNOP,
+    doNoChain,
+    doOptionEnd,
+    doOptionStart,
+    doReverseDir,
+    doRuleChar,
+    doRuleError,
+    doRuleErrorAssignExpr,
+    doScanUnicodeSet,
+    doSlash,
+    doStartAssign,
+    doStartTagValue,
+    doStartVariableName,
+    doTagDigit,
+    doTagExpectedError,
+    doTagValue,
+    doUnaryOpPlus,
+    doUnaryOpQuestion,
+    doUnaryOpStar,
+    doVariableNameExpectedErr,
+    rbbiLastAction};
+
+//-------------------------------------------------------------------------------
+//
+//  RBBIRuleTableEl    represents the structure of a row in the transition table
+//                     for the rule parser state machine.
+//-------------------------------------------------------------------------------
+struct RBBIRuleTableEl {
+    RBBI_RuleParseAction          fAction;
+    uint8_t                       fCharClass;       // 0-127:    an individual ASCII character
+                                                    // 128-255:  character class index
+    uint8_t                       fNextState;       // 0-250:    normal next-stat numbers
+                                                    // 255:      pop next-state from stack.
+    uint8_t                       fPushState;
+    UBool                         fNextChar;
+};
+
+static const struct RBBIRuleTableEl gRuleParseStateTable[] = {
+    {doNOP, 0, 0, 0, TRUE}
+    , {doExprStart, 254, 29, 9, FALSE}     //  1      start
+    , {doNOP, 132, 1,0,  TRUE}     //  2 
+    , {doNoChain, 94 /* ^ */, 12, 9, TRUE}     //  3 
+    , {doExprStart, 36 /* $ */, 88, 98, FALSE}     //  4 
+    , {doNOP, 33 /* ! */, 19,0,  TRUE}     //  5 
+    , {doNOP, 59 /* ; */, 1,0,  TRUE}     //  6 
+    , {doNOP, 252, 0,0,  FALSE}     //  7 
+    , {doExprStart, 255, 29, 9, FALSE}     //  8 
+    , {doEndOfRule, 59 /* ; */, 1,0,  TRUE}     //  9      break-rule-end
+    , {doNOP, 132, 9,0,  TRUE}     //  10 
+    , {doRuleError, 255, 103,0,  FALSE}     //  11 
+    , {doExprStart, 254, 29,0,  FALSE}     //  12      start-after-caret
+    , {doNOP, 132, 12,0,  TRUE}     //  13 
+    , {doRuleError, 94 /* ^ */, 103,0,  FALSE}     //  14 
+    , {doExprStart, 36 /* $ */, 88, 37, FALSE}     //  15 
+    , {doRuleError, 59 /* ; */, 103,0,  FALSE}     //  16 
+    , {doRuleError, 252, 103,0,  FALSE}     //  17 
+    , {doExprStart, 255, 29,0,  FALSE}     //  18 
+    , {doNOP, 33 /* ! */, 21,0,  TRUE}     //  19      rev-option
+    , {doReverseDir, 255, 28, 9, FALSE}     //  20 
+    , {doOptionStart, 130, 23,0,  TRUE}     //  21      option-scan1
+    , {doRuleError, 255, 103,0,  FALSE}     //  22 
+    , {doNOP, 129, 23,0,  TRUE}     //  23      option-scan2
+    , {doOptionEnd, 255, 25,0,  FALSE}     //  24 
+    , {doNOP, 59 /* ; */, 1,0,  TRUE}     //  25      option-scan3
+    , {doNOP, 132, 25,0,  TRUE}     //  26 
+    , {doRuleError, 255, 103,0,  FALSE}     //  27 
+    , {doExprStart, 255, 29, 9, FALSE}     //  28      reverse-rule
+    , {doRuleChar, 254, 38,0,  TRUE}     //  29      term
+    , {doNOP, 132, 29,0,  TRUE}     //  30 
+    , {doRuleChar, 131, 38,0,  TRUE}     //  31 
+    , {doNOP, 91 /* [ */, 94, 38, FALSE}     //  32 
+    , {doLParen, 40 /* ( */, 29, 38, TRUE}     //  33 
+    , {doNOP, 36 /* $ */, 88, 37, FALSE}     //  34 
+    , {doDotAny, 46 /* . */, 38,0,  TRUE}     //  35 
+    , {doRuleError, 255, 103,0,  FALSE}     //  36 
+    , {doCheckVarDef, 255, 38,0,  FALSE}     //  37      term-var-ref
+    , {doNOP, 132, 38,0,  TRUE}     //  38      expr-mod
+    , {doUnaryOpStar, 42 /* * */, 43,0,  TRUE}     //  39 
+    , {doUnaryOpPlus, 43 /* + */, 43,0,  TRUE}     //  40 
+    , {doUnaryOpQuestion, 63 /* ? */, 43,0,  TRUE}     //  41 
+    , {doNOP, 255, 43,0,  FALSE}     //  42 
+    , {doExprCatOperator, 254, 29,0,  FALSE}     //  43      expr-cont
+    , {doNOP, 132, 43,0,  TRUE}     //  44 
+    , {doExprCatOperator, 131, 29,0,  FALSE}     //  45 
+    , {doExprCatOperator, 91 /* [ */, 29,0,  FALSE}     //  46 
+    , {doExprCatOperator, 40 /* ( */, 29,0,  FALSE}     //  47 
+    , {doExprCatOperator, 36 /* $ */, 29,0,  FALSE}     //  48 
+    , {doExprCatOperator, 46 /* . */, 29,0,  FALSE}     //  49 
+    , {doExprCatOperator, 47 /* / */, 55,0,  FALSE}     //  50 
+    , {doExprCatOperator, 123 /* { */, 67,0,  TRUE}     //  51 
+    , {doExprOrOperator, 124 /* | */, 29,0,  TRUE}     //  52 
+    , {doExprRParen, 41 /* ) */, 255,0,  TRUE}     //  53 
+    , {doExprFinished, 255, 255,0,  FALSE}     //  54 
+    , {doSlash, 47 /* / */, 57,0,  TRUE}     //  55      look-ahead
+    , {doNOP, 255, 103,0,  FALSE}     //  56 
+    , {doExprCatOperator, 254, 29,0,  FALSE}     //  57      expr-cont-no-slash
+    , {doNOP, 132, 43,0,  TRUE}     //  58 
+    , {doExprCatOperator, 131, 29,0,  FALSE}     //  59 
+    , {doExprCatOperator, 91 /* [ */, 29,0,  FALSE}     //  60 
+    , {doExprCatOperator, 40 /* ( */, 29,0,  FALSE}     //  61 
+    , {doExprCatOperator, 36 /* $ */, 29,0,  FALSE}     //  62 
+    , {doExprCatOperator, 46 /* . */, 29,0,  FALSE}     //  63 
+    , {doExprOrOperator, 124 /* | */, 29,0,  TRUE}     //  64 
+    , {doExprRParen, 41 /* ) */, 255,0,  TRUE}     //  65 
+    , {doExprFinished, 255, 255,0,  FALSE}     //  66 
+    , {doNOP, 132, 67,0,  TRUE}     //  67      tag-open
+    , {doStartTagValue, 128, 70,0,  FALSE}     //  68 
+    , {doTagExpectedError, 255, 103,0,  FALSE}     //  69 
+    , {doNOP, 132, 74,0,  TRUE}     //  70      tag-value
+    , {doNOP, 125 /* } */, 74,0,  FALSE}     //  71 
+    , {doTagDigit, 128, 70,0,  TRUE}     //  72 
+    , {doTagExpectedError, 255, 103,0,  FALSE}     //  73 
+    , {doNOP, 132, 74,0,  TRUE}     //  74      tag-close
+    , {doTagValue, 125 /* } */, 77,0,  TRUE}     //  75 
+    , {doTagExpectedError, 255, 103,0,  FALSE}     //  76 
+    , {doExprCatOperator, 254, 29,0,  FALSE}     //  77      expr-cont-no-tag
+    , {doNOP, 132, 77,0,  TRUE}     //  78 
+    , {doExprCatOperator, 131, 29,0,  FALSE}     //  79 
+    , {doExprCatOperator, 91 /* [ */, 29,0,  FALSE}     //  80 
+    , {doExprCatOperator, 40 /* ( */, 29,0,  FALSE}     //  81 
+    , {doExprCatOperator, 36 /* $ */, 29,0,  FALSE}     //  82 
+    , {doExprCatOperator, 46 /* . */, 29,0,  FALSE}     //  83 
+    , {doExprCatOperator, 47 /* / */, 55,0,  FALSE}     //  84 
+    , {doExprOrOperator, 124 /* | */, 29,0,  TRUE}     //  85 
+    , {doExprRParen, 41 /* ) */, 255,0,  TRUE}     //  86 
+    , {doExprFinished, 255, 255,0,  FALSE}     //  87 
+    , {doStartVariableName, 36 /* $ */, 90,0,  TRUE}     //  88      scan-var-name
+    , {doNOP, 255, 103,0,  FALSE}     //  89 
+    , {doNOP, 130, 92,0,  TRUE}     //  90      scan-var-start
+    , {doVariableNameExpectedErr, 255, 103,0,  FALSE}     //  91 
+    , {doNOP, 129, 92,0,  TRUE}     //  92      scan-var-body
+    , {doEndVariableName, 255, 255,0,  FALSE}     //  93 
+    , {doScanUnicodeSet, 91 /* [ */, 255,0,  TRUE}     //  94      scan-unicode-set
+    , {doScanUnicodeSet, 112 /* p */, 255,0,  TRUE}     //  95 
+    , {doScanUnicodeSet, 80 /* P */, 255,0,  TRUE}     //  96 
+    , {doNOP, 255, 103,0,  FALSE}     //  97 
+    , {doNOP, 132, 98,0,  TRUE}     //  98      assign-or-rule
+    , {doStartAssign, 61 /* = */, 29, 101, TRUE}     //  99 
+    , {doNOP, 255, 37, 9, FALSE}     //  100 
+    , {doEndAssign, 59 /* ; */, 1,0,  TRUE}     //  101      assign-end
+    , {doRuleErrorAssignExpr, 255, 103,0,  FALSE}     //  102 
+    , {doExit, 255, 103,0,  TRUE}     //  103      errorDeath
+ };
+#ifdef RBBI_DEBUG
+static const char * const RBBIRuleStateNames[] = {    0,
+     "start",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "break-rule-end",
+    0,
+    0,
+     "start-after-caret",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "rev-option",
+    0,
+     "option-scan1",
+    0,
+     "option-scan2",
+    0,
+     "option-scan3",
+    0,
+    0,
+     "reverse-rule",
+     "term",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "term-var-ref",
+     "expr-mod",
+    0,
+    0,
+    0,
+    0,
+     "expr-cont",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "look-ahead",
+    0,
+     "expr-cont-no-slash",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "tag-open",
+    0,
+    0,
+     "tag-value",
+    0,
+    0,
+    0,
+     "tag-close",
+    0,
+    0,
+     "expr-cont-no-tag",
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+     "scan-var-name",
+    0,
+     "scan-var-start",
+    0,
+     "scan-var-body",
+    0,
+     "scan-unicode-set",
+    0,
+    0,
+    0,
+     "assign-or-rule",
+    0,
+    0,
+     "assign-end",
+    0,
+     "errorDeath",
+    0};
+#endif
+
+U_NAMESPACE_END
+#endif
diff --git a/thirdparty/icu4c/common/rbbiscan.cpp b/thirdparty/icu4c/common/rbbiscan.cpp
new file mode 100644
index 0000000000..9c406af671
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbiscan.cpp
@@ -0,0 +1,1281 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  file:  rbbiscan.cpp
+//
+//  Copyright (C) 2002-2016, International Business Machines Corporation and others.
+//  All Rights Reserved.
+//
+//  This file contains the Rule Based Break Iterator Rule Builder functions for
+//   scanning the rules and assembling a parse tree.  This is the first phase
+//   of compiling the rules.
+//
+//  The overall of the rules is managed by class RBBIRuleBuilder, which will
+//  create and use an instance of this class as part of the process.
+//
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/unistr.h"
+#include "unicode/uniset.h"
+#include "unicode/uchar.h"
+#include "unicode/uchriter.h"
+#include "unicode/parsepos.h"
+#include "unicode/parseerr.h"
+#include "cmemory.h"
+#include "cstring.h"
+
+#include "rbbirpt.h"   // Contains state table for the rbbi rules parser.
+                       //   generated by a Perl script.
+#include "rbbirb.h"
+#include "rbbinode.h"
+#include "rbbiscan.h"
+#include "rbbitblb.h"
+
+#include "uassert.h"
+
+//------------------------------------------------------------------------------
+//
+// Unicode Set init strings for each of the character classes needed for parsing a rule file.
+//               (Initialized with hex values for portability to EBCDIC based machines.
+//                Really ugly, but there's no good way to avoid it.)
+//
+//              The sets are referred to by name in the rbbirpt.txt, which is the
+//              source form of the state transition table for the RBBI rule parser.
+//
+//------------------------------------------------------------------------------
+static const UChar gRuleSet_rule_char_pattern[]       = {
+ // Characters that may appear as literals in patterns without escaping or quoting.
+ //   [    ^      [    \     p     {      Z     }     \     u    0      0    2      0
+    0x5b, 0x5e, 0x5b, 0x5c, 0x70, 0x7b, 0x5a, 0x7d, 0x5c, 0x75, 0x30, 0x30, 0x32, 0x30,
+ //   -    \      u    0     0     7      f     ]     -     [    \      p
+    0x2d, 0x5c, 0x75, 0x30, 0x30, 0x37, 0x66, 0x5d, 0x2d, 0x5b, 0x5c, 0x70,
+ //   {     L     }    ]     -     [      \     p     {     N    }      ]     ]
+    0x7b, 0x4c, 0x7d, 0x5d, 0x2d, 0x5b, 0x5c, 0x70, 0x7b, 0x4e, 0x7d, 0x5d, 0x5d, 0};
+
+static const UChar gRuleSet_name_char_pattern[]       = {
+//    [    _      \    p     {     L      }     \     p     {    N      }     ]
+    0x5b, 0x5f, 0x5c, 0x70, 0x7b, 0x4c, 0x7d, 0x5c, 0x70, 0x7b, 0x4e, 0x7d, 0x5d, 0};
+
+static const UChar gRuleSet_digit_char_pattern[] = {
+//    [    0      -    9     ]
+    0x5b, 0x30, 0x2d, 0x39, 0x5d, 0};
+
+static const UChar gRuleSet_name_start_char_pattern[] = {
+//    [    _      \    p     {     L      }     ]
+    0x5b, 0x5f, 0x5c, 0x70, 0x7b, 0x4c, 0x7d, 0x5d, 0 };
+
+static const UChar kAny[] = {0x61, 0x6e, 0x79, 0x00};  // "any"
+
+
+U_CDECL_BEGIN
+static void U_CALLCONV RBBISetTable_deleter(void *p) {
+    icu::RBBISetTableEl *px = (icu::RBBISetTableEl *)p;
+    delete px->key;
+    // Note:  px->val is owned by the linked list "fSetsListHead" in scanner.
+    //        Don't delete the value nodes here.
+    uprv_free(px);
+}
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+//------------------------------------------------------------------------------
+//
+//  Constructor.
+//
+//------------------------------------------------------------------------------
+RBBIRuleScanner::RBBIRuleScanner(RBBIRuleBuilder *rb)
+{
+    fRB                 = rb;
+    fScanIndex          = 0;
+    fNextIndex          = 0;
+    fQuoteMode          = FALSE;
+    fLineNum            = 1;
+    fCharNum            = 0;
+    fLastChar           = 0;
+    
+    fStateTable         = NULL;
+    fStack[0]           = 0;
+    fStackPtr           = 0;
+    fNodeStack[0]       = NULL;
+    fNodeStackPtr       = 0;
+
+    fReverseRule        = FALSE;
+    fLookAheadRule      = FALSE;
+    fNoChainInRule      = FALSE;
+
+    fSymbolTable        = NULL;
+    fSetTable           = NULL;
+    fRuleNum            = 0;
+    fOptionStart        = 0;
+
+    // Do not check status until after all critical fields are sufficiently initialized
+    //   that the destructor can run cleanly.
+    if (U_FAILURE(*rb->fStatus)) {
+        return;
+    }
+
+    //
+    //  Set up the constant Unicode Sets.
+    //     Note:  These could be made static, lazily initialized, and shared among
+    //            all instances of RBBIRuleScanners.  BUT this is quite a bit simpler,
+    //            and the time to build these few sets should be small compared to a
+    //            full break iterator build.
+    fRuleSets[kRuleSet_rule_char-128]
+        = UnicodeSet(UnicodeString(gRuleSet_rule_char_pattern),       *rb->fStatus);
+    // fRuleSets[kRuleSet_white_space-128] = [:Pattern_White_Space:]
+    fRuleSets[kRuleSet_white_space-128].
+        add(9, 0xd).add(0x20).add(0x85).add(0x200e, 0x200f).add(0x2028, 0x2029);
+    fRuleSets[kRuleSet_name_char-128]
+        = UnicodeSet(UnicodeString(gRuleSet_name_char_pattern),       *rb->fStatus);
+    fRuleSets[kRuleSet_name_start_char-128]
+        = UnicodeSet(UnicodeString(gRuleSet_name_start_char_pattern), *rb->fStatus);
+    fRuleSets[kRuleSet_digit_char-128]
+        = UnicodeSet(UnicodeString(gRuleSet_digit_char_pattern),      *rb->fStatus);
+    if (*rb->fStatus == U_ILLEGAL_ARGUMENT_ERROR) {
+        // This case happens if ICU's data is missing.  UnicodeSet tries to look up property
+        //   names from the init string, can't find them, and claims an illegal argument.
+        //   Change the error so that the actual problem will be clearer to users.
+        *rb->fStatus = U_BRK_INIT_ERROR;
+    }
+    if (U_FAILURE(*rb->fStatus)) {
+        return;
+    }
+
+    fSymbolTable = new RBBISymbolTable(this, rb->fRules, *rb->fStatus);
+    if (fSymbolTable == NULL) {
+        *rb->fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    fSetTable    = uhash_open(uhash_hashUnicodeString, uhash_compareUnicodeString, NULL, rb->fStatus);
+    if (U_FAILURE(*rb->fStatus)) {
+        return;
+    }
+    uhash_setValueDeleter(fSetTable, RBBISetTable_deleter);
+}
+
+
+
+//------------------------------------------------------------------------------
+//
+//  Destructor
+//
+//------------------------------------------------------------------------------
+RBBIRuleScanner::~RBBIRuleScanner() {
+    delete fSymbolTable;
+    if (fSetTable != NULL) {
+         uhash_close(fSetTable);
+         fSetTable = NULL;
+
+    }
+
+
+    // Node Stack.
+    //   Normally has one entry, which is the entire parse tree for the rules.
+    //   If errors occured, there may be additional subtrees left on the stack.
+    while (fNodeStackPtr > 0) {
+        delete fNodeStack[fNodeStackPtr];
+        fNodeStackPtr--;
+    }
+
+}
+
+//------------------------------------------------------------------------------
+//
+//  doParseAction        Do some action during rule parsing.
+//                       Called by the parse state machine.
+//                       Actions build the parse tree and Unicode Sets,
+//                       and maintain the parse stack for nested expressions.
+//
+//                       TODO:  unify EParseAction and RBBI_RuleParseAction enum types.
+//                              They represent exactly the same thing.  They're separate
+//                              only to work around enum forward declaration restrictions
+//                              in some compilers, while at the same time avoiding multiple
+//                              definitions problems.  I'm sure that there's a better way.
+//
+//------------------------------------------------------------------------------
+UBool RBBIRuleScanner::doParseActions(int32_t action)
+{
+    RBBINode *n       = NULL;
+
+    UBool   returnVal = TRUE;
+
+    switch (action) {
+
+    case doExprStart:
+        pushNewNode(RBBINode::opStart);
+        fRuleNum++;
+        break;
+
+
+    case doNoChain:
+        // Scanned a '^' while on the rule start state.
+        fNoChainInRule = TRUE;
+        break;
+
+
+    case doExprOrOperator:
+        {
+            fixOpStack(RBBINode::precOpCat);
+            RBBINode  *operandNode = fNodeStack[fNodeStackPtr--];
+            RBBINode  *orNode      = pushNewNode(RBBINode::opOr);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            orNode->fLeftChild     = operandNode;
+            operandNode->fParent   = orNode;
+        }
+        break;
+
+    case doExprCatOperator:
+        // concatenation operator.
+        // For the implicit concatenation of adjacent terms in an expression that are
+        //   not separated by any other operator.  Action is invoked between the
+        //   actions for the two terms.
+        {
+            fixOpStack(RBBINode::precOpCat);
+            RBBINode  *operandNode = fNodeStack[fNodeStackPtr--];
+            RBBINode  *catNode     = pushNewNode(RBBINode::opCat);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            catNode->fLeftChild    = operandNode;
+            operandNode->fParent   = catNode;
+        }
+        break;
+
+    case doLParen:
+        // Open Paren.
+        //   The openParen node is a dummy operation type with a low precedence,
+        //     which has the affect of ensuring that any real binary op that
+        //     follows within the parens binds more tightly to the operands than
+        //     stuff outside of the parens.
+        pushNewNode(RBBINode::opLParen);
+        break;
+
+    case doExprRParen:
+        fixOpStack(RBBINode::precLParen);
+        break;
+
+    case doNOP:
+        break;
+
+    case doStartAssign:
+        // We've just scanned "$variable = "
+        // The top of the node stack has the $variable ref node.
+
+        // Save the start position of the RHS text in the StartExpression node
+        //   that precedes the $variableReference node on the stack.
+        //   This will eventually be used when saving the full $variable replacement
+        //   text as a string.
+        n = fNodeStack[fNodeStackPtr-1];
+        n->fFirstPos = fNextIndex;              // move past the '='
+
+        // Push a new start-of-expression node; needed to keep parse of the
+        //   RHS expression happy.
+        pushNewNode(RBBINode::opStart);
+        break;
+
+
+
+
+    case doEndAssign:
+        {
+            // We have reached the end of an assignement statement.
+            //   Current scan char is the ';' that terminates the assignment.
+
+            // Terminate expression, leaves expression parse tree rooted in TOS node.
+            fixOpStack(RBBINode::precStart);
+
+            RBBINode *startExprNode  = fNodeStack[fNodeStackPtr-2];
+            RBBINode *varRefNode     = fNodeStack[fNodeStackPtr-1];
+            RBBINode *RHSExprNode    = fNodeStack[fNodeStackPtr];
+
+            // Save original text of right side of assignment, excluding the terminating ';'
+            //  in the root of the node for the right-hand-side expression.
+            RHSExprNode->fFirstPos = startExprNode->fFirstPos;
+            RHSExprNode->fLastPos  = fScanIndex;
+            fRB->fRules.extractBetween(RHSExprNode->fFirstPos, RHSExprNode->fLastPos, RHSExprNode->fText);
+
+            // Expression parse tree becomes l. child of the $variable reference node.
+            varRefNode->fLeftChild = RHSExprNode;
+            RHSExprNode->fParent   = varRefNode;
+
+            // Make a symbol table entry for the $variableRef node.
+            fSymbolTable->addEntry(varRefNode->fText, varRefNode, *fRB->fStatus);
+            if (U_FAILURE(*fRB->fStatus)) {
+                // This is a round-about way to get the parse position set
+                //  so that duplicate symbols error messages include a line number.
+                UErrorCode t = *fRB->fStatus;
+                *fRB->fStatus = U_ZERO_ERROR;
+                error(t);
+            }
+
+            // Clean up the stack.
+            delete startExprNode;
+            fNodeStackPtr-=3;
+            break;
+        }
+
+    case doEndOfRule:
+        {
+        fixOpStack(RBBINode::precStart);      // Terminate expression, leaves expression
+        if (U_FAILURE(*fRB->fStatus)) {       //   parse tree rooted in TOS node.
+            break;
+        }
+#ifdef RBBI_DEBUG
+        if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "rtree")) {printNodeStack("end of rule");}
+#endif
+        U_ASSERT(fNodeStackPtr == 1);
+        RBBINode *thisRule = fNodeStack[fNodeStackPtr];
+
+        // If this rule includes a look-ahead '/', add a endMark node to the
+        //   expression tree.
+        if (fLookAheadRule) {
+            RBBINode  *endNode        = pushNewNode(RBBINode::endMark);
+            RBBINode  *catNode        = pushNewNode(RBBINode::opCat);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            fNodeStackPtr -= 2;
+            catNode->fLeftChild       = thisRule;
+            catNode->fRightChild      = endNode;
+            fNodeStack[fNodeStackPtr] = catNode;
+            endNode->fVal             = fRuleNum;
+            endNode->fLookAheadEnd    = TRUE;
+            thisRule                  = catNode;
+
+            // TODO: Disable chaining out of look-ahead (hard break) rules.
+            //   The break on rule match is forced, so there is no point in building up
+            //   the state table to chain into another rule for a longer match.
+        }
+
+        // Mark this node as being the root of a rule.
+        thisRule->fRuleRoot = TRUE;
+
+        // Flag if chaining into this rule is wanted.
+        //    
+        if (fRB->fChainRules &&         // If rule chaining is enabled globally via !!chain
+                !fNoChainInRule) {      //     and no '^' chain-in inhibit was on this rule
+            thisRule->fChainIn = TRUE;
+        }
+
+
+        // All rule expressions are ORed together.
+        // The ';' that terminates an expression really just functions as a '|' with
+        //   a low operator prededence.
+        //
+        // Each of the four sets of rules are collected separately.
+        //  (forward, reverse, safe_forward, safe_reverse)
+        //  OR this rule into the appropriate group of them.
+        //
+        RBBINode **destRules = (fReverseRule? &fRB->fSafeRevTree : fRB->fDefaultTree);
+
+        if (*destRules != NULL) {
+            // This is not the first rule encounted.
+            // OR previous stuff  (from *destRules)
+            // with the current rule expression (on the Node Stack)
+            //  with the resulting OR expression going to *destRules
+            //
+                       thisRule    = fNodeStack[fNodeStackPtr];
+            RBBINode  *prevRules   = *destRules;
+            RBBINode  *orNode      = pushNewNode(RBBINode::opOr);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            orNode->fLeftChild     = prevRules;
+            prevRules->fParent     = orNode;
+            orNode->fRightChild    = thisRule;
+            thisRule->fParent      = orNode;
+            *destRules             = orNode;
+        }
+        else
+        {
+            // This is the first rule encountered (for this direction).
+            // Just move its parse tree from the stack to *destRules.
+            *destRules = fNodeStack[fNodeStackPtr];
+        }
+        fReverseRule   = FALSE;   // in preparation for the next rule.
+        fLookAheadRule = FALSE;
+        fNoChainInRule = FALSE;
+        fNodeStackPtr  = 0;
+        }
+        break;
+
+
+    case doRuleError:
+        error(U_BRK_RULE_SYNTAX);
+        returnVal = FALSE;
+        break;
+
+
+    case doVariableNameExpectedErr:
+        error(U_BRK_RULE_SYNTAX);
+        break;
+
+
+    //
+    //  Unary operands  + ? *
+    //    These all appear after the operand to which they apply.
+    //    When we hit one, the operand (may be a whole sub expression)
+    //    will be on the top of the stack.
+    //    Unary Operator becomes TOS, with the old TOS as its one child.
+    case doUnaryOpPlus:
+        {
+            RBBINode  *operandNode = fNodeStack[fNodeStackPtr--];
+            RBBINode  *plusNode    = pushNewNode(RBBINode::opPlus);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            plusNode->fLeftChild   = operandNode;
+            operandNode->fParent   = plusNode;
+        }
+        break;
+
+    case doUnaryOpQuestion:
+        {
+            RBBINode  *operandNode = fNodeStack[fNodeStackPtr--];
+            RBBINode  *qNode       = pushNewNode(RBBINode::opQuestion);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            qNode->fLeftChild      = operandNode;
+            operandNode->fParent   = qNode;
+        }
+        break;
+
+    case doUnaryOpStar:
+        {
+            RBBINode  *operandNode = fNodeStack[fNodeStackPtr--];
+            RBBINode  *starNode    = pushNewNode(RBBINode::opStar);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            starNode->fLeftChild   = operandNode;
+            operandNode->fParent   = starNode;
+        }
+        break;
+
+    case doRuleChar:
+        // A "Rule Character" is any single character that is a literal part
+        // of the regular expression.  Like a, b and c in the expression "(abc*) | [:L:]"
+        // These are pretty uncommon in break rules; the terms are more commonly
+        //  sets.  To keep things uniform, treat these characters like as
+        // sets that just happen to contain only one character.
+        {
+            n = pushNewNode(RBBINode::setRef);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            findSetFor(UnicodeString(fC.fChar), n);
+            n->fFirstPos = fScanIndex;
+            n->fLastPos  = fNextIndex;
+            fRB->fRules.extractBetween(n->fFirstPos, n->fLastPos, n->fText);
+            break;
+        }
+
+    case doDotAny:
+        // scanned a ".", meaning match any single character.
+        {
+            n = pushNewNode(RBBINode::setRef);
+            if (U_FAILURE(*fRB->fStatus)) {
+                break;
+            }
+            findSetFor(UnicodeString(TRUE, kAny, 3), n);
+            n->fFirstPos = fScanIndex;
+            n->fLastPos  = fNextIndex;
+            fRB->fRules.extractBetween(n->fFirstPos, n->fLastPos, n->fText);
+            break;
+        }
+
+    case doSlash:
+        // Scanned a '/', which identifies a look-ahead break position in a rule.
+        n = pushNewNode(RBBINode::lookAhead);
+        if (U_FAILURE(*fRB->fStatus)) {
+            break;
+        }
+        n->fVal      = fRuleNum;
+        n->fFirstPos = fScanIndex;
+        n->fLastPos  = fNextIndex;
+        fRB->fRules.extractBetween(n->fFirstPos, n->fLastPos, n->fText);
+        fLookAheadRule = TRUE;
+        break;
+
+
+    case doStartTagValue:
+        // Scanned a '{', the opening delimiter for a tag value within a rule.
+        n = pushNewNode(RBBINode::tag);
+        if (U_FAILURE(*fRB->fStatus)) {
+            break;
+        }
+        n->fVal      = 0;
+        n->fFirstPos = fScanIndex;
+        n->fLastPos  = fNextIndex;
+        break;
+
+    case doTagDigit:
+        // Just scanned a decimal digit that's part of a tag value
+        {
+            n = fNodeStack[fNodeStackPtr];
+            uint32_t v = u_charDigitValue(fC.fChar);
+            U_ASSERT(v < 10);
+            n->fVal = n->fVal*10 + v;
+            break;
+        }
+
+    case doTagValue:
+        n = fNodeStack[fNodeStackPtr];
+        n->fLastPos = fNextIndex;
+        fRB->fRules.extractBetween(n->fFirstPos, n->fLastPos, n->fText);
+        break;
+
+    case doTagExpectedError:
+        error(U_BRK_MALFORMED_RULE_TAG);
+        returnVal = FALSE;
+        break;
+
+    case doOptionStart:
+        // Scanning a !!option.   At the start of string.
+        fOptionStart = fScanIndex;
+        break;
+
+    case doOptionEnd:
+        {
+            UnicodeString opt(fRB->fRules, fOptionStart, fScanIndex-fOptionStart);
+            if (opt == UNICODE_STRING("chain", 5)) {
+                fRB->fChainRules = TRUE;
+            } else if (opt == UNICODE_STRING("LBCMNoChain", 11)) {
+                fRB->fLBCMNoChain = TRUE;
+            } else if (opt == UNICODE_STRING("forward", 7)) {
+                fRB->fDefaultTree   = &fRB->fForwardTree;
+            } else if (opt == UNICODE_STRING("reverse", 7)) {
+                fRB->fDefaultTree   = &fRB->fReverseTree;
+            } else if (opt == UNICODE_STRING("safe_forward", 12)) {
+                fRB->fDefaultTree   = &fRB->fSafeFwdTree;
+            } else if (opt == UNICODE_STRING("safe_reverse", 12)) {
+                fRB->fDefaultTree   = &fRB->fSafeRevTree;
+            } else if (opt == UNICODE_STRING("lookAheadHardBreak", 18)) {
+                fRB->fLookAheadHardBreak = TRUE;
+            } else if (opt == UNICODE_STRING("quoted_literals_only", 20)) {
+                fRuleSets[kRuleSet_rule_char-128].clear();
+            } else if (opt == UNICODE_STRING("unquoted_literals",  17)) {
+                fRuleSets[kRuleSet_rule_char-128].applyPattern(UnicodeString(gRuleSet_rule_char_pattern), *fRB->fStatus);
+            } else {
+                error(U_BRK_UNRECOGNIZED_OPTION);
+            }
+        }
+        break;
+
+    case doReverseDir:
+        fReverseRule = TRUE;
+        break;
+
+    case doStartVariableName:
+        n = pushNewNode(RBBINode::varRef);
+        if (U_FAILURE(*fRB->fStatus)) {
+            break;
+        }
+        n->fFirstPos = fScanIndex;
+        break;
+
+    case doEndVariableName:
+        n = fNodeStack[fNodeStackPtr];
+        if (n==NULL || n->fType != RBBINode::varRef) {
+            error(U_BRK_INTERNAL_ERROR);
+            break;
+        }
+        n->fLastPos = fScanIndex;
+        fRB->fRules.extractBetween(n->fFirstPos+1, n->fLastPos, n->fText);
+        // Look the newly scanned name up in the symbol table
+        //   If there's an entry, set the l. child of the var ref to the replacement expression.
+        //   (We also pass through here when scanning assignments, but no harm is done, other
+        //    than a slight wasted effort that seems hard to avoid.  Lookup will be null)
+        n->fLeftChild = fSymbolTable->lookupNode(n->fText);
+        break;
+
+    case doCheckVarDef:
+        n = fNodeStack[fNodeStackPtr];
+        if (n->fLeftChild == NULL) {
+            error(U_BRK_UNDEFINED_VARIABLE);
+            returnVal = FALSE;
+        }
+        break;
+
+    case doExprFinished:
+        break;
+
+    case doRuleErrorAssignExpr:
+        error(U_BRK_ASSIGN_ERROR);
+        returnVal = FALSE;
+        break;
+
+    case doExit:
+        returnVal = FALSE;
+        break;
+
+    case doScanUnicodeSet:
+        scanSet();
+        break;
+
+    default:
+        error(U_BRK_INTERNAL_ERROR);
+        returnVal = FALSE;
+        break;
+    }
+    return returnVal && U_SUCCESS(*fRB->fStatus);
+}
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//  Error         Report a rule parse error.
+//                Only report it if no previous error has been recorded.
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::error(UErrorCode e) {
+    if (U_SUCCESS(*fRB->fStatus)) {
+        *fRB->fStatus = e;
+        if (fRB->fParseError) {
+            fRB->fParseError->line  = fLineNum;
+            fRB->fParseError->offset = fCharNum;
+            fRB->fParseError->preContext[0] = 0;
+            fRB->fParseError->postContext[0] = 0;
+        }
+    }
+}
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//  fixOpStack   The parse stack holds partially assembled chunks of the parse tree.
+//               An entry on the stack may be as small as a single setRef node,
+//               or as large as the parse tree
+//               for an entire expression (this will be the one item left on the stack
+//               when the parsing of an RBBI rule completes.
+//
+//               This function is called when a binary operator is encountered.
+//               It looks back up the stack for operators that are not yet associated
+//               with a right operand, and if the precedence of the stacked operator >=
+//               the precedence of the current operator, binds the operand left,
+//               to the previously encountered operator.
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::fixOpStack(RBBINode::OpPrecedence p) {
+    RBBINode *n;
+    // printNodeStack("entering fixOpStack()");
+    for (;;) {
+        n = fNodeStack[fNodeStackPtr-1];   // an operator node
+        if (n->fPrecedence == 0) {
+            RBBIDebugPuts("RBBIRuleScanner::fixOpStack, bad operator node");
+            error(U_BRK_INTERNAL_ERROR);
+            return;
+        }
+
+        if (n->fPrecedence < p || n->fPrecedence <= RBBINode::precLParen) {
+            // The most recent operand goes with the current operator,
+            //   not with the previously stacked one.
+            break;
+        }
+            // Stack operator is a binary op  ( '|' or concatenation)
+            //   TOS operand becomes right child of this operator.
+            //   Resulting subexpression becomes the TOS operand.
+            n->fRightChild = fNodeStack[fNodeStackPtr];
+            fNodeStack[fNodeStackPtr]->fParent = n;
+            fNodeStackPtr--;
+        // printNodeStack("looping in fixOpStack()   ");
+    }
+
+    if (p <= RBBINode::precLParen) {
+        // Scan is at a right paren or end of expression.
+        //  The scanned item must match the stack, or else there was an error.
+        //  Discard the left paren (or start expr) node from the stack,
+            //  leaving the completed (sub)expression as TOS.
+            if (n->fPrecedence != p) {
+                // Right paren encountered matched start of expression node, or
+                // end of expression matched with a left paren node.
+                error(U_BRK_MISMATCHED_PAREN);
+            }
+            fNodeStack[fNodeStackPtr-1] = fNodeStack[fNodeStackPtr];
+            fNodeStackPtr--;
+            // Delete the now-discarded LParen or Start node.
+            delete n;
+    }
+    // printNodeStack("leaving fixOpStack()");
+}
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//   findSetFor    given a UnicodeString,
+//                  - find the corresponding Unicode Set  (uset node)
+//                         (create one if necessary)
+//                  - Set fLeftChild of the caller's node (should be a setRef node)
+//                         to the uset node
+//                 Maintain a hash table of uset nodes, so the same one is always used
+//                    for the same string.
+//                 If a "to adopt" set is provided and we haven't seen this key before,
+//                    add the provided set to the hash table.
+//                 If the string is one (32 bit) char in length, the set contains
+//                    just one element which is the char in question.
+//                 If the string is "any", return a set containing all chars.
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::findSetFor(const UnicodeString &s, RBBINode *node, UnicodeSet *setToAdopt) {
+
+    RBBISetTableEl   *el;
+
+    // First check whether we've already cached a set for this string.
+    // If so, just use the cached set in the new node.
+    //   delete any set provided by the caller, since we own it.
+    el = (RBBISetTableEl *)uhash_get(fSetTable, &s);
+    if (el != NULL) {
+        delete setToAdopt;
+        node->fLeftChild = el->val;
+        U_ASSERT(node->fLeftChild->fType == RBBINode::uset);
+        return;
+    }
+
+    // Haven't seen this set before.
+    // If the caller didn't provide us with a prebuilt set,
+    //   create a new UnicodeSet now.
+    if (setToAdopt == NULL) {
+        if (s.compare(kAny, -1) == 0) {
+            setToAdopt = new UnicodeSet(0x000000, 0x10ffff);
+        } else {
+            UChar32 c;
+            c = s.char32At(0);
+            setToAdopt = new UnicodeSet(c, c);
+        }
+    }
+
+    //
+    // Make a new uset node to refer to this UnicodeSet
+    // This new uset node becomes the child of the caller's setReference node.
+    //
+    RBBINode *usetNode    = new RBBINode(RBBINode::uset);
+    if (usetNode == NULL) {
+        error(U_MEMORY_ALLOCATION_ERROR);
+        return;
+    }
+    usetNode->fInputSet   = setToAdopt;
+    usetNode->fParent     = node;
+    node->fLeftChild      = usetNode;
+    usetNode->fText = s;
+
+
+    //
+    // Add the new uset node to the list of all uset nodes.
+    //
+    fRB->fUSetNodes->addElement(usetNode, *fRB->fStatus);
+
+
+    //
+    // Add the new set to the set hash table.
+    //
+    el      = (RBBISetTableEl *)uprv_malloc(sizeof(RBBISetTableEl));
+    UnicodeString *tkey = new UnicodeString(s);
+    if (tkey == NULL || el == NULL || setToAdopt == NULL) {
+        // Delete to avoid memory leak
+        delete tkey;
+        tkey = NULL;
+        uprv_free(el);
+        el = NULL;
+        delete setToAdopt;
+        setToAdopt = NULL;
+
+        error(U_MEMORY_ALLOCATION_ERROR);
+        return;
+    }
+    el->key = tkey;
+    el->val = usetNode;
+    uhash_put(fSetTable, el->key, el, fRB->fStatus);
+
+    return;
+}
+
+
+
+//
+//  Assorted Unicode character constants.
+//     Numeric because there is no portable way to enter them as literals.
+//     (Think EBCDIC).
+//
+static const UChar      chCR        = 0x0d;      // New lines, for terminating comments.
+static const UChar      chLF        = 0x0a;
+static const UChar      chNEL       = 0x85;      //    NEL newline variant
+static const UChar      chLS        = 0x2028;    //    Unicode Line Separator
+static const UChar      chApos      = 0x27;      //  single quote, for quoted chars.
+static const UChar      chPound     = 0x23;      // '#', introduces a comment.
+static const UChar      chBackSlash = 0x5c;      // '\'  introduces a char escape
+static const UChar      chLParen    = 0x28;
+static const UChar      chRParen    = 0x29;
+
+
+//------------------------------------------------------------------------------
+//
+//  stripRules    Return a rules string without extra spaces.
+//                (Comments are removed separately, during rule parsing.)
+//
+//------------------------------------------------------------------------------
+UnicodeString RBBIRuleScanner::stripRules(const UnicodeString &rules) {
+    UnicodeString strippedRules;
+    int32_t rulesLength = rules.length();
+
+    for (int32_t idx=0; idx<rulesLength; idx = rules.moveIndex32(idx, 1)) {
+        UChar32 cp = rules.char32At(idx);
+        bool whiteSpace = u_hasBinaryProperty(cp, UCHAR_PATTERN_WHITE_SPACE);
+        if (whiteSpace) {
+            continue;
+        }
+        strippedRules.append(cp);
+    }
+    return strippedRules;
+}
+
+
+//------------------------------------------------------------------------------
+//
+//  nextCharLL    Low Level Next Char from rule input source.
+//                Get a char from the input character iterator,
+//                keep track of input position for error reporting.
+//
+//------------------------------------------------------------------------------
+UChar32  RBBIRuleScanner::nextCharLL() {
+    UChar32  ch;
+
+    if (fNextIndex >= fRB->fRules.length()) {
+        return (UChar32)-1;
+    }
+    ch         = fRB->fRules.char32At(fNextIndex);
+    fNextIndex = fRB->fRules.moveIndex32(fNextIndex, 1);
+
+    if (ch == chCR ||
+        ch == chNEL ||
+        ch == chLS   ||
+        (ch == chLF && fLastChar != chCR)) {
+        // Character is starting a new line.  Bump up the line number, and
+        //  reset the column to 0.
+        fLineNum++;
+        fCharNum=0;
+        if (fQuoteMode) {
+            error(U_BRK_NEW_LINE_IN_QUOTED_STRING);
+            fQuoteMode = FALSE;
+        }
+    }
+    else {
+        // Character is not starting a new line.  Except in the case of a
+        //   LF following a CR, increment the column position.
+        if (ch != chLF) {
+            fCharNum++;
+        }
+    }
+    fLastChar = ch;
+    return ch;
+}
+
+
+//------------------------------------------------------------------------------
+//
+//   nextChar     for rules scanning.  At this level, we handle stripping
+//                out comments and processing backslash character escapes.
+//                The rest of the rules grammar is handled at the next level up.
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::nextChar(RBBIRuleChar &c) {
+
+    // Unicode Character constants needed for the processing done by nextChar(),
+    //   in hex because literals wont work on EBCDIC machines.
+
+    fScanIndex = fNextIndex;
+    c.fChar    = nextCharLL();
+    c.fEscaped = FALSE;
+
+    //
+    //  check for '' sequence.
+    //  These are recognized in all contexts, whether in quoted text or not.
+    //
+    if (c.fChar == chApos) {
+        if (fRB->fRules.char32At(fNextIndex) == chApos) {
+            c.fChar    = nextCharLL();        // get nextChar officially so character counts
+            c.fEscaped = TRUE;                //   stay correct.
+        }
+        else
+        {
+            // Single quote, by itself.
+            //   Toggle quoting mode.
+            //   Return either '('  or ')', because quotes cause a grouping of the quoted text.
+            fQuoteMode = !fQuoteMode;
+            if (fQuoteMode == TRUE) {
+                c.fChar = chLParen;
+            } else {
+                c.fChar = chRParen;
+            }
+            c.fEscaped = FALSE;      // The paren that we return is not escaped.
+            return;
+        }
+    }
+
+    if (fQuoteMode) {
+        c.fEscaped = TRUE;
+    }
+    else
+    {
+        // We are not in a 'quoted region' of the source.
+        //
+        if (c.fChar == chPound) {
+            // Start of a comment.  Consume the rest of it.
+            //  The new-line char that terminates the comment is always returned.
+            //  It will be treated as white-space, and serves to break up anything
+            //    that might otherwise incorrectly clump together with a comment in
+            //    the middle (a variable name, for example.)
+            int32_t commentStart = fScanIndex;
+            for (;;) {
+                c.fChar = nextCharLL();
+                if (c.fChar == (UChar32)-1 ||  // EOF
+                    c.fChar == chCR     ||
+                    c.fChar == chLF     ||
+                    c.fChar == chNEL    ||
+                    c.fChar == chLS)       {break;}
+            }
+            for (int32_t i=commentStart; i<fNextIndex-1; ++i) {
+                fRB->fStrippedRules.setCharAt(i, u' ');
+            }
+        }
+        if (c.fChar == (UChar32)-1) {
+            return;
+        }
+
+        //
+        //  check for backslash escaped characters.
+        //  Use UnicodeString::unescapeAt() to handle them.
+        //
+        if (c.fChar == chBackSlash) {
+            c.fEscaped = TRUE;
+            int32_t startX = fNextIndex;
+            c.fChar = fRB->fRules.unescapeAt(fNextIndex);
+            if (fNextIndex == startX) {
+                error(U_BRK_HEX_DIGITS_EXPECTED);
+            }
+            fCharNum += fNextIndex-startX;
+        }
+    }
+    // putc(c.fChar, stdout);
+}
+
+//------------------------------------------------------------------------------
+//
+//  Parse RBBI rules.   The state machine for rules parsing is here.
+//                      The state tables are hand-written in the file rbbirpt.txt,
+//                      and converted to the form used here by a perl
+//                      script rbbicst.pl
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::parse() {
+    uint16_t                state;
+    const RBBIRuleTableEl  *tableEl;
+
+    if (U_FAILURE(*fRB->fStatus)) {
+        return;
+    }
+
+    state = 1;
+    nextChar(fC);
+    //
+    // Main loop for the rule parsing state machine.
+    //   Runs once per state transition.
+    //   Each time through optionally performs, depending on the state table,
+    //      - an advance to the the next input char
+    //      - an action to be performed.
+    //      - pushing or popping a state to/from the local state return stack.
+    //
+    for (;;) {
+        //  Bail out if anything has gone wrong.
+        //  RBBI rule file parsing stops on the first error encountered.
+        if (U_FAILURE(*fRB->fStatus)) {
+            break;
+        }
+
+        // Quit if state == 0.  This is the normal way to exit the state machine.
+        //
+        if (state == 0) {
+            break;
+        }
+
+        // Find the state table element that matches the input char from the rule, or the
+        //    class of the input character.  Start with the first table row for this
+        //    state, then linearly scan forward until we find a row that matches the
+        //    character.  The last row for each state always matches all characters, so
+        //    the search will stop there, if not before.
+        //
+        tableEl = &gRuleParseStateTable[state];
+        #ifdef RBBI_DEBUG
+            if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "scan")) {
+                RBBIDebugPrintf("char, line, col = (\'%c\', %d, %d)    state=%s ",
+                    fC.fChar, fLineNum, fCharNum, RBBIRuleStateNames[state]);
+            }
+        #endif
+
+        for (;;) {
+            #ifdef RBBI_DEBUG
+                if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "scan")) { RBBIDebugPrintf("."); fflush(stdout);}
+            #endif
+            if (tableEl->fCharClass < 127 && fC.fEscaped == FALSE &&   tableEl->fCharClass == fC.fChar) {
+                // Table row specified an individual character, not a set, and
+                //   the input character is not escaped, and
+                //   the input character matched it.
+                break;
+            }
+            if (tableEl->fCharClass == 255) {
+                // Table row specified default, match anything character class.
+                break;
+            }
+            if (tableEl->fCharClass == 254 && fC.fEscaped)  {
+                // Table row specified "escaped" and the char was escaped.
+                break;
+            }
+            if (tableEl->fCharClass == 253 && fC.fEscaped &&
+                (fC.fChar == 0x50 || fC.fChar == 0x70 ))  {
+                // Table row specified "escaped P" and the char is either 'p' or 'P'.
+                break;
+            }
+            if (tableEl->fCharClass == 252 && fC.fChar == (UChar32)-1)  {
+                // Table row specified eof and we hit eof on the input.
+                break;
+            }
+
+            if (tableEl->fCharClass >= 128 && tableEl->fCharClass < 240 &&   // Table specs a char class &&
+                fC.fEscaped == FALSE &&                                      //   char is not escaped &&
+                fC.fChar != (UChar32)-1) {                                   //   char is not EOF
+                U_ASSERT((tableEl->fCharClass-128) < UPRV_LENGTHOF(fRuleSets));
+                if (fRuleSets[tableEl->fCharClass-128].contains(fC.fChar)) {
+                    // Table row specified a character class, or set of characters,
+                    //   and the current char matches it.
+                    break;
+                }
+            }
+
+            // No match on this row, advance to the next  row for this state,
+            tableEl++;
+        }
+        if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "scan")) { RBBIDebugPuts("");}
+
+        //
+        // We've found the row of the state table that matches the current input
+        //   character from the rules string.
+        // Perform any action specified  by this row in the state table.
+        if (doParseActions((int32_t)tableEl->fAction) == FALSE) {
+            // Break out of the state machine loop if the
+            //   the action signalled some kind of error, or
+            //   the action was to exit, occurs on normal end-of-rules-input.
+            break;
+        }
+
+        if (tableEl->fPushState != 0) {
+            fStackPtr++;
+            if (fStackPtr >= kStackSize) {
+                error(U_BRK_INTERNAL_ERROR);
+                RBBIDebugPuts("RBBIRuleScanner::parse() - state stack overflow.");
+                fStackPtr--;
+            }
+            fStack[fStackPtr] = tableEl->fPushState;
+        }
+
+        if (tableEl->fNextChar) {
+            nextChar(fC);
+        }
+
+        // Get the next state from the table entry, or from the
+        //   state stack if the next state was specified as "pop".
+        if (tableEl->fNextState != 255) {
+            state = tableEl->fNextState;
+        } else {
+            state = fStack[fStackPtr];
+            fStackPtr--;
+            if (fStackPtr < 0) {
+                error(U_BRK_INTERNAL_ERROR);
+                RBBIDebugPuts("RBBIRuleScanner::parse() - state stack underflow.");
+                fStackPtr++;
+            }
+        }
+
+    }
+
+    if (U_FAILURE(*fRB->fStatus)) {
+        return;
+    }
+    
+    // If there are no forward rules set an error.
+    //
+    if (fRB->fForwardTree == NULL) {
+        error(U_BRK_RULE_SYNTAX);
+        return;
+    }
+
+    //
+    // Parsing of the input RBBI rules is complete.
+    // We now have a parse tree for the rule expressions
+    // and a list of all UnicodeSets that are referenced.
+    //
+#ifdef RBBI_DEBUG
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "symbols")) {fSymbolTable->rbbiSymtablePrint();}
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "ptree")) {
+        RBBIDebugPrintf("Completed Forward Rules Parse Tree...\n");
+        RBBINode::printTree(fRB->fForwardTree, TRUE);
+        RBBIDebugPrintf("\nCompleted Reverse Rules Parse Tree...\n");
+        RBBINode::printTree(fRB->fReverseTree, TRUE);
+        RBBIDebugPrintf("\nCompleted Safe Point Forward Rules Parse Tree...\n");
+        RBBINode::printTree(fRB->fSafeFwdTree, TRUE);
+        RBBIDebugPrintf("\nCompleted Safe Point Reverse Rules Parse Tree...\n");
+        RBBINode::printTree(fRB->fSafeRevTree, TRUE);
+    }
+#endif
+}
+
+
+//------------------------------------------------------------------------------
+//
+//  printNodeStack     for debugging...
+//
+//------------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBIRuleScanner::printNodeStack(const char *title) {
+    int i;
+    RBBIDebugPrintf("%s.  Dumping node stack...\n", title);
+    for (i=fNodeStackPtr; i>0; i--) {RBBINode::printTree(fNodeStack[i], TRUE);}
+}
+#endif
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//  pushNewNode   create a new RBBINode of the specified type and push it
+//                onto the stack of nodes.
+//
+//------------------------------------------------------------------------------
+RBBINode  *RBBIRuleScanner::pushNewNode(RBBINode::NodeType  t) {
+    if (U_FAILURE(*fRB->fStatus)) {
+        return NULL;
+    }
+    if (fNodeStackPtr >= kStackSize - 1) {
+        error(U_BRK_RULE_SYNTAX);
+        RBBIDebugPuts("RBBIRuleScanner::pushNewNode - stack overflow.");
+        return NULL;
+    }
+    fNodeStackPtr++;
+    fNodeStack[fNodeStackPtr] = new RBBINode(t);
+    if (fNodeStack[fNodeStackPtr] == NULL) {
+        *fRB->fStatus = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return fNodeStack[fNodeStackPtr];
+}
+
+
+
+//------------------------------------------------------------------------------
+//
+//  scanSet    Construct a UnicodeSet from the text at the current scan
+//             position.  Advance the scan position to the first character
+//             after the set.
+//
+//             A new RBBI setref node referring to the set is pushed onto the node
+//             stack.
+//
+//             The scan position is normally under the control of the state machine
+//             that controls rule parsing.  UnicodeSets, however, are parsed by
+//             the UnicodeSet constructor, not by the RBBI rule parser.
+//
+//------------------------------------------------------------------------------
+void RBBIRuleScanner::scanSet() {
+    UnicodeSet    *uset;
+    ParsePosition  pos;
+    int            startPos;
+    int            i;
+
+    if (U_FAILURE(*fRB->fStatus)) {
+        return;
+    }
+
+    pos.setIndex(fScanIndex);
+    startPos = fScanIndex;
+    UErrorCode localStatus = U_ZERO_ERROR;
+    uset = new UnicodeSet();
+    if (uset == NULL) {
+        localStatus = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        uset->applyPatternIgnoreSpace(fRB->fRules, pos, fSymbolTable, localStatus);
+    }
+    if (U_FAILURE(localStatus)) {
+        //  TODO:  Get more accurate position of the error from UnicodeSet's return info.
+        //         UnicodeSet appears to not be reporting correctly at this time.
+        #ifdef RBBI_DEBUG
+            RBBIDebugPrintf("UnicodeSet parse postion.ErrorIndex = %d\n", pos.getIndex());
+        #endif
+        error(localStatus);
+        delete uset;
+        return;
+    }
+
+    // Verify that the set contains at least one code point.
+    //
+    U_ASSERT(uset!=NULL);
+    if (uset->isEmpty()) {
+        // This set is empty.
+        //  Make it an error, because it almost certainly is not what the user wanted.
+        //  Also, avoids having to think about corner cases in the tree manipulation code
+        //   that occurs later on.
+        error(U_BRK_RULE_EMPTY_SET);
+        delete uset;
+        return;
+    }
+
+
+    // Advance the RBBI parse postion over the UnicodeSet pattern.
+    //   Don't just set fScanIndex because the line/char positions maintained
+    //   for error reporting would be thrown off.
+    i = pos.getIndex();
+    for (;;) {
+        if (fNextIndex >= i) {
+            break;
+        }
+        nextCharLL();
+    }
+
+    if (U_SUCCESS(*fRB->fStatus)) {
+        RBBINode         *n;
+
+        n = pushNewNode(RBBINode::setRef);
+        if (U_FAILURE(*fRB->fStatus)) {
+            return;
+        }
+        n->fFirstPos = startPos;
+        n->fLastPos  = fNextIndex;
+        fRB->fRules.extractBetween(n->fFirstPos, n->fLastPos, n->fText);
+        //  findSetFor() serves several purposes here:
+        //     - Adopts storage for the UnicodeSet, will be responsible for deleting.
+        //     - Mantains collection of all sets in use, needed later for establishing
+        //          character categories for run time engine.
+        //     - Eliminates mulitiple instances of the same set.
+        //     - Creates a new uset node if necessary (if this isn't a duplicate.)
+        findSetFor(n->fText, n, uset);
+    }
+
+}
+
+int32_t RBBIRuleScanner::numRules() {
+    return fRuleNum;
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbiscan.h b/thirdparty/icu4c/common/rbbiscan.h
new file mode 100644
index 0000000000..58022002c5
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbiscan.h
@@ -0,0 +1,167 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  rbbiscan.h
+//
+//  Copyright (C) 2002-2016, International Business Machines Corporation and others.
+//  All Rights Reserved.
+//
+//  This file contains declarations for class RBBIRuleScanner
+//
+
+
+#ifndef RBBISCAN_H
+#define RBBISCAN_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "unicode/rbbi.h"
+#include "unicode/uniset.h"
+#include "unicode/parseerr.h"
+#include "uhash.h"
+#include "uvector.h"
+#include "unicode/symtable.h"// For UnicodeSet parsing, is the interface that
+                          //    looks up references to $variables within a set.
+#include "rbbinode.h"
+#include "rbbirpt.h"
+
+U_NAMESPACE_BEGIN
+
+class   RBBIRuleBuilder;
+class   RBBISymbolTable;
+
+
+//--------------------------------------------------------------------------------
+//
+//  class RBBIRuleScanner does the lowest level, character-at-a-time
+//                        scanning of break iterator rules.  
+//
+//                        The output of the scanner is parse trees for
+//                        the rule expressions and a list of all Unicode Sets
+//                        encountered.
+//
+//--------------------------------------------------------------------------------
+
+class RBBIRuleScanner : public UMemory {
+public:
+
+    enum {
+        kStackSize = 100            // The size of the state stack for
+    };                              //   rules parsing.  Corresponds roughly
+                                    //   to the depth of parentheses nesting
+                                    //   that is allowed in the rules.
+
+    struct RBBIRuleChar {
+        UChar32             fChar;
+        UBool               fEscaped;
+        RBBIRuleChar() : fChar(0), fEscaped(false) {}
+    };
+
+    RBBIRuleScanner(RBBIRuleBuilder  *rb);
+
+
+    virtual    ~RBBIRuleScanner();
+
+    void        nextChar(RBBIRuleChar &c);          // Get the next char from the input stream.
+                                                    // Return false if at end.
+
+    UBool       push(const RBBIRuleChar &c);        // Push (unget) one character.
+                                                    //   Only a single character may be pushed.
+
+    void        parse();                            // Parse the rules, generating two parse
+                                                    //   trees, one each for the forward and
+                                                    //   reverse rules,
+                                                    //   and a list of UnicodeSets encountered.
+
+    int32_t     numRules();                         // Return the number of rules that have been seen.
+
+    /**
+     * Return a rules string without unnecessary
+     * characters.
+     */
+    static UnicodeString stripRules(const UnicodeString &rules);
+private:
+
+    UBool       doParseActions(int32_t a);
+    void        error(UErrorCode e);                   // error reporting convenience function.
+    void        fixOpStack(RBBINode::OpPrecedence p);
+                                                       //   a character.
+    void        findSetFor(const UnicodeString &s, RBBINode *node, UnicodeSet *setToAdopt = NULL);
+
+    UChar32     nextCharLL();
+#ifdef RBBI_DEBUG
+    void        printNodeStack(const char *title);
+#endif
+    RBBINode    *pushNewNode(RBBINode::NodeType  t);
+    void        scanSet();
+
+
+    RBBIRuleBuilder               *fRB;              // The rule builder that we are part of.
+
+    int32_t                       fScanIndex;        // Index of current character being processed
+                                                     //   in the rule input string.
+    int32_t                       fNextIndex;        // Index of the next character, which
+                                                     //   is the first character not yet scanned.
+    UBool                         fQuoteMode;        // Scan is in a 'quoted region'
+    int32_t                       fLineNum;          // Line number in input file.
+    int32_t                       fCharNum;          // Char position within the line.
+    UChar32                       fLastChar;         // Previous char, needed to count CR-LF
+                                                     //   as a single line, not two.
+
+    RBBIRuleChar                  fC;                // Current char for parse state machine
+                                                     //   processing.
+    UnicodeString                 fVarName;          // $variableName, valid when we've just
+                                                     //   scanned one.
+
+    RBBIRuleTableEl               **fStateTable;     // State Transition Table for RBBI Rule
+                                                     //   parsing.  index by p[state][char-class]
+
+    uint16_t                      fStack[kStackSize];  // State stack, holds state pushes
+    int32_t                       fStackPtr;           //  and pops as specified in the state
+                                                       //  transition rules.
+
+    RBBINode                      *fNodeStack[kStackSize]; // Node stack, holds nodes created
+                                                           //  during the parse of a rule
+    int32_t                        fNodeStackPtr;
+
+
+    UBool                          fReverseRule;     // True if the rule currently being scanned
+                                                     //  is a reverse direction rule (if it
+                                                     //  starts with a '!')
+
+    UBool                          fLookAheadRule;   // True if the rule includes a '/'
+                                                     //   somewhere within it.
+
+    UBool                          fNoChainInRule;   // True if the current rule starts with a '^'.
+
+    RBBISymbolTable               *fSymbolTable;     // symbol table, holds definitions of
+                                                     //   $variable symbols.
+
+    UHashtable                    *fSetTable;        // UnicocodeSet hash table, holds indexes to
+                                                     //   the sets created while parsing rules.
+                                                     //   The key is the string used for creating
+                                                     //   the set.
+
+    UnicodeSet                     fRuleSets[10];    // Unicode Sets that are needed during
+                                                     //  the scanning of RBBI rules.  The
+                                                     //  indicies for these are assigned by the
+                                                     //  perl script that builds the state tables.
+                                                     //  See rbbirpt.h.
+
+    int32_t                        fRuleNum;         // Counts each rule as it is scanned.
+
+    int32_t                        fOptionStart;     // Input index of start of a !!option
+                                                     //   keyword, while being scanned.
+
+    UnicodeSet *gRuleSet_rule_char;
+    UnicodeSet *gRuleSet_white_space;
+    UnicodeSet *gRuleSet_name_char;
+    UnicodeSet *gRuleSet_name_start_char;
+
+    RBBIRuleScanner(const RBBIRuleScanner &other); // forbid copying of this class
+    RBBIRuleScanner &operator=(const RBBIRuleScanner &other); // forbid copying of this class
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/rbbisetb.cpp b/thirdparty/icu4c/common/rbbisetb.cpp
new file mode 100644
index 0000000000..29faeb8c45
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbisetb.cpp
@@ -0,0 +1,694 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  rbbisetb.cpp
+//
+/*
+***************************************************************************
+*   Copyright (C) 2002-2008 International Business Machines Corporation   *
+*   and others. All rights reserved.                                      *
+***************************************************************************
+*/
+//
+//  RBBISetBuilder   Handles processing of Unicode Sets from RBBI rules
+//                   (part of the rule building process.)
+//
+//      Starting with the rules parse tree from the scanner,
+//
+//                   -  Enumerate the set of UnicodeSets that are referenced
+//                      by the RBBI rules.
+//                   -  compute a set of non-overlapping character ranges
+//                      with all characters within a range belonging to the same
+//                      set of input unicode sets.
+//                   -  Derive a set of non-overlapping UnicodeSet (like things)
+//                      that will correspond to columns in the state table for
+//                      the RBBI execution engine.  All characters within one
+//                      of these sets belong to the same set of the original
+//                      UnicodeSets from the user's rules.
+//                   -  construct the trie table that maps input characters
+//                      to the index of the matching non-overlapping set of set from
+//                      the previous step.
+//
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/uniset.h"
+#include "uvector.h"
+#include "uassert.h"
+#include "cmemory.h"
+#include "cstring.h"
+
+#include "rbbisetb.h"
+#include "rbbinode.h"
+
+U_NAMESPACE_BEGIN
+
+const int32_t kMaxCharCategoriesFor8BitsTrie = 255;
+//------------------------------------------------------------------------
+//
+//   Constructor
+//
+//------------------------------------------------------------------------
+RBBISetBuilder::RBBISetBuilder(RBBIRuleBuilder *rb)
+{
+    fRB             = rb;
+    fStatus         = rb->fStatus;
+    fRangeList      = nullptr;
+    fMutableTrie    = nullptr;
+    fTrie           = nullptr;
+    fTrieSize       = 0;
+    fGroupCount     = 0;
+    fSawBOF         = false;
+}
+
+
+//------------------------------------------------------------------------
+//
+//   Destructor
+//
+//------------------------------------------------------------------------
+RBBISetBuilder::~RBBISetBuilder()
+{
+    RangeDescriptor   *nextRangeDesc;
+
+    // Walk through & delete the linked list of RangeDescriptors
+    for (nextRangeDesc = fRangeList; nextRangeDesc!=NULL;) {
+        RangeDescriptor *r = nextRangeDesc;
+        nextRangeDesc      = r->fNext;
+        delete r;
+    }
+
+    ucptrie_close(fTrie);
+    umutablecptrie_close(fMutableTrie);
+}
+
+
+
+
+//------------------------------------------------------------------------
+//
+//   build          Build the list of non-overlapping character ranges
+//                  from the Unicode Sets.
+//
+//------------------------------------------------------------------------
+void RBBISetBuilder::buildRanges() {
+    RBBINode        *usetNode;
+    RangeDescriptor *rlRange;
+
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "usets")) {printSets();}
+
+    //
+    //  Initialize the process by creating a single range encompassing all characters
+    //  that is in no sets.
+    //
+    fRangeList                = new RangeDescriptor(*fStatus); // will check for status here
+    if (fRangeList == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    fRangeList->fStartChar    = 0;
+    fRangeList->fEndChar      = 0x10ffff;
+
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    //
+    //  Find the set of non-overlapping ranges of characters
+    //
+    int  ni;
+    for (ni=0; ; ni++) {        // Loop over each of the UnicodeSets encountered in the input rules
+        usetNode = (RBBINode *)this->fRB->fUSetNodes->elementAt(ni);
+        if (usetNode==NULL) {
+            break;
+        }
+
+        UnicodeSet      *inputSet             = usetNode->fInputSet;
+        int32_t          inputSetRangeCount   = inputSet->getRangeCount();
+        int              inputSetRangeIndex   = 0;
+                         rlRange              = fRangeList;
+
+        for (;;) {
+            if (inputSetRangeIndex >= inputSetRangeCount) {
+                break;
+            }
+            UChar32      inputSetRangeBegin  = inputSet->getRangeStart(inputSetRangeIndex);
+            UChar32      inputSetRangeEnd    = inputSet->getRangeEnd(inputSetRangeIndex);
+
+            // skip over ranges from the range list that are completely
+            //   below the current range from the input unicode set.
+            while (rlRange->fEndChar < inputSetRangeBegin) {
+                rlRange = rlRange->fNext;
+            }
+
+            // If the start of the range from the range list is before with
+            //   the start of the range from the unicode set, split the range list range
+            //   in two, with one part being before (wholly outside of) the unicode set
+            //   and the other containing the rest.
+            //   Then continue the loop; the post-split current range will then be skipped
+            //     over
+            if (rlRange->fStartChar < inputSetRangeBegin) {
+                rlRange->split(inputSetRangeBegin, *fStatus);
+                if (U_FAILURE(*fStatus)) {
+                    return;
+                }
+                continue;
+            }
+
+            // Same thing at the end of the ranges...
+            // If the end of the range from the range list doesn't coincide with
+            //   the end of the range from the unicode set, split the range list
+            //   range in two.  The first part of the split range will be
+            //   wholly inside the Unicode set.
+            if (rlRange->fEndChar > inputSetRangeEnd) {
+                rlRange->split(inputSetRangeEnd+1, *fStatus);
+                if (U_FAILURE(*fStatus)) {
+                    return;
+                }
+            }
+
+            // The current rlRange is now entirely within the UnicodeSet range.
+            // Add this unicode set to the list of sets for this rlRange
+            if (rlRange->fIncludesSets->indexOf(usetNode) == -1) {
+                rlRange->fIncludesSets->addElement(usetNode, *fStatus);
+                if (U_FAILURE(*fStatus)) {
+                    return;
+                }
+            }
+
+            // Advance over ranges that we are finished with.
+            if (inputSetRangeEnd == rlRange->fEndChar) {
+                inputSetRangeIndex++;
+            }
+            rlRange = rlRange->fNext;
+        }
+    }
+
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "range")) { printRanges();}
+
+    //
+    //  Group the above ranges, with each group consisting of one or more
+    //    ranges that are in exactly the same set of original UnicodeSets.
+    //    The groups are numbered, and these group numbers are the set of
+    //    input symbols recognized by the run-time state machine.
+    //
+    //    Numbering: # 0  (state table column 0) is unused.
+    //               # 1  is reserved - table column 1 is for end-of-input
+    //               # 2  is reserved - table column 2 is for beginning-of-input
+    //               # 3  is the first range list.
+    //
+    RangeDescriptor *rlSearchRange;
+    int32_t dictGroupCount = 0;
+
+    for (rlRange = fRangeList; rlRange!=nullptr; rlRange=rlRange->fNext) {
+        for (rlSearchRange=fRangeList; rlSearchRange != rlRange; rlSearchRange=rlSearchRange->fNext) {
+            if (rlRange->fIncludesSets->equals(*rlSearchRange->fIncludesSets)) {
+                rlRange->fNum = rlSearchRange->fNum;
+                rlRange->fIncludesDict = rlSearchRange->fIncludesDict;
+                break;
+            }
+        }
+        if (rlRange->fNum == 0) {
+            rlRange->fFirstInGroup = true;
+            if (rlRange->isDictionaryRange()) {
+                rlRange->fNum = ++dictGroupCount;
+                rlRange->fIncludesDict = true;
+            } else {
+                fGroupCount++;
+                rlRange->fNum = fGroupCount+2;
+                addValToSets(rlRange->fIncludesSets, rlRange->fNum);
+            }
+        }
+    }
+
+    // Move the character category numbers for any dictionary ranges up, so that they
+    // immediately follow the non-dictionary ranges.
+
+    fDictCategoriesStart = fGroupCount + 3;
+    for (rlRange = fRangeList; rlRange!=nullptr; rlRange=rlRange->fNext) {
+        if (rlRange->fIncludesDict) {
+            rlRange->fNum += fDictCategoriesStart - 1;
+            if (rlRange->fFirstInGroup) {
+                addValToSets(rlRange->fIncludesSets, rlRange->fNum);
+            }
+        }
+    }
+    fGroupCount += dictGroupCount;
+
+
+    // Handle input sets that contain the special string {eof}.
+    //   Column 1 of the state table is reserved for EOF on input.
+    //   Column 2 is reserved for before-the-start-input.
+    //            (This column can be optimized away later if there are no rule
+    //             references to {bof}.)
+    //   Add this column value (1 or 2) to the equivalent expression
+    //     subtree for each UnicodeSet that contains the string {eof}
+    //   Because {bof} and {eof} are not characters in the normal sense,
+    //   they don't affect the computation of the ranges or TRIE.
+
+    UnicodeString eofString(u"eof");
+    UnicodeString bofString(u"bof");
+    for (ni=0; ; ni++) {        // Loop over each of the UnicodeSets encountered in the input rules
+        usetNode = (RBBINode *)this->fRB->fUSetNodes->elementAt(ni);
+        if (usetNode==NULL) {
+            break;
+        }
+        UnicodeSet      *inputSet = usetNode->fInputSet;
+        if (inputSet->contains(eofString)) {
+            addValToSet(usetNode, 1);
+        }
+        if (inputSet->contains(bofString)) {
+            addValToSet(usetNode, 2);
+            fSawBOF = TRUE;
+        }
+    }
+
+
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "rgroup")) {printRangeGroups();}
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "esets")) {printSets();}
+}
+
+
+//
+// Build the Trie table for mapping UChar32 values to the corresponding
+// range group number.
+//
+void RBBISetBuilder::buildTrie() {
+    fMutableTrie = umutablecptrie_open(
+                        0,       //  Initial value for all code points.
+                        0,       //  Error value for out-of-range input.
+                        fStatus);
+
+    for (RangeDescriptor *range = fRangeList; range!=nullptr && U_SUCCESS(*fStatus); range=range->fNext) {
+        umutablecptrie_setRange(fMutableTrie,
+                                range->fStartChar,     // Range start
+                                range->fEndChar,       // Range end (inclusive)
+                                range->fNum,           // value for range
+                                fStatus);
+    }
+}
+
+
+void RBBISetBuilder::mergeCategories(IntPair categories) {
+    U_ASSERT(categories.first >= 1);
+    U_ASSERT(categories.second > categories.first);
+    U_ASSERT((categories.first <  fDictCategoriesStart && categories.second <  fDictCategoriesStart) ||
+             (categories.first >= fDictCategoriesStart && categories.second >= fDictCategoriesStart));
+
+    for (RangeDescriptor *rd = fRangeList; rd != nullptr; rd = rd->fNext) {
+        int32_t rangeNum = rd->fNum;
+        if (rangeNum == categories.second) {
+            rd->fNum = categories.first;
+        } else if (rangeNum > categories.second) {
+            rd->fNum--;
+        }
+    }
+    --fGroupCount;
+    if (categories.second <= fDictCategoriesStart) {
+        --fDictCategoriesStart;
+    }
+}
+
+
+//-----------------------------------------------------------------------------------
+//
+//  getTrieSize()    Return the size that will be required to serialize the Trie.
+//
+//-----------------------------------------------------------------------------------
+int32_t RBBISetBuilder::getTrieSize()  {
+    if (U_FAILURE(*fStatus)) {
+        return 0;
+    }
+    if (fTrie == nullptr) {
+        bool use8Bits = getNumCharCategories() <= kMaxCharCategoriesFor8BitsTrie;
+        fTrie = umutablecptrie_buildImmutable(
+            fMutableTrie,
+            UCPTRIE_TYPE_FAST,
+            use8Bits ? UCPTRIE_VALUE_BITS_8 : UCPTRIE_VALUE_BITS_16,
+            fStatus);
+        fTrieSize = ucptrie_toBinary(fTrie, nullptr, 0, fStatus);
+        if (*fStatus == U_BUFFER_OVERFLOW_ERROR) {
+            *fStatus = U_ZERO_ERROR;
+        }
+    }
+    return fTrieSize;
+}
+
+
+//-----------------------------------------------------------------------------------
+//
+//  serializeTrie()   Put the serialized trie at the specified address.
+//                    Trust the caller to have given us enough memory.
+//                    getTrieSize() MUST be called first.
+//
+//-----------------------------------------------------------------------------------
+void RBBISetBuilder::serializeTrie(uint8_t *where) {
+    ucptrie_toBinary(fTrie,
+                     where,                // Buffer
+                     fTrieSize,            // Capacity
+                     fStatus);
+}
+
+//------------------------------------------------------------------------
+//
+//  addValToSets     Add a runtime-mapped input value to each uset from a
+//                   list of uset nodes. (val corresponds to a state table column.)
+//                   For each of the original Unicode sets - which correspond
+//                   directly to uset nodes - a logically equivalent expression
+//                   is constructed in terms of the remapped runtime input
+//                   symbol set.  This function adds one runtime input symbol to
+//                   a list of sets.
+//
+//                   The "logically equivalent expression" is the tree for an
+//                   or-ing together of all of the symbols that go into the set.
+//
+//------------------------------------------------------------------------
+void  RBBISetBuilder::addValToSets(UVector *sets, uint32_t val) {
+    int32_t       ix;
+
+    for (ix=0; ix<sets->size(); ix++) {
+        RBBINode *usetNode = (RBBINode *)sets->elementAt(ix);
+        addValToSet(usetNode, val);
+    }
+}
+
+void  RBBISetBuilder::addValToSet(RBBINode *usetNode, uint32_t val) {
+    RBBINode *leafNode = new RBBINode(RBBINode::leafChar);
+    if (leafNode == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    leafNode->fVal = (unsigned short)val;
+    if (usetNode->fLeftChild == NULL) {
+        usetNode->fLeftChild = leafNode;
+        leafNode->fParent    = usetNode;
+    } else {
+        // There are already input symbols present for this set.
+        // Set up an OR node, with the previous stuff as the left child
+        //   and the new value as the right child.
+        RBBINode *orNode = new RBBINode(RBBINode::opOr);
+        if (orNode == NULL) {
+            *fStatus = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        orNode->fLeftChild  = usetNode->fLeftChild;
+        orNode->fRightChild = leafNode;
+        orNode->fLeftChild->fParent  = orNode;
+        orNode->fRightChild->fParent = orNode;
+        usetNode->fLeftChild = orNode;
+        orNode->fParent = usetNode;
+    }
+}
+
+
+//------------------------------------------------------------------------
+//
+//   getNumCharCategories
+//
+//------------------------------------------------------------------------
+int32_t  RBBISetBuilder::getNumCharCategories() const {
+    return fGroupCount + 3;
+}
+
+
+//------------------------------------------------------------------------
+//
+//   getDictCategoriesStart
+//
+//------------------------------------------------------------------------
+int32_t  RBBISetBuilder::getDictCategoriesStart() const {
+    return fDictCategoriesStart;
+}
+
+
+//------------------------------------------------------------------------
+//
+//   sawBOF
+//
+//------------------------------------------------------------------------
+UBool  RBBISetBuilder::sawBOF() const {
+    return fSawBOF;
+}
+
+
+//------------------------------------------------------------------------
+//
+//   getFirstChar      Given a runtime RBBI character category, find
+//                     the first UChar32 that is in the set of chars 
+//                     in the category.
+//------------------------------------------------------------------------
+UChar32  RBBISetBuilder::getFirstChar(int32_t category) const {
+    RangeDescriptor   *rlRange;
+    UChar32            retVal = (UChar32)-1;
+    for (rlRange = fRangeList; rlRange!=nullptr; rlRange=rlRange->fNext) {
+        if (rlRange->fNum == category) {
+            retVal = rlRange->fStartChar;
+            break;
+        }
+    }
+    return retVal;
+}
+
+
+//------------------------------------------------------------------------
+//
+//   printRanges        A debugging function.
+//                      dump out all of the range definitions.
+//
+//------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBISetBuilder::printRanges() {
+    RangeDescriptor       *rlRange;
+    int                    i;
+
+    RBBIDebugPrintf("\n\n Nonoverlapping Ranges ...\n");
+    for (rlRange = fRangeList; rlRange!=nullptr; rlRange=rlRange->fNext) {
+        RBBIDebugPrintf("%4x-%4x  ", rlRange->fStartChar, rlRange->fEndChar);
+
+        for (i=0; i<rlRange->fIncludesSets->size(); i++) {
+            RBBINode       *usetNode    = (RBBINode *)rlRange->fIncludesSets->elementAt(i);
+            UnicodeString   setName {u"anon"};
+            RBBINode       *setRef = usetNode->fParent;
+            if (setRef != nullptr) {
+                RBBINode *varRef = setRef->fParent;
+                if (varRef != nullptr  &&  varRef->fType == RBBINode::varRef) {
+                    setName = varRef->fText;
+                }
+            }
+            RBBI_DEBUG_printUnicodeString(setName); RBBIDebugPrintf("  ");
+        }
+        RBBIDebugPrintf("\n");
+    }
+}
+#endif
+
+
+//------------------------------------------------------------------------
+//
+//   printRangeGroups     A debugging function.
+//                        dump out all of the range groups.
+//
+//------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBISetBuilder::printRangeGroups() {
+    int                    i;
+
+    RBBIDebugPrintf("\nRanges grouped by Unicode Set Membership...\n");
+    for (RangeDescriptor *rlRange = fRangeList; rlRange!=nullptr; rlRange=rlRange->fNext) {
+        if (rlRange->fFirstInGroup) {
+            int groupNum = rlRange->fNum;
+            RBBIDebugPrintf("%2i  ", groupNum);
+
+            if (groupNum >= fDictCategoriesStart) { RBBIDebugPrintf(" <DICT> ");}
+
+            for (i=0; i<rlRange->fIncludesSets->size(); i++) {
+                RBBINode       *usetNode    = (RBBINode *)rlRange->fIncludesSets->elementAt(i);
+                UnicodeString   setName = UNICODE_STRING("anon", 4);
+                RBBINode       *setRef = usetNode->fParent;
+                if (setRef != NULL) {
+                    RBBINode *varRef = setRef->fParent;
+                    if (varRef != NULL  &&  varRef->fType == RBBINode::varRef) {
+                        setName = varRef->fText;
+                    }
+                }
+                RBBI_DEBUG_printUnicodeString(setName); RBBIDebugPrintf(" ");
+            }
+
+            i = 0;
+            for (RangeDescriptor *tRange = rlRange; tRange != nullptr; tRange = tRange->fNext) {
+                if (tRange->fNum == rlRange->fNum) {
+                    if (i++ % 5 == 0) {
+                        RBBIDebugPrintf("\n    ");
+                    }
+                    RBBIDebugPrintf("  %05x-%05x", tRange->fStartChar, tRange->fEndChar);
+                }
+            }
+            RBBIDebugPrintf("\n");
+        }
+    }
+    RBBIDebugPrintf("\n");
+}
+#endif
+
+
+//------------------------------------------------------------------------
+//
+//   printSets          A debugging function.
+//                      dump out all of the set definitions.
+//
+//------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBISetBuilder::printSets() {
+    int                   i;
+
+    RBBIDebugPrintf("\n\nUnicode Sets List\n------------------\n");
+    for (i=0; ; i++) {
+        RBBINode        *usetNode;
+        RBBINode        *setRef;
+        RBBINode        *varRef;
+        UnicodeString    setName;
+
+        usetNode = (RBBINode *)fRB->fUSetNodes->elementAt(i);
+        if (usetNode == NULL) {
+            break;
+        }
+
+        RBBIDebugPrintf("%3d    ", i);
+        setName = UNICODE_STRING("anonymous", 9);
+        setRef = usetNode->fParent;
+        if (setRef != NULL) {
+            varRef = setRef->fParent;
+            if (varRef != NULL  &&  varRef->fType == RBBINode::varRef) {
+                setName = varRef->fText;
+            }
+        }
+        RBBI_DEBUG_printUnicodeString(setName);
+        RBBIDebugPrintf("   ");
+        RBBI_DEBUG_printUnicodeString(usetNode->fText);
+        RBBIDebugPrintf("\n");
+        if (usetNode->fLeftChild != NULL) {
+            RBBINode::printTree(usetNode->fLeftChild, TRUE);
+        }
+    }
+    RBBIDebugPrintf("\n");
+}
+#endif
+
+
+
+//-------------------------------------------------------------------------------------
+//
+//  RangeDescriptor copy constructor
+//
+//-------------------------------------------------------------------------------------
+
+RangeDescriptor::RangeDescriptor(const RangeDescriptor &other, UErrorCode &status) :
+        fStartChar(other.fStartChar), fEndChar {other.fEndChar}, fNum {other.fNum},
+        fIncludesDict{other.fIncludesDict}, fFirstInGroup{other.fFirstInGroup} {
+
+    if (U_FAILURE(status)) {
+        return;
+    }
+    fIncludesSets = new UVector(status);
+    if (this->fIncludesSets == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_FAILURE(status)) {
+        return;
+    }
+
+    for (int32_t i=0; i<other.fIncludesSets->size(); i++) {
+        this->fIncludesSets->addElement(other.fIncludesSets->elementAt(i), status);
+    }
+}
+
+
+//-------------------------------------------------------------------------------------
+//
+//  RangeDesriptor default constructor
+//
+//-------------------------------------------------------------------------------------
+RangeDescriptor::RangeDescriptor(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    fIncludesSets = new UVector(status);
+    if (fIncludesSets == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+
+//-------------------------------------------------------------------------------------
+//
+//  RangeDesriptor Destructor
+//
+//-------------------------------------------------------------------------------------
+RangeDescriptor::~RangeDescriptor() {
+    delete  fIncludesSets;
+    fIncludesSets = nullptr;
+}
+
+//-------------------------------------------------------------------------------------
+//
+//  RangeDesriptor::split()
+//
+//-------------------------------------------------------------------------------------
+void RangeDescriptor::split(UChar32 where, UErrorCode &status) {
+    U_ASSERT(where>fStartChar && where<=fEndChar);
+    RangeDescriptor *nr = new RangeDescriptor(*this, status);
+    if(nr == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    if (U_FAILURE(status)) {
+        delete nr;
+        return;
+    }
+    //  RangeDescriptor copy constructor copies all fields.
+    //  Only need to update those that are different after the split.
+    nr->fStartChar = where;
+    this->fEndChar = where-1;
+    nr->fNext      = this->fNext;
+    this->fNext    = nr;
+}
+
+
+//-------------------------------------------------------------------------------------
+//
+//   RangeDescriptor::isDictionaryRange
+//
+//            Test whether this range includes characters from
+//            the original Unicode Set named "dictionary".
+//
+//            This function looks through the Unicode Sets that
+//            the range includes, checking for one named "dictionary"
+//
+//            TODO:  a faster way would be to find the set node for
+//                   "dictionary" just once, rather than looking it
+//                   up by name every time.
+//
+//-------------------------------------------------------------------------------------
+bool RangeDescriptor::isDictionaryRange() {
+    static const char16_t *dictionary = u"dictionary";
+    for (int32_t i=0; i<fIncludesSets->size(); i++) {
+        RBBINode *usetNode  = (RBBINode *)fIncludesSets->elementAt(i);
+        RBBINode *setRef = usetNode->fParent;
+        if (setRef != nullptr) {
+            RBBINode *varRef = setRef->fParent;
+            if (varRef && varRef->fType == RBBINode::varRef) {
+                const UnicodeString *setName = &varRef->fText;
+                if (setName->compare(dictionary, -1) == 0) {
+                    return true;
+                }
+            }
+        }
+    }
+    return false;
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbisetb.h b/thirdparty/icu4c/common/rbbisetb.h
new file mode 100644
index 0000000000..6409a4ea57
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbisetb.h
@@ -0,0 +1,147 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  rbbisetb.h
+/*
+**********************************************************************
+*   Copyright (c) 2001-2005, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef RBBISETB_H
+#define RBBISETB_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/ucptrie.h"
+#include "unicode/umutablecptrie.h"
+#include "unicode/uobject.h"
+#include "rbbirb.h"
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+//
+//  RBBISetBuilder   Derives the character categories used by the runtime RBBI engine
+//                   from the Unicode Sets appearing in the source  RBBI rules, and
+//                   creates the TRIE table used to map from Unicode to the
+//                   character categories.
+//
+
+
+//
+//  RangeDescriptor
+//
+//     Each of the non-overlapping character ranges gets one of these descriptors.
+//     All of them are strung together in a linked list, which is kept in order
+//     (by character)
+//
+class RangeDescriptor : public UMemory {
+public:
+    UChar32            fStartChar {};            // Start of range, unicode 32 bit value.
+    UChar32            fEndChar {};              // End of range, unicode 32 bit value.
+    int32_t            fNum {0};                 // runtime-mapped input value for this range.
+    bool               fIncludesDict {false};    // True if the range includes $dictionary.
+    bool               fFirstInGroup {false};    // True if first range in a group with the same fNum.
+    UVector           *fIncludesSets {nullptr};  // vector of the the original
+                                                 //   Unicode sets that include this range.
+                                                 //    (Contains ptrs to uset nodes)
+    RangeDescriptor   *fNext {nullptr};          // Next RangeDescriptor in the linked list.
+
+    RangeDescriptor(UErrorCode &status);
+    RangeDescriptor(const RangeDescriptor &other, UErrorCode &status);
+    ~RangeDescriptor();
+    void split(UChar32 where, UErrorCode &status);   // Spit this range in two at "where", with
+                                        //   where appearing in the second (higher) part.
+    bool isDictionaryRange();           // Check whether this range appears as part of
+                                        //   the Unicode set named "dictionary"
+
+    RangeDescriptor(const RangeDescriptor &other) = delete; // forbid default copying of this class
+    RangeDescriptor &operator=(const RangeDescriptor &other) = delete; // forbid assigning of this class
+};
+
+
+//
+//  RBBISetBuilder   Handles processing of Unicode Sets from RBBI rules.
+//
+//      Starting with the rules parse tree from the scanner,
+//
+//                   -  Enumerate the set of UnicodeSets that are referenced
+//                      by the RBBI rules.
+//                   -  compute a derived set of non-overlapping UnicodeSets
+//                      that will correspond to columns in the state table for
+//                      the RBBI execution engine.
+//                   -  construct the trie table that maps input characters
+//                      to set numbers in the non-overlapping set of sets.
+//
+
+
+class RBBISetBuilder : public UMemory {
+public:
+    RBBISetBuilder(RBBIRuleBuilder *rb);
+    ~RBBISetBuilder();
+
+    void     buildRanges();
+    void     buildTrie();
+    void     addValToSets(UVector *sets,      uint32_t val);
+    void     addValToSet (RBBINode *usetNode, uint32_t val);
+    int32_t  getNumCharCategories() const;   // CharCategories are the same as input symbol set to the
+                                             //    runtime state machine, which are the same as
+                                             //    columns in the DFA state table
+    int32_t  getDictCategoriesStart() const; // First char category that includes $dictionary, or
+                                             // last category + 1 if there are no dictionary categories.
+    int32_t  getTrieSize() /*const*/;        // Size in bytes of the serialized Trie.
+    void     serializeTrie(uint8_t *where);  // write out the serialized Trie.
+    UChar32  getFirstChar(int32_t  val) const;
+    UBool    sawBOF() const;                 // Indicate whether any references to the {bof} pseudo
+                                             //   character were encountered.
+    /**
+     * Merge two character categories that have been identified as having equivalent behavior.
+     * The ranges belonging to the second category (table column) will be added to the first.
+     * @param categories the pair of categories to be merged.
+     */
+    void     mergeCategories(IntPair categories);
+
+#ifdef RBBI_DEBUG
+    void     printSets();
+    void     printRanges();
+    void     printRangeGroups();
+#else
+    #define printSets()
+    #define printRanges()
+    #define printRangeGroups()
+#endif
+
+private:
+    RBBIRuleBuilder       *fRB;             // The RBBI Rule Compiler that owns us.
+    UErrorCode            *fStatus;
+
+    RangeDescriptor       *fRangeList;      // Head of the linked list of RangeDescriptors
+
+    UMutableCPTrie        *fMutableTrie;    // The mapping TRIE that is the end result of processing
+    UCPTrie               *fTrie;           //  the Unicode Sets.
+    uint32_t               fTrieSize;
+
+    // Number of range groups, which are groups of ranges that are in the same original UnicodeSets.
+    int32_t               fGroupCount;
+
+    // The number of the first dictionary char category.
+    // If there are no Dictionary categories, set to the last category + 1.
+    int32_t               fDictCategoriesStart;
+
+    UBool                 fSawBOF;
+
+    RBBISetBuilder(const RBBISetBuilder &other); // forbid copying of this class
+    RBBISetBuilder &operator=(const RBBISetBuilder &other); // forbid copying of this class
+};
+
+
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif
diff --git a/thirdparty/icu4c/common/rbbistbl.cpp b/thirdparty/icu4c/common/rbbistbl.cpp
new file mode 100644
index 0000000000..5303f76096
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbistbl.cpp
@@ -0,0 +1,270 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  file:  rbbistbl.cpp    Implementation of the ICU RBBISymbolTable class
+//
+/*
+***************************************************************************
+*   Copyright (C) 2002-2014 International Business Machines Corporation
+*   and others. All rights reserved.
+***************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/unistr.h"
+#include "unicode/uniset.h"
+#include "unicode/uchar.h"
+#include "unicode/parsepos.h"
+
+#include "cstr.h"
+#include "rbbinode.h"
+#include "rbbirb.h"
+#include "umutex.h"
+
+
+//
+//  RBBISymbolTableEntry_deleter    Used by the UHashTable to delete the contents
+//                                  when the hash table is deleted.
+//
+U_CDECL_BEGIN
+static void U_CALLCONV RBBISymbolTableEntry_deleter(void *p) {
+    icu::RBBISymbolTableEntry *px = (icu::RBBISymbolTableEntry *)p;
+    delete px;
+}
+U_CDECL_END
+
+
+
+U_NAMESPACE_BEGIN
+
+RBBISymbolTable::RBBISymbolTable(RBBIRuleScanner *rs, const UnicodeString &rules, UErrorCode &status)
+    :fRules(rules), fRuleScanner(rs), ffffString(UChar(0xffff))
+{
+    fHashTable       = NULL;
+    fCachedSetLookup = NULL;
+    
+    fHashTable = uhash_open(uhash_hashUnicodeString, uhash_compareUnicodeString, NULL, &status);
+    // uhash_open checks status
+    if (U_FAILURE(status)) {
+        return;
+    }
+    uhash_setValueDeleter(fHashTable, RBBISymbolTableEntry_deleter);
+}
+
+
+
+RBBISymbolTable::~RBBISymbolTable()
+{
+    uhash_close(fHashTable);
+}
+
+
+//
+//  RBBISymbolTable::lookup       This function from the abstract symbol table inteface
+//                                looks up a variable name and returns a UnicodeString
+//                                containing the substitution text.
+//
+//                                The variable name does NOT include the leading $.
+//
+const UnicodeString  *RBBISymbolTable::lookup(const UnicodeString& s) const
+{
+    RBBISymbolTableEntry  *el;
+    RBBINode              *varRefNode;
+    RBBINode              *exprNode;
+    RBBINode              *usetNode;
+    const UnicodeString   *retString;
+    RBBISymbolTable       *This = (RBBISymbolTable *)this;   // cast off const
+
+    el = (RBBISymbolTableEntry *)uhash_get(fHashTable, &s);
+    if (el == NULL) {
+        return NULL;
+    }
+
+    varRefNode = el->val;
+    exprNode   = varRefNode->fLeftChild;     // Root node of expression for variable
+    if (exprNode->fType == RBBINode::setRef) {
+        // The $variable refers to a single UnicodeSet
+        //   return the ffffString, which will subsequently be interpreted as a
+        //   stand-in character for the set by RBBISymbolTable::lookupMatcher()
+        usetNode = exprNode->fLeftChild;
+        This->fCachedSetLookup = usetNode->fInputSet;
+        retString = &ffffString;
+    }
+    else
+    {
+        // The variable refers to something other than just a set.
+        // return the original source string for the expression
+        retString = &exprNode->fText;
+        This->fCachedSetLookup = NULL;
+    }
+    return retString;
+}
+
+
+
+//
+//  RBBISymbolTable::lookupMatcher   This function from the abstract symbol table
+//                                   interface maps a single stand-in character to a
+//                                   pointer to a Unicode Set.   The Unicode Set code uses this
+//                                   mechanism to get all references to the same $variable
+//                                   name to refer to a single common Unicode Set instance.
+//
+//    This implementation cheats a little, and does not maintain a map of stand-in chars
+//    to sets.  Instead, it takes advantage of the fact that  the UnicodeSet
+//    constructor will always call this function right after calling lookup(),
+//    and we just need to remember what set to return between these two calls.
+const UnicodeFunctor *RBBISymbolTable::lookupMatcher(UChar32 ch) const
+{
+    UnicodeSet *retVal = NULL;
+    RBBISymbolTable *This = (RBBISymbolTable *)this;   // cast off const
+    if (ch == 0xffff) {
+        retVal = fCachedSetLookup;
+        This->fCachedSetLookup = 0;
+    }
+    return retVal;
+}
+
+//
+// RBBISymbolTable::parseReference   This function from the abstract symbol table interface
+//                                   looks for a $variable name in the source text.
+//                                   It does not look it up, only scans for it.
+//                                   It is used by the UnicodeSet parser.
+//
+//                                   This implementation is lifted pretty much verbatim
+//                                   from the rules based transliterator implementation.
+//                                   I didn't see an obvious way of sharing it.
+//
+UnicodeString   RBBISymbolTable::parseReference(const UnicodeString& text,
+                                                ParsePosition& pos, int32_t limit) const
+{
+    int32_t start = pos.getIndex();
+    int32_t i = start;
+    UnicodeString result;
+    while (i < limit) {
+        UChar c = text.charAt(i);
+        if ((i==start && !u_isIDStart(c)) || !u_isIDPart(c)) {
+            break;
+        }
+        ++i;
+    }
+    if (i == start) { // No valid name chars
+        return result; // Indicate failure with empty string
+    }
+    pos.setIndex(i);
+    text.extractBetween(start, i, result);
+    return result;
+}
+
+
+
+//
+// RBBISymbolTable::lookupNode      Given a key (a variable name), return the
+//                                  corresponding RBBI Node.  If there is no entry
+//                                  in the table for this name, return NULL.
+//
+RBBINode       *RBBISymbolTable::lookupNode(const UnicodeString &key) const{
+
+    RBBINode             *retNode = NULL;
+    RBBISymbolTableEntry *el;
+
+    el = (RBBISymbolTableEntry *)uhash_get(fHashTable, &key);
+    if (el != NULL) {
+        retNode = el->val;
+    }
+    return retNode;
+}
+
+
+//
+//    RBBISymbolTable::addEntry     Add a new entry to the symbol table.
+//                                  Indicate an error if the name already exists -
+//                                    this will only occur in the case of duplicate
+//                                    variable assignments.
+//
+void            RBBISymbolTable::addEntry  (const UnicodeString &key, RBBINode *val, UErrorCode &err) {
+    RBBISymbolTableEntry *e;
+    /* test for buffer overflows */
+    if (U_FAILURE(err)) {
+        return;
+    }
+    e = (RBBISymbolTableEntry *)uhash_get(fHashTable, &key);
+    if (e != NULL) {
+        err = U_BRK_VARIABLE_REDFINITION;
+        return;
+    }
+
+    e = new RBBISymbolTableEntry;
+    if (e == NULL) {
+        err = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    e->key = key;
+    e->val = val;
+    uhash_put( fHashTable, &e->key, e, &err);
+}
+
+
+RBBISymbolTableEntry::RBBISymbolTableEntry() : UMemory(), key(), val(NULL) {}
+
+RBBISymbolTableEntry::~RBBISymbolTableEntry() {
+    // The "val" of a symbol table entry is a variable reference node.
+    // The l. child of the val is the rhs expression from the assignment.
+    // Unlike other node types, children of variable reference nodes are not
+    //    automatically recursively deleted.  We do it manually here.
+    delete val->fLeftChild;
+    val->fLeftChild = NULL;
+
+    delete  val;
+
+    // Note: the key UnicodeString is destructed by virtue of being in the object by value.
+}
+
+
+//
+//  RBBISymbolTable::print    Debugging function, dump out the symbol table contents.
+//
+#ifdef RBBI_DEBUG
+void RBBISymbolTable::rbbiSymtablePrint() const {
+    RBBIDebugPrintf("Variable Definitions Symbol Table\n"
+           "Name                  Node         serial  String Val\n"
+           "-------------------------------------------------------------------\n");
+
+    int32_t pos = UHASH_FIRST;
+    const UHashElement  *e   = NULL;
+    for (;;) {
+        e = uhash_nextElement(fHashTable,  &pos);
+        if (e == NULL ) {
+            break;
+        }
+        RBBISymbolTableEntry  *s   = (RBBISymbolTableEntry *)e->value.pointer;
+
+        RBBIDebugPrintf("%-19s   %8p %7d ", CStr(s->key)(), (void *)s->val, s->val->fSerialNum);
+        RBBIDebugPrintf(" %s\n", CStr(s->val->fLeftChild->fText)());
+    }
+
+    RBBIDebugPrintf("\nParsed Variable Definitions\n");
+    pos = -1;
+    for (;;) {
+        e = uhash_nextElement(fHashTable,  &pos);
+        if (e == NULL ) {
+            break;
+        }
+        RBBISymbolTableEntry  *s   = (RBBISymbolTableEntry *)e->value.pointer;
+        RBBIDebugPrintf("%s\n", CStr(s->key)());
+        RBBINode::printTree(s->val, TRUE);
+        RBBINode::printTree(s->val->fLeftChild, FALSE);
+        RBBIDebugPrintf("\n");
+    }
+}
+#endif
+
+
+
+
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbitblb.cpp b/thirdparty/icu4c/common/rbbitblb.cpp
new file mode 100644
index 0000000000..bcbdab9227
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbitblb.cpp
@@ -0,0 +1,1793 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+//
+//  rbbitblb.cpp
+//
+
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/unistr.h"
+#include "rbbitblb.h"
+#include "rbbirb.h"
+#include "rbbiscan.h"
+#include "rbbisetb.h"
+#include "rbbidata.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "uvectr32.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+const int32_t kMaxStateFor8BitsTable = 255;
+
+RBBITableBuilder::RBBITableBuilder(RBBIRuleBuilder *rb, RBBINode **rootNode, UErrorCode &status) :
+        fRB(rb),
+        fTree(*rootNode),
+        fStatus(&status),
+        fDStates(nullptr),
+        fSafeTable(nullptr) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    // fDStates is UVector<RBBIStateDescriptor *>
+    fDStates = new UVector(status);
+    if (U_SUCCESS(status) && fDStates == nullptr ) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+
+
+RBBITableBuilder::~RBBITableBuilder() {
+    int i;
+    for (i=0; i<fDStates->size(); i++) {
+        delete (RBBIStateDescriptor *)fDStates->elementAt(i);
+    }
+    delete fDStates;
+    delete fSafeTable;
+    delete fLookAheadRuleMap;
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//   RBBITableBuilder::buildForwardTable  -  This is the main function for building
+//                               the DFA state transition table from the RBBI rules parse tree.
+//
+//-----------------------------------------------------------------------------
+void  RBBITableBuilder::buildForwardTable() {
+
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    // If there were no rules, just return.  This situation can easily arise
+    //   for the reverse rules.
+    if (fTree==NULL) {
+        return;
+    }
+
+    //
+    // Walk through the tree, replacing any references to $variables with a copy of the
+    //   parse tree for the substition expression.
+    //
+    fTree = fTree->flattenVariables();
+#ifdef RBBI_DEBUG
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "ftree")) {
+        RBBIDebugPuts("\nParse tree after flattening variable references.");
+        RBBINode::printTree(fTree, TRUE);
+    }
+#endif
+
+    //
+    // If the rules contained any references to {bof} 
+    //   add a {bof} <cat> <former root of tree> to the
+    //   tree.  Means that all matches must start out with the 
+    //   {bof} fake character.
+    // 
+    if (fRB->fSetBuilder->sawBOF()) {
+        RBBINode *bofTop    = new RBBINode(RBBINode::opCat);
+        RBBINode *bofLeaf   = new RBBINode(RBBINode::leafChar);
+        // Delete and exit if memory allocation failed.
+        if (bofTop == NULL || bofLeaf == NULL) {
+            *fStatus = U_MEMORY_ALLOCATION_ERROR;
+            delete bofTop;
+            delete bofLeaf;
+            return;
+        }
+        bofTop->fLeftChild  = bofLeaf;
+        bofTop->fRightChild = fTree;
+        bofLeaf->fParent    = bofTop;
+        bofLeaf->fVal       = 2;      // Reserved value for {bof}.
+        fTree               = bofTop;
+    }
+
+    //
+    // Add a unique right-end marker to the expression.
+    //   Appears as a cat-node, left child being the original tree,
+    //   right child being the end marker.
+    //
+    RBBINode *cn = new RBBINode(RBBINode::opCat);
+    // Exit if memory allocation failed.
+    if (cn == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    cn->fLeftChild = fTree;
+    fTree->fParent = cn;
+    RBBINode *endMarkerNode = cn->fRightChild = new RBBINode(RBBINode::endMark);
+    // Delete and exit if memory allocation failed.
+    if (cn->fRightChild == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        delete cn;
+        return;
+    }
+    cn->fRightChild->fParent = cn;
+    fTree = cn;
+
+    //
+    //  Replace all references to UnicodeSets with the tree for the equivalent
+    //      expression.
+    //
+    fTree->flattenSets();
+#ifdef RBBI_DEBUG
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "stree")) {
+        RBBIDebugPuts("\nParse tree after flattening Unicode Set references.");
+        RBBINode::printTree(fTree, TRUE);
+    }
+#endif
+
+
+    //
+    // calculate the functions nullable, firstpos, lastpos and followpos on
+    // nodes in the parse tree.
+    //    See the alogrithm description in Aho.
+    //    Understanding how this works by looking at the code alone will be
+    //       nearly impossible.
+    //
+    calcNullable(fTree);
+    calcFirstPos(fTree);
+    calcLastPos(fTree);
+    calcFollowPos(fTree);
+    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "pos")) {
+        RBBIDebugPuts("\n");
+        printPosSets(fTree);
+    }
+
+    //
+    //  For "chained" rules, modify the followPos sets
+    //
+    if (fRB->fChainRules) {
+        calcChainedFollowPos(fTree, endMarkerNode);
+    }
+
+    //
+    //  BOF (start of input) test fixup.
+    //
+    if (fRB->fSetBuilder->sawBOF()) {
+        bofFixup();
+    }
+
+    //
+    // Build the DFA state transition tables.
+    //
+    buildStateTable();
+    mapLookAheadRules();
+    flagAcceptingStates();
+    flagLookAheadStates();
+    flagTaggedStates();
+
+    //
+    // Update the global table of rule status {tag} values
+    // The rule builder has a global vector of status values that are common
+    //    for all tables.  Merge the ones from this table into the global set.
+    //
+    mergeRuleStatusVals();
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   calcNullable.    Impossible to explain succinctly.  See Aho, section 3.9
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::calcNullable(RBBINode *n) {
+    if (n == NULL) {
+        return;
+    }
+    if (n->fType == RBBINode::setRef ||
+        n->fType == RBBINode::endMark ) {
+        // These are non-empty leaf node types.
+        n->fNullable = FALSE;
+        return;
+    }
+
+    if (n->fType == RBBINode::lookAhead || n->fType == RBBINode::tag) {
+        // Lookahead marker node.  It's a leaf, so no recursion on children.
+        // It's nullable because it does not match any literal text from the input stream.
+        n->fNullable = TRUE;
+        return;
+    }
+
+
+    // The node is not a leaf.
+    //  Calculate nullable on its children.
+    calcNullable(n->fLeftChild);
+    calcNullable(n->fRightChild);
+
+    // Apply functions from table 3.40 in Aho
+    if (n->fType == RBBINode::opOr) {
+        n->fNullable = n->fLeftChild->fNullable || n->fRightChild->fNullable;
+    }
+    else if (n->fType == RBBINode::opCat) {
+        n->fNullable = n->fLeftChild->fNullable && n->fRightChild->fNullable;
+    }
+    else if (n->fType == RBBINode::opStar || n->fType == RBBINode::opQuestion) {
+        n->fNullable = TRUE;
+    }
+    else {
+        n->fNullable = FALSE;
+    }
+}
+
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   calcFirstPos.    Impossible to explain succinctly.  See Aho, section 3.9
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::calcFirstPos(RBBINode *n) {
+    if (n == NULL) {
+        return;
+    }
+    if (n->fType == RBBINode::leafChar  ||
+        n->fType == RBBINode::endMark   ||
+        n->fType == RBBINode::lookAhead ||
+        n->fType == RBBINode::tag) {
+        // These are non-empty leaf node types.
+        // Note: In order to maintain the sort invariant on the set,
+        // this function should only be called on a node whose set is
+        // empty to start with.
+        n->fFirstPosSet->addElement(n, *fStatus);
+        return;
+    }
+
+    // The node is not a leaf.
+    //  Calculate firstPos on its children.
+    calcFirstPos(n->fLeftChild);
+    calcFirstPos(n->fRightChild);
+
+    // Apply functions from table 3.40 in Aho
+    if (n->fType == RBBINode::opOr) {
+        setAdd(n->fFirstPosSet, n->fLeftChild->fFirstPosSet);
+        setAdd(n->fFirstPosSet, n->fRightChild->fFirstPosSet);
+    }
+    else if (n->fType == RBBINode::opCat) {
+        setAdd(n->fFirstPosSet, n->fLeftChild->fFirstPosSet);
+        if (n->fLeftChild->fNullable) {
+            setAdd(n->fFirstPosSet, n->fRightChild->fFirstPosSet);
+        }
+    }
+    else if (n->fType == RBBINode::opStar ||
+             n->fType == RBBINode::opQuestion ||
+             n->fType == RBBINode::opPlus) {
+        setAdd(n->fFirstPosSet, n->fLeftChild->fFirstPosSet);
+    }
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   calcLastPos.    Impossible to explain succinctly.  See Aho, section 3.9
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::calcLastPos(RBBINode *n) {
+    if (n == NULL) {
+        return;
+    }
+    if (n->fType == RBBINode::leafChar  ||
+        n->fType == RBBINode::endMark   ||
+        n->fType == RBBINode::lookAhead ||
+        n->fType == RBBINode::tag) {
+        // These are non-empty leaf node types.
+        // Note: In order to maintain the sort invariant on the set,
+        // this function should only be called on a node whose set is
+        // empty to start with.
+        n->fLastPosSet->addElement(n, *fStatus);
+        return;
+    }
+
+    // The node is not a leaf.
+    //  Calculate lastPos on its children.
+    calcLastPos(n->fLeftChild);
+    calcLastPos(n->fRightChild);
+
+    // Apply functions from table 3.40 in Aho
+    if (n->fType == RBBINode::opOr) {
+        setAdd(n->fLastPosSet, n->fLeftChild->fLastPosSet);
+        setAdd(n->fLastPosSet, n->fRightChild->fLastPosSet);
+    }
+    else if (n->fType == RBBINode::opCat) {
+        setAdd(n->fLastPosSet, n->fRightChild->fLastPosSet);
+        if (n->fRightChild->fNullable) {
+            setAdd(n->fLastPosSet, n->fLeftChild->fLastPosSet);
+        }
+    }
+    else if (n->fType == RBBINode::opStar     ||
+             n->fType == RBBINode::opQuestion ||
+             n->fType == RBBINode::opPlus) {
+        setAdd(n->fLastPosSet, n->fLeftChild->fLastPosSet);
+    }
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   calcFollowPos.    Impossible to explain succinctly.  See Aho, section 3.9
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::calcFollowPos(RBBINode *n) {
+    if (n == NULL ||
+        n->fType == RBBINode::leafChar ||
+        n->fType == RBBINode::endMark) {
+        return;
+    }
+
+    calcFollowPos(n->fLeftChild);
+    calcFollowPos(n->fRightChild);
+
+    // Aho rule #1
+    if (n->fType == RBBINode::opCat) {
+        RBBINode *i;   // is 'i' in Aho's description
+        uint32_t     ix;
+
+        UVector *LastPosOfLeftChild = n->fLeftChild->fLastPosSet;
+
+        for (ix=0; ix<(uint32_t)LastPosOfLeftChild->size(); ix++) {
+            i = (RBBINode *)LastPosOfLeftChild->elementAt(ix);
+            setAdd(i->fFollowPos, n->fRightChild->fFirstPosSet);
+        }
+    }
+
+    // Aho rule #2
+    if (n->fType == RBBINode::opStar ||
+        n->fType == RBBINode::opPlus) {
+        RBBINode   *i;  // again, n and i are the names from Aho's description.
+        uint32_t    ix;
+
+        for (ix=0; ix<(uint32_t)n->fLastPosSet->size(); ix++) {
+            i = (RBBINode *)n->fLastPosSet->elementAt(ix);
+            setAdd(i->fFollowPos, n->fFirstPosSet);
+        }
+    }
+
+
+
+}
+
+//-----------------------------------------------------------------------------
+//
+//    addRuleRootNodes    Recursively walk a parse tree, adding all nodes flagged
+//                        as roots of a rule to a destination vector.
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::addRuleRootNodes(UVector *dest, RBBINode *node) {
+    if (node == NULL || U_FAILURE(*fStatus)) {
+        return;
+    }
+    if (node->fRuleRoot) {
+        dest->addElement(node, *fStatus);
+        // Note: rules cannot nest. If we found a rule start node,
+        //       no child node can also be a start node.
+        return;
+    }
+    addRuleRootNodes(dest, node->fLeftChild);
+    addRuleRootNodes(dest, node->fRightChild);
+}
+
+//-----------------------------------------------------------------------------
+//
+//   calcChainedFollowPos.    Modify the previously calculated followPos sets
+//                            to implement rule chaining.  NOT described by Aho
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::calcChainedFollowPos(RBBINode *tree, RBBINode *endMarkNode) {
+
+    UVector         leafNodes(*fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    // get a list all leaf nodes
+    tree->findNodes(&leafNodes, RBBINode::leafChar, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    // Collect all leaf nodes that can start matches for rules
+    // with inbound chaining enabled, which is the union of the 
+    // firstPosition sets from each of the rule root nodes.
+    
+    UVector ruleRootNodes(*fStatus);
+    addRuleRootNodes(&ruleRootNodes, tree);
+
+    UVector matchStartNodes(*fStatus);
+    for (int j=0; j<ruleRootNodes.size(); ++j) {
+        RBBINode *node = static_cast<RBBINode *>(ruleRootNodes.elementAt(j));
+        if (node->fChainIn) {
+            setAdd(&matchStartNodes, node->fFirstPosSet);
+        }
+    }
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    int32_t  endNodeIx;
+    int32_t  startNodeIx;
+
+    for (endNodeIx=0; endNodeIx<leafNodes.size(); endNodeIx++) {
+        RBBINode *endNode   = (RBBINode *)leafNodes.elementAt(endNodeIx);
+
+        // Identify leaf nodes that correspond to overall rule match positions.
+        // These include the endMarkNode in their followPos sets.
+        //
+        // Note: do not consider other end marker nodes, those that are added to
+        //       look-ahead rules. These can't chain; a match immediately stops
+        //       further matching. This leaves exactly one end marker node, the one
+        //       at the end of the complete tree.
+
+        if (!endNode->fFollowPos->contains(endMarkNode)) {
+            continue;
+        }
+
+        // We've got a node that can end a match.
+
+        // !!LBCMNoChain implementation:  If this node's val correspond to
+        // the Line Break $CM char class, don't chain from it.
+        // TODO:  Remove this. !!LBCMNoChain is deprecated, and is not used
+        //        by any of the standard ICU rules.
+        if (fRB->fLBCMNoChain) {
+            UChar32 c = this->fRB->fSetBuilder->getFirstChar(endNode->fVal);
+            if (c != -1) {
+                // c == -1 occurs with sets containing only the {eof} marker string.
+                ULineBreak cLBProp = (ULineBreak)u_getIntPropertyValue(c, UCHAR_LINE_BREAK);
+                if (cLBProp == U_LB_COMBINING_MARK) {
+                    continue;
+                }
+            }
+        }
+
+        // Now iterate over the nodes that can start a match, looking for ones
+        //   with the same char class as our ending node.
+        RBBINode *startNode;
+        for (startNodeIx = 0; startNodeIx<matchStartNodes.size(); startNodeIx++) {
+            startNode = (RBBINode *)matchStartNodes.elementAt(startNodeIx);
+            if (startNode->fType != RBBINode::leafChar) {
+                continue;
+            }
+
+            if (endNode->fVal == startNode->fVal) {
+                // The end val (character class) of one possible match is the
+                //   same as the start of another.
+
+                // Add all nodes from the followPos of the start node to the
+                //  followPos set of the end node, which will have the effect of
+                //  letting matches transition from a match state at endNode
+                //  to the second char of a match starting with startNode.
+                setAdd(endNode->fFollowPos, startNode->fFollowPos);
+            }
+        }
+    }
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//   bofFixup.    Fixup for state tables that include {bof} beginning of input testing.
+//                Do an swizzle similar to chaining, modifying the followPos set of
+//                the bofNode to include the followPos nodes from other {bot} nodes
+//                scattered through the tree.
+//
+//                This function has much in common with calcChainedFollowPos().
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::bofFixup() {
+
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    //   The parse tree looks like this ...
+    //         fTree root  --->       <cat>
+    //                               /     \       .
+    //                            <cat>   <#end node>
+    //                           /     \  .
+    //                     <bofNode>   rest
+    //                               of tree
+    //
+    //    We will be adding things to the followPos set of the <bofNode>
+    //
+    RBBINode  *bofNode = fTree->fLeftChild->fLeftChild;
+    U_ASSERT(bofNode->fType == RBBINode::leafChar);
+    U_ASSERT(bofNode->fVal == 2);
+
+    // Get all nodes that can be the start a match of the user-written rules
+    //  (excluding the fake bofNode)
+    //  We want the nodes that can start a match in the
+    //     part labeled "rest of tree"
+    // 
+    UVector *matchStartNodes = fTree->fLeftChild->fRightChild->fFirstPosSet;
+
+    RBBINode *startNode;
+    int       startNodeIx;
+    for (startNodeIx = 0; startNodeIx<matchStartNodes->size(); startNodeIx++) {
+        startNode = (RBBINode *)matchStartNodes->elementAt(startNodeIx);
+        if (startNode->fType != RBBINode::leafChar) {
+            continue;
+        }
+
+        if (startNode->fVal == bofNode->fVal) {
+            //  We found a leaf node corresponding to a {bof} that was
+            //    explicitly written into a rule.
+            //  Add everything from the followPos set of this node to the
+            //    followPos set of the fake bofNode at the start of the tree.
+            //  
+            setAdd(bofNode->fFollowPos, startNode->fFollowPos);
+        }
+    }
+}
+
+//-----------------------------------------------------------------------------
+//
+//   buildStateTable()    Determine the set of runtime DFA states and the
+//                        transition tables for these states, by the algorithm
+//                        of fig. 3.44 in Aho.
+//
+//                        Most of the comments are quotes of Aho's psuedo-code.
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::buildStateTable() {
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    RBBIStateDescriptor *failState;
+    // Set it to NULL to avoid uninitialized warning
+    RBBIStateDescriptor *initialState = NULL; 
+    //
+    // Add a dummy state 0 - the stop state.  Not from Aho.
+    int      lastInputSymbol = fRB->fSetBuilder->getNumCharCategories() - 1;
+    failState = new RBBIStateDescriptor(lastInputSymbol, fStatus);
+    if (failState == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        goto ExitBuildSTdeleteall;
+    }
+    failState->fPositions = new UVector(*fStatus);
+    if (failState->fPositions == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (failState->fPositions == NULL || U_FAILURE(*fStatus)) {
+        goto ExitBuildSTdeleteall;
+    }
+    fDStates->addElement(failState, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        goto ExitBuildSTdeleteall;
+    }
+
+    // initially, the only unmarked state in Dstates is firstpos(root),
+    //       where toot is the root of the syntax tree for (r)#;
+    initialState = new RBBIStateDescriptor(lastInputSymbol, fStatus);
+    if (initialState == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_FAILURE(*fStatus)) {
+        goto ExitBuildSTdeleteall;
+    }
+    initialState->fPositions = new UVector(*fStatus);
+    if (initialState->fPositions == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_FAILURE(*fStatus)) {
+        goto ExitBuildSTdeleteall;
+    }
+    setAdd(initialState->fPositions, fTree->fFirstPosSet);
+    fDStates->addElement(initialState, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        goto ExitBuildSTdeleteall;
+    }
+
+    // while there is an unmarked state T in Dstates do begin
+    for (;;) {
+        RBBIStateDescriptor *T = NULL;
+        int32_t              tx;
+        for (tx=1; tx<fDStates->size(); tx++) {
+            RBBIStateDescriptor *temp;
+            temp = (RBBIStateDescriptor *)fDStates->elementAt(tx);
+            if (temp->fMarked == FALSE) {
+                T = temp;
+                break;
+            }
+        }
+        if (T == NULL) {
+            break;
+        }
+
+        // mark T;
+        T->fMarked = TRUE;
+
+        // for each input symbol a do begin
+        int32_t  a;
+        for (a = 1; a<=lastInputSymbol; a++) {
+            // let U be the set of positions that are in followpos(p)
+            //    for some position p in T
+            //    such that the symbol at position p is a;
+            UVector    *U = NULL;
+            RBBINode   *p;
+            int32_t     px;
+            for (px=0; px<T->fPositions->size(); px++) {
+                p = (RBBINode *)T->fPositions->elementAt(px);
+                if ((p->fType == RBBINode::leafChar) &&  (p->fVal == a)) {
+                    if (U == NULL) {
+                        U = new UVector(*fStatus);
+                        if (U == NULL) {
+                        	*fStatus = U_MEMORY_ALLOCATION_ERROR;
+                        	goto ExitBuildSTdeleteall;
+                        }
+                    }
+                    setAdd(U, p->fFollowPos);
+                }
+            }
+
+            // if U is not empty and not in DStates then
+            int32_t  ux = 0;
+            UBool    UinDstates = FALSE;
+            if (U != NULL) {
+                U_ASSERT(U->size() > 0);
+                int  ix;
+                for (ix=0; ix<fDStates->size(); ix++) {
+                    RBBIStateDescriptor *temp2;
+                    temp2 = (RBBIStateDescriptor *)fDStates->elementAt(ix);
+                    if (setEquals(U, temp2->fPositions)) {
+                        delete U;
+                        U  = temp2->fPositions;
+                        ux = ix;
+                        UinDstates = TRUE;
+                        break;
+                    }
+                }
+
+                // Add U as an unmarked state to Dstates
+                if (!UinDstates)
+                {
+                    RBBIStateDescriptor *newState = new RBBIStateDescriptor(lastInputSymbol, fStatus);
+                    if (newState == NULL) {
+                    	*fStatus = U_MEMORY_ALLOCATION_ERROR;
+                    }
+                    if (U_FAILURE(*fStatus)) {
+                        goto ExitBuildSTdeleteall;
+                    }
+                    newState->fPositions = U;
+                    fDStates->addElement(newState, *fStatus);
+                    if (U_FAILURE(*fStatus)) {
+                        return;
+                    }
+                    ux = fDStates->size()-1;
+                }
+
+                // Dtran[T, a] := U;
+                T->fDtran->setElementAt(ux, a);
+            }
+        }
+    }
+    return;
+    // delete local pointers only if error occured.
+ExitBuildSTdeleteall:
+    delete initialState;
+    delete failState;
+}
+
+
+/**
+ * mapLookAheadRules
+ *
+ */
+void RBBITableBuilder::mapLookAheadRules() {
+    fLookAheadRuleMap =  new UVector32(fRB->fScanner->numRules() + 1, *fStatus);
+    if (fLookAheadRuleMap == nullptr) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    fLookAheadRuleMap->setSize(fRB->fScanner->numRules() + 1);
+
+    for (int32_t n=0; n<fDStates->size(); n++) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+        int32_t laSlotForState = 0;
+
+        // Establish the look-ahead slot for this state, if the state covers
+        // any look-ahead nodes - corresponding to the '/' in look-ahead rules.
+
+        // If any of the look-ahead nodes already have a slot assigned, use it,
+        // otherwise assign a new one.
+
+        bool sawLookAheadNode = false;
+        for (int32_t ipos=0; ipos<sd->fPositions->size(); ++ipos) {
+            RBBINode *node = static_cast<RBBINode *>(sd->fPositions->elementAt(ipos));
+            if (node->fType != RBBINode::NodeType::lookAhead) {
+                continue;
+            }
+            sawLookAheadNode = true;
+            int32_t ruleNum = node->fVal;     // Set when rule was originally parsed.
+            U_ASSERT(ruleNum < fLookAheadRuleMap->size());
+            U_ASSERT(ruleNum > 0);
+            int32_t laSlot = fLookAheadRuleMap->elementAti(ruleNum);
+            if (laSlot != 0) {
+                if (laSlotForState == 0) {
+                    laSlotForState = laSlot;
+                } else {
+                    // TODO: figure out if this can fail, change to setting an error code if so.
+                    U_ASSERT(laSlot == laSlotForState);
+                }
+            }
+        }
+        if (!sawLookAheadNode) {
+            continue;
+        }
+
+        if (laSlotForState == 0) {
+            laSlotForState = ++fLASlotsInUse;
+        }
+
+        // For each look ahead node covered by this state,
+        // set the mapping from the node's rule number to the look ahead slot.
+        // There can be multiple nodes/rule numbers going to the same la slot.
+
+        for (int32_t ipos=0; ipos<sd->fPositions->size(); ++ipos) {
+            RBBINode *node = static_cast<RBBINode *>(sd->fPositions->elementAt(ipos));
+            if (node->fType != RBBINode::NodeType::lookAhead) {
+                continue;
+            }
+            int32_t ruleNum = node->fVal;     // Set when rule was originally parsed.
+            int32_t existingVal = fLookAheadRuleMap->elementAti(ruleNum);
+            (void)existingVal;
+            U_ASSERT(existingVal == 0 || existingVal == laSlotForState);
+            fLookAheadRuleMap->setElementAt(laSlotForState, ruleNum);
+        }
+    }
+
+}
+
+//-----------------------------------------------------------------------------
+//
+//   flagAcceptingStates    Identify accepting states.
+//                          First get a list of all of the end marker nodes.
+//                          Then, for each state s,
+//                              if s contains one of the end marker nodes in its list of tree positions then
+//                                  s is an accepting state.
+//
+//-----------------------------------------------------------------------------
+void     RBBITableBuilder::flagAcceptingStates() {
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    UVector     endMarkerNodes(*fStatus);
+    RBBINode    *endMarker;
+    int32_t     i;
+    int32_t     n;
+
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    fTree->findNodes(&endMarkerNodes, RBBINode::endMark, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+
+    for (i=0; i<endMarkerNodes.size(); i++) {
+        endMarker = (RBBINode *)endMarkerNodes.elementAt(i);
+        for (n=0; n<fDStates->size(); n++) {
+            RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+            if (sd->fPositions->indexOf(endMarker) >= 0) {
+                // Any non-zero value for fAccepting means this is an accepting node.
+                // The value is what will be returned to the user as the break status.
+                // If no other value was specified, force it to ACCEPTING_UNCONDITIONAL (1).
+
+                if (sd->fAccepting==0) {
+                    // State hasn't been marked as accepting yet.  Do it now.
+                    sd->fAccepting = fLookAheadRuleMap->elementAti(endMarker->fVal);
+                    if (sd->fAccepting == 0) {
+                        sd->fAccepting = ACCEPTING_UNCONDITIONAL;
+                    }
+                }
+                if (sd->fAccepting==ACCEPTING_UNCONDITIONAL && endMarker->fVal != 0) {
+                    // Both lookahead and non-lookahead accepting for this state.
+                    // Favor the look-ahead, because a look-ahead match needs to
+                    // immediately stop the run-time engine. First match, not longest.
+                    sd->fAccepting = fLookAheadRuleMap->elementAti(endMarker->fVal);
+                }
+                // implicit else:
+                // if sd->fAccepting already had a value other than 0 or 1, leave it be.
+            }
+        }
+    }
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//    flagLookAheadStates   Very similar to flagAcceptingStates, above.
+//
+//-----------------------------------------------------------------------------
+void     RBBITableBuilder::flagLookAheadStates() {
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    UVector     lookAheadNodes(*fStatus);
+    RBBINode    *lookAheadNode;
+    int32_t     i;
+    int32_t     n;
+
+    fTree->findNodes(&lookAheadNodes, RBBINode::lookAhead, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    for (i=0; i<lookAheadNodes.size(); i++) {
+        lookAheadNode = (RBBINode *)lookAheadNodes.elementAt(i);
+        U_ASSERT(lookAheadNode->fType == RBBINode::NodeType::lookAhead);
+
+        for (n=0; n<fDStates->size(); n++) {
+            RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+            int32_t positionsIdx = sd->fPositions->indexOf(lookAheadNode);
+            if (positionsIdx >= 0) {
+                U_ASSERT(lookAheadNode == sd->fPositions->elementAt(positionsIdx));
+                uint32_t lookaheadSlot = fLookAheadRuleMap->elementAti(lookAheadNode->fVal);
+                U_ASSERT(sd->fLookAhead == 0 || sd->fLookAhead == lookaheadSlot);
+                // if (sd->fLookAhead != 0 && sd->fLookAhead != lookaheadSlot) {
+                //     printf("%s:%d Bingo. sd->fLookAhead:%d   lookaheadSlot:%d\n",
+                //            __FILE__, __LINE__, sd->fLookAhead, lookaheadSlot);
+                // }
+                sd->fLookAhead = lookaheadSlot;
+            }
+        }
+    }
+}
+
+
+
+
+//-----------------------------------------------------------------------------
+//
+//    flagTaggedStates
+//
+//-----------------------------------------------------------------------------
+void     RBBITableBuilder::flagTaggedStates() {
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    UVector     tagNodes(*fStatus);
+    RBBINode    *tagNode;
+    int32_t     i;
+    int32_t     n;
+
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    fTree->findNodes(&tagNodes, RBBINode::tag, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    for (i=0; i<tagNodes.size(); i++) {                   // For each tag node t (all of 'em)
+        tagNode = (RBBINode *)tagNodes.elementAt(i);
+
+        for (n=0; n<fDStates->size(); n++) {              //    For each state  s (row in the state table)
+            RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+            if (sd->fPositions->indexOf(tagNode) >= 0) {  //       if  s include the tag node t
+                sortedAdd(&sd->fTagVals, tagNode->fVal);
+            }
+        }
+    }
+}
+
+
+
+
+//-----------------------------------------------------------------------------
+//
+//  mergeRuleStatusVals
+//
+//      Update the global table of rule status {tag} values
+//      The rule builder has a global vector of status values that are common
+//      for all tables.  Merge the ones from this table into the global set.
+//
+//-----------------------------------------------------------------------------
+void  RBBITableBuilder::mergeRuleStatusVals() {
+    //
+    //  The basic outline of what happens here is this...
+    //
+    //    for each state in this state table
+    //       if the status tag list for this state is in the global statuses list
+    //           record where and
+    //           continue with the next state
+    //       else
+    //           add the tag list for this state to the global list.
+    //
+    int i;
+    int n;
+
+    // Pre-set a single tag of {0} into the table.
+    //   We will need this as a default, for rule sets with no explicit tagging.
+    if (fRB->fRuleStatusVals->size() == 0) {
+        fRB->fRuleStatusVals->addElement(1, *fStatus);  // Num of statuses in group
+        fRB->fRuleStatusVals->addElement((int32_t)0, *fStatus);  //   and our single status of zero
+    }
+
+    //    For each state
+    for (n=0; n<fDStates->size(); n++) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+        UVector *thisStatesTagValues = sd->fTagVals;
+        if (thisStatesTagValues == NULL) {
+            // No tag values are explicitly associated with this state.
+            //   Set the default tag value.
+            sd->fTagsIdx = 0;
+            continue;
+        }
+
+        // There are tag(s) associated with this state.
+        //   fTagsIdx will be the index into the global tag list for this state's tag values.
+        //   Initial value of -1 flags that we haven't got it set yet.
+        sd->fTagsIdx = -1;
+        int32_t  thisTagGroupStart = 0;   // indexes into the global rule status vals list
+        int32_t  nextTagGroupStart = 0;
+
+        // Loop runs once per group of tags in the global list
+        while (nextTagGroupStart < fRB->fRuleStatusVals->size()) {
+            thisTagGroupStart = nextTagGroupStart;
+            nextTagGroupStart += fRB->fRuleStatusVals->elementAti(thisTagGroupStart) + 1;
+            if (thisStatesTagValues->size() != fRB->fRuleStatusVals->elementAti(thisTagGroupStart)) {
+                // The number of tags for this state is different from
+                //    the number of tags in this group from the global list.
+                //    Continue with the next group from the global list.
+                continue;
+            }
+            // The lengths match, go ahead and compare the actual tag values
+            //    between this state and the group from the global list.
+            for (i=0; i<thisStatesTagValues->size(); i++) {
+                if (thisStatesTagValues->elementAti(i) !=
+                    fRB->fRuleStatusVals->elementAti(thisTagGroupStart + 1 + i) ) {
+                    // Mismatch.
+                    break;
+                }
+            }
+
+            if (i == thisStatesTagValues->size()) {
+                // We found a set of tag values in the global list that match
+                //   those for this state.  Use them.
+                sd->fTagsIdx = thisTagGroupStart;
+                break;
+            }
+        }
+
+        if (sd->fTagsIdx == -1) {
+            // No suitable entry in the global tag list already.  Add one
+            sd->fTagsIdx = fRB->fRuleStatusVals->size();
+            fRB->fRuleStatusVals->addElement(thisStatesTagValues->size(), *fStatus);
+            for (i=0; i<thisStatesTagValues->size(); i++) {
+                fRB->fRuleStatusVals->addElement(thisStatesTagValues->elementAti(i), *fStatus);
+            }
+        }
+    }
+}
+
+
+
+
+
+
+
+//-----------------------------------------------------------------------------
+//
+//  sortedAdd  Add a value to a vector of sorted values (ints).
+//             Do not replicate entries; if the value is already there, do not
+//                add a second one.
+//             Lazily create the vector if it does not already exist.
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::sortedAdd(UVector **vector, int32_t val) {
+    int32_t i;
+
+    if (*vector == NULL) {
+        *vector = new UVector(*fStatus);
+    }
+    if (*vector == NULL || U_FAILURE(*fStatus)) {
+        return;
+    }
+    UVector *vec = *vector;
+    int32_t  vSize = vec->size();
+    for (i=0; i<vSize; i++) {
+        int32_t valAtI = vec->elementAti(i);
+        if (valAtI == val) {
+            // The value is already in the vector.  Don't add it again.
+            return;
+        }
+        if (valAtI > val) {
+            break;
+        }
+    }
+    vec->insertElementAt(val, i, *fStatus);
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//  setAdd     Set operation on UVector
+//             dest = dest union source
+//             Elements may only appear once and must be sorted.
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::setAdd(UVector *dest, UVector *source) {
+    int32_t destOriginalSize = dest->size();
+    int32_t sourceSize       = source->size();
+    int32_t di           = 0;
+    MaybeStackArray<void *, 16> destArray, sourceArray;  // Handle small cases without malloc
+    void **destPtr, **sourcePtr;
+    void **destLim, **sourceLim;
+
+    if (destOriginalSize > destArray.getCapacity()) {
+        if (destArray.resize(destOriginalSize) == NULL) {
+            return;
+        }
+    }
+    destPtr = destArray.getAlias();
+    destLim = destPtr + destOriginalSize;  // destArray.getArrayLimit()?
+
+    if (sourceSize > sourceArray.getCapacity()) {
+        if (sourceArray.resize(sourceSize) == NULL) {
+            return;
+        }
+    }
+    sourcePtr = sourceArray.getAlias();
+    sourceLim = sourcePtr + sourceSize;  // sourceArray.getArrayLimit()?
+
+    // Avoid multiple "get element" calls by getting the contents into arrays
+    (void) dest->toArray(destPtr);
+    (void) source->toArray(sourcePtr);
+
+    dest->setSize(sourceSize+destOriginalSize, *fStatus);
+
+    while (sourcePtr < sourceLim && destPtr < destLim) {
+        if (*destPtr == *sourcePtr) {
+            dest->setElementAt(*sourcePtr++, di++);
+            destPtr++;
+        }
+        // This check is required for machines with segmented memory, like i5/OS.
+        // Direct pointer comparison is not recommended.
+        else if (uprv_memcmp(destPtr, sourcePtr, sizeof(void *)) < 0) {
+            dest->setElementAt(*destPtr++, di++);
+        }
+        else { /* *sourcePtr < *destPtr */
+            dest->setElementAt(*sourcePtr++, di++);
+        }
+    }
+
+    // At most one of these two cleanup loops will execute
+    while (destPtr < destLim) {
+        dest->setElementAt(*destPtr++, di++);
+    }
+    while (sourcePtr < sourceLim) {
+        dest->setElementAt(*sourcePtr++, di++);
+    }
+
+    dest->setSize(di, *fStatus);
+}
+
+
+
+//-----------------------------------------------------------------------------
+//
+//  setEqual    Set operation on UVector.
+//              Compare for equality.
+//              Elements must be sorted.
+//
+//-----------------------------------------------------------------------------
+UBool RBBITableBuilder::setEquals(UVector *a, UVector *b) {
+    return a->equals(*b);
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//  printPosSets   Debug function.  Dump Nullable, firstpos, lastpos and followpos
+//                 for each node in the tree.
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBITableBuilder::printPosSets(RBBINode *n) {
+    if (n==NULL) {
+        return;
+    }
+    printf("\n");
+    RBBINode::printNodeHeader();
+    RBBINode::printNode(n);
+    RBBIDebugPrintf("         Nullable:  %s\n", n->fNullable?"TRUE":"FALSE");
+
+    RBBIDebugPrintf("         firstpos:  ");
+    printSet(n->fFirstPosSet);
+
+    RBBIDebugPrintf("         lastpos:   ");
+    printSet(n->fLastPosSet);
+
+    RBBIDebugPrintf("         followpos: ");
+    printSet(n->fFollowPos);
+
+    printPosSets(n->fLeftChild);
+    printPosSets(n->fRightChild);
+}
+#endif
+
+//
+//    findDuplCharClassFrom()
+//
+bool RBBITableBuilder::findDuplCharClassFrom(IntPair *categories) {
+    int32_t numStates = fDStates->size();
+    int32_t numCols = fRB->fSetBuilder->getNumCharCategories();
+
+    for (; categories->first < numCols-1; categories->first++) {
+        // Note: dictionary & non-dictionary columns cannot be merged.
+        //       The limitSecond value prevents considering mixed pairs.
+        //       Dictionary categories are >= DictCategoriesStart.
+        //       Non dict categories are   <  DictCategoriesStart.
+        int limitSecond = categories->first < fRB->fSetBuilder->getDictCategoriesStart() ?
+            fRB->fSetBuilder->getDictCategoriesStart() : numCols;
+        for (categories->second=categories->first+1; categories->second < limitSecond; categories->second++) {
+            // Initialized to different values to prevent returning true if numStates = 0 (implies no duplicates).
+            uint16_t table_base = 0;
+            uint16_t table_dupl = 1;
+            for (int32_t state=0; state<numStates; state++) {
+                RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(state);
+                table_base = (uint16_t)sd->fDtran->elementAti(categories->first);
+                table_dupl = (uint16_t)sd->fDtran->elementAti(categories->second);
+                if (table_base != table_dupl) {
+                    break;
+                }
+            }
+            if (table_base == table_dupl) {
+                return true;
+            }
+        }
+    }
+    return false;
+}
+
+
+//
+//    removeColumn()
+//
+void RBBITableBuilder::removeColumn(int32_t column) {
+    int32_t numStates = fDStates->size();
+    for (int32_t state=0; state<numStates; state++) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(state);
+        U_ASSERT(column < sd->fDtran->size());
+        sd->fDtran->removeElementAt(column);
+    }
+}
+
+/*
+ * findDuplicateState
+ */
+bool RBBITableBuilder::findDuplicateState(IntPair *states) {
+    int32_t numStates = fDStates->size();
+    int32_t numCols = fRB->fSetBuilder->getNumCharCategories();
+
+    for (; states->first<numStates-1; states->first++) {
+        RBBIStateDescriptor *firstSD = (RBBIStateDescriptor *)fDStates->elementAt(states->first);
+        for (states->second=states->first+1; states->second<numStates; states->second++) {
+            RBBIStateDescriptor *duplSD = (RBBIStateDescriptor *)fDStates->elementAt(states->second);
+            if (firstSD->fAccepting != duplSD->fAccepting ||
+                firstSD->fLookAhead != duplSD->fLookAhead ||
+                firstSD->fTagsIdx   != duplSD->fTagsIdx) {
+                continue;
+            }
+            bool rowsMatch = true;
+            for (int32_t col=0; col < numCols; ++col) {
+                int32_t firstVal = firstSD->fDtran->elementAti(col);
+                int32_t duplVal = duplSD->fDtran->elementAti(col);
+                if (!((firstVal == duplVal) ||
+                        ((firstVal == states->first || firstVal == states->second) &&
+                        (duplVal  == states->first || duplVal  == states->second)))) {
+                    rowsMatch = false;
+                    break;
+                }
+            }
+            if (rowsMatch) {
+                return true;
+            }
+        }
+    }
+    return false;
+}
+
+
+bool RBBITableBuilder::findDuplicateSafeState(IntPair *states) {
+    int32_t numStates = fSafeTable->size();
+
+    for (; states->first<numStates-1; states->first++) {
+        UnicodeString *firstRow = static_cast<UnicodeString *>(fSafeTable->elementAt(states->first));
+        for (states->second=states->first+1; states->second<numStates; states->second++) {
+            UnicodeString *duplRow = static_cast<UnicodeString *>(fSafeTable->elementAt(states->second));
+            bool rowsMatch = true;
+            int32_t numCols = firstRow->length();
+            for (int32_t col=0; col < numCols; ++col) {
+                int32_t firstVal = firstRow->charAt(col);
+                int32_t duplVal = duplRow->charAt(col);
+                if (!((firstVal == duplVal) ||
+                        ((firstVal == states->first || firstVal == states->second) &&
+                        (duplVal  == states->first || duplVal  == states->second)))) {
+                    rowsMatch = false;
+                    break;
+                }
+            }
+            if (rowsMatch) {
+                return true;
+            }
+        }
+    }
+    return false;
+}
+
+
+void RBBITableBuilder::removeState(IntPair duplStates) {
+    const int32_t keepState = duplStates.first;
+    const int32_t duplState = duplStates.second;
+    U_ASSERT(keepState < duplState);
+    U_ASSERT(duplState < fDStates->size());
+
+    RBBIStateDescriptor *duplSD = (RBBIStateDescriptor *)fDStates->elementAt(duplState);
+    fDStates->removeElementAt(duplState);
+    delete duplSD;
+
+    int32_t numStates = fDStates->size();
+    int32_t numCols = fRB->fSetBuilder->getNumCharCategories();
+    for (int32_t state=0; state<numStates; ++state) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(state);
+        for (int32_t col=0; col<numCols; col++) {
+            int32_t existingVal = sd->fDtran->elementAti(col);
+            int32_t newVal = existingVal;
+            if (existingVal == duplState) {
+                newVal = keepState;
+            } else if (existingVal > duplState) {
+                newVal = existingVal - 1;
+            }
+            sd->fDtran->setElementAt(newVal, col);
+        }
+    }
+}
+
+void RBBITableBuilder::removeSafeState(IntPair duplStates) {
+    const int32_t keepState = duplStates.first;
+    const int32_t duplState = duplStates.second;
+    U_ASSERT(keepState < duplState);
+    U_ASSERT(duplState < fSafeTable->size());
+
+    fSafeTable->removeElementAt(duplState);   // Note that fSafeTable has a deleter function
+                                              // and will auto-delete the removed element.
+    int32_t numStates = fSafeTable->size();
+    for (int32_t state=0; state<numStates; ++state) {
+        UnicodeString *sd = (UnicodeString *)fSafeTable->elementAt(state);
+        int32_t numCols = sd->length();
+        for (int32_t col=0; col<numCols; col++) {
+            int32_t existingVal = sd->charAt(col);
+            int32_t newVal = existingVal;
+            if (existingVal == duplState) {
+                newVal = keepState;
+            } else if (existingVal > duplState) {
+                newVal = existingVal - 1;
+            }
+            sd->setCharAt(col, static_cast<char16_t>(newVal));
+        }
+    }
+}
+
+
+/*
+ * RemoveDuplicateStates
+ */
+int32_t RBBITableBuilder::removeDuplicateStates() {
+    IntPair dupls = {3, 0};
+    int32_t numStatesRemoved = 0;
+
+    while (findDuplicateState(&dupls)) {
+        // printf("Removing duplicate states (%d, %d)\n", dupls.first, dupls.second);
+        removeState(dupls);
+        ++numStatesRemoved;
+    }
+    return numStatesRemoved;
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//   getTableSize()    Calculate the size of the runtime form of this
+//                     state transition table.
+//
+//-----------------------------------------------------------------------------
+int32_t  RBBITableBuilder::getTableSize() const {
+    int32_t    size = 0;
+    int32_t    numRows;
+    int32_t    numCols;
+    int32_t    rowSize;
+
+    if (fTree == NULL) {
+        return 0;
+    }
+
+    size    = offsetof(RBBIStateTable, fTableData);    // The header, with no rows to the table.
+
+    numRows = fDStates->size();
+    numCols = fRB->fSetBuilder->getNumCharCategories();
+
+    if (use8BitsForTable()) {
+        rowSize = offsetof(RBBIStateTableRow8, fNextState) + sizeof(int8_t)*numCols;
+    } else {
+        rowSize = offsetof(RBBIStateTableRow16, fNextState) + sizeof(int16_t)*numCols;
+    }
+    size   += numRows * rowSize;
+    return size;
+}
+
+bool RBBITableBuilder::use8BitsForTable() const {
+    return fDStates->size() <= kMaxStateFor8BitsTable;
+}
+
+//-----------------------------------------------------------------------------
+//
+//   exportTable()    export the state transition table in the format required
+//                    by the runtime engine.  getTableSize() bytes of memory
+//                    must be available at the output address "where".
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::exportTable(void *where) {
+    RBBIStateTable    *table = (RBBIStateTable *)where;
+    uint32_t           state;
+    int                col;
+
+    if (U_FAILURE(*fStatus) || fTree == NULL) {
+        return;
+    }
+
+    int32_t catCount = fRB->fSetBuilder->getNumCharCategories();
+    if (catCount > 0x7fff ||
+        fDStates->size() > 0x7fff) {
+        *fStatus = U_BRK_INTERNAL_ERROR;
+        return;
+    }
+
+    table->fNumStates = fDStates->size();
+    table->fDictCategoriesStart = fRB->fSetBuilder->getDictCategoriesStart();
+    table->fLookAheadResultsSize = fLASlotsInUse == ACCEPTING_UNCONDITIONAL ? 0 : fLASlotsInUse + 1;
+    table->fFlags     = 0;
+    if (use8BitsForTable()) {
+        table->fRowLen    = offsetof(RBBIStateTableRow8, fNextState) + sizeof(uint8_t) * catCount;
+        table->fFlags  |= RBBI_8BITS_ROWS;
+    } else {
+        table->fRowLen    = offsetof(RBBIStateTableRow16, fNextState) + sizeof(int16_t) * catCount;
+    }
+    if (fRB->fLookAheadHardBreak) {
+        table->fFlags  |= RBBI_LOOKAHEAD_HARD_BREAK;
+    }
+    if (fRB->fSetBuilder->sawBOF()) {
+        table->fFlags  |= RBBI_BOF_REQUIRED;
+    }
+
+    for (state=0; state<table->fNumStates; state++) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(state);
+        RBBIStateTableRow   *row = (RBBIStateTableRow *)(table->fTableData + state*table->fRowLen);
+        if (use8BitsForTable()) {
+            U_ASSERT (sd->fAccepting <= 255);
+            U_ASSERT (sd->fLookAhead <= 255);
+            U_ASSERT (0 <= sd->fTagsIdx && sd->fTagsIdx <= 255);
+            row->r8.fAccepting = sd->fAccepting;
+            row->r8.fLookAhead = sd->fLookAhead;
+            row->r8.fTagsIdx   = sd->fTagsIdx;
+            for (col=0; col<catCount; col++) {
+                U_ASSERT (sd->fDtran->elementAti(col) <= kMaxStateFor8BitsTable);
+                row->r8.fNextState[col] = sd->fDtran->elementAti(col);
+            }
+        } else {
+            U_ASSERT (sd->fAccepting <= 0xffff);
+            U_ASSERT (sd->fLookAhead <= 0xffff);
+            U_ASSERT (0 <= sd->fTagsIdx && sd->fTagsIdx <= 0xffff);
+            row->r16.fAccepting = sd->fAccepting;
+            row->r16.fLookAhead = sd->fLookAhead;
+            row->r16.fTagsIdx   = sd->fTagsIdx;
+            for (col=0; col<catCount; col++) {
+                row->r16.fNextState[col] = sd->fDtran->elementAti(col);
+            }
+        }
+    }
+}
+
+
+/**
+ *   Synthesize a safe state table from the main state table.
+ */
+void RBBITableBuilder::buildSafeReverseTable(UErrorCode &status) {
+    // The safe table creation has three steps:
+
+    // 1. Identifiy pairs of character classes that are "safe." Safe means that boundaries
+    // following the pair do not depend on context or state before the pair. To test
+    // whether a pair is safe, run it through the main forward state table, starting
+    // from each state. If the the final state is the same, no matter what the starting state,
+    // the pair is safe.
+    //
+    // 2. Build a state table that recognizes the safe pairs. It's similar to their
+    // forward table, with a column for each input character [class], and a row for
+    // each state. Row 1 is the start state, and row 0 is the stop state. Initially
+    // create an additional state for each input character category; being in
+    // one of these states means that the character has been seen, and is potentially
+    // the first of a pair. In each of these rows, the entry for the second character
+    // of a safe pair is set to the stop state (0), indicating that a match was found.
+    // All other table entries are set to the state corresponding the current input
+    // character, allowing that charcter to be the of a start following pair.
+    //
+    // Because the safe rules are to be run in reverse, moving backwards in the text,
+    // the first and second pair categories are swapped when building the table.
+    //
+    // 3. Compress the table. There are typically many rows (states) that are
+    // equivalent - that have zeroes (match completed) in the same columns -
+    // and can be folded together.
+
+    // Each safe pair is stored as two UChars in the safePair string.
+    UnicodeString safePairs;
+
+    int32_t numCharClasses = fRB->fSetBuilder->getNumCharCategories();
+    int32_t numStates = fDStates->size();
+
+    for (int32_t c1=0; c1<numCharClasses; ++c1) {
+        for (int32_t c2=0; c2 < numCharClasses; ++c2) {
+            int32_t wantedEndState = -1;
+            int32_t endState = 0;
+            for (int32_t startState = 1; startState < numStates; ++startState) {
+                RBBIStateDescriptor *startStateD = static_cast<RBBIStateDescriptor *>(fDStates->elementAt(startState));
+                int32_t s2 = startStateD->fDtran->elementAti(c1);
+                RBBIStateDescriptor *s2StateD = static_cast<RBBIStateDescriptor *>(fDStates->elementAt(s2));
+                endState = s2StateD->fDtran->elementAti(c2);
+                if (wantedEndState < 0) {
+                    wantedEndState = endState;
+                } else {
+                    if (wantedEndState != endState) {
+                        break;
+                    }
+                }
+            }
+            if (wantedEndState == endState) {
+                safePairs.append((char16_t)c1);
+                safePairs.append((char16_t)c2);
+                // printf("(%d, %d) ", c1, c2);
+            }
+        }
+        // printf("\n");
+    }
+
+    // Populate the initial safe table.
+    // The table as a whole is UVector<UnicodeString>
+    // Each row is represented by a UnicodeString, being used as a Vector<int16>.
+    // Row 0 is the stop state.
+    // Row 1 is the start sate.
+    // Row 2 and beyond are other states, initially one per char class, but
+    //   after initial construction, many of the states will be combined, compacting the table.
+    // The String holds the nextState data only. The four leading fields of a row, fAccepting,
+    // fLookAhead, etc. are not needed for the safe table, and are omitted at this stage of building.
+
+    U_ASSERT(fSafeTable == nullptr);
+    fSafeTable = new UVector(uprv_deleteUObject, uhash_compareUnicodeString, numCharClasses + 2, status);
+    for (int32_t row=0; row<numCharClasses + 2; ++row) {
+        fSafeTable->addElement(new UnicodeString(numCharClasses, 0, numCharClasses+4), status);
+    }
+
+    // From the start state, each input char class transitions to the state for that input.
+    UnicodeString &startState = *static_cast<UnicodeString *>(fSafeTable->elementAt(1));
+    for (int32_t charClass=0; charClass < numCharClasses; ++charClass) {
+        // Note: +2 for the start & stop state.
+        startState.setCharAt(charClass, static_cast<char16_t>(charClass+2));
+    }
+
+    // Initially make every other state table row look like the start state row,
+    for (int32_t row=2; row<numCharClasses+2; ++row) {
+        UnicodeString &rowState = *static_cast<UnicodeString *>(fSafeTable->elementAt(row));
+        rowState = startState;   // UnicodeString assignment, copies contents.
+    }
+
+    // Run through the safe pairs, set the next state to zero when pair has been seen.
+    // Zero being the stop state, meaning we found a safe point.
+    for (int32_t pairIdx=0; pairIdx<safePairs.length(); pairIdx+=2) {
+        int32_t c1 = safePairs.charAt(pairIdx);
+        int32_t c2 = safePairs.charAt(pairIdx + 1);
+
+        UnicodeString &rowState = *static_cast<UnicodeString *>(fSafeTable->elementAt(c2 + 2));
+        rowState.setCharAt(c1, 0);
+    }
+
+    // Remove duplicate or redundant rows from the table.
+    IntPair states = {1, 0};
+    while (findDuplicateSafeState(&states)) {
+        // printf("Removing duplicate safe states (%d, %d)\n", states.first, states.second);
+        removeSafeState(states);
+    }
+}
+
+
+//-----------------------------------------------------------------------------
+//
+//   getSafeTableSize()    Calculate the size of the runtime form of this
+//                         safe state table.
+//
+//-----------------------------------------------------------------------------
+int32_t  RBBITableBuilder::getSafeTableSize() const {
+    int32_t    size = 0;
+    int32_t    numRows;
+    int32_t    numCols;
+    int32_t    rowSize;
+
+    if (fSafeTable == nullptr) {
+        return 0;
+    }
+
+    size    = offsetof(RBBIStateTable, fTableData);    // The header, with no rows to the table.
+
+    numRows = fSafeTable->size();
+    numCols = fRB->fSetBuilder->getNumCharCategories();
+
+    if (use8BitsForSafeTable()) {
+        rowSize = offsetof(RBBIStateTableRow8, fNextState) + sizeof(int8_t)*numCols;
+    } else {
+        rowSize = offsetof(RBBIStateTableRow16, fNextState) + sizeof(int16_t)*numCols;
+    }
+    size   += numRows * rowSize;
+    return size;
+}
+
+bool RBBITableBuilder::use8BitsForSafeTable() const {
+    return fSafeTable->size() <= kMaxStateFor8BitsTable;
+}
+
+//-----------------------------------------------------------------------------
+//
+//   exportSafeTable()   export the state transition table in the format required
+//                       by the runtime engine.  getTableSize() bytes of memory
+//                       must be available at the output address "where".
+//
+//-----------------------------------------------------------------------------
+void RBBITableBuilder::exportSafeTable(void *where) {
+    RBBIStateTable    *table = (RBBIStateTable *)where;
+    uint32_t           state;
+    int                col;
+
+    if (U_FAILURE(*fStatus) || fSafeTable == nullptr) {
+        return;
+    }
+
+    int32_t catCount = fRB->fSetBuilder->getNumCharCategories();
+    if (catCount > 0x7fff ||
+            fSafeTable->size() > 0x7fff) {
+        *fStatus = U_BRK_INTERNAL_ERROR;
+        return;
+    }
+
+    table->fNumStates = fSafeTable->size();
+    table->fFlags     = 0;
+    if (use8BitsForSafeTable()) {
+        table->fRowLen    = offsetof(RBBIStateTableRow8, fNextState) + sizeof(uint8_t) * catCount;
+        table->fFlags  |= RBBI_8BITS_ROWS;
+    } else {
+        table->fRowLen    = offsetof(RBBIStateTableRow16, fNextState) + sizeof(int16_t) * catCount;
+    }
+
+    for (state=0; state<table->fNumStates; state++) {
+        UnicodeString *rowString = (UnicodeString *)fSafeTable->elementAt(state);
+        RBBIStateTableRow   *row = (RBBIStateTableRow *)(table->fTableData + state*table->fRowLen);
+        if (use8BitsForSafeTable()) {
+            row->r8.fAccepting = 0;
+            row->r8.fLookAhead = 0;
+            row->r8.fTagsIdx    = 0;
+            for (col=0; col<catCount; col++) {
+                U_ASSERT(rowString->charAt(col) <= kMaxStateFor8BitsTable);
+                row->r8.fNextState[col] = static_cast<uint8_t>(rowString->charAt(col));
+            }
+        } else {
+            row->r16.fAccepting = 0;
+            row->r16.fLookAhead = 0;
+            row->r16.fTagsIdx    = 0;
+            for (col=0; col<catCount; col++) {
+                row->r16.fNextState[col] = rowString->charAt(col);
+            }
+        }
+    }
+}
+
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   printSet    Debug function.   Print the contents of a UVector
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBITableBuilder::printSet(UVector *s) {
+    int32_t  i;
+    for (i=0; i<s->size(); i++) {
+        const RBBINode *v = static_cast<const RBBINode *>(s->elementAt(i));
+        RBBIDebugPrintf("%5d", v==NULL? -1 : v->fSerialNum);
+    }
+    RBBIDebugPrintf("\n");
+}
+#endif
+
+
+//-----------------------------------------------------------------------------
+//
+//   printStates    Debug Function.  Dump the fully constructed state transition table.
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBITableBuilder::printStates() {
+    int     c;    // input "character"
+    int     n;    // state number
+
+    RBBIDebugPrintf("state |           i n p u t     s y m b o l s \n");
+    RBBIDebugPrintf("      | Acc  LA    Tag");
+    for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+        RBBIDebugPrintf(" %3d", c);
+    }
+    RBBIDebugPrintf("\n");
+    RBBIDebugPrintf("      |---------------");
+    for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+        RBBIDebugPrintf("----");
+    }
+    RBBIDebugPrintf("\n");
+
+    for (n=0; n<fDStates->size(); n++) {
+        RBBIStateDescriptor *sd = (RBBIStateDescriptor *)fDStates->elementAt(n);
+        RBBIDebugPrintf("  %3d | " , n);
+        RBBIDebugPrintf("%3d %3d %5d ", sd->fAccepting, sd->fLookAhead, sd->fTagsIdx);
+        for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+            RBBIDebugPrintf(" %3d", sd->fDtran->elementAti(c));
+        }
+        RBBIDebugPrintf("\n");
+    }
+    RBBIDebugPrintf("\n\n");
+}
+#endif
+
+
+//-----------------------------------------------------------------------------
+//
+//   printSafeTable    Debug Function.  Dump the fully constructed safe table.
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBITableBuilder::printReverseTable() {
+    int     c;    // input "character"
+    int     n;    // state number
+
+    RBBIDebugPrintf("    Safe Reverse Table \n");
+    if (fSafeTable == nullptr) {
+        RBBIDebugPrintf("   --- nullptr ---\n");
+        return;
+    }
+    RBBIDebugPrintf("state |           i n p u t     s y m b o l s \n");
+    RBBIDebugPrintf("      | Acc  LA    Tag");
+    for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+        RBBIDebugPrintf(" %2d", c);
+    }
+    RBBIDebugPrintf("\n");
+    RBBIDebugPrintf("      |---------------");
+    for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+        RBBIDebugPrintf("---");
+    }
+    RBBIDebugPrintf("\n");
+
+    for (n=0; n<fSafeTable->size(); n++) {
+        UnicodeString *rowString = (UnicodeString *)fSafeTable->elementAt(n);
+        RBBIDebugPrintf("  %3d | " , n);
+        RBBIDebugPrintf("%3d %3d %5d ", 0, 0, 0);  // Accepting, LookAhead, Tags
+        for (c=0; c<fRB->fSetBuilder->getNumCharCategories(); c++) {
+            RBBIDebugPrintf(" %2d", rowString->charAt(c));
+        }
+        RBBIDebugPrintf("\n");
+    }
+    RBBIDebugPrintf("\n\n");
+}
+#endif
+
+
+
+//-----------------------------------------------------------------------------
+//
+//   printRuleStatusTable    Debug Function.  Dump the common rule status table
+//
+//-----------------------------------------------------------------------------
+#ifdef RBBI_DEBUG
+void RBBITableBuilder::printRuleStatusTable() {
+    int32_t  thisRecord = 0;
+    int32_t  nextRecord = 0;
+    int      i;
+    UVector  *tbl = fRB->fRuleStatusVals;
+
+    RBBIDebugPrintf("index |  tags \n");
+    RBBIDebugPrintf("-------------------\n");
+
+    while (nextRecord < tbl->size()) {
+        thisRecord = nextRecord;
+        nextRecord = thisRecord + tbl->elementAti(thisRecord) + 1;
+        RBBIDebugPrintf("%4d   ", thisRecord);
+        for (i=thisRecord+1; i<nextRecord; i++) {
+            RBBIDebugPrintf("  %5d", tbl->elementAti(i));
+        }
+        RBBIDebugPrintf("\n");
+    }
+    RBBIDebugPrintf("\n\n");
+}
+#endif
+
+
+//-----------------------------------------------------------------------------
+//
+//   RBBIStateDescriptor     Methods.  This is a very struct-like class
+//                           Most access is directly to the fields.
+//
+//-----------------------------------------------------------------------------
+
+RBBIStateDescriptor::RBBIStateDescriptor(int lastInputSymbol, UErrorCode *fStatus) {
+    fMarked    = FALSE;
+    fAccepting = 0;
+    fLookAhead = 0;
+    fTagsIdx   = 0;
+    fTagVals   = NULL;
+    fPositions = NULL;
+    fDtran     = NULL;
+
+    fDtran     = new UVector32(lastInputSymbol+1, *fStatus);
+    if (U_FAILURE(*fStatus)) {
+        return;
+    }
+    if (fDtran == NULL) {
+        *fStatus = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    fDtran->setSize(lastInputSymbol+1);    // fDtran needs to be pre-sized.
+                                           //   It is indexed by input symbols, and will
+                                           //   hold  the next state number for each
+                                           //   symbol.
+}
+
+
+RBBIStateDescriptor::~RBBIStateDescriptor() {
+    delete       fPositions;
+    delete       fDtran;
+    delete       fTagVals;
+    fPositions = NULL;
+    fDtran     = NULL;
+    fTagVals   = NULL;
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/rbbitblb.h b/thirdparty/icu4c/common/rbbitblb.h
new file mode 100644
index 0000000000..fe3db8d7bf
--- /dev/null
+++ b/thirdparty/icu4c/common/rbbitblb.h
@@ -0,0 +1,232 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+//  rbbitblb.h
+//
+
+/*
+**********************************************************************
+*   Copyright (c) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef RBBITBLB_H
+#define RBBITBLB_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/uobject.h"
+#include "unicode/rbbi.h"
+#include "rbbidata.h"
+#include "rbbirb.h"
+#include "rbbinode.h"
+
+
+U_NAMESPACE_BEGIN
+
+class RBBIRuleScanner;
+class RBBIRuleBuilder;
+class UVector32;
+
+//
+//  class RBBITableBuilder is part of the RBBI rule compiler.
+//                         It builds the state transition table used by the RBBI runtime
+//                         from the expression syntax tree generated by the rule scanner.
+//
+//                         This class is part of the RBBI implementation only.
+//                         There is no user-visible public API here.
+//
+
+class RBBITableBuilder : public UMemory {
+public:
+    RBBITableBuilder(RBBIRuleBuilder *rb, RBBINode **rootNode, UErrorCode &status);
+    ~RBBITableBuilder();
+
+    void     buildForwardTable();
+
+    /** Return the runtime size in bytes of the built state table.  */
+    int32_t  getTableSize() const;
+
+    /** Fill in the runtime state table. Sufficient memory must exist at the specified location.
+     */
+    void     exportTable(void *where);
+
+    /** Use 8 bits to encode the forward table */
+    bool     use8BitsForTable() const;
+
+    /**
+     *  Find duplicate (redundant) character classes. Begin looking with categories.first.
+     *  Duplicate, if found are returned in the categories parameter.
+     *  This is an iterator-like function, used to identify character classes
+     *  (state table columns) that can be eliminated.
+     *  @param categories in/out parameter, specifies where to start looking for duplicates,
+     *                and returns the first pair of duplicates found, if any.
+     *  @return true if duplicate char classes were found, false otherwise.
+     */
+    bool     findDuplCharClassFrom(IntPair *categories);
+
+    /** Remove a column from the state table. Used when two character categories
+     *  have been found equivalent, and merged together, to eliminate the uneeded table column.
+     */
+    void     removeColumn(int32_t column);
+
+    /**
+     * Check for, and remove dupicate states (table rows).
+     * @return the number of states removed.
+     */
+    int32_t  removeDuplicateStates();
+
+    /** Build the safe reverse table from the already-constructed forward table. */
+    void     buildSafeReverseTable(UErrorCode &status);
+
+    /** Return the runtime size in bytes of the built safe reverse state table. */
+    int32_t  getSafeTableSize() const;
+
+    /** Fill in the runtime safe state table. Sufficient memory must exist at the specified location.
+     */
+    void     exportSafeTable(void *where);
+
+    /** Use 8 bits to encode the safe reverse table */
+    bool     use8BitsForSafeTable() const;
+
+private:
+    void     calcNullable(RBBINode *n);
+    void     calcFirstPos(RBBINode *n);
+    void     calcLastPos(RBBINode  *n);
+    void     calcFollowPos(RBBINode *n);
+    void     calcChainedFollowPos(RBBINode *n, RBBINode *endMarkNode);
+    void     bofFixup();
+    void     buildStateTable();
+    void     mapLookAheadRules();
+    void     flagAcceptingStates();
+    void     flagLookAheadStates();
+    void     flagTaggedStates();
+    void     mergeRuleStatusVals();
+
+    /**
+     * Merge redundant state table columns, eliminating character classes with identical behavior.
+     * Done after the state tables are generated, just before converting to their run-time format.
+     */
+    int32_t  mergeColumns();
+
+    void     addRuleRootNodes(UVector *dest, RBBINode *node);
+
+    /**
+     *  Find duplicate (redundant) states, beginning at the specified pair,
+     *  within this state table. This is an iterator-like function, used to
+     *  identify states (state table rows) that can be eliminated.
+     *  @param states in/out parameter, specifies where to start looking for duplicates,
+     *                and returns the first pair of duplicates found, if any.
+     *  @return true if duplicate states were found, false otherwise.
+     */
+    bool findDuplicateState(IntPair *states);
+
+    /** Remove a duplicate state.
+     * @param duplStates The duplicate states. The first is kept, the second is removed.
+     *                   All references to the second in the state table are retargeted
+     *                   to the first.
+     */
+    void removeState(IntPair duplStates);
+
+    /** Find the next duplicate state in the safe reverse table. An iterator function.
+     *  @param states in/out parameter, specifies where to start looking for duplicates,
+     *                and returns the first pair of duplicates found, if any.
+     *  @return true if a duplicate pair of states was found.
+     */
+    bool findDuplicateSafeState(IntPair *states);
+
+    /** Remove a duplicate state from the safe table.
+     * @param duplStates The duplicate states. The first is kept, the second is removed.
+     *                   All references to the second in the state table are retargeted
+     *                   to the first.
+     */
+    void removeSafeState(IntPair duplStates);
+
+    // Set functions for UVector.
+    //   TODO:  make a USet subclass of UVector
+
+    void     setAdd(UVector *dest, UVector *source);
+    UBool    setEquals(UVector *a, UVector *b);
+
+    void     sortedAdd(UVector **dest, int32_t val);
+
+public:
+#ifdef RBBI_DEBUG
+    void     printSet(UVector *s);
+    void     printPosSets(RBBINode *n /* = NULL*/);
+    void     printStates();
+    void     printRuleStatusTable();
+    void     printReverseTable();
+#else
+    #define  printSet(s)
+    #define  printPosSets(n)
+    #define  printStates()
+    #define  printRuleStatusTable()
+    #define  printReverseTable()
+#endif
+
+private:
+    RBBIRuleBuilder  *fRB;
+    RBBINode         *&fTree;              // The root node of the parse tree to build a
+                                           //   table for.
+    UErrorCode       *fStatus;
+
+    /** State Descriptors, UVector<RBBIStateDescriptor> */
+    UVector          *fDStates;            //  D states (Aho's terminology)
+                                           //  Index is state number
+                                           //  Contents are RBBIStateDescriptor pointers.
+
+    /** Synthesized safe table, UVector of UnicodeString, one string per table row.   */
+    UVector          *fSafeTable;
+
+    /** Map from rule number (fVal in look ahead nodes) to sequential lookahead index. */
+    UVector32        *fLookAheadRuleMap = nullptr;
+
+    /* Counter used when assigning lookahead rule numbers.
+     * Contains the last look-ahead number already in use.
+     * The first look-ahead number is 2; Number 1 (ACCEPTING_UNCONDITIONAL) is reserved
+     * for non-lookahead accepting states. See the declarations of RBBIStateTableRowT.   */
+    int32_t          fLASlotsInUse = ACCEPTING_UNCONDITIONAL;
+
+
+    RBBITableBuilder(const RBBITableBuilder &other) = delete; // forbid copying of this class
+    RBBITableBuilder &operator=(const RBBITableBuilder &other) = delete; // forbid copying of this class
+};
+
+//
+//  RBBIStateDescriptor - The DFA is constructed as a set of these descriptors,
+//                        one for each state.
+class RBBIStateDescriptor : public UMemory {
+public:
+    UBool            fMarked;
+    uint32_t         fAccepting;
+    uint32_t         fLookAhead;
+    UVector          *fTagVals;
+    int32_t          fTagsIdx;
+    UVector          *fPositions;          // Set of parse tree positions associated
+                                           //   with this state.  Unordered (it's a set).
+                                           //   UVector contents are RBBINode *
+
+    UVector32        *fDtran;              // Transitions out of this state.
+                                           //   indexed by input character
+                                           //   contents is int index of dest state
+                                           //   in RBBITableBuilder.fDStates
+
+    RBBIStateDescriptor(int maxInputSymbol,  UErrorCode *fStatus);
+    ~RBBIStateDescriptor();
+
+private:
+    RBBIStateDescriptor(const RBBIStateDescriptor &other); // forbid copying of this class
+    RBBIStateDescriptor &operator=(const RBBIStateDescriptor &other); // forbid copying of this class
+};
+
+
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif
diff --git a/thirdparty/icu4c/common/resbund.cpp b/thirdparty/icu4c/common/resbund.cpp
new file mode 100644
index 0000000000..7c5063b211
--- /dev/null
+++ b/thirdparty/icu4c/common/resbund.cpp
@@ -0,0 +1,399 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File resbund.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   02/05/97    aliu        Fixed bug in chopLocale.  Added scanForLocaleInFile
+*                           based on code taken from scanForLocale.  Added
+*                           constructor which attempts to read resource bundle
+*                           from a specific file, without searching other files.
+*   02/11/97    aliu        Added UErrorCode return values to constructors. Fixed
+*                           infinite loops in scanForFile and scanForLocale.
+*                           Modified getRawResourceData to not delete storage in
+*                           localeData and resourceData which it doesn't own.
+*                           Added Mac compatibility #ifdefs for tellp() and
+*                           ios::nocreate.
+*   03/04/97    aliu        Modified to use ExpandingDataSink objects instead of
+*                           the highly inefficient ostrstream objects.
+*   03/13/97    aliu        Rewrote to load in entire resource bundle and store
+*                           it as a Hashtable of ResourceBundleData objects.
+*                           Added state table to govern parsing of files.
+*                           Modified to load locale index out of new file distinct
+*                           from default.txt.
+*   03/25/97    aliu        Modified to support 2-d arrays, needed for timezone data.
+*                           Added support for custom file suffixes.  Again, needed
+*                           to support timezone data.  Improved error handling to
+*                           detect duplicate tags and subtags.
+*   04/07/97    aliu        Fixed bug in getHashtableForLocale().  Fixed handling
+*                           of failing UErrorCode values on entry to API methods.
+*                           Fixed bugs in getArrayItem() for negative indices.
+*   04/29/97    aliu        Update to use new Hashtable deletion protocol.
+*   05/06/97    aliu        Flattened kTransitionTable for HP compiler.
+*                           Fixed usage of CharString.
+* 06/11/99      stephen     Removed parsing of .txt files.
+*                           Reworked to use new binary format.
+*                           Cleaned up.
+* 06/14/99      stephen     Removed methods taking a filename suffix.
+* 06/22/99      stephen     Added missing T_FileStream_close in parse()
+* 11/09/99      weiv        Added getLocale(), rewritten constructForLocale()
+* March 2000    weiv        complete overhaul.
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/resbund.h"
+
+#include "cmemory.h"
+#include "mutex.h"
+#include "uassert.h"
+#include "umutex.h"
+
+#include "uresimp.h"
+
+U_NAMESPACE_BEGIN
+
+/*-----------------------------------------------------------------------------
+ * Implementation Notes
+ *
+ * Resource bundles are read in once, and thereafter cached.
+ * ResourceBundle statically keeps track of which files have been
+ * read, so we are guaranteed that each file is read at most once.
+ * Resource bundles can be loaded from different data directories and
+ * will be treated as distinct, even if they are for the same locale.
+ *
+ * Resource bundles are lightweight objects, which have pointers to
+ * one or more shared Hashtable objects containing all the data.
+ * Copying would be cheap, but there is no copy constructor, since
+ * there wasn't one in the original API.
+ *
+ * The ResourceBundle parsing mechanism is implemented as a transition
+ * network, for easy maintenance and modification.  The network is
+ * implemented as a matrix (instead of in code) to make this even
+ * easier.  The matrix contains Transition objects.  Each Transition
+ * object describes a destination node and an action to take before
+ * moving to the destination node.  The source node is encoded by the
+ * index of the object in the array that contains it.  The pieces
+ * needed to understand the transition network are the enums for node
+ * IDs and actions, the parse() method, which walks through the
+ * network and implements the actions, and the network itself.  The
+ * network guarantees certain conditions, for example, that a new
+ * resource will not be closed until one has been opened first; or
+ * that data will not be stored into a TaggedList until a TaggedList
+ * has been created.  Nonetheless, the code in parse() does some
+ * consistency checks as it runs the network, and fails with an
+ * U_INTERNAL_PROGRAM_ERROR if one of these checks fails.  If the input
+ * data has a bad format, an U_INVALID_FORMAT_ERROR is returned.  If you
+ * see an U_INTERNAL_PROGRAM_ERROR the transition matrix has a bug in
+ * it.
+ *
+ * Old functionality of multiple locales in a single file is still
+ * supported.  For this reason, LOCALE names override FILE names.  If
+ * data for en_US is located in the en.txt file, once it is loaded,
+ * the code will not care where it came from (other than remembering
+ * which directory it came from).  However, if there is an en_US
+ * resource in en_US.txt, that will take precedence.  There is no
+ * limit to the number or type of resources that can be stored in a
+ * file, however, files are only searched in a specific way.  If
+ * en_US_CA is requested, then first en_US_CA.txt is searched, then
+ * en_US.txt, then en.txt, then default.txt.  So it only makes sense
+ * to put certain locales in certain files.  In this example, it would
+ * be logical to put en_US_CA, en_US, and en into the en.txt file,
+ * since they would be found there if asked for.  The extreme example
+ * is to place all locale resources into default.txt, which should
+ * also work.
+ *
+ * Inheritance is implemented.  For example, xx_YY_zz inherits as
+ * follows: xx_YY_zz, xx_YY, xx, default.  Inheritance is implemented
+ * as an array of hashtables.  There will be from 1 to 4 hashtables in
+ * the array.
+ *
+ * Fallback files are implemented.  The fallback pattern is Language
+ * Country Variant (LCV) -> LC -> L.  Fallback is first done for the
+ * requested locale.  Then it is done for the default locale, as
+ * returned by Locale::getDefault().  Then the special file
+ * default.txt is searched for the default locale.  The overall FILE
+ * fallback path is LCV -> LC -> L -> dLCV -> dLC -> dL -> default.
+ *
+ * Note that although file name searching includes the default locale,
+ * once a ResourceBundle object is constructed, the inheritance path
+ * no longer includes the default locale.  The path is LCV -> LC -> L
+ * -> default.
+ *
+ * File parsing is lazy.  Nothing is parsed unless it is called for by
+ * someone.  So when a ResourceBundle for xx_YY_zz is constructed,
+ * only that locale is parsed (along with anything else in the same
+ * file).  Later, if the FooBar tag is asked for, and if it isn't
+ * found in xx_YY_zz, then xx_YY.txt will be parsed and checked, and
+ * so forth, until the chain is exhausted or the tag is found.
+ *
+ * Thread-safety is implemented around caches, both the cache that
+ * stores all the resouce data, and the cache that stores flags
+ * indicating whether or not a file has been visited.  These caches
+ * delete their storage at static cleanup time, when the process
+ * quits.
+ *
+ * ResourceBundle supports TableCollation as a special case.  This
+ * involves having special ResourceBundle objects which DO own their
+ * data, since we don't want large collation rule strings in the
+ * ResourceBundle cache (these are already cached in the
+ * TableCollation cache).  TableCollation files (.ctx files) have the
+ * same format as normal resource data files, with a different
+ * interpretation, from the standpoint of ResourceBundle.  .ctx files
+ * are loaded into otherwise ordinary ResourceBundle objects.  They
+ * don't inherit (that's implemented by TableCollation) and they own
+ * their data (as mentioned above).  However, they still support
+ * possible multiple locales in a single .ctx file.  (This is in
+ * practice a bad idea, since you only want the one locale you're
+ * looking for, and only one tag will be present
+ * ("CollationElements"), so you don't need an inheritance chain of
+ * multiple locales.)  Up to 4 locale resources will be loaded from a
+ * .ctx file; everything after the first 4 is ignored (parsed and
+ * deleted).  (Normal .txt files have no limit.)  Instead of being
+ * loaded into the cache, and then looked up as needed, the locale
+ * resources are read straight into the ResourceBundle object.
+ *
+ * The Index, which used to reside in default.txt, has been moved to a
+ * new file, index.txt.  This file contains a slightly modified format
+ * with the addition of the "InstalledLocales" tag; it looks like:
+ *
+ * Index {
+ *   InstalledLocales {
+ *     ar
+ *     ..
+ *     zh_TW
+ *   }
+ * }
+ */
+//-----------------------------------------------------------------------------
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ResourceBundle)
+
+ResourceBundle::ResourceBundle(UErrorCode &err)
+                                :UObject(), fLocale(NULL)
+{
+    fResource = ures_open(0, Locale::getDefault().getName(), &err);
+}
+
+ResourceBundle::ResourceBundle(const ResourceBundle &other)
+                              :UObject(other), fLocale(NULL)
+{
+    UErrorCode status = U_ZERO_ERROR;
+
+    if (other.fResource) {
+        fResource = ures_copyResb(0, other.fResource, &status);
+    } else {
+        /* Copying a bad resource bundle */
+        fResource = NULL;
+    }
+}
+
+ResourceBundle::ResourceBundle(UResourceBundle *res, UErrorCode& err)
+                               :UObject(), fLocale(NULL)
+{
+    if (res) {
+        fResource = ures_copyResb(0, res, &err);
+    } else {
+        /* Copying a bad resource bundle */
+        fResource = NULL;
+    }
+}
+
+ResourceBundle::ResourceBundle(const char* path, const Locale& locale, UErrorCode& err) 
+                               :UObject(), fLocale(NULL)
+{
+    fResource = ures_open(path, locale.getName(), &err);
+}
+
+
+ResourceBundle& ResourceBundle::operator=(const ResourceBundle& other)
+{
+    if(this == &other) {
+        return *this;
+    }
+    if(fResource != 0) {
+        ures_close(fResource);
+        fResource = NULL;
+    }
+    if (fLocale != NULL) {
+        delete fLocale;
+        fLocale = NULL;
+    }
+    UErrorCode status = U_ZERO_ERROR;
+    if (other.fResource) {
+        fResource = ures_copyResb(0, other.fResource, &status);
+    } else {
+        /* Copying a bad resource bundle */
+        fResource = NULL;
+    }
+    return *this;
+}
+
+ResourceBundle::~ResourceBundle()
+{
+    if(fResource != 0) {
+        ures_close(fResource);
+    }
+    if(fLocale != NULL) {
+      delete(fLocale);
+    }
+}
+
+ResourceBundle *
+ResourceBundle::clone() const {
+    return new ResourceBundle(*this);
+}
+
+UnicodeString ResourceBundle::getString(UErrorCode& status) const {
+    int32_t len = 0;
+    const UChar *r = ures_getString(fResource, &len, &status);
+    return UnicodeString(TRUE, r, len);
+}
+
+const uint8_t *ResourceBundle::getBinary(int32_t& len, UErrorCode& status) const {
+    return ures_getBinary(fResource, &len, &status);
+}
+
+const int32_t *ResourceBundle::getIntVector(int32_t& len, UErrorCode& status) const {
+    return ures_getIntVector(fResource, &len, &status);
+}
+
+uint32_t ResourceBundle::getUInt(UErrorCode& status) const {
+    return ures_getUInt(fResource, &status);
+}
+
+int32_t ResourceBundle::getInt(UErrorCode& status) const {
+    return ures_getInt(fResource, &status);
+}
+
+const char *ResourceBundle::getName(void) const {
+    return ures_getName(fResource);
+}
+
+const char *ResourceBundle::getKey(void) const {
+    return ures_getKey(fResource);
+}
+
+UResType ResourceBundle::getType(void) const {
+    return ures_getType(fResource);
+}
+
+int32_t ResourceBundle::getSize(void) const {
+    return ures_getSize(fResource);
+}
+
+UBool ResourceBundle::hasNext(void) const {
+    return ures_hasNext(fResource);
+}
+
+void ResourceBundle::resetIterator(void) {
+    ures_resetIterator(fResource);
+}
+
+ResourceBundle ResourceBundle::getNext(UErrorCode& status) {
+    UResourceBundle r;
+
+    ures_initStackObject(&r);
+    ures_getNextResource(fResource, &r, &status);
+    ResourceBundle res(&r, status);
+    if (U_SUCCESS(status)) {
+        ures_close(&r);
+    }
+    return res;
+}
+
+UnicodeString ResourceBundle::getNextString(UErrorCode& status) {
+    int32_t len = 0;
+    const UChar* r = ures_getNextString(fResource, &len, 0, &status);
+    return UnicodeString(TRUE, r, len);
+}
+
+UnicodeString ResourceBundle::getNextString(const char ** key, UErrorCode& status) {
+    int32_t len = 0;
+    const UChar* r = ures_getNextString(fResource, &len, key, &status);
+    return UnicodeString(TRUE, r, len);
+}
+
+ResourceBundle ResourceBundle::get(int32_t indexR, UErrorCode& status) const {
+    UResourceBundle r;
+
+    ures_initStackObject(&r);
+    ures_getByIndex(fResource, indexR, &r, &status);
+    ResourceBundle res(&r, status);
+    if (U_SUCCESS(status)) {
+        ures_close(&r);
+    }
+    return res;
+}
+
+UnicodeString ResourceBundle::getStringEx(int32_t indexS, UErrorCode& status) const {
+    int32_t len = 0;
+    const UChar* r = ures_getStringByIndex(fResource, indexS, &len, &status);
+    return UnicodeString(TRUE, r, len);
+}
+
+ResourceBundle ResourceBundle::get(const char* key, UErrorCode& status) const {
+    UResourceBundle r;
+
+    ures_initStackObject(&r);
+    ures_getByKey(fResource, key, &r, &status);
+    ResourceBundle res(&r, status);
+    if (U_SUCCESS(status)) {
+        ures_close(&r);
+    }
+    return res;
+}
+
+ResourceBundle ResourceBundle::getWithFallback(const char* key, UErrorCode& status){
+    UResourceBundle r;
+    ures_initStackObject(&r);
+    ures_getByKeyWithFallback(fResource, key, &r, &status);
+    ResourceBundle res(&r, status);
+    if(U_SUCCESS(status)){
+        ures_close(&r);
+    }
+    return res;
+}
+UnicodeString ResourceBundle::getStringEx(const char* key, UErrorCode& status) const {
+    int32_t len = 0;
+    const UChar* r = ures_getStringByKey(fResource, key, &len, &status);
+    return UnicodeString(TRUE, r, len);
+}
+
+const char*
+ResourceBundle::getVersionNumber()  const
+{
+    return ures_getVersionNumberInternal(fResource);
+}
+
+void ResourceBundle::getVersion(UVersionInfo versionInfo) const {
+    ures_getVersion(fResource, versionInfo);
+}
+
+const Locale &ResourceBundle::getLocale(void) const {
+    static UMutex gLocaleLock;
+    Mutex lock(&gLocaleLock);
+    if (fLocale != NULL) {
+        return *fLocale;
+    }
+    UErrorCode status = U_ZERO_ERROR;
+    const char *localeName = ures_getLocaleInternal(fResource, &status);
+    ResourceBundle *ncThis = const_cast<ResourceBundle *>(this);
+    ncThis->fLocale = new Locale(localeName);
+    return ncThis->fLocale != NULL ? *ncThis->fLocale : Locale::getDefault();
+}
+
+const Locale ResourceBundle::getLocale(ULocDataLocaleType type, UErrorCode &status) const
+{
+  return ures_getLocaleByType(fResource, type, &status);
+}
+
+U_NAMESPACE_END
+//eof
diff --git a/thirdparty/icu4c/common/resbund_cnv.cpp b/thirdparty/icu4c/common/resbund_cnv.cpp
new file mode 100644
index 0000000000..45c0b399bf
--- /dev/null
+++ b/thirdparty/icu4c/common/resbund_cnv.cpp
@@ -0,0 +1,57 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  resbund_cnv.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug25
+*   created by: Markus W. Scherer
+*
+*   Character conversion functions moved here from resbund.cpp
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/resbund.h"
+#include "uinvchar.h"
+
+U_NAMESPACE_BEGIN
+
+ResourceBundle::ResourceBundle( const UnicodeString&    path,
+                                const Locale&           locale,
+                                UErrorCode&              error)
+                                :UObject(), fLocale(NULL)
+{
+    constructForLocale(path, locale, error);
+}
+
+ResourceBundle::ResourceBundle( const UnicodeString&    path,
+                                UErrorCode&              error)
+                                :UObject(), fLocale(NULL)
+{
+    constructForLocale(path, Locale::getDefault(), error);
+}
+
+void 
+ResourceBundle::constructForLocale(const UnicodeString& path,
+                                   const Locale& locale,
+                                   UErrorCode& error)
+{
+    if (path.isEmpty()) {
+        fResource = ures_open(NULL, locale.getName(), &error);
+    }
+    else {
+        UnicodeString nullTerminatedPath(path);
+        nullTerminatedPath.append((UChar)0);
+        fResource = ures_openU(nullTerminatedPath.getBuffer(), locale.getName(), &error);
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/resource.cpp b/thirdparty/icu4c/common/resource.cpp
new file mode 100644
index 0000000000..3d41a16029
--- /dev/null
+++ b/thirdparty/icu4c/common/resource.cpp
@@ -0,0 +1,22 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2015-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* resource.cpp
+*
+* created on: 2015nov04
+* created by: Markus W. Scherer
+*/
+
+#include "resource.h"
+
+U_NAMESPACE_BEGIN
+
+ResourceValue::~ResourceValue() {}
+
+ResourceSink::~ResourceSink() {}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/resource.h b/thirdparty/icu4c/common/resource.h
new file mode 100644
index 0000000000..3795694412
--- /dev/null
+++ b/thirdparty/icu4c/common/resource.h
@@ -0,0 +1,293 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2015-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+*******************************************************************************
+* resource.h
+*
+* created on: 2015nov04
+* created by: Markus W. Scherer
+*/
+
+#ifndef __URESOURCE_H__
+#define __URESOURCE_H__
+
+/**
+ * \file
+ * \brief ICU resource bundle key and value types.
+ */
+
+// Note: Ported from ICU4J class UResource and its nested classes,
+// but the C++ classes are separate, not nested.
+
+// We use the Resource prefix for C++ classes, as usual.
+// The UResource prefix would be used for C types.
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/ures.h"
+#include "restrace.h"
+
+struct ResourceData;
+
+U_NAMESPACE_BEGIN
+
+class ResourceValue;
+
+// Note: In C++, we use const char * pointers for keys,
+// rather than an abstraction like Java UResource.Key.
+
+/**
+ * Interface for iterating over a resource bundle array resource.
+ */
+class U_COMMON_API ResourceArray {
+public:
+    /** Constructs an empty array object. */
+    ResourceArray() : items16(NULL), items32(NULL), length(0) {}
+
+    /** Only for implementation use. @internal */
+    ResourceArray(const uint16_t *i16, const uint32_t *i32, int32_t len,
+                  const ResourceTracer& traceInfo) :
+            items16(i16), items32(i32), length(len),
+            fTraceInfo(traceInfo) {}
+
+    /**
+     * @return The number of items in the array resource.
+     */
+    int32_t getSize() const { return length; }
+    /**
+     * @param i Array item index.
+     * @param value Output-only, receives the value of the i'th item.
+     * @return true if i is non-negative and less than getSize().
+     */
+    UBool getValue(int32_t i, ResourceValue &value) const;
+
+    /** Only for implementation use. @internal */
+    uint32_t internalGetResource(const ResourceData *pResData, int32_t i) const;
+
+private:
+    const uint16_t *items16;
+    const uint32_t *items32;
+    int32_t length;
+    ResourceTracer fTraceInfo;
+};
+
+/**
+ * Interface for iterating over a resource bundle table resource.
+ */
+class U_COMMON_API ResourceTable {
+public:
+    /** Constructs an empty table object. */
+    ResourceTable() : keys16(NULL), keys32(NULL), items16(NULL), items32(NULL), length(0) {}
+
+    /** Only for implementation use. @internal */
+    ResourceTable(const uint16_t *k16, const int32_t *k32,
+                  const uint16_t *i16, const uint32_t *i32, int32_t len,
+                  const ResourceTracer& traceInfo) :
+            keys16(k16), keys32(k32), items16(i16), items32(i32), length(len),
+            fTraceInfo(traceInfo) {}
+
+    /**
+     * @return The number of items in the array resource.
+     */
+    int32_t getSize() const { return length; }
+    /**
+     * @param i Table item index.
+     * @param key Output-only, receives the key of the i'th item.
+     * @param value Output-only, receives the value of the i'th item.
+     * @return true if i is non-negative and less than getSize().
+     */
+    UBool getKeyAndValue(int32_t i, const char *&key, ResourceValue &value) const;
+
+    /**
+     * @param key Key string to find in the table.
+     * @param value Output-only, receives the value of the item with that key.
+     * @return true if the table contains the key.
+     */
+    UBool findValue(const char *key, ResourceValue &value) const;
+
+private:
+    const uint16_t *keys16;
+    const int32_t *keys32;
+    const uint16_t *items16;
+    const uint32_t *items32;
+    int32_t length;
+    ResourceTracer fTraceInfo;
+};
+
+/**
+ * Represents a resource bundle item's value.
+ * Avoids object creations as much as possible.
+ * Mutable, not thread-safe.
+ */
+class U_COMMON_API ResourceValue : public UObject {
+public:
+    virtual ~ResourceValue();
+
+    /**
+     * @return ICU resource type, for example, URES_STRING
+     */
+    virtual UResType getType() const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not a string resource.
+     *
+     * @see ures_getString()
+     */
+    virtual const UChar *getString(int32_t &length, UErrorCode &errorCode) const = 0;
+
+    inline UnicodeString getUnicodeString(UErrorCode &errorCode) const {
+        int32_t len = 0;
+        const UChar *r = getString(len, errorCode);
+        return UnicodeString(true, r, len);
+    }
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an alias resource.
+     */
+    virtual const UChar *getAliasString(int32_t &length, UErrorCode &errorCode) const = 0;
+
+    inline UnicodeString getAliasUnicodeString(UErrorCode &errorCode) const {
+        int32_t len = 0;
+        const UChar *r = getAliasString(len, errorCode);
+        return UnicodeString(true, r, len);
+    }
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an integer resource.
+     *
+     * @see ures_getInt()
+     */
+    virtual int32_t getInt(UErrorCode &errorCode) const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an integer resource.
+     *
+     * @see ures_getUInt()
+     */
+    virtual uint32_t getUInt(UErrorCode &errorCode) const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an intvector resource.
+     *
+     * @see ures_getIntVector()
+     */
+    virtual const int32_t *getIntVector(int32_t &length, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not a binary-blob resource.
+     *
+     * @see ures_getBinary()
+     */
+    virtual const uint8_t *getBinary(int32_t &length, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an array resource
+     */
+    virtual ResourceArray getArray(UErrorCode &errorCode) const = 0;
+
+    /**
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not a table resource
+     */
+    virtual ResourceTable getTable(UErrorCode &errorCode) const = 0;
+
+    /**
+     * Is this a no-fallback/no-inheritance marker string?
+     * Such a marker is used for
+     * CLDR no-fallback data values of (three empty-set symbols)=={2205, 2205, 2205}
+     * when enumerating tables with fallback from the specific resource bundle to root.
+     *
+     * @return true if this is a no-inheritance marker string
+     */
+    virtual UBool isNoInheritanceMarker() const = 0;
+
+    /**
+     * Sets the dest strings from the string values in this array resource.
+     *
+     * @return the number of strings in this array resource.
+     *     If greater than capacity, then an overflow error is set.
+     *
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is not an array resource
+     *     or if any of the array items is not a string
+     */
+    virtual int32_t getStringArray(UnicodeString *dest, int32_t capacity,
+                                   UErrorCode &errorCode) const = 0;
+
+    /**
+     * Same as
+     * <pre>
+     * if (getType() == URES_STRING) {
+     *     return new String[] { getString(); }
+     * } else {
+     *     return getStringArray();
+     * }
+     * </pre>
+     *
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is
+     *     neither a string resource nor an array resource containing strings
+     * @see getString()
+     * @see getStringArray()
+     */
+    virtual int32_t getStringArrayOrStringAsArray(UnicodeString *dest, int32_t capacity,
+                                                  UErrorCode &errorCode) const = 0;
+
+    /**
+     * Same as
+     * <pre>
+     * if (getType() == URES_STRING) {
+     *     return getString();
+     * } else {
+     *     return getStringArray()[0];
+     * }
+     * </pre>
+     *
+     * Sets U_RESOURCE_TYPE_MISMATCH if this is
+     *     neither a string resource nor an array resource containing strings
+     * @see getString()
+     * @see getStringArray()
+     */
+    virtual UnicodeString getStringOrFirstOfArray(UErrorCode &errorCode) const = 0;
+
+protected:
+    ResourceValue() {}
+
+private:
+    ResourceValue(const ResourceValue &);  // no copy constructor
+    ResourceValue &operator=(const ResourceValue &);  // no assignment operator
+};
+
+/**
+ * Sink for ICU resource bundle contents.
+ */
+class U_COMMON_API ResourceSink : public UObject {
+public:
+    ResourceSink() {}
+    virtual ~ResourceSink();
+
+    /**
+     * Called once for each bundle (child-parent-...-root).
+     * The value is normally an array or table resource,
+     * and implementations of this method normally iterate over the
+     * tree of resource items stored there.
+     *
+     * @param key The key string of the enumeration-start resource.
+     *     Empty if the enumeration starts at the top level of the bundle.
+     * @param value Call getArray() or getTable() as appropriate.
+     *     Then reuse for output values from Array and Table getters.
+     * @param noFallback true if the bundle has no parent;
+     *     that is, its top-level table has the nofallback attribute,
+     *     or it is the root bundle of a locale tree.
+     */
+    virtual void put(const char *key, ResourceValue &value, UBool noFallback,
+                     UErrorCode &errorCode) = 0;
+
+private:
+    ResourceSink(const ResourceSink &);  // no copy constructor
+    ResourceSink &operator=(const ResourceSink &);  // no assignment operator
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/restrace.cpp b/thirdparty/icu4c/common/restrace.cpp
new file mode 100644
index 0000000000..5c6498850e
--- /dev/null
+++ b/thirdparty/icu4c/common/restrace.cpp
@@ -0,0 +1,130 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+#include "unicode/utypes.h"
+
+#if U_ENABLE_TRACING
+
+#include "restrace.h"
+#include "charstr.h"
+#include "cstring.h"
+#include "utracimp.h"
+#include "uresimp.h"
+#include "uassert.h"
+#include "util.h"
+
+U_NAMESPACE_BEGIN
+
+ResourceTracer::~ResourceTracer() = default;
+
+void ResourceTracer::trace(const char* resType) const {
+    U_ASSERT(fResB || fParent);
+    UTRACE_ENTRY(UTRACE_UDATA_RESOURCE);
+    UErrorCode status = U_ZERO_ERROR;
+
+    CharString filePath;
+    getFilePath(filePath, status);
+
+    CharString resPath;
+    getResPath(resPath, status);
+
+    // The longest type ("intvector") is 9 chars
+    const char kSpaces[] = "         ";
+    CharString format;
+    format.append(kSpaces, sizeof(kSpaces) - 1 - uprv_strlen(resType), status);
+    format.append("(%s) %s @ %s", status);
+
+    UTRACE_DATA3(UTRACE_VERBOSE,
+        format.data(),
+        resType,
+        filePath.data(),
+        resPath.data());
+    UTRACE_EXIT_STATUS(status);
+}
+
+void ResourceTracer::traceOpen() const {
+    U_ASSERT(fResB);
+    UTRACE_ENTRY(UTRACE_UDATA_BUNDLE);
+    UErrorCode status = U_ZERO_ERROR;
+
+    CharString filePath;
+    UTRACE_DATA1(UTRACE_VERBOSE, "%s", getFilePath(filePath, status).data());
+    UTRACE_EXIT_STATUS(status);
+}
+
+CharString& ResourceTracer::getFilePath(CharString& output, UErrorCode& status) const {
+    if (fResB) {
+        output.append(fResB->fData->fPath, status);
+        output.append('/', status);
+        output.append(fResB->fData->fName, status);
+        output.append(".res", status);
+    } else {
+        fParent->getFilePath(output, status);
+    }
+    return output;
+}
+
+CharString& ResourceTracer::getResPath(CharString& output, UErrorCode& status) const {
+    if (fResB) {
+        output.append('/', status);
+        output.append(fResB->fResPath, status);
+        // removing the trailing /
+        U_ASSERT(output[output.length()-1] == '/');
+        output.truncate(output.length()-1);
+    } else {
+        fParent->getResPath(output, status);
+    }
+    if (fKey) {
+        output.append('/', status);
+        output.append(fKey, status);
+    }
+    if (fIndex != -1) {
+        output.append('[', status);
+        UnicodeString indexString;
+        ICU_Utility::appendNumber(indexString, fIndex);
+        output.appendInvariantChars(indexString, status);
+        output.append(']', status);
+    }
+    return output;
+}
+
+void FileTracer::traceOpen(const char* path, const char* type, const char* name) {
+    if (uprv_strcmp(type, "res") == 0) {
+        traceOpenResFile(path, name);
+    } else {
+        traceOpenDataFile(path, type, name);
+    }
+}
+
+void FileTracer::traceOpenDataFile(const char* path, const char* type, const char* name) {
+    UTRACE_ENTRY(UTRACE_UDATA_DATA_FILE);
+    UErrorCode status = U_ZERO_ERROR;
+
+    CharString filePath;
+    filePath.append(path, status);
+    filePath.append('/', status);
+    filePath.append(name, status);
+    filePath.append('.', status);
+    filePath.append(type, status);
+
+    UTRACE_DATA1(UTRACE_VERBOSE, "%s", filePath.data());
+    UTRACE_EXIT_STATUS(status);
+}
+
+void FileTracer::traceOpenResFile(const char* path, const char* name) {
+    UTRACE_ENTRY(UTRACE_UDATA_RES_FILE);
+    UErrorCode status = U_ZERO_ERROR;
+
+    CharString filePath;
+    filePath.append(path, status);
+    filePath.append('/', status);
+    filePath.append(name, status);
+    filePath.append(".res", status);
+
+    UTRACE_DATA1(UTRACE_VERBOSE, "%s", filePath.data());
+    UTRACE_EXIT_STATUS(status);
+}
+
+U_NAMESPACE_END
+
+#endif // U_ENABLE_TRACING
diff --git a/thirdparty/icu4c/common/restrace.h b/thirdparty/icu4c/common/restrace.h
new file mode 100644
index 0000000000..ef29eaed57
--- /dev/null
+++ b/thirdparty/icu4c/common/restrace.h
@@ -0,0 +1,147 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+#ifndef __RESTRACE_H__
+#define __RESTRACE_H__
+
+#include "unicode/utypes.h"
+
+#if U_ENABLE_TRACING
+
+struct UResourceBundle;
+
+U_NAMESPACE_BEGIN
+
+class CharString;
+
+/**
+ * Instances of this class store information used to trace reads from resource
+ * bundles when ICU is built with --enable-tracing.
+ *
+ * All arguments of type const UResourceBundle*, const char*, and
+ * const ResourceTracer& are stored as pointers. The caller must retain
+ * ownership for the lifetime of this ResourceTracer.
+ *
+ * Exported as U_COMMON_API for Windows because it is a value field
+ * in other exported types.
+ */
+class U_COMMON_API ResourceTracer {
+public:
+    ResourceTracer() :
+        fResB(nullptr),
+        fParent(nullptr),
+        fKey(nullptr),
+        fIndex(-1) {}
+
+    ResourceTracer(const UResourceBundle* resB) :
+        fResB(resB),
+        fParent(nullptr),
+        fKey(nullptr),
+        fIndex(-1) {}
+
+    ResourceTracer(const UResourceBundle* resB, const char* key) :
+        fResB(resB),
+        fParent(nullptr),
+        fKey(key),
+        fIndex(-1) {}
+
+    ResourceTracer(const UResourceBundle* resB, int32_t index) :
+        fResB(resB),
+        fParent(nullptr),
+        fKey(nullptr),
+        fIndex(index) {}
+
+    ResourceTracer(const ResourceTracer& parent, const char* key) :
+        fResB(nullptr),
+        fParent(&parent),
+        fKey(key),
+        fIndex(-1) {}
+
+    ResourceTracer(const ResourceTracer& parent, int32_t index) :
+        fResB(nullptr),
+        fParent(&parent),
+        fKey(nullptr),
+        fIndex(index) {}
+
+    ~ResourceTracer();
+
+    void trace(const char* type) const;
+    void traceOpen() const;
+
+    /**
+     * Calls trace() if the resB or parent provided to the constructor was
+     * non-null; otherwise, does nothing.
+     */
+    void maybeTrace(const char* type) const {
+        if (fResB || fParent) {
+            trace(type);
+        }
+    }
+
+private:
+    const UResourceBundle* fResB;
+    const ResourceTracer* fParent;
+    const char* fKey;
+    int32_t fIndex;
+
+    CharString& getFilePath(CharString& output, UErrorCode& status) const;
+
+    CharString& getResPath(CharString& output, UErrorCode& status) const;
+};
+
+/**
+ * This class provides methods to trace data file reads when ICU is built
+ * with --enable-tracing.
+ */
+class FileTracer {
+public:
+    static void traceOpen(const char* path, const char* type, const char* name);
+
+private:
+    static void traceOpenDataFile(const char* path, const char* type, const char* name);
+    static void traceOpenResFile(const char* path, const char* name);
+};
+
+U_NAMESPACE_END
+
+#else // U_ENABLE_TRACING
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Default trivial implementation when --enable-tracing is not used.
+ */
+class U_COMMON_API ResourceTracer {
+public:
+    ResourceTracer() {}
+
+    ResourceTracer(const void*) {}
+
+    ResourceTracer(const void*, const char*) {}
+
+    ResourceTracer(const void*, int32_t) {}
+
+    ResourceTracer(const ResourceTracer&, const char*) {}
+
+    ResourceTracer(const ResourceTracer&, int32_t) {}
+
+    void trace(const char*) const {}
+
+    void traceOpen() const {}
+
+    void maybeTrace(const char*) const {}
+};
+
+/**
+ * Default trivial implementation when --enable-tracing is not used.
+ */
+class FileTracer {
+public:
+    static void traceOpen(const char*, const char*, const char*) {}
+};
+
+U_NAMESPACE_END
+
+#endif // U_ENABLE_TRACING
+
+#endif //__RESTRACE_H__
diff --git a/thirdparty/icu4c/common/ruleiter.cpp b/thirdparty/icu4c/common/ruleiter.cpp
new file mode 100644
index 0000000000..41eea23c0d
--- /dev/null
+++ b/thirdparty/icu4c/common/ruleiter.cpp
@@ -0,0 +1,162 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2003-2011, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: September 24 2003
+* Since: ICU 2.8
+**********************************************************************
+*/
+#include "ruleiter.h"
+#include "unicode/parsepos.h"
+#include "unicode/symtable.h"
+#include "unicode/unistr.h"
+#include "unicode/utf16.h"
+#include "patternprops.h"
+
+/* \U87654321 or \ud800\udc00 */
+#define MAX_U_NOTATION_LEN 12
+
+U_NAMESPACE_BEGIN
+
+RuleCharacterIterator::RuleCharacterIterator(const UnicodeString& theText, const SymbolTable* theSym,
+                      ParsePosition& thePos) :
+    text(theText),
+    pos(thePos),
+    sym(theSym),
+    buf(0),
+    bufPos(0)
+{}
+
+UBool RuleCharacterIterator::atEnd() const {
+    return buf == 0 && pos.getIndex() == text.length();
+}
+
+UChar32 RuleCharacterIterator::next(int32_t options, UBool& isEscaped, UErrorCode& ec) {
+    if (U_FAILURE(ec)) return DONE;
+
+    UChar32 c = DONE;
+    isEscaped = FALSE;
+
+    for (;;) {
+        c = _current();
+        _advance(U16_LENGTH(c));
+
+        if (c == SymbolTable::SYMBOL_REF && buf == 0 &&
+            (options & PARSE_VARIABLES) != 0 && sym != 0) {
+            UnicodeString name = sym->parseReference(text, pos, text.length());
+            // If name is empty there was an isolated SYMBOL_REF;
+            // return it.  Caller must be prepared for this.
+            if (name.length() == 0) {
+                break;
+            }
+            bufPos = 0;
+            buf = sym->lookup(name);
+            if (buf == 0) {
+                ec = U_UNDEFINED_VARIABLE;
+                return DONE;
+            }
+            // Handle empty variable value
+            if (buf->length() == 0) {
+                buf = 0;
+            }
+            continue;
+        }
+
+        if ((options & SKIP_WHITESPACE) != 0 && PatternProps::isWhiteSpace(c)) {
+            continue;
+        }
+
+        if (c == 0x5C /*'\\'*/ && (options & PARSE_ESCAPES) != 0) {
+            UnicodeString tempEscape;
+            int32_t offset = 0;
+            c = lookahead(tempEscape, MAX_U_NOTATION_LEN).unescapeAt(offset);
+            jumpahead(offset);
+            isEscaped = TRUE;
+            if (c < 0) {
+                ec = U_MALFORMED_UNICODE_ESCAPE;
+                return DONE;
+            }
+        }
+
+        break;
+    }
+
+    return c;
+}
+
+void RuleCharacterIterator::getPos(RuleCharacterIterator::Pos& p) const {
+    p.buf = buf;
+    p.pos = pos.getIndex();
+    p.bufPos = bufPos;
+}
+
+void RuleCharacterIterator::setPos(const RuleCharacterIterator::Pos& p) {
+    buf = p.buf;
+    pos.setIndex(p.pos);
+    bufPos = p.bufPos;
+}
+
+void RuleCharacterIterator::skipIgnored(int32_t options) {
+    if ((options & SKIP_WHITESPACE) != 0) {
+        for (;;) {
+            UChar32 a = _current();
+            if (!PatternProps::isWhiteSpace(a)) break;
+            _advance(U16_LENGTH(a));
+        }
+    }
+}
+
+UnicodeString& RuleCharacterIterator::lookahead(UnicodeString& result, int32_t maxLookAhead) const {
+    if (maxLookAhead < 0) {
+        maxLookAhead = 0x7FFFFFFF;
+    }
+    if (buf != 0) {
+        buf->extract(bufPos, maxLookAhead, result);
+    } else {
+        text.extract(pos.getIndex(), maxLookAhead, result);
+    }
+    return result;
+}
+
+void RuleCharacterIterator::jumpahead(int32_t count) {
+    _advance(count);
+}
+
+/*
+UnicodeString& RuleCharacterIterator::toString(UnicodeString& result) const {
+    int32_t b = pos.getIndex();
+    text.extract(0, b, result);
+    return result.append((UChar) 0x7C).append(text, b, 0x7FFFFFFF); // Insert '|' at index
+}
+*/
+
+UChar32 RuleCharacterIterator::_current() const {
+    if (buf != 0) {
+        return buf->char32At(bufPos);
+    } else {
+        int i = pos.getIndex();
+        return (i < text.length()) ? text.char32At(i) : (UChar32)DONE;
+    }
+}
+
+void RuleCharacterIterator::_advance(int32_t count) {
+    if (buf != 0) {
+        bufPos += count;
+        if (bufPos == buf->length()) {
+            buf = 0;
+        }
+    } else {
+        pos.setIndex(pos.getIndex() + count);
+        if (pos.getIndex() > text.length()) {
+            pos.setIndex(text.length());
+        }
+    }
+}
+
+U_NAMESPACE_END
+
+//eof
diff --git a/thirdparty/icu4c/common/ruleiter.h b/thirdparty/icu4c/common/ruleiter.h
new file mode 100644
index 0000000000..28e2ca5753
--- /dev/null
+++ b/thirdparty/icu4c/common/ruleiter.h
@@ -0,0 +1,233 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2003-2011, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: September 24 2003
+* Since: ICU 2.8
+**********************************************************************
+*/
+#ifndef _RULEITER_H_
+#define _RULEITER_H_
+
+#include "unicode/uobject.h"
+
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+class ParsePosition;
+class SymbolTable;
+
+/**
+ * An iterator that returns 32-bit code points.  This class is deliberately
+ * <em>not</em> related to any of the ICU character iterator classes
+ * in order to minimize complexity.
+ * @author Alan Liu
+ * @since ICU 2.8
+ */
+class RuleCharacterIterator : public UMemory {
+
+    // TODO: Ideas for later.  (Do not implement if not needed, lest the
+    // code coverage numbers go down due to unused methods.)
+    // 1. Add a copy constructor, operator==() method.
+    // 2. Rather than return DONE, throw an exception if the end
+    // is reached -- this is an alternate usage model, probably not useful.
+
+private:
+    /**
+     * Text being iterated.
+     */    
+    const UnicodeString& text;
+
+    /**
+     * Position of iterator.
+     */
+    ParsePosition& pos;
+
+    /**
+     * Symbol table used to parse and dereference variables.  May be 0.
+     */
+    const SymbolTable* sym;
+    
+    /**
+     * Current variable expansion, or 0 if none.
+     */
+    const UnicodeString* buf;
+
+    /**
+     * Position within buf.  Meaningless if buf == 0.
+     */
+    int32_t bufPos;
+
+public:
+    /**
+     * Value returned when there are no more characters to iterate.
+     */
+    enum { DONE = -1 };
+
+    /**
+     * Bitmask option to enable parsing of variable names.  If (options &
+     * PARSE_VARIABLES) != 0, then an embedded variable will be expanded to
+     * its value.  Variables are parsed using the SymbolTable API.
+     */
+    enum { PARSE_VARIABLES = 1 };
+
+    /**
+     * Bitmask option to enable parsing of escape sequences.  If (options &
+     * PARSE_ESCAPES) != 0, then an embedded escape sequence will be expanded
+     * to its value.  Escapes are parsed using Utility.unescapeAt().
+     */
+    enum { PARSE_ESCAPES   = 2 };
+
+    /**
+     * Bitmask option to enable skipping of whitespace.  If (options &
+     * SKIP_WHITESPACE) != 0, then Pattern_White_Space characters will be silently
+     * skipped, as if they were not present in the input.
+     */
+    enum { SKIP_WHITESPACE = 4 };
+
+    /**
+     * Constructs an iterator over the given text, starting at the given
+     * position.
+     * @param text the text to be iterated
+     * @param sym the symbol table, or null if there is none.  If sym is null,
+     * then variables will not be deferenced, even if the PARSE_VARIABLES
+     * option is set.
+     * @param pos upon input, the index of the next character to return.  If a
+     * variable has been dereferenced, then pos will <em>not</em> increment as
+     * characters of the variable value are iterated.
+     */
+    RuleCharacterIterator(const UnicodeString& text, const SymbolTable* sym,
+                          ParsePosition& pos);
+    
+    /**
+     * Returns true if this iterator has no more characters to return.
+     */
+    UBool atEnd() const;
+
+    /**
+     * Returns the next character using the given options, or DONE if there
+     * are no more characters, and advance the position to the next
+     * character.
+     * @param options one or more of the following options, bitwise-OR-ed
+     * together: PARSE_VARIABLES, PARSE_ESCAPES, SKIP_WHITESPACE.
+     * @param isEscaped output parameter set to true if the character
+     * was escaped
+     * @param ec input-output error code.  An error will only be set by
+     * this routing if options includes PARSE_VARIABLES and an unknown
+     * variable name is seen, or if options includes PARSE_ESCAPES and
+     * an invalid escape sequence is seen.
+     * @return the current 32-bit code point, or DONE
+     */
+    UChar32 next(int32_t options, UBool& isEscaped, UErrorCode& ec);
+
+    /**
+     * Returns true if this iterator is currently within a variable expansion.
+     */
+    inline UBool inVariable() const;
+
+    /**
+     * An opaque object representing the position of a RuleCharacterIterator.
+     */
+    struct Pos : public UMemory {
+    private:
+        const UnicodeString* buf;
+        int32_t pos;
+        int32_t bufPos;
+        friend class RuleCharacterIterator;
+    };
+
+    /**
+     * Sets an object which, when later passed to setPos(), will
+     * restore this iterator's position.  Usage idiom:
+     *
+     * RuleCharacterIterator iterator = ...;
+     * RuleCharacterIterator::Pos pos;
+     * iterator.getPos(pos);
+     * for (;;) {
+     *   iterator.getPos(pos);
+     *   int c = iterator.next(...);
+     *   ...
+     * }
+     * iterator.setPos(pos);
+     *
+     * @param p a position object to be set to this iterator's
+     * current position.
+     */
+    void getPos(Pos& p) const;
+
+    /**
+     * Restores this iterator to the position it had when getPos()
+     * set the given object.
+     * @param p a position object previously set by getPos()
+     */
+    void setPos(const Pos& p);
+
+    /**
+     * Skips ahead past any ignored characters, as indicated by the given
+     * options.  This is useful in conjunction with the lookahead() method.
+     *
+     * Currently, this only has an effect for SKIP_WHITESPACE.
+     * @param options one or more of the following options, bitwise-OR-ed
+     * together: PARSE_VARIABLES, PARSE_ESCAPES, SKIP_WHITESPACE.
+     */
+    void skipIgnored(int32_t options);
+
+    /**
+     * Returns a string containing the remainder of the characters to be
+     * returned by this iterator, without any option processing.  If the
+     * iterator is currently within a variable expansion, this will only
+     * extend to the end of the variable expansion.  This method is provided
+     * so that iterators may interoperate with string-based APIs.  The typical
+     * sequence of calls is to call skipIgnored(), then call lookahead(), then
+     * parse the string returned by lookahead(), then call jumpahead() to
+     * resynchronize the iterator.
+     * @param result a string to receive the characters to be returned
+     * by future calls to next()
+     * @param maxLookAhead The maximum to copy into the result.
+     * @return a reference to result
+     */
+    UnicodeString& lookahead(UnicodeString& result, int32_t maxLookAhead = -1) const;
+
+    /**
+     * Advances the position by the given number of 16-bit code units.
+     * This is useful in conjunction with the lookahead() method.
+     * @param count the number of 16-bit code units to jump over
+     */
+    void jumpahead(int32_t count);
+
+    /**
+     * Returns a string representation of this object, consisting of the
+     * characters being iterated, with a '|' marking the current position.
+     * Position within an expanded variable is <em>not</em> indicated.
+     * @param result output parameter to receive a string
+     * representation of this object
+     */
+//    UnicodeString& toString(UnicodeString& result) const;
+    
+private:
+    /**
+     * Returns the current 32-bit code point without parsing escapes, parsing
+     * variables, or skipping whitespace.
+     * @return the current 32-bit code point
+     */
+    UChar32 _current() const;
+    
+    /**
+     * Advances the position by the given amount.
+     * @param count the number of 16-bit code units to advance past
+     */
+    void _advance(int32_t count);
+};
+
+inline UBool RuleCharacterIterator::inVariable() const {
+    return buf != 0;
+}
+
+U_NAMESPACE_END
+
+#endif // _RULEITER_H_
+//eof
diff --git a/thirdparty/icu4c/common/schriter.cpp b/thirdparty/icu4c/common/schriter.cpp
new file mode 100644
index 0000000000..17b68aee9d
--- /dev/null
+++ b/thirdparty/icu4c/common/schriter.cpp
@@ -0,0 +1,119 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1998-2012, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File schriter.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*  05/05/99     stephen     Cleaned up.
+******************************************************************************
+*/
+
+#include "utypeinfo.h"  // for 'typeid' to work
+
+#include "unicode/chariter.h"
+#include "unicode/schriter.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(StringCharacterIterator)
+
+StringCharacterIterator::StringCharacterIterator()
+  : UCharCharacterIterator(),
+    text()
+{
+  // NEVER DEFAULT CONSTRUCT!
+}
+
+StringCharacterIterator::StringCharacterIterator(const UnicodeString& textStr)
+  : UCharCharacterIterator(textStr.getBuffer(), textStr.length()),
+    text(textStr)
+{
+    // we had set the input parameter's array, now we need to set our copy's array
+    UCharCharacterIterator::text = this->text.getBuffer();
+}
+
+StringCharacterIterator::StringCharacterIterator(const UnicodeString& textStr,
+                                                 int32_t textPos)
+  : UCharCharacterIterator(textStr.getBuffer(), textStr.length(), textPos),
+    text(textStr)
+{
+    // we had set the input parameter's array, now we need to set our copy's array
+    UCharCharacterIterator::text = this->text.getBuffer();
+}
+
+StringCharacterIterator::StringCharacterIterator(const UnicodeString& textStr,
+                                                 int32_t textBegin,
+                                                 int32_t textEnd,
+                                                 int32_t textPos)
+  : UCharCharacterIterator(textStr.getBuffer(), textStr.length(), textBegin, textEnd, textPos),
+    text(textStr)
+{
+    // we had set the input parameter's array, now we need to set our copy's array
+    UCharCharacterIterator::text = this->text.getBuffer();
+}
+
+StringCharacterIterator::StringCharacterIterator(const StringCharacterIterator& that)
+  : UCharCharacterIterator(that),
+    text(that.text)
+{
+    // we had set the input parameter's array, now we need to set our copy's array
+    UCharCharacterIterator::text = this->text.getBuffer();
+}
+
+StringCharacterIterator::~StringCharacterIterator() {
+}
+
+StringCharacterIterator&
+StringCharacterIterator::operator=(const StringCharacterIterator& that) {
+    UCharCharacterIterator::operator=(that);
+    text = that.text;
+    // we had set the input parameter's array, now we need to set our copy's array
+    UCharCharacterIterator::text = this->text.getBuffer();
+    return *this;
+}
+
+UBool
+StringCharacterIterator::operator==(const ForwardCharacterIterator& that) const {
+    if (this == &that) {
+        return TRUE;
+    }
+
+    // do not call UCharCharacterIterator::operator==()
+    // because that checks for array pointer equality
+    // while we compare UnicodeString objects
+
+    if (typeid(*this) != typeid(that)) {
+        return FALSE;
+    }
+
+    StringCharacterIterator&    realThat = (StringCharacterIterator&)that;
+
+    return text == realThat.text
+        && pos == realThat.pos
+        && begin == realThat.begin
+        && end == realThat.end;
+}
+
+StringCharacterIterator*
+StringCharacterIterator::clone() const {
+    return new StringCharacterIterator(*this);
+}
+
+void
+StringCharacterIterator::setText(const UnicodeString& newText) {
+    text = newText;
+    UCharCharacterIterator::setText(text.getBuffer(), text.length());
+}
+
+void
+StringCharacterIterator::getText(UnicodeString& result) {
+    result = text;
+}
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/serv.cpp b/thirdparty/icu4c/common/serv.cpp
new file mode 100644
index 0000000000..ce545b9dbd
--- /dev/null
+++ b/thirdparty/icu4c/common/serv.cpp
@@ -0,0 +1,982 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+*******************************************************************************
+* Copyright (C) 2001-2014, International Business Machines Corporation.
+* All Rights Reserved.
+*******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "serv.h"
+#include "umutex.h"
+
+#undef SERVICE_REFCOUNT
+
+// in case we use the refcount stuff
+
+U_NAMESPACE_BEGIN
+
+/*
+******************************************************************
+*/
+
+const UChar ICUServiceKey::PREFIX_DELIMITER = 0x002F;   /* '/' */
+
+ICUServiceKey::ICUServiceKey(const UnicodeString& id) 
+: _id(id) {
+}
+
+ICUServiceKey::~ICUServiceKey() 
+{
+}
+
+const UnicodeString& 
+ICUServiceKey::getID() const 
+{
+    return _id;
+}
+
+UnicodeString& 
+ICUServiceKey::canonicalID(UnicodeString& result) const 
+{
+    return result.append(_id);
+}
+
+UnicodeString& 
+ICUServiceKey::currentID(UnicodeString& result) const 
+{
+    return canonicalID(result);
+}
+
+UnicodeString& 
+ICUServiceKey::currentDescriptor(UnicodeString& result) const 
+{
+    prefix(result);
+    result.append(PREFIX_DELIMITER);
+    return currentID(result);
+}
+
+UBool 
+ICUServiceKey::fallback() 
+{
+    return FALSE;
+}
+
+UBool 
+ICUServiceKey::isFallbackOf(const UnicodeString& id) const 
+{
+    return id == _id;
+}
+
+UnicodeString& 
+ICUServiceKey::prefix(UnicodeString& result) const 
+{
+    return result;
+}
+
+UnicodeString& 
+ICUServiceKey::parsePrefix(UnicodeString& result) 
+{
+    int32_t n = result.indexOf(PREFIX_DELIMITER);
+    if (n < 0) {
+        n = 0;
+    }
+    result.remove(n);
+    return result;
+}
+
+UnicodeString& 
+ICUServiceKey::parseSuffix(UnicodeString& result) 
+{
+    int32_t n = result.indexOf(PREFIX_DELIMITER);
+    if (n >= 0) {
+        result.remove(0, n+1);
+    }
+    return result;
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString& 
+ICUServiceKey::debug(UnicodeString& result) const 
+{
+    debugClass(result);
+    result.append((UnicodeString)" id: ");
+    result.append(_id);
+    return result;
+}
+
+UnicodeString& 
+ICUServiceKey::debugClass(UnicodeString& result) const 
+{
+    return result.append((UnicodeString)"ICUServiceKey");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ICUServiceKey)
+
+/*
+******************************************************************
+*/
+
+ICUServiceFactory::~ICUServiceFactory() {}
+
+SimpleFactory::SimpleFactory(UObject* instanceToAdopt, const UnicodeString& id, UBool visible) 
+: _instance(instanceToAdopt), _id(id), _visible(visible)
+{
+}
+
+SimpleFactory::~SimpleFactory() 
+{
+    delete _instance;
+}
+
+UObject* 
+SimpleFactory::create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const 
+{
+    if (U_SUCCESS(status)) {
+        UnicodeString temp;
+        if (_id == key.currentID(temp)) {
+            return service->cloneInstance(_instance); 
+        }
+    }
+    return NULL;
+}
+
+void 
+SimpleFactory::updateVisibleIDs(Hashtable& result, UErrorCode& status) const 
+{
+    if (_visible) {
+        result.put(_id, (void*)this, status); // cast away const
+    } else {
+        result.remove(_id);
+    }
+}
+
+UnicodeString& 
+SimpleFactory::getDisplayName(const UnicodeString& id, const Locale& /* locale */, UnicodeString& result) const 
+{
+    if (_visible && _id == id) {
+        result = _id;
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString& 
+SimpleFactory::debug(UnicodeString& toAppendTo) const 
+{
+    debugClass(toAppendTo);
+    toAppendTo.append((UnicodeString)" id: ");
+    toAppendTo.append(_id);
+    toAppendTo.append((UnicodeString)", visible: ");
+    toAppendTo.append(_visible ? (UnicodeString)"T" : (UnicodeString)"F");
+    return toAppendTo;
+}
+
+UnicodeString& 
+SimpleFactory::debugClass(UnicodeString& toAppendTo) const 
+{
+    return toAppendTo.append((UnicodeString)"SimpleFactory");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(SimpleFactory)
+
+/*
+******************************************************************
+*/
+
+ServiceListener::~ServiceListener() {}
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ServiceListener)
+
+/*
+******************************************************************
+*/
+
+// Record the actual id for this service in the cache, so we can return it
+// even if we succeed later with a different id.
+class CacheEntry : public UMemory {
+private:
+    int32_t refcount;
+
+public:
+    UnicodeString actualDescriptor;
+    UObject* service;
+
+    /**
+    * Releases a reference to the shared resource.
+    */
+    ~CacheEntry() {
+        delete service;
+    }
+
+    CacheEntry(const UnicodeString& _actualDescriptor, UObject* _service) 
+        : refcount(1), actualDescriptor(_actualDescriptor), service(_service) {
+    }
+
+    /**
+    * Instantiation creates an initial reference, so don't call this
+    * unless you're creating a new pointer to this.  Management of
+    * that pointer will have to know how to deal with refcounts.  
+    * Return true if the resource has not already been released.
+    */
+    CacheEntry* ref() {
+        ++refcount;
+        return this;
+    }
+
+    /**
+    * Destructions removes a reference, so don't call this unless
+    * you're removing pointer to this somewhere.  Management of that
+    * pointer will have to know how to deal with refcounts.  Once
+    * the refcount drops to zero, the resource is released.  Return
+    * false if the resouce has been released.
+    */
+    CacheEntry* unref() {
+        if ((--refcount) == 0) {
+            delete this;
+            return NULL;
+        }
+        return this;
+    }
+
+    /**
+    * Return TRUE if there is at least one reference to this and the
+    * resource has not been released.
+    */
+    UBool isShared() const {
+        return refcount > 1;
+    }
+};
+
+// UObjectDeleter for serviceCache
+U_CDECL_BEGIN
+static void U_CALLCONV
+cacheDeleter(void* obj) {
+    U_NAMESPACE_USE ((CacheEntry*)obj)->unref();
+}
+
+/**
+* Deleter for UObjects
+*/
+static void U_CALLCONV
+deleteUObject(void *obj) {
+    U_NAMESPACE_USE delete (UObject*) obj;
+}
+U_CDECL_END
+
+/*
+******************************************************************
+*/
+
+class DNCache : public UMemory {
+public:
+    Hashtable cache;
+    const Locale locale;
+
+    DNCache(const Locale& _locale) 
+        : cache(), locale(_locale) 
+    {
+        // cache.setKeyDeleter(uprv_deleteUObject);
+    }
+};
+
+
+/*
+******************************************************************
+*/
+
+StringPair* 
+StringPair::create(const UnicodeString& displayName, 
+                   const UnicodeString& id,
+                   UErrorCode& status)
+{
+    if (U_SUCCESS(status)) {
+        StringPair* sp = new StringPair(displayName, id);
+        if (sp == NULL || sp->isBogus()) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            delete sp;
+            return NULL;
+        }
+        return sp;
+    }
+    return NULL;
+}
+
+UBool 
+StringPair::isBogus() const {
+    return displayName.isBogus() || id.isBogus();
+}
+
+StringPair::StringPair(const UnicodeString& _displayName, 
+                       const UnicodeString& _id)
+: displayName(_displayName)
+, id(_id)
+{
+}
+
+U_CDECL_BEGIN
+static void U_CALLCONV
+userv_deleteStringPair(void *obj) {
+    U_NAMESPACE_USE delete (StringPair*) obj;
+}
+U_CDECL_END
+
+/*
+******************************************************************
+*/
+
+static UMutex lock;
+
+ICUService::ICUService()
+: name()
+, timestamp(0)
+, factories(NULL)
+, serviceCache(NULL)
+, idCache(NULL)
+, dnCache(NULL)
+{
+}
+
+ICUService::ICUService(const UnicodeString& newName) 
+: name(newName)
+, timestamp(0)
+, factories(NULL)
+, serviceCache(NULL)
+, idCache(NULL)
+, dnCache(NULL)
+{
+}
+
+ICUService::~ICUService()
+{
+    {
+        Mutex mutex(&lock);
+        clearCaches();
+        delete factories;
+        factories = NULL;
+    }
+}
+
+UObject* 
+ICUService::get(const UnicodeString& descriptor, UErrorCode& status) const 
+{
+    return get(descriptor, NULL, status);
+}
+
+UObject* 
+ICUService::get(const UnicodeString& descriptor, UnicodeString* actualReturn, UErrorCode& status) const 
+{
+    UObject* result = NULL;
+    ICUServiceKey* key = createKey(&descriptor, status);
+    if (key) {
+        result = getKey(*key, actualReturn, status);
+        delete key;
+    }
+    return result;
+}
+
+UObject* 
+ICUService::getKey(ICUServiceKey& key, UErrorCode& status) const 
+{
+    return getKey(key, NULL, status);
+}
+
+// this is a vector that subclasses of ICUService can override to further customize the result object
+// before returning it.  All other public get functions should call this one.
+
+UObject* 
+ICUService::getKey(ICUServiceKey& key, UnicodeString* actualReturn, UErrorCode& status) const 
+{
+    return getKey(key, actualReturn, NULL, status);
+}
+
+// make it possible to call reentrantly on systems that don't have reentrant mutexes.
+// we can use this simple approach since we know the situation where we're calling
+// reentrantly even without knowing the thread.
+class XMutex : public UMemory {
+public:
+    inline XMutex(UMutex *mutex, UBool reentering) 
+        : fMutex(mutex)
+        , fActive(!reentering) 
+    {
+        if (fActive) umtx_lock(fMutex);
+    }
+    inline ~XMutex() {
+        if (fActive) umtx_unlock(fMutex);
+    }
+
+private:
+    UMutex  *fMutex;
+    UBool fActive;
+};
+
+struct UVectorDeleter {
+    UVector* _obj;
+    UVectorDeleter() : _obj(NULL) {}
+    ~UVectorDeleter() { delete _obj; }
+};
+
+// called only by factories, treat as private
+UObject* 
+ICUService::getKey(ICUServiceKey& key, UnicodeString* actualReturn, const ICUServiceFactory* factory, UErrorCode& status) const 
+{
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+
+    if (isDefault()) {
+        return handleDefault(key, actualReturn, status);
+    }
+
+    ICUService* ncthis = (ICUService*)this; // cast away semantic const
+
+    CacheEntry* result = NULL;
+    {
+        // The factory list can't be modified until we're done, 
+        // otherwise we might update the cache with an invalid result.
+        // The cache has to stay in synch with the factory list.
+        // ICU doesn't have monitors so we can't use rw locks, so 
+        // we single-thread everything using this service, for now.
+
+        // if factory is not null, we're calling from within the mutex,
+        // and since some unix machines don't have reentrant mutexes we
+        // need to make sure not to try to lock it again.
+        XMutex mutex(&lock, factory != NULL);
+
+        if (serviceCache == NULL) {
+            ncthis->serviceCache = new Hashtable(status);
+            if (ncthis->serviceCache == NULL) {
+                return NULL;
+            }
+            if (U_FAILURE(status)) {
+                delete serviceCache;
+                return NULL;
+            }
+            serviceCache->setValueDeleter(cacheDeleter);
+        }
+
+        UnicodeString currentDescriptor;
+        UVectorDeleter cacheDescriptorList;
+        UBool putInCache = FALSE;
+
+        int32_t startIndex = 0;
+        int32_t limit = factories->size();
+        UBool cacheResult = TRUE;
+
+        if (factory != NULL) {
+            for (int32_t i = 0; i < limit; ++i) {
+                if (factory == (const ICUServiceFactory*)factories->elementAt(i)) {
+                    startIndex = i + 1;
+                    break;
+                }
+            }
+            if (startIndex == 0) {
+                // throw new InternalError("Factory " + factory + "not registered with service: " + this);
+                status = U_ILLEGAL_ARGUMENT_ERROR;
+                return NULL;
+            }
+            cacheResult = FALSE;
+        }
+
+        do {
+            currentDescriptor.remove();
+            key.currentDescriptor(currentDescriptor);
+            result = (CacheEntry*)serviceCache->get(currentDescriptor);
+            if (result != NULL) {
+                break;
+            }
+
+            // first test of cache failed, so we'll have to update
+            // the cache if we eventually succeed-- that is, if we're 
+            // going to update the cache at all.
+            putInCache = TRUE;
+
+            int32_t index = startIndex;
+            while (index < limit) {
+                ICUServiceFactory* f = (ICUServiceFactory*)factories->elementAt(index++);
+                UObject* service = f->create(key, this, status);
+                if (U_FAILURE(status)) {
+                    delete service;
+                    return NULL;
+                }
+                if (service != NULL) {
+                    result = new CacheEntry(currentDescriptor, service);
+                    if (result == NULL) {
+                        delete service;
+                        status = U_MEMORY_ALLOCATION_ERROR;
+                        return NULL;
+                    }
+
+                    goto outerEnd;
+                }
+            }
+
+            // prepare to load the cache with all additional ids that 
+            // will resolve to result, assuming we'll succeed.  We
+            // don't want to keep querying on an id that's going to
+            // fallback to the one that succeeded, we want to hit the
+            // cache the first time next goaround.
+            if (cacheDescriptorList._obj == NULL) {
+                cacheDescriptorList._obj = new UVector(uprv_deleteUObject, NULL, 5, status);
+                if (U_FAILURE(status)) {
+                    return NULL;
+                }
+            }
+            UnicodeString* idToCache = new UnicodeString(currentDescriptor);
+            if (idToCache == NULL || idToCache->isBogus()) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+                return NULL;
+            }
+
+            cacheDescriptorList._obj->addElement(idToCache, status);
+            if (U_FAILURE(status)) {
+                return NULL;
+            }
+        } while (key.fallback());
+outerEnd:
+
+        if (result != NULL) {
+            if (putInCache && cacheResult) {
+                serviceCache->put(result->actualDescriptor, result, status);
+                if (U_FAILURE(status)) {
+                    return NULL;
+                }
+
+                if (cacheDescriptorList._obj != NULL) {
+                    for (int32_t i = cacheDescriptorList._obj->size(); --i >= 0;) {
+                        UnicodeString* desc = (UnicodeString*)cacheDescriptorList._obj->elementAt(i);
+
+                        serviceCache->put(*desc, result, status);
+                        if (U_FAILURE(status)) {
+                            return NULL;
+                        }
+
+                        result->ref();
+                        cacheDescriptorList._obj->removeElementAt(i);
+                    }
+                }
+            }
+
+            if (actualReturn != NULL) {
+                // strip null prefix
+                if (result->actualDescriptor.indexOf((UChar)0x2f) == 0) { // U+002f=slash (/)
+                    actualReturn->remove();
+                    actualReturn->append(result->actualDescriptor, 
+                        1, 
+                        result->actualDescriptor.length() - 1);
+                } else {
+                    *actualReturn = result->actualDescriptor;
+                }
+
+                if (actualReturn->isBogus()) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                    delete result;
+                    return NULL;
+                }
+            }
+
+            UObject* service = cloneInstance(result->service);
+            if (putInCache && !cacheResult) {
+                delete result;
+            }
+            return service;
+        }
+    }
+
+    return handleDefault(key, actualReturn, status);
+}
+
+UObject* 
+ICUService::handleDefault(const ICUServiceKey& /* key */, UnicodeString* /* actualIDReturn */, UErrorCode& /* status */) const 
+{
+    return NULL;
+}
+
+UVector& 
+ICUService::getVisibleIDs(UVector& result, UErrorCode& status) const {
+    return getVisibleIDs(result, NULL, status);
+}
+
+UVector& 
+ICUService::getVisibleIDs(UVector& result, const UnicodeString* matchID, UErrorCode& status) const 
+{
+    result.removeAllElements();
+
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    {
+        Mutex mutex(&lock);
+        const Hashtable* map = getVisibleIDMap(status);
+        if (map != NULL) {
+            ICUServiceKey* fallbackKey = createKey(matchID, status);
+
+            for (int32_t pos = UHASH_FIRST;;) {
+                const UHashElement* e = map->nextElement(pos);
+                if (e == NULL) {
+                    break;
+                }
+
+                const UnicodeString* id = (const UnicodeString*)e->key.pointer;
+                if (fallbackKey != NULL) {
+                    if (!fallbackKey->isFallbackOf(*id)) {
+                        continue;
+                    }
+                }
+
+                UnicodeString* idClone = new UnicodeString(*id);
+                if (idClone == NULL || idClone->isBogus()) {
+                    delete idClone;
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                    break;
+                }
+                result.addElement(idClone, status);
+                if (U_FAILURE(status)) {
+                    delete idClone;
+                    break;
+                }
+            }
+            delete fallbackKey;
+        }
+    }
+    if (U_FAILURE(status)) {
+        result.removeAllElements();
+    }
+    return result;
+}
+
+const Hashtable* 
+ICUService::getVisibleIDMap(UErrorCode& status) const {
+    if (U_FAILURE(status)) return NULL;
+
+    // must only be called when lock is already held
+
+    ICUService* ncthis = (ICUService*)this; // cast away semantic const
+    if (idCache == NULL) {
+        ncthis->idCache = new Hashtable(status);
+        if (idCache == NULL) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+        } else if (factories != NULL) {
+            for (int32_t pos = factories->size(); --pos >= 0;) {
+                ICUServiceFactory* f = (ICUServiceFactory*)factories->elementAt(pos);
+                f->updateVisibleIDs(*idCache, status);
+            }
+            if (U_FAILURE(status)) {
+                delete idCache;
+                ncthis->idCache = NULL;
+            }
+        }
+    }
+
+    return idCache;
+}
+
+
+UnicodeString& 
+ICUService::getDisplayName(const UnicodeString& id, UnicodeString& result) const 
+{
+    return getDisplayName(id, result, Locale::getDefault());
+}
+
+UnicodeString& 
+ICUService::getDisplayName(const UnicodeString& id, UnicodeString& result, const Locale& locale) const 
+{
+    {
+        UErrorCode status = U_ZERO_ERROR;
+        Mutex mutex(&lock);
+        const Hashtable* map = getVisibleIDMap(status);
+        if (map != NULL) {
+            ICUServiceFactory* f = (ICUServiceFactory*)map->get(id);
+            if (f != NULL) {
+                f->getDisplayName(id, locale, result);
+                return result;
+            }
+
+            // fallback
+            status = U_ZERO_ERROR;
+            ICUServiceKey* fallbackKey = createKey(&id, status);
+            while (fallbackKey != NULL && fallbackKey->fallback()) {
+                UnicodeString us;
+                fallbackKey->currentID(us);
+                f = (ICUServiceFactory*)map->get(us);
+                if (f != NULL) {
+                    f->getDisplayName(id, locale, result);
+                    delete fallbackKey;
+                    return result;
+                }
+            }
+            delete fallbackKey;
+        }
+    }
+    result.setToBogus();
+    return result;
+}
+
+UVector& 
+ICUService::getDisplayNames(UVector& result, UErrorCode& status) const 
+{
+    return getDisplayNames(result, Locale::getDefault(), NULL, status);
+}
+
+
+UVector& 
+ICUService::getDisplayNames(UVector& result, const Locale& locale, UErrorCode& status) const 
+{
+    return getDisplayNames(result, locale, NULL, status);
+}
+
+UVector& 
+ICUService::getDisplayNames(UVector& result, 
+                            const Locale& locale, 
+                            const UnicodeString* matchID, 
+                            UErrorCode& status) const 
+{
+    result.removeAllElements();
+    result.setDeleter(userv_deleteStringPair);
+    if (U_SUCCESS(status)) {
+        ICUService* ncthis = (ICUService*)this; // cast away semantic const
+        Mutex mutex(&lock);
+
+        if (dnCache != NULL && dnCache->locale != locale) {
+            delete dnCache;
+            ncthis->dnCache = NULL;
+        }
+
+        if (dnCache == NULL) {
+            const Hashtable* m = getVisibleIDMap(status);
+            if (U_FAILURE(status)) {
+                return result;
+            }
+            ncthis->dnCache = new DNCache(locale); 
+            if (dnCache == NULL) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+                return result;
+            }
+
+            int32_t pos = UHASH_FIRST;
+            const UHashElement* entry = NULL;
+            while ((entry = m->nextElement(pos)) != NULL) {
+                const UnicodeString* id = (const UnicodeString*)entry->key.pointer;
+                ICUServiceFactory* f = (ICUServiceFactory*)entry->value.pointer;
+                UnicodeString dname;
+                f->getDisplayName(*id, locale, dname);
+                if (dname.isBogus()) {
+                    status = U_MEMORY_ALLOCATION_ERROR;
+                } else {
+                    dnCache->cache.put(dname, (void*)id, status); // share pointer with visibleIDMap
+                    if (U_SUCCESS(status)) {
+                        continue;
+                    }
+                }
+                delete dnCache;
+                ncthis->dnCache = NULL;
+                return result;
+            }
+        }
+    }
+
+    ICUServiceKey* matchKey = createKey(matchID, status);
+    /* To ensure that all elements in the hashtable are iterated, set pos to -1.
+     * nextElement(pos) will skip the position at pos and begin the iteration
+     * at the next position, which in this case will be 0.
+     */
+    int32_t pos = UHASH_FIRST; 
+    const UHashElement *entry = NULL;
+    while ((entry = dnCache->cache.nextElement(pos)) != NULL) {
+        const UnicodeString* id = (const UnicodeString*)entry->value.pointer;
+        if (matchKey != NULL && !matchKey->isFallbackOf(*id)) {
+            continue;
+        }
+        const UnicodeString* dn = (const UnicodeString*)entry->key.pointer;
+        StringPair* sp = StringPair::create(*id, *dn, status);
+        result.addElement(sp, status);
+        if (U_FAILURE(status)) {
+            result.removeAllElements();
+            break;
+        }
+    }
+    delete matchKey;
+
+    return result;
+}
+
+URegistryKey
+ICUService::registerInstance(UObject* objToAdopt, const UnicodeString& id, UErrorCode& status) 
+{
+    return registerInstance(objToAdopt, id, TRUE, status);
+}
+
+URegistryKey
+ICUService::registerInstance(UObject* objToAdopt, const UnicodeString& id, UBool visible, UErrorCode& status) 
+{
+    ICUServiceKey* key = createKey(&id, status);
+    if (key != NULL) {
+        UnicodeString canonicalID;
+        key->canonicalID(canonicalID);
+        delete key;
+
+        ICUServiceFactory* f = createSimpleFactory(objToAdopt, canonicalID, visible, status);
+        if (f != NULL) {
+            return registerFactory(f, status);
+        }
+    }
+    delete objToAdopt;
+    return NULL;
+}
+
+ICUServiceFactory* 
+ICUService::createSimpleFactory(UObject* objToAdopt, const UnicodeString& id, UBool visible, UErrorCode& status)
+{
+    if (U_SUCCESS(status)) {
+        if ((objToAdopt != NULL) && (!id.isBogus())) {
+            return new SimpleFactory(objToAdopt, id, visible);
+        }
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return NULL;
+}
+
+URegistryKey
+ICUService::registerFactory(ICUServiceFactory* factoryToAdopt, UErrorCode& status) 
+{
+    if (U_SUCCESS(status) && factoryToAdopt != NULL) {
+        Mutex mutex(&lock);
+
+        if (factories == NULL) {
+            factories = new UVector(deleteUObject, NULL, status);
+            if (U_FAILURE(status)) {
+                delete factories;
+                return NULL;
+            }
+        }
+        factories->insertElementAt(factoryToAdopt, 0, status);
+        if (U_SUCCESS(status)) {
+            clearCaches();
+        } else {
+            delete factoryToAdopt;
+            factoryToAdopt = NULL;
+        }
+    }
+
+    if (factoryToAdopt != NULL) {
+        notifyChanged();
+    }
+
+    return (URegistryKey)factoryToAdopt;
+}
+
+UBool 
+ICUService::unregister(URegistryKey rkey, UErrorCode& status) 
+{
+    ICUServiceFactory *factory = (ICUServiceFactory*)rkey;
+    UBool result = FALSE;
+    if (factory != NULL && factories != NULL) {
+        Mutex mutex(&lock);
+
+        if (factories->removeElement(factory)) {
+            clearCaches();
+            result = TRUE;
+        } else {
+            status = U_ILLEGAL_ARGUMENT_ERROR;
+            delete factory;
+        }
+    }
+    if (result) {
+        notifyChanged();
+    }
+    return result;
+}
+
+void 
+ICUService::reset() 
+{
+    {
+        Mutex mutex(&lock);
+        reInitializeFactories();
+        clearCaches();
+    }
+    notifyChanged();
+}
+
+void 
+ICUService::reInitializeFactories() 
+{
+    if (factories != NULL) {
+        factories->removeAllElements();
+    }
+}
+
+UBool 
+ICUService::isDefault() const 
+{
+    return countFactories() == 0;
+}
+
+ICUServiceKey* 
+ICUService::createKey(const UnicodeString* id, UErrorCode& status) const 
+{
+    return (U_FAILURE(status) || id == NULL) ? NULL : new ICUServiceKey(*id);
+}
+
+void 
+ICUService::clearCaches() 
+{
+    // callers synchronize before use
+    ++timestamp;
+    delete dnCache;
+    dnCache = NULL;
+    delete idCache;
+    idCache = NULL;
+    delete serviceCache; serviceCache = NULL;
+}
+
+void 
+ICUService::clearServiceCache() 
+{
+    // callers synchronize before use
+    delete serviceCache; serviceCache = NULL;
+}
+
+UBool 
+ICUService::acceptsListener(const EventListener& l) const 
+{
+    return dynamic_cast<const ServiceListener*>(&l) != NULL;
+}
+
+void 
+ICUService::notifyListener(EventListener& l) const 
+{
+    ((ServiceListener&)l).serviceChanged(*this);
+}
+
+UnicodeString&
+ICUService::getName(UnicodeString& result) const 
+{
+    return result.append(name);
+}
+
+int32_t 
+ICUService::countFactories() const 
+{
+    return factories == NULL ? 0 : factories->size();
+}
+
+int32_t
+ICUService::getTimestamp() const
+{
+    return timestamp;
+}
+
+U_NAMESPACE_END
+
+/* UCONFIG_NO_SERVICE */
+#endif
diff --git a/thirdparty/icu4c/common/serv.h b/thirdparty/icu4c/common/serv.h
new file mode 100644
index 0000000000..ca070b6c6d
--- /dev/null
+++ b/thirdparty/icu4c/common/serv.h
@@ -0,0 +1,996 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2011, International Business Machines Corporation.       *
+ * All Rights Reserved.                                                        *
+ *******************************************************************************
+ */
+
+#ifndef ICUSERV_H
+#define ICUSERV_H
+
+#include "unicode/utypes.h"
+
+#if UCONFIG_NO_SERVICE
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Allow the declaration of APIs with pointers to ICUService
+ * even when service is removed from the build.
+ */
+class ICUService;
+
+U_NAMESPACE_END
+
+#else
+
+#include "unicode/unistr.h"
+#include "unicode/locid.h"
+#include "unicode/umisc.h"
+
+#include "hash.h"
+#include "uvector.h"
+#include "servnotf.h"
+
+class ICUServiceTest;
+
+U_NAMESPACE_BEGIN
+
+class ICUServiceKey;
+class ICUServiceFactory;
+class SimpleFactory;
+class ServiceListener;
+class ICUService;
+
+class DNCache;
+
+/*******************************************************************
+ * ICUServiceKey
+ */
+
+/**
+ * <p>ICUServiceKeys are used to communicate with factories to
+ * generate an instance of the service.  ICUServiceKeys define how
+ * ids are canonicalized, provide both a current id and a current
+ * descriptor to use in querying the cache and factories, and
+ * determine the fallback strategy.</p>
+ *
+ * <p>ICUServiceKeys provide both a currentDescriptor and a currentID.
+ * The descriptor contains an optional prefix, followed by '/'
+ * and the currentID.  Factories that handle complex keys,
+ * for example number format factories that generate multiple
+ * kinds of formatters for the same locale, use the descriptor 
+ * to provide a fully unique identifier for the service object, 
+ * while using the currentID (in this case, the locale string),
+ * as the visible IDs that can be localized.</p>
+ *
+ * <p>The default implementation of ICUServiceKey has no fallbacks and
+ * has no custom descriptors.</p> 
+ */
+class U_COMMON_API ICUServiceKey : public UObject {
+ private: 
+  const UnicodeString _id;
+
+ protected:
+  static const UChar PREFIX_DELIMITER;
+
+ public:
+
+  /**
+   * <p>Construct a key from an id.</p>
+   *
+   * @param id the ID from which to construct the key.
+   */
+  ICUServiceKey(const UnicodeString& id);
+
+  /**
+   * <p>Virtual destructor.</p>
+   */
+  virtual ~ICUServiceKey();
+
+ /**
+  * <p>Return the original ID used to construct this key.</p>
+  *
+  * @return the ID used to construct this key.
+  */
+  virtual const UnicodeString& getID() const;
+
+ /**
+  * <p>Return the canonical version of the original ID.  This implementation
+  * appends the original ID to result.  Result is returned as a convenience.</p>
+  *
+  * @param result the output parameter to which the id will be appended.
+  * @return the modified result.
+  */
+  virtual UnicodeString& canonicalID(UnicodeString& result) const;
+
+ /**
+  * <p>Return the (canonical) current ID.  This implementation appends
+  * the canonical ID to result.  Result is returned as a convenience.</p>
+  *
+  * @param result the output parameter to which the current id will be appended.
+  * @return the modified result.  
+  */
+  virtual UnicodeString& currentID(UnicodeString& result) const;
+
+ /**
+  * <p>Return the current descriptor.  This implementation appends
+  * the current descriptor to result.  Result is returned as a convenience.</p>
+  *
+  * <p>The current descriptor is used to fully
+  * identify an instance of the service in the cache.  A
+  * factory may handle all descriptors for an ID, or just a
+  * particular descriptor.  The factory can either parse the
+  * descriptor or use custom API on the key in order to
+  * instantiate the service.</p>
+  *
+  * @param result the output parameter to which the current id will be appended.
+  * @return the modified result.  
+  */
+  virtual UnicodeString& currentDescriptor(UnicodeString& result) const;
+
+ /**
+  * <p>If the key has a fallback, modify the key and return true,
+  * otherwise return false.  The current ID will change if there
+  * is a fallback.  No currentIDs should be repeated, and fallback
+  * must eventually return false.  This implementation has no fallbacks
+  * and always returns false.</p>
+  *
+  * @return true if the ICUServiceKey changed to a valid fallback value.
+  */
+  virtual UBool fallback();
+
+ /**
+  * <p>Return true if a key created from id matches, or would eventually
+  * fallback to match, the canonical ID of this ICUServiceKey.</p>
+  *
+  * @param id the id to test.
+  * @return true if this ICUServiceKey's canonical ID is a fallback of id.
+  */
+  virtual UBool isFallbackOf(const UnicodeString& id) const;
+
+ /**
+  * <p>Return the prefix.  This implementation leaves result unchanged.
+  * Result is returned as a convenience.</p>
+  *
+  * @param result the output parameter to which the prefix will be appended.
+  * @return the modified result.
+  */
+  virtual UnicodeString& prefix(UnicodeString& result) const;
+
+ /**
+  * <p>A utility to parse the prefix out of a descriptor string.  Only
+  * the (undelimited) prefix, if any, remains in result.  Result is returned as a 
+  * convenience.</p>
+  *
+  * @param result an input/output parameter that on entry is a descriptor, and 
+  * on exit is the prefix of that descriptor.
+  * @return the modified result.
+  */
+  static UnicodeString& parsePrefix(UnicodeString& result);
+
+  /**
+  * <p>A utility to parse the suffix out of a descriptor string.  Only
+  * the (undelimited) suffix, if any, remains in result.  Result is returned as a 
+  * convenience.</p>
+  *
+  * @param result an input/output parameter that on entry is a descriptor, and 
+  * on exit is the suffix of that descriptor.
+  * @return the modified result.
+  */
+  static UnicodeString& parseSuffix(UnicodeString& result);
+
+public:
+  /**
+   * UObject RTTI boilerplate.
+   */
+  static UClassID U_EXPORT2 getStaticClassID();
+
+  /**
+   * UObject RTTI boilerplate.
+   */
+  virtual UClassID getDynamicClassID() const;
+
+#ifdef SERVICE_DEBUG
+ public:
+  virtual UnicodeString& debug(UnicodeString& result) const;
+  virtual UnicodeString& debugClass(UnicodeString& result) const;
+#endif
+
+};
+
+ /*******************************************************************
+  * ICUServiceFactory
+  */
+
+ /**
+  * <p>An implementing ICUServiceFactory generates the service objects maintained by the
+  * service.  A factory generates a service object from a key,
+  * updates id->factory mappings, and returns the display name for
+  * a supported id.</p>
+  */
+class U_COMMON_API ICUServiceFactory : public UObject {
+ public:
+    virtual ~ICUServiceFactory();
+
+    /**
+     * <p>Create a service object from the key, if this factory
+     * supports the key.  Otherwise, return NULL.</p>
+     *
+     * <p>If the factory supports the key, then it can call
+     * the service's getKey(ICUServiceKey, String[], ICUServiceFactory) method
+     * passing itself as the factory to get the object that
+     * the service would have created prior to the factory's
+     * registration with the service.  This can change the
+     * key, so any information required from the key should
+     * be extracted before making such a callback.</p>
+     *
+     * @param key the service key.
+     * @param service the service with which this factory is registered.
+     * @param status the error code status.
+     * @return the service object, or NULL if the factory does not support the key.
+     */
+    virtual UObject* create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const = 0;
+
+    /**
+     * <p>Update result to reflect the IDs (not descriptors) that this
+     * factory publicly handles.  Result contains mappings from ID to
+     * factory.  On entry it will contain all (visible) mappings from
+     * previously-registered factories.</p>
+     *
+     * <p>This function, together with getDisplayName, are used to
+     * support ICUService::getDisplayNames.  The factory determines
+     * which IDs (of those it supports) it will make visible, and of
+     * those, which it will provide localized display names for.  In
+     * most cases it will register mappings from all IDs it supports
+     * to itself.</p>
+     *
+     * @param result the mapping table to update.
+     * @param status the error code status.
+     */
+    virtual void updateVisibleIDs(Hashtable& result, UErrorCode& status) const = 0;
+
+    /**
+     * <p>Return, in result, the display name of the id in the provided locale.
+     * This is an id, not a descriptor.  If the id is 
+     * not visible, sets result to bogus.  If the
+     * incoming result is bogus, it remains bogus.  Result is returned as a
+     * convenience.  Results are not defined if id is not one supported by this
+         * factory.</p>
+     *
+     * @param id a visible id supported by this factory.
+     * @param locale the locale for which to generate the corresponding localized display name.
+     * @param result output parameter to hold the display name.
+     * @return result.
+     */
+    virtual UnicodeString& getDisplayName(const UnicodeString& id, const Locale& locale, UnicodeString& result) const = 0;
+};
+
+/*
+ ******************************************************************
+ */
+
+ /**
+  * <p>A default implementation of factory.  This provides default
+  * implementations for subclasses, and implements a singleton
+  * factory that matches a single ID and returns a single
+  * (possibly deferred-initialized) instance.  This implements
+  * updateVisibleIDs to add a mapping from its ID to itself
+  * if visible is true, or to remove any existing mapping
+  * for its ID if visible is false.  No localization of display
+  * names is performed.</p>
+  */
+class U_COMMON_API SimpleFactory : public ICUServiceFactory {
+ protected:
+  UObject* _instance;
+  const UnicodeString _id;
+  const UBool _visible;
+
+ public:
+  /**
+   * <p>Construct a SimpleFactory that maps a single ID to a single 
+   * service instance.  If visible is true, the ID will be visible.
+   * The instance must not be NULL.  The SimpleFactory will adopt
+   * the instance, which must not be changed subsequent to this call.</p>
+   *
+   * @param instanceToAdopt the service instance to adopt.
+   * @param id the ID to assign to this service instance.
+   * @param visible if true, the ID will be visible.
+   */
+  SimpleFactory(UObject* instanceToAdopt, const UnicodeString& id, UBool visible = true);
+
+  /**
+   * <p>Destructor.</p>
+   */
+  virtual ~SimpleFactory();
+
+  /**
+   * <p>This implementation returns a clone of the service instance if the factory's ID is equal to
+   * the key's currentID.  Service and prefix are ignored.</p>
+   *
+   * @param key the service key.
+   * @param service the service with which this factory is registered.
+   * @param status the error code status.
+   * @return the service object, or NULL if the factory does not support the key.
+   */
+  virtual UObject* create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const;
+
+  /**
+   * <p>This implementation adds a mapping from ID -> this to result if visible is true, 
+   * otherwise it removes ID from result.</p>
+   *
+   * @param result the mapping table to update.
+   * @param status the error code status.
+   */
+  virtual void updateVisibleIDs(Hashtable& result, UErrorCode& status) const;
+
+  /**
+   * <p>This implementation returns the factory ID if it equals id and visible is true,
+   * otherwise it returns the empty string.  (This implementation provides
+   * no localized id information.)</p>
+   *
+   * @param id a visible id supported by this factory.
+   * @param locale the locale for which to generate the corresponding localized display name.
+   * @param result output parameter to hold the display name.
+   * @return result.
+   */
+  virtual UnicodeString& getDisplayName(const UnicodeString& id, const Locale& locale, UnicodeString& result) const;
+
+public:
+ /**
+  * UObject RTTI boilerplate.
+  */
+  static UClassID U_EXPORT2 getStaticClassID();
+
+ /**
+  * UObject RTTI boilerplate.
+  */
+  virtual UClassID getDynamicClassID() const;
+
+#ifdef SERVICE_DEBUG
+ public:
+  virtual UnicodeString& debug(UnicodeString& toAppendTo) const;
+  virtual UnicodeString& debugClass(UnicodeString& toAppendTo) const;
+#endif
+
+};
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * <p>ServiceListener is the listener that ICUService provides by default.
+ * ICUService will notifiy this listener when factories are added to
+ * or removed from the service.  Subclasses can provide
+ * different listener interfaces that extend EventListener, and modify
+ * acceptsListener and notifyListener as appropriate.</p>
+ */
+class U_COMMON_API ServiceListener : public EventListener {
+public:
+    virtual ~ServiceListener();
+
+    /**
+     * <p>This method is called when the service changes. At the time of the
+     * call this listener is registered with the service.  It must
+     * not modify the notifier in the context of this call.</p>
+     * 
+     * @param service the service that changed.
+     */
+    virtual void serviceChanged(const ICUService& service) const = 0;
+    
+public:
+    /**
+     * UObject RTTI boilerplate.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+    
+    /**
+     * UObject RTTI boilerplate.
+     */
+    virtual UClassID getDynamicClassID() const;
+    
+};
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * <p>A StringPair holds a displayName/ID pair.  ICUService uses it
+ * as the array elements returned by getDisplayNames.
+ */
+class U_COMMON_API StringPair : public UMemory {
+public:
+  /**
+   * <p>The display name of the pair.</p>
+   */
+  const UnicodeString displayName;
+
+  /**
+   * <p>The ID of the pair.</p>
+   */
+  const UnicodeString id;
+
+  /**
+   * <p>Creates a string pair from a displayName and an ID.</p>
+   *
+   * @param displayName the displayName.
+   * @param id the ID.
+   * @param status the error code status.
+   * @return a StringPair if the creation was successful, otherwise NULL.
+   */
+  static StringPair* create(const UnicodeString& displayName, 
+                            const UnicodeString& id,
+                            UErrorCode& status);
+
+  /**
+   * <p>Return true if either string of the pair is bogus.</p>
+   * @return true if either string of the pair is bogus.
+   */
+  UBool isBogus() const;
+
+private:
+  StringPair(const UnicodeString& displayName, const UnicodeString& id);
+};
+
+/*******************************************************************
+ * ICUService
+ */
+
+ /**
+ * <p>A Service provides access to service objects that implement a
+ * particular service, e.g. transliterators.  Users provide a String
+ * id (for example, a locale string) to the service, and get back an
+ * object for that id.  Service objects can be any kind of object.  A
+ * new service object is returned for each query. The caller is
+ * responsible for deleting it.</p>
+ *
+ * <p>Services 'canonicalize' the query ID and use the canonical ID to
+ * query for the service.  The service also defines a mechanism to
+ * 'fallback' the ID multiple times.  Clients can optionally request
+ * the actual ID that was matched by a query when they use an ID to
+ * retrieve a service object.</p>
+ *
+ * <p>Service objects are instantiated by ICUServiceFactory objects
+ * registered with the service.  The service queries each
+ * ICUServiceFactory in turn, from most recently registered to
+ * earliest registered, until one returns a service object.  If none
+ * responds with a service object, a fallback ID is generated, and the
+ * process repeats until a service object is returned or until the ID
+ * has no further fallbacks.</p>
+ *
+ * <p>In ICU 2.4, UObject (the base class of service instances) does
+ * not define a polymorphic clone function.  ICUService uses clones to
+ * manage ownership.  Thus, for now, ICUService defines an abstract
+ * method, cloneInstance, that clients must implement to create clones
+ * of the service instances.  This may change in future releases of
+ * ICU.</p>
+ *
+ * <p>ICUServiceFactories can be dynamically registered and
+ * unregistered with the service.  When registered, an
+ * ICUServiceFactory is installed at the head of the factory list, and
+ * so gets 'first crack' at any keys or fallback keys.  When
+ * unregistered, it is removed from the service and can no longer be
+ * located through it.  Service objects generated by this factory and
+ * held by the client are unaffected.</p>
+ *
+ * <p>If a service has variants (e.g., the different variants of
+ * BreakIterator) an ICUServiceFactory can use the prefix of the
+ * ICUServiceKey to determine the variant of a service to generate.
+ * If it does not support all variants, it can request
+ * previously-registered factories to handle the ones it does not
+ * support.</p>
+ *
+ * <p>ICUService uses ICUServiceKeys to query factories and perform
+ * fallback.  The ICUServiceKey defines the canonical form of the ID,
+ * and implements the fallback strategy.  Custom ICUServiceKeys can be
+ * defined that parse complex IDs into components that
+ * ICUServiceFactories can more easily use.  The ICUServiceKey can
+ * cache the results of this parsing to save repeated effort.
+ * ICUService provides convenience APIs that take UnicodeStrings and
+ * generate default ICUServiceKeys for use in querying.</p>
+ *
+ * <p>ICUService provides API to get the list of IDs publicly
+ * supported by the service (although queries aren't restricted to
+ * this list).  This list contains only 'simple' IDs, and not fully
+ * unique IDs.  ICUServiceFactories are associated with each simple ID
+ * and the responsible factory can also return a human-readable
+ * localized version of the simple ID, for use in user interfaces.
+ * ICUService can also provide an array of the all the localized
+ * visible IDs and their corresponding internal IDs.</p>
+ *
+ * <p>ICUService implements ICUNotifier, so that clients can register
+ * to receive notification when factories are added or removed from
+ * the service.  ICUService provides a default EventListener
+ * subinterface, ServiceListener, which can be registered with the
+ * service.  When the service changes, the ServiceListener's
+ * serviceChanged method is called with the service as the
+ * argument.</p>
+ *
+ * <p>The ICUService API is both rich and generic, and it is expected
+ * that most implementations will statically 'wrap' ICUService to
+ * present a more appropriate API-- for example, to declare the type
+ * of the objects returned from get, to limit the factories that can
+ * be registered with the service, or to define their own listener
+ * interface with a custom callback method.  They might also customize
+ * ICUService by overriding it, for example, to customize the
+ * ICUServiceKey and fallback strategy.  ICULocaleService is a
+ * subclass of ICUService that uses Locale names as IDs and uses
+ * ICUServiceKeys that implement the standard resource bundle fallback
+ * strategy.  Most clients will wish to subclass it instead of
+ * ICUService.</p> 
+ */
+class U_COMMON_API ICUService : public ICUNotifier {
+ protected: 
+    /**
+     * Name useful for debugging.
+     */
+    const UnicodeString name;
+
+ private:
+
+    /**
+     * Timestamp so iterators can be fail-fast.
+     */
+    uint32_t timestamp;
+
+    /**
+     * All the factories registered with this service.
+     */
+    UVector* factories;
+
+    /**
+     * The service cache.
+     */
+    Hashtable* serviceCache;
+
+    /**
+     * The ID cache.
+     */
+    Hashtable* idCache;
+
+    /**
+     * The name cache.
+     */
+    DNCache* dnCache;
+
+    /**
+     * Constructor.
+     */
+ public:
+    /**
+     * <p>Construct a new ICUService.</p>
+     */
+    ICUService();
+
+    /**
+     * <p>Construct with a name (useful for debugging).</p>
+     *
+     * @param name a name to use in debugging.
+     */
+    ICUService(const UnicodeString& name);
+
+    /**
+     * <p>Destructor.</p>
+     */
+    virtual ~ICUService();
+
+    /**
+     * <p>Return the name of this service. This will be the empty string if none was assigned.
+     * Returns result as a convenience.</p>
+     *
+     * @param result an output parameter to contain the name of this service.
+     * @return the name of this service.
+     */
+    UnicodeString& getName(UnicodeString& result) const;
+
+    /**
+     * <p>Convenience override for get(ICUServiceKey&, UnicodeString*). This uses
+     * createKey to create a key for the provided descriptor.</p>
+     *
+     * @param descriptor the descriptor.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    UObject* get(const UnicodeString& descriptor, UErrorCode& status) const;
+
+    /**
+     * <p>Convenience override for get(ICUServiceKey&, UnicodeString*).  This uses
+     * createKey to create a key from the provided descriptor.</p>
+     *
+     * @param descriptor the descriptor.
+     * @param actualReturn a pointer to a UnicodeString to hold the matched descriptor, or NULL.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    UObject* get(const UnicodeString& descriptor, UnicodeString* actualReturn, UErrorCode& status) const;
+
+    /**
+     * <p>Convenience override for get(ICUServiceKey&, UnicodeString*).</p>
+     *
+     * @param key the key.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    UObject* getKey(ICUServiceKey& key, UErrorCode& status) const;
+
+    /**
+     * <p>Given a key, return a service object, and, if actualReturn
+     * is not NULL, the descriptor with which it was found in the
+     * first element of actualReturn.  If no service object matches
+     * this key, returns NULL and leaves actualReturn unchanged.</p>
+     *
+     * <p>This queries the cache using the key's descriptor, and if no
+     * object in the cache matches, tries the key on each
+     * registered factory, in order.  If none generates a service
+     * object for the key, repeats the process with each fallback of
+     * the key, until either a factory returns a service object, or the key
+     * has no fallback.  If no object is found, the result of handleDefault
+     * is returned.</p>
+     *
+     * <p>Subclasses can override this method to further customize the 
+     * result before returning it.
+     *
+     * @param key the key.
+     * @param actualReturn a pointer to a UnicodeString to hold the matched descriptor, or NULL.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    virtual UObject* getKey(ICUServiceKey& key, UnicodeString* actualReturn, UErrorCode& status) const;
+
+    /**
+     * <p>This version of getKey is only called by ICUServiceFactories within the scope
+     * of a previous getKey call, to determine what previously-registered factories would
+     * have returned.  For details, see getKey(ICUServiceKey&, UErrorCode&).  Subclasses
+     * should not call it directly, but call through one of the other get functions.</p>
+     * 
+     * @param key the key.
+     * @param actualReturn a pointer to a UnicodeString to hold the matched descriptor, or NULL.
+     * @param factory the factory making the recursive call.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    UObject* getKey(ICUServiceKey& key, UnicodeString* actualReturn, const ICUServiceFactory* factory, UErrorCode& status) const;
+
+    /**
+     * <p>Convenience override for getVisibleIDs(String) that passes null
+     * as the fallback, thus returning all visible IDs.</p>
+     *
+     * @param result a vector to hold the returned IDs.
+     * @param status the error code status.
+     * @return the result vector.
+     */
+    UVector& getVisibleIDs(UVector& result, UErrorCode& status) const;
+
+    /**
+     * <p>Return a snapshot of the visible IDs for this service.  This
+     * list will not change as ICUServiceFactories are added or removed, but the
+     * supported IDs will, so there is no guarantee that all and only
+     * the IDs in the returned list will be visible and supported by the
+     * service in subsequent calls.</p>
+     *
+     * <p>The IDs are returned as pointers to UnicodeStrings.  The
+     * caller owns the IDs.  Previous contents of result are discarded before
+     * new elements, if any, are added.</p>
+     *
+     * <p>matchID is passed to createKey to create a key.  If the key
+     * is not NULL, its isFallbackOf method is used to filter out IDs
+     * that don't match the key or have it as a fallback.</p>
+     *
+     * @param result a vector to hold the returned IDs.
+     * @param matchID an ID used to filter the result, or NULL if all IDs are desired.
+     * @param status the error code status.
+     * @return the result vector.
+     */
+    UVector& getVisibleIDs(UVector& result, const UnicodeString* matchID, UErrorCode& status) const;
+
+    /**
+     * <p>Convenience override for getDisplayName(const UnicodeString&, const Locale&, UnicodeString&) that
+     * uses the current default locale.</p>
+     *
+     * @param id the ID for which to retrieve the localized displayName.
+     * @param result an output parameter to hold the display name.
+     * @return the modified result.
+     */
+    UnicodeString& getDisplayName(const UnicodeString& id, UnicodeString& result) const;
+
+    /**
+     * <p>Given a visible ID, return the display name in the requested locale.
+     * If there is no directly supported ID corresponding to this ID, result is
+     * set to bogus.</p>
+     *
+     * @param id the ID for which to retrieve the localized displayName.
+     * @param result an output parameter to hold the display name.
+     * @param locale the locale in which to localize the ID.
+     * @return the modified result.
+     */
+    UnicodeString& getDisplayName(const UnicodeString& id, UnicodeString& result, const Locale& locale) const;
+
+    /**
+     * <p>Convenience override of getDisplayNames(const Locale&, const UnicodeString*) that 
+     * uses the current default Locale as the locale and NULL for
+     * the matchID.</p>
+     *
+     * @param result a vector to hold the returned displayName/id StringPairs.
+     * @param status the error code status.
+     * @return the modified result vector.
+     */
+    UVector& getDisplayNames(UVector& result, UErrorCode& status) const;
+
+    /**
+     * <p>Convenience override of getDisplayNames(const Locale&, const UnicodeString*) that 
+     * uses NULL for the matchID.</p>
+     *
+     * @param result a vector to hold the returned displayName/id StringPairs.
+     * @param locale the locale in which to localize the ID.
+     * @param status the error code status.
+     * @return the modified result vector.
+     */
+    UVector& getDisplayNames(UVector& result, const Locale& locale, UErrorCode& status) const;
+
+    /**
+     * <p>Return a snapshot of the mapping from display names to visible
+     * IDs for this service.  This set will not change as factories
+     * are added or removed, but the supported IDs will, so there is
+     * no guarantee that all and only the IDs in the returned map will
+     * be visible and supported by the service in subsequent calls,
+     * nor is there any guarantee that the current display names match
+     * those in the result.</p>
+     *
+     * <p>The names are returned as pointers to StringPairs, which
+     * contain both the displayName and the corresponding ID.  The
+     * caller owns the StringPairs.  Previous contents of result are
+     * discarded before new elements, if any, are added.</p>
+     *
+     * <p>matchID is passed to createKey to create a key.  If the key
+     * is not NULL, its isFallbackOf method is used to filter out IDs
+     * that don't match the key or have it as a fallback.</p>
+     *
+     * @param result a vector to hold the returned displayName/id StringPairs.
+     * @param locale the locale in which to localize the ID.
+     * @param matchID an ID used to filter the result, or NULL if all IDs are desired.
+     * @param status the error code status.
+     * @return the result vector.  */
+    UVector& getDisplayNames(UVector& result,
+                             const Locale& locale, 
+                             const UnicodeString* matchID, 
+                             UErrorCode& status) const;
+
+    /**
+     * <p>A convenience override of registerInstance(UObject*, const UnicodeString&, UBool)
+     * that defaults visible to true.</p>
+     *
+     * @param objToAdopt the object to register and adopt.
+     * @param id the ID to assign to this object.
+     * @param status the error code status.
+     * @return a registry key that can be passed to unregister to unregister
+     * (and discard) this instance.
+     */
+    URegistryKey registerInstance(UObject* objToAdopt, const UnicodeString& id, UErrorCode& status);
+
+    /**
+     * <p>Register a service instance with the provided ID.  The ID will be 
+     * canonicalized.  The canonicalized ID will be returned by
+     * getVisibleIDs if visible is true.  The service instance will be adopted and
+     * must not be modified subsequent to this call.</p>
+     *
+     * <p>This issues a serviceChanged notification to registered listeners.</p>
+     *
+     * <p>This implementation wraps the object using
+     * createSimpleFactory, and calls registerFactory.</p>
+     *
+     * @param objToAdopt the object to register and adopt.
+     * @param id the ID to assign to this object.
+     * @param visible true if getVisibleIDs is to return this ID.
+     * @param status the error code status.
+     * @return a registry key that can be passed to unregister() to unregister
+     * (and discard) this instance.
+     */
+    virtual URegistryKey registerInstance(UObject* objToAdopt, const UnicodeString& id, UBool visible, UErrorCode& status);
+
+    /**
+     * <p>Register an ICUServiceFactory.  Returns a registry key that
+     * can be used to unregister the factory.  The factory
+     * must not be modified subsequent to this call.  The service owns
+     * all registered factories. In case of an error, the factory is
+     * deleted.</p>
+     *
+     * <p>This issues a serviceChanged notification to registered listeners.</p>
+     *
+     * <p>The default implementation accepts all factories.</p>
+     *
+     * @param factoryToAdopt the factory to register and adopt.
+     * @param status the error code status.
+     * @return a registry key that can be passed to unregister to unregister
+     * (and discard) this factory.
+     */
+    virtual URegistryKey registerFactory(ICUServiceFactory* factoryToAdopt, UErrorCode& status);
+
+    /**
+     * <p>Unregister a factory using a registry key returned by
+     * registerInstance or registerFactory.  After a successful call,
+     * the factory will be removed from the service factory list and
+     * deleted, and the key becomes invalid.</p>
+     *
+     * <p>This issues a serviceChanged notification to registered
+     * listeners.</p>
+     *
+     * @param rkey the registry key.
+     * @param status the error code status.  
+     * @return true if the call successfully unregistered the factory.
+     */
+    virtual UBool unregister(URegistryKey rkey, UErrorCode& status);
+
+    /**
+     * </p>Reset the service to the default factories.  The factory
+     * lock is acquired and then reInitializeFactories is called.</p>
+     *
+     * <p>This issues a serviceChanged notification to registered listeners.</p>
+     */
+    virtual void reset(void);
+
+    /**
+     * <p>Return true if the service is in its default state.</p>
+     *
+     * <p>The default implementation returns true if there are no 
+     * factories registered.</p>
+     */
+    virtual UBool isDefault(void) const;
+
+    /**
+     * <p>Create a key from an ID.  If ID is NULL, returns NULL.</p>
+     *
+     * <p>The default implementation creates an ICUServiceKey instance.
+     * Subclasses can override to define more useful keys appropriate
+     * to the factories they accept.</p>
+     *
+     * @param a pointer to the ID for which to create a default ICUServiceKey.
+     * @param status the error code status.
+     * @return the ICUServiceKey corresponding to ID, or NULL.
+     */
+    virtual ICUServiceKey* createKey(const UnicodeString* id, UErrorCode& status) const;
+
+    /**
+     * <p>Clone object so that caller can own the copy.  In ICU2.4, UObject doesn't define
+     * clone, so we need an instance-aware method that knows how to do this.
+     * This is public so factories can call it, but should really be protected.</p>
+     *
+     * @param instance the service instance to clone.
+     * @return a clone of the passed-in instance, or NULL if cloning was unsuccessful.
+     */
+    virtual UObject* cloneInstance(UObject* instance) const = 0;
+
+
+    /************************************************************************
+     * Subclassing API
+     */
+
+ protected:
+
+    /**
+     * <p>Create a factory that wraps a single service object.  Called by registerInstance.</p>
+     *
+     * <p>The default implementation returns an instance of SimpleFactory.</p>
+     *
+     * @param instanceToAdopt the service instance to adopt.
+     * @param id the ID to assign to this service instance.
+     * @param visible if true, the ID will be visible.
+     * @param status the error code status.
+     * @return an instance of ICUServiceFactory that maps this instance to the provided ID.
+     */
+    virtual ICUServiceFactory* createSimpleFactory(UObject* instanceToAdopt, const UnicodeString& id, UBool visible, UErrorCode& status);
+
+    /**
+     * <p>Reinitialize the factory list to its default state.  After this call, isDefault()
+     * must return true.</p>
+     *
+     * <p>This issues a serviceChanged notification to registered listeners.</p>
+     *
+     * <p>The default implementation clears the factory list.
+     * Subclasses can override to provide other default initialization
+     * of the factory list.  Subclasses must not call this method
+     * directly, since it must only be called while holding write
+     * access to the factory list.</p>
+     */
+    virtual void reInitializeFactories(void);
+
+    /**
+     * <p>Default handler for this service if no factory in the factory list
+     * handled the key passed to getKey.</p>
+     *
+     * <p>The default implementation returns NULL.</p>
+     *
+     * @param key the key.
+     * @param actualReturn a pointer to a UnicodeString to hold the matched descriptor, or NULL.
+     * @param status the error code status.
+     * @return the service instance, or NULL.
+     */
+    virtual UObject* handleDefault(const ICUServiceKey& key, UnicodeString* actualReturn, UErrorCode& status) const;
+
+    /**
+     * <p>Clear caches maintained by this service.</p>
+     *
+     * <p>Subclasses can override if they implement additional caches
+     * that need to be cleared when the service changes.  Subclasses
+     * should generally not call this method directly, as it must only
+     * be called while synchronized on the factory lock.</p>
+     */
+    virtual void clearCaches(void);
+
+    /**
+     * <p>Return true if the listener is accepted.</p>
+     *
+     * <p>The default implementation accepts the listener if it is
+     * a ServiceListener.  Subclasses can override this to accept
+     * different listeners.</p>
+     *
+     * @param l the listener to test.
+     * @return true if the service accepts the listener.
+     */
+    virtual UBool acceptsListener(const EventListener& l) const;
+
+    /**
+     * <p>Notify the listener of a service change.</p>
+     *
+     * <p>The default implementation assumes a ServiceListener.
+     * If acceptsListener has been overridden to accept different
+     * listeners, this should be overridden as well.</p>
+     *
+     * @param l the listener to notify.
+     */
+    virtual void notifyListener(EventListener& l) const;
+
+    /************************************************************************
+     * Utilities for subclasses.
+     */
+
+    /**
+     * <p>Clear only the service cache.</p>
+     *
+     * <p>This can be called by subclasses when a change affects the service
+     * cache but not the ID caches, e.g., when the default locale changes
+     * the resolution of IDs also changes, requiring the cache to be
+     * flushed, but not the visible IDs themselves.</p>
+     */
+    void clearServiceCache(void);
+
+    /**
+     * <p>Return a map from visible IDs to factories.
+     * This must only be called when the mutex is held.</p>
+     *
+     * @param status the error code status.
+     * @return a Hashtable containing mappings from visible
+     * IDs to factories.
+     */
+    const Hashtable* getVisibleIDMap(UErrorCode& status) const;
+
+    /**
+     * <p>Allow subclasses to read the time stamp.</p>
+     *
+     * @return the timestamp.
+     */
+    int32_t getTimestamp(void) const;
+
+    /**
+     * <p>Return the number of registered factories.</p>
+     *
+     * @return the number of factories registered at the time of the call.
+     */
+    int32_t countFactories(void) const;
+
+private:
+
+    friend class ::ICUServiceTest; // give tests access to countFactories.
+};
+
+U_NAMESPACE_END
+
+    /* UCONFIG_NO_SERVICE */
+#endif
+
+    /* ICUSERV_H */
+#endif
+
diff --git a/thirdparty/icu4c/common/servlk.cpp b/thirdparty/icu4c/common/servlk.cpp
new file mode 100644
index 0000000000..538982ca36
--- /dev/null
+++ b/thirdparty/icu4c/common/servlk.cpp
@@ -0,0 +1,188 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "unicode/resbund.h"
+#include "uresimp.h"
+#include "cmemory.h"
+#include "servloc.h"
+#include "ustrfmt.h"
+#include "uhash.h"
+#include "charstr.h"
+#include "uassert.h"
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+U_NAMESPACE_BEGIN
+
+LocaleKey*
+LocaleKey::createWithCanonicalFallback(const UnicodeString* primaryID,
+                                       const UnicodeString* canonicalFallbackID,
+                                       UErrorCode& status)
+{
+    return LocaleKey::createWithCanonicalFallback(primaryID, canonicalFallbackID, KIND_ANY, status);
+}
+
+LocaleKey*
+LocaleKey::createWithCanonicalFallback(const UnicodeString* primaryID,
+                                       const UnicodeString* canonicalFallbackID,
+                                       int32_t kind,
+                                       UErrorCode& status)
+{
+    if (primaryID == NULL || U_FAILURE(status)) {
+        return NULL;
+    }
+    UnicodeString canonicalPrimaryID;
+    LocaleUtility::canonicalLocaleString(primaryID, canonicalPrimaryID);
+    return new LocaleKey(*primaryID, canonicalPrimaryID, canonicalFallbackID, kind);
+}
+
+LocaleKey::LocaleKey(const UnicodeString& primaryID,
+                     const UnicodeString& canonicalPrimaryID,
+                     const UnicodeString* canonicalFallbackID,
+                     int32_t kind)
+  : ICUServiceKey(primaryID)
+  , _kind(kind)
+  , _primaryID(canonicalPrimaryID)
+  , _fallbackID()
+  , _currentID()
+{
+    _fallbackID.setToBogus();
+    if (_primaryID.length() != 0) {
+        if (canonicalFallbackID != NULL && _primaryID != *canonicalFallbackID) {
+            _fallbackID = *canonicalFallbackID;
+        }
+    }
+
+    _currentID = _primaryID;
+}
+
+LocaleKey::~LocaleKey() {}
+
+UnicodeString&
+LocaleKey::prefix(UnicodeString& result) const {
+    if (_kind != KIND_ANY) {
+        UChar buffer[64];
+        uprv_itou(buffer, 64, _kind, 10, 0);
+        UnicodeString temp(buffer);
+        result.append(temp);
+    }
+    return result;
+}
+
+int32_t
+LocaleKey::kind() const {
+    return _kind;
+}
+
+UnicodeString&
+LocaleKey::canonicalID(UnicodeString& result) const {
+    return result.append(_primaryID);
+}
+
+UnicodeString&
+LocaleKey::currentID(UnicodeString& result) const {
+    if (!_currentID.isBogus()) {
+        result.append(_currentID);
+    }
+    return result;
+}
+
+UnicodeString&
+LocaleKey::currentDescriptor(UnicodeString& result) const {
+    if (!_currentID.isBogus()) {
+        prefix(result).append(PREFIX_DELIMITER).append(_currentID);
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+Locale&
+LocaleKey::canonicalLocale(Locale& result) const {
+    return LocaleUtility::initLocaleFromName(_primaryID, result);
+}
+
+Locale&
+LocaleKey::currentLocale(Locale& result) const {
+    return LocaleUtility::initLocaleFromName(_currentID, result);
+}
+
+UBool
+LocaleKey::fallback() {
+    if (!_currentID.isBogus()) {
+        int x = _currentID.lastIndexOf(UNDERSCORE_CHAR);
+        if (x != -1) {
+            _currentID.remove(x); // truncate current or fallback, whichever we're pointing to
+            return TRUE;
+        }
+
+        if (!_fallbackID.isBogus()) {
+            _currentID = _fallbackID;
+            _fallbackID.setToBogus();
+            return TRUE;
+        }
+
+        if (_currentID.length() > 0) {
+            _currentID.remove(0); // completely truncate
+            return TRUE;
+        }
+
+        _currentID.setToBogus();
+    }
+
+    return FALSE;
+}
+
+UBool
+LocaleKey::isFallbackOf(const UnicodeString& id) const {
+    UnicodeString temp(id);
+    parseSuffix(temp);
+    return temp.indexOf(_primaryID) == 0 &&
+        (temp.length() == _primaryID.length() ||
+        temp.charAt(_primaryID.length()) == UNDERSCORE_CHAR);
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString&
+LocaleKey::debug(UnicodeString& result) const
+{
+    ICUServiceKey::debug(result);
+    result.append((UnicodeString)" kind: ");
+    result.append(_kind);
+    result.append((UnicodeString)" primaryID: ");
+    result.append(_primaryID);
+    result.append((UnicodeString)" fallbackID: ");
+    result.append(_fallbackID);
+    result.append((UnicodeString)" currentID: ");
+    result.append(_currentID);
+    return result;
+}
+
+UnicodeString&
+LocaleKey::debugClass(UnicodeString& result) const
+{
+    return result.append((UnicodeString)"LocaleKey ");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(LocaleKey)
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/servlkf.cpp b/thirdparty/icu4c/common/servlkf.cpp
new file mode 100644
index 0000000000..84f2347cdd
--- /dev/null
+++ b/thirdparty/icu4c/common/servlkf.cpp
@@ -0,0 +1,152 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "unicode/resbund.h"
+#include "uresimp.h"
+#include "cmemory.h"
+#include "servloc.h"
+#include "ustrfmt.h"
+#include "uhash.h"
+#include "charstr.h"
+#include "ucln_cmn.h"
+#include "uassert.h"
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+
+U_NAMESPACE_BEGIN
+
+LocaleKeyFactory::LocaleKeyFactory(int32_t coverage)
+  : _name()
+  , _coverage(coverage)
+{
+}
+
+LocaleKeyFactory::LocaleKeyFactory(int32_t coverage, const UnicodeString& name)
+  : _name(name)
+  , _coverage(coverage)
+{
+}
+
+LocaleKeyFactory::~LocaleKeyFactory() {
+}
+
+UObject*
+LocaleKeyFactory::create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const {
+    if (handlesKey(key, status)) {
+        const LocaleKey& lkey = (const LocaleKey&)key;
+        int32_t kind = lkey.kind();
+        Locale loc;
+        lkey.currentLocale(loc);
+
+        return handleCreate(loc, kind, service, status);
+    }
+    return NULL;
+}
+
+UBool
+LocaleKeyFactory::handlesKey(const ICUServiceKey& key, UErrorCode& status) const {
+    const Hashtable* supported = getSupportedIDs(status);
+    if (supported) {
+        UnicodeString id;
+        key.currentID(id);
+        return supported->get(id) != NULL;
+    }
+    return FALSE;
+}
+
+void
+LocaleKeyFactory::updateVisibleIDs(Hashtable& result, UErrorCode& status) const {
+    const Hashtable* supported = getSupportedIDs(status);
+    if (supported) {
+        UBool visible = (_coverage & 0x1) == 0;
+        const UHashElement* elem = NULL;
+        int32_t pos = UHASH_FIRST;
+        while ((elem = supported->nextElement(pos)) != NULL) {
+            const UnicodeString& id = *((const UnicodeString*)elem->key.pointer);
+            if (!visible) {
+                result.remove(id);
+            } else {
+                result.put(id, (void*)this, status); // this is dummy non-void marker used for set semantics
+                if (U_FAILURE(status)) {
+                    break;
+                }
+            }
+        }
+    }
+}
+
+UnicodeString&
+LocaleKeyFactory::getDisplayName(const UnicodeString& id, const Locale& locale, UnicodeString& result) const {
+    if ((_coverage & 0x1) == 0) {
+        //UErrorCode status = U_ZERO_ERROR;
+        // assume if this is called on us, we support some fallback of this id
+        // if (isSupportedID(id, status)) {
+            Locale loc;
+            LocaleUtility::initLocaleFromName(id, loc);
+            return loc.getDisplayName(locale, result);
+        // }
+    }
+    result.setToBogus();
+    return result;
+}
+
+UObject*
+LocaleKeyFactory::handleCreate(const Locale& /* loc */, 
+                   int32_t /* kind */, 
+                   const ICUService* /* service */, 
+                   UErrorCode& /* status */) const {
+    return NULL;
+}
+
+//UBool
+//LocaleKeyFactory::isSupportedID(const UnicodeString& id, UErrorCode& status) const {
+//    const Hashtable* ids = getSupportedIDs(status);
+//    return ids && ids->get(id);
+//}
+
+const Hashtable*
+LocaleKeyFactory::getSupportedIDs(UErrorCode& /* status */) const {
+    return NULL;
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString&
+LocaleKeyFactory::debug(UnicodeString& result) const
+{
+    debugClass(result);
+    result.append((UnicodeString)", name: ");
+    result.append(_name);
+    result.append((UnicodeString)", coverage: ");
+    result.append(_coverage);
+    return result;
+}
+
+UnicodeString&
+LocaleKeyFactory::debugClass(UnicodeString& result) const
+{
+  return result.append((UnicodeString)"LocaleKeyFactory");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(LocaleKeyFactory)
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/servloc.h b/thirdparty/icu4c/common/servloc.h
new file mode 100644
index 0000000000..ccf6433379
--- /dev/null
+++ b/thirdparty/icu4c/common/servloc.h
@@ -0,0 +1,551 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2011, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#ifndef ICULSERV_H
+#define ICULSERV_H
+
+#include "unicode/utypes.h"
+
+#if UCONFIG_NO_SERVICE
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Allow the declaration of APIs with pointers to ICUService
+ * even when service is removed from the build.
+ */
+class ICULocaleService;
+
+U_NAMESPACE_END
+
+#else
+
+#include "unicode/unistr.h"
+#include "unicode/locid.h"
+#include "unicode/strenum.h"
+
+#include "hash.h"
+#include "uvector.h"
+
+#include "serv.h"
+#include "locutil.h"
+
+U_NAMESPACE_BEGIN
+
+class ICULocaleService;
+
+class LocaleKey;
+class LocaleKeyFactory;
+class SimpleLocaleKeyFactory;
+class ServiceListener;
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * A subclass of Key that implements a locale fallback mechanism.
+ * The first locale to search for is the locale provided by the
+ * client, and the fallback locale to search for is the current
+ * default locale.  If a prefix is present, the currentDescriptor
+ * includes it before the locale proper, separated by "/".  This
+ * is the default key instantiated by ICULocaleService.</p>
+ *
+ * <p>Canonicalization adjusts the locale string so that the
+ * section before the first understore is in lower case, and the rest
+ * is in upper case, with no trailing underscores.</p> 
+ */
+
+class U_COMMON_API LocaleKey : public ICUServiceKey {
+  private: 
+    int32_t _kind;
+    UnicodeString _primaryID;
+    UnicodeString _fallbackID;
+    UnicodeString _currentID;
+
+  public:
+    enum {
+        KIND_ANY = -1
+    };
+
+    /**
+     * Create a LocaleKey with canonical primary and fallback IDs.
+     */
+    static LocaleKey* createWithCanonicalFallback(const UnicodeString* primaryID, 
+                                                  const UnicodeString* canonicalFallbackID,
+                                                  UErrorCode& status);
+
+    /**
+     * Create a LocaleKey with canonical primary and fallback IDs.
+     */
+    static LocaleKey* createWithCanonicalFallback(const UnicodeString* primaryID, 
+                                                  const UnicodeString* canonicalFallbackID, 
+                                                  int32_t kind,
+                                                  UErrorCode& status);
+
+  protected:
+    /**
+     * PrimaryID is the user's requested locale string,
+     * canonicalPrimaryID is this string in canonical form,
+     * fallbackID is the current default locale's string in
+     * canonical form.
+     */
+    LocaleKey(const UnicodeString& primaryID, 
+              const UnicodeString& canonicalPrimaryID, 
+              const UnicodeString* canonicalFallbackID, 
+              int32_t kind);
+
+ public:
+    /**
+     * Append the prefix associated with the kind, or nothing if the kind is KIND_ANY.
+     */
+    virtual UnicodeString& prefix(UnicodeString& result) const;
+
+    /**
+     * Return the kind code associated with this key.
+     */
+    virtual int32_t kind() const;
+
+    /**
+     * Return the canonicalID.
+     */
+    virtual UnicodeString& canonicalID(UnicodeString& result) const;
+
+    /**
+     * Return the currentID.
+     */
+    virtual UnicodeString& currentID(UnicodeString& result) const;
+
+    /**
+     * Return the (canonical) current descriptor, or null if no current id.
+     */
+    virtual UnicodeString& currentDescriptor(UnicodeString& result) const;
+
+    /**
+     * Convenience method to return the locale corresponding to the (canonical) original ID.
+     */
+    virtual Locale& canonicalLocale(Locale& result) const;
+
+    /**
+     * Convenience method to return the locale corresponding to the (canonical) current ID.
+     */
+    virtual Locale& currentLocale(Locale& result) const;
+
+    /**
+     * If the key has a fallback, modify the key and return true,
+     * otherwise return false.</p>
+     *
+     * <p>First falls back through the primary ID, then through
+     * the fallbackID.  The final fallback is the empty string,
+     * unless the primary id was the empty string, in which case
+     * there is no fallback.  
+     */
+    virtual UBool fallback();
+
+    /**
+     * Return true if a key created from id matches, or would eventually
+     * fallback to match, the canonical ID of this key.  
+     */
+    virtual UBool isFallbackOf(const UnicodeString& id) const;
+    
+ public:
+    /**
+     * UObject boilerplate.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    virtual UClassID getDynamicClassID() const;
+
+    /**
+     * Destructor.
+     */
+    virtual ~LocaleKey();
+
+#ifdef SERVICE_DEBUG
+ public:
+    virtual UnicodeString& debug(UnicodeString& result) const;
+    virtual UnicodeString& debugClass(UnicodeString& result) const;
+#endif
+
+};
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * A subclass of ICUServiceFactory that uses LocaleKeys, and is able to
+ * 'cover' more specific locales with more general locales that it
+ * supports.  
+ *
+ * <p>Coverage may be either of the values VISIBLE or INVISIBLE.
+ *
+ * <p>'Visible' indicates that the specific locale(s) supported by
+ * the factory are registered in getSupportedIDs, 'Invisible'
+ * indicates that they are not.
+ *
+ * <p>Localization of visible ids is handled
+ * by the handling factory, regardless of kind.
+ */
+class U_COMMON_API LocaleKeyFactory : public ICUServiceFactory {
+protected:
+    const UnicodeString _name;
+    const int32_t _coverage;
+
+public:
+    enum {
+        /**
+         * Coverage value indicating that the factory makes
+         * its locales visible, and does not cover more specific 
+         * locales.
+         */
+        VISIBLE = 0,
+
+        /**
+         * Coverage value indicating that the factory does not make
+         * its locales visible, and does not cover more specific
+         * locales.
+         */
+        INVISIBLE = 1
+    };
+
+    /**
+     * Destructor.
+     */
+    virtual ~LocaleKeyFactory();
+
+protected:
+    /**
+     * Constructor used by subclasses.
+     */
+    LocaleKeyFactory(int32_t coverage);
+
+    /**
+     * Constructor used by subclasses.
+     */
+    LocaleKeyFactory(int32_t coverage, const UnicodeString& name);
+
+    /**
+     * Implement superclass abstract method.  This checks the currentID of
+     * the key against the supported IDs, and passes the canonicalLocale and
+     * kind off to handleCreate (which subclasses must implement).
+     */
+public:
+    virtual UObject* create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const;
+
+protected:
+    virtual UBool handlesKey(const ICUServiceKey& key, UErrorCode& status) const;
+
+public:
+    /**
+     * Override of superclass method.  This adjusts the result based
+     * on the coverage rule for this factory.
+     */
+    virtual void updateVisibleIDs(Hashtable& result, UErrorCode& status) const;
+
+    /**
+     * Return a localized name for the locale represented by id.
+     */
+    virtual UnicodeString& getDisplayName(const UnicodeString& id, const Locale& locale, UnicodeString& result) const;
+
+protected:
+    /**
+     * Utility method used by create(ICUServiceKey, ICUService).  Subclasses can implement
+     * this instead of create.  The default returns NULL.
+     */
+    virtual UObject* handleCreate(const Locale& loc, int32_t kind, const ICUService* service, UErrorCode& status) const;
+
+   /**
+     * Return true if this id is one the factory supports (visible or 
+     * otherwise).
+     */
+ //   virtual UBool isSupportedID(const UnicodeString& id, UErrorCode& status) const;
+
+   /**
+     * Return the set of ids that this factory supports (visible or 
+     * otherwise).  This can be called often and might need to be
+     * cached if it is expensive to create.
+     */
+    virtual const Hashtable* getSupportedIDs(UErrorCode& status) const;
+
+public:
+    /**
+     * UObject boilerplate.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    virtual UClassID getDynamicClassID() const;
+
+#ifdef SERVICE_DEBUG
+ public:
+    virtual UnicodeString& debug(UnicodeString& result) const;
+    virtual UnicodeString& debugClass(UnicodeString& result) const;
+#endif
+
+};
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * A LocaleKeyFactory that just returns a single object for a kind/locale.
+ */
+
+class U_COMMON_API SimpleLocaleKeyFactory : public LocaleKeyFactory {
+ private:
+    UObject* _obj;
+    UnicodeString _id;
+    const int32_t _kind;
+
+ public:
+    SimpleLocaleKeyFactory(UObject* objToAdopt, 
+                           const UnicodeString& locale, 
+                           int32_t kind, 
+                           int32_t coverage);
+
+    SimpleLocaleKeyFactory(UObject* objToAdopt, 
+                           const Locale& locale, 
+                           int32_t kind, 
+                           int32_t coverage);
+
+    /**
+     * Destructor.
+     */
+    virtual ~SimpleLocaleKeyFactory();
+
+    /**
+     * Override of superclass method.  Returns the service object if kind/locale match.  Service is not used.
+     */
+    virtual UObject* create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const;
+
+    /**
+     * Override of superclass method.  This adjusts the result based
+     * on the coverage rule for this factory.
+     */
+    virtual void updateVisibleIDs(Hashtable& result, UErrorCode& status) const;
+
+ protected:
+    /**
+     * Return true if this id is equal to the locale name.
+     */
+    //virtual UBool isSupportedID(const UnicodeString& id, UErrorCode& status) const;
+
+
+public:
+    /**
+     * UObject boilerplate.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    virtual UClassID getDynamicClassID() const;
+
+#ifdef SERVICE_DEBUG
+ public:
+    virtual UnicodeString& debug(UnicodeString& result) const;
+    virtual UnicodeString& debugClass(UnicodeString& result) const;
+#endif
+
+};
+
+/*
+ ******************************************************************
+ */
+
+/**
+ * A LocaleKeyFactory that creates a service based on the ICU locale data.
+ * This is a base class for most ICU factories.  Subclasses instantiate it
+ * with a constructor that takes a bundle name, which determines the supported
+ * IDs.  Subclasses then override handleCreate to create the actual service
+ * object.  The default implementation returns a resource bundle.
+ */
+class U_COMMON_API ICUResourceBundleFactory : public LocaleKeyFactory 
+{
+ protected:
+    UnicodeString _bundleName;
+
+ public:
+    /**
+     * Convenience constructor that uses the main ICU bundle name.
+     */
+    ICUResourceBundleFactory();
+
+    /**
+     * A service factory based on ICU resource data in resources with
+     * the given name.  This should be a 'path' that can be passed to
+     * ures_openAvailableLocales, such as U_ICUDATA or U_ICUDATA_COLL.
+     * The empty string is equivalent to U_ICUDATA.
+     */
+    ICUResourceBundleFactory(const UnicodeString& bundleName);
+
+    /**
+     * Destructor
+     */
+    virtual ~ICUResourceBundleFactory();
+
+protected:
+    /**
+     * Return the supported IDs.  This is the set of all locale names in ICULocaleData.
+     */
+    virtual const Hashtable* getSupportedIDs(UErrorCode& status) const;
+
+    /**
+     * Create the service.  The default implementation returns the resource bundle
+     * for the locale, ignoring kind, and service.
+     */
+    virtual UObject* handleCreate(const Locale& loc, int32_t kind, const ICUService* service, UErrorCode& status) const;
+
+public:
+    /**
+     * UObject boilerplate.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+    virtual UClassID getDynamicClassID() const;
+
+
+#ifdef SERVICE_DEBUG
+ public:
+    virtual UnicodeString& debug(UnicodeString& result) const;
+    virtual UnicodeString& debugClass(UnicodeString& result) const;
+#endif
+
+};
+
+/*
+ ******************************************************************
+ */
+
+class U_COMMON_API ICULocaleService : public ICUService 
+{
+ private:
+  Locale fallbackLocale;
+  UnicodeString fallbackLocaleName;
+
+ public:
+  /**
+   * Construct an ICULocaleService.
+   */
+  ICULocaleService();
+
+  /**
+   * Construct an ICULocaleService with a name (useful for debugging).
+   */
+  ICULocaleService(const UnicodeString& name);
+
+  /**
+   * Destructor.
+   */
+  virtual ~ICULocaleService();
+
+#if 0
+  // redeclare because of overload resolution rules?
+  // no, causes ambiguities since both UnicodeString and Locale have constructors that take a const char*
+  // need some compiler flag to remove warnings 
+  UObject* get(const UnicodeString& descriptor, UErrorCode& status) const {
+    return ICUService::get(descriptor, status);
+  }
+
+  UObject* get(const UnicodeString& descriptor, UnicodeString* actualReturn, UErrorCode& status) const {
+    return ICUService::get(descriptor, actualReturn, status);
+  }
+#endif
+
+  /**
+   * Convenience override for callers using locales.  This calls
+   * get(Locale, int, Locale[]) with KIND_ANY for kind and null for
+   * actualReturn.
+   */
+  UObject* get(const Locale& locale, UErrorCode& status) const;
+
+  /**
+   * Convenience override for callers using locales.  This calls
+   * get(Locale, int, Locale[]) with a null actualReturn.
+   */
+  UObject* get(const Locale& locale, int32_t kind, UErrorCode& status) const;
+
+  /**
+   * Convenience override for callers using locales. This calls
+   * get(Locale, String, Locale[]) with a null kind.
+   */
+  UObject* get(const Locale& locale, Locale* actualReturn, UErrorCode& status) const;
+                   
+  /**
+   * Convenience override for callers using locales.  This uses
+   * createKey(Locale.toString(), kind) to create a key, calls getKey, and then
+   * if actualReturn is not null, returns the actualResult from
+   * getKey (stripping any prefix) into a Locale.  
+   */
+  UObject* get(const Locale& locale, int32_t kind, Locale* actualReturn, UErrorCode& status) const;
+
+  /**
+   * Convenience override for callers using locales.  This calls
+   * registerObject(Object, Locale, int32_t kind, int coverage)
+   * passing KIND_ANY for the kind, and VISIBLE for the coverage.
+   */
+  virtual URegistryKey registerInstance(UObject* objToAdopt, const Locale& locale, UErrorCode& status);
+
+  /**
+   * Convenience function for callers using locales.  This calls
+   * registerObject(Object, Locale, int kind, int coverage)
+   * passing VISIBLE for the coverage.
+   */
+  virtual URegistryKey registerInstance(UObject* objToAdopt, const Locale& locale, int32_t kind, UErrorCode& status);
+
+  /**
+   * Convenience function for callers using locales.  This  instantiates
+   * a SimpleLocaleKeyFactory, and registers the factory.
+   */
+  virtual URegistryKey registerInstance(UObject* objToAdopt, const Locale& locale, int32_t kind, int32_t coverage, UErrorCode& status);
+
+
+  /**
+   * (Stop compiler from complaining about hidden overrides.)
+   * Since both UnicodeString and Locale have constructors that take const char*, adding a public
+   * method that takes UnicodeString causes ambiguity at call sites that use const char*.
+   * We really need a flag that is understood by all compilers that will suppress the warning about
+   * hidden overrides.
+   */
+  virtual URegistryKey registerInstance(UObject* objToAdopt, const UnicodeString& locale, UBool visible, UErrorCode& status);
+
+  /**
+   * Convenience method for callers using locales.  This returns the standard
+   * service ID enumeration.
+   */
+  virtual StringEnumeration* getAvailableLocales(void) const;
+
+ protected:
+
+  /**
+   * Return the name of the current fallback locale.  If it has changed since this was
+   * last accessed, the service cache is cleared.
+   */
+  const UnicodeString& validateFallbackLocale() const;
+
+  /**
+   * Override superclass createKey method.
+   */
+  virtual ICUServiceKey* createKey(const UnicodeString* id, UErrorCode& status) const;
+
+  /**
+   * Additional createKey that takes a kind.
+   */
+  virtual ICUServiceKey* createKey(const UnicodeString* id, int32_t kind, UErrorCode& status) const;
+
+  friend class ServiceEnumeration;
+};
+
+U_NAMESPACE_END
+
+    /* UCONFIG_NO_SERVICE */
+#endif
+
+    /* ICULSERV_H */
+#endif
+
diff --git a/thirdparty/icu4c/common/servls.cpp b/thirdparty/icu4c/common/servls.cpp
new file mode 100644
index 0000000000..81dc4f750e
--- /dev/null
+++ b/thirdparty/icu4c/common/servls.cpp
@@ -0,0 +1,295 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "unicode/resbund.h"
+#include "uresimp.h"
+#include "cmemory.h"
+#include "servloc.h"
+#include "ustrfmt.h"
+#include "charstr.h"
+#include "uassert.h"
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+U_NAMESPACE_BEGIN
+
+ICULocaleService::ICULocaleService()
+  : fallbackLocale(Locale::getDefault())
+{
+}
+
+ICULocaleService::ICULocaleService(const UnicodeString& dname)
+  : ICUService(dname)
+  , fallbackLocale(Locale::getDefault())
+{
+}
+
+ICULocaleService::~ICULocaleService()
+{
+}
+
+UObject*
+ICULocaleService::get(const Locale& locale, UErrorCode& status) const
+{
+    return get(locale, LocaleKey::KIND_ANY, NULL, status);
+}
+
+UObject*
+ICULocaleService::get(const Locale& locale, int32_t kind, UErrorCode& status) const
+{
+    return get(locale, kind, NULL, status);
+}
+
+UObject*
+ICULocaleService::get(const Locale& locale, Locale* actualReturn, UErrorCode& status) const
+{
+    return get(locale, LocaleKey::KIND_ANY, actualReturn, status);
+}
+
+UObject*
+ICULocaleService::get(const Locale& locale, int32_t kind, Locale* actualReturn, UErrorCode& status) const
+{
+    UObject* result = NULL;
+    if (U_FAILURE(status)) {
+        return result;
+    }
+
+    UnicodeString locName(locale.getName(), -1, US_INV);
+    if (locName.isBogus()) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        ICUServiceKey* key = createKey(&locName, kind, status);
+        if (key) {
+            if (actualReturn == NULL) {
+                result = getKey(*key, status);
+            } else {
+                UnicodeString temp;
+                result = getKey(*key, &temp, status);
+
+                if (result != NULL) {
+                    key->parseSuffix(temp);
+                    LocaleUtility::initLocaleFromName(temp, *actualReturn);
+                }
+            }
+            delete key;
+        }
+    }
+    return result;
+}
+
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const UnicodeString& locale, 
+    UBool visible, UErrorCode& status)
+{
+    Locale loc;
+    LocaleUtility::initLocaleFromName(locale, loc);
+    return registerInstance(objToAdopt, loc, LocaleKey::KIND_ANY, 
+        visible ? LocaleKeyFactory::VISIBLE : LocaleKeyFactory::INVISIBLE, status);
+}
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const Locale& locale, UErrorCode& status)
+{
+    return registerInstance(objToAdopt, locale, LocaleKey::KIND_ANY, LocaleKeyFactory::VISIBLE, status);
+}
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const Locale& locale, int32_t kind, UErrorCode& status)
+{
+    return registerInstance(objToAdopt, locale, kind, LocaleKeyFactory::VISIBLE, status);
+}
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const Locale& locale, int32_t kind, int32_t coverage, UErrorCode& status)
+{
+    ICUServiceFactory * factory = new SimpleLocaleKeyFactory(objToAdopt, locale, kind, coverage);
+    if (factory != NULL) {
+        return registerFactory(factory, status);
+    }
+    delete objToAdopt;
+    return NULL;
+}
+
+#if 0
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const UnicodeString& locale, UErrorCode& status)
+{
+    return registerInstance(objToAdopt, locale, LocaleKey::KIND_ANY, LocaleKeyFactory::VISIBLE, status);
+}
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const UnicodeString& locale, UBool visible, UErrorCode& status)
+{
+    return registerInstance(objToAdopt, locale, LocaleKey::KIND_ANY,
+                            visible ? LocaleKeyFactory::VISIBLE : LocaleKeyFactory::INVISIBLE,
+                            status);
+}
+
+URegistryKey
+ICULocaleService::registerInstance(UObject* objToAdopt, const UnicodeString& locale, int32_t kind, int32_t coverage, UErrorCode& status)
+{
+    ICUServiceFactory * factory = new SimpleLocaleKeyFactory(objToAdopt, locale, kind, coverage);
+    if (factory != NULL) {
+        return registerFactory(factory, status);
+    }
+    delete objToAdopt;
+    return NULL;
+}
+#endif
+
+class ServiceEnumeration : public StringEnumeration {
+private:
+    const ICULocaleService* _service;
+    int32_t _timestamp;
+    UVector _ids;
+    int32_t _pos;
+
+private:
+    ServiceEnumeration(const ICULocaleService* service, UErrorCode &status)
+        : _service(service)
+        , _timestamp(service->getTimestamp())
+        , _ids(uprv_deleteUObject, NULL, status)
+        , _pos(0)
+    {
+        _service->getVisibleIDs(_ids, status);
+    }
+
+    ServiceEnumeration(const ServiceEnumeration &other, UErrorCode &status)
+        : _service(other._service)
+        , _timestamp(other._timestamp)
+        , _ids(uprv_deleteUObject, NULL, status)
+        , _pos(0)
+    {
+        if(U_SUCCESS(status)) {
+            int32_t i, length;
+
+            length = other._ids.size();
+            for(i = 0; i < length; ++i) {
+                _ids.addElement(((UnicodeString *)other._ids.elementAt(i))->clone(), status);
+            }
+
+            if(U_SUCCESS(status)) {
+                _pos = other._pos;
+            }
+        }
+    }
+
+public:
+    static ServiceEnumeration* create(const ICULocaleService* service) {
+        UErrorCode status = U_ZERO_ERROR;
+        ServiceEnumeration* result = new ServiceEnumeration(service, status);
+        if (U_SUCCESS(status)) {
+            return result;
+        }
+        delete result;
+        return NULL;
+    }
+
+    virtual ~ServiceEnumeration();
+
+    virtual StringEnumeration *clone() const {
+        UErrorCode status = U_ZERO_ERROR;
+        ServiceEnumeration *cl = new ServiceEnumeration(*this, status);
+        if(U_FAILURE(status)) {
+            delete cl;
+            cl = NULL;
+        }
+        return cl;
+    }
+
+    UBool upToDate(UErrorCode& status) const {
+        if (U_SUCCESS(status)) {
+            if (_timestamp == _service->getTimestamp()) {
+                return TRUE;
+            }
+            status = U_ENUM_OUT_OF_SYNC_ERROR;
+        }
+        return FALSE;
+    }
+
+    virtual int32_t count(UErrorCode& status) const {
+        return upToDate(status) ? _ids.size() : 0;
+    }
+
+    virtual const UnicodeString* snext(UErrorCode& status) {
+        if (upToDate(status) && (_pos < _ids.size())) {
+            return (const UnicodeString*)_ids[_pos++];
+        }
+        return NULL;
+    }
+
+    virtual void reset(UErrorCode& status) {
+        if (status == U_ENUM_OUT_OF_SYNC_ERROR) {
+            status = U_ZERO_ERROR;
+        }
+        if (U_SUCCESS(status)) {
+            _timestamp = _service->getTimestamp();
+            _pos = 0;
+            _service->getVisibleIDs(_ids, status);
+        }
+    }
+
+public:
+    static UClassID U_EXPORT2 getStaticClassID(void);
+    virtual UClassID getDynamicClassID(void) const;
+};
+
+ServiceEnumeration::~ServiceEnumeration() {}
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ServiceEnumeration)
+
+StringEnumeration*
+ICULocaleService::getAvailableLocales(void) const
+{
+    return ServiceEnumeration::create(this);
+}
+
+const UnicodeString&
+ICULocaleService::validateFallbackLocale() const
+{
+    const Locale&     loc    = Locale::getDefault();
+    ICULocaleService* ncThis = (ICULocaleService*)this;
+    static UMutex llock;
+    {
+        Mutex mutex(&llock);
+        if (loc != fallbackLocale) {
+            ncThis->fallbackLocale = loc;
+            LocaleUtility::initNameFromLocale(loc, ncThis->fallbackLocaleName);
+            ncThis->clearServiceCache();
+        }
+    }
+    return fallbackLocaleName;
+}
+
+ICUServiceKey*
+ICULocaleService::createKey(const UnicodeString* id, UErrorCode& status) const
+{
+    return LocaleKey::createWithCanonicalFallback(id, &validateFallbackLocale(), status);
+}
+
+ICUServiceKey*
+ICULocaleService::createKey(const UnicodeString* id, int32_t kind, UErrorCode& status) const
+{
+    return LocaleKey::createWithCanonicalFallback(id, &validateFallbackLocale(), kind, status);
+}
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/servnotf.cpp b/thirdparty/icu4c/common/servnotf.cpp
new file mode 100644
index 0000000000..f577795cae
--- /dev/null
+++ b/thirdparty/icu4c/common/servnotf.cpp
@@ -0,0 +1,120 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2012, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "servnotf.h"
+#ifdef NOTIFIER_DEBUG
+#include <stdio.h>
+#endif
+
+U_NAMESPACE_BEGIN
+
+EventListener::~EventListener() {}
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(EventListener)
+
+static UMutex notifyLock;
+
+ICUNotifier::ICUNotifier(void) 
+: listeners(NULL) 
+{
+}
+
+ICUNotifier::~ICUNotifier(void) {
+    {
+        Mutex lmx(&notifyLock);
+        delete listeners;
+        listeners = NULL;
+    }
+}
+
+
+void 
+ICUNotifier::addListener(const EventListener* l, UErrorCode& status) 
+{
+    if (U_SUCCESS(status)) {
+        if (l == NULL) {
+            status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+
+        if (acceptsListener(*l)) {
+            Mutex lmx(&notifyLock);
+            if (listeners == NULL) {
+                listeners = new UVector(5, status);
+            } else {
+                for (int i = 0, e = listeners->size(); i < e; ++i) {
+                    const EventListener* el = (const EventListener*)(listeners->elementAt(i));
+                    if (l == el) {
+                        return;
+                    }
+                }
+            }
+
+            listeners->addElement((void*)l, status); // cast away const
+        }
+#ifdef NOTIFIER_DEBUG
+        else {
+            fprintf(stderr, "Listener invalid for this notifier.");
+            exit(1);
+        }
+#endif
+    }
+}
+
+void 
+ICUNotifier::removeListener(const EventListener *l, UErrorCode& status) 
+{
+    if (U_SUCCESS(status)) {
+        if (l == NULL) {
+            status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+
+        {
+            Mutex lmx(&notifyLock);
+            if (listeners != NULL) {
+                // identity equality check
+                for (int i = 0, e = listeners->size(); i < e; ++i) {
+                    const EventListener* el = (const EventListener*)listeners->elementAt(i);
+                    if (l == el) {
+                        listeners->removeElementAt(i);
+                        if (listeners->size() == 0) {
+                            delete listeners;
+                            listeners = NULL;
+                        }
+                        return;
+                    }
+                }
+            }
+        }
+    }
+}
+
+void 
+ICUNotifier::notifyChanged(void) 
+{
+    if (listeners != NULL) {
+        Mutex lmx(&notifyLock);
+        if (listeners != NULL) {
+            for (int i = 0, e = listeners->size(); i < e; ++i) {
+                EventListener* el = (EventListener*)listeners->elementAt(i);
+                notifyListener(*el);
+            }
+        }
+    }
+}
+
+U_NAMESPACE_END
+
+/* UCONFIG_NO_SERVICE */
+#endif
+
diff --git a/thirdparty/icu4c/common/servnotf.h b/thirdparty/icu4c/common/servnotf.h
new file mode 100644
index 0000000000..305570c1e6
--- /dev/null
+++ b/thirdparty/icu4c/common/servnotf.h
@@ -0,0 +1,125 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ */
+#ifndef ICUNOTIF_H
+#define ICUNOTIF_H
+
+#include "unicode/utypes.h"
+
+#if UCONFIG_NO_SERVICE
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Allow the declaration of APIs with pointers to BreakIterator
+ * even when break iteration is removed from the build.
+ */
+class ICUNotifier;
+
+U_NAMESPACE_END
+
+#else
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+
+#include "mutex.h"
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+class U_COMMON_API EventListener : public UObject {
+public: 
+    virtual ~EventListener();
+
+public:
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    virtual UClassID getDynamicClassID() const;
+
+public:
+#ifdef SERVICE_DEBUG
+    virtual UnicodeString& debug(UnicodeString& result) const {
+      return debugClass(result);
+    }
+
+    virtual UnicodeString& debugClass(UnicodeString& result) const {
+      return result.append((UnicodeString)"Key");
+    }
+#endif
+};
+
+/**
+ * <p>Abstract implementation of a notification facility.  Clients add
+ * EventListeners with addListener and remove them with removeListener.
+ * Notifiers call notifyChanged when they wish to notify listeners.
+ * This queues the listener list on the notification thread, which
+ * eventually dequeues the list and calls notifyListener on each
+ * listener in the list.</p>
+ *
+ * <p>Subclasses override acceptsListener and notifyListener 
+ * to add type-safe notification.  AcceptsListener should return
+ * true if the listener is of the appropriate type; ICUNotifier
+ * itself will ensure the listener is non-null and that the
+ * identical listener is not already registered with the Notifier.
+ * NotifyListener should cast the listener to the appropriate 
+ * type and call the appropriate method on the listener.
+ */
+
+class U_COMMON_API ICUNotifier : public UMemory  {
+private: UVector* listeners;
+         
+public: 
+    ICUNotifier(void);
+    
+    virtual ~ICUNotifier(void);
+    
+    /**
+     * Add a listener to be notified when notifyChanged is called.
+     * The listener must not be null. AcceptsListener must return
+     * true for the listener.  Attempts to concurrently
+     * register the identical listener more than once will be
+     * silently ignored.  
+     */
+    virtual void addListener(const EventListener* l, UErrorCode& status);
+    
+    /**
+     * Stop notifying this listener.  The listener must
+     * not be null.  Attemps to remove a listener that is
+     * not registered will be silently ignored.
+     */
+    virtual void removeListener(const EventListener* l, UErrorCode& status);
+    
+    /**
+     * ICU doesn't spawn its own threads.  All listeners are notified in
+     * the thread of the caller.  Misbehaved listeners can therefore
+     * indefinitely block the calling thread.  Callers should beware of
+     * deadlock situations.  
+     */
+    virtual void notifyChanged(void);
+    
+protected: 
+    /**
+     * Subclasses implement this to return true if the listener is
+     * of the appropriate type.
+     */
+    virtual UBool acceptsListener(const EventListener& l) const = 0;
+    
+    /**
+     * Subclasses implement this to notify the listener.
+     */
+    virtual void notifyListener(EventListener& l) const = 0;
+};
+
+U_NAMESPACE_END
+
+/* UCONFIG_NO_SERVICE */
+#endif
+
+/* ICUNOTIF_H */
+#endif
diff --git a/thirdparty/icu4c/common/servrbf.cpp b/thirdparty/icu4c/common/servrbf.cpp
new file mode 100644
index 0000000000..94279ab3a1
--- /dev/null
+++ b/thirdparty/icu4c/common/servrbf.cpp
@@ -0,0 +1,96 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "unicode/resbund.h"
+#include "uresimp.h"
+#include "cmemory.h"
+#include "servloc.h"
+#include "ustrfmt.h"
+#include "uhash.h"
+#include "charstr.h"
+#include "ucln_cmn.h"
+#include "uassert.h"
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+U_NAMESPACE_BEGIN
+
+ICUResourceBundleFactory::ICUResourceBundleFactory()
+  : LocaleKeyFactory(VISIBLE)
+  , _bundleName()
+{
+}
+
+ICUResourceBundleFactory::ICUResourceBundleFactory(const UnicodeString& bundleName)
+  : LocaleKeyFactory(VISIBLE)
+  , _bundleName(bundleName)
+{
+}
+
+ICUResourceBundleFactory::~ICUResourceBundleFactory() {}
+
+const Hashtable*
+ICUResourceBundleFactory::getSupportedIDs(UErrorCode& status) const
+{
+    if (U_SUCCESS(status)) {
+        return LocaleUtility::getAvailableLocaleNames(_bundleName);
+    }
+    return NULL;
+}
+
+UObject*
+ICUResourceBundleFactory::handleCreate(const Locale& loc, int32_t /* kind */, const ICUService* /* service */, UErrorCode& status) const
+{
+    if (U_SUCCESS(status)) {
+        // _bundleName is a package name
+        // and should only contain invariant characters
+                // ??? is it always true that the max length of the bundle name is 19?
+                // who made this change? -- dlf
+        char pkg[20];
+        int32_t length;
+        length=_bundleName.extract(0, INT32_MAX, pkg, (int32_t)sizeof(pkg), US_INV);
+        if(length>=(int32_t)sizeof(pkg)) {
+            return NULL;
+        }
+        return new ResourceBundle(pkg, loc, status);
+    }
+    return NULL;
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString&
+ICUResourceBundleFactory::debug(UnicodeString& result) const
+{
+    LocaleKeyFactory::debug(result);
+    result.append((UnicodeString)", bundle: ");
+    return result.append(_bundleName);
+}
+
+UnicodeString&
+ICUResourceBundleFactory::debugClass(UnicodeString& result) const
+{
+    return result.append((UnicodeString)"ICUResourceBundleFactory");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(ICUResourceBundleFactory)
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/servslkf.cpp b/thirdparty/icu4c/common/servslkf.cpp
new file mode 100644
index 0000000000..09154d1b91
--- /dev/null
+++ b/thirdparty/icu4c/common/servslkf.cpp
@@ -0,0 +1,123 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/**
+ *******************************************************************************
+ * Copyright (C) 2001-2014, International Business Machines Corporation and    *
+ * others. All Rights Reserved.                                                *
+ *******************************************************************************
+ *
+ *******************************************************************************
+ */
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_SERVICE
+
+#include "unicode/resbund.h"
+#include "uresimp.h"
+#include "cmemory.h"
+#include "servloc.h"
+#include "ustrfmt.h"
+#include "uhash.h"
+#include "charstr.h"
+#include "uassert.h"
+
+#define UNDERSCORE_CHAR ((UChar)0x005f)
+#define AT_SIGN_CHAR    ((UChar)64)
+#define PERIOD_CHAR     ((UChar)46)
+
+U_NAMESPACE_BEGIN
+
+/*
+ ******************************************************************
+ */
+
+SimpleLocaleKeyFactory::SimpleLocaleKeyFactory(UObject* objToAdopt,
+                                               const UnicodeString& locale,
+                                               int32_t kind,
+                                               int32_t coverage)
+  : LocaleKeyFactory(coverage)
+  , _obj(objToAdopt)
+  , _id(locale)
+  , _kind(kind)
+{
+}
+
+SimpleLocaleKeyFactory::SimpleLocaleKeyFactory(UObject* objToAdopt,
+                                               const Locale& locale,
+                                               int32_t kind,
+                                               int32_t coverage)
+  : LocaleKeyFactory(coverage)
+  , _obj(objToAdopt)
+  , _id()
+  , _kind(kind)
+{
+    LocaleUtility::initNameFromLocale(locale, _id);
+}
+
+SimpleLocaleKeyFactory::~SimpleLocaleKeyFactory()
+{
+  delete _obj;
+  _obj = NULL;
+}
+
+UObject*
+SimpleLocaleKeyFactory::create(const ICUServiceKey& key, const ICUService* service, UErrorCode& status) const
+{
+    if (U_SUCCESS(status)) {
+        const LocaleKey& lkey = (const LocaleKey&)key;
+        if (_kind == LocaleKey::KIND_ANY || _kind == lkey.kind()) {
+            UnicodeString keyID;
+            lkey.currentID(keyID);
+            if (_id == keyID) {
+                return service->cloneInstance(_obj);
+            }
+        }
+    }
+    return NULL;
+}
+
+//UBool
+//SimpleLocaleKeyFactory::isSupportedID(const UnicodeString& id, UErrorCode& /* status */) const
+//{
+//    return id == _id;
+//}
+
+void
+SimpleLocaleKeyFactory::updateVisibleIDs(Hashtable& result, UErrorCode& status) const
+{
+    if (U_SUCCESS(status)) {
+        if (_coverage & 0x1) {
+            result.remove(_id);
+        } else {
+            result.put(_id, (void*)this, status);
+        }
+    }
+}
+
+#ifdef SERVICE_DEBUG
+UnicodeString&
+SimpleLocaleKeyFactory::debug(UnicodeString& result) const
+{
+    LocaleKeyFactory::debug(result);
+    result.append((UnicodeString)", id: ");
+    result.append(_id);
+    result.append((UnicodeString)", kind: ");
+    result.append(_kind);
+    return result;
+}
+
+UnicodeString&
+SimpleLocaleKeyFactory::debugClass(UnicodeString& result) const
+{
+    return result.append((UnicodeString)"SimpleLocaleKeyFactory");
+}
+#endif
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(SimpleLocaleKeyFactory)
+
+U_NAMESPACE_END
+
+/* !UCONFIG_NO_SERVICE */
+#endif
+
+
diff --git a/thirdparty/icu4c/common/sharedobject.cpp b/thirdparty/icu4c/common/sharedobject.cpp
new file mode 100644
index 0000000000..6eeca8605f
--- /dev/null
+++ b/thirdparty/icu4c/common/sharedobject.cpp
@@ -0,0 +1,62 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015, International Business Machines
+* Corporation and others.  All Rights Reserved.
+******************************************************************************
+* sharedobject.cpp
+*/
+#include "sharedobject.h"
+#include "mutex.h"
+#include "uassert.h"
+#include "umutex.h"
+#include "unifiedcache.h"
+
+U_NAMESPACE_BEGIN
+
+SharedObject::~SharedObject() {}
+
+UnifiedCacheBase::~UnifiedCacheBase() {}
+
+void
+SharedObject::addRef() const {
+    umtx_atomic_inc(&hardRefCount);
+}
+
+// removeRef Decrement the reference count and delete if it is zero.
+//           Note that SharedObjects with a non-null cachePtr are owned by the
+//           unified cache, and the cache will be responsible for the actual deletion.
+//           The deletion could be as soon as immediately following the
+//           update to the reference count, if another thread is running
+//           a cache eviction cycle concurrently.
+//           NO ACCESS TO *this PERMITTED AFTER REFERENCE COUNT == 0 for cached objects.
+//           THE OBJECT MAY ALREADY BE GONE.
+void
+SharedObject::removeRef() const {
+    const UnifiedCacheBase *cache = this->cachePtr;
+    int32_t updatedRefCount = umtx_atomic_dec(&hardRefCount);
+    U_ASSERT(updatedRefCount >= 0);
+    if (updatedRefCount == 0) {
+        if (cache) {
+            cache->handleUnreferencedObject();
+        } else {
+            delete this;
+        }
+    }
+}
+
+
+int32_t
+SharedObject::getRefCount() const {
+    return umtx_loadAcquire(hardRefCount);
+}
+
+void
+SharedObject::deleteIfZeroRefCount() const {
+    if (this->cachePtr == nullptr && getRefCount() == 0) {
+        delete this;
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/sharedobject.h b/thirdparty/icu4c/common/sharedobject.h
new file mode 100644
index 0000000000..6ccfb27b01
--- /dev/null
+++ b/thirdparty/icu4c/common/sharedobject.h
@@ -0,0 +1,184 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+******************************************************************************
+* sharedobject.h
+*/
+
+#ifndef __SHAREDOBJECT_H__
+#define __SHAREDOBJECT_H__
+
+
+#include "unicode/uobject.h"
+#include "umutex.h"
+
+U_NAMESPACE_BEGIN
+
+class SharedObject;
+
+/**
+ * Base class for unified cache exposing enough methods to SharedObject
+ * instances to allow their addRef() and removeRef() methods to
+ * update cache metrics. No other part of ICU, except for SharedObject,
+ * should directly call the methods of this base class.
+ */
+class U_COMMON_API UnifiedCacheBase : public UObject {
+public:
+    UnifiedCacheBase() { }
+
+    /**
+     * Notify the cache implementation that an object was seen transitioning to
+     * zero hard references. The cache may use this to keep track the number of
+     * unreferenced SharedObjects, and to trigger evictions.
+     */
+    virtual void handleUnreferencedObject() const = 0;
+
+    virtual ~UnifiedCacheBase();
+private:
+    UnifiedCacheBase(const UnifiedCacheBase &);
+    UnifiedCacheBase &operator=(const UnifiedCacheBase &);
+};
+
+/**
+ * Base class for shared, reference-counted, auto-deleted objects.
+ * Subclasses can be immutable.
+ * If they are mutable, then they must implement their copy constructor
+ * so that copyOnWrite() works.
+ *
+ * Either stack-allocate, use LocalPointer, or use addRef()/removeRef().
+ * Sharing requires reference-counting.
+ */
+class U_COMMON_API SharedObject : public UObject {
+public:
+    /** Initializes totalRefCount, softRefCount to 0. */
+    SharedObject() :
+            softRefCount(0),
+            hardRefCount(0),
+            cachePtr(NULL) {}
+
+    /** Initializes totalRefCount, softRefCount to 0. */
+    SharedObject(const SharedObject &other) :
+            UObject(other),
+            softRefCount(0),
+            hardRefCount(0),
+            cachePtr(NULL) {}
+
+    virtual ~SharedObject();
+
+    /**
+     * Increments the number of hard references to this object. Thread-safe.
+     * Not for use from within the Unified Cache implementation.
+     */
+    void addRef() const;
+
+    /**
+     * Decrements the number of hard references to this object, and
+     * arrange for possible cache-eviction and/or deletion if ref
+     * count goes to zero. Thread-safe.
+     * 
+     * Not for use from within the UnifiedCache implementation.
+     */
+    void removeRef() const;
+
+    /**
+     * Returns the number of hard references for this object.
+     * Uses a memory barrier.
+     */
+    int32_t getRefCount() const;
+
+    /**
+     * If noHardReferences() == true then this object has no hard references.
+     * Must be called only from within the internals of UnifiedCache.
+     */
+    inline UBool noHardReferences() const { return getRefCount() == 0; }
+
+    /**
+     * If hasHardReferences() == true then this object has hard references.
+     * Must be called only from within the internals of UnifiedCache.
+     */
+    inline UBool hasHardReferences() const { return getRefCount() != 0; }
+
+    /**
+     * Deletes this object if it has no references.
+     * Available for non-cached SharedObjects only. Ownership of cached objects
+     * is with the UnifiedCache, which is solely responsible for eviction and deletion.
+     */
+    void deleteIfZeroRefCount() const;
+
+        
+    /**
+     * Returns a writable version of ptr.
+     * If there is exactly one owner, then ptr itself is returned as a
+     *  non-const pointer.
+     * If there are multiple owners, then ptr is replaced with a 
+     * copy-constructed clone,
+     * and that is returned.
+     * Returns NULL if cloning failed.
+     *
+     * T must be a subclass of SharedObject.
+     */
+    template<typename T>
+    static T *copyOnWrite(const T *&ptr) {
+        const T *p = ptr;
+        if(p->getRefCount() <= 1) { return const_cast<T *>(p); }
+        T *p2 = new T(*p);
+        if(p2 == NULL) { return NULL; }
+        p->removeRef();
+        ptr = p2;
+        p2->addRef();
+        return p2;
+    }
+
+    /**
+     * Makes dest an owner of the object pointed to by src while adjusting
+     * reference counts and deleting the previous object dest pointed to
+     * if necessary. Before this call is made, dest must either be NULL or
+     * be included in the reference count of the object it points to. 
+     *
+     * T must be a subclass of SharedObject.
+     */
+    template<typename T>
+    static void copyPtr(const T *src, const T *&dest) {
+        if(src != dest) {
+            if(dest != NULL) { dest->removeRef(); }
+            dest = src;
+            if(src != NULL) { src->addRef(); }
+        }
+    }
+
+    /**
+     * Equivalent to copyPtr(NULL, dest).
+     */
+    template<typename T>
+    static void clearPtr(const T *&ptr) {
+        if (ptr != NULL) {
+            ptr->removeRef();
+            ptr = NULL;
+        }
+    }
+
+private:
+    /**
+     * The number of references from the UnifiedCache, which is
+     * the number of times that the sharedObject is stored as a hash table value.
+     * For use by UnifiedCache implementation code only.
+     * All access is synchronized by UnifiedCache's gCacheMutex
+     */
+    mutable int32_t softRefCount;
+    friend class UnifiedCache;
+
+    /**
+     * Reference count, excluding references from within the UnifiedCache implementation.
+     */
+    mutable u_atomic_int32_t hardRefCount;
+    
+    mutable const UnifiedCacheBase *cachePtr;
+
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/simpleformatter.cpp b/thirdparty/icu4c/common/simpleformatter.cpp
new file mode 100644
index 0000000000..f7f7aead61
--- /dev/null
+++ b/thirdparty/icu4c/common/simpleformatter.cpp
@@ -0,0 +1,325 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2014-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+******************************************************************************
+* simpleformatter.cpp
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/simpleformatter.h"
+#include "unicode/unistr.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+/**
+ * Argument numbers must be smaller than this limit.
+ * Text segment lengths are offset by this much.
+ * This is currently the only unused char value in compiled patterns,
+ * except it is the maximum value of the first unit (max arg +1).
+ */
+const int32_t ARG_NUM_LIMIT = 0x100;
+/**
+ * Initial and maximum char/UChar value set for a text segment.
+ * Segment length char values are from ARG_NUM_LIMIT+1 to this value here.
+ * Normally 0xffff, but can be as small as ARG_NUM_LIMIT+1 for testing.
+ */
+const UChar SEGMENT_LENGTH_PLACEHOLDER_CHAR = 0xffff;
+/**
+ * Maximum length of a text segment. Longer segments are split into shorter ones.
+ */
+const int32_t MAX_SEGMENT_LENGTH = SEGMENT_LENGTH_PLACEHOLDER_CHAR - ARG_NUM_LIMIT;
+
+enum {
+    APOS = 0x27,
+    DIGIT_ZERO = 0x30,
+    DIGIT_ONE = 0x31,
+    DIGIT_NINE = 0x39,
+    OPEN_BRACE = 0x7b,
+    CLOSE_BRACE = 0x7d
+};
+
+inline UBool isInvalidArray(const void *array, int32_t length) {
+   return (length < 0 || (array == NULL && length != 0));
+}
+
+}  // namespace
+
+SimpleFormatter &SimpleFormatter::operator=(const SimpleFormatter& other) {
+    if (this == &other) {
+        return *this;
+    }
+    compiledPattern = other.compiledPattern;
+    return *this;
+}
+
+SimpleFormatter::~SimpleFormatter() {}
+
+UBool SimpleFormatter::applyPatternMinMaxArguments(
+        const UnicodeString &pattern,
+        int32_t min, int32_t max,
+        UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    // Parse consistent with MessagePattern, but
+    // - support only simple numbered arguments
+    // - build a simple binary structure into the result string
+    const UChar *patternBuffer = pattern.getBuffer();
+    int32_t patternLength = pattern.length();
+    // Reserve the first char for the number of arguments.
+    compiledPattern.setTo((UChar)0);
+    int32_t textLength = 0;
+    int32_t maxArg = -1;
+    UBool inQuote = FALSE;
+    for (int32_t i = 0; i < patternLength;) {
+        UChar c = patternBuffer[i++];
+        if (c == APOS) {
+            if (i < patternLength && (c = patternBuffer[i]) == APOS) {
+                // double apostrophe, skip the second one
+                ++i;
+            } else if (inQuote) {
+                // skip the quote-ending apostrophe
+                inQuote = FALSE;
+                continue;
+            } else if (c == OPEN_BRACE || c == CLOSE_BRACE) {
+                // Skip the quote-starting apostrophe, find the end of the quoted literal text.
+                ++i;
+                inQuote = TRUE;
+            } else {
+                // The apostrophe is part of literal text.
+                c = APOS;
+            }
+        } else if (!inQuote && c == OPEN_BRACE) {
+            if (textLength > 0) {
+                compiledPattern.setCharAt(compiledPattern.length() - textLength - 1,
+                                          (UChar)(ARG_NUM_LIMIT + textLength));
+                textLength = 0;
+            }
+            int32_t argNumber;
+            if ((i + 1) < patternLength &&
+                    0 <= (argNumber = patternBuffer[i] - DIGIT_ZERO) && argNumber <= 9 &&
+                    patternBuffer[i + 1] == CLOSE_BRACE) {
+                i += 2;
+            } else {
+                // Multi-digit argument number (no leading zero) or syntax error.
+                // MessagePattern permits PatternProps.skipWhiteSpace(pattern, index)
+                // around the number, but this class does not.
+                argNumber = -1;
+                if (i < patternLength && DIGIT_ONE <= (c = patternBuffer[i++]) && c <= DIGIT_NINE) {
+                    argNumber = c - DIGIT_ZERO;
+                    while (i < patternLength &&
+                            DIGIT_ZERO <= (c = patternBuffer[i++]) && c <= DIGIT_NINE) {
+                        argNumber = argNumber * 10 + (c - DIGIT_ZERO);
+                        if (argNumber >= ARG_NUM_LIMIT) {
+                            break;
+                        }
+                    }
+                }
+                if (argNumber < 0 || c != CLOSE_BRACE) {
+                    errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+                    return FALSE;
+                }
+            }
+            if (argNumber > maxArg) {
+                maxArg = argNumber;
+            }
+            compiledPattern.append((UChar)argNumber);
+            continue;
+        }  // else: c is part of literal text
+        // Append c and track the literal-text segment length.
+        if (textLength == 0) {
+            // Reserve a char for the length of a new text segment, preset the maximum length.
+            compiledPattern.append(SEGMENT_LENGTH_PLACEHOLDER_CHAR);
+        }
+        compiledPattern.append(c);
+        if (++textLength == MAX_SEGMENT_LENGTH) {
+            textLength = 0;
+        }
+    }
+    if (textLength > 0) {
+        compiledPattern.setCharAt(compiledPattern.length() - textLength - 1,
+                                  (UChar)(ARG_NUM_LIMIT + textLength));
+    }
+    int32_t argCount = maxArg + 1;
+    if (argCount < min || max < argCount) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    compiledPattern.setCharAt(0, (UChar)argCount);
+    return TRUE;
+}
+
+UnicodeString& SimpleFormatter::format(
+        const UnicodeString &value0,
+        UnicodeString &appendTo, UErrorCode &errorCode) const {
+    const UnicodeString *values[] = { &value0 };
+    return formatAndAppend(values, 1, appendTo, NULL, 0, errorCode);
+}
+
+UnicodeString& SimpleFormatter::format(
+        const UnicodeString &value0,
+        const UnicodeString &value1,
+        UnicodeString &appendTo, UErrorCode &errorCode) const {
+    const UnicodeString *values[] = { &value0, &value1 };
+    return formatAndAppend(values, 2, appendTo, NULL, 0, errorCode);
+}
+
+UnicodeString& SimpleFormatter::format(
+        const UnicodeString &value0,
+        const UnicodeString &value1,
+        const UnicodeString &value2,
+        UnicodeString &appendTo, UErrorCode &errorCode) const {
+    const UnicodeString *values[] = { &value0, &value1, &value2 };
+    return formatAndAppend(values, 3, appendTo, NULL, 0, errorCode);
+}
+
+UnicodeString& SimpleFormatter::formatAndAppend(
+        const UnicodeString *const *values, int32_t valuesLength,
+        UnicodeString &appendTo,
+        int32_t *offsets, int32_t offsetsLength, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return appendTo;
+    }
+    if (isInvalidArray(values, valuesLength) || isInvalidArray(offsets, offsetsLength) ||
+            valuesLength < getArgumentLimit()) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return appendTo;
+    }
+    return format(compiledPattern.getBuffer(), compiledPattern.length(), values,
+                  appendTo, NULL, TRUE,
+                  offsets, offsetsLength, errorCode);
+}
+
+UnicodeString &SimpleFormatter::formatAndReplace(
+        const UnicodeString *const *values, int32_t valuesLength,
+        UnicodeString &result,
+        int32_t *offsets, int32_t offsetsLength, UErrorCode &errorCode) const {
+    if (U_FAILURE(errorCode)) {
+        return result;
+    }
+    if (isInvalidArray(values, valuesLength) || isInvalidArray(offsets, offsetsLength)) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return result;
+    }
+    const UChar *cp = compiledPattern.getBuffer();
+    int32_t cpLength = compiledPattern.length();
+    if (valuesLength < getArgumentLimit(cp, cpLength)) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return result;
+    }
+
+    // If the pattern starts with an argument whose value is the same object
+    // as the result, then we keep the result contents and append to it.
+    // Otherwise we replace its contents.
+    int32_t firstArg = -1;
+    // If any non-initial argument value is the same object as the result,
+    // then we first copy its contents and use that instead while formatting.
+    UnicodeString resultCopy;
+    if (getArgumentLimit(cp, cpLength) > 0) {
+        for (int32_t i = 1; i < cpLength;) {
+            int32_t n = cp[i++];
+            if (n < ARG_NUM_LIMIT) {
+                if (values[n] == &result) {
+                    if (i == 2) {
+                        firstArg = n;
+                    } else if (resultCopy.isEmpty() && !result.isEmpty()) {
+                        resultCopy = result;
+                    }
+                }
+            } else {
+                i += n - ARG_NUM_LIMIT;
+            }
+        }
+    }
+    if (firstArg < 0) {
+        result.remove();
+    }
+    return format(cp, cpLength, values,
+                  result, &resultCopy, FALSE,
+                  offsets, offsetsLength, errorCode);
+}
+
+UnicodeString SimpleFormatter::getTextWithNoArguments(
+        const UChar *compiledPattern,
+        int32_t compiledPatternLength,
+        int32_t* offsets,
+        int32_t offsetsLength) {
+    for (int32_t i = 0; i < offsetsLength; i++) {
+        offsets[i] = -1;
+    }
+    int32_t capacity = compiledPatternLength - 1 -
+            getArgumentLimit(compiledPattern, compiledPatternLength);
+    UnicodeString sb(capacity, 0, 0);  // Java: StringBuilder
+    for (int32_t i = 1; i < compiledPatternLength;) {
+        int32_t n = compiledPattern[i++];
+        if (n > ARG_NUM_LIMIT) {
+            n -= ARG_NUM_LIMIT;
+            sb.append(compiledPattern + i, n);
+            i += n;
+        } else if (n < offsetsLength) {
+            // TODO(ICU-20406): This does not distinguish between "{0}{1}" and "{1}{0}".
+            // Consider removing this function and replacing it with an iterator interface.
+            offsets[n] = sb.length();
+        }
+    }
+    return sb;
+}
+
+UnicodeString &SimpleFormatter::format(
+        const UChar *compiledPattern, int32_t compiledPatternLength,
+        const UnicodeString *const *values,
+        UnicodeString &result, const UnicodeString *resultCopy, UBool forbidResultAsValue,
+        int32_t *offsets, int32_t offsetsLength,
+        UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return result;
+    }
+    for (int32_t i = 0; i < offsetsLength; i++) {
+        offsets[i] = -1;
+    }
+    for (int32_t i = 1; i < compiledPatternLength;) {
+        int32_t n = compiledPattern[i++];
+        if (n < ARG_NUM_LIMIT) {
+            const UnicodeString *value = values[n];
+            if (value == NULL) {
+                errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+                return result;
+            }
+            if (value == &result) {
+                if (forbidResultAsValue) {
+                    errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+                    return result;
+                }
+                if (i == 2) {
+                    // We are appending to result which is also the first value object.
+                    if (n < offsetsLength) {
+                        offsets[n] = 0;
+                    }
+                } else {
+                    if (n < offsetsLength) {
+                        offsets[n] = result.length();
+                    }
+                    result.append(*resultCopy);
+                }
+            } else {
+                if (n < offsetsLength) {
+                    offsets[n] = result.length();
+                }
+                result.append(*value);
+            }
+        } else {
+            int32_t length = n - ARG_NUM_LIMIT;
+            result.append(compiledPattern + i, length);
+            i += length;
+        }
+    }
+    return result;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/sprpimpl.h b/thirdparty/icu4c/common/sprpimpl.h
new file mode 100644
index 0000000000..ca0bcdb516
--- /dev/null
+++ b/thirdparty/icu4c/common/sprpimpl.h
@@ -0,0 +1,130 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ *
+ *   Copyright (C) 2003-2006, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *******************************************************************************
+ *   file name:  sprpimpl.h
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2003feb1
+ *   created by: Ram Viswanadha
+ */
+
+#ifndef SPRPIMPL_H
+#define SPRPIMPL_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/ustring.h"
+#include "unicode/parseerr.h"
+#include "unicode/usprep.h"
+#include "unicode/udata.h"
+#include "utrie.h"
+#include "udataswp.h"
+#include "ubidi_props.h"
+
+#define _SPREP_DATA_TYPE "spp"
+
+enum UStringPrepType{
+    USPREP_UNASSIGNED           = 0x0000 ,
+    USPREP_MAP                  = 0x0001 ,
+    USPREP_PROHIBITED           = 0x0002 , 
+    USPREP_DELETE               = 0x0003 ,
+    USPREP_TYPE_LIMIT           = 0x0004  
+};
+
+typedef enum UStringPrepType UStringPrepType;
+
+#ifdef USPREP_TYPE_NAMES_ARRAY
+static const char* usprepTypeNames[] ={
+    "UNASSIGNED" ,          
+    "MAP" , 
+    "PROHIBITED" ,        
+    "DELETE",
+    "TYPE_LIMIT" 
+};
+#endif
+
+enum{
+    _SPREP_NORMALIZATION_ON = 0x0001,
+    _SPREP_CHECK_BIDI_ON    = 0x0002
+};
+
+enum{
+    _SPREP_TYPE_THRESHOLD       = 0xFFF0,
+    _SPREP_MAX_INDEX_VALUE      = 0x3FBF,   /*16139*/ 
+    _SPREP_MAX_INDEX_TOP_LENGTH = 0x0003
+};
+
+/* indexes[] value names */
+enum {
+    _SPREP_INDEX_TRIE_SIZE                  = 0, /* number of bytes in StringPrep trie */
+    _SPREP_INDEX_MAPPING_DATA_SIZE          = 1, /* The array that contains the mapping   */
+    _SPREP_NORM_CORRECTNS_LAST_UNI_VERSION  = 2, /* The index of Unicode version of last entry in NormalizationCorrections.txt */ 
+    _SPREP_ONE_UCHAR_MAPPING_INDEX_START    = 3, /* The starting index of 1 UChar mapping index in the mapping data array */
+    _SPREP_TWO_UCHARS_MAPPING_INDEX_START   = 4, /* The starting index of 2 UChars mapping index in the mapping data array */
+    _SPREP_THREE_UCHARS_MAPPING_INDEX_START = 5, /* The starting index of 3 UChars mapping index in the mapping data array */
+    _SPREP_FOUR_UCHARS_MAPPING_INDEX_START  = 6, /* The starting index of 4 UChars mapping index in the mapping data array */
+    _SPREP_OPTIONS                          = 7, /* Bit set of options to turn on in the profile */
+    _SPREP_INDEX_TOP=16                          /* changing this requires a new formatVersion */
+};
+
+typedef struct UStringPrepKey UStringPrepKey;
+
+
+struct UStringPrepKey{
+    char* name;
+    char* path;
+};
+
+struct UStringPrepProfile{
+    int32_t indexes[_SPREP_INDEX_TOP];
+    UTrie sprepTrie;
+    const uint16_t* mappingData;
+    UDataMemory* sprepData;
+    int32_t refCount;
+    UBool isDataLoaded;
+    UBool doNFKC;
+    UBool checkBiDi;
+};
+
+/**
+ * Helper function for populating the UParseError struct
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+uprv_syntaxError(const UChar* rules, 
+                 int32_t pos,
+                 int32_t rulesLen,
+                 UParseError* parseError);
+
+
+/**
+ * Swap StringPrep .spp profile data. See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+usprep_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode);
+
+#endif /* #if !UCONFIG_NO_IDNA */
+
+#endif
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/static_unicode_sets.cpp b/thirdparty/icu4c/common/static_unicode_sets.cpp
new file mode 100644
index 0000000000..5dab3931a7
--- /dev/null
+++ b/thirdparty/icu4c/common/static_unicode_sets.cpp
@@ -0,0 +1,245 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+// Allow implicit conversion from char16_t* to UnicodeString for this file:
+// Helpful in toString methods and elsewhere.
+#define UNISTR_FROM_STRING_EXPLICIT
+
+#include "static_unicode_sets.h"
+#include "umutex.h"
+#include "ucln_cmn.h"
+#include "unicode/uniset.h"
+#include "uresimp.h"
+#include "cstring.h"
+#include "uassert.h"
+
+using namespace icu;
+using namespace icu::unisets;
+
+
+namespace {
+
+UnicodeSet* gUnicodeSets[UNISETS_KEY_COUNT] = {};
+
+// Save the empty instance in static memory to have well-defined behavior if a
+// regular UnicodeSet cannot be allocated.
+alignas(UnicodeSet)
+char gEmptyUnicodeSet[sizeof(UnicodeSet)];
+
+// Whether the gEmptyUnicodeSet is initialized and ready to use.
+UBool gEmptyUnicodeSetInitialized = FALSE;
+
+inline UnicodeSet* getImpl(Key key) {
+    UnicodeSet* candidate = gUnicodeSets[key];
+    if (candidate == nullptr) {
+        return reinterpret_cast<UnicodeSet*>(gEmptyUnicodeSet);
+    }
+    return candidate;
+}
+
+UnicodeSet* computeUnion(Key k1, Key k2) {
+    UnicodeSet* result = new UnicodeSet();
+    if (result == nullptr) {
+        return nullptr;
+    }
+    result->addAll(*getImpl(k1));
+    result->addAll(*getImpl(k2));
+    result->freeze();
+    return result;
+}
+
+UnicodeSet* computeUnion(Key k1, Key k2, Key k3) {
+    UnicodeSet* result = new UnicodeSet();
+    if (result == nullptr) {
+        return nullptr;
+    }
+    result->addAll(*getImpl(k1));
+    result->addAll(*getImpl(k2));
+    result->addAll(*getImpl(k3));
+    result->freeze();
+    return result;
+}
+
+
+void saveSet(Key key, const UnicodeString& unicodeSetPattern, UErrorCode& status) {
+    // assert unicodeSets.get(key) == null;
+    gUnicodeSets[key] = new UnicodeSet(unicodeSetPattern, status);
+}
+
+class ParseDataSink : public ResourceSink {
+  public:
+    void put(const char* key, ResourceValue& value, UBool /*noFallback*/, UErrorCode& status) U_OVERRIDE {
+        ResourceTable contextsTable = value.getTable(status);
+        if (U_FAILURE(status)) { return; }
+        for (int i = 0; contextsTable.getKeyAndValue(i, key, value); i++) {
+            if (uprv_strcmp(key, "date") == 0) {
+                // ignore
+            } else {
+                ResourceTable strictnessTable = value.getTable(status);
+                if (U_FAILURE(status)) { return; }
+                for (int j = 0; strictnessTable.getKeyAndValue(j, key, value); j++) {
+                    bool isLenient = (uprv_strcmp(key, "lenient") == 0);
+                    ResourceArray array = value.getArray(status);
+                    if (U_FAILURE(status)) { return; }
+                    for (int k = 0; k < array.getSize(); k++) {
+                        array.getValue(k, value);
+                        UnicodeString str = value.getUnicodeString(status);
+                        if (U_FAILURE(status)) { return; }
+                        // There is both lenient and strict data for comma/period,
+                        // but not for any of the other symbols.
+                        if (str.indexOf(u'.') != -1) {
+                            saveSet(isLenient ? PERIOD : STRICT_PERIOD, str, status);
+                        } else if (str.indexOf(u',') != -1) {
+                            saveSet(isLenient ? COMMA : STRICT_COMMA, str, status);
+                        } else if (str.indexOf(u'+') != -1) {
+                            saveSet(PLUS_SIGN, str, status);
+                        } else if (str.indexOf(u'-') != -1) {
+                            saveSet(MINUS_SIGN, str, status);
+                        } else if (str.indexOf(u'$') != -1) {
+                            saveSet(DOLLAR_SIGN, str, status);
+                        } else if (str.indexOf(u'£') != -1) {
+                            saveSet(POUND_SIGN, str, status);
+                        } else if (str.indexOf(u'₹') != -1) {
+                            saveSet(RUPEE_SIGN, str, status);
+                        } else if (str.indexOf(u'¥') != -1) {
+                            saveSet(YEN_SIGN, str, status);
+                        } else if (str.indexOf(u'₩') != -1) {
+                            saveSet(WON_SIGN, str, status);
+                        } else if (str.indexOf(u'%') != -1) {
+                            saveSet(PERCENT_SIGN, str, status);
+                        } else if (str.indexOf(u'‰') != -1) {
+                            saveSet(PERMILLE_SIGN, str, status);
+                        } else if (str.indexOf(u'’') != -1) {
+                            saveSet(APOSTROPHE_SIGN, str, status);
+                        } else {
+                            // Unknown class of parse lenients
+                            // TODO(ICU-20428): Make ICU automatically accept new classes?
+                            U_ASSERT(FALSE);
+                        }
+                        if (U_FAILURE(status)) { return; }
+                    }
+                }
+            }
+        }
+    }
+};
+
+
+icu::UInitOnce gNumberParseUniSetsInitOnce = U_INITONCE_INITIALIZER;
+
+UBool U_CALLCONV cleanupNumberParseUniSets() {
+    if (gEmptyUnicodeSetInitialized) {
+        reinterpret_cast<UnicodeSet*>(gEmptyUnicodeSet)->~UnicodeSet();
+        gEmptyUnicodeSetInitialized = FALSE;
+    }
+    for (int32_t i = 0; i < UNISETS_KEY_COUNT; i++) {
+        delete gUnicodeSets[i];
+        gUnicodeSets[i] = nullptr;
+    }
+    gNumberParseUniSetsInitOnce.reset();
+    return TRUE;
+}
+
+void U_CALLCONV initNumberParseUniSets(UErrorCode& status) {
+    ucln_common_registerCleanup(UCLN_COMMON_NUMPARSE_UNISETS, cleanupNumberParseUniSets);
+
+    // Initialize the empty instance for well-defined fallback behavior
+    new(gEmptyUnicodeSet) UnicodeSet();
+    reinterpret_cast<UnicodeSet*>(gEmptyUnicodeSet)->freeze();
+    gEmptyUnicodeSetInitialized = TRUE;
+
+    // These sets were decided after discussion with icu-design@. See tickets #13084 and #13309.
+    // Zs+TAB is "horizontal whitespace" according to UTS #18 (blank property).
+    gUnicodeSets[DEFAULT_IGNORABLES] = new UnicodeSet(
+            u"[[:Zs:][\\u0009][:Bidi_Control:][:Variation_Selector:]]", status);
+    gUnicodeSets[STRICT_IGNORABLES] = new UnicodeSet(u"[[:Bidi_Control:]]", status);
+
+    LocalUResourceBundlePointer rb(ures_open(nullptr, "root", &status));
+    if (U_FAILURE(status)) { return; }
+    ParseDataSink sink;
+    ures_getAllItemsWithFallback(rb.getAlias(), "parse", sink, status);
+    if (U_FAILURE(status)) { return; }
+
+    // NOTE: It is OK for these assertions to fail if there was a no-data build.
+    U_ASSERT(gUnicodeSets[COMMA] != nullptr);
+    U_ASSERT(gUnicodeSets[STRICT_COMMA] != nullptr);
+    U_ASSERT(gUnicodeSets[PERIOD] != nullptr);
+    U_ASSERT(gUnicodeSets[STRICT_PERIOD] != nullptr);
+    U_ASSERT(gUnicodeSets[APOSTROPHE_SIGN] != nullptr);
+
+    LocalPointer<UnicodeSet> otherGrouping(new UnicodeSet(
+        u"[٬‘\\u0020\\u00A0\\u2000-\\u200A\\u202F\\u205F\\u3000]",
+        status
+    ), status);
+    if (U_FAILURE(status)) { return; }
+    otherGrouping->addAll(*gUnicodeSets[APOSTROPHE_SIGN]);
+    gUnicodeSets[OTHER_GROUPING_SEPARATORS] = otherGrouping.orphan();
+    gUnicodeSets[ALL_SEPARATORS] = computeUnion(COMMA, PERIOD, OTHER_GROUPING_SEPARATORS);
+    gUnicodeSets[STRICT_ALL_SEPARATORS] = computeUnion(
+            STRICT_COMMA, STRICT_PERIOD, OTHER_GROUPING_SEPARATORS);
+
+    U_ASSERT(gUnicodeSets[MINUS_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[PLUS_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[PERCENT_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[PERMILLE_SIGN] != nullptr);
+
+    gUnicodeSets[INFINITY_SIGN] = new UnicodeSet(u"[∞]", status);
+    if (U_FAILURE(status)) { return; }
+
+    U_ASSERT(gUnicodeSets[DOLLAR_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[POUND_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[RUPEE_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[YEN_SIGN] != nullptr);
+    U_ASSERT(gUnicodeSets[WON_SIGN] != nullptr);
+
+    gUnicodeSets[DIGITS] = new UnicodeSet(u"[:digit:]", status);
+    if (U_FAILURE(status)) { return; }
+    gUnicodeSets[DIGITS_OR_ALL_SEPARATORS] = computeUnion(DIGITS, ALL_SEPARATORS);
+    gUnicodeSets[DIGITS_OR_STRICT_ALL_SEPARATORS] = computeUnion(DIGITS, STRICT_ALL_SEPARATORS);
+
+    for (auto* uniset : gUnicodeSets) {
+        if (uniset != nullptr) {
+            uniset->freeze();
+        }
+    }
+}
+
+}
+
+const UnicodeSet* unisets::get(Key key) {
+    UErrorCode localStatus = U_ZERO_ERROR;
+    umtx_initOnce(gNumberParseUniSetsInitOnce, &initNumberParseUniSets, localStatus);
+    if (U_FAILURE(localStatus)) {
+        return reinterpret_cast<UnicodeSet*>(gEmptyUnicodeSet);
+    }
+    return getImpl(key);
+}
+
+Key unisets::chooseFrom(UnicodeString str, Key key1) {
+    return get(key1)->contains(str) ? key1 : NONE;
+}
+
+Key unisets::chooseFrom(UnicodeString str, Key key1, Key key2) {
+    return get(key1)->contains(str) ? key1 : chooseFrom(str, key2);
+}
+
+//Key unisets::chooseCurrency(UnicodeString str) {
+//    if (get(DOLLAR_SIGN)->contains(str)) {
+//        return DOLLAR_SIGN;
+//    } else if (get(POUND_SIGN)->contains(str)) {
+//        return POUND_SIGN;
+//    } else if (get(RUPEE_SIGN)->contains(str)) {
+//        return RUPEE_SIGN;
+//    } else if (get(YEN_SIGN)->contains(str)) {
+//        return YEN_SIGN;
+//    } else {
+//        return NONE;
+//    }
+//}
+
+
+#endif /* #if !UCONFIG_NO_FORMATTING */
diff --git a/thirdparty/icu4c/common/static_unicode_sets.h b/thirdparty/icu4c/common/static_unicode_sets.h
new file mode 100644
index 0000000000..5d90ce5908
--- /dev/null
+++ b/thirdparty/icu4c/common/static_unicode_sets.h
@@ -0,0 +1,140 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// This file contains utilities to deal with static-allocated UnicodeSets.
+//
+// Common use case: you write a "private static final" UnicodeSet in Java, and
+// want something similarly easy in C++.  Originally written for number
+// parsing, but this header can be used for other applications.
+//
+// Main entrypoint: `unisets::get(unisets::MY_SET_ID_HERE)`
+//
+// This file is in common instead of i18n because it is needed by ucurr.cpp.
+//
+// Author: sffc
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+#ifndef __STATIC_UNICODE_SETS_H__
+#define __STATIC_UNICODE_SETS_H__
+
+#include "unicode/uniset.h"
+#include "unicode/unistr.h"
+
+U_NAMESPACE_BEGIN
+namespace unisets {
+
+enum Key {
+    // NONE is used to indicate null in chooseFrom().
+    // EMPTY is used to get an empty UnicodeSet.
+    NONE = -1,
+    EMPTY = 0,
+
+    // Ignorables
+    DEFAULT_IGNORABLES,
+    STRICT_IGNORABLES,
+
+    // Separators
+    // Notes:
+    // - COMMA is a superset of STRICT_COMMA
+    // - PERIOD is a superset of SCRICT_PERIOD
+    // - ALL_SEPARATORS is the union of COMMA, PERIOD, and OTHER_GROUPING_SEPARATORS
+    // - STRICT_ALL_SEPARATORS is the union of STRICT_COMMA, STRICT_PERIOD, and OTHER_GRP_SEPARATORS
+    COMMA,
+    PERIOD,
+    STRICT_COMMA,
+    STRICT_PERIOD,
+    APOSTROPHE_SIGN,
+    OTHER_GROUPING_SEPARATORS,
+    ALL_SEPARATORS,
+    STRICT_ALL_SEPARATORS,
+
+    // Symbols
+    MINUS_SIGN,
+    PLUS_SIGN,
+    PERCENT_SIGN,
+    PERMILLE_SIGN,
+    INFINITY_SIGN,
+
+    // Currency Symbols
+    DOLLAR_SIGN,
+    POUND_SIGN,
+    RUPEE_SIGN,
+    YEN_SIGN,
+    WON_SIGN,
+
+    // Other
+    DIGITS,
+
+    // Combined Separators with Digits (for lead code points)
+    DIGITS_OR_ALL_SEPARATORS,
+    DIGITS_OR_STRICT_ALL_SEPARATORS,
+
+    // The number of elements in the enum.
+    UNISETS_KEY_COUNT
+};
+
+/**
+ * Gets the static-allocated UnicodeSet according to the provided key. The
+ * pointer will be deleted during u_cleanup(); the caller should NOT delete it.
+ *
+ * Exported as U_COMMON_API for ucurr.cpp
+ *
+ * This method is always safe and OK to chain: in the case of a memory or other
+ * error, it returns an empty set from static memory.
+ * 
+ * Example:
+ * 
+ *     UBool hasIgnorables = unisets::get(unisets::DEFAULT_IGNORABLES)->contains(...);
+ *
+ * @param key The desired UnicodeSet according to the enum in this file.
+ * @return The requested UnicodeSet. Guaranteed to be frozen and non-null, but
+ *         may be empty if an error occurred during data loading.
+ */
+U_COMMON_API const UnicodeSet* get(Key key);
+
+/**
+ * Checks if the UnicodeSet given by key1 contains the given string.
+ *
+ * Exported as U_COMMON_API for numparse_decimal.cpp
+ *
+ * @param str The string to check.
+ * @param key1 The set to check.
+ * @return key1 if the set contains str, or NONE if not.
+ */
+U_COMMON_API Key chooseFrom(UnicodeString str, Key key1);
+
+/**
+ * Checks if the UnicodeSet given by either key1 or key2 contains the string.
+ *
+ * Exported as U_COMMON_API for numparse_decimal.cpp
+ *
+ * @param str The string to check.
+ * @param key1 The first set to check.
+ * @param key2 The second set to check.
+ * @return key1 if that set contains str; key2 if that set contains str; or
+ *         NONE if neither set contains str.
+ */
+U_COMMON_API Key chooseFrom(UnicodeString str, Key key1, Key key2);
+
+// TODO: Load these from data: ICU-20108
+// Unused in C++:
+// Key chooseCurrency(UnicodeString str);
+// Used instead:
+static const struct {
+    Key key;
+    UChar32 exemplar;
+} kCurrencyEntries[] = {
+    {DOLLAR_SIGN, u'$'},
+    {POUND_SIGN, u'£'},
+    {RUPEE_SIGN, u'₹'},
+    {YEN_SIGN, u'¥'},
+    {WON_SIGN, u'₩'},
+};
+
+} // namespace unisets
+U_NAMESPACE_END
+
+#endif //__STATIC_UNICODE_SETS_H__
+#endif /* #if !UCONFIG_NO_FORMATTING */
diff --git a/thirdparty/icu4c/common/stringpiece.cpp b/thirdparty/icu4c/common/stringpiece.cpp
new file mode 100644
index 0000000000..99089e08ef
--- /dev/null
+++ b/thirdparty/icu4c/common/stringpiece.cpp
@@ -0,0 +1,116 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+// Copyright (C) 2009-2013, International Business Machines
+// Corporation and others. All Rights Reserved.
+//
+// Copyright 2004 and onwards Google Inc.
+//
+// Author: wilsonh@google.com (Wilson Hsieh)
+//
+
+#include "unicode/utypes.h"
+#include "unicode/stringpiece.h"
+#include "cstring.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+StringPiece::StringPiece(const char* str)
+    : ptr_(str), length_((str == NULL) ? 0 : static_cast<int32_t>(uprv_strlen(str))) { }
+
+StringPiece::StringPiece(const StringPiece& x, int32_t pos) {
+  if (pos < 0) {
+    pos = 0;
+  } else if (pos > x.length_) {
+    pos = x.length_;
+  }
+  ptr_ = x.ptr_ + pos;
+  length_ = x.length_ - pos;
+}
+
+StringPiece::StringPiece(const StringPiece& x, int32_t pos, int32_t len) {
+  if (pos < 0) {
+    pos = 0;
+  } else if (pos > x.length_) {
+    pos = x.length_;
+  }
+  if (len < 0) {
+    len = 0;
+  } else if (len > x.length_ - pos) {
+    len = x.length_ - pos;
+  }
+  ptr_ = x.ptr_ + pos;
+  length_ = len;
+}
+
+void StringPiece::set(const char* str) {
+  ptr_ = str;
+  if (str != NULL)
+    length_ = static_cast<int32_t>(uprv_strlen(str));
+  else
+    length_ = 0;
+}
+
+int32_t StringPiece::find(StringPiece needle, int32_t offset) {
+  if (length() == 0 && needle.length() == 0) {
+    return 0;
+  }
+  // TODO: Improve to be better than O(N^2)?
+  for (int32_t i = offset; i < length(); i++) {
+    int32_t j = 0;
+    for (; j < needle.length(); i++, j++) {
+      if (data()[i] != needle.data()[j]) {
+        i -= j;
+        goto outer_end;
+      }
+    }
+    return i - j;
+    outer_end: void();
+  }
+  return -1;
+}
+
+int32_t StringPiece::compare(StringPiece other) {
+  int32_t i = 0;
+  for (; i < length(); i++) {
+    if (i == other.length()) {
+      // this is longer
+      return 1;
+    }
+    char a = data()[i];
+    char b = other.data()[i];
+    if (a < b) {
+      return -1;
+    } else if (a > b) {
+      return 1;
+    }
+  }
+  if (i < other.length()) {
+    // other is longer
+    return -1;
+  }
+  return 0;
+}
+
+U_EXPORT UBool U_EXPORT2
+operator==(const StringPiece& x, const StringPiece& y) {
+  int32_t len = x.size();
+  if (len != y.size()) {
+    return false;
+  }
+  if (len == 0) {
+    return true;
+  }
+  const char* p = x.data();
+  const char* p2 = y.data();
+  // Test last byte in case strings share large common prefix
+  --len;
+  if (p[len] != p2[len]) return false;
+  // At this point we can, but don't have to, ignore the last byte.
+  return uprv_memcmp(p, p2, len) == 0;
+}
+
+
+const int32_t StringPiece::npos = 0x7fffffff;
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/stringtriebuilder.cpp b/thirdparty/icu4c/common/stringtriebuilder.cpp
new file mode 100644
index 0000000000..6f9cc2e5c2
--- /dev/null
+++ b/thirdparty/icu4c/common/stringtriebuilder.cpp
@@ -0,0 +1,618 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  stringtriebuilder.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010dec24
+*   created by: Markus W. Scherer
+*/
+
+#include "utypeinfo.h"  // for 'typeid' to work
+#include "unicode/utypes.h"
+#include "unicode/stringtriebuilder.h"
+#include "uassert.h"
+#include "uhash.h"
+
+U_CDECL_BEGIN
+
+static int32_t U_CALLCONV
+hashStringTrieNode(const UHashTok key) {
+    return icu::StringTrieBuilder::hashNode(key.pointer);
+}
+
+static UBool U_CALLCONV
+equalStringTrieNodes(const UHashTok key1, const UHashTok key2) {
+    return icu::StringTrieBuilder::equalNodes(key1.pointer, key2.pointer);
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+StringTrieBuilder::StringTrieBuilder() : nodes(NULL) {}
+
+StringTrieBuilder::~StringTrieBuilder() {
+    deleteCompactBuilder();
+}
+
+void
+StringTrieBuilder::createCompactBuilder(int32_t sizeGuess, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    nodes=uhash_openSize(hashStringTrieNode, equalStringTrieNodes, NULL,
+                         sizeGuess, &errorCode);
+    if(U_SUCCESS(errorCode)) {
+        if(nodes==NULL) {
+          errorCode=U_MEMORY_ALLOCATION_ERROR;
+        } else {
+          uhash_setKeyDeleter(nodes, uprv_deleteUObject);
+        }
+    }
+}
+
+void
+StringTrieBuilder::deleteCompactBuilder() {
+    uhash_close(nodes);
+    nodes=NULL;
+}
+
+void
+StringTrieBuilder::build(UStringTrieBuildOption buildOption, int32_t elementsLength,
+                       UErrorCode &errorCode) {
+    if(buildOption==USTRINGTRIE_BUILD_FAST) {
+        writeNode(0, elementsLength, 0);
+    } else /* USTRINGTRIE_BUILD_SMALL */ {
+        createCompactBuilder(2*elementsLength, errorCode);
+        Node *root=makeNode(0, elementsLength, 0, errorCode);
+        if(U_SUCCESS(errorCode)) {
+            root->markRightEdgesFirst(-1);
+            root->write(*this);
+        }
+        deleteCompactBuilder();
+    }
+}
+
+// Requires start<limit,
+// and all strings of the [start..limit[ elements must be sorted and
+// have a common prefix of length unitIndex.
+int32_t
+StringTrieBuilder::writeNode(int32_t start, int32_t limit, int32_t unitIndex) {
+    UBool hasValue=FALSE;
+    int32_t value=0;
+    int32_t type;
+    if(unitIndex==getElementStringLength(start)) {
+        // An intermediate or final value.
+        value=getElementValue(start++);
+        if(start==limit) {
+            return writeValueAndFinal(value, TRUE);  // final-value node
+        }
+        hasValue=TRUE;
+    }
+    // Now all [start..limit[ strings are longer than unitIndex.
+    int32_t minUnit=getElementUnit(start, unitIndex);
+    int32_t maxUnit=getElementUnit(limit-1, unitIndex);
+    if(minUnit==maxUnit) {
+        // Linear-match node: All strings have the same character at unitIndex.
+        int32_t lastUnitIndex=getLimitOfLinearMatch(start, limit-1, unitIndex);
+        writeNode(start, limit, lastUnitIndex);
+        // Break the linear-match sequence into chunks of at most kMaxLinearMatchLength.
+        int32_t length=lastUnitIndex-unitIndex;
+        int32_t maxLinearMatchLength=getMaxLinearMatchLength();
+        while(length>maxLinearMatchLength) {
+            lastUnitIndex-=maxLinearMatchLength;
+            length-=maxLinearMatchLength;
+            writeElementUnits(start, lastUnitIndex, maxLinearMatchLength);
+            write(getMinLinearMatch()+maxLinearMatchLength-1);
+        }
+        writeElementUnits(start, unitIndex, length);
+        type=getMinLinearMatch()+length-1;
+    } else {
+        // Branch node.
+        int32_t length=countElementUnits(start, limit, unitIndex);
+        // length>=2 because minUnit!=maxUnit.
+        writeBranchSubNode(start, limit, unitIndex, length);
+        if(--length<getMinLinearMatch()) {
+            type=length;
+        } else {
+            write(length);
+            type=0;
+        }
+    }
+    return writeValueAndType(hasValue, value, type);
+}
+
+// start<limit && all strings longer than unitIndex &&
+// length different units at unitIndex
+int32_t
+StringTrieBuilder::writeBranchSubNode(int32_t start, int32_t limit, int32_t unitIndex, int32_t length) {
+    UChar middleUnits[kMaxSplitBranchLevels];
+    int32_t lessThan[kMaxSplitBranchLevels];
+    int32_t ltLength=0;
+    while(length>getMaxBranchLinearSubNodeLength()) {
+        // Branch on the middle unit.
+        // First, find the middle unit.
+        int32_t i=skipElementsBySomeUnits(start, unitIndex, length/2);
+        // Encode the less-than branch first.
+        middleUnits[ltLength]=getElementUnit(i, unitIndex);  // middle unit
+        lessThan[ltLength]=writeBranchSubNode(start, i, unitIndex, length/2);
+        ++ltLength;
+        // Continue for the greater-or-equal branch.
+        start=i;
+        length=length-length/2;
+    }
+    // For each unit, find its elements array start and whether it has a final value.
+    int32_t starts[kMaxBranchLinearSubNodeLength];
+    UBool isFinal[kMaxBranchLinearSubNodeLength-1];
+    int32_t unitNumber=0;
+    do {
+        int32_t i=starts[unitNumber]=start;
+        UChar unit=getElementUnit(i++, unitIndex);
+        i=indexOfElementWithNextUnit(i, unitIndex, unit);
+        isFinal[unitNumber]= start==i-1 && unitIndex+1==getElementStringLength(start);
+        start=i;
+    } while(++unitNumber<length-1);
+    // unitNumber==length-1, and the maxUnit elements range is [start..limit[
+    starts[unitNumber]=start;
+
+    // Write the sub-nodes in reverse order: The jump lengths are deltas from
+    // after their own positions, so if we wrote the minUnit sub-node first,
+    // then its jump delta would be larger.
+    // Instead we write the minUnit sub-node last, for a shorter delta.
+    int32_t jumpTargets[kMaxBranchLinearSubNodeLength-1];
+    do {
+        --unitNumber;
+        if(!isFinal[unitNumber]) {
+            jumpTargets[unitNumber]=writeNode(starts[unitNumber], starts[unitNumber+1], unitIndex+1);
+        }
+    } while(unitNumber>0);
+    // The maxUnit sub-node is written as the very last one because we do
+    // not jump for it at all.
+    unitNumber=length-1;
+    writeNode(start, limit, unitIndex+1);
+    int32_t offset=write(getElementUnit(start, unitIndex));
+    // Write the rest of this node's unit-value pairs.
+    while(--unitNumber>=0) {
+        start=starts[unitNumber];
+        int32_t value;
+        if(isFinal[unitNumber]) {
+            // Write the final value for the one string ending with this unit.
+            value=getElementValue(start);
+        } else {
+            // Write the delta to the start position of the sub-node.
+            value=offset-jumpTargets[unitNumber];
+        }
+        writeValueAndFinal(value, isFinal[unitNumber]);
+        offset=write(getElementUnit(start, unitIndex));
+    }
+    // Write the split-branch nodes.
+    while(ltLength>0) {
+        --ltLength;
+        writeDeltaTo(lessThan[ltLength]);
+        offset=write(middleUnits[ltLength]);
+    }
+    return offset;
+}
+
+// Requires start<limit,
+// and all strings of the [start..limit[ elements must be sorted and
+// have a common prefix of length unitIndex.
+StringTrieBuilder::Node *
+StringTrieBuilder::makeNode(int32_t start, int32_t limit, int32_t unitIndex, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    UBool hasValue=FALSE;
+    int32_t value=0;
+    if(unitIndex==getElementStringLength(start)) {
+        // An intermediate or final value.
+        value=getElementValue(start++);
+        if(start==limit) {
+            return registerFinalValue(value, errorCode);
+        }
+        hasValue=TRUE;
+    }
+    Node *node;
+    // Now all [start..limit[ strings are longer than unitIndex.
+    int32_t minUnit=getElementUnit(start, unitIndex);
+    int32_t maxUnit=getElementUnit(limit-1, unitIndex);
+    if(minUnit==maxUnit) {
+        // Linear-match node: All strings have the same character at unitIndex.
+        int32_t lastUnitIndex=getLimitOfLinearMatch(start, limit-1, unitIndex);
+        Node *nextNode=makeNode(start, limit, lastUnitIndex, errorCode);
+        // Break the linear-match sequence into chunks of at most kMaxLinearMatchLength.
+        int32_t length=lastUnitIndex-unitIndex;
+        int32_t maxLinearMatchLength=getMaxLinearMatchLength();
+        while(length>maxLinearMatchLength) {
+            lastUnitIndex-=maxLinearMatchLength;
+            length-=maxLinearMatchLength;
+            node=createLinearMatchNode(start, lastUnitIndex, maxLinearMatchLength, nextNode);
+            nextNode=registerNode(node, errorCode);
+        }
+        node=createLinearMatchNode(start, unitIndex, length, nextNode);
+    } else {
+        // Branch node.
+        int32_t length=countElementUnits(start, limit, unitIndex);
+        // length>=2 because minUnit!=maxUnit.
+        Node *subNode=makeBranchSubNode(start, limit, unitIndex, length, errorCode);
+        node=new BranchHeadNode(length, subNode);
+    }
+    if(hasValue && node!=NULL) {
+        if(matchNodesCanHaveValues()) {
+            ((ValueNode *)node)->setValue(value);
+        } else {
+            node=new IntermediateValueNode(value, registerNode(node, errorCode));
+        }
+    }
+    return registerNode(node, errorCode);
+}
+
+// start<limit && all strings longer than unitIndex &&
+// length different units at unitIndex
+StringTrieBuilder::Node *
+StringTrieBuilder::makeBranchSubNode(int32_t start, int32_t limit, int32_t unitIndex,
+                                   int32_t length, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    UChar middleUnits[kMaxSplitBranchLevels];
+    Node *lessThan[kMaxSplitBranchLevels];
+    int32_t ltLength=0;
+    while(length>getMaxBranchLinearSubNodeLength()) {
+        // Branch on the middle unit.
+        // First, find the middle unit.
+        int32_t i=skipElementsBySomeUnits(start, unitIndex, length/2);
+        // Create the less-than branch.
+        middleUnits[ltLength]=getElementUnit(i, unitIndex);  // middle unit
+        lessThan[ltLength]=makeBranchSubNode(start, i, unitIndex, length/2, errorCode);
+        ++ltLength;
+        // Continue for the greater-or-equal branch.
+        start=i;
+        length=length-length/2;
+    }
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    ListBranchNode *listNode=new ListBranchNode();
+    if(listNode==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    // For each unit, find its elements array start and whether it has a final value.
+    int32_t unitNumber=0;
+    do {
+        int32_t i=start;
+        UChar unit=getElementUnit(i++, unitIndex);
+        i=indexOfElementWithNextUnit(i, unitIndex, unit);
+        if(start==i-1 && unitIndex+1==getElementStringLength(start)) {
+            listNode->add(unit, getElementValue(start));
+        } else {
+            listNode->add(unit, makeNode(start, i, unitIndex+1, errorCode));
+        }
+        start=i;
+    } while(++unitNumber<length-1);
+    // unitNumber==length-1, and the maxUnit elements range is [start..limit[
+    UChar unit=getElementUnit(start, unitIndex);
+    if(start==limit-1 && unitIndex+1==getElementStringLength(start)) {
+        listNode->add(unit, getElementValue(start));
+    } else {
+        listNode->add(unit, makeNode(start, limit, unitIndex+1, errorCode));
+    }
+    Node *node=registerNode(listNode, errorCode);
+    // Create the split-branch nodes.
+    while(ltLength>0) {
+        --ltLength;
+        node=registerNode(
+            new SplitBranchNode(middleUnits[ltLength], lessThan[ltLength], node), errorCode);
+    }
+    return node;
+}
+
+StringTrieBuilder::Node *
+StringTrieBuilder::registerNode(Node *newNode, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        delete newNode;
+        return NULL;
+    }
+    if(newNode==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    const UHashElement *old=uhash_find(nodes, newNode);
+    if(old!=NULL) {
+        delete newNode;
+        return (Node *)old->key.pointer;
+    }
+    // If uhash_puti() returns a non-zero value from an equivalent, previously
+    // registered node, then uhash_find() failed to find that and we will leak newNode.
+#if U_DEBUG
+    int32_t oldValue=  // Only in debug mode to avoid a compiler warning about unused oldValue.
+#endif
+    uhash_puti(nodes, newNode, 1, &errorCode);
+    U_ASSERT(oldValue==0);
+    if(U_FAILURE(errorCode)) {
+        delete newNode;
+        return NULL;
+    }
+    return newNode;
+}
+
+StringTrieBuilder::Node *
+StringTrieBuilder::registerFinalValue(int32_t value, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    FinalValueNode key(value);
+    const UHashElement *old=uhash_find(nodes, &key);
+    if(old!=NULL) {
+        return (Node *)old->key.pointer;
+    }
+    Node *newNode=new FinalValueNode(value);
+    if(newNode==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    // If uhash_puti() returns a non-zero value from an equivalent, previously
+    // registered node, then uhash_find() failed to find that and we will leak newNode.
+#if U_DEBUG
+    int32_t oldValue=  // Only in debug mode to avoid a compiler warning about unused oldValue.
+#endif
+    uhash_puti(nodes, newNode, 1, &errorCode);
+    U_ASSERT(oldValue==0);
+    if(U_FAILURE(errorCode)) {
+        delete newNode;
+        return NULL;
+    }
+    return newNode;
+}
+
+int32_t
+StringTrieBuilder::hashNode(const void *node) {
+    return ((const Node *)node)->hashCode();
+}
+
+UBool
+StringTrieBuilder::equalNodes(const void *left, const void *right) {
+    return *(const Node *)left==*(const Node *)right;
+}
+
+UBool
+StringTrieBuilder::Node::operator==(const Node &other) const {
+    return this==&other || (typeid(*this)==typeid(other) && hash==other.hash);
+}
+
+int32_t
+StringTrieBuilder::Node::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        offset=edgeNumber;
+    }
+    return edgeNumber;
+}
+
+UBool
+StringTrieBuilder::FinalValueNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!Node::operator==(other)) {
+        return FALSE;
+    }
+    const FinalValueNode &o=(const FinalValueNode &)other;
+    return value==o.value;
+}
+
+void
+StringTrieBuilder::FinalValueNode::write(StringTrieBuilder &builder) {
+    offset=builder.writeValueAndFinal(value, TRUE);
+}
+
+UBool
+StringTrieBuilder::ValueNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!Node::operator==(other)) {
+        return FALSE;
+    }
+    const ValueNode &o=(const ValueNode &)other;
+    return hasValue==o.hasValue && (!hasValue || value==o.value);
+}
+
+UBool
+StringTrieBuilder::IntermediateValueNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!ValueNode::operator==(other)) {
+        return FALSE;
+    }
+    const IntermediateValueNode &o=(const IntermediateValueNode &)other;
+    return next==o.next;
+}
+
+int32_t
+StringTrieBuilder::IntermediateValueNode::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        offset=edgeNumber=next->markRightEdgesFirst(edgeNumber);
+    }
+    return edgeNumber;
+}
+
+void
+StringTrieBuilder::IntermediateValueNode::write(StringTrieBuilder &builder) {
+    next->write(builder);
+    offset=builder.writeValueAndFinal(value, FALSE);
+}
+
+UBool
+StringTrieBuilder::LinearMatchNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!ValueNode::operator==(other)) {
+        return FALSE;
+    }
+    const LinearMatchNode &o=(const LinearMatchNode &)other;
+    return length==o.length && next==o.next;
+}
+
+int32_t
+StringTrieBuilder::LinearMatchNode::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        offset=edgeNumber=next->markRightEdgesFirst(edgeNumber);
+    }
+    return edgeNumber;
+}
+
+UBool
+StringTrieBuilder::ListBranchNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!Node::operator==(other)) {
+        return FALSE;
+    }
+    const ListBranchNode &o=(const ListBranchNode &)other;
+    for(int32_t i=0; i<length; ++i) {
+        if(units[i]!=o.units[i] || values[i]!=o.values[i] || equal[i]!=o.equal[i]) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+int32_t
+StringTrieBuilder::ListBranchNode::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        firstEdgeNumber=edgeNumber;
+        int32_t step=0;
+        int32_t i=length;
+        do {
+            Node *edge=equal[--i];
+            if(edge!=NULL) {
+                edgeNumber=edge->markRightEdgesFirst(edgeNumber-step);
+            }
+            // For all but the rightmost edge, decrement the edge number.
+            step=1;
+        } while(i>0);
+        offset=edgeNumber;
+    }
+    return edgeNumber;
+}
+
+void
+StringTrieBuilder::ListBranchNode::write(StringTrieBuilder &builder) {
+    // Write the sub-nodes in reverse order: The jump lengths are deltas from
+    // after their own positions, so if we wrote the minUnit sub-node first,
+    // then its jump delta would be larger.
+    // Instead we write the minUnit sub-node last, for a shorter delta.
+    int32_t unitNumber=length-1;
+    Node *rightEdge=equal[unitNumber];
+    int32_t rightEdgeNumber= rightEdge==NULL ? firstEdgeNumber : rightEdge->getOffset();
+    do {
+        --unitNumber;
+        if(equal[unitNumber]!=NULL) {
+            equal[unitNumber]->writeUnlessInsideRightEdge(firstEdgeNumber, rightEdgeNumber, builder);
+        }
+    } while(unitNumber>0);
+    // The maxUnit sub-node is written as the very last one because we do
+    // not jump for it at all.
+    unitNumber=length-1;
+    if(rightEdge==NULL) {
+        builder.writeValueAndFinal(values[unitNumber], TRUE);
+    } else {
+        rightEdge->write(builder);
+    }
+    offset=builder.write(units[unitNumber]);
+    // Write the rest of this node's unit-value pairs.
+    while(--unitNumber>=0) {
+        int32_t value;
+        UBool isFinal;
+        if(equal[unitNumber]==NULL) {
+            // Write the final value for the one string ending with this unit.
+            value=values[unitNumber];
+            isFinal=TRUE;
+        } else {
+            // Write the delta to the start position of the sub-node.
+            U_ASSERT(equal[unitNumber]->getOffset()>0);
+            value=offset-equal[unitNumber]->getOffset();
+            isFinal=FALSE;
+        }
+        builder.writeValueAndFinal(value, isFinal);
+        offset=builder.write(units[unitNumber]);
+    }
+}
+
+UBool
+StringTrieBuilder::SplitBranchNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!Node::operator==(other)) {
+        return FALSE;
+    }
+    const SplitBranchNode &o=(const SplitBranchNode &)other;
+    return unit==o.unit && lessThan==o.lessThan && greaterOrEqual==o.greaterOrEqual;
+}
+
+int32_t
+StringTrieBuilder::SplitBranchNode::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        firstEdgeNumber=edgeNumber;
+        edgeNumber=greaterOrEqual->markRightEdgesFirst(edgeNumber);
+        offset=edgeNumber=lessThan->markRightEdgesFirst(edgeNumber-1);
+    }
+    return edgeNumber;
+}
+
+void
+StringTrieBuilder::SplitBranchNode::write(StringTrieBuilder &builder) {
+    // Encode the less-than branch first.
+    lessThan->writeUnlessInsideRightEdge(firstEdgeNumber, greaterOrEqual->getOffset(), builder);
+    // Encode the greater-or-equal branch last because we do not jump for it at all.
+    greaterOrEqual->write(builder);
+    // Write this node.
+    U_ASSERT(lessThan->getOffset()>0);
+    builder.writeDeltaTo(lessThan->getOffset());  // less-than
+    offset=builder.write(unit);
+}
+
+UBool
+StringTrieBuilder::BranchHeadNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!ValueNode::operator==(other)) {
+        return FALSE;
+    }
+    const BranchHeadNode &o=(const BranchHeadNode &)other;
+    return length==o.length && next==o.next;
+}
+
+int32_t
+StringTrieBuilder::BranchHeadNode::markRightEdgesFirst(int32_t edgeNumber) {
+    if(offset==0) {
+        offset=edgeNumber=next->markRightEdgesFirst(edgeNumber);
+    }
+    return edgeNumber;
+}
+
+void
+StringTrieBuilder::BranchHeadNode::write(StringTrieBuilder &builder) {
+    next->write(builder);
+    if(length<=builder.getMinLinearMatch()) {
+        offset=builder.writeValueAndType(hasValue, value, length-1);
+    } else {
+        builder.write(length-1);
+        offset=builder.writeValueAndType(hasValue, value, 0);
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/uarrsort.cpp b/thirdparty/icu4c/common/uarrsort.cpp
new file mode 100644
index 0000000000..c17dbb2e2b
--- /dev/null
+++ b/thirdparty/icu4c/common/uarrsort.cpp
@@ -0,0 +1,274 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uarrsort.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003aug04
+*   created by: Markus W. Scherer
+*
+*   Internal function for sorting arrays.
+*/
+
+#include <cstddef>
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "uarrsort.h"
+
+enum {
+    /**
+     * "from Knuth"
+     *
+     * A binary search over 8 items performs 4 comparisons:
+     * log2(8)=3 to subdivide, +1 to check for equality.
+     * A linear search over 8 items on average also performs 4 comparisons.
+     */
+    MIN_QSORT=9,
+    STACK_ITEM_SIZE=200
+};
+
+static constexpr int32_t sizeInMaxAlignTs(int32_t sizeInBytes) {
+    return (sizeInBytes + sizeof(std::max_align_t) - 1) / sizeof(std::max_align_t);
+}
+
+/* UComparator convenience implementations ---------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+uprv_uint16Comparator(const void *context, const void *left, const void *right) {
+    (void)context;
+    return (int32_t)*(const uint16_t *)left - (int32_t)*(const uint16_t *)right;
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_int32Comparator(const void *context, const void *left, const void *right) {
+    (void)context;
+    return *(const int32_t *)left - *(const int32_t *)right;
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_uint32Comparator(const void *context, const void *left, const void *right) {
+    (void)context;
+    uint32_t l=*(const uint32_t *)left, r=*(const uint32_t *)right;
+
+    /* compare directly because (l-r) would overflow the int32_t result */
+    if(l<r) {
+        return -1;
+    } else if(l==r) {
+        return 0;
+    } else /* l>r */ {
+        return 1;
+    }
+}
+
+/* Insertion sort using binary search --------------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+uprv_stableBinarySearch(char *array, int32_t limit, void *item, int32_t itemSize,
+                        UComparator *cmp, const void *context) {
+    int32_t start=0;
+    UBool found=FALSE;
+
+    /* Binary search until we get down to a tiny sub-array. */
+    while((limit-start)>=MIN_QSORT) {
+        int32_t i=(start+limit)/2;
+        int32_t diff=cmp(context, item, array+i*itemSize);
+        if(diff==0) {
+            /*
+             * Found the item. We look for the *last* occurrence of such
+             * an item, for stable sorting.
+             * If we knew that there will be only few equal items,
+             * we could break now and enter the linear search.
+             * However, if there are many equal items, then it should be
+             * faster to continue with the binary search.
+             * It seems likely that we either have all unique items
+             * (where found will never become TRUE in the insertion sort)
+             * or potentially many duplicates.
+             */
+            found=TRUE;
+            start=i+1;
+        } else if(diff<0) {
+            limit=i;
+        } else {
+            start=i;
+        }
+    }
+
+    /* Linear search over the remaining tiny sub-array. */
+    while(start<limit) {
+        int32_t diff=cmp(context, item, array+start*itemSize);
+        if(diff==0) {
+            found=TRUE;
+        } else if(diff<0) {
+            break;
+        }
+        ++start;
+    }
+    return found ? (start-1) : ~start;
+}
+
+static void
+doInsertionSort(char *array, int32_t length, int32_t itemSize,
+                UComparator *cmp, const void *context, void *pv) {
+    int32_t j;
+
+    for(j=1; j<length; ++j) {
+        char *item=array+j*itemSize;
+        int32_t insertionPoint=uprv_stableBinarySearch(array, j, item, itemSize, cmp, context);
+        if(insertionPoint<0) {
+            insertionPoint=~insertionPoint;
+        } else {
+            ++insertionPoint;  /* one past the last equal item */
+        }
+        if(insertionPoint<j) {
+            char *dest=array+insertionPoint*itemSize;
+            uprv_memcpy(pv, item, itemSize);  /* v=array[j] */
+            uprv_memmove(dest+itemSize, dest, (j-insertionPoint)*(size_t)itemSize);
+            uprv_memcpy(dest, pv, itemSize);  /* array[insertionPoint]=v */
+        }
+    }
+}
+
+static void
+insertionSort(char *array, int32_t length, int32_t itemSize,
+              UComparator *cmp, const void *context, UErrorCode *pErrorCode) {
+
+    icu::MaybeStackArray<std::max_align_t, sizeInMaxAlignTs(STACK_ITEM_SIZE)> v;
+    if (sizeInMaxAlignTs(itemSize) > v.getCapacity() &&
+            v.resize(sizeInMaxAlignTs(itemSize)) == nullptr) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    doInsertionSort(array, length, itemSize, cmp, context, v.getAlias());
+}
+
+/* QuickSort ---------------------------------------------------------------- */
+
+/*
+ * This implementation is semi-recursive:
+ * It recurses for the smaller sub-array to shorten the recursion depth,
+ * and loops for the larger sub-array.
+ *
+ * Loosely after QuickSort algorithms in
+ * Niklaus Wirth
+ * Algorithmen und Datenstrukturen mit Modula-2
+ * B.G. Teubner Stuttgart
+ * 4. Auflage 1986
+ * ISBN 3-519-02260-5
+ */
+static void
+subQuickSort(char *array, int32_t start, int32_t limit, int32_t itemSize,
+             UComparator *cmp, const void *context,
+             void *px, void *pw) {
+    int32_t left, right;
+
+    /* start and left are inclusive, limit and right are exclusive */
+    do {
+        if((start+MIN_QSORT)>=limit) {
+            doInsertionSort(array+start*itemSize, limit-start, itemSize, cmp, context, px);
+            break;
+        }
+
+        left=start;
+        right=limit;
+
+        /* x=array[middle] */
+        uprv_memcpy(px, array+(size_t)((start+limit)/2)*itemSize, itemSize);
+
+        do {
+            while(/* array[left]<x */
+                  cmp(context, array+left*itemSize, px)<0
+            ) {
+                ++left;
+            }
+            while(/* x<array[right-1] */
+                  cmp(context, px, array+(right-1)*itemSize)<0
+            ) {
+                --right;
+            }
+
+            /* swap array[left] and array[right-1] via w; ++left; --right */
+            if(left<right) {
+                --right;
+
+                if(left<right) {
+                    uprv_memcpy(pw, array+(size_t)left*itemSize, itemSize);
+                    uprv_memcpy(array+(size_t)left*itemSize, array+(size_t)right*itemSize, itemSize);
+                    uprv_memcpy(array+(size_t)right*itemSize, pw, itemSize);
+                }
+
+                ++left;
+            }
+        } while(left<right);
+
+        /* sort sub-arrays */
+        if((right-start)<(limit-left)) {
+            /* sort [start..right[ */
+            if(start<(right-1)) {
+                subQuickSort(array, start, right, itemSize, cmp, context, px, pw);
+            }
+
+            /* sort [left..limit[ */
+            start=left;
+        } else {
+            /* sort [left..limit[ */
+            if(left<(limit-1)) {
+                subQuickSort(array, left, limit, itemSize, cmp, context, px, pw);
+            }
+
+            /* sort [start..right[ */
+            limit=right;
+        }
+    } while(start<(limit-1));
+}
+
+static void
+quickSort(char *array, int32_t length, int32_t itemSize,
+            UComparator *cmp, const void *context, UErrorCode *pErrorCode) {
+    /* allocate two intermediate item variables (x and w) */
+    icu::MaybeStackArray<std::max_align_t, sizeInMaxAlignTs(STACK_ITEM_SIZE) * 2> xw;
+    if(sizeInMaxAlignTs(itemSize)*2 > xw.getCapacity() &&
+            xw.resize(sizeInMaxAlignTs(itemSize) * 2) == nullptr) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    subQuickSort(array, 0, length, itemSize, cmp, context,
+                 xw.getAlias(), xw.getAlias() + sizeInMaxAlignTs(itemSize));
+}
+
+/* uprv_sortArray() API ----------------------------------------------------- */
+
+/*
+ * Check arguments, select an appropriate implementation,
+ * cast the array to char * so that array+i*itemSize works.
+ */
+U_CAPI void U_EXPORT2
+uprv_sortArray(void *array, int32_t length, int32_t itemSize,
+               UComparator *cmp, const void *context,
+               UBool sortStable, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if((length>0 && array==NULL) || length<0 || itemSize<=0 || cmp==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if(length<=1) {
+        return;
+    } else if(length<MIN_QSORT || sortStable) {
+        insertionSort((char *)array, length, itemSize, cmp, context, pErrorCode);
+    } else {
+        quickSort((char *)array, length, itemSize, cmp, context, pErrorCode);
+    }
+}
diff --git a/thirdparty/icu4c/common/uarrsort.h b/thirdparty/icu4c/common/uarrsort.h
new file mode 100644
index 0000000000..a55dca5b9e
--- /dev/null
+++ b/thirdparty/icu4c/common/uarrsort.h
@@ -0,0 +1,103 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uarrsort.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003aug04
+*   created by: Markus W. Scherer
+*
+*   Internal function for sorting arrays.
+*/
+
+#ifndef __UARRSORT_H__
+#define __UARRSORT_H__
+
+#include "unicode/utypes.h"
+
+U_CDECL_BEGIN
+/**
+ * Function type for comparing two items as part of sorting an array or similar.
+ * Callback function for uprv_sortArray().
+ *
+ * @param context Application-specific pointer, passed through by uprv_sortArray().
+ * @param left    Pointer to the "left" item.
+ * @param right   Pointer to the "right" item.
+ * @return 32-bit signed integer comparison result:
+ *                <0 if left<right
+ *               ==0 if left==right
+ *                >0 if left>right
+ *
+ * @internal
+ */
+typedef int32_t U_CALLCONV
+UComparator(const void *context, const void *left, const void *right);
+U_CDECL_END
+
+/**
+ * Array sorting function.
+ * Uses a UComparator for comparing array items to each other, and simple
+ * memory copying to move items.
+ *
+ * @param array      The array to be sorted.
+ * @param length     The number of items in the array.
+ * @param itemSize   The size in bytes of each array item.
+ * @param cmp        UComparator function used to compare two items each.
+ * @param context    Application-specific pointer, passed through to the UComparator.
+ * @param sortStable If true, a stable sorting algorithm must be used.
+ * @param pErrorCode ICU in/out UErrorCode parameter.
+ *
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+uprv_sortArray(void *array, int32_t length, int32_t itemSize,
+               UComparator *cmp, const void *context,
+               UBool sortStable, UErrorCode *pErrorCode);
+
+/**
+ * Convenience UComparator implementation for uint16_t arrays.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_uint16Comparator(const void *context, const void *left, const void *right);
+
+/**
+ * Convenience UComparator implementation for int32_t arrays.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_int32Comparator(const void *context, const void *left, const void *right);
+
+/**
+ * Convenience UComparator implementation for uint32_t arrays.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_uint32Comparator(const void *context, const void *left, const void *right);
+
+/**
+ * Much like Java Collections.binarySearch(list, key, comparator).
+ *
+ * Except: Java documents "If the list contains multiple elements equal to
+ * the specified object, there is no guarantee which one will be found."
+ *
+ * This version here will return the largest index of any equal item,
+ * for use in stable sorting.
+ *
+ * @return the index>=0 where the item was found:
+ *         the largest such index, if multiple, for stable sorting;
+ *         or the index<0 for inserting the item at ~index in sorted order
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_stableBinarySearch(char *array, int32_t length, void *item, int32_t itemSize,
+                        UComparator *cmp, const void *context);
+
+#endif
diff --git a/thirdparty/icu4c/common/uassert.h b/thirdparty/icu4c/common/uassert.h
new file mode 100644
index 0000000000..afd31eeffd
--- /dev/null
+++ b/thirdparty/icu4c/common/uassert.h
@@ -0,0 +1,51 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2002-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File uassert.h
+*
+*  Contains the U_ASSERT and UPRV_UNREACHABLE macros
+*
+******************************************************************************
+*/
+#ifndef U_ASSERT_H
+#define U_ASSERT_H
+
+/* utypes.h is included to get the proper define for uint8_t */
+#include "unicode/utypes.h"
+/* for abort */
+#include <stdlib.h>
+
+/**
+ * \def U_ASSERT
+ * By default, U_ASSERT just wraps the C library assert macro.
+ * By changing the definition here, the assert behavior for ICU can be changed
+ * without affecting other non - ICU uses of the C library assert().
+*/
+#if U_DEBUG
+#   include <assert.h>
+#   define U_ASSERT(exp) assert(exp)
+#elif U_CPLUSPLUS_VERSION
+#   define U_ASSERT(exp) (void)0
+#else
+#   define U_ASSERT(exp)
+#endif
+
+/**
+ * \def UPRV_UNREACHABLE
+ * This macro is used to unconditionally abort if unreachable code is ever executed.
+ * @internal
+*/
+#if defined(UPRV_UNREACHABLE)
+    // Use the predefined value.
+#else
+#   define UPRV_UNREACHABLE abort()
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/ubidi.cpp b/thirdparty/icu4c/common/ubidi.cpp
new file mode 100644
index 0000000000..3ddb45721e
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidi.cpp
@@ -0,0 +1,3036 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ubidi.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999jul27
+*   created by: Markus W. Scherer, updated by Matitiahu Allouche
+*
+*/
+
+#include "cmemory.h"
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/uchar.h"
+#include "unicode/ubidi.h"
+#include "unicode/utf16.h"
+#include "ubidi_props.h"
+#include "ubidiimp.h"
+#include "uassert.h"
+
+/*
+ * General implementation notes:
+ *
+ * Throughout the implementation, there are comments like (W2) that refer to
+ * rules of the BiDi algorithm, in this example to the second rule of the
+ * resolution of weak types.
+ *
+ * For handling surrogate pairs, where two UChar's form one "abstract" (or UTF-32)
+ * character according to UTF-16, the second UChar gets the directional property of
+ * the entire character assigned, while the first one gets a BN, a boundary
+ * neutral, type, which is ignored by most of the algorithm according to
+ * rule (X9) and the implementation suggestions of the BiDi algorithm.
+ *
+ * Later, adjustWSLevels() will set the level for each BN to that of the
+ * following character (UChar), which results in surrogate pairs getting the
+ * same level on each of their surrogates.
+ *
+ * In a UTF-8 implementation, the same thing could be done: the last byte of
+ * a multi-byte sequence would get the "real" property, while all previous
+ * bytes of that sequence would get BN.
+ *
+ * It is not possible to assign all those parts of a character the same real
+ * property because this would fail in the resolution of weak types with rules
+ * that look at immediately surrounding types.
+ *
+ * As a related topic, this implementation does not remove Boundary Neutral
+ * types from the input, but ignores them wherever this is relevant.
+ * For example, the loop for the resolution of the weak types reads
+ * types until it finds a non-BN.
+ * Also, explicit embedding codes are neither changed into BN nor removed.
+ * They are only treated the same way real BNs are.
+ * As stated before, adjustWSLevels() takes care of them at the end.
+ * For the purpose of conformance, the levels of all these codes
+ * do not matter.
+ *
+ * Note that this implementation modifies the dirProps
+ * after the initial setup, when applying X5c (replace FSI by LRI or RLI),
+ * X6, N0 (replace paired brackets by L or R).
+ *
+ * In this implementation, the resolution of weak types (W1 to W6),
+ * neutrals (N1 and N2), and the assignment of the resolved level (In)
+ * are all done in one single loop, in resolveImplicitLevels().
+ * Changes of dirProp values are done on the fly, without writing
+ * them back to the dirProps array.
+ *
+ *
+ * This implementation contains code that allows to bypass steps of the
+ * algorithm that are not needed on the specific paragraph
+ * in order to speed up the most common cases considerably,
+ * like text that is entirely LTR, or RTL text without numbers.
+ *
+ * Most of this is done by setting a bit for each directional property
+ * in a flags variable and later checking for whether there are
+ * any LTR characters or any RTL characters, or both, whether
+ * there are any explicit embedding codes, etc.
+ *
+ * If the (Xn) steps are performed, then the flags are re-evaluated,
+ * because they will then not contain the embedding codes any more
+ * and will be adjusted for override codes, so that subsequently
+ * more bypassing may be possible than what the initial flags suggested.
+ *
+ * If the text is not mixed-directional, then the
+ * algorithm steps for the weak type resolution are not performed,
+ * and all levels are set to the paragraph level.
+ *
+ * If there are no explicit embedding codes, then the (Xn) steps
+ * are not performed.
+ *
+ * If embedding levels are supplied as a parameter, then all
+ * explicit embedding codes are ignored, and the (Xn) steps
+ * are not performed.
+ *
+ * White Space types could get the level of the run they belong to,
+ * and are checked with a test of (flags&MASK_EMBEDDING) to
+ * consider if the paragraph direction should be considered in
+ * the flags variable.
+ *
+ * If there are no White Space types in the paragraph, then
+ * (L1) is not necessary in adjustWSLevels().
+ */
+
+/* to avoid some conditional statements, use tiny constant arrays */
+static const Flags flagLR[2]={ DIRPROP_FLAG(L), DIRPROP_FLAG(R) };
+static const Flags flagE[2]={ DIRPROP_FLAG(LRE), DIRPROP_FLAG(RLE) };
+static const Flags flagO[2]={ DIRPROP_FLAG(LRO), DIRPROP_FLAG(RLO) };
+
+#define DIRPROP_FLAG_LR(level) flagLR[(level)&1]
+#define DIRPROP_FLAG_E(level)  flagE[(level)&1]
+#define DIRPROP_FLAG_O(level)  flagO[(level)&1]
+
+#define DIR_FROM_STRONG(strong) ((strong)==L ? L : R)
+
+#define NO_OVERRIDE(level)  ((level)&~UBIDI_LEVEL_OVERRIDE)
+
+/* UBiDi object management -------------------------------------------------- */
+
+U_CAPI UBiDi * U_EXPORT2
+ubidi_open(void)
+{
+    UErrorCode errorCode=U_ZERO_ERROR;
+    return ubidi_openSized(0, 0, &errorCode);
+}
+
+U_CAPI UBiDi * U_EXPORT2
+ubidi_openSized(int32_t maxLength, int32_t maxRunCount, UErrorCode *pErrorCode) {
+    UBiDi *pBiDi;
+
+    /* check the argument values */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    } else if(maxLength<0 || maxRunCount<0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;    /* invalid arguments */
+    }
+
+    /* allocate memory for the object */
+    pBiDi=(UBiDi *)uprv_malloc(sizeof(UBiDi));
+    if(pBiDi==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+
+    /* reset the object, all pointers NULL, all flags FALSE, all sizes 0 */
+    uprv_memset(pBiDi, 0, sizeof(UBiDi));
+
+    /* allocate memory for arrays as requested */
+    if(maxLength>0) {
+        if( !getInitialDirPropsMemory(pBiDi, maxLength) ||
+            !getInitialLevelsMemory(pBiDi, maxLength)
+        ) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        }
+    } else {
+        pBiDi->mayAllocateText=TRUE;
+    }
+
+    if(maxRunCount>0) {
+        if(maxRunCount==1) {
+            /* use simpleRuns[] */
+            pBiDi->runsSize=sizeof(Run);
+        } else if(!getInitialRunsMemory(pBiDi, maxRunCount)) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        }
+    } else {
+        pBiDi->mayAllocateRuns=TRUE;
+    }
+
+    if(U_SUCCESS(*pErrorCode)) {
+        return pBiDi;
+    } else {
+        ubidi_close(pBiDi);
+        return NULL;
+    }
+}
+
+/*
+ * We are allowed to allocate memory if memory==NULL or
+ * mayAllocate==TRUE for each array that we need.
+ * We also try to grow memory as needed if we
+ * allocate it.
+ *
+ * Assume sizeNeeded>0.
+ * If *pMemory!=NULL, then assume *pSize>0.
+ *
+ * ### this realloc() may unnecessarily copy the old data,
+ * which we know we don't need any more;
+ * is this the best way to do this??
+ */
+U_CFUNC UBool
+ubidi_getMemory(BidiMemoryForAllocation *bidiMem, int32_t *pSize, UBool mayAllocate, int32_t sizeNeeded) {
+    void **pMemory = (void **)bidiMem;
+    /* check for existing memory */
+    if(*pMemory==NULL) {
+        /* we need to allocate memory */
+        if(mayAllocate && (*pMemory=uprv_malloc(sizeNeeded))!=NULL) {
+            *pSize=sizeNeeded;
+            return TRUE;
+        } else {
+            return FALSE;
+        }
+    } else {
+        if(sizeNeeded<=*pSize) {
+            /* there is already enough memory */
+            return TRUE;
+        }
+        else if(!mayAllocate) {
+            /* not enough memory, and we must not allocate */
+            return FALSE;
+        } else {
+            /* we try to grow */
+            void *memory;
+            /* in most cases, we do not need the copy-old-data part of
+             * realloc, but it is needed when adding runs using getRunsMemory()
+             * in setParaRunsOnly()
+             */
+            if((memory=uprv_realloc(*pMemory, sizeNeeded))!=NULL) {
+                *pMemory=memory;
+                *pSize=sizeNeeded;
+                return TRUE;
+            } else {
+                /* we failed to grow */
+                return FALSE;
+            }
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_close(UBiDi *pBiDi) {
+    if(pBiDi!=NULL) {
+        pBiDi->pParaBiDi=NULL;          /* in case one tries to reuse this block */
+        if(pBiDi->dirPropsMemory!=NULL) {
+            uprv_free(pBiDi->dirPropsMemory);
+        }
+        if(pBiDi->levelsMemory!=NULL) {
+            uprv_free(pBiDi->levelsMemory);
+        }
+        if(pBiDi->openingsMemory!=NULL) {
+            uprv_free(pBiDi->openingsMemory);
+        }
+        if(pBiDi->parasMemory!=NULL) {
+            uprv_free(pBiDi->parasMemory);
+        }
+        if(pBiDi->runsMemory!=NULL) {
+            uprv_free(pBiDi->runsMemory);
+        }
+        if(pBiDi->isolatesMemory!=NULL) {
+            uprv_free(pBiDi->isolatesMemory);
+        }
+        if(pBiDi->insertPoints.points!=NULL) {
+            uprv_free(pBiDi->insertPoints.points);
+        }
+
+        uprv_free(pBiDi);
+    }
+}
+
+/* set to approximate "inverse BiDi" ---------------------------------------- */
+
+U_CAPI void U_EXPORT2
+ubidi_setInverse(UBiDi *pBiDi, UBool isInverse) {
+    if(pBiDi!=NULL) {
+        pBiDi->isInverse=isInverse;
+        pBiDi->reorderingMode = isInverse ? UBIDI_REORDER_INVERSE_NUMBERS_AS_L
+                                          : UBIDI_REORDER_DEFAULT;
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+ubidi_isInverse(UBiDi *pBiDi) {
+    if(pBiDi!=NULL) {
+        return pBiDi->isInverse;
+    } else {
+        return FALSE;
+    }
+}
+
+/* FOOD FOR THOUGHT: currently the reordering modes are a mixture of
+ * algorithm for direct BiDi, algorithm for inverse BiDi and the bizarre
+ * concept of RUNS_ONLY which is a double operation.
+ * It could be advantageous to divide this into 3 concepts:
+ * a) Operation: direct / inverse / RUNS_ONLY
+ * b) Direct algorithm: default / NUMBERS_SPECIAL / GROUP_NUMBERS_WITH_R
+ * c) Inverse algorithm: default / INVERSE_LIKE_DIRECT / NUMBERS_SPECIAL
+ * This would allow combinations not possible today like RUNS_ONLY with
+ * NUMBERS_SPECIAL.
+ * Also allow to set INSERT_MARKS for the direct step of RUNS_ONLY and
+ * REMOVE_CONTROLS for the inverse step.
+ * Not all combinations would be supported, and probably not all do make sense.
+ * This would need to document which ones are supported and what are the
+ * fallbacks for unsupported combinations.
+ */
+U_CAPI void U_EXPORT2
+ubidi_setReorderingMode(UBiDi *pBiDi, UBiDiReorderingMode reorderingMode) {
+    if ((pBiDi!=NULL) && (reorderingMode >= UBIDI_REORDER_DEFAULT)
+                        && (reorderingMode < UBIDI_REORDER_COUNT)) {
+        pBiDi->reorderingMode = reorderingMode;
+        pBiDi->isInverse = (UBool)(reorderingMode == UBIDI_REORDER_INVERSE_NUMBERS_AS_L);
+    }
+}
+
+U_CAPI UBiDiReorderingMode U_EXPORT2
+ubidi_getReorderingMode(UBiDi *pBiDi) {
+    if (pBiDi!=NULL) {
+        return pBiDi->reorderingMode;
+    } else {
+        return UBIDI_REORDER_DEFAULT;
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_setReorderingOptions(UBiDi *pBiDi, uint32_t reorderingOptions) {
+    if (reorderingOptions & UBIDI_OPTION_REMOVE_CONTROLS) {
+        reorderingOptions&=~UBIDI_OPTION_INSERT_MARKS;
+    }
+    if (pBiDi!=NULL) {
+        pBiDi->reorderingOptions=reorderingOptions;
+    }
+}
+
+U_CAPI uint32_t U_EXPORT2
+ubidi_getReorderingOptions(UBiDi *pBiDi) {
+    if (pBiDi!=NULL) {
+        return pBiDi->reorderingOptions;
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getBaseDirection(const UChar *text,
+int32_t length){
+
+    int32_t i;
+    UChar32 uchar;
+    UCharDirection dir;
+
+    if( text==NULL || length<-1 ){
+        return UBIDI_NEUTRAL;
+    }
+
+    if(length==-1) {
+        length=u_strlen(text);
+    }
+
+    for( i = 0 ; i < length; ) {
+        /* i is incremented by U16_NEXT */
+        U16_NEXT(text, i, length, uchar);
+        dir = u_charDirection(uchar);
+        if( dir == U_LEFT_TO_RIGHT )
+                return UBIDI_LTR;
+        if( dir == U_RIGHT_TO_LEFT || dir ==U_RIGHT_TO_LEFT_ARABIC )
+                return UBIDI_RTL;
+    }
+    return UBIDI_NEUTRAL;
+}
+
+/* perform (P2)..(P3) ------------------------------------------------------- */
+
+/**
+ * Returns the directionality of the first strong character
+ * after the last B in prologue, if any.
+ * Requires prologue!=null.
+ */
+static DirProp
+firstL_R_AL(UBiDi *pBiDi) {
+    const UChar *text=pBiDi->prologue;
+    int32_t length=pBiDi->proLength;
+    int32_t i;
+    UChar32 uchar;
+    DirProp dirProp, result=ON;
+    for(i=0; i<length; ) {
+        /* i is incremented by U16_NEXT */
+        U16_NEXT(text, i, length, uchar);
+        dirProp=(DirProp)ubidi_getCustomizedClass(pBiDi, uchar);
+        if(result==ON) {
+            if(dirProp==L || dirProp==R || dirProp==AL) {
+                result=dirProp;
+            }
+        } else {
+            if(dirProp==B) {
+                result=ON;
+            }
+        }
+    }
+    return result;
+}
+
+/*
+ * Check that there are enough entries in the array pointed to by pBiDi->paras
+ */
+static UBool
+checkParaCount(UBiDi *pBiDi) {
+    int32_t count=pBiDi->paraCount;
+    if(pBiDi->paras==pBiDi->simpleParas) {
+        if(count<=SIMPLE_PARAS_COUNT)
+            return TRUE;
+        if(!getInitialParasMemory(pBiDi, SIMPLE_PARAS_COUNT * 2))
+            return FALSE;
+        pBiDi->paras=pBiDi->parasMemory;
+        uprv_memcpy(pBiDi->parasMemory, pBiDi->simpleParas, SIMPLE_PARAS_COUNT * sizeof(Para));
+        return TRUE;
+    }
+    if(!getInitialParasMemory(pBiDi, count * 2))
+        return FALSE;
+    pBiDi->paras=pBiDi->parasMemory;
+    return TRUE;
+}
+
+/*
+ * Get the directional properties for the text, calculate the flags bit-set, and
+ * determine the paragraph level if necessary (in pBiDi->paras[i].level).
+ * FSI initiators are also resolved and their dirProp replaced with LRI or RLI.
+ * When encountering an FSI, it is initially replaced with an LRI, which is the
+ * default. Only if a strong R or AL is found within its scope will the LRI be
+ * replaced by an RLI.
+ */
+static UBool
+getDirProps(UBiDi *pBiDi) {
+    const UChar *text=pBiDi->text;
+    DirProp *dirProps=pBiDi->dirPropsMemory;    /* pBiDi->dirProps is const */
+
+    int32_t i=0, originalLength=pBiDi->originalLength;
+    Flags flags=0;      /* collect all directionalities in the text */
+    UChar32 uchar;
+    DirProp dirProp=0, defaultParaLevel=0;  /* initialize to avoid compiler warnings */
+    UBool isDefaultLevel=IS_DEFAULT_LEVEL(pBiDi->paraLevel);
+    /* for inverse BiDi, the default para level is set to RTL if there is a
+       strong R or AL character at either end of the text                            */
+    UBool isDefaultLevelInverse=isDefaultLevel && (UBool)
+            (pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_LIKE_DIRECT ||
+             pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL);
+    int32_t lastArabicPos=-1;
+    int32_t controlCount=0;
+    UBool removeBiDiControls = (UBool)(pBiDi->reorderingOptions &
+                                       UBIDI_OPTION_REMOVE_CONTROLS);
+
+    enum State {
+         NOT_SEEKING_STRONG,            /* 0: not contextual paraLevel, not after FSI */
+         SEEKING_STRONG_FOR_PARA,       /* 1: looking for first strong char in para */
+         SEEKING_STRONG_FOR_FSI,        /* 2: looking for first strong after FSI */
+         LOOKING_FOR_PDI                /* 3: found strong after FSI, looking for PDI */
+    };
+    State state;
+    DirProp lastStrong=ON;              /* for default level & inverse BiDi */
+    /* The following stacks are used to manage isolate sequences. Those
+       sequences may be nested, but obviously never more deeply than the
+       maximum explicit embedding level.
+       lastStack is the index of the last used entry in the stack. A value of -1
+       means that there is no open isolate sequence.
+       lastStack is reset to -1 on paragraph boundaries. */
+    /* The following stack contains the position of the initiator of
+       each open isolate sequence */
+    int32_t isolateStartStack[UBIDI_MAX_EXPLICIT_LEVEL+1];
+    /* The following stack contains the last known state before
+       encountering the initiator of an isolate sequence */
+    State  previousStateStack[UBIDI_MAX_EXPLICIT_LEVEL+1];
+    int32_t stackLast=-1;
+
+    if(pBiDi->reorderingOptions & UBIDI_OPTION_STREAMING)
+        pBiDi->length=0;
+    defaultParaLevel=pBiDi->paraLevel&1;
+    if(isDefaultLevel) {
+        pBiDi->paras[0].level=defaultParaLevel;
+        lastStrong=defaultParaLevel;
+        if(pBiDi->proLength>0 &&                    /* there is a prologue */
+           (dirProp=firstL_R_AL(pBiDi))!=ON) {  /* with a strong character */
+            if(dirProp==L)
+                pBiDi->paras[0].level=0;    /* set the default para level */
+            else
+                pBiDi->paras[0].level=1;    /* set the default para level */
+            state=NOT_SEEKING_STRONG;
+        } else {
+            state=SEEKING_STRONG_FOR_PARA;
+        }
+    } else {
+        pBiDi->paras[0].level=pBiDi->paraLevel;
+        state=NOT_SEEKING_STRONG;
+    }
+    /* count paragraphs and determine the paragraph level (P2..P3) */
+    /*
+     * see comment in ubidi.h:
+     * the UBIDI_DEFAULT_XXX values are designed so that
+     * their bit 0 alone yields the intended default
+     */
+    for( /* i=0 above */ ; i<originalLength; ) {
+        /* i is incremented by U16_NEXT */
+        U16_NEXT(text, i, originalLength, uchar);
+        flags|=DIRPROP_FLAG(dirProp=(DirProp)ubidi_getCustomizedClass(pBiDi, uchar));
+        dirProps[i-1]=dirProp;
+        if(uchar>0xffff) {  /* set the lead surrogate's property to BN */
+            flags|=DIRPROP_FLAG(BN);
+            dirProps[i-2]=BN;
+        }
+        if(removeBiDiControls && IS_BIDI_CONTROL_CHAR(uchar))
+            controlCount++;
+        if(dirProp==L) {
+            if(state==SEEKING_STRONG_FOR_PARA) {
+                pBiDi->paras[pBiDi->paraCount-1].level=0;
+                state=NOT_SEEKING_STRONG;
+            }
+            else if(state==SEEKING_STRONG_FOR_FSI) {
+                if(stackLast<=UBIDI_MAX_EXPLICIT_LEVEL) {
+                    /* no need for next statement, already set by default */
+                    /* dirProps[isolateStartStack[stackLast]]=LRI; */
+                    flags|=DIRPROP_FLAG(LRI);
+                }
+                state=LOOKING_FOR_PDI;
+            }
+            lastStrong=L;
+            continue;
+        }
+        if(dirProp==R || dirProp==AL) {
+            if(state==SEEKING_STRONG_FOR_PARA) {
+                pBiDi->paras[pBiDi->paraCount-1].level=1;
+                state=NOT_SEEKING_STRONG;
+            }
+            else if(state==SEEKING_STRONG_FOR_FSI) {
+                if(stackLast<=UBIDI_MAX_EXPLICIT_LEVEL) {
+                    dirProps[isolateStartStack[stackLast]]=RLI;
+                    flags|=DIRPROP_FLAG(RLI);
+                }
+                state=LOOKING_FOR_PDI;
+            }
+            lastStrong=R;
+            if(dirProp==AL)
+                lastArabicPos=i-1;
+            continue;
+        }
+        if(dirProp>=FSI && dirProp<=RLI) {  /* FSI, LRI or RLI */
+            stackLast++;
+            if(stackLast<=UBIDI_MAX_EXPLICIT_LEVEL) {
+                isolateStartStack[stackLast]=i-1;
+                previousStateStack[stackLast]=state;
+            }
+            if(dirProp==FSI) {
+                dirProps[i-1]=LRI;      /* default if no strong char */
+                state=SEEKING_STRONG_FOR_FSI;
+            }
+            else
+                state=LOOKING_FOR_PDI;
+            continue;
+        }
+        if(dirProp==PDI) {
+            if(state==SEEKING_STRONG_FOR_FSI) {
+                if(stackLast<=UBIDI_MAX_EXPLICIT_LEVEL) {
+                    /* no need for next statement, already set by default */
+                    /* dirProps[isolateStartStack[stackLast]]=LRI; */
+                    flags|=DIRPROP_FLAG(LRI);
+                }
+            }
+            if(stackLast>=0) {
+                if(stackLast<=UBIDI_MAX_EXPLICIT_LEVEL)
+                    state=previousStateStack[stackLast];
+                stackLast--;
+            }
+            continue;
+        }
+        if(dirProp==B) {
+            if(i<originalLength && uchar==CR && text[i]==LF) /* do nothing on the CR */
+                continue;
+            pBiDi->paras[pBiDi->paraCount-1].limit=i;
+            if(isDefaultLevelInverse && lastStrong==R)
+                pBiDi->paras[pBiDi->paraCount-1].level=1;
+            if(pBiDi->reorderingOptions & UBIDI_OPTION_STREAMING) {
+                /* When streaming, we only process whole paragraphs
+                   thus some updates are only done on paragraph boundaries */
+                pBiDi->length=i;        /* i is index to next character */
+                pBiDi->controlCount=controlCount;
+            }
+            if(i<originalLength) {              /* B not last char in text */
+                pBiDi->paraCount++;
+                if(checkParaCount(pBiDi)==FALSE)    /* not enough memory for a new para entry */
+                    return FALSE;
+                if(isDefaultLevel) {
+                    pBiDi->paras[pBiDi->paraCount-1].level=defaultParaLevel;
+                    state=SEEKING_STRONG_FOR_PARA;
+                    lastStrong=defaultParaLevel;
+                } else {
+                    pBiDi->paras[pBiDi->paraCount-1].level=pBiDi->paraLevel;
+                    state=NOT_SEEKING_STRONG;
+                }
+                stackLast=-1;
+            }
+            continue;
+        }
+    }
+    /* Ignore still open isolate sequences with overflow */
+    if(stackLast>UBIDI_MAX_EXPLICIT_LEVEL) {
+        stackLast=UBIDI_MAX_EXPLICIT_LEVEL;
+        state=SEEKING_STRONG_FOR_FSI;   /* to be on the safe side */
+    }
+    /* Resolve direction of still unresolved open FSI sequences */
+    while(stackLast>=0) {
+        if(state==SEEKING_STRONG_FOR_FSI) {
+            /* no need for next statement, already set by default */
+            /* dirProps[isolateStartStack[stackLast]]=LRI; */
+            flags|=DIRPROP_FLAG(LRI);
+            break;
+        }
+        state=previousStateStack[stackLast];
+        stackLast--;
+    }
+    /* When streaming, ignore text after the last paragraph separator */
+    if(pBiDi->reorderingOptions & UBIDI_OPTION_STREAMING) {
+        if(pBiDi->length<originalLength)
+            pBiDi->paraCount--;
+    } else {
+        pBiDi->paras[pBiDi->paraCount-1].limit=originalLength;
+        pBiDi->controlCount=controlCount;
+    }
+    /* For inverse bidi, default para direction is RTL if there is
+       a strong R or AL at either end of the paragraph */
+    if(isDefaultLevelInverse && lastStrong==R) {
+        pBiDi->paras[pBiDi->paraCount-1].level=1;
+    }
+    if(isDefaultLevel) {
+        pBiDi->paraLevel=static_cast<UBiDiLevel>(pBiDi->paras[0].level);
+    }
+    /* The following is needed to resolve the text direction for default level
+       paragraphs containing no strong character */
+    for(i=0; i<pBiDi->paraCount; i++)
+        flags|=DIRPROP_FLAG_LR(pBiDi->paras[i].level);
+
+    if(pBiDi->orderParagraphsLTR && (flags&DIRPROP_FLAG(B))) {
+        flags|=DIRPROP_FLAG(L);
+    }
+    pBiDi->flags=flags;
+    pBiDi->lastArabicPos=lastArabicPos;
+    return TRUE;
+}
+
+/* determine the paragraph level at position index */
+U_CFUNC UBiDiLevel
+ubidi_getParaLevelAtIndex(const UBiDi *pBiDi, int32_t pindex) {
+    int32_t i;
+    for(i=0; i<pBiDi->paraCount; i++)
+        if(pindex<pBiDi->paras[i].limit)
+            break;
+    if(i>=pBiDi->paraCount)
+        i=pBiDi->paraCount-1;
+    return (UBiDiLevel)(pBiDi->paras[i].level);
+}
+
+/* Functions for handling paired brackets ----------------------------------- */
+
+/* In the isoRuns array, the first entry is used for text outside of any
+   isolate sequence.  Higher entries are used for each more deeply nested
+   isolate sequence. isoRunLast is the index of the last used entry.  The
+   openings array is used to note the data of opening brackets not yet
+   matched by a closing bracket, or matched but still susceptible to change
+   level.
+   Each isoRun entry contains the index of the first and
+   one-after-last openings entries for pending opening brackets it
+   contains.  The next openings entry to use is the one-after-last of the
+   most deeply nested isoRun entry.
+   isoRun entries also contain their current embedding level and the last
+   encountered strong character, since these will be needed to resolve
+   the level of paired brackets.  */
+
+static void
+bracketInit(UBiDi *pBiDi, BracketData *bd) {
+    bd->pBiDi=pBiDi;
+    bd->isoRunLast=0;
+    bd->isoRuns[0].start=0;
+    bd->isoRuns[0].limit=0;
+    bd->isoRuns[0].level=GET_PARALEVEL(pBiDi, 0);
+    UBiDiLevel t = GET_PARALEVEL(pBiDi, 0) & 1;
+    bd->isoRuns[0].lastStrong = bd->isoRuns[0].lastBase = t;
+    bd->isoRuns[0].contextDir = (UBiDiDirection)t;
+    bd->isoRuns[0].contextPos=0;
+    if(pBiDi->openingsMemory) {
+        bd->openings=pBiDi->openingsMemory;
+        bd->openingsCount=pBiDi->openingsSize / sizeof(Opening);
+    } else {
+        bd->openings=bd->simpleOpenings;
+        bd->openingsCount=SIMPLE_OPENINGS_COUNT;
+    }
+    bd->isNumbersSpecial=bd->pBiDi->reorderingMode==UBIDI_REORDER_NUMBERS_SPECIAL ||
+                         bd->pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL;
+}
+
+/* paragraph boundary */
+static void
+bracketProcessB(BracketData *bd, UBiDiLevel level) {
+    bd->isoRunLast=0;
+    bd->isoRuns[0].limit=0;
+    bd->isoRuns[0].level=level;
+    bd->isoRuns[0].lastStrong=bd->isoRuns[0].lastBase=level&1;
+    bd->isoRuns[0].contextDir=(UBiDiDirection)(level&1);
+    bd->isoRuns[0].contextPos=0;
+}
+
+/* LRE, LRO, RLE, RLO, PDF */
+static void
+bracketProcessBoundary(BracketData *bd, int32_t lastCcPos,
+                       UBiDiLevel contextLevel, UBiDiLevel embeddingLevel) {
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    DirProp *dirProps=bd->pBiDi->dirProps;
+    if(DIRPROP_FLAG(dirProps[lastCcPos])&MASK_ISO)  /* after an isolate */
+        return;
+    if(NO_OVERRIDE(embeddingLevel)>NO_OVERRIDE(contextLevel))   /* not a PDF */
+        contextLevel=embeddingLevel;
+    pLastIsoRun->limit=pLastIsoRun->start;
+    pLastIsoRun->level=embeddingLevel;
+    pLastIsoRun->lastStrong=pLastIsoRun->lastBase=contextLevel&1;
+    pLastIsoRun->contextDir=(UBiDiDirection)(contextLevel&1);
+    pLastIsoRun->contextPos=(UBiDiDirection)lastCcPos;
+}
+
+/* LRI or RLI */
+static void
+bracketProcessLRI_RLI(BracketData *bd, UBiDiLevel level) {
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    int16_t lastLimit;
+    pLastIsoRun->lastBase=ON;
+    lastLimit=pLastIsoRun->limit;
+    bd->isoRunLast++;
+    pLastIsoRun++;
+    pLastIsoRun->start=pLastIsoRun->limit=lastLimit;
+    pLastIsoRun->level=level;
+    pLastIsoRun->lastStrong=pLastIsoRun->lastBase=level&1;
+    pLastIsoRun->contextDir=(UBiDiDirection)(level&1);
+    pLastIsoRun->contextPos=0;
+}
+
+/* PDI */
+static void
+bracketProcessPDI(BracketData *bd) {
+    IsoRun *pLastIsoRun;
+    bd->isoRunLast--;
+    pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    pLastIsoRun->lastBase=ON;
+}
+
+/* newly found opening bracket: create an openings entry */
+static UBool                            /* return TRUE if success */
+bracketAddOpening(BracketData *bd, UChar match, int32_t position) {
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    Opening *pOpening;
+    if(pLastIsoRun->limit>=bd->openingsCount) {  /* no available new entry */
+        UBiDi *pBiDi=bd->pBiDi;
+        if(!getInitialOpeningsMemory(pBiDi, pLastIsoRun->limit * 2))
+            return FALSE;
+        if(bd->openings==bd->simpleOpenings)
+            uprv_memcpy(pBiDi->openingsMemory, bd->simpleOpenings,
+                        SIMPLE_OPENINGS_COUNT * sizeof(Opening));
+        bd->openings=pBiDi->openingsMemory;     /* may have changed */
+        bd->openingsCount=pBiDi->openingsSize / sizeof(Opening);
+    }
+    pOpening=&bd->openings[pLastIsoRun->limit];
+    pOpening->position=position;
+    pOpening->match=match;
+    pOpening->contextDir=pLastIsoRun->contextDir;
+    pOpening->contextPos=pLastIsoRun->contextPos;
+    pOpening->flags=0;
+    pLastIsoRun->limit++;
+    return TRUE;
+}
+
+/* change N0c1 to N0c2 when a preceding bracket is assigned the embedding level */
+static void
+fixN0c(BracketData *bd, int32_t openingIndex, int32_t newPropPosition, DirProp newProp) {
+    /* This function calls itself recursively */
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    Opening *qOpening;
+    DirProp *dirProps=bd->pBiDi->dirProps;
+    int32_t k, openingPosition, closingPosition;
+    for(k=openingIndex+1, qOpening=&bd->openings[k]; k<pLastIsoRun->limit; k++, qOpening++) {
+        if(qOpening->match>=0)      /* not an N0c match */
+            continue;
+        if(newPropPosition<qOpening->contextPos)
+            break;
+        if(newPropPosition>=qOpening->position)
+            continue;
+        if(newProp==qOpening->contextDir)
+            break;
+        openingPosition=qOpening->position;
+        dirProps[openingPosition]=newProp;
+        closingPosition=-(qOpening->match);
+        dirProps[closingPosition]=newProp;
+        qOpening->match=0;                      /* prevent further changes */
+        fixN0c(bd, k, openingPosition, newProp);
+        fixN0c(bd, k, closingPosition, newProp);
+    }
+}
+
+/* process closing bracket */
+static DirProp              /* return L or R if N0b or N0c, ON if N0d */
+bracketProcessClosing(BracketData *bd, int32_t openIdx, int32_t position) {
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    Opening *pOpening, *qOpening;
+    UBiDiDirection direction;
+    UBool stable;
+    DirProp newProp;
+    pOpening=&bd->openings[openIdx];
+    direction=(UBiDiDirection)(pLastIsoRun->level&1);
+    stable=TRUE;            /* assume stable until proved otherwise */
+
+    /* The stable flag is set when brackets are paired and their
+       level is resolved and cannot be changed by what will be
+       found later in the source string.
+       An unstable match can occur only when applying N0c, where
+       the resolved level depends on the preceding context, and
+       this context may be affected by text occurring later.
+       Example: RTL paragraph containing:  abc[(latin) HEBREW]
+       When the closing parenthesis is encountered, it appears
+       that N0c1 must be applied since 'abc' sets an opposite
+       direction context and both parentheses receive level 2.
+       However, when the closing square bracket is processed,
+       N0b applies because of 'HEBREW' being included within the
+       brackets, thus the square brackets are treated like R and
+       receive level 1. However, this changes the preceding
+       context of the opening parenthesis, and it now appears
+       that N0c2 must be applied to the parentheses rather than
+       N0c1. */
+
+    if((direction==0 && pOpening->flags&FOUND_L) ||
+       (direction==1 && pOpening->flags&FOUND_R)) {                         /* N0b */
+        newProp=static_cast<DirProp>(direction);
+    }
+    else if(pOpening->flags&(FOUND_L|FOUND_R)) {                            /* N0c */
+        /* it is stable if there is no containing pair or in
+           conditions too complicated and not worth checking */
+        stable=(openIdx==pLastIsoRun->start);
+        if(direction!=pOpening->contextDir)
+            newProp= static_cast<DirProp>(pOpening->contextDir);           /* N0c1 */
+        else
+            newProp= static_cast<DirProp>(direction);                      /* N0c2 */
+    } else {
+        /* forget this and any brackets nested within this pair */
+        pLastIsoRun->limit= static_cast<uint16_t>(openIdx);
+        return ON;                                                          /* N0d */
+    }
+    bd->pBiDi->dirProps[pOpening->position]=newProp;
+    bd->pBiDi->dirProps[position]=newProp;
+    /* Update nested N0c pairs that may be affected */
+    fixN0c(bd, openIdx, pOpening->position, newProp);
+    if(stable) {
+        pLastIsoRun->limit= static_cast<uint16_t>(openIdx); /* forget any brackets nested within this pair */
+        /* remove lower located synonyms if any */
+        while(pLastIsoRun->limit>pLastIsoRun->start &&
+              bd->openings[pLastIsoRun->limit-1].position==pOpening->position)
+            pLastIsoRun->limit--;
+    } else {
+        int32_t k;
+        pOpening->match=-position;
+        /* neutralize lower located synonyms if any */
+        k=openIdx-1;
+        while(k>=pLastIsoRun->start &&
+              bd->openings[k].position==pOpening->position)
+            bd->openings[k--].match=0;
+        /* neutralize any unmatched opening between the current pair;
+           this will also neutralize higher located synonyms if any */
+        for(k=openIdx+1; k<pLastIsoRun->limit; k++) {
+            qOpening=&bd->openings[k];
+            if(qOpening->position>=position)
+                break;
+            if(qOpening->match>0)
+                qOpening->match=0;
+        }
+    }
+    return newProp;
+}
+
+/* handle strong characters, digits and candidates for closing brackets */
+static UBool                            /* return TRUE if success */
+bracketProcessChar(BracketData *bd, int32_t position) {
+    IsoRun *pLastIsoRun=&bd->isoRuns[bd->isoRunLast];
+    DirProp *dirProps, dirProp, newProp;
+    UBiDiLevel level;
+    dirProps=bd->pBiDi->dirProps;
+    dirProp=dirProps[position];
+    if(dirProp==ON) {
+        UChar c, match;
+        int32_t idx;
+        /* First see if it is a matching closing bracket. Hopefully, this is
+           more efficient than checking if it is a closing bracket at all */
+        c=bd->pBiDi->text[position];
+        for(idx=pLastIsoRun->limit-1; idx>=pLastIsoRun->start; idx--) {
+            if(bd->openings[idx].match!=c)
+                continue;
+            /* We have a match */
+            newProp=bracketProcessClosing(bd, idx, position);
+            if(newProp==ON) {           /* N0d */
+                c=0;        /* prevent handling as an opening */
+                break;
+            }
+            pLastIsoRun->lastBase=ON;
+            pLastIsoRun->contextDir=(UBiDiDirection)newProp;
+            pLastIsoRun->contextPos=position;
+            level=bd->pBiDi->levels[position];
+            if(level&UBIDI_LEVEL_OVERRIDE) {    /* X4, X5 */
+                uint16_t flag;
+                int32_t i;
+                newProp=level&1;
+                pLastIsoRun->lastStrong=newProp;
+                flag=DIRPROP_FLAG(newProp);
+                for(i=pLastIsoRun->start; i<idx; i++)
+                    bd->openings[i].flags|=flag;
+                /* matching brackets are not overridden by LRO/RLO */
+                bd->pBiDi->levels[position]&=~UBIDI_LEVEL_OVERRIDE;
+            }
+            /* matching brackets are not overridden by LRO/RLO */
+            bd->pBiDi->levels[bd->openings[idx].position]&=~UBIDI_LEVEL_OVERRIDE;
+            return TRUE;
+        }
+        /* We get here only if the ON character is not a matching closing
+           bracket or it is a case of N0d */
+        /* Now see if it is an opening bracket */
+        if(c)
+            match= static_cast<UChar>(u_getBidiPairedBracket(c));    /* get the matching char */
+        else
+            match=0;
+        if(match!=c &&                  /* has a matching char */
+           ubidi_getPairedBracketType(c)==U_BPT_OPEN) { /* opening bracket */
+            /* special case: process synonyms
+               create an opening entry for each synonym */
+            if(match==0x232A) {     /* RIGHT-POINTING ANGLE BRACKET */
+                if(!bracketAddOpening(bd, 0x3009, position))
+                    return FALSE;
+            }
+            else if(match==0x3009) {         /* RIGHT ANGLE BRACKET */
+                if(!bracketAddOpening(bd, 0x232A, position))
+                    return FALSE;
+            }
+            if(!bracketAddOpening(bd, match, position))
+                return FALSE;
+        }
+    }
+    level=bd->pBiDi->levels[position];
+    if(level&UBIDI_LEVEL_OVERRIDE) {    /* X4, X5 */
+        newProp=level&1;
+        if(dirProp!=S && dirProp!=WS && dirProp!=ON)
+            dirProps[position]=newProp;
+        pLastIsoRun->lastBase=newProp;
+        pLastIsoRun->lastStrong=newProp;
+        pLastIsoRun->contextDir=(UBiDiDirection)newProp;
+        pLastIsoRun->contextPos=position;
+    }
+    else if(dirProp<=R || dirProp==AL) {
+        newProp= static_cast<DirProp>(DIR_FROM_STRONG(dirProp));
+        pLastIsoRun->lastBase=dirProp;
+        pLastIsoRun->lastStrong=dirProp;
+        pLastIsoRun->contextDir=(UBiDiDirection)newProp;
+        pLastIsoRun->contextPos=position;
+    }
+    else if(dirProp==EN) {
+        pLastIsoRun->lastBase=EN;
+        if(pLastIsoRun->lastStrong==L) {
+            newProp=L;                  /* W7 */
+            if(!bd->isNumbersSpecial)
+                dirProps[position]=ENL;
+            pLastIsoRun->contextDir=(UBiDiDirection)L;
+            pLastIsoRun->contextPos=position;
+        }
+        else {
+            newProp=R;                  /* N0 */
+            if(pLastIsoRun->lastStrong==AL)
+                dirProps[position]=AN;  /* W2 */
+            else
+                dirProps[position]=ENR;
+            pLastIsoRun->contextDir=(UBiDiDirection)R;
+            pLastIsoRun->contextPos=position;
+        }
+    }
+    else if(dirProp==AN) {
+        newProp=R;                      /* N0 */
+        pLastIsoRun->lastBase=AN;
+        pLastIsoRun->contextDir=(UBiDiDirection)R;
+        pLastIsoRun->contextPos=position;
+    }
+    else if(dirProp==NSM) {
+        /* if the last real char was ON, change NSM to ON so that it
+           will stay ON even if the last real char is a bracket which
+           may be changed to L or R */
+        newProp=pLastIsoRun->lastBase;
+        if(newProp==ON)
+            dirProps[position]=newProp;
+    }
+    else {
+        newProp=dirProp;
+        pLastIsoRun->lastBase=dirProp;
+    }
+    if(newProp<=R || newProp==AL) {
+        int32_t i;
+        uint16_t flag=DIRPROP_FLAG(DIR_FROM_STRONG(newProp));
+        for(i=pLastIsoRun->start; i<pLastIsoRun->limit; i++)
+            if(position>bd->openings[i].position)
+                bd->openings[i].flags|=flag;
+    }
+    return TRUE;
+}
+
+/* perform (X1)..(X9) ------------------------------------------------------- */
+
+/* determine if the text is mixed-directional or single-directional */
+static UBiDiDirection
+directionFromFlags(UBiDi *pBiDi) {
+    Flags flags=pBiDi->flags;
+    /* if the text contains AN and neutrals, then some neutrals may become RTL */
+    if(!(flags&MASK_RTL || ((flags&DIRPROP_FLAG(AN)) && (flags&MASK_POSSIBLE_N)))) {
+        return UBIDI_LTR;
+    } else if(!(flags&MASK_LTR)) {
+        return UBIDI_RTL;
+    } else {
+        return UBIDI_MIXED;
+    }
+}
+
+/*
+ * Resolve the explicit levels as specified by explicit embedding codes.
+ * Recalculate the flags to have them reflect the real properties
+ * after taking the explicit embeddings into account.
+ *
+ * The BiDi algorithm is designed to result in the same behavior whether embedding
+ * levels are externally specified (from "styled text", supposedly the preferred
+ * method) or set by explicit embedding codes (LRx, RLx, PDF, FSI, PDI) in the plain text.
+ * That is why (X9) instructs to remove all not-isolate explicit codes (and BN).
+ * However, in a real implementation, the removal of these codes and their index
+ * positions in the plain text is undesirable since it would result in
+ * reallocated, reindexed text.
+ * Instead, this implementation leaves the codes in there and just ignores them
+ * in the subsequent processing.
+ * In order to get the same reordering behavior, positions with a BN or a not-isolate
+ * explicit embedding code just get the same level assigned as the last "real"
+ * character.
+ *
+ * Some implementations, not this one, then overwrite some of these
+ * directionality properties at "real" same-level-run boundaries by
+ * L or R codes so that the resolution of weak types can be performed on the
+ * entire paragraph at once instead of having to parse it once more and
+ * perform that resolution on same-level-runs.
+ * This limits the scope of the implicit rules in effectively
+ * the same way as the run limits.
+ *
+ * Instead, this implementation does not modify these codes, except for
+ * paired brackets whose properties (ON) may be replaced by L or R.
+ * On one hand, the paragraph has to be scanned for same-level-runs, but
+ * on the other hand, this saves another loop to reset these codes,
+ * or saves making and modifying a copy of dirProps[].
+ *
+ *
+ * Note that (Pn) and (Xn) changed significantly from version 4 of the BiDi algorithm.
+ *
+ *
+ * Handling the stack of explicit levels (Xn):
+ *
+ * With the BiDi stack of explicit levels, as pushed with each
+ * LRE, RLE, LRO, RLO, LRI, RLI and FSI and popped with each PDF and PDI,
+ * the explicit level must never exceed UBIDI_MAX_EXPLICIT_LEVEL.
+ *
+ * In order to have a correct push-pop semantics even in the case of overflows,
+ * overflow counters and a valid isolate counter are used as described in UAX#9
+ * section 3.3.2 "Explicit Levels and Directions".
+ *
+ * This implementation assumes that UBIDI_MAX_EXPLICIT_LEVEL is odd.
+ *
+ * Returns normally the direction; -1 if there was a memory shortage
+ *
+ */
+static UBiDiDirection
+resolveExplicitLevels(UBiDi *pBiDi, UErrorCode *pErrorCode) {
+    DirProp *dirProps=pBiDi->dirProps;
+    UBiDiLevel *levels=pBiDi->levels;
+    const UChar *text=pBiDi->text;
+
+    int32_t i=0, length=pBiDi->length;
+    Flags flags=pBiDi->flags;       /* collect all directionalities in the text */
+    DirProp dirProp;
+    UBiDiLevel level=GET_PARALEVEL(pBiDi, 0);
+    UBiDiDirection direction;
+    pBiDi->isolateCount=0;
+
+    if(U_FAILURE(*pErrorCode)) { return UBIDI_LTR; }
+
+    /* determine if the text is mixed-directional or single-directional */
+    direction=directionFromFlags(pBiDi);
+
+    /* we may not need to resolve any explicit levels */
+    if((direction!=UBIDI_MIXED)) {
+        /* not mixed directionality: levels don't matter - trailingWSStart will be 0 */
+        return direction;
+    }
+    if(pBiDi->reorderingMode > UBIDI_REORDER_LAST_LOGICAL_TO_VISUAL) {
+        /* inverse BiDi: mixed, but all characters are at the same embedding level */
+        /* set all levels to the paragraph level */
+        int32_t paraIndex, start, limit;
+        for(paraIndex=0; paraIndex<pBiDi->paraCount; paraIndex++) {
+            if(paraIndex==0)
+                start=0;
+            else
+                start=pBiDi->paras[paraIndex-1].limit;
+            limit=pBiDi->paras[paraIndex].limit;
+            level= static_cast<UBiDiLevel>(pBiDi->paras[paraIndex].level);
+            for(i=start; i<limit; i++)
+                levels[i]=level;
+        }
+        return direction;   /* no bracket matching for inverse BiDi */
+    }
+    if(!(flags&(MASK_EXPLICIT|MASK_ISO))) {
+        /* no embeddings, set all levels to the paragraph level */
+        /* we still have to perform bracket matching */
+        int32_t paraIndex, start, limit;
+        BracketData bracketData;
+        bracketInit(pBiDi, &bracketData);
+        for(paraIndex=0; paraIndex<pBiDi->paraCount; paraIndex++) {
+            if(paraIndex==0)
+                start=0;
+            else
+                start=pBiDi->paras[paraIndex-1].limit;
+            limit=pBiDi->paras[paraIndex].limit;
+            level= static_cast<UBiDiLevel>(pBiDi->paras[paraIndex].level);
+            for(i=start; i<limit; i++) {
+                levels[i]=level;
+                dirProp=dirProps[i];
+                if(dirProp==BN)
+                    continue;
+                if(dirProp==B) {
+                    if((i+1)<length) {
+                        if(text[i]==CR && text[i+1]==LF)
+                            continue;   /* skip CR when followed by LF */
+                        bracketProcessB(&bracketData, level);
+                    }
+                    continue;
+                }
+                if(!bracketProcessChar(&bracketData, i)) {
+                    *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return UBIDI_LTR;
+                }
+            }
+        }
+        return direction;
+    }
+    {
+        /* continue to perform (Xn) */
+
+        /* (X1) level is set for all codes, embeddingLevel keeps track of the push/pop operations */
+        /* both variables may carry the UBIDI_LEVEL_OVERRIDE flag to indicate the override status */
+        UBiDiLevel embeddingLevel=level, newLevel;
+        UBiDiLevel previousLevel=level;     /* previous level for regular (not CC) characters */
+        int32_t lastCcPos=0;                /* index of last effective LRx,RLx, PDx */
+
+        /* The following stack remembers the embedding level and the ISOLATE flag of level runs.
+           stackLast points to its current entry. */
+        uint16_t stack[UBIDI_MAX_EXPLICIT_LEVEL+2];   /* we never push anything >=UBIDI_MAX_EXPLICIT_LEVEL
+                                                        but we need one more entry as base */
+        uint32_t stackLast=0;
+        int32_t overflowIsolateCount=0;
+        int32_t overflowEmbeddingCount=0;
+        int32_t validIsolateCount=0;
+        BracketData bracketData;
+        bracketInit(pBiDi, &bracketData);
+        stack[0]=level;     /* initialize base entry to para level, no override, no isolate */
+
+        /* recalculate the flags */
+        flags=0;
+
+        for(i=0; i<length; ++i) {
+            dirProp=dirProps[i];
+            switch(dirProp) {
+            case LRE:
+            case RLE:
+            case LRO:
+            case RLO:
+                /* (X2, X3, X4, X5) */
+                flags|=DIRPROP_FLAG(BN);
+                levels[i]=previousLevel;
+                if (dirProp==LRE || dirProp==LRO)
+                    /* least greater even level */
+                    newLevel=(UBiDiLevel)((embeddingLevel+2)&~(UBIDI_LEVEL_OVERRIDE|1));
+                else
+                    /* least greater odd level */
+                    newLevel=(UBiDiLevel)((NO_OVERRIDE(embeddingLevel)+1)|1);
+                if(newLevel<=UBIDI_MAX_EXPLICIT_LEVEL && overflowIsolateCount==0 &&
+                                                         overflowEmbeddingCount==0) {
+                    lastCcPos=i;
+                    embeddingLevel=newLevel;
+                    if(dirProp==LRO || dirProp==RLO)
+                        embeddingLevel|=UBIDI_LEVEL_OVERRIDE;
+                    stackLast++;
+                    stack[stackLast]=embeddingLevel;
+                    /* we don't need to set UBIDI_LEVEL_OVERRIDE off for LRE and RLE
+                       since this has already been done for newLevel which is
+                       the source for embeddingLevel.
+                     */
+                } else {
+                    if(overflowIsolateCount==0)
+                        overflowEmbeddingCount++;
+                }
+                break;
+            case PDF:
+                /* (X7) */
+                flags|=DIRPROP_FLAG(BN);
+                levels[i]=previousLevel;
+                /* handle all the overflow cases first */
+                if(overflowIsolateCount) {
+                    break;
+                }
+                if(overflowEmbeddingCount) {
+                    overflowEmbeddingCount--;
+                    break;
+                }
+                if(stackLast>0 && stack[stackLast]<ISOLATE) {   /* not an isolate entry */
+                    lastCcPos=i;
+                    stackLast--;
+                    embeddingLevel=(UBiDiLevel)stack[stackLast];
+                }
+                break;
+            case LRI:
+            case RLI:
+                flags|=(DIRPROP_FLAG(ON)|DIRPROP_FLAG_LR(embeddingLevel));
+                levels[i]=NO_OVERRIDE(embeddingLevel);
+                if(NO_OVERRIDE(embeddingLevel)!=NO_OVERRIDE(previousLevel)) {
+                    bracketProcessBoundary(&bracketData, lastCcPos,
+                                           previousLevel, embeddingLevel);
+                    flags|=DIRPROP_FLAG_MULTI_RUNS;
+                }
+                previousLevel=embeddingLevel;
+                /* (X5a, X5b) */
+                if(dirProp==LRI)
+                    /* least greater even level */
+                    newLevel=(UBiDiLevel)((embeddingLevel+2)&~(UBIDI_LEVEL_OVERRIDE|1));
+                else
+                    /* least greater odd level */
+                    newLevel=(UBiDiLevel)((NO_OVERRIDE(embeddingLevel)+1)|1);
+                if(newLevel<=UBIDI_MAX_EXPLICIT_LEVEL && overflowIsolateCount==0 &&
+                                                         overflowEmbeddingCount==0) {
+                    flags|=DIRPROP_FLAG(dirProp);
+                    lastCcPos=i;
+                    validIsolateCount++;
+                    if(validIsolateCount>pBiDi->isolateCount)
+                        pBiDi->isolateCount=validIsolateCount;
+                    embeddingLevel=newLevel;
+                    /* we can increment stackLast without checking because newLevel
+                       will exceed UBIDI_MAX_EXPLICIT_LEVEL before stackLast overflows */
+                    stackLast++;
+                    stack[stackLast]=embeddingLevel+ISOLATE;
+                    bracketProcessLRI_RLI(&bracketData, embeddingLevel);
+                } else {
+                    /* make it WS so that it is handled by adjustWSLevels() */
+                    dirProps[i]=WS;
+                    overflowIsolateCount++;
+                }
+                break;
+            case PDI:
+                if(NO_OVERRIDE(embeddingLevel)!=NO_OVERRIDE(previousLevel)) {
+                    bracketProcessBoundary(&bracketData, lastCcPos,
+                                           previousLevel, embeddingLevel);
+                    flags|=DIRPROP_FLAG_MULTI_RUNS;
+                }
+                /* (X6a) */
+                if(overflowIsolateCount) {
+                    overflowIsolateCount--;
+                    /* make it WS so that it is handled by adjustWSLevels() */
+                    dirProps[i]=WS;
+                }
+                else if(validIsolateCount) {
+                    flags|=DIRPROP_FLAG(PDI);
+                    lastCcPos=i;
+                    overflowEmbeddingCount=0;
+                    while(stack[stackLast]<ISOLATE) /* pop embedding entries */
+                        stackLast--;                /* until the last isolate entry */
+                    stackLast--;                    /* pop also the last isolate entry */
+                    validIsolateCount--;
+                    bracketProcessPDI(&bracketData);
+                } else
+                    /* make it WS so that it is handled by adjustWSLevels() */
+                    dirProps[i]=WS;
+                embeddingLevel=(UBiDiLevel)stack[stackLast]&~ISOLATE;
+                flags|=(DIRPROP_FLAG(ON)|DIRPROP_FLAG_LR(embeddingLevel));
+                previousLevel=embeddingLevel;
+                levels[i]=NO_OVERRIDE(embeddingLevel);
+                break;
+            case B:
+                flags|=DIRPROP_FLAG(B);
+                levels[i]=GET_PARALEVEL(pBiDi, i);
+                if((i+1)<length) {
+                    if(text[i]==CR && text[i+1]==LF)
+                        break;          /* skip CR when followed by LF */
+                    overflowEmbeddingCount=overflowIsolateCount=0;
+                    validIsolateCount=0;
+                    stackLast=0;
+                    previousLevel=embeddingLevel=GET_PARALEVEL(pBiDi, i+1);
+                    stack[0]=embeddingLevel; /* initialize base entry to para level, no override, no isolate */
+                    bracketProcessB(&bracketData, embeddingLevel);
+                }
+                break;
+            case BN:
+                /* BN, LRE, RLE, and PDF are supposed to be removed (X9) */
+                /* they will get their levels set correctly in adjustWSLevels() */
+                levels[i]=previousLevel;
+                flags|=DIRPROP_FLAG(BN);
+                break;
+            default:
+                /* all other types are normal characters and get the "real" level */
+                if(NO_OVERRIDE(embeddingLevel)!=NO_OVERRIDE(previousLevel)) {
+                    bracketProcessBoundary(&bracketData, lastCcPos,
+                                           previousLevel, embeddingLevel);
+                    flags|=DIRPROP_FLAG_MULTI_RUNS;
+                    if(embeddingLevel&UBIDI_LEVEL_OVERRIDE)
+                        flags|=DIRPROP_FLAG_O(embeddingLevel);
+                    else
+                        flags|=DIRPROP_FLAG_E(embeddingLevel);
+                }
+                previousLevel=embeddingLevel;
+                levels[i]=embeddingLevel;
+                if(!bracketProcessChar(&bracketData, i))
+                    return (UBiDiDirection)-1;
+                /* the dirProp may have been changed in bracketProcessChar() */
+                flags|=DIRPROP_FLAG(dirProps[i]);
+                break;
+            }
+        }
+        if(flags&MASK_EMBEDDING)
+            flags|=DIRPROP_FLAG_LR(pBiDi->paraLevel);
+        if(pBiDi->orderParagraphsLTR && (flags&DIRPROP_FLAG(B)))
+            flags|=DIRPROP_FLAG(L);
+        /* again, determine if the text is mixed-directional or single-directional */
+        pBiDi->flags=flags;
+        direction=directionFromFlags(pBiDi);
+    }
+    return direction;
+}
+
+/*
+ * Use a pre-specified embedding levels array:
+ *
+ * Adjust the directional properties for overrides (->LEVEL_OVERRIDE),
+ * ignore all explicit codes (X9),
+ * and check all the preset levels.
+ *
+ * Recalculate the flags to have them reflect the real properties
+ * after taking the explicit embeddings into account.
+ */
+static UBiDiDirection
+checkExplicitLevels(UBiDi *pBiDi, UErrorCode *pErrorCode) {
+    DirProp *dirProps=pBiDi->dirProps;
+    UBiDiLevel *levels=pBiDi->levels;
+    int32_t isolateCount=0;
+
+    int32_t length=pBiDi->length;
+    Flags flags=0;  /* collect all directionalities in the text */
+    pBiDi->isolateCount=0;
+
+    int32_t currentParaIndex = 0;
+    int32_t currentParaLimit = pBiDi->paras[0].limit;
+    int32_t currentParaLevel = pBiDi->paraLevel;
+
+    for(int32_t i=0; i<length; ++i) {
+        UBiDiLevel level=levels[i];
+        DirProp dirProp=dirProps[i];
+        if(dirProp==LRI || dirProp==RLI) {
+            isolateCount++;
+            if(isolateCount>pBiDi->isolateCount)
+                pBiDi->isolateCount=isolateCount;
+        }
+        else if(dirProp==PDI)
+            isolateCount--;
+        else if(dirProp==B)
+            isolateCount=0;
+
+        // optimized version of  int32_t currentParaLevel = GET_PARALEVEL(pBiDi, i);
+        if (pBiDi->defaultParaLevel != 0 &&
+                i == currentParaLimit && (currentParaIndex + 1) < pBiDi->paraCount) {
+            currentParaLevel = pBiDi->paras[++currentParaIndex].level;
+            currentParaLimit = pBiDi->paras[currentParaIndex].limit;
+        }
+
+        UBiDiLevel overrideFlag = level & UBIDI_LEVEL_OVERRIDE;
+        level &= ~UBIDI_LEVEL_OVERRIDE;
+        if (level < currentParaLevel || UBIDI_MAX_EXPLICIT_LEVEL < level) {
+            if (level == 0) {
+                if (dirProp == B) {
+                    // Paragraph separators are ok with explicit level 0.
+                    // Prevents reordering of paragraphs.
+                } else {
+                    // Treat explicit level 0 as a wildcard for the paragraph level.
+                    // Avoid making the caller guess what the paragraph level would be.
+                    level = (UBiDiLevel)currentParaLevel;
+                    levels[i] = level | overrideFlag;
+                }
+            } else {
+                // 1 <= level < currentParaLevel or UBIDI_MAX_EXPLICIT_LEVEL < level
+                /* level out of bounds */
+                *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+                return UBIDI_LTR;
+            }
+        }
+        if (overrideFlag != 0) {
+            /* keep the override flag in levels[i] but adjust the flags */
+            flags|=DIRPROP_FLAG_O(level);
+        } else {
+            /* set the flags */
+            flags|=DIRPROP_FLAG_E(level)|DIRPROP_FLAG(dirProp);
+        }
+    }
+    if(flags&MASK_EMBEDDING)
+        flags|=DIRPROP_FLAG_LR(pBiDi->paraLevel);
+    /* determine if the text is mixed-directional or single-directional */
+    pBiDi->flags=flags;
+    return directionFromFlags(pBiDi);
+}
+
+/******************************************************************
+ The Properties state machine table
+*******************************************************************
+
+ All table cells are 8 bits:
+      bits 0..4:  next state
+      bits 5..7:  action to perform (if > 0)
+
+ Cells may be of format "n" where n represents the next state
+ (except for the rightmost column).
+ Cells may also be of format "s(x,y)" where x represents an action
+ to perform and y represents the next state.
+
+*******************************************************************
+ Definitions and type for properties state table
+*******************************************************************
+*/
+#define IMPTABPROPS_COLUMNS 16
+#define IMPTABPROPS_RES (IMPTABPROPS_COLUMNS - 1)
+#define GET_STATEPROPS(cell) ((cell)&0x1f)
+#define GET_ACTIONPROPS(cell) ((cell)>>5)
+#define s(action, newState) ((uint8_t)(newState+(action<<5)))
+
+static const uint8_t groupProp[] =          /* dirProp regrouped */
+{
+/*  L   R   EN  ES  ET  AN  CS  B   S   WS  ON  LRE LRO AL  RLE RLO PDF NSM BN  FSI LRI RLI PDI ENL ENR */
+    0,  1,  2,  7,  8,  3,  9,  6,  5,  4,  4,  10, 10, 12, 10, 10, 10, 11, 10, 4,  4,  4,  4,  13, 14
+};
+enum { DirProp_L=0, DirProp_R=1, DirProp_EN=2, DirProp_AN=3, DirProp_ON=4, DirProp_S=5, DirProp_B=6 }; /* reduced dirProp */
+
+/******************************************************************
+
+      PROPERTIES  STATE  TABLE
+
+ In table impTabProps,
+      - the ON column regroups ON and WS, FSI, RLI, LRI and PDI
+      - the BN column regroups BN, LRE, RLE, LRO, RLO, PDF
+      - the Res column is the reduced property assigned to a run
+
+ Action 1: process current run1, init new run1
+        2: init new run2
+        3: process run1, process run2, init new run1
+        4: process run1, set run1=run2, init new run2
+
+ Notes:
+  1) This table is used in resolveImplicitLevels().
+  2) This table triggers actions when there is a change in the Bidi
+     property of incoming characters (action 1).
+  3) Most such property sequences are processed immediately (in
+     fact, passed to processPropertySeq().
+  4) However, numbers are assembled as one sequence. This means
+     that undefined situations (like CS following digits, until
+     it is known if the next char will be a digit) are held until
+     following chars define them.
+     Example: digits followed by CS, then comes another CS or ON;
+              the digits will be processed, then the CS assigned
+              as the start of an ON sequence (action 3).
+  5) There are cases where more than one sequence must be
+     processed, for instance digits followed by CS followed by L:
+     the digits must be processed as one sequence, and the CS
+     must be processed as an ON sequence, all this before starting
+     assembling chars for the opening L sequence.
+
+
+*/
+static const uint8_t impTabProps[][IMPTABPROPS_COLUMNS] =
+{
+/*                        L ,     R ,    EN ,    AN ,    ON ,     S ,     B ,    ES ,    ET ,    CS ,    BN ,   NSM ,    AL ,   ENL ,   ENR , Res */
+/* 0 Init        */ {     1 ,     2 ,     4 ,     5 ,     7 ,    15 ,    17 ,     7 ,     9 ,     7 ,     0 ,     7 ,     3 ,    18 ,    21 , DirProp_ON },
+/* 1 L           */ {     1 , s(1,2), s(1,4), s(1,5), s(1,7),s(1,15),s(1,17), s(1,7), s(1,9), s(1,7),     1 ,     1 , s(1,3),s(1,18),s(1,21),  DirProp_L },
+/* 2 R           */ { s(1,1),     2 , s(1,4), s(1,5), s(1,7),s(1,15),s(1,17), s(1,7), s(1,9), s(1,7),     2 ,     2 , s(1,3),s(1,18),s(1,21),  DirProp_R },
+/* 3 AL          */ { s(1,1), s(1,2), s(1,6), s(1,6), s(1,8),s(1,16),s(1,17), s(1,8), s(1,8), s(1,8),     3 ,     3 ,     3 ,s(1,18),s(1,21),  DirProp_R },
+/* 4 EN          */ { s(1,1), s(1,2),     4 , s(1,5), s(1,7),s(1,15),s(1,17),s(2,10),    11 ,s(2,10),     4 ,     4 , s(1,3),    18 ,    21 , DirProp_EN },
+/* 5 AN          */ { s(1,1), s(1,2), s(1,4),     5 , s(1,7),s(1,15),s(1,17), s(1,7), s(1,9),s(2,12),     5 ,     5 , s(1,3),s(1,18),s(1,21), DirProp_AN },
+/* 6 AL:EN/AN    */ { s(1,1), s(1,2),     6 ,     6 , s(1,8),s(1,16),s(1,17), s(1,8), s(1,8),s(2,13),     6 ,     6 , s(1,3),    18 ,    21 , DirProp_AN },
+/* 7 ON          */ { s(1,1), s(1,2), s(1,4), s(1,5),     7 ,s(1,15),s(1,17),     7 ,s(2,14),     7 ,     7 ,     7 , s(1,3),s(1,18),s(1,21), DirProp_ON },
+/* 8 AL:ON       */ { s(1,1), s(1,2), s(1,6), s(1,6),     8 ,s(1,16),s(1,17),     8 ,     8 ,     8 ,     8 ,     8 , s(1,3),s(1,18),s(1,21), DirProp_ON },
+/* 9 ET          */ { s(1,1), s(1,2),     4 , s(1,5),     7 ,s(1,15),s(1,17),     7 ,     9 ,     7 ,     9 ,     9 , s(1,3),    18 ,    21 , DirProp_ON },
+/*10 EN+ES/CS    */ { s(3,1), s(3,2),     4 , s(3,5), s(4,7),s(3,15),s(3,17), s(4,7),s(4,14), s(4,7),    10 , s(4,7), s(3,3),    18 ,    21 , DirProp_EN },
+/*11 EN+ET       */ { s(1,1), s(1,2),     4 , s(1,5), s(1,7),s(1,15),s(1,17), s(1,7),    11 , s(1,7),    11 ,    11 , s(1,3),    18 ,    21 , DirProp_EN },
+/*12 AN+CS       */ { s(3,1), s(3,2), s(3,4),     5 , s(4,7),s(3,15),s(3,17), s(4,7),s(4,14), s(4,7),    12 , s(4,7), s(3,3),s(3,18),s(3,21), DirProp_AN },
+/*13 AL:EN/AN+CS */ { s(3,1), s(3,2),     6 ,     6 , s(4,8),s(3,16),s(3,17), s(4,8), s(4,8), s(4,8),    13 , s(4,8), s(3,3),    18 ,    21 , DirProp_AN },
+/*14 ON+ET       */ { s(1,1), s(1,2), s(4,4), s(1,5),     7 ,s(1,15),s(1,17),     7 ,    14 ,     7 ,    14 ,    14 , s(1,3),s(4,18),s(4,21), DirProp_ON },
+/*15 S           */ { s(1,1), s(1,2), s(1,4), s(1,5), s(1,7),    15 ,s(1,17), s(1,7), s(1,9), s(1,7),    15 , s(1,7), s(1,3),s(1,18),s(1,21),  DirProp_S },
+/*16 AL:S        */ { s(1,1), s(1,2), s(1,6), s(1,6), s(1,8),    16 ,s(1,17), s(1,8), s(1,8), s(1,8),    16 , s(1,8), s(1,3),s(1,18),s(1,21),  DirProp_S },
+/*17 B           */ { s(1,1), s(1,2), s(1,4), s(1,5), s(1,7),s(1,15),    17 , s(1,7), s(1,9), s(1,7),    17 , s(1,7), s(1,3),s(1,18),s(1,21),  DirProp_B },
+/*18 ENL         */ { s(1,1), s(1,2),    18 , s(1,5), s(1,7),s(1,15),s(1,17),s(2,19),    20 ,s(2,19),    18 ,    18 , s(1,3),    18 ,    21 ,  DirProp_L },
+/*19 ENL+ES/CS   */ { s(3,1), s(3,2),    18 , s(3,5), s(4,7),s(3,15),s(3,17), s(4,7),s(4,14), s(4,7),    19 , s(4,7), s(3,3),    18 ,    21 ,  DirProp_L },
+/*20 ENL+ET      */ { s(1,1), s(1,2),    18 , s(1,5), s(1,7),s(1,15),s(1,17), s(1,7),    20 , s(1,7),    20 ,    20 , s(1,3),    18 ,    21 ,  DirProp_L },
+/*21 ENR         */ { s(1,1), s(1,2),    21 , s(1,5), s(1,7),s(1,15),s(1,17),s(2,22),    23 ,s(2,22),    21 ,    21 , s(1,3),    18 ,    21 , DirProp_AN },
+/*22 ENR+ES/CS   */ { s(3,1), s(3,2),    21 , s(3,5), s(4,7),s(3,15),s(3,17), s(4,7),s(4,14), s(4,7),    22 , s(4,7), s(3,3),    18 ,    21 , DirProp_AN },
+/*23 ENR+ET      */ { s(1,1), s(1,2),    21 , s(1,5), s(1,7),s(1,15),s(1,17), s(1,7),    23 , s(1,7),    23 ,    23 , s(1,3),    18 ,    21 , DirProp_AN }
+};
+
+/*  we must undef macro s because the levels tables have a different
+ *  structure (4 bits for action and 4 bits for next state.
+ */
+#undef s
+
+/******************************************************************
+ The levels state machine tables
+*******************************************************************
+
+ All table cells are 8 bits:
+      bits 0..3:  next state
+      bits 4..7:  action to perform (if > 0)
+
+ Cells may be of format "n" where n represents the next state
+ (except for the rightmost column).
+ Cells may also be of format "s(x,y)" where x represents an action
+ to perform and y represents the next state.
+
+ This format limits each table to 16 states each and to 15 actions.
+
+*******************************************************************
+ Definitions and type for levels state tables
+*******************************************************************
+*/
+#define IMPTABLEVELS_COLUMNS (DirProp_B + 2)
+#define IMPTABLEVELS_RES (IMPTABLEVELS_COLUMNS - 1)
+#define GET_STATE(cell) ((cell)&0x0f)
+#define GET_ACTION(cell) ((cell)>>4)
+#define s(action, newState) ((uint8_t)(newState+(action<<4)))
+
+typedef uint8_t ImpTab[][IMPTABLEVELS_COLUMNS];
+typedef uint8_t ImpAct[];
+
+/* FOOD FOR THOUGHT: each ImpTab should have its associated ImpAct,
+ * instead of having a pair of ImpTab and a pair of ImpAct.
+ */
+typedef struct ImpTabPair {
+    const void * pImpTab[2];
+    const void * pImpAct[2];
+} ImpTabPair;
+
+/******************************************************************
+
+      LEVELS  STATE  TABLES
+
+ In all levels state tables,
+      - state 0 is the initial state
+      - the Res column is the increment to add to the text level
+        for this property sequence.
+
+ The impAct arrays for each table of a pair map the local action
+ numbers of the table to the total list of actions. For instance,
+ action 2 in a given table corresponds to the action number which
+ appears in entry [2] of the impAct array for that table.
+ The first entry of all impAct arrays must be 0.
+
+ Action 1: init conditional sequence
+        2: prepend conditional sequence to current sequence
+        3: set ON sequence to new level - 1
+        4: init EN/AN/ON sequence
+        5: fix EN/AN/ON sequence followed by R
+        6: set previous level sequence to level 2
+
+ Notes:
+  1) These tables are used in processPropertySeq(). The input
+     is property sequences as determined by resolveImplicitLevels.
+  2) Most such property sequences are processed immediately
+     (levels are assigned).
+  3) However, some sequences cannot be assigned a final level till
+     one or more following sequences are received. For instance,
+     ON following an R sequence within an even-level paragraph.
+     If the following sequence is R, the ON sequence will be
+     assigned basic run level+1, and so will the R sequence.
+  4) S is generally handled like ON, since its level will be fixed
+     to paragraph level in adjustWSLevels().
+
+*/
+
+static const ImpTab impTabL_DEFAULT =   /* Even paragraph level */
+/*  In this table, conditional sequences receive the lower possible level
+    until proven otherwise.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     0 ,     1 ,     0 ,     2 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : R          */ {     0 ,     1 ,     3 ,     3 , s(1,4), s(1,4),     0 ,  1 },
+/* 2 : AN         */ {     0 ,     1 ,     0 ,     2 , s(1,5), s(1,5),     0 ,  2 },
+/* 3 : R+EN/AN    */ {     0 ,     1 ,     3 ,     3 , s(1,4), s(1,4),     0 ,  2 },
+/* 4 : R+ON       */ {     0 , s(2,1), s(3,3), s(3,3),     4 ,     4 ,     0 ,  0 },
+/* 5 : AN+ON      */ {     0 , s(2,1),     0 , s(3,2),     5 ,     5 ,     0 ,  0 }
+};
+static const ImpTab impTabR_DEFAULT =   /* Odd  paragraph level */
+/*  In this table, conditional sequences receive the lower possible level
+    until proven otherwise.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     1 ,     0 ,     2 ,     2 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : L          */ {     1 ,     0 ,     1 ,     3 , s(1,4), s(1,4),     0 ,  1 },
+/* 2 : EN/AN      */ {     1 ,     0 ,     2 ,     2 ,     0 ,     0 ,     0 ,  1 },
+/* 3 : L+AN       */ {     1 ,     0 ,     1 ,     3 ,     5 ,     5 ,     0 ,  1 },
+/* 4 : L+ON       */ { s(2,1),     0 , s(2,1),     3 ,     4 ,     4 ,     0 ,  0 },
+/* 5 : L+AN+ON    */ {     1 ,     0 ,     1 ,     3 ,     5 ,     5 ,     0 ,  0 }
+};
+static const ImpAct impAct0 = {0,1,2,3,4};
+static const ImpTabPair impTab_DEFAULT = {{&impTabL_DEFAULT,
+                                           &impTabR_DEFAULT},
+                                          {&impAct0, &impAct0}};
+
+static const ImpTab impTabL_NUMBERS_SPECIAL =   /* Even paragraph level */
+/*  In this table, conditional sequences receive the lower possible level
+    until proven otherwise.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     0 ,     2 , s(1,1), s(1,1),     0 ,     0 ,     0 ,  0 },
+/* 1 : L+EN/AN    */ {     0 , s(4,2),     1 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 2 : R          */ {     0 ,     2 ,     4 ,     4 , s(1,3), s(1,3),     0 ,  1 },
+/* 3 : R+ON       */ {     0 , s(2,2), s(3,4), s(3,4),     3 ,     3 ,     0 ,  0 },
+/* 4 : R+EN/AN    */ {     0 ,     2 ,     4 ,     4 , s(1,3), s(1,3),     0 ,  2 }
+};
+static const ImpTabPair impTab_NUMBERS_SPECIAL = {{&impTabL_NUMBERS_SPECIAL,
+                                                   &impTabR_DEFAULT},
+                                                  {&impAct0, &impAct0}};
+
+static const ImpTab impTabL_GROUP_NUMBERS_WITH_R =
+/*  In this table, EN/AN+ON sequences receive levels as if associated with R
+    until proven that there is L or sor/eor on both sides. AN is handled like EN.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 init         */ {     0 ,     3 , s(1,1), s(1,1),     0 ,     0 ,     0 ,  0 },
+/* 1 EN/AN        */ { s(2,0),     3 ,     1 ,     1 ,     2 , s(2,0), s(2,0),  2 },
+/* 2 EN/AN+ON     */ { s(2,0),     3 ,     1 ,     1 ,     2 , s(2,0), s(2,0),  1 },
+/* 3 R            */ {     0 ,     3 ,     5 ,     5 , s(1,4),     0 ,     0 ,  1 },
+/* 4 R+ON         */ { s(2,0),     3 ,     5 ,     5 ,     4 , s(2,0), s(2,0),  1 },
+/* 5 R+EN/AN      */ {     0 ,     3 ,     5 ,     5 , s(1,4),     0 ,     0 ,  2 }
+};
+static const ImpTab impTabR_GROUP_NUMBERS_WITH_R =
+/*  In this table, EN/AN+ON sequences receive levels as if associated with R
+    until proven that there is L on both sides. AN is handled like EN.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 init         */ {     2 ,     0 ,     1 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 1 EN/AN        */ {     2 ,     0 ,     1 ,     1 ,     0 ,     0 ,     0 ,  1 },
+/* 2 L            */ {     2 ,     0 , s(1,4), s(1,4), s(1,3),     0 ,     0 ,  1 },
+/* 3 L+ON         */ { s(2,2),     0 ,     4 ,     4 ,     3 ,     0 ,     0 ,  0 },
+/* 4 L+EN/AN      */ { s(2,2),     0 ,     4 ,     4 ,     3 ,     0 ,     0 ,  1 }
+};
+static const ImpTabPair impTab_GROUP_NUMBERS_WITH_R = {
+                        {&impTabL_GROUP_NUMBERS_WITH_R,
+                         &impTabR_GROUP_NUMBERS_WITH_R},
+                        {&impAct0, &impAct0}};
+
+
+static const ImpTab impTabL_INVERSE_NUMBERS_AS_L =
+/*  This table is identical to the Default LTR table except that EN and AN are
+    handled like L.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     0 ,     1 ,     0 ,     0 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : R          */ {     0 ,     1 ,     0 ,     0 , s(1,4), s(1,4),     0 ,  1 },
+/* 2 : AN         */ {     0 ,     1 ,     0 ,     0 , s(1,5), s(1,5),     0 ,  2 },
+/* 3 : R+EN/AN    */ {     0 ,     1 ,     0 ,     0 , s(1,4), s(1,4),     0 ,  2 },
+/* 4 : R+ON       */ { s(2,0),     1 , s(2,0), s(2,0),     4 ,     4 , s(2,0),  1 },
+/* 5 : AN+ON      */ { s(2,0),     1 , s(2,0), s(2,0),     5 ,     5 , s(2,0),  1 }
+};
+static const ImpTab impTabR_INVERSE_NUMBERS_AS_L =
+/*  This table is identical to the Default RTL table except that EN and AN are
+    handled like L.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     1 ,     0 ,     1 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : L          */ {     1 ,     0 ,     1 ,     1 , s(1,4), s(1,4),     0 ,  1 },
+/* 2 : EN/AN      */ {     1 ,     0 ,     1 ,     1 ,     0 ,     0 ,     0 ,  1 },
+/* 3 : L+AN       */ {     1 ,     0 ,     1 ,     1 ,     5 ,     5 ,     0 ,  1 },
+/* 4 : L+ON       */ { s(2,1),     0 , s(2,1), s(2,1),     4 ,     4 ,     0 ,  0 },
+/* 5 : L+AN+ON    */ {     1 ,     0 ,     1 ,     1 ,     5 ,     5 ,     0 ,  0 }
+};
+static const ImpTabPair impTab_INVERSE_NUMBERS_AS_L = {
+                        {&impTabL_INVERSE_NUMBERS_AS_L,
+                         &impTabR_INVERSE_NUMBERS_AS_L},
+                        {&impAct0, &impAct0}};
+
+static const ImpTab impTabR_INVERSE_LIKE_DIRECT =   /* Odd  paragraph level */
+/*  In this table, conditional sequences receive the lower possible level
+    until proven otherwise.
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     1 ,     0 ,     2 ,     2 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : L          */ {     1 ,     0 ,     1 ,     2 , s(1,3), s(1,3),     0 ,  1 },
+/* 2 : EN/AN      */ {     1 ,     0 ,     2 ,     2 ,     0 ,     0 ,     0 ,  1 },
+/* 3 : L+ON       */ { s(2,1), s(3,0),     6 ,     4 ,     3 ,     3 , s(3,0),  0 },
+/* 4 : L+ON+AN    */ { s(2,1), s(3,0),     6 ,     4 ,     5 ,     5 , s(3,0),  3 },
+/* 5 : L+AN+ON    */ { s(2,1), s(3,0),     6 ,     4 ,     5 ,     5 , s(3,0),  2 },
+/* 6 : L+ON+EN    */ { s(2,1), s(3,0),     6 ,     4 ,     3 ,     3 , s(3,0),  1 }
+};
+static const ImpAct impAct1 = {0,1,13,14};
+/* FOOD FOR THOUGHT: in LTR table below, check case "JKL 123abc"
+ */
+static const ImpTabPair impTab_INVERSE_LIKE_DIRECT = {
+                        {&impTabL_DEFAULT,
+                         &impTabR_INVERSE_LIKE_DIRECT},
+                        {&impAct0, &impAct1}};
+
+static const ImpTab impTabL_INVERSE_LIKE_DIRECT_WITH_MARKS =
+/*  The case handled in this table is (visually):  R EN L
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     0 , s(6,3),     0 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : L+AN       */ {     0 , s(6,3),     0 ,     1 , s(1,2), s(3,0),     0 ,  4 },
+/* 2 : L+AN+ON    */ { s(2,0), s(6,3), s(2,0),     1 ,     2 , s(3,0), s(2,0),  3 },
+/* 3 : R          */ {     0 , s(6,3), s(5,5), s(5,6), s(1,4), s(3,0),     0 ,  3 },
+/* 4 : R+ON       */ { s(3,0), s(4,3), s(5,5), s(5,6),     4 , s(3,0), s(3,0),  3 },
+/* 5 : R+EN       */ { s(3,0), s(4,3),     5 , s(5,6), s(1,4), s(3,0), s(3,0),  4 },
+/* 6 : R+AN       */ { s(3,0), s(4,3), s(5,5),     6 , s(1,4), s(3,0), s(3,0),  4 }
+};
+static const ImpTab impTabR_INVERSE_LIKE_DIRECT_WITH_MARKS =
+/*  The cases handled in this table are (visually):  R EN L
+                                                     R L AN L
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ { s(1,3),     0 ,     1 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : R+EN/AN    */ { s(2,3),     0 ,     1 ,     1 ,     2 , s(4,0),     0 ,  1 },
+/* 2 : R+EN/AN+ON */ { s(2,3),     0 ,     1 ,     1 ,     2 , s(4,0),     0 ,  0 },
+/* 3 : L          */ {     3 ,     0 ,     3 , s(3,6), s(1,4), s(4,0),     0 ,  1 },
+/* 4 : L+ON       */ { s(5,3), s(4,0),     5 , s(3,6),     4 , s(4,0), s(4,0),  0 },
+/* 5 : L+ON+EN    */ { s(5,3), s(4,0),     5 , s(3,6),     4 , s(4,0), s(4,0),  1 },
+/* 6 : L+AN       */ { s(5,3), s(4,0),     6 ,     6 ,     4 , s(4,0), s(4,0),  3 }
+};
+static const ImpAct impAct2 = {0,1,2,5,6,7,8};
+static const ImpAct impAct3 = {0,1,9,10,11,12};
+static const ImpTabPair impTab_INVERSE_LIKE_DIRECT_WITH_MARKS = {
+                        {&impTabL_INVERSE_LIKE_DIRECT_WITH_MARKS,
+                         &impTabR_INVERSE_LIKE_DIRECT_WITH_MARKS},
+                        {&impAct2, &impAct3}};
+
+static const ImpTabPair impTab_INVERSE_FOR_NUMBERS_SPECIAL = {
+                        {&impTabL_NUMBERS_SPECIAL,
+                         &impTabR_INVERSE_LIKE_DIRECT},
+                        {&impAct0, &impAct1}};
+
+static const ImpTab impTabL_INVERSE_FOR_NUMBERS_SPECIAL_WITH_MARKS =
+/*  The case handled in this table is (visually):  R EN L
+*/
+{
+/*                         L ,     R ,    EN ,    AN ,    ON ,     S ,     B , Res */
+/* 0 : init       */ {     0 , s(6,2),     1 ,     1 ,     0 ,     0 ,     0 ,  0 },
+/* 1 : L+EN/AN    */ {     0 , s(6,2),     1 ,     1 ,     0 , s(3,0),     0 ,  4 },
+/* 2 : R          */ {     0 , s(6,2), s(5,4), s(5,4), s(1,3), s(3,0),     0 ,  3 },
+/* 3 : R+ON       */ { s(3,0), s(4,2), s(5,4), s(5,4),     3 , s(3,0), s(3,0),  3 },
+/* 4 : R+EN/AN    */ { s(3,0), s(4,2),     4 ,     4 , s(1,3), s(3,0), s(3,0),  4 }
+};
+static const ImpTabPair impTab_INVERSE_FOR_NUMBERS_SPECIAL_WITH_MARKS = {
+                        {&impTabL_INVERSE_FOR_NUMBERS_SPECIAL_WITH_MARKS,
+                         &impTabR_INVERSE_LIKE_DIRECT_WITH_MARKS},
+                        {&impAct2, &impAct3}};
+
+#undef s
+
+typedef struct {
+    const ImpTab * pImpTab;             /* level table pointer          */
+    const ImpAct * pImpAct;             /* action map array             */
+    int32_t startON;                    /* start of ON sequence         */
+    int32_t startL2EN;                  /* start of level 2 sequence    */
+    int32_t lastStrongRTL;              /* index of last found R or AL  */
+    int32_t state;                      /* current state                */
+    int32_t runStart;                   /* start position of the run    */
+    UBiDiLevel runLevel;                /* run level before implicit solving */
+} LevState;
+
+/*------------------------------------------------------------------------*/
+
+static void
+addPoint(UBiDi *pBiDi, int32_t pos, int32_t flag)
+  /* param pos:     position where to insert
+     param flag:    one of LRM_BEFORE, LRM_AFTER, RLM_BEFORE, RLM_AFTER
+  */
+{
+#define FIRSTALLOC  10
+    Point point;
+    InsertPoints * pInsertPoints=&(pBiDi->insertPoints);
+
+    if (pInsertPoints->capacity == 0)
+    {
+        pInsertPoints->points=static_cast<Point *>(uprv_malloc(sizeof(Point)*FIRSTALLOC));
+        if (pInsertPoints->points == NULL)
+        {
+            pInsertPoints->errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        pInsertPoints->capacity=FIRSTALLOC;
+    }
+    if (pInsertPoints->size >= pInsertPoints->capacity) /* no room for new point */
+    {
+        Point * savePoints=pInsertPoints->points;
+        pInsertPoints->points=static_cast<Point *>(uprv_realloc(pInsertPoints->points,
+                                           pInsertPoints->capacity*2*sizeof(Point)));
+        if (pInsertPoints->points == NULL)
+        {
+            pInsertPoints->points=savePoints;
+            pInsertPoints->errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        else  pInsertPoints->capacity*=2;
+    }
+    point.pos=pos;
+    point.flag=flag;
+    pInsertPoints->points[pInsertPoints->size]=point;
+    pInsertPoints->size++;
+#undef FIRSTALLOC
+}
+
+static void
+setLevelsOutsideIsolates(UBiDi *pBiDi, int32_t start, int32_t limit, UBiDiLevel level)
+{
+    DirProp *dirProps=pBiDi->dirProps, dirProp;
+    UBiDiLevel *levels=pBiDi->levels;
+    int32_t isolateCount=0, k;
+    for(k=start; k<limit; k++) {
+        dirProp=dirProps[k];
+        if(dirProp==PDI)
+            isolateCount--;
+        if(isolateCount==0)
+            levels[k]=level;
+        if(dirProp==LRI || dirProp==RLI)
+            isolateCount++;
+    }
+}
+
+/* perform rules (Wn), (Nn), and (In) on a run of the text ------------------ */
+
+/*
+ * This implementation of the (Wn) rules applies all rules in one pass.
+ * In order to do so, it needs a look-ahead of typically 1 character
+ * (except for W5: sequences of ET) and keeps track of changes
+ * in a rule Wp that affect a later Wq (p<q).
+ *
+ * The (Nn) and (In) rules are also performed in that same single loop,
+ * but effectively one iteration behind for white space.
+ *
+ * Since all implicit rules are performed in one step, it is not necessary
+ * to actually store the intermediate directional properties in dirProps[].
+ */
+
+static void
+processPropertySeq(UBiDi *pBiDi, LevState *pLevState, uint8_t _prop,
+                   int32_t start, int32_t limit) {
+    uint8_t cell, oldStateSeq, actionSeq;
+    const ImpTab * pImpTab=pLevState->pImpTab;
+    const ImpAct * pImpAct=pLevState->pImpAct;
+    UBiDiLevel * levels=pBiDi->levels;
+    UBiDiLevel level, addLevel;
+    InsertPoints * pInsertPoints;
+    int32_t start0, k;
+
+    start0=start;                           /* save original start position */
+    oldStateSeq=(uint8_t)pLevState->state;
+    cell=(*pImpTab)[oldStateSeq][_prop];
+    pLevState->state=GET_STATE(cell);       /* isolate the new state */
+    actionSeq=(*pImpAct)[GET_ACTION(cell)]; /* isolate the action */
+    addLevel=(*pImpTab)[pLevState->state][IMPTABLEVELS_RES];
+
+    if(actionSeq) {
+        switch(actionSeq) {
+        case 1:                         /* init ON seq */
+            pLevState->startON=start0;
+            break;
+
+        case 2:                         /* prepend ON seq to current seq */
+            start=pLevState->startON;
+            break;
+
+        case 3:                         /* EN/AN after R+ON */
+            level=pLevState->runLevel+1;
+            setLevelsOutsideIsolates(pBiDi, pLevState->startON, start0, level);
+            break;
+
+        case 4:                         /* EN/AN before R for NUMBERS_SPECIAL */
+            level=pLevState->runLevel+2;
+            setLevelsOutsideIsolates(pBiDi, pLevState->startON, start0, level);
+            break;
+
+        case 5:                         /* L or S after possible relevant EN/AN */
+            /* check if we had EN after R/AL */
+            if (pLevState->startL2EN >= 0) {
+                addPoint(pBiDi, pLevState->startL2EN, LRM_BEFORE);
+            }
+            pLevState->startL2EN=-1;  /* not within previous if since could also be -2 */
+            /* check if we had any relevant EN/AN after R/AL */
+            pInsertPoints=&(pBiDi->insertPoints);
+            if ((pInsertPoints->capacity == 0) ||
+                (pInsertPoints->size <= pInsertPoints->confirmed))
+            {
+                /* nothing, just clean up */
+                pLevState->lastStrongRTL=-1;
+                /* check if we have a pending conditional segment */
+                level=(*pImpTab)[oldStateSeq][IMPTABLEVELS_RES];
+                if ((level & 1) && (pLevState->startON > 0)) {  /* after ON */
+                    start=pLevState->startON;   /* reset to basic run level */
+                }
+                if (_prop == DirProp_S)                /* add LRM before S */
+                {
+                    addPoint(pBiDi, start0, LRM_BEFORE);
+                    pInsertPoints->confirmed=pInsertPoints->size;
+                }
+                break;
+            }
+            /* reset previous RTL cont to level for LTR text */
+            for (k=pLevState->lastStrongRTL+1; k<start0; k++)
+            {
+                /* reset odd level, leave runLevel+2 as is */
+                levels[k]=(levels[k] - 2) & ~1;
+            }
+            /* mark insert points as confirmed */
+            pInsertPoints->confirmed=pInsertPoints->size;
+            pLevState->lastStrongRTL=-1;
+            if (_prop == DirProp_S)            /* add LRM before S */
+            {
+                addPoint(pBiDi, start0, LRM_BEFORE);
+                pInsertPoints->confirmed=pInsertPoints->size;
+            }
+            break;
+
+        case 6:                         /* R/AL after possible relevant EN/AN */
+            /* just clean up */
+            pInsertPoints=&(pBiDi->insertPoints);
+            if (pInsertPoints->capacity > 0)
+                /* remove all non confirmed insert points */
+                pInsertPoints->size=pInsertPoints->confirmed;
+            pLevState->startON=-1;
+            pLevState->startL2EN=-1;
+            pLevState->lastStrongRTL=limit - 1;
+            break;
+
+        case 7:                         /* EN/AN after R/AL + possible cont */
+            /* check for real AN */
+            if ((_prop == DirProp_AN) && (pBiDi->dirProps[start0] == AN) &&
+                (pBiDi->reorderingMode!=UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL))
+            {
+                /* real AN */
+                if (pLevState->startL2EN == -1) /* if no relevant EN already found */
+                {
+                    /* just note the righmost digit as a strong RTL */
+                    pLevState->lastStrongRTL=limit - 1;
+                    break;
+                }
+                if (pLevState->startL2EN >= 0)  /* after EN, no AN */
+                {
+                    addPoint(pBiDi, pLevState->startL2EN, LRM_BEFORE);
+                    pLevState->startL2EN=-2;
+                }
+                /* note AN */
+                addPoint(pBiDi, start0, LRM_BEFORE);
+                break;
+            }
+            /* if first EN/AN after R/AL */
+            if (pLevState->startL2EN == -1) {
+                pLevState->startL2EN=start0;
+            }
+            break;
+
+        case 8:                         /* note location of latest R/AL */
+            pLevState->lastStrongRTL=limit - 1;
+            pLevState->startON=-1;
+            break;
+
+        case 9:                         /* L after R+ON/EN/AN */
+            /* include possible adjacent number on the left */
+            for (k=start0-1; k>=0 && !(levels[k]&1); k--);
+            if(k>=0) {
+                addPoint(pBiDi, k, RLM_BEFORE);             /* add RLM before */
+                pInsertPoints=&(pBiDi->insertPoints);
+                pInsertPoints->confirmed=pInsertPoints->size;   /* confirm it */
+            }
+            pLevState->startON=start0;
+            break;
+
+        case 10:                        /* AN after L */
+            /* AN numbers between L text on both sides may be trouble. */
+            /* tentatively bracket with LRMs; will be confirmed if followed by L */
+            addPoint(pBiDi, start0, LRM_BEFORE);    /* add LRM before */
+            addPoint(pBiDi, start0, LRM_AFTER);     /* add LRM after  */
+            break;
+
+        case 11:                        /* R after L+ON/EN/AN */
+            /* false alert, infirm LRMs around previous AN */
+            pInsertPoints=&(pBiDi->insertPoints);
+            pInsertPoints->size=pInsertPoints->confirmed;
+            if (_prop == DirProp_S)            /* add RLM before S */
+            {
+                addPoint(pBiDi, start0, RLM_BEFORE);
+                pInsertPoints->confirmed=pInsertPoints->size;
+            }
+            break;
+
+        case 12:                        /* L after L+ON/AN */
+            level=pLevState->runLevel + addLevel;
+            for(k=pLevState->startON; k<start0; k++) {
+                if (levels[k]<level)
+                    levels[k]=level;
+            }
+            pInsertPoints=&(pBiDi->insertPoints);
+            pInsertPoints->confirmed=pInsertPoints->size;   /* confirm inserts */
+            pLevState->startON=start0;
+            break;
+
+        case 13:                        /* L after L+ON+EN/AN/ON */
+            level=pLevState->runLevel;
+            for(k=start0-1; k>=pLevState->startON; k--) {
+                if(levels[k]==level+3) {
+                    while(levels[k]==level+3) {
+                        levels[k--]-=2;
+                    }
+                    while(levels[k]==level) {
+                        k--;
+                    }
+                }
+                if(levels[k]==level+2) {
+                    levels[k]=level;
+                    continue;
+                }
+                levels[k]=level+1;
+            }
+            break;
+
+        case 14:                        /* R after L+ON+EN/AN/ON */
+            level=pLevState->runLevel+1;
+            for(k=start0-1; k>=pLevState->startON; k--) {
+                if(levels[k]>level) {
+                    levels[k]-=2;
+                }
+            }
+            break;
+
+        default:                        /* we should never get here */
+            UPRV_UNREACHABLE;
+        }
+    }
+    if((addLevel) || (start < start0)) {
+        level=pLevState->runLevel + addLevel;
+        if(start>=pLevState->runStart) {
+            for(k=start; k<limit; k++) {
+                levels[k]=level;
+            }
+        } else {
+            setLevelsOutsideIsolates(pBiDi, start, limit, level);
+        }
+    }
+}
+
+/**
+ * Returns the directionality of the last strong character at the end of the prologue, if any.
+ * Requires prologue!=null.
+ */
+static DirProp
+lastL_R_AL(UBiDi *pBiDi) {
+    const UChar *text=pBiDi->prologue;
+    int32_t length=pBiDi->proLength;
+    int32_t i;
+    UChar32 uchar;
+    DirProp dirProp;
+    for(i=length; i>0; ) {
+        /* i is decremented by U16_PREV */
+        U16_PREV(text, 0, i, uchar);
+        dirProp=(DirProp)ubidi_getCustomizedClass(pBiDi, uchar);
+        if(dirProp==L) {
+            return DirProp_L;
+        }
+        if(dirProp==R || dirProp==AL) {
+            return DirProp_R;
+        }
+        if(dirProp==B) {
+            return DirProp_ON;
+        }
+    }
+    return DirProp_ON;
+}
+
+/**
+ * Returns the directionality of the first strong character, or digit, in the epilogue, if any.
+ * Requires epilogue!=null.
+ */
+static DirProp
+firstL_R_AL_EN_AN(UBiDi *pBiDi) {
+    const UChar *text=pBiDi->epilogue;
+    int32_t length=pBiDi->epiLength;
+    int32_t i;
+    UChar32 uchar;
+    DirProp dirProp;
+    for(i=0; i<length; ) {
+        /* i is incremented by U16_NEXT */
+        U16_NEXT(text, i, length, uchar);
+        dirProp=(DirProp)ubidi_getCustomizedClass(pBiDi, uchar);
+        if(dirProp==L) {
+            return DirProp_L;
+        }
+        if(dirProp==R || dirProp==AL) {
+            return DirProp_R;
+        }
+        if(dirProp==EN) {
+            return DirProp_EN;
+        }
+        if(dirProp==AN) {
+            return DirProp_AN;
+        }
+    }
+    return DirProp_ON;
+}
+
+static void
+resolveImplicitLevels(UBiDi *pBiDi,
+                      int32_t start, int32_t limit,
+                      DirProp sor, DirProp eor) {
+    const DirProp *dirProps=pBiDi->dirProps;
+    DirProp dirProp;
+    LevState levState;
+    int32_t i, start1, start2;
+    uint16_t oldStateImp, stateImp, actionImp;
+    uint8_t gprop, resProp, cell;
+    UBool inverseRTL;
+    DirProp nextStrongProp=R;
+    int32_t nextStrongPos=-1;
+
+    /* check for RTL inverse BiDi mode */
+    /* FOOD FOR THOUGHT: in case of RTL inverse BiDi, it would make sense to
+     * loop on the text characters from end to start.
+     * This would need a different properties state table (at least different
+     * actions) and different levels state tables (maybe very similar to the
+     * LTR corresponding ones.
+     */
+    inverseRTL=(UBool)
+        ((start<pBiDi->lastArabicPos) && (GET_PARALEVEL(pBiDi, start) & 1) &&
+         (pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_LIKE_DIRECT  ||
+          pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL));
+
+    /* initialize for property and levels state tables */
+    levState.startL2EN=-1;              /* used for INVERSE_LIKE_DIRECT_WITH_MARKS */
+    levState.lastStrongRTL=-1;          /* used for INVERSE_LIKE_DIRECT_WITH_MARKS */
+    levState.runStart=start;
+    levState.runLevel=pBiDi->levels[start];
+    levState.pImpTab=(const ImpTab*)((pBiDi->pImpTabPair)->pImpTab)[levState.runLevel&1];
+    levState.pImpAct=(const ImpAct*)((pBiDi->pImpTabPair)->pImpAct)[levState.runLevel&1];
+    if(start==0 && pBiDi->proLength>0) {
+        DirProp lastStrong=lastL_R_AL(pBiDi);
+        if(lastStrong!=DirProp_ON) {
+            sor=lastStrong;
+        }
+    }
+    /* The isolates[] entries contain enough information to
+       resume the bidi algorithm in the same state as it was
+       when it was interrupted by an isolate sequence. */
+    if(dirProps[start]==PDI  && pBiDi->isolateCount >= 0) {
+        levState.startON=pBiDi->isolates[pBiDi->isolateCount].startON;
+        start1=pBiDi->isolates[pBiDi->isolateCount].start1;
+        stateImp=pBiDi->isolates[pBiDi->isolateCount].stateImp;
+        levState.state=pBiDi->isolates[pBiDi->isolateCount].state;
+        pBiDi->isolateCount--;
+    } else {
+        levState.startON=-1;
+        start1=start;
+        if(dirProps[start]==NSM)
+            stateImp = 1 + sor;
+        else
+            stateImp=0;
+        levState.state=0;
+        processPropertySeq(pBiDi, &levState, sor, start, start);
+    }
+    start2=start;                       /* to make Java compiler happy */
+
+    for(i=start; i<=limit; i++) {
+        if(i>=limit) {
+            int32_t k;
+            for(k=limit-1; k>start&&(DIRPROP_FLAG(dirProps[k])&MASK_BN_EXPLICIT); k--);
+            dirProp=dirProps[k];
+            if(dirProp==LRI || dirProp==RLI)
+                break;      /* no forced closing for sequence ending with LRI/RLI */
+            gprop=eor;
+        } else {
+            DirProp prop, prop1;
+            prop=dirProps[i];
+            if(prop==B) {
+                pBiDi->isolateCount=-1; /* current isolates stack entry == none */
+            }
+            if(inverseRTL) {
+                if(prop==AL) {
+                    /* AL before EN does not make it AN */
+                    prop=R;
+                } else if(prop==EN) {
+                    if(nextStrongPos<=i) {
+                        /* look for next strong char (L/R/AL) */
+                        int32_t j;
+                        nextStrongProp=R;   /* set default */
+                        nextStrongPos=limit;
+                        for(j=i+1; j<limit; j++) {
+                            prop1=dirProps[j];
+                            if(prop1==L || prop1==R || prop1==AL) {
+                                nextStrongProp=prop1;
+                                nextStrongPos=j;
+                                break;
+                            }
+                        }
+                    }
+                    if(nextStrongProp==AL) {
+                        prop=AN;
+                    }
+                }
+            }
+            gprop=groupProp[prop];
+        }
+        oldStateImp=stateImp;
+        cell=impTabProps[oldStateImp][gprop];
+        stateImp=GET_STATEPROPS(cell);      /* isolate the new state */
+        actionImp=GET_ACTIONPROPS(cell);    /* isolate the action */
+        if((i==limit) && (actionImp==0)) {
+            /* there is an unprocessed sequence if its property == eor   */
+            actionImp=1;                    /* process the last sequence */
+        }
+        if(actionImp) {
+            resProp=impTabProps[oldStateImp][IMPTABPROPS_RES];
+            switch(actionImp) {
+            case 1:             /* process current seq1, init new seq1 */
+                processPropertySeq(pBiDi, &levState, resProp, start1, i);
+                start1=i;
+                break;
+            case 2:             /* init new seq2 */
+                start2=i;
+                break;
+            case 3:             /* process seq1, process seq2, init new seq1 */
+                processPropertySeq(pBiDi, &levState, resProp, start1, start2);
+                processPropertySeq(pBiDi, &levState, DirProp_ON, start2, i);
+                start1=i;
+                break;
+            case 4:             /* process seq1, set seq1=seq2, init new seq2 */
+                processPropertySeq(pBiDi, &levState, resProp, start1, start2);
+                start1=start2;
+                start2=i;
+                break;
+            default:            /* we should never get here */
+                UPRV_UNREACHABLE;
+            }
+        }
+    }
+
+    /* flush possible pending sequence, e.g. ON */
+    if(limit==pBiDi->length && pBiDi->epiLength>0) {
+        DirProp firstStrong=firstL_R_AL_EN_AN(pBiDi);
+        if(firstStrong!=DirProp_ON) {
+            eor=firstStrong;
+        }
+    }
+
+    /* look for the last char not a BN or LRE/RLE/LRO/RLO/PDF */
+    for(i=limit-1; i>start&&(DIRPROP_FLAG(dirProps[i])&MASK_BN_EXPLICIT); i--);
+    dirProp=dirProps[i];
+    if((dirProp==LRI || dirProp==RLI) && limit<pBiDi->length) {
+        pBiDi->isolateCount++;
+        pBiDi->isolates[pBiDi->isolateCount].stateImp=stateImp;
+        pBiDi->isolates[pBiDi->isolateCount].state=levState.state;
+        pBiDi->isolates[pBiDi->isolateCount].start1=start1;
+        pBiDi->isolates[pBiDi->isolateCount].startON=levState.startON;
+    }
+    else
+        processPropertySeq(pBiDi, &levState, eor, limit, limit);
+}
+
+/* perform (L1) and (X9) ---------------------------------------------------- */
+
+/*
+ * Reset the embedding levels for some non-graphic characters (L1).
+ * This function also sets appropriate levels for BN, and
+ * explicit embedding types that are supposed to have been removed
+ * from the paragraph in (X9).
+ */
+static void
+adjustWSLevels(UBiDi *pBiDi) {
+    const DirProp *dirProps=pBiDi->dirProps;
+    UBiDiLevel *levels=pBiDi->levels;
+    int32_t i;
+
+    if(pBiDi->flags&MASK_WS) {
+        UBool orderParagraphsLTR=pBiDi->orderParagraphsLTR;
+        Flags flag;
+
+        i=pBiDi->trailingWSStart;
+        while(i>0) {
+            /* reset a sequence of WS/BN before eop and B/S to the paragraph paraLevel */
+            while(i>0 && (flag=DIRPROP_FLAG(dirProps[--i]))&MASK_WS) {
+                if(orderParagraphsLTR&&(flag&DIRPROP_FLAG(B))) {
+                    levels[i]=0;
+                } else {
+                    levels[i]=GET_PARALEVEL(pBiDi, i);
+                }
+            }
+
+            /* reset BN to the next character's paraLevel until B/S, which restarts above loop */
+            /* here, i+1 is guaranteed to be <length */
+            while(i>0) {
+                flag=DIRPROP_FLAG(dirProps[--i]);
+                if(flag&MASK_BN_EXPLICIT) {
+                    levels[i]=levels[i+1];
+                } else if(orderParagraphsLTR&&(flag&DIRPROP_FLAG(B))) {
+                    levels[i]=0;
+                    break;
+                } else if(flag&MASK_B_S) {
+                    levels[i]=GET_PARALEVEL(pBiDi, i);
+                    break;
+                }
+            }
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_setContext(UBiDi *pBiDi,
+                 const UChar *prologue, int32_t proLength,
+                 const UChar *epilogue, int32_t epiLength,
+                 UErrorCode *pErrorCode) {
+    /* check the argument values */
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    if(pBiDi==NULL || proLength<-1 || epiLength<-1 ||
+       (prologue==NULL && proLength!=0) || (epilogue==NULL && epiLength!=0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if(proLength==-1) {
+        pBiDi->proLength=u_strlen(prologue);
+    } else {
+        pBiDi->proLength=proLength;
+    }
+    if(epiLength==-1) {
+        pBiDi->epiLength=u_strlen(epilogue);
+    } else {
+        pBiDi->epiLength=epiLength;
+    }
+    pBiDi->prologue=prologue;
+    pBiDi->epilogue=epilogue;
+}
+
+static void
+setParaSuccess(UBiDi *pBiDi) {
+    pBiDi->proLength=0;                 /* forget the last context */
+    pBiDi->epiLength=0;
+    pBiDi->pParaBiDi=pBiDi;             /* mark successful setPara */
+}
+
+#define BIDI_MIN(x, y)   ((x)<(y) ? (x) : (y))
+#define BIDI_ABS(x)      ((x)>=0  ? (x) : (-(x)))
+
+static void
+setParaRunsOnly(UBiDi *pBiDi, const UChar *text, int32_t length,
+                UBiDiLevel paraLevel, UErrorCode *pErrorCode) {
+    int32_t *runsOnlyMemory = NULL;
+    int32_t *visualMap;
+    UChar *visualText;
+    int32_t saveLength, saveTrailingWSStart;
+    const UBiDiLevel *levels;
+    UBiDiLevel *saveLevels;
+    UBiDiDirection saveDirection;
+    UBool saveMayAllocateText;
+    Run *runs;
+    int32_t visualLength, i, j, visualStart, logicalStart,
+            runCount, runLength, addedRuns, insertRemove,
+            start, limit, step, indexOddBit, logicalPos,
+            index0, index1;
+    uint32_t saveOptions;
+
+    pBiDi->reorderingMode=UBIDI_REORDER_DEFAULT;
+    if(length==0) {
+        ubidi_setPara(pBiDi, text, length, paraLevel, NULL, pErrorCode);
+        goto cleanup3;
+    }
+    /* obtain memory for mapping table and visual text */
+    runsOnlyMemory=static_cast<int32_t *>(uprv_malloc(length*(sizeof(int32_t)+sizeof(UChar)+sizeof(UBiDiLevel))));
+    if(runsOnlyMemory==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        goto cleanup3;
+    }
+    visualMap=runsOnlyMemory;
+    visualText=(UChar *)&visualMap[length];
+    saveLevels=(UBiDiLevel *)&visualText[length];
+    saveOptions=pBiDi->reorderingOptions;
+    if(saveOptions & UBIDI_OPTION_INSERT_MARKS) {
+        pBiDi->reorderingOptions&=~UBIDI_OPTION_INSERT_MARKS;
+        pBiDi->reorderingOptions|=UBIDI_OPTION_REMOVE_CONTROLS;
+    }
+    paraLevel&=1;                       /* accept only 0 or 1 */
+    ubidi_setPara(pBiDi, text, length, paraLevel, NULL, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        goto cleanup3;
+    }
+    /* we cannot access directly pBiDi->levels since it is not yet set if
+     * direction is not MIXED
+     */
+    levels=ubidi_getLevels(pBiDi, pErrorCode);
+    uprv_memcpy(saveLevels, levels, (size_t)pBiDi->length*sizeof(UBiDiLevel));
+    saveTrailingWSStart=pBiDi->trailingWSStart;
+    saveLength=pBiDi->length;
+    saveDirection=pBiDi->direction;
+
+    /* FOOD FOR THOUGHT: instead of writing the visual text, we could use
+     * the visual map and the dirProps array to drive the second call
+     * to ubidi_setPara (but must make provision for possible removal of
+     * BiDi controls.  Alternatively, only use the dirProps array via
+     * customized classifier callback.
+     */
+    visualLength=ubidi_writeReordered(pBiDi, visualText, length,
+                                      UBIDI_DO_MIRRORING, pErrorCode);
+    ubidi_getVisualMap(pBiDi, visualMap, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        goto cleanup2;
+    }
+    pBiDi->reorderingOptions=saveOptions;
+
+    pBiDi->reorderingMode=UBIDI_REORDER_INVERSE_LIKE_DIRECT;
+    paraLevel^=1;
+    /* Because what we did with reorderingOptions, visualText may be shorter
+     * than the original text. But we don't want the levels memory to be
+     * reallocated shorter than the original length, since we need to restore
+     * the levels as after the first call to ubidi_setpara() before returning.
+     * We will force mayAllocateText to FALSE before the second call to
+     * ubidi_setpara(), and will restore it afterwards.
+     */
+    saveMayAllocateText=pBiDi->mayAllocateText;
+    pBiDi->mayAllocateText=FALSE;
+    ubidi_setPara(pBiDi, visualText, visualLength, paraLevel, NULL, pErrorCode);
+    pBiDi->mayAllocateText=saveMayAllocateText;
+    ubidi_getRuns(pBiDi, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        goto cleanup1;
+    }
+    /* check if some runs must be split, count how many splits */
+    addedRuns=0;
+    runCount=pBiDi->runCount;
+    runs=pBiDi->runs;
+    visualStart=0;
+    for(i=0; i<runCount; i++, visualStart+=runLength) {
+        runLength=runs[i].visualLimit-visualStart;
+        if(runLength<2) {
+            continue;
+        }
+        logicalStart=GET_INDEX(runs[i].logicalStart);
+        for(j=logicalStart+1; j<logicalStart+runLength; j++) {
+            index0=visualMap[j];
+            index1=visualMap[j-1];
+            if((BIDI_ABS(index0-index1)!=1) || (saveLevels[index0]!=saveLevels[index1])) {
+                addedRuns++;
+            }
+        }
+    }
+    if(addedRuns) {
+        if(getRunsMemory(pBiDi, runCount+addedRuns)) {
+            if(runCount==1) {
+                /* because we switch from UBiDi.simpleRuns to UBiDi.runs */
+                pBiDi->runsMemory[0]=runs[0];
+            }
+            runs=pBiDi->runs=pBiDi->runsMemory;
+            pBiDi->runCount+=addedRuns;
+        } else {
+            goto cleanup1;
+        }
+    }
+    /* split runs which are not consecutive in source text */
+    for(i=runCount-1; i>=0; i--) {
+        runLength= i==0 ? runs[0].visualLimit :
+                          runs[i].visualLimit-runs[i-1].visualLimit;
+        logicalStart=runs[i].logicalStart;
+        indexOddBit=GET_ODD_BIT(logicalStart);
+        logicalStart=GET_INDEX(logicalStart);
+        if(runLength<2) {
+            if(addedRuns) {
+                runs[i+addedRuns]=runs[i];
+            }
+            logicalPos=visualMap[logicalStart];
+            runs[i+addedRuns].logicalStart=MAKE_INDEX_ODD_PAIR(logicalPos,
+                                            saveLevels[logicalPos]^indexOddBit);
+            continue;
+        }
+        if(indexOddBit) {
+            start=logicalStart;
+            limit=logicalStart+runLength-1;
+            step=1;
+        } else {
+            start=logicalStart+runLength-1;
+            limit=logicalStart;
+            step=-1;
+        }
+        for(j=start; j!=limit; j+=step) {
+            index0=visualMap[j];
+            index1=visualMap[j+step];
+            if((BIDI_ABS(index0-index1)!=1) || (saveLevels[index0]!=saveLevels[index1])) {
+                logicalPos=BIDI_MIN(visualMap[start], index0);
+                runs[i+addedRuns].logicalStart=MAKE_INDEX_ODD_PAIR(logicalPos,
+                                            saveLevels[logicalPos]^indexOddBit);
+                runs[i+addedRuns].visualLimit=runs[i].visualLimit;
+                runs[i].visualLimit-=BIDI_ABS(j-start)+1;
+                insertRemove=runs[i].insertRemove&(LRM_AFTER|RLM_AFTER);
+                runs[i+addedRuns].insertRemove=insertRemove;
+                runs[i].insertRemove&=~insertRemove;
+                start=j+step;
+                addedRuns--;
+            }
+        }
+        if(addedRuns) {
+            runs[i+addedRuns]=runs[i];
+        }
+        logicalPos=BIDI_MIN(visualMap[start], visualMap[limit]);
+        runs[i+addedRuns].logicalStart=MAKE_INDEX_ODD_PAIR(logicalPos,
+                                            saveLevels[logicalPos]^indexOddBit);
+    }
+
+  cleanup1:
+    /* restore initial paraLevel */
+    pBiDi->paraLevel^=1;
+  cleanup2:
+    /* restore real text */
+    pBiDi->text=text;
+    pBiDi->length=saveLength;
+    pBiDi->originalLength=length;
+    pBiDi->direction=saveDirection;
+    /* the saved levels should never excess levelsSize, but we check anyway */
+    if(saveLength>pBiDi->levelsSize) {
+        saveLength=pBiDi->levelsSize;
+    }
+    uprv_memcpy(pBiDi->levels, saveLevels, (size_t)saveLength*sizeof(UBiDiLevel));
+    pBiDi->trailingWSStart=saveTrailingWSStart;
+    if(pBiDi->runCount>1) {
+        pBiDi->direction=UBIDI_MIXED;
+    }
+  cleanup3:
+    /* free memory for mapping table and visual text */
+    uprv_free(runsOnlyMemory);
+
+    pBiDi->reorderingMode=UBIDI_REORDER_RUNS_ONLY;
+}
+
+/* ubidi_setPara ------------------------------------------------------------ */
+
+U_CAPI void U_EXPORT2
+ubidi_setPara(UBiDi *pBiDi, const UChar *text, int32_t length,
+              UBiDiLevel paraLevel, UBiDiLevel *embeddingLevels,
+              UErrorCode *pErrorCode) {
+    UBiDiDirection direction;
+    DirProp *dirProps;
+
+    /* check the argument values */
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    if(pBiDi==NULL || text==NULL || length<-1 ||
+       (paraLevel>UBIDI_MAX_EXPLICIT_LEVEL && paraLevel<UBIDI_DEFAULT_LTR)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if(length==-1) {
+        length=u_strlen(text);
+    }
+
+    /* special treatment for RUNS_ONLY mode */
+    if(pBiDi->reorderingMode==UBIDI_REORDER_RUNS_ONLY) {
+        setParaRunsOnly(pBiDi, text, length, paraLevel, pErrorCode);
+        return;
+    }
+
+    /* initialize the UBiDi structure */
+    pBiDi->pParaBiDi=NULL;          /* mark unfinished setPara */
+    pBiDi->text=text;
+    pBiDi->length=pBiDi->originalLength=pBiDi->resultLength=length;
+    pBiDi->paraLevel=paraLevel;
+    pBiDi->direction=(UBiDiDirection)(paraLevel&1);
+    pBiDi->paraCount=1;
+
+    pBiDi->dirProps=NULL;
+    pBiDi->levels=NULL;
+    pBiDi->runs=NULL;
+    pBiDi->insertPoints.size=0;         /* clean up from last call */
+    pBiDi->insertPoints.confirmed=0;    /* clean up from last call */
+
+    /*
+     * Save the original paraLevel if contextual; otherwise, set to 0.
+     */
+    pBiDi->defaultParaLevel=IS_DEFAULT_LEVEL(paraLevel);
+
+    if(length==0) {
+        /*
+         * For an empty paragraph, create a UBiDi object with the paraLevel and
+         * the flags and the direction set but without allocating zero-length arrays.
+         * There is nothing more to do.
+         */
+        if(IS_DEFAULT_LEVEL(paraLevel)) {
+            pBiDi->paraLevel&=1;
+            pBiDi->defaultParaLevel=0;
+        }
+        pBiDi->flags=DIRPROP_FLAG_LR(paraLevel);
+        pBiDi->runCount=0;
+        pBiDi->paraCount=0;
+        setParaSuccess(pBiDi);          /* mark successful setPara */
+        return;
+    }
+
+    pBiDi->runCount=-1;
+
+    /* allocate paras memory */
+    if(pBiDi->parasMemory)
+        pBiDi->paras=pBiDi->parasMemory;
+    else
+        pBiDi->paras=pBiDi->simpleParas;
+
+    /*
+     * Get the directional properties,
+     * the flags bit-set, and
+     * determine the paragraph level if necessary.
+     */
+    if(getDirPropsMemory(pBiDi, length)) {
+        pBiDi->dirProps=pBiDi->dirPropsMemory;
+        if(!getDirProps(pBiDi)) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    } else {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    dirProps=pBiDi->dirProps;
+    /* the processed length may have changed if UBIDI_OPTION_STREAMING */
+    length= pBiDi->length;
+    pBiDi->trailingWSStart=length;  /* the levels[] will reflect the WS run */
+
+    /* are explicit levels specified? */
+    if(embeddingLevels==NULL) {
+        /* no: determine explicit levels according to the (Xn) rules */\
+        if(getLevelsMemory(pBiDi, length)) {
+            pBiDi->levels=pBiDi->levelsMemory;
+            direction=resolveExplicitLevels(pBiDi, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                return;
+            }
+        } else {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+    } else {
+        /* set BN for all explicit codes, check that all levels are 0 or paraLevel..UBIDI_MAX_EXPLICIT_LEVEL */
+        pBiDi->levels=embeddingLevels;
+        direction=checkExplicitLevels(pBiDi, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return;
+        }
+    }
+
+    /* allocate isolate memory */
+    if(pBiDi->isolateCount<=SIMPLE_ISOLATES_COUNT)
+        pBiDi->isolates=pBiDi->simpleIsolates;
+    else
+        if((int32_t)(pBiDi->isolateCount*sizeof(Isolate))<=pBiDi->isolatesSize)
+            pBiDi->isolates=pBiDi->isolatesMemory;
+        else {
+            if(getInitialIsolatesMemory(pBiDi, pBiDi->isolateCount)) {
+                pBiDi->isolates=pBiDi->isolatesMemory;
+            } else {
+                *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+        }
+    pBiDi->isolateCount=-1;             /* current isolates stack entry == none */
+
+    /*
+     * The steps after (X9) in the UBiDi algorithm are performed only if
+     * the paragraph text has mixed directionality!
+     */
+    pBiDi->direction=direction;
+    switch(direction) {
+    case UBIDI_LTR:
+        /* all levels are implicitly at paraLevel (important for ubidi_getLevels()) */
+        pBiDi->trailingWSStart=0;
+        break;
+    case UBIDI_RTL:
+        /* all levels are implicitly at paraLevel (important for ubidi_getLevels()) */
+        pBiDi->trailingWSStart=0;
+        break;
+    default:
+        /*
+         *  Choose the right implicit state table
+         */
+        switch(pBiDi->reorderingMode) {
+        case UBIDI_REORDER_DEFAULT:
+            pBiDi->pImpTabPair=&impTab_DEFAULT;
+            break;
+        case UBIDI_REORDER_NUMBERS_SPECIAL:
+            pBiDi->pImpTabPair=&impTab_NUMBERS_SPECIAL;
+            break;
+        case UBIDI_REORDER_GROUP_NUMBERS_WITH_R:
+            pBiDi->pImpTabPair=&impTab_GROUP_NUMBERS_WITH_R;
+            break;
+        case UBIDI_REORDER_INVERSE_NUMBERS_AS_L:
+            pBiDi->pImpTabPair=&impTab_INVERSE_NUMBERS_AS_L;
+            break;
+        case UBIDI_REORDER_INVERSE_LIKE_DIRECT:
+            if (pBiDi->reorderingOptions & UBIDI_OPTION_INSERT_MARKS) {
+                pBiDi->pImpTabPair=&impTab_INVERSE_LIKE_DIRECT_WITH_MARKS;
+            } else {
+                pBiDi->pImpTabPair=&impTab_INVERSE_LIKE_DIRECT;
+            }
+            break;
+        case UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL:
+            if (pBiDi->reorderingOptions & UBIDI_OPTION_INSERT_MARKS) {
+                pBiDi->pImpTabPair=&impTab_INVERSE_FOR_NUMBERS_SPECIAL_WITH_MARKS;
+            } else {
+                pBiDi->pImpTabPair=&impTab_INVERSE_FOR_NUMBERS_SPECIAL;
+            }
+            break;
+        default:
+            /* we should never get here */
+            UPRV_UNREACHABLE;
+        }
+        /*
+         * If there are no external levels specified and there
+         * are no significant explicit level codes in the text,
+         * then we can treat the entire paragraph as one run.
+         * Otherwise, we need to perform the following rules on runs of
+         * the text with the same embedding levels. (X10)
+         * "Significant" explicit level codes are ones that actually
+         * affect non-BN characters.
+         * Examples for "insignificant" ones are empty embeddings
+         * LRE-PDF, LRE-RLE-PDF-PDF, etc.
+         */
+        if(embeddingLevels==NULL && pBiDi->paraCount<=1 &&
+                                   !(pBiDi->flags&DIRPROP_FLAG_MULTI_RUNS)) {
+            resolveImplicitLevels(pBiDi, 0, length,
+                                    GET_LR_FROM_LEVEL(GET_PARALEVEL(pBiDi, 0)),
+                                    GET_LR_FROM_LEVEL(GET_PARALEVEL(pBiDi, length-1)));
+        } else {
+            /* sor, eor: start and end types of same-level-run */
+            UBiDiLevel *levels=pBiDi->levels;
+            int32_t start, limit=0;
+            UBiDiLevel level, nextLevel;
+            DirProp sor, eor;
+
+            /* determine the first sor and set eor to it because of the loop body (sor=eor there) */
+            level=GET_PARALEVEL(pBiDi, 0);
+            nextLevel=levels[0];
+            if(level<nextLevel) {
+                eor=GET_LR_FROM_LEVEL(nextLevel);
+            } else {
+                eor=GET_LR_FROM_LEVEL(level);
+            }
+
+            do {
+                /* determine start and limit of the run (end points just behind the run) */
+
+                /* the values for this run's start are the same as for the previous run's end */
+                start=limit;
+                level=nextLevel;
+                if((start>0) && (dirProps[start-1]==B)) {
+                    /* except if this is a new paragraph, then set sor = para level */
+                    sor=GET_LR_FROM_LEVEL(GET_PARALEVEL(pBiDi, start));
+                } else {
+                    sor=eor;
+                }
+
+                /* search for the limit of this run */
+                while((++limit<length) &&
+                      ((levels[limit]==level) ||
+                       (DIRPROP_FLAG(dirProps[limit])&MASK_BN_EXPLICIT))) {}
+
+                /* get the correct level of the next run */
+                if(limit<length) {
+                    nextLevel=levels[limit];
+                } else {
+                    nextLevel=GET_PARALEVEL(pBiDi, length-1);
+                }
+
+                /* determine eor from max(level, nextLevel); sor is last run's eor */
+                if(NO_OVERRIDE(level)<NO_OVERRIDE(nextLevel)) {
+                    eor=GET_LR_FROM_LEVEL(nextLevel);
+                } else {
+                    eor=GET_LR_FROM_LEVEL(level);
+                }
+
+                /* if the run consists of overridden directional types, then there
+                   are no implicit types to be resolved */
+                if(!(level&UBIDI_LEVEL_OVERRIDE)) {
+                    resolveImplicitLevels(pBiDi, start, limit, sor, eor);
+                } else {
+                    /* remove the UBIDI_LEVEL_OVERRIDE flags */
+                    do {
+                        levels[start++]&=~UBIDI_LEVEL_OVERRIDE;
+                    } while(start<limit);
+                }
+            } while(limit<length);
+        }
+        /* check if we got any memory shortage while adding insert points */
+        if (U_FAILURE(pBiDi->insertPoints.errorCode))
+        {
+            *pErrorCode=pBiDi->insertPoints.errorCode;
+            return;
+        }
+        /* reset the embedding levels for some non-graphic characters (L1), (X9) */
+        adjustWSLevels(pBiDi);
+        break;
+    }
+    /* add RLM for inverse Bidi with contextual orientation resolving
+     * to RTL which would not round-trip otherwise
+     */
+    if((pBiDi->defaultParaLevel>0) &&
+       (pBiDi->reorderingOptions & UBIDI_OPTION_INSERT_MARKS) &&
+       ((pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_LIKE_DIRECT) ||
+        (pBiDi->reorderingMode==UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL))) {
+        int32_t i, j, start, last;
+        UBiDiLevel level;
+        DirProp dirProp;
+        for(i=0; i<pBiDi->paraCount; i++) {
+            last=(pBiDi->paras[i].limit)-1;
+            level= static_cast<UBiDiLevel>(pBiDi->paras[i].level);
+            if(level==0)
+                continue;           /* LTR paragraph */
+            start= i==0 ? 0 : pBiDi->paras[i-1].limit;
+            for(j=last; j>=start; j--) {
+                dirProp=dirProps[j];
+                if(dirProp==L) {
+                    if(j<last) {
+                        while(dirProps[last]==B) {
+                            last--;
+                        }
+                    }
+                    addPoint(pBiDi, last, RLM_BEFORE);
+                    break;
+                }
+                if(DIRPROP_FLAG(dirProp) & MASK_R_AL) {
+                    break;
+                }
+            }
+        }
+    }
+
+    if(pBiDi->reorderingOptions & UBIDI_OPTION_REMOVE_CONTROLS) {
+        pBiDi->resultLength -= pBiDi->controlCount;
+    } else {
+        pBiDi->resultLength += pBiDi->insertPoints.size;
+    }
+    setParaSuccess(pBiDi);              /* mark successful setPara */
+}
+
+U_CAPI void U_EXPORT2
+ubidi_orderParagraphsLTR(UBiDi *pBiDi, UBool orderParagraphsLTR) {
+    if(pBiDi!=NULL) {
+        pBiDi->orderParagraphsLTR=orderParagraphsLTR;
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+ubidi_isOrderParagraphsLTR(UBiDi *pBiDi) {
+    if(pBiDi!=NULL) {
+        return pBiDi->orderParagraphsLTR;
+    } else {
+        return FALSE;
+    }
+}
+
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getDirection(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->direction;
+    } else {
+        return UBIDI_LTR;
+    }
+}
+
+U_CAPI const UChar * U_EXPORT2
+ubidi_getText(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->text;
+    } else {
+        return NULL;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getLength(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->originalLength;
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getProcessedLength(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->length;
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getResultLength(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->resultLength;
+    } else {
+        return 0;
+    }
+}
+
+/* paragraphs API functions ------------------------------------------------- */
+
+U_CAPI UBiDiLevel U_EXPORT2
+ubidi_getParaLevel(const UBiDi *pBiDi) {
+    if(IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return pBiDi->paraLevel;
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_countParagraphs(UBiDi *pBiDi) {
+    if(!IS_VALID_PARA_OR_LINE(pBiDi)) {
+        return 0;
+    } else {
+        return pBiDi->paraCount;
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_getParagraphByIndex(const UBiDi *pBiDi, int32_t paraIndex,
+                          int32_t *pParaStart, int32_t *pParaLimit,
+                          UBiDiLevel *pParaLevel, UErrorCode *pErrorCode) {
+    int32_t paraStart;
+
+    /* check the argument values */
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    RETURN_VOID_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode);
+    RETURN_VOID_IF_BAD_RANGE(paraIndex, 0, pBiDi->paraCount, *pErrorCode);
+
+    pBiDi=pBiDi->pParaBiDi;             /* get Para object if Line object */
+    if(paraIndex) {
+        paraStart=pBiDi->paras[paraIndex-1].limit;
+    } else {
+        paraStart=0;
+    }
+    if(pParaStart!=NULL) {
+        *pParaStart=paraStart;
+    }
+    if(pParaLimit!=NULL) {
+        *pParaLimit=pBiDi->paras[paraIndex].limit;
+    }
+    if(pParaLevel!=NULL) {
+        *pParaLevel=GET_PARALEVEL(pBiDi, paraStart);
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getParagraph(const UBiDi *pBiDi, int32_t charIndex,
+                          int32_t *pParaStart, int32_t *pParaLimit,
+                          UBiDiLevel *pParaLevel, UErrorCode *pErrorCode) {
+    int32_t paraIndex;
+
+    /* check the argument values */
+    /* pErrorCode will be checked by the call to ubidi_getParagraphByIndex */
+    RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrorCode, -1);
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode, -1);
+    pBiDi=pBiDi->pParaBiDi;             /* get Para object if Line object */
+    RETURN_IF_BAD_RANGE(charIndex, 0, pBiDi->length, *pErrorCode, -1);
+
+    for(paraIndex=0; charIndex>=pBiDi->paras[paraIndex].limit; paraIndex++);
+    ubidi_getParagraphByIndex(pBiDi, paraIndex, pParaStart, pParaLimit, pParaLevel, pErrorCode);
+    return paraIndex;
+}
+
+U_CAPI void U_EXPORT2
+ubidi_setClassCallback(UBiDi *pBiDi, UBiDiClassCallback *newFn,
+                       const void *newContext, UBiDiClassCallback **oldFn,
+                       const void **oldContext, UErrorCode *pErrorCode)
+{
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    if(pBiDi==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if( oldFn )
+    {
+        *oldFn = pBiDi->fnClassCallback;
+    }
+    if( oldContext )
+    {
+        *oldContext = pBiDi->coClassCallback;
+    }
+    pBiDi->fnClassCallback = newFn;
+    pBiDi->coClassCallback = newContext;
+}
+
+U_CAPI void U_EXPORT2
+ubidi_getClassCallback(UBiDi *pBiDi, UBiDiClassCallback **fn, const void **context)
+{
+    if(pBiDi==NULL) {
+        return;
+    }
+    if( fn )
+    {
+        *fn = pBiDi->fnClassCallback;
+    }
+    if( context )
+    {
+        *context = pBiDi->coClassCallback;
+    }
+}
+
+U_CAPI UCharDirection U_EXPORT2
+ubidi_getCustomizedClass(UBiDi *pBiDi, UChar32 c)
+{
+    UCharDirection dir;
+
+    if( pBiDi->fnClassCallback == NULL ||
+        (dir = (*pBiDi->fnClassCallback)(pBiDi->coClassCallback, c)) == U_BIDI_CLASS_DEFAULT )
+    {
+        dir = ubidi_getClass(c);
+    }
+    if(dir >= U_CHAR_DIRECTION_COUNT) {
+        dir = (UCharDirection)ON;
+    }
+    return dir;
+}
diff --git a/thirdparty/icu4c/common/ubidi_props.cpp b/thirdparty/icu4c/common/ubidi_props.cpp
new file mode 100644
index 0000000000..afcc4aaf4f
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidi_props.cpp
@@ -0,0 +1,254 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ubidi_props.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004dec30
+*   created by: Markus W. Scherer
+*
+*   Low-level Unicode bidi/shaping properties access.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+#include "unicode/udata.h" /* UDataInfo */
+#include "ucmndata.h" /* DataHeader */
+#include "udatamem.h"
+#include "uassert.h"
+#include "cmemory.h"
+#include "utrie2.h"
+#include "ubidi_props.h"
+#include "ucln_cmn.h"
+
+struct UBiDiProps {
+    UDataMemory *mem;
+    const int32_t *indexes;
+    const uint32_t *mirrors;
+    const uint8_t *jgArray;
+    const uint8_t *jgArray2;
+
+    UTrie2 trie;
+    uint8_t formatVersion[4];
+};
+
+/* ubidi_props_data.h is machine-generated by genbidi --csource */
+#define INCLUDED_FROM_UBIDI_PROPS_C
+#include "ubidi_props_data.h"
+
+/* set of property starts for UnicodeSet ------------------------------------ */
+
+static UBool U_CALLCONV
+_enumPropertyStartsRange(const void *context, UChar32 start, UChar32 end, uint32_t value) {
+    (void)end;
+    (void)value;
+    /* add the start code point to the USet */
+    const USetAdder *sa=(const USetAdder *)context;
+    sa->add(sa->set, start);
+    return TRUE;
+}
+
+U_CFUNC void
+ubidi_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode) {
+    int32_t i, length;
+    UChar32 c, start, limit;
+
+    const uint8_t *jgArray;
+    uint8_t prev, jg;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /* add the start code point of each same-value range of the trie */
+    utrie2_enum(&ubidi_props_singleton.trie, NULL, _enumPropertyStartsRange, sa);
+
+    /* add the code points from the bidi mirroring table */
+    length=ubidi_props_singleton.indexes[UBIDI_IX_MIRROR_LENGTH];
+    for(i=0; i<length; ++i) {
+        c=UBIDI_GET_MIRROR_CODE_POINT(ubidi_props_singleton.mirrors[i]);
+        sa->addRange(sa->set, c, c+1);
+    }
+
+    /* add the code points from the Joining_Group array where the value changes */
+    start=ubidi_props_singleton.indexes[UBIDI_IX_JG_START];
+    limit=ubidi_props_singleton.indexes[UBIDI_IX_JG_LIMIT];
+    jgArray=ubidi_props_singleton.jgArray;
+    for(;;) {
+        prev=0;
+        while(start<limit) {
+            jg=*jgArray++;
+            if(jg!=prev) {
+                sa->add(sa->set, start);
+                prev=jg;
+            }
+            ++start;
+        }
+        if(prev!=0) {
+            /* add the limit code point if the last value was not 0 (it is now start==limit) */
+            sa->add(sa->set, limit);
+        }
+        if(limit==ubidi_props_singleton.indexes[UBIDI_IX_JG_LIMIT]) {
+            /* switch to the second Joining_Group range */
+            start=ubidi_props_singleton.indexes[UBIDI_IX_JG_START2];
+            limit=ubidi_props_singleton.indexes[UBIDI_IX_JG_LIMIT2];
+            jgArray=ubidi_props_singleton.jgArray2;
+        } else {
+            break;
+        }
+    }
+
+    /* add code points with hardcoded properties, plus the ones following them */
+
+    /* (none right now) */
+}
+
+/* property access functions ------------------------------------------------ */
+
+U_CFUNC int32_t
+ubidi_getMaxValue(UProperty which) {
+    int32_t max=ubidi_props_singleton.indexes[UBIDI_MAX_VALUES_INDEX];
+    switch(which) {
+    case UCHAR_BIDI_CLASS:
+        return (max&UBIDI_CLASS_MASK);
+    case UCHAR_JOINING_GROUP:
+        return (max&UBIDI_MAX_JG_MASK)>>UBIDI_MAX_JG_SHIFT;
+    case UCHAR_JOINING_TYPE:
+        return (max&UBIDI_JT_MASK)>>UBIDI_JT_SHIFT;
+    case UCHAR_BIDI_PAIRED_BRACKET_TYPE:
+        return (max&UBIDI_BPT_MASK)>>UBIDI_BPT_SHIFT;
+    default:
+        return -1; /* undefined */
+    }
+}
+
+U_CAPI UCharDirection
+ubidi_getClass(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UCharDirection)UBIDI_GET_CLASS(props);
+}
+
+U_CFUNC UBool
+ubidi_isMirrored(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UBool)UBIDI_GET_FLAG(props, UBIDI_IS_MIRRORED_SHIFT);
+}
+
+static UChar32
+getMirror(UChar32 c, uint16_t props) {
+    int32_t delta=UBIDI_GET_MIRROR_DELTA(props);
+    if(delta!=UBIDI_ESC_MIRROR_DELTA) {
+        return c+delta;
+    } else {
+        /* look for mirror code point in the mirrors[] table */
+        const uint32_t *mirrors;
+        uint32_t m;
+        int32_t i, length;
+        UChar32 c2;
+
+        mirrors=ubidi_props_singleton.mirrors;
+        length=ubidi_props_singleton.indexes[UBIDI_IX_MIRROR_LENGTH];
+
+        /* linear search */
+        for(i=0; i<length; ++i) {
+            m=mirrors[i];
+            c2=UBIDI_GET_MIRROR_CODE_POINT(m);
+            if(c==c2) {
+                /* found c, return its mirror code point using the index in m */
+                return UBIDI_GET_MIRROR_CODE_POINT(mirrors[UBIDI_GET_MIRROR_INDEX(m)]);
+            } else if(c<c2) {
+                break;
+            }
+        }
+
+        /* c not found, return it itself */
+        return c;
+    }
+}
+
+U_CFUNC UChar32
+ubidi_getMirror(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return getMirror(c, props);
+}
+
+U_CFUNC UBool
+ubidi_isBidiControl(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UBool)UBIDI_GET_FLAG(props, UBIDI_BIDI_CONTROL_SHIFT);
+}
+
+U_CFUNC UBool
+ubidi_isJoinControl(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UBool)UBIDI_GET_FLAG(props, UBIDI_JOIN_CONTROL_SHIFT);
+}
+
+U_CFUNC UJoiningType
+ubidi_getJoiningType(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UJoiningType)((props&UBIDI_JT_MASK)>>UBIDI_JT_SHIFT);
+}
+
+U_CFUNC UJoiningGroup
+ubidi_getJoiningGroup(UChar32 c) {
+    UChar32 start, limit;
+
+    start=ubidi_props_singleton.indexes[UBIDI_IX_JG_START];
+    limit=ubidi_props_singleton.indexes[UBIDI_IX_JG_LIMIT];
+    if(start<=c && c<limit) {
+        return (UJoiningGroup)ubidi_props_singleton.jgArray[c-start];
+    }
+    start=ubidi_props_singleton.indexes[UBIDI_IX_JG_START2];
+    limit=ubidi_props_singleton.indexes[UBIDI_IX_JG_LIMIT2];
+    if(start<=c && c<limit) {
+        return (UJoiningGroup)ubidi_props_singleton.jgArray2[c-start];
+    }
+    return U_JG_NO_JOINING_GROUP;
+}
+
+U_CFUNC UBidiPairedBracketType
+ubidi_getPairedBracketType(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    return (UBidiPairedBracketType)((props&UBIDI_BPT_MASK)>>UBIDI_BPT_SHIFT);
+}
+
+U_CFUNC UChar32
+ubidi_getPairedBracket(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ubidi_props_singleton.trie, c);
+    if((props&UBIDI_BPT_MASK)==0) {
+        return c;
+    } else {
+        return getMirror(c, props);
+    }
+}
+
+/* public API (see uchar.h) ------------------------------------------------- */
+
+U_CFUNC UCharDirection
+u_charDirection(UChar32 c) {
+    return ubidi_getClass(c);
+}
+
+U_CFUNC UBool
+u_isMirrored(UChar32 c) {
+    return ubidi_isMirrored(c);
+}
+
+U_CFUNC UChar32
+u_charMirror(UChar32 c) {
+    return ubidi_getMirror(c);
+}
+
+U_CAPI UChar32 U_EXPORT2
+u_getBidiPairedBracket(UChar32 c) {
+    return ubidi_getPairedBracket(c);
+}
diff --git a/thirdparty/icu4c/common/ubidi_props.h b/thirdparty/icu4c/common/ubidi_props.h
new file mode 100644
index 0000000000..698ee9c52b
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidi_props.h
@@ -0,0 +1,148 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ubidi_props.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004dec30
+*   created by: Markus W. Scherer
+*
+*   Low-level Unicode bidi/shaping properties access.
+*/
+
+#ifndef __UBIDI_PROPS_H__
+#define __UBIDI_PROPS_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+#include "putilimp.h"
+#include "uset_imp.h"
+#include "udataswp.h"
+
+U_CDECL_BEGIN
+
+/* library API -------------------------------------------------------------- */
+
+U_CFUNC void
+ubidi_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);
+
+/* property access functions */
+
+U_CFUNC int32_t
+ubidi_getMaxValue(UProperty which);
+
+U_CAPI UCharDirection
+ubidi_getClass(UChar32 c);
+
+U_CFUNC UBool
+ubidi_isMirrored(UChar32 c);
+
+U_CFUNC UChar32
+ubidi_getMirror(UChar32 c);
+
+U_CFUNC UBool
+ubidi_isBidiControl(UChar32 c);
+
+U_CFUNC UBool
+ubidi_isJoinControl(UChar32 c);
+
+U_CFUNC UJoiningType
+ubidi_getJoiningType(UChar32 c);
+
+U_CFUNC UJoiningGroup
+ubidi_getJoiningGroup(UChar32 c);
+
+U_CFUNC UBidiPairedBracketType
+ubidi_getPairedBracketType(UChar32 c);
+
+U_CFUNC UChar32
+ubidi_getPairedBracket(UChar32 c);
+
+/* file definitions --------------------------------------------------------- */
+
+#define UBIDI_DATA_NAME "ubidi"
+#define UBIDI_DATA_TYPE "icu"
+
+/* format "BiDi" */
+#define UBIDI_FMT_0 0x42
+#define UBIDI_FMT_1 0x69
+#define UBIDI_FMT_2 0x44
+#define UBIDI_FMT_3 0x69
+
+/* indexes into indexes[] */
+enum {
+    UBIDI_IX_INDEX_TOP,
+    UBIDI_IX_LENGTH,
+    UBIDI_IX_TRIE_SIZE,
+    UBIDI_IX_MIRROR_LENGTH,
+
+    UBIDI_IX_JG_START,
+    UBIDI_IX_JG_LIMIT,
+    UBIDI_IX_JG_START2,  /* new in format version 2.2, ICU 54 */
+    UBIDI_IX_JG_LIMIT2,
+
+    UBIDI_MAX_VALUES_INDEX=15,
+    UBIDI_IX_TOP=16
+};
+
+/* definitions for 16-bit bidi/shaping properties word ---------------------- */
+
+enum {
+ /* UBIDI_CLASS_SHIFT=0, */     /* bidi class: 5 bits (4..0) */
+    UBIDI_JT_SHIFT=5,           /* joining type: 3 bits (7..5) */
+
+    UBIDI_BPT_SHIFT=8,          /* Bidi_Paired_Bracket_Type(bpt): 2 bits (9..8) */
+
+    UBIDI_JOIN_CONTROL_SHIFT=10,
+    UBIDI_BIDI_CONTROL_SHIFT=11,
+
+    UBIDI_IS_MIRRORED_SHIFT=12,         /* 'is mirrored' */
+    UBIDI_MIRROR_DELTA_SHIFT=13,        /* bidi mirroring delta: 3 bits (15..13) */
+
+    UBIDI_MAX_JG_SHIFT=16               /* max JG value in indexes[UBIDI_MAX_VALUES_INDEX] bits 23..16 */
+};
+
+#define UBIDI_CLASS_MASK        0x0000001f
+#define UBIDI_JT_MASK           0x000000e0
+#define UBIDI_BPT_MASK          0x00000300
+
+#define UBIDI_MAX_JG_MASK       0x00ff0000
+
+#define UBIDI_GET_CLASS(props) ((props)&UBIDI_CLASS_MASK)
+#define UBIDI_GET_FLAG(props, shift) (((props)>>(shift))&1)
+
+#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+#   define UBIDI_GET_MIRROR_DELTA(props) ((int16_t)(props)>>UBIDI_MIRROR_DELTA_SHIFT)
+#else
+#   define UBIDI_GET_MIRROR_DELTA(props) (int16_t)(((props)&0x8000) ? (((props)>>UBIDI_MIRROR_DELTA_SHIFT)|0xe000) : ((props)>>UBIDI_MIRROR_DELTA_SHIFT))
+#endif
+
+enum {
+    UBIDI_ESC_MIRROR_DELTA=-4,
+    UBIDI_MIN_MIRROR_DELTA=-3,
+    UBIDI_MAX_MIRROR_DELTA=3
+};
+
+/* definitions for 32-bit mirror table entry -------------------------------- */
+
+enum {
+    /* the source Unicode code point takes 21 bits (20..0) */
+    UBIDI_MIRROR_INDEX_SHIFT=21,
+    UBIDI_MAX_MIRROR_INDEX=0x7ff
+};
+
+#define UBIDI_GET_MIRROR_CODE_POINT(m) (UChar32)((m)&0x1fffff)
+
+#define UBIDI_GET_MIRROR_INDEX(m) ((m)>>UBIDI_MIRROR_INDEX_SHIFT)
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/ubidi_props_data.h b/thirdparty/icu4c/common/ubidi_props_data.h
new file mode 100644
index 0000000000..7a34870bd8
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidi_props_data.h
@@ -0,0 +1,922 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+// Copyright (C) 1999-2016, International Business Machines
+// Corporation and others.  All Rights Reserved.
+//
+// file name: ubidi_props_data.h
+//
+// machine-generated by: icu/tools/unicode/c/genprops/bidipropsbuilder.cpp
+
+
+#ifdef INCLUDED_FROM_UBIDI_PROPS_C
+
+static const UVersionInfo ubidi_props_dataVersion={0xd,0,0,0};
+
+static const int32_t ubidi_props_indexes[UBIDI_IX_TOP]={0x10,0x67ec,0x6200,0x28,0x620,0x8c8,0x10ac0,0x10d24,0,0,0,0,0,0,0,0x6502b6};
+
+static const uint16_t ubidi_props_trieIndex[12536]={
+0x37c,0x384,0x38c,0x394,0x3ac,0x3b4,0x3bc,0x3c4,0x39c,0x3a4,0x39c,0x3a4,0x39c,0x3a4,0x39c,0x3a4,
+0x39c,0x3a4,0x39c,0x3a4,0x3ca,0x3d2,0x3da,0x3e2,0x3ea,0x3f2,0x3ee,0x3f6,0x3fe,0x406,0x401,0x409,
+0x39c,0x3a4,0x39c,0x3a4,0x411,0x419,0x39c,0x3a4,0x39c,0x3a4,0x39c,0x3a4,0x41f,0x427,0x42f,0x437,
+0x43f,0x447,0x44f,0x457,0x45d,0x465,0x46d,0x475,0x47d,0x485,0x48b,0x493,0x49b,0x4a3,0x4ab,0x4b3,
+0x4bf,0x4bb,0x4c7,0x4cf,0x431,0x4df,0x4e6,0x4d7,0x4ee,0x4f0,0x4f8,0x500,0x508,0x509,0x511,0x519,
+0x521,0x509,0x529,0x52e,0x521,0x509,0x536,0x53e,0x508,0x546,0x54e,0x500,0x556,0x39c,0x55e,0x562,
+0x56a,0x56c,0x574,0x57c,0x508,0x584,0x58c,0x500,0x413,0x590,0x511,0x500,0x508,0x39c,0x598,0x39c,
+0x39c,0x59e,0x5a6,0x39c,0x39c,0x5aa,0x5b2,0x39c,0x5b6,0x5bd,0x39c,0x5c5,0x5cd,0x5d4,0x555,0x39c,
+0x39c,0x5dc,0x5e4,0x5ec,0x5f4,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x5fc,0x39c,0x604,0x39c,0x39c,0x39c,
+0x60c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x614,0x39c,0x39c,0x39c,0x61c,0x61c,0x515,0x515,0x39c,0x622,0x62a,0x604,
+0x640,0x632,0x632,0x648,0x64f,0x638,0x39c,0x39c,0x39c,0x657,0x65f,0x39c,0x39c,0x39c,0x661,0x669,
+0x671,0x39c,0x678,0x680,0x39c,0x688,0x56b,0x39c,0x545,0x690,0x556,0x698,0x413,0x6a0,0x39c,0x6a7,
+0x39c,0x6ac,0x39c,0x39c,0x39c,0x39c,0x6b2,0x6ba,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x3ea,0x6c2,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x6ca,0x6d2,0x6d6,
+0x6ee,0x6f4,0x6de,0x6e6,0x6fc,0x704,0x708,0x5d7,0x710,0x718,0x720,0x39c,0x728,0x669,0x669,0x669,
+0x738,0x740,0x748,0x750,0x755,0x75d,0x765,0x730,0x76d,0x775,0x39c,0x77b,0x782,0x669,0x669,0x669,
+0x669,0x582,0x788,0x669,0x790,0x39c,0x39c,0x666,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,
+0x669,0x669,0x669,0x669,0x669,0x798,0x669,0x669,0x669,0x669,0x669,0x79e,0x669,0x669,0x7a6,0x7ae,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x669,0x669,0x669,0x669,0x7be,0x7c6,0x7ce,0x7b6,
+0x7de,0x7e6,0x7ee,0x7f5,0x7fc,0x804,0x808,0x7d6,0x669,0x669,0x669,0x810,0x816,0x669,0x669,0x81c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x824,0x39c,0x39c,0x39c,0x82c,0x39c,0x39c,0x39c,0x3ea,
+0x834,0x83c,0x840,0x39c,0x848,0x669,0x669,0x66c,0x669,0x669,0x669,0x669,0x669,0x669,0x84f,0x855,
+0x865,0x85d,0x39c,0x39c,0x86d,0x60c,0x39c,0x3c3,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x669,0x82b,
+0x3d1,0x39c,0x875,0x87d,0x39c,0x885,0x88d,0x39c,0x39c,0x39c,0x39c,0x891,0x39c,0x39c,0x661,0x3c2,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x669,0x669,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x875,0x669,0x582,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x898,0x39c,0x39c,0x89d,0x56c,0x39c,0x39c,0x5b8,0x669,0x660,0x39c,0x39c,0x8a5,0x39c,0x39c,0x39c,
+0x8ad,0x8b4,0x632,0x8bc,0x39c,0x39c,0x58e,0x8c4,0x39c,0x8cc,0x8d3,0x39c,0x4ee,0x8d8,0x39c,0x507,
+0x39c,0x8e0,0x8e8,0x509,0x39c,0x8ec,0x508,0x8f4,0x39c,0x39c,0x39c,0x8fa,0x39c,0x39c,0x39c,0x901,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x915,0x909,0x90d,0x49b,0x49b,0x49b,0x49b,0x49b,
+0x49b,0x49b,0x49b,0x49b,0x49b,0x49b,0x49b,0x49b,0x49b,0x91d,0x49b,0x49b,0x49b,0x49b,0x925,0x929,
+0x931,0x939,0x93d,0x945,0x49b,0x49b,0x49b,0x949,0x951,0x38c,0x959,0x961,0x39c,0x39c,0x39c,0x969,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0xe70,0xe70,0xeb0,0xef0,0xe70,0xe70,0xe70,0xe70,0xe70,0xe70,0xf28,0xf68,0xfa8,0xfb8,0xff8,0x1004,
+0xe70,0xe70,0x1044,0xe70,0xe70,0xe70,0x107c,0x10bc,0x10fc,0x113c,0x1174,0x11b4,0x11f4,0x122c,0x126c,0x12ac,
+0xa40,0xa80,0xac0,0xaff,0x1a0,0x1a0,0xb3f,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xb68,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xba8,0x1a0,0x1a0,0xbdd,0xc1d,0xc5d,0xc9d,0xcdd,0xd1d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0xd9d,0xdad,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,
+0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0x1a0,0xd5d,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x971,0x39c,0x669,0x669,0x979,0x60c,0x39c,0x501,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x981,0x39c,0x39c,0x39c,0x988,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x990,0x431,0x431,0x431,0x431,0x431,0x431,0x431,
+0x998,0x99c,0x431,0x431,0x431,0x431,0x9ac,0x9a4,0x431,0x9b4,0x431,0x431,0x9bc,0x9c2,0x431,0x431,
+0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x9d2,0x9ca,0x431,0x431,0x431,0x431,0x431,0x431,
+0x431,0x431,0x431,0x9da,0x431,0x9e2,0x431,0x431,0x431,0x9e6,0x9ed,0x9f3,0x431,0x9f7,0x9ff,0x431,
+0x508,0xa07,0xa0e,0xa15,0x413,0xa18,0x39c,0x39c,0x4ee,0xa1f,0x39c,0xa25,0x413,0xa2a,0xa32,0x39c,
+0x39c,0xa37,0x39c,0x39c,0x39c,0x39c,0x82c,0xa3f,0x413,0x590,0x56b,0xa46,0x39c,0x39c,0x39c,0x39c,
+0x39c,0xa07,0xa4e,0x39c,0x39c,0xa56,0xa5e,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xa62,0xa6a,0x39c,
+0x39c,0xa72,0x56b,0xa7a,0x39c,0xa80,0x39c,0x39c,0x5fc,0xa88,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0xa8d,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xa94,0xa9c,0x39c,0x39c,0x39c,0xa9f,0x56b,0xaa7,
+0xaab,0xab3,0x39c,0xaba,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0xac1,0x39c,0x39c,0xacf,0xac9,0x39c,0x39c,0x39c,0xad7,0xadf,0x39c,0xae3,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x592,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xaf0,0xaeb,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0xaf8,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xaff,
+0x39c,0xb05,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0xa26,0x39c,0xb0b,0x39c,0x39c,0xb13,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x522,0xb1b,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xb22,0xb2a,0xb30,0x39c,0x39c,0x669,0x669,0xb38,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x669,0x669,0x83f,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0xb3a,0x39c,0xb41,0x39c,0xb3d,0x39c,0xb44,0x39c,0xb4c,0xb50,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x3ea,0xb58,0x3ea,
+0xb5f,0xb66,0xb6e,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xb76,0xb7e,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xb05,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0xb83,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x431,0x431,0x431,
+0x431,0x431,0x431,0xb8b,0x431,0xb93,0xb93,0xb9a,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,
+0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,
+0x90d,0x49b,0x49b,0x431,0x431,0x49b,0x49b,0x9f3,0x431,0x431,0x431,0x431,0x431,0x49b,0x49b,0x49b,
+0x49b,0x49b,0x49b,0x49b,0xba2,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x431,0x669,0xbaa,0x669,
+0x669,0x66c,0xbaf,0xbb3,0x84f,0xbbb,0x3be,0x39c,0xbc1,0x39c,0xbc6,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x779,0x39c,0x39c,0x39c,0x39c,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,
+0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,0x669,
+0x669,0x669,0x669,0x66b,0x979,0x669,0x669,0x669,0x66c,0x669,0x669,0xbce,0x66e,0xbaa,0x669,0xbd6,
+0x669,0xbde,0xbe3,0x39c,0x39c,0x669,0x669,0x669,0xbeb,0x669,0x669,0x798,0x669,0x669,0x669,0x66c,
+0xbf2,0xbfa,0xc00,0xc05,0x39c,0x669,0x669,0x669,0x669,0xc0d,0x669,0x788,0xc15,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xc1c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,
+0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0x39c,0xc1c,0xc2c,0xc24,0xc24,
+0xc24,0xc2d,0xc2d,0xc2d,0xc2d,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0xc35,0xc2d,0xc2d,0xc2d,
+0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,
+0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,
+0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,
+0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0xc2d,0x37b,0x37b,0x37b,
+0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,8,7,8,9,7,0x12,0x12,
+0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,7,7,7,8,
+9,0xa,0xa,4,4,4,0xa,0xa,0x310a,0xf20a,0xa,3,6,3,6,6,
+2,2,2,2,2,2,2,2,2,2,6,0xa,0x500a,0xa,0xd00a,0xa,
+0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0x510a,0xa,0xd20a,0xa,0xa,
+0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0x510a,0xa,0xd20a,0xa,0x12,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x12,0x12,0x12,0x12,0x12,7,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+6,0xa,4,4,4,4,0xa,0xa,0xa,0xa,0,0x900a,0xa,0xb2,0xa,0xa,
+4,4,2,2,0xa,0,0xa,0xa,0xa,2,0,0x900a,0xa,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0xa,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0xa,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0xa,0xa,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0xa,0xa,0,0,
+0,0,0,0,0,0,0xa,0,0,0,0,0,0xa,0xa,0,0xa,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xa,0,0,0,0,0,
+0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xa,0,0,0xa,0xa,4,1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,1,0xb1,1,0xb1,0xb1,1,
+0xb1,0xb1,1,0xb1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,5,5,5,5,
+5,5,0xa,0xa,0xd,4,4,0xd,6,0xd,0xa,0xa,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xd,0x8ad,0xd,0xd,0xd,0x4d,0xd,0x8d,0x8d,
+0x8d,0x8d,0x4d,0x8d,0x4d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x8d,0x8d,0x8d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x2d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x8d,0x4d,0x4d,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,5,5,5,5,
+5,5,5,5,5,5,4,5,5,0xd,0x4d,0x4d,0xb1,0x8d,0x8d,0x8d,
+0xd,0x8d,0x8d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x8d,0x8d,0x8d,
+0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x8d,0x4d,0x4d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,
+0x4d,0x8d,0x4d,0x8d,0x4d,0x4d,0x8d,0x8d,0xd,0x8d,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,5,0xa,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xd,0xd,0xb1,0xb1,0xa,0xb1,0xb1,
+0xb1,0xb1,0x8d,0x8d,2,2,2,2,2,2,2,2,2,2,0x4d,0x4d,
+0x4d,0xd,0xd,0x4d,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xad,0x8d,0xb1,0x4d,0x4d,0x4d,0x8d,0x8d,0x8d,0x8d,0x8d,0x4d,0x4d,
+0x4d,0x4d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x4d,0x8d,0x4d,
+0x8d,0x4d,0x4d,0x8d,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xd,0xd,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x8d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x8d,0x4d,0x4d,0x4d,0x4d,0x8d,0x4d,0x8d,
+0x8d,0x4d,0x4d,0x4d,0x8d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,1,1,1,1,
+1,1,1,1,1,1,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,
+0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,
+0x41,0x41,0x41,0x41,0x41,0x41,0x41,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+1,1,0xa,0xa,0xa,0xa,0x21,1,1,0xb1,1,1,0xb1,0xb1,0xb1,0xb1,
+1,0xb1,0xb1,0xb1,1,0xb1,0xb1,0xb1,0xb1,0xb1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0xb1,0xb1,0xb1,0xb1,1,0xb1,0xb1,0xb1,0xb1,0xb1,0x81,0x41,0x41,0x41,
+0x41,0x41,0x81,0x81,0x41,0x81,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,
+0x81,0x41,0x81,0x81,0x81,0xb1,0xb1,0xb1,1,1,1,1,0x4d,0xd,0x4d,0x4d,
+0x4d,0x4d,0xd,0x8d,0x4d,0x8d,0x8d,0xd,0xd,0xd,0xd,0xd,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0xb1,0xb1,5,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x8d,0x8d,0x8d,0xd,0x8d,0x4d,0x4d,0x8d,0x8d,0x4d,
+0x4d,0xd,0x4d,0x4d,0x4d,0x8d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0,0xb1,0,0,0,
+0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0xb1,0,0,
+0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,
+0,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,4,0,0,0,0,0,0,0,4,
+0,0,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xb1,0xb1,0,0,0,0,0xb1,0xb1,0,0,0xb1,
+0xb1,0xb1,0,0,0,0xb1,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,0,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,
+0xb1,0,0,0,0,0xb1,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0,0,0xb1,0,0xb1,0xb1,0xb1,0xb1,0,0,0,
+0,0,0,0,0,0xb1,0,0,0,0,0,0,0,0xb1,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0,0,0,0,0,0,0,
+0,0,0,0,0,0xb1,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,
+0xa,4,0xa,0,0,0,0,0,0xb1,0,0,0,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,
+0xb1,0,0,0,0,0,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0,0,0xa0,
+0,0,0,0,0,0,0xa0,0,0,0,0,0,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0,0,0,4,0,0,0,0,0,0,0,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0,0xb1,0,0xb1,0x310a,0xf20a,
+0x310a,0xf20a,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0,0,0,0,
+0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,
+0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0,0,0xb1,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,0,0,0xb1,0xb1,
+0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,
+0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+9,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0x310a,0xf20a,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0xb1,0,
+0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,
+0,0,0,4,0,0xb1,0,0,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0xb1,0x40,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x4a,0xa,0xa,0x2a,0xb1,0xb1,0xb1,0x12,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0,0,0,0,0,0,0,
+0,0xb1,0xb1,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0xb1,0xb1,0xb1,0,
+0,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0xb1,0,
+0,0,0,0,0,0xb1,0xb1,0xb1,0,0,0,0,0xa,0,0,0,
+0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,
+0xb1,0,0xb1,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0xb1,0,0xb1,0xb1,0,0,0,0xb1,0,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0,0,0,0,0xb1,0,0,0,0,0,0,0xb1,0,0,0,
+0xb1,0xb1,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0,0xa,0xa,0xa,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0xa,0xa,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0xa,0xa,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0xa,0,0xa,0xa,0xa,0xa,6,0x310a,0xf20a,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,9,0xb2,0xb2,0xb2,0xb2,0xb2,0x12,0x814,0x815,
+0x813,0x816,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,2,0,0,0,2,2,2,2,
+2,2,3,3,0xa,0x310a,0xf20a,0,9,9,9,9,9,9,9,9,
+9,9,9,0xb2,0x412,0x432,0x8a0,0x8a1,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,9,7,0x8ab,0x8ae,0x8b0,0x8ac,0x8af,6,
+4,4,4,4,4,0xa,0xa,0xa,0xa,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,
+2,2,2,2,2,2,2,2,2,2,3,3,0xa,0x310a,0xf20a,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xa,0xa,0,0xa,0xa,0xa,0xa,0,0xa,0xa,0,0,0,0,0,0,
+0,0,0,0,0xa,0,0xa,0xa,0xa,0,0,0,0,0,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0xa,0,0xa,0,0xa,0,0,0,0,4,0,
+0,0,0,0,0,0,0,0,0,0,0xa,0xa,0,0,0,0,
+0x100a,0xa,0xa,0xa,0xa,0,0,0,0,0,0xa,0xa,0xa,0xa,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,
+0x300a,0xf00a,0x900a,0x900a,0x900a,0x100a,0x900a,0x900a,0x100a,0x100a,0x900a,0x900a,0x900a,0x900a,0x900a,0x100a,
+0xa,0x100a,0x100a,0x100a,0x100a,0xa,0xa,0xa,0x700a,0x700a,0x700a,0xb00a,0xb00a,0xb00a,0xa,0xa,
+0xa,0x100a,3,4,0xa,0x900a,0x100a,0xa,0xa,0xa,0x100a,0x100a,0x100a,0x100a,0xa,0x900a,
+0x900a,0x900a,0x900a,0xa,0x900a,0xa,0x100a,0xa,0xa,0xa,0xa,0x100a,0x100a,0x100a,0x100a,0x100a,
+0x100a,0x100a,0x100a,0x100a,0xa,0xa,0xa,0xa,0xa,0x100a,0xa,0x100a,0x300a,0xf00a,0x100a,0x100a,
+0x100a,0x100a,0x100a,0x900a,0x100a,0x900a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x900a,0xa,0xa,0xa,
+0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x100a,
+0x100a,0xa,0x100a,0xa,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0x300a,0xf00a,
+0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,
+0x100a,0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0x900a,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0xa,0xa,0x900a,0x100a,0x900a,0x900a,0x100a,0x900a,
+0x100a,0x100a,0x100a,0x100a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x900a,0xa,0xa,0xa,
+0xa,0xa,0x100a,0x100a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0x300a,
+0xf00a,0x900a,0xa,0xa,0x300a,0xf00a,0xa,0xa,0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,
+0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x310a,0xf20a,0x310a,0xf20a,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0x100a,0x100a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x310a,0xf20a,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x100a,0xa,0xa,0x300a,0xf00a,0x310a,0xf20a,0xa,
+0x300a,0xf00a,0xa,0x500a,0x100a,0xd00a,0xa,0xa,0xa,0xa,0xa,0x100a,0x100a,0x300a,0xf00a,0xa,
+0xa,0xa,0xa,0xa,0x900a,0x300a,0xf00a,0xa,0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0x310a,0xf20a,
+0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x100a,0xa,0x100a,0x100a,0x100a,0xa,0xa,
+0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x100a,0x900a,0x100a,0x100a,
+0x300a,0xf00a,0xa,0xa,0x310a,0xf20a,0xa,0xa,0xa,0xa,0xa,0x310a,0xf20a,0x310a,0xf20a,0x310a,
+0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x710a,0x320a,0xf10a,0xb20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,
+0xf20a,0xa,0xa,0x900a,0x100a,0x100a,0x100a,0x100a,0x900a,0xa,0x100a,0x900a,0x300a,0xf00a,0x100a,0x100a,
+0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0x900a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0x100a,0x100a,0x300a,0xf00a,0xa,0xa,
+0xa,0x100a,0xa,0xa,0xa,0xa,0x100a,0x300a,0xf00a,0x300a,0xf00a,0xa,0x300a,0xf00a,0xa,0xa,
+0x310a,0xf20a,0x310a,0xf20a,0x100a,0xa,0xa,0xa,0xa,0xa,0x100a,0x900a,0x900a,0x900a,0x100a,0xa,
+0xa,0xa,0xa,0xa,0x300a,0xf00a,0x900a,0xa,0xa,0xa,0xa,0x100a,0xa,0xa,0xa,0x300a,
+0xf00a,0x300a,0xf00a,0x100a,0xa,0x100a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,0x100a,
+0x100a,0x100a,0x100a,0x100a,0x100a,0xa,0x100a,0x100a,0x100a,0x100a,0xa,0xa,0x100a,0xa,0x100a,0xa,
+0xa,0x100a,0xa,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0xa,0xa,
+0xa,0xa,0xa,0xa,0x300a,0xf00a,0x100a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x100a,
+0x100a,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0xa,0xa,0xa,0xa,0x100a,0x100a,
+0x100a,0x100a,0xa,0x100a,0x100a,0xa,0xa,0x100a,0x100a,0xa,0xa,0xa,0xa,0x300a,0xf00a,0x300a,
+0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,
+0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,
+0xf00a,0x300a,0xf00a,0x100a,0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0xa,0x300a,
+0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,
+0xf00a,0x300a,0xf00a,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0xa,0xa,0x100a,0xa,0x900a,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x900a,0xa,
+0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0xb1,
+0xb1,0xb1,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,
+0xa,0xa,0x300a,0xf00a,0x300a,0xf00a,0xa,0xa,0xa,0x300a,0xf00a,0xa,0x300a,0xf00a,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0x300a,0xf00a,0xa,0xa,
+0x300a,0xf00a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0,0,0,0,0xa,0,0,0,0,0,0,0,0,0,0xb1,0xb1,
+0xb1,0xb1,0,0,0xa,0,0,0,0,0,0xa,0xa,0,0,0,0,
+0,0xa,0xa,0xa,9,0xa,0xa,0xa,0xa,0,0,0,0x310a,0xf20a,0x310a,0xf20a,
+0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0xa,0xa,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,0x310a,0xf20a,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xa,
+0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xa,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xa,0xa,0,0,0,0,0,0,0,0,0xa,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xb1,0,0,0,0xb1,0,0,0,0,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xb1,0xb1,0,0xa,0xa,0xa,0xa,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,4,4,0,0,0,0,0,0,
+0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,0x40,
+0x40,0x40,0x60,0,0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,
+0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0,0,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,
+0,0xb1,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0,0,0,0,0,0,0,0,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0xb1,0xb1,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0xb1,0,0,0xb1,0,0,0,
+0,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,1,1,1,1,1,1,1,1,1,3,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,1,0xb1,1,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xa,0xa,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xa,0xd,0xd,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,
+0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,6,0xa,6,0,0xa,6,0xa,0xa,0xa,0x310a,0xf20a,0x310a,
+0xf20a,0x310a,0xf20a,4,0xa,0xa,3,3,0x300a,0xf00a,0xa,0,0xa,4,4,0xa,
+0,0,0,0,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xb2,0,0xa,0xa,4,4,4,0xa,0xa,0x310a,0xf20a,0xa,3,
+6,3,6,6,2,2,2,2,2,2,2,2,2,2,6,0xa,
+0x500a,0xa,0xd00a,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x510a,
+0xa,0xd20a,0xa,0x310a,0xf20a,0xa,0x310a,0xf20a,0xa,0xa,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,0xa,0xa,0xa,4,4,0,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0xaa,0xaa,0xaa,
+0xa,0xa,0x12,0x12,0,0xa,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0,0,0,0xb1,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0xa,
+1,0xb1,0xb1,0xb1,1,0xb1,0xb1,1,1,1,1,1,0xb1,0xb1,0xb1,0xb1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xb1,0xb1,0xb1,1,1,1,1,0xb1,
+0x41,0x81,1,1,0x81,0xb1,0xb1,1,1,1,1,0x41,0x41,0x41,0x41,0x81,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+0x41,0x41,0x41,0x41,0x41,0x81,1,0x81,1,0x81,0x81,1,1,0x61,0x81,0x81,
+0x81,0x81,0x81,0x41,0x41,0x41,0x41,0x61,0x41,0x41,0x41,0x41,0x41,0x81,0x41,0x41,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0x41,0x81,0x41,0x81,0x81,0x81,0x41,0x41,0x41,0x81,0x41,0x41,0x81,0x41,0x81,0x81,
+0x41,0x81,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,0x81,0x81,0x81,0x81,0x41,0x41,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0x4d,0x4d,0x8d,0x4d,0xb1,0xb1,0xb1,0xb1,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,5,5,5,5,5,5,5,5,
+5,5,0xd,0xd,0xd,0xd,0xd,0xd,0x6d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,1,1,1,1,1,1,1,1,1,
+1,1,1,0xb1,0xb1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0x4d,0x4d,0x4d,0x8d,0x4d,0x4d,0x4d,0x4d,
+0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0x4d,0xd,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0x4d,0x4d,0x4d,0x8d,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0x41,1,0x41,0x41,
+0x81,0x81,0x81,1,0x41,0x81,0x81,0x41,0x41,0x81,0x41,0x41,1,0x41,0x81,0x81,
+0x41,1,1,1,1,0x81,0x41,0x61,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0,0,
+0xb1,0,0xb1,0xb1,0,0,0,0,0,0,0xb1,0,0,0,0,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0,0xb1,0,0,0,0,0xb1,0xb1,0,0xb1,0xb1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0xb1,0xb1,0,0xb1,0xb1,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0,0xb1,0,0xb1,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0,0xb1,0,0,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0xb1,0,
+0,0,0,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0,0,0,0,0,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xa0,0xa0,0xb1,0xb1,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,
+0,0,0,0xb1,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,
+0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0,0xa0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0,0xb1,0xb1,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0,0,0xb1,0,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0,0,
+0,0xb1,0,0xb1,0,0,0,0,0,0,0,0,4,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,4,4,4,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa0,0xa0,0xa0,0xa0,0xa0,0xa0,0xa0,0xa0,0xa0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,
+0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xa,0,
+0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xb2,0xb2,0xb2,0xb2,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xb1,
+0xb1,0xb1,0,0,0,0,0,0,0,0,0,0xb2,0xb2,0xb2,0xb2,0xb2,
+0xb2,0xb2,0xb2,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xa,0xa,0xb1,0xb1,0xb1,0xa,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x100a,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0x100a,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x100a,0,0,0,0,0,0,0,0,0,0,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,
+0,0xb1,0,0,0,0,0,0,0,0,0,0,0xb1,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,
+0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0,0,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0,0xb1,0xb1,0,0xb1,0xb1,
+0xb1,0xb1,0xb1,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xb1,0xb1,0xb1,0xb1,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,4,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,1,1,1,1,1,1,1,1,1,0x41,0x41,0x41,0x41,
+0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,
+0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0x41,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xa1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xa,0xa,0xd,0xd,0xd,0xd,0xd,0xd,
+0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xd,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,2,2,2,2,
+2,2,2,2,2,2,2,0xa,0xa,0xa,0xa,0xa,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0xa,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0xa,0xa,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0,0,0xa,0xa,0xa,0xa,0xa,0,0,0,
+0xa,0xa,0xa,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,
+0,0,0,0,0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0,0,0,0,
+0,0,0,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0xa,0xa,0xa,0,0xa,0xa,0xa,0xa,0xa,0xa,0xa,0xa,
+0xa,0xa,0xa,0xa,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,2,2,2,2,2,2,2,2,2,2,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x12,0x12,
+0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,
+0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,0xb2,
+0x12,0xb2,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0,0,0,0
+};
+
+static const uint32_t ubidi_props_mirrors[40]={
+0x2000ab,0xbb,0x4202215,0x4e0221f,0x3e02220,0x3a02221,0x3c02222,0x4c02224,0x2202243,0x1402245,0x120224c,0x4002298,0x44022a6,0x48022a8,0x46022a9,0x4a022ab,
+0x38022b8,0x10022cd,0x2e022f2,0x30022f3,0x32022f4,0x34022f6,0x36022f7,0x24022fa,0x26022fb,0x28022fc,0x2a022fd,0x2c022fe,0x20027dc,0xa0299b,0xc029a0,0x8029a3,
+0x16029b8,0x4029f5,0x1802ade,0x1c02ae3,0x1a02ae4,0x1e02ae5,0xe02aee,0x602bfe
+};
+
+static const uint8_t ubidi_props_jgArray[680]={
+0x2d,0,3,3,0x2c,3,0x2d,3,4,0x2a,4,4,0xd,0xd,0xd,6,
+6,0x1f,0x1f,0x23,0x23,0x21,0x21,0x28,0x28,1,1,0xb,0xb,0x37,0x37,0x37,
+0,9,0x1d,0x13,0x16,0x18,0x1a,0x10,0x2c,0x2d,0x2d,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,0x1d,
+0,3,3,3,0,3,0x2c,0x2c,0x2d,4,4,4,4,4,4,4,
+4,0xd,0xd,0xd,0xd,0xd,0xd,0xd,6,6,6,6,6,6,6,6,
+6,0x1f,0x1f,0x1f,0x1f,0x1f,0x1f,0x1f,0x1f,0x1f,0x23,0x23,0x23,0x21,0x21,0x28,
+1,9,9,9,9,9,9,0x1d,0x1d,0xb,0x26,0xb,0x13,0x13,0x13,0xb,
+0xb,0xb,0xb,0xb,0xb,0x16,0x16,0x16,0x16,0x1a,0x1a,0x1a,0x1a,0x38,0x15,0xd,
+0x2a,0x11,0x11,0xe,0x2c,0x2c,0x2c,0x2c,0x2c,0x2c,0x2c,0x2c,0x37,0x2f,0x37,0x2c,
+0x2d,0x2d,0x2e,0x2e,0,0x2a,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,6,0x1f,
+0,0,0,0,0,0,0,0,0,0,0x23,0x21,1,0,0,0x15,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+2,0,5,0xc,0xc,7,7,0xf,0x27,0x32,0x12,0x2b,0x2b,0x30,0x31,0x14,
+0x17,0x19,0x1b,0x24,0xa,8,0x1c,0x20,0x22,0x1e,7,0x25,0x29,5,0xc,7,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0x35,0x34,0x33,
+4,4,4,4,4,4,4,0xd,0xd,6,6,0x1f,0x23,1,1,1,
+9,9,0xb,0xb,0xb,0x18,0x18,0x1a,0x1a,0x1a,0x16,0x1f,0x1f,0x23,0xd,0xd,
+0x23,0x1f,0xd,3,3,0x37,0x37,0x2d,0x2c,0x2c,0x36,0x36,0xd,0x23,0x23,0x13,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x5d,0x5a,0x60,0x63,0x5e,0x5f,0x59,0x61,0x5b,0x5c,0x62,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,0xd,0x28,9,0x1d,0x16,0x18,0x2d,0x2d,0x1f,0x2c,0x39,0,6,0x21,
+0xb,0x55,0x1f,1,0x13,0,4,4,4,0x1f,0x2d,0x56,0x58,0x57,4,4,
+4,0xd,0xb,1,0x58,0xd,0xd,0x16
+};
+
+static const uint8_t ubidi_props_jgArray2[612]={
+0x3a,0x3c,0x3c,0x40,0x40,0x3d,0,0x52,0,0x54,0x54,0,0,0x41,0x4f,0x53,
+0x43,0x43,0x43,0x44,0x3e,0x50,0x45,0x46,0x4c,0x3b,0x3b,0x48,0x48,0x4b,0x49,0x49,
+0x49,0x4a,0,0,0x4d,0,0,0,0,0,0,0x47,0x3f,0x4e,0x51,0x42,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x65,0,0,0,0,0,0,0x65,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0x64,0,0,0x65,0,0x64,0,
+0x64,0,0,0x64
+};
+
+static const UBiDiProps ubidi_props_singleton={
+  NULL,
+  ubidi_props_indexes,
+  ubidi_props_mirrors,
+  ubidi_props_jgArray,
+  ubidi_props_jgArray2,
+  {
+    ubidi_props_trieIndex,
+    ubidi_props_trieIndex+3568,
+    NULL,
+    3568,
+    8968,
+    0x1a0,
+    0xe70,
+    0x0,
+    0x0,
+    0x110000,
+    0x30f4,
+    NULL, 0, FALSE, FALSE, 0, NULL
+  },
+  { 2,2,0,0 }
+};
+
+#endif  // INCLUDED_FROM_UBIDI_PROPS_C
diff --git a/thirdparty/icu4c/common/ubidiimp.h b/thirdparty/icu4c/common/ubidiimp.h
new file mode 100644
index 0000000000..e48fc6f941
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidiimp.h
@@ -0,0 +1,484 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ubidiimp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999aug06
+*   created by: Markus W. Scherer, updated by Matitiahu Allouche
+*/
+
+#ifndef UBIDIIMP_H
+#define UBIDIIMP_H
+
+#include "unicode/utypes.h"
+#include "unicode/ubidi.h"
+#include "unicode/uchar.h"
+#include "ubidi_props.h"
+
+/* miscellaneous definitions ---------------------------------------------- */
+
+// ICU-20853=ICU-20935 Solaris #defines CS and ES in sys/regset.h
+#ifdef CS
+#   undef CS
+#endif
+#ifdef ES
+#   undef ES
+#endif
+
+typedef uint8_t DirProp;
+typedef uint32_t Flags;
+
+/*  Comparing the description of the BiDi algorithm with this implementation
+    is easier with the same names for the BiDi types in the code as there.
+    See UCharDirection in uchar.h .
+*/
+enum {
+    L=  U_LEFT_TO_RIGHT,                /*  0 */
+    R=  U_RIGHT_TO_LEFT,                /*  1 */
+    EN= U_EUROPEAN_NUMBER,              /*  2 */
+    ES= U_EUROPEAN_NUMBER_SEPARATOR,    /*  3 */
+    ET= U_EUROPEAN_NUMBER_TERMINATOR,   /*  4 */
+    AN= U_ARABIC_NUMBER,                /*  5 */
+    CS= U_COMMON_NUMBER_SEPARATOR,      /*  6 */
+    B=  U_BLOCK_SEPARATOR,              /*  7 */
+    S=  U_SEGMENT_SEPARATOR,            /*  8 */
+    WS= U_WHITE_SPACE_NEUTRAL,          /*  9 */
+    ON= U_OTHER_NEUTRAL,                /* 10 */
+    LRE=U_LEFT_TO_RIGHT_EMBEDDING,      /* 11 */
+    LRO=U_LEFT_TO_RIGHT_OVERRIDE,       /* 12 */
+    AL= U_RIGHT_TO_LEFT_ARABIC,         /* 13 */
+    RLE=U_RIGHT_TO_LEFT_EMBEDDING,      /* 14 */
+    RLO=U_RIGHT_TO_LEFT_OVERRIDE,       /* 15 */
+    PDF=U_POP_DIRECTIONAL_FORMAT,       /* 16 */
+    NSM=U_DIR_NON_SPACING_MARK,         /* 17 */
+    BN= U_BOUNDARY_NEUTRAL,             /* 18 */
+    FSI=U_FIRST_STRONG_ISOLATE,         /* 19 */
+    LRI=U_LEFT_TO_RIGHT_ISOLATE,        /* 20 */
+    RLI=U_RIGHT_TO_LEFT_ISOLATE,        /* 21 */
+    PDI=U_POP_DIRECTIONAL_ISOLATE,      /* 22 */
+    ENL,    /* EN after W7 */           /* 23 */
+    ENR,    /* EN not subject to W7 */  /* 24 */
+    dirPropCount
+};
+
+/*  Sometimes, bit values are more appropriate
+    to deal with directionality properties.
+    Abbreviations in these macro names refer to names
+    used in the BiDi algorithm.
+*/
+#define DIRPROP_FLAG(dir) (1UL<<(dir))
+#define PURE_DIRPROP(prop)  ((prop)&~0xE0)    ?????????????????????????
+
+/* special flag for multiple runs from explicit embedding codes */
+#define DIRPROP_FLAG_MULTI_RUNS (1UL<<31)
+
+/* are there any characters that are LTR or RTL? */
+#define MASK_LTR (DIRPROP_FLAG(L)|DIRPROP_FLAG(EN)|DIRPROP_FLAG(ENL)|DIRPROP_FLAG(ENR)|DIRPROP_FLAG(AN)|DIRPROP_FLAG(LRE)|DIRPROP_FLAG(LRO)|DIRPROP_FLAG(LRI))
+#define MASK_RTL (DIRPROP_FLAG(R)|DIRPROP_FLAG(AL)|DIRPROP_FLAG(RLE)|DIRPROP_FLAG(RLO)|DIRPROP_FLAG(RLI))
+#define MASK_R_AL (DIRPROP_FLAG(R)|DIRPROP_FLAG(AL))
+#define MASK_STRONG_EN_AN (DIRPROP_FLAG(L)|DIRPROP_FLAG(R)|DIRPROP_FLAG(AL)|DIRPROP_FLAG(EN)|DIRPROP_FLAG(AN))
+
+/* explicit embedding codes */
+#define MASK_EXPLICIT (DIRPROP_FLAG(LRE)|DIRPROP_FLAG(LRO)|DIRPROP_FLAG(RLE)|DIRPROP_FLAG(RLO)|DIRPROP_FLAG(PDF))
+
+/* explicit isolate codes */
+#define MASK_ISO (DIRPROP_FLAG(LRI)|DIRPROP_FLAG(RLI)|DIRPROP_FLAG(FSI)|DIRPROP_FLAG(PDI))
+
+#define MASK_BN_EXPLICIT (DIRPROP_FLAG(BN)|MASK_EXPLICIT)
+
+/* paragraph and segment separators */
+#define MASK_B_S (DIRPROP_FLAG(B)|DIRPROP_FLAG(S))
+
+/* all types that are counted as White Space or Neutral in some steps */
+#define MASK_WS (MASK_B_S|DIRPROP_FLAG(WS)|MASK_BN_EXPLICIT|MASK_ISO)
+
+/* types that are neutrals or could becomes neutrals in (Wn) */
+#define MASK_POSSIBLE_N (DIRPROP_FLAG(ON)|DIRPROP_FLAG(CS)|DIRPROP_FLAG(ES)|DIRPROP_FLAG(ET)|MASK_WS)
+
+/*
+ *  These types may be changed to "e",
+ *  the embedding type (L or R) of the run,
+ *  in the BiDi algorithm (N2)
+ */
+#define MASK_EMBEDDING (DIRPROP_FLAG(NSM)|MASK_POSSIBLE_N)
+
+/* the dirProp's L and R are defined to 0 and 1 values in UCharDirection */
+#define GET_LR_FROM_LEVEL(level) ((DirProp)((level)&1))
+
+#define IS_DEFAULT_LEVEL(level) ((level)>=0xfe)
+
+/*
+ *  The following bit is used for the directional isolate status.
+ *  Stack entries corresponding to isolate sequences are greater than ISOLATE.
+ */
+#define ISOLATE  0x0100
+
+U_CFUNC UBiDiLevel
+ubidi_getParaLevelAtIndex(const UBiDi *pBiDi, int32_t index);
+
+#define GET_PARALEVEL(ubidi, index) \
+            ((UBiDiLevel)(!(ubidi)->defaultParaLevel || (index)<(ubidi)->paras[0].limit ? \
+                         (ubidi)->paraLevel : ubidi_getParaLevelAtIndex((ubidi), (index))))
+
+/* number of paras entries allocated initially without malloc */
+#define SIMPLE_PARAS_COUNT      10
+/* number of isolate entries allocated initially without malloc */
+#define SIMPLE_ISOLATES_COUNT   5
+/* number of isolate run entries for paired brackets allocated initially without malloc */
+#define SIMPLE_OPENINGS_COUNT   20
+
+#define CR  0x000D
+#define LF  0x000A
+
+/* Run structure for reordering --------------------------------------------- */
+enum {
+    LRM_BEFORE=1,
+    LRM_AFTER=2,
+    RLM_BEFORE=4,
+    RLM_AFTER=8
+};
+
+typedef struct Para {
+    int32_t limit;
+    int32_t level;
+} Para;
+
+enum {                                  /* flags for Opening.flags */
+    FOUND_L=DIRPROP_FLAG(L),
+    FOUND_R=DIRPROP_FLAG(R)
+};
+
+typedef struct Opening {
+    int32_t position;                   /* position of opening bracket */
+    int32_t match;                      /* matching char or -position of closing bracket */
+    int32_t contextPos;                 /* position of last strong char found before opening */
+    uint16_t flags;                     /* bits for L or R/AL found within the pair */
+    UBiDiDirection contextDir;          /* L or R according to last strong char before opening */
+    uint8_t filler;                     /* to complete a nice multiple of 4 chars */
+} Opening;
+
+typedef struct IsoRun {
+    int32_t  contextPos;                /* position of char determining context */
+    uint16_t start;                     /* index of first opening entry for this run */
+    uint16_t limit;                     /* index after last opening entry for this run */
+    UBiDiLevel level;                   /* level of this run */
+    DirProp lastStrong;                 /* bidi class of last strong char found in this run */
+    DirProp lastBase;                   /* bidi class of last base char found in this run */
+    UBiDiDirection contextDir;          /* L or R to use as context for following openings */
+} IsoRun;
+
+typedef struct BracketData {
+    UBiDi   *pBiDi;
+    /* array of opening entries which should be enough in most cases; no malloc() */
+    Opening simpleOpenings[SIMPLE_OPENINGS_COUNT];
+    Opening *openings;                  /* pointer to current array of entries */
+    int32_t openingsCount;              /* number of allocated entries */
+    int32_t isoRunLast;                 /* index of last used entry */
+    /* array of nested isolated sequence entries; can never excess UBIDI_MAX_EXPLICIT_LEVEL
+       + 1 for index 0, + 1 for before the first isolated sequence */
+    IsoRun  isoRuns[UBIDI_MAX_EXPLICIT_LEVEL+2];
+    UBool isNumbersSpecial;             /* reordering mode for NUMBERS_SPECIAL */
+} BracketData;
+
+typedef struct Isolate {
+    int32_t startON;
+    int32_t start1;
+    int32_t state;
+    int16_t stateImp;
+} Isolate;
+
+typedef struct Run {
+    int32_t logicalStart,   /* first character of the run; b31 indicates even/odd level */
+            visualLimit,    /* last visual position of the run +1 */
+            insertRemove;   /* if >0, flags for inserting LRM/RLM before/after run,
+                               if <0, count of bidi controls within run            */
+} Run;
+
+/* in a Run, logicalStart will get this bit set if the run level is odd */
+#define INDEX_ODD_BIT (1UL<<31)
+
+#define MAKE_INDEX_ODD_PAIR(index, level) ((index)|((int32_t)((level)&1)<<31))
+#define ADD_ODD_BIT_FROM_LEVEL(x, level)  ((x)|=((int32_t)((level)&1)<<31))
+#define REMOVE_ODD_BIT(x)                 ((x)&=~INDEX_ODD_BIT)
+
+#define GET_INDEX(x)   ((x)&~INDEX_ODD_BIT)
+#define GET_ODD_BIT(x) ((uint32_t)(x)>>31)
+#define IS_ODD_RUN(x)  ((UBool)(((x)&INDEX_ODD_BIT)!=0))
+#define IS_EVEN_RUN(x) ((UBool)(((x)&INDEX_ODD_BIT)==0))
+
+U_CFUNC UBool
+ubidi_getRuns(UBiDi *pBiDi, UErrorCode *pErrorCode);
+
+/** BiDi control code points */
+enum {
+    ZWNJ_CHAR=0x200c,
+    ZWJ_CHAR,
+    LRM_CHAR,
+    RLM_CHAR,
+    LRE_CHAR=0x202a,
+    RLE_CHAR,
+    PDF_CHAR,
+    LRO_CHAR,
+    RLO_CHAR,
+    LRI_CHAR=0x2066,
+    RLI_CHAR,
+    FSI_CHAR,
+    PDI_CHAR
+};
+
+#define IS_BIDI_CONTROL_CHAR(c) (((uint32_t)(c)&0xfffffffc)==ZWNJ_CHAR || (uint32_t)((c)-LRE_CHAR)<5 || (uint32_t)((c)-LRI_CHAR)<4)
+
+/* InsertPoints structure for noting where to put BiDi marks ---------------- */
+
+typedef struct Point {
+    int32_t pos;            /* position in text */
+    int32_t flag;           /* flag for LRM/RLM, before/after */
+} Point;
+
+typedef struct InsertPoints {
+    int32_t capacity;       /* number of points allocated */
+    int32_t size;           /* number of points used */
+    int32_t confirmed;      /* number of points confirmed */
+    UErrorCode errorCode;   /* for eventual memory shortage */
+    Point *points;          /* pointer to array of points */
+} InsertPoints;
+
+
+/* UBiDi structure ----------------------------------------------------------- */
+
+struct UBiDi {
+    /* pointer to parent paragraph object (pointer to self if this object is
+     * a paragraph object); set to NULL in a newly opened object; set to a
+     * real value after a successful execution of ubidi_setPara or ubidi_setLine
+     */
+    const UBiDi * pParaBiDi;
+
+    /* alias pointer to the current text */
+    const UChar *text;
+
+    /* length of the current text */
+    int32_t originalLength;
+
+    /* if the UBIDI_OPTION_STREAMING option is set, this is the length
+     * of text actually processed by ubidi_setPara, which may be shorter than
+     * the original length.
+     * Otherwise, it is identical to the original length.
+     */
+    int32_t length;
+
+    /* if the UBIDI_OPTION_REMOVE_CONTROLS option is set, and/or
+     * marks are allowed to be inserted in one of the reordering mode, the
+     * length of the result string may be different from the processed length.
+     */
+    int32_t resultLength;
+
+    /* memory sizes in bytes */
+    int32_t dirPropsSize, levelsSize, openingsSize, parasSize, runsSize, isolatesSize;
+
+    /* allocated memory */
+    DirProp *dirPropsMemory;
+    UBiDiLevel *levelsMemory;
+    Opening *openingsMemory;
+    Para *parasMemory;
+    Run *runsMemory;
+    Isolate *isolatesMemory;
+
+    /* indicators for whether memory may be allocated after ubidi_open() */
+    UBool mayAllocateText, mayAllocateRuns;
+
+    /* arrays with one value per text-character */
+    DirProp *dirProps;
+    UBiDiLevel *levels;
+
+    /* are we performing an approximation of the "inverse BiDi" algorithm? */
+    UBool isInverse;
+
+    /* are we using the basic algorithm or its variation? */
+    UBiDiReorderingMode reorderingMode;
+
+    /* UBIDI_REORDER_xxx values must be ordered so that all the regular
+     * logical to visual modes come first, and all inverse BiDi modes
+     * come last.
+     */
+    #define UBIDI_REORDER_LAST_LOGICAL_TO_VISUAL    UBIDI_REORDER_NUMBERS_SPECIAL
+
+    /* bitmask for reordering options */
+    uint32_t reorderingOptions;
+
+    /* must block separators receive level 0? */
+    UBool orderParagraphsLTR;
+
+    /* the paragraph level */
+    UBiDiLevel paraLevel;
+    /* original paraLevel when contextual */
+    /* must be one of UBIDI_DEFAULT_xxx or 0 if not contextual */
+    UBiDiLevel defaultParaLevel;
+
+    /* context data */
+    const UChar *prologue;
+    int32_t proLength;
+    const UChar *epilogue;
+    int32_t epiLength;
+
+    /* the following is set in ubidi_setPara, used in processPropertySeq */
+    const struct ImpTabPair * pImpTabPair;  /* pointer to levels state table pair */
+
+    /* the overall paragraph or line directionality - see UBiDiDirection */
+    UBiDiDirection direction;
+
+    /* flags is a bit set for which directional properties are in the text */
+    Flags flags;
+
+    /* lastArabicPos is index to the last AL in the text, -1 if none */
+    int32_t lastArabicPos;
+
+    /* characters after trailingWSStart are WS and are */
+    /* implicitly at the paraLevel (rule (L1)) - levels may not reflect that */
+    int32_t trailingWSStart;
+
+    /* fields for paragraph handling */
+    int32_t paraCount;                  /* set in getDirProps() */
+    /* filled in getDirProps() */
+    Para *paras;
+
+    /* for relatively short text, we only need a tiny array of paras (no malloc()) */
+    Para simpleParas[SIMPLE_PARAS_COUNT];
+
+    /* fields for line reordering */
+    int32_t runCount;     /* ==-1: runs not set up yet */
+    Run *runs;
+
+    /* for non-mixed text, we only need a tiny array of runs (no malloc()) */
+    Run simpleRuns[1];
+
+    /* maximum or current nesting depth of isolate sequences */
+    /* Within resolveExplicitLevels() and checkExplicitLevels(), this is the maximal
+       nesting encountered.
+       Within resolveImplicitLevels(), this is the index of the current isolates
+       stack entry. */
+    int32_t isolateCount;
+    Isolate *isolates;
+
+    /* for simple text, have a small stack (no malloc()) */
+    Isolate simpleIsolates[SIMPLE_ISOLATES_COUNT];
+
+    /* for inverse Bidi with insertion of directional marks */
+    InsertPoints insertPoints;
+
+    /* for option UBIDI_OPTION_REMOVE_CONTROLS */
+    int32_t controlCount;
+
+    /* for Bidi class callback */
+    UBiDiClassCallback *fnClassCallback;    /* action pointer */
+    const void *coClassCallback;            /* context pointer */
+};
+
+#define IS_VALID_PARA(x) ((x) && ((x)->pParaBiDi==(x)))
+#define IS_VALID_PARA_OR_LINE(x) ((x) && ((x)->pParaBiDi==(x) || (((x)->pParaBiDi) && (x)->pParaBiDi->pParaBiDi==(x)->pParaBiDi)))
+
+typedef union {
+    DirProp *dirPropsMemory;
+    UBiDiLevel *levelsMemory;
+    Opening *openingsMemory;
+    Para *parasMemory;
+    Run *runsMemory;
+    Isolate *isolatesMemory;
+} BidiMemoryForAllocation;
+
+/* Macros for initial checks at function entry */
+#define RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrcode, retvalue) UPRV_BLOCK_MACRO_BEGIN { \
+    if((pErrcode)==NULL || U_FAILURE(*pErrcode)) return retvalue; \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_IF_NOT_VALID_PARA(bidi, errcode, retvalue) UPRV_BLOCK_MACRO_BEGIN { \
+    if(!IS_VALID_PARA(bidi)) { \
+        errcode=U_INVALID_STATE_ERROR; \
+        return retvalue; \
+    } \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_IF_NOT_VALID_PARA_OR_LINE(bidi, errcode, retvalue) UPRV_BLOCK_MACRO_BEGIN { \
+    if(!IS_VALID_PARA_OR_LINE(bidi)) { \
+        errcode=U_INVALID_STATE_ERROR; \
+        return retvalue; \
+    } \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_IF_BAD_RANGE(arg, start, limit, errcode, retvalue) UPRV_BLOCK_MACRO_BEGIN { \
+    if((arg)<(start) || (arg)>=(limit)) { \
+        (errcode)=U_ILLEGAL_ARGUMENT_ERROR; \
+        return retvalue; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#define RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrcode) UPRV_BLOCK_MACRO_BEGIN { \
+    if((pErrcode)==NULL || U_FAILURE(*pErrcode)) return; \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_VOID_IF_NOT_VALID_PARA(bidi, errcode) UPRV_BLOCK_MACRO_BEGIN { \
+    if(!IS_VALID_PARA(bidi)) { \
+        errcode=U_INVALID_STATE_ERROR; \
+        return; \
+    } \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_VOID_IF_NOT_VALID_PARA_OR_LINE(bidi, errcode) UPRV_BLOCK_MACRO_BEGIN { \
+    if(!IS_VALID_PARA_OR_LINE(bidi)) { \
+        errcode=U_INVALID_STATE_ERROR; \
+        return; \
+    } \
+} UPRV_BLOCK_MACRO_END
+#define RETURN_VOID_IF_BAD_RANGE(arg, start, limit, errcode) UPRV_BLOCK_MACRO_BEGIN { \
+    if((arg)<(start) || (arg)>=(limit)) { \
+        (errcode)=U_ILLEGAL_ARGUMENT_ERROR; \
+        return; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* helper function to (re)allocate memory if allowed */
+U_CFUNC UBool
+ubidi_getMemory(BidiMemoryForAllocation *pMemory, int32_t *pSize, UBool mayAllocate, int32_t sizeNeeded);
+
+/* helper macros for each allocated array in UBiDi */
+#define getDirPropsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->dirPropsMemory, &(pBiDi)->dirPropsSize, \
+                        (pBiDi)->mayAllocateText, (length))
+
+#define getLevelsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->levelsMemory, &(pBiDi)->levelsSize, \
+                        (pBiDi)->mayAllocateText, (length))
+
+#define getRunsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->runsMemory, &(pBiDi)->runsSize, \
+                        (pBiDi)->mayAllocateRuns, (length)*sizeof(Run))
+
+/* additional macros used by ubidi_open() - always allow allocation */
+#define getInitialDirPropsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->dirPropsMemory, &(pBiDi)->dirPropsSize, \
+                        true, (length))
+
+#define getInitialLevelsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->levelsMemory, &(pBiDi)->levelsSize, \
+                        true, (length))
+
+#define getInitialOpeningsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->openingsMemory, &(pBiDi)->openingsSize, \
+                        true, (length)*sizeof(Opening))
+
+#define getInitialParasMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->parasMemory, &(pBiDi)->parasSize, \
+                        true, (length)*sizeof(Para))
+
+#define getInitialRunsMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->runsMemory, &(pBiDi)->runsSize, \
+                        true, (length)*sizeof(Run))
+
+#define getInitialIsolatesMemory(pBiDi, length) \
+        ubidi_getMemory((BidiMemoryForAllocation *)&(pBiDi)->isolatesMemory, &(pBiDi)->isolatesSize, \
+                        true, (length)*sizeof(Isolate))
+
+#endif
diff --git a/thirdparty/icu4c/common/ubidiln.cpp b/thirdparty/icu4c/common/ubidiln.cpp
new file mode 100644
index 0000000000..3545f4e111
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidiln.cpp
@@ -0,0 +1,1347 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ubidiln.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999aug06
+*   created by: Markus W. Scherer, updated by Matitiahu Allouche
+*/
+
+#include "cmemory.h"
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/uchar.h"
+#include "unicode/ubidi.h"
+#include "ubidiimp.h"
+#include "uassert.h"
+
+/*
+ * General remarks about the functions in this file:
+ *
+ * These functions deal with the aspects of potentially mixed-directional
+ * text in a single paragraph or in a line of a single paragraph
+ * which has already been processed according to
+ * the Unicode 6.3 BiDi algorithm as defined in
+ * http://www.unicode.org/unicode/reports/tr9/ , version 28,
+ * also described in The Unicode Standard, Version 6.3.0 .
+ *
+ * This means that there is a UBiDi object with a levels
+ * and a dirProps array.
+ * paraLevel and direction are also set.
+ * Only if the length of the text is zero, then levels==dirProps==NULL.
+ *
+ * The overall directionality of the paragraph
+ * or line is used to bypass the reordering steps if possible.
+ * Even purely RTL text does not need reordering there because
+ * the ubidi_getLogical/VisualIndex() functions can compute the
+ * index on the fly in such a case.
+ *
+ * The implementation of the access to same-level-runs and of the reordering
+ * do attempt to provide better performance and less memory usage compared to
+ * a direct implementation of especially rule (L2) with an array of
+ * one (32-bit) integer per text character.
+ *
+ * Here, the levels array is scanned as soon as necessary, and a vector of
+ * same-level-runs is created. Reordering then is done on this vector.
+ * For each run of text positions that were resolved to the same level,
+ * only 8 bytes are stored: the first text position of the run and the visual
+ * position behind the run after reordering.
+ * One sign bit is used to hold the directionality of the run.
+ * This is inefficient if there are many very short runs. If the average run
+ * length is <2, then this uses more memory.
+ *
+ * In a further attempt to save memory, the levels array is never changed
+ * after all the resolution rules (Xn, Wn, Nn, In).
+ * Many functions have to consider the field trailingWSStart:
+ * if it is less than length, then there is an implicit trailing run
+ * at the paraLevel,
+ * which is not reflected in the levels array.
+ * This allows a line UBiDi object to use the same levels array as
+ * its paragraph parent object.
+ *
+ * When a UBiDi object is created for a line of a paragraph, then the
+ * paragraph's levels and dirProps arrays are reused by way of setting
+ * a pointer into them, not by copying. This again saves memory and forbids to
+ * change the now shared levels for (L1).
+ */
+
+/* handle trailing WS (L1) -------------------------------------------------- */
+
+/*
+ * setTrailingWSStart() sets the start index for a trailing
+ * run of WS in the line. This is necessary because we do not modify
+ * the paragraph's levels array that we just point into.
+ * Using trailingWSStart is another form of performing (L1).
+ *
+ * To make subsequent operations easier, we also include the run
+ * before the WS if it is at the paraLevel - we merge the two here.
+ *
+ * This function is called only from ubidi_setLine(), so pBiDi->paraLevel is
+ * set correctly for the line even when contextual multiple paragraphs.
+ */
+static void
+setTrailingWSStart(UBiDi *pBiDi) {
+    /* pBiDi->direction!=UBIDI_MIXED */
+
+    const DirProp *dirProps=pBiDi->dirProps;
+    UBiDiLevel *levels=pBiDi->levels;
+    int32_t start=pBiDi->length;
+    UBiDiLevel paraLevel=pBiDi->paraLevel;
+
+    /* If the line is terminated by a block separator, all preceding WS etc...
+       are already set to paragraph level.
+       Setting trailingWSStart to pBidi->length will avoid changing the
+       level of B chars from 0 to paraLevel in ubidi_getLevels when
+       orderParagraphsLTR==TRUE.
+     */
+    if(dirProps[start-1]==B) {
+        pBiDi->trailingWSStart=start;   /* currently == pBiDi->length */
+        return;
+    }
+    /* go backwards across all WS, BN, explicit codes */
+    while(start>0 && DIRPROP_FLAG(dirProps[start-1])&MASK_WS) {
+        --start;
+    }
+
+    /* if the WS run can be merged with the previous run then do so here */
+    while(start>0 && levels[start-1]==paraLevel) {
+        --start;
+    }
+
+    pBiDi->trailingWSStart=start;
+}
+
+/* ubidi_setLine ------------------------------------------------------------ */
+
+U_CAPI void U_EXPORT2
+ubidi_setLine(const UBiDi *pParaBiDi,
+              int32_t start, int32_t limit,
+              UBiDi *pLineBiDi,
+              UErrorCode *pErrorCode) {
+    int32_t length;
+
+    /* check the argument values */
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    RETURN_VOID_IF_NOT_VALID_PARA(pParaBiDi, *pErrorCode);
+    RETURN_VOID_IF_BAD_RANGE(start, 0, limit, *pErrorCode);
+    RETURN_VOID_IF_BAD_RANGE(limit, 0, pParaBiDi->length+1, *pErrorCode);
+    if(pLineBiDi==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if(ubidi_getParagraph(pParaBiDi, start, NULL, NULL, NULL, pErrorCode) !=
+       ubidi_getParagraph(pParaBiDi, limit-1, NULL, NULL, NULL, pErrorCode)) {
+        /* the line crosses a paragraph boundary */
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    /* set the values in pLineBiDi from its pParaBiDi parent */
+    pLineBiDi->pParaBiDi=NULL;          /* mark unfinished setLine */
+    pLineBiDi->text=pParaBiDi->text+start;
+    length=pLineBiDi->length=limit-start;
+    pLineBiDi->resultLength=pLineBiDi->originalLength=length;
+    pLineBiDi->paraLevel=GET_PARALEVEL(pParaBiDi, start);
+    pLineBiDi->paraCount=pParaBiDi->paraCount;
+    pLineBiDi->runs=NULL;
+    pLineBiDi->flags=0;
+    pLineBiDi->reorderingMode=pParaBiDi->reorderingMode;
+    pLineBiDi->reorderingOptions=pParaBiDi->reorderingOptions;
+    pLineBiDi->controlCount=0;
+    if(pParaBiDi->controlCount>0) {
+        int32_t j;
+        for(j=start; j<limit; j++) {
+            if(IS_BIDI_CONTROL_CHAR(pParaBiDi->text[j])) {
+                pLineBiDi->controlCount++;
+            }
+        }
+        pLineBiDi->resultLength-=pLineBiDi->controlCount;
+    }
+
+    pLineBiDi->dirProps=pParaBiDi->dirProps+start;
+    pLineBiDi->levels=pParaBiDi->levels+start;
+    pLineBiDi->runCount=-1;
+
+    if(pParaBiDi->direction!=UBIDI_MIXED) {
+        /* the parent is already trivial */
+        pLineBiDi->direction=pParaBiDi->direction;
+
+        /*
+         * The parent's levels are all either
+         * implicitly or explicitly ==paraLevel;
+         * do the same here.
+         */
+        if(pParaBiDi->trailingWSStart<=start) {
+            pLineBiDi->trailingWSStart=0;
+        } else if(pParaBiDi->trailingWSStart<limit) {
+            pLineBiDi->trailingWSStart=pParaBiDi->trailingWSStart-start;
+        } else {
+            pLineBiDi->trailingWSStart=length;
+        }
+    } else {
+        const UBiDiLevel *levels=pLineBiDi->levels;
+        int32_t i, trailingWSStart;
+        UBiDiLevel level;
+
+        setTrailingWSStart(pLineBiDi);
+        trailingWSStart=pLineBiDi->trailingWSStart;
+
+        /* recalculate pLineBiDi->direction */
+        if(trailingWSStart==0) {
+            /* all levels are at paraLevel */
+            pLineBiDi->direction=(UBiDiDirection)(pLineBiDi->paraLevel&1);
+        } else {
+            /* get the level of the first character */
+            level=(UBiDiLevel)(levels[0]&1);
+
+            /* if there is anything of a different level, then the line is mixed */
+            if(trailingWSStart<length && (pLineBiDi->paraLevel&1)!=level) {
+                /* the trailing WS is at paraLevel, which differs from levels[0] */
+                pLineBiDi->direction=UBIDI_MIXED;
+            } else {
+                /* see if levels[1..trailingWSStart-1] have the same direction as levels[0] and paraLevel */
+                i=1;
+                for(;;) {
+                    if(i==trailingWSStart) {
+                        /* the direction values match those in level */
+                        pLineBiDi->direction=(UBiDiDirection)level;
+                        break;
+                    } else if((levels[i]&1)!=level) {
+                        pLineBiDi->direction=UBIDI_MIXED;
+                        break;
+                    }
+                    ++i;
+                }
+            }
+        }
+
+        switch(pLineBiDi->direction) {
+        case UBIDI_LTR:
+            /* make sure paraLevel is even */
+            pLineBiDi->paraLevel=(UBiDiLevel)((pLineBiDi->paraLevel+1)&~1);
+
+            /* all levels are implicitly at paraLevel (important for ubidi_getLevels()) */
+            pLineBiDi->trailingWSStart=0;
+            break;
+        case UBIDI_RTL:
+            /* make sure paraLevel is odd */
+            pLineBiDi->paraLevel|=1;
+
+            /* all levels are implicitly at paraLevel (important for ubidi_getLevels()) */
+            pLineBiDi->trailingWSStart=0;
+            break;
+        default:
+            break;
+        }
+    }
+    pLineBiDi->pParaBiDi=pParaBiDi;     /* mark successful setLine */
+    return;
+}
+
+U_CAPI UBiDiLevel U_EXPORT2
+ubidi_getLevelAt(const UBiDi *pBiDi, int32_t charIndex) {
+    /* return paraLevel if in the trailing WS run, otherwise the real level */
+    if(!IS_VALID_PARA_OR_LINE(pBiDi) || charIndex<0 || pBiDi->length<=charIndex) {
+        return 0;
+    } else if(pBiDi->direction!=UBIDI_MIXED || charIndex>=pBiDi->trailingWSStart) {
+        return GET_PARALEVEL(pBiDi, charIndex);
+    } else {
+        return pBiDi->levels[charIndex];
+    }
+}
+
+U_CAPI const UBiDiLevel * U_EXPORT2
+ubidi_getLevels(UBiDi *pBiDi, UErrorCode *pErrorCode) {
+    int32_t start, length;
+
+    RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrorCode, NULL);
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode, NULL);
+    if((length=pBiDi->length)<=0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    if((start=pBiDi->trailingWSStart)==length) {
+        /* the current levels array reflects the WS run */
+        return pBiDi->levels;
+    }
+
+    /*
+     * After the previous if(), we know that the levels array
+     * has an implicit trailing WS run and therefore does not fully
+     * reflect itself all the levels.
+     * This must be a UBiDi object for a line, and
+     * we need to create a new levels array.
+     */
+    if(getLevelsMemory(pBiDi, length)) {
+        UBiDiLevel *levels=pBiDi->levelsMemory;
+
+        if(start>0 && levels!=pBiDi->levels) {
+            uprv_memcpy(levels, pBiDi->levels, start);
+        }
+        /* pBiDi->paraLevel is ok even if contextual multiple paragraphs,
+           since pBidi is a line object                                     */
+        uprv_memset(levels+start, pBiDi->paraLevel, length-start);
+
+        /* this new levels array is set for the line and reflects the WS run */
+        pBiDi->trailingWSStart=length;
+        return pBiDi->levels=levels;
+    } else {
+        /* out of memory */
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_getLogicalRun(const UBiDi *pBiDi, int32_t logicalPosition,
+                    int32_t *pLogicalLimit, UBiDiLevel *pLevel) {
+    UErrorCode errorCode;
+    int32_t runCount, visualStart, logicalLimit, logicalFirst, i;
+    Run iRun;
+
+    errorCode=U_ZERO_ERROR;
+    RETURN_VOID_IF_BAD_RANGE(logicalPosition, 0, pBiDi->length, errorCode);
+    /* ubidi_countRuns will check VALID_PARA_OR_LINE */
+    runCount=ubidi_countRuns((UBiDi *)pBiDi, &errorCode);
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    /* this is done based on runs rather than on levels since levels have
+       a special interpretation when UBIDI_REORDER_RUNS_ONLY
+     */
+    visualStart=logicalLimit=0;
+    iRun=pBiDi->runs[0];
+
+    for(i=0; i<runCount; i++) {
+        iRun = pBiDi->runs[i];
+        logicalFirst=GET_INDEX(iRun.logicalStart);
+        logicalLimit=logicalFirst+iRun.visualLimit-visualStart;
+        if((logicalPosition>=logicalFirst) &&
+           (logicalPosition<logicalLimit)) {
+            break;
+        }
+        visualStart = iRun.visualLimit;
+    }
+    if(pLogicalLimit) {
+        *pLogicalLimit=logicalLimit;
+    }
+    if(pLevel) {
+        if(pBiDi->reorderingMode==UBIDI_REORDER_RUNS_ONLY) {
+            *pLevel=(UBiDiLevel)GET_ODD_BIT(iRun.logicalStart);
+        }
+        else if(pBiDi->direction!=UBIDI_MIXED || logicalPosition>=pBiDi->trailingWSStart) {
+            *pLevel=GET_PARALEVEL(pBiDi, logicalPosition);
+        } else {
+        *pLevel=pBiDi->levels[logicalPosition];
+        }
+    }
+}
+
+/* runs API functions ------------------------------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+ubidi_countRuns(UBiDi *pBiDi, UErrorCode *pErrorCode) {
+    RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrorCode, -1);
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode, -1);
+    ubidi_getRuns(pBiDi, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return -1;
+    }
+    return pBiDi->runCount;
+}
+
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getVisualRun(UBiDi *pBiDi, int32_t runIndex,
+                   int32_t *pLogicalStart, int32_t *pLength)
+{
+    int32_t start;
+    UErrorCode errorCode = U_ZERO_ERROR;
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, errorCode, UBIDI_LTR);
+    ubidi_getRuns(pBiDi, &errorCode);
+    if(U_FAILURE(errorCode)) {
+        return UBIDI_LTR;
+    }
+    RETURN_IF_BAD_RANGE(runIndex, 0, pBiDi->runCount, errorCode, UBIDI_LTR);
+
+    start=pBiDi->runs[runIndex].logicalStart;
+    if(pLogicalStart!=NULL) {
+        *pLogicalStart=GET_INDEX(start);
+    }
+    if(pLength!=NULL) {
+        if(runIndex>0) {
+            *pLength=pBiDi->runs[runIndex].visualLimit-
+                     pBiDi->runs[runIndex-1].visualLimit;
+        } else {
+            *pLength=pBiDi->runs[0].visualLimit;
+        }
+    }
+    return (UBiDiDirection)GET_ODD_BIT(start);
+}
+
+/* in trivial cases there is only one trivial run; called by ubidi_getRuns() */
+static void
+getSingleRun(UBiDi *pBiDi, UBiDiLevel level) {
+    /* simple, single-run case */
+    pBiDi->runs=pBiDi->simpleRuns;
+    pBiDi->runCount=1;
+
+    /* fill and reorder the single run */
+    pBiDi->runs[0].logicalStart=MAKE_INDEX_ODD_PAIR(0, level);
+    pBiDi->runs[0].visualLimit=pBiDi->length;
+    pBiDi->runs[0].insertRemove=0;
+}
+
+/* reorder the runs array (L2) ---------------------------------------------- */
+
+/*
+ * Reorder the same-level runs in the runs array.
+ * Here, runCount>1 and maxLevel>=minLevel>=paraLevel.
+ * All the visualStart fields=logical start before reordering.
+ * The "odd" bits are not set yet.
+ *
+ * Reordering with this data structure lends itself to some handy shortcuts:
+ *
+ * Since each run is moved but not modified, and since at the initial maxLevel
+ * each sequence of same-level runs consists of only one run each, we
+ * don't need to do anything there and can predecrement maxLevel.
+ * In many simple cases, the reordering is thus done entirely in the
+ * index mapping.
+ * Also, reordering occurs only down to the lowest odd level that occurs,
+ * which is minLevel|1. However, if the lowest level itself is odd, then
+ * in the last reordering the sequence of the runs at this level or higher
+ * will be all runs, and we don't need the elaborate loop to search for them.
+ * This is covered by ++minLevel instead of minLevel|=1 followed
+ * by an extra reorder-all after the reorder-some loop.
+ * About a trailing WS run:
+ * Such a run would need special treatment because its level is not
+ * reflected in levels[] if this is not a paragraph object.
+ * Instead, all characters from trailingWSStart on are implicitly at
+ * paraLevel.
+ * However, for all maxLevel>paraLevel, this run will never be reordered
+ * and does not need to be taken into account. maxLevel==paraLevel is only reordered
+ * if minLevel==paraLevel is odd, which is done in the extra segment.
+ * This means that for the main reordering loop we don't need to consider
+ * this run and can --runCount. If it is later part of the all-runs
+ * reordering, then runCount is adjusted accordingly.
+ */
+static void
+reorderLine(UBiDi *pBiDi, UBiDiLevel minLevel, UBiDiLevel maxLevel) {
+    Run *runs, tempRun;
+    UBiDiLevel *levels;
+    int32_t firstRun, endRun, limitRun, runCount;
+
+    /* nothing to do? */
+    if(maxLevel<=(minLevel|1)) {
+        return;
+    }
+
+    /*
+     * Reorder only down to the lowest odd level
+     * and reorder at an odd minLevel in a separate, simpler loop.
+     * See comments above for why minLevel is always incremented.
+     */
+    ++minLevel;
+
+    runs=pBiDi->runs;
+    levels=pBiDi->levels;
+    runCount=pBiDi->runCount;
+
+    /* do not include the WS run at paraLevel<=old minLevel except in the simple loop */
+    if(pBiDi->trailingWSStart<pBiDi->length) {
+        --runCount;
+    }
+
+    while(--maxLevel>=minLevel) {
+        firstRun=0;
+
+        /* loop for all sequences of runs */
+        for(;;) {
+            /* look for a sequence of runs that are all at >=maxLevel */
+            /* look for the first run of such a sequence */
+            while(firstRun<runCount && levels[runs[firstRun].logicalStart]<maxLevel) {
+                ++firstRun;
+            }
+            if(firstRun>=runCount) {
+                break;  /* no more such runs */
+            }
+
+            /* look for the limit run of such a sequence (the run behind it) */
+            for(limitRun=firstRun; ++limitRun<runCount && levels[runs[limitRun].logicalStart]>=maxLevel;) {}
+
+            /* Swap the entire sequence of runs from firstRun to limitRun-1. */
+            endRun=limitRun-1;
+            while(firstRun<endRun) {
+                tempRun = runs[firstRun];
+                runs[firstRun]=runs[endRun];
+                runs[endRun]=tempRun;
+                ++firstRun;
+                --endRun;
+            }
+
+            if(limitRun==runCount) {
+                break;  /* no more such runs */
+            } else {
+                firstRun=limitRun+1;
+            }
+        }
+    }
+
+    /* now do maxLevel==old minLevel (==odd!), see above */
+    if(!(minLevel&1)) {
+        firstRun=0;
+
+        /* include the trailing WS run in this complete reordering */
+        if(pBiDi->trailingWSStart==pBiDi->length) {
+            --runCount;
+        }
+
+        /* Swap the entire sequence of all runs. (endRun==runCount) */
+        while(firstRun<runCount) {
+            tempRun=runs[firstRun];
+            runs[firstRun]=runs[runCount];
+            runs[runCount]=tempRun;
+            ++firstRun;
+            --runCount;
+        }
+    }
+}
+
+/* compute the runs array --------------------------------------------------- */
+
+static int32_t getRunFromLogicalIndex(UBiDi *pBiDi, int32_t logicalIndex) {
+    Run *runs=pBiDi->runs;
+    int32_t runCount=pBiDi->runCount, visualStart=0, i, length, logicalStart;
+
+    for(i=0; i<runCount; i++) {
+        length=runs[i].visualLimit-visualStart;
+        logicalStart=GET_INDEX(runs[i].logicalStart);
+        if((logicalIndex>=logicalStart) && (logicalIndex<(logicalStart+length))) {
+            return i;
+        }
+        visualStart+=length;
+    }
+    /* we should never get here */
+    UPRV_UNREACHABLE;
+}
+
+/*
+ * Compute the runs array from the levels array.
+ * After ubidi_getRuns() returns TRUE, runCount is guaranteed to be >0
+ * and the runs are reordered.
+ * Odd-level runs have visualStart on their visual right edge and
+ * they progress visually to the left.
+ * If option UBIDI_OPTION_INSERT_MARKS is set, insertRemove will contain the
+ * sum of appropriate LRM/RLM_BEFORE/AFTER flags.
+ * If option UBIDI_OPTION_REMOVE_CONTROLS is set, insertRemove will contain the
+ * negative number of BiDi control characters within this run.
+ */
+U_CFUNC UBool
+ubidi_getRuns(UBiDi *pBiDi, UErrorCode*) {
+    /*
+     * This method returns immediately if the runs are already set. This
+     * includes the case of length==0 (handled in setPara)..
+     */
+    if (pBiDi->runCount>=0) {
+        return TRUE;
+    }
+
+    if(pBiDi->direction!=UBIDI_MIXED) {
+        /* simple, single-run case - this covers length==0 */
+        /* pBiDi->paraLevel is ok even for contextual multiple paragraphs */
+        getSingleRun(pBiDi, pBiDi->paraLevel);
+    } else /* UBIDI_MIXED, length>0 */ {
+        /* mixed directionality */
+        int32_t length=pBiDi->length, limit;
+        UBiDiLevel *levels=pBiDi->levels;
+        int32_t i, runCount;
+        UBiDiLevel level=UBIDI_DEFAULT_LTR;   /* initialize with no valid level */
+        /*
+         * If there are WS characters at the end of the line
+         * and the run preceding them has a level different from
+         * paraLevel, then they will form their own run at paraLevel (L1).
+         * Count them separately.
+         * We need some special treatment for this in order to not
+         * modify the levels array which a line UBiDi object shares
+         * with its paragraph parent and its other line siblings.
+         * In other words, for the trailing WS, it may be
+         * levels[]!=paraLevel but we have to treat it like it were so.
+         */
+        limit=pBiDi->trailingWSStart;
+        /* count the runs, there is at least one non-WS run, and limit>0 */
+        runCount=0;
+        for(i=0; i<limit; ++i) {
+            /* increment runCount at the start of each run */
+            if(levels[i]!=level) {
+                ++runCount;
+                level=levels[i];
+            }
+        }
+
+        /*
+         * We don't need to see if the last run can be merged with a trailing
+         * WS run because setTrailingWSStart() would have done that.
+         */
+        if(runCount==1 && limit==length) {
+            /* There is only one non-WS run and no trailing WS-run. */
+            getSingleRun(pBiDi, levels[0]);
+        } else /* runCount>1 || limit<length */ {
+            /* allocate and set the runs */
+            Run *runs;
+            int32_t runIndex, start;
+            UBiDiLevel minLevel=UBIDI_MAX_EXPLICIT_LEVEL+1, maxLevel=0;
+
+            /* now, count a (non-mergeable) WS run */
+            if(limit<length) {
+                ++runCount;
+            }
+
+            /* runCount>1 */
+            if(getRunsMemory(pBiDi, runCount)) {
+                runs=pBiDi->runsMemory;
+            } else {
+                return FALSE;
+            }
+
+            /* set the runs */
+            /* FOOD FOR THOUGHT: this could be optimized, e.g.:
+             * 464->444, 484->444, 575->555, 595->555
+             * However, that would take longer. Check also how it would
+             * interact with BiDi control removal and inserting Marks.
+             */
+            runIndex=0;
+
+            /* search for the run limits and initialize visualLimit values with the run lengths */
+            i=0;
+            do {
+                /* prepare this run */
+                start=i;
+                level=levels[i];
+                if(level<minLevel) {
+                    minLevel=level;
+                }
+                if(level>maxLevel) {
+                    maxLevel=level;
+                }
+
+                /* look for the run limit */
+                while(++i<limit && levels[i]==level) {}
+
+                /* i is another run limit */
+                runs[runIndex].logicalStart=start;
+                runs[runIndex].visualLimit=i-start;
+                runs[runIndex].insertRemove=0;
+                ++runIndex;
+            } while(i<limit);
+
+            if(limit<length) {
+                /* there is a separate WS run */
+                runs[runIndex].logicalStart=limit;
+                runs[runIndex].visualLimit=length-limit;
+                /* For the trailing WS run, pBiDi->paraLevel is ok even
+                   if contextual multiple paragraphs.                   */
+                if(pBiDi->paraLevel<minLevel) {
+                    minLevel=pBiDi->paraLevel;
+                }
+            }
+
+            /* set the object fields */
+            pBiDi->runs=runs;
+            pBiDi->runCount=runCount;
+
+            reorderLine(pBiDi, minLevel, maxLevel);
+
+            /* now add the direction flags and adjust the visualLimit's to be just that */
+            /* this loop will also handle the trailing WS run */
+            limit=0;
+            for(i=0; i<runCount; ++i) {
+                ADD_ODD_BIT_FROM_LEVEL(runs[i].logicalStart, levels[runs[i].logicalStart]);
+                limit+=runs[i].visualLimit;
+                runs[i].visualLimit=limit;
+            }
+
+            /* Set the "odd" bit for the trailing WS run. */
+            /* For a RTL paragraph, it will be the *first* run in visual order. */
+            /* For the trailing WS run, pBiDi->paraLevel is ok even if
+               contextual multiple paragraphs.                          */
+            if(runIndex<runCount) {
+                int32_t trailingRun = ((pBiDi->paraLevel & 1) != 0)? 0 : runIndex;
+
+                ADD_ODD_BIT_FROM_LEVEL(runs[trailingRun].logicalStart, pBiDi->paraLevel);
+            }
+        }
+    }
+
+    /* handle insert LRM/RLM BEFORE/AFTER run */
+    if(pBiDi->insertPoints.size>0) {
+        Point *point, *start=pBiDi->insertPoints.points,
+                      *limit=start+pBiDi->insertPoints.size;
+        int32_t runIndex;
+        for(point=start; point<limit; point++) {
+            runIndex=getRunFromLogicalIndex(pBiDi, point->pos);
+            pBiDi->runs[runIndex].insertRemove|=point->flag;
+        }
+    }
+
+    /* handle remove BiDi control characters */
+    if(pBiDi->controlCount>0) {
+        int32_t runIndex;
+        const UChar *start=pBiDi->text, *limit=start+pBiDi->length, *pu;
+        for(pu=start; pu<limit; pu++) {
+            if(IS_BIDI_CONTROL_CHAR(*pu)) {
+                runIndex=getRunFromLogicalIndex(pBiDi, (int32_t)(pu-start));
+                pBiDi->runs[runIndex].insertRemove--;
+            }
+        }
+    }
+
+    return TRUE;
+}
+
+static UBool
+prepareReorder(const UBiDiLevel *levels, int32_t length,
+               int32_t *indexMap,
+               UBiDiLevel *pMinLevel, UBiDiLevel *pMaxLevel) {
+    int32_t start;
+    UBiDiLevel level, minLevel, maxLevel;
+
+    if(levels==NULL || length<=0) {
+        return FALSE;
+    }
+
+    /* determine minLevel and maxLevel */
+    minLevel=UBIDI_MAX_EXPLICIT_LEVEL+1;
+    maxLevel=0;
+    for(start=length; start>0;) {
+        level=levels[--start];
+        if(level>UBIDI_MAX_EXPLICIT_LEVEL+1) {
+            return FALSE;
+        }
+        if(level<minLevel) {
+            minLevel=level;
+        }
+        if(level>maxLevel) {
+            maxLevel=level;
+        }
+    }
+    *pMinLevel=minLevel;
+    *pMaxLevel=maxLevel;
+
+    /* initialize the index map */
+    for(start=length; start>0;) {
+        --start;
+        indexMap[start]=start;
+    }
+
+    return TRUE;
+}
+
+/* reorder a line based on a levels array (L2) ------------------------------ */
+
+U_CAPI void U_EXPORT2
+ubidi_reorderLogical(const UBiDiLevel *levels, int32_t length, int32_t *indexMap) {
+    int32_t start, limit, sumOfSosEos;
+    UBiDiLevel minLevel = 0, maxLevel = 0;
+
+    if(indexMap==NULL || !prepareReorder(levels, length, indexMap, &minLevel, &maxLevel)) {
+        return;
+    }
+
+    /* nothing to do? */
+    if(minLevel==maxLevel && (minLevel&1)==0) {
+        return;
+    }
+
+    /* reorder only down to the lowest odd level */
+    minLevel|=1;
+
+    /* loop maxLevel..minLevel */
+    do {
+        start=0;
+
+        /* loop for all sequences of levels to reorder at the current maxLevel */
+        for(;;) {
+            /* look for a sequence of levels that are all at >=maxLevel */
+            /* look for the first index of such a sequence */
+            while(start<length && levels[start]<maxLevel) {
+                ++start;
+            }
+            if(start>=length) {
+                break;  /* no more such sequences */
+            }
+
+            /* look for the limit of such a sequence (the index behind it) */
+            for(limit=start; ++limit<length && levels[limit]>=maxLevel;) {}
+
+            /*
+             * sos=start of sequence, eos=end of sequence
+             *
+             * The closed (inclusive) interval from sos to eos includes all the logical
+             * and visual indexes within this sequence. They are logically and
+             * visually contiguous and in the same range.
+             *
+             * For each run, the new visual index=sos+eos-old visual index;
+             * we pre-add sos+eos into sumOfSosEos ->
+             * new visual index=sumOfSosEos-old visual index;
+             */
+            sumOfSosEos=start+limit-1;
+
+            /* reorder each index in the sequence */
+            do {
+                indexMap[start]=sumOfSosEos-indexMap[start];
+            } while(++start<limit);
+
+            /* start==limit */
+            if(limit==length) {
+                break;  /* no more such sequences */
+            } else {
+                start=limit+1;
+            }
+        }
+    } while(--maxLevel>=minLevel);
+}
+
+U_CAPI void U_EXPORT2
+ubidi_reorderVisual(const UBiDiLevel *levels, int32_t length, int32_t *indexMap) {
+    int32_t start, end, limit, temp;
+    UBiDiLevel minLevel = 0, maxLevel = 0;
+
+    if(indexMap==NULL || !prepareReorder(levels, length, indexMap, &minLevel, &maxLevel)) {
+        return;
+    }
+
+    /* nothing to do? */
+    if(minLevel==maxLevel && (minLevel&1)==0) {
+        return;
+    }
+
+    /* reorder only down to the lowest odd level */
+    minLevel|=1;
+
+    /* loop maxLevel..minLevel */
+    do {
+        start=0;
+
+        /* loop for all sequences of levels to reorder at the current maxLevel */
+        for(;;) {
+            /* look for a sequence of levels that are all at >=maxLevel */
+            /* look for the first index of such a sequence */
+            while(start<length && levels[start]<maxLevel) {
+                ++start;
+            }
+            if(start>=length) {
+                break;  /* no more such runs */
+            }
+
+            /* look for the limit of such a sequence (the index behind it) */
+            for(limit=start; ++limit<length && levels[limit]>=maxLevel;) {}
+
+            /*
+             * Swap the entire interval of indexes from start to limit-1.
+             * We don't need to swap the levels for the purpose of this
+             * algorithm: the sequence of levels that we look at does not
+             * move anyway.
+             */
+            end=limit-1;
+            while(start<end) {
+                temp=indexMap[start];
+                indexMap[start]=indexMap[end];
+                indexMap[end]=temp;
+
+                ++start;
+                --end;
+            }
+
+            if(limit==length) {
+                break;  /* no more such sequences */
+            } else {
+                start=limit+1;
+            }
+        }
+    } while(--maxLevel>=minLevel);
+}
+
+/* API functions for logical<->visual mapping ------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getVisualIndex(UBiDi *pBiDi, int32_t logicalIndex, UErrorCode *pErrorCode) {
+    int32_t visualIndex=UBIDI_MAP_NOWHERE;
+    RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrorCode, -1);
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode, -1);
+    RETURN_IF_BAD_RANGE(logicalIndex, 0, pBiDi->length, *pErrorCode, -1);
+
+    /* we can do the trivial cases without the runs array */
+    switch(pBiDi->direction) {
+    case UBIDI_LTR:
+        visualIndex=logicalIndex;
+        break;
+    case UBIDI_RTL:
+        visualIndex=pBiDi->length-logicalIndex-1;
+        break;
+    default:
+        if(!ubidi_getRuns(pBiDi, pErrorCode)) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return -1;
+        } else {
+            Run *runs=pBiDi->runs;
+            int32_t i, visualStart=0, offset, length;
+
+            /* linear search for the run, search on the visual runs */
+            for(i=0; i<pBiDi->runCount; ++i) {
+                length=runs[i].visualLimit-visualStart;
+                offset=logicalIndex-GET_INDEX(runs[i].logicalStart);
+                if(offset>=0 && offset<length) {
+                    if(IS_EVEN_RUN(runs[i].logicalStart)) {
+                        /* LTR */
+                        visualIndex=visualStart+offset;
+                    } else {
+                        /* RTL */
+                        visualIndex=visualStart+length-offset-1;
+                    }
+                    break;          /* exit for loop */
+                }
+                visualStart+=length;
+            }
+            if(i>=pBiDi->runCount) {
+                return UBIDI_MAP_NOWHERE;
+            }
+        }
+    }
+
+    if(pBiDi->insertPoints.size>0) {
+        /* add the number of added marks until the calculated visual index */
+        Run *runs=pBiDi->runs;
+        int32_t i, length, insertRemove;
+        int32_t visualStart=0, markFound=0;
+        for(i=0; ; i++, visualStart+=length) {
+            length=runs[i].visualLimit-visualStart;
+            insertRemove=runs[i].insertRemove;
+            if(insertRemove & (LRM_BEFORE|RLM_BEFORE)) {
+                markFound++;
+            }
+            /* is it the run containing the visual index? */
+            if(visualIndex<runs[i].visualLimit) {
+                return visualIndex+markFound;
+            }
+            if(insertRemove & (LRM_AFTER|RLM_AFTER)) {
+                markFound++;
+            }
+        }
+    }
+    else if(pBiDi->controlCount>0) {
+        /* subtract the number of controls until the calculated visual index */
+        Run *runs=pBiDi->runs;
+        int32_t i, j, start, limit, length, insertRemove;
+        int32_t visualStart=0, controlFound=0;
+        UChar uchar=pBiDi->text[logicalIndex];
+        /* is the logical index pointing to a control ? */
+        if(IS_BIDI_CONTROL_CHAR(uchar)) {
+            return UBIDI_MAP_NOWHERE;
+        }
+        /* loop on runs */
+        for(i=0; ; i++, visualStart+=length) {
+            length=runs[i].visualLimit-visualStart;
+            insertRemove=runs[i].insertRemove;
+            /* calculated visual index is beyond this run? */
+            if(visualIndex>=runs[i].visualLimit) {
+                controlFound-=insertRemove;
+                continue;
+            }
+            /* calculated visual index must be within current run */
+            if(insertRemove==0) {
+                return visualIndex-controlFound;
+            }
+            if(IS_EVEN_RUN(runs[i].logicalStart)) {
+                /* LTR: check from run start to logical index */
+                start=runs[i].logicalStart;
+                limit=logicalIndex;
+            } else {
+                /* RTL: check from logical index to run end */
+                start=logicalIndex+1;
+                limit=GET_INDEX(runs[i].logicalStart)+length;
+            }
+            for(j=start; j<limit; j++) {
+                uchar=pBiDi->text[j];
+                if(IS_BIDI_CONTROL_CHAR(uchar)) {
+                    controlFound++;
+                }
+            }
+            return visualIndex-controlFound;
+        }
+    }
+
+    return visualIndex;
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_getLogicalIndex(UBiDi *pBiDi, int32_t visualIndex, UErrorCode *pErrorCode) {
+    Run *runs;
+    int32_t i, runCount, start;
+    RETURN_IF_NULL_OR_FAILING_ERRCODE(pErrorCode, -1);
+    RETURN_IF_NOT_VALID_PARA_OR_LINE(pBiDi, *pErrorCode, -1);
+    RETURN_IF_BAD_RANGE(visualIndex, 0, pBiDi->resultLength, *pErrorCode, -1);
+    /* we can do the trivial cases without the runs array */
+    if(pBiDi->insertPoints.size==0 && pBiDi->controlCount==0) {
+        if(pBiDi->direction==UBIDI_LTR) {
+            return visualIndex;
+        }
+        else if(pBiDi->direction==UBIDI_RTL) {
+            return pBiDi->length-visualIndex-1;
+        }
+    }
+    if(!ubidi_getRuns(pBiDi, pErrorCode)) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return -1;
+    }
+
+    runs=pBiDi->runs;
+    runCount=pBiDi->runCount;
+    if(pBiDi->insertPoints.size>0) {
+        /* handle inserted LRM/RLM */
+        int32_t markFound=0, insertRemove;
+        int32_t visualStart=0, length;
+        runs=pBiDi->runs;
+        /* subtract number of marks until visual index */
+        for(i=0; ; i++, visualStart+=length) {
+            length=runs[i].visualLimit-visualStart;
+            insertRemove=runs[i].insertRemove;
+            if(insertRemove&(LRM_BEFORE|RLM_BEFORE)) {
+                if(visualIndex<=(visualStart+markFound)) {
+                    return UBIDI_MAP_NOWHERE;
+                }
+                markFound++;
+            }
+            /* is adjusted visual index within this run? */
+            if(visualIndex<(runs[i].visualLimit+markFound)) {
+                visualIndex-=markFound;
+                break;
+            }
+            if(insertRemove&(LRM_AFTER|RLM_AFTER)) {
+                if(visualIndex==(visualStart+length+markFound)) {
+                    return UBIDI_MAP_NOWHERE;
+                }
+                markFound++;
+            }
+        }
+    }
+    else if(pBiDi->controlCount>0) {
+        /* handle removed BiDi control characters */
+        int32_t controlFound=0, insertRemove, length;
+        int32_t logicalStart, logicalEnd, visualStart=0, j, k;
+        UChar uchar;
+        UBool evenRun;
+        /* add number of controls until visual index */
+        for(i=0; ; i++, visualStart+=length) {
+            length=runs[i].visualLimit-visualStart;
+            insertRemove=runs[i].insertRemove;
+            /* is adjusted visual index beyond current run? */
+            if(visualIndex>=(runs[i].visualLimit-controlFound+insertRemove)) {
+                controlFound-=insertRemove;
+                continue;
+            }
+            /* adjusted visual index is within current run */
+            if(insertRemove==0) {
+                visualIndex+=controlFound;
+                break;
+            }
+            /* count non-control chars until visualIndex */
+            logicalStart=runs[i].logicalStart;
+            evenRun=IS_EVEN_RUN(logicalStart);
+            REMOVE_ODD_BIT(logicalStart);
+            logicalEnd=logicalStart+length-1;
+            for(j=0; j<length; j++) {
+                k= evenRun ? logicalStart+j : logicalEnd-j;
+                uchar=pBiDi->text[k];
+                if(IS_BIDI_CONTROL_CHAR(uchar)) {
+                    controlFound++;
+                }
+                if((visualIndex+controlFound)==(visualStart+j)) {
+                    break;
+                }
+            }
+            visualIndex+=controlFound;
+            break;
+        }
+    }
+    /* handle all cases */
+    if(runCount<=10) {
+        /* linear search for the run */
+        for(i=0; visualIndex>=runs[i].visualLimit; ++i) {}
+    } else {
+        /* binary search for the run */
+        int32_t begin=0, limit=runCount;
+
+        /* the middle if() is guaranteed to find the run, we don't need a loop limit */
+        for(;;) {
+            i=(begin+limit)/2;
+            if(visualIndex>=runs[i].visualLimit) {
+                begin=i+1;
+            } else if(i==0 || visualIndex>=runs[i-1].visualLimit) {
+                break;
+            } else {
+                limit=i;
+            }
+        }
+    }
+
+    start=runs[i].logicalStart;
+    if(IS_EVEN_RUN(start)) {
+        /* LTR */
+        /* the offset in runs[i] is visualIndex-runs[i-1].visualLimit */
+        if(i>0) {
+            visualIndex-=runs[i-1].visualLimit;
+        }
+        return start+visualIndex;
+    } else {
+        /* RTL */
+        return GET_INDEX(start)+runs[i].visualLimit-visualIndex-1;
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_getLogicalMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode) {
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    /* ubidi_countRuns() checks for VALID_PARA_OR_LINE */
+    ubidi_countRuns(pBiDi, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        /* no op */
+    } else if(indexMap==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else {
+        /* fill a logical-to-visual index map using the runs[] */
+        int32_t visualStart, visualLimit, i, j, k;
+        int32_t logicalStart, logicalLimit;
+        Run *runs=pBiDi->runs;
+        if (pBiDi->length<=0) {
+            return;
+        }
+        if (pBiDi->length>pBiDi->resultLength) {
+            uprv_memset(indexMap, 0xFF, pBiDi->length*sizeof(int32_t));
+        }
+
+        visualStart=0;
+        for(j=0; j<pBiDi->runCount; ++j) {
+            logicalStart=GET_INDEX(runs[j].logicalStart);
+            visualLimit=runs[j].visualLimit;
+            if(IS_EVEN_RUN(runs[j].logicalStart)) {
+                do { /* LTR */
+                    indexMap[logicalStart++]=visualStart++;
+                } while(visualStart<visualLimit);
+            } else {
+                logicalStart+=visualLimit-visualStart;  /* logicalLimit */
+                do { /* RTL */
+                    indexMap[--logicalStart]=visualStart++;
+                } while(visualStart<visualLimit);
+            }
+            /* visualStart==visualLimit; */
+        }
+
+        if(pBiDi->insertPoints.size>0) {
+            int32_t markFound=0, runCount=pBiDi->runCount;
+            int32_t length, insertRemove;
+            visualStart=0;
+            /* add number of marks found until each index */
+            for(i=0; i<runCount; i++, visualStart+=length) {
+                length=runs[i].visualLimit-visualStart;
+                insertRemove=runs[i].insertRemove;
+                if(insertRemove&(LRM_BEFORE|RLM_BEFORE)) {
+                    markFound++;
+                }
+                if(markFound>0) {
+                    logicalStart=GET_INDEX(runs[i].logicalStart);
+                    logicalLimit=logicalStart+length;
+                    for(j=logicalStart; j<logicalLimit; j++) {
+                        indexMap[j]+=markFound;
+                    }
+                }
+                if(insertRemove&(LRM_AFTER|RLM_AFTER)) {
+                    markFound++;
+                }
+            }
+        }
+        else if(pBiDi->controlCount>0) {
+            int32_t controlFound=0, runCount=pBiDi->runCount;
+            int32_t length, insertRemove;
+            UBool evenRun;
+            UChar uchar;
+            visualStart=0;
+            /* subtract number of controls found until each index */
+            for(i=0; i<runCount; i++, visualStart+=length) {
+                length=runs[i].visualLimit-visualStart;
+                insertRemove=runs[i].insertRemove;
+                /* no control found within previous runs nor within this run */
+                if((controlFound-insertRemove)==0) {
+                    continue;
+                }
+                logicalStart=runs[i].logicalStart;
+                evenRun=IS_EVEN_RUN(logicalStart);
+                REMOVE_ODD_BIT(logicalStart);
+                logicalLimit=logicalStart+length;
+                /* if no control within this run */
+                if(insertRemove==0) {
+                    for(j=logicalStart; j<logicalLimit; j++) {
+                        indexMap[j]-=controlFound;
+                    }
+                    continue;
+                }
+                for(j=0; j<length; j++) {
+                    k= evenRun ? logicalStart+j : logicalLimit-j-1;
+                    uchar=pBiDi->text[k];
+                    if(IS_BIDI_CONTROL_CHAR(uchar)) {
+                        controlFound++;
+                        indexMap[k]=UBIDI_MAP_NOWHERE;
+                        continue;
+                    }
+                    indexMap[k]-=controlFound;
+                }
+            }
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_getVisualMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode) {
+    RETURN_VOID_IF_NULL_OR_FAILING_ERRCODE(pErrorCode);
+    if(indexMap==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    /* ubidi_countRuns() checks for VALID_PARA_OR_LINE */
+    ubidi_countRuns(pBiDi, pErrorCode);
+    if(U_SUCCESS(*pErrorCode)) {
+        /* fill a visual-to-logical index map using the runs[] */
+        Run *runs=pBiDi->runs, *runsLimit=runs+pBiDi->runCount;
+        int32_t logicalStart, visualStart, visualLimit, *pi=indexMap;
+
+        if (pBiDi->resultLength<=0) {
+            return;
+        }
+        visualStart=0;
+        for(; runs<runsLimit; ++runs) {
+            logicalStart=runs->logicalStart;
+            visualLimit=runs->visualLimit;
+            if(IS_EVEN_RUN(logicalStart)) {
+                do { /* LTR */
+                    *pi++ = logicalStart++;
+                } while(++visualStart<visualLimit);
+            } else {
+                REMOVE_ODD_BIT(logicalStart);
+                logicalStart+=visualLimit-visualStart;  /* logicalLimit */
+                do { /* RTL */
+                    *pi++ = --logicalStart;
+                } while(++visualStart<visualLimit);
+            }
+            /* visualStart==visualLimit; */
+        }
+
+        if(pBiDi->insertPoints.size>0) {
+            int32_t markFound=0, runCount=pBiDi->runCount;
+            int32_t insertRemove, i, j, k;
+            runs=pBiDi->runs;
+            /* count all inserted marks */
+            for(i=0; i<runCount; i++) {
+                insertRemove=runs[i].insertRemove;
+                if(insertRemove&(LRM_BEFORE|RLM_BEFORE)) {
+                    markFound++;
+                }
+                if(insertRemove&(LRM_AFTER|RLM_AFTER)) {
+                    markFound++;
+                }
+            }
+            /* move back indexes by number of preceding marks */
+            k=pBiDi->resultLength;
+            for(i=runCount-1; i>=0 && markFound>0; i--) {
+                insertRemove=runs[i].insertRemove;
+                if(insertRemove&(LRM_AFTER|RLM_AFTER)) {
+                    indexMap[--k]= UBIDI_MAP_NOWHERE;
+                    markFound--;
+                }
+                visualStart= i>0 ? runs[i-1].visualLimit : 0;
+                for(j=runs[i].visualLimit-1; j>=visualStart && markFound>0; j--) {
+                    indexMap[--k]=indexMap[j];
+                }
+                if(insertRemove&(LRM_BEFORE|RLM_BEFORE)) {
+                    indexMap[--k]= UBIDI_MAP_NOWHERE;
+                    markFound--;
+                }
+            }
+        }
+        else if(pBiDi->controlCount>0) {
+            int32_t runCount=pBiDi->runCount, logicalEnd;
+            int32_t insertRemove, length, i, j, k, m;
+            UChar uchar;
+            UBool evenRun;
+            runs=pBiDi->runs;
+            visualStart=0;
+            /* move forward indexes by number of preceding controls */
+            k=0;
+            for(i=0; i<runCount; i++, visualStart+=length) {
+                length=runs[i].visualLimit-visualStart;
+                insertRemove=runs[i].insertRemove;
+                /* if no control found yet, nothing to do in this run */
+                if((insertRemove==0)&&(k==visualStart)) {
+                    k+=length;
+                    continue;
+                }
+                /* if no control in this run */
+                if(insertRemove==0) {
+                    visualLimit=runs[i].visualLimit;
+                    for(j=visualStart; j<visualLimit; j++) {
+                        indexMap[k++]=indexMap[j];
+                    }
+                    continue;
+                }
+                logicalStart=runs[i].logicalStart;
+                evenRun=IS_EVEN_RUN(logicalStart);
+                REMOVE_ODD_BIT(logicalStart);
+                logicalEnd=logicalStart+length-1;
+                for(j=0; j<length; j++) {
+                    m= evenRun ? logicalStart+j : logicalEnd-j;
+                    uchar=pBiDi->text[m];
+                    if(!IS_BIDI_CONTROL_CHAR(uchar)) {
+                        indexMap[k++]=m;
+                    }
+                }
+            }
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+ubidi_invertMap(const int32_t *srcMap, int32_t *destMap, int32_t length) {
+    if(srcMap!=NULL && destMap!=NULL && length>0) {
+        const int32_t *pi;
+        int32_t destLength=-1, count=0;
+        /* find highest value and count positive indexes in srcMap */
+        pi=srcMap+length;
+        while(pi>srcMap) {
+            if(*--pi>destLength) {
+                destLength=*pi;
+            }
+            if(*pi>=0) {
+                count++;
+            }
+        }
+        destLength++;           /* add 1 for origin 0 */
+        if(count<destLength) {
+            /* we must fill unmatched destMap entries with -1 */
+            uprv_memset(destMap, 0xFF, destLength*sizeof(int32_t));
+        }
+        pi=srcMap+length;
+        while(length>0) {
+            if(*--pi>=0) {
+                destMap[*pi]=--length;
+            } else {
+                --length;
+            }
+        }
+    }
+}
diff --git a/thirdparty/icu4c/common/ubiditransform.cpp b/thirdparty/icu4c/common/ubiditransform.cpp
new file mode 100644
index 0000000000..d56bf1518b
--- /dev/null
+++ b/thirdparty/icu4c/common/ubiditransform.cpp
@@ -0,0 +1,530 @@
+/*
+******************************************************************************
+*
+* © 2016 and later: Unicode, Inc. and others.
+* License & terms of use: http://www.unicode.org/copyright.html
+*
+******************************************************************************
+*   file name:  ubiditransform.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2016jul24
+*   created by: Lina Kemmel
+*
+*/
+
+#include "cmemory.h"
+#include "unicode/ubidi.h"
+#include "unicode/ustring.h"
+#include "unicode/ushape.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+#include "unicode/ubiditransform.h"
+
+/* Some convenience defines */
+#define LTR                     UBIDI_LTR
+#define RTL                     UBIDI_RTL
+#define LOGICAL                 UBIDI_LOGICAL
+#define VISUAL                  UBIDI_VISUAL
+#define SHAPE_LOGICAL           U_SHAPE_TEXT_DIRECTION_LOGICAL
+#define SHAPE_VISUAL            U_SHAPE_TEXT_DIRECTION_VISUAL_LTR
+
+#define CHECK_LEN(STR, LEN, ERROR) UPRV_BLOCK_MACRO_BEGIN { \
+    if (LEN == 0) return 0; \
+    if (LEN < -1) { *(ERROR) = U_ILLEGAL_ARGUMENT_ERROR; return 0; } \
+    if (LEN == -1) LEN = u_strlen(STR); \
+} UPRV_BLOCK_MACRO_END
+
+#define MAX_ACTIONS     7
+
+/**
+ * Typedef for a pointer to a function, which performs some operation (such as
+ * reordering, setting "inverse" mode, character mirroring, etc.). Return value
+ * indicates whether the text was changed in the course of this operation or
+ * not.
+ */
+typedef UBool (*UBiDiAction)(UBiDiTransform *, UErrorCode *);
+
+/**
+ * Structure that holds a predefined reordering scheme, including the following
+ * information:
+ * <ul>
+ * <li>an input base direction,</li>
+ * <li>an input order,</li>
+ * <li>an output base direction,</li>
+ * <li>an output order,</li>
+ * <li>a digit shaping direction,</li>
+ * <li>a letter shaping direction,</li>
+ * <li>a base direction that should be applied when the reordering engine is
+ *     invoked (which can not always be derived from the caller-defined
+ *     options),</li>
+ * <li>an array of pointers to functions that accomplish the bidi layout
+ *     transformation.</li>
+ * </ul>
+ */
+typedef struct {
+    UBiDiLevel        inLevel;               /* input level */
+    UBiDiOrder        inOrder;               /* input order */
+    UBiDiLevel        outLevel;              /* output level */
+    UBiDiOrder        outOrder;              /* output order */
+    uint32_t          digitsDir;             /* digit shaping direction */
+    uint32_t          lettersDir;            /* letter shaping direction */
+    UBiDiLevel        baseLevel;             /* paragraph level to be used with setPara */
+    const UBiDiAction actions[MAX_ACTIONS];  /* array of pointers to functions carrying out the transformation */
+} ReorderingScheme;
+
+struct UBiDiTransform {
+    UBiDi                   *pBidi;             /* pointer to a UBiDi object */
+    const ReorderingScheme  *pActiveScheme;     /* effective reordering scheme */
+    UChar                   *src;               /* input text */
+    UChar                   *dest;              /* output text */
+    uint32_t                srcLength;          /* input text length - not really needed as we are zero-terminated and can u_strlen */
+    uint32_t                srcSize;            /* input text capacity excluding the trailing zero */
+    uint32_t                destSize;           /* output text capacity */
+    uint32_t                *pDestLength;       /* number of UChars written to dest */
+    uint32_t                reorderingOptions;  /* reordering options - currently only suppot DO_MIRRORING */
+    uint32_t                digits;             /* digit option for ArabicShaping */
+    uint32_t                letters;            /* letter option for ArabicShaping */
+};
+
+U_CAPI UBiDiTransform* U_EXPORT2
+ubiditransform_open(UErrorCode *pErrorCode)
+{
+    UBiDiTransform *pBiDiTransform = NULL;
+    if (U_SUCCESS(*pErrorCode)) {
+        pBiDiTransform = (UBiDiTransform*) uprv_calloc(1, sizeof(UBiDiTransform));
+        if (pBiDiTransform == NULL) {
+            *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+    return pBiDiTransform;
+}
+
+U_CAPI void U_EXPORT2
+ubiditransform_close(UBiDiTransform *pBiDiTransform)
+{
+    if (pBiDiTransform != NULL) {
+        if (pBiDiTransform->pBidi != NULL) {
+            ubidi_close(pBiDiTransform->pBidi);
+        }
+        if (pBiDiTransform->src != NULL) {
+            uprv_free(pBiDiTransform->src);
+        }
+        uprv_free(pBiDiTransform);
+    }
+}
+
+/**
+ * Performs Bidi resolution of text.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_resolve(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    ubidi_setPara(pTransform->pBidi, pTransform->src, pTransform->srcLength,
+            pTransform->pActiveScheme->baseLevel, NULL, pErrorCode);
+    return FALSE;
+}
+
+/**
+ * Performs basic reordering of text (Logical -> Visual LTR).
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_reorder(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    ubidi_writeReordered(pTransform->pBidi, pTransform->dest, pTransform->destSize,
+            static_cast<uint16_t>(pTransform->reorderingOptions), pErrorCode);
+
+    *pTransform->pDestLength = pTransform->srcLength;
+    pTransform->reorderingOptions = UBIDI_REORDER_DEFAULT;
+    return TRUE;
+}
+
+/**
+ * Sets "inverse" mode on the <code>UBiDi</code> object.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_setInverse(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    (void)pErrorCode;
+    ubidi_setInverse(pTransform->pBidi, TRUE);
+    ubidi_setReorderingMode(pTransform->pBidi, UBIDI_REORDER_INVERSE_LIKE_DIRECT);
+    return FALSE;
+}
+
+/**
+ * Sets "runs only" reordering mode indicating a Logical LTR <-> Logical RTL
+ * transformation.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_setRunsOnly(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    (void)pErrorCode;
+    ubidi_setReorderingMode(pTransform->pBidi, UBIDI_REORDER_RUNS_ONLY);
+    return FALSE;
+}
+
+/**
+ * Performs string reverse.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_reverse(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    ubidi_writeReverse(pTransform->src, pTransform->srcLength,
+            pTransform->dest, pTransform->destSize,
+            UBIDI_REORDER_DEFAULT, pErrorCode);
+    *pTransform->pDestLength = pTransform->srcLength;
+    return TRUE;
+}
+
+/**
+ * Applies a new value to the text that serves as input at the current
+ * processing step. This value is identical to the original one when we begin
+ * the processing, but usually changes as the transformation progresses.
+ * 
+ * @param pTransform A pointer to the <code>UBiDiTransform</code> structure.
+ * @param newSrc A pointer whose value is to be used as input text.
+ * @param newLength A length of the new text in <code>UChar</code>s.
+ * @param newSize A new source capacity in <code>UChar</code>s.
+ * @param pErrorCode Pointer to the error code value.
+ */
+static void
+updateSrc(UBiDiTransform *pTransform, const UChar *newSrc, uint32_t newLength,
+        uint32_t newSize, UErrorCode *pErrorCode)
+{
+    if (newSize < newLength) {
+        *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+        return;
+    }
+    if (newSize > pTransform->srcSize) {
+        newSize += 50; // allocate slightly more than needed right now
+        if (pTransform->src != NULL) {
+            uprv_free(pTransform->src);
+            pTransform->src = NULL;
+        }
+        pTransform->src = (UChar *)uprv_malloc(newSize * sizeof(UChar));
+        if (pTransform->src == NULL) {
+            *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+            //pTransform->srcLength = pTransform->srcSize = 0;
+            return;
+        }
+        pTransform->srcSize = newSize;
+    }
+    u_strncpy(pTransform->src, newSrc, newLength);
+    pTransform->srcLength = u_terminateUChars(pTransform->src,
+    		pTransform->srcSize, newLength, pErrorCode);
+}
+
+/**
+ * Calls a lower level shaping function.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param options Shaping options.
+ * @param pErrorCode Pointer to the error code value.
+ */
+static void
+doShape(UBiDiTransform *pTransform, uint32_t options, UErrorCode *pErrorCode)
+{
+    *pTransform->pDestLength = u_shapeArabic(pTransform->src,
+            pTransform->srcLength, pTransform->dest, pTransform->destSize,
+            options, pErrorCode);
+}
+
+/**
+ * Performs digit and letter shaping.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_shapeArabic(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    if ((pTransform->letters | pTransform->digits) == 0) {
+        return FALSE;
+    }
+    if (pTransform->pActiveScheme->lettersDir == pTransform->pActiveScheme->digitsDir) {
+        doShape(pTransform, pTransform->letters | pTransform->digits | pTransform->pActiveScheme->lettersDir,
+                pErrorCode);
+    } else {
+        doShape(pTransform, pTransform->digits | pTransform->pActiveScheme->digitsDir, pErrorCode);
+        if (U_SUCCESS(*pErrorCode)) {
+            updateSrc(pTransform, pTransform->dest, *pTransform->pDestLength,
+                    *pTransform->pDestLength, pErrorCode);
+            doShape(pTransform, pTransform->letters | pTransform->pActiveScheme->lettersDir,
+                    pErrorCode);
+        }
+    }
+    return TRUE;
+}
+
+/**
+ * Performs character mirroring.
+ * 
+ * @param pTransform Pointer to the <code>UBiDiTransform</code> structure.
+ * @param pErrorCode Pointer to the error code value.
+ *
+ * @return Whether or not this function modifies the text. Besides the return
+ * value, the caller should also check <code>U_SUCCESS(*pErrorCode)</code>.
+ */
+static UBool
+action_mirror(UBiDiTransform *pTransform, UErrorCode *pErrorCode)
+{
+    UChar32 c;
+    uint32_t i = 0, j = 0;
+    if (0 == (pTransform->reorderingOptions & UBIDI_DO_MIRRORING)) {
+        return FALSE;
+    }
+    if (pTransform->destSize < pTransform->srcLength) {
+        *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+        return FALSE;
+    }
+    do {
+        UBool isOdd = ubidi_getLevelAt(pTransform->pBidi, i) & 1;
+        U16_NEXT(pTransform->src, i, pTransform->srcLength, c); 
+        U16_APPEND_UNSAFE(pTransform->dest, j, isOdd ? u_charMirror(c) : c);
+    } while (i < pTransform->srcLength);
+    
+    *pTransform->pDestLength = pTransform->srcLength;
+    pTransform->reorderingOptions = UBIDI_REORDER_DEFAULT;
+    return TRUE;
+}
+
+/**
+ * All possible reordering schemes.
+ *
+ */
+static const ReorderingScheme Schemes[] =
+{
+    /* 0: Logical LTR => Visual LTR */
+    {LTR, LOGICAL, LTR, VISUAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_shapeArabic, action_resolve, action_reorder, NULL}},
+    /* 1: Logical RTL => Visual LTR */
+    {RTL, LOGICAL, LTR, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, RTL,
+            {action_resolve, action_reorder, action_shapeArabic, NULL}},
+    /* 2: Logical LTR => Visual RTL */
+    {LTR, LOGICAL, RTL, VISUAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_shapeArabic, action_resolve, action_reorder, action_reverse, NULL}},
+    /* 3: Logical RTL => Visual RTL */
+    {RTL, LOGICAL, RTL, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, RTL,
+            {action_resolve, action_reorder, action_shapeArabic, action_reverse, NULL}},
+    /* 4: Visual LTR => Logical RTL */
+    {LTR, VISUAL, RTL, LOGICAL, SHAPE_LOGICAL, SHAPE_VISUAL, RTL,
+            {action_shapeArabic, action_setInverse, action_resolve, action_reorder, NULL}},
+    /* 5: Visual RTL => Logical RTL */
+    {RTL, VISUAL, RTL, LOGICAL, SHAPE_LOGICAL, SHAPE_VISUAL, RTL,
+            {action_reverse, action_shapeArabic, action_setInverse, action_resolve, action_reorder, NULL}},
+    /* 6: Visual LTR => Logical LTR */
+    {LTR, VISUAL, LTR, LOGICAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_setInverse, action_resolve, action_reorder, action_shapeArabic, NULL}},
+    /* 7: Visual RTL => Logical LTR */
+    {RTL, VISUAL, LTR, LOGICAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_reverse, action_setInverse, action_resolve, action_reorder, action_shapeArabic, NULL}},
+    /* 8: Logical LTR => Logical RTL */
+    {LTR, LOGICAL, RTL, LOGICAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_shapeArabic, action_resolve, action_mirror, action_setRunsOnly, action_resolve, action_reorder, NULL}},
+    /* 9: Logical RTL => Logical LTR */
+    {RTL, LOGICAL, LTR, LOGICAL, SHAPE_LOGICAL, SHAPE_LOGICAL, RTL,
+            {action_resolve, action_mirror, action_setRunsOnly, action_resolve, action_reorder, action_shapeArabic, NULL}},
+    /* 10: Visual LTR => Visual RTL */
+    {LTR, VISUAL, RTL, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, LTR,
+            {action_shapeArabic, action_setInverse, action_resolve, action_mirror, action_reverse, NULL}},
+    /* 11: Visual RTL => Visual LTR */
+    {RTL, VISUAL, LTR, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, LTR,
+            {action_reverse, action_shapeArabic, action_setInverse, action_resolve, action_mirror, NULL}},
+    /* 12: Logical LTR => Logical LTR */
+    {LTR, LOGICAL, LTR, LOGICAL, SHAPE_LOGICAL, SHAPE_LOGICAL, LTR,
+            {action_resolve, action_mirror, action_shapeArabic, NULL}},
+    /* 13: Logical RTL => Logical RTL */
+    {RTL, LOGICAL, RTL, LOGICAL, SHAPE_VISUAL, SHAPE_LOGICAL, RTL,
+            {action_resolve, action_mirror, action_shapeArabic, NULL}},
+    /* 14: Visual LTR => Visual LTR */
+    {LTR, VISUAL, LTR, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, LTR,
+            {action_resolve, action_mirror, action_shapeArabic, NULL}},
+    /* 15: Visual RTL => Visual RTL */
+    {RTL, VISUAL, RTL, VISUAL, SHAPE_LOGICAL, SHAPE_VISUAL, LTR,
+            {action_reverse, action_resolve, action_mirror, action_shapeArabic, action_reverse, NULL}}
+};
+
+static const uint32_t nSchemes = sizeof(Schemes) / sizeof(*Schemes);
+
+/**
+ * When the direction option is <code>UBIDI_DEFAULT_LTR</code> or
+ * <code>UBIDI_DEFAULT_RTL</code>, resolve the base direction according to that
+ * of the first strong bidi character.
+ */
+static void
+resolveBaseDirection(const UChar *text, uint32_t length,
+        UBiDiLevel *pInLevel, UBiDiLevel *pOutLevel)
+{
+    switch (*pInLevel) {
+        case UBIDI_DEFAULT_LTR:
+        case UBIDI_DEFAULT_RTL: {
+            UBiDiLevel level = static_cast<UBiDiLevel>(ubidi_getBaseDirection(text, length));
+            *pInLevel = static_cast<UBiDiLevel>(level != UBIDI_NEUTRAL) ? level
+                    : *pInLevel == UBIDI_DEFAULT_RTL ? static_cast<UBiDiLevel>(RTL) : static_cast<UBiDiLevel>(LTR);
+            break;
+        }
+        default:
+            *pInLevel &= 1;
+            break;
+    }
+    switch (*pOutLevel) {
+        case UBIDI_DEFAULT_LTR:
+        case UBIDI_DEFAULT_RTL:
+            *pOutLevel = *pInLevel;
+            break;
+        default:
+            *pOutLevel &= 1;
+            break;
+    }
+}
+
+/**
+ * Finds a valid <code>ReorderingScheme</code> matching the
+ * caller-defined scheme.
+ * 
+ * @return A valid <code>ReorderingScheme</code> object or NULL
+ */
+static const ReorderingScheme*
+findMatchingScheme(UBiDiLevel inLevel, UBiDiLevel outLevel,
+        UBiDiOrder inOrder, UBiDiOrder outOrder)
+{
+    uint32_t i;
+    for (i = 0; i < nSchemes; i++) {
+        const ReorderingScheme *pScheme = Schemes + i;
+        if (inLevel == pScheme->inLevel && outLevel == pScheme->outLevel
+                && inOrder == pScheme->inOrder && outOrder == pScheme->outOrder) {
+            return pScheme;
+        }
+    }
+    return NULL;
+}
+
+U_CAPI uint32_t U_EXPORT2
+ubiditransform_transform(UBiDiTransform *pBiDiTransform,
+            const UChar *src, int32_t srcLength,
+            UChar *dest, int32_t destSize,
+            UBiDiLevel inParaLevel, UBiDiOrder inOrder,
+            UBiDiLevel outParaLevel, UBiDiOrder outOrder,
+            UBiDiMirroring doMirroring, uint32_t shapingOptions,
+            UErrorCode *pErrorCode)
+{
+    uint32_t destLength = 0;
+    UBool textChanged = FALSE;
+    const UBiDiTransform *pOrigTransform = pBiDiTransform;
+    const UBiDiAction *action = NULL;
+
+    if (U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if (src == NULL || dest == NULL) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    CHECK_LEN(src, srcLength, pErrorCode);
+    CHECK_LEN(dest, destSize, pErrorCode);
+
+    if (pBiDiTransform == NULL) {
+        pBiDiTransform = ubiditransform_open(pErrorCode);
+        if (U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+    }
+    /* Current limitation: in multiple paragraphs will be resolved according
+       to the 1st paragraph */
+    resolveBaseDirection(src, srcLength, &inParaLevel, &outParaLevel);
+
+    pBiDiTransform->pActiveScheme = findMatchingScheme(inParaLevel, outParaLevel,
+            inOrder, outOrder);
+    if (pBiDiTransform->pActiveScheme == NULL) {
+        goto cleanup;
+    }
+    pBiDiTransform->reorderingOptions = doMirroring ? UBIDI_DO_MIRRORING
+            : UBIDI_REORDER_DEFAULT;
+
+    /* Ignore TEXT_DIRECTION_* flags, as we apply our own depending on the text
+       scheme at the time shaping is invoked. */
+    shapingOptions &= ~U_SHAPE_TEXT_DIRECTION_MASK;
+    pBiDiTransform->digits = shapingOptions & ~U_SHAPE_LETTERS_MASK;
+    pBiDiTransform->letters = shapingOptions & ~U_SHAPE_DIGITS_MASK;
+
+    updateSrc(pBiDiTransform, src, srcLength, destSize > srcLength ? destSize : srcLength, pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {
+        goto cleanup;
+    }
+    if (pBiDiTransform->pBidi == NULL) {
+        pBiDiTransform->pBidi = ubidi_openSized(0, 0, pErrorCode);
+        if (U_FAILURE(*pErrorCode)) {
+            goto cleanup;
+        }
+    }
+    pBiDiTransform->dest = dest;
+    pBiDiTransform->destSize = destSize;
+    pBiDiTransform->pDestLength = &destLength;
+
+    /* Checking for U_SUCCESS() within the loop to bail out on first failure. */
+    for (action = pBiDiTransform->pActiveScheme->actions; *action && U_SUCCESS(*pErrorCode); action++) {
+        if ((*action)(pBiDiTransform, pErrorCode)) {
+            if (action + 1) {
+                updateSrc(pBiDiTransform, pBiDiTransform->dest, *pBiDiTransform->pDestLength,
+                        *pBiDiTransform->pDestLength, pErrorCode);
+            }
+            textChanged = TRUE;
+        }
+    }
+    ubidi_setInverse(pBiDiTransform->pBidi, FALSE);
+
+    if (!textChanged && U_SUCCESS(*pErrorCode)) {
+        /* Text was not changed - just copy src to dest */
+        if (destSize < srcLength) {
+            *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+        } else {
+            u_strncpy(dest, src, srcLength);
+            destLength = srcLength;
+        }
+    }
+cleanup:
+    if (pOrigTransform != pBiDiTransform) {
+        ubiditransform_close(pBiDiTransform);
+    } else {
+        pBiDiTransform->dest = NULL;
+        pBiDiTransform->pDestLength = NULL;
+        pBiDiTransform->srcLength = 0;
+        pBiDiTransform->destSize = 0;
+    }
+    return U_FAILURE(*pErrorCode) ? 0 : destLength;
+}
diff --git a/thirdparty/icu4c/common/ubidiwrt.cpp b/thirdparty/icu4c/common/ubidiwrt.cpp
new file mode 100644
index 0000000000..a69c0a4b8b
--- /dev/null
+++ b/thirdparty/icu4c/common/ubidiwrt.cpp
@@ -0,0 +1,650 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2000-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ubidiwrt.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999aug06
+*   created by: Markus W. Scherer, updated by Matitiahu Allouche
+*
+* This file contains implementations for BiDi functions that use
+* the core algorithm and core API to write reordered text.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/uchar.h"
+#include "unicode/ubidi.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "ustr_imp.h"
+#include "ubidiimp.h"
+
+/*
+ * The function implementations in this file are designed
+ * for UTF-16 and UTF-32, not for UTF-8.
+ *
+ * Assumptions that are not true for UTF-8:
+ * - Any code point always needs the same number of code units
+ *   ("minimum-length-problem" of UTF-8)
+ * - The BiDi control characters need only one code unit each
+ *
+ * Further assumptions for all UTFs:
+ * - u_charMirror(c) needs the same number of code units as c
+ */
+#if defined(UTF_SIZE) && UTF_SIZE==8
+# error reimplement ubidi_writeReordered() for UTF-8, see comment above
+#endif
+
+#define IS_COMBINING(type) ((1UL<<(type))&(1UL<<U_NON_SPACING_MARK|1UL<<U_COMBINING_SPACING_MARK|1UL<<U_ENCLOSING_MARK))
+
+/*
+ * When we have UBIDI_OUTPUT_REVERSE set on ubidi_writeReordered(), then we
+ * semantically write RTL runs in reverse and later reverse them again.
+ * Instead, we actually write them in forward order to begin with.
+ * However, if the RTL run was to be mirrored, we need to mirror here now
+ * since the implicit second reversal must not do it.
+ * It looks strange to do mirroring in LTR output, but it is only because
+ * we are writing RTL output in reverse.
+ */
+static int32_t
+doWriteForward(const UChar *src, int32_t srcLength,
+               UChar *dest, int32_t destSize,
+               uint16_t options,
+               UErrorCode *pErrorCode) {
+    /* optimize for several combinations of options */
+    switch(options&(UBIDI_REMOVE_BIDI_CONTROLS|UBIDI_DO_MIRRORING)) {
+    case 0: {
+        /* simply copy the LTR run to the destination */
+        int32_t length=srcLength;
+        if(destSize<length) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return srcLength;
+        }
+        do {
+            *dest++=*src++;
+        } while(--length>0);
+        return srcLength;
+    }
+    case UBIDI_DO_MIRRORING: {
+        /* do mirroring */
+        int32_t i=0, j=0;
+        UChar32 c;
+
+        if(destSize<srcLength) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return srcLength;
+        }
+        do {
+            U16_NEXT(src, i, srcLength, c);
+            c=u_charMirror(c);
+            U16_APPEND_UNSAFE(dest, j, c);
+        } while(i<srcLength);
+        return srcLength;
+    }
+    case UBIDI_REMOVE_BIDI_CONTROLS: {
+        /* copy the LTR run and remove any BiDi control characters */
+        int32_t remaining=destSize;
+        UChar c;
+        do {
+            c=*src++;
+            if(!IS_BIDI_CONTROL_CHAR(c)) {
+                if(--remaining<0) {
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+
+                    /* preflight the length */
+                    while(--srcLength>0) {
+                        c=*src++;
+                        if(!IS_BIDI_CONTROL_CHAR(c)) {
+                            --remaining;
+                        }
+                    }
+                    return destSize-remaining;
+                }
+                *dest++=c;
+            }
+        } while(--srcLength>0);
+        return destSize-remaining;
+    }
+    default: {
+        /* remove BiDi control characters and do mirroring */
+        int32_t remaining=destSize;
+        int32_t i, j=0;
+        UChar32 c;
+        do {
+            i=0;
+            U16_NEXT(src, i, srcLength, c);
+            src+=i;
+            srcLength-=i;
+            if(!IS_BIDI_CONTROL_CHAR(c)) {
+                remaining-=i;
+                if(remaining<0) {
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+
+                    /* preflight the length */
+                    while(srcLength>0) {
+                        c=*src++;
+                        if(!IS_BIDI_CONTROL_CHAR(c)) {
+                            --remaining;
+                        }
+                        --srcLength;
+                    }
+                    return destSize-remaining;
+                }
+                c=u_charMirror(c);
+                U16_APPEND_UNSAFE(dest, j, c);
+            }
+        } while(srcLength>0);
+        return j;
+    }
+    } /* end of switch */
+}
+
+static int32_t
+doWriteReverse(const UChar *src, int32_t srcLength,
+               UChar *dest, int32_t destSize,
+               uint16_t options,
+               UErrorCode *pErrorCode) {
+    /*
+     * RTL run -
+     *
+     * RTL runs need to be copied to the destination in reverse order
+     * of code points, not code units, to keep Unicode characters intact.
+     *
+     * The general strategy for this is to read the source text
+     * in backward order, collect all code units for a code point
+     * (and optionally following combining characters, see below),
+     * and copy all these code units in ascending order
+     * to the destination for this run.
+     *
+     * Several options request whether combining characters
+     * should be kept after their base characters,
+     * whether BiDi control characters should be removed, and
+     * whether characters should be replaced by their mirror-image
+     * equivalent Unicode characters.
+     */
+    int32_t i, j;
+    UChar32 c;
+
+    /* optimize for several combinations of options */
+    switch(options&(UBIDI_REMOVE_BIDI_CONTROLS|UBIDI_DO_MIRRORING|UBIDI_KEEP_BASE_COMBINING)) {
+    case 0:
+        /*
+         * With none of the "complicated" options set, the destination
+         * run will have the same length as the source run,
+         * and there is no mirroring and no keeping combining characters
+         * with their base characters.
+         */
+        if(destSize<srcLength) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return srcLength;
+        }
+        destSize=srcLength;
+
+        /* preserve character integrity */
+        do {
+            /* i is always after the last code unit known to need to be kept in this segment */
+            i=srcLength;
+
+            /* collect code units for one base character */
+            U16_BACK_1(src, 0, srcLength);
+
+            /* copy this base character */
+            j=srcLength;
+            do {
+                *dest++=src[j++];
+            } while(j<i);
+        } while(srcLength>0);
+        break;
+    case UBIDI_KEEP_BASE_COMBINING:
+        /*
+         * Here, too, the destination
+         * run will have the same length as the source run,
+         * and there is no mirroring.
+         * We do need to keep combining characters with their base characters.
+         */
+        if(destSize<srcLength) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return srcLength;
+        }
+        destSize=srcLength;
+
+        /* preserve character integrity */
+        do {
+            /* i is always after the last code unit known to need to be kept in this segment */
+            i=srcLength;
+
+            /* collect code units and modifier letters for one base character */
+            do {
+                U16_PREV(src, 0, srcLength, c);
+            } while(srcLength>0 && IS_COMBINING(u_charType(c)));
+
+            /* copy this "user character" */
+            j=srcLength;
+            do {
+                *dest++=src[j++];
+            } while(j<i);
+        } while(srcLength>0);
+        break;
+    default:
+        /*
+         * With several "complicated" options set, this is the most
+         * general and the slowest copying of an RTL run.
+         * We will do mirroring, remove BiDi controls, and
+         * keep combining characters with their base characters
+         * as requested.
+         */
+        if(!(options&UBIDI_REMOVE_BIDI_CONTROLS)) {
+            i=srcLength;
+        } else {
+            /* we need to find out the destination length of the run,
+               which will not include the BiDi control characters */
+            int32_t length=srcLength;
+            UChar ch;
+
+            i=0;
+            do {
+                ch=*src++;
+                if(!IS_BIDI_CONTROL_CHAR(ch)) {
+                    ++i;
+                }
+            } while(--length>0);
+            src-=srcLength;
+        }
+
+        if(destSize<i) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return i;
+        }
+        destSize=i;
+
+        /* preserve character integrity */
+        do {
+            /* i is always after the last code unit known to need to be kept in this segment */
+            i=srcLength;
+
+            /* collect code units for one base character */
+            U16_PREV(src, 0, srcLength, c);
+            if(options&UBIDI_KEEP_BASE_COMBINING) {
+                /* collect modifier letters for this base character */
+                while(srcLength>0 && IS_COMBINING(u_charType(c))) {
+                    U16_PREV(src, 0, srcLength, c);
+                }
+            }
+
+            if(options&UBIDI_REMOVE_BIDI_CONTROLS && IS_BIDI_CONTROL_CHAR(c)) {
+                /* do not copy this BiDi control character */
+                continue;
+            }
+
+            /* copy this "user character" */
+            j=srcLength;
+            if(options&UBIDI_DO_MIRRORING) {
+                /* mirror only the base character */
+                int32_t k=0;
+                c=u_charMirror(c);
+                U16_APPEND_UNSAFE(dest, k, c);
+                dest+=k;
+                j+=k;
+            }
+            while(j<i) {
+                *dest++=src[j++];
+            }
+        } while(srcLength>0);
+        break;
+    } /* end of switch */
+
+    return destSize;
+}
+
+U_CAPI int32_t U_EXPORT2
+ubidi_writeReverse(const UChar *src, int32_t srcLength,
+                   UChar *dest, int32_t destSize,
+                   uint16_t options,
+                   UErrorCode *pErrorCode) {
+    int32_t destLength;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* more error checking */
+    if( src==NULL || srcLength<-1 ||
+        destSize<0 || (destSize>0 && dest==NULL))
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* do input and output overlap? */
+    if( dest!=NULL &&
+        ((src>=dest && src<dest+destSize) ||
+         (dest>=src && dest<src+srcLength)))
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(srcLength==-1) {
+        srcLength=u_strlen(src);
+    }
+    if(srcLength>0) {
+        destLength=doWriteReverse(src, srcLength, dest, destSize, options, pErrorCode);
+    } else {
+        /* nothing to do */
+        destLength=0;
+    }
+
+    return u_terminateUChars(dest, destSize, destLength, pErrorCode);
+}
+
+// Ticket 20907 - The optimizer in MSVC/Visual Studio versions below 16.4 has trouble with this
+// function on Windows ARM64. As a work-around, we disable optimizations for this function.
+// This work-around could/should be removed once the following versions of Visual Studio are no
+// longer supported: All versions of VS2017, and versions of VS2019 below 16.4.
+#if (defined(_MSC_VER) && (defined(_M_ARM64)) && (_MSC_VER < 1924))
+#pragma optimize( "", off )
+#endif
+U_CAPI int32_t U_EXPORT2
+ubidi_writeReordered(UBiDi *pBiDi,
+                     UChar *dest, int32_t destSize,
+                     uint16_t options,
+                     UErrorCode *pErrorCode) {
+    const UChar *text;
+    UChar *saveDest;
+    int32_t length, destCapacity;
+    int32_t run, runCount, logicalStart, runLength;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* more error checking */
+    if( pBiDi==NULL ||
+        (text=pBiDi->text)==NULL || (length=pBiDi->length)<0 ||
+        destSize<0 || (destSize>0 && dest==NULL))
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* do input and output overlap? */
+    if( dest!=NULL &&
+        ((text>=dest && text<dest+destSize) ||
+         (dest>=text && dest<text+pBiDi->originalLength)))
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(length==0) {
+        /* nothing to do */
+        return u_terminateUChars(dest, destSize, 0, pErrorCode);
+    }
+
+    runCount=ubidi_countRuns(pBiDi, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* destSize shrinks, later destination length=destCapacity-destSize */
+    saveDest=dest;
+    destCapacity=destSize;
+
+    /*
+     * Option "insert marks" implies UBIDI_INSERT_LRM_FOR_NUMERIC if the
+     * reordering mode (checked below) is appropriate.
+     */
+    if(pBiDi->reorderingOptions & UBIDI_OPTION_INSERT_MARKS) {
+        options|=UBIDI_INSERT_LRM_FOR_NUMERIC;
+        options&=~UBIDI_REMOVE_BIDI_CONTROLS;
+    }
+    /*
+     * Option "remove controls" implies UBIDI_REMOVE_BIDI_CONTROLS
+     * and cancels UBIDI_INSERT_LRM_FOR_NUMERIC.
+     */
+    if(pBiDi->reorderingOptions & UBIDI_OPTION_REMOVE_CONTROLS) {
+        options|=UBIDI_REMOVE_BIDI_CONTROLS;
+        options&=~UBIDI_INSERT_LRM_FOR_NUMERIC;
+    }
+    /*
+     * If we do not perform the "inverse BiDi" algorithm, then we
+     * don't need to insert any LRMs, and don't need to test for it.
+     */
+    if((pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_NUMBERS_AS_L) &&
+       (pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_LIKE_DIRECT)  &&
+       (pBiDi->reorderingMode != UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL) &&
+       (pBiDi->reorderingMode != UBIDI_REORDER_RUNS_ONLY)) {
+        options&=~UBIDI_INSERT_LRM_FOR_NUMERIC;
+    }
+    /*
+     * Iterate through all visual runs and copy the run text segments to
+     * the destination, according to the options.
+     *
+     * The tests for where to insert LRMs ignore the fact that there may be
+     * BN codes or non-BMP code points at the beginning and end of a run;
+     * they may insert LRMs unnecessarily but the tests are faster this way
+     * (this would have to be improved for UTF-8).
+     *
+     * Note that the only errors that are set by doWriteXY() are buffer overflow
+     * errors. Ignore them until the end, and continue for preflighting.
+     */
+    if(!(options&UBIDI_OUTPUT_REVERSE)) {
+        /* forward output */
+        if(!(options&UBIDI_INSERT_LRM_FOR_NUMERIC)) {
+            /* do not insert BiDi controls */
+            for(run=0; run<runCount; ++run) {
+                if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength)) {
+                    runLength=doWriteForward(text+logicalStart, runLength,
+                                             dest, destSize,
+                                             (uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
+                } else {
+                    runLength=doWriteReverse(text+logicalStart, runLength,
+                                             dest, destSize,
+                                             options, pErrorCode);
+                }
+                if(dest!=NULL) {
+                  dest+=runLength;
+                }
+                destSize-=runLength;
+            }
+        } else {
+            /* insert BiDi controls for "inverse BiDi" */
+            const DirProp *dirProps=pBiDi->dirProps;
+            const UChar *src;
+            UChar uc;
+            UBiDiDirection dir;
+            int32_t markFlag;
+
+            for(run=0; run<runCount; ++run) {
+                dir=ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength);
+                src=text+logicalStart;
+                /* check if something relevant in insertPoints */
+                markFlag=pBiDi->runs[run].insertRemove;
+                if(markFlag<0) {        /* BiDi controls count */
+                    markFlag=0;
+                }
+
+                if(UBIDI_LTR==dir) {
+                    if((pBiDi->isInverse) &&
+                       (/*run>0 &&*/ dirProps[logicalStart]!=L)) {
+                        markFlag |= LRM_BEFORE;
+                    }
+                    if (markFlag & LRM_BEFORE) {
+                        uc=LRM_CHAR;
+                    }
+                    else if (markFlag & RLM_BEFORE) {
+                        uc=RLM_CHAR;
+                    }
+                    else  uc=0;
+                    if(uc) {
+                        if(destSize>0) {
+                            *dest++=uc;
+                        }
+                        --destSize;
+                    }
+
+                    runLength=doWriteForward(src, runLength,
+                                             dest, destSize,
+                                             (uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
+                    if(dest!=NULL) {
+                      dest+=runLength;
+                    }
+                    destSize-=runLength;
+
+                    if((pBiDi->isInverse) &&
+                       (/*run<runCount-1 &&*/ dirProps[logicalStart+runLength-1]!=L)) {
+                        markFlag |= LRM_AFTER;
+                    }
+                    if (markFlag & LRM_AFTER) {
+                        uc=LRM_CHAR;
+                    }
+                    else if (markFlag & RLM_AFTER) {
+                        uc=RLM_CHAR;
+                    }
+                    else  uc=0;
+                    if(uc) {
+                        if(destSize>0) {
+                            *dest++=uc;
+                        }
+                        --destSize;
+                    }
+                } else {                /* RTL run */
+                    if((pBiDi->isInverse) &&
+                       (/*run>0 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart+runLength-1])))) {
+                        markFlag |= RLM_BEFORE;
+                    }
+                    if (markFlag & LRM_BEFORE) {
+                        uc=LRM_CHAR;
+                    }
+                    else if (markFlag & RLM_BEFORE) {
+                        uc=RLM_CHAR;
+                    }
+                    else  uc=0;
+                    if(uc) {
+                        if(destSize>0) {
+                            *dest++=uc;
+                        }
+                        --destSize;
+                    }
+
+                    runLength=doWriteReverse(src, runLength,
+                                             dest, destSize,
+                                             options, pErrorCode);
+                    if(dest!=NULL) {
+                      dest+=runLength;
+                    }
+                    destSize-=runLength;
+
+                    if((pBiDi->isInverse) &&
+                       (/*run<runCount-1 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart])))) {
+                        markFlag |= RLM_AFTER;
+                    }
+                    if (markFlag & LRM_AFTER) {
+                        uc=LRM_CHAR;
+                    }
+                    else if (markFlag & RLM_AFTER) {
+                        uc=RLM_CHAR;
+                    }
+                    else  uc=0;
+                    if(uc) {
+                        if(destSize>0) {
+                            *dest++=uc;
+                        }
+                        --destSize;
+                    }
+                }
+            }
+        }
+    } else {
+        /* reverse output */
+        if(!(options&UBIDI_INSERT_LRM_FOR_NUMERIC)) {
+            /* do not insert BiDi controls */
+            for(run=runCount; --run>=0;) {
+                if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength)) {
+                    runLength=doWriteReverse(text+logicalStart, runLength,
+                                             dest, destSize,
+                                             (uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
+                } else {
+                    runLength=doWriteForward(text+logicalStart, runLength,
+                                             dest, destSize,
+                                             options, pErrorCode);
+                }
+                if(dest!=NULL) {
+                  dest+=runLength;
+                }
+                destSize-=runLength;
+            }
+        } else {
+            /* insert BiDi controls for "inverse BiDi" */
+            const DirProp *dirProps=pBiDi->dirProps;
+            const UChar *src;
+            UBiDiDirection dir;
+
+            for(run=runCount; --run>=0;) {
+                /* reverse output */
+                dir=ubidi_getVisualRun(pBiDi, run, &logicalStart, &runLength);
+                src=text+logicalStart;
+
+                if(UBIDI_LTR==dir) {
+                    if(/*run<runCount-1 &&*/ dirProps[logicalStart+runLength-1]!=L) {
+                        if(destSize>0) {
+                            *dest++=LRM_CHAR;
+                        }
+                        --destSize;
+                    }
+
+                    runLength=doWriteReverse(src, runLength,
+                                             dest, destSize,
+                                             (uint16_t)(options&~UBIDI_DO_MIRRORING), pErrorCode);
+                    if(dest!=NULL) {
+                      dest+=runLength;
+                    }
+                    destSize-=runLength;
+
+                    if(/*run>0 &&*/ dirProps[logicalStart]!=L) {
+                        if(destSize>0) {
+                            *dest++=LRM_CHAR;
+                        }
+                        --destSize;
+                    }
+                } else {
+                    if(/*run<runCount-1 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart]))) {
+                        if(destSize>0) {
+                            *dest++=RLM_CHAR;
+                        }
+                        --destSize;
+                    }
+
+                    runLength=doWriteForward(src, runLength,
+                                             dest, destSize,
+                                             options, pErrorCode);
+                    if(dest!=NULL) {
+                      dest+=runLength;
+                    }
+                    destSize-=runLength;
+
+                    if(/*run>0 &&*/ !(MASK_R_AL&DIRPROP_FLAG(dirProps[logicalStart+runLength-1]))) {
+                        if(destSize>0) {
+                            *dest++=RLM_CHAR;
+                        }
+                        --destSize;
+                    }
+                }
+            }
+        }
+    }
+
+    return u_terminateUChars(saveDest, destCapacity, destCapacity-destSize, pErrorCode);
+}
+#if (defined(_MSC_VER) && (defined(_M_ARM64)) && (_MSC_VER < 1924))
+#pragma optimize( "", on )
+#endif
diff --git a/thirdparty/icu4c/common/ubrk.cpp b/thirdparty/icu4c/common/ubrk.cpp
new file mode 100644
index 0000000000..f8bdf5a6b6
--- /dev/null
+++ b/thirdparty/icu4c/common/ubrk.cpp
@@ -0,0 +1,357 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 1996-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/ubrk.h"
+
+#include "unicode/brkiter.h"
+#include "unicode/uloc.h"
+#include "unicode/ustring.h"
+#include "unicode/uchriter.h"
+#include "unicode/rbbi.h"
+#include "rbbirb.h"
+#include "uassert.h"
+#include "cmemory.h"
+
+U_NAMESPACE_USE
+
+//------------------------------------------------------------------------------
+//
+//    ubrk_open      Create a canned type of break iterator based on type (word, line, etc.)
+//                   and locale.
+//
+//------------------------------------------------------------------------------
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_open(UBreakIteratorType type,
+      const char *locale,
+      const UChar *text,
+      int32_t textLength,
+      UErrorCode *status)
+{
+
+  if(U_FAILURE(*status)) return 0;
+
+  BreakIterator *result = 0;
+
+  switch(type) {
+
+  case UBRK_CHARACTER:
+    result = BreakIterator::createCharacterInstance(Locale(locale), *status);
+    break;
+
+  case UBRK_WORD:
+    result = BreakIterator::createWordInstance(Locale(locale), *status);
+    break;
+
+  case UBRK_LINE:
+    result = BreakIterator::createLineInstance(Locale(locale), *status);
+    break;
+
+  case UBRK_SENTENCE:
+    result = BreakIterator::createSentenceInstance(Locale(locale), *status);
+    break;
+
+  case UBRK_TITLE:
+    result = BreakIterator::createTitleInstance(Locale(locale), *status);
+    break;
+
+  default:
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+  }
+
+  // check for allocation error
+  if (U_FAILURE(*status)) {
+     return 0;
+  }
+  if(result == 0) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return 0;
+  }
+
+
+  UBreakIterator *uBI = (UBreakIterator *)result;
+  if (text != NULL) {
+      ubrk_setText(uBI, text, textLength, status);
+  }
+  return uBI;
+}
+
+
+
+//------------------------------------------------------------------------------
+//
+//   ubrk_openRules      open a break iterator from a set of break rules.
+//                       Invokes the rule builder.
+//
+//------------------------------------------------------------------------------
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_openRules(  const UChar        *rules,
+                       int32_t       rulesLength,
+                 const UChar        *text,
+                       int32_t       textLength,
+                       UParseError  *parseErr,
+                       UErrorCode   *status)  {
+
+    if (status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+
+    BreakIterator *result = 0;
+    UnicodeString ruleString(rules, rulesLength);
+    result = RBBIRuleBuilder::createRuleBasedBreakIterator(ruleString, parseErr, *status);
+    if(U_FAILURE(*status)) {
+        return 0;
+    }
+
+    UBreakIterator *uBI = (UBreakIterator *)result;
+    if (text != NULL) {
+        ubrk_setText(uBI, text, textLength, status);
+    }
+    return uBI;
+}
+
+
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_openBinaryRules(const uint8_t *binaryRules, int32_t rulesLength,
+                     const UChar *  text, int32_t textLength,
+                     UErrorCode *   status)
+{
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (rulesLength < 0) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    LocalPointer<RuleBasedBreakIterator> lpRBBI(new RuleBasedBreakIterator(binaryRules, rulesLength, *status), *status);
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    UBreakIterator *uBI = reinterpret_cast<UBreakIterator *>(lpRBBI.orphan());
+    if (text != NULL) {
+        ubrk_setText(uBI, text, textLength, status);
+    }
+    return uBI;
+}
+
+
+U_CAPI UBreakIterator * U_EXPORT2
+ubrk_safeClone(
+          const UBreakIterator *bi,
+          void * /*stackBuffer*/,
+          int32_t *pBufferSize,
+          UErrorCode *status)
+{
+    if (status == NULL || U_FAILURE(*status)){
+        return NULL;
+    }
+    if (bi == NULL) {
+       *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    if (pBufferSize != NULL) {
+        int32_t inputSize = *pBufferSize;
+        *pBufferSize = 1;
+        if (inputSize == 0) {
+            return NULL;  // preflighting for deprecated functionality
+        }
+    }
+    BreakIterator *newBI = ((BreakIterator *)bi)->clone();
+    if (newBI == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        *status = U_SAFECLONE_ALLOCATED_WARNING;
+    }
+    return (UBreakIterator *)newBI;
+}
+
+
+
+U_CAPI void U_EXPORT2
+ubrk_close(UBreakIterator *bi)
+{
+    delete (BreakIterator *)bi;
+}
+
+U_CAPI void U_EXPORT2
+ubrk_setText(UBreakIterator* bi,
+             const UChar*    text,
+             int32_t         textLength,
+             UErrorCode*     status)
+{
+    UText  ut = UTEXT_INITIALIZER;
+    utext_openUChars(&ut, text, textLength, status);
+    ((BreakIterator*)bi)->setText(&ut, *status);
+    // A stack allocated UText wrapping a UChar * string
+    //   can be dumped without explicitly closing it.
+}
+
+
+
+U_CAPI void U_EXPORT2
+ubrk_setUText(UBreakIterator *bi,
+             UText          *text,
+             UErrorCode     *status)
+{
+  ((BreakIterator*)bi)->setText(text, *status);
+}
+
+
+
+
+
+U_CAPI int32_t U_EXPORT2
+ubrk_current(const UBreakIterator *bi)
+{
+
+  return ((BreakIterator*)bi)->current();
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_next(UBreakIterator *bi)
+{
+
+  return ((BreakIterator*)bi)->next();
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_previous(UBreakIterator *bi)
+{
+
+  return ((BreakIterator*)bi)->previous();
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_first(UBreakIterator *bi)
+{
+
+  return ((BreakIterator*)bi)->first();
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_last(UBreakIterator *bi)
+{
+
+  return ((BreakIterator*)bi)->last();
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_preceding(UBreakIterator *bi,
+           int32_t offset)
+{
+
+  return ((BreakIterator*)bi)->preceding(offset);
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_following(UBreakIterator *bi,
+           int32_t offset)
+{
+
+  return ((BreakIterator*)bi)->following(offset);
+}
+
+U_CAPI const char* U_EXPORT2
+ubrk_getAvailable(int32_t index)
+{
+
+  return uloc_getAvailable(index);
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_countAvailable()
+{
+
+  return uloc_countAvailable();
+}
+
+
+U_CAPI  UBool U_EXPORT2
+ubrk_isBoundary(UBreakIterator *bi, int32_t offset)
+{
+    return ((BreakIterator*)bi)->isBoundary(offset);
+}
+
+
+U_CAPI  int32_t U_EXPORT2
+ubrk_getRuleStatus(UBreakIterator *bi)
+{
+    return ((BreakIterator*)bi)->getRuleStatus();
+}
+
+U_CAPI  int32_t U_EXPORT2
+ubrk_getRuleStatusVec(UBreakIterator *bi, int32_t *fillInVec, int32_t capacity, UErrorCode *status)
+{
+    return ((BreakIterator*)bi)->getRuleStatusVec(fillInVec, capacity, *status);
+}
+
+
+U_CAPI const char* U_EXPORT2
+ubrk_getLocaleByType(const UBreakIterator *bi,
+                     ULocDataLocaleType type,
+                     UErrorCode* status)
+{
+    if (bi == NULL) {
+        if (U_SUCCESS(*status)) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return NULL;
+    }
+    return ((BreakIterator*)bi)->getLocaleID(type, *status);
+}
+
+
+U_CAPI void U_EXPORT2
+ubrk_refreshUText(UBreakIterator *bi,
+                       UText          *text,
+                       UErrorCode     *status)
+{
+    BreakIterator *bii = reinterpret_cast<BreakIterator *>(bi);
+    bii->refreshInputText(text, *status);
+}
+
+U_CAPI int32_t U_EXPORT2
+ubrk_getBinaryRules(UBreakIterator *bi,
+                    uint8_t *       binaryRules, int32_t rulesCapacity,
+                    UErrorCode *    status)
+{
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+    if ((binaryRules == NULL && rulesCapacity > 0) || rulesCapacity < 0) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    RuleBasedBreakIterator* rbbi;
+    if ((rbbi = dynamic_cast<RuleBasedBreakIterator*>(reinterpret_cast<BreakIterator*>(bi))) == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    uint32_t rulesLength;
+    const uint8_t * returnedRules = rbbi->getBinaryRules(rulesLength);
+    if (rulesLength > INT32_MAX) {
+        *status = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    if (binaryRules != NULL) { // if not preflighting
+        // Here we know rulesLength <= INT32_MAX and rulesCapacity >= 0, can cast safely
+        if ((int32_t)rulesLength > rulesCapacity) {
+            *status = U_BUFFER_OVERFLOW_ERROR;
+        } else {
+            uprv_memcpy(binaryRules, returnedRules, rulesLength);
+        }
+    }
+    return (int32_t)rulesLength;
+}
+
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
diff --git a/thirdparty/icu4c/common/ubrkimpl.h b/thirdparty/icu4c/common/ubrkimpl.h
new file mode 100644
index 0000000000..8197f66339
--- /dev/null
+++ b/thirdparty/icu4c/common/ubrkimpl.h
@@ -0,0 +1,15 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef UBRKIMPL_H
+#define UBRKIMPL_H
+
+#define U_ICUDATA_BRKITR U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "brkitr"
+
+#endif /*UBRKIMPL_H*/
diff --git a/thirdparty/icu4c/common/ucase.cpp b/thirdparty/icu4c/common/ucase.cpp
new file mode 100644
index 0000000000..2b142f5bc2
--- /dev/null
+++ b/thirdparty/icu4c/common/ucase.cpp
@@ -0,0 +1,1608 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucase.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug30
+*   created by: Markus W. Scherer
+*
+*   Low-level Unicode character/string case mapping code.
+*   Much code moved here (and modified) from uchar.c.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/uset.h"
+#include "unicode/udata.h" /* UDataInfo */
+#include "unicode/utf16.h"
+#include "ucmndata.h" /* DataHeader */
+#include "udatamem.h"
+#include "umutex.h"
+#include "uassert.h"
+#include "cmemory.h"
+#include "utrie2.h"
+#include "ucase.h"
+
+struct UCaseProps {
+    UDataMemory *mem;
+    const int32_t *indexes;
+    const uint16_t *exceptions;
+    const uint16_t *unfold;
+
+    UTrie2 trie;
+    uint8_t formatVersion[4];
+};
+
+/* ucase_props_data.h is machine-generated by gencase --csource */
+#define INCLUDED_FROM_UCASE_CPP
+#include "ucase_props_data.h"
+
+/* set of property starts for UnicodeSet ------------------------------------ */
+
+static UBool U_CALLCONV
+_enumPropertyStartsRange(const void *context, UChar32 start, UChar32 /*end*/, uint32_t /*value*/) {
+    /* add the start code point to the USet */
+    const USetAdder *sa=(const USetAdder *)context;
+    sa->add(sa->set, start);
+    return TRUE;
+}
+
+U_CFUNC void U_EXPORT2
+ucase_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /* add the start code point of each same-value range of the trie */
+    utrie2_enum(&ucase_props_singleton.trie, NULL, _enumPropertyStartsRange, sa);
+
+    /* add code points with hardcoded properties, plus the ones following them */
+
+    /* (none right now, see comment below) */
+
+    /*
+     * Omit code points with hardcoded specialcasing properties
+     * because we do not build property UnicodeSets for them right now.
+     */
+}
+
+/* data access primitives --------------------------------------------------- */
+
+U_CFUNC const UTrie2 * U_EXPORT2
+ucase_getTrie() {
+    return &ucase_props_singleton.trie;
+}
+
+#define GET_EXCEPTIONS(csp, props) ((csp)->exceptions+((props)>>UCASE_EXC_SHIFT))
+
+/* number of bits in an 8-bit integer value */
+static const uint8_t flagsOffset[256]={
+    0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
+    1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
+    1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
+    1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
+    2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
+    3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
+    3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
+    4, 5, 5, 6, 5, 6, 6, 7, 5, 6, 6, 7, 6, 7, 7, 8
+};
+
+#define HAS_SLOT(flags, idx) ((flags)&(1<<(idx)))
+#define SLOT_OFFSET(flags, idx) flagsOffset[(flags)&((1<<(idx))-1)]
+
+/*
+ * Get the value of an optional-value slot where HAS_SLOT(excWord, idx).
+ *
+ * @param excWord (in) initial exceptions word
+ * @param idx (in) desired slot index
+ * @param pExc16 (in/out) const uint16_t * after excWord=*pExc16++;
+ *               moved to the last uint16_t of the value, use +1 for beginning of next slot
+ * @param value (out) int32_t or uint32_t output if hasSlot, otherwise not modified
+ */
+#define GET_SLOT_VALUE(excWord, idx, pExc16, value) UPRV_BLOCK_MACRO_BEGIN { \
+    if(((excWord)&UCASE_EXC_DOUBLE_SLOTS)==0) { \
+        (pExc16)+=SLOT_OFFSET(excWord, idx); \
+        (value)=*pExc16; \
+    } else { \
+        (pExc16)+=2*SLOT_OFFSET(excWord, idx); \
+        (value)=*pExc16++; \
+        (value)=((value)<<16)|*pExc16; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* simple case mappings ----------------------------------------------------- */
+
+U_CAPI UChar32 U_EXPORT2
+ucase_tolower(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_IS_UPPER_OR_TITLE(props)) {
+            c+=UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        uint16_t excWord=*pe++;
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_IS_UPPER_OR_TITLE(props)) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_LOWER)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_LOWER, pe, c);
+        }
+    }
+    return c;
+}
+
+U_CAPI UChar32 U_EXPORT2
+ucase_toupper(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            c+=UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        uint16_t excWord=*pe++;
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_UPPER)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_UPPER, pe, c);
+        }
+    }
+    return c;
+}
+
+U_CAPI UChar32 U_EXPORT2
+ucase_totitle(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            c+=UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        uint16_t excWord=*pe++;
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        int32_t idx;
+        if(HAS_SLOT(excWord, UCASE_EXC_TITLE)) {
+            idx=UCASE_EXC_TITLE;
+        } else if(HAS_SLOT(excWord, UCASE_EXC_UPPER)) {
+            idx=UCASE_EXC_UPPER;
+        } else {
+            return c;
+        }
+        GET_SLOT_VALUE(excWord, idx, pe, c);
+    }
+    return c;
+}
+
+static const UChar iDot[2] = { 0x69, 0x307 };
+static const UChar jDot[2] = { 0x6a, 0x307 };
+static const UChar iOgonekDot[3] = { 0x12f, 0x307 };
+static const UChar iDotGrave[3] = { 0x69, 0x307, 0x300 };
+static const UChar iDotAcute[3] = { 0x69, 0x307, 0x301 };
+static const UChar iDotTilde[3] = { 0x69, 0x307, 0x303 };
+
+
+U_CFUNC void U_EXPORT2
+ucase_addCaseClosure(UChar32 c, const USetAdder *sa) {
+    uint16_t props;
+
+    /*
+     * Hardcode the case closure of i and its relatives and ignore the
+     * data file data for these characters.
+     * The Turkic dotless i and dotted I with their case mapping conditions
+     * and case folding option make the related characters behave specially.
+     * This code matches their closure behavior to their case folding behavior.
+     */
+
+    switch(c) {
+    case 0x49:
+        /* regular i and I are in one equivalence class */
+        sa->add(sa->set, 0x69);
+        return;
+    case 0x69:
+        sa->add(sa->set, 0x49);
+        return;
+    case 0x130:
+        /* dotted I is in a class with <0069 0307> (for canonical equivalence with <0049 0307>) */
+        sa->addString(sa->set, iDot, 2);
+        return;
+    case 0x131:
+        /* dotless i is in a class by itself */
+        return;
+    default:
+        /* otherwise use the data file data */
+        break;
+    }
+
+    props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_GET_TYPE(props)!=UCASE_NONE) {
+            /* add the one simple case mapping, no matter what type it is */
+            int32_t delta=UCASE_GET_DELTA(props);
+            if(delta!=0) {
+                sa->add(sa->set, c+delta);
+            }
+        }
+    } else {
+        /*
+         * c has exceptions, so there may be multiple simple and/or
+         * full case mappings. Add them all.
+         */
+        const uint16_t *pe0, *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        const UChar *closure;
+        uint16_t excWord=*pe++;
+        int32_t idx, closureLength, fullLength, length;
+
+        pe0=pe;
+
+        /* add all simple case mappings */
+        for(idx=UCASE_EXC_LOWER; idx<=UCASE_EXC_TITLE; ++idx) {
+            if(HAS_SLOT(excWord, idx)) {
+                pe=pe0;
+                GET_SLOT_VALUE(excWord, idx, pe, c);
+                sa->add(sa->set, c);
+            }
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA)) {
+            pe=pe0;
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe, delta);
+            sa->add(sa->set, (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta);
+        }
+
+        /* get the closure string pointer & length */
+        if(HAS_SLOT(excWord, UCASE_EXC_CLOSURE)) {
+            pe=pe0;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_CLOSURE, pe, closureLength);
+            closureLength&=UCASE_CLOSURE_MAX_LENGTH; /* higher bits are reserved */
+            closure=(const UChar *)pe+1; /* behind this slot, unless there are full case mappings */
+        } else {
+            closureLength=0;
+            closure=NULL;
+        }
+
+        /* add the full case folding */
+        if(HAS_SLOT(excWord, UCASE_EXC_FULL_MAPPINGS)) {
+            pe=pe0;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_FULL_MAPPINGS, pe, fullLength);
+
+            /* start of full case mapping strings */
+            ++pe;
+
+            fullLength&=0xffff; /* bits 16 and higher are reserved */
+
+            /* skip the lowercase result string */
+            pe+=fullLength&UCASE_FULL_LOWER;
+            fullLength>>=4;
+
+            /* add the full case folding string */
+            length=fullLength&0xf;
+            if(length!=0) {
+                sa->addString(sa->set, (const UChar *)pe, length);
+                pe+=length;
+            }
+
+            /* skip the uppercase and titlecase strings */
+            fullLength>>=4;
+            pe+=fullLength&0xf;
+            fullLength>>=4;
+            pe+=fullLength;
+
+            closure=(const UChar *)pe; /* behind full case mappings */
+        }
+
+        /* add each code point in the closure string */
+        for(idx=0; idx<closureLength;) {
+            U16_NEXT_UNSAFE(closure, idx, c);
+            sa->add(sa->set, c);
+        }
+    }
+}
+
+/*
+ * compare s, which has a length, with t, which has a maximum length or is NUL-terminated
+ * must be length>0 and max>0 and length<=max
+ */
+static inline int32_t
+strcmpMax(const UChar *s, int32_t length, const UChar *t, int32_t max) {
+    int32_t c1, c2;
+
+    max-=length; /* we require length<=max, so no need to decrement max in the loop */
+    do {
+        c1=*s++;
+        c2=*t++;
+        if(c2==0) {
+            return 1; /* reached the end of t but not of s */
+        }
+        c1-=c2;
+        if(c1!=0) {
+            return c1; /* return difference result */
+        }
+    } while(--length>0);
+    /* ends with length==0 */
+
+    if(max==0 || *t==0) {
+        return 0; /* equal to length of both strings */
+    } else {
+        return -max; /* return lengh difference */
+    }
+}
+
+U_CFUNC UBool U_EXPORT2
+ucase_addStringCaseClosure(const UChar *s, int32_t length, const USetAdder *sa) {
+    int32_t i, start, limit, result, unfoldRows, unfoldRowWidth, unfoldStringWidth;
+
+    if(ucase_props_singleton.unfold==NULL || s==NULL) {
+        return FALSE; /* no reverse case folding data, or no string */
+    }
+    if(length<=1) {
+        /* the string is too short to find any match */
+        /*
+         * more precise would be:
+         * if(!u_strHasMoreChar32Than(s, length, 1))
+         * but this does not make much practical difference because
+         * a single supplementary code point would just not be found
+         */
+        return FALSE;
+    }
+
+    const uint16_t *unfold=ucase_props_singleton.unfold;
+    unfoldRows=unfold[UCASE_UNFOLD_ROWS];
+    unfoldRowWidth=unfold[UCASE_UNFOLD_ROW_WIDTH];
+    unfoldStringWidth=unfold[UCASE_UNFOLD_STRING_WIDTH];
+    unfold+=unfoldRowWidth;
+
+    if(length>unfoldStringWidth) {
+        /* the string is too long to find any match */
+        return FALSE;
+    }
+
+    /* do a binary search for the string */
+    start=0;
+    limit=unfoldRows;
+    while(start<limit) {
+        i=(start+limit)/2;
+        const UChar *p=reinterpret_cast<const UChar *>(unfold+(i*unfoldRowWidth));
+        result=strcmpMax(s, length, p, unfoldStringWidth);
+
+        if(result==0) {
+            /* found the string: add each code point, and its case closure */
+            UChar32 c;
+
+            for(i=unfoldStringWidth; i<unfoldRowWidth && p[i]!=0;) {
+                U16_NEXT_UNSAFE(p, i, c);
+                sa->add(sa->set, c);
+                ucase_addCaseClosure(c, sa);
+            }
+            return TRUE;
+        } else if(result<0) {
+            limit=i;
+        } else /* result>0 */ {
+            start=i+1;
+        }
+    }
+
+    return FALSE; /* string not found */
+}
+
+U_NAMESPACE_BEGIN
+
+FullCaseFoldingIterator::FullCaseFoldingIterator()
+        : unfold(reinterpret_cast<const UChar *>(ucase_props_singleton.unfold)),
+          unfoldRows(unfold[UCASE_UNFOLD_ROWS]),
+          unfoldRowWidth(unfold[UCASE_UNFOLD_ROW_WIDTH]),
+          unfoldStringWidth(unfold[UCASE_UNFOLD_STRING_WIDTH]),
+          currentRow(0),
+          rowCpIndex(unfoldStringWidth) {
+    unfold+=unfoldRowWidth;
+}
+
+UChar32
+FullCaseFoldingIterator::next(UnicodeString &full) {
+    // Advance past the last-delivered code point.
+    const UChar *p=unfold+(currentRow*unfoldRowWidth);
+    if(rowCpIndex>=unfoldRowWidth || p[rowCpIndex]==0) {
+        ++currentRow;
+        p+=unfoldRowWidth;
+        rowCpIndex=unfoldStringWidth;
+    }
+    if(currentRow>=unfoldRows) { return U_SENTINEL; }
+    // Set "full" to the NUL-terminated string in the first unfold column.
+    int32_t length=unfoldStringWidth;
+    while(length>0 && p[length-1]==0) { --length; }
+    full.setTo(FALSE, p, length);
+    // Return the code point.
+    UChar32 c;
+    U16_NEXT_UNSAFE(p, rowCpIndex, c);
+    return c;
+}
+
+namespace LatinCase {
+
+const int8_t TO_LOWER_NORMAL[LIMIT] = {
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+    32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, EXC, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32,
+    32, 32, 32, 32, 32, 32, 32, 0, 32, 32, 32, 32, 32, 32, 32, EXC,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    EXC, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1,
+
+    0, 1, 0, 1, 0, 1, 0, 1, 0, EXC, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, -121, 1, 0, 1, 0, 1, 0, EXC
+};
+
+const int8_t TO_LOWER_TR_LT[LIMIT] = {
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 32, 32, 32, 32, 32, 32, 32, 32, EXC, EXC, 32, 32, 32, 32, 32,
+    32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, EXC, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, EXC, EXC, 32, 32,
+    32, 32, 32, 32, 32, 32, 32, 0, 32, 32, 32, 32, 32, 32, 32, EXC,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, EXC, 0, 1, 0, 1, 0, EXC, 0,
+    EXC, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1,
+
+    0, 1, 0, 1, 0, 1, 0, 1, 0, EXC, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0,
+    1, 0, 1, 0, 1, 0, 1, 0, -121, 1, 0, 1, 0, 1, 0, EXC
+};
+
+const int8_t TO_UPPER_NORMAL[LIMIT] = {
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32,
+    -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, EXC, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, EXC,
+    -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32,
+    -32, -32, -32, -32, -32, -32, -32, 0, -32, -32, -32, -32, -32, -32, -32, 121,
+
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, EXC, 0, -1, 0, -1, 0, -1, 0, 0, -1, 0, -1, 0, -1, 0,
+
+    -1, 0, -1, 0, -1, 0, -1, 0, -1, EXC, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, 0, -1, 0, -1, 0, -1, EXC
+};
+
+const int8_t TO_UPPER_TR[LIMIT] = {
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, -32, -32, -32, -32, -32, -32, -32, -32, EXC, -32, -32, -32, -32, -32, -32,
+    -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, EXC, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, EXC,
+    -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32, -32,
+    -32, -32, -32, -32, -32, -32, -32, 0, -32, -32, -32, -32, -32, -32, -32, 121,
+
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, EXC, 0, -1, 0, -1, 0, -1, 0, 0, -1, 0, -1, 0, -1, 0,
+
+    -1, 0, -1, 0, -1, 0, -1, 0, -1, EXC, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1, 0, -1,
+    0, -1, 0, -1, 0, -1, 0, -1, 0, 0, -1, 0, -1, 0, -1, EXC
+};
+
+}  // namespace LatinCase
+
+U_NAMESPACE_END
+
+/** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
+U_CAPI int32_t U_EXPORT2
+ucase_getType(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    return UCASE_GET_TYPE(props);
+}
+
+/** @return same as ucase_getType() and set bit 2 if c is case-ignorable */
+U_CAPI int32_t U_EXPORT2
+ucase_getTypeOrIgnorable(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    return UCASE_GET_TYPE_AND_IGNORABLE(props);
+}
+
+/** @return UCASE_NO_DOT, UCASE_SOFT_DOTTED, UCASE_ABOVE, UCASE_OTHER_ACCENT */
+static inline int32_t
+getDotType(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        return props&UCASE_DOT_MASK;
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        return (*pe>>UCASE_EXC_DOT_SHIFT)&UCASE_DOT_MASK;
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+ucase_isSoftDotted(UChar32 c) {
+    return (UBool)(getDotType(c)==UCASE_SOFT_DOTTED);
+}
+
+U_CAPI UBool U_EXPORT2
+ucase_isCaseSensitive(UChar32 c) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        return (UBool)((props&UCASE_SENSITIVE)!=0);
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        return (UBool)((*pe&UCASE_EXC_SENSITIVE)!=0);
+    }
+}
+
+/* string casing ------------------------------------------------------------ */
+
+/*
+ * These internal functions form the core of string case mappings.
+ * They map single code points to result code points or strings and take
+ * all necessary conditions (context, locale ID, options) into account.
+ *
+ * They do not iterate over the source or write to the destination
+ * so that the same functions are useful for non-standard string storage,
+ * such as in a Replaceable (for Transliterator) or UTF-8/32 strings etc.
+ * For the same reason, the "surrounding text" context is passed in as a
+ * UCaseContextIterator which does not make any assumptions about
+ * the underlying storage.
+ *
+ * This section contains helper functions that check for conditions
+ * in the input text surrounding the current code point
+ * according to SpecialCasing.txt.
+ *
+ * Each helper function gets the index
+ * - after the current code point if it looks at following text
+ * - before the current code point if it looks at preceding text
+ *
+ * Unicode 3.2 UAX 21 "Case Mappings" defines the conditions as follows:
+ *
+ * Final_Sigma
+ *   C is preceded by a sequence consisting of
+ *     a cased letter and a case-ignorable sequence,
+ *   and C is not followed by a sequence consisting of
+ *     an ignorable sequence and then a cased letter.
+ *
+ * More_Above
+ *   C is followed by one or more characters of combining class 230 (ABOVE)
+ *   in the combining character sequence.
+ *
+ * After_Soft_Dotted
+ *   The last preceding character with combining class of zero before C
+ *   was Soft_Dotted,
+ *   and there is no intervening combining character class 230 (ABOVE).
+ *
+ * Before_Dot
+ *   C is followed by combining dot above (U+0307).
+ *   Any sequence of characters with a combining class that is neither 0 nor 230
+ *   may intervene between the current character and the combining dot above.
+ *
+ * The erratum from 2002-10-31 adds the condition
+ *
+ * After_I
+ *   The last preceding base character was an uppercase I, and there is no
+ *   intervening combining character class 230 (ABOVE).
+ *
+ *   (See Jitterbug 2344 and the comments on After_I below.)
+ *
+ * Helper definitions in Unicode 3.2 UAX 21:
+ *
+ * D1. A character C is defined to be cased
+ *     if it meets any of the following criteria:
+ *
+ *   - The general category of C is Titlecase Letter (Lt)
+ *   - In [CoreProps], C has one of the properties Uppercase, or Lowercase
+ *   - Given D = NFD(C), then it is not the case that:
+ *     D = UCD_lower(D) = UCD_upper(D) = UCD_title(D)
+ *     (This third criterium does not add any characters to the list
+ *      for Unicode 3.2. Ignored.)
+ *
+ * D2. A character C is defined to be case-ignorable
+ *     if it meets either of the following criteria:
+ *
+ *   - The general category of C is
+ *     Nonspacing Mark (Mn), or Enclosing Mark (Me), or Format Control (Cf), or
+ *     Letter Modifier (Lm), or Symbol Modifier (Sk)
+ *   - C is one of the following characters 
+ *     U+0027 APOSTROPHE
+ *     U+00AD SOFT HYPHEN (SHY)
+ *     U+2019 RIGHT SINGLE QUOTATION MARK
+ *            (the preferred character for apostrophe)
+ *
+ * D3. A case-ignorable sequence is a sequence of
+ *     zero or more case-ignorable characters.
+ */
+
+#define is_d(c) ((c)=='d' || (c)=='D')
+#define is_e(c) ((c)=='e' || (c)=='E')
+#define is_i(c) ((c)=='i' || (c)=='I')
+#define is_l(c) ((c)=='l' || (c)=='L')
+#define is_r(c) ((c)=='r' || (c)=='R')
+#define is_t(c) ((c)=='t' || (c)=='T')
+#define is_u(c) ((c)=='u' || (c)=='U')
+#define is_y(c) ((c)=='y' || (c)=='Y')
+#define is_z(c) ((c)=='z' || (c)=='Z')
+
+/* separator? */
+#define is_sep(c) ((c)=='_' || (c)=='-' || (c)==0)
+
+/**
+ * Requires non-NULL locale ID but otherwise does the equivalent of
+ * checking for language codes as if uloc_getLanguage() were called:
+ * Accepts both 2- and 3-letter codes and accepts case variants.
+ */
+U_CFUNC int32_t
+ucase_getCaseLocale(const char *locale) {
+    /*
+     * This function used to use uloc_getLanguage(), but the current code
+     * removes the dependency of this low-level code on uloc implementation code
+     * and is faster because not the whole locale ID has to be
+     * examined and copied/transformed.
+     *
+     * Because this code does not want to depend on uloc, the caller must
+     * pass in a non-NULL locale, i.e., may need to call uloc_getDefault().
+     */
+    char c=*locale++;
+    // Fastpath for English "en" which is often used for default (=root locale) case mappings,
+    // and for Chinese "zh": Very common but no special case mapping behavior.
+    // Then check lowercase vs. uppercase to reduce the number of comparisons
+    // for other locales without special behavior.
+    if(c=='e') {
+        /* el or ell? */
+        c=*locale++;
+        if(is_l(c)) {
+            c=*locale++;
+            if(is_l(c)) {
+                c=*locale;
+            }
+            if(is_sep(c)) {
+                return UCASE_LOC_GREEK;
+            }
+        }
+        // en, es, ... -> root
+    } else if(c=='z') {
+        return UCASE_LOC_ROOT;
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+    } else if(c>='a') {  // ASCII a-z = 0x61..0x7a, after A-Z
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    } else if(c<='z') {  // EBCDIC a-z = 0x81..0xa9 with two gaps, before A-Z
+#else
+#   error Unknown charset family!
+#endif
+        // lowercase c
+        if(c=='t') {
+            /* tr or tur? */
+            c=*locale++;
+            if(is_u(c)) {
+                c=*locale++;
+            }
+            if(is_r(c)) {
+                c=*locale;
+                if(is_sep(c)) {
+                    return UCASE_LOC_TURKISH;
+                }
+            }
+        } else if(c=='a') {
+            /* az or aze? */
+            c=*locale++;
+            if(is_z(c)) {
+                c=*locale++;
+                if(is_e(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_TURKISH;
+                }
+            }
+        } else if(c=='l') {
+            /* lt or lit? */
+            c=*locale++;
+            if(is_i(c)) {
+                c=*locale++;
+            }
+            if(is_t(c)) {
+                c=*locale;
+                if(is_sep(c)) {
+                    return UCASE_LOC_LITHUANIAN;
+                }
+            }
+        } else if(c=='n') {
+            /* nl or nld? */
+            c=*locale++;
+            if(is_l(c)) {
+                c=*locale++;
+                if(is_d(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_DUTCH;
+                }
+            }
+        } else if(c=='h') {
+            /* hy or hye? *not* hyw */
+            c=*locale++;
+            if(is_y(c)) {
+                c=*locale++;
+                if(is_e(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_ARMENIAN;
+                }
+            }
+        }
+    } else {
+        // uppercase c
+        // Same code as for lowercase c but also check for 'E'.
+        if(c=='T') {
+            /* tr or tur? */
+            c=*locale++;
+            if(is_u(c)) {
+                c=*locale++;
+            }
+            if(is_r(c)) {
+                c=*locale;
+                if(is_sep(c)) {
+                    return UCASE_LOC_TURKISH;
+                }
+            }
+        } else if(c=='A') {
+            /* az or aze? */
+            c=*locale++;
+            if(is_z(c)) {
+                c=*locale++;
+                if(is_e(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_TURKISH;
+                }
+            }
+        } else if(c=='L') {
+            /* lt or lit? */
+            c=*locale++;
+            if(is_i(c)) {
+                c=*locale++;
+            }
+            if(is_t(c)) {
+                c=*locale;
+                if(is_sep(c)) {
+                    return UCASE_LOC_LITHUANIAN;
+                }
+            }
+        } else if(c=='E') {
+            /* el or ell? */
+            c=*locale++;
+            if(is_l(c)) {
+                c=*locale++;
+                if(is_l(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_GREEK;
+                }
+            }
+        } else if(c=='N') {
+            /* nl or nld? */
+            c=*locale++;
+            if(is_l(c)) {
+                c=*locale++;
+                if(is_d(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_DUTCH;
+                }
+            }
+        } else if(c=='H') {
+            /* hy or hye? *not* hyw */
+            c=*locale++;
+            if(is_y(c)) {
+                c=*locale++;
+                if(is_e(c)) {
+                    c=*locale;
+                }
+                if(is_sep(c)) {
+                    return UCASE_LOC_ARMENIAN;
+                }
+            }
+        }
+    }
+    return UCASE_LOC_ROOT;
+}
+
+/*
+ * Is followed by
+ *   {case-ignorable}* cased
+ * ?
+ * (dir determines looking forward/backward)
+ * If a character is case-ignorable, it is skipped regardless of whether
+ * it is also cased or not.
+ */
+static UBool
+isFollowedByCasedLetter(UCaseContextIterator *iter, void *context, int8_t dir) {
+    UChar32 c;
+
+    if(iter==NULL) {
+        return FALSE;
+    }
+
+    for(/* dir!=0 sets direction */; (c=iter(context, dir))>=0; dir=0) {
+        int32_t type=ucase_getTypeOrIgnorable(c);
+        if(type&4) {
+            /* case-ignorable, continue with the loop */
+        } else if(type!=UCASE_NONE) {
+            return TRUE; /* followed by cased letter */
+        } else {
+            return FALSE; /* uncased and not case-ignorable */
+        }
+    }
+
+    return FALSE; /* not followed by cased letter */
+}
+
+/* Is preceded by Soft_Dotted character with no intervening cc=230 ? */
+static UBool
+isPrecededBySoftDotted(UCaseContextIterator *iter, void *context) {
+    UChar32 c;
+    int32_t dotType;
+    int8_t dir;
+
+    if(iter==NULL) {
+        return FALSE;
+    }
+
+    for(dir=-1; (c=iter(context, dir))>=0; dir=0) {
+        dotType=getDotType(c);
+        if(dotType==UCASE_SOFT_DOTTED) {
+            return TRUE; /* preceded by TYPE_i */
+        } else if(dotType!=UCASE_OTHER_ACCENT) {
+            return FALSE; /* preceded by different base character (not TYPE_i), or intervening cc==230 */
+        }
+    }
+
+    return FALSE; /* not preceded by TYPE_i */
+}
+
+/*
+ * See Jitterbug 2344:
+ * The condition After_I for Turkic-lowercasing of U+0307 combining dot above
+ * is checked in ICU 2.0, 2.1, 2.6 but was not in 2.2 & 2.4 because
+ * we made those releases compatible with Unicode 3.2 which had not fixed
+ * a related bug in SpecialCasing.txt.
+ *
+ * From the Jitterbug 2344 text:
+ * ... this bug is listed as a Unicode erratum
+ * from 2002-10-31 at http://www.unicode.org/uni2errata/UnicodeErrata.html
+ * <quote>
+ * There are two errors in SpecialCasing.txt.
+ * 1. Missing semicolons on two lines. ... [irrelevant for ICU]
+ * 2. An incorrect context definition. Correct as follows:
+ * < 0307; ; 0307; 0307; tr After_Soft_Dotted; # COMBINING DOT ABOVE
+ * < 0307; ; 0307; 0307; az After_Soft_Dotted; # COMBINING DOT ABOVE
+ * ---
+ * > 0307; ; 0307; 0307; tr After_I; # COMBINING DOT ABOVE
+ * > 0307; ; 0307; 0307; az After_I; # COMBINING DOT ABOVE
+ * where the context After_I is defined as:
+ * The last preceding base character was an uppercase I, and there is no
+ * intervening combining character class 230 (ABOVE).
+ * </quote>
+ *
+ * Note that SpecialCasing.txt even in Unicode 3.2 described the condition as:
+ *
+ * # When lowercasing, remove dot_above in the sequence I + dot_above, which will turn into i.
+ * # This matches the behavior of the canonically equivalent I-dot_above
+ *
+ * See also the description in this place in older versions of uchar.c (revision 1.100).
+ *
+ * Markus W. Scherer 2003-feb-15
+ */
+
+/* Is preceded by base character 'I' with no intervening cc=230 ? */
+static UBool
+isPrecededBy_I(UCaseContextIterator *iter, void *context) {
+    UChar32 c;
+    int32_t dotType;
+    int8_t dir;
+
+    if(iter==NULL) {
+        return FALSE;
+    }
+
+    for(dir=-1; (c=iter(context, dir))>=0; dir=0) {
+        if(c==0x49) {
+            return TRUE; /* preceded by I */
+        }
+        dotType=getDotType(c);
+        if(dotType!=UCASE_OTHER_ACCENT) {
+            return FALSE; /* preceded by different base character (not I), or intervening cc==230 */
+        }
+    }
+
+    return FALSE; /* not preceded by I */
+}
+
+/* Is followed by one or more cc==230 ? */
+static UBool
+isFollowedByMoreAbove(UCaseContextIterator *iter, void *context) {
+    UChar32 c;
+    int32_t dotType;
+    int8_t dir;
+
+    if(iter==NULL) {
+        return FALSE;
+    }
+
+    for(dir=1; (c=iter(context, dir))>=0; dir=0) {
+        dotType=getDotType(c);
+        if(dotType==UCASE_ABOVE) {
+            return TRUE; /* at least one cc==230 following */
+        } else if(dotType!=UCASE_OTHER_ACCENT) {
+            return FALSE; /* next base character, no more cc==230 following */
+        }
+    }
+
+    return FALSE; /* no more cc==230 following */
+}
+
+/* Is followed by a dot above (without cc==230 in between) ? */
+static UBool
+isFollowedByDotAbove(UCaseContextIterator *iter, void *context) {
+    UChar32 c;
+    int32_t dotType;
+    int8_t dir;
+
+    if(iter==NULL) {
+        return FALSE;
+    }
+
+    for(dir=1; (c=iter(context, dir))>=0; dir=0) {
+        if(c==0x307) {
+            return TRUE;
+        }
+        dotType=getDotType(c);
+        if(dotType!=UCASE_OTHER_ACCENT) {
+            return FALSE; /* next base character or cc==230 in between */
+        }
+    }
+
+    return FALSE; /* no dot above following */
+}
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullLower(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t loc) {
+    // The sign of the result has meaning, input must be non-negative so that it can be returned as is.
+    U_ASSERT(c >= 0);
+    UChar32 result=c;
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_IS_UPPER_OR_TITLE(props)) {
+            result=c+UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props), *pe2;
+        uint16_t excWord=*pe++;
+        int32_t full;
+
+        pe2=pe;
+
+        if(excWord&UCASE_EXC_CONDITIONAL_SPECIAL) {
+            /* use hardcoded conditions and mappings */
+
+            /*
+             * Test for conditional mappings first
+             *   (otherwise the unconditional default mappings are always taken),
+             * then test for characters that have unconditional mappings in SpecialCasing.txt,
+             * then get the UnicodeData.txt mappings.
+             */
+            if( loc==UCASE_LOC_LITHUANIAN &&
+                    /* base characters, find accents above */
+                    (((c==0x49 || c==0x4a || c==0x12e) &&
+                        isFollowedByMoreAbove(iter, context)) ||
+                    /* precomposed with accent above, no need to find one */
+                    (c==0xcc || c==0xcd || c==0x128))
+            ) {
+                /*
+                    # Lithuanian
+
+                    # Lithuanian retains the dot in a lowercase i when followed by accents.
+
+                    # Introduce an explicit dot above when lowercasing capital I's and J's
+                    # whenever there are more accents above.
+                    # (of the accents used in Lithuanian: grave, acute, tilde above, and ogonek)
+
+                    0049; 0069 0307; 0049; 0049; lt More_Above; # LATIN CAPITAL LETTER I
+                    004A; 006A 0307; 004A; 004A; lt More_Above; # LATIN CAPITAL LETTER J
+                    012E; 012F 0307; 012E; 012E; lt More_Above; # LATIN CAPITAL LETTER I WITH OGONEK
+                    00CC; 0069 0307 0300; 00CC; 00CC; lt; # LATIN CAPITAL LETTER I WITH GRAVE
+                    00CD; 0069 0307 0301; 00CD; 00CD; lt; # LATIN CAPITAL LETTER I WITH ACUTE
+                    0128; 0069 0307 0303; 0128; 0128; lt; # LATIN CAPITAL LETTER I WITH TILDE
+                 */
+                switch(c) {
+                case 0x49:  /* LATIN CAPITAL LETTER I */
+                    *pString=iDot;
+                    return 2;
+                case 0x4a:  /* LATIN CAPITAL LETTER J */
+                    *pString=jDot;
+                    return 2;
+                case 0x12e: /* LATIN CAPITAL LETTER I WITH OGONEK */
+                    *pString=iOgonekDot;
+                    return 2;
+                case 0xcc:  /* LATIN CAPITAL LETTER I WITH GRAVE */
+                    *pString=iDotGrave;
+                    return 3;
+                case 0xcd:  /* LATIN CAPITAL LETTER I WITH ACUTE */
+                    *pString=iDotAcute;
+                    return 3;
+                case 0x128: /* LATIN CAPITAL LETTER I WITH TILDE */
+                    *pString=iDotTilde;
+                    return 3;
+                default:
+                    return 0; /* will not occur */
+                }
+            /* # Turkish and Azeri */
+            } else if(loc==UCASE_LOC_TURKISH && c==0x130) {
+                /*
+                    # I and i-dotless; I-dot and i are case pairs in Turkish and Azeri
+                    # The following rules handle those cases.
+
+                    0130; 0069; 0130; 0130; tr # LATIN CAPITAL LETTER I WITH DOT ABOVE
+                    0130; 0069; 0130; 0130; az # LATIN CAPITAL LETTER I WITH DOT ABOVE
+                 */
+                return 0x69;
+            } else if(loc==UCASE_LOC_TURKISH && c==0x307 && isPrecededBy_I(iter, context)) {
+                /*
+                    # When lowercasing, remove dot_above in the sequence I + dot_above, which will turn into i.
+                    # This matches the behavior of the canonically equivalent I-dot_above
+
+                    0307; ; 0307; 0307; tr After_I; # COMBINING DOT ABOVE
+                    0307; ; 0307; 0307; az After_I; # COMBINING DOT ABOVE
+                 */
+                *pString=nullptr;
+                return 0; /* remove the dot (continue without output) */
+            } else if(loc==UCASE_LOC_TURKISH && c==0x49 && !isFollowedByDotAbove(iter, context)) {
+                /*
+                    # When lowercasing, unless an I is before a dot_above, it turns into a dotless i.
+
+                    0049; 0131; 0049; 0049; tr Not_Before_Dot; # LATIN CAPITAL LETTER I
+                    0049; 0131; 0049; 0049; az Not_Before_Dot; # LATIN CAPITAL LETTER I
+                 */
+                return 0x131;
+            } else if(c==0x130) {
+                /*
+                    # Preserve canonical equivalence for I with dot. Turkic is handled below.
+
+                    0130; 0069 0307; 0130; 0130; # LATIN CAPITAL LETTER I WITH DOT ABOVE
+                 */
+                *pString=iDot;
+                return 2;
+            } else if(  c==0x3a3 &&
+                        !isFollowedByCasedLetter(iter, context, 1) &&
+                        isFollowedByCasedLetter(iter, context, -1) /* -1=preceded */
+            ) {
+                /* greek capital sigma maps depending on surrounding cased letters (see SpecialCasing.txt) */
+                /*
+                    # Special case for final form of sigma
+
+                    03A3; 03C2; 03A3; 03A3; Final_Sigma; # GREEK CAPITAL LETTER SIGMA
+                 */
+                return 0x3c2; /* greek small final sigma */
+            } else {
+                /* no known conditional special case mapping, use a normal mapping */
+            }
+        } else if(HAS_SLOT(excWord, UCASE_EXC_FULL_MAPPINGS)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_FULL_MAPPINGS, pe, full);
+            full&=UCASE_FULL_LOWER;
+            if(full!=0) {
+                /* set the output pointer to the lowercase mapping */
+                *pString=reinterpret_cast<const UChar *>(pe+1);
+
+                /* return the string length */
+                return full;
+            }
+        }
+
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_IS_UPPER_OR_TITLE(props)) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe2, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_LOWER)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_LOWER, pe2, result);
+        }
+    }
+
+    return (result==c) ? ~result : result;
+}
+
+/* internal */
+static int32_t
+toUpperOrTitle(UChar32 c,
+               UCaseContextIterator *iter, void *context,
+               const UChar **pString,
+               int32_t loc,
+               UBool upperNotTitle) {
+    // The sign of the result has meaning, input must be non-negative so that it can be returned as is.
+    U_ASSERT(c >= 0);
+    UChar32 result=c;
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            result=c+UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props), *pe2;
+        uint16_t excWord=*pe++;
+        int32_t full, idx;
+
+        pe2=pe;
+
+        if(excWord&UCASE_EXC_CONDITIONAL_SPECIAL) {
+            /* use hardcoded conditions and mappings */
+            if(loc==UCASE_LOC_TURKISH && c==0x69) {
+                /*
+                    # Turkish and Azeri
+
+                    # I and i-dotless; I-dot and i are case pairs in Turkish and Azeri
+                    # The following rules handle those cases.
+
+                    # When uppercasing, i turns into a dotted capital I
+
+                    0069; 0069; 0130; 0130; tr; # LATIN SMALL LETTER I
+                    0069; 0069; 0130; 0130; az; # LATIN SMALL LETTER I
+                */
+                return 0x130;
+            } else if(loc==UCASE_LOC_LITHUANIAN && c==0x307 && isPrecededBySoftDotted(iter, context)) {
+                /*
+                    # Lithuanian
+
+                    # Lithuanian retains the dot in a lowercase i when followed by accents.
+
+                    # Remove DOT ABOVE after "i" with upper or titlecase
+
+                    0307; 0307; ; ; lt After_Soft_Dotted; # COMBINING DOT ABOVE
+                 */
+                *pString=nullptr;
+                return 0; /* remove the dot (continue without output) */
+            } else if(c==0x0587) {
+                // See ICU-13416:
+                // և ligature ech-yiwn
+                // uppercases to ԵՒ=ech+yiwn by default and in Western Armenian,
+                // but to ԵՎ=ech+vew in Eastern Armenian.
+                if(loc==UCASE_LOC_ARMENIAN) {
+                    *pString=upperNotTitle ? u"ԵՎ" : u"Եվ";
+                } else {
+                    *pString=upperNotTitle ? u"ԵՒ" : u"Եւ";
+                }
+                return 2;
+            } else {
+                /* no known conditional special case mapping, use a normal mapping */
+            }
+        } else if(HAS_SLOT(excWord, UCASE_EXC_FULL_MAPPINGS)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_FULL_MAPPINGS, pe, full);
+
+            /* start of full case mapping strings */
+            ++pe;
+
+            /* skip the lowercase and case-folding result strings */
+            pe+=full&UCASE_FULL_LOWER;
+            full>>=4;
+            pe+=full&0xf;
+            full>>=4;
+
+            if(upperNotTitle) {
+                full&=0xf;
+            } else {
+                /* skip the uppercase result string */
+                pe+=full&0xf;
+                full=(full>>4)&0xf;
+            }
+
+            if(full!=0) {
+                /* set the output pointer to the result string */
+                *pString=reinterpret_cast<const UChar *>(pe);
+
+                /* return the string length */
+                return full;
+            }
+        }
+
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_GET_TYPE(props)==UCASE_LOWER) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe2, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(!upperNotTitle && HAS_SLOT(excWord, UCASE_EXC_TITLE)) {
+            idx=UCASE_EXC_TITLE;
+        } else if(HAS_SLOT(excWord, UCASE_EXC_UPPER)) {
+            /* here, titlecase is same as uppercase */
+            idx=UCASE_EXC_UPPER;
+        } else {
+            return ~c;
+        }
+        GET_SLOT_VALUE(excWord, idx, pe2, result);
+    }
+
+    return (result==c) ? ~result : result;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullUpper(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t caseLocale) {
+    return toUpperOrTitle(c, iter, context, pString, caseLocale, TRUE);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullTitle(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t caseLocale) {
+    return toUpperOrTitle(c, iter, context, pString, caseLocale, FALSE);
+}
+
+/* case folding ------------------------------------------------------------- */
+
+/*
+ * Case folding is similar to lowercasing.
+ * The result may be a simple mapping, i.e., a single code point, or
+ * a full mapping, i.e., a string.
+ * If the case folding for a code point is the same as its simple (1:1) lowercase mapping,
+ * then only the lowercase mapping is stored.
+ *
+ * Some special cases are hardcoded because their conditions cannot be
+ * parsed and processed from CaseFolding.txt.
+ *
+ * Unicode 3.2 CaseFolding.txt specifies for its status field:
+
+# C: common case folding, common mappings shared by both simple and full mappings.
+# F: full case folding, mappings that cause strings to grow in length. Multiple characters are separated by spaces.
+# S: simple case folding, mappings to single characters where different from F.
+# T: special case for uppercase I and dotted uppercase I
+#    - For non-Turkic languages, this mapping is normally not used.
+#    - For Turkic languages (tr, az), this mapping can be used instead of the normal mapping for these characters.
+#
+# Usage:
+#  A. To do a simple case folding, use the mappings with status C + S.
+#  B. To do a full case folding, use the mappings with status C + F.
+#
+#    The mappings with status T can be used or omitted depending on the desired case-folding
+#    behavior. (The default option is to exclude them.)
+
+ * Unicode 3.2 has 'T' mappings as follows:
+
+0049; T; 0131; # LATIN CAPITAL LETTER I
+0130; T; 0069; # LATIN CAPITAL LETTER I WITH DOT ABOVE
+
+ * while the default mappings for these code points are:
+
+0049; C; 0069; # LATIN CAPITAL LETTER I
+0130; F; 0069 0307; # LATIN CAPITAL LETTER I WITH DOT ABOVE
+
+ * U+0130 has no simple case folding (simple-case-folds to itself).
+ */
+
+/* return the simple case folding mapping for c */
+U_CAPI UChar32 U_EXPORT2
+ucase_fold(UChar32 c, uint32_t options) {
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_IS_UPPER_OR_TITLE(props)) {
+            c+=UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props);
+        uint16_t excWord=*pe++;
+        int32_t idx;
+        if(excWord&UCASE_EXC_CONDITIONAL_FOLD) {
+            /* special case folding mappings, hardcoded */
+            if((options&_FOLD_CASE_OPTIONS_MASK)==U_FOLD_CASE_DEFAULT) {
+                /* default mappings */
+                if(c==0x49) {
+                    /* 0049; C; 0069; # LATIN CAPITAL LETTER I */
+                    return 0x69;
+                } else if(c==0x130) {
+                    /* no simple case folding for U+0130 */
+                    return c;
+                }
+            } else {
+                /* Turkic mappings */
+                if(c==0x49) {
+                    /* 0049; T; 0131; # LATIN CAPITAL LETTER I */
+                    return 0x131;
+                } else if(c==0x130) {
+                    /* 0130; T; 0069; # LATIN CAPITAL LETTER I WITH DOT ABOVE */
+                    return 0x69;
+                }
+            }
+        }
+        if((excWord&UCASE_EXC_NO_SIMPLE_CASE_FOLDING)!=0) {
+            return c;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_IS_UPPER_OR_TITLE(props)) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_FOLD)) {
+            idx=UCASE_EXC_FOLD;
+        } else if(HAS_SLOT(excWord, UCASE_EXC_LOWER)) {
+            idx=UCASE_EXC_LOWER;
+        } else {
+            return c;
+        }
+        GET_SLOT_VALUE(excWord, idx, pe, c);
+    }
+    return c;
+}
+
+/*
+ * Issue for canonical caseless match (UAX #21):
+ * Turkic casefolding (using "T" mappings in CaseFolding.txt) does not preserve
+ * canonical equivalence, unlike default-option casefolding.
+ * For example, I-grave and I + grave fold to strings that are not canonically
+ * equivalent.
+ * For more details, see the comment in unorm_compare() in unorm.cpp
+ * and the intermediate prototype changes for Jitterbug 2021.
+ * (For example, revision 1.104 of uchar.c and 1.4 of CaseFolding.txt.)
+ *
+ * This did not get fixed because it appears that it is not possible to fix
+ * it for uppercase and lowercase characters (I-grave vs. i-grave)
+ * together in a way that they still fold to common result strings.
+ */
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullFolding(UChar32 c,
+                    const UChar **pString,
+                    uint32_t options) {
+    // The sign of the result has meaning, input must be non-negative so that it can be returned as is.
+    U_ASSERT(c >= 0);
+    UChar32 result=c;
+    uint16_t props=UTRIE2_GET16(&ucase_props_singleton.trie, c);
+    if(!UCASE_HAS_EXCEPTION(props)) {
+        if(UCASE_IS_UPPER_OR_TITLE(props)) {
+            result=c+UCASE_GET_DELTA(props);
+        }
+    } else {
+        const uint16_t *pe=GET_EXCEPTIONS(&ucase_props_singleton, props), *pe2;
+        uint16_t excWord=*pe++;
+        int32_t full, idx;
+
+        pe2=pe;
+
+        if(excWord&UCASE_EXC_CONDITIONAL_FOLD) {
+            /* use hardcoded conditions and mappings */
+            if((options&_FOLD_CASE_OPTIONS_MASK)==U_FOLD_CASE_DEFAULT) {
+                /* default mappings */
+                if(c==0x49) {
+                    /* 0049; C; 0069; # LATIN CAPITAL LETTER I */
+                    return 0x69;
+                } else if(c==0x130) {
+                    /* 0130; F; 0069 0307; # LATIN CAPITAL LETTER I WITH DOT ABOVE */
+                    *pString=iDot;
+                    return 2;
+                }
+            } else {
+                /* Turkic mappings */
+                if(c==0x49) {
+                    /* 0049; T; 0131; # LATIN CAPITAL LETTER I */
+                    return 0x131;
+                } else if(c==0x130) {
+                    /* 0130; T; 0069; # LATIN CAPITAL LETTER I WITH DOT ABOVE */
+                    return 0x69;
+                }
+            }
+        } else if(HAS_SLOT(excWord, UCASE_EXC_FULL_MAPPINGS)) {
+            GET_SLOT_VALUE(excWord, UCASE_EXC_FULL_MAPPINGS, pe, full);
+
+            /* start of full case mapping strings */
+            ++pe;
+
+            /* skip the lowercase result string */
+            pe+=full&UCASE_FULL_LOWER;
+            full=(full>>4)&0xf;
+
+            if(full!=0) {
+                /* set the output pointer to the result string */
+                *pString=reinterpret_cast<const UChar *>(pe);
+
+                /* return the string length */
+                return full;
+            }
+        }
+
+        if((excWord&UCASE_EXC_NO_SIMPLE_CASE_FOLDING)!=0) {
+            return ~c;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_DELTA) && UCASE_IS_UPPER_OR_TITLE(props)) {
+            int32_t delta;
+            GET_SLOT_VALUE(excWord, UCASE_EXC_DELTA, pe2, delta);
+            return (excWord&UCASE_EXC_DELTA_IS_NEGATIVE)==0 ? c+delta : c-delta;
+        }
+        if(HAS_SLOT(excWord, UCASE_EXC_FOLD)) {
+            idx=UCASE_EXC_FOLD;
+        } else if(HAS_SLOT(excWord, UCASE_EXC_LOWER)) {
+            idx=UCASE_EXC_LOWER;
+        } else {
+            return ~c;
+        }
+        GET_SLOT_VALUE(excWord, idx, pe2, result);
+    }
+
+    return (result==c) ? ~result : result;
+}
+
+/* case mapping properties API ---------------------------------------------- */
+
+/* public API (see uchar.h) */
+
+U_CAPI UBool U_EXPORT2
+u_isULowercase(UChar32 c) {
+    return (UBool)(UCASE_LOWER==ucase_getType(c));
+}
+
+U_CAPI UBool U_EXPORT2
+u_isUUppercase(UChar32 c) {
+    return (UBool)(UCASE_UPPER==ucase_getType(c));
+}
+
+/* Transforms the Unicode character to its lower case equivalent.*/
+U_CAPI UChar32 U_EXPORT2
+u_tolower(UChar32 c) {
+    return ucase_tolower(c);
+}
+    
+/* Transforms the Unicode character to its upper case equivalent.*/
+U_CAPI UChar32 U_EXPORT2
+u_toupper(UChar32 c) {
+    return ucase_toupper(c);
+}
+
+/* Transforms the Unicode character to its title case equivalent.*/
+U_CAPI UChar32 U_EXPORT2
+u_totitle(UChar32 c) {
+    return ucase_totitle(c);
+}
+
+/* return the simple case folding mapping for c */
+U_CAPI UChar32 U_EXPORT2
+u_foldCase(UChar32 c, uint32_t options) {
+    return ucase_fold(c, options);
+}
+
+U_CFUNC int32_t U_EXPORT2
+ucase_hasBinaryProperty(UChar32 c, UProperty which) {
+    /* case mapping properties */
+    const UChar *resultString;
+    switch(which) {
+    case UCHAR_LOWERCASE:
+        return (UBool)(UCASE_LOWER==ucase_getType(c));
+    case UCHAR_UPPERCASE:
+        return (UBool)(UCASE_UPPER==ucase_getType(c));
+    case UCHAR_SOFT_DOTTED:
+        return ucase_isSoftDotted(c);
+    case UCHAR_CASE_SENSITIVE:
+        return ucase_isCaseSensitive(c);
+    case UCHAR_CASED:
+        return (UBool)(UCASE_NONE!=ucase_getType(c));
+    case UCHAR_CASE_IGNORABLE:
+        return (UBool)(ucase_getTypeOrIgnorable(c)>>2);
+    /*
+     * Note: The following Changes_When_Xyz are defined as testing whether
+     * the NFD form of the input changes when Xyz-case-mapped.
+     * However, this simpler implementation of these properties,
+     * ignoring NFD, passes the tests.
+     * The implementation needs to be changed if the tests start failing.
+     * When that happens, optimizations should be used to work with the
+     * per-single-code point ucase_toFullXyz() functions unless
+     * the NFD form has more than one code point,
+     * and the property starts set needs to be the union of the
+     * start sets for normalization and case mappings.
+     */
+    case UCHAR_CHANGES_WHEN_LOWERCASED:
+        return (UBool)(ucase_toFullLower(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0);
+    case UCHAR_CHANGES_WHEN_UPPERCASED:
+        return (UBool)(ucase_toFullUpper(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0);
+    case UCHAR_CHANGES_WHEN_TITLECASED:
+        return (UBool)(ucase_toFullTitle(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0);
+    /* case UCHAR_CHANGES_WHEN_CASEFOLDED: -- in uprops.c */
+    case UCHAR_CHANGES_WHEN_CASEMAPPED:
+        return (UBool)(
+            ucase_toFullLower(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0 ||
+            ucase_toFullUpper(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0 ||
+            ucase_toFullTitle(c, NULL, NULL, &resultString, UCASE_LOC_ROOT)>=0);
+    default:
+        return FALSE;
+    }
+}
diff --git a/thirdparty/icu4c/common/ucase.h b/thirdparty/icu4c/common/ucase.h
new file mode 100644
index 0000000000..a018f82b81
--- /dev/null
+++ b/thirdparty/icu4c/common/ucase.h
@@ -0,0 +1,445 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucase.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug30
+*   created by: Markus W. Scherer
+*
+*   Low-level Unicode character/string case mapping code.
+*/
+
+#ifndef __UCASE_H__
+#define __UCASE_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+#include "putilimp.h"
+#include "uset_imp.h"
+#include "udataswp.h"
+#include "utrie2.h"
+
+#ifdef __cplusplus
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+
+U_NAMESPACE_END
+#endif
+
+/* library API -------------------------------------------------------------- */
+
+U_CFUNC void U_EXPORT2
+ucase_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);
+
+/**
+ * Requires non-NULL locale ID but otherwise does the equivalent of
+ * checking for language codes as if uloc_getLanguage() were called:
+ * Accepts both 2- and 3-letter codes and accepts case variants.
+ */
+U_CFUNC int32_t
+ucase_getCaseLocale(const char *locale);
+
+/* Casing locale types for ucase_getCaseLocale */
+enum {
+    UCASE_LOC_UNKNOWN,
+    UCASE_LOC_ROOT,
+    UCASE_LOC_TURKISH,
+    UCASE_LOC_LITHUANIAN,
+    UCASE_LOC_GREEK,
+    UCASE_LOC_DUTCH,
+    UCASE_LOC_ARMENIAN
+};
+
+/**
+ * Bit mask for getting just the options from a string compare options word
+ * that are relevant for case-insensitive string comparison.
+ * See stringoptions.h. Also include _STRNCMP_STYLE and U_COMPARE_CODE_POINT_ORDER.
+ * @internal
+ */
+#define _STRCASECMP_OPTIONS_MASK 0xffff
+
+/**
+ * Bit mask for getting just the options from a string compare options word
+ * that are relevant for case folding (of a single string or code point).
+ *
+ * Currently only bit 0 for U_FOLD_CASE_EXCLUDE_SPECIAL_I.
+ * It is conceivable that at some point we might use one more bit for using uppercase sharp s.
+ * It is conceivable that at some point we might want the option to use only simple case foldings
+ * when operating on strings.
+ *
+ * See stringoptions.h.
+ * @internal
+ */
+#define _FOLD_CASE_OPTIONS_MASK 7
+
+/* single-code point functions */
+
+U_CAPI UChar32 U_EXPORT2
+ucase_tolower(UChar32 c);
+
+U_CAPI UChar32 U_EXPORT2
+ucase_toupper(UChar32 c);
+
+U_CAPI UChar32 U_EXPORT2
+ucase_totitle(UChar32 c);
+
+U_CAPI UChar32 U_EXPORT2
+ucase_fold(UChar32 c, uint32_t options);
+
+/**
+ * Adds all simple case mappings and the full case folding for c to sa,
+ * and also adds special case closure mappings.
+ * c itself is not added.
+ * For example, the mappings
+ * - for s include long s
+ * - for sharp s include ss
+ * - for k include the Kelvin sign
+ */
+U_CFUNC void U_EXPORT2
+ucase_addCaseClosure(UChar32 c, const USetAdder *sa);
+
+/**
+ * Maps the string to single code points and adds the associated case closure
+ * mappings.
+ * The string is mapped to code points if it is their full case folding string.
+ * In other words, this performs a reverse full case folding and then
+ * adds the case closure items of the resulting code points.
+ * If the string is found and its closure applied, then
+ * the string itself is added as well as part of its code points' closure.
+ * It must be length>=0.
+ *
+ * @return true if the string was found
+ */
+U_CFUNC UBool U_EXPORT2
+ucase_addStringCaseClosure(const UChar *s, int32_t length, const USetAdder *sa);
+
+#ifdef __cplusplus
+U_NAMESPACE_BEGIN
+
+/**
+ * Iterator over characters with more than one code point in the full default Case_Folding.
+ */
+class U_COMMON_API FullCaseFoldingIterator {
+public:
+    /** Constructor. */
+    FullCaseFoldingIterator();
+    /**
+     * Returns the next (cp, full) pair where "full" is cp's full default Case_Folding.
+     * Returns a negative cp value at the end of the iteration.
+     */
+    UChar32 next(UnicodeString &full);
+private:
+    FullCaseFoldingIterator(const FullCaseFoldingIterator &);  // no copy
+    FullCaseFoldingIterator &operator=(const FullCaseFoldingIterator &);  // no assignment
+
+    const UChar *unfold;
+    int32_t unfoldRows;
+    int32_t unfoldRowWidth;
+    int32_t unfoldStringWidth;
+    int32_t currentRow;
+    int32_t rowCpIndex;
+};
+
+/**
+ * Fast case mapping data for ASCII/Latin.
+ * Linear arrays of delta bytes: 0=no mapping; EXC=exception.
+ * Deltas must not cross the ASCII boundary, or else they cannot be easily used
+ * in simple UTF-8 code.
+ */
+namespace LatinCase {
+
+/** Case mapping/folding data for code points up to U+017F. */
+constexpr UChar LIMIT = 0x180;
+/** U+017F case-folds and uppercases crossing the ASCII boundary. */
+constexpr UChar LONG_S = 0x17f;
+/** Exception: Complex mapping, or too-large delta. */
+constexpr int8_t EXC = -0x80;
+
+/** Deltas for lowercasing for most locales, and default case folding. */
+extern const int8_t TO_LOWER_NORMAL[LIMIT];
+/** Deltas for lowercasing for tr/az/lt, and Turkic case folding. */
+extern const int8_t TO_LOWER_TR_LT[LIMIT];
+
+/** Deltas for uppercasing for most locales. */
+extern const int8_t TO_UPPER_NORMAL[LIMIT];
+/** Deltas for uppercasing for tr/az. */
+extern const int8_t TO_UPPER_TR[LIMIT];
+
+}  // namespace LatinCase
+
+U_NAMESPACE_END
+#endif
+
+/** @return UCASE_NONE, UCASE_LOWER, UCASE_UPPER, UCASE_TITLE */
+U_CAPI int32_t U_EXPORT2
+ucase_getType(UChar32 c);
+
+/** @return like ucase_getType() but also sets UCASE_IGNORABLE if c is case-ignorable */
+U_CAPI int32_t U_EXPORT2
+ucase_getTypeOrIgnorable(UChar32 c);
+
+U_CAPI UBool U_EXPORT2
+ucase_isSoftDotted(UChar32 c);
+
+U_CAPI UBool U_EXPORT2
+ucase_isCaseSensitive(UChar32 c);
+
+/* string case mapping functions */
+
+U_CDECL_BEGIN
+
+/**
+ * Iterator function for string case mappings, which need to look at the
+ * context (surrounding text) of a given character for conditional mappings.
+ *
+ * The iterator only needs to go backward or forward away from the
+ * character in question. It does not use any indexes on this interface.
+ * It does not support random access or an arbitrary change of
+ * iteration direction.
+ *
+ * The code point being case-mapped itself is never returned by
+ * this iterator.
+ *
+ * @param context A pointer to the iterator's working data.
+ * @param dir If <0 then start iterating backward from the character;
+ *            if >0 then start iterating forward from the character;
+ *            if 0 then continue iterating in the current direction.
+ * @return Next code point, or <0 when the iteration is done.
+ */
+typedef UChar32 U_CALLCONV
+UCaseContextIterator(void *context, int8_t dir);
+
+/**
+ * Sample struct which may be used by some implementations of
+ * UCaseContextIterator.
+ */
+struct UCaseContext {
+    void *p;
+    int32_t start, index, limit;
+    int32_t cpStart, cpLimit;
+    int8_t dir;
+    int8_t b1, b2, b3;
+};
+typedef struct UCaseContext UCaseContext;
+
+U_CDECL_END
+
+#define UCASECONTEXT_INITIALIZER { NULL,  0, 0, 0,  0, 0,  0,  0, 0, 0 }
+
+enum {
+    /**
+     * For string case mappings, a single character (a code point) is mapped
+     * either to itself (in which case in-place mapping functions do nothing),
+     * or to another single code point, or to a string.
+     * Aside from the string contents, these are indicated with a single int32_t
+     * value as follows:
+     *
+     * Mapping to self: Negative values (~self instead of -self to support U+0000)
+     *
+     * Mapping to another code point: Positive values >UCASE_MAX_STRING_LENGTH
+     *
+     * Mapping to a string: The string length (0..UCASE_MAX_STRING_LENGTH) is
+     * returned. Note that the string result may indeed have zero length.
+     */
+    UCASE_MAX_STRING_LENGTH=0x1f
+};
+
+/**
+ * Get the full lowercase mapping for c.
+ *
+ * @param csp Case mapping properties.
+ * @param c Character to be mapped.
+ * @param iter Character iterator, used for context-sensitive mappings.
+ *             See UCaseContextIterator for details.
+ *             If iter==NULL then a context-independent result is returned.
+ * @param context Pointer to be passed into iter.
+ * @param pString If the mapping result is a string, then the pointer is
+ *                written to *pString.
+ * @param caseLocale Case locale value from ucase_getCaseLocale().
+ * @return Output code point or string length, see UCASE_MAX_STRING_LENGTH.
+ *
+ * @see UCaseContextIterator
+ * @see UCASE_MAX_STRING_LENGTH
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucase_toFullLower(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t caseLocale);
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullUpper(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t caseLocale);
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullTitle(UChar32 c,
+                  UCaseContextIterator *iter, void *context,
+                  const UChar **pString,
+                  int32_t caseLocale);
+
+U_CAPI int32_t U_EXPORT2
+ucase_toFullFolding(UChar32 c,
+                    const UChar **pString,
+                    uint32_t options);
+
+U_CFUNC int32_t U_EXPORT2
+ucase_hasBinaryProperty(UChar32 c, UProperty which);
+
+
+U_CDECL_BEGIN
+
+/**
+ * @internal
+ */
+typedef int32_t U_CALLCONV
+UCaseMapFull(UChar32 c,
+             UCaseContextIterator *iter, void *context,
+             const UChar **pString,
+             int32_t caseLocale);
+
+U_CDECL_END
+
+/* file definitions --------------------------------------------------------- */
+
+#define UCASE_DATA_NAME "ucase"
+#define UCASE_DATA_TYPE "icu"
+
+/* format "cAsE" */
+#define UCASE_FMT_0 0x63
+#define UCASE_FMT_1 0x41
+#define UCASE_FMT_2 0x53
+#define UCASE_FMT_3 0x45
+
+/* indexes into indexes[] */
+enum {
+    UCASE_IX_INDEX_TOP,
+    UCASE_IX_LENGTH,
+    UCASE_IX_TRIE_SIZE,
+    UCASE_IX_EXC_LENGTH,
+    UCASE_IX_UNFOLD_LENGTH,
+
+    UCASE_IX_MAX_FULL_LENGTH=15,
+    UCASE_IX_TOP=16
+};
+
+/* definitions for 16-bit case properties word ------------------------------ */
+
+U_CFUNC const UTrie2 * U_EXPORT2
+ucase_getTrie();
+
+/* 2-bit constants for types of cased characters */
+#define UCASE_TYPE_MASK     3
+enum {
+    UCASE_NONE,
+    UCASE_LOWER,
+    UCASE_UPPER,
+    UCASE_TITLE
+};
+
+#define UCASE_GET_TYPE(props) ((props)&UCASE_TYPE_MASK)
+#define UCASE_GET_TYPE_AND_IGNORABLE(props) ((props)&7)
+
+#define UCASE_IS_UPPER_OR_TITLE(props) ((props)&2)
+
+#define UCASE_IGNORABLE         4
+#define UCASE_EXCEPTION         8
+#define UCASE_SENSITIVE         0x10
+
+#define UCASE_HAS_EXCEPTION(props) ((props)&UCASE_EXCEPTION)
+
+#define UCASE_DOT_MASK      0x60
+enum {
+    UCASE_NO_DOT=0,         /* normal characters with cc=0 */
+    UCASE_SOFT_DOTTED=0x20, /* soft-dotted characters with cc=0 */
+    UCASE_ABOVE=0x40,       /* "above" accents with cc=230 */
+    UCASE_OTHER_ACCENT=0x60 /* other accent character (0<cc!=230) */
+};
+
+/* no exception: bits 15..7 are a 9-bit signed case mapping delta */
+#define UCASE_DELTA_SHIFT   7
+#define UCASE_DELTA_MASK    0xff80
+#define UCASE_MAX_DELTA     0xff
+#define UCASE_MIN_DELTA     (-UCASE_MAX_DELTA-1)
+
+#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+#   define UCASE_GET_DELTA(props) ((int16_t)(props)>>UCASE_DELTA_SHIFT)
+#else
+#   define UCASE_GET_DELTA(props) (int16_t)(((props)&0x8000) ? (((props)>>UCASE_DELTA_SHIFT)|0xfe00) : ((uint16_t)(props)>>UCASE_DELTA_SHIFT))
+#endif
+
+/* exception: bits 15..4 are an unsigned 12-bit index into the exceptions array */
+#define UCASE_EXC_SHIFT     4
+#define UCASE_EXC_MASK      0xfff0
+#define UCASE_MAX_EXCEPTIONS ((UCASE_EXC_MASK>>UCASE_EXC_SHIFT)+1)
+
+/* definitions for 16-bit main exceptions word ------------------------------ */
+
+/* first 8 bits indicate values in optional slots */
+enum {
+    UCASE_EXC_LOWER,
+    UCASE_EXC_FOLD,
+    UCASE_EXC_UPPER,
+    UCASE_EXC_TITLE,
+    UCASE_EXC_DELTA,
+    UCASE_EXC_5,            /* reserved */
+    UCASE_EXC_CLOSURE,
+    UCASE_EXC_FULL_MAPPINGS,
+    UCASE_EXC_ALL_SLOTS     /* one past the last slot */
+};
+
+/* each slot is 2 uint16_t instead of 1 */
+#define UCASE_EXC_DOUBLE_SLOTS      0x100
+
+enum {
+    UCASE_EXC_NO_SIMPLE_CASE_FOLDING=0x200,
+    UCASE_EXC_DELTA_IS_NEGATIVE=0x400,
+    UCASE_EXC_SENSITIVE=0x800
+};
+
+/* UCASE_EXC_DOT_MASK=UCASE_DOT_MASK<<UCASE_EXC_DOT_SHIFT */
+#define UCASE_EXC_DOT_SHIFT     7
+
+/* normally stored in the main word, but pushed out for larger exception indexes */
+#define UCASE_EXC_DOT_MASK      0x3000
+enum {
+    UCASE_EXC_NO_DOT=0,
+    UCASE_EXC_SOFT_DOTTED=0x1000,
+    UCASE_EXC_ABOVE=0x2000,         /* "above" accents with cc=230 */
+    UCASE_EXC_OTHER_ACCENT=0x3000   /* other character (0<cc!=230) */
+};
+
+/* complex/conditional mappings */
+#define UCASE_EXC_CONDITIONAL_SPECIAL   0x4000
+#define UCASE_EXC_CONDITIONAL_FOLD      0x8000
+
+/* definitions for lengths word for full case mappings */
+#define UCASE_FULL_LOWER    0xf
+#define UCASE_FULL_FOLDING  0xf0
+#define UCASE_FULL_UPPER    0xf00
+#define UCASE_FULL_TITLE    0xf000
+
+/* maximum lengths */
+#define UCASE_FULL_MAPPINGS_MAX_LENGTH (4*0xf)
+#define UCASE_CLOSURE_MAX_LENGTH 0xf
+
+/* constants for reverse case folding ("unfold") data */
+enum {
+    UCASE_UNFOLD_ROWS,
+    UCASE_UNFOLD_ROW_WIDTH,
+    UCASE_UNFOLD_STRING_WIDTH
+};
+
+#endif
diff --git a/thirdparty/icu4c/common/ucase_props_data.h b/thirdparty/icu4c/common/ucase_props_data.h
new file mode 100644
index 0000000000..aead6d58d1
--- /dev/null
+++ b/thirdparty/icu4c/common/ucase_props_data.h
@@ -0,0 +1,951 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+// Copyright (C) 1999-2016, International Business Machines
+// Corporation and others.  All Rights Reserved.
+//
+// file name: ucase_props_data.h
+//
+// machine-generated by: icu/tools/unicode/c/genprops/casepropsbuilder.cpp
+
+
+#ifdef INCLUDED_FROM_UCASE_CPP
+
+static const UVersionInfo ucase_props_dataVersion={0xd,0,0,0};
+
+static const int32_t ucase_props_indexes[UCASE_IX_TOP]={0x10,0x70c2,0x6098,0x683,0x172,0,0,0,0,0,0,0,0,0,0,3};
+
+static const uint16_t ucase_props_trieIndex[12356]={
+0x336,0x33e,0x346,0x34e,0x35c,0x364,0x36c,0x374,0x37c,0x384,0x38b,0x393,0x39b,0x3a3,0x3ab,0x3b3,
+0x3b9,0x3c1,0x3c9,0x3d1,0x3d9,0x3e1,0x3e9,0x3f1,0x3f9,0x401,0x409,0x411,0x419,0x421,0x429,0x431,
+0x439,0x441,0x449,0x451,0x459,0x461,0x469,0x471,0x46d,0x475,0x47a,0x482,0x489,0x491,0x499,0x4a1,
+0x4a9,0x4b1,0x4b9,0x4c1,0x355,0x35d,0x4c6,0x4ce,0x4d3,0x4db,0x4e3,0x4eb,0x4ea,0x4f2,0x4f7,0x4ff,
+0x507,0x50e,0x512,0x355,0x355,0x355,0x519,0x521,0x529,0x52b,0x533,0x53b,0x53f,0x540,0x548,0x550,
+0x558,0x540,0x560,0x565,0x558,0x540,0x56d,0x575,0x53f,0x57d,0x585,0x58d,0x595,0x355,0x59d,0x355,
+0x5a5,0x4ec,0x5ad,0x58d,0x53f,0x57d,0x5b4,0x58d,0x5bc,0x5be,0x548,0x58d,0x53f,0x355,0x5c6,0x355,
+0x355,0x5cc,0x5d3,0x355,0x355,0x5d7,0x5df,0x355,0x5e3,0x5ea,0x355,0x5f1,0x5f9,0x600,0x608,0x355,
+0x355,0x60d,0x615,0x61d,0x625,0x62d,0x634,0x63c,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x644,0x355,0x355,0x654,0x654,0x64c,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x65c,0x65c,0x54c,0x54c,0x355,0x662,0x66a,0x355,
+0x672,0x355,0x67a,0x355,0x681,0x687,0x355,0x355,0x355,0x68f,0x355,0x355,0x355,0x355,0x355,0x355,
+0x696,0x355,0x69d,0x6a5,0x355,0x6ad,0x6b5,0x355,0x57c,0x6b8,0x6c0,0x6c6,0x5bc,0x6ce,0x355,0x6d5,
+0x355,0x6da,0x355,0x6e0,0x6e8,0x6ec,0x6f4,0x6fc,0x704,0x709,0x70c,0x714,0x724,0x71c,0x734,0x72c,
+0x37c,0x73c,0x37c,0x744,0x747,0x37c,0x74f,0x37c,0x757,0x75f,0x767,0x76f,0x777,0x77f,0x787,0x78f,
+0x797,0x79e,0x355,0x7a6,0x7ae,0x355,0x7b6,0x7be,0x7c6,0x7ce,0x7d6,0x7de,0x7e6,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x7e9,0x7ef,0x7f5,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x7fd,0x802,0x806,0x80e,0x37c,0x37c,0x37c,0x816,0x81e,0x825,0x355,0x82a,0x355,0x355,0x355,0x832,
+0x355,0x677,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x53e,0x83a,0x355,0x355,0x841,0x355,0x355,0x849,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x851,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x6e0,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x857,0x355,0x85f,0x864,0x86c,0x355,0x355,0x874,0x87c,0x884,0x37c,0x889,0x891,0x897,0x89f,0x8a2,
+0x8aa,0x8b1,0x355,0x355,0x355,0x355,0x8b8,0x8c0,0x355,0x8c8,0x8cf,0x355,0x529,0x8d4,0x8dc,0x681,
+0x355,0x8e2,0x8ea,0x8ee,0x355,0x8f6,0x8fe,0x906,0x355,0x90c,0x910,0x918,0x928,0x920,0x355,0x930,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x938,0x355,0x355,0x355,0x355,0x940,0x5bc,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x945,0x94d,0x951,0x355,0x355,0x355,0x355,0x338,0x33e,0x959,0x961,0x968,0x4ec,0x355,0x355,0x970,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0xd58,0xd58,0xd70,0xdb0,0xdf0,0xe2c,0xe6c,0xeac,0xee4,0xf24,0xf64,0xfa4,0xfe4,0x1024,0x1064,0x10a4,
+0x10e4,0x1124,0x1164,0x11a4,0x11b4,0x11e8,0x1224,0x1264,0x12a4,0x12e4,0xd54,0x1318,0x134c,0x138c,0x13a8,0x13dc,
+0x9e1,0xa11,0xa51,0xa90,0x188,0x188,0xac8,0x188,0x188,0x188,0x188,0x188,0x188,0xaf1,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0xb31,0x188,0x188,0xb66,0xba5,0xbe5,0xc1f,0xc56,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,0x188,
+0xc96,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x977,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x6b5,0x355,0x355,0x355,0x97f,0x355,0x355,0x355,
+0x355,0x987,0x98d,0x991,0x355,0x355,0x995,0x999,0x99f,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x9a7,0x9ab,0x355,0x355,0x355,0x355,0x355,0x9b3,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x9bb,0x9bf,0x9c7,0x9cb,0x355,0x9d2,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x9d8,0x355,0x355,0x355,0x355,0x9df,0x355,0x355,0x355,0x355,
+0x355,0x53f,0x9e4,0x9eb,0x5bd,0x5bc,0x9ef,0x53c,0x355,0x9f7,0x9fe,0x355,0xa04,0x5bc,0xa09,0xa11,
+0x355,0x355,0xa16,0x355,0x355,0x355,0x355,0x338,0xa1e,0x5bc,0x5be,0xa26,0xa2d,0x355,0x355,0x355,
+0x355,0x355,0x9e4,0xa35,0x355,0x355,0xa3d,0xa45,0x355,0x355,0x355,0x355,0x355,0x355,0xa49,0xa51,
+0x355,0x355,0xa59,0x4b0,0x355,0x355,0xa61,0x355,0x355,0xa67,0xa6f,0x355,0x355,0x355,0x355,0x355,
+0x355,0xa74,0x355,0x355,0x355,0xa7c,0xa84,0x355,0x355,0xa8c,0xa94,0x355,0x355,0x355,0xa97,0x6b5,
+0xa9f,0xaa3,0xaab,0x355,0xab2,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0xab9,0x355,0x355,0x940,0xac1,0x355,0x355,0x355,0xac7,0xacf,0x355,0xad3,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0xad9,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0xadf,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0xae6,0x355,0xaec,0x57c,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0xa7c,0xa84,0x355,0x355,0x355,0x355,0x355,0x355,0x677,0x355,0xaf2,0x355,0x355,
+0xafa,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0xaff,0x57c,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0xb07,0xb0f,0xb15,0x355,0x355,0x355,0x355,0xb1d,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0xb25,0xb2d,0xb32,0xb38,0xb40,0xb48,0xb50,0xb29,0xb58,0xb60,
+0xb68,0xb6f,0xb2a,0xb25,0xb2d,0xb28,0xb38,0xb2b,0xb26,0xb77,0xb29,0xb7f,0xb87,0xb8f,0xb96,0xb82,
+0xb8a,0xb92,0xb99,0xb85,0xba1,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x87c,0xba9,0x87c,0xbb0,0xbb7,0xbbf,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0xbc7,0xbcf,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0xbd3,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x9d0,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0xbdb,0x355,0xbe3,0xbeb,0xbf2,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0xb21,
+0xbfa,0xbfa,0xc00,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x9f9,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x53f,0x87c,0x87c,0x87c,0x355,0x355,0x355,0x355,0x87c,0x87c,
+0x87c,0x87c,0x87c,0x87c,0x87c,0xc08,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,0x355,
+0x355,0x355,0x355,0x355,0x355,0x355,0x335,0x335,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+0,0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,
+0,0,4,0,0,0,0,0,0,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,
+0x1012,0xa,0x5a,0x7a,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0xba,0x1012,0x1012,0x1012,0x1012,
+0x1012,0x1012,0x1012,0,0,0,4,0,4,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf9,0xf031,0x149,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0x189,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf011,0xf011,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,0,1,0,0,4,0,4,
+0,0,0,0,4,0x1c9,0,4,4,0,1,0,0,0,0,0,
+0x1012,0x1012,0x1012,0x1012,0x1012,0x1fa,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x5a,0x5a,0x1012,0x1012,
+0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x239,
+0xf011,0xf011,0xf011,0xf011,0xf011,0x2d9,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0x3c91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x31a,0xff91,0x92,0xff91,0x92,0xff91,0x31a,0xffb1,
+0x33a,0x389,0x92,0xff91,0x92,0xff91,0x92,0xff91,1,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,
+0xff91,0x92,0xff91,0x92,0xff91,0x3d9,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0xc392,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x459,0x6191,0x6912,0x92,0xff91,
+0x92,0xff91,0x6712,0x92,0xff91,0x6692,0x6692,0x92,0xff91,1,0x2792,0x6512,0x6592,0x92,0xff91,0x6692,
+0x6792,0x3091,0x6992,0x6892,0x92,0xff91,0x5191,1,0x6992,0x6a92,0x4111,0x6b12,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x6d12,0x92,0xff91,0x6d12,1,1,0x92,0xff91,0x6d12,0x92,0xff91,0x6c92,0x6c92,0x92,
+0xff91,0x92,0xff91,0x6d92,0x92,0xff91,1,0,0x92,0xff91,1,0x1c11,0,0,0,0,
+0x48a,0x4bb,0x4f9,0x52a,0x55b,0x599,0x5ca,0x5fb,0x639,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,
+0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0xd891,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x669,0x6ea,0x71b,0x759,
+0x92,0xff91,0xcf92,0xe412,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0xbf12,1,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,1,1,1,1,1,1,0x78a,0x92,
+0xff91,0xae92,0x7aa,0x7c9,0x7c9,0x92,0xff91,0x9e92,0x2292,0x2392,0x92,0xff91,0x92,0xffb1,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x7e9,0x809,0x829,0x9711,0x9911,1,0x9991,0x9991,1,0x9b11,1,0x9a91,
+0x849,1,1,1,0x9991,0x869,1,0x9891,1,0x889,0x8a9,1,0x97b1,0x9691,0x8a9,0x8c9,
+0x8e9,1,1,0x9691,1,0x909,0x9591,1,1,0x9511,1,1,1,1,1,1,
+1,0x929,1,1,0x9311,1,0x949,0x9311,1,1,1,0x969,0x9311,0xdd91,0x9391,0x9391,
+0xdc91,1,1,1,1,1,0x9291,1,0,1,1,1,1,1,1,1,
+1,0x989,0x9a9,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,5,5,0x25,5,5,5,5,5,5,4,4,4,
+0x14,4,0x14,4,5,5,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,5,5,5,5,5,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,0x54,0x54,0x44,0x44,0x44,0x44,0x44,0x9cc,0x54,0x44,0x54,0x44,
+0x54,0x44,0x44,0x44,0x44,0x44,0x44,0x54,0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x64,
+0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,
+0x64,0x64,0x64,0x64,0x64,0x74,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,
+0x64,0x44,0x44,0x44,0x44,0x44,0x54,0x44,0x44,0x9dd,0x44,0x64,0x64,0x64,0x44,0x44,
+0x44,0x64,0x64,4,0x44,0x44,0x44,0x64,0x64,0x64,0x64,0x44,0x64,0x64,0x64,0x44,
+0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x92,0xff91,0x92,0xff91,4,4,0x92,0xff91,0,0,5,0x4111,
+0x4111,0x4111,0,0x3a12,0,0,0,0,4,4,0x1312,4,0x1292,0x1292,0x1292,0,
+0x2012,0,0x1f92,0x1f92,0xa29,0x1012,0xafa,0x1012,0x1012,0xb3a,0x1012,0x1012,0xb7a,0xbca,0xc1a,0x1012,
+0xc5a,0x1012,0x1012,0x1012,0xc9a,0xcda,0,0xd1a,0x1012,0x1012,0xd5a,0x1012,0x1012,0xd9a,0x1012,0x1012,
+0xed11,0xed91,0xed91,0xed91,0xdd9,0xf011,0xea9,0xf011,0xf011,0xee9,0xf011,0xf011,0xf29,0xf79,0xfc9,0xf011,
+0x1009,0xf011,0xf011,0xf011,0x1049,0x1089,0x10c9,0x10f9,0xf011,0xf011,0x1139,0xf011,0xf011,0x1179,0xf011,0xf011,
+0xe011,0xe091,0xe091,0x412,0x11b9,0x11e9,2,2,2,0x1239,0x1269,0xfc11,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x1299,0x12c9,0x391,0xc631,0x12fa,0x1349,0,0x92,0xff91,0xfc92,0x92,0xff91,
+1,0xbf12,0xbf12,0xbf12,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,0x2812,
+0x2812,0x2812,0x2812,0x2812,0x1012,0x1012,0x137a,0x1012,0x13ba,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,
+0x1012,0x1012,0x13fa,0x1012,0x1012,0x143a,0x147a,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x14ca,0x1012,
+0x1012,0x1012,0x1012,0x1012,0xf011,0xf011,0x1509,0xf011,0x1549,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf011,0x1589,0xf011,0xf011,0x15c9,0x1609,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0x1659,0xf011,
+0xf011,0xf011,0xf011,0xf011,0xd811,0xd811,0xd811,0xd811,0xd811,0xd811,0xd831,0xd811,0xd831,0xd811,0xd811,0xd811,
+0xd811,0xd811,0xd811,0xd811,0x92,0xff91,0x169a,0x16d9,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0,0x44,0x44,0x44,0x44,0x44,4,4,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x792,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,
+0xff91,0x92,0xff91,0xf891,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,
+0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0,
+0,4,0,0,0,0,0,4,1,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,
+0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,
+0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0x1719,1,0,0,0,
+0,0,0,0,0,0x64,0x44,0x44,0x44,0x44,0x64,0x44,0x44,0x44,0x64,0x64,
+0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0x44,0x64,0x44,
+0x44,0x64,0x64,0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,
+0x64,0x64,0,0x64,0,0x64,0x64,0,0x44,0x64,0,0x64,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0,
+0,0,0,0,4,4,4,4,4,4,0,0,0,0,0,0,
+0,0,0,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0x64,0,
+4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0x64,
+0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0x44,0x64,0x64,0x44,0x44,0x44,0x44,0x44,
+0x64,0x44,0x44,0x64,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x64,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,4,0,0x44,0x44,0x44,0x44,0x64,0x44,4,4,0x44,
+0x44,0,0x64,0x44,0x44,0x64,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,4,0,0x64,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x64,0x44,0x44,
+0x64,0x44,0x44,0x64,0x64,0x64,0x44,0x64,0x64,0x44,0x64,0x44,0x44,0x44,0x64,0x44,
+0x64,0x44,0x64,0x44,0x64,0x44,0x44,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,4,
+4,4,4,4,4,4,4,4,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x44,
+4,4,0,0,0,0,4,0,0,0x64,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x44,0x44,0x44,0x44,4,0x44,0x44,0x44,0x44,0x44,4,0x44,0x44,0x44,
+4,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0x64,0x64,0x64,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,0x44,4,0x64,0x44,0x44,0x64,0x44,0x44,0x64,0x44,0x44,
+0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0x44,0x44,0x64,0x44,0x44,0x64,0x64,0x44,
+0x44,0x44,0x44,0x44,4,4,4,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,0,0x64,0,0,0,0,4,4,4,
+4,4,4,4,4,0,0,0,0,0x64,0,0,0,0x44,0x64,0x44,
+0x44,4,4,4,0,0,0,0,0,0,0,0,0,0,4,4,
+0,0,0,0,0,0,0,0,0,0,0,0,0,4,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x64,0,0,0,
+0,4,4,4,4,0,0,0,0,0,0,0,0,0x64,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,4,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x44,0,
+0,4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,4,4,0,0,0,0,4,4,0,0,4,4,0x64,0,0,
+0,4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,0,0,0,4,0,0,0,0,0,0,
+0,0,0,0,0,4,4,4,4,4,0,4,4,0,0,0,
+0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,4,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,4,
+4,4,4,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x64,0,0,4,0,4,4,4,4,0,0,0,0,0,0,0,
+0,0x64,0,0,0,0,0,0,0,4,4,0,0,0,0,0,
+0,0,0,0,0,0,4,4,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,
+0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,0,0,0,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,0,0,0,0,0,4,4,4,0,4,4,
+4,0x64,0,0,0,0,0,0,0,0x64,0x64,0,0,0,0,0,
+0,0,0,0,0,0,4,0,0,0,0,0,4,0x64,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x64,0x64,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x64,0,0,0,0,0,0,0,4,4,4,0,4,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,4,0,0,4,4,4,4,0x64,0x64,0x64,0,0,0,0,0,
+0,0,4,4,0x64,0x64,0x64,0x64,4,4,4,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,4,0,0,
+4,4,4,4,0x64,0x64,0x64,4,4,0,0,0,0,0,0,0,
+0,0,4,0,0x64,0x64,0x64,0x64,4,4,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x64,0x64,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0x64,0,0x64,
+0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0x64,0x64,4,0x64,4,4,4,4,4,0x64,0x64,
+0x64,0x64,4,0,0x64,4,0x44,0x44,0x64,0,0x44,0x44,0,0,0,0,
+0,4,4,4,4,4,4,4,4,4,4,4,0,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,0,0,0,
+0,0,0,0,0,0,0x64,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,4,4,4,4,0,4,4,4,4,4,0x64,0,0x64,0x64,0,
+0,4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,4,4,0,0,
+0,0,4,4,4,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,4,4,4,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,0,0,4,4,0,0,0,0,0,
+0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,4,0,0,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,
+0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,
+0x175a,0x175a,0x175a,0x175a,0x175a,0x175a,0,0x175a,0,0,0,0,0,0x175a,0,0,
+0x1779,0x17a9,0x17d9,0x1809,0x1839,0x1869,0x1899,0x18c9,0x18f9,0x1929,0x1959,0x1989,0x19b9,0x19e9,0x1a19,0x1a49,
+0x1a79,0x1aa9,0x1ad9,0x1b09,0x1b39,0x1b69,0x1b99,0x1bc9,0x1bf9,0x1c29,0x1c59,0x1c89,0x1cb9,0x1ce9,0x1d19,0x1d49,
+0x1d79,0x1da9,0x1dd9,0x1e09,0x1e39,0x1e69,0x1e99,0x1ec9,0x1ef9,0x1f29,0x1f59,0,4,0x1f89,0x1fb9,0x1fe9,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,0x44,
+0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,
+0x203a,0x203a,0x203a,0x203a,0x203a,0x203a,0,0,0x2059,0x2089,0x20b9,0x20e9,0x2119,0x2149,0,0,
+0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,
+0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,0x201a,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,4,4,0x64,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,4,4,0,4,
+4,4,4,4,4,4,0,0,0,0,0,0,0,0,4,0,
+0,4,4,4,4,4,4,4,4,4,0x64,4,0,0,0,4,
+0,0,0,0,0,0x44,0,0,0,0,0,0,0,0,0,0,
+0,0,0,4,4,4,4,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,4,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0x64,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,4,4,4,0,
+0,0,0,4,4,0,0,0,0,0,0,0,0,0,4,0,
+0,0,0,0,0,0x64,0x44,0x64,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x44,
+0x64,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,4,0,4,4,4,4,
+4,4,4,0,0x64,0,4,0,0,4,4,4,4,4,4,4,
+4,0,0,0,0,0,0,4,4,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0,0,0x64,0,0,0,0,0,0,0,4,0,0,0,0,
+0,0,0,0,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x44,
+0x44,0x64,4,0x64,0x64,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x64,0,4,4,4,4,4,0,4,0,0,0,
+0,0,4,0,0x60,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x44,0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,4,4,4,4,0,0,
+4,4,0x60,0x64,4,4,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0x64,0,4,4,0,0,
+0,4,0,4,4,4,0x60,0x60,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,4,4,4,4,4,4,0,0,4,0x64,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,4,4,4,4,4,0,0,
+0x2179,0x21a9,0x21d9,0x2209,0x2239,0x2289,0x22d9,0x2309,0x2339,0,0,0,0,0,0,0,
+0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,
+0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0x236a,0,0,0x236a,0x236a,0x236a,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x44,0x44,0x44,0,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0x44,0x64,0x64,0x64,0x64,
+0x44,0,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0,0,0,0,0x64,0,0,
+0,0,0,0,0x44,0,0,0,0x44,0x44,0,0,0,0,0,0,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,0x25,5,5,5,5,5,5,5,5,1,1,1,1,1,
+1,1,1,1,1,1,1,1,5,0x2389,1,1,1,0x23a9,1,1,
+5,5,5,5,0x25,5,5,5,0x25,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x23c9,1,
+1,1,1,1,1,1,0x21,1,1,1,1,5,5,5,5,5,
+0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0x64,0x64,0,0x44,0x64,0x64,0x44,0x64,
+0x44,0x44,0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x44,0x44,0x64,0x64,0x64,
+0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xffb1,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x23ea,0x2429,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x2469,0x24e9,0x2569,0x25e9,0x2669,0x26e9,1,1,0x271a,1,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xffb1,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x411,0x411,0x411,0x411,
+0x411,0x411,0x411,0x411,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0x411,0x411,0x411,0x411,
+0x411,0x411,0,0,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0,0,0x411,0x411,0x411,0x411,
+0x411,0x411,0x411,0x411,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0x411,0x411,0x411,0x411,
+0x411,0x411,0x411,0x411,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0x411,0x411,0x411,0x411,
+0x411,0x411,0,0,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0,0,0x2769,0x411,0x27e9,0x411,
+0x2899,0x411,0x2949,0x411,0,0xfc12,0,0xfc12,0,0xfc12,0,0xfc12,0x411,0x411,0x411,0x411,
+0x411,0x411,0x411,0x411,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0xfc12,0x2511,0x2511,0x2b11,0x2b11,
+0x2b11,0x2b11,0x3211,0x3211,0x4011,0x4011,0x3811,0x3811,0x3f11,0x3f11,0,0,0x29f9,0x2a69,0x2ad9,0x2b49,
+0x2bb9,0x2c29,0x2c99,0x2d09,0x2d7b,0x2deb,0x2e5b,0x2ecb,0x2f3b,0x2fab,0x301b,0x308b,0x30f9,0x3169,0x31d9,0x3249,
+0x32b9,0x3329,0x3399,0x3409,0x347b,0x34eb,0x355b,0x35cb,0x363b,0x36ab,0x371b,0x378b,0x37f9,0x3869,0x38d9,0x3949,
+0x39b9,0x3a29,0x3a99,0x3b09,0x3b7b,0x3beb,0x3c5b,0x3ccb,0x3d3b,0x3dab,0x3e1b,0x3e8b,0x411,0x411,0x3ef9,0x3f79,
+0x3fe9,0,0x4069,0x40e9,0xfc12,0xfc12,0xdb12,0xdb12,0x419b,4,0x4209,4,4,4,0x4259,0x42d9,
+0x4349,0,0x43c9,0x4449,0xd512,0xd512,0xd512,0xd512,0x44fb,4,4,4,0x411,0x411,0x4569,0x4619,
+0,0,0x46e9,0x4769,0xfc12,0xfc12,0xce12,0xce12,0,4,4,4,0x411,0x411,0x4819,0x48c9,
+0x4999,0x391,0x4a19,0x4a99,0xfc12,0xfc12,0xc812,0xc812,0xfc92,4,4,4,0,0,0x4b49,0x4bc9,
+0x4c39,0,0x4cb9,0x4d39,0xc012,0xc012,0xc112,0xc112,0x4deb,4,4,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,4,4,0,0,0,0,
+0,0,0,0,4,4,0,0,0,0,0,0,4,0,0,4,
+0,0,4,4,4,4,4,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,4,4,4,4,0,4,4,
+4,4,4,4,4,4,4,4,0,0x25,0,0,0,0,0,0,
+0,0,0,0,0,0,0,5,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x44,0x44,0x64,0x64,0x44,0x44,0x44,0x44,
+0x64,0x64,0x64,0x44,0x44,4,4,4,4,0x44,4,4,4,0x64,0x64,0x44,
+0x64,0x44,0x64,0x64,0x64,0x64,0x64,0x64,0x44,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,2,0,0,0,0,2,
+0,0,1,2,2,2,1,1,2,2,2,1,0,2,0,0,
+0,2,2,2,2,2,0,0,0,0,0,0,2,0,0x4e5a,0,
+2,0,0x4e9a,0x4eda,2,2,0,1,2,2,0xe12,2,1,0,0,0,
+0,1,0,0,1,1,2,2,0,0,0,0,0,2,1,1,
+0x21,0x21,0,0,0,0,0xf211,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x812,0x812,0x812,0x812,0x812,0x812,0x812,0x812,
+0x812,0x812,0x812,0x812,0x812,0x812,0x812,0x812,0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,
+0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,0xf811,0,0,0,0x92,0xff91,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,
+0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xd12,0xf311,0xf311,0xf311,0xf311,
+0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0xf311,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,
+0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,
+0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0x1812,0,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,
+0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,
+0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0xe811,0,0x92,0xff91,0x4f1a,0x4f3a,0x4f5a,0x4f79,0x4f99,0x92,
+0xff91,0x92,0xff91,0x92,0xff91,0x4fba,0x4fda,0x4ffa,0x501a,1,0x92,0xff91,1,0x92,0xff91,1,
+1,1,1,1,0x25,5,0x503a,0x503a,0x92,0xff91,0x92,0xff91,1,0,0,0,
+0,0,0,0x92,0xff91,0x92,0xff91,0x44,0x44,0x44,0x92,0xff91,0,0,0,0,
+0,0,0,0,0,0,0,0,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,
+0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,
+0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0x5059,0,0x5059,0,0,0,0,
+0,0x5059,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0x64,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0,0,0,
+0,0,0x64,0x64,0x64,0x64,0x60,0x60,0,4,4,4,4,4,0,0,
+0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0x64,0x64,4,
+4,4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x507a,0x50b9,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0,0x44,
+4,4,4,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0,4,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,5,5,0x44,0x44,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x44,0x44,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+1,1,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,5,1,1,1,1,1,1,1,1,0x92,0xff91,0x92,
+0xff91,0x50fa,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,4,4,4,0x92,
+0xff91,0x511a,1,0,0x92,0xff91,0x92,0xff91,0x1811,1,0x92,0xff91,0x92,0xff91,0x92,0xff91,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x513a,0x515a,0x517a,0x519a,0x513a,1,0x51ba,0x51da,0x51fa,0x521a,
+0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0x92,0xff91,0,0,0x92,0xff91,
+0xe812,0x523a,0x525a,0x92,0xff91,0x92,0xff91,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0x92,0xff91,0,
+5,5,1,0,0,0,0,0,0,0,4,0,0,0,0x64,0,
+0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,4,0,0,0,0,0,
+0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x64,4,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+0,0,0,0,0,0,4,4,4,4,4,0x64,0x64,0x64,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,4,4,4,4,4,4,4,4,4,4,4,0,0x60,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x64,0,0,4,4,4,4,0,0,4,4,0,0,
+0x60,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,4,4,4,4,4,4,0,0,4,4,0,0,4,4,0,
+0,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0,
+0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0,
+0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x44,0,0x44,0x44,0x64,0,0,0x44,
+0x44,0,0,0,0,0,0x44,0x44,0,0x44,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,0,0,0,0,0,4,4,0,0x64,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,0x5279,1,1,1,1,1,1,1,4,5,5,5,5,
+1,1,1,1,1,1,1,1,1,4,4,4,0,0,0,0,
+0x5299,0x52c9,0x52f9,0x5329,0x5359,0x5389,0x53b9,0x53e9,0x5419,0x5449,0x5479,0x54a9,0x54d9,0x5509,0x5539,0x5569,
+0x5b99,0x5bc9,0x5bf9,0x5c29,0x5c59,0x5c89,0x5cb9,0x5ce9,0x5d19,0x5d49,0x5d79,0x5da9,0x5dd9,0x5e09,0x5e39,0x5e69,
+0x5e99,0x5ec9,0x5ef9,0x5f29,0x5f59,0x5f89,0x5fb9,0x5fe9,0x6019,0x6049,0x6079,0x60a9,0x60d9,0x6109,0x6139,0x6169,
+0x5599,0x55c9,0x55f9,0x5629,0x5659,0x5689,0x56b9,0x56e9,0x5719,0x5749,0x5779,0x57a9,0x57d9,0x5809,0x5839,0x5869,
+0x5899,0x58c9,0x58f9,0x5929,0x5959,0x5989,0x59b9,0x59e9,0x5a19,0x5a49,0x5a79,0x5aa9,0x5ad9,0x5b09,0x5b39,0x5b69,
+0,0,0,0,0,4,0,0,4,0,0,0,0,0x64,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x6199,0x6219,0x6299,0x6319,0x63c9,0x6479,0x6519,0,0,0,0,0,0,0,0,0,
+0,0,0,0x65b9,0x6639,0x66b9,0x6739,0x67b9,0,0,0,0,0,0,0x64,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,0,0,0,4,0,0,0,0,0,0,0,0,
+0,0,0,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0x64,0x64,0x64,
+0x64,0x64,0x44,0x44,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,0,0,4,0,0,0,0,0,0,
+0,0,0,0,0,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,
+0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0,
+0,0,4,0,4,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,4,4,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0x64,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0x1412,0x1412,0x1412,0x1412,
+0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,
+0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0xec11,0xec11,0xec11,0xec11,
+0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,
+0xec11,0xec11,0xec11,0xec11,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,
+0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0x1412,0,0,0,0,0xec11,0xec11,0xec11,0xec11,
+0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,
+0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0xec11,0,0,0,0,0,4,4,4,
+0,4,4,0,0,0,0,0,4,0x64,4,0x44,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x44,0x64,0x64,0,0,0,0,0x64,0,0,0,0,
+0,0x44,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x2012,0x2012,0x2012,0x2012,
+0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,
+0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0x2012,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0xe011,0xe011,0xe011,0xe011,
+0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,
+0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0xe011,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,0x44,0x44,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x64,0x64,0x44,0x44,0x44,0x64,0x44,0x64,0x64,0x64,0x64,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,4,4,4,4,4,4,4,
+4,4,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+4,4,4,0,0,0x64,0x64,0,0,4,0,0,0x44,0x44,0x44,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+4,4,4,4,0,4,4,4,4,4,4,0x64,0x64,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,4,4,4,4,4,4,
+4,4,4,0,0x60,0,0,0,0,0,0,0,0,4,0x64,4,
+4,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,0,0,4,0x60,0x64,4,
+0,0,0,0,0,0,4,0,0,0,0,4,4,4,4,4,
+4,0x64,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0x60,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0,0,0,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0x64,4,4,0,0x64,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0x44,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,4,4,4,0,4,0,
+0,0,0,4,4,0,0x64,0x64,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,4,4,4,4,0,0,0,0,0,0,
+4,4,0,0x64,0x64,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,4,4,4,4,4,0,
+0,4,0,0x64,0,0,0,0,0,0,0,0,0,0,0,4,
+0,4,0,0,4,4,4,4,4,4,0x60,0x64,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,4,4,0,0,4,4,
+4,4,0,4,4,4,4,0x64,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+4,4,4,4,4,4,4,4,0,0x64,0x64,0,0,0,0,0,
+0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,
+0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,0x1012,
+0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,0xf011,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,4,4,0x60,0x64,0,
+0,0,0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,4,4,0,0,4,4,0,0,0,0,0,4,4,4,
+4,4,4,4,4,4,4,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,4,
+0x64,4,4,4,4,0,0,4,4,4,4,0,0,0,0,0,
+0,0,0,0x64,0,0,0,0,0,0,0,0,0,4,4,4,
+4,4,4,0,0,4,4,4,0,0,0,0,0,0,0,0,
+0,0,4,4,4,4,4,4,4,4,4,4,4,4,4,0,
+4,0x64,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,4,4,4,4,4,0,4,4,4,4,
+4,4,0,0x64,4,4,4,4,4,4,4,4,0,0,4,4,
+4,4,4,4,4,0,4,4,0,4,4,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,4,4,4,
+4,4,4,0,0,0,4,0,4,4,0,4,4,4,0x64,4,
+0x64,0x64,0,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,4,4,0,0,
+0,4,0,0x64,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,4,4,4,4,
+4,4,4,4,4,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x64,0x64,0x64,0x64,0x64,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x44,0x44,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,0,4,4,0,0,0,
+0,0,0,0,0,0,0,0,0x60,0x60,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,4,0x64,0,0,0,0,0,
+0,0x60,0x60,0x64,0x64,0x64,0,0,0,0x60,0x60,0x60,0x60,0x60,0x60,4,
+4,4,4,4,4,4,4,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64,0,
+0,0x44,0x44,0x44,0x44,0x44,0x64,0x64,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,
+0x44,0x44,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0x44,0x44,0x44,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,1,1,
+1,1,1,1,1,1,0x21,0x21,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,1,1,1,1,1,1,1,0,0x21,0x21,
+1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,1,1,1,1,1,1,1,1,0x21,0x21,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,2,0,2,2,
+0,0,2,0,0,2,2,0,0,2,2,2,2,0,2,2,
+2,2,2,2,2,2,1,1,1,1,0,1,0,1,0x21,0x21,
+1,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+1,1,1,1,2,2,0,2,2,2,2,0,0,2,2,2,
+2,2,2,2,2,0,2,2,2,2,2,2,2,0,1,1,
+1,1,1,1,1,1,0x21,0x21,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,2,2,0,2,2,2,2,0,
+2,2,2,2,2,0,2,0,0,0,2,2,2,2,2,2,
+2,0,1,1,1,1,1,1,1,1,0x21,0x21,1,1,1,1,
+1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,0,0,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,0,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0,1,1,1,1,1,1,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,0,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0,1,1,
+1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,1,1,1,0,1,1,1,1,1,1,2,1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,0,0,0,0,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,0,0,0,
+0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,
+4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,4,4,4,4,4,0,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,0x44,0x44,0,0,0x44,0x44,0x44,0x44,0x44,0x44,0x44,0,0x44,
+0x44,0,0x44,0x44,0x44,0x44,0x44,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x44,0x44,0x44,0x44,
+0x44,0x44,0x44,4,4,4,4,4,4,4,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x64,0x64,0x64,0x64,
+0x64,0x64,0x64,0,0,0,0,0,0,0,0,0,0x1112,0x1112,0x1112,0x1112,
+0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,
+0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0x1112,0xef11,0xef11,
+0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,
+0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0xef11,0x44,0x44,0x44,0x44,
+0x44,0x44,0x64,4,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,2,2,2,2,2,2,2,2,
+2,2,0,0,0,0,0,0,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0
+};
+
+static const uint16_t ucase_props_exceptions[1667]={
+0xc850,0x20,2,0x130,0x131,0x4810,0x20,0x841,0x6b,1,0x212a,0x841,0x73,1,0x17f,0x5c50,
+0x20,2,0x130,0x131,0x844,0x4b,1,0x212a,0x844,0x53,1,0x17f,0x806,0x3bc,0x39c,0x841,
+0xe5,1,0x212b,0x8c0,1,0x2220,0x73,0x73,0x53,0x53,0x53,0x73,0x1e9e,0x844,0xc5,1,
+0x212b,0x4810,1,0xce50,0xc7,2,0x49,0x131,0x844,0x49,2,0x69,0x130,0x880,0x2220,0x2bc,
+0x6e,0x2bc,0x4e,0x2bc,0x4e,0x806,0x73,0x53,0x809,0x1c6,0x1c5,0x80d,0x1c6,0x1c4,0x1c5,0x80c,
+0x1c4,0x1c5,0x809,0x1c9,0x1c8,0x80d,0x1c9,0x1c7,0x1c8,0x80c,0x1c7,0x1c8,0x809,0x1cc,0x1cb,0x80d,
+0x1cc,0x1ca,0x1cb,0x80c,0x1ca,0x1cb,0x880,0x2220,0x6a,0x30c,0x4a,0x30c,0x4a,0x30c,0x809,0x1f3,
+0x1f2,0x80d,0x1f3,0x1f1,0x1f2,0x80c,0x1f1,0x1f2,0x810,0x2a2b,0x810,0x2a28,0x810,0x2a3f,0x810,0x2a1f,
+0x810,0x2a1c,0x810,0x2a1e,0x810,0xa54f,0x810,0xa54b,0x810,0xa528,0x810,0xa544,0x810,0x29f7,0x810,0xa541,
+0x810,0x29fd,0x810,0x29e7,0x810,0xa543,0x810,0xa52a,0x1810,0xa515,0x810,0xa512,0x6800,0x3846,0x3b9,0x399,
+1,0x1fbe,0x8c0,1,0x3330,0x3b9,0x308,0x301,0x399,0x308,0x301,0x399,0x308,0x301,0x1fd3,0x841,
+0x3b2,1,0x3d0,0x841,0x3b5,1,0x3f5,0x841,0x3b8,2,0x3d1,0x3f4,0x841,0x3b9,2,0x345,
+0x1fbe,0x841,0x3ba,1,0x3f0,0x841,0x3bc,1,0xb5,0x841,0x3c0,1,0x3d6,0x841,0x3c1,1,
+0x3f1,0x4850,0x20,1,0x3c2,0x841,0x3c6,1,0x3d5,0x841,0x3c9,1,0x2126,0x8c0,1,0x3330,
+0x3c5,0x308,0x301,0x3a5,0x308,0x301,0x3a5,0x308,0x301,0x1fe3,0x844,0x392,1,0x3d0,0x844,0x395,
+1,0x3f5,0x844,0x398,2,0x3d1,0x3f4,0x844,0x399,2,0x345,0x1fbe,0x844,0x39a,1,0x3f0,
+0x844,0x39c,1,0xb5,0x844,0x3a0,1,0x3d6,0x844,0x3a1,1,0x3f1,0x806,0x3c3,0x3a3,0x844,
+0x3a3,1,0x3c2,0x844,0x3a6,1,0x3d5,0x844,0x3a9,1,0x2126,0x806,0x3b2,0x392,0x846,0x3b8,
+0x398,1,0x3f4,0x806,0x3c6,0x3a6,0x806,0x3c0,0x3a0,0x806,0x3ba,0x39a,0x806,0x3c1,0x3a1,0x841,
+0x3b8,2,0x398,0x3d1,0x806,0x3b5,0x395,0x841,0x432,1,0x1c80,0x841,0x434,1,0x1c81,0x841,
+0x43e,1,0x1c82,0x841,0x441,1,0x1c83,0x841,0x442,2,0x1c84,0x1c85,0x841,0x44a,1,0x1c86,
+0x844,0x412,1,0x1c80,0x844,0x414,1,0x1c81,0x844,0x41e,1,0x1c82,0x844,0x421,1,0x1c83,
+0x844,0x422,2,0x1c84,0x1c85,0x844,0x42a,1,0x1c86,0x841,0x463,1,0x1c87,0x844,0x462,1,
+0x1c87,0x4880,0x20,0x565,0x582,0x810,0x1c60,0x80c,0x1c90,0x10d0,0x80c,0x1c91,0x10d1,0x80c,0x1c92,0x10d2,
+0x80c,0x1c93,0x10d3,0x80c,0x1c94,0x10d4,0x80c,0x1c95,0x10d5,0x80c,0x1c96,0x10d6,0x80c,0x1c97,0x10d7,0x80c,
+0x1c98,0x10d8,0x80c,0x1c99,0x10d9,0x80c,0x1c9a,0x10da,0x80c,0x1c9b,0x10db,0x80c,0x1c9c,0x10dc,0x80c,0x1c9d,
+0x10dd,0x80c,0x1c9e,0x10de,0x80c,0x1c9f,0x10df,0x80c,0x1ca0,0x10e0,0x80c,0x1ca1,0x10e1,0x80c,0x1ca2,0x10e2,
+0x80c,0x1ca3,0x10e3,0x80c,0x1ca4,0x10e4,0x80c,0x1ca5,0x10e5,0x80c,0x1ca6,0x10e6,0x80c,0x1ca7,0x10e7,0x80c,
+0x1ca8,0x10e8,0x80c,0x1ca9,0x10e9,0x80c,0x1caa,0x10ea,0x80c,0x1cab,0x10eb,0x80c,0x1cac,0x10ec,0x80c,0x1cad,
+0x10ed,0x80c,0x1cae,0x10ee,0x80c,0x1caf,0x10ef,0x80c,0x1cb0,0x10f0,0x80c,0x1cb1,0x10f1,0x80c,0x1cb2,0x10f2,
+0x80c,0x1cb3,0x10f3,0x80c,0x1cb4,0x10f4,0x80c,0x1cb5,0x10f5,0x80c,0x1cb6,0x10f6,0x80c,0x1cb7,0x10f7,0x80c,
+0x1cb8,0x10f8,0x80c,0x1cb9,0x10f9,0x80c,0x1cba,0x10fa,0x80c,0x1cbd,0x10fd,0x80c,0x1cbe,0x10fe,0x80c,0x1cbf,
+0x10ff,0xa10,0x97d0,0xa10,8,0x806,0x13f0,0x13f0,0x806,0x13f1,0x13f1,0x806,0x13f2,0x13f2,0x806,0x13f3,
+0x13f3,0x806,0x13f4,0x13f4,0x806,0x13f5,0x13f5,0x806,0x432,0x412,0x806,0x434,0x414,0x806,0x43e,0x41e,
+0x806,0x441,0x421,0x846,0x442,0x422,1,0x1c85,0x846,0x442,0x422,1,0x1c84,0x806,0x44a,0x42a,
+0x806,0x463,0x462,0x806,0xa64b,0xa64a,0xc10,0xbc0,0x810,0x8a04,0x810,0xee6,0x810,0x8a38,0x841,0x1e61,
+1,0x1e9b,0x844,0x1e60,1,0x1e9b,0x880,0x2220,0x68,0x331,0x48,0x331,0x48,0x331,0x880,0x2220,
+0x74,0x308,0x54,0x308,0x54,0x308,0x880,0x2220,0x77,0x30a,0x57,0x30a,0x57,0x30a,0x880,0x2220,
+0x79,0x30a,0x59,0x30a,0x59,0x30a,0x880,0x2220,0x61,0x2be,0x41,0x2be,0x41,0x2be,0x806,0x1e61,
+0x1e60,0xc90,0x1dbf,0x20,0x73,0x73,0x880,0x2220,0x3c5,0x313,0x3a5,0x313,0x3a5,0x313,0x880,0x3330,
+0x3c5,0x313,0x300,0x3a5,0x313,0x300,0x3a5,0x313,0x300,0x880,0x3330,0x3c5,0x313,0x301,0x3a5,0x313,
+0x301,0x3a5,0x313,0x301,0x880,0x3330,0x3c5,0x313,0x342,0x3a5,0x313,0x342,0x3a5,0x313,0x342,0x890,
+8,0x220,0x1f00,0x3b9,0x1f08,0x399,0x890,8,0x220,0x1f01,0x3b9,0x1f09,0x399,0x890,8,0x220,
+0x1f02,0x3b9,0x1f0a,0x399,0x890,8,0x220,0x1f03,0x3b9,0x1f0b,0x399,0x890,8,0x220,0x1f04,0x3b9,
+0x1f0c,0x399,0x890,8,0x220,0x1f05,0x3b9,0x1f0d,0x399,0x890,8,0x220,0x1f06,0x3b9,0x1f0e,0x399,
+0x890,8,0x220,0x1f07,0x3b9,0x1f0f,0x399,0xc90,8,0x220,0x1f00,0x3b9,0x1f08,0x399,0xc90,8,
+0x220,0x1f01,0x3b9,0x1f09,0x399,0xc90,8,0x220,0x1f02,0x3b9,0x1f0a,0x399,0xc90,8,0x220,0x1f03,
+0x3b9,0x1f0b,0x399,0xc90,8,0x220,0x1f04,0x3b9,0x1f0c,0x399,0xc90,8,0x220,0x1f05,0x3b9,0x1f0d,
+0x399,0xc90,8,0x220,0x1f06,0x3b9,0x1f0e,0x399,0xc90,8,0x220,0x1f07,0x3b9,0x1f0f,0x399,0x890,
+8,0x220,0x1f20,0x3b9,0x1f28,0x399,0x890,8,0x220,0x1f21,0x3b9,0x1f29,0x399,0x890,8,0x220,
+0x1f22,0x3b9,0x1f2a,0x399,0x890,8,0x220,0x1f23,0x3b9,0x1f2b,0x399,0x890,8,0x220,0x1f24,0x3b9,
+0x1f2c,0x399,0x890,8,0x220,0x1f25,0x3b9,0x1f2d,0x399,0x890,8,0x220,0x1f26,0x3b9,0x1f2e,0x399,
+0x890,8,0x220,0x1f27,0x3b9,0x1f2f,0x399,0xc90,8,0x220,0x1f20,0x3b9,0x1f28,0x399,0xc90,8,
+0x220,0x1f21,0x3b9,0x1f29,0x399,0xc90,8,0x220,0x1f22,0x3b9,0x1f2a,0x399,0xc90,8,0x220,0x1f23,
+0x3b9,0x1f2b,0x399,0xc90,8,0x220,0x1f24,0x3b9,0x1f2c,0x399,0xc90,8,0x220,0x1f25,0x3b9,0x1f2d,
+0x399,0xc90,8,0x220,0x1f26,0x3b9,0x1f2e,0x399,0xc90,8,0x220,0x1f27,0x3b9,0x1f2f,0x399,0x890,
+8,0x220,0x1f60,0x3b9,0x1f68,0x399,0x890,8,0x220,0x1f61,0x3b9,0x1f69,0x399,0x890,8,0x220,
+0x1f62,0x3b9,0x1f6a,0x399,0x890,8,0x220,0x1f63,0x3b9,0x1f6b,0x399,0x890,8,0x220,0x1f64,0x3b9,
+0x1f6c,0x399,0x890,8,0x220,0x1f65,0x3b9,0x1f6d,0x399,0x890,8,0x220,0x1f66,0x3b9,0x1f6e,0x399,
+0x890,8,0x220,0x1f67,0x3b9,0x1f6f,0x399,0xc90,8,0x220,0x1f60,0x3b9,0x1f68,0x399,0xc90,8,
+0x220,0x1f61,0x3b9,0x1f69,0x399,0xc90,8,0x220,0x1f62,0x3b9,0x1f6a,0x399,0xc90,8,0x220,0x1f63,
+0x3b9,0x1f6b,0x399,0xc90,8,0x220,0x1f64,0x3b9,0x1f6c,0x399,0xc90,8,0x220,0x1f65,0x3b9,0x1f6d,
+0x399,0xc90,8,0x220,0x1f66,0x3b9,0x1f6e,0x399,0xc90,8,0x220,0x1f67,0x3b9,0x1f6f,0x399,0x880,
+0x2220,0x1f70,0x3b9,0x1fba,0x399,0x1fba,0x345,0x890,9,0x220,0x3b1,0x3b9,0x391,0x399,0x880,0x2220,
+0x3ac,0x3b9,0x386,0x399,0x386,0x345,0x880,0x2220,0x3b1,0x342,0x391,0x342,0x391,0x342,0x880,0x3330,
+0x3b1,0x342,0x3b9,0x391,0x342,0x399,0x391,0x342,0x345,0xc90,9,0x220,0x3b1,0x3b9,0x391,0x399,
+0x846,0x3b9,0x399,1,0x345,0x880,0x2220,0x1f74,0x3b9,0x1fca,0x399,0x1fca,0x345,0x890,9,0x220,
+0x3b7,0x3b9,0x397,0x399,0x880,0x2220,0x3ae,0x3b9,0x389,0x399,0x389,0x345,0x880,0x2220,0x3b7,0x342,
+0x397,0x342,0x397,0x342,0x880,0x3330,0x3b7,0x342,0x3b9,0x397,0x342,0x399,0x397,0x342,0x345,0xc90,
+9,0x220,0x3b7,0x3b9,0x397,0x399,0x880,0x3330,0x3b9,0x308,0x300,0x399,0x308,0x300,0x399,0x308,
+0x300,0x8c0,1,0x3330,0x3b9,0x308,0x301,0x399,0x308,0x301,0x399,0x308,0x301,0x390,0x880,0x2220,
+0x3b9,0x342,0x399,0x342,0x399,0x342,0x880,0x3330,0x3b9,0x308,0x342,0x399,0x308,0x342,0x399,0x308,
+0x342,0x880,0x3330,0x3c5,0x308,0x300,0x3a5,0x308,0x300,0x3a5,0x308,0x300,0x8c0,1,0x3330,0x3c5,
+0x308,0x301,0x3a5,0x308,0x301,0x3a5,0x308,0x301,0x3b0,0x880,0x2220,0x3c1,0x313,0x3a1,0x313,0x3a1,
+0x313,0x880,0x2220,0x3c5,0x342,0x3a5,0x342,0x3a5,0x342,0x880,0x3330,0x3c5,0x308,0x342,0x3a5,0x308,
+0x342,0x3a5,0x308,0x342,0x880,0x2220,0x1f7c,0x3b9,0x1ffa,0x399,0x1ffa,0x345,0x890,9,0x220,0x3c9,
+0x3b9,0x3a9,0x399,0x880,0x2220,0x3ce,0x3b9,0x38f,0x399,0x38f,0x345,0x880,0x2220,0x3c9,0x342,0x3a9,
+0x342,0x3a9,0x342,0x880,0x3330,0x3c9,0x342,0x3b9,0x3a9,0x342,0x399,0x3a9,0x342,0x345,0xc90,9,
+0x220,0x3c9,0x3b9,0x3a9,0x399,0xc50,0x1d5d,1,0x3a9,0xc50,0x20bf,1,0x4b,0xc50,0x2046,1,
+0xc5,0xc10,0x29f7,0xc10,0xee6,0xc10,0x29e7,0xc10,0x2a2b,0xc10,0x2a28,0xc10,0x2a1c,0xc10,0x29fd,0xc10,
+0x2a1f,0xc10,0x2a1e,0xc10,0x2a3f,0xc10,0x1c60,0x841,0xa64b,1,0x1c88,0x844,0xa64a,1,0x1c88,0xc10,
+0x8a04,0xc10,0xa528,0xc10,0xa544,0xc10,0xa54f,0xc10,0xa54b,0xc10,0xa541,0xc10,0xa512,0xc10,0xa52a,0xc10,
+0xa515,0x810,0x3a0,0xc10,0xa543,0xc10,0x8a38,0xc10,0x3a0,0x806,0x13a0,0x13a0,0x806,0x13a1,0x13a1,0x806,
+0x13a2,0x13a2,0x806,0x13a3,0x13a3,0x806,0x13a4,0x13a4,0x806,0x13a5,0x13a5,0x806,0x13a6,0x13a6,0x806,0x13a7,
+0x13a7,0x806,0x13a8,0x13a8,0x806,0x13a9,0x13a9,0x806,0x13aa,0x13aa,0x806,0x13ab,0x13ab,0x806,0x13ac,0x13ac,
+0x806,0x13ad,0x13ad,0x806,0x13ae,0x13ae,0x806,0x13af,0x13af,0x806,0x13b0,0x13b0,0x806,0x13b1,0x13b1,0x806,
+0x13b2,0x13b2,0x806,0x13b3,0x13b3,0x806,0x13b4,0x13b4,0x806,0x13b5,0x13b5,0x806,0x13b6,0x13b6,0x806,0x13b7,
+0x13b7,0x806,0x13b8,0x13b8,0x806,0x13b9,0x13b9,0x806,0x13ba,0x13ba,0x806,0x13bb,0x13bb,0x806,0x13bc,0x13bc,
+0x806,0x13bd,0x13bd,0x806,0x13be,0x13be,0x806,0x13bf,0x13bf,0x806,0x13c0,0x13c0,0x806,0x13c1,0x13c1,0x806,
+0x13c2,0x13c2,0x806,0x13c3,0x13c3,0x806,0x13c4,0x13c4,0x806,0x13c5,0x13c5,0x806,0x13c6,0x13c6,0x806,0x13c7,
+0x13c7,0x806,0x13c8,0x13c8,0x806,0x13c9,0x13c9,0x806,0x13ca,0x13ca,0x806,0x13cb,0x13cb,0x806,0x13cc,0x13cc,
+0x806,0x13cd,0x13cd,0x806,0x13ce,0x13ce,0x806,0x13cf,0x13cf,0x806,0x13d0,0x13d0,0x806,0x13d1,0x13d1,0x806,
+0x13d2,0x13d2,0x806,0x13d3,0x13d3,0x806,0x13d4,0x13d4,0x806,0x13d5,0x13d5,0x806,0x13d6,0x13d6,0x806,0x13d7,
+0x13d7,0x806,0x13d8,0x13d8,0x806,0x13d9,0x13d9,0x806,0x13da,0x13da,0x806,0x13db,0x13db,0x806,0x13dc,0x13dc,
+0x806,0x13dd,0x13dd,0x806,0x13de,0x13de,0x806,0x13df,0x13df,0x806,0x13e0,0x13e0,0x806,0x13e1,0x13e1,0x806,
+0x13e2,0x13e2,0x806,0x13e3,0x13e3,0x806,0x13e4,0x13e4,0x806,0x13e5,0x13e5,0x806,0x13e6,0x13e6,0x806,0x13e7,
+0x13e7,0x806,0x13e8,0x13e8,0x806,0x13e9,0x13e9,0x806,0x13ea,0x13ea,0x806,0x13eb,0x13eb,0x806,0x13ec,0x13ec,
+0x806,0x13ed,0x13ed,0x806,0x13ee,0x13ee,0x806,0x13ef,0x13ef,0x880,0x2220,0x66,0x66,0x46,0x46,0x46,
+0x66,0x880,0x2220,0x66,0x69,0x46,0x49,0x46,0x69,0x880,0x2220,0x66,0x6c,0x46,0x4c,0x46,
+0x6c,0x880,0x3330,0x66,0x66,0x69,0x46,0x46,0x49,0x46,0x66,0x69,0x880,0x3330,0x66,0x66,
+0x6c,0x46,0x46,0x4c,0x46,0x66,0x6c,0x8c0,1,0x2220,0x73,0x74,0x53,0x54,0x53,0x74,
+0xfb06,0x8c0,1,0x2220,0x73,0x74,0x53,0x54,0x53,0x74,0xfb05,0x880,0x2220,0x574,0x576,0x544,
+0x546,0x544,0x576,0x880,0x2220,0x574,0x565,0x544,0x535,0x544,0x565,0x880,0x2220,0x574,0x56b,0x544,
+0x53b,0x544,0x56b,0x880,0x2220,0x57e,0x576,0x54e,0x546,0x54e,0x576,0x880,0x2220,0x574,0x56d,0x544,
+0x53d,0x544,0x56d
+};
+
+static const uint16_t ucase_props_unfold[370]={
+0x49,5,3,0,0,0x61,0x2be,0,0x1e9a,0,0x66,0x66,0,0xfb00,0,0x66,
+0x66,0x69,0xfb03,0,0x66,0x66,0x6c,0xfb04,0,0x66,0x69,0,0xfb01,0,0x66,0x6c,
+0,0xfb02,0,0x68,0x331,0,0x1e96,0,0x69,0x307,0,0x130,0,0x6a,0x30c,0,
+0x1f0,0,0x73,0x73,0,0xdf,0x1e9e,0x73,0x74,0,0xfb05,0xfb06,0x74,0x308,0,0x1e97,
+0,0x77,0x30a,0,0x1e98,0,0x79,0x30a,0,0x1e99,0,0x2bc,0x6e,0,0x149,0,
+0x3ac,0x3b9,0,0x1fb4,0,0x3ae,0x3b9,0,0x1fc4,0,0x3b1,0x342,0,0x1fb6,0,0x3b1,
+0x342,0x3b9,0x1fb7,0,0x3b1,0x3b9,0,0x1fb3,0x1fbc,0x3b7,0x342,0,0x1fc6,0,0x3b7,0x342,
+0x3b9,0x1fc7,0,0x3b7,0x3b9,0,0x1fc3,0x1fcc,0x3b9,0x308,0x300,0x1fd2,0,0x3b9,0x308,0x301,
+0x390,0x1fd3,0x3b9,0x308,0x342,0x1fd7,0,0x3b9,0x342,0,0x1fd6,0,0x3c1,0x313,0,0x1fe4,
+0,0x3c5,0x308,0x300,0x1fe2,0,0x3c5,0x308,0x301,0x3b0,0x1fe3,0x3c5,0x308,0x342,0x1fe7,0,
+0x3c5,0x313,0,0x1f50,0,0x3c5,0x313,0x300,0x1f52,0,0x3c5,0x313,0x301,0x1f54,0,0x3c5,
+0x313,0x342,0x1f56,0,0x3c5,0x342,0,0x1fe6,0,0x3c9,0x342,0,0x1ff6,0,0x3c9,0x342,
+0x3b9,0x1ff7,0,0x3c9,0x3b9,0,0x1ff3,0x1ffc,0x3ce,0x3b9,0,0x1ff4,0,0x565,0x582,0,
+0x587,0,0x574,0x565,0,0xfb14,0,0x574,0x56b,0,0xfb15,0,0x574,0x56d,0,0xfb17,
+0,0x574,0x576,0,0xfb13,0,0x57e,0x576,0,0xfb16,0,0x1f00,0x3b9,0,0x1f80,0x1f88,
+0x1f01,0x3b9,0,0x1f81,0x1f89,0x1f02,0x3b9,0,0x1f82,0x1f8a,0x1f03,0x3b9,0,0x1f83,0x1f8b,0x1f04,
+0x3b9,0,0x1f84,0x1f8c,0x1f05,0x3b9,0,0x1f85,0x1f8d,0x1f06,0x3b9,0,0x1f86,0x1f8e,0x1f07,0x3b9,
+0,0x1f87,0x1f8f,0x1f20,0x3b9,0,0x1f90,0x1f98,0x1f21,0x3b9,0,0x1f91,0x1f99,0x1f22,0x3b9,0,
+0x1f92,0x1f9a,0x1f23,0x3b9,0,0x1f93,0x1f9b,0x1f24,0x3b9,0,0x1f94,0x1f9c,0x1f25,0x3b9,0,0x1f95,
+0x1f9d,0x1f26,0x3b9,0,0x1f96,0x1f9e,0x1f27,0x3b9,0,0x1f97,0x1f9f,0x1f60,0x3b9,0,0x1fa0,0x1fa8,
+0x1f61,0x3b9,0,0x1fa1,0x1fa9,0x1f62,0x3b9,0,0x1fa2,0x1faa,0x1f63,0x3b9,0,0x1fa3,0x1fab,0x1f64,
+0x3b9,0,0x1fa4,0x1fac,0x1f65,0x3b9,0,0x1fa5,0x1fad,0x1f66,0x3b9,0,0x1fa6,0x1fae,0x1f67,0x3b9,
+0,0x1fa7,0x1faf,0x1f70,0x3b9,0,0x1fb2,0,0x1f74,0x3b9,0,0x1fc2,0,0x1f7c,0x3b9,0,
+0x1ff2,0
+};
+
+static const UCaseProps ucase_props_singleton={
+  NULL,
+  ucase_props_indexes,
+  ucase_props_exceptions,
+  ucase_props_unfold,
+  {
+    ucase_props_trieIndex,
+    ucase_props_trieIndex+3288,
+    NULL,
+    3288,
+    9068,
+    0x188,
+    0xd54,
+    0x0,
+    0x0,
+    0xe0800,
+    0x3040,
+    NULL, 0, FALSE, FALSE, 0, NULL
+  },
+  { 4,0,0,0 }
+};
+
+#endif  // INCLUDED_FROM_UCASE_CPP
diff --git a/thirdparty/icu4c/common/ucasemap.cpp b/thirdparty/icu4c/common/ucasemap.cpp
new file mode 100644
index 0000000000..ed72bda828
--- /dev/null
+++ b/thirdparty/icu4c/common/ucasemap.cpp
@@ -0,0 +1,953 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2005-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucasemap.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2005may06
+*   created by: Markus W. Scherer
+*
+*   Case mapping service object and functions using it.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/brkiter.h"
+#include "unicode/bytestream.h"
+#include "unicode/casemap.h"
+#include "unicode/edits.h"
+#include "unicode/stringoptions.h"
+#include "unicode/stringpiece.h"
+#include "unicode/ubrk.h"
+#include "unicode/uloc.h"
+#include "unicode/ustring.h"
+#include "unicode/ucasemap.h"
+#if !UCONFIG_NO_BREAK_ITERATION
+#include "unicode/utext.h"
+#endif
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "bytesinkutil.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "ucase.h"
+#include "ucasemap_imp.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_USE
+
+/* UCaseMap service object -------------------------------------------------- */
+
+UCaseMap::UCaseMap(const char *localeID, uint32_t opts, UErrorCode *pErrorCode) :
+#if !UCONFIG_NO_BREAK_ITERATION
+        iter(NULL),
+#endif
+        caseLocale(UCASE_LOC_UNKNOWN), options(opts) {
+    ucasemap_setLocale(this, localeID, pErrorCode);
+}
+
+UCaseMap::~UCaseMap() {
+#if !UCONFIG_NO_BREAK_ITERATION
+    delete iter;
+#endif
+}
+
+U_CAPI UCaseMap * U_EXPORT2
+ucasemap_open(const char *locale, uint32_t options, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    UCaseMap *csm = new UCaseMap(locale, options, pErrorCode);
+    if(csm==NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    } else if (U_FAILURE(*pErrorCode)) {
+        delete csm;
+        return NULL;
+    }
+    return csm;
+}
+
+U_CAPI void U_EXPORT2
+ucasemap_close(UCaseMap *csm) {
+    delete csm;
+}
+
+U_CAPI const char * U_EXPORT2
+ucasemap_getLocale(const UCaseMap *csm) {
+    return csm->locale;
+}
+
+U_CAPI uint32_t U_EXPORT2
+ucasemap_getOptions(const UCaseMap *csm) {
+    return csm->options;
+}
+
+U_CAPI void U_EXPORT2
+ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if (locale != NULL && *locale == 0) {
+        csm->locale[0] = 0;
+        csm->caseLocale = UCASE_LOC_ROOT;
+        return;
+    }
+
+    int32_t length=uloc_getName(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
+    if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || length==sizeof(csm->locale)) {
+        *pErrorCode=U_ZERO_ERROR;
+        /* we only really need the language code for case mappings */
+        length=uloc_getLanguage(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode);
+    }
+    if(length==sizeof(csm->locale)) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+    if(U_SUCCESS(*pErrorCode)) {
+        csm->caseLocale=UCASE_LOC_UNKNOWN;
+        csm->caseLocale = ucase_getCaseLocale(csm->locale);
+    } else {
+        csm->locale[0]=0;
+        csm->caseLocale = UCASE_LOC_ROOT;
+    }
+}
+
+U_CAPI void U_EXPORT2
+ucasemap_setOptions(UCaseMap *csm, uint32_t options, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    csm->options=options;
+}
+
+/* UTF-8 string case mappings ----------------------------------------------- */
+
+/* TODO(markus): Move to a new, separate utf8case.cpp file. */
+
+namespace {
+
+/* append a full case mapping result, see UCASE_MAX_STRING_LENGTH */
+inline UBool
+appendResult(int32_t cpLength, int32_t result, const UChar *s,
+             ByteSink &sink, uint32_t options, icu::Edits *edits, UErrorCode &errorCode) {
+    U_ASSERT(U_SUCCESS(errorCode));
+
+    /* decode the result */
+    if(result<0) {
+        /* (not) original code point */
+        if(edits!=NULL) {
+            edits->addUnchanged(cpLength);
+        }
+        if((options & U_OMIT_UNCHANGED_TEXT) == 0) {
+            ByteSinkUtil::appendCodePoint(cpLength, ~result, sink);
+        }
+    } else {
+        if(result<=UCASE_MAX_STRING_LENGTH) {
+            // string: "result" is the UTF-16 length
+            return ByteSinkUtil::appendChange(cpLength, s, result, sink, edits, errorCode);
+        } else {
+            ByteSinkUtil::appendCodePoint(cpLength, result, sink, edits);
+        }
+    }
+    return TRUE;
+}
+
+// See unicode/utf8.h U8_APPEND_UNSAFE().
+inline uint8_t getTwoByteLead(UChar32 c) { return (uint8_t)((c >> 6) | 0xc0); }
+inline uint8_t getTwoByteTrail(UChar32 c) { return (uint8_t)((c & 0x3f) | 0x80); }
+
+UChar32 U_CALLCONV
+utf8_caseContextIterator(void *context, int8_t dir) {
+    UCaseContext *csc=(UCaseContext *)context;
+    UChar32 c;
+
+    if(dir<0) {
+        /* reset for backward iteration */
+        csc->index=csc->cpStart;
+        csc->dir=dir;
+    } else if(dir>0) {
+        /* reset for forward iteration */
+        csc->index=csc->cpLimit;
+        csc->dir=dir;
+    } else {
+        /* continue current iteration direction */
+        dir=csc->dir;
+    }
+
+    if(dir<0) {
+        if(csc->start<csc->index) {
+            U8_PREV((const uint8_t *)csc->p, csc->start, csc->index, c);
+            return c;
+        }
+    } else {
+        if(csc->index<csc->limit) {
+            U8_NEXT((const uint8_t *)csc->p, csc->index, csc->limit, c);
+            return c;
+        }
+    }
+    return U_SENTINEL;
+}
+
+/**
+ * caseLocale >= 0: Lowercases [srcStart..srcLimit[ but takes context [0..srcLength[ into account.
+ * caseLocale < 0: Case-folds [srcStart..srcLimit[.
+ */
+void toLower(int32_t caseLocale, uint32_t options,
+             const uint8_t *src, UCaseContext *csc, int32_t srcStart, int32_t srcLimit,
+             icu::ByteSink &sink, icu::Edits *edits, UErrorCode &errorCode) {
+    const int8_t *latinToLower;
+    if (caseLocale == UCASE_LOC_ROOT ||
+            (caseLocale >= 0 ?
+                !(caseLocale == UCASE_LOC_TURKISH || caseLocale == UCASE_LOC_LITHUANIAN) :
+                (options & _FOLD_CASE_OPTIONS_MASK) == U_FOLD_CASE_DEFAULT)) {
+        latinToLower = LatinCase::TO_LOWER_NORMAL;
+    } else {
+        latinToLower = LatinCase::TO_LOWER_TR_LT;
+    }
+    const UTrie2 *trie = ucase_getTrie();
+    int32_t prev = srcStart;
+    int32_t srcIndex = srcStart;
+    for (;;) {
+        // fast path for simple cases
+        int32_t cpStart;
+        UChar32 c;
+        for (;;) {
+            if (U_FAILURE(errorCode) || srcIndex >= srcLimit) {
+                c = U_SENTINEL;
+                break;
+            }
+            uint8_t lead = src[srcIndex++];
+            if (lead <= 0x7f) {
+                int8_t d = latinToLower[lead];
+                if (d == LatinCase::EXC) {
+                    cpStart = srcIndex - 1;
+                    c = lead;
+                    break;
+                }
+                if (d == 0) { continue; }
+                ByteSinkUtil::appendUnchanged(src + prev, srcIndex - 1 - prev,
+                                              sink, options, edits, errorCode);
+                char ascii = (char)(lead + d);
+                sink.Append(&ascii, 1);
+                if (edits != nullptr) {
+                    edits->addReplace(1, 1);
+                }
+                prev = srcIndex;
+                continue;
+            } else if (lead < 0xe3) {
+                uint8_t t;
+                if (0xc2 <= lead && lead <= 0xc5 && srcIndex < srcLimit &&
+                        (t = src[srcIndex] - 0x80) <= 0x3f) {
+                    // U+0080..U+017F
+                    ++srcIndex;
+                    c = ((lead - 0xc0) << 6) | t;
+                    int8_t d = latinToLower[c];
+                    if (d == LatinCase::EXC) {
+                        cpStart = srcIndex - 2;
+                        break;
+                    }
+                    if (d == 0) { continue; }
+                    ByteSinkUtil::appendUnchanged(src + prev, srcIndex - 2 - prev,
+                                                  sink, options, edits, errorCode);
+                    ByteSinkUtil::appendTwoBytes(c + d, sink);
+                    if (edits != nullptr) {
+                        edits->addReplace(2, 2);
+                    }
+                    prev = srcIndex;
+                    continue;
+                }
+            } else if ((lead <= 0xe9 || lead == 0xeb || lead == 0xec) &&
+                    (srcIndex + 2) <= srcLimit &&
+                    U8_IS_TRAIL(src[srcIndex]) && U8_IS_TRAIL(src[srcIndex + 1])) {
+                // most of CJK: no case mappings
+                srcIndex += 2;
+                continue;
+            }
+            cpStart = --srcIndex;
+            U8_NEXT(src, srcIndex, srcLimit, c);
+            if (c < 0) {
+                // ill-formed UTF-8
+                continue;
+            }
+            uint16_t props = UTRIE2_GET16(trie, c);
+            if (UCASE_HAS_EXCEPTION(props)) { break; }
+            int32_t delta;
+            if (!UCASE_IS_UPPER_OR_TITLE(props) || (delta = UCASE_GET_DELTA(props)) == 0) {
+                continue;
+            }
+            ByteSinkUtil::appendUnchanged(src + prev, cpStart - prev,
+                                          sink, options, edits, errorCode);
+            ByteSinkUtil::appendCodePoint(srcIndex - cpStart, c + delta, sink, edits);
+            prev = srcIndex;
+        }
+        if (c < 0) {
+            break;
+        }
+        // slow path
+        const UChar *s;
+        if (caseLocale >= 0) {
+            csc->cpStart = cpStart;
+            csc->cpLimit = srcIndex;
+            c = ucase_toFullLower(c, utf8_caseContextIterator, csc, &s, caseLocale);
+        } else {
+            c = ucase_toFullFolding(c, &s, options);
+        }
+        if (c >= 0) {
+            ByteSinkUtil::appendUnchanged(src + prev, cpStart - prev,
+                                          sink, options, edits, errorCode);
+            appendResult(srcIndex - cpStart, c, s, sink, options, edits, errorCode);
+            prev = srcIndex;
+        }
+    }
+    ByteSinkUtil::appendUnchanged(src + prev, srcIndex - prev,
+                                  sink, options, edits, errorCode);
+}
+
+void toUpper(int32_t caseLocale, uint32_t options,
+             const uint8_t *src, UCaseContext *csc, int32_t srcLength,
+             icu::ByteSink &sink, icu::Edits *edits, UErrorCode &errorCode) {
+    const int8_t *latinToUpper;
+    if (caseLocale == UCASE_LOC_TURKISH) {
+        latinToUpper = LatinCase::TO_UPPER_TR;
+    } else {
+        latinToUpper = LatinCase::TO_UPPER_NORMAL;
+    }
+    const UTrie2 *trie = ucase_getTrie();
+    int32_t prev = 0;
+    int32_t srcIndex = 0;
+    for (;;) {
+        // fast path for simple cases
+        int32_t cpStart;
+        UChar32 c;
+        for (;;) {
+            if (U_FAILURE(errorCode) || srcIndex >= srcLength) {
+                c = U_SENTINEL;
+                break;
+            }
+            uint8_t lead = src[srcIndex++];
+            if (lead <= 0x7f) {
+                int8_t d = latinToUpper[lead];
+                if (d == LatinCase::EXC) {
+                    cpStart = srcIndex - 1;
+                    c = lead;
+                    break;
+                }
+                if (d == 0) { continue; }
+                ByteSinkUtil::appendUnchanged(src + prev, srcIndex - 1 - prev,
+                                              sink, options, edits, errorCode);
+                char ascii = (char)(lead + d);
+                sink.Append(&ascii, 1);
+                if (edits != nullptr) {
+                    edits->addReplace(1, 1);
+                }
+                prev = srcIndex;
+                continue;
+            } else if (lead < 0xe3) {
+                uint8_t t;
+                if (0xc2 <= lead && lead <= 0xc5 && srcIndex < srcLength &&
+                        (t = src[srcIndex] - 0x80) <= 0x3f) {
+                    // U+0080..U+017F
+                    ++srcIndex;
+                    c = ((lead - 0xc0) << 6) | t;
+                    int8_t d = latinToUpper[c];
+                    if (d == LatinCase::EXC) {
+                        cpStart = srcIndex - 2;
+                        break;
+                    }
+                    if (d == 0) { continue; }
+                    ByteSinkUtil::appendUnchanged(src + prev, srcIndex - 2 - prev,
+                                                  sink, options, edits, errorCode);
+                    ByteSinkUtil::appendTwoBytes(c + d, sink);
+                    if (edits != nullptr) {
+                        edits->addReplace(2, 2);
+                    }
+                    prev = srcIndex;
+                    continue;
+                }
+            } else if ((lead <= 0xe9 || lead == 0xeb || lead == 0xec) &&
+                    (srcIndex + 2) <= srcLength &&
+                    U8_IS_TRAIL(src[srcIndex]) && U8_IS_TRAIL(src[srcIndex + 1])) {
+                // most of CJK: no case mappings
+                srcIndex += 2;
+                continue;
+            }
+            cpStart = --srcIndex;
+            U8_NEXT(src, srcIndex, srcLength, c);
+            if (c < 0) {
+                // ill-formed UTF-8
+                continue;
+            }
+            uint16_t props = UTRIE2_GET16(trie, c);
+            if (UCASE_HAS_EXCEPTION(props)) { break; }
+            int32_t delta;
+            if (UCASE_GET_TYPE(props) != UCASE_LOWER || (delta = UCASE_GET_DELTA(props)) == 0) {
+                continue;
+            }
+            ByteSinkUtil::appendUnchanged(src + prev, cpStart - prev,
+                                          sink, options, edits, errorCode);
+            ByteSinkUtil::appendCodePoint(srcIndex - cpStart, c + delta, sink, edits);
+            prev = srcIndex;
+        }
+        if (c < 0) {
+            break;
+        }
+        // slow path
+        csc->cpStart = cpStart;
+        csc->cpLimit = srcIndex;
+        const UChar *s;
+        c = ucase_toFullUpper(c, utf8_caseContextIterator, csc, &s, caseLocale);
+        if (c >= 0) {
+            ByteSinkUtil::appendUnchanged(src + prev, cpStart - prev,
+                                          sink, options, edits, errorCode);
+            appendResult(srcIndex - cpStart, c, s, sink, options, edits, errorCode);
+            prev = srcIndex;
+        }
+    }
+    ByteSinkUtil::appendUnchanged(src + prev, srcIndex - prev,
+                                  sink, options, edits, errorCode);
+}
+
+}  // namespace
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+U_CFUNC void U_CALLCONV
+ucasemap_internalUTF8ToTitle(
+        int32_t caseLocale, uint32_t options, BreakIterator *iter,
+        const uint8_t *src, int32_t srcLength,
+        ByteSink &sink, icu::Edits *edits,
+        UErrorCode &errorCode) {
+    if (!ustrcase_checkTitleAdjustmentOptions(options, errorCode)) {
+        return;
+    }
+
+    /* set up local variables */
+    UCaseContext csc=UCASECONTEXT_INITIALIZER;
+    csc.p=(void *)src;
+    csc.limit=srcLength;
+    int32_t prev=0;
+    UBool isFirstIndex=TRUE;
+
+    /* titlecasing loop */
+    while(prev<srcLength) {
+        /* find next index where to titlecase */
+        int32_t index;
+        if(isFirstIndex) {
+            isFirstIndex=FALSE;
+            index=iter->first();
+        } else {
+            index=iter->next();
+        }
+        if(index==UBRK_DONE || index>srcLength) {
+            index=srcLength;
+        }
+
+        /*
+         * Segment [prev..index[ into 3 parts:
+         * a) skipped characters (copy as-is) [prev..titleStart[
+         * b) first letter (titlecase)              [titleStart..titleLimit[
+         * c) subsequent characters (lowercase)                 [titleLimit..index[
+         */
+        if(prev<index) {
+            /* find and copy skipped characters [prev..titleStart[ */
+            int32_t titleStart=prev;
+            int32_t titleLimit=prev;
+            UChar32 c;
+            U8_NEXT(src, titleLimit, index, c);
+            if ((options&U_TITLECASE_NO_BREAK_ADJUSTMENT)==0) {
+                // Adjust the titlecasing index to the next cased character,
+                // or to the next letter/number/symbol/private use.
+                // Stop with titleStart<titleLimit<=index
+                // if there is a character to be titlecased,
+                // or else stop with titleStart==titleLimit==index.
+                UBool toCased = (options&U_TITLECASE_ADJUST_TO_CASED) != 0;
+                while (toCased ? UCASE_NONE==ucase_getType(c) : !ustrcase_isLNS(c)) {
+                    titleStart=titleLimit;
+                    if(titleLimit==index) {
+                        break;
+                    }
+                    U8_NEXT(src, titleLimit, index, c);
+                }
+                if (prev < titleStart) {
+                    if (!ByteSinkUtil::appendUnchanged(src+prev, titleStart-prev,
+                                                       sink, options, edits, errorCode)) {
+                        return;
+                    }
+                }
+            }
+
+            if(titleStart<titleLimit) {
+                /* titlecase c which is from [titleStart..titleLimit[ */
+                if(c>=0) {
+                    csc.cpStart=titleStart;
+                    csc.cpLimit=titleLimit;
+                    const UChar *s;
+                    c=ucase_toFullTitle(c, utf8_caseContextIterator, &csc, &s, caseLocale);
+                    if (!appendResult(titleLimit-titleStart, c, s, sink, options, edits, errorCode)) {
+                        return;
+                    }
+                } else {
+                    // Malformed UTF-8.
+                    if (!ByteSinkUtil::appendUnchanged(src+titleStart, titleLimit-titleStart,
+                                                       sink, options, edits, errorCode)) {
+                        return;
+                    }
+                }
+
+                /* Special case Dutch IJ titlecasing */
+                if (titleStart+1 < index &&
+                        caseLocale == UCASE_LOC_DUTCH &&
+                        (src[titleStart] == 0x0049 || src[titleStart] == 0x0069)) {
+                    if (src[titleStart+1] == 0x006A) {
+                        ByteSinkUtil::appendCodePoint(1, 0x004A, sink, edits);
+                        titleLimit++;
+                    } else if (src[titleStart+1] == 0x004A) {
+                        // Keep the capital J from getting lowercased.
+                        if (!ByteSinkUtil::appendUnchanged(src+titleStart+1, 1,
+                                                           sink, options, edits, errorCode)) {
+                            return;
+                        }
+                        titleLimit++;
+                    }
+                }
+
+                /* lowercase [titleLimit..index[ */
+                if(titleLimit<index) {
+                    if((options&U_TITLECASE_NO_LOWERCASE)==0) {
+                        /* Normal operation: Lowercase the rest of the word. */
+                        toLower(caseLocale, options,
+                                src, &csc, titleLimit, index,
+                                sink, edits, errorCode);
+                        if(U_FAILURE(errorCode)) {
+                            return;
+                        }
+                    } else {
+                        /* Optionally just copy the rest of the word unchanged. */
+                        if (!ByteSinkUtil::appendUnchanged(src+titleLimit, index-titleLimit,
+                                                           sink, options, edits, errorCode)) {
+                            return;
+                        }
+                    }
+                }
+            }
+        }
+
+        prev=index;
+    }
+}
+
+#endif
+
+U_NAMESPACE_BEGIN
+namespace GreekUpper {
+
+UBool isFollowedByCasedLetter(const uint8_t *s, int32_t i, int32_t length) {
+    while (i < length) {
+        UChar32 c;
+        U8_NEXT(s, i, length, c);
+        int32_t type = ucase_getTypeOrIgnorable(c);
+        if ((type & UCASE_IGNORABLE) != 0) {
+            // Case-ignorable, continue with the loop.
+        } else if (type != UCASE_NONE) {
+            return TRUE;  // Followed by cased letter.
+        } else {
+            return FALSE;  // Uncased and not case-ignorable.
+        }
+    }
+    return FALSE;  // Not followed by cased letter.
+}
+
+// Keep this consistent with the UTF-16 version in ustrcase.cpp and the Java version in CaseMap.java.
+void toUpper(uint32_t options,
+             const uint8_t *src, int32_t srcLength,
+             ByteSink &sink, Edits *edits,
+             UErrorCode &errorCode) {
+    uint32_t state = 0;
+    for (int32_t i = 0; i < srcLength;) {
+        int32_t nextIndex = i;
+        UChar32 c;
+        U8_NEXT(src, nextIndex, srcLength, c);
+        uint32_t nextState = 0;
+        int32_t type = ucase_getTypeOrIgnorable(c);
+        if ((type & UCASE_IGNORABLE) != 0) {
+            // c is case-ignorable
+            nextState |= (state & AFTER_CASED);
+        } else if (type != UCASE_NONE) {
+            // c is cased
+            nextState |= AFTER_CASED;
+        }
+        uint32_t data = getLetterData(c);
+        if (data > 0) {
+            uint32_t upper = data & UPPER_MASK;
+            // Add a dialytika to this iota or ypsilon vowel
+            // if we removed a tonos from the previous vowel,
+            // and that previous vowel did not also have (or gain) a dialytika.
+            // Adding one only to the final vowel in a longer sequence
+            // (which does not occur in normal writing) would require lookahead.
+            // Set the same flag as for preserving an existing dialytika.
+            if ((data & HAS_VOWEL) != 0 && (state & AFTER_VOWEL_WITH_ACCENT) != 0 &&
+                    (upper == 0x399 || upper == 0x3A5)) {
+                data |= HAS_DIALYTIKA;
+            }
+            int32_t numYpogegrammeni = 0;  // Map each one to a trailing, spacing, capital iota.
+            if ((data & HAS_YPOGEGRAMMENI) != 0) {
+                numYpogegrammeni = 1;
+            }
+            // Skip combining diacritics after this Greek letter.
+            int32_t nextNextIndex = nextIndex;
+            while (nextIndex < srcLength) {
+                UChar32 c2;
+                U8_NEXT(src, nextNextIndex, srcLength, c2);
+                uint32_t diacriticData = getDiacriticData(c2);
+                if (diacriticData != 0) {
+                    data |= diacriticData;
+                    if ((diacriticData & HAS_YPOGEGRAMMENI) != 0) {
+                        ++numYpogegrammeni;
+                    }
+                    nextIndex = nextNextIndex;
+                } else {
+                    break;  // not a Greek diacritic
+                }
+            }
+            if ((data & HAS_VOWEL_AND_ACCENT_AND_DIALYTIKA) == HAS_VOWEL_AND_ACCENT) {
+                nextState |= AFTER_VOWEL_WITH_ACCENT;
+            }
+            // Map according to Greek rules.
+            UBool addTonos = FALSE;
+            if (upper == 0x397 &&
+                    (data & HAS_ACCENT) != 0 &&
+                    numYpogegrammeni == 0 &&
+                    (state & AFTER_CASED) == 0 &&
+                    !isFollowedByCasedLetter(src, nextIndex, srcLength)) {
+                // Keep disjunctive "or" with (only) a tonos.
+                // We use the same "word boundary" conditions as for the Final_Sigma test.
+                if (i == nextIndex) {
+                    upper = 0x389;  // Preserve the precomposed form.
+                } else {
+                    addTonos = TRUE;
+                }
+            } else if ((data & HAS_DIALYTIKA) != 0) {
+                // Preserve a vowel with dialytika in precomposed form if it exists.
+                if (upper == 0x399) {
+                    upper = 0x3AA;
+                    data &= ~HAS_EITHER_DIALYTIKA;
+                } else if (upper == 0x3A5) {
+                    upper = 0x3AB;
+                    data &= ~HAS_EITHER_DIALYTIKA;
+                }
+            }
+
+            UBool change;
+            if (edits == nullptr && (options & U_OMIT_UNCHANGED_TEXT) == 0) {
+                change = TRUE;  // common, simple usage
+            } else {
+                // Find out first whether we are changing the text.
+                U_ASSERT(0x370 <= upper && upper <= 0x3ff);  // 2-byte UTF-8, main Greek block
+                change = (i + 2) > nextIndex ||
+                        src[i] != getTwoByteLead(upper) || src[i + 1] != getTwoByteTrail(upper) ||
+                        numYpogegrammeni > 0;
+                int32_t i2 = i + 2;
+                if ((data & HAS_EITHER_DIALYTIKA) != 0) {
+                    change |= (i2 + 2) > nextIndex ||
+                            src[i2] != (uint8_t)u8"\u0308"[0] ||
+                            src[i2 + 1] != (uint8_t)u8"\u0308"[1];
+                    i2 += 2;
+                }
+                if (addTonos) {
+                    change |= (i2 + 2) > nextIndex ||
+                            src[i2] != (uint8_t)u8"\u0301"[0] ||
+                            src[i2 + 1] != (uint8_t)u8"\u0301"[1];
+                    i2 += 2;
+                }
+                int32_t oldLength = nextIndex - i;
+                int32_t newLength = (i2 - i) + numYpogegrammeni * 2;  // 2 bytes per U+0399
+                change |= oldLength != newLength;
+                if (change) {
+                    if (edits != NULL) {
+                        edits->addReplace(oldLength, newLength);
+                    }
+                } else {
+                    if (edits != NULL) {
+                        edits->addUnchanged(oldLength);
+                    }
+                    // Write unchanged text?
+                    change = (options & U_OMIT_UNCHANGED_TEXT) == 0;
+                }
+            }
+
+            if (change) {
+                ByteSinkUtil::appendTwoBytes(upper, sink);
+                if ((data & HAS_EITHER_DIALYTIKA) != 0) {
+                    sink.AppendU8(u8"\u0308", 2);  // restore or add a dialytika
+                }
+                if (addTonos) {
+                    sink.AppendU8(u8"\u0301", 2);
+                }
+                while (numYpogegrammeni > 0) {
+                    sink.AppendU8(u8"\u0399", 2);
+                    --numYpogegrammeni;
+                }
+            }
+        } else if(c>=0) {
+            const UChar *s;
+            c=ucase_toFullUpper(c, NULL, NULL, &s, UCASE_LOC_GREEK);
+            if (!appendResult(nextIndex - i, c, s, sink, options, edits, errorCode)) {
+                return;
+            }
+        } else {
+            // Malformed UTF-8.
+            if (!ByteSinkUtil::appendUnchanged(src+i, nextIndex-i,
+                                               sink, options, edits, errorCode)) {
+                return;
+            }
+        }
+        i = nextIndex;
+        state = nextState;
+    }
+}
+
+}  // namespace GreekUpper
+U_NAMESPACE_END
+
+static void U_CALLCONV
+ucasemap_internalUTF8ToLower(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                             const uint8_t *src, int32_t srcLength,
+                             icu::ByteSink &sink, icu::Edits *edits,
+                             UErrorCode &errorCode) {
+    UCaseContext csc=UCASECONTEXT_INITIALIZER;
+    csc.p=(void *)src;
+    csc.limit=srcLength;
+    toLower(
+        caseLocale, options,
+        src, &csc, 0, srcLength,
+        sink, edits, errorCode);
+}
+
+static void U_CALLCONV
+ucasemap_internalUTF8ToUpper(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                             const uint8_t *src, int32_t srcLength,
+                             icu::ByteSink &sink, icu::Edits *edits,
+                             UErrorCode &errorCode) {
+    if (caseLocale == UCASE_LOC_GREEK) {
+        GreekUpper::toUpper(options, src, srcLength, sink, edits, errorCode);
+    } else {
+        UCaseContext csc=UCASECONTEXT_INITIALIZER;
+        csc.p=(void *)src;
+        csc.limit=srcLength;
+        toUpper(
+            caseLocale, options,
+            src, &csc, srcLength,
+            sink, edits, errorCode);
+    }
+}
+
+static void U_CALLCONV
+ucasemap_internalUTF8Fold(int32_t /* caseLocale */, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                          const uint8_t *src, int32_t srcLength,
+                          icu::ByteSink &sink, icu::Edits *edits,
+                          UErrorCode &errorCode) {
+    toLower(
+        -1, options,
+        src, nullptr, 0, srcLength,
+        sink, edits, errorCode);
+}
+
+void
+ucasemap_mapUTF8(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                 const char *src, int32_t srcLength,
+                 UTF8CaseMapper *stringCaseMapper,
+                 icu::ByteSink &sink, icu::Edits *edits,
+                 UErrorCode &errorCode) {
+    /* check argument values */
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if ((src == nullptr && srcLength != 0) || srcLength < -1) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    // Get the string length.
+    if (srcLength == -1) {
+        srcLength = (int32_t)uprv_strlen((const char *)src);
+    }
+
+    if (edits != nullptr && (options & U_EDITS_NO_RESET) == 0) {
+        edits->reset();
+    }
+    stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                     (const uint8_t *)src, srcLength, sink, edits, errorCode);
+    sink.Flush();
+    if (U_SUCCESS(errorCode)) {
+        if (edits != nullptr) {
+            edits->copyErrorTo(errorCode);
+        }
+    }
+}
+
+int32_t
+ucasemap_mapUTF8(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                 char *dest, int32_t destCapacity,
+                 const char *src, int32_t srcLength,
+                 UTF8CaseMapper *stringCaseMapper,
+                 icu::Edits *edits,
+                 UErrorCode &errorCode) {
+    /* check argument values */
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if( destCapacity<0 ||
+        (dest==NULL && destCapacity>0) ||
+        (src==NULL && srcLength!=0) || srcLength<-1
+    ) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* get the string length */
+    if(srcLength==-1) {
+        srcLength=(int32_t)uprv_strlen((const char *)src);
+    }
+
+    /* check for overlapping source and destination */
+    if( dest!=NULL &&
+        ((src>=dest && src<(dest+destCapacity)) ||
+         (dest>=src && dest<(src+srcLength)))
+    ) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    CheckedArrayByteSink sink(dest, destCapacity);
+    if (edits != nullptr && (options & U_EDITS_NO_RESET) == 0) {
+        edits->reset();
+    }
+    stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                     (const uint8_t *)src, srcLength, sink, edits, errorCode);
+    sink.Flush();
+    if (U_SUCCESS(errorCode)) {
+        if (sink.Overflowed()) {
+            errorCode = U_BUFFER_OVERFLOW_ERROR;
+        } else if (edits != nullptr) {
+            edits->copyErrorTo(errorCode);
+        }
+    }
+    return u_terminateChars(dest, destCapacity, sink.NumberOfBytesAppended(), &errorCode);
+}
+
+/* public API functions */
+
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToLower(const UCaseMap *csm,
+                     char *dest, int32_t destCapacity,
+                     const char *src, int32_t srcLength,
+                     UErrorCode *pErrorCode) {
+    return ucasemap_mapUTF8(
+        csm->caseLocale, csm->options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8ToLower, NULL, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToUpper(const UCaseMap *csm,
+                     char *dest, int32_t destCapacity,
+                     const char *src, int32_t srcLength,
+                     UErrorCode *pErrorCode) {
+    return ucasemap_mapUTF8(
+        csm->caseLocale, csm->options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8ToUpper, NULL, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8FoldCase(const UCaseMap *csm,
+                      char *dest, int32_t destCapacity,
+                      const char *src, int32_t srcLength,
+                      UErrorCode *pErrorCode) {
+    return ucasemap_mapUTF8(
+        UCASE_LOC_ROOT, csm->options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8Fold, NULL, *pErrorCode);
+}
+
+U_NAMESPACE_BEGIN
+
+void CaseMap::utf8ToLower(
+        const char *locale, uint32_t options,
+        StringPiece src, ByteSink &sink, Edits *edits,
+        UErrorCode &errorCode) {
+    ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        src.data(), src.length(),
+        ucasemap_internalUTF8ToLower, sink, edits, errorCode);
+}
+
+void CaseMap::utf8ToUpper(
+        const char *locale, uint32_t options,
+        StringPiece src, ByteSink &sink, Edits *edits,
+        UErrorCode &errorCode) {
+    ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        src.data(), src.length(),
+        ucasemap_internalUTF8ToUpper, sink, edits, errorCode);
+}
+
+void CaseMap::utf8Fold(
+        uint32_t options,
+        StringPiece src, ByteSink &sink, Edits *edits,
+        UErrorCode &errorCode) {
+    ucasemap_mapUTF8(
+        UCASE_LOC_ROOT, options, UCASEMAP_BREAK_ITERATOR_NULL
+        src.data(), src.length(),
+        ucasemap_internalUTF8Fold, sink, edits, errorCode);
+}
+
+int32_t CaseMap::utf8ToLower(
+        const char *locale, uint32_t options,
+        const char *src, int32_t srcLength,
+        char *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8ToLower, edits, errorCode);
+}
+
+int32_t CaseMap::utf8ToUpper(
+        const char *locale, uint32_t options,
+        const char *src, int32_t srcLength,
+        char *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8ToUpper, edits, errorCode);
+}
+
+int32_t CaseMap::utf8Fold(
+        uint32_t options,
+        const char *src, int32_t srcLength,
+        char *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ucasemap_mapUTF8(
+        UCASE_LOC_ROOT, options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8Fold, edits, errorCode);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ucasemap_imp.h b/thirdparty/icu4c/common/ucasemap_imp.h
new file mode 100644
index 0000000000..e17a0ae5a3
--- /dev/null
+++ b/thirdparty/icu4c/common/ucasemap_imp.h
@@ -0,0 +1,282 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ucasemap_imp.h
+// created: 2017feb08 Markus W. Scherer
+
+#ifndef __UCASEMAP_IMP_H__
+#define __UCASEMAP_IMP_H__
+
+#include "unicode/utypes.h"
+#include "unicode/ucasemap.h"
+#include "unicode/uchar.h"
+#include "ucase.h"
+
+/**
+ * Bit mask for the titlecasing iterator options bit field.
+ * Currently only 3 out of 8 values are used:
+ * 0 (words), U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
+ * See stringoptions.h.
+ * @internal
+ */
+#define U_TITLECASE_ITERATOR_MASK 0xe0
+
+/**
+ * Bit mask for the titlecasing index adjustment options bit set.
+ * Currently two bits are defined:
+ * U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED.
+ * See stringoptions.h.
+ * @internal
+ */
+#define U_TITLECASE_ADJUSTMENT_MASK 0x600
+
+/**
+ * Internal API, used by u_strcasecmp() etc.
+ * Compare strings case-insensitively,
+ * in code point order or code unit order.
+ */
+U_CFUNC int32_t
+u_strcmpFold(const UChar *s1, int32_t length1,
+             const UChar *s2, int32_t length2,
+             uint32_t options,
+             UErrorCode *pErrorCode);
+
+/**
+ * Internal API, used for detecting length of
+ * shared prefix case-insensitively.
+ * @param s1            input string 1
+ * @param length1       length of string 1, or -1 (NULL terminated)
+ * @param s2            input string 2
+ * @param length2       length of string 2, or -1 (NULL terminated)
+ * @param options       compare options
+ * @param matchLen1     (output) length of partial prefix match in s1
+ * @param matchLen2     (output) length of partial prefix match in s2
+ * @param pErrorCode    receives error status
+ */
+U_CAPI void
+u_caseInsensitivePrefixMatch(const UChar *s1, int32_t length1,
+                             const UChar *s2, int32_t length2,
+                             uint32_t options,
+                             int32_t *matchLen1, int32_t *matchLen2,
+                             UErrorCode *pErrorCode);
+
+#ifdef __cplusplus
+
+U_NAMESPACE_BEGIN
+
+class BreakIterator;        // unicode/brkiter.h
+class ByteSink;
+class Locale;               // unicode/locid.h
+
+/** Returns true if the options are valid. Otherwise false, and sets an error. */
+inline UBool ustrcase_checkTitleAdjustmentOptions(uint32_t options, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return false; }
+    if ((options & U_TITLECASE_ADJUSTMENT_MASK) == U_TITLECASE_ADJUSTMENT_MASK) {
+        // Both options together.
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return false;
+    }
+    return true;
+}
+
+inline UBool ustrcase_isLNS(UChar32 c) {
+    // Letter, number, symbol,
+    // or a private use code point because those are typically used as letters or numbers.
+    // Consider modifier letters only if they are cased.
+    const uint32_t LNS = (U_GC_L_MASK|U_GC_N_MASK|U_GC_S_MASK|U_GC_CO_MASK) & ~U_GC_LM_MASK;
+    int gc = u_charType(c);
+    return (U_MASK(gc) & LNS) != 0 || (gc == U_MODIFIER_LETTER && ucase_getType(c) != UCASE_NONE);
+}
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/** Returns nullptr if error. Pass in either locale or locID, not both. */
+U_CFUNC
+BreakIterator *ustrcase_getTitleBreakIterator(
+        const Locale *locale, const char *locID, uint32_t options, BreakIterator *iter,
+        LocalPointer<BreakIterator> &ownedIter, UErrorCode &errorCode);
+
+#endif
+
+U_NAMESPACE_END
+
+#include "unicode/unistr.h"  // for UStringCaseMapper
+
+/*
+ * Internal string casing functions implementing
+ * ustring.h/ustrcase.cpp and UnicodeString case mapping functions.
+ */
+
+struct UCaseMap : public icu::UMemory {
+    /** Implements most of ucasemap_open(). */
+    UCaseMap(const char *localeID, uint32_t opts, UErrorCode *pErrorCode);
+    ~UCaseMap();
+
+#if !UCONFIG_NO_BREAK_ITERATION
+    icu::BreakIterator *iter;  /* We adopt the iterator, so we own it. */
+#endif
+    char locale[32];
+    int32_t caseLocale;
+    uint32_t options;
+};
+
+#if UCONFIG_NO_BREAK_ITERATION
+#   define UCASEMAP_BREAK_ITERATOR_PARAM
+#   define UCASEMAP_BREAK_ITERATOR_UNUSED
+#   define UCASEMAP_BREAK_ITERATOR
+#   define UCASEMAP_BREAK_ITERATOR_NULL
+#else
+#   define UCASEMAP_BREAK_ITERATOR_PARAM icu::BreakIterator *iter,
+#   define UCASEMAP_BREAK_ITERATOR_UNUSED icu::BreakIterator *,
+#   define UCASEMAP_BREAK_ITERATOR iter,
+#   define UCASEMAP_BREAK_ITERATOR_NULL NULL,
+#endif
+
+U_CFUNC int32_t
+ustrcase_getCaseLocale(const char *locale);
+
+// TODO: swap src / dest if approved for new public api
+/** Implements UStringCaseMapper. */
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToLower(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode);
+
+/** Implements UStringCaseMapper. */
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToUpper(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/** Implements UStringCaseMapper. */
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToTitle(int32_t caseLocale, uint32_t options,
+                         icu::BreakIterator *iter,
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode);
+
+#endif
+
+/** Implements UStringCaseMapper. */
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalFold(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                      UChar *dest, int32_t destCapacity,
+                      const UChar *src, int32_t srcLength,
+                      icu::Edits *edits,
+                      UErrorCode &errorCode);
+
+/**
+ * Common string case mapping implementation for ucasemap_toXyz() and UnicodeString::toXyz().
+ * Implements argument checking.
+ */
+U_CFUNC int32_t
+ustrcase_map(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+             UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             UStringCaseMapper *stringCaseMapper,
+             icu::Edits *edits,
+             UErrorCode &errorCode);
+
+/**
+ * Common string case mapping implementation for old-fashioned u_strToXyz() functions
+ * that allow the source string to overlap the destination buffer.
+ * Implements argument checking and internally works with an intermediate buffer if necessary.
+ */
+U_CFUNC int32_t
+ustrcase_mapWithOverlap(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                        UChar *dest, int32_t destCapacity,
+                        const UChar *src, int32_t srcLength,
+                        UStringCaseMapper *stringCaseMapper,
+                        UErrorCode &errorCode);
+
+/**
+ * UTF-8 string case mapping function type, used by ucasemap_mapUTF8().
+ * UTF-8 version of UStringCaseMapper.
+ * All error checking must be done.
+ * The UCaseMap must be fully initialized, with locale and/or iter set as needed.
+ */
+typedef void U_CALLCONV
+UTF8CaseMapper(int32_t caseLocale, uint32_t options,
+#if !UCONFIG_NO_BREAK_ITERATION
+               icu::BreakIterator *iter,
+#endif
+               const uint8_t *src, int32_t srcLength,
+               icu::ByteSink &sink, icu::Edits *edits,
+               UErrorCode &errorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/** Implements UTF8CaseMapper. */
+U_CFUNC void U_CALLCONV
+ucasemap_internalUTF8ToTitle(int32_t caseLocale, uint32_t options,
+        icu::BreakIterator *iter,
+        const uint8_t *src, int32_t srcLength,
+        icu::ByteSink &sink, icu::Edits *edits,
+        UErrorCode &errorCode);
+
+#endif
+
+void
+ucasemap_mapUTF8(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                 const char *src, int32_t srcLength,
+                 UTF8CaseMapper *stringCaseMapper,
+                 icu::ByteSink &sink, icu::Edits *edits,
+                 UErrorCode &errorCode);
+
+/**
+ * Implements argument checking and buffer handling
+ * for UTF-8 string case mapping as a common function.
+ */
+int32_t
+ucasemap_mapUTF8(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                 char *dest, int32_t destCapacity,
+                 const char *src, int32_t srcLength,
+                 UTF8CaseMapper *stringCaseMapper,
+                 icu::Edits *edits,
+                 UErrorCode &errorCode);
+
+U_NAMESPACE_BEGIN
+namespace GreekUpper {
+
+// Data bits.
+static const uint32_t UPPER_MASK = 0x3ff;
+static const uint32_t HAS_VOWEL = 0x1000;
+static const uint32_t HAS_YPOGEGRAMMENI = 0x2000;
+static const uint32_t HAS_ACCENT = 0x4000;
+static const uint32_t HAS_DIALYTIKA = 0x8000;
+// Further bits during data building and processing, not stored in the data map.
+static const uint32_t HAS_COMBINING_DIALYTIKA = 0x10000;
+static const uint32_t HAS_OTHER_GREEK_DIACRITIC = 0x20000;
+
+static const uint32_t HAS_VOWEL_AND_ACCENT = HAS_VOWEL | HAS_ACCENT;
+static const uint32_t HAS_VOWEL_AND_ACCENT_AND_DIALYTIKA =
+        HAS_VOWEL_AND_ACCENT | HAS_DIALYTIKA;
+static const uint32_t HAS_EITHER_DIALYTIKA = HAS_DIALYTIKA | HAS_COMBINING_DIALYTIKA;
+
+// State bits.
+static const uint32_t AFTER_CASED = 1;
+static const uint32_t AFTER_VOWEL_WITH_ACCENT = 2;
+
+uint32_t getLetterData(UChar32 c);
+
+/**
+ * Returns a non-zero value for each of the Greek combining diacritics
+ * listed in The Unicode Standard, version 8, chapter 7.2 Greek,
+ * plus some perispomeni look-alikes.
+ */
+uint32_t getDiacriticData(UChar32 c);
+
+}  // namespace GreekUpper
+U_NAMESPACE_END
+
+#endif  // __cplusplus
+
+#endif  // __UCASEMAP_IMP_H__
diff --git a/thirdparty/icu4c/common/ucasemap_titlecase_brkiter.cpp b/thirdparty/icu4c/common/ucasemap_titlecase_brkiter.cpp
new file mode 100644
index 0000000000..c21dfb7698
--- /dev/null
+++ b/thirdparty/icu4c/common/ucasemap_titlecase_brkiter.cpp
@@ -0,0 +1,134 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucasemap_titlecase_brkiter.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011jun02
+*   created by: Markus W. Scherer
+*
+*   Titlecasing functions that are based on BreakIterator
+*   were moved here to break dependency cycles among parts of the common library.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/brkiter.h"
+#include "unicode/ubrk.h"
+#include "unicode/casemap.h"
+#include "unicode/ucasemap.h"
+#include "cmemory.h"
+#include "ucase.h"
+#include "ucasemap_imp.h"
+
+U_NAMESPACE_BEGIN
+
+void CaseMap::utf8ToTitle(
+        const char *locale, uint32_t options, BreakIterator *iter,
+        StringPiece src, ByteSink &sink, Edits *edits,
+        UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    UText utext = UTEXT_INITIALIZER;
+    utext_openUTF8(&utext, src.data(), src.length(), &errorCode);
+    LocalPointer<BreakIterator> ownedIter;
+    iter = ustrcase_getTitleBreakIterator(nullptr, locale, options, iter, ownedIter, errorCode);
+    if (iter == nullptr) {
+        utext_close(&utext);
+        return;
+    }
+    iter->setText(&utext, errorCode);
+    ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, iter,
+        src.data(), src.length(),
+        ucasemap_internalUTF8ToTitle, sink, edits, errorCode);
+    utext_close(&utext);
+}
+
+int32_t CaseMap::utf8ToTitle(
+        const char *locale, uint32_t options, BreakIterator *iter,
+        const char *src, int32_t srcLength,
+        char *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return 0;
+    }
+    UText utext=UTEXT_INITIALIZER;
+    utext_openUTF8(&utext, src, srcLength, &errorCode);
+    LocalPointer<BreakIterator> ownedIter;
+    iter = ustrcase_getTitleBreakIterator(nullptr, locale, options, iter, ownedIter, errorCode);
+    if(iter==NULL) {
+        utext_close(&utext);
+        return 0;
+    }
+    iter->setText(&utext, errorCode);
+    int32_t length=ucasemap_mapUTF8(
+        ustrcase_getCaseLocale(locale), options, iter,
+        dest, destCapacity,
+        src, srcLength,
+        ucasemap_internalUTF8ToTitle, edits, errorCode);
+    utext_close(&utext);
+    return length;
+}
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI const UBreakIterator * U_EXPORT2
+ucasemap_getBreakIterator(const UCaseMap *csm) {
+    return reinterpret_cast<UBreakIterator *>(csm->iter);
+}
+
+U_CAPI void U_EXPORT2
+ucasemap_setBreakIterator(UCaseMap *csm, UBreakIterator *iterToAdopt, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    delete csm->iter;
+    csm->iter=reinterpret_cast<BreakIterator *>(iterToAdopt);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToTitle(UCaseMap *csm,
+                     char *dest, int32_t destCapacity,
+                     const char *src, int32_t srcLength,
+                     UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    UText utext=UTEXT_INITIALIZER;
+    utext_openUTF8(&utext, (const char *)src, srcLength, pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(csm->iter==NULL) {
+        LocalPointer<BreakIterator> ownedIter;
+        BreakIterator *iter = ustrcase_getTitleBreakIterator(
+            nullptr, csm->locale, csm->options, nullptr, ownedIter, *pErrorCode);
+        if (iter == nullptr) {
+            utext_close(&utext);
+            return 0;
+        }
+        csm->iter = ownedIter.orphan();
+    }
+    csm->iter->setText(&utext, *pErrorCode);
+    int32_t length=ucasemap_mapUTF8(
+            csm->caseLocale, csm->options, csm->iter,
+            dest, destCapacity,
+            src, srcLength,
+            ucasemap_internalUTF8ToTitle, NULL, *pErrorCode);
+    utext_close(&utext);
+    return length;
+}
+
+#endif  // !UCONFIG_NO_BREAK_ITERATION
diff --git a/thirdparty/icu4c/common/ucat.cpp b/thirdparty/icu4c/common/ucat.cpp
new file mode 100644
index 0000000000..dac56eeb5c
--- /dev/null
+++ b/thirdparty/icu4c/common/ucat.cpp
@@ -0,0 +1,78 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2003, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: March 19 2003
+* Since: ICU 2.6
+**********************************************************************
+*/
+#include "unicode/ucat.h"
+#include "unicode/ustring.h"
+#include "cstring.h"
+#include "uassert.h"
+
+/* Separator between set_num and msg_num */
+static const char SEPARATOR = '%';
+
+/* Maximum length of a set_num/msg_num key, incl. terminating zero.
+ * Longest possible key is "-2147483648%-2147483648" */
+#define MAX_KEY_LEN (24)
+
+/**
+ * Fill in buffer with a set_num/msg_num key string, given the numeric
+ * values. Numeric values must be >= 0. Buffer must be of length
+ * MAX_KEY_LEN or more.
+ */
+static char*
+_catkey(char* buffer, int32_t set_num, int32_t msg_num) {
+    int32_t i = 0;
+    i = T_CString_integerToString(buffer, set_num, 10);
+    buffer[i++] = SEPARATOR;
+    T_CString_integerToString(buffer+i, msg_num, 10);
+    return buffer;
+}
+
+U_CAPI u_nl_catd U_EXPORT2
+u_catopen(const char* name, const char* locale, UErrorCode* ec) {
+    return (u_nl_catd) ures_open(name, locale, ec);
+}
+
+U_CAPI void U_EXPORT2
+u_catclose(u_nl_catd catd) {
+    ures_close((UResourceBundle*) catd); /* may be NULL */
+}
+
+U_CAPI const UChar* U_EXPORT2
+u_catgets(u_nl_catd catd, int32_t set_num, int32_t msg_num,
+          const UChar* s,
+          int32_t* len, UErrorCode* ec) {
+
+    char key[MAX_KEY_LEN];
+    const UChar* result;
+
+    if (ec == NULL || U_FAILURE(*ec)) {
+        goto ERROR;
+    }
+
+    result = ures_getStringByKey((const UResourceBundle*) catd,
+                                 _catkey(key, set_num, msg_num),
+                                 len, ec);
+    if (U_FAILURE(*ec)) {
+        goto ERROR;
+    }
+
+    return result;
+
+ ERROR:
+    /* In case of any failure, return s */
+    if (len != NULL) {
+        *len = u_strlen(s);
+    }
+    return s;
+}
+
+/*eof*/
diff --git a/thirdparty/icu4c/common/uchar.cpp b/thirdparty/icu4c/common/uchar.cpp
new file mode 100644
index 0000000000..eb14e4c75d
--- /dev/null
+++ b/thirdparty/icu4c/common/uchar.cpp
@@ -0,0 +1,730 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 1996-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*
+* File UCHAR.C
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/02/97    aliu        Creation.
+*   4/15/99     Madhu       Updated all the function definitions for C Implementation
+*   5/20/99     Madhu       Added the function u_getVersion()
+*   8/19/1999   srl         Upgraded scripts to Unicode3.0 
+*   11/11/1999  weiv        added u_isalnum(), cleaned comments
+*   01/11/2000  helena      Renamed u_getVersion to u_getUnicodeVersion.
+*   06/20/2000  helena      OS/400 port changes; mostly typecast.
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+#include "unicode/uscript.h"
+#include "unicode/udata.h"
+#include "uassert.h"
+#include "cmemory.h"
+#include "ucln_cmn.h"
+#include "utrie2.h"
+#include "udataswp.h"
+#include "uprops.h"
+#include "ustr_imp.h"
+
+/* uchar_props_data.h is machine-generated by genprops --csource */
+#define INCLUDED_FROM_UCHAR_C
+#include "uchar_props_data.h"
+
+/* constants and macros for access to the data ------------------------------ */
+
+/* getting a uint32_t properties word from the data */
+#define GET_PROPS(c, result) ((result)=UTRIE2_GET16(&propsTrie, c))
+
+/* API functions ------------------------------------------------------------ */
+
+/* Gets the Unicode character's general category.*/
+U_CAPI int8_t U_EXPORT2
+u_charType(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (int8_t)GET_CATEGORY(props);
+}
+
+/* Enumerate all code points with their general categories. */
+struct _EnumTypeCallback {
+    UCharEnumTypeRange *enumRange;
+    const void *context;
+};
+
+static uint32_t U_CALLCONV
+_enumTypeValue(const void *context, uint32_t value) {
+    (void)context;
+    return GET_CATEGORY(value);
+}
+
+static UBool U_CALLCONV
+_enumTypeRange(const void *context, UChar32 start, UChar32 end, uint32_t value) {
+    /* just cast the value to UCharCategory */
+    return ((struct _EnumTypeCallback *)context)->
+        enumRange(((struct _EnumTypeCallback *)context)->context,
+                  start, end+1, (UCharCategory)value);
+}
+
+U_CAPI void U_EXPORT2
+u_enumCharTypes(UCharEnumTypeRange *enumRange, const void *context) {
+    struct _EnumTypeCallback callback;
+
+    if(enumRange==NULL) {
+        return;
+    }
+
+    callback.enumRange=enumRange;
+    callback.context=context;
+    utrie2_enum(&propsTrie, _enumTypeValue, _enumTypeRange, &callback);
+}
+
+/* Checks if ch is a lower case letter.*/
+U_CAPI UBool U_EXPORT2
+u_islower(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)==U_LOWERCASE_LETTER);
+}
+
+/* Checks if ch is an upper case letter.*/
+U_CAPI UBool U_EXPORT2
+u_isupper(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)==U_UPPERCASE_LETTER);
+}
+
+/* Checks if ch is a title case letter; usually upper case letters.*/
+U_CAPI UBool U_EXPORT2
+u_istitle(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)==U_TITLECASE_LETTER);
+}
+
+/* Checks if ch is a decimal digit. */
+U_CAPI UBool U_EXPORT2
+u_isdigit(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)==U_DECIMAL_DIGIT_NUMBER);
+}
+
+U_CAPI UBool U_EXPORT2
+u_isxdigit(UChar32 c) {
+    uint32_t props;
+
+    /* check ASCII and Fullwidth ASCII a-fA-F */
+    if(
+        (c<=0x66 && c>=0x41 && (c<=0x46 || c>=0x61)) ||
+        (c>=0xff21 && c<=0xff46 && (c<=0xff26 || c>=0xff41))
+    ) {
+        return TRUE;
+    }
+
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)==U_DECIMAL_DIGIT_NUMBER);
+}
+
+/* Checks if the Unicode character is a letter.*/
+U_CAPI UBool U_EXPORT2
+u_isalpha(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&U_GC_L_MASK)!=0);
+}
+
+U_CAPI UBool U_EXPORT2
+u_isUAlphabetic(UChar32 c) {
+    return (u_getUnicodeProperties(c, 1)&U_MASK(UPROPS_ALPHABETIC))!=0;
+}
+
+/* Checks if c is a letter or a decimal digit */
+U_CAPI UBool U_EXPORT2
+u_isalnum(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&(U_GC_L_MASK|U_GC_ND_MASK))!=0);
+}
+
+/**
+ * Checks if c is alphabetic, or a decimal digit; implements UCHAR_POSIX_ALNUM.
+ * @internal
+ */
+U_CFUNC UBool
+u_isalnumPOSIX(UChar32 c) {
+    return (UBool)(u_isUAlphabetic(c) || u_isdigit(c));
+}
+
+/* Checks if ch is a unicode character with assigned character type.*/
+U_CAPI UBool U_EXPORT2
+u_isdefined(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(GET_CATEGORY(props)!=0);
+}
+
+/* Checks if the Unicode character is a base form character that can take a diacritic.*/
+U_CAPI UBool U_EXPORT2
+u_isbase(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&(U_GC_L_MASK|U_GC_N_MASK|U_GC_MC_MASK|U_GC_ME_MASK))!=0);
+}
+
+/* Checks if the Unicode character is a control character.*/
+U_CAPI UBool U_EXPORT2
+u_iscntrl(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&(U_GC_CC_MASK|U_GC_CF_MASK|U_GC_ZL_MASK|U_GC_ZP_MASK))!=0);
+}
+
+U_CAPI UBool U_EXPORT2
+u_isISOControl(UChar32 c) {
+    return (uint32_t)c<=0x9f && (c<=0x1f || c>=0x7f);
+}
+
+/* Some control characters that are used as space. */
+#define IS_THAT_CONTROL_SPACE(c) \
+    (c<=0x9f && ((c>=TAB && c<=CR) || (c>=0x1c && c <=0x1f) || c==NL))
+
+/* Java has decided that U+0085 New Line is not whitespace any more. */
+#define IS_THAT_ASCII_CONTROL_SPACE(c) \
+    (c<=0x1f && c>=TAB && (c<=CR || c>=0x1c))
+
+/* Checks if the Unicode character is a space character.*/
+U_CAPI UBool U_EXPORT2
+u_isspace(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&U_GC_Z_MASK)!=0 || IS_THAT_CONTROL_SPACE(c));
+}
+
+U_CAPI UBool U_EXPORT2
+u_isJavaSpaceChar(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&U_GC_Z_MASK)!=0);
+}
+
+/* Checks if the Unicode character is a whitespace character.*/
+U_CAPI UBool U_EXPORT2
+u_isWhitespace(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(
+                ((CAT_MASK(props)&U_GC_Z_MASK)!=0 &&
+                    c!=NBSP && c!=FIGURESP && c!=NNBSP) || /* exclude no-break spaces */
+                IS_THAT_ASCII_CONTROL_SPACE(c)
+           );
+}
+
+U_CAPI UBool U_EXPORT2
+u_isblank(UChar32 c) {
+    if((uint32_t)c<=0x9f) {
+        return c==9 || c==0x20; /* TAB or SPACE */
+    } else {
+        /* Zs */
+        uint32_t props;
+        GET_PROPS(c, props);
+        return (UBool)(GET_CATEGORY(props)==U_SPACE_SEPARATOR);
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+u_isUWhiteSpace(UChar32 c) {
+    return (u_getUnicodeProperties(c, 1)&U_MASK(UPROPS_WHITE_SPACE))!=0;
+}
+
+/* Checks if the Unicode character is printable.*/
+U_CAPI UBool U_EXPORT2
+u_isprint(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    /* comparing ==0 returns FALSE for the categories mentioned */
+    return (UBool)((CAT_MASK(props)&U_GC_C_MASK)==0);
+}
+
+/**
+ * Checks if c is in \p{graph}\p{blank} - \p{cntrl}.
+ * Implements UCHAR_POSIX_PRINT.
+ * @internal
+ */
+U_CFUNC UBool
+u_isprintPOSIX(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    /*
+     * The only cntrl character in graph+blank is TAB (in blank).
+     * Here we implement (blank-TAB)=Zs instead of calling u_isblank().
+     */
+    return (UBool)((GET_CATEGORY(props)==U_SPACE_SEPARATOR) || u_isgraphPOSIX(c));
+}
+
+U_CAPI UBool U_EXPORT2
+u_isgraph(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    /* comparing ==0 returns FALSE for the categories mentioned */
+    return (UBool)((CAT_MASK(props)&
+                    (U_GC_CC_MASK|U_GC_CF_MASK|U_GC_CS_MASK|U_GC_CN_MASK|U_GC_Z_MASK))
+                   ==0);
+}
+
+/**
+ * Checks if c is in
+ * [^\p{space}\p{gc=Control}\p{gc=Surrogate}\p{gc=Unassigned}]
+ * with space=\p{Whitespace} and Control=Cc.
+ * Implements UCHAR_POSIX_GRAPH.
+ * @internal
+ */
+U_CFUNC UBool
+u_isgraphPOSIX(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    /* \p{space}\p{gc=Control} == \p{gc=Z}\p{Control} */
+    /* comparing ==0 returns FALSE for the categories mentioned */
+    return (UBool)((CAT_MASK(props)&
+                    (U_GC_CC_MASK|U_GC_CS_MASK|U_GC_CN_MASK|U_GC_Z_MASK))
+                   ==0);
+}
+
+U_CAPI UBool U_EXPORT2
+u_ispunct(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&U_GC_P_MASK)!=0);
+}
+
+/* Checks if the Unicode character can start a Unicode identifier.*/
+U_CAPI UBool U_EXPORT2
+u_isIDStart(UChar32 c) {
+    /* same as u_isalpha() */
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&(U_GC_L_MASK|U_GC_NL_MASK))!=0);
+}
+
+/* Checks if the Unicode character can be a Unicode identifier part other than starting the
+ identifier.*/
+U_CAPI UBool U_EXPORT2
+u_isIDPart(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(
+           (CAT_MASK(props)&
+            (U_GC_ND_MASK|U_GC_NL_MASK|
+             U_GC_L_MASK|
+             U_GC_PC_MASK|U_GC_MC_MASK|U_GC_MN_MASK)
+           )!=0 ||
+           u_isIDIgnorable(c));
+}
+
+/*Checks if the Unicode character can be ignorable in a Java or Unicode identifier.*/
+U_CAPI UBool U_EXPORT2
+u_isIDIgnorable(UChar32 c) {
+    if(c<=0x9f) {
+        return u_isISOControl(c) && !IS_THAT_ASCII_CONTROL_SPACE(c);
+    } else {
+        uint32_t props;
+        GET_PROPS(c, props);
+        return (UBool)(GET_CATEGORY(props)==U_FORMAT_CHAR);
+    }
+}
+
+/*Checks if the Unicode character can start a Java identifier.*/
+U_CAPI UBool U_EXPORT2
+u_isJavaIDStart(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)((CAT_MASK(props)&(U_GC_L_MASK|U_GC_SC_MASK|U_GC_PC_MASK))!=0);
+}
+
+/*Checks if the Unicode character can be a Java identifier part other than starting the
+ * identifier.
+ */
+U_CAPI UBool U_EXPORT2
+u_isJavaIDPart(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return (UBool)(
+           (CAT_MASK(props)&
+            (U_GC_ND_MASK|U_GC_NL_MASK|
+             U_GC_L_MASK|
+             U_GC_SC_MASK|U_GC_PC_MASK|
+             U_GC_MC_MASK|U_GC_MN_MASK)
+           )!=0 ||
+           u_isIDIgnorable(c));
+}
+
+U_CAPI int32_t U_EXPORT2
+u_charDigitValue(UChar32 c) {
+    uint32_t props;
+    int32_t value;
+    GET_PROPS(c, props);
+    value=(int32_t)GET_NUMERIC_TYPE_VALUE(props)-UPROPS_NTV_DECIMAL_START;
+    if(value<=9) {
+        return value;
+    } else {
+        return -1;
+    }
+}
+
+U_CAPI double U_EXPORT2
+u_getNumericValue(UChar32 c) {
+    uint32_t props;
+    int32_t ntv;
+    GET_PROPS(c, props);
+    ntv=(int32_t)GET_NUMERIC_TYPE_VALUE(props);
+
+    if(ntv==UPROPS_NTV_NONE) {
+        return U_NO_NUMERIC_VALUE;
+    } else if(ntv<UPROPS_NTV_DIGIT_START) {
+        /* decimal digit */
+        return ntv-UPROPS_NTV_DECIMAL_START;
+    } else if(ntv<UPROPS_NTV_NUMERIC_START) {
+        /* other digit */
+        return ntv-UPROPS_NTV_DIGIT_START;
+    } else if(ntv<UPROPS_NTV_FRACTION_START) {
+        /* small integer */
+        return ntv-UPROPS_NTV_NUMERIC_START;
+    } else if(ntv<UPROPS_NTV_LARGE_START) {
+        /* fraction */
+        int32_t numerator=(ntv>>4)-12;
+        int32_t denominator=(ntv&0xf)+1;
+        return (double)numerator/denominator;
+    } else if(ntv<UPROPS_NTV_BASE60_START) {
+        /* large, single-significant-digit integer */
+        double numValue;
+        int32_t mant=(ntv>>5)-14;
+        int32_t exp=(ntv&0x1f)+2;
+        numValue=mant;
+
+        /* multiply by 10^exp without math.h */
+        while(exp>=4) {
+            numValue*=10000.;
+            exp-=4;
+        }
+        switch(exp) {
+        case 3:
+            numValue*=1000.;
+            break;
+        case 2:
+            numValue*=100.;
+            break;
+        case 1:
+            numValue*=10.;
+            break;
+        case 0:
+        default:
+            break;
+        }
+
+        return numValue;
+    } else if(ntv<UPROPS_NTV_FRACTION20_START) {
+        /* sexagesimal (base 60) integer */
+        int32_t numValue=(ntv>>2)-0xbf;
+        int32_t exp=(ntv&3)+1;
+
+        switch(exp) {
+        case 4:
+            numValue*=60*60*60*60;
+            break;
+        case 3:
+            numValue*=60*60*60;
+            break;
+        case 2:
+            numValue*=60*60;
+            break;
+        case 1:
+            numValue*=60;
+            break;
+        case 0:
+        default:
+            break;
+        }
+
+        return numValue;
+    } else if(ntv<UPROPS_NTV_FRACTION32_START) {
+        // fraction-20 e.g. 3/80
+        int32_t frac20=ntv-UPROPS_NTV_FRACTION20_START;  // 0..0x17
+        int32_t numerator=2*(frac20&3)+1;
+        int32_t denominator=20<<(frac20>>2);
+        return (double)numerator/denominator;
+    } else if(ntv<UPROPS_NTV_RESERVED_START) {
+        // fraction-32 e.g. 3/64
+        int32_t frac32=ntv-UPROPS_NTV_FRACTION32_START;  // 0..15
+        int32_t numerator=2*(frac32&3)+1;
+        int32_t denominator=32<<(frac32>>2);
+        return (double)numerator/denominator;
+    } else {
+        /* reserved */
+        return U_NO_NUMERIC_VALUE;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+u_digit(UChar32 ch, int8_t radix) {
+    int8_t value;
+    if((uint8_t)(radix-2)<=(36-2)) {
+        value=(int8_t)u_charDigitValue(ch);
+        if(value<0) {
+            /* ch is not a decimal digit, try latin letters */
+            if(ch>=0x61 && ch<=0x7A) {
+                value=(int8_t)(ch-0x57);  /* ch - 'a' + 10 */
+            } else if(ch>=0x41 && ch<=0x5A) {
+                value=(int8_t)(ch-0x37);  /* ch - 'A' + 10 */
+            } else if(ch>=0xFF41 && ch<=0xFF5A) {
+                value=(int8_t)(ch-0xFF37);  /* fullwidth ASCII a-z */
+            } else if(ch>=0xFF21 && ch<=0xFF3A) {
+                value=(int8_t)(ch-0xFF17);  /* fullwidth ASCII A-Z */
+            }
+        }
+    } else {
+        value=-1;   /* invalid radix */
+    }
+    return (int8_t)((value<radix) ? value : -1);
+}
+
+U_CAPI UChar32 U_EXPORT2
+u_forDigit(int32_t digit, int8_t radix) {
+    if((uint8_t)(radix-2)>(36-2) || (uint32_t)digit>=(uint32_t)radix) {
+        return 0;
+    } else if(digit<10) {
+        return (UChar32)(0x30+digit);
+    } else {
+        return (UChar32)((0x61-10)+digit);
+    }
+}
+
+/* miscellaneous, and support for uprops.cpp -------------------------------- */
+
+U_CAPI void U_EXPORT2
+u_getUnicodeVersion(UVersionInfo versionArray) {
+    if(versionArray!=NULL) {
+        uprv_memcpy(versionArray, dataVersion, U_MAX_VERSION_LENGTH);
+    }
+}
+
+U_CFUNC uint32_t
+u_getMainProperties(UChar32 c) {
+    uint32_t props;
+    GET_PROPS(c, props);
+    return props;
+}
+
+U_CFUNC uint32_t
+u_getUnicodeProperties(UChar32 c, int32_t column) {
+    U_ASSERT(column>=0);
+    if(column>=propsVectorsColumns) {
+        return 0;
+    } else {
+        uint16_t vecIndex=UTRIE2_GET16(&propsVectorsTrie, c);
+        return propsVectors[vecIndex+column];
+    }
+}
+
+U_CFUNC int32_t
+uprv_getMaxValues(int32_t column) {
+    switch(column) {
+    case 0:
+        return indexes[UPROPS_MAX_VALUES_INDEX];
+    case 2:
+        return indexes[UPROPS_MAX_VALUES_2_INDEX];
+    default:
+        return 0;
+    }
+}
+
+U_CAPI void U_EXPORT2
+u_charAge(UChar32 c, UVersionInfo versionArray) {
+    if(versionArray!=NULL) {
+        uint32_t version=u_getUnicodeProperties(c, 0)>>UPROPS_AGE_SHIFT;
+        versionArray[0]=(uint8_t)(version>>4);
+        versionArray[1]=(uint8_t)(version&0xf);
+        versionArray[2]=versionArray[3]=0;
+    }
+}
+
+U_CAPI UScriptCode U_EXPORT2
+uscript_getScript(UChar32 c, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return USCRIPT_INVALID_CODE;
+    }
+    if((uint32_t)c>0x10ffff) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return USCRIPT_INVALID_CODE;
+    }
+    uint32_t scriptX=u_getUnicodeProperties(c, 0)&UPROPS_SCRIPT_X_MASK;
+    uint32_t codeOrIndex=uprops_mergeScriptCodeOrIndex(scriptX);
+    if(scriptX<UPROPS_SCRIPT_X_WITH_COMMON) {
+        return (UScriptCode)codeOrIndex;
+    } else if(scriptX<UPROPS_SCRIPT_X_WITH_INHERITED) {
+        return USCRIPT_COMMON;
+    } else if(scriptX<UPROPS_SCRIPT_X_WITH_OTHER) {
+        return USCRIPT_INHERITED;
+    } else {
+        return (UScriptCode)scriptExtensions[codeOrIndex];
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+uscript_hasScript(UChar32 c, UScriptCode sc) {
+    uint32_t scriptX=u_getUnicodeProperties(c, 0)&UPROPS_SCRIPT_X_MASK;
+    uint32_t codeOrIndex=uprops_mergeScriptCodeOrIndex(scriptX);
+    if(scriptX<UPROPS_SCRIPT_X_WITH_COMMON) {
+        return sc==(UScriptCode)codeOrIndex;
+    }
+
+    const uint16_t *scx=scriptExtensions+codeOrIndex;
+    if(scriptX>=UPROPS_SCRIPT_X_WITH_OTHER) {
+        scx=scriptExtensions+scx[1];
+    }
+    uint32_t sc32=sc;
+    if(sc32>0x7fff) {
+        /* Guard against bogus input that would make us go past the Script_Extensions terminator. */
+        return FALSE;
+    }
+    while(sc32>*scx) {
+        ++scx;
+    }
+    return sc32==(*scx&0x7fff);
+}
+
+U_CAPI int32_t U_EXPORT2
+uscript_getScriptExtensions(UChar32 c,
+                            UScriptCode *scripts, int32_t capacity,
+                            UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(capacity<0 || (capacity>0 && scripts==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    uint32_t scriptX=u_getUnicodeProperties(c, 0)&UPROPS_SCRIPT_X_MASK;
+    uint32_t codeOrIndex=uprops_mergeScriptCodeOrIndex(scriptX);
+    if(scriptX<UPROPS_SCRIPT_X_WITH_COMMON) {
+        if(capacity==0) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        } else {
+            scripts[0]=(UScriptCode)codeOrIndex;
+        }
+        return 1;
+    }
+
+    const uint16_t *scx=scriptExtensions+codeOrIndex;
+    if(scriptX>=UPROPS_SCRIPT_X_WITH_OTHER) {
+        scx=scriptExtensions+scx[1];
+    }
+    int32_t length=0;
+    uint16_t sx;
+    do {
+        sx=*scx++;
+        if(length<capacity) {
+            scripts[length]=(UScriptCode)(sx&0x7fff);
+        }
+        ++length;
+    } while(sx<0x8000);
+    if(length>capacity) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+    return length;
+}
+
+U_CAPI UBlockCode U_EXPORT2
+ublock_getCode(UChar32 c) {
+    return (UBlockCode)((u_getUnicodeProperties(c, 0)&UPROPS_BLOCK_MASK)>>UPROPS_BLOCK_SHIFT);
+}
+
+/* property starts for UnicodeSet ------------------------------------------- */
+
+static UBool U_CALLCONV
+_enumPropertyStartsRange(const void *context, UChar32 start, UChar32 end, uint32_t value) {
+    /* add the start code point to the USet */
+    const USetAdder *sa=(const USetAdder *)context;
+    sa->add(sa->set, start);
+    (void)end;
+    (void)value;
+    return TRUE;
+}
+
+#define USET_ADD_CP_AND_NEXT(sa, cp) sa->add(sa->set, cp); sa->add(sa->set, cp+1)
+
+U_CFUNC void U_EXPORT2
+uchar_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /* add the start code point of each same-value range of the main trie */
+    utrie2_enum(&propsTrie, NULL, _enumPropertyStartsRange, sa);
+
+    /* add code points with hardcoded properties, plus the ones following them */
+
+    /* add for u_isblank() */
+    USET_ADD_CP_AND_NEXT(sa, TAB);
+
+    /* add for IS_THAT_CONTROL_SPACE() */
+    sa->add(sa->set, CR+1); /* range TAB..CR */
+    sa->add(sa->set, 0x1c);
+    sa->add(sa->set, 0x1f+1);
+    USET_ADD_CP_AND_NEXT(sa, NL);
+
+    /* add for u_isIDIgnorable() what was not added above */
+    sa->add(sa->set, DEL); /* range DEL..NBSP-1, NBSP added below */
+    sa->add(sa->set, HAIRSP);
+    sa->add(sa->set, RLM+1);
+    sa->add(sa->set, INHSWAP);
+    sa->add(sa->set, NOMDIG+1);
+    USET_ADD_CP_AND_NEXT(sa, ZWNBSP);
+
+    /* add no-break spaces for u_isWhitespace() what was not added above */
+    USET_ADD_CP_AND_NEXT(sa, NBSP);
+    USET_ADD_CP_AND_NEXT(sa, FIGURESP);
+    USET_ADD_CP_AND_NEXT(sa, NNBSP);
+
+    /* add for u_digit() */
+    sa->add(sa->set, U_a);
+    sa->add(sa->set, U_z+1);
+    sa->add(sa->set, U_A);
+    sa->add(sa->set, U_Z+1);
+    sa->add(sa->set, U_FW_a);
+    sa->add(sa->set, U_FW_z+1);
+    sa->add(sa->set, U_FW_A);
+    sa->add(sa->set, U_FW_Z+1);
+
+    /* add for u_isxdigit() */
+    sa->add(sa->set, U_f+1);
+    sa->add(sa->set, U_F+1);
+    sa->add(sa->set, U_FW_f+1);
+    sa->add(sa->set, U_FW_F+1);
+
+    /* add for UCHAR_DEFAULT_IGNORABLE_CODE_POINT what was not added above */
+    sa->add(sa->set, WJ); /* range WJ..NOMDIG */
+    sa->add(sa->set, 0xfff0);
+    sa->add(sa->set, 0xfffb+1);
+    sa->add(sa->set, 0xe0000);
+    sa->add(sa->set, 0xe0fff+1);
+
+    /* add for UCHAR_GRAPHEME_BASE and others */
+    USET_ADD_CP_AND_NEXT(sa, CGJ);
+}
+
+U_CFUNC void U_EXPORT2
+upropsvec_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /* add the start code point of each same-value range of the properties vectors trie */
+    utrie2_enum(&propsVectorsTrie, NULL, _enumPropertyStartsRange, sa);
+}
diff --git a/thirdparty/icu4c/common/uchar_props_data.h b/thirdparty/icu4c/common/uchar_props_data.h
new file mode 100644
index 0000000000..9a78918204
--- /dev/null
+++ b/thirdparty/icu4c/common/uchar_props_data.h
@@ -0,0 +1,3860 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+//
+// Copyright (C) 1999-2016, International Business Machines
+// Corporation and others.  All Rights Reserved.
+//
+// file name: uchar_props_data.h
+//
+// machine-generated by: icu/tools/unicode/c/genprops/corepropsbuilder.cpp
+
+
+#ifdef INCLUDED_FROM_UCHAR_C
+
+static const UVersionInfo dataVersion={0xd,0,0,0};
+
+static const uint16_t propsTrie_index[22276]={
+0x46d,0x475,0x47d,0x485,0x49d,0x4a5,0x4ad,0x4b5,0x4bd,0x4c5,0x4cb,0x4d3,0x4db,0x4e3,0x4eb,0x4f3,
+0x4f9,0x501,0x509,0x511,0x514,0x51c,0x524,0x52c,0x534,0x53c,0x538,0x540,0x548,0x550,0x555,0x55d,
+0x565,0x56d,0x571,0x579,0x581,0x589,0x591,0x599,0x595,0x59d,0x5a2,0x5aa,0x5b0,0x5b8,0x5c0,0x5c8,
+0x5d0,0x5d8,0x5e0,0x5e8,0x5ed,0x5f5,0x5f8,0x600,0x608,0x610,0x616,0x61e,0x61d,0x625,0x62d,0x635,
+0x645,0x63d,0x64d,0x655,0x48d,0x665,0x66b,0x65d,0x67b,0x67d,0x685,0x673,0x695,0x69b,0x6a3,0x68d,
+0x6b3,0x6b9,0x6c1,0x6ab,0x6d1,0x6d7,0x6df,0x6c9,0x6ef,0x6f5,0x6fd,0x6e7,0x70d,0x715,0x71d,0x705,
+0x72d,0x733,0x73b,0x725,0x74b,0x751,0x759,0x743,0x769,0x76e,0x776,0x761,0x786,0x78d,0x795,0x77e,
+0x619,0x79d,0x7a5,0x48d,0x7ad,0x7b4,0x7bc,0x48d,0x7c4,0x7cc,0x7d4,0x7d9,0x7e1,0x7e8,0x7f0,0x48d,
+0x5d8,0x7f8,0x800,0x808,0x810,0x565,0x820,0x818,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x828,0x5d8,0x830,0x834,0x83c,0x5d8,0x842,0x5d8,0x848,0x850,0x858,0x565,0x565,0x860,
+0x868,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x86d,0x875,0x5d8,0x5d8,0x87d,0x885,0x88d,0x895,0x89d,0x5d8,0x8a5,0x8ad,0x8b5,
+0x8c5,0x5d8,0x8cd,0x8cf,0x8d7,0x8bd,0x5d8,0x8da,0x8ee,0x8e2,0x8ea,0x8f6,0x5d8,0x8fe,0x904,0x90c,
+0x914,0x5d8,0x924,0x92c,0x934,0x91c,0x944,0x48d,0x94c,0x94f,0x957,0x93c,0x967,0x95f,0x5d8,0x96e,
+0x5d8,0x97d,0x976,0x985,0x98d,0x991,0x999,0x9a1,0x50d,0x9a9,0x9ac,0x9b2,0x9b9,0x9ac,0x534,0x9c1,
+0x4bd,0x4bd,0x4bd,0x4bd,0x9c9,0x4bd,0x4bd,0x4bd,0x9d9,0x9e1,0x9e9,0x9f1,0x9f9,0x9fd,0xa05,0x9d1,
+0xa1d,0xa25,0xa0d,0xa15,0xa2d,0xa35,0xa3d,0xa45,0xa5d,0xa4d,0xa55,0xa65,0xa6d,0xa7c,0xa81,0xa74,
+0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa91,0xa99,0x90c,0xa9c,0xaa4,0xaab,0xab0,0xab8,
+0x90c,0xabf,0xabe,0xacf,0xad2,0x90c,0x90c,0xac7,0x90c,0x90c,0x90c,0x90c,0x90c,0xae1,0xae9,0xad9,
+0x90c,0x90c,0x90c,0xaee,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0xaf4,0xafc,0x90c,0xb04,0xb0b,
+0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0xa89,0xa89,0xa89,0xa89,0xb13,0xa89,0xb1a,0xb21,
+0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0x90c,0xb29,0xb30,0xb34,0xb3a,0x90c,0x90c,0x90c,
+0x565,0xb4a,0xb42,0xb52,0x4bd,0x4bd,0x4bd,0xb5a,0x50d,0xb62,0x5d8,0xb68,0xb78,0xb70,0xb70,0x534,
+0xb80,0xb88,0xb90,0x48d,0xb98,0x90c,0x90c,0xb9f,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0xba7,0xbad,
+0xbbd,0xbb5,0x619,0x5d8,0xbc5,0x868,0x5d8,0xbcd,0xbd5,0xbd9,0x5d8,0x5d8,0xbde,0x5d8,0x90c,0xbe5,
+0xab9,0xbed,0xbf3,0x90c,0xbed,0xbfb,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,
+0xc03,0x5d8,0x5d8,0x5d8,0xc0b,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0xc11,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc16,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x90c,0x90c,
+0xc1e,0x5d8,0xc21,0x5d8,0xc29,0xc2f,0xc37,0xc3f,0xc44,0x5d8,0x5d8,0xc48,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc4f,0x5d8,0xc56,0xc5c,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc64,0x5d8,0x5d8,0x5d8,0xc6c,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc6e,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc75,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0xc7c,0x5d8,0x5d8,0x5d8,0xc83,0xc8b,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc90,0x5d8,0x5d8,0xc98,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc9c,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xc9f,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xca2,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0xca8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0xcb0,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0xcb5,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xcba,0x5d8,0x5d8,0x5d8,0xcbf,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0xcc7,0xcce,0xcd2,0x5d8,0x5d8,0x5d8,0xcd9,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x8ce,
+0xce7,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0xcdf,0x90c,0xcef,0x985,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0xcf4,0xcfc,0x4bd,0xd0c,0xd04,0x5d8,0x5d8,0xd14,0xd1c,0xd2c,0x4bd,0xd31,0xd39,0xd3f,0xd47,0xd24,
+0xd4f,0xd57,0x5d8,0xd5f,0xd6f,0xd72,0xd67,0xd7a,0x62d,0xd82,0xd89,0x8ce,0x67b,0xd99,0xd91,0xda1,
+0x5d8,0xda9,0xdb1,0xdb9,0x5d8,0xdc1,0xdc9,0xdd1,0xdd9,0xde1,0xde5,0xded,0x50d,0x50d,0x5d8,0xdf5,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xdfd,0xe09,0xe01,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,
+0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0x5d8,0x5d8,0x5d8,0xe21,0x5d8,0xcda,0xe28,0xe2d,
+0x5d8,0x5d8,0x5d8,0xe35,0x5d8,0x5d8,0x8d9,0x48d,0xe4b,0xe3b,0xe43,0x5d8,0x5d8,0xe53,0xe5b,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xe60,0xe68,0x5d8,0xe6c,0x5d8,0xe72,0xe76,
+0xe7e,0xe86,0xe8d,0xe95,0x5d8,0x5d8,0x5d8,0xe9b,0xeb3,0x47d,0xebb,0xec3,0xec8,0x8ee,0xea3,0xeab,
+0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,
+0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,0xe11,
+0x1234,0x1234,0x1274,0x12b4,0x12f4,0x132c,0x136c,0x13ac,0x13e4,0x1424,0x1450,0x1490,0x14d0,0x14e0,0x1520,0x1554,
+0x1594,0x15c4,0x1604,0x1644,0x1654,0x1688,0x16c0,0x1700,0x1740,0x1780,0x17b4,0x17e0,0x1820,0x1858,0x1874,0x18b4,
+0xa80,0xac0,0xb00,0xb40,0xb80,0xa40,0xbc0,0xa40,0xbe2,0xa40,0xa40,0xa40,0xa40,0xc22,0x1db,0x1db,
+0xc62,0xca2,0xa40,0xa40,0xa40,0xa40,0xce2,0xd02,0xa40,0xa40,0xd42,0xd82,0xdc2,0xe02,0xe42,0xe82,
+0xec2,0xef9,0x1db,0x1db,0xf1d,0xf51,0x1db,0xf79,0x1db,0x1db,0x1db,0x1db,0xfa6,0x1db,0x1db,0x1db,
+0x1db,0x1db,0x1db,0x1db,0xfba,0x1db,0xff2,0x1032,0x1db,0x103d,0x1db,0x1db,0x1db,0x1073,0xa40,0x10b3,
+0x1db,0x1db,0x10f3,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0x1133,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,
+0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x1173,
+0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,
+0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x700,0x1173,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0xed0,0xed7,0xedf,0x48d,0x5d8,0x5d8,0x5d8,0xee7,0xef7,0xeef,0xf0e,0xeff,0xf06,0xf16,0xf1a,0xf1e,
+0x48d,0x48d,0x48d,0x48d,0x8ce,0x5d8,0xf26,0xf2e,0x5d8,0xf36,0xf3e,0xf42,0xf4a,0x5d8,0xf52,0x48d,
+0x565,0x56f,0xf5a,0x5d8,0xf5e,0xf66,0xf76,0xf6e,0x5d8,0xf7e,0x5d8,0xf85,0x48d,0x48d,0x48d,0x48d,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xb78,0x8da,0xe72,0x48d,0x48d,0x48d,0x48d,
+0xf95,0xf8d,0xf98,0xfa0,0x8ee,0xfa8,0x48d,0xfb0,0xfb8,0xfc0,0x48d,0x48d,0x5d8,0xfd0,0xfd8,0xfc8,
+0xfe8,0xfef,0xfe0,0xff7,0xfff,0x48d,0x100f,0x1007,0x5d8,0x1012,0x101a,0x1022,0x102a,0x1032,0x48d,0x48d,
+0x5d8,0x5d8,0x103a,0x48d,0x565,0x1042,0x50d,0x104a,0x5d8,0x1052,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x105a,0x5d8,0x1062,0x48d,0x48d,0x106a,0x1072,0x1079,0x48d,0x48d,0xe68,0x1081,0xb78,
+0x1091,0x60e,0x1099,0x1089,0x967,0x10a1,0x10a9,0x10af,0x10c7,0x10b7,0x10bf,0x10cb,0x967,0x10db,0x10d3,0x10e3,
+0x10f3,0x10eb,0x48d,0x48d,0x10fa,0x1102,0x630,0x110a,0x111a,0x1120,0x1128,0x1112,0x48d,0x48d,0x48d,0x48d,
+0x5d8,0x1130,0x1138,0x1140,0x5d8,0x1148,0x1150,0x48d,0x48d,0x48d,0x48d,0x48d,0x5d8,0x1158,0x1160,0x48d,
+0x5d8,0x1168,0x1170,0x1178,0x5d8,0x1188,0x1180,0x48d,0x848,0x1190,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x5d8,0x1198,0x48d,0x48d,0x48d,0x565,0x50d,0x11a0,0x11b0,0x11b6,0x11a8,0x48d,0x48d,0x11c6,0x11ca,0x11be,
+0x11e2,0x11d2,0x11da,0x5d8,0x11f2,0x11ea,0x5d8,0x8cf,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x1208,0x120d,0x11fa,0x1202,0x121d,0x1215,0x48d,0x48d,0x122c,0x1230,0x1224,0x1240,0x1238,0x1180,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x1244,0x48d,0x48d,0x48d,0x48d,0x48d,0x124b,0x125b,0x1253,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x8d9,0x48d,0x48d,0x48d,
+0x126b,0x1273,0x127b,0x1263,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x1283,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x128b,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x1293,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x8cf,0x8ee,0x129b,0x48d,0x48d,0xe68,0x12a3,0x5d8,0x12b3,0x12bb,0x12c3,0x12ab,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x565,0x50d,0x12cb,0x48d,0x48d,0x48d,0x5d8,0x5d8,0x12d3,0x12d8,0x12de,0x48d,
+0x48d,0x12e6,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x12ee,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x8da,0x48d,0x103a,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x8ee,0x48d,0x12f4,0x12fb,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xe01,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x1301,0x1306,0x130e,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0xba7,0x90c,0x1316,0x90c,0x131d,0x1325,0x132b,
+0x90c,0x1331,0x90c,0x90c,0x1339,0x48d,0x48d,0x48d,0x48d,0x1341,0x90c,0x90c,0xabb,0x1349,0x48d,0x48d,
+0x48d,0x48d,0x1359,0x1360,0x1365,0x136b,0x1373,0x137b,0x1383,0x135d,0x138b,0x1393,0x139b,0x13a0,0x1372,0x1359,
+0x1360,0x135c,0x136b,0x13a8,0x135a,0x13ab,0x135d,0x13b3,0x13bb,0x13c3,0x13ca,0x13b6,0x13be,0x13c6,0x13cd,0x13b9,
+0x13d5,0x1351,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,
+0x90c,0x90c,0x534,0x13e5,0x534,0x13ec,0x13f3,0x13dd,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x13fa,0x1402,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x5d8,0x1412,0x140a,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x5d8,0x141a,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x1422,0x48d,0x565,0x1432,0x142a,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x143a,0x144a,0x1442,0x48d,0x48d,0x145a,0x1452,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x146a,0x1472,0x147a,0x1482,0x148a,0x1492,0x48d,0x1462,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x90c,0x149a,0x90c,0x90c,0xb9f,0x149f,0x14a3,0xba7,0x14ab,0x90c,0x90c,0x90c,0x90c,0xba9,
+0x48d,0x14b3,0x14bb,0x14bf,0x14c7,0x14cf,0x48d,0x48d,0x48d,0x48d,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,
+0x90c,0x14d7,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,
+0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x90c,0x14df,0x14e7,0x90c,0x90c,0x90c,0xb9f,0x90c,0x90c,
+0x14ef,0x14f7,0x149a,0x90c,0x14ff,0x90c,0x1507,0x150c,0x48d,0x48d,0x90c,0x90c,0x90c,0x1514,0x90c,0x90c,
+0x151b,0x90c,0x90c,0x90c,0xb9f,0x1520,0x1528,0x152e,0x1533,0x48d,0x90c,0x90c,0x90c,0x90c,0x153b,0x90c,
+0xabe,0x117c,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x1543,0x5d8,0x5d8,0x154a,0x5d8,0x5d8,0x5d8,0x1552,0x5d8,0x155a,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0xc80,0x5d8,0x5d8,0x1562,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x156a,0x1572,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0xcbf,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x1579,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x1580,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x1587,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0xf5e,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x158b,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0xf5e,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x1066,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x1590,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x1598,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0xf5e,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,
+0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x5d8,0x655,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x15a8,0x15a0,0x15a0,0x15a0,0x48d,0x48d,0x48d,0x48d,0x534,0x534,0x534,0x534,0x534,
+0x534,0x534,0x15b0,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,0x48d,
+0x48d,0x48d,0x48d,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,0xe19,
+0xe19,0xe19,0x15b8,0x46c,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0xc,0x17,0x17,0x17,0x19,0x17,0x17,0x17,0x14,0x15,0x17,0x18,
+0x17,0x13,0x17,0x17,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,
+0x18,0x18,0x18,0x17,0x17,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0x14,
+0x17,0x15,0x1a,0x16,0x1a,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,0x14,
+0x18,0x15,0x18,0xf,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0xc,0x17,0x19,0x19,0x19,0x19,0x1b,0x17,0x1a,0x1b,5,0x1c,
+0x18,0x10,0x1b,0x1a,0x1b,0x18,0x34b,0x38b,0x1a,2,0x17,0x17,0x1a,0x30b,5,0x1d,
+0x34cb,0x344b,0x3ccb,0x17,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0x18,1,1,1,1,
+1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,0x18,2,2,2,2,
+2,2,2,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,2,1,2,1,
+2,1,2,1,2,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,1,2,1,2,1,2,2,2,1,1,2,
+1,2,1,1,2,1,1,1,2,2,1,1,1,1,2,1,
+1,2,1,1,1,2,2,2,1,1,2,1,1,2,1,2,
+1,2,1,1,2,1,2,2,1,2,1,1,2,1,1,1,
+2,1,2,1,1,2,2,5,1,2,2,2,5,5,5,5,
+1,3,2,1,3,2,1,3,2,1,2,1,2,1,2,1,
+2,1,2,1,2,1,2,1,2,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,2,1,3,2,
+1,2,1,1,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,2,2,2,2,2,2,1,1,
+2,1,1,2,2,1,2,1,1,1,1,2,1,2,1,2,
+1,2,1,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,5,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,0x1a,0x1a,0x1a,0x1a,4,4,4,4,4,4,4,4,4,4,
+4,4,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,
+4,4,4,4,4,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,4,0x1a,4,0x1a,
+0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+1,2,1,2,4,0x1a,1,2,0,0,4,2,2,2,0x17,1,
+0,0,0,0,0x1a,0x1a,1,0x17,1,1,1,0,1,0,1,1,
+2,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,0,1,1,1,1,1,1,1,1,1,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,1,2,2,1,1,1,2,2,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,2,2,2,2,1,2,0x18,1,2,1,1,2,
+2,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,0x1b,6,6,6,6,6,7,7,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,1,2,1,2,1,2,1,2,1,2,1,
+2,1,2,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,0,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,
+0,4,0x17,0x17,0x17,0x17,0x17,0x17,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,0x17,0x13,0,0,0x1b,0x1b,0x19,
+0,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,0x13,6,
+0x17,6,6,0x17,6,6,0x17,6,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0,0,0,0,5,
+5,5,5,0x17,0x17,0,0,0,0,0,0,0,0,0,0,0,
+0x10,0x10,0x10,0x10,0x10,0x10,0x18,0x18,0x18,0x17,0x17,0x19,0x17,0x17,0x1b,0x1b,
+6,6,6,6,6,6,6,6,6,6,6,0x17,0x10,0,0x17,0x17,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+4,5,5,5,5,5,5,5,5,5,5,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0x17,0x17,5,5,
+6,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0x17,5,6,6,6,6,6,6,6,0x10,0x1b,6,
+6,6,6,6,6,4,4,6,6,0x1b,6,6,6,6,5,5,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,5,0x1b,0x1b,5,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0,0x10,
+5,6,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,0,0,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,6,6,6,6,6,6,
+6,6,6,6,6,5,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,6,
+6,6,6,6,6,6,6,6,4,4,0x1b,0x17,0x17,0x17,4,0,
+0,6,0x19,0x19,6,6,6,6,4,6,6,6,4,6,6,6,
+6,6,0,0,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,6,6,6,6,4,6,
+6,6,6,6,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,6,6,6,
+0,0,0x17,0,5,5,5,5,5,5,5,5,5,5,5,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,6,6,0x10,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,5,5,
+5,5,5,5,0,0,0,0,0,0,0,0,0,0,0,6,
+6,6,6,6,6,6,6,6,6,6,6,6,5,5,6,6,
+0x17,0x17,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,4,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,6,6,6,8,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,6,8,
+6,5,8,8,8,6,6,6,6,6,6,6,6,8,8,8,
+8,6,8,8,5,6,6,6,6,6,6,6,5,5,5,5,
+5,5,5,5,5,5,6,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,5,5,0x19,0x19,0x37cb,0x35cb,0x3fcb,0x34cb,0x3ccb,0x94b,0x1b,0x19,
+5,0x17,6,0,5,6,8,8,0,5,5,5,5,5,5,5,
+5,0,0,5,5,0,0,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,5,5,5,5,5,5,5,0,5,0,
+0,0,5,5,5,5,0,0,6,5,8,8,8,6,6,6,
+6,0,0,8,8,0,0,8,8,6,5,0,0,0,0,0,
+0,0,0,8,0,0,0,0,5,5,0,5,0,0,0,0,
+0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,6,6,5,5,
+5,6,0x17,0,0,0,0,0,0,0,0,0,0,6,6,8,
+0,5,5,5,5,5,5,0,0,0,0,5,5,0,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,0,5,5,
+5,5,5,5,5,0,5,5,0,5,5,0,5,5,0,0,
+6,0,8,8,8,6,6,0,0,0,0,6,6,0,0,6,
+6,6,0,0,0,6,0,0,0,0,0,0,0,5,5,5,
+5,0,5,0,5,5,6,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,0x17,0x19,0,0,0,0,0,0,0,5,6,6,
+6,6,6,6,0,6,6,8,0,5,5,5,5,5,5,5,
+5,5,0,5,5,5,0,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,5,5,5,5,5,5,5,0,5,5,
+0,5,5,5,5,5,0,0,6,5,8,8,8,6,6,6,
+6,6,0,6,6,8,0,8,8,6,0,0,5,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,5,6,6,
+0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x1b,5,0x34cb,0x344b,
+0x3ccb,0x37cb,0x35cb,0x3fcb,0,0,0,0,0,0,0,0,0,6,8,8,
+0,5,5,5,5,5,5,5,5,0,0,5,5,0,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,0,5,5,
+5,5,5,5,5,0,5,5,0,5,5,5,5,5,0,0,
+6,5,8,6,8,6,6,6,6,0,0,8,8,0,0,8,
+8,6,0,0,0,0,0,0,0,6,6,8,0,0,0,0,
+5,5,0,5,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,0x7cb,0x1e4b,0x784b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x19,0x1b,0,
+0,0,0,0,0,0,6,5,0,5,5,5,5,5,5,0,
+0,0,5,5,5,0,5,5,5,5,0,0,0,5,5,0,
+5,0,5,5,0,0,0,5,5,0,0,0,5,5,5,0,
+0,0,5,5,5,5,5,5,5,5,5,5,5,5,0,0,
+0,0,8,8,6,8,8,0,0,0,8,8,8,0,8,8,
+8,6,0,0,5,0,0,0,0,0,0,8,0,0,0,0,
+0,0,0,0,5,5,6,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,0,0x17,0x54b,0x58b,0x5cb,0x60b,
+0x58b,0x5cb,0x60b,0x1b,6,8,8,8,6,5,5,5,5,5,5,5,
+5,0,5,5,5,0,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,0,0,5,6,6,6,8,8,8,
+8,0,6,6,6,0,6,6,6,6,0,0,0,0,0,0,
+0,6,6,0,5,5,5,0,0,0,0,0,5,5,6,6,
+0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,5,5,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,6,8,8,
+0x17,5,5,5,5,5,5,5,5,0,5,5,5,0,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,0,5,5,
+5,5,5,5,5,5,5,5,0,5,5,5,5,5,0,0,
+6,5,8,6,8,8,8,8,8,0,6,8,8,0,8,8,
+6,6,0,0,0,0,0,0,0,8,8,0,0,0,0,0,
+0,0,5,0,5,5,6,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,0x7cb,0x1e4b,0x784b,0x34cb,0x344b,0x3ccb,0x37cb,0x35cb,0x3fcb,0x1b,5,5,
+5,5,5,5,6,6,8,8,5,5,5,5,5,5,5,5,
+5,0,5,5,5,0,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,6,6,5,8,8,8,6,6,6,6,0,8,8,
+8,0,8,8,8,6,5,0x1b,0,0,0,0,5,5,5,8,
+0xcc0b,0xca0b,0xcb4b,0xc90b,0x364b,0xc94b,0x350b,5,0,0,0,0,0,0,0x49,0x89,
+0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,8,8,0x17,0,0,0,
+0,0,0,0,0,0,0,0,0,6,8,8,0,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,
+0,0,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,5,5,5,5,5,5,5,5,5,
+0,5,0,0,5,5,5,5,5,5,5,0,0,0,6,0,
+0,0,0,8,8,8,6,6,6,0,6,0,8,8,8,8,
+8,8,8,8,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,6,5,5,6,6,6,6,6,6,6,0,
+0,0,0,0x19,5,5,5,5,5,5,4,6,6,6,6,6,
+6,6,6,0x17,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,
+0,0,0,0,0,5,5,0,5,0,5,5,5,5,5,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,5,0,5,5,5,5,5,5,5,5,5,
+5,6,5,5,6,6,6,6,6,6,6,6,6,5,0,0,
+5,5,5,5,5,0,4,0,6,6,6,6,6,6,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,5,5,5,5,
+5,0x1b,0x1b,0x1b,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0x1b,0x17,0x1b,0x1b,0x1b,6,6,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x344b,0x3c4b,0x444b,0x4c4b,0x544b,0x5c4b,
+0x644b,0x6c4b,0x744b,0x2c4b,0x1b,6,0x1b,6,0x1b,6,0x14,0x15,0x14,0x15,8,8,
+5,5,5,5,5,5,5,5,0,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,0,0,0,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,8,6,6,6,6,6,0x17,6,6,5,5,5,5,
+5,6,6,6,6,6,6,6,6,6,6,6,0,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,0,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,6,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,0x1b,
+0x17,0x17,0x17,0x17,0x17,0x1b,0x1b,0x1b,0x1b,0x17,0x17,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,8,8,6,6,6,
+6,8,6,6,6,6,6,6,8,6,6,8,8,6,6,5,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0x17,0x17,0x17,0x17,
+5,5,5,5,5,5,8,8,6,6,5,5,5,5,6,6,
+6,5,8,8,8,5,5,8,8,8,8,8,8,8,5,5,
+5,6,6,6,6,5,5,5,5,5,5,5,5,5,5,5,
+5,5,6,8,8,6,6,8,8,8,8,8,8,6,5,8,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,8,8,8,6,0x1b,0x1b,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,0x17,4,2,2,2,
+1,1,1,1,1,1,0,1,0,0,0,0,0,1,0,0,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,0,0,
+5,5,5,5,5,5,5,0,5,0,5,5,5,5,0,0,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,5,5,5,5,0,0,5,5,5,5,5,5,5,0,
+5,0,5,5,5,5,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0,0,6,6,6,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,
+0x4cb,0x50b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x788b,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,0,0,2,2,2,2,2,2,0,0,
+0x13,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x1b,0x17,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0xc,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x14,
+0x15,0,0,0,5,5,5,5,5,5,5,5,5,5,5,0x17,
+0x17,0x17,0x98a,0x9ca,0xa0a,5,5,5,5,5,5,5,5,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,5,5,5,5,6,6,6,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,6,6,6,0x17,0x17,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,6,6,0,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,5,5,5,0,6,6,0,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,6,6,8,6,6,6,6,6,
+6,6,8,8,8,8,8,8,8,8,6,8,8,6,6,6,
+6,6,6,6,6,6,6,6,0x17,0x17,0x17,4,0x17,0x17,0x17,0x19,
+5,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,
+0,0,0,0,0x54b,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,6,5,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0x17,0x17,0x17,0x17,0x17,0x17,0x13,0x17,0x17,0x17,0x17,6,
+6,6,0x10,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,
+0,0,0,0,5,5,5,4,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,0,0,0,0,0,0,5,5,5,5,
+5,6,6,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,0,
+0,0,0,0,0,0,0,0,6,6,6,8,8,8,8,6,
+6,8,8,8,0,0,0,0,8,8,6,8,8,8,8,8,
+8,6,6,6,0,0,0,0,0x1b,0,0,0,0x17,0x17,0x49,0x89,
+0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,0,5,5,5,5,5,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x30b,0,0,0,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,6,6,8,8,6,0,0,0x17,0x17,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,4,0x17,0x17,0x17,0x17,0x17,0x17,0,0,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,7,6,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,8,6,8,6,6,6,6,6,6,6,0,
+6,8,6,8,8,6,6,6,6,6,6,6,6,8,8,8,
+8,8,8,6,6,6,6,6,6,6,6,6,6,0,0,6,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+0x17,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,6,6,6,6,6,
+6,6,6,6,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,
+6,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+6,6,6,6,8,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+6,8,6,6,6,6,6,8,6,8,8,8,8,8,6,8,
+8,5,5,5,5,5,5,5,0,0,0,0,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0x17,0x17,0x17,0x17,5,8,6,6,
+6,6,8,8,6,6,8,6,6,6,5,5,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,5,5,5,5,6,6,8,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,6,8,
+6,6,8,8,8,6,8,6,6,6,8,8,0,0,0,0,
+0,0,0,0,0x17,0x17,0x17,0x17,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,5,5,5,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,5,5,5,5,5,5,8,8,8,8,8,8,8,8,
+6,6,6,6,6,6,6,6,8,8,6,6,0,0,0,0x17,
+0x17,0x17,0x17,0x17,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,4,4,4,4,
+4,4,0x17,0x17,2,2,2,2,2,2,2,2,2,0,0,0,
+0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,
+0,1,1,1,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0,0,0,0,
+0,0,0,0,6,6,6,0x17,6,6,6,6,6,6,6,6,
+6,6,6,6,6,8,6,6,6,6,6,6,6,5,5,5,
+5,6,5,5,5,5,5,5,6,5,5,8,6,6,5,0,
+0,0,0,0,2,2,2,2,2,2,2,2,2,2,2,2,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,
+4,4,4,2,2,2,2,2,2,2,2,2,2,2,2,2,
+4,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,4,
+4,4,4,4,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,0,6,
+6,6,6,6,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,2,2,2,2,2,2,
+2,2,1,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,0x1a,0x1a,0x1a,0,0,2,2,2,0,2,2,1,1,1,1,
+3,0x1a,0x1a,0,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,2,2,2,2,2,2,0,0,1,1,1,1,
+1,1,0,0,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,2,2,2,2,2,2,0,0,1,1,1,1,
+1,1,0,0,2,2,2,2,2,2,2,2,0,1,0,1,
+0,1,0,1,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,0,0,2,2,2,2,2,2,2,2,3,3,3,3,
+3,3,3,3,2,2,2,2,2,2,2,2,3,3,3,3,
+3,3,3,3,2,2,2,2,2,0,2,2,1,1,1,1,
+3,0x1a,2,0x1a,0x1a,0x1a,2,2,2,0,2,2,1,1,1,1,
+3,0x1a,0x1a,0x1a,2,2,2,2,0,0,2,2,1,1,1,1,
+0,0x1a,0x1a,0x1a,0x16,0x17,0x17,0x17,0x18,0x14,0x15,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0x17,0x17,0x17,0x18,0x17,0x16,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0xc,0x10,0x10,0x10,0x10,0x10,0,0x10,0x10,0x10,0x10,0x10,0x10,
+0x10,0x10,0x10,0x10,0x2cb,4,0,0,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x18,0x18,
+0x18,0x14,0x15,4,0xc,0xc,0xc,0xc,0xc,0xc,0xc,0xc,0xc,0xc,0xc,0x10,
+0x10,0x10,0x10,0x10,0x13,0x13,0x13,0x13,0x13,0x13,0x17,0x17,0x1c,0x1d,0x14,0x1c,
+0x1c,0x1d,0x14,0x1c,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0xd,0xe,0x10,0x10,
+0x10,0x10,0x10,0xc,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x1c,0x1d,0x17,
+0x17,0x17,0x17,0x16,0x2cb,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x18,0x18,
+0x18,0x14,0x15,0,4,4,4,4,4,4,4,4,4,4,4,4,
+4,0,0,0,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,
+0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,0x19,
+0x19,0x19,0x19,0x19,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,6,6,6,6,6,6,6,6,6,6,6,6,
+6,7,7,7,7,6,7,7,7,6,6,6,6,6,6,6,
+6,6,6,6,6,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x1b,0x1b,0x1b,0x1b,1,0x1b,1,0x1b,1,0x1b,1,1,
+1,1,0x1b,2,1,1,1,1,2,5,5,5,5,2,0x1b,0x1b,
+2,2,1,1,0x18,0x18,0x18,0x18,0x18,1,2,2,2,2,0x1b,0x18,
+0x1b,0x1b,2,0x1b,0x358b,0x360b,0x364b,0x348b,0x388b,0x350b,0x390b,0x3d0b,0x410b,0x354b,0x454b,0x35cb,
+0x3dcb,0x45cb,0x4dcb,0x58b,0x1b,0x1b,1,0x1b,0x1b,0x1b,0x1b,1,0x1b,0x1b,2,1,
+1,1,2,2,1,1,1,2,0x1b,1,0x1b,0x1b,0x18,1,1,1,
+1,1,0x1b,0x1b,0x58a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x7ca,0x80a,0x84a,
+0x11ca,0x1e4a,0x980a,0x784a,0x58a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x7ca,0x80a,0x84a,
+0x11ca,0x1e4a,0x980a,0x784a,0x784a,0x984a,0x788a,1,2,0x6ca,0x11ca,0x988a,0x78ca,0x54b,0x1b,0x1b,
+0,0,0,0,0x18,0x18,0x18,0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x18,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x1b,0x1b,0x18,0x1b,0x1b,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x18,0x18,0x1b,0x1b,0x18,0x1b,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x14,0x15,0x14,0x15,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x14,0x15,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,
+0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x2cb,0x80b,0x84b,0x88b,0x8cb,0x90b,0x94b,0x98b,0x9cb,0xa0b,
+0xa4b,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,0x2cb,0x30b,0x34b,0x38b,0x3cb,
+0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,0x80b,0x84b,0x88b,0x8cb,0x90b,0x94b,0x98b,0x9cb,0xa0b,0xa4b,
+0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,0x80b,0x84b,0x88b,0x8cb,0x90b,0x94b,
+0x98b,0x9cb,0xa0b,0xa4b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,
+0x14,0x15,0x14,0x15,0x14,0x15,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,
+0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,
+0x48b,0x4cb,0x50b,0x7cb,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x18,0x18,0x18,0x18,0x18,0x14,0x15,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x14,
+0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,
+0x15,0x14,0x15,0x14,0x15,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x14,0x15,0x14,0x15,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x14,0x15,0x18,0x18,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0x18,0x18,0x18,0x18,0x18,0x1b,0x1b,0x18,0x18,0x18,0x18,0x18,0x18,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,0,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,1,2,1,1,1,2,2,1,
+2,1,2,1,2,1,1,1,1,2,1,2,2,1,2,2,
+2,2,2,2,4,4,1,1,1,2,1,2,2,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,1,2,1,2,6,6,6,1,2,0,0,0,0,
+0,0x17,0x17,0x17,0x17,0x344b,0x17,0x17,2,2,2,2,2,2,0,2,
+0,0,0,0,0,2,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0,0,0,0,0,0,0,4,
+0x17,0,0,0,0,0,0,0,0,0,0,0,0,0,0,6,
+5,5,5,5,5,5,5,0,5,5,5,5,5,5,5,0,
+5,5,5,5,5,5,5,0,5,5,5,5,5,5,5,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,0,0,0,0,0,0,0,0,
+0x17,0x17,0x1c,0x1d,0x1c,0x1d,0x17,0x17,0x17,0x1c,0x1d,0x17,0x1c,0x1d,0x17,0x17,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x13,0x17,0x17,0x13,0x17,0x1c,0x1d,0x17,0x17,
+0x1c,0x1d,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x17,0x17,0x17,0x17,0x17,4,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x13,0x13,0x17,0x17,0x17,0x17,
+0x13,0x17,0x14,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x1b,0x1b,0x17,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0,0,0,0,0x1b,0x58a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,6,6,
+6,6,8,8,0x13,4,4,4,4,4,0x1b,0x1b,0x7ca,0xa4a,0xcca,4,
+5,0x17,0x1b,0x1b,0xc,0x17,0x17,0x17,0x1b,4,5,0x54a,0x14,0x15,0x14,0x15,
+0x14,0x15,0x14,0x15,0x14,0x15,0x1b,0x1b,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,
+0x13,0x14,0x15,0x15,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0,0,6,6,0x1a,
+0x1a,4,4,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x17,
+4,4,4,5,0,0,0,0,0,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,0x1b,0x1b,0x58b,0x5cb,0x60b,0x64b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1b,0xa8b,0xacb,0xb0b,
+0xb4b,0xb8b,0xbcb,0xc0b,0xc4b,0xc8b,0xccb,0xd0b,0xd4b,0xd8b,0xdcb,0xe0b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0xe4b,0xe8b,0xecb,
+0xf0b,0xf4b,0xf8b,0xfcb,0x100b,0x104b,0x108b,0x10cb,0x110b,0x114b,0x118b,0x11cb,5,5,5,5,
+5,0x685,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x5c5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x685,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0x705,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0x585,5,5,0x705,5,5,5,0x7885,
+5,0x605,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x785,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x5c5,5,5,5,5,5,5,5,0x685,5,0x645,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0x7985,0x7c5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0x7845,5,5,5,5,
+5,5,5,5,0x605,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,0x685,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x1e45,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x7985,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x7a85,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x5c5,5,0x745,5,0x6c5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x7c5,5,0x7845,0xa45,0xcc5,5,5,5,5,5,5,0xf45,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x605,0x605,0x605,0x605,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,0x645,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0x585,5,5,5,5,5,5,5,0x585,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0x585,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0x785,0xa45,5,5,5,5,
+5,5,5,5,5,5,5,5,0x585,0x5c5,0x605,5,0x5c5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x7c5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,0x745,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,0x705,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x785,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x1e45,5,
+5,5,5,5,5,5,0x645,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x7885,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0x5c5,5,5,5,5,0x5c5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,0x5c5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,0x7845,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x6c5,5,
+5,5,5,5,0x1e45,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x6c5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0x545,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,4,5,5,5,5,5,5,5,5,5,5,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,4,0x17,0x17,0x17,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,2,1,2,1,2,4,4,6,6,
+1,2,1,2,1,2,1,2,1,2,1,2,1,2,5,6,
+7,7,7,0x17,6,6,6,6,6,6,6,6,6,6,0x17,4,
+5,5,5,5,5,5,0x58a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x54a,
+6,6,0x17,0x17,0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,0,
+0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,
+0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,4,4,4,4,4,4,4,4,4,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,1,2,5,4,4,2,5,5,5,5,5,
+0x1a,0x1a,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+2,2,1,2,1,2,1,2,1,2,1,2,1,2,1,2,
+1,2,1,2,4,2,2,2,2,2,2,2,2,1,2,1,
+2,1,1,2,1,2,1,2,1,2,1,2,4,0x1a,0x1a,1,
+2,1,2,5,1,2,1,2,2,2,1,2,1,2,1,2,
+1,2,1,2,1,2,1,1,1,1,1,2,1,1,1,1,
+1,2,1,2,1,2,1,2,1,2,1,2,0,0,1,2,
+1,1,1,1,2,1,2,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,5,6,5,
+5,5,6,5,5,5,5,6,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,8,
+8,6,6,8,0x1b,0x1b,0x1b,0x1b,6,0,0,0,0x34cb,0x344b,0x3ccb,0x37cb,
+0x35cb,0x3fcb,0x1b,0x1b,0x19,0x1b,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,0,8,8,8,8,
+6,6,0,0,0,0,0,0,0,0,0x17,0x17,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,8,8,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,8,8,8,8,
+8,8,8,8,8,8,8,8,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,5,5,5,5,5,5,
+0x17,0x17,0x17,5,0x17,5,5,6,5,5,5,5,5,5,6,6,
+6,6,6,6,6,6,0x17,0x17,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,6,6,6,6,6,
+6,6,6,6,6,6,8,8,0,0,0,0,0,0,0,0,
+0,0,0,0x17,8,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0,4,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,
+0,0,0x17,0x17,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,6,8,8,6,6,6,6,8,8,
+6,6,8,8,5,5,5,5,5,6,4,5,5,5,5,5,
+5,5,5,5,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,5,5,
+5,5,5,0,5,5,5,5,5,5,5,5,5,6,6,6,
+6,6,6,8,8,6,6,8,8,6,6,0,0,0,0,0,
+0,0,0,0,5,5,5,6,5,5,5,5,5,5,5,5,
+6,8,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,
+0x17,0x17,0x17,0x17,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,4,5,5,5,5,5,5,0x1b,0x1b,0x1b,5,8,
+6,8,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,6,5,6,6,6,5,5,6,6,5,5,5,
+5,5,6,6,5,6,5,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,5,
+5,4,0x17,0x17,5,5,5,5,5,5,5,5,5,5,5,8,
+6,6,8,8,0x17,0x17,5,4,4,8,6,0,0,0,0,0,
+0,0,0,0,0,5,5,5,5,5,5,0,0,5,5,5,
+5,5,5,0,0,5,5,5,5,5,5,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,0,5,5,5,5,
+5,5,5,0,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,0x1a,
+4,4,4,4,2,2,2,2,2,2,2,2,2,4,0x1a,0x1a,
+0,0,0,0,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,5,5,5,8,8,6,8,8,6,8,8,0x17,
+8,6,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,
+0,0,0,0,5,5,5,5,0,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0,0,0,0,5,5,5,5,5,5,5,0,0,0,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12,
+0x12,0x12,0x12,0x12,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,
+0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,
+0x11,0x11,0x11,0x11,5,5,5,5,5,5,5,5,5,5,5,0x605,
+5,5,5,5,5,5,5,0x7c5,5,5,5,5,0x5c5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x6c5,5,0x6c5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x7c5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0x18,5,5,5,5,5,5,5,5,5,5,
+5,5,5,0,5,5,5,5,5,0,5,0,5,5,0,5,
+5,0,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,2,2,2,2,
+2,2,2,0,0,0,0,0,0,0,0,0,0,0,0,2,
+2,2,2,2,0,0,0,0,0,5,6,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x1a,0x1a,
+0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0x15,0x14,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0,0,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,0x19,0x1b,0,0,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x14,
+0x15,0x17,0,0,0,0,0,0,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,0x17,0x13,0x13,0x16,0x16,0x14,0x15,0x14,
+0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x14,0x15,0x17,0x17,0x14,0x15,0x17,0x17,0x17,
+0x17,0x16,0x16,0x16,0x17,0x17,0x17,0,0x17,0x17,0x17,0x17,0x13,0x14,0x15,0x14,
+0x15,0x14,0x15,0x17,0x17,0x17,0x18,0x13,0x18,0x18,0x18,0,0x17,0x19,0x17,0x17,
+0,0,0,0,5,5,5,5,5,0,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,0,0x10,0,0,5,5,
+5,5,5,5,0,0,5,5,5,5,5,5,0,0,5,5,
+5,5,5,5,0,0,5,5,5,0,0,0,0x19,0x19,0x18,0x1a,
+0x1b,0x19,0x19,0,0x1b,0x18,0x18,0x18,0x18,0x1b,0x1b,0,0,0,0,0,
+0,0,0,0,0,0x10,0x10,0x10,0x1b,0x1b,0,0,0,0x17,0x17,0x17,
+0x19,0x17,0x17,0x17,0x14,0x15,0x17,0x18,0x17,0x13,0x17,0x17,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0x18,0x18,0x18,0x17,0x1a,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,0x14,0x18,0x15,0x18,0x14,0x15,0x17,0x14,0x15,
+0x17,0x17,5,5,5,5,5,5,5,5,5,5,4,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,4,4,
+5,5,5,5,5,5,5,5,5,5,5,5,0,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,5,5,0,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0,0,0,0,0,0xb00b,0xb80b,0x784b,0x804b,
+0x884b,0x904b,0x984b,0xa04b,0xa84b,0xb04b,0xb84b,0x788b,0x808b,0x888b,0x908b,0x988b,0xa08b,0xa88b,0xb08b,0xb88b,
+0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x17,0x17,0x17,0,
+0,0,0,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,
+0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x800b,0x880b,0x900b,0x980b,0xa00b,0xa80b,0x7ca,0x7ca,0x7ca,0x7ca,
+0x7ca,0xcca,0x11ca,0x11ca,0x11ca,0x11ca,0x1e4a,0x880a,0x980a,0x980a,0x980a,0x980a,0x980a,0x784a,0x984a,0x68a,
+0x11ca,0x344b,0x344b,0x388b,0x3ccb,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x54b,0x34cb,0x1b,0x1b,0x1b,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0x34ca,0x344a,0x58a,0x68a,0x11ca,0x980a,0x984a,0x988a,
+0x68a,0x7ca,0x11ca,0x1e4a,0x980a,0x784a,0x984a,0x68a,0x7ca,0x11ca,0x1e4a,0x980a,0x784a,0x788a,0x988a,0x7ca,
+0x58a,0x58a,0x58a,0x5ca,0x5ca,0x5ca,0x5ca,0x68a,0x1b,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,6,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,6,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,
+0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x800b,0x880b,0x900b,0x980b,
+0xa00b,0xa80b,0xb00b,0xb80b,0,0,0,0,0x58b,0x68b,0x7cb,0x11cb,0,0,0,0,
+0,0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x1bca,5,5,5,5,5,5,
+5,5,0xb80a,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,6,6,
+6,6,6,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,0x17,5,5,5,5,0,0,0,0,
+5,5,5,5,5,5,5,5,0x17,0x58a,0x5ca,0x7ca,0xa4a,0x1e4a,0,0,
+0,0,0,0,0,0,0,0,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,0,0,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,0,0,0,0,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,
+2,2,2,2,2,2,2,2,5,5,5,5,5,5,5,5,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0,0,0,0,0,0,0,0,
+0,0,0,0x17,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,5,5,0,0,0,
+5,0,0,5,5,5,5,5,5,5,0,0,5,0,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,0x17,0x58b,0x5cb,0x60b,0x7cb,0xa4b,0x1e4b,0x784b,0x788b,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,0x1b,0x1b,0x58b,0x5cb,0x60b,0x64b,0x68b,0x7cb,0xa4b,
+0,0,0,0,0,0,0,0x58b,0x5cb,0x60b,0x64b,0x64b,0x68b,0x7cb,0xa4b,0x1e4b,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,0,5,5,0,0,0,0,0,0x58b,0x68b,0x7cb,0xa4b,0x1e4b,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0x58b,0x7cb,0xa4b,0x1e4b,0x5cb,0x60b,0,0,0,0x17,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,0,0,0,0x17,
+0xa04b,0xa84b,0xb04b,0xb84b,0x788b,0x808b,0x888b,0x908b,0x988b,0xa08b,0xa88b,0xb08b,0xb88b,0x78cb,0x80cb,0x88cb,
+0x90cb,0x98cb,0xa0cb,0xa8cb,0xb0cb,0xb8cb,0x36cb,0x354b,0x34cb,0x348b,0x46cb,0x344b,0x4ecb,0x388b,0x3ccb,0x454b,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0,0,0,0,0x5ecb,0x344b,5,5,
+0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,
+0,0,0x1e4b,0x800b,0x880b,0x900b,0x980b,0xa00b,0xa80b,0xb00b,0xb80b,0x784b,0x804b,0x884b,0x904b,0x984b,
+0x30b,0x34b,0x38b,0x3cb,0x7cb,0xa4b,0x1e4b,0x784b,0x344b,0,0,0,0,0,0,0,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,
+5,6,6,6,0,6,6,0,0,0,0,0,6,6,6,6,
+5,5,5,5,0,5,5,5,0,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,0,0,6,6,6,0,0,0,0,6,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x58b,0x11cb,0x17,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x58b,0x7cb,0xa4b,5,5,5,5,
+5,6,6,0,0,0,0,0x58b,0x68b,0x7cb,0xa4b,0x1e4b,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,0x1b,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,0,
+0,0x17,0x17,0x17,0x17,0x17,0x17,0x17,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,0,
+0x58b,0x5cb,0x60b,0x64b,0x7cb,0xa4b,0x1e4b,0x784b,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0,0,0,0,0,
+0x58b,0x5cb,0x60b,0x64b,0x7cb,0xa4b,0x1e4b,0x784b,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,0,0,0,0,
+0,0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,0,0,0,0,
+0,0x58b,0x5cb,0x60b,0x64b,0x7cb,0xa4b,0x1e4b,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,
+0,0,0,0,0,0,0,0,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,0,0,0,0,0,
+0,0,0x58b,0x68b,0x7cb,0x11cb,0x1e4b,0x784b,5,5,5,5,6,6,6,6,
+0,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,0,0,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,
+0x50b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x800b,0x880b,0x900b,0x980b,0xa00b,
+0xa80b,0xb00b,0xb80b,0x344b,0x34cb,0x348b,0x388b,0,5,5,5,5,5,5,5,5,
+5,5,0,6,6,0x13,0,0,5,5,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0x58b,0x5cb,0x60b,0x64b,0x68b,0x7cb,0xa4b,0xccb,0x1e4b,0x344b,5,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,6,6,6,6,6,6,
+6,6,6,6,6,0x58b,0x7cb,0xa4b,0x1e4b,0x17,0x17,0x17,0x17,0x17,0,0,
+0,0,0,0,5,5,5,5,5,0x58b,0x5cb,0x60b,0x64b,0x7cb,0xa4b,0x1e4b,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x784b,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,6,8,6,8,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,6,6,6,6,6,6,6,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0,0,0,0,0x30b,0x34b,0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x7cb,
+0xa4b,0xccb,0xf4b,0x11cb,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,8,8,8,6,6,6,6,8,8,6,6,0x17,
+0x17,0x10,0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,0,0,0,
+0,0x10,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,5,5,5,5,
+5,5,5,6,6,6,6,6,8,6,6,6,6,6,6,6,
+6,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0x17,0x17,
+5,8,8,5,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,6,6,6,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,6,
+0x17,0x17,5,0,0,0,0,0,0,0,0,0,8,5,5,5,
+5,0x17,0x17,0x17,0x17,6,6,6,6,0x17,8,6,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,5,0x17,5,0x17,0x17,0x17,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,8,
+8,8,6,6,6,6,6,6,6,6,6,8,0,0x58b,0x5cb,0x60b,
+0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,
+0x784b,0,0,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,8,8,8,6,6,6,8,8,
+6,8,6,6,0x17,0x17,0x17,0x17,0x17,0x17,6,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,
+5,0,5,5,5,5,0,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,0,5,5,5,5,5,5,5,5,5,
+5,0x17,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,8,8,8,6,6,6,6,6,
+6,6,6,0,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,0,0,5,5,8,8,0,0,6,6,
+6,6,6,6,6,0,0,0,6,6,6,6,6,0,0,0,
+0,0,0,0,0,0,0,0,6,6,8,8,0,5,5,5,
+5,5,5,5,5,0,0,5,5,0,0,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,5,5,
+5,0,5,5,0,5,5,5,5,5,0,6,6,5,8,8,
+6,8,8,8,8,0,0,8,8,0,0,8,8,8,0,0,
+5,0,0,0,0,0,0,8,0,0,0,0,0,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,8,8,8,6,6,6,6,6,6,6,6,
+8,8,6,6,6,8,6,5,5,5,5,0x17,0x17,0x17,0x17,0x17,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x17,0x17,0,0x17,6,5,
+5,5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+8,8,8,6,6,6,6,6,6,8,6,8,8,8,8,6,
+6,8,6,6,5,5,0x17,5,0,0,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,8,
+8,8,6,6,6,6,0,0,8,8,8,8,6,6,8,6,
+6,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,5,5,5,5,6,6,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+8,8,8,6,6,6,6,6,6,6,6,8,8,6,8,6,
+6,0x17,0x17,0x17,5,0,0,0,0,0,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,6,8,6,8,8,
+6,6,6,6,6,6,8,6,5,0,0,0,0,0,0,0,
+8,8,6,6,6,6,8,6,6,6,6,6,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x7cb,0xa4b,0x17,0x17,0x17,0x1b,
+5,5,5,5,5,5,5,5,5,5,5,5,8,8,8,6,
+6,6,6,6,6,6,6,6,8,6,6,0x17,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,
+0x16cb,0x194b,0x1bcb,0,0,0,0,0,0,0,0,0,0,0,0,5,
+8,5,8,6,0x17,0x17,0x17,0,0,0,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+5,5,5,5,5,5,5,0,0,5,0,0,5,5,5,5,
+5,5,5,5,0,5,5,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,8,8,8,8,8,8,0,8,
+8,0,0,6,6,8,6,5,6,5,0x17,5,8,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+0,0,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,8,8,8,6,6,6,6,
+0,0,6,6,8,8,8,8,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,6,6,6,6,6,
+6,8,5,6,6,6,6,0x17,0x17,0x17,0x17,0x17,0x17,0x17,0x17,6,
+0,0,0,0,0,0,0,0,5,6,6,6,6,6,6,8,
+8,6,6,6,5,5,5,5,5,6,6,6,6,6,6,6,
+6,6,6,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0x17,0x17,0x17,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,6,6,6,6,6,6,6,6,6,6,6,6,6,8,
+6,6,0x17,0x17,0x17,5,0x17,0x17,5,0x17,0x17,0x17,0x17,0x17,0,0,
+0,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,
+0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0,0,0,0x17,0x17,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,8,6,6,6,6,6,6,6,0,6,6,6,6,
+6,6,8,6,6,6,6,6,6,6,6,6,0,8,6,6,
+6,6,6,6,6,8,6,6,8,6,6,0,0,0,0,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,0,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,5,6,0,0,0,0,0,0,0,0,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0,0,
+5,5,5,5,5,5,5,0,5,5,0,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,6,6,6,6,6,6,0,0,0,6,0,6,6,0,6,
+5,5,5,5,5,5,5,5,5,5,8,8,8,8,8,0,
+6,6,0,8,8,6,8,6,5,0,0,0,0,0,0,0,
+5,5,5,5,5,5,0,5,5,0,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,6,6,8,8,0x17,0x17,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x19,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0x17,0xcd0b,0xcc0b,0xcb0b,0xd00b,
+0xca0b,0xcf0b,0xcb4b,0xd04b,0xc90b,0x37cb,0x37cb,0x364b,0x35cb,0xc94b,0x3fcb,0x350b,0x34cb,0x344b,0x344b,0x3ccb,
+0xcd0b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x19,0x19,0x19,0x34ca,0x354a,0x34ca,0x34ca,
+0x344a,0x348a,0x388a,0xf4a,0x11ca,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0,0x17,0x17,0x17,0x17,
+0x17,0,0,0,0,0,0,0,0,0,0,0,0x5ca,0x60a,0x64a,0x68a,
+0x6ca,0x70a,0x74a,0x78a,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x64a,0x68a,0x6ca,0x70a,0x74a,
+0x78a,0x58a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x58a,0x5ca,0x60a,0x64a,0x68a,0x5ca,
+0x60a,0x60a,0x64a,0x68a,0x6ca,0x70a,0x74a,0x78a,0x58a,0x5ca,0x60a,0x60a,0x64a,0x68a,0xc08a,0xc18a,
+0x58a,0x5ca,0x60a,0x60a,0x64a,0x68a,0x60a,0x60a,0x64a,0x64a,0x64a,0x64a,0x6ca,0x70a,0x70a,0x70a,
+0x74a,0x74a,0x78a,0x78a,0x78a,0x78a,0x5ca,0x60a,0x64a,0x68a,0x6ca,0x58a,0x5ca,0x60a,0x64a,0x64a,
+0x68a,0x68a,0x5ca,0x60a,0x58a,0x5ca,0x348a,0x388a,0x454a,0x348a,0x388a,0x35ca,5,5,5,5,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0,0x10,0x10,0x10,0x10,
+0x10,0x10,0x10,0x10,0x10,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0,0,0,0,0x17,0x17,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0,0,6,6,6,6,
+6,0x17,0,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,6,6,6,6,
+6,6,6,0x17,0x17,0x17,0x17,0x17,0x1b,0x1b,0x1b,0x1b,4,4,4,4,
+0x17,0x1b,0,0,0,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0x249,0x289,0,0x7cb,0x1e4b,0x788b,0x790b,0x798b,0x7a0b,0x7a8b,0,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,0,0,0,0,5,5,5,0x54b,0x58b,0x5cb,0x60b,
+0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0x80b,0x84b,0x88b,0x8cb,0x90b,0x94b,0x98b,0x9cb,0xa0b,
+0x58b,0x5cb,0x60b,0x17,0x17,0x17,0x17,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,0,0,0,0,6,5,8,8,8,
+8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,
+0,0,0,0,0,0,0,6,6,6,6,4,4,4,4,4,
+4,4,4,4,4,4,4,4,4,4,0x17,4,6,0,0,0,
+0,0,0,0,0,0,0,0,8,8,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,0,0,0,0,0,0,0,0,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,0,
+0,0,0,0,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,0,0,0x1b,6,6,0x17,0x10,0x10,0x10,0x10,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,
+0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,8,8,6,6,6,0x1b,0x1b,
+0x1b,8,8,8,8,8,8,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,6,
+6,6,6,6,6,6,6,0x1b,0x1b,6,6,6,6,6,6,6,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,6,6,6,6,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x1b,0x1b,6,6,6,0x1b,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0x54b,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0x80b,
+0x84b,0x88b,0x8cb,0x90b,0x94b,0x98b,0x9cb,0xa0b,0,0,0,0,0,0,0,0,
+0,0,0,0,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,0xa4b,0xccb,
+0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x58b,0x5cb,0x60b,0x64b,0x68b,0x58b,0x68b,0,0,0,
+0,0,0,0,0x249,0x289,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,
+0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x49,0x89,0xc9,0x109,0x149,0x189,
+0x1c9,0x209,0x249,0x289,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,2,2,2,2,2,2,2,0,2,2,2,2,2,2,
+2,2,2,2,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,1,0,1,1,0,0,1,0,
+0,1,1,0,0,1,1,1,1,0,1,1,1,1,1,1,
+1,1,2,2,2,2,0,2,0,2,2,2,2,2,2,2,
+0,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,
+1,1,0,1,1,1,1,0,0,1,1,1,1,1,1,1,
+1,0,1,1,1,1,1,1,1,0,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,1,1,0,1,1,1,1,0,1,1,1,1,
+1,0,1,0,0,0,1,1,1,1,1,1,1,0,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,
+2,2,0,0,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,0x18,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,0x18,2,2,2,2,2,2,1,1,
+1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,1,1,1,0x18,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,0x18,2,2,
+2,2,2,2,1,1,1,1,1,1,1,1,1,1,1,1,
+1,1,1,1,2,2,2,0x18,2,2,2,2,2,2,1,2,
+0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,0x249,0x289,0x49,0x89,0xc9,0x109,
+0x149,0x189,0x1c9,0x209,0,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,0x1b,0x1b,0x1b,0x1b,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,6,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+6,0x1b,0x1b,0x17,0x17,0x17,0x17,0x17,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,6,6,6,6,6,6,6,6,0,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,0,0,6,6,6,6,6,6,6,0,6,6,0,6,6,
+6,6,6,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,5,0x1b,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0,0,0,6,6,6,6,6,6,6,4,
+4,4,4,4,4,4,0,0,5,5,5,5,5,5,5,5,
+5,5,5,5,6,6,6,6,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,0,0x19,5,5,5,5,5,0,0,0x58b,
+0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,6,6,6,6,6,6,6,0,
+0,0,0,0,0,0,0,0,2,2,2,2,6,6,6,6,
+6,6,6,4,0,0,0,0,0x49,0x89,0xc9,0x109,0x149,0x189,0x1c9,0x209,
+0x249,0x289,0,0,0,0,0x17,0x17,1,1,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+2,2,2,2,2,2,2,2,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,
+0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x78cb,0x794b,0x814b,0x58b,0x5cb,0x60b,0x64b,0x68b,
+0x6cb,0x70b,0x74b,0x78b,0x1b,0x34cb,0x344b,0x3ccb,0x19,0x58b,0x5cb,0x788b,0x78cb,0,0,0,
+0,0,0,0,0,0,0,0,0x16cb,0x194b,0x1bcb,0x1e4b,0x800b,0x880b,0x900b,0x980b,
+0xa00b,0xa80b,0xb00b,0xb80b,0x784b,0x804b,0x884b,0x904b,0x984b,0xa04b,0xa84b,0xb04b,0xb84b,0x788b,0x808b,0x888b,
+0x908b,0x988b,0xa08b,0xa88b,0xb08b,0xb88b,0x78cb,0x80cb,0x984b,0xa04b,0xa84b,0xb04b,0xb84b,0x788b,0x808b,0x888b,
+0x908b,0x988b,0xa08b,0xa88b,0xb08b,0xb88b,0x1b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,0x74b,0x78b,0x7cb,
+0x900b,0xa00b,0x804b,0x788b,0x344b,0x354b,0,0,0,0x58b,0x5cb,0x60b,0x64b,0x68b,0x6cb,0x70b,
+0x74b,0x78b,0x7cb,0xa4b,0xccb,0xf4b,0x11cb,0x144b,0x16cb,0x194b,0x1bcb,0x1e4b,0x800b,0x880b,0x900b,0x980b,
+0xa00b,0xa80b,0xb00b,0xb80b,0x784b,0x804b,0x884b,0x904b,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0x18,0x18,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,5,5,5,5,0,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,0,5,5,0,5,0,0,5,
+0,5,5,5,5,5,5,5,5,5,5,0,5,5,5,5,
+0,5,0,5,0,0,0,0,0,0,5,0,0,0,0,5,
+0,5,0,5,0,5,5,5,0,5,5,0,5,0,0,5,
+0,5,0,5,0,5,0,5,0,5,5,0,5,0,0,5,
+5,5,5,0,5,5,5,5,5,5,5,0,5,5,5,5,
+0,5,5,5,5,0,5,0,5,5,5,5,5,5,5,5,
+5,5,0,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,0,0,0,0,5,5,5,0,5,5,5,
+5,5,0,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x2cb,0x2cb,0x30b,0x34b,
+0x38b,0x3cb,0x40b,0x44b,0x48b,0x4cb,0x50b,0x54b,0x54b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,
+0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0x1b,0x1b,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1a,0x1a,0x1a,0x1a,0x1a,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,
+0x1b,0x1b,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0x1b,0x1b,0x1b,0,0,0,0,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0,0,0,0,0,0,0,0,0,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0,
+0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,0x1b,5,0x705,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,0x645,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,0x645,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x685,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0xcc5,5,5,5,5,
+5,5,5,5,0xf45,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,0xf45,5,5,5,5,5,5,5,5,5,5,5,
+5,5,0x6c5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,0x605,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0x605,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,0x605,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,0x605,5,5,5,5,
+5,5,5,5,5,5,5,5,5,0x645,5,5,5,5,5,5,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x785,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
+0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,
+0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,0x10,
+0,0x10,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,
+0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0x11,0,0,
+0,0,0,0
+};
+
+static const UTrie2 propsTrie={
+    propsTrie_index,
+    propsTrie_index+4532,
+    NULL,
+    4532,
+    17744,
+    0xa40,
+    0x1234,
+    0x0,
+    0x0,
+    0x110000,
+    0x5700,
+    NULL, 0, FALSE, FALSE, 0, NULL
+};
+
+static const uint16_t propsVectorsTrie_index[31228]={
+0x4e8,0x4f0,0x4f8,0x500,0x518,0x520,0x528,0x530,0x538,0x540,0x548,0x550,0x558,0x560,0x568,0x570,
+0x577,0x57f,0x587,0x58f,0x592,0x59a,0x5a2,0x5aa,0x5b2,0x5ba,0x5c2,0x5ca,0x5d2,0x5da,0x5e2,0x5ea,
+0x5f2,0x5fa,0x601,0x609,0x611,0x619,0x621,0x629,0x631,0x639,0x63e,0x646,0x64d,0x655,0x65d,0x665,
+0x66d,0x675,0x67d,0x685,0x68c,0x694,0x69c,0x6a4,0x6ac,0x6b4,0x6bc,0x6c4,0x6cc,0x6d4,0x6dc,0x6e4,
+0x1a38,0xd5e,0xe35,0x6ec,0x508,0xe9c,0xea4,0x1bf2,0x1300,0x1310,0x12f8,0x1308,0x7c5,0x7cb,0x7d3,0x7db,
+0x7e3,0x7e9,0x7f1,0x7f9,0x801,0x807,0x80f,0x817,0x81f,0x825,0x82d,0x835,0x83d,0x845,0x84d,0x854,
+0x85c,0x862,0x86a,0x872,0x87a,0x880,0x888,0x890,0x898,0x1318,0x8a0,0x8a8,0x8b0,0x8b7,0x8bf,0x8c7,
+0x8cf,0x8d3,0x8db,0x8e2,0x8ea,0x8f2,0x8fa,0x902,0x162c,0x1634,0x90a,0x912,0x91a,0x922,0x92a,0x931,
+0x1692,0x1682,0x168a,0x1973,0x197b,0x1328,0x939,0x1320,0x1572,0x1572,0x1574,0x133c,0x133d,0x1330,0x1332,0x1334,
+0x169a,0x169c,0x941,0x169c,0x949,0x94e,0x956,0x16a1,0x95c,0x169c,0x962,0x96a,0xc39,0x16a9,0x16a9,0x972,
+0x16b9,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,0x16ba,
+0x16ba,0x16ba,0x16ba,0x16b1,0x97a,0x16c2,0x16c2,0x982,0xb59,0xb61,0xb69,0xb71,0x16d2,0x16ca,0x98a,0x992,
+0x99a,0x16dc,0x16e4,0x9a2,0x16da,0x9aa,0x1a40,0xd66,0xb79,0xb81,0xb89,0xb8e,0x18e1,0xc6c,0xc73,0x1849,
+0xc09,0x1a48,0xd6e,0xd76,0xd7e,0xd86,0xf47,0xf48,0x1939,0x193e,0xca8,0xcb0,0x19af,0x19b7,0x1b11,0xe3d,
+0x19bf,0xcf2,0xcfa,0x19c7,0x10f6,0x1196,0xf27,0xd8e,0x1869,0x1851,0x1861,0x1859,0x18f9,0x18f1,0x18b9,0xc19,
+0x1345,0x1345,0x1345,0x1345,0x1348,0x1345,0x1345,0x1350,0x9b2,0x1358,0x9b6,0x9be,0x1358,0x9c6,0x9ce,0x9d6,
+0x1368,0x1360,0x1370,0x9de,0x9e6,0x1378,0x9ee,0x9f6,0x1380,0x1388,0x1390,0x1398,0x9fe,0x13a0,0x13a7,0x13af,
+0x13b7,0x13bf,0x13c7,0x13cf,0x13d7,0x13de,0x13e6,0x13ee,0x13f6,0x13fe,0x1401,0x1403,0x16ec,0x17dc,0x17e2,0x1929,
+0x140b,0xa06,0xa0e,0x1525,0x152a,0x152d,0x1535,0x1413,0x153d,0x153d,0x1423,0x141b,0x142b,0x1433,0x143b,0x1443,
+0x144b,0x1453,0x145b,0x1463,0x17ea,0x1841,0x1983,0x1ad9,0x1473,0x147a,0x1482,0x148a,0x146b,0x1492,0x17f2,0x17f9,
+0x16f4,0x16f4,0x16f4,0x16f4,0x16f4,0x16f4,0x16f4,0x16f4,0x1801,0x1804,0x1801,0x1801,0x180c,0x1813,0x1815,0x181c,
+0x1824,0x1828,0x1828,0x182b,0x1828,0x1828,0x1831,0x1828,0x1871,0x1931,0x198b,0xb96,0xb9c,0x1c36,0x1c3e,0x1d15,
+0x18d1,0xc49,0xc4d,0x1946,0x18c1,0x18c1,0x18c1,0xc21,0x18c9,0xc41,0x1911,0xc98,0xc29,0xc31,0xc31,0x19cf,
+0x1901,0x1993,0xc83,0xc88,0xa16,0x16fc,0x16fc,0xa1e,0x1704,0x1704,0x1704,0x1704,0x1704,0x1704,0xa26,0x6f0,
+0x155a,0x157c,0xa2e,0x1584,0xa36,0x158c,0x1594,0x159c,0xa3e,0xa43,0x15a4,0x15ab,0xa48,0x170c,0x1921,0xc11,
+0xa50,0x1606,0x160d,0x15b3,0x1615,0x161c,0x15bb,0x15bf,0x15d8,0x15d8,0x15da,0x15c7,0x15cf,0x15cf,0x15d0,0x1624,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,
+0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1714,0x1717,0x1879,0x1879,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,
+0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e2,0x15e9,0x1a30,0x12b5,
+0x171f,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,
+0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,0x1725,
+0x1725,0x1725,0x1725,0x1725,0xa58,0x172d,0xa60,0x1a50,0x19db,0x19db,0x19db,0x19db,0x19db,0x19db,0x19db,0x19db,
+0x19d7,0xd02,0x19eb,0x19e3,0x19ed,0x1a58,0x1a58,0xd96,0x18d9,0x194e,0x19a3,0x19a7,0x199b,0x1b09,0xcb8,0xcbb,
+0x1909,0xc90,0x1956,0xcc3,0x19f5,0x19f8,0xd0a,0x1a60,0x1a08,0x1a00,0xd12,0xd9e,0x1a68,0x1a6c,0xda6,0xff0,
+0x1a10,0xd1a,0xd22,0x1a74,0x1a84,0x1a7c,0xdae,0xef7,0xe45,0xe4d,0x1c85,0xfa8,0x1d32,0x1d32,0x1a8c,0xdb6,
+0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,
+0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,
+0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,
+0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,
+0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,
+0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,
+0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,
+0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,
+0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,
+0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,
+0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,
+0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,
+0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,
+0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,
+0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,
+0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,
+0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,
+0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,
+0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,
+0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,
+0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,
+0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0x167a,0x1674,0x1675,0x1676,0x1677,0x1678,0x1679,0xa68,0xdbe,0xdc1,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,
+0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,0x164c,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,
+0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x1545,0x15f1,0x15f1,0x15f1,0x15f1,0x15f1,0x15f1,0x15f1,0x15f1,
+0x15f6,0x15fe,0x1839,0x12bd,0x1919,0x1919,0x12c1,0x12c8,0xa70,0xa78,0xa80,0x14b2,0x14b9,0x14c1,0xa88,0x14c9,
+0x14fa,0x14fa,0x14a2,0x14aa,0x14d1,0x14f1,0x14f2,0x1502,0x14d9,0x149a,0xa90,0x14e1,0xa98,0x14e9,0xaa0,0xaa4,
+0xca0,0x150a,0xaac,0xab4,0x1512,0x1518,0x151d,0xabc,0xacc,0x1562,0x156a,0x154d,0x1552,0xad4,0xadc,0xac4,
+0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,
+0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x163c,0x1644,0x1644,0x1644,0x1644,
+0x1420,0x1420,0x1460,0x14a0,0x14e0,0x1520,0x1560,0x15a0,0x15dc,0x161c,0x1648,0x1688,0x16c8,0x1708,0x1748,0x1788,
+0x17c8,0x1804,0x1844,0x1884,0x18c4,0x18f8,0x1934,0x1974,0x19b4,0x19f4,0x1a30,0x1a70,0x1ab0,0x1af0,0x1b30,0x1b70,
+0xa80,0xac0,0xb00,0xb40,0xb80,0xa40,0xe75,0xa40,0xe97,0xa40,0xa40,0xa40,0xa40,0xbc0,0x12dd,0x12dd,
+0xed7,0xc00,0xa40,0xa40,0xa40,0xa40,0xf17,0xc2d,0xa40,0xa40,0xc6d,0xcad,0xced,0xd2d,0xe35,0xda5,
+0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,0x121d,
+0x121d,0x121d,0x121d,0x121d,0xf57,0x125d,0x1092,0x10d2,0x129d,0x10dd,0x131d,0x131d,0x131d,0xf97,0xfb7,0xff7,
+0x135d,0x135d,0x1037,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,
+0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0xfb7,0x1052,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0xde5,0xdf5,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,
+0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xa40,0xd65,
+0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,
+0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x119d,0x111d,
+0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,
+0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x11dd,0x115d,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0xba4,0xbab,0xbb3,0xbbb,0x1881,0x1881,0x1881,0xbc3,0xbcb,0xbce,0x18b1,0x18a9,0xc01,0xd2a,0xd2e,0xd32,
+0x508,0x508,0x508,0x508,0xd3a,0x1a18,0xd42,0xf3f,0x1735,0xae4,0xaea,0x1000,0xbd6,0x18e9,0xc7b,0x508,
+0x174a,0x173d,0x1742,0x1889,0xbde,0xbe6,0x1134,0x113a,0x1c6d,0xf5d,0x1c5d,0x6f8,0x508,0x508,0x508,0x508,
+0x1c8d,0x1c8d,0x1c8d,0x1c8d,0x1c8d,0x1c8d,0x1c8d,0x1c8d,0x1c8d,0xfb0,0xfb8,0xfc0,0x508,0x508,0x508,0x508,
+0xbee,0xbf1,0xdc9,0x1cd5,0xff8,0x700,0x508,0x1092,0xccb,0xd4a,0x508,0x508,0x1c02,0xeff,0xf07,0x1d1d,
+0xc55,0xc5c,0xc64,0x1a94,0x1cb5,0x508,0x1c95,0xfd0,0x1a9c,0xdd1,0xdd9,0xde1,0x1020,0x708,0x508,0x508,
+0x1aa4,0x1aa4,0x710,0x508,0x1d4a,0x10aa,0x1d42,0x10b2,0x1e0e,0x11ac,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0xde9,0x1e66,0x1291,0x508,0x508,0x1e2e,0x11d4,0x11db,0x718,0x508,0x71c,0x1248,0x11e3,
+0x1b19,0x1b1b,0xe55,0xe5c,0x1aac,0x1ab4,0xdf1,0xf1f,0x1bfa,0xee7,0xeef,0xfc8,0x1c1a,0x1c1e,0x1c26,0x1040,
+0xf93,0xf98,0x724,0x508,0x109a,0x10a2,0x1c7d,0xfa0,0xf75,0xf7b,0xf83,0xf8b,0x508,0x508,0x508,0x508,
+0x1daa,0x1da2,0x1124,0x112c,0x1cfd,0x1cf5,0x1068,0x508,0x508,0x508,0x508,0x508,0x1ce5,0x1028,0x1030,0x1038,
+0x1cad,0x1ca5,0xfe0,0x111c,0x1c2e,0xf2f,0x72c,0x508,0x1078,0x1080,0x508,0x508,0x508,0x508,0x508,0x508,
+0x1e06,0x118e,0x734,0x508,0x508,0x1d0d,0x1d05,0x1070,0x1250,0x1256,0x125e,0x508,0x508,0x11eb,0x11ef,0x11f7,
+0x1dde,0x1dd6,0x1176,0x1dce,0x1dc6,0x73c,0x1cdd,0x1018,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x10da,0x10df,0x10e7,0x10ee,0x110e,0x1114,0x508,0x508,0x115a,0x115e,0x1166,0x119e,0x11a4,0x744,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x11bc,0x508,0x508,0x508,0x508,0x508,0x748,0x1e4e,0x1238,
+0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,
+0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x195e,0x1963,0xcd3,0xcda,0xcda,0xcda,
+0x196b,0x196b,0x196b,0xce2,0x1d3a,0x1d3a,0x1d3a,0x1d3a,0x1d3a,0x1d3a,0x750,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,0x1b23,
+0x1b23,0xe64,0xfe8,0x758,0x508,0x508,0x75c,0xf37,0x1ccd,0x1cc5,0x1008,0x1010,0x764,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x1e26,0x1e1e,0x11cc,0x508,0x508,0x508,0x1c12,0x1c12,0xf0f,0x1c0a,0xf17,0x508,0x508,0x1106,
+0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,
+0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dba,0x1dbe,0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,
+0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,0x1e76,0x1266,0x126c,0x1286,0x1289,0x1289,0x1289,0x76c,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x1c65,0x1c65,0x1c65,
+0xf50,0xf55,0x774,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x1752,0x1752,0x1752,
+0x1752,0x1752,0x1752,0x1752,0xaf2,0x1762,0xafa,0x1763,0x175a,0x176b,0x1771,0x1779,0xb02,0x18a1,0x18a1,0x77c,
+0x508,0x508,0x508,0x508,0x11c4,0x1891,0x1891,0xbf9,0xcea,0x508,0x508,0x508,0x508,0x17aa,0x17b1,0xb0a,
+0x17b4,0xb12,0xb1a,0xb22,0x17ae,0xb2a,0xb32,0xb3a,0x17b3,0x17bb,0x17aa,0x17b1,0x17ad,0x17b4,0x17bc,0x17ab,
+0x17b2,0x17ae,0xb41,0x1781,0x1789,0x1790,0x1797,0x1784,0x178c,0x1793,0x179a,0xb49,0x17a2,0x1d62,0x1d62,0x1d62,
+0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d62,0x1d52,0x1d55,0x1d52,
+0x1d5c,0x10ca,0x784,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x10fe,0x78c,0x508,
+0x508,0x508,0x508,0x508,0x508,0x1e46,0x11ff,0x794,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x1e56,0x1240,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x1c9d,0x1c9d,0x1c9d,
+0x1c9d,0x1c9d,0x1c9d,0xfd8,0x508,0x1d9a,0x1d92,0x10d2,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x798,0x1e16,0x11b4,0x508,0x508,0x1207,0x1208,0x7a0,0x508,0x508,0x508,0x508,0x508,0xeac,0xeb4,0xebc,
+0xec4,0xecc,0xed4,0xedb,0xedf,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x7a4,0x1048,0x1ced,0x104e,0x1ced,0x1056,0x105b,0x1060,0x1060,0x1d72,0x1d82,0x1d8a,
+0x10ba,0x1d7a,0x1e36,0x10c2,0x1dee,0x1e3e,0x1e3e,0x117e,0x1186,0x121f,0x1225,0x122a,0x1230,0x1e5e,0x1e5e,0x1e5e,
+0x1e5e,0x1274,0x1e5e,0x127a,0x127e,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,
+0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,0x7ac,
+0x7ac,0x7ac,0x7ac,0x7ac,0x7ad,0xb51,0x17c4,0x17c4,0x17c4,0x7b5,0x7b5,0x7b5,0x7b5,0x1899,0x1899,0x1899,
+0x1899,0x1899,0x1899,0x1899,0x7bd,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,
+0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,
+0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,
+0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,
+0x7b5,0x7b5,0x7b5,0x7b5,0x7b5,0x1a20,0xd52,0x1a28,0x1a28,0xd56,0xe6c,0xe74,0xe7c,0x1ae9,0x1ad1,0x1af1,
+0x1af9,0x1ae1,0xe01,0xe05,0xe0c,0xe14,0xe1b,0xe23,0xe2b,0xe2d,0xe2d,0xe2d,0xe2d,0x1b5a,0x1b62,0x1b5a,
+0x1b68,0x1b70,0x1b3b,0x1b78,0x1b80,0x1b5a,0x1b88,0x1b90,0x1b97,0x1b9f,0x1b43,0x1b5a,0x1ba4,0x1b4b,0x1b52,0x1bac,
+0x1bb2,0x1c4e,0x1c55,0x1c46,0x1bba,0x1bc2,0x1bca,0x1bd2,0x1cbd,0x1bda,0x1be2,0xe84,0xe8c,0x1b2b,0x1b2b,0x1b2b,
+0xe94,0x1c75,0x1c75,0xf65,0xf6d,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,
+0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abe,0x1abc,0x1ac6,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,
+0x1ac9,0x1abc,0x1abc,0x1abc,0x1abc,0x1abc,0xdf9,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,
+0x1d25,0x1d25,0x1d25,0x1d25,0x1d25,0x1d2a,0x1d25,0x1d25,0x1d25,0x1088,0x108a,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1142,0x1b33,0x1de6,0x1de6,0x1de6,0x1de6,0x1de6,0x1de6,0x1de6,0x114a,
+0x1152,0x1210,0x1217,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x1dfe,0x116e,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x508,
+0x508,0x508,0x508,0x508,0x508,0x508,0x508,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x12d0,0x1299,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x12a1,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,
+0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x17d4,0x12d8,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x12a5,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x12ad,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,0x1299,
+0x1299,0x12a5,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x12e0,0x1bea,0x1bea,0x1bea,0x1bea,
+0x1bea,0x1bea,0x12e8,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x12f0,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1654,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x165c,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,
+0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x1664,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,
+0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x166c,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,
+0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x17cc,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,
+0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1b01,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,
+0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1d6a,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,
+0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1db2,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,
+0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1df6,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,
+0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x1e6e,0x4e7,0x4e7,0x4e7,
+0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2ca,0x2d3,0x2cd,0x2cd,0x2d0,0x2c7,0x2c7,
+0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,0x2c7,
+0x7fb,0x7f5,0x7da,0x7d1,0x7c8,0x7c5,0x7bc,0x7d7,0x7c2,0x7ce,0x7d1,0x7ec,0x7e3,0x7d4,0x7f8,0x7cb,
+0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7b9,0x7e0,0x7dd,0x7e6,0x7e6,0x7e6,0x7f5,
+0x7bc,0x807,0x807,0x807,0x807,0x807,0x807,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,
+0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x801,0x7c2,0x7c8,0x7ce,0x7f2,0x7b6,
+0x7ef,0x804,0x804,0x804,0x804,0x804,0x804,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,
+0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7fe,0x7c2,0x7e9,0x7bf,0x7e6,0x2c7,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2e5,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,
+0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,0x2d6,
+0x2d9,0x651,0x810,0x813,0x657,0x813,0x80d,0x64e,0x645,0x2df,0x663,0x2e2,0x816,0x63c,0x65a,0x80a,
+0x654,0x660,0x642,0x642,0x648,0x2dc,0x64e,0x64b,0x645,0x642,0x663,0x2e2,0x63f,0x63f,0x63f,0x651,
+0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x66c,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,
+0x66c,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x65d,0x66c,0x2eb,0x2eb,0x2eb,0x2eb,0x2eb,0x66c,0x666,
+0x669,0x669,0x2e8,0x2e8,0x2e8,0x2e8,0x666,0x2e8,0x669,0x669,0x669,0x2e8,0x669,0x669,0x2e8,0x2e8,
+0x666,0x2e8,0x669,0x669,0x2e8,0x2e8,0x2e8,0x65d,0x666,0x669,0x669,0x2e8,0x669,0x2e8,0x666,0x2e8,
+0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,
+0x2f4,0x66f,0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x2ee,
+0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x678,0x66f,0x2f7,0x2ee,0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x2ee,
+0x2f7,0x66f,0x67b,0x675,0x2f7,0x2ee,0x2f7,0x2ee,0x66f,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x67b,
+0x675,0x678,0x66f,0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x672,0x67e,0x678,0x66f,0x2f7,0x672,0x2f7,0x2ee,
+0x2f7,0x2ee,0x678,0x66f,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,
+0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x678,0x66f,0x2f7,0x2ee,0x2f7,0x672,0x2f7,0x2ee,0x2f7,0x2ee,
+0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2f7,0x2ee,0x2f7,0x2ee,0x2f7,0x2ee,0x2f1,
+0x2fa,0x306,0x306,0x2fa,0x306,0x2fa,0x306,0x306,0x2fa,0x306,0x306,0x306,0x2fa,0x2fa,0x306,0x306,
+0x306,0x306,0x2fa,0x306,0x306,0x2fa,0x306,0x306,0x306,0x2fa,0x2fa,0x2fa,0x306,0x306,0x2fa,0x306,
+0x309,0x2fd,0x306,0x2fa,0x306,0x2fa,0x306,0x306,0x2fa,0x306,0x2fa,0x2fa,0x306,0x2fa,0x306,0x309,
+0x2fd,0x306,0x306,0x306,0x2fa,0x306,0x2fa,0x306,0x306,0x2fa,0x2fa,0x303,0x306,0x2fa,0x2fa,0x2fa,
+0x303,0x303,0x303,0x303,0x30c,0x30c,0x300,0x30c,0x30c,0x300,0x30c,0x30c,0x300,0x309,0x681,0x309,
+0x681,0x309,0x681,0x309,0x681,0x309,0x681,0x309,0x681,0x309,0x681,0x309,0x681,0x2fa,0x309,0x2fd,
+0x309,0x2fd,0x309,0x2fd,0x306,0x2fa,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,
+0x2fd,0x30c,0x30c,0x300,0x309,0x2fd,0x9ea,0x9ea,0x9ed,0x9e7,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,
+0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,0x309,0x2fd,
+0x309,0x2fd,0x309,0x2fd,0x9ed,0x9e7,0x9ed,0x9e7,0x9ea,0x9e4,0x9ed,0x9e7,0xbaf,0xcb7,0x9ea,0x9e4,
+0x9ea,0x9e4,0x9ed,0x9e7,0x9ed,0x9e7,0x9ed,0x9e7,0x9ed,0x9e7,0x9ed,0x9e7,0x9ed,0x9e7,0x9ed,0x9e7,
+0xcb7,0xcb7,0xcb7,0xdb6,0xdb6,0xdb6,0xdb9,0xdb9,0xdb6,0xdb9,0xdb9,0xdb6,0xdb6,0xdb9,0xefa,0xefd,
+0xefd,0xefd,0xefd,0xefa,0xefd,0xefa,0xefd,0xefa,0xefd,0xefa,0xefd,0xefa,0x30f,0x684,0x30f,0x30f,
+0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x684,0x30f,0x30f,
+0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,
+0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x312,0x30f,0x30f,0x30f,
+0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,0x30f,
+0x30f,0x9f0,0x9f0,0x9f0,0x9f0,0x9f0,0xcba,0xcba,0x327,0x327,0x327,0x327,0x327,0x327,0x327,0x327,
+0x327,0x31e,0x31e,0x31e,0x31e,0x31e,0x31e,0x31e,0x31b,0x31b,0x318,0x318,0x68a,0x318,0x31e,0x68d,
+0x321,0x68d,0x68d,0x68d,0x321,0x68d,0x31e,0x31e,0x690,0x324,0x318,0x318,0x318,0x318,0x318,0x318,
+0x687,0x687,0x687,0x687,0x315,0x687,0x318,0xb25,0x327,0x327,0x327,0x327,0x327,0x318,0x318,0x318,
+0x318,0x318,0x9f9,0x9f9,0x9f6,0x9f3,0x9f6,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,
+0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0xcbd,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,
+0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,
+0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,
+0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,
+0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x693,0x696,0x696,0x94b,0x696,0x696,0x94e,0xb28,0xb28,
+0xb28,0xb28,0xb28,0xb28,0xb28,0xb28,0xb28,0xc6c,0xd83,0xd83,0xd83,0xd83,0xd83,0xd83,0xd83,0xd83,
+0xebe,0xebe,0xebe,0xebe,0xec1,0xd86,0xd86,0xd86,0x699,0x699,0xb2b,0xcb4,0xcb4,0xcb4,0xcb4,0xcb4,
+0xcb4,0xcb4,0xcb4,0xcb4,0xcb4,0xcb4,0xcb4,0xcb4,0xfa8,0xfa5,0xfa8,0xfa5,0x333,0x33c,0xfa8,0xfa5,
+9,9,0x342,0xf00,0xf00,0xf00,0x32a,0x14fd,9,9,9,9,0x33f,0x32d,0x351,0x330,
+0x351,0x351,0x351,9,0x351,9,0x351,0x351,0x348,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,
+0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,0x69f,9,0x69f,0x69f,0x69f,0x69f,0x69f,
+0x69f,0x69f,0x351,0x351,0x348,0x348,0x348,0x348,0x348,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,
+0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x69c,0x345,0x69c,0x69c,0x69c,0x69c,0x69c,
+0x69c,0x69c,0x348,0x348,0x348,0x348,0x348,0xfa8,0x354,0x354,0x357,0x351,0x351,0x354,0x34b,0x9fc,
+0xbb8,0xbb5,0x34e,0x9fc,0x34e,0x9fc,0x34e,0x9fc,0x34e,0x9fc,0x339,0x336,0x339,0x336,0x339,0x336,
+0x339,0x336,0x339,0x336,0x339,0x336,0x339,0x336,0x354,0x354,0x34b,0x345,0xb67,0xb64,0xbb2,0xcc3,
+0xcc0,0xcc6,0xcc3,0xcc0,0xdbc,0xdbf,0xdbf,0xdbf,0xa0b,0x6ab,0x363,0x366,0x363,0x363,0x363,0x366,
+0x363,0x363,0x363,0x363,0x366,0xa0b,0x366,0x363,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,
+0x6a8,0x6ab,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,
+0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a8,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,
+0x6a2,0x6a5,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,0x6a2,
+0x6a2,0x6a2,0x6a2,0x6a2,0xa05,0x6a5,0x35d,0x360,0x35d,0x35d,0x35d,0x360,0x35d,0x35d,0x35d,0x35d,
+0x360,0xa05,0x360,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,
+0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x366,0x360,0x363,0x35d,0x363,0x35d,
+0x363,0x35d,0x363,0x35d,0x363,0x35d,0x35a,0x957,0x95a,0x93c,0x93c,0x114f,0x9ff,0x9ff,0xbbe,0xbbb,
+0xa08,0xa02,0xa08,0xa02,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,
+0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,
+0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,0x363,0x35d,
+0x363,0x35d,0x363,0x35d,0x363,0x366,0x360,0x363,0x35d,0xbbe,0xbbb,0x363,0x35d,0xbbe,0xbbb,0x363,
+0x35d,0xbbe,0xbbb,0xf03,0x366,0x360,0x366,0x360,0x363,0x35d,0x366,0x360,0x363,0x35d,0x366,0x360,
+0x366,0x360,0x366,0x360,0x363,0x35d,0x366,0x360,0x366,0x360,0x366,0x360,0x363,0x35d,0x366,0x360,
+0xa0b,0xa05,0x366,0x360,0x366,0x360,0x366,0x360,0x366,0x360,0xdc5,0xdc2,0x366,0x360,0xf06,0xf03,
+0xf06,0xf03,0xf06,0xf03,0xc2d,0xc2a,0xc2d,0xc2a,0xc2d,0xc2a,0xc2d,0xc2a,0xc2d,0xc2a,0xc2d,0xc2a,
+0xc2d,0xc2a,0xc2d,0xc2a,0xf33,0xf30,0xf33,0xf30,0x1023,0x1020,0x1023,0x1020,0x1023,0x1020,0x1023,0x1020,
+0x1023,0x1020,0x1023,0x1020,0x1023,0x1020,0x1023,0x1020,0x1188,0x1185,0x1371,0x136e,0x1536,0x1533,0x1536,0x1533,
+0x1536,0x1533,0x1536,0x1533,0xc,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,
+0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0x378,0xc,
+0xc,0x37b,0x369,0x369,0x369,0x36f,0x369,0x36c,0x1941,0x372,0x372,0x372,0x372,0x372,0x372,0x372,
+0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,
+0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x372,0x375,0x1941,0x37e,0xa0e,0xc,
+0xc,0x1500,0x1500,0x141c,0xf,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,
+0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0xdc8,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,0x97e,
+0x97e,0x97e,0x97e,0x97e,0x381,0x381,0x381,0x381,0x381,0x381,0x381,0x381,0x381,0x381,0xf09,0x381,
+0x381,0x381,0x38d,0x381,0x384,0x381,0x381,0x390,0x981,0xdcb,0xdce,0xdcb,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,
+0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0x393,0xf,
+0xf,0xf,0xf,0x1944,0x393,0x393,0x393,0x38a,0x387,0xf,0xf,0xf,0xf,0xf,0xf,0xf,
+0xf,0xf,0xf,0xf,0xcdb,0xcdb,0xcdb,0xcdb,0x141f,0x1503,0xfb1,0xfb1,0xfb1,0xfae,0xfae,0xdd4,
+0x8c7,0xcd5,0xcd2,0xcd2,0xcc9,0xcc9,0xcc9,0xcc9,0xcc9,0xcc9,0xfab,0xfab,0xfab,0xfab,0xfab,0x8c4,
+0x14fa,0x12,0xdd7,0x8ca,0x1338,0x3ae,0x3b1,0x3b1,0x3b1,0x3b1,0x3b1,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,
+0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0xfb4,
+0xfb4,0xfb4,0xfb4,0xfb4,0x8cd,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x942,
+0x942,0x942,0x942,0x942,0x942,0x942,0x942,0xb5e,0xb5e,0xb5e,0xcc9,0xccf,0xccc,0xdd1,0xdd1,0xdd1,
+0xdd1,0xdd1,0xdd1,0x1335,0x960,0x960,0x960,0x960,0x960,0x960,0x960,0x960,0x960,0x960,0x3a8,0x3a5,
+0x3a2,0x39f,0xbc1,0xbc1,0x93f,0x3ae,0x3ae,0x3ba,0x3ae,0x3b4,0x3b4,0x3b4,0x3b4,0x3ae,0x3ae,0x3ae,
+0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,
+0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,
+0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,
+0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0xa14,0xa14,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0xa14,
+0x3b1,0x3ae,0x3b1,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0x3ae,0xa14,
+0x3ae,0x3ae,0x3ae,0x3b1,0x95d,0x3ae,0x399,0x399,0x399,0x399,0x399,0x399,0x399,0x396,0x39f,0x39c,
+0x39c,0x399,0x399,0x399,0x399,0x3b7,0x3b7,0x399,0x399,0x39f,0x39c,0x39c,0x39c,0x399,0xcd8,0xcd8,
+0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0x3ab,0xa14,0xa14,0xa14,0xa11,0xa11,0xcd8,
+0xa29,0xa29,0xa29,0xa23,0xa23,0xa23,0xa23,0xa23,0xa23,0xa23,0xa23,0xa20,0xa23,0xa20,0x15,0xa2c,
+0xa26,0xa17,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,
+0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xa26,0xcde,0xcde,0xcde,
+0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,0xa1d,
+0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0xa1a,0x15,0x15,0xcde,0xcde,0xcde,
+0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,
+0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0xe37,0x1035,0x1035,
+0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,0x1035,
+0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,
+0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,
+0xa32,0xa32,0xa32,0xa32,0xa32,0xa32,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,0xa2f,
+0xa2f,0xbc4,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,0x18,
+0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4b,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,
+0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,
+0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf4e,0xf42,0xf42,0xf42,0xf42,0xf42,
+0xf42,0xf42,0xf42,0xf42,0xf51,0xf51,0xf45,0xf45,0xf48,0xf57,0xf54,0x10b,0x10b,0x1968,0x196b,0x196b,
+0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x18f9,0x252,0x252,0x252,0x252,0x252,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xb37,0xb37,0xb3a,0xb3a,0xb37,0xb37,0xb37,0xb37,0xb37,0xb37,0xb37,0xb37,0x72,0x72,0x72,0x72,
+0x15ba,0x15ba,0x15ba,0x15ba,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x1bc,0x15b7,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1ef,0x1ef,0x1ef,0x1ef,0x1ef,0x1ef,0x1ef,0x1674,0x1674,0x1674,0x1674,0x1674,0x1674,0x1674,0x1674,0x1674,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x16ad,0x16ad,0x16ad,0x16ad,0x16ad,0x16ad,0x16ad,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x174,0x174,0x174,0x174,0x174,0x174,0x174,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,0x276,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,
+0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,0x1d1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x14d6,0x1b6,0x1b6,0x1b6,0x1b6,0x1b6,0x1b6,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,0x25b,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x18e7,0x18ea,0x18ea,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,0x24f,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x19b6,0x261,0x261,0x261,0x261,0x261,0x261,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1b1b,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,0x29d,
+0x17a9,0x17a9,0x17a9,0x17a9,0x216,0x216,0x216,0x216,0x216,0x216,0x216,0x216,0x216,0x216,0x216,0x216,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x165f,0x165f,0x165f,0x165f,0x165f,0x165f,0x165f,0x165f,0x165f,0x165f,0x1e9,0x1e9,0x1e9,0x1e9,0x1665,0x1665,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,
+0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1713,0x1713,0x1713,0x1713,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,0x1fe,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0xe2e,0xe2e,0xe2b,0xe2b,0xe2b,0xe2e,0xd8,0xd8,0xd8,0xd8,0xd8,0xd8,0xd8,0xd8,0xd8,0xd8,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x225,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1845,0x1845,0x231,0x1845,0x1845,0x231,0x1845,0x1845,0x1845,0x1845,0x1845,0x231,0x231,0x231,0x231,0x231,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x1a6d,0x282,0x282,0x282,0x282,0x1a70,0x1a6a,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0x267,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,
+0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,0x285,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
+0,0,0,0,0,0,0,0,0,0,0,0,0,0,0x97b,0x97b,
+3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,
+3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,
+3,3,0x97b,0x97b,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,
+0xd8c,0xd8c,0xd8c,0xd8c,6,6,6,6,6,6,6,6,6,6,6,6,
+6,6,6,6,0x1509,0x3d5,0x3e4,0x3e4,0x1b,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,
+0x3ea,0x1b,0x1b,0x3ea,0x3ea,0x1b,0x1b,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,
+0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x1b,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x3ea,0x1b,0x3ea,0x1b,
+0x1b,0x1b,0x3ea,0x3ea,0x3ea,0x3ea,0x1b,0x1b,0x3d8,0xce4,0x3d5,0x3e4,0x3e4,0x3d5,0x3d5,0x3d5,
+0x3d5,0x1b,0x1b,0x3e4,0x3e4,0x1b,0x1b,0x3e7,0x3e7,0x3db,0xddd,0x1b,0x1b,0x1b,0x1b,0x1b,
+0x1b,0x1b,0x1b,0x3d5,0x1b,0x1b,0x1b,0x1b,0x3ed,0x3ed,0x1b,0x3ed,0x3ea,0x3ea,0x3d5,0x3d5,
+0x1b,0x1b,0x966,0x966,0x966,0x966,0x966,0x966,0x966,0x966,0x966,0x966,0x3ea,0x3ea,0x3e1,0x3e1,
+0x3de,0x3de,0x3de,0x3de,0x3de,0x3e1,0x3de,0x115e,0x18a2,0x189f,0x1947,0x1b,0x1e,0xce7,0x3f0,0xcea,
+0x1e,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x1e,0x1e,0x1e,0x1e,0x3fc,0x3fc,0x1e,0x1e,0x3fc,
+0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x1e,0x3fc,0x3fc,
+0x3fc,0x3fc,0x3fc,0x3fc,0x3fc,0x1e,0x3fc,0x3ff,0x1e,0x3fc,0x3ff,0x1e,0x3fc,0x3fc,0x1e,0x1e,
+0x3f3,0x1e,0x3f9,0x3f9,0x3f9,0x3f0,0x3f0,0x1e,0x1e,0x1e,0x1e,0x3f0,0x3f0,0x1e,0x1e,0x3f0,
+0x3f0,0x3f6,0x1e,0x1e,0x1e,0xfbd,0x1e,0x1e,0x1e,0x1e,0x1e,0x1e,0x1e,0x3ff,0x3ff,0x3ff,
+0x3fc,0x1e,0x3ff,0x1e,0x1e,0x1e,0x1e,0x1e,0x1e,0x1e,0x969,0x969,0x969,0x969,0x969,0x969,
+0x969,0x969,0x969,0x969,0x3f0,0x3f0,0x3fc,0x3fc,0x3fc,0xfbd,0x194a,0x1e,0x1e,0x1e,0x1e,0x1e,
+0x1e,0x1e,0x1e,0x1e,0x21,0x402,0x402,0x40b,0x21,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,
+0xcf3,0x40e,0x21,0x40e,0x40e,0x40e,0x21,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,
+0x40e,0x40e,0x40e,0x40e,0x40e,0x21,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x40e,0x21,0x40e,0x40e,
+0x21,0x40e,0x40e,0x40e,0x40e,0x40e,0x21,0x21,0x405,0x40e,0x40b,0x40b,0x40b,0x402,0x402,0x402,
+0x402,0x402,0x21,0x402,0x402,0x40b,0x21,0x40b,0x40b,0x408,0x21,0x21,0x40e,0x21,0x21,0x21,
+0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x21,0x40e,0xcf3,0xced,0xced,
+0x21,0x21,0x96c,0x96c,0x96c,0x96c,0x96c,0x96c,0x96c,0x96c,0x96c,0x96c,0x1422,0xcf0,0x21,0x21,
+0x21,0x21,0x21,0x21,0x21,0x1725,0x18a5,0x18a5,0x18a5,0x18a8,0x18a8,0x18a8,0x24,0x411,0x420,0x420,
+0x24,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x24,0x24,0x426,0x426,0x24,0x24,0x426,
+0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x426,0x24,0x426,0x426,
+0x426,0x426,0x426,0x426,0x426,0x24,0x426,0x426,0x24,0xcf6,0x426,0x426,0x426,0x426,0x24,0x24,
+0x414,0x426,0x411,0x411,0x420,0x411,0x411,0x411,0xfc0,0x24,0x24,0x420,0x423,0x24,0x24,0x423,
+0x423,0x417,0x24,0x24,0x24,0x24,0x24,0x24,0x24,0x1ab5,0x411,0x411,0x24,0x24,0x24,0x24,
+0x429,0x429,0x24,0x426,0x426,0x426,0xfc0,0xfc0,0x24,0x24,0x41d,0x41d,0x41d,0x41d,0x41d,0x41d,
+0x41d,0x41d,0x41d,0x41d,0x41a,0xcf6,0x1344,0x1344,0x1344,0x1344,0x1344,0x1344,0x24,0x24,0x24,0x24,
+0x24,0x24,0x24,0x24,0x27,0x27,0x42c,0x438,0x27,0x438,0x438,0x438,0x438,0x438,0x438,0x27,
+0x27,0x27,0x438,0x438,0x438,0x27,0x438,0x438,0x43b,0x438,0x27,0x27,0x27,0x438,0x438,0x27,
+0x438,0x27,0x438,0x438,0x27,0x27,0x27,0x438,0x438,0x27,0x27,0x27,0x438,0x438,0x438,0x27,
+0x27,0x27,0x438,0x438,0x438,0x438,0x438,0x438,0x438,0x438,0xde0,0x438,0x438,0x438,0x27,0x27,
+0x27,0x27,0x42c,0x432,0x42c,0x432,0x432,0x27,0x27,0x27,0x432,0x432,0x432,0x27,0x435,0x435,
+0x435,0x42f,0x27,0x27,0xfc3,0x27,0x27,0x27,0x27,0x27,0x27,0x42c,0x27,0x27,0x27,0x27,
+0x27,0x27,0x27,0x27,0x27,0x27,0xef7,0x972,0x972,0x972,0x972,0x972,0x972,0x972,0x972,0x972,
+0x96f,0x96f,0x96f,0xdb0,0xcf9,0xcf9,0xcf9,0xcf9,0xcf9,0xcfc,0xcf9,0x27,0x27,0x27,0x27,0x27,
+0x150c,0x44a,0x44a,0x44a,0x194d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x2a,0x44d,0x44d,
+0x44d,0x2a,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,
+0x44d,0x2a,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x44d,0x150f,0x44d,0x44d,0x44d,
+0x44d,0x44d,0x2a,0x2a,0x2a,0xfcc,0x43e,0x43e,0x43e,0x44a,0x44a,0x44a,0x44a,0x2a,0x43e,0x43e,
+0x441,0x2a,0x43e,0x43e,0x43e,0x444,0x2a,0x2a,0x2a,0x2a,0x2a,0x2a,0x2a,0x43e,0x43e,0x2a,
+0xfcc,0xfcc,0x1728,0x2a,0x2a,0x2a,0x2a,0x2a,0x44d,0x44d,0xfc6,0xfc6,0x2a,0x2a,0x447,0x447,
+0x447,0x447,0x447,0x447,0x447,0x447,0x447,0x447,0x2a,0x2a,0x2a,0x2a,0x2a,0x2a,0x2a,0x1a19,
+0xfc9,0xfc9,0xfc9,0xfc9,0xfc9,0xfc9,0xfc9,0xfc9,0x17e5,0x1512,0x456,0x456,0x1950,0x45c,0x45c,0x45c,
+0x45c,0x45c,0x45c,0x45c,0x45c,0x2d,0x45c,0x45c,0x45c,0x2d,0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,
+0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,0x2d,0x45c,0x45c,0x45c,0x45c,0x45c,0x45c,
+0x45c,0x45c,0x45c,0x45c,0x2d,0x45c,0x45c,0x45c,0x45c,0x45c,0x2d,0x2d,0xcff,0xd02,0x456,0x450,
+0x459,0x456,0x450,0x456,0x456,0x2d,0x450,0x459,0x459,0x2d,0x459,0x459,0x450,0x453,0x2d,0x2d,
+0x2d,0x2d,0x2d,0x2d,0x2d,0x450,0x450,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x45c,0x2d,
+0x45c,0x45c,0xf0f,0xf0f,0x2d,0x2d,0x975,0x975,0x975,0x975,0x975,0x975,0x975,0x975,0x975,0x975,
+0x2d,0xf12,0xf12,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,0x2d,
+0x18ab,0x1515,0x468,0x468,0x1ab8,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x30,0x46e,0x46e,
+0x46e,0x30,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,
+0x468,0x45f,0x45f,0x45f,0xfcf,0x30,0x468,0x468,0x468,0x30,0x46b,0x46b,0x46b,0x462,0x134a,0x17e8,
+0x30,0x30,0x30,0x30,0x17eb,0x17eb,0x17eb,0x45f,0x17e8,0x17e8,0x17e8,0x17e8,0x17e8,0x17e8,0x17e8,0x172b,
+0x46e,0x46e,0xfcf,0xfcf,0x30,0x30,0x465,0x465,0x465,0x465,0x465,0x465,0x465,0x465,0x465,0x465,
+0xfd2,0xfd2,0xfd2,0xfd2,0xfd2,0xfd2,0x17e8,0x17e8,0x17e8,0xfd5,0xfd8,0xfd8,0xfd8,0xfd8,0xfd8,0xfd8,
+0x33,0x1abb,0xa3e,0xa3e,0x33,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,
+0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0x33,0x33,0x33,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,
+0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0x33,0xa44,
+0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0xa44,0x33,0xa44,0x33,0x33,0xa44,0xa44,0xa44,0xa44,
+0xa44,0xa44,0xa44,0x33,0x33,0x33,0xa38,0x33,0x33,0x33,0x33,0xa35,0xa3e,0xa3e,0xa35,0xa35,
+0xa35,0x33,0xa35,0x33,0xa3e,0xa3e,0xa41,0xa3e,0xa41,0xa41,0xa41,0xa35,0x33,0x33,0x33,0x33,
+0x33,0x33,0x1518,0x1518,0x1518,0x1518,0x1518,0x1518,0x1518,0x1518,0x1518,0x1518,0x33,0x33,0xa3e,0xa3e,
+0xa3b,0x33,0x33,0x33,0x33,0x33,0x33,0x33,0x33,0x33,0x33,0x33,0x36,0x489,0x489,0x489,
+0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,
+0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x489,0x474,0x489,0x486,
+0x474,0x474,0x474,0x474,0x474,0x474,0x47a,0x36,0x36,0x36,0x36,0x471,0x48f,0x48f,0x48f,0x48f,
+0x48f,0x489,0x48c,0x477,0x477,0x477,0x477,0x477,0x477,0x474,0x477,0x47d,0x483,0x483,0x483,0x483,
+0x483,0x483,0x483,0x483,0x483,0x483,0x480,0x480,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,
+0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,
+0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x36,0x39,0x49e,0x49e,0x39,0x49e,0x39,0x1a1f,0x49e,
+0x49e,0x1a1f,0x49e,0x39,0x1a1f,0x49e,0x1a1f,0x1a1f,0x1a1f,0x1a1f,0x1a1f,0x1a1f,0x49e,0x49e,0x49e,0x49e,
+0x1a1f,0x49e,0x49e,0x49e,0x49e,0x49e,0x49e,0x49e,0x1a1f,0x49e,0x49e,0x49e,0x39,0x49e,0x39,0x49e,
+0x1a1f,0x1a1f,0x49e,0x49e,0x1a1f,0x49e,0x49e,0x49e,0x49e,0x492,0x49e,0x49b,0x492,0x492,0x492,0x492,
+0x492,0x492,0x1a1c,0x492,0x492,0x49e,0x39,0x39,0x4a7,0x4a7,0x4a7,0x4a7,0x4a7,0x39,0x4a4,0x39,
+0x495,0x495,0x495,0x495,0x495,0x492,0x39,0x39,0x498,0x498,0x498,0x498,0x498,0x498,0x498,0x498,
+0x498,0x498,0x39,0x39,0x4a1,0x4a1,0x1425,0x1425,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,
+0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,
+0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x39,0x9b7,0x9b7,0x9b7,0x9ba,0x9b7,0x9b7,0x9b7,0x9b7,
+0x3c,0x9b7,0x9b7,0x9b7,0x9b7,0x9ba,0x9b7,0x9b7,0x9b7,0x9b7,0x9ba,0x9b7,0x9b7,0x9b7,0x9b7,0x9ba,
+0x9b7,0x9b7,0x9b7,0x9b7,0x9ba,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,0x9b7,
+0x9b7,0x9ba,0xa53,0xfe4,0xfe4,0x3c,0x3c,0x3c,0x3c,0x984,0x984,0x987,0x984,0x987,0x987,0x990,
+0x987,0x990,0x984,0x984,0x984,0x984,0x984,0x9b1,0x984,0x987,0x98a,0x98a,0x98d,0x996,0x98a,0x98a,
+0x9b7,0x9b7,0x9b7,0x9b7,0x1353,0x134d,0x134d,0x134d,0x984,0x984,0x984,0x987,0x984,0x984,0xa47,0x984,
+0x3c,0x984,0x984,0x984,0x984,0x987,0x984,0x984,0x984,0x984,0x987,0x984,0x984,0x984,0x984,0x987,
+0x984,0x984,0x984,0x984,0x987,0x984,0xa47,0xa47,0xa47,0x984,0x984,0x984,0x984,0x984,0x984,0x984,
+0xa47,0x987,0xa47,0xa47,0xa47,0x3c,0xa50,0xa50,0xa4d,0xa4d,0xa4d,0xa4d,0xa4d,0xa4d,0xa4a,0xa4d,
+0xa4d,0xa4d,0xa4d,0xa4d,0xa4d,0x3c,0xfdb,0xa4d,0xde3,0xde3,0xfde,0xfe1,0xfdb,0x1161,0x1161,0x1161,
+0x1161,0x1350,0x1350,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,
+0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,0x3c,
+0x3c,0x3c,0x3c,0x3c,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x3f,0x142b,0x3f,0x3f,0x3f,0x3f,
+0x3f,0x142b,0x3f,0x3f,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,
+0x4aa,0x4aa,0x4aa,0x4aa,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,0xa7d,0x42,0xa7d,0xa7d,
+0xa7d,0xa7d,0x42,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0x42,0xa7d,0x42,0xa7d,0xa7d,
+0xa7d,0xa7d,0x42,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,0xa7d,0x42,0xa7d,0xa7d,
+0xa7d,0xa7d,0x42,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,0xa7d,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0x42,0x42,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0x42,0xa7d,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0x42,0x42,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0x42,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,
+0xa7d,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0x42,0x42,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,
+0xa7d,0xa7d,0xa7d,0x42,0x42,0x1356,0x1356,0xdec,0xdef,0xa77,0xa80,0xa74,0xa74,0xa74,0xa74,0xa80,
+0xa80,0xa7a,0xa7a,0xa7a,0xa7a,0xa7a,0xa7a,0xa7a,0xa7a,0xa7a,0xa71,0xa71,0xa71,0xa71,0xa71,0xa71,
+0xa71,0xa71,0xa71,0xa71,0xa71,0x42,0x42,0x42,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,
+0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0x1731,0x45,0x45,
+0x172e,0x172e,0x172e,0x172e,0x172e,0x172e,0x45,0x45,0xa95,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,
+0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,0xa98,
+0xa98,0xa98,0xa98,0xa92,0xa8f,0x48,0x48,0x48,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,
+0xa9e,0xa9e,0xa9e,0xa9b,0xa9b,0xa9b,0xa9e,0xa9e,0xa9e,0x151b,0x151b,0x151b,0x151b,0x151b,0x151b,0x151b,
+0x151b,0x4b,0x4b,0x4b,0x4b,0x4b,0x4b,0x4b,0xabf,0xabf,0xabf,0xabf,0xabf,0xabf,0xaa1,0xabf,
+0xabf,0xaa4,0xaa4,0xaa4,0xaa4,0xaa4,0xaa4,0xaa4,0xaa4,0xaa4,0xaa7,0xaa4,0xab6,0xab6,0xab9,0xac2,
+0xab0,0xaad,0xab6,0xab3,0xac2,0xd05,0x4e,0x4e,0xabc,0xabc,0xabc,0xabc,0xabc,0xabc,0xabc,0xabc,
+0xabc,0xabc,0x4e,0x4e,0x4e,0x4e,0x4e,0x4e,0xd08,0xd08,0xd08,0xd08,0xd08,0xd08,0xd08,0xd08,
+0xd08,0xd08,0x4e,0x4e,0x4e,0x4e,0x4e,0x4e,0xad1,0xad1,0xb52,0xb55,0xad7,0xb4f,0xad4,0xad1,
+0xada,0xae9,0xadd,0xaec,0xaec,0xaec,0xac8,0x51,0xae0,0xae0,0xae0,0xae0,0xae0,0xae0,0xae0,0xae0,
+0xae0,0xae0,0x51,0x51,0x51,0x51,0x51,0x51,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0x1953,0x51,0x51,0x51,0x51,0x51,0x51,0x51,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xae3,0xacb,0x1002,0x51,0x51,0x51,0x51,0x51,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,
+0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x54,0x54,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x54,0x54,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x54,0x4ce,0x54,0x4ce,0x54,0x4ce,0x54,0x4ce,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x54,0x54,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x54,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4c5,0x4cb,0x4c5,0x4c5,0x4c2,0x4cb,0x4cb,0x4cb,0x54,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4c2,0x4c2,0x4c2,0x4cb,0x4cb,0x4cb,0x4cb,0x54,0x54,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x54,0x4c2,0x4c2,0x4c2,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4c2,0x4c2,0x4c2,0x54,0x54,0x4cb,0x4cb,0x4cb,0x54,0x4cb,0x4cb,
+0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4c8,0x4c5,0x54,0xbca,0xbcd,0xbcd,0xbcd,0x100b,0x57,0x14f7,0x14f7,
+0x14f7,0x14f7,0x4d7,0x4d7,0x4d7,0x4d7,0x4d7,0x4d7,0x522,0xbdf,0x5a,0x5a,0x6e1,0x522,0x522,0x522,
+0x522,0x522,0x528,0x53a,0x528,0x534,0x52e,0x6e4,0x51f,0x6de,0x6de,0x6de,0x6de,0x51f,0x51f,0x51f,
+0x51f,0x51f,0x525,0x537,0x525,0x531,0x52b,0x5a,0xdfb,0xdfb,0xdfb,0xdfb,0xdfb,0x1359,0x1359,0x1359,
+0x1359,0x1359,0x1359,0x1359,0x1359,0x5a,0x5a,0x5a,0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,
+0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,0x5d,0x549,0x549,0x549,0x549,0x549,0x549,0x549,0x549,
+0x549,0x549,0x549,0x549,0x549,0x546,0x546,0x546,0x546,0x549,0xafb,0xafe,0xbe5,0xbeb,0xbeb,0xbe8,
+0xbe8,0xbe8,0xbe8,0xe01,0xf15,0xf15,0xf15,0xf15,0x114c,0x60,0x60,0x60,0x60,0x60,0x60,0x60,
+0x60,0x60,0x60,0x60,0x60,0x60,0x60,0x60,0x579,0x579,0x579,0xb07,0xf1e,0x1011,0x1011,0x1011,
+0x1011,0x12ab,0x1737,0x1737,0x63,0x63,0x63,0x63,0x70b,0x70b,0x70b,0x70b,0x70e,0x70e,0x70e,0x70e,
+0x70e,0x70e,0x585,0x585,0x582,0x582,0x582,0x582,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0xb13,0xb13,0x66,
+0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,
+0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x66,0x5af,0x5af,0x5af,0x5af,0x5af,0x5af,0x5af,0x5af,
+0x5af,0x5af,0x5af,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,
+0x69,0x69,0x69,0x69,0x69,0x69,0x69,0x69,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,
+0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,
+0xb2e,0xb2e,0x6c,0xb2e,0xb2e,0xb2e,0xb2e,0xb31,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,
+0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb31,0x6c,0x6c,0x6c,0x6c,
+0x6c,0x6c,0x6c,0x6c,0x6c,0x6c,0x6c,0x6c,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,
+0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0x6f,0x6f,
+0x6f,0x6f,0x6f,0x6f,0x6f,0x6f,0x6f,0x6f,0x75,0x843,0x83d,0x843,0x83d,0x843,0x83d,0x843,
+0x83d,0x843,0x83d,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,
+0x840,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,0x83d,0x83d,0x83d,0x843,0x83d,0x843,0x83d,0x843,
+0x83d,0x83d,0x83d,0x83d,0x83d,0x83d,0x843,0x83d,0x83d,0x83d,0x83d,0x83d,0x840,0xc93,0xc93,0x75,
+0x75,0x954,0x954,0x91e,0x91e,0x846,0x849,0xc90,0x78,0x78,0x78,0x78,0x78,0x85b,0x85b,0x85b,
+0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,
+0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x85b,0x113a,0x191a,0x1a01,0x7b,0x85e,0x85e,0x85e,
+0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x7b,
+0x927,0x927,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,0x92a,
+0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,
+0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0x867,0xd98,0xd98,0x7e,
+0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0x81,0x81,0x81,
+0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,
+0xb4c,0xc9c,0xb4c,0xb4c,0xb4c,0xc9c,0xb4c,0x84,0x84,0x84,0x84,0x84,0x84,0x84,0x84,0x84,
+0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,
+0x9db,0x9db,0x9db,0x9db,0x87,0x87,0x87,0x87,0x87,0x87,0x87,0x87,0x87,0x87,0x87,0x87,
+0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,
+0x60c,0x60c,0x60c,0x60c,0x60c,0x60c,0x60c,0x8a,0x8a,0x8a,0x8a,0x8a,0x8a,0x8a,0x8a,0x8a,
+0x8a,0x8a,0x8a,0x5fa,0x5fa,0x5fa,0x5fa,0x5fa,0x8a,0x8a,0x8a,0x8a,0x8a,0xb1f,0x5fd,0x603,
+0x609,0x609,0x609,0x609,0x609,0x609,0x609,0x609,0x609,0x600,0x603,0x603,0x603,0x603,0x603,0x603,
+0x603,0x603,0x603,0x603,0x603,0x603,0x603,0x8a,0x603,0x603,0x603,0x603,0x603,0x8a,0x603,0x8a,
+0x603,0x603,0x8a,0x603,0x603,0x8a,0x603,0x603,0x603,0x603,0x603,0x603,0x603,0x603,0x603,0x606,
+0x61e,0x618,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,
+0x136b,0x136b,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,
+0x8d,0x8d,0x8d,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x61e,0x618,
+0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,
+0x61b,0x618,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x618,0x61b,0x618,0x618,0x61b,0x61b,0x618,0x618,
+0x618,0x618,0x618,0x61b,0x618,0x618,0x61b,0x618,0x61b,0x61b,0x61b,0x618,0x61b,0x61b,0x61b,0x61b,
+0x8d,0x8d,0x61b,0x61b,0x61b,0x61b,0x618,0x618,0x61b,0x618,0x618,0x618,0x618,0x61b,0x618,0x618,
+0x618,0x618,0x618,0x61b,0x61b,0x61b,0x618,0x618,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,0x8d,
+0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,0xb6a,
+0x61e,0x61e,0x978,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x615,0x615,0xc24,0xdb3,0x8d,0x8d,
+0x87f,0x891,0x88e,0x891,0x88e,0xcb1,0xcb1,0xda4,0xda1,0x882,0x882,0x882,0x882,0x894,0x894,0x894,
+0x8ac,0x8af,0x8be,0x90,0x8b2,0x8b5,0x8c1,0x8c1,0x8a9,0x8a0,0x89a,0x8a0,0x89a,0x8a0,0x89a,0x89d,
+0x89d,0x8b8,0x8b8,0x8bb,0x8b8,0x8b8,0x8b8,0x90,0x8b8,0x8a6,0x8a3,0x89d,0x90,0x90,0x90,0x90,
+0x62a,0x636,0x62a,0xc27,0x62a,0x93,0x62a,0x636,0x62a,0x636,0x62a,0x636,0x62a,0x636,0x62a,0x636,
+0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x633,
+0x62d,0x633,0x62d,0x630,0x636,0x633,0x62d,0x633,0x62d,0x633,0x62d,0x633,0x62d,0x93,0x93,0x627,
+0x77d,0x780,0x795,0x798,0x777,0x780,0x780,0x99,0x75f,0x762,0x762,0x762,0x762,0x75f,0x75f,0x99,
+0x96,0x96,0x96,0x96,0x96,0x96,0x96,0x96,0x96,0xb22,0xb22,0xb22,0x9de,0x759,0x639,0x639,
+0x99,0x7a7,0x786,0x777,0x780,0x77d,0x777,0x789,0x77a,0x774,0x777,0x795,0x78c,0x783,0x7a4,0x777,
+0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x7a1,0x792,0x78f,0x795,0x795,0x795,0x7a7,
+0x768,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,
+0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x765,0x99,
+0x99,0x99,0x765,0x765,0x765,0x765,0x765,0x765,0x99,0x99,0x765,0x765,0x765,0x765,0x765,0x765,
+0x99,0x99,0x765,0x765,0x765,0x765,0x765,0x765,0x99,0x99,0x765,0x765,0x765,0x99,0x99,0x99,
+0xb6d,0xb6d,0xb6d,0xb6d,0x9c,0x9c,0x9c,0x9c,0x9c,0x9c,0x9c,0x9c,0x9c,0x18b7,0x18b7,0x18b7,
+0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,0xb73,
+0xb73,0xb73,0xb73,0x9f,0x9f,0x9f,0x9f,0x9f,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,
+0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,
+0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xa2,0xa2,
+0xa2,0xa2,0xa2,0xa2,0xa2,0xa2,0xa2,0xa2,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xa5,
+0xa5,0x101d,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0x173d,0x173d,0x173d,0x173d,0x173d,0x173d,0x173d,0x173d,
+0x173d,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,
+0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xa5,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xa8,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xa8,0xba0,0xba0,0xa8,0xa8,0xba0,0xa8,0xa8,0xba0,0xba0,0xa8,
+0xa8,0xba0,0xba0,0xba0,0xba0,0xa8,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xb9d,0xb9d,
+0xb9d,0xb9d,0xa8,0xb9d,0xa8,0xb9d,0xb9d,0xb9d,0xb9d,0xd29,0xb9d,0xb9d,0xa8,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xa8,0xba0,
+0xba0,0xba0,0xba0,0xa8,0xa8,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xa8,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xa8,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xba0,0xba0,0xa8,0xba0,0xba0,0xba0,0xba0,0xa8,0xba0,0xba0,0xba0,0xba0,0xba0,0xa8,0xba0,0xa8,
+0xa8,0xa8,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xa8,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xe16,0xe16,0xa8,0xa8,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb97,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xf2d,0xf2a,
+0xa8,0xa8,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,
+0xb9a,0xb9a,0xb9a,0xb9a,0xab,0xba6,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,
+0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,0xab,
+0xab,0xab,0xab,0xab,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,0xc36,
+0xc36,0xae,0xc36,0xc36,0xc36,0xc36,0xc30,0xc30,0xc33,0xae,0xae,0xae,0xae,0xae,0xae,0xae,
+0xae,0xae,0xae,0xae,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,
+0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc3f,0xc39,0xc39,0xc3c,0xca5,0xca5,0xb1,0xb1,0xb1,0xb1,0xb1,
+0xb1,0xb1,0xb1,0xb1,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,
+0xc45,0xc45,0xc45,0xc45,0xc45,0xc45,0xc42,0xc42,0xb4,0xb4,0xb4,0xb4,0xb4,0xb4,0xb4,0xb4,
+0xb4,0xb4,0xb4,0xb4,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,0xc4b,
+0xc4b,0xb7,0xc4b,0xc4b,0xc4b,0xb7,0xc48,0xc48,0xb7,0xb7,0xb7,0xb7,0xb7,0xb7,0xb7,0xb7,
+0xb7,0xb7,0xb7,0xb7,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,
+0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,0xd3b,
+0xd3b,0x1539,0x1539,0xba,0xd2c,0xd2c,0xd2c,0xd38,0xd38,0xd38,0xd38,0xd2c,0xd2c,0xd38,0xd38,0xd38,
+0xba,0xba,0xba,0xba,0xd38,0xd38,0xd2c,0xd38,0xd38,0xd38,0xd38,0xd38,0xd38,0xd2f,0xd2f,0xd2f,
+0xba,0xba,0xba,0xba,0xd32,0xba,0xba,0xba,0xd3e,0xd3e,0xd35,0xd35,0xd35,0xd35,0xd35,0xd35,
+0xd35,0xd35,0xd35,0xd35,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,
+0xd41,0xd41,0xd41,0xd41,0xd41,0xd41,0xbd,0xbd,0xd41,0xd41,0xd41,0xd41,0xd41,0xbd,0xbd,0xbd,
+0xbd,0xbd,0xbd,0xbd,0xbd,0xbd,0xbd,0xbd,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0xc0,0xc0,0x153c,0x153c,
+0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0xc0,0x1abe,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xc3,0xd68,0xd68,0xd68,
+0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,
+0xd68,0xd68,0xd68,0xc3,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,
+0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xc3,0xd68,0xd68,0xc3,0xd68,0xd68,0xd68,0xd68,0xd68,
+0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xc3,0xc3,0xd68,0xd68,0xd68,0xd68,
+0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xd68,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,
+0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,
+0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xc3,0xd6b,0xd6b,0xd6b,0xd6b,
+0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,
+0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xc6,0xc6,0xc6,0xc6,0xc6,0xdad,0xdad,0xdad,0xc9,
+0xc9,0xc9,0xc9,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,
+0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xda7,0xc9,0xc9,0xc9,0xdaa,
+0xdaa,0xdaa,0xdaa,0xdaa,0xdaa,0xdaa,0xdaa,0xdaa,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,
+0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,
+0xd71,0xd71,0xd71,0xd71,0xd71,0xd71,0xcc,0xd6e,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,
+0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,
+0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xd7a,0xcf,0xcf,0xd77,0xd77,0xd77,0xd77,0xd77,0xd77,0xd77,0xd77,
+0xd77,0xd77,0xcf,0xcf,0xcf,0xcf,0xcf,0xcf,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,
+0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd2,0xd2,
+0xd7d,0xd2,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,
+0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd7d,0xd2,0xd7d,0xd7d,0xd2,0xd2,0xd2,
+0xd7d,0xd2,0xd2,0xd7d,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,
+0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd5,0xd5,0xd5,0xd5,0xd5,
+0xd5,0xd5,0xd5,0xd5,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0x153f,
+0x153f,0x17f1,0x17f1,0xdb,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,0x1119,
+0x1acd,0x132,0x132,0x132,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,
+0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe43,0xe3a,0xe3a,0xe40,0xe40,0xe3a,
+0xde,0xde,0xe3d,0xe3d,0x1149,0x1149,0x1149,0x1149,0xe1,0xe1,0xe1,0xe1,0xe1,0xe1,0xe1,0xe1,
+0xe1,0xe1,0xe1,0xe1,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,0xca2,
+0xca2,0xca2,0xca2,0xca2,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1542,0x1542,0x1542,0x1542,0x1542,
+0x1542,0x1542,0x1542,0x1542,0x1542,0x1542,0x1542,0x1542,0x1542,0x1545,0x18bd,0x18bd,0x193e,0x18bd,0xe4,0x17f4,
+0x1377,0x118b,0xf3c,0xf3c,0xe55,0xe52,0xe55,0xe52,0xe52,0xe49,0xe49,0xe49,0xe49,0xe49,0xe49,0x1194,
+0x1191,0x1194,0x1191,0x118e,0x118e,0x118e,0x1434,0x1431,0xe7,0xe7,0xe7,0xe7,0xe7,0xe4f,0xe4c,0xe4c,
+0xe4c,0xe49,0xe4f,0xe4c,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,
+0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xea,0xea,0xea,0xea,0xea,
+0xea,0xea,0xea,0xea,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xea,0xe58,0xe58,0xe58,0xe58,
+0xe58,0xe58,0xe58,0xea,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xe58,0xea,0xe58,0xe58,0xe58,0xe58,
+0xe58,0xe58,0xe58,0xea,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,0xe5e,
+0xe5e,0xe5e,0xe5e,0xe5e,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xe5b,0xed,0xed,
+0xed,0xed,0xed,0xed,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xf0,0x1437,0xf0,0xf0,0xf0,0xf0,
+0xf0,0x1437,0xf0,0xf0,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,
+0xeb8,0xeb8,0xeb8,0xeb8,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,
+0xe67,0xe67,0xe67,0xf3,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,
+0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,0xe64,
+0xe64,0xe64,0xe64,0xf3,0xe79,0xe6d,0xe6d,0xe6d,0xf6,0xe6d,0xe6d,0xf6,0xf6,0xf6,0xf6,0xf6,
+0xe6d,0xe6d,0xe6d,0xe6d,0xe79,0xe79,0xe79,0xe79,0xf6,0xe79,0xe79,0xe79,0xf6,0xe79,0xe79,0xe79,
+0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,0xe79,
+0xe79,0xe79,0xe79,0xe79,0x195c,0x195c,0xf6,0xf6,0xe6a,0xe6a,0xe6a,0xf6,0xf6,0xf6,0xf6,0xe70,
+0xe73,0xe73,0xe73,0xe73,0xe73,0xe73,0xe73,0xe73,0x1959,0xf6,0xf6,0xf6,0xf6,0xf6,0xf6,0xf6,
+0xe76,0xe76,0xe76,0xe76,0xe76,0xe76,0xe7c,0xe7c,0xe73,0xf6,0xf6,0xf6,0xf6,0xf6,0xf6,0xf6,
+0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0x119a,0x119a,0xf9,0xf9,0xf9,0xf9,
+0xe88,0xe88,0xe88,0xe88,0xe88,0xe8b,0xe8b,0xe8b,0xe88,0xe88,0xe8b,0xe88,0xe88,0xe88,0xe88,0xe88,
+0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xf9,0xf9,0xf9,0xf9,0xf9,0xf9,0xe85,0xe85,0xe85,0xe85,
+0xe85,0xe85,0xe85,0xe85,0xe85,0xe85,0x1197,0xf9,0xf9,0xf9,0xe82,0xe82,0xe91,0xe91,0xe91,0xe91,
+0xfc,0xfc,0xfc,0xfc,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe8e,0xe91,0xe91,0xe91,
+0xe91,0xe91,0xfc,0xfc,0xfc,0xfc,0xfc,0xfc,0xfc,0xfc,0xfc,0xfc,0x154e,0x1554,0x1551,0x189c,
+0x17f7,0x18c0,0x18c0,0x18c0,0x18c0,0x18c0,0x1962,0x195f,0x1965,0x195f,0x1965,0x1a25,0x1ac1,0x1ac1,0x1ac1,0xff,
+0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
+0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
+0xeb5,0xeb5,0xeb5,0xeb2,0xeb2,0xea9,0xea9,0xeb2,0xeaf,0xeaf,0xeaf,0xeaf,0x1ac4,0x102,0x102,0x102,
+0x1314,0x1314,0x1314,0x1317,0x1317,0x1317,0x130e,0x130e,0x1311,0x130e,0x156,0x156,0x156,0x156,0x156,0x156,
+0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0x1443,0x1443,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0xebb,
+0x137d,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x105,0x137a,
+0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc72,0xc75,
+0xee8,0xed9,0xed3,0xee5,0xee2,0xedc,0xedc,0xeeb,0xed6,0xedf,0x108,0x108,0x108,0x108,0x108,0x108,
+0xf6f,0xf6f,0xf5a,0xf6f,0xf72,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0x10e,0x10e,0x10e,0x10e,
+0xf69,0xf69,0xf69,0xf69,0xf69,0xf69,0xf69,0xf69,0xf69,0xf69,0xf7b,0xf7b,0xf60,0xf66,0xf7b,0xf7b,
+0xf63,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf5d,0xf5d,0xf5d,0xf5d,0xf5d,
+0xf5d,0xf5d,0xf5d,0xf5d,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0xf60,0x10e,0x10e,0x10e,
+0x111,0x111,0x1a2b,0x1a28,0x1a2b,0x1a2b,0x1a2b,0x1aca,0x1ac7,0x1aca,0x1ac7,0x111,0x111,0x111,0x111,0x111,
+0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,0x111,
+0x111,0x1aca,0x1ac7,0x155d,0x144c,0x144c,0x1380,0x1074,0x1074,0x1074,0x1074,0x1074,0xf8a,0xf8a,0xf8a,0xf8a,
+0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,
+0xf87,0xf87,0xf8d,0xf8d,0x114,0x114,0x114,0x114,0x114,0x114,0x114,0x114,0xf96,0xf96,0xf96,0xf96,
+0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,0xf96,
+0xf96,0xf96,0xf90,0xf90,0xf90,0xf90,0x11a3,0x11a3,0x117,0x117,0x117,0xf93,0x1563,0x1563,0x1563,0x1563,
+0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,
+0x1563,0x1563,0x1563,0x1563,0x1563,0x174c,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,
+0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,
+0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0x11a,0xf9f,0xf9f,0xf9f,0x1569,0x1569,0x1569,0x1569,0x1569,
+0x1569,0x1569,0x1569,0x1569,0x1569,0x1569,0x1569,0x11d,0xf9c,0xf9c,0xf9c,0xf9c,0x1566,0x11d,0x11d,0x11d,
+0x11d,0x11d,0x11d,0x11d,0x11d,0x11d,0x11d,0x11d,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,
+0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0xfa2,0x1974,0x1974,0x1974,0x1974,0x1974,0x1974,
+0x1974,0x120,0x120,0x120,0x120,0x120,0x120,0x120,0x109b,0x109b,0x109b,0x109b,0x1098,0x1098,0x1098,0x1098,
+0x1098,0x1098,0x1098,0x1098,0x1089,0x1089,0x1089,0x1089,0x1089,0x1089,0x1089,0x1089,0x1098,0x1098,0x108f,0x108c,
+0x123,0x123,0x123,0x109e,0x109e,0x1092,0x1092,0x1092,0x1095,0x1095,0x1095,0x1095,0x1095,0x1095,0x1095,0x1095,
+0x1095,0x1095,0x123,0x123,0x123,0x109b,0x109b,0x109b,0x10a1,0x10a1,0x10a1,0x10a1,0x10a1,0x10a1,0x10a1,0x10a1,
+0x10a1,0x10a1,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10b6,0x10b6,0x10b6,0x10b6,0x10b6,0x10b6,0x10b6,0x10b6,
+0x10b6,0x10b6,0x10b9,0x10b9,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,
+0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x126,0x10e0,0x10e0,0x10e0,0x10e0,0x10da,0x17fd,0x129,0x129,
+0x129,0x129,0x129,0x129,0x129,0x129,0x10e6,0x10e6,0x10dd,0x10dd,0x10dd,0x10dd,0x10dd,0x10dd,0x10dd,0x10dd,
+0x10dd,0x10dd,0x129,0x129,0x129,0x129,0x129,0x129,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x10f8,
+0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10f8,0x10fe,0x1101,0x12c,0x12c,0x12c,0x12c,
+0x12c,0x12c,0x12c,0x12c,0x12c,0x12c,0x12c,0x10fb,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,
+0x1113,0x1107,0x1107,0x1107,0x1107,0x1107,0x1107,0x1110,0x1110,0x1107,0x1107,0x1110,0x1110,0x1107,0x1107,0x12f,
+0x12f,0x12f,0x12f,0x12f,0x12f,0x12f,0x12f,0x12f,0x1113,0x1113,0x1113,0x1107,0x1113,0x1113,0x1113,0x1113,
+0x1113,0x1113,0x1113,0x1113,0x1107,0x1110,0x12f,0x12f,0x110d,0x110d,0x110d,0x110d,0x110d,0x110d,0x110d,0x110d,
+0x110d,0x110d,0x12f,0x12f,0x110a,0x1116,0x1116,0x1116,0x1575,0x132,0x132,0x132,0x132,0x132,0x132,0x132,
+0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,
+0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x132,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,
+0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,0x111c,
+0x111c,0x111c,0x111c,0x111c,0x111c,0x111f,0x135,0x135,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,
+0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,0x1122,
+0x1122,0x1122,0x1122,0x1122,0x1122,0x138,0x138,0x138,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,
+0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x13b,0x13b,0x13b,0x13b,0x13b,0x13b,0x13b,
+0x13b,0x13b,0x13b,0x13b,0x13b,0x13b,0x13b,0x13b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,
+0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,0x112b,
+0x112b,0x112b,0x13e,0x13e,0x13e,0x13e,0x13e,0x1128,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,
+0x112e,0x112e,0x112e,0x112e,0x141,0x141,0x141,0x141,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,
+0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x144,0x144,0x144,0x144,
+0x144,0x144,0x144,0x144,0x144,0x144,0x144,0x144,0x11a9,0x11a9,0x11a9,0x11a9,0x11b2,0x11a9,0x11a9,0x11a9,
+0x11b2,0x11a9,0x11a9,0x11a9,0x11a9,0x11a6,0x147,0x147,0x11af,0x11af,0x11af,0x11af,0x11af,0x11af,0x11af,0x11b5,
+0x11af,0x11b5,0x11af,0x11af,0x11af,0x11b5,0x11b5,0x147,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,
+0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x14a,0x14a,
+0x14a,0x14a,0x14a,0x14a,0x14a,0x14a,0x14a,0x14a,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,
+0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d0,0x11bb,0x11d0,
+0x11bb,0x11bb,0x11bb,0x11bb,0x11bb,0x11bb,0x11bb,0x14d,0x11c4,0x11cd,0x11bb,0x11cd,0x11cd,0x11bb,0x11bb,0x11bb,
+0x11bb,0x11bb,0x11bb,0x11bb,0x11bb,0x11d0,0x11d0,0x11d0,0x11d0,0x11d0,0x11d0,0x11bb,0x11bb,0x11c1,0x11c1,0x11c1,
+0x11c1,0x11c1,0x11c1,0x11c1,0x11c1,0x14d,0x14d,0x11be,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,
+0x11ca,0x11ca,0x14d,0x14d,0x14d,0x14d,0x14d,0x14d,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,0x11ca,
+0x11ca,0x11ca,0x14d,0x14d,0x14d,0x14d,0x14d,0x14d,0x11c7,0x11c7,0x11c7,0x11c7,0x11c7,0x11c7,0x11c7,0x11d6,
+0x11d9,0x11d9,0x11d9,0x11d9,0x11c7,0x11c7,0x14d,0x14d,0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,
+0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,0x15c0,0x15bd,0x1adf,0x1329,0x1302,0x1320,0x1320,0x1320,0x1320,0x1320,0x1320,
+0x1320,0x1308,0x1305,0x12fc,0x12fc,0x1326,0x12fc,0x12fc,0x12fc,0x12fc,0x130b,0x14eb,0x14f1,0x14ee,0x14ee,0x193b,
+0x1716,0x1716,0x1aac,0x150,0x150,0x150,0x150,0x150,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,
+0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11e5,0x11e5,0x11e8,0x11f1,0x11eb,0x11eb,0x11eb,0x11f1,
+0x153,0x153,0x153,0x153,0x153,0x153,0x153,0x153,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,
+0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,
+0x12ea,0x12ea,0x12ea,0x12ea,0x12ea,0x159,0x159,0x159,0x120f,0x1203,0x1203,0x1203,0x1203,0x1203,0x1203,0x1206,
+0x1215,0x1215,0x1203,0x1203,0x1203,0x1203,0x15c,0x131a,0x1209,0x1209,0x1209,0x1209,0x1209,0x1209,0x1209,0x1209,
+0x1209,0x1209,0x15c,0x15c,0x15c,0x15c,0x1203,0x1203,0x1233,0x1227,0x1233,0x15f,0x15f,0x15f,0x15f,0x15f,
+0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,0x15f,
+0x15f,0x15f,0x15f,0x1230,0x1230,0x1236,0x122a,0x122d,0x124b,0x124b,0x124b,0x1245,0x1245,0x123c,0x1245,0x1245,
+0x123c,0x1245,0x1245,0x124e,0x1248,0x123f,0x162,0x162,0x1242,0x1242,0x1242,0x1242,0x1242,0x1242,0x1242,0x1242,
+0x1242,0x1242,0x162,0x162,0x162,0x162,0x162,0x162,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x1254,0x165,
+0x165,0x165,0x165,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,
+0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,0x1251,
+0x165,0x165,0x165,0x165,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,
+0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x125d,0x168,0x125a,0x1257,0x1257,0x1257,0x1257,
+0x1257,0x1257,0x1257,0x1257,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,
+0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x16b,0x16b,0x16b,0x1266,0x1269,0x1269,
+0x1269,0x1269,0x1269,0x1269,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,
+0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x1272,0x16e,0x16e,0x126f,0x126f,0x126f,0x126f,
+0x126f,0x126f,0x126f,0x126f,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,
+0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x1278,0x171,0x171,0x171,0x171,0x171,0x1275,0x1275,0x1275,0x1275,
+0x1275,0x1275,0x1275,0x1275,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,
+0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,0x127e,
+0x127e,0x127e,0x127e,0x177,0x1299,0x1299,0x17a,0x17a,0x17a,0x17a,0x17a,0x17a,0x17a,0x17a,0x17a,0x17a,
+0x17a,0x197d,0x17a,0x17a,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,
+0x14ca,0x14ca,0x14ca,0x14ca,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,
+0x129f,0x129f,0x129f,0x17d,0x1a4f,0x1a4f,0x1a4f,0x1a4f,0x1a4f,0x1a4f,0x1a4f,0x1a52,0x1a4c,0x279,0x279,0x279,
+0x279,0x279,0x279,0x279,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,0x187e,
+0x187e,0x1ad0,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,
+0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,0x180,
+0x180,0x180,0x180,0x180,0x180,0x180,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,
+0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,0x138c,
+0x12f6,0x13f5,0x13f2,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,
+0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f3,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,
+0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f0,0x12f3,0x12f0,0x12f0,0x13f5,0x13f5,
+0x13f5,0x13f5,0x13f5,0x13f2,0x13f5,0x13f5,0x13f5,0x1881,0x183,0x183,0x183,0x183,0x12ed,0x12ed,0x12ed,0x12ed,
+0x12ed,0x12ed,0x12ed,0x12ed,0x12ed,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x1419,0x1419,0x183,0x183,
+0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x1920,0x1920,0x1920,0x1920,
+0x1920,0x1920,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,
+0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,0x183,
+0x183,0x183,0x183,0x183,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,
+0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x1395,0x138f,0x138f,0x138f,
+0x186,0x186,0x1392,0x186,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x1398,0x13a1,0x139b,0x139b,0x13a1,0x13a1,
+0x13a1,0x139b,0x13a1,0x139b,0x139b,0x139b,0x13a4,0x13a4,0x189,0x189,0x189,0x189,0x189,0x189,0x189,0x189,
+0x139e,0x139e,0x139e,0x139e,0x18c,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x18c,0x18c,0x13aa,0x13aa,0x13aa,
+0x13aa,0x13aa,0x13aa,0x18c,0x18c,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x18c,0x18c,0x18c,0x18c,0x18c,
+0x18c,0x18c,0x18c,0x18c,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x13aa,0x18c,0x13aa,0x13aa,0x13aa,0x13aa,
+0x13aa,0x13aa,0x13aa,0x18c,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,
+0x161d,0x161d,0x161d,0x161d,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13b0,0x13c2,0x13c2,0x13b6,0x13b6,0x13b6,
+0x13b6,0x13b6,0x18f,0x18f,0x18f,0x18f,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,
+0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b3,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,0x13b9,
+0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x18f,0x1584,
+0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,
+0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x192,0x192,0x192,0x192,0x192,0x192,0x192,
+0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x195,
+0x195,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x1587,
+0x195,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13fe,
+0x195,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,0x13c8,
+0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,
+0x1587,0x1587,0x1587,0x1587,0x1587,0x1587,0x195,0x195,0x195,0x195,0x195,0x195,0x195,0x195,0x195,0x195,
+0x1413,0x1410,0x1410,0x1410,0x1410,0x1410,0x159c,0x159c,0x159c,0x159c,0x159c,0x159f,0x170d,0x159f,0x159f,0x159f,
+0x17d9,0x188a,0x188a,0x18c3,0x18c3,0x1a8e,0x1b39,0x1b39,0x198,0x198,0x198,0x198,0x198,0x198,0x198,0x198,
+0x159f,0x159f,0x159f,0x159f,0x159f,0x159f,0x159c,0x159c,0x159c,0x159f,0x159c,0x170a,0x170a,0x198,0x198,0x198,
+0x159f,0x159c,0x159c,0x159f,0x188a,0x188a,0x188a,0x1926,0x1926,0x1a07,0x1a8e,0x1b39,0x1b39,0x198,0x198,0x198,
+0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,
+0x13cb,0x13cb,0x13cb,0x13cb,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,0x19b,
+0x1467,0x15a5,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x1467,0x15a5,0x15a5,0x15a5,
+0x15a5,0x15a5,0x15a5,0x175e,0x175e,0x19e,0x1809,0x1809,0x1809,0x1809,0x1809,0x1809,0x1809,0x1809,0x1ad3,0x1ad3,
+0x1ad3,0x1ad3,0x1ad3,0x1ad3,0x1ad3,0x1ad3,0x1ad3,0x1ad3,0x19e,0x19e,0x19e,0x19e,0x19e,0x19e,0x19e,0x19e,
+0x19e,0x19e,0x19e,0x1983,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,0x1806,
+0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,
+0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,
+0x1a1,0x146d,0x146d,0x1a1,0x146d,0x1a1,0x1a1,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,
+0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x1a1,0x1a1,0x1a1,
+0x1a1,0x1a1,0x146d,0x1a1,0x1a1,0x1a1,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x146d,0x146d,0x146d,
+0x1a1,0x146d,0x146d,0x1a1,0x146d,0x1a1,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x146d,0x1a1,0x146d,
+0x1a1,0x146d,0x146d,0x1a1,0x146d,0x1a1,0x1a1,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,
+0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x1a1,
+0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x146d,
+0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x1a1,0x1a1,0x1a1,0x1a1,
+0x1a1,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x146d,0x1a1,0x146d,0x146d,0x146d,0x146d,0x146d,
+0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x146d,0x1a1,0x1a1,0x1a1,0x1a1,
+0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,
+0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x146a,0x146a,0x1a1,0x1a1,
+0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1a1,0x1482,0x1482,0x1482,0x1482,
+0x1482,0x1482,0x1482,0x1470,0x1470,0x1470,0x1470,0x1470,0x147f,0x1470,0x1473,0x1473,0x1470,0x1470,0x1470,0x1476,
+0x1476,0x1a4,0x147c,0x147c,0x147c,0x147c,0x147c,0x147c,0x147c,0x147c,0x147c,0x147c,0x1479,0x1485,0x1485,0x1485,
+0x1989,0x1986,0x1986,0x1ad6,0x1a4,0x1a4,0x1a4,0x1a4,0x1a4,0x1a4,0x1a4,0x1a4,0x162f,0x162f,0x162f,0x162f,
+0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x1491,0x1491,0x1491,0x1491,
+0x1491,0x1491,0x1491,0x1491,0x1491,0x1491,0x1491,0x148e,0x1488,0x1488,0x148e,0x148e,0x1497,0x1497,0x1491,0x1494,
+0x1494,0x148e,0x148b,0x1a7,0x1a7,0x1a7,0x1a7,0x1a7,0x1a7,0x1a7,0x1a7,0x1a7,0x149a,0x149a,0x149a,0x149a,
+0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,0x149a,
+0x149a,0x149a,0x149a,0x149a,0x1aa,0x1aa,0x1aa,0x1aa,0x1761,0x1761,0x149a,0x149a,0x1761,0x1761,0x1761,0x1761,
+0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1aa,0x1aa,0x1761,0x1761,
+0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x14a6,0x14a6,0x14a6,0x14a6,
+0x14a6,0x1a37,0x1a37,0x1a37,0x1a37,0x1a37,0x1a37,0x1ad,0x1ad,0x1ad,0x1ad,0x1a31,0x14a6,0x14a3,0x14a3,0x14a3,
+0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x1a34,0x1a34,0x1a34,0x1a34,
+0x1a34,0x1a34,0x1a34,0x1a34,0x1ad,0x1ad,0x1ad,0x1ad,0x1ad,0x1ad,0x1ad,0x14a0,0x14a0,0x14a0,0x14a0,0x14a9,
+0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14a9,0x14ca,0x14ca,0x14ca,0x14ca,
+0x14ca,0x14ca,0x14ca,0x14ca,0x14ca,0x1b0,0x1b0,0x1b0,0x1b0,0x1b0,0x1b0,0x1b0,0x14c7,0x14c7,0x14c7,0x14c7,
+0x14c7,0x14c7,0x14c7,0x14c7,0x14c7,0x14c7,0x1b0,0x1b0,0x1b0,0x1b0,0x1b0,0x1b0,0x14cd,0x14cd,0x14cd,0x14cd,
+0x14cd,0x14cd,0x14cd,0x14cd,0x1b3,0x1b3,0x1b3,0x1b3,0x1b3,0x1b3,0x1b3,0x1b3,0x1323,0x1320,0x1323,0x12ff,
+0x1320,0x1326,0x1326,0x1329,0x1326,0x1329,0x132c,0x1320,0x1329,0x1329,0x1320,0x1320,0x14df,0x14df,0x14df,0x14df,
+0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14d0,0x14d9,0x14d0,0x14d9,0x14d9,0x14d0,0x14d0,0x14d0,0x14d0,
+0x14d0,0x14d0,0x14dc,0x14d3,0x1a3a,0x1b6,0x1b6,0x1b6,0x1b6,0x1b6,0x1b6,0x1b6,0x15b1,0x15b1,0x15b1,0x15b1,
+0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x15b1,0x1b9,0x1b9,0x15ae,0x15ae,0x15ae,0x15ae,
+0x15ae,0x15b4,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1b9,0x1719,0x1710,0x1710,0x1710,
+0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,
+0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1710,0x1bf,0x1bf,0x1bf,0x1bf,0x1adf,0x1c2,0x1c2,0x1c2,
+0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,
+0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,0x1c2,
+0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x1c5,0x1c5,0x1c5,0x1c5,0x1c5,
+0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x1c5,0x1c5,0x1c5,
+0x1c5,0x1c5,0x1c5,0x1c5,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x1c5,0x1c5,
+0x15c9,0x15c3,0x15c6,0x15cf,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x1c8,0x1c8,0x1c8,0x1c8,
+0x1c8,0x1c8,0x1c8,0x1c8,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,
+0x15ba,0x15ba,0x15ba,0x15ba,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,
+0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x198c,0x198c,0x198c,0x198c,0x1cb,0x1cb,0x1cb,
+0x1cb,0x1cb,0x1cb,0x1cb,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,0x1a91,
+0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,0x1cb,
+0x1cb,0x1cb,0x1cb,0x1cb,0x1776,0x171c,0x15de,0x1722,0x1ce,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,
+0x15e7,0x1ce,0x1ce,0x15e7,0x15e7,0x1ce,0x1ce,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,
+0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x1ce,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x1ce,0x15e7,0x15e7,
+0x1ce,0x15e7,0x15e7,0x15e7,0x15e7,0x15e7,0x1ce,0x1a16,0x171f,0x15e7,0x15d8,0x15de,0x15d8,0x15de,0x15de,0x15de,
+0x15de,0x1ce,0x1ce,0x15de,0x15de,0x1ce,0x1ce,0x15e1,0x15e1,0x15e4,0x1ce,0x1ce,0x1779,0x1ce,0x1ce,0x1ce,
+0x1ce,0x1ce,0x1ce,0x15d8,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x15ea,0x15e7,0x15e7,0x15e7,0x15e7,0x15de,0x15de,
+0x1ce,0x1ce,0x15db,0x15db,0x15db,0x15db,0x15db,0x15db,0x15db,0x1ce,0x1ce,0x1ce,0x15db,0x15db,0x15db,0x15db,
+0x15db,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x1ce,0x15ff,0x15ff,0x15ff,0x15ff,
+0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x1d1,0x15ff,
+0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15ff,0x15f9,0x15f9,0x15f9,0x15ed,
+0x15ed,0x15ed,0x15f9,0x15f9,0x15ed,0x15fc,0x15f0,0x15ed,0x1602,0x1602,0x15f6,0x1602,0x1602,0x15f3,0x180c,0x1d1,
+0x1611,0x1611,0x1611,0x1605,0x1605,0x1605,0x1605,0x1605,0x1605,0x1608,0x160b,0x1d4,0x1d4,0x1d4,0x1d4,0x1d4,
+0x160e,0x160e,0x160e,0x160e,0x160e,0x160e,0x160e,0x160e,0x160e,0x160e,0x1d4,0x1d4,0x1d4,0x1d4,0x1d4,0x1d4,
+0x177c,0x177c,0x177c,0x177c,0x161d,0x161a,0x1a3d,0x1a3d,0x1ae5,0x1ae8,0x1ae2,0x1ae2,0x1d7,0x1d7,0x1d7,0x1d7,
+0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,
+0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,0x1da,
+0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,0x162f,
+0x162f,0x162f,0x162f,0x1626,0x1629,0x162c,0x162f,0x1dd,0x1dd,0x1dd,0x1dd,0x1dd,0x1dd,0x1dd,0x1dd,0x1dd,
+0x163e,0x163e,0x163e,0x163e,0x163e,0x1632,0x1632,0x1e0,0x1e0,0x1e0,0x1e0,0x1635,0x1635,0x1635,0x1635,0x1635,
+0x163b,0x163b,0x163b,0x163b,0x163b,0x163b,0x1638,0x1e0,0x1e0,0x1e0,0x1e0,0x1e0,0x1e0,0x1e0,0x1e0,0x1e0,
+0x1647,0x1647,0x1647,0x1647,0x1647,0x1e3,0x1e3,0x1644,0x1644,0x1644,0x1644,0x1644,0x1644,0x1644,0x1644,0x1644,
+0x1641,0x1641,0x1641,0x1641,0x1641,0x1641,0x1641,0x1e3,0x1e3,0x1e3,0x1e3,0x1e3,0x1e3,0x1e3,0x1e3,0x1e3,
+0x164a,0x165c,0x165c,0x1650,0x1659,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,
+0x1653,0x1653,0x1653,0x1653,0x1653,0x1653,0x1653,0x1653,0x1653,0x1653,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,0x1e6,
+0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,
+0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1662,0x1e9,
+0x166e,0x166e,0x166e,0x166e,0x166e,0x1668,0x1671,0x166e,0x166e,0x166e,0x166e,0x166e,0x166e,0x166e,0x166e,0x166e,
+0x166b,0x166b,0x166b,0x166b,0x166b,0x166b,0x166b,0x166b,0x166b,0x166b,0x166e,0x166e,0x166e,0x166e,0x166e,0x1ec,
+0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,
+0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1677,0x1ef,
+0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,
+0x1683,0x1683,0x1683,0x1683,0x1683,0x1683,0x1680,0x1680,0x1680,0x1680,0x1680,0x1f2,0x1f2,0x1f2,0x1f2,0x1f2,
+0x169b,0x169b,0x169e,0x169e,0x16a1,0x1692,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,
+0x1698,0x1698,0x1698,0x1698,0x1698,0x1698,0x1698,0x1698,0x1698,0x1698,0x1f5,0x1692,0x1692,0x1692,0x1692,0x1692,
+0x1692,0x1692,0x1f5,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,
+0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x1f5,0x1f5,0x1f5,0x1f5,0x1f5,0x169b,0x169b,0x169b,
+0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,
+0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x1f8,0x1f8,0x1f8,0x1f8,0x1f8,0x1f8,0x1f8,
+0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,0x16b3,
+0x16b3,0x16b3,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x1fb,0x16b0,0x16b0,0x16b0,0x16b0,0x1fb,0x1fb,0x1fb,
+0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16b6,
+0x16c8,0x16c8,0x16b6,0x16b6,0x16b6,0x16b6,0x201,0x201,0x16c8,0x16c8,0x16cb,0x16cb,0x16b6,0x16b6,0x16c8,0x16bc,
+0x16b9,0x16bf,0x16d1,0x16d1,0x16c2,0x16c2,0x16c5,0x16c5,0x16c5,0x16d1,0x1785,0x1785,0x1785,0x1785,0x1785,0x1785,
+0x1785,0x1785,0x1785,0x1785,0x1785,0x1785,0x1785,0x1785,0x1782,0x1782,0x1782,0x1782,0x177f,0x177f,0x201,0x201,
+0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,
+0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,0x201,
+0x204,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,
+0x16d4,0x16d4,0x16d4,0x16d4,0x16d4,0x204,0x204,0x204,0x204,0x204,0x204,0x204,0x204,0x204,0x204,0x204,
+0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x207,0x207,0x207,0x207,
+0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x16d7,0x16d7,0x207,0x207,0x207,0x207,0x207,0x207,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x207,0x207,0x1aeb,0x1aeb,0x207,0x207,
+0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,
+0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,0x207,
+0x16da,0x16e9,0x16e0,0x16dd,0x16ef,0x16ef,0x16e3,0x16ef,0x20a,0x20a,0x20a,0x20a,0x20a,0x20a,0x20a,0x20a,
+0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x16e6,0x20a,0x20a,0x20a,0x20a,0x20a,0x20a,
+0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f5,0x16f2,0x16f2,0x16f2,0x16f2,0x16f2,0x16f2,
+0x16f2,0x16f2,0x16f2,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x20d,0x16fb,
+0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,
+0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x1797,0x198f,0x210,0x210,0x1788,0x1788,0x1788,
+0x1794,0x1794,0x1788,0x1788,0x1788,0x1788,0x1794,0x1788,0x1788,0x1788,0x1788,0x178b,0x210,0x210,0x210,0x210,
+0x1791,0x1791,0x1791,0x1791,0x1791,0x1791,0x1791,0x1791,0x1791,0x1791,0x178e,0x178e,0x179a,0x179a,0x179a,0x178e,
+0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,
+0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,
+0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x213,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,
+0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x17af,0x219,0x17af,0x17af,0x219,0x219,
+0x219,0x219,0x219,0x17ac,0x17ac,0x17ac,0x17ac,0x17ac,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x21c,
+0x17b2,0x21c,0x17b2,0x17b2,0x17b2,0x17b2,0x21c,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,
+0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x21c,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,0x17b2,
+0x17b2,0x17b5,0x21c,0x21c,0x21c,0x21c,0x21c,0x21c,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,
+0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,
+0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x21f,0x21f,0x21f,0x21f,0x21f,
+0x21f,0x21f,0x21f,0x21f,0x21f,0x21f,0x21f,0x21f,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,
+0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x21f,0x21f,0x21f,0x21f,0x21f,
+0x21f,0x21f,0x17b8,0x17b8,0x17b8,0x17b8,0x17b8,0x17b8,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,
+0x1929,0x1929,0x1929,0x1929,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a94,0x1b3c,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1b3f,
+0x1b3c,0x222,0x1a0a,0x1a94,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x17df,0x1a0a,0x1a0a,0x1a94,0x1a94,0x1a94,0x1a94,0x1a94,
+0x1a94,0x1a94,0x1a94,0x1b3c,0x222,0x1a97,0x1a97,0x1a97,0x1929,0x192c,0x192c,0x192c,0x192c,0x192c,0x192c,0x192c,
+0x192c,0x192c,0x192c,0x192c,0x192c,0x192c,0x1929,0x1929,0x17c4,0x17c4,0x17c4,0x17c4,0x17c1,0x17c4,0x17c4,0x17c7,
+0x17ca,0x17c7,0x17c7,0x17c4,0x225,0x225,0x225,0x225,0x225,0x225,0x225,0x225,0x225,0x225,0x225,0x225,
+0x225,0x225,0x225,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x1821,0x1821,0x1821,0x1821,0x1818,0x1818,0x1818,0x1812,
+0x1815,0x1815,0x1815,0x1a40,0x228,0x228,0x228,0x228,0x181e,0x181e,0x181e,0x181e,0x181e,0x181e,0x181e,0x181e,
+0x181e,0x181e,0x228,0x228,0x228,0x228,0x181b,0x181b,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,
+0x183c,0x22b,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,
+0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x183c,0x1839,0x1827,0x1827,0x1827,0x1827,
+0x1827,0x1827,0x1827,0x22b,0x1827,0x1827,0x1827,0x1827,0x1827,0x1827,0x1839,0x182a,0x183c,0x183f,0x183f,0x1833,
+0x1830,0x1830,0x22b,0x22b,0x22b,0x22b,0x22b,0x22b,0x22b,0x22b,0x22b,0x22b,0x1836,0x1836,0x1836,0x1836,
+0x1836,0x1836,0x1836,0x1836,0x1836,0x1836,0x182d,0x182d,0x182d,0x182d,0x182d,0x182d,0x182d,0x182d,0x182d,0x182d,
+0x182d,0x182d,0x182d,0x182d,0x182d,0x22b,0x22b,0x22b,0x184b,0x184e,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,
+0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1842,0x1842,0x1842,0x1842,0x1842,0x1842,0x1842,0x1842,
+0x1842,0x22e,0x22e,0x22e,0x22e,0x22e,0x22e,0x22e,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,
+0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x231,
+0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,0x1845,
+0x1845,0x231,0x231,0x1845,0x1845,0x1845,0x1845,0x1845,0x1893,0x192f,0x1a9a,0x1a9d,0x1b45,0x234,0x234,0x234,
+0x234,0x234,0x234,0x234,0x234,0x234,0x234,0x234,0x1b42,0x1b42,0x234,0x234,0x234,0x234,0x234,0x234,
+0x234,0x234,0x234,0x234,0x234,0x234,0x234,0x234,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,
+0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x1854,0x237,0x237,0x1848,0x1848,0x1848,0x1848,0x1848,0x1848,
+0x1848,0x1848,0x1848,0x1848,0x1848,0x1848,0x1848,0x1848,0x237,0x1851,0x1848,0x1848,0x1848,0x1848,0x1848,0x1848,
+0x1848,0x1851,0x1848,0x1848,0x1851,0x1848,0x1848,0x237,0x237,0x237,0x237,0x237,0x237,0x237,0x237,0x237,
+0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x1857,0x23a,0x23a,0x23a,
+0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,0x23a,
+0x186f,0x186f,0x1860,0x185a,0x185a,0x186f,0x185d,0x1872,0x1872,0x1872,0x1872,0x1875,0x1875,0x1869,0x1866,0x1863,
+0x186c,0x186c,0x186c,0x186c,0x186c,0x186c,0x186c,0x186c,0x186c,0x186c,0x1aee,0x1869,0x23d,0x1863,0x1992,0x1a43,
+0x1af1,0x1af1,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,
+0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,0x23d,
+0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,0x187b,
+0x187b,0x187b,0x187b,0x187b,0x240,0x240,0x240,0x240,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,
+0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,0x1878,
+0x1878,0x1878,0x1878,0x1878,0x240,0x240,0x240,0x240,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,
+0x1896,0x1896,0x1896,0x1896,0x1896,0x1a13,0x1a13,0x1a13,0x1a13,0x1a13,0x1aa0,0x1aa0,0x1aa0,0x1aa0,0x1aa0,0x1aa0,
+0x243,0x243,0x243,0x243,0x243,0x243,0x243,0x243,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,
+0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,
+0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x1935,0x246,0x246,0x246,0x246,0x246,0x246,0x246,0x246,0x246,
+0x246,0x246,0x246,0x246,0x246,0x246,0x246,0x246,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,
+0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x249,
+0x18d5,0x18d5,0x249,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,
+0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18d5,0x18c9,0x18c9,0x18c9,0x18c9,0x18c9,0x18c9,0x249,
+0x249,0x249,0x18c9,0x249,0x18c9,0x18c9,0x249,0x18c9,0x18c9,0x18c9,0x18cc,0x18c9,0x18cf,0x18cf,0x18d8,0x18c9,
+0x249,0x249,0x249,0x249,0x249,0x249,0x249,0x249,0x18d2,0x18d2,0x18d2,0x18d2,0x18d2,0x18d2,0x18d2,0x18d2,
+0x18d2,0x18d2,0x249,0x249,0x249,0x249,0x249,0x249,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,
+0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,
+0x1938,0x1938,0x1938,0x1938,0x24c,0x24c,0x24c,0x24c,0x1905,0x1908,0x1917,0x1917,0x1908,0x190b,0x1905,0x1902,
+0x255,0x255,0x255,0x255,0x255,0x255,0x255,0x255,0x18f0,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18ed,
+0x18ed,0x18db,0x18db,0x18db,0x18f0,0x18f0,0x18f0,0x18f0,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,
+0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x258,0x258,0x258,0x258,
+0x258,0x258,0x258,0x258,0x258,0x258,0x258,0x258,0x1998,0x1998,0x1998,0x1998,0x1998,0x1998,0x1998,0x1998,
+0x1998,0x1998,0x1998,0x1998,0x1998,0x1998,0x258,0x258,0x1aa9,0x1aa9,0x1aa9,0x1aa9,0x1b4b,0x28b,0x28b,0x28b,
+0x1aa9,0x1aa9,0x1aa9,0x28b,0x28b,0x28b,0x28b,0x28b,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,
+0x19aa,0x19aa,0x19aa,0x19aa,0x19a7,0x19a7,0x19a7,0x199b,0x199b,0x199b,0x199b,0x199b,0x199b,0x199b,0x199b,0x199b,
+0x19a7,0x19a1,0x199e,0x19a4,0x25b,0x25b,0x25b,0x25b,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,
+0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,0x19ad,
+0x19ad,0x19ad,0x19ad,0x25e,0x25e,0x19ad,0x19ad,0x19ad,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x261,0x19bc,
+0x19bc,0x261,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,
+0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19bc,0x19b9,0x19b9,0x19b9,0x19b9,0x19b9,0x261,
+0x19b0,0x19b0,0x261,0x19b9,0x19b9,0x19b0,0x19b9,0x19b3,0x19bc,0x261,0x261,0x261,0x261,0x261,0x261,0x261,
+0x19c5,0x19c5,0x19c8,0x19c8,0x19bf,0x19bf,0x19bf,0x19bf,0x264,0x264,0x264,0x264,0x264,0x264,0x264,0x264,
+0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x19c2,0x264,0x264,0x264,0x264,0x264,0x264,
+0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19ce,0x19cb,0x19cb,0x19cb,
+0x19ce,0x19cb,0x19cb,0x19cb,0x19cb,0x267,0x267,0x267,0x267,0x267,0x267,0x267,0x267,0x267,0x267,0x267,
+0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,0x19d7,
+0x19d7,0x19d7,0x19d7,0x19d1,0x19d1,0x19d4,0x19d4,0x19da,0x19da,0x26a,0x26a,0x26a,0x26a,0x26a,0x26a,0x26a,
+0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,0x19dd,
+0x19dd,0x19dd,0x19dd,0x19dd,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,0x26d,
+0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,
+0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e0,0x19e3,0x19ec,0x19e0,0x19e0,0x270,0x270,0x270,0x270,0x270,
+0x19ef,0x19ef,0x19ef,0x19ef,0x19ef,0x19ef,0x19ef,0x19f2,0x273,0x273,0x273,0x273,0x273,0x273,0x273,0x273,
+0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,0x19fb,
+0x19fb,0x19fb,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f5,0x19f8,0x19f8,0x19f8,
+0x19f8,0x19fe,0x19fe,0x19fe,0x19fe,0x19fe,0x276,0x276,0x276,0x276,0x276,0x276,0x1a55,0x1a55,0x1a55,0x1a55,
+0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,0x1a55,
+0x1a55,0x1a55,0x1a55,0x27c,0x27c,0x27c,0x27c,0x27c,0x27c,0x27c,0x27c,0x27c,0x1a64,0x1a64,0x1a64,0x1a64,
+0x1a64,0x1a64,0x1a64,0x1a64,0x27f,0x27f,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,
+0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a64,0x1a61,0x1a61,0x1a61,
+0x1a58,0x1a58,0x1a58,0x1a58,0x27f,0x27f,0x1a58,0x1a58,0x1a61,0x1a61,0x1a61,0x1a61,0x1a5b,0x1a64,0x1a5e,0x1a64,
+0x1a61,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,
+0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x27f,0x1a70,0x1a70,0x1a70,0x1a70,
+0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x282,0x282,0x282,0x1a67,0x1a67,0x1a67,0x1a67,
+0x1a67,0x1a67,0x1a67,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a73,0x1a73,0x282,0x282,0x285,0x1a76,0x1a76,0x1a76,
+0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,
+0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x1a76,0x285,0x285,
+0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,
+0x1aa3,0x1aa3,0x1aa3,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,
+0x1aa6,0x1aa6,0x1aa6,0x1aa6,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x288,0x1938,0x1938,0x1938,0x1938,
+0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1aa9,0x1aa9,0x1aa9,0x1b4b,
+0x1b4b,0x1b4b,0x1b4b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x1aa9,0x1aa9,0x1aa9,0x1aa9,
+0x1aa9,0x1aa9,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x28b,0x28b,0x28b,
+0x28b,0x28b,0x28b,0x28b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x28b,0x28b,0x28b,0x28b,0x28b,
+0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x1b4b,0x28b,
+0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,
+0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,0x28b,
+0x1a7f,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,
+0x1a79,0x1a79,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x28e,0x1a7c,
+0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a82,0x1a82,0x1a82,0x1a82,
+0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x1a88,0x291,0x291,0x291,0x291,0x291,0x1a85,
+0x1af7,0x1af7,0x1af7,0x1af7,0x1af7,0x1af4,0x1af4,0x1af4,0x1af4,0x1af4,0x1af4,0x1af4,0x294,0x294,0x294,0x294,
+0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,0x294,
+0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x297,0x297,0x1b12,0x297,0x297,0x1b12,0x1b12,0x1b12,0x1b12,
+0x1b12,0x1b12,0x1b12,0x1b12,0x297,0x1b12,0x1b12,0x297,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,
+0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1b12,0x1afa,0x1b09,0x1b09,0x1b09,0x1b09,0x1b09,0x297,0x1b09,
+0x1b0c,0x297,0x297,0x1afa,0x1afa,0x1b0f,0x1b00,0x1b15,0x1b09,0x1b15,0x1b09,0x1afd,0x1b18,0x1b03,0x1b18,0x297,
+0x297,0x297,0x297,0x297,0x297,0x297,0x297,0x297,0x1b06,0x1b06,0x1b06,0x1b06,0x1b06,0x1b06,0x1b06,0x1b06,
+0x1b06,0x1b06,0x297,0x297,0x297,0x297,0x297,0x297,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,
+0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x29a,0x29a,
+0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,
+0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,0x29a,
+0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,
+0x1b1e,0x1b1e,0x1b1e,0x2a0,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,
+0x1b1e,0x1b1e,0x1b1e,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,
+0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x1b21,0x1b21,0x1b21,0x1b21,0x1b21,0x1b21,0x1b21,0x1b21,
+0x1b21,0x1b21,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x2a0,0x1b54,0x1b54,0x1b54,0x1b54,0x1b54,0x1b54,0x1b54,0x1b54,
+0x1b54,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,
+0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,0x2a3,
+0x2a3,0x2a3,0x2a3,0x2a3,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x2a6,0x1b24,
+0x1b24,0x1b27,0x2a6,0x2a6,0x1b2a,0x1b2a,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,0x2a6,
+0x2a6,0x2a6,0x2a6,0x2a6,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,
+0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,
+0x2a9,0x2a9,0x2a9,0x2a9,0x1932,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,0x2c1,
+0x2c1,0x2c1,0x2c1,0x2c1,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,
+0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,
+0x2a9,0x2a9,0x97b,0x97b,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x2c4,
+0x2c4,0x2c4,0x2c4,0x2c4,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,0x2a9,
+0x2a9,0x2a9,0x2a9,0x2a9,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x1a04,
+0x1a04,0x1a04,0x1a04,0x1a04,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,0x1b33,
+0x1b33,0x2ac,0x2ac,0x2ac,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0x12e4,
+0x12e4,0x12e4,0x2af,0x2af,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,
+0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0x2af,0x2af,
+0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,
+0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,0x2af,
+0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,
+0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0x1b36,0x1b36,0x1b36,0x1b36,0x1b36,0x1b36,0x1b36,0x2b2,0x2b2,
+0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,
+0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0x2b5,0x2b5,
+0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,
+0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,0x2b8,
+0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,
+0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x2bb,0x2bb,
+0x17dc,0x17dc,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,0x2be,
+0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,
+0x3cc,0x3c0,0x3c0,0x3c0,0x3c0,0x3c0,0x3c0,0x3c0,0x3c0,0x3cc,0x3cc,0x3cc,0x3cc,0x3c6,0x1158,0x133e,
+0x3cf,0x945,0x948,0x3bd,0x3bd,0x1155,0x133b,0x133b,0x3d2,0x3d2,0x3d2,0x3d2,0x3d2,0x3d2,0x3d2,0x3d2,
+0x1155,0x3c0,0x3c0,0x3cc,0xce1,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,
+0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,
+0x3cf,0x3cf,0x3c0,0x3c0,0x8d0,0x8d3,0x963,0x963,0x963,0x963,0x963,0x963,0x963,0x963,0x963,0x963,
+0x3c9,0xfba,0xfb7,0x1341,0x1341,0x1341,0x1341,0x1341,0x1506,0x115b,0x115b,0xf0c,0xf0c,0xdda,0xf0c,0xf0c,
+0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3d2,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,
+0x3cf,0x3d2,0x3cf,0x3cf,0x3d2,0x3cf,0x3cf,0x3cf,0x3cf,0x3cf,0x133b,0x133e,0x3c3,0x3cf,0x3cc,0x3cc,
+0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x1347,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,
+0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x46e,0x1347,0x18ae,0x18ae,0xfd8,0x45f,0x468,
+0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,
+0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0x4aa,0xbc7,0xbc7,0xde6,0xde6,0x8d6,0xde9,0x1428,0x1428,0x1428,
+0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,
+0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,0x4ad,
+0x4b3,0x4b3,0x4b3,0x1170,0x1170,0x1170,0x1170,0x1170,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,
+0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,
+0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x4b0,0x116d,0x116d,0x116d,0x116d,0x116d,0x116d,
+0x4b6,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,
+0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,0x4b3,
+0x4b3,0x4b3,0x4b3,0x4b3,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,
+0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,
+0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4b9,0x4b9,0x4b9,0x4b9,0x4bc,0x9bd,0x1005,0x1005,0x1008,0x1005,
+0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,
+0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x4bf,0x4b9,0x1008,0x1005,0x1008,0x1005,0x1008,0x1005,
+0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,
+0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4cb,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,0x4ce,
+0x6ae,0x6ae,0x6b1,0x4e9,0x6bd,0x6ba,0x6ba,0x6b7,0x513,0x513,0x4d1,0x4d1,0x4d1,0x4d1,0x4d1,0xb58,
+0x6c0,0x4f5,0x6d8,0x6db,0x50a,0x6c0,0x4f8,0x4f8,0x4e9,0x504,0x504,0x6ae,0x510,0x50d,0x6b4,0x4e3,
+0x4da,0x4da,0x4dd,0x4dd,0x4dd,0x4dd,0x4dd,0x4e0,0x4dd,0x4dd,0x4dd,0x4d4,0x51c,0x519,0x516,0x516,
+0x6cc,0x4fe,0x4fb,0x6c9,0x6c6,0x6c3,0x6d5,0x4ec,0x6d2,0x6d2,0x501,0x504,0x6cf,0x6cf,0x501,0x504,
+0x4e6,0x4e9,0x4e9,0x4e9,0x507,0x4f2,0x4ef,0xbdc,0xaf2,0xaf5,0xaef,0xaef,0xaef,0xaef,0xbd3,0xbd3,
+0xbd3,0xbd3,0xbd9,0xd0e,0xd0b,0xdf5,0xdf8,0xbd6,0xdf8,0xdf8,0xdf8,0xdf8,0xdf5,0xdf8,0xdf8,0xbd0,
+0x540,0x540,0x540,0x540,0x540,0x540,0x540,0x53d,0x543,0x75c,0x540,0x9c0,0x9e1,0xaf8,0xaf8,0xaf8,
+0xbe2,0xbe2,0xdfe,0xdfe,0xdfe,0xdfe,0x1179,0x117c,0x117c,0x135c,0x14f4,0x151e,0x1521,0x1521,0x1734,0x18b1,
+0x54f,0x54f,0x567,0x6ea,0x54c,0x6e7,0x54f,0x564,0x54c,0x6ea,0x55e,0x567,0x567,0x567,0x55e,0x55e,
+0x567,0x567,0x567,0x6f3,0x54c,0x567,0x6ed,0x54c,0x55b,0x567,0x567,0x567,0x567,0x567,0x54c,0x54c,
+0x552,0x6e7,0x6f0,0x54c,0x567,0x54c,0x6f6,0x54c,0x567,0x555,0x56d,0x6f9,0x567,0x567,0x558,0x55e,
+0x567,0x567,0x56a,0x567,0x55e,0x561,0x561,0x561,0x561,0xb04,0xb01,0xd11,0xe07,0xbf7,0xbfa,0xbfa,
+0xbf4,0xbf1,0xbf1,0xbf1,0xbf1,0xbfa,0xbf7,0xbf7,0xbf7,0xbf7,0xbee,0xbf1,0xe04,0xf18,0xf1b,0x100e,
+0x117f,0x117f,0x117f,0x6ff,0x6fc,0x570,0x573,0x573,0x573,0x573,0x573,0x6fc,0x6ff,0x6ff,0x6fc,0x573,
+0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x705,0x57c,0x57c,0x57c,0x57c,
+0x702,0x702,0x702,0x702,0x702,0x702,0x702,0x702,0x702,0x702,0x576,0x576,0x576,0x576,0x576,0x576,
+0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x57f,0x588,0x588,0x582,0x582,0x582,0x585,0x57f,
+0x582,0x582,0x57f,0x57f,0x57f,0x57f,0x582,0x582,0x708,0x708,0x57f,0x57f,0x582,0x582,0x582,0x582,
+0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x585,0x585,0x585,0x582,0x582,0x70b,0x582,
+0x70b,0x582,0x582,0x582,0x582,0x582,0x582,0x582,0x57f,0x582,0x57f,0x57f,0x57f,0x57f,0x57f,0x57f,
+0x582,0x582,0x57f,0x708,0x57f,0x57f,0x57f,0xb0a,0xb0a,0xb0a,0xb0a,0xb0a,0xb0a,0xb0a,0xb0a,0xb0a,
+0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0xbfd,0x711,0x58b,0x711,0x711,
+0x58e,0x58b,0x58b,0x711,0x711,0x58e,0x58b,0x711,0x58e,0x58b,0x58b,0x711,0x58b,0x711,0x59a,0x597,
+0x58b,0x711,0x58b,0x58b,0x58b,0x58b,0x711,0x58b,0x58b,0x711,0x711,0x711,0x711,0x58b,0x58b,0x711,
+0x58e,0x711,0x58e,0x711,0x711,0x711,0x711,0x711,0x717,0x591,0x711,0x591,0x591,0x58b,0x58b,0x58b,
+0x711,0x711,0x711,0x711,0x58b,0x58b,0x58b,0x58b,0x711,0x711,0x58b,0x58b,0x58b,0x58e,0x58b,0x58b,
+0x58e,0x58b,0x58b,0x58e,0x711,0x58e,0x58b,0x58b,0x711,0x58b,0x58b,0x58b,0x58b,0x58b,0x711,0x58b,
+0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x714,0x711,0x58e,0x58b,
+0x711,0x711,0x711,0x711,0x58b,0x58b,0x711,0x711,0x58b,0x58e,0x714,0x714,0x58e,0x58e,0x58b,0x58b,
+0x58e,0x58e,0x58b,0x58b,0x58e,0x58e,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58e,0x58e,0x711,0x711,
+0x58e,0x58e,0x711,0x711,0x58e,0x58e,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,
+0x58b,0x711,0x58b,0x58b,0x58b,0x711,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x711,0x58b,0x58b,
+0x58b,0x58b,0x58b,0x58b,0x58e,0x58e,0x58e,0x58e,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,
+0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x711,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,
+0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,
+0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58b,0x58e,0x58e,0x58e,0x58e,0x58b,0x58b,0x58b,0x58b,
+0x58b,0x58b,0x58e,0x58e,0x58e,0x58e,0x58b,0x594,0x58b,0x58b,0xc00,0xc00,0xc00,0xc00,0xc00,0xc00,
+0xc00,0xc00,0xc00,0xc00,0xc00,0xc00,0xc00,0xc00,0x59d,0xb0d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,
+0x5a9,0x5a6,0x5a9,0x5a6,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x71a,0x59d,0x59d,0x59d,0x59d,0x59d,
+0x59d,0x59d,0x81f,0x81f,0x59d,0x59d,0x59d,0x59d,0x5a3,0x5a3,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,
+0x5a0,0x825,0x822,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,
+0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,
+0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0x59d,0xb0d,0xc06,0xb0d,0xb0d,0xb0d,0x5ac,0x5ac,0x5ac,0x5ac,
+0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,
+0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x5ac,0x723,0x723,0x723,0x723,
+0x723,0x723,0x723,0x723,0x723,0x723,0x5b2,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,
+0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xc6f,0xd89,0x72c,0x72c,0x72c,0x72c,
+0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,
+0x5b5,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x5b8,0x72c,0x72c,0x72c,0x72c,
+0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x5b8,0x5b8,0x5b8,0x5b8,0x72c,0x72c,0x72c,0x72c,
+0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72f,0x72f,0x72f,0x72f,
+0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x72f,0x5bb,0x5bb,0x72f,0x72f,
+0x72f,0x72f,0xc09,0xc09,0xc09,0xc09,0xc09,0xc09,0xc09,0xc09,0xc09,0xc09,0x735,0x735,0x5be,0x732,
+0x732,0x732,0x732,0x732,0x732,0x732,0x5c1,0x5c1,0x5be,0x5be,0x5c4,0x5c4,0x5c4,0x5c4,0x735,0x735,
+0x5c4,0x5c4,0x738,0x735,0x5be,0x5be,0x5be,0x5be,0x735,0x735,0x5c4,0x5c4,0x738,0x735,0x5be,0x5be,
+0x5be,0x5be,0x735,0x735,0x732,0x5be,0x5c4,0x735,0x5be,0x5be,0x732,0x735,0x735,0x735,0x5c4,0x5c4,
+0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x5be,0x735,0x732,
+0x735,0x732,0x5be,0x5c4,0x5c4,0x5c4,0x5c4,0x5c4,0x5c4,0x5be,0x5be,0x732,0xb16,0xb16,0xb16,0xb16,
+0xb16,0xb16,0xb16,0xb16,0xc0c,0xc0c,0xc0c,0xc0f,0xc0f,0xc8a,0xc8a,0xc0c,0x5d3,0x5d3,0x5d3,0x5d3,
+0x5d0,0x74a,0x747,0x5ca,0x5ca,0x73b,0x5ca,0x5ca,0x5ca,0x5ca,0x741,0x73b,0x5ca,0x5d0,0x5ca,0x5c7,
+0xd92,0xd92,0xc15,0xc15,0xe13,0xb19,0x5cd,0x5cd,0x73e,0x5d6,0x73e,0x5cd,0x5d0,0x5ca,0x5d0,0x5d0,
+0x5ca,0x5ca,0x5d0,0x5ca,0x5ca,0x5ca,0x5d0,0x5ca,0x5ca,0x5ca,0x5d0,0x5d0,0x5ca,0x5ca,0x5ca,0x5ca,
+0x5ca,0x5ca,0x5ca,0x5ca,0x5d0,0x5d3,0x5d3,0x5cd,0x5ca,0x5ca,0x5ca,0x5ca,0x74d,0x5ca,0x74d,0x5ca,
+0x5ca,0x5ca,0x5ca,0x5ca,0x828,0x828,0x828,0x828,0x828,0x828,0x828,0x828,0x828,0x828,0x828,0x828,
+0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5ca,0x5d0,0x74d,0x74a,0x5d9,0x74d,
+0x73b,0x741,0x5d0,0x73b,0x744,0x73b,0x73b,0x5ca,0x73b,0x74a,0x5d9,0x74a,0xb19,0xb19,0xc18,0xc18,
+0xc18,0xc18,0xc18,0xc18,0xc18,0xc18,0xc18,0xc1b,0xc18,0xc18,0xe10,0xec7,0x5dc,0x5dc,0x5dc,0x5dc,
+0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,0x5dc,
+0x5df,0x13e6,0x13e6,0x13e6,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x152a,0x5eb,0x5f4,0x5eb,
+0x5eb,0x13e6,0x5df,0x5df,0x5f4,0x5f4,0x13e9,0x13e9,0x5f7,0x5f7,0x5e8,0x5ee,0x5e8,0x5e8,0x5ee,0x5df,
+0x5ee,0x5df,0x5ee,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5ee,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,
+0x13e6,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5ee,0x5ee,0x5df,0x5df,0x5df,
+0x5df,0x5df,0x5df,0x5df,0x5df,0x753,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5ee,0x5df,0x5df,0x5ee,
+0x5df,0x5df,0x5df,0x5df,0x13e6,0x5df,0x13e6,0x5df,0x5df,0x5df,0x5df,0x13e6,0x13e6,0x13e6,0x5df,0x12de,
+0x5df,0x5df,0x5df,0x5e5,0x5e5,0x5e5,0x5e5,0x1368,0x1368,0x5df,0x5e2,0x5f1,0x5f4,0x5e8,0x5e8,0x5e8,
+0xc21,0xc1e,0xc21,0xc1e,0xc21,0xc1e,0xc21,0xc1e,0xc21,0xc1e,0xc21,0xc1e,0xc21,0xc1e,0x750,0x750,
+0x750,0x750,0x750,0x750,0x750,0x750,0x750,0x750,0x5df,0x5ee,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,
+0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x13e6,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,
+0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x5df,0x13e6,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x618,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x621,0x621,0x621,0x621,
+0x621,0x621,0x621,0x621,0x618,0x61e,0x60f,0x612,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x615,0x615,0x615,0x615,0x615,0x615,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,
+0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,
+0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x618,
+0x61e,0x618,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61b,0x621,0x61e,0x618,
+0x61b,0x621,0x61e,0x618,0x61e,0x618,0x61b,0x621,0x61e,0x618,0x61e,0x618,0x61b,0x621,0x61e,0x618,
+0x61b,0x621,0x61e,0x618,0x61e,0x618,0x136b,0x136b,0x136b,0x136b,0x136b,0x136b,0x136b,0x136b,0x136b,0x136b,
+0x136b,0x136b,0x136b,0x136b,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61b,0x621,0x61b,0x621,0x61e,0x618,
+0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61e,0x618,0x61b,0x61e,0x618,0x61b,
+0x61e,0x618,0x61b,0x621,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x61b,0x61b,0x61b,0x61b,0x61b,
+0x61b,0x61b,0x61b,0x61b,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x61b,0x61b,0x618,0x61b,0x618,0x61b,0x618,0x618,0x61b,0x618,0x618,0x61b,
+0x618,0x61b,0x618,0x618,0x61b,0x618,0x61b,0x61b,0x618,0x618,0x618,0x61b,0x618,0x618,0x618,0x618,
+0x618,0x61b,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,
+0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x618,0x61b,0x61b,0x618,0x618,0x61b,0x618,0x61b,0x618,
+0x618,0x618,0x618,0x618,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,
+0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,
+0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x61b,0x621,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,
+0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x621,0x61e,0x61e,0x61e,
+0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x61e,0x624,0x624,0x624,0x624,0x101a,0x101a,0x101a,0x152d,
+0x152d,0x152d,0x152d,0x152d,0x152d,0x152d,0x173a,0x173a,0x885,0x88b,0x88b,0x897,0x897,0x888,0x87f,0x888,
+0x87f,0x888,0x87f,0x888,0x87f,0x888,0x87f,0x888,0x633,0x633,0x62d,0x633,0x62d,0x633,0x62d,0x633,
+0x62d,0x633,0x62d,0x630,0x636,0x633,0x62d,0x633,0x62d,0x630,0x636,0x633,0x62d,0x633,0x62d,0x630,
+0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x633,0x62d,0x633,0x62d,0x633,
+0x62d,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,
+0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,0x636,0x633,0x62d,0x630,
+0x636,0x633,0x62d,0x630,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,
+0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x720,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,
+0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,
+0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x71d,0x726,0x726,0x726,0x726,0x726,0x726,
+0x726,0x726,0x726,0x726,0x726,0x726,0x729,0x726,0x726,0x726,0x726,0x726,0x726,0x726,0x726,0x726,
+0x726,0x726,0x726,0x726,0x723,0x723,0x723,0x723,0x723,0x723,0x723,0x723,0x723,0x723,0x723,0x723,
+0x723,0x723,0x723,0x723,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,
+0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,0x72c,
+0x72c,0x72c,0x72c,0x72c,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,
+0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,0x756,
+0x756,0x756,0x756,0x756,0xc78,0x8e8,0x8e2,0x8df,0x8e5,0x8dc,0x76b,0x76e,0x76e,0x76e,0x76e,0x76e,
+0x76e,0x76e,0x76e,0x76e,0x8ee,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,
+0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,
+0x76b,0x76b,0x76b,0x76b,0x76b,0x76b,0x8eb,0x8eb,0x771,0x8fd,0x900,0x906,0x82b,0x837,0x91b,0x834,
+0x8f4,0x8f1,0x8f4,0x8f1,0x8fa,0x8f7,0x8fa,0x8f7,0x8f4,0x8f1,0x831,0x906,0x8f4,0x8f1,0x8f4,0x8f1,
+0x8f4,0x8f1,0x8f4,0x8f1,0x909,0x912,0x90f,0x90f,0x777,0x7b3,0x7b3,0x7b3,0x7b3,0x7b3,0x7b3,0x7ad,
+0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,0x7ad,
+0x7ad,0x7ad,0x7ad,0x77a,0x795,0x774,0x79b,0x79e,0x798,0x7b0,0x7b0,0x7b0,0x7b0,0x7b0,0x7b0,0x7aa,
+0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,0x7aa,
+0x7aa,0x7aa,0x7aa,0x77a,0x795,0x774,0x795,0xc7b,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,
+0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,
+0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x819,0x12d8,0x12d8,0x12d8,0x12d8,0x12d8,0x81c,
+0x831,0x834,0x834,0x834,0x834,0x834,0x834,0x834,0x834,0x834,0x951,0x951,0x951,0x951,0x83a,0x83a,
+0x90c,0x918,0x918,0x918,0x918,0x915,0x82e,0x903,0xb3d,0xb3d,0xb3d,0xc8d,0xcab,0xca8,0xb5b,0x8d9,
+0x840,0x83d,0x840,0x843,0x83d,0x840,0x83d,0x840,0x83d,0x840,0x83d,0x83d,0x83d,0x83d,0x83d,0x83d,
+0x840,0x840,0x83d,0x840,0x840,0x83d,0x840,0x840,0x83d,0x840,0x840,0x83d,0x840,0x840,0x83d,0x83d,
+0xcae,0x852,0x84c,0x852,0x84c,0x852,0x84c,0x852,0x84c,0x852,0x84c,0x84c,0x84f,0x84c,0x84f,0x84c,
+0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,
+0x84f,0x84c,0x84f,0x852,0x84c,0x84f,0x84c,0x84f,0x84c,0x84f,0x84c,0x84c,0x84c,0x84c,0x84c,0x84c,
+0x84f,0x84f,0x84c,0x84f,0x84f,0x84c,0x84f,0x84f,0x84c,0x84f,0x84f,0x84c,0x84f,0x84f,0x84c,0x84c,
+0x84c,0x84c,0x84c,0x852,0x84c,0x852,0x84c,0x852,0x84c,0x84c,0x84c,0x84c,0x84c,0x84c,0x852,0x84c,
+0x84c,0x84c,0x84c,0x84c,0x84f,0x852,0x852,0x84f,0x84f,0x84f,0x84f,0x921,0x924,0x855,0x858,0xc96,
+0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,
+0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,
+0x861,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,
+0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x85e,0x86a,0x86a,0x86a,0x86a,
+0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,
+0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0x86a,0xd9b,0xd9b,0xeca,0x864,0x92d,0x92d,0x92d,0x92d,
+0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0xd95,0xd95,0xd95,0xd95,0x86d,0x86d,0x86d,0x86d,
+0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,
+0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x86d,0x1ab2,0x936,0x936,0x936,0x936,
+0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x870,0x870,0x870,
+0x870,0x870,0x870,0xd9e,0xd9e,0xd9e,0xd9e,0x939,0x939,0x939,0x939,0x939,0x870,0x870,0x870,0x870,
+0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,
+0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0x870,0xd9e,0xd9e,
+0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,
+0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,0x873,
+0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,
+0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,
+0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0x876,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,
+0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,0xecd,
+0x113d,0x113d,0x113d,0x113d,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,
+0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x879,
+0x879,0x879,0x879,0x879,0x879,0x879,0x87c,0x87c,0x879,0x87c,0x879,0x87c,0x87c,0x879,0x879,0x879,
+0x879,0x879,0x879,0x879,0x879,0x879,0x879,0x87c,0x879,0x87c,0x879,0x87c,0x87c,0x879,0x879,0x87c,
+0x87c,0x87c,0x879,0x879,0x879,0x879,0x14e5,0x14e5,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,
+0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,
+0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,
+0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x92d,0x131d,0x131d,0x131d,0x131d,0x12bd,0x12bd,0x12bd,0x12bd,
+0x12bd,0x12bd,0x12bd,0x12bd,0xd95,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,
+0xc99,0xc99,0xc99,0xc99,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,
+0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x933,0x930,0x933,0x930,0x930,
+0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,0x930,
+0x930,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,0xc99,
+0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,
+0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0x936,0xd9e,
+0x9b7,0x999,0x999,0x999,0x999,0x993,0x999,0x999,0x9ab,0x999,0x999,0x996,0x9a2,0x9a8,0x9a8,0x9a8,
+0x9a8,0x9a8,0x9ab,0x993,0x99f,0x993,0x993,0x993,0x98a,0x98a,0x993,0x993,0x993,0x993,0x993,0x993,
+0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x9ae,0x993,0x993,0x993,0x993,0x993,0x993,
+0x993,0x993,0x993,0x993,0x996,0x98a,0x993,0x98a,0x993,0x98a,0x9a5,0x99c,0x9a5,0x99c,0x9b4,0x9b4,
+0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,
+0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,0x9c3,
+0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,
+0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,0x9c6,
+0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,
+0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,0x9c9,
+0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,
+0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9cc,0x9cc,
+0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,
+0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9cf,0x9cf,
+0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,
+0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,0x9d2,
+0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,
+0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,0x9d5,
+0x9d8,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,
+0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9d8,0x9db,0x9db,0x9db,
+0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,
+0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0x9db,0xa68,0xa68,0xfff,0xa68,0xa68,0xa68,0xa6b,0xa68,
+0xfff,0xa68,0xa68,0xff6,0xa62,0xa56,0xa56,0xa56,0xa56,0xa65,0xa56,0xfe7,0xfe7,0xfe7,0xa56,0xa59,
+0xa62,0xa5c,0xfed,0xff9,0xff9,0xfe7,0xfe7,0xfff,0xb61,0xb61,0xb61,0xb61,0xb61,0xb61,0xb61,0xb61,
+0xb61,0xb61,0xa6e,0xa6e,0xa5f,0xa5f,0xa5f,0xa5f,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa65,0xa65,
+0xa56,0xa56,0xfff,0xfff,0xfff,0xfff,0xfe7,0xfe7,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,
+0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,
+0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa68,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xdf2,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,
+0xa7d,0xa7d,0xa7d,0xdf2,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,0xa7d,
+0xa7d,0xa7d,0xa7d,0xa7d,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,
+0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,0xa83,
+0xa83,0xa83,0xa83,0xa83,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,
+0xa89,0xa86,0xa8c,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0x1176,0x1176,0x1176,0x1176,0x1176,
+0x1176,0x1176,0x1176,0x1176,0x1173,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,
+0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,
+0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa89,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,
+0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,
+0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xa9e,0xac2,0xac2,0xac2,0xac5,0xac5,0xac2,0xac2,0xac2,
+0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xaaa,0xaaa,0xabf,0xaa1,
+0xaa1,0xaa1,0xaa1,0xaa1,0xaa1,0xaa1,0xabf,0xabf,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,
+0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,
+0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xac2,0xae3,0xae3,0xae3,0xae3,0xae3,0xace,0xace,0xae3,
+0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xae3,0xae3,0xae3,0xae6,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,0xae3,
+0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb10,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,
+0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xb0d,0xc06,0xc06,0xc06,0xc06,0xc06,
+0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,
+0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,0xb1c,
+0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,
+0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,0xb2e,
+0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,
+0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,0xb34,
+0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,
+0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0xb40,0x13ec,0x13ec,0x13ec,0x1b2d,0x1b2d,0x1b2d,0x1b2d,0x1b2d,
+0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,
+0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,0xb43,
+0xb43,0xb43,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0x1b30,0xb46,0xb46,0xb46,0xb46,
+0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,
+0xb46,0xb49,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,
+0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,0xb46,
+0xb46,0xb46,0xb46,0xb46,0xb4c,0xb4c,0xc9c,0xc9c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,
+0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xc9c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,0xb4c,
+0xb4c,0xb4c,0xb4c,0xb4c,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,
+0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,0xb70,
+0xb70,0xb70,0xb70,0x1530,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xd26,0xd26,0xb76,0xb76,0xb76,0xb76,
+0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,
+0xb76,0xb76,0xb76,0xb76,0xb76,0xb76,0xd23,0xd23,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,
+0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,
+0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,
+0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb79,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,
+0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,
+0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb7c,0xb8b,0xb8b,0xb8b,0xb8b,0xb8b,0xb82,0xb8e,0xb94,
+0xb94,0xb94,0xb88,0xb88,0xb88,0xb91,0xb85,0xb85,0xb85,0xb85,0xb85,0xb7f,0xb7f,0xb7f,0xb7f,0xb7f,
+0xb7f,0xb7f,0xb7f,0xb94,0xb94,0xb94,0xb94,0xb94,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb8b,0xb8b,0xb94,0xb94,0xb94,0xb88,
+0xb88,0xb94,0xb94,0xb94,0xb94,0xb94,0xb94,0xb94,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb94,0xb94,
+0xb94,0xb94,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb8b,
+0xb8b,0xb8b,0xb8b,0xb8b,0xb8b,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,0xb88,
+0xb88,0xb88,0x173d,0x173d,0xba0,0xb97,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb97,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xb97,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb97,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xb97,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,
+0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,
+0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xb9a,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,0xba0,
+0xba0,0xba0,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,
+0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xb9d,0xba0,0xba0,0xba0,0xba0,
+0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,
+0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,0xba3,
+0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,
+0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,0xba9,
+0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,
+0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,0xbac,
+0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,
+0xc06,0xc06,0xc06,0xc06,0xc06,0xc06,0xc03,0xc06,0xc03,0xc03,0xc03,0xc03,0xc03,0xc03,0xc03,0xc03,
+0xc03,0xc03,0xc03,0xc03,0xc03,0xc03,0xc03,0xd14,0xd17,0xe0a,0xe0a,0xe0a,0xe0a,0xe0a,0xe0a,0xe0a,
+0xe0a,0xe0a,0xe0a,0xe0a,0xf24,0xf24,0xf24,0xf24,0xc18,0xc18,0xc18,0xc18,0xc18,0xc18,0xc12,0xc12,
+0xc12,0xc12,0xd1a,0xd1a,0xd1a,0xd1a,0xd1a,0xd1a,0xd1d,0xd1d,0xe10,0xec4,0xe10,0xe10,0xe10,0xe10,
+0xe0d,0xe10,0xe0d,0xe10,0xe10,0x1014,0x12ae,0x12ae,0xe19,0xe19,0xe19,0xe19,0xe19,0xe1f,0xe1c,0xf36,
+0xf36,0xf36,0xf36,0x142e,0x1026,0x142e,0x1374,0x1374,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,
+0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc4e,0xc81,0xc7e,0xc81,0xc7e,0xc81,0xc7e,
+0x1137,0x1134,0x102c,0x1029,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,0xc51,
+0xc51,0xc51,0xc51,0xc51,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,
+0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,
+0xc54,0xc54,0xc54,0xc54,0xc57,0xc57,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,0xc54,
+0xc5a,0xc5a,0xc5a,0xc60,0xc5d,0xc87,0xc84,0xc60,0xc5d,0xc60,0xc5d,0xc60,0xc5d,0xc60,0xc5d,0xc60,
+0xc5d,0xc60,0xc5d,0xc60,0xc5d,0xc60,0xc5d,0xc60,0xc5d,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,
+0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,
+0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc60,0xc5d,0xc60,0xc5d,
+0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,
+0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc5a,0xc60,0xc5d,0xc5a,0xc5a,
+0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc69,0xc63,0xc63,0xc63,
+0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,
+0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,
+0xc69,0xc69,0xc69,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,
+0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,0xc63,
+0xc66,0xc63,0xc63,0xc63,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,
+0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,0xc9f,
+0xc9f,0xc9f,0xc9f,0xc9f,0xd20,0xd8f,0xe0d,0xe0d,0xe0d,0xe0d,0xe0d,0xe10,0xe0d,0xe0d,0xec4,0xec4,
+0xe0d,0xe0d,0xe0d,0xe0d,0xe10,0xe10,0xf27,0x1014,0x1014,0x1014,0x1014,0x1014,0x1014,0x1014,0x1014,0x1014,
+0x1014,0x12db,0x12db,0x12b1,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,
+0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,0xd44,
+0xd44,0xd44,0xd44,0xd44,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd4a,0xd4a,0xd4a,0xd4a,0xd4a,0xd47,
+0xd5c,0xd5c,0xd5c,0xd56,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd56,
+0xd5c,0xd5c,0xd5c,0xd5c,0xd50,0xd50,0xd59,0xd59,0xd59,0xd59,0xd4d,0xd4d,0xd4d,0xd4d,0xd4d,0xd53,
+0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe25,0xe22,0xe25,0xe25,0xe25,
+0xe25,0xe25,0xe25,0xe25,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,
+0xd5c,0xd5c,0xd56,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,0xd5c,
+0xd5c,0xd50,0xd50,0xd50,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,
+0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,0xd53,
+0xd53,0xd53,0xd53,0xd53,0xd5f,0xd5f,0xd5f,0xd5f,0xd5f,0xd62,0xd62,0xd62,0xd5f,0xd5f,0xd5f,0xd5f,
+0xd5f,0xd5f,0xe28,0xe28,0xe28,0xe28,0xe28,0xe28,0xf39,0xf39,0xf39,0xf39,0xf39,0xf39,0xf39,0x1140,
+0x1140,0x102f,0x102f,0x102f,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,
+0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,0xd65,
+0xd65,0xd65,0xd65,0xd65,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,
+0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,0xd6b,
+0xd6b,0xd6b,0xd6b,0xd6b,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,
+0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,0xd74,
+0xd74,0xd74,0xd74,0xd74,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,
+0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,0xd80,
+0xd80,0xd80,0xd80,0xd80,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,
+0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,0xd8c,
+0xd8c,0xd8c,0xd8c,0xd8c,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,
+0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,0xe2e,
+0xe2e,0xe2e,0xe2e,0xe2e,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,
+0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,0xe31,
+0xe31,0xe31,0xe31,0xe31,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,
+0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,0xe34,
+0xe34,0xe34,0xe34,0xe34,0xef4,0xef4,0xe46,0xe46,0xf3c,0xf3c,0xf3c,0xf3c,0xf3c,0xf3c,0xf3c,0x103b,
+0x103b,0x103b,0x103b,0x103b,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,0x1038,
+0x1038,0x1038,0x1038,0x1038,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,
+0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,0xe55,0xe52,
+0xe55,0xe52,0xe55,0xe52,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,
+0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,0xe61,
+0xe61,0xe61,0xe61,0xe61,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,
+0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,0xe67,
+0xe67,0xe67,0xe67,0xe67,0xeee,0xeee,0xeee,0xeee,0xeee,0xeee,0xeee,0xeee,0xe7f,0xe7f,0xe7f,0xe7f,
+0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xe7f,0xf3f,0xf3f,0xf3f,0xf3f,0x103e,
+0x103e,0x103e,0x103e,0x103e,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,
+0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,0xe88,
+0xe88,0xe88,0xe88,0xe88,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,
+0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,0xe91,
+0xe91,0xe91,0xe91,0xe91,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,
+0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,0xe9a,
+0xe9a,0xe9a,0xe9a,0xe94,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,
+0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe97,0xe9a,
+0xe9a,0xe9a,0xe9a,0xe9a,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,0xea3,
+0xea3,0xea3,0xea0,0xea0,0xea0,0xea0,0xea0,0xea0,0xea0,0xea0,0xe9d,0xea6,0x104a,0x1044,0x1053,0x1041,
+0xea3,0xea3,0x1041,0x1041,0xeb5,0xeb5,0xea9,0xeb5,0xeb5,0xeb5,0xeac,0xeb5,0xeb5,0xeb5,0xeb5,0xea9,
+0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,0xeb5,
+0xeb5,0xeb5,0xeb5,0xeb5,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,
+0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,0xeb8,
+0xeb8,0xeb8,0xeb8,0xeb8,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,
+0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,0xed0,
+0xed0,0xed0,0xed0,0xed0,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,0xef1,
+0xef1,0xef1,0xef1,0xef1,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,0x1149,
+0x1149,0x1149,0x1149,0x1149,0xf24,0xf24,0xf24,0xf21,0xf21,0xf21,0xf21,0xf21,0x1182,0x13dd,0x13dd,0x13dd,
+0x13dd,0x135f,0x135f,0x135f,0x13e0,0x1362,0x1362,0x13e0,0x1524,0x1524,0x1524,0x1524,0x1527,0x1527,0x1527,0x17ee,
+0x17ee,0x17ee,0x17ee,0x18b4,0xf39,0xf39,0xf39,0xf39,0x102f,0x102f,0x102f,0x102f,0x102f,0x102f,0x102f,0x102f,
+0x102f,0x102f,0x102f,0x102f,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,
+0x1032,0x1032,0x1032,0x1032,0xf5a,0xf5a,0xf5a,0xf5a,0xf6c,0xf75,0xf78,0xf75,0xf78,0xf75,0xf78,0xf75,
+0xf78,0xf75,0xf78,0xf75,0xf75,0xf75,0xf78,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,
+0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf75,0xf5d,0xf5a,0xf5a,0xf5a,
+0xf5a,0xf5a,0xf5a,0xf6f,0xf5a,0xf6f,0xf6c,0xf6c,0xf81,0xf7e,0xf81,0xf81,0xf81,0xf7e,0xf7e,0xf81,
+0xf7e,0xf81,0xf7e,0xf81,0xf7e,0x1065,0x1065,0x1065,0x11a0,0x105c,0x1065,0x105c,0xf7e,0xf81,0xf7e,0xf7e,
+0x105c,0x105c,0x105c,0x105c,0x105f,0x1062,0x11a0,0x11a0,0xf84,0xf84,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,
+0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x106e,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,
+0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,
+0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,
+0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf8a,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,
+0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,
+0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0xf99,0x1563,0x1563,0x1563,0x1563,0x1563,
+0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0x1563,0xf9f,0xf9f,0xf9f,0xf9f,
+0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,
+0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xf9f,0xfe7,0xfff,0xff6,0xffc,
+0xffc,0xfff,0xfff,0xff6,0xff6,0xffc,0xffc,0xffc,0xffc,0xffc,0xfff,0xfff,0xfff,0xfe7,0xfe7,0xfe7,
+0xfe7,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfff,0xfe7,0xff6,
+0xff9,0xfe7,0xfe7,0xffc,0xffc,0xffc,0xffc,0xffc,0xffc,0xfea,0xfff,0xffc,0xff3,0xff3,0xff3,0xff3,
+0xff3,0xff3,0xff3,0xff3,0xff3,0xff3,0x116a,0x116a,0x1167,0x1164,0xff0,0xff0,0x1017,0x1017,0x1017,0x1017,
+0x12db,0x12db,0x12b1,0x12b1,0x12b7,0x12ae,0x12ae,0x12ae,0x12ae,0x12b1,0x13e3,0x12b7,0x12b1,0x12b7,0x12ae,0x12b7,
+0x12db,0x12ae,0x12ae,0x12ae,0x12b1,0x12b1,0x12ae,0x12ae,0x12b1,0x12ae,0x12ae,0x12b1,0x1032,0x1032,0x1032,0x1032,
+0x1032,0x102f,0x102f,0x1032,0x1032,0x1032,0x1032,0x1032,0x1032,0x153c,0x153c,0x153c,0x1140,0x102f,0x102f,0x102f,
+0x102f,0x12e7,0x12c0,0x12c0,0x12c0,0x12c0,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x1050,0x1050,0x104d,0x1047,
+0x104d,0x1047,0x104d,0x1047,0x104d,0x1047,0x1044,0x1044,0x1044,0x1044,0x1059,0x1056,0x1044,0x119d,0x143a,0x143d,
+0x143d,0x143a,0x143a,0x143a,0x143a,0x143a,0x1440,0x1440,0x1557,0x154b,0x154b,0x1548,0x1077,0x106e,0x1077,0x106e,
+0x1077,0x106e,0x1077,0x106e,0x106b,0x1068,0x1068,0x1077,0x106e,0x1383,0x1380,0x1746,0x1383,0x1380,0x1449,0x1446,
+0x155a,0x155a,0x1560,0x155a,0x1560,0x155a,0x1560,0x155a,0x1560,0x155a,0x1560,0x155a,0x1077,0x106e,0x1077,0x106e,
+0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,
+0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x106e,0x1071,0x106e,0x106e,0x106e,
+0x106e,0x106e,0x106e,0x106e,0x106e,0x1077,0x106e,0x1077,0x106e,0x1077,0x1077,0x106e,0x107a,0x107a,0x1080,0x1086,
+0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,
+0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1086,0x1080,0x107a,0x107a,
+0x107a,0x107a,0x1080,0x1080,0x107a,0x107a,0x1083,0x1452,0x144f,0x144f,0x1086,0x1086,0x107d,0x107d,0x107d,0x107d,
+0x107d,0x107d,0x107d,0x107d,0x107d,0x107d,0x1455,0x1455,0x1455,0x1455,0x1455,0x1455,0x109b,0x109b,0x109b,0x109b,
+0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,
+0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x109b,0x10a4,0x10a4,0x10a4,0x10a4,
+0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,0x10a4,
+0x10a4,0x10a4,0x10a4,0x10a4,0x10a7,0x10a7,0x10a7,0x10aa,0x10a7,0x10a7,0x10ad,0x10ad,0x10b0,0x10b0,0x10b0,0x10b0,
+0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,
+0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b0,0x10b9,0x10b9,0x10b9,0x10b9,
+0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10bc,0x10b3,0x10c2,0x10bf,0x10b9,0x10b9,0x10b9,0x10b9,
+0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,
+0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x10b9,0x1389,0x1386,0x10d4,0x10ce,
+0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d1,0x1152,0x10c5,0x10c5,0x10c5,0x10cb,
+0x1458,0x1458,0x1458,0x1458,0x1458,0x1458,0x1458,0x1458,0x10c8,0x10c8,0x10cb,0x10d7,0x10d4,0x10ce,0x10d4,0x10ce,
+0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,
+0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x10d4,0x10ce,0x156f,0x156c,0x156f,0x156c,
+0x1572,0x1572,0x174f,0x1458,0x10e0,0x10e0,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,
+0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,0x10e3,
+0x10e3,0x10e3,0x10e3,0x10e3,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,0x10e0,
+0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10e9,0x10e9,0x10e9,0x10e9,0x10e9,0x10ec,0x10ec,0x10ec,0x1146,0x10f5,
+0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,0x1104,
+0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10ef,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,
+0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,0x10f2,
+0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,
+0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,0x1113,
+0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,
+0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,0x1125,
+0x112e,0x112e,0x112e,0x112e,0x1143,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,
+0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,0x112e,
+0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,
+0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,0x1131,
+0x113d,0x113d,0x113d,0x113d,0x12e1,0x12e1,0x12e1,0x12e1,0x12e1,0x12e1,0x12e1,0x12e1,0x14e2,0x17cd,0x17cd,0x17cd,
+0x17cd,0x17cd,0x17cd,0x17cd,0x17cd,0x17cd,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,0x191d,
+0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,
+0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11b2,0x11a9,0x11a9,0x11ac,0x11ac,0x11b2,0x11a9,0x11a9,0x11a9,0x11a9,0x11a9,
+0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,
+0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,0x11b8,
+0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,
+0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,0x11d3,
+0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,
+0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11df,0x11dc,0x11e2,
+0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,
+0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,0x11ee,
+0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,0x11f4,
+0x11f4,0x132f,0x11fa,0x1332,0x11fa,0x11fa,0x11fa,0x11fa,0x11f7,0x11f7,0x11f7,0x11fa,0x1752,0x1755,0x197a,0x1977,
+0x11fd,0x11fd,0x11fd,0x120c,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,
+0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,0x1212,
+0x1212,0x1212,0x1212,0x1200,0x120c,0x120c,0x11fd,0x11fd,0x11fd,0x11fd,0x120c,0x120c,0x11fd,0x11fd,0x120c,0x120c,
+0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,
+0x1221,0x121e,0x121e,0x121e,0x121e,0x121e,0x121e,0x1218,0x1218,0x1218,0x121e,0x121b,0x1578,0x157b,0x157e,0x157e,
+0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,
+0x1224,0x1230,0x1224,0x1224,0x1224,0x1239,0x1239,0x1224,0x1224,0x1239,0x1230,0x1239,0x1239,0x1230,0x1224,0x1227,
+0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,
+0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,0x1230,
+0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,
+0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,0x124b,
+0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,
+0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1263,0x1260,0x1260,0x1260,
+0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,
+0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,0x126c,
+0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,
+0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,0x127b,
+0x1281,0x1281,0x1290,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,
+0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1296,0x1293,0x1296,0x1293,0x1293,0x1293,
+0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1293,0x1296,0x1293,0x1293,0x1293,0x1293,
+0x1290,0x1290,0x1290,0x1284,0x1284,0x1284,0x1284,0x1290,0x1290,0x128a,0x1287,0x128d,0x128d,0x129c,0x1299,0x1299,
+0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,
+0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,
+0x12a5,0x12a5,0x12a5,0x12a2,0x12a2,0x12a2,0x129f,0x129f,0x129f,0x129f,0x12a2,0x129f,0x129f,0x129f,0x12a5,0x12a2,
+0x12a5,0x12a2,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,
+0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x129f,0x12a5,0x12a2,0x12a2,
+0x129f,0x129f,0x129f,0x129f,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c9,
+0x12c9,0x12c9,0x12a8,0x1980,0x13d7,0x12d2,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,0x13d7,
+0x13d7,0x12d2,0x13d7,0x12d2,0x12b1,0x12b1,0x1365,0x12ae,0x1365,0x1365,0x1365,0x1365,0x12ae,0x12b4,0x12db,0x12ae,
+0x12ae,0x12ae,0x12ae,0x12ae,0x12b4,0x12b7,0x12db,0x12db,0x12b7,0x12db,0x12ae,0x12b7,0x12b7,0x12ba,0x12db,0x12ae,
+0x12ae,0x12db,0x12b1,0x12b1,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x12c3,0x12c3,
+0x12c3,0x12c3,0x13ef,0x13ce,0x12cc,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x13ef,0x187e,
+0x187e,0x187e,0x187e,0x187e,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x1581,
+0x1581,0x1ad0,0x1ad0,0x1ad0,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,0x12c6,
+0x12c6,0x12c6,0x12c6,0x12c6,0x13d7,0x13d7,0x12d2,0x13d7,0x13d7,0x13d7,0x12d2,0x13d7,0x13d7,0x13d7,0x12cc,0x12cc,
+0x12cc,0x12cc,0x12cc,0x13d1,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x12cf,0x13d4,0x13d4,0x13d4,0x13d4,
+0x13d4,0x13d4,0x13d4,0x12cf,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x145b,0x145b,
+0x1a2e,0x1ad0,0x1ad0,0x1ad0,0x13da,0x13da,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x13d4,0x12cf,0x13d4,0x12cf,
+0x12cf,0x13d4,0x13da,0x12d5,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,
+0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,0x12f9,
+0x12f9,0x12f9,0x12f9,0x12f9,0x1383,0x1380,0x1383,0x1380,0x1383,0x1380,0x1383,0x1380,0x1383,0x1380,0x1449,0x1560,
+0x1560,0x1560,0x17fa,0x196e,0x1560,0x1560,0x1749,0x1749,0x1749,0x1743,0x1749,0x1743,0x1971,0x196e,0x1a2b,0x1a28,
+0x1a2b,0x1a28,0x1a2b,0x1a28,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,
+0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,0x13a7,
+0x13a7,0x13a7,0x13a7,0x13a7,0x13bc,0x13ad,0x13bc,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,
+0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,0x13bf,
+0x13bf,0x13bf,0x13bf,0x13bf,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13ad,0x13c5,0x13c5,0x13c5,0x13c5,
+0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,
+0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13c5,0x13cb,0x13cb,0x13cb,0x13cb,
+0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,
+0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13cb,0x13fb,0x13f8,0x1923,0x1923,
+0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,
+0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1401,0x1401,0x1404,0x1404,0x1404,0x1404,0x1404,0x1401,0x1404,0x1404,0x1404,0x1401,0x1404,0x1401,0x1404,
+0x1401,0x1404,0x1404,0x1404,0x1404,0x1404,0x1407,0x1404,0x1404,0x1404,0x1404,0x1401,0x1404,0x1401,0x1401,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1401,0x1401,0x1401,0x1401,
+0x1401,0x1401,0x1401,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,
+0x1401,0x1401,0x1401,0x1401,0x1401,0x1401,0x158a,0x158a,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1596,0x1590,0x1590,0x1596,0x1596,0x1596,0x1596,
+0x1596,0x1596,0x1596,0x1596,0x1596,0x17d0,0x17d0,0x17d0,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1596,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1596,0x17d0,0x17d0,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1407,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1590,0x1590,0x1596,0x1596,0x1590,0x1596,0x1596,0x1596,0x158d,0x158d,0x1596,0x1596,
+0x1404,0x1404,0x1407,0x1407,0x1407,0x1701,0x1404,0x1407,0x1404,0x1404,0x1407,0x1599,0x1599,0x1596,0x1596,0x17d0,
+0x17d0,0x17d0,0x17d0,0x17d0,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,0x1596,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1590,0x1590,0x1596,0x1701,0x1596,0x1590,0x1596,0x17d0,0x17d0,0x17d0,0x17d3,0x17d3,0x17d3,0x17d3,0x17d3,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1596,
+0x1404,0x1596,0x1407,0x1407,0x1404,0x1404,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,
+0x1407,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,0x1407,
+0x1407,0x1407,0x1407,0x1407,0x1407,0x1404,0x1404,0x1404,0x1407,0x1404,0x1404,0x1404,0x1404,0x1407,0x1407,0x1407,
+0x1404,0x1407,0x1407,0x1407,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1407,0x1404,0x1407,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1701,0x1404,0x1404,0x1404,0x1404,0x1596,0x1590,0x17d0,
+0x145e,0x145e,0x145e,0x145e,0x158a,0x158a,0x158d,0x158d,0x158d,0x1593,0x1596,0x17d0,0x17d0,0x17d0,0x17d0,0x1758,
+0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,0x1404,
+0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,0x1596,0x1590,0x1590,0x1596,0x1599,0x1599,0x1596,0x1596,
+0x1596,0x1596,0x1887,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,0x1590,0x1596,0x1590,0x1590,0x1590,0x1590,
+0x1596,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,0x1590,0x1590,0x1590,0x1596,0x158d,0x158d,0x158d,0x158d,
+0x158d,0x158d,0x1596,0x1404,0x1404,0x1404,0x1404,0x1404,0x14e8,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,
+0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x14e8,0x140a,0x140a,0x140a,0x14e8,0x140a,0x14e8,
+0x140a,0x14e8,0x140a,0x14e8,0x140a,0x140a,0x140a,0x14e8,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x14e8,0x14e8,
+0x140a,0x140a,0x140a,0x140a,0x14e8,0x140a,0x14e8,0x14e8,0x140a,0x140a,0x140a,0x140a,0x14e8,0x140a,0x140a,0x140a,
+0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x140a,0x1707,0x1707,0x17d6,0x17d6,0x140d,0x140d,0x140d,
+0x140a,0x140a,0x140a,0x140d,0x140d,0x140d,0x140d,0x140d,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,
+0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,
+0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,
+0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1413,0x1410,0x1410,0x1410,0x1410,
+0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1413,0x1413,0x1413,0x1410,
+0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1410,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,
+0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,
+0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1416,0x1803,0x1803,0x1800,0x175b,0x1464,0x1464,0x1464,0x1464,
+0x1464,0x1464,0x1461,0x1461,0x1461,0x1461,0x1461,0x1461,0x1464,0x1464,0x1464,0x1464,0x1464,0x1464,0x1464,0x1464,
+0x1464,0x1464,0x1464,0x1464,0x1464,0x1464,0x1464,0x15a2,0x1470,0x1470,0x1470,0x1482,0x1482,0x1482,0x1482,0x1482,
+0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,
+0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x1482,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,
+0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,
+0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x149d,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,
+0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,
+0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x14a3,0x1a34,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,
+0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,
+0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14a6,0x14ac,0x14ac,0x14b8,0x14be,0x14be,0x14be,0x14be,0x14be,
+0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,
+0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14be,0x14b8,0x14b8,0x14b8,0x14ac,0x14ac,
+0x14ac,0x14ac,0x14ac,0x14ac,0x14ac,0x14ac,0x14ac,0x14b8,0x14bb,0x14be,0x14c1,0x14c1,0x14be,0x14c4,0x14c4,0x14af,
+0x14b2,0x1764,0x1767,0x1767,0x1767,0x15ab,0x1adc,0x1ad9,0x14b5,0x14b5,0x14b5,0x14b5,0x14b5,0x14b5,0x14b5,0x14b5,
+0x14b5,0x14b5,0x15a8,0x176d,0x1770,0x176a,0x1773,0x1773,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,
+0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,
+0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x14df,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0x153c,0x153c,0x1956,0x1956,0x1956,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,
+0x153c,0x1a22,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x153c,0x18ba,0x1956,0x1956,0x1956,0x1956,0x1956,
+0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1590,0x1590,0x1596,0x1596,0x1596,0x1590,0x1590,0x1590,
+0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,0x1596,0x1596,0x158d,0x158d,0x158d,0x158d,
+0x158d,0x158d,0x158d,0x158d,0x1596,0x1596,0x1596,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,
+0x1590,0x1590,0x1596,0x1596,0x1596,0x1596,0x1590,0x1590,0x1599,0x1590,0x1590,0x1590,0x1590,0x1704,0x1704,0x1590,
+0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1884,0x1596,0x1590,0x1590,0x1596,0x1590,0x1590,0x1590,
+0x1590,0x1590,0x1590,0x1590,0x1590,0x1596,0x1596,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,0x1590,
+0x1596,0x1590,0x1590,0x1590,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,
+0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,0x15ba,
+0x15ba,0x15ba,0x15ba,0x15ba,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,
+0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,0x15cc,
+0x15cc,0x15cc,0x15cc,0x15cc,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,
+0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,0x15d2,
+0x15d2,0x15d2,0x15d2,0x15d2,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,
+0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,0x15d5,
+0x15d5,0x15d5,0x15d5,0x15d5,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,
+0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,0x1614,
+0x1614,0x1614,0x1614,0x1605,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,
+0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x161d,0x1617,
+0x1620,0x1620,0x1620,0x1620,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,
+0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,0x1623,
+0x1623,0x1623,0x1623,0x1623,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x1635,0x163e,0x163e,0x163e,
+0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,0x163e,
+0x163e,0x163e,0x163e,0x163e,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,
+0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,0x1647,
+0x1647,0x1647,0x1647,0x1647,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,
+0x1659,0x1659,0x1659,0x1659,0x1656,0x1656,0x1656,0x164a,0x164a,0x164a,0x164a,0x164a,0x164a,0x164a,0x164a,0x1656,
+0x1656,0x164a,0x1656,0x164d,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,
+0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,0x1659,
+0x1659,0x1659,0x1659,0x1659,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,
+0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,0x167d,
+0x167d,0x167a,0x167a,0x167a,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,
+0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x1686,0x168c,0x168c,0x168c,0x1689,0x1689,0x1689,
+0x1686,0x1686,0x1686,0x1686,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,
+0x169b,0x169b,0x169b,0x169b,0x168f,0x168f,0x168f,0x168f,0x168f,0x168f,0x168f,0x16a1,0x16a1,0x1695,0x1692,0x1692,
+0x1692,0x1692,0x1692,0x1692,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,
+0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,0x169b,
+0x169b,0x169b,0x169b,0x169b,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,
+0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a7,0x16a4,0x16a4,0x16a4,0x16a4,0x16a4,
+0x16a4,0x16a4,0x16a4,0x16a4,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,
+0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,0x16aa,
+0x16aa,0x16aa,0x16aa,0x16aa,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,
+0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,0x16ce,
+0x16ce,0x16ce,0x16ce,0x16ce,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,0x16d7,
+0x16d7,0x16d7,0x16d7,0x16d7,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,
+0x16ef,0x16ef,0x16ef,0x16ef,0x16da,0x16e9,0x16e9,0x16da,0x16da,0x16da,0x16da,0x16da,0x16da,0x16e9,0x16da,0x16ec,
+0x16ec,0x16da,0x16ec,0x16da,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,
+0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,0x16ef,
+0x16ef,0x16ef,0x16ef,0x16ef,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,
+0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,0x16f8,
+0x16f8,0x16f8,0x16f8,0x16f8,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,
+0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,0x16fe,
+0x16fe,0x16fe,0x16fe,0x16fe,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,
+0x1740,0x1740,0x1740,0x1740,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,0x1956,
+0x1956,0x1956,0x1956,0x1a22,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,
+0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,0x1761,
+0x1761,0x1761,0x1761,0x1761,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,
+0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,
+0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x17a3,0x17a0,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,
+0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x179d,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,
+0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,
+0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a6,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,
+0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,
+0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17a9,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,
+0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,
+0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17bb,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,
+0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,
+0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17be,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,
+0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,
+0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c4,0x17c4,0x17c4,0x17c4,0x17c1,
+0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c1,0x17c4,0x17c4,0x17c4,
+0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c1,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,
+0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,
+0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17c4,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,
+0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,
+0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x17dc,0x18c6,0x18c6,0x18c6,0x18c6,0x18c6,0x18c6,0x18c6,0x18c6,
+0x18c6,0x18c6,0x18c6,0x18c6,0x1b3f,0x1a94,0x1a94,0x1a97,0x17df,0x17df,0x17df,0x17df,0x17df,0x17df,0x17df,0x17df,
+0x17e2,0x1890,0x1890,0x1890,0x1890,0x1890,0x1890,0x192c,0x17df,0x17df,0x17df,0x17df,0x17df,0x188d,0x188d,0x188d,
+0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,
+0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x1890,0x188d,
+0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1890,0x192c,0x192c,0x1890,0x1890,0x1890,0x1890,0x1890,
+0x1890,0x1890,0x188d,0x180f,0x1890,0x1890,0x1890,0x1a94,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x180f,0x188d,
+0x188d,0x188d,0x188d,0x188d,0x1929,0x1a0a,0x1a0a,0x1a0a,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,
+0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x188d,0x1929,0x1824,0x1824,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,
+0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,
+0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1821,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,
+0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,
+0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1824,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,
+0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x186f,0x186f,0x186f,
+0x185a,0x185a,0x185a,0x185a,0x185a,0x185a,0x185a,0x185a,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,
+0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,
+0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1872,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,
+0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,
+0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1896,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,
+0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,
+0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1899,0x1b48,0x1b48,0x1b48,0x1b48,0x1b48,
+0x1b48,0x1b48,0x1b48,0x1b48,0x1b48,0x1b48,0x1b48,0x1b48,0x18f0,0x18f0,0x18f0,0x18f0,0x1a46,0x1a46,0x18f3,0x18f3,
+0x18f3,0x18f3,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18db,0x18ed,
+0x18de,0x18e1,0x18e4,0x18f6,0x18f6,0x1995,0x18e7,0x18e7,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,
+0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,
+0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x18f0,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,
+0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x18fc,0x1902,0x18ff,0x18ff,0x18ff,
+0x18ff,0x190e,0x1914,0x18ff,0x18ff,0x18ff,0x18ff,0x190b,0x1911,0x18ff,0x18ff,0x18ff,0x18ff,0x18ff,0x18ff,0x18ff,
+0x18ff,0x18ff,0x18ff,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,
+0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1911,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,
+0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,
+0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1923,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1929,0x1a0a,
+0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,
+0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1a0a,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,
+0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,
+0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1932,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,
+0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,
+0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x1938,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,
+0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,
+0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19aa,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,
+0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,
+0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19c5,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,
+0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,
+0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19cb,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,
+0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,
+0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e6,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,
+0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,
+0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19e9,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,
+0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,
+0x19f2,0x19f2,0x19f2,0x19f2,0x19f2,0x19ef,0x19ef,0x19ef,0x1a0a,0x1a0a,0x1a0a,0x1b3c,0x1b3c,0x1a94,0x1a94,0x1a94,
+0x1a94,0x1a94,0x1a94,0x1b3c,0x1b3c,0x1b3c,0x1a94,0x1a94,0x1a0d,0x1a0d,0x1a0d,0x1a0d,0x1a0a,0x1a10,0x1a10,0x1a0a,
+0x1a10,0x1a10,0x1a94,0x1a97,0x1a94,0x1a94,0x1a94,0x1a94,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,
+0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,
+0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a49,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,
+0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,
+0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a70,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,
+0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1aaf,0x1aaf,0x1a79,0x1aaf,0x1a79,0x1a79,0x1a79,0x1a79,
+0x1a79,0x1a79,0x1a79,0x1a79,0x1a79,0x1a7f,0x1a7f,0x1a7f,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,
+0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,
+0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1a8b,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,
+0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,
+0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b1e,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,
+0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,
+0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b2a,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,
+0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,
+0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b4e,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,
+0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,
+0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0x1b51,0,0,0,0
+};
+
+static const UTrie2 propsVectorsTrie={
+    propsVectorsTrie_index,
+    propsVectorsTrie_index+5024,
+    NULL,
+    5024,
+    26204,
+    0xa40,
+    0x1420,
+    0x0,
+    0x0,
+    0x110000,
+    0x79f8,
+    NULL, 0, FALSE, FALSE, 0, NULL
+};
+
+static const uint32_t propsVectors[6999]={
+0x67,0,0,0x67,0,0x4e00000,0x67,0x80000,0x20,0x867,0,0,0xa67,0,0,0xb67,
+0,0,0xc67,0,0,0xd67,0,0,0xe67,0,0,0x1067,0,0,0x1167,0,
+0,0x1267,0,0,0x1367,0,0,0x1467,0,0,0x1567,0,0,0x1667,0,0,
+0x1767,0,0,0x1867,0,0,0x1967,0,0,0x1a67,0,0,0x1b67,0,0,0x1d67,
+0,0,0x1f67,0,0,0x2067,0,0,0x2267,0,0,0x2367,0,0,0x2467,0,
+0,0x2567,0,0,0x2767,0,0,0x2867,0x80000,0x20,0x2967,0,0,0x2a67,0,0x1600000,
+0x2b67,0,0,0x2d67,0,0,0x3167,0x20000000,0,0x3267,0x20000000,0,0x3a67,0,0,0x3b67,
+0,0,0x3c67,0,0,0x3e67,0,0,0x4067,0,0,0x4167,0,0,0x4467,0,
+0,0x4867,0,0,0x4967,0,0,0x4a67,0,0,0x5067,0,0,0x5167,0,0,
+0x5467,0,0,0x5567,0,0,0x5667,0x80000,0x20,0x5767,0,0,0x5867,0,0,0x5967,
+0,0,0x5b67,0,0,0x5c67,0,0,0x5d67,0,0,0x6067,0x80000,0x20,0x6267,0,
+0,0x6367,0,0,0x6467,0,0,0x6567,0,0,0x6f67,0,0,0x7067,0,0,
+0x7367,0x20000000,0,0x7567,0,0,0x7667,0,0,0x7767,0,0,0x7867,0,0,0x7a67,
+0,0,0x7b67,0,0,0x7c67,0,0,0x7e67,0,0,0x7f67,0,0,0x8167,0,
+0,0x8267,0,0,0x8367,0,0,0x8467,0,0,0x8567,0,0,0x8667,0,0,
+0x8767,0,0,0x8867,0,0,0x8967,0,0,0x8b67,0,0,0x8c67,0,0,0x8e67,
+0x20000000,0,0x8f67,0,0,0x9067,0,0,0x9167,0,0,0x9267,0,0,0x9367,0,
+0,0x9567,0,0,0x9667,0,0,0x9767,0,0,0x9867,0,0,0x9967,0,0,
+0x9a67,0,0,0x9c67,0,0,0x9f67,0,0,0xa167,0,0,0xa367,0,0,0xa467,
+0,0,0xa567,0,0,0xa667,0,0,0xa767,0,0,0xa867,0,0,0xa967,0,
+0,0xaa67,0,0x4e00000,0xab67,0,0x4e00000,0xac67,0,0,0xad67,0,0,0xae67,0,0,
+0xaf67,0,0,0xb167,0,0,0xb267,0,0,0xb467,0,0,0xb567,0,0,0xb767,
+0,0,0xb867,0,0,0xb967,0,0,0xba67,0,0,0xbc67,0,0,0xbd67,0,
+0,0xbe67,0,0,0xbf67,0,0,0xc067,0,0,0xc167,0,0,0xc267,0,0,
+0xc367,0,0x4e00000,0xc467,0,0x4e00000,0xc667,0,0,0xc767,0,0,0xc867,0,0,0xc967,
+0,0,0xca67,0,0,0xcc67,0,0x4e00000,0xcf67,0,0x4e00000,0xd067,0,0x4e00000,0xd267,0,
+0,0xd367,0,0,0xd467,0,0,0xd567,0,0,0xd667,0,0,0xd867,0,0,
+0xda67,0,0,0xdb67,0,0,0xdc67,0,0,0xdd67,0,0,0xde67,0,0,0xdf67,
+0,0,0xe067,0,0,0xe167,0,0,0xe267,0,0,0xe367,0,0x4e00000,0xe467,0,
+0,0xe567,0,0,0xe667,0,0,0xe767,0,0,0xe867,0,0,0xe967,0,0,
+0xea67,0,0,0xeb67,0,0,0xec67,0,0,0xed67,0,0,0xee67,0,0,0xef67,
+0,0,0xf167,0,0,0xf367,0,0,0xf567,0,0,0xf667,0,0,0xf767,0,
+0,0xf867,0,0,0xf967,0,0,0xfa67,0,0x4e00000,0xfb67,0,0,0xfc67,0,0,
+0xfd67,0,0,0xfe67,0,0,0x10167,0,0,0x10267,0,0,0x10367,0,0,0x10467,
+0,0,0x10567,0,0x4e00000,0x10667,0,0,0x10767,0,0,0x10867,0,0,0x10967,0,
+0,0x10a67,0,0,0x10b67,0,0,0x10c67,0,0,0x10d67,0,0,0x10e67,0,0,
+0x10f67,0,0,0x11067,0,0,0x11367,0,0,0x11467,0,0,0x11567,0,0,0x11667,
+0,0,0x11767,0,0,0x11867,0,0,0x11967,0,0x4e00000,0x11a67,0,0,0x11b67,0,
+0,0x11c67,0,0,0x11d67,0,0,0x11e67,0,0,0x11f67,0,0,0x12067,0,0,
+0x12167,0,0,0x12267,0,0,0x12367,0,0,0x12467,0,0,0x12567,0,0,0x12667,
+0,0,0x12767,0,0,0x12867,0,0,0x12967,0,0,0x12a67,0,0x4e00000,0x12b67,0,
+0,0x12c67,0,0,0x12d67,0,0,0x12f67,0,0,0x13067,0,0,0x13167,0,0,
+0x13267,0,0,0x13367,0,0,0x13467,0,0,0xa0067,0,0xe00000,0xa4767,0,0xe00000,0xa4f67,
+0,0xe00000,0xa5e67,0,0xe00000,0xa5f67,0,0xe00000,0xac567,0,0xe00000,0xad167,0,0xe00000,0xb0067,0,
+0xe00000,0xb1267,0,0xe00000,0xb2e67,0,0xe00000,0x11000100,0,0x900020,0x11000100,0x40000001,0x440020,0x11000100,0x40000001,0x643020,
+0x11000100,0x40000001,0xa5a040,0x11000100,0x40000001,0x116a8a0,0x11000200,0,0x900020,0x11000200,0x4000001,0xc4000b,0x11000200,0x7c00100,0x220402,0x11000200,
+0x24000000,0x14200000,0x11000200,0x24000008,0x1710000,0x11000200,0x40000001,0x1d3b020,0x11000219,0x7c00100,0x220401,0x11000219,0x7c00100,0x250401,0x11000319,0x7c00100,
+0x220401,0x11000319,0x7c00100,0x220402,0x11000319,0x7c00100,0x250400,0x11000319,0x7c00100,0x250401,0x11000419,0x7c00100,0x220400,0x11000419,0x7c00100,0x220401,
+0x11000419,0x7c00100,0x220402,0x11000419,0x7c00100,0x230400,0x11000419,0x7c00100,0x250400,0x11000419,0x7c00100,0x250401,0x11000419,0x7c00100,0x250402,0x11000519,
+0x7c00100,0x220400,0x11000519,0x7c00100,0x230400,0x11000600,0x4000400,0x200002,0x11000600,0x4000400,0x200400,0x11000600,0x7c00500,0x220400,0x11000600,0x7c00500,
+0x230400,0x11000600,0x7c00500,0x530400,0x11000600,0x7c00d00,0x230400,0x11000619,0x7c00500,0x22040f,0x11000800,0x4000010,0x1001401,0x11000800,0x4000400,0x200001,
+0x11000800,0x6800010,0x201001,0x11000800,0x7c00500,0x230401,0x11000807,0x7c00100,0x220400,0x11000807,0x7c00100,0x250400,0x1100080e,0x4000400,0x200000,0x1100080e,
+0x4000400,0x200002,0x1100080e,0x7000500,0x220402,0x1100080e,0x7c00100,0x220400,0x1100080e,0x7c00100,0x220401,0x1100080e,0x7c00100,0x220402,0x1100080e,0x7c00100,
+0x250400,0x1100080e,0x7c00100,0x250401,0x1100080e,0x7c00120,0x220402,0x1100080e,0x7c00120,0x250402,0x11000908,0x4000000,0x200000,0x11000908,0x7c00100,0x220400,
+0x11000908,0x7c00100,0x220401,0x11000908,0x7c00100,0x250400,0x11000908,0x7c00100,0x250401,0x11000a03,0x4000000,0x200400,0x11000a03,0x4000000,0x201000,0x11000a03,
+0x4000000,0x270000,0x11000a03,0x7c00100,0x220400,0x11000a03,0x7c00100,0x220402,0x11000a03,0x7c00100,0x250400,0x11000a03,0x7c00500,0x230400,0x11000a03,0xc000010,
+0x1049400,0x11000b13,0x2802500,0x962460,0x11000b13,0x4000000,0x200000,0x11000b13,0x4000000,0x201000,0x11000b13,0x4000000,0x230400,0x11000b13,0x4000002,0x400000,
+0x11000b13,0x4000010,0x200000,0x11000b13,0x7c00100,0x2633800,0x11000c00,0x80000000,0x218960,0x11000c02,0x2802100,0x962460,0x11000c02,0x2802400,0x962460,0x11000c02,
+0x4000000,0x200000,0x11000c02,0x4000000,0x1329400,0x11000c02,0x4000000,0x1329800,0x11000c02,0x4000000,0x1500000,0x11000c02,0x6800000,0x1329800,0x11000c02,0x7c00100,
+0x230400,0x11000c02,0x7c00100,0x230401,0x11000c02,0x7c00100,0x230402,0x11000c02,0x7c00500,0x230400,0x11000c02,0x7d00100,0x230400,0x11000f01,0x2802400,0x962460,
+0x11000f0a,0x2802100,0x962460,0x11000f0a,0x2802400,0x962460,0x11000f0a,0x2806400,0x962460,0x11000f0a,0x4000000,0x200000,0x11000f0a,0x6800100,0x962540,0x11000f0a,
+0x7c00100,0x230400,0x11000f0a,0x7c00100,0x230401,0x11001004,0x2802100,0x962460,0x11001004,0x2802400,0x962460,0x11001004,0x2806400,0x962460,0x11001004,0x4000000,
+0x200000,0x11001004,0x4000000,0x1500000,0x11001004,0x6800100,0x962540,0x11001004,0x6800100,0x962541,0x11001004,0x7c00100,0x230400,0x11001004,0x7c00100,0x230401,
+0x11001110,0x2802100,0x962460,0x11001110,0x2802400,0x962460,0x11001110,0x2806400,0x962460,0x11001110,0x6800100,0x962540,0x11001110,0x7c00100,0x230400,0x11001110,
+0x7c00100,0x230401,0x1100120f,0x2802100,0x962460,0x1100120f,0x2802400,0x962460,0x1100120f,0x2806400,0x962460,0x1100120f,0x6800100,0x962540,0x1100120f,0x7c00100,
+0x230400,0x1100131f,0x2802100,0x962460,0x1100131f,0x2802400,0x962460,0x1100131f,0x2806400,0x962460,0x1100131f,0x4000000,0x200000,0x1100131f,0x6800000,0x1329800,
+0x1100131f,0x6800100,0x962540,0x1100131f,0x6800100,0x962541,0x1100131f,0x7c00100,0x230400,0x1100131f,0x7c00100,0x230401,0x11001423,0x2802100,0x962460,0x11001423,
+0x2806400,0x962460,0x11001423,0x6800100,0x962540,0x11001423,0x6800100,0x962541,0x11001423,0x7c00100,0x230400,0x11001423,0x7c00100,0x230401,0x11001524,0x2802100,
+0x962460,0x11001524,0x2802100,0x962461,0x11001524,0x2806400,0x962460,0x11001524,0x6800000,0x1329800,0x11001524,0x6800100,0x962540,0x11001524,0x7c00100,0x230400,
+0x11001615,0x2802100,0x962460,0x11001615,0x2806400,0x962460,0x11001615,0x6800100,0x962540,0x11001615,0x6800100,0x962541,0x11001615,0x7c00100,0x230400,0x1100171a,
+0x2802100,0x962460,0x1100171a,0x2806400,0x962460,0x1100171a,0x6800000,0x1329800,0x1100171a,0x6800100,0x962540,0x1100171a,0x6800100,0x962541,0x1100171a,0x7c00100,
+0x230400,0x11001900,0x4000000,0x1600000,0x11001926,0x2802100,0x1862460,0x11001926,0x2802400,0x1862460,0x11001926,0x2806100,0x1862460,0x11001926,0x4000000,0x200000,
+0x11001926,0x4000010,0x400000,0x11001926,0x6800000,0x1329800,0x11001926,0x7800100,0x1830142,0x11001926,0x7c00100,0x1830000,0x11001926,0x7c00900,0x1830000,0x11001926,
+0x7e00100,0x1830000,0x11001a18,0x2802100,0x1862460,0x11001a18,0x2802400,0x1862460,0x11001a18,0x6800000,0x1329800,0x11001a18,0x7800100,0x1830142,0x11001a18,0x7c00100,
+0x1830000,0x11001a18,0x7c00100,0x1830002,0x11001a18,0x7c00900,0x1830000,0x11001a18,0x7e00100,0x1830000,0x11001d0c,0x7c00100,0x230400,0x11001d0c,0x7c00100,0x250400,
+0x11001e12,0x7c00100,0x2230500,0x11001e12,0x7c00100,0x2330520,0x11001e12,0x7c80100,0x2330520,0x11002619,0x7c00100,0x220401,0x11002619,0x7c00100,0x220402,0x11002619,
+0x7c00100,0x250401,0x1100270e,0x4000400,0x200001,0x1100270e,0x4000400,0x200002,0x1100270e,0x4000400,0x500001,0x1100270e,0x7c00100,0x220401,0x1100270e,0x7c00100,
+0x250401,0x11002800,0x80000,0x918820,0x11002800,0x80000,0x1c18020,0x11002800,0x180000,0x918820,0x11002800,0x4000001,0x445801,0x11002800,0x4000001,0x445802,
+0x11002800,0x4000001,0xc4000b,0x11002800,0x6800000,0x201c00,0x11002800,0x6800020,0x201c00,0x11002800,0x24000000,0x200000,0x11002800,0x24000000,0x200002,0x11002800,
+0x24000000,0x810000,0x11002800,0x24000000,0x1410000,0x11002800,0x24000000,0x1500000,0x11002800,0x24000000,0x1500002,0x11002800,0x24000002,0x400000,0x11002800,0x24000006,
+0xc0000b,0x11002800,0x24000008,0x1410000,0x11002800,0x24000008,0x1710000,0x11002800,0x24000020,0x1001400,0x11002800,0x24000020,0x1500002,0x11002800,0x2c000010,0x1248000,
+0x11002800,0x2c000010,0x15248002,0x11002800,0x40000001,0x63b020,0x11002800,0x40080000,0x918820,0x11002801,0x80000,0xaa65620,0x11002801,0x82000,0x962460,0x11002900,
+0x4000000,0x20000e,0x11002900,0x4000000,0x20000f,0x11002900,0x4000020,0x20000e,0x11002900,0x4000020,0x20000f,0x11002900,0x4000020,0x81000e,0x11002900,0x4000020,
+0x81000f,0x11002900,0x4000020,0x141000e,0x11002900,0x4000020,0x141000f,0x11002900,0x4000022,0x20000e,0x11002900,0x4000022,0x20000f,0x11002a00,0x4000000,0x1500000,
+0x11002a00,0x4000000,0x1600000,0x11002a00,0x4000000,0x1600002,0x11002b01,0x2000,0x962460,0x11002b01,0x2802020,0x962460,0x11002c00,0x4000000,0x200000,0x11002c00,
+0x4000000,0x200002,0x11002c00,0x4000000,0x20000f,0x11002c00,0x4000020,0x200000,0x11002c00,0x7c00000,0x200000,0x11002c00,0x7c00020,0x200000,0x11002c00,0x7c00120,
+0x220405,0x11002c00,0x7c00120,0x230402,0x11002c00,0x7c00120,0x250402,0x11002c00,0x7c00120,0x250405,0x11002c19,0x7c00100,0x250400,0x11002c19,0x7c00100,0x250401,
+0x11002d00,0x4000000,0x100006,0x11002d00,0x4000000,0x200006,0x11002d19,0x7c00100,0x220402,0x11002d19,0x7c00100,0x230400,0x11002d19,0x7c00100,0x250402,0x11002e00,
+0x24000000,0x200000,0x11002e00,0x24000020,0x200000,0x11002e00,0x24000020,0x200001,0x11002e00,0x24000020,0x14200000,0x11002f00,0x24000020,0x200000,0x11002f00,0x24000020,
+0x200001,0x11002f00,0x24000020,0x200002,0x11002f00,0x24000020,0xf00000,0x11002f00,0x24000020,0x1600000,0x11002f00,0x24000022,0x1600000,0x11003000,0x24000000,0x200000,
+0x11003000,0x24000000,0x14200000,0x11003000,0x24000020,0x200000,0x11003000,0x24000020,0x810000,0x11003000,0x24000020,0x1410000,0x11003100,0x24000000,0x200000,0x11003200,
+0x24000000,0x200000,0x11003300,0x4000000,0x100003,0x11003400,0x24000000,0x100000,0x11003400,0x24000000,0x200000,0x11003500,0x24000000,0x200000,0x11003600,0x24000000,
+0x200000,0x11003600,0x24000000,0x14200000,0x11003600,0x24000020,0x200000,0x11003700,0x24000000,0x200000,0x11003700,0x24000000,0x4200000,0x11003700,0x24000000,0x4e00000,
+0x11003700,0x24000000,0x14200000,0x11003700,0x24000000,0x14e00000,0x11003700,0x24000000,0x96800000,0x11003700,0x24000020,0x4200000,0x11003800,0x4000000,0x100000,0x11003800,
+0x24000000,0x200000,0x11003800,0x24000000,0xb00000,0x11003800,0x24000000,0x1710000,0x11003800,0x24000000,0x4200000,0x11003800,0x24000000,0x4e00000,0x11003800,0x24000000,
+0x14200000,0x11003800,0x24000000,0x14b00000,0x11003800,0x24000000,0x14e00000,0x11003800,0x24000000,0x96800000,0x11005003,0x7c00100,0x220402,0x11005013,0x2802500,0x962460,
+0x11005013,0x4000020,0x200005,0x11005013,0x7c00100,0x2633801,0x11005013,0x7c00100,0x2633802,0x11005013,0x7c00100,0x2633805,0x11005019,0x7c00100,0x220402,0x11005100,
+0x24000000,0x810000,0x11005100,0x24000000,0x1410000,0x11005102,0x7000100,0x230408,0x11005102,0x7c00100,0x230404,0x11005102,0x7c00100,0x230407,0x11005102,0x7c00100,
+0x230408,0x11005102,0x7c00100,0x230409,0x11005201,0x2802400,0x962460,0x11005500,0x80000,0x1e18820,0x11005502,0x7000100,0x230408,0x11005502,0x7c00100,0x230404,
+0x11005502,0x7c00100,0x230407,0x11005502,0x7c00100,0x230408,0x11005502,0x7c00100,0x230409,0x11005667,0x1000,0,0x11020200,0x80004,0x418820,0x11020200,
+0x4000000,0x100006,0x11020200,0x4000000,0x10000f,0x11020200,0x4000400,0x100002,0x11020200,0x4000400,0x500002,0x11020200,0x6800c00,0x101000,0x11020200,0x24000000,
+0x100000,0x11020200,0x24000000,0x1400000,0x11020200,0x24000000,0x1500000,0x11020200,0x24000000,0x1600000,0x11020200,0x24000000,0x14200000,0x11020200,0x24000020,0x100000,
+0x11020200,0x24000020,0x1600000,0x11020219,0x7c00100,0x12040f,0x11020219,0x7c00100,0x220400,0x11020219,0x7c00100,0x220401,0x11020219,0x7c00100,0x250400,0x11020319,
+0x7c00100,0x220400,0x11020319,0x7c00100,0x220401,0x11020319,0x7c00100,0x220402,0x11020319,0x7c00100,0x250400,0x11020319,0x7c00100,0x250402,0x11020319,0x7d00100,
+0x220402,0x11020419,0x7c00100,0x220401,0x11020519,0x7c00100,0x220400,0x11020600,0x4000400,0x100002,0x11020600,0x4000400,0x200400,0x11020600,0x7c00500,0x130400,
+0x11020600,0x7c00d00,0x130400,0x11020701,0x2802400,0x962460,0x11020701,0x2802400,0x962461,0x11020701,0x2802400,0xc62460,0x1102080e,0x7c00100,0x220400,0x1102080e,
+0x7c00100,0x250400,0x11020908,0x7c00100,0x220400,0x11020908,0x7c00100,0x220401,0x11020908,0x7c00100,0x250400,0x11020908,0x7c00100,0x250401,0x11022800,0x24000000,
+0x100000,0x11022800,0x24000000,0x200000,0x11022800,0x24000000,0x200002,0x11022800,0x24000000,0x401000,0x11022800,0x24000000,0xf00002,0x11022800,0x24000000,0xf0ac02,
+0x11022800,0x24000000,0x1500000,0x11022800,0x24000002,0x100000,0x11022800,0x24000002,0x370000,0x11022800,0x24000002,0x470000,0x11022800,0x24000006,0x400000,0x11022800,
+0x24000008,0x1710000,0x11022800,0x24000008,0x1712c00,0x11022800,0x24000020,0x100000,0x11022800,0x24000020,0x1500000,0x11022800,0x24000020,0x1500002,0x11022900,0x4000000,
+0x10000e,0x11022900,0x4000000,0x10000f,0x11022919,0x7c00100,0x12040f,0x11022c00,0x4000000,0x100002,0x11022c00,0x4000000,0x1500002,0x11022c00,0x4000000,0x1600002,
+0x11022c00,0x4000000,0x1410000f,0x11022c00,0x7c00120,0x120405,0x11022c0e,0x7c00100,0x250401,0x11022c19,0x7c00100,0x150401,0x11022d00,0x4000000,0x100006,0x11022d00,
+0x4000000,0x200006,0x11022d19,0x7c00100,0x120402,0x11022d19,0x7c00100,0x150402,0x11022e00,0x24000000,0x200000,0x11022e00,0x24000020,0x100000,0x11022e00,0x24000020,
+0x14100000,0x11022f00,0x24000020,0x100000,0x11022f00,0x24000020,0x100001,0x11022f00,0x24000020,0x100002,0x11023000,0x24000000,0x100000,0x11023300,0x4000000,0x100002,
+0x11023300,0x4000000,0x100003,0x11023300,0x4000100,0x120403,0x11023300,0x4000100,0x150403,0x11023300,0x4000100,0x14150403,0x11023400,0x24000000,0x100000,0x11023500,
+0x24000000,0x100000,0x11023600,0x24000000,0x100000,0x11023600,0x24000020,0x100000,0x11023600,0x24000020,0x14100000,0x11023700,0x24000000,0x4100000,0x11023700,0x24000000,
+0x4e00000,0x11023700,0x24000000,0x14100000,0x11023700,0x24000000,0x14e00000,0x11023700,0x24000020,0x100000,0x11023700,0x24000020,0x4100000,0x11023700,0x24000020,0x14100000,
+0x11023800,0x4000000,0x100000,0x11023800,0x24000000,0x200000,0x11024e67,0,0,0x11025600,0x4000000,0x100000,0x11042a00,0x4000000,0x1600000,0x11045700,
+0x4000000,0x20000a,0x11045700,0x4000020,0x20000a,0x11045712,0x7c00100,0xe3040a,0x11045712,0x7c80100,0xe3040a,0x11045716,0x7c00100,0xe30c0a,0x11045716,0x7c00100,
+0x2530c0a,0x11063d00,0x4000001,0x445811,0x11065700,0x4000000,0x810011,0x11065700,0x4000000,0xe00011,0x11065700,0x4000000,0x1410011,0x11065700,0x4000000,0x1500011,
+0x11065700,0x4000000,0x1600011,0x11065700,0x4000006,0xe70011,0x11065700,0x4000008,0xe00011,0x11065700,0x4000008,0xe02c11,0x11065700,0x4000010,0x871411,0x11065700,
+0x4000010,0x1201411,0x11065700,0x4000010,0x1271011,0x11065700,0x4000020,0xe00011,0x11065700,0x4000400,0xe00011,0x11065700,0x4000420,0xe00011,0x11065700,0x6800000,
+0xe01c11,0x11065700,0x6800040,0xe29811,0x11065700,0xc000010,0x80ac11,0x11065700,0xc000010,0xb48011,0x11065719,0x7c00100,0xe20411,0x11065719,0x7c00100,0xe50411,
+0x11065719,0x7c00140,0xe20411,0x11065719,0x7c00140,0xe50411,0x11080100,0x6800000,0x201c00,0x11080100,0x68000c0,0x19329800,0x11080100,0x24000000,0x200000,0x11080100,
+0x24000000,0x810000,0x11080100,0x24000000,0x1410000,0x11080100,0x24000000,0x1500000,0x11080100,0x24000000,0x1600000,0x11080100,0x24000000,0x1b00000,0x11080100,0x24000000,
+0x2410000,0x11080100,0x24000000,0x18200000,0x11080100,0x24000006,0xd70000,0x11080100,0x24000008,0x1713c00,0x11080100,0x24000008,0x1714000,0x11080100,0x24000010,0x1001400,
+0x11080100,0x24000010,0x1071000,0x11080100,0x24000010,0x1071400,0x11080100,0x24000020,0x200000,0x11080100,0x24000020,0x400000,0x11080100,0x24000020,0x1600000,0x11080100,
+0x24000400,0x200000,0x11080100,0x24000420,0x200000,0x11080100,0x2c000010,0xb48000,0x11080100,0x2c000010,0x100ac00,0x11080100,0x44000001,0x1a45800,0x11080119,0x7c00100,
+0x220400,0x11080119,0x7c00100,0x250400,0x11080119,0x7c001c0,0x220400,0x11080119,0x7c001c0,0x250400,0x11080200,0x4000400,0x200002,0x11080200,0x24000000,0x200000,
+0x11080200,0x24000000,0x1500000,0x11080200,0x24000000,0x1600000,0x11080200,0x24000020,0x200000,0x110a1e12,0x7c00100,0x2130480,0x110a1e12,0x7c80100,0x2130480,0x110a3000,
+0x24000000,0x34e00000,0x110a3000,0x24100000,0x810001,0x110a3000,0x24100000,0x1410001,0x110a3700,0x24000000,0x34200000,0x110a3d00,0x4000000,0xe00000,0x110a3d00,0x4000000,
+0xe00002,0x110a3d00,0x24000000,0xe00000,0x110a3d11,0x7c00300,0xe30000,0x110a3d11,0x7c00900,0x1230400,0x110a3d12,0x2802400,0x962460,0x110a3e14,0x7c00100,0xe30000,
+0x110a3e14,0x7c00100,0xe30001,0x110a3e14,0x7c00100,0x2530000,0x110a3e14,0x7c00900,0x1230000,0x110a3e14,0x7c00900,0x1230001,0x110a3f16,0x7c00100,0xe30c00,0x110a3f16,
+0x7c00100,0xe30c01,0x110a3f16,0x7c00100,0x2530c00,0x110a3f16,0x7c00900,0x1230c00,0x110a3f16,0x7c00900,0x1230c01,0x110a4005,0x7c00100,0xe30400,0x110a4112,0x7c00100,
+0xe30402,0x110a4112,0x7c80100,0xe30402,0x110a4400,0x4000000,0xe00000,0x110a4412,0x4000000,0xe00002,0x110a4412,0x4000000,0xe00003,0x110a4416,0x4000000,0xe00c03,
+0x110a4500,0x4000000,0xe0000d,0x110a4516,0x4000000,0xe00c0d,0x110a4711,0x7c40300,0xe30000,0x110a4f11,0x7c00300,0xe30001,0x110a4f11,0x7c40300,0xe30000,0x110a5300,
+0x4000000,0x810010,0x110a5300,0x4000000,0xe00002,0x110a5300,0x4000000,0xe00010,0x110a5300,0x4000000,0x1410010,0x110a5300,0x4000002,0xe70010,0x110a5300,0x4000008,
+0x810010,0x110a5300,0x4000008,0x1410010,0x110a5300,0x6800000,0xe01c02,0x110a5300,0x6800000,0xe01c10,0x110a5400,0x4000000,0x81000c,0x110a5400,0x4000000,0xe0000c,
+0x110a5400,0x4000000,0x141000c,0x110a5400,0x4000000,0x150000c,0x110a5400,0x4000000,0x160000c,0x110a5400,0x4000002,0xe7000c,0x110a5400,0x4000010,0x87140c,0x110a5400,
+0x4000010,0xe7000c,0x110a5400,0x4000010,0x120140c,0x110a5400,0x4000010,0x127100c,0x110a5400,0x4000020,0xe0000c,0x110a5400,0x4000026,0xe7000c,0x110a5400,0xc000010,
+0x80ac0c,0x110a5400,0xc000010,0xb4800c,0x11400c0c,0x4000010,0xb00000,0x11400c0c,0x4000010,0x1071400,0x11400c0c,0xc000010,0xb48000,0x11400c16,0x7c00900,0x230400,
+0x11400f40,0xc000010,0x448000,0x11400f54,0xc000010,0x448000,0x11401d89,0x4000000,0x200000,0x11403dbf,0x4000000,0xe00000,0x114457b4,0x4000004,0x120000a,0x114457b4,
+0x4000008,0x81000a,0x114457b4,0x4000008,0x141000a,0x114457b4,0x4000010,0x87000a,0x114457b4,0xc000010,0x84800a,0x114457bd,0x3802500,0x126246a,0x114457bd,0x7c00d00,
+0x2530c0a,0x114a3db4,0x24000000,0x810000,0x114a3db4,0x24000000,0x1410000,0x114a3db4,0x24000008,0x810000,0x114a3db4,0x24000008,0x1410000,0x114a3db4,0x24000010,0x870000,
+0x114a3db4,0x2c000010,0x848000,0x114a3dba,0x4000000,0xe00000,0x114a3dba,0x24000000,0xe00000,0x114a3dba,0x24000002,0x1200000,0x114a3dba,0x24000002,0x14e00000,0x114a3dba,
+0x24000008,0x810000,0x114a3dba,0x24000008,0x1410000,0x114a3dbd,0x7c00900,0x930c00,0x114a3dbd,0x7c00900,0xe30c00,0x114a3dbf,0x7c00300,0xe30000,0x114a3ebd,0x7000400,
+0x1200c02,0x114a3fb4,0x4000004,0x1200000,0x114a3fbd,0x7c00d00,0x2530c00,0x114a42bf,0x4000000,0xe00000,0x114a42bf,0x4000000,0xe0000f,0x114a44bf,0x4000000,0xe00002,
+0x114a44bf,0x4000000,0xe00003,0x114a44bf,0x4000000,0x14e00003,0x114a45bf,0x4000000,0xe00002,0x114a45bf,0x4000000,0xe0000d,0x1180090a,0x2802400,0x962460,0x11800c1e,
+0x2802100,0x962460,0x11800c1e,0x2802500,0x962460,0x11800f27,0x2802400,0x962460,0x11800f34,0x2802400,0x962460,0x11820700,0x2802400,0x962460,0x11820700,0x2802500,
+0x962460,0x118a3dc0,0x2802400,0x962460,0x118a3ebd,0x2802400,0x962460,0x11c00904,0x2802400,0x962460,0x11c00908,0x2802400,0x962460,0x11c00c20,0xc000010,0xb48000,
+0x11c00c23,0x6800000,0x1329800,0x11c00f6d,0x6800000,0x1329800,0x11c01072,0x6800000,0x1329800,0x11c01176,0x6800000,0x1329800,0x11c0127a,0x6800000,0x1329800,0x11c0147e,
+0x4000000,0x200000,0x11c0147e,0x6800000,0x1329800,0x11c01682,0x6800000,0x1329800,0x11c051fa,0x7c00100,0x230408,0x20000067,0x1000,0,0x20000b13,0x2802400,
+0x962460,0x20000b13,0x2802500,0x962460,0x20001b27,0x2802100,0x962460,0x20001b27,0x2802100,0x962461,0x20001b27,0x2802400,0x962460,0x20001b27,0x2806400,0x962460,
+0x20001b27,0x2902100,0x962462,0x20001b27,0x4000000,0x200000,0x20001b27,0x4000000,0x400000,0x20001b27,0x4000000,0x500000,0x20001b27,0x4000000,0x810000,0x20001b27,
+0x4000000,0xb00000,0x20001b27,0x4000000,0xc0000b,0x20001b27,0x4000000,0x1410000,0x20001b27,0x4000010,0xb00000,0x20001b27,0x4000010,0xc00000,0x20001b27,0x6800000,
+0x1329800,0x20001b27,0x6800100,0x462540,0x20001b27,0x6800400,0x962540,0x20001b27,0x7c00100,0x230400,0x20001b27,0x7c00100,0x230401,0x20002619,0x7c00100,0x220401,
+0x20002a00,0x4000000,0x1600000,0x20004b67,0,0x1900000,0x20004c67,0,0x1900000,0x20004d67,0,0x1900000,0x20006d67,0x1000,0,0x20006e67,
+0x1000,0,0x20026d67,0,0,0x20026e67,0,0,0x200a4a12,0x7c00100,0x1f304c1,0x200a4a12,0x7c00100,0x20304e1,0x21005600,0x4000000,
+0x700000,0x21022a00,0x4000000,0x1600000,0x30000419,0x7c00100,0x220400,0x30000419,0x7c00100,0x220401,0x30000419,0x7c00100,0x250400,0x30000419,0x7c00100,0x250401,
+0x30000519,0x7c00100,0x220400,0x30000600,0x4000400,0x200400,0x30000600,0x7c00500,0x230400,0x30000605,0x4000400,0x200400,0x3000080e,0x7c00100,0x220400,0x30000908,
+0x2000,0x962460,0x30000908,0x7c00100,0x220400,0x30000908,0x7c00100,0x220401,0x30000908,0x7c00100,0x250400,0x30000908,0x7c00100,0x250401,0x30000a03,0x4000006,
+0x400400,0x30000c02,0x4000000,0x200000,0x30000c02,0x7c00100,0x230400,0x30000d22,0x2802100,0x962460,0x30000d22,0x2802400,0x962460,0x30000d22,0x2802500,0x962460,
+0x30000d22,0x4000000,0x200000,0x30000d22,0x4000010,0x200000,0x30000d22,0x7c00100,0x230400,0x30000d22,0xc000010,0x248000,0x30000d22,0x80000000,0x218960,0x30000e25,
+0x2802500,0x962460,0x30000e25,0x7c00100,0x230400,0x30001821,0x2802100,0x962460,0x30001821,0x2806400,0x962460,0x30001821,0x4000000,0x200000,0x30001821,0x6800100,
+0x962540,0x30001821,0x6800100,0x962541,0x30001821,0x7c00100,0x230400,0x30001b27,0x2802100,0x962460,0x30001b27,0x2802400,0x962460,0x30001b27,0x4000000,0x200000,
+0x30001b27,0x4000000,0x400000,0x30001b27,0x7c00100,0x230400,0x30001c1c,0x2802100,0x1862460,0x30001c1c,0x2802400,0x1862460,0x30001c1c,0x2806400,0x1862460,0x30001c1c,
+0x4000000,0x200000,0x30001c1c,0x6800100,0x1862400,0x30001c1c,0x6800100,0x1862540,0x30001c1c,0x7c00100,0x1830000,0x30001c1c,0x7c00100,0x1830001,0x30001c1c,0xc000010,
+0x448000,0x30001f0b,0x4000000,0x200000,0x30001f0b,0x4000010,0x200000,0x30001f0b,0x4000010,0x400000,0x30001f0b,0x6800000,0x200000,0x30001f0b,0x7c00100,0x230400,
+0x30001f0b,0xc000010,0x248000,0x30002006,0x7c00100,0x250400,0x30002128,0x4000000,0x200000,0x30002128,0x7c00100,0x230400,0x30002128,0xc000010,0x248000,0x3000221d,
+0x4000000,0x810000,0x3000221d,0x4000000,0x1410000,0x3000221d,0x4000001,0x445800,0x3000221d,0x7c00100,0x230400,0x30002300,0x4000010,0x400000,0x30002320,0x7c00100,
+0x230400,0x30002417,0x2802100,0x1862460,0x30002417,0x2802400,0x1862460,0x30002417,0x2806400,0x1862460,0x30002417,0x2882000,0x1862460,0x30002417,0x4000000,0x200000,
+0x30002417,0x4000000,0x400000,0x30002417,0x4000000,0x1600000,0x30002417,0x4000010,0x400000,0x30002417,0x4000010,0x1200000,0x30002417,0x6800000,0x1329800,0x30002417,
+0x6800100,0x1862540,0x30002417,0x7c00100,0x1830000,0x30002417,0x7d00100,0x1830000,0x3000251b,0x80000,0xc18820,0x3000251b,0x2802100,0x962460,0x3000251b,0x3c02100,
+0x962460,0x3000251b,0x4000000,0x200000,0x3000251b,0x4000006,0x500000,0x3000251b,0x4000010,0x400000,0x3000251b,0x4000010,0xb70000,0x3000251b,0x4000800,0x200000,
+0x3000251b,0x6800000,0x1329800,0x3000251b,0x7c00100,0x230400,0x3000251b,0x7c00900,0x230400,0x3000251b,0xc000010,0xb48000,0x3000251b,0x12882000,0x962460,0x30002800,
+0x24000000,0x200000,0x30002800,0x2c000010,0x1248002,0x30002800,0x2c000010,0x15248002,0x30002a00,0x4000000,0x1600000,0x30002b01,0x2000,0x962460,0x30002b01,0x2000,
+0x8962460,0x30002c00,0x4000000,0x200000,0x30002c00,0x7c00100,0x14220405,0x30002d19,0x7c00100,0x250400,0x30002e00,0x24000000,0x200000,0x30003000,0x24000000,0x200000,
+0x30003000,0x24000000,0x4200000,0x30003100,0x24000000,0x200000,0x30003600,0x24000000,0x200000,0x30003700,0x24000000,0x4200000,0x3000392e,0x24000000,0x200000,0x30005013,
+0x7c00100,0x2633801,0x30005600,0,0x918820,0x30020600,0x4000400,0x500400,0x30020701,0x2802400,0x962460,0x30020701,0x2802400,0xc62460,0x300a3a11,0x4020000,
+0xe00000,0x300a3a11,0x4020000,0xe00002,0x300a3b11,0x4020000,0xe00002,0x300a3c00,0x4008000,0xe00000,0x300a3c00,0x4010000,0xe00000,0x300a3d11,0x7c00300,0xe30002,
+0x300a4305,0x7c00100,0xe30400,0x300a4611,0x7c40300,0xe30000,0x300a4829,0x7c00100,0xe30400,0x300a4829,0x7c00900,0x1230400,0x300a4929,0x4000000,0xe00000,0x3040258f,
+0x4000010,0x400000,0x3040258f,0x4000010,0xb70000,0x3040258f,0xc000010,0xb48000,0x304028af,0x4000001,0xc41c0b,0x304a3dbf,0x4000000,0xe00000,0x30800c1e,0x2802100,
+0x962460,0x30c01c87,0x6800000,0x1329800,0x3100080e,0x7c00120,0x220402,0x3100080e,0x7c00120,0x250402,0x31005167,0x1000,0,0x3100581e,0x4000000,0x200000,
+0x3100581e,0x7c00100,0x230400,0x3100590d,0x7c00100,0x230400,0x31005a09,0x7c00100,0x220400,0x31005a09,0x7c00100,0x250400,0x31005b00,0x4000000,0x200000,0x31005c00,
+0x80000,0x918820,0x31005c00,0x2802000,0x962460,0x31005c00,0x2802400,0x962460,0x31005c00,0x4000000,0x200000,0x31005c00,0x4000000,0x200001,0x31005c00,0x6800000,
+0x962540,0x31005c00,0x6800400,0x962540,0x31005c01,0x2802400,0x962460,0x31005d00,0x4000020,0x200005,0x31005d00,0x6800020,0x1329805,0x31005d00,0x7c00120,0x220405,
+0x31005d00,0x7c00120,0x250405,0x31006000,0x82000,0x8962460,0x31006000,0x180000,0x918820,0x310a5e11,0x7c40300,0xe30000,0x310a5f11,0x7c00300,0xe30001,0x32000419,
+0x7c00100,0x250400,0x3200080e,0x4000020,0x200000,0x3200080e,0x7c00100,0x220400,0x3200080e,0x7c00100,0x250400,0x32000908,0x7c00100,0x220400,0x32000908,0x7c00100,
+0x250400,0x32000c02,0x7c00100,0x230400,0x32000e25,0x7c00100,0x230400,0x32001d0c,0x7c00100,0x230400,0x32002800,0x80000,0x1e18820,0x32002800,0x80020,0x218820,
+0x32002800,0x4000001,0x445802,0x32002800,0x24000000,0x200000,0x32002800,0x24000000,0x200002,0x32002800,0x24000020,0x200000,0x32002800,0x2c000010,0x1248002,0x32002919,
+0x7c00100,0x22040f,0x32002a00,0x4000000,0x1600000,0x32002b01,0x2000,0x962460,0x32002b01,0x2802000,0x962460,0x32002b01,0x2802020,0x962460,0x32002c00,0x4000000,
+0x200000,0x32002c00,0x4000020,0x200000,0x32002c00,0x4000020,0x200005,0x32002c00,0x7c00120,0x220405,0x32002c00,0x7c00120,0x250405,0x32002e00,0x24000020,0x200000,
+0x32002f00,0x24000020,0x200000,0x32003000,0x24000000,0x200000,0x32003000,0x24000020,0x200000,0x32003500,0x24000000,0x200000,0x32003600,0x24000020,0x200000,0x32003600,
+0x24000020,0x14200000,0x32003700,0x24000000,0x200000,0x32003700,0x24000000,0x4100000,0x32003700,0x24000000,0x4200000,0x32003700,0x24000000,0x14200000,0x32003800,0x24000000,
+0x810000,0x32003800,0x24000000,0x1410000,0x32005102,0x4000000,0x1500008,0x32005502,0x7c00100,0x230400,0x32006108,0x7c00100,0x220400,0x32006108,0x7c00100,0x250400,
+0x3200622a,0x2802100,0x962460,0x3200622a,0x2806000,0x962460,0x3200622a,0x7c00100,0x230400,0x3200632b,0x2802100,0x962460,0x3200632b,0x2806000,0x962460,0x3200632b,
+0x7c00100,0x230400,0x3200642c,0x2802100,0x962460,0x3200642c,0x7c00100,0x230400,0x3200652d,0x2802100,0x962460,0x3200652d,0x7c00100,0x230400,0x32006600,0x24000020,
+0x200000,0x32006700,0x24000020,0x200000,0x32006800,0x24000020,0x200000,0x32006800,0x24000020,0x14200000,0x32006900,0x24000020,0x200000,0x32006900,0x24000020,0x810000,
+0x32006900,0x24000020,0x1410000,0x32006a00,0x24000020,0x200000,0x32006a00,0x24000020,0x200001,0x32006a00,0x24000020,0x200002,0x32020701,0x2882000,0xc62460,0x32023300,
+0x4000000,0x100000,0x32026c01,0x12882000,0x962460,0x32026c01,0x12882000,0x8962460,0x32065700,0x4000000,0x810011,0x32065700,0x4000000,0x1410011,0x32086600,0x24000020,
+0x810000,0x32086600,0x24000020,0x1410000,0x32086900,0x24000020,0x810000,0x32086900,0x24000020,0x1410000,0x320a3600,0x24000020,0x34200000,0x320a3d11,0x7c00100,0x1230400,
+0x320a3e14,0x7c00100,0xe30010,0x320a3e14,0x7c00100,0x2530000,0x320a3f16,0x7c00100,0xe30c10,0x320a4400,0x4000000,0xe00003,0x320a4929,0x4000000,0xe00000,0x320a4f11,
+0x7c00300,0xe30001,0x320a6b16,0x7c00100,0x2530c00,0x3240638b,0xc000010,0x448000,0x324a3dc2,0x4000000,0x14e00000,0x324a3dc2,0x7c00100,0x1230400,0x324a3fbd,0x4000002,
+0x1200c00,0x324a53ba,0x24000000,0xe00000,0x32820701,0x2802000,0x962460,0x40000419,0x7c00100,0x220400,0x40000519,0x7c00100,0x220400,0x40000600,0x4000400,0x200400,
+0x4000080e,0x7c00100,0x220400,0x4000080e,0x7c00100,0x250400,0x4000080e,0x7c00100,0x250402,0x40000c02,0x2802100,0x962460,0x40000c02,0x2802400,0x962460,0x40000c02,
+0x2802500,0x962460,0x40000c02,0x4000000,0x200000,0x40000c02,0x4000000,0x1071400,0x40000c02,0x7c00100,0x230400,0x40000c02,0x80000000,0x218960,0x40000d22,0x7c00100,
+0x230400,0x40000f0a,0x7c00100,0x230400,0x40001004,0x7c00100,0x230400,0x40001110,0x2802100,0x962460,0x40001110,0x6800100,0x962540,0x4000120f,0x2802100,0x962460,
+0x4000120f,0x4000000,0x1600000,0x4000120f,0x7c00100,0x230400,0x4000131f,0x7c00100,0x230400,0x40001423,0x4000000,0x200000,0x40001423,0x4000000,0x1600000,0x40001615,
+0x2802400,0x962460,0x40001615,0x7c00100,0x230400,0x40002417,0x2802400,0x1862460,0x40002417,0x4000000,0x200000,0x40002800,0x6800000,0x201c00,0x40002800,0x24000002,
+0x200000,0x40002c00,0x4000000,0x200002,0x40003000,0x24000000,0x14200000,0x40003000,0x24000020,0x200000,0x40003700,0x24000000,0x200000,0x40003700,0x24000000,0x4200000,
+0x40003700,0x24000000,0x14200000,0x40005a09,0x7c00100,0x220400,0x40005a09,0x7c00100,0x250400,0x40005d00,0x7c00120,0x220405,0x40006f30,0x2802100,0x962460,0x40006f30,
+0x2802400,0x962460,0x40006f30,0x4000000,0x200000,0x40006f30,0x6800000,0x1329800,0x40006f30,0x6800100,0x962540,0x40006f30,0x7c00100,0x230400,0x40006f30,0xc000010,
+0xb48000,0x40007034,0x7c00100,0x1830000,0x40007117,0x4000000,0x200000,0x40007208,0x7c00100,0x220400,0x4000720e,0x7c00100,0x220400,0x4000720e,0x7c00500,0x22040e,
+0x4000720e,0x7c00500,0x22040f,0x40007219,0x7c00100,0x220400,0x40007219,0x7c00500,0x220400,0x40007219,0x7c00500,0x22040e,0x40007219,0x7c00500,0x22040f,0x40007300,
+0x24000000,0x200000,0x40007300,0x24000000,0x14200000,0x40007400,0x4000000,0x200000,0x40007531,0x7c00100,0x230400,0x40007631,0x7c00100,0x230400,0x40007835,0x4000010,
+0x400000,0x40007835,0x7c00100,0x230400,0x40007933,0x7c00100,0x230400,0x40007a32,0x6800000,0x1329800,0x40007a32,0x7c00100,0x230400,0x40007b2f,0x7c00100,0x230400,
+0x40007c00,0x4000000,0x200000,0x40020701,0x2802400,0x962460,0x40020701,0x2802400,0xc62460,0x40023300,0x4000000,0x200000,0x40027d01,0x12882000,0x962460,0x400a3700,
+0x24000000,0x34200000,0x400a3700,0x24000000,0x34e00000,0x400a4400,0x4000000,0xe0000d,0x400a4412,0x4000000,0xe00002,0x400a4412,0x4000000,0xe00003,0x400a4500,0x4000000,
+0xe0000d,0x400a5300,0x4000000,0x810010,0x400a5300,0x4000000,0x1410010,0x404077fc,0x4000000,0x200000,0x404077ff,0x4000000,0x200000,0x404077ff,0x4000000,0x400000,
+0x40c0147e,0x4000000,0x200000,0x40c051fa,0x4000000,0x200000,0x41000419,0x7c00100,0x220400,0x41000419,0x7c00100,0x250400,0x4100080e,0x7c00100,0x220400,0x4100080e,
+0x7c00100,0x250400,0x41000908,0x7c00100,0x220400,0x41000908,0x7c00100,0x250400,0x41000b13,0x2802000,0x962460,0x41000b13,0x2802100,0x962460,0x41000b13,0x4000000,
+0xb00000,0x41000c02,0x2802100,0x962460,0x41000c02,0x4000000,0x1500000,0x41000c02,0xc000010,0xb48000,0x41000f0a,0x7c00100,0x230400,0x41001004,0x7c00100,0x230400,
+0x41001423,0x7c00100,0x230400,0x41001b27,0x4000000,0x500000,0x41001d0c,0x7c00100,0x230400,0x41001d0c,0x7c00100,0x23040f,0x41001f0b,0x2802400,0x962460,0x41001f0b,
+0x4000000,0x200000,0x41001f0b,0x7c00100,0x230400,0x41002800,0x24000000,0x200000,0x41002800,0x24000000,0x400000,0x41002919,0x7c00100,0x22040e,0x41002a00,0x4000000,
+0x1600000,0x41002b01,0x2802020,0x962460,0x41002c00,0x4000000,0x200000,0x41002c00,0x7c00120,0x220405,0x41003000,0x24000000,0x200000,0x41003700,0x24000000,0x4200000,
+0x41003700,0x24000000,0x14200000,0x41003700,0x24000000,0x14e00000,0x41005d00,0x7c00120,0x220405,0x41006600,0x24000020,0x200000,0x41006600,0x24000020,0x810000,0x41006600,
+0x24000020,0x1410000,0x41007208,0x7c00100,0x22040f,0x41007219,0x7c00100,0x220400,0x41007300,0x24000000,0x200000,0x41007e0e,0x2802000,0x962460,0x41007e0e,0x4000000,
+0x200000,0x41007f0e,0x4000000,0x200000,0x41007f0e,0x7c00100,0x230400,0x41008002,0x7c00100,0x230400,0x41008137,0x2802100,0x962460,0x41008137,0x4000000,0x200000,
+0x41008137,0x6800100,0x962540,0x41008137,0x7c00100,0x230400,0x41008301,0x2802000,0x962460,0x41008407,0x4000000,0x200000,0x41008407,0x4000000,0x400000,0x41008407,
+0x4000000,0xb00000,0x41008407,0x7c00100,0x220400,0x41008407,0x7c00100,0x250400,0x4100850b,0x7c00100,0x230400,0x4100860b,0x4000000,0x200000,0x4100860b,0x7c00100,
+0x230400,0x4100870c,0x7c00100,0x220400,0x41008838,0x7c00100,0x220400,0x41008838,0x7c00100,0x250400,0x41008939,0x2802000,0x962460,0x41008939,0x2802100,0x962460,
+0x41008939,0x2806000,0x962460,0x41008939,0x4000000,0x200000,0x41008939,0x4000000,0x400000,0x41008939,0x7c00100,0x230400,0x41008939,0xc000010,0x448000,0x41008a00,
+0x4000400,0x200400,0x41008b3b,0x4000000,0x1800000,0x41008b3b,0x6800000,0x1329800,0x41008b3b,0x7c00100,0x1830000,0x41008b3b,0x7e00100,0x1830000,0x41008c3d,0x4000010,
+0x400000,0x41008c3d,0x7c00100,0x230400,0x41008d0e,0x7c00100,0x22040f,0x41008d19,0x7c00100,0x220400,0x41008d19,0x7c00100,0x22040f,0x41008e00,0x24000000,0x200000,
+0x41008e00,0x24000000,0x400000,0x41008e00,0x24000000,0x1710000,0x41008e00,0x24000006,0x400000,0x41008f3a,0x2802100,0x962460,0x41008f3a,0x2806000,0x962460,0x41008f3a,
+0x4000000,0x200000,0x41008f3a,0x6800100,0x962540,0x41008f3a,0x7c00100,0x230400,0x4100903c,0x7c00100,0x230400,0x4100903c,0x7c00100,0x23040f,0x41020701,0x2802000,
+0x962460,0x41020701,0x2802000,0xc62460,0x410a3700,0x24000000,0x34200000,0x410a3700,0x24000000,0x34e00000,0x410a4412,0x4000000,0xe00003,0x410a4711,0x7c40300,0xe30000,
+0x410a4f11,0x7c00300,0xe30001,0x410a9100,0x4000000,0x800010,0x410a9100,0x4000000,0x810010,0x410a9100,0x4000000,0x870010,0x410a9100,0x4000000,0xb00010,0x410a9100,
+0x4000000,0xf00010,0x410a9100,0x4000000,0x1001410,0x410a9100,0x4000000,0x1071010,0x410a9100,0x4000000,0x1071410,0x410a9100,0x4000000,0x1410010,0x41408ac5,0x4000400,
+0x200000,0x414a82bf,0x4000000,0xe00000,0x41808300,0x2802000,0x962460,0x41c0147e,0x6800000,0x1329800,0x50000419,0x7c00100,0x220400,0x50000419,0x7c00100,0x250400,
+0x5000080e,0x7c00100,0x220400,0x50000908,0x7c00100,0x220400,0x50000908,0x7c00100,0x250400,0x50000b13,0x2802500,0x962460,0x50000f0a,0x7c00100,0x230400,0x50001615,
+0x2802100,0x962460,0x50001615,0x7c00100,0x230400,0x50002b01,0x2802020,0x962460,0x50002c00,0x4000000,0x200000,0x50002c19,0x7c00100,0x220400,0x50002d19,0x7c00100,
+0x220400,0x50003000,0x24000000,0x200000,0x50003000,0x24000020,0x200000,0x50003700,0x24000000,0x4200000,0x50005d00,0x7c00120,0x220405,0x50005d00,0x7c00120,0x250405,
+0x50006108,0x7c00100,0x220400,0x50006108,0x7c00100,0x250400,0x50006600,0x24000020,0x200000,0x50007300,0x24000000,0x200000,0x50008301,0x2802400,0x962460,0x50008a00,
+0x7c00500,0x230400,0x50009257,0x2802400,0x962460,0x50009257,0x4000000,0x200000,0x50009257,0x4000010,0x1071400,0x50009257,0x6800000,0x1329800,0x50009257,0x7c00100,
+0x230400,0x50009257,0x7c00500,0x230400,0x50009257,0x7c00900,0x230400,0x50009257,0xc000010,0xb48000,0x5000933e,0x2802100,0x962460,0x5000933e,0x2802400,0x962460,
+0x5000933e,0x4000000,0x200000,0x5000933e,0x4000000,0x400000,0x5000933e,0x4000010,0x400000,0x5000933e,0x6800000,0x1329800,0x5000933e,0x6800100,0x962540,0x5000933e,
+0x6800100,0x962541,0x5000933e,0x6804400,0x962540,0x5000933e,0x7c00100,0x230400,0x5000933e,0x7c00100,0x230401,0x5000933e,0xc000010,0x448000,0x50009419,0x7c00100,
+0x220400,0x50009419,0x7c00100,0x250400,0x50009500,0x4000400,0x200400,0x5000965a,0x4000000,0x500000,0x5000965a,0x7c00100,0x230400,0x5000965a,0xc000010,0xb48000,
+0x5000975b,0x4000000,0x200000,0x5000975b,0x4000010,0x400000,0x5000975b,0x7c00100,0x230400,0x50009865,0x7c00100,0x230400,0x50009965,0x4000010,0x400000,0x50009965,
+0x7c00100,0x230400,0x50409abf,0x4000000,0x200000,0x5100080e,0x7c00100,0x220400,0x5100080e,0x7c00100,0x250400,0x51000c02,0x2802100,0x962460,0x51000c02,0x4000000,
+0x1500000,0x51000c02,0x4000020,0x200000,0x51000c02,0x7c00100,0x230400,0x51000f0a,0x7c00100,0x230400,0x51000f0a,0x7c00500,0x230400,0x51001110,0x2802100,0x962460,
+0x5100131f,0x2802100,0x962460,0x51001423,0x7c00100,0x230400,0x51001524,0x2802100,0x962460,0x51001524,0x4000000,0x200000,0x51001524,0x7c00100,0x230400,0x5100171a,
+0x2802100,0x962460,0x5100171a,0x4000000,0x200000,0x5100171a,0x4000000,0x1500000,0x5100171a,0x7c00100,0x230400,0x51001b27,0x4000000,0x200000,0x51001b27,0x4000000,
+0x400000,0x51001b27,0x4000000,0x500000,0x51001b27,0x7c00100,0x230400,0x51001c1c,0x2802100,0x1862460,0x51001c1c,0x2802500,0x1862460,0x51001c1c,0x2806400,0x1862460,
+0x51001c1c,0x4000000,0x1800000,0x51001c1c,0x6800000,0x1329800,0x51001c1c,0x6800100,0x1862400,0x51001c1c,0x6800100,0x1862540,0x51001c1c,0x6800500,0x1862400,0x51001c1c,
+0x7c00100,0x1830000,0x5100251b,0x7c00100,0x230400,0x51002619,0x7c00100,0x220400,0x51002619,0x7c00100,0x250400,0x51002800,0x80020,0x218820,0x51002c00,0x4000000,
+0x200000,0x51002d19,0x7c00100,0x230400,0x51003700,0x24000000,0x4200000,0x51003700,0x24000000,0x4e00000,0x51005201,0x2802400,0x962460,0x51005c00,0x4000000,0x200000,
+0x51006108,0x7c00100,0x220400,0x51006108,0x7c00100,0x250400,0x51006600,0x24000020,0x200000,0x51006600,0x24000020,0x810000,0x51006600,0x24000020,0x1410000,0x51007300,
+0x24000000,0x200000,0x51007300,0x24000020,0x200000,0x51008002,0x7c00100,0x230400,0x51008301,0x2802000,0x962460,0x51008301,0x2802400,0x962460,0x51008a00,0x7c00500,
+0x230400,0x51008e00,0x24000000,0x200000,0x51008e00,0x24000000,0x400000,0x51008e00,0x24000000,0x810000,0x51008e00,0x24000000,0x1400000,0x51008e00,0x24000000,0x1410000,
+0x51008e00,0x24000000,0x1710000,0x51008e00,0x24000002,0x200000,0x51008e00,0x24000500,0x230400,0x51008e00,0x2c000010,0xb48000,0x51009419,0x7c00100,0x220400,0x51009419,
+0x7c00100,0x22040e,0x51009419,0x7c00100,0x22040f,0x51009419,0x7c00100,0x250400,0x51009500,0x4000400,0x200400,0x51009500,0x7c00500,0x230400,0x51009519,0x7c00100,
+0x220400,0x51009519,0x7c00100,0x22040f,0x51009519,0x7c00100,0x230400,0x51009519,0x7c00100,0x250400,0x51009b71,0x2802100,0x962460,0x51009b71,0x6800000,0x1329800,
+0x51009b71,0x6800100,0x962540,0x51009b71,0x6804400,0x962540,0x51009b71,0x7c00100,0x230400,0x51009c52,0x2802100,0x962460,0x51009c52,0x2802400,0x962460,0x51009c52,
+0x2802d00,0x962460,0x51009c52,0x4000010,0x400000,0x51009c52,0x6800000,0x1329800,0x51009c52,0x6800100,0x962540,0x51009c52,0x7c00100,0x230400,0x51009c52,0xc000010,
+0x448000,0x51009d6d,0x6800000,0x1329800,0x51009d6d,0x7c00100,0x230400,0x51009d6d,0x7c00500,0x230400,0x51009d6d,0x7c00d00,0x230400,0x51009d6d,0xc000010,0x448000,
+0x51009e08,0x2802100,0x962460,0x51009f63,0x4000010,0x400000,0x51009f63,0x6800000,0x1329800,0x51009f63,0x7c00100,0x230400,0x51009f63,0x7c00900,0x230400,0x51009f63,
+0xc000010,0x448000,0x51009f63,0xc000010,0xb48000,0x5100a008,0x2000,0x962460,0x5100a008,0x2802400,0x962460,0x5100a008,0x4000000,0x200000,0x5100a008,0x7c00100,
+0x220400,0x5100a008,0x7c00100,0x230400,0x5100a008,0x7c00100,0x250400,0x5100a008,0x7c00500,0x230400,0x5100a16f,0x2806400,0x962460,0x5100a16f,0x6800000,0x1329800,
+0x5100a16f,0x6800100,0x962540,0x5100a16f,0x7c00100,0x230400,0x5100a16f,0xc000010,0x448000,0x5100a24f,0x2802100,0x962460,0x5100a24f,0x2802400,0x962460,0x5100a24f,
+0x6800000,0x1329800,0x5100a24f,0x7c00100,0x230400,0x5100a24f,0xc000010,0x448000,0x5100a36e,0x2802100,0x962460,0x5100a36e,0x4000000,0x200000,0x5100a36e,0x6800100,
+0x962540,0x5100a36e,0x6804400,0x962540,0x5100a36e,0x7c00100,0x230400,0x5100a442,0x2802100,0x962460,0x5100a442,0x4000000,0x200000,0x5100a442,0x6800000,0x1329800,
+0x5100a442,0x6800100,0x962540,0x5100a442,0x7c00100,0x230400,0x5100a442,0xc000010,0x448000,0x5100a500,0x4000000,0x200000,0x5100a600,0x4000000,0x200000,0x5100a601,
+0x2802000,0x962460,0x5100a76b,0x7c00100,0x230400,0x5100a868,0x7c00100,0x230400,0x5100a96c,0x4000000,0x200000,0x5100a96c,0x7c00100,0x230400,0x5100aa00,0x4000000,
+0x4e00000,0x5100ab00,0x4000000,0x4e00000,0x51086600,0x24000020,0x810000,0x51086600,0x24000020,0x1410000,0x510a4005,0x7c00100,0xe30400,0x510a4711,0x7c40300,0xe30000,
+0x510a7300,0x24000000,0x34200000,0x510aaa00,0x4000000,0x34e00000,0x5140a2f3,0x4000400,0x400000,0x514a82bf,0x4000000,0xe00000,0x51802bb1,0x2802000,0x962460,0x51c00908,
+0x2802400,0x962460,0x51c0a008,0x2802400,0x962460,0x52000f0a,0x2802100,0x962460,0x52000f0a,0x6800100,0x962540,0x52000f0a,0x7c00100,0x230400,0x52001004,0x4000000,
+0x1600000,0x52001b00,0x4000000,0x200000,0x52001c1c,0x2802100,0x1862460,0x52001c1c,0x6800100,0x1862400,0x52001c1c,0x6800500,0x1862400,0x52001e12,0x7c00100,0x2230500,
+0x52001e12,0x7c00100,0x2330520,0x52002128,0x4000002,0x400000,0x52002128,0x7c00100,0x230400,0x52002a00,0x4000000,0x1500000,0x52002a00,0x4000000,0x1600000,0x52002d00,
+0x4000000,0x200006,0x52003000,0x24000000,0x200000,0x52006108,0x7c00100,0x220400,0x52006108,0x7c00100,0x250400,0x52008301,0x2802400,0x962460,0x52008407,0x2802400,
+0x962460,0x52008407,0x7c00100,0x220400,0x52008407,0x7c00100,0x250400,0x52008b3b,0x6800000,0x1800000,0x52008b3b,0x7c00100,0x1830000,0x52008e00,0x24000000,0x400000,
+0x52009419,0x7c00100,0x250400,0x5200975b,0x4000000,0x200000,0x5200ac7e,0x2802000,0x962460,0x5200ac7e,0x2802100,0x962460,0x5200ac7e,0x2802400,0x962460,0x5200ac7e,
+0x4000010,0x200000,0x5200ac7e,0x7c00100,0x230400,0x5200ac7e,0xc000010,0x248000,0x5200ad28,0x7c00100,0x230400,0x5200ae6a,0x2802100,0x1862460,0x5200ae6a,0x2802400,
+0x962460,0x5200ae6a,0x2802400,0x1862460,0x5200ae6a,0x2806000,0x1862460,0x5200ae6a,0x4000000,0x1800000,0x5200ae6a,0x6800000,0x1329800,0x5200ae6a,0x6800100,0x1862400,
+0x5200ae6a,0x6800100,0x1862540,0x5200ae6a,0x7c00100,0x1830000,0x5200ae6a,0x7c00900,0x1830000,0x5200ae6a,0xc000010,0x1848000,0x5200b083,0x4000010,0x400000,0x5200b083,
+0x7c00100,0x230400,0x5200b083,0xc000010,0x448000,0x5200b182,0x2802400,0x962460,0x5200b182,0x4000000,0x200000,0x5200b182,0x4000010,0x400000,0x5200b182,0x7c00100,
+0x230400,0x5200b182,0xc000010,0x448000,0x5200b30a,0x2802400,0x962460,0x5200b30a,0x4000000,0x200000,0x5200b30a,0x7c00100,0x230400,0x5200b54e,0x2802100,0x962460,
+0x5200b54e,0x2802400,0x962460,0x5200b54e,0x4000000,0x200000,0x5200b54e,0x4000010,0x400000,0x5200b54e,0x6800000,0x1329800,0x5200b54e,0x6800100,0x962540,0x5200b54e,
+0x6804400,0x962540,0x5200b54e,0x7c00100,0x230400,0x5200b54e,0xc000010,0x448000,0x5200b61c,0x4000000,0x1800000,0x5200b61c,0x6800500,0x1862400,0x5200b61c,0x7c00100,
+0x1830000,0x5200b61c,0x7c00900,0x1830000,0x5200b77f,0x2802100,0x1862460,0x5200b77f,0x2802400,0x1862460,0x5200b77f,0x4000000,0x1800000,0x5200b77f,0x4000010,0x1800000,
+0x5200b77f,0x7c00100,0x1830000,0x5200b77f,0x7c00500,0x1830000,0x5200b77f,0x7c00900,0x1830000,0x5200b77f,0x7e00100,0x1830000,0x5200b873,0x2802100,0x962460,0x5200b873,
+0x2806400,0x962460,0x5200b873,0x6800000,0x1329800,0x5200b873,0x6800100,0x962540,0x5200b873,0x6800400,0x962540,0x5200b873,0x7c00100,0x230400,0x5200b873,0xc000010,
+0x448000,0x5200b912,0x7c00100,0x2230500,0x5200b912,0x7c00100,0x2330520,0x5200ba74,0x4000000,0x200000,0x5200ba74,0x4000010,0x400000,0x5200ba74,0x7c00100,0x230400,
+0x5200bb85,0x4000000,0x200000,0x5200bb85,0x7c00100,0x230400,0x5200bc75,0x4000000,0x400000,0x5200bc75,0x4000010,0x400000,0x5200bc75,0x7c00100,0x230400,0x5200bd7d,
+0x4000000,0x200000,0x5200bd7d,0x7c00100,0x230400,0x5200be7a,0x4000000,0x200000,0x5200be7a,0x7c00100,0x230400,0x5200bf58,0x7c00100,0x230400,0x5200c002,0x4000000,
+0x200000,0x5200c178,0x2802000,0x962460,0x5200c178,0x2802100,0x962460,0x5200c178,0x2802400,0x962460,0x5200c178,0x2806400,0x962460,0x5200c178,0x4000000,0x200000,
+0x5200c178,0x6800100,0x962540,0x5200c178,0x7c00100,0x230400,0x5200c178,0x7c00100,0x230401,0x5200c178,0xc000010,0x448000,0x5200c178,0x80000000,0x218960,0x5200c247,
+0x7c00100,0x230400,0x5200c247,0x7c00100,0x830400,0x5200c247,0x7c00100,0x1430400,0x5200c300,0x4000000,0x200003,0x52022d00,0x4000000,0x100006,0x52023700,0x24000000,
+0x4100000,0x52023700,0x24000000,0x4e00000,0x52023700,0x24000000,0x14100000,0x52023700,0x24000000,0x14e00000,0x52023700,0x24000000,0x96800000,0x52024400,0x4000000,0x100000,
+0x52027300,0x24000000,0x100000,0x5202c300,0x4000000,0x100000,0x5202c300,0x4000000,0x100002,0x5202c300,0x4000000,0x100003,0x5202c300,0x4000000,0x10000d,0x5202c300,
+0x4000100,0x150400,0x5202c300,0x4000100,0x15040d,0x5202c300,0x4000100,0x14150400,0x520a1e12,0x7c00100,0x2130480,0x520a3700,0x24000000,0x34e00000,0x520a3800,0x24000000,
+0x34100000,0x520a4711,0x7c40300,0xe30000,0x520a4f11,0x7c00300,0xe30001,0x520a7300,0x24000000,0x34100000,0x520ab412,0x7c00100,0x2130480,0x520ac400,0x4000000,0xe00002,
+0x520ac400,0x4000000,0xe0000d,0x520ac400,0x4000000,0x34e0000d,0x520ac414,0x4000000,0xe0000d,0x520ac511,0x7c40300,0xe30000,0x5240af91,0x7c00100,0x230400,0x5240af96,
+0x4000400,0x200000,0x5240af98,0x6800400,0x962540,0x5240af98,0x7c00100,0x230400,0x5240afa2,0x7c00100,0x230400,0x5240afa4,0x7c00100,0x230400,0x5240b2c7,0x4000000,
+0x200000,0x5240b2c7,0x4000000,0x1500000,0x5240b2d2,0x4000000,0x200000,0x5240b2e0,0x4000000,0x200000,0x5240b5f6,0x7c00900,0x230400,0x524a44bf,0x4000000,0xe00003,
+0x5280af91,0x2802400,0x962460,0x5280af92,0x2802400,0x962460,0x5280af98,0x2802400,0x962460,0x5280af9a,0x2802400,0x962460,0x5280af9c,0x2802400,0x962460,0x52c0b3ed,
+0x2802400,0x962460,0x52c0b3f1,0x7c00100,0x230400,0x60000c02,0x2802100,0x962460,0x60000c02,0x7c00100,0x230400,0x60000f0a,0x2802100,0x962460,0x60000f0a,0x6800100,
+0x962540,0x60000f0a,0x7c00100,0x230400,0x6000131f,0x4000000,0x200000,0x6000171a,0x7c00100,0x230400,0x6000171a,0x7c00100,0x230560,0x60001b27,0x2802100,0x962460,
+0x60001b27,0x4000000,0xc00000,0x60001b27,0x7c00100,0x230400,0x60001f0b,0x2802400,0x962460,0x60002919,0x7c00100,0x22040e,0x60002a00,0x4000000,0x1600000,0x60003000,
+0x24000000,0x14200000,0x60003000,0x24000000,0x14e00000,0x60003700,0x24000000,0x4200000,0x60003800,0x24000000,0x1710000,0x60005102,0x4000000,0x200000,0x60006108,0x7c00100,
+0x220400,0x60006108,0x7c00100,0x250400,0x60006600,0x24000020,0x200000,0x60008301,0x2802000,0x962460,0x6000903c,0x2806000,0x962460,0x6000903c,0x4000000,0x400000,
+0x60009519,0x7c00100,0x220400,0x60009519,0x7c00100,0x250400,0x6000a008,0x7c00100,0x220400,0x6000a008,0x7c00100,0x250400,0x6000c300,0x4000000,0x3a703580,0x6000c654,
+0x2802000,0x962460,0x6000c654,0x4000010,0x200000,0x6000c654,0x7c00100,0x230400,0x6000c73f,0x2802000,0x962460,0x6000c73f,0x2802100,0x962460,0x6000c73f,0x4000000,
+0x200000,0x6000c73f,0x6800100,0x962540,0x6000c73f,0x6804000,0x962540,0x6000c73f,0x7c00100,0x230400,0x6000c80b,0x7c00100,0x230400,0x6000c941,0x2802100,0x962460,
+0x6000c941,0x2806000,0x962460,0x6000c941,0x4000000,0x200000,0x6000c941,0x4000010,0x200000,0x6000c941,0x6800000,0x1329800,0x6000c941,0x6800100,0x962540,0x6000c941,
+0x7c00100,0x230400,0x6000c941,0xc000010,0x448000,0x6000ca82,0x7c00100,0x230400,0x6000cc00,0x4000000,0x4e00000,0x6000d000,0x4000000,0x200000,0x6002c300,0x4000000,
+0x100000,0x6002c300,0x4000000,0x10000d,0x6002c300,0x4000100,0x150400,0x6002c300,0x4000100,0x15040d,0x6002c300,0x4000100,0x14150400,0x600a3000,0x24000000,0x34200000,
+0x600a3000,0x24000000,0x34e00000,0x600a3700,0x24000000,0x34200000,0x600a3800,0x24000000,0x34200000,0x600a3800,0x24000000,0xb6800000,0x600a4305,0x7c00100,0xe30400,0x600ac300,
+0x4000000,0x34100000,0x600ac400,0x4000000,0x14e0000d,0x600ac400,0x4000000,0x34e0000d,0x600acb14,0x7c00100,0xe30000,0x600acb16,0x7c00100,0xe30c00,0x600acc00,0x4000000,
+0x34e00000,0x600acd00,0x4000000,0x34200000,0x600acd00,0x4000000,0x34e00000,0x600acd00,0x4000000,0xb6800000,0x600ace00,0x4000000,0x34e00000,0x600ace00,0x4000000,0xb6800000,
+0x600acf00,0x4000000,0x34e00000,0x600acf00,0x4000000,0xb6800000,0x600ad111,0x7c40300,0xe30000,0x604ac4bf,0x4000000,0x34e00003,0x61000a03,0x4000000,0x1600000,0x61000c02,
+0x80000000,0x218960,0x6100120f,0x4000000,0x200000,0x61001a18,0x7c00100,0x1830000,0x61001d0c,0x7c00100,0x230400,0x61001d0c,0x7c00100,0x250400,0x61006600,0x24000020,
+0x200000,0x61008407,0x7c00100,0x220400,0x61008407,0x7c00100,0x250400,0x6100870c,0x7c00100,0x220400,0x61008e00,0x24000000,0x200000,0x61008e00,0x24000000,0x400000,
+0x61008e00,0x24000002,0x300000,0x6100903c,0x7c00100,0x230400,0x61009519,0x7c00100,0x220400,0x61009519,0x7c00100,0x250400,0x61009519,0x7c00500,0x22040f,0x61009b71,
+0x2802100,0x962460,0x61009b71,0x2806400,0x962460,0x61009b71,0x7c00100,0x230400,0x6100a008,0x2802100,0x962460,0x6100c300,0x4000000,0x20000f,0x6100cd00,0x4000000,
+0x200000,0x6100d202,0x2802400,0x962460,0x6100d202,0x2802500,0x962460,0x6100d202,0x7c00100,0x230400,0x6100d302,0x4000020,0x200000,0x6100d302,0x7c00120,0x230405,
+0x6100d476,0x2802100,0x962460,0x6100d476,0x2802100,0x962461,0x6100d476,0x2806400,0x962460,0x6100d476,0x4000000,0x400000,0x6100d476,0x6800000,0x1329800,0x6100d476,
+0x6800100,0x962540,0x6100d476,0x7c00100,0x230400,0x6100d476,0xc000010,0x448000,0x6100d573,0x2802100,0x962460,0x6100d573,0x2806400,0x962460,0x6100d573,0x6800100,
+0x962540,0x6100d573,0x7c00100,0x230400,0x6100d573,0x7c00900,0x230400,0x6100d573,0xc000010,0x448000,0x6100d68d,0x7c00100,0x230400,0x6100d756,0x7c00100,0x230400,
+0x6100d85c,0x2802500,0x962460,0x6100d85c,0x6800100,0x962540,0x6100d85c,0x7c00100,0x230400,0x6100d85c,0x7c00500,0x230400,0x6100d997,0x2802100,0x962460,0x6100d997,
+0x4000000,0x200000,0x6100d997,0x4000000,0x400000,0x6100d997,0x6800000,0x1329800,0x6100d997,0x6800100,0x962540,0x6100d997,0x6804400,0x962540,0x6100d997,0x7c00100,
+0x230400,0x6100d997,0x7c00100,0x230560,0x6100d997,0xc000010,0x448000,0x6100da98,0x6800000,0x1329800,0x6100da98,0x7c00100,0x230400,0x6100db71,0x4000000,0x200000,
+0x6100dc99,0x2802100,0x962460,0x6100dc99,0x2802400,0x962460,0x6100dc99,0x6800000,0x1329800,0x6100dc99,0x6800100,0x962540,0x6100dc99,0x6804400,0x962540,0x6100dc99,
+0x7c00100,0x230400,0x610a4711,0x7c40300,0xe30000,0x610a4f11,0x7c00300,0xe30001,0x610ace00,0x4000000,0x34e00000,0x6140af96,0x7c00100,0x230400,0x6140af98,0x7c00100,
+0x230400,0x6180af93,0x2802400,0x962460,0x62002a00,0x4000000,0x1600000,0x63002800,0x80000,0x918820,0x63c00c14,0x80000,0x918820,0x7000080e,0x7c00100,0x250400,
+0x70000a03,0x4000000,0x200000,0x70000c00,0x80000000,0x218960,0x70000f0a,0x7c00100,0x230400,0x70001004,0x7c00100,0x230400,0x70001524,0x2802100,0x962460,0x70001524,
+0x7c00100,0x230400,0x70001615,0x2802100,0x962460,0x7000171a,0x2802100,0x962460,0x70001821,0x6800000,0x1329800,0x70002320,0x7c00100,0x230400,0x70002a00,0x4000000,
+0x1500000,0x70002a00,0x4000000,0x1600000,0x70003000,0x24000000,0x200000,0x70003000,0x24000000,0x14200000,0x70003800,0x24000000,0x4e00000,0x70005201,0x2802400,0x962460,
+0x7000581e,0x7c00100,0x230400,0x70006108,0x7c00100,0x220400,0x70006108,0x7c00100,0x250400,0x70006f30,0x7c00100,0x230400,0x70007300,0x24000000,0x200000,0x70007f0e,
+0x4000000,0x200000,0x70008301,0x2802100,0x962460,0x70008301,0x2802400,0x962460,0x70008e00,0x24000000,0x200000,0x70008e00,0x24000000,0x400000,0x70008e00,0x24000002,
+0x400000,0x70008e00,0x24000008,0x1410000,0x70008e00,0x24000010,0x400000,0x70008e00,0x2c000010,0x448000,0x70009519,0x7c00100,0x220400,0x70009519,0x7c00100,0x230400,
+0x70009519,0x7c00100,0x250400,0x70009865,0x7c00100,0x230400,0x70009965,0x4000010,0x400000,0x70009965,0x7c00100,0x230400,0x7000a008,0x7c00100,0x220400,0x7000a008,
+0x7c00100,0x250400,0x7000a008,0x7c00500,0x22040f,0x7000a50e,0x4000000,0x200000,0x7000b61c,0x2802500,0x1862460,0x7000b61c,0x6800500,0x1862400,0x7000b61c,0x7c00100,
+0x1830000,0x7000c300,0x4000000,0x100000,0x7000c941,0x2806000,0x962460,0x7000cc00,0x4000000,0x4e00000,0x7000cd00,0x4000000,0x200000,0x7000cd00,0x4000000,0x4200000,
+0x7000cd00,0x4000000,0x4e00000,0x7000cd00,0x4000000,0x14200000,0x7000cd00,0x4000000,0x14e00000,0x7000cd00,0x4000000,0x96800000,0x7000cf00,0x4000000,0x4e00000,0x7000cf00,
+0x4000000,0x14e00000,0x7000d202,0x2802100,0x962460,0x7000d202,0x7c00100,0x230400,0x7000d997,0x7c00100,0x230400,0x7000d997,0xc000010,0x248000,0x7000dd86,0x2802400,
+0x962460,0x7000dd86,0x7c00100,0x230400,0x7000dd86,0xc000010,0x448000,0x7000de9f,0x4000000,0x200000,0x7000de9f,0x7c00100,0x230400,0x7000e001,0x2000,0x962460,
+0x7000e001,0x2802400,0x962460,0x7000e187,0x2802000,0x962460,0x7000e187,0x2802100,0x962460,0x7000e187,0x4000000,0x200000,0x7000e187,0x7c00100,0x230400,0x7000e187,
+0xc000010,0x448000,0x7000e288,0x7c00100,0x230400,0x7000e300,0x4000000,0x200000,0x7000e489,0x2802100,0x962460,0x7000e489,0x2802400,0x962460,0x7000e489,0x6800100,
+0x962540,0x7000e489,0x6800100,0x962541,0x7000e489,0x6804400,0x962540,0x7000e489,0x7c00100,0x230400,0x7000e489,0x7c00900,0x230400,0x7000e59d,0x2802100,0x962460,
+0x7000e59d,0x2802400,0x962460,0x7000e59d,0x4000000,0x200000,0x7000e59d,0x4000010,0x200000,0x7000e59d,0x6800100,0x962540,0x7000e59d,0x6804400,0x962540,0x7000e59d,
+0x7c00100,0x230400,0x7000e59d,0xc000010,0x448000,0x7000e691,0x2802100,0x962460,0x7000e691,0x2802400,0x962460,0x7000e691,0x2806400,0x962460,0x7000e691,0x6800000,
+0x1329800,0x7000e691,0x6800100,0x962540,0x7000e691,0x7c00100,0x230400,0x7000e700,0x4000400,0x200400,0x7000e70e,0x7c00100,0x220400,0x7000e719,0x7c00100,0x220400,
+0x7000e719,0x7c00500,0x22040f,0x7000e853,0x7c00100,0x230400,0x7000e9a0,0x2802400,0x962460,0x7000e9a0,0x4000000,0x200000,0x7000e9a0,0x4000000,0x500000,0x7000e9a0,
+0x7c00100,0x230400,0x7000ea79,0x2802400,0x962460,0x7000ea79,0x4000000,0x200000,0x7000ea79,0x4000000,0xf00000,0x7000ea79,0x4000010,0x400000,0x7000ea79,0x7c00100,
+0x230400,0x7000eb8c,0x2802400,0x962460,0x7000eb8c,0x4000000,0x200000,0x7000eb8c,0x7c00100,0x230400,0x7000eca3,0x2802100,0x962460,0x7000eca3,0x2806400,0x962460,
+0x7000eca3,0x4000000,0x200000,0x7000eca3,0x6800000,0x1329800,0x7000eca3,0x6800100,0x962540,0x7000eca3,0x7c00100,0x230400,0x7000eca3,0xc000010,0x448000,0x7000ed95,
+0x6800000,0x1329800,0x7000ed95,0x7c00100,0x230400,0x7000ed95,0xc000010,0x448000,0x7000ee1c,0x2802500,0x1862460,0x7000ee1c,0x6800000,0x1329800,0x7000ee1c,0x7c00100,
+0x1830000,0x7000ee1c,0x7c00900,0x1830000,0x7000ef8f,0x4000000,0x200000,0x7000ef8f,0x7c00100,0x230400,0x7000f08e,0x4000000,0x200000,0x7000f08e,0x7c00100,0x230400,
+0x7000f159,0x2802100,0x962460,0x7000f159,0x7c00100,0x230400,0x7000f200,0x4000000,0x200000,0x7000f200,0x4000000,0x1200000,0x7000f200,0x4000000,0x1710000,0x7000f34b,
+0x2802400,0x962460,0x7000f34b,0x4000000,0x200000,0x7000f34b,0x4000010,0x400000,0x7000f34b,0x6800000,0x1329800,0x7000f34b,0x7c00100,0x230400,0x7000f34b,0x7c00900,
+0x230400,0x7000f34b,0xc000010,0x448000,0x7000f490,0x4000000,0x200000,0x7000f490,0x7c00100,0x230400,0x7000f5a5,0x7c00100,0x230400,0x7000f67b,0x4000000,0x200000,
+0x7000f67b,0x4000010,0x200000,0x7000f67b,0x7c00100,0x230400,0x7000f8a6,0x2802100,0x962460,0x7000f8a6,0x2802400,0x962460,0x7000f8a6,0x2806400,0x962460,0x7000f8a6,
+0x4000000,0x500000,0x7000f8a6,0x4000010,0xb00000,0x7000f8a6,0x4000800,0x200000,0x7000f8a6,0x6800100,0x962540,0x7000f8a6,0x6800100,0x962541,0x7000f8a6,0x7c00100,
+0x230400,0x7000f8a6,0xc000010,0x448000,0x7000f921,0x4000000,0x200000,0x7000fa00,0x4000000,0x200000,0x7000fb9e,0x2802100,0x962460,0x7000fb9e,0x2802400,0x962460,
+0x7000fb9e,0x2806400,0x962460,0x7000fb9e,0x4000000,0x200000,0x7000fb9e,0x6800000,0x1329800,0x7000fb9e,0x6800100,0x962540,0x7000fb9e,0x6800100,0x962541,0x7000fb9e,
+0x7c00100,0x230400,0x7000fc92,0x4000000,0x200000,0x7000fc92,0x6800000,0x1329800,0x7000fc92,0x7c00100,0x220400,0x7000fc92,0x7c00100,0x230400,0x7000fc92,0x7c00100,
+0x250400,0x700acd00,0x4000000,0x34e00000,0x700acd00,0x4000000,0xb6800000,0x700ace00,0x4000000,0x34e00000,0x700acf00,0x4000000,0x34e00000,0x700acf00,0x4000000,0xb6800000,
+0x7050df01,0x4000000,0x200000,0x7050f705,0x80000,0x918820,0x7080af96,0x2802400,0x962460,0x7090df01,0x2802400,0x962460,0x70d0e403,0x2802100,0x962460,0x70d0e403,
+0x2802400,0x962460,0x70d0e403,0x6800100,0x962540,0x8000120f,0x7c00100,0x230400,0x80001524,0x7c00100,0x230400,0x8000171a,0x7c00100,0x230400,0x80002006,0x7c00100,
+0x220400,0x80002006,0x7c00100,0x250400,0x80002a00,0x4000000,0x1500000,0x80002d00,0x4000000,0x200000,0x80005208,0x2802400,0x962460,0x80005c00,0x4000000,0x200000,
+0x80007300,0x24000000,0x200000,0x80009519,0x7c00100,0x220400,0x80009519,0x7c00100,0x230400,0x80009519,0x7c00100,0x250400,0x80009865,0x7c00100,0x230400,0x8000a008,
+0x2802100,0x962460,0x8000b30a,0x4000000,0x500000,0x8000b30a,0x7c00100,0x230400,0x8000cd00,0x4000000,0x4e00000,0x8000d202,0x2802500,0x962460,0x8000d202,0x7c00100,
+0x230400,0x8000d68d,0x4000000,0x200000,0x8000d997,0x2802000,0x962460,0x8000d997,0x2802400,0x962460,0x8000d997,0x4000000,0x400000,0x8000d997,0x4000000,0x500000,
+0x8000d997,0x7c00100,0x230400,0x8000d997,0xc000010,0x448000,0x8000e489,0x2802100,0x962460,0x8000e489,0x7c00100,0x230400,0x8000e719,0x7c00100,0x220400,0x8000f8a6,
+0x2802100,0x962460,0x8000f8a6,0x7c00100,0x230400,0x8000f8a6,0xc000010,0x448000,0x8000fda1,0x2802100,0x1862460,0x8000fda1,0x2806400,0x1862460,0x8000fda1,0x4000000,
+0x1800000,0x8000fda1,0x6800000,0x1329800,0x8000fda1,0x6800100,0x1862540,0x8000fda1,0x7c00100,0x1830000,0x8000fda1,0xc000010,0x448000,0x8000fe9c,0x7c00100,0x230400,
+0x8000fe9c,0x7c00100,0x830400,0x8000fe9c,0x7c00100,0x1430400,0x8000ff06,0x7c00100,0x220400,0x80010165,0x7c00100,0x230400,0x800102a2,0x4000000,0x200000,0x800102a2,
+0x7c00100,0x230400,0x800103a4,0x7c00100,0x230400,0x800103a4,0xc000010,0x448000,0x8001044c,0x4000000,0x200000,0x8001044c,0x7c00100,0x220400,0x8001044c,0x7c00100,
+0x250400,0x80010670,0x2802000,0x962460,0x80010670,0x4000000,0x200000,0x80010670,0x4000010,0x400000,0x80010670,0xc000010,0x448000,0x800a4711,0x7c40300,0xe30000,
+0x800acd00,0x4000000,0x34e00000,0x800acd00,0x4000000,0x7a902460,0x800ace00,0x4000000,0x34e00000,0x800acf00,0x4000000,0x34e00000,0x800b0011,0x7c40300,0xe30000,0x800b0500,
+0x4000000,0x34e00000,0x800b0500,0x4000000,0xb6800000,0x90001615,0x7c00100,0x230400,0x9000171a,0x4000000,0x200000,0x9000171a,0x7c00100,0x230400,0x90003000,0x24000000,
+0x200000,0x90007f0e,0x4000000,0x200000,0x90008301,0x2802000,0x962460,0x90008e00,0x24000000,0x400000,0x90009519,0x7c00100,0x250400,0x9000a16f,0x2802100,0x962460,
+0x9000d200,0x80000000,0x218960,0x9000d202,0x2802000,0x962460,0x9000d202,0x2802100,0x962460,0x9000d202,0x7c00100,0x230400,0x9000e59d,0x2802100,0x962460,0x90010500,
+0x4000000,0xe00000,0x900107a7,0x2802100,0x962460,0x900107a7,0x2802400,0x962460,0x900107a7,0x2802c00,0x962460,0x900107a7,0x4000000,0x1400000,0x900107a7,0x6800000,
+0x1329800,0x900107a7,0x7c00100,0x220400,0x900107a7,0x7c00100,0x250400,0x900108a8,0x2802100,0x962460,0x900108a8,0x2806400,0x962460,0x900108a8,0x4000000,0x200000,
+0x900108a8,0x4000000,0x400000,0x900108a8,0x4000010,0x400000,0x900108a8,0x6800000,0x1329800,0x900108a8,0x6800100,0x962540,0x900108a8,0x7c00100,0x230400,0x900108a8,
+0xc000010,0x448000,0x90010908,0x7c00100,0x220400,0x90010a38,0x2802100,0x962460,0x90010ca9,0x2802100,0x962460,0x90010ca9,0x4000000,0x500000,0x90010ca9,0x4000010,
+0xb00000,0x90010ca9,0x6800100,0x962540,0x90010ca9,0x7c00100,0x230400,0x90010d1b,0x4000000,0x500000,0x90010eaa,0x2802100,0x962460,0x90010eaa,0x2802400,0x962460,
+0x90010eaa,0x2806400,0x962460,0x90010eaa,0x4000000,0x200000,0x90010eaa,0x4000000,0x400000,0x90010eaa,0x4000010,0x400000,0x90010eaa,0x6800000,0x1329800,0x90010eaa,
+0x6800100,0x962540,0x90010eaa,0x7c00100,0x230400,0x90010eaa,0xc000010,0x448000,0x90010fab,0x7c00100,0x220400,0x90010fab,0x7c00100,0x250400,0x9002c300,0x4000000,
+0x100000,0x900ac400,0x4000000,0xe0000d,0x900acd00,0x4000000,0x34e00000,0x900acd00,0x4000000,0xb6800000,0x900acf00,0x4000000,0x34e00000,0x900b0500,0x4000000,0x34e00000,
+0x900b0500,0x4000000,0xb6800000,0x900b0b9a,0x7c00900,0x1230400,0x900b109a,0x7c00300,0xe30000,0x900b119a,0x7c00300,0xe30000,0x90408e06,0x24000000,0x400000,0xa0001004,
+0x4000000,0x200000,0xa0001004,0x7c00100,0x230400,0xa000120f,0x2802100,0x962460,0xa000120f,0x2802400,0x962460,0xa000171a,0x2802100,0x962460,0xa000171a,0x2806400,
+0x962460,0xa0002a00,0x4000000,0x1600000,0xa0003000,0x24000000,0x200000,0xa000581e,0x7c00100,0x230400,0xa0007300,0x24000000,0x200000,0xa0008301,0x2802400,0x962460,
+0xa0008e00,0x24000000,0x400000,0xa000cf00,0x4000000,0x4e00000,0xa0010500,0x4000000,0x200000,0xa00114af,0x2802100,0x962460,0xa00114af,0x2802400,0x962460,0xa00114af,
+0x2806400,0x962460,0xa00114af,0x6800000,0x1329800,0xa00114af,0x7c00100,0x230400,0xa00114af,0x7c00100,0x230560,0xa00116b0,0x2802100,0x962460,0xa00116b0,0x2802800,
+0x962460,0xa00116b0,0x2806400,0x962460,0xa00116b0,0x4000000,0x400000,0xa00116b0,0x4000000,0x500000,0xa00116b0,0x4000010,0x400000,0xa00116b0,0x6800100,0x962540,
+0xa00116b0,0x7c00100,0x230400,0xa00116b0,0x7c00100,0x230560,0xa00116b0,0xc000010,0x448000,0xa0011722,0x7c00100,0x230400,0xa00118b1,0x2802000,0x962460,0xa00118b1,
+0x2802100,0x962460,0xa00118b1,0x2806400,0x962460,0xa00118b1,0x4000000,0x200000,0xa00118b1,0x4000000,0x400000,0xa00118b1,0x4000000,0x500000,0xa00118b1,0x6800100,
+0x962540,0xa00118b1,0x7c00100,0x230400,0xa00118b1,0x7c00100,0x230560,0xa00118b1,0xc000010,0x448000,0xa00a4005,0x7c00100,0xe30400,0xa00a4711,0x7c40300,0xe30000,
+0xa00ac400,0x4000000,0x4e00000,0xa00acb14,0x7c00100,0xe30000,0xa00acf00,0x4000000,0x34e00000,0xa00b0500,0x4000000,0x34e00000,0xa00b0500,0x4000000,0xb6800000,0xa00b0b96,
+0x7c00900,0x1230400,0xa00b1211,0x7c40300,0xe30000,0xa00b1314,0x7c00100,0xe30000,0xa00b1596,0x7c00300,0xe30000,0xa040afac,0x6800400,0x962540,0xa08083ad,0x2802400,
+0x962460,0xb0000a03,0x7c00100,0x220400,0xb0000b13,0x7c00100,0x2633800,0xb0001004,0x2802000,0x962460,0xb0001110,0x4000000,0x200000,0xb0001524,0x2802000,0x962460,
+0xb0001615,0x4000000,0x500000,0xb000251b,0x7c00100,0x230400,0xb0007300,0x24000000,0x200000,0xb0008939,0x4000000,0x200000,0xb0008939,0x7c00100,0x230400,0xb0008e00,
+0x24000000,0x200000,0xb0008e00,0x24000000,0x400000,0xb0008e00,0x24000010,0x400000,0xb0009257,0x2802000,0x962460,0xb0009257,0x4000000,0x1600000,0xb0009519,0x7c00100,
+0x220400,0xb0009519,0x7c00100,0x250400,0xb0009a00,0x4000000,0x200000,0xb000b30a,0x2802100,0x962460,0xb000b30a,0x7c00100,0x230400,0xb000c178,0x80000000,0x218960,
+0xb000c300,0x4000000,0x4200000,0xb000d202,0x2802000,0x962460,0xb000d476,0x6800100,0x962540,0xb000d476,0x7c00100,0x230400,0xb000e300,0x4000000,0x4e00000,0xb000fda1,
+0x7c00100,0x1830000,0xb0010eaa,0x2802000,0x962460,0xb00116b0,0x7c00100,0x230400,0xb0011900,0x4000000,0x4e00000,0xb0011ab2,0x2802100,0x962460,0xb0011ab2,0x2802400,
+0x962460,0xb0011ab2,0x2806400,0x962460,0xb0011ab2,0x4000000,0x200000,0xb0011ab2,0x6800100,0x962540,0xb0011ab2,0x7c00100,0x230400,0xb0011b0c,0x7c00100,0x230400,
+0xb0011cb3,0x2802100,0x962460,0xb0011cb3,0x2806400,0x962460,0xb0011cb3,0x6800000,0x1329800,0xb0011cb3,0x6800100,0x962540,0xb0011cb3,0x7c00100,0x230400,0xb0011db6,
+0x2802500,0x962460,0xb0011db6,0x6800000,0x1329800,0xb0011db6,0x7c00100,0x230400,0xb0011db6,0x7c00500,0x230400,0xb0011e00,0x4000000,0x200000,0xb0011e00,0x4000000,
+0x1500000,0xb0011fb4,0x2802100,0x962460,0xb0011fb4,0x6800100,0x962540,0xb0011fb4,0x7c00100,0x230400,0xb0011fb4,0xc000010,0x248000,0xb0012000,0x4000000,0x200000,
+0xb00121b5,0x4000000,0x200000,0xb00121b5,0x4000010,0x400000,0xb00121b5,0x7c00100,0x220400,0xb00121b5,0x7c00100,0x250400,0xb00121b5,0xc000010,0x448000,0xb00122b8,
+0x4000000,0x200000,0xb00122b8,0x7c00100,0x230400,0xb00123b7,0x2802400,0x962460,0xb00123b7,0x4000000,0x200000,0xb00123b7,0x7c00100,0x230400,0xb00123b7,0xc000010,
+0x248000,0xb00a4005,0x7c00100,0xe30400,0xb00a4711,0x7c40300,0xe30000,0xb00acf00,0x4000000,0x34e00000,0xb00b0500,0x4000000,0x34e00000,0xb00b0500,0x4000000,0x3ce00000,
+0xb00b0500,0x4000000,0xb6800000,0xb00b109a,0x7c00300,0xe30000,0xb080e47c,0x2802000,0x962460,0xc0001524,0x4000000,0x500000,0xc0001a18,0x2806400,0x1862460,0xc0001a18,
+0x7c00100,0x1830000,0xc0007300,0x24000000,0x200000,0xc0008e00,0x24000010,0x400000,0xc0009519,0x7c00100,0x220400,0xc0009519,0x7c00100,0x250400,0xc000c300,0x4000000,
+0x420000f,0xc000d85c,0x2802100,0x962460,0xc000d85c,0x6800100,0x962540,0xc000d85c,0x7c00100,0x230400,0xc000dc99,0x7c00100,0x230400,0xc000e719,0x7c00100,0x220400,
+0xc00107a7,0x7c00100,0x230400,0xc0010eaa,0x7c00100,0x230400,0xc00116b0,0x7c00100,0x230560,0xc0011900,0x4000000,0x4200000,0xc0012447,0,0x818820,0xc0012447,
+0,0xc18820,0xc0012447,0,0x1418820,0xc00125b9,0x7c00100,0x230400,0xc00126bb,0x2802100,0x962460,0xc00126bb,0x2806400,0x962460,0xc00126bb,0x4000000,
+0x500000,0xc00126bb,0x6800100,0x962540,0xc00126bb,0x7c00100,0x230400,0xc00127ba,0x2802400,0x962460,0xc00127ba,0x4000000,0x200000,0xc00127ba,0x6800000,0x1329800,
+0xc00127ba,0x7c00100,0x230400,0xc00127ba,0x7c00900,0x230400,0xc0012800,0x4000000,0x200000,0xc0012b23,0x4000000,0x200000,0xc0012b23,0x4000000,0x400000,0xc0012b23,
+0x4000000,0x1500000,0xc0012cbc,0x2802400,0x962460,0xc0012cbc,0x4000000,0x1600000,0xc0012cbc,0x6800000,0x1329800,0xc0012cbc,0x7c00100,0x230400,0xc00acf00,0x4000000,
+0x34e00000,0xc00ae300,0x4000000,0x34e00000,0xc00b0500,0x4000000,0x34e00000,0xc00b0500,0x4000000,0xb6800000,0xc00b0b00,0x4000000,0x1200000,0xc00b0b00,0x7c00900,0x1230400,
+0xc00b109a,0x7c00300,0xe30000,0xc00b2914,0x7c00100,0x2530000,0xc00b2916,0x7c00100,0x2530c00,0xc00b2a00,0x4000000,0x34e00000,0xc040af53,0x7c00100,0x230400,0xc0c12b7e,
+0x4000000,0x200000,0xc14a44bf,0x4000000,0xe0000d,0xd000131f,0x2802c00,0x962460,0xd000171a,0x7c00100,0x230400,0xd0001821,0x2802100,0x962460,0xd0007300,0x24000000,
+0x200000,0xd0008e00,0x24000000,0x200000,0xd0008f3a,0x2806000,0x962460,0xd0009519,0x7c00100,0x220400,0xd0009519,0x7c00100,0x250400,0xd000a500,0x4000000,0x200000,
+0xd000c300,0x4000000,0x4e00000,0xd000d202,0x7c00100,0x230400,0xd000d476,0x7c00100,0x230400,0xd000d997,0x2802100,0x962460,0xd000d997,0x6800100,0x962540,0xd000e001,
+0x2802100,0x962460,0xd000e700,0x4000400,0x200000,0xd000e719,0x7c00100,0x220400,0xd000e719,0x7c00500,0x23040f,0xd000fa00,0x4000000,0x4e00000,0xd0010eaa,0x4000010,
+0x400000,0xd0010eaa,0x7c00100,0x230400,0xd0012dbd,0x4000000,0x200000,0xd0012dbd,0x7c00100,0x230400,0xd0012fbe,0x2802100,0x962460,0xd0012fbe,0x2802400,0x962460,
+0xd0012fbe,0x2806400,0x962460,0xd0012fbe,0x4000000,0x400000,0xd0012fbe,0x6800000,0x1329800,0xd0012fbe,0x6800100,0x962540,0xd0012fbe,0x6800100,0x962541,0xd0012fbe,
+0x6804400,0x962540,0xd0012fbe,0x7c00100,0x230400,0xd0012fbe,0x7c00100,0x230560,0xd0012fbe,0xc000010,0x448000,0xd0013183,0x7c00100,0x230400,0xd0013200,0x4000000,
+0x200000,0xd0013200,0x6800000,0x1329805,0xd00134c0,0x2802100,0x962460,0xd00134c0,0x4000002,0x400000,0xd00134c0,0x7c00100,0x230400,0xd00a4305,0x7c00100,0xe30400,
+0xd00a4611,0x7c40300,0xe30000,0xd00a4711,0x7c40300,0xe30000,0xd00a5e11,0x7c40300,0xe30000,0xd00acf00,0x4000000,0x34e00000,0xd00b0500,0x4000000,0x34e00000,0xd00b0500,
+0x4000000,0xb6800000,0xd00b0b11,0x6800500,0x962540,0xd00b0bbf,0x2802200,0xc62460,0xd00b119a,0x7c00300,0xe30000,0xd00b2a00,0x4000000,0x34e00000,0xd00b2e11,0x7c40300,
+0xe30000,0xd00b30bf,0x7c00300,0x230000,0xd00b339a,0x7c00300,0xe30000};
+
+static const int32_t countPropsVectors=6999;
+static const int32_t propsVectorsColumns=3;
+static const uint16_t scriptExtensions[262]={
+0x800e,0x8019,8,0x8059,8,2,8,0x8038,8,6,8,0x8019,2,0x22,0x25,0xb6,
+0x80c0,2,0x22,0x8025,2,0x11,2,0x22,0x54,0x79,0x7b,0xa7,0xb6,0x80b7,2,0x8022,
+2,0x25,0x80c0,2,0x20,2,0x80b6,4,0xa,0xf,0x10,0x15,0x19,0x1a,0x1f,0x23,
+0x24,0x89,0x97,0x809e,4,0xa,0xf,0x10,0x15,0x19,0x1a,0x1f,0x23,0x24,0x89,0x809e,
+4,0xa,0xf,0x10,0x15,0x1a,0x1f,0x21,0x23,0x24,0x3a,0x89,0x91,0x99,0x9e,0xa0,
+0xaf,0xb2,0xb3,0x80bb,4,0xa,0xf,0x10,0x15,0x1a,0x1f,0x21,0x23,0x24,0x30,0x3a,
+0x89,0x91,0x99,0x9e,0xa0,0xaf,0xb2,0xb3,0x80bb,0xa,0x78,0xa0,0x80b2,0xa,0x69,4,
+0x3a,0x8076,4,0x6f,0x10,0x80a4,0x10,0x74,0xf,0x809d,0xf,0x78,0x23,0x8089,0x23,0x7c,
+0x15,0x80bb,0x15,0x80,0x1c,0x34,0x8076,0x1c,0x84,0xc,0x8019,0x2a,0x2b,0x2c,0x802d,0x1b,
+0x805a,0x800a,4,0xa,0x15,0x8089,0xa,0x8089,4,0x800a,0xa,0x8097,0xa,0x15,0x1a,0x1f,
+0x23,0x8024,0xa,0x80bb,4,0xa,0x15,0x1f,0x24,0x89,0x9e,0x80bb,0x8004,8,0x8022,0x19,
+0x801b,0xa,0x19,0x8089,5,0x11,0x12,0x14,0x16,0x8029,5,0x11,0x12,0x14,0x8016,0x8011,
+5,0x8011,0x11,0x14,0x8016,0x11,0x8019,0xa,0xf,0x10,0x78,0x91,0x99,0x9d,0x9e,0xa0,
+0xa3,0x80b2,0xa,0xf,0x10,0x15,0x1a,0x78,0x91,0x99,0x9d,0x9e,0xa0,0xa3,0xb2,0x80bb,
+0xa,0xf,0x10,0x15,0x78,0x91,0x99,0x9d,0x9e,0xa0,0xa3,0xb2,0x80bb,0xa,0x98,0xa,
+0x8023,0xa,0xef,0x19,0x1c,0x804f,0x37,0x804e,2,0x8025,2,0xf8,0x2f,0x31,0x8053,0x2f,
+0x8031,2,0x8007,0x89,0x7c,0x8087};
+
+static const int32_t indexes[UPROPS_INDEX_COUNT]={0x2b96,0x2b96,0x2b96,0x2b96,0x6898,3,0x83ef,0x8472,0x8472,0x8472,0xb34c0,0x2a75a31,0,0,0,0};
+
+#endif  // INCLUDED_FROM_UCHAR_C
diff --git a/thirdparty/icu4c/common/ucharstrie.cpp b/thirdparty/icu4c/common/ucharstrie.cpp
new file mode 100644
index 0000000000..e0b33af519
--- /dev/null
+++ b/thirdparty/icu4c/common/ucharstrie.cpp
@@ -0,0 +1,414 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucharstrie.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov14
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/appendable.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/uobject.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+UCharsTrie::~UCharsTrie() {
+    uprv_free(ownedArray_);
+}
+
+UStringTrieResult
+UCharsTrie::current() const {
+    const UChar *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    } else {
+        int32_t node;
+        return (remainingMatchLength_<0 && (node=*pos)>=kMinValueLead) ?
+                valueResult(node) : USTRINGTRIE_NO_VALUE;
+    }
+}
+
+UStringTrieResult
+UCharsTrie::firstForCodePoint(UChar32 cp) {
+    return cp<=0xffff ?
+        first(cp) :
+        (USTRINGTRIE_HAS_NEXT(first(U16_LEAD(cp))) ?
+            next(U16_TRAIL(cp)) :
+            USTRINGTRIE_NO_MATCH);
+}
+
+UStringTrieResult
+UCharsTrie::nextForCodePoint(UChar32 cp) {
+    return cp<=0xffff ?
+        next(cp) :
+        (USTRINGTRIE_HAS_NEXT(next(U16_LEAD(cp))) ?
+            next(U16_TRAIL(cp)) :
+            USTRINGTRIE_NO_MATCH);
+}
+
+UStringTrieResult
+UCharsTrie::branchNext(const UChar *pos, int32_t length, int32_t uchar) {
+    // Branch according to the current unit.
+    if(length==0) {
+        length=*pos++;
+    }
+    ++length;
+    // The length of the branch is the number of units to select from.
+    // The data structure encodes a binary search.
+    while(length>kMaxBranchLinearSubNodeLength) {
+        if(uchar<*pos++) {
+            length>>=1;
+            pos=jumpByDelta(pos);
+        } else {
+            length=length-(length>>1);
+            pos=skipDelta(pos);
+        }
+    }
+    // Drop down to linear search for the last few units.
+    // length>=2 because the loop body above sees length>kMaxBranchLinearSubNodeLength>=3
+    // and divides length by 2.
+    do {
+        if(uchar==*pos++) {
+            UStringTrieResult result;
+            int32_t node=*pos;
+            if(node&kValueIsFinal) {
+                // Leave the final value for getValue() to read.
+                result=USTRINGTRIE_FINAL_VALUE;
+            } else {
+                // Use the non-final value as the jump delta.
+                ++pos;
+                // int32_t delta=readValue(pos, node);
+                int32_t delta;
+                if(node<kMinTwoUnitValueLead) {
+                    delta=node;
+                } else if(node<kThreeUnitValueLead) {
+                    delta=((node-kMinTwoUnitValueLead)<<16)|*pos++;
+                } else {
+                    delta=(pos[0]<<16)|pos[1];
+                    pos+=2;
+                }
+                // end readValue()
+                pos+=delta;
+                node=*pos;
+                result= node>=kMinValueLead ? valueResult(node) : USTRINGTRIE_NO_VALUE;
+            }
+            pos_=pos;
+            return result;
+        }
+        --length;
+        pos=skipValue(pos);
+    } while(length>1);
+    if(uchar==*pos++) {
+        pos_=pos;
+        int32_t node=*pos;
+        return node>=kMinValueLead ? valueResult(node) : USTRINGTRIE_NO_VALUE;
+    } else {
+        stop();
+        return USTRINGTRIE_NO_MATCH;
+    }
+}
+
+UStringTrieResult
+UCharsTrie::nextImpl(const UChar *pos, int32_t uchar) {
+    int32_t node=*pos++;
+    for(;;) {
+        if(node<kMinLinearMatch) {
+            return branchNext(pos, node, uchar);
+        } else if(node<kMinValueLead) {
+            // Match the first of length+1 units.
+            int32_t length=node-kMinLinearMatch;  // Actual match length minus 1.
+            if(uchar==*pos++) {
+                remainingMatchLength_=--length;
+                pos_=pos;
+                return (length<0 && (node=*pos)>=kMinValueLead) ?
+                        valueResult(node) : USTRINGTRIE_NO_VALUE;
+            } else {
+                // No match.
+                break;
+            }
+        } else if(node&kValueIsFinal) {
+            // No further matching units.
+            break;
+        } else {
+            // Skip intermediate value.
+            pos=skipNodeValue(pos, node);
+            node&=kNodeTypeMask;
+        }
+    }
+    stop();
+    return USTRINGTRIE_NO_MATCH;
+}
+
+UStringTrieResult
+UCharsTrie::next(int32_t uchar) {
+    const UChar *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    if(length>=0) {
+        // Remaining part of a linear-match node.
+        if(uchar==*pos++) {
+            remainingMatchLength_=--length;
+            pos_=pos;
+            int32_t node;
+            return (length<0 && (node=*pos)>=kMinValueLead) ?
+                    valueResult(node) : USTRINGTRIE_NO_VALUE;
+        } else {
+            stop();
+            return USTRINGTRIE_NO_MATCH;
+        }
+    }
+    return nextImpl(pos, uchar);
+}
+
+UStringTrieResult
+UCharsTrie::next(ConstChar16Ptr ptr, int32_t sLength) {
+    const UChar *s=ptr;
+    if(sLength<0 ? *s==0 : sLength==0) {
+        // Empty input.
+        return current();
+    }
+    const UChar *pos=pos_;
+    if(pos==NULL) {
+        return USTRINGTRIE_NO_MATCH;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    for(;;) {
+        // Fetch the next input unit, if there is one.
+        // Continue a linear-match node without rechecking sLength<0.
+        int32_t uchar;
+        if(sLength<0) {
+            for(;;) {
+                if((uchar=*s++)==0) {
+                    remainingMatchLength_=length;
+                    pos_=pos;
+                    int32_t node;
+                    return (length<0 && (node=*pos)>=kMinValueLead) ?
+                            valueResult(node) : USTRINGTRIE_NO_VALUE;
+                }
+                if(length<0) {
+                    remainingMatchLength_=length;
+                    break;
+                }
+                if(uchar!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+            }
+        } else {
+            for(;;) {
+                if(sLength==0) {
+                    remainingMatchLength_=length;
+                    pos_=pos;
+                    int32_t node;
+                    return (length<0 && (node=*pos)>=kMinValueLead) ?
+                            valueResult(node) : USTRINGTRIE_NO_VALUE;
+                }
+                uchar=*s++;
+                --sLength;
+                if(length<0) {
+                    remainingMatchLength_=length;
+                    break;
+                }
+                if(uchar!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+            }
+        }
+        int32_t node=*pos++;
+        for(;;) {
+            if(node<kMinLinearMatch) {
+                UStringTrieResult result=branchNext(pos, node, uchar);
+                if(result==USTRINGTRIE_NO_MATCH) {
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                // Fetch the next input unit, if there is one.
+                if(sLength<0) {
+                    if((uchar=*s++)==0) {
+                        return result;
+                    }
+                } else {
+                    if(sLength==0) {
+                        return result;
+                    }
+                    uchar=*s++;
+                    --sLength;
+                }
+                if(result==USTRINGTRIE_FINAL_VALUE) {
+                    // No further matching units.
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                pos=pos_;  // branchNext() advanced pos and wrote it to pos_ .
+                node=*pos++;
+            } else if(node<kMinValueLead) {
+                // Match length+1 units.
+                length=node-kMinLinearMatch;  // Actual match length minus 1.
+                if(uchar!=*pos) {
+                    stop();
+                    return USTRINGTRIE_NO_MATCH;
+                }
+                ++pos;
+                --length;
+                break;
+            } else if(node&kValueIsFinal) {
+                // No further matching units.
+                stop();
+                return USTRINGTRIE_NO_MATCH;
+            } else {
+                // Skip intermediate value.
+                pos=skipNodeValue(pos, node);
+                node&=kNodeTypeMask;
+            }
+        }
+    }
+}
+
+const UChar *
+UCharsTrie::findUniqueValueFromBranch(const UChar *pos, int32_t length,
+                                      UBool haveUniqueValue, int32_t &uniqueValue) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison unit
+        if(NULL==findUniqueValueFromBranch(jumpByDelta(pos), length>>1, haveUniqueValue, uniqueValue)) {
+            return NULL;
+        }
+        length=length-(length>>1);
+        pos=skipDelta(pos);
+    }
+    do {
+        ++pos;  // ignore a comparison unit
+        // handle its value
+        int32_t node=*pos++;
+        UBool isFinal=(UBool)(node>>15);
+        node&=0x7fff;
+        int32_t value=readValue(pos, node);
+        pos=skipValue(pos, node);
+        if(isFinal) {
+            if(haveUniqueValue) {
+                if(value!=uniqueValue) {
+                    return NULL;
+                }
+            } else {
+                uniqueValue=value;
+                haveUniqueValue=TRUE;
+            }
+        } else {
+            if(!findUniqueValue(pos+value, haveUniqueValue, uniqueValue)) {
+                return NULL;
+            }
+            haveUniqueValue=TRUE;
+        }
+    } while(--length>1);
+    return pos+1;  // ignore the last comparison unit
+}
+
+UBool
+UCharsTrie::findUniqueValue(const UChar *pos, UBool haveUniqueValue, int32_t &uniqueValue) {
+    int32_t node=*pos++;
+    for(;;) {
+        if(node<kMinLinearMatch) {
+            if(node==0) {
+                node=*pos++;
+            }
+            pos=findUniqueValueFromBranch(pos, node+1, haveUniqueValue, uniqueValue);
+            if(pos==NULL) {
+                return FALSE;
+            }
+            haveUniqueValue=TRUE;
+            node=*pos++;
+        } else if(node<kMinValueLead) {
+            // linear-match node
+            pos+=node-kMinLinearMatch+1;  // Ignore the match units.
+            node=*pos++;
+        } else {
+            UBool isFinal=(UBool)(node>>15);
+            int32_t value;
+            if(isFinal) {
+                value=readValue(pos, node&0x7fff);
+            } else {
+                value=readNodeValue(pos, node);
+            }
+            if(haveUniqueValue) {
+                if(value!=uniqueValue) {
+                    return FALSE;
+                }
+            } else {
+                uniqueValue=value;
+                haveUniqueValue=TRUE;
+            }
+            if(isFinal) {
+                return TRUE;
+            }
+            pos=skipNodeValue(pos, node);
+            node&=kNodeTypeMask;
+        }
+    }
+}
+
+int32_t
+UCharsTrie::getNextUChars(Appendable &out) const {
+    const UChar *pos=pos_;
+    if(pos==NULL) {
+        return 0;
+    }
+    if(remainingMatchLength_>=0) {
+        out.appendCodeUnit(*pos);  // Next unit of a pending linear-match node.
+        return 1;
+    }
+    int32_t node=*pos++;
+    if(node>=kMinValueLead) {
+        if(node&kValueIsFinal) {
+            return 0;
+        } else {
+            pos=skipNodeValue(pos, node);
+            node&=kNodeTypeMask;
+        }
+    }
+    if(node<kMinLinearMatch) {
+        if(node==0) {
+            node=*pos++;
+        }
+        out.reserveAppendCapacity(++node);
+        getNextBranchUChars(pos, node, out);
+        return node;
+    } else {
+        // First unit of the linear-match node.
+        out.appendCodeUnit(*pos);
+        return 1;
+    }
+}
+
+void
+UCharsTrie::getNextBranchUChars(const UChar *pos, int32_t length, Appendable &out) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison unit
+        getNextBranchUChars(jumpByDelta(pos), length>>1, out);
+        length=length-(length>>1);
+        pos=skipDelta(pos);
+    }
+    do {
+        out.appendCodeUnit(*pos++);
+        pos=skipValue(pos);
+    } while(--length>1);
+    out.appendCodeUnit(*pos);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ucharstriebuilder.cpp b/thirdparty/icu4c/common/ucharstriebuilder.cpp
new file mode 100644
index 0000000000..049997a275
--- /dev/null
+++ b/thirdparty/icu4c/common/ucharstriebuilder.cpp
@@ -0,0 +1,443 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucharstriebuilder.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov14
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/ucharstriebuilder.h"
+#include "unicode/unistr.h"
+#include "unicode/ustring.h"
+#include "cmemory.h"
+#include "uarrsort.h"
+#include "uassert.h"
+#include "uhash.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Note: This builder implementation stores (string, value) pairs with full copies
+ * of the 16-bit-unit sequences, until the UCharsTrie is built.
+ * It might(!) take less memory if we collected the data in a temporary, dynamic trie.
+ */
+
+class UCharsTrieElement : public UMemory {
+public:
+    // Use compiler's default constructor, initializes nothing.
+
+    void setTo(const UnicodeString &s, int32_t val, UnicodeString &strings, UErrorCode &errorCode);
+
+    UnicodeString getString(const UnicodeString &strings) const {
+        int32_t length=strings[stringOffset];
+        return strings.tempSubString(stringOffset+1, length);
+    }
+    int32_t getStringLength(const UnicodeString &strings) const {
+        return strings[stringOffset];
+    }
+
+    UChar charAt(int32_t index, const UnicodeString &strings) const {
+        return strings[stringOffset+1+index];
+    }
+
+    int32_t getValue() const { return value; }
+
+    int32_t compareStringTo(const UCharsTrieElement &o, const UnicodeString &strings) const;
+
+private:
+    // The first strings unit contains the string length.
+    // (Compared with a stringLength field here, this saves 2 bytes per string.)
+    int32_t stringOffset;
+    int32_t value;
+};
+
+void
+UCharsTrieElement::setTo(const UnicodeString &s, int32_t val,
+                         UnicodeString &strings, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    int32_t length=s.length();
+    if(length>0xffff) {
+        // Too long: We store the length in 1 unit.
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+    stringOffset=strings.length();
+    strings.append((UChar)length);
+    value=val;
+    strings.append(s);
+}
+
+int32_t
+UCharsTrieElement::compareStringTo(const UCharsTrieElement &other, const UnicodeString &strings) const {
+    return getString(strings).compare(other.getString(strings));
+}
+
+UCharsTrieBuilder::UCharsTrieBuilder(UErrorCode & /*errorCode*/)
+        : elements(NULL), elementsCapacity(0), elementsLength(0),
+          uchars(NULL), ucharsCapacity(0), ucharsLength(0) {}
+
+UCharsTrieBuilder::~UCharsTrieBuilder() {
+    delete[] elements;
+    uprv_free(uchars);
+}
+
+UCharsTrieBuilder &
+UCharsTrieBuilder::add(const UnicodeString &s, int32_t value, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return *this;
+    }
+    if(ucharsLength>0) {
+        // Cannot add elements after building.
+        errorCode=U_NO_WRITE_PERMISSION;
+        return *this;
+    }
+    if(elementsLength==elementsCapacity) {
+        int32_t newCapacity;
+        if(elementsCapacity==0) {
+            newCapacity=1024;
+        } else {
+            newCapacity=4*elementsCapacity;
+        }
+        UCharsTrieElement *newElements=new UCharsTrieElement[newCapacity];
+        if(newElements==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return *this;
+        }
+        if(elementsLength>0) {
+            uprv_memcpy(newElements, elements, (size_t)elementsLength*sizeof(UCharsTrieElement));
+        }
+        delete[] elements;
+        elements=newElements;
+        elementsCapacity=newCapacity;
+    }
+    elements[elementsLength++].setTo(s, value, strings, errorCode);
+    if(U_SUCCESS(errorCode) && strings.isBogus()) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+    return *this;
+}
+
+U_CDECL_BEGIN
+
+static int32_t U_CALLCONV
+compareElementStrings(const void *context, const void *left, const void *right) {
+    const UnicodeString *strings=static_cast<const UnicodeString *>(context);
+    const UCharsTrieElement *leftElement=static_cast<const UCharsTrieElement *>(left);
+    const UCharsTrieElement *rightElement=static_cast<const UCharsTrieElement *>(right);
+    return leftElement->compareStringTo(*rightElement, *strings);
+}
+
+U_CDECL_END
+
+UCharsTrie *
+UCharsTrieBuilder::build(UStringTrieBuildOption buildOption, UErrorCode &errorCode) {
+    buildUChars(buildOption, errorCode);
+    UCharsTrie *newTrie=NULL;
+    if(U_SUCCESS(errorCode)) {
+        newTrie=new UCharsTrie(uchars, uchars+(ucharsCapacity-ucharsLength));
+        if(newTrie==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            uchars=NULL;  // The new trie now owns the array.
+            ucharsCapacity=0;
+        }
+    }
+    return newTrie;
+}
+
+UnicodeString &
+UCharsTrieBuilder::buildUnicodeString(UStringTrieBuildOption buildOption, UnicodeString &result,
+                                      UErrorCode &errorCode) {
+    buildUChars(buildOption, errorCode);
+    if(U_SUCCESS(errorCode)) {
+        result.setTo(FALSE, uchars+(ucharsCapacity-ucharsLength), ucharsLength);
+    }
+    return result;
+}
+
+void
+UCharsTrieBuilder::buildUChars(UStringTrieBuildOption buildOption, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    if(uchars!=NULL && ucharsLength>0) {
+        // Already built.
+        return;
+    }
+    if(ucharsLength==0) {
+        if(elementsLength==0) {
+            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return;
+        }
+        if(strings.isBogus()) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        uprv_sortArray(elements, elementsLength, (int32_t)sizeof(UCharsTrieElement),
+                      compareElementStrings, &strings,
+                      FALSE,  // need not be a stable sort
+                      &errorCode);
+        if(U_FAILURE(errorCode)) {
+            return;
+        }
+        // Duplicate strings are not allowed.
+        UnicodeString prev=elements[0].getString(strings);
+        for(int32_t i=1; i<elementsLength; ++i) {
+            UnicodeString current=elements[i].getString(strings);
+            if(prev==current) {
+                errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+                return;
+            }
+            prev.fastCopyFrom(current);
+        }
+    }
+    // Create and UChar-serialize the trie for the elements.
+    ucharsLength=0;
+    int32_t capacity=strings.length();
+    if(capacity<1024) {
+        capacity=1024;
+    }
+    if(ucharsCapacity<capacity) {
+        uprv_free(uchars);
+        uchars=static_cast<UChar *>(uprv_malloc(capacity*2));
+        if(uchars==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            ucharsCapacity=0;
+            return;
+        }
+        ucharsCapacity=capacity;
+    }
+    StringTrieBuilder::build(buildOption, elementsLength, errorCode);
+    if(uchars==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+int32_t
+UCharsTrieBuilder::getElementStringLength(int32_t i) const {
+    return elements[i].getStringLength(strings);
+}
+
+UChar
+UCharsTrieBuilder::getElementUnit(int32_t i, int32_t unitIndex) const {
+    return elements[i].charAt(unitIndex, strings);
+}
+
+int32_t
+UCharsTrieBuilder::getElementValue(int32_t i) const {
+    return elements[i].getValue();
+}
+
+int32_t
+UCharsTrieBuilder::getLimitOfLinearMatch(int32_t first, int32_t last, int32_t unitIndex) const {
+    const UCharsTrieElement &firstElement=elements[first];
+    const UCharsTrieElement &lastElement=elements[last];
+    int32_t minStringLength=firstElement.getStringLength(strings);
+    while(++unitIndex<minStringLength &&
+            firstElement.charAt(unitIndex, strings)==
+            lastElement.charAt(unitIndex, strings)) {}
+    return unitIndex;
+}
+
+int32_t
+UCharsTrieBuilder::countElementUnits(int32_t start, int32_t limit, int32_t unitIndex) const {
+    int32_t length=0;  // Number of different units at unitIndex.
+    int32_t i=start;
+    do {
+        UChar unit=elements[i++].charAt(unitIndex, strings);
+        while(i<limit && unit==elements[i].charAt(unitIndex, strings)) {
+            ++i;
+        }
+        ++length;
+    } while(i<limit);
+    return length;
+}
+
+int32_t
+UCharsTrieBuilder::skipElementsBySomeUnits(int32_t i, int32_t unitIndex, int32_t count) const {
+    do {
+        UChar unit=elements[i++].charAt(unitIndex, strings);
+        while(unit==elements[i].charAt(unitIndex, strings)) {
+            ++i;
+        }
+    } while(--count>0);
+    return i;
+}
+
+int32_t
+UCharsTrieBuilder::indexOfElementWithNextUnit(int32_t i, int32_t unitIndex, UChar unit) const {
+    while(unit==elements[i].charAt(unitIndex, strings)) {
+        ++i;
+    }
+    return i;
+}
+
+UCharsTrieBuilder::UCTLinearMatchNode::UCTLinearMatchNode(const UChar *units, int32_t len, Node *nextNode)
+        : LinearMatchNode(len, nextNode), s(units) {
+    hash=hash*37u+ustr_hashUCharsN(units, len);
+}
+
+UBool
+UCharsTrieBuilder::UCTLinearMatchNode::operator==(const Node &other) const {
+    if(this==&other) {
+        return TRUE;
+    }
+    if(!LinearMatchNode::operator==(other)) {
+        return FALSE;
+    }
+    const UCTLinearMatchNode &o=(const UCTLinearMatchNode &)other;
+    return 0==u_memcmp(s, o.s, length);
+}
+
+void
+UCharsTrieBuilder::UCTLinearMatchNode::write(StringTrieBuilder &builder) {
+    UCharsTrieBuilder &b=(UCharsTrieBuilder &)builder;
+    next->write(builder);
+    b.write(s, length);
+    offset=b.writeValueAndType(hasValue, value, b.getMinLinearMatch()+length-1);
+}
+
+StringTrieBuilder::Node *
+UCharsTrieBuilder::createLinearMatchNode(int32_t i, int32_t unitIndex, int32_t length,
+                                         Node *nextNode) const {
+    return new UCTLinearMatchNode(
+            elements[i].getString(strings).getBuffer()+unitIndex,
+            length,
+            nextNode);
+}
+
+UBool
+UCharsTrieBuilder::ensureCapacity(int32_t length) {
+    if(uchars==NULL) {
+        return FALSE;  // previous memory allocation had failed
+    }
+    if(length>ucharsCapacity) {
+        int32_t newCapacity=ucharsCapacity;
+        do {
+            newCapacity*=2;
+        } while(newCapacity<=length);
+        UChar *newUChars=static_cast<UChar *>(uprv_malloc(newCapacity*2));
+        if(newUChars==NULL) {
+            // unable to allocate memory
+            uprv_free(uchars);
+            uchars=NULL;
+            ucharsCapacity=0;
+            return FALSE;
+        }
+        u_memcpy(newUChars+(newCapacity-ucharsLength),
+                 uchars+(ucharsCapacity-ucharsLength), ucharsLength);
+        uprv_free(uchars);
+        uchars=newUChars;
+        ucharsCapacity=newCapacity;
+    }
+    return TRUE;
+}
+
+int32_t
+UCharsTrieBuilder::write(int32_t unit) {
+    int32_t newLength=ucharsLength+1;
+    if(ensureCapacity(newLength)) {
+        ucharsLength=newLength;
+        uchars[ucharsCapacity-ucharsLength]=(UChar)unit;
+    }
+    return ucharsLength;
+}
+
+int32_t
+UCharsTrieBuilder::write(const UChar *s, int32_t length) {
+    int32_t newLength=ucharsLength+length;
+    if(ensureCapacity(newLength)) {
+        ucharsLength=newLength;
+        u_memcpy(uchars+(ucharsCapacity-ucharsLength), s, length);
+    }
+    return ucharsLength;
+}
+
+int32_t
+UCharsTrieBuilder::writeElementUnits(int32_t i, int32_t unitIndex, int32_t length) {
+    return write(elements[i].getString(strings).getBuffer()+unitIndex, length);
+}
+
+int32_t
+UCharsTrieBuilder::writeValueAndFinal(int32_t i, UBool isFinal) {
+    if(0<=i && i<=UCharsTrie::kMaxOneUnitValue) {
+        return write(i|(isFinal<<15));
+    }
+    UChar intUnits[3];
+    int32_t length;
+    if(i<0 || i>UCharsTrie::kMaxTwoUnitValue) {
+        intUnits[0]=(UChar)(UCharsTrie::kThreeUnitValueLead);
+        intUnits[1]=(UChar)((uint32_t)i>>16);
+        intUnits[2]=(UChar)i;
+        length=3;
+    // } else if(i<=UCharsTrie::kMaxOneUnitValue) {
+    //     intUnits[0]=(UChar)(i);
+    //     length=1;
+    } else {
+        intUnits[0]=(UChar)(UCharsTrie::kMinTwoUnitValueLead+(i>>16));
+        intUnits[1]=(UChar)i;
+        length=2;
+    }
+    intUnits[0]=(UChar)(intUnits[0]|(isFinal<<15));
+    return write(intUnits, length);
+}
+
+int32_t
+UCharsTrieBuilder::writeValueAndType(UBool hasValue, int32_t value, int32_t node) {
+    if(!hasValue) {
+        return write(node);
+    }
+    UChar intUnits[3];
+    int32_t length;
+    if(value<0 || value>UCharsTrie::kMaxTwoUnitNodeValue) {
+        intUnits[0]=(UChar)(UCharsTrie::kThreeUnitNodeValueLead);
+        intUnits[1]=(UChar)((uint32_t)value>>16);
+        intUnits[2]=(UChar)value;
+        length=3;
+    } else if(value<=UCharsTrie::kMaxOneUnitNodeValue) {
+        intUnits[0]=(UChar)((value+1)<<6);
+        length=1;
+    } else {
+        intUnits[0]=(UChar)(UCharsTrie::kMinTwoUnitNodeValueLead+((value>>10)&0x7fc0));
+        intUnits[1]=(UChar)value;
+        length=2;
+    }
+    intUnits[0]|=(UChar)node;
+    return write(intUnits, length);
+}
+
+int32_t
+UCharsTrieBuilder::writeDeltaTo(int32_t jumpTarget) {
+    int32_t i=ucharsLength-jumpTarget;
+    U_ASSERT(i>=0);
+    if(i<=UCharsTrie::kMaxOneUnitDelta) {
+        return write(i);
+    }
+    UChar intUnits[3];
+    int32_t length;
+    if(i<=UCharsTrie::kMaxTwoUnitDelta) {
+        intUnits[0]=(UChar)(UCharsTrie::kMinTwoUnitDeltaLead+(i>>16));
+        length=1;
+    } else {
+        intUnits[0]=(UChar)(UCharsTrie::kThreeUnitDeltaLead);
+        intUnits[1]=(UChar)(i>>16);
+        length=2;
+    }
+    intUnits[length++]=(UChar)i;
+    return write(intUnits, length);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ucharstrieiterator.cpp b/thirdparty/icu4c/common/ucharstrieiterator.cpp
new file mode 100644
index 0000000000..b3132241fe
--- /dev/null
+++ b/thirdparty/icu4c/common/ucharstrieiterator.cpp
@@ -0,0 +1,215 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucharstrieiterator.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov15
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/unistr.h"
+#include "uvectr32.h"
+
+U_NAMESPACE_BEGIN
+
+UCharsTrie::Iterator::Iterator(ConstChar16Ptr trieUChars, int32_t maxStringLength,
+                               UErrorCode &errorCode)
+        : uchars_(trieUChars),
+          pos_(uchars_), initialPos_(uchars_),
+          remainingMatchLength_(-1), initialRemainingMatchLength_(-1),
+          skipValue_(FALSE),
+          maxLength_(maxStringLength), value_(0), stack_(NULL) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    // stack_ is a pointer so that it's easy to turn ucharstrie.h into
+    // a public API header for which we would want it to depend only on
+    // other public headers.
+    // Unlike UCharsTrie itself, its Iterator performs memory allocations anyway
+    // via the UnicodeString and UVector32 implementations, so this additional
+    // cost is minimal.
+    stack_=new UVector32(errorCode);
+    if(stack_==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+UCharsTrie::Iterator::Iterator(const UCharsTrie &trie, int32_t maxStringLength,
+                               UErrorCode &errorCode)
+        : uchars_(trie.uchars_), pos_(trie.pos_), initialPos_(trie.pos_),
+          remainingMatchLength_(trie.remainingMatchLength_),
+          initialRemainingMatchLength_(trie.remainingMatchLength_),
+          skipValue_(FALSE),
+          maxLength_(maxStringLength), value_(0), stack_(NULL) {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    stack_=new UVector32(errorCode);
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    if(stack_==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    int32_t length=remainingMatchLength_;  // Actual remaining match length minus 1.
+    if(length>=0) {
+        // Pending linear-match node, append remaining UChars to str_.
+        ++length;
+        if(maxLength_>0 && length>maxLength_) {
+            length=maxLength_;  // This will leave remainingMatchLength>=0 as a signal.
+        }
+        str_.append(pos_, length);
+        pos_+=length;
+        remainingMatchLength_-=length;
+    }
+}
+
+UCharsTrie::Iterator::~Iterator() {
+    delete stack_;
+}
+
+UCharsTrie::Iterator &
+UCharsTrie::Iterator::reset() {
+    pos_=initialPos_;
+    remainingMatchLength_=initialRemainingMatchLength_;
+    skipValue_=FALSE;
+    int32_t length=remainingMatchLength_+1;  // Remaining match length.
+    if(maxLength_>0 && length>maxLength_) {
+        length=maxLength_;
+    }
+    str_.truncate(length);
+    pos_+=length;
+    remainingMatchLength_-=length;
+    stack_->setSize(0);
+    return *this;
+}
+
+UBool
+UCharsTrie::Iterator::hasNext() const { return pos_!=NULL || !stack_->isEmpty(); }
+
+UBool
+UCharsTrie::Iterator::next(UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    const UChar *pos=pos_;
+    if(pos==NULL) {
+        if(stack_->isEmpty()) {
+            return FALSE;
+        }
+        // Pop the state off the stack and continue with the next outbound edge of
+        // the branch node.
+        int32_t stackSize=stack_->size();
+        int32_t length=stack_->elementAti(stackSize-1);
+        pos=uchars_+stack_->elementAti(stackSize-2);
+        stack_->setSize(stackSize-2);
+        str_.truncate(length&0xffff);
+        length=(int32_t)((uint32_t)length>>16);
+        if(length>1) {
+            pos=branchNext(pos, length, errorCode);
+            if(pos==NULL) {
+                return TRUE;  // Reached a final value.
+            }
+        } else {
+            str_.append(*pos++);
+        }
+    }
+    if(remainingMatchLength_>=0) {
+        // We only get here if we started in a pending linear-match node
+        // with more than maxLength remaining units.
+        return truncateAndStop();
+    }
+    for(;;) {
+        int32_t node=*pos++;
+        if(node>=kMinValueLead) {
+            if(skipValue_) {
+                pos=skipNodeValue(pos, node);
+                node&=kNodeTypeMask;
+                skipValue_=FALSE;
+            } else {
+                // Deliver value for the string so far.
+                UBool isFinal=(UBool)(node>>15);
+                if(isFinal) {
+                    value_=readValue(pos, node&0x7fff);
+                } else {
+                    value_=readNodeValue(pos, node);
+                }
+                if(isFinal || (maxLength_>0 && str_.length()==maxLength_)) {
+                    pos_=NULL;
+                } else {
+                    // We cannot skip the value right here because it shares its
+                    // lead unit with a match node which we have to evaluate
+                    // next time.
+                    // Instead, keep pos_ on the node lead unit itself.
+                    pos_=pos-1;
+                    skipValue_=TRUE;
+                }
+                return TRUE;
+            }
+        }
+        if(maxLength_>0 && str_.length()==maxLength_) {
+            return truncateAndStop();
+        }
+        if(node<kMinLinearMatch) {
+            if(node==0) {
+                node=*pos++;
+            }
+            pos=branchNext(pos, node+1, errorCode);
+            if(pos==NULL) {
+                return TRUE;  // Reached a final value.
+            }
+        } else {
+            // Linear-match node, append length units to str_.
+            int32_t length=node-kMinLinearMatch+1;
+            if(maxLength_>0 && str_.length()+length>maxLength_) {
+                str_.append(pos, maxLength_-str_.length());
+                return truncateAndStop();
+            }
+            str_.append(pos, length);
+            pos+=length;
+        }
+    }
+}
+
+// Branch node, needs to take the first outbound edge and push state for the rest.
+const UChar *
+UCharsTrie::Iterator::branchNext(const UChar *pos, int32_t length, UErrorCode &errorCode) {
+    while(length>kMaxBranchLinearSubNodeLength) {
+        ++pos;  // ignore the comparison unit
+        // Push state for the greater-or-equal edge.
+        stack_->addElement((int32_t)(skipDelta(pos)-uchars_), errorCode);
+        stack_->addElement(((length-(length>>1))<<16)|str_.length(), errorCode);
+        // Follow the less-than edge.
+        length>>=1;
+        pos=jumpByDelta(pos);
+    }
+    // List of key-value pairs where values are either final values or jump deltas.
+    // Read the first (key, value) pair.
+    UChar trieUnit=*pos++;
+    int32_t node=*pos++;
+    UBool isFinal=(UBool)(node>>15);
+    int32_t value=readValue(pos, node&=0x7fff);
+    pos=skipValue(pos, node);
+    stack_->addElement((int32_t)(pos-uchars_), errorCode);
+    stack_->addElement(((length-1)<<16)|str_.length(), errorCode);
+    str_.append(trieUnit);
+    if(isFinal) {
+        pos_=NULL;
+        value_=value;
+        return NULL;
+    } else {
+        return pos+value;
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/uchriter.cpp b/thirdparty/icu4c/common/uchriter.cpp
new file mode 100644
index 0000000000..bedbabc74c
--- /dev/null
+++ b/thirdparty/icu4c/common/uchriter.cpp
@@ -0,0 +1,367 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1998-2012, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*/
+
+#include "utypeinfo.h"  // for 'typeid' to work
+
+#include "unicode/uchriter.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UCharCharacterIterator)
+
+UCharCharacterIterator::UCharCharacterIterator()
+  : CharacterIterator(),
+  text(0)
+{
+    // never default construct!
+}
+
+UCharCharacterIterator::UCharCharacterIterator(ConstChar16Ptr textPtr,
+                                               int32_t length)
+  : CharacterIterator(textPtr != 0 ? (length>=0 ? length : u_strlen(textPtr)) : 0),
+  text(textPtr)
+{
+}
+
+UCharCharacterIterator::UCharCharacterIterator(ConstChar16Ptr textPtr,
+                                               int32_t length,
+                                               int32_t position)
+  : CharacterIterator(textPtr != 0 ? (length>=0 ? length : u_strlen(textPtr)) : 0, position),
+  text(textPtr)
+{
+}
+
+UCharCharacterIterator::UCharCharacterIterator(ConstChar16Ptr textPtr,
+                                               int32_t length,
+                                               int32_t textBegin,
+                                               int32_t textEnd,
+                                               int32_t position)
+  : CharacterIterator(textPtr != 0 ? (length>=0 ? length : u_strlen(textPtr)) : 0, textBegin, textEnd, position),
+  text(textPtr)
+{
+}
+
+UCharCharacterIterator::UCharCharacterIterator(const UCharCharacterIterator& that)
+: CharacterIterator(that),
+  text(that.text)
+{
+}
+
+UCharCharacterIterator&
+UCharCharacterIterator::operator=(const UCharCharacterIterator& that) {
+    CharacterIterator::operator=(that);
+    text = that.text;
+    return *this;
+}
+
+UCharCharacterIterator::~UCharCharacterIterator() {
+}
+
+UBool
+UCharCharacterIterator::operator==(const ForwardCharacterIterator& that) const {
+    if (this == &that) {
+        return TRUE;
+    }
+    if (typeid(*this) != typeid(that)) {
+        return FALSE;
+    }
+
+    UCharCharacterIterator&    realThat = (UCharCharacterIterator&)that;
+
+    return text == realThat.text
+        && textLength == realThat.textLength
+        && pos == realThat.pos
+        && begin == realThat.begin
+        && end == realThat.end;
+}
+
+int32_t
+UCharCharacterIterator::hashCode() const {
+    return ustr_hashUCharsN(text, textLength) ^ pos ^ begin ^ end;
+}
+
+UCharCharacterIterator*
+UCharCharacterIterator::clone() const {
+    return new UCharCharacterIterator(*this);
+}
+
+UChar
+UCharCharacterIterator::first() {
+    pos = begin;
+    if(pos < end) {
+        return text[pos];
+    } else {
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::firstPostInc() {
+    pos = begin;
+    if(pos < end) {
+        return text[pos++];
+    } else {
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::last() {
+    pos = end;
+    if(pos > begin) {
+        return text[--pos];
+    } else {
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::setIndex(int32_t position) {
+    if(position < begin) {
+        pos = begin;
+    } else if(position > end) {
+        pos = end;
+    } else {
+        pos = position;
+    }
+    if(pos < end) {
+        return text[pos];
+    } else {
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::current() const {
+    if (pos >= begin && pos < end) {
+        return text[pos];
+    } else {
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::next() {
+    if (pos + 1 < end) {
+        return text[++pos];
+    } else {
+        /* make current() return DONE */
+        pos = end;
+        return DONE;
+    }
+}
+
+UChar
+UCharCharacterIterator::nextPostInc() {
+    if (pos < end) {
+        return text[pos++];
+    } else {
+        return DONE;
+    }
+}
+
+UBool
+UCharCharacterIterator::hasNext() {
+    return (UBool)(pos < end ? TRUE : FALSE);
+}
+
+UChar
+UCharCharacterIterator::previous() {
+    if (pos > begin) {
+        return text[--pos];
+    } else {
+        return DONE;
+    }
+}
+
+UBool
+UCharCharacterIterator::hasPrevious() {
+    return (UBool)(pos > begin ? TRUE : FALSE);
+}
+
+UChar32
+UCharCharacterIterator::first32() {
+    pos = begin;
+    if(pos < end) {
+        int32_t i = pos;
+        UChar32 c;
+        U16_NEXT(text, i, end, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::first32PostInc() {
+    pos = begin;
+    if(pos < end) {
+        UChar32 c;
+        U16_NEXT(text, pos, end, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::last32() {
+    pos = end;
+    if(pos > begin) {
+        UChar32 c;
+        U16_PREV(text, begin, pos, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::setIndex32(int32_t position) {
+    if(position < begin) {
+        position = begin;
+    } else if(position > end) {
+        position = end;
+    }
+    if(position < end) {
+        U16_SET_CP_START(text, begin, position);
+        int32_t i = this->pos = position;
+        UChar32 c;
+        U16_NEXT(text, i, end, c);
+        return c;
+    } else {
+        this->pos = position;
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::current32() const {
+    if (pos >= begin && pos < end) {
+        UChar32 c;
+        U16_GET(text, begin, pos, end, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::next32() {
+    if (pos < end) {
+        U16_FWD_1(text, pos, end);
+        if(pos < end) {
+            int32_t i = pos;
+            UChar32 c;
+            U16_NEXT(text, i, end, c);
+            return c;
+        }
+    }
+    /* make current() return DONE */
+    pos = end;
+    return DONE;
+}
+
+UChar32
+UCharCharacterIterator::next32PostInc() {
+    if (pos < end) {
+        UChar32 c;
+        U16_NEXT(text, pos, end, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+UChar32
+UCharCharacterIterator::previous32() {
+    if (pos > begin) {
+        UChar32 c;
+        U16_PREV(text, begin, pos, c);
+        return c;
+    } else {
+        return DONE;
+    }
+}
+
+int32_t
+UCharCharacterIterator::move(int32_t delta, CharacterIterator::EOrigin origin) {
+    switch(origin) {
+    case kStart:
+        pos = begin + delta;
+        break;
+    case kCurrent:
+        pos += delta;
+        break;
+    case kEnd:
+        pos = end + delta;
+        break;
+    default:
+        break;
+    }
+
+    if(pos < begin) {
+        pos = begin;
+    } else if(pos > end) {
+        pos = end;
+    }
+
+    return pos;
+}
+
+int32_t
+UCharCharacterIterator::move32(int32_t delta, CharacterIterator::EOrigin origin) {
+    // this implementation relies on the "safe" version of the UTF macros
+    // (or the trustworthiness of the caller)
+    switch(origin) {
+    case kStart:
+        pos = begin;
+        if(delta > 0) {
+            U16_FWD_N(text, pos, end, delta);
+        }
+        break;
+    case kCurrent:
+        if(delta > 0) {
+            U16_FWD_N(text, pos, end, delta);
+        } else {
+            U16_BACK_N(text, begin, pos, -delta);
+        }
+        break;
+    case kEnd:
+        pos = end;
+        if(delta < 0) {
+            U16_BACK_N(text, begin, pos, -delta);
+        }
+        break;
+    default:
+        break;
+    }
+
+    return pos;
+}
+
+void UCharCharacterIterator::setText(ConstChar16Ptr newText,
+                                     int32_t      newTextLength) {
+    text = newText;
+    if(newText == 0 || newTextLength < 0) {
+        newTextLength = 0;
+    }
+    end = textLength = newTextLength;
+    pos = begin = 0;
+}
+
+void
+UCharCharacterIterator::getText(UnicodeString& result) {
+    result = UnicodeString(text, textLength);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ucln.h b/thirdparty/icu4c/common/ucln.h
new file mode 100644
index 0000000000..fe6666efed
--- /dev/null
+++ b/thirdparty/icu4c/common/ucln.h
@@ -0,0 +1,91 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+* Copyright (C) 2001-2013, International Business Machines
+*                Corporation and others. All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucln.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001July05
+*   created by: George Rhoten
+*/
+
+#ifndef __UCLN_H__
+#define __UCLN_H__
+
+#include "unicode/utypes.h"
+
+/** These are the functions used to register a library's memory cleanup
+ * functions.  Each library should define a single library register function
+ * to call this API.  In the i18n library, it is ucln_i18n_registerCleanup().
+ *
+ * None of the cleanup functions should use a mutex to clean up an API's
+ * allocated memory because a cleanup function is not meant to be thread safe,
+ * and plenty of data cannot be reference counted in order to make sure that
+ * no one else needs the allocated data.
+ *
+ * In order to make a cleanup function get called when u_cleanup is called,
+ * You should add your function to the library specific cleanup function.
+ * If the cleanup function is not in the common library, the code that
+ * allocates the memory should call the library specific cleanup function.
+ * For instance, in the i18n library, any memory allocated statically must
+ * call ucln_i18n_registerCleanup() from the ucln_in.h header.  These library
+ * cleanup functions are needed in order to prevent a circular dependency
+ * between the common library and any other library.
+ *
+ * The order of the cleanup is very important.  In general, an API that
+ * depends on a second API should be cleaned up before the second API.
+ * For instance, the default converter in ustring depends upon the converter
+ * API.  So the default converter should be closed before the converter API
+ * has its cache flushed.  This will prevent any memory leaks due to
+ * reference counting.
+ *
+ * Please see common/ucln_cmn.{h,c} and i18n/ucln_in.{h,c} for examples.
+ */
+
+/**
+ * Data Type for cleanup function selector. These roughly correspond to libraries.
+ */
+typedef enum ECleanupLibraryType {
+    UCLN_START = -1,
+    UCLN_UPLUG,     /* ICU plugins */
+    UCLN_CUSTOM,    /* Custom is for anyone else. */
+    UCLN_CTESTFW,
+    UCLN_TOOLUTIL,
+    UCLN_LAYOUTEX,
+    UCLN_LAYOUT,
+    UCLN_IO,
+    UCLN_I18N,
+    UCLN_COMMON /* This must be the last one to cleanup. */
+} ECleanupLibraryType;
+
+/**
+ * Data type for cleanup function pointer
+ */
+U_CDECL_BEGIN
+typedef UBool U_CALLCONV cleanupFunc(void);
+typedef void U_CALLCONV initFunc(UErrorCode *);
+U_CDECL_END
+
+/**
+ * Register a cleanup function
+ * @param type which library to register for.
+ * @param func the function pointer
+ */
+U_CAPI void U_EXPORT2 ucln_registerCleanup(ECleanupLibraryType type,
+                                           cleanupFunc *func);
+
+/**
+ * Request cleanup for one specific library.
+ * Not thread safe.
+ * @param type which library to cleanup
+ */
+U_CAPI void U_EXPORT2 ucln_cleanupOne(ECleanupLibraryType type);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucln_cmn.cpp b/thirdparty/icu4c/common/ucln_cmn.cpp
new file mode 100644
index 0000000000..f3e07c6b89
--- /dev/null
+++ b/thirdparty/icu4c/common/ucln_cmn.cpp
@@ -0,0 +1,124 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2001-2014, International Business Machines
+*                Corporation and others. All Rights Reserved.
+******************************************************************************
+*   file name:  ucln_cmn.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001July05
+*   created by: George Rhoten
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uclean.h"
+#include "cmemory.h"
+#include "mutex.h"
+#include "uassert.h"
+#include "ucln.h"
+#include "ucln_cmn.h"
+#include "utracimp.h"
+#include "umutex.h"
+
+/**  Auto-client for UCLN_COMMON **/
+#define UCLN_TYPE_IS_COMMON
+#include "ucln_imp.h"
+
+static cleanupFunc *gCommonCleanupFunctions[UCLN_COMMON_COUNT];
+static cleanupFunc *gLibCleanupFunctions[UCLN_COMMON];
+
+
+/************************************************
+ The cleanup order is important in this function.
+ Please be sure that you have read ucln.h
+ ************************************************/
+U_CAPI void U_EXPORT2
+u_cleanup(void)
+{
+    UTRACE_ENTRY_OC(UTRACE_U_CLEANUP);
+    icu::umtx_lock(NULL);     /* Force a memory barrier, so that we are sure to see   */
+    icu::umtx_unlock(NULL);   /*   all state left around by any other threads.        */
+
+    ucln_lib_cleanup();
+
+    cmemory_cleanup();       /* undo any heap functions set by u_setMemoryFunctions(). */
+    UTRACE_EXIT();           /* Must be before utrace_cleanup(), which turns off tracing. */
+/*#if U_ENABLE_TRACING*/
+    utrace_cleanup();
+/*#endif*/
+}
+
+U_CAPI void U_EXPORT2 ucln_cleanupOne(ECleanupLibraryType libType) 
+{
+    if (gLibCleanupFunctions[libType])
+    {
+        gLibCleanupFunctions[libType]();
+        gLibCleanupFunctions[libType] = NULL;
+    }
+}
+
+U_CFUNC void
+ucln_common_registerCleanup(ECleanupCommonType type,
+                            cleanupFunc *func)
+{
+    // Thread safety messiness: From ticket 10295, calls to registerCleanup() may occur
+    // concurrently. Although such cases should be storing the same value, they raise errors
+    // from the thread sanity checker. Doing the store within a mutex avoids those.
+    // BUT that can trigger a recursive entry into std::call_once() in umutex.cpp when this code,
+    // running from the call_once function, tries to grab the ICU global mutex, which
+    // re-enters the mutex init path. So, work-around by special casing UCLN_COMMON_MUTEX, not
+    // using the ICU global mutex for it.
+    //
+    // No other point in ICU uses std::call_once().
+
+    U_ASSERT(UCLN_COMMON_START < type && type < UCLN_COMMON_COUNT);
+    if (type == UCLN_COMMON_MUTEX) {
+        gCommonCleanupFunctions[type] = func;
+    } else if (UCLN_COMMON_START < type && type < UCLN_COMMON_COUNT)  {
+        icu::Mutex m;     // See ticket 10295 for discussion.
+        gCommonCleanupFunctions[type] = func;
+    }
+#if !UCLN_NO_AUTO_CLEANUP && (defined(UCLN_AUTO_ATEXIT) || defined(UCLN_AUTO_LOCAL))
+    ucln_registerAutomaticCleanup();
+#endif
+}
+
+// Note: ucln_registerCleanup() is called with the ICU global mutex locked.
+//       Be aware if adding anything to the function.
+//       See ticket 10295 for discussion.
+
+U_CAPI void U_EXPORT2
+ucln_registerCleanup(ECleanupLibraryType type,
+                     cleanupFunc *func)
+{
+    U_ASSERT(UCLN_START < type && type < UCLN_COMMON);
+    if (UCLN_START < type && type < UCLN_COMMON)
+    {
+        gLibCleanupFunctions[type] = func;
+    }
+}
+
+U_CFUNC UBool ucln_lib_cleanup(void) {
+    int32_t libType = UCLN_START;
+    int32_t commonFunc = UCLN_COMMON_START;
+
+    for (libType++; libType<UCLN_COMMON; libType++) {
+        ucln_cleanupOne(static_cast<ECleanupLibraryType>(libType));
+    }
+
+    for (commonFunc++; commonFunc<UCLN_COMMON_COUNT; commonFunc++) {
+        if (gCommonCleanupFunctions[commonFunc])
+        {
+            gCommonCleanupFunctions[commonFunc]();
+            gCommonCleanupFunctions[commonFunc] = NULL;
+        }
+    }
+#if !UCLN_NO_AUTO_CLEANUP && (defined(UCLN_AUTO_ATEXIT) || defined(UCLN_AUTO_LOCAL))
+    ucln_unRegisterAutomaticCleanup();
+#endif
+    return TRUE;
+}
diff --git a/thirdparty/icu4c/common/ucln_cmn.h b/thirdparty/icu4c/common/ucln_cmn.h
new file mode 100644
index 0000000000..44b73e94da
--- /dev/null
+++ b/thirdparty/icu4c/common/ucln_cmn.h
@@ -0,0 +1,77 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2001-2016, International Business Machines
+*                Corporation and others. All Rights Reserved.
+******************************************************************************
+*   file name:  ucln_cmn.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001July05
+*   created by: George Rhoten
+*/
+
+#ifndef __UCLN_CMN_H__
+#define __UCLN_CMN_H__
+
+#include "unicode/utypes.h"
+#include "ucln.h"
+
+/* These are the cleanup functions for various APIs. */
+/* @return true if cleanup complete successfully.*/
+U_CFUNC UBool utrace_cleanup(void);
+
+U_CFUNC UBool ucln_lib_cleanup(void);
+
+/*
+Please keep the order of enums declared in same order
+as the cleanup functions are suppose to be called. */
+typedef enum ECleanupCommonType {
+    UCLN_COMMON_START = -1,
+    UCLN_COMMON_NUMPARSE_UNISETS,
+    UCLN_COMMON_USPREP,
+    UCLN_COMMON_BREAKITERATOR,
+    UCLN_COMMON_RBBI,
+    UCLN_COMMON_SERVICE,
+    UCLN_COMMON_LOCALE_KEY_TYPE,
+    UCLN_COMMON_LOCALE,
+    UCLN_COMMON_LOCALE_ALIAS,
+    UCLN_COMMON_LOCALE_KNOWN_CANONICALIZED,
+    UCLN_COMMON_LOCALE_AVAILABLE,
+    UCLN_COMMON_LIKELY_SUBTAGS,
+    UCLN_COMMON_LOCALE_DISTANCE,
+    UCLN_COMMON_ULOC,
+    UCLN_COMMON_CURRENCY,
+    UCLN_COMMON_LOADED_NORMALIZER2,
+    UCLN_COMMON_NORMALIZER2,
+    UCLN_COMMON_CHARACTERPROPERTIES,
+    UCLN_COMMON_USET,
+    UCLN_COMMON_UNAMES,
+    UCLN_COMMON_UPROPS,
+    UCLN_COMMON_UCNV,
+    UCLN_COMMON_UCNV_IO,
+    UCLN_COMMON_UDATA,
+    UCLN_COMMON_PUTIL,
+    UCLN_COMMON_UINIT,
+
+    /*
+       Unified caches caches collation stuff. Collation data structures
+       contain resource bundles which means that unified cache cleanup
+       must happen before resource bundle clean up.
+    */
+    UCLN_COMMON_UNIFIED_CACHE,
+    UCLN_COMMON_URES,
+    UCLN_COMMON_MUTEX,    // Mutexes should be the last to be cleaned up.
+    UCLN_COMMON_COUNT /* This must be last */
+} ECleanupCommonType;
+
+/* Main library cleanup registration function. */
+/* See common/ucln.h for details on adding a cleanup function. */
+/* Note: the global mutex must not be held when calling this function. */
+U_CFUNC void U_EXPORT2 ucln_common_registerCleanup(ECleanupCommonType type,
+                                                   cleanupFunc *func);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucln_imp.h b/thirdparty/icu4c/common/ucln_imp.h
new file mode 100644
index 0000000000..63a54c86f6
--- /dev/null
+++ b/thirdparty/icu4c/common/ucln_imp.h
@@ -0,0 +1,182 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+* Copyright (C) 2009-2011, International Business Machines
+*                Corporation and others. All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucln_imp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   This file contains the platform specific implementation of per-library cleanup.
+*
+*/
+
+
+#ifndef __UCLN_IMP_H__
+#define __UCLN_IMP_H__
+
+#include "ucln.h"
+#include <stdlib.h>
+
+/**
+ * Auto cleanup of ICU libraries
+ * There are several methods in per library cleanup of icu libraries:
+ * 1) Compiler/Platform based cleanup:
+ *   a) Windows MSVC uses DllMain()
+ *   b) GCC uses destructor function attribute
+ *   c) Sun Studio, AIX VA, and HP-UX aCC uses a linker option to set the exit function
+ * 2) Using atexit()
+ * 3) Implementing own automatic cleanup functions
+ *
+ * For option 1, ensure that UCLN_NO_AUTO_CLEANUP is set to 0 by using --enable-auto-cleanup
+ * configure option or by otherwise setting UCLN_NO_AUTO_CLEANUP to 0
+ * For option 2, follow option 1 and also define UCLN_AUTO_ATEXIT
+ * For option 3, follow option 1 and also define UCLN_AUTO_LOCAL (see below for more information)
+ */
+
+#if !UCLN_NO_AUTO_CLEANUP
+
+/*
+ * The following declarations are for when UCLN_AUTO_LOCAL or UCLN_AUTO_ATEXIT
+ * are defined. They are commented out because they are static and will be defined
+ * later. The information is still here to provide some guidance for the developer
+ * who chooses to use UCLN_AUTO_LOCAL.
+ */
+/**
+ * Give the library an opportunity to register an automatic cleanup.
+ * This may be called more than once.
+ */
+/*static void ucln_registerAutomaticCleanup();*/
+/**
+ * Unregister an automatic cleanup, if possible. Called from cleanup.
+ */
+/*static void ucln_unRegisterAutomaticCleanup();*/
+
+#ifdef UCLN_TYPE_IS_COMMON
+#   define UCLN_CLEAN_ME_UP u_cleanup()
+#else
+#   define UCLN_CLEAN_ME_UP ucln_cleanupOne(UCLN_TYPE)
+#endif
+
+/* ------------ automatic cleanup: registration. Choose ONE ------- */
+#if defined(UCLN_AUTO_LOCAL)
+/* To use:
+ *  1. define UCLN_AUTO_LOCAL,
+ *  2. create ucln_local_hook.c containing implementations of
+ *           static void ucln_registerAutomaticCleanup()
+ *           static void ucln_unRegisterAutomaticCleanup()
+ */
+#include "ucln_local_hook.c"
+
+#elif defined(UCLN_AUTO_ATEXIT)
+/*
+ * Use the ANSI C 'atexit' function. Note that this mechanism does not
+ * guarantee the order of cleanup relative to other users of ICU!
+ */
+static UBool gAutoCleanRegistered = false;
+
+static void ucln_atexit_handler()
+{
+    UCLN_CLEAN_ME_UP;
+}
+
+static void ucln_registerAutomaticCleanup()
+{
+    if(!gAutoCleanRegistered) {
+        gAutoCleanRegistered = true;
+        atexit(&ucln_atexit_handler);
+    }
+}
+
+static void ucln_unRegisterAutomaticCleanup () {
+}
+/* ------------end of automatic cleanup: registration. ------- */
+
+#elif defined (UCLN_FINI)
+/**
+ * If UCLN_FINI is defined, it is the (versioned, etc) name of a cleanup
+ * entrypoint. Add a stub to call ucln_cleanupOne
+ * Used on AIX, Solaris, and HP-UX
+ */
+U_CAPI void U_EXPORT2 UCLN_FINI (void);
+
+U_CAPI void U_EXPORT2 UCLN_FINI ()
+{
+    /* This function must be defined, if UCLN_FINI is defined, else link error. */
+     UCLN_CLEAN_ME_UP;
+}
+
+/* Windows: DllMain */
+#elif U_PLATFORM_HAS_WIN32_API
+/* 
+ * ICU's own DllMain.
+ */
+
+/* these are from putil.c */
+/* READ READ READ READ!    Are you getting compilation errors from windows.h?
+          Any source file which includes this (ucln_imp.h) header MUST 
+          be defined with language extensions ON. */
+#ifndef WIN32_LEAN_AND_MEAN
+#   define WIN32_LEAN_AND_MEAN
+#endif
+#   define VC_EXTRALEAN
+#   define NOUSER
+#   define NOSERVICE
+#   define NOIME
+#   define NOMCX
+#   include <windows.h>
+/*
+ * This is a stub DllMain function with icu specific process handling code.
+ */
+BOOL WINAPI DllMain(HINSTANCE hinstDLL, DWORD fdwReason, LPVOID lpvReserved)
+{
+    BOOL status = true;
+
+    switch(fdwReason) {
+        case DLL_PROCESS_ATTACH:
+             /* ICU does not trap process attach, but must pass these through properly. */
+            /* ICU specific process attach could go here */
+            break;
+
+        case DLL_PROCESS_DETACH:
+            /* Here is the one we actually care about. */
+
+            UCLN_CLEAN_ME_UP;
+
+            break;
+
+        case DLL_THREAD_ATTACH:
+            /* ICU does not trap thread attach, but must pass these through properly. */
+            /* ICU specific thread attach could go here */
+            break;
+
+        case DLL_THREAD_DETACH:
+            /* ICU does not trap thread detach, but must pass these through properly. */
+            /* ICU specific thread detach could go here */
+            break;
+
+    }
+    return status;
+}
+
+#elif defined(__GNUC__)
+/* GCC - use __attribute((destructor)) */
+static void ucln_destructor()   __attribute__((destructor)) ;
+
+static void ucln_destructor() 
+{
+    UCLN_CLEAN_ME_UP;
+}
+
+#endif
+
+#endif /* UCLN_NO_AUTO_CLEANUP */
+
+#else
+#error This file can only be included once.
+#endif
diff --git a/thirdparty/icu4c/common/ucmndata.cpp b/thirdparty/icu4c/common/ucmndata.cpp
new file mode 100644
index 0000000000..ba2310bb7a
--- /dev/null
+++ b/thirdparty/icu4c/common/ucmndata.cpp
@@ -0,0 +1,393 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+
+/*------------------------------------------------------------------------------
+ *
+ *   UCommonData   An abstract interface for dealing with ICU Common Data Files.
+ *                 ICU Common Data Files are a grouping of a number of individual
+ *                 data items (resources, converters, tables, anything) into a
+ *                 single file or dll.  The combined format includes a table of
+ *                 contents for locating the individual items by name.
+ *
+ *                 Two formats for the table of contents are supported, which is
+ *                 why there is an abstract inteface involved.
+ *
+ */
+
+#include "unicode/utypes.h"
+#include "unicode/udata.h"
+#include "cstring.h"
+#include "ucmndata.h"
+#include "udatamem.h"
+
+#if defined(UDATA_DEBUG) || defined(UDATA_DEBUG_DUMP)
+#   include <stdio.h>
+#endif
+
+U_CFUNC uint16_t
+udata_getHeaderSize(const DataHeader *udh) {
+    if(udh==NULL) {
+        return 0;
+    } else if(udh->info.isBigEndian==U_IS_BIG_ENDIAN) {
+        /* same endianness */
+        return udh->dataHeader.headerSize;
+    } else {
+        /* opposite endianness */
+        uint16_t x=udh->dataHeader.headerSize;
+        return (uint16_t)((x<<8)|(x>>8));
+    }
+}
+
+U_CFUNC uint16_t
+udata_getInfoSize(const UDataInfo *info) {
+    if(info==NULL) {
+        return 0;
+    } else if(info->isBigEndian==U_IS_BIG_ENDIAN) {
+        /* same endianness */
+        return info->size;
+    } else {
+        /* opposite endianness */
+        uint16_t x=info->size;
+        return (uint16_t)((x<<8)|(x>>8));
+    }
+}
+
+/*-----------------------------------------------------------------------------*
+ *                                                                             *
+ *  Pointer TOCs.   TODO: This form of table-of-contents should be removed     *
+ *                  because DLLs must be relocated on loading to correct the   *
+ *                  pointer values and this operation makes shared memory      *
+ *                  mapping of the data much less likely to work.              *
+ *                                                                             *
+ *-----------------------------------------------------------------------------*/
+typedef struct {
+    const char       *entryName;
+    const DataHeader *pHeader;
+} PointerTOCEntry;
+
+
+typedef struct  {
+    uint32_t          count;
+    uint32_t          reserved;
+    /**
+     * Variable-length array declared with length 1 to disable bounds checkers.
+     * The actual array length is in the count field.
+     */
+    PointerTOCEntry   entry[1];
+}  PointerTOC;
+
+
+/* definition of OffsetTOC struct types moved to ucmndata.h */
+
+/*-----------------------------------------------------------------------------*
+ *                                                                             *
+ *    entry point lookup implementations                                       *
+ *                                                                             *
+ *-----------------------------------------------------------------------------*/
+
+#ifndef MIN
+#define MIN(a,b) (((a)<(b)) ? (a) : (b))
+#endif
+
+/**
+ * Compare strings where we know the shared prefix length,
+ * and advance the prefix length as we find that the strings share even more characters.
+ */
+static int32_t
+strcmpAfterPrefix(const char *s1, const char *s2, int32_t *pPrefixLength) {
+    int32_t pl=*pPrefixLength;
+    int32_t cmp=0;
+    s1+=pl;
+    s2+=pl;
+    for(;;) {
+        int32_t c1=(uint8_t)*s1++;
+        int32_t c2=(uint8_t)*s2++;
+        cmp=c1-c2;
+        if(cmp!=0 || c1==0) {  /* different or done */
+            break;
+        }
+        ++pl;  /* increment shared same-prefix length */
+    }
+    *pPrefixLength=pl;
+    return cmp;
+}
+
+static int32_t
+offsetTOCPrefixBinarySearch(const char *s, const char *names,
+                            const UDataOffsetTOCEntry *toc, int32_t count) {
+    int32_t start=0;
+    int32_t limit=count;
+    /*
+     * Remember the shared prefix between s, start and limit,
+     * and don't compare that shared prefix again.
+     * The shared prefix should get longer as we narrow the [start, limit[ range.
+     */
+    int32_t startPrefixLength=0;
+    int32_t limitPrefixLength=0;
+    if(count==0) {
+        return -1;
+    }
+    /*
+     * Prime the prefix lengths so that we don't keep prefixLength at 0 until
+     * both the start and limit indexes have moved.
+     * At the same time, we find if s is one of the start and (limit-1) names,
+     * and if not, exclude them from the actual binary search.
+     */
+    if(0==strcmpAfterPrefix(s, names+toc[0].nameOffset, &startPrefixLength)) {
+        return 0;
+    }
+    ++start;
+    --limit;
+    if(0==strcmpAfterPrefix(s, names+toc[limit].nameOffset, &limitPrefixLength)) {
+        return limit;
+    }
+    while(start<limit) {
+        int32_t i=(start+limit)/2;
+        int32_t prefixLength=MIN(startPrefixLength, limitPrefixLength);
+        int32_t cmp=strcmpAfterPrefix(s, names+toc[i].nameOffset, &prefixLength);
+        if(cmp<0) {
+            limit=i;
+            limitPrefixLength=prefixLength;
+        } else if(cmp==0) {
+            return i;
+        } else {
+            start=i+1;
+            startPrefixLength=prefixLength;
+        }
+    }
+    return -1;
+}
+
+static int32_t
+pointerTOCPrefixBinarySearch(const char *s, const PointerTOCEntry *toc, int32_t count) {
+    int32_t start=0;
+    int32_t limit=count;
+    /*
+     * Remember the shared prefix between s, start and limit,
+     * and don't compare that shared prefix again.
+     * The shared prefix should get longer as we narrow the [start, limit[ range.
+     */
+    int32_t startPrefixLength=0;
+    int32_t limitPrefixLength=0;
+    if(count==0) {
+        return -1;
+    }
+    /*
+     * Prime the prefix lengths so that we don't keep prefixLength at 0 until
+     * both the start and limit indexes have moved.
+     * At the same time, we find if s is one of the start and (limit-1) names,
+     * and if not, exclude them from the actual binary search.
+     */
+    if(0==strcmpAfterPrefix(s, toc[0].entryName, &startPrefixLength)) {
+        return 0;
+    }
+    ++start;
+    --limit;
+    if(0==strcmpAfterPrefix(s, toc[limit].entryName, &limitPrefixLength)) {
+        return limit;
+    }
+    while(start<limit) {
+        int32_t i=(start+limit)/2;
+        int32_t prefixLength=MIN(startPrefixLength, limitPrefixLength);
+        int32_t cmp=strcmpAfterPrefix(s, toc[i].entryName, &prefixLength);
+        if(cmp<0) {
+            limit=i;
+            limitPrefixLength=prefixLength;
+        } else if(cmp==0) {
+            return i;
+        } else {
+            start=i+1;
+            startPrefixLength=prefixLength;
+        }
+    }
+    return -1;
+}
+
+U_CDECL_BEGIN
+static uint32_t U_CALLCONV
+offsetTOCEntryCount(const UDataMemory *pData) {
+    int32_t          retVal=0;
+    const UDataOffsetTOC *toc = (UDataOffsetTOC *)pData->toc;
+    if (toc != NULL) {
+        retVal = toc->count;
+    }
+    return retVal;
+}
+
+static const DataHeader * U_CALLCONV
+offsetTOCLookupFn(const UDataMemory *pData,
+                  const char *tocEntryName,
+                  int32_t *pLength,
+                  UErrorCode *pErrorCode) {
+    (void)pErrorCode;
+    const UDataOffsetTOC  *toc = (UDataOffsetTOC *)pData->toc;
+    if(toc!=NULL) {
+        const char *base=(const char *)toc;
+        int32_t number, count=(int32_t)toc->count;
+
+        /* perform a binary search for the data in the common data's table of contents */
+#if defined (UDATA_DEBUG_DUMP)
+        /* list the contents of the TOC each time .. not recommended */
+        for(number=0; number<count; ++number) {
+            fprintf(stderr, "\tx%d: %s\n", number, &base[toc->entry[number].nameOffset]);
+        }
+#endif
+        number=offsetTOCPrefixBinarySearch(tocEntryName, base, toc->entry, count);
+        if(number>=0) {
+            /* found it */
+            const UDataOffsetTOCEntry *entry=toc->entry+number;
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "%s: Found.\n", tocEntryName);
+#endif
+            if((number+1) < count) {
+                *pLength = (int32_t)(entry[1].dataOffset - entry->dataOffset);
+            } else {
+                *pLength = -1;
+            }
+            return (const DataHeader *)(base+entry->dataOffset);
+        } else {
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "%s: Not found.\n", tocEntryName);
+#endif
+            return NULL;
+        }
+    } else {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "returning header\n");
+#endif
+
+        return pData->pHeader;
+    }
+}
+
+
+static uint32_t U_CALLCONV pointerTOCEntryCount(const UDataMemory *pData) {
+    const PointerTOC *toc = (PointerTOC *)pData->toc;
+    return (uint32_t)((toc != NULL) ? (toc->count) : 0);
+}
+
+static const DataHeader * U_CALLCONV pointerTOCLookupFn(const UDataMemory *pData,
+                   const char *name,
+                   int32_t *pLength,
+                   UErrorCode *pErrorCode) {
+    (void)pErrorCode;
+    if(pData->toc!=NULL) {
+        const PointerTOC *toc = (PointerTOC *)pData->toc;
+        int32_t number, count=(int32_t)toc->count;
+
+#if defined (UDATA_DEBUG_DUMP)
+        /* list the contents of the TOC each time .. not recommended */
+        for(number=0; number<count; ++number) {
+            fprintf(stderr, "\tx%d: %s\n", number, toc->entry[number].entryName);
+        }
+#endif
+        number=pointerTOCPrefixBinarySearch(name, toc->entry, count);
+        if(number>=0) {
+            /* found it */
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "%s: Found.\n", toc->entry[number].entryName);
+#endif
+            *pLength=-1;
+            return UDataMemory_normalizeDataPointer(toc->entry[number].pHeader);
+        } else {
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "%s: Not found.\n", name);
+#endif
+            return NULL;
+        }
+    } else {
+        return pData->pHeader;
+    }
+}
+U_CDECL_END
+
+
+static const commonDataFuncs CmnDFuncs = {offsetTOCLookupFn,  offsetTOCEntryCount};
+static const commonDataFuncs ToCPFuncs = {pointerTOCLookupFn, pointerTOCEntryCount};
+
+
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *  checkCommonData   Validate the format of a common data file.        *
+ *                    Fill in the virtual function ptr based on TOC type *
+ *                    If the data is invalid, close the UDataMemory     *
+ *                    and set the appropriate error code.               *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+U_CFUNC void udata_checkCommonData(UDataMemory *udm, UErrorCode *err) {
+    if (U_FAILURE(*err)) {
+        return;
+    }
+
+    if(udm==NULL || udm->pHeader==NULL) {
+      *err=U_INVALID_FORMAT_ERROR;
+    } else if(!(udm->pHeader->dataHeader.magic1==0xda &&
+        udm->pHeader->dataHeader.magic2==0x27 &&
+        udm->pHeader->info.isBigEndian==U_IS_BIG_ENDIAN &&
+        udm->pHeader->info.charsetFamily==U_CHARSET_FAMILY)
+        ) {
+        /* header not valid */
+        *err=U_INVALID_FORMAT_ERROR;
+    }
+    else if (udm->pHeader->info.dataFormat[0]==0x43 &&
+        udm->pHeader->info.dataFormat[1]==0x6d &&
+        udm->pHeader->info.dataFormat[2]==0x6e &&
+        udm->pHeader->info.dataFormat[3]==0x44 &&
+        udm->pHeader->info.formatVersion[0]==1
+        ) {
+        /* dataFormat="CmnD" */
+        udm->vFuncs = &CmnDFuncs;
+        udm->toc=(const char *)udm->pHeader+udata_getHeaderSize(udm->pHeader);
+    }
+    else if(udm->pHeader->info.dataFormat[0]==0x54 &&
+        udm->pHeader->info.dataFormat[1]==0x6f &&
+        udm->pHeader->info.dataFormat[2]==0x43 &&
+        udm->pHeader->info.dataFormat[3]==0x50 &&
+        udm->pHeader->info.formatVersion[0]==1
+        ) {
+        /* dataFormat="ToCP" */
+        udm->vFuncs = &ToCPFuncs;
+        udm->toc=(const char *)udm->pHeader+udata_getHeaderSize(udm->pHeader);
+    }
+    else {
+        /* dataFormat not recognized */
+        *err=U_INVALID_FORMAT_ERROR;
+    }
+
+    if (U_FAILURE(*err)) {
+        /* If the data is no good and we memory-mapped it ourselves,
+         *  close the memory mapping so it doesn't leak.  Note that this has
+         *  no effect on non-memory mapped data, other than clearing fields in udm.
+         */
+        udata_close(udm);
+    }
+}
+
+/*
+ * TODO: Add a udata_swapPackageHeader() function that swaps an ICU .dat package
+ * header but not its sub-items.
+ * This function will be needed for automatic runtime swapping.
+ * Sub-items should not be swapped to limit the swapping to the parts of the
+ * package that are actually used.
+ *
+ * Since lengths of items are implicit in the order and offsets of their
+ * ToC entries, and since offsets are relative to the start of the ToC,
+ * a swapped version may need to generate a different data structure
+ * with pointers to the original data items and with their lengths
+ * (-1 for the last one if it is not known), and maybe even pointers to the
+ * swapped versions of the items.
+ * These pointers to swapped versions would establish a cache;
+ * instead, each open data item could simply own the storage for its swapped
+ * data. This fits better with the current design.
+ *
+ * markus 2003sep18 Jitterbug 2235
+ */
diff --git a/thirdparty/icu4c/common/ucmndata.h b/thirdparty/icu4c/common/ucmndata.h
new file mode 100644
index 0000000000..c3eba9f4d0
--- /dev/null
+++ b/thirdparty/icu4c/common/ucmndata.h
@@ -0,0 +1,117 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+
+/*----------------------------------------------------------------------------------
+ *
+ *   UCommonData   An abstract interface for dealing with ICU Common Data Files.
+ *                 ICU Common Data Files are a grouping of a number of individual
+ *                 data items (resources, converters, tables, anything) into a
+ *                 single file or dll.  The combined format includes a table of
+ *                 contents for locating the individual items by name.
+ *
+ *                 Two formats for the table of contents are supported, which is
+ *                 why there is an abstract inteface involved.
+ *
+ *                 These functions are part of the ICU internal implementation, and
+ *                 are not inteded to be used directly by applications.
+ */
+
+#ifndef __UCMNDATA_H__
+#define __UCMNDATA_H__
+
+#include "unicode/udata.h"
+#include "umapfile.h"
+
+
+#define COMMON_DATA_NAME U_ICUDATA_NAME
+
+typedef struct  {
+    uint16_t    headerSize;
+    uint8_t     magic1;
+    uint8_t     magic2;
+} MappedData;
+
+
+typedef struct  {
+    MappedData  dataHeader;
+    UDataInfo   info;
+} DataHeader;
+
+typedef struct {
+    uint32_t nameOffset;
+    uint32_t dataOffset;
+} UDataOffsetTOCEntry;
+
+typedef struct {
+    uint32_t count;
+    /**
+     * Variable-length array declared with length 1 to disable bounds checkers.
+     * The actual array length is in the count field.
+     */
+    UDataOffsetTOCEntry entry[1];
+} UDataOffsetTOC;
+
+/**
+ * Get the header size from a const DataHeader *udh.
+ * Handles opposite-endian data.
+ *
+ * @internal
+ */
+U_CFUNC uint16_t
+udata_getHeaderSize(const DataHeader *udh);
+
+/**
+ * Get the UDataInfo.size from a const UDataInfo *info.
+ * Handles opposite-endian data.
+ *
+ * @internal
+ */
+U_CFUNC uint16_t
+udata_getInfoSize(const UDataInfo *info);
+
+U_CDECL_BEGIN
+/*
+ *  "Virtual" functions for data lookup.
+ *  To call one, given a UDataMemory *p, the code looks like this:
+ *     p->vFuncs.Lookup(p, tocEntryName, pErrorCode);
+ *          (I sure do wish this was written in C++, not C)
+ */
+
+typedef const DataHeader *
+(U_CALLCONV * LookupFn)(const UDataMemory *pData,
+                        const char *tocEntryName,
+                        int32_t *pLength,
+                        UErrorCode *pErrorCode);
+
+typedef uint32_t
+(U_CALLCONV * NumEntriesFn)(const UDataMemory *pData);
+
+U_CDECL_END
+
+typedef struct {
+    LookupFn      Lookup;
+    NumEntriesFn  NumEntries; 
+} commonDataFuncs;
+
+
+/*
+ *  Functions to check whether a UDataMemory refers to memory containing 
+ *     a recognizable header and table of contents a Common Data Format
+ *
+ *     If a valid header and TOC are found,
+ *         set the CommonDataFuncs function dispatch vector in the UDataMemory
+ *             to point to the right functions for the TOC type.
+ *     otherwise
+ *         set an errorcode.
+ */
+U_CFUNC void udata_checkCommonData(UDataMemory *pData, UErrorCode *pErrorCode);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv.cpp b/thirdparty/icu4c/common/ucnv.cpp
new file mode 100644
index 0000000000..5dcf35e043
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv.cpp
@@ -0,0 +1,2910 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1998-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  ucnv.c:
+*  Implements APIs for the ICU's codeset conversion library;
+*  mostly calls through internal functions;
+*  created by Bertrand A. Damiba
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/04/99    helena      Fixed internal header inclusion.
+*   05/09/00    helena      Added implementation to handle fallback mappings.
+*   06/20/2000  helena      OS/400 port changes; mostly typecast.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include <memory>
+
+#include "unicode/ustring.h"
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_err.h"
+#include "unicode/uset.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "putilimp.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "utracimp.h"
+#include "ustr_imp.h"
+#include "ucnv_imp.h"
+#include "ucnv_cnv.h"
+#include "ucnv_bld.h"
+
+/* size of intermediate and preflighting buffers in ucnv_convert() */
+#define CHUNK_SIZE 1024
+
+typedef struct UAmbiguousConverter {
+    const char *name;
+    const UChar variant5c;
+} UAmbiguousConverter;
+
+static const UAmbiguousConverter ambiguousConverters[]={
+    { "ibm-897_P100-1995", 0xa5 },
+    { "ibm-942_P120-1999", 0xa5 },
+    { "ibm-943_P130-1999", 0xa5 },
+    { "ibm-946_P100-1995", 0xa5 },
+    { "ibm-33722_P120-1999", 0xa5 },
+    { "ibm-1041_P100-1995", 0xa5 },
+    /*{ "ibm-54191_P100-2006", 0xa5 },*/
+    /*{ "ibm-62383_P100-2007", 0xa5 },*/
+    /*{ "ibm-891_P100-1995", 0x20a9 },*/
+    { "ibm-944_P100-1995", 0x20a9 },
+    { "ibm-949_P110-1999", 0x20a9 },
+    { "ibm-1363_P110-1997", 0x20a9 },
+    { "ISO_2022,locale=ko,version=0", 0x20a9 },
+    { "ibm-1088_P100-1995", 0x20a9 }
+};
+
+/*Calls through createConverter */
+U_CAPI UConverter* U_EXPORT2
+ucnv_open (const char *name,
+                       UErrorCode * err)
+{
+    UConverter *r;
+
+    if (err == NULL || U_FAILURE (*err)) {
+        return NULL;
+    }
+
+    r =  ucnv_createConverter(NULL, name, err);
+    return r;
+}
+
+U_CAPI UConverter* U_EXPORT2 
+ucnv_openPackage   (const char *packageName, const char *converterName, UErrorCode * err)
+{
+    return ucnv_createConverterFromPackage(packageName, converterName,  err);
+}
+
+/*Extracts the UChar* to a char* and calls through createConverter */
+U_CAPI UConverter*   U_EXPORT2
+ucnv_openU (const UChar * name,
+                         UErrorCode * err)
+{
+    char asciiName[UCNV_MAX_CONVERTER_NAME_LENGTH];
+
+    if (err == NULL || U_FAILURE(*err))
+        return NULL;
+    if (name == NULL)
+        return ucnv_open (NULL, err);
+    if (u_strlen(name) >= UCNV_MAX_CONVERTER_NAME_LENGTH)
+    {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    return ucnv_open(u_austrcpy(asciiName, name), err);
+}
+
+/* Copy the string that is represented by the UConverterPlatform enum
+ * @param platformString An output buffer
+ * @param platform An enum representing a platform
+ * @return the length of the copied string.
+ */
+static int32_t
+ucnv_copyPlatformString(char *platformString, UConverterPlatform pltfrm)
+{
+    switch (pltfrm)
+    {
+    case UCNV_IBM:
+        uprv_strcpy(platformString, "ibm-");
+        return 4;
+    case UCNV_UNKNOWN:
+        break;
+    }
+
+    /* default to empty string */
+    *platformString = 0;
+    return 0;
+}
+
+/*Assumes a $platform-#codepage.$CONVERTER_FILE_EXTENSION scheme and calls
+ *through createConverter*/
+U_CAPI UConverter*   U_EXPORT2
+ucnv_openCCSID (int32_t codepage,
+                UConverterPlatform platform,
+                UErrorCode * err)
+{
+    char myName[UCNV_MAX_CONVERTER_NAME_LENGTH];
+    int32_t myNameLen;
+
+    if (err == NULL || U_FAILURE (*err))
+        return NULL;
+
+    /* ucnv_copyPlatformString could return "ibm-" or "cp" */
+    myNameLen = ucnv_copyPlatformString(myName, platform);
+    T_CString_integerToString(myName + myNameLen, codepage, 10);
+
+    return ucnv_createConverter(NULL, myName, err);
+}
+
+/* Creating a temporary stack-based object that can be used in one thread, 
+and created from a converter that is shared across threads.
+*/
+
+U_CAPI UConverter* U_EXPORT2
+ucnv_safeClone(const UConverter* cnv, void *stackBuffer, int32_t *pBufferSize, UErrorCode *status)
+{
+    UConverter *localConverter, *allocatedConverter;
+    int32_t stackBufferSize;
+    int32_t bufferSizeNeeded;
+    UErrorCode cbErr;
+    UConverterToUnicodeArgs toUArgs = {
+        sizeof(UConverterToUnicodeArgs),
+            TRUE,
+            NULL,
+            NULL,
+            NULL,
+            NULL,
+            NULL,
+            NULL
+    };
+    UConverterFromUnicodeArgs fromUArgs = {
+        sizeof(UConverterFromUnicodeArgs),
+            TRUE,
+            NULL,
+            NULL,
+            NULL,
+            NULL,
+            NULL,
+            NULL
+    };
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_CLONE);
+
+    if (status == NULL || U_FAILURE(*status)){
+        UTRACE_EXIT_STATUS(status? *status: U_ILLEGAL_ARGUMENT_ERROR);
+        return NULL;
+    }
+
+    if (cnv == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        UTRACE_EXIT_STATUS(*status);
+        return NULL;
+    }
+
+    UTRACE_DATA3(UTRACE_OPEN_CLOSE, "clone converter %s at %p into stackBuffer %p",
+                                    ucnv_getName(cnv, status), cnv, stackBuffer);
+
+    if (cnv->sharedData->impl->safeClone != NULL) {
+        /* call the custom safeClone function for sizing */
+        bufferSizeNeeded = 0;
+        cnv->sharedData->impl->safeClone(cnv, NULL, &bufferSizeNeeded, status);
+        if (U_FAILURE(*status)) {
+            UTRACE_EXIT_STATUS(*status);
+            return NULL;
+        }
+    }
+    else
+    {
+        /* inherent sizing */
+        bufferSizeNeeded = sizeof(UConverter);
+    }
+
+    if (pBufferSize == NULL) {
+        stackBufferSize = 1;
+        pBufferSize = &stackBufferSize;
+    } else {
+        stackBufferSize = *pBufferSize;
+        if (stackBufferSize <= 0){ /* 'preflighting' request - set needed size into *pBufferSize */
+            *pBufferSize = bufferSizeNeeded;
+            UTRACE_EXIT_VALUE(bufferSizeNeeded);
+            return NULL;
+        }
+    }
+
+    /* Adjust (if necessary) the stackBuffer pointer to be aligned correctly for a UConverter.
+     * TODO(Jira ICU-20736) Redo this using std::align() once g++4.9 compatibility is no longer needed.
+     */
+    if (stackBuffer) {
+        uintptr_t p = reinterpret_cast<uintptr_t>(stackBuffer);
+        uintptr_t aligned_p = (p + alignof(UConverter) - 1) & ~(alignof(UConverter) - 1);
+        ptrdiff_t pointerAdjustment = aligned_p - p;
+        if (bufferSizeNeeded + pointerAdjustment <= stackBufferSize) {
+            stackBuffer = reinterpret_cast<void *>(aligned_p);
+            stackBufferSize -= static_cast<int32_t>(pointerAdjustment);
+        } else {
+            /* prevent using the stack buffer but keep the size > 0 so that we do not just preflight */
+            stackBufferSize = 1;
+        }
+    }
+
+    /* Now, see if we must allocate any memory */
+    if (stackBufferSize < bufferSizeNeeded || stackBuffer == NULL)
+    {
+        /* allocate one here...*/
+        localConverter = allocatedConverter = (UConverter *) uprv_malloc (bufferSizeNeeded);
+
+        if(localConverter == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            UTRACE_EXIT_STATUS(*status);
+            return NULL;
+        }
+        *status = U_SAFECLONE_ALLOCATED_WARNING;
+
+        /* record the fact that memory was allocated */
+        *pBufferSize = bufferSizeNeeded;
+    } else {
+        /* just use the stack buffer */
+        localConverter = (UConverter*) stackBuffer;
+        allocatedConverter = NULL;
+    }
+
+    uprv_memset(localConverter, 0, bufferSizeNeeded);
+
+    /* Copy initial state */
+    uprv_memcpy(localConverter, cnv, sizeof(UConverter));
+    localConverter->isCopyLocal = localConverter->isExtraLocal = FALSE;
+
+    /* copy the substitution string */
+    if (cnv->subChars == (uint8_t *)cnv->subUChars) {
+        localConverter->subChars = (uint8_t *)localConverter->subUChars;
+    } else {
+        localConverter->subChars = (uint8_t *)uprv_malloc(UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
+        if (localConverter->subChars == NULL) {
+            uprv_free(allocatedConverter);
+            UTRACE_EXIT_STATUS(*status);
+            return NULL;
+        }
+        uprv_memcpy(localConverter->subChars, cnv->subChars, UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
+    }
+
+    /* now either call the safeclone fcn or not */
+    if (cnv->sharedData->impl->safeClone != NULL) {
+        /* call the custom safeClone function */
+        localConverter = cnv->sharedData->impl->safeClone(cnv, localConverter, pBufferSize, status);
+    }
+
+    if(localConverter==NULL || U_FAILURE(*status)) {
+        if (allocatedConverter != NULL && allocatedConverter->subChars != (uint8_t *)allocatedConverter->subUChars) {
+            uprv_free(allocatedConverter->subChars);
+        }
+        uprv_free(allocatedConverter);
+        UTRACE_EXIT_STATUS(*status);
+        return NULL;
+    }
+
+    /* increment refcount of shared data if needed */
+    if (cnv->sharedData->isReferenceCounted) {
+        ucnv_incrementRefCount(cnv->sharedData);
+    }
+
+    if(localConverter == (UConverter*)stackBuffer) {
+        /* we're using user provided data - set to not destroy */
+        localConverter->isCopyLocal = TRUE;
+    }
+
+    /* allow callback functions to handle any memory allocation */
+    toUArgs.converter = fromUArgs.converter = localConverter;
+    cbErr = U_ZERO_ERROR;
+    cnv->fromCharErrorBehaviour(cnv->toUContext, &toUArgs, NULL, 0, UCNV_CLONE, &cbErr);
+    cbErr = U_ZERO_ERROR;
+    cnv->fromUCharErrorBehaviour(cnv->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_CLONE, &cbErr);
+
+    UTRACE_EXIT_PTR_STATUS(localConverter, *status);
+    return localConverter;
+}
+
+
+
+/*Decreases the reference counter in the shared immutable section of the object
+ *and frees the mutable part*/
+
+U_CAPI void  U_EXPORT2
+ucnv_close (UConverter * converter)
+{
+    UErrorCode errorCode = U_ZERO_ERROR;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_CLOSE);
+
+    if (converter == NULL)
+    {
+        UTRACE_EXIT();
+        return;
+    }
+
+    UTRACE_DATA3(UTRACE_OPEN_CLOSE, "close converter %s at %p, isCopyLocal=%b",
+        ucnv_getName(converter, &errorCode), converter, converter->isCopyLocal);
+
+    /* In order to speed up the close, only call the callbacks when they have been changed.
+    This performance check will only work when the callbacks are set within a shared library
+    or from user code that statically links this code. */
+    /* first, notify the callback functions that the converter is closed */
+    if (converter->fromCharErrorBehaviour != UCNV_TO_U_DEFAULT_CALLBACK) {
+        UConverterToUnicodeArgs toUArgs = {
+            sizeof(UConverterToUnicodeArgs),
+                TRUE,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL
+        };
+
+        toUArgs.converter = converter;
+        errorCode = U_ZERO_ERROR;
+        converter->fromCharErrorBehaviour(converter->toUContext, &toUArgs, NULL, 0, UCNV_CLOSE, &errorCode);
+    }
+    if (converter->fromUCharErrorBehaviour != UCNV_FROM_U_DEFAULT_CALLBACK) {
+        UConverterFromUnicodeArgs fromUArgs = {
+            sizeof(UConverterFromUnicodeArgs),
+                TRUE,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL
+        };
+        fromUArgs.converter = converter;
+        errorCode = U_ZERO_ERROR;
+        converter->fromUCharErrorBehaviour(converter->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_CLOSE, &errorCode);
+    }
+
+    if (converter->sharedData->impl->close != NULL) {
+        converter->sharedData->impl->close(converter);
+    }
+
+    if (converter->subChars != (uint8_t *)converter->subUChars) {
+        uprv_free(converter->subChars);
+    }
+
+    if (converter->sharedData->isReferenceCounted) {
+        ucnv_unloadSharedDataIfReady(converter->sharedData);
+    }
+
+    if(!converter->isCopyLocal){
+        uprv_free(converter);
+    }
+
+    UTRACE_EXIT();
+}
+
+/*returns a single Name from the list, will return NULL if out of bounds
+ */
+U_CAPI const char*   U_EXPORT2
+ucnv_getAvailableName (int32_t n)
+{
+    if (0 <= n && n <= 0xffff) {
+        UErrorCode err = U_ZERO_ERROR;
+        const char *name = ucnv_bld_getAvailableConverter((uint16_t)n, &err);
+        if (U_SUCCESS(err)) {
+            return name;
+        }
+    }
+    return NULL;
+}
+
+U_CAPI int32_t   U_EXPORT2
+ucnv_countAvailable ()
+{
+    UErrorCode err = U_ZERO_ERROR;
+    return ucnv_bld_countAvailableConverters(&err);
+}
+
+U_CAPI void    U_EXPORT2
+ucnv_getSubstChars (const UConverter * converter,
+                    char *mySubChar,
+                    int8_t * len,
+                    UErrorCode * err)
+{
+    if (U_FAILURE (*err))
+        return;
+
+    if (converter->subCharLen <= 0) {
+        /* Unicode string or empty string from ucnv_setSubstString(). */
+        *len = 0;
+        return;
+    }
+
+    if (*len < converter->subCharLen) /*not enough space in subChars */
+    {
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+
+    uprv_memcpy (mySubChar, converter->subChars, converter->subCharLen);   /*fills in the subchars */
+    *len = converter->subCharLen; /*store # of bytes copied to buffer */
+}
+
+U_CAPI void    U_EXPORT2
+ucnv_setSubstChars (UConverter * converter,
+                    const char *mySubChar,
+                    int8_t len,
+                    UErrorCode * err)
+{
+    if (U_FAILURE (*err))
+        return;
+    
+    /*Makes sure that the subChar is within the codepages char length boundaries */
+    if ((len > converter->sharedData->staticData->maxBytesPerChar)
+     || (len < converter->sharedData->staticData->minBytesPerChar))
+    {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    
+    uprv_memcpy (converter->subChars, mySubChar, len); /*copies the subchars */
+    converter->subCharLen = len;  /*sets the new len */
+
+    /*
+    * There is currently (2001Feb) no separate API to set/get subChar1.
+    * In order to always have subChar written after it is explicitly set,
+    * we set subChar1 to 0.
+    */
+    converter->subChar1 = 0;
+    
+    return;
+}
+
+U_CAPI void U_EXPORT2
+ucnv_setSubstString(UConverter *cnv,
+                    const UChar *s,
+                    int32_t length,
+                    UErrorCode *err) {
+    alignas(UConverter) char cloneBuffer[U_CNV_SAFECLONE_BUFFERSIZE];
+    char chars[UCNV_ERROR_BUFFER_LENGTH];
+
+    UConverter *clone;
+    uint8_t *subChars;
+    int32_t cloneSize, length8;
+
+    /* Let the following functions check all arguments. */
+    cloneSize = sizeof(cloneBuffer);
+    clone = ucnv_safeClone(cnv, cloneBuffer, &cloneSize, err);
+    ucnv_setFromUCallBack(clone, UCNV_FROM_U_CALLBACK_STOP, NULL, NULL, NULL, err);
+    length8 = ucnv_fromUChars(clone, chars, (int32_t)sizeof(chars), s, length, err);
+    ucnv_close(clone);
+    if (U_FAILURE(*err)) {
+        return;
+    }
+
+    if (cnv->sharedData->impl->writeSub == NULL
+#if !UCONFIG_NO_LEGACY_CONVERSION
+        || (cnv->sharedData->staticData->conversionType == UCNV_MBCS &&
+         ucnv_MBCSGetType(cnv) != UCNV_EBCDIC_STATEFUL)
+#endif
+    ) {
+        /* The converter is not stateful. Store the charset bytes as a fixed string. */
+        subChars = (uint8_t *)chars;
+    } else {
+        /*
+         * The converter has a non-default writeSub() function, indicating
+         * that it is stateful.
+         * Store the Unicode string for on-the-fly conversion for correct
+         * state handling.
+         */
+        if (length > UCNV_ERROR_BUFFER_LENGTH) {
+            /*
+             * Should not occur. The converter should output at least one byte
+             * per UChar, which means that ucnv_fromUChars() should catch all
+             * overflows.
+             */
+            *err = U_BUFFER_OVERFLOW_ERROR;
+            return;
+        }
+        subChars = (uint8_t *)s;
+        if (length < 0) {
+            length = u_strlen(s);
+        }
+        length8 = length * U_SIZEOF_UCHAR;
+    }
+
+    /*
+     * For storing the substitution string, select either the small buffer inside
+     * UConverter or allocate a subChars buffer.
+     */
+    if (length8 > UCNV_MAX_SUBCHAR_LEN) {
+        /* Use a separate buffer for the string. Outside UConverter to not make it too large. */
+        if (cnv->subChars == (uint8_t *)cnv->subUChars) {
+            /* Allocate a new buffer for the string. */
+            cnv->subChars = (uint8_t *)uprv_malloc(UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
+            if (cnv->subChars == NULL) {
+                cnv->subChars = (uint8_t *)cnv->subUChars;
+                *err = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            uprv_memset(cnv->subChars, 0, UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
+        }
+    }
+
+    /* Copy the substitution string into the UConverter or its subChars buffer. */
+    if (length8 == 0) {
+        cnv->subCharLen = 0;
+    } else {
+        uprv_memcpy(cnv->subChars, subChars, length8);
+        if (subChars == (uint8_t *)chars) {
+            cnv->subCharLen = (int8_t)length8;
+        } else /* subChars == s */ {
+            cnv->subCharLen = (int8_t)-length;
+        }
+    }
+
+    /* See comment in ucnv_setSubstChars(). */
+    cnv->subChar1 = 0;
+}
+
+/*resets the internal states of a converter
+ *goal : have the same behaviour than a freshly created converter
+ */
+static void _reset(UConverter *converter, UConverterResetChoice choice,
+                   UBool callCallback) {
+    if(converter == NULL) {
+        return;
+    }
+
+    if(callCallback) {
+        /* first, notify the callback functions that the converter is reset */
+        UErrorCode errorCode;
+
+        if(choice<=UCNV_RESET_TO_UNICODE && converter->fromCharErrorBehaviour != UCNV_TO_U_DEFAULT_CALLBACK) {
+            UConverterToUnicodeArgs toUArgs = {
+                sizeof(UConverterToUnicodeArgs),
+                TRUE,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL
+            };
+            toUArgs.converter = converter;
+            errorCode = U_ZERO_ERROR;
+            converter->fromCharErrorBehaviour(converter->toUContext, &toUArgs, NULL, 0, UCNV_RESET, &errorCode);
+        }
+        if(choice!=UCNV_RESET_TO_UNICODE && converter->fromUCharErrorBehaviour != UCNV_FROM_U_DEFAULT_CALLBACK) {
+            UConverterFromUnicodeArgs fromUArgs = {
+                sizeof(UConverterFromUnicodeArgs),
+                TRUE,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL,
+                NULL
+            };
+            fromUArgs.converter = converter;
+            errorCode = U_ZERO_ERROR;
+            converter->fromUCharErrorBehaviour(converter->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_RESET, &errorCode);
+        }
+    }
+
+    /* now reset the converter itself */
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        converter->toUnicodeStatus = converter->sharedData->toUnicodeStatus;
+        converter->mode = 0;
+        converter->toULength = 0;
+        converter->invalidCharLength = converter->UCharErrorBufferLength = 0;
+        converter->preToULength = 0;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        converter->fromUnicodeStatus = 0;
+        converter->fromUChar32 = 0;
+        converter->invalidUCharLength = converter->charErrorBufferLength = 0;
+        converter->preFromUFirstCP = U_SENTINEL;
+        converter->preFromULength = 0;
+    }
+
+    if (converter->sharedData->impl->reset != NULL) {
+        /* call the custom reset function */
+        converter->sharedData->impl->reset(converter, choice);
+    }
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_reset(UConverter *converter)
+{
+    _reset(converter, UCNV_RESET_BOTH, TRUE);
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_resetToUnicode(UConverter *converter)
+{
+    _reset(converter, UCNV_RESET_TO_UNICODE, TRUE);
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_resetFromUnicode(UConverter *converter)
+{
+    _reset(converter, UCNV_RESET_FROM_UNICODE, TRUE);
+}
+
+U_CAPI int8_t   U_EXPORT2
+ucnv_getMaxCharSize (const UConverter * converter)
+{
+    return converter->maxBytesPerUChar;
+}
+
+
+U_CAPI int8_t   U_EXPORT2
+ucnv_getMinCharSize (const UConverter * converter)
+{
+    return converter->sharedData->staticData->minBytesPerChar;
+}
+
+U_CAPI const char*   U_EXPORT2
+ucnv_getName (const UConverter * converter, UErrorCode * err)
+     
+{
+    if (U_FAILURE (*err))
+        return NULL;
+    if(converter->sharedData->impl->getName){
+        const char* temp= converter->sharedData->impl->getName(converter);
+        if(temp)
+            return temp;
+    }
+    return converter->sharedData->staticData->name;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_getCCSID(const UConverter * converter,
+              UErrorCode * err)
+{
+    int32_t ccsid;
+    if (U_FAILURE (*err))
+        return -1;
+
+    ccsid = converter->sharedData->staticData->codepage;
+    if (ccsid == 0) {
+        /* Rare case. This is for cases like gb18030,
+        which doesn't have an IBM canonical name, but does have an IBM alias. */
+        const char *standardName = ucnv_getStandardName(ucnv_getName(converter, err), "IBM", err);
+        if (U_SUCCESS(*err) && standardName) {
+            const char *ccsidStr = uprv_strchr(standardName, '-');
+            if (ccsidStr) {
+                ccsid = (int32_t)atol(ccsidStr+1);  /* +1 to skip '-' */
+            }
+        }
+    }
+    return ccsid;
+}
+
+
+U_CAPI UConverterPlatform   U_EXPORT2
+ucnv_getPlatform (const UConverter * converter,
+                                      UErrorCode * err)
+{
+    if (U_FAILURE (*err))
+        return UCNV_UNKNOWN;
+
+    return (UConverterPlatform)converter->sharedData->staticData->platform;
+}
+
+U_CAPI void U_EXPORT2
+    ucnv_getToUCallBack (const UConverter * converter,
+                         UConverterToUCallback *action,
+                         const void **context)
+{
+    *action = converter->fromCharErrorBehaviour;
+    *context = converter->toUContext;
+}
+
+U_CAPI void U_EXPORT2
+    ucnv_getFromUCallBack (const UConverter * converter,
+                           UConverterFromUCallback *action,
+                           const void **context)
+{
+    *action = converter->fromUCharErrorBehaviour;
+    *context = converter->fromUContext;
+}
+
+U_CAPI void    U_EXPORT2
+ucnv_setToUCallBack (UConverter * converter,
+                            UConverterToUCallback newAction,
+                            const void* newContext,
+                            UConverterToUCallback *oldAction,
+                            const void** oldContext,
+                            UErrorCode * err)
+{
+    if (U_FAILURE (*err))
+        return;
+    if (oldAction) *oldAction = converter->fromCharErrorBehaviour;
+    converter->fromCharErrorBehaviour = newAction;
+    if (oldContext) *oldContext = converter->toUContext;
+    converter->toUContext = newContext;
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_setFromUCallBack (UConverter * converter,
+                            UConverterFromUCallback newAction,
+                            const void* newContext,
+                            UConverterFromUCallback *oldAction,
+                            const void** oldContext,
+                            UErrorCode * err)
+{
+    if (U_FAILURE (*err))
+        return;
+    if (oldAction) *oldAction = converter->fromUCharErrorBehaviour;
+    converter->fromUCharErrorBehaviour = newAction;
+    if (oldContext) *oldContext = converter->fromUContext;
+    converter->fromUContext = newContext;
+}
+
+static void
+_updateOffsets(int32_t *offsets, int32_t length,
+               int32_t sourceIndex, int32_t errorInputLength) {
+    int32_t *limit;
+    int32_t delta, offset;
+
+    if(sourceIndex>=0) {
+        /*
+         * adjust each offset by adding the previous sourceIndex
+         * minus the length of the input sequence that caused an
+         * error, if any
+         */
+        delta=sourceIndex-errorInputLength;
+    } else {
+        /*
+         * set each offset to -1 because this conversion function
+         * does not handle offsets
+         */
+        delta=-1;
+    }
+
+    limit=offsets+length;
+    if(delta==0) {
+        /* most common case, nothing to do */
+    } else if(delta>0) {
+        /* add the delta to each offset (but not if the offset is <0) */
+        while(offsets<limit) {
+            offset=*offsets;
+            if(offset>=0) {
+                *offsets=offset+delta;
+            }
+            ++offsets;
+        }
+    } else /* delta<0 */ {
+        /*
+         * set each offset to -1 because this conversion function
+         * does not handle offsets
+         * or the error input sequence started in a previous buffer
+         */
+        while(offsets<limit) {
+            *offsets++=-1;
+        }
+    }
+}
+
+/* ucnv_fromUnicode --------------------------------------------------------- */
+
+/*
+ * Implementation note for m:n conversions
+ *
+ * While collecting source units to find the longest match for m:n conversion,
+ * some source units may need to be stored for a partial match.
+ * When a second buffer does not yield a match on all of the previously stored
+ * source units, then they must be "replayed", i.e., fed back into the converter.
+ *
+ * The code relies on the fact that replaying will not nest -
+ * converting a replay buffer will not result in a replay.
+ * This is because a replay is necessary only after the _continuation_ of a
+ * partial match failed, but a replay buffer is converted as a whole.
+ * It may result in some of its units being stored again for a partial match,
+ * but there will not be a continuation _during_ the replay which could fail.
+ *
+ * It is conceivable that a callback function could call the converter
+ * recursively in a way that causes another replay to be stored, but that
+ * would be an error in the callback function.
+ * Such violations will cause assertion failures in a debug build,
+ * and wrong output, but they will not cause a crash.
+ */
+
+static void
+_fromUnicodeWithCallback(UConverterFromUnicodeArgs *pArgs, UErrorCode *err) {
+    UConverterFromUnicode fromUnicode;
+    UConverter *cnv;
+    const UChar *s;
+    char *t;
+    int32_t *offsets;
+    int32_t sourceIndex;
+    int32_t errorInputLength;
+    UBool converterSawEndOfInput, calledCallback;
+
+    /* variables for m:n conversion */
+    UChar replay[UCNV_EXT_MAX_UCHARS];
+    const UChar *realSource, *realSourceLimit;
+    int32_t realSourceIndex;
+    UBool realFlush;
+
+    cnv=pArgs->converter;
+    s=pArgs->source;
+    t=pArgs->target;
+    offsets=pArgs->offsets;
+
+    /* get the converter implementation function */
+    sourceIndex=0;
+    if(offsets==NULL) {
+        fromUnicode=cnv->sharedData->impl->fromUnicode;
+    } else {
+        fromUnicode=cnv->sharedData->impl->fromUnicodeWithOffsets;
+        if(fromUnicode==NULL) {
+            /* there is no WithOffsets implementation */
+            fromUnicode=cnv->sharedData->impl->fromUnicode;
+            /* we will write -1 for each offset */
+            sourceIndex=-1;
+        }
+    }
+
+    if(cnv->preFromULength>=0) {
+        /* normal mode */
+        realSource=NULL;
+
+        /* avoid compiler warnings - not otherwise necessary, and the values do not matter */
+        realSourceLimit=NULL;
+        realFlush=FALSE;
+        realSourceIndex=0;
+    } else {
+        /*
+         * Previous m:n conversion stored source units from a partial match
+         * and failed to consume all of them.
+         * We need to "replay" them from a temporary buffer and convert them first.
+         */
+        realSource=pArgs->source;
+        realSourceLimit=pArgs->sourceLimit;
+        realFlush=pArgs->flush;
+        realSourceIndex=sourceIndex;
+
+        uprv_memcpy(replay, cnv->preFromU, -cnv->preFromULength*U_SIZEOF_UCHAR);
+        pArgs->source=replay;
+        pArgs->sourceLimit=replay-cnv->preFromULength;
+        pArgs->flush=FALSE;
+        sourceIndex=-1;
+
+        cnv->preFromULength=0;
+    }
+
+    /*
+     * loop for conversion and error handling
+     *
+     * loop {
+     *   convert
+     *   loop {
+     *     update offsets
+     *     handle end of input
+     *     handle errors/call callback
+     *   }
+     * }
+     */
+    for(;;) {
+        if(U_SUCCESS(*err)) {
+            /* convert */
+            fromUnicode(pArgs, err);
+
+            /*
+             * set a flag for whether the converter
+             * successfully processed the end of the input
+             *
+             * need not check cnv->preFromULength==0 because a replay (<0) will cause
+             * s<sourceLimit before converterSawEndOfInput is checked
+             */
+            converterSawEndOfInput=
+                (UBool)(U_SUCCESS(*err) &&
+                        pArgs->flush && pArgs->source==pArgs->sourceLimit &&
+                        cnv->fromUChar32==0);
+        } else {
+            /* handle error from ucnv_convertEx() */
+            converterSawEndOfInput=FALSE;
+        }
+
+        /* no callback called yet for this iteration */
+        calledCallback=FALSE;
+
+        /* no sourceIndex adjustment for conversion, only for callback output */
+        errorInputLength=0;
+
+        /*
+         * loop for offsets and error handling
+         *
+         * iterates at most 3 times:
+         * 1. to clean up after the conversion function
+         * 2. after the callback
+         * 3. after the callback again if there was truncated input
+         */
+        for(;;) {
+            /* update offsets if we write any */
+            if(offsets!=NULL) {
+                int32_t length=(int32_t)(pArgs->target-t);
+                if(length>0) {
+                    _updateOffsets(offsets, length, sourceIndex, errorInputLength);
+
+                    /*
+                     * if a converter handles offsets and updates the offsets
+                     * pointer at the end, then pArgs->offset should not change
+                     * here;
+                     * however, some converters do not handle offsets at all
+                     * (sourceIndex<0) or may not update the offsets pointer
+                     */
+                    pArgs->offsets=offsets+=length;
+                }
+
+                if(sourceIndex>=0) {
+                    sourceIndex+=(int32_t)(pArgs->source-s);
+                }
+            }
+
+            if(cnv->preFromULength<0) {
+                /*
+                 * switch the source to new replay units (cannot occur while replaying)
+                 * after offset handling and before end-of-input and callback handling
+                 */
+                if(realSource==NULL) {
+                    realSource=pArgs->source;
+                    realSourceLimit=pArgs->sourceLimit;
+                    realFlush=pArgs->flush;
+                    realSourceIndex=sourceIndex;
+
+                    uprv_memcpy(replay, cnv->preFromU, -cnv->preFromULength*U_SIZEOF_UCHAR);
+                    pArgs->source=replay;
+                    pArgs->sourceLimit=replay-cnv->preFromULength;
+                    pArgs->flush=FALSE;
+                    if((sourceIndex+=cnv->preFromULength)<0) {
+                        sourceIndex=-1;
+                    }
+
+                    cnv->preFromULength=0;
+                } else {
+                    /* see implementation note before _fromUnicodeWithCallback() */
+                    U_ASSERT(realSource==NULL);
+                    *err=U_INTERNAL_PROGRAM_ERROR;
+                }
+            }
+
+            /* update pointers */
+            s=pArgs->source;
+            t=pArgs->target;
+
+            if(U_SUCCESS(*err)) {
+                if(s<pArgs->sourceLimit) {
+                    /*
+                     * continue with the conversion loop while there is still input left
+                     * (continue converting by breaking out of only the inner loop)
+                     */
+                    break;
+                } else if(realSource!=NULL) {
+                    /* switch back from replaying to the real source and continue */
+                    pArgs->source=realSource;
+                    pArgs->sourceLimit=realSourceLimit;
+                    pArgs->flush=realFlush;
+                    sourceIndex=realSourceIndex;
+
+                    realSource=NULL;
+                    break;
+                } else if(pArgs->flush && cnv->fromUChar32!=0) {
+                    /*
+                     * the entire input stream is consumed
+                     * and there is a partial, truncated input sequence left
+                     */
+
+                    /* inject an error and continue with callback handling */
+                    *err=U_TRUNCATED_CHAR_FOUND;
+                    calledCallback=FALSE; /* new error condition */
+                } else {
+                    /* input consumed */
+                    if(pArgs->flush) {
+                        /*
+                         * return to the conversion loop once more if the flush
+                         * flag is set and the conversion function has not
+                         * successfully processed the end of the input yet
+                         *
+                         * (continue converting by breaking out of only the inner loop)
+                         */
+                        if(!converterSawEndOfInput) {
+                            break;
+                        }
+
+                        /* reset the converter without calling the callback function */
+                        _reset(cnv, UCNV_RESET_FROM_UNICODE, FALSE);
+                    }
+
+                    /* done successfully */
+                    return;
+                }
+            }
+
+            /* U_FAILURE(*err) */
+            {
+                UErrorCode e;
+
+                if( calledCallback ||
+                    (e=*err)==U_BUFFER_OVERFLOW_ERROR ||
+                    (e!=U_INVALID_CHAR_FOUND &&
+                     e!=U_ILLEGAL_CHAR_FOUND &&
+                     e!=U_TRUNCATED_CHAR_FOUND)
+                ) {
+                    /*
+                     * the callback did not or cannot resolve the error:
+                     * set output pointers and return
+                     *
+                     * the check for buffer overflow is redundant but it is
+                     * a high-runner case and hopefully documents the intent
+                     * well
+                     *
+                     * if we were replaying, then the replay buffer must be
+                     * copied back into the UConverter
+                     * and the real arguments must be restored
+                     */
+                    if(realSource!=NULL) {
+                        int32_t length;
+
+                        U_ASSERT(cnv->preFromULength==0);
+
+                        length=(int32_t)(pArgs->sourceLimit-pArgs->source);
+                        if(length>0) {
+                            u_memcpy(cnv->preFromU, pArgs->source, length);
+                            cnv->preFromULength=(int8_t)-length;
+                        }
+
+                        pArgs->source=realSource;
+                        pArgs->sourceLimit=realSourceLimit;
+                        pArgs->flush=realFlush;
+                    }
+
+                    return;
+                }
+            }
+
+            /* callback handling */
+            {
+                UChar32 codePoint;
+
+                /* get and write the code point */
+                codePoint=cnv->fromUChar32;
+                errorInputLength=0;
+                U16_APPEND_UNSAFE(cnv->invalidUCharBuffer, errorInputLength, codePoint);
+                cnv->invalidUCharLength=(int8_t)errorInputLength;
+
+                /* set the converter state to deal with the next character */
+                cnv->fromUChar32=0;
+
+                /* call the callback function */
+                cnv->fromUCharErrorBehaviour(cnv->fromUContext, pArgs,
+                    cnv->invalidUCharBuffer, errorInputLength, codePoint,
+                    *err==U_INVALID_CHAR_FOUND ? UCNV_UNASSIGNED : UCNV_ILLEGAL,
+                    err);
+            }
+
+            /*
+             * loop back to the offset handling
+             *
+             * this flag will indicate after offset handling
+             * that a callback was called;
+             * if the callback did not resolve the error, then we return
+             */
+            calledCallback=TRUE;
+        }
+    }
+}
+
+/*
+ * Output the fromUnicode overflow buffer.
+ * Call this function if(cnv->charErrorBufferLength>0).
+ * @return TRUE if overflow
+ */
+static UBool
+ucnv_outputOverflowFromUnicode(UConverter *cnv,
+                               char **target, const char *targetLimit,
+                               int32_t **pOffsets,
+                               UErrorCode *err) {
+    int32_t *offsets;
+    char *overflow, *t;
+    int32_t i, length;
+
+    t=*target;
+    if(pOffsets!=NULL) {
+        offsets=*pOffsets;
+    } else {
+        offsets=NULL;
+    }
+
+    overflow=(char *)cnv->charErrorBuffer;
+    length=cnv->charErrorBufferLength;
+    i=0;
+    while(i<length) {
+        if(t==targetLimit) {
+            /* the overflow buffer contains too much, keep the rest */
+            int32_t j=0;
+
+            do {
+                overflow[j++]=overflow[i++];
+            } while(i<length);
+
+            cnv->charErrorBufferLength=(int8_t)j;
+            *target=t;
+            if(offsets!=NULL) {
+                *pOffsets=offsets;
+            }
+            *err=U_BUFFER_OVERFLOW_ERROR;
+            return TRUE;
+        }
+
+        /* copy the overflow contents to the target */
+        *t++=overflow[i++];
+        if(offsets!=NULL) {
+            *offsets++=-1; /* no source index available for old output */
+        }
+    }
+
+    /* the overflow buffer is completely copied to the target */
+    cnv->charErrorBufferLength=0;
+    *target=t;
+    if(offsets!=NULL) {
+        *pOffsets=offsets;
+    }
+    return FALSE;
+}
+
+U_CAPI void U_EXPORT2
+ucnv_fromUnicode(UConverter *cnv,
+                 char **target, const char *targetLimit,
+                 const UChar **source, const UChar *sourceLimit,
+                 int32_t *offsets,
+                 UBool flush,
+                 UErrorCode *err) {
+    UConverterFromUnicodeArgs args;
+    const UChar *s;
+    char *t;
+
+    /* check parameters */
+    if(err==NULL || U_FAILURE(*err)) {
+        return;
+    }
+
+    if(cnv==NULL || target==NULL || source==NULL) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    s=*source;
+    t=*target;
+
+    if ((const void *)U_MAX_PTR(sourceLimit) == (const void *)sourceLimit) {
+        /*
+        Prevent code from going into an infinite loop in case we do hit this
+        limit. The limit pointer is expected to be on a UChar * boundary.
+        This also prevents the next argument check from failing.
+        */
+        sourceLimit = (const UChar *)(((const char *)sourceLimit) - 1);
+    }
+
+    /*
+     * All these conditions should never happen.
+     *
+     * 1) Make sure that the limits are >= to the address source or target
+     *
+     * 2) Make sure that the buffer sizes do not exceed the number range for
+     * int32_t because some functions use the size (in units or bytes)
+     * rather than comparing pointers, and because offsets are int32_t values.
+     *
+     * size_t is guaranteed to be unsigned and large enough for the job.
+     *
+     * Return with an error instead of adjusting the limits because we would
+     * not be able to maintain the semantics that either the source must be
+     * consumed or the target filled (unless an error occurs).
+     * An adjustment would be targetLimit=t+0x7fffffff; for example.
+     *
+     * 3) Make sure that the user didn't incorrectly cast a UChar * pointer
+     * to a char * pointer and provide an incomplete UChar code unit.
+     */
+    if (sourceLimit<s || targetLimit<t ||
+        ((size_t)(sourceLimit-s)>(size_t)0x3fffffff && sourceLimit>s) ||
+        ((size_t)(targetLimit-t)>(size_t)0x7fffffff && targetLimit>t) ||
+        (((const char *)sourceLimit-(const char *)s) & 1) != 0)
+    {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    
+    /* output the target overflow buffer */
+    if( cnv->charErrorBufferLength>0 &&
+        ucnv_outputOverflowFromUnicode(cnv, target, targetLimit, &offsets, err)
+    ) {
+        /* U_BUFFER_OVERFLOW_ERROR */
+        return;
+    }
+    /* *target may have moved, therefore stop using t */
+
+    if(!flush && s==sourceLimit && cnv->preFromULength>=0) {
+        /* the overflow buffer is emptied and there is no new input: we are done */
+        return;
+    }
+
+    /*
+     * Do not simply return with a buffer overflow error if
+     * !flush && t==targetLimit
+     * because it is possible that the source will not generate any output.
+     * For example, the skip callback may be called;
+     * it does not output anything.
+     */
+
+    /* prepare the converter arguments */
+    args.converter=cnv;
+    args.flush=flush;
+    args.offsets=offsets;
+    args.source=s;
+    args.sourceLimit=sourceLimit;
+    args.target=*target;
+    args.targetLimit=targetLimit;
+    args.size=sizeof(args);
+
+    _fromUnicodeWithCallback(&args, err);
+
+    *source=args.source;
+    *target=args.target;
+}
+
+/* ucnv_toUnicode() --------------------------------------------------------- */
+
+static void
+_toUnicodeWithCallback(UConverterToUnicodeArgs *pArgs, UErrorCode *err) {
+    UConverterToUnicode toUnicode;
+    UConverter *cnv;
+    const char *s;
+    UChar *t;
+    int32_t *offsets;
+    int32_t sourceIndex;
+    int32_t errorInputLength;
+    UBool converterSawEndOfInput, calledCallback;
+
+    /* variables for m:n conversion */
+    char replay[UCNV_EXT_MAX_BYTES];
+    const char *realSource, *realSourceLimit;
+    int32_t realSourceIndex;
+    UBool realFlush;
+
+    cnv=pArgs->converter;
+    s=pArgs->source;
+    t=pArgs->target;
+    offsets=pArgs->offsets;
+
+    /* get the converter implementation function */
+    sourceIndex=0;
+    if(offsets==NULL) {
+        toUnicode=cnv->sharedData->impl->toUnicode;
+    } else {
+        toUnicode=cnv->sharedData->impl->toUnicodeWithOffsets;
+        if(toUnicode==NULL) {
+            /* there is no WithOffsets implementation */
+            toUnicode=cnv->sharedData->impl->toUnicode;
+            /* we will write -1 for each offset */
+            sourceIndex=-1;
+        }
+    }
+
+    if(cnv->preToULength>=0) {
+        /* normal mode */
+        realSource=NULL;
+
+        /* avoid compiler warnings - not otherwise necessary, and the values do not matter */
+        realSourceLimit=NULL;
+        realFlush=FALSE;
+        realSourceIndex=0;
+    } else {
+        /*
+         * Previous m:n conversion stored source units from a partial match
+         * and failed to consume all of them.
+         * We need to "replay" them from a temporary buffer and convert them first.
+         */
+        realSource=pArgs->source;
+        realSourceLimit=pArgs->sourceLimit;
+        realFlush=pArgs->flush;
+        realSourceIndex=sourceIndex;
+
+        uprv_memcpy(replay, cnv->preToU, -cnv->preToULength);
+        pArgs->source=replay;
+        pArgs->sourceLimit=replay-cnv->preToULength;
+        pArgs->flush=FALSE;
+        sourceIndex=-1;
+
+        cnv->preToULength=0;
+    }
+
+    /*
+     * loop for conversion and error handling
+     *
+     * loop {
+     *   convert
+     *   loop {
+     *     update offsets
+     *     handle end of input
+     *     handle errors/call callback
+     *   }
+     * }
+     */
+    for(;;) {
+        if(U_SUCCESS(*err)) {
+            /* convert */
+            toUnicode(pArgs, err);
+
+            /*
+             * set a flag for whether the converter
+             * successfully processed the end of the input
+             *
+             * need not check cnv->preToULength==0 because a replay (<0) will cause
+             * s<sourceLimit before converterSawEndOfInput is checked
+             */
+            converterSawEndOfInput=
+                (UBool)(U_SUCCESS(*err) &&
+                        pArgs->flush && pArgs->source==pArgs->sourceLimit &&
+                        cnv->toULength==0);
+        } else {
+            /* handle error from getNextUChar() or ucnv_convertEx() */
+            converterSawEndOfInput=FALSE;
+        }
+
+        /* no callback called yet for this iteration */
+        calledCallback=FALSE;
+
+        /* no sourceIndex adjustment for conversion, only for callback output */
+        errorInputLength=0;
+
+        /*
+         * loop for offsets and error handling
+         *
+         * iterates at most 3 times:
+         * 1. to clean up after the conversion function
+         * 2. after the callback
+         * 3. after the callback again if there was truncated input
+         */
+        for(;;) {
+            /* update offsets if we write any */
+            if(offsets!=NULL) {
+                int32_t length=(int32_t)(pArgs->target-t);
+                if(length>0) {
+                    _updateOffsets(offsets, length, sourceIndex, errorInputLength);
+
+                    /*
+                     * if a converter handles offsets and updates the offsets
+                     * pointer at the end, then pArgs->offset should not change
+                     * here;
+                     * however, some converters do not handle offsets at all
+                     * (sourceIndex<0) or may not update the offsets pointer
+                     */
+                    pArgs->offsets=offsets+=length;
+                }
+
+                if(sourceIndex>=0) {
+                    sourceIndex+=(int32_t)(pArgs->source-s);
+                }
+            }
+
+            if(cnv->preToULength<0) {
+                /*
+                 * switch the source to new replay units (cannot occur while replaying)
+                 * after offset handling and before end-of-input and callback handling
+                 */
+                if(realSource==NULL) {
+                    realSource=pArgs->source;
+                    realSourceLimit=pArgs->sourceLimit;
+                    realFlush=pArgs->flush;
+                    realSourceIndex=sourceIndex;
+
+                    uprv_memcpy(replay, cnv->preToU, -cnv->preToULength);
+                    pArgs->source=replay;
+                    pArgs->sourceLimit=replay-cnv->preToULength;
+                    pArgs->flush=FALSE;
+                    if((sourceIndex+=cnv->preToULength)<0) {
+                        sourceIndex=-1;
+                    }
+
+                    cnv->preToULength=0;
+                } else {
+                    /* see implementation note before _fromUnicodeWithCallback() */
+                    U_ASSERT(realSource==NULL);
+                    *err=U_INTERNAL_PROGRAM_ERROR;
+                }
+            }
+
+            /* update pointers */
+            s=pArgs->source;
+            t=pArgs->target;
+
+            if(U_SUCCESS(*err)) {
+                if(s<pArgs->sourceLimit) {
+                    /*
+                     * continue with the conversion loop while there is still input left
+                     * (continue converting by breaking out of only the inner loop)
+                     */
+                    break;
+                } else if(realSource!=NULL) {
+                    /* switch back from replaying to the real source and continue */
+                    pArgs->source=realSource;
+                    pArgs->sourceLimit=realSourceLimit;
+                    pArgs->flush=realFlush;
+                    sourceIndex=realSourceIndex;
+
+                    realSource=NULL;
+                    break;
+                } else if(pArgs->flush && cnv->toULength>0) {
+                    /*
+                     * the entire input stream is consumed
+                     * and there is a partial, truncated input sequence left
+                     */
+
+                    /* inject an error and continue with callback handling */
+                    *err=U_TRUNCATED_CHAR_FOUND;
+                    calledCallback=FALSE; /* new error condition */
+                } else {
+                    /* input consumed */
+                    if(pArgs->flush) {
+                        /*
+                         * return to the conversion loop once more if the flush
+                         * flag is set and the conversion function has not
+                         * successfully processed the end of the input yet
+                         *
+                         * (continue converting by breaking out of only the inner loop)
+                         */
+                        if(!converterSawEndOfInput) {
+                            break;
+                        }
+
+                        /* reset the converter without calling the callback function */
+                        _reset(cnv, UCNV_RESET_TO_UNICODE, FALSE);
+                    }
+
+                    /* done successfully */
+                    return;
+                }
+            }
+
+            /* U_FAILURE(*err) */
+            {
+                UErrorCode e;
+
+                if( calledCallback ||
+                    (e=*err)==U_BUFFER_OVERFLOW_ERROR ||
+                    (e!=U_INVALID_CHAR_FOUND &&
+                     e!=U_ILLEGAL_CHAR_FOUND &&
+                     e!=U_TRUNCATED_CHAR_FOUND &&
+                     e!=U_ILLEGAL_ESCAPE_SEQUENCE &&
+                     e!=U_UNSUPPORTED_ESCAPE_SEQUENCE)
+                ) {
+                    /*
+                     * the callback did not or cannot resolve the error:
+                     * set output pointers and return
+                     *
+                     * the check for buffer overflow is redundant but it is
+                     * a high-runner case and hopefully documents the intent
+                     * well
+                     *
+                     * if we were replaying, then the replay buffer must be
+                     * copied back into the UConverter
+                     * and the real arguments must be restored
+                     */
+                    if(realSource!=NULL) {
+                        int32_t length;
+
+                        U_ASSERT(cnv->preToULength==0);
+
+                        length=(int32_t)(pArgs->sourceLimit-pArgs->source);
+                        if(length>0) {
+                            uprv_memcpy(cnv->preToU, pArgs->source, length);
+                            cnv->preToULength=(int8_t)-length;
+                        }
+
+                        pArgs->source=realSource;
+                        pArgs->sourceLimit=realSourceLimit;
+                        pArgs->flush=realFlush;
+                    }
+
+                    return;
+                }
+            }
+
+            /* copy toUBytes[] to invalidCharBuffer[] */
+            errorInputLength=cnv->invalidCharLength=cnv->toULength;
+            if(errorInputLength>0) {
+                uprv_memcpy(cnv->invalidCharBuffer, cnv->toUBytes, errorInputLength);
+            }
+
+            /* set the converter state to deal with the next character */
+            cnv->toULength=0;
+
+            /* call the callback function */
+            if(cnv->toUCallbackReason==UCNV_ILLEGAL && *err==U_INVALID_CHAR_FOUND) {
+                cnv->toUCallbackReason = UCNV_UNASSIGNED;
+            }
+            cnv->fromCharErrorBehaviour(cnv->toUContext, pArgs,
+                cnv->invalidCharBuffer, errorInputLength,
+                cnv->toUCallbackReason,
+                err);
+            cnv->toUCallbackReason = UCNV_ILLEGAL; /* reset to default value */
+
+            /*
+             * loop back to the offset handling
+             *
+             * this flag will indicate after offset handling
+             * that a callback was called;
+             * if the callback did not resolve the error, then we return
+             */
+            calledCallback=TRUE;
+        }
+    }
+}
+
+/*
+ * Output the toUnicode overflow buffer.
+ * Call this function if(cnv->UCharErrorBufferLength>0).
+ * @return TRUE if overflow
+ */
+static UBool
+ucnv_outputOverflowToUnicode(UConverter *cnv,
+                             UChar **target, const UChar *targetLimit,
+                             int32_t **pOffsets,
+                             UErrorCode *err) {
+    int32_t *offsets;
+    UChar *overflow, *t;
+    int32_t i, length;
+
+    t=*target;
+    if(pOffsets!=NULL) {
+        offsets=*pOffsets;
+    } else {
+        offsets=NULL;
+    }
+
+    overflow=cnv->UCharErrorBuffer;
+    length=cnv->UCharErrorBufferLength;
+    i=0;
+    while(i<length) {
+        if(t==targetLimit) {
+            /* the overflow buffer contains too much, keep the rest */
+            int32_t j=0;
+
+            do {
+                overflow[j++]=overflow[i++];
+            } while(i<length);
+
+            cnv->UCharErrorBufferLength=(int8_t)j;
+            *target=t;
+            if(offsets!=NULL) {
+                *pOffsets=offsets;
+            }
+            *err=U_BUFFER_OVERFLOW_ERROR;
+            return TRUE;
+        }
+
+        /* copy the overflow contents to the target */
+        *t++=overflow[i++];
+        if(offsets!=NULL) {
+            *offsets++=-1; /* no source index available for old output */
+        }
+    }
+
+    /* the overflow buffer is completely copied to the target */
+    cnv->UCharErrorBufferLength=0;
+    *target=t;
+    if(offsets!=NULL) {
+        *pOffsets=offsets;
+    }
+    return FALSE;
+}
+
+U_CAPI void U_EXPORT2
+ucnv_toUnicode(UConverter *cnv,
+               UChar **target, const UChar *targetLimit,
+               const char **source, const char *sourceLimit,
+               int32_t *offsets,
+               UBool flush,
+               UErrorCode *err) {
+    UConverterToUnicodeArgs args;
+    const char *s;
+    UChar *t;
+
+    /* check parameters */
+    if(err==NULL || U_FAILURE(*err)) {
+        return;
+    }
+
+    if(cnv==NULL || target==NULL || source==NULL) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    s=*source;
+    t=*target;
+
+    if ((const void *)U_MAX_PTR(targetLimit) == (const void *)targetLimit) {
+        /*
+        Prevent code from going into an infinite loop in case we do hit this
+        limit. The limit pointer is expected to be on a UChar * boundary.
+        This also prevents the next argument check from failing.
+        */
+        targetLimit = (const UChar *)(((const char *)targetLimit) - 1);
+    }
+
+    /*
+     * All these conditions should never happen.
+     *
+     * 1) Make sure that the limits are >= to the address source or target
+     *
+     * 2) Make sure that the buffer sizes do not exceed the number range for
+     * int32_t because some functions use the size (in units or bytes)
+     * rather than comparing pointers, and because offsets are int32_t values.
+     *
+     * size_t is guaranteed to be unsigned and large enough for the job.
+     *
+     * Return with an error instead of adjusting the limits because we would
+     * not be able to maintain the semantics that either the source must be
+     * consumed or the target filled (unless an error occurs).
+     * An adjustment would be sourceLimit=t+0x7fffffff; for example.
+     *
+     * 3) Make sure that the user didn't incorrectly cast a UChar * pointer
+     * to a char * pointer and provide an incomplete UChar code unit.
+     */
+    if (sourceLimit<s || targetLimit<t ||
+        ((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s) ||
+        ((size_t)(targetLimit-t)>(size_t)0x3fffffff && targetLimit>t) ||
+        (((const char *)targetLimit-(const char *)t) & 1) != 0
+    ) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    
+    /* output the target overflow buffer */
+    if( cnv->UCharErrorBufferLength>0 &&
+        ucnv_outputOverflowToUnicode(cnv, target, targetLimit, &offsets, err)
+    ) {
+        /* U_BUFFER_OVERFLOW_ERROR */
+        return;
+    }
+    /* *target may have moved, therefore stop using t */
+
+    if(!flush && s==sourceLimit && cnv->preToULength>=0) {
+        /* the overflow buffer is emptied and there is no new input: we are done */
+        return;
+    }
+
+    /*
+     * Do not simply return with a buffer overflow error if
+     * !flush && t==targetLimit
+     * because it is possible that the source will not generate any output.
+     * For example, the skip callback may be called;
+     * it does not output anything.
+     */
+
+    /* prepare the converter arguments */
+    args.converter=cnv;
+    args.flush=flush;
+    args.offsets=offsets;
+    args.source=s;
+    args.sourceLimit=sourceLimit;
+    args.target=*target;
+    args.targetLimit=targetLimit;
+    args.size=sizeof(args);
+
+    _toUnicodeWithCallback(&args, err);
+
+    *source=args.source;
+    *target=args.target;
+}
+
+/* ucnv_to/fromUChars() ----------------------------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+ucnv_fromUChars(UConverter *cnv,
+                char *dest, int32_t destCapacity,
+                const UChar *src, int32_t srcLength,
+                UErrorCode *pErrorCode) {
+    const UChar *srcLimit;
+    char *originalDest, *destLimit;
+    int32_t destLength;
+
+    /* check arguments */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( cnv==NULL ||
+        destCapacity<0 || (destCapacity>0 && dest==NULL) ||
+        srcLength<-1 || (srcLength!=0 && src==NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* initialize */
+    ucnv_resetFromUnicode(cnv);
+    originalDest=dest;
+    if(srcLength==-1) {
+        srcLength=u_strlen(src);
+    }
+    if(srcLength>0) {
+        srcLimit=src+srcLength;
+        destCapacity=pinCapacity(dest, destCapacity);
+        destLimit=dest+destCapacity;
+
+        /* perform the conversion */
+        ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
+        destLength=(int32_t)(dest-originalDest);
+
+        /* if an overflow occurs, then get the preflighting length */
+        if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
+            char buffer[1024];
+
+            destLimit=buffer+sizeof(buffer);
+            do {
+                dest=buffer;
+                *pErrorCode=U_ZERO_ERROR;
+                ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
+                destLength+=(int32_t)(dest-buffer);
+            } while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
+        }
+    } else {
+        destLength=0;
+    }
+
+    return u_terminateChars(originalDest, destCapacity, destLength, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_toUChars(UConverter *cnv,
+              UChar *dest, int32_t destCapacity,
+              const char *src, int32_t srcLength,
+              UErrorCode *pErrorCode) {
+    const char *srcLimit;
+    UChar *originalDest, *destLimit;
+    int32_t destLength;
+
+    /* check arguments */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( cnv==NULL ||
+        destCapacity<0 || (destCapacity>0 && dest==NULL) ||
+        srcLength<-1 || (srcLength!=0 && src==NULL))
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* initialize */
+    ucnv_resetToUnicode(cnv);
+    originalDest=dest;
+    if(srcLength==-1) {
+        srcLength=(int32_t)uprv_strlen(src);
+    }
+    if(srcLength>0) {
+        srcLimit=src+srcLength;
+        destCapacity=pinCapacity(dest, destCapacity);
+        destLimit=dest+destCapacity;
+
+        /* perform the conversion */
+        ucnv_toUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
+        destLength=(int32_t)(dest-originalDest);
+
+        /* if an overflow occurs, then get the preflighting length */
+        if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR)
+        {
+            UChar buffer[1024];
+
+            destLimit=buffer+UPRV_LENGTHOF(buffer);
+            do {
+                dest=buffer;
+                *pErrorCode=U_ZERO_ERROR;
+                ucnv_toUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
+                destLength+=(int32_t)(dest-buffer);
+            }
+            while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
+        }
+    } else {
+        destLength=0;
+    }
+
+    return u_terminateUChars(originalDest, destCapacity, destLength, pErrorCode);
+}
+
+/* ucnv_getNextUChar() ------------------------------------------------------ */
+
+U_CAPI UChar32 U_EXPORT2
+ucnv_getNextUChar(UConverter *cnv,
+                  const char **source, const char *sourceLimit,
+                  UErrorCode *err) {
+    UConverterToUnicodeArgs args;
+    UChar buffer[U16_MAX_LENGTH];
+    const char *s;
+    UChar32 c;
+    int32_t i, length;
+
+    /* check parameters */
+    if(err==NULL || U_FAILURE(*err)) {
+        return 0xffff;
+    }
+
+    if(cnv==NULL || source==NULL) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0xffff;
+    }
+
+    s=*source;
+    if(sourceLimit<s) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0xffff;
+    }
+
+    /*
+     * Make sure that the buffer sizes do not exceed the number range for
+     * int32_t because some functions use the size (in units or bytes)
+     * rather than comparing pointers, and because offsets are int32_t values.
+     *
+     * size_t is guaranteed to be unsigned and large enough for the job.
+     *
+     * Return with an error instead of adjusting the limits because we would
+     * not be able to maintain the semantics that either the source must be
+     * consumed or the target filled (unless an error occurs).
+     * An adjustment would be sourceLimit=t+0x7fffffff; for example.
+     */
+    if(((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s)) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0xffff;
+    }
+
+    c=U_SENTINEL;
+
+    /* flush the target overflow buffer */
+    if(cnv->UCharErrorBufferLength>0) {
+        UChar *overflow;
+
+        overflow=cnv->UCharErrorBuffer;
+        i=0;
+        length=cnv->UCharErrorBufferLength;
+        U16_NEXT(overflow, i, length, c);
+
+        /* move the remaining overflow contents up to the beginning */
+        if((cnv->UCharErrorBufferLength=(int8_t)(length-i))>0) {
+            uprv_memmove(cnv->UCharErrorBuffer, cnv->UCharErrorBuffer+i,
+                         cnv->UCharErrorBufferLength*U_SIZEOF_UCHAR);
+        }
+
+        if(!U16_IS_LEAD(c) || i<length) {
+            return c;
+        }
+        /*
+         * Continue if the overflow buffer contained only a lead surrogate,
+         * in case the converter outputs single surrogates from complete
+         * input sequences.
+         */
+    }
+
+    /*
+     * flush==TRUE is implied for ucnv_getNextUChar()
+     *
+     * do not simply return even if s==sourceLimit because the converter may
+     * not have seen flush==TRUE before
+     */
+
+    /* prepare the converter arguments */
+    args.converter=cnv;
+    args.flush=TRUE;
+    args.offsets=NULL;
+    args.source=s;
+    args.sourceLimit=sourceLimit;
+    args.target=buffer;
+    args.targetLimit=buffer+1;
+    args.size=sizeof(args);
+
+    if(c<0) {
+        /*
+         * call the native getNextUChar() implementation if we are
+         * at a character boundary (toULength==0)
+         *
+         * unlike with _toUnicode(), getNextUChar() implementations must set
+         * U_TRUNCATED_CHAR_FOUND for truncated input,
+         * in addition to setting toULength/toUBytes[]
+         */
+        if(cnv->toULength==0 && cnv->sharedData->impl->getNextUChar!=NULL) {
+            c=cnv->sharedData->impl->getNextUChar(&args, err);
+            *source=s=args.source;
+            if(*err==U_INDEX_OUTOFBOUNDS_ERROR) {
+                /* reset the converter without calling the callback function */
+                _reset(cnv, UCNV_RESET_TO_UNICODE, FALSE);
+                return 0xffff; /* no output */
+            } else if(U_SUCCESS(*err) && c>=0) {
+                return c;
+            /*
+             * else fall through to use _toUnicode() because
+             *   UCNV_GET_NEXT_UCHAR_USE_TO_U: the native function did not want to handle it after all
+             *   U_FAILURE: call _toUnicode() for callback handling (do not output c)
+             */
+            }
+        }
+
+        /* convert to one UChar in buffer[0], or handle getNextUChar() errors */
+        _toUnicodeWithCallback(&args, err);
+
+        if(*err==U_BUFFER_OVERFLOW_ERROR) {
+            *err=U_ZERO_ERROR;
+        }
+
+        i=0;
+        length=(int32_t)(args.target-buffer);
+    } else {
+        /* write the lead surrogate from the overflow buffer */
+        buffer[0]=(UChar)c;
+        args.target=buffer+1;
+        i=0;
+        length=1;
+    }
+
+    /* buffer contents starts at i and ends before length */
+
+    if(U_FAILURE(*err)) {
+        c=0xffff; /* no output */
+    } else if(length==0) {
+        /* no input or only state changes */
+        *err=U_INDEX_OUTOFBOUNDS_ERROR;
+        /* no need to reset explicitly because _toUnicodeWithCallback() did it */
+        c=0xffff; /* no output */
+    } else {
+        c=buffer[0];
+        i=1;
+        if(!U16_IS_LEAD(c)) {
+            /* consume c=buffer[0], done */
+        } else {
+            /* got a lead surrogate, see if a trail surrogate follows */
+            UChar c2;
+
+            if(cnv->UCharErrorBufferLength>0) {
+                /* got overflow output from the conversion */
+                if(U16_IS_TRAIL(c2=cnv->UCharErrorBuffer[0])) {
+                    /* got a trail surrogate, too */
+                    c=U16_GET_SUPPLEMENTARY(c, c2);
+
+                    /* move the remaining overflow contents up to the beginning */
+                    if((--cnv->UCharErrorBufferLength)>0) {
+                        uprv_memmove(cnv->UCharErrorBuffer, cnv->UCharErrorBuffer+1,
+                                     cnv->UCharErrorBufferLength*U_SIZEOF_UCHAR);
+                    }
+                } else {
+                    /* c is an unpaired lead surrogate, just return it */
+                }
+            } else if(args.source<sourceLimit) {
+                /* convert once more, to buffer[1] */
+                args.targetLimit=buffer+2;
+                _toUnicodeWithCallback(&args, err);
+                if(*err==U_BUFFER_OVERFLOW_ERROR) {
+                    *err=U_ZERO_ERROR;
+                }
+
+                length=(int32_t)(args.target-buffer);
+                if(U_SUCCESS(*err) && length==2 && U16_IS_TRAIL(c2=buffer[1])) {
+                    /* got a trail surrogate, too */
+                    c=U16_GET_SUPPLEMENTARY(c, c2);
+                    i=2;
+                }
+            }
+        }
+    }
+
+    /*
+     * move leftover output from buffer[i..length[
+     * into the beginning of the overflow buffer
+     */
+    if(i<length) {
+        /* move further overflow back */
+        int32_t delta=length-i;
+        if((length=cnv->UCharErrorBufferLength)>0) {
+            uprv_memmove(cnv->UCharErrorBuffer+delta, cnv->UCharErrorBuffer,
+                         length*U_SIZEOF_UCHAR);
+        }
+        cnv->UCharErrorBufferLength=(int8_t)(length+delta);
+
+        cnv->UCharErrorBuffer[0]=buffer[i++];
+        if(delta>1) {
+            cnv->UCharErrorBuffer[1]=buffer[i];
+        }
+    }
+
+    *source=args.source;
+    return c;
+}
+
+/* ucnv_convert() and siblings ---------------------------------------------- */
+
+U_CAPI void U_EXPORT2
+ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
+               char **target, const char *targetLimit,
+               const char **source, const char *sourceLimit,
+               UChar *pivotStart, UChar **pivotSource,
+               UChar **pivotTarget, const UChar *pivotLimit,
+               UBool reset, UBool flush,
+               UErrorCode *pErrorCode) {
+    UChar pivotBuffer[CHUNK_SIZE];
+    const UChar *myPivotSource;
+    UChar *myPivotTarget;
+    const char *s;
+    char *t;
+
+    UConverterToUnicodeArgs toUArgs;
+    UConverterFromUnicodeArgs fromUArgs;
+    UConverterConvert convert;
+
+    /* error checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    if( targetCnv==NULL || sourceCnv==NULL ||
+        source==NULL || *source==NULL ||
+        target==NULL || *target==NULL || targetLimit==NULL
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    s=*source;
+    t=*target;
+    if((sourceLimit!=NULL && sourceLimit<s) || targetLimit<t) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    /*
+     * Make sure that the buffer sizes do not exceed the number range for
+     * int32_t. See ucnv_toUnicode() for a more detailed comment.
+     */
+    if(
+        (sourceLimit!=NULL && ((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s)) ||
+        ((size_t)(targetLimit-t)>(size_t)0x7fffffff && targetLimit>t)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    
+    if(pivotStart==NULL) {
+        if(!flush) {
+            /* streaming conversion requires an explicit pivot buffer */
+            *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+
+        /* use the stack pivot buffer */
+        myPivotSource=myPivotTarget=pivotStart=pivotBuffer;
+        pivotSource=(UChar **)&myPivotSource;
+        pivotTarget=&myPivotTarget;
+        pivotLimit=pivotBuffer+CHUNK_SIZE;
+    } else if(  pivotStart>=pivotLimit ||
+                pivotSource==NULL || *pivotSource==NULL ||
+                pivotTarget==NULL || *pivotTarget==NULL ||
+                pivotLimit==NULL
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if(sourceLimit==NULL) {
+        /* get limit of single-byte-NUL-terminated source string */
+        sourceLimit=uprv_strchr(*source, 0);
+    }
+
+    if(reset) {
+        ucnv_resetToUnicode(sourceCnv);
+        ucnv_resetFromUnicode(targetCnv);
+        *pivotSource=*pivotTarget=pivotStart;
+    } else if(targetCnv->charErrorBufferLength>0) {
+        /* output the targetCnv overflow buffer */
+        if(ucnv_outputOverflowFromUnicode(targetCnv, target, targetLimit, NULL, pErrorCode)) {
+            /* U_BUFFER_OVERFLOW_ERROR */
+            return;
+        }
+        /* *target has moved, therefore stop using t */
+
+        if( !flush &&
+            targetCnv->preFromULength>=0 && *pivotSource==*pivotTarget &&
+            sourceCnv->UCharErrorBufferLength==0 && sourceCnv->preToULength>=0 && s==sourceLimit
+        ) {
+            /* the fromUnicode overflow buffer is emptied and there is no new input: we are done */
+            return;
+        }
+    }
+
+    /* Is direct-UTF-8 conversion available? */
+    if( sourceCnv->sharedData->staticData->conversionType==UCNV_UTF8 &&
+        targetCnv->sharedData->impl->fromUTF8!=NULL
+    ) {
+        convert=targetCnv->sharedData->impl->fromUTF8;
+    } else if( targetCnv->sharedData->staticData->conversionType==UCNV_UTF8 &&
+               sourceCnv->sharedData->impl->toUTF8!=NULL
+    ) {
+        convert=sourceCnv->sharedData->impl->toUTF8;
+    } else {
+        convert=NULL;
+    }
+
+    /*
+     * If direct-UTF-8 conversion is available, then we use a smaller
+     * pivot buffer for error handling and partial matches
+     * so that we quickly return to direct conversion.
+     *
+     * 32 is large enough for UCNV_EXT_MAX_UCHARS and UCNV_ERROR_BUFFER_LENGTH.
+     *
+     * We could reduce the pivot buffer size further, at the cost of
+     * buffer overflows from callbacks.
+     * The pivot buffer should not be smaller than the maximum number of
+     * fromUnicode extension table input UChars
+     * (for m:n conversion, see
+     * targetCnv->sharedData->mbcs.extIndexes[UCNV_EXT_COUNT_UCHARS])
+     * or 2 for surrogate pairs.
+     *
+     * Too small a buffer can cause thrashing between pivoting and direct
+     * conversion, with function call overhead outweighing the benefits
+     * of direct conversion.
+     */
+    if(convert!=NULL && (pivotLimit-pivotStart)>32) {
+        pivotLimit=pivotStart+32;
+    }
+
+    /* prepare the converter arguments */
+    fromUArgs.converter=targetCnv;
+    fromUArgs.flush=FALSE;
+    fromUArgs.offsets=NULL;
+    fromUArgs.target=*target;
+    fromUArgs.targetLimit=targetLimit;
+    fromUArgs.size=sizeof(fromUArgs);
+
+    toUArgs.converter=sourceCnv;
+    toUArgs.flush=flush;
+    toUArgs.offsets=NULL;
+    toUArgs.source=s;
+    toUArgs.sourceLimit=sourceLimit;
+    toUArgs.targetLimit=pivotLimit;
+    toUArgs.size=sizeof(toUArgs);
+
+    /*
+     * TODO: Consider separating this function into two functions,
+     * extracting exactly the conversion loop,
+     * for readability and to reduce the set of visible variables.
+     *
+     * Otherwise stop using s and t from here on.
+     */
+    s=t=NULL;
+
+    /*
+     * conversion loop
+     *
+     * The sequence of steps in the loop may appear backward,
+     * but the principle is simple:
+     * In the chain of
+     *   source - sourceCnv overflow - pivot - targetCnv overflow - target
+     * empty out later buffers before refilling them from earlier ones.
+     *
+     * The targetCnv overflow buffer is flushed out only once before the loop.
+     */
+    for(;;) {
+        /*
+         * if(pivot not empty or error or replay or flush fromUnicode) {
+         *   fromUnicode(pivot -> target);
+         * }
+         *
+         * For pivoting conversion; and for direct conversion for
+         * error callback handling and flushing the replay buffer.
+         */
+        if( *pivotSource<*pivotTarget ||
+            U_FAILURE(*pErrorCode) ||
+            targetCnv->preFromULength<0 ||
+            fromUArgs.flush
+        ) {
+            fromUArgs.source=*pivotSource;
+            fromUArgs.sourceLimit=*pivotTarget;
+            _fromUnicodeWithCallback(&fromUArgs, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                /* target overflow, or conversion error */
+                *pivotSource=(UChar *)fromUArgs.source;
+                break;
+            }
+
+            /*
+             * _fromUnicodeWithCallback() must have consumed the pivot contents
+             * (*pivotSource==*pivotTarget) since it returned with U_SUCCESS()
+             */
+        }
+
+        /* The pivot buffer is empty; reset it so we start at pivotStart. */
+        *pivotSource=*pivotTarget=pivotStart;
+
+        /*
+         * if(sourceCnv overflow buffer not empty) {
+         *     move(sourceCnv overflow buffer -> pivot);
+         *     continue;
+         * }
+         */
+        /* output the sourceCnv overflow buffer */
+        if(sourceCnv->UCharErrorBufferLength>0) {
+            if(ucnv_outputOverflowToUnicode(sourceCnv, pivotTarget, pivotLimit, NULL, pErrorCode)) {
+                /* U_BUFFER_OVERFLOW_ERROR */
+                *pErrorCode=U_ZERO_ERROR;
+            }
+            continue;
+        }
+
+        /*
+         * check for end of input and break if done
+         *
+         * Checking both flush and fromUArgs.flush ensures that the converters
+         * have been called with the flush flag set if the ucnv_convertEx()
+         * caller set it.
+         */
+        if( toUArgs.source==sourceLimit &&
+            sourceCnv->preToULength>=0 && sourceCnv->toULength==0 &&
+            (!flush || fromUArgs.flush)
+        ) {
+            /* done successfully */
+            break;
+        }
+
+        /*
+         * use direct conversion if available
+         * but not if continuing a partial match
+         * or flushing the toUnicode replay buffer
+         */
+        if(convert!=NULL && targetCnv->preFromUFirstCP<0 && sourceCnv->preToULength==0) {
+            if(*pErrorCode==U_USING_DEFAULT_WARNING) {
+                /* remove a warning that may be set by this function */
+                *pErrorCode=U_ZERO_ERROR;
+            }
+            convert(&fromUArgs, &toUArgs, pErrorCode);
+            if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
+                break;
+            } else if(U_FAILURE(*pErrorCode)) {
+                if(sourceCnv->toULength>0) {
+                    /*
+                     * Fall through to calling _toUnicodeWithCallback()
+                     * for callback handling.
+                     *
+                     * The pivot buffer will be reset with
+                     *   *pivotSource=*pivotTarget=pivotStart;
+                     * which indicates a toUnicode error to the caller
+                     * (*pivotSource==pivotStart shows no pivot UChars consumed).
+                     */
+                } else {
+                    /*
+                     * Indicate a fromUnicode error to the caller
+                     * (*pivotSource>pivotStart shows some pivot UChars consumed).
+                     */
+                    *pivotSource=*pivotTarget=pivotStart+1;
+                    /*
+                     * Loop around to calling _fromUnicodeWithCallbacks()
+                     * for callback handling.
+                     */
+                    continue;
+                }
+            } else if(*pErrorCode==U_USING_DEFAULT_WARNING) {
+                /*
+                 * No error, but the implementation requested to temporarily
+                 * fall back to pivoting.
+                 */
+                *pErrorCode=U_ZERO_ERROR;
+            /*
+             * The following else branches are almost identical to the end-of-input
+             * handling in _toUnicodeWithCallback().
+             * Avoid calling it just for the end of input.
+             */
+            } else if(flush && sourceCnv->toULength>0) { /* flush==toUArgs.flush */
+                /*
+                 * the entire input stream is consumed
+                 * and there is a partial, truncated input sequence left
+                 */
+
+                /* inject an error and continue with callback handling */
+                *pErrorCode=U_TRUNCATED_CHAR_FOUND;
+            } else {
+                /* input consumed */
+                if(flush) {
+                    /* reset the converters without calling the callback functions */
+                    _reset(sourceCnv, UCNV_RESET_TO_UNICODE, FALSE);
+                    _reset(targetCnv, UCNV_RESET_FROM_UNICODE, FALSE);
+                }
+
+                /* done successfully */
+                break;
+            }
+        }
+        
+        /*
+         * toUnicode(source -> pivot);
+         *
+         * For pivoting conversion; and for direct conversion for
+         * error callback handling, continuing partial matches
+         * and flushing the replay buffer.
+         *
+         * The pivot buffer is empty and reset.
+         */
+        toUArgs.target=pivotStart; /* ==*pivotTarget */
+        /* toUArgs.targetLimit=pivotLimit; already set before the loop */
+        _toUnicodeWithCallback(&toUArgs, pErrorCode);
+        *pivotTarget=toUArgs.target;
+        if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
+            /* pivot overflow: continue with the conversion loop */
+            *pErrorCode=U_ZERO_ERROR;
+        } else if(U_FAILURE(*pErrorCode) || (!flush && *pivotTarget==pivotStart)) {
+            /* conversion error, or there was nothing left to convert */
+            break;
+        }
+        /*
+         * else:
+         * _toUnicodeWithCallback() wrote into the pivot buffer,
+         * continue with fromUnicode conversion.
+         *
+         * Set the fromUnicode flush flag if we flush and if toUnicode has
+         * processed the end of the input.
+         */
+        if( flush && toUArgs.source==sourceLimit &&
+            sourceCnv->preToULength>=0 &&
+            sourceCnv->UCharErrorBufferLength==0
+        ) {
+            fromUArgs.flush=TRUE;
+        }
+    }
+
+    /*
+     * The conversion loop is exited when one of the following is true:
+     * - the entire source text has been converted successfully to the target buffer
+     * - a target buffer overflow occurred
+     * - a conversion error occurred
+     */
+
+    *source=toUArgs.source;
+    *target=fromUArgs.target;
+
+    /* terminate the target buffer if possible */
+    if(flush && U_SUCCESS(*pErrorCode)) {
+        if(*target!=targetLimit) {
+            **target=0;
+            if(*pErrorCode==U_STRING_NOT_TERMINATED_WARNING) {
+                *pErrorCode=U_ZERO_ERROR;
+            }
+        } else {
+            *pErrorCode=U_STRING_NOT_TERMINATED_WARNING;
+        }
+    }
+}
+
+/* internal implementation of ucnv_convert() etc. with preflighting */
+static int32_t
+ucnv_internalConvert(UConverter *outConverter, UConverter *inConverter,
+                     char *target, int32_t targetCapacity,
+                     const char *source, int32_t sourceLength,
+                     UErrorCode *pErrorCode) {
+    UChar pivotBuffer[CHUNK_SIZE];
+    UChar *pivot, *pivot2;
+
+    char *myTarget;
+    const char *sourceLimit;
+    const char *targetLimit;
+    int32_t targetLength=0;
+
+    /* set up */
+    if(sourceLength<0) {
+        sourceLimit=uprv_strchr(source, 0);
+    } else {
+        sourceLimit=source+sourceLength;
+    }
+
+    /* if there is no input data, we're done */
+    if(source==sourceLimit) {
+        return u_terminateChars(target, targetCapacity, 0, pErrorCode);
+    }
+
+    pivot=pivot2=pivotBuffer;
+    myTarget=target;
+    targetLength=0;
+
+    if(targetCapacity>0) {
+        /* perform real conversion */
+        targetLimit=target+targetCapacity;
+        ucnv_convertEx(outConverter, inConverter,
+                       &myTarget, targetLimit,
+                       &source, sourceLimit,
+                       pivotBuffer, &pivot, &pivot2, pivotBuffer+CHUNK_SIZE,
+                       FALSE,
+                       TRUE,
+                       pErrorCode);
+        targetLength=(int32_t)(myTarget-target);
+    }
+
+    /*
+     * If the output buffer is exhausted (or we are only "preflighting"), we need to stop writing
+     * to it but continue the conversion in order to store in targetCapacity
+     * the number of bytes that was required.
+     */
+    if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || targetCapacity==0)
+    {
+        char targetBuffer[CHUNK_SIZE];
+
+        targetLimit=targetBuffer+CHUNK_SIZE;
+        do {
+            *pErrorCode=U_ZERO_ERROR;
+            myTarget=targetBuffer;
+            ucnv_convertEx(outConverter, inConverter,
+                           &myTarget, targetLimit,
+                           &source, sourceLimit,
+                           pivotBuffer, &pivot, &pivot2, pivotBuffer+CHUNK_SIZE,
+                           FALSE,
+                           TRUE,
+                           pErrorCode);
+            targetLength+=(int32_t)(myTarget-targetBuffer);
+        } while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
+
+        /* done with preflighting, set warnings and errors as appropriate */
+        return u_terminateChars(target, targetCapacity, targetLength, pErrorCode);
+    }
+
+    /* no need to call u_terminateChars() because ucnv_convertEx() took care of that */
+    return targetLength;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_convert(const char *toConverterName, const char *fromConverterName,
+             char *target, int32_t targetCapacity,
+             const char *source, int32_t sourceLength,
+             UErrorCode *pErrorCode) {
+    UConverter in, out; /* stack-allocated */
+    UConverter *inConverter, *outConverter;
+    int32_t targetLength;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( source==NULL || sourceLength<-1 ||
+        targetCapacity<0 || (targetCapacity>0 && target==NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* if there is no input data, we're done */
+    if(sourceLength==0 || (sourceLength<0 && *source==0)) {
+        return u_terminateChars(target, targetCapacity, 0, pErrorCode);
+    }
+
+    /* create the converters */
+    inConverter=ucnv_createConverter(&in, fromConverterName, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    outConverter=ucnv_createConverter(&out, toConverterName, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        ucnv_close(inConverter);
+        return 0;
+    }
+
+    targetLength=ucnv_internalConvert(outConverter, inConverter,
+                                      target, targetCapacity,
+                                      source, sourceLength,
+                                      pErrorCode);
+
+    ucnv_close(inConverter);
+    ucnv_close(outConverter);
+
+    return targetLength;
+}
+
+/* @internal */
+static int32_t
+ucnv_convertAlgorithmic(UBool convertToAlgorithmic,
+                        UConverterType algorithmicType,
+                        UConverter *cnv,
+                        char *target, int32_t targetCapacity,
+                        const char *source, int32_t sourceLength,
+                        UErrorCode *pErrorCode) {
+    UConverter algoConverterStatic; /* stack-allocated */
+    UConverter *algoConverter, *to, *from;
+    int32_t targetLength;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( cnv==NULL || source==NULL || sourceLength<-1 ||
+        targetCapacity<0 || (targetCapacity>0 && target==NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* if there is no input data, we're done */
+    if(sourceLength==0 || (sourceLength<0 && *source==0)) {
+        return u_terminateChars(target, targetCapacity, 0, pErrorCode);
+    }
+
+    /* create the algorithmic converter */
+    algoConverter=ucnv_createAlgorithmicConverter(&algoConverterStatic, algorithmicType,
+                                                  "", 0, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* reset the other converter */
+    if(convertToAlgorithmic) {
+        /* cnv->Unicode->algo */
+        ucnv_resetToUnicode(cnv);
+        to=algoConverter;
+        from=cnv;
+    } else {
+        /* algo->Unicode->cnv */
+        ucnv_resetFromUnicode(cnv);
+        from=algoConverter;
+        to=cnv;
+    }
+
+    targetLength=ucnv_internalConvert(to, from,
+                                      target, targetCapacity,
+                                      source, sourceLength,
+                                      pErrorCode);
+
+    ucnv_close(algoConverter);
+
+    return targetLength;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_toAlgorithmic(UConverterType algorithmicType,
+                   UConverter *cnv,
+                   char *target, int32_t targetCapacity,
+                   const char *source, int32_t sourceLength,
+                   UErrorCode *pErrorCode) {
+    return ucnv_convertAlgorithmic(TRUE, algorithmicType, cnv,
+                                   target, targetCapacity,
+                                   source, sourceLength,
+                                   pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_fromAlgorithmic(UConverter *cnv,
+                     UConverterType algorithmicType,
+                     char *target, int32_t targetCapacity,
+                     const char *source, int32_t sourceLength,
+                     UErrorCode *pErrorCode) {
+    return ucnv_convertAlgorithmic(FALSE, algorithmicType, cnv,
+                                   target, targetCapacity,
+                                   source, sourceLength,
+                                   pErrorCode);
+}
+
+U_CAPI UConverterType  U_EXPORT2
+ucnv_getType(const UConverter* converter)
+{
+    int8_t type = converter->sharedData->staticData->conversionType;
+#if !UCONFIG_NO_LEGACY_CONVERSION
+    if(type == UCNV_MBCS) {
+        return ucnv_MBCSGetType(converter);
+    }
+#endif
+    return (UConverterType)type;
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_getStarters(const UConverter* converter, 
+                 UBool starters[256],
+                 UErrorCode* err)
+{
+    if (err == NULL || U_FAILURE(*err)) {
+        return;
+    }
+
+    if(converter->sharedData->impl->getStarters != NULL) {
+        converter->sharedData->impl->getStarters(converter, starters, err);
+    } else {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static const UAmbiguousConverter *ucnv_getAmbiguous(const UConverter *cnv)
+{
+    UErrorCode errorCode;
+    const char *name;
+    int32_t i;
+
+    if(cnv==NULL) {
+        return NULL;
+    }
+
+    errorCode=U_ZERO_ERROR;
+    name=ucnv_getName(cnv, &errorCode);
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+
+    for(i=0; i<UPRV_LENGTHOF(ambiguousConverters); ++i)
+    {
+        if(0==uprv_strcmp(name, ambiguousConverters[i].name))
+        {
+            return ambiguousConverters+i;
+        }
+    }
+
+    return NULL;
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_fixFileSeparator(const UConverter *cnv, 
+                      UChar* source, 
+                      int32_t sourceLength) {
+    const UAmbiguousConverter *a;
+    int32_t i;
+    UChar variant5c;
+
+    if(cnv==NULL || source==NULL || sourceLength<=0 || (a=ucnv_getAmbiguous(cnv))==NULL)
+    {
+        return;
+    }
+
+    variant5c=a->variant5c;
+    for(i=0; i<sourceLength; ++i) {
+        if(source[i]==variant5c) {
+            source[i]=0x5c;
+        }
+    }
+}
+
+U_CAPI UBool  U_EXPORT2
+ucnv_isAmbiguous(const UConverter *cnv) {
+    return (UBool)(ucnv_getAmbiguous(cnv)!=NULL);
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_setFallback(UConverter *cnv, UBool usesFallback)
+{
+    cnv->useFallback = usesFallback;
+}
+
+U_CAPI UBool  U_EXPORT2
+ucnv_usesFallback(const UConverter *cnv)
+{
+    return cnv->useFallback;
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_getInvalidChars (const UConverter * converter,
+                      char *errBytes,
+                      int8_t * len,
+                      UErrorCode * err)
+{
+    if (err == NULL || U_FAILURE(*err))
+    {
+        return;
+    }
+    if (len == NULL || errBytes == NULL || converter == NULL)
+    {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if (*len < converter->invalidCharLength)
+    {
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+    if ((*len = converter->invalidCharLength) > 0)
+    {
+        uprv_memcpy (errBytes, converter->invalidCharBuffer, *len);
+    }
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_getInvalidUChars (const UConverter * converter,
+                       UChar *errChars,
+                       int8_t * len,
+                       UErrorCode * err)
+{
+    if (err == NULL || U_FAILURE(*err))
+    {
+        return;
+    }
+    if (len == NULL || errChars == NULL || converter == NULL)
+    {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if (*len < converter->invalidUCharLength)
+    {
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+    if ((*len = converter->invalidUCharLength) > 0)
+    {
+        u_memcpy (errChars, converter->invalidUCharBuffer, *len);
+    }
+}
+
+#define SIG_MAX_LEN 5
+
+U_CAPI const char* U_EXPORT2
+ucnv_detectUnicodeSignature( const char* source,
+                             int32_t sourceLength,
+                             int32_t* signatureLength,
+                             UErrorCode* pErrorCode) {
+    int32_t dummy;
+
+    /* initial 0xa5 bytes: make sure that if we read <SIG_MAX_LEN
+     * bytes we don't misdetect something 
+     */
+    char start[SIG_MAX_LEN]={ '\xa5', '\xa5', '\xa5', '\xa5', '\xa5' };
+    int i = 0;
+
+    if((pErrorCode==NULL) || U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+    
+    if(source == NULL || sourceLength < -1){
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(signatureLength == NULL) {
+        signatureLength = &dummy;
+    }
+
+    if(sourceLength==-1){
+        sourceLength=(int32_t)uprv_strlen(source);
+    }
+
+    
+    while(i<sourceLength&& i<SIG_MAX_LEN){
+        start[i]=source[i];
+        i++;
+    }
+
+    if(start[0] == '\xFE' && start[1] == '\xFF') {
+        *signatureLength=2;
+        return  "UTF-16BE";
+    } else if(start[0] == '\xFF' && start[1] == '\xFE') {
+        if(start[2] == '\x00' && start[3] =='\x00') {
+            *signatureLength=4;
+            return "UTF-32LE";
+        } else {
+            *signatureLength=2;
+            return  "UTF-16LE";
+        }
+    } else if(start[0] == '\xEF' && start[1] == '\xBB' && start[2] == '\xBF') {
+        *signatureLength=3;
+        return  "UTF-8";
+    } else if(start[0] == '\x00' && start[1] == '\x00' && 
+              start[2] == '\xFE' && start[3]=='\xFF') {
+        *signatureLength=4;
+        return  "UTF-32BE";
+    } else if(start[0] == '\x0E' && start[1] == '\xFE' && start[2] == '\xFF') {
+        *signatureLength=3;
+        return "SCSU";
+    } else if(start[0] == '\xFB' && start[1] == '\xEE' && start[2] == '\x28') {
+        *signatureLength=3;
+        return "BOCU-1";
+    } else if(start[0] == '\x2B' && start[1] == '\x2F' && start[2] == '\x76') {
+        /*
+         * UTF-7: Initial U+FEFF is encoded as +/v8  or  +/v9  or  +/v+  or  +/v/
+         * depending on the second UTF-16 code unit.
+         * Detect the entire, closed Unicode mode sequence +/v8- for only U+FEFF
+         * if it occurs.
+         *
+         * So far we have +/v
+         */
+        if(start[3] == '\x38' && start[4] == '\x2D') {
+            /* 5 bytes +/v8- */
+            *signatureLength=5;
+            return "UTF-7";
+        } else if(start[3] == '\x38' || start[3] == '\x39' || start[3] == '\x2B' || start[3] == '\x2F') {
+            /* 4 bytes +/v8  or  +/v9  or  +/v+  or  +/v/ */
+            *signatureLength=4;
+            return "UTF-7";
+        }
+    }else if(start[0]=='\xDD' && start[1]== '\x73'&& start[2]=='\x66' && start[3]=='\x73'){
+        *signatureLength=4;
+        return "UTF-EBCDIC";
+    }
+
+
+    /* no known Unicode signature byte sequence recognized */
+    *signatureLength=0;
+    return NULL;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_fromUCountPending(const UConverter* cnv, UErrorCode* status)
+{
+    if(status == NULL || U_FAILURE(*status)){
+        return -1;
+    }
+    if(cnv == NULL){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return -1;
+    }
+
+    if(cnv->preFromUFirstCP >= 0){
+        return U16_LENGTH(cnv->preFromUFirstCP)+cnv->preFromULength ;
+    }else if(cnv->preFromULength < 0){
+        return -cnv->preFromULength ;
+    }else if(cnv->fromUChar32 > 0){
+        return 1;
+    }
+    return 0; 
+
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_toUCountPending(const UConverter* cnv, UErrorCode* status){
+
+    if(status == NULL || U_FAILURE(*status)){
+        return -1;
+    }
+    if(cnv == NULL){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return -1;
+    }
+
+    if(cnv->preToULength > 0){
+        return cnv->preToULength ;
+    }else if(cnv->preToULength < 0){
+        return -cnv->preToULength;
+    }else if(cnv->toULength > 0){
+        return cnv->toULength;
+    }
+    return 0;
+}
+
+U_CAPI UBool U_EXPORT2
+ucnv_isFixedWidth(UConverter *cnv, UErrorCode *status){
+    if (U_FAILURE(*status)) {
+        return FALSE;
+    }
+
+    if (cnv == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+
+    switch (ucnv_getType(cnv)) {
+        case UCNV_SBCS:
+        case UCNV_DBCS:
+        case UCNV_UTF32_BigEndian:
+        case UCNV_UTF32_LittleEndian:
+        case UCNV_UTF32:
+        case UCNV_US_ASCII:
+            return TRUE;
+        default:
+            return FALSE;
+    }
+}
+#endif
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/ucnv2022.cpp b/thirdparty/icu4c/common/ucnv2022.cpp
new file mode 100644
index 0000000000..169ad4c526
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv2022.cpp
@@ -0,0 +1,3973 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv2022.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000feb03
+*   created by: Markus W. Scherer
+*
+*   Change history:
+*
+*   06/29/2000  helena  Major rewrite of the callback APIs.
+*   08/08/2000  Ram     Included support for ISO-2022-JP-2
+*                       Changed implementation of toUnicode
+*                       function
+*   08/21/2000  Ram     Added support for ISO-2022-KR
+*   08/29/2000  Ram     Seperated implementation of EBCDIC to
+*                       ucnvebdc.c
+*   09/20/2000  Ram     Added support for ISO-2022-CN
+*                       Added implementations for getNextUChar()
+*                       for specific 2022 country variants.
+*   10/31/2000  Ram     Implemented offsets logic functions
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/uset.h"
+#include "unicode/ucnv_err.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/utf16.h"
+#include "ucnv_imp.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "ucnvmbcs.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+/*
+ * I am disabling the generic ISO-2022 converter after proposing to do so on
+ * the icu mailing list two days ago.
+ *
+ * Reasons:
+ * 1. It does not fully support the ISO-2022/ECMA-35 specification with all of
+ *    its designation sequences, single shifts with return to the previous state,
+ *    switch-with-no-return to UTF-16BE or similar, etc.
+ *    This is unlike the language-specific variants like ISO-2022-JP which
+ *    require a much smaller repertoire of ISO-2022 features.
+ *    These variants continue to be supported.
+ * 2. I believe that no one is really using the generic ISO-2022 converter
+ *    but rather always one of the language-specific variants.
+ *    Note that ICU's generic ISO-2022 converter has always output one escape
+ *    sequence followed by UTF-8 for the whole stream.
+ * 3. Switching between subcharsets is extremely slow, because each time
+ *    the previous converter is closed and a new one opened,
+ *    without any kind of caching, least-recently-used list, etc.
+ * 4. The code is currently buggy, and given the above it does not seem
+ *    reasonable to spend the time on maintenance.
+ * 5. ISO-2022 subcharsets should normally be used with 7-bit byte encodings.
+ *    This means, for example, that when ISO-8859-7 is designated, the following
+ *    ISO-2022 bytes 00..7f should be interpreted as ISO-8859-7 bytes 80..ff.
+ *    The ICU ISO-2022 converter does not handle this - and has no information
+ *    about which subconverter would have to be shifted vs. which is designed
+ *    for 7-bit ISO-2022.
+ *
+ * Markus Scherer 2003-dec-03
+ */
+#endif
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+static const char SHIFT_IN_STR[]  = "\x0F";
+// static const char SHIFT_OUT_STR[] = "\x0E";
+#endif
+
+#define CR      0x0D
+#define LF      0x0A
+#define H_TAB   0x09
+#define V_TAB   0x0B
+#define SPACE   0x20
+
+enum {
+    HWKANA_START=0xff61,
+    HWKANA_END=0xff9f
+};
+
+/*
+ * 94-character sets with native byte values A1..FE are encoded in ISO 2022
+ * as bytes 21..7E. (Subtract 0x80.)
+ * 96-character sets with native byte values A0..FF are encoded in ISO 2022
+ * as bytes 20..7F. (Subtract 0x80.)
+ * Do not encode C1 control codes with native bytes 80..9F
+ * as bytes 00..1F (C0 control codes).
+ */
+enum {
+    GR94_START=0xa1,
+    GR94_END=0xfe,
+    GR96_START=0xa0,
+    GR96_END=0xff
+};
+
+/*
+ * ISO 2022 control codes must not be converted from Unicode
+ * because they would mess up the byte stream.
+ * The bit mask 0x0800c000 has bits set at bit positions 0xe, 0xf, 0x1b
+ * corresponding to SO, SI, and ESC.
+ */
+#define IS_2022_CONTROL(c) (((c)<0x20) && (((uint32_t)1<<(c))&0x0800c000)!=0)
+
+/* for ISO-2022-JP and -CN implementations */
+typedef enum  {
+        /* shared values */
+        INVALID_STATE=-1,
+        ASCII = 0,
+
+        SS2_STATE=0x10,
+        SS3_STATE,
+
+        /* JP */
+        ISO8859_1 = 1 ,
+        ISO8859_7 = 2 ,
+        JISX201  = 3,
+        JISX208 = 4,
+        JISX212 = 5,
+        GB2312  =6,
+        KSC5601 =7,
+        HWKANA_7BIT=8,    /* Halfwidth Katakana 7 bit */
+
+        /* CN */
+        /* the first few enum constants must keep their values because they correspond to myConverterArray[] */
+        GB2312_1=1,
+        ISO_IR_165=2,
+        CNS_11643=3,
+
+        /*
+         * these are used in StateEnum and ISO2022State variables,
+         * but CNS_11643 must be used to index into myConverterArray[]
+         */
+        CNS_11643_0=0x20,
+        CNS_11643_1,
+        CNS_11643_2,
+        CNS_11643_3,
+        CNS_11643_4,
+        CNS_11643_5,
+        CNS_11643_6,
+        CNS_11643_7
+} StateEnum;
+
+/* is the StateEnum charset value for a DBCS charset? */
+#if UCONFIG_ONLY_HTML_CONVERSION
+#define IS_JP_DBCS(cs) (JISX208==(cs))
+#else
+#define IS_JP_DBCS(cs) (JISX208<=(cs) && (cs)<=KSC5601)
+#endif
+
+#define CSM(cs) ((uint16_t)1<<(cs))
+
+/*
+ * Each of these charset masks (with index x) contains a bit for a charset in exact correspondence
+ * to whether that charset is used in the corresponding version x of ISO_2022,locale=ja,version=x
+ *
+ * Note: The converter uses some leniency:
+ * - The escape sequence ESC ( I for half-width 7-bit Katakana is recognized in
+ *   all versions, not just JIS7 and JIS8.
+ * - ICU does not distinguish between different versions of JIS X 0208.
+ */
+#if UCONFIG_ONLY_HTML_CONVERSION
+enum { MAX_JA_VERSION=0 };
+#else
+enum { MAX_JA_VERSION=4 };
+#endif
+static const uint16_t jpCharsetMasks[MAX_JA_VERSION+1]={
+    CSM(ASCII)|CSM(JISX201)|CSM(JISX208)|CSM(HWKANA_7BIT),
+#if !UCONFIG_ONLY_HTML_CONVERSION
+    CSM(ASCII)|CSM(JISX201)|CSM(JISX208)|CSM(HWKANA_7BIT)|CSM(JISX212),
+    CSM(ASCII)|CSM(JISX201)|CSM(JISX208)|CSM(HWKANA_7BIT)|CSM(JISX212)|CSM(GB2312)|CSM(KSC5601)|CSM(ISO8859_1)|CSM(ISO8859_7),
+    CSM(ASCII)|CSM(JISX201)|CSM(JISX208)|CSM(HWKANA_7BIT)|CSM(JISX212)|CSM(GB2312)|CSM(KSC5601)|CSM(ISO8859_1)|CSM(ISO8859_7),
+    CSM(ASCII)|CSM(JISX201)|CSM(JISX208)|CSM(HWKANA_7BIT)|CSM(JISX212)|CSM(GB2312)|CSM(KSC5601)|CSM(ISO8859_1)|CSM(ISO8859_7)
+#endif
+};
+
+typedef enum {
+        ASCII1=0,
+        LATIN1,
+        SBCS,
+        DBCS,
+        MBCS,
+        HWKANA
+}Cnv2022Type;
+
+typedef struct ISO2022State {
+    int8_t cs[4];       /* charset number for SI (G0)/SO (G1)/SS2 (G2)/SS3 (G3) */
+    int8_t g;           /* 0..3 for G0..G3 (SI/SO/SS2/SS3) */
+    int8_t prevG;       /* g before single shift (SS2 or SS3) */
+} ISO2022State;
+
+#define UCNV_OPTIONS_VERSION_MASK 0xf
+#define UCNV_2022_MAX_CONVERTERS 10
+
+typedef struct{
+    UConverterSharedData *myConverterArray[UCNV_2022_MAX_CONVERTERS];
+    UConverter *currentConverter;
+    Cnv2022Type currentType;
+    ISO2022State toU2022State, fromU2022State;
+    uint32_t key;
+    uint32_t version;
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    UBool isFirstBuffer;
+#endif
+    UBool isEmptySegment;
+    char name[30];
+    char locale[3];
+}UConverterDataISO2022;
+
+/* Protos */
+/* ISO-2022 ----------------------------------------------------------------- */
+
+/*Forward declaration */
+U_CFUNC void U_CALLCONV
+ucnv_fromUnicode_UTF8(UConverterFromUnicodeArgs * args,
+                      UErrorCode * err);
+U_CFUNC void U_CALLCONV
+ucnv_fromUnicode_UTF8_OFFSETS_LOGIC(UConverterFromUnicodeArgs * args,
+                                    UErrorCode * err);
+
+#define ESC_2022 0x1B /*ESC*/
+
+typedef enum
+{
+        INVALID_2022 = -1, /*Doesn't correspond to a valid iso 2022 escape sequence*/
+        VALID_NON_TERMINAL_2022 = 0, /*so far corresponds to a valid iso 2022 escape sequence*/
+        VALID_TERMINAL_2022 = 1, /*corresponds to a valid iso 2022 escape sequence*/
+        VALID_MAYBE_TERMINAL_2022 = 2 /*so far matches one iso 2022 escape sequence, but by adding more characters might match another escape sequence*/
+} UCNV_TableStates_2022;
+
+/*
+* The way these state transition arrays work is:
+* ex : ESC$B is the sequence for JISX208
+*      a) First Iteration: char is ESC
+*          i) Get the value of ESC from normalize_esq_chars_2022[] with int value of ESC as index
+*             int x = normalize_esq_chars_2022[27] which is equal to 1
+*         ii) Search for this value in escSeqStateTable_Key_2022[]
+*             value of x is stored at escSeqStateTable_Key_2022[0]
+*        iii) Save this index as offset
+*         iv) Get state of this sequence from escSeqStateTable_Value_2022[]
+*             escSeqStateTable_Value_2022[offset], which is VALID_NON_TERMINAL_2022
+*     b) Switch on this state and continue to next char
+*          i) Get the value of $ from normalize_esq_chars_2022[] with int value of $ as index
+*             which is normalize_esq_chars_2022[36] == 4
+*         ii) x is currently 1(from above)
+*               x<<=5 -- x is now 32
+*               x+=normalize_esq_chars_2022[36]
+*               now x is 36
+*        iii) Search for this value in escSeqStateTable_Key_2022[]
+*             value of x is stored at escSeqStateTable_Key_2022[2], so offset is 2
+*         iv) Get state of this sequence from escSeqStateTable_Value_2022[]
+*             escSeqStateTable_Value_2022[offset], which is VALID_NON_TERMINAL_2022
+*     c) Switch on this state and continue to next char
+*        i)  Get the value of B from normalize_esq_chars_2022[] with int value of B as index
+*        ii) x is currently 36 (from above)
+*            x<<=5 -- x is now 1152
+*            x+=normalize_esq_chars_2022[66]
+*            now x is 1161
+*       iii) Search for this value in escSeqStateTable_Key_2022[]
+*            value of x is stored at escSeqStateTable_Key_2022[21], so offset is 21
+*        iv) Get state of this sequence from escSeqStateTable_Value_2022[21]
+*            escSeqStateTable_Value_2022[offset], which is VALID_TERMINAL_2022
+*         v) Get the converter name form escSeqStateTable_Result_2022[21] which is JISX208
+*/
+
+
+/*Below are the 3 arrays depicting a state transition table*/
+static const int8_t normalize_esq_chars_2022[256] = {
+/*       0      1       2       3       4      5       6        7       8       9           */
+
+         0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,1      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,4      ,7      ,29      ,0
+        ,2     ,24     ,26     ,27     ,0      ,3      ,23     ,6      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,5      ,8      ,9      ,10     ,11     ,12
+        ,13    ,14     ,15     ,16     ,17     ,18     ,19     ,20     ,25     ,28
+        ,0     ,0      ,21     ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,22    ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0      ,0
+        ,0     ,0      ,0      ,0      ,0      ,0
+};
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+/*
+ * When the generic ISO-2022 converter is completely removed, not just disabled
+ * per #ifdef, then the following state table and the associated tables that are
+ * dimensioned with MAX_STATES_2022 should be trimmed.
+ *
+ * Especially, VALID_MAYBE_TERMINAL_2022 will not be used any more, and all of
+ * the associated escape sequences starting with ESC ( B should be removed.
+ * This includes the ones with key values 1097 and all of the ones above 1000000.
+ *
+ * For the latter, the tables can simply be truncated.
+ * For the former, since the tables must be kept parallel, it is probably best
+ * to simply duplicate an adjacent table cell, parallel in all tables.
+ *
+ * It may make sense to restructure the tables, especially by using small search
+ * tables for the variants instead of indexing them parallel to the table here.
+ */
+#endif
+
+#define MAX_STATES_2022 74
+static const int32_t escSeqStateTable_Key_2022[MAX_STATES_2022] = {
+/*   0           1           2           3           4           5           6           7           8           9           */
+
+     1          ,34         ,36         ,39         ,55         ,57         ,60         ,61         ,1093       ,1096
+    ,1097       ,1098       ,1099       ,1100       ,1101       ,1102       ,1103       ,1104       ,1105       ,1106
+    ,1109       ,1154       ,1157       ,1160       ,1161       ,1176       ,1178       ,1179       ,1254       ,1257
+    ,1768       ,1773       ,1957       ,35105      ,36933      ,36936      ,36937      ,36938      ,36939      ,36940
+    ,36942      ,36943      ,36944      ,36945      ,36946      ,36947      ,36948      ,37640      ,37642      ,37644
+    ,37646      ,37711      ,37744      ,37745      ,37746      ,37747      ,37748      ,40133      ,40136      ,40138
+    ,40139      ,40140      ,40141      ,1123363    ,35947624   ,35947625   ,35947626   ,35947627   ,35947629   ,35947630
+    ,35947631   ,35947635   ,35947636   ,35947638
+};
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+
+static const char* const escSeqStateTable_Result_2022[MAX_STATES_2022] = {
+ /*  0                      1                        2                      3                   4                   5                        6                      7                       8                       9    */
+
+     NULL                   ,NULL                   ,NULL                   ,NULL               ,NULL               ,NULL                   ,NULL                   ,NULL                   ,"latin1"               ,"latin1"
+    ,"latin1"               ,"ibm-865"              ,"ibm-865"              ,"ibm-865"          ,"ibm-865"          ,"ibm-865"              ,"ibm-865"              ,"JISX0201"             ,"JISX0201"             ,"latin1"
+    ,"latin1"               ,NULL                   ,"JISX-208"             ,"ibm-5478"         ,"JISX-208"         ,NULL                   ,NULL                   ,NULL                   ,NULL                   ,"UTF8"
+    ,"ISO-8859-1"           ,"ISO-8859-7"           ,"JIS-X-208"            ,NULL               ,"ibm-955"          ,"ibm-367"              ,"ibm-952"              ,"ibm-949"              ,"JISX-212"             ,"ibm-1383"
+    ,"ibm-952"              ,"ibm-964"              ,"ibm-964"              ,"ibm-964"          ,"ibm-964"          ,"ibm-964"              ,"ibm-964"              ,"ibm-5478"         ,"ibm-949"              ,"ISO-IR-165"
+    ,"CNS-11643-1992,1"     ,"CNS-11643-1992,2"     ,"CNS-11643-1992,3"     ,"CNS-11643-1992,4" ,"CNS-11643-1992,5" ,"CNS-11643-1992,6"     ,"CNS-11643-1992,7"     ,"UTF16_PlatformEndian" ,"UTF16_PlatformEndian" ,"UTF16_PlatformEndian"
+    ,"UTF16_PlatformEndian" ,"UTF16_PlatformEndian" ,"UTF16_PlatformEndian" ,NULL               ,"latin1"           ,"ibm-912"              ,"ibm-913"              ,"ibm-914"              ,"ibm-813"              ,"ibm-1089"
+    ,"ibm-920"              ,"ibm-915"              ,"ibm-915"              ,"latin1"
+};
+
+#endif
+
+static const int8_t escSeqStateTable_Value_2022[MAX_STATES_2022] = {
+/*          0                           1                         2                             3                           4                           5                               6                        7                          8                           9       */
+     VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022     ,VALID_NON_TERMINAL_2022   ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_NON_TERMINAL_2022    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_MAYBE_TERMINAL_2022  ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_NON_TERMINAL_2022    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022    ,VALID_NON_TERMINAL_2022    ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_NON_TERMINAL_2022    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_NON_TERMINAL_2022    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+    ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022        ,VALID_TERMINAL_2022
+};
+
+/* Type def for refactoring changeState_2022 code*/
+typedef enum{
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    ISO_2022=0,
+#endif
+    ISO_2022_JP=1,
+#if !UCONFIG_ONLY_HTML_CONVERSION
+    ISO_2022_KR=2,
+    ISO_2022_CN=3
+#endif
+} Variant2022;
+
+/*********** ISO 2022 Converter Protos ***********/
+static void U_CALLCONV
+_ISO2022Open(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode);
+
+static void U_CALLCONV
+ _ISO2022Close(UConverter *converter);
+
+static void U_CALLCONV
+_ISO2022Reset(UConverter *converter, UConverterResetChoice choice);
+
+U_CDECL_BEGIN
+static const char * U_CALLCONV
+_ISO2022getName(const UConverter* cnv);
+U_CDECL_END
+
+static void  U_CALLCONV
+_ISO_2022_WriteSub(UConverterFromUnicodeArgs *args, int32_t offsetIndex, UErrorCode *err);
+
+U_CDECL_BEGIN
+static UConverter * U_CALLCONV
+_ISO_2022_SafeClone(const UConverter *cnv, void *stackBuffer, int32_t *pBufferSize, UErrorCode *status);
+
+U_CDECL_END
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+static void U_CALLCONV
+T_UConverter_toUnicode_ISO_2022_OFFSETS_LOGIC(UConverterToUnicodeArgs* args, UErrorCode* err);
+#endif
+
+namespace {
+
+/*const UConverterSharedData _ISO2022Data;*/
+extern const UConverterSharedData _ISO2022JPData;
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+extern const UConverterSharedData _ISO2022KRData;
+extern const UConverterSharedData _ISO2022CNData;
+#endif
+
+}  // namespace
+
+/*************** Converter implementations ******************/
+
+/* The purpose of this function is to get around gcc compiler warnings. */
+static inline void
+fromUWriteUInt8(UConverter *cnv,
+                 const char *bytes, int32_t length,
+                 uint8_t **target, const char *targetLimit,
+                 int32_t **offsets,
+                 int32_t sourceIndex,
+                 UErrorCode *pErrorCode)
+{
+    char *targetChars = (char *)*target;
+    ucnv_fromUWriteBytes(cnv, bytes, length, &targetChars, targetLimit,
+                         offsets, sourceIndex, pErrorCode);
+    *target = (uint8_t*)targetChars;
+
+}
+
+static inline void
+setInitialStateToUnicodeKR(UConverter* /*converter*/, UConverterDataISO2022 *myConverterData){
+    if(myConverterData->version == 1) {
+        UConverter *cnv = myConverterData->currentConverter;
+
+        cnv->toUnicodeStatus=0;     /* offset */
+        cnv->mode=0;                /* state */
+        cnv->toULength=0;           /* byteIndex */
+    }
+}
+
+static inline void
+setInitialStateFromUnicodeKR(UConverter* converter,UConverterDataISO2022 *myConverterData){
+   /* in ISO-2022-KR the designator sequence appears only once
+    * in a file so we append it only once
+    */
+    if( converter->charErrorBufferLength==0){
+
+        converter->charErrorBufferLength = 4;
+        converter->charErrorBuffer[0] = 0x1b;
+        converter->charErrorBuffer[1] = 0x24;
+        converter->charErrorBuffer[2] = 0x29;
+        converter->charErrorBuffer[3] = 0x43;
+    }
+    if(myConverterData->version == 1) {
+        UConverter *cnv = myConverterData->currentConverter;
+
+        cnv->fromUChar32=0;
+        cnv->fromUnicodeStatus=1;   /* prevLength */
+    }
+}
+
+static void U_CALLCONV
+_ISO2022Open(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode){
+
+    char myLocale[7]={' ',' ',' ',' ',' ',' ', '\0'};
+
+    cnv->extraInfo = uprv_malloc (sizeof (UConverterDataISO2022));
+    if(cnv->extraInfo != NULL) {
+        UConverterNamePieces stackPieces;
+        UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+        UConverterDataISO2022 *myConverterData=(UConverterDataISO2022 *) cnv->extraInfo;
+        uint32_t version;
+
+        stackArgs.onlyTestIsLoadable = pArgs->onlyTestIsLoadable;
+
+        uprv_memset(myConverterData, 0, sizeof(UConverterDataISO2022));
+        myConverterData->currentType = ASCII1;
+        cnv->fromUnicodeStatus =FALSE;
+        if(pArgs->locale){
+            uprv_strncpy(myLocale, pArgs->locale, sizeof(myLocale)-1);
+        }
+        version = pArgs->options & UCNV_OPTIONS_VERSION_MASK;
+        myConverterData->version = version;
+        if(myLocale[0]=='j' && (myLocale[1]=='a'|| myLocale[1]=='p') &&
+            (myLocale[2]=='_' || myLocale[2]=='\0'))
+        {
+            /* open the required converters and cache them */
+            if(version>MAX_JA_VERSION) {
+                // ICU 55 fails to open a converter for an unsupported version.
+                // Previously, it fell back to version 0, but that would yield
+                // unexpected behavior.
+                *errorCode = U_MISSING_RESOURCE_ERROR;
+                return;
+            }
+            if(jpCharsetMasks[version]&CSM(ISO8859_7)) {
+                myConverterData->myConverterArray[ISO8859_7] =
+                    ucnv_loadSharedData("ISO8859_7", &stackPieces, &stackArgs, errorCode);
+            }
+            myConverterData->myConverterArray[JISX208] =
+                ucnv_loadSharedData("Shift-JIS", &stackPieces, &stackArgs, errorCode);
+            if(jpCharsetMasks[version]&CSM(JISX212)) {
+                myConverterData->myConverterArray[JISX212] =
+                    ucnv_loadSharedData("jisx-212", &stackPieces, &stackArgs, errorCode);
+            }
+            if(jpCharsetMasks[version]&CSM(GB2312)) {
+                myConverterData->myConverterArray[GB2312] =
+                    ucnv_loadSharedData("ibm-5478", &stackPieces, &stackArgs, errorCode);   /* gb_2312_80-1 */
+            }
+            if(jpCharsetMasks[version]&CSM(KSC5601)) {
+                myConverterData->myConverterArray[KSC5601] =
+                    ucnv_loadSharedData("ksc_5601", &stackPieces, &stackArgs, errorCode);
+            }
+
+            /* set the function pointers to appropriate funtions */
+            cnv->sharedData=(UConverterSharedData*)(&_ISO2022JPData);
+            uprv_strcpy(myConverterData->locale,"ja");
+
+            (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=ja,version=");
+            size_t len = uprv_strlen(myConverterData->name);
+            myConverterData->name[len]=(char)(myConverterData->version+(int)'0');
+            myConverterData->name[len+1]='\0';
+        }
+#if !UCONFIG_ONLY_HTML_CONVERSION
+        else if(myLocale[0]=='k' && (myLocale[1]=='o'|| myLocale[1]=='r') &&
+            (myLocale[2]=='_' || myLocale[2]=='\0'))
+        {
+            if(version>1) {
+                // ICU 55 fails to open a converter for an unsupported version.
+                // Previously, it fell back to version 0, but that would yield
+                // unexpected behavior.
+                *errorCode = U_MISSING_RESOURCE_ERROR;
+                return;
+            }
+            const char *cnvName;
+            if(version==1) {
+                cnvName="icu-internal-25546";
+            } else {
+                cnvName="ibm-949";
+                myConverterData->version=version=0;
+            }
+            if(pArgs->onlyTestIsLoadable) {
+                ucnv_canCreateConverter(cnvName, errorCode);  /* errorCode carries result */
+                uprv_free(cnv->extraInfo);
+                cnv->extraInfo=NULL;
+                return;
+            } else {
+                myConverterData->currentConverter=ucnv_open(cnvName, errorCode);
+                if (U_FAILURE(*errorCode)) {
+                    _ISO2022Close(cnv);
+                    return;
+                }
+
+                if(version==1) {
+                    (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=ko,version=1");
+                    uprv_memcpy(cnv->subChars, myConverterData->currentConverter->subChars, 4);
+                    cnv->subCharLen = myConverterData->currentConverter->subCharLen;
+                }else{
+                    (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=ko,version=0");
+                }
+
+                /* initialize the state variables */
+                setInitialStateToUnicodeKR(cnv, myConverterData);
+                setInitialStateFromUnicodeKR(cnv, myConverterData);
+
+                /* set the function pointers to appropriate funtions */
+                cnv->sharedData=(UConverterSharedData*)&_ISO2022KRData;
+                uprv_strcpy(myConverterData->locale,"ko");
+            }
+        }
+        else if(((myLocale[0]=='z' && myLocale[1]=='h') || (myLocale[0]=='c'&& myLocale[1]=='n'))&&
+            (myLocale[2]=='_' || myLocale[2]=='\0'))
+        {
+            if(version>2) {
+                // ICU 55 fails to open a converter for an unsupported version.
+                // Previously, it fell back to version 0, but that would yield
+                // unexpected behavior.
+                *errorCode = U_MISSING_RESOURCE_ERROR;
+                return;
+            }
+
+            /* open the required converters and cache them */
+            myConverterData->myConverterArray[GB2312_1] =
+                ucnv_loadSharedData("ibm-5478", &stackPieces, &stackArgs, errorCode);
+            if(version==1) {
+                myConverterData->myConverterArray[ISO_IR_165] =
+                    ucnv_loadSharedData("iso-ir-165", &stackPieces, &stackArgs, errorCode);
+            }
+            myConverterData->myConverterArray[CNS_11643] =
+                ucnv_loadSharedData("cns-11643-1992", &stackPieces, &stackArgs, errorCode);
+
+
+            /* set the function pointers to appropriate funtions */
+            cnv->sharedData=(UConverterSharedData*)&_ISO2022CNData;
+            uprv_strcpy(myConverterData->locale,"cn");
+
+            if (version==0){
+                myConverterData->version = 0;
+                (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=zh,version=0");
+            }else if (version==1){
+                myConverterData->version = 1;
+                (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=zh,version=1");
+            }else {
+                myConverterData->version = 2;
+                (void)uprv_strcpy(myConverterData->name,"ISO_2022,locale=zh,version=2");
+            }
+        }
+#endif  // !UCONFIG_ONLY_HTML_CONVERSION
+        else{
+#ifdef U_ENABLE_GENERIC_ISO_2022
+            myConverterData->isFirstBuffer = TRUE;
+
+            /* append the UTF-8 escape sequence */
+            cnv->charErrorBufferLength = 3;
+            cnv->charErrorBuffer[0] = 0x1b;
+            cnv->charErrorBuffer[1] = 0x25;
+            cnv->charErrorBuffer[2] = 0x42;
+
+            cnv->sharedData=(UConverterSharedData*)&_ISO2022Data;
+            /* initialize the state variables */
+            uprv_strcpy(myConverterData->name,"ISO_2022");
+#else
+            *errorCode = U_MISSING_RESOURCE_ERROR;
+            // Was U_UNSUPPORTED_ERROR but changed in ICU 55 to a more standard
+            // data loading error code.
+            return;
+#endif
+        }
+
+        cnv->maxBytesPerUChar=cnv->sharedData->staticData->maxBytesPerChar;
+
+        if(U_FAILURE(*errorCode) || pArgs->onlyTestIsLoadable) {
+            _ISO2022Close(cnv);
+        }
+    } else {
+        *errorCode = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+
+static void U_CALLCONV
+_ISO2022Close(UConverter *converter) {
+    UConverterDataISO2022* myData =(UConverterDataISO2022 *) (converter->extraInfo);
+    UConverterSharedData **array = myData->myConverterArray;
+    int32_t i;
+
+    if (converter->extraInfo != NULL) {
+        /*close the array of converter pointers and free the memory*/
+        for (i=0; i<UCNV_2022_MAX_CONVERTERS; i++) {
+            if(array[i]!=NULL) {
+                ucnv_unloadSharedDataIfReady(array[i]);
+            }
+        }
+
+        ucnv_close(myData->currentConverter);
+
+        if(!converter->isExtraLocal){
+            uprv_free (converter->extraInfo);
+            converter->extraInfo = NULL;
+        }
+    }
+}
+
+static void U_CALLCONV
+_ISO2022Reset(UConverter *converter, UConverterResetChoice choice) {
+    UConverterDataISO2022 *myConverterData=(UConverterDataISO2022 *) (converter->extraInfo);
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        uprv_memset(&myConverterData->toU2022State, 0, sizeof(ISO2022State));
+        myConverterData->key = 0;
+        myConverterData->isEmptySegment = FALSE;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        uprv_memset(&myConverterData->fromU2022State, 0, sizeof(ISO2022State));
+    }
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    if(myConverterData->locale[0] == 0){
+        if(choice<=UCNV_RESET_TO_UNICODE) {
+            myConverterData->isFirstBuffer = TRUE;
+            myConverterData->key = 0;
+            if (converter->mode == UCNV_SO){
+                ucnv_close (myConverterData->currentConverter);
+                myConverterData->currentConverter=NULL;
+            }
+            converter->mode = UCNV_SI;
+        }
+        if(choice!=UCNV_RESET_TO_UNICODE) {
+            /* re-append UTF-8 escape sequence */
+            converter->charErrorBufferLength = 3;
+            converter->charErrorBuffer[0] = 0x1b;
+            converter->charErrorBuffer[1] = 0x28;
+            converter->charErrorBuffer[2] = 0x42;
+        }
+    }
+    else
+#endif
+    {
+        /* reset the state variables */
+        if(myConverterData->locale[0] == 'k'){
+            if(choice<=UCNV_RESET_TO_UNICODE) {
+                setInitialStateToUnicodeKR(converter, myConverterData);
+            }
+            if(choice!=UCNV_RESET_TO_UNICODE) {
+                setInitialStateFromUnicodeKR(converter, myConverterData);
+            }
+        }
+    }
+}
+
+U_CDECL_BEGIN
+
+static const char * U_CALLCONV
+_ISO2022getName(const UConverter* cnv){
+    if(cnv->extraInfo){
+        UConverterDataISO2022* myData= (UConverterDataISO2022*)cnv->extraInfo;
+        return myData->name;
+    }
+    return NULL;
+}
+
+U_CDECL_END
+
+
+/*************** to unicode *******************/
+/****************************************************************************
+ * Recognized escape sequences are
+ * <ESC>(B  ASCII
+ * <ESC>.A  ISO-8859-1
+ * <ESC>.F  ISO-8859-7
+ * <ESC>(J  JISX-201
+ * <ESC>(I  JISX-201
+ * <ESC>$B  JISX-208
+ * <ESC>$@  JISX-208
+ * <ESC>$(D JISX-212
+ * <ESC>$A  GB2312
+ * <ESC>$(C KSC5601
+ */
+static const int8_t nextStateToUnicodeJP[MAX_STATES_2022]= {
+/*      0                1               2               3               4               5               6               7               8               9    */
+    INVALID_STATE   ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,SS2_STATE      ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,ASCII          ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,JISX201        ,HWKANA_7BIT    ,JISX201        ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,JISX208        ,GB2312         ,JISX208        ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,ISO8859_1      ,ISO8859_7      ,JISX208        ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,KSC5601        ,JISX212        ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+};
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+/*************** to unicode *******************/
+static const int8_t nextStateToUnicodeCN[MAX_STATES_2022]= {
+/*      0                1               2               3               4               5               6               7               8               9    */
+     INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,SS2_STATE      ,SS3_STATE      ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,GB2312_1       ,INVALID_STATE  ,ISO_IR_165
+    ,CNS_11643_1    ,CNS_11643_2    ,CNS_11643_3    ,CNS_11643_4    ,CNS_11643_5    ,CNS_11643_6    ,CNS_11643_7    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+    ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE  ,INVALID_STATE
+};
+#endif
+
+
+static UCNV_TableStates_2022
+getKey_2022(char c,int32_t* key,int32_t* offset){
+    int32_t togo;
+    int32_t low = 0;
+    int32_t hi = MAX_STATES_2022;
+    int32_t oldmid=0;
+
+    togo = normalize_esq_chars_2022[(uint8_t)c];
+    if(togo == 0) {
+        /* not a valid character anywhere in an escape sequence */
+        *key = 0;
+        *offset = 0;
+        return INVALID_2022;
+    }
+    togo = (*key << 5) + togo;
+
+    while (hi != low)  /*binary search*/{
+
+        int32_t mid = (hi+low) >> 1; /*Finds median*/
+
+        if (mid == oldmid)
+            break;
+
+        if (escSeqStateTable_Key_2022[mid] > togo){
+            hi = mid;
+        }
+        else if (escSeqStateTable_Key_2022[mid] < togo){
+            low = mid;
+        }
+        else /*we found it*/{
+            *key = togo;
+            *offset = mid;
+            return (UCNV_TableStates_2022)escSeqStateTable_Value_2022[mid];
+        }
+        oldmid = mid;
+
+    }
+
+    *key = 0;
+    *offset = 0;
+    return INVALID_2022;
+}
+
+/*runs through a state machine to determine the escape sequence - codepage correspondance
+ */
+static void
+changeState_2022(UConverter* _this,
+                const char** source,
+                const char* sourceLimit,
+                Variant2022 var,
+                UErrorCode* err){
+    UCNV_TableStates_2022 value;
+    UConverterDataISO2022* myData2022 = ((UConverterDataISO2022*)_this->extraInfo);
+    uint32_t key = myData2022->key;
+    int32_t offset = 0;
+    int8_t initialToULength = _this->toULength;
+    char c;
+
+    value = VALID_NON_TERMINAL_2022;
+    while (*source < sourceLimit) {
+        c = *(*source)++;
+        _this->toUBytes[_this->toULength++]=(uint8_t)c;
+        value = getKey_2022(c,(int32_t *) &key, &offset);
+
+        switch (value){
+
+        case VALID_NON_TERMINAL_2022 :
+            /* continue with the loop */
+            break;
+
+        case VALID_TERMINAL_2022:
+            key = 0;
+            goto DONE;
+
+        case INVALID_2022:
+            goto DONE;
+
+        case VALID_MAYBE_TERMINAL_2022:
+#ifdef U_ENABLE_GENERIC_ISO_2022
+            /* ESC ( B is ambiguous only for ISO_2022 itself */
+            if(var == ISO_2022) {
+                /* discard toUBytes[] for ESC ( B because this sequence is correct and complete */
+                _this->toULength = 0;
+
+                /* TODO need to indicate that ESC ( B was seen; if failure, then need to replay from source or from MBCS-style replay */
+
+                /* continue with the loop */
+                value = VALID_NON_TERMINAL_2022;
+                break;
+            } else
+#endif
+            {
+                /* not ISO_2022 itself, finish here */
+                value = VALID_TERMINAL_2022;
+                key = 0;
+                goto DONE;
+            }
+        }
+    }
+
+DONE:
+    myData2022->key = key;
+
+    if (value == VALID_NON_TERMINAL_2022) {
+        /* indicate that the escape sequence is incomplete: key!=0 */
+        return;
+    } else if (value == INVALID_2022 ) {
+        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+    } else /* value == VALID_TERMINAL_2022 */ {
+        switch(var){
+#ifdef U_ENABLE_GENERIC_ISO_2022
+        case ISO_2022:
+        {
+            const char *chosenConverterName = escSeqStateTable_Result_2022[offset];
+            if(chosenConverterName == NULL) {
+                /* SS2 or SS3 */
+                *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                _this->toUCallbackReason = UCNV_UNASSIGNED;
+                return;
+            }
+
+            _this->mode = UCNV_SI;
+            ucnv_close(myData2022->currentConverter);
+            myData2022->currentConverter = myUConverter = ucnv_open(chosenConverterName, err);
+            if(U_SUCCESS(*err)) {
+                myUConverter->fromCharErrorBehaviour = UCNV_TO_U_CALLBACK_STOP;
+                _this->mode = UCNV_SO;
+            }
+            break;
+        }
+#endif
+        case ISO_2022_JP:
+            {
+                StateEnum tempState=(StateEnum)nextStateToUnicodeJP[offset];
+                switch(tempState) {
+                case INVALID_STATE:
+                    *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                    break;
+                case SS2_STATE:
+                    if(myData2022->toU2022State.cs[2]!=0) {
+                        if(myData2022->toU2022State.g<2) {
+                            myData2022->toU2022State.prevG=myData2022->toU2022State.g;
+                        }
+                        myData2022->toU2022State.g=2;
+                    } else {
+                        /* illegal to have SS2 before a matching designator */
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    }
+                    break;
+                /* case SS3_STATE: not used in ISO-2022-JP-x */
+                case ISO8859_1:
+                case ISO8859_7:
+                    if((jpCharsetMasks[myData2022->version] & CSM(tempState)) == 0) {
+                        *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                    } else {
+                        /* G2 charset for SS2 */
+                        myData2022->toU2022State.cs[2]=(int8_t)tempState;
+                    }
+                    break;
+                default:
+                    if((jpCharsetMasks[myData2022->version] & CSM(tempState)) == 0) {
+                        *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                    } else {
+                        /* G0 charset */
+                        myData2022->toU2022State.cs[0]=(int8_t)tempState;
+                    }
+                    break;
+                }
+            }
+            break;
+#if !UCONFIG_ONLY_HTML_CONVERSION
+        case ISO_2022_CN:
+            {
+                StateEnum tempState=(StateEnum)nextStateToUnicodeCN[offset];
+                switch(tempState) {
+                case INVALID_STATE:
+                    *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                    break;
+                case SS2_STATE:
+                    if(myData2022->toU2022State.cs[2]!=0) {
+                        if(myData2022->toU2022State.g<2) {
+                            myData2022->toU2022State.prevG=myData2022->toU2022State.g;
+                        }
+                        myData2022->toU2022State.g=2;
+                    } else {
+                        /* illegal to have SS2 before a matching designator */
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    }
+                    break;
+                case SS3_STATE:
+                    if(myData2022->toU2022State.cs[3]!=0) {
+                        if(myData2022->toU2022State.g<2) {
+                            myData2022->toU2022State.prevG=myData2022->toU2022State.g;
+                        }
+                        myData2022->toU2022State.g=3;
+                    } else {
+                        /* illegal to have SS3 before a matching designator */
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    }
+                    break;
+                case ISO_IR_165:
+                    if(myData2022->version==0) {
+                        *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                        break;
+                    }
+                    U_FALLTHROUGH;
+                case GB2312_1:
+                    U_FALLTHROUGH;
+                case CNS_11643_1:
+                    myData2022->toU2022State.cs[1]=(int8_t)tempState;
+                    break;
+                case CNS_11643_2:
+                    myData2022->toU2022State.cs[2]=(int8_t)tempState;
+                    break;
+                default:
+                    /* other CNS 11643 planes */
+                    if(myData2022->version==0) {
+                        *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+                    } else {
+                       myData2022->toU2022State.cs[3]=(int8_t)tempState;
+                    }
+                    break;
+                }
+            }
+            break;
+        case ISO_2022_KR:
+            if(offset==0x30){
+                /* nothing to be done, just accept this one escape sequence */
+            } else {
+                *err = U_UNSUPPORTED_ESCAPE_SEQUENCE;
+            }
+            break;
+#endif  // !UCONFIG_ONLY_HTML_CONVERSION
+
+        default:
+            *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+            break;
+        }
+    }
+    if(U_SUCCESS(*err)) {
+        _this->toULength = 0;
+    } else if(*err==U_ILLEGAL_ESCAPE_SEQUENCE) {
+        if(_this->toULength>1) {
+            /*
+             * Ticket 5691: consistent illegal sequences:
+             * - We include at least the first byte (ESC) in the illegal sequence.
+             * - If any of the non-initial bytes could be the start of a character,
+             *   we stop the illegal sequence before the first one of those.
+             *   In escape sequences, all following bytes are "printable", that is,
+             *   unless they are completely illegal (>7f in SBCS, outside 21..7e in DBCS),
+             *   they are valid single/lead bytes.
+             *   For simplicity, we always only report the initial ESC byte as the
+             *   illegal sequence and back out all other bytes we looked at.
+             */
+            /* Back out some bytes. */
+            int8_t backOutDistance=_this->toULength-1;
+            int8_t bytesFromThisBuffer=_this->toULength-initialToULength;
+            if(backOutDistance<=bytesFromThisBuffer) {
+                /* same as initialToULength<=1 */
+                *source-=backOutDistance;
+            } else {
+                /* Back out bytes from the previous buffer: Need to replay them. */
+                _this->preToULength=(int8_t)(bytesFromThisBuffer-backOutDistance);
+                /* same as -(initialToULength-1) */
+                /* preToULength is negative! */
+                uprv_memcpy(_this->preToU, _this->toUBytes+1, -_this->preToULength);
+                *source-=bytesFromThisBuffer;
+            }
+            _this->toULength=1;
+        }
+    } else if(*err==U_UNSUPPORTED_ESCAPE_SEQUENCE) {
+        _this->toUCallbackReason = UCNV_UNASSIGNED;
+    }
+}
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+/*Checks the characters of the buffer against valid 2022 escape sequences
+*if the match we return a pointer to the initial start of the sequence otherwise
+*we return sourceLimit
+*/
+/*for 2022 looks ahead in the stream
+ *to determine the longest possible convertible
+ *data stream
+ */
+static inline const char*
+getEndOfBuffer_2022(const char** source,
+                   const char* sourceLimit,
+                   UBool /*flush*/){
+
+    const char* mySource = *source;
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    if (*source >= sourceLimit)
+        return sourceLimit;
+
+    do{
+
+        if (*mySource == ESC_2022){
+            int8_t i;
+            int32_t key = 0;
+            int32_t offset;
+            UCNV_TableStates_2022 value = VALID_NON_TERMINAL_2022;
+
+            /* Kludge: I could not
+            * figure out the reason for validating an escape sequence
+            * twice - once here and once in changeState_2022().
+            * is it possible to have an ESC character in a ISO2022
+            * byte stream which is valid in a code page? Is it legal?
+            */
+            for (i=0;
+            (mySource+i < sourceLimit)&&(value == VALID_NON_TERMINAL_2022);
+            i++) {
+                value =  getKey_2022(*(mySource+i), &key, &offset);
+            }
+            if (value > 0 || *mySource==ESC_2022)
+                return mySource;
+
+            if ((value == VALID_NON_TERMINAL_2022)&&(!flush) )
+                return sourceLimit;
+        }
+    }while (++mySource < sourceLimit);
+
+    return sourceLimit;
+#else
+    while(mySource < sourceLimit && *mySource != ESC_2022) {
+        ++mySource;
+    }
+    return mySource;
+#endif
+}
+#endif
+
+/* This inline function replicates code in _MBCSFromUChar32() function in ucnvmbcs.c
+ * any future change in _MBCSFromUChar32() function should be reflected here.
+ * @return number of bytes in *value; negative number if fallback; 0 if no mapping
+ */
+static inline int32_t
+MBCS_FROM_UCHAR32_ISO2022(UConverterSharedData* sharedData,
+                                         UChar32 c,
+                                         uint32_t* value,
+                                         UBool useFallback,
+                                         int outputType)
+{
+    const int32_t *cx;
+    const uint16_t *table;
+    uint32_t stage2Entry;
+    uint32_t myValue;
+    int32_t length;
+    const uint8_t *p;
+    /*
+     * TODO(markus): Use and require new, faster MBCS conversion table structures.
+     * Use internal version of ucnv_open() that verifies that the new structures are available,
+     * else U_INTERNAL_PROGRAM_ERROR.
+     */
+    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+    if(c<0x10000 || (sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+        table=sharedData->mbcs.fromUnicodeTable;
+        stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+        /* get the bytes and the length for the output */
+        if(outputType==MBCS_OUTPUT_2){
+            myValue=MBCS_VALUE_2_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+            if(myValue<=0xff) {
+                length=1;
+            } else {
+                length=2;
+            }
+        } else /* outputType==MBCS_OUTPUT_3 */ {
+            p=MBCS_POINTER_3_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+            myValue=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+            if(myValue<=0xff) {
+                length=1;
+            } else if(myValue<=0xffff) {
+                length=2;
+            } else {
+                length=3;
+            }
+        }
+        /* is this code point assigned, or do we use fallbacks? */
+        if((stage2Entry&(1<<(16+(c&0xf))))!=0) {
+            /* assigned */
+            *value=myValue;
+            return length;
+        } else if(FROM_U_USE_FALLBACK(useFallback, c) && myValue!=0) {
+            /*
+             * We allow a 0 byte output if the "assigned" bit is set for this entry.
+             * There is no way with this data structure for fallback output
+             * to be a zero byte.
+             */
+            *value=myValue;
+            return -length;
+        }
+    }
+
+    cx=sharedData->mbcs.extIndexes;
+    if(cx!=NULL) {
+        return ucnv_extSimpleMatchFromU(cx, c, value, useFallback);
+    }
+
+    /* unassigned */
+    return 0;
+}
+
+/* This inline function replicates code in _MBCSSingleFromUChar32() function in ucnvmbcs.c
+ * any future change in _MBCSSingleFromUChar32() function should be reflected here.
+ * @param retval pointer to output byte
+ * @return 1 roundtrip byte  0 no mapping  -1 fallback byte
+ */
+static inline int32_t
+MBCS_SINGLE_FROM_UCHAR32(UConverterSharedData* sharedData,
+                                       UChar32 c,
+                                       uint32_t* retval,
+                                       UBool useFallback)
+{
+    const uint16_t *table;
+    int32_t value;
+    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+    if(c>=0x10000 && !(sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+        return 0;
+    }
+    /* convert the Unicode code point in c into codepage bytes (same as in _MBCSFromUnicodeWithOffsets) */
+    table=sharedData->mbcs.fromUnicodeTable;
+    /* get the byte for the output */
+    value=MBCS_SINGLE_RESULT_FROM_U(table, (uint16_t *)sharedData->mbcs.fromUnicodeBytes, c);
+    /* is this code point assigned, or do we use fallbacks? */
+    *retval=(uint32_t)(value&0xff);
+    if(value>=0xf00) {
+        return 1;  /* roundtrip */
+    } else if(useFallback ? value>=0x800 : value>=0xc00) {
+        return -1;  /* fallback taken */
+    } else {
+        return 0;  /* no mapping */
+    }
+}
+
+/*
+ * Check that the result is a 2-byte value with each byte in the range A1..FE
+ * (strict EUC DBCS) before accepting it and subtracting 0x80 from each byte
+ * to move it to the ISO 2022 range 21..7E.
+ * Return 0 if out of range.
+ */
+static inline uint32_t
+_2022FromGR94DBCS(uint32_t value) {
+    if( (uint16_t)(value - 0xa1a1) <= (0xfefe - 0xa1a1) &&
+        (uint8_t)(value - 0xa1) <= (0xfe - 0xa1)
+    ) {
+        return value - 0x8080;  /* shift down to 21..7e byte range */
+    } else {
+        return 0;  /* not valid for ISO 2022 */
+    }
+}
+
+#if 0 /* 5691: Call sites now check for validity. They can just += 0x8080 after that. */
+/*
+ * This method does the reverse of _2022FromGR94DBCS(). Given the 2022 code point, it returns the
+ * 2 byte value that is in the range A1..FE for each byte. Otherwise it returns the 2022 code point
+ * unchanged. 
+ */
+static inline uint32_t
+_2022ToGR94DBCS(uint32_t value) {
+    uint32_t returnValue = value + 0x8080;
+    if( (uint16_t)(returnValue - 0xa1a1) <= (0xfefe - 0xa1a1) &&
+        (uint8_t)(returnValue - 0xa1) <= (0xfe - 0xa1)) {
+        return returnValue;
+    } else {
+        return value;
+    }
+}
+#endif
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+
+/**********************************************************************************
+*  ISO-2022 Converter
+*
+*
+*/
+
+static void U_CALLCONV
+T_UConverter_toUnicode_ISO_2022_OFFSETS_LOGIC(UConverterToUnicodeArgs* args,
+                                                           UErrorCode* err){
+    const char* mySourceLimit, *realSourceLimit;
+    const char* sourceStart;
+    const UChar* myTargetStart;
+    UConverter* saveThis;
+    UConverterDataISO2022* myData;
+    int8_t length;
+
+    saveThis = args->converter;
+    myData=((UConverterDataISO2022*)(saveThis->extraInfo));
+
+    realSourceLimit = args->sourceLimit;
+    while (args->source < realSourceLimit) {
+        if(myData->key == 0) { /* are we in the middle of an escape sequence? */
+            /*Find the end of the buffer e.g : Next Escape Seq | end of Buffer*/
+            mySourceLimit = getEndOfBuffer_2022(&(args->source), realSourceLimit, args->flush);
+
+            if(args->source < mySourceLimit) {
+                if(myData->currentConverter==NULL) {
+                    myData->currentConverter = ucnv_open("ASCII",err);
+                    if(U_FAILURE(*err)){
+                        return;
+                    }
+
+                    myData->currentConverter->fromCharErrorBehaviour = UCNV_TO_U_CALLBACK_STOP;
+                    saveThis->mode = UCNV_SO;
+                }
+
+                /* convert to before the ESC or until the end of the buffer */
+                myData->isFirstBuffer=FALSE;
+                sourceStart = args->source;
+                myTargetStart = args->target;
+                args->converter = myData->currentConverter;
+                ucnv_toUnicode(args->converter,
+                    &args->target,
+                    args->targetLimit,
+                    &args->source,
+                    mySourceLimit,
+                    args->offsets,
+                    (UBool)(args->flush && mySourceLimit == realSourceLimit),
+                    err);
+                args->converter = saveThis;
+
+                if (*err == U_BUFFER_OVERFLOW_ERROR) {
+                    /* move the overflow buffer */
+                    length = saveThis->UCharErrorBufferLength = myData->currentConverter->UCharErrorBufferLength;
+                    myData->currentConverter->UCharErrorBufferLength = 0;
+                    if(length > 0) {
+                        uprv_memcpy(saveThis->UCharErrorBuffer,
+                                    myData->currentConverter->UCharErrorBuffer,
+                                    length*U_SIZEOF_UCHAR);
+                    }
+                    return;
+                }
+
+                /*
+                 * At least one of:
+                 * -Error while converting
+                 * -Done with entire buffer
+                 * -Need to write offsets or update the current offset
+                 *  (leave that up to the code in ucnv.c)
+                 *
+                 * or else we just stopped at an ESC byte and continue with changeState_2022()
+                 */
+                if (U_FAILURE(*err) ||
+                    (args->source == realSourceLimit) ||
+                    (args->offsets != NULL && (args->target != myTargetStart || args->source != sourceStart) ||
+                    (mySourceLimit < realSourceLimit && myData->currentConverter->toULength > 0))
+                ) {
+                    /* copy partial or error input for truncated detection and error handling */
+                    if(U_FAILURE(*err)) {
+                        length = saveThis->invalidCharLength = myData->currentConverter->invalidCharLength;
+                        if(length > 0) {
+                            uprv_memcpy(saveThis->invalidCharBuffer, myData->currentConverter->invalidCharBuffer, length);
+                        }
+                    } else {
+                        length = saveThis->toULength = myData->currentConverter->toULength;
+                        if(length > 0) {
+                            uprv_memcpy(saveThis->toUBytes, myData->currentConverter->toUBytes, length);
+                            if(args->source < mySourceLimit) {
+                                *err = U_TRUNCATED_CHAR_FOUND; /* truncated input before ESC */
+                            }
+                        }
+                    }
+                    return;
+                }
+            }
+        }
+
+        sourceStart = args->source;
+        changeState_2022(args->converter,
+               &(args->source),
+               realSourceLimit,
+               ISO_2022,
+               err);
+        if (U_FAILURE(*err) || (args->source != sourceStart && args->offsets != NULL)) {
+            /* let the ucnv.c code update its current offset */
+            return;
+        }
+    }
+}
+
+#endif
+
+/*
+ * To Unicode Callback helper function
+ */
+static void
+toUnicodeCallback(UConverter *cnv,
+                  const uint32_t sourceChar, const uint32_t targetUniChar,
+                  UErrorCode* err){
+    if(sourceChar>0xff){
+        cnv->toUBytes[0] = (uint8_t)(sourceChar>>8);
+        cnv->toUBytes[1] = (uint8_t)sourceChar;
+        cnv->toULength = 2;
+    }
+    else{
+        cnv->toUBytes[0] =(char) sourceChar;
+        cnv->toULength = 1;
+    }
+
+    if(targetUniChar == (missingCharMarker-1/*0xfffe*/)){
+        *err = U_INVALID_CHAR_FOUND;
+    }
+    else{
+        *err = U_ILLEGAL_CHAR_FOUND;
+    }
+}
+
+/**************************************ISO-2022-JP*************************************************/
+
+/************************************** IMPORTANT **************************************************
+* The UConverter_fromUnicode_ISO2022_JP converter does not use ucnv_fromUnicode() functions for SBCS,DBCS and
+* MBCS; instead, the values are obtained directly by calling _MBCSFromUChar32().
+* The converter iterates over each Unicode codepoint
+* to obtain the equivalent codepoints from the codepages supported. Since the source buffer is
+* processed one char at a time it would make sense to reduce the extra processing a canned converter
+* would do as far as possible.
+*
+* If the implementation of these macros or structure of sharedData struct change in the future, make
+* sure that ISO-2022 is also changed.
+***************************************************************************************************
+*/
+
+/***************************************************************************************************
+* Rules for ISO-2022-jp encoding
+* (i)   Escape sequences must be fully contained within a line they should not
+*       span new lines or CRs
+* (ii)  If the last character on a line is represented by two bytes then an ASCII or
+*       JIS-Roman character escape sequence should follow before the line terminates
+* (iii) If the first character on the line is represented by two bytes then a two
+*       byte character escape sequence should precede it
+* (iv)  If no escape sequence is encountered then the characters are ASCII
+* (v)   Latin(ISO-8859-1) and Greek(ISO-8859-7) characters must be designated to G2,
+*       and invoked with SS2 (ESC N).
+* (vi)  If there is any G0 designation in text, there must be a switch to
+*       ASCII or to JIS X 0201-Roman before a space character (but not
+*       necessarily before "ESC 4/14 2/0" or "ESC N ' '") or control
+*       characters such as tab or CRLF.
+* (vi)  Supported encodings:
+*          ASCII, JISX201, JISX208, JISX212, GB2312, KSC5601, ISO-8859-1,ISO-8859-7
+*
+*  source : RFC-1554
+*
+*          JISX201, JISX208,JISX212 : new .cnv data files created
+*          KSC5601 : alias to ibm-949 mapping table
+*          GB2312 : alias to ibm-1386 mapping table
+*          ISO-8859-1 : Algorithmic implemented as LATIN1 case
+*          ISO-8859-7 : alisas to ibm-9409 mapping table
+*/
+
+/* preference order of JP charsets */
+static const StateEnum jpCharsetPref[]={
+    ASCII,
+    JISX201,
+    ISO8859_1,
+    JISX208,
+    ISO8859_7,
+    JISX212,
+    GB2312,
+    KSC5601,
+    HWKANA_7BIT
+};
+
+/*
+ * The escape sequences must be in order of the enum constants like JISX201  = 3,
+ * not in order of jpCharsetPref[]!
+ */
+static const char escSeqChars[][6] ={
+    "\x1B\x28\x42",         /* <ESC>(B  ASCII       */
+    "\x1B\x2E\x41",         /* <ESC>.A  ISO-8859-1  */
+    "\x1B\x2E\x46",         /* <ESC>.F  ISO-8859-7  */
+    "\x1B\x28\x4A",         /* <ESC>(J  JISX-201    */
+    "\x1B\x24\x42",         /* <ESC>$B  JISX-208    */
+    "\x1B\x24\x28\x44",     /* <ESC>$(D JISX-212    */
+    "\x1B\x24\x41",         /* <ESC>$A  GB2312      */
+    "\x1B\x24\x28\x43",     /* <ESC>$(C KSC5601     */
+    "\x1B\x28\x49"          /* <ESC>(I  HWKANA_7BIT */
+
+};
+static  const int8_t escSeqCharsLen[] ={
+    3, /* length of <ESC>(B  ASCII       */
+    3, /* length of <ESC>.A  ISO-8859-1  */
+    3, /* length of <ESC>.F  ISO-8859-7  */
+    3, /* length of <ESC>(J  JISX-201    */
+    3, /* length of <ESC>$B  JISX-208    */
+    4, /* length of <ESC>$(D JISX-212    */
+    3, /* length of <ESC>$A  GB2312      */
+    4, /* length of <ESC>$(C KSC5601     */
+    3  /* length of <ESC>(I  HWKANA_7BIT */
+};
+
+/*
+* The iteration over various code pages works this way:
+* i)   Get the currentState from myConverterData->currentState
+* ii)  Check if the character is mapped to a valid character in the currentState
+*      Yes ->  a) set the initIterState to currentState
+*       b) remain in this state until an invalid character is found
+*      No  ->  a) go to the next code page and find the character
+* iii) Before changing the state increment the current state check if the current state
+*      is equal to the intitIteration state
+*      Yes ->  A character that cannot be represented in any of the supported encodings
+*       break and return a U_INVALID_CHARACTER error
+*      No  ->  Continue and find the character in next code page
+*
+*
+* TODO: Implement a priority technique where the users are allowed to set the priority of code pages
+*/
+
+/* Map 00..7F to Unicode according to JIS X 0201. */
+static inline uint32_t
+jisx201ToU(uint32_t value) {
+    if(value < 0x5c) {
+        return value;
+    } else if(value == 0x5c) {
+        return 0xa5;
+    } else if(value == 0x7e) {
+        return 0x203e;
+    } else /* value <= 0x7f */ {
+        return value;
+    }
+}
+
+/* Map Unicode to 00..7F according to JIS X 0201. Return U+FFFE if unmappable. */
+static inline uint32_t
+jisx201FromU(uint32_t value) {
+    if(value<=0x7f) {
+        if(value!=0x5c && value!=0x7e) {
+            return value;
+        }
+    } else if(value==0xa5) {
+        return 0x5c;
+    } else if(value==0x203e) {
+        return 0x7e;
+    }
+    return 0xfffe;
+}
+
+/*
+ * Take a valid Shift-JIS byte pair, check that it is in the range corresponding
+ * to JIS X 0208, and convert it to a pair of 21..7E bytes.
+ * Return 0 if the byte pair is out of range.
+ */
+static inline uint32_t
+_2022FromSJIS(uint32_t value) {
+    uint8_t trail;
+
+    if(value > 0xEFFC) {
+        return 0;  /* beyond JIS X 0208 */
+    }
+
+    trail = (uint8_t)value;
+
+    value &= 0xff00;  /* lead byte */
+    if(value <= 0x9f00) {
+        value -= 0x7000;
+    } else /* 0xe000 <= value <= 0xef00 */ {
+        value -= 0xb000;
+    }
+    value <<= 1;
+
+    if(trail <= 0x9e) {
+        value -= 0x100;
+        if(trail <= 0x7e) {
+            value |= trail - 0x1f;
+        } else {
+            value |= trail - 0x20;
+        }
+    } else /* trail <= 0xfc */ {
+        value |= trail - 0x7e;
+    }
+    return value;
+}
+
+/*
+ * Convert a pair of JIS X 0208 21..7E bytes to Shift-JIS.
+ * If either byte is outside 21..7E make sure that the result is not valid
+ * for Shift-JIS so that the converter catches it.
+ * Some invalid byte values already turn into equally invalid Shift-JIS
+ * byte values and need not be tested explicitly.
+ */
+static inline void
+_2022ToSJIS(uint8_t c1, uint8_t c2, char bytes[2]) {
+    if(c1&1) {
+        ++c1;
+        if(c2 <= 0x5f) {
+            c2 += 0x1f;
+        } else if(c2 <= 0x7e) {
+            c2 += 0x20;
+        } else {
+            c2 = 0;  /* invalid */
+        }
+    } else {
+        if((uint8_t)(c2-0x21) <= ((0x7e)-0x21)) {
+            c2 += 0x7e;
+        } else {
+            c2 = 0;  /* invalid */
+        }
+    }
+    c1 >>= 1;
+    if(c1 <= 0x2f) {
+        c1 += 0x70;
+    } else if(c1 <= 0x3f) {
+        c1 += 0xb0;
+    } else {
+        c1 = 0;  /* invalid */
+    }
+    bytes[0] = (char)c1;
+    bytes[1] = (char)c2;
+}
+
+/*
+ * JIS X 0208 has fallbacks from Unicode half-width Katakana to full-width (DBCS)
+ * Katakana.
+ * Now that we use a Shift-JIS table for JIS X 0208 we need to hardcode these fallbacks
+ * because Shift-JIS roundtrips half-width Katakana to single bytes.
+ * These were the only fallbacks in ICU's jisx-208.ucm file.
+ */
+static const uint16_t hwkana_fb[HWKANA_END - HWKANA_START + 1] = {
+    0x2123,  /* U+FF61 */
+    0x2156,
+    0x2157,
+    0x2122,
+    0x2126,
+    0x2572,
+    0x2521,
+    0x2523,
+    0x2525,
+    0x2527,
+    0x2529,
+    0x2563,
+    0x2565,
+    0x2567,
+    0x2543,
+    0x213C,  /* U+FF70 */
+    0x2522,
+    0x2524,
+    0x2526,
+    0x2528,
+    0x252A,
+    0x252B,
+    0x252D,
+    0x252F,
+    0x2531,
+    0x2533,
+    0x2535,
+    0x2537,
+    0x2539,
+    0x253B,
+    0x253D,
+    0x253F,  /* U+FF80 */
+    0x2541,
+    0x2544,
+    0x2546,
+    0x2548,
+    0x254A,
+    0x254B,
+    0x254C,
+    0x254D,
+    0x254E,
+    0x254F,
+    0x2552,
+    0x2555,
+    0x2558,
+    0x255B,
+    0x255E,
+    0x255F,  /* U+FF90 */
+    0x2560,
+    0x2561,
+    0x2562,
+    0x2564,
+    0x2566,
+    0x2568,
+    0x2569,
+    0x256A,
+    0x256B,
+    0x256C,
+    0x256D,
+    0x256F,
+    0x2573,
+    0x212B,
+    0x212C   /* U+FF9F */
+};
+
+static void U_CALLCONV
+UConverter_fromUnicode_ISO_2022_JP_OFFSETS_LOGIC(UConverterFromUnicodeArgs* args, UErrorCode* err) {
+    UConverter *cnv = args->converter;
+    UConverterDataISO2022 *converterData;
+    ISO2022State *pFromU2022State;
+    uint8_t *target = (uint8_t *) args->target;
+    const uint8_t *targetLimit = (const uint8_t *) args->targetLimit;
+    const UChar* source = args->source;
+    const UChar* sourceLimit = args->sourceLimit;
+    int32_t* offsets = args->offsets;
+    UChar32 sourceChar;
+    char buffer[8];
+    int32_t len, outLen;
+    int8_t choices[10];
+    int32_t choiceCount;
+    uint32_t targetValue = 0;
+    UBool useFallback;
+
+    int32_t i;
+    int8_t cs, g;
+
+    /* set up the state */
+    converterData     = (UConverterDataISO2022*)cnv->extraInfo;
+    pFromU2022State   = &converterData->fromU2022State;
+
+    choiceCount = 0;
+
+    /* check if the last codepoint of previous buffer was a lead surrogate*/
+    if((sourceChar = cnv->fromUChar32)!=0 && target< targetLimit) {
+        goto getTrail;
+    }
+
+    while(source < sourceLimit) {
+        if(target < targetLimit) {
+
+            sourceChar  = *(source++);
+            /*check if the char is a First surrogate*/
+            if(U16_IS_SURROGATE(sourceChar)) {
+                if(U16_IS_SURROGATE_LEAD(sourceChar)) {
+getTrail:
+                    /*look ahead to find the trail surrogate*/
+                    if(source < sourceLimit) {
+                        /* test the following code unit */
+                        UChar trail=(UChar) *source;
+                        if(U16_IS_TRAIL(trail)) {
+                            source++;
+                            sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
+                            cnv->fromUChar32=0x00;
+                            /* convert this supplementary code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *err=U_ILLEGAL_CHAR_FOUND;
+                            cnv->fromUChar32=sourceChar;
+                            break;
+                        }
+                    } else {
+                        /* no more input */
+                        cnv->fromUChar32=sourceChar;
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *err=U_ILLEGAL_CHAR_FOUND;
+                    cnv->fromUChar32=sourceChar;
+                    break;
+                }
+            }
+
+            /* do not convert SO/SI/ESC */
+            if(IS_2022_CONTROL(sourceChar)) {
+                /* callback(illegal) */
+                *err=U_ILLEGAL_CHAR_FOUND;
+                cnv->fromUChar32=sourceChar;
+                break;
+            }
+
+            /* do the conversion */
+
+            if(choiceCount == 0) {
+                uint16_t csm;
+
+                /*
+                 * The csm variable keeps track of which charsets are allowed
+                 * and not used yet while building the choices[].
+                 */
+                csm = jpCharsetMasks[converterData->version];
+                choiceCount = 0;
+
+                /* JIS7/8: try single-byte half-width Katakana before JISX208 */
+                if(converterData->version == 3 || converterData->version == 4) {
+                    choices[choiceCount++] = (int8_t)HWKANA_7BIT;
+                }
+                /* Do not try single-byte half-width Katakana for other versions. */
+                csm &= ~CSM(HWKANA_7BIT);
+
+                /* try the current G0 charset */
+                choices[choiceCount++] = cs = pFromU2022State->cs[0];
+                csm &= ~CSM(cs);
+
+                /* try the current G2 charset */
+                if((cs = pFromU2022State->cs[2]) != 0) {
+                    choices[choiceCount++] = cs;
+                    csm &= ~CSM(cs);
+                }
+
+                /* try all the other possible charsets */
+                for(i = 0; i < UPRV_LENGTHOF(jpCharsetPref); ++i) {
+                    cs = (int8_t)jpCharsetPref[i];
+                    if(CSM(cs) & csm) {
+                        choices[choiceCount++] = cs;
+                        csm &= ~CSM(cs);
+                    }
+                }
+            }
+
+            cs = g = 0;
+            /*
+             * len==0: no mapping found yet
+             * len<0: found a fallback result: continue looking for a roundtrip but no further fallbacks
+             * len>0: found a roundtrip result, done
+             */
+            len = 0;
+            /*
+             * We will turn off useFallback after finding a fallback,
+             * but we still get fallbacks from PUA code points as usual.
+             * Therefore, we will also need to check that we don't overwrite
+             * an early fallback with a later one.
+             */
+            useFallback = cnv->useFallback;
+
+            for(i = 0; i < choiceCount && len <= 0; ++i) {
+                uint32_t value;
+                int32_t len2;
+                int8_t cs0 = choices[i];
+                switch(cs0) {
+                case ASCII:
+                    if(sourceChar <= 0x7f) {
+                        targetValue = (uint32_t)sourceChar;
+                        len = 1;
+                        cs = cs0;
+                        g = 0;
+                    }
+                    break;
+                case ISO8859_1:
+                    if(GR96_START <= sourceChar && sourceChar <= GR96_END) {
+                        targetValue = (uint32_t)sourceChar - 0x80;
+                        len = 1;
+                        cs = cs0;
+                        g = 2;
+                    }
+                    break;
+                case HWKANA_7BIT:
+                    if((uint32_t)(sourceChar - HWKANA_START) <= (HWKANA_END - HWKANA_START)) {
+                        if(converterData->version==3) {
+                            /* JIS7: use G1 (SO) */
+                            /* Shift U+FF61..U+FF9F to bytes 21..5F. */
+                            targetValue = (uint32_t)(sourceChar - (HWKANA_START - 0x21));
+                            len = 1;
+                            pFromU2022State->cs[1] = cs = cs0; /* do not output an escape sequence */
+                            g = 1;
+                        } else if(converterData->version==4) {
+                            /* JIS8: use 8-bit bytes with any single-byte charset, see escape sequence output below */
+                            /* Shift U+FF61..U+FF9F to bytes A1..DF. */
+                            targetValue = (uint32_t)(sourceChar - (HWKANA_START - 0xa1));
+                            len = 1;
+
+                            cs = pFromU2022State->cs[0];
+                            if(IS_JP_DBCS(cs)) {
+                                /* switch from a DBCS charset to JISX201 */
+                                cs = (int8_t)JISX201;
+                            }
+                            /* else stay in the current G0 charset */
+                            g = 0;
+                        }
+                        /* else do not use HWKANA_7BIT with other versions */
+                    }
+                    break;
+                case JISX201:
+                    /* G0 SBCS */
+                    value = jisx201FromU(sourceChar);
+                    if(value <= 0x7f) {
+                        targetValue = value;
+                        len = 1;
+                        cs = cs0;
+                        g = 0;
+                        useFallback = FALSE;
+                    }
+                    break;
+                case JISX208:
+                    /* G0 DBCS from Shift-JIS table */
+                    len2 = MBCS_FROM_UCHAR32_ISO2022(
+                                converterData->myConverterArray[cs0],
+                                sourceChar, &value,
+                                useFallback, MBCS_OUTPUT_2);
+                    if(len2 == 2 || (len2 == -2 && len == 0)) {  /* only accept DBCS: abs(len)==2 */
+                        value = _2022FromSJIS(value);
+                        if(value != 0) {
+                            targetValue = value;
+                            len = len2;
+                            cs = cs0;
+                            g = 0;
+                            useFallback = FALSE;
+                        }
+                    } else if(len == 0 && useFallback &&
+                              (uint32_t)(sourceChar - HWKANA_START) <= (HWKANA_END - HWKANA_START)) {
+                        targetValue = hwkana_fb[sourceChar - HWKANA_START];
+                        len = -2;
+                        cs = cs0;
+                        g = 0;
+                        useFallback = FALSE;
+                    }
+                    break;
+                case ISO8859_7:
+                    /* G0 SBCS forced to 7-bit output */
+                    len2 = MBCS_SINGLE_FROM_UCHAR32(
+                                converterData->myConverterArray[cs0],
+                                sourceChar, &value,
+                                useFallback);
+                    if(len2 != 0 && !(len2 < 0 && len != 0) && GR96_START <= value && value <= GR96_END) {
+                        targetValue = value - 0x80;
+                        len = len2;
+                        cs = cs0;
+                        g = 2;
+                        useFallback = FALSE;
+                    }
+                    break;
+                default:
+                    /* G0 DBCS */
+                    len2 = MBCS_FROM_UCHAR32_ISO2022(
+                                converterData->myConverterArray[cs0],
+                                sourceChar, &value,
+                                useFallback, MBCS_OUTPUT_2);
+                    if(len2 == 2 || (len2 == -2 && len == 0)) {  /* only accept DBCS: abs(len)==2 */
+                        if(cs0 == KSC5601) {
+                            /*
+                             * Check for valid bytes for the encoding scheme.
+                             * This is necessary because the sub-converter (windows-949)
+                             * has a broader encoding scheme than is valid for 2022.
+                             */
+                            value = _2022FromGR94DBCS(value);
+                            if(value == 0) {
+                                break;
+                            }
+                        }
+                        targetValue = value;
+                        len = len2;
+                        cs = cs0;
+                        g = 0;
+                        useFallback = FALSE;
+                    }
+                    break;
+                }
+            }
+
+            if(len != 0) {
+                if(len < 0) {
+                    len = -len;  /* fallback */
+                }
+                outLen = 0; /* count output bytes */
+
+                /* write SI if necessary (only for JIS7) */
+                if(pFromU2022State->g == 1 && g == 0) {
+                    buffer[outLen++] = UCNV_SI;
+                    pFromU2022State->g = 0;
+                }
+
+                /* write the designation sequence if necessary */
+                if(cs != pFromU2022State->cs[g]) {
+                    int32_t escLen = escSeqCharsLen[cs];
+                    uprv_memcpy(buffer + outLen, escSeqChars[cs], escLen);
+                    outLen += escLen;
+                    pFromU2022State->cs[g] = cs;
+
+                    /* invalidate the choices[] */
+                    choiceCount = 0;
+                }
+
+                /* write the shift sequence if necessary */
+                if(g != pFromU2022State->g) {
+                    switch(g) {
+                    /* case 0 handled before writing escapes */
+                    case 1:
+                        buffer[outLen++] = UCNV_SO;
+                        pFromU2022State->g = 1;
+                        break;
+                    default: /* case 2 */
+                        buffer[outLen++] = 0x1b;
+                        buffer[outLen++] = 0x4e;
+                        break;
+                    /* no case 3: no SS3 in ISO-2022-JP-x */
+                    }
+                }
+
+                /* write the output bytes */
+                if(len == 1) {
+                    buffer[outLen++] = (char)targetValue;
+                } else /* len == 2 */ {
+                    buffer[outLen++] = (char)(targetValue >> 8);
+                    buffer[outLen++] = (char)targetValue;
+                }
+            } else {
+                /*
+                 * if we cannot find the character after checking all codepages
+                 * then this is an error
+                 */
+                *err = U_INVALID_CHAR_FOUND;
+                cnv->fromUChar32=sourceChar;
+                break;
+            }
+
+            if(sourceChar == CR || sourceChar == LF) {
+                /* reset the G2 state at the end of a line (conversion got us into ASCII or JISX201 already) */
+                pFromU2022State->cs[2] = 0;
+                choiceCount = 0;
+            }
+
+            /* output outLen>0 bytes in buffer[] */
+            if(outLen == 1) {
+                *target++ = buffer[0];
+                if(offsets) {
+                    *offsets++ = (int32_t)(source - args->source - 1); /* -1: known to be ASCII */
+                }
+            } else if(outLen == 2 && (target + 2) <= targetLimit) {
+                *target++ = buffer[0];
+                *target++ = buffer[1];
+                if(offsets) {
+                    int32_t sourceIndex = (int32_t)(source - args->source - U16_LENGTH(sourceChar));
+                    *offsets++ = sourceIndex;
+                    *offsets++ = sourceIndex;
+                }
+            } else {
+                fromUWriteUInt8(
+                    cnv,
+                    buffer, outLen,
+                    &target, (const char *)targetLimit,
+                    &offsets, (int32_t)(source - args->source - U16_LENGTH(sourceChar)),
+                    err);
+                if(U_FAILURE(*err)) {
+                    break;
+                }
+            }
+        } /* end if(myTargetIndex<myTargetLength) */
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+    }/* end while(mySourceIndex<mySourceLength) */
+
+    /*
+     * the end of the input stream and detection of truncated input
+     * are handled by the framework, but for ISO-2022-JP conversion
+     * we need to be in ASCII mode at the very end
+     *
+     * conditions:
+     *   successful
+     *   in SO mode or not in ASCII mode
+     *   end of input and no truncated input
+     */
+    if( U_SUCCESS(*err) &&
+        (pFromU2022State->g!=0 || pFromU2022State->cs[0]!=ASCII) &&
+        args->flush && source>=sourceLimit && cnv->fromUChar32==0
+    ) {
+        int32_t sourceIndex;
+
+        outLen = 0;
+
+        if(pFromU2022State->g != 0) {
+            buffer[outLen++] = UCNV_SI;
+            pFromU2022State->g = 0;
+        }
+
+        if(pFromU2022State->cs[0] != ASCII) {
+            int32_t escLen = escSeqCharsLen[ASCII];
+            uprv_memcpy(buffer + outLen, escSeqChars[ASCII], escLen);
+            outLen += escLen;
+            pFromU2022State->cs[0] = (int8_t)ASCII;
+        }
+
+        /* get the source index of the last input character */
+        /*
+         * TODO this would be simpler and more reliable if we used a pair
+         * of sourceIndex/prevSourceIndex like in ucnvmbcs.c
+         * so that we could simply use the prevSourceIndex here;
+         * this code gives an incorrect result for the rare case of an unmatched
+         * trail surrogate that is alone in the last buffer of the text stream
+         */
+        sourceIndex=(int32_t)(source-args->source);
+        if(sourceIndex>0) {
+            --sourceIndex;
+            if( U16_IS_TRAIL(args->source[sourceIndex]) &&
+                (sourceIndex==0 || U16_IS_LEAD(args->source[sourceIndex-1]))
+            ) {
+                --sourceIndex;
+            }
+        } else {
+            sourceIndex=-1;
+        }
+
+        fromUWriteUInt8(
+            cnv,
+            buffer, outLen,
+            &target, (const char *)targetLimit,
+            &offsets, sourceIndex,
+            err);
+    }
+
+    /*save the state and return */
+    args->source = source;
+    args->target = (char*)target;
+}
+
+/*************** to unicode *******************/
+
+static void U_CALLCONV
+UConverter_toUnicode_ISO_2022_JP_OFFSETS_LOGIC(UConverterToUnicodeArgs *args,
+                                               UErrorCode* err){
+    char tempBuf[2];
+    const char *mySource = (char *) args->source;
+    UChar *myTarget = args->target;
+    const char *mySourceLimit = args->sourceLimit;
+    uint32_t targetUniChar = 0x0000;
+    uint32_t mySourceChar = 0x0000;
+    uint32_t tmpSourceChar = 0x0000;
+    UConverterDataISO2022* myData;
+    ISO2022State *pToU2022State;
+    StateEnum cs;
+
+    myData=(UConverterDataISO2022*)(args->converter->extraInfo);
+    pToU2022State = &myData->toU2022State;
+
+    if(myData->key != 0) {
+        /* continue with a partial escape sequence */
+        goto escape;
+    } else if(args->converter->toULength == 1 && mySource < mySourceLimit && myTarget < args->targetLimit) {
+        /* continue with a partial double-byte character */
+        mySourceChar = args->converter->toUBytes[0];
+        args->converter->toULength = 0;
+        cs = (StateEnum)pToU2022State->cs[pToU2022State->g];
+        targetUniChar = missingCharMarker;
+        goto getTrailByte;
+    }
+
+    while(mySource < mySourceLimit){
+
+        targetUniChar =missingCharMarker;
+
+        if(myTarget < args->targetLimit){
+
+            mySourceChar= (unsigned char) *mySource++;
+
+            switch(mySourceChar) {
+            case UCNV_SI:
+                if(myData->version==3) {
+                    pToU2022State->g=0;
+                    continue;
+                } else {
+                    /* only JIS7 uses SI/SO, not ISO-2022-JP-x */
+                    myData->isEmptySegment = FALSE;	/* reset this, we have a different error */
+                    break;
+                }
+
+            case UCNV_SO:
+                if(myData->version==3) {
+                    /* JIS7: switch to G1 half-width Katakana */
+                    pToU2022State->cs[1] = (int8_t)HWKANA_7BIT;
+                    pToU2022State->g=1;
+                    continue;
+                } else {
+                    /* only JIS7 uses SI/SO, not ISO-2022-JP-x */
+                    myData->isEmptySegment = FALSE;	/* reset this, we have a different error */
+                    break;
+                }
+
+            case ESC_2022:
+                mySource--;
+escape:
+                {
+                    const char * mySourceBefore = mySource;
+                    int8_t toULengthBefore = args->converter->toULength;
+
+                    changeState_2022(args->converter,&(mySource),
+                        mySourceLimit, ISO_2022_JP,err);
+
+                    /* If in ISO-2022-JP only and we successully completed an escape sequence, but previous segment was empty, create an error */
+                    if(myData->version==0 && myData->key==0 && U_SUCCESS(*err) && myData->isEmptySegment) {
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                        args->converter->toUCallbackReason = UCNV_IRREGULAR;
+                        args->converter->toULength = (int8_t)(toULengthBefore + (mySource - mySourceBefore));
+                    }
+                }
+
+                /* invalid or illegal escape sequence */
+                if(U_FAILURE(*err)){
+                    args->target = myTarget;
+                    args->source = mySource;
+                    myData->isEmptySegment = FALSE;	/* Reset to avoid future spurious errors */
+                    return;
+                }
+                /* If we successfully completed an escape sequence, we begin a new segment, empty so far */
+                if(myData->key==0) {
+                    myData->isEmptySegment = TRUE;
+                }
+                continue;
+
+            /* ISO-2022-JP does not use single-byte (C1) SS2 and SS3 */
+
+            case CR:
+            case LF:
+                /* automatically reset to single-byte mode */
+                if((StateEnum)pToU2022State->cs[0] != ASCII && (StateEnum)pToU2022State->cs[0] != JISX201) {
+                    pToU2022State->cs[0] = (int8_t)ASCII;
+                }
+                pToU2022State->cs[2] = 0;
+                pToU2022State->g = 0;
+                U_FALLTHROUGH;
+            default:
+                /* convert one or two bytes */
+                myData->isEmptySegment = FALSE;
+                cs = (StateEnum)pToU2022State->cs[pToU2022State->g];
+                if( (uint8_t)(mySourceChar - 0xa1) <= (0xdf - 0xa1) && myData->version==4 &&
+                    !IS_JP_DBCS(cs)
+                ) {
+                    /* 8-bit halfwidth katakana in any single-byte mode for JIS8 */
+                    targetUniChar = mySourceChar + (HWKANA_START - 0xa1);
+
+                    /* return from a single-shift state to the previous one */
+                    if(pToU2022State->g >= 2) {
+                        pToU2022State->g=pToU2022State->prevG;
+                    }
+                } else switch(cs) {
+                case ASCII:
+                    if(mySourceChar <= 0x7f) {
+                        targetUniChar = mySourceChar;
+                    }
+                    break;
+                case ISO8859_1:
+                    if(mySourceChar <= 0x7f) {
+                        targetUniChar = mySourceChar + 0x80;
+                    }
+                    /* return from a single-shift state to the previous one */
+                    pToU2022State->g=pToU2022State->prevG;
+                    break;
+                case ISO8859_7:
+                    if(mySourceChar <= 0x7f) {
+                        /* convert mySourceChar+0x80 to use a normal 8-bit table */
+                        targetUniChar =
+                            _MBCS_SINGLE_SIMPLE_GET_NEXT_BMP(
+                                myData->myConverterArray[cs],
+                                mySourceChar + 0x80);
+                    }
+                    /* return from a single-shift state to the previous one */
+                    pToU2022State->g=pToU2022State->prevG;
+                    break;
+                case JISX201:
+                    if(mySourceChar <= 0x7f) {
+                        targetUniChar = jisx201ToU(mySourceChar);
+                    }
+                    break;
+                case HWKANA_7BIT:
+                    if((uint8_t)(mySourceChar - 0x21) <= (0x5f - 0x21)) {
+                        /* 7-bit halfwidth Katakana */
+                        targetUniChar = mySourceChar + (HWKANA_START - 0x21);
+                    }
+                    break;
+                default:
+                    /* G0 DBCS */
+                    if(mySource < mySourceLimit) {
+                        int leadIsOk, trailIsOk;
+                        uint8_t trailByte;
+getTrailByte:
+                        trailByte = (uint8_t)*mySource;
+                        /*
+                         * Ticket 5691: consistent illegal sequences:
+                         * - We include at least the first byte in the illegal sequence.
+                         * - If any of the non-initial bytes could be the start of a character,
+                         *   we stop the illegal sequence before the first one of those.
+                         *
+                         * In ISO-2022 DBCS, if the second byte is in the 21..7e range or is
+                         * an ESC/SO/SI, we report only the first byte as the illegal sequence.
+                         * Otherwise we convert or report the pair of bytes.
+                         */
+                        leadIsOk = (uint8_t)(mySourceChar - 0x21) <= (0x7e - 0x21);
+                        trailIsOk = (uint8_t)(trailByte - 0x21) <= (0x7e - 0x21);
+                        if (leadIsOk && trailIsOk) {
+                            ++mySource;
+                            tmpSourceChar = (mySourceChar << 8) | trailByte;
+                            if(cs == JISX208) {
+                                _2022ToSJIS((uint8_t)mySourceChar, trailByte, tempBuf);
+                                mySourceChar = tmpSourceChar;
+                            } else {
+                                /* Copy before we modify tmpSourceChar so toUnicodeCallback() sees the correct bytes. */
+                                mySourceChar = tmpSourceChar;
+                                if (cs == KSC5601) {
+                                    tmpSourceChar += 0x8080;  /* = _2022ToGR94DBCS(tmpSourceChar) */
+                                }
+                                tempBuf[0] = (char)(tmpSourceChar >> 8);
+                                tempBuf[1] = (char)(tmpSourceChar);
+                            }
+                            targetUniChar = ucnv_MBCSSimpleGetNextUChar(myData->myConverterArray[cs], tempBuf, 2, FALSE);
+                        } else if (!(trailIsOk || IS_2022_CONTROL(trailByte))) {
+                            /* report a pair of illegal bytes if the second byte is not a DBCS starter */
+                            ++mySource;
+                            /* add another bit so that the code below writes 2 bytes in case of error */
+                            mySourceChar = 0x10000 | (mySourceChar << 8) | trailByte;
+                        }
+                    } else {
+                        args->converter->toUBytes[0] = (uint8_t)mySourceChar;
+                        args->converter->toULength = 1;
+                        goto endloop;
+                    }
+                }  /* End of inner switch */
+                break;
+            }  /* End of outer switch */
+            if(targetUniChar < (missingCharMarker-1/*0xfffe*/)){
+                if(args->offsets){
+                    args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                }
+                *(myTarget++)=(UChar)targetUniChar;
+            }
+            else if(targetUniChar > missingCharMarker){
+                /* disassemble the surrogate pair and write to output*/
+                targetUniChar-=0x0010000;
+                *myTarget = (UChar)(0xd800+(UChar)(targetUniChar>>10));
+                if(args->offsets){
+                    args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                }
+                ++myTarget;
+                if(myTarget< args->targetLimit){
+                    *myTarget = (UChar)(0xdc00+(UChar)(targetUniChar&0x3ff));
+                    if(args->offsets){
+                        args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                    }
+                    ++myTarget;
+                }else{
+                    args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]=
+                                    (UChar)(0xdc00+(UChar)(targetUniChar&0x3ff));
+                }
+
+            }
+            else{
+                /* Call the callback function*/
+                toUnicodeCallback(args->converter,mySourceChar,targetUniChar,err);
+                break;
+            }
+        }
+        else{    /* goes with "if(myTarget < args->targetLimit)"  way up near top of function */
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+endloop:
+    args->target = myTarget;
+    args->source = mySource;
+}
+
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+/***************************************************************
+*   Rules for ISO-2022-KR encoding
+*   i) The KSC5601 designator sequence should appear only once in a file,
+*      at the begining of a line before any KSC5601 characters. This usually
+*      means that it appears by itself on the first line of the file
+*  ii) There are only 2 shifting sequences SO to shift into double byte mode
+*      and SI to shift into single byte mode
+*/
+static void U_CALLCONV
+UConverter_fromUnicode_ISO_2022_KR_OFFSETS_LOGIC_IBM(UConverterFromUnicodeArgs* args, UErrorCode* err){
+
+    UConverter* saveConv = args->converter;
+    UConverterDataISO2022 *myConverterData=(UConverterDataISO2022*)saveConv->extraInfo;
+    args->converter=myConverterData->currentConverter;
+
+    myConverterData->currentConverter->fromUChar32 = saveConv->fromUChar32;
+    ucnv_MBCSFromUnicodeWithOffsets(args,err);
+    saveConv->fromUChar32 = myConverterData->currentConverter->fromUChar32;
+
+    if(*err == U_BUFFER_OVERFLOW_ERROR) {
+        if(myConverterData->currentConverter->charErrorBufferLength > 0) {
+            uprv_memcpy(
+                saveConv->charErrorBuffer,
+                myConverterData->currentConverter->charErrorBuffer,
+                myConverterData->currentConverter->charErrorBufferLength);
+        }
+        saveConv->charErrorBufferLength = myConverterData->currentConverter->charErrorBufferLength;
+        myConverterData->currentConverter->charErrorBufferLength = 0;
+    }
+    args->converter=saveConv;
+}
+
+static void U_CALLCONV
+UConverter_fromUnicode_ISO_2022_KR_OFFSETS_LOGIC(UConverterFromUnicodeArgs* args, UErrorCode* err){
+
+    const UChar *source = args->source;
+    const UChar *sourceLimit = args->sourceLimit;
+    unsigned char *target = (unsigned char *) args->target;
+    unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    int32_t* offsets = args->offsets;
+    uint32_t targetByteUnit = 0x0000;
+    UChar32 sourceChar = 0x0000;
+    UBool isTargetByteDBCS;
+    UBool oldIsTargetByteDBCS;
+    UConverterDataISO2022 *converterData;
+    UConverterSharedData* sharedData;
+    UBool useFallback;
+    int32_t length =0;
+
+    converterData=(UConverterDataISO2022*)args->converter->extraInfo;
+    /* if the version is 1 then the user is requesting
+     * conversion with ibm-25546 pass the arguments to
+     * MBCS converter and return
+     */
+    if(converterData->version==1){
+        UConverter_fromUnicode_ISO_2022_KR_OFFSETS_LOGIC_IBM(args,err);
+        return;
+    }
+
+    /* initialize data */
+    sharedData = converterData->currentConverter->sharedData;
+    useFallback = args->converter->useFallback;
+    isTargetByteDBCS=(UBool)args->converter->fromUnicodeStatus;
+    oldIsTargetByteDBCS = isTargetByteDBCS;
+
+    isTargetByteDBCS   = (UBool) args->converter->fromUnicodeStatus;
+    if((sourceChar = args->converter->fromUChar32)!=0 && target <targetLimit) {
+        goto getTrail;
+    }
+    while(source < sourceLimit){
+
+        targetByteUnit = missingCharMarker;
+
+        if(target < (unsigned char*) args->targetLimit){
+            sourceChar = *source++;
+
+            /* do not convert SO/SI/ESC */
+            if(IS_2022_CONTROL(sourceChar)) {
+                /* callback(illegal) */
+                *err=U_ILLEGAL_CHAR_FOUND;
+                args->converter->fromUChar32=sourceChar;
+                break;
+            }
+
+            length = MBCS_FROM_UCHAR32_ISO2022(sharedData,sourceChar,&targetByteUnit,useFallback,MBCS_OUTPUT_2);
+            if(length < 0) {
+                length = -length;  /* fallback */
+            }
+            /* only DBCS or SBCS characters are expected*/
+            /* DB characters with high bit set to 1 are expected */
+            if( length > 2 || length==0 ||
+                (length == 1 && targetByteUnit > 0x7f) ||
+                (length == 2 &&
+                    ((uint16_t)(targetByteUnit - 0xa1a1) > (0xfefe - 0xa1a1) ||
+                    (uint8_t)(targetByteUnit - 0xa1) > (0xfe - 0xa1)))
+            ) {
+                targetByteUnit=missingCharMarker;
+            }
+            if (targetByteUnit != missingCharMarker){
+
+                oldIsTargetByteDBCS = isTargetByteDBCS;
+                isTargetByteDBCS = (UBool)(targetByteUnit>0x00FF);
+                  /* append the shift sequence */
+                if (oldIsTargetByteDBCS != isTargetByteDBCS ){
+
+                    if (isTargetByteDBCS)
+                        *target++ = UCNV_SO;
+                    else
+                        *target++ = UCNV_SI;
+                    if(offsets)
+                        *(offsets++) = (int32_t)(source - args->source-1);
+                }
+                /* write the targetUniChar  to target */
+                if(targetByteUnit <= 0x00FF){
+                    if( target < targetLimit){
+                        *(target++) = (unsigned char) targetByteUnit;
+                        if(offsets){
+                            *(offsets++) = (int32_t)(source - args->source-1);
+                        }
+
+                    }else{
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (unsigned char) (targetByteUnit);
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }else{
+                    if(target < targetLimit){
+                        *(target++) =(unsigned char) ((targetByteUnit>>8) -0x80);
+                        if(offsets){
+                            *(offsets++) = (int32_t)(source - args->source-1);
+                        }
+                        if(target < targetLimit){
+                            *(target++) =(unsigned char) (targetByteUnit -0x80);
+                            if(offsets){
+                                *(offsets++) = (int32_t)(source - args->source-1);
+                            }
+                        }else{
+                            args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (unsigned char) (targetByteUnit -0x80);
+                            *err = U_BUFFER_OVERFLOW_ERROR;
+                        }
+                    }else{
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (unsigned char) ((targetByteUnit>>8) -0x80);
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (unsigned char) (targetByteUnit-0x80);
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }
+
+            }
+            else{
+                /* oops.. the code point is unassingned
+                 * set the error and reason
+                 */
+
+                /*check if the char is a First surrogate*/
+                if(U16_IS_SURROGATE(sourceChar)) {
+                    if(U16_IS_SURROGATE_LEAD(sourceChar)) {
+getTrail:
+                        /*look ahead to find the trail surrogate*/
+                        if(source <  sourceLimit) {
+                            /* test the following code unit */
+                            UChar trail=(UChar) *source;
+                            if(U16_IS_TRAIL(trail)) {
+                                source++;
+                                sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
+                                *err = U_INVALID_CHAR_FOUND;
+                                /* convert this surrogate code point */
+                                /* exit this condition tree */
+                            } else {
+                                /* this is an unmatched lead code unit (1st surrogate) */
+                                /* callback(illegal) */
+                                *err=U_ILLEGAL_CHAR_FOUND;
+                            }
+                        } else {
+                            /* no more input */
+                            *err = U_ZERO_ERROR;
+                        }
+                    } else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err=U_ILLEGAL_CHAR_FOUND;
+                    }
+                } else {
+                    /* callback(unassigned) for a BMP code point */
+                    *err = U_INVALID_CHAR_FOUND;
+                }
+
+                args->converter->fromUChar32=sourceChar;
+                break;
+            }
+        } /* end if(myTargetIndex<myTargetLength) */
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+    }/* end while(mySourceIndex<mySourceLength) */
+
+    /*
+     * the end of the input stream and detection of truncated input
+     * are handled by the framework, but for ISO-2022-KR conversion
+     * we need to be in ASCII mode at the very end
+     *
+     * conditions:
+     *   successful
+     *   not in ASCII mode
+     *   end of input and no truncated input
+     */
+    if( U_SUCCESS(*err) &&
+        isTargetByteDBCS &&
+        args->flush && source>=sourceLimit && args->converter->fromUChar32==0
+    ) {
+        int32_t sourceIndex;
+
+        /* we are switching to ASCII */
+        isTargetByteDBCS=FALSE;
+
+        /* get the source index of the last input character */
+        /*
+         * TODO this would be simpler and more reliable if we used a pair
+         * of sourceIndex/prevSourceIndex like in ucnvmbcs.c
+         * so that we could simply use the prevSourceIndex here;
+         * this code gives an incorrect result for the rare case of an unmatched
+         * trail surrogate that is alone in the last buffer of the text stream
+         */
+        sourceIndex=(int32_t)(source-args->source);
+        if(sourceIndex>0) {
+            --sourceIndex;
+            if( U16_IS_TRAIL(args->source[sourceIndex]) &&
+                (sourceIndex==0 || U16_IS_LEAD(args->source[sourceIndex-1]))
+            ) {
+                --sourceIndex;
+            }
+        } else {
+            sourceIndex=-1;
+        }
+
+        fromUWriteUInt8(
+            args->converter,
+            SHIFT_IN_STR, 1,
+            &target, (const char *)targetLimit,
+            &offsets, sourceIndex,
+            err);
+    }
+
+    /*save the state and return */
+    args->source = source;
+    args->target = (char*)target;
+    args->converter->fromUnicodeStatus = (uint32_t)isTargetByteDBCS;
+}
+
+/************************ To Unicode ***************************************/
+
+static void U_CALLCONV
+UConverter_toUnicode_ISO_2022_KR_OFFSETS_LOGIC_IBM(UConverterToUnicodeArgs *args,
+                                                            UErrorCode* err){
+    char const* sourceStart;
+    UConverterDataISO2022* myData=(UConverterDataISO2022*)(args->converter->extraInfo);
+
+    UConverterToUnicodeArgs subArgs;
+    int32_t minArgsSize;
+
+    /* set up the subconverter arguments */
+    if(args->size<sizeof(UConverterToUnicodeArgs)) {
+        minArgsSize = args->size;
+    } else {
+        minArgsSize = (int32_t)sizeof(UConverterToUnicodeArgs);
+    }
+
+    uprv_memcpy(&subArgs, args, minArgsSize);
+    subArgs.size = (uint16_t)minArgsSize;
+    subArgs.converter = myData->currentConverter;
+
+    /* remember the original start of the input for offsets */
+    sourceStart = args->source;
+
+    if(myData->key != 0) {
+        /* continue with a partial escape sequence */
+        goto escape;
+    }
+
+    while(U_SUCCESS(*err) && args->source < args->sourceLimit) {
+        /*Find the end of the buffer e.g : Next Escape Seq | end of Buffer*/
+        subArgs.source = args->source;
+        subArgs.sourceLimit = getEndOfBuffer_2022(&(args->source), args->sourceLimit, args->flush);
+        if(subArgs.source != subArgs.sourceLimit) {
+            /*
+             * get the current partial byte sequence
+             *
+             * it needs to be moved between the public and the subconverter
+             * so that the conversion framework, which only sees the public
+             * converter, can handle truncated and illegal input etc.
+             */
+            if(args->converter->toULength > 0) {
+                uprv_memcpy(subArgs.converter->toUBytes, args->converter->toUBytes, args->converter->toULength);
+            }
+            subArgs.converter->toULength = args->converter->toULength;
+
+            /*
+             * Convert up to the end of the input, or to before the next escape character.
+             * Does not handle conversion extensions because the preToU[] state etc.
+             * is not copied.
+             */
+            ucnv_MBCSToUnicodeWithOffsets(&subArgs, err);
+
+            if(args->offsets != NULL && sourceStart != args->source) {
+                /* update offsets to base them on the actual start of the input */
+                int32_t *offsets = args->offsets;
+                UChar *target = args->target;
+                int32_t delta = (int32_t)(args->source - sourceStart);
+                while(target < subArgs.target) {
+                    if(*offsets >= 0) {
+                        *offsets += delta;
+                    }
+                    ++offsets;
+                    ++target;
+                }
+            }
+            args->source = subArgs.source;
+            args->target = subArgs.target;
+            args->offsets = subArgs.offsets;
+
+            /* copy input/error/overflow buffers */
+            if(subArgs.converter->toULength > 0) {
+                uprv_memcpy(args->converter->toUBytes, subArgs.converter->toUBytes, subArgs.converter->toULength);
+            }
+            args->converter->toULength = subArgs.converter->toULength;
+
+            if(*err == U_BUFFER_OVERFLOW_ERROR) {
+                if(subArgs.converter->UCharErrorBufferLength > 0) {
+                    uprv_memcpy(args->converter->UCharErrorBuffer, subArgs.converter->UCharErrorBuffer,
+                                subArgs.converter->UCharErrorBufferLength);
+                }
+                args->converter->UCharErrorBufferLength=subArgs.converter->UCharErrorBufferLength;
+                subArgs.converter->UCharErrorBufferLength = 0;
+            }
+        }
+
+        if (U_FAILURE(*err) || (args->source == args->sourceLimit)) {
+            return;
+        }
+
+escape:
+        changeState_2022(args->converter,
+               &(args->source),
+               args->sourceLimit,
+               ISO_2022_KR,
+               err);
+    }
+}
+
+static void U_CALLCONV
+UConverter_toUnicode_ISO_2022_KR_OFFSETS_LOGIC(UConverterToUnicodeArgs *args,
+                                                            UErrorCode* err){
+    char tempBuf[2];
+    const char *mySource = ( char *) args->source;
+    UChar *myTarget = args->target;
+    const char *mySourceLimit = args->sourceLimit;
+    UChar32 targetUniChar = 0x0000;
+    UChar mySourceChar = 0x0000;
+    UConverterDataISO2022* myData;
+    UConverterSharedData* sharedData ;
+    UBool useFallback;
+
+    myData=(UConverterDataISO2022*)(args->converter->extraInfo);
+    if(myData->version==1){
+        UConverter_toUnicode_ISO_2022_KR_OFFSETS_LOGIC_IBM(args,err);
+        return;
+    }
+
+    /* initialize state */
+    sharedData = myData->currentConverter->sharedData;
+    useFallback = args->converter->useFallback;
+
+    if(myData->key != 0) {
+        /* continue with a partial escape sequence */
+        goto escape;
+    } else if(args->converter->toULength == 1 && mySource < mySourceLimit && myTarget < args->targetLimit) {
+        /* continue with a partial double-byte character */
+        mySourceChar = args->converter->toUBytes[0];
+        args->converter->toULength = 0;
+        goto getTrailByte;
+    }
+
+    while(mySource< mySourceLimit){
+
+        if(myTarget < args->targetLimit){
+
+            mySourceChar= (unsigned char) *mySource++;
+
+            if(mySourceChar==UCNV_SI){
+                myData->toU2022State.g = 0;
+                if (myData->isEmptySegment) {
+                    myData->isEmptySegment = FALSE;	/* we are handling it, reset to avoid future spurious errors */
+                    *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    args->converter->toUCallbackReason = UCNV_IRREGULAR;
+                    args->converter->toUBytes[0] = (uint8_t)mySourceChar;
+                    args->converter->toULength = 1;
+                    args->target = myTarget;
+                    args->source = mySource;
+                    return;
+                }
+                /*consume the source */
+                continue;
+            }else if(mySourceChar==UCNV_SO){
+                myData->toU2022State.g = 1;
+                myData->isEmptySegment = TRUE;	/* Begin a new segment, empty so far */
+                /*consume the source */
+                continue;
+            }else if(mySourceChar==ESC_2022){
+                mySource--;
+escape:
+                myData->isEmptySegment = FALSE;	/* Any invalid ESC sequences will be detected separately, so just reset this */
+                changeState_2022(args->converter,&(mySource),
+                                mySourceLimit, ISO_2022_KR, err);
+                if(U_FAILURE(*err)){
+                    args->target = myTarget;
+                    args->source = mySource;
+                    return;
+                }
+                continue;
+            }
+
+            myData->isEmptySegment = FALSE;	/* Any invalid char errors will be detected separately, so just reset this */
+            if(myData->toU2022State.g == 1) {
+                if(mySource < mySourceLimit) {
+                    int leadIsOk, trailIsOk;
+                    uint8_t trailByte;
+getTrailByte:
+                    targetUniChar = missingCharMarker;
+                    trailByte = (uint8_t)*mySource;
+                    /*
+                     * Ticket 5691: consistent illegal sequences:
+                     * - We include at least the first byte in the illegal sequence.
+                     * - If any of the non-initial bytes could be the start of a character,
+                     *   we stop the illegal sequence before the first one of those.
+                     *
+                     * In ISO-2022 DBCS, if the second byte is in the 21..7e range or is
+                     * an ESC/SO/SI, we report only the first byte as the illegal sequence.
+                     * Otherwise we convert or report the pair of bytes.
+                     */
+                    leadIsOk = (uint8_t)(mySourceChar - 0x21) <= (0x7e - 0x21);
+                    trailIsOk = (uint8_t)(trailByte - 0x21) <= (0x7e - 0x21);
+                    if (leadIsOk && trailIsOk) {
+                        ++mySource;
+                        tempBuf[0] = (char)(mySourceChar + 0x80);
+                        tempBuf[1] = (char)(trailByte + 0x80);
+                        targetUniChar = ucnv_MBCSSimpleGetNextUChar(sharedData, tempBuf, 2, useFallback);
+                        mySourceChar = (mySourceChar << 8) | trailByte;
+                    } else if (!(trailIsOk || IS_2022_CONTROL(trailByte))) {
+                        /* report a pair of illegal bytes if the second byte is not a DBCS starter */
+                        ++mySource;
+                        /* add another bit so that the code below writes 2 bytes in case of error */
+                        mySourceChar = static_cast<UChar>(0x10000 | (mySourceChar << 8) | trailByte);
+                    }
+                } else {
+                    args->converter->toUBytes[0] = (uint8_t)mySourceChar;
+                    args->converter->toULength = 1;
+                    break;
+                }
+            }
+            else if(mySourceChar <= 0x7f) {
+                targetUniChar = ucnv_MBCSSimpleGetNextUChar(sharedData, mySource - 1, 1, useFallback);
+            } else {
+                targetUniChar = 0xffff;
+            }
+            if(targetUniChar < 0xfffe){
+                if(args->offsets) {
+                    args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                }
+                *(myTarget++)=(UChar)targetUniChar;
+            }
+            else {
+                /* Call the callback function*/
+                toUnicodeCallback(args->converter,mySourceChar,targetUniChar,err);
+                break;
+            }
+        }
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+    args->target = myTarget;
+    args->source = mySource;
+}
+
+/*************************** END ISO2022-KR *********************************/
+
+/*************************** ISO-2022-CN *********************************
+*
+* Rules for ISO-2022-CN Encoding:
+* i)   The designator sequence must appear once on a line before any instance
+*      of character set it designates.
+* ii)  If two lines contain characters from the same character set, both lines
+*      must include the designator sequence.
+* iii) Once the designator sequence is known, a shifting sequence has to be found
+*      to invoke the  shifting
+* iv)  All lines start in ASCII and end in ASCII.
+* v)   Four shifting sequences are employed for this purpose:
+*
+*      Sequcence   ASCII Eq    Charsets
+*      ----------  -------    ---------
+*      SI           <SI>        US-ASCII
+*      SO           <SO>        CNS-11643-1992 Plane 1, GB2312, ISO-IR-165
+*      SS2          <ESC>N      CNS-11643-1992 Plane 2
+*      SS3          <ESC>O      CNS-11643-1992 Planes 3-7
+*
+* vi)
+*      SOdesignator  : ESC "$" ")" finalchar_for_SO
+*      SS2designator : ESC "$" "*" finalchar_for_SS2
+*      SS3designator : ESC "$" "+" finalchar_for_SS3
+*
+*      ESC $ ) A       Indicates the bytes following SO are Chinese
+*       characters as defined in GB 2312-80, until
+*       another SOdesignation appears
+*
+*
+*      ESC $ ) E       Indicates the bytes following SO are as defined
+*       in ISO-IR-165 (for details, see section 2.1),
+*       until another SOdesignation appears
+*
+*      ESC $ ) G       Indicates the bytes following SO are as defined
+*       in CNS 11643-plane-1, until another
+*       SOdesignation appears
+*
+*      ESC $ * H       Indicates the two bytes immediately following
+*       SS2 is a Chinese character as defined in CNS
+*       11643-plane-2, until another SS2designation
+*       appears
+*       (Meaning <ESC>N must preceed every 2 byte
+*        sequence.)
+*
+*      ESC $ + I       Indicates the immediate two bytes following SS3
+*       is a Chinese character as defined in CNS
+*       11643-plane-3, until another SS3designation
+*       appears
+*       (Meaning <ESC>O must preceed every 2 byte
+*        sequence.)
+*
+*      ESC $ + J       Indicates the immediate two bytes following SS3
+*       is a Chinese character as defined in CNS
+*       11643-plane-4, until another SS3designation
+*       appears
+*       (In English: <ESC>O must preceed every 2 byte
+*        sequence.)
+*
+*      ESC $ + K       Indicates the immediate two bytes following SS3
+*       is a Chinese character as defined in CNS
+*       11643-plane-5, until another SS3designation
+*       appears
+*
+*      ESC $ + L       Indicates the immediate two bytes following SS3
+*       is a Chinese character as defined in CNS
+*       11643-plane-6, until another SS3designation
+*       appears
+*
+*      ESC $ + M       Indicates the immediate two bytes following SS3
+*       is a Chinese character as defined in CNS
+*       11643-plane-7, until another SS3designation
+*       appears
+*
+*       As in ISO-2022-CN, each line starts in ASCII, and ends in ASCII, and
+*       has its own designation information before any Chinese characters
+*       appear
+*
+*/
+
+/* The following are defined this way to make the strings truly readonly */
+static const char GB_2312_80_STR[] = "\x1B\x24\x29\x41";
+static const char ISO_IR_165_STR[] = "\x1B\x24\x29\x45";
+static const char CNS_11643_1992_Plane_1_STR[] = "\x1B\x24\x29\x47";
+static const char CNS_11643_1992_Plane_2_STR[] = "\x1B\x24\x2A\x48";
+static const char CNS_11643_1992_Plane_3_STR[] = "\x1B\x24\x2B\x49";
+static const char CNS_11643_1992_Plane_4_STR[] = "\x1B\x24\x2B\x4A";
+static const char CNS_11643_1992_Plane_5_STR[] = "\x1B\x24\x2B\x4B";
+static const char CNS_11643_1992_Plane_6_STR[] = "\x1B\x24\x2B\x4C";
+static const char CNS_11643_1992_Plane_7_STR[] = "\x1B\x24\x2B\x4D";
+
+/********************** ISO2022-CN Data **************************/
+static const char* const escSeqCharsCN[10] ={
+        SHIFT_IN_STR,                   /* 0 ASCII */
+        GB_2312_80_STR,                 /* 1 GB2312_1 */
+        ISO_IR_165_STR,                 /* 2 ISO_IR_165 */
+        CNS_11643_1992_Plane_1_STR,
+        CNS_11643_1992_Plane_2_STR,
+        CNS_11643_1992_Plane_3_STR,
+        CNS_11643_1992_Plane_4_STR,
+        CNS_11643_1992_Plane_5_STR,
+        CNS_11643_1992_Plane_6_STR,
+        CNS_11643_1992_Plane_7_STR
+};
+
+static void U_CALLCONV
+UConverter_fromUnicode_ISO_2022_CN_OFFSETS_LOGIC(UConverterFromUnicodeArgs* args, UErrorCode* err){
+    UConverter *cnv = args->converter;
+    UConverterDataISO2022 *converterData;
+    ISO2022State *pFromU2022State;
+    uint8_t *target = (uint8_t *) args->target;
+    const uint8_t *targetLimit = (const uint8_t *) args->targetLimit;
+    const UChar* source = args->source;
+    const UChar* sourceLimit = args->sourceLimit;
+    int32_t* offsets = args->offsets;
+    UChar32 sourceChar;
+    char buffer[8];
+    int32_t len;
+    int8_t choices[3];
+    int32_t choiceCount;
+    uint32_t targetValue = 0;
+    UBool useFallback;
+
+    /* set up the state */
+    converterData     = (UConverterDataISO2022*)cnv->extraInfo;
+    pFromU2022State   = &converterData->fromU2022State;
+
+    choiceCount = 0;
+
+    /* check if the last codepoint of previous buffer was a lead surrogate*/
+    if((sourceChar = cnv->fromUChar32)!=0 && target< targetLimit) {
+        goto getTrail;
+    }
+
+    while( source < sourceLimit){
+        if(target < targetLimit){
+
+            sourceChar  = *(source++);
+            /*check if the char is a First surrogate*/
+             if(U16_IS_SURROGATE(sourceChar)) {
+                if(U16_IS_SURROGATE_LEAD(sourceChar)) {
+getTrail:
+                    /*look ahead to find the trail surrogate*/
+                    if(source < sourceLimit) {
+                        /* test the following code unit */
+                        UChar trail=(UChar) *source;
+                        if(U16_IS_TRAIL(trail)) {
+                            source++;
+                            sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
+                            cnv->fromUChar32=0x00;
+                            /* convert this supplementary code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *err=U_ILLEGAL_CHAR_FOUND;
+                            cnv->fromUChar32=sourceChar;
+                            break;
+                        }
+                    } else {
+                        /* no more input */
+                        cnv->fromUChar32=sourceChar;
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *err=U_ILLEGAL_CHAR_FOUND;
+                    cnv->fromUChar32=sourceChar;
+                    break;
+                }
+            }
+
+            /* do the conversion */
+            if(sourceChar <= 0x007f ){
+                /* do not convert SO/SI/ESC */
+                if(IS_2022_CONTROL(sourceChar)) {
+                    /* callback(illegal) */
+                    *err=U_ILLEGAL_CHAR_FOUND;
+                    cnv->fromUChar32=sourceChar;
+                    break;
+                }
+
+                /* US-ASCII */
+                if(pFromU2022State->g == 0) {
+                    buffer[0] = (char)sourceChar;
+                    len = 1;
+                } else {
+                    buffer[0] = UCNV_SI;
+                    buffer[1] = (char)sourceChar;
+                    len = 2;
+                    pFromU2022State->g = 0;
+                    choiceCount = 0;
+                }
+                if(sourceChar == CR || sourceChar == LF) {
+                    /* reset the state at the end of a line */
+                    uprv_memset(pFromU2022State, 0, sizeof(ISO2022State));
+                    choiceCount = 0;
+                }
+            }
+            else{
+                /* convert U+0080..U+10ffff */
+                int32_t i;
+                int8_t cs, g;
+
+                if(choiceCount == 0) {
+                    /* try the current SO/G1 converter first */
+                    choices[0] = pFromU2022State->cs[1];
+
+                    /* default to GB2312_1 if none is designated yet */
+                    if(choices[0] == 0) {
+                        choices[0] = GB2312_1;
+                    }
+
+                    if(converterData->version == 0) {
+                        /* ISO-2022-CN */
+
+                        /* try the other SO/G1 converter; a CNS_11643_1 lookup may result in any plane */
+                        if(choices[0] == GB2312_1) {
+                            choices[1] = (int8_t)CNS_11643_1;
+                        } else {
+                            choices[1] = (int8_t)GB2312_1;
+                        }
+
+                        choiceCount = 2;
+                    } else if (converterData->version == 1) {
+                        /* ISO-2022-CN-EXT */
+
+                        /* try one of the other converters */
+                        switch(choices[0]) {
+                        case GB2312_1:
+                            choices[1] = (int8_t)CNS_11643_1;
+                            choices[2] = (int8_t)ISO_IR_165;
+                            break;
+                        case ISO_IR_165:
+                            choices[1] = (int8_t)GB2312_1;
+                            choices[2] = (int8_t)CNS_11643_1;
+                            break;
+                        default: /* CNS_11643_x */
+                            choices[1] = (int8_t)GB2312_1;
+                            choices[2] = (int8_t)ISO_IR_165;
+                            break;
+                        }
+
+                        choiceCount = 3;
+                    } else {
+                        choices[0] = (int8_t)CNS_11643_1;
+                        choices[1] = (int8_t)GB2312_1;
+                    }
+                }
+
+                cs = g = 0;
+                /*
+                 * len==0: no mapping found yet
+                 * len<0: found a fallback result: continue looking for a roundtrip but no further fallbacks
+                 * len>0: found a roundtrip result, done
+                 */
+                len = 0;
+                /*
+                 * We will turn off useFallback after finding a fallback,
+                 * but we still get fallbacks from PUA code points as usual.
+                 * Therefore, we will also need to check that we don't overwrite
+                 * an early fallback with a later one.
+                 */
+                useFallback = cnv->useFallback;
+
+                for(i = 0; i < choiceCount && len <= 0; ++i) {
+                    int8_t cs0 = choices[i];
+                    if(cs0 > 0) {
+                        uint32_t value;
+                        int32_t len2;
+                        if(cs0 >= CNS_11643_0) {
+                            len2 = MBCS_FROM_UCHAR32_ISO2022(
+                                        converterData->myConverterArray[CNS_11643],
+                                        sourceChar,
+                                        &value,
+                                        useFallback,
+                                        MBCS_OUTPUT_3);
+                            if(len2 == 3 || (len2 == -3 && len == 0)) {
+                                targetValue = value;
+                                cs = (int8_t)(CNS_11643_0 + (value >> 16) - 0x80);
+                                if(len2 >= 0) {
+                                    len = 2;
+                                } else {
+                                    len = -2;
+                                    useFallback = FALSE;
+                                }
+                                if(cs == CNS_11643_1) {
+                                    g = 1;
+                                } else if(cs == CNS_11643_2) {
+                                    g = 2;
+                                } else /* plane 3..7 */ if(converterData->version == 1) {
+                                    g = 3;
+                                } else {
+                                    /* ISO-2022-CN (without -EXT) does not support plane 3..7 */
+                                    len = 0;
+                                }
+                            }
+                        } else {
+                            /* GB2312_1 or ISO-IR-165 */
+                            U_ASSERT(cs0<UCNV_2022_MAX_CONVERTERS);
+                            len2 = MBCS_FROM_UCHAR32_ISO2022(
+                                        converterData->myConverterArray[cs0],
+                                        sourceChar,
+                                        &value,
+                                        useFallback,
+                                        MBCS_OUTPUT_2);
+                            if(len2 == 2 || (len2 == -2 && len == 0)) {
+                                targetValue = value;
+                                len = len2;
+                                cs = cs0;
+                                g = 1;
+                                useFallback = FALSE;
+                            }
+                        }
+                    }
+                }
+
+                if(len != 0) {
+                    len = 0; /* count output bytes; it must have been abs(len) == 2 */
+
+                    /* write the designation sequence if necessary */
+                    if(cs != pFromU2022State->cs[g]) {
+                        if(cs < CNS_11643) {
+                            uprv_memcpy(buffer, escSeqCharsCN[cs], 4);
+                        } else {
+                            U_ASSERT(cs >= CNS_11643_1);
+                            uprv_memcpy(buffer, escSeqCharsCN[CNS_11643 + (cs - CNS_11643_1)], 4);
+                        }
+                        len = 4;
+                        pFromU2022State->cs[g] = cs;
+                        if(g == 1) {
+                            /* changing the SO/G1 charset invalidates the choices[] */
+                            choiceCount = 0;
+                        }
+                    }
+
+                    /* write the shift sequence if necessary */
+                    if(g != pFromU2022State->g) {
+                        switch(g) {
+                        case 1:
+                            buffer[len++] = UCNV_SO;
+
+                            /* set the new state only if it is the locking shift SO/G1, not for SS2 or SS3 */
+                            pFromU2022State->g = 1;
+                            break;
+                        case 2:
+                            buffer[len++] = 0x1b;
+                            buffer[len++] = 0x4e;
+                            break;
+                        default: /* case 3 */
+                            buffer[len++] = 0x1b;
+                            buffer[len++] = 0x4f;
+                            break;
+                        }
+                    }
+
+                    /* write the two output bytes */
+                    buffer[len++] = (char)(targetValue >> 8);
+                    buffer[len++] = (char)targetValue;
+                } else {
+                    /* if we cannot find the character after checking all codepages
+                     * then this is an error
+                     */
+                    *err = U_INVALID_CHAR_FOUND;
+                    cnv->fromUChar32=sourceChar;
+                    break;
+                }
+            }
+
+            /* output len>0 bytes in buffer[] */
+            if(len == 1) {
+                *target++ = buffer[0];
+                if(offsets) {
+                    *offsets++ = (int32_t)(source - args->source - 1); /* -1: known to be ASCII */
+                }
+            } else if(len == 2 && (target + 2) <= targetLimit) {
+                *target++ = buffer[0];
+                *target++ = buffer[1];
+                if(offsets) {
+                    int32_t sourceIndex = (int32_t)(source - args->source - U16_LENGTH(sourceChar));
+                    *offsets++ = sourceIndex;
+                    *offsets++ = sourceIndex;
+                }
+            } else {
+                fromUWriteUInt8(
+                    cnv,
+                    buffer, len,
+                    &target, (const char *)targetLimit,
+                    &offsets, (int32_t)(source - args->source - U16_LENGTH(sourceChar)),
+                    err);
+                if(U_FAILURE(*err)) {
+                    break;
+                }
+            }
+        } /* end if(myTargetIndex<myTargetLength) */
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+    }/* end while(mySourceIndex<mySourceLength) */
+
+    /*
+     * the end of the input stream and detection of truncated input
+     * are handled by the framework, but for ISO-2022-CN conversion
+     * we need to be in ASCII mode at the very end
+     *
+     * conditions:
+     *   successful
+     *   not in ASCII mode
+     *   end of input and no truncated input
+     */
+    if( U_SUCCESS(*err) &&
+        pFromU2022State->g!=0 &&
+        args->flush && source>=sourceLimit && cnv->fromUChar32==0
+    ) {
+        int32_t sourceIndex;
+
+        /* we are switching to ASCII */
+        pFromU2022State->g=0;
+
+        /* get the source index of the last input character */
+        /*
+         * TODO this would be simpler and more reliable if we used a pair
+         * of sourceIndex/prevSourceIndex like in ucnvmbcs.c
+         * so that we could simply use the prevSourceIndex here;
+         * this code gives an incorrect result for the rare case of an unmatched
+         * trail surrogate that is alone in the last buffer of the text stream
+         */
+        sourceIndex=(int32_t)(source-args->source);
+        if(sourceIndex>0) {
+            --sourceIndex;
+            if( U16_IS_TRAIL(args->source[sourceIndex]) &&
+                (sourceIndex==0 || U16_IS_LEAD(args->source[sourceIndex-1]))
+            ) {
+                --sourceIndex;
+            }
+        } else {
+            sourceIndex=-1;
+        }
+
+        fromUWriteUInt8(
+            cnv,
+            SHIFT_IN_STR, 1,
+            &target, (const char *)targetLimit,
+            &offsets, sourceIndex,
+            err);
+    }
+
+    /*save the state and return */
+    args->source = source;
+    args->target = (char*)target;
+}
+
+
+static void U_CALLCONV
+UConverter_toUnicode_ISO_2022_CN_OFFSETS_LOGIC(UConverterToUnicodeArgs *args,
+                                               UErrorCode* err){
+    char tempBuf[3];
+    const char *mySource = (char *) args->source;
+    UChar *myTarget = args->target;
+    const char *mySourceLimit = args->sourceLimit;
+    uint32_t targetUniChar = 0x0000;
+    uint32_t mySourceChar = 0x0000;
+    UConverterDataISO2022* myData;
+    ISO2022State *pToU2022State;
+
+    myData=(UConverterDataISO2022*)(args->converter->extraInfo);
+    pToU2022State = &myData->toU2022State;
+
+    if(myData->key != 0) {
+        /* continue with a partial escape sequence */
+        goto escape;
+    } else if(args->converter->toULength == 1 && mySource < mySourceLimit && myTarget < args->targetLimit) {
+        /* continue with a partial double-byte character */
+        mySourceChar = args->converter->toUBytes[0];
+        args->converter->toULength = 0;
+        targetUniChar = missingCharMarker;
+        goto getTrailByte;
+    }
+
+    while(mySource < mySourceLimit){
+
+        targetUniChar =missingCharMarker;
+
+        if(myTarget < args->targetLimit){
+
+            mySourceChar= (unsigned char) *mySource++;
+
+            switch(mySourceChar){
+            case UCNV_SI:
+                pToU2022State->g=0;
+                if (myData->isEmptySegment) {
+                    myData->isEmptySegment = FALSE;	/* we are handling it, reset to avoid future spurious errors */
+                    *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    args->converter->toUCallbackReason = UCNV_IRREGULAR;
+                    args->converter->toUBytes[0] = static_cast<uint8_t>(mySourceChar);
+                    args->converter->toULength = 1;
+                    args->target = myTarget;
+                    args->source = mySource;
+                    return;
+                }
+                continue;
+
+            case UCNV_SO:
+                if(pToU2022State->cs[1] != 0) {
+                    pToU2022State->g=1;
+                    myData->isEmptySegment = TRUE;	/* Begin a new segment, empty so far */
+                    continue;
+                } else {
+                    /* illegal to have SO before a matching designator */
+                    myData->isEmptySegment = FALSE;	/* Handling a different error, reset this to avoid future spurious errs */
+                    break;
+                }
+
+            case ESC_2022:
+                mySource--;
+escape:
+                {
+                    const char * mySourceBefore = mySource;
+                    int8_t toULengthBefore = args->converter->toULength;
+
+                    changeState_2022(args->converter,&(mySource),
+                        mySourceLimit, ISO_2022_CN,err);
+
+                    /* After SO there must be at least one character before a designator (designator error handled separately) */
+                    if(myData->key==0 && U_SUCCESS(*err) && myData->isEmptySegment) {
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                        args->converter->toUCallbackReason = UCNV_IRREGULAR;
+                        args->converter->toULength = (int8_t)(toULengthBefore + (mySource - mySourceBefore));
+                    }
+                }
+
+                /* invalid or illegal escape sequence */
+                if(U_FAILURE(*err)){
+                    args->target = myTarget;
+                    args->source = mySource;
+                    myData->isEmptySegment = FALSE;	/* Reset to avoid future spurious errors */
+                    return;
+                }
+                continue;
+
+            /* ISO-2022-CN does not use single-byte (C1) SS2 and SS3 */
+
+            case CR:
+            case LF:
+                uprv_memset(pToU2022State, 0, sizeof(ISO2022State));
+                U_FALLTHROUGH;
+            default:
+                /* convert one or two bytes */
+                myData->isEmptySegment = FALSE;
+                if(pToU2022State->g != 0) {
+                    if(mySource < mySourceLimit) {
+                        UConverterSharedData *cnv;
+                        StateEnum tempState;
+                        int32_t tempBufLen;
+                        int leadIsOk, trailIsOk;
+                        uint8_t trailByte;
+getTrailByte:
+                        trailByte = (uint8_t)*mySource;
+                        /*
+                         * Ticket 5691: consistent illegal sequences:
+                         * - We include at least the first byte in the illegal sequence.
+                         * - If any of the non-initial bytes could be the start of a character,
+                         *   we stop the illegal sequence before the first one of those.
+                         *
+                         * In ISO-2022 DBCS, if the second byte is in the 21..7e range or is
+                         * an ESC/SO/SI, we report only the first byte as the illegal sequence.
+                         * Otherwise we convert or report the pair of bytes.
+                         */
+                        leadIsOk = (uint8_t)(mySourceChar - 0x21) <= (0x7e - 0x21);
+                        trailIsOk = (uint8_t)(trailByte - 0x21) <= (0x7e - 0x21);
+                        if (leadIsOk && trailIsOk) {
+                            ++mySource;
+                            tempState = (StateEnum)pToU2022State->cs[pToU2022State->g];
+                            if(tempState >= CNS_11643_0) {
+                                cnv = myData->myConverterArray[CNS_11643];
+                                tempBuf[0] = (char) (0x80+(tempState-CNS_11643_0));
+                                tempBuf[1] = (char) (mySourceChar);
+                                tempBuf[2] = (char) trailByte;
+                                tempBufLen = 3;
+
+                            }else{
+                                U_ASSERT(tempState<UCNV_2022_MAX_CONVERTERS);
+                                cnv = myData->myConverterArray[tempState];
+                                tempBuf[0] = (char) (mySourceChar);
+                                tempBuf[1] = (char) trailByte;
+                                tempBufLen = 2;
+                            }
+                            targetUniChar = ucnv_MBCSSimpleGetNextUChar(cnv, tempBuf, tempBufLen, FALSE);
+                            mySourceChar = (mySourceChar << 8) | trailByte;
+                        } else if (!(trailIsOk || IS_2022_CONTROL(trailByte))) {
+                            /* report a pair of illegal bytes if the second byte is not a DBCS starter */
+                            ++mySource;
+                            /* add another bit so that the code below writes 2 bytes in case of error */
+                            mySourceChar = 0x10000 | (mySourceChar << 8) | trailByte;
+                        }
+                        if(pToU2022State->g>=2) {
+                            /* return from a single-shift state to the previous one */
+                            pToU2022State->g=pToU2022State->prevG;
+                        }
+                    } else {
+                        args->converter->toUBytes[0] = (uint8_t)mySourceChar;
+                        args->converter->toULength = 1;
+                        goto endloop;
+                    }
+                }
+                else{
+                    if(mySourceChar <= 0x7f) {
+                        targetUniChar = (UChar) mySourceChar;
+                    }
+                }
+                break;
+            }
+            if(targetUniChar < (missingCharMarker-1/*0xfffe*/)){
+                if(args->offsets){
+                    args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                }
+                *(myTarget++)=(UChar)targetUniChar;
+            }
+            else if(targetUniChar > missingCharMarker){
+                /* disassemble the surrogate pair and write to output*/
+                targetUniChar-=0x0010000;
+                *myTarget = (UChar)(0xd800+(UChar)(targetUniChar>>10));
+                if(args->offsets){
+                    args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                }
+                ++myTarget;
+                if(myTarget< args->targetLimit){
+                    *myTarget = (UChar)(0xdc00+(UChar)(targetUniChar&0x3ff));
+                    if(args->offsets){
+                        args->offsets[myTarget - args->target] = (int32_t)(mySource - args->source - (mySourceChar <= 0xff ? 1 : 2));
+                    }
+                    ++myTarget;
+                }else{
+                    args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]=
+                                    (UChar)(0xdc00+(UChar)(targetUniChar&0x3ff));
+                }
+
+            }
+            else{
+                /* Call the callback function*/
+                toUnicodeCallback(args->converter,mySourceChar,targetUniChar,err);
+                break;
+            }
+        }
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+endloop:
+    args->target = myTarget;
+    args->source = mySource;
+}
+#endif /* #if !UCONFIG_ONLY_HTML_CONVERSION */
+
+static void U_CALLCONV
+_ISO_2022_WriteSub(UConverterFromUnicodeArgs *args, int32_t offsetIndex, UErrorCode *err) {
+    UConverter *cnv = args->converter;
+    UConverterDataISO2022 *myConverterData=(UConverterDataISO2022 *) cnv->extraInfo;
+    ISO2022State *pFromU2022State=&myConverterData->fromU2022State;
+    char *p, *subchar;
+    char buffer[8];
+    int32_t length;
+
+    subchar=(char *)cnv->subChars;
+    length=cnv->subCharLen; /* assume length==1 for most variants */
+
+    p = buffer;
+    switch(myConverterData->locale[0]){
+    case 'j':
+        {
+            int8_t cs;
+
+            if(pFromU2022State->g == 1) {
+                /* JIS7: switch from G1 to G0 */
+                pFromU2022State->g = 0;
+                *p++ = UCNV_SI;
+            }
+
+            cs = pFromU2022State->cs[0];
+            if(cs != ASCII && cs != JISX201) {
+                /* not in ASCII or JIS X 0201: switch to ASCII */
+                pFromU2022State->cs[0] = (int8_t)ASCII;
+                *p++ = '\x1b';
+                *p++ = '\x28';
+                *p++ = '\x42';
+            }
+
+            *p++ = subchar[0];
+            break;
+        }
+    case 'c':
+        if(pFromU2022State->g != 0) {
+            /* not in ASCII mode: switch to ASCII */
+            pFromU2022State->g = 0;
+            *p++ = UCNV_SI;
+        }
+        *p++ = subchar[0];
+        break;
+    case 'k':
+        if(myConverterData->version == 0) {
+            if(length == 1) {
+                if(args->converter->fromUnicodeStatus) {
+                    /* in DBCS mode: switch to SBCS */
+                    args->converter->fromUnicodeStatus = 0;
+                    *p++ = UCNV_SI;
+                }
+                *p++ = subchar[0];
+            } else /* length == 2*/ {
+                if(!args->converter->fromUnicodeStatus) {
+                    /* in SBCS mode: switch to DBCS */
+                    args->converter->fromUnicodeStatus = 1;
+                    *p++ = UCNV_SO;
+                }
+                *p++ = subchar[0];
+                *p++ = subchar[1];
+            }
+            break;
+        } else {
+            /* save the subconverter's substitution string */
+            uint8_t *currentSubChars = myConverterData->currentConverter->subChars;
+            int8_t currentSubCharLen = myConverterData->currentConverter->subCharLen;
+
+            /* set our substitution string into the subconverter */
+            myConverterData->currentConverter->subChars = (uint8_t *)subchar;
+            myConverterData->currentConverter->subCharLen = (int8_t)length;
+
+            /* let the subconverter write the subchar, set/retrieve fromUChar32 state */
+            args->converter = myConverterData->currentConverter;
+            myConverterData->currentConverter->fromUChar32 = cnv->fromUChar32;
+            ucnv_cbFromUWriteSub(args, 0, err);
+            cnv->fromUChar32 = myConverterData->currentConverter->fromUChar32;
+            args->converter = cnv;
+
+            /* restore the subconverter's substitution string */
+            myConverterData->currentConverter->subChars = currentSubChars;
+            myConverterData->currentConverter->subCharLen = currentSubCharLen;
+
+            if(*err == U_BUFFER_OVERFLOW_ERROR) {
+                if(myConverterData->currentConverter->charErrorBufferLength > 0) {
+                    uprv_memcpy(
+                        cnv->charErrorBuffer,
+                        myConverterData->currentConverter->charErrorBuffer,
+                        myConverterData->currentConverter->charErrorBufferLength);
+                }
+                cnv->charErrorBufferLength = myConverterData->currentConverter->charErrorBufferLength;
+                myConverterData->currentConverter->charErrorBufferLength = 0;
+            }
+            return;
+        }
+    default:
+        /* not expected */
+        break;
+    }
+    ucnv_cbFromUWriteBytes(args,
+                           buffer, (int32_t)(p - buffer),
+                           offsetIndex, err);
+}
+
+/*
+ * Structure for cloning an ISO 2022 converter into a single memory block.
+ */
+struct cloneStruct
+{
+    UConverter cnv;
+    UConverter currentConverter;
+    UConverterDataISO2022 mydata;
+};
+
+
+U_CDECL_BEGIN
+
+static UConverter * U_CALLCONV
+_ISO_2022_SafeClone(
+            const UConverter *cnv,
+            void *stackBuffer,
+            int32_t *pBufferSize,
+            UErrorCode *status)
+{
+    struct cloneStruct * localClone;
+    UConverterDataISO2022 *cnvData;
+    int32_t i, size;
+
+    if (U_FAILURE(*status)){
+        return nullptr;
+    }
+
+    if (*pBufferSize == 0) { /* 'preflighting' request - set needed size into *pBufferSize */
+        *pBufferSize = (int32_t)sizeof(struct cloneStruct);
+        return NULL;
+    }
+
+    cnvData = (UConverterDataISO2022 *)cnv->extraInfo;
+    localClone = (struct cloneStruct *)stackBuffer;
+
+    /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
+
+    uprv_memcpy(&localClone->mydata, cnvData, sizeof(UConverterDataISO2022));
+    localClone->cnv.extraInfo = &localClone->mydata; /* set pointer to extra data */
+    localClone->cnv.isExtraLocal = TRUE;
+
+    /* share the subconverters */
+
+    if(cnvData->currentConverter != NULL) {
+        size = (int32_t)sizeof(UConverter);
+        localClone->mydata.currentConverter =
+            ucnv_safeClone(cnvData->currentConverter,
+                            &localClone->currentConverter,
+                            &size, status);
+        if(U_FAILURE(*status)) {
+            return NULL;
+        }
+    }
+
+    for(i=0; i<UCNV_2022_MAX_CONVERTERS; ++i) {
+        if(cnvData->myConverterArray[i] != NULL) {
+            ucnv_incrementRefCount(cnvData->myConverterArray[i]);
+        }
+    }
+
+    return &localClone->cnv;
+}
+
+U_CDECL_END
+
+static void U_CALLCONV
+_ISO_2022_GetUnicodeSet(const UConverter *cnv,
+                    const USetAdder *sa,
+                    UConverterUnicodeSet which,
+                    UErrorCode *pErrorCode)
+{
+    int32_t i;
+    UConverterDataISO2022* cnvData;
+
+    if (U_FAILURE(*pErrorCode)) {
+        return;
+    }
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    if (cnv->sharedData == &_ISO2022Data) {
+        /* We use UTF-8 in this case */
+        sa->addRange(sa->set, 0, 0xd7FF);
+        sa->addRange(sa->set, 0xE000, 0x10FFFF);
+        return;
+    }
+#endif
+
+    cnvData = (UConverterDataISO2022*)cnv->extraInfo;
+
+    /* open a set and initialize it with code points that are algorithmically round-tripped */
+    switch(cnvData->locale[0]){
+    case 'j':
+        /* include JIS X 0201 which is hardcoded */
+        sa->add(sa->set, 0xa5);
+        sa->add(sa->set, 0x203e);
+        if(jpCharsetMasks[cnvData->version]&CSM(ISO8859_1)) {
+            /* include Latin-1 for some variants of JP */
+            sa->addRange(sa->set, 0, 0xff);
+        } else {
+            /* include ASCII for JP */
+            sa->addRange(sa->set, 0, 0x7f);
+        }
+        if(cnvData->version==3 || cnvData->version==4 || which==UCNV_ROUNDTRIP_AND_FALLBACK_SET) {
+            /*
+             * Do not test (jpCharsetMasks[cnvData->version]&CSM(HWKANA_7BIT))!=0
+             * because the bit is on for all JP versions although only versions 3 & 4 (JIS7 & JIS8)
+             * use half-width Katakana.
+             * This is because all ISO-2022-JP variants are lenient in that they accept (in toUnicode)
+             * half-width Katakana via the ESC ( I sequence.
+             * However, we only emit (fromUnicode) half-width Katakana according to the
+             * definition of each variant.
+             *
+             * When including fallbacks,
+             * we need to include half-width Katakana Unicode code points for all JP variants because
+             * JIS X 0208 has hardcoded fallbacks for them (which map to full-width Katakana).
+             */
+            /* include half-width Katakana for JP */
+            sa->addRange(sa->set, HWKANA_START, HWKANA_END);
+        }
+        break;
+#if !UCONFIG_ONLY_HTML_CONVERSION
+    case 'c':
+    case 'z':
+        /* include ASCII for CN */
+        sa->addRange(sa->set, 0, 0x7f);
+        break;
+    case 'k':
+        /* there is only one converter for KR, and it is not in the myConverterArray[] */
+        cnvData->currentConverter->sharedData->impl->getUnicodeSet(
+                cnvData->currentConverter, sa, which, pErrorCode);
+        /* the loop over myConverterArray[] will simply not find another converter */
+        break;
+#endif
+    default:
+        break;
+    }
+
+#if 0  /* Replaced by ucnv_MBCSGetFilteredUnicodeSetForUnicode() until we implement ucnv_getUnicodeSet() with reverse fallbacks. */
+            if( (cnvData->locale[0]=='c' || cnvData->locale[0]=='z') &&
+                cnvData->version==0 && i==CNS_11643
+            ) {
+                /* special handling for non-EXT ISO-2022-CN: add only code points for CNS planes 1 and 2 */
+                ucnv_MBCSGetUnicodeSetForBytes(
+                        cnvData->myConverterArray[i],
+                        sa, UCNV_ROUNDTRIP_SET,
+                        0, 0x81, 0x82,
+                        pErrorCode);
+            }
+#endif
+
+    for (i=0; i<UCNV_2022_MAX_CONVERTERS; i++) {
+        UConverterSetFilter filter;
+        if(cnvData->myConverterArray[i]!=NULL) {
+            if(cnvData->locale[0]=='j' && i==JISX208) {
+                /*
+                 * Only add code points that map to Shift-JIS codes
+                 * corresponding to JIS X 0208.
+                 */
+                filter=UCNV_SET_FILTER_SJIS;
+#if !UCONFIG_ONLY_HTML_CONVERSION
+            } else if( (cnvData->locale[0]=='c' || cnvData->locale[0]=='z') &&
+                       cnvData->version==0 && i==CNS_11643) {
+                /*
+                 * Version-specific for CN:
+                 * CN version 0 does not map CNS planes 3..7 although
+                 * they are all available in the CNS conversion table;
+                 * CN version 1 (-EXT) does map them all.
+                 * The two versions create different Unicode sets.
+                 */
+                filter=UCNV_SET_FILTER_2022_CN;
+            } else if(i==KSC5601) {
+                /*
+                 * Some of the KSC 5601 tables (convrtrs.txt has this aliases on multiple tables)
+                 * are broader than GR94.
+                 */
+                filter=UCNV_SET_FILTER_GR94DBCS;
+#endif
+            } else {
+                filter=UCNV_SET_FILTER_NONE;
+            }
+            ucnv_MBCSGetFilteredUnicodeSetForUnicode(cnvData->myConverterArray[i], sa, which, filter, pErrorCode);
+        }
+    }
+
+    /*
+     * ISO 2022 converters must not convert SO/SI/ESC despite what
+     * sub-converters do by themselves.
+     * Remove these characters from the set.
+     */
+    sa->remove(sa->set, 0x0e);
+    sa->remove(sa->set, 0x0f);
+    sa->remove(sa->set, 0x1b);
+
+    /* ISO 2022 converters do not convert C1 controls either */
+    sa->removeRange(sa->set, 0x80, 0x9f);
+}
+
+static const UConverterImpl _ISO2022Impl={
+    UCNV_ISO_2022,
+
+    NULL,
+    NULL,
+
+    _ISO2022Open,
+    _ISO2022Close,
+    _ISO2022Reset,
+
+#ifdef U_ENABLE_GENERIC_ISO_2022
+    T_UConverter_toUnicode_ISO_2022_OFFSETS_LOGIC,
+    T_UConverter_toUnicode_ISO_2022_OFFSETS_LOGIC,
+    ucnv_fromUnicode_UTF8,
+    ucnv_fromUnicode_UTF8_OFFSETS_LOGIC,
+#else
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+#endif
+    NULL,
+
+    NULL,
+    _ISO2022getName,
+    _ISO_2022_WriteSub,
+    _ISO_2022_SafeClone,
+    _ISO_2022_GetUnicodeSet,
+
+    NULL,
+    NULL
+};
+static const UConverterStaticData _ISO2022StaticData={
+    sizeof(UConverterStaticData),
+    "ISO_2022",
+    2022,
+    UCNV_IBM,
+    UCNV_ISO_2022,
+    1,
+    3, /* max 3 bytes per UChar from UTF-8 (4 bytes from surrogate _pair_) */
+    { 0x1a, 0, 0, 0 },
+    1,
+    FALSE,
+    FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+const UConverterSharedData _ISO2022Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ISO2022StaticData, &_ISO2022Impl);
+
+/*************JP****************/
+static const UConverterImpl _ISO2022JPImpl={
+    UCNV_ISO_2022,
+
+    NULL,
+    NULL,
+
+    _ISO2022Open,
+    _ISO2022Close,
+    _ISO2022Reset,
+
+    UConverter_toUnicode_ISO_2022_JP_OFFSETS_LOGIC,
+    UConverter_toUnicode_ISO_2022_JP_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_JP_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_JP_OFFSETS_LOGIC,
+    NULL,
+
+    NULL,
+    _ISO2022getName,
+    _ISO_2022_WriteSub,
+    _ISO_2022_SafeClone,
+    _ISO_2022_GetUnicodeSet,
+
+    NULL,
+    NULL
+};
+static const UConverterStaticData _ISO2022JPStaticData={
+    sizeof(UConverterStaticData),
+    "ISO_2022_JP",
+    0,
+    UCNV_IBM,
+    UCNV_ISO_2022,
+    1,
+    6, /* max 6 bytes per UChar: 4-byte escape sequence + DBCS */
+    { 0x1a, 0, 0, 0 },
+    1,
+    FALSE,
+    FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+namespace {
+
+const UConverterSharedData _ISO2022JPData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ISO2022JPStaticData, &_ISO2022JPImpl);
+
+}  // namespace
+
+#if !UCONFIG_ONLY_HTML_CONVERSION
+/************* KR ***************/
+static const UConverterImpl _ISO2022KRImpl={
+    UCNV_ISO_2022,
+
+    NULL,
+    NULL,
+
+    _ISO2022Open,
+    _ISO2022Close,
+    _ISO2022Reset,
+
+    UConverter_toUnicode_ISO_2022_KR_OFFSETS_LOGIC,
+    UConverter_toUnicode_ISO_2022_KR_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_KR_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_KR_OFFSETS_LOGIC,
+    NULL,
+
+    NULL,
+    _ISO2022getName,
+    _ISO_2022_WriteSub,
+    _ISO_2022_SafeClone,
+    _ISO_2022_GetUnicodeSet,
+
+    NULL,
+    NULL
+};
+static const UConverterStaticData _ISO2022KRStaticData={
+    sizeof(UConverterStaticData),
+    "ISO_2022_KR",
+    0,
+    UCNV_IBM,
+    UCNV_ISO_2022,
+    1,
+    8, /* max 8 bytes per UChar */
+    { 0x1a, 0, 0, 0 },
+    1,
+    FALSE,
+    FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+namespace {
+
+const UConverterSharedData _ISO2022KRData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ISO2022KRStaticData, &_ISO2022KRImpl);
+
+}  // namespace
+
+/*************** CN ***************/
+static const UConverterImpl _ISO2022CNImpl={
+
+    UCNV_ISO_2022,
+
+    NULL,
+    NULL,
+
+    _ISO2022Open,
+    _ISO2022Close,
+    _ISO2022Reset,
+
+    UConverter_toUnicode_ISO_2022_CN_OFFSETS_LOGIC,
+    UConverter_toUnicode_ISO_2022_CN_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_CN_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISO_2022_CN_OFFSETS_LOGIC,
+    NULL,
+
+    NULL,
+    _ISO2022getName,
+    _ISO_2022_WriteSub,
+    _ISO_2022_SafeClone,
+    _ISO_2022_GetUnicodeSet,
+
+    NULL,
+    NULL
+};
+static const UConverterStaticData _ISO2022CNStaticData={
+    sizeof(UConverterStaticData),
+    "ISO_2022_CN",
+    0,
+    UCNV_IBM,
+    UCNV_ISO_2022,
+    1,
+    8, /* max 8 bytes per UChar: 4-byte CNS designator + 2 bytes for SS2/SS3 + DBCS */
+    { 0x1a, 0, 0, 0 },
+    1,
+    FALSE,
+    FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+namespace {
+
+const UConverterSharedData _ISO2022CNData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ISO2022CNStaticData, &_ISO2022CNImpl);
+
+}  // namespace
+#endif /* #if !UCONFIG_ONLY_HTML_CONVERSION */
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnv_bld.cpp b/thirdparty/icu4c/common/ucnv_bld.cpp
new file mode 100644
index 0000000000..0e198892f1
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_bld.cpp
@@ -0,0 +1,1689 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ ********************************************************************
+ * COPYRIGHT:
+ * Copyright (c) 1996-2016, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ ********************************************************************
+ *
+ *  ucnv_bld.cpp:
+ *
+ *  Defines functions that are used in the creation/initialization/deletion
+ *  of converters and related structures.
+ *  uses uconv_io.h routines to access disk information
+ *  is used by ucnv.h to implement public API create/delete/flushCache routines
+ * Modification History:
+ *
+ *   Date        Name        Description
+ *
+ *   06/20/2000  helena      OS/400 port changes; mostly typecast.
+ *   06/29/2000  helena      Major rewrite of the callback interface.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/putil.h"
+#include "unicode/udata.h"
+#include "unicode/ucnv.h"
+#include "unicode/uloc.h"
+#include "mutex.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include "utracimp.h"
+#include "ucnv_io.h"
+#include "ucnv_bld.h"
+#include "ucnvmbcs.h"
+#include "ucnv_ext.h"
+#include "ucnv_cnv.h"
+#include "ucnv_imp.h"
+#include "uhash.h"
+#include "umutex.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "ucln_cmn.h"
+#include "ustr_cnv.h"
+
+
+#if 0
+#include <stdio.h>
+extern void UCNV_DEBUG_LOG(char *what, char *who, void *p, int l);
+#define UCNV_DEBUG_LOG(x,y,z) UCNV_DEBUG_LOG(x,y,z,__LINE__)
+#else
+# define UCNV_DEBUG_LOG(x,y,z)
+#endif
+
+static const UConverterSharedData * const
+converterData[UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES]={
+    NULL, NULL,
+
+#if UCONFIG_NO_LEGACY_CONVERSION
+    NULL,
+#else
+    &_MBCSData,
+#endif
+
+    &_Latin1Data,
+    &_UTF8Data, &_UTF16BEData, &_UTF16LEData,
+#if UCONFIG_ONLY_HTML_CONVERSION
+    NULL, NULL,
+#else
+    &_UTF32BEData, &_UTF32LEData,
+#endif
+    NULL,
+
+#if UCONFIG_NO_LEGACY_CONVERSION
+    NULL,
+#else
+    &_ISO2022Data,
+#endif
+
+#if UCONFIG_NO_LEGACY_CONVERSION || UCONFIG_ONLY_HTML_CONVERSION
+    NULL, NULL, NULL, NULL, NULL, NULL,
+    NULL, NULL, NULL, NULL, NULL, NULL,
+    NULL,
+#else
+    &_LMBCSData1,&_LMBCSData2, &_LMBCSData3, &_LMBCSData4, &_LMBCSData5, &_LMBCSData6,
+    &_LMBCSData8,&_LMBCSData11,&_LMBCSData16,&_LMBCSData17,&_LMBCSData18,&_LMBCSData19,
+    &_HZData,
+#endif
+
+#if UCONFIG_ONLY_HTML_CONVERSION
+    NULL,
+#else
+    &_SCSUData,
+#endif
+
+
+#if UCONFIG_NO_LEGACY_CONVERSION || UCONFIG_ONLY_HTML_CONVERSION
+    NULL,
+#else
+    &_ISCIIData,
+#endif
+
+    &_ASCIIData,
+#if UCONFIG_ONLY_HTML_CONVERSION
+    NULL, NULL, &_UTF16Data, NULL, NULL, NULL,
+#else
+    &_UTF7Data, &_Bocu1Data, &_UTF16Data, &_UTF32Data, &_CESU8Data, &_IMAPData,
+#endif
+
+#if UCONFIG_NO_LEGACY_CONVERSION || UCONFIG_ONLY_HTML_CONVERSION
+    NULL,
+#else
+    &_CompoundTextData
+#endif
+};
+
+/* Please keep this in binary sorted order for getAlgorithmicTypeFromName.
+   Also the name should be in lower case and all spaces, dashes and underscores
+   removed
+*/
+static struct {
+  const char *name;
+  const UConverterType type;
+} const cnvNameType[] = {
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "bocu1", UCNV_BOCU1 },
+  { "cesu8", UCNV_CESU8 },
+#endif
+#if !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+  { "hz",UCNV_HZ },
+#endif
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "imapmailboxname", UCNV_IMAP_MAILBOX },
+#endif
+#if !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+  { "iscii", UCNV_ISCII },
+#endif
+#if !UCONFIG_NO_LEGACY_CONVERSION
+  { "iso2022", UCNV_ISO_2022 },
+#endif
+  { "iso88591", UCNV_LATIN_1 },
+#if !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+  { "lmbcs1", UCNV_LMBCS_1 },
+  { "lmbcs11",UCNV_LMBCS_11 },
+  { "lmbcs16",UCNV_LMBCS_16 },
+  { "lmbcs17",UCNV_LMBCS_17 },
+  { "lmbcs18",UCNV_LMBCS_18 },
+  { "lmbcs19",UCNV_LMBCS_19 },
+  { "lmbcs2", UCNV_LMBCS_2 },
+  { "lmbcs3", UCNV_LMBCS_3 },
+  { "lmbcs4", UCNV_LMBCS_4 },
+  { "lmbcs5", UCNV_LMBCS_5 },
+  { "lmbcs6", UCNV_LMBCS_6 },
+  { "lmbcs8", UCNV_LMBCS_8 },
+#endif
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "scsu", UCNV_SCSU },
+#endif
+  { "usascii", UCNV_US_ASCII },
+  { "utf16", UCNV_UTF16 },
+  { "utf16be", UCNV_UTF16_BigEndian },
+  { "utf16le", UCNV_UTF16_LittleEndian },
+#if U_IS_BIG_ENDIAN
+  { "utf16oppositeendian", UCNV_UTF16_LittleEndian },
+  { "utf16platformendian", UCNV_UTF16_BigEndian },
+#else
+  { "utf16oppositeendian", UCNV_UTF16_BigEndian},
+  { "utf16platformendian", UCNV_UTF16_LittleEndian },
+#endif
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "utf32", UCNV_UTF32 },
+  { "utf32be", UCNV_UTF32_BigEndian },
+  { "utf32le", UCNV_UTF32_LittleEndian },
+#if U_IS_BIG_ENDIAN
+  { "utf32oppositeendian", UCNV_UTF32_LittleEndian },
+  { "utf32platformendian", UCNV_UTF32_BigEndian },
+#else
+  { "utf32oppositeendian", UCNV_UTF32_BigEndian },
+  { "utf32platformendian", UCNV_UTF32_LittleEndian },
+#endif
+#endif
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "utf7", UCNV_UTF7 },
+#endif
+  { "utf8", UCNV_UTF8 },
+#if !UCONFIG_ONLY_HTML_CONVERSION
+  { "x11compoundtext", UCNV_COMPOUND_TEXT}
+#endif
+};
+
+
+/*initializes some global variables */
+static UHashtable *SHARED_DATA_HASHTABLE = NULL;
+static icu::UMutex cnvCacheMutex;
+/*  Note:  the global mutex is used for      */
+/*         reference count updates.          */
+
+static const char **gAvailableConverters = NULL;
+static uint16_t gAvailableConverterCount = 0;
+static icu::UInitOnce gAvailableConvertersInitOnce = U_INITONCE_INITIALIZER;
+
+#if !U_CHARSET_IS_UTF8
+
+/* This contains the resolved converter name. So no further alias lookup is needed again. */
+static char gDefaultConverterNameBuffer[UCNV_MAX_CONVERTER_NAME_LENGTH + 1]; /* +1 for NULL */
+static const char *gDefaultConverterName = NULL;
+
+/*
+If the default converter is an algorithmic converter, this is the cached value.
+We don't cache a full UConverter and clone it because ucnv_clone doesn't have
+less overhead than an algorithmic open. We don't cache non-algorithmic converters
+because ucnv_flushCache must be able to unload the default converter and its table.
+*/
+static const UConverterSharedData *gDefaultAlgorithmicSharedData = NULL;
+
+/* Does gDefaultConverterName have a converter option and require extra parsing? */
+static UBool gDefaultConverterContainsOption;
+
+#endif  /* !U_CHARSET_IS_UTF8 */
+
+static const char DATA_TYPE[] = "cnv";
+
+/* ucnv_flushAvailableConverterCache. This is only called from ucnv_cleanup().
+ *                       If it is ever to be called from elsewhere, synchronization
+ *                       will need to be considered.
+ */
+static void
+ucnv_flushAvailableConverterCache() {
+    gAvailableConverterCount = 0;
+    if (gAvailableConverters) {
+        uprv_free((char **)gAvailableConverters);
+        gAvailableConverters = NULL;
+    }
+    gAvailableConvertersInitOnce.reset();
+}
+
+/* ucnv_cleanup - delete all storage held by the converter cache, except any  */
+/*                in use by open converters.                                  */
+/*                Not thread safe.                                            */
+/*                Not supported API.                                          */
+static UBool U_CALLCONV ucnv_cleanup(void) {
+    ucnv_flushCache();
+    if (SHARED_DATA_HASHTABLE != NULL && uhash_count(SHARED_DATA_HASHTABLE) == 0) {
+        uhash_close(SHARED_DATA_HASHTABLE);
+        SHARED_DATA_HASHTABLE = NULL;
+    }
+
+    /* Isn't called from flushCache because other threads may have preexisting references to the table. */
+    ucnv_flushAvailableConverterCache();
+
+#if !U_CHARSET_IS_UTF8
+    gDefaultConverterName = NULL;
+    gDefaultConverterNameBuffer[0] = 0;
+    gDefaultConverterContainsOption = FALSE;
+    gDefaultAlgorithmicSharedData = NULL;
+#endif
+
+    return (SHARED_DATA_HASHTABLE == NULL);
+}
+
+U_CAPI void U_EXPORT2
+ucnv_enableCleanup(void) {
+    ucln_common_registerCleanup(UCLN_COMMON_UCNV, ucnv_cleanup);
+}
+
+static UBool U_CALLCONV
+isCnvAcceptable(void * /*context*/,
+                const char * /*type*/, const char * /*name*/,
+                const UDataInfo *pInfo) {
+    return (UBool)(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->sizeofUChar==U_SIZEOF_UCHAR &&
+        pInfo->dataFormat[0]==0x63 &&   /* dataFormat="cnvt" */
+        pInfo->dataFormat[1]==0x6e &&
+        pInfo->dataFormat[2]==0x76 &&
+        pInfo->dataFormat[3]==0x74 &&
+        pInfo->formatVersion[0]==6);  /* Everything will be version 6 */
+}
+
+/**
+ * Un flatten shared data from a UDATA..
+ */
+static UConverterSharedData*
+ucnv_data_unFlattenClone(UConverterLoadArgs *pArgs, UDataMemory *pData, UErrorCode *status)
+{
+    /* UDataInfo info; -- necessary only if some converters have different formatVersion */
+    const uint8_t *raw = (const uint8_t *)udata_getMemory(pData);
+    const UConverterStaticData *source = (const UConverterStaticData *) raw;
+    UConverterSharedData *data;
+    UConverterType type = (UConverterType)source->conversionType;
+
+    if(U_FAILURE(*status))
+        return NULL;
+
+    if( (uint16_t)type >= UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES ||
+        converterData[type] == NULL ||
+        !converterData[type]->isReferenceCounted ||
+        converterData[type]->referenceCounter != 1 ||
+        source->structSize != sizeof(UConverterStaticData))
+    {
+        *status = U_INVALID_TABLE_FORMAT;
+        return NULL;
+    }
+
+    data = (UConverterSharedData *)uprv_malloc(sizeof(UConverterSharedData));
+    if(data == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+
+    /* copy initial values from the static structure for this type */
+    uprv_memcpy(data, converterData[type], sizeof(UConverterSharedData));
+
+    data->staticData = source;
+
+    data->sharedDataCached = FALSE;
+
+    /* fill in fields from the loaded data */
+    data->dataMemory = (void*)pData; /* for future use */
+
+    if(data->impl->load != NULL) {
+        data->impl->load(data, pArgs, raw + source->structSize, status);
+        if(U_FAILURE(*status)) {
+            uprv_free(data);
+            return NULL;
+        }
+    }
+    return data;
+}
+
+/*Takes an alias name gets an actual converter file name
+ *goes to disk and opens it.
+ *allocates the memory and returns a new UConverter object
+ */
+static UConverterSharedData *createConverterFromFile(UConverterLoadArgs *pArgs, UErrorCode * err)
+{
+    UDataMemory *data;
+    UConverterSharedData *sharedData;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_LOAD);
+
+    if (U_FAILURE (*err)) {
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    UTRACE_DATA2(UTRACE_OPEN_CLOSE, "load converter %s from package %s", pArgs->name, pArgs->pkg);
+
+    data = udata_openChoice(pArgs->pkg, DATA_TYPE, pArgs->name, isCnvAcceptable, NULL, err);
+    if(U_FAILURE(*err))
+    {
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    sharedData = ucnv_data_unFlattenClone(pArgs, data, err);
+    if(U_FAILURE(*err))
+    {
+        udata_close(data);
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    /*
+     * TODO Store pkg in a field in the shared data so that delta-only converters
+     * can load base converters from the same package.
+     * If the pkg name is longer than the field, then either do not load the converter
+     * in the first place, or just set the pkg field to "".
+     */
+
+    UTRACE_EXIT_PTR_STATUS(sharedData, *err);
+    return sharedData;
+}
+
+/*returns a converter type from a string
+ */
+static const UConverterSharedData *
+getAlgorithmicTypeFromName(const char *realName)
+{
+    uint32_t mid, start, limit;
+    uint32_t lastMid;
+    int result;
+    char strippedName[UCNV_MAX_CONVERTER_NAME_LENGTH];
+
+    /* Lower case and remove ignoreable characters. */
+    ucnv_io_stripForCompare(strippedName, realName);
+
+    /* do a binary search for the alias */
+    start = 0;
+    limit = UPRV_LENGTHOF(cnvNameType);
+    mid = limit;
+    lastMid = UINT32_MAX;
+
+    for (;;) {
+        mid = (uint32_t)((start + limit) / 2);
+        if (lastMid == mid) {   /* Have we moved? */
+            break;  /* We haven't moved, and it wasn't found. */
+        }
+        lastMid = mid;
+        result = uprv_strcmp(strippedName, cnvNameType[mid].name);
+
+        if (result < 0) {
+            limit = mid;
+        } else if (result > 0) {
+            start = mid;
+        } else {
+            return converterData[cnvNameType[mid].type];
+        }
+    }
+
+    return NULL;
+}
+
+/*
+* Based on the number of known converters, this determines how many times larger
+* the shared data hash table should be. When on small platforms, or just a couple
+* of converters are used, this number should be 2. When memory is plentiful, or
+* when ucnv_countAvailable is ever used with a lot of available converters,
+* this should be 4.
+* Larger numbers reduce the number of hash collisions, but use more memory.
+*/
+#define UCNV_CACHE_LOAD_FACTOR 2
+
+/* Puts the shared data in the static hashtable SHARED_DATA_HASHTABLE */
+/*   Will always be called with the cnvCacheMutex alrady being held   */
+/*     by the calling function.                                       */
+/* Stores the shared data in the SHARED_DATA_HASHTABLE
+ * @param data The shared data
+ */
+static void
+ucnv_shareConverterData(UConverterSharedData * data)
+{
+    UErrorCode err = U_ZERO_ERROR;
+    /*Lazy evaluates the Hashtable itself */
+    /*void *sanity = NULL;*/
+
+    if (SHARED_DATA_HASHTABLE == NULL)
+    {
+        SHARED_DATA_HASHTABLE = uhash_openSize(uhash_hashChars, uhash_compareChars, NULL,
+                            ucnv_io_countKnownConverters(&err)*UCNV_CACHE_LOAD_FACTOR,
+                            &err);
+        ucnv_enableCleanup();
+
+        if (U_FAILURE(err))
+            return;
+    }
+
+    /* ### check to see if the element is not already there! */
+
+    /*
+    sanity =   ucnv_getSharedConverterData (data->staticData->name);
+    if(sanity != NULL)
+    {
+    UCNV_DEBUG_LOG("put:overwrite!",data->staticData->name,sanity);
+    }
+    UCNV_DEBUG_LOG("put:chk",data->staticData->name,sanity);
+    */
+
+    /* Mark it shared */
+    data->sharedDataCached = TRUE;
+
+    uhash_put(SHARED_DATA_HASHTABLE,
+            (void*) data->staticData->name, /* Okay to cast away const as long as
+            keyDeleter == NULL */
+            data,
+            &err);
+    UCNV_DEBUG_LOG("put", data->staticData->name,data);
+
+}
+
+/*  Look up a converter name in the shared data cache.                    */
+/*    cnvCacheMutex must be held by the caller to protect the hash table. */
+/* gets the shared data from the SHARED_DATA_HASHTABLE (might return NULL if it isn't there)
+ * @param name The name of the shared data
+ * @return the shared data from the SHARED_DATA_HASHTABLE
+ */
+static UConverterSharedData *
+ucnv_getSharedConverterData(const char *name)
+{
+    /*special case when no Table has yet been created we return NULL */
+    if (SHARED_DATA_HASHTABLE == NULL)
+    {
+        return NULL;
+    }
+    else
+    {
+        UConverterSharedData *rc;
+
+        rc = (UConverterSharedData*)uhash_get(SHARED_DATA_HASHTABLE, name);
+        UCNV_DEBUG_LOG("get",name,rc);
+        return rc;
+    }
+}
+
+/*frees the string of memory blocks associates with a sharedConverter
+ *if and only if the referenceCounter == 0
+ */
+/* Deletes (frees) the Shared data it's passed. first it checks the referenceCounter to
+ * see if anyone is using it, if not it frees all the memory stemming from sharedConverterData and
+ * returns TRUE,
+ * otherwise returns FALSE
+ * @param sharedConverterData The shared data
+ * @return if not it frees all the memory stemming from sharedConverterData and
+ * returns TRUE, otherwise returns FALSE
+ */
+static UBool
+ucnv_deleteSharedConverterData(UConverterSharedData * deadSharedData)
+{
+    UTRACE_ENTRY_OC(UTRACE_UCNV_UNLOAD);
+    UTRACE_DATA2(UTRACE_OPEN_CLOSE, "unload converter %s shared data %p", deadSharedData->staticData->name, deadSharedData);
+
+    if (deadSharedData->referenceCounter > 0) {
+        UTRACE_EXIT_VALUE((int32_t)FALSE);
+        return FALSE;
+    }
+
+    if (deadSharedData->impl->unload != NULL) {
+        deadSharedData->impl->unload(deadSharedData);
+    }
+
+    if(deadSharedData->dataMemory != NULL)
+    {
+        UDataMemory *data = (UDataMemory*)deadSharedData->dataMemory;
+        udata_close(data);
+    }
+
+    uprv_free(deadSharedData);
+
+    UTRACE_EXIT_VALUE((int32_t)TRUE);
+    return TRUE;
+}
+
+/**
+ * Load a non-algorithmic converter.
+ * If pkg==NULL, then this function must be called inside umtx_lock(&cnvCacheMutex).
+ */
+UConverterSharedData *
+ucnv_load(UConverterLoadArgs *pArgs, UErrorCode *err) {
+    UConverterSharedData *mySharedConverterData;
+
+    if(err == NULL || U_FAILURE(*err)) {
+        return NULL;
+    }
+
+    if(pArgs->pkg != NULL && *pArgs->pkg != 0) {
+        /* application-provided converters are not currently cached */
+        return createConverterFromFile(pArgs, err);
+    }
+
+    mySharedConverterData = ucnv_getSharedConverterData(pArgs->name);
+    if (mySharedConverterData == NULL)
+    {
+        /*Not cached, we need to stream it in from file */
+        mySharedConverterData = createConverterFromFile(pArgs, err);
+        if (U_FAILURE (*err) || (mySharedConverterData == NULL))
+        {
+            return NULL;
+        }
+        else if (!pArgs->onlyTestIsLoadable)
+        {
+            /* share it with other library clients */
+            ucnv_shareConverterData(mySharedConverterData);
+        }
+    }
+    else
+    {
+        /* The data for this converter was already in the cache.            */
+        /* Update the reference counter on the shared data: one more client */
+        mySharedConverterData->referenceCounter++;
+    }
+
+    return mySharedConverterData;
+}
+
+/**
+ * Unload a non-algorithmic converter.
+ * It must be sharedData->isReferenceCounted
+ * and this function must be called inside umtx_lock(&cnvCacheMutex).
+ */
+U_CAPI void
+ucnv_unload(UConverterSharedData *sharedData) {
+    if(sharedData != NULL) {
+        if (sharedData->referenceCounter > 0) {
+            sharedData->referenceCounter--;
+        }
+
+        if((sharedData->referenceCounter <= 0)&&(sharedData->sharedDataCached == FALSE)) {
+            ucnv_deleteSharedConverterData(sharedData);
+        }
+    }
+}
+
+U_CFUNC void
+ucnv_unloadSharedDataIfReady(UConverterSharedData *sharedData)
+{
+    if(sharedData != NULL && sharedData->isReferenceCounted) {
+        umtx_lock(&cnvCacheMutex);
+        ucnv_unload(sharedData);
+        umtx_unlock(&cnvCacheMutex);
+    }
+}
+
+U_CFUNC void
+ucnv_incrementRefCount(UConverterSharedData *sharedData)
+{
+    if(sharedData != NULL && sharedData->isReferenceCounted) {
+        umtx_lock(&cnvCacheMutex);
+        sharedData->referenceCounter++;
+        umtx_unlock(&cnvCacheMutex);
+    }
+}
+
+/*
+ * *pPieces must be initialized.
+ * The name without options will be copied to pPieces->cnvName.
+ * The locale and options will be copied to pPieces only if present in inName,
+ * otherwise the existing values in pPieces remain.
+ * *pArgs will be set to the pPieces values.
+ */
+static void
+parseConverterOptions(const char *inName,
+                      UConverterNamePieces *pPieces,
+                      UConverterLoadArgs *pArgs,
+                      UErrorCode *err)
+{
+    char *cnvName = pPieces->cnvName;
+    char c;
+    int32_t len = 0;
+
+    pArgs->name=inName;
+    pArgs->locale=pPieces->locale;
+    pArgs->options=pPieces->options;
+
+    /* copy the converter name itself to cnvName */
+    while((c=*inName)!=0 && c!=UCNV_OPTION_SEP_CHAR) {
+        if (++len>=UCNV_MAX_CONVERTER_NAME_LENGTH) {
+            *err = U_ILLEGAL_ARGUMENT_ERROR;    /* bad name */
+            pPieces->cnvName[0]=0;
+            return;
+        }
+        *cnvName++=c;
+        inName++;
+    }
+    *cnvName=0;
+    pArgs->name=pPieces->cnvName;
+
+    /* parse options. No more name copying should occur. */
+    while((c=*inName)!=0) {
+        if(c==UCNV_OPTION_SEP_CHAR) {
+            ++inName;
+        }
+
+        /* inName is behind an option separator */
+        if(uprv_strncmp(inName, "locale=", 7)==0) {
+            /* do not modify locale itself in case we have multiple locale options */
+            char *dest=pPieces->locale;
+
+            /* copy the locale option value */
+            inName+=7;
+            len=0;
+            while((c=*inName)!=0 && c!=UCNV_OPTION_SEP_CHAR) {
+                ++inName;
+
+                if(++len>=ULOC_FULLNAME_CAPACITY) {
+                    *err=U_ILLEGAL_ARGUMENT_ERROR;    /* bad name */
+                    pPieces->locale[0]=0;
+                    return;
+                }
+
+                *dest++=c;
+            }
+            *dest=0;
+        } else if(uprv_strncmp(inName, "version=", 8)==0) {
+            /* copy the version option value into bits 3..0 of pPieces->options */
+            inName+=8;
+            c=*inName;
+            if(c==0) {
+                pArgs->options=(pPieces->options&=~UCNV_OPTION_VERSION);
+                return;
+            } else if((uint8_t)(c-'0')<10) {
+                pArgs->options=pPieces->options=(pPieces->options&~UCNV_OPTION_VERSION)|(uint32_t)(c-'0');
+                ++inName;
+            }
+        } else if(uprv_strncmp(inName, "swaplfnl", 8)==0) {
+            inName+=8;
+            pArgs->options=(pPieces->options|=UCNV_OPTION_SWAP_LFNL);
+        /* add processing for new options here with another } else if(uprv_strncmp(inName, "option-name=", XX)==0) { */
+        } else {
+            /* ignore any other options until we define some */
+            while(((c = *inName++) != 0) && (c != UCNV_OPTION_SEP_CHAR)) {
+            }
+            if(c==0) {
+                return;
+            }
+        }
+    }
+}
+
+/*Logic determines if the converter is Algorithmic AND/OR cached
+ *depending on that:
+ * -we either go to get data from disk and cache it (Data=TRUE, Cached=False)
+ * -Get it from a Hashtable (Data=X, Cached=TRUE)
+ * -Call dataConverter initializer (Data=TRUE, Cached=TRUE)
+ * -Call AlgorithmicConverter initializer (Data=FALSE, Cached=TRUE)
+ */
+U_CFUNC UConverterSharedData *
+ucnv_loadSharedData(const char *converterName,
+                    UConverterNamePieces *pPieces,
+                    UConverterLoadArgs *pArgs,
+                    UErrorCode * err) {
+    UConverterNamePieces stackPieces;
+    UConverterLoadArgs stackArgs;
+    UConverterSharedData *mySharedConverterData = NULL;
+    UErrorCode internalErrorCode = U_ZERO_ERROR;
+    UBool mayContainOption = TRUE;
+    UBool checkForAlgorithmic = TRUE;
+
+    if (U_FAILURE (*err)) {
+        return NULL;
+    }
+
+    if(pPieces == NULL) {
+        if(pArgs != NULL) {
+            /*
+             * Bad: We may set pArgs pointers to stackPieces fields
+             * which will be invalid after this function returns.
+             */
+            *err = U_INTERNAL_PROGRAM_ERROR;
+            return NULL;
+        }
+        pPieces = &stackPieces;
+    }
+    if(pArgs == NULL) {
+        uprv_memset(&stackArgs, 0, sizeof(stackArgs));
+        stackArgs.size = (int32_t)sizeof(stackArgs);
+        pArgs = &stackArgs;
+    }
+
+    pPieces->cnvName[0] = 0;
+    pPieces->locale[0] = 0;
+    pPieces->options = 0;
+
+    pArgs->name = converterName;
+    pArgs->locale = pPieces->locale;
+    pArgs->options = pPieces->options;
+
+    /* In case "name" is NULL we want to open the default converter. */
+    if (converterName == NULL) {
+#if U_CHARSET_IS_UTF8
+        pArgs->name = "UTF-8";
+        return (UConverterSharedData *)converterData[UCNV_UTF8];
+#else
+        /* Call ucnv_getDefaultName first to query the name from the OS. */
+        pArgs->name = ucnv_getDefaultName();
+        if (pArgs->name == NULL) {
+            *err = U_MISSING_RESOURCE_ERROR;
+            return NULL;
+        }
+        mySharedConverterData = (UConverterSharedData *)gDefaultAlgorithmicSharedData;
+        checkForAlgorithmic = FALSE;
+        mayContainOption = gDefaultConverterContainsOption;
+        /* the default converter name is already canonical */
+#endif
+    }
+    else if(UCNV_FAST_IS_UTF8(converterName)) {
+        /* fastpath for UTF-8 */
+        pArgs->name = "UTF-8";
+        return (UConverterSharedData *)converterData[UCNV_UTF8];
+    }
+    else {
+        /* separate the converter name from the options */
+        parseConverterOptions(converterName, pPieces, pArgs, err);
+        if (U_FAILURE(*err)) {
+            /* Very bad name used. */
+            return NULL;
+        }
+
+        /* get the canonical converter name */
+        pArgs->name = ucnv_io_getConverterName(pArgs->name, &mayContainOption, &internalErrorCode);
+        if (U_FAILURE(internalErrorCode) || pArgs->name == NULL) {
+            /*
+            * set the input name in case the converter was added
+            * without updating the alias table, or when there is no alias table
+            */
+            pArgs->name = pPieces->cnvName;
+        } else if (internalErrorCode == U_AMBIGUOUS_ALIAS_WARNING) {
+            *err = U_AMBIGUOUS_ALIAS_WARNING;
+        }
+    }
+
+    /* separate the converter name from the options */
+    if(mayContainOption && pArgs->name != pPieces->cnvName) {
+        parseConverterOptions(pArgs->name, pPieces, pArgs, err);
+    }
+
+    /* get the shared data for an algorithmic converter, if it is one */
+    if (checkForAlgorithmic) {
+        mySharedConverterData = (UConverterSharedData *)getAlgorithmicTypeFromName(pArgs->name);
+    }
+    if (mySharedConverterData == NULL)
+    {
+        /* it is a data-based converter, get its shared data.               */
+        /* Hold the cnvCacheMutex through the whole process of checking the */
+        /*   converter data cache, and adding new entries to the cache      */
+        /*   to prevent other threads from modifying the cache during the   */
+        /*   process.                                                       */
+        pArgs->nestedLoads=1;
+        pArgs->pkg=NULL;
+
+        umtx_lock(&cnvCacheMutex);
+        mySharedConverterData = ucnv_load(pArgs, err);
+        umtx_unlock(&cnvCacheMutex);
+        if (U_FAILURE (*err) || (mySharedConverterData == NULL))
+        {
+            return NULL;
+        }
+    }
+
+    return mySharedConverterData;
+}
+
+U_CAPI UConverter *
+ucnv_createConverter(UConverter *myUConverter, const char *converterName, UErrorCode * err)
+{
+    UConverterNamePieces stackPieces;
+    UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+    UConverterSharedData *mySharedConverterData;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_OPEN);
+
+    if(U_SUCCESS(*err)) {
+        UTRACE_DATA1(UTRACE_OPEN_CLOSE, "open converter %s", converterName);
+
+        mySharedConverterData = ucnv_loadSharedData(converterName, &stackPieces, &stackArgs, err);
+
+        myUConverter = ucnv_createConverterFromSharedData(
+            myUConverter, mySharedConverterData,
+            &stackArgs,
+            err);
+
+        if(U_SUCCESS(*err)) {
+            UTRACE_EXIT_PTR_STATUS(myUConverter, *err);
+            return myUConverter;
+        }
+    }
+
+    /* exit with error */
+    UTRACE_EXIT_STATUS(*err);
+    return NULL;
+}
+
+U_CFUNC UBool
+ucnv_canCreateConverter(const char *converterName, UErrorCode *err) {
+    UConverter myUConverter;
+    UConverterNamePieces stackPieces;
+    UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+    UConverterSharedData *mySharedConverterData;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_OPEN);
+
+    if(U_SUCCESS(*err)) {
+        UTRACE_DATA1(UTRACE_OPEN_CLOSE, "test if can open converter %s", converterName);
+
+        stackArgs.onlyTestIsLoadable=TRUE;
+        mySharedConverterData = ucnv_loadSharedData(converterName, &stackPieces, &stackArgs, err);
+        ucnv_createConverterFromSharedData(
+            &myUConverter, mySharedConverterData,
+            &stackArgs,
+            err);
+        ucnv_unloadSharedDataIfReady(mySharedConverterData);
+    }
+
+    UTRACE_EXIT_STATUS(*err);
+    return U_SUCCESS(*err);
+}
+
+UConverter *
+ucnv_createAlgorithmicConverter(UConverter *myUConverter,
+                                UConverterType type,
+                                const char *locale, uint32_t options,
+                                UErrorCode *err) {
+    UConverter *cnv;
+    const UConverterSharedData *sharedData;
+    UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_OPEN_ALGORITHMIC);
+    UTRACE_DATA1(UTRACE_OPEN_CLOSE, "open algorithmic converter type %d", (int32_t)type);
+
+    if(type<0 || UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES<=type) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        UTRACE_EXIT_STATUS(U_ILLEGAL_ARGUMENT_ERROR);
+        return NULL;
+    }
+
+    sharedData = converterData[type];
+    if(sharedData == NULL || sharedData->isReferenceCounted) {
+        /* not a valid type, or not an algorithmic converter */
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        UTRACE_EXIT_STATUS(U_ILLEGAL_ARGUMENT_ERROR);
+        return NULL;
+    }
+
+    stackArgs.name = "";
+    stackArgs.options = options;
+    stackArgs.locale=locale;
+    cnv = ucnv_createConverterFromSharedData(
+            myUConverter, (UConverterSharedData *)sharedData,
+            &stackArgs, err);
+
+    UTRACE_EXIT_PTR_STATUS(cnv, *err);
+    return cnv;
+}
+
+U_CFUNC UConverter*
+ucnv_createConverterFromPackage(const char *packageName, const char *converterName, UErrorCode * err)
+{
+    UConverter *myUConverter;
+    UConverterSharedData *mySharedConverterData;
+    UConverterNamePieces stackPieces;
+    UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_OPEN_PACKAGE);
+
+    if(U_FAILURE(*err)) {
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    UTRACE_DATA2(UTRACE_OPEN_CLOSE, "open converter %s from package %s", converterName, packageName);
+
+    /* first, get the options out of the converterName string */
+    stackPieces.cnvName[0] = 0;
+    stackPieces.locale[0] = 0;
+    stackPieces.options = 0;
+    parseConverterOptions(converterName, &stackPieces, &stackArgs, err);
+    if (U_FAILURE(*err)) {
+        /* Very bad name used. */
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+    stackArgs.nestedLoads=1;
+    stackArgs.pkg=packageName;
+
+    /* open the data, unflatten the shared structure */
+    mySharedConverterData = createConverterFromFile(&stackArgs, err);
+
+    if (U_FAILURE(*err)) {
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    /* create the actual converter */
+    myUConverter = ucnv_createConverterFromSharedData(NULL, mySharedConverterData, &stackArgs, err);
+
+    if (U_FAILURE(*err)) {
+        ucnv_close(myUConverter);
+        UTRACE_EXIT_STATUS(*err);
+        return NULL;
+    }
+
+    UTRACE_EXIT_PTR_STATUS(myUConverter, *err);
+    return myUConverter;
+}
+
+
+U_CFUNC UConverter*
+ucnv_createConverterFromSharedData(UConverter *myUConverter,
+                                   UConverterSharedData *mySharedConverterData,
+                                   UConverterLoadArgs *pArgs,
+                                   UErrorCode *err)
+{
+    UBool isCopyLocal;
+
+    if(U_FAILURE(*err)) {
+        ucnv_unloadSharedDataIfReady(mySharedConverterData);
+        return myUConverter;
+    }
+    if(myUConverter == NULL)
+    {
+        myUConverter = (UConverter *) uprv_malloc (sizeof (UConverter));
+        if(myUConverter == NULL)
+        {
+            *err = U_MEMORY_ALLOCATION_ERROR;
+            ucnv_unloadSharedDataIfReady(mySharedConverterData);
+            return NULL;
+        }
+        isCopyLocal = FALSE;
+    } else {
+        isCopyLocal = TRUE;
+    }
+
+    /* initialize the converter */
+    uprv_memset(myUConverter, 0, sizeof(UConverter));
+    myUConverter->isCopyLocal = isCopyLocal;
+    /*myUConverter->isExtraLocal = FALSE;*/ /* Set by the memset call */
+    myUConverter->sharedData = mySharedConverterData;
+    myUConverter->options = pArgs->options;
+    if(!pArgs->onlyTestIsLoadable) {
+        myUConverter->preFromUFirstCP = U_SENTINEL;
+        myUConverter->fromCharErrorBehaviour = UCNV_TO_U_DEFAULT_CALLBACK;
+        myUConverter->fromUCharErrorBehaviour = UCNV_FROM_U_DEFAULT_CALLBACK;
+        myUConverter->toUnicodeStatus = mySharedConverterData->toUnicodeStatus;
+        myUConverter->maxBytesPerUChar = mySharedConverterData->staticData->maxBytesPerChar;
+        myUConverter->subChar1 = mySharedConverterData->staticData->subChar1;
+        myUConverter->subCharLen = mySharedConverterData->staticData->subCharLen;
+        myUConverter->subChars = (uint8_t *)myUConverter->subUChars;
+        uprv_memcpy(myUConverter->subChars, mySharedConverterData->staticData->subChar, myUConverter->subCharLen);
+        myUConverter->toUCallbackReason = UCNV_ILLEGAL; /* default reason to invoke (*fromCharErrorBehaviour) */
+    }
+
+    if(mySharedConverterData->impl->open != NULL) {
+        mySharedConverterData->impl->open(myUConverter, pArgs, err);
+        if(U_FAILURE(*err) && !pArgs->onlyTestIsLoadable) {
+            /* don't ucnv_close() if onlyTestIsLoadable because not fully initialized */
+            ucnv_close(myUConverter);
+            return NULL;
+        }
+    }
+
+    return myUConverter;
+}
+
+/*Frees all shared immutable objects that aren't referred to (reference count = 0)
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_flushCache ()
+{
+    UConverterSharedData *mySharedData = NULL;
+    int32_t pos;
+    int32_t tableDeletedNum = 0;
+    const UHashElement *e;
+    /*UErrorCode status = U_ILLEGAL_ARGUMENT_ERROR;*/
+    int32_t i, remaining;
+
+    UTRACE_ENTRY_OC(UTRACE_UCNV_FLUSH_CACHE);
+
+    /* Close the default converter without creating a new one so that everything will be flushed. */
+    u_flushDefaultConverter();
+
+    /*if shared data hasn't even been lazy evaluated yet
+    * return 0
+    */
+    if (SHARED_DATA_HASHTABLE == NULL) {
+        UTRACE_EXIT_VALUE((int32_t)0);
+        return 0;
+    }
+
+    /*creates an enumeration to iterate through every element in the
+    * table
+    *
+    * Synchronization:  holding cnvCacheMutex will prevent any other thread from
+    *                   accessing or modifying the hash table during the iteration.
+    *                   The reference count of an entry may be decremented by
+    *                   ucnv_close while the iteration is in process, but this is
+    *                   benign.  It can't be incremented (in ucnv_createConverter())
+    *                   because the sequence of looking up in the cache + incrementing
+    *                   is protected by cnvCacheMutex.
+    */
+    umtx_lock(&cnvCacheMutex);
+    /*
+     * double loop: A delta/extension-only converter has a pointer to its base table's
+     * shared data; the first iteration of the outer loop may see the delta converter
+     * before the base converter, and unloading the delta converter may get the base
+     * converter's reference counter down to 0.
+     */
+    i = 0;
+    do {
+        remaining = 0;
+        pos = UHASH_FIRST;
+        while ((e = uhash_nextElement (SHARED_DATA_HASHTABLE, &pos)) != NULL)
+        {
+            mySharedData = (UConverterSharedData *) e->value.pointer;
+            /*deletes only if reference counter == 0 */
+            if (mySharedData->referenceCounter == 0)
+            {
+                tableDeletedNum++;
+
+                UCNV_DEBUG_LOG("del",mySharedData->staticData->name,mySharedData);
+
+                uhash_removeElement(SHARED_DATA_HASHTABLE, e);
+                mySharedData->sharedDataCached = FALSE;
+                ucnv_deleteSharedConverterData (mySharedData);
+            } else {
+                ++remaining;
+            }
+        }
+    } while(++i == 1 && remaining > 0);
+    umtx_unlock(&cnvCacheMutex);
+
+    UTRACE_DATA1(UTRACE_INFO, "ucnv_flushCache() exits with %d converters remaining", remaining);
+
+    UTRACE_EXIT_VALUE(tableDeletedNum);
+    return tableDeletedNum;
+}
+
+/* available converters list --------------------------------------------------- */
+
+static void U_CALLCONV initAvailableConvertersList(UErrorCode &errCode) {
+    U_ASSERT(gAvailableConverterCount == 0);
+    U_ASSERT(gAvailableConverters == NULL);
+
+    ucnv_enableCleanup();
+    UEnumeration *allConvEnum = ucnv_openAllNames(&errCode);
+    int32_t allConverterCount = uenum_count(allConvEnum, &errCode);
+    if (U_FAILURE(errCode)) {
+        return;
+    }
+
+    /* We can't have more than "*converterTable" converters to open */
+    gAvailableConverters = (const char **) uprv_malloc(allConverterCount * sizeof(char*));
+    if (!gAvailableConverters) {
+        errCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    /* Open the default converter to make sure that it has first dibs in the hash table. */
+    UErrorCode localStatus = U_ZERO_ERROR;
+    UConverter tempConverter;
+    ucnv_close(ucnv_createConverter(&tempConverter, NULL, &localStatus));
+
+    gAvailableConverterCount = 0;
+
+    for (int32_t idx = 0; idx < allConverterCount; idx++) {
+        localStatus = U_ZERO_ERROR;
+        const char *converterName = uenum_next(allConvEnum, NULL, &localStatus);
+        if (ucnv_canCreateConverter(converterName, &localStatus)) {
+            gAvailableConverters[gAvailableConverterCount++] = converterName;
+        }
+    }
+
+    uenum_close(allConvEnum);
+}
+
+
+static UBool haveAvailableConverterList(UErrorCode *pErrorCode) {
+    umtx_initOnce(gAvailableConvertersInitOnce, &initAvailableConvertersList, *pErrorCode);
+    return U_SUCCESS(*pErrorCode);
+}
+
+U_CFUNC uint16_t
+ucnv_bld_countAvailableConverters(UErrorCode *pErrorCode) {
+    if (haveAvailableConverterList(pErrorCode)) {
+        return gAvailableConverterCount;
+    }
+    return 0;
+}
+
+U_CFUNC const char *
+ucnv_bld_getAvailableConverter(uint16_t n, UErrorCode *pErrorCode) {
+    if (haveAvailableConverterList(pErrorCode)) {
+        if (n < gAvailableConverterCount) {
+            return gAvailableConverters[n];
+        }
+        *pErrorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+    }
+    return NULL;
+}
+
+/* default converter name --------------------------------------------------- */
+
+#if !U_CHARSET_IS_UTF8
+/*
+Copy the canonical converter name.
+ucnv_getDefaultName must be thread safe, which can call this function.
+
+ucnv_setDefaultName calls this function and it doesn't have to be
+thread safe because there is no reliable/safe way to reset the
+converter in use in all threads. If you did reset the converter, you
+would not be sure that retrieving a default converter for one string
+would be the same type of default converter for a successive string.
+Since the name is a returned via ucnv_getDefaultName without copying,
+you shouldn't be modifying or deleting the string from a separate thread.
+*/
+static inline void
+internalSetName(const char *name, UErrorCode *status) {
+    UConverterNamePieces stackPieces;
+    UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+    int32_t length=(int32_t)(uprv_strlen(name));
+    UBool containsOption = (UBool)(uprv_strchr(name, UCNV_OPTION_SEP_CHAR) != NULL);
+    const UConverterSharedData *algorithmicSharedData;
+
+    stackArgs.name = name;
+    if(containsOption) {
+        stackPieces.cnvName[0] = 0;
+        stackPieces.locale[0] = 0;
+        stackPieces.options = 0;
+        parseConverterOptions(name, &stackPieces, &stackArgs, status);
+        if(U_FAILURE(*status)) {
+            return;
+        }
+    }
+    algorithmicSharedData = getAlgorithmicTypeFromName(stackArgs.name);
+
+    umtx_lock(&cnvCacheMutex);
+
+    gDefaultAlgorithmicSharedData = algorithmicSharedData;
+    gDefaultConverterContainsOption = containsOption;
+    uprv_memcpy(gDefaultConverterNameBuffer, name, length);
+    gDefaultConverterNameBuffer[length]=0;
+
+    /* gDefaultConverterName MUST be the last global var set by this function.  */
+    /*    It is the variable checked in ucnv_getDefaultName() to see if initialization is required. */
+    //    But there is nothing here preventing that from being reordered, either by the compiler
+    //             or hardware. I'm adding the mutex to ucnv_getDefaultName for now. UMTX_CHECK is not enough.
+    //             -- Andy
+    gDefaultConverterName = gDefaultConverterNameBuffer;
+
+    ucnv_enableCleanup();
+
+    umtx_unlock(&cnvCacheMutex);
+}
+#endif
+
+/*
+ * In order to be really thread-safe, the get function would have to take
+ * a buffer parameter and copy the current string inside a mutex block.
+ * This implementation only tries to be really thread-safe while
+ * setting the name.
+ * It assumes that setting a pointer is atomic.
+ */
+
+U_CAPI const char*  U_EXPORT2
+ucnv_getDefaultName() {
+#if U_CHARSET_IS_UTF8
+    return "UTF-8";
+#else
+    /* local variable to be thread-safe */
+    const char *name;
+
+    /*
+    Concurrent calls to ucnv_getDefaultName must be thread safe,
+    but ucnv_setDefaultName is not thread safe.
+    */
+    {
+        icu::Mutex lock(&cnvCacheMutex);
+        name = gDefaultConverterName;
+    }
+    if(name==NULL) {
+        UErrorCode errorCode = U_ZERO_ERROR;
+        UConverter *cnv = NULL;
+
+        name = uprv_getDefaultCodepage();
+
+        /* if the name is there, test it out and get the canonical name with options */
+        if(name != NULL) {
+            cnv = ucnv_open(name, &errorCode);
+            if(U_SUCCESS(errorCode) && cnv != NULL) {
+                name = ucnv_getName(cnv, &errorCode);
+            }
+        }
+
+        if(name == NULL || name[0] == 0
+            || U_FAILURE(errorCode) || cnv == NULL
+            || uprv_strlen(name)>=sizeof(gDefaultConverterNameBuffer))
+        {
+            /* Panic time, let's use a fallback. */
+#if (U_CHARSET_FAMILY == U_ASCII_FAMILY)
+            name = "US-ASCII";
+            /* there is no 'algorithmic' converter for EBCDIC */
+#elif U_PLATFORM == U_PF_OS390
+            name = "ibm-1047_P100-1995" UCNV_SWAP_LFNL_OPTION_STRING;
+#else
+            name = "ibm-37_P100-1995";
+#endif
+        }
+
+        internalSetName(name, &errorCode);
+
+        /* The close may make the current name go away. */
+        ucnv_close(cnv);
+    }
+
+    return name;
+#endif
+}
+
+#if U_CHARSET_IS_UTF8
+U_CAPI void U_EXPORT2 ucnv_setDefaultName(const char *) {}
+#else
+/*
+This function is not thread safe, and it can't be thread safe.
+See internalSetName or the API reference for details.
+*/
+U_CAPI void U_EXPORT2
+ucnv_setDefaultName(const char *converterName) {
+    if(converterName==NULL) {
+        /* reset to the default codepage */
+        gDefaultConverterName=NULL;
+    } else {
+        UErrorCode errorCode = U_ZERO_ERROR;
+        UConverter *cnv = NULL;
+        const char *name = NULL;
+
+        /* if the name is there, test it out and get the canonical name with options */
+        cnv = ucnv_open(converterName, &errorCode);
+        if(U_SUCCESS(errorCode) && cnv != NULL) {
+            name = ucnv_getName(cnv, &errorCode);
+        }
+
+        if(U_SUCCESS(errorCode) && name!=NULL) {
+            internalSetName(name, &errorCode);
+        }
+        /* else this converter is bad to use. Don't change it to a bad value. */
+
+        /* The close may make the current name go away. */
+        ucnv_close(cnv);
+
+        /* reset the converter cache */
+        u_flushDefaultConverter();
+    }
+}
+#endif
+
+/* data swapping ------------------------------------------------------------ */
+
+/* most of this might belong more properly into ucnvmbcs.c, but that is so large */
+
+#if !UCONFIG_NO_LEGACY_CONVERSION
+
+U_CAPI int32_t U_EXPORT2
+ucnv_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    uint32_t offset, count, staticDataSize;
+    int32_t size;
+
+    const UConverterStaticData *inStaticData;
+    UConverterStaticData *outStaticData;
+
+    const _MBCSHeader *inMBCSHeader;
+    _MBCSHeader *outMBCSHeader;
+    _MBCSHeader mbcsHeader;
+    uint32_t mbcsHeaderLength;
+    UBool noFromU=FALSE;
+
+    uint8_t outputType;
+
+    int32_t maxFastUChar, mbcsIndexLength;
+
+    const int32_t *inExtIndexes;
+    int32_t extOffset;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x63 &&   /* dataFormat="cnvt" */
+        pInfo->dataFormat[1]==0x6e &&
+        pInfo->dataFormat[2]==0x76 &&
+        pInfo->dataFormat[3]==0x74 &&
+        pInfo->formatVersion[0]==6 &&
+        pInfo->formatVersion[1]>=2
+    )) {
+        udata_printError(ds, "ucnv_swap(): data format %02x.%02x.%02x.%02x (format version %02x.%02x) is not recognized as an ICU .cnv conversion table\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0], pInfo->formatVersion[1]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData+headerSize;
+    outBytes=(uint8_t *)outData+headerSize;
+
+    /* read the initial UConverterStaticData structure after the UDataInfo header */
+    inStaticData=(const UConverterStaticData *)inBytes;
+    outStaticData=(UConverterStaticData *)outBytes;
+
+    if(length<0) {
+        staticDataSize=ds->readUInt32(inStaticData->structSize);
+    } else {
+        length-=headerSize;
+        if( length<(int32_t)sizeof(UConverterStaticData) ||
+            (uint32_t)length<(staticDataSize=ds->readUInt32(inStaticData->structSize))
+        ) {
+            udata_printError(ds, "ucnv_swap(): too few bytes (%d after header) for an ICU .cnv conversion table\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    if(length>=0) {
+        /* swap the static data */
+        if(inStaticData!=outStaticData) {
+            uprv_memcpy(outStaticData, inStaticData, staticDataSize);
+        }
+
+        ds->swapArray32(ds, &inStaticData->structSize, 4,
+                           &outStaticData->structSize, pErrorCode);
+        ds->swapArray32(ds, &inStaticData->codepage, 4,
+                           &outStaticData->codepage, pErrorCode);
+
+        ds->swapInvChars(ds, inStaticData->name, (int32_t)uprv_strlen(inStaticData->name),
+                            outStaticData->name, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            udata_printError(ds, "ucnv_swap(): error swapping converter name\n");
+            return 0;
+        }
+    }
+
+    inBytes+=staticDataSize;
+    outBytes+=staticDataSize;
+    if(length>=0) {
+        length-=(int32_t)staticDataSize;
+    }
+
+    /* check for supported conversionType values */
+    if(inStaticData->conversionType==UCNV_MBCS) {
+        /* swap MBCS data */
+        inMBCSHeader=(const _MBCSHeader *)inBytes;
+        outMBCSHeader=(_MBCSHeader *)outBytes;
+
+        if(0<=length && length<(int32_t)sizeof(_MBCSHeader)) {
+            udata_printError(ds, "ucnv_swap(): too few bytes (%d after headers) for an ICU MBCS .cnv conversion table\n",
+                                length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+        if(inMBCSHeader->version[0]==4 && inMBCSHeader->version[1]>=1) {
+            mbcsHeaderLength=MBCS_HEADER_V4_LENGTH;
+        } else if(inMBCSHeader->version[0]==5 && inMBCSHeader->version[1]>=3 &&
+                  ((mbcsHeader.options=ds->readUInt32(inMBCSHeader->options))&
+                   MBCS_OPT_UNKNOWN_INCOMPATIBLE_MASK)==0
+        ) {
+            mbcsHeaderLength=mbcsHeader.options&MBCS_OPT_LENGTH_MASK;
+            noFromU=(UBool)((mbcsHeader.options&MBCS_OPT_NO_FROM_U)!=0);
+        } else {
+            udata_printError(ds, "ucnv_swap(): unsupported _MBCSHeader.version %d.%d\n",
+                             inMBCSHeader->version[0], inMBCSHeader->version[1]);
+            *pErrorCode=U_UNSUPPORTED_ERROR;
+            return 0;
+        }
+
+        uprv_memcpy(mbcsHeader.version, inMBCSHeader->version, 4);
+        mbcsHeader.countStates=         ds->readUInt32(inMBCSHeader->countStates);
+        mbcsHeader.countToUFallbacks=   ds->readUInt32(inMBCSHeader->countToUFallbacks);
+        mbcsHeader.offsetToUCodeUnits=  ds->readUInt32(inMBCSHeader->offsetToUCodeUnits);
+        mbcsHeader.offsetFromUTable=    ds->readUInt32(inMBCSHeader->offsetFromUTable);
+        mbcsHeader.offsetFromUBytes=    ds->readUInt32(inMBCSHeader->offsetFromUBytes);
+        mbcsHeader.flags=               ds->readUInt32(inMBCSHeader->flags);
+        mbcsHeader.fromUBytesLength=    ds->readUInt32(inMBCSHeader->fromUBytesLength);
+        /* mbcsHeader.options have been read above */
+
+        extOffset=(int32_t)(mbcsHeader.flags>>8);
+        outputType=(uint8_t)mbcsHeader.flags;
+        if(noFromU && outputType==MBCS_OUTPUT_1) {
+            udata_printError(ds, "ucnv_swap(): unsupported combination of makeconv --small with SBCS\n");
+            *pErrorCode=U_UNSUPPORTED_ERROR;
+            return 0;
+        }
+
+        /* make sure that the output type is known */
+        switch(outputType) {
+        case MBCS_OUTPUT_1:
+        case MBCS_OUTPUT_2:
+        case MBCS_OUTPUT_3:
+        case MBCS_OUTPUT_4:
+        case MBCS_OUTPUT_3_EUC:
+        case MBCS_OUTPUT_4_EUC:
+        case MBCS_OUTPUT_2_SISO:
+        case MBCS_OUTPUT_EXT_ONLY:
+            /* OK */
+            break;
+        default:
+            udata_printError(ds, "ucnv_swap(): unsupported MBCS output type 0x%x\n",
+                             outputType);
+            *pErrorCode=U_UNSUPPORTED_ERROR;
+            return 0;
+        }
+
+        /* calculate the length of the MBCS data */
+
+        /*
+         * utf8Friendly MBCS files (mbcsHeader.version 4.3)
+         * contain an additional mbcsIndex table:
+         *   uint16_t[(maxFastUChar+1)>>6];
+         * where maxFastUChar=((mbcsHeader.version[2]<<8)|0xff).
+         */
+        maxFastUChar=0;
+        mbcsIndexLength=0;
+        if( outputType!=MBCS_OUTPUT_EXT_ONLY && outputType!=MBCS_OUTPUT_1 &&
+            mbcsHeader.version[1]>=3 && (maxFastUChar=mbcsHeader.version[2])!=0
+        ) {
+            maxFastUChar=(maxFastUChar<<8)|0xff;
+            mbcsIndexLength=((maxFastUChar+1)>>6)*2;  /* number of bytes */
+        }
+
+        if(extOffset==0) {
+            size=(int32_t)(mbcsHeader.offsetFromUBytes+mbcsIndexLength);
+            if(!noFromU) {
+                size+=(int32_t)mbcsHeader.fromUBytesLength;
+            }
+
+            /* avoid compiler warnings - not otherwise necessary, and the value does not matter */
+            inExtIndexes=NULL;
+        } else {
+            /* there is extension data after the base data, see ucnv_ext.h */
+            if(length>=0 && length<(extOffset+UCNV_EXT_INDEXES_MIN_LENGTH*4)) {
+                udata_printError(ds, "ucnv_swap(): too few bytes (%d after headers) for an ICU MBCS .cnv conversion table with extension data\n",
+                                 length);
+                *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+
+            inExtIndexes=(const int32_t *)(inBytes+extOffset);
+            size=extOffset+udata_readInt32(ds, inExtIndexes[UCNV_EXT_SIZE]);
+        }
+
+        if(length>=0) {
+            if(length<size) {
+                udata_printError(ds, "ucnv_swap(): too few bytes (%d after headers) for an ICU MBCS .cnv conversion table\n",
+                                 length);
+                *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+
+            /* copy the data for inaccessible bytes */
+            if(inBytes!=outBytes) {
+                uprv_memcpy(outBytes, inBytes, size);
+            }
+
+            /* swap the MBCSHeader, except for the version field */
+            count=mbcsHeaderLength*4;
+            ds->swapArray32(ds, &inMBCSHeader->countStates, count-4,
+                               &outMBCSHeader->countStates, pErrorCode);
+
+            if(outputType==MBCS_OUTPUT_EXT_ONLY) {
+                /*
+                 * extension-only file,
+                 * contains a base name instead of normal base table data
+                 */
+
+                /* swap the base name, between the header and the extension data */
+                const char *inBaseName=(const char *)inBytes+count;
+                char *outBaseName=(char *)outBytes+count;
+                ds->swapInvChars(ds, inBaseName, (int32_t)uprv_strlen(inBaseName),
+                                    outBaseName, pErrorCode);
+            } else {
+                /* normal file with base table data */
+
+                /* swap the state table, 1kB per state */
+                offset=count;
+                count=mbcsHeader.countStates*1024;
+                ds->swapArray32(ds, inBytes+offset, (int32_t)count,
+                                   outBytes+offset, pErrorCode);
+
+                /* swap the toUFallbacks[] */
+                offset+=count;
+                count=mbcsHeader.countToUFallbacks*8;
+                ds->swapArray32(ds, inBytes+offset, (int32_t)count,
+                                   outBytes+offset, pErrorCode);
+
+                /* swap the unicodeCodeUnits[] */
+                offset=mbcsHeader.offsetToUCodeUnits;
+                count=mbcsHeader.offsetFromUTable-offset;
+                ds->swapArray16(ds, inBytes+offset, (int32_t)count,
+                                   outBytes+offset, pErrorCode);
+
+                /* offset to the stage 1 table, independent of the outputType */
+                offset=mbcsHeader.offsetFromUTable;
+
+                if(outputType==MBCS_OUTPUT_1) {
+                    /* SBCS: swap the fromU tables, all 16 bits wide */
+                    count=(mbcsHeader.offsetFromUBytes-offset)+mbcsHeader.fromUBytesLength;
+                    ds->swapArray16(ds, inBytes+offset, (int32_t)count,
+                                       outBytes+offset, pErrorCode);
+                } else {
+                    /* otherwise: swap the stage tables separately */
+
+                    /* stage 1 table: uint16_t[0x440 or 0x40] */
+                    if(inStaticData->unicodeMask&UCNV_HAS_SUPPLEMENTARY) {
+                        count=0x440*2; /* for all of Unicode */
+                    } else {
+                        count=0x40*2; /* only BMP */
+                    }
+                    ds->swapArray16(ds, inBytes+offset, (int32_t)count,
+                                       outBytes+offset, pErrorCode);
+
+                    /* stage 2 table: uint32_t[] */
+                    offset+=count;
+                    count=mbcsHeader.offsetFromUBytes-offset;
+                    ds->swapArray32(ds, inBytes+offset, (int32_t)count,
+                                       outBytes+offset, pErrorCode);
+
+                    /* stage 3/result bytes: sometimes uint16_t[] or uint32_t[] */
+                    offset=mbcsHeader.offsetFromUBytes;
+                    count= noFromU ? 0 : mbcsHeader.fromUBytesLength;
+                    switch(outputType) {
+                    case MBCS_OUTPUT_2:
+                    case MBCS_OUTPUT_3_EUC:
+                    case MBCS_OUTPUT_2_SISO:
+                        ds->swapArray16(ds, inBytes+offset, (int32_t)count,
+                                           outBytes+offset, pErrorCode);
+                        break;
+                    case MBCS_OUTPUT_4:
+                        ds->swapArray32(ds, inBytes+offset, (int32_t)count,
+                                           outBytes+offset, pErrorCode);
+                        break;
+                    default:
+                        /* just uint8_t[], nothing to swap */
+                        break;
+                    }
+
+                    if(mbcsIndexLength!=0) {
+                        offset+=count;
+                        count=mbcsIndexLength;
+                        ds->swapArray16(ds, inBytes+offset, (int32_t)count,
+                                           outBytes+offset, pErrorCode);
+                    }
+                }
+            }
+
+            if(extOffset!=0) {
+                /* swap the extension data */
+                inBytes+=extOffset;
+                outBytes+=extOffset;
+
+                /* swap toUTable[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_TO_U_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_TO_U_LENGTH]);
+                ds->swapArray32(ds, inBytes+offset, length*4, outBytes+offset, pErrorCode);
+
+                /* swap toUUChars[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_TO_U_UCHARS_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_TO_U_UCHARS_LENGTH]);
+                ds->swapArray16(ds, inBytes+offset, length*2, outBytes+offset, pErrorCode);
+
+                /* swap fromUTableUChars[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_UCHARS_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_LENGTH]);
+                ds->swapArray16(ds, inBytes+offset, length*2, outBytes+offset, pErrorCode);
+
+                /* swap fromUTableValues[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_VALUES_INDEX]);
+                /* same length as for fromUTableUChars[] */
+                ds->swapArray32(ds, inBytes+offset, length*4, outBytes+offset, pErrorCode);
+
+                /* no need to swap fromUBytes[] */
+
+                /* swap fromUStage12[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_12_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_12_LENGTH]);
+                ds->swapArray16(ds, inBytes+offset, length*2, outBytes+offset, pErrorCode);
+
+                /* swap fromUStage3[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_3_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_3_LENGTH]);
+                ds->swapArray16(ds, inBytes+offset, length*2, outBytes+offset, pErrorCode);
+
+                /* swap fromUStage3b[] */
+                offset=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_3B_INDEX]);
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_FROM_U_STAGE_3B_LENGTH]);
+                ds->swapArray32(ds, inBytes+offset, length*4, outBytes+offset, pErrorCode);
+
+                /* swap indexes[] */
+                length=udata_readInt32(ds, inExtIndexes[UCNV_EXT_INDEXES_LENGTH]);
+                ds->swapArray32(ds, inBytes, length*4, outBytes, pErrorCode);
+            }
+        }
+    } else {
+        udata_printError(ds, "ucnv_swap(): unknown conversionType=%d!=UCNV_MBCS\n",
+                         inStaticData->conversionType);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    return headerSize+(int32_t)staticDataSize+size;
+}
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_bld.h b/thirdparty/icu4c/common/ucnv_bld.h
new file mode 100644
index 0000000000..43e6c09ac0
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_bld.h
@@ -0,0 +1,296 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2015 International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+*
+*  ucnv_bld.h:
+*  Contains internal data structure definitions
+* Created by Bertrand A. Damiba
+*
+*   Change history:
+*
+*   06/29/2000  helena      Major rewrite of the callback APIs.
+*/
+
+#ifndef UCNV_BLD_H
+#define UCNV_BLD_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_err.h"
+#include "unicode/utf16.h"
+#include "ucnv_cnv.h"
+#include "ucnvmbcs.h"
+#include "ucnv_ext.h"
+#include "udataswp.h"
+
+/* size of the overflow buffers in UConverter, enough for escaping callbacks */
+#define UCNV_ERROR_BUFFER_LENGTH 32
+
+/* at most 4 bytes per substitution character (part of .cnv file format! see UConverterStaticData) */
+#define UCNV_MAX_SUBCHAR_LEN 4
+
+/* at most 8 bytes per character in toUBytes[] (UTF-8 uses up to 6) */
+#define UCNV_MAX_CHAR_LEN 8
+
+/* converter options bits */
+#define UCNV_OPTION_VERSION     0xf
+#define UCNV_OPTION_SWAP_LFNL   0x10
+
+#define UCNV_GET_VERSION(cnv) ((cnv)->options&UCNV_OPTION_VERSION)
+
+U_CDECL_BEGIN /* We must declare the following as 'extern "C"' so that if ucnv
+                 itself is compiled under C++, the linkage of the funcptrs will
+                 work.
+              */
+
+union UConverterTable {
+    UConverterMBCSTable mbcs;
+};
+
+typedef union UConverterTable UConverterTable;
+
+struct UConverterImpl;
+typedef struct UConverterImpl UConverterImpl;
+
+/** values for the unicodeMask */
+#define UCNV_HAS_SUPPLEMENTARY 1
+#define UCNV_HAS_SURROGATES    2
+
+typedef struct UConverterStaticData {   /* +offset: size */
+    uint32_t structSize;                /* +0: 4 Size of this structure */
+
+    char name
+      [UCNV_MAX_CONVERTER_NAME_LENGTH]; /* +4: 60  internal name of the converter- invariant chars */
+
+    int32_t codepage;               /* +64: 4 codepage # (now IBM-$codepage) */
+
+    int8_t platform;                /* +68: 1 platform of the converter (only IBM now) */
+    int8_t conversionType;          /* +69: 1 conversion type */
+
+    int8_t minBytesPerChar;         /* +70: 1 Minimum # bytes per char in this codepage */
+    int8_t maxBytesPerChar;         /* +71: 1 Maximum # bytes output per UChar in this codepage */
+
+    uint8_t subChar[UCNV_MAX_SUBCHAR_LEN]; /* +72: 4  [note:  4 and 8 byte boundary] */
+    int8_t subCharLen;              /* +76: 1 */
+
+    uint8_t hasToUnicodeFallback;   /* +77: 1 UBool needs to be changed to UBool to be consistent across platform */
+    uint8_t hasFromUnicodeFallback; /* +78: 1 */
+    uint8_t unicodeMask;            /* +79: 1  bit 0: has supplementary  bit 1: has single surrogates */
+    uint8_t subChar1;               /* +80: 1  single-byte substitution character for IBM MBCS (0 if none) */
+    uint8_t reserved[19];           /* +81: 19 to round out the structure */
+                                    /* total size: 100 */
+} UConverterStaticData;
+
+/*
+ * Defines the UConverterSharedData struct,
+ * the immutable, shared part of UConverter.
+ */
+struct UConverterSharedData {
+    uint32_t structSize;            /* Size of this structure */
+    uint32_t referenceCounter;      /* used to count number of clients, unused for static/immutable SharedData */
+
+    const void *dataMemory;         /* from udata_openChoice() - for cleanup */
+
+    const UConverterStaticData *staticData; /* pointer to the static (non changing) data. */
+
+    UBool                sharedDataCached;   /* true:  shared data is in cache, don't destroy on ucnv_close() if 0 ref.  false: shared data isn't in the cache, do attempt to clean it up if the ref is 0 */
+    /** If false, then referenceCounter is not used. Must not change after initialization. */
+    UBool isReferenceCounted;
+
+    const UConverterImpl *impl;     /* vtable-style struct of mostly function pointers */
+
+    /*initial values of some members of the mutable part of object */
+    uint32_t toUnicodeStatus;
+
+    /*
+     * Shared data structures currently come in two flavors:
+     * - readonly for built-in algorithmic converters
+     * - allocated for MBCS, with a pointer to an allocated UConverterTable
+     *   which always has a UConverterMBCSTable
+     *
+     * To eliminate one allocation, I am making the UConverterMBCSTable
+     * a member of the shared data.
+     *
+     * markus 2003-nov-07
+     */
+    UConverterMBCSTable mbcs;
+};
+
+/** UConverterSharedData initializer for static, non-reference-counted converters. */
+#define UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(pStaticData, pImpl) \
+    { \
+        sizeof(UConverterSharedData), ~((uint32_t)0), \
+        NULL, pStaticData, false, false, pImpl, \
+        0, UCNV_MBCS_TABLE_INITIALIZER \
+    }
+
+/* Defines a UConverter, the lightweight mutable part the user sees */
+
+struct UConverter {
+    /*
+     * Error function pointer called when conversion issues
+     * occur during a ucnv_fromUnicode call
+     */
+    void (U_EXPORT2 *fromUCharErrorBehaviour) (const void *context,
+                                     UConverterFromUnicodeArgs *args,
+                                     const UChar *codeUnits,
+                                     int32_t length,
+                                     UChar32 codePoint,
+                                     UConverterCallbackReason reason,
+                                     UErrorCode *);
+    /*
+     * Error function pointer called when conversion issues
+     * occur during a ucnv_toUnicode call
+     */
+    void (U_EXPORT2 *fromCharErrorBehaviour) (const void *context,
+                                    UConverterToUnicodeArgs *args,
+                                    const char *codeUnits,
+                                    int32_t length,
+                                    UConverterCallbackReason reason,
+                                    UErrorCode *);
+
+    /*
+     * Pointer to additional data that depends on the converter type.
+     * Used by ISO 2022, SCSU, GB 18030 converters, possibly more.
+     */
+    void *extraInfo;
+
+    const void *fromUContext;
+    const void *toUContext;
+
+    /*
+     * Pointer to charset bytes for substitution string if subCharLen>0,
+     * or pointer to Unicode string (UChar *) if subCharLen<0.
+     * subCharLen==0 is equivalent to using a skip callback.
+     * If the pointer is !=subUChars then it is allocated with
+     * UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR bytes.
+     * The subUChars field is declared as UChar[] not uint8_t[] to
+     * guarantee alignment for UChars.
+     */
+    uint8_t *subChars;
+
+    UConverterSharedData *sharedData;   /* Pointer to the shared immutable part of the converter object */
+
+    uint32_t options; /* options flags from UConverterOpen, may contain additional bits */
+
+    UBool sharedDataIsCached;  /* true:  shared data is in cache, don't destroy on ucnv_close() if 0 ref.  false: shared data isn't in the cache, do attempt to clean it up if the ref is 0 */
+    UBool isCopyLocal;  /* true if UConverter is not owned and not released in ucnv_close() (stack-allocated, safeClone(), etc.) */
+    UBool isExtraLocal; /* true if extraInfo is not owned and not released in ucnv_close() (stack-allocated, safeClone(), etc.) */
+
+    UBool  useFallback;
+    int8_t toULength;                   /* number of bytes in toUBytes */
+    uint8_t toUBytes[UCNV_MAX_CHAR_LEN-1];/* more "toU status"; keeps the bytes of the current character */
+    uint32_t toUnicodeStatus;           /* Used to internalize stream status information */
+    int32_t mode;
+    uint32_t fromUnicodeStatus;
+
+    /*
+     * More fromUnicode() status. Serves 3 purposes:
+     * - keeps a lead surrogate between buffers (similar to toUBytes[])
+     * - keeps a lead surrogate at the end of the stream,
+     *   which the framework handles as truncated input
+     * - if the fromUnicode() implementation returns to the framework
+     *   (ucnv.c ucnv_fromUnicode()), then the framework calls the callback
+     *   for this code point
+     */
+    UChar32 fromUChar32;
+
+    /*
+     * value for ucnv_getMaxCharSize()
+     *
+     * usually simply copied from the static data, but ucnvmbcs.c modifies
+     * the value depending on the converter type and options
+     */
+    int8_t maxBytesPerUChar;
+
+    int8_t subCharLen;                  /* length of the codepage specific character sequence */
+    int8_t invalidCharLength;
+    int8_t charErrorBufferLength;       /* number of valid bytes in charErrorBuffer */
+
+    int8_t invalidUCharLength;
+    int8_t UCharErrorBufferLength;      /* number of valid UChars in charErrorBuffer */
+
+    uint8_t subChar1;                                   /* single-byte substitution character if different from subChar */
+    UBool useSubChar1;
+    char invalidCharBuffer[UCNV_MAX_CHAR_LEN];          /* bytes from last error/callback situation */
+    uint8_t charErrorBuffer[UCNV_ERROR_BUFFER_LENGTH];  /* codepage output from Error functions */
+    UChar subUChars[UCNV_MAX_SUBCHAR_LEN/U_SIZEOF_UCHAR]; /* see subChars documentation */
+
+    UChar invalidUCharBuffer[U16_MAX_LENGTH];           /* UChars from last error/callback situation */
+    UChar UCharErrorBuffer[UCNV_ERROR_BUFFER_LENGTH];   /* unicode output from Error functions */
+
+    /* fields for conversion extension */
+
+    /* store previous UChars/chars to continue partial matches */
+    UChar32 preFromUFirstCP;                /* >=0: partial match */
+    UChar preFromU[UCNV_EXT_MAX_UCHARS];
+    char preToU[UCNV_EXT_MAX_BYTES];
+    int8_t preFromULength, preToULength;    /* negative: replay */
+    int8_t preToUFirstLength;               /* length of first character */
+
+    /* new fields for ICU 4.0 */
+    UConverterCallbackReason toUCallbackReason; /* (*fromCharErrorBehaviour) reason, set when error is detected */
+};
+
+U_CDECL_END /* end of UConverter */
+
+#define CONVERTER_FILE_EXTENSION ".cnv"
+
+
+/**
+ * Return the number of all converter names.
+ * @param pErrorCode The error code
+ * @return the number of all converter names
+ */
+U_CFUNC uint16_t
+ucnv_bld_countAvailableConverters(UErrorCode *pErrorCode);
+
+/**
+ * Return the (n)th converter name in mixed case, or NULL
+ * if there is none (typically, if the data cannot be loaded).
+ * 0<=index<ucnv_io_countAvailableConverters().
+ * @param n The number specifies which converter name to get
+ * @param pErrorCode The error code
+ * @return the (n)th converter name in mixed case, or NULL if there is none.
+ */
+U_CFUNC const char *
+ucnv_bld_getAvailableConverter(uint16_t n, UErrorCode *pErrorCode);
+
+/**
+ * Load a non-algorithmic converter.
+ * If pkg==NULL, then this function must be called inside umtx_lock(&cnvCacheMutex).
+ */
+U_CAPI UConverterSharedData *
+ucnv_load(UConverterLoadArgs *pArgs, UErrorCode *err);
+
+/**
+ * Unload a non-algorithmic converter.
+ * It must be sharedData->isReferenceCounted
+ * and this function must be called inside umtx_lock(&cnvCacheMutex).
+ */
+U_CAPI void
+ucnv_unload(UConverterSharedData *sharedData);
+
+/**
+ * Swap ICU .cnv conversion tables. See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode);
+
+U_CAPI void U_EXPORT2
+ucnv_enableCleanup(void);
+
+#endif
+
+#endif /* _UCNV_BLD */
diff --git a/thirdparty/icu4c/common/ucnv_cb.cpp b/thirdparty/icu4c/common/ucnv_cb.cpp
new file mode 100644
index 0000000000..1bb0012014
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_cb.cpp
@@ -0,0 +1,261 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2000-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+ *  ucnv_cb.c:
+ *  External APIs for the ICU's codeset conversion library
+ *  Helena Shih
+ *
+ * Modification History:
+ *
+ *   Date        Name        Description
+ *   7/28/2000   srl         Implementation
+ */
+
+/**
+ * @name Character Conversion C API
+ *
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv_cb.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+
+/* need to update the offsets when the target moves. */
+/* Note: Recursion may occur in the cb functions, be sure to update the offsets correctly
+if you don't use ucnv_cbXXX functions.  Make sure you don't use the same callback within
+the same call stack if the complexity arises. */
+U_CAPI void  U_EXPORT2
+ucnv_cbFromUWriteBytes (UConverterFromUnicodeArgs *args,
+                       const char* source,
+                       int32_t length,
+                       int32_t offsetIndex,
+                       UErrorCode * err)
+{
+    if(U_FAILURE(*err)) {
+        return;
+    }
+
+    ucnv_fromUWriteBytes(
+        args->converter,
+        source, length,
+        &args->target, args->targetLimit,
+        &args->offsets, offsetIndex,
+        err);
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_cbFromUWriteUChars(UConverterFromUnicodeArgs *args,
+                             const UChar** source,
+                             const UChar*  sourceLimit,
+                             int32_t offsetIndex,
+                             UErrorCode * err)
+{
+    /*
+    This is a fun one.  Recursion can occur - we're basically going to
+    just retry shoving data through the same converter. Note, if you got
+    here through some kind of invalid sequence, you maybe should emit a
+    reset sequence of some kind and/or call ucnv_reset().  Since this
+    IS an actual conversion, take care that you've changed the callback
+    or the data, or you'll get an infinite loop.
+
+    Please set the err value to something reasonable before calling
+    into this.
+    */
+
+    char *oldTarget;
+
+    if(U_FAILURE(*err))
+    {
+        return;
+    }
+
+    oldTarget = args->target;
+
+    ucnv_fromUnicode(args->converter,
+        &args->target,
+        args->targetLimit,
+        source,
+        sourceLimit,
+        NULL, /* no offsets */
+        FALSE, /* no flush */
+        err);
+
+    if(args->offsets)
+    {
+        while (args->target != oldTarget)  /* if it moved at all.. */
+        {
+            *(args->offsets)++ = offsetIndex;
+            oldTarget++;
+        }
+    }
+
+    /*
+    Note, if you did something like used a Stop subcallback, things would get interesting.
+    In fact, here's where we want to return the partially consumed in-source!
+    */
+    if(*err == U_BUFFER_OVERFLOW_ERROR)
+    /* && (*source < sourceLimit && args->target >= args->targetLimit)
+    -- S. Hrcek */
+    {
+        /* Overflowed the target.  Now, we'll write into the charErrorBuffer.
+        It's a fixed size. If we overflow it... Hmm */
+        char *newTarget;
+        const char *newTargetLimit;
+        UErrorCode err2 = U_ZERO_ERROR;
+
+        int8_t errBuffLen;
+
+        errBuffLen  = args->converter->charErrorBufferLength;
+
+        /* start the new target at the first free slot in the errbuff.. */
+        newTarget = (char *)(args->converter->charErrorBuffer + errBuffLen);
+
+        newTargetLimit = (char *)(args->converter->charErrorBuffer +
+            sizeof(args->converter->charErrorBuffer));
+
+        if(newTarget >= newTargetLimit)
+        {
+            *err = U_INTERNAL_PROGRAM_ERROR;
+            return;
+        }
+
+        /* We're going to tell the converter that the errbuff len is empty.
+        This prevents the existing errbuff from being 'flushed' out onto
+        itself.  If the errbuff is needed by the converter this time,
+        we're hosed - we're out of space! */
+
+        args->converter->charErrorBufferLength = 0;
+
+        ucnv_fromUnicode(args->converter,
+                         &newTarget,
+                         newTargetLimit,
+                         source,
+                         sourceLimit,
+                         NULL,
+                         FALSE,
+                         &err2);
+
+        /* We can go ahead and overwrite the  length here. We know just how
+        to recalculate it. */
+
+        args->converter->charErrorBufferLength = (int8_t)(
+            newTarget - (char*)args->converter->charErrorBuffer);
+
+        if((newTarget >= newTargetLimit) || (err2 == U_BUFFER_OVERFLOW_ERROR))
+        {
+            /* now we're REALLY in trouble.
+            Internal program error - callback shouldn't have written this much
+            data!
+            */
+            *err = U_INTERNAL_PROGRAM_ERROR;
+            return;
+        }
+        /*else {*/
+            /* sub errs could be invalid/truncated/illegal chars or w/e.
+            These might want to be passed on up.. But the problem is, we already
+            need to pass U_BUFFER_OVERFLOW_ERROR. That has to override these
+            other errs.. */
+
+            /*
+            if(U_FAILURE(err2))
+            ??
+            */
+        /*}*/
+    }
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_cbFromUWriteSub (UConverterFromUnicodeArgs *args,
+                           int32_t offsetIndex,
+                           UErrorCode * err)
+{
+    UConverter *converter;
+    int32_t length;
+
+    if(U_FAILURE(*err)) {
+        return;
+    }
+    converter = args->converter;
+    length = converter->subCharLen;
+
+    if(length == 0) {
+        return;
+    }
+
+    if(length < 0) {
+        /*
+         * Write/convert the substitution string. Its real length is -length.
+         * Unlike the escape callback, we need not change the converter's
+         * callback function because ucnv_setSubstString() verified that
+         * the string can be converted, so we will not get a conversion error
+         * and will not recurse.
+         * At worst we should get a U_BUFFER_OVERFLOW_ERROR.
+         */
+        const UChar *source = (const UChar *)converter->subChars;
+        ucnv_cbFromUWriteUChars(args, &source, source - length, offsetIndex, err);
+        return;
+    }
+
+    if(converter->sharedData->impl->writeSub!=NULL) {
+        converter->sharedData->impl->writeSub(args, offsetIndex, err);
+    }
+    else if(converter->subChar1!=0 && (uint16_t)converter->invalidUCharBuffer[0]<=(uint16_t)0xffu) {
+        /*
+        TODO: Is this untestable because the MBCS converter has a writeSub function to call
+        and the other converters don't use subChar1?
+        */
+        ucnv_cbFromUWriteBytes(args,
+                               (const char *)&converter->subChar1, 1,
+                               offsetIndex, err);
+    }
+    else {
+        ucnv_cbFromUWriteBytes(args,
+                               (const char *)converter->subChars, length,
+                               offsetIndex, err);
+    }
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_cbToUWriteUChars (UConverterToUnicodeArgs *args,
+                            const UChar* source,
+                            int32_t length,
+                            int32_t offsetIndex,
+                            UErrorCode * err)
+{
+    if(U_FAILURE(*err)) {
+        return;
+    }
+
+    ucnv_toUWriteUChars(
+        args->converter,
+        source, length,
+        &args->target, args->targetLimit,
+        &args->offsets, offsetIndex,
+        err);
+}
+
+U_CAPI void  U_EXPORT2
+ucnv_cbToUWriteSub (UConverterToUnicodeArgs *args,
+                         int32_t offsetIndex,
+                       UErrorCode * err)
+{
+    static const UChar kSubstituteChar1 = 0x1A, kSubstituteChar = 0xFFFD;
+
+    /* could optimize this case, just one uchar */
+    if(args->converter->invalidCharLength == 1 && args->converter->subChar1 != 0) {
+        ucnv_cbToUWriteUChars(args, &kSubstituteChar1, 1, offsetIndex, err);
+    } else {
+        ucnv_cbToUWriteUChars(args, &kSubstituteChar, 1, offsetIndex, err);
+    }
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_cnv.cpp b/thirdparty/icu4c/common/ucnv_cnv.cpp
new file mode 100644
index 0000000000..ea71acf92c
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_cnv.cpp
@@ -0,0 +1,182 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2004, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*   uconv_cnv.c:
+*   Implements all the low level conversion functions
+*   T_UnicodeConverter_{to,from}Unicode_$ConversionType
+*
+*   Change history:
+*
+*   06/29/2000  helena      Major rewrite of the callback APIs.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv_err.h"
+#include "unicode/ucnv.h"
+#include "unicode/uset.h"
+#include "ucnv_cnv.h"
+#include "ucnv_bld.h"
+#include "cmemory.h"
+
+U_CFUNC void
+ucnv_getCompleteUnicodeSet(const UConverter *cnv,
+                   const USetAdder *sa,
+                   UConverterUnicodeSet which,
+                   UErrorCode *pErrorCode) {
+    (void)cnv;
+    (void)which;
+    (void)pErrorCode;
+    sa->addRange(sa->set, 0, 0x10ffff);
+}
+
+U_CFUNC void
+ucnv_getNonSurrogateUnicodeSet(const UConverter *cnv,
+                               const USetAdder *sa,
+                               UConverterUnicodeSet which,
+                               UErrorCode *pErrorCode) {
+    (void)cnv;
+    (void)which;
+    (void)pErrorCode;
+    sa->addRange(sa->set, 0, 0xd7ff);
+    sa->addRange(sa->set, 0xe000, 0x10ffff);
+}
+
+U_CFUNC void
+ucnv_fromUWriteBytes(UConverter *cnv,
+                     const char *bytes, int32_t length,
+                     char **target, const char *targetLimit,
+                     int32_t **offsets,
+                     int32_t sourceIndex,
+                     UErrorCode *pErrorCode) {
+    char *t=*target;
+    int32_t *o;
+
+    /* write bytes */
+    if(offsets==NULL || (o=*offsets)==NULL) {
+        while(length>0 && t<targetLimit) {
+            *t++=*bytes++;
+            --length;
+        }
+    } else {
+        /* output with offsets */
+        while(length>0 && t<targetLimit) {
+            *t++=*bytes++;
+            *o++=sourceIndex;
+            --length;
+        }
+        *offsets=o;
+    }
+    *target=t;
+
+    /* write overflow */
+    if(length>0) {
+        if(cnv!=NULL) {
+            t=(char *)cnv->charErrorBuffer;
+            cnv->charErrorBufferLength=(int8_t)length;
+            do {
+                *t++=(uint8_t)*bytes++;
+            } while(--length>0);
+        }
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+}
+
+U_CFUNC void
+ucnv_toUWriteUChars(UConverter *cnv,
+                    const UChar *uchars, int32_t length,
+                    UChar **target, const UChar *targetLimit,
+                    int32_t **offsets,
+                    int32_t sourceIndex,
+                    UErrorCode *pErrorCode) {
+    UChar *t=*target;
+    int32_t *o;
+
+    /* write UChars */
+    if(offsets==NULL || (o=*offsets)==NULL) {
+        while(length>0 && t<targetLimit) {
+            *t++=*uchars++;
+            --length;
+        }
+    } else {
+        /* output with offsets */
+        while(length>0 && t<targetLimit) {
+            *t++=*uchars++;
+            *o++=sourceIndex;
+            --length;
+        }
+        *offsets=o;
+    }
+    *target=t;
+
+    /* write overflow */
+    if(length>0) {
+        if(cnv!=NULL) {
+            t=cnv->UCharErrorBuffer;
+            cnv->UCharErrorBufferLength=(int8_t)length;
+            do {
+                *t++=*uchars++;
+            } while(--length>0);
+        }
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+}
+
+U_CFUNC void
+ucnv_toUWriteCodePoint(UConverter *cnv,
+                       UChar32 c,
+                       UChar **target, const UChar *targetLimit,
+                       int32_t **offsets,
+                       int32_t sourceIndex,
+                       UErrorCode *pErrorCode) {
+    UChar *t;
+    int32_t *o;
+
+    t=*target;
+
+    if(t<targetLimit) {
+        if(c<=0xffff) {
+            *t++=(UChar)c;
+            c=U_SENTINEL;
+        } else /* c is a supplementary code point */ {
+            *t++=U16_LEAD(c);
+            c=U16_TRAIL(c);
+            if(t<targetLimit) {
+                *t++=(UChar)c;
+                c=U_SENTINEL;
+            }
+        }
+
+        /* write offsets */
+        if(offsets!=NULL && (o=*offsets)!=NULL) {
+            *o++=sourceIndex;
+            if((*target+1)<t) {
+                *o++=sourceIndex;
+            }
+            *offsets=o;
+        }
+    }
+
+    *target=t;
+
+    /* write overflow from c */
+    if(c>=0) {
+        if(cnv!=NULL) {
+            int8_t i=0;
+            U16_APPEND_UNSAFE(cnv->UCharErrorBuffer, i, c);
+            cnv->UCharErrorBufferLength=i;
+        }
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_cnv.h b/thirdparty/icu4c/common/ucnv_cnv.h
new file mode 100644
index 0000000000..59be8bdb37
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_cnv.h
@@ -0,0 +1,323 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+*   ucnv_cnv.h:
+*   Definitions for converter implementations.
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   05/09/00    helena      Added implementation to handle fallback mappings.
+*   06/29/2000  helena      Major rewrite of the callback APIs.
+*/
+
+#ifndef UCNV_CNV_H
+#define UCNV_CNV_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_err.h"
+#include "unicode/uset.h"
+#include "uset_imp.h"
+
+U_CDECL_BEGIN
+
+/* this is used in fromUnicode DBCS tables as an "unassigned" marker */
+#define missingCharMarker 0xFFFF
+
+/*
+ * #define missingUCharMarker 0xfffe
+ *
+ * commented out because there are actually two values used in toUnicode tables:
+ * U+fffe "unassigned"
+ * U+ffff "illegal"
+ */
+
+/** Forward declaration, see ucnv_bld.h */
+struct UConverterSharedData;
+typedef struct UConverterSharedData UConverterSharedData;
+
+/* function types for UConverterImpl ---------------------------------------- */
+
+/* struct with arguments for UConverterLoad and ucnv_load() */
+typedef struct {
+    int32_t size;               /* sizeof(UConverterLoadArgs) */
+    int32_t nestedLoads;        /* count nested ucnv_load() calls */
+    UBool onlyTestIsLoadable;   /* input: don't actually load */
+    UBool reserved0;            /* reserved - for good alignment of the pointers */
+    int16_t reserved;           /* reserved - for good alignment of the pointers */
+    uint32_t options;
+    const char *pkg, *name, *locale;
+} UConverterLoadArgs;
+
+#define UCNV_LOAD_ARGS_INITIALIZER \
+    { (int32_t)sizeof(UConverterLoadArgs), 0, false, false, 0, 0, NULL, NULL, NULL }
+
+typedef void (*UConverterLoad) (UConverterSharedData *sharedData,
+                                UConverterLoadArgs *pArgs,
+                                const uint8_t *raw, UErrorCode *pErrorCode);
+typedef void (*UConverterUnload) (UConverterSharedData *sharedData);
+
+typedef void (*UConverterOpen) (UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *pErrorCode);
+typedef void (*UConverterClose) (UConverter *cnv);
+
+typedef enum UConverterResetChoice {
+    UCNV_RESET_BOTH,
+    UCNV_RESET_TO_UNICODE,
+    UCNV_RESET_FROM_UNICODE
+} UConverterResetChoice;
+
+typedef void (*UConverterReset) (UConverter *cnv, UConverterResetChoice choice);
+
+/*
+ * Converter implementation function(s) for ucnv_toUnicode().
+ * If the toUnicodeWithOffsets function pointer is NULL,
+ * then the toUnicode function will be used and the offsets will be set to -1.
+ *
+ * Must maintain state across buffers. Use toUBytes[toULength] for partial input
+ * sequences; it will be checked in ucnv.c at the end of the input stream
+ * to detect truncated input.
+ * Some converters may need additional detection and may then set U_TRUNCATED_CHAR_FOUND.
+ *
+ * The toUnicodeWithOffsets must write exactly as many offset values as target
+ * units. Write offset values of -1 for when the source index corresponding to
+ * the output unit is not known (e.g., the character started in an earlier buffer).
+ * The pArgs->offsets pointer need not be moved forward.
+ *
+ * At function return, either one of the following conditions must be true:
+ * - U_BUFFER_OVERFLOW_ERROR and the target is full: target==targetLimit
+ * - another error code with toUBytes[toULength] set to the offending input
+ * - no error, and the source is consumed: source==sourceLimit
+ *
+ * The ucnv.c code will handle the end of the input (reset)
+ * (reset, and truncation detection) and callbacks.
+ */
+typedef void (*UConverterToUnicode) (UConverterToUnicodeArgs *, UErrorCode *);
+
+/*
+ * Same rules as for UConverterToUnicode.
+ * A lead surrogate is kept in fromUChar32 across buffers, and if an error
+ * occurs, then the offending input code point must be put into fromUChar32
+ * as well.
+ */
+typedef void (*UConverterFromUnicode) (UConverterFromUnicodeArgs *, UErrorCode *);
+
+/*
+ * Converter implementation function for ucnv_convertEx(), for direct conversion
+ * between two charsets without pivoting through UTF-16.
+ * The rules are the same as for UConverterToUnicode and UConverterFromUnicode.
+ * In addition,
+ * - The toUnicode side must behave and keep state exactly like the
+ *   UConverterToUnicode implementation for the same source charset.
+ * - A U_USING_DEFAULT_WARNING can be set to request to temporarily fall back
+ *   to pivoting. When this function is called, the conversion framework makes
+ *   sure that this warning is not set on input.
+ * - Continuing a partial match and flushing the toUnicode replay buffer
+ *   are handled by pivoting, using the toUnicode and fromUnicode functions.
+ */
+typedef void (*UConverterConvert) (UConverterFromUnicodeArgs *pFromUArgs,
+                                   UConverterToUnicodeArgs *pToUArgs,
+                                   UErrorCode *pErrorCode);
+
+/*
+ * Converter implementation function for ucnv_getNextUChar().
+ * If the function pointer is NULL, then the toUnicode function will be used.
+ *
+ * Will be called at a character boundary (toULength==0).
+ * May return with
+ * - U_INDEX_OUTOFBOUNDS_ERROR if there was no output for the input
+ *   (the return value will be ignored)
+ * - U_TRUNCATED_CHAR_FOUND or another error code (never U_BUFFER_OVERFLOW_ERROR!)
+ *   with toUBytes[toULength] set to the offending input
+ *   (the return value will be ignored)
+ * - return UCNV_GET_NEXT_UCHAR_USE_TO_U, without moving the source pointer,
+ *   to indicate that the ucnv.c code shall call the toUnicode function instead
+ * - return a real code point result
+ *
+ * Unless UCNV_GET_NEXT_UCHAR_USE_TO_U is returned, the source bytes must be consumed.
+ *
+ * The ucnv.c code will handle the end of the input (reset)
+ * (except for truncation detection!) and callbacks.
+ */
+typedef UChar32 (*UConverterGetNextUChar) (UConverterToUnicodeArgs *, UErrorCode *);
+
+typedef void (*UConverterGetStarters)(const UConverter* converter,
+                                      UBool starters[256],
+                                      UErrorCode *pErrorCode);
+
+/* If this function pointer is null or if the function returns null
+ * the name field in static data struct should be returned by 
+ * ucnv_getName() API function
+ */
+typedef const char * (*UConverterGetName) (const UConverter *cnv);
+
+/**
+ * Write the codepage substitution character.
+ * If this function is not set, then ucnv_cbFromUWriteSub() writes
+ * the substitution character from UConverter.
+ * For stateful converters, it is typically necessary to handle this
+ * specificially for the converter in order to properly maintain the state.
+ */
+typedef void (*UConverterWriteSub) (UConverterFromUnicodeArgs *pArgs, int32_t offsetIndex, UErrorCode *pErrorCode);
+
+/**
+ * For converter-specific safeClone processing
+ * If this function is not set, then ucnv_safeClone assumes that the converter has no private data that changes
+ * after the converter is done opening.
+ * If this function is set, then it is called just after a memcpy() of
+ * converter data to the new, empty converter, and is expected to set up
+ * the initial state of the converter.  It is not expected to increment the
+ * reference counts of the standard data types such as the shared data.
+ */
+typedef UConverter * (*UConverterSafeClone) (const UConverter   *cnv, 
+                                             void               *stackBuffer,
+                                             int32_t            *pBufferSize, 
+                                             UErrorCode         *status);
+
+/**
+ * Filters for some ucnv_getUnicodeSet() implementation code.
+ */
+typedef enum UConverterSetFilter {
+    UCNV_SET_FILTER_NONE,
+    UCNV_SET_FILTER_DBCS_ONLY,
+    UCNV_SET_FILTER_2022_CN,
+    UCNV_SET_FILTER_SJIS,
+    UCNV_SET_FILTER_GR94DBCS,
+    UCNV_SET_FILTER_HZ,
+    UCNV_SET_FILTER_COUNT
+} UConverterSetFilter;
+
+/**
+ * Fills the set of Unicode code points that can be converted by an ICU converter.
+ * The API function ucnv_getUnicodeSet() clears the USet before calling
+ * the converter's getUnicodeSet() implementation; the converter should only
+ * add the appropriate code points to allow recursive use.
+ * For example, the ISO-2022-JP converter will call each subconverter's
+ * getUnicodeSet() implementation to consecutively add code points to
+ * the same USet, which will result in a union of the sets of all subconverters.
+ *
+ * For more documentation, see ucnv_getUnicodeSet() in ucnv.h.
+ */
+typedef void (*UConverterGetUnicodeSet) (const UConverter *cnv,
+                                         const USetAdder *sa,
+                                         UConverterUnicodeSet which,
+                                         UErrorCode *pErrorCode);
+
+UBool CONVERSION_U_SUCCESS (UErrorCode err);
+
+/**
+ * UConverterImpl contains all the data and functions for a converter type.
+ * Its function pointers work much like a C++ vtable.
+ * Many converter types need to define only a subset of the functions;
+ * when a function pointer is NULL, then a default action will be performed.
+ *
+ * Every converter type must implement toUnicode, fromUnicode, and getNextUChar,
+ * otherwise the converter may crash.
+ * Every converter type that has variable-length codepage sequences should
+ * also implement toUnicodeWithOffsets and fromUnicodeWithOffsets for
+ * correct offset handling.
+ * All other functions may or may not be implemented - it depends only on
+ * whether the converter type needs them.
+ *
+ * When open() fails, then close() will be called, if present.
+ */
+struct UConverterImpl {
+    UConverterType type;
+
+    UConverterLoad load;
+    UConverterUnload unload;
+
+    UConverterOpen open;
+    UConverterClose close;
+    UConverterReset reset;
+
+    UConverterToUnicode toUnicode;
+    UConverterToUnicode toUnicodeWithOffsets;
+    UConverterFromUnicode fromUnicode;
+    UConverterFromUnicode fromUnicodeWithOffsets;
+    UConverterGetNextUChar getNextUChar;
+
+    UConverterGetStarters getStarters;
+    UConverterGetName getName;
+    UConverterWriteSub writeSub;
+    UConverterSafeClone safeClone;
+    UConverterGetUnicodeSet getUnicodeSet;
+
+    UConverterConvert toUTF8;
+    UConverterConvert fromUTF8;
+};
+
+extern const UConverterSharedData
+    _MBCSData, _Latin1Data,
+    _UTF8Data, _UTF16BEData, _UTF16LEData, _UTF32BEData, _UTF32LEData,
+    _ISO2022Data, 
+    _LMBCSData1,_LMBCSData2, _LMBCSData3, _LMBCSData4, _LMBCSData5, _LMBCSData6,
+    _LMBCSData8,_LMBCSData11,_LMBCSData16,_LMBCSData17,_LMBCSData18,_LMBCSData19,
+    _HZData,_ISCIIData, _SCSUData, _ASCIIData,
+    _UTF7Data, _Bocu1Data, _UTF16Data, _UTF32Data, _CESU8Data, _IMAPData, _CompoundTextData;
+
+U_CDECL_END
+
+/** Always use fallbacks from codepage to Unicode */
+#define TO_U_USE_FALLBACK(useFallback) true
+#define UCNV_TO_U_USE_FALLBACK(cnv) true
+
+/** Use fallbacks from Unicode to codepage when cnv->useFallback or for private-use code points */
+#define IS_PRIVATE_USE(c) ((uint32_t)((c)-0xe000)<0x1900 || (uint32_t)((c)-0xf0000)<0x20000)
+#define FROM_U_USE_FALLBACK(useFallback, c) ((useFallback) || IS_PRIVATE_USE(c))
+#define UCNV_FROM_U_USE_FALLBACK(cnv, c) FROM_U_USE_FALLBACK((cnv)->useFallback, c)
+
+/**
+ * Magic number for ucnv_getNextUChar(), returned by a
+ * getNextUChar() implementation to indicate to use the converter's toUnicode()
+ * instead of the native function.
+ * @internal
+ */
+#define UCNV_GET_NEXT_UCHAR_USE_TO_U -9
+
+U_CFUNC void
+ucnv_getCompleteUnicodeSet(const UConverter *cnv,
+                   const USetAdder *sa,
+                   UConverterUnicodeSet which,
+                   UErrorCode *pErrorCode);
+
+U_CFUNC void
+ucnv_getNonSurrogateUnicodeSet(const UConverter *cnv,
+                               const USetAdder *sa,
+                               UConverterUnicodeSet which,
+                               UErrorCode *pErrorCode);
+
+U_CFUNC void
+ucnv_fromUWriteBytes(UConverter *cnv,
+                     const char *bytes, int32_t length,
+                     char **target, const char *targetLimit,
+                     int32_t **offsets,
+                     int32_t sourceIndex,
+                     UErrorCode *pErrorCode);
+U_CFUNC void
+ucnv_toUWriteUChars(UConverter *cnv,
+                    const UChar *uchars, int32_t length,
+                    UChar **target, const UChar *targetLimit,
+                    int32_t **offsets,
+                    int32_t sourceIndex,
+                    UErrorCode *pErrorCode);
+
+U_CFUNC void
+ucnv_toUWriteCodePoint(UConverter *cnv,
+                       UChar32 c,
+                       UChar **target, const UChar *targetLimit,
+                       int32_t **offsets,
+                       int32_t sourceIndex,
+                       UErrorCode *pErrorCode);
+
+#endif
+
+#endif /* UCNV_CNV */
diff --git a/thirdparty/icu4c/common/ucnv_ct.cpp b/thirdparty/icu4c/common/ucnv_ct.cpp
new file mode 100644
index 0000000000..b40e1b2c97
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_ct.cpp
@@ -0,0 +1,646 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2010-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_ct.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010Dec09
+*   created by: Michael Ow
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/uset.h"
+#include "unicode/ucnv_err.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/utf16.h"
+#include "ucnv_imp.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "ucnvmbcs.h"
+#include "cstring.h"
+#include "cmemory.h"
+
+typedef enum {
+    INVALID = -2,
+    DO_SEARCH = -1,
+
+    COMPOUND_TEXT_SINGLE_0 = 0,
+    COMPOUND_TEXT_SINGLE_1 = 1,
+    COMPOUND_TEXT_SINGLE_2 = 2,
+    COMPOUND_TEXT_SINGLE_3 = 3,
+
+    COMPOUND_TEXT_DOUBLE_1 = 4,
+    COMPOUND_TEXT_DOUBLE_2 = 5,
+    COMPOUND_TEXT_DOUBLE_3 = 6,
+    COMPOUND_TEXT_DOUBLE_4 = 7,
+    COMPOUND_TEXT_DOUBLE_5 = 8,
+    COMPOUND_TEXT_DOUBLE_6 = 9,
+    COMPOUND_TEXT_DOUBLE_7 = 10,
+
+    COMPOUND_TEXT_TRIPLE_DOUBLE = 11,
+
+    IBM_915 = 12,
+    IBM_916 = 13,
+    IBM_914 = 14,
+    IBM_874 = 15,
+    IBM_912 = 16,
+    IBM_913 = 17,
+    ISO_8859_14 = 18,
+    IBM_923 = 19,
+    NUM_OF_CONVERTERS = 20
+} COMPOUND_TEXT_CONVERTERS;
+
+#define SEARCH_LENGTH 12
+
+static const uint8_t escSeqCompoundText[NUM_OF_CONVERTERS][5] = {
+    /* Single */
+    { 0x1B, 0x2D, 0x41, 0, 0 },
+    { 0x1B, 0x2D, 0x4D, 0, 0 },
+    { 0x1B, 0x2D, 0x46, 0, 0 },
+    { 0x1B, 0x2D, 0x47, 0, 0 },
+
+    /* Double */
+    { 0x1B, 0x24, 0x29, 0x41, 0 },
+    { 0x1B, 0x24, 0x29, 0x42, 0 },
+    { 0x1B, 0x24, 0x29, 0x43, 0 },
+    { 0x1B, 0x24, 0x29, 0x44, 0 },
+    { 0x1B, 0x24, 0x29, 0x47, 0 },
+    { 0x1B, 0x24, 0x29, 0x48, 0 },
+    { 0x1B, 0x24, 0x29, 0x49, 0 },
+
+    /* Triple/Double */
+    { 0x1B, 0x25, 0x47, 0, 0 },
+
+    /*IBM-915*/
+    { 0x1B, 0x2D, 0x4C, 0, 0 },
+    /*IBM-916*/
+    { 0x1B, 0x2D, 0x48, 0, 0 },
+    /*IBM-914*/
+    { 0x1B, 0x2D, 0x44, 0, 0 },
+    /*IBM-874*/
+    { 0x1B, 0x2D, 0x54, 0, 0 },
+    /*IBM-912*/
+    { 0x1B, 0x2D, 0x42, 0, 0 },
+    /* IBM-913 */
+    { 0x1B, 0x2D, 0x43, 0, 0 },
+    /* ISO-8859_14 */
+    { 0x1B, 0x2D, 0x5F, 0, 0 },
+    /* IBM-923 */
+    { 0x1B, 0x2D, 0x62, 0, 0 },
+};
+
+#define ESC_START 0x1B
+
+#define isASCIIRange(codepoint) \
+        ((codepoint == 0x0000) || (codepoint == 0x0009) || (codepoint == 0x000A) || \
+         (codepoint >= 0x0020 && codepoint <= 0x007f) || (codepoint >= 0x00A0 && codepoint <= 0x00FF))
+
+#define isIBM915(codepoint) \
+        ((codepoint >= 0x0401 && codepoint <= 0x045F) || (codepoint == 0x2116))
+
+#define isIBM916(codepoint) \
+        ((codepoint >= 0x05D0 && codepoint <= 0x05EA) || (codepoint == 0x2017) || (codepoint == 0x203E))
+
+#define isCompoundS3(codepoint) \
+        ((codepoint == 0x060C) || (codepoint == 0x061B) || (codepoint == 0x061F) || (codepoint >= 0x0621 && codepoint <= 0x063A) || \
+         (codepoint >= 0x0640 && codepoint <= 0x0652) || (codepoint >= 0x0660 && codepoint <= 0x066D) || (codepoint == 0x200B) || \
+         (codepoint >= 0x0FE70 && codepoint <= 0x0FE72) || (codepoint == 0x0FE74) || (codepoint >= 0x0FE76 && codepoint <= 0x0FEBE))
+
+#define isCompoundS2(codepoint) \
+        ((codepoint == 0x02BC) || (codepoint == 0x02BD) || (codepoint >= 0x0384 && codepoint <= 0x03CE) || (codepoint == 0x2015))
+
+#define isIBM914(codepoint) \
+        ((codepoint == 0x0100) || (codepoint == 0x0101) || (codepoint == 0x0112) || (codepoint == 0x0113) || (codepoint == 0x0116) || (codepoint == 0x0117) || \
+         (codepoint == 0x0122) || (codepoint == 0x0123) || (codepoint >= 0x0128 && codepoint <= 0x012B) || (codepoint == 0x012E) || (codepoint == 0x012F) || \
+         (codepoint >= 0x0136 && codepoint <= 0x0138) || (codepoint == 0x013B) || (codepoint == 0x013C) || (codepoint == 0x0145) || (codepoint ==  0x0146) || \
+         (codepoint >= 0x014A && codepoint <= 0x014D) || (codepoint == 0x0156) || (codepoint == 0x0157) || (codepoint >= 0x0166 && codepoint <= 0x016B) || \
+         (codepoint == 0x0172) || (codepoint == 0x0173))
+
+#define isIBM874(codepoint) \
+        ((codepoint >= 0x0E01 && codepoint <= 0x0E3A) || (codepoint >= 0x0E3F && codepoint <= 0x0E5B))
+
+#define isIBM912(codepoint) \
+        ((codepoint >= 0x0102 && codepoint <= 0x0107) || (codepoint >= 0x010C && codepoint <= 0x0111) || (codepoint >= 0x0118 && codepoint <= 0x011B) || \
+         (codepoint == 0x0139) || (codepoint == 0x013A) || (codepoint == 0x013D) || (codepoint == 0x013E) || (codepoint >= 0x0141 && codepoint <= 0x0144) || \
+         (codepoint == 0x0147) || (codepoint == 0x0147) || (codepoint == 0x0150) || (codepoint == 0x0151) || (codepoint == 0x0154) || (codepoint == 0x0155) || \
+         (codepoint >= 0x0158 && codepoint <= 0x015B) || (codepoint == 0x015E) || (codepoint == 0x015F) || (codepoint >= 0x0160 && codepoint <= 0x0165) || \
+         (codepoint == 0x016E) || (codepoint == 0x016F) || (codepoint == 0x0170) || (codepoint ==  0x0171) || (codepoint >= 0x0179 && codepoint <= 0x017E) || \
+         (codepoint == 0x02C7) || (codepoint == 0x02D8) || (codepoint == 0x02D9) || (codepoint == 0x02DB) || (codepoint == 0x02DD))
+
+#define isIBM913(codepoint) \
+        ((codepoint >= 0x0108 && codepoint <= 0x010B) || (codepoint == 0x011C) || \
+         (codepoint == 0x011D) || (codepoint == 0x0120) || (codepoint == 0x0121) || \
+         (codepoint >= 0x0124 && codepoint <= 0x0127) || (codepoint == 0x0134) || (codepoint == 0x0135) || \
+         (codepoint == 0x015C) || (codepoint == 0x015D) || (codepoint == 0x016C) || (codepoint ==  0x016D))
+
+#define isCompoundS1(codepoint) \
+        ((codepoint == 0x011E) || (codepoint == 0x011F) || (codepoint == 0x0130) || \
+         (codepoint == 0x0131) || (codepoint >= 0x0218 && codepoint <= 0x021B))
+
+#define isISO8859_14(codepoint) \
+        ((codepoint >= 0x0174 && codepoint <= 0x0177) || (codepoint == 0x1E0A) || \
+         (codepoint == 0x1E0B) || (codepoint == 0x1E1E) || (codepoint == 0x1E1F) || \
+         (codepoint == 0x1E40) || (codepoint == 0x1E41) || (codepoint == 0x1E56) || \
+         (codepoint == 0x1E57) || (codepoint == 0x1E60) || (codepoint == 0x1E61) || \
+         (codepoint == 0x1E6A) || (codepoint == 0x1E6B) || (codepoint == 0x1EF2) || \
+         (codepoint == 0x1EF3) || (codepoint >= 0x1E80 && codepoint <= 0x1E85))
+
+#define isIBM923(codepoint) \
+        ((codepoint == 0x0152) || (codepoint == 0x0153) || (codepoint == 0x0178) || (codepoint == 0x20AC))
+
+
+typedef struct{
+    UConverterSharedData *myConverterArray[NUM_OF_CONVERTERS];
+    COMPOUND_TEXT_CONVERTERS state;
+} UConverterDataCompoundText;
+
+/*********** Compound Text Converter Protos ***********/
+U_CDECL_BEGIN
+static void U_CALLCONV
+_CompoundTextOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode);
+
+static void U_CALLCONV
+ _CompoundTextClose(UConverter *converter);
+
+static void U_CALLCONV
+_CompoundTextReset(UConverter *converter, UConverterResetChoice choice);
+
+static const char* U_CALLCONV
+_CompoundTextgetName(const UConverter* cnv);
+
+
+static int32_t findNextEsc(const char *source, const char *sourceLimit) {
+    int32_t length = static_cast<int32_t>(sourceLimit - source);
+    int32_t i;
+    for (i = 1; i < length; i++) {
+        if (*(source + i) == 0x1B) {
+            return i;
+        }
+    }
+
+    return length;
+}
+
+static COMPOUND_TEXT_CONVERTERS getState(int codepoint) {
+    COMPOUND_TEXT_CONVERTERS state = DO_SEARCH;
+
+    if (isASCIIRange(codepoint)) {
+        state = COMPOUND_TEXT_SINGLE_0;
+    } else if (isIBM912(codepoint)) {
+        state = IBM_912;
+    }else if (isIBM913(codepoint)) {
+        state = IBM_913;
+    } else if (isISO8859_14(codepoint)) {
+        state = ISO_8859_14;
+    } else if (isIBM923(codepoint)) {
+        state = IBM_923;
+    } else if (isIBM874(codepoint)) {
+        state = IBM_874;
+    } else if (isIBM914(codepoint)) {
+        state = IBM_914;
+    } else if (isCompoundS2(codepoint)) {
+        state = COMPOUND_TEXT_SINGLE_2;
+    } else if (isCompoundS3(codepoint)) {
+        state = COMPOUND_TEXT_SINGLE_3;
+    } else if (isIBM916(codepoint)) {
+        state = IBM_916;
+    } else if (isIBM915(codepoint)) {
+        state = IBM_915;
+    } else if (isCompoundS1(codepoint)) {
+        state = COMPOUND_TEXT_SINGLE_1;
+    }
+
+    return state;
+}
+
+static COMPOUND_TEXT_CONVERTERS findStateFromEscSeq(const char* source, const char* sourceLimit, const uint8_t* toUBytesBuffer, int32_t toUBytesBufferLength, UErrorCode *err) {
+    COMPOUND_TEXT_CONVERTERS state = INVALID;
+    UBool matchFound = FALSE;
+    int32_t i, n, offset = toUBytesBufferLength;
+
+    for (i = 0; i < NUM_OF_CONVERTERS; i++) {
+        matchFound = TRUE;
+        for (n = 0; escSeqCompoundText[i][n] != 0; n++) {
+            if (n < toUBytesBufferLength) {
+                if (toUBytesBuffer[n] != escSeqCompoundText[i][n]) {
+                    matchFound = FALSE;
+                    break;
+                }
+            } else if ((source + (n - offset)) >= sourceLimit) {
+                *err = U_TRUNCATED_CHAR_FOUND;
+                matchFound = FALSE;
+                break;
+            } else if (*(source + (n - offset)) != escSeqCompoundText[i][n]) {
+                matchFound = FALSE;
+                break;
+            }
+        }
+
+        if (matchFound) {
+            break;
+        }
+    }
+
+    if (matchFound) {
+        state = (COMPOUND_TEXT_CONVERTERS)i;
+    }
+
+    return state;
+}
+
+static void U_CALLCONV
+_CompoundTextOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode){
+    cnv->extraInfo = uprv_malloc (sizeof (UConverterDataCompoundText));
+    if (cnv->extraInfo != NULL) {
+        UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) cnv->extraInfo;
+
+        UConverterNamePieces stackPieces;
+        UConverterLoadArgs stackArgs=UCNV_LOAD_ARGS_INITIALIZER;
+
+        myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_0] = NULL;
+        myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_1] = ucnv_loadSharedData("icu-internal-compound-s1", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_2] = ucnv_loadSharedData("icu-internal-compound-s2", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_3] = ucnv_loadSharedData("icu-internal-compound-s3", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_1] = ucnv_loadSharedData("icu-internal-compound-d1", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_2] = ucnv_loadSharedData("icu-internal-compound-d2", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_3] = ucnv_loadSharedData("icu-internal-compound-d3", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_4] = ucnv_loadSharedData("icu-internal-compound-d4", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_5] = ucnv_loadSharedData("icu-internal-compound-d5", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_6] = ucnv_loadSharedData("icu-internal-compound-d6", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_7] = ucnv_loadSharedData("icu-internal-compound-d7", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[COMPOUND_TEXT_TRIPLE_DOUBLE] = ucnv_loadSharedData("icu-internal-compound-t", &stackPieces, &stackArgs, errorCode);
+
+        myConverterData->myConverterArray[IBM_915] = ucnv_loadSharedData("ibm-915_P100-1995", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_916] = ucnv_loadSharedData("ibm-916_P100-1995", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_914] = ucnv_loadSharedData("ibm-914_P100-1995", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_874] = ucnv_loadSharedData("ibm-874_P100-1995", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_912] = ucnv_loadSharedData("ibm-912_P100-1995", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_913] = ucnv_loadSharedData("ibm-913_P100-2000", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[ISO_8859_14] = ucnv_loadSharedData("iso-8859_14-1998", &stackPieces, &stackArgs, errorCode);
+        myConverterData->myConverterArray[IBM_923] = ucnv_loadSharedData("ibm-923_P100-1998", &stackPieces, &stackArgs, errorCode);
+
+        if (U_FAILURE(*errorCode) || pArgs->onlyTestIsLoadable) {
+            _CompoundTextClose(cnv);
+            return;
+        }
+
+        myConverterData->state = (COMPOUND_TEXT_CONVERTERS)0;
+    } else {
+        *errorCode = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+
+static void U_CALLCONV
+_CompoundTextClose(UConverter *converter) {
+    UConverterDataCompoundText* myConverterData = (UConverterDataCompoundText*)(converter->extraInfo);
+    int32_t i;
+
+    if (converter->extraInfo != NULL) {
+        /*close the array of converter pointers and free the memory*/
+        for (i = 0; i < NUM_OF_CONVERTERS; i++) {
+            if (myConverterData->myConverterArray[i] != NULL) {
+                ucnv_unloadSharedDataIfReady(myConverterData->myConverterArray[i]);
+            }
+        }
+
+        uprv_free(converter->extraInfo);
+        converter->extraInfo = NULL;
+    }
+}
+
+static void U_CALLCONV
+_CompoundTextReset(UConverter *converter, UConverterResetChoice choice) {
+    (void)converter;
+    (void)choice;
+}
+
+static const char* U_CALLCONV
+_CompoundTextgetName(const UConverter* cnv){
+    (void)cnv;
+    return "x11-compound-text";
+}
+
+static void U_CALLCONV
+UConverter_fromUnicode_CompoundText_OFFSETS(UConverterFromUnicodeArgs* args, UErrorCode* err){
+    UConverter *cnv = args->converter;
+    uint8_t *target = (uint8_t *) args->target;
+    const uint8_t *targetLimit = (const uint8_t *) args->targetLimit;
+    const UChar* source = args->source;
+    const UChar* sourceLimit = args->sourceLimit;
+    /* int32_t* offsets = args->offsets; */
+    UChar32 sourceChar;
+    UBool useFallback = cnv->useFallback;
+    uint8_t tmpTargetBuffer[7];
+    int32_t tmpTargetBufferLength = 0;
+    COMPOUND_TEXT_CONVERTERS currentState, tmpState;
+    uint32_t pValue;
+    int32_t pValueLength = 0;
+    int32_t i, n, j;
+
+    UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) cnv->extraInfo;
+
+    currentState = myConverterData->state;
+
+    /* check if the last codepoint of previous buffer was a lead surrogate*/
+    if((sourceChar = cnv->fromUChar32)!=0 && target< targetLimit) {
+        goto getTrail;
+    }
+
+    while( source < sourceLimit){
+        if(target < targetLimit){
+
+            sourceChar  = *(source++);
+            /*check if the char is a First surrogate*/
+             if(U16_IS_SURROGATE(sourceChar)) {
+                if(U16_IS_SURROGATE_LEAD(sourceChar)) {
+getTrail:
+                    /*look ahead to find the trail surrogate*/
+                    if(source < sourceLimit) {
+                        /* test the following code unit */
+                        UChar trail=(UChar) *source;
+                        if(U16_IS_TRAIL(trail)) {
+                            source++;
+                            sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
+                            cnv->fromUChar32=0x00;
+                            /* convert this supplementary code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *err=U_ILLEGAL_CHAR_FOUND;
+                            cnv->fromUChar32=sourceChar;
+                            break;
+                        }
+                    } else {
+                        /* no more input */
+                        cnv->fromUChar32=sourceChar;
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *err=U_ILLEGAL_CHAR_FOUND;
+                    cnv->fromUChar32=sourceChar;
+                    break;
+                }
+            }
+
+             tmpTargetBufferLength = 0;
+             tmpState = getState(sourceChar);
+
+             if (tmpState != DO_SEARCH && currentState != tmpState) {
+                 /* Get escape sequence if necessary */
+                 currentState = tmpState;
+                 for (i = 0; escSeqCompoundText[currentState][i] != 0; i++) {
+                     tmpTargetBuffer[tmpTargetBufferLength++] = escSeqCompoundText[currentState][i];
+                 }
+             }
+
+             if (tmpState == DO_SEARCH) {
+                 /* Test all available converters */
+                 for (i = 1; i < SEARCH_LENGTH; i++) {
+                     pValueLength = ucnv_MBCSFromUChar32(myConverterData->myConverterArray[i], sourceChar, &pValue, useFallback);
+                     if (pValueLength > 0) {
+                         tmpState = (COMPOUND_TEXT_CONVERTERS)i;
+                         if (currentState != tmpState) {
+                             currentState = tmpState;
+                             for (j = 0; escSeqCompoundText[currentState][j] != 0; j++) {
+                                 tmpTargetBuffer[tmpTargetBufferLength++] = escSeqCompoundText[currentState][j];
+                             }
+                         }
+                         for (n = (pValueLength - 1); n >= 0; n--) {
+                             tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)(pValue >> (n * 8));
+                         }
+                         break;
+                     }
+                 }
+             } else if (tmpState == COMPOUND_TEXT_SINGLE_0) {
+                 tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)sourceChar;
+             } else {
+                 pValueLength = ucnv_MBCSFromUChar32(myConverterData->myConverterArray[currentState], sourceChar, &pValue, useFallback);
+                 if (pValueLength > 0) {
+                     for (n = (pValueLength - 1); n >= 0; n--) {
+                         tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)(pValue >> (n * 8));
+                     }
+                 }
+             }
+
+             for (i = 0; i < tmpTargetBufferLength; i++) {
+                 if (target < targetLimit) {
+                     *target++ = tmpTargetBuffer[i];
+                 } else {
+                     *err = U_BUFFER_OVERFLOW_ERROR;
+                     break;
+                 }
+             }
+
+             if (*err == U_BUFFER_OVERFLOW_ERROR) {
+                 for (; i < tmpTargetBufferLength; i++) {
+                     args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = tmpTargetBuffer[i];
+                 }
+             }
+        } else {
+            *err = U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /*save the state and return */
+    myConverterData->state = currentState;
+    args->source = source;
+    args->target = (char*)target;
+}
+
+
+static void U_CALLCONV
+UConverter_toUnicode_CompoundText_OFFSETS(UConverterToUnicodeArgs *args,
+                                               UErrorCode* err){
+    const char *mySource = (char *) args->source;
+    UChar *myTarget = args->target;
+    const char *mySourceLimit = args->sourceLimit;
+    const char *tmpSourceLimit = mySourceLimit;
+    uint32_t mySourceChar = 0x0000;
+    COMPOUND_TEXT_CONVERTERS currentState, tmpState;
+    int32_t sourceOffset = 0;
+    UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) args->converter->extraInfo;
+    UConverterSharedData* savedSharedData = NULL;
+
+    UConverterToUnicodeArgs subArgs;
+    int32_t minArgsSize;
+
+    /* set up the subconverter arguments */
+    if(args->size<sizeof(UConverterToUnicodeArgs)) {
+        minArgsSize = args->size;
+    } else {
+        minArgsSize = (int32_t)sizeof(UConverterToUnicodeArgs);
+    }
+
+    uprv_memcpy(&subArgs, args, minArgsSize);
+    subArgs.size = (uint16_t)minArgsSize;
+
+    currentState = tmpState =  myConverterData->state;
+
+    while(mySource < mySourceLimit){
+        if(myTarget < args->targetLimit){
+            if (args->converter->toULength > 0) {
+                mySourceChar = args->converter->toUBytes[0];
+            } else {
+                mySourceChar = (uint8_t)*mySource;
+            }
+
+            if (mySourceChar == ESC_START) {
+                tmpState = findStateFromEscSeq(mySource, mySourceLimit, args->converter->toUBytes, args->converter->toULength, err);
+
+                if (*err == U_TRUNCATED_CHAR_FOUND) {
+                    for (; mySource < mySourceLimit;) {
+                        args->converter->toUBytes[args->converter->toULength++] = *mySource++;
+                    }
+                    *err = U_ZERO_ERROR;
+                    break;
+                } else if (tmpState == INVALID) {
+                    if (args->converter->toULength == 0) {
+                        mySource++; /* skip over the 0x1b byte */
+                    }
+                    *err = U_ILLEGAL_CHAR_FOUND;
+                    break;
+                }
+
+                if (tmpState != currentState) {
+                    currentState = tmpState;
+                }
+
+                sourceOffset = static_cast<int32_t>(uprv_strlen((char*)escSeqCompoundText[currentState]) - args->converter->toULength);
+
+                mySource += sourceOffset;
+
+                args->converter->toULength = 0;
+            }
+
+            if (currentState == COMPOUND_TEXT_SINGLE_0) {
+                while (mySource < mySourceLimit) {
+                    if (*mySource == ESC_START) {
+                        break;
+                    }
+                    if (myTarget < args->targetLimit) {
+                        *myTarget++ = 0x00ff&(*mySource++);
+                    } else {
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                        break;
+                    }
+                }
+            } else if (mySource < mySourceLimit){
+                sourceOffset = findNextEsc(mySource, mySourceLimit);
+
+                tmpSourceLimit = mySource + sourceOffset;
+
+                subArgs.source = mySource;
+                subArgs.sourceLimit = tmpSourceLimit;
+                subArgs.target = myTarget;
+                savedSharedData = subArgs.converter->sharedData;
+                subArgs.converter->sharedData = myConverterData->myConverterArray[currentState];
+
+                ucnv_MBCSToUnicodeWithOffsets(&subArgs, err);
+
+                subArgs.converter->sharedData = savedSharedData;
+
+                mySource = subArgs.source;
+                myTarget = subArgs.target;
+
+                if (U_FAILURE(*err)) {
+                    if(*err == U_BUFFER_OVERFLOW_ERROR) {
+                        if(subArgs.converter->UCharErrorBufferLength > 0) {
+                            uprv_memcpy(args->converter->UCharErrorBuffer, subArgs.converter->UCharErrorBuffer,
+                                        subArgs.converter->UCharErrorBufferLength);
+                        }
+                        args->converter->UCharErrorBufferLength=subArgs.converter->UCharErrorBufferLength;
+                        subArgs.converter->UCharErrorBufferLength = 0;
+                    }
+                    break;
+                }
+            }
+        } else {
+            *err = U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+    myConverterData->state = currentState;
+    args->target = myTarget;
+    args->source = mySource;
+}
+
+static void U_CALLCONV
+_CompoundText_GetUnicodeSet(const UConverter *cnv,
+                    const USetAdder *sa,
+                    UConverterUnicodeSet which,
+                    UErrorCode *pErrorCode) {
+    UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *)cnv->extraInfo;
+    int32_t i;
+
+    for (i = 1; i < NUM_OF_CONVERTERS; i++) {
+        ucnv_MBCSGetUnicodeSetForUnicode(myConverterData->myConverterArray[i], sa, which, pErrorCode);
+    }
+    sa->add(sa->set, 0x0000);
+    sa->add(sa->set, 0x0009);
+    sa->add(sa->set, 0x000A);
+    sa->addRange(sa->set, 0x0020, 0x007F);
+    sa->addRange(sa->set, 0x00A0, 0x00FF);
+}
+U_CDECL_END
+
+static const UConverterImpl _CompoundTextImpl = {
+
+    UCNV_COMPOUND_TEXT,
+
+    NULL,
+    NULL,
+
+    _CompoundTextOpen,
+    _CompoundTextClose,
+    _CompoundTextReset,
+
+    UConverter_toUnicode_CompoundText_OFFSETS,
+    UConverter_toUnicode_CompoundText_OFFSETS,
+    UConverter_fromUnicode_CompoundText_OFFSETS,
+    UConverter_fromUnicode_CompoundText_OFFSETS,
+    NULL,
+
+    NULL,
+    _CompoundTextgetName,
+    NULL,
+    NULL,
+    _CompoundText_GetUnicodeSet,
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _CompoundTextStaticData = {
+    sizeof(UConverterStaticData),
+    "COMPOUND_TEXT",
+    0,
+    UCNV_IBM,
+    UCNV_COMPOUND_TEXT,
+    1,
+    6,
+    { 0xef, 0, 0, 0 },
+    1,
+    FALSE,
+    FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+const UConverterSharedData _CompoundTextData =
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_CompoundTextStaticData, &_CompoundTextImpl);
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnv_err.cpp b/thirdparty/icu4c/common/ucnv_err.cpp
new file mode 100644
index 0000000000..6b738face5
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_err.cpp
@@ -0,0 +1,486 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *****************************************************************************
+ *
+ *   Copyright (C) 1998-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *****************************************************************************
+ *
+ *  ucnv_err.c
+ *  Implements error behaviour functions called by T_UConverter_{from,to}Unicode
+ *
+ *
+*   Change history:
+*
+*   06/29/2000  helena      Major rewrite of the callback APIs.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv_err.h"
+#include "unicode/ucnv_cb.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+#include "unicode/ucnv.h"
+#include "ustrfmt.h"
+
+#define VALUE_STRING_LENGTH 48
+/*Magic # 32 = 4(number of char in value string) * 8(max number of bytes per char for any converter) */
+#define UNICODE_PERCENT_SIGN_CODEPOINT  0x0025
+#define UNICODE_U_CODEPOINT             0x0055
+#define UNICODE_X_CODEPOINT             0x0058
+#define UNICODE_RS_CODEPOINT            0x005C
+#define UNICODE_U_LOW_CODEPOINT         0x0075
+#define UNICODE_X_LOW_CODEPOINT         0x0078
+#define UNICODE_AMP_CODEPOINT           0x0026
+#define UNICODE_HASH_CODEPOINT          0x0023
+#define UNICODE_SEMICOLON_CODEPOINT     0x003B
+#define UNICODE_PLUS_CODEPOINT          0x002B
+#define UNICODE_LEFT_CURLY_CODEPOINT    0x007B
+#define UNICODE_RIGHT_CURLY_CODEPOINT   0x007D
+#define UNICODE_SPACE_CODEPOINT         0x0020
+#define UCNV_PRV_ESCAPE_ICU         0
+#define UCNV_PRV_ESCAPE_C           'C'
+#define UCNV_PRV_ESCAPE_XML_DEC     'D'
+#define UCNV_PRV_ESCAPE_XML_HEX     'X'
+#define UCNV_PRV_ESCAPE_JAVA        'J'
+#define UCNV_PRV_ESCAPE_UNICODE     'U'
+#define UCNV_PRV_ESCAPE_CSS2        'S'
+#define UCNV_PRV_STOP_ON_ILLEGAL    'i'
+
+/*
+ * IS_DEFAULT_IGNORABLE_CODE_POINT
+ * This is to check if a code point has the default ignorable unicode property.
+ * As such, this list needs to be updated if the ignorable code point list ever
+ * changes.
+ * To avoid dependency on other code, this list is hard coded here.
+ * When an ignorable code point is found and is unmappable, the default callbacks
+ * will ignore them.
+ * For a list of the default ignorable code points, use this link:
+ * https://unicode.org/cldr/utility/list-unicodeset.jsp?a=%5B%3ADI%3A%5D&abb=on&g=&i=
+ *
+ * This list should be sync with the one in CharsetCallback.java
+ */
+#define IS_DEFAULT_IGNORABLE_CODE_POINT(c) ( \
+    (c == 0x00AD) || \
+    (c == 0x034F) || \
+    (c == 0x061C) || \
+    (c == 0x115F) || \
+    (c == 0x1160) || \
+    (0x17B4 <= c && c <= 0x17B5) || \
+    (0x180B <= c && c <= 0x180E) || \
+    (0x200B <= c && c <= 0x200F) || \
+    (0x202A <= c && c <= 0x202E) || \
+    (0x2060 <= c && c <= 0x206F) || \
+    (c == 0x3164) || \
+    (0xFE00 <= c && c <= 0xFE0F) || \
+    (c == 0xFEFF) || \
+    (c == 0xFFA0) || \
+    (0xFFF0 <= c && c <= 0xFFF8) || \
+    (0x1BCA0 <= c && c <= 0x1BCA3) || \
+    (0x1D173 <= c && c <= 0x1D17A) || \
+    (0xE0000 <= c && c <= 0xE0FFF))
+
+
+/*Function Pointer STOPS at the ILLEGAL_SEQUENCE */
+U_CAPI void    U_EXPORT2
+UCNV_FROM_U_CALLBACK_STOP (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err)
+{
+    (void)context;
+    (void)fromUArgs;
+    (void)codeUnits;
+    (void)length;
+    if (reason == UCNV_UNASSIGNED && IS_DEFAULT_IGNORABLE_CODE_POINT(codePoint))
+    {
+        /*
+         * Skip if the codepoint has unicode property of default ignorable.
+         */
+        *err = U_ZERO_ERROR;
+    }
+    /* the caller must have set the error code accordingly */
+    return;
+}
+
+
+/*Function Pointer STOPS at the ILLEGAL_SEQUENCE */
+U_CAPI void    U_EXPORT2
+UCNV_TO_U_CALLBACK_STOP (
+                   const void *context,
+                   UConverterToUnicodeArgs *toUArgs,
+                   const char* codePoints,
+                   int32_t length,
+                   UConverterCallbackReason reason,
+                   UErrorCode * err)
+{
+    /* the caller must have set the error code accordingly */
+    (void)context; (void)toUArgs; (void)codePoints; (void)length; (void)reason; (void)err;
+    return;
+}
+
+U_CAPI void    U_EXPORT2
+UCNV_FROM_U_CALLBACK_SKIP (                  
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err)
+{
+    (void)fromUArgs;
+    (void)codeUnits;
+    (void)length;
+    if (reason <= UCNV_IRREGULAR)
+    {
+        if (reason == UCNV_UNASSIGNED && IS_DEFAULT_IGNORABLE_CODE_POINT(codePoint))
+        {
+            /*
+             * Skip if the codepoint has unicode property of default ignorable.
+             */
+            *err = U_ZERO_ERROR;
+        }
+        else if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
+        {
+            *err = U_ZERO_ERROR;
+        }
+        /* else the caller must have set the error code accordingly. */
+    }
+    /* else ignore the reset, close and clone calls. */
+}
+
+U_CAPI void    U_EXPORT2
+UCNV_FROM_U_CALLBACK_SUBSTITUTE (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err)
+{
+    (void)codeUnits;
+    (void)length;
+    if (reason <= UCNV_IRREGULAR)
+    {
+        if (reason == UCNV_UNASSIGNED && IS_DEFAULT_IGNORABLE_CODE_POINT(codePoint))
+        {
+            /*
+             * Skip if the codepoint has unicode property of default ignorable.
+             */
+            *err = U_ZERO_ERROR;
+        }
+        else if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
+        {
+            *err = U_ZERO_ERROR;
+            ucnv_cbFromUWriteSub(fromArgs, 0, err);
+        }
+        /* else the caller must have set the error code accordingly. */
+    }
+    /* else ignore the reset, close and clone calls. */
+}
+
+/*uses uprv_itou to get a unicode escape sequence of the offensive sequence,
+ *uses a clean copy (resetted) of the converter, to convert that unicode
+ *escape sequence to the target codepage (if conversion failure happens then
+ *we revert to substituting with subchar)
+ */
+U_CAPI void    U_EXPORT2
+UCNV_FROM_U_CALLBACK_ESCAPE (
+                         const void *context,
+                         UConverterFromUnicodeArgs *fromArgs,
+                         const UChar *codeUnits,
+                         int32_t length,
+                         UChar32 codePoint,
+                         UConverterCallbackReason reason,
+                         UErrorCode * err)
+{
+
+  UChar valueString[VALUE_STRING_LENGTH];
+  int32_t valueStringLength = 0;
+  int32_t i = 0;
+
+  const UChar *myValueSource = NULL;
+  UErrorCode err2 = U_ZERO_ERROR;
+  UConverterFromUCallback original = NULL;
+  const void *originalContext;
+
+  UConverterFromUCallback ignoredCallback = NULL;
+  const void *ignoredContext;
+  
+  if (reason > UCNV_IRREGULAR)
+  {
+      return;
+  }
+  else if (reason == UCNV_UNASSIGNED && IS_DEFAULT_IGNORABLE_CODE_POINT(codePoint))
+  {
+      /*
+       * Skip if the codepoint has unicode property of default ignorable.
+       */
+      *err = U_ZERO_ERROR;
+      return;
+  }
+
+  ucnv_setFromUCallBack (fromArgs->converter,
+                     (UConverterFromUCallback) UCNV_FROM_U_CALLBACK_SUBSTITUTE,
+                     NULL,
+                     &original,
+                     &originalContext,
+                     &err2);
+  
+  if (U_FAILURE (err2))
+  {
+    *err = err2;
+    return;
+  } 
+  if(context==NULL)
+  { 
+      while (i < length)
+      {
+        valueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT;  /* adding % */
+        valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT; /* adding U */
+        valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
+      }
+  }
+  else
+  {
+      switch(*((char*)context))
+      {
+      case UCNV_PRV_ESCAPE_JAVA:
+          while (i < length)
+          {
+              valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
+              valueString[valueStringLength++] = (UChar) UNICODE_U_LOW_CODEPOINT; /* adding u */
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
+          }
+          break;
+
+      case UCNV_PRV_ESCAPE_C:
+          valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
+
+          if(length==2){
+              valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT; /* adding U */
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 8);
+
+          }
+          else{
+              valueString[valueStringLength++] = (UChar) UNICODE_U_LOW_CODEPOINT; /* adding u */
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 4);
+          }
+          break;
+
+      case UCNV_PRV_ESCAPE_XML_DEC:
+
+          valueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
+          valueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
+          if(length==2){
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 10, 0);
+          }
+          else{
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 10, 0);
+          }
+          valueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
+          break;
+
+      case UCNV_PRV_ESCAPE_XML_HEX:
+
+          valueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
+          valueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
+          valueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
+          if(length==2){
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 0);
+          }
+          else{
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 0);
+          }
+          valueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
+          break;
+
+      case UCNV_PRV_ESCAPE_UNICODE:
+          valueString[valueStringLength++] = (UChar) UNICODE_LEFT_CURLY_CODEPOINT;    /* adding { */
+          valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT;    /* adding U */
+          valueString[valueStringLength++] = (UChar) UNICODE_PLUS_CODEPOINT; /* adding + */
+          if (length == 2) {
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 4);
+          } else {
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 4);
+          }
+          valueString[valueStringLength++] = (UChar) UNICODE_RIGHT_CURLY_CODEPOINT;    /* adding } */
+          break;
+
+      case UCNV_PRV_ESCAPE_CSS2:
+          valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
+          valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 0);
+          /* Always add space character, becase the next character might be whitespace,
+             which would erroneously be considered the termination of the escape sequence. */
+          valueString[valueStringLength++] = (UChar) UNICODE_SPACE_CODEPOINT;
+          break;
+
+      default:
+          while (i < length)
+          {
+              valueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT;  /* adding % */
+              valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT;             /* adding U */
+              valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
+          }
+      }
+  }  
+  myValueSource = valueString;
+
+  /* reset the error */
+  *err = U_ZERO_ERROR;
+
+  ucnv_cbFromUWriteUChars(fromArgs, &myValueSource, myValueSource+valueStringLength, 0, err);
+
+  ucnv_setFromUCallBack (fromArgs->converter,
+                         original,
+                         originalContext,
+                         &ignoredCallback,
+                         &ignoredContext,
+                         &err2);
+  if (U_FAILURE (err2))
+  {
+      *err = err2;
+      return;
+  }
+
+  return;
+}
+
+
+
+U_CAPI void  U_EXPORT2
+UCNV_TO_U_CALLBACK_SKIP (
+                 const void *context,
+                 UConverterToUnicodeArgs *toArgs,
+                 const char* codeUnits,
+                 int32_t length,
+                 UConverterCallbackReason reason,
+                 UErrorCode * err)
+{
+    (void)toArgs;
+    (void)codeUnits;
+    (void)length;
+    if (reason <= UCNV_IRREGULAR)
+    {
+        if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
+        {
+            *err = U_ZERO_ERROR;
+        }
+        /* else the caller must have set the error code accordingly. */
+    }
+    /* else ignore the reset, close and clone calls. */
+}
+
+U_CAPI void    U_EXPORT2
+UCNV_TO_U_CALLBACK_SUBSTITUTE (
+                 const void *context,
+                 UConverterToUnicodeArgs *toArgs,
+                 const char* codeUnits,
+                 int32_t length,
+                 UConverterCallbackReason reason,
+                 UErrorCode * err)
+{
+    (void)codeUnits;
+    (void)length;
+    if (reason <= UCNV_IRREGULAR)
+    {
+        if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
+        {
+            *err = U_ZERO_ERROR;
+            ucnv_cbToUWriteSub(toArgs,0,err);
+        }
+        /* else the caller must have set the error code accordingly. */
+    }
+    /* else ignore the reset, close and clone calls. */
+}
+
+/*uses uprv_itou to get a unicode escape sequence of the offensive sequence,
+ *and uses that as the substitution sequence
+ */
+U_CAPI void   U_EXPORT2
+UCNV_TO_U_CALLBACK_ESCAPE (
+                 const void *context,
+                 UConverterToUnicodeArgs *toArgs,
+                 const char* codeUnits,
+                 int32_t length,
+                 UConverterCallbackReason reason,
+                 UErrorCode * err)
+{
+    UChar uniValueString[VALUE_STRING_LENGTH];
+    int32_t valueStringLength = 0;
+    int32_t i = 0;
+
+    if (reason > UCNV_IRREGULAR)
+    {
+        return;
+    }
+
+    if(context==NULL)
+    {    
+        while (i < length)
+        {
+            uniValueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */
+            uniValueString[valueStringLength++] = (UChar) UNICODE_X_CODEPOINT;    /* adding X */
+            valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t) codeUnits[i++], 16, 2);
+        }
+    }
+    else
+    {
+        switch(*((char*)context))
+        {
+        case UCNV_PRV_ESCAPE_XML_DEC:
+            while (i < length)
+            {
+                uniValueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
+                uniValueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
+                valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 10, 0);
+                uniValueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
+            }
+            break;
+
+        case UCNV_PRV_ESCAPE_XML_HEX:
+            while (i < length)
+            {
+                uniValueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
+                uniValueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
+                uniValueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
+                valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 16, 0);
+                uniValueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
+            }
+            break;
+        case UCNV_PRV_ESCAPE_C:
+            while (i < length)
+            {
+                uniValueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
+                uniValueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
+                valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 16, 2);
+            }
+            break;
+        default:
+            while (i < length)
+            {
+                uniValueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */
+                uniValueString[valueStringLength++] = (UChar) UNICODE_X_CODEPOINT;    /* adding X */
+                uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t) codeUnits[i++], 16, 2);
+                valueStringLength += 2;
+            }
+        }
+    }
+    /* reset the error */
+    *err = U_ZERO_ERROR;
+
+    ucnv_cbToUWriteUChars(toArgs, uniValueString, valueStringLength, 0, err);
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_ext.cpp b/thirdparty/icu4c/common/ucnv_ext.cpp
new file mode 100644
index 0000000000..7dea4eef41
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_ext.cpp
@@ -0,0 +1,1143 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2003-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnv_ext.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003jun13
+*   created by: Markus W. Scherer
+*
+*   Conversion extensions
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION
+
+#include "unicode/uset.h"
+#include "unicode/ustring.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "ucnv_ext.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+/* to Unicode --------------------------------------------------------------- */
+
+/*
+ * @return lookup value for the byte, if found; else 0
+ */
+static inline uint32_t
+ucnv_extFindToU(const uint32_t *toUSection, int32_t length, uint8_t byte) {
+    uint32_t word0, word;
+    int32_t i, start, limit;
+
+    /* check the input byte against the lowest and highest section bytes */
+    start=(int32_t)UCNV_EXT_TO_U_GET_BYTE(toUSection[0]);
+    limit=(int32_t)UCNV_EXT_TO_U_GET_BYTE(toUSection[length-1]);
+    if(byte<start || limit<byte) {
+        return 0; /* the byte is out of range */
+    }
+
+    if(length==((limit-start)+1)) {
+        /* direct access on a linear array */
+        return UCNV_EXT_TO_U_GET_VALUE(toUSection[byte-start]); /* could be 0 */
+    }
+
+    /* word0 is suitable for <=toUSection[] comparison, word for <toUSection[] */
+    word0=UCNV_EXT_TO_U_MAKE_WORD(byte, 0);
+
+    /*
+     * Shift byte once instead of each section word and add 0xffffff.
+     * We will compare the shifted/added byte (bbffffff) against
+     * section words which have byte values in the same bit position.
+     * If and only if byte bb < section byte ss then bbffffff<ssvvvvvv
+     * for all v=0..f
+     * so we need not mask off the lower 24 bits of each section word.
+     */
+    word=word0|UCNV_EXT_TO_U_VALUE_MASK;
+
+    /* binary search */
+    start=0;
+    limit=length;
+    for(;;) {
+        i=limit-start;
+        if(i<=1) {
+            break; /* done */
+        }
+        /* start<limit-1 */
+
+        if(i<=4) {
+            /* linear search for the last part */
+            if(word0<=toUSection[start]) {
+                break;
+            }
+            if(++start<limit && word0<=toUSection[start]) {
+                break;
+            }
+            if(++start<limit && word0<=toUSection[start]) {
+                break;
+            }
+            /* always break at start==limit-1 */
+            ++start;
+            break;
+        }
+
+        i=(start+limit)/2;
+        if(word<toUSection[i]) {
+            limit=i;
+        } else {
+            start=i;
+        }
+    }
+
+    /* did we really find it? */
+    if(start<limit && byte==UCNV_EXT_TO_U_GET_BYTE(word=toUSection[start])) {
+        return UCNV_EXT_TO_U_GET_VALUE(word); /* never 0 */
+    } else {
+        return 0; /* not found */
+    }
+}
+
+/*
+ * TRUE if not an SI/SO stateful converter,
+ * or if the match length fits with the current converter state
+ */
+#define UCNV_EXT_TO_U_VERIFY_SISO_MATCH(sisoState, match) \
+    ((sisoState)<0 || ((sisoState)==0) == (match==1))
+
+/*
+ * this works like ucnv_extMatchFromU() except
+ * - the first character is in pre
+ * - no trie is used
+ * - the returned matchLength is not offset by 2
+ */
+static int32_t
+ucnv_extMatchToU(const int32_t *cx, int8_t sisoState,
+                 const char *pre, int32_t preLength,
+                 const char *src, int32_t srcLength,
+                 uint32_t *pMatchValue,
+                 UBool /*useFallback*/, UBool flush) {
+    const uint32_t *toUTable, *toUSection;
+
+    uint32_t value, matchValue;
+    int32_t i, j, idx, length, matchLength;
+    uint8_t b;
+
+    if(cx==NULL || cx[UCNV_EXT_TO_U_LENGTH]<=0) {
+        return 0; /* no extension data, no match */
+    }
+
+    /* initialize */
+    toUTable=UCNV_EXT_ARRAY(cx, UCNV_EXT_TO_U_INDEX, uint32_t);
+    idx=0;
+
+    matchValue=0;
+    i=j=matchLength=0;
+
+    if(sisoState==0) {
+        /* SBCS state of an SI/SO stateful converter, look at only exactly 1 byte */
+        if(preLength>1) {
+            return 0; /* no match of a DBCS sequence in SBCS mode */
+        } else if(preLength==1) {
+            srcLength=0;
+        } else /* preLength==0 */ {
+            if(srcLength>1) {
+                srcLength=1;
+            }
+        }
+        flush=TRUE;
+    }
+
+    /* we must not remember fallback matches when not using fallbacks */
+
+    /* match input units until there is a full match or the input is consumed */
+    for(;;) {
+        /* go to the next section */
+        toUSection=toUTable+idx;
+
+        /* read first pair of the section */
+        value=*toUSection++;
+        length=UCNV_EXT_TO_U_GET_BYTE(value);
+        value=UCNV_EXT_TO_U_GET_VALUE(value);
+        if( value!=0 &&
+            (UCNV_EXT_TO_U_IS_ROUNDTRIP(value) ||
+             TO_U_USE_FALLBACK(useFallback)) &&
+            UCNV_EXT_TO_U_VERIFY_SISO_MATCH(sisoState, i+j)
+        ) {
+            /* remember longest match so far */
+            matchValue=value;
+            matchLength=i+j;
+        }
+
+        /* match pre[] then src[] */
+        if(i<preLength) {
+            b=(uint8_t)pre[i++];
+        } else if(j<srcLength) {
+            b=(uint8_t)src[j++];
+        } else {
+            /* all input consumed, partial match */
+            if(flush || (length=(i+j))>UCNV_EXT_MAX_BYTES) {
+                /*
+                 * end of the entire input stream, stop with the longest match so far
+                 * or: partial match must not be longer than UCNV_EXT_MAX_BYTES
+                 * because it must fit into state buffers
+                 */
+                break;
+            } else {
+                /* continue with more input next time */
+                return -length;
+            }
+        }
+
+        /* search for the current UChar */
+        value=ucnv_extFindToU(toUSection, length, b);
+        if(value==0) {
+            /* no match here, stop with the longest match so far */
+            break;
+        } else {
+            if(UCNV_EXT_TO_U_IS_PARTIAL(value)) {
+                /* partial match, continue */
+                idx=(int32_t)UCNV_EXT_TO_U_GET_PARTIAL_INDEX(value);
+            } else {
+                if( (UCNV_EXT_TO_U_IS_ROUNDTRIP(value) ||
+                     TO_U_USE_FALLBACK(useFallback)) &&
+                    UCNV_EXT_TO_U_VERIFY_SISO_MATCH(sisoState, i+j)
+                ) {
+                    /* full match, stop with result */
+                    matchValue=value;
+                    matchLength=i+j;
+                } else {
+                    /* full match on fallback not taken, stop with the longest match so far */
+                }
+                break;
+            }
+        }
+    }
+
+    if(matchLength==0) {
+        /* no match at all */
+        return 0;
+    }
+
+    /* return result */
+    *pMatchValue=UCNV_EXT_TO_U_MASK_ROUNDTRIP(matchValue);
+    return matchLength;
+}
+
+static inline void
+ucnv_extWriteToU(UConverter *cnv, const int32_t *cx,
+                 uint32_t value,
+                 UChar **target, const UChar *targetLimit,
+                 int32_t **offsets, int32_t srcIndex,
+                 UErrorCode *pErrorCode) {
+    /* output the result */
+    if(UCNV_EXT_TO_U_IS_CODE_POINT(value)) {
+        /* output a single code point */
+        ucnv_toUWriteCodePoint(
+            cnv, UCNV_EXT_TO_U_GET_CODE_POINT(value),
+            target, targetLimit,
+            offsets, srcIndex,
+            pErrorCode);
+    } else {
+        /* output a string - with correct data we have resultLength>0 */
+        ucnv_toUWriteUChars(
+            cnv,
+            UCNV_EXT_ARRAY(cx, UCNV_EXT_TO_U_UCHARS_INDEX, UChar)+
+                UCNV_EXT_TO_U_GET_INDEX(value),
+            UCNV_EXT_TO_U_GET_LENGTH(value),
+            target, targetLimit,
+            offsets, srcIndex,
+            pErrorCode);
+    }
+}
+
+/*
+ * get the SI/SO toU state (state 0 is for SBCS, 1 for DBCS),
+ * or 1 for DBCS-only,
+ * or -1 if the converter is not SI/SO stateful
+ *
+ * Note: For SI/SO stateful converters getting here,
+ * cnv->mode==0 is equivalent to firstLength==1.
+ */
+#define UCNV_SISO_STATE(cnv) \
+    ((cnv)->sharedData->mbcs.outputType==MBCS_OUTPUT_2_SISO ? (int8_t)(cnv)->mode : \
+     (cnv)->sharedData->mbcs.outputType==MBCS_OUTPUT_DBCS_ONLY ? 1 : -1)
+
+/*
+ * target<targetLimit; set error code for overflow
+ */
+U_CFUNC UBool
+ucnv_extInitialMatchToU(UConverter *cnv, const int32_t *cx,
+                        int32_t firstLength,
+                        const char **src, const char *srcLimit,
+                        UChar **target, const UChar *targetLimit,
+                        int32_t **offsets, int32_t srcIndex,
+                        UBool flush,
+                        UErrorCode *pErrorCode) {
+    uint32_t value = 0;  /* initialize output-only param to 0 to silence gcc */
+    int32_t match;
+
+    /* try to match */
+    match=ucnv_extMatchToU(cx, (int8_t)UCNV_SISO_STATE(cnv),
+                           (const char *)cnv->toUBytes, firstLength,
+                           *src, (int32_t)(srcLimit-*src),
+                           &value,
+                           cnv->useFallback, flush);
+    if(match>0) {
+        /* advance src pointer for the consumed input */
+        *src+=match-firstLength;
+
+        /* write result to target */
+        ucnv_extWriteToU(cnv, cx,
+                         value,
+                         target, targetLimit,
+                         offsets, srcIndex,
+                         pErrorCode);
+        return TRUE;
+    } else if(match<0) {
+        /* save state for partial match */
+        const char *s;
+        int32_t j;
+
+        /* copy the first code point */
+        s=(const char *)cnv->toUBytes;
+        cnv->preToUFirstLength=(int8_t)firstLength;
+        for(j=0; j<firstLength; ++j) {
+            cnv->preToU[j]=*s++;
+        }
+
+        /* now copy the newly consumed input */
+        s=*src;
+        match=-match;
+        for(; j<match; ++j) {
+            cnv->preToU[j]=*s++;
+        }
+        *src=s; /* same as *src=srcLimit; because we reached the end of input */
+        cnv->preToULength=(int8_t)match;
+        return TRUE;
+    } else /* match==0 no match */ {
+        return FALSE;
+    }
+}
+
+U_CFUNC UChar32
+ucnv_extSimpleMatchToU(const int32_t *cx,
+                       const char *source, int32_t length,
+                       UBool useFallback) {
+    uint32_t value = 0;  /* initialize output-only param to 0 to silence gcc */
+    int32_t match;
+
+    if(length<=0) {
+        return 0xffff;
+    }
+
+    /* try to match */
+    match=ucnv_extMatchToU(cx, -1,
+                           source, length,
+                           NULL, 0,
+                           &value,
+                           useFallback, TRUE);
+    if(match==length) {
+        /* write result for simple, single-character conversion */
+        if(UCNV_EXT_TO_U_IS_CODE_POINT(value)) {
+            return UCNV_EXT_TO_U_GET_CODE_POINT(value);
+        }
+    }
+
+    /*
+     * return no match because
+     * - match>0 && value points to string: simple conversion cannot handle multiple code points
+     * - match>0 && match!=length: not all input consumed, forbidden for this function
+     * - match==0: no match found in the first place
+     * - match<0: partial match, not supported for simple conversion (and flush==TRUE)
+     */
+    return 0xfffe;
+}
+
+/*
+ * continue partial match with new input
+ * never called for simple, single-character conversion
+ */
+U_CFUNC void
+ucnv_extContinueMatchToU(UConverter *cnv,
+                         UConverterToUnicodeArgs *pArgs, int32_t srcIndex,
+                         UErrorCode *pErrorCode) {
+    uint32_t value = 0;  /* initialize output-only param to 0 to silence gcc */
+    int32_t match, length;
+
+    match=ucnv_extMatchToU(cnv->sharedData->mbcs.extIndexes, (int8_t)UCNV_SISO_STATE(cnv),
+                           cnv->preToU, cnv->preToULength,
+                           pArgs->source, (int32_t)(pArgs->sourceLimit-pArgs->source),
+                           &value,
+                           cnv->useFallback, pArgs->flush);
+    if(match>0) {
+        if(match>=cnv->preToULength) {
+            /* advance src pointer for the consumed input */
+            pArgs->source+=match-cnv->preToULength;
+            cnv->preToULength=0;
+        } else {
+            /* the match did not use all of preToU[] - keep the rest for replay */
+            length=cnv->preToULength-match;
+            uprv_memmove(cnv->preToU, cnv->preToU+match, length);
+            cnv->preToULength=(int8_t)-length;
+        }
+
+        /* write result */
+        ucnv_extWriteToU(cnv, cnv->sharedData->mbcs.extIndexes,
+                         value,
+                         &pArgs->target, pArgs->targetLimit,
+                         &pArgs->offsets, srcIndex,
+                         pErrorCode);
+    } else if(match<0) {
+        /* save state for partial match */
+        const char *s;
+        int32_t j;
+
+        /* just _append_ the newly consumed input to preToU[] */
+        s=pArgs->source;
+        match=-match;
+        for(j=cnv->preToULength; j<match; ++j) {
+            cnv->preToU[j]=*s++;
+        }
+        pArgs->source=s; /* same as *src=srcLimit; because we reached the end of input */
+        cnv->preToULength=(int8_t)match;
+    } else /* match==0 */ {
+        /*
+         * no match
+         *
+         * We need to split the previous input into two parts:
+         *
+         * 1. The first codepage character is unmappable - that's how we got into
+         *    trying the extension data in the first place.
+         *    We need to move it from the preToU buffer
+         *    to the error buffer, set an error code,
+         *    and prepare the rest of the previous input for 2.
+         *
+         * 2. The rest of the previous input must be converted once we
+         *    come back from the callback for the first character.
+         *    At that time, we have to try again from scratch to convert
+         *    these input characters.
+         *    The replay will be handled by the ucnv.c conversion code.
+         */
+
+        /* move the first codepage character to the error field */
+        uprv_memcpy(cnv->toUBytes, cnv->preToU, cnv->preToUFirstLength);
+        cnv->toULength=cnv->preToUFirstLength;
+
+        /* move the rest up inside the buffer */
+        length=cnv->preToULength-cnv->preToUFirstLength;
+        if(length>0) {
+            uprv_memmove(cnv->preToU, cnv->preToU+cnv->preToUFirstLength, length);
+        }
+
+        /* mark preToU for replay */
+        cnv->preToULength=(int8_t)-length;
+
+        /* set the error code for unassigned */
+        *pErrorCode=U_INVALID_CHAR_FOUND;
+    }
+}
+
+/* from Unicode ------------------------------------------------------------- */
+
+// Use roundtrips, "good one-way" mappings, and some normal fallbacks.
+static inline UBool
+extFromUUseMapping(UBool useFallback, uint32_t value, UChar32 firstCP) {
+    return
+        ((value&UCNV_EXT_FROM_U_STATUS_MASK)!=0 ||
+            FROM_U_USE_FALLBACK(useFallback, firstCP)) &&
+        (value&UCNV_EXT_FROM_U_RESERVED_MASK)==0;
+}
+
+/*
+ * @return index of the UChar, if found; else <0
+ */
+static inline int32_t
+ucnv_extFindFromU(const UChar *fromUSection, int32_t length, UChar u) {
+    int32_t i, start, limit;
+
+    /* binary search */
+    start=0;
+    limit=length;
+    for(;;) {
+        i=limit-start;
+        if(i<=1) {
+            break; /* done */
+        }
+        /* start<limit-1 */
+
+        if(i<=4) {
+            /* linear search for the last part */
+            if(u<=fromUSection[start]) {
+                break;
+            }
+            if(++start<limit && u<=fromUSection[start]) {
+                break;
+            }
+            if(++start<limit && u<=fromUSection[start]) {
+                break;
+            }
+            /* always break at start==limit-1 */
+            ++start;
+            break;
+        }
+
+        i=(start+limit)/2;
+        if(u<fromUSection[i]) {
+            limit=i;
+        } else {
+            start=i;
+        }
+    }
+
+    /* did we really find it? */
+    if(start<limit && u==fromUSection[start]) {
+        return start;
+    } else {
+        return -1; /* not found */
+    }
+}
+
+/*
+ * @param cx pointer to extension data; if NULL, returns 0
+ * @param firstCP the first code point before all the other UChars
+ * @param pre UChars that must match; !initialMatch: partial match with them
+ * @param preLength length of pre, >=0
+ * @param src UChars that can be used to complete a match
+ * @param srcLength length of src, >=0
+ * @param pMatchValue [out] output result value for the match from the data structure
+ * @param useFallback "use fallback" flag, usually from cnv->useFallback
+ * @param flush TRUE if the end of the input stream is reached
+ * @return >1: matched, return value=total match length (number of input units matched)
+ *          1: matched, no mapping but request for <subchar1>
+ *             (only for the first code point)
+ *          0: no match
+ *         <0: partial match, return value=negative total match length
+ *             (partial matches are never returned for flush==TRUE)
+ *             (partial matches are never returned as being longer than UCNV_EXT_MAX_UCHARS)
+ *         the matchLength is 2 if only firstCP matched, and >2 if firstCP and
+ *         further code units matched
+ */
+static int32_t
+ucnv_extMatchFromU(const int32_t *cx,
+                   UChar32 firstCP,
+                   const UChar *pre, int32_t preLength,
+                   const UChar *src, int32_t srcLength,
+                   uint32_t *pMatchValue,
+                   UBool useFallback, UBool flush) {
+    const uint16_t *stage12, *stage3;
+    const uint32_t *stage3b;
+
+    const UChar *fromUTableUChars, *fromUSectionUChars;
+    const uint32_t *fromUTableValues, *fromUSectionValues;
+
+    uint32_t value, matchValue;
+    int32_t i, j, idx, length, matchLength;
+    UChar c;
+
+    if(cx==NULL) {
+        return 0; /* no extension data, no match */
+    }
+
+    /* trie lookup of firstCP */
+    idx=firstCP>>10; /* stage 1 index */
+    if(idx>=cx[UCNV_EXT_FROM_U_STAGE_1_LENGTH]) {
+        return 0; /* the first code point is outside the trie */
+    }
+
+    stage12=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_12_INDEX, uint16_t);
+    stage3=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_3_INDEX, uint16_t);
+    idx=UCNV_EXT_FROM_U(stage12, stage3, idx, firstCP);
+
+    stage3b=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_3B_INDEX, uint32_t);
+    value=stage3b[idx];
+    if(value==0) {
+        return 0;
+    }
+
+    /*
+     * Tests for (value&UCNV_EXT_FROM_U_RESERVED_MASK)==0:
+     * Do not interpret values with reserved bits used, for forward compatibility,
+     * and do not even remember intermediate results with reserved bits used.
+     */
+
+    if(UCNV_EXT_TO_U_IS_PARTIAL(value)) {
+        /* partial match, enter the loop below */
+        idx=(int32_t)UCNV_EXT_FROM_U_GET_PARTIAL_INDEX(value);
+
+        /* initialize */
+        fromUTableUChars=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_UCHARS_INDEX, UChar);
+        fromUTableValues=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_VALUES_INDEX, uint32_t);
+
+        matchValue=0;
+        i=j=matchLength=0;
+
+        /* we must not remember fallback matches when not using fallbacks */
+
+        /* match input units until there is a full match or the input is consumed */
+        for(;;) {
+            /* go to the next section */
+            fromUSectionUChars=fromUTableUChars+idx;
+            fromUSectionValues=fromUTableValues+idx;
+
+            /* read first pair of the section */
+            length=*fromUSectionUChars++;
+            value=*fromUSectionValues++;
+            if(value!=0 && extFromUUseMapping(useFallback, value, firstCP)) {
+                /* remember longest match so far */
+                matchValue=value;
+                matchLength=2+i+j;
+            }
+
+            /* match pre[] then src[] */
+            if(i<preLength) {
+                c=pre[i++];
+            } else if(j<srcLength) {
+                c=src[j++];
+            } else {
+                /* all input consumed, partial match */
+                if(flush || (length=(i+j))>UCNV_EXT_MAX_UCHARS) {
+                    /*
+                     * end of the entire input stream, stop with the longest match so far
+                     * or: partial match must not be longer than UCNV_EXT_MAX_UCHARS
+                     * because it must fit into state buffers
+                     */
+                    break;
+                } else {
+                    /* continue with more input next time */
+                    return -(2+length);
+                }
+            }
+
+            /* search for the current UChar */
+            idx=ucnv_extFindFromU(fromUSectionUChars, length, c);
+            if(idx<0) {
+                /* no match here, stop with the longest match so far */
+                break;
+            } else {
+                value=fromUSectionValues[idx];
+                if(UCNV_EXT_FROM_U_IS_PARTIAL(value)) {
+                    /* partial match, continue */
+                    idx=(int32_t)UCNV_EXT_FROM_U_GET_PARTIAL_INDEX(value);
+                } else {
+                    if(extFromUUseMapping(useFallback, value, firstCP)) {
+                        /* full match, stop with result */
+                        matchValue=value;
+                        matchLength=2+i+j;
+                    } else {
+                        /* full match on fallback not taken, stop with the longest match so far */
+                    }
+                    break;
+                }
+            }
+        }
+
+        if(matchLength==0) {
+            /* no match at all */
+            return 0;
+        }
+    } else /* result from firstCP trie lookup */ {
+        if(extFromUUseMapping(useFallback, value, firstCP)) {
+            /* full match, stop with result */
+            matchValue=value;
+            matchLength=2;
+        } else {
+            /* fallback not taken */
+            return 0;
+        }
+    }
+
+    /* return result */
+    if(matchValue==UCNV_EXT_FROM_U_SUBCHAR1) {
+        return 1; /* assert matchLength==2 */
+    }
+
+    *pMatchValue=matchValue;
+    return matchLength;
+}
+
+/*
+ * @param value fromUnicode mapping table value; ignores roundtrip and reserved bits
+ */
+static inline void
+ucnv_extWriteFromU(UConverter *cnv, const int32_t *cx,
+                   uint32_t value,
+                   char **target, const char *targetLimit,
+                   int32_t **offsets, int32_t srcIndex,
+                   UErrorCode *pErrorCode) {
+    uint8_t buffer[1+UCNV_EXT_MAX_BYTES];
+    const uint8_t *result;
+    int32_t length, prevLength;
+
+    length=UCNV_EXT_FROM_U_GET_LENGTH(value);
+    value=(uint32_t)UCNV_EXT_FROM_U_GET_DATA(value);
+
+    /* output the result */
+    if(length<=UCNV_EXT_FROM_U_MAX_DIRECT_LENGTH) {
+        /*
+         * Generate a byte array and then write it below.
+         * This is not the fastest possible way, but it should be ok for
+         * extension mappings, and it is much simpler.
+         * Offset and overflow handling are only done once this way.
+         */
+        uint8_t *p=buffer+1; /* reserve buffer[0] for shiftByte below */
+        switch(length) {
+        case 3:
+            *p++=(uint8_t)(value>>16);
+            U_FALLTHROUGH;
+        case 2:
+            *p++=(uint8_t)(value>>8);
+            U_FALLTHROUGH;
+        case 1:
+            *p++=(uint8_t)value;
+            U_FALLTHROUGH;
+        default:
+            break; /* will never occur */
+        }
+        result=buffer+1;
+    } else {
+        result=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_BYTES_INDEX, uint8_t)+value;
+    }
+
+    /* with correct data we have length>0 */
+
+    if((prevLength=cnv->fromUnicodeStatus)!=0) {
+        /* handle SI/SO stateful output */
+        uint8_t shiftByte;
+
+        if(prevLength>1 && length==1) {
+            /* change from double-byte mode to single-byte */
+            shiftByte=(uint8_t)UCNV_SI;
+            cnv->fromUnicodeStatus=1;
+        } else if(prevLength==1 && length>1) {
+            /* change from single-byte mode to double-byte */
+            shiftByte=(uint8_t)UCNV_SO;
+            cnv->fromUnicodeStatus=2;
+        } else {
+            shiftByte=0;
+        }
+
+        if(shiftByte!=0) {
+            /* prepend the shift byte to the result bytes */
+            buffer[0]=shiftByte;
+            if(result!=buffer+1) {
+                uprv_memcpy(buffer+1, result, length);
+            }
+            result=buffer;
+            ++length;
+        }
+    }
+
+    ucnv_fromUWriteBytes(cnv, (const char *)result, length,
+                         target, targetLimit,
+                         offsets, srcIndex,
+                         pErrorCode);
+}
+
+/*
+ * target<targetLimit; set error code for overflow
+ */
+U_CFUNC UBool
+ucnv_extInitialMatchFromU(UConverter *cnv, const int32_t *cx,
+                          UChar32 cp,
+                          const UChar **src, const UChar *srcLimit,
+                          char **target, const char *targetLimit,
+                          int32_t **offsets, int32_t srcIndex,
+                          UBool flush,
+                          UErrorCode *pErrorCode) {
+    uint32_t value = 0;  /* initialize output-only param to 0 to silence gcc */
+    int32_t match;
+
+    /* try to match */
+    match=ucnv_extMatchFromU(cx, cp,
+                             NULL, 0,
+                             *src, (int32_t)(srcLimit-*src),
+                             &value,
+                             cnv->useFallback, flush);
+
+    /* reject a match if the result is a single byte for DBCS-only */
+    if( match>=2 &&
+        !(UCNV_EXT_FROM_U_GET_LENGTH(value)==1 &&
+          cnv->sharedData->mbcs.outputType==MBCS_OUTPUT_DBCS_ONLY)
+    ) {
+        /* advance src pointer for the consumed input */
+        *src+=match-2; /* remove 2 for the initial code point */
+
+        /* write result to target */
+        ucnv_extWriteFromU(cnv, cx,
+                           value,
+                           target, targetLimit,
+                           offsets, srcIndex,
+                           pErrorCode);
+        return TRUE;
+    } else if(match<0) {
+        /* save state for partial match */
+        const UChar *s;
+        int32_t j;
+
+        /* copy the first code point */
+        cnv->preFromUFirstCP=cp;
+
+        /* now copy the newly consumed input */
+        s=*src;
+        match=-match-2; /* remove 2 for the initial code point */
+        for(j=0; j<match; ++j) {
+            cnv->preFromU[j]=*s++;
+        }
+        *src=s; /* same as *src=srcLimit; because we reached the end of input */
+        cnv->preFromULength=(int8_t)match;
+        return TRUE;
+    } else if(match==1) {
+        /* matched, no mapping but request for <subchar1> */
+        cnv->useSubChar1=TRUE;
+        return FALSE;
+    } else /* match==0 no match */ {
+        return FALSE;
+    }
+}
+
+/*
+ * Used by ISO 2022 implementation.
+ * @return number of bytes in *pValue; negative number if fallback; 0 for no mapping
+ */
+U_CFUNC int32_t
+ucnv_extSimpleMatchFromU(const int32_t *cx,
+                         UChar32 cp, uint32_t *pValue,
+                         UBool useFallback) {
+    uint32_t value;
+    int32_t match;
+
+    /* try to match */
+    match=ucnv_extMatchFromU(cx,
+                             cp,
+                             NULL, 0,
+                             NULL, 0,
+                             &value,
+                             useFallback, TRUE);
+    if(match>=2) {
+        /* write result for simple, single-character conversion */
+        int32_t length;
+        int isRoundtrip;
+
+        isRoundtrip=UCNV_EXT_FROM_U_IS_ROUNDTRIP(value);
+        length=UCNV_EXT_FROM_U_GET_LENGTH(value);
+        value=(uint32_t)UCNV_EXT_FROM_U_GET_DATA(value);
+
+        if(length<=UCNV_EXT_FROM_U_MAX_DIRECT_LENGTH) {
+            *pValue=value;
+            return isRoundtrip ? length : -length;
+#if 0 /* not currently used */
+        } else if(length==4) {
+            /* de-serialize a 4-byte result */
+            const uint8_t *result=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_BYTES_INDEX, uint8_t)+value;
+            *pValue=
+                ((uint32_t)result[0]<<24)|
+                ((uint32_t)result[1]<<16)|
+                ((uint32_t)result[2]<<8)|
+                result[3];
+            return isRoundtrip ? 4 : -4;
+#endif
+        }
+    }
+
+    /*
+     * return no match because
+     * - match>1 && resultLength>4: result too long for simple conversion
+     * - match==1: no match found, <subchar1> preferred
+     * - match==0: no match found in the first place
+     * - match<0: partial match, not supported for simple conversion (and flush==TRUE)
+     */
+    return 0;
+}
+
+/*
+ * continue partial match with new input, requires cnv->preFromUFirstCP>=0
+ * never called for simple, single-character conversion
+ */
+U_CFUNC void
+ucnv_extContinueMatchFromU(UConverter *cnv,
+                           UConverterFromUnicodeArgs *pArgs, int32_t srcIndex,
+                           UErrorCode *pErrorCode) {
+    uint32_t value = 0;  /* initialize output-only param to 0 to silence gcc */
+    int32_t match;
+
+    match=ucnv_extMatchFromU(cnv->sharedData->mbcs.extIndexes,
+                             cnv->preFromUFirstCP,
+                             cnv->preFromU, cnv->preFromULength,
+                             pArgs->source, (int32_t)(pArgs->sourceLimit-pArgs->source),
+                             &value,
+                             cnv->useFallback, pArgs->flush);
+    if(match>=2) {
+        match-=2; /* remove 2 for the initial code point */
+
+        if(match>=cnv->preFromULength) {
+            /* advance src pointer for the consumed input */
+            pArgs->source+=match-cnv->preFromULength;
+            cnv->preFromULength=0;
+        } else {
+            /* the match did not use all of preFromU[] - keep the rest for replay */
+            int32_t length=cnv->preFromULength-match;
+            u_memmove(cnv->preFromU, cnv->preFromU+match, length);
+            cnv->preFromULength=(int8_t)-length;
+        }
+
+        /* finish the partial match */
+        cnv->preFromUFirstCP=U_SENTINEL;
+
+        /* write result */
+        ucnv_extWriteFromU(cnv, cnv->sharedData->mbcs.extIndexes,
+                           value,
+                           &pArgs->target, pArgs->targetLimit,
+                           &pArgs->offsets, srcIndex,
+                           pErrorCode);
+    } else if(match<0) {
+        /* save state for partial match */
+        const UChar *s;
+        int32_t j;
+
+        /* just _append_ the newly consumed input to preFromU[] */
+        s=pArgs->source;
+        match=-match-2; /* remove 2 for the initial code point */
+        for(j=cnv->preFromULength; j<match; ++j) {
+            U_ASSERT(j>=0);
+            cnv->preFromU[j]=*s++;
+        }
+        pArgs->source=s; /* same as *src=srcLimit; because we reached the end of input */
+        cnv->preFromULength=(int8_t)match;
+    } else /* match==0 or 1 */ {
+        /*
+         * no match
+         *
+         * We need to split the previous input into two parts:
+         *
+         * 1. The first code point is unmappable - that's how we got into
+         *    trying the extension data in the first place.
+         *    We need to move it from the preFromU buffer
+         *    to the error buffer, set an error code,
+         *    and prepare the rest of the previous input for 2.
+         *
+         * 2. The rest of the previous input must be converted once we
+         *    come back from the callback for the first code point.
+         *    At that time, we have to try again from scratch to convert
+         *    these input characters.
+         *    The replay will be handled by the ucnv.c conversion code.
+         */
+
+        if(match==1) {
+            /* matched, no mapping but request for <subchar1> */
+            cnv->useSubChar1=TRUE;
+        }
+
+        /* move the first code point to the error field */
+        cnv->fromUChar32=cnv->preFromUFirstCP;
+        cnv->preFromUFirstCP=U_SENTINEL;
+
+        /* mark preFromU for replay */
+        cnv->preFromULength=-cnv->preFromULength;
+
+        /* set the error code for unassigned */
+        *pErrorCode=U_INVALID_CHAR_FOUND;
+    }
+}
+
+static UBool
+extSetUseMapping(UConverterUnicodeSet which, int32_t minLength, uint32_t value) {
+    if(which==UCNV_ROUNDTRIP_SET) {
+        // Add only code points for which the roundtrip flag is set.
+        // Do not add any fallbacks, even if ucnv_fromUnicode() would use them
+        // (fallbacks from PUA). See the API docs for ucnv_getUnicodeSet().
+        //
+        // By analogy, also do not add "good one-way" mappings.
+        //
+        // Do not add entries with reserved bits set.
+        if(((value&(UCNV_EXT_FROM_U_ROUNDTRIP_FLAG|UCNV_EXT_FROM_U_RESERVED_MASK))!=
+                UCNV_EXT_FROM_U_ROUNDTRIP_FLAG)) {
+            return FALSE;
+        }
+    } else /* UCNV_ROUNDTRIP_AND_FALLBACK_SET */ {
+        // Do not add entries with reserved bits set.
+        if((value&UCNV_EXT_FROM_U_RESERVED_MASK)!=0) {
+            return FALSE;
+        }
+    }
+    // Do not add <subchar1> entries or other (future?) pseudo-entries
+    // with an output length of 0.
+    return UCNV_EXT_FROM_U_GET_LENGTH(value)>=minLength;
+}
+
+static void
+ucnv_extGetUnicodeSetString(const UConverterSharedData *sharedData,
+                            const int32_t *cx,
+                            const USetAdder *sa,
+                            UConverterUnicodeSet which,
+                            int32_t minLength,
+                            UChar32 firstCP,
+                            UChar s[UCNV_EXT_MAX_UCHARS], int32_t length,
+                            int32_t sectionIndex,
+                            UErrorCode *pErrorCode) {
+    const UChar *fromUSectionUChars;
+    const uint32_t *fromUSectionValues;
+
+    uint32_t value;
+    int32_t i, count;
+
+    fromUSectionUChars=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_UCHARS_INDEX, UChar)+sectionIndex;
+    fromUSectionValues=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_VALUES_INDEX, uint32_t)+sectionIndex;
+
+    /* read first pair of the section */
+    count=*fromUSectionUChars++;
+    value=*fromUSectionValues++;
+
+    if(extSetUseMapping(which, minLength, value)) {
+        if(length==U16_LENGTH(firstCP)) {
+            /* add the initial code point */
+            sa->add(sa->set, firstCP);
+        } else {
+            /* add the string so far */
+            sa->addString(sa->set, s, length);
+        }
+    }
+
+    for(i=0; i<count; ++i) {
+        /* append this code unit and recurse or add the string */
+        s[length]=fromUSectionUChars[i];
+        value=fromUSectionValues[i];
+
+        if(value==0) {
+            /* no mapping, do nothing */
+        } else if(UCNV_EXT_FROM_U_IS_PARTIAL(value)) {
+            ucnv_extGetUnicodeSetString(
+                sharedData, cx, sa, which, minLength,
+                firstCP, s, length+1,
+                (int32_t)UCNV_EXT_FROM_U_GET_PARTIAL_INDEX(value),
+                pErrorCode);
+        } else if(extSetUseMapping(which, minLength, value)) {
+            sa->addString(sa->set, s, length+1);
+        }
+    }
+}
+
+U_CFUNC void
+ucnv_extGetUnicodeSet(const UConverterSharedData *sharedData,
+                      const USetAdder *sa,
+                      UConverterUnicodeSet which,
+                      UConverterSetFilter filter,
+                      UErrorCode *pErrorCode) {
+    const int32_t *cx;
+    const uint16_t *stage12, *stage3, *ps2, *ps3;
+    const uint32_t *stage3b;
+
+    uint32_t value;
+    int32_t st1, stage1Length, st2, st3, minLength;
+
+    UChar s[UCNV_EXT_MAX_UCHARS];
+    UChar32 c;
+    int32_t length;
+
+    cx=sharedData->mbcs.extIndexes;
+    if(cx==NULL) {
+        return;
+    }
+
+    stage12=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_12_INDEX, uint16_t);
+    stage3=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_3_INDEX, uint16_t);
+    stage3b=UCNV_EXT_ARRAY(cx, UCNV_EXT_FROM_U_STAGE_3B_INDEX, uint32_t);
+
+    stage1Length=cx[UCNV_EXT_FROM_U_STAGE_1_LENGTH];
+
+    /* enumerate the from-Unicode trie table */
+    c=0; /* keep track of the current code point while enumerating */
+
+    if(filter==UCNV_SET_FILTER_2022_CN) {
+        minLength=3;
+    } else if( sharedData->mbcs.outputType==MBCS_OUTPUT_DBCS_ONLY ||
+               filter!=UCNV_SET_FILTER_NONE
+    ) {
+        /* DBCS-only, ignore single-byte results */
+        minLength=2;
+    } else {
+        minLength=1;
+    }
+
+    /*
+     * the trie enumeration is almost the same as
+     * in MBCSGetUnicodeSet() for MBCS_OUTPUT_1
+     */
+    for(st1=0; st1<stage1Length; ++st1) {
+        st2=stage12[st1];
+        if(st2>stage1Length) {
+            ps2=stage12+st2;
+            for(st2=0; st2<64; ++st2) {
+                if((st3=(int32_t)ps2[st2]<<UCNV_EXT_STAGE_2_LEFT_SHIFT)!=0) {
+                    /* read the stage 3 block */
+                    ps3=stage3+st3;
+
+                    do {
+                        value=stage3b[*ps3++];
+                        if(value==0) {
+                            /* no mapping, do nothing */
+                        } else if(UCNV_EXT_FROM_U_IS_PARTIAL(value)) {
+                            // Recurse for partial results.
+                            length=0;
+                            U16_APPEND_UNSAFE(s, length, c);
+                            ucnv_extGetUnicodeSetString(
+                                sharedData, cx, sa, which, minLength,
+                                c, s, length,
+                                (int32_t)UCNV_EXT_FROM_U_GET_PARTIAL_INDEX(value),
+                                pErrorCode);
+                        } else if(extSetUseMapping(which, minLength, value)) {
+                            switch(filter) {
+                            case UCNV_SET_FILTER_2022_CN:
+                                if(!(UCNV_EXT_FROM_U_GET_LENGTH(value)==3 && UCNV_EXT_FROM_U_GET_DATA(value)<=0x82ffff)) {
+                                    continue;
+                                }
+                                break;
+                            case UCNV_SET_FILTER_SJIS:
+                                if(!(UCNV_EXT_FROM_U_GET_LENGTH(value)==2 && (value=UCNV_EXT_FROM_U_GET_DATA(value))>=0x8140 && value<=0xeffc)) {
+                                    continue;
+                                }
+                                break;
+                            case UCNV_SET_FILTER_GR94DBCS:
+                                if(!(UCNV_EXT_FROM_U_GET_LENGTH(value)==2 &&
+                                     (uint16_t)((value=UCNV_EXT_FROM_U_GET_DATA(value))-0xa1a1)<=(0xfefe - 0xa1a1) &&
+                                     (uint8_t)(value-0xa1)<=(0xfe - 0xa1))) {
+                                    continue;
+                                }
+                                break;
+                            case UCNV_SET_FILTER_HZ:
+                                if(!(UCNV_EXT_FROM_U_GET_LENGTH(value)==2 &&
+                                     (uint16_t)((value=UCNV_EXT_FROM_U_GET_DATA(value))-0xa1a1)<=(0xfdfe - 0xa1a1) &&
+                                     (uint8_t)(value-0xa1)<=(0xfe - 0xa1))) {
+                                    continue;
+                                }
+                                break;
+                            default:
+                                /*
+                                 * UCNV_SET_FILTER_NONE,
+                                 * or UCNV_SET_FILTER_DBCS_ONLY which is handled via minLength
+                                 */
+                                break;
+                            }
+                            sa->add(sa->set, c);
+                        }
+                    } while((++c&0xf)!=0);
+                } else {
+                    c+=16; /* empty stage 3 block */
+                }
+            }
+        } else {
+            c+=1024; /* empty stage 2 block */
+        }
+    }
+}
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnv_ext.h b/thirdparty/icu4c/common/ucnv_ext.h
new file mode 100644
index 0000000000..dceea7ef12
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_ext.h
@@ -0,0 +1,481 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2003-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnv_ext.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003jun13
+*   created by: Markus W. Scherer
+*
+*   Conversion extensions
+*/
+
+#ifndef __UCNV_EXT_H__
+#define __UCNV_EXT_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "ucnv_cnv.h"
+
+/*
+ * See icuhtml/design/conversion/conversion_extensions.html
+ *
+ * Conversion extensions serve three purposes:
+ * 1. They support m:n mappings.
+ * 2. They support extension-only conversion files that are used together
+ *    with the regular conversion data in base files.
+ * 3. They support mappings with more complicated meta data,
+ *    for example "good one-way" mappings (|4).
+ *
+ * A base file may contain an extension table (explicitly requested or
+ * implicitly generated for m:n mappings), but its extension table is not
+ * used when an extension-only file is used.
+ *
+ * It is an error if a base file contains any regular (not extension) mapping
+ * from the same sequence as a mapping in the extension file
+ * because the base mapping would hide the extension mapping.
+ *
+ *
+ * Data for conversion extensions:
+ *
+ * One set of data structures per conversion direction (to/from Unicode).
+ * The data structures are sorted by input units to allow for binary search.
+ * Input sequences of more than one unit are handled like contraction tables
+ * in collation:
+ * The lookup value of a unit points to another table that is to be searched
+ * for the next unit, recursively.
+ *
+ * For conversion from Unicode, the initial code point is looked up in
+ * a 3-stage trie for speed,
+ * with an additional table of unique results to save space.
+ *
+ * Long output strings are stored in separate arrays, with length and index
+ * in the lookup tables.
+ * Output results also include a flag distinguishing roundtrip from
+ * (reverse) fallback mappings.
+ *
+ * Input Unicode strings must not begin or end with unpaired surrogates
+ * to avoid problems with matches on parts of surrogate pairs.
+ *
+ * Mappings from multiple characters (code points or codepage state
+ * table sequences) must be searched preferring the longest match.
+ * For this to work and be efficient, the variable-width table must contain
+ * all mappings that contain prefixes of the multiple characters.
+ * If an extension table is built on top of a base table in another file
+ * and a base table entry is a prefix of a multi-character mapping, then
+ * this is an error.
+ *
+ *
+ * Implementation note:
+ *
+ * Currently, the parser and several checks in the code limit the number
+ * of UChars or bytes in a mapping to
+ * UCNV_EXT_MAX_UCHARS and UCNV_EXT_MAX_BYTES, respectively,
+ * which are output value limits in the data structure.
+ *
+ * For input, this is not strictly necessary - it is a hard limit only for the
+ * buffers in UConverter that are used to store partial matches.
+ *
+ * Input sequences could otherwise be arbitrarily long if partial matches
+ * need not be stored (i.e., if a sequence does not span several buffers with too
+ * many units before the last buffer), although then results would differ
+ * depending on whether partial matches exceed the limits or not,
+ * which depends on the pattern of buffer sizes.
+ *
+ *
+ * Data structure:
+ *
+ * int32_t indexes[>=32];
+ *
+ *   Array of indexes and lengths etc. The length of the array is at least 32.
+ *   The actual length is stored in indexes[0] to be forward compatible.
+ *
+ *   Each index to another array is the number of bytes from indexes[].
+ *   Each length of an array is the number of array base units in that array.
+ *
+ *   Some of the structures may not be present, in which case their indexes
+ *   and lengths are 0.
+ *
+ *   Usage of indexes[i]:
+ *   [0]  length of indexes[]
+ *
+ *   // to Unicode table
+ *   [1]  index of toUTable[] (array of uint32_t)
+ *   [2]  length of toUTable[]
+ *   [3]  index of toUUChars[] (array of UChar)
+ *   [4]  length of toUUChars[]
+ *
+ *   // from Unicode table, not for the initial code point
+ *   [5]  index of fromUTableUChars[] (array of UChar)
+ *   [6]  index of fromUTableValues[] (array of uint32_t)
+ *   [7]  length of fromUTableUChars[] and fromUTableValues[]
+ *   [8]  index of fromUBytes[] (array of char)
+ *   [9]  length of fromUBytes[]
+ *
+ *   // from Unicode trie for initial-code point lookup
+ *   [10] index of fromUStage12[] (combined array of uint16_t for stages 1 & 2)
+ *   [11] length of stage 1 portion of fromUStage12[]
+ *   [12] length of fromUStage12[]
+ *   [13] index of fromUStage3[] (array of uint16_t indexes into fromUStage3b[])
+ *   [14] length of fromUStage3[]
+ *   [15] index of fromUStage3b[] (array of uint32_t like fromUTableValues[])
+ *   [16] length of fromUStage3b[]
+ *
+ *   [17] Bit field containing numbers of bytes:
+ *        31..24 reserved, 0
+ *        23..16 maximum input bytes
+ *        15.. 8 maximum output bytes
+ *         7.. 0 maximum bytes per UChar
+ *
+ *   [18] Bit field containing numbers of UChars:
+ *        31..24 reserved, 0
+ *        23..16 maximum input UChars
+ *        15.. 8 maximum output UChars
+ *         7.. 0 maximum UChars per byte
+ *
+ *   [19] Bit field containing flags:
+ *               (extension table unicodeMask)
+ *         1     UCNV_HAS_SURROGATES flag for the extension table
+ *         0     UCNV_HAS_SUPPLEMENTARY flag for the extension table
+ *
+ *   [20]..[30] reserved, 0
+ *   [31] number of bytes for the entire extension structure
+ *   [>31] reserved; there are indexes[0] indexes
+ *
+ *
+ * uint32_t toUTable[];
+ *
+ *   Array of byte/value pairs for lookups for toUnicode conversion.
+ *   The array is partitioned into sections like collation contraction tables.
+ *   Each section contains one word with the number of following words and
+ *   a default value for when the lookup in this section yields no match.
+ *
+ *   A section is sorted in ascending order of input bytes,
+ *   allowing for fast linear or binary searches.
+ *   The builder may store entries for a contiguous range of byte values
+ *   (compare difference between the first and last one with count),
+ *   which then allows for direct array access.
+ *   The builder should always do this for the initial table section.
+ *
+ *   Entries may have 0 values, see below.
+ *   No two entries in a section have the same byte values.
+ *
+ *   Each uint32_t contains an input byte value in bits 31..24 and the
+ *   corresponding lookup value in bits 23..0.
+ *   Interpret the value as follows:
+ *     if(value==0) {
+ *       no match, see below
+ *     } else if(value<0x1f0000) {
+ *       partial match - use value as index to the next toUTable section
+ *       and match the next unit; (value indexes toUTable[value])
+ *     } else {
+ *       if(bit 23 set) {
+ *         roundtrip;
+ *       } else {
+ *         fallback;
+ *       }
+ *       unset value bit 23;
+ *       if(value<=0x2fffff) {
+ *         (value-0x1f0000) is a code point; (BMP: value<=0x1fffff)
+ *       } else {
+ *         bits 17..0 (value&0x3ffff) is an index to
+ *           the result UChars in toUUChars[]; (0 indexes toUUChars[0])
+ *         length of the result=((value>>18)-12); (length=0..19)
+ *       }
+ *     }
+ *
+ *   The first word in a section contains the number of following words in the
+ *   input byte position (bits 31..24, number=1..0xff).
+ *   The value of the initial word is used when the current byte is not found
+ *   in this section.
+ *   If the value is not 0, then it represents a result as above.
+ *   If the value is 0, then the search has to return a shorter match with an
+ *   earlier default value as the result, or result in "unmappable" even for the
+ *   initial bytes.
+ *   If the value is 0 for the initial toUTable entry, then the initial byte
+ *   does not start any mapping input.
+ *
+ *
+ * UChar toUUChars[];
+ *
+ *   Contains toUnicode mapping results, stored as sequences of UChars.
+ *   Indexes and lengths stored in the toUTable[].
+ *
+ *
+ * UChar fromUTableUChars[];
+ * uint32_t fromUTableValues[];
+ *
+ *   The fromUTable is split into two arrays, but works otherwise much like
+ *   the toUTable. The array is partitioned into sections like collation
+ *   contraction tables and toUTable.
+ *   A row in the table consists of same-index entries in fromUTableUChars[]
+ *   and fromUTableValues[].
+ *
+ *   Interpret a value as follows:
+ *     if(value==0) {
+ *       no match, see below
+ *     } else if(value<=0xffffff) { (bits 31..24 are 0)
+ *       partial match - use value as index to the next fromUTable section
+ *       and match the next unit; (value indexes fromUTable[value])
+ *     } else {
+ *       if(value==0x80000001) {
+ *         return no mapping, but request for <subchar1>;
+ *       }
+ *       if(bit 31 set) {
+ *         roundtrip (|0);
+ *       } else if(bit 30 set) {
+ *         "good one-way" mapping (|4); -- new in ICU4C 51, _MBCSHeader.version 5.4/4.4
+ *       } else {
+ *         normal fallback (|1);
+ *       }
+ *       // bit 29 reserved, 0
+ *       length=(value>>24)&0x1f; (bits 28..24)
+ *       if(length==1..3) {
+ *         bits 23..0 contain 1..3 bytes, padded with 00s on the left;
+ *       } else {
+ *         bits 23..0 (value&0xffffff) is an index to
+ *           the result bytes in fromUBytes[]; (0 indexes fromUBytes[0])
+ *       }
+ *     }
+ *       
+ *   The first pair in a section contains the number of following pairs in the
+ *   UChar position (16 bits, number=1..0xffff).
+ *   The value of the initial pair is used when the current UChar is not found
+ *   in this section.
+ *   If the value is not 0, then it represents a result as above.
+ *   If the value is 0, then the search has to return a shorter match with an
+ *   earlier default value as the result, or result in "unmappable" even for the
+ *   initial UChars.
+ *
+ *   If the from Unicode trie is present, then the from Unicode search tables
+ *   are not used for initial code points.
+ *   In this case, the first entries (index 0) in the tables are not used
+ *   (reserved, set to 0) because a value of 0 is used in trie results
+ *   to indicate no mapping.
+ *
+ *
+ * uint16_t fromUStage12[];
+ *
+ *   Stages 1 & 2 of a trie that maps an initial code point.
+ *   Indexes in stage 1 are all offset by the length of stage 1 so that the
+ *   same array pointer can be used for both stages.
+ *   If (c>>10)>=(length of stage 1) then c does not start any mapping.
+ *   Same bit distribution as for regular conversion tries.
+ *
+ *
+ * uint16_t fromUStage3[];
+ * uint32_t fromUStage3b[];
+ *
+ *   Stage 3 of the trie. The first array simply contains indexes to the second,
+ *   which contains words in the same format as fromUTableValues[].
+ *   Use a stage 3 granularity of 4, which allows for 256k stage 3 entries,
+ *   and 16-bit entries in stage 3 allow for 64k stage 3b entries.
+ *   The stage 3 granularity means that the stage 2 entry needs to be left-shifted.
+ *
+ *   Two arrays are used because it is expected that more than half of the stage 3
+ *   entries will be zero. The 16-bit index stage 3 array saves space even
+ *   considering storing a total of 6 bytes per non-zero entry in both arrays
+ *   together.
+ *   Using a stage 3 granularity of >1 diminishes the compactability in that stage
+ *   but provides a larger effective addressing space in stage 2.
+ *   All but the final result stage use 16-bit entries to save space.
+ *
+ *   fromUStage3b[] contains a zero for "no mapping" at its index 0,
+ *   and may contain UCNV_EXT_FROM_U_SUBCHAR1 at index 1 for "<subchar1> SUB mapping"
+ *   (i.e., "no mapping" with preference for <subchar1> rather than <subchar>),
+ *   and all other items are unique non-zero results.
+ *
+ *   The default value of a fromUTableValues[] section that is referenced
+ *   _directly_ from a fromUStage3b[] item may also be UCNV_EXT_FROM_U_SUBCHAR1,
+ *   but this value must not occur anywhere else in fromUTableValues[]
+ *   because "no mapping" is always a property of a single code point,
+ *   never of multiple.
+ *
+ *
+ * char fromUBytes[];
+ *
+ *   Contains fromUnicode mapping results, stored as sequences of chars.
+ *   Indexes and lengths stored in the fromUTableValues[].
+ */
+enum {
+    UCNV_EXT_INDEXES_LENGTH,            /* 0 */
+
+    UCNV_EXT_TO_U_INDEX,                /* 1 */
+    UCNV_EXT_TO_U_LENGTH,
+    UCNV_EXT_TO_U_UCHARS_INDEX,
+    UCNV_EXT_TO_U_UCHARS_LENGTH,
+
+    UCNV_EXT_FROM_U_UCHARS_INDEX,       /* 5 */
+    UCNV_EXT_FROM_U_VALUES_INDEX,
+    UCNV_EXT_FROM_U_LENGTH,
+    UCNV_EXT_FROM_U_BYTES_INDEX,
+    UCNV_EXT_FROM_U_BYTES_LENGTH,
+
+    UCNV_EXT_FROM_U_STAGE_12_INDEX,     /* 10 */
+    UCNV_EXT_FROM_U_STAGE_1_LENGTH,
+    UCNV_EXT_FROM_U_STAGE_12_LENGTH,
+    UCNV_EXT_FROM_U_STAGE_3_INDEX,
+    UCNV_EXT_FROM_U_STAGE_3_LENGTH,
+    UCNV_EXT_FROM_U_STAGE_3B_INDEX,
+    UCNV_EXT_FROM_U_STAGE_3B_LENGTH,
+
+    UCNV_EXT_COUNT_BYTES,               /* 17 */
+    UCNV_EXT_COUNT_UCHARS,
+    UCNV_EXT_FLAGS,
+
+    UCNV_EXT_RESERVED_INDEX,            /* 20, moves with additional indexes */
+
+    UCNV_EXT_SIZE=31,
+    UCNV_EXT_INDEXES_MIN_LENGTH=32
+};
+
+/* get the pointer to an extension array from indexes[index] */
+#define UCNV_EXT_ARRAY(indexes, index, itemType) \
+    ((const itemType *)((const char *)(indexes)+(indexes)[index]))
+
+#define UCNV_GET_MAX_BYTES_PER_UCHAR(indexes) \
+    ((indexes)[UCNV_EXT_COUNT_BYTES]&0xff)
+
+/* internal API ------------------------------------------------------------- */
+
+U_CFUNC UBool
+ucnv_extInitialMatchToU(UConverter *cnv, const int32_t *cx,
+                        int32_t firstLength,
+                        const char **src, const char *srcLimit,
+                        UChar **target, const UChar *targetLimit,
+                        int32_t **offsets, int32_t srcIndex,
+                        UBool flush,
+                        UErrorCode *pErrorCode);
+
+U_CFUNC UChar32
+ucnv_extSimpleMatchToU(const int32_t *cx,
+                       const char *source, int32_t length,
+                       UBool useFallback);
+
+U_CFUNC void
+ucnv_extContinueMatchToU(UConverter *cnv,
+                         UConverterToUnicodeArgs *pArgs, int32_t srcIndex,
+                         UErrorCode *pErrorCode);
+
+
+U_CFUNC UBool
+ucnv_extInitialMatchFromU(UConverter *cnv, const int32_t *cx,
+                          UChar32 cp,
+                          const UChar **src, const UChar *srcLimit,
+                          char **target, const char *targetLimit,
+                          int32_t **offsets, int32_t srcIndex,
+                          UBool flush,
+                          UErrorCode *pErrorCode);
+
+U_CFUNC int32_t
+ucnv_extSimpleMatchFromU(const int32_t *cx,
+                         UChar32 cp, uint32_t *pValue,
+                         UBool useFallback);
+
+U_CFUNC void
+ucnv_extContinueMatchFromU(UConverter *cnv,
+                           UConverterFromUnicodeArgs *pArgs, int32_t srcIndex,
+                           UErrorCode *pErrorCode);
+
+/*
+ * Add code points and strings to the set according to the extension mappings.
+ * Limitation on the UConverterSetFilter:
+ * The filters currently assume that they are used with 1:1 mappings.
+ * They only apply to single input code points, and then they pass through
+ * only mappings with single-charset-code results.
+ * For example, the Shift-JIS filter only works for 2-byte results and tests
+ * that those 2 bytes are in the JIS X 0208 range of Shift-JIS.
+ */
+U_CFUNC void
+ucnv_extGetUnicodeSet(const UConverterSharedData *sharedData,
+                      const USetAdder *sa,
+                      UConverterUnicodeSet which,
+                      UConverterSetFilter filter,
+                      UErrorCode *pErrorCode);
+
+/* toUnicode helpers -------------------------------------------------------- */
+
+#define UCNV_EXT_TO_U_BYTE_SHIFT 24
+#define UCNV_EXT_TO_U_VALUE_MASK 0xffffff
+#define UCNV_EXT_TO_U_MIN_CODE_POINT 0x1f0000
+#define UCNV_EXT_TO_U_MAX_CODE_POINT 0x2fffff
+#define UCNV_EXT_TO_U_ROUNDTRIP_FLAG ((uint32_t)1<<23)
+#define UCNV_EXT_TO_U_INDEX_MASK 0x3ffff
+#define UCNV_EXT_TO_U_LENGTH_SHIFT 18
+#define UCNV_EXT_TO_U_LENGTH_OFFSET 12
+
+/* maximum number of indexed UChars */
+#define UCNV_EXT_MAX_UCHARS 19
+
+#define UCNV_EXT_TO_U_MAKE_WORD(byte, value) (((uint32_t)(byte)<<UCNV_EXT_TO_U_BYTE_SHIFT)|(value))
+
+#define UCNV_EXT_TO_U_GET_BYTE(word) ((word)>>UCNV_EXT_TO_U_BYTE_SHIFT)
+#define UCNV_EXT_TO_U_GET_VALUE(word) ((word)&UCNV_EXT_TO_U_VALUE_MASK)
+
+#define UCNV_EXT_TO_U_IS_PARTIAL(value) ((value)<UCNV_EXT_TO_U_MIN_CODE_POINT)
+#define UCNV_EXT_TO_U_GET_PARTIAL_INDEX(value) (value)
+
+#define UCNV_EXT_TO_U_IS_ROUNDTRIP(value) (((value)&UCNV_EXT_TO_U_ROUNDTRIP_FLAG)!=0)
+#define UCNV_EXT_TO_U_MASK_ROUNDTRIP(value) ((value)&~UCNV_EXT_TO_U_ROUNDTRIP_FLAG)
+
+/* use after masking off the roundtrip flag */
+#define UCNV_EXT_TO_U_IS_CODE_POINT(value) ((value)<=UCNV_EXT_TO_U_MAX_CODE_POINT)
+#define UCNV_EXT_TO_U_GET_CODE_POINT(value) ((value)-UCNV_EXT_TO_U_MIN_CODE_POINT)
+
+#define UCNV_EXT_TO_U_GET_INDEX(value) ((value)&UCNV_EXT_TO_U_INDEX_MASK)
+#define UCNV_EXT_TO_U_GET_LENGTH(value) (((value)>>UCNV_EXT_TO_U_LENGTH_SHIFT)-UCNV_EXT_TO_U_LENGTH_OFFSET)
+
+/* fromUnicode helpers ------------------------------------------------------ */
+
+/* most trie constants are shared with ucnvmbcs.h */
+
+/* see similar utrie.h UTRIE_INDEX_SHIFT and UTRIE_DATA_GRANULARITY */
+#define UCNV_EXT_STAGE_2_LEFT_SHIFT 2
+#define UCNV_EXT_STAGE_3_GRANULARITY 4
+
+/* trie access, returns the stage 3 value=index to stage 3b; s1Index=c>>10 */
+#define UCNV_EXT_FROM_U(stage12, stage3, s1Index, c) \
+    (stage3)[ ((int32_t)(stage12)[ (stage12)[s1Index] +(((c)>>4)&0x3f) ]<<UCNV_EXT_STAGE_2_LEFT_SHIFT) +((c)&0xf) ]
+
+#define UCNV_EXT_FROM_U_LENGTH_SHIFT 24
+#define UCNV_EXT_FROM_U_ROUNDTRIP_FLAG ((uint32_t)1<<31)
+#define UCNV_EXT_FROM_U_GOOD_ONE_WAY_FLAG 0x40000000
+#define UCNV_EXT_FROM_U_STATUS_MASK 0xc0000000
+#define UCNV_EXT_FROM_U_RESERVED_MASK 0x20000000
+#define UCNV_EXT_FROM_U_DATA_MASK 0xffffff
+
+/* special value for "no mapping" to <subchar1> (impossible roundtrip to 0 bytes, value 01) */
+#define UCNV_EXT_FROM_U_SUBCHAR1 0x80000001
+
+/* at most 3 bytes in the lower part of the value */
+#define UCNV_EXT_FROM_U_MAX_DIRECT_LENGTH 3
+
+/* maximum number of indexed bytes */
+#define UCNV_EXT_MAX_BYTES 0x1f
+
+#define UCNV_EXT_FROM_U_IS_PARTIAL(value) (((value)>>UCNV_EXT_FROM_U_LENGTH_SHIFT)==0)
+#define UCNV_EXT_FROM_U_GET_PARTIAL_INDEX(value) (value)
+
+#define UCNV_EXT_FROM_U_IS_ROUNDTRIP(value) (((value)&UCNV_EXT_FROM_U_ROUNDTRIP_FLAG)!=0)
+#define UCNV_EXT_FROM_U_MASK_ROUNDTRIP(value) ((value)&~UCNV_EXT_FROM_U_ROUNDTRIP_FLAG)
+
+/* get length; masks away all other bits */
+#define UCNV_EXT_FROM_U_GET_LENGTH(value) (int32_t)(((value)>>UCNV_EXT_FROM_U_LENGTH_SHIFT)&UCNV_EXT_MAX_BYTES)
+
+/* get bytes or bytes index */
+#define UCNV_EXT_FROM_U_GET_DATA(value) ((value)&UCNV_EXT_FROM_U_DATA_MASK)
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_imp.h b/thirdparty/icu4c/common/ucnv_imp.h
new file mode 100644
index 0000000000..c5e6aeb47e
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_imp.h
@@ -0,0 +1,139 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+*
+*  ucnv_imp.h:
+*  Contains all internal and external data structure definitions
+* Created & Maitained by Bertrand A. Damiba
+*
+*
+*
+* ATTENTION:
+* ---------
+* Although the data structures in this file are open and stack allocatable
+* we reserve the right to hide them in further releases.
+*/
+
+#ifndef UCNV_IMP_H
+#define UCNV_IMP_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/uloc.h"
+#include "ucnv_bld.h"
+
+/*
+ * Fast check for whether a charset name is "UTF-8".
+ * This does not recognize all of the variations that ucnv_open()
+ * and other functions recognize, but it covers most cases.
+ * @param name const char * charset name
+ * @return
+ */
+#define UCNV_FAST_IS_UTF8(name) \
+    (((name[0]=='U' ? \
+      (                name[1]=='T' && name[2]=='F') : \
+      (name[0]=='u' && name[1]=='t' && name[2]=='f'))) \
+  && (name[3]=='-' ? \
+     (name[4]=='8' && name[5]==0) : \
+     (name[3]=='8' && name[4]==0)))
+
+typedef struct {
+    char cnvName[UCNV_MAX_CONVERTER_NAME_LENGTH];
+    char locale[ULOC_FULLNAME_CAPACITY];
+    uint32_t options;
+} UConverterNamePieces;
+
+U_CFUNC UBool
+ucnv_canCreateConverter(const char *converterName, UErrorCode *err);
+
+/* figures out if we need to go to file to read in the data tables.
+ * @param converterName The name of the converter
+ * @param err The error code
+ * @return the newly created converter
+ */
+U_CAPI UConverter *
+ucnv_createConverter(UConverter *myUConverter, const char *converterName, UErrorCode * err);
+
+/*
+ * Open a purely algorithmic converter, specified by a type constant.
+ * @param myUConverter  NULL, or pre-allocated UConverter structure to avoid
+ *                      a memory allocation
+ * @param type          requested converter type
+ * @param locale        locale parameter, or ""
+ * @param options       converter options bit set (default 0)
+ * @param err           ICU error code, not tested for U_FAILURE on input
+ *                      because this is an internal function
+ * @internal
+ */
+U_CFUNC UConverter *
+ucnv_createAlgorithmicConverter(UConverter *myUConverter,
+                                UConverterType type,
+                                const char *locale, uint32_t options,
+                                UErrorCode *err);
+
+/*
+ * Creates a converter from shared data.
+ * Adopts mySharedConverterData: No matter what happens, the caller must not
+ * unload mySharedConverterData, except via ucnv_close(return value)
+ * if this function is successful.
+ */
+U_CFUNC UConverter *
+ucnv_createConverterFromSharedData(UConverter *myUConverter,
+                                   UConverterSharedData *mySharedConverterData,
+                                   UConverterLoadArgs *pArgs,
+                                   UErrorCode *err);
+
+U_CFUNC UConverter *
+ucnv_createConverterFromPackage(const char *packageName, const char *converterName, UErrorCode *err);
+
+/**
+ * Load a converter but do not create a UConverter object.
+ * Simply return the UConverterSharedData.
+ * Performs alias lookup etc.
+ * The UConverterNamePieces need not be initialized
+ * before calling this function.
+ * The UConverterLoadArgs must be initialized
+ * before calling this function.
+ * If the args are passed in, then the pieces must be passed in too.
+ * In other words, the following combinations are allowed:
+ * - pieces==NULL && args==NULL
+ * - pieces!=NULL && args==NULL
+ * - pieces!=NULL && args!=NULL
+ * @internal
+ */
+U_CFUNC UConverterSharedData *
+ucnv_loadSharedData(const char *converterName,
+                    UConverterNamePieces *pieces,
+                    UConverterLoadArgs *pArgs,
+                    UErrorCode * err);
+
+/**
+ * This may unload the shared data in a thread safe manner.
+ * This will only unload the data if no other converters are sharing it.
+ */
+U_CFUNC void
+ucnv_unloadSharedDataIfReady(UConverterSharedData *sharedData);
+
+/**
+ * This is a thread safe way to increment the reference count.
+ */
+U_CFUNC void
+ucnv_incrementRefCount(UConverterSharedData *sharedData);
+
+/**
+ * These are the default error handling callbacks for the charset conversion framework.
+ * For performance reasons, they are only called to handle an error (not normally called for a reset or close).
+ */
+#define UCNV_TO_U_DEFAULT_CALLBACK ((UConverterToUCallback) UCNV_TO_U_CALLBACK_SUBSTITUTE)
+#define UCNV_FROM_U_DEFAULT_CALLBACK ((UConverterFromUCallback) UCNV_FROM_U_CALLBACK_SUBSTITUTE)
+
+#endif
+
+#endif /* _UCNV_IMP */
diff --git a/thirdparty/icu4c/common/ucnv_io.cpp b/thirdparty/icu4c/common/ucnv_io.cpp
new file mode 100644
index 0000000000..7a95a3f1e6
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_io.cpp
@@ -0,0 +1,1360 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*
+*  ucnv_io.cpp:
+*  initializes global variables and defines functions pertaining to converter 
+*  name resolution aspect of the conversion code.
+*
+*   new implementation:
+*
+*   created on: 1999nov22
+*   created by: Markus W. Scherer
+*
+*   Use the binary cnvalias.icu (created from convrtrs.txt) to work
+*   with aliases for converter names.
+*
+*   Date        Name        Description
+*   11/22/1999  markus      Created
+*   06/28/2002  grhoten     Major overhaul of the converter alias design.
+*                           Now an alias can map to different converters
+*                           depending on the specified standard.
+*******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/udata.h"
+
+#include "umutex.h"
+#include "uarrsort.h"
+#include "uassert.h"
+#include "udataswp.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "ucnv_io.h"
+#include "uenumimp.h"
+#include "ucln_cmn.h"
+
+/* Format of cnvalias.icu -----------------------------------------------------
+ *
+ * cnvalias.icu is a binary, memory-mappable form of convrtrs.txt.
+ * This binary form contains several tables. All indexes are to uint16_t
+ * units, and not to the bytes (uint8_t units). Addressing everything on
+ * 16-bit boundaries allows us to store more information with small index
+ * numbers, which are also 16-bit in size. The majority of the table (except
+ * the string table) are 16-bit numbers.
+ *
+ * First there is the size of the Table of Contents (TOC). The TOC
+ * entries contain the size of each section. In order to find the offset
+ * you just need to sum up the previous offsets.
+ * The TOC length and entries are an array of uint32_t values.
+ * The first section after the TOC starts immediately after the TOC.
+ *
+ * 1) This section contains a list of converters. This list contains indexes
+ * into the string table for the converter name. The index of this list is
+ * also used by other sections, which are mentioned later on.
+ * This list is not sorted.
+ *
+ * 2) This section contains a list of tags. This list contains indexes
+ * into the string table for the tag name. The index of this list is
+ * also used by other sections, which are mentioned later on.
+ * This list is in priority order of standards.
+ *
+ * 3) This section contains a list of sorted unique aliases. This
+ * list contains indexes into the string table for the alias name. The
+ * index of this list is also used by other sections, like the 4th section.
+ * The index for the 3rd and 4th section is used to get the
+ * alias -> converter name mapping. Section 3 and 4 form a two column table.
+ * Some of the most significant bits of each index may contain other
+ * information (see findConverter for details).
+ *
+ * 4) This section contains a list of mapped converter names. Consider this
+ * as a table that maps the 3rd section to the 1st section. This list contains
+ * indexes into the 1st section. The index of this list is the same index in
+ * the 3rd section. There is also some extra information in the high bits of
+ * each converter index in this table. Currently it's only used to say that
+ * an alias mapped to this converter is ambiguous. See UCNV_CONVERTER_INDEX_MASK
+ * and UCNV_AMBIGUOUS_ALIAS_MAP_BIT for more information. This section is
+ * the predigested form of the 5th section so that an alias lookup can be fast.
+ *
+ * 5) This section contains a 2D array with indexes to the 6th section. This
+ * section is the full form of all alias mappings. The column index is the
+ * index into the converter list (column header). The row index is the index
+ * to tag list (row header). This 2D array is the top part a 3D array. The
+ * third dimension is in the 6th section.
+ *
+ * 6) This is blob of variable length arrays. Each array starts with a size,
+ * and is followed by indexes to alias names in the string table. This is
+ * the third dimension to the section 5. No other section should be referencing
+ * this section.
+ *
+ * 7) Starting in ICU 3.6, this can be a UConverterAliasOptions struct. Its
+ * presence indicates that a section 9 exists. UConverterAliasOptions specifies
+ * what type of string normalization is used among other potential things in the
+ * future.
+ *
+ * 8) This is the string table. All strings are indexed on an even address.
+ * There are two reasons for this. First many chip architectures locate strings
+ * faster on even address boundaries. Second, since all indexes are 16-bit
+ * numbers, this string table can be 128KB in size instead of 64KB when we
+ * only have strings starting on an even address.
+ *
+ * 9) When present this is a set of prenormalized strings from section 8. This
+ * table contains normalized strings with the dashes and spaces stripped out,
+ * and all strings lowercased. In the future, the options in section 7 may state
+ * other types of normalization.
+ *
+ * Here is the concept of section 5 and 6. It's a 3D cube. Each tag
+ * has a unique alias among all converters. That same alias can
+ * be mentioned in other standards on different converters,
+ * but only one alias per tag can be unique.
+ *
+ *
+ *              Converter Names (Usually in TR22 form)
+ *           -------------------------------------------.
+ *     T    /                                          /|
+ *     a   /                                          / |
+ *     g  /                                          /  |
+ *     s /                                          /   |
+ *      /                                          /    |
+ *      ------------------------------------------/     |
+ *    A |                                         |     |
+ *    l |                                         |     |
+ *    i |                                         |    /
+ *    a |                                         |   /
+ *    s |                                         |  /
+ *    e |                                         | /
+ *    s |                                         |/
+ *      -------------------------------------------
+ *
+ *
+ *
+ * Here is what it really looks like. It's like swiss cheese.
+ * There are holes. Some converters aren't recognized by
+ * a standard, or they are really old converters that the
+ * standard doesn't recognize anymore.
+ *
+ *              Converter Names (Usually in TR22 form)
+ *           -------------------------------------------.
+ *     T    /##########################################/|
+ *     a   /     #            #                       /#
+ *     g  /  #      ##     ##     ### # ### ### ### #/
+ *     s / #             #####  ####        ##  ## #/#
+ *      / ### # # ##  #  #   #          ### # #   #/##
+ *      ------------------------------------------/# #
+ *    A |### # # ##  #  #   #          ### # #   #|# #
+ *    l |# # #    #     #               ## #     #|# #
+ *    i |# # #    #     #                #       #|#
+ *    a |#                                       #|#
+ *    s |                                        #|#
+ *    e
+ *    s
+ *
+ */
+
+/**
+ * Used by the UEnumeration API
+ */
+typedef struct UAliasContext {
+    uint32_t listOffset;
+    uint32_t listIdx;
+} UAliasContext;
+
+static const char DATA_NAME[] = "cnvalias";
+static const char DATA_TYPE[] = "icu";
+
+static UDataMemory *gAliasData=NULL;
+static icu::UInitOnce gAliasDataInitOnce = U_INITONCE_INITIALIZER;
+
+enum {
+    tocLengthIndex=0,
+    converterListIndex=1,
+    tagListIndex=2,
+    aliasListIndex=3,
+    untaggedConvArrayIndex=4,
+    taggedAliasArrayIndex=5,
+    taggedAliasListsIndex=6,
+    tableOptionsIndex=7,
+    stringTableIndex=8,
+    normalizedStringTableIndex=9,
+    offsetsCount,    /* length of the swapper's temporary offsets[] */
+    minTocLength=8 /* min. tocLength in the file, does not count the tocLengthIndex! */
+};
+
+static const UConverterAliasOptions defaultTableOptions = {
+    UCNV_IO_UNNORMALIZED,
+    0 /* containsCnvOptionInfo */
+};
+static UConverterAlias gMainTable;
+
+#define GET_STRING(idx) (const char *)(gMainTable.stringTable + (idx))
+#define GET_NORMALIZED_STRING(idx) (const char *)(gMainTable.normalizedStringTable + (idx))
+
+static UBool U_CALLCONV
+isAcceptable(void * /*context*/,
+             const char * /*type*/, const char * /*name*/,
+             const UDataInfo *pInfo) {
+    return (UBool)(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->dataFormat[0]==0x43 &&   /* dataFormat="CvAl" */
+        pInfo->dataFormat[1]==0x76 &&
+        pInfo->dataFormat[2]==0x41 &&
+        pInfo->dataFormat[3]==0x6c &&
+        pInfo->formatVersion[0]==3);
+}
+
+static UBool U_CALLCONV ucnv_io_cleanup(void)
+{
+    if (gAliasData) {
+        udata_close(gAliasData);
+        gAliasData = NULL;
+    }
+    gAliasDataInitOnce.reset();
+
+    uprv_memset(&gMainTable, 0, sizeof(gMainTable));
+
+    return TRUE;                   /* Everything was cleaned up */
+}
+
+static void U_CALLCONV initAliasData(UErrorCode &errCode) {
+    UDataMemory *data;
+    const uint16_t *table;
+    const uint32_t *sectionSizes;
+    uint32_t tableStart;
+    uint32_t currOffset;
+
+    ucln_common_registerCleanup(UCLN_COMMON_UCNV_IO, ucnv_io_cleanup);
+
+    U_ASSERT(gAliasData == NULL);
+    data = udata_openChoice(NULL, DATA_TYPE, DATA_NAME, isAcceptable, NULL, &errCode);
+    if(U_FAILURE(errCode)) {
+        return;
+    }
+
+    sectionSizes = (const uint32_t *)udata_getMemory(data);
+    table = (const uint16_t *)sectionSizes;
+
+    tableStart      = sectionSizes[0];
+    if (tableStart < minTocLength) {
+        errCode = U_INVALID_FORMAT_ERROR;
+        udata_close(data);
+        return;
+    }
+    gAliasData = data;
+
+    gMainTable.converterListSize      = sectionSizes[1];
+    gMainTable.tagListSize            = sectionSizes[2];
+    gMainTable.aliasListSize          = sectionSizes[3];
+    gMainTable.untaggedConvArraySize  = sectionSizes[4];
+    gMainTable.taggedAliasArraySize   = sectionSizes[5];
+    gMainTable.taggedAliasListsSize   = sectionSizes[6];
+    gMainTable.optionTableSize        = sectionSizes[7];
+    gMainTable.stringTableSize        = sectionSizes[8];
+
+    if (tableStart > 8) {
+        gMainTable.normalizedStringTableSize = sectionSizes[9];
+    }
+
+    currOffset = tableStart * (sizeof(uint32_t)/sizeof(uint16_t)) + (sizeof(uint32_t)/sizeof(uint16_t));
+    gMainTable.converterList = table + currOffset;
+
+    currOffset += gMainTable.converterListSize;
+    gMainTable.tagList = table + currOffset;
+
+    currOffset += gMainTable.tagListSize;
+    gMainTable.aliasList = table + currOffset;
+
+    currOffset += gMainTable.aliasListSize;
+    gMainTable.untaggedConvArray = table + currOffset;
+
+    currOffset += gMainTable.untaggedConvArraySize;
+    gMainTable.taggedAliasArray = table + currOffset;
+
+    /* aliasLists is a 1's based array, but it has a padding character */
+    currOffset += gMainTable.taggedAliasArraySize;
+    gMainTable.taggedAliasLists = table + currOffset;
+
+    currOffset += gMainTable.taggedAliasListsSize;
+    if (gMainTable.optionTableSize > 0
+        && ((const UConverterAliasOptions *)(table + currOffset))->stringNormalizationType < UCNV_IO_NORM_TYPE_COUNT)
+    {
+        /* Faster table */
+        gMainTable.optionTable = (const UConverterAliasOptions *)(table + currOffset);
+    }
+    else {
+        /* Smaller table, or I can't handle this normalization mode!
+        Use the original slower table lookup. */
+        gMainTable.optionTable = &defaultTableOptions;
+    }
+
+    currOffset += gMainTable.optionTableSize;
+    gMainTable.stringTable = table + currOffset;
+
+    currOffset += gMainTable.stringTableSize;
+    gMainTable.normalizedStringTable = ((gMainTable.optionTable->stringNormalizationType == UCNV_IO_UNNORMALIZED)
+        ? gMainTable.stringTable : (table + currOffset));
+}
+
+
+static UBool
+haveAliasData(UErrorCode *pErrorCode) {
+    umtx_initOnce(gAliasDataInitOnce, &initAliasData, *pErrorCode);
+    return U_SUCCESS(*pErrorCode);
+}
+
+static inline UBool
+isAlias(const char *alias, UErrorCode *pErrorCode) {
+    if(alias==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    return (UBool)(*alias!=0);
+}
+
+static uint32_t getTagNumber(const char *tagname) {
+    if (gMainTable.tagList) {
+        uint32_t tagNum;
+        for (tagNum = 0; tagNum < gMainTable.tagListSize; tagNum++) {
+            if (!uprv_stricmp(GET_STRING(gMainTable.tagList[tagNum]), tagname)) {
+                return tagNum;
+            }
+        }
+    }
+
+    return UINT32_MAX;
+}
+
+/* character types relevant for ucnv_compareNames() */
+enum {
+    UIGNORE,
+    ZERO,
+    NONZERO,
+    MINLETTER /* any values from here on are lowercase letter mappings */
+};
+
+/* character types for ASCII 00..7F */
+static const uint8_t asciiTypes[128] = {
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    ZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, 0, 0, 0, 0, 0, 0,
+    0, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
+    0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0, 0, 0, 0, 0,
+    0, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
+    0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0, 0, 0, 0, 0
+};
+
+#define GET_ASCII_TYPE(c) ((int8_t)(c) >= 0 ? asciiTypes[(uint8_t)c] : (uint8_t)UIGNORE)
+
+/* character types for EBCDIC 80..FF */
+static const uint8_t ebcdicTypes[128] = {
+    0,    0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89, 0, 0, 0, 0, 0, 0,
+    0,    0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, 0x98, 0x99, 0, 0, 0, 0, 0, 0,
+    0,    0,    0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0,    0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89, 0, 0, 0, 0, 0, 0,
+    0,    0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, 0x98, 0x99, 0, 0, 0, 0, 0, 0,
+    0,    0,    0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0, 0, 0, 0, 0, 0,
+    ZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, NONZERO, 0, 0, 0, 0, 0, 0
+};
+
+#define GET_EBCDIC_TYPE(c) ((int8_t)(c) < 0 ? ebcdicTypes[(c)&0x7f] : (uint8_t)UIGNORE)
+
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define GET_CHAR_TYPE(c) GET_ASCII_TYPE(c)
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define GET_CHAR_TYPE(c) GET_EBCDIC_TYPE(c)
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+
+
+/* @see ucnv_compareNames */
+U_CAPI char * U_CALLCONV
+ucnv_io_stripASCIIForCompare(char *dst, const char *name) {
+    char *dstItr = dst;
+    uint8_t type, nextType;
+    char c1;
+    UBool afterDigit = FALSE;
+
+    while ((c1 = *name++) != 0) {
+        type = GET_ASCII_TYPE(c1);
+        switch (type) {
+        case UIGNORE:
+            afterDigit = FALSE;
+            continue; /* ignore all but letters and digits */
+        case ZERO:
+            if (!afterDigit) {
+                nextType = GET_ASCII_TYPE(*name);
+                if (nextType == ZERO || nextType == NONZERO) {
+                    continue; /* ignore leading zero before another digit */
+                }
+            }
+            break;
+        case NONZERO:
+            afterDigit = TRUE;
+            break;
+        default:
+            c1 = (char)type; /* lowercased letter */
+            afterDigit = FALSE;
+            break;
+        }
+        *dstItr++ = c1;
+    }
+    *dstItr = 0;
+    return dst;
+}
+
+U_CAPI char * U_CALLCONV
+ucnv_io_stripEBCDICForCompare(char *dst, const char *name) {
+    char *dstItr = dst;
+    uint8_t type, nextType;
+    char c1;
+    UBool afterDigit = FALSE;
+
+    while ((c1 = *name++) != 0) {
+        type = GET_EBCDIC_TYPE(c1);
+        switch (type) {
+        case UIGNORE:
+            afterDigit = FALSE;
+            continue; /* ignore all but letters and digits */
+        case ZERO:
+            if (!afterDigit) {
+                nextType = GET_EBCDIC_TYPE(*name);
+                if (nextType == ZERO || nextType == NONZERO) {
+                    continue; /* ignore leading zero before another digit */
+                }
+            }
+            break;
+        case NONZERO:
+            afterDigit = TRUE;
+            break;
+        default:
+            c1 = (char)type; /* lowercased letter */
+            afterDigit = FALSE;
+            break;
+        }
+        *dstItr++ = c1;
+    }
+    *dstItr = 0;
+    return dst;
+}
+
+/**
+ * Do a fuzzy compare of two converter/alias names.
+ * The comparison is case-insensitive, ignores leading zeroes if they are not
+ * followed by further digits, and ignores all but letters and digits.
+ * Thus the strings "UTF-8", "utf_8", "u*T@f08" and "Utf 8" are exactly equivalent.
+ * See section 1.4, Charset Alias Matching in Unicode Technical Standard #22
+ * at http://www.unicode.org/reports/tr22/
+ *
+ * This is a symmetrical (commutative) operation; order of arguments
+ * is insignificant.  This is an important property for sorting the
+ * list (when the list is preprocessed into binary form) and for
+ * performing binary searches on it at run time.
+ *
+ * @param name1 a converter name or alias, zero-terminated
+ * @param name2 a converter name or alias, zero-terminated
+ * @return 0 if the names match, or a negative value if the name1
+ * lexically precedes name2, or a positive value if the name1
+ * lexically follows name2.
+ *
+ * @see ucnv_io_stripForCompare
+ */
+U_CAPI int U_EXPORT2
+ucnv_compareNames(const char *name1, const char *name2) {
+    int rc;
+    uint8_t type, nextType;
+    char c1, c2;
+    UBool afterDigit1 = FALSE, afterDigit2 = FALSE;
+
+    for (;;) {
+        while ((c1 = *name1++) != 0) {
+            type = GET_CHAR_TYPE(c1);
+            switch (type) {
+            case UIGNORE:
+                afterDigit1 = FALSE;
+                continue; /* ignore all but letters and digits */
+            case ZERO:
+                if (!afterDigit1) {
+                    nextType = GET_CHAR_TYPE(*name1);
+                    if (nextType == ZERO || nextType == NONZERO) {
+                        continue; /* ignore leading zero before another digit */
+                    }
+                }
+                break;
+            case NONZERO:
+                afterDigit1 = TRUE;
+                break;
+            default:
+                c1 = (char)type; /* lowercased letter */
+                afterDigit1 = FALSE;
+                break;
+            }
+            break; /* deliver c1 */
+        }
+        while ((c2 = *name2++) != 0) {
+            type = GET_CHAR_TYPE(c2);
+            switch (type) {
+            case UIGNORE:
+                afterDigit2 = FALSE;
+                continue; /* ignore all but letters and digits */
+            case ZERO:
+                if (!afterDigit2) {
+                    nextType = GET_CHAR_TYPE(*name2);
+                    if (nextType == ZERO || nextType == NONZERO) {
+                        continue; /* ignore leading zero before another digit */
+                    }
+                }
+                break;
+            case NONZERO:
+                afterDigit2 = TRUE;
+                break;
+            default:
+                c2 = (char)type; /* lowercased letter */
+                afterDigit2 = FALSE;
+                break;
+            }
+            break; /* deliver c2 */
+        }
+
+        /* If we reach the ends of both strings then they match */
+        if ((c1|c2)==0) {
+            return 0;
+        }
+
+        /* Case-insensitive comparison */
+        rc = (int)(unsigned char)c1 - (int)(unsigned char)c2;
+        if (rc != 0) {
+            return rc;
+        }
+    }
+}
+
+/*
+ * search for an alias
+ * return the converter number index for gConverterList
+ */
+static inline uint32_t
+findConverter(const char *alias, UBool *containsOption, UErrorCode *pErrorCode) {
+    uint32_t mid, start, limit;
+    uint32_t lastMid;
+    int result;
+    int isUnnormalized = (gMainTable.optionTable->stringNormalizationType == UCNV_IO_UNNORMALIZED);
+    char strippedName[UCNV_MAX_CONVERTER_NAME_LENGTH];
+
+    if (!isUnnormalized) {
+        if (uprv_strlen(alias) >= UCNV_MAX_CONVERTER_NAME_LENGTH) {
+            *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+            return UINT32_MAX;
+        }
+
+        /* Lower case and remove ignoreable characters. */
+        ucnv_io_stripForCompare(strippedName, alias);
+        alias = strippedName;
+    }
+
+    /* do a binary search for the alias */
+    start = 0;
+    limit = gMainTable.untaggedConvArraySize;
+    mid = limit;
+    lastMid = UINT32_MAX;
+
+    for (;;) {
+        mid = (uint32_t)((start + limit) / 2);
+        if (lastMid == mid) {   /* Have we moved? */
+            break;  /* We haven't moved, and it wasn't found. */
+        }
+        lastMid = mid;
+        if (isUnnormalized) {
+            result = ucnv_compareNames(alias, GET_STRING(gMainTable.aliasList[mid]));
+        }
+        else {
+            result = uprv_strcmp(alias, GET_NORMALIZED_STRING(gMainTable.aliasList[mid]));
+        }
+
+        if (result < 0) {
+            limit = mid;
+        } else if (result > 0) {
+            start = mid;
+        } else {
+            /* Since the gencnval tool folds duplicates into one entry,
+             * this alias in gAliasList is unique, but different standards
+             * may map an alias to different converters.
+             */
+            if (gMainTable.untaggedConvArray[mid] & UCNV_AMBIGUOUS_ALIAS_MAP_BIT) {
+                *pErrorCode = U_AMBIGUOUS_ALIAS_WARNING;
+            }
+            /* State whether the canonical converter name contains an option.
+            This information is contained in this list in order to maintain backward & forward compatibility. */
+            if (containsOption) {
+                UBool containsCnvOptionInfo = (UBool)gMainTable.optionTable->containsCnvOptionInfo;
+                *containsOption = (UBool)((containsCnvOptionInfo
+                    && ((gMainTable.untaggedConvArray[mid] & UCNV_CONTAINS_OPTION_BIT) != 0))
+                    || !containsCnvOptionInfo);
+            }
+            return gMainTable.untaggedConvArray[mid] & UCNV_CONVERTER_INDEX_MASK;
+        }
+    }
+
+    return UINT32_MAX;
+}
+
+/*
+ * Is this alias in this list?
+ * alias and listOffset should be non-NULL.
+ */
+static inline UBool
+isAliasInList(const char *alias, uint32_t listOffset) {
+    if (listOffset) {
+        uint32_t currAlias;
+        uint32_t listCount = gMainTable.taggedAliasLists[listOffset];
+        /* +1 to skip listCount */
+        const uint16_t *currList = gMainTable.taggedAliasLists + listOffset + 1;
+        for (currAlias = 0; currAlias < listCount; currAlias++) {
+            if (currList[currAlias]
+                && ucnv_compareNames(alias, GET_STRING(currList[currAlias]))==0)
+            {
+                return TRUE;
+            }
+        }
+    }
+    return FALSE;
+}
+
+/*
+ * Search for an standard name of an alias (what is the default name
+ * that this standard uses?)
+ * return the listOffset for gTaggedAliasLists. If it's 0,
+ * the it couldn't be found, but the parameters are valid.
+ */
+static uint32_t
+findTaggedAliasListsOffset(const char *alias, const char *standard, UErrorCode *pErrorCode) {
+    uint32_t idx;
+    uint32_t listOffset;
+    uint32_t convNum;
+    UErrorCode myErr = U_ZERO_ERROR;
+    uint32_t tagNum = getTagNumber(standard);
+
+    /* Make a quick guess. Hopefully they used a TR22 canonical alias. */
+    convNum = findConverter(alias, NULL, &myErr);
+    if (myErr != U_ZERO_ERROR) {
+        *pErrorCode = myErr;
+    }
+
+    if (tagNum < (gMainTable.tagListSize - UCNV_NUM_HIDDEN_TAGS) && convNum < gMainTable.converterListSize) {
+        listOffset = gMainTable.taggedAliasArray[tagNum*gMainTable.converterListSize + convNum];
+        if (listOffset && gMainTable.taggedAliasLists[listOffset + 1]) {
+            return listOffset;
+        }
+        if (myErr == U_AMBIGUOUS_ALIAS_WARNING) {
+            /* Uh Oh! They used an ambiguous alias.
+               We have to search the whole swiss cheese starting
+               at the highest standard affinity.
+               This may take a while.
+            */
+            for (idx = 0; idx < gMainTable.taggedAliasArraySize; idx++) {
+                listOffset = gMainTable.taggedAliasArray[idx];
+                if (listOffset && isAliasInList(alias, listOffset)) {
+                    uint32_t currTagNum = idx/gMainTable.converterListSize;
+                    uint32_t currConvNum = (idx - currTagNum*gMainTable.converterListSize);
+                    uint32_t tempListOffset = gMainTable.taggedAliasArray[tagNum*gMainTable.converterListSize + currConvNum];
+                    if (tempListOffset && gMainTable.taggedAliasLists[tempListOffset + 1]) {
+                        return tempListOffset;
+                    }
+                    /* else keep on looking */
+                    /* We could speed this up by starting on the next row
+                       because an alias is unique per row, right now.
+                       This would change if alias versioning appears. */
+                }
+            }
+            /* The standard doesn't know about the alias */
+        }
+        /* else no default name */
+        return 0;
+    }
+    /* else converter or tag not found */
+
+    return UINT32_MAX;
+}
+
+/* Return the canonical name */
+static uint32_t
+findTaggedConverterNum(const char *alias, const char *standard, UErrorCode *pErrorCode) {
+    uint32_t idx;
+    uint32_t listOffset;
+    uint32_t convNum;
+    UErrorCode myErr = U_ZERO_ERROR;
+    uint32_t tagNum = getTagNumber(standard);
+
+    /* Make a quick guess. Hopefully they used a TR22 canonical alias. */
+    convNum = findConverter(alias, NULL, &myErr);
+    if (myErr != U_ZERO_ERROR) {
+        *pErrorCode = myErr;
+    }
+
+    if (tagNum < (gMainTable.tagListSize - UCNV_NUM_HIDDEN_TAGS) && convNum < gMainTable.converterListSize) {
+        listOffset = gMainTable.taggedAliasArray[tagNum*gMainTable.converterListSize + convNum];
+        if (listOffset && isAliasInList(alias, listOffset)) {
+            return convNum;
+        }
+        if (myErr == U_AMBIGUOUS_ALIAS_WARNING) {
+            /* Uh Oh! They used an ambiguous alias.
+               We have to search one slice of the swiss cheese.
+               We search only in the requested tag, not the whole thing.
+               This may take a while.
+            */
+            uint32_t convStart = (tagNum)*gMainTable.converterListSize;
+            uint32_t convLimit = (tagNum+1)*gMainTable.converterListSize;
+            for (idx = convStart; idx < convLimit; idx++) {
+                listOffset = gMainTable.taggedAliasArray[idx];
+                if (listOffset && isAliasInList(alias, listOffset)) {
+                    return idx-convStart;
+                }
+            }
+            /* The standard doesn't know about the alias */
+        }
+        /* else no canonical name */
+    }
+    /* else converter or tag not found */
+
+    return UINT32_MAX;
+}
+
+U_CAPI const char *
+ucnv_io_getConverterName(const char *alias, UBool *containsOption, UErrorCode *pErrorCode) {
+    const char *aliasTmp = alias;
+    int32_t i = 0;
+    for (i = 0; i < 2; i++) {
+        if (i == 1) {
+            /*
+             * After the first unsuccess converter lookup, check to see if
+             * the name begins with 'x-'. If it does, strip it off and try
+             * again.  This behaviour is similar to how ICU4J does it.
+             */
+            if (aliasTmp[0] == 'x' && aliasTmp[1] == '-') {
+                aliasTmp = aliasTmp+2;
+            } else {
+                break;
+            }
+        }
+        if(haveAliasData(pErrorCode) && isAlias(aliasTmp, pErrorCode)) {
+            uint32_t convNum = findConverter(aliasTmp, containsOption, pErrorCode);
+            if (convNum < gMainTable.converterListSize) {
+                return GET_STRING(gMainTable.converterList[convNum]);
+            }
+            /* else converter not found */
+        } else {
+            break;
+        }
+    }
+
+    return NULL;
+}
+
+U_CDECL_BEGIN
+
+
+static int32_t U_CALLCONV
+ucnv_io_countStandardAliases(UEnumeration *enumerator, UErrorCode * /*pErrorCode*/) {
+    int32_t value = 0;
+    UAliasContext *myContext = (UAliasContext *)(enumerator->context);
+    uint32_t listOffset = myContext->listOffset;
+
+    if (listOffset) {
+        value = gMainTable.taggedAliasLists[listOffset];
+    }
+    return value;
+}
+
+static const char * U_CALLCONV
+ucnv_io_nextStandardAliases(UEnumeration *enumerator,
+                            int32_t* resultLength,
+                            UErrorCode * /*pErrorCode*/)
+{
+    UAliasContext *myContext = (UAliasContext *)(enumerator->context);
+    uint32_t listOffset = myContext->listOffset;
+
+    if (listOffset) {
+        uint32_t listCount = gMainTable.taggedAliasLists[listOffset];
+        const uint16_t *currList = gMainTable.taggedAliasLists + listOffset + 1;
+
+        if (myContext->listIdx < listCount) {
+            const char *myStr = GET_STRING(currList[myContext->listIdx++]);
+            if (resultLength) {
+                *resultLength = (int32_t)uprv_strlen(myStr);
+            }
+            return myStr;
+        }
+    }
+    /* Either we accessed a zero length list, or we enumerated too far. */
+    if (resultLength) {
+        *resultLength = 0;
+    }
+    return NULL;
+}
+
+static void U_CALLCONV
+ucnv_io_resetStandardAliases(UEnumeration *enumerator, UErrorCode * /*pErrorCode*/) {
+    ((UAliasContext *)(enumerator->context))->listIdx = 0;
+}
+
+static void U_CALLCONV
+ucnv_io_closeUEnumeration(UEnumeration *enumerator) {
+    uprv_free(enumerator->context);
+    uprv_free(enumerator);
+}
+
+U_CDECL_END
+
+/* Enumerate the aliases for the specified converter and standard tag */
+static const UEnumeration gEnumAliases = {
+    NULL,
+    NULL,
+    ucnv_io_closeUEnumeration,
+    ucnv_io_countStandardAliases,
+    uenum_unextDefault,
+    ucnv_io_nextStandardAliases,
+    ucnv_io_resetStandardAliases
+};
+
+U_CAPI UEnumeration * U_EXPORT2
+ucnv_openStandardNames(const char *convName,
+                       const char *standard,
+                       UErrorCode *pErrorCode)
+{
+    UEnumeration *myEnum = NULL;
+    if (haveAliasData(pErrorCode) && isAlias(convName, pErrorCode)) {
+        uint32_t listOffset = findTaggedAliasListsOffset(convName, standard, pErrorCode);
+
+        /* When listOffset == 0, we want to acknowledge that the
+           converter name and standard are okay, but there
+           is nothing to enumerate. */
+        if (listOffset < gMainTable.taggedAliasListsSize) {
+            UAliasContext *myContext;
+
+            myEnum = static_cast<UEnumeration *>(uprv_malloc(sizeof(UEnumeration)));
+            if (myEnum == NULL) {
+                *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                return NULL;
+            }
+            uprv_memcpy(myEnum, &gEnumAliases, sizeof(UEnumeration));
+            myContext = static_cast<UAliasContext *>(uprv_malloc(sizeof(UAliasContext)));
+            if (myContext == NULL) {
+                *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                uprv_free(myEnum);
+                return NULL;
+            }
+            myContext->listOffset = listOffset;
+            myContext->listIdx = 0;
+            myEnum->context = myContext;
+        }
+        /* else converter or tag not found */
+    }
+    return myEnum;
+}
+
+static uint16_t
+ucnv_io_countAliases(const char *alias, UErrorCode *pErrorCode) {
+    if(haveAliasData(pErrorCode) && isAlias(alias, pErrorCode)) {
+        uint32_t convNum = findConverter(alias, NULL, pErrorCode);
+        if (convNum < gMainTable.converterListSize) {
+            /* tagListNum - 1 is the ALL tag */
+            int32_t listOffset = gMainTable.taggedAliasArray[(gMainTable.tagListSize - 1)*gMainTable.converterListSize + convNum];
+
+            if (listOffset) {
+                return gMainTable.taggedAliasLists[listOffset];
+            }
+            /* else this shouldn't happen. internal program error */
+        }
+        /* else converter not found */
+    }
+    return 0;
+}
+
+static uint16_t
+ucnv_io_getAliases(const char *alias, uint16_t start, const char **aliases, UErrorCode *pErrorCode) {
+    if(haveAliasData(pErrorCode) && isAlias(alias, pErrorCode)) {
+        uint32_t currAlias;
+        uint32_t convNum = findConverter(alias, NULL, pErrorCode);
+        if (convNum < gMainTable.converterListSize) {
+            /* tagListNum - 1 is the ALL tag */
+            int32_t listOffset = gMainTable.taggedAliasArray[(gMainTable.tagListSize - 1)*gMainTable.converterListSize + convNum];
+
+            if (listOffset) {
+                uint32_t listCount = gMainTable.taggedAliasLists[listOffset];
+                /* +1 to skip listCount */
+                const uint16_t *currList = gMainTable.taggedAliasLists + listOffset + 1;
+
+                for (currAlias = start; currAlias < listCount; currAlias++) {
+                    aliases[currAlias] = GET_STRING(currList[currAlias]);
+                }
+            }
+            /* else this shouldn't happen. internal program error */
+        }
+        /* else converter not found */
+    }
+    return 0;
+}
+
+static const char *
+ucnv_io_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode) {
+    if(haveAliasData(pErrorCode) && isAlias(alias, pErrorCode)) {
+        uint32_t convNum = findConverter(alias, NULL, pErrorCode);
+        if (convNum < gMainTable.converterListSize) {
+            /* tagListNum - 1 is the ALL tag */
+            int32_t listOffset = gMainTable.taggedAliasArray[(gMainTable.tagListSize - 1)*gMainTable.converterListSize + convNum];
+
+            if (listOffset) {
+                uint32_t listCount = gMainTable.taggedAliasLists[listOffset];
+                /* +1 to skip listCount */
+                const uint16_t *currList = gMainTable.taggedAliasLists + listOffset + 1;
+
+                if (n < listCount)  {
+                    return GET_STRING(currList[n]);
+                }
+                *pErrorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+            }
+            /* else this shouldn't happen. internal program error */
+        }
+        /* else converter not found */
+    }
+    return NULL;
+}
+
+static uint16_t
+ucnv_io_countStandards(UErrorCode *pErrorCode) {
+    if (haveAliasData(pErrorCode)) {
+        /* Don't include the empty list */
+        return (uint16_t)(gMainTable.tagListSize - UCNV_NUM_HIDDEN_TAGS);
+    }
+
+    return 0;
+}
+
+U_CAPI const char * U_EXPORT2
+ucnv_getStandard(uint16_t n, UErrorCode *pErrorCode) {
+    if (haveAliasData(pErrorCode)) {
+        if (n < gMainTable.tagListSize - UCNV_NUM_HIDDEN_TAGS) {
+            return GET_STRING(gMainTable.tagList[n]);
+        }
+        *pErrorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+    }
+
+    return NULL;
+}
+
+U_CAPI const char * U_EXPORT2
+ucnv_getStandardName(const char *alias, const char *standard, UErrorCode *pErrorCode) {
+    if (haveAliasData(pErrorCode) && isAlias(alias, pErrorCode)) {
+        uint32_t listOffset = findTaggedAliasListsOffset(alias, standard, pErrorCode);
+
+        if (0 < listOffset && listOffset < gMainTable.taggedAliasListsSize) {
+            const uint16_t *currList = gMainTable.taggedAliasLists + listOffset + 1;
+
+            /* Get the preferred name from this list */
+            if (currList[0]) {
+                return GET_STRING(currList[0]);
+            }
+            /* else someone screwed up the alias table. */
+            /* *pErrorCode = U_INVALID_FORMAT_ERROR */
+        }
+    }
+
+    return NULL;
+}
+
+U_CAPI uint16_t U_EXPORT2
+ucnv_countAliases(const char *alias, UErrorCode *pErrorCode)
+{
+    return ucnv_io_countAliases(alias, pErrorCode);
+}
+
+
+U_CAPI const char* U_EXPORT2
+ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode)
+{
+    return ucnv_io_getAlias(alias, n, pErrorCode);
+}
+
+U_CAPI void U_EXPORT2
+ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode)
+{
+    ucnv_io_getAliases(alias, 0, aliases, pErrorCode);
+}
+
+U_CAPI uint16_t U_EXPORT2
+ucnv_countStandards(void)
+{
+    UErrorCode err = U_ZERO_ERROR;
+    return ucnv_io_countStandards(&err);
+}
+
+U_CAPI const char * U_EXPORT2
+ucnv_getCanonicalName(const char *alias, const char *standard, UErrorCode *pErrorCode) {
+    if (haveAliasData(pErrorCode) && isAlias(alias, pErrorCode)) {
+        uint32_t convNum = findTaggedConverterNum(alias, standard, pErrorCode);
+
+        if (convNum < gMainTable.converterListSize) {
+            return GET_STRING(gMainTable.converterList[convNum]);
+        }
+    }
+
+    return NULL;
+}
+
+U_CDECL_BEGIN
+
+
+static int32_t U_CALLCONV
+ucnv_io_countAllConverters(UEnumeration * /*enumerator*/, UErrorCode * /*pErrorCode*/) {
+    return gMainTable.converterListSize;
+}
+
+static const char * U_CALLCONV
+ucnv_io_nextAllConverters(UEnumeration *enumerator,
+                            int32_t* resultLength,
+                            UErrorCode * /*pErrorCode*/)
+{
+    uint16_t *myContext = (uint16_t *)(enumerator->context);
+
+    if (*myContext < gMainTable.converterListSize) {
+        const char *myStr = GET_STRING(gMainTable.converterList[(*myContext)++]);
+        if (resultLength) {
+            *resultLength = (int32_t)uprv_strlen(myStr);
+        }
+        return myStr;
+    }
+    /* Either we accessed a zero length list, or we enumerated too far. */
+    if (resultLength) {
+        *resultLength = 0;
+    }
+    return NULL;
+}
+
+static void U_CALLCONV
+ucnv_io_resetAllConverters(UEnumeration *enumerator, UErrorCode * /*pErrorCode*/) {
+    *((uint16_t *)(enumerator->context)) = 0;
+}
+U_CDECL_END
+static const UEnumeration gEnumAllConverters = {
+    NULL,
+    NULL,
+    ucnv_io_closeUEnumeration,
+    ucnv_io_countAllConverters,
+    uenum_unextDefault,
+    ucnv_io_nextAllConverters,
+    ucnv_io_resetAllConverters
+};
+
+U_CAPI UEnumeration * U_EXPORT2
+ucnv_openAllNames(UErrorCode *pErrorCode) {
+    UEnumeration *myEnum = NULL;
+    if (haveAliasData(pErrorCode)) {
+        uint16_t *myContext;
+
+        myEnum = static_cast<UEnumeration *>(uprv_malloc(sizeof(UEnumeration)));
+        if (myEnum == NULL) {
+            *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+        uprv_memcpy(myEnum, &gEnumAllConverters, sizeof(UEnumeration));
+        myContext = static_cast<uint16_t *>(uprv_malloc(sizeof(uint16_t)));
+        if (myContext == NULL) {
+            *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+            uprv_free(myEnum);
+            return NULL;
+        }
+        *myContext = 0;
+        myEnum->context = myContext;
+    }
+    return myEnum;
+}
+
+U_CAPI uint16_t
+ucnv_io_countKnownConverters(UErrorCode *pErrorCode) {
+    if (haveAliasData(pErrorCode)) {
+        return (uint16_t)gMainTable.converterListSize;
+    }
+    return 0;
+}
+
+/* alias table swapping ----------------------------------------------------- */
+
+U_CDECL_BEGIN
+
+typedef char * U_CALLCONV StripForCompareFn(char *dst, const char *name);
+U_CDECL_END
+
+
+/*
+ * row of a temporary array
+ *
+ * gets platform-endian charset string indexes and sorting indexes;
+ * after sorting this array by strings, the actual arrays are permutated
+ * according to the sorting indexes
+ */
+typedef struct TempRow {
+    uint16_t strIndex, sortIndex;
+} TempRow;
+
+typedef struct TempAliasTable {
+    const char *chars;
+    TempRow *rows;
+    uint16_t *resort;
+    StripForCompareFn *stripForCompare;
+} TempAliasTable;
+
+enum {
+    STACK_ROW_CAPACITY=500
+};
+
+static int32_t U_CALLCONV
+io_compareRows(const void *context, const void *left, const void *right) {
+    char strippedLeft[UCNV_MAX_CONVERTER_NAME_LENGTH],
+         strippedRight[UCNV_MAX_CONVERTER_NAME_LENGTH];
+
+    TempAliasTable *tempTable=(TempAliasTable *)context;
+    const char *chars=tempTable->chars;
+
+    return (int32_t)uprv_strcmp(tempTable->stripForCompare(strippedLeft, chars+2*((const TempRow *)left)->strIndex),
+                                tempTable->stripForCompare(strippedRight, chars+2*((const TempRow *)right)->strIndex));
+}
+
+U_CAPI int32_t U_EXPORT2
+ucnv_swapAliases(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint16_t *inTable;
+    const uint32_t *inSectionSizes;
+    uint32_t toc[offsetsCount];
+    uint32_t offsets[offsetsCount]; /* 16-bit-addressed offsets from inTable/outTable */
+    uint32_t i, count, tocLength, topOffset;
+
+    TempRow rows[STACK_ROW_CAPACITY];
+    uint16_t resort[STACK_ROW_CAPACITY];
+    TempAliasTable tempTable;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x43 &&   /* dataFormat="CvAl" */
+        pInfo->dataFormat[1]==0x76 &&
+        pInfo->dataFormat[2]==0x41 &&
+        pInfo->dataFormat[3]==0x6c &&
+        pInfo->formatVersion[0]==3
+    )) {
+        udata_printError(ds, "ucnv_swapAliases(): data format %02x.%02x.%02x.%02x (format version %02x) is not an alias table\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    /* an alias table must contain at least the table of contents array */
+    if(length>=0 && (length-headerSize)<4*(1+minTocLength)) {
+        udata_printError(ds, "ucnv_swapAliases(): too few bytes (%d after header) for an alias table\n",
+                         length-headerSize);
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    inSectionSizes=(const uint32_t *)((const char *)inData+headerSize);
+    inTable=(const uint16_t *)inSectionSizes;
+    uprv_memset(toc, 0, sizeof(toc));
+    toc[tocLengthIndex]=tocLength=ds->readUInt32(inSectionSizes[tocLengthIndex]);
+    if(tocLength<minTocLength || offsetsCount<=tocLength) {
+        udata_printError(ds, "ucnv_swapAliases(): table of contents contains unsupported number of sections (%u sections)\n", tocLength);
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    /* read the known part of the table of contents */
+    for(i=converterListIndex; i<=tocLength; ++i) {
+        toc[i]=ds->readUInt32(inSectionSizes[i]);
+    }
+
+    /* compute offsets */
+    uprv_memset(offsets, 0, sizeof(offsets));
+    offsets[converterListIndex]=2*(1+tocLength); /* count two 16-bit units per toc entry */
+    for(i=tagListIndex; i<=tocLength; ++i) {
+        offsets[i]=offsets[i-1]+toc[i-1];
+    }
+
+    /* compute the overall size of the after-header data, in numbers of 16-bit units */
+    topOffset=offsets[i-1]+toc[i-1];
+
+    if(length>=0) {
+        uint16_t *outTable;
+        const uint16_t *p, *p2;
+        uint16_t *q, *q2;
+        uint16_t oldIndex;
+
+        if((length-headerSize)<(2*(int32_t)topOffset)) {
+            udata_printError(ds, "ucnv_swapAliases(): too few bytes (%d after header) for an alias table\n",
+                             length-headerSize);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        outTable=(uint16_t *)((char *)outData+headerSize);
+
+        /* swap the entire table of contents */
+        ds->swapArray32(ds, inTable, 4*(1+tocLength), outTable, pErrorCode);
+
+        /* swap unormalized strings & normalized strings */
+        ds->swapInvChars(ds, inTable+offsets[stringTableIndex], 2*(int32_t)(toc[stringTableIndex]+toc[normalizedStringTableIndex]),
+                             outTable+offsets[stringTableIndex], pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            udata_printError(ds, "ucnv_swapAliases().swapInvChars(charset names) failed\n");
+            return 0;
+        }
+
+        if(ds->inCharset==ds->outCharset) {
+            /* no need to sort, just swap all 16-bit values together */
+            ds->swapArray16(ds,
+                            inTable+offsets[converterListIndex],
+                            2*(int32_t)(offsets[stringTableIndex]-offsets[converterListIndex]),
+                            outTable+offsets[converterListIndex],
+                            pErrorCode);
+        } else {
+            /* allocate the temporary table for sorting */
+            count=toc[aliasListIndex];
+
+            tempTable.chars=(const char *)(outTable+offsets[stringTableIndex]); /* sort by outCharset */
+
+            if(count<=STACK_ROW_CAPACITY) {
+                tempTable.rows=rows;
+                tempTable.resort=resort;
+            } else {
+                tempTable.rows=(TempRow *)uprv_malloc(count*sizeof(TempRow)+count*2);
+                if(tempTable.rows==NULL) {
+                    udata_printError(ds, "ucnv_swapAliases(): unable to allocate memory for sorting tables (max length: %u)\n",
+                                     count);
+                    *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return 0;
+                }
+                tempTable.resort=(uint16_t *)(tempTable.rows+count);
+            }
+
+            if(ds->outCharset==U_ASCII_FAMILY) {
+                tempTable.stripForCompare=ucnv_io_stripASCIIForCompare;
+            } else /* U_EBCDIC_FAMILY */ {
+                tempTable.stripForCompare=ucnv_io_stripEBCDICForCompare;
+            }
+
+            /*
+             * Sort unique aliases+mapped names.
+             *
+             * We need to sort the list again by outCharset strings because they
+             * sort differently for different charset families.
+             * First we set up a temporary table with the string indexes and
+             * sorting indexes and sort that.
+             * Then we permutate and copy/swap the actual values.
+             */
+            p=inTable+offsets[aliasListIndex];
+            q=outTable+offsets[aliasListIndex];
+
+            p2=inTable+offsets[untaggedConvArrayIndex];
+            q2=outTable+offsets[untaggedConvArrayIndex];
+
+            for(i=0; i<count; ++i) {
+                tempTable.rows[i].strIndex=ds->readUInt16(p[i]);
+                tempTable.rows[i].sortIndex=(uint16_t)i;
+            }
+
+            uprv_sortArray(tempTable.rows, (int32_t)count, sizeof(TempRow),
+                           io_compareRows, &tempTable,
+                           FALSE, pErrorCode);
+
+            if(U_SUCCESS(*pErrorCode)) {
+                /* copy/swap/permutate items */
+                if(p!=q) {
+                    for(i=0; i<count; ++i) {
+                        oldIndex=tempTable.rows[i].sortIndex;
+                        ds->swapArray16(ds, p+oldIndex, 2, q+i, pErrorCode);
+                        ds->swapArray16(ds, p2+oldIndex, 2, q2+i, pErrorCode);
+                    }
+                } else {
+                    /*
+                     * If we swap in-place, then the permutation must use another
+                     * temporary array (tempTable.resort)
+                     * before the results are copied to the outBundle.
+                     */
+                    uint16_t *r=tempTable.resort;
+
+                    for(i=0; i<count; ++i) {
+                        oldIndex=tempTable.rows[i].sortIndex;
+                        ds->swapArray16(ds, p+oldIndex, 2, r+i, pErrorCode);
+                    }
+                    uprv_memcpy(q, r, 2*(size_t)count);
+
+                    for(i=0; i<count; ++i) {
+                        oldIndex=tempTable.rows[i].sortIndex;
+                        ds->swapArray16(ds, p2+oldIndex, 2, r+i, pErrorCode);
+                    }
+                    uprv_memcpy(q2, r, 2*(size_t)count);
+                }
+            }
+
+            if(tempTable.rows!=rows) {
+                uprv_free(tempTable.rows);
+            }
+
+            if(U_FAILURE(*pErrorCode)) {
+                udata_printError(ds, "ucnv_swapAliases().uprv_sortArray(%u items) failed\n",
+                                 count);
+                return 0;
+            }
+
+            /* swap remaining 16-bit values */
+            ds->swapArray16(ds,
+                            inTable+offsets[converterListIndex],
+                            2*(int32_t)(offsets[aliasListIndex]-offsets[converterListIndex]),
+                            outTable+offsets[converterListIndex],
+                            pErrorCode);
+            ds->swapArray16(ds,
+                            inTable+offsets[taggedAliasArrayIndex],
+                            2*(int32_t)(offsets[stringTableIndex]-offsets[taggedAliasArrayIndex]),
+                            outTable+offsets[taggedAliasArrayIndex],
+                            pErrorCode);
+        }
+    }
+
+    return headerSize+2*(int32_t)topOffset;
+}
+
+#endif
+
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/ucnv_io.h b/thirdparty/icu4c/common/ucnv_io.h
new file mode 100644
index 0000000000..8f2d7b5a02
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_io.h
@@ -0,0 +1,127 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ **********************************************************************
+ *   Copyright (C) 1999-2006, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ **********************************************************************
+ *
+ *
+ *  ucnv_io.h:
+ *  defines  variables and functions pertaining to converter name resolution
+ *  aspect of the conversion code
+ */
+
+#ifndef UCNV_IO_H
+#define UCNV_IO_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "udataswp.h"
+
+#define UCNV_AMBIGUOUS_ALIAS_MAP_BIT 0x8000
+#define UCNV_CONTAINS_OPTION_BIT 0x4000
+#define UCNV_CONVERTER_INDEX_MASK 0xFFF
+#define UCNV_NUM_RESERVED_TAGS 2
+#define UCNV_NUM_HIDDEN_TAGS 1
+
+enum {
+    UCNV_IO_UNNORMALIZED,
+    UCNV_IO_STD_NORMALIZED,
+    UCNV_IO_NORM_TYPE_COUNT
+};
+
+typedef struct {
+    uint16_t stringNormalizationType;
+    uint16_t containsCnvOptionInfo;
+} UConverterAliasOptions;
+
+typedef struct UConverterAlias {
+    const uint16_t *converterList;
+    const uint16_t *tagList;
+    const uint16_t *aliasList;
+    const uint16_t *untaggedConvArray;
+    const uint16_t *taggedAliasArray;
+    const uint16_t *taggedAliasLists;
+    const UConverterAliasOptions *optionTable;
+    const uint16_t *stringTable;
+    const uint16_t *normalizedStringTable;
+
+    uint32_t converterListSize;
+    uint32_t tagListSize;
+    uint32_t aliasListSize;
+    uint32_t untaggedConvArraySize;
+    uint32_t taggedAliasArraySize;
+    uint32_t taggedAliasListsSize;
+    uint32_t optionTableSize;
+    uint32_t stringTableSize;
+    uint32_t normalizedStringTableSize;
+} UConverterAlias;
+
+/**
+ * \var ucnv_io_stripForCompare
+ * Remove the underscores, dashes and spaces from the name, and convert
+ * the name to lower case.
+ * @param dst The destination buffer, which is <= the buffer of name.
+ * @param dst The destination buffer, which is <= the buffer of name.
+ * @see ucnv_compareNames
+ * @return the destination buffer.
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define ucnv_io_stripForCompare ucnv_io_stripASCIIForCompare
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define ucnv_io_stripForCompare ucnv_io_stripEBCDICForCompare
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+
+U_CAPI char * U_CALLCONV
+ucnv_io_stripASCIIForCompare(char *dst, const char *name);
+
+U_CAPI char * U_CALLCONV
+ucnv_io_stripEBCDICForCompare(char *dst, const char *name);
+
+/**
+ * Map a converter alias name to a canonical converter name.
+ * The alias is searched for case-insensitively, the converter name
+ * is returned in mixed-case.
+ * Returns NULL if the alias is not found.
+ * @param alias The alias name to be searched.
+ * @param containsOption A return value stating whether the returned converter name contains an option (a comma)
+ * @param pErrorCode The error code
+ * @return the converter name in mixed-case, return NULL if the alias is not found.
+ */
+U_CAPI const char *
+ucnv_io_getConverterName(const char *alias, UBool *containsOption, UErrorCode *pErrorCode);
+
+/**
+ * Return the number of all known converter names (no aliases).
+ * @param pErrorCode The error code
+ * @return the number of all aliases
+ */
+U_CAPI uint16_t
+ucnv_io_countKnownConverters(UErrorCode *pErrorCode);
+
+/**
+ * Swap an ICU converter alias table. See implementation for details.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_swapAliases(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode);
+
+#endif
+
+#endif /* _UCNV_IO */
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/ucnv_lmb.cpp b/thirdparty/icu4c/common/ucnv_lmb.cpp
new file mode 100644
index 0000000000..168392837b
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_lmb.cpp
@@ -0,0 +1,1388 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_lmb.cpp
+*   encoding:   UTF-8
+*   tab size:   4 (not used)
+*   indentation:4
+*
+*   created on: 2000feb09
+*   created by: Brendan Murray
+*   extensively hacked up by: Jim Snyder-Grant
+*
+* Modification History:
+* 
+*   Date        Name             Description
+* 
+*   06/20/2000  helena           OS/400 port changes; mostly typecast.
+*   06/27/2000  Jim Snyder-Grant Deal with partial characters and small buffers.
+*                                Add comments to document LMBCS format and implementation
+*                                restructured order & breakdown of functions
+*   06/28/2000  helena           Major rewrite for the callback API changes.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv_err.h"
+#include "unicode/ucnv.h"
+#include "unicode/uset.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "ucnv_imp.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+
+#ifdef EBCDIC_RTL
+    #include "ascii_a.h"
+#endif
+
+/*
+  LMBCS
+
+  (Lotus Multi-Byte Character Set)
+
+  LMBCS was invented in the late 1980's and is primarily used in Lotus Notes 
+  databases and in Lotus 1-2-3 files. Programmers who work with the APIs 
+  into these products will sometimes need to deal with strings in this format.
+
+  The code in this file provides an implementation for an ICU converter of 
+  LMBCS to and from Unicode. 
+
+  Since the LMBCS character set is only sparsely documented in existing 
+  printed or online material, we have added  extensive annotation to this 
+  file to serve as a guide to understanding LMBCS. 
+
+  LMBCS was originally designed with these four sometimes-competing design goals:
+
+  -Provide encodings for the characters in 12 existing national standards
+   (plus a few other characters)
+  -Minimal memory footprint
+  -Maximal speed of conversion into the existing national character sets
+  -No need to track a changing state as you interpret a string.
+
+
+  All of the national character sets LMBCS was trying to encode are 'ANSI'
+  based, in that the bytes from 0x20 - 0x7F are almost exactly the 
+  same common Latin unaccented characters and symbols in all character sets. 
+
+  So, in order to help meet the speed & memory design goals, the common ANSI 
+  bytes from 0x20-0x7F are represented by the same single-byte values in LMBCS. 
+
+  The general LMBCS code unit is from 1-3 bytes. We can describe the 3 bytes as
+  follows:
+
+  [G] D1 [D2]
+
+  That is, a sometimes-optional 'group' byte, followed by 1 and sometimes 2
+  data bytes. The maximum size of a LMBCS chjaracter is 3 bytes:
+*/
+#define ULMBCS_CHARSIZE_MAX      3
+/*
+  The single-byte values from 0x20 to 0x7F are examples of single D1 bytes.
+  We often have to figure out if byte values are below or above this, so we 
+  use the ANSI nomenclature 'C0' and 'C1' to refer to the range of control 
+  characters just above & below the common lower-ANSI  range */
+#define ULMBCS_C0END           0x1F   
+#define ULMBCS_C1START         0x80   
+/*
+  Since LMBCS is always dealing in byte units. we create a local type here for 
+  dealing with these units of LMBCS code units:
+
+*/  
+typedef uint8_t ulmbcs_byte_t;
+
+/* 
+   Most of the values less than 0x20 are reserved in LMBCS to announce 
+   which national  character standard is being used for the 'D' bytes. 
+   In the comments we show the common name and the IBM character-set ID
+   for these character-set announcers:
+*/
+
+#define ULMBCS_GRP_L1         0x01   /* Latin-1    :ibm-850  */
+#define ULMBCS_GRP_GR         0x02   /* Greek      :ibm-851  */
+#define ULMBCS_GRP_HE         0x03   /* Hebrew     :ibm-1255 */
+#define ULMBCS_GRP_AR         0x04   /* Arabic     :ibm-1256 */
+#define ULMBCS_GRP_RU         0x05   /* Cyrillic   :ibm-1251 */
+#define ULMBCS_GRP_L2         0x06   /* Latin-2    :ibm-852  */
+#define ULMBCS_GRP_TR         0x08   /* Turkish    :ibm-1254 */
+#define ULMBCS_GRP_TH         0x0B   /* Thai       :ibm-874  */
+#define ULMBCS_GRP_JA         0x10   /* Japanese   :ibm-943  */
+#define ULMBCS_GRP_KO         0x11   /* Korean     :ibm-1261 */
+#define ULMBCS_GRP_TW         0x12   /* Chinese SC :ibm-950  */
+#define ULMBCS_GRP_CN         0x13   /* Chinese TC :ibm-1386 */
+
+/*
+   So, the beginning of understanding LMBCS is that IF the first byte of a LMBCS 
+   character is one of those 12 values, you can interpret the remaining bytes of 
+   that character as coming from one of those character sets. Since the lower 
+   ANSI bytes already are represented in single bytes, using one of the character 
+   set announcers is used to announce a character that starts with a byte of 
+   0x80 or greater.
+
+   The character sets are  arranged so that the single byte sets all appear 
+   before the multi-byte character sets. When we need to tell whether a 
+   group byte is for a single byte char set or not we use this define: */
+
+#define ULMBCS_DOUBLEOPTGROUP_START  0x10   
+
+/* 
+However, to fully understand LMBCS, you must also understand a series of 
+exceptions & optimizations made in service of the design goals. 
+
+First, those of you who are character set mavens may have noticed that
+the 'double-byte' character sets are actually multi-byte character sets 
+that can have 1 or two bytes, even in the upper-ascii range. To force
+each group byte to introduce a fixed-width encoding (to make it faster to 
+count characters), we use a convention of doubling up on the group byte 
+to introduce any single-byte character > 0x80 in an otherwise double-byte
+character set. So, for example, the LMBCS sequence x10 x10 xAE is the 
+same as '0xAE' in the Japanese code page 943.
+
+Next, you will notice that the list of group bytes has some gaps. 
+These are used in various ways.
+
+We reserve a few special single byte values for common control 
+characters. These are in the same place as their ANSI eqivalents for speed.
+*/
+                     
+#define ULMBCS_HT    0x09   /* Fixed control char - Horizontal Tab */
+#define ULMBCS_LF    0x0A   /* Fixed control char - Line Feed */
+#define ULMBCS_CR    0x0D   /* Fixed control char - Carriage Return */
+
+/* Then, 1-2-3 reserved a special single-byte character to put at the 
+beginning of internal 'system' range names: */
+
+#define ULMBCS_123SYSTEMRANGE  0x19   
+
+/* Then we needed a place to put all the other ansi control characters 
+that must be moved to different values because LMBCS reserves those 
+values for other purposes. To represent the control characters, we start 
+with a first byte of 0xF & add the control chaarcter value as the 
+second byte */
+#define ULMBCS_GRP_CTRL       0x0F   
+
+/* For the C0 controls (less than 0x20), we add 0x20 to preserve the 
+useful doctrine that any byte less than 0x20 in a LMBCS char must be 
+the first byte of a character:*/
+#define ULMBCS_CTRLOFFSET      0x20   
+
+/* 
+Where to put the characters that aren't part of any of the 12 national 
+character sets? The first thing that was done, in the earlier years of 
+LMBCS, was to use up the spaces of the form
+
+  [G] D1, 
+  
+ where  'G' was one of the single-byte character groups, and
+ D1 was less than 0x80. These sequences are gathered together 
+ into a Lotus-invented doublebyte character set to represent a 
+ lot of stray values. Internally, in this implementation, we track this 
+ as group '0', as a place to tuck this exceptions list.*/
+
+#define ULMBCS_GRP_EXCEPT     0x00    
+/*
+ Finally, as the durability and usefulness of UNICODE became clear, 
+ LOTUS added a new group 0x14 to hold Unicode values not otherwise 
+ represented in LMBCS: */
+#define ULMBCS_GRP_UNICODE    0x14   
+/* The two bytes appearing after a 0x14 are intrepreted as UFT-16 BE
+(Big-Endian) characters. The exception comes when the UTF16 
+representation would have a zero as the second byte. In that case,
+'F6' is used in its place, and the bytes are swapped. (This prevents 
+LMBCS from encoding any Unicode values of the form U+F6xx, but that's OK:
+0xF6xx is in the middle of the Private Use Area.)*/
+#define ULMBCS_UNICOMPATZERO   0xF6   
+
+/* It is also useful in our code to have a constant for the size of 
+a LMBCS char that holds a literal Unicode value */
+#define ULMBCS_UNICODE_SIZE      3    
+
+/* 
+To squish the LMBCS representations down even further, and to make 
+translations even faster,sometimes the optimization group byte can be dropped 
+from a LMBCS character. This is decided on a process-by-process basis. The 
+group byte that is dropped is called the 'optimization group'.
+
+For Notes, the optimzation group is always 0x1.*/
+#define ULMBCS_DEFAULTOPTGROUP 0x1    
+/* For 1-2-3 files, the optimzation group is stored in the header of the 1-2-3 
+file. 
+
+ In any case, when using ICU, you either pass in the 
+optimization group as part of the name of the converter (LMBCS-1, LMBCS-2, 
+etc.). Using plain 'LMBCS' as the name of the converter will give you 
+LMBCS-1.
+
+
+*** Implementation strategy ***
+
+
+Because of the extensive use of other character sets, the LMBCS converter
+keeps a mapping between optimization groups and IBM character sets, so that
+ICU converters can be created and used as needed. */
+
+/* As you can see, even though any byte below 0x20 could be an optimization 
+byte, only those at 0x13 or below can map to an actual converter. To limit
+some loops and searches, we define a value for that last group converter:*/
+
+#define ULMBCS_GRP_LAST       0x13   /* last LMBCS group that has a converter */
+
+static const char * const OptGroupByteToCPName[ULMBCS_GRP_LAST + 1] = {
+   /* 0x0000 */ "lmb-excp", /* internal home for the LOTUS exceptions list */
+   /* 0x0001 */ "ibm-850",
+   /* 0x0002 */ "ibm-851",
+   /* 0x0003 */ "windows-1255",
+   /* 0x0004 */ "windows-1256",
+   /* 0x0005 */ "windows-1251",
+   /* 0x0006 */ "ibm-852",
+   /* 0x0007 */ NULL,      /* Unused */
+   /* 0x0008 */ "windows-1254",
+   /* 0x0009 */ NULL,      /* Control char HT */
+   /* 0x000A */ NULL,      /* Control char LF */
+   /* 0x000B */ "windows-874",
+   /* 0x000C */ NULL,      /* Unused */
+   /* 0x000D */ NULL,      /* Control char CR */
+   /* 0x000E */ NULL,      /* Unused */
+   /* 0x000F */ NULL,      /* Control chars: 0x0F20 + C0/C1 character: algorithmic */
+   /* 0x0010 */ "windows-932",
+   /* 0x0011 */ "windows-949",
+   /* 0x0012 */ "windows-950",
+   /* 0x0013 */ "windows-936"
+
+   /* The rest are null, including the 0x0014 Unicode compatibility region
+   and 0x0019, the 1-2-3 system range control char */      
+};
+
+
+/* That's approximately all the data that's needed for translating 
+  LMBCS to Unicode. 
+
+
+However, to translate Unicode to LMBCS, we need some more support.
+
+That's because there are often more than one possible mappings from a Unicode
+code point back into LMBCS. The first thing we do is look up into a table
+to figure out if there are more than one possible mappings. This table,
+arranged by Unicode values (including ranges) either lists which group 
+to use, or says that it could go into one or more of the SBCS sets, or
+into one or more of the DBCS sets.  (If the character exists in both DBCS & 
+SBCS, the table will place it in the SBCS sets, to make the LMBCS code point 
+length as small as possible. Here's the two special markers we use to indicate
+ambiguous mappings: */
+
+#define ULMBCS_AMBIGUOUS_SBCS   0x80   /* could fit in more than one 
+                                          LMBCS sbcs native encoding 
+                                          (example: most accented latin) */
+#define ULMBCS_AMBIGUOUS_MBCS   0x81   /* could fit in more than one 
+                                          LMBCS mbcs native encoding 
+                                          (example: Unihan) */
+#define ULMBCS_AMBIGUOUS_ALL   0x82
+/* And here's a simple way to see if a group falls in an appropriate range */
+#define ULMBCS_AMBIGUOUS_MATCH(agroup, xgroup) \
+                  ((((agroup) == ULMBCS_AMBIGUOUS_SBCS) && \
+                  (xgroup) < ULMBCS_DOUBLEOPTGROUP_START) || \
+                  (((agroup) == ULMBCS_AMBIGUOUS_MBCS) && \
+                  (xgroup) >= ULMBCS_DOUBLEOPTGROUP_START)) || \
+                  ((agroup) == ULMBCS_AMBIGUOUS_ALL)
+
+
+/* The table & some code to use it: */
+
+
+static const struct _UniLMBCSGrpMap  
+{
+   const UChar uniStartRange;
+   const UChar uniEndRange;
+   const ulmbcs_byte_t  GrpType;
+} UniLMBCSGrpMap[]
+=
+{
+
+    {0x0001, 0x001F,  ULMBCS_GRP_CTRL},
+    {0x0080, 0x009F,  ULMBCS_GRP_CTRL},
+    {0x00A0, 0x00A6,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00A7, 0x00A8,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00A9, 0x00AF,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00B0, 0x00B1,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00B2, 0x00B3,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00B4, 0x00B4,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00B5, 0x00B5,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00B6, 0x00B6,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00B7, 0x00D6,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00D7, 0x00D7,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00D8, 0x00F6,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x00F7, 0x00F7,  ULMBCS_AMBIGUOUS_ALL},
+    {0x00F8, 0x01CD,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x01CE, 0x01CE,  ULMBCS_GRP_TW },
+    {0x01CF, 0x02B9,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x02BA, 0x02BA,  ULMBCS_GRP_CN},
+    {0x02BC, 0x02C8,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x02C9, 0x02D0,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x02D8, 0x02DD,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x0384, 0x0390,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x0391, 0x03A9,  ULMBCS_AMBIGUOUS_ALL},
+    {0x03AA, 0x03B0,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x03B1, 0x03C9,  ULMBCS_AMBIGUOUS_ALL},
+    {0x03CA, 0x03CE,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x0400, 0x0400,  ULMBCS_GRP_RU},
+    {0x0401, 0x0401,  ULMBCS_AMBIGUOUS_ALL},
+    {0x0402, 0x040F,  ULMBCS_GRP_RU},
+    {0x0410, 0x0431,  ULMBCS_AMBIGUOUS_ALL},
+    {0x0432, 0x044E,  ULMBCS_GRP_RU},
+    {0x044F, 0x044F,  ULMBCS_AMBIGUOUS_ALL},
+    {0x0450, 0x0491,  ULMBCS_GRP_RU},
+    {0x05B0, 0x05F2,  ULMBCS_GRP_HE},
+    {0x060C, 0x06AF,  ULMBCS_GRP_AR},
+    {0x0E01, 0x0E5B,  ULMBCS_GRP_TH},
+    {0x200C, 0x200F,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2010, 0x2010,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2013, 0x2014,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2015, 0x2015,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2016, 0x2016,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2017, 0x2017,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2018, 0x2019,  ULMBCS_AMBIGUOUS_ALL},
+    {0x201A, 0x201B,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x201C, 0x201D,  ULMBCS_AMBIGUOUS_ALL},
+    {0x201E, 0x201F,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2020, 0x2021,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2022, 0x2024,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2025, 0x2025,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2026, 0x2026,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2027, 0x2027,  ULMBCS_GRP_TW},
+    {0x2030, 0x2030,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2031, 0x2031,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2032, 0x2033,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2035, 0x2035,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2039, 0x203A,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x203B, 0x203B,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x203C, 0x203C,  ULMBCS_GRP_EXCEPT},
+    {0x2074, 0x2074,  ULMBCS_GRP_KO},
+    {0x207F, 0x207F,  ULMBCS_GRP_EXCEPT},
+    {0x2081, 0x2084,  ULMBCS_GRP_KO},
+    {0x20A4, 0x20AC,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2103, 0x2109,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2111, 0x2120,  ULMBCS_AMBIGUOUS_SBCS},
+    /*zhujin: upgrade, for regressiont test, spr HKIA4YHTSU*/
+    {0x2121, 0x2121,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2122, 0x2126,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x212B, 0x212B,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2135, 0x2135,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2153, 0x2154,  ULMBCS_GRP_KO},
+    {0x215B, 0x215E,  ULMBCS_GRP_EXCEPT},
+    {0x2160, 0x2179,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2190, 0x2193,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2194, 0x2195,  ULMBCS_GRP_EXCEPT},
+    {0x2196, 0x2199,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x21A8, 0x21A8,  ULMBCS_GRP_EXCEPT},
+    {0x21B8, 0x21B9,  ULMBCS_GRP_CN},
+    {0x21D0, 0x21D1,  ULMBCS_GRP_EXCEPT},
+    {0x21D2, 0x21D2,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x21D3, 0x21D3,  ULMBCS_GRP_EXCEPT},
+    {0x21D4, 0x21D4,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x21D5, 0x21D5,  ULMBCS_GRP_EXCEPT},
+    {0x21E7, 0x21E7,  ULMBCS_GRP_CN},
+    {0x2200, 0x2200,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2201, 0x2201,  ULMBCS_GRP_EXCEPT},
+    {0x2202, 0x2202,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2203, 0x2203,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2204, 0x2206,  ULMBCS_GRP_EXCEPT},
+    {0x2207, 0x2208,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2209, 0x220A,  ULMBCS_GRP_EXCEPT},
+    {0x220B, 0x220B,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x220F, 0x2215,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2219, 0x2219,  ULMBCS_GRP_EXCEPT},
+    {0x221A, 0x221A,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x221B, 0x221C,  ULMBCS_GRP_EXCEPT},
+    {0x221D, 0x221E,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x221F, 0x221F,  ULMBCS_GRP_EXCEPT},
+    {0x2220, 0x2220,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2223, 0x222A,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x222B, 0x223D,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2245, 0x2248,  ULMBCS_GRP_EXCEPT},
+    {0x224C, 0x224C,  ULMBCS_GRP_TW},
+    {0x2252, 0x2252,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2260, 0x2261,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2262, 0x2265,  ULMBCS_GRP_EXCEPT},
+    {0x2266, 0x226F,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2282, 0x2283,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2284, 0x2285,  ULMBCS_GRP_EXCEPT},
+    {0x2286, 0x2287,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2288, 0x2297,  ULMBCS_GRP_EXCEPT},
+    {0x2299, 0x22BF,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x22C0, 0x22C0,  ULMBCS_GRP_EXCEPT},
+    {0x2310, 0x2310,  ULMBCS_GRP_EXCEPT},
+    {0x2312, 0x2312,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2318, 0x2321,  ULMBCS_GRP_EXCEPT},
+    {0x2318, 0x2321,  ULMBCS_GRP_CN},
+    {0x2460, 0x24E9,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2500, 0x2500,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2501, 0x2501,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2502, 0x2502,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2503, 0x2503,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x2504, 0x2505,  ULMBCS_GRP_TW},
+    {0x2506, 0x2665,  ULMBCS_AMBIGUOUS_ALL},
+    {0x2666, 0x2666,  ULMBCS_GRP_EXCEPT},
+    {0x2667, 0x2669,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x266A, 0x266A,  ULMBCS_AMBIGUOUS_ALL},
+    {0x266B, 0x266C,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x266D, 0x266D,  ULMBCS_AMBIGUOUS_MBCS},
+    {0x266E, 0x266E,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x266F, 0x266F,  ULMBCS_GRP_JA},
+    {0x2670, 0x2E7F,  ULMBCS_AMBIGUOUS_SBCS},
+    {0x2E80, 0xF861,  ULMBCS_AMBIGUOUS_MBCS},
+    {0xF862, 0xF8FF,  ULMBCS_GRP_EXCEPT},
+    {0xF900, 0xFA2D,  ULMBCS_AMBIGUOUS_MBCS},
+    {0xFB00, 0xFEFF,  ULMBCS_AMBIGUOUS_SBCS},
+    {0xFF01, 0xFFEE,  ULMBCS_AMBIGUOUS_MBCS},
+    {0xFFFF, 0xFFFF,  ULMBCS_GRP_UNICODE}
+};
+   
+static ulmbcs_byte_t 
+FindLMBCSUniRange(UChar uniChar)
+{
+   const struct _UniLMBCSGrpMap * pTable = UniLMBCSGrpMap;
+
+   while (uniChar > pTable->uniEndRange) 
+   {
+      pTable++;
+   }
+
+   if (uniChar >= pTable->uniStartRange) 
+   {
+      return pTable->GrpType;
+   }
+   return ULMBCS_GRP_UNICODE;
+}
+
+/* 
+We also ask the creator of a converter to send in a preferred locale 
+that we can use in resolving ambiguous mappings. They send the locale
+in as a string, and we map it, if possible, to one of the 
+LMBCS groups. We use this table, and the associated code, to 
+do the lookup: */
+
+/**************************************************
+  This table maps locale ID's to LMBCS opt groups.
+  The default return is group 0x01. Note that for
+  performance reasons, the table is sorted in
+  increasing alphabetic order, with the notable
+  exception of zhTW. This is to force the check
+  for Traditonal Chinese before dropping back to
+  Simplified.
+
+  Note too that the Latin-1 groups have been
+  commented out because it's the default, and
+  this shortens the table, allowing a serial
+  search to go quickly.
+ *************************************************/
+
+static const struct _LocaleLMBCSGrpMap
+{
+   const char    *LocaleID;
+   const ulmbcs_byte_t OptGroup;
+} LocaleLMBCSGrpMap[] =
+{
+    {"ar", ULMBCS_GRP_AR},
+    {"be", ULMBCS_GRP_RU},
+    {"bg", ULMBCS_GRP_L2},
+   /* {"ca", ULMBCS_GRP_L1}, */
+    {"cs", ULMBCS_GRP_L2},
+   /* {"da", ULMBCS_GRP_L1}, */
+   /* {"de", ULMBCS_GRP_L1}, */
+    {"el", ULMBCS_GRP_GR},
+   /* {"en", ULMBCS_GRP_L1}, */
+   /* {"es", ULMBCS_GRP_L1}, */
+   /* {"et", ULMBCS_GRP_L1}, */
+   /* {"fi", ULMBCS_GRP_L1}, */
+   /* {"fr", ULMBCS_GRP_L1}, */
+    {"he", ULMBCS_GRP_HE},
+    {"hu", ULMBCS_GRP_L2},
+   /* {"is", ULMBCS_GRP_L1}, */
+   /* {"it", ULMBCS_GRP_L1}, */
+    {"iw", ULMBCS_GRP_HE},
+    {"ja", ULMBCS_GRP_JA},
+    {"ko", ULMBCS_GRP_KO},
+   /* {"lt", ULMBCS_GRP_L1}, */
+   /* {"lv", ULMBCS_GRP_L1}, */
+    {"mk", ULMBCS_GRP_RU},
+   /* {"nl", ULMBCS_GRP_L1}, */
+   /* {"no", ULMBCS_GRP_L1}, */
+    {"pl", ULMBCS_GRP_L2},
+   /* {"pt", ULMBCS_GRP_L1}, */
+    {"ro", ULMBCS_GRP_L2},
+    {"ru", ULMBCS_GRP_RU},
+    {"sh", ULMBCS_GRP_L2},
+    {"sk", ULMBCS_GRP_L2},
+    {"sl", ULMBCS_GRP_L2},
+    {"sq", ULMBCS_GRP_L2},
+    {"sr", ULMBCS_GRP_RU},
+   /* {"sv", ULMBCS_GRP_L1}, */
+    {"th", ULMBCS_GRP_TH},
+    {"tr", ULMBCS_GRP_TR},
+    {"uk", ULMBCS_GRP_RU},
+   /* {"vi", ULMBCS_GRP_L1}, */
+    {"zhTW", ULMBCS_GRP_TW},
+    {"zh", ULMBCS_GRP_CN},
+    {NULL, ULMBCS_GRP_L1}
+};
+
+
+static ulmbcs_byte_t 
+FindLMBCSLocale(const char *LocaleID)
+{
+   const struct _LocaleLMBCSGrpMap *pTable = LocaleLMBCSGrpMap;
+
+   if ((!LocaleID) || (!*LocaleID)) 
+   {
+      return 0;
+   }
+
+   while (pTable->LocaleID)
+   {
+      if (*pTable->LocaleID == *LocaleID) /* Check only first char for speed */
+      {
+         /* First char matches - check whole name, for entry-length */
+         if (uprv_strncmp(pTable->LocaleID, LocaleID, strlen(pTable->LocaleID)) == 0)
+            return pTable->OptGroup;
+      }
+      else
+      if (*pTable->LocaleID > *LocaleID) /* Sorted alphabetically - exit */
+         break;
+      pTable++;
+   }
+   return ULMBCS_GRP_L1;
+}
+
+
+/* 
+  Before we get to the main body of code, here's how we hook up to the rest 
+  of ICU. ICU converters are required to define a structure that includes 
+  some function pointers, and some common data, in the style of a C++
+  vtable. There is also room in there for converter-specific data. LMBCS
+  uses that converter-specific data to keep track of the 12 subconverters
+  we use, the optimization group, and the group (if any) that matches the 
+  locale. We have one structure instantiated for each of the 12 possible
+  optimization groups. To avoid typos & to avoid boring the reader, we 
+  put the declarations of these structures and functions into macros. To see 
+  the definitions of these structures, see unicode\ucnv_bld.h
+*/
+
+typedef struct
+  {
+    UConverterSharedData *OptGrpConverter[ULMBCS_GRP_LAST+1];    /* Converter per Opt. grp. */
+    uint8_t    OptGroup;                  /* default Opt. grp. for this LMBCS session */
+    uint8_t    localeConverterIndex;      /* reasonable locale match for index */
+  }
+UConverterDataLMBCS;
+
+U_CDECL_BEGIN
+static void  U_CALLCONV _LMBCSClose(UConverter * _this);
+U_CDECL_END
+
+#define DECLARE_LMBCS_DATA(n) \
+static const UConverterImpl _LMBCSImpl##n={\
+    UCNV_LMBCS_##n,\
+    NULL,NULL,\
+    _LMBCSOpen##n,\
+    _LMBCSClose,\
+    NULL,\
+    _LMBCSToUnicodeWithOffsets,\
+    _LMBCSToUnicodeWithOffsets,\
+    _LMBCSFromUnicode,\
+    _LMBCSFromUnicode,\
+    NULL,\
+    NULL,\
+    NULL,\
+    NULL,\
+    _LMBCSSafeClone,\
+    ucnv_getCompleteUnicodeSet,\
+    NULL,\
+    NULL\
+};\
+static const UConverterStaticData _LMBCSStaticData##n={\
+  sizeof(UConverterStaticData),\
+ "LMBCS-"  #n,\
+    0, UCNV_IBM, UCNV_LMBCS_##n, 1, 3,\
+    { 0x3f, 0, 0, 0 },1,FALSE,FALSE,0,0,{0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0} \
+};\
+const UConverterSharedData _LMBCSData##n= \
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_LMBCSStaticData##n, &_LMBCSImpl##n);
+
+ /* The only function we needed to duplicate 12 times was the 'open'
+function, which will do basically the same thing except set a  different
+optimization group. So, we put the common stuff into a worker function, 
+and set up another macro to stamp out the 12 open functions:*/
+#define DEFINE_LMBCS_OPEN(n) \
+static void U_CALLCONV \
+   _LMBCSOpen##n(UConverter* _this, UConverterLoadArgs* pArgs, UErrorCode* err) \
+{ _LMBCSOpenWorker(_this, pArgs, err, n); }
+
+
+
+/* Here's the open worker & the common close function */
+static void 
+_LMBCSOpenWorker(UConverter*  _this,
+                 UConverterLoadArgs *pArgs,
+                 UErrorCode*  err,
+                 ulmbcs_byte_t OptGroup)
+{
+    UConverterDataLMBCS * extraInfo = (UConverterDataLMBCS*)uprv_malloc (sizeof (UConverterDataLMBCS));
+    _this->extraInfo = extraInfo;
+    if(extraInfo != NULL)
+    {
+        UConverterNamePieces stackPieces;
+        UConverterLoadArgs stackArgs= UCNV_LOAD_ARGS_INITIALIZER;
+        ulmbcs_byte_t i;
+
+        uprv_memset(extraInfo, 0, sizeof(UConverterDataLMBCS));
+
+        stackArgs.onlyTestIsLoadable = pArgs->onlyTestIsLoadable;
+
+        for (i=0; i <= ULMBCS_GRP_LAST && U_SUCCESS(*err); i++)         
+        {
+            if(OptGroupByteToCPName[i] != NULL) {
+                extraInfo->OptGrpConverter[i] = ucnv_loadSharedData(OptGroupByteToCPName[i], &stackPieces, &stackArgs, err);
+            }
+        }
+
+        if(U_FAILURE(*err) || pArgs->onlyTestIsLoadable) {
+            _LMBCSClose(_this);
+            return;
+        }
+        extraInfo->OptGroup = OptGroup;
+        extraInfo->localeConverterIndex = FindLMBCSLocale(pArgs->locale);
+    }
+    else
+    {
+        *err = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+U_CDECL_BEGIN
+static void  U_CALLCONV
+_LMBCSClose(UConverter *   _this) 
+{
+    if (_this->extraInfo != NULL)
+    {
+        ulmbcs_byte_t Ix;
+        UConverterDataLMBCS * extraInfo = (UConverterDataLMBCS *) _this->extraInfo;
+
+        for (Ix=0; Ix <= ULMBCS_GRP_LAST; Ix++)
+        {
+           if (extraInfo->OptGrpConverter[Ix] != NULL)
+              ucnv_unloadSharedDataIfReady(extraInfo->OptGrpConverter[Ix]);
+        }
+        if (!_this->isExtraLocal) {
+            uprv_free (_this->extraInfo);
+            _this->extraInfo = NULL;
+        }
+    }
+}
+
+typedef struct LMBCSClone {
+    UConverter cnv;
+    UConverterDataLMBCS lmbcs;
+} LMBCSClone;
+
+static UConverter *  U_CALLCONV
+_LMBCSSafeClone(const UConverter *cnv, 
+                void *stackBuffer, 
+                int32_t *pBufferSize, 
+                UErrorCode *status) {
+    (void)status;
+    LMBCSClone *newLMBCS;
+    UConverterDataLMBCS *extraInfo;
+    int32_t i;
+
+    if(*pBufferSize<=0) {
+        *pBufferSize=(int32_t)sizeof(LMBCSClone);
+        return NULL;
+    }
+
+    extraInfo=(UConverterDataLMBCS *)cnv->extraInfo;
+    newLMBCS=(LMBCSClone *)stackBuffer;
+
+    /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
+
+    uprv_memcpy(&newLMBCS->lmbcs, extraInfo, sizeof(UConverterDataLMBCS));
+
+    /* share the subconverters */
+    for(i = 0; i <= ULMBCS_GRP_LAST; ++i) {
+        if(extraInfo->OptGrpConverter[i] != NULL) {
+            ucnv_incrementRefCount(extraInfo->OptGrpConverter[i]);
+        }
+    }
+
+    newLMBCS->cnv.extraInfo = &newLMBCS->lmbcs;
+    newLMBCS->cnv.isExtraLocal = TRUE;
+    return &newLMBCS->cnv;
+}
+
+/*
+ * There used to be a _LMBCSGetUnicodeSet() function here (up to svn revision 20117)
+ * which added all code points except for U+F6xx
+ * because those cannot be represented in the Unicode group.
+ * However, it turns out that windows-950 has roundtrips for all of U+F6xx
+ * which means that LMBCS can convert all Unicode code points after all.
+ * We now simply use ucnv_getCompleteUnicodeSet().
+ *
+ * This may need to be looked at again as Lotus uses _LMBCSGetUnicodeSet(). (091216)
+ */
+
+/* 
+   Here's the basic helper function that we use when converting from
+   Unicode to LMBCS, and we suspect that a Unicode character will fit into 
+   one of the 12 groups. The return value is the number of bytes written 
+   starting at pStartLMBCS (if any).
+*/
+
+static size_t
+LMBCSConversionWorker (
+   UConverterDataLMBCS * extraInfo,    /* subconverters, opt & locale groups */
+   ulmbcs_byte_t group,                /* The group to try */
+   ulmbcs_byte_t  * pStartLMBCS,              /* where to put the results */
+   UChar * pUniChar,                   /* The input unicode character */
+   ulmbcs_byte_t * lastConverterIndex, /* output: track last successful group used */
+   UBool * groups_tried                /* output: track any unsuccessful groups */
+)   
+{
+   ulmbcs_byte_t  * pLMBCS = pStartLMBCS;
+   UConverterSharedData * xcnv = extraInfo->OptGrpConverter[group];
+
+   int bytesConverted;
+   uint32_t value;
+   ulmbcs_byte_t firstByte;
+
+   U_ASSERT(xcnv);
+   U_ASSERT(group<ULMBCS_GRP_UNICODE);
+
+   bytesConverted = ucnv_MBCSFromUChar32(xcnv, *pUniChar, &value, FALSE);
+
+   /* get the first result byte */
+   if(bytesConverted > 0) {
+      firstByte = (ulmbcs_byte_t)(value >> ((bytesConverted - 1) * 8));
+   } else {
+      /* most common failure mode is an unassigned character */
+      groups_tried[group] = TRUE;
+      return 0;
+   }
+
+   *lastConverterIndex = group;
+
+   /* All initial byte values in lower ascii range should have been caught by now,
+      except with the exception group.
+    */
+   U_ASSERT((firstByte <= ULMBCS_C0END) || (firstByte >= ULMBCS_C1START) || (group == ULMBCS_GRP_EXCEPT));
+   
+   /* use converted data: first write 0, 1 or two group bytes */
+   if (group != ULMBCS_GRP_EXCEPT && extraInfo->OptGroup != group)
+   {
+      *pLMBCS++ = group;
+      if (bytesConverted == 1 && group >= ULMBCS_DOUBLEOPTGROUP_START)
+      {
+         *pLMBCS++ = group;
+      }
+   }
+
+  /* don't emit control chars */
+   if ( bytesConverted == 1 && firstByte < 0x20 )
+      return 0;
+
+
+   /* then move over the converted data */
+   switch(bytesConverted)
+   {
+   case 4:
+      *pLMBCS++ = (ulmbcs_byte_t)(value >> 24);
+      U_FALLTHROUGH;
+   case 3:
+      *pLMBCS++ = (ulmbcs_byte_t)(value >> 16);
+      U_FALLTHROUGH;
+   case 2:
+      *pLMBCS++ = (ulmbcs_byte_t)(value >> 8);
+      U_FALLTHROUGH;
+   case 1:
+      *pLMBCS++ = (ulmbcs_byte_t)value;
+      U_FALLTHROUGH;
+   default:
+      /* will never occur */
+      break;
+   }
+
+   return (pLMBCS - pStartLMBCS);
+}
+
+
+/* This is a much simpler version of above, when we 
+know we are writing LMBCS using the Unicode group
+*/
+static size_t 
+LMBCSConvertUni(ulmbcs_byte_t * pLMBCS, UChar uniChar)  
+{
+     /* encode into LMBCS Unicode range */
+   uint8_t LowCh =   (uint8_t)(uniChar & 0x00FF);
+   uint8_t HighCh  = (uint8_t)(uniChar >> 8);
+
+   *pLMBCS++ = ULMBCS_GRP_UNICODE;
+
+   if (LowCh == 0)
+   {
+      *pLMBCS++ = ULMBCS_UNICOMPATZERO;
+      *pLMBCS++ = HighCh;
+   }
+   else
+   {
+      *pLMBCS++ = HighCh;
+      *pLMBCS++ = LowCh;
+   }
+   return ULMBCS_UNICODE_SIZE;
+}
+
+
+
+/* The main Unicode to LMBCS conversion function */
+static void  U_CALLCONV
+_LMBCSFromUnicode(UConverterFromUnicodeArgs*     args,
+                  UErrorCode*     err)
+{
+   ulmbcs_byte_t lastConverterIndex = 0;
+   UChar uniChar;
+   ulmbcs_byte_t  LMBCS[ULMBCS_CHARSIZE_MAX];
+   ulmbcs_byte_t  * pLMBCS;
+   int32_t bytes_written;
+   UBool groups_tried[ULMBCS_GRP_LAST+1];
+   UConverterDataLMBCS * extraInfo = (UConverterDataLMBCS *) args->converter->extraInfo;
+   int sourceIndex = 0; 
+
+   /* Basic strategy: attempt to fill in local LMBCS 1-char buffer.(LMBCS)
+      If that succeeds, see if it will all fit into the target & copy it over 
+      if it does.
+
+      We try conversions in the following order:
+
+      1. Single-byte ascii & special fixed control chars (&null)
+      2. Look up group in table & try that (could be 
+            A) Unicode group
+            B) control group,
+            C) national encoding, 
+               or ambiguous SBCS or MBCS group (on to step 4...)
+        
+      3. If its ambiguous, try this order:
+         A) The optimization group
+         B) The locale group
+         C) The last group that succeeded with this string.
+         D) every other group that's relevent (single or double)
+         E) If its single-byte ambiguous, try the exceptions group
+
+      4. And as a grand fallback: Unicode
+   */
+
+    /*Fix for SPR#DJOE66JFN3 (Lotus)*/
+    ulmbcs_byte_t OldConverterIndex = 0;
+
+   while (args->source < args->sourceLimit && !U_FAILURE(*err))
+   {
+      /*Fix for SPR#DJOE66JFN3 (Lotus)*/
+      OldConverterIndex = extraInfo->localeConverterIndex;
+
+      if (args->target >= args->targetLimit)
+      {
+         *err = U_BUFFER_OVERFLOW_ERROR;
+         break;
+      }
+      uniChar = *(args->source);
+      bytes_written = 0;
+      pLMBCS = LMBCS;
+
+      /* check cases in rough order of how common they are, for speed */
+
+      /* single byte matches: strategy 1 */
+      /*Fix for SPR#DJOE66JFN3 (Lotus)*/
+      if((uniChar>=0x80) && (uniChar<=0xff)
+      /*Fix for SPR#JUYA6XAERU and TSAO7GL5NK (Lotus)*/ &&(uniChar!=0xB1) &&(uniChar!=0xD7) &&(uniChar!=0xF7)
+        &&(uniChar!=0xB0) &&(uniChar!=0xB4) &&(uniChar!=0xB6) &&(uniChar!=0xA7) &&(uniChar!=0xA8))
+      {
+            extraInfo->localeConverterIndex = ULMBCS_GRP_L1;
+      }
+      if (((uniChar > ULMBCS_C0END) && (uniChar < ULMBCS_C1START)) ||
+          uniChar == 0 || uniChar == ULMBCS_HT || uniChar == ULMBCS_CR || 
+          uniChar == ULMBCS_LF || uniChar == ULMBCS_123SYSTEMRANGE 
+          )
+      {
+         *pLMBCS++ = (ulmbcs_byte_t ) uniChar;
+         bytes_written = 1;
+      }
+
+
+      if (!bytes_written) 
+      {
+         /* Check by UNICODE range (Strategy 2) */
+         ulmbcs_byte_t group = FindLMBCSUniRange(uniChar);
+         
+         if (group == ULMBCS_GRP_UNICODE)  /* (Strategy 2A) */
+         {
+            pLMBCS += LMBCSConvertUni(pLMBCS,uniChar);
+            
+            bytes_written = (int32_t)(pLMBCS - LMBCS);
+         }
+         else if (group == ULMBCS_GRP_CTRL)  /* (Strategy 2B) */
+         {
+            /* Handle control characters here */
+            if (uniChar <= ULMBCS_C0END)
+            {
+               *pLMBCS++ = ULMBCS_GRP_CTRL;
+               *pLMBCS++ = (ulmbcs_byte_t)(ULMBCS_CTRLOFFSET + uniChar);
+            }
+            else if (uniChar >= ULMBCS_C1START && uniChar <= ULMBCS_C1START + ULMBCS_CTRLOFFSET)
+            {
+               *pLMBCS++ = ULMBCS_GRP_CTRL;
+               *pLMBCS++ = (ulmbcs_byte_t ) (uniChar & 0x00FF);
+            }
+            bytes_written = (int32_t)(pLMBCS - LMBCS);
+         }
+         else if (group < ULMBCS_GRP_UNICODE)  /* (Strategy 2C) */
+         {
+            /* a specific converter has been identified - use it */
+            bytes_written = (int32_t)LMBCSConversionWorker (
+                              extraInfo, group, pLMBCS, &uniChar, 
+                              &lastConverterIndex, groups_tried);
+         }
+         if (!bytes_written)    /* the ambiguous group cases  (Strategy 3) */
+         {
+            uprv_memset(groups_tried, 0, sizeof(groups_tried));
+
+            /* check for non-default optimization group (Strategy 3A )*/
+            if ((extraInfo->OptGroup != 1) && (ULMBCS_AMBIGUOUS_MATCH(group, extraInfo->OptGroup)))
+            {
+                /*zhujin: upgrade, merge #39299 here (Lotus) */
+                /*To make R5 compatible translation, look for exceptional group first for non-DBCS*/
+
+                if(extraInfo->localeConverterIndex < ULMBCS_DOUBLEOPTGROUP_START)
+                {
+                  bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                     ULMBCS_GRP_L1, pLMBCS, &uniChar,
+                     &lastConverterIndex, groups_tried);
+
+                  if(!bytes_written)
+                  {
+                     bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                         ULMBCS_GRP_EXCEPT, pLMBCS, &uniChar,
+                         &lastConverterIndex, groups_tried);
+                  }
+                  if(!bytes_written)
+                  {
+                      bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                          extraInfo->localeConverterIndex, pLMBCS, &uniChar,
+                          &lastConverterIndex, groups_tried);
+                  }
+                }
+                else
+                {
+                     bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                         extraInfo->localeConverterIndex, pLMBCS, &uniChar,
+                         &lastConverterIndex, groups_tried);
+                }
+            }
+            /* check for locale optimization group (Strategy 3B) */
+            if (!bytes_written && (extraInfo->localeConverterIndex) && (ULMBCS_AMBIGUOUS_MATCH(group, extraInfo->localeConverterIndex)))
+            {
+                bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                        extraInfo->localeConverterIndex, pLMBCS, &uniChar, &lastConverterIndex, groups_tried);
+            }
+            /* check for last optimization group used for this string (Strategy 3C) */
+            if (!bytes_written && (lastConverterIndex) && (ULMBCS_AMBIGUOUS_MATCH(group, lastConverterIndex)))
+            {
+                bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                        lastConverterIndex, pLMBCS, &uniChar, &lastConverterIndex, groups_tried);
+            }
+            if (!bytes_written)
+            {
+               /* just check every possible matching converter (Strategy 3D) */ 
+               ulmbcs_byte_t grp_start;
+               ulmbcs_byte_t grp_end;  
+               ulmbcs_byte_t grp_ix;
+               grp_start = (ulmbcs_byte_t)((group == ULMBCS_AMBIGUOUS_MBCS) 
+                        ? ULMBCS_DOUBLEOPTGROUP_START 
+                        :  ULMBCS_GRP_L1);
+               grp_end = (ulmbcs_byte_t)((group == ULMBCS_AMBIGUOUS_MBCS) 
+                        ? ULMBCS_GRP_LAST 
+                        :  ULMBCS_GRP_TH);
+               if(group == ULMBCS_AMBIGUOUS_ALL)
+               {
+                   grp_start = ULMBCS_GRP_L1;
+                   grp_end = ULMBCS_GRP_LAST;
+               }
+               for (grp_ix = grp_start;
+                   grp_ix <= grp_end && !bytes_written; 
+                    grp_ix++)
+               {
+                  if (extraInfo->OptGrpConverter [grp_ix] && !groups_tried [grp_ix])
+                  {
+                     bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                       grp_ix, pLMBCS, &uniChar, 
+                       &lastConverterIndex, groups_tried);
+                  }
+               }
+                /* a final conversion fallback to the exceptions group if its likely 
+                     to be single byte  (Strategy 3E) */
+               if (!bytes_written && grp_start == ULMBCS_GRP_L1)
+               {
+                  bytes_written = (int32_t)LMBCSConversionWorker (extraInfo,
+                     ULMBCS_GRP_EXCEPT, pLMBCS, &uniChar, 
+                     &lastConverterIndex, groups_tried);
+               }
+            }
+            /* all of our other strategies failed. Fallback to Unicode. (Strategy 4)*/
+            if (!bytes_written)
+            {
+
+               pLMBCS += LMBCSConvertUni(pLMBCS, uniChar);
+               bytes_written = (int32_t)(pLMBCS - LMBCS);
+            }
+         }
+      }
+  
+      /* we have a translation. increment source and write as much as posible to target */
+      args->source++;
+      pLMBCS = LMBCS;
+      while (args->target < args->targetLimit && bytes_written--)
+      {
+         *(args->target)++ = *pLMBCS++;
+         if (args->offsets)
+         {
+            *(args->offsets)++ = sourceIndex;
+         }
+      }
+      sourceIndex++;
+      if (bytes_written > 0)
+      {
+         /* write any bytes that didn't fit in target to the error buffer,
+            common code will move this to target if we get called back with
+            enough target room
+         */
+         uint8_t * pErrorBuffer = args->converter->charErrorBuffer;
+         *err = U_BUFFER_OVERFLOW_ERROR;
+         args->converter->charErrorBufferLength = (int8_t)bytes_written;
+         while (bytes_written--)
+         {
+            *pErrorBuffer++ = *pLMBCS++;
+         }
+      }
+      /*Fix for SPR#DJOE66JFN3 (Lotus)*/
+      extraInfo->localeConverterIndex = OldConverterIndex;
+   }     
+}
+
+
+/* Now, the Unicode from LMBCS section */
+
+
+/* A function to call when we are looking at the Unicode group byte in LMBCS */
+static UChar
+GetUniFromLMBCSUni(char const ** ppLMBCSin)  /* Called with LMBCS-style Unicode byte stream */
+{
+   uint8_t  HighCh = *(*ppLMBCSin)++;  /* Big-endian Unicode in LMBCS compatibility group*/
+   uint8_t  LowCh  = *(*ppLMBCSin)++;
+
+   if (HighCh == ULMBCS_UNICOMPATZERO ) 
+   {
+      HighCh = LowCh;
+      LowCh = 0; /* zero-byte in LSB special character */
+   }
+   return (UChar)((HighCh << 8) | LowCh);
+}
+
+
+
+/* CHECK_SOURCE_LIMIT: Helper macro to verify that there are at least'index' 
+   bytes left in source up to  sourceLimit.Errors appropriately if not.
+   If we reach the limit, then update the source pointer to there to consume
+   all input as required by ICU converter semantics.
+*/
+
+#define CHECK_SOURCE_LIMIT(index) UPRV_BLOCK_MACRO_BEGIN { \
+    if (args->source+index > args->sourceLimit) { \
+        *err = U_TRUNCATED_CHAR_FOUND; \
+        args->source = args->sourceLimit; \
+        return 0xffff; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* Return the Unicode representation for the current LMBCS character */
+
+static UChar32  U_CALLCONV
+_LMBCSGetNextUCharWorker(UConverterToUnicodeArgs*   args,
+                         UErrorCode*   err)
+{
+    UChar32 uniChar = 0;    /* an output UNICODE char */
+    ulmbcs_byte_t   CurByte; /* A byte from the input stream */
+
+    /* error check */
+    if (args->source >= args->sourceLimit)
+    {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0xffff;
+    }
+    /* Grab first byte & save address for error recovery */
+    CurByte = *((ulmbcs_byte_t  *) (args->source++));
+   
+    /*
+    * at entry of each if clause:
+    * 1. 'CurByte' points at the first byte of a LMBCS character
+    * 2. '*source'points to the next byte of the source stream after 'CurByte' 
+    *
+    * the job of each if clause is:
+    * 1. set '*source' to point at the beginning of next char (nop if LMBCS char is only 1 byte)
+    * 2. set 'uniChar' up with the right Unicode value, or set 'err' appropriately
+    */
+   
+    /* First lets check the simple fixed values. */
+
+    if(((CurByte > ULMBCS_C0END) && (CurByte < ULMBCS_C1START)) /* ascii range */
+    ||  (CurByte == 0) 
+    ||  CurByte == ULMBCS_HT || CurByte == ULMBCS_CR 
+    ||  CurByte == ULMBCS_LF || CurByte == ULMBCS_123SYSTEMRANGE)
+    {
+        uniChar = CurByte;
+    }
+    else  
+    {
+        UConverterDataLMBCS * extraInfo;
+        ulmbcs_byte_t group; 
+        UConverterSharedData *cnv; 
+        
+        if (CurByte == ULMBCS_GRP_CTRL)  /* Control character group - no opt group update */
+        {
+            ulmbcs_byte_t  C0C1byte;
+            CHECK_SOURCE_LIMIT(1);
+            C0C1byte = *(args->source)++;
+            uniChar = (C0C1byte < ULMBCS_C1START) ? C0C1byte - ULMBCS_CTRLOFFSET : C0C1byte;
+        }
+        else 
+        if (CurByte == ULMBCS_GRP_UNICODE) /* Unicode compatibility group: BigEndian UTF16 */
+        {
+            CHECK_SOURCE_LIMIT(2);
+     
+            /* don't check for error indicators fffe/ffff below */
+            return GetUniFromLMBCSUni(&(args->source));
+        }
+        else if (CurByte <= ULMBCS_CTRLOFFSET)  
+        {
+            group = CurByte;                   /* group byte is in the source */
+            extraInfo = (UConverterDataLMBCS *) args->converter->extraInfo;
+            if (group > ULMBCS_GRP_LAST || (cnv = extraInfo->OptGrpConverter[group]) == NULL)
+            {
+                /* this is not a valid group byte - no converter*/
+                *err = U_INVALID_CHAR_FOUND;
+            }      
+            else if (group >= ULMBCS_DOUBLEOPTGROUP_START)    /* double byte conversion */
+            {
+
+                CHECK_SOURCE_LIMIT(2);
+
+                /* check for LMBCS doubled-group-byte case */
+                if (*args->source == group) {
+                    /* single byte */
+                    ++args->source;
+                    uniChar = ucnv_MBCSSimpleGetNextUChar(cnv, args->source, 1, FALSE);
+                    ++args->source;
+                } else {
+                    /* double byte */
+                    uniChar = ucnv_MBCSSimpleGetNextUChar(cnv, args->source, 2, FALSE);
+                    args->source += 2;
+                }
+            }
+            else {                                  /* single byte conversion */
+                CHECK_SOURCE_LIMIT(1);
+                CurByte = *(args->source)++;
+        
+                if (CurByte >= ULMBCS_C1START)
+                {
+                    uniChar = _MBCS_SINGLE_SIMPLE_GET_NEXT_BMP(cnv, CurByte);
+                }
+                else
+                {
+                    /* The non-optimizable oddballs where there is an explicit byte 
+                    * AND the second byte is not in the upper ascii range
+                    */
+                    char bytes[2];
+
+                    extraInfo = (UConverterDataLMBCS *) args->converter->extraInfo;
+                    cnv = extraInfo->OptGrpConverter [ULMBCS_GRP_EXCEPT];  
+        
+                    /* Lookup value must include opt group */
+                    bytes[0] = group;
+                    bytes[1] = CurByte;
+                    uniChar = ucnv_MBCSSimpleGetNextUChar(cnv, bytes, 2, FALSE);
+                }
+            }
+        }
+        else if (CurByte >= ULMBCS_C1START) /* group byte is implicit */
+        {
+            extraInfo = (UConverterDataLMBCS *) args->converter->extraInfo;
+            group = extraInfo->OptGroup;
+            cnv = extraInfo->OptGrpConverter[group];
+            if (group >= ULMBCS_DOUBLEOPTGROUP_START)    /* double byte conversion */
+            {
+                if (!ucnv_MBCSIsLeadByte(cnv, CurByte))
+                {
+                    CHECK_SOURCE_LIMIT(0);
+
+                    /* let the MBCS conversion consume CurByte again */
+                    uniChar = ucnv_MBCSSimpleGetNextUChar(cnv, args->source - 1, 1, FALSE);
+                }
+                else
+                {
+                    CHECK_SOURCE_LIMIT(1);
+                    /* let the MBCS conversion consume CurByte again */
+                    uniChar = ucnv_MBCSSimpleGetNextUChar(cnv, args->source - 1, 2, FALSE);
+                    ++args->source;
+                }
+            }
+            else                                   /* single byte conversion */
+            {
+                uniChar = _MBCS_SINGLE_SIMPLE_GET_NEXT_BMP(cnv, CurByte);
+            }
+        }
+    }
+    return uniChar;
+}
+
+
+/* The exported function that converts lmbcs to one or more
+   UChars - currently UTF-16
+*/
+static void  U_CALLCONV
+_LMBCSToUnicodeWithOffsets(UConverterToUnicodeArgs*    args,
+                     UErrorCode*    err)
+{
+   char LMBCS [ULMBCS_CHARSIZE_MAX];
+   UChar uniChar;    /* one output UNICODE char */
+   const char * saveSource; /* beginning of current code point */
+   const char * pStartLMBCS = args->source;  /* beginning of whole string */
+   const char * errSource = NULL; /* pointer to actual input in case an error occurs */
+   int8_t savebytes = 0;
+
+   /* Process from source to limit, or until error */
+   while (U_SUCCESS(*err) && args->sourceLimit > args->source && args->targetLimit > args->target)
+   {
+      saveSource = args->source; /* beginning of current code point */
+
+      if (args->converter->toULength) /* reassemble char from previous call */
+      {
+        const char *saveSourceLimit; 
+        size_t size_old = args->converter->toULength;
+
+         /* limit from source is either remainder of temp buffer, or user limit on source */
+        size_t size_new_maybe_1 = sizeof(LMBCS) - size_old;
+        size_t size_new_maybe_2 = args->sourceLimit - args->source;
+        size_t size_new = (size_new_maybe_1 < size_new_maybe_2) ? size_new_maybe_1 : size_new_maybe_2;
+         
+      
+        uprv_memcpy(LMBCS, args->converter->toUBytes, size_old);
+        uprv_memcpy(LMBCS + size_old, args->source, size_new);
+        saveSourceLimit = args->sourceLimit;
+        args->source = errSource = LMBCS;
+        args->sourceLimit = LMBCS+size_old+size_new;
+        savebytes = (int8_t)(size_old+size_new);
+        uniChar = (UChar) _LMBCSGetNextUCharWorker(args, err);
+        args->source = saveSource + ((args->source - LMBCS) - size_old);
+        args->sourceLimit = saveSourceLimit;
+
+        if (*err == U_TRUNCATED_CHAR_FOUND)
+        {
+            /* evil special case: source buffers so small a char spans more than 2 buffers */
+            args->converter->toULength = savebytes;
+            uprv_memcpy(args->converter->toUBytes, LMBCS, savebytes);
+            args->source = args->sourceLimit;
+            *err = U_ZERO_ERROR;
+            return;
+         }
+         else
+         {
+            /* clear the partial-char marker */
+            args->converter->toULength = 0;
+         }
+      }
+      else
+      {
+         errSource = saveSource;
+         uniChar = (UChar) _LMBCSGetNextUCharWorker(args, err);
+         savebytes = (int8_t)(args->source - saveSource);
+      }
+      if (U_SUCCESS(*err))
+      {
+         if (uniChar < 0xfffe)
+         {
+            *(args->target)++ = uniChar;
+            if(args->offsets)
+            {
+               *(args->offsets)++ = (int32_t)(saveSource - pStartLMBCS);
+            }
+         }
+         else if (uniChar == 0xfffe)
+         {
+            *err = U_INVALID_CHAR_FOUND;
+         }
+         else /* if (uniChar == 0xffff) */
+         {
+            *err = U_ILLEGAL_CHAR_FOUND;
+         }
+      }
+   }
+   /* if target ran out before source, return U_BUFFER_OVERFLOW_ERROR */
+   if (U_SUCCESS(*err) && args->sourceLimit > args->source && args->targetLimit <= args->target)
+   {
+      *err = U_BUFFER_OVERFLOW_ERROR;
+   }
+   else if (U_FAILURE(*err)) 
+   {
+      /* If character incomplete or unmappable/illegal, store it in toUBytes[] */
+      args->converter->toULength = savebytes;
+      if (savebytes > 0) {
+         uprv_memcpy(args->converter->toUBytes, errSource, savebytes);
+      }
+      if (*err == U_TRUNCATED_CHAR_FOUND) {
+         *err = U_ZERO_ERROR;
+      }
+   }
+}
+
+/* And now, the macroized declarations of data & functions: */
+DEFINE_LMBCS_OPEN(1)
+DEFINE_LMBCS_OPEN(2)
+DEFINE_LMBCS_OPEN(3)
+DEFINE_LMBCS_OPEN(4)
+DEFINE_LMBCS_OPEN(5)
+DEFINE_LMBCS_OPEN(6)
+DEFINE_LMBCS_OPEN(8)
+DEFINE_LMBCS_OPEN(11)
+DEFINE_LMBCS_OPEN(16)
+DEFINE_LMBCS_OPEN(17)
+DEFINE_LMBCS_OPEN(18)
+DEFINE_LMBCS_OPEN(19)
+
+
+DECLARE_LMBCS_DATA(1)
+DECLARE_LMBCS_DATA(2)
+DECLARE_LMBCS_DATA(3)
+DECLARE_LMBCS_DATA(4)
+DECLARE_LMBCS_DATA(5)
+DECLARE_LMBCS_DATA(6)
+DECLARE_LMBCS_DATA(8)
+DECLARE_LMBCS_DATA(11)
+DECLARE_LMBCS_DATA(16)
+DECLARE_LMBCS_DATA(17)
+DECLARE_LMBCS_DATA(18)
+DECLARE_LMBCS_DATA(19)
+
+U_CDECL_END
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnv_set.cpp b/thirdparty/icu4c/common/ucnv_set.cpp
new file mode 100644
index 0000000000..926cee0de8
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_set.cpp
@@ -0,0 +1,70 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2007, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucnv_set.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004sep07
+*   created by: Markus W. Scherer
+*
+*   Conversion API functions using USet (ucnv_getUnicodeSet())
+*   moved here from ucnv.c for removing the dependency of other ucnv_
+*   implementation functions on the USet implementation.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+#include "unicode/ucnv.h"
+#include "ucnv_bld.h"
+#include "uset_imp.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+U_CAPI void U_EXPORT2
+ucnv_getUnicodeSet(const UConverter *cnv,
+                   USet *setFillIn,
+                   UConverterUnicodeSet whichSet,
+                   UErrorCode *pErrorCode) {
+    /* argument checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if(cnv==NULL || setFillIn==NULL || whichSet<UCNV_ROUNDTRIP_SET || UCNV_SET_COUNT<=whichSet) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    /* does this converter support this function? */
+    if(cnv->sharedData->impl->getUnicodeSet==NULL) {
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return;
+    }
+
+    {
+        USetAdder sa={
+            NULL,
+            uset_add,
+            uset_addRange,
+            uset_addString,
+            uset_remove,
+            uset_removeRange
+        };
+        sa.set=setFillIn;
+
+        /* empty the set */
+        uset_clear(setFillIn);
+
+        /* call the converter to add the code points it supports */
+        cnv->sharedData->impl->getUnicodeSet(cnv, &sa, whichSet, pErrorCode);
+    }
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_u16.cpp b/thirdparty/icu4c/common/ucnv_u16.cpp
new file mode 100644
index 0000000000..a5e8367400
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_u16.cpp
@@ -0,0 +1,1579 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2002-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_u16.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jul01
+*   created by: Markus W. Scherer
+*
+*   UTF-16 converter implementation. Used to be in ucnv_utf.c.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/uversion.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+
+enum {
+    UCNV_NEED_TO_WRITE_BOM=1
+};
+
+U_CDECL_BEGIN
+/*
+ * The UTF-16 toUnicode implementation is also used for the Java-specific
+ * "with BOM" variants of UTF-16BE and UTF-16LE.
+ */
+static void  U_CALLCONV
+_UTF16ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                           UErrorCode *pErrorCode);
+
+/* UTF-16BE ----------------------------------------------------------------- */
+
+#if U_IS_BIG_ENDIAN
+#   define _UTF16PEFromUnicodeWithOffsets   _UTF16BEFromUnicodeWithOffsets
+#else
+#   define _UTF16PEFromUnicodeWithOffsets   _UTF16LEFromUnicodeWithOffsets
+#endif
+
+
+static void  U_CALLCONV
+_UTF16BEFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                               UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source;
+    char *target;
+    int32_t *offsets;
+
+    uint32_t targetCapacity, length, sourceIndex;
+    UChar c, trail;
+    char overflow[4];
+
+    source=pArgs->source;
+    length=(int32_t)(pArgs->sourceLimit-source);
+    if(length<=0) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    cnv=pArgs->converter;
+
+    /* write the BOM if necessary */
+    if(cnv->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ (char)0xfeu, (char)0xffu };
+        ucnv_fromUWriteBytes(cnv,
+                             bom, 2,
+                             &pArgs->target, pArgs->targetLimit,
+                             &pArgs->offsets, -1,
+                             pErrorCode);
+        cnv->fromUnicodeStatus=0;
+    }
+
+    target=pArgs->target;
+    if(target >= pArgs->targetLimit) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        return;
+    }
+
+    targetCapacity=(uint32_t)(pArgs->targetLimit-target);
+    offsets=pArgs->offsets;
+    sourceIndex=0;
+
+    /* c!=0 indicates in several places outside the main loops that a surrogate was found */
+
+    if((c=(UChar)cnv->fromUChar32)!=0 && U16_IS_TRAIL(trail=*source) && targetCapacity>=4) {
+        /* the last buffer ended with a lead surrogate, output the surrogate pair */
+        ++source;
+        --length;
+        target[0]=(uint8_t)(c>>8);
+        target[1]=(uint8_t)c;
+        target[2]=(uint8_t)(trail>>8);
+        target[3]=(uint8_t)trail;
+        target+=4;
+        targetCapacity-=4;
+        if(offsets!=NULL) {
+            *offsets++=-1;
+            *offsets++=-1;
+            *offsets++=-1;
+            *offsets++=-1;
+        }
+        sourceIndex=1;
+        cnv->fromUChar32=c=0;
+    }
+
+    if(c==0) {
+        /* copy an even number of bytes for complete UChars */
+        uint32_t count=2*length;
+        if(count>targetCapacity) {
+            count=targetCapacity&~1;
+        }
+        /* count is even */
+        targetCapacity-=count;
+        count>>=1;
+        length-=count;
+
+        if(offsets==NULL) {
+            while(count>0) {
+                c=*source++;
+                if(U16_IS_SINGLE(c)) {
+                    target[0]=(uint8_t)(c>>8);
+                    target[1]=(uint8_t)c;
+                    target+=2;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 && U16_IS_TRAIL(trail=*source)) {
+                    ++source;
+                    --count;
+                    target[0]=(uint8_t)(c>>8);
+                    target[1]=(uint8_t)c;
+                    target[2]=(uint8_t)(trail>>8);
+                    target[3]=(uint8_t)trail;
+                    target+=4;
+                } else {
+                    break;
+                }
+                --count;
+            }
+        } else {
+            while(count>0) {
+                c=*source++;
+                if(U16_IS_SINGLE(c)) {
+                    target[0]=(uint8_t)(c>>8);
+                    target[1]=(uint8_t)c;
+                    target+=2;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex++;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 && U16_IS_TRAIL(trail=*source)) {
+                    ++source;
+                    --count;
+                    target[0]=(uint8_t)(c>>8);
+                    target[1]=(uint8_t)c;
+                    target[2]=(uint8_t)(trail>>8);
+                    target[3]=(uint8_t)trail;
+                    target+=4;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=2;
+                } else {
+                    break;
+                }
+                --count;
+            }
+        }
+
+        if(count==0) {
+            /* done with the loop for complete UChars */
+            if(length>0 && targetCapacity>0) {
+                /*
+                 * there is more input and some target capacity -
+                 * it must be targetCapacity==1 because otherwise
+                 * the above would have copied more;
+                 * prepare for overflow output
+                 */
+                if(U16_IS_SINGLE(c=*source++)) {
+                    overflow[0]=(char)(c>>8);
+                    overflow[1]=(char)c;
+                    length=2; /* 2 bytes to output */
+                    c=0;
+                /* } else { keep c for surrogate handling, length will be set there */
+                }
+            } else {
+                length=0;
+                c=0;
+            }
+        } else {
+            /* keep c for surrogate handling, length will be set there */
+            targetCapacity+=2*count;
+        }
+    } else {
+        length=0; /* from here on, length counts the bytes in overflow[] */
+    }
+    
+    if(c!=0) {
+        /*
+         * c is a surrogate, and
+         * - source or target too short
+         * - or the surrogate is unmatched
+         */
+        length=0;
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(source<pArgs->sourceLimit) {
+                if(U16_IS_TRAIL(trail=*source)) {
+                    /* output the surrogate pair, will overflow (see conditions comment above) */
+                    ++source;
+                    overflow[0]=(char)(c>>8);
+                    overflow[1]=(char)c;
+                    overflow[2]=(char)(trail>>8);
+                    overflow[3]=(char)trail;
+                    length=4; /* 4 bytes to output */
+                    c=0;
+                } else {
+                    /* unmatched lead surrogate */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else {
+                /* see if the trail surrogate is in the next buffer */
+            }
+        } else {
+            /* unmatched trail surrogate */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        }
+        cnv->fromUChar32=c;
+    }
+
+    if(length>0) {
+        /* output length bytes with overflow (length>targetCapacity>0) */
+        ucnv_fromUWriteBytes(cnv,
+                             overflow, length,
+                             (char **)&target, pArgs->targetLimit,
+                             &offsets, sourceIndex,
+                             pErrorCode);
+        targetCapacity=(uint32_t)(pArgs->targetLimit-(char *)target);
+    }
+
+    if(U_SUCCESS(*pErrorCode) && source<pArgs->sourceLimit && targetCapacity==0) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+static void  U_CALLCONV
+_UTF16BEToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                             UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source;
+    UChar *target;
+    int32_t *offsets;
+
+    uint32_t targetCapacity, length, count, sourceIndex;
+    UChar c, trail;
+
+    if(pArgs->converter->mode<8) {
+        _UTF16ToUnicodeWithOffsets(pArgs, pErrorCode);
+        return;
+    }
+
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    length=(int32_t)((const uint8_t *)pArgs->sourceLimit-source);
+    if(length<=0 && cnv->toUnicodeStatus==0) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    target=pArgs->target;
+    if(target >= pArgs->targetLimit) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        return;
+    }
+
+    targetCapacity=(uint32_t)(pArgs->targetLimit-target);
+    offsets=pArgs->offsets;
+    sourceIndex=0;
+    c=0;
+
+    /* complete a partial UChar or pair from the last call */
+    if(cnv->toUnicodeStatus!=0) {
+        /*
+         * special case: single byte from a previous buffer,
+         * where the byte turned out not to belong to a trail surrogate
+         * and the preceding, unmatched lead surrogate was put into toUBytes[]
+         * for error handling
+         */
+        cnv->toUBytes[0]=(uint8_t)cnv->toUnicodeStatus;
+        cnv->toULength=1;
+        cnv->toUnicodeStatus=0;
+    }
+    if((count=cnv->toULength)!=0) {
+        uint8_t *p=cnv->toUBytes;
+        do {
+            p[count++]=*source++;
+            ++sourceIndex;
+            --length;
+            if(count==2) {
+                c=((UChar)p[0]<<8)|p[1];
+                if(U16_IS_SINGLE(c)) {
+                    /* output the BMP code point */
+                    *target++=c;
+                    if(offsets!=NULL) {
+                        *offsets++=-1;
+                    }
+                    --targetCapacity;
+                    count=0;
+                    c=0;
+                    break;
+                } else if(U16_IS_SURROGATE_LEAD(c)) {
+                    /* continue collecting bytes for the trail surrogate */
+                    c=0; /* avoid unnecessary surrogate handling below */
+                } else {
+                    /* fall through to error handling for an unmatched trail surrogate */
+                    break;
+                }
+            } else if(count==4) {
+                c=((UChar)p[0]<<8)|p[1];
+                trail=((UChar)p[2]<<8)|p[3];
+                if(U16_IS_TRAIL(trail)) {
+                    /* output the surrogate pair */
+                    *target++=c;
+                    if(targetCapacity>=2) {
+                        *target++=trail;
+                        if(offsets!=NULL) {
+                            *offsets++=-1;
+                            *offsets++=-1;
+                        }
+                        targetCapacity-=2;
+                    } else /* targetCapacity==1 */ {
+                        targetCapacity=0;
+                        cnv->UCharErrorBuffer[0]=trail;
+                        cnv->UCharErrorBufferLength=1;
+                        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    }
+                    count=0;
+                    c=0;
+                    break;
+                } else {
+                    /* unmatched lead surrogate, handle here for consistent toUBytes[] */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+
+                    /* back out reading the code unit after it */
+                    if(((const uint8_t *)pArgs->source-source)>=2) {
+                        source-=2;
+                    } else {
+                        /*
+                         * if the trail unit's first byte was in a previous buffer, then
+                         * we need to put it into a special place because toUBytes[] will be
+                         * used for the lead unit's bytes
+                         */
+                        cnv->toUnicodeStatus=0x100|p[2];
+                        --source;
+                    }
+                    cnv->toULength=2;
+
+                    /* write back the updated pointers */
+                    pArgs->source=(const char *)source;
+                    pArgs->target=target;
+                    pArgs->offsets=offsets;
+                    return;
+                }
+            }
+        } while(length>0);
+        cnv->toULength=(int8_t)count;
+    }
+
+    /* copy an even number of bytes for complete UChars */
+    count=2*targetCapacity;
+    if(count>length) {
+        count=length&~1;
+    }
+    if(c==0 && count>0) {
+        length-=count;
+        count>>=1;
+        targetCapacity-=count;
+        if(offsets==NULL) {
+            do {
+                c=((UChar)source[0]<<8)|source[1];
+                source+=2;
+                if(U16_IS_SINGLE(c)) {
+                    *target++=c;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 &&
+                          U16_IS_TRAIL(trail=((UChar)source[0]<<8)|source[1])
+                ) {
+                    source+=2;
+                    --count;
+                    *target++=c;
+                    *target++=trail;
+                } else {
+                    break;
+                }
+            } while(--count>0);
+        } else {
+            do {
+                c=((UChar)source[0]<<8)|source[1];
+                source+=2;
+                if(U16_IS_SINGLE(c)) {
+                    *target++=c;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=2;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 &&
+                          U16_IS_TRAIL(trail=((UChar)source[0]<<8)|source[1])
+                ) {
+                    source+=2;
+                    --count;
+                    *target++=c;
+                    *target++=trail;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=4;
+                } else {
+                    break;
+                }
+            } while(--count>0);
+        }
+
+        if(count==0) {
+            /* done with the loop for complete UChars */
+            c=0;
+        } else {
+            /* keep c for surrogate handling, trail will be set there */
+            length+=2*(count-1); /* one more byte pair was consumed than count decremented */
+            targetCapacity+=count;
+        }
+    }
+
+    if(c!=0) {
+        /*
+         * c is a surrogate, and
+         * - source or target too short
+         * - or the surrogate is unmatched
+         */
+        cnv->toUBytes[0]=(uint8_t)(c>>8);
+        cnv->toUBytes[1]=(uint8_t)c;
+        cnv->toULength=2;
+
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(length>=2) {
+                if(U16_IS_TRAIL(trail=((UChar)source[0]<<8)|source[1])) {
+                    /* output the surrogate pair, will overflow (see conditions comment above) */
+                    source+=2;
+                    length-=2;
+                    *target++=c;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    cnv->UCharErrorBuffer[0]=trail;
+                    cnv->UCharErrorBufferLength=1;
+                    cnv->toULength=0;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                } else {
+                    /* unmatched lead surrogate */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else {
+                /* see if the trail surrogate is in the next buffer */
+            }
+        } else {
+            /* unmatched trail surrogate */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode)) {
+        /* check for a remaining source byte */
+        if(length>0) {
+            if(targetCapacity==0) {
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            } else {
+                /* it must be length==1 because otherwise the above would have copied more */
+                cnv->toUBytes[cnv->toULength++]=*source++;
+            }
+        }
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+static UChar32  U_CALLCONV
+_UTF16BEGetNextUChar(UConverterToUnicodeArgs *pArgs, UErrorCode *err) {
+    const uint8_t *s, *sourceLimit;
+    UChar32 c;
+
+    if(pArgs->converter->mode<8) {
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    }
+
+    s=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+
+    if(s>=sourceLimit) {
+        /* no input */
+        *err=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0xffff;
+    }
+
+    if(s+2>sourceLimit) {
+        /* only one byte: truncated UChar */
+        pArgs->converter->toUBytes[0]=*s++;
+        pArgs->converter->toULength=1;
+        pArgs->source=(const char *)s;
+        *err = U_TRUNCATED_CHAR_FOUND;
+        return 0xffff;
+    }
+
+    /* get one UChar */
+    c=((UChar32)*s<<8)|s[1];
+    s+=2;
+
+    /* check for a surrogate pair */
+    if(U_IS_SURROGATE(c)) {
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(s+2<=sourceLimit) {
+                UChar trail;
+
+                /* get a second UChar and see if it is a trail surrogate */
+                trail=((UChar)*s<<8)|s[1];
+                if(U16_IS_TRAIL(trail)) {
+                    c=U16_GET_SUPPLEMENTARY(c, trail);
+                    s+=2;
+                } else {
+                    /* unmatched lead surrogate */
+                    c=-2;
+                }
+            } else {
+                /* too few (2 or 3) bytes for a surrogate pair: truncated code point */
+                uint8_t *bytes=pArgs->converter->toUBytes;
+                s-=2;
+                pArgs->converter->toULength=(int8_t)(sourceLimit-s);
+                do {
+                    *bytes++=*s++;
+                } while(s<sourceLimit);
+
+                c=0xffff;
+                *err=U_TRUNCATED_CHAR_FOUND;
+            }
+        } else {
+            /* unmatched trail surrogate */
+            c=-2;
+        }
+
+        if(c<0) {
+            /* write the unmatched surrogate */
+            uint8_t *bytes=pArgs->converter->toUBytes;
+            pArgs->converter->toULength=2;
+            *bytes=*(s-2);
+            bytes[1]=*(s-1);
+
+            c=0xffff;
+            *err=U_ILLEGAL_CHAR_FOUND;
+        }
+    }
+
+    pArgs->source=(const char *)s;
+    return c;
+} 
+
+static void  U_CALLCONV
+_UTF16BEReset(UConverter *cnv, UConverterResetChoice choice) {
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode state */
+        if(UCNV_GET_VERSION(cnv)==0) {
+            cnv->mode=8; /* no BOM handling */
+        } else {
+            cnv->mode=0; /* Java-specific "UnicodeBig" requires BE BOM or no BOM */
+        }
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE && UCNV_GET_VERSION(cnv)==1) {
+        /* reset fromUnicode for "UnicodeBig": prepare to output the UTF-16BE BOM */
+        cnv->fromUnicodeStatus=UCNV_NEED_TO_WRITE_BOM;
+    }
+}
+
+static void  U_CALLCONV
+_UTF16BEOpen(UConverter *cnv,
+             UConverterLoadArgs *pArgs,
+             UErrorCode *pErrorCode) {
+    (void)pArgs;
+    if(UCNV_GET_VERSION(cnv)<=1) {
+        _UTF16BEReset(cnv, UCNV_RESET_BOTH);
+    } else {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static const char *  U_CALLCONV
+_UTF16BEGetName(const UConverter *cnv) {
+    if(UCNV_GET_VERSION(cnv)==0) {
+        return "UTF-16BE";
+    } else {
+        return "UTF-16BE,version=1";
+    }
+}
+U_CDECL_END
+
+static const UConverterImpl _UTF16BEImpl={
+    UCNV_UTF16_BigEndian,
+
+    NULL,
+    NULL,
+
+    _UTF16BEOpen,
+    NULL,
+    _UTF16BEReset,
+
+    _UTF16BEToUnicodeWithOffsets,
+    _UTF16BEToUnicodeWithOffsets,
+    _UTF16BEFromUnicodeWithOffsets,
+    _UTF16BEFromUnicodeWithOffsets,
+    _UTF16BEGetNextUChar,
+
+    NULL,
+    _UTF16BEGetName,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _UTF16BEStaticData={
+    sizeof(UConverterStaticData),
+    "UTF-16BE",
+    1200, UCNV_IBM, UCNV_UTF16_BigEndian, 2, 2,
+    { 0xff, 0xfd, 0, 0 },2,FALSE,FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+
+const UConverterSharedData _UTF16BEData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF16BEStaticData, &_UTF16BEImpl);
+
+/* UTF-16LE ----------------------------------------------------------------- */
+U_CDECL_BEGIN
+static void  U_CALLCONV
+_UTF16LEFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                               UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source;
+    char *target;
+    int32_t *offsets;
+
+    uint32_t targetCapacity, length, sourceIndex;
+    UChar c, trail;
+    char overflow[4];
+
+    source=pArgs->source;
+    length=(int32_t)(pArgs->sourceLimit-source);
+    if(length<=0) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    cnv=pArgs->converter;
+
+    /* write the BOM if necessary */
+    if(cnv->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ (char)0xffu, (char)0xfeu };
+        ucnv_fromUWriteBytes(cnv,
+                             bom, 2,
+                             &pArgs->target, pArgs->targetLimit,
+                             &pArgs->offsets, -1,
+                             pErrorCode);
+        cnv->fromUnicodeStatus=0;
+    }
+
+    target=pArgs->target;
+    if(target >= pArgs->targetLimit) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        return;
+    }
+
+    targetCapacity=(uint32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+    sourceIndex=0;
+
+    /* c!=0 indicates in several places outside the main loops that a surrogate was found */
+
+    if((c=(UChar)cnv->fromUChar32)!=0 && U16_IS_TRAIL(trail=*source) && targetCapacity>=4) {
+        /* the last buffer ended with a lead surrogate, output the surrogate pair */
+        ++source;
+        --length;
+        target[0]=(uint8_t)c;
+        target[1]=(uint8_t)(c>>8);
+        target[2]=(uint8_t)trail;
+        target[3]=(uint8_t)(trail>>8);
+        target+=4;
+        targetCapacity-=4;
+        if(offsets!=NULL) {
+            *offsets++=-1;
+            *offsets++=-1;
+            *offsets++=-1;
+            *offsets++=-1;
+        }
+        sourceIndex=1;
+        cnv->fromUChar32=c=0;
+    }
+
+    if(c==0) {
+        /* copy an even number of bytes for complete UChars */
+        uint32_t count=2*length;
+        if(count>targetCapacity) {
+            count=targetCapacity&~1;
+        }
+        /* count is even */
+        targetCapacity-=count;
+        count>>=1;
+        length-=count;
+
+        if(offsets==NULL) {
+            while(count>0) {
+                c=*source++;
+                if(U16_IS_SINGLE(c)) {
+                    target[0]=(uint8_t)c;
+                    target[1]=(uint8_t)(c>>8);
+                    target+=2;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 && U16_IS_TRAIL(trail=*source)) {
+                    ++source;
+                    --count;
+                    target[0]=(uint8_t)c;
+                    target[1]=(uint8_t)(c>>8);
+                    target[2]=(uint8_t)trail;
+                    target[3]=(uint8_t)(trail>>8);
+                    target+=4;
+                } else {
+                    break;
+                }
+                --count;
+            }
+        } else {
+            while(count>0) {
+                c=*source++;
+                if(U16_IS_SINGLE(c)) {
+                    target[0]=(uint8_t)c;
+                    target[1]=(uint8_t)(c>>8);
+                    target+=2;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex++;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 && U16_IS_TRAIL(trail=*source)) {
+                    ++source;
+                    --count;
+                    target[0]=(uint8_t)c;
+                    target[1]=(uint8_t)(c>>8);
+                    target[2]=(uint8_t)trail;
+                    target[3]=(uint8_t)(trail>>8);
+                    target+=4;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=2;
+                } else {
+                    break;
+                }
+                --count;
+            }
+        }
+
+        if(count==0) {
+            /* done with the loop for complete UChars */
+            if(length>0 && targetCapacity>0) {
+                /*
+                 * there is more input and some target capacity -
+                 * it must be targetCapacity==1 because otherwise
+                 * the above would have copied more;
+                 * prepare for overflow output
+                 */
+                if(U16_IS_SINGLE(c=*source++)) {
+                    overflow[0]=(char)c;
+                    overflow[1]=(char)(c>>8);
+                    length=2; /* 2 bytes to output */
+                    c=0;
+                /* } else { keep c for surrogate handling, length will be set there */
+                }
+            } else {
+                length=0;
+                c=0;
+            }
+        } else {
+            /* keep c for surrogate handling, length will be set there */
+            targetCapacity+=2*count;
+        }
+    } else {
+        length=0; /* from here on, length counts the bytes in overflow[] */
+    }
+    
+    if(c!=0) {
+        /*
+         * c is a surrogate, and
+         * - source or target too short
+         * - or the surrogate is unmatched
+         */
+        length=0;
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(source<pArgs->sourceLimit) {
+                if(U16_IS_TRAIL(trail=*source)) {
+                    /* output the surrogate pair, will overflow (see conditions comment above) */
+                    ++source;
+                    overflow[0]=(char)c;
+                    overflow[1]=(char)(c>>8);
+                    overflow[2]=(char)trail;
+                    overflow[3]=(char)(trail>>8);
+                    length=4; /* 4 bytes to output */
+                    c=0;
+                } else {
+                    /* unmatched lead surrogate */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else {
+                /* see if the trail surrogate is in the next buffer */
+            }
+        } else {
+            /* unmatched trail surrogate */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        }
+        cnv->fromUChar32=c;
+    }
+
+    if(length>0) {
+        /* output length bytes with overflow (length>targetCapacity>0) */
+        ucnv_fromUWriteBytes(cnv,
+                             overflow, length,
+                             &target, pArgs->targetLimit,
+                             &offsets, sourceIndex,
+                             pErrorCode);
+        targetCapacity=(uint32_t)(pArgs->targetLimit-(char *)target);
+    }
+
+    if(U_SUCCESS(*pErrorCode) && source<pArgs->sourceLimit && targetCapacity==0) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+static void  U_CALLCONV
+_UTF16LEToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                             UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source;
+    UChar *target;
+    int32_t *offsets;
+
+    uint32_t targetCapacity, length, count, sourceIndex;
+    UChar c, trail;
+
+    if(pArgs->converter->mode<8) {
+        _UTF16ToUnicodeWithOffsets(pArgs, pErrorCode);
+        return;
+    }
+
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    length=(int32_t)((const uint8_t *)pArgs->sourceLimit-source);
+    if(length<=0 && cnv->toUnicodeStatus==0) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    target=pArgs->target;
+    if(target >= pArgs->targetLimit) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        return;
+    }
+
+    targetCapacity=(uint32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+    sourceIndex=0;
+    c=0;
+
+    /* complete a partial UChar or pair from the last call */
+    if(cnv->toUnicodeStatus!=0) {
+        /*
+         * special case: single byte from a previous buffer,
+         * where the byte turned out not to belong to a trail surrogate
+         * and the preceding, unmatched lead surrogate was put into toUBytes[]
+         * for error handling
+         */
+        cnv->toUBytes[0]=(uint8_t)cnv->toUnicodeStatus;
+        cnv->toULength=1;
+        cnv->toUnicodeStatus=0;
+    }
+    if((count=cnv->toULength)!=0) {
+        uint8_t *p=cnv->toUBytes;
+        do {
+            p[count++]=*source++;
+            ++sourceIndex;
+            --length;
+            if(count==2) {
+                c=((UChar)p[1]<<8)|p[0];
+                if(U16_IS_SINGLE(c)) {
+                    /* output the BMP code point */
+                    *target++=c;
+                    if(offsets!=NULL) {
+                        *offsets++=-1;
+                    }
+                    --targetCapacity;
+                    count=0;
+                    c=0;
+                    break;
+                } else if(U16_IS_SURROGATE_LEAD(c)) {
+                    /* continue collecting bytes for the trail surrogate */
+                    c=0; /* avoid unnecessary surrogate handling below */
+                } else {
+                    /* fall through to error handling for an unmatched trail surrogate */
+                    break;
+                }
+            } else if(count==4) {
+                c=((UChar)p[1]<<8)|p[0];
+                trail=((UChar)p[3]<<8)|p[2];
+                if(U16_IS_TRAIL(trail)) {
+                    /* output the surrogate pair */
+                    *target++=c;
+                    if(targetCapacity>=2) {
+                        *target++=trail;
+                        if(offsets!=NULL) {
+                            *offsets++=-1;
+                            *offsets++=-1;
+                        }
+                        targetCapacity-=2;
+                    } else /* targetCapacity==1 */ {
+                        targetCapacity=0;
+                        cnv->UCharErrorBuffer[0]=trail;
+                        cnv->UCharErrorBufferLength=1;
+                        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    }
+                    count=0;
+                    c=0;
+                    break;
+                } else {
+                    /* unmatched lead surrogate, handle here for consistent toUBytes[] */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+
+                    /* back out reading the code unit after it */
+                    if(((const uint8_t *)pArgs->source-source)>=2) {
+                        source-=2;
+                    } else {
+                        /*
+                         * if the trail unit's first byte was in a previous buffer, then
+                         * we need to put it into a special place because toUBytes[] will be
+                         * used for the lead unit's bytes
+                         */
+                        cnv->toUnicodeStatus=0x100|p[2];
+                        --source;
+                    }
+                    cnv->toULength=2;
+
+                    /* write back the updated pointers */
+                    pArgs->source=(const char *)source;
+                    pArgs->target=target;
+                    pArgs->offsets=offsets;
+                    return;
+                }
+            }
+        } while(length>0);
+        cnv->toULength=(int8_t)count;
+    }
+
+    /* copy an even number of bytes for complete UChars */
+    count=2*targetCapacity;
+    if(count>length) {
+        count=length&~1;
+    }
+    if(c==0 && count>0) {
+        length-=count;
+        count>>=1;
+        targetCapacity-=count;
+        if(offsets==NULL) {
+            do {
+                c=((UChar)source[1]<<8)|source[0];
+                source+=2;
+                if(U16_IS_SINGLE(c)) {
+                    *target++=c;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 &&
+                          U16_IS_TRAIL(trail=((UChar)source[1]<<8)|source[0])
+                ) {
+                    source+=2;
+                    --count;
+                    *target++=c;
+                    *target++=trail;
+                } else {
+                    break;
+                }
+            } while(--count>0);
+        } else {
+            do {
+                c=((UChar)source[1]<<8)|source[0];
+                source+=2;
+                if(U16_IS_SINGLE(c)) {
+                    *target++=c;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=2;
+                } else if(U16_IS_SURROGATE_LEAD(c) && count>=2 &&
+                          U16_IS_TRAIL(trail=((UChar)source[1]<<8)|source[0])
+                ) {
+                    source+=2;
+                    --count;
+                    *target++=c;
+                    *target++=trail;
+                    *offsets++=sourceIndex;
+                    *offsets++=sourceIndex;
+                    sourceIndex+=4;
+                } else {
+                    break;
+                }
+            } while(--count>0);
+        }
+
+        if(count==0) {
+            /* done with the loop for complete UChars */
+            c=0;
+        } else {
+            /* keep c for surrogate handling, trail will be set there */
+            length+=2*(count-1); /* one more byte pair was consumed than count decremented */
+            targetCapacity+=count;
+        }
+    }
+
+    if(c!=0) {
+        /*
+         * c is a surrogate, and
+         * - source or target too short
+         * - or the surrogate is unmatched
+         */
+        cnv->toUBytes[0]=(uint8_t)c;
+        cnv->toUBytes[1]=(uint8_t)(c>>8);
+        cnv->toULength=2;
+
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(length>=2) {
+                if(U16_IS_TRAIL(trail=((UChar)source[1]<<8)|source[0])) {
+                    /* output the surrogate pair, will overflow (see conditions comment above) */
+                    source+=2;
+                    length-=2;
+                    *target++=c;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    cnv->UCharErrorBuffer[0]=trail;
+                    cnv->UCharErrorBufferLength=1;
+                    cnv->toULength=0;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                } else {
+                    /* unmatched lead surrogate */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else {
+                /* see if the trail surrogate is in the next buffer */
+            }
+        } else {
+            /* unmatched trail surrogate */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode)) {
+        /* check for a remaining source byte */
+        if(length>0) {
+            if(targetCapacity==0) {
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            } else {
+                /* it must be length==1 because otherwise the above would have copied more */
+                cnv->toUBytes[cnv->toULength++]=*source++;
+            }
+        }
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+static UChar32  U_CALLCONV
+_UTF16LEGetNextUChar(UConverterToUnicodeArgs *pArgs, UErrorCode *err) {
+    const uint8_t *s, *sourceLimit;
+    UChar32 c;
+
+    if(pArgs->converter->mode<8) {
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    }
+
+    s=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+
+    if(s>=sourceLimit) {
+        /* no input */
+        *err=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0xffff;
+    }
+
+    if(s+2>sourceLimit) {
+        /* only one byte: truncated UChar */
+        pArgs->converter->toUBytes[0]=*s++;
+        pArgs->converter->toULength=1;
+        pArgs->source=(const char *)s;
+        *err = U_TRUNCATED_CHAR_FOUND;
+        return 0xffff;
+    }
+
+    /* get one UChar */
+    c=((UChar32)s[1]<<8)|*s;
+    s+=2;
+
+    /* check for a surrogate pair */
+    if(U_IS_SURROGATE(c)) {
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            if(s+2<=sourceLimit) {
+                UChar trail;
+
+                /* get a second UChar and see if it is a trail surrogate */
+                trail=((UChar)s[1]<<8)|*s;
+                if(U16_IS_TRAIL(trail)) {
+                    c=U16_GET_SUPPLEMENTARY(c, trail);
+                    s+=2;
+                } else {
+                    /* unmatched lead surrogate */
+                    c=-2;
+                }
+            } else {
+                /* too few (2 or 3) bytes for a surrogate pair: truncated code point */
+                uint8_t *bytes=pArgs->converter->toUBytes;
+                s-=2;
+                pArgs->converter->toULength=(int8_t)(sourceLimit-s);
+                do {
+                    *bytes++=*s++;
+                } while(s<sourceLimit);
+
+                c=0xffff;
+                *err=U_TRUNCATED_CHAR_FOUND;
+            }
+        } else {
+            /* unmatched trail surrogate */
+            c=-2;
+        }
+
+        if(c<0) {
+            /* write the unmatched surrogate */
+            uint8_t *bytes=pArgs->converter->toUBytes;
+            pArgs->converter->toULength=2;
+            *bytes=*(s-2);
+            bytes[1]=*(s-1);
+
+            c=0xffff;
+            *err=U_ILLEGAL_CHAR_FOUND;
+        }
+    }
+
+    pArgs->source=(const char *)s;
+    return c;
+} 
+
+static void  U_CALLCONV
+_UTF16LEReset(UConverter *cnv, UConverterResetChoice choice) {
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode state */
+        if(UCNV_GET_VERSION(cnv)==0) {
+            cnv->mode=8; /* no BOM handling */
+        } else {
+            cnv->mode=0; /* Java-specific "UnicodeLittle" requires LE BOM or no BOM */
+        }
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE && UCNV_GET_VERSION(cnv)==1) {
+        /* reset fromUnicode for "UnicodeLittle": prepare to output the UTF-16LE BOM */
+        cnv->fromUnicodeStatus=UCNV_NEED_TO_WRITE_BOM;
+    }
+}
+
+static void  U_CALLCONV
+_UTF16LEOpen(UConverter *cnv,
+             UConverterLoadArgs *pArgs,
+             UErrorCode *pErrorCode) {
+    (void)pArgs;
+    if(UCNV_GET_VERSION(cnv)<=1) {
+        _UTF16LEReset(cnv, UCNV_RESET_BOTH);
+    } else {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static const char *  U_CALLCONV
+_UTF16LEGetName(const UConverter *cnv) {
+    if(UCNV_GET_VERSION(cnv)==0) {
+        return "UTF-16LE";
+    } else {
+        return "UTF-16LE,version=1";
+    }
+}
+U_CDECL_END
+
+static const UConverterImpl _UTF16LEImpl={
+    UCNV_UTF16_LittleEndian,
+
+    NULL,
+    NULL,
+
+    _UTF16LEOpen,
+    NULL,
+    _UTF16LEReset,
+
+    _UTF16LEToUnicodeWithOffsets,
+    _UTF16LEToUnicodeWithOffsets,
+    _UTF16LEFromUnicodeWithOffsets,
+    _UTF16LEFromUnicodeWithOffsets,
+    _UTF16LEGetNextUChar,
+
+    NULL,
+    _UTF16LEGetName,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+
+static const UConverterStaticData _UTF16LEStaticData={
+    sizeof(UConverterStaticData),
+    "UTF-16LE",
+    1202, UCNV_IBM, UCNV_UTF16_LittleEndian, 2, 2,
+    { 0xfd, 0xff, 0, 0 },2,FALSE,FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+
+const UConverterSharedData _UTF16LEData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF16LEStaticData, &_UTF16LEImpl);
+
+/* UTF-16 (Detect BOM) ------------------------------------------------------ */
+
+/*
+ * Detect a BOM at the beginning of the stream and select UTF-16BE or UTF-16LE
+ * accordingly.
+ * This is a simpler version of the UTF-32 converter, with
+ * fewer states for shorter BOMs.
+ *
+ * State values:
+ * 0    initial state
+ * 1    saw first byte
+ * 2..5 -
+ * 6..7 see _UTF16ToUnicodeWithOffsets() comments in state 1
+ * 8    UTF-16BE mode
+ * 9    UTF-16LE mode
+ *
+ * During detection: state==number of initial bytes seen so far.
+ *
+ * On output, emit U+FEFF as the first code point.
+ *
+ * Variants:
+ * - UTF-16,version=1 (Java "Unicode" encoding) treats a missing BOM as an error.
+ * - UTF-16BE,version=1 (Java "UnicodeBig" encoding) and
+ *   UTF-16LE,version=1 (Java "UnicodeLittle" encoding) treat a reverse BOM as an error.
+ */
+U_CDECL_BEGIN
+static void  U_CALLCONV
+_UTF16Reset(UConverter *cnv, UConverterResetChoice choice) {
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode: state=0 */
+        cnv->mode=0;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        /* reset fromUnicode: prepare to output the UTF-16PE BOM */
+        cnv->fromUnicodeStatus=UCNV_NEED_TO_WRITE_BOM;
+    }
+}
+U_CDECL_END
+extern const UConverterSharedData _UTF16v2Data;
+U_CDECL_BEGIN
+static void U_CALLCONV
+_UTF16Open(UConverter *cnv,
+           UConverterLoadArgs *pArgs,
+           UErrorCode *pErrorCode) {
+    if(UCNV_GET_VERSION(cnv)<=2) {
+        if(UCNV_GET_VERSION(cnv)==2 && !pArgs->onlyTestIsLoadable) {
+            /*
+             * Switch implementation, and switch the staticData that's different
+             * and was copied into the UConverter.
+             * (See ucnv_createConverterFromSharedData() in ucnv_bld.c.)
+             * UTF-16,version=2 fromUnicode() always writes a big-endian byte stream.
+             */
+            cnv->sharedData=(UConverterSharedData*)&_UTF16v2Data;
+            uprv_memcpy(cnv->subChars, _UTF16v2Data.staticData->subChar, UCNV_MAX_SUBCHAR_LEN);
+        }
+        _UTF16Reset(cnv, UCNV_RESET_BOTH);
+    } else {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static const char *  U_CALLCONV
+_UTF16GetName(const UConverter *cnv) {
+    if(UCNV_GET_VERSION(cnv)==0) {
+        return "UTF-16";
+    } else if(UCNV_GET_VERSION(cnv)==1) {
+        return "UTF-16,version=1";
+    } else {
+        return "UTF-16,version=2";
+    }
+}
+U_CDECL_END
+extern const UConverterSharedData _UTF16Data;
+
+static inline bool IS_UTF16BE(const UConverter *cnv) {
+    return ((cnv)->sharedData == &_UTF16BEData);
+}
+
+static inline bool IS_UTF16LE(const UConverter *cnv) {
+    return ((cnv)->sharedData == &_UTF16LEData);
+}
+
+static inline bool IS_UTF16(const UConverter *cnv) {
+    return ((cnv)->sharedData==&_UTF16Data) || ((cnv)->sharedData == &_UTF16v2Data);
+}
+
+U_CDECL_BEGIN
+static void U_CALLCONV
+_UTF16ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                           UErrorCode *pErrorCode) {
+    UConverter *cnv=pArgs->converter;
+    const char *source=pArgs->source;
+    const char *sourceLimit=pArgs->sourceLimit;
+    int32_t *offsets=pArgs->offsets;
+
+    int32_t state, offsetDelta;
+    uint8_t b;
+
+    state=cnv->mode;
+
+    /*
+     * If we detect a BOM in this buffer, then we must add the BOM size to the
+     * offsets because the actual converter function will not see and count the BOM.
+     * offsetDelta will have the number of the BOM bytes that are in the current buffer.
+     */
+    offsetDelta=0;
+
+    while(source<sourceLimit && U_SUCCESS(*pErrorCode)) {
+        switch(state) {
+        case 0:
+            cnv->toUBytes[0]=(uint8_t)*source++;
+            cnv->toULength=1;
+            state=1;
+            break;
+        case 1:
+            /*
+             * Only inside this switch case can the state variable
+             * temporarily take two additional values:
+             * 6: BOM error, continue with BE
+             * 7: BOM error, continue with LE
+             */
+            b=*source;
+            if(cnv->toUBytes[0]==0xfe && b==0xff) {
+                if(IS_UTF16LE(cnv)) {
+                    state=7; /* illegal reverse BOM for Java "UnicodeLittle" */
+                } else {
+                    state=8; /* detect UTF-16BE */
+                }
+            } else if(cnv->toUBytes[0]==0xff && b==0xfe) {
+                if(IS_UTF16BE(cnv)) {
+                    state=6; /* illegal reverse BOM for Java "UnicodeBig" */
+                } else {
+                    state=9; /* detect UTF-16LE */
+                }
+            } else if((IS_UTF16(cnv) && UCNV_GET_VERSION(cnv)==1)) {
+                state=6; /* illegal missing BOM for Java "Unicode" */
+            }
+            if(state>=8) {
+                /* BOM detected, consume it */
+                ++source;
+                cnv->toULength=0;
+                offsetDelta=(int32_t)(source-pArgs->source);
+            } else if(state<6) {
+                /* ok: no BOM, and not a reverse BOM */
+                if(source!=pArgs->source) {
+                    /* reset the source for a correct first offset */
+                    source=pArgs->source;
+                    cnv->toULength=0;
+                }
+                if(IS_UTF16LE(cnv)) {
+                    /* Make Java "UnicodeLittle" default to LE. */
+                    state=9;
+                } else {
+                    /* Make standard UTF-16 and Java "UnicodeBig" default to BE. */
+                    state=8;
+                }
+            } else {
+                /*
+                 * error: missing BOM, or reverse BOM
+                 * UTF-16,version=1: Java-specific "Unicode" requires a BOM.
+                 * UTF-16BE,version=1: Java-specific "UnicodeBig" requires a BE BOM or no BOM.
+                 * UTF-16LE,version=1: Java-specific "UnicodeLittle" requires an LE BOM or no BOM.
+                 */
+                /* report the non-BOM or reverse BOM as an illegal sequence */
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                pArgs->source=source+1;
+                /* continue with conversion if the callback resets the error */
+                /*
+                 * Make Java "Unicode" default to BE like standard UTF-16.
+                 * Make Java "UnicodeBig" and "UnicodeLittle" default
+                 * to their normal endiannesses.
+                 */
+                cnv->mode=state+2;
+                *pErrorCode=U_ILLEGAL_ESCAPE_SEQUENCE;
+                return;
+            }
+            /* convert the rest of the stream */
+            cnv->mode=state;
+            continue;
+        case 8:
+            /* call UTF-16BE */
+            pArgs->source=source;
+            _UTF16BEToUnicodeWithOffsets(pArgs, pErrorCode);
+            source=pArgs->source;
+            break;
+        case 9:
+            /* call UTF-16LE */
+            pArgs->source=source;
+            _UTF16LEToUnicodeWithOffsets(pArgs, pErrorCode);
+            source=pArgs->source;
+            break;
+        default:
+            break; /* does not occur */
+        }
+    }
+
+    /* add BOM size to offsets - see comment at offsetDelta declaration */
+    if(offsets!=NULL && offsetDelta!=0) {
+        int32_t *offsetsLimit=pArgs->offsets;
+        while(offsets<offsetsLimit) {
+            *offsets++ += offsetDelta;
+        }
+    }
+
+    pArgs->source=source;
+
+    if(source==sourceLimit && pArgs->flush) {
+        /* handle truncated input */
+        switch(state) {
+        case 0:
+            break; /* no input at all, nothing to do */
+        case 8:
+            _UTF16BEToUnicodeWithOffsets(pArgs, pErrorCode);
+            break;
+        case 9:
+            _UTF16LEToUnicodeWithOffsets(pArgs, pErrorCode);
+            break;
+        default:
+            /* 0<state<8: framework will report truncation, nothing to do here */
+            break;
+        }
+    }
+
+    cnv->mode=state;
+}
+
+static UChar32 U_CALLCONV
+_UTF16GetNextUChar(UConverterToUnicodeArgs *pArgs,
+                   UErrorCode *pErrorCode) {
+    switch(pArgs->converter->mode) {
+    case 8:
+        return _UTF16BEGetNextUChar(pArgs, pErrorCode);
+    case 9:
+        return _UTF16LEGetNextUChar(pArgs, pErrorCode);
+    default:
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    }
+}
+U_CDECL_END
+
+static const UConverterImpl _UTF16Impl = {
+    UCNV_UTF16,
+
+    NULL,
+    NULL,
+
+    _UTF16Open,
+    NULL,
+    _UTF16Reset,
+
+    _UTF16ToUnicodeWithOffsets,
+    _UTF16ToUnicodeWithOffsets,
+    _UTF16PEFromUnicodeWithOffsets,
+    _UTF16PEFromUnicodeWithOffsets,
+    _UTF16GetNextUChar,
+
+    NULL, /* ### TODO implement getStarters for all Unicode encodings?! */
+    _UTF16GetName,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _UTF16StaticData = {
+    sizeof(UConverterStaticData),
+    "UTF-16",
+    1204, /* CCSID for BOM sensitive UTF-16 */
+    UCNV_IBM, UCNV_UTF16, 2, 2,
+#if U_IS_BIG_ENDIAN
+    { 0xff, 0xfd, 0, 0 }, 2,
+#else
+    { 0xfd, 0xff, 0, 0 }, 2,
+#endif
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _UTF16Data =
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF16StaticData, &_UTF16Impl);
+
+static const UConverterImpl _UTF16v2Impl = {
+    UCNV_UTF16,
+
+    NULL,
+    NULL,
+
+    _UTF16Open,
+    NULL,
+    _UTF16Reset,
+
+    _UTF16ToUnicodeWithOffsets,
+    _UTF16ToUnicodeWithOffsets,
+    _UTF16BEFromUnicodeWithOffsets,
+    _UTF16BEFromUnicodeWithOffsets,
+    _UTF16GetNextUChar,
+
+    NULL, /* ### TODO implement getStarters for all Unicode encodings?! */
+    _UTF16GetName,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _UTF16v2StaticData = {
+    sizeof(UConverterStaticData),
+    "UTF-16,version=2",
+    1204, /* CCSID for BOM sensitive UTF-16 */
+    UCNV_IBM, UCNV_UTF16, 2, 2,
+    { 0xff, 0xfd, 0, 0 }, 2,
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _UTF16v2Data =
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF16v2StaticData, &_UTF16v2Impl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_u32.cpp b/thirdparty/icu4c/common/ucnv_u32.cpp
new file mode 100644
index 0000000000..9f98914b9d
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_u32.cpp
@@ -0,0 +1,1253 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2002-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_u32.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jul01
+*   created by: Markus W. Scherer
+*
+*   UTF-32 converter implementation. Used to be in ucnv_utf.c.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/utf.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+
+#define MAXIMUM_UCS2            0x0000FFFF
+#define MAXIMUM_UTF             0x0010FFFF
+#define HALF_SHIFT              10
+#define HALF_BASE               0x0010000
+#define HALF_MASK               0x3FF
+#define SURROGATE_HIGH_START    0xD800
+#define SURROGATE_LOW_START     0xDC00
+
+/* -SURROGATE_LOW_START + HALF_BASE */
+#define SURROGATE_LOW_BASE      9216
+
+enum {
+    UCNV_NEED_TO_WRITE_BOM=1
+};
+
+/* UTF-32BE ----------------------------------------------------------------- */
+U_CDECL_BEGIN
+static void U_CALLCONV
+T_UConverter_toUnicode_UTF32_BE(UConverterToUnicodeArgs * args,
+                                UErrorCode * err)
+{
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = args->converter->toUBytes;
+    uint32_t ch, i;
+
+    /* Restore state of current sequence */
+    if (args->converter->toULength > 0 && myTarget < targetLimit) {
+        i = args->converter->toULength;       /* restore # of bytes consumed */
+        args->converter->toULength = 0;
+
+        ch = args->converter->toUnicodeStatus - 1;/*Stores the previously calculated ch from a previous call*/
+        args->converter->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit) {
+        i = 0;
+        ch = 0;
+morebytes:
+        while (i < sizeof(uint32_t)) {
+            if (mySource < sourceLimit) {
+                ch = (ch << 8) | (uint8_t)(*mySource);
+                toUBytes[i++] = (char) *(mySource++);
+            }
+            else {
+                /* stores a partially calculated target*/
+                /* + 1 to make 0 a valid character */
+                args->converter->toUnicodeStatus = ch + 1;
+                args->converter->toULength = (int8_t) i;
+                goto donefornow;
+            }
+        }
+
+        if (ch <= MAXIMUM_UTF && !U_IS_SURROGATE(ch)) {
+            /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+            if (ch <= MAXIMUM_UCS2) 
+            {
+                /* fits in 16 bits */
+                *(myTarget++) = (UChar) ch;
+            }
+            else {
+                /* write out the surrogates */
+                *(myTarget++) = U16_LEAD(ch);
+                ch = U16_TRAIL(ch);
+                if (myTarget < targetLimit) {
+                    *(myTarget++) = (UChar)ch;
+                }
+                else {
+                    /* Put in overflow buffer (not handled here) */
+                    args->converter->UCharErrorBuffer[0] = (UChar) ch;
+                    args->converter->UCharErrorBufferLength = 1;
+                    *err = U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        }
+        else {
+            args->converter->toULength = (int8_t)i;
+            *err = U_ILLEGAL_CHAR_FOUND;
+            break;
+        }
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err)) {
+        /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+}
+
+static void U_CALLCONV
+T_UConverter_toUnicode_UTF32_BE_OFFSET_LOGIC(UConverterToUnicodeArgs * args,
+                                             UErrorCode * err)
+{
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    int32_t *myOffsets = args->offsets;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = args->converter->toUBytes;
+    uint32_t ch, i;
+    int32_t offsetNum = 0;
+
+    /* Restore state of current sequence */
+    if (args->converter->toULength > 0 && myTarget < targetLimit) {
+        i = args->converter->toULength;       /* restore # of bytes consumed */
+        args->converter->toULength = 0;
+
+        ch = args->converter->toUnicodeStatus - 1;/*Stores the previously calculated ch from a previous call*/
+        args->converter->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit) {
+        i = 0;
+        ch = 0;
+morebytes:
+        while (i < sizeof(uint32_t)) {
+            if (mySource < sourceLimit) {
+                ch = (ch << 8) | (uint8_t)(*mySource);
+                toUBytes[i++] = (char) *(mySource++);
+            }
+            else {
+                /* stores a partially calculated target*/
+                /* + 1 to make 0 a valid character */
+                args->converter->toUnicodeStatus = ch + 1;
+                args->converter->toULength = (int8_t) i;
+                goto donefornow;
+            }
+        }
+
+        if (ch <= MAXIMUM_UTF && !U_IS_SURROGATE(ch)) {
+            /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+            if (ch <= MAXIMUM_UCS2) {
+                /* fits in 16 bits */
+                *(myTarget++) = (UChar) ch;
+                *(myOffsets++) = offsetNum;
+            }
+            else {
+                /* write out the surrogates */
+                *(myTarget++) = U16_LEAD(ch);
+                *myOffsets++ = offsetNum;
+                ch = U16_TRAIL(ch);
+                if (myTarget < targetLimit)
+                {
+                    *(myTarget++) = (UChar)ch;
+                    *(myOffsets++) = offsetNum;
+                }
+                else {
+                    /* Put in overflow buffer (not handled here) */
+                    args->converter->UCharErrorBuffer[0] = (UChar) ch;
+                    args->converter->UCharErrorBufferLength = 1;
+                    *err = U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        }
+        else {
+            args->converter->toULength = (int8_t)i;
+            *err = U_ILLEGAL_CHAR_FOUND;
+            break;
+        }
+        offsetNum += i;
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+    args->offsets = myOffsets;
+}
+
+static void U_CALLCONV
+T_UConverter_fromUnicode_UTF32_BE(UConverterFromUnicodeArgs * args,
+                                  UErrorCode * err)
+{
+    const UChar *mySource = args->source;
+    unsigned char *myTarget;
+    const UChar *sourceLimit = args->sourceLimit;
+    const unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    UChar32 ch, ch2;
+    unsigned int indexToWrite;
+    unsigned char temp[sizeof(uint32_t)];
+
+    if(mySource >= sourceLimit) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    /* write the BOM if necessary */
+    if(args->converter->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ 0, 0, (char)0xfeu, (char)0xffu };
+        ucnv_fromUWriteBytes(args->converter,
+                             bom, 4,
+                             &args->target, args->targetLimit,
+                             &args->offsets, -1,
+                             err);
+        args->converter->fromUnicodeStatus=0;
+    }
+
+    myTarget = (unsigned char *) args->target;
+    temp[0] = 0;
+
+    if (args->converter->fromUChar32) {
+        ch = args->converter->fromUChar32;
+        args->converter->fromUChar32 = 0;
+        goto lowsurogate;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit) {
+        ch = *(mySource++);
+
+        if (U_IS_SURROGATE(ch)) {
+            if (U_IS_LEAD(ch)) {
+lowsurogate:
+                if (mySource < sourceLimit) {
+                    ch2 = *mySource;
+                    if (U_IS_TRAIL(ch2)) {
+                        ch = ((ch - SURROGATE_HIGH_START) << HALF_SHIFT) + ch2 + SURROGATE_LOW_BASE;
+                        mySource++;
+                    }
+                    else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        args->converter->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* ran out of source */
+                    args->converter->fromUChar32 = ch;
+                    if (args->flush) {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                    }
+                    break;
+                }
+            }
+            else {
+                /* this is an unmatched trail code unit (2nd surrogate) */
+                /* callback(illegal) */
+                args->converter->fromUChar32 = ch;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+
+        /* We cannot get any larger than 10FFFF because we are coming from UTF-16 */
+        temp[1] = (uint8_t) (ch >> 16 & 0x1F);
+        temp[2] = (uint8_t) (ch >> 8);  /* unsigned cast implicitly does (ch & FF) */
+        temp[3] = (uint8_t) (ch);       /* unsigned cast implicitly does (ch & FF) */
+
+        for (indexToWrite = 0; indexToWrite <= sizeof(uint32_t) - 1; indexToWrite++) {
+            if (myTarget < targetLimit) {
+                *(myTarget++) = temp[indexToWrite];
+            }
+            else {
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = temp[indexToWrite];
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err)) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+}
+
+static void U_CALLCONV
+T_UConverter_fromUnicode_UTF32_BE_OFFSET_LOGIC(UConverterFromUnicodeArgs * args,
+                                               UErrorCode * err)
+{
+    const UChar *mySource = args->source;
+    unsigned char *myTarget;
+    int32_t *myOffsets;
+    const UChar *sourceLimit = args->sourceLimit;
+    const unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    UChar32 ch, ch2;
+    int32_t offsetNum = 0;
+    unsigned int indexToWrite;
+    unsigned char temp[sizeof(uint32_t)];
+
+    if(mySource >= sourceLimit) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    /* write the BOM if necessary */
+    if(args->converter->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ 0, 0, (char)0xfeu, (char)0xffu };
+        ucnv_fromUWriteBytes(args->converter,
+                             bom, 4,
+                             &args->target, args->targetLimit,
+                             &args->offsets, -1,
+                             err);
+        args->converter->fromUnicodeStatus=0;
+    }
+
+    myTarget = (unsigned char *) args->target;
+    myOffsets = args->offsets;
+    temp[0] = 0;
+
+    if (args->converter->fromUChar32) {
+        ch = args->converter->fromUChar32;
+        args->converter->fromUChar32 = 0;
+        goto lowsurogate;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit) {
+        ch = *(mySource++);
+
+        if (U_IS_SURROGATE(ch)) {
+            if (U_IS_LEAD(ch)) {
+lowsurogate:
+                if (mySource < sourceLimit) {
+                    ch2 = *mySource;
+                    if (U_IS_TRAIL(ch2)) {
+                        ch = ((ch - SURROGATE_HIGH_START) << HALF_SHIFT) + ch2 + SURROGATE_LOW_BASE;
+                        mySource++;
+                    }
+                    else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        args->converter->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* ran out of source */
+                    args->converter->fromUChar32 = ch;
+                    if (args->flush) {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                    }
+                    break;
+                }
+            }
+            else {
+                /* this is an unmatched trail code unit (2nd surrogate) */
+                /* callback(illegal) */
+                args->converter->fromUChar32 = ch;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+
+        /* We cannot get any larger than 10FFFF because we are coming from UTF-16 */
+        temp[1] = (uint8_t) (ch >> 16 & 0x1F);
+        temp[2] = (uint8_t) (ch >> 8);  /* unsigned cast implicitly does (ch & FF) */
+        temp[3] = (uint8_t) (ch);       /* unsigned cast implicitly does (ch & FF) */
+
+        for (indexToWrite = 0; indexToWrite <= sizeof(uint32_t) - 1; indexToWrite++) {
+            if (myTarget < targetLimit) {
+                *(myTarget++) = temp[indexToWrite];
+                *(myOffsets++) = offsetNum;
+            }
+            else {
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = temp[indexToWrite];
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        offsetNum = offsetNum + 1 + (temp[1] != 0);
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err)) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+    args->offsets = myOffsets;
+}
+
+static UChar32 U_CALLCONV
+T_UConverter_getNextUChar_UTF32_BE(UConverterToUnicodeArgs* args,
+                                   UErrorCode* err)
+{
+    const uint8_t *mySource;
+    UChar32 myUChar;
+    int32_t length;
+
+    mySource = (const uint8_t *)args->source;
+    if (mySource >= (const uint8_t *)args->sourceLimit)
+    {
+        /* no input */
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0xffff;
+    }
+
+    length = (int32_t)((const uint8_t *)args->sourceLimit - mySource);
+    if (length < 4) 
+    {
+        /* got a partial character */
+        uprv_memcpy(args->converter->toUBytes, mySource, length);
+        args->converter->toULength = (int8_t)length;
+        args->source = (const char *)(mySource + length);
+        *err = U_TRUNCATED_CHAR_FOUND;
+        return 0xffff;
+    }
+
+    /* Don't even try to do a direct cast because the value may be on an odd address. */
+    myUChar = ((UChar32)mySource[0] << 24)
+            | ((UChar32)mySource[1] << 16)
+            | ((UChar32)mySource[2] << 8)
+            | ((UChar32)mySource[3]);
+
+    args->source = (const char *)(mySource + 4);
+    if ((uint32_t)myUChar <= MAXIMUM_UTF && !U_IS_SURROGATE(myUChar)) {
+        return myUChar;
+    }
+
+    uprv_memcpy(args->converter->toUBytes, mySource, 4);
+    args->converter->toULength = 4;
+
+    *err = U_ILLEGAL_CHAR_FOUND;
+    return 0xffff;
+}
+U_CDECL_END
+static const UConverterImpl _UTF32BEImpl = {
+    UCNV_UTF32_BigEndian,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    T_UConverter_toUnicode_UTF32_BE,
+    T_UConverter_toUnicode_UTF32_BE_OFFSET_LOGIC,
+    T_UConverter_fromUnicode_UTF32_BE,
+    T_UConverter_fromUnicode_UTF32_BE_OFFSET_LOGIC,
+    T_UConverter_getNextUChar_UTF32_BE,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+/* The 1232 CCSID refers to any version of Unicode with any endianess of UTF-32 */
+static const UConverterStaticData _UTF32BEStaticData = {
+    sizeof(UConverterStaticData),
+    "UTF-32BE",
+    1232,
+    UCNV_IBM, UCNV_UTF32_BigEndian, 4, 4,
+    { 0, 0, 0xff, 0xfd }, 4, FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _UTF32BEData =
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF32BEStaticData, &_UTF32BEImpl);
+
+/* UTF-32LE ---------------------------------------------------------- */
+U_CDECL_BEGIN
+static void U_CALLCONV
+T_UConverter_toUnicode_UTF32_LE(UConverterToUnicodeArgs * args,
+                                UErrorCode * err)
+{
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = args->converter->toUBytes;
+    uint32_t ch, i;
+
+    /* Restore state of current sequence */
+    if (args->converter->toULength > 0 && myTarget < targetLimit)
+    {
+        i = args->converter->toULength;       /* restore # of bytes consumed */
+        args->converter->toULength = 0;
+
+        /* Stores the previously calculated ch from a previous call*/
+        ch = args->converter->toUnicodeStatus - 1;
+        args->converter->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        i = 0;
+        ch = 0;
+morebytes:
+        while (i < sizeof(uint32_t))
+        {
+            if (mySource < sourceLimit)
+            {
+                ch |= ((uint8_t)(*mySource)) << (i * 8);
+                toUBytes[i++] = (char) *(mySource++);
+            }
+            else
+            {
+                /* stores a partially calculated target*/
+                /* + 1 to make 0 a valid character */
+                args->converter->toUnicodeStatus = ch + 1;
+                args->converter->toULength = (int8_t) i;
+                goto donefornow;
+            }
+        }
+
+        if (ch <= MAXIMUM_UTF && !U_IS_SURROGATE(ch)) {
+            /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+            if (ch <= MAXIMUM_UCS2) {
+                /* fits in 16 bits */
+                *(myTarget++) = (UChar) ch;
+            }
+            else {
+                /* write out the surrogates */
+                *(myTarget++) = U16_LEAD(ch);
+                ch = U16_TRAIL(ch);
+                if (myTarget < targetLimit) {
+                    *(myTarget++) = (UChar)ch;
+                }
+                else {
+                    /* Put in overflow buffer (not handled here) */
+                    args->converter->UCharErrorBuffer[0] = (UChar) ch;
+                    args->converter->UCharErrorBufferLength = 1;
+                    *err = U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        }
+        else {
+            args->converter->toULength = (int8_t)i;
+            *err = U_ILLEGAL_CHAR_FOUND;
+            break;
+        }
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+}
+
+static void U_CALLCONV
+T_UConverter_toUnicode_UTF32_LE_OFFSET_LOGIC(UConverterToUnicodeArgs * args,
+                                             UErrorCode * err)
+{
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    int32_t *myOffsets = args->offsets;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = args->converter->toUBytes;
+    uint32_t ch, i;
+    int32_t offsetNum = 0;
+
+    /* Restore state of current sequence */
+    if (args->converter->toULength > 0 && myTarget < targetLimit)
+    {
+        i = args->converter->toULength;       /* restore # of bytes consumed */
+        args->converter->toULength = 0;
+
+        /* Stores the previously calculated ch from a previous call*/
+        ch = args->converter->toUnicodeStatus - 1;
+        args->converter->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        i = 0;
+        ch = 0;
+morebytes:
+        while (i < sizeof(uint32_t))
+        {
+            if (mySource < sourceLimit)
+            {
+                ch |= ((uint8_t)(*mySource)) << (i * 8);
+                toUBytes[i++] = (char) *(mySource++);
+            }
+            else
+            {
+                /* stores a partially calculated target*/
+                /* + 1 to make 0 a valid character */
+                args->converter->toUnicodeStatus = ch + 1;
+                args->converter->toULength = (int8_t) i;
+                goto donefornow;
+            }
+        }
+
+        if (ch <= MAXIMUM_UTF && !U_IS_SURROGATE(ch))
+        {
+            /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+            if (ch <= MAXIMUM_UCS2) 
+            {
+                /* fits in 16 bits */
+                *(myTarget++) = (UChar) ch;
+                *(myOffsets++) = offsetNum;
+            }
+            else {
+                /* write out the surrogates */
+                *(myTarget++) = U16_LEAD(ch);
+                *(myOffsets++) = offsetNum;
+                ch = U16_TRAIL(ch);
+                if (myTarget < targetLimit)
+                {
+                    *(myTarget++) = (UChar)ch;
+                    *(myOffsets++) = offsetNum;
+                }
+                else
+                {
+                    /* Put in overflow buffer (not handled here) */
+                    args->converter->UCharErrorBuffer[0] = (UChar) ch;
+                    args->converter->UCharErrorBufferLength = 1;
+                    *err = U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        }
+        else
+        {
+            args->converter->toULength = (int8_t)i;
+            *err = U_ILLEGAL_CHAR_FOUND;
+            break;
+        }
+        offsetNum += i;
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+    args->offsets = myOffsets;
+}
+
+static void U_CALLCONV
+T_UConverter_fromUnicode_UTF32_LE(UConverterFromUnicodeArgs * args,
+                                  UErrorCode * err)
+{
+    const UChar *mySource = args->source;
+    unsigned char *myTarget;
+    const UChar *sourceLimit = args->sourceLimit;
+    const unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    UChar32 ch, ch2;
+    unsigned int indexToWrite;
+    unsigned char temp[sizeof(uint32_t)];
+
+    if(mySource >= sourceLimit) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    /* write the BOM if necessary */
+    if(args->converter->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ (char)0xffu, (char)0xfeu, 0, 0 };
+        ucnv_fromUWriteBytes(args->converter,
+                             bom, 4,
+                             &args->target, args->targetLimit,
+                             &args->offsets, -1,
+                             err);
+        args->converter->fromUnicodeStatus=0;
+    }
+
+    myTarget = (unsigned char *) args->target;
+    temp[3] = 0;
+
+    if (args->converter->fromUChar32)
+    {
+        ch = args->converter->fromUChar32;
+        args->converter->fromUChar32 = 0;
+        goto lowsurogate;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+
+        if (U16_IS_SURROGATE(ch)) {
+            if (U16_IS_LEAD(ch))
+            {
+lowsurogate:
+                if (mySource < sourceLimit)
+                {
+                    ch2 = *mySource;
+                    if (U16_IS_TRAIL(ch2)) {
+                        ch = ((ch - SURROGATE_HIGH_START) << HALF_SHIFT) + ch2 + SURROGATE_LOW_BASE;
+                        mySource++;
+                    }
+                    else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        args->converter->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* ran out of source */
+                    args->converter->fromUChar32 = ch;
+                    if (args->flush) {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                    }
+                    break;
+                }
+            }
+            else {
+                /* this is an unmatched trail code unit (2nd surrogate) */
+                /* callback(illegal) */
+                args->converter->fromUChar32 = ch;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+
+        /* We cannot get any larger than 10FFFF because we are coming from UTF-16 */
+        temp[2] = (uint8_t) (ch >> 16 & 0x1F);
+        temp[1] = (uint8_t) (ch >> 8);  /* unsigned cast implicitly does (ch & FF) */
+        temp[0] = (uint8_t) (ch);       /* unsigned cast implicitly does (ch & FF) */
+
+        for (indexToWrite = 0; indexToWrite <= sizeof(uint32_t) - 1; indexToWrite++)
+        {
+            if (myTarget < targetLimit)
+            {
+                *(myTarget++) = temp[indexToWrite];
+            }
+            else
+            {
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = temp[indexToWrite];
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+}
+
+static void U_CALLCONV
+T_UConverter_fromUnicode_UTF32_LE_OFFSET_LOGIC(UConverterFromUnicodeArgs * args,
+                                               UErrorCode * err)
+{
+    const UChar *mySource = args->source;
+    unsigned char *myTarget;
+    int32_t *myOffsets;
+    const UChar *sourceLimit = args->sourceLimit;
+    const unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    UChar32 ch, ch2;
+    unsigned int indexToWrite;
+    unsigned char temp[sizeof(uint32_t)];
+    int32_t offsetNum = 0;
+
+    if(mySource >= sourceLimit) {
+        /* no input, nothing to do */
+        return;
+    }
+
+    /* write the BOM if necessary */
+    if(args->converter->fromUnicodeStatus==UCNV_NEED_TO_WRITE_BOM) {
+        static const char bom[]={ (char)0xffu, (char)0xfeu, 0, 0 };
+        ucnv_fromUWriteBytes(args->converter,
+                             bom, 4,
+                             &args->target, args->targetLimit,
+                             &args->offsets, -1,
+                             err);
+        args->converter->fromUnicodeStatus=0;
+    }
+
+    myTarget = (unsigned char *) args->target;
+    myOffsets = args->offsets;
+    temp[3] = 0;
+
+    if (args->converter->fromUChar32)
+    {
+        ch = args->converter->fromUChar32;
+        args->converter->fromUChar32 = 0;
+        goto lowsurogate;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+
+        if (U16_IS_SURROGATE(ch)) {
+            if (U16_IS_LEAD(ch))
+            {
+lowsurogate:
+                if (mySource < sourceLimit)
+                {
+                    ch2 = *mySource;
+                    if (U16_IS_TRAIL(ch2))
+                    {
+                        ch = ((ch - SURROGATE_HIGH_START) << HALF_SHIFT) + ch2 + SURROGATE_LOW_BASE;
+                        mySource++;
+                    }
+                    else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        args->converter->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* ran out of source */
+                    args->converter->fromUChar32 = ch;
+                    if (args->flush) {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                    }
+                    break;
+                }
+            }
+            else {
+                /* this is an unmatched trail code unit (2nd surrogate) */
+                /* callback(illegal) */
+                args->converter->fromUChar32 = ch;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+
+        /* We cannot get any larger than 10FFFF because we are coming from UTF-16 */
+        temp[2] = (uint8_t) (ch >> 16 & 0x1F);
+        temp[1] = (uint8_t) (ch >> 8);  /* unsigned cast implicitly does (ch & FF) */
+        temp[0] = (uint8_t) (ch);       /* unsigned cast implicitly does (ch & FF) */
+
+        for (indexToWrite = 0; indexToWrite <= sizeof(uint32_t) - 1; indexToWrite++)
+        {
+            if (myTarget < targetLimit)
+            {
+                *(myTarget++) = temp[indexToWrite];
+                *(myOffsets++) = offsetNum;
+            }
+            else
+            {
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = temp[indexToWrite];
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        offsetNum = offsetNum + 1 + (temp[2] != 0);
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+    args->offsets = myOffsets;
+}
+
+static UChar32 U_CALLCONV
+T_UConverter_getNextUChar_UTF32_LE(UConverterToUnicodeArgs* args,
+                                   UErrorCode* err)
+{
+    const uint8_t *mySource;
+    UChar32 myUChar;
+    int32_t length;
+
+    mySource = (const uint8_t *)args->source;
+    if (mySource >= (const uint8_t *)args->sourceLimit)
+    {
+        /* no input */
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0xffff;
+    }
+
+    length = (int32_t)((const uint8_t *)args->sourceLimit - mySource);
+    if (length < 4) 
+    {
+        /* got a partial character */
+        uprv_memcpy(args->converter->toUBytes, mySource, length);
+        args->converter->toULength = (int8_t)length;
+        args->source = (const char *)(mySource + length);
+        *err = U_TRUNCATED_CHAR_FOUND;
+        return 0xffff;
+    }
+
+    /* Don't even try to do a direct cast because the value may be on an odd address. */
+    myUChar = ((UChar32)mySource[3] << 24)
+            | ((UChar32)mySource[2] << 16)
+            | ((UChar32)mySource[1] << 8)
+            | ((UChar32)mySource[0]);
+
+    args->source = (const char *)(mySource + 4);
+    if ((uint32_t)myUChar <= MAXIMUM_UTF && !U_IS_SURROGATE(myUChar)) {
+        return myUChar;
+    }
+
+    uprv_memcpy(args->converter->toUBytes, mySource, 4);
+    args->converter->toULength = 4;
+
+    *err = U_ILLEGAL_CHAR_FOUND;
+    return 0xffff;
+}
+U_CDECL_END
+static const UConverterImpl _UTF32LEImpl = {
+    UCNV_UTF32_LittleEndian,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    T_UConverter_toUnicode_UTF32_LE,
+    T_UConverter_toUnicode_UTF32_LE_OFFSET_LOGIC,
+    T_UConverter_fromUnicode_UTF32_LE,
+    T_UConverter_fromUnicode_UTF32_LE_OFFSET_LOGIC,
+    T_UConverter_getNextUChar_UTF32_LE,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+/* The 1232 CCSID refers to any version of Unicode with any endianess of UTF-32 */
+static const UConverterStaticData _UTF32LEStaticData = {
+    sizeof(UConverterStaticData),
+    "UTF-32LE",
+    1234,
+    UCNV_IBM, UCNV_UTF32_LittleEndian, 4, 4,
+    { 0xfd, 0xff, 0, 0 }, 4, FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+
+const UConverterSharedData _UTF32LEData =
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF32LEStaticData, &_UTF32LEImpl);
+
+/* UTF-32 (Detect BOM) ------------------------------------------------------ */
+
+/*
+ * Detect a BOM at the beginning of the stream and select UTF-32BE or UTF-32LE
+ * accordingly.
+ *
+ * State values:
+ * 0    initial state
+ * 1    saw 00
+ * 2    saw 00 00
+ * 3    saw 00 00 FE
+ * 4    -
+ * 5    saw FF
+ * 6    saw FF FE
+ * 7    saw FF FE 00
+ * 8    UTF-32BE mode
+ * 9    UTF-32LE mode
+ *
+ * During detection: state&3==number of matching bytes so far.
+ *
+ * On output, emit U+FEFF as the first code point.
+ */
+U_CDECL_BEGIN
+static void U_CALLCONV
+_UTF32Reset(UConverter *cnv, UConverterResetChoice choice) {
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode: state=0 */
+        cnv->mode=0;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        /* reset fromUnicode: prepare to output the UTF-32PE BOM */
+        cnv->fromUnicodeStatus=UCNV_NEED_TO_WRITE_BOM;
+    }
+}
+
+static void U_CALLCONV
+_UTF32Open(UConverter *cnv,
+           UConverterLoadArgs *pArgs,
+           UErrorCode *pErrorCode) {
+    (void)pArgs;
+    (void)pErrorCode;
+    _UTF32Reset(cnv, UCNV_RESET_BOTH);
+}
+
+static const char utf32BOM[8]={ 0, 0, (char)0xfeu, (char)0xffu, (char)0xffu, (char)0xfeu, 0, 0 };
+
+static void U_CALLCONV
+_UTF32ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                           UErrorCode *pErrorCode) {
+    UConverter *cnv=pArgs->converter;
+    const char *source=pArgs->source;
+    const char *sourceLimit=pArgs->sourceLimit;
+    int32_t *offsets=pArgs->offsets;
+
+    int32_t state, offsetDelta;
+    char b;
+
+    state=cnv->mode;
+
+    /*
+     * If we detect a BOM in this buffer, then we must add the BOM size to the
+     * offsets because the actual converter function will not see and count the BOM.
+     * offsetDelta will have the number of the BOM bytes that are in the current buffer.
+     */
+    offsetDelta=0;
+
+    while(source<sourceLimit && U_SUCCESS(*pErrorCode)) {
+        switch(state) {
+        case 0:
+            b=*source;
+            if(b==0) {
+                state=1; /* could be 00 00 FE FF */
+            } else if(b==(char)0xffu) {
+                state=5; /* could be FF FE 00 00 */
+            } else {
+                state=8; /* default to UTF-32BE */
+                continue;
+            }
+            ++source;
+            break;
+        case 1:
+        case 2:
+        case 3:
+        case 5:
+        case 6:
+        case 7:
+            if(*source==utf32BOM[state]) {
+                ++state;
+                ++source;
+                if(state==4) {
+                    state=8; /* detect UTF-32BE */
+                    offsetDelta=(int32_t)(source-pArgs->source);
+                } else if(state==8) {
+                    state=9; /* detect UTF-32LE */
+                    offsetDelta=(int32_t)(source-pArgs->source);
+                }
+            } else {
+                /* switch to UTF-32BE and pass the previous bytes */
+                int32_t count=(int32_t)(source-pArgs->source); /* number of bytes from this buffer */
+
+                /* reset the source */
+                source=pArgs->source;
+
+                if(count==(state&3)) {
+                    /* simple: all in the same buffer, just reset source */
+                } else {
+                    UBool oldFlush=pArgs->flush;
+
+                    /* some of the bytes are from a previous buffer, replay those first */
+                    pArgs->source=utf32BOM+(state&4); /* select the correct BOM */
+                    pArgs->sourceLimit=pArgs->source+((state&3)-count); /* replay previous bytes */
+                    pArgs->flush=FALSE; /* this sourceLimit is not the real source stream limit */
+
+                    /* no offsets: bytes from previous buffer, and not enough for output */
+                    T_UConverter_toUnicode_UTF32_BE(pArgs, pErrorCode);
+
+                    /* restore real pointers; pArgs->source will be set in case 8/9 */
+                    pArgs->sourceLimit=sourceLimit;
+                    pArgs->flush=oldFlush;
+                }
+                state=8;
+                continue;
+            }
+            break;
+        case 8:
+            /* call UTF-32BE */
+            pArgs->source=source;
+            if(offsets==NULL) {
+                T_UConverter_toUnicode_UTF32_BE(pArgs, pErrorCode);
+            } else {
+                T_UConverter_toUnicode_UTF32_BE_OFFSET_LOGIC(pArgs, pErrorCode);
+            }
+            source=pArgs->source;
+            break;
+        case 9:
+            /* call UTF-32LE */
+            pArgs->source=source;
+            if(offsets==NULL) {
+                T_UConverter_toUnicode_UTF32_LE(pArgs, pErrorCode);
+            } else {
+                T_UConverter_toUnicode_UTF32_LE_OFFSET_LOGIC(pArgs, pErrorCode);
+            }
+            source=pArgs->source;
+            break;
+        default:
+            break; /* does not occur */
+        }
+    }
+
+    /* add BOM size to offsets - see comment at offsetDelta declaration */
+    if(offsets!=NULL && offsetDelta!=0) {
+        int32_t *offsetsLimit=pArgs->offsets;
+        while(offsets<offsetsLimit) {
+            *offsets++ += offsetDelta;
+        }
+    }
+
+    pArgs->source=source;
+
+    if(source==sourceLimit && pArgs->flush) {
+        /* handle truncated input */
+        switch(state) {
+        case 0:
+            break; /* no input at all, nothing to do */
+        case 8:
+            T_UConverter_toUnicode_UTF32_BE(pArgs, pErrorCode);
+            break;
+        case 9:
+            T_UConverter_toUnicode_UTF32_LE(pArgs, pErrorCode);
+            break;
+        default:
+            /* handle 0<state<8: call UTF-32BE with too-short input */
+            pArgs->source=utf32BOM+(state&4); /* select the correct BOM */
+            pArgs->sourceLimit=pArgs->source+(state&3); /* replay bytes */
+
+            /* no offsets: not enough for output */
+            T_UConverter_toUnicode_UTF32_BE(pArgs, pErrorCode);
+            pArgs->source=source;
+            pArgs->sourceLimit=sourceLimit;
+            state=8;
+            break;
+        }
+    }
+
+    cnv->mode=state;
+}
+
+static UChar32 U_CALLCONV
+_UTF32GetNextUChar(UConverterToUnicodeArgs *pArgs,
+                   UErrorCode *pErrorCode) {
+    switch(pArgs->converter->mode) {
+    case 8:
+        return T_UConverter_getNextUChar_UTF32_BE(pArgs, pErrorCode);
+    case 9:
+        return T_UConverter_getNextUChar_UTF32_LE(pArgs, pErrorCode);
+    default:
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    }
+}
+U_CDECL_END
+static const UConverterImpl _UTF32Impl = {
+    UCNV_UTF32,
+
+    NULL,
+    NULL,
+
+    _UTF32Open,
+    NULL,
+    _UTF32Reset,
+
+    _UTF32ToUnicodeWithOffsets,
+    _UTF32ToUnicodeWithOffsets,
+#if U_IS_BIG_ENDIAN
+    T_UConverter_fromUnicode_UTF32_BE,
+    T_UConverter_fromUnicode_UTF32_BE_OFFSET_LOGIC,
+#else
+    T_UConverter_fromUnicode_UTF32_LE,
+    T_UConverter_fromUnicode_UTF32_LE_OFFSET_LOGIC,
+#endif
+    _UTF32GetNextUChar,
+
+    NULL, /* ### TODO implement getStarters for all Unicode encodings?! */
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+/* The 1236 CCSID refers to any version of Unicode with a BOM sensitive endianess of UTF-32 */
+static const UConverterStaticData _UTF32StaticData = {
+    sizeof(UConverterStaticData),
+    "UTF-32",
+    1236,
+    UCNV_IBM, UCNV_UTF32, 4, 4,
+#if U_IS_BIG_ENDIAN
+    { 0, 0, 0xff, 0xfd }, 4,
+#else
+    { 0xfd, 0xff, 0, 0 }, 4,
+#endif
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _UTF32Data = 
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF32StaticData, &_UTF32Impl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_u7.cpp b/thirdparty/icu4c/common/ucnv_u7.cpp
new file mode 100644
index 0000000000..87ba8cf37e
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_u7.cpp
@@ -0,0 +1,1491 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_u7.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jul01
+*   created by: Markus W. Scherer
+*
+*   UTF-7 converter implementation. Used to be in ucnv_utf.c.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "cmemory.h"
+#include "unicode/ucnv.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "uassert.h"
+
+/* UTF-7 -------------------------------------------------------------------- */
+
+/*
+ * UTF-7 is a stateful encoding of Unicode.
+ * It is defined in RFC 2152. (http://www.ietf.org/rfc/rfc2152.txt)
+ * It was intended for use in Internet email systems, using in its bytewise
+ * encoding only a subset of 7-bit US-ASCII.
+ * UTF-7 is deprecated in favor of UTF-8/16/32 and SCSU, but still
+ * occasionally used.
+ *
+ * For converting Unicode to UTF-7, the RFC allows to encode some US-ASCII
+ * characters directly or in base64. Especially, the characters in set O
+ * as defined in the RFC (see below) may be encoded directly but are not
+ * allowed in, e.g., email headers.
+ * By default, the ICU UTF-7 converter encodes set O directly.
+ * By choosing the option "version=1", set O will be escaped instead.
+ * For example:
+ *     utf7Converter=ucnv_open("UTF-7,version=1");
+ *
+ * For details about email headers see RFC 2047.
+ */
+
+/*
+ * Tests for US-ASCII characters belonging to character classes
+ * defined in UTF-7.
+ *
+ * Set D (directly encoded characters) consists of the following
+ * characters: the upper and lower case letters A through Z
+ * and a through z, the 10 digits 0-9, and the following nine special
+ * characters (note that "+" and "=" are omitted):
+ *     '(),-./:?
+ *
+ * Set O (optional direct characters) consists of the following
+ * characters (note that "\" and "~" are omitted):
+ *     !"#$%&*;<=>@[]^_`{|}
+ *
+ * According to the rules in RFC 2152, the byte values for the following
+ * US-ASCII characters are not used in UTF-7 and are therefore illegal:
+ * - all C0 control codes except for CR LF TAB
+ * - BACKSLASH
+ * - TILDE
+ * - DEL
+ * - all codes beyond US-ASCII, i.e. all >127
+ */
+#define inSetD(c) \
+    ((uint8_t)((c)-97)<26 || (uint8_t)((c)-65)<26 || /* letters */ \
+     (uint8_t)((c)-48)<10 ||    /* digits */ \
+     (uint8_t)((c)-39)<3 ||     /* '() */ \
+     (uint8_t)((c)-44)<4 ||     /* ,-./ */ \
+     (c)==58 || (c)==63         /* :? */ \
+    )
+
+#define inSetO(c) \
+    ((uint8_t)((c)-33)<6 ||         /* !"#$%& */ \
+     (uint8_t)((c)-59)<4 ||         /* ;<=> */ \
+     (uint8_t)((c)-93)<4 ||         /* ]^_` */ \
+     (uint8_t)((c)-123)<3 ||        /* {|} */ \
+     (c)==42 || (c)==64 || (c)==91  /* *@[ */ \
+    )
+
+#define isCRLFTAB(c) ((c)==13 || (c)==10 || (c)==9)
+#define isCRLFSPTAB(c) ((c)==32 || (c)==13 || (c)==10 || (c)==9)
+
+#define PLUS  43
+#define MINUS 45
+#define BACKSLASH 92
+#define TILDE 126
+
+/* legal byte values: all US-ASCII graphic characters from space to before tilde, and CR LF TAB */
+#define isLegalUTF7(c) (((uint8_t)((c)-32)<94 && (c)!=BACKSLASH) || isCRLFTAB(c))
+
+/* encode directly sets D and O and CR LF SP TAB */
+static const UBool encodeDirectlyMaximum[128]={
+ /* 0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f */
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1,
+
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0
+};
+
+/* encode directly set D and CR LF SP TAB but not set O */
+static const UBool encodeDirectlyRestricted[128]={
+ /* 0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f */
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,
+
+    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
+
+    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0
+};
+
+static const uint8_t
+toBase64[64]={
+    /* A-Z */
+    65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77,
+    78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90,
+    /* a-z */
+    97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109,
+    110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122,
+    /* 0-9 */
+    48, 49, 50, 51, 52, 53, 54, 55, 56, 57,
+    /* +/ */
+    43, 47
+};
+
+static const int8_t
+fromBase64[128]={
+    /* C0 controls, -1 for legal ones (CR LF TAB), -3 for illegal ones */
+    -3, -3, -3, -3, -3, -3, -3, -3, -3, -1, -1, -3, -3, -1, -3, -3,
+    -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3, -3,
+
+    /* general punctuation with + and / and a special value (-2) for - */
+    -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, -2, -1, 63,
+    /* digits */
+    52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -1, -1, -1,
+
+    /* A-Z */
+    -1,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14,
+    15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -3, -1, -1, -1,
+
+    /* a-z */
+    -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
+    41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -3, -3
+};
+
+/*
+ * converter status values:
+ *
+ * toUnicodeStatus:
+ *     24 inDirectMode (boolean)
+ * 23..16 base64Counter (-1..7)
+ * 15..0  bits (up to 14 bits incoming base64)
+ *
+ * fromUnicodeStatus:
+ * 31..28 version (0: set O direct  1: set O escaped)
+ *     24 inDirectMode (boolean)
+ * 23..16 base64Counter (0..2)
+ *  7..0  bits (6 bits outgoing base64)
+ *
+ */
+
+U_CDECL_BEGIN
+static void U_CALLCONV
+_UTF7Reset(UConverter *cnv, UConverterResetChoice choice) {
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode */
+        cnv->toUnicodeStatus=0x1000000; /* inDirectMode=TRUE */
+        cnv->toULength=0;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        /* reset fromUnicode */
+        cnv->fromUnicodeStatus=(cnv->fromUnicodeStatus&0xf0000000)|0x1000000; /* keep version, inDirectMode=TRUE */
+    }
+}
+
+static void U_CALLCONV
+_UTF7Open(UConverter *cnv,
+          UConverterLoadArgs *pArgs,
+          UErrorCode *pErrorCode) {
+    (void)pArgs;
+    if(UCNV_GET_VERSION(cnv)<=1) {
+        /* TODO(markus): Should just use cnv->options rather than copying the version number. */
+        cnv->fromUnicodeStatus=UCNV_GET_VERSION(cnv)<<28;
+        _UTF7Reset(cnv, UCNV_RESET_BOTH);
+    } else {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+}
+
+static void U_CALLCONV
+_UTF7ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                          UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+
+    uint8_t *bytes;
+    uint8_t byteIndex;
+
+    int32_t length, targetCapacity;
+
+    /* UTF-7 state */
+    uint16_t bits;
+    int8_t base64Counter;
+    UBool inDirectMode;
+
+    int8_t base64Value;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    uint8_t b;
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+    /* get the state machine state */
+    {
+        uint32_t status=cnv->toUnicodeStatus;
+        inDirectMode=(UBool)((status>>24)&1);
+        base64Counter=(int8_t)(status>>16);
+        bits=(uint16_t)status;
+    }
+    bytes=cnv->toUBytes;
+    byteIndex=cnv->toULength;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=byteIndex==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    if(inDirectMode) {
+directMode:
+        /*
+         * In Direct Mode, most US-ASCII characters are encoded directly, i.e.,
+         * with their US-ASCII byte values.
+         * Backslash and Tilde and most control characters are not allowed in UTF-7.
+         * A plus sign starts Unicode (or "escape") Mode.
+         *
+         * In Direct Mode, only the sourceIndex is used.
+         */
+        byteIndex=0;
+        length=(int32_t)(sourceLimit-source);
+        targetCapacity=(int32_t)(targetLimit-target);
+        if(length>targetCapacity) {
+            length=targetCapacity;
+        }
+        while(length>0) {
+            b=*source++;
+            if(!isLegalUTF7(b)) {
+                /* illegal */
+                bytes[0]=b;
+                byteIndex=1;
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                break;
+            } else if(b!=PLUS) {
+                /* write directly encoded character */
+                *target++=b;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex++;
+                }
+            } else /* PLUS */ {
+                /* switch to Unicode mode */
+                nextSourceIndex=++sourceIndex;
+                inDirectMode=FALSE;
+                byteIndex=0;
+                bits=0;
+                base64Counter=-1;
+                goto unicodeMode;
+            }
+            --length;
+        }
+        if(source<sourceLimit && target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        }
+    } else {
+unicodeMode:
+        /*
+         * In Unicode (or "escape") Mode, UTF-16BE is base64-encoded.
+         * The base64 sequence ends with any character that is not in the base64 alphabet.
+         * A terminating minus sign is consumed.
+         *
+         * In Unicode Mode, the sourceIndex has the index to the start of the current
+         * base64 bytes, while nextSourceIndex is precisely parallel to source,
+         * keeping the index to the following byte.
+         * Note that in 2 out of 3 cases, UChars overlap within a base64 byte.
+         */
+        while(source<sourceLimit) {
+            if(target<targetLimit) {
+                bytes[byteIndex++]=b=*source++;
+                ++nextSourceIndex;
+                base64Value = -3; /* initialize as illegal */
+                if(b>=126 || (base64Value=fromBase64[b])==-3 || base64Value==-1) {
+                    /* either
+                     * base64Value==-1 for any legal character except base64 and minus sign, or
+                     * base64Value==-3 for illegal characters:
+                     * 1. In either case, leave Unicode mode.
+                     * 2.1. If we ended with an incomplete UChar or none after the +, then
+                     *      generate an error for the preceding erroneous sequence and deal with
+                     *      the current (possibly illegal) character next time through.
+                     * 2.2. Else the current char comes after a complete UChar, which was already
+                     *      pushed to the output buf, so:
+                     * 2.2.1. If the current char is legal, just save it for processing next time.
+                     *        It may be for example, a plus which we need to deal with in direct mode.
+                     * 2.2.2. Else if the current char is illegal, we might as well deal with it here.
+                     */
+                    inDirectMode=TRUE;
+                    if(base64Counter==-1) {
+                        /* illegal: + immediately followed by something other than base64 or minus sign */
+                        /* include the plus sign in the reported sequence, but not the subsequent char */
+                        --source;
+                        bytes[0]=PLUS;
+                        byteIndex=1;
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    } else if(bits!=0) {
+                        /* bits are illegally left over, a UChar is incomplete */
+                        /* don't include current char (legal or illegal) in error seq */
+                        --source;
+                        --byteIndex;
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    } else {
+                        /* previous UChar was complete */
+                        if(base64Value==-3) {
+                            /* current character is illegal, deal with it here */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            break;
+                        } else {
+                            /* un-read the current character in case it is a plus sign */
+                            --source;
+                            sourceIndex=nextSourceIndex-1;
+                            goto directMode;
+                        }
+                    }
+                } else if(base64Value>=0) {
+                    /* collect base64 bytes into UChars */
+                    switch(base64Counter) {
+                    case -1: /* -1 is immediately after the + */
+                    case 0:
+                        bits=base64Value;
+                        base64Counter=1;
+                        break;
+                    case 1:
+                    case 3:
+                    case 4:
+                    case 6:
+                        bits=(uint16_t)((bits<<6)|base64Value);
+                        ++base64Counter;
+                        break;
+                    case 2:
+                        *target++=(UChar)((bits<<4)|(base64Value>>2));
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex-1;
+                        }
+                        bytes[0]=b; /* keep this byte in case an error occurs */
+                        byteIndex=1;
+                        bits=(uint16_t)(base64Value&3);
+                        base64Counter=3;
+                        break;
+                    case 5:
+                        *target++=(UChar)((bits<<2)|(base64Value>>4));
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex-1;
+                        }
+                        bytes[0]=b; /* keep this byte in case an error occurs */
+                        byteIndex=1;
+                        bits=(uint16_t)(base64Value&15);
+                        base64Counter=6;
+                        break;
+                    case 7:
+                        *target++=(UChar)((bits<<6)|base64Value);
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex;
+                        }
+                        byteIndex=0;
+                        bits=0;
+                        base64Counter=0;
+                        break;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                } else /*base64Value==-2*/ {
+                    /* minus sign terminates the base64 sequence */
+                    inDirectMode=TRUE;
+                    if(base64Counter==-1) {
+                        /* +- i.e. a minus immediately following a plus */
+                        *target++=PLUS;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex-1;
+                        }
+                    } else {
+                        /* absorb the minus and leave the Unicode Mode */
+                        if(bits!=0) {
+                            /* bits are illegally left over, a UChar is incomplete */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            break;
+                        }
+                    }
+                    sourceIndex=nextSourceIndex;
+                    goto directMode;
+                }
+            } else {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode) && pArgs->flush && source==sourceLimit && bits==0) {
+        /*
+         * if we are in Unicode mode, then the byteIndex might not be 0,
+         * but that is ok if bits==0
+         * -> we set byteIndex=0 at the end of the stream to avoid a truncated error
+         * (not true for IMAP-mailbox-name where we must end in direct mode)
+         */
+        byteIndex=0;
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->toUnicodeStatus=((uint32_t)inDirectMode<<24)|((uint32_t)((uint8_t)base64Counter)<<16)|(uint32_t)bits;
+    cnv->toULength=byteIndex;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+    return;
+}
+
+static void U_CALLCONV
+_UTF7FromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target, *targetLimit;
+    int32_t *offsets;
+
+    int32_t length, targetCapacity, sourceIndex;
+    UChar c;
+
+    /* UTF-7 state */
+    const UBool *encodeDirectly;
+    uint8_t bits;
+    int8_t base64Counter;
+    UBool inDirectMode;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetLimit=(uint8_t *)pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    /* get the state machine state */
+    {
+        uint32_t status=cnv->fromUnicodeStatus;
+        encodeDirectly= status<0x10000000 ? encodeDirectlyMaximum : encodeDirectlyRestricted;
+        inDirectMode=(UBool)((status>>24)&1);
+        base64Counter=(int8_t)(status>>16);
+        bits=(uint8_t)status;
+        U_ASSERT(bits<=UPRV_LENGTHOF(toBase64));
+    }
+
+    /* UTF-7 always encodes UTF-16 code units, therefore we need only a simple sourceIndex */
+    sourceIndex=0;
+
+    if(inDirectMode) {
+directMode:
+        length=(int32_t)(sourceLimit-source);
+        targetCapacity=(int32_t)(targetLimit-target);
+        if(length>targetCapacity) {
+            length=targetCapacity;
+        }
+        while(length>0) {
+            c=*source++;
+            /* currently always encode CR LF SP TAB directly */
+            if(c<=127 && encodeDirectly[c]) {
+                /* encode directly */
+                *target++=(uint8_t)c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex++;
+                }
+            } else if(c==PLUS) {
+                /* output +- for + */
+                *target++=PLUS;
+                if(target<targetLimit) {
+                    *target++=MINUS;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                        *offsets++=sourceIndex++;
+                    }
+                    /* realign length and targetCapacity */
+                    goto directMode;
+                } else {
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex++;
+                    }
+                    cnv->charErrorBuffer[0]=MINUS;
+                    cnv->charErrorBufferLength=1;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            } else {
+                /* un-read this character and switch to Unicode Mode */
+                --source;
+                *target++=PLUS;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                inDirectMode=FALSE;
+                base64Counter=0;
+                goto unicodeMode;
+            }
+            --length;
+        }
+        if(source<sourceLimit && target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        }
+    } else {
+unicodeMode:
+        while(source<sourceLimit) {
+            if(target<targetLimit) {
+                c=*source++;
+                if(c<=127 && encodeDirectly[c]) {
+                    /* encode directly */
+                    inDirectMode=TRUE;
+
+                    /* trick: back out this character to make this easier */
+                    --source;
+
+                    /* terminate the base64 sequence */
+                    if(base64Counter!=0) {
+                        /* write remaining bits for the previous character */
+                        *target++=toBase64[bits];
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex-1;
+                        }
+                    }
+                    if(fromBase64[c]!=-1) {
+                        /* need to terminate with a minus */
+                        if(target<targetLimit) {
+                            *target++=MINUS;
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex-1;
+                            }
+                        } else {
+                            cnv->charErrorBuffer[0]=MINUS;
+                            cnv->charErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            break;
+                        }
+                    }
+                    goto directMode;
+                } else {
+                    /*
+                     * base64 this character:
+                     * Output 2 or 3 base64 bytes for the remaining bits of the previous character
+                     * and the bits of this character, each implicitly in UTF-16BE.
+                     *
+                     * Here, bits is an 8-bit variable because only 6 bits need to be kept from one
+                     * character to the next. The actual 2 or 4 bits are shifted to the left edge
+                     * of the 6-bits field 5..0 to make the termination of the base64 sequence easier.
+                     */
+                    switch(base64Counter) {
+                    case 0:
+                        *target++=toBase64[c>>10];
+                        if(target<targetLimit) {
+                            *target++=toBase64[(c>>4)&0x3f];
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                *offsets++=sourceIndex++;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            cnv->charErrorBuffer[0]=toBase64[(c>>4)&0x3f];
+                            cnv->charErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=(uint8_t)((c&15)<<2);
+                        base64Counter=1;
+                        break;
+                    case 1:
+                        *target++=toBase64[bits|(c>>14)];
+                        if(target<targetLimit) {
+                            *target++=toBase64[(c>>8)&0x3f];
+                            if(target<targetLimit) {
+                                *target++=toBase64[(c>>2)&0x3f];
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                            } else {
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                                cnv->charErrorBuffer[0]=toBase64[(c>>2)&0x3f];
+                                cnv->charErrorBufferLength=1;
+                                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            cnv->charErrorBuffer[0]=toBase64[(c>>8)&0x3f];
+                            cnv->charErrorBuffer[1]=toBase64[(c>>2)&0x3f];
+                            cnv->charErrorBufferLength=2;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=(uint8_t)((c&3)<<4);
+                        base64Counter=2;
+                        break;
+                    case 2:
+                        *target++=toBase64[bits|(c>>12)];
+                        if(target<targetLimit) {
+                            *target++=toBase64[(c>>6)&0x3f];
+                            if(target<targetLimit) {
+                                *target++=toBase64[c&0x3f];
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                            } else {
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                                cnv->charErrorBuffer[0]=toBase64[c&0x3f];
+                                cnv->charErrorBufferLength=1;
+                                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            cnv->charErrorBuffer[0]=toBase64[(c>>6)&0x3f];
+                            cnv->charErrorBuffer[1]=toBase64[c&0x3f];
+                            cnv->charErrorBufferLength=2;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=0;
+                        base64Counter=0;
+                        break;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                }
+            } else {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+    }
+
+    if(pArgs->flush && source>=sourceLimit) {
+        /* flush remaining bits to the target */
+        if(!inDirectMode) {
+            if (base64Counter!=0) {
+                if(target<targetLimit) {
+                    *target++=toBase64[bits];
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex-1;
+                    }
+                } else {
+                    cnv->charErrorBuffer[cnv->charErrorBufferLength++]=toBase64[bits];
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                }
+            }
+            /* Add final MINUS to terminate unicodeMode */
+            if(target<targetLimit) {
+                *target++=MINUS;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex-1;
+                }
+            } else {
+                cnv->charErrorBuffer[cnv->charErrorBufferLength++]=MINUS;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        /* reset the state for the next conversion */
+        cnv->fromUnicodeStatus=(cnv->fromUnicodeStatus&0xf0000000)|0x1000000; /* keep version, inDirectMode=TRUE */
+    } else {
+        /* set the converter state back into UConverter */
+        cnv->fromUnicodeStatus=
+            (cnv->fromUnicodeStatus&0xf0000000)|    /* keep version*/
+            ((uint32_t)inDirectMode<<24)|((uint32_t)base64Counter<<16)|(uint32_t)bits;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+    return;
+}
+
+static const char * U_CALLCONV
+_UTF7GetName(const UConverter *cnv) {
+    switch(cnv->fromUnicodeStatus>>28) {
+    case 1:
+        return "UTF-7,version=1";
+    default:
+        return "UTF-7";
+    }
+}
+U_CDECL_END
+
+static const UConverterImpl _UTF7Impl={
+    UCNV_UTF7,
+
+    NULL,
+    NULL,
+
+    _UTF7Open,
+    NULL,
+    _UTF7Reset,
+
+    _UTF7ToUnicodeWithOffsets,
+    _UTF7ToUnicodeWithOffsets,
+    _UTF7FromUnicodeWithOffsets,
+    _UTF7FromUnicodeWithOffsets,
+    NULL,
+
+    NULL,
+    _UTF7GetName,
+    NULL, /* we don't need writeSub() because we never call a callback at fromUnicode() */
+    NULL,
+    ucnv_getCompleteUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _UTF7StaticData={
+    sizeof(UConverterStaticData),
+    "UTF-7",
+    0, /* TODO CCSID for UTF-7 */
+    UCNV_IBM, UCNV_UTF7,
+    1, 4,
+    { 0x3f, 0, 0, 0 }, 1, /* the subchar is not used */
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _UTF7Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF7StaticData, &_UTF7Impl);
+
+/* IMAP mailbox name encoding ----------------------------------------------- */
+
+/*
+ * RFC 2060: INTERNET MESSAGE ACCESS PROTOCOL - VERSION 4rev1
+ * http://www.ietf.org/rfc/rfc2060.txt
+ *
+ * 5.1.3.  Mailbox International Naming Convention
+ *
+ * By convention, international mailbox names are specified using a
+ * modified version of the UTF-7 encoding described in [UTF-7].  The
+ * purpose of these modifications is to correct the following problems
+ * with UTF-7:
+ *
+ *    1) UTF-7 uses the "+" character for shifting; this conflicts with
+ *       the common use of "+" in mailbox names, in particular USENET
+ *       newsgroup names.
+ *
+ *    2) UTF-7's encoding is BASE64 which uses the "/" character; this
+ *       conflicts with the use of "/" as a popular hierarchy delimiter.
+ *
+ *    3) UTF-7 prohibits the unencoded usage of "\"; this conflicts with
+ *       the use of "\" as a popular hierarchy delimiter.
+ *
+ *    4) UTF-7 prohibits the unencoded usage of "~"; this conflicts with
+ *       the use of "~" in some servers as a home directory indicator.
+ *
+ *    5) UTF-7 permits multiple alternate forms to represent the same
+ *       string; in particular, printable US-ASCII chararacters can be
+ *       represented in encoded form.
+ *
+ * In modified UTF-7, printable US-ASCII characters except for "&"
+ * represent themselves; that is, characters with octet values 0x20-0x25
+ * and 0x27-0x7e.  The character "&" (0x26) is represented by the two-
+ * octet sequence "&-".
+ *
+ * All other characters (octet values 0x00-0x1f, 0x7f-0xff, and all
+ * Unicode 16-bit octets) are represented in modified BASE64, with a
+ * further modification from [UTF-7] that "," is used instead of "/".
+ * Modified BASE64 MUST NOT be used to represent any printing US-ASCII
+ * character which can represent itself.
+ *
+ * "&" is used to shift to modified BASE64 and "-" to shift back to US-
+ * ASCII.  All names start in US-ASCII, and MUST end in US-ASCII (that
+ * is, a name that ends with a Unicode 16-bit octet MUST end with a "-
+ * ").
+ *
+ * For example, here is a mailbox name which mixes English, Japanese,
+ * and Chinese text: ~peter/mail/&ZeVnLIqe-/&U,BTFw-
+ */
+
+/*
+ * Tests for US-ASCII characters belonging to character classes
+ * defined in UTF-7.
+ *
+ * Set D (directly encoded characters) consists of the following
+ * characters: the upper and lower case letters A through Z
+ * and a through z, the 10 digits 0-9, and the following nine special
+ * characters (note that "+" and "=" are omitted):
+ *     '(),-./:?
+ *
+ * Set O (optional direct characters) consists of the following
+ * characters (note that "\" and "~" are omitted):
+ *     !"#$%&*;<=>@[]^_`{|}
+ *
+ * According to the rules in RFC 2152, the byte values for the following
+ * US-ASCII characters are not used in UTF-7 and are therefore illegal:
+ * - all C0 control codes except for CR LF TAB
+ * - BACKSLASH
+ * - TILDE
+ * - DEL
+ * - all codes beyond US-ASCII, i.e. all >127
+ */
+
+/* uses '&' not '+' to start a base64 sequence */
+#define AMPERSAND 0x26
+#define COMMA 0x2c
+#define SLASH 0x2f
+
+/* legal byte values: all US-ASCII graphic characters 0x20..0x7e */
+#define isLegalIMAP(c) (0x20<=(c) && (c)<=0x7e)
+
+/* direct-encode all of printable ASCII 0x20..0x7e except '&' 0x26 */
+#define inSetDIMAP(c) (isLegalIMAP(c) && c!=AMPERSAND)
+
+#define TO_BASE64_IMAP(n) ((n)<63 ? toBase64[n] : COMMA)
+#define FROM_BASE64_IMAP(c) ((c)==COMMA ? 63 : (c)==SLASH ? -1 : fromBase64[c])
+
+/*
+ * converter status values:
+ *
+ * toUnicodeStatus:
+ *     24 inDirectMode (boolean)
+ * 23..16 base64Counter (-1..7)
+ * 15..0  bits (up to 14 bits incoming base64)
+ *
+ * fromUnicodeStatus:
+ *     24 inDirectMode (boolean)
+ * 23..16 base64Counter (0..2)
+ *  7..0  bits (6 bits outgoing base64)
+ *
+ * ignore bits 31..25
+ */
+
+U_CDECL_BEGIN
+static void U_CALLCONV
+_IMAPToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                          UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+
+    uint8_t *bytes;
+    uint8_t byteIndex;
+
+    int32_t length, targetCapacity;
+
+    /* UTF-7 state */
+    uint16_t bits;
+    int8_t base64Counter;
+    UBool inDirectMode;
+
+    int8_t base64Value;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    UChar c;
+    uint8_t b;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+    /* get the state machine state */
+    {
+        uint32_t status=cnv->toUnicodeStatus;
+        inDirectMode=(UBool)((status>>24)&1);
+        base64Counter=(int8_t)(status>>16);
+        bits=(uint16_t)status;
+    }
+    bytes=cnv->toUBytes;
+    byteIndex=cnv->toULength;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=byteIndex==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    if(inDirectMode) {
+directMode:
+        /*
+         * In Direct Mode, US-ASCII characters are encoded directly, i.e.,
+         * with their US-ASCII byte values.
+         * An ampersand starts Unicode (or "escape") Mode.
+         *
+         * In Direct Mode, only the sourceIndex is used.
+         */
+        byteIndex=0;
+        length=(int32_t)(sourceLimit-source);
+        targetCapacity=(int32_t)(targetLimit-target);
+        if(length>targetCapacity) {
+            length=targetCapacity;
+        }
+        while(length>0) {
+            b=*source++;
+            if(!isLegalIMAP(b)) {
+                /* illegal */
+                bytes[0]=b;
+                byteIndex=1;
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                break;
+            } else if(b!=AMPERSAND) {
+                /* write directly encoded character */
+                *target++=b;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex++;
+                }
+            } else /* AMPERSAND */ {
+                /* switch to Unicode mode */
+                nextSourceIndex=++sourceIndex;
+                inDirectMode=FALSE;
+                byteIndex=0;
+                bits=0;
+                base64Counter=-1;
+                goto unicodeMode;
+            }
+            --length;
+        }
+        if(source<sourceLimit && target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        }
+    } else {
+unicodeMode:
+        /*
+         * In Unicode (or "escape") Mode, UTF-16BE is base64-encoded.
+         * The base64 sequence ends with any character that is not in the base64 alphabet.
+         * A terminating minus sign is consumed.
+         * US-ASCII must not be base64-ed.
+         *
+         * In Unicode Mode, the sourceIndex has the index to the start of the current
+         * base64 bytes, while nextSourceIndex is precisely parallel to source,
+         * keeping the index to the following byte.
+         * Note that in 2 out of 3 cases, UChars overlap within a base64 byte.
+         */
+        while(source<sourceLimit) {
+            if(target<targetLimit) {
+                bytes[byteIndex++]=b=*source++;
+                ++nextSourceIndex;
+                if(b>0x7e) {
+                    /* illegal - test other illegal US-ASCII values by base64Value==-3 */
+                    inDirectMode=TRUE;
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    break;
+                } else if((base64Value=FROM_BASE64_IMAP(b))>=0) {
+                    /* collect base64 bytes into UChars */
+                    switch(base64Counter) {
+                    case -1: /* -1 is immediately after the & */
+                    case 0:
+                        bits=base64Value;
+                        base64Counter=1;
+                        break;
+                    case 1:
+                    case 3:
+                    case 4:
+                    case 6:
+                        bits=(uint16_t)((bits<<6)|base64Value);
+                        ++base64Counter;
+                        break;
+                    case 2:
+                        c=(UChar)((bits<<4)|(base64Value>>2));
+                        if(isLegalIMAP(c)) {
+                            /* illegal */
+                            inDirectMode=TRUE;
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                        *target++=c;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex-1;
+                        }
+                        bytes[0]=b; /* keep this byte in case an error occurs */
+                        byteIndex=1;
+                        bits=(uint16_t)(base64Value&3);
+                        base64Counter=3;
+                        break;
+                    case 5:
+                        c=(UChar)((bits<<2)|(base64Value>>4));
+                        if(isLegalIMAP(c)) {
+                            /* illegal */
+                            inDirectMode=TRUE;
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                        *target++=c;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex-1;
+                        }
+                        bytes[0]=b; /* keep this byte in case an error occurs */
+                        byteIndex=1;
+                        bits=(uint16_t)(base64Value&15);
+                        base64Counter=6;
+                        break;
+                    case 7:
+                        c=(UChar)((bits<<6)|base64Value);
+                        if(isLegalIMAP(c)) {
+                            /* illegal */
+                            inDirectMode=TRUE;
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                        *target++=c;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                            sourceIndex=nextSourceIndex;
+                        }
+                        byteIndex=0;
+                        bits=0;
+                        base64Counter=0;
+                        break;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                } else if(base64Value==-2) {
+                    /* minus sign terminates the base64 sequence */
+                    inDirectMode=TRUE;
+                    if(base64Counter==-1) {
+                        /* &- i.e. a minus immediately following an ampersand */
+                        *target++=AMPERSAND;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex-1;
+                        }
+                    } else {
+                        /* absorb the minus and leave the Unicode Mode */
+                        if(bits!=0 || (base64Counter!=0 && base64Counter!=3 && base64Counter!=6)) {
+                            /* bits are illegally left over, a UChar is incomplete */
+                            /* base64Counter other than 0, 3, 6 means non-minimal zero-padding, also illegal */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            break;
+                        }
+                    }
+                    sourceIndex=nextSourceIndex;
+                    goto directMode;
+                } else {
+                    if(base64Counter==-1) {
+                        /* illegal: & immediately followed by something other than base64 or minus sign */
+                        /* include the ampersand in the reported sequence */
+                        --sourceIndex;
+                        bytes[0]=AMPERSAND;
+                        bytes[1]=b;
+                        byteIndex=2;
+                    }
+                    /* base64Value==-1 for characters that are illegal only in Unicode mode */
+                    /* base64Value==-3 for illegal characters */
+                    /* illegal */
+                    inDirectMode=TRUE;
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    break;
+                }
+            } else {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+    }
+endloop:
+
+    /*
+     * the end of the input stream and detection of truncated input
+     * are handled by the framework, but here we must check if we are in Unicode
+     * mode and byteIndex==0 because we must end in direct mode
+     *
+     * conditions:
+     *   successful
+     *   in Unicode mode and byteIndex==0
+     *   end of input and no truncated input
+     */
+    if( U_SUCCESS(*pErrorCode) &&
+        !inDirectMode && byteIndex==0 &&
+        pArgs->flush && source>=sourceLimit
+    ) {
+        if(base64Counter==-1) {
+            /* & at the very end of the input */
+            /* make the ampersand the reported sequence */
+            bytes[0]=AMPERSAND;
+            byteIndex=1;
+        }
+        /* else if(base64Counter!=-1) byteIndex remains 0 because there is no particular byte sequence */
+
+        inDirectMode=TRUE; /* avoid looping */
+        *pErrorCode=U_TRUNCATED_CHAR_FOUND;
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->toUnicodeStatus=((uint32_t)inDirectMode<<24)|((uint32_t)((uint8_t)base64Counter)<<16)|(uint32_t)bits;
+    cnv->toULength=byteIndex;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+    return;
+}
+
+static void U_CALLCONV
+_IMAPFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target, *targetLimit;
+    int32_t *offsets;
+
+    int32_t length, targetCapacity, sourceIndex;
+    UChar c;
+    uint8_t b;
+
+    /* UTF-7 state */
+    uint8_t bits;
+    int8_t base64Counter;
+    UBool inDirectMode;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetLimit=(uint8_t *)pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    /* get the state machine state */
+    {
+        uint32_t status=cnv->fromUnicodeStatus;
+        inDirectMode=(UBool)((status>>24)&1);
+        base64Counter=(int8_t)(status>>16);
+        bits=(uint8_t)status;
+    }
+
+    /* UTF-7 always encodes UTF-16 code units, therefore we need only a simple sourceIndex */
+    sourceIndex=0;
+
+    if(inDirectMode) {
+directMode:
+        length=(int32_t)(sourceLimit-source);
+        targetCapacity=(int32_t)(targetLimit-target);
+        if(length>targetCapacity) {
+            length=targetCapacity;
+        }
+        while(length>0) {
+            c=*source++;
+            /* encode 0x20..0x7e except '&' directly */
+            if(inSetDIMAP(c)) {
+                /* encode directly */
+                *target++=(uint8_t)c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex++;
+                }
+            } else if(c==AMPERSAND) {
+                /* output &- for & */
+                *target++=AMPERSAND;
+                if(target<targetLimit) {
+                    *target++=MINUS;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                        *offsets++=sourceIndex++;
+                    }
+                    /* realign length and targetCapacity */
+                    goto directMode;
+                } else {
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex++;
+                    }
+                    cnv->charErrorBuffer[0]=MINUS;
+                    cnv->charErrorBufferLength=1;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            } else {
+                /* un-read this character and switch to Unicode Mode */
+                --source;
+                *target++=AMPERSAND;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                inDirectMode=FALSE;
+                base64Counter=0;
+                goto unicodeMode;
+            }
+            --length;
+        }
+        if(source<sourceLimit && target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        }
+    } else {
+unicodeMode:
+        while(source<sourceLimit) {
+            if(target<targetLimit) {
+                c=*source++;
+                if(isLegalIMAP(c)) {
+                    /* encode directly */
+                    inDirectMode=TRUE;
+
+                    /* trick: back out this character to make this easier */
+                    --source;
+
+                    /* terminate the base64 sequence */
+                    if(base64Counter!=0) {
+                        /* write remaining bits for the previous character */
+                        *target++=TO_BASE64_IMAP(bits);
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex-1;
+                        }
+                    }
+                    /* need to terminate with a minus */
+                    if(target<targetLimit) {
+                        *target++=MINUS;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex-1;
+                        }
+                    } else {
+                        cnv->charErrorBuffer[0]=MINUS;
+                        cnv->charErrorBufferLength=1;
+                        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        break;
+                    }
+                    goto directMode;
+                } else {
+                    /*
+                     * base64 this character:
+                     * Output 2 or 3 base64 bytes for the remaining bits of the previous character
+                     * and the bits of this character, each implicitly in UTF-16BE.
+                     *
+                     * Here, bits is an 8-bit variable because only 6 bits need to be kept from one
+                     * character to the next. The actual 2 or 4 bits are shifted to the left edge
+                     * of the 6-bits field 5..0 to make the termination of the base64 sequence easier.
+                     */
+                    switch(base64Counter) {
+                    case 0:
+                        b=(uint8_t)(c>>10);
+                        *target++=TO_BASE64_IMAP(b);
+                        if(target<targetLimit) {
+                            b=(uint8_t)((c>>4)&0x3f);
+                            *target++=TO_BASE64_IMAP(b);
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                *offsets++=sourceIndex++;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            b=(uint8_t)((c>>4)&0x3f);
+                            cnv->charErrorBuffer[0]=TO_BASE64_IMAP(b);
+                            cnv->charErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=(uint8_t)((c&15)<<2);
+                        base64Counter=1;
+                        break;
+                    case 1:
+                        b=(uint8_t)(bits|(c>>14));
+                        *target++=TO_BASE64_IMAP(b);
+                        if(target<targetLimit) {
+                            b=(uint8_t)((c>>8)&0x3f);
+                            *target++=TO_BASE64_IMAP(b);
+                            if(target<targetLimit) {
+                                b=(uint8_t)((c>>2)&0x3f);
+                                *target++=TO_BASE64_IMAP(b);
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                            } else {
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                                b=(uint8_t)((c>>2)&0x3f);
+                                cnv->charErrorBuffer[0]=TO_BASE64_IMAP(b);
+                                cnv->charErrorBufferLength=1;
+                                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            b=(uint8_t)((c>>8)&0x3f);
+                            cnv->charErrorBuffer[0]=TO_BASE64_IMAP(b);
+                            b=(uint8_t)((c>>2)&0x3f);
+                            cnv->charErrorBuffer[1]=TO_BASE64_IMAP(b);
+                            cnv->charErrorBufferLength=2;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=(uint8_t)((c&3)<<4);
+                        base64Counter=2;
+                        break;
+                    case 2:
+                        b=(uint8_t)(bits|(c>>12));
+                        *target++=TO_BASE64_IMAP(b);
+                        if(target<targetLimit) {
+                            b=(uint8_t)((c>>6)&0x3f);
+                            *target++=TO_BASE64_IMAP(b);
+                            if(target<targetLimit) {
+                                b=(uint8_t)(c&0x3f);
+                                *target++=TO_BASE64_IMAP(b);
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                            } else {
+                                if(offsets!=NULL) {
+                                    *offsets++=sourceIndex;
+                                    *offsets++=sourceIndex++;
+                                }
+                                b=(uint8_t)(c&0x3f);
+                                cnv->charErrorBuffer[0]=TO_BASE64_IMAP(b);
+                                cnv->charErrorBufferLength=1;
+                                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            }
+                        } else {
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex++;
+                            }
+                            b=(uint8_t)((c>>6)&0x3f);
+                            cnv->charErrorBuffer[0]=TO_BASE64_IMAP(b);
+                            b=(uint8_t)(c&0x3f);
+                            cnv->charErrorBuffer[1]=TO_BASE64_IMAP(b);
+                            cnv->charErrorBufferLength=2;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                        }
+                        bits=0;
+                        base64Counter=0;
+                        break;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                }
+            } else {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+    }
+
+    if(pArgs->flush && source>=sourceLimit) {
+        /* flush remaining bits to the target */
+        if(!inDirectMode) {
+            if(base64Counter!=0) {
+                if(target<targetLimit) {
+                    *target++=TO_BASE64_IMAP(bits);
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex-1;
+                    }
+                } else {
+                    cnv->charErrorBuffer[cnv->charErrorBufferLength++]=TO_BASE64_IMAP(bits);
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                }
+            }
+            /* need to terminate with a minus */
+            if(target<targetLimit) {
+                *target++=MINUS;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex-1;
+                }
+            } else {
+                cnv->charErrorBuffer[cnv->charErrorBufferLength++]=MINUS;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        /* reset the state for the next conversion */
+        cnv->fromUnicodeStatus=(cnv->fromUnicodeStatus&0xf0000000)|0x1000000; /* keep version, inDirectMode=TRUE */
+    } else {
+        /* set the converter state back into UConverter */
+        cnv->fromUnicodeStatus=
+            (cnv->fromUnicodeStatus&0xf0000000)|    /* keep version*/
+            ((uint32_t)inDirectMode<<24)|((uint32_t)base64Counter<<16)|(uint32_t)bits;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+    return;
+}
+U_CDECL_END
+
+static const UConverterImpl _IMAPImpl={
+    UCNV_IMAP_MAILBOX,
+
+    NULL,
+    NULL,
+
+    _UTF7Open,
+    NULL,
+    _UTF7Reset,
+
+    _IMAPToUnicodeWithOffsets,
+    _IMAPToUnicodeWithOffsets,
+    _IMAPFromUnicodeWithOffsets,
+    _IMAPFromUnicodeWithOffsets,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL, /* we don't need writeSub() because we never call a callback at fromUnicode() */
+    NULL,
+    ucnv_getCompleteUnicodeSet,
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _IMAPStaticData={
+    sizeof(UConverterStaticData),
+    "IMAP-mailbox-name",
+    0, /* TODO CCSID for IMAP-mailbox-name */
+    UCNV_IBM, UCNV_IMAP_MAILBOX,
+    1, 4,
+    { 0x3f, 0, 0, 0 }, 1, /* the subchar is not used */
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _IMAPData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_IMAPStaticData, &_IMAPImpl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnv_u8.cpp b/thirdparty/icu4c/common/ucnv_u8.cpp
new file mode 100644
index 0000000000..1ef7fa2f02
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnv_u8.cpp
@@ -0,0 +1,944 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnv_u8.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jul01
+*   created by: Markus W. Scherer
+*
+*   UTF-8 converter implementation. Used to be in ucnv_utf.c.
+*
+*   Also, CESU-8 implementation, see UTR 26.
+*   The CESU-8 converter uses all the same functions as the
+*   UTF-8 converter, with a branch for converting supplementary code points.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "uassert.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+#include "ustr_imp.h"
+
+/* Prototypes --------------------------------------------------------------- */
+
+/* Keep these here to make finicky compilers happy */
+
+U_CFUNC void ucnv_fromUnicode_UTF8(UConverterFromUnicodeArgs *args,
+                                           UErrorCode *err);
+U_CFUNC void ucnv_fromUnicode_UTF8_OFFSETS_LOGIC(UConverterFromUnicodeArgs *args,
+                                                        UErrorCode *err);
+
+
+/* UTF-8 -------------------------------------------------------------------- */
+
+#define MAXIMUM_UCS2            0x0000FFFF
+
+static const uint32_t offsetsFromUTF8[5] = {0,
+  (uint32_t) 0x00000000, (uint32_t) 0x00003080, (uint32_t) 0x000E2080,
+  (uint32_t) 0x03C82080
+};
+
+static UBool hasCESU8Data(const UConverter *cnv)
+{
+#if UCONFIG_ONLY_HTML_CONVERSION
+    return FALSE;
+#else
+    return (UBool)(cnv->sharedData == &_CESU8Data);
+#endif
+}
+U_CDECL_BEGIN
+static void  U_CALLCONV ucnv_toUnicode_UTF8 (UConverterToUnicodeArgs * args,
+                                  UErrorCode * err)
+{
+    UConverter *cnv = args->converter;
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = cnv->toUBytes;
+    UBool isCESU8 = hasCESU8Data(cnv);
+    uint32_t ch, ch2 = 0;
+    int32_t i, inBytes;
+
+    /* Restore size of current sequence */
+    if (cnv->toULength > 0 && myTarget < targetLimit)
+    {
+        inBytes = cnv->mode;            /* restore # of bytes to consume */
+        i = cnv->toULength;             /* restore # of bytes consumed */
+        cnv->toULength = 0;
+
+        ch = cnv->toUnicodeStatus;/*Stores the previously calculated ch from a previous call*/
+        cnv->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+        if (U8_IS_SINGLE(ch))        /* Simple case */
+        {
+            *(myTarget++) = (UChar) ch;
+        }
+        else
+        {
+            /* store the first char */
+            toUBytes[0] = (char)ch;
+            inBytes = U8_COUNT_BYTES_NON_ASCII(ch); /* lookup current sequence length */
+            i = 1;
+
+morebytes:
+            while (i < inBytes)
+            {
+                if (mySource < sourceLimit)
+                {
+                    toUBytes[i] = (char) (ch2 = *mySource);
+                    if (!icu::UTF8::isValidTrail(ch, static_cast<uint8_t>(ch2), i, inBytes) &&
+                            !(isCESU8 && i == 1 && ch == 0xed && U8_IS_TRAIL(ch2)))
+                    {
+                        break; /* i < inBytes */
+                    }
+                    ch = (ch << 6) + ch2;
+                    ++mySource;
+                    i++;
+                }
+                else
+                {
+                    /* stores a partially calculated target*/
+                    cnv->toUnicodeStatus = ch;
+                    cnv->mode = inBytes;
+                    cnv->toULength = (int8_t) i;
+                    goto donefornow;
+                }
+            }
+
+            // In CESU-8, only surrogates, not supplementary code points, are encoded directly.
+            if (i == inBytes && (!isCESU8 || i <= 3))
+            {
+                /* Remove the accumulated high bits */
+                ch -= offsetsFromUTF8[inBytes];
+
+                /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+                if (ch <= MAXIMUM_UCS2) 
+                {
+                    /* fits in 16 bits */
+                    *(myTarget++) = (UChar) ch;
+                }
+                else
+                {
+                    /* write out the surrogates */
+                    *(myTarget++) = U16_LEAD(ch);
+                    ch = U16_TRAIL(ch);
+                    if (myTarget < targetLimit)
+                    {
+                        *(myTarget++) = (UChar)ch;
+                    }
+                    else
+                    {
+                        /* Put in overflow buffer (not handled here) */
+                        cnv->UCharErrorBuffer[0] = (UChar) ch;
+                        cnv->UCharErrorBufferLength = 1;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                        break;
+                    }
+                }
+            }
+            else
+            {
+                cnv->toULength = (int8_t)i;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+}
+
+static void  U_CALLCONV ucnv_toUnicode_UTF8_OFFSETS_LOGIC (UConverterToUnicodeArgs * args,
+                                                UErrorCode * err)
+{
+    UConverter *cnv = args->converter;
+    const unsigned char *mySource = (unsigned char *) args->source;
+    UChar *myTarget = args->target;
+    int32_t *myOffsets = args->offsets;
+    int32_t offsetNum = 0;
+    const unsigned char *sourceLimit = (unsigned char *) args->sourceLimit;
+    const UChar *targetLimit = args->targetLimit;
+    unsigned char *toUBytes = cnv->toUBytes;
+    UBool isCESU8 = hasCESU8Data(cnv);
+    uint32_t ch, ch2 = 0;
+    int32_t i, inBytes;
+
+    /* Restore size of current sequence */
+    if (cnv->toULength > 0 && myTarget < targetLimit)
+    {
+        inBytes = cnv->mode;            /* restore # of bytes to consume */
+        i = cnv->toULength;             /* restore # of bytes consumed */
+        cnv->toULength = 0;
+
+        ch = cnv->toUnicodeStatus;/*Stores the previously calculated ch from a previous call*/
+        cnv->toUnicodeStatus = 0;
+        goto morebytes;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+        if (U8_IS_SINGLE(ch))        /* Simple case */
+        {
+            *(myTarget++) = (UChar) ch;
+            *(myOffsets++) = offsetNum++;
+        }
+        else
+        {
+            toUBytes[0] = (char)ch;
+            inBytes = U8_COUNT_BYTES_NON_ASCII(ch);
+            i = 1;
+
+morebytes:
+            while (i < inBytes)
+            {
+                if (mySource < sourceLimit)
+                {
+                    toUBytes[i] = (char) (ch2 = *mySource);
+                    if (!icu::UTF8::isValidTrail(ch, static_cast<uint8_t>(ch2), i, inBytes) &&
+                            !(isCESU8 && i == 1 && ch == 0xed && U8_IS_TRAIL(ch2)))
+                    {
+                        break; /* i < inBytes */
+                    }
+                    ch = (ch << 6) + ch2;
+                    ++mySource;
+                    i++;
+                }
+                else
+                {
+                    cnv->toUnicodeStatus = ch;
+                    cnv->mode = inBytes;
+                    cnv->toULength = (int8_t)i;
+                    goto donefornow;
+                }
+            }
+
+            // In CESU-8, only surrogates, not supplementary code points, are encoded directly.
+            if (i == inBytes && (!isCESU8 || i <= 3))
+            {
+                /* Remove the accumulated high bits */
+                ch -= offsetsFromUTF8[inBytes];
+
+                /* Normal valid byte when the loop has not prematurely terminated (i < inBytes) */
+                if (ch <= MAXIMUM_UCS2) 
+                {
+                    /* fits in 16 bits */
+                    *(myTarget++) = (UChar) ch;
+                    *(myOffsets++) = offsetNum;
+                }
+                else
+                {
+                    /* write out the surrogates */
+                    *(myTarget++) = U16_LEAD(ch);
+                    *(myOffsets++) = offsetNum;
+                    ch = U16_TRAIL(ch);
+                    if (myTarget < targetLimit)
+                    {
+                        *(myTarget++) = (UChar)ch;
+                        *(myOffsets++) = offsetNum;
+                    }
+                    else
+                    {
+                        cnv->UCharErrorBuffer[0] = (UChar) ch;
+                        cnv->UCharErrorBufferLength = 1;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }
+                offsetNum += i;
+            }
+            else
+            {
+                cnv->toULength = (int8_t)i;
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+    }
+
+donefornow:
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {   /* End of target buffer */
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = myTarget;
+    args->source = (const char *) mySource;
+    args->offsets = myOffsets;
+}
+U_CDECL_END
+
+U_CFUNC void  U_CALLCONV ucnv_fromUnicode_UTF8 (UConverterFromUnicodeArgs * args,
+                                    UErrorCode * err)
+{
+    UConverter *cnv = args->converter;
+    const UChar *mySource = args->source;
+    const UChar *sourceLimit = args->sourceLimit;
+    uint8_t *myTarget = (uint8_t *) args->target;
+    const uint8_t *targetLimit = (uint8_t *) args->targetLimit;
+    uint8_t *tempPtr;
+    UChar32 ch;
+    uint8_t tempBuf[4];
+    int32_t indexToWrite;
+    UBool isNotCESU8 = !hasCESU8Data(cnv);
+
+    if (cnv->fromUChar32 && myTarget < targetLimit)
+    {
+        ch = cnv->fromUChar32;
+        cnv->fromUChar32 = 0;
+        goto lowsurrogate;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+
+        if (ch < 0x80)        /* Single byte */
+        {
+            *(myTarget++) = (uint8_t) ch;
+        }
+        else if (ch < 0x800)  /* Double byte */
+        {
+            *(myTarget++) = (uint8_t) ((ch >> 6) | 0xc0);
+            if (myTarget < targetLimit)
+            {
+                *(myTarget++) = (uint8_t) ((ch & 0x3f) | 0x80);
+            }
+            else
+            {
+                cnv->charErrorBuffer[0] = (uint8_t) ((ch & 0x3f) | 0x80);
+                cnv->charErrorBufferLength = 1;
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        else {
+            /* Check for surrogates */
+            if(U16_IS_SURROGATE(ch) && isNotCESU8) {
+lowsurrogate:
+                if (mySource < sourceLimit) {
+                    /* test both code units */
+                    if(U16_IS_SURROGATE_LEAD(ch) && U16_IS_TRAIL(*mySource)) {
+                        /* convert and consume this supplementary code point */
+                        ch=U16_GET_SUPPLEMENTARY(ch, *mySource);
+                        ++mySource;
+                        /* exit this condition tree */
+                    }
+                    else {
+                        /* this is an unpaired trail or lead code unit */
+                        /* callback(illegal) */
+                        cnv->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* no more input */
+                    cnv->fromUChar32 = ch;
+                    break;
+                }
+            }
+
+            /* Do we write the buffer directly for speed,
+            or do we have to be careful about target buffer space? */
+            tempPtr = (((targetLimit - myTarget) >= 4) ? myTarget : tempBuf);
+
+            if (ch <= MAXIMUM_UCS2) {
+                indexToWrite = 2;
+                tempPtr[0] = (uint8_t) ((ch >> 12) | 0xe0);
+            }
+            else {
+                indexToWrite = 3;
+                tempPtr[0] = (uint8_t) ((ch >> 18) | 0xf0);
+                tempPtr[1] = (uint8_t) (((ch >> 12) & 0x3f) | 0x80);
+            }
+            tempPtr[indexToWrite-1] = (uint8_t) (((ch >> 6) & 0x3f) | 0x80);
+            tempPtr[indexToWrite] = (uint8_t) ((ch & 0x3f) | 0x80);
+
+            if (tempPtr == myTarget) {
+                /* There was enough space to write the codepoint directly. */
+                myTarget += (indexToWrite + 1);
+            }
+            else {
+                /* We might run out of room soon. Write it slowly. */
+                for (; tempPtr <= (tempBuf + indexToWrite); tempPtr++) {
+                    if (myTarget < targetLimit) {
+                        *(myTarget++) = *tempPtr;
+                    }
+                    else {
+                        cnv->charErrorBuffer[cnv->charErrorBufferLength++] = *tempPtr;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }
+            }
+        }
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+}
+
+U_CFUNC void  U_CALLCONV ucnv_fromUnicode_UTF8_OFFSETS_LOGIC (UConverterFromUnicodeArgs * args,
+                                                  UErrorCode * err)
+{
+    UConverter *cnv = args->converter;
+    const UChar *mySource = args->source;
+    int32_t *myOffsets = args->offsets;
+    const UChar *sourceLimit = args->sourceLimit;
+    uint8_t *myTarget = (uint8_t *) args->target;
+    const uint8_t *targetLimit = (uint8_t *) args->targetLimit;
+    uint8_t *tempPtr;
+    UChar32 ch;
+    int32_t offsetNum, nextSourceIndex;
+    int32_t indexToWrite;
+    uint8_t tempBuf[4];
+    UBool isNotCESU8 = !hasCESU8Data(cnv);
+
+    if (cnv->fromUChar32 && myTarget < targetLimit)
+    {
+        ch = cnv->fromUChar32;
+        cnv->fromUChar32 = 0;
+        offsetNum = -1;
+        nextSourceIndex = 0;
+        goto lowsurrogate;
+    } else {
+        offsetNum = 0;
+    }
+
+    while (mySource < sourceLimit && myTarget < targetLimit)
+    {
+        ch = *(mySource++);
+
+        if (ch < 0x80)        /* Single byte */
+        {
+            *(myOffsets++) = offsetNum++;
+            *(myTarget++) = (char) ch;
+        }
+        else if (ch < 0x800)  /* Double byte */
+        {
+            *(myOffsets++) = offsetNum;
+            *(myTarget++) = (uint8_t) ((ch >> 6) | 0xc0);
+            if (myTarget < targetLimit)
+            {
+                *(myOffsets++) = offsetNum++;
+                *(myTarget++) = (uint8_t) ((ch & 0x3f) | 0x80);
+            }
+            else
+            {
+                cnv->charErrorBuffer[0] = (uint8_t) ((ch & 0x3f) | 0x80);
+                cnv->charErrorBufferLength = 1;
+                *err = U_BUFFER_OVERFLOW_ERROR;
+            }
+        }
+        else
+        /* Check for surrogates */
+        {
+            nextSourceIndex = offsetNum + 1;
+
+            if(U16_IS_SURROGATE(ch) && isNotCESU8) {
+lowsurrogate:
+                if (mySource < sourceLimit) {
+                    /* test both code units */
+                    if(U16_IS_SURROGATE_LEAD(ch) && U16_IS_TRAIL(*mySource)) {
+                        /* convert and consume this supplementary code point */
+                        ch=U16_GET_SUPPLEMENTARY(ch, *mySource);
+                        ++mySource;
+                        ++nextSourceIndex;
+                        /* exit this condition tree */
+                    }
+                    else {
+                        /* this is an unpaired trail or lead code unit */
+                        /* callback(illegal) */
+                        cnv->fromUChar32 = ch;
+                        *err = U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+                else {
+                    /* no more input */
+                    cnv->fromUChar32 = ch;
+                    break;
+                }
+            }
+
+            /* Do we write the buffer directly for speed,
+            or do we have to be careful about target buffer space? */
+            tempPtr = (((targetLimit - myTarget) >= 4) ? myTarget : tempBuf);
+
+            if (ch <= MAXIMUM_UCS2) {
+                indexToWrite = 2;
+                tempPtr[0] = (uint8_t) ((ch >> 12) | 0xe0);
+            }
+            else {
+                indexToWrite = 3;
+                tempPtr[0] = (uint8_t) ((ch >> 18) | 0xf0);
+                tempPtr[1] = (uint8_t) (((ch >> 12) & 0x3f) | 0x80);
+            }
+            tempPtr[indexToWrite-1] = (uint8_t) (((ch >> 6) & 0x3f) | 0x80);
+            tempPtr[indexToWrite] = (uint8_t) ((ch & 0x3f) | 0x80);
+
+            if (tempPtr == myTarget) {
+                /* There was enough space to write the codepoint directly. */
+                myTarget += (indexToWrite + 1);
+                myOffsets[0] = offsetNum;
+                myOffsets[1] = offsetNum;
+                myOffsets[2] = offsetNum;
+                if (indexToWrite >= 3) {
+                    myOffsets[3] = offsetNum;
+                }
+                myOffsets += (indexToWrite + 1);
+            }
+            else {
+                /* We might run out of room soon. Write it slowly. */
+                for (; tempPtr <= (tempBuf + indexToWrite); tempPtr++) {
+                    if (myTarget < targetLimit)
+                    {
+                        *(myOffsets++) = offsetNum;
+                        *(myTarget++) = *tempPtr;
+                    }
+                    else
+                    {
+                        cnv->charErrorBuffer[cnv->charErrorBufferLength++] = *tempPtr;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }
+            }
+            offsetNum = nextSourceIndex;
+        }
+    }
+
+    if (mySource < sourceLimit && myTarget >= targetLimit && U_SUCCESS(*err))
+    {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    args->target = (char *) myTarget;
+    args->source = mySource;
+    args->offsets = myOffsets;
+}
+
+U_CDECL_BEGIN
+static UChar32 U_CALLCONV ucnv_getNextUChar_UTF8(UConverterToUnicodeArgs *args,
+                                               UErrorCode *err) {
+    UConverter *cnv;
+    const uint8_t *sourceInitial;
+    const uint8_t *source;
+    uint8_t myByte;
+    UChar32 ch;
+    int8_t i;
+
+    /* UTF-8 only here, the framework handles CESU-8 to combine surrogate pairs */
+
+    cnv = args->converter;
+    sourceInitial = source = (const uint8_t *)args->source;
+    if (source >= (const uint8_t *)args->sourceLimit)
+    {
+        /* no input */
+        *err = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0xffff;
+    }
+
+    myByte = (uint8_t)*(source++);
+    if (U8_IS_SINGLE(myByte))
+    {
+        args->source = (const char *)source;
+        return (UChar32)myByte;
+    }
+
+    uint16_t countTrailBytes = U8_COUNT_TRAIL_BYTES(myByte);
+    if (countTrailBytes == 0) {
+        cnv->toUBytes[0] = myByte;
+        cnv->toULength = 1;
+        *err = U_ILLEGAL_CHAR_FOUND;
+        args->source = (const char *)source;
+        return 0xffff;
+    }
+
+    /*The byte sequence is longer than the buffer area passed*/
+    if (((const char *)source + countTrailBytes) > args->sourceLimit)
+    {
+        /* check if all of the remaining bytes are trail bytes */
+        uint16_t extraBytesToWrite = countTrailBytes + 1;
+        cnv->toUBytes[0] = myByte;
+        i = 1;
+        *err = U_TRUNCATED_CHAR_FOUND;
+        while(source < (const uint8_t *)args->sourceLimit) {
+            uint8_t b = *source;
+            if(icu::UTF8::isValidTrail(myByte, b, i, extraBytesToWrite)) {
+                cnv->toUBytes[i++] = b;
+                ++source;
+            } else {
+                /* error even before we run out of input */
+                *err = U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        }
+        cnv->toULength = i;
+        args->source = (const char *)source;
+        return 0xffff;
+    }
+
+    ch = myByte << 6;
+    if(countTrailBytes == 2) {
+        uint8_t t1 = *source, t2;
+        if(U8_IS_VALID_LEAD3_AND_T1(myByte, t1) && U8_IS_TRAIL(t2 = *++source)) {
+            args->source = (const char *)(source + 1);
+            return (((ch + t1) << 6) + t2) - offsetsFromUTF8[3];
+        }
+    } else if(countTrailBytes == 1) {
+        uint8_t t1 = *source;
+        if(U8_IS_TRAIL(t1)) {
+            args->source = (const char *)(source + 1);
+            return (ch + t1) - offsetsFromUTF8[2];
+        }
+    } else {  // countTrailBytes == 3
+        uint8_t t1 = *source, t2, t3;
+        if(U8_IS_VALID_LEAD4_AND_T1(myByte, t1) && U8_IS_TRAIL(t2 = *++source) &&
+                U8_IS_TRAIL(t3 = *++source)) {
+            args->source = (const char *)(source + 1);
+            return (((((ch + t1) << 6) + t2) << 6) + t3) - offsetsFromUTF8[4];
+        }
+    }
+    args->source = (const char *)source;
+
+    for(i = 0; sourceInitial < source; ++i) {
+        cnv->toUBytes[i] = *sourceInitial++;
+    }
+    cnv->toULength = i;
+    *err = U_ILLEGAL_CHAR_FOUND;
+    return 0xffff;
+} 
+U_CDECL_END
+
+/* UTF-8-from-UTF-8 conversion functions ------------------------------------ */
+
+U_CDECL_BEGIN
+/* "Convert" UTF-8 to UTF-8: Validate and copy. Modified from ucnv_DBCSFromUTF8(). */
+static void U_CALLCONV
+ucnv_UTF8FromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                  UConverterToUnicodeArgs *pToUArgs,
+                  UErrorCode *pErrorCode) {
+    UConverter *utf8;
+    const uint8_t *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t count;
+
+    int8_t oldToULength, toULength, toULimit;
+
+    UChar32 c;
+    uint8_t b, t1, t2;
+
+    /* set up the local pointers */
+    utf8=pToUArgs->converter;
+    source=(uint8_t *)pToUArgs->source;
+    sourceLimit=(uint8_t *)pToUArgs->sourceLimit;
+    target=(uint8_t *)pFromUArgs->target;
+    targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target);
+
+    /* get the converter state from the UTF-8 UConverter */
+    if(utf8->toULength > 0) {
+        toULength=oldToULength=utf8->toULength;
+        toULimit=(int8_t)utf8->mode;
+        c=(UChar32)utf8->toUnicodeStatus;
+    } else {
+        toULength=oldToULength=toULimit=0;
+        c = 0;
+    }
+
+    count=(int32_t)(sourceLimit-source)+oldToULength;
+    if(count<toULimit) {
+        /*
+         * Not enough input to complete the partial character.
+         * Jump to moreBytes below - it will not output to target.
+         */
+    } else if(targetCapacity<toULimit) {
+        /*
+         * Not enough target capacity to output the partial character.
+         * Let the standard converter handle this.
+         */
+        *pErrorCode=U_USING_DEFAULT_WARNING;
+        return;
+    } else {
+        // Use a single counter for source and target, counting the minimum of
+        // the source length and the target capacity.
+        // Let the standard converter handle edge cases.
+        if(count>targetCapacity) {
+            count=targetCapacity;
+        }
+
+        // The conversion loop checks count>0 only once per character.
+        // If the buffer ends with a truncated sequence,
+        // then we reduce the count to stop before that,
+        // and collect the remaining bytes after the conversion loop.
+
+        // Do not go back into the bytes that will be read for finishing a partial
+        // sequence from the previous buffer.
+        int32_t length=count-toULength;
+        U8_TRUNCATE_IF_INCOMPLETE(source, 0, length);
+        count=toULength+length;
+    }
+
+    if(c!=0) {
+        utf8->toUnicodeStatus=0;
+        utf8->toULength=0;
+        goto moreBytes;
+        /* See note in ucnv_SBCSFromUTF8() about this goto. */
+    }
+
+    /* conversion loop */
+    while(count>0) {
+        b=*source++;
+        if(U8_IS_SINGLE(b)) {
+            /* convert ASCII */
+            *target++=b;
+            --count;
+            continue;
+        } else {
+            if(b>=0xe0) {
+                if( /* handle U+0800..U+FFFF inline */
+                    b<0xf0 &&
+                    U8_IS_VALID_LEAD3_AND_T1(b, t1=source[0]) &&
+                    U8_IS_TRAIL(t2=source[1])
+                ) {
+                    source+=2;
+                    *target++=b;
+                    *target++=t1;
+                    *target++=t2;
+                    count-=3;
+                    continue;
+                }
+            } else {
+                if( /* handle U+0080..U+07FF inline */
+                    b>=0xc2 &&
+                    U8_IS_TRAIL(t1=*source)
+                ) {
+                    ++source;
+                    *target++=b;
+                    *target++=t1;
+                    count-=2;
+                    continue;
+                }
+            }
+
+            /* handle "complicated" and error cases, and continuing partial characters */
+            oldToULength=0;
+            toULength=1;
+            toULimit=U8_COUNT_BYTES_NON_ASCII(b);
+            c=b;
+moreBytes:
+            while(toULength<toULimit) {
+                if(source<sourceLimit) {
+                    b=*source;
+                    if(icu::UTF8::isValidTrail(c, b, toULength, toULimit)) {
+                        ++source;
+                        ++toULength;
+                        c=(c<<6)+b;
+                    } else {
+                        break; /* sequence too short, stop with toULength<toULimit */
+                    }
+                } else {
+                    /* store the partial UTF-8 character, compatible with the regular UTF-8 converter */
+                    source-=(toULength-oldToULength);
+                    while(oldToULength<toULength) {
+                        utf8->toUBytes[oldToULength++]=*source++;
+                    }
+                    utf8->toUnicodeStatus=c;
+                    utf8->toULength=toULength;
+                    utf8->mode=toULimit;
+                    pToUArgs->source=(char *)source;
+                    pFromUArgs->target=(char *)target;
+                    return;
+                }
+            }
+
+            if(toULength!=toULimit) {
+                /* error handling: illegal UTF-8 byte sequence */
+                source-=(toULength-oldToULength);
+                while(oldToULength<toULength) {
+                    utf8->toUBytes[oldToULength++]=*source++;
+                }
+                utf8->toULength=toULength;
+                pToUArgs->source=(char *)source;
+                pFromUArgs->target=(char *)target;
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                return;
+            }
+
+            /* copy the legal byte sequence to the target */
+            {
+                int8_t i;
+
+                for(i=0; i<oldToULength; ++i) {
+                    *target++=utf8->toUBytes[i];
+                }
+                source-=(toULength-oldToULength);
+                for(; i<toULength; ++i) {
+                    *target++=*source++;
+                }
+                count-=toULength;
+            }
+        }
+    }
+    U_ASSERT(count>=0);
+
+    if(U_SUCCESS(*pErrorCode) && source<sourceLimit) {
+        if(target==(const uint8_t *)pFromUArgs->targetLimit) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        } else {
+            b=*source;
+            toULimit=U8_COUNT_BYTES(b);
+            if(toULimit>(sourceLimit-source)) {
+                /* collect a truncated byte sequence */
+                toULength=0;
+                c=b;
+                for(;;) {
+                    utf8->toUBytes[toULength++]=b;
+                    if(++source==sourceLimit) {
+                        /* partial byte sequence at end of source */
+                        utf8->toUnicodeStatus=c;
+                        utf8->toULength=toULength;
+                        utf8->mode=toULimit;
+                        break;
+                    } else if(!icu::UTF8::isValidTrail(c, b=*source, toULength, toULimit)) {
+                        utf8->toULength=toULength;
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                    c=(c<<6)+b;
+                }
+            } else {
+                /* partial-sequence target overflow: fall back to the pivoting implementation */
+                *pErrorCode=U_USING_DEFAULT_WARNING;
+            }
+        }
+    }
+
+    /* write back the updated pointers */
+    pToUArgs->source=(char *)source;
+    pFromUArgs->target=(char *)target;
+}
+
+U_CDECL_END
+
+/* UTF-8 converter data ----------------------------------------------------- */
+
+static const UConverterImpl _UTF8Impl={
+    UCNV_UTF8,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    ucnv_toUnicode_UTF8,
+    ucnv_toUnicode_UTF8_OFFSETS_LOGIC,
+    ucnv_fromUnicode_UTF8,
+    ucnv_fromUnicode_UTF8_OFFSETS_LOGIC,
+    ucnv_getNextUChar_UTF8,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getNonSurrogateUnicodeSet,
+
+    ucnv_UTF8FromUTF8,
+    ucnv_UTF8FromUTF8
+};
+
+/* The 1208 CCSID refers to any version of Unicode of UTF-8 */
+static const UConverterStaticData _UTF8StaticData={
+    sizeof(UConverterStaticData),
+    "UTF-8",
+    1208, UCNV_IBM, UCNV_UTF8,
+    1, 3, /* max 3 bytes per UChar from UTF-8 (4 bytes from surrogate _pair_) */
+    { 0xef, 0xbf, 0xbd, 0 },3,FALSE,FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+
+const UConverterSharedData _UTF8Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_UTF8StaticData, &_UTF8Impl);
+
+/* CESU-8 converter data ---------------------------------------------------- */
+
+static const UConverterImpl _CESU8Impl={
+    UCNV_CESU8,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    ucnv_toUnicode_UTF8,
+    ucnv_toUnicode_UTF8_OFFSETS_LOGIC,
+    ucnv_fromUnicode_UTF8,
+    ucnv_fromUnicode_UTF8_OFFSETS_LOGIC,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getCompleteUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _CESU8StaticData={
+    sizeof(UConverterStaticData),
+    "CESU-8",
+    9400, /* CCSID for CESU-8 */
+    UCNV_UNKNOWN, UCNV_CESU8, 1, 3,
+    { 0xef, 0xbf, 0xbd, 0 },3,FALSE,FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+
+const UConverterSharedData _CESU8Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_CESU8StaticData, &_CESU8Impl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnvbocu.cpp b/thirdparty/icu4c/common/ucnvbocu.cpp
new file mode 100644
index 0000000000..7c2aab5655
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvbocu.cpp
@@ -0,0 +1,1413 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnvbocu.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002mar27
+*   created by: Markus W. Scherer
+*
+*   This is an implementation of the Binary Ordered Compression for Unicode,
+*   in its MIME-friendly form as defined in http://www.unicode.org/notes/tn6/
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/utf16.h"
+#include "putilimp.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "uassert.h"
+
+/* BOCU-1 constants and macros ---------------------------------------------- */
+
+/*
+ * BOCU-1 encodes the code points of a Unicode string as
+ * a sequence of byte-encoded differences (slope detection),
+ * preserving lexical order.
+ *
+ * Optimize the difference-taking for runs of Unicode text within
+ * small scripts:
+ *
+ * Most small scripts are allocated within aligned 128-blocks of Unicode
+ * code points. Lexical order is preserved if the "previous code point" state
+ * is always moved into the middle of such a block.
+ *
+ * Additionally, "prev" is moved from anywhere in the Unihan and Hangul
+ * areas into the middle of those areas.
+ *
+ * C0 control codes and space are encoded with their US-ASCII bytes.
+ * "prev" is reset for C0 controls but not for space.
+ */
+
+/* initial value for "prev": middle of the ASCII range */
+#define BOCU1_ASCII_PREV        0x40
+
+/* bounding byte values for differences */
+#define BOCU1_MIN               0x21
+#define BOCU1_MIDDLE            0x90
+#define BOCU1_MAX_LEAD          0xfe
+#define BOCU1_MAX_TRAIL         0xff
+#define BOCU1_RESET             0xff
+
+/* number of lead bytes */
+#define BOCU1_COUNT             (BOCU1_MAX_LEAD-BOCU1_MIN+1)
+
+/* adjust trail byte counts for the use of some C0 control byte values */
+#define BOCU1_TRAIL_CONTROLS_COUNT  20
+#define BOCU1_TRAIL_BYTE_OFFSET     (BOCU1_MIN-BOCU1_TRAIL_CONTROLS_COUNT)
+
+/* number of trail bytes */
+#define BOCU1_TRAIL_COUNT       ((BOCU1_MAX_TRAIL-BOCU1_MIN+1)+BOCU1_TRAIL_CONTROLS_COUNT)
+
+/*
+ * number of positive and negative single-byte codes
+ * (counting 0==BOCU1_MIDDLE among the positive ones)
+ */
+#define BOCU1_SINGLE            64
+
+/* number of lead bytes for positive and negative 2/3/4-byte sequences */
+#define BOCU1_LEAD_2            43
+#define BOCU1_LEAD_3            3
+#define BOCU1_LEAD_4            1
+
+/* The difference value range for single-byters. */
+#define BOCU1_REACH_POS_1   (BOCU1_SINGLE-1)
+#define BOCU1_REACH_NEG_1   (-BOCU1_SINGLE)
+
+/* The difference value range for double-byters. */
+#define BOCU1_REACH_POS_2   (BOCU1_REACH_POS_1+BOCU1_LEAD_2*BOCU1_TRAIL_COUNT)
+#define BOCU1_REACH_NEG_2   (BOCU1_REACH_NEG_1-BOCU1_LEAD_2*BOCU1_TRAIL_COUNT)
+
+/* The difference value range for 3-byters. */
+#define BOCU1_REACH_POS_3   \
+    (BOCU1_REACH_POS_2+BOCU1_LEAD_3*BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT)
+
+#define BOCU1_REACH_NEG_3   (BOCU1_REACH_NEG_2-BOCU1_LEAD_3*BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT)
+
+/* The lead byte start values. */
+#define BOCU1_START_POS_2   (BOCU1_MIDDLE+BOCU1_REACH_POS_1+1)
+#define BOCU1_START_POS_3   (BOCU1_START_POS_2+BOCU1_LEAD_2)
+#define BOCU1_START_POS_4   (BOCU1_START_POS_3+BOCU1_LEAD_3)
+     /* ==BOCU1_MAX_LEAD */
+
+#define BOCU1_START_NEG_2   (BOCU1_MIDDLE+BOCU1_REACH_NEG_1)
+#define BOCU1_START_NEG_3   (BOCU1_START_NEG_2-BOCU1_LEAD_2)
+#define BOCU1_START_NEG_4   (BOCU1_START_NEG_3-BOCU1_LEAD_3)
+     /* ==BOCU1_MIN+1 */
+
+/* The length of a byte sequence, according to the lead byte (!=BOCU1_RESET). */
+#define BOCU1_LENGTH_FROM_LEAD(lead) \
+    ((BOCU1_START_NEG_2<=(lead) && (lead)<BOCU1_START_POS_2) ? 1 : \
+     (BOCU1_START_NEG_3<=(lead) && (lead)<BOCU1_START_POS_3) ? 2 : \
+     (BOCU1_START_NEG_4<=(lead) && (lead)<BOCU1_START_POS_4) ? 3 : 4)
+
+/* The length of a byte sequence, according to its packed form. */
+#define BOCU1_LENGTH_FROM_PACKED(packed) \
+    ((uint32_t)(packed)<0x04000000 ? (packed)>>24 : 4)
+
+/*
+ * 12 commonly used C0 control codes (and space) are only used to encode
+ * themselves directly,
+ * which makes BOCU-1 MIME-usable and reasonably safe for
+ * ASCII-oriented software.
+ *
+ * These controls are
+ *  0   NUL
+ *
+ *  7   BEL
+ *  8   BS
+ *
+ *  9   TAB
+ *  a   LF
+ *  b   VT
+ *  c   FF
+ *  d   CR
+ *
+ *  e   SO
+ *  f   SI
+ *
+ * 1a   SUB
+ * 1b   ESC
+ *
+ * The other 20 C0 controls are also encoded directly (to preserve order)
+ * but are also used as trail bytes in difference encoding
+ * (for better compression).
+ */
+#define BOCU1_TRAIL_TO_BYTE(t) ((t)>=BOCU1_TRAIL_CONTROLS_COUNT ? (t)+BOCU1_TRAIL_BYTE_OFFSET : bocu1TrailToByte[t])
+
+/*
+ * Byte value map for control codes,
+ * from external byte values 0x00..0x20
+ * to trail byte values 0..19 (0..0x13) as used in the difference calculation.
+ * External byte values that are illegal as trail bytes are mapped to -1.
+ */
+static const int8_t
+bocu1ByteToTrail[BOCU1_MIN]={
+/*  0     1     2     3     4     5     6     7    */
+    -1,   0x00, 0x01, 0x02, 0x03, 0x04, 0x05, -1,
+
+/*  8     9     a     b     c     d     e     f    */
+    -1,   -1,   -1,   -1,   -1,   -1,   -1,   -1,
+
+/*  10    11    12    13    14    15    16    17   */
+    0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d,
+
+/*  18    19    1a    1b    1c    1d    1e    1f   */
+    0x0e, 0x0f, -1,   -1,   0x10, 0x11, 0x12, 0x13,
+
+/*  20   */
+    -1
+};
+
+/*
+ * Byte value map for control codes,
+ * from trail byte values 0..19 (0..0x13) as used in the difference calculation
+ * to external byte values 0x00..0x20.
+ */
+static const int8_t
+bocu1TrailToByte[BOCU1_TRAIL_CONTROLS_COUNT]={
+/*  0     1     2     3     4     5     6     7    */
+    0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x10, 0x11,
+
+/*  8     9     a     b     c     d     e     f    */
+    0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19,
+
+/*  10    11    12    13   */
+    0x1c, 0x1d, 0x1e, 0x1f
+};
+
+/**
+ * Integer division and modulo with negative numerators
+ * yields negative modulo results and quotients that are one more than
+ * what we need here.
+ * This macro adjust the results so that the modulo-value m is always >=0.
+ *
+ * For positive n, the if() condition is always FALSE.
+ *
+ * @param n Number to be split into quotient and rest.
+ *          Will be modified to contain the quotient.
+ * @param d Divisor.
+ * @param m Output variable for the rest (modulo result).
+ */
+#define NEGDIVMOD(n, d, m) UPRV_BLOCK_MACRO_BEGIN { \
+    (m)=(n)%(d); \
+    (n)/=(d); \
+    if((m)<0) { \
+        --(n); \
+        (m)+=(d); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* Faster versions of packDiff() for single-byte-encoded diff values. */
+
+/** Is a diff value encodable in a single byte? */
+#define DIFF_IS_SINGLE(diff) (BOCU1_REACH_NEG_1<=(diff) && (diff)<=BOCU1_REACH_POS_1)
+
+/** Encode a diff value in a single byte. */
+#define PACK_SINGLE_DIFF(diff) (BOCU1_MIDDLE+(diff))
+
+/** Is a diff value encodable in two bytes? */
+#define DIFF_IS_DOUBLE(diff) (BOCU1_REACH_NEG_2<=(diff) && (diff)<=BOCU1_REACH_POS_2)
+
+/* BOCU-1 implementation functions ------------------------------------------ */
+
+#define BOCU1_SIMPLE_PREV(c) (((c)&~0x7f)+BOCU1_ASCII_PREV)
+
+/**
+ * Compute the next "previous" value for differencing
+ * from the current code point.
+ *
+ * @param c current code point, 0x3040..0xd7a3 (rest handled by macro below)
+ * @return "previous code point" state value
+ */
+static inline int32_t
+bocu1Prev(int32_t c) {
+    /* compute new prev */
+    if(/* 0x3040<=c && */ c<=0x309f) {
+        /* Hiragana is not 128-aligned */
+        return 0x3070;
+    } else if(0x4e00<=c && c<=0x9fa5) {
+        /* CJK Unihan */
+        return 0x4e00-BOCU1_REACH_NEG_2;
+    } else if(0xac00<=c /* && c<=0xd7a3 */) {
+        /* Korean Hangul */
+        return (0xd7a3+0xac00)/2;
+    } else {
+        /* mostly small scripts */
+        return BOCU1_SIMPLE_PREV(c);
+    }
+}
+
+/** Fast version of bocu1Prev() for most scripts. */
+#define BOCU1_PREV(c) ((c)<0x3040 || (c)>0xd7a3 ? BOCU1_SIMPLE_PREV(c) : bocu1Prev(c))
+
+/*
+ * The BOCU-1 converter uses the standard setup code in ucnv.c/ucnv_bld.c.
+ * The UConverter fields are used as follows:
+ *
+ * fromUnicodeStatus    encoder's prev (0 will be interpreted as BOCU1_ASCII_PREV)
+ *
+ * toUnicodeStatus      decoder's prev (0 will be interpreted as BOCU1_ASCII_PREV)
+ * mode                 decoder's incomplete (diff<<2)|count (ignored when toULength==0)
+ */
+
+/* BOCU-1-from-Unicode conversion functions --------------------------------- */
+
+/**
+ * Encode a difference -0x10ffff..0x10ffff in 1..4 bytes
+ * and return a packed integer with them.
+ *
+ * The encoding favors small absolute differences with short encodings
+ * to compress runs of same-script characters.
+ *
+ * Optimized version with unrolled loops and fewer floating-point operations
+ * than the standard packDiff().
+ *
+ * @param diff difference value -0x10ffff..0x10ffff
+ * @return
+ *      0x010000zz for 1-byte sequence zz
+ *      0x0200yyzz for 2-byte sequence yy zz
+ *      0x03xxyyzz for 3-byte sequence xx yy zz
+ *      0xwwxxyyzz for 4-byte sequence ww xx yy zz (ww>0x03)
+ */
+static int32_t
+packDiff(int32_t diff) {
+    int32_t result, m;
+
+    U_ASSERT(!DIFF_IS_SINGLE(diff)); /* assume we won't be called where diff==BOCU1_REACH_NEG_1=-64 */
+    if(diff>=BOCU1_REACH_NEG_1) {
+        /* mostly positive differences, and single-byte negative ones */
+#if 0   /* single-byte case handled in macros, see below */
+        if(diff<=BOCU1_REACH_POS_1) {
+            /* single byte */
+            return 0x01000000|(BOCU1_MIDDLE+diff);
+        } else
+#endif
+        if(diff<=BOCU1_REACH_POS_2) {
+            /* two bytes */
+            diff-=BOCU1_REACH_POS_1+1;
+            result=0x02000000;
+
+            m=diff%BOCU1_TRAIL_COUNT;
+            diff/=BOCU1_TRAIL_COUNT;
+            result|=BOCU1_TRAIL_TO_BYTE(m);
+
+            result|=(BOCU1_START_POS_2+diff)<<8;
+        } else if(diff<=BOCU1_REACH_POS_3) {
+            /* three bytes */
+            diff-=BOCU1_REACH_POS_2+1;
+            result=0x03000000;
+
+            m=diff%BOCU1_TRAIL_COUNT;
+            diff/=BOCU1_TRAIL_COUNT;
+            result|=BOCU1_TRAIL_TO_BYTE(m);
+
+            m=diff%BOCU1_TRAIL_COUNT;
+            diff/=BOCU1_TRAIL_COUNT;
+            result|=BOCU1_TRAIL_TO_BYTE(m)<<8;
+
+            result|=(BOCU1_START_POS_3+diff)<<16;
+        } else {
+            /* four bytes */
+            diff-=BOCU1_REACH_POS_3+1;
+
+            m=diff%BOCU1_TRAIL_COUNT;
+            diff/=BOCU1_TRAIL_COUNT;
+            result=BOCU1_TRAIL_TO_BYTE(m);
+
+            m=diff%BOCU1_TRAIL_COUNT;
+            diff/=BOCU1_TRAIL_COUNT;
+            result|=BOCU1_TRAIL_TO_BYTE(m)<<8;
+
+            /*
+             * We know that / and % would deliver quotient 0 and rest=diff.
+             * Avoid division and modulo for performance.
+             */
+            result|=BOCU1_TRAIL_TO_BYTE(diff)<<16;
+
+            result|=((uint32_t)BOCU1_START_POS_4)<<24;
+        }
+    } else {
+        /* two- to four-byte negative differences */
+        if(diff>=BOCU1_REACH_NEG_2) {
+            /* two bytes */
+            diff-=BOCU1_REACH_NEG_1;
+            result=0x02000000;
+
+            NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+            result|=BOCU1_TRAIL_TO_BYTE(m);
+
+            result|=(BOCU1_START_NEG_2+diff)<<8;
+        } else if(diff>=BOCU1_REACH_NEG_3) {
+            /* three bytes */
+            diff-=BOCU1_REACH_NEG_2;
+            result=0x03000000;
+
+            NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+            result|=BOCU1_TRAIL_TO_BYTE(m);
+
+            NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+            result|=BOCU1_TRAIL_TO_BYTE(m)<<8;
+
+            result|=(BOCU1_START_NEG_3+diff)<<16;
+        } else {
+            /* four bytes */
+            diff-=BOCU1_REACH_NEG_3;
+
+            NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+            result=BOCU1_TRAIL_TO_BYTE(m);
+
+            NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+            result|=BOCU1_TRAIL_TO_BYTE(m)<<8;
+
+            /*
+             * We know that NEGDIVMOD would deliver
+             * quotient -1 and rest=diff+BOCU1_TRAIL_COUNT.
+             * Avoid division and modulo for performance.
+             */
+            m=diff+BOCU1_TRAIL_COUNT;
+            result|=BOCU1_TRAIL_TO_BYTE(m)<<16;
+
+            result|=BOCU1_MIN<<24;
+        }
+    }
+    return result;
+}
+
+
+static void U_CALLCONV
+_Bocu1FromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                             UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t *offsets;
+
+    int32_t prev, c, diff;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+    prev=(int32_t)cnv->fromUnicodeStatus;
+    if(prev==0) {
+        prev=BOCU1_ASCII_PREV;
+    }
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= c==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* conversion loop */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+fastSingle:
+    /* fast loop for single-byte differences */
+    /* use only one loop counter variable, targetCapacity, not also source */
+    diff=(int32_t)(sourceLimit-source);
+    if(targetCapacity>diff) {
+        targetCapacity=diff;
+    }
+    while(targetCapacity>0 && (c=*source)<0x3000) {
+        if(c<=0x20) {
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(uint8_t)c;
+            *offsets++=nextSourceIndex++;
+            ++source;
+            --targetCapacity;
+        } else {
+            diff=c-prev;
+            if(DIFF_IS_SINGLE(diff)) {
+                prev=BOCU1_SIMPLE_PREV(c);
+                *target++=(uint8_t)PACK_SINGLE_DIFF(diff);
+                *offsets++=nextSourceIndex++;
+                ++source;
+                --targetCapacity;
+            } else {
+                break;
+            }
+        }
+    }
+    /* restore real values */
+    targetCapacity=(int32_t)((const uint8_t *)pArgs->targetLimit-target);
+    sourceIndex=nextSourceIndex; /* wrong if offsets==NULL but does not matter */
+
+    /* regular loop for all cases */
+    while(source<sourceLimit) {
+        if(targetCapacity>0) {
+            c=*source++;
+            ++nextSourceIndex;
+
+            if(c<=0x20) {
+                /*
+                 * ISO C0 control & space:
+                 * Encode directly for MIME compatibility,
+                 * and reset state except for space, to not disrupt compression.
+                 */
+                if(c!=0x20) {
+                    prev=BOCU1_ASCII_PREV;
+                }
+                *target++=(uint8_t)c;
+                *offsets++=sourceIndex;
+                --targetCapacity;
+
+                sourceIndex=nextSourceIndex;
+                continue;
+            }
+
+            if(U16_IS_LEAD(c)) {
+getTrail:
+                if(source<sourceLimit) {
+                    /* test the following code unit */
+                    UChar trail=*source;
+                    if(U16_IS_TRAIL(trail)) {
+                        ++source;
+                        ++nextSourceIndex;
+                        c=U16_GET_SUPPLEMENTARY(c, trail);
+                    }
+                } else {
+                    /* no more input */
+                    c=-c; /* negative lead surrogate as "incomplete" indicator to avoid c=0 everywhere else */
+                    break;
+                }
+            }
+
+            /*
+             * all other Unicode code points c==U+0021..U+10ffff
+             * are encoded with the difference c-prev
+             *
+             * a new prev is computed from c,
+             * placed in the middle of a 0x80-block (for most small scripts) or
+             * in the middle of the Unihan and Hangul blocks
+             * to statistically minimize the following difference
+             */
+            diff=c-prev;
+            prev=BOCU1_PREV(c);
+            if(DIFF_IS_SINGLE(diff)) {
+                *target++=(uint8_t)PACK_SINGLE_DIFF(diff);
+                *offsets++=sourceIndex;
+                --targetCapacity;
+                sourceIndex=nextSourceIndex;
+                if(c<0x3000) {
+                    goto fastSingle;
+                }
+            } else if(DIFF_IS_DOUBLE(diff) && 2<=targetCapacity) {
+                /* optimize 2-byte case */
+                int32_t m;
+
+                if(diff>=0) {
+                    diff-=BOCU1_REACH_POS_1+1;
+                    m=diff%BOCU1_TRAIL_COUNT;
+                    diff/=BOCU1_TRAIL_COUNT;
+                    diff+=BOCU1_START_POS_2;
+                } else {
+                    diff-=BOCU1_REACH_NEG_1;
+                    NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+                    diff+=BOCU1_START_NEG_2;
+                }
+                *target++=(uint8_t)diff;
+                *target++=(uint8_t)BOCU1_TRAIL_TO_BYTE(m);
+                *offsets++=sourceIndex;
+                *offsets++=sourceIndex;
+                targetCapacity-=2;
+                sourceIndex=nextSourceIndex;
+            } else {
+                int32_t length; /* will be 2..4 */
+
+                diff=packDiff(diff);
+                length=BOCU1_LENGTH_FROM_PACKED(diff);
+
+                /* write the output character bytes from diff and length */
+                /* from the first if in the loop we know that targetCapacity>0 */
+                if(length<=targetCapacity) {
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 4:
+                        *target++=(uint8_t)(diff>>24);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 3:
+                        *target++=(uint8_t)(diff>>16);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 2:
+                        *target++=(uint8_t)(diff>>8);
+                        *offsets++=sourceIndex;
+                    /* case 1: handled above */
+                        *target++=(uint8_t)diff;
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                    targetCapacity-=length;
+                    sourceIndex=nextSourceIndex;
+                } else {
+                    uint8_t *charErrorBuffer;
+
+                    /*
+                     * We actually do this backwards here:
+                     * In order to save an intermediate variable, we output
+                     * first to the overflow buffer what does not fit into the
+                     * regular target.
+                     */
+                    /* we know that 1<=targetCapacity<length<=4 */
+                    length-=targetCapacity;
+                    charErrorBuffer=(uint8_t *)cnv->charErrorBuffer;
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 3:
+                        *charErrorBuffer++=(uint8_t)(diff>>16);
+                        U_FALLTHROUGH;
+                    case 2:
+                        *charErrorBuffer++=(uint8_t)(diff>>8);
+                        U_FALLTHROUGH;
+                    case 1:
+                        *charErrorBuffer=(uint8_t)diff;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                    cnv->charErrorBufferLength=(int8_t)length;
+
+                    /* now output what fits into the regular target */
+                    diff>>=8*length; /* length was reduced by targetCapacity */
+                    switch(targetCapacity) {
+                        /* each branch falls through to the next one */
+                    case 3:
+                        *target++=(uint8_t)(diff>>16);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 2:
+                        *target++=(uint8_t)(diff>>8);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 1:
+                        *target++=(uint8_t)diff;
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+
+                    /* target overflow */
+                    targetCapacity=0;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32= c<0 ? -c : 0;
+    cnv->fromUnicodeStatus=(uint32_t)prev;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+/*
+ * Identical to _Bocu1FromUnicodeWithOffsets but without offset handling.
+ * If a change is made in the original function, then either
+ * change this function the same way or
+ * re-copy the original function and remove the variables
+ * offsets, sourceIndex, and nextSourceIndex.
+ */
+static void U_CALLCONV
+_Bocu1FromUnicode(UConverterFromUnicodeArgs *pArgs,
+                  UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+
+    int32_t prev, c, diff;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+    prev=(int32_t)cnv->fromUnicodeStatus;
+    if(prev==0) {
+        prev=BOCU1_ASCII_PREV;
+    }
+
+    /* conversion loop */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+fastSingle:
+    /* fast loop for single-byte differences */
+    /* use only one loop counter variable, targetCapacity, not also source */
+    diff=(int32_t)(sourceLimit-source);
+    if(targetCapacity>diff) {
+        targetCapacity=diff;
+    }
+    while(targetCapacity>0 && (c=*source)<0x3000) {
+        if(c<=0x20) {
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(uint8_t)c;
+        } else {
+            diff=c-prev;
+            if(DIFF_IS_SINGLE(diff)) {
+                prev=BOCU1_SIMPLE_PREV(c);
+                *target++=(uint8_t)PACK_SINGLE_DIFF(diff);
+            } else {
+                break;
+            }
+        }
+        ++source;
+        --targetCapacity;
+    }
+    /* restore real values */
+    targetCapacity=(int32_t)((const uint8_t *)pArgs->targetLimit-target);
+
+    /* regular loop for all cases */
+    while(source<sourceLimit) {
+        if(targetCapacity>0) {
+            c=*source++;
+
+            if(c<=0x20) {
+                /*
+                 * ISO C0 control & space:
+                 * Encode directly for MIME compatibility,
+                 * and reset state except for space, to not disrupt compression.
+                 */
+                if(c!=0x20) {
+                    prev=BOCU1_ASCII_PREV;
+                }
+                *target++=(uint8_t)c;
+                --targetCapacity;
+                continue;
+            }
+
+            if(U16_IS_LEAD(c)) {
+getTrail:
+                if(source<sourceLimit) {
+                    /* test the following code unit */
+                    UChar trail=*source;
+                    if(U16_IS_TRAIL(trail)) {
+                        ++source;
+                        c=U16_GET_SUPPLEMENTARY(c, trail);
+                    }
+                } else {
+                    /* no more input */
+                    c=-c; /* negative lead surrogate as "incomplete" indicator to avoid c=0 everywhere else */
+                    break;
+                }
+            }
+
+            /*
+             * all other Unicode code points c==U+0021..U+10ffff
+             * are encoded with the difference c-prev
+             *
+             * a new prev is computed from c,
+             * placed in the middle of a 0x80-block (for most small scripts) or
+             * in the middle of the Unihan and Hangul blocks
+             * to statistically minimize the following difference
+             */
+            diff=c-prev;
+            prev=BOCU1_PREV(c);
+            if(DIFF_IS_SINGLE(diff)) {
+                *target++=(uint8_t)PACK_SINGLE_DIFF(diff);
+                --targetCapacity;
+                if(c<0x3000) {
+                    goto fastSingle;
+                }
+            } else if(DIFF_IS_DOUBLE(diff) && 2<=targetCapacity) {
+                /* optimize 2-byte case */
+                int32_t m;
+
+                if(diff>=0) {
+                    diff-=BOCU1_REACH_POS_1+1;
+                    m=diff%BOCU1_TRAIL_COUNT;
+                    diff/=BOCU1_TRAIL_COUNT;
+                    diff+=BOCU1_START_POS_2;
+                } else {
+                    diff-=BOCU1_REACH_NEG_1;
+                    NEGDIVMOD(diff, BOCU1_TRAIL_COUNT, m);
+                    diff+=BOCU1_START_NEG_2;
+                }
+                *target++=(uint8_t)diff;
+                *target++=(uint8_t)BOCU1_TRAIL_TO_BYTE(m);
+                targetCapacity-=2;
+            } else {
+                int32_t length; /* will be 2..4 */
+
+                diff=packDiff(diff);
+                length=BOCU1_LENGTH_FROM_PACKED(diff);
+
+                /* write the output character bytes from diff and length */
+                /* from the first if in the loop we know that targetCapacity>0 */
+                if(length<=targetCapacity) {
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 4:
+                        *target++=(uint8_t)(diff>>24);
+                        U_FALLTHROUGH;
+                    case 3:
+                        *target++=(uint8_t)(diff>>16);
+                    /* case 2: handled above */
+                        *target++=(uint8_t)(diff>>8);
+                    /* case 1: handled above */
+                        *target++=(uint8_t)diff;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                    targetCapacity-=length;
+                } else {
+                    uint8_t *charErrorBuffer;
+
+                    /*
+                     * We actually do this backwards here:
+                     * In order to save an intermediate variable, we output
+                     * first to the overflow buffer what does not fit into the
+                     * regular target.
+                     */
+                    /* we know that 1<=targetCapacity<length<=4 */
+                    length-=targetCapacity;
+                    charErrorBuffer=(uint8_t *)cnv->charErrorBuffer;
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 3:
+                        *charErrorBuffer++=(uint8_t)(diff>>16);
+                        U_FALLTHROUGH;
+                    case 2:
+                        *charErrorBuffer++=(uint8_t)(diff>>8);
+                        U_FALLTHROUGH;
+                    case 1:
+                        *charErrorBuffer=(uint8_t)diff;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                    cnv->charErrorBufferLength=(int8_t)length;
+
+                    /* now output what fits into the regular target */
+                    diff>>=8*length; /* length was reduced by targetCapacity */
+                    switch(targetCapacity) {
+                        /* each branch falls through to the next one */
+                    case 3:
+                        *target++=(uint8_t)(diff>>16);
+                        U_FALLTHROUGH;
+                    case 2:
+                        *target++=(uint8_t)(diff>>8);
+                        U_FALLTHROUGH;
+                    case 1:
+                        *target++=(uint8_t)diff;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+
+                    /* target overflow */
+                    targetCapacity=0;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32= c<0 ? -c : 0;
+    cnv->fromUnicodeStatus=(uint32_t)prev;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+}
+
+/* BOCU-1-to-Unicode conversion functions ----------------------------------- */
+
+/**
+ * Function for BOCU-1 decoder; handles multi-byte lead bytes.
+ *
+ * @param b lead byte;
+ *          BOCU1_MIN<=b<BOCU1_START_NEG_2 or BOCU1_START_POS_2<=b<BOCU1_MAX_LEAD
+ * @return (diff<<2)|count
+ */
+static inline int32_t
+decodeBocu1LeadByte(int32_t b) {
+    int32_t diff, count;
+
+    if(b>=BOCU1_START_NEG_2) {
+        /* positive difference */
+        if(b<BOCU1_START_POS_3) {
+            /* two bytes */
+            diff=((int32_t)b-BOCU1_START_POS_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_POS_1+1;
+            count=1;
+        } else if(b<BOCU1_START_POS_4) {
+            /* three bytes */
+            diff=((int32_t)b-BOCU1_START_POS_3)*BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT+BOCU1_REACH_POS_2+1;
+            count=2;
+        } else {
+            /* four bytes */
+            diff=BOCU1_REACH_POS_3+1;
+            count=3;
+        }
+    } else {
+        /* negative difference */
+        if(b>=BOCU1_START_NEG_3) {
+            /* two bytes */
+            diff=((int32_t)b-BOCU1_START_NEG_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_NEG_1;
+            count=1;
+        } else if(b>BOCU1_MIN) {
+            /* three bytes */
+            diff=((int32_t)b-BOCU1_START_NEG_3)*BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT+BOCU1_REACH_NEG_2;
+            count=2;
+        } else {
+            /* four bytes */
+            diff=-BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT+BOCU1_REACH_NEG_3;
+            count=3;
+        }
+    }
+
+    /* return the state for decoding the trail byte(s) */
+    return (diff<<2)|count;
+}
+
+/**
+ * Function for BOCU-1 decoder; handles multi-byte trail bytes.
+ *
+ * @param count number of remaining trail bytes including this one
+ * @param b trail byte
+ * @return new delta for diff including b - <0 indicates an error
+ *
+ * @see decodeBocu1
+ */
+static inline int32_t
+decodeBocu1TrailByte(int32_t count, int32_t b) {
+    if(b<=0x20) {
+        /* skip some C0 controls and make the trail byte range contiguous */
+        b=bocu1ByteToTrail[b];
+        /* b<0 for an illegal trail byte value will result in return<0 below */
+#if BOCU1_MAX_TRAIL<0xff
+    } else if(b>BOCU1_MAX_TRAIL) {
+        return -99;
+#endif
+    } else {
+        b-=BOCU1_TRAIL_BYTE_OFFSET;
+    }
+
+    /* add trail byte into difference and decrement count */
+    if(count==1) {
+        return b;
+    } else if(count==2) {
+        return b*BOCU1_TRAIL_COUNT;
+    } else /* count==3 */ {
+        return b*(BOCU1_TRAIL_COUNT*BOCU1_TRAIL_COUNT);
+    }
+}
+
+static void U_CALLCONV
+_Bocu1ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                           UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+
+    int32_t prev, count, diff, c;
+
+    int8_t byteIndex;
+    uint8_t *bytes;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    /* get the converter state from UConverter */
+    prev=(int32_t)cnv->toUnicodeStatus;
+    if(prev==0) {
+        prev=BOCU1_ASCII_PREV;
+    }
+    diff=cnv->mode; /* mode may be set to UCNV_SI by ucnv_bld.c but then toULength==0 */
+    count=diff&3;
+    diff>>=2;
+
+    byteIndex=cnv->toULength;
+    bytes=cnv->toUBytes;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=byteIndex==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* conversion "loop" similar to _SCSUToUnicodeWithOffsets() */
+    if(count>0 && byteIndex>0 && target<targetLimit) {
+        goto getTrail;
+    }
+
+fastSingle:
+    /* fast loop for single-byte differences */
+    /* use count as the only loop counter variable */
+    diff=(int32_t)(sourceLimit-source);
+    count=(int32_t)(pArgs->targetLimit-target);
+    if(count>diff) {
+        count=diff;
+    }
+    while(count>0) {
+        if(BOCU1_START_NEG_2<=(c=*source) && c<BOCU1_START_POS_2) {
+            c=prev+(c-BOCU1_MIDDLE);
+            if(c<0x3000) {
+                *target++=(UChar)c;
+                *offsets++=nextSourceIndex++;
+                prev=BOCU1_SIMPLE_PREV(c);
+            } else {
+                break;
+            }
+        } else if(c<=0x20) {
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(UChar)c;
+            *offsets++=nextSourceIndex++;
+        } else {
+            break;
+        }
+        ++source;
+        --count;
+    }
+    sourceIndex=nextSourceIndex; /* wrong if offsets==NULL but does not matter */
+
+    /* decode a sequence of single and lead bytes */
+    while(source<sourceLimit) {
+        if(target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+        ++nextSourceIndex;
+        c=*source++;
+        if(BOCU1_START_NEG_2<=c && c<BOCU1_START_POS_2) {
+            /* Write a code point directly from a single-byte difference. */
+            c=prev+(c-BOCU1_MIDDLE);
+            if(c<0x3000) {
+                *target++=(UChar)c;
+                *offsets++=sourceIndex;
+                prev=BOCU1_SIMPLE_PREV(c);
+                sourceIndex=nextSourceIndex;
+                goto fastSingle;
+            }
+        } else if(c<=0x20) {
+            /*
+             * Direct-encoded C0 control code or space.
+             * Reset prev for C0 control codes but not for space.
+             */
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(UChar)c;
+            *offsets++=sourceIndex;
+            sourceIndex=nextSourceIndex;
+            continue;
+        } else if(BOCU1_START_NEG_3<=c && c<BOCU1_START_POS_3 && source<sourceLimit) {
+            /* Optimize two-byte case. */
+            if(c>=BOCU1_MIDDLE) {
+                diff=((int32_t)c-BOCU1_START_POS_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_POS_1+1;
+            } else {
+                diff=((int32_t)c-BOCU1_START_NEG_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_NEG_1;
+            }
+
+            /* trail byte */
+            ++nextSourceIndex;
+            c=decodeBocu1TrailByte(1, *source++);
+            if(c<0 || (uint32_t)(c=prev+diff+c)>0x10ffff) {
+                bytes[0]=source[-2];
+                bytes[1]=source[-1];
+                byteIndex=2;
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        } else if(c==BOCU1_RESET) {
+            /* only reset the state, no code point */
+            prev=BOCU1_ASCII_PREV;
+            sourceIndex=nextSourceIndex;
+            continue;
+        } else {
+            /*
+             * For multi-byte difference lead bytes, set the decoder state
+             * with the partial difference value from the lead byte and
+             * with the number of trail bytes.
+             */
+            bytes[0]=(uint8_t)c;
+            byteIndex=1;
+
+            diff=decodeBocu1LeadByte(c);
+            count=diff&3;
+            diff>>=2;
+getTrail:
+            for(;;) {
+                if(source>=sourceLimit) {
+                    goto endloop;
+                }
+                ++nextSourceIndex;
+                c=bytes[byteIndex++]=*source++;
+
+                /* trail byte in any position */
+                c=decodeBocu1TrailByte(count, c);
+                if(c<0) {
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                diff+=c;
+                if(--count==0) {
+                    /* final trail byte, deliver a code point */
+                    byteIndex=0;
+                    c=prev+diff;
+                    if((uint32_t)c>0x10ffff) {
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        goto endloop;
+                    }
+                    break;
+                }
+            }
+        }
+
+        /* calculate the next prev and output c */
+        prev=BOCU1_PREV(c);
+        if(c<=0xffff) {
+            *target++=(UChar)c;
+            *offsets++=sourceIndex;
+        } else {
+            /* output surrogate pair */
+            *target++=U16_LEAD(c);
+            if(target<targetLimit) {
+                *target++=U16_TRAIL(c);
+                *offsets++=sourceIndex;
+                *offsets++=sourceIndex;
+            } else {
+                /* target overflow */
+                *offsets++=sourceIndex;
+                cnv->UCharErrorBuffer[0]=U16_TRAIL(c);
+                cnv->UCharErrorBufferLength=1;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+        sourceIndex=nextSourceIndex;
+    }
+endloop:
+
+    if(*pErrorCode==U_ILLEGAL_CHAR_FOUND) {
+        /* set the converter state in UConverter to deal with the next character */
+        cnv->toUnicodeStatus=BOCU1_ASCII_PREV;
+        cnv->mode=0;
+    } else {
+        /* set the converter state back into UConverter */
+        cnv->toUnicodeStatus=(uint32_t)prev;
+        cnv->mode=(diff<<2)|count;
+    }
+    cnv->toULength=byteIndex;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+    return;
+}
+
+/*
+ * Identical to _Bocu1ToUnicodeWithOffsets but without offset handling.
+ * If a change is made in the original function, then either
+ * change this function the same way or
+ * re-copy the original function and remove the variables
+ * offsets, sourceIndex, and nextSourceIndex.
+ */
+static void U_CALLCONV
+_Bocu1ToUnicode(UConverterToUnicodeArgs *pArgs,
+                UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+
+    int32_t prev, count, diff, c;
+
+    int8_t byteIndex;
+    uint8_t *bytes;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+
+    /* get the converter state from UConverter */
+    prev=(int32_t)cnv->toUnicodeStatus;
+    if(prev==0) {
+        prev=BOCU1_ASCII_PREV;
+    }
+    diff=cnv->mode; /* mode may be set to UCNV_SI by ucnv_bld.c but then toULength==0 */
+    count=diff&3;
+    diff>>=2;
+
+    byteIndex=cnv->toULength;
+    bytes=cnv->toUBytes;
+
+    /* conversion "loop" similar to _SCSUToUnicodeWithOffsets() */
+    if(count>0 && byteIndex>0 && target<targetLimit) {
+        goto getTrail;
+    }
+
+fastSingle:
+    /* fast loop for single-byte differences */
+    /* use count as the only loop counter variable */
+    diff=(int32_t)(sourceLimit-source);
+    count=(int32_t)(pArgs->targetLimit-target);
+    if(count>diff) {
+        count=diff;
+    }
+    while(count>0) {
+        if(BOCU1_START_NEG_2<=(c=*source) && c<BOCU1_START_POS_2) {
+            c=prev+(c-BOCU1_MIDDLE);
+            if(c<0x3000) {
+                *target++=(UChar)c;
+                prev=BOCU1_SIMPLE_PREV(c);
+            } else {
+                break;
+            }
+        } else if(c<=0x20) {
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(UChar)c;
+        } else {
+            break;
+        }
+        ++source;
+        --count;
+    }
+
+    /* decode a sequence of single and lead bytes */
+    while(source<sourceLimit) {
+        if(target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+        c=*source++;
+        if(BOCU1_START_NEG_2<=c && c<BOCU1_START_POS_2) {
+            /* Write a code point directly from a single-byte difference. */
+            c=prev+(c-BOCU1_MIDDLE);
+            if(c<0x3000) {
+                *target++=(UChar)c;
+                prev=BOCU1_SIMPLE_PREV(c);
+                goto fastSingle;
+            }
+        } else if(c<=0x20) {
+            /*
+             * Direct-encoded C0 control code or space.
+             * Reset prev for C0 control codes but not for space.
+             */
+            if(c!=0x20) {
+                prev=BOCU1_ASCII_PREV;
+            }
+            *target++=(UChar)c;
+            continue;
+        } else if(BOCU1_START_NEG_3<=c && c<BOCU1_START_POS_3 && source<sourceLimit) {
+            /* Optimize two-byte case. */
+            if(c>=BOCU1_MIDDLE) {
+                diff=((int32_t)c-BOCU1_START_POS_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_POS_1+1;
+            } else {
+                diff=((int32_t)c-BOCU1_START_NEG_2)*BOCU1_TRAIL_COUNT+BOCU1_REACH_NEG_1;
+            }
+
+            /* trail byte */
+            c=decodeBocu1TrailByte(1, *source++);
+            if(c<0 || (uint32_t)(c=prev+diff+c)>0x10ffff) {
+                bytes[0]=source[-2];
+                bytes[1]=source[-1];
+                byteIndex=2;
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                break;
+            }
+        } else if(c==BOCU1_RESET) {
+            /* only reset the state, no code point */
+            prev=BOCU1_ASCII_PREV;
+            continue;
+        } else {
+            /*
+             * For multi-byte difference lead bytes, set the decoder state
+             * with the partial difference value from the lead byte and
+             * with the number of trail bytes.
+             */
+            bytes[0]=(uint8_t)c;
+            byteIndex=1;
+
+            diff=decodeBocu1LeadByte(c);
+            count=diff&3;
+            diff>>=2;
+getTrail:
+            for(;;) {
+                if(source>=sourceLimit) {
+                    goto endloop;
+                }
+                c=bytes[byteIndex++]=*source++;
+
+                /* trail byte in any position */
+                c=decodeBocu1TrailByte(count, c);
+                if(c<0) {
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                diff+=c;
+                if(--count==0) {
+                    /* final trail byte, deliver a code point */
+                    byteIndex=0;
+                    c=prev+diff;
+                    if((uint32_t)c>0x10ffff) {
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        goto endloop;
+                    }
+                    break;
+                }
+            }
+        }
+
+        /* calculate the next prev and output c */
+        prev=BOCU1_PREV(c);
+        if(c<=0xffff) {
+            *target++=(UChar)c;
+        } else {
+            /* output surrogate pair */
+            *target++=U16_LEAD(c);
+            if(target<targetLimit) {
+                *target++=U16_TRAIL(c);
+            } else {
+                /* target overflow */
+                cnv->UCharErrorBuffer[0]=U16_TRAIL(c);
+                cnv->UCharErrorBufferLength=1;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+        }
+    }
+endloop:
+
+    if(*pErrorCode==U_ILLEGAL_CHAR_FOUND) {
+        /* set the converter state in UConverter to deal with the next character */
+        cnv->toUnicodeStatus=BOCU1_ASCII_PREV;
+        cnv->mode=0;
+    } else {
+        /* set the converter state back into UConverter */
+        cnv->toUnicodeStatus=(uint32_t)prev;
+        cnv->mode=(diff<<2)|count;
+    }
+    cnv->toULength=byteIndex;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    return;
+}
+
+/* miscellaneous ------------------------------------------------------------ */
+
+static const UConverterImpl _Bocu1Impl={
+    UCNV_BOCU1,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    _Bocu1ToUnicode,
+    _Bocu1ToUnicodeWithOffsets,
+    _Bocu1FromUnicode,
+    _Bocu1FromUnicodeWithOffsets,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    ucnv_getCompleteUnicodeSet,
+
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _Bocu1StaticData={
+    sizeof(UConverterStaticData),
+    "BOCU-1",
+    1214, /* CCSID for BOCU-1 */
+    UCNV_IBM, UCNV_BOCU1,
+    1, 4, /* one UChar generates at least 1 byte and at most 4 bytes */
+    { 0x1a, 0, 0, 0 }, 1, /* BOCU-1 never needs to write a subchar */
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _Bocu1Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_Bocu1StaticData, &_Bocu1Impl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnvdisp.cpp b/thirdparty/icu4c/common/ucnvdisp.cpp
new file mode 100644
index 0000000000..ac86b98597
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvdisp.cpp
@@ -0,0 +1,88 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1998-2004, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  ucnvdisp.c:
+*  Implements APIs for the ICU's codeset conversion library display names.
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/04/99    helena      Fixed internal header inclusion.
+*   05/09/00    helena      Added implementation to handle fallback mappings.
+*   06/20/2000  helena      OS/400 port changes; mostly typecast.
+*   09/08/2004  grhoten     split from ucnv.c
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ustring.h"
+#include "unicode/ures.h"
+#include "unicode/ucnv.h"
+#include "cstring.h"
+#include "ustr_imp.h"
+#include "ucnv_imp.h"
+#include "putilimp.h"
+
+U_CAPI int32_t U_EXPORT2
+ucnv_getDisplayName(const UConverter *cnv,
+                    const char *displayLocale,
+                    UChar *displayName, int32_t displayNameCapacity,
+                    UErrorCode *pErrorCode) {
+    UResourceBundle *rb;
+    const UChar *name;
+    int32_t length;
+    UErrorCode localStatus = U_ZERO_ERROR;
+
+    /* check arguments */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(cnv==NULL || displayNameCapacity<0 || (displayNameCapacity>0 && displayName==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* open the resource bundle and get the display name string */
+    rb=ures_open(NULL, displayLocale, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* use the internal name as the key */
+    name=ures_getStringByKey(rb, cnv->sharedData->staticData->name, &length, &localStatus);
+    ures_close(rb);
+
+    if(U_SUCCESS(localStatus)) {
+        /* copy the string */
+        if (*pErrorCode == U_ZERO_ERROR) {
+            *pErrorCode = localStatus;
+        }
+        u_memcpy(displayName, name, uprv_min(length, displayNameCapacity)*U_SIZEOF_UCHAR);
+    } else {
+        /* convert the internal name into a Unicode string */
+        length=(int32_t)uprv_strlen(cnv->sharedData->staticData->name);
+        u_charsToUChars(cnv->sharedData->staticData->name, displayName, uprv_min(length, displayNameCapacity));
+    }
+    return u_terminateUChars(displayName, displayNameCapacity, length, pErrorCode);
+}
+
+#endif
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/ucnvhz.cpp b/thirdparty/icu4c/common/ucnvhz.cpp
new file mode 100644
index 0000000000..6b2f5faaf0
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvhz.cpp
@@ -0,0 +1,625 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2000-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnvhz.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000oct16
+*   created by: Ram Viswanadha
+*   10/31/2000  Ram     Implemented offsets logic function
+*   
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "cmemory.h"
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/uset.h"
+#include "unicode/utf16.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "ucnv_imp.h"
+
+#define UCNV_TILDE 0x7E          /* ~ */
+#define UCNV_OPEN_BRACE 0x7B     /* { */
+#define UCNV_CLOSE_BRACE 0x7D   /* } */
+#define SB_ESCAPE    "\x7E\x7D"
+#define DB_ESCAPE    "\x7E\x7B"
+#define TILDE_ESCAPE "\x7E\x7E"
+#define ESC_LEN       2
+
+
+#define CONCAT_ESCAPE_MACRO(args, targetIndex,targetLength,strToAppend, err, len,sourceIndex) UPRV_BLOCK_MACRO_BEGIN {      \
+    while(len-->0){                                                                                                         \
+        if(targetIndex < targetLength){                                                                                     \
+            args->target[targetIndex] = (unsigned char) *strToAppend;                                                       \
+            if(args->offsets!=NULL){                                                                                        \
+                *(offsets++) = sourceIndex-1;                                                                               \
+            }                                                                                                               \
+            targetIndex++;                                                                                                  \
+        }                                                                                                                   \
+        else{                                                                                                               \
+            args->converter->charErrorBuffer[(int)args->converter->charErrorBufferLength++] = (unsigned char) *strToAppend; \
+            *err =U_BUFFER_OVERFLOW_ERROR;                                                                                  \
+        }                                                                                                                   \
+        strToAppend++;                                                                                                      \
+    }                                                                                                                       \
+} UPRV_BLOCK_MACRO_END
+
+
+typedef struct{
+    UConverter* gbConverter;
+    int32_t targetIndex;
+    int32_t sourceIndex;
+    UBool isEscapeAppended;
+    UBool isStateDBCS;
+    UBool isTargetUCharDBCS;
+    UBool isEmptySegment;
+}UConverterDataHZ;
+
+
+U_CDECL_BEGIN
+static void  U_CALLCONV
+_HZOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode){
+    UConverter *gbConverter;
+    if(pArgs->onlyTestIsLoadable) {
+        ucnv_canCreateConverter("GBK", errorCode);  /* errorCode carries result */
+        return;
+    }
+    gbConverter = ucnv_open("GBK", errorCode);
+    if(U_FAILURE(*errorCode)) {
+        return;
+    }
+    cnv->toUnicodeStatus = 0;
+    cnv->fromUnicodeStatus= 0;
+    cnv->mode=0;
+    cnv->fromUChar32=0x0000;
+    cnv->extraInfo = uprv_calloc(1, sizeof(UConverterDataHZ));
+    if(cnv->extraInfo != NULL){
+        ((UConverterDataHZ*)cnv->extraInfo)->gbConverter = gbConverter;
+    }
+    else {
+        ucnv_close(gbConverter);
+        *errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+}
+
+static void  U_CALLCONV
+_HZClose(UConverter *cnv){
+    if(cnv->extraInfo != NULL) {
+        ucnv_close (((UConverterDataHZ *) (cnv->extraInfo))->gbConverter);
+        if(!cnv->isExtraLocal) {
+            uprv_free(cnv->extraInfo);
+        }
+        cnv->extraInfo = NULL;
+    }
+}
+
+static void  U_CALLCONV
+_HZReset(UConverter *cnv, UConverterResetChoice choice){
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        cnv->toUnicodeStatus = 0;
+        cnv->mode=0;
+        if(cnv->extraInfo != NULL){
+            ((UConverterDataHZ*)cnv->extraInfo)->isStateDBCS = FALSE;
+            ((UConverterDataHZ*)cnv->extraInfo)->isEmptySegment = FALSE;
+        }
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        cnv->fromUnicodeStatus= 0;
+        cnv->fromUChar32=0x0000; 
+        if(cnv->extraInfo != NULL){
+            ((UConverterDataHZ*)cnv->extraInfo)->isEscapeAppended = FALSE;
+            ((UConverterDataHZ*)cnv->extraInfo)->targetIndex = 0;
+            ((UConverterDataHZ*)cnv->extraInfo)->sourceIndex = 0;
+            ((UConverterDataHZ*)cnv->extraInfo)->isTargetUCharDBCS = FALSE;
+        }
+    }
+}
+
+/**************************************HZ Encoding*************************************************
+* Rules for HZ encoding
+* 
+*   In ASCII mode, a byte is interpreted as an ASCII character, unless a
+*   '~' is encountered. The character '~' is an escape character. By
+*   convention, it must be immediately followed ONLY by '~', '{' or '\n'
+*   (<LF>), with the following special meaning.
+
+*   1. The escape sequence '~~' is interpreted as a '~'.
+*   2. The escape-to-GB sequence '~{' switches the mode from ASCII to GB.
+*   3. The escape sequence '~\n' is a line-continuation marker to be
+*     consumed with no output produced.
+*   In GB mode, characters are interpreted two bytes at a time as (pure)
+*   GB codes until the escape-from-GB code '~}' is read. This code
+*   switches the mode from GB back to ASCII.  (Note that the escape-
+*   from-GB code '~}' ($7E7D) is outside the defined GB range.)
+*
+*   Source: RFC 1842
+*
+*   Note that the formal syntax in RFC 1842 is invalid. I assume that the
+*   intended definition of single-byte-segment is as follows (pedberg):
+*   single-byte-segment = single-byte-seq 1*single-byte-char
+*/
+
+
+static void  U_CALLCONV
+UConverter_toUnicode_HZ_OFFSETS_LOGIC(UConverterToUnicodeArgs *args,
+                                                            UErrorCode* err){
+    char tempBuf[2];
+    const char *mySource = ( char *) args->source;
+    UChar *myTarget = args->target;
+    const char *mySourceLimit = args->sourceLimit;
+    UChar32 targetUniChar = 0x0000;
+    int32_t mySourceChar = 0x0000;
+    UConverterDataHZ* myData=(UConverterDataHZ*)(args->converter->extraInfo);
+    tempBuf[0]=0; 
+    tempBuf[1]=0;
+
+    /* Calling code already handles this situation. */
+    /*if ((args->converter == NULL) || (args->targetLimit < args->target) || (mySourceLimit < args->source)){
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }*/
+    
+    while(mySource< mySourceLimit){
+        
+        if(myTarget < args->targetLimit){
+            
+            mySourceChar= (unsigned char) *mySource++;
+
+            if(args->converter->mode == UCNV_TILDE) {
+                /* second byte after ~ */
+                args->converter->mode=0;
+                switch(mySourceChar) {
+                case 0x0A:
+                    /* no output for ~\n (line-continuation marker) */
+                    continue;
+                case UCNV_TILDE:
+                    if(args->offsets) {
+                        args->offsets[myTarget - args->target]=(int32_t)(mySource - args->source - 2);
+                    }
+                    *(myTarget++)=(UChar)mySourceChar;
+                    myData->isEmptySegment = FALSE;
+                    continue;
+                case UCNV_OPEN_BRACE:
+                case UCNV_CLOSE_BRACE:
+                    myData->isStateDBCS = (mySourceChar == UCNV_OPEN_BRACE);
+                    if (myData->isEmptySegment) {
+                        myData->isEmptySegment = FALSE; /* we are handling it, reset to avoid future spurious errors */
+                        *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                        args->converter->toUCallbackReason = UCNV_IRREGULAR;
+                        args->converter->toUBytes[0] = UCNV_TILDE;
+                        args->converter->toUBytes[1] = static_cast<uint8_t>(mySourceChar);
+                        args->converter->toULength = 2;
+                        args->target = myTarget;
+                        args->source = mySource;
+                        return;
+                    }
+                    myData->isEmptySegment = TRUE;
+                    continue;
+                default:
+                     /* if the first byte is equal to TILDE and the trail byte
+                     * is not a valid byte then it is an error condition
+                     */
+                    /*
+                     * Ticket 5691: consistent illegal sequences:
+                     * - We include at least the first byte in the illegal sequence.
+                     * - If any of the non-initial bytes could be the start of a character,
+                     *   we stop the illegal sequence before the first one of those.
+                     */
+                    myData->isEmptySegment = FALSE; /* different error here, reset this to avoid spurious future error */
+                    *err = U_ILLEGAL_ESCAPE_SEQUENCE;
+                    args->converter->toUBytes[0] = UCNV_TILDE;
+                    if( myData->isStateDBCS ?
+                            (0x21 <= mySourceChar && mySourceChar <= 0x7e) :
+                            mySourceChar <= 0x7f
+                    ) {
+                        /* The current byte could be the start of a character: Back it out. */
+                        args->converter->toULength = 1;
+                        --mySource;
+                    } else {
+                        /* Include the current byte in the illegal sequence. */
+                        args->converter->toUBytes[1] = static_cast<uint8_t>(mySourceChar);
+                        args->converter->toULength = 2;
+                    }
+                    args->target = myTarget;
+                    args->source = mySource;
+                    return;
+                }
+            } else if(myData->isStateDBCS) {
+                if(args->converter->toUnicodeStatus == 0x00){
+                    /* lead byte */
+                    if(mySourceChar == UCNV_TILDE) {
+                        args->converter->mode = UCNV_TILDE;
+                    } else {
+                        /* add another bit to distinguish a 0 byte from not having seen a lead byte */
+                        args->converter->toUnicodeStatus = (uint32_t) (mySourceChar | 0x100);
+                        myData->isEmptySegment = FALSE; /* the segment has something, either valid or will produce a different error, so reset this */
+                    }
+                    continue;
+                }
+                else{
+                    /* trail byte */
+                    int leadIsOk, trailIsOk;
+                    uint32_t leadByte = args->converter->toUnicodeStatus & 0xff;
+                    targetUniChar = 0xffff;
+                    /*
+                     * Ticket 5691: consistent illegal sequences:
+                     * - We include at least the first byte in the illegal sequence.
+                     * - If any of the non-initial bytes could be the start of a character,
+                     *   we stop the illegal sequence before the first one of those.
+                     *
+                     * In HZ DBCS, if the second byte is in the 21..7e range,
+                     * we report only the first byte as the illegal sequence.
+                     * Otherwise we convert or report the pair of bytes.
+                     */
+                    leadIsOk = (uint8_t)(leadByte - 0x21) <= (0x7d - 0x21);
+                    trailIsOk = (uint8_t)(mySourceChar - 0x21) <= (0x7e - 0x21);
+                    if (leadIsOk && trailIsOk) {
+                        tempBuf[0] = (char) (leadByte+0x80) ;
+                        tempBuf[1] = (char) (mySourceChar+0x80);
+                        targetUniChar = ucnv_MBCSSimpleGetNextUChar(myData->gbConverter->sharedData,
+                            tempBuf, 2, args->converter->useFallback);
+                        mySourceChar= (leadByte << 8) | mySourceChar;
+                    } else if (trailIsOk) {
+                        /* report a single illegal byte and continue with the following DBCS starter byte */
+                        --mySource;
+                        mySourceChar = (int32_t)leadByte;
+                    } else {
+                        /* report a pair of illegal bytes if the second byte is not a DBCS starter */
+                        /* add another bit so that the code below writes 2 bytes in case of error */
+                        mySourceChar= 0x10000 | (leadByte << 8) | mySourceChar;
+                    }
+                    args->converter->toUnicodeStatus =0x00;
+                }
+            }
+            else{
+                if(mySourceChar == UCNV_TILDE) {
+                    args->converter->mode = UCNV_TILDE;
+                    continue;
+                } else if(mySourceChar <= 0x7f) {
+                    targetUniChar = (UChar)mySourceChar;  /* ASCII */
+                    myData->isEmptySegment = FALSE; /* the segment has something valid */
+                } else {
+                    targetUniChar = 0xffff;
+                    myData->isEmptySegment = FALSE; /* different error here, reset this to avoid spurious future error */
+                }
+            }
+            if(targetUniChar < 0xfffe){
+                if(args->offsets) {
+                    args->offsets[myTarget - args->target]=(int32_t)(mySource - args->source - 1-(myData->isStateDBCS));
+                }
+
+                *(myTarget++)=(UChar)targetUniChar;
+            }
+            else /* targetUniChar>=0xfffe */ {
+                if(targetUniChar == 0xfffe){
+                    *err = U_INVALID_CHAR_FOUND;
+                }
+                else{
+                    *err = U_ILLEGAL_CHAR_FOUND;
+                }
+                if(mySourceChar > 0xff){
+                    args->converter->toUBytes[0] = (uint8_t)(mySourceChar >> 8);
+                    args->converter->toUBytes[1] = (uint8_t)mySourceChar;
+                    args->converter->toULength=2;
+                }
+                else{
+                    args->converter->toUBytes[0] = (uint8_t)mySourceChar;
+                    args->converter->toULength=1;
+                }
+                break;
+            }
+        }
+        else{
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    args->target = myTarget;
+    args->source = mySource;
+}
+
+
+static void  U_CALLCONV
+UConverter_fromUnicode_HZ_OFFSETS_LOGIC (UConverterFromUnicodeArgs * args,
+                                                      UErrorCode * err){
+    const UChar *mySource = args->source;
+    char *myTarget = args->target;
+    int32_t* offsets = args->offsets;
+    int32_t mySourceIndex = 0;
+    int32_t myTargetIndex = 0;
+    int32_t targetLength = (int32_t)(args->targetLimit - myTarget);
+    int32_t mySourceLength = (int32_t)(args->sourceLimit - args->source);
+    uint32_t targetUniChar = 0x0000;
+    UChar32 mySourceChar = 0x0000;
+    UConverterDataHZ *myConverterData=(UConverterDataHZ*)args->converter->extraInfo;
+    UBool isTargetUCharDBCS = (UBool) myConverterData->isTargetUCharDBCS;
+    UBool oldIsTargetUCharDBCS;
+    int len =0;
+    const char* escSeq=NULL;
+    
+    /* Calling code already handles this situation. */
+    /*if ((args->converter == NULL) || (args->targetLimit < myTarget) || (args->sourceLimit < args->source)){
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }*/
+    if(args->converter->fromUChar32!=0 && myTargetIndex < targetLength) {
+        goto getTrail;
+    }
+    /*writing the char to the output stream */
+    while (mySourceIndex < mySourceLength){
+        targetUniChar = missingCharMarker;
+        if (myTargetIndex < targetLength){
+            
+            mySourceChar = (UChar) mySource[mySourceIndex++];
+            
+
+            oldIsTargetUCharDBCS = isTargetUCharDBCS;
+            if(mySourceChar ==UCNV_TILDE){
+                /*concatEscape(args, &myTargetIndex, &targetLength,"\x7E\x7E",err,2,&mySourceIndex);*/
+                len = ESC_LEN;
+                escSeq = TILDE_ESCAPE;
+                CONCAT_ESCAPE_MACRO(args, myTargetIndex, targetLength, escSeq,err,len,mySourceIndex);
+                continue;
+            } else if(mySourceChar <= 0x7f) {
+                targetUniChar = mySourceChar;
+            } else {
+                int32_t length= ucnv_MBCSFromUChar32(myConverterData->gbConverter->sharedData,
+                    mySourceChar,&targetUniChar,args->converter->useFallback);
+                /* we can only use lead bytes 21..7D and trail bytes 21..7E */
+                if( length == 2 &&
+                    (uint16_t)(targetUniChar - 0xa1a1) <= (0xfdfe - 0xa1a1) &&
+                    (uint8_t)(targetUniChar - 0xa1) <= (0xfe - 0xa1)
+                ) {
+                    targetUniChar -= 0x8080;
+                } else {
+                    targetUniChar = missingCharMarker;
+                }
+            }
+            if (targetUniChar != missingCharMarker){
+               myConverterData->isTargetUCharDBCS = isTargetUCharDBCS = (UBool)(targetUniChar>0x00FF);     
+                 if(oldIsTargetUCharDBCS != isTargetUCharDBCS || !myConverterData->isEscapeAppended ){
+                    /*Shifting from a double byte to single byte mode*/
+                    if(!isTargetUCharDBCS){
+                        len =ESC_LEN;
+                        escSeq = SB_ESCAPE;
+                        CONCAT_ESCAPE_MACRO(args, myTargetIndex, targetLength, escSeq,err,len,mySourceIndex);
+                        myConverterData->isEscapeAppended = TRUE;
+                    }
+                    else{ /* Shifting from a single byte to double byte mode*/
+                        len =ESC_LEN;
+                        escSeq = DB_ESCAPE;
+                        CONCAT_ESCAPE_MACRO(args, myTargetIndex, targetLength, escSeq,err,len,mySourceIndex);
+                        myConverterData->isEscapeAppended = TRUE;
+                        
+                    }
+                }
+            
+                if(isTargetUCharDBCS){
+                    if( myTargetIndex <targetLength){
+                        myTarget[myTargetIndex++] =(char) (targetUniChar >> 8);
+                        if(offsets){
+                            *(offsets++) = mySourceIndex-1;
+                        }
+                        if(myTargetIndex < targetLength){
+                            myTarget[myTargetIndex++] =(char) targetUniChar;
+                            if(offsets){
+                                *(offsets++) = mySourceIndex-1;
+                            }
+                        }else{
+                            args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (char) targetUniChar;
+                            *err = U_BUFFER_OVERFLOW_ERROR;
+                        } 
+                    }else{
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =(char) (targetUniChar >> 8);
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (char) targetUniChar;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+
+                }else{
+                    if( myTargetIndex <targetLength){
+                        myTarget[myTargetIndex++] = (char) (targetUniChar );
+                        if(offsets){
+                            *(offsets++) = mySourceIndex-1;
+                        }
+                        
+                    }else{
+                        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = (char) targetUniChar;
+                        *err = U_BUFFER_OVERFLOW_ERROR;
+                    }
+                }
+
+            }
+            else{
+                /* oops.. the code point is unassigned */
+                /*Handle surrogates */
+                /*check if the char is a First surrogate*/
+                if(U16_IS_SURROGATE(mySourceChar)) {
+                    if(U16_IS_SURROGATE_LEAD(mySourceChar)) {
+                        args->converter->fromUChar32=mySourceChar;
+getTrail:
+                        /*look ahead to find the trail surrogate*/
+                        if(mySourceIndex <  mySourceLength) {
+                            /* test the following code unit */
+                            UChar trail=(UChar) args->source[mySourceIndex];
+                            if(U16_IS_TRAIL(trail)) {
+                                ++mySourceIndex;
+                                mySourceChar=U16_GET_SUPPLEMENTARY(args->converter->fromUChar32, trail);
+                                args->converter->fromUChar32=0x00;
+                                /* there are no surrogates in GB2312*/
+                                *err = U_INVALID_CHAR_FOUND;
+                                /* exit this condition tree */
+                            } else {
+                                /* this is an unmatched lead code unit (1st surrogate) */
+                                /* callback(illegal) */
+                                *err=U_ILLEGAL_CHAR_FOUND;
+                            }
+                        } else {
+                            /* no more input */
+                            *err = U_ZERO_ERROR;
+                        }
+                    } else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *err=U_ILLEGAL_CHAR_FOUND;
+                    }
+                } else {
+                    /* callback(unassigned) for a BMP code point */
+                    *err = U_INVALID_CHAR_FOUND;
+                }
+
+                args->converter->fromUChar32=mySourceChar;
+                break;
+            }
+        }
+        else{
+            *err = U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+        targetUniChar=missingCharMarker;
+    }
+
+    args->target += myTargetIndex;
+    args->source += mySourceIndex;
+    myConverterData->isTargetUCharDBCS = isTargetUCharDBCS;
+}
+
+static void U_CALLCONV
+_HZ_WriteSub(UConverterFromUnicodeArgs *args, int32_t offsetIndex, UErrorCode *err) {
+    UConverter *cnv = args->converter;
+    UConverterDataHZ *convData=(UConverterDataHZ *) cnv->extraInfo;
+    char *p;
+    char buffer[4];
+    p = buffer;
+    
+    if( convData->isTargetUCharDBCS){
+        *p++= UCNV_TILDE;
+        *p++= UCNV_CLOSE_BRACE;
+        convData->isTargetUCharDBCS=FALSE;
+    }
+    *p++= (char)cnv->subChars[0];
+
+    ucnv_cbFromUWriteBytes(args,
+                           buffer, (int32_t)(p - buffer),
+                           offsetIndex, err);
+}
+
+/*
+ * Structure for cloning an HZ converter into a single memory block.
+ */
+struct cloneHZStruct
+{
+    UConverter cnv;
+    UConverter subCnv;
+    UConverterDataHZ mydata;
+};
+
+
+static UConverter *  U_CALLCONV
+_HZ_SafeClone(const UConverter *cnv, 
+              void *stackBuffer, 
+              int32_t *pBufferSize, 
+              UErrorCode *status)
+{
+    struct cloneHZStruct * localClone;
+    int32_t size, bufferSizeNeeded = sizeof(struct cloneHZStruct);
+
+    if (U_FAILURE(*status)){
+        return nullptr;
+    }
+
+    if (*pBufferSize == 0){ /* 'preflighting' request - set needed size into *pBufferSize */
+        *pBufferSize = bufferSizeNeeded;
+        return nullptr;
+    }
+
+    localClone = (struct cloneHZStruct *)stackBuffer;
+    /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
+
+    uprv_memcpy(&localClone->mydata, cnv->extraInfo, sizeof(UConverterDataHZ));
+    localClone->cnv.extraInfo = &localClone->mydata;
+    localClone->cnv.isExtraLocal = TRUE;
+
+    /* deep-clone the sub-converter */
+    size = (int32_t)sizeof(UConverter);
+    ((UConverterDataHZ*)localClone->cnv.extraInfo)->gbConverter =
+        ucnv_safeClone(((UConverterDataHZ*)cnv->extraInfo)->gbConverter, &localClone->subCnv, &size, status);
+
+    return &localClone->cnv;
+}
+
+static void U_CALLCONV
+_HZ_GetUnicodeSet(const UConverter *cnv,
+                  const USetAdder *sa,
+                  UConverterUnicodeSet which,
+                  UErrorCode *pErrorCode) {
+    /* HZ converts all of ASCII */
+    sa->addRange(sa->set, 0, 0x7f);
+
+    /* add all of the code points that the sub-converter handles */
+    ucnv_MBCSGetFilteredUnicodeSetForUnicode(
+        ((UConverterDataHZ*)cnv->extraInfo)->gbConverter->sharedData,
+        sa, which, UCNV_SET_FILTER_HZ,
+        pErrorCode);
+}
+U_CDECL_END
+static const UConverterImpl _HZImpl={
+
+    UCNV_HZ,
+    
+    NULL,
+    NULL,
+    
+    _HZOpen,
+    _HZClose,
+    _HZReset,
+    
+    UConverter_toUnicode_HZ_OFFSETS_LOGIC,
+    UConverter_toUnicode_HZ_OFFSETS_LOGIC,
+    UConverter_fromUnicode_HZ_OFFSETS_LOGIC,
+    UConverter_fromUnicode_HZ_OFFSETS_LOGIC,
+    NULL,
+    
+    NULL,
+    NULL,
+    _HZ_WriteSub,
+    _HZ_SafeClone,
+    _HZ_GetUnicodeSet,
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _HZStaticData={
+    sizeof(UConverterStaticData),
+        "HZ",
+         0, 
+         UCNV_IBM, 
+         UCNV_HZ, 
+         1, 
+         4,
+        { 0x1a, 0, 0, 0 },
+        1,
+        FALSE, 
+        FALSE,
+        0,
+        0,
+        { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 }, /* reserved */
+
+};
+
+const UConverterSharedData _HZData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_HZStaticData, &_HZImpl);
+
+#endif /* #if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnvisci.cpp b/thirdparty/icu4c/common/ucnvisci.cpp
new file mode 100644
index 0000000000..44a7c05a3c
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvisci.cpp
@@ -0,0 +1,1635 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnvisci.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001JUN26
+*   created by: Ram Viswanadha
+*
+*   Date        Name        Description
+*   24/7/2001   Ram         Added support for EXT character handling
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cstring.h"
+#include "uassert.h"
+
+#define UCNV_OPTIONS_VERSION_MASK 0xf
+#define NUKTA               0x093c
+#define HALANT              0x094d
+#define ZWNJ                0x200c /* Zero Width Non Joiner */
+#define ZWJ                 0x200d /* Zero width Joiner */
+#define INVALID_CHAR        0xffff
+#define ATR                 0xEF   /* Attribute code */
+#define EXT                 0xF0   /* Extension code */
+#define DANDA               0x0964
+#define DOUBLE_DANDA        0x0965
+#define ISCII_NUKTA         0xE9
+#define ISCII_HALANT        0xE8
+#define ISCII_DANDA         0xEA
+#define ISCII_INV           0xD9
+#define ISCII_VOWEL_SIGN_E  0xE0
+#define INDIC_BLOCK_BEGIN   0x0900
+#define INDIC_BLOCK_END     0x0D7F
+#define INDIC_RANGE         (INDIC_BLOCK_END - INDIC_BLOCK_BEGIN)
+#define VOCALLIC_RR         0x0931
+#define LF                  0x0A
+#define ASCII_END           0xA0
+#define NO_CHAR_MARKER      0xFFFE
+#define TELUGU_DELTA        DELTA * TELUGU
+#define DEV_ABBR_SIGN       0x0970
+#define DEV_ANUDATTA        0x0952
+#define EXT_RANGE_BEGIN     0xA1
+#define EXT_RANGE_END       0xEE
+
+#define PNJ_DELTA           0x0100
+#define PNJ_BINDI           0x0A02
+#define PNJ_TIPPI           0x0A70
+#define PNJ_SIGN_VIRAMA     0x0A4D
+#define PNJ_ADHAK           0x0A71
+#define PNJ_HA              0x0A39
+#define PNJ_RRA             0x0A5C
+
+typedef enum {
+    DEVANAGARI =0,
+    BENGALI,
+    GURMUKHI,
+    GUJARATI,
+    ORIYA,
+    TAMIL,
+    TELUGU,
+    KANNADA,
+    MALAYALAM,
+    DELTA=0x80
+}UniLang;
+
+/**
+ * Enumeration for switching code pages if <ATR>+<one of below values>
+ * is encountered
+ */
+typedef enum {
+    DEF = 0x40,
+    RMN = 0x41,
+    DEV = 0x42,
+    BNG = 0x43,
+    TML = 0x44,
+    TLG = 0x45,
+    ASM = 0x46,
+    ORI = 0x47,
+    KND = 0x48,
+    MLM = 0x49,
+    GJR = 0x4A,
+    PNJ = 0x4B,
+    ARB = 0x71,
+    PES = 0x72,
+    URD = 0x73,
+    SND = 0x74,
+    KSM = 0x75,
+    PST = 0x76
+}ISCIILang;
+
+typedef enum {
+    DEV_MASK =0x80,
+    PNJ_MASK =0x40,
+    GJR_MASK =0x20,
+    ORI_MASK =0x10,
+    BNG_MASK =0x08,
+    KND_MASK =0x04,
+    MLM_MASK =0x02,
+    TML_MASK =0x01,
+    ZERO =0x00
+}MaskEnum;
+
+#define ISCII_CNV_PREFIX "ISCII,version="
+
+typedef struct {
+    UChar contextCharToUnicode;         /* previous Unicode codepoint for contextual analysis */
+    UChar contextCharFromUnicode;       /* previous Unicode codepoint for contextual analysis */
+    uint16_t defDeltaToUnicode;         /* delta for switching to default state when DEF is encountered  */
+    uint16_t currentDeltaFromUnicode;   /* current delta in Indic block */
+    uint16_t currentDeltaToUnicode;     /* current delta in Indic block */
+    MaskEnum currentMaskFromUnicode;    /* mask for current state in toUnicode */
+    MaskEnum currentMaskToUnicode;      /* mask for current state in toUnicode */
+    MaskEnum defMaskToUnicode;          /* mask for default state in toUnicode */
+    UBool isFirstBuffer;                /* boolean for fromUnicode to see if we need to announce the first script */
+    UBool resetToDefaultToUnicode;      /* boolean for reseting to default delta and mask when a newline is encountered*/
+    char name[sizeof(ISCII_CNV_PREFIX) + 1];
+    UChar32 prevToUnicodeStatus;        /* Hold the previous toUnicodeStatus. This is necessary because we may need to know the last two code points. */
+} UConverterDataISCII;
+
+typedef struct LookupDataStruct {
+    UniLang uniLang;
+    MaskEnum maskEnum;
+    ISCIILang isciiLang;
+} LookupDataStruct;
+
+static const LookupDataStruct lookupInitialData[]={
+    { DEVANAGARI, DEV_MASK,  DEV },
+    { BENGALI,    BNG_MASK,  BNG },
+    { GURMUKHI,   PNJ_MASK,  PNJ },
+    { GUJARATI,   GJR_MASK,  GJR },
+    { ORIYA,      ORI_MASK,  ORI },
+    { TAMIL,      TML_MASK,  TML },
+    { TELUGU,     KND_MASK,  TLG },
+    { KANNADA,    KND_MASK,  KND },
+    { MALAYALAM,  MLM_MASK,  MLM }
+};
+
+/*
+ * For special handling of certain Gurmukhi characters.
+ * Bit 0 (value 1): PNJ consonant
+ * Bit 1 (value 2): PNJ Bindi Tippi
+ */
+static const uint8_t pnjMap[80] = {
+    /* 0A00..0A0F */
+    0, 0, 0, 0, 0, 2, 0, 2,  0, 0, 0, 0, 0, 0, 0, 0,
+    /* 0A10..0A1F */
+    0, 0, 0, 0, 0, 3, 3, 3,  3, 3, 3, 3, 3, 3, 3, 3,
+    /* 0A20..0A2F */
+    3, 3, 3, 3, 3, 3, 3, 3,  3, 0, 3, 3, 3, 3, 3, 3,
+    /* 0A30..0A3F */
+    3, 0, 0, 0, 0, 3, 3, 0,  3, 3, 0, 0, 0, 0, 0, 2,
+    /* 0A40..0A4F */
+    0, 2, 2, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0
+};
+
+static UBool
+isPNJConsonant(UChar32 c) {
+    if (c < 0xa00 || 0xa50 <= c) {
+        return FALSE;
+    } else {
+        return (UBool)(pnjMap[c - 0xa00] & 1);
+    }
+}
+
+static UBool
+isPNJBindiTippi(UChar32 c) {
+    if (c < 0xa00 || 0xa50 <= c) {
+        return FALSE;
+    } else {
+        return (UBool)(pnjMap[c - 0xa00] >> 1);
+    }
+}
+U_CDECL_BEGIN
+static void  U_CALLCONV
+_ISCIIOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode) {
+    if(pArgs->onlyTestIsLoadable) {
+        return;
+    }
+
+    cnv->extraInfo = uprv_malloc(sizeof(UConverterDataISCII));
+
+    if (cnv->extraInfo != NULL) {
+        int32_t len=0;
+        UConverterDataISCII *converterData=
+                (UConverterDataISCII *) cnv->extraInfo;
+        converterData->contextCharToUnicode=NO_CHAR_MARKER;
+        cnv->toUnicodeStatus = missingCharMarker;
+        converterData->contextCharFromUnicode=0x0000;
+        converterData->resetToDefaultToUnicode=FALSE;
+        /* check if the version requested is supported */
+        if ((pArgs->options & UCNV_OPTIONS_VERSION_MASK) < 9) {
+            /* initialize state variables */
+            converterData->currentDeltaFromUnicode
+                    = converterData->currentDeltaToUnicode
+                            = converterData->defDeltaToUnicode = (uint16_t)(lookupInitialData[pArgs->options & UCNV_OPTIONS_VERSION_MASK].uniLang * DELTA);
+
+            converterData->currentMaskFromUnicode
+                    = converterData->currentMaskToUnicode
+                            = converterData->defMaskToUnicode = lookupInitialData[pArgs->options & UCNV_OPTIONS_VERSION_MASK].maskEnum;
+            
+            converterData->isFirstBuffer=TRUE;
+            (void)uprv_strcpy(converterData->name, ISCII_CNV_PREFIX);
+            len = (int32_t)uprv_strlen(converterData->name);
+            converterData->name[len]= (char)((pArgs->options & UCNV_OPTIONS_VERSION_MASK) + '0');
+            converterData->name[len+1]=0;
+            
+            converterData->prevToUnicodeStatus = 0x0000;
+        } else {
+            uprv_free(cnv->extraInfo);
+            cnv->extraInfo = NULL;
+            *errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+
+    } else {
+        *errorCode =U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+static void U_CALLCONV
+_ISCIIClose(UConverter *cnv) {
+    if (cnv->extraInfo!=NULL) {
+        if (!cnv->isExtraLocal) {
+            uprv_free(cnv->extraInfo);
+        }
+        cnv->extraInfo=NULL;
+    }
+}
+
+static const char*  U_CALLCONV
+_ISCIIgetName(const UConverter* cnv) {
+    if (cnv->extraInfo) {
+        UConverterDataISCII* myData= (UConverterDataISCII*)cnv->extraInfo;
+        return myData->name;
+    }
+    return NULL;
+}
+
+static void U_CALLCONV
+_ISCIIReset(UConverter *cnv, UConverterResetChoice choice) {
+    UConverterDataISCII* data =(UConverterDataISCII *) (cnv->extraInfo);
+    if (choice<=UCNV_RESET_TO_UNICODE) {
+        cnv->toUnicodeStatus = missingCharMarker;
+        cnv->mode=0;
+        data->currentDeltaToUnicode=data->defDeltaToUnicode;
+        data->currentMaskToUnicode = data->defMaskToUnicode;
+        data->contextCharToUnicode=NO_CHAR_MARKER;
+        data->prevToUnicodeStatus = 0x0000;
+    }
+    if (choice!=UCNV_RESET_TO_UNICODE) {
+        cnv->fromUChar32=0x0000;
+        data->contextCharFromUnicode=0x00;
+        data->currentMaskFromUnicode=data->defMaskToUnicode;
+        data->currentDeltaFromUnicode=data->defDeltaToUnicode;
+        data->isFirstBuffer=TRUE;
+        data->resetToDefaultToUnicode=FALSE;
+    }
+}
+
+/**
+ * The values in validity table are indexed by the lower bits of Unicode
+ * range 0x0900 - 0x09ff. The values have a structure like:
+ *       ---------------------------------------------------------------
+ *      | DEV   | PNJ   | GJR   | ORI   | BNG   | TLG   | MLM   | TML   |
+ *      |       |       |       |       | ASM   | KND   |       |       |
+ *       ---------------------------------------------------------------
+ * If a code point is valid in a particular script
+ * then that bit is turned on
+ *
+ * Unicode does not distinguish between Bengali and Assamese so we use 1 bit for
+ * to represent these languages
+ *
+ * Telugu and Kannada have same codepoints except for Vocallic_RR which we special case
+ * and combine and use 1 bit to represent these languages.
+ *
+ * TODO: It is probably easier to understand and maintain to change this
+ * to use uint16_t and give each of the 9 Unicode/script blocks its own bit.
+ */
+
+static const uint8_t validityTable[128] = {
+/* This state table is tool generated please do not edit unless you know exactly what you are doing */
+/* Note: This table was edited to mirror the Windows XP implementation */
+/*ISCII:Valid:Unicode */
+/*0xa0 : 0x00: 0x900  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xa1 : 0xb8: 0x901  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0xa2 : 0xfe: 0x902  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa3 : 0xbf: 0x903  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0x00 : 0x00: 0x904  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xa4 : 0xff: 0x905  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa5 : 0xff: 0x906  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa6 : 0xff: 0x907  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa7 : 0xff: 0x908  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa8 : 0xff: 0x909  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xa9 : 0xff: 0x90a  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xaa : 0xfe: 0x90b  */ DEV_MASK + ZERO     + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0x00 : 0x00: 0x90c  */ DEV_MASK + ZERO     + ZERO     + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xae : 0x80: 0x90d  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xab : 0x87: 0x90e  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xac : 0xff: 0x90f  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xad : 0xff: 0x910  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb2 : 0x80: 0x911  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xaf : 0x87: 0x912  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb0 : 0xff: 0x913  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb1 : 0xff: 0x914  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb3 : 0xff: 0x915  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb4 : 0xfe: 0x916  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xb5 : 0xfe: 0x917  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xb6 : 0xfe: 0x918  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xb7 : 0xff: 0x919  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb8 : 0xff: 0x91a  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xb9 : 0xfe: 0x91b  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xba : 0xff: 0x91c  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xbb : 0xfe: 0x91d  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xbc : 0xff: 0x91e  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xbd : 0xff: 0x91f  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xbe : 0xfe: 0x920  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xbf : 0xfe: 0x921  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xc0 : 0xfe: 0x922  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xc1 : 0xff: 0x923  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xc2 : 0xff: 0x924  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xc3 : 0xfe: 0x925  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xc4 : 0xfe: 0x926  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xc5 : 0xfe: 0x927  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xc6 : 0xff: 0x928  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xc7 : 0x81: 0x929  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + TML_MASK ,
+/*0xc8 : 0xff: 0x92a  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xc9 : 0xfe: 0x92b  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xca : 0xfe: 0x92c  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xcb : 0xfe: 0x92d  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xcc : 0xfe: 0x92e  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xcd : 0xff: 0x92f  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xcf : 0xff: 0x930  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd0 : 0x87: 0x931  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + MLM_MASK + TML_MASK ,
+/*0xd1 : 0xff: 0x932  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd2 : 0xb7: 0x933  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd3 : 0x83: 0x934  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + MLM_MASK + TML_MASK ,
+/*0xd4 : 0xff: 0x935  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd5 : 0xfe: 0x936  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0xd6 : 0xbf: 0x937  */ DEV_MASK + ZERO     + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd7 : 0xff: 0x938  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xd8 : 0xff: 0x939  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0x00 : 0x00: 0x93A  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x93B  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xe9 : 0xda: 0x93c  */ DEV_MASK + PNJ_MASK + ZERO     + ORI_MASK + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x93d  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xda : 0xff: 0x93e  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xdb : 0xff: 0x93f  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xdc : 0xff: 0x940  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xdd : 0xff: 0x941  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xde : 0xff: 0x942  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xdf : 0xbe: 0x943  */ DEV_MASK + ZERO     + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0x00 : 0x00: 0x944  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + BNG_MASK + KND_MASK + ZERO     + ZERO     ,
+/*0xe3 : 0x80: 0x945  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xe0 : 0x87: 0x946  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe1 : 0xff: 0x947  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe2 : 0xff: 0x948  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe7 : 0x80: 0x949  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xe4 : 0x87: 0x94a  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe5 : 0xff: 0x94b  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe6 : 0xff: 0x94c  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xe8 : 0xff: 0x94d  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xec : 0x00: 0x94e  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xed : 0x00: 0x94f  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x950  */ DEV_MASK + ZERO     + GJR_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x951  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x952  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x953  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x954  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x955  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + KND_MASK + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x956  */ ZERO     + ZERO     + ZERO     + ORI_MASK + ZERO     + KND_MASK + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x957  */ ZERO     + ZERO     + ZERO     + ORI_MASK + BNG_MASK + ZERO     + MLM_MASK + ZERO     ,
+/*0x00 : 0x00: 0x958  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x959  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x95a  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x95b  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x95c  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x95d  */ DEV_MASK + ZERO     + ZERO     + ORI_MASK + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x95e  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xce : 0x98: 0x95f  */ DEV_MASK + ZERO     + ZERO     + ORI_MASK + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x960  */ DEV_MASK + ZERO     + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0x00 : 0x00: 0x961  */ DEV_MASK + ZERO     + ZERO     + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO     ,
+/*0x00 : 0x00: 0x962  */ DEV_MASK + ZERO     + ZERO     + ZERO     + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0x00 : 0x00: 0x963  */ DEV_MASK + ZERO     + ZERO     + ZERO     + BNG_MASK + ZERO     + ZERO     + ZERO     ,
+/*0xea : 0xf8: 0x964  */ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xeaea : 0x00: 0x965*/ DEV_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*0xf1 : 0xff: 0x966  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf2 : 0xff: 0x967  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf3 : 0xff: 0x968  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf4 : 0xff: 0x969  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf5 : 0xff: 0x96a  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf6 : 0xff: 0x96b  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf7 : 0xff: 0x96c  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf8 : 0xff: 0x96d  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xf9 : 0xff: 0x96e  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0xfa : 0xff: 0x96f  */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
+/*0x00 : 0x80: 0x970  */ DEV_MASK + PNJ_MASK + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     ,
+/*
+ * The length of the array is 128 to provide values for 0x900..0x97f.
+ * The last 15 entries for 0x971..0x97f of the validity table are all zero
+ * because no Indic script uses such Unicode code points.
+ */
+/*0x00 : 0x00: 0x9yz  */ ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO     + ZERO
+};
+
+static const uint16_t fromUnicodeTable[128]={
+    0x00a0 ,/* 0x0900 */
+    0x00a1 ,/* 0x0901 */
+    0x00a2 ,/* 0x0902 */
+    0x00a3 ,/* 0x0903 */
+    0xa4e0 ,/* 0x0904 */
+    0x00a4 ,/* 0x0905 */
+    0x00a5 ,/* 0x0906 */
+    0x00a6 ,/* 0x0907 */
+    0x00a7 ,/* 0x0908 */
+    0x00a8 ,/* 0x0909 */
+    0x00a9 ,/* 0x090a */
+    0x00aa ,/* 0x090b */
+    0xA6E9 ,/* 0x090c */
+    0x00ae ,/* 0x090d */
+    0x00ab ,/* 0x090e */
+    0x00ac ,/* 0x090f */
+    0x00ad ,/* 0x0910 */
+    0x00b2 ,/* 0x0911 */
+    0x00af ,/* 0x0912 */
+    0x00b0 ,/* 0x0913 */
+    0x00b1 ,/* 0x0914 */
+    0x00b3 ,/* 0x0915 */
+    0x00b4 ,/* 0x0916 */
+    0x00b5 ,/* 0x0917 */
+    0x00b6 ,/* 0x0918 */
+    0x00b7 ,/* 0x0919 */
+    0x00b8 ,/* 0x091a */
+    0x00b9 ,/* 0x091b */
+    0x00ba ,/* 0x091c */
+    0x00bb ,/* 0x091d */
+    0x00bc ,/* 0x091e */
+    0x00bd ,/* 0x091f */
+    0x00be ,/* 0x0920 */
+    0x00bf ,/* 0x0921 */
+    0x00c0 ,/* 0x0922 */
+    0x00c1 ,/* 0x0923 */
+    0x00c2 ,/* 0x0924 */
+    0x00c3 ,/* 0x0925 */
+    0x00c4 ,/* 0x0926 */
+    0x00c5 ,/* 0x0927 */
+    0x00c6 ,/* 0x0928 */
+    0x00c7 ,/* 0x0929 */
+    0x00c8 ,/* 0x092a */
+    0x00c9 ,/* 0x092b */
+    0x00ca ,/* 0x092c */
+    0x00cb ,/* 0x092d */
+    0x00cc ,/* 0x092e */
+    0x00cd ,/* 0x092f */
+    0x00cf ,/* 0x0930 */
+    0x00d0 ,/* 0x0931 */
+    0x00d1 ,/* 0x0932 */
+    0x00d2 ,/* 0x0933 */
+    0x00d3 ,/* 0x0934 */
+    0x00d4 ,/* 0x0935 */
+    0x00d5 ,/* 0x0936 */
+    0x00d6 ,/* 0x0937 */
+    0x00d7 ,/* 0x0938 */
+    0x00d8 ,/* 0x0939 */
+    0xFFFF ,/* 0x093A */
+    0xFFFF ,/* 0x093B */
+    0x00e9 ,/* 0x093c */
+    0xEAE9 ,/* 0x093d */
+    0x00da ,/* 0x093e */
+    0x00db ,/* 0x093f */
+    0x00dc ,/* 0x0940 */
+    0x00dd ,/* 0x0941 */
+    0x00de ,/* 0x0942 */
+    0x00df ,/* 0x0943 */
+    0xDFE9 ,/* 0x0944 */
+    0x00e3 ,/* 0x0945 */
+    0x00e0 ,/* 0x0946 */
+    0x00e1 ,/* 0x0947 */
+    0x00e2 ,/* 0x0948 */
+    0x00e7 ,/* 0x0949 */
+    0x00e4 ,/* 0x094a */
+    0x00e5 ,/* 0x094b */
+    0x00e6 ,/* 0x094c */
+    0x00e8 ,/* 0x094d */
+    0x00ec ,/* 0x094e */
+    0x00ed ,/* 0x094f */
+    0xA1E9 ,/* 0x0950 */ /* OM Symbol */
+    0xFFFF ,/* 0x0951 */
+    0xF0B8 ,/* 0x0952 */
+    0xFFFF ,/* 0x0953 */
+    0xFFFF ,/* 0x0954 */
+    0xFFFF ,/* 0x0955 */
+    0xFFFF ,/* 0x0956 */
+    0xFFFF ,/* 0x0957 */
+    0xb3e9 ,/* 0x0958 */
+    0xb4e9 ,/* 0x0959 */
+    0xb5e9 ,/* 0x095a */
+    0xbae9 ,/* 0x095b */
+    0xbfe9 ,/* 0x095c */
+    0xC0E9 ,/* 0x095d */
+    0xc9e9 ,/* 0x095e */
+    0x00ce ,/* 0x095f */
+    0xAAe9 ,/* 0x0960 */
+    0xA7E9 ,/* 0x0961 */
+    0xDBE9 ,/* 0x0962 */
+    0xDCE9 ,/* 0x0963 */
+    0x00ea ,/* 0x0964 */
+    0xeaea ,/* 0x0965 */
+    0x00f1 ,/* 0x0966 */
+    0x00f2 ,/* 0x0967 */
+    0x00f3 ,/* 0x0968 */
+    0x00f4 ,/* 0x0969 */
+    0x00f5 ,/* 0x096a */
+    0x00f6 ,/* 0x096b */
+    0x00f7 ,/* 0x096c */
+    0x00f8 ,/* 0x096d */
+    0x00f9 ,/* 0x096e */
+    0x00fa ,/* 0x096f */
+    0xF0BF ,/* 0x0970 */
+    0xFFFF ,/* 0x0971 */
+    0xFFFF ,/* 0x0972 */
+    0xFFFF ,/* 0x0973 */
+    0xFFFF ,/* 0x0974 */
+    0xFFFF ,/* 0x0975 */
+    0xFFFF ,/* 0x0976 */
+    0xFFFF ,/* 0x0977 */
+    0xFFFF ,/* 0x0978 */
+    0xFFFF ,/* 0x0979 */
+    0xFFFF ,/* 0x097a */
+    0xFFFF ,/* 0x097b */
+    0xFFFF ,/* 0x097c */
+    0xFFFF ,/* 0x097d */
+    0xFFFF ,/* 0x097e */
+    0xFFFF ,/* 0x097f */
+};
+static const uint16_t toUnicodeTable[256]={
+    0x0000,/* 0x00 */
+    0x0001,/* 0x01 */
+    0x0002,/* 0x02 */
+    0x0003,/* 0x03 */
+    0x0004,/* 0x04 */
+    0x0005,/* 0x05 */
+    0x0006,/* 0x06 */
+    0x0007,/* 0x07 */
+    0x0008,/* 0x08 */
+    0x0009,/* 0x09 */
+    0x000a,/* 0x0a */
+    0x000b,/* 0x0b */
+    0x000c,/* 0x0c */
+    0x000d,/* 0x0d */
+    0x000e,/* 0x0e */
+    0x000f,/* 0x0f */
+    0x0010,/* 0x10 */
+    0x0011,/* 0x11 */
+    0x0012,/* 0x12 */
+    0x0013,/* 0x13 */
+    0x0014,/* 0x14 */
+    0x0015,/* 0x15 */
+    0x0016,/* 0x16 */
+    0x0017,/* 0x17 */
+    0x0018,/* 0x18 */
+    0x0019,/* 0x19 */
+    0x001a,/* 0x1a */
+    0x001b,/* 0x1b */
+    0x001c,/* 0x1c */
+    0x001d,/* 0x1d */
+    0x001e,/* 0x1e */
+    0x001f,/* 0x1f */
+    0x0020,/* 0x20 */
+    0x0021,/* 0x21 */
+    0x0022,/* 0x22 */
+    0x0023,/* 0x23 */
+    0x0024,/* 0x24 */
+    0x0025,/* 0x25 */
+    0x0026,/* 0x26 */
+    0x0027,/* 0x27 */
+    0x0028,/* 0x28 */
+    0x0029,/* 0x29 */
+    0x002a,/* 0x2a */
+    0x002b,/* 0x2b */
+    0x002c,/* 0x2c */
+    0x002d,/* 0x2d */
+    0x002e,/* 0x2e */
+    0x002f,/* 0x2f */
+    0x0030,/* 0x30 */
+    0x0031,/* 0x31 */
+    0x0032,/* 0x32 */
+    0x0033,/* 0x33 */
+    0x0034,/* 0x34 */
+    0x0035,/* 0x35 */
+    0x0036,/* 0x36 */
+    0x0037,/* 0x37 */
+    0x0038,/* 0x38 */
+    0x0039,/* 0x39 */
+    0x003A,/* 0x3A */
+    0x003B,/* 0x3B */
+    0x003c,/* 0x3c */
+    0x003d,/* 0x3d */
+    0x003e,/* 0x3e */
+    0x003f,/* 0x3f */
+    0x0040,/* 0x40 */
+    0x0041,/* 0x41 */
+    0x0042,/* 0x42 */
+    0x0043,/* 0x43 */
+    0x0044,/* 0x44 */
+    0x0045,/* 0x45 */
+    0x0046,/* 0x46 */
+    0x0047,/* 0x47 */
+    0x0048,/* 0x48 */
+    0x0049,/* 0x49 */
+    0x004a,/* 0x4a */
+    0x004b,/* 0x4b */
+    0x004c,/* 0x4c */
+    0x004d,/* 0x4d */
+    0x004e,/* 0x4e */
+    0x004f,/* 0x4f */
+    0x0050,/* 0x50 */
+    0x0051,/* 0x51 */
+    0x0052,/* 0x52 */
+    0x0053,/* 0x53 */
+    0x0054,/* 0x54 */
+    0x0055,/* 0x55 */
+    0x0056,/* 0x56 */
+    0x0057,/* 0x57 */
+    0x0058,/* 0x58 */
+    0x0059,/* 0x59 */
+    0x005a,/* 0x5a */
+    0x005b,/* 0x5b */
+    0x005c,/* 0x5c */
+    0x005d,/* 0x5d */
+    0x005e,/* 0x5e */
+    0x005f,/* 0x5f */
+    0x0060,/* 0x60 */
+    0x0061,/* 0x61 */
+    0x0062,/* 0x62 */
+    0x0063,/* 0x63 */
+    0x0064,/* 0x64 */
+    0x0065,/* 0x65 */
+    0x0066,/* 0x66 */
+    0x0067,/* 0x67 */
+    0x0068,/* 0x68 */
+    0x0069,/* 0x69 */
+    0x006a,/* 0x6a */
+    0x006b,/* 0x6b */
+    0x006c,/* 0x6c */
+    0x006d,/* 0x6d */
+    0x006e,/* 0x6e */
+    0x006f,/* 0x6f */
+    0x0070,/* 0x70 */
+    0x0071,/* 0x71 */
+    0x0072,/* 0x72 */
+    0x0073,/* 0x73 */
+    0x0074,/* 0x74 */
+    0x0075,/* 0x75 */
+    0x0076,/* 0x76 */
+    0x0077,/* 0x77 */
+    0x0078,/* 0x78 */
+    0x0079,/* 0x79 */
+    0x007a,/* 0x7a */
+    0x007b,/* 0x7b */
+    0x007c,/* 0x7c */
+    0x007d,/* 0x7d */
+    0x007e,/* 0x7e */
+    0x007f,/* 0x7f */
+    0x0080,/* 0x80 */
+    0x0081,/* 0x81 */
+    0x0082,/* 0x82 */
+    0x0083,/* 0x83 */
+    0x0084,/* 0x84 */
+    0x0085,/* 0x85 */
+    0x0086,/* 0x86 */
+    0x0087,/* 0x87 */
+    0x0088,/* 0x88 */
+    0x0089,/* 0x89 */
+    0x008a,/* 0x8a */
+    0x008b,/* 0x8b */
+    0x008c,/* 0x8c */
+    0x008d,/* 0x8d */
+    0x008e,/* 0x8e */
+    0x008f,/* 0x8f */
+    0x0090,/* 0x90 */
+    0x0091,/* 0x91 */
+    0x0092,/* 0x92 */
+    0x0093,/* 0x93 */
+    0x0094,/* 0x94 */
+    0x0095,/* 0x95 */
+    0x0096,/* 0x96 */
+    0x0097,/* 0x97 */
+    0x0098,/* 0x98 */
+    0x0099,/* 0x99 */
+    0x009a,/* 0x9a */
+    0x009b,/* 0x9b */
+    0x009c,/* 0x9c */
+    0x009d,/* 0x9d */
+    0x009e,/* 0x9e */
+    0x009f,/* 0x9f */
+    0x00A0,/* 0xa0 */
+    0x0901,/* 0xa1 */
+    0x0902,/* 0xa2 */
+    0x0903,/* 0xa3 */
+    0x0905,/* 0xa4 */
+    0x0906,/* 0xa5 */
+    0x0907,/* 0xa6 */
+    0x0908,/* 0xa7 */
+    0x0909,/* 0xa8 */
+    0x090a,/* 0xa9 */
+    0x090b,/* 0xaa */
+    0x090e,/* 0xab */
+    0x090f,/* 0xac */
+    0x0910,/* 0xad */
+    0x090d,/* 0xae */
+    0x0912,/* 0xaf */
+    0x0913,/* 0xb0 */
+    0x0914,/* 0xb1 */
+    0x0911,/* 0xb2 */
+    0x0915,/* 0xb3 */
+    0x0916,/* 0xb4 */
+    0x0917,/* 0xb5 */
+    0x0918,/* 0xb6 */
+    0x0919,/* 0xb7 */
+    0x091a,/* 0xb8 */
+    0x091b,/* 0xb9 */
+    0x091c,/* 0xba */
+    0x091d,/* 0xbb */
+    0x091e,/* 0xbc */
+    0x091f,/* 0xbd */
+    0x0920,/* 0xbe */
+    0x0921,/* 0xbf */
+    0x0922,/* 0xc0 */
+    0x0923,/* 0xc1 */
+    0x0924,/* 0xc2 */
+    0x0925,/* 0xc3 */
+    0x0926,/* 0xc4 */
+    0x0927,/* 0xc5 */
+    0x0928,/* 0xc6 */
+    0x0929,/* 0xc7 */
+    0x092a,/* 0xc8 */
+    0x092b,/* 0xc9 */
+    0x092c,/* 0xca */
+    0x092d,/* 0xcb */
+    0x092e,/* 0xcc */
+    0x092f,/* 0xcd */
+    0x095f,/* 0xce */
+    0x0930,/* 0xcf */
+    0x0931,/* 0xd0 */
+    0x0932,/* 0xd1 */
+    0x0933,/* 0xd2 */
+    0x0934,/* 0xd3 */
+    0x0935,/* 0xd4 */
+    0x0936,/* 0xd5 */
+    0x0937,/* 0xd6 */
+    0x0938,/* 0xd7 */
+    0x0939,/* 0xd8 */
+    0x200D,/* 0xd9 */
+    0x093e,/* 0xda */
+    0x093f,/* 0xdb */
+    0x0940,/* 0xdc */
+    0x0941,/* 0xdd */
+    0x0942,/* 0xde */
+    0x0943,/* 0xdf */
+    0x0946,/* 0xe0 */
+    0x0947,/* 0xe1 */
+    0x0948,/* 0xe2 */
+    0x0945,/* 0xe3 */
+    0x094a,/* 0xe4 */
+    0x094b,/* 0xe5 */
+    0x094c,/* 0xe6 */
+    0x0949,/* 0xe7 */
+    0x094d,/* 0xe8 */
+    0x093c,/* 0xe9 */
+    0x0964,/* 0xea */
+    0xFFFF,/* 0xeb */
+    0xFFFF,/* 0xec */
+    0xFFFF,/* 0xed */
+    0xFFFF,/* 0xee */
+    0xFFFF,/* 0xef */
+    0xFFFF,/* 0xf0 */
+    0x0966,/* 0xf1 */
+    0x0967,/* 0xf2 */
+    0x0968,/* 0xf3 */
+    0x0969,/* 0xf4 */
+    0x096a,/* 0xf5 */
+    0x096b,/* 0xf6 */
+    0x096c,/* 0xf7 */
+    0x096d,/* 0xf8 */
+    0x096e,/* 0xf9 */
+    0x096f,/* 0xfa */
+    0xFFFF,/* 0xfb */
+    0xFFFF,/* 0xfc */
+    0xFFFF,/* 0xfd */
+    0xFFFF,/* 0xfe */
+    0xFFFF /* 0xff */
+};
+
+static const uint16_t vowelSignESpecialCases[][2]={
+	{ 2 /*length of array*/    , 0      },
+	{ 0xA4 , 0x0904 },
+};
+
+static const uint16_t nuktaSpecialCases[][2]={
+    { 16 /*length of array*/   , 0      },
+    { 0xA6 , 0x090c },
+    { 0xEA , 0x093D },
+    { 0xDF , 0x0944 },
+    { 0xA1 , 0x0950 },
+    { 0xb3 , 0x0958 },
+    { 0xb4 , 0x0959 },
+    { 0xb5 , 0x095a },
+    { 0xba , 0x095b },
+    { 0xbf , 0x095c },
+    { 0xC0 , 0x095d },
+    { 0xc9 , 0x095e },
+    { 0xAA , 0x0960 },
+    { 0xA7 , 0x0961 },
+    { 0xDB , 0x0962 },
+    { 0xDC , 0x0963 },
+};
+
+
+#define WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t offset = (int32_t)(source - args->source-1);                                        \
+      /* write the targetUniChar  to target */                                                  \
+    if(target < targetLimit){                                                                   \
+        if(targetByteUnit <= 0xFF){                                                             \
+            *(target)++ = (uint8_t)(targetByteUnit);                                            \
+            if(offsets){                                                                        \
+                *(offsets++) = offset;                                                          \
+            }                                                                                   \
+        }else{                                                                                  \
+            if (targetByteUnit > 0xFFFF) {                                                      \
+                *(target)++ = (uint8_t)(targetByteUnit>>16);                                    \
+                if (offsets) {                                                                  \
+                    --offset;                                                                   \
+                    *(offsets++) = offset;                                                      \
+                }                                                                               \
+            }                                                                                   \
+            if (!(target < targetLimit)) {                                                      \
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =    \
+                                (uint8_t)(targetByteUnit >> 8);                                 \
+                args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =    \
+                                (uint8_t)targetByteUnit;                                        \
+                *err = U_BUFFER_OVERFLOW_ERROR;                                                 \
+            } else {                                                                            \
+                *(target)++ = (uint8_t)(targetByteUnit>>8);                                     \
+                if(offsets){                                                                    \
+                    *(offsets++) = offset;                                                      \
+                }                                                                               \
+                if(target < targetLimit){                                                       \
+                    *(target)++ = (uint8_t)  targetByteUnit;                                    \
+                    if(offsets){                                                                \
+                        *(offsets++) = offset                            ;                      \
+                    }                                                                           \
+                }else{                                                                          \
+                    args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =\
+                                (uint8_t) (targetByteUnit);                                     \
+                    *err = U_BUFFER_OVERFLOW_ERROR;                                             \
+                }                                                                               \
+            }                                                                                   \
+        }                                                                                       \
+    }else{                                                                                      \
+        if (targetByteUnit & 0xFF0000) {                                                        \
+            args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =        \
+                        (uint8_t) (targetByteUnit >>16);                                        \
+        }                                                                                       \
+        if(targetByteUnit & 0xFF00){                                                            \
+            args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =        \
+                        (uint8_t) (targetByteUnit >>8);                                         \
+        }                                                                                       \
+        args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =            \
+                        (uint8_t) (targetByteUnit);                                             \
+        *err = U_BUFFER_OVERFLOW_ERROR;                                                         \
+    }                                                                                           \
+} UPRV_BLOCK_MACRO_END
+
+/* Rules:
+ *    Explicit Halant :
+ *                      <HALANT> + <ZWNJ>
+ *    Soft Halant :
+ *                      <HALANT> + <ZWJ>
+ */
+static void U_CALLCONV
+UConverter_fromUnicode_ISCII_OFFSETS_LOGIC(
+        UConverterFromUnicodeArgs * args, UErrorCode * err) {
+    const UChar *source = args->source;
+    const UChar *sourceLimit = args->sourceLimit;
+    unsigned char *target = (unsigned char *) args->target;
+    unsigned char *targetLimit = (unsigned char *) args->targetLimit;
+    int32_t* offsets = args->offsets;
+    uint32_t targetByteUnit = 0x0000;
+    UChar32 sourceChar = 0x0000;
+    UChar32 tempContextFromUnicode = 0x0000;    /* For special handling of the Gurmukhi script. */
+    UConverterDataISCII *converterData;
+    uint16_t newDelta=0;
+    uint16_t range = 0;
+    UBool deltaChanged = FALSE;
+
+    if ((args->converter == NULL) || (args->targetLimit < args->target) || (args->sourceLimit < args->source)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    /* initialize data */
+    converterData=(UConverterDataISCII*)args->converter->extraInfo;
+    newDelta=converterData->currentDeltaFromUnicode;
+    range = (uint16_t)(newDelta/DELTA);
+
+    if ((sourceChar = args->converter->fromUChar32)!=0) {
+        goto getTrail;
+    }
+
+    /*writing the char to the output stream */
+    while (source < sourceLimit) {
+        /* Write the language code following LF only if LF is not the last character. */
+        if (args->converter->fromUnicodeStatus == LF) {
+            targetByteUnit = ATR<<8;
+            targetByteUnit += (uint8_t) lookupInitialData[range].isciiLang;
+            args->converter->fromUnicodeStatus = 0x0000;
+            /* now append ATR and language code */
+            WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err);
+            if (U_FAILURE(*err)) {
+                break;
+            }
+        }
+        
+        sourceChar = *source++;
+        tempContextFromUnicode = converterData->contextCharFromUnicode;
+        
+        targetByteUnit = missingCharMarker;
+        
+        /*check if input is in ASCII and C0 control codes range*/
+        if (sourceChar <= ASCII_END) {
+            args->converter->fromUnicodeStatus = sourceChar;
+            WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,sourceChar,err);
+            if (U_FAILURE(*err)) {
+                break;
+            }
+            continue;
+        }
+        switch (sourceChar) {
+        case ZWNJ:
+            /* contextChar has HALANT */
+            if (converterData->contextCharFromUnicode) {
+                converterData->contextCharFromUnicode = 0x00;
+                targetByteUnit = ISCII_HALANT;
+            } else {
+                /* consume ZWNJ and continue */
+                converterData->contextCharFromUnicode = 0x00;
+                continue;
+            }
+            break;
+        case ZWJ:
+            /* contextChar has HALANT */
+            if (converterData->contextCharFromUnicode) {
+                targetByteUnit = ISCII_NUKTA;
+            } else {
+                targetByteUnit =ISCII_INV;
+            }
+            converterData->contextCharFromUnicode = 0x00;
+            break;
+        default:
+            /* is the sourceChar in the INDIC_RANGE? */
+            if ((uint16_t)(INDIC_BLOCK_END-sourceChar) <= INDIC_RANGE) {
+                /* Danda and Double Danda are valid in Northern scripts.. since Unicode
+                 * does not include these codepoints in all Northern scrips we need to
+                 * filter them out
+                 */
+                if (sourceChar!= DANDA && sourceChar != DOUBLE_DANDA) {
+                    /* find out to which block the souceChar belongs*/
+                    range =(uint16_t)((sourceChar-INDIC_BLOCK_BEGIN)/DELTA);
+                    newDelta =(uint16_t)(range*DELTA);
+
+                    /* Now are we in the same block as the previous? */
+                    if (newDelta!= converterData->currentDeltaFromUnicode || converterData->isFirstBuffer) {
+                        converterData->currentDeltaFromUnicode = newDelta;
+                        converterData->currentMaskFromUnicode = lookupInitialData[range].maskEnum;
+                        deltaChanged =TRUE;
+                        converterData->isFirstBuffer=FALSE;
+                    }
+                    
+                    if (converterData->currentDeltaFromUnicode == PNJ_DELTA) { 
+                        if (sourceChar == PNJ_TIPPI) {
+                            /* Make sure Tippi is converterd to Bindi. */
+                            sourceChar = PNJ_BINDI;
+                        } else if (sourceChar == PNJ_ADHAK) {
+                            /* This is for consonant cluster handling. */
+                            converterData->contextCharFromUnicode = PNJ_ADHAK;
+                        }
+                        
+                    }
+                    /* Normalize all Indic codepoints to Devanagari and map them to ISCII */
+                    /* now subtract the new delta from sourceChar*/
+                    sourceChar -= converterData->currentDeltaFromUnicode;
+                }
+
+                /* get the target byte unit */
+                targetByteUnit=fromUnicodeTable[(uint8_t)sourceChar];
+
+                /* is the code point valid in current script? */
+                if ((validityTable[(uint8_t)sourceChar] & converterData->currentMaskFromUnicode)==0) {
+                    /* Vocallic RR is assigned in ISCII Telugu and Unicode */
+                    if (converterData->currentDeltaFromUnicode!=(TELUGU_DELTA) || sourceChar!=VOCALLIC_RR) {
+                        targetByteUnit=missingCharMarker;
+                    }
+                }
+
+                if (deltaChanged) {
+                    /* we are in a script block which is different than
+                     * previous sourceChar's script block write ATR and language codes
+                     */
+                    uint32_t temp=0;
+                    temp =(uint16_t)(ATR<<8);
+                    temp += (uint16_t)((uint8_t) lookupInitialData[range].isciiLang);
+                    /* reset */
+                    deltaChanged=FALSE;
+                    /* now append ATR and language code */
+                    WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,temp,err);
+                    if (U_FAILURE(*err)) {
+                        break;
+                    }
+                }
+                
+                if (converterData->currentDeltaFromUnicode == PNJ_DELTA && (sourceChar + PNJ_DELTA) == PNJ_ADHAK) {
+                    continue;
+                }
+            }
+            /* reset context char */
+            converterData->contextCharFromUnicode = 0x00;
+            break;
+        }
+        if (converterData->currentDeltaFromUnicode == PNJ_DELTA && tempContextFromUnicode == PNJ_ADHAK && isPNJConsonant((sourceChar + PNJ_DELTA))) {
+            /* If the previous codepoint is Adhak and the current codepoint is a consonant, the targetByteUnit should be C + Halant + C. */
+            /* reset context char */
+            converterData->contextCharFromUnicode = 0x0000;
+            targetByteUnit = targetByteUnit << 16 | ISCII_HALANT << 8 | targetByteUnit;
+            /* write targetByteUnit to target */
+            WRITE_TO_TARGET_FROM_U(args, offsets, source, target, targetLimit, targetByteUnit,err);
+            if (U_FAILURE(*err)) {
+                break;
+            }
+        } else if (targetByteUnit != missingCharMarker) {
+            if (targetByteUnit==ISCII_HALANT) {
+                converterData->contextCharFromUnicode = (UChar)targetByteUnit;
+            }
+            /* write targetByteUnit to target*/
+            WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err);
+            if (U_FAILURE(*err)) {
+                break;
+            }
+        } else {
+            /* oops.. the code point is unassigned */
+            /*check if the char is a First surrogate*/
+            if (U16_IS_SURROGATE(sourceChar)) {
+                if (U16_IS_SURROGATE_LEAD(sourceChar)) {
+getTrail:
+                    /*look ahead to find the trail surrogate*/
+                    if (source < sourceLimit) {
+                        /* test the following code unit */
+                        UChar trail= (*source);
+                        if (U16_IS_TRAIL(trail)) {
+                            source++;
+                            sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
+                            *err =U_INVALID_CHAR_FOUND;
+                            /* convert this surrogate code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *err=U_ILLEGAL_CHAR_FOUND;
+                        }
+                    } else {
+                        /* no more input */
+                        *err = U_ZERO_ERROR;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *err=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else {
+                /* callback(unassigned) for a BMP code point */
+                *err = U_INVALID_CHAR_FOUND;
+            }
+
+            args->converter->fromUChar32=sourceChar;
+            break;
+        }
+    }/* end while(mySourceIndex<mySourceLength) */
+
+    /*save the state and return */
+    args->source = source;
+    args->target = (char*)target;
+}
+
+static const uint16_t lookupTable[][2]={
+    { ZERO,       ZERO     },     /*DEFALT*/
+    { ZERO,       ZERO     },     /*ROMAN*/
+    { DEVANAGARI, DEV_MASK },
+    { BENGALI,    BNG_MASK },
+    { TAMIL,      TML_MASK },
+    { TELUGU,     KND_MASK },
+    { BENGALI,    BNG_MASK },
+    { ORIYA,      ORI_MASK },
+    { KANNADA,    KND_MASK },
+    { MALAYALAM,  MLM_MASK },
+    { GUJARATI,   GJR_MASK },
+    { GURMUKHI,   PNJ_MASK }
+};
+
+#define WRITE_TO_TARGET_TO_U(args,source,target,offsets,offset,targetUniChar,delta, err) UPRV_BLOCK_MACRO_BEGIN { \
+    /* add offset to current Indic Block */                                              \
+    if(targetUniChar>ASCII_END &&                                                        \
+           targetUniChar != ZWJ &&                                                       \
+           targetUniChar != ZWNJ &&                                                      \
+           targetUniChar != DANDA &&                                                     \
+           targetUniChar != DOUBLE_DANDA){                                               \
+                                                                                         \
+           targetUniChar+=(uint16_t)(delta);                                             \
+    }                                                                                    \
+    /* now write the targetUniChar */                                                    \
+    if(target<args->targetLimit){                                                        \
+        *(target)++ = (UChar)targetUniChar;                                              \
+        if(offsets){                                                                     \
+            *(offsets)++ = (int32_t)(offset);                                            \
+        }                                                                                \
+    }else{                                                                               \
+        args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++] =   \
+            (UChar)targetUniChar;                                                        \
+        *err = U_BUFFER_OVERFLOW_ERROR;                                                  \
+    }                                                                                    \
+} UPRV_BLOCK_MACRO_END
+
+#define GET_MAPPING(sourceChar,targetUniChar,data) UPRV_BLOCK_MACRO_BEGIN {              \
+    targetUniChar = toUnicodeTable[(sourceChar)] ;                                       \
+    /* is the code point valid in current script? */                                     \
+    if(sourceChar> ASCII_END &&                                                          \
+            (validityTable[(targetUniChar & 0x7F)] & data->currentMaskToUnicode)==0){    \
+        /* Vocallic RR is assigne in ISCII Telugu and Unicode */                         \
+        if(data->currentDeltaToUnicode!=(TELUGU_DELTA) ||                                \
+                    targetUniChar!=VOCALLIC_RR){                                         \
+            targetUniChar=missingCharMarker;                                             \
+        }                                                                                \
+    }                                                                                    \
+} UPRV_BLOCK_MACRO_END
+
+/***********
+ *  Rules for ISCII to Unicode converter
+ *  ISCII is stateful encoding. To convert ISCII bytes to Unicode,
+ *  which has both precomposed and decomposed forms characters
+ *  pre-context and post-context need to be considered.
+ *
+ *  Post context
+ *  i)  ATR : Attribute code is used to declare the font and script switching.
+ *      Currently we only switch scripts and font codes consumed without generating an error
+ *  ii) EXT : Extention code is used to declare switching to Sanskrit and for obscure,
+ *      obsolete characters
+ *  Pre context
+ *  i)  Halant: if preceeded by a halant then it is a explicit halant
+ *  ii) Nukta :
+ *       a) if preceeded by a halant then it is a soft halant
+ *       b) if preceeded by specific consonants and the ligatures have pre-composed
+ *          characters in Unicode then convert to pre-composed characters
+ *  iii) Danda: If Danda is preceeded by a Danda then convert to Double Danda
+ *
+ */
+
+static void U_CALLCONV
+UConverter_toUnicode_ISCII_OFFSETS_LOGIC(UConverterToUnicodeArgs *args, UErrorCode* err) {
+    const char *source = ( char *) args->source;
+    UChar *target = args->target;
+    const char *sourceLimit = args->sourceLimit;
+    const UChar* targetLimit = args->targetLimit;
+    uint32_t targetUniChar = 0x0000;
+    uint8_t sourceChar = 0x0000;
+    UConverterDataISCII* data;
+    UChar32* toUnicodeStatus=NULL;
+    UChar32 tempTargetUniChar = 0x0000;
+    UChar* contextCharToUnicode= NULL;
+    UBool found;
+    int i; 
+    int offset = 0;
+
+    if ((args->converter == NULL) || (target < args->target) || (source < args->source)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    data = (UConverterDataISCII*)(args->converter->extraInfo);
+    contextCharToUnicode = &data->contextCharToUnicode; /* contains previous ISCII codepoint visited */
+    toUnicodeStatus = (UChar32*)&args->converter->toUnicodeStatus;/* contains the mapping to Unicode of the above codepoint*/
+
+    while (U_SUCCESS(*err) && source<sourceLimit) {
+
+        targetUniChar = missingCharMarker;
+
+        if (target < targetLimit) {
+            sourceChar = (unsigned char)*(source)++;
+
+            /* look at the post-context preform special processing */
+            if (*contextCharToUnicode==ATR) {
+
+                /* If we have ATR in *contextCharToUnicode then we need to change our
+                 * state to the Indic Script specified by sourceChar
+                 */
+
+                /* check if the sourceChar is supported script range*/
+                if ((uint8_t)(PNJ-sourceChar)<=PNJ-DEV) {
+                    data->currentDeltaToUnicode = (uint16_t)(lookupTable[sourceChar & 0x0F][0] * DELTA);
+                    data->currentMaskToUnicode = (MaskEnum)lookupTable[sourceChar & 0x0F][1];
+                } else if (sourceChar==DEF) {
+                    /* switch back to default */
+                    data->currentDeltaToUnicode = data->defDeltaToUnicode;
+                    data->currentMaskToUnicode = data->defMaskToUnicode;
+                } else {
+                    if ((sourceChar >= 0x21 && sourceChar <= 0x3F)) {
+                        /* these are display codes consume and continue */
+                    } else {
+                        *err =U_ILLEGAL_CHAR_FOUND;
+                        /* reset */
+                        *contextCharToUnicode=NO_CHAR_MARKER;
+                        goto CALLBACK;
+                    }
+                }
+
+                /* reset */
+                *contextCharToUnicode=NO_CHAR_MARKER;
+
+                continue;
+
+            } else if (*contextCharToUnicode==EXT) {
+                /* check if sourceChar is in 0xA1-0xEE range */
+                if ((uint8_t) (EXT_RANGE_END - sourceChar) <= (EXT_RANGE_END - EXT_RANGE_BEGIN)) {
+                    /* We currently support only Anudatta and Devanagari abbreviation sign */
+                    if (sourceChar==0xBF || sourceChar == 0xB8) {
+                        targetUniChar = (sourceChar==0xBF) ? DEV_ABBR_SIGN : DEV_ANUDATTA;
+                        
+                        /* find out if the mapping is valid in this state */
+                        if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
+                            *contextCharToUnicode= NO_CHAR_MARKER;
+
+                            /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                            if (data->prevToUnicodeStatus) {
+                                WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                                data->prevToUnicodeStatus = 0x0000;
+                            }
+                            /* write to target */
+                            WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
+
+                            continue;
+                        }
+                    }
+                    /* byte unit is unassigned */
+                    targetUniChar = missingCharMarker;
+                    *err= U_INVALID_CHAR_FOUND;
+                } else {
+                    /* only 0xA1 - 0xEE are legal after EXT char */
+                    *contextCharToUnicode= NO_CHAR_MARKER;
+                    *err = U_ILLEGAL_CHAR_FOUND;
+                }
+                goto CALLBACK;
+            } else if (*contextCharToUnicode==ISCII_INV) {
+                if (sourceChar==ISCII_HALANT) {
+                    targetUniChar = 0x0020; /* replace with space accoding to Indic FAQ */
+                } else {
+                    targetUniChar = ZWJ;
+                }
+
+                /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                if (data->prevToUnicodeStatus) {
+                    WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                    data->prevToUnicodeStatus = 0x0000;
+                }
+                /* write to target */
+                WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
+                /* reset */
+                *contextCharToUnicode=NO_CHAR_MARKER;
+            }
+
+            /* look at the pre-context and perform special processing */
+            switch (sourceChar) {
+            case ISCII_INV:
+            case EXT:
+            case ATR:
+                *contextCharToUnicode = (UChar)sourceChar;
+
+                if (*toUnicodeStatus != missingCharMarker) {
+                    /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                    if (data->prevToUnicodeStatus) {
+                        WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                        data->prevToUnicodeStatus = 0x0000;
+                    }
+                    WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,data->currentDeltaToUnicode,err);
+                    *toUnicodeStatus = missingCharMarker;
+                }
+                continue;
+            case ISCII_DANDA:
+                /* handle double danda*/
+                if (*contextCharToUnicode== ISCII_DANDA) {
+                    targetUniChar = DOUBLE_DANDA;
+                    /* clear the context */
+                    *contextCharToUnicode = NO_CHAR_MARKER;
+                    *toUnicodeStatus = missingCharMarker;
+                } else {
+                    GET_MAPPING(sourceChar,targetUniChar,data);
+                    *contextCharToUnicode = sourceChar;
+                }
+                break;
+            case ISCII_HALANT:
+                /* handle explicit halant */
+                if (*contextCharToUnicode == ISCII_HALANT) {
+                    targetUniChar = ZWNJ;
+                    /* clear the context */
+                    *contextCharToUnicode = NO_CHAR_MARKER;
+                } else {
+                    GET_MAPPING(sourceChar,targetUniChar,data);
+                    *contextCharToUnicode = sourceChar;
+                }
+                break;
+            case 0x0A:
+            case 0x0D:
+                data->resetToDefaultToUnicode = TRUE;
+                GET_MAPPING(sourceChar,targetUniChar,data)
+                ;
+                *contextCharToUnicode = sourceChar;
+                break;
+
+            case ISCII_VOWEL_SIGN_E:
+                i=1;
+                found=FALSE;
+                for (; i<vowelSignESpecialCases[0][0]; i++) {
+                    U_ASSERT(i<UPRV_LENGTHOF(vowelSignESpecialCases));
+                    if (vowelSignESpecialCases[i][0]==(uint8_t)*contextCharToUnicode) {
+                        targetUniChar=vowelSignESpecialCases[i][1];
+                        found=TRUE;
+                        break;
+                    }
+                }
+                if (found) {
+                    /* find out if the mapping is valid in this state */
+                    if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
+                        /*targetUniChar += data->currentDeltaToUnicode ;*/
+                        *contextCharToUnicode= NO_CHAR_MARKER;
+                        *toUnicodeStatus = missingCharMarker;
+                        break;
+                    }
+                }
+                GET_MAPPING(sourceChar,targetUniChar,data);
+                *contextCharToUnicode = sourceChar;
+                break;
+
+            case ISCII_NUKTA:
+                /* handle soft halant */
+                if (*contextCharToUnicode == ISCII_HALANT) {
+                    targetUniChar = ZWJ;
+                    /* clear the context */
+                    *contextCharToUnicode = NO_CHAR_MARKER;
+                    break;
+                } else if (data->currentDeltaToUnicode == PNJ_DELTA && data->contextCharToUnicode == 0xc0) {
+                    /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                    if (data->prevToUnicodeStatus) {
+                        WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                        data->prevToUnicodeStatus = 0x0000;
+                    }
+                    /* We got here because ISCII_NUKTA was preceded by 0xc0 and we are converting Gurmukhi.
+                     * In that case we must convert (0xc0 0xe9) to (\u0a5c\u0a4d\u0a39).
+                     */
+                    targetUniChar = PNJ_RRA;
+                    WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
+                    if (U_SUCCESS(*err)) {
+                        targetUniChar = PNJ_SIGN_VIRAMA;
+                        WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
+                        if (U_SUCCESS(*err)) {
+                            targetUniChar = PNJ_HA;
+                            WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
+                        } else {
+                            args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_HA;
+                        }
+                    } else {
+                        args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_SIGN_VIRAMA;
+                        args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_HA;
+                    }
+                    *toUnicodeStatus = missingCharMarker;
+                    data->contextCharToUnicode = NO_CHAR_MARKER;
+                    continue;
+                } else {
+                    /* try to handle <CHAR> + ISCII_NUKTA special mappings */
+                    i=1;
+                    found =FALSE;
+                    for (; i<nuktaSpecialCases[0][0]; i++) {
+                        if (nuktaSpecialCases[i][0]==(uint8_t)
+                                *contextCharToUnicode) {
+                            targetUniChar=nuktaSpecialCases[i][1];
+                            found =TRUE;
+                            break;
+                        }
+                    }
+                    if (found) {
+                        /* find out if the mapping is valid in this state */
+                        if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
+                            /*targetUniChar += data->currentDeltaToUnicode ;*/
+                            *contextCharToUnicode= NO_CHAR_MARKER;
+                            *toUnicodeStatus = missingCharMarker;
+                            if (data->currentDeltaToUnicode == PNJ_DELTA) {
+                                /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                                if (data->prevToUnicodeStatus) {
+                                    WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                                    data->prevToUnicodeStatus = 0x0000;
+                                }
+                                WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
+                                continue;
+                            }
+                            break;
+                        }
+                        /* else fall through to default */
+                    }
+                    /* else fall through to default */
+                    U_FALLTHROUGH;
+                }
+            default:GET_MAPPING(sourceChar,targetUniChar,data)
+                ;
+                *contextCharToUnicode = sourceChar;
+                break;
+            }
+
+            if (*toUnicodeStatus != missingCharMarker) {
+                /* Check to make sure that consonant clusters are handled correct for Gurmukhi script. */
+                if (data->currentDeltaToUnicode == PNJ_DELTA && data->prevToUnicodeStatus != 0 && isPNJConsonant(data->prevToUnicodeStatus) &&
+                        (*toUnicodeStatus + PNJ_DELTA) == PNJ_SIGN_VIRAMA && ((UChar32)(targetUniChar + PNJ_DELTA) == data->prevToUnicodeStatus)) {
+                    /* Consonant clusters C + HALANT + C should be encoded as ADHAK + C */
+                    offset = (int)(source-args->source - 3);
+                    tempTargetUniChar = PNJ_ADHAK; /* This is necessary to avoid some compiler warnings. */
+                    WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,offset,tempTargetUniChar,0,err);
+                    WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,offset,data->prevToUnicodeStatus,0,err);
+                    data->prevToUnicodeStatus = 0x0000; /* reset the previous unicode code point */
+                    *toUnicodeStatus = missingCharMarker;
+                    continue;
+                } else {
+                    /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
+                    if (data->prevToUnicodeStatus) {
+                        WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
+                        data->prevToUnicodeStatus = 0x0000;
+                    }
+                    /* Check to make sure that Bindi and Tippi are handled correctly for Gurmukhi script. 
+                     * If 0xA2 is preceded by a codepoint in the PNJ_BINDI_TIPPI_SET then the target codepoint should be Tippi instead of Bindi.
+                     */
+                    if (data->currentDeltaToUnicode == PNJ_DELTA && (targetUniChar + PNJ_DELTA) == PNJ_BINDI && isPNJBindiTippi((*toUnicodeStatus + PNJ_DELTA))) {
+                        targetUniChar = PNJ_TIPPI - PNJ_DELTA;
+                        WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,PNJ_DELTA,err);
+                    } else if (data->currentDeltaToUnicode == PNJ_DELTA && (targetUniChar + PNJ_DELTA) == PNJ_SIGN_VIRAMA && isPNJConsonant((*toUnicodeStatus + PNJ_DELTA))) {
+                        /* Store the current toUnicodeStatus code point for later handling of consonant cluster in Gurmukhi. */
+                        data->prevToUnicodeStatus = *toUnicodeStatus + PNJ_DELTA;
+                    } else {
+                        /* write the previously mapped codepoint */
+                        WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,data->currentDeltaToUnicode,err);
+                    }
+                }
+                *toUnicodeStatus = missingCharMarker;
+            }
+
+            if (targetUniChar != missingCharMarker) {
+                /* now save the targetUniChar for delayed write */
+                *toUnicodeStatus = (UChar) targetUniChar;
+                if (data->resetToDefaultToUnicode==TRUE) {
+                    data->currentDeltaToUnicode = data->defDeltaToUnicode;
+                    data->currentMaskToUnicode = data->defMaskToUnicode;
+                    data->resetToDefaultToUnicode=FALSE;
+                }
+            } else {
+
+                /* we reach here only if targetUniChar == missingCharMarker
+                 * so assign codes to reason and err
+                 */
+                *err = U_INVALID_CHAR_FOUND;
+CALLBACK:
+                args->converter->toUBytes[0] = (uint8_t) sourceChar;
+                args->converter->toULength = 1;
+                break;
+            }
+
+        } else {
+            *err =U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    if (U_SUCCESS(*err) && args->flush && source == sourceLimit) {
+        /* end of the input stream */
+        UConverter *cnv = args->converter;
+
+        if (*contextCharToUnicode==ATR || *contextCharToUnicode==EXT || *contextCharToUnicode==ISCII_INV) {
+            /* set toUBytes[] */
+            cnv->toUBytes[0] = (uint8_t)*contextCharToUnicode;
+            cnv->toULength = 1;
+
+            /* avoid looping on truncated sequences */
+            *contextCharToUnicode = NO_CHAR_MARKER;
+        } else {
+            cnv->toULength = 0;
+        }
+
+        if (*toUnicodeStatus != missingCharMarker) {
+            /* output a remaining target character */
+            WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source - args->source -1),*toUnicodeStatus,data->currentDeltaToUnicode,err);
+            *toUnicodeStatus = missingCharMarker;
+        }
+    }
+
+    args->target = target;
+    args->source = source;
+}
+
+/* structure for SafeClone calculations */
+struct cloneISCIIStruct {
+    UConverter cnv;
+    UConverterDataISCII mydata;
+};
+
+static UConverter * U_CALLCONV
+_ISCII_SafeClone(const UConverter *cnv,
+              void *stackBuffer,
+              int32_t *pBufferSize,
+              UErrorCode *status)
+{
+    struct cloneISCIIStruct * localClone;
+    int32_t bufferSizeNeeded = sizeof(struct cloneISCIIStruct);
+
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    if (*pBufferSize == 0) { /* 'preflighting' request - set needed size into *pBufferSize */
+        *pBufferSize = bufferSizeNeeded;
+        return 0;
+    }
+
+    localClone = (struct cloneISCIIStruct *)stackBuffer;
+    /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
+
+    uprv_memcpy(&localClone->mydata, cnv->extraInfo, sizeof(UConverterDataISCII));
+    localClone->cnv.extraInfo = &localClone->mydata;
+    localClone->cnv.isExtraLocal = TRUE;
+
+    return &localClone->cnv;
+}
+
+static void U_CALLCONV
+_ISCIIGetUnicodeSet(const UConverter *cnv,
+                    const USetAdder *sa,
+                    UConverterUnicodeSet which,
+                    UErrorCode *pErrorCode)
+{
+    (void)cnv;
+    (void)which;
+    (void)pErrorCode;
+    int32_t idx, script;
+    uint8_t mask;
+
+    /* Since all ISCII versions allow switching to other ISCII
+    scripts, we add all roundtrippable characters to this set. */
+    sa->addRange(sa->set, 0, ASCII_END);
+    for (script = DEVANAGARI; script <= MALAYALAM; script++) {
+        mask = (uint8_t)(lookupInitialData[script].maskEnum);
+        for (idx = 0; idx < DELTA; idx++) {
+            /* added check for TELUGU character */
+            if ((validityTable[idx] & mask) || (script==TELUGU && idx==0x31)) {
+                sa->add(sa->set, idx + (script * DELTA) + INDIC_BLOCK_BEGIN);
+            }
+        }
+    }
+    sa->add(sa->set, DANDA);
+    sa->add(sa->set, DOUBLE_DANDA);
+    sa->add(sa->set, ZWNJ);
+    sa->add(sa->set, ZWJ);
+}
+U_CDECL_END
+static const UConverterImpl _ISCIIImpl={
+
+    UCNV_ISCII,
+
+    NULL,
+    NULL,
+
+    _ISCIIOpen,
+    _ISCIIClose,
+    _ISCIIReset,
+
+    UConverter_toUnicode_ISCII_OFFSETS_LOGIC,
+    UConverter_toUnicode_ISCII_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISCII_OFFSETS_LOGIC,
+    UConverter_fromUnicode_ISCII_OFFSETS_LOGIC,
+    NULL,
+
+    NULL,
+    _ISCIIgetName,
+    NULL,
+    _ISCII_SafeClone,
+    _ISCIIGetUnicodeSet,
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _ISCIIStaticData={
+    sizeof(UConverterStaticData),
+        "ISCII",
+         0,
+         UCNV_IBM,
+         UCNV_ISCII,
+         1,
+         4,
+        { 0x1a, 0, 0, 0 },
+        0x1,
+        FALSE,
+        FALSE,
+        0x0,
+        0x0,
+        { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 }, /* reserved */
+
+};
+
+const UConverterSharedData _ISCIIData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ISCIIStaticData, &_ISCIIImpl);
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnvlat1.cpp b/thirdparty/icu4c/common/ucnvlat1.cpp
new file mode 100644
index 0000000000..358bc0caa2
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvlat1.cpp
@@ -0,0 +1,756 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/* 
+**********************************************************************
+*   Copyright (C) 2000-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ucnvlat1.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000feb07
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/uset.h"
+#include "unicode/utf8.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "ustr_imp.h"
+
+/* control optimizations according to the platform */
+#define LATIN1_UNROLL_FROM_UNICODE 1
+
+/* ISO 8859-1 --------------------------------------------------------------- */
+
+/* This is a table-less and callback-less version of ucnv_MBCSSingleToBMPWithOffsets(). */
+U_CDECL_BEGIN
+static void U_CALLCONV
+_Latin1ToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    const uint8_t *source;
+    UChar *target;
+    int32_t targetCapacity, length;
+    int32_t *offsets;
+
+    int32_t sourceIndex;
+
+    /* set up the local pointers */
+    source=(const uint8_t *)pArgs->source;
+    target=pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    sourceIndex=0;
+
+    /*
+     * since the conversion here is 1:1 UChar:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)((const uint8_t *)pArgs->sourceLimit-source);
+    if(length<=targetCapacity) {
+        targetCapacity=length;
+    } else {
+        /* target will be full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        length=targetCapacity;
+    }
+
+    if(targetCapacity>=8) {
+        /* This loop is unrolled for speed and improved pipelining. */
+        int32_t count, loops;
+
+        loops=count=targetCapacity>>3;
+        length=targetCapacity&=0x7;
+        do {
+            target[0]=source[0];
+            target[1]=source[1];
+            target[2]=source[2];
+            target[3]=source[3];
+            target[4]=source[4];
+            target[5]=source[5];
+            target[6]=source[6];
+            target[7]=source[7];
+            target+=8;
+            source+=8;
+        } while(--count>0);
+
+        if(offsets!=NULL) {
+            do {
+                offsets[0]=sourceIndex++;
+                offsets[1]=sourceIndex++;
+                offsets[2]=sourceIndex++;
+                offsets[3]=sourceIndex++;
+                offsets[4]=sourceIndex++;
+                offsets[5]=sourceIndex++;
+                offsets[6]=sourceIndex++;
+                offsets[7]=sourceIndex++;
+                offsets+=8;
+            } while(--loops>0);
+        }
+    }
+
+    /* conversion loop */
+    while(targetCapacity>0) {
+        *target++=*source++;
+        --targetCapacity;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+
+    /* set offsets */
+    if(offsets!=NULL) {
+        while(length>0) {
+            *offsets++=sourceIndex++;
+            --length;
+        }
+        pArgs->offsets=offsets;
+    }
+}
+
+/* This is a table-less and callback-less version of ucnv_MBCSSingleGetNextUChar(). */
+static UChar32 U_CALLCONV
+_Latin1GetNextUChar(UConverterToUnicodeArgs *pArgs,
+                    UErrorCode *pErrorCode) {
+    const uint8_t *source=(const uint8_t *)pArgs->source;
+    if(source<(const uint8_t *)pArgs->sourceLimit) {
+        pArgs->source=(const char *)(source+1);
+        return *source;
+    }
+
+    /* no output because of empty input */
+    *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    return 0xffff;
+}
+
+/* This is a table-less version of ucnv_MBCSSingleFromBMPWithOffsets(). */
+static void U_CALLCONV
+_Latin1FromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                              UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target, *oldTarget;
+    int32_t targetCapacity, length;
+    int32_t *offsets;
+
+    UChar32 cp;
+    UChar c, max;
+
+    int32_t sourceIndex;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=oldTarget=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    if(cnv->sharedData==&_Latin1Data) {
+        max=0xff; /* Latin-1 */
+    } else {
+        max=0x7f; /* US-ASCII */
+    }
+
+    /* get the converter state from UConverter */
+    cp=cnv->fromUChar32;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= cp==0 ? 0 : -1;
+
+    /*
+     * since the conversion here is 1:1 UChar:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)(sourceLimit-source);
+    if(length<targetCapacity) {
+        targetCapacity=length;
+    }
+
+    /* conversion loop */
+    if(cp!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+#if LATIN1_UNROLL_FROM_UNICODE
+    /* unroll the loop with the most common case */
+    if(targetCapacity>=16) {
+        int32_t count, loops;
+        UChar u, oredChars;
+
+        loops=count=targetCapacity>>4;
+        do {
+            oredChars=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+            oredChars|=u=*source++;
+            *target++=(uint8_t)u;
+
+            /* were all 16 entries really valid? */
+            if(oredChars>max) {
+                /* no, return to the first of these 16 */
+                source-=16;
+                target-=16;
+                break;
+            }
+        } while(--count>0);
+        count=loops-count;
+        targetCapacity-=16*count;
+
+        if(offsets!=NULL) {
+            oldTarget+=16*count;
+            while(count>0) {
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                --count;
+            }
+        }
+    }
+#endif
+
+    /* conversion loop */
+    c=0;
+    while(targetCapacity>0 && (c=*source++)<=max) {
+        /* convert the Unicode code point */
+        *target++=(uint8_t)c;
+        --targetCapacity;
+    }
+
+    if(c>max) {
+        cp=c;
+        if(!U_IS_SURROGATE(cp)) {
+            /* callback(unassigned) */
+        } else if(U_IS_SURROGATE_LEAD(cp)) {
+getTrail:
+            if(source<sourceLimit) {
+                /* test the following code unit */
+                UChar trail=*source;
+                if(U16_IS_TRAIL(trail)) {
+                    ++source;
+                    cp=U16_GET_SUPPLEMENTARY(cp, trail);
+                    /* this codepage does not map supplementary code points */
+                    /* callback(unassigned) */
+                } else {
+                    /* this is an unmatched lead code unit (1st surrogate) */
+                    /* callback(illegal) */
+                }
+            } else {
+                /* no more input */
+                cnv->fromUChar32=cp;
+                goto noMoreInput;
+            }
+        } else {
+            /* this is an unmatched trail code unit (2nd surrogate) */
+            /* callback(illegal) */
+        }
+
+        *pErrorCode= U_IS_SURROGATE(cp) ? U_ILLEGAL_CHAR_FOUND : U_INVALID_CHAR_FOUND;
+        cnv->fromUChar32=cp;
+    }
+noMoreInput:
+
+    /* set offsets since the start */
+    if(offsets!=NULL) {
+        size_t count=target-oldTarget;
+        while(count>0) {
+            *offsets++=sourceIndex++;
+            --count;
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode) && source<sourceLimit && target>=(uint8_t *)pArgs->targetLimit) {
+        /* target is full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+/* Convert UTF-8 to Latin-1. Adapted from ucnv_SBCSFromUTF8(). */
+static void U_CALLCONV
+ucnv_Latin1FromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                    UConverterToUnicodeArgs *pToUArgs,
+                    UErrorCode *pErrorCode) {
+    UConverter *utf8;
+    const uint8_t *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+
+    UChar32 c;
+    uint8_t b, t1;
+
+    /* set up the local pointers */
+    utf8=pToUArgs->converter;
+    source=(uint8_t *)pToUArgs->source;
+    sourceLimit=(uint8_t *)pToUArgs->sourceLimit;
+    target=(uint8_t *)pFromUArgs->target;
+    targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target);
+
+    /* get the converter state from the UTF-8 UConverter */
+    if (utf8->toULength > 0) {
+        c=(UChar32)utf8->toUnicodeStatus;
+    } else {
+        c = 0;
+    }
+    if(c!=0 && source<sourceLimit) {
+        if(targetCapacity==0) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return;
+        } else if(c>=0xc2 && c<=0xc3 && (t1=(uint8_t)(*source-0x80)) <= 0x3f) {
+            ++source;
+            *target++=(uint8_t)(((c&3)<<6)|t1);
+            --targetCapacity;
+
+            utf8->toUnicodeStatus=0;
+            utf8->toULength=0;
+        } else {
+            /* complicated, illegal or unmappable input: fall back to the pivoting implementation */
+            *pErrorCode=U_USING_DEFAULT_WARNING;
+            return;
+        }
+    }
+
+    /*
+     * Make sure that the last byte sequence before sourceLimit is complete
+     * or runs into a lead byte.
+     * In the conversion loop compare source with sourceLimit only once
+     * per multi-byte character.
+     * For Latin-1, adjust sourceLimit only for 1 trail byte because
+     * the conversion loop handles at most 2-byte sequences.
+     */
+    if(source<sourceLimit && U8_IS_LEAD(*(sourceLimit-1))) {
+        --sourceLimit;
+    }
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        if(targetCapacity>0) {
+            b=*source++;
+            if(U8_IS_SINGLE(b)) {
+                /* convert ASCII */
+                *target++=(uint8_t)b;
+                --targetCapacity;
+            } else if( /* handle U+0080..U+00FF inline */
+                       b>=0xc2 && b<=0xc3 &&
+                       (t1=(uint8_t)(*source-0x80)) <= 0x3f
+            ) {
+                ++source;
+                *target++=(uint8_t)(((b&3)<<6)|t1);
+                --targetCapacity;
+            } else {
+                /* complicated, illegal or unmappable input: fall back to the pivoting implementation */
+                pToUArgs->source=(char *)(source-1);
+                pFromUArgs->target=(char *)target;
+                *pErrorCode=U_USING_DEFAULT_WARNING;
+                return;
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /*
+     * The sourceLimit may have been adjusted before the conversion loop
+     * to stop before a truncated sequence.
+     * If so, then collect the truncated sequence now.
+     * For Latin-1, there is at most exactly one lead byte because of the
+     * smaller sourceLimit adjustment logic.
+     */
+    if(U_SUCCESS(*pErrorCode) && source<(sourceLimit=(uint8_t *)pToUArgs->sourceLimit)) {
+        utf8->toUnicodeStatus=utf8->toUBytes[0]=b=*source++;
+        utf8->toULength=1;
+        utf8->mode=U8_COUNT_BYTES(b);
+    }
+
+    /* write back the updated pointers */
+    pToUArgs->source=(char *)source;
+    pFromUArgs->target=(char *)target;
+}
+
+static void U_CALLCONV
+_Latin1GetUnicodeSet(const UConverter *cnv,
+                     const USetAdder *sa,
+                     UConverterUnicodeSet which,
+                     UErrorCode *pErrorCode) {
+    (void)cnv;
+    (void)which;
+    (void)pErrorCode;
+    sa->addRange(sa->set, 0, 0xff);
+}
+U_CDECL_END
+
+
+static const UConverterImpl _Latin1Impl={
+    UCNV_LATIN_1,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    _Latin1ToUnicodeWithOffsets,
+    _Latin1ToUnicodeWithOffsets,
+    _Latin1FromUnicodeWithOffsets,
+    _Latin1FromUnicodeWithOffsets,
+    _Latin1GetNextUChar,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    _Latin1GetUnicodeSet,
+
+    NULL,
+    ucnv_Latin1FromUTF8
+};
+
+static const UConverterStaticData _Latin1StaticData={
+    sizeof(UConverterStaticData),
+    "ISO-8859-1",
+    819, UCNV_IBM, UCNV_LATIN_1, 1, 1,
+    { 0x1a, 0, 0, 0 }, 1, FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _Latin1Data=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_Latin1StaticData, &_Latin1Impl);
+
+/* US-ASCII ----------------------------------------------------------------- */
+
+U_CDECL_BEGIN
+/* This is a table-less version of ucnv_MBCSSingleToBMPWithOffsets(). */
+static void U_CALLCONV
+_ASCIIToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                           UErrorCode *pErrorCode) {
+    const uint8_t *source, *sourceLimit;
+    UChar *target, *oldTarget;
+    int32_t targetCapacity, length;
+    int32_t *offsets;
+
+    int32_t sourceIndex;
+
+    uint8_t c;
+
+    /* set up the local pointers */
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=oldTarget=pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=0;
+
+    /*
+     * since the conversion here is 1:1 UChar:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)(sourceLimit-source);
+    if(length<targetCapacity) {
+        targetCapacity=length;
+    }
+
+    if(targetCapacity>=8) {
+        /* This loop is unrolled for speed and improved pipelining. */
+        int32_t count, loops;
+        UChar oredChars;
+
+        loops=count=targetCapacity>>3;
+        do {
+            oredChars=target[0]=source[0];
+            oredChars|=target[1]=source[1];
+            oredChars|=target[2]=source[2];
+            oredChars|=target[3]=source[3];
+            oredChars|=target[4]=source[4];
+            oredChars|=target[5]=source[5];
+            oredChars|=target[6]=source[6];
+            oredChars|=target[7]=source[7];
+
+            /* were all 16 entries really valid? */
+            if(oredChars>0x7f) {
+                /* no, return to the first of these 16 */
+                break;
+            }
+            source+=8;
+            target+=8;
+        } while(--count>0);
+        count=loops-count;
+        targetCapacity-=count*8;
+
+        if(offsets!=NULL) {
+            oldTarget+=count*8;
+            while(count>0) {
+                offsets[0]=sourceIndex++;
+                offsets[1]=sourceIndex++;
+                offsets[2]=sourceIndex++;
+                offsets[3]=sourceIndex++;
+                offsets[4]=sourceIndex++;
+                offsets[5]=sourceIndex++;
+                offsets[6]=sourceIndex++;
+                offsets[7]=sourceIndex++;
+                offsets+=8;
+                --count;
+            }
+        }
+    }
+
+    /* conversion loop */
+    c=0;
+    while(targetCapacity>0 && (c=*source++)<=0x7f) {
+        *target++=c;
+        --targetCapacity;
+    }
+
+    if(c>0x7f) {
+        /* callback(illegal); copy the current bytes to toUBytes[] */
+        UConverter *cnv=pArgs->converter;
+        cnv->toUBytes[0]=c;
+        cnv->toULength=1;
+        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+    } else if(source<sourceLimit && target>=pArgs->targetLimit) {
+        /* target is full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* set offsets since the start */
+    if(offsets!=NULL) {
+        size_t count=target-oldTarget;
+        while(count>0) {
+            *offsets++=sourceIndex++;
+            --count;
+        }
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+/* This is a table-less version of ucnv_MBCSSingleGetNextUChar(). */
+static UChar32 U_CALLCONV
+_ASCIIGetNextUChar(UConverterToUnicodeArgs *pArgs,
+                   UErrorCode *pErrorCode) {
+    const uint8_t *source;
+    uint8_t b;
+
+    source=(const uint8_t *)pArgs->source;
+    if(source<(const uint8_t *)pArgs->sourceLimit) {
+        b=*source++;
+        pArgs->source=(const char *)source;
+        if(b<=0x7f) {
+            return b;
+        } else {
+            UConverter *cnv=pArgs->converter;
+            cnv->toUBytes[0]=b;
+            cnv->toULength=1;
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            return 0xffff;
+        }
+    }
+
+    /* no output because of empty input */
+    *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    return 0xffff;
+}
+
+/* "Convert" UTF-8 to US-ASCII: Validate and copy. */
+static void U_CALLCONV
+ucnv_ASCIIFromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                   UConverterToUnicodeArgs *pToUArgs,
+                   UErrorCode *pErrorCode) {
+    const uint8_t *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity, length;
+
+    uint8_t c;
+
+    if(pToUArgs->converter->toULength > 0) {
+        /* no handling of partial UTF-8 characters here, fall back to pivoting */
+        *pErrorCode=U_USING_DEFAULT_WARNING;
+        return;
+    }
+
+    /* set up the local pointers */
+    source=(const uint8_t *)pToUArgs->source;
+    sourceLimit=(const uint8_t *)pToUArgs->sourceLimit;
+    target=(uint8_t *)pFromUArgs->target;
+    targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target);
+
+    /*
+     * since the conversion here is 1:1 uint8_t:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)(sourceLimit-source);
+    if(length<targetCapacity) {
+        targetCapacity=length;
+    }
+
+    /* unroll the loop with the most common case */
+    if(targetCapacity>=16) {
+        int32_t count, loops;
+        uint8_t oredChars;
+
+        loops=count=targetCapacity>>4;
+        do {
+            oredChars=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+            oredChars|=*target++=*source++;
+
+            /* were all 16 entries really valid? */
+            if(oredChars>0x7f) {
+                /* no, return to the first of these 16 */
+                source-=16;
+                target-=16;
+                break;
+            }
+        } while(--count>0);
+        count=loops-count;
+        targetCapacity-=16*count;
+    }
+
+    /* conversion loop */
+    c=0;
+    while(targetCapacity>0 && (c=*source)<=0x7f) {
+        ++source;
+        *target++=c;
+        --targetCapacity;
+    }
+
+    if(c>0x7f) {
+        /* non-ASCII character, handle in standard converter */
+        *pErrorCode=U_USING_DEFAULT_WARNING;
+    } else if(source<sourceLimit && target>=(const uint8_t *)pFromUArgs->targetLimit) {
+        /* target is full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* write back the updated pointers */
+    pToUArgs->source=(const char *)source;
+    pFromUArgs->target=(char *)target;
+}
+
+static void U_CALLCONV
+_ASCIIGetUnicodeSet(const UConverter *cnv,
+                    const USetAdder *sa,
+                    UConverterUnicodeSet which,
+                    UErrorCode *pErrorCode) {
+    (void)cnv;
+    (void)which;
+    (void)pErrorCode;
+    sa->addRange(sa->set, 0, 0x7f);
+}
+U_CDECL_END
+
+static const UConverterImpl _ASCIIImpl={
+    UCNV_US_ASCII,
+
+    NULL,
+    NULL,
+
+    NULL,
+    NULL,
+    NULL,
+
+    _ASCIIToUnicodeWithOffsets,
+    _ASCIIToUnicodeWithOffsets,
+    _Latin1FromUnicodeWithOffsets,
+    _Latin1FromUnicodeWithOffsets,
+    _ASCIIGetNextUChar,
+
+    NULL,
+    NULL,
+    NULL,
+    NULL,
+    _ASCIIGetUnicodeSet,
+
+    NULL,
+    ucnv_ASCIIFromUTF8
+};
+
+static const UConverterStaticData _ASCIIStaticData={
+    sizeof(UConverterStaticData),
+    "US-ASCII",
+    367, UCNV_IBM, UCNV_US_ASCII, 1, 1,
+    { 0x1a, 0, 0, 0 }, 1, FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _ASCIIData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_ASCIIStaticData, &_ASCIIImpl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnvmbcs.cpp b/thirdparty/icu4c/common/ucnvmbcs.cpp
new file mode 100644
index 0000000000..ca9b0a335a
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvmbcs.cpp
@@ -0,0 +1,5723 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnvmbcs.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000jul03
+*   created by: Markus W. Scherer
+*
+*   The current code in this file replaces the previous implementation
+*   of conversion code from multi-byte codepages to Unicode and back.
+*   This implementation supports the following:
+*   - legacy variable-length codepages with up to 4 bytes per character
+*   - all Unicode code points (up to 0x10ffff)
+*   - efficient distinction of unassigned vs. illegal byte sequences
+*   - it is possible in fromUnicode() to directly deal with simple
+*     stateful encodings (used for EBCDIC_STATEFUL)
+*   - it is possible to convert Unicode code points
+*     to a single zero byte (but not as a fallback except for SBCS)
+*
+*   Remaining limitations in fromUnicode:
+*   - byte sequences must not have leading zero bytes
+*   - except for SBCS codepages: no fallback mapping from Unicode to a zero byte
+*   - limitation to up to 4 bytes per character
+*
+*   ICU 2.8 (late 2003) adds a secondary data structure which lifts some of these
+*   limitations and adds m:n character mappings and other features.
+*   See ucnv_ext.h for details.
+*
+*   Change history: 
+*
+*    5/6/2001       Ram       Moved  MBCS_SINGLE_RESULT_FROM_U,MBCS_STAGE_2_FROM_U,
+*                             MBCS_VALUE_2_FROM_STAGE_2, MBCS_VALUE_4_FROM_STAGE_2
+*                             macros to ucnvmbcs.h file
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/udata.h"
+#include "unicode/uset.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "ucnv_bld.h"
+#include "ucnvmbcs.h"
+#include "ucnv_ext.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "umutex.h"
+#include "ustr_imp.h"
+
+/* control optimizations according to the platform */
+#define MBCS_UNROLL_SINGLE_TO_BMP 1
+#define MBCS_UNROLL_SINGLE_FROM_BMP 0
+
+/*
+ * _MBCSHeader versions 5.3 & 4.3
+ * (Note that the _MBCSHeader version is in addition to the converter formatVersion.)
+ *
+ * This version is optional. Version 5 is used for incompatible data format changes.
+ * makeconv will continue to generate version 4 files if possible.
+ *
+ * Changes from version 4:
+ *
+ * The main difference is an additional _MBCSHeader field with
+ * - the length (number of uint32_t) of the _MBCSHeader
+ * - flags for further incompatible data format changes
+ * - flags for further, backward compatible data format changes
+ *
+ * The MBCS_OPT_FROM_U flag indicates that most of the fromUnicode data is omitted from
+ * the file and needs to be reconstituted at load time.
+ * This requires a utf8Friendly format with an additional mbcsIndex table for fast
+ * (and UTF-8-friendly) fromUnicode conversion for Unicode code points up to maxFastUChar.
+ * (For details about these structures see below, and see ucnvmbcs.h.)
+ *
+ *   utf8Friendly also implies that the fromUnicode mappings are stored in ascending order
+ *   of the Unicode code points. (This requires that the .ucm file has the |0 etc.
+ *   precision markers for all mappings.)
+ *
+ *   All fallbacks have been moved to the extension table, leaving only roundtrips in the
+ *   omitted data that can be reconstituted from the toUnicode data.
+ *
+ *   Of the stage 2 table, the part corresponding to maxFastUChar and below is omitted.
+ *   With only roundtrip mappings in the base fromUnicode data, this part is fully
+ *   redundant with the mbcsIndex and will be reconstituted from that (also using the
+ *   stage 1 table which contains the information about how stage 2 was compacted).
+ *
+ *   The rest of the stage 2 table, the part for code points above maxFastUChar,
+ *   is stored in the file and will be appended to the reconstituted part.
+ *
+ *   The entire fromUBytes array is omitted from the file and will be reconstitued.
+ *   This is done by enumerating all toUnicode roundtrip mappings, performing
+ *   each mapping (using the stage 1 and reconstituted stage 2 tables) and
+ *   writing instead of reading the byte values.
+ *
+ * _MBCSHeader version 4.3
+ *
+ * Change from version 4.2:
+ * - Optional utf8Friendly data structures, with 64-entry stage 3 block
+ *   allocation for parts of the BMP, and an additional mbcsIndex in non-SBCS
+ *   files which can be used instead of stages 1 & 2.
+ *   Faster lookups for roundtrips from most commonly used characters,
+ *   and lookups from UTF-8 byte sequences with a natural bit distribution.
+ *   See ucnvmbcs.h for more details.
+ *
+ * Change from version 4.1:
+ * - Added an optional extension table structure at the end of the .cnv file.
+ *   It is present if the upper bits of the header flags field contains a non-zero
+ *   byte offset to it.
+ *   Files that contain only a conversion table and no base table
+ *   use the special outputType MBCS_OUTPUT_EXT_ONLY.
+ *   These contain the base table name between the MBCS header and the extension
+ *   data.
+ *
+ * Change from version 4.0:
+ * - Replace header.reserved with header.fromUBytesLength so that all
+ *   fields in the data have length.
+ *
+ * Changes from version 3 (for performance improvements):
+ * - new bit distribution for state table entries
+ * - reordered action codes
+ * - new data structure for single-byte fromUnicode
+ *   + stage 2 only contains indexes
+ *   + stage 3 stores 16 bits per character with classification bits 15..8
+ * - no multiplier for stage 1 entries
+ * - stage 2 for non-single-byte codepages contains the index and the flags in
+ *   one 32-bit value
+ * - 2-byte and 4-byte fromUnicode results are stored directly as 16/32-bit integers
+ *
+ * For more details about old versions of the MBCS data structure, see
+ * the corresponding versions of this file.
+ *
+ * Converting stateless codepage data ---------------------------------------***
+ * (or codepage data with simple states) to Unicode.
+ *
+ * Data structure and algorithm for converting from complex legacy codepages
+ * to Unicode. (Designed before 2000-may-22.)
+ *
+ * The basic idea is that the structure of legacy codepages can be described
+ * with state tables.
+ * When reading a byte stream, each input byte causes a state transition.
+ * Some transitions result in the output of a code point, some result in
+ * "unassigned" or "illegal" output.
+ * This is used here for character conversion.
+ *
+ * The data structure begins with a state table consisting of a row
+ * per state, with 256 entries (columns) per row for each possible input
+ * byte value.
+ * Each entry is 32 bits wide, with two formats distinguished by
+ * the sign bit (bit 31):
+ *
+ * One format for transitional entries (bit 31 not set) for non-final bytes, and
+ * one format for final entries (bit 31 set).
+ * Both formats contain the number of the next state in the same bit
+ * positions.
+ * State 0 is the initial state.
+ *
+ * Most of the time, the offset values of subsequent states are added
+ * up to a scalar value. This value will eventually be the index of
+ * the Unicode code point in a table that follows the state table.
+ * The effect is that the code points for final state table rows
+ * are contiguous. The code points of final state rows follow each other
+ * in the order of the references to those final states by previous
+ * states, etc.
+ *
+ * For some terminal states, the offset is itself the output Unicode
+ * code point (16 bits for a BMP code point or 20 bits for a supplementary
+ * code point (stored as code point minus 0x10000 so that 20 bits are enough).
+ * For others, the code point in the Unicode table is stored with either
+ * one or two code units: one for BMP code points, two for a pair of
+ * surrogates.
+ * All code points for a final state entry take up the same number of code
+ * units, regardless of whether they all actually _use_ the same number
+ * of code units. This is necessary for simple array access.
+ *
+ * An additional feature comes in with what in ICU is called "fallback"
+ * mappings:
+ *
+ * In addition to round-trippable, precise, 1:1 mappings, there are often
+ * mappings defined between similar, though not the same, characters.
+ * Typically, such mappings occur only in fromUnicode mapping tables because
+ * Unicode has a superset repertoire of most other codepages. However, it
+ * is possible to provide such mappings in the toUnicode tables, too.
+ * In this case, the fallback mappings are partly integrated into the
+ * general state tables because the structure of the encoding includes their
+ * byte sequences.
+ * For final entries in an initial state, fallback mappings are stored in
+ * the entry itself like with roundtrip mappings.
+ * For other final entries, they are stored in the code units table if
+ * the entry is for a pair of code units.
+ * For single-unit results in the code units table, there is no space to
+ * alternatively hold a fallback mapping; in this case, the code unit
+ * is stored as U+fffe (unassigned), and the fallback mapping needs to
+ * be looked up by the scalar offset value in a separate table.
+ *
+ * "Unassigned" state entries really mean "structurally unassigned",
+ * i.e., such a byte sequence will never have a mapping result.
+ *
+ * The interpretation of the bits in each entry is as follows:
+ *
+ * Bit 31 not set, not a terminal entry ("transitional"):
+ * 30..24 next state
+ * 23..0  offset delta, to be added up
+ *
+ * Bit 31 set, terminal ("final") entry:
+ * 30..24 next state (regardless of action code)
+ * 23..20 action code:
+ *        action codes 0 and 1 result in precise-mapping Unicode code points
+ *        0  valid byte sequence
+ *           19..16 not used, 0
+ *           15..0  16-bit Unicode BMP code point
+ *                  never U+fffe or U+ffff
+ *        1  valid byte sequence
+ *           19..0  20-bit Unicode supplementary code point
+ *                  never U+fffe or U+ffff
+ *
+ *        action codes 2 and 3 result in fallback (unidirectional-mapping) Unicode code points
+ *        2  valid byte sequence (fallback)
+ *           19..16 not used, 0
+ *           15..0  16-bit Unicode BMP code point as fallback result
+ *        3  valid byte sequence (fallback)
+ *           19..0  20-bit Unicode supplementary code point as fallback result
+ *
+ *        action codes 4 and 5 may result in roundtrip/fallback/unassigned/illegal results
+ *        depending on the code units they result in
+ *        4  valid byte sequence
+ *           19..9  not used, 0
+ *            8..0  final offset delta
+ *                  pointing to one 16-bit code unit which may be
+ *                  fffe  unassigned -- look for a fallback for this offset
+ *                  ffff  illegal
+ *        5  valid byte sequence
+ *           19..9  not used, 0
+ *            8..0  final offset delta
+ *                  pointing to two 16-bit code units
+ *                  (typically UTF-16 surrogates)
+ *                  the result depends on the first code unit as follows:
+ *                  0000..d7ff  roundtrip BMP code point (1st alone)
+ *                  d800..dbff  roundtrip surrogate pair (1st, 2nd)
+ *                  dc00..dfff  fallback surrogate pair (1st-400, 2nd)
+ *                  e000        roundtrip BMP code point (2nd alone)
+ *                  e001        fallback BMP code point (2nd alone)
+ *                  fffe        unassigned
+ *                  ffff        illegal
+ *           (the final offset deltas are at most 255 * 2,
+ *            times 2 because of storing code unit pairs)
+ *
+ *        6  unassigned byte sequence
+ *           19..16 not used, 0
+ *           15..0  16-bit Unicode BMP code point U+fffe (new with version 2)
+ *                  this does not contain a final offset delta because the main
+ *                  purpose of this action code is to save scalar offset values;
+ *                  therefore, fallback values cannot be assigned to byte
+ *                  sequences that result in this action code
+ *        7  illegal byte sequence
+ *           19..16 not used, 0
+ *           15..0  16-bit Unicode BMP code point U+ffff (new with version 2)
+ *        8  state change only
+ *           19..0  not used, 0
+ *           useful for state changes in simple stateful encodings,
+ *           at Shift-In/Shift-Out codes
+ *
+ *
+ *        9..15 reserved for future use
+ *           current implementations will only perform a state change
+ *           and ignore bits 19..0
+ *
+ * An encoding with contiguous ranges of unassigned byte sequences, like
+ * Shift-JIS and especially EUC-TW, can be stored efficiently by having
+ * at least two states for the trail bytes:
+ * One trail byte state that results in code points, and one that only
+ * has "unassigned" and "illegal" terminal states.
+ *
+ * Note: partly by accident, this data structure supports simple stateful
+ * encodings without any additional logic.
+ * Currently, only simple Shift-In/Shift-Out schemes are handled with
+ * appropriate state tables (especially EBCDIC_STATEFUL!).
+ *
+ * MBCS version 2 added:
+ * unassigned and illegal action codes have U+fffe and U+ffff
+ * instead of unused bits; this is useful for _MBCS_SINGLE_SIMPLE_GET_NEXT_BMP()
+ *
+ * Converting from Unicode to codepage bytes --------------------------------***
+ *
+ * The conversion data structure for fromUnicode is designed for the known
+ * structure of Unicode. It maps from 21-bit code points (0..0x10ffff) to
+ * a sequence of 1..4 bytes, in addition to a flag that indicates if there is
+ * a roundtrip mapping.
+ *
+ * The lookup is done with a 3-stage trie, using 11/6/4 bits for stage 1/2/3
+ * like in the character properties table.
+ * The beginning of the trie is at offsetFromUTable, the beginning of stage 3
+ * with the resulting bytes is at offsetFromUBytes.
+ *
+ * Beginning with version 4, single-byte codepages have a significantly different
+ * trie compared to other codepages.
+ * In all cases, the entry in stage 1 is directly the index of the block of
+ * 64 entries in stage 2.
+ *
+ * Single-byte lookup:
+ *
+ * Stage 2 only contains 16-bit indexes directly to the 16-blocks in stage 3.
+ * Stage 3 contains one 16-bit word per result:
+ * Bits 15..8 indicate the kind of result:
+ *    f  roundtrip result
+ *    c  fallback result from private-use code point
+ *    8  fallback result from other code points
+ *    0  unassigned
+ * Bits 7..0 contain the codepage byte. A zero byte is always possible.
+ *
+ * In version 4.3, the runtime code can build an sbcsIndex for a utf8Friendly
+ * file. For 2-byte UTF-8 byte sequences and some 3-byte sequences the lookup
+ * becomes a 2-stage (single-index) trie lookup with 6 bits for stage 3.
+ * ASCII code points can be looked up with a linear array access into stage 3.
+ * See maxFastUChar and other details in ucnvmbcs.h.
+ *
+ * Multi-byte lookup:
+ *
+ * Stage 2 contains a 32-bit word for each 16-block in stage 3:
+ * Bits 31..16 contain flags for which stage 3 entries contain roundtrip results
+ *             test: MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c)
+ *             If this test is false, then a non-zero result will be interpreted as
+ *             a fallback mapping.
+ * Bits 15..0  contain the index to stage 3, which must be multiplied by 16*(bytes per char)
+ *
+ * Stage 3 contains 2, 3, or 4 bytes per result.
+ * 2 or 4 bytes are stored as uint16_t/uint32_t in platform endianness,
+ * while 3 bytes are stored as bytes in big-endian order.
+ * Leading zero bytes are ignored, and the number of bytes is counted.
+ * A zero byte mapping result is possible as a roundtrip result.
+ * For some output types, the actual result is processed from this;
+ * see ucnv_MBCSFromUnicodeWithOffsets().
+ *
+ * Note that stage 1 always contains 0x440=1088 entries (0x440==0x110000>>10),
+ * or (version 3 and up) for BMP-only codepages, it contains 64 entries.
+ *
+ * In version 4.3, a utf8Friendly file contains an mbcsIndex table.
+ * For 2-byte UTF-8 byte sequences and most 3-byte sequences the lookup
+ * becomes a 2-stage (single-index) trie lookup with 6 bits for stage 3.
+ * ASCII code points can be looked up with a linear array access into stage 3.
+ * See maxFastUChar, mbcsIndex and other details in ucnvmbcs.h.
+ *
+ * In version 3, stage 2 blocks may overlap by multiples of the multiplier
+ * for compaction.
+ * In version 4, stage 2 blocks (and for single-byte codepages, stage 3 blocks)
+ * may overlap by any number of entries.
+ *
+ * MBCS version 2 added:
+ * the converter checks for known output types, which allows
+ * adding new ones without crashing an unaware converter
+ */
+
+/**
+ * Callback from ucnv_MBCSEnumToUnicode(), takes 32 mappings from
+ * consecutive sequences of bytes, starting from the one encoded in value,
+ * to Unicode code points. (Multiple mappings to reduce per-function call overhead.)
+ * Does not currently support m:n mappings or reverse fallbacks.
+ * This function will not be called for sequences of bytes with leading zeros.
+ *
+ * @param context an opaque pointer, as passed into ucnv_MBCSEnumToUnicode()
+ * @param value contains 1..4 bytes of the first byte sequence, right-aligned
+ * @param codePoints resulting Unicode code points, or negative if a byte sequence does
+ *        not map to anything
+ * @return TRUE to continue enumeration, FALSE to stop
+ */
+typedef UBool U_CALLCONV
+UConverterEnumToUCallback(const void *context, uint32_t value, UChar32 codePoints[32]);
+
+static void U_CALLCONV
+ucnv_MBCSLoad(UConverterSharedData *sharedData,
+          UConverterLoadArgs *pArgs,
+          const uint8_t *raw,
+          UErrorCode *pErrorCode);
+
+static void U_CALLCONV
+ucnv_MBCSUnload(UConverterSharedData *sharedData);
+
+static void U_CALLCONV
+ucnv_MBCSOpen(UConverter *cnv,
+              UConverterLoadArgs *pArgs,
+              UErrorCode *pErrorCode);
+
+static UChar32 U_CALLCONV
+ucnv_MBCSGetNextUChar(UConverterToUnicodeArgs *pArgs,
+                  UErrorCode *pErrorCode);
+
+static void U_CALLCONV
+ucnv_MBCSGetStarters(const UConverter* cnv,
+                 UBool starters[256],
+                 UErrorCode *pErrorCode);
+
+U_CDECL_BEGIN
+static const char* U_CALLCONV
+ucnv_MBCSGetName(const UConverter *cnv);
+U_CDECL_END
+
+static void U_CALLCONV
+ucnv_MBCSWriteSub(UConverterFromUnicodeArgs *pArgs,
+              int32_t offsetIndex,
+              UErrorCode *pErrorCode);
+
+static UChar32 U_CALLCONV
+ucnv_MBCSGetNextUChar(UConverterToUnicodeArgs *pArgs,
+                  UErrorCode *pErrorCode);
+
+static void U_CALLCONV
+ucnv_SBCSFromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                  UConverterToUnicodeArgs *pToUArgs,
+                  UErrorCode *pErrorCode);
+
+static void U_CALLCONV
+ucnv_MBCSGetUnicodeSet(const UConverter *cnv,
+                   const USetAdder *sa,
+                   UConverterUnicodeSet which,
+                   UErrorCode *pErrorCode);
+
+static void U_CALLCONV
+ucnv_DBCSFromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                  UConverterToUnicodeArgs *pToUArgs,
+                  UErrorCode *pErrorCode);
+
+static const UConverterImpl _SBCSUTF8Impl={
+    UCNV_MBCS,
+
+    ucnv_MBCSLoad,
+    ucnv_MBCSUnload,
+
+    ucnv_MBCSOpen,
+    NULL,
+    NULL,
+
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSGetNextUChar,
+
+    ucnv_MBCSGetStarters,
+    ucnv_MBCSGetName,
+    ucnv_MBCSWriteSub,
+    NULL,
+    ucnv_MBCSGetUnicodeSet,
+
+    NULL,
+    ucnv_SBCSFromUTF8
+};
+
+static const UConverterImpl _DBCSUTF8Impl={
+    UCNV_MBCS,
+
+    ucnv_MBCSLoad,
+    ucnv_MBCSUnload,
+
+    ucnv_MBCSOpen,
+    NULL,
+    NULL,
+
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSGetNextUChar,
+
+    ucnv_MBCSGetStarters,
+    ucnv_MBCSGetName,
+    ucnv_MBCSWriteSub,
+    NULL,
+    ucnv_MBCSGetUnicodeSet,
+
+    NULL,
+    ucnv_DBCSFromUTF8
+};
+
+static const UConverterImpl _MBCSImpl={
+    UCNV_MBCS,
+
+    ucnv_MBCSLoad,
+    ucnv_MBCSUnload,
+
+    ucnv_MBCSOpen,
+    NULL,
+    NULL,
+
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSToUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSFromUnicodeWithOffsets,
+    ucnv_MBCSGetNextUChar,
+
+    ucnv_MBCSGetStarters,
+    ucnv_MBCSGetName,
+    ucnv_MBCSWriteSub,
+    NULL,
+    ucnv_MBCSGetUnicodeSet,
+    NULL,
+    NULL
+};
+
+/* Static data is in tools/makeconv/ucnvstat.c for data-based
+ * converters. Be sure to update it as well.
+ */
+
+const UConverterSharedData _MBCSData={
+    sizeof(UConverterSharedData), 1,
+    NULL, NULL, FALSE, TRUE, &_MBCSImpl,
+    0, UCNV_MBCS_TABLE_INITIALIZER
+};
+
+
+/* GB 18030 data ------------------------------------------------------------ */
+
+/* helper macros for linear values for GB 18030 four-byte sequences */
+#define LINEAR_18030(a, b, c, d) ((((a)*10+(b))*126L+(c))*10L+(d))
+
+#define LINEAR_18030_BASE LINEAR_18030(0x81, 0x30, 0x81, 0x30)
+
+#define LINEAR(x) LINEAR_18030(x>>24, (x>>16)&0xff, (x>>8)&0xff, x&0xff)
+
+/*
+ * Some ranges of GB 18030 where both the Unicode code points and the
+ * GB four-byte sequences are contiguous and are handled algorithmically by
+ * the special callback functions below.
+ * The values are start & end of Unicode & GB codes.
+ *
+ * Note that single surrogates are not mapped by GB 18030
+ * as of the re-released mapping tables from 2000-nov-30.
+ */
+static const uint32_t
+gb18030Ranges[14][4]={
+    {0x10000, 0x10FFFF, LINEAR(0x90308130), LINEAR(0xE3329A35)},
+    {0x9FA6, 0xD7FF, LINEAR(0x82358F33), LINEAR(0x8336C738)},
+    {0x0452, 0x1E3E, LINEAR(0x8130D330), LINEAR(0x8135F436)},
+    {0x1E40, 0x200F, LINEAR(0x8135F438), LINEAR(0x8136A531)},
+    {0xE865, 0xF92B, LINEAR(0x8336D030), LINEAR(0x84308534)},
+    {0x2643, 0x2E80, LINEAR(0x8137A839), LINEAR(0x8138FD38)},
+    {0xFA2A, 0xFE2F, LINEAR(0x84309C38), LINEAR(0x84318537)},
+    {0x3CE1, 0x4055, LINEAR(0x8231D438), LINEAR(0x8232AF32)},
+    {0x361B, 0x3917, LINEAR(0x8230A633), LINEAR(0x8230F237)},
+    {0x49B8, 0x4C76, LINEAR(0x8234A131), LINEAR(0x8234E733)},
+    {0x4160, 0x4336, LINEAR(0x8232C937), LINEAR(0x8232F837)},
+    {0x478E, 0x4946, LINEAR(0x8233E838), LINEAR(0x82349638)},
+    {0x44D7, 0x464B, LINEAR(0x8233A339), LINEAR(0x8233C931)},
+    {0xFFE6, 0xFFFF, LINEAR(0x8431A234), LINEAR(0x8431A439)}
+};
+
+/* bit flag for UConverter.options indicating GB 18030 special handling */
+#define _MBCS_OPTION_GB18030 0x8000
+
+/* bit flag for UConverter.options indicating KEIS,JEF,JIF special handling */
+#define _MBCS_OPTION_KEIS 0x01000
+#define _MBCS_OPTION_JEF  0x02000
+#define _MBCS_OPTION_JIPS 0x04000
+
+#define KEIS_SO_CHAR_1 0x0A
+#define KEIS_SO_CHAR_2 0x42
+#define KEIS_SI_CHAR_1 0x0A
+#define KEIS_SI_CHAR_2 0x41
+
+#define JEF_SO_CHAR 0x28
+#define JEF_SI_CHAR 0x29
+
+#define JIPS_SO_CHAR_1 0x1A
+#define JIPS_SO_CHAR_2 0x70
+#define JIPS_SI_CHAR_1 0x1A
+#define JIPS_SI_CHAR_2 0x71
+
+enum SISO_Option {
+    SI,
+    SO
+};
+typedef enum SISO_Option SISO_Option;
+
+static int32_t getSISOBytes(SISO_Option option, uint32_t cnvOption, uint8_t *value) {
+    int32_t SISOLength = 0;
+
+    switch (option) {
+        case SI:
+            if ((cnvOption&_MBCS_OPTION_KEIS)!=0) {
+                value[0] = KEIS_SI_CHAR_1;
+                value[1] = KEIS_SI_CHAR_2;
+                SISOLength = 2;
+            } else if ((cnvOption&_MBCS_OPTION_JEF)!=0) {
+                value[0] = JEF_SI_CHAR;
+                SISOLength = 1;
+            } else if ((cnvOption&_MBCS_OPTION_JIPS)!=0) {
+                value[0] = JIPS_SI_CHAR_1;
+                value[1] = JIPS_SI_CHAR_2;
+                SISOLength = 2;
+            } else {
+                value[0] = UCNV_SI;
+                SISOLength = 1;
+            }
+            break;
+        case SO:
+            if ((cnvOption&_MBCS_OPTION_KEIS)!=0) {
+                value[0] = KEIS_SO_CHAR_1;
+                value[1] = KEIS_SO_CHAR_2;
+                SISOLength = 2;
+            } else if ((cnvOption&_MBCS_OPTION_JEF)!=0) {
+                value[0] = JEF_SO_CHAR;
+                SISOLength = 1;
+            } else if ((cnvOption&_MBCS_OPTION_JIPS)!=0) {
+                value[0] = JIPS_SO_CHAR_1;
+                value[1] = JIPS_SO_CHAR_2;
+                SISOLength = 2;
+            } else {
+                value[0] = UCNV_SO;
+                SISOLength = 1;
+            }
+            break;
+        default:
+            /* Should never happen. */
+            break;
+    }
+
+    return SISOLength;
+}
+
+/* Miscellaneous ------------------------------------------------------------ */
+
+/* similar to ucnv_MBCSGetNextUChar() but recursive */
+static UBool
+enumToU(UConverterMBCSTable *mbcsTable, int8_t stateProps[],
+        int32_t state, uint32_t offset,
+        uint32_t value,
+        UConverterEnumToUCallback *callback, const void *context,
+        UErrorCode *pErrorCode) {
+    UChar32 codePoints[32];
+    const int32_t *row;
+    const uint16_t *unicodeCodeUnits;
+    UChar32 anyCodePoints;
+    int32_t b, limit;
+
+    row=mbcsTable->stateTable[state];
+    unicodeCodeUnits=mbcsTable->unicodeCodeUnits;
+
+    value<<=8;
+    anyCodePoints=-1;  /* becomes non-negative if there is a mapping */
+
+    b=(stateProps[state]&0x38)<<2;
+    if(b==0 && stateProps[state]>=0x40) {
+        /* skip byte sequences with leading zeros because they are not stored in the fromUnicode table */
+        codePoints[0]=U_SENTINEL;
+        b=1;
+    }
+    limit=((stateProps[state]&7)+1)<<5;
+    while(b<limit) {
+        int32_t entry=row[b];
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            int32_t nextState=MBCS_ENTRY_TRANSITION_STATE(entry);
+            if(stateProps[nextState]>=0) {
+                /* recurse to a state with non-ignorable actions */
+                if(!enumToU(
+                        mbcsTable, stateProps, nextState,
+                        offset+MBCS_ENTRY_TRANSITION_OFFSET(entry),
+                        value|(uint32_t)b,
+                        callback, context,
+                        pErrorCode)) {
+                    return FALSE;
+                }
+            }
+            codePoints[b&0x1f]=U_SENTINEL;
+        } else {
+            UChar32 c;
+            int32_t action;
+
+            /*
+             * An if-else-if chain provides more reliable performance for
+             * the most common cases compared to a switch.
+             */
+            action=MBCS_ENTRY_FINAL_ACTION(entry);
+            if(action==MBCS_STATE_VALID_DIRECT_16) {
+                /* output BMP code point */
+                c=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            } else if(action==MBCS_STATE_VALID_16) {
+                int32_t finalOffset=offset+MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[finalOffset];
+                if(c<0xfffe) {
+                    /* output BMP code point */
+                } else {
+                    c=U_SENTINEL;
+                }
+            } else if(action==MBCS_STATE_VALID_16_PAIR) {
+                int32_t finalOffset=offset+MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[finalOffset++];
+                if(c<0xd800) {
+                    /* output BMP code point below 0xd800 */
+                } else if(c<=0xdbff) {
+                    /* output roundtrip or fallback supplementary code point */
+                    c=((c&0x3ff)<<10)+unicodeCodeUnits[finalOffset]+(0x10000-0xdc00);
+                } else if(c==0xe000) {
+                    /* output roundtrip BMP code point above 0xd800 or fallback BMP code point */
+                    c=unicodeCodeUnits[finalOffset];
+                } else {
+                    c=U_SENTINEL;
+                }
+            } else if(action==MBCS_STATE_VALID_DIRECT_20) {
+                /* output supplementary code point */
+                c=(UChar32)(MBCS_ENTRY_FINAL_VALUE(entry)+0x10000);
+            } else {
+                c=U_SENTINEL;
+            }
+
+            codePoints[b&0x1f]=c;
+            anyCodePoints&=c;
+        }
+        if(((++b)&0x1f)==0) {
+            if(anyCodePoints>=0) {
+                if(!callback(context, value|(uint32_t)(b-0x20), codePoints)) {
+                    return FALSE;
+                }
+                anyCodePoints=-1;
+            }
+        }
+    }
+    return TRUE;
+}
+
+/*
+ * Only called if stateProps[state]==-1.
+ * A recursive call may do stateProps[state]|=0x40 if this state is the target of an
+ * MBCS_STATE_CHANGE_ONLY.
+ */
+static int8_t
+getStateProp(const int32_t (*stateTable)[256], int8_t stateProps[], int state) {
+    const int32_t *row;
+    int32_t min, max, entry, nextState;
+
+    row=stateTable[state];
+    stateProps[state]=0;
+
+    /* find first non-ignorable state */
+    for(min=0;; ++min) {
+        entry=row[min];
+        nextState=MBCS_ENTRY_STATE(entry);
+        if(stateProps[nextState]==-1) {
+            getStateProp(stateTable, stateProps, nextState);
+        }
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            if(stateProps[nextState]>=0) {
+                break;
+            }
+        } else if(MBCS_ENTRY_FINAL_ACTION(entry)<MBCS_STATE_UNASSIGNED) {
+            break;
+        }
+        if(min==0xff) {
+            stateProps[state]=-0x40;  /* (int8_t)0xc0 */
+            return stateProps[state];
+        }
+    }
+    stateProps[state]|=(int8_t)((min>>5)<<3);
+
+    /* find last non-ignorable state */
+    for(max=0xff; min<max; --max) {
+        entry=row[max];
+        nextState=MBCS_ENTRY_STATE(entry);
+        if(stateProps[nextState]==-1) {
+            getStateProp(stateTable, stateProps, nextState);
+        }
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            if(stateProps[nextState]>=0) {
+                break;
+            }
+        } else if(MBCS_ENTRY_FINAL_ACTION(entry)<MBCS_STATE_UNASSIGNED) {
+            break;
+        }
+    }
+    stateProps[state]|=(int8_t)(max>>5);
+
+    /* recurse further and collect direct-state information */
+    while(min<=max) {
+        entry=row[min];
+        nextState=MBCS_ENTRY_STATE(entry);
+        if(stateProps[nextState]==-1) {
+            getStateProp(stateTable, stateProps, nextState);
+        }
+        if(MBCS_ENTRY_IS_FINAL(entry)) {
+            stateProps[nextState]|=0x40;
+            if(MBCS_ENTRY_FINAL_ACTION(entry)<=MBCS_STATE_FALLBACK_DIRECT_20) {
+                stateProps[state]|=0x40;
+            }
+        }
+        ++min;
+    }
+    return stateProps[state];
+}
+
+/*
+ * Internal function enumerating the toUnicode data of an MBCS converter.
+ * Currently only used for reconstituting data for a MBCS_OPT_NO_FROM_U
+ * table, but could also be used for a future ucnv_getUnicodeSet() option
+ * that includes reverse fallbacks (after updating this function's implementation).
+ * Currently only handles roundtrip mappings.
+ * Does not currently handle extensions.
+ */
+static void
+ucnv_MBCSEnumToUnicode(UConverterMBCSTable *mbcsTable,
+                       UConverterEnumToUCallback *callback, const void *context,
+                       UErrorCode *pErrorCode) {
+    /*
+     * Properties for each state, to speed up the enumeration.
+     * Ignorable actions are unassigned/illegal/state-change-only:
+     * They do not lead to mappings.
+     *
+     * Bits 7..6:
+     * 1 direct/initial state (stateful converters have multiple)
+     * 0 non-initial state with transitions or with non-ignorable result actions
+     * -1 final state with only ignorable actions
+     *
+     * Bits 5..3:
+     * The lowest byte value with non-ignorable actions is
+     * value<<5 (rounded down).
+     *
+     * Bits 2..0:
+     * The highest byte value with non-ignorable actions is
+     * (value<<5)&0x1f (rounded up).
+     */
+    int8_t stateProps[MBCS_MAX_STATE_COUNT];
+    int32_t state;
+
+    uprv_memset(stateProps, -1, sizeof(stateProps));
+
+    /* recurse from state 0 and set all stateProps */
+    getStateProp(mbcsTable->stateTable, stateProps, 0);
+
+    for(state=0; state<mbcsTable->countStates; ++state) {
+        /*if(stateProps[state]==-1) {
+            printf("unused/unreachable <icu:state> %d\n", state);
+        }*/
+        if(stateProps[state]>=0x40) {
+            /* start from each direct state */
+            enumToU(
+                mbcsTable, stateProps, state, 0, 0,
+                callback, context,
+                pErrorCode);
+        }
+    }
+}
+
+U_CFUNC void 
+ucnv_MBCSGetFilteredUnicodeSetForUnicode(const UConverterSharedData *sharedData,
+                                         const USetAdder *sa,
+                                         UConverterUnicodeSet which,
+                                         UConverterSetFilter filter,
+                                         UErrorCode *pErrorCode) {
+    const UConverterMBCSTable *mbcsTable;
+    const uint16_t *table;
+
+    uint32_t st3;
+    uint16_t st1, maxStage1, st2;
+
+    UChar32 c;
+
+    /* enumerate the from-Unicode trie table */
+    mbcsTable=&sharedData->mbcs;
+    table=mbcsTable->fromUnicodeTable;
+    if(mbcsTable->unicodeMask&UCNV_HAS_SUPPLEMENTARY) {
+        maxStage1=0x440;
+    } else {
+        maxStage1=0x40;
+    }
+
+    c=0; /* keep track of the current code point while enumerating */
+
+    if(mbcsTable->outputType==MBCS_OUTPUT_1) {
+        const uint16_t *stage2, *stage3, *results;
+        uint16_t minValue;
+
+        results=(const uint16_t *)mbcsTable->fromUnicodeBytes;
+
+        /*
+         * Set a threshold variable for selecting which mappings to use.
+         * See ucnv_MBCSSingleFromBMPWithOffsets() and
+         * MBCS_SINGLE_RESULT_FROM_U() for details.
+         */
+        if(which==UCNV_ROUNDTRIP_SET) {
+            /* use only roundtrips */
+            minValue=0xf00;
+        } else /* UCNV_ROUNDTRIP_AND_FALLBACK_SET */ {
+            /* use all roundtrip and fallback results */
+            minValue=0x800;
+        }
+
+        for(st1=0; st1<maxStage1; ++st1) {
+            st2=table[st1];
+            if(st2>maxStage1) {
+                stage2=table+st2;
+                for(st2=0; st2<64; ++st2) {
+                    if((st3=stage2[st2])!=0) {
+                        /* read the stage 3 block */
+                        stage3=results+st3;
+
+                        do {
+                            if(*stage3++>=minValue) {
+                                sa->add(sa->set, c);
+                            }
+                        } while((++c&0xf)!=0);
+                    } else {
+                        c+=16; /* empty stage 3 block */
+                    }
+                }
+            } else {
+                c+=1024; /* empty stage 2 block */
+            }
+        }
+    } else {
+        const uint32_t *stage2;
+        const uint8_t *stage3, *bytes;
+        uint32_t st3Multiplier;
+        uint32_t value;
+        UBool useFallback;
+
+        bytes=mbcsTable->fromUnicodeBytes;
+
+        useFallback=(UBool)(which==UCNV_ROUNDTRIP_AND_FALLBACK_SET);
+
+        switch(mbcsTable->outputType) {
+        case MBCS_OUTPUT_3:
+        case MBCS_OUTPUT_4_EUC:
+            st3Multiplier=3;
+            break;
+        case MBCS_OUTPUT_4:
+            st3Multiplier=4;
+            break;
+        default:
+            st3Multiplier=2;
+            break;
+        }
+
+        for(st1=0; st1<maxStage1; ++st1) {
+            st2=table[st1];
+            if(st2>(maxStage1>>1)) {
+                stage2=(const uint32_t *)table+st2;
+                for(st2=0; st2<64; ++st2) {
+                    if((st3=stage2[st2])!=0) {
+                        /* read the stage 3 block */
+                        stage3=bytes+st3Multiplier*16*(uint32_t)(uint16_t)st3;
+
+                        /* get the roundtrip flags for the stage 3 block */
+                        st3>>=16;
+
+                        /*
+                         * Add code points for which the roundtrip flag is set,
+                         * or which map to non-zero bytes if we use fallbacks.
+                         * See ucnv_MBCSFromUnicodeWithOffsets() for details.
+                         */
+                        switch(filter) {
+                        case UCNV_SET_FILTER_NONE:
+                            do {
+                                if(st3&1) {
+                                    sa->add(sa->set, c);
+                                    stage3+=st3Multiplier;
+                                } else if(useFallback) {
+                                    uint8_t b=0;
+                                    switch(st3Multiplier) {
+                                    case 4:
+                                        b|=*stage3++;
+                                        U_FALLTHROUGH;
+                                    case 3:
+                                        b|=*stage3++;
+                                        U_FALLTHROUGH;
+                                    case 2:
+                                        b|=stage3[0]|stage3[1];
+                                        stage3+=2;
+                                        U_FALLTHROUGH;
+                                    default:
+                                        break;
+                                    }
+                                    if(b!=0) {
+                                        sa->add(sa->set, c);
+                                    }
+                                }
+                                st3>>=1;
+                            } while((++c&0xf)!=0);
+                            break;
+                        case UCNV_SET_FILTER_DBCS_ONLY:
+                             /* Ignore single-byte results (<0x100). */
+                            do {
+                                if(((st3&1)!=0 || useFallback) && *((const uint16_t *)stage3)>=0x100) {
+                                    sa->add(sa->set, c);
+                                }
+                                st3>>=1;
+                                stage3+=2;  /* +=st3Multiplier */
+                            } while((++c&0xf)!=0);
+                            break;
+                        case UCNV_SET_FILTER_2022_CN:
+                             /* Only add code points that map to CNS 11643 planes 1 & 2 for non-EXT ISO-2022-CN. */
+                            do {
+                                if(((st3&1)!=0 || useFallback) && ((value=*stage3)==0x81 || value==0x82)) {
+                                    sa->add(sa->set, c);
+                                }
+                                st3>>=1;
+                                stage3+=3;  /* +=st3Multiplier */
+                            } while((++c&0xf)!=0);
+                            break;
+                        case UCNV_SET_FILTER_SJIS:
+                             /* Only add code points that map to Shift-JIS codes corresponding to JIS X 0208. */
+                            do {
+                                if(((st3&1)!=0 || useFallback) && (value=*((const uint16_t *)stage3))>=0x8140 && value<=0xeffc) {
+                                    sa->add(sa->set, c);
+                                }
+                                st3>>=1;
+                                stage3+=2;  /* +=st3Multiplier */
+                            } while((++c&0xf)!=0);
+                            break;
+                        case UCNV_SET_FILTER_GR94DBCS:
+                            /* Only add code points that map to ISO 2022 GR 94 DBCS codes (each byte A1..FE). */
+                            do {
+                                if( ((st3&1)!=0 || useFallback) &&
+                                    (uint16_t)((value=*((const uint16_t *)stage3)) - 0xa1a1)<=(0xfefe - 0xa1a1) &&
+                                    (uint8_t)(value-0xa1)<=(0xfe - 0xa1)
+                                ) {
+                                    sa->add(sa->set, c);
+                                }
+                                st3>>=1;
+                                stage3+=2;  /* +=st3Multiplier */
+                            } while((++c&0xf)!=0);
+                            break;
+                        case UCNV_SET_FILTER_HZ:
+                            /* Only add code points that are suitable for HZ DBCS (lead byte A1..FD). */
+                            do {
+                                if( ((st3&1)!=0 || useFallback) &&
+                                    (uint16_t)((value=*((const uint16_t *)stage3))-0xa1a1)<=(0xfdfe - 0xa1a1) &&
+                                    (uint8_t)(value-0xa1)<=(0xfe - 0xa1)
+                                ) {
+                                    sa->add(sa->set, c);
+                                }
+                                st3>>=1;
+                                stage3+=2;  /* +=st3Multiplier */
+                            } while((++c&0xf)!=0);
+                            break;
+                        default:
+                            *pErrorCode=U_INTERNAL_PROGRAM_ERROR;
+                            return;
+                        }
+                    } else {
+                        c+=16; /* empty stage 3 block */
+                    }
+                }
+            } else {
+                c+=1024; /* empty stage 2 block */
+            }
+        }
+    }
+
+    ucnv_extGetUnicodeSet(sharedData, sa, which, filter, pErrorCode);
+}
+
+U_CFUNC void
+ucnv_MBCSGetUnicodeSetForUnicode(const UConverterSharedData *sharedData,
+                                 const USetAdder *sa,
+                                 UConverterUnicodeSet which,
+                                 UErrorCode *pErrorCode) {
+    ucnv_MBCSGetFilteredUnicodeSetForUnicode(
+        sharedData, sa, which,
+        sharedData->mbcs.outputType==MBCS_OUTPUT_DBCS_ONLY ?
+            UCNV_SET_FILTER_DBCS_ONLY :
+            UCNV_SET_FILTER_NONE,
+        pErrorCode);
+}
+
+static void U_CALLCONV
+ucnv_MBCSGetUnicodeSet(const UConverter *cnv,
+                   const USetAdder *sa,
+                   UConverterUnicodeSet which,
+                   UErrorCode *pErrorCode) {
+    if(cnv->options&_MBCS_OPTION_GB18030) {
+        sa->addRange(sa->set, 0, 0xd7ff);
+        sa->addRange(sa->set, 0xe000, 0x10ffff);
+    } else {
+        ucnv_MBCSGetUnicodeSetForUnicode(cnv->sharedData, sa, which, pErrorCode);
+    }
+}
+
+/* conversion extensions for input not in the main table -------------------- */
+
+/*
+ * Hardcoded extension handling for GB 18030.
+ * Definition of LINEAR macros and gb18030Ranges see near the beginning of the file.
+ *
+ * In the future, conversion extensions may handle m:n mappings and delta tables,
+ * see http://source.icu-project.org/repos/icu/icuhtml/trunk/design/conversion/conversion_extensions.html
+ *
+ * If an input character cannot be mapped, then these functions set an error
+ * code. The framework will then call the callback function.
+ */
+
+/*
+ * @return if(U_FAILURE) return the code point for cnv->fromUChar32
+ *         else return 0 after output has been written to the target
+ */
+static UChar32
+_extFromU(UConverter *cnv, const UConverterSharedData *sharedData,
+          UChar32 cp,
+          const UChar **source, const UChar *sourceLimit,
+          uint8_t **target, const uint8_t *targetLimit,
+          int32_t **offsets, int32_t sourceIndex,
+          UBool flush,
+          UErrorCode *pErrorCode) {
+    const int32_t *cx;
+
+    cnv->useSubChar1=FALSE;
+
+    if( (cx=sharedData->mbcs.extIndexes)!=NULL &&
+        ucnv_extInitialMatchFromU(
+            cnv, cx,
+            cp, source, sourceLimit,
+            (char **)target, (char *)targetLimit,
+            offsets, sourceIndex,
+            flush,
+            pErrorCode)
+    ) {
+        return 0; /* an extension mapping handled the input */
+    }
+
+    /* GB 18030 */
+    if((cnv->options&_MBCS_OPTION_GB18030)!=0) {
+        const uint32_t *range;
+        int32_t i;
+
+        range=gb18030Ranges[0];
+        for(i=0; i<UPRV_LENGTHOF(gb18030Ranges); range+=4, ++i) {
+            if(range[0]<=(uint32_t)cp && (uint32_t)cp<=range[1]) {
+                /* found the Unicode code point, output the four-byte sequence for it */
+                uint32_t linear;
+                char bytes[4];
+
+                /* get the linear value of the first GB 18030 code in this range */
+                linear=range[2]-LINEAR_18030_BASE;
+
+                /* add the offset from the beginning of the range */
+                linear+=((uint32_t)cp-range[0]);
+
+                /* turn this into a four-byte sequence */
+                bytes[3]=(char)(0x30+linear%10); linear/=10;
+                bytes[2]=(char)(0x81+linear%126); linear/=126;
+                bytes[1]=(char)(0x30+linear%10); linear/=10;
+                bytes[0]=(char)(0x81+linear);
+
+                /* output this sequence */
+                ucnv_fromUWriteBytes(cnv,
+                                     bytes, 4, (char **)target, (char *)targetLimit,
+                                     offsets, sourceIndex, pErrorCode);
+                return 0;
+            }
+        }
+    }
+
+    /* no mapping */
+    *pErrorCode=U_INVALID_CHAR_FOUND;
+    return cp;
+}
+
+/*
+ * Input sequence: cnv->toUBytes[0..length[
+ * @return if(U_FAILURE) return the length (toULength, byteIndex) for the input
+ *         else return 0 after output has been written to the target
+ */
+static int8_t
+_extToU(UConverter *cnv, const UConverterSharedData *sharedData,
+        int8_t length,
+        const uint8_t **source, const uint8_t *sourceLimit,
+        UChar **target, const UChar *targetLimit,
+        int32_t **offsets, int32_t sourceIndex,
+        UBool flush,
+        UErrorCode *pErrorCode) {
+    const int32_t *cx;
+
+    if( (cx=sharedData->mbcs.extIndexes)!=NULL &&
+        ucnv_extInitialMatchToU(
+            cnv, cx,
+            length, (const char **)source, (const char *)sourceLimit,
+            target, targetLimit,
+            offsets, sourceIndex,
+            flush,
+            pErrorCode)
+    ) {
+        return 0; /* an extension mapping handled the input */
+    }
+
+    /* GB 18030 */
+    if(length==4 && (cnv->options&_MBCS_OPTION_GB18030)!=0) {
+        const uint32_t *range;
+        uint32_t linear;
+        int32_t i;
+
+        linear=LINEAR_18030(cnv->toUBytes[0], cnv->toUBytes[1], cnv->toUBytes[2], cnv->toUBytes[3]);
+        range=gb18030Ranges[0];
+        for(i=0; i<UPRV_LENGTHOF(gb18030Ranges); range+=4, ++i) {
+            if(range[2]<=linear && linear<=range[3]) {
+                /* found the sequence, output the Unicode code point for it */
+                *pErrorCode=U_ZERO_ERROR;
+
+                /* add the linear difference between the input and start sequences to the start code point */
+                linear=range[0]+(linear-range[2]);
+
+                /* output this code point */
+                ucnv_toUWriteCodePoint(cnv, linear, target, targetLimit, offsets, sourceIndex, pErrorCode);
+
+                return 0;
+            }
+        }
+    }
+
+    /* no mapping */
+    *pErrorCode=U_INVALID_CHAR_FOUND;
+    return length;
+}
+
+/* EBCDIC swap LF<->NL ------------------------------------------------------ */
+
+/*
+ * This code modifies a standard EBCDIC<->Unicode mapping table for
+ * OS/390 (z/OS) Unix System Services (Open Edition).
+ * The difference is in the mapping of Line Feed and New Line control codes:
+ * Standard EBCDIC maps
+ *
+ *   <U000A> \x25 |0
+ *   <U0085> \x15 |0
+ *
+ * but OS/390 USS EBCDIC swaps the control codes for LF and NL,
+ * mapping
+ *
+ *   <U000A> \x15 |0
+ *   <U0085> \x25 |0
+ *
+ * This code modifies a loaded standard EBCDIC<->Unicode mapping table
+ * by copying it into allocated memory and swapping the LF and NL values.
+ * It allows to support the same EBCDIC charset in both versions without
+ * duplicating the entire installed table.
+ */
+
+/* standard EBCDIC codes */
+#define EBCDIC_LF 0x25
+#define EBCDIC_NL 0x15
+
+/* standard EBCDIC codes with roundtrip flag as stored in Unicode-to-single-byte tables */
+#define EBCDIC_RT_LF 0xf25
+#define EBCDIC_RT_NL 0xf15
+
+/* Unicode code points */
+#define U_LF 0x0a
+#define U_NL 0x85
+
+static UBool
+_EBCDICSwapLFNL(UConverterSharedData *sharedData, UErrorCode *pErrorCode) {
+    UConverterMBCSTable *mbcsTable;
+
+    const uint16_t *table, *results;
+    const uint8_t *bytes;
+
+    int32_t (*newStateTable)[256];
+    uint16_t *newResults;
+    uint8_t *p;
+    char *name;
+
+    uint32_t stage2Entry;
+    uint32_t size, sizeofFromUBytes;
+
+    mbcsTable=&sharedData->mbcs;
+
+    table=mbcsTable->fromUnicodeTable;
+    bytes=mbcsTable->fromUnicodeBytes;
+    results=(const uint16_t *)bytes;
+
+    /*
+     * Check that this is an EBCDIC table with SBCS portion -
+     * SBCS or EBCDIC_STATEFUL with standard EBCDIC LF and NL mappings.
+     *
+     * If not, ignore the option. Options are always ignored if they do not apply.
+     */
+    if(!(
+         (mbcsTable->outputType==MBCS_OUTPUT_1 || mbcsTable->outputType==MBCS_OUTPUT_2_SISO) &&
+         mbcsTable->stateTable[0][EBCDIC_LF]==MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_DIRECT_16, U_LF) &&
+         mbcsTable->stateTable[0][EBCDIC_NL]==MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_DIRECT_16, U_NL)
+    )) {
+        return FALSE;
+    }
+
+    if(mbcsTable->outputType==MBCS_OUTPUT_1) {
+        if(!(
+             EBCDIC_RT_LF==MBCS_SINGLE_RESULT_FROM_U(table, results, U_LF) &&
+             EBCDIC_RT_NL==MBCS_SINGLE_RESULT_FROM_U(table, results, U_NL)
+        )) {
+            return FALSE;
+        }
+    } else /* MBCS_OUTPUT_2_SISO */ {
+        stage2Entry=MBCS_STAGE_2_FROM_U(table, U_LF);
+        if(!(
+             MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, U_LF)!=0 &&
+             EBCDIC_LF==MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, U_LF)
+        )) {
+            return FALSE;
+        }
+
+        stage2Entry=MBCS_STAGE_2_FROM_U(table, U_NL);
+        if(!(
+             MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, U_NL)!=0 &&
+             EBCDIC_NL==MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, U_NL)
+        )) {
+            return FALSE;
+        }
+    }
+
+    if(mbcsTable->fromUBytesLength>0) {
+        /*
+         * We _know_ the number of bytes in the fromUnicodeBytes array
+         * starting with header.version 4.1.
+         */
+        sizeofFromUBytes=mbcsTable->fromUBytesLength;
+    } else {
+        /*
+         * Otherwise:
+         * There used to be code to enumerate the fromUnicode
+         * trie and find the highest entry, but it was removed in ICU 3.2
+         * because it was not tested and caused a low code coverage number.
+         * See Jitterbug 3674.
+         * This affects only some .cnv file formats with a header.version
+         * below 4.1, and only when swaplfnl is requested.
+         *
+         * ucnvmbcs.c revision 1.99 is the last one with the
+         * ucnv_MBCSSizeofFromUBytes() function.
+         */
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return FALSE;
+    }
+
+    /*
+     * The table has an appropriate format.
+     * Allocate and build
+     * - a modified to-Unicode state table
+     * - a modified from-Unicode output array
+     * - a converter name string with the swap option appended
+     */
+    size=
+        mbcsTable->countStates*1024+
+        sizeofFromUBytes+
+        UCNV_MAX_CONVERTER_NAME_LENGTH+20;
+    p=(uint8_t *)uprv_malloc(size);
+    if(p==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+
+    /* copy and modify the to-Unicode state table */
+    newStateTable=(int32_t (*)[256])p;
+    uprv_memcpy(newStateTable, mbcsTable->stateTable, mbcsTable->countStates*1024);
+
+    newStateTable[0][EBCDIC_LF]=MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_DIRECT_16, U_NL);
+    newStateTable[0][EBCDIC_NL]=MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_DIRECT_16, U_LF);
+
+    /* copy and modify the from-Unicode result table */
+    newResults=(uint16_t *)newStateTable[mbcsTable->countStates];
+    uprv_memcpy(newResults, bytes, sizeofFromUBytes);
+
+    /* conveniently, the table access macros work on the left side of expressions */
+    if(mbcsTable->outputType==MBCS_OUTPUT_1) {
+        MBCS_SINGLE_RESULT_FROM_U(table, newResults, U_LF)=EBCDIC_RT_NL;
+        MBCS_SINGLE_RESULT_FROM_U(table, newResults, U_NL)=EBCDIC_RT_LF;
+    } else /* MBCS_OUTPUT_2_SISO */ {
+        stage2Entry=MBCS_STAGE_2_FROM_U(table, U_LF);
+        MBCS_VALUE_2_FROM_STAGE_2(newResults, stage2Entry, U_LF)=EBCDIC_NL;
+
+        stage2Entry=MBCS_STAGE_2_FROM_U(table, U_NL);
+        MBCS_VALUE_2_FROM_STAGE_2(newResults, stage2Entry, U_NL)=EBCDIC_LF;
+    }
+
+    /* set the canonical converter name */
+    name=(char *)newResults+sizeofFromUBytes;
+    uprv_strcpy(name, sharedData->staticData->name);
+    uprv_strcat(name, UCNV_SWAP_LFNL_OPTION_STRING);
+
+    /* set the pointers */
+    icu::umtx_lock(NULL);
+    if(mbcsTable->swapLFNLStateTable==NULL) {
+        mbcsTable->swapLFNLStateTable=newStateTable;
+        mbcsTable->swapLFNLFromUnicodeBytes=(uint8_t *)newResults;
+        mbcsTable->swapLFNLName=name;
+
+        newStateTable=NULL;
+    }
+    icu::umtx_unlock(NULL);
+
+    /* release the allocated memory if another thread beat us to it */
+    if(newStateTable!=NULL) {
+        uprv_free(newStateTable);
+    }
+    return TRUE;
+}
+
+/* reconstitute omitted fromUnicode data ------------------------------------ */
+
+/* for details, compare with genmbcs.c MBCSAddFromUnicode() and transformEUC() */
+static UBool U_CALLCONV
+writeStage3Roundtrip(const void *context, uint32_t value, UChar32 codePoints[32]) {
+    UConverterMBCSTable *mbcsTable=(UConverterMBCSTable *)context;
+    const uint16_t *table;
+    uint32_t *stage2;
+    uint8_t *bytes, *p;
+    UChar32 c;
+    int32_t i, st3;
+
+    table=mbcsTable->fromUnicodeTable;
+    bytes=(uint8_t *)mbcsTable->fromUnicodeBytes;
+
+    /* for EUC outputTypes, modify the value like genmbcs.c's transformEUC() */
+    switch(mbcsTable->outputType) {
+    case MBCS_OUTPUT_3_EUC:
+        if(value<=0xffff) {
+            /* short sequences are stored directly */
+            /* code set 0 or 1 */
+        } else if(value<=0x8effff) {
+            /* code set 2 */
+            value&=0x7fff;
+        } else /* first byte is 0x8f */ {
+            /* code set 3 */
+            value&=0xff7f;
+        }
+        break;
+    case MBCS_OUTPUT_4_EUC:
+        if(value<=0xffffff) {
+            /* short sequences are stored directly */
+            /* code set 0 or 1 */
+        } else if(value<=0x8effffff) {
+            /* code set 2 */
+            value&=0x7fffff;
+        } else /* first byte is 0x8f */ {
+            /* code set 3 */
+            value&=0xff7fff;
+        }
+        break;
+    default:
+        break;
+    }
+
+    for(i=0; i<=0x1f; ++value, ++i) {
+        c=codePoints[i];
+        if(c<0) {
+            continue;
+        }
+
+        /* locate the stage 2 & 3 data */
+        stage2=((uint32_t *)table)+table[c>>10]+((c>>4)&0x3f);
+        p=bytes;
+        st3=(int32_t)(uint16_t)*stage2*16+(c&0xf);
+
+        /* write the codepage bytes into stage 3 */
+        switch(mbcsTable->outputType) {
+        case MBCS_OUTPUT_3:
+        case MBCS_OUTPUT_4_EUC:
+            p+=st3*3;
+            p[0]=(uint8_t)(value>>16);
+            p[1]=(uint8_t)(value>>8);
+            p[2]=(uint8_t)value;
+            break;
+        case MBCS_OUTPUT_4:
+            ((uint32_t *)p)[st3]=value;
+            break;
+        default:
+            /* 2 bytes per character */
+            ((uint16_t *)p)[st3]=(uint16_t)value;
+            break;
+        }
+
+        /* set the roundtrip flag */
+        *stage2|=(1UL<<(16+(c&0xf)));
+    }
+    return TRUE;
+ }
+
+static void
+reconstituteData(UConverterMBCSTable *mbcsTable,
+                 uint32_t stage1Length, uint32_t stage2Length,
+                 uint32_t fullStage2Length,  /* lengths are numbers of units, not bytes */
+                 UErrorCode *pErrorCode) {
+    uint16_t *stage1;
+    uint32_t *stage2;
+    uint32_t dataLength=stage1Length*2+fullStage2Length*4+mbcsTable->fromUBytesLength;
+    mbcsTable->reconstitutedData=(uint8_t *)uprv_malloc(dataLength);
+    if(mbcsTable->reconstitutedData==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    uprv_memset(mbcsTable->reconstitutedData, 0, dataLength);
+
+    /* copy existing data and reroute the pointers */
+    stage1=(uint16_t *)mbcsTable->reconstitutedData;
+    uprv_memcpy(stage1, mbcsTable->fromUnicodeTable, stage1Length*2);
+
+    stage2=(uint32_t *)(stage1+stage1Length);
+    uprv_memcpy(stage2+(fullStage2Length-stage2Length),
+                mbcsTable->fromUnicodeTable+stage1Length,
+                stage2Length*4);
+
+    mbcsTable->fromUnicodeTable=stage1;
+    mbcsTable->fromUnicodeBytes=(uint8_t *)(stage2+fullStage2Length);
+
+    /* indexes into stage 2 count from the bottom of the fromUnicodeTable */
+    stage2=(uint32_t *)stage1;
+
+    /* reconstitute the initial part of stage 2 from the mbcsIndex */
+    {
+        int32_t stageUTF8Length=((int32_t)mbcsTable->maxFastUChar+1)>>6;
+        int32_t stageUTF8Index=0;
+        int32_t st1, st2, st3, i;
+
+        for(st1=0; stageUTF8Index<stageUTF8Length; ++st1) {
+            st2=stage1[st1];
+            if(st2!=(int32_t)stage1Length/2) {
+                /* each stage 2 block has 64 entries corresponding to 16 entries in the mbcsIndex */
+                for(i=0; i<16; ++i) {
+                    st3=mbcsTable->mbcsIndex[stageUTF8Index++];
+                    if(st3!=0) {
+                        /* an stage 2 entry's index is per stage 3 16-block, not per stage 3 entry */
+                        st3>>=4;
+                        /*
+                         * 4 stage 2 entries point to 4 consecutive stage 3 16-blocks which are
+                         * allocated together as a single 64-block for access from the mbcsIndex
+                         */
+                        stage2[st2++]=st3++;
+                        stage2[st2++]=st3++;
+                        stage2[st2++]=st3++;
+                        stage2[st2++]=st3;
+                    } else {
+                        /* no stage 3 block, skip */
+                        st2+=4;
+                    }
+                }
+            } else {
+                /* no stage 2 block, skip */
+                stageUTF8Index+=16;
+            }
+        }
+    }
+
+    /* reconstitute fromUnicodeBytes with roundtrips from toUnicode data */
+    ucnv_MBCSEnumToUnicode(mbcsTable, writeStage3Roundtrip, mbcsTable, pErrorCode);
+}
+
+/* MBCS setup functions ----------------------------------------------------- */
+
+static void U_CALLCONV
+ucnv_MBCSLoad(UConverterSharedData *sharedData,
+          UConverterLoadArgs *pArgs,
+          const uint8_t *raw,
+          UErrorCode *pErrorCode) {
+    UDataInfo info;
+    UConverterMBCSTable *mbcsTable=&sharedData->mbcs;
+    _MBCSHeader *header=(_MBCSHeader *)raw;
+    uint32_t offset;
+    uint32_t headerLength;
+    UBool noFromU=FALSE;
+
+    if(header->version[0]==4) {
+        headerLength=MBCS_HEADER_V4_LENGTH;
+    } else if(header->version[0]==5 && header->version[1]>=3 &&
+              (header->options&MBCS_OPT_UNKNOWN_INCOMPATIBLE_MASK)==0) {
+        headerLength=header->options&MBCS_OPT_LENGTH_MASK;
+        noFromU=(UBool)((header->options&MBCS_OPT_NO_FROM_U)!=0);
+    } else {
+        *pErrorCode=U_INVALID_TABLE_FORMAT;
+        return;
+    }
+
+    mbcsTable->outputType=(uint8_t)header->flags;
+    if(noFromU && mbcsTable->outputType==MBCS_OUTPUT_1) {
+        *pErrorCode=U_INVALID_TABLE_FORMAT;
+        return;
+    }
+
+    /* extension data, header version 4.2 and higher */
+    offset=header->flags>>8;
+    if(offset!=0) {
+        mbcsTable->extIndexes=(const int32_t *)(raw+offset);
+    }
+
+    if(mbcsTable->outputType==MBCS_OUTPUT_EXT_ONLY) {
+        UConverterLoadArgs args=UCNV_LOAD_ARGS_INITIALIZER;
+        UConverterSharedData *baseSharedData;
+        const int32_t *extIndexes;
+        const char *baseName;
+
+        /* extension-only file, load the base table and set values appropriately */
+        if((extIndexes=mbcsTable->extIndexes)==NULL) {
+            /* extension-only file without extension */
+            *pErrorCode=U_INVALID_TABLE_FORMAT;
+            return;
+        }
+
+        if(pArgs->nestedLoads!=1) {
+            /* an extension table must not be loaded as a base table */
+            *pErrorCode=U_INVALID_TABLE_FILE;
+            return;
+        }
+
+        /* load the base table */
+        baseName=(const char *)header+headerLength*4;
+        if(0==uprv_strcmp(baseName, sharedData->staticData->name)) {
+            /* forbid loading this same extension-only file */
+            *pErrorCode=U_INVALID_TABLE_FORMAT;
+            return;
+        }
+
+        /* TODO parse package name out of the prefix of the base name in the extension .cnv file? */
+        args.size=sizeof(UConverterLoadArgs);
+        args.nestedLoads=2;
+        args.onlyTestIsLoadable=pArgs->onlyTestIsLoadable;
+        args.reserved=pArgs->reserved;
+        args.options=pArgs->options;
+        args.pkg=pArgs->pkg;
+        args.name=baseName;
+        baseSharedData=ucnv_load(&args, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return;
+        }
+        if( baseSharedData->staticData->conversionType!=UCNV_MBCS ||
+            baseSharedData->mbcs.baseSharedData!=NULL
+        ) {
+            ucnv_unload(baseSharedData);
+            *pErrorCode=U_INVALID_TABLE_FORMAT;
+            return;
+        }
+        if(pArgs->onlyTestIsLoadable) {
+            /*
+             * Exit as soon as we know that we can load the converter
+             * and the format is valid and supported.
+             * The worst that can happen in the following code is a memory
+             * allocation error.
+             */
+            ucnv_unload(baseSharedData);
+            return;
+        }
+
+        /* copy the base table data */
+        uprv_memcpy(mbcsTable, &baseSharedData->mbcs, sizeof(UConverterMBCSTable));
+
+        /* overwrite values with relevant ones for the extension converter */
+        mbcsTable->baseSharedData=baseSharedData;
+        mbcsTable->extIndexes=extIndexes;
+
+        /*
+         * It would be possible to share the swapLFNL data with a base converter,
+         * but the generated name would have to be different, and the memory
+         * would have to be free'd only once.
+         * It is easier to just create the data for the extension converter
+         * separately when it is requested.
+         */
+        mbcsTable->swapLFNLStateTable=NULL;
+        mbcsTable->swapLFNLFromUnicodeBytes=NULL;
+        mbcsTable->swapLFNLName=NULL;
+
+        /*
+         * The reconstitutedData must be deleted only when the base converter
+         * is unloaded.
+         */
+        mbcsTable->reconstitutedData=NULL;
+
+        /*
+         * Set a special, runtime-only outputType if the extension converter
+         * is a DBCS version of a base converter that also maps single bytes.
+         */
+        if( sharedData->staticData->conversionType==UCNV_DBCS ||
+                (sharedData->staticData->conversionType==UCNV_MBCS &&
+                 sharedData->staticData->minBytesPerChar>=2)
+        ) {
+            if(baseSharedData->mbcs.outputType==MBCS_OUTPUT_2_SISO) {
+                /* the base converter is SI/SO-stateful */
+                int32_t entry;
+
+                /* get the dbcs state from the state table entry for SO=0x0e */
+                entry=mbcsTable->stateTable[0][0xe];
+                if( MBCS_ENTRY_IS_FINAL(entry) &&
+                    MBCS_ENTRY_FINAL_ACTION(entry)==MBCS_STATE_CHANGE_ONLY &&
+                    MBCS_ENTRY_FINAL_STATE(entry)!=0
+                ) {
+                    mbcsTable->dbcsOnlyState=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry);
+
+                    mbcsTable->outputType=MBCS_OUTPUT_DBCS_ONLY;
+                }
+            } else if(
+                baseSharedData->staticData->conversionType==UCNV_MBCS &&
+                baseSharedData->staticData->minBytesPerChar==1 &&
+                baseSharedData->staticData->maxBytesPerChar==2 &&
+                mbcsTable->countStates<=127
+            ) {
+                /* non-stateful base converter, need to modify the state table */
+                int32_t (*newStateTable)[256];
+                int32_t *state;
+                int32_t i, count;
+
+                /* allocate a new state table and copy the base state table contents */
+                count=mbcsTable->countStates;
+                newStateTable=(int32_t (*)[256])uprv_malloc((count+1)*1024);
+                if(newStateTable==NULL) {
+                    ucnv_unload(baseSharedData);
+                    *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return;
+                }
+
+                uprv_memcpy(newStateTable, mbcsTable->stateTable, count*1024);
+
+                /* change all final single-byte entries to go to a new all-illegal state */
+                state=newStateTable[0];
+                for(i=0; i<256; ++i) {
+                    if(MBCS_ENTRY_IS_FINAL(state[i])) {
+                        state[i]=MBCS_ENTRY_TRANSITION(count, 0);
+                    }
+                }
+
+                /* build the new all-illegal state */
+                state=newStateTable[count];
+                for(i=0; i<256; ++i) {
+                    state[i]=MBCS_ENTRY_FINAL(0, MBCS_STATE_ILLEGAL, 0);
+                }
+                mbcsTable->stateTable=(const int32_t (*)[256])newStateTable;
+                mbcsTable->countStates=(uint8_t)(count+1);
+                mbcsTable->stateTableOwned=TRUE;
+
+                mbcsTable->outputType=MBCS_OUTPUT_DBCS_ONLY;
+            }
+        }
+
+        /*
+         * unlike below for files with base tables, do not get the unicodeMask
+         * from the sharedData; instead, use the base table's unicodeMask,
+         * which we copied in the memcpy above;
+         * this is necessary because the static data unicodeMask, especially
+         * the UCNV_HAS_SUPPLEMENTARY flag, is part of the base table data
+         */
+    } else {
+        /* conversion file with a base table; an additional extension table is optional */
+        /* make sure that the output type is known */
+        switch(mbcsTable->outputType) {
+        case MBCS_OUTPUT_1:
+        case MBCS_OUTPUT_2:
+        case MBCS_OUTPUT_3:
+        case MBCS_OUTPUT_4:
+        case MBCS_OUTPUT_3_EUC:
+        case MBCS_OUTPUT_4_EUC:
+        case MBCS_OUTPUT_2_SISO:
+            /* OK */
+            break;
+        default:
+            *pErrorCode=U_INVALID_TABLE_FORMAT;
+            return;
+        }
+        if(pArgs->onlyTestIsLoadable) {
+            /*
+             * Exit as soon as we know that we can load the converter
+             * and the format is valid and supported.
+             * The worst that can happen in the following code is a memory
+             * allocation error.
+             */
+            return;
+        }
+
+        mbcsTable->countStates=(uint8_t)header->countStates;
+        mbcsTable->countToUFallbacks=header->countToUFallbacks;
+        mbcsTable->stateTable=(const int32_t (*)[256])(raw+headerLength*4);
+        mbcsTable->toUFallbacks=(const _MBCSToUFallback *)(mbcsTable->stateTable+header->countStates);
+        mbcsTable->unicodeCodeUnits=(const uint16_t *)(raw+header->offsetToUCodeUnits);
+
+        mbcsTable->fromUnicodeTable=(const uint16_t *)(raw+header->offsetFromUTable);
+        mbcsTable->fromUnicodeBytes=(const uint8_t *)(raw+header->offsetFromUBytes);
+        mbcsTable->fromUBytesLength=header->fromUBytesLength;
+
+        /*
+         * converter versions 6.1 and up contain a unicodeMask that is
+         * used here to select the most efficient function implementations
+         */
+        info.size=sizeof(UDataInfo);
+        udata_getInfo((UDataMemory *)sharedData->dataMemory, &info);
+        if(info.formatVersion[0]>6 || (info.formatVersion[0]==6 && info.formatVersion[1]>=1)) {
+            /* mask off possible future extensions to be safe */
+            mbcsTable->unicodeMask=(uint8_t)(sharedData->staticData->unicodeMask&3);
+        } else {
+            /* for older versions, assume worst case: contains anything possible (prevent over-optimizations) */
+            mbcsTable->unicodeMask=UCNV_HAS_SUPPLEMENTARY|UCNV_HAS_SURROGATES;
+        }
+
+        /*
+         * _MBCSHeader.version 4.3 adds utf8Friendly data structures.
+         * Check for the header version, SBCS vs. MBCS, and for whether the
+         * data structures are optimized for code points as high as what the
+         * runtime code is designed for.
+         * The implementation does not handle mapping tables with entries for
+         * unpaired surrogates.
+         */
+        if( header->version[1]>=3 &&
+            (mbcsTable->unicodeMask&UCNV_HAS_SURROGATES)==0 &&
+            (mbcsTable->countStates==1 ?
+                (header->version[2]>=(SBCS_FAST_MAX>>8)) :
+                (header->version[2]>=(MBCS_FAST_MAX>>8))
+            )
+        ) {
+            mbcsTable->utf8Friendly=TRUE;
+
+            if(mbcsTable->countStates==1) {
+                /*
+                 * SBCS: Stage 3 is allocated in 64-entry blocks for U+0000..SBCS_FAST_MAX or higher.
+                 * Build a table with indexes to each block, to be used instead of
+                 * the regular stage 1/2 table.
+                 */
+                int32_t i;
+                for(i=0; i<(SBCS_FAST_LIMIT>>6); ++i) {
+                    mbcsTable->sbcsIndex[i]=mbcsTable->fromUnicodeTable[mbcsTable->fromUnicodeTable[i>>4]+((i<<2)&0x3c)];
+                }
+                /* set SBCS_FAST_MAX to reflect the reach of sbcsIndex[] even if header->version[2]>(SBCS_FAST_MAX>>8) */
+                mbcsTable->maxFastUChar=SBCS_FAST_MAX;
+            } else {
+                /*
+                 * MBCS: Stage 3 is allocated in 64-entry blocks for U+0000..MBCS_FAST_MAX or higher.
+                 * The .cnv file is prebuilt with an additional stage table with indexes
+                 * to each block.
+                 */
+                mbcsTable->mbcsIndex=(const uint16_t *)
+                    (mbcsTable->fromUnicodeBytes+
+                     (noFromU ? 0 : mbcsTable->fromUBytesLength));
+                mbcsTable->maxFastUChar=(((UChar)header->version[2])<<8)|0xff;
+            }
+        }
+
+        /* calculate a bit set of 4 ASCII characters per bit that round-trip to ASCII bytes */
+        {
+            uint32_t asciiRoundtrips=0xffffffff;
+            int32_t i;
+
+            for(i=0; i<0x80; ++i) {
+                if(mbcsTable->stateTable[0][i]!=MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_DIRECT_16, i)) {
+                    asciiRoundtrips&=~((uint32_t)1<<(i>>2));
+                }
+            }
+            mbcsTable->asciiRoundtrips=asciiRoundtrips;
+        }
+
+        if(noFromU) {
+            uint32_t stage1Length=
+                mbcsTable->unicodeMask&UCNV_HAS_SUPPLEMENTARY ?
+                    0x440 : 0x40;
+            uint32_t stage2Length=
+                (header->offsetFromUBytes-header->offsetFromUTable)/4-
+                stage1Length/2;
+            reconstituteData(mbcsTable, stage1Length, stage2Length, header->fullStage2Length, pErrorCode);
+        }
+    }
+
+    /* Set the impl pointer here so that it is set for both extension-only and base tables. */
+    if(mbcsTable->utf8Friendly) {
+        if(mbcsTable->countStates==1) {
+            sharedData->impl=&_SBCSUTF8Impl;
+        } else {
+            if(mbcsTable->outputType==MBCS_OUTPUT_2) {
+                sharedData->impl=&_DBCSUTF8Impl;
+            }
+        }
+    }
+
+    if(mbcsTable->outputType==MBCS_OUTPUT_DBCS_ONLY || mbcsTable->outputType==MBCS_OUTPUT_2_SISO) {
+        /*
+         * MBCS_OUTPUT_DBCS_ONLY: No SBCS mappings, therefore ASCII does not roundtrip.
+         * MBCS_OUTPUT_2_SISO: Bypass the ASCII fastpath to handle prevLength correctly.
+         */
+        mbcsTable->asciiRoundtrips=0;
+    }
+}
+
+static void U_CALLCONV
+ucnv_MBCSUnload(UConverterSharedData *sharedData) {
+    UConverterMBCSTable *mbcsTable=&sharedData->mbcs;
+
+    if(mbcsTable->swapLFNLStateTable!=NULL) {
+        uprv_free(mbcsTable->swapLFNLStateTable);
+    }
+    if(mbcsTable->stateTableOwned) {
+        uprv_free((void *)mbcsTable->stateTable);
+    }
+    if(mbcsTable->baseSharedData!=NULL) {
+        ucnv_unload(mbcsTable->baseSharedData);
+    }
+    if(mbcsTable->reconstitutedData!=NULL) {
+        uprv_free(mbcsTable->reconstitutedData);
+    }
+}
+
+static void U_CALLCONV
+ucnv_MBCSOpen(UConverter *cnv,
+              UConverterLoadArgs *pArgs,
+              UErrorCode *pErrorCode) {
+    UConverterMBCSTable *mbcsTable;
+    const int32_t *extIndexes;
+    uint8_t outputType;
+    int8_t maxBytesPerUChar;
+
+    if(pArgs->onlyTestIsLoadable) {
+        return;
+    }
+
+    mbcsTable=&cnv->sharedData->mbcs;
+    outputType=mbcsTable->outputType;
+
+    if(outputType==MBCS_OUTPUT_DBCS_ONLY) {
+        /* the swaplfnl option does not apply, remove it */
+        cnv->options=pArgs->options&=~UCNV_OPTION_SWAP_LFNL;
+    }
+
+    if((pArgs->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        /* do this because double-checked locking is broken */
+        UBool isCached;
+
+        icu::umtx_lock(NULL);
+        isCached=mbcsTable->swapLFNLStateTable!=NULL;
+        icu::umtx_unlock(NULL);
+
+        if(!isCached) {
+            if(!_EBCDICSwapLFNL(cnv->sharedData, pErrorCode)) {
+                if(U_FAILURE(*pErrorCode)) {
+                    return; /* something went wrong */
+                }
+
+                /* the option does not apply, remove it */
+                cnv->options=pArgs->options&=~UCNV_OPTION_SWAP_LFNL;
+            }
+        }
+    }
+
+    if(uprv_strstr(pArgs->name, "18030")!=NULL) {
+        if(uprv_strstr(pArgs->name, "gb18030")!=NULL || uprv_strstr(pArgs->name, "GB18030")!=NULL) {
+            /* set a flag for GB 18030 mode, which changes the callback behavior */
+            cnv->options|=_MBCS_OPTION_GB18030;
+        }
+    } else if((uprv_strstr(pArgs->name, "KEIS")!=NULL) || (uprv_strstr(pArgs->name, "keis")!=NULL)) {
+        /* set a flag for KEIS converter, which changes the SI/SO character sequence */
+        cnv->options|=_MBCS_OPTION_KEIS;
+    } else if((uprv_strstr(pArgs->name, "JEF")!=NULL) || (uprv_strstr(pArgs->name, "jef")!=NULL)) {
+        /* set a flag for JEF converter, which changes the SI/SO character sequence */
+        cnv->options|=_MBCS_OPTION_JEF;
+    } else if((uprv_strstr(pArgs->name, "JIPS")!=NULL) || (uprv_strstr(pArgs->name, "jips")!=NULL)) {
+        /* set a flag for JIPS converter, which changes the SI/SO character sequence */
+        cnv->options|=_MBCS_OPTION_JIPS;
+    }
+
+    /* fix maxBytesPerUChar depending on outputType and options etc. */
+    if(outputType==MBCS_OUTPUT_2_SISO) {
+        cnv->maxBytesPerUChar=3; /* SO+DBCS */
+    }
+
+    extIndexes=mbcsTable->extIndexes;
+    if(extIndexes!=NULL) {
+        maxBytesPerUChar=(int8_t)UCNV_GET_MAX_BYTES_PER_UCHAR(extIndexes);
+        if(outputType==MBCS_OUTPUT_2_SISO) {
+            ++maxBytesPerUChar; /* SO + multiple DBCS */
+        }
+
+        if(maxBytesPerUChar>cnv->maxBytesPerUChar) {
+            cnv->maxBytesPerUChar=maxBytesPerUChar;
+        }
+    }
+
+#if 0
+    /*
+     * documentation of UConverter fields used for status
+     * all of these fields are (re)set to 0 by ucnv_bld.c and ucnv_reset()
+     */
+
+    /* toUnicode */
+    cnv->toUnicodeStatus=0;     /* offset */
+    cnv->mode=0;                /* state */
+    cnv->toULength=0;           /* byteIndex */
+
+    /* fromUnicode */
+    cnv->fromUChar32=0;
+    cnv->fromUnicodeStatus=1;   /* prevLength */
+#endif
+}
+
+U_CDECL_BEGIN
+
+static const char* U_CALLCONV
+ucnv_MBCSGetName(const UConverter *cnv) {
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0 && cnv->sharedData->mbcs.swapLFNLName!=NULL) {
+        return cnv->sharedData->mbcs.swapLFNLName;
+    } else {
+        return cnv->sharedData->staticData->name;
+    }
+}
+U_CDECL_END
+
+
+/* MBCS-to-Unicode conversion functions ------------------------------------- */
+
+static UChar32 U_CALLCONV
+ucnv_MBCSGetFallback(UConverterMBCSTable *mbcsTable, uint32_t offset) {
+    const _MBCSToUFallback *toUFallbacks;
+    uint32_t i, start, limit;
+
+    limit=mbcsTable->countToUFallbacks;
+    if(limit>0) {
+        /* do a binary search for the fallback mapping */
+        toUFallbacks=mbcsTable->toUFallbacks;
+        start=0;
+        while(start<limit-1) {
+            i=(start+limit)/2;
+            if(offset<toUFallbacks[i].offset) {
+                limit=i;
+            } else {
+                start=i;
+            }
+        }
+
+        /* did we really find it? */
+        if(offset==toUFallbacks[start].offset) {
+            return toUFallbacks[start].codePoint;
+        }
+    }
+
+    return 0xfffe;
+}
+
+/* This version of ucnv_MBCSToUnicodeWithOffsets() is optimized for single-byte, single-state codepages. */
+static void
+ucnv_MBCSSingleToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                                UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+
+    const int32_t (*stateTable)[256];
+
+    int32_t sourceIndex;
+
+    int32_t entry;
+    UChar c;
+    uint8_t action;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        stateTable=(const int32_t (*)[256])cnv->sharedData->mbcs.swapLFNLStateTable;
+    } else {
+        stateTable=cnv->sharedData->mbcs.stateTable;
+    }
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=0;
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        /*
+         * This following test is to see if available input would overflow the output.
+         * It does not catch output of more than one code unit that
+         * overflows as a result of a surrogate pair or callback output
+         * from the last source byte.
+         * Therefore, those situations also test for overflows and will
+         * then break the loop, too.
+         */
+        if(target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+        entry=stateTable[0][*source++];
+        /* MBCS_ENTRY_IS_FINAL(entry) */
+
+        /* test the most common case first */
+        if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+            /* output BMP code point */
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+
+            /* normal end of action codes: prepare for a new character */
+            ++sourceIndex;
+            continue;
+        }
+
+        /*
+         * An if-else-if chain provides more reliable performance for
+         * the most common cases compared to a switch.
+         */
+        action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+        if(action==MBCS_STATE_VALID_DIRECT_20 ||
+           (action==MBCS_STATE_FALLBACK_DIRECT_20 && UCNV_TO_U_USE_FALLBACK(cnv))
+        ) {
+            entry=MBCS_ENTRY_FINAL_VALUE(entry);
+            /* output surrogate pair */
+            *target++=(UChar)(0xd800|(UChar)(entry>>10));
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            c=(UChar)(0xdc00|(UChar)(entry&0x3ff));
+            if(target<targetLimit) {
+                *target++=c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+            } else {
+                /* target overflow */
+                cnv->UCharErrorBuffer[0]=c;
+                cnv->UCharErrorBufferLength=1;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+
+            ++sourceIndex;
+            continue;
+        } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+            if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                /* output BMP code point */
+                *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+
+                ++sourceIndex;
+                continue;
+            }
+        } else if(action==MBCS_STATE_UNASSIGNED) {
+            /* just fall through */
+        } else if(action==MBCS_STATE_ILLEGAL) {
+            /* callback(illegal) */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        } else {
+            /* reserved, must never occur */
+            ++sourceIndex;
+            continue;
+        }
+
+        if(U_FAILURE(*pErrorCode)) {
+            /* callback(illegal) */
+            break;
+        } else /* unassigned sequences indicated with byteIndex>0 */ {
+            /* try an extension mapping */
+            pArgs->source=(const char *)source;
+            cnv->toUBytes[0]=*(source-1);
+            cnv->toULength=_extToU(cnv, cnv->sharedData,
+                                    1, &source, sourceLimit,
+                                    &target, targetLimit,
+                                    &offsets, sourceIndex,
+                                    pArgs->flush,
+                                    pErrorCode);
+            sourceIndex+=1+(int32_t)(source-(const uint8_t *)pArgs->source);
+
+            if(U_FAILURE(*pErrorCode)) {
+                /* not mappable or buffer overflow */
+                break;
+            }
+        }
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+/*
+ * This version of ucnv_MBCSSingleToUnicodeWithOffsets() is optimized for single-byte, single-state codepages
+ * that only map to and from the BMP.
+ * In addition to single-byte optimizations, the offset calculations
+ * become much easier.
+ */
+static void
+ucnv_MBCSSingleToBMPWithOffsets(UConverterToUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit, *lastSource;
+    UChar *target;
+    int32_t targetCapacity, length;
+    int32_t *offsets;
+
+    const int32_t (*stateTable)[256];
+
+    int32_t sourceIndex;
+
+    int32_t entry;
+    uint8_t action;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        stateTable=(const int32_t (*)[256])cnv->sharedData->mbcs.swapLFNLStateTable;
+    } else {
+        stateTable=cnv->sharedData->mbcs.stateTable;
+    }
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=0;
+    lastSource=source;
+
+    /*
+     * since the conversion here is 1:1 UChar:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)(sourceLimit-source);
+    if(length<targetCapacity) {
+        targetCapacity=length;
+    }
+
+#if MBCS_UNROLL_SINGLE_TO_BMP
+    /* unrolling makes it faster on Pentium III/Windows 2000 */
+    /* unroll the loop with the most common case */
+unrolled:
+    if(targetCapacity>=16) {
+        int32_t count, loops, oredEntries;
+
+        loops=count=targetCapacity>>4;
+        do {
+            oredEntries=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            oredEntries|=entry=stateTable[0][*source++];
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+
+            /* were all 16 entries really valid? */
+            if(!MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(oredEntries)) {
+                /* no, return to the first of these 16 */
+                source-=16;
+                target-=16;
+                break;
+            }
+        } while(--count>0);
+        count=loops-count;
+        targetCapacity-=16*count;
+
+        if(offsets!=NULL) {
+            lastSource+=16*count;
+            while(count>0) {
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                --count;
+            }
+        }
+    }
+#endif
+
+    /* conversion loop */
+    while(targetCapacity > 0 && source < sourceLimit) {
+        entry=stateTable[0][*source++];
+        /* MBCS_ENTRY_IS_FINAL(entry) */
+
+        /* test the most common case first */
+        if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+            /* output BMP code point */
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            --targetCapacity;
+            continue;
+        }
+
+        /*
+         * An if-else-if chain provides more reliable performance for
+         * the most common cases compared to a switch.
+         */
+        action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+        if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+            if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                /* output BMP code point */
+                *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                --targetCapacity;
+                continue;
+            }
+        } else if(action==MBCS_STATE_UNASSIGNED) {
+            /* just fall through */
+        } else if(action==MBCS_STATE_ILLEGAL) {
+            /* callback(illegal) */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        } else {
+            /* reserved, must never occur */
+            continue;
+        }
+
+        /* set offsets since the start or the last extension */
+        if(offsets!=NULL) {
+            int32_t count=(int32_t)(source-lastSource);
+
+            /* predecrement: do not set the offset for the callback-causing character */
+            while(--count>0) {
+                *offsets++=sourceIndex++;
+            }
+            /* offset and sourceIndex are now set for the current character */
+        }
+
+        if(U_FAILURE(*pErrorCode)) {
+            /* callback(illegal) */
+            break;
+        } else /* unassigned sequences indicated with byteIndex>0 */ {
+            /* try an extension mapping */
+            lastSource=source;
+            cnv->toUBytes[0]=*(source-1);
+            cnv->toULength=_extToU(cnv, cnv->sharedData,
+                                    1, &source, sourceLimit,
+                                    &target, pArgs->targetLimit,
+                                    &offsets, sourceIndex,
+                                    pArgs->flush,
+                                    pErrorCode);
+            sourceIndex+=1+(int32_t)(source-lastSource);
+
+            if(U_FAILURE(*pErrorCode)) {
+                /* not mappable or buffer overflow */
+                break;
+            }
+
+            /* recalculate the targetCapacity after an extension mapping */
+            targetCapacity=(int32_t)(pArgs->targetLimit-target);
+            length=(int32_t)(sourceLimit-source);
+            if(length<targetCapacity) {
+                targetCapacity=length;
+            }
+        }
+
+#if MBCS_UNROLL_SINGLE_TO_BMP
+        /* unrolling makes it faster on Pentium III/Windows 2000 */
+        goto unrolled;
+#endif
+    }
+
+    if(U_SUCCESS(*pErrorCode) && source<sourceLimit && target>=pArgs->targetLimit) {
+        /* target is full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* set offsets since the start or the last callback */
+    if(offsets!=NULL) {
+        size_t count=source-lastSource;
+        while(count>0) {
+            *offsets++=sourceIndex++;
+            --count;
+        }
+    }
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+static UBool
+hasValidTrailBytes(const int32_t (*stateTable)[256], uint8_t state) {
+    const int32_t *row=stateTable[state];
+    int32_t b, entry;
+    /* First test for final entries in this state for some commonly valid byte values. */
+    entry=row[0xa1];
+    if( !MBCS_ENTRY_IS_TRANSITION(entry) &&
+        MBCS_ENTRY_FINAL_ACTION(entry)!=MBCS_STATE_ILLEGAL
+    ) {
+        return TRUE;
+    }
+    entry=row[0x41];
+    if( !MBCS_ENTRY_IS_TRANSITION(entry) &&
+        MBCS_ENTRY_FINAL_ACTION(entry)!=MBCS_STATE_ILLEGAL
+    ) {
+        return TRUE;
+    }
+    /* Then test for final entries in this state. */
+    for(b=0; b<=0xff; ++b) {
+        entry=row[b];
+        if( !MBCS_ENTRY_IS_TRANSITION(entry) &&
+            MBCS_ENTRY_FINAL_ACTION(entry)!=MBCS_STATE_ILLEGAL
+        ) {
+            return TRUE;
+        }
+    }
+    /* Then recurse for transition entries. */
+    for(b=0; b<=0xff; ++b) {
+        entry=row[b];
+        if( MBCS_ENTRY_IS_TRANSITION(entry) &&
+            hasValidTrailBytes(stateTable, (uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry))
+        ) {
+            return TRUE;
+        }
+    }
+    return FALSE;
+}
+
+/*
+ * Is byte b a single/lead byte in this state?
+ * Recurse for transition states, because here we don't want to say that
+ * b is a lead byte if all byte sequences that start with b are illegal.
+ */
+static UBool
+isSingleOrLead(const int32_t (*stateTable)[256], uint8_t state, UBool isDBCSOnly, uint8_t b) {
+    const int32_t *row=stateTable[state];
+    int32_t entry=row[b];
+    if(MBCS_ENTRY_IS_TRANSITION(entry)) {   /* lead byte */
+        return hasValidTrailBytes(stateTable, (uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry));
+    } else {
+        uint8_t action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+        if(action==MBCS_STATE_CHANGE_ONLY && isDBCSOnly) {
+            return FALSE;   /* SI/SO are illegal for DBCS-only conversion */
+        } else {
+            return action!=MBCS_STATE_ILLEGAL;
+        }
+    }
+}
+
+U_CFUNC void
+ucnv_MBCSToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                          UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+
+    const int32_t (*stateTable)[256];
+    const uint16_t *unicodeCodeUnits;
+
+    uint32_t offset;
+    uint8_t state;
+    int8_t byteIndex;
+    uint8_t *bytes;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    int32_t entry;
+    UChar c;
+    uint8_t action;
+
+    /* use optimized function if possible */
+    cnv=pArgs->converter;
+
+    if(cnv->preToULength>0) {
+        /*
+         * pass sourceIndex=-1 because we continue from an earlier buffer
+         * in the future, this may change with continuous offsets
+         */
+        ucnv_extContinueMatchToU(cnv, pArgs, -1, pErrorCode);
+
+        if(U_FAILURE(*pErrorCode) || cnv->preToULength<0) {
+            return;
+        }
+    }
+
+    if(cnv->sharedData->mbcs.countStates==1) {
+        if(!(cnv->sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+            ucnv_MBCSSingleToBMPWithOffsets(pArgs, pErrorCode);
+        } else {
+            ucnv_MBCSSingleToUnicodeWithOffsets(pArgs, pErrorCode);
+        }
+        return;
+    }
+
+    /* set up the local pointers */
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        stateTable=(const int32_t (*)[256])cnv->sharedData->mbcs.swapLFNLStateTable;
+    } else {
+        stateTable=cnv->sharedData->mbcs.stateTable;
+    }
+    unicodeCodeUnits=cnv->sharedData->mbcs.unicodeCodeUnits;
+
+    /* get the converter state from UConverter */
+    offset=cnv->toUnicodeStatus;
+    byteIndex=cnv->toULength;
+    bytes=cnv->toUBytes;
+
+    /*
+     * if we are in the SBCS state for a DBCS-only converter,
+     * then load the DBCS state from the MBCS data
+     * (dbcsOnlyState==0 if it is not a DBCS-only converter)
+     */
+    if((state=(uint8_t)(cnv->mode))==0) {
+        state=cnv->sharedData->mbcs.dbcsOnlyState;
+    }
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=byteIndex==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        /*
+         * This following test is to see if available input would overflow the output.
+         * It does not catch output of more than one code unit that
+         * overflows as a result of a surrogate pair or callback output
+         * from the last source byte.
+         * Therefore, those situations also test for overflows and will
+         * then break the loop, too.
+         */
+        if(target>=targetLimit) {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+
+        if(byteIndex==0) {
+            /* optimized loop for 1/2-byte input and BMP output */
+            if(offsets==NULL) {
+                do {
+                    entry=stateTable[state][*source];
+                    if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+                        state=(uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry);
+                        offset=MBCS_ENTRY_TRANSITION_OFFSET(entry);
+
+                        ++source;
+                        if( source<sourceLimit &&
+                            MBCS_ENTRY_IS_FINAL(entry=stateTable[state][*source]) &&
+                            MBCS_ENTRY_FINAL_ACTION(entry)==MBCS_STATE_VALID_16 &&
+                            (c=unicodeCodeUnits[offset+MBCS_ENTRY_FINAL_VALUE_16(entry)])<0xfffe
+                        ) {
+                            ++source;
+                            *target++=c;
+                            state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+                            offset=0;
+                        } else {
+                            /* set the state and leave the optimized loop */
+                            bytes[0]=*(source-1);
+                            byteIndex=1;
+                            break;
+                        }
+                    } else {
+                        if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+                            /* output BMP code point */
+                            ++source;
+                            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                            state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+                        } else {
+                            /* leave the optimized loop */
+                            break;
+                        }
+                    }
+                } while(source<sourceLimit && target<targetLimit);
+            } else /* offsets!=NULL */ {
+                do {
+                    entry=stateTable[state][*source];
+                    if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+                        state=(uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry);
+                        offset=MBCS_ENTRY_TRANSITION_OFFSET(entry);
+
+                        ++source;
+                        if( source<sourceLimit &&
+                            MBCS_ENTRY_IS_FINAL(entry=stateTable[state][*source]) &&
+                            MBCS_ENTRY_FINAL_ACTION(entry)==MBCS_STATE_VALID_16 &&
+                            (c=unicodeCodeUnits[offset+MBCS_ENTRY_FINAL_VALUE_16(entry)])<0xfffe
+                        ) {
+                            ++source;
+                            *target++=c;
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                sourceIndex=(nextSourceIndex+=2);
+                            }
+                            state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+                            offset=0;
+                        } else {
+                            /* set the state and leave the optimized loop */
+                            ++nextSourceIndex;
+                            bytes[0]=*(source-1);
+                            byteIndex=1;
+                            break;
+                        }
+                    } else {
+                        if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+                            /* output BMP code point */
+                            ++source;
+                            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                sourceIndex=++nextSourceIndex;
+                            }
+                            state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+                        } else {
+                            /* leave the optimized loop */
+                            break;
+                        }
+                    }
+                } while(source<sourceLimit && target<targetLimit);
+            }
+
+            /*
+             * these tests and break statements could be put inside the loop
+             * if C had "break outerLoop" like Java
+             */
+            if(source>=sourceLimit) {
+                break;
+            }
+            if(target>=targetLimit) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+
+            ++nextSourceIndex;
+            bytes[byteIndex++]=*source++;
+        } else /* byteIndex>0 */ {
+            ++nextSourceIndex;
+            entry=stateTable[state][bytes[byteIndex++]=*source++];
+        }
+
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            state=(uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry);
+            offset+=MBCS_ENTRY_TRANSITION_OFFSET(entry);
+            continue;
+        }
+
+        /* save the previous state for proper extension mapping with SI/SO-stateful converters */
+        cnv->mode=state;
+
+        /* set the next state early so that we can reuse the entry variable */
+        state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+
+        /*
+         * An if-else-if chain provides more reliable performance for
+         * the most common cases compared to a switch.
+         */
+        action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+        if(action==MBCS_STATE_VALID_16) {
+            offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+            c=unicodeCodeUnits[offset];
+            if(c<0xfffe) {
+                /* output BMP code point */
+                *target++=c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                byteIndex=0;
+            } else if(c==0xfffe) {
+                if(UCNV_TO_U_USE_FALLBACK(cnv) && (entry=(int32_t)ucnv_MBCSGetFallback(&cnv->sharedData->mbcs, offset))!=0xfffe) {
+                    /* output fallback BMP code point */
+                    *target++=(UChar)entry;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    byteIndex=0;
+                }
+            } else {
+                /* callback(illegal) */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            }
+        } else if(action==MBCS_STATE_VALID_DIRECT_16) {
+            /* output BMP code point */
+            *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            byteIndex=0;
+        } else if(action==MBCS_STATE_VALID_16_PAIR) {
+            offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+            c=unicodeCodeUnits[offset++];
+            if(c<0xd800) {
+                /* output BMP code point below 0xd800 */
+                *target++=c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                byteIndex=0;
+            } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? c<=0xdfff : c<=0xdbff) {
+                /* output roundtrip or fallback surrogate pair */
+                *target++=(UChar)(c&0xdbff);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                byteIndex=0;
+                if(target<targetLimit) {
+                    *target++=unicodeCodeUnits[offset];
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                } else {
+                    /* target overflow */
+                    cnv->UCharErrorBuffer[0]=unicodeCodeUnits[offset];
+                    cnv->UCharErrorBufferLength=1;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+
+                    offset=0;
+                    break;
+                }
+            } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? (c&0xfffe)==0xe000 : c==0xe000) {
+                /* output roundtrip BMP code point above 0xd800 or fallback BMP code point */
+                *target++=unicodeCodeUnits[offset];
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                byteIndex=0;
+            } else if(c==0xffff) {
+                /* callback(illegal) */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            }
+        } else if(action==MBCS_STATE_VALID_DIRECT_20 ||
+                  (action==MBCS_STATE_FALLBACK_DIRECT_20 && UCNV_TO_U_USE_FALLBACK(cnv))
+        ) {
+            entry=MBCS_ENTRY_FINAL_VALUE(entry);
+            /* output surrogate pair */
+            *target++=(UChar)(0xd800|(UChar)(entry>>10));
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            byteIndex=0;
+            c=(UChar)(0xdc00|(UChar)(entry&0x3ff));
+            if(target<targetLimit) {
+                *target++=c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+            } else {
+                /* target overflow */
+                cnv->UCharErrorBuffer[0]=c;
+                cnv->UCharErrorBufferLength=1;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+
+                offset=0;
+                break;
+            }
+        } else if(action==MBCS_STATE_CHANGE_ONLY) {
+            /*
+             * This serves as a state change without any output.
+             * It is useful for reading simple stateful encodings,
+             * for example using just Shift-In/Shift-Out codes.
+             * The 21 unused bits may later be used for more sophisticated
+             * state transitions.
+             */
+            if(cnv->sharedData->mbcs.dbcsOnlyState==0) {
+                byteIndex=0;
+            } else {
+                /* SI/SO are illegal for DBCS-only conversion */
+                state=(uint8_t)(cnv->mode); /* restore the previous state */
+
+                /* callback(illegal) */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            }
+        } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+            if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                /* output BMP code point */
+                *target++=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                byteIndex=0;
+            }
+        } else if(action==MBCS_STATE_UNASSIGNED) {
+            /* just fall through */
+        } else if(action==MBCS_STATE_ILLEGAL) {
+            /* callback(illegal) */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        } else {
+            /* reserved, must never occur */
+            byteIndex=0;
+        }
+
+        /* end of action codes: prepare for a new character */
+        offset=0;
+
+        if(byteIndex==0) {
+            sourceIndex=nextSourceIndex;
+        } else if(U_FAILURE(*pErrorCode)) {
+            /* callback(illegal) */
+            if(byteIndex>1) {
+                /*
+                 * Ticket 5691: consistent illegal sequences:
+                 * - We include at least the first byte in the illegal sequence.
+                 * - If any of the non-initial bytes could be the start of a character,
+                 *   we stop the illegal sequence before the first one of those.
+                 */
+                UBool isDBCSOnly=(UBool)(cnv->sharedData->mbcs.dbcsOnlyState!=0);
+                int8_t i;
+                for(i=1;
+                    i<byteIndex && !isSingleOrLead(stateTable, state, isDBCSOnly, bytes[i]);
+                    ++i) {}
+                if(i<byteIndex) {
+                    /* Back out some bytes. */
+                    int8_t backOutDistance=byteIndex-i;
+                    int32_t bytesFromThisBuffer=(int32_t)(source-(const uint8_t *)pArgs->source);
+                    byteIndex=i;  /* length of reported illegal byte sequence */
+                    if(backOutDistance<=bytesFromThisBuffer) {
+                        source-=backOutDistance;
+                    } else {
+                        /* Back out bytes from the previous buffer: Need to replay them. */
+                        cnv->preToULength=(int8_t)(bytesFromThisBuffer-backOutDistance);
+                        /* preToULength is negative! */
+                        uprv_memcpy(cnv->preToU, bytes+i, -cnv->preToULength);
+                        source=(const uint8_t *)pArgs->source;
+                    }
+                }
+            }
+            break;
+        } else /* unassigned sequences indicated with byteIndex>0 */ {
+            /* try an extension mapping */
+            pArgs->source=(const char *)source;
+            byteIndex=_extToU(cnv, cnv->sharedData,
+                              byteIndex, &source, sourceLimit,
+                              &target, targetLimit,
+                              &offsets, sourceIndex,
+                              pArgs->flush,
+                              pErrorCode);
+            sourceIndex=nextSourceIndex+=(int32_t)(source-(const uint8_t *)pArgs->source);
+
+            if(U_FAILURE(*pErrorCode)) {
+                /* not mappable or buffer overflow */
+                break;
+            }
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->toUnicodeStatus=offset;
+    cnv->mode=state;
+    cnv->toULength=byteIndex;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+}
+
+/*
+ * This version of ucnv_MBCSGetNextUChar() is optimized for single-byte, single-state codepages.
+ * We still need a conversion loop in case we find reserved action codes, which are to be ignored.
+ */
+static UChar32
+ucnv_MBCSSingleGetNextUChar(UConverterToUnicodeArgs *pArgs,
+                        UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const int32_t (*stateTable)[256];
+    const uint8_t *source, *sourceLimit;
+
+    int32_t entry;
+    uint8_t action;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        stateTable=(const int32_t (*)[256])cnv->sharedData->mbcs.swapLFNLStateTable;
+    } else {
+        stateTable=cnv->sharedData->mbcs.stateTable;
+    }
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        entry=stateTable[0][*source++];
+        /* MBCS_ENTRY_IS_FINAL(entry) */
+
+        /* write back the updated pointer early so that we can return directly */
+        pArgs->source=(const char *)source;
+
+        if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+            /* output BMP code point */
+            return (UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+        }
+
+        /*
+         * An if-else-if chain provides more reliable performance for
+         * the most common cases compared to a switch.
+         */
+        action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+        if( action==MBCS_STATE_VALID_DIRECT_20 ||
+            (action==MBCS_STATE_FALLBACK_DIRECT_20 && UCNV_TO_U_USE_FALLBACK(cnv))
+        ) {
+            /* output supplementary code point */
+            return (UChar32)(MBCS_ENTRY_FINAL_VALUE(entry)+0x10000);
+        } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+            if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                /* output BMP code point */
+                return (UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+            }
+        } else if(action==MBCS_STATE_UNASSIGNED) {
+            /* just fall through */
+        } else if(action==MBCS_STATE_ILLEGAL) {
+            /* callback(illegal) */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+        } else {
+            /* reserved, must never occur */
+            continue;
+        }
+
+        if(U_FAILURE(*pErrorCode)) {
+            /* callback(illegal) */
+            break;
+        } else /* unassigned sequence */ {
+            /* defer to the generic implementation */
+            pArgs->source=(const char *)source-1;
+            return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+        }
+    }
+
+    /* no output because of empty input or only state changes */
+    *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    return 0xffff;
+}
+
+/*
+ * Version of _MBCSToUnicodeWithOffsets() optimized for single-character
+ * conversion without offset handling.
+ *
+ * When a character does not have a mapping to Unicode, then we return to the
+ * generic ucnv_getNextUChar() code for extension/GB 18030 and error/callback
+ * handling.
+ * We also defer to the generic code in other complicated cases and have them
+ * ultimately handled by _MBCSToUnicodeWithOffsets() itself.
+ *
+ * All normal mappings and errors are handled here.
+ */
+static UChar32 U_CALLCONV
+ucnv_MBCSGetNextUChar(UConverterToUnicodeArgs *pArgs,
+                  UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const uint8_t *source, *sourceLimit, *lastSource;
+
+    const int32_t (*stateTable)[256];
+    const uint16_t *unicodeCodeUnits;
+
+    uint32_t offset;
+    uint8_t state;
+
+    int32_t entry;
+    UChar32 c;
+    uint8_t action;
+
+    /* use optimized function if possible */
+    cnv=pArgs->converter;
+
+    if(cnv->preToULength>0) {
+        /* use the generic code in ucnv_getNextUChar() to continue with a partial match */
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    }
+
+    if(cnv->sharedData->mbcs.unicodeMask&UCNV_HAS_SURROGATES) {
+        /*
+         * Using the generic ucnv_getNextUChar() code lets us deal correctly
+         * with the rare case of a codepage that maps single surrogates
+         * without adding the complexity to this already complicated function here.
+         */
+        return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+    } else if(cnv->sharedData->mbcs.countStates==1) {
+        return ucnv_MBCSSingleGetNextUChar(pArgs, pErrorCode);
+    }
+
+    /* set up the local pointers */
+    source=lastSource=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        stateTable=(const int32_t (*)[256])cnv->sharedData->mbcs.swapLFNLStateTable;
+    } else {
+        stateTable=cnv->sharedData->mbcs.stateTable;
+    }
+    unicodeCodeUnits=cnv->sharedData->mbcs.unicodeCodeUnits;
+
+    /* get the converter state from UConverter */
+    offset=cnv->toUnicodeStatus;
+
+    /*
+     * if we are in the SBCS state for a DBCS-only converter,
+     * then load the DBCS state from the MBCS data
+     * (dbcsOnlyState==0 if it is not a DBCS-only converter)
+     */
+    if((state=(uint8_t)(cnv->mode))==0) {
+        state=cnv->sharedData->mbcs.dbcsOnlyState;
+    }
+
+    /* conversion loop */
+    c=U_SENTINEL;
+    while(source<sourceLimit) {
+        entry=stateTable[state][*source++];
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            state=(uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry);
+            offset+=MBCS_ENTRY_TRANSITION_OFFSET(entry);
+
+            /* optimization for 1/2-byte input and BMP output */
+            if( source<sourceLimit &&
+                MBCS_ENTRY_IS_FINAL(entry=stateTable[state][*source]) &&
+                MBCS_ENTRY_FINAL_ACTION(entry)==MBCS_STATE_VALID_16 &&
+                (c=unicodeCodeUnits[offset+MBCS_ENTRY_FINAL_VALUE_16(entry)])<0xfffe
+            ) {
+                ++source;
+                state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+                /* output BMP code point */
+                break;
+            }
+        } else {
+            /* save the previous state for proper extension mapping with SI/SO-stateful converters */
+            cnv->mode=state;
+
+            /* set the next state early so that we can reuse the entry variable */
+            state=(uint8_t)MBCS_ENTRY_FINAL_STATE(entry); /* typically 0 */
+
+            /*
+             * An if-else-if chain provides more reliable performance for
+             * the most common cases compared to a switch.
+             */
+            action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+            if(action==MBCS_STATE_VALID_DIRECT_16) {
+                /* output BMP code point */
+                c=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                break;
+            } else if(action==MBCS_STATE_VALID_16) {
+                offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[offset];
+                if(c<0xfffe) {
+                    /* output BMP code point */
+                    break;
+                } else if(c==0xfffe) {
+                    if(UCNV_TO_U_USE_FALLBACK(cnv) && (c=ucnv_MBCSGetFallback(&cnv->sharedData->mbcs, offset))!=0xfffe) {
+                        break;
+                    }
+                } else {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else if(action==MBCS_STATE_VALID_16_PAIR) {
+                offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[offset++];
+                if(c<0xd800) {
+                    /* output BMP code point below 0xd800 */
+                    break;
+                } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? c<=0xdfff : c<=0xdbff) {
+                    /* output roundtrip or fallback supplementary code point */
+                    c=((c&0x3ff)<<10)+unicodeCodeUnits[offset]+(0x10000-0xdc00);
+                    break;
+                } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? (c&0xfffe)==0xe000 : c==0xe000) {
+                    /* output roundtrip BMP code point above 0xd800 or fallback BMP code point */
+                    c=unicodeCodeUnits[offset];
+                    break;
+                } else if(c==0xffff) {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else if(action==MBCS_STATE_VALID_DIRECT_20 ||
+                      (action==MBCS_STATE_FALLBACK_DIRECT_20 && UCNV_TO_U_USE_FALLBACK(cnv))
+            ) {
+                /* output supplementary code point */
+                c=(UChar32)(MBCS_ENTRY_FINAL_VALUE(entry)+0x10000);
+                break;
+            } else if(action==MBCS_STATE_CHANGE_ONLY) {
+                /*
+                 * This serves as a state change without any output.
+                 * It is useful for reading simple stateful encodings,
+                 * for example using just Shift-In/Shift-Out codes.
+                 * The 21 unused bits may later be used for more sophisticated
+                 * state transitions.
+                 */
+                if(cnv->sharedData->mbcs.dbcsOnlyState!=0) {
+                    /* SI/SO are illegal for DBCS-only conversion */
+                    state=(uint8_t)(cnv->mode); /* restore the previous state */
+
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                }
+            } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+                if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                    /* output BMP code point */
+                    c=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                    break;
+                }
+            } else if(action==MBCS_STATE_UNASSIGNED) {
+                /* just fall through */
+            } else if(action==MBCS_STATE_ILLEGAL) {
+                /* callback(illegal) */
+                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            } else {
+                /* reserved (must never occur), or only state change */
+                offset=0;
+                lastSource=source;
+                continue;
+            }
+
+            /* end of action codes: prepare for a new character */
+            offset=0;
+
+            if(U_FAILURE(*pErrorCode)) {
+                /* callback(illegal) */
+                break;
+            } else /* unassigned sequence */ {
+                /* defer to the generic implementation */
+                cnv->toUnicodeStatus=0;
+                cnv->mode=state;
+                pArgs->source=(const char *)lastSource;
+                return UCNV_GET_NEXT_UCHAR_USE_TO_U;
+            }
+        }
+    }
+
+    if(c<0) {
+        if(U_SUCCESS(*pErrorCode) && source==sourceLimit && lastSource<source) {
+            /* incomplete character byte sequence */
+            uint8_t *bytes=cnv->toUBytes;
+            cnv->toULength=(int8_t)(source-lastSource);
+            do {
+                *bytes++=*lastSource++;
+            } while(lastSource<source);
+            *pErrorCode=U_TRUNCATED_CHAR_FOUND;
+        } else if(U_FAILURE(*pErrorCode)) {
+            /* callback(illegal) */
+            /*
+             * Ticket 5691: consistent illegal sequences:
+             * - We include at least the first byte in the illegal sequence.
+             * - If any of the non-initial bytes could be the start of a character,
+             *   we stop the illegal sequence before the first one of those.
+             */
+            UBool isDBCSOnly=(UBool)(cnv->sharedData->mbcs.dbcsOnlyState!=0);
+            uint8_t *bytes=cnv->toUBytes;
+            *bytes++=*lastSource++;     /* first byte */
+            if(lastSource==source) {
+                cnv->toULength=1;
+            } else /* lastSource<source: multi-byte character */ {
+                int8_t i;
+                for(i=1;
+                    lastSource<source && !isSingleOrLead(stateTable, state, isDBCSOnly, *lastSource);
+                    ++i
+                ) {
+                    *bytes++=*lastSource++;
+                }
+                cnv->toULength=i;
+                source=lastSource;
+            }
+        } else {
+            /* no output because of empty input or only state changes */
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        }
+        c=0xffff;
+    }
+
+    /* set the converter state back into UConverter, ready for a new character */
+    cnv->toUnicodeStatus=0;
+    cnv->mode=state;
+
+    /* write back the updated pointer */
+    pArgs->source=(const char *)source;
+    return c;
+}
+
+#if 0
+/*
+ * Code disabled 2002dec09 (ICU 2.4) because it is not currently used in ICU. markus
+ * Removal improves code coverage.
+ */
+/**
+ * This version of ucnv_MBCSSimpleGetNextUChar() is optimized for single-byte, single-state codepages.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It does not handle conversion extensions (_extToU()).
+ */
+U_CFUNC UChar32
+ucnv_MBCSSingleSimpleGetNextUChar(UConverterSharedData *sharedData,
+                              uint8_t b, UBool useFallback) {
+    int32_t entry;
+    uint8_t action;
+
+    entry=sharedData->mbcs.stateTable[0][b];
+    /* MBCS_ENTRY_IS_FINAL(entry) */
+
+    if(MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry)) {
+        /* output BMP code point */
+        return (UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+    }
+
+    /*
+     * An if-else-if chain provides more reliable performance for
+     * the most common cases compared to a switch.
+     */
+    action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+    if(action==MBCS_STATE_VALID_DIRECT_20) {
+        /* output supplementary code point */
+        return 0x10000+MBCS_ENTRY_FINAL_VALUE(entry);
+    } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+        if(!TO_U_USE_FALLBACK(useFallback)) {
+            return 0xfffe;
+        }
+        /* output BMP code point */
+        return (UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+    } else if(action==MBCS_STATE_FALLBACK_DIRECT_20) {
+        if(!TO_U_USE_FALLBACK(useFallback)) {
+            return 0xfffe;
+        }
+        /* output supplementary code point */
+        return 0x10000+MBCS_ENTRY_FINAL_VALUE(entry);
+    } else if(action==MBCS_STATE_UNASSIGNED) {
+        return 0xfffe;
+    } else if(action==MBCS_STATE_ILLEGAL) {
+        return 0xffff;
+    } else {
+        /* reserved, must never occur */
+        return 0xffff;
+    }
+}
+#endif
+
+/*
+ * This is a simple version of _MBCSGetNextUChar() that is used
+ * by other converter implementations.
+ * It only returns an "assigned" result if it consumes the entire input.
+ * It does not use state from the converter, nor error codes.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It handles conversion extensions but not GB 18030.
+ *
+ * Return value:
+ * U+fffe   unassigned
+ * U+ffff   illegal
+ * otherwise the Unicode code point
+ */
+U_CFUNC UChar32
+ucnv_MBCSSimpleGetNextUChar(UConverterSharedData *sharedData,
+                        const char *source, int32_t length,
+                        UBool useFallback) {
+    const int32_t (*stateTable)[256];
+    const uint16_t *unicodeCodeUnits;
+
+    uint32_t offset;
+    uint8_t state, action;
+
+    UChar32 c;
+    int32_t i, entry;
+
+    if(length<=0) {
+        /* no input at all: "illegal" */
+        return 0xffff;
+    }
+
+#if 0
+/*
+ * Code disabled 2002dec09 (ICU 2.4) because it is not currently used in ICU. markus
+ * TODO In future releases, verify that this function is never called for SBCS
+ * conversions, i.e., that sharedData->mbcs.countStates==1 is still true.
+ * Removal improves code coverage.
+ */
+    /* use optimized function if possible */
+    if(sharedData->mbcs.countStates==1) {
+        if(length==1) {
+            return ucnv_MBCSSingleSimpleGetNextUChar(sharedData, (uint8_t)*source, useFallback);
+        } else {
+            return 0xffff; /* illegal: more than a single byte for an SBCS converter */
+        }
+    }
+#endif
+
+    /* set up the local pointers */
+    stateTable=sharedData->mbcs.stateTable;
+    unicodeCodeUnits=sharedData->mbcs.unicodeCodeUnits;
+
+    /* converter state */
+    offset=0;
+    state=sharedData->mbcs.dbcsOnlyState;
+
+    /* conversion loop */
+    for(i=0;;) {
+        entry=stateTable[state][(uint8_t)source[i++]];
+        if(MBCS_ENTRY_IS_TRANSITION(entry)) {
+            state=(uint8_t)MBCS_ENTRY_TRANSITION_STATE(entry);
+            offset+=MBCS_ENTRY_TRANSITION_OFFSET(entry);
+
+            if(i==length) {
+                return 0xffff; /* truncated character */
+            }
+        } else {
+            /*
+             * An if-else-if chain provides more reliable performance for
+             * the most common cases compared to a switch.
+             */
+            action=(uint8_t)(MBCS_ENTRY_FINAL_ACTION(entry));
+            if(action==MBCS_STATE_VALID_16) {
+                offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[offset];
+                if(c!=0xfffe) {
+                    /* done */
+                } else if(UCNV_TO_U_USE_FALLBACK(cnv)) {
+                    c=ucnv_MBCSGetFallback(&sharedData->mbcs, offset);
+                /* else done with 0xfffe */
+                }
+                break;
+            } else if(action==MBCS_STATE_VALID_DIRECT_16) {
+                /* output BMP code point */
+                c=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                break;
+            } else if(action==MBCS_STATE_VALID_16_PAIR) {
+                offset+=MBCS_ENTRY_FINAL_VALUE_16(entry);
+                c=unicodeCodeUnits[offset++];
+                if(c<0xd800) {
+                    /* output BMP code point below 0xd800 */
+                } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? c<=0xdfff : c<=0xdbff) {
+                    /* output roundtrip or fallback supplementary code point */
+                    c=(UChar32)(((c&0x3ff)<<10)+unicodeCodeUnits[offset]+(0x10000-0xdc00));
+                } else if(UCNV_TO_U_USE_FALLBACK(cnv) ? (c&0xfffe)==0xe000 : c==0xe000) {
+                    /* output roundtrip BMP code point above 0xd800 or fallback BMP code point */
+                    c=unicodeCodeUnits[offset];
+                } else if(c==0xffff) {
+                    return 0xffff;
+                } else {
+                    c=0xfffe;
+                }
+                break;
+            } else if(action==MBCS_STATE_VALID_DIRECT_20) {
+                /* output supplementary code point */
+                c=0x10000+MBCS_ENTRY_FINAL_VALUE(entry);
+                break;
+            } else if(action==MBCS_STATE_FALLBACK_DIRECT_16) {
+                if(!TO_U_USE_FALLBACK(useFallback)) {
+                    c=0xfffe;
+                    break;
+                }
+                /* output BMP code point */
+                c=(UChar)MBCS_ENTRY_FINAL_VALUE_16(entry);
+                break;
+            } else if(action==MBCS_STATE_FALLBACK_DIRECT_20) {
+                if(!TO_U_USE_FALLBACK(useFallback)) {
+                    c=0xfffe;
+                    break;
+                }
+                /* output supplementary code point */
+                c=0x10000+MBCS_ENTRY_FINAL_VALUE(entry);
+                break;
+            } else if(action==MBCS_STATE_UNASSIGNED) {
+                c=0xfffe;
+                break;
+            }
+
+            /*
+             * forbid MBCS_STATE_CHANGE_ONLY for this function,
+             * and MBCS_STATE_ILLEGAL and reserved action codes
+             */
+            return 0xffff;
+        }
+    }
+
+    if(i!=length) {
+        /* illegal for this function: not all input consumed */
+        return 0xffff;
+    }
+
+    if(c==0xfffe) {
+        /* try an extension mapping */
+        const int32_t *cx=sharedData->mbcs.extIndexes;
+        if(cx!=NULL) {
+            return ucnv_extSimpleMatchToU(cx, source, length, useFallback);
+        }
+    }
+
+    return c;
+}
+
+/* MBCS-from-Unicode conversion functions ----------------------------------- */
+
+/* This version of ucnv_MBCSFromUnicodeWithOffsets() is optimized for double-byte codepages. */
+static void
+ucnv_MBCSDoubleFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                                  UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t *offsets;
+
+    const uint16_t *table;
+    const uint16_t *mbcsIndex;
+    const uint8_t *bytes;
+
+    UChar32 c;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    uint32_t stage2Entry;
+    uint32_t asciiRoundtrips;
+    uint32_t value;
+    uint8_t unicodeMask;
+
+    /* use optimized function if possible */
+    cnv=pArgs->converter;
+    unicodeMask=cnv->sharedData->mbcs.unicodeMask;
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    mbcsIndex=cnv->sharedData->mbcs.mbcsIndex;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        bytes=cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        bytes=cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+    asciiRoundtrips=cnv->sharedData->mbcs.asciiRoundtrips;
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= c==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* conversion loop */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+    while(source<sourceLimit) {
+        /*
+         * This following test is to see if available input would overflow the output.
+         * It does not catch output of more than one byte that
+         * overflows as a result of a multi-byte character or callback output
+         * from the last source character.
+         * Therefore, those situations also test for overflows and will
+         * then break the loop, too.
+         */
+        if(targetCapacity>0) {
+            /*
+             * Get a correct Unicode code point:
+             * a single UChar for a BMP code point or
+             * a matched surrogate pair for a "supplementary code point".
+             */
+            c=*source++;
+            ++nextSourceIndex;
+            if(c<=0x7f && IS_ASCII_ROUNDTRIP(c, asciiRoundtrips)) {
+                *target++=(uint8_t)c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                    sourceIndex=nextSourceIndex;
+                }
+                --targetCapacity;
+                c=0;
+                continue;
+            }
+            /*
+             * utf8Friendly table: Test for <=0xd7ff rather than <=MBCS_FAST_MAX
+             * to avoid dealing with surrogates.
+             * MBCS_FAST_MAX must be >=0xd7ff.
+             */
+            if(c<=0xd7ff) {
+                value=DBCS_RESULT_FROM_MOST_BMP(mbcsIndex, (const uint16_t *)bytes, c);
+                /* There are only roundtrips (!=0) and no-mapping (==0) entries. */
+                if(value==0) {
+                    goto unassigned;
+                }
+                /* output the value */
+            } else {
+                /*
+                 * This also tests if the codepage maps single surrogates.
+                 * If it does, then surrogates are not paired but mapped separately.
+                 * Note that in this case unmatched surrogates are not detected.
+                 */
+                if(U16_IS_SURROGATE(c) && !(unicodeMask&UCNV_HAS_SURROGATES)) {
+                    if(U16_IS_SURROGATE_LEAD(c)) {
+getTrail:
+                        if(source<sourceLimit) {
+                            /* test the following code unit */
+                            UChar trail=*source;
+                            if(U16_IS_TRAIL(trail)) {
+                                ++source;
+                                ++nextSourceIndex;
+                                c=U16_GET_SUPPLEMENTARY(c, trail);
+                                if(!(unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+                                    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+                                    /* callback(unassigned) */
+                                    goto unassigned;
+                                }
+                                /* convert this supplementary code point */
+                                /* exit this condition tree */
+                            } else {
+                                /* this is an unmatched lead code unit (1st surrogate) */
+                                /* callback(illegal) */
+                                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                                break;
+                            }
+                        } else {
+                            /* no more input */
+                            break;
+                        }
+                    } else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+
+                /* convert the Unicode code point in c into codepage bytes */
+                stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+
+                /* get the bytes and the length for the output */
+                /* MBCS_OUTPUT_2 */
+                value=MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c);
+
+                /* is this code point assigned, or do we use fallbacks? */
+                if(!(MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c) ||
+                     (UCNV_FROM_U_USE_FALLBACK(cnv, c) && value!=0))
+                ) {
+                    /*
+                     * We allow a 0 byte output if the "assigned" bit is set for this entry.
+                     * There is no way with this data structure for fallback output
+                     * to be a zero byte.
+                     */
+
+unassigned:
+                    /* try an extension mapping */
+                    pArgs->source=source;
+                    c=_extFromU(cnv, cnv->sharedData,
+                                c, &source, sourceLimit,
+                                &target, target+targetCapacity,
+                                &offsets, sourceIndex,
+                                pArgs->flush,
+                                pErrorCode);
+                    nextSourceIndex+=(int32_t)(source-pArgs->source);
+
+                    if(U_FAILURE(*pErrorCode)) {
+                        /* not mappable or buffer overflow */
+                        break;
+                    } else {
+                        /* a mapping was written to the target, continue */
+
+                        /* recalculate the targetCapacity after an extension mapping */
+                        targetCapacity=(int32_t)(pArgs->targetLimit-(char *)target);
+
+                        /* normal end of conversion: prepare for a new character */
+                        sourceIndex=nextSourceIndex;
+                        continue;
+                    }
+                }
+            }
+
+            /* write the output character bytes from value and length */
+            /* from the first if in the loop we know that targetCapacity>0 */
+            if(value<=0xff) {
+                /* this is easy because we know that there is enough space */
+                *target++=(uint8_t)value;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                --targetCapacity;
+            } else /* length==2 */ {
+                *target++=(uint8_t)(value>>8);
+                if(2<=targetCapacity) {
+                    *target++=(uint8_t)value;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                        *offsets++=sourceIndex;
+                    }
+                    targetCapacity-=2;
+                } else {
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    cnv->charErrorBuffer[0]=(char)value;
+                    cnv->charErrorBufferLength=1;
+
+                    /* target overflow */
+                    targetCapacity=0;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    c=0;
+                    break;
+                }
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+            sourceIndex=nextSourceIndex;
+            continue;
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32=c;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+/* This version of ucnv_MBCSFromUnicodeWithOffsets() is optimized for single-byte codepages. */
+static void
+ucnv_MBCSSingleFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                                  UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t *offsets;
+
+    const uint16_t *table;
+    const uint16_t *results;
+
+    UChar32 c;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    uint16_t value, minValue;
+    UBool hasSupplementary;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        results=(uint16_t *)cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        results=(uint16_t *)cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+
+    if(cnv->useFallback) {
+        /* use all roundtrip and fallback results */
+        minValue=0x800;
+    } else {
+        /* use only roundtrips and fallbacks from private-use characters */
+        minValue=0xc00;
+    }
+    hasSupplementary=(UBool)(cnv->sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY);
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= c==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* conversion loop */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+    while(source<sourceLimit) {
+        /*
+         * This following test is to see if available input would overflow the output.
+         * It does not catch output of more than one byte that
+         * overflows as a result of a multi-byte character or callback output
+         * from the last source character.
+         * Therefore, those situations also test for overflows and will
+         * then break the loop, too.
+         */
+        if(targetCapacity>0) {
+            /*
+             * Get a correct Unicode code point:
+             * a single UChar for a BMP code point or
+             * a matched surrogate pair for a "supplementary code point".
+             */
+            c=*source++;
+            ++nextSourceIndex;
+            if(U16_IS_SURROGATE(c)) {
+                if(U16_IS_SURROGATE_LEAD(c)) {
+getTrail:
+                    if(source<sourceLimit) {
+                        /* test the following code unit */
+                        UChar trail=*source;
+                        if(U16_IS_TRAIL(trail)) {
+                            ++source;
+                            ++nextSourceIndex;
+                            c=U16_GET_SUPPLEMENTARY(c, trail);
+                            if(!hasSupplementary) {
+                                /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+                                /* callback(unassigned) */
+                                goto unassigned;
+                            }
+                            /* convert this supplementary code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            break;
+                        }
+                    } else {
+                        /* no more input */
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    break;
+                }
+            }
+
+            /* convert the Unicode code point in c into codepage bytes */
+            value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+
+            /* is this code point assigned, or do we use fallbacks? */
+            if(value>=minValue) {
+                /* assigned, write the output character bytes from value and length */
+                /* length==1 */
+                /* this is easy because we know that there is enough space */
+                *target++=(uint8_t)value;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                --targetCapacity;
+
+                /* normal end of conversion: prepare for a new character */
+                c=0;
+                sourceIndex=nextSourceIndex;
+            } else { /* unassigned */
+unassigned:
+                /* try an extension mapping */
+                pArgs->source=source;
+                c=_extFromU(cnv, cnv->sharedData,
+                            c, &source, sourceLimit,
+                            &target, target+targetCapacity,
+                            &offsets, sourceIndex,
+                            pArgs->flush,
+                            pErrorCode);
+                nextSourceIndex+=(int32_t)(source-pArgs->source);
+
+                if(U_FAILURE(*pErrorCode)) {
+                    /* not mappable or buffer overflow */
+                    break;
+                } else {
+                    /* a mapping was written to the target, continue */
+
+                    /* recalculate the targetCapacity after an extension mapping */
+                    targetCapacity=(int32_t)(pArgs->targetLimit-(char *)target);
+
+                    /* normal end of conversion: prepare for a new character */
+                    sourceIndex=nextSourceIndex;
+                }
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32=c;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+/*
+ * This version of ucnv_MBCSFromUnicode() is optimized for single-byte codepages
+ * that map only to and from the BMP.
+ * In addition to single-byte/state optimizations, the offset calculations
+ * become much easier.
+ * It would be possible to use the sbcsIndex for UTF-8-friendly tables,
+ * but measurements have shown that this diminishes performance
+ * in more cases than it improves it.
+ * See SVN revision 21013 (2007-feb-06) for the last version with #if switches
+ * for various MBCS and SBCS optimizations.
+ */
+static void
+ucnv_MBCSSingleFromBMPWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                              UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit, *lastSource;
+    uint8_t *target;
+    int32_t targetCapacity, length;
+    int32_t *offsets;
+
+    const uint16_t *table;
+    const uint16_t *results;
+
+    UChar32 c;
+
+    int32_t sourceIndex;
+
+    uint32_t asciiRoundtrips;
+    uint16_t value, minValue;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        results=(uint16_t *)cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        results=(uint16_t *)cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+    asciiRoundtrips=cnv->sharedData->mbcs.asciiRoundtrips;
+
+    if(cnv->useFallback) {
+        /* use all roundtrip and fallback results */
+        minValue=0x800;
+    } else {
+        /* use only roundtrips and fallbacks from private-use characters */
+        minValue=0xc00;
+    }
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= c==0 ? 0 : -1;
+    lastSource=source;
+
+    /*
+     * since the conversion here is 1:1 UChar:uint8_t, we need only one counter
+     * for the minimum of the sourceLength and targetCapacity
+     */
+    length=(int32_t)(sourceLimit-source);
+    if(length<targetCapacity) {
+        targetCapacity=length;
+    }
+
+    /* conversion loop */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+#if MBCS_UNROLL_SINGLE_FROM_BMP
+    /* unrolling makes it slower on Pentium III/Windows 2000?! */
+    /* unroll the loop with the most common case */
+unrolled:
+    if(targetCapacity>=4) {
+        int32_t count, loops;
+        uint16_t andedValues;
+
+        loops=count=targetCapacity>>2;
+        do {
+            c=*source++;
+            andedValues=value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+            *target++=(uint8_t)value;
+            c=*source++;
+            andedValues&=value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+            *target++=(uint8_t)value;
+            c=*source++;
+            andedValues&=value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+            *target++=(uint8_t)value;
+            c=*source++;
+            andedValues&=value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+            *target++=(uint8_t)value;
+
+            /* were all 4 entries really valid? */
+            if(andedValues<minValue) {
+                /* no, return to the first of these 4 */
+                source-=4;
+                target-=4;
+                break;
+            }
+        } while(--count>0);
+        count=loops-count;
+        targetCapacity-=4*count;
+
+        if(offsets!=NULL) {
+            lastSource+=4*count;
+            while(count>0) {
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                *offsets++=sourceIndex++;
+                --count;
+            }
+        }
+
+        c=0;
+    }
+#endif
+
+    while(targetCapacity>0) {
+        /*
+         * Get a correct Unicode code point:
+         * a single UChar for a BMP code point or
+         * a matched surrogate pair for a "supplementary code point".
+         */
+        c=*source++;
+        /*
+         * Do not immediately check for single surrogates:
+         * Assume that they are unassigned and check for them in that case.
+         * This speeds up the conversion of assigned characters.
+         */
+        /* convert the Unicode code point in c into codepage bytes */
+        if(c<=0x7f && IS_ASCII_ROUNDTRIP(c, asciiRoundtrips)) {
+            *target++=(uint8_t)c;
+            --targetCapacity;
+            c=0;
+            continue;
+        }
+        value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+        /* is this code point assigned, or do we use fallbacks? */
+        if(value>=minValue) {
+            /* assigned, write the output character bytes from value and length */
+            /* length==1 */
+            /* this is easy because we know that there is enough space */
+            *target++=(uint8_t)value;
+            --targetCapacity;
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+            continue;
+        } else if(!U16_IS_SURROGATE(c)) {
+            /* normal, unassigned BMP character */
+        } else if(U16_IS_SURROGATE_LEAD(c)) {
+getTrail:
+            if(source<sourceLimit) {
+                /* test the following code unit */
+                UChar trail=*source;
+                if(U16_IS_TRAIL(trail)) {
+                    ++source;
+                    c=U16_GET_SUPPLEMENTARY(c, trail);
+                    /* this codepage does not map supplementary code points */
+                    /* callback(unassigned) */
+                } else {
+                    /* this is an unmatched lead code unit (1st surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    break;
+                }
+            } else {
+                /* no more input */
+                if (pArgs->flush) {
+                    *pErrorCode=U_TRUNCATED_CHAR_FOUND;
+                }
+                break;
+            }
+        } else {
+            /* this is an unmatched trail code unit (2nd surrogate) */
+            /* callback(illegal) */
+            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+            break;
+        }
+
+        /* c does not have a mapping */
+
+        /* get the number of code units for c to correctly advance sourceIndex */
+        length=U16_LENGTH(c);
+
+        /* set offsets since the start or the last extension */
+        if(offsets!=NULL) {
+            int32_t count=(int32_t)(source-lastSource);
+
+            /* do not set the offset for this character */
+            count-=length;
+
+            while(count>0) {
+                *offsets++=sourceIndex++;
+                --count;
+            }
+            /* offsets and sourceIndex are now set for the current character */
+        }
+
+        /* try an extension mapping */
+        lastSource=source;
+        c=_extFromU(cnv, cnv->sharedData,
+                    c, &source, sourceLimit,
+                    &target, (const uint8_t *)(pArgs->targetLimit),
+                    &offsets, sourceIndex,
+                    pArgs->flush,
+                    pErrorCode);
+        sourceIndex+=length+(int32_t)(source-lastSource);
+        lastSource=source;
+
+        if(U_FAILURE(*pErrorCode)) {
+            /* not mappable or buffer overflow */
+            break;
+        } else {
+            /* a mapping was written to the target, continue */
+
+            /* recalculate the targetCapacity after an extension mapping */
+            targetCapacity=(int32_t)(pArgs->targetLimit-(char *)target);
+            length=(int32_t)(sourceLimit-source);
+            if(length<targetCapacity) {
+                targetCapacity=length;
+            }
+        }
+
+#if MBCS_UNROLL_SINGLE_FROM_BMP
+        /* unrolling makes it slower on Pentium III/Windows 2000?! */
+        goto unrolled;
+#endif
+    }
+
+    if(U_SUCCESS(*pErrorCode) && source<sourceLimit && target>=(uint8_t *)pArgs->targetLimit) {
+        /* target is full */
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+
+    /* set offsets since the start or the last callback */
+    if(offsets!=NULL) {
+        size_t count=source-lastSource;
+        if (count > 0 && *pErrorCode == U_TRUNCATED_CHAR_FOUND) {
+            /*
+            Caller gave us a partial supplementary character,
+            which this function couldn't convert in any case.
+            The callback will handle the offset.
+            */
+            count--;
+        }
+        while(count>0) {
+            *offsets++=sourceIndex++;
+            --count;
+        }
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32=c;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+U_CFUNC void
+ucnv_MBCSFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t *offsets;
+
+    const uint16_t *table;
+    const uint16_t *mbcsIndex;
+    const uint8_t *p, *bytes;
+    uint8_t outputType;
+
+    UChar32 c;
+
+    int32_t prevSourceIndex, sourceIndex, nextSourceIndex;
+
+    uint32_t stage2Entry;
+    uint32_t asciiRoundtrips;
+    uint32_t value;
+    /* Shift-In and Shift-Out byte sequences differ by encoding scheme. */
+    uint8_t siBytes[2] = {0, 0};
+    uint8_t soBytes[2] = {0, 0};
+    uint8_t siLength, soLength;
+    int32_t length = 0, prevLength;
+    uint8_t unicodeMask;
+
+    cnv=pArgs->converter;
+
+    if(cnv->preFromUFirstCP>=0) {
+        /*
+         * pass sourceIndex=-1 because we continue from an earlier buffer
+         * in the future, this may change with continuous offsets
+         */
+        ucnv_extContinueMatchFromU(cnv, pArgs, -1, pErrorCode);
+
+        if(U_FAILURE(*pErrorCode) || cnv->preFromULength<0) {
+            return;
+        }
+    }
+
+    /* use optimized function if possible */
+    outputType=cnv->sharedData->mbcs.outputType;
+    unicodeMask=cnv->sharedData->mbcs.unicodeMask;
+    if(outputType==MBCS_OUTPUT_1 && !(unicodeMask&UCNV_HAS_SURROGATES)) {
+        if(!(unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+            ucnv_MBCSSingleFromBMPWithOffsets(pArgs, pErrorCode);
+        } else {
+            ucnv_MBCSSingleFromUnicodeWithOffsets(pArgs, pErrorCode);
+        }
+        return;
+    } else if(outputType==MBCS_OUTPUT_2 && cnv->sharedData->mbcs.utf8Friendly) {
+        ucnv_MBCSDoubleFromUnicodeWithOffsets(pArgs, pErrorCode);
+        return;
+    }
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    if(cnv->sharedData->mbcs.utf8Friendly) {
+        mbcsIndex=cnv->sharedData->mbcs.mbcsIndex;
+    } else {
+        mbcsIndex=NULL;
+    }
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        bytes=cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        bytes=cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+    asciiRoundtrips=cnv->sharedData->mbcs.asciiRoundtrips;
+
+    /* get the converter state from UConverter */
+    c=cnv->fromUChar32;
+
+    if(outputType==MBCS_OUTPUT_2_SISO) {
+        prevLength=cnv->fromUnicodeStatus;
+        if(prevLength==0) {
+            /* set the real value */
+            prevLength=1;
+        }
+    } else {
+        /* prevent fromUnicodeStatus from being set to something non-0 */
+        prevLength=0;
+    }
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    prevSourceIndex=-1;
+    sourceIndex= c==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* Get the SI/SO character for the converter */
+    siLength = static_cast<uint8_t>(getSISOBytes(SI, cnv->options, siBytes));
+    soLength = static_cast<uint8_t>(getSISOBytes(SO, cnv->options, soBytes));
+
+    /* conversion loop */
+    /*
+     * This is another piece of ugly code:
+     * A goto into the loop if the converter state contains a first surrogate
+     * from the previous function call.
+     * It saves me to check in each loop iteration a check of if(c==0)
+     * and duplicating the trail-surrogate-handling code in the else
+     * branch of that check.
+     * I could not find any other way to get around this other than
+     * using a function call for the conversion and callback, which would
+     * be even more inefficient.
+     *
+     * Markus Scherer 2000-jul-19
+     */
+    if(c!=0 && targetCapacity>0) {
+        goto getTrail;
+    }
+
+    while(source<sourceLimit) {
+        /*
+         * This following test is to see if available input would overflow the output.
+         * It does not catch output of more than one byte that
+         * overflows as a result of a multi-byte character or callback output
+         * from the last source character.
+         * Therefore, those situations also test for overflows and will
+         * then break the loop, too.
+         */
+        if(targetCapacity>0) {
+            /*
+             * Get a correct Unicode code point:
+             * a single UChar for a BMP code point or
+             * a matched surrogate pair for a "supplementary code point".
+             */
+            c=*source++;
+            ++nextSourceIndex;
+            if(c<=0x7f && IS_ASCII_ROUNDTRIP(c, asciiRoundtrips)) {
+                *target++=(uint8_t)c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                    prevSourceIndex=sourceIndex;
+                    sourceIndex=nextSourceIndex;
+                }
+                --targetCapacity;
+                c=0;
+                continue;
+            }
+            /*
+             * utf8Friendly table: Test for <=0xd7ff rather than <=MBCS_FAST_MAX
+             * to avoid dealing with surrogates.
+             * MBCS_FAST_MAX must be >=0xd7ff.
+             */
+            if(c<=0xd7ff && mbcsIndex!=NULL) {
+                value=mbcsIndex[c>>6];
+
+                /* get the bytes and the length for the output (copied from below and adapted for utf8Friendly data) */
+                /* There are only roundtrips (!=0) and no-mapping (==0) entries. */
+                switch(outputType) {
+                case MBCS_OUTPUT_2:
+                    value=((const uint16_t *)bytes)[value +(c&0x3f)];
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else {
+                            length=1;
+                        }
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_2_SISO:
+                    /* 1/2-byte stateful with Shift-In/Shift-Out */
+                    /*
+                     * Save the old state in the converter object
+                     * right here, then change the local prevLength state variable if necessary.
+                     * Then, if this character turns out to be unassigned or a fallback that
+                     * is not taken, the callback code must not save the new state in the converter
+                     * because the new state is for a character that is not output.
+                     * However, the callback must still restore the state from the converter
+                     * in case the callback function changed it for its output.
+                     */
+                    cnv->fromUnicodeStatus=prevLength; /* save the old state */
+                    value=((const uint16_t *)bytes)[value +(c&0x3f)];
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else if(prevLength<=1) {
+                            length=1;
+                        } else {
+                            /* change from double-byte mode to single-byte */
+                            if (siLength == 1) {
+                                value|=(uint32_t)siBytes[0]<<8;
+                                length = 2;
+                            } else if (siLength == 2) {
+                                value|=(uint32_t)siBytes[1]<<8;
+                                value|=(uint32_t)siBytes[0]<<16;
+                                length = 3;
+                            }
+                            prevLength=1;
+                        }
+                    } else {
+                        if(prevLength==2) {
+                            length=2;
+                        } else {
+                            /* change from single-byte mode to double-byte */
+                            if (soLength == 1) {
+                                value|=(uint32_t)soBytes[0]<<16;
+                                length = 3;
+                            } else if (soLength == 2) {
+                                value|=(uint32_t)soBytes[1]<<16;
+                                value|=(uint32_t)soBytes[0]<<24;
+                                length = 4;
+                            }
+                            prevLength=2;
+                        }
+                    }
+                    break;
+                case MBCS_OUTPUT_DBCS_ONLY:
+                    /* table with single-byte results, but only DBCS mappings used */
+                    value=((const uint16_t *)bytes)[value +(c&0x3f)];
+                    if(value<=0xff) {
+                        /* no mapping or SBCS result, not taken for DBCS-only */
+                        goto unassigned;
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_3:
+                    p=bytes+(value+(c&0x3f))*3;
+                    value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else {
+                            length=1;
+                        }
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else {
+                        length=3;
+                    }
+                    break;
+                case MBCS_OUTPUT_4:
+                    value=((const uint32_t *)bytes)[value +(c&0x3f)];
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else {
+                            length=1;
+                        }
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else if(value<=0xffffff) {
+                        length=3;
+                    } else {
+                        length=4;
+                    }
+                    break;
+                case MBCS_OUTPUT_3_EUC:
+                    value=((const uint16_t *)bytes)[value +(c&0x3f)];
+                    /* EUC 16-bit fixed-length representation */
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else {
+                            length=1;
+                        }
+                    } else if((value&0x8000)==0) {
+                        value|=0x8e8000;
+                        length=3;
+                    } else if((value&0x80)==0) {
+                        value|=0x8f0080;
+                        length=3;
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_4_EUC:
+                    p=bytes+(value+(c&0x3f))*3;
+                    value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                    /* EUC 16-bit fixed-length representation applied to the first two bytes */
+                    if(value<=0xff) {
+                        if(value==0) {
+                            goto unassigned;
+                        } else {
+                            length=1;
+                        }
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else if((value&0x800000)==0) {
+                        value|=0x8e800000;
+                        length=4;
+                    } else if((value&0x8000)==0) {
+                        value|=0x8f008000;
+                        length=4;
+                    } else {
+                        length=3;
+                    }
+                    break;
+                default:
+                    /* must not occur */
+                    /*
+                     * To avoid compiler warnings that value & length may be
+                     * used without having been initialized, we set them here.
+                     * In reality, this is unreachable code.
+                     * Not having a default branch also causes warnings with
+                     * some compilers.
+                     */
+                    value=0;
+                    length=0;
+                    break;
+                }
+                /* output the value */
+            } else {
+                /*
+                 * This also tests if the codepage maps single surrogates.
+                 * If it does, then surrogates are not paired but mapped separately.
+                 * Note that in this case unmatched surrogates are not detected.
+                 */
+                if(U16_IS_SURROGATE(c) && !(unicodeMask&UCNV_HAS_SURROGATES)) {
+                    if(U16_IS_SURROGATE_LEAD(c)) {
+getTrail:
+                        if(source<sourceLimit) {
+                            /* test the following code unit */
+                            UChar trail=*source;
+                            if(U16_IS_TRAIL(trail)) {
+                                ++source;
+                                ++nextSourceIndex;
+                                c=U16_GET_SUPPLEMENTARY(c, trail);
+                                if(!(unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+                                    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+                                    cnv->fromUnicodeStatus=prevLength; /* save the old state */
+                                    /* callback(unassigned) */
+                                    goto unassigned;
+                                }
+                                /* convert this supplementary code point */
+                                /* exit this condition tree */
+                            } else {
+                                /* this is an unmatched lead code unit (1st surrogate) */
+                                /* callback(illegal) */
+                                *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                                break;
+                            }
+                        } else {
+                            /* no more input */
+                            break;
+                        }
+                    } else {
+                        /* this is an unmatched trail code unit (2nd surrogate) */
+                        /* callback(illegal) */
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        break;
+                    }
+                }
+
+                /* convert the Unicode code point in c into codepage bytes */
+
+                /*
+                 * The basic lookup is a triple-stage compact array (trie) lookup.
+                 * For details see the beginning of this file.
+                 *
+                 * Single-byte codepages are handled with a different data structure
+                 * by _MBCSSingle... functions.
+                 *
+                 * The result consists of a 32-bit value from stage 2 and
+                 * a pointer to as many bytes as are stored per character.
+                 * The pointer points to the character's bytes in stage 3.
+                 * Bits 15..0 of the stage 2 entry contain the stage 3 index
+                 * for that pointer, while bits 31..16 are flags for which of
+                 * the 16 characters in the block are roundtrip-assigned.
+                 *
+                 * For 2-byte and 4-byte codepages, the bytes are stored as uint16_t
+                 * respectively as uint32_t, in the platform encoding.
+                 * For 3-byte codepages, the bytes are always stored in big-endian order.
+                 *
+                 * For EUC encodings that use only either 0x8e or 0x8f as the first
+                 * byte of their longest byte sequences, the first two bytes in
+                 * this third stage indicate with their 7th bits whether these bytes
+                 * are to be written directly or actually need to be preceeded by
+                 * one of the two Single-Shift codes. With this, the third stage
+                 * stores one byte fewer per character than the actual maximum length of
+                 * EUC byte sequences.
+                 *
+                 * Other than that, leading zero bytes are removed and the other
+                 * bytes output. A single zero byte may be output if the "assigned"
+                 * bit in stage 2 was on.
+                 * The data structure does not support zero byte output as a fallback,
+                 * and also does not allow output of leading zeros.
+                 */
+                stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+
+                /* get the bytes and the length for the output */
+                switch(outputType) {
+                case MBCS_OUTPUT_2:
+                    value=MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c);
+                    if(value<=0xff) {
+                        length=1;
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_2_SISO:
+                    /* 1/2-byte stateful with Shift-In/Shift-Out */
+                    /*
+                     * Save the old state in the converter object
+                     * right here, then change the local prevLength state variable if necessary.
+                     * Then, if this character turns out to be unassigned or a fallback that
+                     * is not taken, the callback code must not save the new state in the converter
+                     * because the new state is for a character that is not output.
+                     * However, the callback must still restore the state from the converter
+                     * in case the callback function changed it for its output.
+                     */
+                    cnv->fromUnicodeStatus=prevLength; /* save the old state */
+                    value=MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c);
+                    if(value<=0xff) {
+                        if(value==0 && MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c)==0) {
+                            /* no mapping, leave value==0 */
+                            length=0;
+                        } else if(prevLength<=1) {
+                            length=1;
+                        } else {
+                            /* change from double-byte mode to single-byte */
+                            if (siLength == 1) {
+                                value|=(uint32_t)siBytes[0]<<8;
+                                length = 2;
+                            } else if (siLength == 2) {
+                                value|=(uint32_t)siBytes[1]<<8;
+                                value|=(uint32_t)siBytes[0]<<16;
+                                length = 3;
+                            }
+                            prevLength=1;
+                        }
+                    } else {
+                        if(prevLength==2) {
+                            length=2;
+                        } else {
+                            /* change from single-byte mode to double-byte */
+                            if (soLength == 1) {
+                                value|=(uint32_t)soBytes[0]<<16;
+                                length = 3;
+                            } else if (soLength == 2) {
+                                value|=(uint32_t)soBytes[1]<<16;
+                                value|=(uint32_t)soBytes[0]<<24;
+                                length = 4;
+                            }
+                            prevLength=2;
+                        }
+                    }
+                    break;
+                case MBCS_OUTPUT_DBCS_ONLY:
+                    /* table with single-byte results, but only DBCS mappings used */
+                    value=MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c);
+                    if(value<=0xff) {
+                        /* no mapping or SBCS result, not taken for DBCS-only */
+                        value=stage2Entry=0; /* stage2Entry=0 to reset roundtrip flags */
+                        length=0;
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_3:
+                    p=MBCS_POINTER_3_FROM_STAGE_2(bytes, stage2Entry, c);
+                    value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                    if(value<=0xff) {
+                        length=1;
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else {
+                        length=3;
+                    }
+                    break;
+                case MBCS_OUTPUT_4:
+                    value=MBCS_VALUE_4_FROM_STAGE_2(bytes, stage2Entry, c);
+                    if(value<=0xff) {
+                        length=1;
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else if(value<=0xffffff) {
+                        length=3;
+                    } else {
+                        length=4;
+                    }
+                    break;
+                case MBCS_OUTPUT_3_EUC:
+                    value=MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c);
+                    /* EUC 16-bit fixed-length representation */
+                    if(value<=0xff) {
+                        length=1;
+                    } else if((value&0x8000)==0) {
+                        value|=0x8e8000;
+                        length=3;
+                    } else if((value&0x80)==0) {
+                        value|=0x8f0080;
+                        length=3;
+                    } else {
+                        length=2;
+                    }
+                    break;
+                case MBCS_OUTPUT_4_EUC:
+                    p=MBCS_POINTER_3_FROM_STAGE_2(bytes, stage2Entry, c);
+                    value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                    /* EUC 16-bit fixed-length representation applied to the first two bytes */
+                    if(value<=0xff) {
+                        length=1;
+                    } else if(value<=0xffff) {
+                        length=2;
+                    } else if((value&0x800000)==0) {
+                        value|=0x8e800000;
+                        length=4;
+                    } else if((value&0x8000)==0) {
+                        value|=0x8f008000;
+                        length=4;
+                    } else {
+                        length=3;
+                    }
+                    break;
+                default:
+                    /* must not occur */
+                    /*
+                     * To avoid compiler warnings that value & length may be
+                     * used without having been initialized, we set them here.
+                     * In reality, this is unreachable code.
+                     * Not having a default branch also causes warnings with
+                     * some compilers.
+                     */
+                    value=stage2Entry=0; /* stage2Entry=0 to reset roundtrip flags */
+                    length=0;
+                    break;
+                }
+
+                /* is this code point assigned, or do we use fallbacks? */
+                if(!(MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c)!=0 ||
+                     (UCNV_FROM_U_USE_FALLBACK(cnv, c) && value!=0))
+                ) {
+                    /*
+                     * We allow a 0 byte output if the "assigned" bit is set for this entry.
+                     * There is no way with this data structure for fallback output
+                     * to be a zero byte.
+                     */
+
+unassigned:
+                    /* try an extension mapping */
+                    pArgs->source=source;
+                    c=_extFromU(cnv, cnv->sharedData,
+                                c, &source, sourceLimit,
+                                &target, target+targetCapacity,
+                                &offsets, sourceIndex,
+                                pArgs->flush,
+                                pErrorCode);
+                    nextSourceIndex+=(int32_t)(source-pArgs->source);
+                    prevLength=cnv->fromUnicodeStatus; /* restore SISO state */
+
+                    if(U_FAILURE(*pErrorCode)) {
+                        /* not mappable or buffer overflow */
+                        break;
+                    } else {
+                        /* a mapping was written to the target, continue */
+
+                        /* recalculate the targetCapacity after an extension mapping */
+                        targetCapacity=(int32_t)(pArgs->targetLimit-(char *)target);
+
+                        /* normal end of conversion: prepare for a new character */
+                        if(offsets!=NULL) {
+                            prevSourceIndex=sourceIndex;
+                            sourceIndex=nextSourceIndex;
+                        }
+                        continue;
+                    }
+                }
+            }
+
+            /* write the output character bytes from value and length */
+            /* from the first if in the loop we know that targetCapacity>0 */
+            if(length<=targetCapacity) {
+                if(offsets==NULL) {
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 4:
+                        *target++=(uint8_t)(value>>24);
+                        U_FALLTHROUGH;
+                    case 3:
+                        *target++=(uint8_t)(value>>16);
+                        U_FALLTHROUGH;
+                    case 2:
+                        *target++=(uint8_t)(value>>8);
+                        U_FALLTHROUGH;
+                    case 1:
+                        *target++=(uint8_t)value;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                } else {
+                    switch(length) {
+                        /* each branch falls through to the next one */
+                    case 4:
+                        *target++=(uint8_t)(value>>24);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 3:
+                        *target++=(uint8_t)(value>>16);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 2:
+                        *target++=(uint8_t)(value>>8);
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    case 1:
+                        *target++=(uint8_t)value;
+                        *offsets++=sourceIndex;
+                        U_FALLTHROUGH;
+                    default:
+                        /* will never occur */
+                        break;
+                    }
+                }
+                targetCapacity-=length;
+            } else {
+                uint8_t *charErrorBuffer;
+
+                /*
+                 * We actually do this backwards here:
+                 * In order to save an intermediate variable, we output
+                 * first to the overflow buffer what does not fit into the
+                 * regular target.
+                 */
+                /* we know that 1<=targetCapacity<length<=4 */
+                length-=targetCapacity;
+                charErrorBuffer=(uint8_t *)cnv->charErrorBuffer;
+                switch(length) {
+                    /* each branch falls through to the next one */
+                case 3:
+                    *charErrorBuffer++=(uint8_t)(value>>16);
+                    U_FALLTHROUGH;
+                case 2:
+                    *charErrorBuffer++=(uint8_t)(value>>8);
+                    U_FALLTHROUGH;
+                case 1:
+                    *charErrorBuffer=(uint8_t)value;
+                    U_FALLTHROUGH;
+                default:
+                    /* will never occur */
+                    break;
+                }
+                cnv->charErrorBufferLength=(int8_t)length;
+
+                /* now output what fits into the regular target */
+                value>>=8*length; /* length was reduced by targetCapacity */
+                switch(targetCapacity) {
+                    /* each branch falls through to the next one */
+                case 3:
+                    *target++=(uint8_t)(value>>16);
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    U_FALLTHROUGH;
+                case 2:
+                    *target++=(uint8_t)(value>>8);
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    U_FALLTHROUGH;
+                case 1:
+                    *target++=(uint8_t)value;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    U_FALLTHROUGH;
+                default:
+                    /* will never occur */
+                    break;
+                }
+
+                /* target overflow */
+                targetCapacity=0;
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                c=0;
+                break;
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+            if(offsets!=NULL) {
+                prevSourceIndex=sourceIndex;
+                sourceIndex=nextSourceIndex;
+            }
+            continue;
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /*
+     * the end of the input stream and detection of truncated input
+     * are handled by the framework, but for EBCDIC_STATEFUL conversion
+     * we need to emit an SI at the very end
+     *
+     * conditions:
+     *   successful
+     *   EBCDIC_STATEFUL in DBCS mode
+     *   end of input and no truncated input
+     */
+    if( U_SUCCESS(*pErrorCode) &&
+        outputType==MBCS_OUTPUT_2_SISO && prevLength==2 &&
+        pArgs->flush && source>=sourceLimit && c==0
+    ) {
+        /* EBCDIC_STATEFUL ending with DBCS: emit an SI to return the output stream to SBCS */
+        if(targetCapacity>0) {
+            *target++=(uint8_t)siBytes[0];
+            if (siLength == 2) {
+                if (targetCapacity<2) {
+                    cnv->charErrorBuffer[0]=(uint8_t)siBytes[1];
+                    cnv->charErrorBufferLength=1;
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                } else {
+                    *target++=(uint8_t)siBytes[1];
+                }
+            }
+            if(offsets!=NULL) {
+                /* set the last source character's index (sourceIndex points at sourceLimit now) */
+                *offsets++=prevSourceIndex;
+            }
+        } else {
+            /* target is full */
+            cnv->charErrorBuffer[0]=(uint8_t)siBytes[0];
+            if (siLength == 2) {
+                cnv->charErrorBuffer[1]=(uint8_t)siBytes[1];
+            }
+            cnv->charErrorBufferLength=siLength;
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        }
+        prevLength=1; /* we switched into SBCS */
+    }
+
+    /* set the converter state back into UConverter */
+    cnv->fromUChar32=c;
+    cnv->fromUnicodeStatus=prevLength;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+}
+
+/*
+ * This is another simple conversion function for internal use by other
+ * conversion implementations.
+ * It does not use the converter state nor call callbacks.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It handles conversion extensions but not GB 18030.
+ *
+ * It converts one single Unicode code point into codepage bytes, encoded
+ * as one 32-bit value. The function returns the number of bytes in *pValue:
+ * 1..4 the number of bytes in *pValue
+ * 0    unassigned (*pValue undefined)
+ * -1   illegal (currently not used, *pValue undefined)
+ *
+ * *pValue will contain the resulting bytes with the last byte in bits 7..0,
+ * the second to last byte in bits 15..8, etc.
+ * Currently, the function assumes but does not check that 0<=c<=0x10ffff.
+ */
+U_CFUNC int32_t
+ucnv_MBCSFromUChar32(UConverterSharedData *sharedData,
+                 UChar32 c, uint32_t *pValue,
+                 UBool useFallback) {
+    const int32_t *cx;
+    const uint16_t *table;
+#if 0
+/* #if 0 because this is not currently used in ICU - reduce code, increase code coverage */
+    const uint8_t *p;
+#endif
+    uint32_t stage2Entry;
+    uint32_t value;
+    int32_t length;
+
+    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+    if(c<=0xffff || (sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+        table=sharedData->mbcs.fromUnicodeTable;
+
+        /* convert the Unicode code point in c into codepage bytes (same as in _MBCSFromUnicodeWithOffsets) */
+        if(sharedData->mbcs.outputType==MBCS_OUTPUT_1) {
+            value=MBCS_SINGLE_RESULT_FROM_U(table, (uint16_t *)sharedData->mbcs.fromUnicodeBytes, c);
+            /* is this code point assigned, or do we use fallbacks? */
+            if(useFallback ? value>=0x800 : value>=0xc00) {
+                *pValue=value&0xff;
+                return 1;
+            }
+        } else /* outputType!=MBCS_OUTPUT_1 */ {
+            stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+
+            /* get the bytes and the length for the output */
+            switch(sharedData->mbcs.outputType) {
+            case MBCS_OUTPUT_2:
+                value=MBCS_VALUE_2_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                if(value<=0xff) {
+                    length=1;
+                } else {
+                    length=2;
+                }
+                break;
+#if 0
+/* #if 0 because this is not currently used in ICU - reduce code, increase code coverage */
+            case MBCS_OUTPUT_DBCS_ONLY:
+                /* table with single-byte results, but only DBCS mappings used */
+                value=MBCS_VALUE_2_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                if(value<=0xff) {
+                    /* no mapping or SBCS result, not taken for DBCS-only */
+                    value=stage2Entry=0; /* stage2Entry=0 to reset roundtrip flags */
+                    length=0;
+                } else {
+                    length=2;
+                }
+                break;
+            case MBCS_OUTPUT_3:
+                p=MBCS_POINTER_3_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                if(value<=0xff) {
+                    length=1;
+                } else if(value<=0xffff) {
+                    length=2;
+                } else {
+                    length=3;
+                }
+                break;
+            case MBCS_OUTPUT_4:
+                value=MBCS_VALUE_4_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                if(value<=0xff) {
+                    length=1;
+                } else if(value<=0xffff) {
+                    length=2;
+                } else if(value<=0xffffff) {
+                    length=3;
+                } else {
+                    length=4;
+                }
+                break;
+            case MBCS_OUTPUT_3_EUC:
+                value=MBCS_VALUE_2_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                /* EUC 16-bit fixed-length representation */
+                if(value<=0xff) {
+                    length=1;
+                } else if((value&0x8000)==0) {
+                    value|=0x8e8000;
+                    length=3;
+                } else if((value&0x80)==0) {
+                    value|=0x8f0080;
+                    length=3;
+                } else {
+                    length=2;
+                }
+                break;
+            case MBCS_OUTPUT_4_EUC:
+                p=MBCS_POINTER_3_FROM_STAGE_2(sharedData->mbcs.fromUnicodeBytes, stage2Entry, c);
+                value=((uint32_t)*p<<16)|((uint32_t)p[1]<<8)|p[2];
+                /* EUC 16-bit fixed-length representation applied to the first two bytes */
+                if(value<=0xff) {
+                    length=1;
+                } else if(value<=0xffff) {
+                    length=2;
+                } else if((value&0x800000)==0) {
+                    value|=0x8e800000;
+                    length=4;
+                } else if((value&0x8000)==0) {
+                    value|=0x8f008000;
+                    length=4;
+                } else {
+                    length=3;
+                }
+                break;
+#endif
+            default:
+                /* must not occur */
+                return -1;
+            }
+
+            /* is this code point assigned, or do we use fallbacks? */
+            if( MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c) ||
+                (FROM_U_USE_FALLBACK(useFallback, c) && value!=0)
+            ) {
+                /*
+                 * We allow a 0 byte output if the "assigned" bit is set for this entry.
+                 * There is no way with this data structure for fallback output
+                 * to be a zero byte.
+                 */
+                /* assigned */
+                *pValue=value;
+                return length;
+            }
+        }
+    }
+
+    cx=sharedData->mbcs.extIndexes;
+    if(cx!=NULL) {
+        length=ucnv_extSimpleMatchFromU(cx, c, pValue, useFallback);
+        return length>=0 ? length : -length;  /* return abs(length); */
+    }
+
+    /* unassigned */
+    return 0;
+}
+
+
+#if 0
+/*
+ * This function has been moved to ucnv2022.c for inlining.
+ * This implementation is here only for documentation purposes
+ */
+
+/**
+ * This version of ucnv_MBCSFromUChar32() is optimized for single-byte codepages.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It does not handle conversion extensions (_extFromU()).
+ *
+ * It returns the codepage byte for the code point, or -1 if it is unassigned.
+ */
+U_CFUNC int32_t
+ucnv_MBCSSingleFromUChar32(UConverterSharedData *sharedData,
+                       UChar32 c,
+                       UBool useFallback) {
+    const uint16_t *table;
+    int32_t value;
+
+    /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+    if(c>=0x10000 && !(sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY)) {
+        return -1;
+    }
+
+    /* convert the Unicode code point in c into codepage bytes (same as in _MBCSFromUnicodeWithOffsets) */
+    table=sharedData->mbcs.fromUnicodeTable;
+
+    /* get the byte for the output */
+    value=MBCS_SINGLE_RESULT_FROM_U(table, (uint16_t *)sharedData->mbcs.fromUnicodeBytes, c);
+    /* is this code point assigned, or do we use fallbacks? */
+    if(useFallback ? value>=0x800 : value>=0xc00) {
+        return value&0xff;
+    } else {
+        return -1;
+    }
+}
+#endif
+
+/* MBCS-from-UTF-8 conversion functions ------------------------------------- */
+
+/* offsets for n-byte UTF-8 sequences that were calculated with ((lead<<6)+trail)<<6+trail... */
+static const UChar32
+utf8_offsets[5]={ 0, 0, 0x3080, 0xE2080, 0x3C82080 };
+
+static void U_CALLCONV
+ucnv_SBCSFromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                  UConverterToUnicodeArgs *pToUArgs,
+                  UErrorCode *pErrorCode) {
+    UConverter *utf8, *cnv;
+    const uint8_t *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+
+    const uint16_t *table, *sbcsIndex;
+    const uint16_t *results;
+
+    int8_t oldToULength, toULength, toULimit;
+
+    UChar32 c;
+    uint8_t b, t1, t2;
+
+    uint32_t asciiRoundtrips;
+    uint16_t value, minValue = 0;
+    UBool hasSupplementary;
+
+    /* set up the local pointers */
+    utf8=pToUArgs->converter;
+    cnv=pFromUArgs->converter;
+    source=(uint8_t *)pToUArgs->source;
+    sourceLimit=(uint8_t *)pToUArgs->sourceLimit;
+    target=(uint8_t *)pFromUArgs->target;
+    targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target);
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    sbcsIndex=cnv->sharedData->mbcs.sbcsIndex;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        results=(uint16_t *)cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        results=(uint16_t *)cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+    asciiRoundtrips=cnv->sharedData->mbcs.asciiRoundtrips;
+
+    if(cnv->useFallback) {
+        /* use all roundtrip and fallback results */
+        minValue=0x800;
+    } else {
+        /* use only roundtrips and fallbacks from private-use characters */
+        minValue=0xc00;
+    }
+    hasSupplementary=(UBool)(cnv->sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY);
+
+    /* get the converter state from the UTF-8 UConverter */
+    if(utf8->toULength > 0) {
+        toULength=oldToULength=utf8->toULength;
+        toULimit=(int8_t)utf8->mode;
+        c=(UChar32)utf8->toUnicodeStatus;
+    } else {
+        toULength=oldToULength=toULimit=0;
+        c = 0;
+    }
+
+    // The conversion loop checks source<sourceLimit only once per 1/2/3-byte character.
+    // If the buffer ends with a truncated 2- or 3-byte sequence,
+    // then we reduce the sourceLimit to before that,
+    // and collect the remaining bytes after the conversion loop.
+    {
+        // Do not go back into the bytes that will be read for finishing a partial
+        // sequence from the previous buffer.
+        int32_t length=(int32_t)(sourceLimit-source) - (toULimit-oldToULength);
+        if(length>0) {
+            uint8_t b1=*(sourceLimit-1);
+            if(U8_IS_SINGLE(b1)) {
+                // common ASCII character
+            } else if(U8_IS_TRAIL(b1) && length>=2) {
+                uint8_t b2=*(sourceLimit-2);
+                if(0xe0<=b2 && b2<0xf0 && U8_IS_VALID_LEAD3_AND_T1(b2, b1)) {
+                    // truncated 3-byte sequence
+                    sourceLimit-=2;
+                }
+            } else if(0xc2<=b1 && b1<0xf0) {
+                // truncated 2- or 3-byte sequence
+                --sourceLimit;
+            }
+        }
+    }
+
+    if(c!=0 && targetCapacity>0) {
+        utf8->toUnicodeStatus=0;
+        utf8->toULength=0;
+        goto moreBytes;
+        /*
+         * Note: We could avoid the goto by duplicating some of the moreBytes
+         * code, but only up to the point of collecting a complete UTF-8
+         * sequence; then recurse for the toUBytes[toULength]
+         * and then continue with normal conversion.
+         *
+         * If so, move this code to just after initializing the minimum
+         * set of local variables for reading the UTF-8 input
+         * (utf8, source, target, limits but not cnv, table, minValue, etc.).
+         *
+         * Potential advantages:
+         * - avoid the goto
+         * - oldToULength could become a local variable in just those code blocks
+         *   that deal with buffer boundaries
+         * - possibly faster if the goto prevents some compiler optimizations
+         *   (this would need measuring to confirm)
+         * Disadvantage:
+         * - code duplication
+         */
+    }
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        if(targetCapacity>0) {
+            b=*source++;
+            if(U8_IS_SINGLE(b)) {
+                /* convert ASCII */
+                if(IS_ASCII_ROUNDTRIP(b, asciiRoundtrips)) {
+                    *target++=(uint8_t)b;
+                    --targetCapacity;
+                    continue;
+                } else {
+                    c=b;
+                    value=SBCS_RESULT_FROM_UTF8(sbcsIndex, results, 0, c);
+                }
+            } else {
+                if(b<0xe0) {
+                    if( /* handle U+0080..U+07FF inline */
+                        b>=0xc2 &&
+                        (t1=(uint8_t)(*source-0x80)) <= 0x3f
+                    ) {
+                        c=b&0x1f;
+                        ++source;
+                        value=SBCS_RESULT_FROM_UTF8(sbcsIndex, results, c, t1);
+                        if(value>=minValue) {
+                            *target++=(uint8_t)value;
+                            --targetCapacity;
+                            continue;
+                        } else {
+                            c=(c<<6)|t1;
+                        }
+                    } else {
+                        c=-1;
+                    }
+                } else if(b==0xe0) {
+                    if( /* handle U+0800..U+0FFF inline */
+                        (t1=(uint8_t)(source[0]-0x80)) <= 0x3f && t1 >= 0x20 &&
+                        (t2=(uint8_t)(source[1]-0x80)) <= 0x3f
+                    ) {
+                        c=t1;
+                        source+=2;
+                        value=SBCS_RESULT_FROM_UTF8(sbcsIndex, results, c, t2);
+                        if(value>=minValue) {
+                            *target++=(uint8_t)value;
+                            --targetCapacity;
+                            continue;
+                        } else {
+                            c=(c<<6)|t2;
+                        }
+                    } else {
+                        c=-1;
+                    }
+                } else {
+                    c=-1;
+                }
+
+                if(c<0) {
+                    /* handle "complicated" and error cases, and continuing partial characters */
+                    oldToULength=0;
+                    toULength=1;
+                    toULimit=U8_COUNT_BYTES_NON_ASCII(b);
+                    c=b;
+moreBytes:
+                    while(toULength<toULimit) {
+                        /*
+                         * The sourceLimit may have been adjusted before the conversion loop
+                         * to stop before a truncated sequence.
+                         * Here we need to use the real limit in case we have two truncated
+                         * sequences at the end.
+                         * See ticket #7492.
+                         */
+                        if(source<(uint8_t *)pToUArgs->sourceLimit) {
+                            b=*source;
+                            if(icu::UTF8::isValidTrail(c, b, toULength, toULimit)) {
+                                ++source;
+                                ++toULength;
+                                c=(c<<6)+b;
+                            } else {
+                                break; /* sequence too short, stop with toULength<toULimit */
+                            }
+                        } else {
+                            /* store the partial UTF-8 character, compatible with the regular UTF-8 converter */
+                            source-=(toULength-oldToULength);
+                            while(oldToULength<toULength) {
+                                utf8->toUBytes[oldToULength++]=*source++;
+                            }
+                            utf8->toUnicodeStatus=c;
+                            utf8->toULength=toULength;
+                            utf8->mode=toULimit;
+                            pToUArgs->source=(char *)source;
+                            pFromUArgs->target=(char *)target;
+                            return;
+                        }
+                    }
+
+                    if(toULength==toULimit) {
+                        c-=utf8_offsets[toULength];
+                        if(toULength<=3) {  /* BMP */
+                            value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+                        } else {
+                            /* supplementary code point */
+                            if(!hasSupplementary) {
+                                /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+                                value=0;
+                            } else {
+                                value=MBCS_SINGLE_RESULT_FROM_U(table, results, c);
+                            }
+                        }
+                    } else {
+                        /* error handling: illegal UTF-8 byte sequence */
+                        source-=(toULength-oldToULength);
+                        while(oldToULength<toULength) {
+                            utf8->toUBytes[oldToULength++]=*source++;
+                        }
+                        utf8->toULength=toULength;
+                        pToUArgs->source=(char *)source;
+                        pFromUArgs->target=(char *)target;
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        return;
+                    }
+                }
+            }
+
+            if(value>=minValue) {
+                /* output the mapping for c */
+                *target++=(uint8_t)value;
+                --targetCapacity;
+            } else {
+                /* value<minValue means c is unassigned (unmappable) */
+                /*
+                 * Try an extension mapping.
+                 * Pass in no source because we don't have UTF-16 input.
+                 * If we have a partial match on c, we will return and revert
+                 * to UTF-8->UTF-16->charset conversion.
+                 */
+                static const UChar nul=0;
+                const UChar *noSource=&nul;
+                c=_extFromU(cnv, cnv->sharedData,
+                            c, &noSource, noSource,
+                            &target, target+targetCapacity,
+                            NULL, -1,
+                            pFromUArgs->flush,
+                            pErrorCode);
+
+                if(U_FAILURE(*pErrorCode)) {
+                    /* not mappable or buffer overflow */
+                    cnv->fromUChar32=c;
+                    break;
+                } else if(cnv->preFromUFirstCP>=0) {
+                    /*
+                     * Partial match, return and revert to pivoting.
+                     * In normal from-UTF-16 conversion, we would just continue
+                     * but then exit the loop because the extension match would
+                     * have consumed the source.
+                     */
+                    *pErrorCode=U_USING_DEFAULT_WARNING;
+                    break;
+                } else {
+                    /* a mapping was written to the target, continue */
+
+                    /* recalculate the targetCapacity after an extension mapping */
+                    targetCapacity=(int32_t)(pFromUArgs->targetLimit-(char *)target);
+                }
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /*
+     * The sourceLimit may have been adjusted before the conversion loop
+     * to stop before a truncated sequence.
+     * If so, then collect the truncated sequence now.
+     */
+    if(U_SUCCESS(*pErrorCode) &&
+            cnv->preFromUFirstCP<0 &&
+            source<(sourceLimit=(uint8_t *)pToUArgs->sourceLimit)) {
+        c=utf8->toUBytes[0]=b=*source++;
+        toULength=1;
+        toULimit=U8_COUNT_BYTES(b);
+        while(source<sourceLimit) {
+            utf8->toUBytes[toULength++]=b=*source++;
+            c=(c<<6)+b;
+        }
+        utf8->toUnicodeStatus=c;
+        utf8->toULength=toULength;
+        utf8->mode=toULimit;
+    }
+
+    /* write back the updated pointers */
+    pToUArgs->source=(char *)source;
+    pFromUArgs->target=(char *)target;
+}
+
+static void U_CALLCONV
+ucnv_DBCSFromUTF8(UConverterFromUnicodeArgs *pFromUArgs,
+                  UConverterToUnicodeArgs *pToUArgs,
+                  UErrorCode *pErrorCode) {
+    UConverter *utf8, *cnv;
+    const uint8_t *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+
+    const uint16_t *table, *mbcsIndex;
+    const uint16_t *results;
+
+    int8_t oldToULength, toULength, toULimit;
+
+    UChar32 c;
+    uint8_t b, t1, t2;
+
+    uint32_t stage2Entry;
+    uint32_t asciiRoundtrips;
+    uint16_t value = 0;
+    UBool hasSupplementary;
+
+    /* set up the local pointers */
+    utf8=pToUArgs->converter;
+    cnv=pFromUArgs->converter;
+    source=(uint8_t *)pToUArgs->source;
+    sourceLimit=(uint8_t *)pToUArgs->sourceLimit;
+    target=(uint8_t *)pFromUArgs->target;
+    targetCapacity=(int32_t)(pFromUArgs->targetLimit-pFromUArgs->target);
+
+    table=cnv->sharedData->mbcs.fromUnicodeTable;
+    mbcsIndex=cnv->sharedData->mbcs.mbcsIndex;
+    if((cnv->options&UCNV_OPTION_SWAP_LFNL)!=0) {
+        results=(uint16_t *)cnv->sharedData->mbcs.swapLFNLFromUnicodeBytes;
+    } else {
+        results=(uint16_t *)cnv->sharedData->mbcs.fromUnicodeBytes;
+    }
+    asciiRoundtrips=cnv->sharedData->mbcs.asciiRoundtrips;
+
+    hasSupplementary=(UBool)(cnv->sharedData->mbcs.unicodeMask&UCNV_HAS_SUPPLEMENTARY);
+
+    /* get the converter state from the UTF-8 UConverter */
+    if(utf8->toULength > 0) {
+        toULength=oldToULength=utf8->toULength;
+        toULimit=(int8_t)utf8->mode;
+        c=(UChar32)utf8->toUnicodeStatus;
+    } else {
+        toULength=oldToULength=toULimit=0;
+        c = 0;
+    }
+
+    // The conversion loop checks source<sourceLimit only once per 1/2/3-byte character.
+    // If the buffer ends with a truncated 2- or 3-byte sequence,
+    // then we reduce the sourceLimit to before that,
+    // and collect the remaining bytes after the conversion loop.
+    {
+        // Do not go back into the bytes that will be read for finishing a partial
+        // sequence from the previous buffer.
+        int32_t length=(int32_t)(sourceLimit-source) - (toULimit-oldToULength);
+        if(length>0) {
+            uint8_t b1=*(sourceLimit-1);
+            if(U8_IS_SINGLE(b1)) {
+                // common ASCII character
+            } else if(U8_IS_TRAIL(b1) && length>=2) {
+                uint8_t b2=*(sourceLimit-2);
+                if(0xe0<=b2 && b2<0xf0 && U8_IS_VALID_LEAD3_AND_T1(b2, b1)) {
+                    // truncated 3-byte sequence
+                    sourceLimit-=2;
+                }
+            } else if(0xc2<=b1 && b1<0xf0) {
+                // truncated 2- or 3-byte sequence
+                --sourceLimit;
+            }
+        }
+    }
+
+    if(c!=0 && targetCapacity>0) {
+        utf8->toUnicodeStatus=0;
+        utf8->toULength=0;
+        goto moreBytes;
+        /* See note in ucnv_SBCSFromUTF8() about this goto. */
+    }
+
+    /* conversion loop */
+    while(source<sourceLimit) {
+        if(targetCapacity>0) {
+            b=*source++;
+            if(U8_IS_SINGLE(b)) {
+                /* convert ASCII */
+                if(IS_ASCII_ROUNDTRIP(b, asciiRoundtrips)) {
+                    *target++=b;
+                    --targetCapacity;
+                    continue;
+                } else {
+                    value=DBCS_RESULT_FROM_UTF8(mbcsIndex, results, 0, b);
+                    if(value==0) {
+                        c=b;
+                        goto unassigned;
+                    }
+                }
+            } else {
+                if(b>=0xe0) {
+                    if( /* handle U+0800..U+D7FF inline */
+                        b<=0xed &&  // do not assume maxFastUChar>0xd7ff
+                        U8_IS_VALID_LEAD3_AND_T1(b, t1=source[0]) &&
+                        (t2=(uint8_t)(source[1]-0x80)) <= 0x3f
+                    ) {
+                        c=((b&0xf)<<6)|(t1&0x3f);
+                        source+=2;
+                        value=DBCS_RESULT_FROM_UTF8(mbcsIndex, results, c, t2);
+                        if(value==0) {
+                            c=(c<<6)|t2;
+                            goto unassigned;
+                        }
+                    } else {
+                        c=-1;
+                    }
+                } else {
+                    if( /* handle U+0080..U+07FF inline */
+                        b>=0xc2 &&
+                        (t1=(uint8_t)(*source-0x80)) <= 0x3f
+                    ) {
+                        c=b&0x1f;
+                        ++source;
+                        value=DBCS_RESULT_FROM_UTF8(mbcsIndex, results, c, t1);
+                        if(value==0) {
+                            c=(c<<6)|t1;
+                            goto unassigned;
+                        }
+                    } else {
+                        c=-1;
+                    }
+                }
+
+                if(c<0) {
+                    /* handle "complicated" and error cases, and continuing partial characters */
+                    oldToULength=0;
+                    toULength=1;
+                    toULimit=U8_COUNT_BYTES_NON_ASCII(b);
+                    c=b;
+moreBytes:
+                    while(toULength<toULimit) {
+                        /*
+                         * The sourceLimit may have been adjusted before the conversion loop
+                         * to stop before a truncated sequence.
+                         * Here we need to use the real limit in case we have two truncated
+                         * sequences at the end.
+                         * See ticket #7492.
+                         */
+                        if(source<(uint8_t *)pToUArgs->sourceLimit) {
+                            b=*source;
+                            if(icu::UTF8::isValidTrail(c, b, toULength, toULimit)) {
+                                ++source;
+                                ++toULength;
+                                c=(c<<6)+b;
+                            } else {
+                                break; /* sequence too short, stop with toULength<toULimit */
+                            }
+                        } else {
+                            /* store the partial UTF-8 character, compatible with the regular UTF-8 converter */
+                            source-=(toULength-oldToULength);
+                            while(oldToULength<toULength) {
+                                utf8->toUBytes[oldToULength++]=*source++;
+                            }
+                            utf8->toUnicodeStatus=c;
+                            utf8->toULength=toULength;
+                            utf8->mode=toULimit;
+                            pToUArgs->source=(char *)source;
+                            pFromUArgs->target=(char *)target;
+                            return;
+                        }
+                    }
+
+                    if(toULength==toULimit) {
+                        c-=utf8_offsets[toULength];
+                        if(toULength<=3) {  /* BMP */
+                            stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+                        } else {
+                            /* supplementary code point */
+                            if(!hasSupplementary) {
+                                /* BMP-only codepages are stored without stage 1 entries for supplementary code points */
+                                stage2Entry=0;
+                            } else {
+                                stage2Entry=MBCS_STAGE_2_FROM_U(table, c);
+                            }
+                        }
+                    } else {
+                        /* error handling: illegal UTF-8 byte sequence */
+                        source-=(toULength-oldToULength);
+                        while(oldToULength<toULength) {
+                            utf8->toUBytes[oldToULength++]=*source++;
+                        }
+                        utf8->toULength=toULength;
+                        pToUArgs->source=(char *)source;
+                        pFromUArgs->target=(char *)target;
+                        *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                        return;
+                    }
+
+                    /* get the bytes and the length for the output */
+                    /* MBCS_OUTPUT_2 */
+                    value=MBCS_VALUE_2_FROM_STAGE_2(results, stage2Entry, c);
+
+                    /* is this code point assigned, or do we use fallbacks? */
+                    if(!(MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c) ||
+                         (UCNV_FROM_U_USE_FALLBACK(cnv, c) && value!=0))
+                    ) {
+                        goto unassigned;
+                    }
+                }
+            }
+
+            /* write the output character bytes from value and length */
+            /* from the first if in the loop we know that targetCapacity>0 */
+            if(value<=0xff) {
+                /* this is easy because we know that there is enough space */
+                *target++=(uint8_t)value;
+                --targetCapacity;
+            } else /* length==2 */ {
+                *target++=(uint8_t)(value>>8);
+                if(2<=targetCapacity) {
+                    *target++=(uint8_t)value;
+                    targetCapacity-=2;
+                } else {
+                    cnv->charErrorBuffer[0]=(char)value;
+                    cnv->charErrorBufferLength=1;
+
+                    /* target overflow */
+                    *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                    break;
+                }
+            }
+            continue;
+
+unassigned:
+            {
+                /*
+                 * Try an extension mapping.
+                 * Pass in no source because we don't have UTF-16 input.
+                 * If we have a partial match on c, we will return and revert
+                 * to UTF-8->UTF-16->charset conversion.
+                 */
+                static const UChar nul=0;
+                const UChar *noSource=&nul;
+                c=_extFromU(cnv, cnv->sharedData,
+                            c, &noSource, noSource,
+                            &target, target+targetCapacity,
+                            NULL, -1,
+                            pFromUArgs->flush,
+                            pErrorCode);
+
+                if(U_FAILURE(*pErrorCode)) {
+                    /* not mappable or buffer overflow */
+                    cnv->fromUChar32=c;
+                    break;
+                } else if(cnv->preFromUFirstCP>=0) {
+                    /*
+                     * Partial match, return and revert to pivoting.
+                     * In normal from-UTF-16 conversion, we would just continue
+                     * but then exit the loop because the extension match would
+                     * have consumed the source.
+                     */
+                    *pErrorCode=U_USING_DEFAULT_WARNING;
+                    break;
+                } else {
+                    /* a mapping was written to the target, continue */
+
+                    /* recalculate the targetCapacity after an extension mapping */
+                    targetCapacity=(int32_t)(pFromUArgs->targetLimit-(char *)target);
+                    continue;
+                }
+            }
+        } else {
+            /* target is full */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            break;
+        }
+    }
+
+    /*
+     * The sourceLimit may have been adjusted before the conversion loop
+     * to stop before a truncated sequence.
+     * If so, then collect the truncated sequence now.
+     */
+    if(U_SUCCESS(*pErrorCode) &&
+            cnv->preFromUFirstCP<0 &&
+            source<(sourceLimit=(uint8_t *)pToUArgs->sourceLimit)) {
+        c=utf8->toUBytes[0]=b=*source++;
+        toULength=1;
+        toULimit=U8_COUNT_BYTES(b);
+        while(source<sourceLimit) {
+            utf8->toUBytes[toULength++]=b=*source++;
+            c=(c<<6)+b;
+        }
+        utf8->toUnicodeStatus=c;
+        utf8->toULength=toULength;
+        utf8->mode=toULimit;
+    }
+
+    /* write back the updated pointers */
+    pToUArgs->source=(char *)source;
+    pFromUArgs->target=(char *)target;
+}
+
+/* miscellaneous ------------------------------------------------------------ */
+
+static void U_CALLCONV
+ucnv_MBCSGetStarters(const UConverter* cnv,
+                 UBool starters[256],
+                 UErrorCode *) {
+    const int32_t *state0;
+    int i;
+
+    state0=cnv->sharedData->mbcs.stateTable[cnv->sharedData->mbcs.dbcsOnlyState];
+    for(i=0; i<256; ++i) {
+        /* all bytes that cause a state transition from state 0 are lead bytes */
+        starters[i]= (UBool)MBCS_ENTRY_IS_TRANSITION(state0[i]);
+    }
+}
+
+/*
+ * This is an internal function that allows other converter implementations
+ * to check whether a byte is a lead byte.
+ */
+U_CFUNC UBool
+ucnv_MBCSIsLeadByte(UConverterSharedData *sharedData, char byte) {
+    return (UBool)MBCS_ENTRY_IS_TRANSITION(sharedData->mbcs.stateTable[0][(uint8_t)byte]);
+}
+
+static void U_CALLCONV
+ucnv_MBCSWriteSub(UConverterFromUnicodeArgs *pArgs,
+              int32_t offsetIndex,
+              UErrorCode *pErrorCode) {
+    UConverter *cnv=pArgs->converter;
+    char *p, *subchar;
+    char buffer[4];
+    int32_t length;
+
+    /* first, select between subChar and subChar1 */
+    if( cnv->subChar1!=0 &&
+        (cnv->sharedData->mbcs.extIndexes!=NULL ?
+            cnv->useSubChar1 :
+            (cnv->invalidUCharBuffer[0]<=0xff))
+    ) {
+        /* select subChar1 if it is set (not 0) and the unmappable Unicode code point is up to U+00ff (IBM MBCS behavior) */
+        subchar=(char *)&cnv->subChar1;
+        length=1;
+    } else {
+        /* select subChar in all other cases */
+        subchar=(char *)cnv->subChars;
+        length=cnv->subCharLen;
+    }
+
+    /* reset the selector for the next code point */
+    cnv->useSubChar1=FALSE;
+
+    if (cnv->sharedData->mbcs.outputType == MBCS_OUTPUT_2_SISO) {
+        p=buffer;
+
+        /* fromUnicodeStatus contains prevLength */
+        switch(length) {
+        case 1:
+            if(cnv->fromUnicodeStatus==2) {
+                /* DBCS mode and SBCS sub char: change to SBCS */
+                cnv->fromUnicodeStatus=1;
+                *p++=UCNV_SI;
+            }
+            *p++=subchar[0];
+            break;
+        case 2:
+            if(cnv->fromUnicodeStatus<=1) {
+                /* SBCS mode and DBCS sub char: change to DBCS */
+                cnv->fromUnicodeStatus=2;
+                *p++=UCNV_SO;
+            }
+            *p++=subchar[0];
+            *p++=subchar[1];
+            break;
+        default:
+            *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        subchar=buffer;
+        length=(int32_t)(p-buffer);
+    }
+
+    ucnv_cbFromUWriteBytes(pArgs, subchar, length, offsetIndex, pErrorCode);
+}
+
+U_CFUNC UConverterType
+ucnv_MBCSGetType(const UConverter* converter) {
+    /* SBCS, DBCS, and EBCDIC_STATEFUL are replaced by MBCS, but here we cheat a little */
+    if(converter->sharedData->mbcs.countStates==1) {
+        return (UConverterType)UCNV_SBCS;
+    } else if((converter->sharedData->mbcs.outputType&0xff)==MBCS_OUTPUT_2_SISO) {
+        return (UConverterType)UCNV_EBCDIC_STATEFUL;
+    } else if(converter->sharedData->staticData->minBytesPerChar==2 && converter->sharedData->staticData->maxBytesPerChar==2) {
+        return (UConverterType)UCNV_DBCS;
+    }
+    return (UConverterType)UCNV_MBCS;
+}
+
+#endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
diff --git a/thirdparty/icu4c/common/ucnvmbcs.h b/thirdparty/icu4c/common/ucnvmbcs.h
new file mode 100644
index 0000000000..c8f3b89a5e
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvmbcs.h
@@ -0,0 +1,605 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2000-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnvmbcs.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000jul07
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UCNVMBCS_H__
+#define __UCNVMBCS_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "ucnv_cnv.h"
+#include "ucnv_ext.h"
+
+/**
+ * ICU conversion (.cnv) data file structure, following the usual UDataInfo
+ * header.
+ *
+ * Format version: 6.2
+ *
+ * struct UConverterStaticData -- struct containing the converter name, IBM CCSID,
+ *                                min/max bytes per character, etc.
+ *                                see ucnv_bld.h
+ *
+ * --------------------
+ *
+ * The static data is followed by conversionType-specific data structures.
+ * At the moment, there are only variations of MBCS converters. They all have
+ * the same toUnicode structures, while the fromUnicode structures for SBCS
+ * differ from those for other MBCS-style converters.
+ *
+ * _MBCSHeader.version 5 is optional and not backward-compatible
+ * (as usual for changes in the major version field).
+ *
+ * Versions 5.m work like versions 4.m except:
+ * - The _MBCSHeader has variable length (and is always longer than in version 4).
+ *   See the struct _MBCSHeader further description below.
+ * - There is a set of flags which indicate further incompatible changes.
+ *   (Reader code must reject the file if it does not recognize them all.)
+ * - In particular, one of these flags indicates that most of the fromUnicode
+ *   data is missing and must be reconstituted from the toUnicode data
+ *   and from the utf8Friendly mbcsIndex at load time.
+ *   (This only works with a utf8Friendly table.)
+ *   In this case, makeconv may increase maxFastUChar automatically to U+FFFF.
+ *
+ * The first of these versions is 5.3, which is like 4.3 except for the differences above.
+ *
+ * When possible, makeconv continues to generate version 4.m files.
+ *
+ * _MBCSHeader.version 5.4/4.4 supports "good one-way" mappings (|4)
+ * in the extension tables (fromUTableValues bit 30). See ucnv_ext.h for details.
+ *
+ * _MBCSHeader.version 4.3 optionally modifies the fromUnicode data structures
+ * slightly and optionally adds a table for conversion to MBCS (non-SBCS)
+ * charsets.
+ *
+ * The modifications are to make the data utf8Friendly. Not every 4.3 file
+ * file contains utf8Friendly data.
+ * It is utf8Friendly if _MBCSHeader.version[2]!=0.
+ * In this case, the data structures are utf8Friendly up to the code point
+ *   maxFastUChar=((_MBCSHeader.version[2]<<8)|0xff)
+ *
+ * A utf8Friendly file has fromUnicode stage 3 entries for code points up to
+ * maxFastUChar allocated in blocks of 64 for indexing with the 6 bits from
+ * a UTF-8 trail byte. ASCII is allocated linearly with 128 contiguous entries.
+ *
+ * In addition, a utf8Friendly MBCS file contains an additional
+ *   uint16_t mbcsIndex[(maxFastUChar+1)>>6];
+ * which replaces the stage 1 and 2 tables for indexing with bits from the
+ * UTF-8 lead byte and middle trail byte. Unlike the older MBCS stage 2 table,
+ * the mbcsIndex does not contain roundtrip flags. Therefore, all fallbacks
+ * from code points up to maxFastUChar (and roundtrips to 0x00) are moved to
+ * the extension data structure. This also allows for faster roundtrip
+ * conversion from UTF-16.
+ *
+ * SBCS files do not contain an additional sbcsIndex[] array because the
+ * proportional size increase would be noticeable, but the runtime
+ * code builds one for the code point range for which the runtime conversion
+ * code is optimized.
+ *
+ * For SBCS, maxFastUChar should be at least U+0FFF. The initial makeconv
+ * implementation sets it to U+1FFF. Because the sbcsIndex is not stored in
+ * the file, a larger maxFastUChar only affects stage 3 block allocation size
+ * and is free in empty blocks. (Larger blocks with sparse contents cause larger
+ * files.) U+1FFF includes almost all of the small scripts.
+ * U+0FFF covers UTF-8 two-byte sequences and three-byte sequences starting with
+ * 0xe0. This includes most scripts with legacy SBCS charsets.
+ * The initial runtime implementation using 4.3 files only builds an sbcsIndex
+ * for code points up to U+0FFF.
+ *
+ * For MBCS, maxFastUChar should be at least U+D7FF (=initial value).
+ * This boundary is convenient because practically all of the commonly used
+ * characters are below it, and because it is the boundary to surrogate
+ * code points, above which special handling is necessary anyway.
+ * (Surrogate pair assembly for UTF-16, validity checking for UTF-8.)
+ *
+ * maxFastUChar could be up to U+FFFF to cover the whole BMP, which could be
+ * useful especially for conversion from UTF-8 when the input can be assumed
+ * to be valid, because the surrogate range would then not have to be
+ * checked.
+ * (With maxFastUChar=0xffff, makeconv would have to check for mbcsIndex value
+ * overflow because with the all-unassigned block 0 and nearly full mappings
+ * from the BMP it is theoretically possible that an index into stage 3
+ * exceeds 16 bits.)
+ *
+ * _MBCSHeader.version 4.2 adds an optional conversion extension data structure.
+ * If it is present, then an ICU version reading header versions 4.0 or 4.1
+ * will be able to use the base table and ignore the extension.
+ *
+ * The unicodeMask in the static data is part of the base table data structure.
+ * Especially, the UCNV_HAS_SUPPLEMENTARY flag determines the length of the
+ * fromUnicode stage 1 array.
+ * The static data unicodeMask refers only to the base table's properties if
+ * a base table is included.
+ * In an extension-only file, the static data unicodeMask is 0.
+ * The extension data indexes have a separate field with the unicodeMask flags.
+ *
+ * MBCS-style data structure following the static data.
+ * Offsets are counted in bytes from the beginning of the MBCS header structure.
+ * Details about usage in comments in ucnvmbcs.c.
+ *
+ * struct _MBCSHeader (see the definition in this header file below)
+ * contains 32-bit fields as follows:
+ * 8 values:
+ *  0   uint8_t[4]  MBCS version in UVersionInfo format (currently 4.3.x.0)
+ *  1   uint32_t    countStates
+ *  2   uint32_t    countToUFallbacks
+ *  3   uint32_t    offsetToUCodeUnits
+ *  4   uint32_t    offsetFromUTable
+ *  5   uint32_t    offsetFromUBytes
+ *  6   uint32_t    flags, bits:
+ *                      31.. 8 offsetExtension -- _MBCSHeader.version 4.2 (ICU 2.8) and higher
+ *                                                0 for older versions and if
+ *                                                there is not extension structure
+ *                       7.. 0 outputType
+ *  7   uint32_t    fromUBytesLength -- _MBCSHeader.version 4.1 (ICU 2.4) and higher
+ *                  counts bytes in fromUBytes[]
+ *
+ * New and required in version 5:
+ *  8   uint32_t    options, bits:
+ *                      31..16 reserved for flags that can be added without breaking
+ *                                 backward compatibility
+ *                      15.. 6 reserved for flags whose addition will break
+ *                                 backward compatibility
+ *                           6 MBCS_OPT_FROM_U -- if set,
+ *                                 then most of the fromUnicode data is omitted;
+ *                                 fullStage2Length is present and the missing
+ *                                 bottom part of stage 2 must be reconstituted from
+ *                                 the toUnicode data;
+ *                                 stage 3 is missing completely as well;
+ *                                 not used for SBCS tables
+ *                       5.. 0 length of the _MBCSHeader (number of uint32_t)
+ *
+ * New and optional in version 5:
+ *  9   uint32_t    fullStage2Length: used if MBCS_OPT_FROM_U is set
+ *                                 specifies the full length of stage 2
+ *                                 including the omitted part
+ *
+ * if(outputType==MBCS_OUTPUT_EXT_ONLY) {
+ *     -- base table name for extension-only table
+ *     char baseTableName[variable]; -- with NUL plus padding for 4-alignment
+ *
+ *     -- all _MBCSHeader fields except for version and flags are 0
+ * } else {
+ *     -- normal base table with optional extension
+ *
+ *     int32_t stateTable[countStates][256];
+ *    
+ *     struct _MBCSToUFallback { (fallbacks are sorted by offset)
+ *         uint32_t offset;
+ *         UChar32 codePoint;
+ *     } toUFallbacks[countToUFallbacks];
+ *    
+ *     uint16_t unicodeCodeUnits[(offsetFromUTable-offsetToUCodeUnits)/2];
+ *                  (padded to an even number of units)
+ *    
+ *     -- stage 1 tables
+ *     if(staticData.unicodeMask&UCNV_HAS_SUPPLEMENTARY) {
+ *         -- stage 1 table for all of Unicode
+ *         uint16_t fromUTable[0x440]; (32-bit-aligned)
+ *     } else {
+ *         -- BMP-only tables have a smaller stage 1 table
+ *         uint16_t fromUTable[0x40]; (32-bit-aligned)
+ *     }
+ *
+ *     -- stage 2 tables
+ *        length determined by top of stage 1 and bottom of stage 3 tables
+ *     if(outputType==MBCS_OUTPUT_1) {
+ *         -- SBCS: pure indexes
+ *         uint16_t stage 2 indexes[?];
+ *     } else {
+ *         -- DBCS, MBCS, EBCDIC_STATEFUL, ...: roundtrip flags and indexes
+ *         uint32_t stage 2 flags and indexes[?];
+ *         if(options&MBCS_OPT_NO_FROM_U) {
+ *             stage 2 really has length fullStage2Length
+ *             and the omitted lower part must be reconstituted from
+ *             the toUnicode data
+ *         }
+ *     }
+ *
+ *     -- stage 3 tables with byte results
+ *     if(outputType==MBCS_OUTPUT_1) {
+ *         -- SBCS: each 16-bit result contains flags and the result byte, see ucnvmbcs.c
+ *         uint16_t fromUBytes[fromUBytesLength/2];
+ *     } else if(!(options&MBCS_OPT_NO_FROM_U)) {
+ *         -- DBCS, MBCS, EBCDIC_STATEFUL, ... 2/3/4 bytes result, see ucnvmbcs.c
+ *         uint8_t fromUBytes[fromUBytesLength]; or
+ *         uint16_t fromUBytes[fromUBytesLength/2]; or
+ *         uint32_t fromUBytes[fromUBytesLength/4];
+ *     } else {
+ *         fromUBytes[] must be reconstituted from the toUnicode data
+ *     }
+ *
+ *     -- optional utf8Friendly mbcsIndex -- _MBCSHeader.version 4.3 (ICU 3.8) and higher
+ *     if(outputType!=MBCS_OUTPUT_1 &&
+ *        _MBCSHeader.version[1]>=3 &&
+ *        (maxFastUChar=_MBCSHeader.version[2])!=0
+ *     ) {
+ *         maxFastUChar=(maxFastUChar<<8)|0xff;
+ *         uint16_t mbcsIndex[(maxFastUChar+1)>>6];
+ *     }
+ * }
+ *
+ * -- extension table, details see ucnv_ext.h
+ * int32_t indexes[>=32]; ...
+ */
+
+/* MBCS converter data and state -------------------------------------------- */
+
+enum {
+    MBCS_MAX_STATE_COUNT=128
+};
+
+/**
+ * MBCS action codes for conversions to Unicode.
+ * These values are in bits 23..20 of the state table entries.
+ */
+enum {
+    MBCS_STATE_VALID_DIRECT_16,
+    MBCS_STATE_VALID_DIRECT_20,
+
+    MBCS_STATE_FALLBACK_DIRECT_16,
+    MBCS_STATE_FALLBACK_DIRECT_20,
+
+    MBCS_STATE_VALID_16,
+    MBCS_STATE_VALID_16_PAIR,
+
+    MBCS_STATE_UNASSIGNED,
+    MBCS_STATE_ILLEGAL,
+
+    MBCS_STATE_CHANGE_ONLY
+};
+
+/* Macros for state table entries */
+#define MBCS_ENTRY_TRANSITION(state, offset) (int32_t)(((int32_t)(state)<<24L)|(offset))
+#define MBCS_ENTRY_TRANSITION_SET_OFFSET(entry, offset) (int32_t)(((entry)&0xff000000)|(offset))
+#define MBCS_ENTRY_TRANSITION_ADD_OFFSET(entry, offset) (int32_t)((entry)+(offset))
+
+#define MBCS_ENTRY_FINAL(state, action, value) (int32_t)(0x80000000|((int32_t)(state)<<24L)|((action)<<20L)|(value))
+#define MBCS_ENTRY_SET_FINAL(entry) (int32_t)((entry)|0x80000000)
+#define MBCS_ENTRY_FINAL_SET_ACTION(entry, action) (int32_t)(((entry)&0xff0fffff)|((int32_t)(action)<<20L))
+#define MBCS_ENTRY_FINAL_SET_VALUE(entry, value) (int32_t)(((entry)&0xfff00000)|(value))
+#define MBCS_ENTRY_FINAL_SET_ACTION_VALUE(entry, action, value) (int32_t)(((entry)&0xff000000)|((int32_t)(action)<<20L)|(value))
+
+#define MBCS_ENTRY_SET_STATE(entry, state) (int32_t)(((entry)&0x80ffffff)|((int32_t)(state)<<24L))
+
+#define MBCS_ENTRY_STATE(entry) ((((uint32_t)entry)>>24)&0x7f)
+
+#define MBCS_ENTRY_IS_TRANSITION(entry) ((entry)>=0)
+#define MBCS_ENTRY_IS_FINAL(entry) ((entry)<0)
+
+#define MBCS_ENTRY_TRANSITION_STATE(entry) (((uint32_t)entry)>>24)
+#define MBCS_ENTRY_TRANSITION_OFFSET(entry) ((entry)&0xffffff)
+
+#define MBCS_ENTRY_FINAL_STATE(entry) ((((uint32_t)entry)>>24)&0x7f)
+#define MBCS_ENTRY_FINAL_IS_VALID_DIRECT_16(entry) ((entry)<(int32_t)0x80100000)
+#define MBCS_ENTRY_FINAL_ACTION(entry) ((((uint32_t)entry)>>20)&0xf)
+#define MBCS_ENTRY_FINAL_VALUE(entry) ((entry)&0xfffff)
+#define MBCS_ENTRY_FINAL_VALUE_16(entry) (uint16_t)(entry)
+
+#define IS_ASCII_ROUNDTRIP(b, asciiRoundtrips) (((asciiRoundtrips) & (1<<((b)>>2)))!=0)
+
+/* single-byte fromUnicode: get the 16-bit result word */
+#define MBCS_SINGLE_RESULT_FROM_U(table, results, c) (results)[ (table)[ (table)[(c)>>10] +(((c)>>4)&0x3f) ] +((c)&0xf) ]
+
+/* single-byte fromUnicode using the sbcsIndex */
+#define SBCS_RESULT_FROM_LOW_BMP(table, results, c) (results)[ (table)[(c)>>6] +((c)&0x3f) ]
+
+/* single-byte fromUTF8 using the sbcsIndex; l and t must be masked externally; can be l=0 and t<=0x7f */
+#define SBCS_RESULT_FROM_UTF8(table, results, l, t) (results)[ (table)[l] +(t) ]
+
+/* multi-byte fromUnicode: get the 32-bit stage 2 entry */
+#define MBCS_STAGE_2_FROM_U(table, c) ((const uint32_t *)(table))[ (table)[(c)>>10] +(((c)>>4)&0x3f) ]
+#define MBCS_FROM_U_IS_ROUNDTRIP(stage2Entry, c) ( ((stage2Entry) & ((uint32_t)1<< (16+((c)&0xf)) )) !=0)
+
+#define MBCS_VALUE_2_FROM_STAGE_2(bytes, stage2Entry, c) ((uint16_t *)(bytes))[16*(uint32_t)(uint16_t)(stage2Entry)+((c)&0xf)]
+#define MBCS_VALUE_4_FROM_STAGE_2(bytes, stage2Entry, c) ((uint32_t *)(bytes))[16*(uint32_t)(uint16_t)(stage2Entry)+((c)&0xf)]
+
+#define MBCS_POINTER_3_FROM_STAGE_2(bytes, stage2Entry, c) ((bytes)+(16*(uint32_t)(uint16_t)(stage2Entry)+((c)&0xf))*3)
+
+/* double-byte fromUnicode using the mbcsIndex */
+#define DBCS_RESULT_FROM_MOST_BMP(table, results, c) (results)[ (table)[(c)>>6] +((c)&0x3f) ]
+
+/* double-byte fromUTF8 using the mbcsIndex; l and t1 combined into lt1; lt1 and t2 must be masked externally */
+#define DBCS_RESULT_FROM_UTF8(table, results, lt1, t2) (results)[ (table)[lt1] +(t2) ]
+
+
+/**
+ * MBCS output types for conversions from Unicode.
+ * These per-converter types determine the storage method in stage 3 of the lookup table,
+ * mostly how many bytes are stored per entry.
+ */
+enum {
+    MBCS_OUTPUT_1,          /* 0 */
+    MBCS_OUTPUT_2,          /* 1 */
+    MBCS_OUTPUT_3,          /* 2 */
+    MBCS_OUTPUT_4,          /* 3 */
+
+    MBCS_OUTPUT_3_EUC=8,    /* 8 */
+    MBCS_OUTPUT_4_EUC,      /* 9 */
+
+    MBCS_OUTPUT_2_SISO=12,  /* c */
+    MBCS_OUTPUT_2_HZ,       /* d */
+
+    MBCS_OUTPUT_EXT_ONLY,   /* e */
+
+    MBCS_OUTPUT_COUNT,
+
+    MBCS_OUTPUT_DBCS_ONLY=0xdb  /* runtime-only type for DBCS-only handling of SISO tables */
+};
+
+/**
+ * Fallbacks to Unicode are stored outside the normal state table and code point structures
+ * in a vector of items of this type. They are sorted by offset.
+ */
+typedef struct {
+    uint32_t offset;
+    UChar32 codePoint;
+} _MBCSToUFallback;
+
+/** Constants for fast and UTF-8-friendly conversion. */
+enum {
+    SBCS_FAST_MAX=0x0fff,               /* maximum code point with UTF-8-friendly SBCS runtime code, see makeconv SBCS_UTF8_MAX */
+    SBCS_FAST_LIMIT=SBCS_FAST_MAX+1,    /* =0x1000 */
+    MBCS_FAST_MAX=0xd7ff,               /* maximum code point with UTF-8-friendly MBCS runtime code, see makeconv MBCS_UTF8_MAX */
+    MBCS_FAST_LIMIT=MBCS_FAST_MAX+1     /* =0xd800 */
+};
+
+/**
+ * This is the MBCS part of the UConverterTable union (a runtime data structure).
+ * It keeps all the per-converter data and points into the loaded mapping tables.
+ *
+ * utf8Friendly data structures added with _MBCSHeader.version 4.3
+ */
+typedef struct UConverterMBCSTable {
+    /* toUnicode */
+    uint8_t countStates, dbcsOnlyState, stateTableOwned;
+    uint32_t countToUFallbacks;
+
+    const int32_t (*stateTable)/*[countStates]*/[256];
+    int32_t (*swapLFNLStateTable)/*[countStates]*/[256]; /* for swaplfnl */
+    const uint16_t *unicodeCodeUnits/*[countUnicodeResults]*/;
+    const _MBCSToUFallback *toUFallbacks;
+
+    /* fromUnicode */
+    const uint16_t *fromUnicodeTable;
+    const uint16_t *mbcsIndex;              /* for fast conversion from most of BMP to MBCS (utf8Friendly data) */
+    uint16_t sbcsIndex[SBCS_FAST_LIMIT>>6]; /* for fast conversion from low BMP to SBCS (utf8Friendly data) */
+    const uint8_t *fromUnicodeBytes;
+    uint8_t *swapLFNLFromUnicodeBytes;      /* for swaplfnl */
+    uint32_t fromUBytesLength;
+    uint8_t outputType, unicodeMask;
+    UBool utf8Friendly;                     /* for utf8Friendly data */
+    UChar maxFastUChar;                     /* for utf8Friendly data */
+
+    /* roundtrips */
+    uint32_t asciiRoundtrips;
+
+    /* reconstituted data that was omitted from the .cnv file */
+    uint8_t *reconstitutedData;
+
+    /* converter name for swaplfnl */
+    char *swapLFNLName;
+
+    /* extension data */
+    struct UConverterSharedData *baseSharedData;
+    const int32_t *extIndexes;
+} UConverterMBCSTable;
+
+#define UCNV_MBCS_TABLE_INITIALIZER { \
+    /* toUnicode */ \
+    0, 0, 0, \
+    0, \
+     \
+    NULL, \
+    NULL, \
+    NULL, \
+    NULL, \
+     \
+    /* fromUnicode */ \
+    NULL, \
+    NULL, \
+    { 0 }, \
+    NULL, \
+    NULL, \
+    0, \
+    0, 0, \
+    false, \
+    0, \
+     \
+    /* roundtrips */ \
+    0, \
+     \
+    /* reconstituted data that was omitted from the .cnv file */ \
+    NULL, \
+     \
+    /* converter name for swaplfnl */ \
+    NULL, \
+     \
+    /* extension data */ \
+    NULL, \
+    NULL \
+}
+
+enum {
+    MBCS_OPT_LENGTH_MASK=0x3f,
+    MBCS_OPT_NO_FROM_U=0x40,
+    /*
+     * If any of the following options bits are set,
+     * then the file must be rejected.
+     */
+    MBCS_OPT_INCOMPATIBLE_MASK=0xffc0,
+    /*
+     * Remove bits from this mask as more options are recognized
+     * by all implementations that use this constant.
+     */
+    MBCS_OPT_UNKNOWN_INCOMPATIBLE_MASK=0xff80
+};
+
+enum {
+    MBCS_HEADER_V4_LENGTH=8,
+    MBCS_HEADER_V5_MIN_LENGTH=9
+};
+
+/**
+ * MBCS data header. See data format description above.
+ */
+typedef struct {
+    UVersionInfo version;
+    uint32_t countStates,
+             countToUFallbacks,
+             offsetToUCodeUnits,
+             offsetFromUTable,
+             offsetFromUBytes,
+             flags,
+             fromUBytesLength;
+
+    /* new and required in version 5 */
+    uint32_t options;
+
+    /* new and optional in version 5; used if options&MBCS_OPT_NO_FROM_U */
+    uint32_t fullStage2Length;  /* number of 32-bit units */
+} _MBCSHeader;
+
+#define UCNV_MBCS_HEADER_INITIALIZER { { 0 },  0, 0, 0, 0, 0, 0, 0,  0,  0 }
+
+/*
+ * This is a simple version of _MBCSGetNextUChar() that is used
+ * by other converter implementations.
+ * It only returns an "assigned" result if it consumes the entire input.
+ * It does not use state from the converter, nor error codes.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It handles conversion extensions but not GB 18030.
+ *
+ * Return value:
+ * U+fffe   unassigned
+ * U+ffff   illegal
+ * otherwise the Unicode code point
+ */
+U_CFUNC UChar32
+ucnv_MBCSSimpleGetNextUChar(UConverterSharedData *sharedData,
+                        const char *source, int32_t length,
+                        UBool useFallback);
+
+/**
+ * This version of _MBCSSimpleGetNextUChar() is optimized for single-byte, single-state codepages.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It does not handle conversion extensions (_extToU()).
+ */
+U_CFUNC UChar32
+ucnv_MBCSSingleSimpleGetNextUChar(UConverterSharedData *sharedData,
+                              uint8_t b, UBool useFallback);
+
+/**
+ * This macro version of _MBCSSingleSimpleGetNextUChar() gets a code point from a byte.
+ * It works for single-byte, single-state codepages that only map
+ * to and from BMP code points, and it always
+ * returns fallback values.
+ */
+#define _MBCS_SINGLE_SIMPLE_GET_NEXT_BMP(sharedData, b) \
+    (UChar)MBCS_ENTRY_FINAL_VALUE_16((sharedData)->mbcs.stateTable[0][(uint8_t)(b)])
+
+/**
+ * This is an internal function that allows other converter implementations
+ * to check whether a byte is a lead byte.
+ */
+U_CFUNC UBool
+ucnv_MBCSIsLeadByte(UConverterSharedData *sharedData, char byte);
+
+/** This is a macro version of _MBCSIsLeadByte(). */
+#define _MBCS_IS_LEAD_BYTE(sharedData, byte) \
+    (UBool)MBCS_ENTRY_IS_TRANSITION((sharedData)->mbcs.stateTable[0][(uint8_t)(byte)])
+
+/*
+ * This is another simple conversion function for internal use by other
+ * conversion implementations.
+ * It does not use the converter state nor call callbacks.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ * It handles conversion extensions but not GB 18030.
+ *
+ * It converts one single Unicode code point into codepage bytes, encoded
+ * as one 32-bit value. The function returns the number of bytes in *pValue:
+ * 1..4 the number of bytes in *pValue
+ * 0    unassigned (*pValue undefined)
+ * -1   illegal (currently not used, *pValue undefined)
+ *
+ * *pValue will contain the resulting bytes with the last byte in bits 7..0,
+ * the second to last byte in bits 15..8, etc.
+ * Currently, the function assumes but does not check that 0<=c<=0x10ffff.
+ */
+U_CFUNC int32_t
+ucnv_MBCSFromUChar32(UConverterSharedData *sharedData,
+                 UChar32 c, uint32_t *pValue,
+                 UBool useFallback);
+
+/**
+ * This version of _MBCSFromUChar32() is optimized for single-byte codepages.
+ * It does not handle the EBCDIC swaplfnl option (set in UConverter).
+ *
+ * It returns the codepage byte for the code point, or -1 if it is unassigned.
+ */
+U_CFUNC int32_t
+ucnv_MBCSSingleFromUChar32(UConverterSharedData *sharedData,
+                       UChar32 c,
+                       UBool useFallback);
+
+/**
+ * SBCS, DBCS, and EBCDIC_STATEFUL are replaced by MBCS, but
+ * we cheat a little about the type, returning the old types if appropriate.
+ */
+U_CFUNC UConverterType
+ucnv_MBCSGetType(const UConverter* converter);
+
+U_CFUNC void 
+ucnv_MBCSFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode);
+U_CFUNC void 
+ucnv_MBCSToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                          UErrorCode *pErrorCode);
+
+/*
+ * Internal function returning a UnicodeSet for toUnicode() conversion.
+ * Currently only used for ISO-2022-CN, and only handles roundtrip mappings.
+ * In the future, if we add support for fallback sets, this function
+ * needs to be updated.
+ * Handles extensions.
+ * Does not empty the set first.
+ */
+U_CFUNC void
+ucnv_MBCSGetUnicodeSetForUnicode(const UConverterSharedData *sharedData,
+                                 const USetAdder *sa,
+                                 UConverterUnicodeSet which,
+                                 UErrorCode *pErrorCode);
+
+/*
+ * Same as ucnv_MBCSGetUnicodeSetForUnicode() but
+ * the set can be filtered by encoding scheme.
+ * Used by stateful converters which share regular conversion tables
+ * but only use a subset of their mappings.
+ */
+U_CFUNC void
+ucnv_MBCSGetFilteredUnicodeSetForUnicode(const UConverterSharedData *sharedData,
+                                         const USetAdder *sa,
+                                         UConverterUnicodeSet which,
+                                         UConverterSetFilter filter,
+                                         UErrorCode *pErrorCode);
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnvscsu.cpp b/thirdparty/icu4c/common/ucnvscsu.cpp
new file mode 100644
index 0000000000..74b5722b97
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvscsu.cpp
@@ -0,0 +1,2045 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ucnvscsu.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000nov18
+*   created by: Markus W. Scherer
+*
+*   This is an implementation of the Standard Compression Scheme for Unicode
+*   as defined in http://www.unicode.org/unicode/reports/tr6/ .
+*   Reserved commands and window settings are treated as illegal sequences and
+*   will result in callback calls.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION && !UCONFIG_ONLY_HTML_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_cb.h"
+#include "unicode/utf16.h"
+#include "ucnv_bld.h"
+#include "ucnv_cnv.h"
+#include "cmemory.h"
+
+/* SCSU definitions --------------------------------------------------------- */
+
+/* SCSU command byte values */
+enum {
+    SQ0=0x01, /* Quote from window pair 0 */
+    SQ7=0x08, /* Quote from window pair 7 */
+    SDX=0x0B, /* Define a window as extended */
+    Srs=0x0C, /* reserved */
+    SQU=0x0E, /* Quote a single Unicode character */
+    SCU=0x0F, /* Change to Unicode mode */
+    SC0=0x10, /* Select window 0 */
+    SC7=0x17, /* Select window 7 */
+    SD0=0x18, /* Define and select window 0 */
+    SD7=0x1F, /* Define and select window 7 */
+
+    UC0=0xE0, /* Select window 0 */
+    UC7=0xE7, /* Select window 7 */
+    UD0=0xE8, /* Define and select window 0 */
+    UD7=0xEF, /* Define and select window 7 */
+    UQU=0xF0, /* Quote a single Unicode character */
+    UDX=0xF1, /* Define a Window as extended */
+    Urs=0xF2  /* reserved */
+};
+
+enum {
+    /*
+     * Unicode code points from 3400 to E000 are not adressible by
+     * dynamic window, since in these areas no short run alphabets are
+     * found. Therefore add gapOffset to all values from gapThreshold.
+     */
+    gapThreshold=0x68,
+    gapOffset=0xAC00,
+
+    /* values between reservedStart and fixedThreshold are reserved */
+    reservedStart=0xA8,
+
+    /* use table of predefined fixed offsets for values from fixedThreshold */
+    fixedThreshold=0xF9
+};
+
+/* constant offsets for the 8 static windows */
+static const uint32_t staticOffsets[8]={
+    0x0000, /* ASCII for quoted tags */
+    0x0080, /* Latin - 1 Supplement (for access to punctuation) */
+    0x0100, /* Latin Extended-A */
+    0x0300, /* Combining Diacritical Marks */
+    0x2000, /* General Punctuation */
+    0x2080, /* Currency Symbols */
+    0x2100, /* Letterlike Symbols and Number Forms */
+    0x3000  /* CJK Symbols and punctuation */
+};
+
+/* initial offsets for the 8 dynamic (sliding) windows */
+static const uint32_t initialDynamicOffsets[8]={
+    0x0080, /* Latin-1 */
+    0x00C0, /* Latin Extended A */
+    0x0400, /* Cyrillic */
+    0x0600, /* Arabic */
+    0x0900, /* Devanagari */
+    0x3040, /* Hiragana */
+    0x30A0, /* Katakana */
+    0xFF00  /* Fullwidth ASCII */
+};
+
+/* Table of fixed predefined Offsets */
+static const uint32_t fixedOffsets[]={
+    /* 0xF9 */ 0x00C0, /* Latin-1 Letters + half of Latin Extended A */
+    /* 0xFA */ 0x0250, /* IPA extensions */
+    /* 0xFB */ 0x0370, /* Greek */
+    /* 0xFC */ 0x0530, /* Armenian */
+    /* 0xFD */ 0x3040, /* Hiragana */
+    /* 0xFE */ 0x30A0, /* Katakana */
+    /* 0xFF */ 0xFF60  /* Halfwidth Katakana */
+};
+
+/* state values */
+enum {
+    readCommand,
+    quotePairOne,
+    quotePairTwo,
+    quoteOne,
+    definePairOne,
+    definePairTwo,
+    defineOne
+};
+
+typedef struct SCSUData {
+    /* dynamic window offsets, intitialize to default values from initialDynamicOffsets */
+    uint32_t toUDynamicOffsets[8];
+    uint32_t fromUDynamicOffsets[8];
+
+    /* state machine state - toUnicode */
+    UBool toUIsSingleByteMode;
+    uint8_t toUState;
+    int8_t toUQuoteWindow, toUDynamicWindow;
+    uint8_t toUByteOne;
+    uint8_t toUPadding[3];
+
+    /* state machine state - fromUnicode */
+    UBool fromUIsSingleByteMode;
+    int8_t fromUDynamicWindow;
+
+    /*
+     * windowUse[] keeps track of the use of the dynamic windows:
+     * At nextWindowUseIndex there is the least recently used window,
+     * and the following windows (in a wrapping manner) are more and more
+     * recently used.
+     * At nextWindowUseIndex-1 there is the most recently used window.
+     */
+    uint8_t locale;
+    int8_t nextWindowUseIndex;
+    int8_t windowUse[8];
+} SCSUData;
+
+static const int8_t initialWindowUse[8]={ 7, 0, 3, 2, 4, 5, 6, 1 };
+static const int8_t initialWindowUse_ja[8]={ 3, 2, 4, 1, 0, 7, 5, 6 };
+
+enum {
+    lGeneric, l_ja
+};
+
+/* SCSU setup functions ----------------------------------------------------- */
+U_CDECL_BEGIN
+static void U_CALLCONV
+_SCSUReset(UConverter *cnv, UConverterResetChoice choice) {
+    SCSUData *scsu=(SCSUData *)cnv->extraInfo;
+
+    if(choice<=UCNV_RESET_TO_UNICODE) {
+        /* reset toUnicode */
+        uprv_memcpy(scsu->toUDynamicOffsets, initialDynamicOffsets, 32);
+
+        scsu->toUIsSingleByteMode=TRUE;
+        scsu->toUState=readCommand;
+        scsu->toUQuoteWindow=scsu->toUDynamicWindow=0;
+        scsu->toUByteOne=0;
+
+        cnv->toULength=0;
+    }
+    if(choice!=UCNV_RESET_TO_UNICODE) {
+        /* reset fromUnicode */
+        uprv_memcpy(scsu->fromUDynamicOffsets, initialDynamicOffsets, 32);
+
+        scsu->fromUIsSingleByteMode=TRUE;
+        scsu->fromUDynamicWindow=0;
+
+        scsu->nextWindowUseIndex=0;
+        switch(scsu->locale) {
+        case l_ja:
+            uprv_memcpy(scsu->windowUse, initialWindowUse_ja, 8);
+            break;
+        default:
+            uprv_memcpy(scsu->windowUse, initialWindowUse, 8);
+            break;
+        }
+
+        cnv->fromUChar32=0;
+    }
+}
+
+static void U_CALLCONV
+_SCSUOpen(UConverter *cnv,
+          UConverterLoadArgs *pArgs,
+          UErrorCode *pErrorCode) {
+    const char *locale=pArgs->locale;
+    if(pArgs->onlyTestIsLoadable) {
+        return;
+    }
+    cnv->extraInfo=uprv_malloc(sizeof(SCSUData));
+    if(cnv->extraInfo!=NULL) {
+        if(locale!=NULL && locale[0]=='j' && locale[1]=='a' && (locale[2]==0 || locale[2]=='_')) {
+            ((SCSUData *)cnv->extraInfo)->locale=l_ja;
+        } else {
+            ((SCSUData *)cnv->extraInfo)->locale=lGeneric;
+        }
+        _SCSUReset(cnv, UCNV_RESET_BOTH);
+    } else {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+    }
+
+    /* Set the substitution character U+fffd as a Unicode string. */
+    cnv->subUChars[0]=0xfffd;
+    cnv->subCharLen=-1;
+}
+
+static void U_CALLCONV
+_SCSUClose(UConverter *cnv) {
+    if(cnv->extraInfo!=NULL) {
+        if(!cnv->isExtraLocal) {
+            uprv_free(cnv->extraInfo);
+        }
+        cnv->extraInfo=NULL;
+    }
+}
+
+/* SCSU-to-Unicode conversion functions ------------------------------------- */
+
+static void U_CALLCONV
+_SCSUToUnicodeWithOffsets(UConverterToUnicodeArgs *pArgs,
+                          UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    SCSUData *scsu;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    int32_t *offsets;
+    UBool isSingleByteMode;
+    uint8_t state, byteOne;
+    int8_t quoteWindow, dynamicWindow;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    uint8_t b;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    scsu=(SCSUData *)cnv->extraInfo;
+
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+    offsets=pArgs->offsets;
+
+    /* get the state machine state */
+    isSingleByteMode=scsu->toUIsSingleByteMode;
+    state=scsu->toUState;
+    quoteWindow=scsu->toUQuoteWindow;
+    dynamicWindow=scsu->toUDynamicWindow;
+    byteOne=scsu->toUByteOne;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex=state==readCommand ? 0 : -1;
+    nextSourceIndex=0;
+
+    /*
+     * conversion "loop"
+     *
+     * For performance, this is not a normal C loop.
+     * Instead, there are two code blocks for the two SCSU modes.
+     * The function branches to either one, and a change of the mode is done with a goto to
+     * the other branch.
+     *
+     * Each branch has two conventional loops:
+     * - a fast-path loop for the most common codes in the mode
+     * - a loop for all other codes in the mode
+     * When the fast-path runs into a code that it cannot handle, its loop ends and it
+     * runs into the following loop to handle the other codes.
+     * The end of the input or output buffer is also handled by the slower loop.
+     * The slow loop jumps (goto) to the fast-path loop again as soon as possible.
+     *
+     * The callback handling is done by returning with an error code.
+     * The conversion framework actually calls the callback function.
+     */
+    if(isSingleByteMode) {
+        /* fast path for single-byte mode */
+        if(state==readCommand) {
+fastSingle:
+            while(source<sourceLimit && target<targetLimit && (b=*source)>=0x20) {
+                ++source;
+                ++nextSourceIndex;
+                if(b<=0x7f) {
+                    /* write US-ASCII graphic character or DEL */
+                    *target++=(UChar)b;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                } else {
+                    /* write from dynamic window */
+                    uint32_t c=scsu->toUDynamicOffsets[dynamicWindow]+(b&0x7f);
+                    if(c<=0xffff) {
+                        *target++=(UChar)c;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                        }
+                    } else {
+                        /* output surrogate pair */
+                        *target++=(UChar)(0xd7c0+(c>>10));
+                        if(target<targetLimit) {
+                            *target++=(UChar)(0xdc00|(c&0x3ff));
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                *offsets++=sourceIndex;
+                            }
+                        } else {
+                            /* target overflow */
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                            }
+                            cnv->UCharErrorBuffer[0]=(UChar)(0xdc00|(c&0x3ff));
+                            cnv->UCharErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            goto endloop;
+                        }
+                    }
+                }
+                sourceIndex=nextSourceIndex;
+            }
+        }
+
+        /* normal state machine for single-byte mode, minus handling for what fastSingle covers */
+singleByteMode:
+        while(source<sourceLimit) {
+            if(target>=targetLimit) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            b=*source++;
+            ++nextSourceIndex;
+            switch(state) {
+            case readCommand:
+                /* redundant conditions are commented out */
+                /* here: b<0x20 because otherwise we would be in fastSingle */
+                if((1UL<<b)&0x2601 /* binary 0010 0110 0000 0001, check for b==0xd || b==0xa || b==9 || b==0 */) {
+                    /* CR/LF/TAB/NUL */
+                    *target++=(UChar)b;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    sourceIndex=nextSourceIndex;
+                    goto fastSingle;
+                } else if(SC0<=b) {
+                    if(b<=SC7) {
+                        dynamicWindow=(int8_t)(b-SC0);
+                        sourceIndex=nextSourceIndex;
+                        goto fastSingle;
+                    } else /* if(SD0<=b && b<=SD7) */ {
+                        dynamicWindow=(int8_t)(b-SD0);
+                        state=defineOne;
+                    }
+                } else if(/* SQ0<=b && */ b<=SQ7) {
+                    quoteWindow=(int8_t)(b-SQ0);
+                    state=quoteOne;
+                } else if(b==SDX) {
+                    state=definePairOne;
+                } else if(b==SQU) {
+                    state=quotePairOne;
+                } else if(b==SCU) {
+                    sourceIndex=nextSourceIndex;
+                    isSingleByteMode=FALSE;
+                    goto fastUnicode;
+                } else /* Srs */ {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    goto endloop;
+                }
+
+                /* store the first byte of a multibyte sequence in toUBytes[] */
+                cnv->toUBytes[0]=b;
+                cnv->toULength=1;
+                break;
+            case quotePairOne:
+                byteOne=b;
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=quotePairTwo;
+                break;
+            case quotePairTwo:
+                *target++=(UChar)((byteOne<<8)|b);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                sourceIndex=nextSourceIndex;
+                state=readCommand;
+                goto fastSingle;
+            case quoteOne:
+                if(b<0x80) {
+                    /* all static offsets are in the BMP */
+                    *target++=(UChar)(staticOffsets[quoteWindow]+b);
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                } else {
+                    /* write from dynamic window */
+                    uint32_t c=scsu->toUDynamicOffsets[quoteWindow]+(b&0x7f);
+                    if(c<=0xffff) {
+                        *target++=(UChar)c;
+                        if(offsets!=NULL) {
+                            *offsets++=sourceIndex;
+                        }
+                    } else {
+                        /* output surrogate pair */
+                        *target++=(UChar)(0xd7c0+(c>>10));
+                        if(target<targetLimit) {
+                            *target++=(UChar)(0xdc00|(c&0x3ff));
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                                *offsets++=sourceIndex;
+                            }
+                        } else {
+                            /* target overflow */
+                            if(offsets!=NULL) {
+                                *offsets++=sourceIndex;
+                            }
+                            cnv->UCharErrorBuffer[0]=(UChar)(0xdc00|(c&0x3ff));
+                            cnv->UCharErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            goto endloop;
+                        }
+                    }
+                }
+                sourceIndex=nextSourceIndex;
+                state=readCommand;
+                goto fastSingle;
+            case definePairOne:
+                dynamicWindow=(int8_t)((b>>5)&7);
+                byteOne=(uint8_t)(b&0x1f);
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=definePairTwo;
+                break;
+            case definePairTwo:
+                scsu->toUDynamicOffsets[dynamicWindow]=0x10000+(byteOne<<15UL | b<<7UL);
+                sourceIndex=nextSourceIndex;
+                state=readCommand;
+                goto fastSingle;
+            case defineOne:
+                if(b==0) {
+                    /* callback(illegal): Reserved window offset value 0 */
+                    cnv->toUBytes[1]=b;
+                    cnv->toULength=2;
+                    goto endloop;
+                } else if(b<gapThreshold) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=b<<7UL;
+                } else if((uint8_t)(b-gapThreshold)<(reservedStart-gapThreshold)) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=(b<<7UL)+gapOffset;
+                } else if(b>=fixedThreshold) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=fixedOffsets[b-fixedThreshold];
+                } else {
+                    /* callback(illegal): Reserved window offset value 0xa8..0xf8 */
+                    cnv->toUBytes[1]=b;
+                    cnv->toULength=2;
+                    goto endloop;
+                }
+                sourceIndex=nextSourceIndex;
+                state=readCommand;
+                goto fastSingle;
+            }
+        }
+    } else {
+        /* fast path for Unicode mode */
+        if(state==readCommand) {
+fastUnicode:
+            while(source+1<sourceLimit && target<targetLimit && (uint8_t)((b=*source)-UC0)>(Urs-UC0)) {
+                *target++=(UChar)((b<<8)|source[1]);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                sourceIndex=nextSourceIndex;
+                nextSourceIndex+=2;
+                source+=2;
+            }
+        }
+
+        /* normal state machine for Unicode mode */
+/* unicodeByteMode: */
+        while(source<sourceLimit) {
+            if(target>=targetLimit) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            b=*source++;
+            ++nextSourceIndex;
+            switch(state) {
+            case readCommand:
+                if((uint8_t)(b-UC0)>(Urs-UC0)) {
+                    byteOne=b;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=quotePairTwo;
+                } else if(/* UC0<=b && */ b<=UC7) {
+                    dynamicWindow=(int8_t)(b-UC0);
+                    sourceIndex=nextSourceIndex;
+                    isSingleByteMode=TRUE;
+                    goto fastSingle;
+                } else if(/* UD0<=b && */ b<=UD7) {
+                    dynamicWindow=(int8_t)(b-UD0);
+                    isSingleByteMode=TRUE;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=defineOne;
+                    goto singleByteMode;
+                } else if(b==UDX) {
+                    isSingleByteMode=TRUE;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=definePairOne;
+                    goto singleByteMode;
+                } else if(b==UQU) {
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=quotePairOne;
+                } else /* Urs */ {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    goto endloop;
+                }
+                break;
+            case quotePairOne:
+                byteOne=b;
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=quotePairTwo;
+                break;
+            case quotePairTwo:
+                *target++=(UChar)((byteOne<<8)|b);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                sourceIndex=nextSourceIndex;
+                state=readCommand;
+                goto fastUnicode;
+            }
+        }
+    }
+endloop:
+
+    /* set the converter state back into UConverter */
+    if(U_FAILURE(*pErrorCode) && *pErrorCode!=U_BUFFER_OVERFLOW_ERROR) {
+        /* reset to deal with the next character */
+        state=readCommand;
+    } else if(state==readCommand) {
+        /* not in a multi-byte sequence, reset toULength */
+        cnv->toULength=0;
+    }
+    scsu->toUIsSingleByteMode=isSingleByteMode;
+    scsu->toUState=state;
+    scsu->toUQuoteWindow=quoteWindow;
+    scsu->toUDynamicWindow=dynamicWindow;
+    scsu->toUByteOne=byteOne;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    pArgs->offsets=offsets;
+    return;
+}
+
+/*
+ * Identical to _SCSUToUnicodeWithOffsets but without offset handling.
+ * If a change is made in the original function, then either
+ * change this function the same way or
+ * re-copy the original function and remove the variables
+ * offsets, sourceIndex, and nextSourceIndex.
+ */
+static void U_CALLCONV
+_SCSUToUnicode(UConverterToUnicodeArgs *pArgs,
+               UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    SCSUData *scsu;
+    const uint8_t *source, *sourceLimit;
+    UChar *target;
+    const UChar *targetLimit;
+    UBool isSingleByteMode;
+    uint8_t state, byteOne;
+    int8_t quoteWindow, dynamicWindow;
+
+    uint8_t b;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    scsu=(SCSUData *)cnv->extraInfo;
+
+    source=(const uint8_t *)pArgs->source;
+    sourceLimit=(const uint8_t *)pArgs->sourceLimit;
+    target=pArgs->target;
+    targetLimit=pArgs->targetLimit;
+
+    /* get the state machine state */
+    isSingleByteMode=scsu->toUIsSingleByteMode;
+    state=scsu->toUState;
+    quoteWindow=scsu->toUQuoteWindow;
+    dynamicWindow=scsu->toUDynamicWindow;
+    byteOne=scsu->toUByteOne;
+
+    /*
+     * conversion "loop"
+     *
+     * For performance, this is not a normal C loop.
+     * Instead, there are two code blocks for the two SCSU modes.
+     * The function branches to either one, and a change of the mode is done with a goto to
+     * the other branch.
+     *
+     * Each branch has two conventional loops:
+     * - a fast-path loop for the most common codes in the mode
+     * - a loop for all other codes in the mode
+     * When the fast-path runs into a code that it cannot handle, its loop ends and it
+     * runs into the following loop to handle the other codes.
+     * The end of the input or output buffer is also handled by the slower loop.
+     * The slow loop jumps (goto) to the fast-path loop again as soon as possible.
+     *
+     * The callback handling is done by returning with an error code.
+     * The conversion framework actually calls the callback function.
+     */
+    if(isSingleByteMode) {
+        /* fast path for single-byte mode */
+        if(state==readCommand) {
+fastSingle:
+            while(source<sourceLimit && target<targetLimit && (b=*source)>=0x20) {
+                ++source;
+                if(b<=0x7f) {
+                    /* write US-ASCII graphic character or DEL */
+                    *target++=(UChar)b;
+                } else {
+                    /* write from dynamic window */
+                    uint32_t c=scsu->toUDynamicOffsets[dynamicWindow]+(b&0x7f);
+                    if(c<=0xffff) {
+                        *target++=(UChar)c;
+                    } else {
+                        /* output surrogate pair */
+                        *target++=(UChar)(0xd7c0+(c>>10));
+                        if(target<targetLimit) {
+                            *target++=(UChar)(0xdc00|(c&0x3ff));
+                        } else {
+                            /* target overflow */
+                            cnv->UCharErrorBuffer[0]=(UChar)(0xdc00|(c&0x3ff));
+                            cnv->UCharErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            goto endloop;
+                        }
+                    }
+                }
+            }
+        }
+
+        /* normal state machine for single-byte mode, minus handling for what fastSingle covers */
+singleByteMode:
+        while(source<sourceLimit) {
+            if(target>=targetLimit) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            b=*source++;
+            switch(state) {
+            case readCommand:
+                /* redundant conditions are commented out */
+                /* here: b<0x20 because otherwise we would be in fastSingle */
+                if((1UL<<b)&0x2601 /* binary 0010 0110 0000 0001, check for b==0xd || b==0xa || b==9 || b==0 */) {
+                    /* CR/LF/TAB/NUL */
+                    *target++=(UChar)b;
+                    goto fastSingle;
+                } else if(SC0<=b) {
+                    if(b<=SC7) {
+                        dynamicWindow=(int8_t)(b-SC0);
+                        goto fastSingle;
+                    } else /* if(SD0<=b && b<=SD7) */ {
+                        dynamicWindow=(int8_t)(b-SD0);
+                        state=defineOne;
+                    }
+                } else if(/* SQ0<=b && */ b<=SQ7) {
+                    quoteWindow=(int8_t)(b-SQ0);
+                    state=quoteOne;
+                } else if(b==SDX) {
+                    state=definePairOne;
+                } else if(b==SQU) {
+                    state=quotePairOne;
+                } else if(b==SCU) {
+                    isSingleByteMode=FALSE;
+                    goto fastUnicode;
+                } else /* Srs */ {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    goto endloop;
+                }
+
+                /* store the first byte of a multibyte sequence in toUBytes[] */
+                cnv->toUBytes[0]=b;
+                cnv->toULength=1;
+                break;
+            case quotePairOne:
+                byteOne=b;
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=quotePairTwo;
+                break;
+            case quotePairTwo:
+                *target++=(UChar)((byteOne<<8)|b);
+                state=readCommand;
+                goto fastSingle;
+            case quoteOne:
+                if(b<0x80) {
+                    /* all static offsets are in the BMP */
+                    *target++=(UChar)(staticOffsets[quoteWindow]+b);
+                } else {
+                    /* write from dynamic window */
+                    uint32_t c=scsu->toUDynamicOffsets[quoteWindow]+(b&0x7f);
+                    if(c<=0xffff) {
+                        *target++=(UChar)c;
+                    } else {
+                        /* output surrogate pair */
+                        *target++=(UChar)(0xd7c0+(c>>10));
+                        if(target<targetLimit) {
+                            *target++=(UChar)(0xdc00|(c&0x3ff));
+                        } else {
+                            /* target overflow */
+                            cnv->UCharErrorBuffer[0]=(UChar)(0xdc00|(c&0x3ff));
+                            cnv->UCharErrorBufferLength=1;
+                            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                            goto endloop;
+                        }
+                    }
+                }
+                state=readCommand;
+                goto fastSingle;
+            case definePairOne:
+                dynamicWindow=(int8_t)((b>>5)&7);
+                byteOne=(uint8_t)(b&0x1f);
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=definePairTwo;
+                break;
+            case definePairTwo:
+                scsu->toUDynamicOffsets[dynamicWindow]=0x10000+(byteOne<<15UL | b<<7UL);
+                state=readCommand;
+                goto fastSingle;
+            case defineOne:
+                if(b==0) {
+                    /* callback(illegal): Reserved window offset value 0 */
+                    cnv->toUBytes[1]=b;
+                    cnv->toULength=2;
+                    goto endloop;
+                } else if(b<gapThreshold) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=b<<7UL;
+                } else if((uint8_t)(b-gapThreshold)<(reservedStart-gapThreshold)) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=(b<<7UL)+gapOffset;
+                } else if(b>=fixedThreshold) {
+                    scsu->toUDynamicOffsets[dynamicWindow]=fixedOffsets[b-fixedThreshold];
+                } else {
+                    /* callback(illegal): Reserved window offset value 0xa8..0xf8 */
+                    cnv->toUBytes[1]=b;
+                    cnv->toULength=2;
+                    goto endloop;
+                }
+                state=readCommand;
+                goto fastSingle;
+            }
+        }
+    } else {
+        /* fast path for Unicode mode */
+        if(state==readCommand) {
+fastUnicode:
+            while(source+1<sourceLimit && target<targetLimit && (uint8_t)((b=*source)-UC0)>(Urs-UC0)) {
+                *target++=(UChar)((b<<8)|source[1]);
+                source+=2;
+            }
+        }
+
+        /* normal state machine for Unicode mode */
+/* unicodeByteMode: */
+        while(source<sourceLimit) {
+            if(target>=targetLimit) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            b=*source++;
+            switch(state) {
+            case readCommand:
+                if((uint8_t)(b-UC0)>(Urs-UC0)) {
+                    byteOne=b;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=quotePairTwo;
+                } else if(/* UC0<=b && */ b<=UC7) {
+                    dynamicWindow=(int8_t)(b-UC0);
+                    isSingleByteMode=TRUE;
+                    goto fastSingle;
+                } else if(/* UD0<=b && */ b<=UD7) {
+                    dynamicWindow=(int8_t)(b-UD0);
+                    isSingleByteMode=TRUE;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=defineOne;
+                    goto singleByteMode;
+                } else if(b==UDX) {
+                    isSingleByteMode=TRUE;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=definePairOne;
+                    goto singleByteMode;
+                } else if(b==UQU) {
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    state=quotePairOne;
+                } else /* Urs */ {
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    cnv->toUBytes[0]=b;
+                    cnv->toULength=1;
+                    goto endloop;
+                }
+                break;
+            case quotePairOne:
+                byteOne=b;
+                cnv->toUBytes[1]=b;
+                cnv->toULength=2;
+                state=quotePairTwo;
+                break;
+            case quotePairTwo:
+                *target++=(UChar)((byteOne<<8)|b);
+                state=readCommand;
+                goto fastUnicode;
+            }
+        }
+    }
+endloop:
+
+    /* set the converter state back into UConverter */
+    if(U_FAILURE(*pErrorCode) && *pErrorCode!=U_BUFFER_OVERFLOW_ERROR) {
+        /* reset to deal with the next character */
+        state=readCommand;
+    } else if(state==readCommand) {
+        /* not in a multi-byte sequence, reset toULength */
+        cnv->toULength=0;
+    }
+    scsu->toUIsSingleByteMode=isSingleByteMode;
+    scsu->toUState=state;
+    scsu->toUQuoteWindow=quoteWindow;
+    scsu->toUDynamicWindow=dynamicWindow;
+    scsu->toUByteOne=byteOne;
+
+    /* write back the updated pointers */
+    pArgs->source=(const char *)source;
+    pArgs->target=target;
+    return;
+}
+U_CDECL_END
+/* SCSU-from-Unicode conversion functions ----------------------------------- */
+
+/*
+ * This SCSU Encoder is fairly simple but uses all SCSU commands to achieve
+ * reasonable results. The lookahead is minimal.
+ * Many cases are simple:
+ * A character fits directly into the current mode, a dynamic or static window,
+ * or is not compressible. These cases are tested first.
+ * Real compression heuristics are applied to the rest, in code branches for
+ * single/Unicode mode and BMP/supplementary code points.
+ * The heuristics used here are extremely simple.
+ */
+
+/* get the number of the window that this character is in, or -1 */
+static int8_t
+getWindow(const uint32_t offsets[8], uint32_t c) {
+    int i;
+    for(i=0; i<8; ++i) {
+        if((uint32_t)(c-offsets[i])<=0x7f) {
+            return (int8_t)(i);
+        }
+    }
+    return -1;
+}
+
+/* is the character in the dynamic window starting at the offset, or in the direct-encoded range? */
+static UBool
+isInOffsetWindowOrDirect(uint32_t offset, uint32_t c) {
+    return (UBool)(c<=offset+0x7f &&
+          (c>=offset || (c<=0x7f &&
+                        (c>=0x20 || (1UL<<c)&0x2601))));
+                                /* binary 0010 0110 0000 0001,
+                                   check for b==0xd || b==0xa || b==9 || b==0 */
+}
+
+/*
+ * getNextDynamicWindow returns the next dynamic window to be redefined
+ */
+static int8_t
+getNextDynamicWindow(SCSUData *scsu) {
+    int8_t window=scsu->windowUse[scsu->nextWindowUseIndex];
+    if(++scsu->nextWindowUseIndex==8) {
+        scsu->nextWindowUseIndex=0;
+    }
+    return window;
+}
+
+/*
+ * useDynamicWindow() adjusts
+ * windowUse[] and nextWindowUseIndex for the algorithm to choose
+ * the next dynamic window to be defined;
+ * a subclass may override it and provide its own algorithm.
+ */
+static void
+useDynamicWindow(SCSUData *scsu, int8_t window) {
+    /*
+     * move the existing window, which just became the most recently used one,
+     * up in windowUse[] to nextWindowUseIndex-1
+     */
+
+    /* first, find the index of the window - backwards to favor the more recently used windows */
+    int i, j;
+
+    i=scsu->nextWindowUseIndex;
+    do {
+        if(--i<0) {
+            i=7;
+        }
+    } while(scsu->windowUse[i]!=window);
+
+    /* now copy each windowUse[i+1] to [i] */
+    j=i+1;
+    if(j==8) {
+        j=0;
+    }
+    while(j!=scsu->nextWindowUseIndex) {
+        scsu->windowUse[i]=scsu->windowUse[j];
+        i=j;
+        if(++j==8) { j=0; }
+    }
+
+    /* finally, set the window into the most recently used index */
+    scsu->windowUse[i]=window;
+}
+
+/*
+ * calculate the offset and the code for a dynamic window that contains the character
+ * takes fixed offsets into account
+ * the offset of the window is stored in the offset variable,
+ * the code is returned
+ *
+ * return offset code: -1 none  <=0xff code for SDn/UDn  else code for SDX/UDX, subtract 0x200 to get the true code
+ */
+static int
+getDynamicOffset(uint32_t c, uint32_t *pOffset) {
+    int i;
+
+    for(i=0; i<7; ++i) {
+        if((uint32_t)(c-fixedOffsets[i])<=0x7f) {
+            *pOffset=fixedOffsets[i];
+            return 0xf9+i;
+        }
+    }
+
+    if(c<0x80) {
+        /* No dynamic window for US-ASCII. */
+        return -1;
+    } else if(c<0x3400 ||
+              (uint32_t)(c-0x10000)<(0x14000-0x10000) ||
+              (uint32_t)(c-0x1d000)<=(0x1ffff-0x1d000)
+    ) {
+        /* This character is in a code range for a "small", i.e., reasonably windowable, script. */
+        *pOffset=c&0x7fffff80;
+        return (int)(c>>7);
+    } else if(0xe000<=c && c!=0xfeff && c<0xfff0) {
+        /* For these characters we need to take the gapOffset into account. */
+        *pOffset=c&0x7fffff80;
+        return (int)((c-gapOffset)>>7);
+    } else {
+        return -1;
+    }
+}
+U_CDECL_BEGIN
+/*
+ * Idea for compression:
+ *  - save SCSUData and other state before really starting work
+ *  - at endloop, see if compression could be better with just unicode mode
+ *  - don't do this if a callback has been called
+ *  - if unicode mode would be smaller, then override the results with it - may need SCU at the beginning
+ *  - different buffer handling!
+ *
+ * Drawback or need for corrective handling:
+ * it is desirable to encode U+feff as SQU fe ff for the SCSU signature, and
+ * it is desirable to start a document in US-ASCII/Latin-1 for as long as possible
+ * not only for compression but also for HTML/XML documents with following charset/encoding announcers.
+ *
+ * How to achieve both?
+ *  - Only replace the result after an SDX or SCU?
+ */
+
+static void U_CALLCONV
+_SCSUFromUnicodeWithOffsets(UConverterFromUnicodeArgs *pArgs,
+                            UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    SCSUData *scsu;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+    int32_t *offsets;
+
+    UBool isSingleByteMode;
+    uint8_t dynamicWindow;
+    uint32_t currentOffset;
+
+    uint32_t c, delta;
+
+    int32_t sourceIndex, nextSourceIndex;
+
+    int32_t length;
+
+    /* variables for compression heuristics */
+    uint32_t offset;
+    UChar lead, trail;
+    int code;
+    int8_t window;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    scsu=(SCSUData *)cnv->extraInfo;
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+    offsets=pArgs->offsets;
+
+    /* get the state machine state */
+    isSingleByteMode=scsu->fromUIsSingleByteMode;
+    dynamicWindow=scsu->fromUDynamicWindow;
+    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+
+    c=cnv->fromUChar32;
+
+    /* sourceIndex=-1 if the current character began in the previous buffer */
+    sourceIndex= c==0 ? 0 : -1;
+    nextSourceIndex=0;
+
+    /* similar conversion "loop" as in toUnicode */
+loop:
+    if(isSingleByteMode) {
+        if(c!=0 && targetCapacity>0) {
+            goto getTrailSingle;
+        }
+
+        /* state machine for single-byte mode */
+/* singleByteMode: */
+        while(source<sourceLimit) {
+            if(targetCapacity<=0) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            c=*source++;
+            ++nextSourceIndex;
+
+            if((c-0x20)<=0x5f) {
+                /* pass US-ASCII graphic character through */
+                *target++=(uint8_t)c;
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                --targetCapacity;
+            } else if(c<0x20) {
+                if((1UL<<c)&0x2601 /* binary 0010 0110 0000 0001, check for b==0xd || b==0xa || b==9 || b==0 */) {
+                    /* CR/LF/TAB/NUL */
+                    *target++=(uint8_t)c;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    --targetCapacity;
+                } else {
+                    /* quote C0 control character */
+                    c|=SQ0<<8;
+                    length=2;
+                    goto outputBytes;
+                }
+            } else if((delta=c-currentOffset)<=0x7f) {
+                /* use the current dynamic window */
+                *target++=(uint8_t)(delta|0x80);
+                if(offsets!=NULL) {
+                    *offsets++=sourceIndex;
+                }
+                --targetCapacity;
+            } else if(U16_IS_SURROGATE(c)) {
+                if(U16_IS_SURROGATE_LEAD(c)) {
+getTrailSingle:
+                    lead=(UChar)c;
+                    if(source<sourceLimit) {
+                        /* test the following code unit */
+                        trail=*source;
+                        if(U16_IS_TRAIL(trail)) {
+                            ++source;
+                            ++nextSourceIndex;
+                            c=U16_GET_SUPPLEMENTARY(c, trail);
+                            /* convert this surrogate code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                    } else {
+                        /* no more input */
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                /* compress supplementary character U+10000..U+10ffff */
+                if((delta=c-currentOffset)<=0x7f) {
+                    /* use the current dynamic window */
+                    *target++=(uint8_t)(delta|0x80);
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    --targetCapacity;
+                } else if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                    /* there is a dynamic window that contains this character, change to it */
+                    dynamicWindow=window;
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(SC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                    length=2;
+                    goto outputBytes;
+                } else if((code=getDynamicOffset(c, &offset))>=0) {
+                    /* might check if there are more characters in this window to come */
+                    /* define an extended window with this character */
+                    code-=0x200;
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)SDX<<24)|((uint32_t)dynamicWindow<<21)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=4;
+                    goto outputBytes;
+                } else {
+                    /* change to Unicode mode and output this (lead, trail) pair */
+                    isSingleByteMode=FALSE;
+                    *target++=(uint8_t)SCU;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                    }
+                    --targetCapacity;
+                    c=((uint32_t)lead<<16)|trail;
+                    length=4;
+                    goto outputBytes;
+                }
+            } else if(c<0xa0) {
+                /* quote C1 control character */
+                c=(c&0x7f)|(SQ0+1)<<8; /* SQ0+1==SQ1 */
+                length=2;
+                goto outputBytes;
+            } else if(c==0xfeff || c>=0xfff0) {
+                /* quote signature character=byte order mark and specials */
+                c|=SQU<<16;
+                length=3;
+                goto outputBytes;
+            } else {
+                /* compress all other BMP characters */
+                if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                    /* there is a window defined that contains this character - switch to it or quote from it? */
+                    if(source>=sourceLimit || isInOffsetWindowOrDirect(scsu->fromUDynamicOffsets[window], *source)) {
+                        /* change to dynamic window */
+                        dynamicWindow=window;
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(SC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                        length=2;
+                        goto outputBytes;
+                    } else {
+                        /* quote from dynamic window */
+                        c=((uint32_t)(SQ0+window)<<8)|(c-scsu->fromUDynamicOffsets[window])|0x80;
+                        length=2;
+                        goto outputBytes;
+                    }
+                } else if((window=getWindow(staticOffsets, c))>=0) {
+                    /* quote from static window */
+                    c=((uint32_t)(SQ0+window)<<8)|(c-staticOffsets[window]);
+                    length=2;
+                    goto outputBytes;
+                } else if((code=getDynamicOffset(c, &offset))>=0) {
+                    /* define a dynamic window with this character */
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(SD0+dynamicWindow)<<16)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=3;
+                    goto outputBytes;
+                } else if((uint32_t)(c-0x3400)<(0xd800-0x3400) &&
+                          (source>=sourceLimit || (uint32_t)(*source-0x3400)<(0xd800-0x3400))
+                ) {
+                    /*
+                     * this character is not compressible (a BMP ideograph or similar);
+                     * switch to Unicode mode if this is the last character in the block
+                     * or there is at least one more ideograph following immediately
+                     */
+                    isSingleByteMode=FALSE;
+                    c|=SCU<<16;
+                    length=3;
+                    goto outputBytes;
+                } else {
+                    /* quote Unicode */
+                    c|=SQU<<16;
+                    length=3;
+                    goto outputBytes;
+                }
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+            sourceIndex=nextSourceIndex;
+        }
+    } else {
+        if(c!=0 && targetCapacity>0) {
+            goto getTrailUnicode;
+        }
+
+        /* state machine for Unicode mode */
+/* unicodeByteMode: */
+        while(source<sourceLimit) {
+            if(targetCapacity<=0) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            c=*source++;
+            ++nextSourceIndex;
+
+            if((uint32_t)(c-0x3400)<(0xd800-0x3400)) {
+                /* not compressible, write character directly */
+                if(targetCapacity>=2) {
+                    *target++=(uint8_t)(c>>8);
+                    *target++=(uint8_t)c;
+                    if(offsets!=NULL) {
+                        *offsets++=sourceIndex;
+                        *offsets++=sourceIndex;
+                    }
+                    targetCapacity-=2;
+                } else {
+                    length=2;
+                    goto outputBytes;
+                }
+            } else if((uint32_t)(c-0x3400)>=(0xf300-0x3400) /* c<0x3400 || c>=0xf300 */) {
+                /* compress BMP character if the following one is not an uncompressible ideograph */
+                if(!(source<sourceLimit && (uint32_t)(*source-0x3400)<(0xd800-0x3400))) {
+                    if(((uint32_t)(c-0x30)<10 || (uint32_t)(c-0x61)<26 || (uint32_t)(c-0x41)<26)) {
+                        /* ASCII digit or letter */
+                        isSingleByteMode=TRUE;
+                        c|=((uint32_t)(UC0+dynamicWindow)<<8)|c;
+                        length=2;
+                        goto outputBytes;
+                    } else if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                        /* there is a dynamic window that contains this character, change to it */
+                        isSingleByteMode=TRUE;
+                        dynamicWindow=window;
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(UC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                        length=2;
+                        goto outputBytes;
+                    } else if((code=getDynamicOffset(c, &offset))>=0) {
+                        /* define a dynamic window with this character */
+                        isSingleByteMode=TRUE;
+                        dynamicWindow=getNextDynamicWindow(scsu);
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(UD0+dynamicWindow)<<16)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                        length=3;
+                        goto outputBytes;
+                    }
+                }
+
+                /* don't know how to compress this character, just write it directly */
+                length=2;
+                goto outputBytes;
+            } else if(c<0xe000) {
+                /* c is a surrogate */
+                if(U16_IS_SURROGATE_LEAD(c)) {
+getTrailUnicode:
+                    lead=(UChar)c;
+                    if(source<sourceLimit) {
+                        /* test the following code unit */
+                        trail=*source;
+                        if(U16_IS_TRAIL(trail)) {
+                            ++source;
+                            ++nextSourceIndex;
+                            c=U16_GET_SUPPLEMENTARY(c, trail);
+                            /* convert this surrogate code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                    } else {
+                        /* no more input */
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                /* compress supplementary character */
+                if( (window=getWindow(scsu->fromUDynamicOffsets, c))>=0 &&
+                    !(source<sourceLimit && (uint32_t)(*source-0x3400)<(0xd800-0x3400))
+                ) {
+                    /*
+                     * there is a dynamic window that contains this character and
+                     * the following character is not uncompressible,
+                     * change to the window
+                     */
+                    isSingleByteMode=TRUE;
+                    dynamicWindow=window;
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(UC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                    length=2;
+                    goto outputBytes;
+                } else if(source<sourceLimit && lead==*source && /* too lazy to check trail in same window as source[1] */
+                          (code=getDynamicOffset(c, &offset))>=0
+                ) {
+                    /* two supplementary characters in (probably) the same window - define an extended one */
+                    isSingleByteMode=TRUE;
+                    code-=0x200;
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)UDX<<24)|((uint32_t)dynamicWindow<<21)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=4;
+                    goto outputBytes;
+                } else {
+                    /* don't know how to compress this character, just write it directly */
+                    c=((uint32_t)lead<<16)|trail;
+                    length=4;
+                    goto outputBytes;
+                }
+            } else /* 0xe000<=c<0xf300 */ {
+                /* quote to avoid SCSU tags */
+                c|=UQU<<16;
+                length=3;
+                goto outputBytes;
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+            sourceIndex=nextSourceIndex;
+        }
+    }
+endloop:
+
+    /* set the converter state back into UConverter */
+    scsu->fromUIsSingleByteMode=isSingleByteMode;
+    scsu->fromUDynamicWindow=dynamicWindow;
+
+    cnv->fromUChar32=c;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    pArgs->offsets=offsets;
+    return;
+
+outputBytes:
+    /* write the output character bytes from c and length [code copied from ucnvmbcs.c] */
+    /* from the first if in the loop we know that targetCapacity>0 */
+    if(length<=targetCapacity) {
+        if(offsets==NULL) {
+            switch(length) {
+                /* each branch falls through to the next one */
+            case 4:
+                *target++=(uint8_t)(c>>24);
+                U_FALLTHROUGH;
+            case 3:
+                *target++=(uint8_t)(c>>16);
+                U_FALLTHROUGH;
+            case 2:
+                *target++=(uint8_t)(c>>8);
+                U_FALLTHROUGH;
+            case 1:
+                *target++=(uint8_t)c;
+                U_FALLTHROUGH;
+            default:
+                /* will never occur */
+                break;
+            }
+        } else {
+            switch(length) {
+                /* each branch falls through to the next one */
+            case 4:
+                *target++=(uint8_t)(c>>24);
+                *offsets++=sourceIndex;
+                U_FALLTHROUGH;
+            case 3:
+                *target++=(uint8_t)(c>>16);
+                *offsets++=sourceIndex;
+                U_FALLTHROUGH;
+            case 2:
+                *target++=(uint8_t)(c>>8);
+                *offsets++=sourceIndex;
+                U_FALLTHROUGH;
+            case 1:
+                *target++=(uint8_t)c;
+                *offsets++=sourceIndex;
+                U_FALLTHROUGH;
+            default:
+                /* will never occur */
+                break;
+            }
+        }
+        targetCapacity-=length;
+
+        /* normal end of conversion: prepare for a new character */
+        c=0;
+        sourceIndex=nextSourceIndex;
+        goto loop;
+    } else {
+        uint8_t *p;
+
+        /*
+         * We actually do this backwards here:
+         * In order to save an intermediate variable, we output
+         * first to the overflow buffer what does not fit into the
+         * regular target.
+         */
+        /* we know that 0<=targetCapacity<length<=4 */
+        /* targetCapacity==0 when SCU+supplementary where SCU used up targetCapacity==1 */
+        length-=targetCapacity;
+        p=(uint8_t *)cnv->charErrorBuffer;
+        switch(length) {
+            /* each branch falls through to the next one */
+        case 4:
+            *p++=(uint8_t)(c>>24);
+            U_FALLTHROUGH;
+        case 3:
+            *p++=(uint8_t)(c>>16);
+            U_FALLTHROUGH;
+        case 2:
+            *p++=(uint8_t)(c>>8);
+            U_FALLTHROUGH;
+        case 1:
+            *p=(uint8_t)c;
+            U_FALLTHROUGH;
+        default:
+            /* will never occur */
+            break;
+        }
+        cnv->charErrorBufferLength=(int8_t)length;
+
+        /* now output what fits into the regular target */
+        c>>=8*length; /* length was reduced by targetCapacity */
+        switch(targetCapacity) {
+            /* each branch falls through to the next one */
+        case 3:
+            *target++=(uint8_t)(c>>16);
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            U_FALLTHROUGH;
+        case 2:
+            *target++=(uint8_t)(c>>8);
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            U_FALLTHROUGH;
+        case 1:
+            *target++=(uint8_t)c;
+            if(offsets!=NULL) {
+                *offsets++=sourceIndex;
+            }
+            U_FALLTHROUGH;
+        default:
+            break;
+        }
+
+        /* target overflow */
+        targetCapacity=0;
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        c=0;
+        goto endloop;
+    }
+}
+
+/*
+ * Identical to _SCSUFromUnicodeWithOffsets but without offset handling.
+ * If a change is made in the original function, then either
+ * change this function the same way or
+ * re-copy the original function and remove the variables
+ * offsets, sourceIndex, and nextSourceIndex.
+ */
+static void U_CALLCONV
+_SCSUFromUnicode(UConverterFromUnicodeArgs *pArgs,
+                 UErrorCode *pErrorCode) {
+    UConverter *cnv;
+    SCSUData *scsu;
+    const UChar *source, *sourceLimit;
+    uint8_t *target;
+    int32_t targetCapacity;
+
+    UBool isSingleByteMode;
+    uint8_t dynamicWindow;
+    uint32_t currentOffset;
+
+    uint32_t c, delta;
+
+    int32_t length;
+
+    /* variables for compression heuristics */
+    uint32_t offset;
+    UChar lead, trail;
+    int code;
+    int8_t window;
+
+    /* set up the local pointers */
+    cnv=pArgs->converter;
+    scsu=(SCSUData *)cnv->extraInfo;
+
+    /* set up the local pointers */
+    source=pArgs->source;
+    sourceLimit=pArgs->sourceLimit;
+    target=(uint8_t *)pArgs->target;
+    targetCapacity=(int32_t)(pArgs->targetLimit-pArgs->target);
+
+    /* get the state machine state */
+    isSingleByteMode=scsu->fromUIsSingleByteMode;
+    dynamicWindow=scsu->fromUDynamicWindow;
+    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+
+    c=cnv->fromUChar32;
+
+    /* similar conversion "loop" as in toUnicode */
+loop:
+    if(isSingleByteMode) {
+        if(c!=0 && targetCapacity>0) {
+            goto getTrailSingle;
+        }
+
+        /* state machine for single-byte mode */
+/* singleByteMode: */
+        while(source<sourceLimit) {
+            if(targetCapacity<=0) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            c=*source++;
+
+            if((c-0x20)<=0x5f) {
+                /* pass US-ASCII graphic character through */
+                *target++=(uint8_t)c;
+                --targetCapacity;
+            } else if(c<0x20) {
+                if((1UL<<c)&0x2601 /* binary 0010 0110 0000 0001, check for b==0xd || b==0xa || b==9 || b==0 */) {
+                    /* CR/LF/TAB/NUL */
+                    *target++=(uint8_t)c;
+                    --targetCapacity;
+                } else {
+                    /* quote C0 control character */
+                    c|=SQ0<<8;
+                    length=2;
+                    goto outputBytes;
+                }
+            } else if((delta=c-currentOffset)<=0x7f) {
+                /* use the current dynamic window */
+                *target++=(uint8_t)(delta|0x80);
+                --targetCapacity;
+            } else if(U16_IS_SURROGATE(c)) {
+                if(U16_IS_SURROGATE_LEAD(c)) {
+getTrailSingle:
+                    lead=(UChar)c;
+                    if(source<sourceLimit) {
+                        /* test the following code unit */
+                        trail=*source;
+                        if(U16_IS_TRAIL(trail)) {
+                            ++source;
+                            c=U16_GET_SUPPLEMENTARY(c, trail);
+                            /* convert this surrogate code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                    } else {
+                        /* no more input */
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                /* compress supplementary character U+10000..U+10ffff */
+                if((delta=c-currentOffset)<=0x7f) {
+                    /* use the current dynamic window */
+                    *target++=(uint8_t)(delta|0x80);
+                    --targetCapacity;
+                } else if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                    /* there is a dynamic window that contains this character, change to it */
+                    dynamicWindow=window;
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(SC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                    length=2;
+                    goto outputBytes;
+                } else if((code=getDynamicOffset(c, &offset))>=0) {
+                    /* might check if there are more characters in this window to come */
+                    /* define an extended window with this character */
+                    code-=0x200;
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)SDX<<24)|((uint32_t)dynamicWindow<<21)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=4;
+                    goto outputBytes;
+                } else {
+                    /* change to Unicode mode and output this (lead, trail) pair */
+                    isSingleByteMode=FALSE;
+                    *target++=(uint8_t)SCU;
+                    --targetCapacity;
+                    c=((uint32_t)lead<<16)|trail;
+                    length=4;
+                    goto outputBytes;
+                }
+            } else if(c<0xa0) {
+                /* quote C1 control character */
+                c=(c&0x7f)|(SQ0+1)<<8; /* SQ0+1==SQ1 */
+                length=2;
+                goto outputBytes;
+            } else if(c==0xfeff || c>=0xfff0) {
+                /* quote signature character=byte order mark and specials */
+                c|=SQU<<16;
+                length=3;
+                goto outputBytes;
+            } else {
+                /* compress all other BMP characters */
+                if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                    /* there is a window defined that contains this character - switch to it or quote from it? */
+                    if(source>=sourceLimit || isInOffsetWindowOrDirect(scsu->fromUDynamicOffsets[window], *source)) {
+                        /* change to dynamic window */
+                        dynamicWindow=window;
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(SC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                        length=2;
+                        goto outputBytes;
+                    } else {
+                        /* quote from dynamic window */
+                        c=((uint32_t)(SQ0+window)<<8)|(c-scsu->fromUDynamicOffsets[window])|0x80;
+                        length=2;
+                        goto outputBytes;
+                    }
+                } else if((window=getWindow(staticOffsets, c))>=0) {
+                    /* quote from static window */
+                    c=((uint32_t)(SQ0+window)<<8)|(c-staticOffsets[window]);
+                    length=2;
+                    goto outputBytes;
+                } else if((code=getDynamicOffset(c, &offset))>=0) {
+                    /* define a dynamic window with this character */
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(SD0+dynamicWindow)<<16)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=3;
+                    goto outputBytes;
+                } else if((uint32_t)(c-0x3400)<(0xd800-0x3400) &&
+                          (source>=sourceLimit || (uint32_t)(*source-0x3400)<(0xd800-0x3400))
+                ) {
+                    /*
+                     * this character is not compressible (a BMP ideograph or similar);
+                     * switch to Unicode mode if this is the last character in the block
+                     * or there is at least one more ideograph following immediately
+                     */
+                    isSingleByteMode=FALSE;
+                    c|=SCU<<16;
+                    length=3;
+                    goto outputBytes;
+                } else {
+                    /* quote Unicode */
+                    c|=SQU<<16;
+                    length=3;
+                    goto outputBytes;
+                }
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+        }
+    } else {
+        if(c!=0 && targetCapacity>0) {
+            goto getTrailUnicode;
+        }
+
+        /* state machine for Unicode mode */
+/* unicodeByteMode: */
+        while(source<sourceLimit) {
+            if(targetCapacity<=0) {
+                /* target is full */
+                *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                break;
+            }
+            c=*source++;
+
+            if((uint32_t)(c-0x3400)<(0xd800-0x3400)) {
+                /* not compressible, write character directly */
+                if(targetCapacity>=2) {
+                    *target++=(uint8_t)(c>>8);
+                    *target++=(uint8_t)c;
+                    targetCapacity-=2;
+                } else {
+                    length=2;
+                    goto outputBytes;
+                }
+            } else if((uint32_t)(c-0x3400)>=(0xf300-0x3400) /* c<0x3400 || c>=0xf300 */) {
+                /* compress BMP character if the following one is not an uncompressible ideograph */
+                if(!(source<sourceLimit && (uint32_t)(*source-0x3400)<(0xd800-0x3400))) {
+                    if(((uint32_t)(c-0x30)<10 || (uint32_t)(c-0x61)<26 || (uint32_t)(c-0x41)<26)) {
+                        /* ASCII digit or letter */
+                        isSingleByteMode=TRUE;
+                        c|=((uint32_t)(UC0+dynamicWindow)<<8)|c;
+                        length=2;
+                        goto outputBytes;
+                    } else if((window=getWindow(scsu->fromUDynamicOffsets, c))>=0) {
+                        /* there is a dynamic window that contains this character, change to it */
+                        isSingleByteMode=TRUE;
+                        dynamicWindow=window;
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(UC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                        length=2;
+                        goto outputBytes;
+                    } else if((code=getDynamicOffset(c, &offset))>=0) {
+                        /* define a dynamic window with this character */
+                        isSingleByteMode=TRUE;
+                        dynamicWindow=getNextDynamicWindow(scsu);
+                        currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                        useDynamicWindow(scsu, dynamicWindow);
+                        c=((uint32_t)(UD0+dynamicWindow)<<16)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                        length=3;
+                        goto outputBytes;
+                    }
+                }
+
+                /* don't know how to compress this character, just write it directly */
+                length=2;
+                goto outputBytes;
+            } else if(c<0xe000) {
+                /* c is a surrogate */
+                if(U16_IS_SURROGATE_LEAD(c)) {
+getTrailUnicode:
+                    lead=(UChar)c;
+                    if(source<sourceLimit) {
+                        /* test the following code unit */
+                        trail=*source;
+                        if(U16_IS_TRAIL(trail)) {
+                            ++source;
+                            c=U16_GET_SUPPLEMENTARY(c, trail);
+                            /* convert this surrogate code point */
+                            /* exit this condition tree */
+                        } else {
+                            /* this is an unmatched lead code unit (1st surrogate) */
+                            /* callback(illegal) */
+                            *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                            goto endloop;
+                        }
+                    } else {
+                        /* no more input */
+                        break;
+                    }
+                } else {
+                    /* this is an unmatched trail code unit (2nd surrogate) */
+                    /* callback(illegal) */
+                    *pErrorCode=U_ILLEGAL_CHAR_FOUND;
+                    goto endloop;
+                }
+
+                /* compress supplementary character */
+                if( (window=getWindow(scsu->fromUDynamicOffsets, c))>=0 &&
+                    !(source<sourceLimit && (uint32_t)(*source-0x3400)<(0xd800-0x3400))
+                ) {
+                    /*
+                     * there is a dynamic window that contains this character and
+                     * the following character is not uncompressible,
+                     * change to the window
+                     */
+                    isSingleByteMode=TRUE;
+                    dynamicWindow=window;
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow];
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)(UC0+dynamicWindow)<<8)|(c-currentOffset)|0x80;
+                    length=2;
+                    goto outputBytes;
+                } else if(source<sourceLimit && lead==*source && /* too lazy to check trail in same window as source[1] */
+                          (code=getDynamicOffset(c, &offset))>=0
+                ) {
+                    /* two supplementary characters in (probably) the same window - define an extended one */
+                    isSingleByteMode=TRUE;
+                    code-=0x200;
+                    dynamicWindow=getNextDynamicWindow(scsu);
+                    currentOffset=scsu->fromUDynamicOffsets[dynamicWindow]=offset;
+                    useDynamicWindow(scsu, dynamicWindow);
+                    c=((uint32_t)UDX<<24)|((uint32_t)dynamicWindow<<21)|((uint32_t)code<<8)|(c-currentOffset)|0x80;
+                    length=4;
+                    goto outputBytes;
+                } else {
+                    /* don't know how to compress this character, just write it directly */
+                    c=((uint32_t)lead<<16)|trail;
+                    length=4;
+                    goto outputBytes;
+                }
+            } else /* 0xe000<=c<0xf300 */ {
+                /* quote to avoid SCSU tags */
+                c|=UQU<<16;
+                length=3;
+                goto outputBytes;
+            }
+
+            /* normal end of conversion: prepare for a new character */
+            c=0;
+        }
+    }
+endloop:
+
+    /* set the converter state back into UConverter */
+    scsu->fromUIsSingleByteMode=isSingleByteMode;
+    scsu->fromUDynamicWindow=dynamicWindow;
+
+    cnv->fromUChar32=c;
+
+    /* write back the updated pointers */
+    pArgs->source=source;
+    pArgs->target=(char *)target;
+    return;
+
+outputBytes:
+    /* write the output character bytes from c and length [code copied from ucnvmbcs.c] */
+    /* from the first if in the loop we know that targetCapacity>0 */
+    if(length<=targetCapacity) {
+        switch(length) {
+            /* each branch falls through to the next one */
+        case 4:
+            *target++=(uint8_t)(c>>24);
+            U_FALLTHROUGH;
+        case 3:
+            *target++=(uint8_t)(c>>16);
+            U_FALLTHROUGH;
+        case 2:
+            *target++=(uint8_t)(c>>8);
+            U_FALLTHROUGH;
+        case 1:
+            *target++=(uint8_t)c;
+            U_FALLTHROUGH;
+        default:
+            /* will never occur */
+            break;
+        }
+        targetCapacity-=length;
+
+        /* normal end of conversion: prepare for a new character */
+        c=0;
+        goto loop;
+    } else {
+        uint8_t *p;
+
+        /*
+         * We actually do this backwards here:
+         * In order to save an intermediate variable, we output
+         * first to the overflow buffer what does not fit into the
+         * regular target.
+         */
+        /* we know that 0<=targetCapacity<length<=4 */
+        /* targetCapacity==0 when SCU+supplementary where SCU used up targetCapacity==1 */
+        length-=targetCapacity;
+        p=(uint8_t *)cnv->charErrorBuffer;
+        switch(length) {
+            /* each branch falls through to the next one */
+        case 4:
+            *p++=(uint8_t)(c>>24);
+            U_FALLTHROUGH;
+        case 3:
+            *p++=(uint8_t)(c>>16);
+            U_FALLTHROUGH;
+        case 2:
+            *p++=(uint8_t)(c>>8);
+            U_FALLTHROUGH;
+        case 1:
+            *p=(uint8_t)c;
+            U_FALLTHROUGH;
+        default:
+            /* will never occur */
+            break;
+        }
+        cnv->charErrorBufferLength=(int8_t)length;
+
+        /* now output what fits into the regular target */
+        c>>=8*length; /* length was reduced by targetCapacity */
+        switch(targetCapacity) {
+            /* each branch falls through to the next one */
+        case 3:
+            *target++=(uint8_t)(c>>16);
+            U_FALLTHROUGH;
+        case 2:
+            *target++=(uint8_t)(c>>8);
+            U_FALLTHROUGH;
+        case 1:
+            *target++=(uint8_t)c;
+            U_FALLTHROUGH;
+        default:
+            break;
+        }
+
+        /* target overflow */
+        targetCapacity=0;
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        c=0;
+        goto endloop;
+    }
+}
+
+/* miscellaneous ------------------------------------------------------------ */
+
+static const char *  U_CALLCONV
+_SCSUGetName(const UConverter *cnv) {
+    SCSUData *scsu=(SCSUData *)cnv->extraInfo;
+
+    switch(scsu->locale) {
+    case l_ja:
+        return "SCSU,locale=ja";
+    default:
+        return "SCSU";
+    }
+}
+
+/* structure for SafeClone calculations */
+struct cloneSCSUStruct
+{
+    UConverter cnv;
+    SCSUData mydata;
+};
+
+static UConverter *  U_CALLCONV
+_SCSUSafeClone(const UConverter *cnv, 
+               void *stackBuffer, 
+               int32_t *pBufferSize, 
+               UErrorCode *status)
+{
+    struct cloneSCSUStruct * localClone;
+    int32_t bufferSizeNeeded = sizeof(struct cloneSCSUStruct);
+
+    if (U_FAILURE(*status)){
+        return 0;
+    }
+
+    if (*pBufferSize == 0){ /* 'preflighting' request - set needed size into *pBufferSize */
+        *pBufferSize = bufferSizeNeeded;
+        return 0;
+    }
+
+    localClone = (struct cloneSCSUStruct *)stackBuffer;
+    /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
+
+    uprv_memcpy(&localClone->mydata, cnv->extraInfo, sizeof(SCSUData));
+    localClone->cnv.extraInfo = &localClone->mydata;
+    localClone->cnv.isExtraLocal = TRUE;
+
+    return &localClone->cnv;
+}
+U_CDECL_END
+
+static const UConverterImpl _SCSUImpl={
+    UCNV_SCSU,
+
+    NULL,
+    NULL,
+
+    _SCSUOpen,
+    _SCSUClose,
+    _SCSUReset,
+
+    _SCSUToUnicode,
+    _SCSUToUnicodeWithOffsets,
+    _SCSUFromUnicode,
+    _SCSUFromUnicodeWithOffsets,
+    NULL,
+
+    NULL,
+    _SCSUGetName,
+    NULL,
+    _SCSUSafeClone,
+    ucnv_getCompleteUnicodeSet,
+    NULL,
+    NULL
+};
+
+static const UConverterStaticData _SCSUStaticData={
+    sizeof(UConverterStaticData),
+    "SCSU",
+    1212, /* CCSID for SCSU */
+    UCNV_IBM, UCNV_SCSU,
+    1, 3, /* one UChar generates at least 1 byte and at most 3 bytes */
+    /*
+     * The subchar here is ignored because _SCSUOpen() sets U+fffd as a Unicode
+     * substitution string.
+     */
+    { 0x0e, 0xff, 0xfd, 0 }, 3,
+    FALSE, FALSE,
+    0,
+    0,
+    { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
+};
+
+const UConverterSharedData _SCSUData=
+        UCNV_IMMUTABLE_SHARED_DATA_INITIALIZER(&_SCSUStaticData, &_SCSUImpl);
+
+#endif
diff --git a/thirdparty/icu4c/common/ucnvsel.cpp b/thirdparty/icu4c/common/ucnvsel.cpp
new file mode 100644
index 0000000000..2dff5ac1bc
--- /dev/null
+++ b/thirdparty/icu4c/common/ucnvsel.cpp
@@ -0,0 +1,823 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2008-2011, International Business Machines
+*   Corporation, Google and others.  All Rights Reserved.
+*
+*******************************************************************************
+*/
+// Author : eldawy@google.com (Mohamed Eldawy)
+// ucnvsel.cpp
+//
+// Purpose: To generate a list of encodings capable of handling
+// a given Unicode text
+//
+// Started 09-April-2008
+
+/**
+ * \file
+ *
+ * This is an implementation of an encoding selector.
+ * The goal is, given a unicode string, find the encodings
+ * this string can be mapped to. To make processing faster
+ * a trie is built when you call ucnvsel_open() that
+ * stores all encodings a codepoint can map to
+ */
+
+#include "unicode/ucnvsel.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include <string.h>
+
+#include "unicode/uchar.h"
+#include "unicode/uniset.h"
+#include "unicode/ucnv.h"
+#include "unicode/ustring.h"
+#include "unicode/uchriter.h"
+#include "utrie2.h"
+#include "propsvec.h"
+#include "uassert.h"
+#include "ucmndata.h"
+#include "udataswp.h"
+#include "uenumimp.h"
+#include "cmemory.h"
+#include "cstring.h"
+
+U_NAMESPACE_USE
+
+struct UConverterSelector {
+  UTrie2 *trie;              // 16 bit trie containing offsets into pv
+  uint32_t* pv;              // table of bits!
+  int32_t pvCount;
+  char** encodings;          // which encodings did user ask to use?
+  int32_t encodingsCount;
+  int32_t encodingStrLength;
+  uint8_t* swapped;
+  UBool ownPv, ownEncodingStrings;
+};
+
+static void generateSelectorData(UConverterSelector* result,
+                                 UPropsVectors *upvec,
+                                 const USet* excludedCodePoints,
+                                 const UConverterUnicodeSet whichSet,
+                                 UErrorCode* status) {
+  if (U_FAILURE(*status)) {
+    return;
+  }
+
+  int32_t columns = (result->encodingsCount+31)/32;
+
+  // set errorValue to all-ones
+  for (int32_t col = 0; col < columns; col++) {
+    upvec_setValue(upvec, UPVEC_ERROR_VALUE_CP, UPVEC_ERROR_VALUE_CP,
+                   col, static_cast<uint32_t>(~0), static_cast<uint32_t>(~0), status);
+  }
+
+  for (int32_t i = 0; i < result->encodingsCount; ++i) {
+    uint32_t mask;
+    uint32_t column;
+    int32_t item_count;
+    int32_t j;
+    UConverter* test_converter = ucnv_open(result->encodings[i], status);
+    if (U_FAILURE(*status)) {
+      return;
+    }
+    USet* unicode_point_set;
+    unicode_point_set = uset_open(1, 0);  // empty set
+
+    ucnv_getUnicodeSet(test_converter, unicode_point_set,
+                       whichSet, status);
+    if (U_FAILURE(*status)) {
+      ucnv_close(test_converter);
+      return;
+    }
+
+    column = i / 32;
+    mask = 1 << (i%32);
+    // now iterate over intervals on set i!
+    item_count = uset_getItemCount(unicode_point_set);
+
+    for (j = 0; j < item_count; ++j) {
+      UChar32 start_char;
+      UChar32 end_char;
+      UErrorCode smallStatus = U_ZERO_ERROR;
+      uset_getItem(unicode_point_set, j, &start_char, &end_char, NULL, 0,
+                   &smallStatus);
+      if (U_FAILURE(smallStatus)) {
+        // this will be reached for the converters that fill the set with
+        // strings. Those should be ignored by our system
+      } else {
+        upvec_setValue(upvec, start_char, end_char, column, static_cast<uint32_t>(~0), mask,
+                       status);
+      }
+    }
+    ucnv_close(test_converter);
+    uset_close(unicode_point_set);
+    if (U_FAILURE(*status)) {
+      return;
+    }
+  }
+
+  // handle excluded encodings! Simply set their values to all 1's in the upvec
+  if (excludedCodePoints) {
+    int32_t item_count = uset_getItemCount(excludedCodePoints);
+    for (int32_t j = 0; j < item_count; ++j) {
+      UChar32 start_char;
+      UChar32 end_char;
+
+      uset_getItem(excludedCodePoints, j, &start_char, &end_char, NULL, 0,
+                   status);
+      for (int32_t col = 0; col < columns; col++) {
+        upvec_setValue(upvec, start_char, end_char, col, static_cast<uint32_t>(~0), static_cast<uint32_t>(~0),
+                      status);
+      }
+    }
+  }
+
+  // alright. Now, let's put things in the same exact form you'd get when you
+  // unserialize things.
+  result->trie = upvec_compactToUTrie2WithRowIndexes(upvec, status);
+  result->pv = upvec_cloneArray(upvec, &result->pvCount, NULL, status);
+  result->pvCount *= columns;  // number of uint32_t = rows * columns
+  result->ownPv = TRUE;
+}
+
+/* open a selector. If converterListSize is 0, build for all converters.
+   If excludedCodePoints is NULL, don't exclude any codepoints */
+U_CAPI UConverterSelector* U_EXPORT2
+ucnvsel_open(const char* const*  converterList, int32_t converterListSize,
+             const USet* excludedCodePoints,
+             const UConverterUnicodeSet whichSet, UErrorCode* status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+  // ensure args make sense!
+  if (converterListSize < 0 || (converterList == NULL && converterListSize != 0)) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+
+  // allocate a new converter
+  LocalUConverterSelectorPointer newSelector(
+    (UConverterSelector*)uprv_malloc(sizeof(UConverterSelector)));
+  if (newSelector.isNull()) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+  uprv_memset(newSelector.getAlias(), 0, sizeof(UConverterSelector));
+
+  if (converterListSize == 0) {
+    converterList = NULL;
+    converterListSize = ucnv_countAvailable();
+  }
+  newSelector->encodings =
+    (char**)uprv_malloc(converterListSize * sizeof(char*));
+  if (!newSelector->encodings) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+  newSelector->encodings[0] = NULL;  // now we can call ucnvsel_close()
+
+  // make a backup copy of the list of converters
+  int32_t totalSize = 0;
+  int32_t i;
+  for (i = 0; i < converterListSize; i++) {
+    totalSize +=
+      (int32_t)uprv_strlen(converterList != NULL ? converterList[i] : ucnv_getAvailableName(i)) + 1;
+  }
+  // 4-align the totalSize to 4-align the size of the serialized form
+  int32_t encodingStrPadding = totalSize & 3;
+  if (encodingStrPadding != 0) {
+    encodingStrPadding = 4 - encodingStrPadding;
+  }
+  newSelector->encodingStrLength = totalSize += encodingStrPadding;
+  char* allStrings = (char*) uprv_malloc(totalSize);
+  if (!allStrings) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+
+  for (i = 0; i < converterListSize; i++) {
+    newSelector->encodings[i] = allStrings;
+    uprv_strcpy(newSelector->encodings[i],
+                converterList != NULL ? converterList[i] : ucnv_getAvailableName(i));
+    allStrings += uprv_strlen(newSelector->encodings[i]) + 1;
+  }
+  while (encodingStrPadding > 0) {
+    *allStrings++ = 0;
+    --encodingStrPadding;
+  }
+
+  newSelector->ownEncodingStrings = TRUE;
+  newSelector->encodingsCount = converterListSize;
+  UPropsVectors *upvec = upvec_open((converterListSize+31)/32, status);
+  generateSelectorData(newSelector.getAlias(), upvec, excludedCodePoints, whichSet, status);
+  upvec_close(upvec);
+
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+
+  return newSelector.orphan();
+}
+
+/* close opened selector */
+U_CAPI void U_EXPORT2
+ucnvsel_close(UConverterSelector *sel) {
+  if (!sel) {
+    return;
+  }
+  if (sel->ownEncodingStrings) {
+    uprv_free(sel->encodings[0]);
+  }
+  uprv_free(sel->encodings);
+  if (sel->ownPv) {
+    uprv_free(sel->pv);
+  }
+  utrie2_close(sel->trie);
+  uprv_free(sel->swapped);
+  uprv_free(sel);
+}
+
+static const UDataInfo dataInfo = {
+  sizeof(UDataInfo),
+  0,
+
+  U_IS_BIG_ENDIAN,
+  U_CHARSET_FAMILY,
+  U_SIZEOF_UCHAR,
+  0,
+
+  { 0x43, 0x53, 0x65, 0x6c },   /* dataFormat="CSel" */
+  { 1, 0, 0, 0 },               /* formatVersion */
+  { 0, 0, 0, 0 }                /* dataVersion */
+};
+
+enum {
+  UCNVSEL_INDEX_TRIE_SIZE,      // trie size in bytes
+  UCNVSEL_INDEX_PV_COUNT,       // number of uint32_t in the bit vectors
+  UCNVSEL_INDEX_NAMES_COUNT,    // number of encoding names
+  UCNVSEL_INDEX_NAMES_LENGTH,   // number of encoding name bytes including padding
+  UCNVSEL_INDEX_SIZE = 15,      // bytes following the DataHeader
+  UCNVSEL_INDEX_COUNT = 16
+};
+
+/*
+ * Serialized form of a UConverterSelector, formatVersion 1:
+ *
+ * The serialized form begins with a standard ICU DataHeader with a UDataInfo
+ * as the template above.
+ * This is followed by:
+ *   int32_t indexes[UCNVSEL_INDEX_COUNT];          // see index entry constants above
+ *   serialized UTrie2;                             // indexes[UCNVSEL_INDEX_TRIE_SIZE] bytes
+ *   uint32_t pv[indexes[UCNVSEL_INDEX_PV_COUNT]];  // bit vectors
+ *   char* encodingNames[indexes[UCNVSEL_INDEX_NAMES_LENGTH]];  // NUL-terminated strings + padding
+ */
+
+/* serialize a selector */
+U_CAPI int32_t U_EXPORT2
+ucnvsel_serialize(const UConverterSelector* sel,
+                  void* buffer, int32_t bufferCapacity, UErrorCode* status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return 0;
+  }
+  // ensure args make sense!
+  uint8_t *p = (uint8_t *)buffer;
+  if (bufferCapacity < 0 ||
+      (bufferCapacity > 0 && (p == NULL || (U_POINTER_MASK_LSB(p, 3) != 0)))
+  ) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0;
+  }
+  // add up the size of the serialized form
+  int32_t serializedTrieSize = utrie2_serialize(sel->trie, NULL, 0, status);
+  if (*status != U_BUFFER_OVERFLOW_ERROR && U_FAILURE(*status)) {
+    return 0;
+  }
+  *status = U_ZERO_ERROR;
+
+  DataHeader header;
+  uprv_memset(&header, 0, sizeof(header));
+  header.dataHeader.headerSize = (uint16_t)((sizeof(header) + 15) & ~15);
+  header.dataHeader.magic1 = 0xda;
+  header.dataHeader.magic2 = 0x27;
+  uprv_memcpy(&header.info, &dataInfo, sizeof(dataInfo));
+
+  int32_t indexes[UCNVSEL_INDEX_COUNT] = {
+    serializedTrieSize,
+    sel->pvCount,
+    sel->encodingsCount,
+    sel->encodingStrLength
+  };
+
+  int32_t totalSize =
+    header.dataHeader.headerSize +
+    (int32_t)sizeof(indexes) +
+    serializedTrieSize +
+    sel->pvCount * 4 +
+    sel->encodingStrLength;
+  indexes[UCNVSEL_INDEX_SIZE] = totalSize - header.dataHeader.headerSize;
+  if (totalSize > bufferCapacity) {
+    *status = U_BUFFER_OVERFLOW_ERROR;
+    return totalSize;
+  }
+  // ok, save!
+  int32_t length = header.dataHeader.headerSize;
+  uprv_memcpy(p, &header, sizeof(header));
+  uprv_memset(p + sizeof(header), 0, length - sizeof(header));
+  p += length;
+
+  length = (int32_t)sizeof(indexes);
+  uprv_memcpy(p, indexes, length);
+  p += length;
+
+  utrie2_serialize(sel->trie, p, serializedTrieSize, status);
+  p += serializedTrieSize;
+
+  length = sel->pvCount * 4;
+  uprv_memcpy(p, sel->pv, length);
+  p += length;
+
+  uprv_memcpy(p, sel->encodings[0], sel->encodingStrLength);
+  p += sel->encodingStrLength;
+
+  return totalSize;
+}
+
+/**
+ * swap a selector into the desired Endianness and Asciiness of
+ * the system. Just as FYI, selectors are always saved in the format
+ * of the system that created them. They are only converted if used
+ * on another system. In other words, selectors created on different
+ * system can be different even if the params are identical (endianness
+ * and Asciiness differences only)
+ *
+ * @param ds pointer to data swapper containing swapping info
+ * @param inData pointer to incoming data
+ * @param length length of inData in bytes
+ * @param outData pointer to output data. Capacity should
+ *                be at least equal to capacity of inData
+ * @param status an in/out ICU UErrorCode
+ * @return 0 on failure, number of bytes swapped on success
+ *         number of bytes swapped can be smaller than length
+ */
+static int32_t
+ucnvsel_swap(const UDataSwapper *ds,
+             const void *inData, int32_t length,
+             void *outData, UErrorCode *status) {
+  /* udata_swapDataHeader checks the arguments */
+  int32_t headerSize = udata_swapDataHeader(ds, inData, length, outData, status);
+  if(U_FAILURE(*status)) {
+    return 0;
+  }
+
+  /* check data format and format version */
+  const UDataInfo *pInfo = (const UDataInfo *)((const char *)inData + 4);
+  if(!(
+    pInfo->dataFormat[0] == 0x43 &&  /* dataFormat="CSel" */
+    pInfo->dataFormat[1] == 0x53 &&
+    pInfo->dataFormat[2] == 0x65 &&
+    pInfo->dataFormat[3] == 0x6c
+  )) {
+    udata_printError(ds, "ucnvsel_swap(): data format %02x.%02x.%02x.%02x is not recognized as UConverterSelector data\n",
+                     pInfo->dataFormat[0], pInfo->dataFormat[1],
+                     pInfo->dataFormat[2], pInfo->dataFormat[3]);
+    *status = U_INVALID_FORMAT_ERROR;
+    return 0;
+  }
+  if(pInfo->formatVersion[0] != 1) {
+    udata_printError(ds, "ucnvsel_swap(): format version %02x is not supported\n",
+                     pInfo->formatVersion[0]);
+    *status = U_UNSUPPORTED_ERROR;
+    return 0;
+  }
+
+  if(length >= 0) {
+    length -= headerSize;
+    if(length < 16*4) {
+      udata_printError(ds, "ucnvsel_swap(): too few bytes (%d after header) for UConverterSelector data\n",
+                       length);
+      *status = U_INDEX_OUTOFBOUNDS_ERROR;
+      return 0;
+    }
+  }
+
+  const uint8_t *inBytes = (const uint8_t *)inData + headerSize;
+  uint8_t *outBytes = (uint8_t *)outData + headerSize;
+
+  /* read the indexes */
+  const int32_t *inIndexes = (const int32_t *)inBytes;
+  int32_t indexes[16];
+  int32_t i;
+  for(i = 0; i < 16; ++i) {
+    indexes[i] = udata_readInt32(ds, inIndexes[i]);
+  }
+
+  /* get the total length of the data */
+  int32_t size = indexes[UCNVSEL_INDEX_SIZE];
+  if(length >= 0) {
+    if(length < size) {
+      udata_printError(ds, "ucnvsel_swap(): too few bytes (%d after header) for all of UConverterSelector data\n",
+                       length);
+      *status = U_INDEX_OUTOFBOUNDS_ERROR;
+      return 0;
+    }
+
+    /* copy the data for inaccessible bytes */
+    if(inBytes != outBytes) {
+      uprv_memcpy(outBytes, inBytes, size);
+    }
+
+    int32_t offset = 0, count;
+
+    /* swap the int32_t indexes[] */
+    count = UCNVSEL_INDEX_COUNT*4;
+    ds->swapArray32(ds, inBytes, count, outBytes, status);
+    offset += count;
+
+    /* swap the UTrie2 */
+    count = indexes[UCNVSEL_INDEX_TRIE_SIZE];
+    utrie2_swap(ds, inBytes + offset, count, outBytes + offset, status);
+    offset += count;
+
+    /* swap the uint32_t pv[] */
+    count = indexes[UCNVSEL_INDEX_PV_COUNT]*4;
+    ds->swapArray32(ds, inBytes + offset, count, outBytes + offset, status);
+    offset += count;
+
+    /* swap the encoding names */
+    count = indexes[UCNVSEL_INDEX_NAMES_LENGTH];
+    ds->swapInvChars(ds, inBytes + offset, count, outBytes + offset, status);
+    offset += count;
+
+    U_ASSERT(offset == size);
+  }
+
+  return headerSize + size;
+}
+
+/* unserialize a selector */
+U_CAPI UConverterSelector* U_EXPORT2
+ucnvsel_openFromSerialized(const void* buffer, int32_t length, UErrorCode* status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+  // ensure args make sense!
+  const uint8_t *p = (const uint8_t *)buffer;
+  if (length <= 0 ||
+      (length > 0 && (p == NULL || (U_POINTER_MASK_LSB(p, 3) != 0)))
+  ) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+  // header
+  if (length < 32) {
+    // not even enough space for a minimal header
+    *status = U_INDEX_OUTOFBOUNDS_ERROR;
+    return NULL;
+  }
+  const DataHeader *pHeader = (const DataHeader *)p;
+  if (!(
+    pHeader->dataHeader.magic1==0xda &&
+    pHeader->dataHeader.magic2==0x27 &&
+    pHeader->info.dataFormat[0] == 0x43 &&
+    pHeader->info.dataFormat[1] == 0x53 &&
+    pHeader->info.dataFormat[2] == 0x65 &&
+    pHeader->info.dataFormat[3] == 0x6c
+  )) {
+    /* header not valid or dataFormat not recognized */
+    *status = U_INVALID_FORMAT_ERROR;
+    return NULL;
+  }
+  if (pHeader->info.formatVersion[0] != 1) {
+    *status = U_UNSUPPORTED_ERROR;
+    return NULL;
+  }
+  uint8_t* swapped = NULL;
+  if (pHeader->info.isBigEndian != U_IS_BIG_ENDIAN ||
+      pHeader->info.charsetFamily != U_CHARSET_FAMILY
+  ) {
+    // swap the data
+    UDataSwapper *ds =
+      udata_openSwapperForInputData(p, length, U_IS_BIG_ENDIAN, U_CHARSET_FAMILY, status);
+    int32_t totalSize = ucnvsel_swap(ds, p, -1, NULL, status);
+    if (U_FAILURE(*status)) {
+      udata_closeSwapper(ds);
+      return NULL;
+    }
+    if (length < totalSize) {
+      udata_closeSwapper(ds);
+      *status = U_INDEX_OUTOFBOUNDS_ERROR;
+      return NULL;
+    }
+    swapped = (uint8_t*)uprv_malloc(totalSize);
+    if (swapped == NULL) {
+      udata_closeSwapper(ds);
+      *status = U_MEMORY_ALLOCATION_ERROR;
+      return NULL;
+    }
+    ucnvsel_swap(ds, p, length, swapped, status);
+    udata_closeSwapper(ds);
+    if (U_FAILURE(*status)) {
+      uprv_free(swapped);
+      return NULL;
+    }
+    p = swapped;
+    pHeader = (const DataHeader *)p;
+  }
+  if (length < (pHeader->dataHeader.headerSize + 16 * 4)) {
+    // not even enough space for the header and the indexes
+    uprv_free(swapped);
+    *status = U_INDEX_OUTOFBOUNDS_ERROR;
+    return NULL;
+  }
+  p += pHeader->dataHeader.headerSize;
+  length -= pHeader->dataHeader.headerSize;
+  // indexes
+  const int32_t *indexes = (const int32_t *)p;
+  if (length < indexes[UCNVSEL_INDEX_SIZE]) {
+    uprv_free(swapped);
+    *status = U_INDEX_OUTOFBOUNDS_ERROR;
+    return NULL;
+  }
+  p += UCNVSEL_INDEX_COUNT * 4;
+  // create and populate the selector object
+  UConverterSelector* sel = (UConverterSelector*)uprv_malloc(sizeof(UConverterSelector));
+  char **encodings =
+    (char **)uprv_malloc(
+      indexes[UCNVSEL_INDEX_NAMES_COUNT] * sizeof(char *));
+  if (sel == NULL || encodings == NULL) {
+    uprv_free(swapped);
+    uprv_free(sel);
+    uprv_free(encodings);
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+  uprv_memset(sel, 0, sizeof(UConverterSelector));
+  sel->pvCount = indexes[UCNVSEL_INDEX_PV_COUNT];
+  sel->encodings = encodings;
+  sel->encodingsCount = indexes[UCNVSEL_INDEX_NAMES_COUNT];
+  sel->encodingStrLength = indexes[UCNVSEL_INDEX_NAMES_LENGTH];
+  sel->swapped = swapped;
+  // trie
+  sel->trie = utrie2_openFromSerialized(UTRIE2_16_VALUE_BITS,
+                                        p, indexes[UCNVSEL_INDEX_TRIE_SIZE], NULL,
+                                        status);
+  p += indexes[UCNVSEL_INDEX_TRIE_SIZE];
+  if (U_FAILURE(*status)) {
+    ucnvsel_close(sel);
+    return NULL;
+  }
+  // bit vectors
+  sel->pv = (uint32_t *)p;
+  p += sel->pvCount * 4;
+  // encoding names
+  char* s = (char*)p;
+  for (int32_t i = 0; i < sel->encodingsCount; ++i) {
+    sel->encodings[i] = s;
+    s += uprv_strlen(s) + 1;
+  }
+  p += sel->encodingStrLength;
+
+  return sel;
+}
+
+// a bunch of functions for the enumeration thingie! Nothing fancy here. Just
+// iterate over the selected encodings
+struct Enumerator {
+  int16_t* index;
+  int16_t length;
+  int16_t cur;
+  const UConverterSelector* sel;
+};
+
+U_CDECL_BEGIN
+
+static void U_CALLCONV
+ucnvsel_close_selector_iterator(UEnumeration *enumerator) {
+  uprv_free(((Enumerator*)(enumerator->context))->index);
+  uprv_free(enumerator->context);
+  uprv_free(enumerator);
+}
+
+
+static int32_t U_CALLCONV
+ucnvsel_count_encodings(UEnumeration *enumerator, UErrorCode *status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return 0;
+  }
+  return ((Enumerator*)(enumerator->context))->length;
+}
+
+
+static const char* U_CALLCONV ucnvsel_next_encoding(UEnumeration* enumerator,
+                                                 int32_t* resultLength,
+                                                 UErrorCode* status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+
+  int16_t cur = ((Enumerator*)(enumerator->context))->cur;
+  const UConverterSelector* sel;
+  const char* result;
+  if (cur >= ((Enumerator*)(enumerator->context))->length) {
+    return NULL;
+  }
+  sel = ((Enumerator*)(enumerator->context))->sel;
+  result = sel->encodings[((Enumerator*)(enumerator->context))->index[cur] ];
+  ((Enumerator*)(enumerator->context))->cur++;
+  if (resultLength) {
+    *resultLength = (int32_t)uprv_strlen(result);
+  }
+  return result;
+}
+
+static void U_CALLCONV ucnvsel_reset_iterator(UEnumeration* enumerator,
+                                           UErrorCode* status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return ;
+  }
+  ((Enumerator*)(enumerator->context))->cur = 0;
+}
+
+U_CDECL_END
+
+
+static const UEnumeration defaultEncodings = {
+  NULL,
+    NULL,
+    ucnvsel_close_selector_iterator,
+    ucnvsel_count_encodings,
+    uenum_unextDefault,
+    ucnvsel_next_encoding, 
+    ucnvsel_reset_iterator
+};
+
+
+// internal fn to intersect two sets of masks
+// returns whether the mask has reduced to all zeros
+static UBool intersectMasks(uint32_t* dest, const uint32_t* source1, int32_t len) {
+  int32_t i;
+  uint32_t oredDest = 0;
+  for (i = 0 ; i < len ; ++i) {
+    oredDest |= (dest[i] &= source1[i]);
+  }
+  return oredDest == 0;
+}
+
+// internal fn to count how many 1's are there in a mask
+// algorithm taken from  http://graphics.stanford.edu/~seander/bithacks.html
+static int16_t countOnes(uint32_t* mask, int32_t len) {
+  int32_t i, totalOnes = 0;
+  for (i = 0 ; i < len ; ++i) {
+    uint32_t ent = mask[i];
+    for (; ent; totalOnes++)
+    {
+      ent &= ent - 1; // clear the least significant bit set
+    }
+  }
+  return static_cast<int16_t>(totalOnes);
+}
+
+
+/* internal function! */
+static UEnumeration *selectForMask(const UConverterSelector* sel,
+                                   uint32_t *theMask, UErrorCode *status) {
+  LocalMemory<uint32_t> mask(theMask);
+  // this is the context we will use. Store a table of indices to which
+  // encodings are legit.
+  LocalMemory<Enumerator> result(static_cast<Enumerator *>(uprv_malloc(sizeof(Enumerator))));
+  if (result.isNull()) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return nullptr;
+  }
+  result->index = nullptr;  // this will be allocated later!
+  result->length = result->cur = 0;
+  result->sel = sel;
+
+  LocalMemory<UEnumeration> en(static_cast<UEnumeration *>(uprv_malloc(sizeof(UEnumeration))));
+  if (en.isNull()) {
+    // TODO(markus): Combine Enumerator and UEnumeration into one struct.
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return nullptr;
+  }
+  memcpy(en.getAlias(), &defaultEncodings, sizeof(UEnumeration));
+  
+  int32_t columns = (sel->encodingsCount+31)/32;
+  int16_t numOnes = countOnes(mask.getAlias(), columns);
+  // now, we know the exact space we need for index
+  if (numOnes > 0) {
+    result->index = static_cast<int16_t*>(uprv_malloc(numOnes * sizeof(int16_t)));
+    if (result->index == nullptr) {
+      *status = U_MEMORY_ALLOCATION_ERROR;
+      return nullptr;
+    }
+    int32_t i, j;
+    int16_t k = 0;
+    for (j = 0 ; j < columns; j++) {
+      uint32_t v = mask[j];
+      for (i = 0 ; i < 32 && k < sel->encodingsCount; i++, k++) {
+        if ((v & 1) != 0) {
+          result->index[result->length++] = k;
+        }
+        v >>= 1;
+      }
+    }
+  } //otherwise, index will remain NULL (and will never be touched by
+    //the enumerator code anyway)
+  en->context = result.orphan();
+  return en.orphan();
+}
+
+/* check a string against the selector - UTF16 version */
+U_CAPI UEnumeration * U_EXPORT2
+ucnvsel_selectForString(const UConverterSelector* sel,
+                        const UChar *s, int32_t length, UErrorCode *status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+  // ensure args make sense!
+  if (sel == NULL || (s == NULL && length != 0)) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+
+  int32_t columns = (sel->encodingsCount+31)/32;
+  uint32_t* mask = (uint32_t*) uprv_malloc(columns * 4);
+  if (mask == NULL) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+  uprv_memset(mask, ~0, columns *4);
+
+  if(s!=NULL) {
+    const UChar *limit;
+    if (length >= 0) {
+      limit = s + length;
+    } else {
+      limit = NULL;
+    }
+    
+    while (limit == NULL ? *s != 0 : s != limit) {
+      UChar32 c;
+      uint16_t pvIndex;
+      UTRIE2_U16_NEXT16(sel->trie, s, limit, c, pvIndex);
+      if (intersectMasks(mask, sel->pv+pvIndex, columns)) {
+        break;
+      }
+    }
+  }
+  return selectForMask(sel, mask, status);
+}
+
+/* check a string against the selector - UTF8 version */
+U_CAPI UEnumeration * U_EXPORT2
+ucnvsel_selectForUTF8(const UConverterSelector* sel,
+                      const char *s, int32_t length, UErrorCode *status) {
+  // check if already failed
+  if (U_FAILURE(*status)) {
+    return NULL;
+  }
+  // ensure args make sense!
+  if (sel == NULL || (s == NULL && length != 0)) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+
+  int32_t columns = (sel->encodingsCount+31)/32;
+  uint32_t* mask = (uint32_t*) uprv_malloc(columns * 4);
+  if (mask == NULL) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return NULL;
+  }
+  uprv_memset(mask, ~0, columns *4);
+
+  if (length < 0) {
+    length = (int32_t)uprv_strlen(s);
+  }
+
+  if(s!=NULL) {
+    const char *limit = s + length;
+    
+    while (s != limit) {
+      uint16_t pvIndex;
+      UTRIE2_U8_NEXT16(sel->trie, s, limit, pvIndex);
+      if (intersectMasks(mask, sel->pv+pvIndex, columns)) {
+        break;
+      }
+    }
+  }
+  return selectForMask(sel, mask, status);
+}
+
+#endif  // !UCONFIG_NO_CONVERSION
diff --git a/thirdparty/icu4c/common/ucol_data.h b/thirdparty/icu4c/common/ucol_data.h
new file mode 100644
index 0000000000..83f54abba1
--- /dev/null
+++ b/thirdparty/icu4c/common/ucol_data.h
@@ -0,0 +1,89 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2000-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucol_data.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011jul02
+*   created by: Markus Scherer
+*
+* Private implementation header for C/C++ collation.
+* Some file data structure definitions were moved here from i18n/ucol_imp.h
+* so that the common library (via ucol_swp.cpp) need not depend on the i18n library at all.
+*
+* We do not want to move the collation swapper to the i18n library because
+* a) the resource bundle swapper depends on it and would have to move too, and
+* b) we might want to eventually implement runtime data swapping,
+*    which might (or might not) be easier if all swappers are in the common library.
+*/
+
+#ifndef __UCOL_DATA_H__
+#define __UCOL_DATA_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_COLLATION
+
+/* let us know whether reserved fields are reset to zero or junked */
+#define UCOL_HEADER_MAGIC 0x20030618
+
+typedef struct {
+      int32_t size;
+      /* all the offsets are in bytes */
+      /* to get the address add to the header address and cast properly */
+      uint32_t options; /* these are the default options for the collator */
+      uint32_t UCAConsts; /* structure which holds values for indirect positioning and implicit ranges */
+      uint32_t contractionUCACombos;        /* this one is needed only for UCA, to copy the appropriate contractions */
+      uint32_t magic;            /* magic number - lets us know whether reserved data is reset or junked */
+      uint32_t mappingPosition;  /* const uint8_t *mappingPosition; */
+      uint32_t expansion;        /* uint32_t *expansion;            */
+      uint32_t contractionIndex; /* UChar *contractionIndex;        */
+      uint32_t contractionCEs;   /* uint32_t *contractionCEs;       */
+      uint32_t contractionSize;  /* needed for various closures */
+      /*int32_t latinOneMapping;*/ /* this is now handled in the trie itself *//* fast track to latin1 chars      */
+
+      uint32_t endExpansionCE;      /* array of last collation element in
+                                       expansion */
+      uint32_t expansionCESize;     /* array of maximum expansion size
+                                       corresponding to the expansion
+                                       collation elements with last element
+                                       in endExpansionCE*/
+      int32_t  endExpansionCECount; /* size of endExpansionCE */
+      uint32_t unsafeCP;            /* hash table of unsafe code points */
+      uint32_t contrEndCP;          /* hash table of final code points  */
+                                    /*   in contractions.               */
+
+      int32_t contractionUCACombosSize;     /* number of UCA contraction items. */
+                                            /*Length is contractionUCACombosSize*contractionUCACombosWidth*sizeof(UChar) */
+      UBool jamoSpecial;                    /* is jamoSpecial */
+      UBool isBigEndian;                    /* is this data big endian? from the UDataInfo header*/
+      uint8_t charSetFamily;                /* what is the charset family of this data from the UDataInfo header*/
+      uint8_t contractionUCACombosWidth;    /* width of UCA combos field */
+      UVersionInfo version;
+      UVersionInfo UCAVersion;              /* version of the UCA, read from file */
+      UVersionInfo UCDVersion;              /* UCD version, obtained by u_getUnicodeVersion */
+      UVersionInfo formatVersion;           /* format version from the UDataInfo header */
+      uint32_t scriptToLeadByte;            /* offset to script to lead collation byte mapping data */
+      uint32_t leadByteToScript;            /* offset to lead collation byte to script mapping data */
+      uint8_t reserved[76];                 /* for future use */
+} UCATableHeader;
+
+typedef struct {
+  uint32_t byteSize;
+  uint32_t tableSize;
+  uint32_t contsSize;
+  uint32_t table;
+  uint32_t conts;
+  UVersionInfo UCAVersion;              /* version of the UCA, read from file */
+  uint8_t padding[8];
+} InverseUCATableHeader;
+
+#endif  /* !UCONFIG_NO_COLLATION */
+
+#endif  /* __UCOL_DATA_H__ */
diff --git a/thirdparty/icu4c/common/ucol_swp.cpp b/thirdparty/icu4c/common/ucol_swp.cpp
new file mode 100644
index 0000000000..1af19863fa
--- /dev/null
+++ b/thirdparty/icu4c/common/ucol_swp.cpp
@@ -0,0 +1,615 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucol_swp.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003sep10
+*   created by: Markus W. Scherer
+*
+*   Swap collation binaries.
+*/
+
+#include "unicode/udata.h" /* UDataInfo */
+#include "utrie.h"
+#include "utrie2.h"
+#include "udataswp.h"
+#include "cmemory.h"
+#include "ucol_data.h"
+#include "ucol_swp.h"
+
+/* swapping ----------------------------------------------------------------- */
+
+#if !UCONFIG_NO_COLLATION
+
+U_CAPI UBool U_EXPORT2
+ucol_looksLikeCollationBinary(const UDataSwapper *ds,
+                              const void *inData, int32_t length) {
+    if(ds==NULL || inData==NULL || length<-1) {
+        return FALSE;
+    }
+
+    // First check for format version 4+ which has a standard data header.
+    UErrorCode errorCode=U_ZERO_ERROR;
+    (void)udata_swapDataHeader(ds, inData, -1, NULL, &errorCode);
+    if(U_SUCCESS(errorCode)) {
+        const UDataInfo &info=*(const UDataInfo *)((const char *)inData+4);
+        if(info.dataFormat[0]==0x55 &&   // dataFormat="UCol"
+                info.dataFormat[1]==0x43 &&
+                info.dataFormat[2]==0x6f &&
+                info.dataFormat[3]==0x6c) {
+            return TRUE;
+        }
+    }
+
+    // Else check for format version 3.
+    const UCATableHeader *inHeader=(const UCATableHeader *)inData;
+
+    /*
+     * The collation binary must contain at least the UCATableHeader,
+     * starting with its size field.
+     * sizeof(UCATableHeader)==42*4 in ICU 2.8
+     * check the length against the header size before reading the size field
+     */
+    UCATableHeader header;
+    uprv_memset(&header, 0, sizeof(header));
+    if(length<0) {
+        header.size=udata_readInt32(ds, inHeader->size);
+    } else if((length<(42*4) || length<(header.size=udata_readInt32(ds, inHeader->size)))) {
+        return FALSE;
+    }
+
+    header.magic=ds->readUInt32(inHeader->magic);
+    if(!(
+        header.magic==UCOL_HEADER_MAGIC &&
+        inHeader->formatVersion[0]==3 /*&&
+        inHeader->formatVersion[1]>=0*/
+    )) {
+        return FALSE;
+    }
+
+    if(inHeader->isBigEndian!=ds->inIsBigEndian || inHeader->charSetFamily!=ds->inCharset) {
+        return FALSE;
+    }
+
+    return TRUE;
+}
+
+namespace {
+
+/* swap a header-less collation formatVersion=3 binary, inside a resource bundle or ucadata.icu */
+int32_t
+swapFormatVersion3(const UDataSwapper *ds,
+                   const void *inData, int32_t length, void *outData,
+                   UErrorCode *pErrorCode) {
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    const UCATableHeader *inHeader;
+    UCATableHeader *outHeader;
+    UCATableHeader header;
+
+    uint32_t count;
+
+    /* argument checking in case we were not called from ucol_swap() */
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<-1 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData;
+    outBytes=(uint8_t *)outData;
+
+    inHeader=(const UCATableHeader *)inData;
+    outHeader=(UCATableHeader *)outData;
+
+    /*
+     * The collation binary must contain at least the UCATableHeader,
+     * starting with its size field.
+     * sizeof(UCATableHeader)==42*4 in ICU 2.8
+     * check the length against the header size before reading the size field
+     */
+    uprv_memset(&header, 0, sizeof(header));
+    if(length<0) {
+        header.size=udata_readInt32(ds, inHeader->size);
+    } else if((length<(42*4) || length<(header.size=udata_readInt32(ds, inHeader->size)))) {
+        udata_printError(ds, "ucol_swap(formatVersion=3): too few bytes (%d after header) for collation data\n",
+                         length);
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    header.magic=ds->readUInt32(inHeader->magic);
+    if(!(
+        header.magic==UCOL_HEADER_MAGIC &&
+        inHeader->formatVersion[0]==3 /*&&
+        inHeader->formatVersion[1]>=0*/
+    )) {
+        udata_printError(ds, "ucol_swap(formatVersion=3): magic 0x%08x or format version %02x.%02x is not a collation binary\n",
+                         header.magic,
+                         inHeader->formatVersion[0], inHeader->formatVersion[1]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    if(inHeader->isBigEndian!=ds->inIsBigEndian || inHeader->charSetFamily!=ds->inCharset) {
+        udata_printError(ds, "ucol_swap(formatVersion=3): endianness %d or charset %d does not match the swapper\n",
+                         inHeader->isBigEndian, inHeader->charSetFamily);
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    if(length>=0) {
+        /* copy everything, takes care of data that needs no swapping */
+        if(inBytes!=outBytes) {
+            uprv_memcpy(outBytes, inBytes, header.size);
+        }
+
+        /* swap the necessary pieces in the order of their occurrence in the data */
+
+        /* read more of the UCATableHeader (the size field was read above) */
+        header.options=                 ds->readUInt32(inHeader->options);
+        header.UCAConsts=               ds->readUInt32(inHeader->UCAConsts);
+        header.contractionUCACombos=    ds->readUInt32(inHeader->contractionUCACombos);
+        header.mappingPosition=         ds->readUInt32(inHeader->mappingPosition);
+        header.expansion=               ds->readUInt32(inHeader->expansion);
+        header.contractionIndex=        ds->readUInt32(inHeader->contractionIndex);
+        header.contractionCEs=          ds->readUInt32(inHeader->contractionCEs);
+        header.contractionSize=         ds->readUInt32(inHeader->contractionSize);
+        header.endExpansionCE=          ds->readUInt32(inHeader->endExpansionCE);
+        header.expansionCESize=         ds->readUInt32(inHeader->expansionCESize);
+        header.endExpansionCECount=     udata_readInt32(ds, inHeader->endExpansionCECount);
+        header.contractionUCACombosSize=udata_readInt32(ds, inHeader->contractionUCACombosSize);
+        header.scriptToLeadByte=        ds->readUInt32(inHeader->scriptToLeadByte);
+        header.leadByteToScript=        ds->readUInt32(inHeader->leadByteToScript);
+        
+        /* swap the 32-bit integers in the header */
+        ds->swapArray32(ds, inHeader, (int32_t)((const char *)&inHeader->jamoSpecial-(const char *)inHeader),
+                           outHeader, pErrorCode);
+        ds->swapArray32(ds, &(inHeader->scriptToLeadByte), sizeof(header.scriptToLeadByte) + sizeof(header.leadByteToScript),
+                           &(outHeader->scriptToLeadByte), pErrorCode);
+        /* set the output platform properties */
+        outHeader->isBigEndian=ds->outIsBigEndian;
+        outHeader->charSetFamily=ds->outCharset;
+
+        /* swap the options */
+        if(header.options!=0) {
+            ds->swapArray32(ds, inBytes+header.options, header.expansion-header.options,
+                               outBytes+header.options, pErrorCode);
+        }
+
+        /* swap the expansions */
+        if(header.mappingPosition!=0 && header.expansion!=0) {
+            if(header.contractionIndex!=0) {
+                /* expansions bounded by contractions */
+                count=header.contractionIndex-header.expansion;
+            } else {
+                /* no contractions: expansions bounded by the main trie */
+                count=header.mappingPosition-header.expansion;
+            }
+            ds->swapArray32(ds, inBytes+header.expansion, (int32_t)count,
+                               outBytes+header.expansion, pErrorCode);
+        }
+
+        /* swap the contractions */
+        if(header.contractionSize!=0) {
+            /* contractionIndex: UChar[] */
+            ds->swapArray16(ds, inBytes+header.contractionIndex, header.contractionSize*2,
+                               outBytes+header.contractionIndex, pErrorCode);
+
+            /* contractionCEs: CEs[] */
+            ds->swapArray32(ds, inBytes+header.contractionCEs, header.contractionSize*4,
+                               outBytes+header.contractionCEs, pErrorCode);
+        }
+
+        /* swap the main trie */
+        if(header.mappingPosition!=0) {
+            count=header.endExpansionCE-header.mappingPosition;
+            utrie_swap(ds, inBytes+header.mappingPosition, (int32_t)count,
+                          outBytes+header.mappingPosition, pErrorCode);
+        }
+
+        /* swap the max expansion table */
+        if(header.endExpansionCECount!=0) {
+            ds->swapArray32(ds, inBytes+header.endExpansionCE, header.endExpansionCECount*4,
+                               outBytes+header.endExpansionCE, pErrorCode);
+        }
+
+        /* expansionCESize, unsafeCP, contrEndCP: uint8_t[], no need to swap */
+
+        /* swap UCA constants */
+        if(header.UCAConsts!=0) {
+            /*
+             * if UCAConsts!=0 then contractionUCACombos because we are swapping
+             * the UCA data file, and we know that the UCA contains contractions
+             */
+            ds->swapArray32(ds, inBytes+header.UCAConsts, header.contractionUCACombos-header.UCAConsts,
+                               outBytes+header.UCAConsts, pErrorCode);
+        }
+
+        /* swap UCA contractions */
+        if(header.contractionUCACombosSize!=0) {
+            count=header.contractionUCACombosSize*inHeader->contractionUCACombosWidth*U_SIZEOF_UCHAR;
+            ds->swapArray16(ds, inBytes+header.contractionUCACombos, (int32_t)count,
+                               outBytes+header.contractionUCACombos, pErrorCode);
+        }
+        
+        /* swap the script to lead bytes */
+        if(header.scriptToLeadByte!=0) {
+            int indexCount = ds->readUInt16(*((uint16_t*)(inBytes+header.scriptToLeadByte))); // each entry = 2 * uint16
+            int dataCount = ds->readUInt16(*((uint16_t*)(inBytes+header.scriptToLeadByte + 2))); // each entry = uint16
+            ds->swapArray16(ds, inBytes+header.scriptToLeadByte, 
+                                4 + (4 * indexCount) + (2 * dataCount),
+                                outBytes+header.scriptToLeadByte, pErrorCode);
+        }
+        
+        /* swap the lead byte to scripts */
+        if(header.leadByteToScript!=0) {
+            int indexCount = ds->readUInt16(*((uint16_t*)(inBytes+header.leadByteToScript))); // each entry = uint16
+            int dataCount = ds->readUInt16(*((uint16_t*)(inBytes+header.leadByteToScript + 2))); // each entry = uint16
+            ds->swapArray16(ds, inBytes+header.leadByteToScript, 
+                                4 + (2 * indexCount) + (2 * dataCount),
+                                outBytes+header.leadByteToScript, pErrorCode);
+        }
+    }
+
+    return header.size;
+}
+
+// swap formatVersion 4 or 5 ----------------------------------------------- ***
+
+// The following are copied from CollationDataReader, trading an awkward copy of constants
+// for an awkward relocation of the i18n collationdatareader.h file into the common library.
+// Keep them in sync!
+
+enum {
+    IX_INDEXES_LENGTH,  // 0
+    IX_OPTIONS,
+    IX_RESERVED2,
+    IX_RESERVED3,
+
+    IX_JAMO_CE32S_START,  // 4
+    IX_REORDER_CODES_OFFSET,
+    IX_REORDER_TABLE_OFFSET,
+    IX_TRIE_OFFSET,
+
+    IX_RESERVED8_OFFSET,  // 8
+    IX_CES_OFFSET,
+    IX_RESERVED10_OFFSET,
+    IX_CE32S_OFFSET,
+
+    IX_ROOT_ELEMENTS_OFFSET,  // 12
+    IX_CONTEXTS_OFFSET,
+    IX_UNSAFE_BWD_OFFSET,
+    IX_FAST_LATIN_TABLE_OFFSET,
+
+    IX_SCRIPTS_OFFSET,  // 16
+    IX_COMPRESSIBLE_BYTES_OFFSET,
+    IX_RESERVED18_OFFSET,
+    IX_TOTAL_SIZE
+};
+
+int32_t
+swapFormatVersion4(const UDataSwapper *ds,
+                   const void *inData, int32_t length, void *outData,
+                   UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) { return 0; }
+
+    const uint8_t *inBytes=(const uint8_t *)inData;
+    uint8_t *outBytes=(uint8_t *)outData;
+
+    const int32_t *inIndexes=(const int32_t *)inBytes;
+    int32_t indexes[IX_TOTAL_SIZE+1];
+
+    // Need at least IX_INDEXES_LENGTH and IX_OPTIONS.
+    if(0<=length && length<8) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): too few bytes "
+                         "(%d after header) for collation data\n",
+                         length);
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    int32_t indexesLength=indexes[0]=udata_readInt32(ds, inIndexes[0]);
+    if(0<=length && length<(indexesLength*4)) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): too few bytes "
+                         "(%d after header) for collation data\n",
+                         length);
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    for(int32_t i=1; i<=IX_TOTAL_SIZE && i<indexesLength; ++i) {
+        indexes[i]=udata_readInt32(ds, inIndexes[i]);
+    }
+    for(int32_t i=indexesLength; i<=IX_TOTAL_SIZE; ++i) {
+        indexes[i]=-1;
+    }
+    inIndexes=NULL;  // Make sure we do not accidentally use these instead of indexes[].
+
+    // Get the total length of the data.
+    int32_t size;
+    if(indexesLength>IX_TOTAL_SIZE) {
+        size=indexes[IX_TOTAL_SIZE];
+    } else if(indexesLength>IX_REORDER_CODES_OFFSET) {
+        size=indexes[indexesLength-1];
+    } else {
+        size=indexesLength*4;
+    }
+    if(length<0) { return size; }
+
+    if(length<size) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): too few bytes "
+                         "(%d after header) for collation data\n",
+                         length);
+        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    // Copy the data for inaccessible bytes and arrays of bytes.
+    if(inBytes!=outBytes) {
+        uprv_memcpy(outBytes, inBytes, size);
+    }
+
+    // Swap the int32_t indexes[].
+    ds->swapArray32(ds, inBytes, indexesLength * 4, outBytes, &errorCode);
+
+    // The following is a modified version of CollationDataReader::read().
+    // Here we use indexes[] not inIndexes[] because
+    // the inIndexes[] may not be in this machine's endianness.
+    int32_t index;  // one of the indexes[] slots
+    int32_t offset;  // byte offset for the index part
+    // int32_t length;  // number of bytes in the index part
+
+    index = IX_REORDER_CODES_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray32(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    // Skip the IX_REORDER_TABLE_OFFSET byte array.
+
+    index = IX_TRIE_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        utrie2_swap(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_RESERVED8_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): unknown data at IX_RESERVED8_OFFSET\n", length);
+        errorCode = U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    index = IX_CES_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray64(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_RESERVED10_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): unknown data at IX_RESERVED10_OFFSET\n", length);
+        errorCode = U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    index = IX_CE32S_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray32(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_ROOT_ELEMENTS_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray32(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_CONTEXTS_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray16(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_UNSAFE_BWD_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray16(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_FAST_LATIN_TABLE_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray16(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    index = IX_SCRIPTS_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        ds->swapArray16(ds, inBytes + offset, length, outBytes + offset, &errorCode);
+    }
+
+    // Skip the  IX_COMPRESSIBLE_BYTES_OFFSET byte array.
+
+    index = IX_RESERVED18_OFFSET;
+    offset = indexes[index];
+    length = indexes[index + 1] - offset;
+    if(length > 0) {
+        udata_printError(ds, "ucol_swap(formatVersion=4): unknown data at IX_RESERVED18_OFFSET\n", length);
+        errorCode = U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    return size;
+}
+
+}  // namespace
+
+/* swap ICU collation data like ucadata.icu */
+U_CAPI int32_t U_EXPORT2
+ucol_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) { return 0; }
+
+    /* udata_swapDataHeader checks the arguments */
+    int32_t headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        // Try to swap the old format version which did not have a standard data header.
+        *pErrorCode=U_ZERO_ERROR;
+        return swapFormatVersion3(ds, inData, length, outData, pErrorCode);
+    }
+
+    /* check data format and format version */
+    const UDataInfo &info=*(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        info.dataFormat[0]==0x55 &&   // dataFormat="UCol"
+        info.dataFormat[1]==0x43 &&
+        info.dataFormat[2]==0x6f &&
+        info.dataFormat[3]==0x6c &&
+        (3<=info.formatVersion[0] && info.formatVersion[0]<=5)
+    )) {
+        udata_printError(ds, "ucol_swap(): data format %02x.%02x.%02x.%02x "
+                         "(format version %02x.%02x) is not recognized as collation data\n",
+                         info.dataFormat[0], info.dataFormat[1],
+                         info.dataFormat[2], info.dataFormat[3],
+                         info.formatVersion[0], info.formatVersion[1]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inData=(const char *)inData+headerSize;
+    if(length>=0) { length-=headerSize; }
+    outData=(char *)outData+headerSize;
+    int32_t collationSize;
+    if(info.formatVersion[0]>=4) {
+        collationSize=swapFormatVersion4(ds, inData, length, outData, *pErrorCode);
+    } else {
+        collationSize=swapFormatVersion3(ds, inData, length, outData, pErrorCode);
+    }
+    if(U_SUCCESS(*pErrorCode)) {
+        return headerSize+collationSize;
+    } else {
+        return 0;
+    }
+}
+
+/* swap inverse UCA collation data (invuca.icu) */
+U_CAPI int32_t U_EXPORT2
+ucol_swapInverseUCA(const UDataSwapper *ds,
+                    const void *inData, int32_t length, void *outData,
+                    UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    const InverseUCATableHeader *inHeader;
+    InverseUCATableHeader *outHeader;
+    InverseUCATableHeader header={ 0,0,0,0,0,{0,0,0,0},{0,0,0,0,0,0,0,0} };
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x49 &&   /* dataFormat="InvC" */
+        pInfo->dataFormat[1]==0x6e &&
+        pInfo->dataFormat[2]==0x76 &&
+        pInfo->dataFormat[3]==0x43 &&
+        pInfo->formatVersion[0]==2 &&
+        pInfo->formatVersion[1]>=1
+    )) {
+        udata_printError(ds, "ucol_swapInverseUCA(): data format %02x.%02x.%02x.%02x (format version %02x.%02x) is not an inverse UCA collation file\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0], pInfo->formatVersion[1]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData+headerSize;
+    outBytes=(uint8_t *)outData+headerSize;
+
+    inHeader=(const InverseUCATableHeader *)inBytes;
+    outHeader=(InverseUCATableHeader *)outBytes;
+
+    /*
+     * The inverse UCA collation binary must contain at least the InverseUCATableHeader,
+     * starting with its size field.
+     * sizeof(UCATableHeader)==8*4 in ICU 2.8
+     * check the length against the header size before reading the size field
+     */
+    if(length<0) {
+        header.byteSize=udata_readInt32(ds, inHeader->byteSize);
+    } else if(
+        ((length-headerSize)<(8*4) ||
+         (uint32_t)(length-headerSize)<(header.byteSize=udata_readInt32(ds, inHeader->byteSize)))
+    ) {
+        udata_printError(ds, "ucol_swapInverseUCA(): too few bytes (%d after header) for inverse UCA collation data\n",
+                         length);
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    if(length>=0) {
+        /* copy everything, takes care of data that needs no swapping */
+        if(inBytes!=outBytes) {
+            uprv_memcpy(outBytes, inBytes, header.byteSize);
+        }
+
+        /* swap the necessary pieces in the order of their occurrence in the data */
+
+        /* read more of the InverseUCATableHeader (the byteSize field was read above) */
+        header.tableSize=   ds->readUInt32(inHeader->tableSize);
+        header.contsSize=   ds->readUInt32(inHeader->contsSize);
+        header.table=       ds->readUInt32(inHeader->table);
+        header.conts=       ds->readUInt32(inHeader->conts);
+
+        /* swap the 32-bit integers in the header */
+        ds->swapArray32(ds, inHeader, 5*4, outHeader, pErrorCode);
+
+        /* swap the inverse table; tableSize counts uint32_t[3] rows */
+        ds->swapArray32(ds, inBytes+header.table, header.tableSize*3*4,
+                           outBytes+header.table, pErrorCode);
+
+        /* swap the continuation table; contsSize counts UChars */
+        ds->swapArray16(ds, inBytes+header.conts, header.contsSize*U_SIZEOF_UCHAR,
+                           outBytes+header.conts, pErrorCode);
+    }
+
+    return headerSize+header.byteSize;
+}
+
+#endif /* #if !UCONFIG_NO_COLLATION */
diff --git a/thirdparty/icu4c/common/ucol_swp.h b/thirdparty/icu4c/common/ucol_swp.h
new file mode 100644
index 0000000000..0c2990a85e
--- /dev/null
+++ b/thirdparty/icu4c/common/ucol_swp.h
@@ -0,0 +1,58 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucol_swp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003sep10
+*   created by: Markus W. Scherer
+*
+*   Swap collation binaries.
+*/
+
+#ifndef __UCOL_SWP_H__
+#define __UCOL_SWP_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_COLLATION
+
+#include "udataswp.h"
+
+/*
+ * Does the data look like a collation binary?
+ * @internal
+ */
+U_CAPI UBool U_EXPORT2
+ucol_looksLikeCollationBinary(const UDataSwapper *ds,
+                              const void *inData, int32_t length);
+
+/**
+ * Swap ICU collation data like ucadata.icu. See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucol_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode);
+
+/**
+ * Swap inverse UCA collation data (invuca.icu). See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucol_swapInverseUCA(const UDataSwapper *ds,
+                    const void *inData, int32_t length, void *outData,
+                    UErrorCode *pErrorCode);
+
+#endif /* #if !UCONFIG_NO_COLLATION */
+
+#endif
diff --git a/thirdparty/icu4c/common/ucptrie.cpp b/thirdparty/icu4c/common/ucptrie.cpp
new file mode 100644
index 0000000000..0004160a23
--- /dev/null
+++ b/thirdparty/icu4c/common/ucptrie.cpp
@@ -0,0 +1,601 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ucptrie.cpp (modified from utrie2.cpp)
+// created: 2017dec29 Markus W. Scherer
+
+// #define UCPTRIE_DEBUG
+#ifdef UCPTRIE_DEBUG
+#   include <stdio.h>
+#endif
+
+#include "unicode/utypes.h"
+#include "unicode/ucptrie.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "ucptrie_impl.h"
+
+U_CAPI UCPTrie * U_EXPORT2
+ucptrie_openFromBinary(UCPTrieType type, UCPTrieValueWidth valueWidth,
+                       const void *data, int32_t length, int32_t *pActualLength,
+                       UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+
+    if (length <= 0 || (U_POINTER_MASK_LSB(data, 3) != 0) ||
+            type < UCPTRIE_TYPE_ANY || UCPTRIE_TYPE_SMALL < type ||
+            valueWidth < UCPTRIE_VALUE_BITS_ANY || UCPTRIE_VALUE_BITS_8 < valueWidth) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+
+    // Enough data for a trie header?
+    if (length < (int32_t)sizeof(UCPTrieHeader)) {
+        *pErrorCode = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+
+    // Check the signature.
+    const UCPTrieHeader *header = (const UCPTrieHeader *)data;
+    if (header->signature != UCPTRIE_SIG) {
+        *pErrorCode = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+
+    int32_t options = header->options;
+    int32_t typeInt = (options >> 6) & 3;
+    int32_t valueWidthInt = options & UCPTRIE_OPTIONS_VALUE_BITS_MASK;
+    if (typeInt > UCPTRIE_TYPE_SMALL || valueWidthInt > UCPTRIE_VALUE_BITS_8 ||
+            (options & UCPTRIE_OPTIONS_RESERVED_MASK) != 0) {
+        *pErrorCode = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+    UCPTrieType actualType = (UCPTrieType)typeInt;
+    UCPTrieValueWidth actualValueWidth = (UCPTrieValueWidth)valueWidthInt;
+    if (type < 0) {
+        type = actualType;
+    }
+    if (valueWidth < 0) {
+        valueWidth = actualValueWidth;
+    }
+    if (type != actualType || valueWidth != actualValueWidth) {
+        *pErrorCode = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+
+    // Get the length values and offsets.
+    UCPTrie tempTrie;
+    uprv_memset(&tempTrie, 0, sizeof(tempTrie));
+    tempTrie.indexLength = header->indexLength;
+    tempTrie.dataLength =
+        ((options & UCPTRIE_OPTIONS_DATA_LENGTH_MASK) << 4) | header->dataLength;
+    tempTrie.index3NullOffset = header->index3NullOffset;
+    tempTrie.dataNullOffset =
+        ((options & UCPTRIE_OPTIONS_DATA_NULL_OFFSET_MASK) << 8) | header->dataNullOffset;
+
+    tempTrie.highStart = header->shiftedHighStart << UCPTRIE_SHIFT_2;
+    tempTrie.shifted12HighStart = (tempTrie.highStart + 0xfff) >> 12;
+    tempTrie.type = type;
+    tempTrie.valueWidth = valueWidth;
+
+    // Calculate the actual length.
+    int32_t actualLength = (int32_t)sizeof(UCPTrieHeader) + tempTrie.indexLength * 2;
+    if (valueWidth == UCPTRIE_VALUE_BITS_16) {
+        actualLength += tempTrie.dataLength * 2;
+    } else if (valueWidth == UCPTRIE_VALUE_BITS_32) {
+        actualLength += tempTrie.dataLength * 4;
+    } else {
+        actualLength += tempTrie.dataLength;
+    }
+    if (length < actualLength) {
+        *pErrorCode = U_INVALID_FORMAT_ERROR;  // Not enough bytes.
+        return nullptr;
+    }
+
+    // Allocate the trie.
+    UCPTrie *trie = (UCPTrie *)uprv_malloc(sizeof(UCPTrie));
+    if (trie == nullptr) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+    uprv_memcpy(trie, &tempTrie, sizeof(tempTrie));
+#ifdef UCPTRIE_DEBUG
+    trie->name = "fromSerialized";
+#endif
+
+    // Set the pointers to its index and data arrays.
+    const uint16_t *p16 = (const uint16_t *)(header + 1);
+    trie->index = p16;
+    p16 += trie->indexLength;
+
+    // Get the data.
+    int32_t nullValueOffset = trie->dataNullOffset;
+    if (nullValueOffset >= trie->dataLength) {
+        nullValueOffset = trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET;
+    }
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        trie->data.ptr16 = p16;
+        trie->nullValue = trie->data.ptr16[nullValueOffset];
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        trie->data.ptr32 = (const uint32_t *)p16;
+        trie->nullValue = trie->data.ptr32[nullValueOffset];
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        trie->data.ptr8 = (const uint8_t *)p16;
+        trie->nullValue = trie->data.ptr8[nullValueOffset];
+        break;
+    default:
+        // Unreachable because valueWidth was checked above.
+        *pErrorCode = U_INVALID_FORMAT_ERROR;
+        return nullptr;
+    }
+
+    if (pActualLength != nullptr) {
+        *pActualLength = actualLength;
+    }
+    return trie;
+}
+
+U_CAPI void U_EXPORT2
+ucptrie_close(UCPTrie *trie) {
+    uprv_free(trie);
+}
+
+U_CAPI UCPTrieType U_EXPORT2
+ucptrie_getType(const UCPTrie *trie) {
+    return (UCPTrieType)trie->type;
+}
+
+U_CAPI UCPTrieValueWidth U_EXPORT2
+ucptrie_getValueWidth(const UCPTrie *trie) {
+    return (UCPTrieValueWidth)trie->valueWidth;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalSmallIndex(const UCPTrie *trie, UChar32 c) {
+    int32_t i1 = c >> UCPTRIE_SHIFT_1;
+    if (trie->type == UCPTRIE_TYPE_FAST) {
+        U_ASSERT(0xffff < c && c < trie->highStart);
+        i1 += UCPTRIE_BMP_INDEX_LENGTH - UCPTRIE_OMITTED_BMP_INDEX_1_LENGTH;
+    } else {
+        U_ASSERT((uint32_t)c < (uint32_t)trie->highStart && trie->highStart > UCPTRIE_SMALL_LIMIT);
+        i1 += UCPTRIE_SMALL_INDEX_LENGTH;
+    }
+    int32_t i3Block = trie->index[
+        (int32_t)trie->index[i1] + ((c >> UCPTRIE_SHIFT_2) & UCPTRIE_INDEX_2_MASK)];
+    int32_t i3 = (c >> UCPTRIE_SHIFT_3) & UCPTRIE_INDEX_3_MASK;
+    int32_t dataBlock;
+    if ((i3Block & 0x8000) == 0) {
+        // 16-bit indexes
+        dataBlock = trie->index[i3Block + i3];
+    } else {
+        // 18-bit indexes stored in groups of 9 entries per 8 indexes.
+        i3Block = (i3Block & 0x7fff) + (i3 & ~7) + (i3 >> 3);
+        i3 &= 7;
+        dataBlock = ((int32_t)trie->index[i3Block++] << (2 + (2 * i3))) & 0x30000;
+        dataBlock |= trie->index[i3Block + i3];
+    }
+    return dataBlock + (c & UCPTRIE_SMALL_DATA_MASK);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalSmallU8Index(const UCPTrie *trie, int32_t lt1, uint8_t t2, uint8_t t3) {
+    UChar32 c = (lt1 << 12) | (t2 << 6) | t3;
+    if (c >= trie->highStart) {
+        // Possible because the UTF-8 macro compares with shifted12HighStart which may be higher.
+        return trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET;
+    }
+    return ucptrie_internalSmallIndex(trie, c);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalU8PrevIndex(const UCPTrie *trie, UChar32 c,
+                            const uint8_t *start, const uint8_t *src) {
+    int32_t i, length;
+    // Support 64-bit pointers by avoiding cast of arbitrary difference.
+    if ((src - start) <= 7) {
+        i = length = (int32_t)(src - start);
+    } else {
+        i = length = 7;
+        start = src - 7;
+    }
+    c = utf8_prevCharSafeBody(start, 0, &i, c, -1);
+    i = length - i;  // Number of bytes read backward from src.
+    int32_t idx = _UCPTRIE_CP_INDEX(trie, 0xffff, c);
+    return (idx << 3) | i;
+}
+
+namespace {
+
+inline uint32_t getValue(UCPTrieData data, UCPTrieValueWidth valueWidth, int32_t dataIndex) {
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        return data.ptr16[dataIndex];
+    case UCPTRIE_VALUE_BITS_32:
+        return data.ptr32[dataIndex];
+    case UCPTRIE_VALUE_BITS_8:
+        return data.ptr8[dataIndex];
+    default:
+        // Unreachable if the trie is properly initialized.
+        return 0xffffffff;
+    }
+}
+
+}  // namespace
+
+U_CAPI uint32_t U_EXPORT2
+ucptrie_get(const UCPTrie *trie, UChar32 c) {
+    int32_t dataIndex;
+    if ((uint32_t)c <= 0x7f) {
+        // linear ASCII
+        dataIndex = c;
+    } else {
+        UChar32 fastMax = trie->type == UCPTRIE_TYPE_FAST ? 0xffff : UCPTRIE_SMALL_MAX;
+        dataIndex = _UCPTRIE_CP_INDEX(trie, fastMax, c);
+    }
+    return getValue(trie->data, (UCPTrieValueWidth)trie->valueWidth, dataIndex);
+}
+
+namespace {
+
+constexpr int32_t MAX_UNICODE = 0x10ffff;
+
+inline uint32_t maybeFilterValue(uint32_t value, uint32_t trieNullValue, uint32_t nullValue,
+                                 UCPMapValueFilter *filter, const void *context) {
+    if (value == trieNullValue) {
+        value = nullValue;
+    } else if (filter != nullptr) {
+        value = filter(context, value);
+    }
+    return value;
+}
+
+UChar32 getRange(const void *t, UChar32 start,
+                 UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    if ((uint32_t)start > MAX_UNICODE) {
+        return U_SENTINEL;
+    }
+    const UCPTrie *trie = reinterpret_cast<const UCPTrie *>(t);
+    UCPTrieValueWidth valueWidth = (UCPTrieValueWidth)trie->valueWidth;
+    if (start >= trie->highStart) {
+        if (pValue != nullptr) {
+            int32_t di = trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET;
+            uint32_t value = getValue(trie->data, valueWidth, di);
+            if (filter != nullptr) { value = filter(context, value); }
+            *pValue = value;
+        }
+        return MAX_UNICODE;
+    }
+
+    uint32_t nullValue = trie->nullValue;
+    if (filter != nullptr) { nullValue = filter(context, nullValue); }
+    const uint16_t *index = trie->index;
+
+    int32_t prevI3Block = -1;
+    int32_t prevBlock = -1;
+    UChar32 c = start;
+    uint32_t trieValue, value = nullValue;
+    bool haveValue = false;
+    do {
+        int32_t i3Block;
+        int32_t i3;
+        int32_t i3BlockLength;
+        int32_t dataBlockLength;
+        if (c <= 0xffff && (trie->type == UCPTRIE_TYPE_FAST || c <= UCPTRIE_SMALL_MAX)) {
+            i3Block = 0;
+            i3 = c >> UCPTRIE_FAST_SHIFT;
+            i3BlockLength = trie->type == UCPTRIE_TYPE_FAST ?
+                UCPTRIE_BMP_INDEX_LENGTH : UCPTRIE_SMALL_INDEX_LENGTH;
+            dataBlockLength = UCPTRIE_FAST_DATA_BLOCK_LENGTH;
+        } else {
+            // Use the multi-stage index.
+            int32_t i1 = c >> UCPTRIE_SHIFT_1;
+            if (trie->type == UCPTRIE_TYPE_FAST) {
+                U_ASSERT(0xffff < c && c < trie->highStart);
+                i1 += UCPTRIE_BMP_INDEX_LENGTH - UCPTRIE_OMITTED_BMP_INDEX_1_LENGTH;
+            } else {
+                U_ASSERT(c < trie->highStart && trie->highStart > UCPTRIE_SMALL_LIMIT);
+                i1 += UCPTRIE_SMALL_INDEX_LENGTH;
+            }
+            i3Block = trie->index[
+                (int32_t)trie->index[i1] + ((c >> UCPTRIE_SHIFT_2) & UCPTRIE_INDEX_2_MASK)];
+            if (i3Block == prevI3Block && (c - start) >= UCPTRIE_CP_PER_INDEX_2_ENTRY) {
+                // The index-3 block is the same as the previous one, and filled with value.
+                U_ASSERT((c & (UCPTRIE_CP_PER_INDEX_2_ENTRY - 1)) == 0);
+                c += UCPTRIE_CP_PER_INDEX_2_ENTRY;
+                continue;
+            }
+            prevI3Block = i3Block;
+            if (i3Block == trie->index3NullOffset) {
+                // This is the index-3 null block.
+                if (haveValue) {
+                    if (nullValue != value) {
+                        return c - 1;
+                    }
+                } else {
+                    trieValue = trie->nullValue;
+                    value = nullValue;
+                    if (pValue != nullptr) { *pValue = nullValue; }
+                    haveValue = true;
+                }
+                prevBlock = trie->dataNullOffset;
+                c = (c + UCPTRIE_CP_PER_INDEX_2_ENTRY) & ~(UCPTRIE_CP_PER_INDEX_2_ENTRY - 1);
+                continue;
+            }
+            i3 = (c >> UCPTRIE_SHIFT_3) & UCPTRIE_INDEX_3_MASK;
+            i3BlockLength = UCPTRIE_INDEX_3_BLOCK_LENGTH;
+            dataBlockLength = UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+        }
+        // Enumerate data blocks for one index-3 block.
+        do {
+            int32_t block;
+            if ((i3Block & 0x8000) == 0) {
+                block = index[i3Block + i3];
+            } else {
+                // 18-bit indexes stored in groups of 9 entries per 8 indexes.
+                int32_t group = (i3Block & 0x7fff) + (i3 & ~7) + (i3 >> 3);
+                int32_t gi = i3 & 7;
+                block = ((int32_t)index[group++] << (2 + (2 * gi))) & 0x30000;
+                block |= index[group + gi];
+            }
+            if (block == prevBlock && (c - start) >= dataBlockLength) {
+                // The block is the same as the previous one, and filled with value.
+                U_ASSERT((c & (dataBlockLength - 1)) == 0);
+                c += dataBlockLength;
+            } else {
+                int32_t dataMask = dataBlockLength - 1;
+                prevBlock = block;
+                if (block == trie->dataNullOffset) {
+                    // This is the data null block.
+                    if (haveValue) {
+                        if (nullValue != value) {
+                            return c - 1;
+                        }
+                    } else {
+                        trieValue = trie->nullValue;
+                        value = nullValue;
+                        if (pValue != nullptr) { *pValue = nullValue; }
+                        haveValue = true;
+                    }
+                    c = (c + dataBlockLength) & ~dataMask;
+                } else {
+                    int32_t di = block + (c & dataMask);
+                    uint32_t trieValue2 = getValue(trie->data, valueWidth, di);
+                    if (haveValue) {
+                        if (trieValue2 != trieValue) {
+                            if (filter == nullptr ||
+                                    maybeFilterValue(trieValue2, trie->nullValue, nullValue,
+                                                     filter, context) != value) {
+                                return c - 1;
+                            }
+                            trieValue = trieValue2;  // may or may not help
+                        }
+                    } else {
+                        trieValue = trieValue2;
+                        value = maybeFilterValue(trieValue2, trie->nullValue, nullValue,
+                                                 filter, context);
+                        if (pValue != nullptr) { *pValue = value; }
+                        haveValue = true;
+                    }
+                    while ((++c & dataMask) != 0) {
+                        trieValue2 = getValue(trie->data, valueWidth, ++di);
+                        if (trieValue2 != trieValue) {
+                            if (filter == nullptr ||
+                                    maybeFilterValue(trieValue2, trie->nullValue, nullValue,
+                                                     filter, context) != value) {
+                                return c - 1;
+                            }
+                            trieValue = trieValue2;  // may or may not help
+                        }
+                    }
+                }
+            }
+        } while (++i3 < i3BlockLength);
+    } while (c < trie->highStart);
+    U_ASSERT(haveValue);
+    int32_t di = trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET;
+    uint32_t highValue = getValue(trie->data, valueWidth, di);
+    if (maybeFilterValue(highValue, trie->nullValue, nullValue,
+                         filter, context) != value) {
+        return c - 1;
+    } else {
+        return MAX_UNICODE;
+    }
+}
+
+}  // namespace
+
+U_CFUNC UChar32
+ucptrie_internalGetRange(UCPTrieGetRange *getRange,
+                         const void *trie, UChar32 start,
+                         UCPMapRangeOption option, uint32_t surrogateValue,
+                         UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    if (option == UCPMAP_RANGE_NORMAL) {
+        return getRange(trie, start, filter, context, pValue);
+    }
+    uint32_t value;
+    if (pValue == nullptr) {
+        // We need to examine the range value even if the caller does not want it.
+        pValue = &value;
+    }
+    UChar32 surrEnd = option == UCPMAP_RANGE_FIXED_ALL_SURROGATES ? 0xdfff : 0xdbff;
+    UChar32 end = getRange(trie, start, filter, context, pValue);
+    if (end < 0xd7ff || start > surrEnd) {
+        return end;
+    }
+    // The range overlaps with surrogates, or ends just before the first one.
+    if (*pValue == surrogateValue) {
+        if (end >= surrEnd) {
+            // Surrogates followed by a non-surrogateValue range,
+            // or surrogates are part of a larger surrogateValue range.
+            return end;
+        }
+    } else {
+        if (start <= 0xd7ff) {
+            return 0xd7ff;  // Non-surrogateValue range ends before surrogateValue surrogates.
+        }
+        // Start is a surrogate with a non-surrogateValue code *unit* value.
+        // Return a surrogateValue code *point* range.
+        *pValue = surrogateValue;
+        if (end > surrEnd) {
+            return surrEnd;  // Surrogate range ends before non-surrogateValue rest of range.
+        }
+    }
+    // See if the surrogateValue surrogate range can be merged with
+    // an immediately following range.
+    uint32_t value2;
+    UChar32 end2 = getRange(trie, surrEnd + 1, filter, context, &value2);
+    if (value2 == surrogateValue) {
+        return end2;
+    }
+    return surrEnd;
+}
+
+U_CAPI UChar32 U_EXPORT2
+ucptrie_getRange(const UCPTrie *trie, UChar32 start,
+                 UCPMapRangeOption option, uint32_t surrogateValue,
+                 UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    return ucptrie_internalGetRange(getRange, trie, start,
+                                    option, surrogateValue,
+                                    filter, context, pValue);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucptrie_toBinary(const UCPTrie *trie,
+                 void *data, int32_t capacity,
+                 UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    UCPTrieType type = (UCPTrieType)trie->type;
+    UCPTrieValueWidth valueWidth = (UCPTrieValueWidth)trie->valueWidth;
+    if (type < UCPTRIE_TYPE_FAST || UCPTRIE_TYPE_SMALL < type ||
+            valueWidth < UCPTRIE_VALUE_BITS_16 || UCPTRIE_VALUE_BITS_8 < valueWidth ||
+            capacity < 0 ||
+            (capacity > 0 && (data == nullptr || (U_POINTER_MASK_LSB(data, 3) != 0)))) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    int32_t length = (int32_t)sizeof(UCPTrieHeader) + trie->indexLength * 2;
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        length += trie->dataLength * 2;
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        length += trie->dataLength * 4;
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        length += trie->dataLength;
+        break;
+    default:
+        // unreachable
+        break;
+    }
+    if (capacity < length) {
+        *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+        return length;
+    }
+
+    char *bytes = (char *)data;
+    UCPTrieHeader *header = (UCPTrieHeader *)bytes;
+    header->signature = UCPTRIE_SIG;  // "Tri3"
+    header->options = (uint16_t)(
+        ((trie->dataLength & 0xf0000) >> 4) |
+        ((trie->dataNullOffset & 0xf0000) >> 8) |
+        (trie->type << 6) |
+        valueWidth);
+    header->indexLength = (uint16_t)trie->indexLength;
+    header->dataLength = (uint16_t)trie->dataLength;
+    header->index3NullOffset = trie->index3NullOffset;
+    header->dataNullOffset = (uint16_t)trie->dataNullOffset;
+    header->shiftedHighStart = trie->highStart >> UCPTRIE_SHIFT_2;
+    bytes += sizeof(UCPTrieHeader);
+
+    uprv_memcpy(bytes, trie->index, trie->indexLength * 2);
+    bytes += trie->indexLength * 2;
+
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        uprv_memcpy(bytes, trie->data.ptr16, trie->dataLength * 2);
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        uprv_memcpy(bytes, trie->data.ptr32, trie->dataLength * 4);
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        uprv_memcpy(bytes, trie->data.ptr8, trie->dataLength);
+        break;
+    default:
+        // unreachable
+        break;
+    }
+    return length;
+}
+
+namespace {
+
+#ifdef UCPTRIE_DEBUG
+long countNull(const UCPTrie *trie) {
+    uint32_t nullValue=trie->nullValue;
+    int32_t length=trie->dataLength;
+    long count=0;
+    switch (trie->valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        for(int32_t i=0; i<length; ++i) {
+            if(trie->data.ptr16[i]==nullValue) { ++count; }
+        }
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        for(int32_t i=0; i<length; ++i) {
+            if(trie->data.ptr32[i]==nullValue) { ++count; }
+        }
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        for(int32_t i=0; i<length; ++i) {
+            if(trie->data.ptr8[i]==nullValue) { ++count; }
+        }
+        break;
+    default:
+        // unreachable
+        break;
+    }
+    return count;
+}
+
+U_CFUNC void
+ucptrie_printLengths(const UCPTrie *trie, const char *which) {
+    long indexLength=trie->indexLength;
+    long dataLength=(long)trie->dataLength;
+    long totalLength=(long)sizeof(UCPTrieHeader)+indexLength*2+
+            dataLength*(trie->valueWidth==UCPTRIE_VALUE_BITS_16 ? 2 :
+                        trie->valueWidth==UCPTRIE_VALUE_BITS_32 ? 4 : 1);
+    printf("**UCPTrieLengths(%s %s)** index:%6ld  data:%6ld  countNull:%6ld  serialized:%6ld\n",
+           which, trie->name, indexLength, dataLength, countNull(trie), totalLength);
+}
+#endif
+
+}  // namespace
+
+// UCPMap ----
+// Initially, this is the same as UCPTrie. This may well change.
+
+U_CAPI uint32_t U_EXPORT2
+ucpmap_get(const UCPMap *map, UChar32 c) {
+    return ucptrie_get(reinterpret_cast<const UCPTrie *>(map), c);
+}
+
+U_CAPI UChar32 U_EXPORT2
+ucpmap_getRange(const UCPMap *map, UChar32 start,
+                UCPMapRangeOption option, uint32_t surrogateValue,
+                UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    return ucptrie_getRange(reinterpret_cast<const UCPTrie *>(map), start,
+                            option, surrogateValue,
+                            filter, context, pValue);
+}
diff --git a/thirdparty/icu4c/common/ucptrie_impl.h b/thirdparty/icu4c/common/ucptrie_impl.h
new file mode 100644
index 0000000000..1fe6a18ac5
--- /dev/null
+++ b/thirdparty/icu4c/common/ucptrie_impl.h
@@ -0,0 +1,289 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ucptrie_impl.h (modified from utrie2_impl.h)
+// created: 2017dec29 Markus W. Scherer
+
+#ifndef __UCPTRIE_IMPL_H__
+#define __UCPTRIE_IMPL_H__
+
+#include "unicode/ucptrie.h"
+#ifdef UCPTRIE_DEBUG
+#include "unicode/umutablecptrie.h"
+#endif
+
+// UCPTrie signature values, in platform endianness and opposite endianness.
+// The UCPTrie signature ASCII byte values spell "Tri3".
+#define UCPTRIE_SIG     0x54726933
+#define UCPTRIE_OE_SIG  0x33697254
+
+/**
+ * Header data for the binary, memory-mappable representation of a UCPTrie/CodePointTrie.
+ * @internal
+ */
+struct UCPTrieHeader {
+    /** "Tri3" in big-endian US-ASCII (0x54726933) */
+    uint32_t signature;
+
+    /**
+     * Options bit field:
+     * Bits 15..12: Data length bits 19..16.
+     * Bits 11..8: Data null block offset bits 19..16.
+     * Bits 7..6: UCPTrieType
+     * Bits 5..3: Reserved (0).
+     * Bits 2..0: UCPTrieValueWidth
+     */
+    uint16_t options;
+
+    /** Total length of the index tables. */
+    uint16_t indexLength;
+
+    /** Data length bits 15..0. */
+    uint16_t dataLength;
+
+    /** Index-3 null block offset, 0x7fff or 0xffff if none. */
+    uint16_t index3NullOffset;
+
+    /** Data null block offset bits 15..0, 0xfffff if none. */
+    uint16_t dataNullOffset;
+
+    /**
+     * First code point of the single-value range ending with U+10ffff,
+     * rounded up and then shifted right by UCPTRIE_SHIFT_2.
+     */
+    uint16_t shiftedHighStart;
+};
+
+/**
+ * Constants for use with UCPTrieHeader.options.
+ * @internal
+ */
+enum {
+    UCPTRIE_OPTIONS_DATA_LENGTH_MASK = 0xf000,
+    UCPTRIE_OPTIONS_DATA_NULL_OFFSET_MASK = 0xf00,
+    UCPTRIE_OPTIONS_RESERVED_MASK = 0x38,
+    UCPTRIE_OPTIONS_VALUE_BITS_MASK = 7,
+    /**
+     * Value for index3NullOffset which indicates that there is no index-3 null block.
+     * Bit 15 is unused for this value because this bit is used if the index-3 contains
+     * 18-bit indexes.
+     */
+    UCPTRIE_NO_INDEX3_NULL_OFFSET = 0x7fff,
+    UCPTRIE_NO_DATA_NULL_OFFSET = 0xfffff
+};
+
+// Internal constants.
+enum {
+    /** The length of the BMP index table. 1024=0x400 */
+    UCPTRIE_BMP_INDEX_LENGTH = 0x10000 >> UCPTRIE_FAST_SHIFT,
+
+    UCPTRIE_SMALL_LIMIT = 0x1000,
+    UCPTRIE_SMALL_INDEX_LENGTH = UCPTRIE_SMALL_LIMIT >> UCPTRIE_FAST_SHIFT,
+
+    /** Shift size for getting the index-3 table offset. */
+    UCPTRIE_SHIFT_3 = 4,
+
+    /** Shift size for getting the index-2 table offset. */
+    UCPTRIE_SHIFT_2 = 5 + UCPTRIE_SHIFT_3,
+
+    /** Shift size for getting the index-1 table offset. */
+    UCPTRIE_SHIFT_1 = 5 + UCPTRIE_SHIFT_2,
+
+    /**
+     * Difference between two shift sizes,
+     * for getting an index-2 offset from an index-3 offset. 5=9-4
+     */
+    UCPTRIE_SHIFT_2_3 = UCPTRIE_SHIFT_2 - UCPTRIE_SHIFT_3,
+
+    /**
+     * Difference between two shift sizes,
+     * for getting an index-1 offset from an index-2 offset. 5=14-9
+     */
+    UCPTRIE_SHIFT_1_2 = UCPTRIE_SHIFT_1 - UCPTRIE_SHIFT_2,
+
+    /**
+     * Number of index-1 entries for the BMP. (4)
+     * This part of the index-1 table is omitted from the serialized form.
+     */
+    UCPTRIE_OMITTED_BMP_INDEX_1_LENGTH = 0x10000 >> UCPTRIE_SHIFT_1,
+
+    /** Number of entries in an index-2 block. 32=0x20 */
+    UCPTRIE_INDEX_2_BLOCK_LENGTH = 1 << UCPTRIE_SHIFT_1_2,
+
+    /** Mask for getting the lower bits for the in-index-2-block offset. */
+    UCPTRIE_INDEX_2_MASK = UCPTRIE_INDEX_2_BLOCK_LENGTH - 1,
+
+    /** Number of code points per index-2 table entry. 512=0x200 */
+    UCPTRIE_CP_PER_INDEX_2_ENTRY = 1 << UCPTRIE_SHIFT_2,
+
+    /** Number of entries in an index-3 block. 32=0x20 */
+    UCPTRIE_INDEX_3_BLOCK_LENGTH = 1 << UCPTRIE_SHIFT_2_3,
+
+    /** Mask for getting the lower bits for the in-index-3-block offset. */
+    UCPTRIE_INDEX_3_MASK = UCPTRIE_INDEX_3_BLOCK_LENGTH - 1,
+
+    /** Number of entries in a small data block. 16=0x10 */
+    UCPTRIE_SMALL_DATA_BLOCK_LENGTH = 1 << UCPTRIE_SHIFT_3,
+
+    /** Mask for getting the lower bits for the in-small-data-block offset. */
+    UCPTRIE_SMALL_DATA_MASK = UCPTRIE_SMALL_DATA_BLOCK_LENGTH - 1
+};
+
+typedef UChar32
+UCPTrieGetRange(const void *trie, UChar32 start,
+                UCPMapValueFilter *filter, const void *context, uint32_t *pValue);
+
+U_CFUNC UChar32
+ucptrie_internalGetRange(UCPTrieGetRange *getRange,
+                         const void *trie, UChar32 start,
+                         UCPMapRangeOption option, uint32_t surrogateValue,
+                         UCPMapValueFilter *filter, const void *context, uint32_t *pValue);
+
+#ifdef UCPTRIE_DEBUG
+U_CFUNC void
+ucptrie_printLengths(const UCPTrie *trie, const char *which);
+
+U_CFUNC void umutablecptrie_setName(UMutableCPTrie *builder, const char *name);
+#endif
+
+/*
+ * Format of the binary, memory-mappable representation of a UCPTrie/CodePointTrie.
+ * For overview information see http://site.icu-project.org/design/struct/utrie
+ *
+ * The binary trie data should be 32-bit-aligned.
+ * The overall layout is:
+ *
+ * UCPTrieHeader header; -- 16 bytes, see struct definition above
+ * uint16_t index[header.indexLength];
+ * uintXY_t data[header.dataLength];
+ *
+ * The trie data array is an array of uint16_t, uint32_t, or uint8_t,
+ * specified via the UCPTrieValueWidth when building the trie.
+ * The data array is 32-bit-aligned for uint32_t, otherwise 16-bit-aligned.
+ * The overall length of the trie data is a multiple of 4 bytes.
+ * (Padding is added at the end of the index array and/or near the end of the data array as needed.)
+ *
+ * The length of the data array (dataLength) is stored as an integer split across two fields
+ * of the header struct (high bits in header.options).
+ *
+ * The trie type can be "fast" or "small" which determines the index structure,
+ * specified via the UCPTrieType when building the trie.
+ *
+ * The type and valueWidth are stored in the header.options.
+ * There are reserved type and valueWidth values, and reserved header.options bits.
+ * They could be used in future format extensions.
+ * Code reading the trie structure must fail with an error when unknown values or options are set.
+ *
+ * Values for ASCII character (U+0000..U+007F) can always be found at the start of the data array.
+ *
+ * Values for code points below a type-specific fast-indexing limit are found via two-stage lookup.
+ * For a "fast" trie, the limit is the BMP/supplementary boundary at U+10000.
+ * For a "small" trie, the limit is UCPTRIE_SMALL_MAX+1=U+1000.
+ *
+ * All code points in the range highStart..U+10FFFF map to a single highValue
+ * which is stored at the second-to-last position of the data array.
+ * (See UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET.)
+ * The highStart value is header.shiftedHighStart<<UCPTRIE_SHIFT_2.
+ * (UCPTRIE_SHIFT_2=9)
+ *
+ * Values for code points fast_limit..highStart-1 are found via four-stage lookup.
+ * The data block size is smaller for this range than for the fast range.
+ * This together with more index stages with small blocks makes this range
+ * more easily compactable.
+ *
+ * There is also a trie error value stored at the last position of the data array.
+ * (See UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET.)
+ * It is intended to be returned for inputs that are not Unicode code points
+ * (outside U+0000..U+10FFFF), or in string processing for ill-formed input
+ * (unpaired surrogate in UTF-16, ill-formed UTF-8 subsequence).
+ *
+ * For a "fast" trie:
+ *
+ * The index array starts with the BMP index table for BMP code point lookup.
+ * Its length is 1024=0x400.
+ *
+ * The supplementary index-1 table follows the BMP index table.
+ * Variable length, for code points up to highStart-1.
+ * Maximum length 64=0x40=0x100000>>UCPTRIE_SHIFT_1.
+ * (For 0x100000 supplementary code points U+10000..U+10ffff.)
+ *
+ * After this index-1 table follow the variable-length index-3 and index-2 tables.
+ *
+ * The supplementary index tables are omitted completely
+ * if there is only BMP data (highStart<=U+10000).
+ *
+ * For a "small" trie:
+ *
+ * The index array starts with a fast-index table for lookup of code points U+0000..U+0FFF.
+ *
+ * The "supplementary" index tables are always stored.
+ * The index-1 table starts from U+0000, its maximum length is 68=0x44=0x110000>>UCPTRIE_SHIFT_1.
+ *
+ * For both trie types:
+ *
+ * The last index-2 block may be a partial block, storing indexes only for code points
+ * below highStart.
+ *
+ * Lookup for ASCII code point c:
+ *
+ * Linear access from the start of the data array.
+ *
+ * value = data[c];
+ *
+ * Lookup for fast-range code point c:
+ *
+ * Shift the code point right by UCPTRIE_FAST_SHIFT=6 bits,
+ * fetch the index array value at that offset,
+ * add the lower code point bits, index into the data array.
+ *
+ * value = data[index[c>>6] + (c&0x3f)];
+ *
+ * (This works for ASCII as well.)
+ *
+ * Lookup for small-range code point c below highStart:
+ *
+ * Split the code point into four bit fields using several sets of shifts & masks
+ * to read consecutive values from the index-1, index-2, index-3 and data tables.
+ *
+ * If all of the data block offsets in an index-3 block fit within 16 bits (up to 0xffff),
+ * then the data block offsets are stored directly as uint16_t.
+ *
+ * Otherwise (this is very unusual but possible), the index-2 entry for the index-3 block
+ * has bit 15 (0x8000) set, and each set of 8 index-3 entries is preceded by
+ * an additional uint16_t word. Data block offsets are 18 bits wide, with the top 2 bits stored
+ * in the additional word.
+ *
+ * See ucptrie_internalSmallIndex() for details.
+ *
+ * (In a "small" trie, this works for ASCII and below-fast_limit code points as well.)
+ *
+ * Compaction:
+ *
+ * Multiple code point ranges ("blocks") that are aligned on certain boundaries
+ * (determined by the shifting/bit fields of code points) and
+ * map to the same data values normally share a single subsequence of the data array.
+ * Data blocks can also overlap partially.
+ * (Depending on the builder code finding duplicate and overlapping blocks.)
+ *
+ * Iteration over same-value ranges:
+ *
+ * Range iteration (ucptrie_getRange()) walks the structure from a start code point
+ * until some code point is found that maps to a different value;
+ * the end of the returned range is just before that.
+ *
+ * The header.dataNullOffset (split across two header fields, high bits in header.options)
+ * is the offset of a widely shared data block filled with one single value.
+ * It helps quickly skip over large ranges of data with that value.
+ * The builder must ensure that if the start of any data block (fast or small)
+ * matches the dataNullOffset, then the whole block must be filled with the null value.
+ * Special care must be taken if there is no fast null data block
+ * but a small one, which is shorter, and it matches the *start* of some fast data block.
+ *
+ * Similarly, the header.index3NullOffset is the index-array offset of an index-3 block
+ * where all index entries point to the dataNullOffset.
+ * If there is no such data or index-3 block, then these offsets are set to
+ * values that cannot be reached (data offset out of range/reserved index offset),
+ * normally UCPTRIE_NO_DATA_NULL_OFFSET or UCPTRIE_NO_INDEX3_NULL_OFFSET respectively.
+ */
+
+#endif
diff --git a/thirdparty/icu4c/common/ucurr.cpp b/thirdparty/icu4c/common/ucurr.cpp
new file mode 100644
index 0000000000..0e14cddcff
--- /dev/null
+++ b/thirdparty/icu4c/common/ucurr.cpp
@@ -0,0 +1,2701 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/ucurr.h"
+#include "unicode/locid.h"
+#include "unicode/ures.h"
+#include "unicode/ustring.h"
+#include "unicode/parsepos.h"
+#include "unicode/uniset.h"
+#include "unicode/usetiter.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "static_unicode_sets.h"
+#include "uassert.h"
+#include "umutex.h"
+#include "ucln_cmn.h"
+#include "uenumimp.h"
+#include "uhash.h"
+#include "hash.h"
+#include "uinvchar.h"
+#include "uresimp.h"
+#include "ulist.h"
+#include "uresimp.h"
+#include "ureslocs.h"
+#include "ulocimp.h"
+
+using namespace icu;
+
+//#define UCURR_DEBUG_EQUIV 1
+#ifdef UCURR_DEBUG_EQUIV
+#include "stdio.h"
+#endif
+//#define UCURR_DEBUG 1
+#ifdef UCURR_DEBUG
+#include "stdio.h"
+#endif
+
+typedef struct IsoCodeEntry {
+    const UChar *isoCode; /* const because it's a reference to a resource bundle string. */
+    UDate from;
+    UDate to;
+} IsoCodeEntry;
+
+//------------------------------------------------------------
+// Constants
+
+// Default currency meta data of last resort.  We try to use the
+// defaults encoded in the meta data resource bundle.  If there is a
+// configuration/build error and these are not available, we use these
+// hard-coded defaults (which should be identical).
+static const int32_t LAST_RESORT_DATA[] = { 2, 0, 2, 0 };
+
+// POW10[i] = 10^i, i=0..MAX_POW10
+static const int32_t POW10[] = { 1, 10, 100, 1000, 10000, 100000,
+                                 1000000, 10000000, 100000000, 1000000000 };
+
+static const int32_t MAX_POW10 = UPRV_LENGTHOF(POW10) - 1;
+
+#define ISO_CURRENCY_CODE_LENGTH 3
+
+//------------------------------------------------------------
+// Resource tags
+//
+
+static const char CURRENCY_DATA[] = "supplementalData";
+// Tag for meta-data, in root.
+static const char CURRENCY_META[] = "CurrencyMeta";
+
+// Tag for map from countries to currencies, in root.
+static const char CURRENCY_MAP[] = "CurrencyMap";
+
+// Tag for default meta-data, in CURRENCY_META
+static const char DEFAULT_META[] = "DEFAULT";
+
+// Variant delimiter
+static const char VAR_DELIM = '_';
+
+// Tag for localized display names (symbols) of currencies
+static const char CURRENCIES[] = "Currencies";
+static const char CURRENCIES_NARROW[] = "Currencies%narrow";
+static const char CURRENCIES_FORMAL[] = "Currencies%formal";
+static const char CURRENCIES_VARIANT[] = "Currencies%variant";
+static const char CURRENCYPLURALS[] = "CurrencyPlurals";
+
+// ISO codes mapping table
+static const UHashtable* gIsoCodes = NULL;
+static icu::UInitOnce gIsoCodesInitOnce = U_INITONCE_INITIALIZER;
+
+// Currency symbol equivalances
+static const icu::Hashtable* gCurrSymbolsEquiv = NULL;
+static icu::UInitOnce gCurrSymbolsEquivInitOnce = U_INITONCE_INITIALIZER;
+
+U_NAMESPACE_BEGIN
+
+// EquivIterator iterates over all strings that are equivalent to a given
+// string, s. Note that EquivIterator will never yield s itself.
+class EquivIterator : public icu::UMemory {
+public:
+    // Constructor. hash stores the equivalence relationships; s is the string
+    // for which we find equivalent strings.
+    inline EquivIterator(const icu::Hashtable& hash, const icu::UnicodeString& s)
+        : _hash(hash) { 
+        _start = _current = &s;
+    }
+    inline ~EquivIterator() { }
+
+    // next returns the next equivalent string or NULL if there are no more.
+    // If s has no equivalent strings, next returns NULL on the first call.
+    const icu::UnicodeString *next();
+private:
+    const icu::Hashtable& _hash;
+    const icu::UnicodeString* _start;
+    const icu::UnicodeString* _current;
+};
+
+const icu::UnicodeString *
+EquivIterator::next() {
+    const icu::UnicodeString* _next = (const icu::UnicodeString*) _hash.get(*_current);
+    if (_next == NULL) {
+        U_ASSERT(_current == _start);
+        return NULL;
+    }
+    if (*_next == *_start) {
+        return NULL;
+    }
+    _current = _next;
+    return _next;
+}
+
+U_NAMESPACE_END
+
+// makeEquivalent makes lhs and rhs equivalent by updating the equivalence
+// relations in hash accordingly.
+static void makeEquivalent(
+    const icu::UnicodeString &lhs,
+    const icu::UnicodeString &rhs,
+    icu::Hashtable* hash, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (lhs == rhs) {
+        // already equivalent
+        return;
+    }
+    icu::EquivIterator leftIter(*hash, lhs);
+    icu::EquivIterator rightIter(*hash, rhs);
+    const icu::UnicodeString *firstLeft = leftIter.next();
+    const icu::UnicodeString *firstRight = rightIter.next();
+    const icu::UnicodeString *nextLeft = firstLeft;
+    const icu::UnicodeString *nextRight = firstRight;
+    while (nextLeft != NULL && nextRight != NULL) {
+        if (*nextLeft == rhs || *nextRight == lhs) {
+            // Already equivalent
+            return;
+        }
+        nextLeft = leftIter.next();
+        nextRight = rightIter.next();
+    }
+    // Not equivalent. Must join.
+    icu::UnicodeString *newFirstLeft;
+    icu::UnicodeString *newFirstRight;
+    if (firstRight == NULL && firstLeft == NULL) {
+        // Neither lhs or rhs belong to an equivalence circle, so we form
+        // a new equivalnce circle of just lhs and rhs.
+        newFirstLeft = new icu::UnicodeString(rhs);
+        newFirstRight = new icu::UnicodeString(lhs);
+    } else if (firstRight == NULL) {
+        // lhs belongs to an equivalence circle, but rhs does not, so we link
+        // rhs into lhs' circle.
+        newFirstLeft = new icu::UnicodeString(rhs);
+        newFirstRight = new icu::UnicodeString(*firstLeft);
+    } else if (firstLeft == NULL) {
+        // rhs belongs to an equivlance circle, but lhs does not, so we link
+        // lhs into rhs' circle.
+        newFirstLeft = new icu::UnicodeString(*firstRight);
+        newFirstRight = new icu::UnicodeString(lhs);
+    } else {
+        // Both lhs and rhs belong to different equivalnce circles. We link
+        // them together to form one single, larger equivalnce circle.
+        newFirstLeft = new icu::UnicodeString(*firstRight);
+        newFirstRight = new icu::UnicodeString(*firstLeft);
+    }
+    if (newFirstLeft == NULL || newFirstRight == NULL) {
+        delete newFirstLeft;
+        delete newFirstRight;
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    hash->put(lhs, (void *) newFirstLeft, status);
+    hash->put(rhs, (void *) newFirstRight, status);
+}
+
+// countEquivalent counts how many strings are equivalent to s.
+// hash stores all the equivalnce relations.
+// countEquivalent does not include s itself in the count.
+static int32_t countEquivalent(const icu::Hashtable &hash, const icu::UnicodeString &s) {
+    int32_t result = 0;
+    icu::EquivIterator iter(hash, s);
+    while (iter.next() != NULL) {
+        ++result;
+    }
+#ifdef UCURR_DEBUG_EQUIV
+ {
+   char tmp[200];
+   s.extract(0,s.length(),tmp, "UTF-8");
+   printf("CountEquivalent('%s') = %d\n", tmp, result);
+ }
+#endif
+    return result;
+}
+
+static const icu::Hashtable* getCurrSymbolsEquiv();
+
+//------------------------------------------------------------
+// Code
+
+/**
+ * Cleanup callback func
+ */
+static UBool U_CALLCONV 
+isoCodes_cleanup(void)
+{
+    if (gIsoCodes != NULL) {
+        uhash_close(const_cast<UHashtable *>(gIsoCodes));
+        gIsoCodes = NULL;
+    }
+    gIsoCodesInitOnce.reset();
+    return TRUE;
+}
+
+/**
+ * Cleanup callback func
+ */
+static UBool U_CALLCONV 
+currSymbolsEquiv_cleanup(void)
+{
+    delete const_cast<icu::Hashtable *>(gCurrSymbolsEquiv);
+    gCurrSymbolsEquiv = NULL;
+    gCurrSymbolsEquivInitOnce.reset();
+    return TRUE;
+}
+
+/**
+ * Deleter for OlsonToMetaMappingEntry
+ */
+static void U_CALLCONV
+deleteIsoCodeEntry(void *obj) {
+    IsoCodeEntry *entry = (IsoCodeEntry*)obj;
+    uprv_free(entry);
+}
+
+/**
+ * Deleter for gCurrSymbolsEquiv.
+ */
+static void U_CALLCONV
+deleteUnicode(void *obj) {
+    icu::UnicodeString *entry = (icu::UnicodeString*)obj;
+    delete entry;
+}
+
+/**
+ * Unfortunately, we have to convert the UChar* currency code to char*
+ * to use it as a resource key.
+ */
+static inline char*
+myUCharsToChars(char* resultOfLen4, const UChar* currency) {
+    u_UCharsToChars(currency, resultOfLen4, ISO_CURRENCY_CODE_LENGTH);
+    resultOfLen4[ISO_CURRENCY_CODE_LENGTH] = 0;
+    return resultOfLen4;
+}
+
+/**
+ * Internal function to look up currency data.  Result is an array of
+ * four integers.  The first is the fraction digits.  The second is the
+ * rounding increment, or 0 if none.  The rounding increment is in
+ * units of 10^(-fraction_digits).  The third and fourth are the same
+ * except that they are those used in cash transations ( cashDigits
+ * and cashRounding ).
+ */
+static const int32_t*
+_findMetaData(const UChar* currency, UErrorCode& ec) {
+
+    if (currency == 0 || *currency == 0) {
+        if (U_SUCCESS(ec)) {
+            ec = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return LAST_RESORT_DATA;
+    }
+
+    // Get CurrencyMeta resource out of root locale file.  [This may
+    // move out of the root locale file later; if it does, update this
+    // code.]
+    UResourceBundle* currencyData = ures_openDirect(U_ICUDATA_CURR, CURRENCY_DATA, &ec);
+    UResourceBundle* currencyMeta = ures_getByKey(currencyData, CURRENCY_META, currencyData, &ec);
+
+    if (U_FAILURE(ec)) {
+        ures_close(currencyMeta);
+        // Config/build error; return hard-coded defaults
+        return LAST_RESORT_DATA;
+    }
+
+    // Look up our currency, or if that's not available, then DEFAULT
+    char buf[ISO_CURRENCY_CODE_LENGTH+1];
+    UErrorCode ec2 = U_ZERO_ERROR; // local error code: soft failure
+    UResourceBundle* rb = ures_getByKey(currencyMeta, myUCharsToChars(buf, currency), NULL, &ec2);
+      if (U_FAILURE(ec2)) {
+        ures_close(rb);
+        rb = ures_getByKey(currencyMeta,DEFAULT_META, NULL, &ec);
+        if (U_FAILURE(ec)) {
+            ures_close(currencyMeta);
+            ures_close(rb);
+            // Config/build error; return hard-coded defaults
+            return LAST_RESORT_DATA;
+        }
+    }
+
+    int32_t len;
+    const int32_t *data = ures_getIntVector(rb, &len, &ec);
+    if (U_FAILURE(ec) || len != 4) {
+        // Config/build error; return hard-coded defaults
+        if (U_SUCCESS(ec)) {
+            ec = U_INVALID_FORMAT_ERROR;
+        }
+        ures_close(currencyMeta);
+        ures_close(rb);
+        return LAST_RESORT_DATA;
+    }
+
+    ures_close(currencyMeta);
+    ures_close(rb);
+    return data;
+}
+
+// -------------------------------------
+
+static void
+idForLocale(const char* locale, char* countryAndVariant, int capacity, UErrorCode* ec)
+{
+    ulocimp_getRegionForSupplementalData(locale, FALSE, countryAndVariant, capacity, ec);
+}
+
+// ------------------------------------------
+//
+// Registration
+//
+//-------------------------------------------
+
+// don't use ICUService since we don't need fallback
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV currency_cleanup(void);
+U_CDECL_END
+
+#if !UCONFIG_NO_SERVICE
+struct CReg;
+
+static UMutex gCRegLock;
+static CReg* gCRegHead = 0;
+
+struct CReg : public icu::UMemory {
+    CReg *next;
+    UChar iso[ISO_CURRENCY_CODE_LENGTH+1];
+    char  id[ULOC_FULLNAME_CAPACITY];
+
+    CReg(const UChar* _iso, const char* _id)
+        : next(0)
+    {
+        int32_t len = (int32_t)uprv_strlen(_id);
+        if (len > (int32_t)(sizeof(id)-1)) {
+            len = (sizeof(id)-1);
+        }
+        uprv_strncpy(id, _id, len);
+        id[len] = 0;
+        u_memcpy(iso, _iso, ISO_CURRENCY_CODE_LENGTH);
+        iso[ISO_CURRENCY_CODE_LENGTH] = 0;
+    }
+
+    static UCurrRegistryKey reg(const UChar* _iso, const char* _id, UErrorCode* status)
+    {
+        if (status && U_SUCCESS(*status) && _iso && _id) {
+            CReg* n = new CReg(_iso, _id);
+            if (n) {
+                umtx_lock(&gCRegLock);
+                if (!gCRegHead) {
+                    /* register for the first time */
+                    ucln_common_registerCleanup(UCLN_COMMON_CURRENCY, currency_cleanup);
+                }
+                n->next = gCRegHead;
+                gCRegHead = n;
+                umtx_unlock(&gCRegLock);
+                return n;
+            }
+            *status = U_MEMORY_ALLOCATION_ERROR;
+        }
+        return 0;
+    }
+
+    static UBool unreg(UCurrRegistryKey key) {
+        UBool found = FALSE;
+        umtx_lock(&gCRegLock);
+
+        CReg** p = &gCRegHead;
+        while (*p) {
+            if (*p == key) {
+                *p = ((CReg*)key)->next;
+                delete (CReg*)key;
+                found = TRUE;
+                break;
+            }
+            p = &((*p)->next);
+        }
+
+        umtx_unlock(&gCRegLock);
+        return found;
+    }
+
+    static const UChar* get(const char* id) {
+        const UChar* result = NULL;
+        umtx_lock(&gCRegLock);
+        CReg* p = gCRegHead;
+
+        /* register cleanup of the mutex */
+        ucln_common_registerCleanup(UCLN_COMMON_CURRENCY, currency_cleanup);
+        while (p) {
+            if (uprv_strcmp(id, p->id) == 0) {
+                result = p->iso;
+                break;
+            }
+            p = p->next;
+        }
+        umtx_unlock(&gCRegLock);
+        return result;
+    }
+
+    /* This doesn't need to be thread safe. It's for u_cleanup only. */
+    static void cleanup(void) {
+        while (gCRegHead) {
+            CReg* n = gCRegHead;
+            gCRegHead = gCRegHead->next;
+            delete n;
+        }
+    }
+};
+
+// -------------------------------------
+
+U_CAPI UCurrRegistryKey U_EXPORT2
+ucurr_register(const UChar* isoCode, const char* locale, UErrorCode *status)
+{
+    if (status && U_SUCCESS(*status)) {
+        char id[ULOC_FULLNAME_CAPACITY];
+        idForLocale(locale, id, sizeof(id), status);
+        return CReg::reg(isoCode, id, status);
+    }
+    return NULL;
+}
+
+// -------------------------------------
+
+U_CAPI UBool U_EXPORT2
+ucurr_unregister(UCurrRegistryKey key, UErrorCode* status)
+{
+    if (status && U_SUCCESS(*status)) {
+        return CReg::unreg(key);
+    }
+    return FALSE;
+}
+#endif /* UCONFIG_NO_SERVICE */
+
+// -------------------------------------
+
+/**
+ * Release all static memory held by currency.
+ */
+/*The declaration here is needed so currency_cleanup(void)
+ * can call this function.
+ */
+static UBool U_CALLCONV
+currency_cache_cleanup(void);
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV currency_cleanup(void) {
+#if !UCONFIG_NO_SERVICE
+    CReg::cleanup();
+#endif
+    /*
+     * There might be some cached currency data or isoCodes data.
+     */
+    currency_cache_cleanup();
+    isoCodes_cleanup();
+    currSymbolsEquiv_cleanup();
+
+    return TRUE;
+}
+U_CDECL_END
+
+// -------------------------------------
+
+U_CAPI int32_t U_EXPORT2
+ucurr_forLocale(const char* locale,
+                UChar* buff,
+                int32_t buffCapacity,
+                UErrorCode* ec) {
+    if (U_FAILURE(*ec)) { return 0; }
+    if (buffCapacity < 0 || (buff == nullptr && buffCapacity > 0)) {
+        *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    char currency[4];  // ISO currency codes are alpha3 codes.
+    UErrorCode localStatus = U_ZERO_ERROR;
+    int32_t resLen = uloc_getKeywordValue(locale, "currency",
+                                          currency, UPRV_LENGTHOF(currency), &localStatus);
+    if (U_SUCCESS(localStatus) && resLen == 3 && uprv_isInvariantString(currency, resLen)) {
+        if (resLen < buffCapacity) {
+            T_CString_toUpperCase(currency);
+            u_charsToUChars(currency, buff, resLen);
+        }
+        return u_terminateUChars(buff, buffCapacity, resLen, ec);
+    }
+
+    // get country or country_variant in `id'
+    char id[ULOC_FULLNAME_CAPACITY];
+    idForLocale(locale, id, UPRV_LENGTHOF(id), ec);
+    if (U_FAILURE(*ec)) {
+        return 0;
+    }
+
+#if !UCONFIG_NO_SERVICE
+    const UChar* result = CReg::get(id);
+    if (result) {
+        if(buffCapacity > u_strlen(result)) {
+            u_strcpy(buff, result);
+        }
+        resLen = u_strlen(result);
+        return u_terminateUChars(buff, buffCapacity, resLen, ec);
+    }
+#endif
+    // Remove variants, which is only needed for registration.
+    char *idDelim = uprv_strchr(id, VAR_DELIM);
+    if (idDelim) {
+        idDelim[0] = 0;
+    }
+
+    const UChar* s = NULL;  // Currency code from data file.
+    if (id[0] == 0) {
+        // No point looking in the data for an empty string.
+        // This is what we would get.
+        localStatus = U_MISSING_RESOURCE_ERROR;
+    } else {
+        // Look up the CurrencyMap element in the root bundle.
+        localStatus = U_ZERO_ERROR;
+        UResourceBundle *rb = ures_openDirect(U_ICUDATA_CURR, CURRENCY_DATA, &localStatus);
+        UResourceBundle *cm = ures_getByKey(rb, CURRENCY_MAP, rb, &localStatus);
+        UResourceBundle *countryArray = ures_getByKey(rb, id, cm, &localStatus);
+        UResourceBundle *currencyReq = ures_getByIndex(countryArray, 0, NULL, &localStatus);
+        s = ures_getStringByKey(currencyReq, "id", &resLen, &localStatus);
+        ures_close(currencyReq);
+        ures_close(countryArray);
+    }
+
+    if ((U_FAILURE(localStatus)) && strchr(id, '_') != 0) {
+        // We don't know about it.  Check to see if we support the variant.
+        uloc_getParent(locale, id, UPRV_LENGTHOF(id), ec);
+        *ec = U_USING_FALLBACK_WARNING;
+        // TODO: Loop over the shortened id rather than recursing and
+        // looking again for a currency keyword.
+        return ucurr_forLocale(id, buff, buffCapacity, ec);
+    }
+    if (*ec == U_ZERO_ERROR || localStatus != U_ZERO_ERROR) {
+        // There is nothing to fallback to. Report the failure/warning if possible.
+        *ec = localStatus;
+    }
+    if (U_SUCCESS(*ec)) {
+        if(buffCapacity > resLen) {
+            u_strcpy(buff, s);
+        }
+    }
+    return u_terminateUChars(buff, buffCapacity, resLen, ec);
+}
+
+// end registration
+
+/**
+ * Modify the given locale name by removing the rightmost _-delimited
+ * element.  If there is none, empty the string ("" == root).
+ * NOTE: The string "root" is not recognized; do not use it.
+ * @return TRUE if the fallback happened; FALSE if locale is already
+ * root ("").
+ */
+static UBool fallback(char *loc) {
+    if (!*loc) {
+        return FALSE;
+    }
+    UErrorCode status = U_ZERO_ERROR;
+    if (uprv_strcmp(loc, "en_GB") == 0) {
+        // HACK: See #13368.  We need "en_GB" to fall back to "en_001" instead of "en"
+        // in order to consume the correct data strings.  This hack will be removed
+        // when proper data sink loading is implemented here.
+        // NOTE: "001" adds 1 char over "GB".  However, both call sites allocate
+        // arrays with length ULOC_FULLNAME_CAPACITY (plenty of room for en_001).
+        uprv_strcpy(loc + 3, "001");
+    } else {
+        uloc_getParent(loc, loc, (int32_t)uprv_strlen(loc), &status);
+    }
+ /*
+    char *i = uprv_strrchr(loc, '_');
+    if (i == NULL) {
+        i = loc;
+    }
+    *i = 0;
+ */
+    return TRUE;
+}
+
+
+U_CAPI const UChar* U_EXPORT2
+ucurr_getName(const UChar* currency,
+              const char* locale,
+              UCurrNameStyle nameStyle,
+              UBool* isChoiceFormat, // fillin
+              int32_t* len, // fillin
+              UErrorCode* ec) {
+
+    // Look up the Currencies resource for the given locale.  The
+    // Currencies locale data looks like this:
+    //|en {
+    //|  Currencies {
+    //|    USD { "US$", "US Dollar" }
+    //|    CHF { "Sw F", "Swiss Franc" }
+    //|    INR { "=0#Rs|1#Re|1<Rs", "=0#Rupees|1#Rupee|1<Rupees" }
+    //|    //...
+    //|  }
+    //|}
+
+    if (U_FAILURE(*ec)) {
+        return 0;
+    }
+
+    int32_t choice = (int32_t) nameStyle;
+    if (choice < 0 || choice > 4) {
+        *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    // In the future, resource bundles may implement multi-level
+    // fallback.  That is, if a currency is not found in the en_US
+    // Currencies data, then the en Currencies data will be searched.
+    // Currently, if a Currencies datum exists in en_US and en, the
+    // en_US entry hides that in en.
+
+    // We want multi-level fallback for this resource, so we implement
+    // it manually.
+
+    // Use a separate UErrorCode here that does not propagate out of
+    // this function.
+    UErrorCode ec2 = U_ZERO_ERROR;
+
+    char loc[ULOC_FULLNAME_CAPACITY];
+    uloc_getName(locale, loc, sizeof(loc), &ec2);
+    if (U_FAILURE(ec2) || ec2 == U_STRING_NOT_TERMINATED_WARNING) {
+        *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    char buf[ISO_CURRENCY_CODE_LENGTH+1];
+    myUCharsToChars(buf, currency);
+    
+    /* Normalize the keyword value to uppercase */
+    T_CString_toUpperCase(buf);
+    
+    const UChar* s = NULL;
+    ec2 = U_ZERO_ERROR;
+    LocalUResourceBundlePointer rb(ures_open(U_ICUDATA_CURR, loc, &ec2));
+
+    if (nameStyle == UCURR_NARROW_SYMBOL_NAME || nameStyle == UCURR_FORMAL_SYMBOL_NAME || nameStyle == UCURR_VARIANT_SYMBOL_NAME) {
+        CharString key;
+        switch (nameStyle) {
+        case UCURR_NARROW_SYMBOL_NAME:
+            key.append(CURRENCIES_NARROW, ec2);
+            break;
+        case UCURR_FORMAL_SYMBOL_NAME:
+            key.append(CURRENCIES_FORMAL, ec2);
+            break;
+        case UCURR_VARIANT_SYMBOL_NAME:
+            key.append(CURRENCIES_VARIANT, ec2);
+            break;
+        default:
+            *ec = U_UNSUPPORTED_ERROR;
+            return 0;
+        }
+        key.append("/", ec2);
+        key.append(buf, ec2);
+        s = ures_getStringByKeyWithFallback(rb.getAlias(), key.data(), len, &ec2);
+        if (ec2 == U_MISSING_RESOURCE_ERROR) {
+            *ec = U_USING_FALLBACK_WARNING;
+            ec2 = U_ZERO_ERROR;
+            choice = UCURR_SYMBOL_NAME;
+        }
+    }
+    if (s == NULL) {
+        ures_getByKey(rb.getAlias(), CURRENCIES, rb.getAlias(), &ec2);
+        ures_getByKeyWithFallback(rb.getAlias(), buf, rb.getAlias(), &ec2);
+        s = ures_getStringByIndex(rb.getAlias(), choice, len, &ec2);
+    }
+
+    // If we've succeeded we're done.  Otherwise, try to fallback.
+    // If that fails (because we are already at root) then exit.
+    if (U_SUCCESS(ec2)) {
+        if (ec2 == U_USING_DEFAULT_WARNING
+            || (ec2 == U_USING_FALLBACK_WARNING && *ec != U_USING_DEFAULT_WARNING)) {
+            *ec = ec2;
+        }
+    }
+
+    // We no longer support choice format data in names.  Data should not contain
+    // choice patterns.
+    if (isChoiceFormat != NULL) {
+        *isChoiceFormat = FALSE;
+    }
+    if (U_SUCCESS(ec2)) {
+        U_ASSERT(s != NULL);
+        return s;
+    }
+
+    // If we fail to find a match, use the ISO 4217 code
+    *len = u_strlen(currency); // Should == ISO_CURRENCY_CODE_LENGTH, but maybe not...?
+    *ec = U_USING_DEFAULT_WARNING;
+    return currency;
+}
+
+U_CAPI const UChar* U_EXPORT2
+ucurr_getPluralName(const UChar* currency,
+                    const char* locale,
+                    UBool* isChoiceFormat,
+                    const char* pluralCount,
+                    int32_t* len, // fillin
+                    UErrorCode* ec) {
+    // Look up the Currencies resource for the given locale.  The
+    // Currencies locale data looks like this:
+    //|en {
+    //|  CurrencyPlurals {
+    //|    USD{
+    //|      one{"US dollar"}
+    //|      other{"US dollars"}
+    //|    }
+    //|  }
+    //|}
+
+    if (U_FAILURE(*ec)) {
+        return 0;
+    }
+
+    // Use a separate UErrorCode here that does not propagate out of
+    // this function.
+    UErrorCode ec2 = U_ZERO_ERROR;
+
+    char loc[ULOC_FULLNAME_CAPACITY];
+    uloc_getName(locale, loc, sizeof(loc), &ec2);
+    if (U_FAILURE(ec2) || ec2 == U_STRING_NOT_TERMINATED_WARNING) {
+        *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    char buf[ISO_CURRENCY_CODE_LENGTH+1];
+    myUCharsToChars(buf, currency);
+
+    const UChar* s = NULL;
+    ec2 = U_ZERO_ERROR;
+    UResourceBundle* rb = ures_open(U_ICUDATA_CURR, loc, &ec2);
+
+    rb = ures_getByKey(rb, CURRENCYPLURALS, rb, &ec2);
+
+    // Fetch resource with multi-level resource inheritance fallback
+    rb = ures_getByKeyWithFallback(rb, buf, rb, &ec2);
+
+    s = ures_getStringByKeyWithFallback(rb, pluralCount, len, &ec2);
+    if (U_FAILURE(ec2)) {
+        //  fall back to "other"
+        ec2 = U_ZERO_ERROR;
+        s = ures_getStringByKeyWithFallback(rb, "other", len, &ec2);     
+        if (U_FAILURE(ec2)) {
+            ures_close(rb);
+            // fall back to long name in Currencies
+            return ucurr_getName(currency, locale, UCURR_LONG_NAME, 
+                                 isChoiceFormat, len, ec);
+        }
+    }
+    ures_close(rb);
+
+    // If we've succeeded we're done.  Otherwise, try to fallback.
+    // If that fails (because we are already at root) then exit.
+    if (U_SUCCESS(ec2)) {
+        if (ec2 == U_USING_DEFAULT_WARNING
+            || (ec2 == U_USING_FALLBACK_WARNING && *ec != U_USING_DEFAULT_WARNING)) {
+            *ec = ec2;
+        }
+        U_ASSERT(s != NULL);
+        return s;
+    }
+
+    // If we fail to find a match, use the ISO 4217 code
+    *len = u_strlen(currency); // Should == ISO_CURRENCY_CODE_LENGTH, but maybe not...?
+    *ec = U_USING_DEFAULT_WARNING;
+    return currency;
+}
+
+
+//========================================================================
+// Following are structure and function for parsing currency names
+
+#define NEED_TO_BE_DELETED 0x1
+
+// TODO: a better way to define this?
+#define MAX_CURRENCY_NAME_LEN 100
+
+typedef struct {
+    const char* IsoCode;  // key
+    UChar* currencyName;  // value
+    int32_t currencyNameLen;  // value length
+    int32_t flag;  // flags
+} CurrencyNameStruct;
+
+
+#ifndef MIN
+#define MIN(a,b) (((a)<(b)) ? (a) : (b))
+#endif
+
+#ifndef MAX
+#define MAX(a,b) (((a)<(b)) ? (b) : (a))
+#endif
+
+
+// Comparason function used in quick sort.
+static int U_CALLCONV currencyNameComparator(const void* a, const void* b) {
+    const CurrencyNameStruct* currName_1 = (const CurrencyNameStruct*)a;
+    const CurrencyNameStruct* currName_2 = (const CurrencyNameStruct*)b;
+    for (int32_t i = 0; 
+         i < MIN(currName_1->currencyNameLen, currName_2->currencyNameLen);
+         ++i) {
+        if (currName_1->currencyName[i] < currName_2->currencyName[i]) {
+            return -1;
+        }
+        if (currName_1->currencyName[i] > currName_2->currencyName[i]) {
+            return 1;
+        }
+    }
+    if (currName_1->currencyNameLen < currName_2->currencyNameLen) {
+        return -1;
+    } else if (currName_1->currencyNameLen > currName_2->currencyNameLen) {
+        return 1;
+    }
+    return 0;
+}
+
+
+// Give a locale, return the maximum number of currency names associated with
+// this locale.
+// It gets currency names from resource bundles using fallback.
+// It is the maximum number because in the fallback chain, some of the 
+// currency names are duplicated.
+// For example, given locale as "en_US", the currency names get from resource
+// bundle in "en_US" and "en" are duplicated. The fallback mechanism will count
+// all currency names in "en_US" and "en".
+static void
+getCurrencyNameCount(const char* loc, int32_t* total_currency_name_count, int32_t* total_currency_symbol_count) {
+    U_NAMESPACE_USE
+    *total_currency_name_count = 0;
+    *total_currency_symbol_count = 0;
+    const UChar* s = NULL;
+    char locale[ULOC_FULLNAME_CAPACITY] = "";
+    uprv_strcpy(locale, loc);
+    const icu::Hashtable *currencySymbolsEquiv = getCurrSymbolsEquiv();
+    for (;;) {
+        UErrorCode ec2 = U_ZERO_ERROR;
+        // TODO: ures_openDirect?
+        UResourceBundle* rb = ures_open(U_ICUDATA_CURR, locale, &ec2);
+        UResourceBundle* curr = ures_getByKey(rb, CURRENCIES, NULL, &ec2);
+        int32_t n = ures_getSize(curr);
+        for (int32_t i=0; i<n; ++i) {
+            UResourceBundle* names = ures_getByIndex(curr, i, NULL, &ec2);
+            int32_t len;
+            s = ures_getStringByIndex(names, UCURR_SYMBOL_NAME, &len, &ec2);
+            ++(*total_currency_symbol_count);  // currency symbol
+            if (currencySymbolsEquiv != NULL) {
+                *total_currency_symbol_count += countEquivalent(*currencySymbolsEquiv, UnicodeString(TRUE, s, len));
+            }
+            ++(*total_currency_symbol_count); // iso code
+            ++(*total_currency_name_count); // long name
+            ures_close(names);
+        }
+
+        // currency plurals
+        UErrorCode ec3 = U_ZERO_ERROR;
+        UResourceBundle* curr_p = ures_getByKey(rb, CURRENCYPLURALS, NULL, &ec3);
+        n = ures_getSize(curr_p);
+        for (int32_t i=0; i<n; ++i) {
+            UResourceBundle* names = ures_getByIndex(curr_p, i, NULL, &ec3);
+            *total_currency_name_count += ures_getSize(names);
+            ures_close(names);
+        }
+        ures_close(curr_p);
+        ures_close(curr);
+        ures_close(rb);
+
+        if (!fallback(locale)) {
+            break;
+        }
+    }
+}
+
+static UChar* 
+toUpperCase(const UChar* source, int32_t len, const char* locale) {
+    UChar* dest = NULL;
+    UErrorCode ec = U_ZERO_ERROR;
+    int32_t destLen = u_strToUpper(dest, 0, source, len, locale, &ec);
+
+    ec = U_ZERO_ERROR;
+    dest = (UChar*)uprv_malloc(sizeof(UChar) * MAX(destLen, len));
+    u_strToUpper(dest, destLen, source, len, locale, &ec);
+    if (U_FAILURE(ec)) {
+        u_memcpy(dest, source, len);
+    } 
+    return dest;
+}
+
+
+// Collect all available currency names associated with the given locale
+// (enable fallback chain).
+// Read currenc names defined in resource bundle "Currencies" and
+// "CurrencyPlural", enable fallback chain.
+// return the malloc-ed currency name arrays and the total number of currency
+// names in the array.
+static void
+collectCurrencyNames(const char* locale, 
+                     CurrencyNameStruct** currencyNames, 
+                     int32_t* total_currency_name_count, 
+                     CurrencyNameStruct** currencySymbols, 
+                     int32_t* total_currency_symbol_count, 
+                     UErrorCode& ec) {
+    U_NAMESPACE_USE
+    const icu::Hashtable *currencySymbolsEquiv = getCurrSymbolsEquiv();
+    // Look up the Currencies resource for the given locale.
+    UErrorCode ec2 = U_ZERO_ERROR;
+
+    char loc[ULOC_FULLNAME_CAPACITY] = "";
+    uloc_getName(locale, loc, sizeof(loc), &ec2);
+    if (U_FAILURE(ec2) || ec2 == U_STRING_NOT_TERMINATED_WARNING) {
+        ec = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+
+    // Get maximum currency name count first.
+    getCurrencyNameCount(loc, total_currency_name_count, total_currency_symbol_count);
+
+    *currencyNames = (CurrencyNameStruct*)uprv_malloc
+        (sizeof(CurrencyNameStruct) * (*total_currency_name_count));
+    *currencySymbols = (CurrencyNameStruct*)uprv_malloc
+        (sizeof(CurrencyNameStruct) * (*total_currency_symbol_count));
+
+    if(currencyNames == NULL || currencySymbols == NULL) {
+      ec = U_MEMORY_ALLOCATION_ERROR;
+    }
+
+    if (U_FAILURE(ec)) return;
+
+    const UChar* s = NULL;  // currency name
+    char* iso = NULL;  // currency ISO code
+
+    *total_currency_name_count = 0;
+    *total_currency_symbol_count = 0;
+
+    UErrorCode ec3 = U_ZERO_ERROR;
+    UErrorCode ec4 = U_ZERO_ERROR;
+
+    // Using hash to remove duplicates caused by locale fallback
+    UHashtable* currencyIsoCodes = uhash_open(uhash_hashChars, uhash_compareChars, NULL, &ec3);
+    UHashtable* currencyPluralIsoCodes = uhash_open(uhash_hashChars, uhash_compareChars, NULL, &ec4);
+    for (int32_t localeLevel = 0; ; ++localeLevel) {
+        ec2 = U_ZERO_ERROR;
+        // TODO: ures_openDirect
+        UResourceBundle* rb = ures_open(U_ICUDATA_CURR, loc, &ec2);
+        UResourceBundle* curr = ures_getByKey(rb, CURRENCIES, NULL, &ec2);
+        int32_t n = ures_getSize(curr);
+        for (int32_t i=0; i<n; ++i) {
+            UResourceBundle* names = ures_getByIndex(curr, i, NULL, &ec2);
+            int32_t len;
+            s = ures_getStringByIndex(names, UCURR_SYMBOL_NAME, &len, &ec2);
+            // TODO: uhash_put wont change key/value?
+            iso = (char*)ures_getKey(names);
+            if (localeLevel == 0) {
+                uhash_put(currencyIsoCodes, iso, iso, &ec3); 
+            } else {
+                if (uhash_get(currencyIsoCodes, iso) != NULL) {
+                    ures_close(names);
+                    continue;
+                } else {
+                    uhash_put(currencyIsoCodes, iso, iso, &ec3); 
+                }
+            }
+            // Add currency symbol.
+            (*currencySymbols)[*total_currency_symbol_count].IsoCode = iso;
+            (*currencySymbols)[*total_currency_symbol_count].currencyName = (UChar*)s;
+            (*currencySymbols)[*total_currency_symbol_count].flag = 0;
+            (*currencySymbols)[(*total_currency_symbol_count)++].currencyNameLen = len;
+            // Add equivalent symbols
+            if (currencySymbolsEquiv != NULL) {
+                UnicodeString str(TRUE, s, len);
+                icu::EquivIterator iter(*currencySymbolsEquiv, str);
+                const UnicodeString *symbol;
+                while ((symbol = iter.next()) != NULL) {
+                    (*currencySymbols)[*total_currency_symbol_count].IsoCode = iso;
+                    (*currencySymbols)[*total_currency_symbol_count].currencyName =
+                        const_cast<UChar*>(symbol->getBuffer());
+                    (*currencySymbols)[*total_currency_symbol_count].flag = 0;
+                    (*currencySymbols)[(*total_currency_symbol_count)++].currencyNameLen = symbol->length();
+                }
+            }
+
+            // Add currency long name.
+            s = ures_getStringByIndex(names, UCURR_LONG_NAME, &len, &ec2);
+            (*currencyNames)[*total_currency_name_count].IsoCode = iso;
+            UChar* upperName = toUpperCase(s, len, locale);
+            (*currencyNames)[*total_currency_name_count].currencyName = upperName;
+            (*currencyNames)[*total_currency_name_count].flag = NEED_TO_BE_DELETED;
+            (*currencyNames)[(*total_currency_name_count)++].currencyNameLen = len;
+
+            // put (iso, 3, and iso) in to array
+            // Add currency ISO code.
+            (*currencySymbols)[*total_currency_symbol_count].IsoCode = iso;
+            (*currencySymbols)[*total_currency_symbol_count].currencyName = (UChar*)uprv_malloc(sizeof(UChar)*3);
+            // Must convert iso[] into Unicode
+            u_charsToUChars(iso, (*currencySymbols)[*total_currency_symbol_count].currencyName, 3);
+            (*currencySymbols)[*total_currency_symbol_count].flag = NEED_TO_BE_DELETED;
+            (*currencySymbols)[(*total_currency_symbol_count)++].currencyNameLen = 3;
+
+            ures_close(names);
+        }
+
+        // currency plurals
+        UErrorCode ec5 = U_ZERO_ERROR;
+        UResourceBundle* curr_p = ures_getByKey(rb, CURRENCYPLURALS, NULL, &ec5);
+        n = ures_getSize(curr_p);
+        for (int32_t i=0; i<n; ++i) {
+            UResourceBundle* names = ures_getByIndex(curr_p, i, NULL, &ec5);
+            iso = (char*)ures_getKey(names);
+            // Using hash to remove duplicated ISO codes in fallback chain.
+            if (localeLevel == 0) {
+                uhash_put(currencyPluralIsoCodes, iso, iso, &ec4); 
+            } else {
+                if (uhash_get(currencyPluralIsoCodes, iso) != NULL) {
+                    ures_close(names);
+                    continue;
+                } else {
+                    uhash_put(currencyPluralIsoCodes, iso, iso, &ec4); 
+                }
+            }
+            int32_t num = ures_getSize(names);
+            int32_t len;
+            for (int32_t j = 0; j < num; ++j) {
+                // TODO: remove duplicates between singular name and 
+                // currency long name?
+                s = ures_getStringByIndex(names, j, &len, &ec5);
+                (*currencyNames)[*total_currency_name_count].IsoCode = iso;
+                UChar* upperName = toUpperCase(s, len, locale);
+                (*currencyNames)[*total_currency_name_count].currencyName = upperName;
+                (*currencyNames)[*total_currency_name_count].flag = NEED_TO_BE_DELETED;
+                (*currencyNames)[(*total_currency_name_count)++].currencyNameLen = len;
+            }
+            ures_close(names);
+        }
+        ures_close(curr_p);
+        ures_close(curr);
+        ures_close(rb);
+
+        if (!fallback(loc)) {
+            break;
+        }
+    }
+
+    uhash_close(currencyIsoCodes);
+    uhash_close(currencyPluralIsoCodes);
+
+    // quick sort the struct
+    qsort(*currencyNames, *total_currency_name_count, 
+          sizeof(CurrencyNameStruct), currencyNameComparator);
+    qsort(*currencySymbols, *total_currency_symbol_count, 
+          sizeof(CurrencyNameStruct), currencyNameComparator);
+
+#ifdef UCURR_DEBUG
+    printf("currency name count: %d\n", *total_currency_name_count);
+    for (int32_t index = 0; index < *total_currency_name_count; ++index) {
+        printf("index: %d\n", index);
+        printf("iso: %s\n", (*currencyNames)[index].IsoCode);
+        char curNameBuf[1024];
+        memset(curNameBuf, 0, 1024);
+        u_austrncpy(curNameBuf, (*currencyNames)[index].currencyName, (*currencyNames)[index].currencyNameLen);
+        printf("currencyName: %s\n", curNameBuf);
+        printf("len: %d\n", (*currencyNames)[index].currencyNameLen);
+    }
+    printf("currency symbol count: %d\n", *total_currency_symbol_count);
+    for (int32_t index = 0; index < *total_currency_symbol_count; ++index) {
+        printf("index: %d\n", index);
+        printf("iso: %s\n", (*currencySymbols)[index].IsoCode);
+        char curNameBuf[1024];
+        memset(curNameBuf, 0, 1024);
+        u_austrncpy(curNameBuf, (*currencySymbols)[index].currencyName, (*currencySymbols)[index].currencyNameLen);
+        printf("currencySymbol: %s\n", curNameBuf);
+        printf("len: %d\n", (*currencySymbols)[index].currencyNameLen);
+    }
+#endif
+    // fail on hashtable errors
+    if (U_FAILURE(ec3)) {
+      ec = ec3;
+      return;
+    }
+    if (U_FAILURE(ec4)) {
+      ec = ec4;
+      return;
+    }
+}
+
+// @param  currencyNames: currency names array
+// @param  indexInCurrencyNames: the index of the character in currency names 
+//         array against which the comparison is done
+// @param  key: input text char to compare against
+// @param  begin(IN/OUT): the begin index of matching range in currency names array
+// @param  end(IN/OUT): the end index of matching range in currency names array.
+static int32_t
+binarySearch(const CurrencyNameStruct* currencyNames, 
+             int32_t indexInCurrencyNames,
+             const UChar key,
+             int32_t* begin, int32_t* end) {
+#ifdef UCURR_DEBUG
+    printf("key = %x\n", key);
+#endif
+   int32_t first = *begin;
+   int32_t last = *end;
+   while (first <= last) {
+       int32_t mid = (first + last) / 2;  // compute mid point.
+       if (indexInCurrencyNames >= currencyNames[mid].currencyNameLen) {
+           first = mid + 1;
+       } else {
+           if (key > currencyNames[mid].currencyName[indexInCurrencyNames]) {
+               first = mid + 1;
+           }
+           else if (key < currencyNames[mid].currencyName[indexInCurrencyNames]) {
+               last = mid - 1;
+           }
+           else {
+                // Find a match, and looking for ranges
+                // Now do two more binary searches. First, on the left side for
+                // the greatest L such that CurrencyNameStruct[L] < key.
+                int32_t L = *begin;
+                int32_t R = mid;
+
+#ifdef UCURR_DEBUG
+                printf("mid = %d\n", mid);
+#endif
+                while (L < R) {
+                    int32_t M = (L + R) / 2;
+#ifdef UCURR_DEBUG
+                    printf("L = %d, R = %d, M = %d\n", L, R, M);
+#endif
+                    if (indexInCurrencyNames >= currencyNames[M].currencyNameLen) {
+                        L = M + 1;
+                    } else {
+                        if (currencyNames[M].currencyName[indexInCurrencyNames] < key) {
+                            L = M + 1;
+                        } else {
+#ifdef UCURR_DEBUG
+                            U_ASSERT(currencyNames[M].currencyName[indexInCurrencyNames] == key);
+#endif
+                            R = M;
+                        }
+                    }
+                }
+#ifdef UCURR_DEBUG
+                U_ASSERT(L == R);
+#endif
+                *begin = L;
+#ifdef UCURR_DEBUG
+                printf("begin = %d\n", *begin);
+                U_ASSERT(currencyNames[*begin].currencyName[indexInCurrencyNames] == key);
+#endif
+
+                // Now for the second search, finding the least R such that
+                // key < CurrencyNameStruct[R].
+                L = mid;
+                R = *end;
+                while (L < R) {
+                    int32_t M = (L + R) / 2;
+#ifdef UCURR_DEBUG
+                    printf("L = %d, R = %d, M = %d\n", L, R, M);
+#endif
+                    if (currencyNames[M].currencyNameLen < indexInCurrencyNames) {
+                        L = M + 1;
+                    } else {
+                        if (currencyNames[M].currencyName[indexInCurrencyNames] > key) {
+                            R = M;
+                        } else {
+#ifdef UCURR_DEBUG
+                            U_ASSERT(currencyNames[M].currencyName[indexInCurrencyNames] == key);
+#endif
+                            L = M + 1;
+                        }
+                    }
+                }
+#ifdef UCURR_DEBUG
+                U_ASSERT(L == R);
+#endif
+                if (currencyNames[R].currencyName[indexInCurrencyNames] > key) {
+                    *end = R - 1;
+                } else {
+                    *end = R;
+                }
+#ifdef UCURR_DEBUG
+                printf("end = %d\n", *end);
+#endif
+
+                // now, found the range. check whether there is exact match
+                if (currencyNames[*begin].currencyNameLen == indexInCurrencyNames + 1) {
+                    return *begin;  // find range and exact match.
+                }
+                return -1;  // find range, but no exact match.
+           }
+       }
+   }
+   *begin = -1;
+   *end = -1;
+   return -1;    // failed to find range.
+}
+
+
+// Linear search "text" in "currencyNames".
+// @param  begin, end: the begin and end index in currencyNames, within which
+//         range should the search be performed.
+// @param  textLen: the length of the text to be compared
+// @param  maxMatchLen(IN/OUT): passing in the computed max matching length
+//                              pass out the new max  matching length
+// @param  maxMatchIndex: the index in currencyName which has the longest
+//                        match with input text.
+static void
+linearSearch(const CurrencyNameStruct* currencyNames, 
+             int32_t begin, int32_t end,
+             const UChar* text, int32_t textLen,
+             int32_t *partialMatchLen,
+             int32_t *maxMatchLen, int32_t* maxMatchIndex) {
+    int32_t initialPartialMatchLen = *partialMatchLen;
+    for (int32_t index = begin; index <= end; ++index) {
+        int32_t len = currencyNames[index].currencyNameLen;
+        if (len > *maxMatchLen && len <= textLen &&
+            uprv_memcmp(currencyNames[index].currencyName, text, len * sizeof(UChar)) == 0) {
+            *partialMatchLen = MAX(*partialMatchLen, len);
+            *maxMatchIndex = index;
+            *maxMatchLen = len;
+#ifdef UCURR_DEBUG
+            printf("maxMatchIndex = %d, maxMatchLen = %d\n",
+                   *maxMatchIndex, *maxMatchLen);
+#endif
+        } else {
+            // Check for partial matches.
+            for (int32_t i=initialPartialMatchLen; i<MIN(len, textLen); i++) {
+                if (currencyNames[index].currencyName[i] != text[i]) {
+                    break;
+                }
+                *partialMatchLen = MAX(*partialMatchLen, i + 1);
+            }
+        }
+    }
+}
+
+#define LINEAR_SEARCH_THRESHOLD 10
+
+// Find longest match between "text" and currency names in "currencyNames".
+// @param  total_currency_count: total number of currency names in CurrencyNames.
+// @param  textLen: the length of the text to be compared
+// @param  maxMatchLen: passing in the computed max matching length
+//                              pass out the new max  matching length
+// @param  maxMatchIndex: the index in currencyName which has the longest
+//                        match with input text.
+static void
+searchCurrencyName(const CurrencyNameStruct* currencyNames, 
+                   int32_t total_currency_count,
+                   const UChar* text, int32_t textLen,
+                   int32_t *partialMatchLen,
+                   int32_t* maxMatchLen, int32_t* maxMatchIndex) {
+    *maxMatchIndex = -1;
+    *maxMatchLen = 0;
+    int32_t matchIndex = -1;
+    int32_t binarySearchBegin = 0;
+    int32_t binarySearchEnd = total_currency_count - 1;
+    // It is a variant of binary search.
+    // For example, given the currency names in currencyNames array are:
+    // A AB ABC AD AZ B BB BBEX BBEXYZ BS C D E....
+    // and the input text is BBEXST
+    // The first round binary search search "B" in the text against
+    // the first char in currency names, and find the first char matching range
+    // to be "B BB BBEX BBEXYZ BS" (and the maximum matching "B").
+    // The 2nd round binary search search the second "B" in the text against
+    // the 2nd char in currency names, and narrow the matching range to
+    // "BB BBEX BBEXYZ" (and the maximum matching "BB").
+    // The 3rd round returnes the range as "BBEX BBEXYZ" (without changing
+    // maximum matching).
+    // The 4th round returns the same range (the maximum matching is "BBEX").
+    // The 5th round returns no matching range.
+    for (int32_t index = 0; index < textLen; ++index) {
+        // matchIndex saves the one with exact match till the current point.
+        // [binarySearchBegin, binarySearchEnd] saves the matching range.
+        matchIndex = binarySearch(currencyNames, index,
+                                  text[index],
+                                  &binarySearchBegin, &binarySearchEnd);
+        if (binarySearchBegin == -1) { // did not find the range
+            break;
+        }
+        *partialMatchLen = MAX(*partialMatchLen, index + 1);
+        if (matchIndex != -1) { 
+            // find an exact match for text from text[0] to text[index] 
+            // in currencyNames array.
+            *maxMatchLen = index + 1;
+            *maxMatchIndex = matchIndex;
+        }
+        if (binarySearchEnd - binarySearchBegin < LINEAR_SEARCH_THRESHOLD) {
+            // linear search if within threshold.
+            linearSearch(currencyNames, binarySearchBegin, binarySearchEnd,
+                         text, textLen,
+                         partialMatchLen,
+                         maxMatchLen, maxMatchIndex);
+            break;
+        }
+    }
+    return;
+}
+
+//========================= currency name cache =====================
+typedef struct {
+    char locale[ULOC_FULLNAME_CAPACITY];  //key
+    // currency names, case insensitive
+    CurrencyNameStruct* currencyNames;  // value
+    int32_t totalCurrencyNameCount;  // currency name count
+    // currency symbols and ISO code, case sensitive
+    CurrencyNameStruct* currencySymbols; // value
+    int32_t totalCurrencySymbolCount;  // count
+    // reference count.
+    // reference count is set to 1 when an entry is put to cache.
+    // it increases by 1 before accessing, and decreased by 1 after accessing.
+    // The entry is deleted when ref count is zero, which means 
+    // the entry is replaced out of cache and no process is accessing it.
+    int32_t refCount;
+} CurrencyNameCacheEntry;
+
+
+#define CURRENCY_NAME_CACHE_NUM 10
+
+// Reserve 10 cache entries.
+static CurrencyNameCacheEntry* currCache[CURRENCY_NAME_CACHE_NUM] = {NULL};
+// Using an index to indicate which entry to be replaced when cache is full.
+// It is a simple round-robin replacement strategy.
+static int8_t currentCacheEntryIndex = 0;
+
+static UMutex gCurrencyCacheMutex;
+
+// Cache deletion
+static void
+deleteCurrencyNames(CurrencyNameStruct* currencyNames, int32_t count) {
+    for (int32_t index = 0; index < count; ++index) {
+        if ( (currencyNames[index].flag & NEED_TO_BE_DELETED) ) {
+            uprv_free(currencyNames[index].currencyName);
+        }
+    }
+    uprv_free(currencyNames);
+}
+
+
+static void
+deleteCacheEntry(CurrencyNameCacheEntry* entry) {
+    deleteCurrencyNames(entry->currencyNames, entry->totalCurrencyNameCount);
+    deleteCurrencyNames(entry->currencySymbols, entry->totalCurrencySymbolCount);
+    uprv_free(entry);
+}
+
+
+// Cache clean up
+static UBool U_CALLCONV
+currency_cache_cleanup(void) {
+    for (int32_t i = 0; i < CURRENCY_NAME_CACHE_NUM; ++i) {
+        if (currCache[i]) {
+            deleteCacheEntry(currCache[i]);
+            currCache[i] = 0;
+        }
+    }
+    return TRUE;
+}
+
+
+/**
+ * Loads the currency name data from the cache, or from resource bundles if necessary.
+ * The refCount is automatically incremented.  It is the caller's responsibility
+ * to decrement it when done!
+ */
+static CurrencyNameCacheEntry*
+getCacheEntry(const char* locale, UErrorCode& ec) {
+
+    int32_t total_currency_name_count = 0;
+    CurrencyNameStruct* currencyNames = NULL;
+    int32_t total_currency_symbol_count = 0;
+    CurrencyNameStruct* currencySymbols = NULL;
+    CurrencyNameCacheEntry* cacheEntry = NULL;
+
+    umtx_lock(&gCurrencyCacheMutex);
+    // in order to handle racing correctly,
+    // not putting 'search' in a separate function.
+    int8_t found = -1;
+    for (int8_t i = 0; i < CURRENCY_NAME_CACHE_NUM; ++i) {
+        if (currCache[i]!= NULL &&
+            uprv_strcmp(locale, currCache[i]->locale) == 0) {
+            found = i;
+            break;
+        }
+    }
+    if (found != -1) {
+        cacheEntry = currCache[found];
+        ++(cacheEntry->refCount);
+    }
+    umtx_unlock(&gCurrencyCacheMutex);
+    if (found == -1) {
+        collectCurrencyNames(locale, &currencyNames, &total_currency_name_count, &currencySymbols, &total_currency_symbol_count, ec);
+        if (U_FAILURE(ec)) {
+            return NULL;
+        }
+        umtx_lock(&gCurrencyCacheMutex);
+        // check again.
+        for (int8_t i = 0; i < CURRENCY_NAME_CACHE_NUM; ++i) {
+            if (currCache[i]!= NULL &&
+                uprv_strcmp(locale, currCache[i]->locale) == 0) {
+                found = i;
+                break;
+            }
+        }
+        if (found == -1) {
+            // insert new entry to 
+            // currentCacheEntryIndex % CURRENCY_NAME_CACHE_NUM
+            // and remove the existing entry 
+            // currentCacheEntryIndex % CURRENCY_NAME_CACHE_NUM
+            // from cache.
+            cacheEntry = currCache[currentCacheEntryIndex];
+            if (cacheEntry) {
+                --(cacheEntry->refCount);
+                // delete if the ref count is zero
+                if (cacheEntry->refCount == 0) {
+                    deleteCacheEntry(cacheEntry);
+                }
+            }
+            cacheEntry = (CurrencyNameCacheEntry*)uprv_malloc(sizeof(CurrencyNameCacheEntry));
+            currCache[currentCacheEntryIndex] = cacheEntry;
+            uprv_strcpy(cacheEntry->locale, locale);
+            cacheEntry->currencyNames = currencyNames;
+            cacheEntry->totalCurrencyNameCount = total_currency_name_count;
+            cacheEntry->currencySymbols = currencySymbols;
+            cacheEntry->totalCurrencySymbolCount = total_currency_symbol_count;
+            cacheEntry->refCount = 2; // one for cache, one for reference
+            currentCacheEntryIndex = (currentCacheEntryIndex + 1) % CURRENCY_NAME_CACHE_NUM;
+            ucln_common_registerCleanup(UCLN_COMMON_CURRENCY, currency_cleanup);
+        } else {
+            deleteCurrencyNames(currencyNames, total_currency_name_count);
+            deleteCurrencyNames(currencySymbols, total_currency_symbol_count);
+            cacheEntry = currCache[found];
+            ++(cacheEntry->refCount);
+        }
+        umtx_unlock(&gCurrencyCacheMutex);
+    }
+
+    return cacheEntry;
+}
+
+static void releaseCacheEntry(CurrencyNameCacheEntry* cacheEntry) {
+    umtx_lock(&gCurrencyCacheMutex);
+    --(cacheEntry->refCount);
+    if (cacheEntry->refCount == 0) {  // remove
+        deleteCacheEntry(cacheEntry);
+    }
+    umtx_unlock(&gCurrencyCacheMutex);
+}
+
+U_CAPI void
+uprv_parseCurrency(const char* locale,
+                   const icu::UnicodeString& text,
+                   icu::ParsePosition& pos,
+                   int8_t type,
+                   int32_t* partialMatchLen,
+                   UChar* result,
+                   UErrorCode& ec) {
+    U_NAMESPACE_USE
+    if (U_FAILURE(ec)) {
+        return;
+    }
+    CurrencyNameCacheEntry* cacheEntry = getCacheEntry(locale, ec);
+    if (U_FAILURE(ec)) {
+        return;
+    }
+
+    int32_t total_currency_name_count = cacheEntry->totalCurrencyNameCount;
+    CurrencyNameStruct* currencyNames = cacheEntry->currencyNames;
+    int32_t total_currency_symbol_count = cacheEntry->totalCurrencySymbolCount;
+    CurrencyNameStruct* currencySymbols = cacheEntry->currencySymbols;
+
+    int32_t start = pos.getIndex();
+
+    UChar inputText[MAX_CURRENCY_NAME_LEN];  
+    UChar upperText[MAX_CURRENCY_NAME_LEN];  
+    int32_t textLen = MIN(MAX_CURRENCY_NAME_LEN, text.length() - start);
+    text.extract(start, textLen, inputText);
+    UErrorCode ec1 = U_ZERO_ERROR;
+    textLen = u_strToUpper(upperText, MAX_CURRENCY_NAME_LEN, inputText, textLen, locale, &ec1);
+
+    // Make sure partialMatchLen is initialized
+    *partialMatchLen = 0;
+
+    int32_t max = 0;
+    int32_t matchIndex = -1;
+    // case in-sensitive comparision against currency names
+    searchCurrencyName(currencyNames, total_currency_name_count, 
+                       upperText, textLen, partialMatchLen, &max, &matchIndex);
+
+#ifdef UCURR_DEBUG
+    printf("search in names, max = %d, matchIndex = %d\n", max, matchIndex);
+#endif
+
+    int32_t maxInSymbol = 0;
+    int32_t matchIndexInSymbol = -1;
+    if (type != UCURR_LONG_NAME) {  // not name only
+        // case sensitive comparison against currency symbols and ISO code.
+        searchCurrencyName(currencySymbols, total_currency_symbol_count, 
+                           inputText, textLen,
+                           partialMatchLen,
+                           &maxInSymbol, &matchIndexInSymbol);
+    }
+
+#ifdef UCURR_DEBUG
+    printf("search in symbols, maxInSymbol = %d, matchIndexInSymbol = %d\n", maxInSymbol, matchIndexInSymbol);
+    if(matchIndexInSymbol != -1) {
+      printf("== ISO=%s\n", currencySymbols[matchIndexInSymbol].IsoCode);
+    }
+#endif
+
+    if (max >= maxInSymbol && matchIndex != -1) {
+        u_charsToUChars(currencyNames[matchIndex].IsoCode, result, 4);
+        pos.setIndex(start + max);
+    } else if (maxInSymbol >= max && matchIndexInSymbol != -1) {
+        u_charsToUChars(currencySymbols[matchIndexInSymbol].IsoCode, result, 4);
+        pos.setIndex(start + maxInSymbol);
+    }
+
+    // decrease reference count
+    releaseCacheEntry(cacheEntry);
+}
+
+void uprv_currencyLeads(const char* locale, icu::UnicodeSet& result, UErrorCode& ec) {
+    U_NAMESPACE_USE
+    if (U_FAILURE(ec)) {
+        return;
+    }
+    CurrencyNameCacheEntry* cacheEntry = getCacheEntry(locale, ec);
+    if (U_FAILURE(ec)) {
+        return;
+    }
+
+    for (int32_t i=0; i<cacheEntry->totalCurrencySymbolCount; i++) {
+        const CurrencyNameStruct& info = cacheEntry->currencySymbols[i];
+        UChar32 cp;
+        U16_GET(info.currencyName, 0, 0, info.currencyNameLen, cp);
+        result.add(cp);
+    }
+
+    for (int32_t i=0; i<cacheEntry->totalCurrencyNameCount; i++) {
+        const CurrencyNameStruct& info = cacheEntry->currencyNames[i];
+        UChar32 cp;
+        U16_GET(info.currencyName, 0, 0, info.currencyNameLen, cp);
+        result.add(cp);
+    }
+
+    // decrease reference count
+    releaseCacheEntry(cacheEntry);
+}
+
+
+/**
+ * Internal method.  Given a currency ISO code and a locale, return
+ * the "static" currency name.  This is usually the same as the
+ * UCURR_SYMBOL_NAME, but if the latter is a choice format, then the
+ * format is applied to the number 2.0 (to yield the more common
+ * plural) to return a static name.
+ *
+ * This is used for backward compatibility with old currency logic in
+ * DecimalFormat and DecimalFormatSymbols.
+ */
+U_CAPI void
+uprv_getStaticCurrencyName(const UChar* iso, const char* loc,
+                           icu::UnicodeString& result, UErrorCode& ec)
+{
+    U_NAMESPACE_USE
+
+    int32_t len;
+    const UChar* currname = ucurr_getName(iso, loc, UCURR_SYMBOL_NAME,
+                                          nullptr /* isChoiceFormat */, &len, &ec);
+    if (U_SUCCESS(ec)) {
+        result.setTo(currname, len);
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ucurr_getDefaultFractionDigits(const UChar* currency, UErrorCode* ec) {
+    return ucurr_getDefaultFractionDigitsForUsage(currency,UCURR_USAGE_STANDARD,ec);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucurr_getDefaultFractionDigitsForUsage(const UChar* currency, const UCurrencyUsage usage, UErrorCode* ec) {
+    int32_t fracDigits = 0;
+    if (U_SUCCESS(*ec)) {
+        switch (usage) {
+            case UCURR_USAGE_STANDARD:
+                fracDigits = (_findMetaData(currency, *ec))[0];
+                break;
+            case UCURR_USAGE_CASH:
+                fracDigits = (_findMetaData(currency, *ec))[2];
+                break;
+            default:
+                *ec = U_UNSUPPORTED_ERROR;
+        }
+    }
+    return fracDigits;
+}
+
+U_CAPI double U_EXPORT2
+ucurr_getRoundingIncrement(const UChar* currency, UErrorCode* ec) {
+    return ucurr_getRoundingIncrementForUsage(currency, UCURR_USAGE_STANDARD, ec);
+}
+
+U_CAPI double U_EXPORT2
+ucurr_getRoundingIncrementForUsage(const UChar* currency, const UCurrencyUsage usage, UErrorCode* ec) {
+    double result = 0.0;
+
+    const int32_t *data = _findMetaData(currency, *ec);
+    if (U_SUCCESS(*ec)) {
+        int32_t fracDigits;
+        int32_t increment;
+        switch (usage) {
+            case UCURR_USAGE_STANDARD:
+                fracDigits = data[0];
+                increment = data[1];
+                break;
+            case UCURR_USAGE_CASH:
+                fracDigits = data[2];
+                increment = data[3];
+                break;
+            default:
+                *ec = U_UNSUPPORTED_ERROR;
+                return result;
+        }
+
+        // If the meta data is invalid, return 0.0
+        if (fracDigits < 0 || fracDigits > MAX_POW10) {
+            *ec = U_INVALID_FORMAT_ERROR;
+        } else {
+            // A rounding value of 0 or 1 indicates no rounding.
+            if (increment >= 2) {
+                // Return (increment) / 10^(fracDigits).  The only actual rounding data,
+                // as of this writing, is CHF { 2, 5 }.
+                result = double(increment) / POW10[fracDigits];
+            }
+        }
+    }
+
+    return result;
+}
+
+U_CDECL_BEGIN
+
+typedef struct UCurrencyContext {
+    uint32_t currType; /* UCurrCurrencyType */
+    uint32_t listIdx;
+} UCurrencyContext;
+
+/*
+Please keep this list in alphabetical order.
+You can look at the CLDR supplemental data or ISO-4217 for the meaning of some
+of these items.
+ISO-4217: http://www.iso.org/iso/en/prods-services/popstds/currencycodeslist.html
+*/
+static const struct CurrencyList {
+    const char *currency;
+    uint32_t currType;
+} gCurrencyList[] = {
+    {"ADP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AED", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AFA", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AFN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ALK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ALL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AMD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ANG", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AOA", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AOK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AON", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AOR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ARA", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ARL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ARM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ARP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ARS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ATS", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AUD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AWG", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"AZM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"AZN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BAD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BAM", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BAN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BBD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BDT", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BEC", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"BEF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BEL", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"BGL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BGM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BGN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BGO", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BHD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BIF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BMD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BND", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BOB", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BOL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BOP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BOV", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"BRB", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BRC", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BRE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BRL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BRN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BRR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BRZ", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BSD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BTN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BUK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BWP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BYB", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BYN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"BYR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"BZD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CAD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CDF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CHE", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"CHF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CHW", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"CLE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"CLF", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"CLP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CNH", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"CNX", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"CNY", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"COP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"COU", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"CRC", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CSD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"CSK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"CUC", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CUP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CVE", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"CYP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"CZK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"DDM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"DEM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"DJF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"DKK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"DOP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"DZD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ECS", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ECV", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"EEK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"EGP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"EQE", UCURR_COMMON|UCURR_DEPRECATED}, // questionable, remove?
+    {"ERN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ESA", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"ESB", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"ESP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ETB", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"EUR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"FIM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"FJD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"FKP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"FRF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GBP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GEK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GEL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GHC", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GHS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GIP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GMD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GNF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GNS", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GQE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GRD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GTQ", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"GWE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GWP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"GYD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"HKD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"HNL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"HRD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"HRK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"HTG", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"HUF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"IDR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"IEP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ILP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ILR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ILS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"INR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"IQD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"IRR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ISJ", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ISK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ITL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"JMD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"JOD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"JPY", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KES", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KGS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KHR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KMF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KPW", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KRH", UCURR_COMMON|UCURR_DEPRECATED},
+    {"KRO", UCURR_COMMON|UCURR_DEPRECATED},
+    {"KRW", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KWD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KYD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"KZT", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LAK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LBP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LKR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LRD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LSL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"LSM", UCURR_COMMON|UCURR_DEPRECATED}, // questionable, remove?
+    {"LTL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"LTT", UCURR_COMMON|UCURR_DEPRECATED},
+    {"LUC", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"LUF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"LUL", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"LVL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"LVR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"LYD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MAD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MAF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MCF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MDC", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MDL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MGA", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MGF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MKD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MKN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MLF", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MMK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MNT", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MOP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MRO", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MRU", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MTL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MTP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MUR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MVP", UCURR_COMMON|UCURR_DEPRECATED}, // questionable, remove?
+    {"MVR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MWK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MXN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MXP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MXV", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"MYR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"MZE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MZM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"MZN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NAD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NGN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NIC", UCURR_COMMON|UCURR_DEPRECATED},
+    {"NIO", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NLG", UCURR_COMMON|UCURR_DEPRECATED},
+    {"NOK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NPR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"NZD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"OMR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PAB", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PEI", UCURR_COMMON|UCURR_DEPRECATED},
+    {"PEN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PES", UCURR_COMMON|UCURR_DEPRECATED},
+    {"PGK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PHP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PKR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PLN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"PLZ", UCURR_COMMON|UCURR_DEPRECATED},
+    {"PTE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"PYG", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"QAR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"RHD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ROL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"RON", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"RSD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"RUB", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"RUR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"RWF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SAR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SBD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SCR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SDD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SDG", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SDP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SEK", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SGD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SHP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SIT", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SKK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SLL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SOS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SRD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SRG", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SSP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"STD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"STN", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SUR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SVC", UCURR_COMMON|UCURR_DEPRECATED},
+    {"SYP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"SZL", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"THB", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TJR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"TJS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TMM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"TMT", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TND", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TOP", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TPE", UCURR_COMMON|UCURR_DEPRECATED},
+    {"TRL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"TRY", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TTD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TWD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"TZS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"UAH", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"UAK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"UGS", UCURR_COMMON|UCURR_DEPRECATED},
+    {"UGX", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"USD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"USN", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"USS", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"UYI", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"UYP", UCURR_COMMON|UCURR_DEPRECATED},
+    {"UYU", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"UZS", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"VEB", UCURR_COMMON|UCURR_DEPRECATED},
+    {"VEF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"VND", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"VNN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"VUV", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"WST", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"XAF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"XAG", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XAU", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XBA", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XBB", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XBC", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XBD", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XCD", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"XDR", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XEU", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"XFO", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XFU", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XOF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"XPD", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XPF", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"XPT", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XRE", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"XSU", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XTS", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XUA", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"XXX", UCURR_UNCOMMON|UCURR_NON_DEPRECATED},
+    {"YDD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"YER", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"YUD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"YUM", UCURR_COMMON|UCURR_DEPRECATED},
+    {"YUN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"YUR", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZAL", UCURR_UNCOMMON|UCURR_DEPRECATED},
+    {"ZAR", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ZMK", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZMW", UCURR_COMMON|UCURR_NON_DEPRECATED},
+    {"ZRN", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZRZ", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZWD", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZWL", UCURR_COMMON|UCURR_DEPRECATED},
+    {"ZWR", UCURR_COMMON|UCURR_DEPRECATED},
+    { NULL, 0 } // Leave here to denote the end of the list.
+};
+
+#define UCURR_MATCHES_BITMASK(variable, typeToMatch) \
+    ((typeToMatch) == UCURR_ALL || ((variable) & (typeToMatch)) == (typeToMatch))
+
+static int32_t U_CALLCONV
+ucurr_countCurrencyList(UEnumeration *enumerator, UErrorCode * /*pErrorCode*/) {
+    UCurrencyContext *myContext = (UCurrencyContext *)(enumerator->context);
+    uint32_t currType = myContext->currType;
+    int32_t count = 0;
+
+    /* Count the number of items matching the type we are looking for. */
+    for (int32_t idx = 0; gCurrencyList[idx].currency != NULL; idx++) {
+        if (UCURR_MATCHES_BITMASK(gCurrencyList[idx].currType, currType)) {
+            count++;
+        }
+    }
+    return count;
+}
+
+static const char* U_CALLCONV
+ucurr_nextCurrencyList(UEnumeration *enumerator,
+                        int32_t* resultLength,
+                        UErrorCode * /*pErrorCode*/)
+{
+    UCurrencyContext *myContext = (UCurrencyContext *)(enumerator->context);
+
+    /* Find the next in the list that matches the type we are looking for. */
+    while (myContext->listIdx < UPRV_LENGTHOF(gCurrencyList)-1) {
+        const struct CurrencyList *currItem = &gCurrencyList[myContext->listIdx++];
+        if (UCURR_MATCHES_BITMASK(currItem->currType, myContext->currType))
+        {
+            if (resultLength) {
+                *resultLength = 3; /* Currency codes are only 3 chars long */
+            }
+            return currItem->currency;
+        }
+    }
+    /* We enumerated too far. */
+    if (resultLength) {
+        *resultLength = 0;
+    }
+    return NULL;
+}
+
+static void U_CALLCONV
+ucurr_resetCurrencyList(UEnumeration *enumerator, UErrorCode * /*pErrorCode*/) {
+    ((UCurrencyContext *)(enumerator->context))->listIdx = 0;
+}
+
+static void U_CALLCONV
+ucurr_closeCurrencyList(UEnumeration *enumerator) {
+    uprv_free(enumerator->context);
+    uprv_free(enumerator);
+}
+
+static void U_CALLCONV
+ucurr_createCurrencyList(UHashtable *isoCodes, UErrorCode* status){
+    UErrorCode localStatus = U_ZERO_ERROR;
+
+    // Look up the CurrencyMap element in the root bundle.
+    UResourceBundle *rb = ures_openDirect(U_ICUDATA_CURR, CURRENCY_DATA, &localStatus);
+    UResourceBundle *currencyMapArray = ures_getByKey(rb, CURRENCY_MAP, rb, &localStatus);
+
+    if (U_SUCCESS(localStatus)) {
+        // process each entry in currency map 
+        for (int32_t i=0; i<ures_getSize(currencyMapArray); i++) {
+            // get the currency resource
+            UResourceBundle *currencyArray = ures_getByIndex(currencyMapArray, i, NULL, &localStatus);
+            // process each currency 
+            if (U_SUCCESS(localStatus)) {
+                for (int32_t j=0; j<ures_getSize(currencyArray); j++) {
+                    // get the currency resource
+                    UResourceBundle *currencyRes = ures_getByIndex(currencyArray, j, NULL, &localStatus);
+                    IsoCodeEntry *entry = (IsoCodeEntry*)uprv_malloc(sizeof(IsoCodeEntry));
+                    if (entry == NULL) {
+                        *status = U_MEMORY_ALLOCATION_ERROR;
+                        return;
+                    }
+
+                    // get the ISO code
+                    int32_t isoLength = 0;
+                    UResourceBundle *idRes = ures_getByKey(currencyRes, "id", NULL, &localStatus);
+                    if (idRes == NULL) {
+                        continue;
+                    }
+                    const UChar *isoCode = ures_getString(idRes, &isoLength, &localStatus);
+
+                    // get from date
+                    UDate fromDate = U_DATE_MIN;
+                    UResourceBundle *fromRes = ures_getByKey(currencyRes, "from", NULL, &localStatus);
+
+                    if (U_SUCCESS(localStatus)) {
+                        int32_t fromLength = 0;
+                        const int32_t *fromArray = ures_getIntVector(fromRes, &fromLength, &localStatus);
+                        int64_t currDate64 = (int64_t)fromArray[0] << 32;
+                        currDate64 |= ((int64_t)fromArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                        fromDate = (UDate)currDate64;
+                    }
+                    ures_close(fromRes);
+
+                    // get to date
+                    UDate toDate = U_DATE_MAX;
+                    localStatus = U_ZERO_ERROR;
+                    UResourceBundle *toRes = ures_getByKey(currencyRes, "to", NULL, &localStatus);
+
+                    if (U_SUCCESS(localStatus)) {
+                        int32_t toLength = 0;
+                        const int32_t *toArray = ures_getIntVector(toRes, &toLength, &localStatus);
+                        int64_t currDate64 = (int64_t)toArray[0] << 32;
+                        currDate64 |= ((int64_t)toArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                        toDate = (UDate)currDate64;
+                    }
+                    ures_close(toRes);
+
+                    ures_close(idRes);
+                    ures_close(currencyRes);
+
+                    entry->isoCode = isoCode;
+                    entry->from = fromDate;
+                    entry->to = toDate;
+
+                    localStatus = U_ZERO_ERROR;
+                    uhash_put(isoCodes, (UChar *)isoCode, entry, &localStatus);
+                }
+            } else {
+                *status = localStatus;
+            }
+            ures_close(currencyArray);
+        }
+    } else {
+        *status = localStatus;
+    }
+
+    ures_close(currencyMapArray);
+}
+
+static const UEnumeration gEnumCurrencyList = {
+    NULL,
+    NULL,
+    ucurr_closeCurrencyList,
+    ucurr_countCurrencyList,
+    uenum_unextDefault,
+    ucurr_nextCurrencyList,
+    ucurr_resetCurrencyList
+};
+U_CDECL_END
+
+
+static void U_CALLCONV initIsoCodes(UErrorCode &status) {
+    U_ASSERT(gIsoCodes == NULL);
+    ucln_common_registerCleanup(UCLN_COMMON_CURRENCY, currency_cleanup);
+
+    UHashtable *isoCodes = uhash_open(uhash_hashUChars, uhash_compareUChars, NULL, &status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    uhash_setValueDeleter(isoCodes, deleteIsoCodeEntry);
+
+    ucurr_createCurrencyList(isoCodes, &status);
+    if (U_FAILURE(status)) {
+        uhash_close(isoCodes);
+        return;
+    }
+    gIsoCodes = isoCodes;  // Note: gIsoCodes is const. Once set up here it is never altered,
+                           //       and read only access is safe without synchronization.
+}
+
+static void populateCurrSymbolsEquiv(icu::Hashtable *hash, UErrorCode &status) {
+    if (U_FAILURE(status)) { return; }
+    for (auto& entry : unisets::kCurrencyEntries) {
+        UnicodeString exemplar(entry.exemplar);
+        const UnicodeSet* set = unisets::get(entry.key);
+        if (set == nullptr) { return; }
+        UnicodeSetIterator it(*set);
+        while (it.next()) {
+            UnicodeString value = it.getString();
+            if (value == exemplar) {
+                // No need to mark the exemplar character as an equivalent
+                continue;
+            }
+            makeEquivalent(exemplar, value, hash, status);
+            if (U_FAILURE(status)) { return; }
+        }
+    }
+}
+
+static void U_CALLCONV initCurrSymbolsEquiv() {
+    U_ASSERT(gCurrSymbolsEquiv == NULL);
+    UErrorCode status = U_ZERO_ERROR;
+    ucln_common_registerCleanup(UCLN_COMMON_CURRENCY, currency_cleanup);
+    icu::Hashtable *temp = new icu::Hashtable(status);
+    if (temp == NULL) {
+        return;
+    }
+    if (U_FAILURE(status)) {
+        delete temp;
+        return;
+    }
+    temp->setValueDeleter(deleteUnicode);
+    populateCurrSymbolsEquiv(temp, status);
+    if (U_FAILURE(status)) {
+        delete temp;
+        return;
+    }
+    gCurrSymbolsEquiv = temp;
+}
+
+U_CAPI UBool U_EXPORT2
+ucurr_isAvailable(const UChar* isoCode, UDate from, UDate to, UErrorCode* eErrorCode) {
+    umtx_initOnce(gIsoCodesInitOnce, &initIsoCodes, *eErrorCode);
+    if (U_FAILURE(*eErrorCode)) {
+        return FALSE;
+    }
+
+    IsoCodeEntry* result = (IsoCodeEntry *) uhash_get(gIsoCodes, isoCode);
+    if (result == NULL) {
+        return FALSE;
+    } else if (from > to) {
+        *eErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    } else if  ((from > result->to) || (to < result->from)) {
+        return FALSE;
+    }
+    return TRUE;
+}
+
+static const icu::Hashtable* getCurrSymbolsEquiv() {
+    umtx_initOnce(gCurrSymbolsEquivInitOnce, &initCurrSymbolsEquiv);
+    return gCurrSymbolsEquiv;
+}
+
+U_CAPI UEnumeration * U_EXPORT2
+ucurr_openISOCurrencies(uint32_t currType, UErrorCode *pErrorCode) {
+    UEnumeration *myEnum = NULL;
+    UCurrencyContext *myContext;
+
+    myEnum = (UEnumeration*)uprv_malloc(sizeof(UEnumeration));
+    if (myEnum == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memcpy(myEnum, &gEnumCurrencyList, sizeof(UEnumeration));
+    myContext = (UCurrencyContext*)uprv_malloc(sizeof(UCurrencyContext));
+    if (myContext == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        uprv_free(myEnum);
+        return NULL;
+    }
+    myContext->currType = currType;
+    myContext->listIdx = 0;
+    myEnum->context = myContext;
+    return myEnum;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucurr_countCurrencies(const char* locale, 
+                 UDate date, 
+                 UErrorCode* ec)
+{
+    int32_t currCount = 0;
+
+    if (ec != NULL && U_SUCCESS(*ec)) 
+    {
+        // local variables
+        UErrorCode localStatus = U_ZERO_ERROR;
+        char id[ULOC_FULLNAME_CAPACITY];
+
+        // get country or country_variant in `id'
+        idForLocale(locale, id, sizeof(id), ec);
+
+        if (U_FAILURE(*ec))
+        {
+            return 0;
+        }
+
+        // Remove variants, which is only needed for registration.
+        char *idDelim = strchr(id, VAR_DELIM);
+        if (idDelim)
+        {
+            idDelim[0] = 0;
+        }
+
+        // Look up the CurrencyMap element in the root bundle.
+        UResourceBundle *rb = ures_openDirect(U_ICUDATA_CURR, CURRENCY_DATA, &localStatus);
+        UResourceBundle *cm = ures_getByKey(rb, CURRENCY_MAP, rb, &localStatus);
+
+        // Using the id derived from the local, get the currency data
+        UResourceBundle *countryArray = ures_getByKey(rb, id, cm, &localStatus);
+
+        // process each currency to see which one is valid for the given date
+        if (U_SUCCESS(localStatus))
+        {
+            for (int32_t i=0; i<ures_getSize(countryArray); i++)
+            {
+                // get the currency resource
+                UResourceBundle *currencyRes = ures_getByIndex(countryArray, i, NULL, &localStatus);
+
+                // get the from date
+                int32_t fromLength = 0;
+                UResourceBundle *fromRes = ures_getByKey(currencyRes, "from", NULL, &localStatus);
+                const int32_t *fromArray = ures_getIntVector(fromRes, &fromLength, &localStatus);
+
+                int64_t currDate64 = (int64_t)fromArray[0] << 32;
+                currDate64 |= ((int64_t)fromArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                UDate fromDate = (UDate)currDate64;
+
+                if (ures_getSize(currencyRes)> 2)
+                {
+                    int32_t toLength = 0;
+                    UResourceBundle *toRes = ures_getByKey(currencyRes, "to", NULL, &localStatus);
+                    const int32_t *toArray = ures_getIntVector(toRes, &toLength, &localStatus);
+
+                    currDate64 = (int64_t)toArray[0] << 32;
+                    currDate64 |= ((int64_t)toArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                    UDate toDate = (UDate)currDate64;
+
+                    if ((fromDate <= date) && (date < toDate))
+                    {
+                        currCount++;
+                    }
+
+                    ures_close(toRes);
+                }
+                else
+                {
+                    if (fromDate <= date)
+                    {
+                        currCount++;
+                    }
+                }
+
+                // close open resources
+                ures_close(currencyRes);
+                ures_close(fromRes);
+
+            } // end For loop
+        } // end if (U_SUCCESS(localStatus))
+
+        ures_close(countryArray);
+
+        // Check for errors
+        if (*ec == U_ZERO_ERROR || localStatus != U_ZERO_ERROR)
+        {
+            // There is nothing to fallback to. 
+            // Report the failure/warning if possible.
+            *ec = localStatus;
+        }
+
+        if (U_SUCCESS(*ec))
+        {
+            // no errors
+            return currCount;
+        }
+
+    }
+
+    // If we got here, either error code is invalid or
+    // some argument passed is no good.
+    return 0;
+}
+
+U_CAPI int32_t U_EXPORT2 
+ucurr_forLocaleAndDate(const char* locale, 
+                UDate date, 
+                int32_t index,
+                UChar* buff, 
+                int32_t buffCapacity, 
+                UErrorCode* ec)
+{
+    int32_t resLen = 0;
+	int32_t currIndex = 0;
+    const UChar* s = NULL;
+
+    if (ec != NULL && U_SUCCESS(*ec))
+    {
+        // check the arguments passed
+        if ((buff && buffCapacity) || !buffCapacity )
+        {
+            // local variables
+            UErrorCode localStatus = U_ZERO_ERROR;
+            char id[ULOC_FULLNAME_CAPACITY];
+
+            // get country or country_variant in `id'
+            idForLocale(locale, id, sizeof(id), ec);
+            if (U_FAILURE(*ec))
+            {
+                return 0;
+            }
+
+            // Remove variants, which is only needed for registration.
+            char *idDelim = strchr(id, VAR_DELIM);
+            if (idDelim)
+            {
+                idDelim[0] = 0;
+            }
+
+            // Look up the CurrencyMap element in the root bundle.
+            UResourceBundle *rb = ures_openDirect(U_ICUDATA_CURR, CURRENCY_DATA, &localStatus);
+            UResourceBundle *cm = ures_getByKey(rb, CURRENCY_MAP, rb, &localStatus);
+
+            // Using the id derived from the local, get the currency data
+            UResourceBundle *countryArray = ures_getByKey(rb, id, cm, &localStatus);
+
+            // process each currency to see which one is valid for the given date
+            bool matchFound = false;
+            if (U_SUCCESS(localStatus))
+            {
+                if ((index <= 0) || (index> ures_getSize(countryArray)))
+                {
+                    // requested index is out of bounds
+                    ures_close(countryArray);
+                    return 0;
+                }
+
+                for (int32_t i=0; i<ures_getSize(countryArray); i++)
+                {
+                    // get the currency resource
+                    UResourceBundle *currencyRes = ures_getByIndex(countryArray, i, NULL, &localStatus);
+                    s = ures_getStringByKey(currencyRes, "id", &resLen, &localStatus);
+
+                    // get the from date
+                    int32_t fromLength = 0;
+                    UResourceBundle *fromRes = ures_getByKey(currencyRes, "from", NULL, &localStatus);
+                    const int32_t *fromArray = ures_getIntVector(fromRes, &fromLength, &localStatus);
+
+                    int64_t currDate64 = (int64_t)fromArray[0] << 32;
+                    currDate64 |= ((int64_t)fromArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                    UDate fromDate = (UDate)currDate64;
+
+                    if (ures_getSize(currencyRes)> 2)
+                    {
+                        int32_t toLength = 0;
+                        UResourceBundle *toRes = ures_getByKey(currencyRes, "to", NULL, &localStatus);
+                        const int32_t *toArray = ures_getIntVector(toRes, &toLength, &localStatus);
+
+                        currDate64 = (int64_t)toArray[0] << 32;
+                        currDate64 |= ((int64_t)toArray[1] & (int64_t)INT64_C(0x00000000FFFFFFFF));
+                        UDate toDate = (UDate)currDate64;
+
+                        if ((fromDate <= date) && (date < toDate))
+                        {
+                            currIndex++;
+                            if (currIndex == index)
+                            {
+                                matchFound = true;
+                            }
+                        }
+
+                        ures_close(toRes);
+                    }
+                    else
+                    {
+                        if (fromDate <= date)
+                        {
+                            currIndex++;
+                            if (currIndex == index)
+                            {
+                                matchFound = true;
+                            }
+                        }
+                    }
+
+                    // close open resources
+                    ures_close(currencyRes);
+                    ures_close(fromRes);
+
+                    // check for loop exit
+                    if (matchFound)
+                    {
+                        break;
+                    }
+
+                } // end For loop
+            }
+
+            ures_close(countryArray);
+
+            // Check for errors
+            if (*ec == U_ZERO_ERROR || localStatus != U_ZERO_ERROR)
+            {
+                // There is nothing to fallback to. 
+                // Report the failure/warning if possible.
+                *ec = localStatus;
+            }
+
+            if (U_SUCCESS(*ec))
+            {
+                // no errors
+                if((buffCapacity> resLen) && matchFound)
+                {
+                    // write out the currency value
+                    u_strcpy(buff, s);
+                }
+                else
+                {
+                    return 0;
+                }
+            }
+
+            // return null terminated currency string
+            return u_terminateUChars(buff, buffCapacity, resLen, ec);
+        }
+        else
+        {
+            // illegal argument encountered
+            *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+
+    }
+
+    // If we got here, either error code is invalid or
+    // some argument passed is no good.
+    return resLen;
+}
+
+static const UEnumeration defaultKeywordValues = {
+    NULL,
+    NULL,
+    ulist_close_keyword_values_iterator,
+    ulist_count_keyword_values,
+    uenum_unextDefault,
+    ulist_next_keyword_value, 
+    ulist_reset_keyword_values_iterator
+};
+
+U_CAPI UEnumeration *U_EXPORT2 ucurr_getKeywordValuesForLocale(const char *key, const char *locale, UBool commonlyUsed, UErrorCode* status) {
+    // Resolve region
+    char prefRegion[ULOC_COUNTRY_CAPACITY];
+    ulocimp_getRegionForSupplementalData(locale, TRUE, prefRegion, sizeof(prefRegion), status);
+    
+    // Read value from supplementalData
+    UList *values = ulist_createEmptyList(status);
+    UList *otherValues = ulist_createEmptyList(status);
+    UEnumeration *en = (UEnumeration *)uprv_malloc(sizeof(UEnumeration));
+    if (U_FAILURE(*status) || en == NULL) {
+        if (en == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            uprv_free(en);
+        }
+        ulist_deleteList(values);
+        ulist_deleteList(otherValues);
+        return NULL;
+    }
+    memcpy(en, &defaultKeywordValues, sizeof(UEnumeration));
+    en->context = values;
+    
+    UResourceBundle *bundle = ures_openDirect(U_ICUDATA_CURR, "supplementalData", status);
+    ures_getByKey(bundle, "CurrencyMap", bundle, status);
+    UResourceBundle bundlekey, regbndl, curbndl, to;
+    ures_initStackObject(&bundlekey);
+    ures_initStackObject(&regbndl);
+    ures_initStackObject(&curbndl);
+    ures_initStackObject(&to);
+    
+    while (U_SUCCESS(*status) && ures_hasNext(bundle)) {
+        ures_getNextResource(bundle, &bundlekey, status);
+        if (U_FAILURE(*status)) {
+            break;
+        }
+        const char *region = ures_getKey(&bundlekey);
+        UBool isPrefRegion = uprv_strcmp(region, prefRegion) == 0 ? TRUE : FALSE;
+        if (!isPrefRegion && commonlyUsed) {
+            // With commonlyUsed=true, we do not put
+            // currencies for other regions in the
+            // result list.
+            continue;
+        }
+        ures_getByKey(bundle, region, &regbndl, status);
+        if (U_FAILURE(*status)) {
+            break;
+        }
+        while (U_SUCCESS(*status) && ures_hasNext(&regbndl)) {
+            ures_getNextResource(&regbndl, &curbndl, status);
+            if (ures_getType(&curbndl) != URES_TABLE) {
+                // Currently, an empty ARRAY is mixed in.
+                continue;
+            }
+            char *curID = (char *)uprv_malloc(sizeof(char) * ULOC_KEYWORDS_CAPACITY);
+            int32_t curIDLength = ULOC_KEYWORDS_CAPACITY;
+            if (curID == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                break;
+            }
+
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+            ures_getUTF8StringByKey(&curbndl, "id", curID, &curIDLength, TRUE, status);
+            /* optimize - use the utf-8 string */
+#else
+            {
+                       const UChar* defString = ures_getStringByKey(&curbndl, "id", &curIDLength, status);
+                       if(U_SUCCESS(*status)) {
+			   if(curIDLength+1 > ULOC_KEYWORDS_CAPACITY) {
+				*status = U_BUFFER_OVERFLOW_ERROR;
+			   } else {
+                           	u_UCharsToChars(defString, curID, curIDLength+1);
+			   }
+                       }
+            }
+#endif	
+
+            if (U_FAILURE(*status)) {
+                break;
+            }
+            UBool hasTo = FALSE;
+            ures_getByKey(&curbndl, "to", &to, status);
+            if (U_FAILURE(*status)) {
+                // Do nothing here...
+                *status = U_ZERO_ERROR;
+            } else {
+                hasTo = TRUE;
+            }
+            if (isPrefRegion && !hasTo && !ulist_containsString(values, curID, (int32_t)uprv_strlen(curID))) {
+                // Currently active currency for the target country
+                ulist_addItemEndList(values, curID, TRUE, status);
+            } else if (!ulist_containsString(otherValues, curID, (int32_t)uprv_strlen(curID)) && !commonlyUsed) {
+                ulist_addItemEndList(otherValues, curID, TRUE, status);
+            } else {
+                uprv_free(curID);
+            }
+        }
+        
+    }
+    if (U_SUCCESS(*status)) {
+        if (commonlyUsed) {
+            if (ulist_getListSize(values) == 0) {
+                // This could happen if no valid region is supplied in the input
+                // locale. In this case, we use the CLDR's default.
+                uenum_close(en);
+                en = ucurr_getKeywordValuesForLocale(key, "und", TRUE, status);
+            }
+        } else {
+            // Consolidate the list
+            char *value = NULL;
+            ulist_resetList(otherValues);
+            while ((value = (char *)ulist_getNext(otherValues)) != NULL) {
+                if (!ulist_containsString(values, value, (int32_t)uprv_strlen(value))) {
+                    char *tmpValue = (char *)uprv_malloc(sizeof(char) * ULOC_KEYWORDS_CAPACITY);
+                    uprv_memcpy(tmpValue, value, uprv_strlen(value) + 1);
+                    ulist_addItemEndList(values, tmpValue, TRUE, status);
+                    if (U_FAILURE(*status)) {
+                        break;
+                    }
+                }
+            }
+        }
+        
+        ulist_resetList((UList *)(en->context));
+    } else {
+        ulist_deleteList(values);
+        uprv_free(en);
+        values = NULL;
+        en = NULL;
+    }
+    ures_close(&to);
+    ures_close(&curbndl);
+    ures_close(&regbndl);
+    ures_close(&bundlekey);
+    ures_close(bundle);
+    
+    ulist_deleteList(otherValues);
+    
+    return en;
+}
+
+
+U_CAPI int32_t U_EXPORT2
+ucurr_getNumericCode(const UChar* currency) {
+    int32_t code = 0;
+    if (currency && u_strlen(currency) == ISO_CURRENCY_CODE_LENGTH) {
+        UErrorCode status = U_ZERO_ERROR;
+
+        UResourceBundle *bundle = ures_openDirect(0, "currencyNumericCodes", &status);
+        ures_getByKey(bundle, "codeMap", bundle, &status);
+        if (U_SUCCESS(status)) {
+            char alphaCode[ISO_CURRENCY_CODE_LENGTH+1];
+            myUCharsToChars(alphaCode, currency);
+            T_CString_toUpperCase(alphaCode);
+            ures_getByKey(bundle, alphaCode, bundle, &status);
+            int tmpCode = ures_getInt(bundle, &status);
+            if (U_SUCCESS(status)) {
+                code = tmpCode;
+            }
+        }
+        ures_close(bundle);
+    }
+    return code;
+}
+#endif /* #if !UCONFIG_NO_FORMATTING */
+
+//eof
diff --git a/thirdparty/icu4c/common/ucurrimp.h b/thirdparty/icu4c/common/ucurrimp.h
new file mode 100644
index 0000000000..6d9588295d
--- /dev/null
+++ b/thirdparty/icu4c/common/ucurrimp.h
@@ -0,0 +1,78 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef _UCURR_IMP_H_
+#define _UCURR_IMP_H_
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/parsepos.h"
+#include "unicode/uniset.h"
+
+/**
+ * Internal method.  Given a currency ISO code and a locale, return
+ * the "static" currency name.  This is usually the same as the
+ * UCURR_SYMBOL_NAME, but if the latter is a choice format, then the
+ * format is applied to the number 2.0 (to yield the more common
+ * plural) to return a static name.
+ *
+ * This is used for backward compatibility with old currency logic in
+ * DecimalFormat and DecimalFormatSymbols.
+ */
+U_CAPI void
+uprv_getStaticCurrencyName(const UChar* iso, const char* loc,
+                           icu::UnicodeString& result, UErrorCode& ec);
+
+/**
+ * Attempt to parse the given string as a currency, either as a
+ * display name in the given locale, or as a 3-letter ISO 4217
+ * code.  If multiple display names match, then the longest one is
+ * selected.  If both a display name and a 3-letter ISO code
+ * match, then the display name is preferred, unless it's length
+ * is less than 3.
+ *
+ * The parameters must not be NULL.
+ *
+ * @param locale the locale of the display names to match
+ * @param text the text to parse
+ * @param pos input-output position; on input, the position within
+ * text to match; must have 0 <= pos.getIndex() < text.length();
+ * on output, the position after the last matched character. If
+ * the parse fails, the position in unchanged upon output.
+ * @param type currency type to parse against, LONG_NAME only or not
+ * @param partialMatchLen The length of the longest matching prefix;
+ * this may be nonzero even if no full currency was matched.
+ * @return the ISO 4217 code, as a string, of the best match, or
+ * null if there is no match
+ *
+ * @internal
+ */
+U_CAPI void
+uprv_parseCurrency(const char* locale,
+                   const icu::UnicodeString& text,
+                   icu::ParsePosition& pos,
+                   int8_t type,
+                   int32_t* partialMatchLen,
+                   UChar* result,
+                   UErrorCode& ec);
+
+/**
+ * Puts all possible first-characters of a currency into the
+ * specified UnicodeSet.
+ *
+ * @param locale the locale of the display names of interest
+ * @param result the UnicodeSet to which to add the starting characters
+ */
+void uprv_currencyLeads(const char* locale, icu::UnicodeSet& result, UErrorCode& ec);
+
+
+
+#endif /* #ifndef _UCURR_IMP_H_ */
+
+//eof
diff --git a/thirdparty/icu4c/common/udata.cpp b/thirdparty/icu4c/common/udata.cpp
new file mode 100644
index 0000000000..ec9c999cea
--- /dev/null
+++ b/thirdparty/icu4c/common/udata.cpp
@@ -0,0 +1,1460 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  udata.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999oct25
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"  /* U_PLATFORM etc. */
+
+#ifdef __GNUC__
+/* if gcc
+#define ATTRIBUTE_WEAK __attribute__ ((weak))
+might have to #include some other header
+*/
+#endif
+
+#include "unicode/putil.h"
+#include "unicode/udata.h"
+#include "unicode/uversion.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "putilimp.h"
+#include "restrace.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "ucmndata.h"
+#include "udatamem.h"
+#include "uhash.h"
+#include "umapfile.h"
+#include "umutex.h"
+
+/***********************************************************************
+*
+*   Notes on the organization of the ICU data implementation
+*
+*      All of the public API is defined in udata.h
+*
+*      The implementation is split into several files...
+*
+*         - udata.c  (this file) contains higher level code that knows about
+*                     the search paths for locating data, caching opened data, etc.
+*
+*         - umapfile.c  contains the low level platform-specific code for actually loading
+*                     (memory mapping, file reading, whatever) data into memory.
+*
+*         - ucmndata.c  deals with the tables of contents of ICU data items within
+*                     an ICU common format data file.  The implementation includes
+*                     an abstract interface and support for multiple TOC formats.
+*                     All knowledge of any specific TOC format is encapsulated here.
+*
+*         - udatamem.c has code for managing UDataMemory structs.  These are little
+*                     descriptor objects for blocks of memory holding ICU data of
+*                     various types.
+*/
+
+/* configuration ---------------------------------------------------------- */
+
+/* If you are excruciatingly bored turn this on .. */
+/* #define UDATA_DEBUG 1 */
+
+#if defined(UDATA_DEBUG)
+#   include <stdio.h>
+#endif
+
+U_NAMESPACE_USE
+
+/*
+ *  Forward declarations
+ */
+static UDataMemory *udata_findCachedData(const char *path, UErrorCode &err);
+
+/***********************************************************************
+*
+*    static (Global) data
+*
+************************************************************************/
+
+/*
+ * Pointers to the common ICU data.
+ *
+ * We store multiple pointers to ICU data packages and iterate through them
+ * when looking for a data item.
+ *
+ * It is possible to combine this with dependency inversion:
+ * One or more data package libraries may export
+ * functions that each return a pointer to their piece of the ICU data,
+ * and this file would import them as weak functions, without a
+ * strong linker dependency from the common library on the data library.
+ *
+ * Then we can have applications depend on only that part of ICU's data
+ * that they really need, reducing the size of binaries that take advantage
+ * of this.
+ */
+static UDataMemory *gCommonICUDataArray[10] = { NULL };   // Access protected by icu global mutex.
+
+static u_atomic_int32_t gHaveTriedToLoadCommonData = ATOMIC_INT32_T_INITIALIZER(0);  //  See extendICUData().
+
+static UHashtable  *gCommonDataCache = NULL;  /* Global hash table of opened ICU data files.  */
+static icu::UInitOnce gCommonDataCacheInitOnce = U_INITONCE_INITIALIZER;
+
+#if !defined(ICU_DATA_DIR_WINDOWS)
+static UDataFileAccess  gDataFileAccess = UDATA_DEFAULT_ACCESS;  // Access not synchronized.
+                                                                 // Modifying is documented as thread-unsafe.
+#else
+// If we are using the Windows data directory, then look in one spot only.
+static UDataFileAccess  gDataFileAccess = UDATA_NO_FILES;
+#endif
+
+static UBool U_CALLCONV
+udata_cleanup(void)
+{
+    int32_t i;
+
+    if (gCommonDataCache) {             /* Delete the cache of user data mappings.  */
+        uhash_close(gCommonDataCache);  /*   Table owns the contents, and will delete them. */
+        gCommonDataCache = NULL;        /*   Cleanup is not thread safe.                */
+    }
+    gCommonDataCacheInitOnce.reset();
+
+    for (i = 0; i < UPRV_LENGTHOF(gCommonICUDataArray) && gCommonICUDataArray[i] != NULL; ++i) {
+        udata_close(gCommonICUDataArray[i]);
+        gCommonICUDataArray[i] = NULL;
+    }
+    gHaveTriedToLoadCommonData = 0;
+
+    return TRUE;                   /* Everything was cleaned up */
+}
+
+static UBool U_CALLCONV
+findCommonICUDataByName(const char *inBasename, UErrorCode &err)
+{
+    UBool found = FALSE;
+    int32_t i;
+
+    UDataMemory  *pData = udata_findCachedData(inBasename, err);
+    if (U_FAILURE(err) || pData == NULL)
+        return FALSE;
+
+    {
+        Mutex lock;
+        for (i = 0; i < UPRV_LENGTHOF(gCommonICUDataArray); ++i) {
+            if ((gCommonICUDataArray[i] != NULL) && (gCommonICUDataArray[i]->pHeader == pData->pHeader)) {
+                /* The data pointer is already in the array. */
+                found = TRUE;
+                break;
+            }
+        }
+    }
+    return found;
+}
+
+
+/*
+ * setCommonICUData.   Set a UDataMemory to be the global ICU Data
+ */
+static UBool
+setCommonICUData(UDataMemory *pData,     /*  The new common data.  Belongs to caller, we copy it. */
+                 UBool       warn,       /*  If true, set USING_DEFAULT warning if ICUData was    */
+                                         /*    changed by another thread before we got to it.     */
+                 UErrorCode *pErr)
+{
+    UDataMemory  *newCommonData = UDataMemory_createNewInstance(pErr);
+    int32_t i;
+    UBool didUpdate = FALSE;
+    if (U_FAILURE(*pErr)) {
+        return FALSE;
+    }
+
+    /*  For the assignment, other threads must cleanly see either the old            */
+    /*    or the new, not some partially initialized new.  The old can not be        */
+    /*    deleted - someone may still have a pointer to it lying around in           */
+    /*    their locals.                                                              */
+    UDatamemory_assign(newCommonData, pData);
+    umtx_lock(NULL);
+    for (i = 0; i < UPRV_LENGTHOF(gCommonICUDataArray); ++i) {
+        if (gCommonICUDataArray[i] == NULL) {
+            gCommonICUDataArray[i] = newCommonData;
+            didUpdate = TRUE;
+            break;
+        } else if (gCommonICUDataArray[i]->pHeader == pData->pHeader) {
+            /* The same data pointer is already in the array. */
+            break;
+        }
+    }
+    umtx_unlock(NULL);
+
+    if (i == UPRV_LENGTHOF(gCommonICUDataArray) && warn) {
+        *pErr = U_USING_DEFAULT_WARNING;
+    }
+    if (didUpdate) {
+        ucln_common_registerCleanup(UCLN_COMMON_UDATA, udata_cleanup);
+    } else {
+        uprv_free(newCommonData);
+    }
+    return didUpdate;
+}
+
+#if !defined(ICU_DATA_DIR_WINDOWS)
+
+static UBool
+setCommonICUDataPointer(const void *pData, UBool /*warn*/, UErrorCode *pErrorCode) {
+    UDataMemory tData;
+    UDataMemory_init(&tData);
+    UDataMemory_setData(&tData, pData);
+    udata_checkCommonData(&tData, pErrorCode);
+    return setCommonICUData(&tData, FALSE, pErrorCode);
+}
+
+#endif
+
+static const char *
+findBasename(const char *path) {
+    const char *basename=uprv_strrchr(path, U_FILE_SEP_CHAR);
+    if(basename==NULL) {
+        return path;
+    } else {
+        return basename+1;
+    }
+}
+
+#ifdef UDATA_DEBUG
+static const char *
+packageNameFromPath(const char *path)
+{
+    if((path == NULL) || (*path == 0)) {
+        return U_ICUDATA_NAME;
+    }
+
+    path = findBasename(path);
+
+    if((path == NULL) || (*path == 0)) {
+        return U_ICUDATA_NAME;
+    }
+
+    return path;
+}
+#endif
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *   Cache for common data                                              *
+ *      Functions for looking up or adding entries to a cache of        *
+ *      data that has been previously opened.  Avoids a potentially     *
+ *      expensive operation of re-opening the data for subsequent       *
+ *      uses.                                                           *
+ *                                                                      *
+ *      Data remains cached for the duration of the process.            *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+
+typedef struct DataCacheElement {
+    char          *name;
+    UDataMemory   *item;
+} DataCacheElement;
+
+
+
+/*
+ * Deleter function for DataCacheElements.
+ *         udata cleanup function closes the hash table; hash table in turn calls back to
+ *         here for each entry.
+ */
+static void U_CALLCONV DataCacheElement_deleter(void *pDCEl) {
+    DataCacheElement *p = (DataCacheElement *)pDCEl;
+    udata_close(p->item);              /* unmaps storage */
+    uprv_free(p->name);                /* delete the hash key string. */
+    uprv_free(pDCEl);                  /* delete 'this'          */
+}
+
+static void U_CALLCONV udata_initHashTable(UErrorCode &err) {
+    U_ASSERT(gCommonDataCache == NULL);
+    gCommonDataCache = uhash_open(uhash_hashChars, uhash_compareChars, NULL, &err);
+    if (U_FAILURE(err)) {
+       return;
+    }
+    U_ASSERT(gCommonDataCache != NULL);
+    uhash_setValueDeleter(gCommonDataCache, DataCacheElement_deleter);
+    ucln_common_registerCleanup(UCLN_COMMON_UDATA, udata_cleanup);
+}
+
+ /*   udata_getCacheHashTable()
+  *     Get the hash table used to store the data cache entries.
+  *     Lazy create it if it doesn't yet exist.
+  */
+static UHashtable *udata_getHashTable(UErrorCode &err) {
+    umtx_initOnce(gCommonDataCacheInitOnce, &udata_initHashTable, err);
+    return gCommonDataCache;
+}
+
+
+
+static UDataMemory *udata_findCachedData(const char *path, UErrorCode &err)
+{
+    UHashtable        *htable;
+    UDataMemory       *retVal = NULL;
+    DataCacheElement  *el;
+    const char        *baseName;
+
+    htable = udata_getHashTable(err);
+    if (U_FAILURE(err)) {
+        return NULL;
+    }
+
+    baseName = findBasename(path);   /* Cache remembers only the base name, not the full path. */
+    umtx_lock(NULL);
+    el = (DataCacheElement *)uhash_get(htable, baseName);
+    umtx_unlock(NULL);
+    if (el != NULL) {
+        retVal = el->item;
+    }
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "Cache: [%s] -> %p\n", baseName, (void*) retVal);
+#endif
+    return retVal;
+}
+
+
+static UDataMemory *udata_cacheDataItem(const char *path, UDataMemory *item, UErrorCode *pErr) {
+    DataCacheElement *newElement;
+    const char       *baseName;
+    int32_t           nameLen;
+    UHashtable       *htable;
+    DataCacheElement *oldValue = NULL;
+    UErrorCode        subErr = U_ZERO_ERROR;
+
+    htable = udata_getHashTable(*pErr);
+    if (U_FAILURE(*pErr)) {
+        return NULL;
+    }
+
+    /* Create a new DataCacheElement - the thingy we store in the hash table -
+     * and copy the supplied path and UDataMemoryItems into it.
+     */
+    newElement = (DataCacheElement *)uprv_malloc(sizeof(DataCacheElement));
+    if (newElement == NULL) {
+        *pErr = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    newElement->item = UDataMemory_createNewInstance(pErr);
+    if (U_FAILURE(*pErr)) {
+        uprv_free(newElement);
+        return NULL;
+    }
+    UDatamemory_assign(newElement->item, item);
+
+    baseName = findBasename(path);
+    nameLen = (int32_t)uprv_strlen(baseName);
+    newElement->name = (char *)uprv_malloc(nameLen+1);
+    if (newElement->name == NULL) {
+        *pErr = U_MEMORY_ALLOCATION_ERROR;
+        uprv_free(newElement->item);
+        uprv_free(newElement);
+        return NULL;
+    }
+    uprv_strcpy(newElement->name, baseName);
+
+    /* Stick the new DataCacheElement into the hash table.
+    */
+    umtx_lock(NULL);
+    oldValue = (DataCacheElement *)uhash_get(htable, path);
+    if (oldValue != NULL) {
+        subErr = U_USING_DEFAULT_WARNING;
+    }
+    else {
+        uhash_put(
+            htable,
+            newElement->name,               /* Key   */
+            newElement,                     /* Value */
+            &subErr);
+    }
+    umtx_unlock(NULL);
+
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "Cache: [%s] <<< %p : %s. vFunc=%p\n", newElement->name, 
+    (void*) newElement->item, u_errorName(subErr), (void*) newElement->item->vFuncs);
+#endif
+
+    if (subErr == U_USING_DEFAULT_WARNING || U_FAILURE(subErr)) {
+        *pErr = subErr; /* copy sub err unto fillin ONLY if something happens. */
+        uprv_free(newElement->name);
+        uprv_free(newElement->item);
+        uprv_free(newElement);
+        return oldValue ? oldValue->item : NULL;
+    }
+
+    return newElement->item;
+}
+
+/*----------------------------------------------------------------------*==============
+ *                                                                      *
+ *  Path management.  Could be shared with other tools/etc if need be   *
+ * later on.                                                            *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+
+U_NAMESPACE_BEGIN
+
+class UDataPathIterator
+{
+public:
+    UDataPathIterator(const char *path, const char *pkg,
+                      const char *item, const char *suffix, UBool doCheckLastFour,
+                      UErrorCode *pErrorCode);
+    const char *next(UErrorCode *pErrorCode);
+
+private:
+    const char *path;                              /* working path (u_icudata_Dir) */
+    const char *nextPath;                          /* path following this one */
+    const char *basename;                          /* item's basename (icudt22e_mt.res)*/
+
+    StringPiece suffix;                            /* item suffix (can be null) */
+
+    uint32_t    basenameLen;                       /* length of basename */
+
+    CharString  itemPath;                          /* path passed in with item name */
+    CharString  pathBuffer;                        /* output path for this it'ion */
+    CharString  packageStub;                       /* example:  "/icudt28b". Will ignore that leaf in set paths. */
+
+    UBool       checkLastFour;                     /* if TRUE then allow paths such as '/foo/myapp.dat'
+                                                    * to match, checks last 4 chars of suffix with
+                                                    * last 4 of path, then previous chars. */
+};
+
+/**
+ * @param iter    The iterator to be initialized. Its current state does not matter.
+ * @param inPath  The full pathname to be iterated over.  If NULL, defaults to U_ICUDATA_NAME 
+ * @param pkg     Package which is being searched for, ex "icudt28l".  Will ignore leaf directories such as /icudt28l 
+ * @param item    Item to be searched for.  Can include full path, such as /a/b/foo.dat 
+ * @param inSuffix  Optional item suffix, if not-null (ex. ".dat") then 'path' can contain 'item' explicitly.
+ *             Ex:   'stuff.dat' would be found in '/a/foo:/tmp/stuff.dat:/bar/baz' as item #2.   
+ *                   '/blarg/stuff.dat' would also be found.
+ *  Note: inSuffix may also be the 'item' being searched for as well, (ex: "ibm-5348_P100-1997.cnv"), in which case 
+ *        the 'item' parameter is often the same as pkg. (Though sometimes might have a tree part as well, ex: "icudt62l-curr").
+ */
+UDataPathIterator::UDataPathIterator(const char *inPath, const char *pkg,
+                                     const char *item, const char *inSuffix, UBool doCheckLastFour,
+                                     UErrorCode *pErrorCode)
+{
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "SUFFIX1=%s PATH=%s\n", inSuffix, inPath);
+#endif
+    /** Path **/
+    if(inPath == NULL) {
+        path = u_getDataDirectory();
+    } else {
+        path = inPath;
+    }
+
+    /** Package **/
+    if(pkg != NULL) {
+      packageStub.append(U_FILE_SEP_CHAR, *pErrorCode).append(pkg, *pErrorCode);
+#ifdef UDATA_DEBUG
+      fprintf(stderr, "STUB=%s [%d]\n", packageStub.data(), packageStub.length());
+#endif
+    }
+
+    /** Item **/
+    basename = findBasename(item);
+    basenameLen = (int32_t)uprv_strlen(basename);
+
+    /** Item path **/
+    if(basename == item) {
+        nextPath = path;
+    } else {
+        itemPath.append(item, (int32_t)(basename-item), *pErrorCode);
+        nextPath = itemPath.data();
+    }
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "SUFFIX=%s [%p]\n", inSuffix, (void*) inSuffix);
+#endif
+
+    /** Suffix  **/
+    if(inSuffix != NULL) {
+        suffix = inSuffix;
+    } else {
+        suffix = "";
+    }
+
+    checkLastFour = doCheckLastFour;
+
+    /* pathBuffer will hold the output path strings returned by this iterator */
+
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "0: init %s -> [path=%s], [base=%s], [suff=%s], [itempath=%s], [nextpath=%s], [checklast4=%s]\n",
+            item,
+            path,
+            basename,
+            suffix.data(),
+            itemPath.data(),
+            nextPath,
+            checkLastFour?"TRUE":"false");
+#endif
+}
+
+/**
+ * Get the next path on the list.
+ *
+ * @param iter The Iter to be used 
+ * @param len  If set, pointer to the length of the returned path, for convenience. 
+ * @return Pointer to the next path segment, or NULL if there are no more.
+ */
+const char *UDataPathIterator::next(UErrorCode *pErrorCode)
+{
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+
+    const char *currentPath = NULL;
+    int32_t     pathLen = 0;
+    const char *pathBasename;
+
+    do
+    {
+        if( nextPath == NULL ) {
+            break;
+        }
+        currentPath = nextPath;
+
+        if(nextPath == itemPath.data()) { /* we were processing item's path. */
+            nextPath = path; /* start with regular path next tm. */
+            pathLen = (int32_t)uprv_strlen(currentPath);
+        } else {
+            /* fix up next for next time */
+            nextPath = uprv_strchr(currentPath, U_PATH_SEP_CHAR);
+            if(nextPath == NULL) {
+                /* segment: entire path */
+                pathLen = (int32_t)uprv_strlen(currentPath); 
+            } else {
+                /* segment: until next segment */
+                pathLen = (int32_t)(nextPath - currentPath);
+                /* skip divider */
+                nextPath ++;
+            }
+        }
+
+        if(pathLen == 0) {
+            continue;
+        }
+
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "rest of path (IDD) = %s\n", currentPath);
+        fprintf(stderr, "                     ");
+        { 
+            int32_t qqq;
+            for(qqq=0;qqq<pathLen;qqq++)
+            {
+                fprintf(stderr, " ");
+            }
+
+            fprintf(stderr, "^\n");
+        }
+#endif
+        pathBuffer.clear().append(currentPath, pathLen, *pErrorCode);
+
+        /* check for .dat files */
+        pathBasename = findBasename(pathBuffer.data());
+
+        if(checkLastFour == TRUE && 
+           (pathLen>=4) &&
+           uprv_strncmp(pathBuffer.data() +(pathLen-4), suffix.data(), 4)==0 && /* suffix matches */
+           uprv_strncmp(findBasename(pathBuffer.data()), basename, basenameLen)==0  && /* base matches */
+           uprv_strlen(pathBasename)==(basenameLen+4)) { /* base+suffix = full len */
+
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "Have %s file on the path: %s\n", suffix.data(), pathBuffer.data());
+#endif
+            /* do nothing */
+        }
+        else 
+        {       /* regular dir path */
+            if(pathBuffer[pathLen-1] != U_FILE_SEP_CHAR) {
+                if((pathLen>=4) &&
+                   uprv_strncmp(pathBuffer.data()+(pathLen-4), ".dat", 4) == 0)
+                {
+#ifdef UDATA_DEBUG
+                    fprintf(stderr, "skipping non-directory .dat file %s\n", pathBuffer.data());
+#endif
+                    continue;
+                }
+
+                /* Check if it is a directory with the same name as our package */
+                if(!packageStub.isEmpty() &&
+                   (pathLen > packageStub.length()) &&
+                   !uprv_strcmp(pathBuffer.data() + pathLen - packageStub.length(), packageStub.data())) {
+#ifdef UDATA_DEBUG
+                  fprintf(stderr, "Found stub %s (will add package %s of len %d)\n", packageStub.data(), basename, basenameLen);
+#endif
+                  pathBuffer.truncate(pathLen - packageStub.length());
+                }
+                pathBuffer.append(U_FILE_SEP_CHAR, *pErrorCode);
+            }
+
+            /* + basename */
+            pathBuffer.append(packageStub.data()+1, packageStub.length()-1, *pErrorCode);
+
+            if (!suffix.empty())  /* tack on suffix */
+            {
+                if (suffix.length() > 4) {
+                    // If the suffix is actually an item ("ibm-5348_P100-1997.cnv") and not an extension (".res")
+                    // then we need to ensure that the path ends with a separator.
+                    pathBuffer.ensureEndsWithFileSeparator(*pErrorCode);
+                }
+                pathBuffer.append(suffix, *pErrorCode);
+            }
+        }
+
+#ifdef UDATA_DEBUG
+        fprintf(stderr, " -->  %s\n", pathBuffer.data());
+#endif
+
+        return pathBuffer.data();
+
+    } while(path);
+
+    /* fell way off the end */
+    return NULL;
+}
+
+U_NAMESPACE_END
+
+/* ==================================================================================*/
+
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *  Add a static reference to the common data library                   *
+ *   Unless overridden by an explicit udata_setCommonData, this will be *
+ *      our common data.                                                *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+#if !defined(ICU_DATA_DIR_WINDOWS)
+// When using the Windows system data, we expect only a single data file.
+extern "C" const DataHeader U_DATA_API U_ICUDATA_ENTRY_POINT;
+#endif
+
+/*
+ * This would be a good place for weak-linkage declarations of
+ * partial-data-library access functions where each returns a pointer
+ * to its data package, if it is linked in.
+ */
+/*
+extern const void *uprv_getICUData_collation(void) ATTRIBUTE_WEAK;
+extern const void *uprv_getICUData_conversion(void) ATTRIBUTE_WEAK;
+*/
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *   openCommonData   Attempt to open a common format (.dat) file       *
+ *                    Map it into memory (if it's not there already)    *
+ *                    and return a UDataMemory object for it.           *
+ *                                                                      *
+ *                    If the requested data is already open and cached  *
+ *                       just return the cached UDataMem object.        *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+static UDataMemory *
+openCommonData(const char *path,          /*  Path from OpenChoice?          */
+               int32_t commonDataIndex,   /*  ICU Data (index >= 0) if path == NULL */
+               UErrorCode *pErrorCode)
+{
+    UDataMemory tData;
+    const char *pathBuffer;
+    const char *inBasename;
+
+    if (U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+
+    UDataMemory_init(&tData);
+
+    /* ??????? TODO revisit this */ 
+    if (commonDataIndex >= 0) {
+        /* "mini-cache" for common ICU data */
+        if(commonDataIndex >= UPRV_LENGTHOF(gCommonICUDataArray)) {
+            return NULL;
+        }
+        {
+            Mutex lock;
+            if(gCommonICUDataArray[commonDataIndex] != NULL) {
+                return gCommonICUDataArray[commonDataIndex];
+            }
+#if !defined(ICU_DATA_DIR_WINDOWS)
+// When using the Windows system data, we expect only a single data file.
+            int32_t i;
+            for(i = 0; i < commonDataIndex; ++i) {
+                if(gCommonICUDataArray[i]->pHeader == &U_ICUDATA_ENTRY_POINT) {
+                    /* The linked-in data is already in the list. */
+                    return NULL;
+                }
+            }
+#endif
+        }
+
+        /* Add the linked-in data to the list. */
+        /*
+         * This is where we would check and call weakly linked partial-data-library
+         * access functions.
+         */
+        /*
+        if (uprv_getICUData_collation) {
+            setCommonICUDataPointer(uprv_getICUData_collation(), FALSE, pErrorCode);
+        }
+        if (uprv_getICUData_conversion) {
+            setCommonICUDataPointer(uprv_getICUData_conversion(), FALSE, pErrorCode);
+        }
+        */
+#if !defined(ICU_DATA_DIR_WINDOWS)
+// When using the Windows system data, we expect only a single data file.
+        setCommonICUDataPointer(&U_ICUDATA_ENTRY_POINT, FALSE, pErrorCode);
+        {
+            Mutex lock;
+            return gCommonICUDataArray[commonDataIndex];
+        }
+#endif
+    }
+
+
+    /* request is NOT for ICU Data.  */
+
+    /* Find the base name portion of the supplied path.   */
+    /*   inBasename will be left pointing somewhere within the original path string.      */
+    inBasename = findBasename(path);
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "inBasename = %s\n", inBasename);
+#endif
+
+    if(*inBasename==0) {
+        /* no basename.     This will happen if the original path was a directory name,   */
+        /*    like  "a/b/c/".   (Fallback to separate files will still work.)             */
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "ocd: no basename in %s, bailing.\n", path);
+#endif
+        if (U_SUCCESS(*pErrorCode)) {
+            *pErrorCode=U_FILE_ACCESS_ERROR;
+        }
+        return NULL;
+    }
+
+   /* Is the requested common data file already open and cached?                     */
+   /*   Note that the cache is keyed by the base name only.  The rest of the path,   */
+   /*     if any, is not considered.                                                 */
+    UDataMemory  *dataToReturn = udata_findCachedData(inBasename, *pErrorCode);
+    if (dataToReturn != NULL || U_FAILURE(*pErrorCode)) {
+        return dataToReturn;
+    }
+
+    /* Requested item is not in the cache.
+     * Hunt it down, trying all the path locations
+     */
+
+    UDataPathIterator iter(u_getDataDirectory(), inBasename, path, ".dat", TRUE, pErrorCode);
+
+    while ((UDataMemory_isLoaded(&tData)==FALSE) && (pathBuffer = iter.next(pErrorCode)) != NULL)
+    {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "ocd: trying path %s - ", pathBuffer);
+#endif
+        uprv_mapFile(&tData, pathBuffer, pErrorCode);
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "%s\n", UDataMemory_isLoaded(&tData)?"LOADED":"not loaded");
+#endif
+    }
+    if (U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+
+#if defined(OS390_STUBDATA) && defined(OS390BATCH)
+    if (!UDataMemory_isLoaded(&tData)) {
+        char ourPathBuffer[1024];
+        /* One more chance, for extendCommonData() */
+        uprv_strncpy(ourPathBuffer, path, 1019);
+        ourPathBuffer[1019]=0;
+        uprv_strcat(ourPathBuffer, ".dat");
+        uprv_mapFile(&tData, ourPathBuffer, pErrorCode);
+    }
+#endif
+
+    if (U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if (!UDataMemory_isLoaded(&tData)) {
+        /* no common data */
+        *pErrorCode=U_FILE_ACCESS_ERROR;
+        return NULL;
+    }
+
+    /* we have mapped a file, check its header */
+    udata_checkCommonData(&tData, pErrorCode);
+
+
+    /* Cache the UDataMemory struct for this .dat file,
+     *   so we won't need to hunt it down and map it again next time
+     *   something is needed from it.                */
+    return udata_cacheDataItem(inBasename, &tData, pErrorCode);
+}
+
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *   extendICUData   If the full set of ICU data was not loaded at      *
+ *                   program startup, load it now.  This function will  *
+ *                   be called when the lookup of an ICU data item in   *
+ *                   the common ICU data fails.                         *
+ *                                                                      *
+ *                   return true if new data is loaded, false otherwise.*
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+static UBool extendICUData(UErrorCode *pErr)
+{
+    UDataMemory   *pData;
+    UDataMemory   copyPData;
+    UBool         didUpdate = FALSE;
+
+    /*
+     * There is a chance for a race condition here.
+     * Normally, ICU data is loaded from a DLL or via mmap() and
+     * setCommonICUData() will detect if the same address is set twice.
+     * If ICU is built with data loading via fread() then the address will
+     * be different each time the common data is loaded and we may add
+     * multiple copies of the data.
+     * In this case, use a mutex to prevent the race.
+     * Use a specific mutex to avoid nested locks of the global mutex.
+     */
+#if MAP_IMPLEMENTATION==MAP_STDIO
+    static UMutex extendICUDataMutex;
+    umtx_lock(&extendICUDataMutex);
+#endif
+    if(!umtx_loadAcquire(gHaveTriedToLoadCommonData)) {
+        /* See if we can explicitly open a .dat file for the ICUData. */
+        pData = openCommonData(
+                   U_ICUDATA_NAME,            /*  "icudt20l" , for example.          */
+                   -1,                        /*  Pretend we're not opening ICUData  */
+                   pErr);
+
+        /* How about if there is no pData, eh... */
+
+       UDataMemory_init(&copyPData);
+       if(pData != NULL) {
+          UDatamemory_assign(&copyPData, pData);
+          copyPData.map = 0;              /* The mapping for this data is owned by the hash table */
+          copyPData.mapAddr = 0;          /*   which will unmap it when ICU is shut down.         */
+                                          /* CommonICUData is also unmapped when ICU is shut down.*/
+                                          /* To avoid unmapping the data twice, zero out the map  */
+                                          /*   fields in the UDataMemory that we're assigning     */
+                                          /*   to CommonICUData.                                  */
+
+          didUpdate = /* no longer using this result */
+              setCommonICUData(&copyPData,/*  The new common data.                                */
+                       FALSE,             /*  No warnings if write didn't happen                  */
+                       pErr);             /*  setCommonICUData honors errors; NOP if error set    */
+        }
+
+        umtx_storeRelease(gHaveTriedToLoadCommonData, 1);
+    }
+
+    didUpdate = findCommonICUDataByName(U_ICUDATA_NAME, *pErr);  /* Return 'true' when a racing writes out the extended                 */
+                                                          /* data after another thread has failed to see it (in openCommonData), so     */
+                                                          /* extended data can be examined.                                             */
+                                                          /* Also handles a race through here before gHaveTriedToLoadCommonData is set. */
+
+#if MAP_IMPLEMENTATION==MAP_STDIO
+    umtx_unlock(&extendICUDataMutex);
+#endif
+    return didUpdate;               /* Return true if ICUData pointer was updated.   */
+                                    /*   (Could potentially have been done by another thread racing */
+                                    /*   us through here, but that's fine, we still return true    */
+                                    /*   so that current thread will also examine extended data.   */
+}
+
+/*----------------------------------------------------------------------*
+ *                                                                      *
+ *   udata_setCommonData                                                *
+ *                                                                      *
+ *----------------------------------------------------------------------*/
+U_CAPI void U_EXPORT2
+udata_setCommonData(const void *data, UErrorCode *pErrorCode) {
+    UDataMemory dataMemory;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    if(data==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    /* set the data pointer and test for validity */
+    UDataMemory_init(&dataMemory);
+    UDataMemory_setData(&dataMemory, data);
+    udata_checkCommonData(&dataMemory, pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {return;}
+
+    /* we have good data */
+    /* Set it up as the ICU Common Data.  */
+    setCommonICUData(&dataMemory, TRUE, pErrorCode);
+}
+
+/*---------------------------------------------------------------------------
+ *
+ *  udata_setAppData
+ *
+ *---------------------------------------------------------------------------- */
+U_CAPI void U_EXPORT2
+udata_setAppData(const char *path, const void *data, UErrorCode *err)
+{
+    UDataMemory     udm;
+
+    if(err==NULL || U_FAILURE(*err)) {
+        return;
+    }
+    if(data==NULL) {
+        *err=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    UDataMemory_init(&udm);
+    UDataMemory_setData(&udm, data);
+    udata_checkCommonData(&udm, err);
+    udata_cacheDataItem(path, &udm, err);
+}
+
+/*----------------------------------------------------------------------------*
+ *                                                                            *
+ *  checkDataItem     Given a freshly located/loaded data item, either        *
+ *                    an entry in a common file or a separately loaded file,  *
+ *                    sanity check its header, and see if the data is         *
+ *                    acceptable to the app.                                  *
+ *                    If the data is good, create and return a UDataMemory    *
+ *                    object that can be returned to the application.         *
+ *                    Return NULL on any sort of failure.                     *
+ *                                                                            *
+ *----------------------------------------------------------------------------*/
+static UDataMemory *
+checkDataItem
+(
+ const DataHeader         *pHeader,         /* The data item to be checked.                */
+ UDataMemoryIsAcceptable  *isAcceptable,    /* App's call-back function                    */
+ void                     *context,         /*   pass-thru param for above.                */
+ const char               *type,            /*   pass-thru param for above.                */
+ const char               *name,            /*   pass-thru param for above.                */
+ UErrorCode               *nonFatalErr,     /* Error code if this data was not acceptable  */
+                                            /*   but openChoice should continue with       */
+                                            /*   trying to get data from fallback path.    */
+ UErrorCode               *fatalErr         /* Bad error, caller should return immediately */
+ )
+{
+    UDataMemory  *rDataMem = NULL;          /* the new UDataMemory, to be returned.        */
+
+    if (U_FAILURE(*fatalErr)) {
+        return NULL;
+    }
+
+    if(pHeader->dataHeader.magic1==0xda &&
+        pHeader->dataHeader.magic2==0x27 &&
+        (isAcceptable==NULL || isAcceptable(context, type, name, &pHeader->info))
+    ) {
+        rDataMem=UDataMemory_createNewInstance(fatalErr);
+        if (U_FAILURE(*fatalErr)) {
+            return NULL;
+        }
+        rDataMem->pHeader = pHeader;
+    } else {
+        /* the data is not acceptable, look further */
+        /* If we eventually find something good, this errorcode will be */
+        /*    cleared out.                                              */
+        *nonFatalErr=U_INVALID_FORMAT_ERROR;
+    }
+    return rDataMem;
+}
+
+/**
+ * @return 0 if not loaded, 1 if loaded or err 
+ */
+static UDataMemory *doLoadFromIndividualFiles(const char *pkgName, 
+        const char *dataPath, const char *tocEntryPathSuffix,
+            /* following arguments are the same as doOpenChoice itself */
+            const char *path, const char *type, const char *name,
+             UDataMemoryIsAcceptable *isAcceptable, void *context,
+             UErrorCode *subErrorCode,
+             UErrorCode *pErrorCode)
+{
+    const char         *pathBuffer;
+    UDataMemory         dataMemory;
+    UDataMemory *pEntryData;
+
+    /* look in ind. files: package\nam.typ  ========================= */
+    /* init path iterator for individual files */
+    UDataPathIterator iter(dataPath, pkgName, path, tocEntryPathSuffix, FALSE, pErrorCode);
+
+    while ((pathBuffer = iter.next(pErrorCode)) != NULL)
+    {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "UDATA: trying individual file %s\n", pathBuffer);
+#endif
+        if (uprv_mapFile(&dataMemory, pathBuffer, pErrorCode))
+        {
+            pEntryData = checkDataItem(dataMemory.pHeader, isAcceptable, context, type, name, subErrorCode, pErrorCode);
+            if (pEntryData != NULL) {
+                /* Data is good.
+                *  Hand off ownership of the backing memory to the user's UDataMemory.
+                *  and return it.   */
+                pEntryData->mapAddr = dataMemory.mapAddr;
+                pEntryData->map     = dataMemory.map;
+
+#ifdef UDATA_DEBUG
+                fprintf(stderr, "** Mapped file: %s\n", pathBuffer);
+#endif
+                return pEntryData;
+            }
+
+            /* the data is not acceptable, or some error occurred.  Either way, unmap the memory */
+            udata_close(&dataMemory);
+
+            /* If we had a nasty error, bail out completely.  */
+            if (U_FAILURE(*pErrorCode)) {
+                return NULL;
+            }
+
+            /* Otherwise remember that we found data but didn't like it for some reason  */
+            *subErrorCode=U_INVALID_FORMAT_ERROR;
+        }
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "%s\n", UDataMemory_isLoaded(&dataMemory)?"LOADED":"not loaded");
+#endif
+    }
+    return NULL;
+}
+
+/**
+ * @return 0 if not loaded, 1 if loaded or err 
+ */
+static UDataMemory *doLoadFromCommonData(UBool isICUData, const char * /*pkgName*/, 
+        const char * /*dataPath*/, const char * /*tocEntryPathSuffix*/, const char *tocEntryName,
+            /* following arguments are the same as doOpenChoice itself */
+            const char *path, const char *type, const char *name,
+             UDataMemoryIsAcceptable *isAcceptable, void *context,
+             UErrorCode *subErrorCode,
+             UErrorCode *pErrorCode)
+{
+    UDataMemory        *pEntryData;
+    const DataHeader   *pHeader;
+    UDataMemory        *pCommonData;
+    int32_t            commonDataIndex;
+    UBool              checkedExtendedICUData = FALSE;
+    /* try to get common data.  The loop is for platforms such as the 390 that do
+     *  not initially load the full set of ICU data.  If the lookup of an ICU data item
+     *  fails, the full (but slower to load) set is loaded, the and the loop repeats,
+     *  trying the lookup again.  Once the full set of ICU data is loaded, the loop wont
+     *  repeat because the full set will be checked the first time through.
+     *
+     *  The loop also handles the fallback to a .dat file if the application linked
+     *   to the stub data library rather than a real library.
+     */
+    for (commonDataIndex = isICUData ? 0 : -1;;) {
+        pCommonData=openCommonData(path, commonDataIndex, subErrorCode); /** search for pkg **/
+
+        if(U_SUCCESS(*subErrorCode) && pCommonData!=NULL) {
+            int32_t length;
+
+            /* look up the data piece in the common data */
+            pHeader=pCommonData->vFuncs->Lookup(pCommonData, tocEntryName, &length, subErrorCode);
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "%s: pHeader=%p - %s\n", tocEntryName, (void*) pHeader, u_errorName(*subErrorCode));
+#endif
+
+            if(pHeader!=NULL) {
+                pEntryData = checkDataItem(pHeader, isAcceptable, context, type, name, subErrorCode, pErrorCode);
+#ifdef UDATA_DEBUG
+                fprintf(stderr, "pEntryData=%p\n", (void*) pEntryData);
+#endif
+                if (U_FAILURE(*pErrorCode)) {
+                    return NULL;
+                }
+                if (pEntryData != NULL) {
+                    pEntryData->length = length;
+                    return pEntryData;
+                }
+            }
+        }
+        // If we failed due to being out-of-memory, then stop early and report the error.
+        if (*subErrorCode == U_MEMORY_ALLOCATION_ERROR) {
+            *pErrorCode = *subErrorCode;
+            return NULL;
+        }
+        /* Data wasn't found.  If we were looking for an ICUData item and there is
+         * more data available, load it and try again,
+         * otherwise break out of this loop. */
+        if (!isICUData) {
+            return NULL;
+        } else if (pCommonData != NULL) {
+            ++commonDataIndex;  /* try the next data package */
+        } else if ((!checkedExtendedICUData) && extendICUData(subErrorCode)) {
+            checkedExtendedICUData = TRUE;
+            /* try this data package slot again: it changed from NULL to non-NULL */
+        } else {
+            return NULL;
+        }
+    }
+}
+
+/*
+ * Identify the Time Zone resources that are subject to special override data loading.
+ */
+static UBool isTimeZoneFile(const char *name, const char *type) {
+    return ((uprv_strcmp(type, "res") == 0) &&
+            (uprv_strcmp(name, "zoneinfo64") == 0 ||
+             uprv_strcmp(name, "timezoneTypes") == 0 ||
+             uprv_strcmp(name, "windowsZones") == 0 ||
+             uprv_strcmp(name, "metaZones") == 0));
+}
+
+/*
+ *  A note on the ownership of Mapped Memory
+ *
+ *  For common format files, ownership resides with the UDataMemory object
+ *    that lives in the cache of opened common data.  These UDataMemorys are private
+ *    to the udata implementation, and are never seen directly by users.
+ *
+ *    The UDataMemory objects returned to users will have the address of some desired
+ *    data within the mapped region, but they wont have the mapping info itself, and thus
+ *    won't cause anything to be removed from memory when they are closed.
+ *
+ *  For individual data files, the UDataMemory returned to the user holds the
+ *  information necessary to unmap the data on close.  If the user independently
+ *  opens the same data file twice, two completely independent mappings will be made.
+ *  (There is no cache of opened data items from individual files, only a cache of
+ *   opened Common Data files, that is, files containing a collection of data items.)
+ *
+ *  For common data passed in from the user via udata_setAppData() or
+ *  udata_setCommonData(), ownership remains with the user.
+ *
+ *  UDataMemory objects themselves, as opposed to the memory they describe,
+ *  can be anywhere - heap, stack/local or global.
+ *  They have a flag to indicate when they're heap allocated and thus
+ *  must be deleted when closed.
+ */
+
+
+/*----------------------------------------------------------------------------*
+ *                                                                            *
+ * main data loading functions                                                *
+ *                                                                            *
+ *----------------------------------------------------------------------------*/
+static UDataMemory *
+doOpenChoice(const char *path, const char *type, const char *name,
+             UDataMemoryIsAcceptable *isAcceptable, void *context,
+             UErrorCode *pErrorCode)
+{
+    UDataMemory         *retVal = NULL;
+
+    const char         *dataPath;
+
+    int32_t             tocEntrySuffixIndex;
+    const char         *tocEntryPathSuffix;
+    UErrorCode          subErrorCode=U_ZERO_ERROR;
+    const char         *treeChar;
+
+    UBool               isICUData = FALSE;
+
+
+    FileTracer::traceOpen(path, type, name);
+
+
+    /* Is this path ICU data? */
+    if(path == NULL ||
+       !strcmp(path, U_ICUDATA_ALIAS) ||  /* "ICUDATA" */
+       !uprv_strncmp(path, U_ICUDATA_NAME U_TREE_SEPARATOR_STRING, /* "icudt26e-" */
+                     uprv_strlen(U_ICUDATA_NAME U_TREE_SEPARATOR_STRING)) ||  
+       !uprv_strncmp(path, U_ICUDATA_ALIAS U_TREE_SEPARATOR_STRING, /* "ICUDATA-" */
+                     uprv_strlen(U_ICUDATA_ALIAS U_TREE_SEPARATOR_STRING))) {
+      isICUData = TRUE;
+    }
+
+#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)  /* Windows:  try "foo\bar" and "foo/bar" */
+    /* remap from alternate path char to the main one */
+    CharString altSepPath;
+    if(path) {
+        if(uprv_strchr(path,U_FILE_ALT_SEP_CHAR) != NULL) {
+            altSepPath.append(path, *pErrorCode);
+            char *p;
+            while ((p = uprv_strchr(altSepPath.data(), U_FILE_ALT_SEP_CHAR)) != NULL) {
+                *p = U_FILE_SEP_CHAR;
+            }
+#if defined (UDATA_DEBUG)
+            fprintf(stderr, "Changed path from [%s] to [%s]\n", path, altSepPath.s);
+#endif
+            path = altSepPath.data();
+        }
+    }
+#endif
+
+    CharString tocEntryName; /* entry name in tree format. ex:  'icudt28b/coll/ar.res' */
+    CharString tocEntryPath; /* entry name in path format. ex:  'icudt28b\\coll\\ar.res' */
+
+    CharString pkgName;
+    CharString treeName;
+
+    /* ======= Set up strings */
+    if(path==NULL) {
+        pkgName.append(U_ICUDATA_NAME, *pErrorCode);
+    } else {
+        const char *pkg;
+        const char *first;
+        pkg = uprv_strrchr(path, U_FILE_SEP_CHAR);
+        first = uprv_strchr(path, U_FILE_SEP_CHAR);
+        if(uprv_pathIsAbsolute(path) || (pkg != first)) { /* more than one slash in the path- not a tree name */
+            /* see if this is an /absolute/path/to/package  path */
+            if(pkg) {
+                pkgName.append(pkg+1, *pErrorCode);
+            } else {
+                pkgName.append(path, *pErrorCode);
+            }
+        } else {
+            treeChar = uprv_strchr(path, U_TREE_SEPARATOR);
+            if(treeChar) { 
+                treeName.append(treeChar+1, *pErrorCode); /* following '-' */
+                if(isICUData) {
+                    pkgName.append(U_ICUDATA_NAME, *pErrorCode);
+                } else {
+                    pkgName.append(path, (int32_t)(treeChar-path), *pErrorCode);
+                    if (first == NULL) {
+                        /*
+                        This user data has no path, but there is a tree name.
+                        Look up the correct path from the data cache later.
+                        */
+                        path = pkgName.data();
+                    }
+                }
+            } else {
+                if(isICUData) {
+                    pkgName.append(U_ICUDATA_NAME, *pErrorCode);
+                } else {
+                    pkgName.append(path, *pErrorCode);
+                }
+            }
+        }
+    }
+
+#ifdef UDATA_DEBUG
+    fprintf(stderr, " P=%s T=%s\n", pkgName.data(), treeName.data());
+#endif
+
+    /* setting up the entry name and file name 
+     * Make up a full name by appending the type to the supplied
+     *  name, assuming that a type was supplied.
+     */
+
+    /* prepend the package */
+    tocEntryName.append(pkgName, *pErrorCode);
+    tocEntryPath.append(pkgName, *pErrorCode);
+    tocEntrySuffixIndex = tocEntryName.length();
+
+    if(!treeName.isEmpty()) {
+        tocEntryName.append(U_TREE_ENTRY_SEP_CHAR, *pErrorCode).append(treeName, *pErrorCode);
+        tocEntryPath.append(U_FILE_SEP_CHAR, *pErrorCode).append(treeName, *pErrorCode);
+    }
+
+    tocEntryName.append(U_TREE_ENTRY_SEP_CHAR, *pErrorCode).append(name, *pErrorCode);
+    tocEntryPath.append(U_FILE_SEP_CHAR, *pErrorCode).append(name, *pErrorCode);
+    if(type!=NULL && *type!=0) {
+        tocEntryName.append(".", *pErrorCode).append(type, *pErrorCode);
+        tocEntryPath.append(".", *pErrorCode).append(type, *pErrorCode);
+    }
+    // The +1 is for the U_FILE_SEP_CHAR that is always appended above.
+    tocEntryPathSuffix = tocEntryPath.data() + tocEntrySuffixIndex + 1; /* suffix starts here */
+
+#ifdef UDATA_DEBUG
+    fprintf(stderr, " tocEntryName = %s\n", tocEntryName.data());
+    fprintf(stderr, " tocEntryPath = %s\n", tocEntryName.data());
+#endif
+
+#if !defined(ICU_DATA_DIR_WINDOWS)
+    if(path == NULL) {
+        path = COMMON_DATA_NAME; /* "icudt26e" */
+    }
+#else
+    // When using the Windows system data, we expects only a single data file.
+    path = COMMON_DATA_NAME; /* "icudt26e" */
+#endif
+
+    /************************ Begin loop looking for ind. files ***************/
+#ifdef UDATA_DEBUG
+    fprintf(stderr, "IND: inBasename = %s, pkg=%s\n", "(n/a)", packageNameFromPath(path));
+#endif
+
+    /* End of dealing with a null basename */
+    dataPath = u_getDataDirectory();
+
+    /****    Time zone individual files override  */
+    if (isICUData && isTimeZoneFile(name, type)) {
+        const char *tzFilesDir = u_getTimeZoneFilesDirectory(pErrorCode);
+        if (tzFilesDir[0] != 0) {
+#ifdef UDATA_DEBUG
+            fprintf(stderr, "Trying Time Zone Files directory = %s\n", tzFilesDir);
+#endif
+            retVal = doLoadFromIndividualFiles(/* pkgName.data() */ "", tzFilesDir, tocEntryPathSuffix,
+                            /* path */ "", type, name, isAcceptable, context, &subErrorCode, pErrorCode);
+            if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
+                return retVal;
+            }
+        }
+    }
+
+    /****    COMMON PACKAGE  - only if packages are first. */
+    if(gDataFileAccess == UDATA_PACKAGES_FIRST) {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "Trying packages (UDATA_PACKAGES_FIRST)\n");
+#endif
+        /* #2 */
+        retVal = doLoadFromCommonData(isICUData, 
+                            pkgName.data(), dataPath, tocEntryPathSuffix, tocEntryName.data(),
+                            path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
+        if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
+            return retVal;
+        }
+    }
+
+    /****    INDIVIDUAL FILES  */
+    if((gDataFileAccess==UDATA_PACKAGES_FIRST) ||
+       (gDataFileAccess==UDATA_FILES_FIRST)) {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "Trying individual files\n");
+#endif
+        /* Check to make sure that there is a dataPath to iterate over */
+        if ((dataPath && *dataPath) || !isICUData) {
+            retVal = doLoadFromIndividualFiles(pkgName.data(), dataPath, tocEntryPathSuffix,
+                            path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
+            if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
+                return retVal;
+            }
+        }
+    }
+
+    /****    COMMON PACKAGE  */
+    if((gDataFileAccess==UDATA_ONLY_PACKAGES) || 
+       (gDataFileAccess==UDATA_FILES_FIRST)) {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "Trying packages (UDATA_ONLY_PACKAGES || UDATA_FILES_FIRST)\n");
+#endif
+        retVal = doLoadFromCommonData(isICUData,
+                            pkgName.data(), dataPath, tocEntryPathSuffix, tocEntryName.data(),
+                            path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
+        if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
+            return retVal;
+        }
+    }
+    
+    /* Load from DLL.  If we haven't attempted package load, we also haven't had any chance to
+        try a DLL (static or setCommonData/etc)  load.
+         If we ever have a "UDATA_ONLY_FILES", add it to the or list here.  */  
+    if(gDataFileAccess==UDATA_NO_FILES) {
+#ifdef UDATA_DEBUG
+        fprintf(stderr, "Trying common data (UDATA_NO_FILES)\n");
+#endif
+        retVal = doLoadFromCommonData(isICUData,
+                            pkgName.data(), "", tocEntryPathSuffix, tocEntryName.data(),
+                            path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
+        if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
+            return retVal;
+        }
+    }
+
+    /* data not found */
+    if(U_SUCCESS(*pErrorCode)) {
+        if(U_SUCCESS(subErrorCode)) {
+            /* file not found */
+            *pErrorCode=U_FILE_ACCESS_ERROR;
+        } else {
+            /* entry point not found or rejected */
+            *pErrorCode=subErrorCode;
+        }
+    }
+    return retVal;
+}
+
+
+
+/* API ---------------------------------------------------------------------- */
+
+U_CAPI UDataMemory * U_EXPORT2
+udata_open(const char *path, const char *type, const char *name,
+           UErrorCode *pErrorCode) {
+#ifdef UDATA_DEBUG
+  fprintf(stderr, "udata_open(): Opening: %s : %s . %s\n", (path?path:"NULL"), name, type);
+    fflush(stderr);
+#endif
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    } else if(name==NULL || *name==0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    } else {
+        return doOpenChoice(path, type, name, NULL, NULL, pErrorCode);
+    }
+}
+
+
+
+U_CAPI UDataMemory * U_EXPORT2
+udata_openChoice(const char *path, const char *type, const char *name,
+                 UDataMemoryIsAcceptable *isAcceptable, void *context,
+                 UErrorCode *pErrorCode) {
+#ifdef UDATA_DEBUG
+  fprintf(stderr, "udata_openChoice(): Opening: %s : %s . %s\n", (path?path:"NULL"), name, type);
+#endif
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    } else if(name==NULL || *name==0 || isAcceptable==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    } else {
+        return doOpenChoice(path, type, name, isAcceptable, context, pErrorCode);
+    }
+}
+
+
+
+U_CAPI void U_EXPORT2
+udata_getInfo(UDataMemory *pData, UDataInfo *pInfo) {
+    if(pInfo!=NULL) {
+        if(pData!=NULL && pData->pHeader!=NULL) {
+            const UDataInfo *info=&pData->pHeader->info;
+            uint16_t dataInfoSize=udata_getInfoSize(info);
+            if(pInfo->size>dataInfoSize) {
+                pInfo->size=dataInfoSize;
+            }
+            uprv_memcpy((uint16_t *)pInfo+1, (const uint16_t *)info+1, pInfo->size-2);
+            if(info->isBigEndian!=U_IS_BIG_ENDIAN) {
+                /* opposite endianness */
+                uint16_t x=info->reservedWord;
+                pInfo->reservedWord=(uint16_t)((x<<8)|(x>>8));
+            }
+        } else {
+            pInfo->size=0;
+        }
+    }
+}
+
+
+U_CAPI void U_EXPORT2 udata_setFileAccess(UDataFileAccess access, UErrorCode * /*status*/)
+{
+    // Note: this function is documented as not thread safe.
+    gDataFileAccess = access;
+}
diff --git a/thirdparty/icu4c/common/udatamem.cpp b/thirdparty/icu4c/common/udatamem.cpp
new file mode 100644
index 0000000000..6bf7c01235
--- /dev/null
+++ b/thirdparty/icu4c/common/udatamem.cpp
@@ -0,0 +1,161 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+
+/*----------------------------------------------------------------------------------
+ *
+ *  UDataMemory     A class-like struct that serves as a handle to a piece of memory
+ *                  that contains some ICU data (resource, converters, whatever.)
+ *
+ *                  When an application opens ICU data (with udata_open, for example,
+ *                  a UDataMemory * is returned.
+ *
+ *----------------------------------------------------------------------------------*/
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "unicode/udata.h"
+
+#include "udatamem.h"
+
+U_CFUNC void UDataMemory_init(UDataMemory *This) {
+    uprv_memset(This, 0, sizeof(UDataMemory));
+    This->length=-1;
+}
+
+
+U_CFUNC void UDatamemory_assign(UDataMemory *dest, UDataMemory *source) {
+    /* UDataMemory Assignment.  Destination UDataMemory must be initialized first.  */
+    UBool mallocedFlag = dest->heapAllocated;
+    uprv_memcpy(dest, source, sizeof(UDataMemory));
+    dest->heapAllocated = mallocedFlag;
+}
+
+U_CFUNC UDataMemory *UDataMemory_createNewInstance(UErrorCode *pErr) {
+    UDataMemory *This;
+
+    if (U_FAILURE(*pErr)) {
+        return NULL;
+    }
+    This = (UDataMemory *)uprv_malloc(sizeof(UDataMemory));
+    if (This == NULL) {
+        *pErr = U_MEMORY_ALLOCATION_ERROR; }
+    else {
+        UDataMemory_init(This);
+        This->heapAllocated = TRUE;
+    }
+    return This;
+}
+
+
+U_CFUNC const DataHeader *
+UDataMemory_normalizeDataPointer(const void *p) {
+    /* allow the data to be optionally prepended with an alignment-forcing double value */
+    const DataHeader *pdh = (const DataHeader *)p;
+    if(pdh==NULL || (pdh->dataHeader.magic1==0xda && pdh->dataHeader.magic2==0x27)) {
+        return pdh;
+    } else {
+#if U_PLATFORM == U_PF_OS400
+        /*
+        TODO: Fix this once the compiler implements this feature. Keep in sync with genccode.c
+
+        This is here because this platform can't currently put
+        const data into the read-only pages of an object or
+        shared library (service program). Only strings are allowed in read-only
+        pages, so we use char * strings to store the data.
+
+        In order to prevent the beginning of the data from ever matching the
+        magic numbers we must skip the initial double.
+        [grhoten 4/24/2003]
+        */
+        return (const DataHeader *)*((const void **)p+1);
+#else
+        return (const DataHeader *)((const double *)p+1);
+#endif
+    }
+}
+
+
+U_CFUNC void UDataMemory_setData (UDataMemory *This, const void *dataAddr) {
+    This->pHeader = UDataMemory_normalizeDataPointer(dataAddr);
+}
+
+
+U_CAPI void U_EXPORT2
+udata_close(UDataMemory *pData) {
+    if(pData!=NULL) {
+        uprv_unmapFile(pData);
+        if(pData->heapAllocated ) {
+            uprv_free(pData);
+        } else {
+            UDataMemory_init(pData);
+        }
+    }
+}
+
+U_CAPI const void * U_EXPORT2
+udata_getMemory(UDataMemory *pData) {
+    if(pData!=NULL && pData->pHeader!=NULL) {
+        return (char *)(pData->pHeader)+udata_getHeaderSize(pData->pHeader);
+    } else {
+        return NULL;
+    }
+}
+
+/**
+ * Get the length of the data item if possible.
+ * The length may be up to 15 bytes larger than the actual data.
+ *
+ * TODO Consider making this function public.
+ * It would have to return the actual length in more cases.
+ * For example, the length of the last item in a .dat package could be
+ * computed from the size of the whole .dat package minus the offset of the
+ * last item.
+ * The size of a file that was directly memory-mapped could be determined
+ * using some system API.
+ *
+ * In order to get perfect values for all data items, we may have to add a
+ * length field to UDataInfo, but that complicates data generation
+ * and may be overkill.
+ *
+ * @param pData The data item.
+ * @return the length of the data item, or -1 if not known
+ * @internal Currently used only in cintltst/udatatst.c
+ */
+U_CAPI int32_t U_EXPORT2
+udata_getLength(const UDataMemory *pData) {
+    if(pData!=NULL && pData->pHeader!=NULL && pData->length>=0) {
+        /*
+         * subtract the header size,
+         * return only the size of the actual data starting at udata_getMemory()
+         */
+        return pData->length-udata_getHeaderSize(pData->pHeader);
+    } else {
+        return -1;
+    }
+}
+
+/**
+ * Get the memory including the data header.
+ * Used in cintltst/udatatst.c
+ * @internal
+ */
+U_CAPI const void * U_EXPORT2
+udata_getRawMemory(const UDataMemory *pData) {
+    if(pData!=NULL && pData->pHeader!=NULL) {
+        return pData->pHeader;
+    } else {
+        return NULL;
+    }
+}
+
+U_CFUNC UBool UDataMemory_isLoaded(const UDataMemory *This) {
+    return This->pHeader != NULL;
+}
diff --git a/thirdparty/icu4c/common/udatamem.h b/thirdparty/icu4c/common/udatamem.h
new file mode 100644
index 0000000000..a05dd69756
--- /dev/null
+++ b/thirdparty/icu4c/common/udatamem.h
@@ -0,0 +1,61 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2010, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+
+/*----------------------------------------------------------------------------------
+ *
+ *  UDataMemory     A class-like struct that serves as a handle to a piece of memory
+ *                  that contains some ICU data (resource, converters, whatever.)
+ *
+ *                  When an application opens ICU data (with udata_open, for example,
+ *                  a UDataMemory * is returned.
+ *
+ *----------------------------------------------------------------------------------*/
+#ifndef __UDATAMEM_H__
+#define __UDATAMEM_H__
+
+#include "unicode/udata.h"
+#include "ucmndata.h"
+
+struct UDataMemory {
+    const commonDataFuncs  *vFuncs;      /* Function Pointers for accessing TOC             */
+
+    const DataHeader *pHeader;     /* Header of the memory being described by this    */
+                                   /*   UDataMemory object.                           */
+    const void       *toc;         /* For common memory, table of contents for        */
+                                   /*   the pieces within.                            */
+    UBool             heapAllocated;  /* True if this UDataMemory Object is on the    */
+                                   /*  heap and thus needs to be deleted when closed. */
+
+    void             *mapAddr;     /* For mapped or allocated memory, the start addr. */
+                                   /* Only non-null if a close operation should unmap */
+                                   /*  the associated data.                           */
+    void             *map;         /* Handle, or other data, OS dependent.            */
+                                   /* Only non-null if a close operation should unmap */
+                                   /*  the associated data, and additional info       */
+                                   /*   beyond the mapAddr is needed to do that.      */
+    int32_t           length;      /* Length of the data in bytes; -1 if unknown.     */
+};
+
+U_CFUNC UDataMemory *UDataMemory_createNewInstance(UErrorCode *pErr);
+U_CFUNC void         UDatamemory_assign  (UDataMemory *dest, UDataMemory *source);
+U_CFUNC void         UDataMemory_init    (UDataMemory *This);
+U_CFUNC UBool        UDataMemory_isLoaded(const UDataMemory *This);
+U_CFUNC void         UDataMemory_setData (UDataMemory *This, const void *dataAddr);
+
+U_CFUNC const DataHeader *UDataMemory_normalizeDataPointer(const void *p);
+
+U_CAPI int32_t U_EXPORT2
+udata_getLength(const UDataMemory *pData);
+
+U_CAPI const void * U_EXPORT2
+udata_getRawMemory(const UDataMemory *pData);
+
+#endif
diff --git a/thirdparty/icu4c/common/udataswp.cpp b/thirdparty/icu4c/common/udataswp.cpp
new file mode 100644
index 0000000000..86f302bd9c
--- /dev/null
+++ b/thirdparty/icu4c/common/udataswp.cpp
@@ -0,0 +1,473 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  udataswp.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003jun05
+*   created by: Markus W. Scherer
+*
+*   Definitions for ICU data transformations for different platforms,
+*   changing between big- and little-endian data and/or between
+*   charset families (ASCII<->EBCDIC).
+*/
+
+#include <stdarg.h>
+#include "unicode/utypes.h"
+#include "unicode/udata.h" /* UDataInfo */
+#include "ucmndata.h" /* DataHeader */
+#include "cmemory.h"
+#include "udataswp.h"
+
+/* swapping primitives ------------------------------------------------------ */
+
+static int32_t U_CALLCONV
+uprv_swapArray16(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    const uint16_t *p;
+    uint16_t *q;
+    int32_t count;
+    uint16_t x;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&1)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    p=(const uint16_t *)inData;
+    q=(uint16_t *)outData;
+    count=length/2;
+    while(count>0) {
+        x=*p++;
+        *q++=(uint16_t)((x<<8)|(x>>8));
+        --count;
+    }
+
+    return length;
+}
+
+static int32_t U_CALLCONV
+uprv_copyArray16(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&1)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(length>0 && inData!=outData) {
+        uprv_memcpy(outData, inData, length);
+    }
+    return length;
+}
+
+static int32_t U_CALLCONV
+uprv_swapArray32(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    const uint32_t *p;
+    uint32_t *q;
+    int32_t count;
+    uint32_t x;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&3)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    p=(const uint32_t *)inData;
+    q=(uint32_t *)outData;
+    count=length/4;
+    while(count>0) {
+        x=*p++;
+        *q++=(uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
+        --count;
+    }
+
+    return length;
+}
+
+static int32_t U_CALLCONV
+uprv_copyArray32(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&3)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(length>0 && inData!=outData) {
+        uprv_memcpy(outData, inData, length);
+    }
+    return length;
+}
+
+static int32_t U_CALLCONV
+uprv_swapArray64(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    const uint64_t *p;
+    uint64_t *q;
+    int32_t count;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&7)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    p=(const uint64_t *)inData;
+    q=(uint64_t *)outData;
+    count=length/8;
+    while(count>0) {
+        uint64_t x=*p++;
+        x=(x<<56)|((x&0xff00)<<40)|((x&0xff0000)<<24)|((x&0xff000000)<<8)|
+            ((x>>8)&0xff000000)|((x>>24)&0xff0000)|((x>>40)&0xff00)|(x>>56);
+        *q++=x;
+        --count;
+    }
+
+    return length;
+}
+
+static int32_t U_CALLCONV
+uprv_copyArray64(const UDataSwapper *ds,
+                 const void *inData, int32_t length, void *outData,
+                 UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length&7)!=0 || outData==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(length>0 && inData!=outData) {
+        uprv_memcpy(outData, inData, length);
+    }
+    return length;
+}
+
+static uint16_t U_CALLCONV
+uprv_readSwapUInt16(uint16_t x) {
+    return (uint16_t)((x<<8)|(x>>8));
+}
+
+static uint16_t U_CALLCONV
+uprv_readDirectUInt16(uint16_t x) {
+    return x;
+}
+
+static uint32_t U_CALLCONV
+uprv_readSwapUInt32(uint32_t x) {
+    return (uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
+}
+
+static uint32_t U_CALLCONV
+uprv_readDirectUInt32(uint32_t x) {
+    return x;
+}
+
+static void U_CALLCONV
+uprv_writeSwapUInt16(uint16_t *p, uint16_t x) {
+    *p=(uint16_t)((x<<8)|(x>>8));
+}
+
+static void U_CALLCONV
+uprv_writeDirectUInt16(uint16_t *p, uint16_t x) {
+    *p=x;
+}
+
+static void U_CALLCONV
+uprv_writeSwapUInt32(uint32_t *p, uint32_t x) {
+    *p=(uint32_t)((x<<24)|((x<<8)&0xff0000)|((x>>8)&0xff00)|(x>>24));
+}
+
+static void U_CALLCONV
+uprv_writeDirectUInt32(uint32_t *p, uint32_t x) {
+    *p=x;
+}
+
+U_CAPI int16_t U_EXPORT2
+udata_readInt16(const UDataSwapper *ds, int16_t x) {
+    return (int16_t)ds->readUInt16((uint16_t)x);
+}
+
+U_CAPI int32_t U_EXPORT2
+udata_readInt32(const UDataSwapper *ds, int32_t x) {
+    return (int32_t)ds->readUInt32((uint32_t)x);
+}
+
+/**
+ * Swap a block of invariant, NUL-terminated strings, but not padding
+ * bytes after the last string.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+udata_swapInvStringBlock(const UDataSwapper *ds,
+                         const void *inData, int32_t length, void *outData,
+                         UErrorCode *pErrorCode) {
+    const char *inChars;
+    int32_t stringsLength;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* reduce the strings length to not include bytes after the last NUL */
+    inChars=(const char *)inData;
+    stringsLength=length;
+    while(stringsLength>0 && inChars[stringsLength-1]!=0) {
+        --stringsLength;
+    }
+
+    /* swap up to the last NUL */
+    ds->swapInvChars(ds, inData, stringsLength, outData, pErrorCode);
+
+    /* copy the bytes after the last NUL */
+    if(inData!=outData && length>stringsLength) {
+        uprv_memcpy((char *)outData+stringsLength, inChars+stringsLength, length-stringsLength);
+    }
+
+    /* return the length including padding bytes */
+    if(U_SUCCESS(*pErrorCode)) {
+        return length;
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI void U_EXPORT2
+udata_printError(const UDataSwapper *ds,
+                 const char *fmt,
+                 ...) {
+    va_list args;
+
+    if(ds->printError!=NULL) {
+        va_start(args, fmt);
+        ds->printError(ds->printErrorContext, fmt, args);
+        va_end(args);
+    }
+}
+
+/* swap a data header ------------------------------------------------------- */
+
+U_CAPI int32_t U_EXPORT2
+udata_swapDataHeader(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode) {
+    const DataHeader *pHeader;
+    uint16_t headerSize, infoSize;
+
+    /* argument checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<-1 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* check minimum length and magic bytes */
+    pHeader=(const DataHeader *)inData;
+    if( (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
+        pHeader->dataHeader.magic1!=0xda ||
+        pHeader->dataHeader.magic2!=0x27 ||
+        pHeader->info.sizeofUChar!=2
+    ) {
+        udata_printError(ds, "udata_swapDataHeader(): initial bytes do not look like ICU data\n");
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    headerSize=ds->readUInt16(pHeader->dataHeader.headerSize);
+    infoSize=ds->readUInt16(pHeader->info.size);
+
+    if( headerSize<sizeof(DataHeader) ||
+        infoSize<sizeof(UDataInfo) ||
+        headerSize<(sizeof(pHeader->dataHeader)+infoSize) ||
+        (length>=0 && length<headerSize)
+    ) {
+        udata_printError(ds, "udata_swapDataHeader(): header size mismatch - headerSize %d infoSize %d length %d\n",
+                         headerSize, infoSize, length);
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    if(length>0) {
+        DataHeader *outHeader;
+        const char *s;
+        int32_t maxLength;
+
+        /* Most of the fields are just bytes and need no swapping. */
+        if(inData!=outData) {
+            uprv_memcpy(outData, inData, headerSize);
+        }
+        outHeader=(DataHeader *)outData;
+
+        outHeader->info.isBigEndian = ds->outIsBigEndian;
+        outHeader->info.charsetFamily = ds->outCharset;
+
+        /* swap headerSize */
+        ds->swapArray16(ds, &pHeader->dataHeader.headerSize, 2, &outHeader->dataHeader.headerSize, pErrorCode);
+
+        /* swap UDataInfo size and reservedWord */
+        ds->swapArray16(ds, &pHeader->info.size, 4, &outHeader->info.size, pErrorCode);
+
+        /* swap copyright statement after the UDataInfo */
+        infoSize+=sizeof(pHeader->dataHeader);
+        s=(const char *)inData+infoSize;
+        maxLength=headerSize-infoSize;
+        /* get the length of the string */
+        for(length=0; length<maxLength && s[length]!=0; ++length) {}
+        /* swap the string contents */
+        ds->swapInvChars(ds, s, length, (char *)outData+infoSize, pErrorCode);
+    }
+
+    return headerSize;
+}
+
+/* API functions ------------------------------------------------------------ */
+
+U_CAPI UDataSwapper * U_EXPORT2
+udata_openSwapper(UBool inIsBigEndian, uint8_t inCharset,
+                  UBool outIsBigEndian, uint8_t outCharset,
+                  UErrorCode *pErrorCode) {
+    UDataSwapper *swapper;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(inCharset>U_EBCDIC_FAMILY || outCharset>U_EBCDIC_FAMILY) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    /* allocate the swapper */
+    swapper=(UDataSwapper *)uprv_malloc(sizeof(UDataSwapper));
+    if(swapper==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memset(swapper, 0, sizeof(UDataSwapper));
+
+    /* set values and functions pointers according to in/out parameters */
+    swapper->inIsBigEndian=inIsBigEndian;
+    swapper->inCharset=inCharset;
+    swapper->outIsBigEndian=outIsBigEndian;
+    swapper->outCharset=outCharset;
+
+    swapper->readUInt16= inIsBigEndian==U_IS_BIG_ENDIAN ? uprv_readDirectUInt16 : uprv_readSwapUInt16;
+    swapper->readUInt32= inIsBigEndian==U_IS_BIG_ENDIAN ? uprv_readDirectUInt32 : uprv_readSwapUInt32;
+
+    swapper->writeUInt16= outIsBigEndian==U_IS_BIG_ENDIAN ? uprv_writeDirectUInt16 : uprv_writeSwapUInt16;
+    swapper->writeUInt32= outIsBigEndian==U_IS_BIG_ENDIAN ? uprv_writeDirectUInt32 : uprv_writeSwapUInt32;
+
+    swapper->compareInvChars= outCharset==U_ASCII_FAMILY ? uprv_compareInvAscii : uprv_compareInvEbcdic;
+
+    if(inIsBigEndian==outIsBigEndian) {
+        swapper->swapArray16=uprv_copyArray16;
+        swapper->swapArray32=uprv_copyArray32;
+        swapper->swapArray64=uprv_copyArray64;
+    } else {
+        swapper->swapArray16=uprv_swapArray16;
+        swapper->swapArray32=uprv_swapArray32;
+        swapper->swapArray64=uprv_swapArray64;
+    }
+
+    if(inCharset==U_ASCII_FAMILY) {
+        swapper->swapInvChars= outCharset==U_ASCII_FAMILY ? uprv_copyAscii : uprv_ebcdicFromAscii;
+    } else /* U_EBCDIC_FAMILY */ {
+        swapper->swapInvChars= outCharset==U_EBCDIC_FAMILY ? uprv_copyEbcdic : uprv_asciiFromEbcdic;
+    }
+
+    return swapper;
+}
+
+U_CAPI UDataSwapper * U_EXPORT2
+udata_openSwapperForInputData(const void *data, int32_t length,
+                              UBool outIsBigEndian, uint8_t outCharset,
+                              UErrorCode *pErrorCode) {
+    const DataHeader *pHeader;
+    uint16_t headerSize, infoSize;
+    UBool inIsBigEndian;
+    int8_t inCharset;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if( data==NULL ||
+        (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
+        outCharset>U_EBCDIC_FAMILY
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    pHeader=(const DataHeader *)data;
+    if( (length>=0 && length<(int32_t)sizeof(DataHeader)) ||
+        pHeader->dataHeader.magic1!=0xda ||
+        pHeader->dataHeader.magic2!=0x27 ||
+        pHeader->info.sizeofUChar!=2
+    ) {
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inIsBigEndian=(UBool)pHeader->info.isBigEndian;
+    inCharset=pHeader->info.charsetFamily;
+
+    if(inIsBigEndian==U_IS_BIG_ENDIAN) {
+        headerSize=pHeader->dataHeader.headerSize;
+        infoSize=pHeader->info.size;
+    } else {
+        headerSize=uprv_readSwapUInt16(pHeader->dataHeader.headerSize);
+        infoSize=uprv_readSwapUInt16(pHeader->info.size);
+    }
+
+    if( headerSize<sizeof(DataHeader) ||
+        infoSize<sizeof(UDataInfo) ||
+        headerSize<(sizeof(pHeader->dataHeader)+infoSize) ||
+        (length>=0 && length<headerSize)
+    ) {
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    return udata_openSwapper(inIsBigEndian, inCharset, outIsBigEndian, outCharset, pErrorCode);
+}
+
+U_CAPI void U_EXPORT2
+udata_closeSwapper(UDataSwapper *ds) {
+    uprv_free(ds);
+}
diff --git a/thirdparty/icu4c/common/udataswp.h b/thirdparty/icu4c/common/udataswp.h
new file mode 100644
index 0000000000..5e7b043c4c
--- /dev/null
+++ b/thirdparty/icu4c/common/udataswp.h
@@ -0,0 +1,404 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  udataswp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003jun05
+*   created by: Markus W. Scherer
+*
+*   Definitions for ICU data transformations for different platforms,
+*   changing between big- and little-endian data and/or between
+*   charset families (ASCII<->EBCDIC).
+*/
+
+#ifndef __UDATASWP_H__
+#define __UDATASWP_H__
+
+#include <stdarg.h>
+#include "unicode/utypes.h"
+
+/* forward declaration */
+
+U_CDECL_BEGIN
+
+struct UDataSwapper;
+typedef struct UDataSwapper UDataSwapper;
+
+/**
+ * Function type for data transformation.
+ * Transforms data, or just returns the length of the data if
+ * the input length is -1.
+ * Swap functions assume that their data pointers are aligned properly.
+ *
+ * Quick implementation outline:
+ * (best to copy and adapt and existing swapper implementation)
+ * check that the data looks like the expected format
+ * if(length<0) {
+ *   preflight:
+ *   never dereference outData
+ *   read inData and determine the data size
+ *   assume that inData is long enough for this
+ * } else {
+ *   outData can be NULL if length==0
+ *   inData==outData (in-place swapping) possible but not required!
+ *   verify that length>=(actual size)
+ *   if there is a chance that not every byte up to size is reached
+ *     due to padding etc.:
+ *   if(inData!=outData) {
+ *     memcpy(outData, inData, actual size);
+ *   }
+ *   swap contents
+ * }
+ * return actual size
+ *
+ * Further implementation notes:
+ * - read integers from inData before swapping them
+ *   because in-place swapping can make them unreadable
+ * - compareInvChars compares a local Unicode string with already-swapped
+ *   output charset strings
+ *
+ * @param ds Pointer to UDataSwapper containing global data about the
+ *           transformation and function pointers for handling primitive
+ *           types.
+ * @param inData Pointer to the input data to be transformed or examined.
+ * @param length Length of the data, counting bytes. May be -1 for preflighting.
+ *               If length>=0, then transform the data.
+ *               If length==-1, then only determine the length of the data.
+ *               The length cannot be determined from the data itself for all
+ *               types of data (e.g., not for simple arrays of integers).
+ * @param outData Pointer to the output data buffer.
+ *                If length>=0 (transformation), then the output buffer must
+ *                have a capacity of at least length.
+ *                If length==-1, then outData will not be used and can be NULL.
+ * @param pErrorCode ICU UErrorCode parameter, must not be NULL and must
+ *                   fulfill U_SUCCESS on input.
+ * @return The actual length of the data.
+ *
+ * @see UDataSwapper
+ * @internal ICU 2.8
+ */
+typedef int32_t U_CALLCONV
+UDataSwapFn(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode);
+
+/**
+ * Convert one uint16_t from input to platform endianness.
+ * @internal ICU 2.8
+ */
+typedef uint16_t U_CALLCONV
+UDataReadUInt16(uint16_t x);
+
+/**
+ * Convert one uint32_t from input to platform endianness.
+ * @internal ICU 2.8
+ */
+typedef uint32_t U_CALLCONV
+UDataReadUInt32(uint32_t x);
+
+/**
+ * Convert one uint16_t from platform to input endianness.
+ * @internal ICU 2.8
+ */
+typedef void U_CALLCONV
+UDataWriteUInt16(uint16_t *p, uint16_t x);
+
+/**
+ * Convert one uint32_t from platform to input endianness.
+ * @internal ICU 2.8
+ */
+typedef void U_CALLCONV
+UDataWriteUInt32(uint32_t *p, uint32_t x);
+
+/**
+ * Compare invariant-character strings, one in the output data and the
+ * other one caller-provided in Unicode.
+ * An output data string is compared because strings are usually swapped
+ * before the rest of the data, to allow for sorting of string tables
+ * according to the output charset.
+ * You can use -1 for the length parameters of NUL-terminated strings as usual.
+ * Returns Unicode code point order for invariant characters.
+ * @internal ICU 2.8
+ */
+typedef int32_t U_CALLCONV
+UDataCompareInvChars(const UDataSwapper *ds,
+                     const char *outString, int32_t outLength,
+                     const UChar *localString, int32_t localLength);
+
+/**
+ * Function for message output when an error occurs during data swapping.
+ * A format string and variable number of arguments are passed
+ * like for vprintf().
+ *
+ * @param context A function-specific context pointer.
+ * @param fmt The format string.
+ * @param args The arguments for format string inserts.
+ *
+ * @internal ICU 2.8
+ */
+typedef void U_CALLCONV
+UDataPrintError(void *context, const char *fmt, va_list args);
+
+struct UDataSwapper {
+    /** Input endianness. @internal ICU 2.8 */
+    UBool inIsBigEndian;
+    /** Input charset family. @see U_CHARSET_FAMILY @internal ICU 2.8 */
+    uint8_t inCharset;
+    /** Output endianness. @internal ICU 2.8 */
+    UBool outIsBigEndian;
+    /** Output charset family. @see U_CHARSET_FAMILY @internal ICU 2.8 */
+    uint8_t outCharset;
+
+    /* basic functions for reading data values */
+
+    /** Convert one uint16_t from input to platform endianness. @internal ICU 2.8 */
+    UDataReadUInt16 *readUInt16;
+    /** Convert one uint32_t from input to platform endianness. @internal ICU 2.8 */
+    UDataReadUInt32 *readUInt32;
+    /** Compare an invariant-character output string with a local one. @internal ICU 2.8 */
+    UDataCompareInvChars *compareInvChars;
+
+    /* basic functions for writing data values */
+
+    /** Convert one uint16_t from platform to input endianness. @internal ICU 2.8 */
+    UDataWriteUInt16 *writeUInt16;
+    /** Convert one uint32_t from platform to input endianness. @internal ICU 2.8 */
+    UDataWriteUInt32 *writeUInt32;
+
+    /* basic functions for data transformations */
+
+    /** Transform an array of 16-bit integers. @internal ICU 2.8 */
+    UDataSwapFn *swapArray16;
+    /** Transform an array of 32-bit integers. @internal ICU 2.8 */
+    UDataSwapFn *swapArray32;
+    /** Transform an array of 64-bit integers. @internal ICU 53 */
+    UDataSwapFn *swapArray64;
+    /** Transform an invariant-character string. @internal ICU 2.8 */
+    UDataSwapFn *swapInvChars;
+
+    /**
+     * Function for message output when an error occurs during data swapping.
+     * Can be NULL.
+     * @internal ICU 2.8
+     */
+    UDataPrintError *printError;
+    /** Context pointer for printError. @internal ICU 2.8 */
+    void *printErrorContext;
+};
+
+U_CDECL_END
+
+U_CAPI UDataSwapper * U_EXPORT2
+udata_openSwapper(UBool inIsBigEndian, uint8_t inCharset,
+                  UBool outIsBigEndian, uint8_t outCharset,
+                  UErrorCode *pErrorCode);
+
+/**
+ * Open a UDataSwapper for the given input data and the specified output
+ * characteristics.
+ * Values of -1 for any of the characteristics mean the local platform's
+ * characteristics.
+ *
+ * @see udata_swap
+ * @internal ICU 2.8
+ */
+U_CAPI UDataSwapper * U_EXPORT2
+udata_openSwapperForInputData(const void *data, int32_t length,
+                              UBool outIsBigEndian, uint8_t outCharset,
+                              UErrorCode *pErrorCode);
+
+U_CAPI void U_EXPORT2
+udata_closeSwapper(UDataSwapper *ds);
+
+/**
+ * Read the beginning of an ICU data piece, recognize magic bytes,
+ * swap the structure.
+ * Set a U_UNSUPPORTED_ERROR if it does not look like an ICU data piece.
+ *
+ * @return The size of the data header, in bytes.
+ *
+ * @internal ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+udata_swapDataHeader(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Convert one int16_t from input to platform endianness.
+ * @internal ICU 2.8
+ */
+U_CAPI int16_t U_EXPORT2
+udata_readInt16(const UDataSwapper *ds, int16_t x);
+
+/**
+ * Convert one int32_t from input to platform endianness.
+ * @internal ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+udata_readInt32(const UDataSwapper *ds, int32_t x);
+
+/**
+ * Swap a block of invariant, NUL-terminated strings, but not padding
+ * bytes after the last string.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+udata_swapInvStringBlock(const UDataSwapper *ds,
+                         const void *inData, int32_t length, void *outData,
+                         UErrorCode *pErrorCode);
+
+U_CAPI void U_EXPORT2
+udata_printError(const UDataSwapper *ds,
+                 const char *fmt,
+                 ...);
+
+/* internal exports from putil.c -------------------------------------------- */
+
+/* declared here to keep them out of the public putil.h */
+
+/**
+ * Swap invariant char * strings ASCII->EBCDIC.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_ebcdicFromAscii(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Copy invariant ASCII char * strings and verify they are invariant.
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_copyAscii(const UDataSwapper *ds,
+               const void *inData, int32_t length, void *outData,
+               UErrorCode *pErrorCode);
+
+/**
+ * Swap invariant char * strings EBCDIC->ASCII.
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_asciiFromEbcdic(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Copy invariant EBCDIC char * strings and verify they are invariant.
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_copyEbcdic(const UDataSwapper *ds,
+                const void *inData, int32_t length, void *outData,
+                UErrorCode *pErrorCode);
+
+/**
+ * Compare ASCII invariant char * with Unicode invariant UChar *
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_compareInvAscii(const UDataSwapper *ds,
+                     const char *outString, int32_t outLength,
+                     const UChar *localString, int32_t localLength);
+
+/**
+ * Compare EBCDIC invariant char * with Unicode invariant UChar *
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_compareInvEbcdic(const UDataSwapper *ds,
+                      const char *outString, int32_t outLength,
+                      const UChar *localString, int32_t localLength);
+
+/**
+ * \def uprv_compareInvWithUChar
+ * Compare an invariant-character strings with a UChar string
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_compareInvWithUChar uprv_compareInvAscii
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_compareInvWithUChar uprv_compareInvEbcdic
+#else
+#   error Unknown charset family!
+#endif
+
+// utrie_swap.cpp -----------------------------------------------------------***
+
+/**
+ * Swaps a serialized UTrie.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_swap(const UDataSwapper *ds,
+           const void *inData, int32_t length, void *outData,
+           UErrorCode *pErrorCode);
+
+/**
+ * Swaps a serialized UTrie2.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utrie2_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode);
+
+/**
+ * Swaps a serialized UCPTrie.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucptrie_swap(const UDataSwapper *ds,
+             const void *inData, int32_t length, void *outData,
+             UErrorCode *pErrorCode);
+
+/**
+ * Swaps a serialized UTrie, UTrie2, or UCPTrie.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_swapAnyVersion(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode);
+
+/* material... -------------------------------------------------------------- */
+
+#if 0
+
+/* udata.h */
+
+/**
+ * Public API function in udata.c
+ *
+ * Same as udata_openChoice() but automatically swaps the data.
+ * isAcceptable, if not NULL, may accept data with endianness and charset family
+ * different from the current platform's properties.
+ * If the data is acceptable and the platform properties do not match, then
+ * the swap function is called to swap an allocated version of the data.
+ * Preflighting may or may not be performed depending on whether the size of
+ * the loaded data item is known.
+ *
+ * @param isAcceptable Same as for udata_openChoice(). May be NULL.
+ *
+ * @internal ICU 2.8
+ */
+U_CAPI UDataMemory * U_EXPORT2
+udata_openSwap(const char *path, const char *type, const char *name,
+               UDataMemoryIsAcceptable *isAcceptable, void *isAcceptableContext,
+               UDataSwapFn *swap,
+               UDataPrintError *printError, void *printErrorContext,
+               UErrorCode *pErrorCode);
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/uelement.h b/thirdparty/icu4c/common/uelement.h
new file mode 100644
index 0000000000..88dd4d66fb
--- /dev/null
+++ b/thirdparty/icu4c/common/uelement.h
@@ -0,0 +1,91 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 1997-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  uelement.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011jul04
+*   created by: Markus W. Scherer
+*
+*   Common definitions for UHashTable and UVector.
+*   UHashTok moved here from uhash.h and renamed UElement.
+*   This allows users of UVector to avoid the confusing #include of uhash.h.
+*   uhash.h aliases UElement to UHashTok,
+*   so that we need not change all of its code and its users.
+*/
+
+#ifndef __UELEMENT_H__
+#define __UELEMENT_H__
+
+#include "unicode/utypes.h"
+
+U_CDECL_BEGIN
+
+/**
+ * A UVector element, or a key or value within a UHashtable.
+ * It may be either a 32-bit integral value or an opaque void* pointer.
+ * The void* pointer may be smaller than 32 bits (e.g. 24 bits)
+ * or may be larger (e.g. 64 bits).
+ *
+ * Because a UElement is the size of a native pointer or a 32-bit
+ * integer, we pass it around by value.
+ */
+union UElement {
+    void*   pointer;
+    int32_t integer;
+};
+typedef union UElement UElement;
+
+/**
+ * An element-equality (boolean) comparison function.
+ * @param e1 An element (object or integer)
+ * @param e2 An element (object or integer)
+ * @return true if the two elements are equal.
+ */
+typedef UBool U_CALLCONV UElementsAreEqual(const UElement e1, const UElement e2);
+
+/**
+ * An element sorting (three-way) comparison function.
+ * @param e1 An element (object or integer)
+ * @param e2 An element (object or integer)
+ * @return 0 if the two elements are equal, -1 if e1 is < e2, or +1 if e1 is > e2.
+ */
+typedef int8_t U_CALLCONV UElementComparator(UElement e1, UElement e2);
+
+/**
+ * An element assignment function.  It may copy an integer, copy
+ * a pointer, or clone a pointer, as appropriate.
+ * @param dst The element to be assigned to
+ * @param src The element to assign from
+ */
+typedef void U_CALLCONV UElementAssigner(UElement *dst, UElement *src);
+
+U_CDECL_END
+
+/**
+ * Comparator function for UnicodeString* keys. Implements UElementsAreEqual.
+ * @param key1 The string for comparison
+ * @param key2 The string for comparison
+ * @return true if key1 and key2 are equal, return false otherwise.
+ */
+U_CAPI UBool U_EXPORT2 
+uhash_compareUnicodeString(const UElement key1, const UElement key2);
+
+/**
+ * Comparator function for UnicodeString* keys (case insensitive).
+ * Make sure to use together with uhash_hashCaselessUnicodeString.
+ * Implements UElementsAreEqual.
+ * @param key1 The string for comparison
+ * @param key2 The string for comparison
+ * @return true if key1 and key2 are equal, return false otherwise.
+ */
+U_CAPI UBool U_EXPORT2 
+uhash_compareCaselessUnicodeString(const UElement key1, const UElement key2);
+
+#endif  /* __UELEMENT_H__ */
diff --git a/thirdparty/icu4c/common/uenum.cpp b/thirdparty/icu4c/common/uenum.cpp
new file mode 100644
index 0000000000..11d895ebcd
--- /dev/null
+++ b/thirdparty/icu4c/common/uenum.cpp
@@ -0,0 +1,189 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uenum.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2002jul08
+*   created by: Vladimir Weinstein
+*/
+
+#include "unicode/putil.h"
+#include "uenumimp.h"
+#include "cmemory.h"
+
+/* Layout of the baseContext buffer. */
+typedef struct {
+    int32_t len;  /* number of bytes available starting at 'data' */
+    char    data; /* actual data starts here */
+} _UEnumBuffer;
+
+/* Extra bytes to allocate in the baseContext buffer. */
+static const int32_t PAD = 8;
+
+/* Return a pointer to the baseContext buffer, possibly allocating
+   or reallocating it if at least 'capacity' bytes are not available. */
+static void* _getBuffer(UEnumeration* en, int32_t capacity) {
+
+    if (en->baseContext != NULL) {
+        if (((_UEnumBuffer*) en->baseContext)->len < capacity) {
+            capacity += PAD;
+            en->baseContext = uprv_realloc(en->baseContext,
+                                           sizeof(int32_t) + capacity);
+            if (en->baseContext == NULL) {
+                return NULL;
+            }
+            ((_UEnumBuffer*) en->baseContext)->len = capacity;
+        }
+    } else {
+        capacity += PAD;
+        en->baseContext = uprv_malloc(sizeof(int32_t) + capacity);
+        if (en->baseContext == NULL) {
+            return NULL;
+        }
+        ((_UEnumBuffer*) en->baseContext)->len = capacity;
+    }
+    
+    return (void*) & ((_UEnumBuffer*) en->baseContext)->data;
+}
+
+U_CAPI void U_EXPORT2
+uenum_close(UEnumeration* en)
+{
+    if (en) {
+        if (en->close != NULL) {
+            if (en->baseContext) {
+                uprv_free(en->baseContext);
+            }
+            en->close(en);
+        } else { /* this seems dangerous, but we better kill the object */
+            uprv_free(en);
+        }
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uenum_count(UEnumeration* en, UErrorCode* status)
+{
+    if (!en || U_FAILURE(*status)) {
+        return -1;
+    }
+    if (en->count != NULL) {
+        return en->count(en, status);
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+        return -1;
+    }
+}
+
+/* Don't call this directly. Only uenum_unext should be calling this. */
+U_CAPI const UChar* U_EXPORT2
+uenum_unextDefault(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status)
+{
+    UChar *ustr = NULL;
+    int32_t len = 0;
+    if (en->next != NULL) {
+        const char *cstr = en->next(en, &len, status);
+        if (cstr != NULL) {
+            ustr = (UChar*) _getBuffer(en, (len+1) * sizeof(UChar));
+            if (ustr == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+            } else {
+                u_charsToUChars(cstr, ustr, len+1);
+            }
+        }
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+    }
+    if (resultLength) {
+        *resultLength = len;
+    }
+    return ustr;
+}
+
+/* Don't call this directly. Only uenum_next should be calling this. */
+U_CAPI const char* U_EXPORT2
+uenum_nextDefault(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status)
+{
+    if (en->uNext != NULL) {
+        char *tempCharVal;
+        const UChar *tempUCharVal = en->uNext(en, resultLength, status);
+        if (tempUCharVal == NULL) {
+            return NULL;
+        }
+        tempCharVal = (char*)
+            _getBuffer(en, (*resultLength+1) * sizeof(char));
+        if (!tempCharVal) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+        u_UCharsToChars(tempUCharVal, tempCharVal, *resultLength + 1);
+        return tempCharVal;
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+        return NULL;
+    }
+}
+
+U_CAPI const UChar* U_EXPORT2
+uenum_unext(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status)
+{
+    if (!en || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (en->uNext != NULL) {
+        return en->uNext(en, resultLength, status);
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+        return NULL;
+    }
+}
+
+U_CAPI const char* U_EXPORT2
+uenum_next(UEnumeration* en,
+          int32_t* resultLength,
+          UErrorCode* status)
+{
+    if (!en || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (en->next != NULL) {
+        if (resultLength != NULL) {
+            return en->next(en, resultLength, status);
+        }
+        else {
+            int32_t dummyLength=0;
+            return en->next(en, &dummyLength, status);
+        }
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+        return NULL;
+    }
+}
+
+U_CAPI void U_EXPORT2
+uenum_reset(UEnumeration* en, UErrorCode* status)
+{
+    if (!en || U_FAILURE(*status)) {
+        return;
+    }
+    if (en->reset != NULL) {
+        en->reset(en, status);
+    } else {
+        *status = U_UNSUPPORTED_ERROR;
+    }
+}
diff --git a/thirdparty/icu4c/common/uenumimp.h b/thirdparty/icu4c/common/uenumimp.h
new file mode 100644
index 0000000000..9c9df75ae0
--- /dev/null
+++ b/thirdparty/icu4c/common/uenumimp.h
@@ -0,0 +1,155 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uenumimp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2002jul08
+*   created by: Vladimir Weinstein
+*/
+
+#ifndef __UENUMIMP_H
+#define __UENUMIMP_H
+
+#include "unicode/uenum.h"
+
+U_CDECL_BEGIN
+
+/** 
+ * following are the type declarations for 
+ * implementations of APIs. If any of these
+ * functions are NULL, U_UNSUPPORTED_ERROR
+ * is returned. If close is NULL, the enumeration
+ * object is going to be released.
+ * Initial error checking is done in the body
+ * of API function, so the implementations 
+ * need not to check the initial error condition.
+ */
+
+/**
+ * Function type declaration for uenum_close().
+ *
+ * This function should cleanup the enumerator object
+ *
+ * @param en enumeration to be closed
+ */
+typedef void U_CALLCONV
+UEnumClose(UEnumeration *en);
+
+/**
+ * Function type declaration for uenum_count().
+ *
+ * This function should count the number of elements
+ * in this enumeration
+ *
+ * @param en enumeration to be counted
+ * @param status pointer to UErrorCode variable
+ * @return number of elements in enumeration
+ */
+typedef int32_t U_CALLCONV
+UEnumCount(UEnumeration *en, UErrorCode *status);
+
+/**
+ * Function type declaration for uenum_unext().
+ *
+ * This function returns the next element as a UChar *,
+ * or NULL after all elements haven been enumerated.
+ *
+ * @param en enumeration 
+ * @param resultLength pointer to result length
+ * @param status pointer to UErrorCode variable
+ * @return next element as UChar *,
+ *         or NULL after all elements haven been enumerated
+ */
+typedef const UChar* U_CALLCONV 
+UEnumUNext(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status);
+
+/**
+ * Function type declaration for uenum_next().
+ *
+ * This function returns the next element as a char *,
+ * or NULL after all elements haven been enumerated.
+ *
+ * @param en enumeration 
+ * @param resultLength pointer to result length
+ * @param status pointer to UErrorCode variable
+ * @return next element as char *,
+ *         or NULL after all elements haven been enumerated
+ */
+typedef const char* U_CALLCONV 
+UEnumNext(UEnumeration* en,
+           int32_t* resultLength,
+           UErrorCode* status);
+
+/**
+ * Function type declaration for uenum_reset().
+ *
+ * This function should reset the enumeration 
+ * object
+ *
+ * @param en enumeration 
+ * @param status pointer to UErrorCode variable
+ */
+typedef void U_CALLCONV 
+UEnumReset(UEnumeration* en, 
+            UErrorCode* status);
+
+
+struct UEnumeration {
+    /* baseContext. For the base class only. Don't touch! */
+    void *baseContext;
+
+    /* context. Use it for what you need */
+    void *context;
+
+    /** 
+     * these are functions that will 
+     * be used for APIs
+     */
+    /* called from uenum_close */
+    UEnumClose *close;
+    /* called from uenum_count */
+    UEnumCount *count;
+    /* called from uenum_unext */
+    UEnumUNext *uNext;
+    /* called from uenum_next */
+    UEnumNext  *next;
+    /* called from uenum_reset */
+    UEnumReset *reset;
+};
+
+U_CDECL_END
+
+/* This is the default implementation for uenum_unext().
+ * It automatically converts the char * string to UChar *.
+ * Don't call this directly.  This is called internally by uenum_unext
+ * when a UEnumeration is defined with 'uNext' pointing to this
+ * function.
+ */
+U_CAPI const UChar* U_EXPORT2
+uenum_unextDefault(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status);
+
+/* This is the default implementation for uenum_next().
+ * It automatically converts the UChar * string to char *.
+ * Don't call this directly.  This is called internally by uenum_next
+ * when a UEnumeration is defined with 'next' pointing to this
+ * function.
+ */
+U_CAPI const char* U_EXPORT2
+uenum_nextDefault(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status);
+
+#endif
diff --git a/thirdparty/icu4c/common/uhash.cpp b/thirdparty/icu4c/common/uhash.cpp
new file mode 100644
index 0000000000..86311ceb0b
--- /dev/null
+++ b/thirdparty/icu4c/common/uhash.cpp
@@ -0,0 +1,991 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   03/22/00    aliu        Adapted from original C++ ICU Hashtable.
+*   07/06/01    aliu        Modified to support int32_t keys on
+*                           platforms with sizeof(void*) < 32.
+******************************************************************************
+*/
+
+#include "uhash.h"
+#include "unicode/ustring.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "ustr_imp.h"
+
+/* This hashtable is implemented as a double hash.  All elements are
+ * stored in a single array with no secondary storage for collision
+ * resolution (no linked list, etc.).  When there is a hash collision
+ * (when two unequal keys have the same hashcode) we resolve this by
+ * using a secondary hash.  The secondary hash is an increment
+ * computed as a hash function (a different one) of the primary
+ * hashcode.  This increment is added to the initial hash value to
+ * obtain further slots assigned to the same hash code.  For this to
+ * work, the length of the array and the increment must be relatively
+ * prime.  The easiest way to achieve this is to have the length of
+ * the array be prime, and the increment be any value from
+ * 1..length-1.
+ *
+ * Hashcodes are 32-bit integers.  We make sure all hashcodes are
+ * non-negative by masking off the top bit.  This has two effects: (1)
+ * modulo arithmetic is simplified.  If we allowed negative hashcodes,
+ * then when we computed hashcode % length, we could get a negative
+ * result, which we would then have to adjust back into range.  It's
+ * simpler to just make hashcodes non-negative. (2) It makes it easy
+ * to check for empty vs. occupied slots in the table.  We just mark
+ * empty or deleted slots with a negative hashcode.
+ *
+ * The central function is _uhash_find().  This function looks for a
+ * slot matching the given key and hashcode.  If one is found, it
+ * returns a pointer to that slot.  If the table is full, and no match
+ * is found, it returns NULL -- in theory.  This would make the code
+ * more complicated, since all callers of _uhash_find() would then
+ * have to check for a NULL result.  To keep this from happening, we
+ * don't allow the table to fill.  When there is only one
+ * empty/deleted slot left, uhash_put() will refuse to increase the
+ * count, and fail.  This simplifies the code.  In practice, one will
+ * seldom encounter this using default UHashtables.  However, if a
+ * hashtable is set to a U_FIXED resize policy, or if memory is
+ * exhausted, then the table may fill.
+ *
+ * High and low water ratios control rehashing.  They establish levels
+ * of fullness (from 0 to 1) outside of which the data array is
+ * reallocated and repopulated.  Setting the low water ratio to zero
+ * means the table will never shrink.  Setting the high water ratio to
+ * one means the table will never grow.  The ratios should be
+ * coordinated with the ratio between successive elements of the
+ * PRIMES table, so that when the primeIndex is incremented or
+ * decremented during rehashing, it brings the ratio of count / length
+ * back into the desired range (between low and high water ratios).
+ */
+
+/********************************************************************
+ * PRIVATE Constants, Macros
+ ********************************************************************/
+
+/* This is a list of non-consecutive primes chosen such that
+ * PRIMES[i+1] ~ 2*PRIMES[i].  (Currently, the ratio ranges from 1.81
+ * to 2.18; the inverse ratio ranges from 0.459 to 0.552.)  If this
+ * ratio is changed, the low and high water ratios should also be
+ * adjusted to suit.
+ *
+ * These prime numbers were also chosen so that they are the largest
+ * prime number while being less than a power of two.
+ */
+static const int32_t PRIMES[] = {
+    7, 13, 31, 61, 127, 251, 509, 1021, 2039, 4093, 8191, 16381, 32749,
+    65521, 131071, 262139, 524287, 1048573, 2097143, 4194301, 8388593,
+    16777213, 33554393, 67108859, 134217689, 268435399, 536870909,
+    1073741789, 2147483647 /*, 4294967291 */
+};
+
+#define PRIMES_LENGTH UPRV_LENGTHOF(PRIMES)
+#define DEFAULT_PRIME_INDEX 4
+
+/* These ratios are tuned to the PRIMES array such that a resize
+ * places the table back into the zone of non-resizing.  That is,
+ * after a call to _uhash_rehash(), a subsequent call to
+ * _uhash_rehash() should do nothing (should not churn).  This is only
+ * a potential problem with U_GROW_AND_SHRINK.
+ */
+static const float RESIZE_POLICY_RATIO_TABLE[6] = {
+    /* low, high water ratio */
+    0.0F, 0.5F, /* U_GROW: Grow on demand, do not shrink */
+    0.1F, 0.5F, /* U_GROW_AND_SHRINK: Grow and shrink on demand */
+    0.0F, 1.0F  /* U_FIXED: Never change size */
+};
+
+/*
+  Invariants for hashcode values:
+
+  * DELETED < 0
+  * EMPTY < 0
+  * Real hashes >= 0
+
+  Hashcodes may not start out this way, but internally they are
+  adjusted so that they are always positive.  We assume 32-bit
+  hashcodes; adjust these constants for other hashcode sizes.
+*/
+#define HASH_DELETED    ((int32_t) 0x80000000)
+#define HASH_EMPTY      ((int32_t) HASH_DELETED + 1)
+
+#define IS_EMPTY_OR_DELETED(x) ((x) < 0)
+
+/* This macro expects a UHashTok.pointer as its keypointer and
+   valuepointer parameters */
+#define HASH_DELETE_KEY_VALUE(hash, keypointer, valuepointer) UPRV_BLOCK_MACRO_BEGIN { \
+    if (hash->keyDeleter != NULL && keypointer != NULL) { \
+        (*hash->keyDeleter)(keypointer); \
+    } \
+    if (hash->valueDeleter != NULL && valuepointer != NULL) { \
+        (*hash->valueDeleter)(valuepointer); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/*
+ * Constants for hinting whether a key or value is an integer
+ * or a pointer.  If a hint bit is zero, then the associated
+ * token is assumed to be an integer.
+ */
+#define HINT_KEY_POINTER   (1)
+#define HINT_VALUE_POINTER (2)
+
+/********************************************************************
+ * PRIVATE Implementation
+ ********************************************************************/
+
+static UHashTok
+_uhash_setElement(UHashtable *hash, UHashElement* e,
+                  int32_t hashcode,
+                  UHashTok key, UHashTok value, int8_t hint) {
+
+    UHashTok oldValue = e->value;
+    if (hash->keyDeleter != NULL && e->key.pointer != NULL &&
+        e->key.pointer != key.pointer) { /* Avoid double deletion */
+        (*hash->keyDeleter)(e->key.pointer);
+    }
+    if (hash->valueDeleter != NULL) {
+        if (oldValue.pointer != NULL &&
+            oldValue.pointer != value.pointer) { /* Avoid double deletion */
+            (*hash->valueDeleter)(oldValue.pointer);
+        }
+        oldValue.pointer = NULL;
+    }
+    /* Compilers should copy the UHashTok union correctly, but even if
+     * they do, memory heap tools (e.g. BoundsChecker) can get
+     * confused when a pointer is cloaked in a union and then copied.
+     * TO ALLEVIATE THIS, we use hints (based on what API the user is
+     * calling) to copy pointers when we know the user thinks
+     * something is a pointer. */
+    if (hint & HINT_KEY_POINTER) {
+        e->key.pointer = key.pointer;
+    } else {
+        e->key = key;
+    }
+    if (hint & HINT_VALUE_POINTER) {
+        e->value.pointer = value.pointer;
+    } else {
+        e->value = value;
+    }
+    e->hashcode = hashcode;
+    return oldValue;
+}
+
+/**
+ * Assumes that the given element is not empty or deleted.
+ */
+static UHashTok
+_uhash_internalRemoveElement(UHashtable *hash, UHashElement* e) {
+    UHashTok empty;
+    U_ASSERT(!IS_EMPTY_OR_DELETED(e->hashcode));
+    --hash->count;
+    empty.pointer = NULL; empty.integer = 0;
+    return _uhash_setElement(hash, e, HASH_DELETED, empty, empty, 0);
+}
+
+static void
+_uhash_internalSetResizePolicy(UHashtable *hash, enum UHashResizePolicy policy) {
+    U_ASSERT(hash != NULL);
+    U_ASSERT(((int32_t)policy) >= 0);
+    U_ASSERT(((int32_t)policy) < 3);
+    hash->lowWaterRatio  = RESIZE_POLICY_RATIO_TABLE[policy * 2];
+    hash->highWaterRatio = RESIZE_POLICY_RATIO_TABLE[policy * 2 + 1];
+}
+
+/**
+ * Allocate internal data array of a size determined by the given
+ * prime index.  If the index is out of range it is pinned into range.
+ * If the allocation fails the status is set to
+ * U_MEMORY_ALLOCATION_ERROR and all array storage is freed.  In
+ * either case the previous array pointer is overwritten.
+ *
+ * Caller must ensure primeIndex is in range 0..PRIME_LENGTH-1.
+ */
+static void
+_uhash_allocate(UHashtable *hash,
+                int32_t primeIndex,
+                UErrorCode *status) {
+
+    UHashElement *p, *limit;
+    UHashTok emptytok;
+
+    if (U_FAILURE(*status)) return;
+
+    U_ASSERT(primeIndex >= 0 && primeIndex < PRIMES_LENGTH);
+
+    hash->primeIndex = static_cast<int8_t>(primeIndex);
+    hash->length = PRIMES[primeIndex];
+
+    p = hash->elements = (UHashElement*)
+        uprv_malloc(sizeof(UHashElement) * hash->length);
+
+    if (hash->elements == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    emptytok.pointer = NULL; /* Only one of these two is needed */
+    emptytok.integer = 0;    /* but we don't know which one. */
+
+    limit = p + hash->length;
+    while (p < limit) {
+        p->key = emptytok;
+        p->value = emptytok;
+        p->hashcode = HASH_EMPTY;
+        ++p;
+    }
+
+    hash->count = 0;
+    hash->lowWaterMark = (int32_t)(hash->length * hash->lowWaterRatio);
+    hash->highWaterMark = (int32_t)(hash->length * hash->highWaterRatio);
+}
+
+static UHashtable*
+_uhash_init(UHashtable *result,
+              UHashFunction *keyHash,
+              UKeyComparator *keyComp,
+              UValueComparator *valueComp,
+              int32_t primeIndex,
+              UErrorCode *status)
+{
+    if (U_FAILURE(*status)) return NULL;
+    U_ASSERT(keyHash != NULL);
+    U_ASSERT(keyComp != NULL);
+
+    result->keyHasher       = keyHash;
+    result->keyComparator   = keyComp;
+    result->valueComparator = valueComp;
+    result->keyDeleter      = NULL;
+    result->valueDeleter    = NULL;
+    result->allocated       = FALSE;
+    _uhash_internalSetResizePolicy(result, U_GROW);
+
+    _uhash_allocate(result, primeIndex, status);
+
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    return result;
+}
+
+static UHashtable*
+_uhash_create(UHashFunction *keyHash,
+              UKeyComparator *keyComp,
+              UValueComparator *valueComp,
+              int32_t primeIndex,
+              UErrorCode *status) {
+    UHashtable *result;
+
+    if (U_FAILURE(*status)) return NULL;
+
+    result = (UHashtable*) uprv_malloc(sizeof(UHashtable));
+    if (result == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+
+    _uhash_init(result, keyHash, keyComp, valueComp, primeIndex, status);
+    result->allocated       = TRUE;
+
+    if (U_FAILURE(*status)) {
+        uprv_free(result);
+        return NULL;
+    }
+
+    return result;
+}
+
+/**
+ * Look for a key in the table, or if no such key exists, the first
+ * empty slot matching the given hashcode.  Keys are compared using
+ * the keyComparator function.
+ *
+ * First find the start position, which is the hashcode modulo
+ * the length.  Test it to see if it is:
+ *
+ * a. identical:  First check the hash values for a quick check,
+ *    then compare keys for equality using keyComparator.
+ * b. deleted
+ * c. empty
+ *
+ * Stop if it is identical or empty, otherwise continue by adding a
+ * "jump" value (moduloing by the length again to keep it within
+ * range) and retesting.  For efficiency, there need enough empty
+ * values so that the searchs stop within a reasonable amount of time.
+ * This can be changed by changing the high/low water marks.
+ *
+ * In theory, this function can return NULL, if it is full (no empty
+ * or deleted slots) and if no matching key is found.  In practice, we
+ * prevent this elsewhere (in uhash_put) by making sure the last slot
+ * in the table is never filled.
+ *
+ * The size of the table should be prime for this algorithm to work;
+ * otherwise we are not guaranteed that the jump value (the secondary
+ * hash) is relatively prime to the table length.
+ */
+static UHashElement*
+_uhash_find(const UHashtable *hash, UHashTok key,
+            int32_t hashcode) {
+
+    int32_t firstDeleted = -1;  /* assume invalid index */
+    int32_t theIndex, startIndex;
+    int32_t jump = 0; /* lazy evaluate */
+    int32_t tableHash;
+    UHashElement *elements = hash->elements;
+
+    hashcode &= 0x7FFFFFFF; /* must be positive */
+    startIndex = theIndex = (hashcode ^ 0x4000000) % hash->length;
+
+    do {
+        tableHash = elements[theIndex].hashcode;
+        if (tableHash == hashcode) {          /* quick check */
+            if ((*hash->keyComparator)(key, elements[theIndex].key)) {
+                return &(elements[theIndex]);
+            }
+        } else if (!IS_EMPTY_OR_DELETED(tableHash)) {
+            /* We have hit a slot which contains a key-value pair,
+             * but for which the hash code does not match.  Keep
+             * looking.
+             */
+        } else if (tableHash == HASH_EMPTY) { /* empty, end o' the line */
+            break;
+        } else if (firstDeleted < 0) { /* remember first deleted */
+            firstDeleted = theIndex;
+        }
+        if (jump == 0) { /* lazy compute jump */
+            /* The jump value must be relatively prime to the table
+             * length.  As long as the length is prime, then any value
+             * 1..length-1 will be relatively prime to it.
+             */
+            jump = (hashcode % (hash->length - 1)) + 1;
+        }
+        theIndex = (theIndex + jump) % hash->length;
+    } while (theIndex != startIndex);
+
+    if (firstDeleted >= 0) {
+        theIndex = firstDeleted; /* reset if had deleted slot */
+    } else if (tableHash != HASH_EMPTY) {
+        /* We get to this point if the hashtable is full (no empty or
+         * deleted slots), and we've failed to find a match.  THIS
+         * WILL NEVER HAPPEN as long as uhash_put() makes sure that
+         * count is always < length.
+         */
+        UPRV_UNREACHABLE;
+    }
+    return &(elements[theIndex]);
+}
+
+/**
+ * Attempt to grow or shrink the data arrays in order to make the
+ * count fit between the high and low water marks.  hash_put() and
+ * hash_remove() call this method when the count exceeds the high or
+ * low water marks.  This method may do nothing, if memory allocation
+ * fails, or if the count is already in range, or if the length is
+ * already at the low or high limit.  In any case, upon return the
+ * arrays will be valid.
+ */
+static void
+_uhash_rehash(UHashtable *hash, UErrorCode *status) {
+
+    UHashElement *old = hash->elements;
+    int32_t oldLength = hash->length;
+    int32_t newPrimeIndex = hash->primeIndex;
+    int32_t i;
+
+    if (hash->count > hash->highWaterMark) {
+        if (++newPrimeIndex >= PRIMES_LENGTH) {
+            return;
+        }
+    } else if (hash->count < hash->lowWaterMark) {
+        if (--newPrimeIndex < 0) {
+            return;
+        }
+    } else {
+        return;
+    }
+
+    _uhash_allocate(hash, newPrimeIndex, status);
+
+    if (U_FAILURE(*status)) {
+        hash->elements = old;
+        hash->length = oldLength;
+        return;
+    }
+
+    for (i = oldLength - 1; i >= 0; --i) {
+        if (!IS_EMPTY_OR_DELETED(old[i].hashcode)) {
+            UHashElement *e = _uhash_find(hash, old[i].key, old[i].hashcode);
+            U_ASSERT(e != NULL);
+            U_ASSERT(e->hashcode == HASH_EMPTY);
+            e->key = old[i].key;
+            e->value = old[i].value;
+            e->hashcode = old[i].hashcode;
+            ++hash->count;
+        }
+    }
+
+    uprv_free(old);
+}
+
+static UHashTok
+_uhash_remove(UHashtable *hash,
+              UHashTok key) {
+    /* First find the position of the key in the table.  If the object
+     * has not been removed already, remove it.  If the user wanted
+     * keys deleted, then delete it also.  We have to put a special
+     * hashcode in that position that means that something has been
+     * deleted, since when we do a find, we have to continue PAST any
+     * deleted values.
+     */
+    UHashTok result;
+    UHashElement* e = _uhash_find(hash, key, hash->keyHasher(key));
+    U_ASSERT(e != NULL);
+    result.pointer = NULL;
+    result.integer = 0;
+    if (!IS_EMPTY_OR_DELETED(e->hashcode)) {
+        result = _uhash_internalRemoveElement(hash, e);
+        if (hash->count < hash->lowWaterMark) {
+            UErrorCode status = U_ZERO_ERROR;
+            _uhash_rehash(hash, &status);
+        }
+    }
+    return result;
+}
+
+static UHashTok
+_uhash_put(UHashtable *hash,
+           UHashTok key,
+           UHashTok value,
+           int8_t hint,
+           UErrorCode *status) {
+
+    /* Put finds the position in the table for the new value.  If the
+     * key is already in the table, it is deleted, if there is a
+     * non-NULL keyDeleter.  Then the key, the hash and the value are
+     * all put at the position in their respective arrays.
+     */
+    int32_t hashcode;
+    UHashElement* e;
+    UHashTok emptytok;
+
+    if (U_FAILURE(*status)) {
+        goto err;
+    }
+    U_ASSERT(hash != NULL);
+    /* Cannot always check pointer here or iSeries sees NULL every time. */
+    if ((hint & HINT_VALUE_POINTER) && value.pointer == NULL) {
+        /* Disallow storage of NULL values, since NULL is returned by
+         * get() to indicate an absent key.  Storing NULL == removing.
+         */
+        return _uhash_remove(hash, key);
+    }
+    if (hash->count > hash->highWaterMark) {
+        _uhash_rehash(hash, status);
+        if (U_FAILURE(*status)) {
+            goto err;
+        }
+    }
+
+    hashcode = (*hash->keyHasher)(key);
+    e = _uhash_find(hash, key, hashcode);
+    U_ASSERT(e != NULL);
+
+    if (IS_EMPTY_OR_DELETED(e->hashcode)) {
+        /* Important: We must never actually fill the table up.  If we
+         * do so, then _uhash_find() will return NULL, and we'll have
+         * to check for NULL after every call to _uhash_find().  To
+         * avoid this we make sure there is always at least one empty
+         * or deleted slot in the table.  This only is a problem if we
+         * are out of memory and rehash isn't working.
+         */
+        ++hash->count;
+        if (hash->count == hash->length) {
+            /* Don't allow count to reach length */
+            --hash->count;
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            goto err;
+        }
+    }
+
+    /* We must in all cases handle storage properly.  If there was an
+     * old key, then it must be deleted (if the deleter != NULL).
+     * Make hashcodes stored in table positive.
+     */
+    return _uhash_setElement(hash, e, hashcode & 0x7FFFFFFF, key, value, hint);
+
+ err:
+    /* If the deleters are non-NULL, this method adopts its key and/or
+     * value arguments, and we must be sure to delete the key and/or
+     * value in all cases, even upon failure.
+     */
+    HASH_DELETE_KEY_VALUE(hash, key.pointer, value.pointer);
+    emptytok.pointer = NULL; emptytok.integer = 0;
+    return emptytok;
+}
+
+
+/********************************************************************
+ * PUBLIC API
+ ********************************************************************/
+
+U_CAPI UHashtable* U_EXPORT2
+uhash_open(UHashFunction *keyHash,
+           UKeyComparator *keyComp,
+           UValueComparator *valueComp,
+           UErrorCode *status) {
+
+    return _uhash_create(keyHash, keyComp, valueComp, DEFAULT_PRIME_INDEX, status);
+}
+
+U_CAPI UHashtable* U_EXPORT2
+uhash_openSize(UHashFunction *keyHash,
+               UKeyComparator *keyComp,
+               UValueComparator *valueComp,
+               int32_t size,
+               UErrorCode *status) {
+
+    /* Find the smallest index i for which PRIMES[i] >= size. */
+    int32_t i = 0;
+    while (i<(PRIMES_LENGTH-1) && PRIMES[i]<size) {
+        ++i;
+    }
+
+    return _uhash_create(keyHash, keyComp, valueComp, i, status);
+}
+
+U_CAPI UHashtable* U_EXPORT2
+uhash_init(UHashtable *fillinResult,
+           UHashFunction *keyHash,
+           UKeyComparator *keyComp,
+           UValueComparator *valueComp,
+           UErrorCode *status) {
+
+    return _uhash_init(fillinResult, keyHash, keyComp, valueComp, DEFAULT_PRIME_INDEX, status);
+}
+
+U_CAPI UHashtable* U_EXPORT2
+uhash_initSize(UHashtable *fillinResult,
+               UHashFunction *keyHash,
+               UKeyComparator *keyComp,
+               UValueComparator *valueComp,
+               int32_t size,
+               UErrorCode *status) {
+
+    // Find the smallest index i for which PRIMES[i] >= size.
+    int32_t i = 0;
+    while (i<(PRIMES_LENGTH-1) && PRIMES[i]<size) {
+        ++i;
+    }
+    return _uhash_init(fillinResult, keyHash, keyComp, valueComp, i, status);
+}
+
+U_CAPI void U_EXPORT2
+uhash_close(UHashtable *hash) {
+    if (hash == NULL) {
+        return;
+    }
+    if (hash->elements != NULL) {
+        if (hash->keyDeleter != NULL || hash->valueDeleter != NULL) {
+            int32_t pos=UHASH_FIRST;
+            UHashElement *e;
+            while ((e = (UHashElement*) uhash_nextElement(hash, &pos)) != NULL) {
+                HASH_DELETE_KEY_VALUE(hash, e->key.pointer, e->value.pointer);
+            }
+        }
+        uprv_free(hash->elements);
+        hash->elements = NULL;
+    }
+    if (hash->allocated) {
+        uprv_free(hash);
+    }
+}
+
+U_CAPI UHashFunction *U_EXPORT2
+uhash_setKeyHasher(UHashtable *hash, UHashFunction *fn) {
+    UHashFunction *result = hash->keyHasher;
+    hash->keyHasher = fn;
+    return result;
+}
+
+U_CAPI UKeyComparator *U_EXPORT2
+uhash_setKeyComparator(UHashtable *hash, UKeyComparator *fn) {
+    UKeyComparator *result = hash->keyComparator;
+    hash->keyComparator = fn;
+    return result;
+}
+U_CAPI UValueComparator *U_EXPORT2
+uhash_setValueComparator(UHashtable *hash, UValueComparator *fn){
+    UValueComparator *result = hash->valueComparator;
+    hash->valueComparator = fn;
+    return result;
+}
+
+U_CAPI UObjectDeleter *U_EXPORT2
+uhash_setKeyDeleter(UHashtable *hash, UObjectDeleter *fn) {
+    UObjectDeleter *result = hash->keyDeleter;
+    hash->keyDeleter = fn;
+    return result;
+}
+
+U_CAPI UObjectDeleter *U_EXPORT2
+uhash_setValueDeleter(UHashtable *hash, UObjectDeleter *fn) {
+    UObjectDeleter *result = hash->valueDeleter;
+    hash->valueDeleter = fn;
+    return result;
+}
+
+U_CAPI void U_EXPORT2
+uhash_setResizePolicy(UHashtable *hash, enum UHashResizePolicy policy) {
+    UErrorCode status = U_ZERO_ERROR;
+    _uhash_internalSetResizePolicy(hash, policy);
+    hash->lowWaterMark  = (int32_t)(hash->length * hash->lowWaterRatio);
+    hash->highWaterMark = (int32_t)(hash->length * hash->highWaterRatio);
+    _uhash_rehash(hash, &status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_count(const UHashtable *hash) {
+    return hash->count;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_get(const UHashtable *hash,
+          const void* key) {
+    UHashTok keyholder;
+    keyholder.pointer = (void*) key;
+    return _uhash_find(hash, keyholder, hash->keyHasher(keyholder))->value.pointer;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_iget(const UHashtable *hash,
+           int32_t key) {
+    UHashTok keyholder;
+    keyholder.integer = key;
+    return _uhash_find(hash, keyholder, hash->keyHasher(keyholder))->value.pointer;
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_geti(const UHashtable *hash,
+           const void* key) {
+    UHashTok keyholder;
+    keyholder.pointer = (void*) key;
+    return _uhash_find(hash, keyholder, hash->keyHasher(keyholder))->value.integer;
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_igeti(const UHashtable *hash,
+           int32_t key) {
+    UHashTok keyholder;
+    keyholder.integer = key;
+    return _uhash_find(hash, keyholder, hash->keyHasher(keyholder))->value.integer;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_put(UHashtable *hash,
+          void* key,
+          void* value,
+          UErrorCode *status) {
+    UHashTok keyholder, valueholder;
+    keyholder.pointer = key;
+    valueholder.pointer = value;
+    return _uhash_put(hash, keyholder, valueholder,
+                      HINT_KEY_POINTER | HINT_VALUE_POINTER,
+                      status).pointer;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_iput(UHashtable *hash,
+           int32_t key,
+           void* value,
+           UErrorCode *status) {
+    UHashTok keyholder, valueholder;
+    keyholder.integer = key;
+    valueholder.pointer = value;
+    return _uhash_put(hash, keyholder, valueholder,
+                      HINT_VALUE_POINTER,
+                      status).pointer;
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_puti(UHashtable *hash,
+           void* key,
+           int32_t value,
+           UErrorCode *status) {
+    UHashTok keyholder, valueholder;
+    keyholder.pointer = key;
+    valueholder.integer = value;
+    return _uhash_put(hash, keyholder, valueholder,
+                      HINT_KEY_POINTER,
+                      status).integer;
+}
+
+
+U_CAPI int32_t U_EXPORT2
+uhash_iputi(UHashtable *hash,
+           int32_t key,
+           int32_t value,
+           UErrorCode *status) {
+    UHashTok keyholder, valueholder;
+    keyholder.integer = key;
+    valueholder.integer = value;
+    return _uhash_put(hash, keyholder, valueholder,
+                      0, /* neither is a ptr */
+                      status).integer;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_remove(UHashtable *hash,
+             const void* key) {
+    UHashTok keyholder;
+    keyholder.pointer = (void*) key;
+    return _uhash_remove(hash, keyholder).pointer;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_iremove(UHashtable *hash,
+              int32_t key) {
+    UHashTok keyholder;
+    keyholder.integer = key;
+    return _uhash_remove(hash, keyholder).pointer;
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_removei(UHashtable *hash,
+              const void* key) {
+    UHashTok keyholder;
+    keyholder.pointer = (void*) key;
+    return _uhash_remove(hash, keyholder).integer;
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_iremovei(UHashtable *hash,
+               int32_t key) {
+    UHashTok keyholder;
+    keyholder.integer = key;
+    return _uhash_remove(hash, keyholder).integer;
+}
+
+U_CAPI void U_EXPORT2
+uhash_removeAll(UHashtable *hash) {
+    int32_t pos = UHASH_FIRST;
+    const UHashElement *e;
+    U_ASSERT(hash != NULL);
+    if (hash->count != 0) {
+        while ((e = uhash_nextElement(hash, &pos)) != NULL) {
+            uhash_removeElement(hash, e);
+        }
+    }
+    U_ASSERT(hash->count == 0);
+}
+
+U_CAPI const UHashElement* U_EXPORT2
+uhash_find(const UHashtable *hash, const void* key) {
+    UHashTok keyholder;
+    const UHashElement *e;
+    keyholder.pointer = (void*) key;
+    e = _uhash_find(hash, keyholder, hash->keyHasher(keyholder));
+    return IS_EMPTY_OR_DELETED(e->hashcode) ? NULL : e;
+}
+
+U_CAPI const UHashElement* U_EXPORT2
+uhash_nextElement(const UHashtable *hash, int32_t *pos) {
+    /* Walk through the array until we find an element that is not
+     * EMPTY and not DELETED.
+     */
+    int32_t i;
+    U_ASSERT(hash != NULL);
+    for (i = *pos + 1; i < hash->length; ++i) {
+        if (!IS_EMPTY_OR_DELETED(hash->elements[i].hashcode)) {
+            *pos = i;
+            return &(hash->elements[i]);
+        }
+    }
+
+    /* No more elements */
+    return NULL;
+}
+
+U_CAPI void* U_EXPORT2
+uhash_removeElement(UHashtable *hash, const UHashElement* e) {
+    U_ASSERT(hash != NULL);
+    U_ASSERT(e != NULL);
+    if (!IS_EMPTY_OR_DELETED(e->hashcode)) {
+        UHashElement *nce = (UHashElement *)e;
+        return _uhash_internalRemoveElement(hash, nce).pointer;
+    }
+    return NULL;
+}
+
+/********************************************************************
+ * UHashTok convenience
+ ********************************************************************/
+
+/**
+ * Return a UHashTok for an integer.
+ */
+/*U_CAPI UHashTok U_EXPORT2
+uhash_toki(int32_t i) {
+    UHashTok tok;
+    tok.integer = i;
+    return tok;
+}*/
+
+/**
+ * Return a UHashTok for a pointer.
+ */
+/*U_CAPI UHashTok U_EXPORT2
+uhash_tokp(void* p) {
+    UHashTok tok;
+    tok.pointer = p;
+    return tok;
+}*/
+
+/********************************************************************
+ * PUBLIC Key Hash Functions
+ ********************************************************************/
+
+U_CAPI int32_t U_EXPORT2
+uhash_hashUChars(const UHashTok key) {
+    const UChar *s = (const UChar *)key.pointer;
+    return s == NULL ? 0 : ustr_hashUCharsN(s, u_strlen(s));
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_hashChars(const UHashTok key) {
+    const char *s = (const char *)key.pointer;
+    return s == NULL ? 0 : static_cast<int32_t>(ustr_hashCharsN(s, static_cast<int32_t>(uprv_strlen(s))));
+}
+
+U_CAPI int32_t U_EXPORT2
+uhash_hashIChars(const UHashTok key) {
+    const char *s = (const char *)key.pointer;
+    return s == NULL ? 0 : ustr_hashICharsN(s, static_cast<int32_t>(uprv_strlen(s)));
+}
+
+U_CAPI UBool U_EXPORT2
+uhash_equals(const UHashtable* hash1, const UHashtable* hash2){
+    int32_t count1, count2, pos, i;
+
+    if(hash1==hash2){
+        return TRUE;
+    }
+
+    /*
+     * Make sure that we are comparing 2 valid hashes of the same type
+     * with valid comparison functions.
+     * Without valid comparison functions, a binary comparison
+     * of the hash values will yield random results on machines
+     * with 64-bit pointers and 32-bit integer hashes.
+     * A valueComparator is normally optional.
+     */
+    if (hash1==NULL || hash2==NULL ||
+        hash1->keyComparator != hash2->keyComparator ||
+        hash1->valueComparator != hash2->valueComparator ||
+        hash1->valueComparator == NULL)
+    {
+        /*
+        Normally we would return an error here about incompatible hash tables,
+        but we return FALSE instead.
+        */
+        return FALSE;
+    }
+
+    count1 = uhash_count(hash1);
+    count2 = uhash_count(hash2);
+    if(count1!=count2){
+        return FALSE;
+    }
+
+    pos=UHASH_FIRST;
+    for(i=0; i<count1; i++){
+        const UHashElement* elem1 = uhash_nextElement(hash1, &pos);
+        const UHashTok key1 = elem1->key;
+        const UHashTok val1 = elem1->value;
+        /* here the keys are not compared, instead the key form hash1 is used to fetch
+         * value from hash2. If the hashes are equal then then both hashes should
+         * contain equal values for the same key!
+         */
+        const UHashElement* elem2 = _uhash_find(hash2, key1, hash2->keyHasher(key1));
+        const UHashTok val2 = elem2->value;
+        if(hash1->valueComparator(val1, val2)==FALSE){
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+/********************************************************************
+ * PUBLIC Comparator Functions
+ ********************************************************************/
+
+U_CAPI UBool U_EXPORT2
+uhash_compareUChars(const UHashTok key1, const UHashTok key2) {
+    const UChar *p1 = (const UChar*) key1.pointer;
+    const UChar *p2 = (const UChar*) key2.pointer;
+    if (p1 == p2) {
+        return TRUE;
+    }
+    if (p1 == NULL || p2 == NULL) {
+        return FALSE;
+    }
+    while (*p1 != 0 && *p1 == *p2) {
+        ++p1;
+        ++p2;
+    }
+    return (UBool)(*p1 == *p2);
+}
+
+U_CAPI UBool U_EXPORT2
+uhash_compareChars(const UHashTok key1, const UHashTok key2) {
+    const char *p1 = (const char*) key1.pointer;
+    const char *p2 = (const char*) key2.pointer;
+    if (p1 == p2) {
+        return TRUE;
+    }
+    if (p1 == NULL || p2 == NULL) {
+        return FALSE;
+    }
+    while (*p1 != 0 && *p1 == *p2) {
+        ++p1;
+        ++p2;
+    }
+    return (UBool)(*p1 == *p2);
+}
+
+U_CAPI UBool U_EXPORT2
+uhash_compareIChars(const UHashTok key1, const UHashTok key2) {
+    const char *p1 = (const char*) key1.pointer;
+    const char *p2 = (const char*) key2.pointer;
+    if (p1 == p2) {
+        return TRUE;
+    }
+    if (p1 == NULL || p2 == NULL) {
+        return FALSE;
+    }
+    while (*p1 != 0 && uprv_tolower(*p1) == uprv_tolower(*p2)) {
+        ++p1;
+        ++p2;
+    }
+    return (UBool)(*p1 == *p2);
+}
+
+/********************************************************************
+ * PUBLIC int32_t Support Functions
+ ********************************************************************/
+
+U_CAPI int32_t U_EXPORT2
+uhash_hashLong(const UHashTok key) {
+    return key.integer;
+}
+
+U_CAPI UBool U_EXPORT2
+uhash_compareLong(const UHashTok key1, const UHashTok key2) {
+    return (UBool)(key1.integer == key2.integer);
+}
diff --git a/thirdparty/icu4c/common/uhash.h b/thirdparty/icu4c/common/uhash.h
new file mode 100644
index 0000000000..b59d2711bb
--- /dev/null
+++ b/thirdparty/icu4c/common/uhash.h
@@ -0,0 +1,718 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 1997-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   03/22/00    aliu        Adapted from original C++ ICU Hashtable.
+*   07/06/01    aliu        Modified to support int32_t keys on
+*                           platforms with sizeof(void*) < 32.
+******************************************************************************
+*/
+
+#ifndef UHASH_H
+#define UHASH_H
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "uelement.h"
+#include "unicode/localpointer.h"
+
+/**
+ * UHashtable stores key-value pairs and does moderately fast lookup
+ * based on keys.  It provides a good tradeoff between access time and
+ * storage space.  As elements are added to it, it grows to accomodate
+ * them.  By default, the table never shrinks, even if all elements
+ * are removed from it.
+ *
+ * Keys and values are stored as void* pointers.  These void* pointers
+ * may be actual pointers to strings, objects, or any other structure
+ * in memory, or they may simply be integral values cast to void*.
+ * UHashtable doesn't care and manipulates them via user-supplied
+ * functions.  These functions hash keys, compare keys, delete keys,
+ * and delete values.  Some function pointers are optional (may be
+ * NULL); others must be supplied.  Several prebuilt functions exist
+ * to handle common key types.
+ *
+ * UHashtable ownership of keys and values is flexible, and controlled
+ * by whether or not the key deleter and value deleter functions are
+ * set.  If a void* key is actually a pointer to a deletable object,
+ * then UHashtable can be made to delete that object by setting the
+ * key deleter function pointer to a non-NULL value.  If this is done,
+ * then keys passed to uhash_put() are owned by the hashtable and will
+ * be deleted by it at some point, either as keys are replaced, or
+ * when uhash_close() is finally called.  The same is true of values
+ * and the value deleter function pointer.  Keys passed to methods
+ * other than uhash_put() are never owned by the hashtable.
+ *
+ * NULL values are not allowed.  uhash_get() returns NULL to indicate
+ * a key that is not in the table, and having a NULL value in the
+ * table would generate an ambiguous result.  If a key and a NULL
+ * value is passed to uhash_put(), this has the effect of doing a
+ * uhash_remove() on that key.  This keeps uhash_get(), uhash_count(),
+ * and uhash_nextElement() consistent with one another.
+ *
+ * To see everything in a hashtable, use uhash_nextElement() to
+ * iterate through its contents.  Each call to this function returns a
+ * UHashElement pointer.  A hash element contains a key, value, and
+ * hashcode.  During iteration an element may be deleted by calling
+ * uhash_removeElement(); iteration may safely continue thereafter.
+ * The uhash_remove() function may also be safely called in
+ * mid-iteration.  If uhash_put() is called during iteration,
+ * the iteration is still guaranteed to terminate reasonably, but
+ * there is no guarantee that every element will be returned or that
+ * some won't be returned more than once.
+ *
+ * Under no circumstances should the UHashElement returned by
+ * uhash_nextElement be modified directly.
+ *
+ * By default, the hashtable grows when necessary, but never shrinks,
+ * even if all items are removed.  For most applications this is
+ * optimal.  However, in a highly dynamic usage where memory is at a
+ * premium, the table can be set to both grow and shrink by calling
+ * uhash_setResizePolicy() with the policy U_GROW_AND_SHRINK.  In a
+ * situation where memory is critical and the client wants a table
+ * that does not grow at all, the constant U_FIXED can be used.
+ */
+
+/********************************************************************
+ * Data Structures
+ ********************************************************************/
+
+U_CDECL_BEGIN
+
+/**
+ * A key or value within a UHashtable.
+ * The hashing and comparison functions take a pointer to a
+ * UHashTok, but the deleter receives the void* pointer within it.
+ */
+typedef UElement UHashTok;
+
+/**
+ * This is a single hash element.
+ */
+struct UHashElement {
+    /* Reorder these elements to pack nicely if necessary */
+    int32_t  hashcode;
+    UHashTok value;
+    UHashTok key;
+};
+typedef struct UHashElement UHashElement;
+
+/**
+ * A hashing function.
+ * @param key A key stored in a hashtable
+ * @return A NON-NEGATIVE hash code for parm.
+ */
+typedef int32_t U_CALLCONV UHashFunction(const UHashTok key);
+
+/**
+ * A key equality (boolean) comparison function.
+ */
+typedef UElementsAreEqual UKeyComparator;
+
+/**
+ * A value equality (boolean) comparison function.
+ */
+typedef UElementsAreEqual UValueComparator;
+
+/* see cmemory.h for UObjectDeleter and uprv_deleteUObject() */
+
+/**
+ * This specifies whether or not, and how, the hastable resizes itself.
+ * See uhash_setResizePolicy().
+ */
+enum UHashResizePolicy {
+    U_GROW,            /* Grow on demand, do not shrink */
+    U_GROW_AND_SHRINK, /* Grow and shrink on demand */
+    U_FIXED            /* Never change size */
+};
+
+/**
+ * The UHashtable struct.  Clients should treat this as an opaque data
+ * type and manipulate it only through the uhash_... API.
+ */
+struct UHashtable {
+
+    /* Main key-value pair storage array */
+
+    UHashElement *elements;
+
+    /* Function pointers */
+
+    UHashFunction *keyHasher;      /* Computes hash from key.
+                                   * Never null. */
+    UKeyComparator *keyComparator; /* Compares keys for equality.
+                                   * Never null. */
+    UValueComparator *valueComparator; /* Compares the values for equality */
+
+    UObjectDeleter *keyDeleter;    /* Deletes keys when required.
+                                   * If NULL won't do anything */
+    UObjectDeleter *valueDeleter;  /* Deletes values when required.
+                                   * If NULL won't do anything */
+
+    /* Size parameters */
+
+    int32_t     count;      /* The number of key-value pairs in this table.
+                             * 0 <= count <= length.  In practice we
+                             * never let count == length (see code). */
+    int32_t     length;     /* The physical size of the arrays hashes, keys
+                             * and values.  Must be prime. */
+
+    /* Rehashing thresholds */
+
+    int32_t     highWaterMark;  /* If count > highWaterMark, rehash */
+    int32_t     lowWaterMark;   /* If count < lowWaterMark, rehash */
+    float       highWaterRatio; /* 0..1; high water as a fraction of length */
+    float       lowWaterRatio;  /* 0..1; low water as a fraction of length */
+
+    int8_t      primeIndex;     /* Index into our prime table for length.
+                                 * length == PRIMES[primeIndex] */
+    UBool       allocated; /* Was this UHashtable allocated? */
+};
+typedef struct UHashtable UHashtable;
+
+U_CDECL_END
+
+/********************************************************************
+ * API
+ ********************************************************************/
+
+/**
+ * Initialize a new UHashtable.
+ * @param keyHash A pointer to the key hashing function.  Must not be
+ * NULL.
+ * @param keyComp A pointer to the function that compares keys.  Must
+ * not be NULL.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return A pointer to a UHashtable, or 0 if an error occurred.
+ * @see uhash_openSize
+ */
+U_CAPI UHashtable* U_EXPORT2
+uhash_open(UHashFunction *keyHash,
+           UKeyComparator *keyComp,
+           UValueComparator *valueComp,
+           UErrorCode *status);
+
+/**
+ * Initialize a new UHashtable with a given initial size.
+ * @param keyHash A pointer to the key hashing function.  Must not be
+ * NULL.
+ * @param keyComp A pointer to the function that compares keys.  Must
+ * not be NULL.
+ * @param size The initial capacity of this hash table.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return A pointer to a UHashtable, or 0 if an error occurred.
+ * @see uhash_open
+ */
+U_CAPI UHashtable* U_EXPORT2
+uhash_openSize(UHashFunction *keyHash,
+               UKeyComparator *keyComp,
+               UValueComparator *valueComp,
+               int32_t size,
+               UErrorCode *status);
+
+/**
+ * Initialize an existing UHashtable.
+ * @param keyHash A pointer to the key hashing function.  Must not be
+ * NULL.
+ * @param keyComp A pointer to the function that compares keys.  Must
+ * not be NULL.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return A pointer to a UHashtable, or 0 if an error occurred.
+ * @see uhash_openSize
+ */
+U_CAPI UHashtable* U_EXPORT2
+uhash_init(UHashtable *hash,
+           UHashFunction *keyHash,
+           UKeyComparator *keyComp,
+           UValueComparator *valueComp,
+           UErrorCode *status);
+
+/**
+ * Initialize an existing UHashtable.
+ * @param keyHash A pointer to the key hashing function.  Must not be
+ * NULL.
+ * @param keyComp A pointer to the function that compares keys.  Must
+ * not be NULL.
+ * @param size The initial capacity of this hash table.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return A pointer to a UHashtable, or 0 if an error occurred.
+ * @see uhash_openSize
+ */
+U_CAPI UHashtable* U_EXPORT2
+uhash_initSize(UHashtable *hash,
+               UHashFunction *keyHash,
+               UKeyComparator *keyComp,
+               UValueComparator *valueComp,
+               int32_t size,
+               UErrorCode *status);
+
+/**
+ * Close a UHashtable, releasing the memory used.
+ * @param hash The UHashtable to close. If hash is NULL no operation is performed.
+ */
+U_CAPI void U_EXPORT2
+uhash_close(UHashtable *hash);
+
+
+
+/**
+ * Set the function used to hash keys.
+ * @param hash The UHashtable to set
+ * @param fn the function to be used hash keys; must not be NULL
+ * @return the previous key hasher; non-NULL
+ */
+U_CAPI UHashFunction *U_EXPORT2
+uhash_setKeyHasher(UHashtable *hash, UHashFunction *fn);
+
+/**
+ * Set the function used to compare keys.  The default comparison is a
+ * void* pointer comparison.
+ * @param hash The UHashtable to set
+ * @param fn the function to be used compare keys; must not be NULL
+ * @return the previous key comparator; non-NULL
+ */
+U_CAPI UKeyComparator *U_EXPORT2
+uhash_setKeyComparator(UHashtable *hash, UKeyComparator *fn);
+
+/**
+ * Set the function used to compare values.  The default comparison is a
+ * void* pointer comparison.
+ * @param hash The UHashtable to set
+ * @param fn the function to be used compare keys; must not be NULL
+ * @return the previous key comparator; non-NULL
+ */
+U_CAPI UValueComparator *U_EXPORT2
+uhash_setValueComparator(UHashtable *hash, UValueComparator *fn);
+
+/**
+ * Set the function used to delete keys.  If this function pointer is
+ * NULL, this hashtable does not delete keys.  If it is non-NULL, this
+ * hashtable does delete keys.  This function should be set once
+ * before any elements are added to the hashtable and should not be
+ * changed thereafter.
+ * @param hash The UHashtable to set
+ * @param fn the function to be used delete keys, or NULL
+ * @return the previous key deleter; may be NULL
+ */
+U_CAPI UObjectDeleter *U_EXPORT2
+uhash_setKeyDeleter(UHashtable *hash, UObjectDeleter *fn);
+
+/**
+ * Set the function used to delete values.  If this function pointer
+ * is NULL, this hashtable does not delete values.  If it is non-NULL,
+ * this hashtable does delete values.  This function should be set
+ * once before any elements are added to the hashtable and should not
+ * be changed thereafter.
+ * @param hash The UHashtable to set
+ * @param fn the function to be used delete values, or NULL
+ * @return the previous value deleter; may be NULL
+ */
+U_CAPI UObjectDeleter *U_EXPORT2
+uhash_setValueDeleter(UHashtable *hash, UObjectDeleter *fn);
+
+/**
+ * Specify whether or not, and how, the hastable resizes itself.
+ * By default, tables grow but do not shrink (policy U_GROW).
+ * See enum UHashResizePolicy.
+ * @param hash The UHashtable to set
+ * @param policy The way the hashtable resizes itself, {U_GROW, U_GROW_AND_SHRINK, U_FIXED}
+ */
+U_CAPI void U_EXPORT2
+uhash_setResizePolicy(UHashtable *hash, enum UHashResizePolicy policy);
+
+/**
+ * Get the number of key-value pairs stored in a UHashtable.
+ * @param hash The UHashtable to query.
+ * @return The number of key-value pairs stored in hash.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_count(const UHashtable *hash);
+
+/**
+ * Put a (key=pointer, value=pointer) item in a UHashtable.  If the
+ * keyDeleter is non-NULL, then the hashtable owns 'key' after this
+ * call.  If the valueDeleter is non-NULL, then the hashtable owns
+ * 'value' after this call.  Storing a NULL value is the same as
+ * calling uhash_remove().
+ * @param hash The target UHashtable.
+ * @param key The key to store.
+ * @param value The value to store, may be NULL (see above).
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return The previous value, or NULL if none.
+ * @see uhash_get
+ */
+U_CAPI void* U_EXPORT2
+uhash_put(UHashtable *hash,
+          void *key,
+          void *value,
+          UErrorCode *status);
+
+/**
+ * Put a (key=integer, value=pointer) item in a UHashtable.
+ * keyDeleter must be NULL.  If the valueDeleter is non-NULL, then the
+ * hashtable owns 'value' after this call.  Storing a NULL value is
+ * the same as calling uhash_remove().
+ * @param hash The target UHashtable.
+ * @param key The integer key to store.
+ * @param value The value to store, may be NULL (see above).
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return The previous value, or NULL if none.
+ * @see uhash_get
+ */
+U_CAPI void* U_EXPORT2
+uhash_iput(UHashtable *hash,
+           int32_t key,
+           void* value,
+           UErrorCode *status);
+
+/**
+ * Put a (key=pointer, value=integer) item in a UHashtable.  If the
+ * keyDeleter is non-NULL, then the hashtable owns 'key' after this
+ * call.  valueDeleter must be NULL.  Storing a 0 value is the same as
+ * calling uhash_remove().
+ * @param hash The target UHashtable.
+ * @param key The key to store.
+ * @param value The integer value to store.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return The previous value, or 0 if none.
+ * @see uhash_get
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_puti(UHashtable *hash,
+           void* key,
+           int32_t value,
+           UErrorCode *status);
+
+/**
+ * Put a (key=integer, value=integer) item in a UHashtable.  If the
+ * keyDeleter is non-NULL, then the hashtable owns 'key' after this
+ * call.  valueDeleter must be NULL.  Storing a 0 value is the same as
+ * calling uhash_remove().
+ * @param hash The target UHashtable.
+ * @param key The key to store.
+ * @param value The integer value to store.
+ * @param status A pointer to an UErrorCode to receive any errors.
+ * @return The previous value, or 0 if none.
+ * @see uhash_get
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_iputi(UHashtable *hash,
+           int32_t key,
+           int32_t value,
+           UErrorCode *status);
+
+/**
+ * Retrieve a pointer value from a UHashtable using a pointer key,
+ * as previously stored by uhash_put().
+ * @param hash The target UHashtable.
+ * @param key A pointer key stored in a hashtable
+ * @return The requested item, or NULL if not found.
+ */
+U_CAPI void* U_EXPORT2
+uhash_get(const UHashtable *hash,
+          const void *key);
+
+/**
+ * Retrieve a pointer value from a UHashtable using a integer key,
+ * as previously stored by uhash_iput().
+ * @param hash The target UHashtable.
+ * @param key An integer key stored in a hashtable
+ * @return The requested item, or NULL if not found.
+ */
+U_CAPI void* U_EXPORT2
+uhash_iget(const UHashtable *hash,
+           int32_t key);
+
+/**
+ * Retrieve an integer value from a UHashtable using a pointer key,
+ * as previously stored by uhash_puti().
+ * @param hash The target UHashtable.
+ * @param key A pointer key stored in a hashtable
+ * @return The requested item, or 0 if not found.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_geti(const UHashtable *hash,
+           const void* key);
+/**
+ * Retrieve an integer value from a UHashtable using an integer key,
+ * as previously stored by uhash_iputi().
+ * @param hash The target UHashtable.
+ * @param key An integer key stored in a hashtable
+ * @return The requested item, or 0 if not found.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_igeti(const UHashtable *hash,
+           int32_t key);
+
+/**
+ * Remove an item from a UHashtable stored by uhash_put().
+ * @param hash The target UHashtable.
+ * @param key A key stored in a hashtable
+ * @return The item removed, or NULL if not found.
+ */
+U_CAPI void* U_EXPORT2
+uhash_remove(UHashtable *hash,
+             const void *key);
+
+/**
+ * Remove an item from a UHashtable stored by uhash_iput().
+ * @param hash The target UHashtable.
+ * @param key An integer key stored in a hashtable
+ * @return The item removed, or NULL if not found.
+ */
+U_CAPI void* U_EXPORT2
+uhash_iremove(UHashtable *hash,
+              int32_t key);
+
+/**
+ * Remove an item from a UHashtable stored by uhash_puti().
+ * @param hash The target UHashtable.
+ * @param key An key stored in a hashtable
+ * @return The item removed, or 0 if not found.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_removei(UHashtable *hash,
+              const void* key);
+
+/**
+ * Remove an item from a UHashtable stored by uhash_iputi().
+ * @param hash The target UHashtable.
+ * @param key An integer key stored in a hashtable
+ * @return The item removed, or 0 if not found.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_iremovei(UHashtable *hash,
+               int32_t key);
+
+/**
+ * Remove all items from a UHashtable.
+ * @param hash The target UHashtable.
+ */
+U_CAPI void U_EXPORT2
+uhash_removeAll(UHashtable *hash);
+
+/**
+ * Locate an element of a UHashtable.  The caller must not modify the
+ * returned object.  The primary use of this function is to obtain the
+ * stored key when it may not be identical to the search key.  For
+ * example, if the compare function is a case-insensitive string
+ * compare, then the hash key may be desired in order to obtain the
+ * canonical case corresponding to a search key.
+ * @param hash The target UHashtable.
+ * @param key A key stored in a hashtable
+ * @return a hash element, or NULL if the key is not found.
+ */
+U_CAPI const UHashElement* U_EXPORT2
+uhash_find(const UHashtable *hash, const void* key);
+
+/**
+ * \def UHASH_FIRST
+ * Constant for use with uhash_nextElement
+ * @see uhash_nextElement
+ */
+#define UHASH_FIRST (-1)
+
+/**
+ * Iterate through the elements of a UHashtable.  The caller must not
+ * modify the returned object.  However, uhash_removeElement() may be
+ * called during iteration to remove an element from the table.
+ * Iteration may safely be resumed afterwards.  If uhash_put() is
+ * called during iteration the iteration will then be out of sync and
+ * should be restarted.
+ * @param hash The target UHashtable.
+ * @param pos This should be set to UHASH_FIRST initially, and left untouched
+ * thereafter.
+ * @return a hash element, or NULL if no further key-value pairs
+ * exist in the table.
+ */
+U_CAPI const UHashElement* U_EXPORT2
+uhash_nextElement(const UHashtable *hash,
+                  int32_t *pos);
+
+/**
+ * Remove an element, returned by uhash_nextElement(), from the table.
+ * Iteration may be safely continued afterwards.
+ * @param hash The hashtable
+ * @param e The element, returned by uhash_nextElement(), to remove.
+ * Must not be NULL.  Must not be an empty or deleted element (as long
+ * as this was returned by uhash_nextElement() it will not be empty or
+ * deleted).  Note: Although this parameter is const, it will be
+ * modified.
+ * @return the value that was removed.
+ */
+U_CAPI void* U_EXPORT2
+uhash_removeElement(UHashtable *hash, const UHashElement* e);
+
+/********************************************************************
+ * UHashTok convenience
+ ********************************************************************/
+
+/**
+ * Return a UHashTok for an integer.
+ * @param i The given integer
+ * @return a UHashTok for an integer.
+ */
+/*U_CAPI UHashTok U_EXPORT2
+uhash_toki(int32_t i);*/
+
+/**
+ * Return a UHashTok for a pointer.
+ * @param p The given pointer
+ * @return a UHashTok for a pointer.
+ */
+/*U_CAPI UHashTok U_EXPORT2
+uhash_tokp(void* p);*/
+
+/********************************************************************
+ * UChar* and char* Support Functions
+ ********************************************************************/
+
+/**
+ * Generate a hash code for a null-terminated UChar* string.  If the
+ * string is not null-terminated do not use this function.  Use
+ * together with uhash_compareUChars.
+ * @param key The string (const UChar*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashUChars(const UHashTok key);
+
+/**
+ * Generate a hash code for a null-terminated char* string.  If the
+ * string is not null-terminated do not use this function.  Use
+ * together with uhash_compareChars.
+ * @param key The string (const char*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashChars(const UHashTok key);
+
+/**
+ * Generate a case-insensitive hash code for a null-terminated char*
+ * string.  If the string is not null-terminated do not use this
+ * function.  Use together with uhash_compareIChars.
+ * @param key The string (const char*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashIChars(const UHashTok key);
+
+/**
+ * Comparator for null-terminated UChar* strings.  Use together with
+ * uhash_hashUChars.
+ * @param key1 The string for comparison
+ * @param key2 The string for comparison
+ * @return true if key1 and key2 are equal, return false otherwise.
+ */
+U_CAPI UBool U_EXPORT2
+uhash_compareUChars(const UHashTok key1, const UHashTok key2);
+
+/**
+ * Comparator for null-terminated char* strings.  Use together with
+ * uhash_hashChars.
+ * @param key1 The string for comparison
+ * @param key2 The string for comparison
+ * @return true if key1 and key2 are equal, return false otherwise.
+ */
+U_CAPI UBool U_EXPORT2
+uhash_compareChars(const UHashTok key1, const UHashTok key2);
+
+/**
+ * Case-insensitive comparator for null-terminated char* strings.  Use
+ * together with uhash_hashIChars.
+ * @param key1 The string for comparison
+ * @param key2 The string for comparison
+ * @return true if key1 and key2 are equal, return false otherwise.
+ */
+U_CAPI UBool U_EXPORT2
+uhash_compareIChars(const UHashTok key1, const UHashTok key2);
+
+/********************************************************************
+ * UnicodeString Support Functions
+ ********************************************************************/
+
+/**
+ * Hash function for UnicodeString* keys.
+ * @param key The string (const char*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashUnicodeString(const UElement key);
+
+/**
+ * Hash function for UnicodeString* keys (case insensitive).
+ * Make sure to use together with uhash_compareCaselessUnicodeString.
+ * @param key The string (const char*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashCaselessUnicodeString(const UElement key);
+
+/********************************************************************
+ * int32_t Support Functions
+ ********************************************************************/
+
+/**
+ * Hash function for 32-bit integer keys.
+ * @param key The string (const char*) to hash.
+ * @return A hash code for the key.
+ */
+U_CAPI int32_t U_EXPORT2
+uhash_hashLong(const UHashTok key);
+
+/**
+ * Comparator function for 32-bit integer keys.
+ * @param key1 The integer for comparison
+ * @param Key2 The integer for comparison
+ * @return true if key1 and key2 are equal, return false otherwise
+ */
+U_CAPI UBool U_EXPORT2
+uhash_compareLong(const UHashTok key1, const UHashTok key2);
+
+/********************************************************************
+ * Other Support Functions
+ ********************************************************************/
+
+/**
+ * Deleter for Hashtable objects.
+ * @param obj The object to be deleted
+ */
+U_CAPI void U_EXPORT2
+uhash_deleteHashtable(void *obj);
+
+/* Use uprv_free() itself as a deleter for any key or value allocated using uprv_malloc. */
+
+/**
+ * Checks if the given hash tables are equal or not.
+ * @param hash1
+ * @param hash2
+ * @return true if the hashtables are equal and false if not.
+ */
+U_CAPI UBool U_EXPORT2
+uhash_equals(const UHashtable* hash1, const UHashtable* hash2);
+
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUHashtablePointer
+ * "Smart pointer" class, closes a UHashtable via uhash_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUHashtablePointer, UHashtable, uhash_close);
+
+U_NAMESPACE_END
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/uhash_us.cpp b/thirdparty/icu4c/common/uhash_us.cpp
new file mode 100644
index 0000000000..ef482c2746
--- /dev/null
+++ b/thirdparty/icu4c/common/uhash_us.cpp
@@ -0,0 +1,26 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 1997-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   03/22/00    aliu        Creation.
+*   07/06/01    aliu        Modified to support int32_t keys on
+*                           platforms with sizeof(void*) < 32.
+******************************************************************************
+*/
+
+#include "hash.h"
+
+/**
+ * Deleter for Hashtable objects.
+ */
+U_CAPI void U_EXPORT2
+uhash_deleteHashtable(void *obj) {
+    U_NAMESPACE_USE
+    delete (Hashtable*) obj;
+}
+
+//eof
diff --git a/thirdparty/icu4c/common/uidna.cpp b/thirdparty/icu4c/common/uidna.cpp
new file mode 100644
index 0000000000..ac2f9c3c8c
--- /dev/null
+++ b/thirdparty/icu4c/common/uidna.cpp
@@ -0,0 +1,921 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ *
+ *   Copyright (C) 2003-2014, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *******************************************************************************
+ *   file name:  uidna.cpp
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2003feb1
+ *   created by: Ram Viswanadha
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/uidna.h"
+#include "unicode/ustring.h"
+#include "unicode/usprep.h"
+#include "punycode.h"
+#include "ustr_imp.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "sprpimpl.h"
+
+/* it is official IDNA ACE Prefix is "xn--" */
+static const UChar ACE_PREFIX[] ={ 0x0078,0x006E,0x002d,0x002d } ;
+#define ACE_PREFIX_LENGTH 4
+
+#define MAX_LABEL_LENGTH 63
+/* The Max length of the labels should not be more than MAX_LABEL_LENGTH */
+#define MAX_LABEL_BUFFER_SIZE 100
+
+#define MAX_DOMAIN_NAME_LENGTH 255
+/* The Max length of the domain names should not be more than MAX_DOMAIN_NAME_LENGTH */
+#define MAX_IDN_BUFFER_SIZE   MAX_DOMAIN_NAME_LENGTH+1
+
+#define LOWER_CASE_DELTA 0x0020
+#define HYPHEN           0x002D
+#define FULL_STOP        0x002E
+#define CAPITAL_A        0x0041
+#define CAPITAL_Z        0x005A
+
+inline static UChar 
+toASCIILower(UChar ch){
+    if(CAPITAL_A <= ch && ch <= CAPITAL_Z){
+        return ch + LOWER_CASE_DELTA;
+    }
+    return ch;
+}
+
+inline static UBool 
+startsWithPrefix(const UChar* src , int32_t srcLength){
+    if(srcLength < ACE_PREFIX_LENGTH){
+        return FALSE;
+    }
+
+    for(int8_t i=0; i< ACE_PREFIX_LENGTH; i++){
+        if(toASCIILower(src[i]) != ACE_PREFIX[i]){
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+
+inline static int32_t
+compareCaseInsensitiveASCII(const UChar* s1, int32_t s1Len, 
+                            const UChar* s2, int32_t s2Len){
+    
+    int32_t minLength;
+    int32_t lengthResult;
+
+    // are we comparing different lengths?
+    if(s1Len != s2Len) {
+        if(s1Len < s2Len) {
+            minLength = s1Len;
+            lengthResult = -1;
+        } else {
+            minLength = s2Len;
+            lengthResult = 1;
+        }
+    } else {
+        // ok the lengths are equal
+        minLength = s1Len;
+        lengthResult = 0;
+    }
+
+    UChar c1,c2;
+    int32_t rc;
+
+    for(int32_t i =0;/* no condition */;i++) {
+
+        /* If we reach the ends of both strings then they match */
+        if(i == minLength) {
+            return lengthResult;
+        }
+        
+        c1 = s1[i];
+        c2 = s2[i];
+        
+        /* Case-insensitive comparison */
+        if(c1!=c2) {
+            rc=(int32_t)toASCIILower(c1)-(int32_t)toASCIILower(c2);
+            if(rc!=0) {
+                lengthResult=rc;
+                break;
+            }
+        }
+    }
+    return lengthResult;
+}
+
+
+/**
+ * Ascertain if the given code point is a label separator as 
+ * defined by the IDNA RFC
+ * 
+ * @param ch The code point to be ascertained
+ * @return true if the char is a label separator
+ * @stable ICU 2.8
+ */
+static inline UBool isLabelSeparator(UChar ch){
+    switch(ch){
+        case 0x002e:
+        case 0x3002:
+        case 0xFF0E:
+        case 0xFF61:
+            return TRUE;
+        default:
+            return FALSE;           
+    }
+}
+
+// returns the length of the label excluding the separator
+// if *limit == separator then the length returned does not include 
+// the separtor.
+static inline int32_t
+getNextSeparator(UChar *src, int32_t srcLength,
+                 UChar **limit, UBool *done){
+    if(srcLength == -1){
+        int32_t i;
+        for(i=0 ; ;i++){
+            if(src[i] == 0){
+                *limit = src + i; // point to null
+                *done = TRUE;
+                return i;
+            }
+            if(isLabelSeparator(src[i])){
+                *limit = src + (i+1); // go past the delimiter
+                return i;
+                
+            }
+        }
+    }else{
+        int32_t i;
+        for(i=0;i<srcLength;i++){
+            if(isLabelSeparator(src[i])){
+                *limit = src + (i+1); // go past the delimiter
+                return i;
+            }
+        }
+        // we have not found the delimiter
+        // if(i==srcLength)
+        *limit = src+srcLength;
+        *done = TRUE;
+
+        return i;
+    }
+}
+static inline UBool isLDHChar(UChar ch){
+    // high runner case
+    if(ch>0x007A){
+        return FALSE;
+    }
+    //[\\u002D \\u0030-\\u0039 \\u0041-\\u005A \\u0061-\\u007A]
+    if( (ch==0x002D) || 
+        (0x0030 <= ch && ch <= 0x0039) ||
+        (0x0041 <= ch && ch <= 0x005A) ||
+        (0x0061 <= ch && ch <= 0x007A)
+      ){
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static int32_t 
+_internal_toASCII(const UChar* src, int32_t srcLength, 
+                  UChar* dest, int32_t destCapacity,
+                  int32_t options,
+                  UStringPrepProfile* nameprep,
+                  UParseError* parseError,
+                  UErrorCode* status)
+{
+
+    // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
+    UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE];
+    //initialize pointers to stack buffers
+    UChar  *b1 = b1Stack, *b2 = b2Stack;
+    int32_t b1Len=0, b2Len, 
+            b1Capacity = MAX_LABEL_BUFFER_SIZE, 
+            b2Capacity = MAX_LABEL_BUFFER_SIZE ,
+            reqLength=0;
+
+    int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0;
+    UBool* caseFlags = NULL;
+    
+    // the source contains all ascii codepoints
+    UBool srcIsASCII  = TRUE;
+    // assume the source contains all LDH codepoints
+    UBool srcIsLDH = TRUE; 
+
+    int32_t j=0;
+
+    //get the options
+    UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
+
+    int32_t failPos = -1;
+    
+    if(srcLength == -1){
+        srcLength = u_strlen(src);
+    }
+    
+    if(srcLength > b1Capacity){
+        b1 = (UChar*) uprv_malloc(srcLength * U_SIZEOF_UCHAR);
+        if(b1==NULL){
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            goto CLEANUP;
+        }
+        b1Capacity = srcLength;
+    }
+
+    // step 1 
+    for( j=0;j<srcLength;j++){
+        if(src[j] > 0x7F){
+            srcIsASCII = FALSE;
+        }
+        b1[b1Len++] = src[j];
+    }
+    
+    // step 2 is performed only if the source contains non ASCII
+    if(srcIsASCII == FALSE){
+        
+        // step 2    
+        b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
+
+        if(*status == U_BUFFER_OVERFLOW_ERROR){
+            // redo processing of string
+            // we do not have enough room so grow the buffer
+            if(b1 != b1Stack){
+                uprv_free(b1);
+            }
+            b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
+            if(b1==NULL){
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                goto CLEANUP;
+            }
+
+            *status = U_ZERO_ERROR; // reset error
+            
+            b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
+        }
+    }
+    // error bail out
+    if(U_FAILURE(*status)){
+        goto CLEANUP;
+    }
+    if(b1Len == 0){
+        *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
+        goto CLEANUP;
+    }
+
+    // for step 3 & 4
+    srcIsASCII = TRUE;
+    for( j=0;j<b1Len;j++){
+        // check if output of usprep_prepare is all ASCII 
+        if(b1[j] > 0x7F){
+            srcIsASCII = FALSE;
+        }else if(isLDHChar(b1[j])==FALSE){  // if the char is in ASCII range verify that it is an LDH character
+            srcIsLDH = FALSE;
+            failPos = j;
+        }
+    }
+    if(useSTD3ASCIIRules == TRUE){
+        // verify 3a and 3b
+        // 3(a) Verify the absence of non-LDH ASCII code points; that is, the
+        //  absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F.
+        // 3(b) Verify the absence of leading and trailing hyphen-minus; that
+        //  is, the absence of U+002D at the beginning and end of the
+        //  sequence.
+        if( srcIsLDH == FALSE /* source at this point should not contain anyLDH characters */
+            || b1[0] ==  HYPHEN || b1[b1Len-1] == HYPHEN){
+            *status = U_IDNA_STD3_ASCII_RULES_ERROR;
+
+            /* populate the parseError struct */
+            if(srcIsLDH==FALSE){
+                // failPos is always set the index of failure
+                uprv_syntaxError(b1,failPos, b1Len,parseError);
+            }else if(b1[0] == HYPHEN){
+                // fail position is 0 
+                uprv_syntaxError(b1,0,b1Len,parseError);
+            }else{
+                // the last index in the source is always length-1
+                uprv_syntaxError(b1, (b1Len>0) ? b1Len-1 : b1Len, b1Len,parseError);
+            }
+
+            goto CLEANUP;
+        }
+    }
+    // Step 4: if the source is ASCII then proceed to step 8
+    if(srcIsASCII){
+        if(b1Len <= destCapacity){
+            u_memmove(dest, b1, b1Len);
+            reqLength = b1Len;
+        }else{
+            reqLength = b1Len;
+            goto CLEANUP;
+        }
+    }else{
+        // step 5 : verify the sequence does not begin with ACE prefix
+        if(!startsWithPrefix(b1,b1Len)){
+
+            //step 6: encode the sequence with punycode
+
+            // do not preserve the case flags for now!
+            // TODO: Preserve the case while implementing the RFE
+            // caseFlags = (UBool*) uprv_malloc(b1Len * sizeof(UBool));
+            // uprv_memset(caseFlags,TRUE,b1Len);
+
+            b2Len = u_strToPunycode(b1,b1Len,b2,b2Capacity,caseFlags, status);
+
+            if(*status == U_BUFFER_OVERFLOW_ERROR){
+                // redo processing of string
+                /* we do not have enough room so grow the buffer*/
+                b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR); 
+                if(b2 == NULL){
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    goto CLEANUP;
+                }
+
+                *status = U_ZERO_ERROR; // reset error
+                
+                b2Len = u_strToPunycode(b1,b1Len,b2,b2Len,caseFlags, status);
+            }
+            //error bail out
+            if(U_FAILURE(*status)){
+                goto CLEANUP;
+            }
+            // TODO : Reconsider while implementing the case preserve RFE
+            // convert all codepoints to lower case ASCII
+            // toASCIILower(b2,b2Len);
+            reqLength = b2Len+ACE_PREFIX_LENGTH;
+
+            if(reqLength > destCapacity){
+                *status = U_BUFFER_OVERFLOW_ERROR;
+                goto CLEANUP;
+            }
+            //Step 7: prepend the ACE prefix
+            u_memcpy(dest, ACE_PREFIX, ACE_PREFIX_LENGTH);
+            //Step 6: copy the contents in b2 into dest
+            u_memcpy(dest+ACE_PREFIX_LENGTH, b2, b2Len);
+
+        }else{
+            *status = U_IDNA_ACE_PREFIX_ERROR; 
+            //position of failure is 0
+            uprv_syntaxError(b1,0,b1Len,parseError);
+            goto CLEANUP;
+        }
+    }
+    // step 8: verify the length of label
+    if(reqLength > MAX_LABEL_LENGTH){
+        *status = U_IDNA_LABEL_TOO_LONG_ERROR;
+    }
+
+CLEANUP:
+    if(b1 != b1Stack){
+        uprv_free(b1);
+    }
+    if(b2 != b2Stack){
+        uprv_free(b2);
+    }
+    uprv_free(caseFlags);
+    
+    return u_terminateUChars(dest, destCapacity, reqLength, status);
+}
+
+static int32_t
+_internal_toUnicode(const UChar* src, int32_t srcLength,
+                    UChar* dest, int32_t destCapacity,
+                    int32_t options,
+                    UStringPrepProfile* nameprep,
+                    UParseError* parseError,
+                    UErrorCode* status)
+{
+
+    //get the options
+    //UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
+    int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0; 
+
+    // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
+    UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE], b3Stack[MAX_LABEL_BUFFER_SIZE];
+
+    //initialize pointers to stack buffers
+    UChar  *b1 = b1Stack, *b2 = b2Stack, *b1Prime=NULL, *b3=b3Stack;
+    int32_t b1Len = 0, b2Len, b1PrimeLen, b3Len,
+            b1Capacity = MAX_LABEL_BUFFER_SIZE, 
+            b2Capacity = MAX_LABEL_BUFFER_SIZE,
+            b3Capacity = MAX_LABEL_BUFFER_SIZE,
+            reqLength=0;
+
+    UBool* caseFlags = NULL;
+
+    UBool srcIsASCII = TRUE;
+    /*UBool srcIsLDH = TRUE;
+    int32_t failPos =0;*/
+
+    // step 1: find out if all the codepoints in src are ASCII  
+    if(srcLength==-1){
+        srcLength = 0;
+        for(;src[srcLength]!=0;){
+            if(src[srcLength]> 0x7f){
+                srcIsASCII = FALSE;
+            }/*else if(isLDHChar(src[srcLength])==FALSE){
+                // here we do not assemble surrogates
+                // since we know that LDH code points
+                // are in the ASCII range only
+                srcIsLDH = FALSE;
+                failPos = srcLength;
+            }*/
+            srcLength++;
+        }
+    }else if(srcLength > 0){
+        for(int32_t j=0; j<srcLength; j++){
+            if(src[j]> 0x7f){
+                srcIsASCII = FALSE;
+                break;
+            }/*else if(isLDHChar(src[j])==FALSE){
+                // here we do not assemble surrogates
+                // since we know that LDH code points
+                // are in the ASCII range only
+                srcIsLDH = FALSE;
+                failPos = j;
+            }*/
+        }
+    }else{
+        return 0;
+    }
+    
+    if(srcIsASCII == FALSE){
+        // step 2: process the string
+        b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
+        if(*status == U_BUFFER_OVERFLOW_ERROR){
+            // redo processing of string
+            /* we do not have enough room so grow the buffer*/
+            b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
+            if(b1==NULL){
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                goto CLEANUP;
+            }
+
+            *status = U_ZERO_ERROR; // reset error
+            
+            b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
+        }
+        //bail out on error
+        if(U_FAILURE(*status)){
+            goto CLEANUP;
+        }
+    }else{
+
+        //just point src to b1
+        b1 = (UChar*) src;
+        b1Len = srcLength;
+    }
+
+    // The RFC states that 
+    // <quote>
+    // ToUnicode never fails. If any step fails, then the original input
+    // is returned immediately in that step.
+    // </quote>
+
+    //step 3: verify ACE Prefix
+    if(startsWithPrefix(b1,b1Len)){
+
+        //step 4: Remove the ACE Prefix
+        b1Prime = b1 + ACE_PREFIX_LENGTH;
+        b1PrimeLen  = b1Len - ACE_PREFIX_LENGTH;
+
+        //step 5: Decode using punycode
+        b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Capacity, caseFlags,status);
+
+        if(*status == U_BUFFER_OVERFLOW_ERROR){
+            // redo processing of string
+            /* we do not have enough room so grow the buffer*/
+            b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
+            if(b2==NULL){
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                goto CLEANUP;
+            }
+
+            *status = U_ZERO_ERROR; // reset error
+
+            b2Len =  u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Len, caseFlags, status);
+        }
+
+
+        //step 6:Apply toASCII
+        b3Len = uidna_toASCII(b2, b2Len, b3, b3Capacity, options, parseError, status);
+
+        if(*status == U_BUFFER_OVERFLOW_ERROR){
+            // redo processing of string
+            /* we do not have enough room so grow the buffer*/
+            b3 = (UChar*) uprv_malloc(b3Len * U_SIZEOF_UCHAR);
+            if(b3==NULL){
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                goto CLEANUP;
+            }
+
+            *status = U_ZERO_ERROR; // reset error
+
+            b3Len =  uidna_toASCII(b2,b2Len,b3,b3Len,options,parseError, status);
+
+        }
+        //bail out on error
+        if(U_FAILURE(*status)){
+            goto CLEANUP;
+        }
+
+        //step 7: verify
+        if(compareCaseInsensitiveASCII(b1, b1Len, b3, b3Len) !=0){
+            // Cause the original to be returned.
+            *status = U_IDNA_VERIFICATION_ERROR;
+            goto CLEANUP;
+        }
+
+        //step 8: return output of step 5
+        reqLength = b2Len;
+        if(b2Len <= destCapacity) {
+            u_memmove(dest, b2, b2Len);
+        }
+    }
+    else{
+        // See the start of this if statement for why this is commented out.
+        // verify that STD3 ASCII rules are satisfied
+        /*if(useSTD3ASCIIRules == TRUE){
+            if( srcIsLDH == FALSE // source contains some non-LDH characters
+                || src[0] ==  HYPHEN || src[srcLength-1] == HYPHEN){
+                *status = U_IDNA_STD3_ASCII_RULES_ERROR;
+
+                // populate the parseError struct
+                if(srcIsLDH==FALSE){
+                    // failPos is always set the index of failure
+                    uprv_syntaxError(src,failPos, srcLength,parseError);
+                }else if(src[0] == HYPHEN){
+                    // fail position is 0 
+                    uprv_syntaxError(src,0,srcLength,parseError);
+                }else{
+                    // the last index in the source is always length-1
+                    uprv_syntaxError(src, (srcLength>0) ? srcLength-1 : srcLength, srcLength,parseError);
+                }
+
+                goto CLEANUP;
+            }
+        }*/
+        // just return the source
+        //copy the source to destination
+        if(srcLength <= destCapacity){
+            u_memmove(dest, src, srcLength);
+        }
+        reqLength = srcLength;
+    }
+
+
+CLEANUP:
+
+    if(b1 != b1Stack && b1!=src){
+        uprv_free(b1);
+    }
+    if(b2 != b2Stack){
+        uprv_free(b2);
+    }
+    uprv_free(caseFlags);
+
+    // The RFC states that 
+    // <quote>
+    // ToUnicode never fails. If any step fails, then the original input
+    // is returned immediately in that step.
+    // </quote>
+    // So if any step fails lets copy source to destination
+    if(U_FAILURE(*status)){
+        //copy the source to destination
+        if(dest && srcLength <= destCapacity){
+            // srcLength should have already been set earlier.
+            U_ASSERT(srcLength >= 0);
+            u_memmove(dest, src, srcLength);
+        }
+        reqLength = srcLength;
+        *status = U_ZERO_ERROR;
+    }
+
+    return u_terminateUChars(dest, destCapacity, reqLength, status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_toASCII(const UChar* src, int32_t srcLength, 
+              UChar* dest, int32_t destCapacity,
+              int32_t options,
+              UParseError* parseError,
+              UErrorCode* status){
+    
+    if(status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+    if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
+    
+    if(U_FAILURE(*status)){
+        return -1;
+    }
+    
+    int32_t retLen = _internal_toASCII(src, srcLength, dest, destCapacity, options, nameprep, parseError, status);
+    
+    /* close the profile*/
+    usprep_close(nameprep);
+    
+    return retLen;
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_toUnicode(const UChar* src, int32_t srcLength,
+                UChar* dest, int32_t destCapacity,
+                int32_t options,
+                UParseError* parseError,
+                UErrorCode* status){
+
+    if(status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+    if( (src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }  
+
+    UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
+    
+    if(U_FAILURE(*status)){
+        return -1;
+    }
+    
+    int32_t retLen = _internal_toUnicode(src, srcLength, dest, destCapacity, options, nameprep, parseError, status);
+
+    usprep_close(nameprep);
+    
+    return retLen;
+}
+
+
+U_CAPI int32_t U_EXPORT2
+uidna_IDNToASCII(  const UChar *src, int32_t srcLength,
+                   UChar* dest, int32_t destCapacity,
+                   int32_t options,
+                   UParseError *parseError,
+                   UErrorCode *status){
+
+    if(status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+    if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    int32_t reqLength = 0;
+
+    UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
+    
+    if(U_FAILURE(*status)){
+        return 0;
+    }
+
+    //initialize pointers 
+    UChar *delimiter = (UChar*)src;
+    UChar *labelStart = (UChar*)src;
+    UChar *currentDest = (UChar*) dest;
+    int32_t remainingLen = srcLength;
+    int32_t remainingDestCapacity = destCapacity;
+    int32_t labelLen = 0, labelReqLength = 0;
+    UBool done = FALSE;
+
+
+    for(;;){
+
+        labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
+        labelReqLength = 0;
+        if(!(labelLen==0 && done)){// make sure this is not a root label separator.
+        
+            labelReqLength = _internal_toASCII( labelStart, labelLen, 
+                                                currentDest, remainingDestCapacity, 
+                                                options, nameprep, 
+                                                parseError, status);
+    
+            if(*status == U_BUFFER_OVERFLOW_ERROR){
+                
+                *status = U_ZERO_ERROR; // reset error
+                remainingDestCapacity = 0;
+            }
+        }
+
+    
+        if(U_FAILURE(*status)){
+            break;
+        }
+        
+        reqLength +=labelReqLength;
+        // adjust the destination pointer
+        if(labelReqLength < remainingDestCapacity){
+            currentDest = currentDest + labelReqLength;
+            remainingDestCapacity -= labelReqLength;
+        }else{
+            // should never occur
+            remainingDestCapacity = 0;
+        }
+
+        if(done == TRUE){
+            break;
+        }
+
+        // add the label separator
+        if(remainingDestCapacity > 0){
+            *currentDest++ = FULL_STOP;
+            remainingDestCapacity--;
+        }
+        reqLength++;
+
+        labelStart = delimiter;
+        if(remainingLen >0 ){
+            remainingLen = (int32_t)(srcLength - (delimiter - src));
+        }
+
+    }
+
+    if(reqLength > MAX_DOMAIN_NAME_LENGTH){
+        *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
+    }
+
+    usprep_close(nameprep);
+    
+    return u_terminateUChars(dest, destCapacity, reqLength, status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_IDNToUnicode(  const UChar* src, int32_t srcLength,
+                     UChar* dest, int32_t destCapacity,
+                     int32_t options,
+                     UParseError* parseError,
+                     UErrorCode* status){
+    
+    if(status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+    if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    int32_t reqLength = 0;
+
+    UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
+    
+    if(U_FAILURE(*status)){
+        return 0;
+    }
+
+    //initialize pointers
+    UChar *delimiter = (UChar*)src;
+    UChar *labelStart = (UChar*)src;
+    UChar *currentDest = (UChar*) dest;
+    int32_t remainingLen = srcLength;
+    int32_t remainingDestCapacity = destCapacity;
+    int32_t labelLen = 0, labelReqLength = 0;
+    UBool done = FALSE;
+
+    for(;;){
+
+        labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
+        
+        // The RFC states that 
+        // <quote>
+        // ToUnicode never fails. If any step fails, then the original input
+        // is returned immediately in that step.
+        // </quote>
+        // _internal_toUnicode will copy the label.
+        /*if(labelLen==0 && done==FALSE){ 
+            *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
+            break;
+        }*/
+        
+        labelReqLength = _internal_toUnicode(labelStart, labelLen, 
+                                             currentDest, remainingDestCapacity, 
+                                             options, nameprep, 
+                                             parseError, status);
+
+        if(*status == U_BUFFER_OVERFLOW_ERROR){
+            *status = U_ZERO_ERROR; // reset error
+            remainingDestCapacity = 0;
+        }
+
+        if(U_FAILURE(*status)){
+            break;
+        }
+        
+        reqLength +=labelReqLength;
+        // adjust the destination pointer
+        if(labelReqLength < remainingDestCapacity){
+            currentDest = currentDest + labelReqLength;
+            remainingDestCapacity -= labelReqLength;
+        }else{
+            // should never occur
+            remainingDestCapacity = 0;
+        }
+
+        if(done == TRUE){
+            break;
+        }
+
+        // add the label separator
+        // Unlike the ToASCII operation we don't normalize the label separators
+        if(remainingDestCapacity > 0){
+            *currentDest++ = *(labelStart + labelLen);
+            remainingDestCapacity--;
+        }
+        reqLength++;
+
+        labelStart = delimiter;
+        if(remainingLen >0 ){
+            remainingLen = (int32_t)(srcLength - (delimiter - src));
+        }
+
+    }
+
+    if(reqLength > MAX_DOMAIN_NAME_LENGTH){
+        *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
+    }
+
+    usprep_close(nameprep);
+    
+    return u_terminateUChars(dest, destCapacity, reqLength, status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_compare(  const UChar *s1, int32_t length1,
+                const UChar *s2, int32_t length2,
+                int32_t options,
+                UErrorCode* status){
+
+    if(status == NULL || U_FAILURE(*status)){
+        return -1;
+    }
+
+    UChar b1Stack[MAX_IDN_BUFFER_SIZE], b2Stack[MAX_IDN_BUFFER_SIZE];
+    UChar *b1 = b1Stack, *b2 = b2Stack;
+    int32_t b1Len, b2Len, b1Capacity = MAX_IDN_BUFFER_SIZE, b2Capacity = MAX_IDN_BUFFER_SIZE;
+    int32_t result=-1;
+    
+    UParseError parseError;
+
+    b1Len = uidna_IDNToASCII(s1, length1, b1, b1Capacity, options, &parseError, status);
+    if(*status == U_BUFFER_OVERFLOW_ERROR){
+        // redo processing of string
+        b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
+        if(b1==NULL){
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            goto CLEANUP;
+        }
+
+        *status = U_ZERO_ERROR; // reset error
+        
+        b1Len = uidna_IDNToASCII(s1,length1,b1,b1Len, options, &parseError, status);
+        
+    }
+
+    b2Len = uidna_IDNToASCII(s2,length2, b2,b2Capacity, options, &parseError, status);
+    if(*status == U_BUFFER_OVERFLOW_ERROR){
+        // redo processing of string
+        b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
+        if(b2==NULL){
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            goto CLEANUP;
+        }
+
+        *status = U_ZERO_ERROR; // reset error
+        
+        b2Len = uidna_IDNToASCII(s2, length2, b2, b2Len, options, &parseError, status);
+        
+    }
+    // when toASCII is applied all label separators are replaced with FULL_STOP
+    result = compareCaseInsensitiveASCII(b1,b1Len,b2,b2Len);
+
+CLEANUP:
+    if(b1 != b1Stack){
+        uprv_free(b1);
+    }
+
+    if(b2 != b2Stack){
+        uprv_free(b2);
+    }
+
+    return result;
+}
+
+#endif /* #if !UCONFIG_NO_IDNA */
diff --git a/thirdparty/icu4c/common/uinit.cpp b/thirdparty/icu4c/common/uinit.cpp
new file mode 100644
index 0000000000..624431be02
--- /dev/null
+++ b/thirdparty/icu4c/common/uinit.cpp
@@ -0,0 +1,74 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2001-2015, International Business Machines
+*                Corporation and others. All Rights Reserved.
+******************************************************************************
+*   file name:  uinit.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001July05
+*   created by: George Rhoten
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/icuplug.h"
+#include "unicode/uclean.h"
+#include "cmemory.h"
+#include "icuplugimp.h"
+#include "ucln_cmn.h"
+#include "ucnv_io.h"
+#include "umutex.h"
+#include "utracimp.h"
+
+U_NAMESPACE_BEGIN
+
+static UInitOnce gICUInitOnce = U_INITONCE_INITIALIZER;
+
+static UBool U_CALLCONV uinit_cleanup() {
+    gICUInitOnce.reset();
+    return TRUE;
+}
+
+static void U_CALLCONV
+initData(UErrorCode &status)
+{
+#if UCONFIG_ENABLE_PLUGINS
+    /* initialize plugins */
+    uplug_init(&status);
+#endif
+
+#if !UCONFIG_NO_CONVERSION
+    /*
+     * 2005-may-02
+     *
+     * ICU4C 3.4 (jitterbug 4497) hardcodes the data for Unicode character
+     * properties for APIs that want to be fast.
+     * Therefore, we need not load them here nor check for errors.
+     * Instead, we load the converter alias table to see if any ICU data
+     * is available.
+     * Users should really open the service objects they need and check
+     * for errors there, to make sure that the actual items they need are
+     * available.
+     */
+    ucnv_io_countKnownConverters(&status);
+#endif
+    ucln_common_registerCleanup(UCLN_COMMON_UINIT, uinit_cleanup);
+}
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+/*
+ * ICU Initialization Function. Need not be called.
+ */
+U_CAPI void U_EXPORT2
+u_init(UErrorCode *status) {
+    UTRACE_ENTRY_OC(UTRACE_U_INIT);
+    umtx_initOnce(gICUInitOnce, &initData, *status);
+    UTRACE_EXIT_STATUS(*status);
+}
diff --git a/thirdparty/icu4c/common/uinvchar.cpp b/thirdparty/icu4c/common/uinvchar.cpp
new file mode 100644
index 0000000000..52b8906568
--- /dev/null
+++ b/thirdparty/icu4c/common/uinvchar.cpp
@@ -0,0 +1,627 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2010, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uinvchar.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2004sep14
+*   created by: Markus W. Scherer
+*
+*   Functions for handling invariant characters, moved here from putil.c
+*   for better modularization.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "udataswp.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "uinvchar.h"
+
+/* invariant-character handling --------------------------------------------- */
+
+/*
+ * These maps for ASCII to/from EBCDIC map invariant characters (see utypes.h)
+ * appropriately for most EBCDIC codepages.
+ *
+ * They currently also map most other ASCII graphic characters,
+ * appropriately for codepages 37 and 1047.
+ * Exceptions: The characters for []^ have different codes in 37 & 1047.
+ * Both versions are mapped to ASCII.
+ *
+ *    ASCII 37 1047
+ * [     5B BA   AD
+ * ]     5D BB   BD
+ * ^     5E B0   5F
+ *
+ * There are no mappings for variant characters from Unicode to EBCDIC.
+ *
+ * Currently, C0 control codes are also included in these maps.
+ * Exceptions: S/390 Open Edition swaps LF and NEL codes compared with other
+ * EBCDIC platforms; both codes (15 and 25) are mapped to ASCII LF (0A),
+ * but there is no mapping for ASCII LF back to EBCDIC.
+ *
+ *    ASCII EBCDIC S/390-OE
+ * LF    0A     25       15
+ * NEL   85     15       25
+ *
+ * The maps below explicitly exclude the variant
+ * control and graphical characters that are in ASCII-based
+ * codepages at 0x80 and above.
+ * "No mapping" is expressed by mapping to a 00 byte.
+ *
+ * These tables do not establish a converter or a codepage.
+ */
+
+static const uint8_t asciiFromEbcdic[256]={
+    0x00, 0x01, 0x02, 0x03, 0x00, 0x09, 0x00, 0x7f, 0x00, 0x00, 0x00, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+    0x10, 0x11, 0x12, 0x13, 0x00, 0x0a, 0x08, 0x00, 0x18, 0x19, 0x00, 0x00, 0x1c, 0x1d, 0x1e, 0x1f,
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x17, 0x1b, 0x00, 0x00, 0x00, 0x00, 0x00, 0x05, 0x06, 0x07,
+    0x00, 0x00, 0x16, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x14, 0x15, 0x00, 0x1a,
+
+    0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x2e, 0x3c, 0x28, 0x2b, 0x7c,
+    0x26, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x21, 0x24, 0x2a, 0x29, 0x3b, 0x5e,
+    0x2d, 0x2f, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x2c, 0x25, 0x5f, 0x3e, 0x3f,
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x60, 0x3a, 0x23, 0x40, 0x27, 0x3d, 0x22,
+
+    0x00, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x00, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x00, 0x7e, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x00, 0x00, 0x00, 0x5b, 0x00, 0x00,
+    0x5e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x5b, 0x5d, 0x00, 0x5d, 0x00, 0x00,
+
+    0x7b, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x7d, 0x4a, 0x4b, 0x4c, 0x4d, 0x4e, 0x4f, 0x50, 0x51, 0x52, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x5c, 0x00, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59, 0x5a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00
+};
+
+static const uint8_t ebcdicFromAscii[256]={
+    0x00, 0x01, 0x02, 0x03, 0x37, 0x2d, 0x2e, 0x2f, 0x16, 0x05, 0x00, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+    0x10, 0x11, 0x12, 0x13, 0x3c, 0x3d, 0x32, 0x26, 0x18, 0x19, 0x3f, 0x27, 0x1c, 0x1d, 0x1e, 0x1f,
+    0x40, 0x00, 0x7f, 0x00, 0x00, 0x6c, 0x50, 0x7d, 0x4d, 0x5d, 0x5c, 0x4e, 0x6b, 0x60, 0x4b, 0x61,
+    0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0x7a, 0x5e, 0x4c, 0x7e, 0x6e, 0x6f,
+
+    0x00, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, 0xc8, 0xc9, 0xd1, 0xd2, 0xd3, 0xd4, 0xd5, 0xd6,
+    0xd7, 0xd8, 0xd9, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9, 0x00, 0x00, 0x00, 0x00, 0x6d,
+    0x00, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96,
+    0x97, 0x98, 0x99, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0x00, 0x00, 0x00, 0x00, 0x07,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+};
+
+/* Same as asciiFromEbcdic[] except maps all letters to lowercase. */
+static const uint8_t lowercaseAsciiFromEbcdic[256]={
+    0x00, 0x01, 0x02, 0x03, 0x00, 0x09, 0x00, 0x7f, 0x00, 0x00, 0x00, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+    0x10, 0x11, 0x12, 0x13, 0x00, 0x0a, 0x08, 0x00, 0x18, 0x19, 0x00, 0x00, 0x1c, 0x1d, 0x1e, 0x1f,
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x17, 0x1b, 0x00, 0x00, 0x00, 0x00, 0x00, 0x05, 0x06, 0x07,
+    0x00, 0x00, 0x16, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x14, 0x15, 0x00, 0x1a,
+
+    0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x2e, 0x3c, 0x28, 0x2b, 0x7c,
+    0x26, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x21, 0x24, 0x2a, 0x29, 0x3b, 0x5e,
+    0x2d, 0x2f, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x2c, 0x25, 0x5f, 0x3e, 0x3f,
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x60, 0x3a, 0x23, 0x40, 0x27, 0x3d, 0x22,
+
+    0x00, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x00, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x00, 0x7e, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x00, 0x00, 0x00, 0x5b, 0x00, 0x00,
+    0x5e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x5b, 0x5d, 0x00, 0x5d, 0x00, 0x00,
+
+    0x7b, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x7d, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x7c, 0x00, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+    0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00
+};
+
+/*
+ * Bit sets indicating which characters of the ASCII repertoire
+ * (by ASCII/Unicode code) are "invariant".
+ * See utypes.h for more details.
+ *
+ * As invariant are considered the characters of the ASCII repertoire except
+ * for the following:
+ * 21  '!' <exclamation mark>
+ * 23  '#' <number sign>
+ * 24  '$' <dollar sign>
+ *
+ * 40  '@' <commercial at>
+ *
+ * 5b  '[' <left bracket>
+ * 5c  '\' <backslash>
+ * 5d  ']' <right bracket>
+ * 5e  '^' <circumflex>
+ *
+ * 60  '`' <grave accent>
+ *
+ * 7b  '{' <left brace>
+ * 7c  '|' <vertical line>
+ * 7d  '}' <right brace>
+ * 7e  '~' <tilde>
+ */
+static const uint32_t invariantChars[4]={
+    0xfffffbff, /* 00..1f but not 0a */
+    0xffffffe5, /* 20..3f but not 21 23 24 */
+    0x87fffffe, /* 40..5f but not 40 5b..5e */
+    0x87fffffe  /* 60..7f but not 60 7b..7e */
+};
+
+/*
+ * test unsigned types (or values known to be non-negative) for invariant characters,
+ * tests ASCII-family character values
+ */
+#define UCHAR_IS_INVARIANT(c) (((c)<=0x7f) && (invariantChars[(c)>>5]&((uint32_t)1<<((c)&0x1f)))!=0)
+
+/* test signed types for invariant characters, adds test for positive values */
+#define SCHAR_IS_INVARIANT(c) ((0<=(c)) && UCHAR_IS_INVARIANT(c))
+
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#define CHAR_TO_UCHAR(c) c
+#define UCHAR_TO_CHAR(c) c
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#define CHAR_TO_UCHAR(u) asciiFromEbcdic[u]
+#define UCHAR_TO_CHAR(u) ebcdicFromAscii[u]
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+
+
+U_CAPI void U_EXPORT2
+u_charsToUChars(const char *cs, UChar *us, int32_t length) {
+    UChar u;
+    uint8_t c;
+
+    /*
+     * Allow the entire ASCII repertoire to be mapped _to_ Unicode.
+     * For EBCDIC systems, this works for characters with codes from
+     * codepages 37 and 1047 or compatible.
+     */
+    while(length>0) {
+        c=(uint8_t)(*cs++);
+        u=(UChar)CHAR_TO_UCHAR(c);
+        U_ASSERT((u!=0 || c==0)); /* only invariant chars converted? */
+        *us++=u;
+        --length;
+    }
+}
+
+U_CAPI void U_EXPORT2
+u_UCharsToChars(const UChar *us, char *cs, int32_t length) {
+    UChar u;
+
+    while(length>0) {
+        u=*us++;
+        if(!UCHAR_IS_INVARIANT(u)) {
+            U_ASSERT(FALSE); /* Variant characters were used. These are not portable in ICU. */
+            u=0;
+        }
+        *cs++=(char)UCHAR_TO_CHAR(u);
+        --length;
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_isInvariantString(const char *s, int32_t length) {
+    uint8_t c;
+
+    for(;;) {
+        if(length<0) {
+            /* NUL-terminated */
+            c=(uint8_t)*s++;
+            if(c==0) {
+                break;
+            }
+        } else {
+            /* count length */
+            if(length==0) {
+                break;
+            }
+            --length;
+            c=(uint8_t)*s++;
+            if(c==0) {
+                continue; /* NUL is invariant */
+            }
+        }
+        /* c!=0 now, one branch below checks c==0 for variant characters */
+
+        /*
+         * no assertions here because these functions are legitimately called
+         * for strings with variant characters
+         */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+        if(!UCHAR_IS_INVARIANT(c)) {
+            return FALSE; /* found a variant char */
+        }
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+        c=CHAR_TO_UCHAR(c);
+        if(c==0 || !UCHAR_IS_INVARIANT(c)) {
+            return FALSE; /* found a variant char */
+        }
+#else
+#   error U_CHARSET_FAMILY is not valid
+#endif
+    }
+    return TRUE;
+}
+
+U_CAPI UBool U_EXPORT2
+uprv_isInvariantUString(const UChar *s, int32_t length) {
+    UChar c;
+
+    for(;;) {
+        if(length<0) {
+            /* NUL-terminated */
+            c=*s++;
+            if(c==0) {
+                break;
+            }
+        } else {
+            /* count length */
+            if(length==0) {
+                break;
+            }
+            --length;
+            c=*s++;
+        }
+
+        /*
+         * no assertions here because these functions are legitimately called
+         * for strings with variant characters
+         */
+        if(!UCHAR_IS_INVARIANT(c)) {
+            return FALSE; /* found a variant char */
+        }
+    }
+    return TRUE;
+}
+
+/* UDataSwapFn implementations used in udataswp.c ------- */
+
+/* convert ASCII to EBCDIC and verify that all characters are invariant */
+U_CAPI int32_t U_EXPORT2
+uprv_ebcdicFromAscii(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode) {
+    const uint8_t *s;
+    uint8_t *t;
+    uint8_t c;
+
+    int32_t count;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    s=(const uint8_t *)inData;
+    t=(uint8_t *)outData;
+    count=length;
+    while(count>0) {
+        c=*s++;
+        if(!UCHAR_IS_INVARIANT(c)) {
+            udata_printError(ds, "uprv_ebcdicFromAscii() string[%d] contains a variant character in position %d\n",
+                             length, length-count);
+            *pErrorCode=U_INVALID_CHAR_FOUND;
+            return 0;
+        }
+        *t++=ebcdicFromAscii[c];
+        --count;
+    }
+
+    return length;
+}
+
+/* this function only checks and copies ASCII strings without conversion */
+U_CFUNC int32_t
+uprv_copyAscii(const UDataSwapper *ds,
+               const void *inData, int32_t length, void *outData,
+               UErrorCode *pErrorCode) {
+    const uint8_t *s;
+    uint8_t c;
+
+    int32_t count;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and checking */
+    s=(const uint8_t *)inData;
+    count=length;
+    while(count>0) {
+        c=*s++;
+        if(!UCHAR_IS_INVARIANT(c)) {
+            udata_printError(ds, "uprv_copyFromAscii() string[%d] contains a variant character in position %d\n",
+                             length, length-count);
+            *pErrorCode=U_INVALID_CHAR_FOUND;
+            return 0;
+        }
+        --count;
+    }
+
+    if(length>0 && inData!=outData) {
+        uprv_memcpy(outData, inData, length);
+    }
+
+    return length;
+}
+
+/* convert EBCDIC to ASCII and verify that all characters are invariant */
+U_CFUNC int32_t
+uprv_asciiFromEbcdic(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode) {
+    const uint8_t *s;
+    uint8_t *t;
+    uint8_t c;
+
+    int32_t count;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 ||  (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    s=(const uint8_t *)inData;
+    t=(uint8_t *)outData;
+    count=length;
+    while(count>0) {
+        c=*s++;
+        if(c!=0 && ((c=asciiFromEbcdic[c])==0 || !UCHAR_IS_INVARIANT(c))) {
+            udata_printError(ds, "uprv_asciiFromEbcdic() string[%d] contains a variant character in position %d\n",
+                             length, length-count);
+            *pErrorCode=U_INVALID_CHAR_FOUND;
+            return 0;
+        }
+        *t++=c;
+        --count;
+    }
+
+    return length;
+}
+
+/* this function only checks and copies EBCDIC strings without conversion */
+U_CFUNC int32_t
+uprv_copyEbcdic(const UDataSwapper *ds,
+                const void *inData, int32_t length, void *outData,
+                UErrorCode *pErrorCode) {
+    const uint8_t *s;
+    uint8_t c;
+
+    int32_t count;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || length<0 || (length>0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and checking */
+    s=(const uint8_t *)inData;
+    count=length;
+    while(count>0) {
+        c=*s++;
+        if(c!=0 && ((c=asciiFromEbcdic[c])==0 || !UCHAR_IS_INVARIANT(c))) {
+            udata_printError(ds, "uprv_copyEbcdic() string[%] contains a variant character in position %d\n",
+                             length, length-count);
+            *pErrorCode=U_INVALID_CHAR_FOUND;
+            return 0;
+        }
+        --count;
+    }
+
+    if(length>0 && inData!=outData) {
+        uprv_memcpy(outData, inData, length);
+    }
+
+    return length;
+}
+
+U_CFUNC UBool
+uprv_isEbcdicAtSign(char c) {
+    static const uint8_t ebcdicAtSigns[] = {
+        0x7C, 0x44, 0x66, 0x80, 0xAC, 0xAE, 0xAF, 0xB5, 0xEC, 0xEF, 0x00 };
+    return c != 0 && uprv_strchr((const char *)ebcdicAtSigns, c) != nullptr;
+}
+
+/* compare invariant strings; variant characters compare less than others and unlike each other */
+U_CFUNC int32_t
+uprv_compareInvAscii(const UDataSwapper *ds,
+                     const char *outString, int32_t outLength,
+                     const UChar *localString, int32_t localLength) {
+    (void)ds;
+    int32_t minLength;
+    UChar32 c1, c2;
+    uint8_t c;
+
+    if(outString==NULL || outLength<-1 || localString==NULL || localLength<-1) {
+        return 0;
+    }
+
+    if(outLength<0) {
+        outLength=(int32_t)uprv_strlen(outString);
+    }
+    if(localLength<0) {
+        localLength=u_strlen(localString);
+    }
+
+    minLength= outLength<localLength ? outLength : localLength;
+
+    while(minLength>0) {
+        c=(uint8_t)*outString++;
+        if(UCHAR_IS_INVARIANT(c)) {
+            c1=c;
+        } else {
+            c1=-1;
+        }
+
+        c2=*localString++;
+        if(!UCHAR_IS_INVARIANT(c2)) {
+            c2=-2;
+        }
+
+        if((c1-=c2)!=0) {
+            return c1;
+        }
+
+        --minLength;
+    }
+
+    /* strings start with same prefix, compare lengths */
+    return outLength-localLength;
+}
+
+U_CFUNC int32_t
+uprv_compareInvEbcdic(const UDataSwapper *ds,
+                      const char *outString, int32_t outLength,
+                      const UChar *localString, int32_t localLength) {
+    (void)ds;
+    int32_t minLength;
+    UChar32 c1, c2;
+    uint8_t c;
+
+    if(outString==NULL || outLength<-1 || localString==NULL || localLength<-1) {
+        return 0;
+    }
+
+    if(outLength<0) {
+        outLength=(int32_t)uprv_strlen(outString);
+    }
+    if(localLength<0) {
+        localLength=u_strlen(localString);
+    }
+
+    minLength= outLength<localLength ? outLength : localLength;
+
+    while(minLength>0) {
+        c=(uint8_t)*outString++;
+        if(c==0) {
+            c1=0;
+        } else if((c1=asciiFromEbcdic[c])!=0 && UCHAR_IS_INVARIANT(c1)) {
+            /* c1 is set */
+        } else {
+            c1=-1;
+        }
+
+        c2=*localString++;
+        if(!UCHAR_IS_INVARIANT(c2)) {
+            c2=-2;
+        }
+
+        if((c1-=c2)!=0) {
+            return c1;
+        }
+
+        --minLength;
+    }
+
+    /* strings start with same prefix, compare lengths */
+    return outLength-localLength;
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_compareInvEbcdicAsAscii(const char *s1, const char *s2) {
+    int32_t c1, c2;
+
+    for(;; ++s1, ++s2) {
+        c1=(uint8_t)*s1;
+        c2=(uint8_t)*s2;
+        if(c1!=c2) {
+            if(c1!=0 && ((c1=asciiFromEbcdic[c1])==0 || !UCHAR_IS_INVARIANT(c1))) {
+                c1=-(int32_t)(uint8_t)*s1;
+            }
+            if(c2!=0 && ((c2=asciiFromEbcdic[c2])==0 || !UCHAR_IS_INVARIANT(c2))) {
+                c2=-(int32_t)(uint8_t)*s2;
+            }
+            return c1-c2;
+        } else if(c1==0) {
+            return 0;
+        }
+    }
+}
+
+U_CAPI char U_EXPORT2
+uprv_ebcdicToAscii(char c) {
+    return (char)asciiFromEbcdic[(uint8_t)c];
+}
+
+U_CAPI char U_EXPORT2
+uprv_ebcdicToLowercaseAscii(char c) {
+    return (char)lowercaseAsciiFromEbcdic[(uint8_t)c];
+}
+
+U_CAPI uint8_t* U_EXPORT2
+uprv_aestrncpy(uint8_t *dst, const uint8_t *src, int32_t n)
+{
+  uint8_t *orig_dst = dst;
+
+  if(n==-1) { 
+    n = static_cast<int32_t>(uprv_strlen((const char*)src)+1); /* copy NUL */
+  }
+  /* copy non-null */
+  while(*src && n>0) {
+    *(dst++) = asciiFromEbcdic[*(src++)];
+    n--;
+  }
+  /* pad */
+  while(n>0) {
+    *(dst++) = 0;
+    n--;
+  }
+  return orig_dst;
+}
+
+U_CAPI uint8_t* U_EXPORT2
+uprv_eastrncpy(uint8_t *dst, const uint8_t *src, int32_t n)
+{
+  uint8_t *orig_dst = dst;
+
+  if(n==-1) { 
+    n = static_cast<int32_t>(uprv_strlen((const char*)src)+1); /* copy NUL */
+  }
+  /* copy non-null */
+  while(*src && n>0) {
+    char ch = ebcdicFromAscii[*(src++)];
+    if(ch == 0) {
+      ch = ebcdicFromAscii[0x3f]; /* questionmark (subchar) */
+    }
+    *(dst++) = ch;
+    n--;
+  }
+  /* pad */
+  while(n>0) {
+    *(dst++) = 0;
+    n--;
+  }
+  return orig_dst;
+}
+
diff --git a/thirdparty/icu4c/common/uinvchar.h b/thirdparty/icu4c/common/uinvchar.h
new file mode 100644
index 0000000000..9b7a9bd114
--- /dev/null
+++ b/thirdparty/icu4c/common/uinvchar.h
@@ -0,0 +1,219 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uinvchar.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2004sep14
+*   created by: Markus W. Scherer
+*
+*   Definitions for handling invariant characters, moved here from putil.c
+*   for better modularization.
+*/
+
+#ifndef __UINVCHAR_H__
+#define __UINVCHAR_H__
+
+#include "unicode/utypes.h"
+#ifdef __cplusplus
+#include "unicode/unistr.h"
+#endif
+
+/**
+ * Check if a char string only contains invariant characters.
+ * See utypes.h for details.
+ *
+ * @param s Input string pointer.
+ * @param length Length of the string, can be -1 if NUL-terminated.
+ * @return true if s contains only invariant characters.
+ *
+ * @internal (ICU 2.8)
+ */
+U_CAPI UBool U_EXPORT2
+uprv_isInvariantString(const char *s, int32_t length);
+
+/**
+ * Check if a Unicode string only contains invariant characters.
+ * See utypes.h for details.
+ *
+ * @param s Input string pointer.
+ * @param length Length of the string, can be -1 if NUL-terminated.
+ * @return true if s contains only invariant characters.
+ *
+ * @internal (ICU 2.8)
+ */
+U_CAPI UBool U_EXPORT2
+uprv_isInvariantUString(const UChar *s, int32_t length);
+
+/**
+ * \def U_UPPER_ORDINAL
+ * Get the ordinal number of an uppercase invariant character
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define U_UPPER_ORDINAL(x) ((x)-'A')
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define U_UPPER_ORDINAL(x) (((x) < 'J') ? ((x)-'A') : \
+                              (((x) < 'S') ? ((x)-'J'+9) : \
+                               ((x)-'S'+18)))
+#else
+#   error Unknown charset family!
+#endif
+
+#ifdef __cplusplus
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Like U_UPPER_ORDINAL(x) but with validation.
+ * Returns 0..25 for A..Z else a value outside 0..25.
+ */
+inline int32_t uprv_upperOrdinal(int32_t c) {
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+    return c - 'A';
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    // EBCDIC: A-Z (26 letters) is split into three ranges A-I (9 letters), J-R (9), S-Z (8).
+    // https://en.wikipedia.org/wiki/EBCDIC_037#Codepage_layout
+    if (c <= 'I') { return c - 'A'; }  // A-I --> 0-8
+    if (c < 'J') { return -1; }
+    if (c <= 'R') { return c - 'J' + 9; }  // J-R --> 9..17
+    if (c < 'S') { return -1; }
+    return c - 'S' + 18;  // S-Z --> 18..25
+#else
+#   error Unknown charset family!
+#endif
+}
+
+// Like U_UPPER_ORDINAL(x) but for lowercase and with validation.
+// Returns 0..25 for a..z else a value outside 0..25.
+inline int32_t uprv_lowerOrdinal(int32_t c) {
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+    return c - 'a';
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+    // EBCDIC: a-z (26 letters) is split into three ranges a-i (9 letters), j-r (9), s-z (8).
+    // https://en.wikipedia.org/wiki/EBCDIC_037#Codepage_layout
+    if (c <= 'i') { return c - 'a'; }  // a-i --> 0-8
+    if (c < 'j') { return -1; }
+    if (c <= 'r') { return c - 'j' + 9; }  // j-r --> 9..17
+    if (c < 's') { return -1; }
+    return c - 's' + 18;  // s-z --> 18..25
+#else
+#   error Unknown charset family!
+#endif
+}
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Returns true if c == '@' is possible.
+ * The @ sign is variant, and the @ sign used on one
+ * EBCDIC machine won't be compiled the same way on other EBCDIC based machines.
+ * @internal
+ */
+U_CFUNC UBool
+uprv_isEbcdicAtSign(char c);
+
+/**
+ * \def uprv_isAtSign
+ * Returns true if c == '@' is possible.
+ * For ASCII, checks for exactly '@'. For EBCDIC, calls uprv_isEbcdicAtSign().
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_isAtSign(c) ((c)=='@')
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_isAtSign(c) uprv_isEbcdicAtSign(c)
+#else
+#   error Unknown charset family!
+#endif
+
+/**
+ * Compare two EBCDIC invariant-character strings in ASCII order.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_compareInvEbcdicAsAscii(const char *s1, const char *s2);
+
+/**
+ * \def uprv_compareInvCharsAsAscii
+ * Compare two invariant-character strings in ASCII order.
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_compareInvCharsAsAscii(s1, s2) uprv_strcmp(s1, s2)
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_compareInvCharsAsAscii(s1, s2) uprv_compareInvEbcdicAsAscii(s1, s2)
+#else
+#   error Unknown charset family!
+#endif
+
+/**
+ * Converts an EBCDIC invariant character to ASCII.
+ * @internal
+ */
+U_CAPI char U_EXPORT2
+uprv_ebcdicToAscii(char c);
+
+/**
+ * \def uprv_invCharToAscii
+ * Converts an invariant character to ASCII.
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_invCharToAscii(c) (c)
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_invCharToAscii(c) uprv_ebcdicToAscii(c)
+#else
+#   error Unknown charset family!
+#endif
+
+/**
+ * Converts an EBCDIC invariant character to lowercase ASCII.
+ * @internal
+ */
+U_CAPI char U_EXPORT2
+uprv_ebcdicToLowercaseAscii(char c);
+
+/**
+ * \def uprv_invCharToLowercaseAscii
+ * Converts an invariant character to lowercase ASCII.
+ * @internal
+ */
+#if U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define uprv_invCharToLowercaseAscii uprv_asciitolower
+#elif U_CHARSET_FAMILY==U_EBCDIC_FAMILY
+#   define uprv_invCharToLowercaseAscii uprv_ebcdicToLowercaseAscii
+#else
+#   error Unknown charset family!
+#endif
+
+/**
+ * Copy EBCDIC to ASCII
+ * @internal
+ * @see uprv_strncpy
+ */
+U_CAPI uint8_t* U_EXPORT2
+uprv_aestrncpy(uint8_t *dst, const uint8_t *src, int32_t n);
+
+
+/**
+ * Copy ASCII to EBCDIC
+ * @internal
+ * @see uprv_strncpy
+ */
+U_CAPI uint8_t* U_EXPORT2
+uprv_eastrncpy(uint8_t *dst, const uint8_t *src, int32_t n);
+
+
+
+#endif
diff --git a/thirdparty/icu4c/common/uiter.cpp b/thirdparty/icu4c/common/uiter.cpp
new file mode 100644
index 0000000000..b9252d81c2
--- /dev/null
+++ b/thirdparty/icu4c/common/uiter.cpp
@@ -0,0 +1,1108 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uiter.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jan18
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/chariter.h"
+#include "unicode/rep.h"
+#include "unicode/uiter.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cstring.h"
+
+U_NAMESPACE_USE
+
+#define IS_EVEN(n) (((n)&1)==0)
+#define IS_POINTER_EVEN(p) IS_EVEN((size_t)p)
+
+U_CDECL_BEGIN
+
+/* No-Op UCharIterator implementation for illegal input --------------------- */
+
+static int32_t U_CALLCONV
+noopGetIndex(UCharIterator * /*iter*/, UCharIteratorOrigin /*origin*/) {
+    return 0;
+}
+
+static int32_t U_CALLCONV
+noopMove(UCharIterator * /*iter*/, int32_t /*delta*/, UCharIteratorOrigin /*origin*/) {
+    return 0;
+}
+
+static UBool U_CALLCONV
+noopHasNext(UCharIterator * /*iter*/) {
+    return FALSE;
+}
+
+static UChar32 U_CALLCONV
+noopCurrent(UCharIterator * /*iter*/) {
+    return U_SENTINEL;
+}
+
+static uint32_t U_CALLCONV
+noopGetState(const UCharIterator * /*iter*/) {
+    return UITER_NO_STATE;
+}
+
+static void U_CALLCONV
+noopSetState(UCharIterator * /*iter*/, uint32_t /*state*/, UErrorCode *pErrorCode) {
+    *pErrorCode=U_UNSUPPORTED_ERROR;
+}
+
+static const UCharIterator noopIterator={
+    0, 0, 0, 0, 0, 0,
+    noopGetIndex,
+    noopMove,
+    noopHasNext,
+    noopHasNext,
+    noopCurrent,
+    noopCurrent,
+    noopCurrent,
+    NULL,
+    noopGetState,
+    noopSetState
+};
+
+/* UCharIterator implementation for simple strings -------------------------- */
+
+/*
+ * This is an implementation of a code unit (UChar) iterator
+ * for UChar * strings.
+ *
+ * The UCharIterator.context field holds a pointer to the string.
+ */
+
+static int32_t U_CALLCONV
+stringIteratorGetIndex(UCharIterator *iter, UCharIteratorOrigin origin) {
+    switch(origin) {
+    case UITER_ZERO:
+        return 0;
+    case UITER_START:
+        return iter->start;
+    case UITER_CURRENT:
+        return iter->index;
+    case UITER_LIMIT:
+        return iter->limit;
+    case UITER_LENGTH:
+        return iter->length;
+    default:
+        /* not a valid origin */
+        /* Should never get here! */
+        return -1;
+    }
+}
+
+static int32_t U_CALLCONV
+stringIteratorMove(UCharIterator *iter, int32_t delta, UCharIteratorOrigin origin) {
+    int32_t pos;
+
+    switch(origin) {
+    case UITER_ZERO:
+        pos=delta;
+        break;
+    case UITER_START:
+        pos=iter->start+delta;
+        break;
+    case UITER_CURRENT:
+        pos=iter->index+delta;
+        break;
+    case UITER_LIMIT:
+        pos=iter->limit+delta;
+        break;
+    case UITER_LENGTH:
+        pos=iter->length+delta;
+        break;
+    default:
+        return -1;  /* Error */
+    }
+
+    if(pos<iter->start) {
+        pos=iter->start;
+    } else if(pos>iter->limit) {
+        pos=iter->limit;
+    }
+
+    return iter->index=pos;
+}
+
+static UBool U_CALLCONV
+stringIteratorHasNext(UCharIterator *iter) {
+    return iter->index<iter->limit;
+}
+
+static UBool U_CALLCONV
+stringIteratorHasPrevious(UCharIterator *iter) {
+    return iter->index>iter->start;
+}
+
+static UChar32 U_CALLCONV
+stringIteratorCurrent(UCharIterator *iter) {
+    if(iter->index<iter->limit) {
+        return ((const UChar *)(iter->context))[iter->index];
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+stringIteratorNext(UCharIterator *iter) {
+    if(iter->index<iter->limit) {
+        return ((const UChar *)(iter->context))[iter->index++];
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+stringIteratorPrevious(UCharIterator *iter) {
+    if(iter->index>iter->start) {
+        return ((const UChar *)(iter->context))[--iter->index];
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static uint32_t U_CALLCONV
+stringIteratorGetState(const UCharIterator *iter) {
+    return (uint32_t)iter->index;
+}
+
+static void U_CALLCONV
+stringIteratorSetState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        /* do nothing */
+    } else if(iter==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else if((int32_t)state<iter->start || iter->limit<(int32_t)state) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    } else {
+        iter->index=(int32_t)state;
+    }
+}
+
+static const UCharIterator stringIterator={
+    0, 0, 0, 0, 0, 0,
+    stringIteratorGetIndex,
+    stringIteratorMove,
+    stringIteratorHasNext,
+    stringIteratorHasPrevious,
+    stringIteratorCurrent,
+    stringIteratorNext,
+    stringIteratorPrevious,
+    NULL,
+    stringIteratorGetState,
+    stringIteratorSetState
+};
+
+U_CAPI void U_EXPORT2
+uiter_setString(UCharIterator *iter, const UChar *s, int32_t length) {
+    if(iter!=0) {
+        if(s!=0 && length>=-1) {
+            *iter=stringIterator;
+            iter->context=s;
+            if(length>=0) {
+                iter->length=length;
+            } else {
+                iter->length=u_strlen(s);
+            }
+            iter->limit=iter->length;
+        } else {
+            *iter=noopIterator;
+        }
+    }
+}
+
+/* UCharIterator implementation for UTF-16BE strings ------------------------ */
+
+/*
+ * This is an implementation of a code unit (UChar) iterator
+ * for UTF-16BE strings, i.e., strings in byte-vectors where
+ * each UChar is stored as a big-endian pair of bytes.
+ *
+ * The UCharIterator.context field holds a pointer to the string.
+ * Everything works just like with a normal UChar iterator (uiter_setString),
+ * except that UChars are assembled from byte pairs.
+ */
+
+/* internal helper function */
+static inline UChar32
+utf16BEIteratorGet(UCharIterator *iter, int32_t index) {
+    const uint8_t *p=(const uint8_t *)iter->context;
+    return ((UChar)p[2*index]<<8)|(UChar)p[2*index+1];
+}
+
+static UChar32 U_CALLCONV
+utf16BEIteratorCurrent(UCharIterator *iter) {
+    int32_t index;
+
+    if((index=iter->index)<iter->limit) {
+        return utf16BEIteratorGet(iter, index);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+utf16BEIteratorNext(UCharIterator *iter) {
+    int32_t index;
+
+    if((index=iter->index)<iter->limit) {
+        iter->index=index+1;
+        return utf16BEIteratorGet(iter, index);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+utf16BEIteratorPrevious(UCharIterator *iter) {
+    int32_t index;
+
+    if((index=iter->index)>iter->start) {
+        iter->index=--index;
+        return utf16BEIteratorGet(iter, index);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static const UCharIterator utf16BEIterator={
+    0, 0, 0, 0, 0, 0,
+    stringIteratorGetIndex,
+    stringIteratorMove,
+    stringIteratorHasNext,
+    stringIteratorHasPrevious,
+    utf16BEIteratorCurrent,
+    utf16BEIteratorNext,
+    utf16BEIteratorPrevious,
+    NULL,
+    stringIteratorGetState,
+    stringIteratorSetState
+};
+
+/*
+ * Count the number of UChars in a UTF-16BE string before a terminating UChar NUL,
+ * i.e., before a pair of 0 bytes where the first 0 byte is at an even
+ * offset from s.
+ */
+static int32_t
+utf16BE_strlen(const char *s) {
+    if(IS_POINTER_EVEN(s)) {
+        /*
+         * even-aligned, call u_strlen(s)
+         * we are probably on a little-endian machine, but searching for UChar NUL
+         * does not care about endianness
+         */
+        return u_strlen((const UChar *)s);
+    } else {
+        /* odd-aligned, search for pair of 0 bytes */
+        const char *p=s;
+
+        while(!(*p==0 && p[1]==0)) {
+            p+=2;
+        }
+        return (int32_t)((p-s)/2);
+    }
+}
+
+U_CAPI void U_EXPORT2
+uiter_setUTF16BE(UCharIterator *iter, const char *s, int32_t length) {
+    if(iter!=NULL) {
+        /* allow only even-length strings (the input length counts bytes) */
+        if(s!=NULL && (length==-1 || (length>=0 && IS_EVEN(length)))) {
+            /* length/=2, except that >>=1 also works for -1 (-1/2==0, -1>>1==-1) */
+            length>>=1;
+
+            if(U_IS_BIG_ENDIAN && IS_POINTER_EVEN(s)) {
+                /* big-endian machine and 2-aligned UTF-16BE string: use normal UChar iterator */
+                uiter_setString(iter, (const UChar *)s, length);
+                return;
+            }
+
+            *iter=utf16BEIterator;
+            iter->context=s;
+            if(length>=0) {
+                iter->length=length;
+            } else {
+                iter->length=utf16BE_strlen(s);
+            }
+            iter->limit=iter->length;
+        } else {
+            *iter=noopIterator;
+        }
+    }
+}
+
+/* UCharIterator wrapper around CharacterIterator --------------------------- */
+
+/*
+ * This is wrapper code around a C++ CharacterIterator to
+ * look like a C UCharIterator.
+ *
+ * The UCharIterator.context field holds a pointer to the CharacterIterator.
+ */
+
+static int32_t U_CALLCONV
+characterIteratorGetIndex(UCharIterator *iter, UCharIteratorOrigin origin) {
+    switch(origin) {
+    case UITER_ZERO:
+        return 0;
+    case UITER_START:
+        return ((CharacterIterator *)(iter->context))->startIndex();
+    case UITER_CURRENT:
+        return ((CharacterIterator *)(iter->context))->getIndex();
+    case UITER_LIMIT:
+        return ((CharacterIterator *)(iter->context))->endIndex();
+    case UITER_LENGTH:
+        return ((CharacterIterator *)(iter->context))->getLength();
+    default:
+        /* not a valid origin */
+        /* Should never get here! */
+        return -1;
+    }
+}
+
+static int32_t U_CALLCONV
+characterIteratorMove(UCharIterator *iter, int32_t delta, UCharIteratorOrigin origin) {
+    switch(origin) {
+    case UITER_ZERO:
+        ((CharacterIterator *)(iter->context))->setIndex(delta);
+        return ((CharacterIterator *)(iter->context))->getIndex();
+    case UITER_START:
+    case UITER_CURRENT:
+    case UITER_LIMIT:
+        return ((CharacterIterator *)(iter->context))->move(delta, (CharacterIterator::EOrigin)origin);
+    case UITER_LENGTH:
+        ((CharacterIterator *)(iter->context))->setIndex(((CharacterIterator *)(iter->context))->getLength()+delta);
+        return ((CharacterIterator *)(iter->context))->getIndex();
+    default:
+        /* not a valid origin */
+        /* Should never get here! */
+        return -1;
+    }
+}
+
+static UBool U_CALLCONV
+characterIteratorHasNext(UCharIterator *iter) {
+    return ((CharacterIterator *)(iter->context))->hasNext();
+}
+
+static UBool U_CALLCONV
+characterIteratorHasPrevious(UCharIterator *iter) {
+    return ((CharacterIterator *)(iter->context))->hasPrevious();
+}
+
+static UChar32 U_CALLCONV
+characterIteratorCurrent(UCharIterator *iter) {
+    UChar32 c;
+
+    c=((CharacterIterator *)(iter->context))->current();
+    if(c!=0xffff || ((CharacterIterator *)(iter->context))->hasNext()) {
+        return c;
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+characterIteratorNext(UCharIterator *iter) {
+    if(((CharacterIterator *)(iter->context))->hasNext()) {
+        return ((CharacterIterator *)(iter->context))->nextPostInc();
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+characterIteratorPrevious(UCharIterator *iter) {
+    if(((CharacterIterator *)(iter->context))->hasPrevious()) {
+        return ((CharacterIterator *)(iter->context))->previous();
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static uint32_t U_CALLCONV
+characterIteratorGetState(const UCharIterator *iter) {
+    return ((CharacterIterator *)(iter->context))->getIndex();
+}
+
+static void U_CALLCONV
+characterIteratorSetState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        /* do nothing */
+    } else if(iter==NULL || iter->context==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else if((int32_t)state<((CharacterIterator *)(iter->context))->startIndex() || ((CharacterIterator *)(iter->context))->endIndex()<(int32_t)state) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    } else {
+        ((CharacterIterator *)(iter->context))->setIndex((int32_t)state);
+    }
+}
+
+static const UCharIterator characterIteratorWrapper={
+    0, 0, 0, 0, 0, 0,
+    characterIteratorGetIndex,
+    characterIteratorMove,
+    characterIteratorHasNext,
+    characterIteratorHasPrevious,
+    characterIteratorCurrent,
+    characterIteratorNext,
+    characterIteratorPrevious,
+    NULL,
+    characterIteratorGetState,
+    characterIteratorSetState
+};
+
+U_CAPI void U_EXPORT2
+uiter_setCharacterIterator(UCharIterator *iter, CharacterIterator *charIter) {
+    if(iter!=0) {
+        if(charIter!=0) {
+            *iter=characterIteratorWrapper;
+            iter->context=charIter;
+        } else {
+            *iter=noopIterator;
+        }
+    }
+}
+
+/* UCharIterator wrapper around Replaceable --------------------------------- */
+
+/*
+ * This is an implementation of a code unit (UChar) iterator
+ * based on a Replaceable object.
+ *
+ * The UCharIterator.context field holds a pointer to the Replaceable.
+ * UCharIterator.length and UCharIterator.index hold Replaceable.length()
+ * and the iteration index.
+ */
+
+static UChar32 U_CALLCONV
+replaceableIteratorCurrent(UCharIterator *iter) {
+    if(iter->index<iter->limit) {
+        return ((Replaceable *)(iter->context))->charAt(iter->index);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+replaceableIteratorNext(UCharIterator *iter) {
+    if(iter->index<iter->limit) {
+        return ((Replaceable *)(iter->context))->charAt(iter->index++);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+replaceableIteratorPrevious(UCharIterator *iter) {
+    if(iter->index>iter->start) {
+        return ((Replaceable *)(iter->context))->charAt(--iter->index);
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static const UCharIterator replaceableIterator={
+    0, 0, 0, 0, 0, 0,
+    stringIteratorGetIndex,
+    stringIteratorMove,
+    stringIteratorHasNext,
+    stringIteratorHasPrevious,
+    replaceableIteratorCurrent,
+    replaceableIteratorNext,
+    replaceableIteratorPrevious,
+    NULL,
+    stringIteratorGetState,
+    stringIteratorSetState
+};
+
+U_CAPI void U_EXPORT2
+uiter_setReplaceable(UCharIterator *iter, const Replaceable *rep) {
+    if(iter!=0) {
+        if(rep!=0) {
+            *iter=replaceableIterator;
+            iter->context=rep;
+            iter->limit=iter->length=rep->length();
+        } else {
+            *iter=noopIterator;
+        }
+    }
+}
+
+/* UCharIterator implementation for UTF-8 strings --------------------------- */
+
+/*
+ * Possible, probably necessary only for an implementation for arbitrary
+ * converters:
+ * Maintain a buffer (ring buffer?) for a piece of converted 16-bit text.
+ * This would require to turn reservedFn into a close function and
+ * to introduce a uiter_close(iter).
+ */
+
+#define UITER_CNV_CAPACITY 16
+
+/*
+ * Minimal implementation:
+ * Maintain a single-UChar buffer for an additional surrogate.
+ * The caller must not modify start and limit because they are used internally.
+ *
+ * Use UCharIterator fields as follows:
+ *   context        pointer to UTF-8 string
+ *   length         UTF-16 length of the string; -1 until lazy evaluation
+ *   start          current UTF-8 index
+ *   index          current UTF-16 index; may be -1="unknown" after setState()
+ *   limit          UTF-8 length of the string
+ *   reservedField  supplementary code point
+ *
+ * Since UCharIterator delivers 16-bit code units, the iteration can be
+ * currently in the middle of the byte sequence for a supplementary code point.
+ * In this case, reservedField will contain that code point and start will
+ * point to after the corresponding byte sequence. The UTF-16 index will be
+ * one less than what it would otherwise be corresponding to the UTF-8 index.
+ * Otherwise, reservedField will be 0.
+ */
+
+/*
+ * Possible optimization for NUL-terminated UTF-8 and UTF-16 strings:
+ * Add implementations that do not call strlen() for iteration but check for NUL.
+ */
+
+static int32_t U_CALLCONV
+utf8IteratorGetIndex(UCharIterator *iter, UCharIteratorOrigin origin) {
+    switch(origin) {
+    case UITER_ZERO:
+    case UITER_START:
+        return 0;
+    case UITER_CURRENT:
+        if(iter->index<0) {
+            /* the current UTF-16 index is unknown after setState(), count from the beginning */
+            const uint8_t *s;
+            UChar32 c;
+            int32_t i, limit, index;
+
+            s=(const uint8_t *)iter->context;
+            i=index=0;
+            limit=iter->start; /* count up to the UTF-8 index */
+            while(i<limit) {
+                U8_NEXT_OR_FFFD(s, i, limit, c);
+                index+=U16_LENGTH(c);
+            }
+
+            iter->start=i; /* just in case setState() did not get us to a code point boundary */
+            if(i==iter->limit) {
+                iter->length=index; /* in case it was <0 or wrong */
+            }
+            if(iter->reservedField!=0) {
+                --index; /* we are in the middle of a supplementary code point */
+            }
+            iter->index=index;
+        }
+        return iter->index;
+    case UITER_LIMIT:
+    case UITER_LENGTH:
+        if(iter->length<0) {
+            const uint8_t *s;
+            UChar32 c;
+            int32_t i, limit, length;
+
+            s=(const uint8_t *)iter->context;
+            if(iter->index<0) {
+                /*
+                 * the current UTF-16 index is unknown after setState(),
+                 * we must first count from the beginning to here
+                 */
+                i=length=0;
+                limit=iter->start;
+
+                /* count from the beginning to the current index */
+                while(i<limit) {
+                    U8_NEXT_OR_FFFD(s, i, limit, c);
+                    length+=U16_LENGTH(c);
+                }
+
+                /* assume i==limit==iter->start, set the UTF-16 index */
+                iter->start=i; /* just in case setState() did not get us to a code point boundary */
+                iter->index= iter->reservedField!=0 ? length-1 : length;
+            } else {
+                i=iter->start;
+                length=iter->index;
+                if(iter->reservedField!=0) {
+                    ++length;
+                }
+            }
+
+            /* count from the current index to the end */
+            limit=iter->limit;
+            while(i<limit) {
+                U8_NEXT_OR_FFFD(s, i, limit, c);
+                length+=U16_LENGTH(c);
+            }
+            iter->length=length;
+        }
+        return iter->length;
+    default:
+        /* not a valid origin */
+        /* Should never get here! */
+        return -1;
+    }
+}
+
+static int32_t U_CALLCONV
+utf8IteratorMove(UCharIterator *iter, int32_t delta, UCharIteratorOrigin origin) {
+    const uint8_t *s;
+    UChar32 c;
+    int32_t pos; /* requested UTF-16 index */
+    int32_t i; /* UTF-8 index */
+    UBool havePos;
+
+    /* calculate the requested UTF-16 index */
+    switch(origin) {
+    case UITER_ZERO:
+    case UITER_START:
+        pos=delta;
+        havePos=TRUE;
+        /* iter->index<0 (unknown) is possible */
+        break;
+    case UITER_CURRENT:
+        if(iter->index>=0) {
+            pos=iter->index+delta;
+            havePos=TRUE;
+        } else {
+            /* the current UTF-16 index is unknown after setState(), use only delta */
+            pos=0;
+            havePos=FALSE;
+        }
+        break;
+    case UITER_LIMIT:
+    case UITER_LENGTH:
+        if(iter->length>=0) {
+            pos=iter->length+delta;
+            havePos=TRUE;
+        } else {
+            /* pin to the end, avoid counting the length */
+            iter->index=-1;
+            iter->start=iter->limit;
+            iter->reservedField=0;
+            if(delta>=0) {
+                return UITER_UNKNOWN_INDEX;
+            } else {
+                /* the current UTF-16 index is unknown, use only delta */
+                pos=0;
+                havePos=FALSE;
+            }
+        }
+        break;
+    default:
+        return -1;  /* Error */
+    }
+
+    if(havePos) {
+        /* shortcuts: pinning to the edges of the string */
+        if(pos<=0) {
+            iter->index=iter->start=iter->reservedField=0;
+            return 0;
+        } else if(iter->length>=0 && pos>=iter->length) {
+            iter->index=iter->length;
+            iter->start=iter->limit;
+            iter->reservedField=0;
+            return iter->index;
+        }
+
+        /* minimize the number of U8_NEXT/PREV operations */
+        if(iter->index<0 || pos<iter->index/2) {
+            /* go forward from the start instead of backward from the current index */
+            iter->index=iter->start=iter->reservedField=0;
+        } else if(iter->length>=0 && (iter->length-pos)<(pos-iter->index)) {
+            /*
+             * if we have the UTF-16 index and length and the new position is
+             * closer to the end than the current index,
+             * then go backward from the end instead of forward from the current index
+             */
+            iter->index=iter->length;
+            iter->start=iter->limit;
+            iter->reservedField=0;
+        }
+
+        delta=pos-iter->index;
+        if(delta==0) {
+            return iter->index; /* nothing to do */
+        }
+    } else {
+        /* move relative to unknown UTF-16 index */
+        if(delta==0) {
+            return UITER_UNKNOWN_INDEX; /* nothing to do */
+        } else if(-delta>=iter->start) {
+            /* moving backwards by more UChars than there are UTF-8 bytes, pin to 0 */
+            iter->index=iter->start=iter->reservedField=0;
+            return 0;
+        } else if(delta>=(iter->limit-iter->start)) {
+            /* moving forward by more UChars than the remaining UTF-8 bytes, pin to the end */
+            iter->index=iter->length; /* may or may not be <0 (unknown) */
+            iter->start=iter->limit;
+            iter->reservedField=0;
+            return iter->index>=0 ? iter->index : (int32_t)UITER_UNKNOWN_INDEX;
+        }
+    }
+
+    /* delta!=0 */
+
+    /* move towards the requested position, pin to the edges of the string */
+    s=(const uint8_t *)iter->context;
+    pos=iter->index; /* could be <0 (unknown) */
+    i=iter->start;
+    if(delta>0) {
+        /* go forward */
+        int32_t limit=iter->limit;
+        if(iter->reservedField!=0) {
+            iter->reservedField=0;
+            ++pos;
+            --delta;
+        }
+        while(delta>0 && i<limit) {
+            U8_NEXT_OR_FFFD(s, i, limit, c);
+            if(c<=0xffff) {
+                ++pos;
+                --delta;
+            } else if(delta>=2) {
+                pos+=2;
+                delta-=2;
+            } else /* delta==1 */ {
+                /* stop in the middle of a supplementary code point */
+                iter->reservedField=c;
+                ++pos;
+                break; /* delta=0; */
+            }
+        }
+        if(i==limit) {
+            if(iter->length<0 && iter->index>=0) {
+                iter->length= iter->reservedField==0 ? pos : pos+1;
+            } else if(iter->index<0 && iter->length>=0) {
+                iter->index= iter->reservedField==0 ? iter->length : iter->length-1;
+            }
+        }
+    } else /* delta<0 */ {
+        /* go backward */
+        if(iter->reservedField!=0) {
+            iter->reservedField=0;
+            i-=4; /* we stayed behind the supplementary code point; go before it now */
+            --pos;
+            ++delta;
+        }
+        while(delta<0 && i>0) {
+            U8_PREV_OR_FFFD(s, 0, i, c);
+            if(c<=0xffff) {
+                --pos;
+                ++delta;
+            } else if(delta<=-2) {
+                pos-=2;
+                delta+=2;
+            } else /* delta==-1 */ {
+                /* stop in the middle of a supplementary code point */
+                i+=4; /* back to behind this supplementary code point for consistent state */
+                iter->reservedField=c;
+                --pos;
+                break; /* delta=0; */
+            }
+        }
+    }
+
+    iter->start=i;
+    if(iter->index>=0) {
+        return iter->index=pos;
+    } else {
+        /* we started with index<0 (unknown) so pos is bogus */
+        if(i<=1) {
+            return iter->index=i; /* reached the beginning */
+        } else {
+            /* we still don't know the UTF-16 index */
+            return UITER_UNKNOWN_INDEX;
+        }
+    }
+}
+
+static UBool U_CALLCONV
+utf8IteratorHasNext(UCharIterator *iter) {
+    return iter->start<iter->limit || iter->reservedField!=0;
+}
+
+static UBool U_CALLCONV
+utf8IteratorHasPrevious(UCharIterator *iter) {
+    return iter->start>0;
+}
+
+static UChar32 U_CALLCONV
+utf8IteratorCurrent(UCharIterator *iter) {
+    if(iter->reservedField!=0) {
+        return U16_TRAIL(iter->reservedField);
+    } else if(iter->start<iter->limit) {
+        const uint8_t *s=(const uint8_t *)iter->context;
+        UChar32 c;
+        int32_t i=iter->start;
+
+        U8_NEXT_OR_FFFD(s, i, iter->limit, c);
+        if(c<=0xffff) {
+            return c;
+        } else {
+            return U16_LEAD(c);
+        }
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+utf8IteratorNext(UCharIterator *iter) {
+    int32_t index;
+
+    if(iter->reservedField!=0) {
+        UChar trail=U16_TRAIL(iter->reservedField);
+        iter->reservedField=0;
+        if((index=iter->index)>=0) {
+            iter->index=index+1;
+        }
+        return trail;
+    } else if(iter->start<iter->limit) {
+        const uint8_t *s=(const uint8_t *)iter->context;
+        UChar32 c;
+
+        U8_NEXT_OR_FFFD(s, iter->start, iter->limit, c);
+        if((index=iter->index)>=0) {
+            iter->index=++index;
+            if(iter->length<0 && iter->start==iter->limit) {
+                iter->length= c<=0xffff ? index : index+1;
+            }
+        } else if(iter->start==iter->limit && iter->length>=0) {
+            iter->index= c<=0xffff ? iter->length : iter->length-1;
+        }
+        if(c<=0xffff) {
+            return c;
+        } else {
+            iter->reservedField=c;
+            return U16_LEAD(c);
+        }
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static UChar32 U_CALLCONV
+utf8IteratorPrevious(UCharIterator *iter) {
+    int32_t index;
+
+    if(iter->reservedField!=0) {
+        UChar lead=U16_LEAD(iter->reservedField);
+        iter->reservedField=0;
+        iter->start-=4; /* we stayed behind the supplementary code point; go before it now */
+        if((index=iter->index)>0) {
+            iter->index=index-1;
+        }
+        return lead;
+    } else if(iter->start>0) {
+        const uint8_t *s=(const uint8_t *)iter->context;
+        UChar32 c;
+
+        U8_PREV_OR_FFFD(s, 0, iter->start, c);
+        if((index=iter->index)>0) {
+            iter->index=index-1;
+        } else if(iter->start<=1) {
+            iter->index= c<=0xffff ? iter->start : iter->start+1;
+        }
+        if(c<=0xffff) {
+            return c;
+        } else {
+            iter->start+=4; /* back to behind this supplementary code point for consistent state */
+            iter->reservedField=c;
+            return U16_TRAIL(c);
+        }
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+static uint32_t U_CALLCONV
+utf8IteratorGetState(const UCharIterator *iter) {
+    uint32_t state=(uint32_t)(iter->start<<1);
+    if(iter->reservedField!=0) {
+        state|=1;
+    }
+    return state;
+}
+
+static void U_CALLCONV
+utf8IteratorSetState(UCharIterator *iter,
+                     uint32_t state,
+                     UErrorCode *pErrorCode)
+{
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        /* do nothing */
+    } else if(iter==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else if(state==utf8IteratorGetState(iter)) {
+        /* setting to the current state: no-op */
+    } else {
+        int32_t index=(int32_t)(state>>1); /* UTF-8 index */
+        state&=1; /* 1 if in surrogate pair, must be index>=4 */
+
+        if((state==0 ? index<0 : index<4) || iter->limit<index) {
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        } else {
+            iter->start=index; /* restore UTF-8 byte index */
+            if(index<=1) {
+                iter->index=index;
+            } else {
+                iter->index=-1; /* unknown UTF-16 index */
+            }
+            if(state==0) {
+                iter->reservedField=0;
+            } else {
+                /* verified index>=4 above */
+                UChar32 c;
+                U8_PREV_OR_FFFD((const uint8_t *)iter->context, 0, index, c);
+                if(c<=0xffff) {
+                    *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                } else {
+                    iter->reservedField=c;
+                }
+            }
+        }
+    }
+}
+
+static const UCharIterator utf8Iterator={
+    0, 0, 0, 0, 0, 0,
+    utf8IteratorGetIndex,
+    utf8IteratorMove,
+    utf8IteratorHasNext,
+    utf8IteratorHasPrevious,
+    utf8IteratorCurrent,
+    utf8IteratorNext,
+    utf8IteratorPrevious,
+    NULL,
+    utf8IteratorGetState,
+    utf8IteratorSetState
+};
+
+U_CAPI void U_EXPORT2
+uiter_setUTF8(UCharIterator *iter, const char *s, int32_t length) {
+    if(iter!=0) {
+        if(s!=0 && length>=-1) {
+            *iter=utf8Iterator;
+            iter->context=s;
+            if(length>=0) {
+                iter->limit=length;
+            } else {
+                iter->limit=(int32_t)uprv_strlen(s);
+            }
+            iter->length= iter->limit<=1 ? iter->limit : -1;
+        } else {
+            *iter=noopIterator;
+        }
+    }
+}
+
+/* Helper functions --------------------------------------------------------- */
+
+U_CAPI UChar32 U_EXPORT2
+uiter_current32(UCharIterator *iter) {
+    UChar32 c, c2;
+
+    c=iter->current(iter);
+    if(U16_IS_SURROGATE(c)) {
+        if(U16_IS_SURROGATE_LEAD(c)) {
+            /*
+             * go to the next code unit
+             * we know that we are not at the limit because c!=U_SENTINEL
+             */
+            iter->move(iter, 1, UITER_CURRENT);
+            if(U16_IS_TRAIL(c2=iter->current(iter))) {
+                c=U16_GET_SUPPLEMENTARY(c, c2);
+            }
+
+            /* undo index movement */
+            iter->move(iter, -1, UITER_CURRENT);
+        } else {
+            if(U16_IS_LEAD(c2=iter->previous(iter))) {
+                c=U16_GET_SUPPLEMENTARY(c2, c);
+            }
+            if(c2>=0) {
+                /* undo index movement */
+                iter->move(iter, 1, UITER_CURRENT);
+            }
+        }
+    }
+    return c;
+}
+
+U_CAPI UChar32 U_EXPORT2
+uiter_next32(UCharIterator *iter) {
+    UChar32 c, c2;
+
+    c=iter->next(iter);
+    if(U16_IS_LEAD(c)) {
+        if(U16_IS_TRAIL(c2=iter->next(iter))) {
+            c=U16_GET_SUPPLEMENTARY(c, c2);
+        } else if(c2>=0) {
+            /* unmatched first surrogate, undo index movement */
+            iter->move(iter, -1, UITER_CURRENT);
+        }
+    }
+    return c;
+}
+
+U_CAPI UChar32 U_EXPORT2
+uiter_previous32(UCharIterator *iter) {
+    UChar32 c, c2;
+
+    c=iter->previous(iter);
+    if(U16_IS_TRAIL(c)) {
+        if(U16_IS_LEAD(c2=iter->previous(iter))) {
+            c=U16_GET_SUPPLEMENTARY(c2, c);
+        } else if(c2>=0) {
+            /* unmatched second surrogate, undo index movement */
+            iter->move(iter, 1, UITER_CURRENT);
+        }
+    }
+    return c;
+}
+
+U_CAPI uint32_t U_EXPORT2
+uiter_getState(const UCharIterator *iter) {
+    if(iter==NULL || iter->getState==NULL) {
+        return UITER_NO_STATE;
+    } else {
+        return iter->getState(iter);
+    }
+}
+
+U_CAPI void U_EXPORT2
+uiter_setState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        /* do nothing */
+    } else if(iter==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else if(iter->setState==NULL) {
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+    } else {
+        iter->setState(iter, state, pErrorCode);
+    }
+}
+
+U_CDECL_END
diff --git a/thirdparty/icu4c/common/ulayout_props.h b/thirdparty/icu4c/common/ulayout_props.h
new file mode 100644
index 0000000000..c0f028c713
--- /dev/null
+++ b/thirdparty/icu4c/common/ulayout_props.h
@@ -0,0 +1,46 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ulayout_props.h
+// created: 2019feb12 Markus W. Scherer
+
+#ifndef __ULAYOUT_PROPS_H__
+#define __ULAYOUT_PROPS_H__
+
+#include "unicode/utypes.h"
+
+// file definitions ------------------------------------------------------------
+
+#define ULAYOUT_DATA_NAME "ulayout"
+#define ULAYOUT_DATA_TYPE "icu"
+
+// data format "Layo"
+#define ULAYOUT_FMT_0 0x4c
+#define ULAYOUT_FMT_1 0x61
+#define ULAYOUT_FMT_2 0x79
+#define ULAYOUT_FMT_3 0x6f
+
+// indexes into indexes[]
+enum {
+    // Element 0 stores the length of the indexes[] array.
+    ULAYOUT_IX_INDEXES_LENGTH,
+    // Elements 1..7 store the tops of consecutive code point tries.
+    // No trie is stored if the difference between two of these is less than 16.
+    ULAYOUT_IX_INPC_TRIE_TOP,
+    ULAYOUT_IX_INSC_TRIE_TOP,
+    ULAYOUT_IX_VO_TRIE_TOP,
+    ULAYOUT_IX_RESERVED_TOP,
+
+    ULAYOUT_IX_TRIES_TOP = 7,
+
+    ULAYOUT_IX_MAX_VALUES = 9,
+
+    // Length of indexes[]. Multiple of 4 to 16-align the tries.
+    ULAYOUT_IX_COUNT = 12
+};
+
+constexpr int32_t ULAYOUT_MAX_INPC_SHIFT = 24;
+constexpr int32_t ULAYOUT_MAX_INSC_SHIFT = 16;
+constexpr int32_t ULAYOUT_MAX_VO_SHIFT = 8;
+
+#endif  // __ULAYOUT_PROPS_H__
diff --git a/thirdparty/icu4c/common/ulist.cpp b/thirdparty/icu4c/common/ulist.cpp
new file mode 100644
index 0000000000..c5180431c3
--- /dev/null
+++ b/thirdparty/icu4c/common/ulist.cpp
@@ -0,0 +1,270 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 2009-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*/
+
+#include "ulist.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uenumimp.h"
+
+typedef struct UListNode UListNode;
+struct UListNode {
+    void *data;
+    
+    UListNode *next;
+    UListNode *previous;
+    
+    /* When data is created with uprv_malloc, needs to be freed during deleteList function. */
+    UBool forceDelete;
+};
+
+struct UList {
+    UListNode *curr;
+    UListNode *head;
+    UListNode *tail;
+    
+    int32_t size;
+};
+
+static void ulist_addFirstItem(UList *list, UListNode *newItem);
+
+U_CAPI UList *U_EXPORT2 ulist_createEmptyList(UErrorCode *status) {
+    UList *newList = NULL;
+    
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    
+    newList = (UList *)uprv_malloc(sizeof(UList));
+    if (newList == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    
+    newList->curr = NULL;
+    newList->head = NULL;
+    newList->tail = NULL;
+    newList->size = 0;
+    
+    return newList;
+}
+
+/*
+ * Function called by addItemEndList or addItemBeginList when the first item is added to the list.
+ * This function properly sets the pointers for the first item added.
+ */
+static void ulist_addFirstItem(UList *list, UListNode *newItem) {
+    newItem->next = NULL;
+    newItem->previous = NULL;
+    list->head = newItem;
+    list->tail = newItem;
+}
+
+static void ulist_removeItem(UList *list, UListNode *p) {
+    if (p->previous == NULL) {
+        // p is the list head.
+        list->head = p->next;
+    } else {
+        p->previous->next = p->next;
+    }
+    if (p->next == NULL) {
+        // p is the list tail.
+        list->tail = p->previous;
+    } else {
+        p->next->previous = p->previous;
+    }
+    if (p == list->curr) {
+        list->curr = p->next;
+    }
+    --list->size;
+    if (p->forceDelete) {
+        uprv_free(p->data);
+    }
+    uprv_free(p);
+}
+
+U_CAPI void U_EXPORT2 ulist_addItemEndList(UList *list, const void *data, UBool forceDelete, UErrorCode *status) {
+    UListNode *newItem = NULL;
+    
+    if (U_FAILURE(*status) || list == NULL || data == NULL) {
+        if (forceDelete) {
+            uprv_free((void *)data);
+        }
+        return;
+    }
+    
+    newItem = (UListNode *)uprv_malloc(sizeof(UListNode));
+    if (newItem == NULL) {
+        if (forceDelete) {
+            uprv_free((void *)data);
+        }
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    newItem->data = (void *)(data);
+    newItem->forceDelete = forceDelete;
+    
+    if (list->size == 0) {
+        ulist_addFirstItem(list, newItem);
+    } else {
+        newItem->next = NULL;
+        newItem->previous = list->tail;
+        list->tail->next = newItem;
+        list->tail = newItem;
+    }
+    
+    list->size++;
+}
+
+U_CAPI void U_EXPORT2 ulist_addItemBeginList(UList *list, const void *data, UBool forceDelete, UErrorCode *status) {
+    UListNode *newItem = NULL;
+    
+    if (U_FAILURE(*status) || list == NULL || data == NULL) {
+        if (forceDelete) {
+            uprv_free((void *)data);
+        }
+        return;
+    }
+    
+    newItem = (UListNode *)uprv_malloc(sizeof(UListNode));
+    if (newItem == NULL) {
+        if (forceDelete) {
+            uprv_free((void *)data);
+        }
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    newItem->data = (void *)(data);
+    newItem->forceDelete = forceDelete;
+    
+    if (list->size == 0) {
+        ulist_addFirstItem(list, newItem);
+    } else {
+        newItem->previous = NULL;
+        newItem->next = list->head;
+        list->head->previous = newItem;
+        list->head = newItem;
+    }
+    
+    list->size++;
+}
+
+U_CAPI UBool U_EXPORT2 ulist_containsString(const UList *list, const char *data, int32_t length) {
+    if (list != NULL) {
+        const UListNode *pointer;
+        for (pointer = list->head; pointer != NULL; pointer = pointer->next) {
+            if (length == (int32_t)uprv_strlen((const char *)pointer->data)) {
+                if (uprv_memcmp(data, pointer->data, length) == 0) {
+                    return TRUE;
+                }
+            }
+        }
+    }
+    return FALSE;
+}
+
+U_CAPI UBool U_EXPORT2 ulist_removeString(UList *list, const char *data) {
+    if (list != NULL) {
+        UListNode *pointer;
+        for (pointer = list->head; pointer != NULL; pointer = pointer->next) {
+            if (uprv_strcmp(data, (const char *)pointer->data) == 0) {
+                ulist_removeItem(list, pointer);
+                // Remove only the first occurrence, like Java LinkedList.remove(Object).
+                return TRUE;
+            }
+        }
+    }
+    return FALSE;
+}
+
+U_CAPI void *U_EXPORT2 ulist_getNext(UList *list) {
+    UListNode *curr = NULL;
+    
+    if (list == NULL || list->curr == NULL) {
+        return NULL;
+    }
+    
+    curr = list->curr;
+    list->curr = curr->next;
+    
+    return curr->data;
+}
+
+U_CAPI int32_t U_EXPORT2 ulist_getListSize(const UList *list) {
+    if (list != NULL) {
+        return list->size;
+    }
+    
+    return -1;
+}
+
+U_CAPI void U_EXPORT2 ulist_resetList(UList *list) {
+    if (list != NULL) {
+        list->curr = list->head;
+    }
+}
+
+U_CAPI void U_EXPORT2 ulist_deleteList(UList *list) {
+    UListNode *listHead = NULL;
+
+    if (list != NULL) {
+        listHead = list->head;
+        while (listHead != NULL) {
+            UListNode *listPointer = listHead->next;
+
+            if (listHead->forceDelete) {
+                uprv_free(listHead->data);
+            }
+
+            uprv_free(listHead);
+            listHead = listPointer;
+        }
+        uprv_free(list);
+        list = NULL;
+    }
+}
+
+U_CAPI void U_EXPORT2 ulist_close_keyword_values_iterator(UEnumeration *en) {
+    if (en != NULL) {
+        ulist_deleteList((UList *)(en->context));
+        uprv_free(en);
+    }
+}
+
+U_CAPI int32_t U_EXPORT2 ulist_count_keyword_values(UEnumeration *en, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return -1;
+    }
+    
+    return ulist_getListSize((UList *)(en->context));
+}
+
+U_CAPI const char * U_EXPORT2 ulist_next_keyword_value(UEnumeration *en, int32_t *resultLength, UErrorCode *status) {
+    const char *s;
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    s = (const char *)ulist_getNext((UList *)(en->context));
+    if (s != NULL && resultLength != NULL) {
+        *resultLength = static_cast<int32_t>(uprv_strlen(s));
+    }
+    return s;
+}
+
+U_CAPI void U_EXPORT2 ulist_reset_keyword_values_iterator(UEnumeration *en, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return ;
+    }
+    
+    ulist_resetList((UList *)(en->context));
+}
+
+U_CAPI UList * U_EXPORT2 ulist_getListFromEnum(UEnumeration *en) {
+    return (UList *)(en->context);
+}
diff --git a/thirdparty/icu4c/common/ulist.h b/thirdparty/icu4c/common/ulist.h
new file mode 100644
index 0000000000..de58a4ad02
--- /dev/null
+++ b/thirdparty/icu4c/common/ulist.h
@@ -0,0 +1,50 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 2009-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*/
+
+#ifndef ULIST_H
+#define ULIST_H
+
+#include "unicode/utypes.h"
+#include "unicode/uenum.h"
+
+struct UList;
+typedef struct UList UList;
+
+U_CAPI UList * U_EXPORT2 ulist_createEmptyList(UErrorCode *status);
+
+U_CAPI void U_EXPORT2 ulist_addItemEndList(UList *list, const void *data, UBool forceDelete, UErrorCode *status);
+
+U_CAPI void U_EXPORT2 ulist_addItemBeginList(UList *list, const void *data, UBool forceDelete, UErrorCode *status);
+
+U_CAPI UBool U_EXPORT2 ulist_containsString(const UList *list, const char *data, int32_t length);
+
+U_CAPI UBool U_EXPORT2 ulist_removeString(UList *list, const char *data);
+
+U_CAPI void *U_EXPORT2 ulist_getNext(UList *list);
+
+U_CAPI int32_t U_EXPORT2 ulist_getListSize(const UList *list);
+
+U_CAPI void U_EXPORT2 ulist_resetList(UList *list);
+
+U_CAPI void U_EXPORT2 ulist_deleteList(UList *list);
+
+/*
+ * The following are for use when creating UEnumeration object backed by UList.
+ */
+U_CAPI void U_EXPORT2 ulist_close_keyword_values_iterator(UEnumeration *en);
+
+U_CAPI int32_t U_EXPORT2 ulist_count_keyword_values(UEnumeration *en, UErrorCode *status);
+
+U_CAPI const char * U_EXPORT2 ulist_next_keyword_value(UEnumeration* en, int32_t *resultLength, UErrorCode* status);
+
+U_CAPI void U_EXPORT2 ulist_reset_keyword_values_iterator(UEnumeration* en, UErrorCode* status);
+
+U_CAPI UList * U_EXPORT2 ulist_getListFromEnum(UEnumeration *en);
+
+#endif
diff --git a/thirdparty/icu4c/common/uloc.cpp b/thirdparty/icu4c/common/uloc.cpp
new file mode 100644
index 0000000000..522f33dbe2
--- /dev/null
+++ b/thirdparty/icu4c/common/uloc.cpp
@@ -0,0 +1,2176 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File ULOC.CPP
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/01/97    aliu        Creation.
+*   08/21/98    stephen     JDK 1.2 sync
+*   12/08/98    rtg         New Locale implementation and C API
+*   03/15/99    damiba      overhaul.
+*   04/06/99    stephen     changed setDefault() to realloc and copy
+*   06/14/99    stephen     Changed calls to ures_open for new params
+*   07/21/99    stephen     Modified setDefault() to propagate to C++
+*   05/14/04    alan        7 years later: refactored, cleaned up, fixed bugs,
+*                           brought canonicalization code into line with spec
+*****************************************************************************/
+
+/*
+   POSIX's locale format, from putil.c: [no spaces]
+
+     ll [ _CC ] [ . MM ] [ @ VV]
+
+     l = lang, C = ctry, M = charmap, V = variant
+*/
+
+#include "unicode/bytestream.h"
+#include "unicode/errorcode.h"
+#include "unicode/stringpiece.h"
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/uloc.h"
+
+#include "bytesinkutil.h"
+#include "putilimp.h"
+#include "ustr_imp.h"
+#include "ulocimp.h"
+#include "umutex.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "locmap.h"
+#include "uarrsort.h"
+#include "uenumimp.h"
+#include "uassert.h"
+#include "charstr.h"
+
+U_NAMESPACE_USE
+
+/* ### Declarations **************************************************/
+
+/* Locale stuff from locid.cpp */
+U_CFUNC void locale_set_default(const char *id);
+U_CFUNC const char *locale_get_default(void);
+
+/* ### Data tables **************************************************/
+
+/**
+ * Table of language codes, both 2- and 3-letter, with preference
+ * given to 2-letter codes where possible.  Includes 3-letter codes
+ * that lack a 2-letter equivalent.
+ *
+ * This list must be in sorted order.  This list is returned directly
+ * to the user by some API.
+ *
+ * This list must be kept in sync with LANGUAGES_3, with corresponding
+ * entries matched.
+ *
+ * This table should be terminated with a NULL entry, followed by a
+ * second list, and another NULL entry.  The first list is visible to
+ * user code when this array is returned by API.  The second list
+ * contains codes we support, but do not expose through user API.
+ *
+ * Notes
+ *
+ * Tables updated per http://lcweb.loc.gov/standards/iso639-2/ to
+ * include the revisions up to 2001/7/27 *CWB*
+ *
+ * The 3 character codes are the terminology codes like RFC 3066.  This
+ * is compatible with prior ICU codes
+ *
+ * "in" "iw" "ji" "jw" & "sh" have been withdrawn but are still in the
+ * table but now at the end of the table because 3 character codes are
+ * duplicates.  This avoids bad searches going from 3 to 2 character
+ * codes.
+ *
+ * The range qaa-qtz is reserved for local use
+ */
+/* Generated using org.unicode.cldr.icu.GenerateISO639LanguageTables */
+/* ISO639 table version is 20150505 */
+/* Subsequent hand addition of selected languages */
+static const char * const LANGUAGES[] = {
+    "aa",  "ab",  "ace", "ach", "ada", "ady", "ae",  "aeb",
+    "af",  "afh", "agq", "ain", "ak",  "akk", "akz", "ale",
+    "aln", "alt", "am",  "an",  "ang", "anp", "ar",  "arc",
+    "arn", "aro", "arp", "arq", "ars", "arw", "ary", "arz", "as",
+    "asa", "ase", "ast", "av",  "avk", "awa", "ay",  "az",
+    "ba",  "bal", "ban", "bar", "bas", "bax", "bbc", "bbj",
+    "be",  "bej", "bem", "bew", "bez", "bfd", "bfq", "bg",
+    "bgn", "bho", "bi",  "bik", "bin", "bjn", "bkm", "bla",
+    "bm",  "bn",  "bo",  "bpy", "bqi", "br",  "bra", "brh",
+    "brx", "bs",  "bss", "bua", "bug", "bum", "byn", "byv",
+    "ca",  "cad", "car", "cay", "cch", "ccp", "ce",  "ceb", "cgg",
+    "ch",  "chb", "chg", "chk", "chm", "chn", "cho", "chp",
+    "chr", "chy", "ckb", "co",  "cop", "cps", "cr",  "crh",
+    "cs",  "csb", "cu",  "cv",  "cy",
+    "da",  "dak", "dar", "dav", "de",  "del", "den", "dgr",
+    "din", "dje", "doi", "dsb", "dtp", "dua", "dum", "dv",
+    "dyo", "dyu", "dz",  "dzg",
+    "ebu", "ee",  "efi", "egl", "egy", "eka", "el",  "elx",
+    "en",  "enm", "eo",  "es",  "esu", "et",  "eu",  "ewo",
+    "ext",
+    "fa",  "fan", "fat", "ff",  "fi",  "fil", "fit", "fj",
+    "fo",  "fon", "fr",  "frc", "frm", "fro", "frp", "frr",
+    "frs", "fur", "fy",
+    "ga",  "gaa", "gag", "gan", "gay", "gba", "gbz", "gd",
+    "gez", "gil", "gl",  "glk", "gmh", "gn",  "goh", "gom",
+    "gon", "gor", "got", "grb", "grc", "gsw", "gu",  "guc",
+    "gur", "guz", "gv",  "gwi",
+    "ha",  "hai", "hak", "haw", "he",  "hi",  "hif", "hil",
+    "hit", "hmn", "ho",  "hr",  "hsb", "hsn", "ht",  "hu",
+    "hup", "hy",  "hz",
+    "ia",  "iba", "ibb", "id",  "ie",  "ig",  "ii",  "ik",
+    "ilo", "inh", "io",  "is",  "it",  "iu",  "izh",
+    "ja",  "jam", "jbo", "jgo", "jmc", "jpr", "jrb", "jut",
+    "jv",
+    "ka",  "kaa", "kab", "kac", "kaj", "kam", "kaw", "kbd",
+    "kbl", "kcg", "kde", "kea", "ken", "kfo", "kg",  "kgp",
+    "kha", "kho", "khq", "khw", "ki",  "kiu", "kj",  "kk",
+    "kkj", "kl",  "kln", "km",  "kmb", "kn",  "ko",  "koi",
+    "kok", "kos", "kpe", "kr",  "krc", "kri", "krj", "krl",
+    "kru", "ks",  "ksb", "ksf", "ksh", "ku",  "kum", "kut",
+    "kv",  "kw",  "ky",
+    "la",  "lad", "lag", "lah", "lam", "lb",  "lez", "lfn",
+    "lg",  "li",  "lij", "liv", "lkt", "lmo", "ln",  "lo",
+    "lol", "loz", "lrc", "lt",  "ltg", "lu",  "lua", "lui",
+    "lun", "luo", "lus", "luy", "lv",  "lzh", "lzz",
+    "mad", "maf", "mag", "mai", "mak", "man", "mas", "mde",
+    "mdf", "mdh", "mdr", "men", "mer", "mfe", "mg",  "mga",
+    "mgh", "mgo", "mh",  "mi",  "mic", "min", "mis", "mk",
+    "ml",  "mn",  "mnc", "mni", "mo",
+    "moh", "mos", "mr",  "mrj",
+    "ms",  "mt",  "mua", "mul", "mus", "mwl", "mwr", "mwv",
+    "my",  "mye", "myv", "mzn",
+    "na",  "nan", "nap", "naq", "nb",  "nd",  "nds", "ne",
+    "new", "ng",  "nia", "niu", "njo", "nl",  "nmg", "nn",
+    "nnh", "no",  "nog", "non", "nov", "nqo", "nr",  "nso",
+    "nus", "nv",  "nwc", "ny",  "nym", "nyn", "nyo", "nzi",
+    "oc",  "oj",  "om",  "or",  "os",  "osa", "ota",
+    "pa",  "pag", "pal", "pam", "pap", "pau", "pcd", "pcm", "pdc",
+    "pdt", "peo", "pfl", "phn", "pi",  "pl",  "pms", "pnt",
+    "pon", "prg", "pro", "ps",  "pt",
+    "qu",  "quc", "qug",
+    "raj", "rap", "rar", "rgn", "rif", "rm",  "rn",  "ro",
+    "rof", "rom", "rtm", "ru",  "rue", "rug", "rup",
+    "rw",  "rwk",
+    "sa",  "sad", "sah", "sam", "saq", "sas", "sat", "saz",
+    "sba", "sbp", "sc",  "scn", "sco", "sd",  "sdc", "sdh",
+    "se",  "see", "seh", "sei", "sel", "ses", "sg",  "sga",
+    "sgs", "shi", "shn", "shu", "si",  "sid", "sk",
+    "sl",  "sli", "sly", "sm",  "sma", "smj", "smn", "sms",
+    "sn",  "snk", "so",  "sog", "sq",  "sr",  "srn", "srr",
+    "ss",  "ssy", "st",  "stq", "su",  "suk", "sus", "sux",
+    "sv",  "sw",  "swb", "swc", "syc", "syr", "szl",
+    "ta",  "tcy", "te",  "tem", "teo", "ter", "tet", "tg",
+    "th",  "ti",  "tig", "tiv", "tk",  "tkl", "tkr", "tl",
+    "tlh", "tli", "tly", "tmh", "tn",  "to",  "tog", "tpi",
+    "tr",  "tru", "trv", "ts",  "tsd", "tsi", "tt",  "ttt",
+    "tum", "tvl", "tw",  "twq", "ty",  "tyv", "tzm",
+    "udm", "ug",  "uga", "uk",  "umb", "und", "ur",  "uz",
+    "vai", "ve",  "vec", "vep", "vi",  "vls", "vmf", "vo",
+    "vot", "vro", "vun",
+    "wa",  "wae", "wal", "war", "was", "wbp", "wo",  "wuu",
+    "xal", "xh",  "xmf", "xog",
+    "yao", "yap", "yav", "ybb", "yi",  "yo",  "yrl", "yue",
+    "za",  "zap", "zbl", "zea", "zen", "zgh", "zh",  "zu",
+    "zun", "zxx", "zza",
+NULL,
+    "in",  "iw",  "ji",  "jw",  "sh",    /* obsolete language codes */
+NULL
+};
+
+static const char* const DEPRECATED_LANGUAGES[]={
+    "in", "iw", "ji", "jw", NULL, NULL
+};
+static const char* const REPLACEMENT_LANGUAGES[]={
+    "id", "he", "yi", "jv", NULL, NULL
+};
+
+/**
+ * Table of 3-letter language codes.
+ *
+ * This is a lookup table used to convert 3-letter language codes to
+ * their 2-letter equivalent, where possible.  It must be kept in sync
+ * with LANGUAGES.  For all valid i, LANGUAGES[i] must refer to the
+ * same language as LANGUAGES_3[i].  The commented-out lines are
+ * copied from LANGUAGES to make eyeballing this baby easier.
+ *
+ * Where a 3-letter language code has no 2-letter equivalent, the
+ * 3-letter code occupies both LANGUAGES[i] and LANGUAGES_3[i].
+ *
+ * This table should be terminated with a NULL entry, followed by a
+ * second list, and another NULL entry.  The two lists correspond to
+ * the two lists in LANGUAGES.
+ */
+/* Generated using org.unicode.cldr.icu.GenerateISO639LanguageTables */
+/* ISO639 table version is 20150505 */
+/* Subsequent hand addition of selected languages */
+static const char * const LANGUAGES_3[] = {
+    "aar", "abk", "ace", "ach", "ada", "ady", "ave", "aeb",
+    "afr", "afh", "agq", "ain", "aka", "akk", "akz", "ale",
+    "aln", "alt", "amh", "arg", "ang", "anp", "ara", "arc",
+    "arn", "aro", "arp", "arq", "ars", "arw", "ary", "arz", "asm",
+    "asa", "ase", "ast", "ava", "avk", "awa", "aym", "aze",
+    "bak", "bal", "ban", "bar", "bas", "bax", "bbc", "bbj",
+    "bel", "bej", "bem", "bew", "bez", "bfd", "bfq", "bul",
+    "bgn", "bho", "bis", "bik", "bin", "bjn", "bkm", "bla",
+    "bam", "ben", "bod", "bpy", "bqi", "bre", "bra", "brh",
+    "brx", "bos", "bss", "bua", "bug", "bum", "byn", "byv",
+    "cat", "cad", "car", "cay", "cch", "ccp", "che", "ceb", "cgg",
+    "cha", "chb", "chg", "chk", "chm", "chn", "cho", "chp",
+    "chr", "chy", "ckb", "cos", "cop", "cps", "cre", "crh",
+    "ces", "csb", "chu", "chv", "cym",
+    "dan", "dak", "dar", "dav", "deu", "del", "den", "dgr",
+    "din", "dje", "doi", "dsb", "dtp", "dua", "dum", "div",
+    "dyo", "dyu", "dzo", "dzg",
+    "ebu", "ewe", "efi", "egl", "egy", "eka", "ell", "elx",
+    "eng", "enm", "epo", "spa", "esu", "est", "eus", "ewo",
+    "ext",
+    "fas", "fan", "fat", "ful", "fin", "fil", "fit", "fij",
+    "fao", "fon", "fra", "frc", "frm", "fro", "frp", "frr",
+    "frs", "fur", "fry",
+    "gle", "gaa", "gag", "gan", "gay", "gba", "gbz", "gla",
+    "gez", "gil", "glg", "glk", "gmh", "grn", "goh", "gom",
+    "gon", "gor", "got", "grb", "grc", "gsw", "guj", "guc",
+    "gur", "guz", "glv", "gwi",
+    "hau", "hai", "hak", "haw", "heb", "hin", "hif", "hil",
+    "hit", "hmn", "hmo", "hrv", "hsb", "hsn", "hat", "hun",
+    "hup", "hye", "her",
+    "ina", "iba", "ibb", "ind", "ile", "ibo", "iii", "ipk",
+    "ilo", "inh", "ido", "isl", "ita", "iku", "izh",
+    "jpn", "jam", "jbo", "jgo", "jmc", "jpr", "jrb", "jut",
+    "jav",
+    "kat", "kaa", "kab", "kac", "kaj", "kam", "kaw", "kbd",
+    "kbl", "kcg", "kde", "kea", "ken", "kfo", "kon", "kgp",
+    "kha", "kho", "khq", "khw", "kik", "kiu", "kua", "kaz",
+    "kkj", "kal", "kln", "khm", "kmb", "kan", "kor", "koi",
+    "kok", "kos", "kpe", "kau", "krc", "kri", "krj", "krl",
+    "kru", "kas", "ksb", "ksf", "ksh", "kur", "kum", "kut",
+    "kom", "cor", "kir",
+    "lat", "lad", "lag", "lah", "lam", "ltz", "lez", "lfn",
+    "lug", "lim", "lij", "liv", "lkt", "lmo", "lin", "lao",
+    "lol", "loz", "lrc", "lit", "ltg", "lub", "lua", "lui",
+    "lun", "luo", "lus", "luy", "lav", "lzh", "lzz",
+    "mad", "maf", "mag", "mai", "mak", "man", "mas", "mde",
+    "mdf", "mdh", "mdr", "men", "mer", "mfe", "mlg", "mga",
+    "mgh", "mgo", "mah", "mri", "mic", "min", "mis", "mkd",
+    "mal", "mon", "mnc", "mni", "mol",
+    "moh", "mos", "mar", "mrj",
+    "msa", "mlt", "mua", "mul", "mus", "mwl", "mwr", "mwv",
+    "mya", "mye", "myv", "mzn",
+    "nau", "nan", "nap", "naq", "nob", "nde", "nds", "nep",
+    "new", "ndo", "nia", "niu", "njo", "nld", "nmg", "nno",
+    "nnh", "nor", "nog", "non", "nov", "nqo", "nbl", "nso",
+    "nus", "nav", "nwc", "nya", "nym", "nyn", "nyo", "nzi",
+    "oci", "oji", "orm", "ori", "oss", "osa", "ota",
+    "pan", "pag", "pal", "pam", "pap", "pau", "pcd", "pcm", "pdc",
+    "pdt", "peo", "pfl", "phn", "pli", "pol", "pms", "pnt",
+    "pon", "prg", "pro", "pus", "por",
+    "que", "quc", "qug",
+    "raj", "rap", "rar", "rgn", "rif", "roh", "run", "ron",
+    "rof", "rom", "rtm", "rus", "rue", "rug", "rup",
+    "kin", "rwk",
+    "san", "sad", "sah", "sam", "saq", "sas", "sat", "saz",
+    "sba", "sbp", "srd", "scn", "sco", "snd", "sdc", "sdh",
+    "sme", "see", "seh", "sei", "sel", "ses", "sag", "sga",
+    "sgs", "shi", "shn", "shu", "sin", "sid", "slk",
+    "slv", "sli", "sly", "smo", "sma", "smj", "smn", "sms",
+    "sna", "snk", "som", "sog", "sqi", "srp", "srn", "srr",
+    "ssw", "ssy", "sot", "stq", "sun", "suk", "sus", "sux",
+    "swe", "swa", "swb", "swc", "syc", "syr", "szl",
+    "tam", "tcy", "tel", "tem", "teo", "ter", "tet", "tgk",
+    "tha", "tir", "tig", "tiv", "tuk", "tkl", "tkr", "tgl",
+    "tlh", "tli", "tly", "tmh", "tsn", "ton", "tog", "tpi",
+    "tur", "tru", "trv", "tso", "tsd", "tsi", "tat", "ttt",
+    "tum", "tvl", "twi", "twq", "tah", "tyv", "tzm",
+    "udm", "uig", "uga", "ukr", "umb", "und", "urd", "uzb",
+    "vai", "ven", "vec", "vep", "vie", "vls", "vmf", "vol",
+    "vot", "vro", "vun",
+    "wln", "wae", "wal", "war", "was", "wbp", "wol", "wuu",
+    "xal", "xho", "xmf", "xog",
+    "yao", "yap", "yav", "ybb", "yid", "yor", "yrl", "yue",
+    "zha", "zap", "zbl", "zea", "zen", "zgh", "zho", "zul",
+    "zun", "zxx", "zza",
+NULL,
+/*  "in",  "iw",  "ji",  "jw",  "sh",                          */
+    "ind", "heb", "yid", "jaw", "srp",
+NULL
+};
+
+/**
+ * Table of 2-letter country codes.
+ *
+ * This list must be in sorted order.  This list is returned directly
+ * to the user by some API.
+ *
+ * This list must be kept in sync with COUNTRIES_3, with corresponding
+ * entries matched.
+ *
+ * This table should be terminated with a NULL entry, followed by a
+ * second list, and another NULL entry.  The first list is visible to
+ * user code when this array is returned by API.  The second list
+ * contains codes we support, but do not expose through user API.
+ *
+ * Notes:
+ *
+ * ZR(ZAR) is now CD(COD) and FX(FXX) is PS(PSE) as per
+ * http://www.evertype.com/standards/iso3166/iso3166-1-en.html added
+ * new codes keeping the old ones for compatibility updated to include
+ * 1999/12/03 revisions *CWB*
+ *
+ * RO(ROM) is now RO(ROU) according to
+ * http://www.iso.org/iso/en/prods-services/iso3166ma/03updates-on-iso-3166/nlv3e-rou.html
+ */
+static const char * const COUNTRIES[] = {
+    "AD",  "AE",  "AF",  "AG",  "AI",  "AL",  "AM",
+    "AO",  "AQ",  "AR",  "AS",  "AT",  "AU",  "AW",  "AX",  "AZ",
+    "BA",  "BB",  "BD",  "BE",  "BF",  "BG",  "BH",  "BI",
+    "BJ",  "BL",  "BM",  "BN",  "BO",  "BQ",  "BR",  "BS",  "BT",  "BV",
+    "BW",  "BY",  "BZ",  "CA",  "CC",  "CD",  "CF",  "CG",
+    "CH",  "CI",  "CK",  "CL",  "CM",  "CN",  "CO",  "CR",
+    "CU",  "CV",  "CW",  "CX",  "CY",  "CZ",  "DE",  "DJ",  "DK",
+    "DM",  "DO",  "DZ",  "EC",  "EE",  "EG",  "EH",  "ER",
+    "ES",  "ET",  "FI",  "FJ",  "FK",  "FM",  "FO",  "FR",
+    "GA",  "GB",  "GD",  "GE",  "GF",  "GG",  "GH",  "GI",  "GL",
+    "GM",  "GN",  "GP",  "GQ",  "GR",  "GS",  "GT",  "GU",
+    "GW",  "GY",  "HK",  "HM",  "HN",  "HR",  "HT",  "HU",
+    "ID",  "IE",  "IL",  "IM",  "IN",  "IO",  "IQ",  "IR",  "IS",
+    "IT",  "JE",  "JM",  "JO",  "JP",  "KE",  "KG",  "KH",  "KI",
+    "KM",  "KN",  "KP",  "KR",  "KW",  "KY",  "KZ",  "LA",
+    "LB",  "LC",  "LI",  "LK",  "LR",  "LS",  "LT",  "LU",
+    "LV",  "LY",  "MA",  "MC",  "MD",  "ME",  "MF",  "MG",  "MH",  "MK",
+    "ML",  "MM",  "MN",  "MO",  "MP",  "MQ",  "MR",  "MS",
+    "MT",  "MU",  "MV",  "MW",  "MX",  "MY",  "MZ",  "NA",
+    "NC",  "NE",  "NF",  "NG",  "NI",  "NL",  "NO",  "NP",
+    "NR",  "NU",  "NZ",  "OM",  "PA",  "PE",  "PF",  "PG",
+    "PH",  "PK",  "PL",  "PM",  "PN",  "PR",  "PS",  "PT",
+    "PW",  "PY",  "QA",  "RE",  "RO",  "RS",  "RU",  "RW",  "SA",
+    "SB",  "SC",  "SD",  "SE",  "SG",  "SH",  "SI",  "SJ",
+    "SK",  "SL",  "SM",  "SN",  "SO",  "SR",  "SS",  "ST",  "SV",
+    "SX",  "SY",  "SZ",  "TC",  "TD",  "TF",  "TG",  "TH",  "TJ",
+    "TK",  "TL",  "TM",  "TN",  "TO",  "TR",  "TT",  "TV",
+    "TW",  "TZ",  "UA",  "UG",  "UM",  "US",  "UY",  "UZ",
+    "VA",  "VC",  "VE",  "VG",  "VI",  "VN",  "VU",  "WF",
+    "WS",  "YE",  "YT",  "ZA",  "ZM",  "ZW",
+NULL,
+    "AN",  "BU", "CS", "FX", "RO", "SU", "TP", "YD", "YU", "ZR",   /* obsolete country codes */
+NULL
+};
+
+static const char* const DEPRECATED_COUNTRIES[] = {
+    "AN", "BU", "CS", "DD", "DY", "FX", "HV", "NH", "RH", "SU", "TP", "UK", "VD", "YD", "YU", "ZR", NULL, NULL /* deprecated country list */
+};
+static const char* const REPLACEMENT_COUNTRIES[] = {
+/*  "AN", "BU", "CS", "DD", "DY", "FX", "HV", "NH", "RH", "SU", "TP", "UK", "VD", "YD", "YU", "ZR" */
+    "CW", "MM", "RS", "DE", "BJ", "FR", "BF", "VU", "ZW", "RU", "TL", "GB", "VN", "YE", "RS", "CD", NULL, NULL  /* replacement country codes */
+};
+
+/**
+ * Table of 3-letter country codes.
+ *
+ * This is a lookup table used to convert 3-letter country codes to
+ * their 2-letter equivalent.  It must be kept in sync with COUNTRIES.
+ * For all valid i, COUNTRIES[i] must refer to the same country as
+ * COUNTRIES_3[i].  The commented-out lines are copied from COUNTRIES
+ * to make eyeballing this baby easier.
+ *
+ * This table should be terminated with a NULL entry, followed by a
+ * second list, and another NULL entry.  The two lists correspond to
+ * the two lists in COUNTRIES.
+ */
+static const char * const COUNTRIES_3[] = {
+/*  "AD",  "AE",  "AF",  "AG",  "AI",  "AL",  "AM",      */
+    "AND", "ARE", "AFG", "ATG", "AIA", "ALB", "ARM",
+/*  "AO",  "AQ",  "AR",  "AS",  "AT",  "AU",  "AW",  "AX",  "AZ",     */
+    "AGO", "ATA", "ARG", "ASM", "AUT", "AUS", "ABW", "ALA", "AZE",
+/*  "BA",  "BB",  "BD",  "BE",  "BF",  "BG",  "BH",  "BI",     */
+    "BIH", "BRB", "BGD", "BEL", "BFA", "BGR", "BHR", "BDI",
+/*  "BJ",  "BL",  "BM",  "BN",  "BO",  "BQ",  "BR",  "BS",  "BT",  "BV",     */
+    "BEN", "BLM", "BMU", "BRN", "BOL", "BES", "BRA", "BHS", "BTN", "BVT",
+/*  "BW",  "BY",  "BZ",  "CA",  "CC",  "CD",  "CF",  "CG",     */
+    "BWA", "BLR", "BLZ", "CAN", "CCK", "COD", "CAF", "COG",
+/*  "CH",  "CI",  "CK",  "CL",  "CM",  "CN",  "CO",  "CR",     */
+    "CHE", "CIV", "COK", "CHL", "CMR", "CHN", "COL", "CRI",
+/*  "CU",  "CV",  "CW",  "CX",  "CY",  "CZ",  "DE",  "DJ",  "DK",     */
+    "CUB", "CPV", "CUW", "CXR", "CYP", "CZE", "DEU", "DJI", "DNK",
+/*  "DM",  "DO",  "DZ",  "EC",  "EE",  "EG",  "EH",  "ER",     */
+    "DMA", "DOM", "DZA", "ECU", "EST", "EGY", "ESH", "ERI",
+/*  "ES",  "ET",  "FI",  "FJ",  "FK",  "FM",  "FO",  "FR",     */
+    "ESP", "ETH", "FIN", "FJI", "FLK", "FSM", "FRO", "FRA",
+/*  "GA",  "GB",  "GD",  "GE",  "GF",  "GG",  "GH",  "GI",  "GL",     */
+    "GAB", "GBR", "GRD", "GEO", "GUF", "GGY", "GHA", "GIB", "GRL",
+/*  "GM",  "GN",  "GP",  "GQ",  "GR",  "GS",  "GT",  "GU",     */
+    "GMB", "GIN", "GLP", "GNQ", "GRC", "SGS", "GTM", "GUM",
+/*  "GW",  "GY",  "HK",  "HM",  "HN",  "HR",  "HT",  "HU",     */
+    "GNB", "GUY", "HKG", "HMD", "HND", "HRV", "HTI", "HUN",
+/*  "ID",  "IE",  "IL",  "IM",  "IN",  "IO",  "IQ",  "IR",  "IS" */
+    "IDN", "IRL", "ISR", "IMN", "IND", "IOT", "IRQ", "IRN", "ISL",
+/*  "IT",  "JE",  "JM",  "JO",  "JP",  "KE",  "KG",  "KH",  "KI",     */
+    "ITA", "JEY", "JAM", "JOR", "JPN", "KEN", "KGZ", "KHM", "KIR",
+/*  "KM",  "KN",  "KP",  "KR",  "KW",  "KY",  "KZ",  "LA",     */
+    "COM", "KNA", "PRK", "KOR", "KWT", "CYM", "KAZ", "LAO",
+/*  "LB",  "LC",  "LI",  "LK",  "LR",  "LS",  "LT",  "LU",     */
+    "LBN", "LCA", "LIE", "LKA", "LBR", "LSO", "LTU", "LUX",
+/*  "LV",  "LY",  "MA",  "MC",  "MD",  "ME",  "MF",  "MG",  "MH",  "MK",     */
+    "LVA", "LBY", "MAR", "MCO", "MDA", "MNE", "MAF", "MDG", "MHL", "MKD",
+/*  "ML",  "MM",  "MN",  "MO",  "MP",  "MQ",  "MR",  "MS",     */
+    "MLI", "MMR", "MNG", "MAC", "MNP", "MTQ", "MRT", "MSR",
+/*  "MT",  "MU",  "MV",  "MW",  "MX",  "MY",  "MZ",  "NA",     */
+    "MLT", "MUS", "MDV", "MWI", "MEX", "MYS", "MOZ", "NAM",
+/*  "NC",  "NE",  "NF",  "NG",  "NI",  "NL",  "NO",  "NP",     */
+    "NCL", "NER", "NFK", "NGA", "NIC", "NLD", "NOR", "NPL",
+/*  "NR",  "NU",  "NZ",  "OM",  "PA",  "PE",  "PF",  "PG",     */
+    "NRU", "NIU", "NZL", "OMN", "PAN", "PER", "PYF", "PNG",
+/*  "PH",  "PK",  "PL",  "PM",  "PN",  "PR",  "PS",  "PT",     */
+    "PHL", "PAK", "POL", "SPM", "PCN", "PRI", "PSE", "PRT",
+/*  "PW",  "PY",  "QA",  "RE",  "RO",  "RS",  "RU",  "RW",  "SA",     */
+    "PLW", "PRY", "QAT", "REU", "ROU", "SRB", "RUS", "RWA", "SAU",
+/*  "SB",  "SC",  "SD",  "SE",  "SG",  "SH",  "SI",  "SJ",     */
+    "SLB", "SYC", "SDN", "SWE", "SGP", "SHN", "SVN", "SJM",
+/*  "SK",  "SL",  "SM",  "SN",  "SO",  "SR",  "SS",  "ST",  "SV",     */
+    "SVK", "SLE", "SMR", "SEN", "SOM", "SUR", "SSD", "STP", "SLV",
+/*  "SX",  "SY",  "SZ",  "TC",  "TD",  "TF",  "TG",  "TH",  "TJ",     */
+    "SXM", "SYR", "SWZ", "TCA", "TCD", "ATF", "TGO", "THA", "TJK",
+/*  "TK",  "TL",  "TM",  "TN",  "TO",  "TR",  "TT",  "TV",     */
+    "TKL", "TLS", "TKM", "TUN", "TON", "TUR", "TTO", "TUV",
+/*  "TW",  "TZ",  "UA",  "UG",  "UM",  "US",  "UY",  "UZ",     */
+    "TWN", "TZA", "UKR", "UGA", "UMI", "USA", "URY", "UZB",
+/*  "VA",  "VC",  "VE",  "VG",  "VI",  "VN",  "VU",  "WF",     */
+    "VAT", "VCT", "VEN", "VGB", "VIR", "VNM", "VUT", "WLF",
+/*  "WS",  "YE",  "YT",  "ZA",  "ZM",  "ZW",          */
+    "WSM", "YEM", "MYT", "ZAF", "ZMB", "ZWE",
+NULL,
+/*  "AN",  "BU",  "CS",  "FX",  "RO", "SU",  "TP",  "YD",  "YU",  "ZR" */
+    "ANT", "BUR", "SCG", "FXX", "ROM", "SUN", "TMP", "YMD", "YUG", "ZAR",
+NULL
+};
+
+typedef struct CanonicalizationMap {
+    const char *id;          /* input ID */
+    const char *canonicalID; /* canonicalized output ID */
+} CanonicalizationMap;
+
+/**
+ * A map to canonicalize locale IDs.  This handles a variety of
+ * different semantic kinds of transformations.
+ */
+static const CanonicalizationMap CANONICALIZE_MAP[] = {
+    { "art__LOJBAN",    "jbo" }, /* registered name */
+    { "hy__AREVELA",    "hy" }, /* Registered IANA variant */
+    { "hy__AREVMDA",    "hyw" }, /* Registered IANA variant */
+    { "zh__GUOYU",      "zh" }, /* registered name */
+    { "zh__HAKKA",      "hak" }, /* registered name */
+    { "zh__XIANG",      "hsn" }, /* registered name */
+    // subtags with 3 chars won't be treated as variants.
+    { "zh_GAN",         "gan" }, /* registered name */
+    { "zh_MIN_NAN",     "nan" }, /* registered name */
+    { "zh_WUU",         "wuu" }, /* registered name */
+    { "zh_YUE",         "yue" }, /* registered name */
+};
+
+/* ### BCP47 Conversion *******************************************/
+/* Test if the locale id has BCP47 u extension and does not have '@' */
+#define _hasBCP47Extension(id) (id && uprv_strstr(id, "@") == NULL && getShortestSubtagLength(localeID) == 1)
+/* Converts the BCP47 id to Unicode id. Does nothing to id if conversion fails */
+#define _ConvertBCP47(finalID, id, buffer, length,err) UPRV_BLOCK_MACRO_BEGIN { \
+    if (uloc_forLanguageTag(id, buffer, length, NULL, err) <= 0 || \
+            U_FAILURE(*err) || *err == U_STRING_NOT_TERMINATED_WARNING) { \
+        finalID=id; \
+        if (*err == U_STRING_NOT_TERMINATED_WARNING) { *err = U_BUFFER_OVERFLOW_ERROR; } \
+    } else { \
+        finalID=buffer; \
+    } \
+} UPRV_BLOCK_MACRO_END
+/* Gets the size of the shortest subtag in the given localeID. */
+static int32_t getShortestSubtagLength(const char *localeID) {
+    int32_t localeIDLength = static_cast<int32_t>(uprv_strlen(localeID));
+    int32_t length = localeIDLength;
+    int32_t tmpLength = 0;
+    int32_t i;
+    UBool reset = TRUE;
+
+    for (i = 0; i < localeIDLength; i++) {
+        if (localeID[i] != '_' && localeID[i] != '-') {
+            if (reset) {
+                tmpLength = 0;
+                reset = FALSE;
+            }
+            tmpLength++;
+        } else {
+            if (tmpLength != 0 && tmpLength < length) {
+                length = tmpLength;
+            }
+            reset = TRUE;
+        }
+    }
+
+    return length;
+}
+
+/* ### Keywords **************************************************/
+#define UPRV_ISDIGIT(c) (((c) >= '0') && ((c) <= '9'))
+#define UPRV_ISALPHANUM(c) (uprv_isASCIILetter(c) || UPRV_ISDIGIT(c) )
+/* Punctuation/symbols allowed in legacy key values */
+#define UPRV_OK_VALUE_PUNCTUATION(c) ((c) == '_' || (c) == '-' || (c) == '+' || (c) == '/')
+
+#define ULOC_KEYWORD_BUFFER_LEN 25
+#define ULOC_MAX_NO_KEYWORDS 25
+
+U_CAPI const char * U_EXPORT2
+locale_getKeywordsStart(const char *localeID) {
+    const char *result = NULL;
+    if((result = uprv_strchr(localeID, '@')) != NULL) {
+        return result;
+    }
+#if (U_CHARSET_FAMILY == U_EBCDIC_FAMILY)
+    else {
+        /* We do this because the @ sign is variant, and the @ sign used on one
+        EBCDIC machine won't be compiled the same way on other EBCDIC based
+        machines. */
+        static const uint8_t ebcdicSigns[] = { 0x7C, 0x44, 0x66, 0x80, 0xAC, 0xAE, 0xAF, 0xB5, 0xEC, 0xEF, 0x00 };
+        const uint8_t *charToFind = ebcdicSigns;
+        while(*charToFind) {
+            if((result = uprv_strchr(localeID, *charToFind)) != NULL) {
+                return result;
+            }
+            charToFind++;
+        }
+    }
+#endif
+    return NULL;
+}
+
+/**
+ * @param buf buffer of size [ULOC_KEYWORD_BUFFER_LEN]
+ * @param keywordName incoming name to be canonicalized
+ * @param status return status (keyword too long)
+ * @return length of the keyword name
+ */
+static int32_t locale_canonKeywordName(char *buf, const char *keywordName, UErrorCode *status)
+{
+  int32_t keywordNameLen = 0;
+
+  for (; *keywordName != 0; keywordName++) {
+    if (!UPRV_ISALPHANUM(*keywordName)) {
+      *status = U_ILLEGAL_ARGUMENT_ERROR; /* malformed keyword name */
+      return 0;
+    }
+    if (keywordNameLen < ULOC_KEYWORD_BUFFER_LEN - 1) {
+      buf[keywordNameLen++] = uprv_tolower(*keywordName);
+    } else {
+      /* keyword name too long for internal buffer */
+      *status = U_INTERNAL_PROGRAM_ERROR;
+      return 0;
+    }
+  }
+  if (keywordNameLen == 0) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR; /* empty keyword name */
+    return 0;
+  }
+  buf[keywordNameLen] = 0; /* terminate */
+
+  return keywordNameLen;
+}
+
+typedef struct {
+    char keyword[ULOC_KEYWORD_BUFFER_LEN];
+    int32_t keywordLen;
+    const char *valueStart;
+    int32_t valueLen;
+} KeywordStruct;
+
+static int32_t U_CALLCONV
+compareKeywordStructs(const void * /*context*/, const void *left, const void *right) {
+    const char* leftString = ((const KeywordStruct *)left)->keyword;
+    const char* rightString = ((const KeywordStruct *)right)->keyword;
+    return uprv_strcmp(leftString, rightString);
+}
+
+U_CFUNC void
+ulocimp_getKeywords(const char *localeID,
+                    char prev,
+                    ByteSink& sink,
+                    UBool valuesToo,
+                    UErrorCode *status)
+{
+    KeywordStruct keywordList[ULOC_MAX_NO_KEYWORDS];
+
+    int32_t maxKeywords = ULOC_MAX_NO_KEYWORDS;
+    int32_t numKeywords = 0;
+    const char* pos = localeID;
+    const char* equalSign = NULL;
+    const char* semicolon = NULL;
+    int32_t i = 0, j, n;
+
+    if(prev == '@') { /* start of keyword definition */
+        /* we will grab pairs, trim spaces, lowercase keywords, sort and return */
+        do {
+            UBool duplicate = FALSE;
+            /* skip leading spaces */
+            while(*pos == ' ') {
+                pos++;
+            }
+            if (!*pos) { /* handle trailing "; " */
+                break;
+            }
+            if(numKeywords == maxKeywords) {
+                *status = U_INTERNAL_PROGRAM_ERROR;
+                return;
+            }
+            equalSign = uprv_strchr(pos, '=');
+            semicolon = uprv_strchr(pos, ';');
+            /* lack of '=' [foo@currency] is illegal */
+            /* ';' before '=' [foo@currency;collation=pinyin] is illegal */
+            if(!equalSign || (semicolon && semicolon<equalSign)) {
+                *status = U_INVALID_FORMAT_ERROR;
+                return;
+            }
+            /* need to normalize both keyword and keyword name */
+            if(equalSign - pos >= ULOC_KEYWORD_BUFFER_LEN) {
+                /* keyword name too long for internal buffer */
+                *status = U_INTERNAL_PROGRAM_ERROR;
+                return;
+            }
+            for(i = 0, n = 0; i < equalSign - pos; ++i) {
+                if (pos[i] != ' ') {
+                    keywordList[numKeywords].keyword[n++] = uprv_tolower(pos[i]);
+                }
+            }
+
+            /* zero-length keyword is an error. */
+            if (n == 0) {
+                *status = U_INVALID_FORMAT_ERROR;
+                return;
+            }
+
+            keywordList[numKeywords].keyword[n] = 0;
+            keywordList[numKeywords].keywordLen = n;
+            /* now grab the value part. First we skip the '=' */
+            equalSign++;
+            /* then we leading spaces */
+            while(*equalSign == ' ') {
+                equalSign++;
+            }
+
+            /* Premature end or zero-length value */
+            if (!*equalSign || equalSign == semicolon) {
+                *status = U_INVALID_FORMAT_ERROR;
+                return;
+            }
+
+            keywordList[numKeywords].valueStart = equalSign;
+
+            pos = semicolon;
+            i = 0;
+            if(pos) {
+                while(*(pos - i - 1) == ' ') {
+                    i++;
+                }
+                keywordList[numKeywords].valueLen = (int32_t)(pos - equalSign - i);
+                pos++;
+            } else {
+                i = (int32_t)uprv_strlen(equalSign);
+                while(i && equalSign[i-1] == ' ') {
+                    i--;
+                }
+                keywordList[numKeywords].valueLen = i;
+            }
+            /* If this is a duplicate keyword, then ignore it */
+            for (j=0; j<numKeywords; ++j) {
+                if (uprv_strcmp(keywordList[j].keyword, keywordList[numKeywords].keyword) == 0) {
+                    duplicate = TRUE;
+                    break;
+                }
+            }
+            if (!duplicate) {
+                ++numKeywords;
+            }
+        } while(pos);
+
+        /* now we have a list of keywords */
+        /* we need to sort it */
+        uprv_sortArray(keywordList, numKeywords, sizeof(KeywordStruct), compareKeywordStructs, NULL, FALSE, status);
+
+        /* Now construct the keyword part */
+        for(i = 0; i < numKeywords; i++) {
+            sink.Append(keywordList[i].keyword, keywordList[i].keywordLen);
+            if(valuesToo) {
+                sink.Append("=", 1);
+                sink.Append(keywordList[i].valueStart, keywordList[i].valueLen);
+                if(i < numKeywords - 1) {
+                    sink.Append(";", 1);
+                }
+            } else {
+                sink.Append("\0", 1);
+            }
+        }
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getKeywordValue(const char* localeID,
+                     const char* keywordName,
+                     char* buffer, int32_t bufferCapacity,
+                     UErrorCode* status)
+{
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    CheckedArrayByteSink sink(buffer, bufferCapacity);
+    ulocimp_getKeywordValue(localeID, keywordName, sink, status);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*status)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(buffer, bufferCapacity, reslen, status);
+    }
+
+    return reslen;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_getKeywordValue(const char* localeID,
+                        const char* keywordName,
+                        icu::ByteSink& sink,
+                        UErrorCode* status)
+{
+    const char* startSearchHere = NULL;
+    const char* nextSeparator = NULL;
+    char keywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
+    char localeKeywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
+
+    if(status && U_SUCCESS(*status) && localeID) {
+      char tempBuffer[ULOC_FULLNAME_CAPACITY];
+      const char* tmpLocaleID;
+
+      if (keywordName == NULL || keywordName[0] == 0) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+      }
+
+      locale_canonKeywordName(keywordNameBuffer, keywordName, status);
+      if(U_FAILURE(*status)) {
+        return;
+      }
+
+      if (_hasBCP47Extension(localeID)) {
+          _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), status);
+      } else {
+          tmpLocaleID=localeID;
+      }
+
+      startSearchHere = locale_getKeywordsStart(tmpLocaleID);
+      if(startSearchHere == NULL) {
+          /* no keywords, return at once */
+          return;
+      }
+
+      /* find the first keyword */
+      while(startSearchHere) {
+          const char* keyValueTail;
+          int32_t keyValueLen;
+
+          startSearchHere++; /* skip @ or ; */
+          nextSeparator = uprv_strchr(startSearchHere, '=');
+          if(!nextSeparator) {
+              *status = U_ILLEGAL_ARGUMENT_ERROR; /* key must have =value */
+              return;
+          }
+          /* strip leading & trailing spaces (TC decided to tolerate these) */
+          while(*startSearchHere == ' ') {
+              startSearchHere++;
+          }
+          keyValueTail = nextSeparator;
+          while (keyValueTail > startSearchHere && *(keyValueTail-1) == ' ') {
+              keyValueTail--;
+          }
+          /* now keyValueTail points to first char after the keyName */
+          /* copy & normalize keyName from locale */
+          if (startSearchHere == keyValueTail) {
+              *status = U_ILLEGAL_ARGUMENT_ERROR; /* empty keyword name in passed-in locale */
+              return;
+          }
+          keyValueLen = 0;
+          while (startSearchHere < keyValueTail) {
+            if (!UPRV_ISALPHANUM(*startSearchHere)) {
+              *status = U_ILLEGAL_ARGUMENT_ERROR; /* malformed keyword name */
+              return;
+            }
+            if (keyValueLen < ULOC_KEYWORD_BUFFER_LEN - 1) {
+              localeKeywordNameBuffer[keyValueLen++] = uprv_tolower(*startSearchHere++);
+            } else {
+              /* keyword name too long for internal buffer */
+              *status = U_INTERNAL_PROGRAM_ERROR;
+              return;
+            }
+          }
+          localeKeywordNameBuffer[keyValueLen] = 0; /* terminate */
+
+          startSearchHere = uprv_strchr(nextSeparator, ';');
+
+          if(uprv_strcmp(keywordNameBuffer, localeKeywordNameBuffer) == 0) {
+               /* current entry matches the keyword. */
+             nextSeparator++; /* skip '=' */
+              /* First strip leading & trailing spaces (TC decided to tolerate these) */
+              while(*nextSeparator == ' ') {
+                nextSeparator++;
+              }
+              keyValueTail = (startSearchHere)? startSearchHere: nextSeparator + uprv_strlen(nextSeparator);
+              while(keyValueTail > nextSeparator && *(keyValueTail-1) == ' ') {
+                keyValueTail--;
+              }
+              /* Now copy the value, but check well-formedness */
+              if (nextSeparator == keyValueTail) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR; /* empty key value name in passed-in locale */
+                return;
+              }
+              while (nextSeparator < keyValueTail) {
+                if (!UPRV_ISALPHANUM(*nextSeparator) && !UPRV_OK_VALUE_PUNCTUATION(*nextSeparator)) {
+                  *status = U_ILLEGAL_ARGUMENT_ERROR; /* malformed key value */
+                  return;
+                }
+                /* Should we lowercase value to return here? Tests expect as-is. */
+                sink.Append(nextSeparator++, 1);
+              }
+              return;
+          }
+      }
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_setKeywordValue(const char* keywordName,
+                     const char* keywordValue,
+                     char* buffer, int32_t bufferCapacity,
+                     UErrorCode* status)
+{
+    /* TODO: sorting. removal. */
+    int32_t keywordNameLen;
+    int32_t keywordValueLen;
+    int32_t bufLen;
+    int32_t needLen = 0;
+    char keywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
+    char keywordValueBuffer[ULOC_KEYWORDS_CAPACITY+1];
+    char localeKeywordNameBuffer[ULOC_KEYWORD_BUFFER_LEN];
+    int32_t rc;
+    char* nextSeparator = NULL;
+    char* nextEqualsign = NULL;
+    char* startSearchHere = NULL;
+    char* keywordStart = NULL;
+    CharString updatedKeysAndValues;
+    UBool handledInputKeyAndValue = FALSE;
+    char keyValuePrefix = '@';
+
+    if(U_FAILURE(*status)) {
+        return -1;
+    }
+    if (keywordName == NULL || keywordName[0] == 0 || bufferCapacity <= 1) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    bufLen = (int32_t)uprv_strlen(buffer);
+    if(bufferCapacity<bufLen) {
+        /* The capacity is less than the length?! Is this NULL terminated? */
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    keywordNameLen = locale_canonKeywordName(keywordNameBuffer, keywordName, status);
+    if(U_FAILURE(*status)) {
+        return 0;
+    }
+
+    keywordValueLen = 0;
+    if(keywordValue) {
+        while (*keywordValue != 0) {
+            if (!UPRV_ISALPHANUM(*keywordValue) && !UPRV_OK_VALUE_PUNCTUATION(*keywordValue)) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR; /* malformed key value */
+                return 0;
+            }
+            if (keywordValueLen < ULOC_KEYWORDS_CAPACITY) {
+                /* Should we force lowercase in value to set? */
+                keywordValueBuffer[keywordValueLen++] = *keywordValue++;
+            } else {
+                /* keywordValue too long for internal buffer */
+                *status = U_INTERNAL_PROGRAM_ERROR;
+                return 0;
+            }
+        }
+    }
+    keywordValueBuffer[keywordValueLen] = 0; /* terminate */
+
+    startSearchHere = (char*)locale_getKeywordsStart(buffer);
+    if(startSearchHere == NULL || (startSearchHere[1]==0)) {
+        if(keywordValueLen == 0) { /* no keywords = nothing to remove */
+            return bufLen;
+        }
+
+        needLen = bufLen+1+keywordNameLen+1+keywordValueLen;
+        if(startSearchHere) { /* had a single @ */
+            needLen--; /* already had the @ */
+            /* startSearchHere points at the @ */
+        } else {
+            startSearchHere=buffer+bufLen;
+        }
+        if(needLen >= bufferCapacity) {
+            *status = U_BUFFER_OVERFLOW_ERROR;
+            return needLen; /* no change */
+        }
+        *startSearchHere++ = '@';
+        uprv_strcpy(startSearchHere, keywordNameBuffer);
+        startSearchHere += keywordNameLen;
+        *startSearchHere++ = '=';
+        uprv_strcpy(startSearchHere, keywordValueBuffer);
+        return needLen;
+    } /* end shortcut - no @ */
+
+    keywordStart = startSearchHere;
+    /* search for keyword */
+    while(keywordStart) {
+        const char* keyValueTail;
+        int32_t keyValueLen;
+
+        keywordStart++; /* skip @ or ; */
+        nextEqualsign = uprv_strchr(keywordStart, '=');
+        if (!nextEqualsign) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR; /* key must have =value */
+            return 0;
+        }
+        /* strip leading & trailing spaces (TC decided to tolerate these) */
+        while(*keywordStart == ' ') {
+            keywordStart++;
+        }
+        keyValueTail = nextEqualsign;
+        while (keyValueTail > keywordStart && *(keyValueTail-1) == ' ') {
+            keyValueTail--;
+        }
+        /* now keyValueTail points to first char after the keyName */
+        /* copy & normalize keyName from locale */
+        if (keywordStart == keyValueTail) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR; /* empty keyword name in passed-in locale */
+            return 0;
+        }
+        keyValueLen = 0;
+        while (keywordStart < keyValueTail) {
+            if (!UPRV_ISALPHANUM(*keywordStart)) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR; /* malformed keyword name */
+                return 0;
+            }
+            if (keyValueLen < ULOC_KEYWORD_BUFFER_LEN - 1) {
+                localeKeywordNameBuffer[keyValueLen++] = uprv_tolower(*keywordStart++);
+            } else {
+                /* keyword name too long for internal buffer */
+                *status = U_INTERNAL_PROGRAM_ERROR;
+                return 0;
+            }
+        }
+        localeKeywordNameBuffer[keyValueLen] = 0; /* terminate */
+
+        nextSeparator = uprv_strchr(nextEqualsign, ';');
+
+        /* start processing the value part */
+        nextEqualsign++; /* skip '=' */
+        /* First strip leading & trailing spaces (TC decided to tolerate these) */
+        while(*nextEqualsign == ' ') {
+            nextEqualsign++;
+        }
+        keyValueTail = (nextSeparator)? nextSeparator: nextEqualsign + uprv_strlen(nextEqualsign);
+        while(keyValueTail > nextEqualsign && *(keyValueTail-1) == ' ') {
+            keyValueTail--;
+        }
+        if (nextEqualsign == keyValueTail) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR; /* empty key value in passed-in locale */
+            return 0;
+        }
+
+        rc = uprv_strcmp(keywordNameBuffer, localeKeywordNameBuffer);
+        if(rc == 0) {
+            /* Current entry matches the input keyword. Update the entry */
+            if(keywordValueLen > 0) { /* updating a value */
+                updatedKeysAndValues.append(keyValuePrefix, *status);
+                keyValuePrefix = ';'; /* for any subsequent key-value pair */
+                updatedKeysAndValues.append(keywordNameBuffer, keywordNameLen, *status);
+                updatedKeysAndValues.append('=', *status);
+                updatedKeysAndValues.append(keywordValueBuffer, keywordValueLen, *status);
+            } /* else removing this entry, don't emit anything */
+            handledInputKeyAndValue = TRUE;
+        } else {
+           /* input keyword sorts earlier than current entry, add before current entry */
+            if (rc < 0 && keywordValueLen > 0 && !handledInputKeyAndValue) {
+                /* insert new entry at this location */
+                updatedKeysAndValues.append(keyValuePrefix, *status);
+                keyValuePrefix = ';'; /* for any subsequent key-value pair */
+                updatedKeysAndValues.append(keywordNameBuffer, keywordNameLen, *status);
+                updatedKeysAndValues.append('=', *status);
+                updatedKeysAndValues.append(keywordValueBuffer, keywordValueLen, *status);
+                handledInputKeyAndValue = TRUE;
+            }
+            /* copy the current entry */
+            updatedKeysAndValues.append(keyValuePrefix, *status);
+            keyValuePrefix = ';'; /* for any subsequent key-value pair */
+            updatedKeysAndValues.append(localeKeywordNameBuffer, keyValueLen, *status);
+            updatedKeysAndValues.append('=', *status);
+            updatedKeysAndValues.append(nextEqualsign, static_cast<int32_t>(keyValueTail-nextEqualsign), *status);
+        }
+        if (!nextSeparator && keywordValueLen > 0 && !handledInputKeyAndValue) {
+            /* append new entry at the end, it sorts later than existing entries */
+            updatedKeysAndValues.append(keyValuePrefix, *status);
+            /* skip keyValuePrefix update, no subsequent key-value pair */
+            updatedKeysAndValues.append(keywordNameBuffer, keywordNameLen, *status);
+            updatedKeysAndValues.append('=', *status);
+            updatedKeysAndValues.append(keywordValueBuffer, keywordValueLen, *status);
+            handledInputKeyAndValue = TRUE;
+        }
+        keywordStart = nextSeparator;
+    } /* end loop searching */
+
+    /* Any error from updatedKeysAndValues.append above would be internal and not due to
+     * problems with the passed-in locale. So if we did encounter problems with the
+     * passed-in locale above, those errors took precedence and overrode any error
+     * status from updatedKeysAndValues.append, and also caused a return of 0. If there
+     * are errors here they are from updatedKeysAndValues.append; they do cause an
+     * error return but the passed-in locale is unmodified and the original bufLen is
+     * returned.
+     */
+    if (!handledInputKeyAndValue || U_FAILURE(*status)) {
+        /* if input key/value specified removal of a keyword not present in locale, or
+         * there was an error in CharString.append, leave original locale alone. */
+        return bufLen;
+    }
+
+    // needLen = length of the part before '@'
+    needLen = (int32_t)(startSearchHere - buffer);
+    return needLen + updatedKeysAndValues.extract(
+                         startSearchHere, bufferCapacity - needLen, *status);
+}
+
+/* ### ID parsing implementation **************************************************/
+
+#define _isPrefixLetter(a) ((a=='x')||(a=='X')||(a=='i')||(a=='I'))
+
+/*returns TRUE if one of the special prefixes is here (s=string)
+  'x-' or 'i-' */
+#define _isIDPrefix(s) (_isPrefixLetter(s[0])&&_isIDSeparator(s[1]))
+
+/* Dot terminates it because of POSIX form  where dot precedes the codepage
+ * except for variant
+ */
+#define _isTerminator(a)  ((a==0)||(a=='.')||(a=='@'))
+
+/**
+ * Lookup 'key' in the array 'list'.  The array 'list' should contain
+ * a NULL entry, followed by more entries, and a second NULL entry.
+ *
+ * The 'list' param should be LANGUAGES, LANGUAGES_3, COUNTRIES, or
+ * COUNTRIES_3.
+ */
+static int16_t _findIndex(const char* const* list, const char* key)
+{
+    const char* const* anchor = list;
+    int32_t pass = 0;
+
+    /* Make two passes through two NULL-terminated arrays at 'list' */
+    while (pass++ < 2) {
+        while (*list) {
+            if (uprv_strcmp(key, *list) == 0) {
+                return (int16_t)(list - anchor);
+            }
+            list++;
+        }
+        ++list;     /* skip final NULL *CWB*/
+    }
+    return -1;
+}
+
+U_CFUNC const char*
+uloc_getCurrentCountryID(const char* oldID){
+    int32_t offset = _findIndex(DEPRECATED_COUNTRIES, oldID);
+    if (offset >= 0) {
+        return REPLACEMENT_COUNTRIES[offset];
+    }
+    return oldID;
+}
+U_CFUNC const char*
+uloc_getCurrentLanguageID(const char* oldID){
+    int32_t offset = _findIndex(DEPRECATED_LANGUAGES, oldID);
+    if (offset >= 0) {
+        return REPLACEMENT_LANGUAGES[offset];
+    }
+    return oldID;
+}
+/*
+ * the internal functions _getLanguage(), _getCountry(), _getVariant()
+ * avoid duplicating code to handle the earlier locale ID pieces
+ * in the functions for the later ones by
+ * setting the *pEnd pointer to where they stopped parsing
+ *
+ * TODO try to use this in Locale
+ */
+CharString U_EXPORT2
+ulocimp_getLanguage(const char *localeID,
+                    const char **pEnd,
+                    UErrorCode &status) {
+    CharString result;
+
+    if (uprv_stricmp(localeID, "root") == 0) {
+        localeID += 4;
+    } else if (uprv_strnicmp(localeID, "und", 3) == 0 &&
+               (localeID[3] == '\0' ||
+                localeID[3] == '-' ||
+                localeID[3] == '_' ||
+                localeID[3] == '@')) {
+        localeID += 3;
+    }
+
+    /* if it starts with i- or x- then copy that prefix */
+    if(_isIDPrefix(localeID)) {
+        result.append((char)uprv_tolower(*localeID), status);
+        result.append('-', status);
+        localeID+=2;
+    }
+
+    /* copy the language as far as possible and count its length */
+    while(!_isTerminator(*localeID) && !_isIDSeparator(*localeID)) {
+        result.append((char)uprv_tolower(*localeID), status);
+        localeID++;
+    }
+
+    if(result.length()==3) {
+        /* convert 3 character code to 2 character code if possible *CWB*/
+        int32_t offset = _findIndex(LANGUAGES_3, result.data());
+        if(offset>=0) {
+            result.clear();
+            result.append(LANGUAGES[offset], status);
+        }
+    }
+
+    if(pEnd!=NULL) {
+        *pEnd=localeID;
+    }
+
+    return result;
+}
+
+CharString U_EXPORT2
+ulocimp_getScript(const char *localeID,
+                  const char **pEnd,
+                  UErrorCode &status) {
+    CharString result;
+    int32_t idLen = 0;
+
+    if (pEnd != NULL) {
+        *pEnd = localeID;
+    }
+
+    /* copy the second item as far as possible and count its length */
+    while(!_isTerminator(localeID[idLen]) && !_isIDSeparator(localeID[idLen])
+            && uprv_isASCIILetter(localeID[idLen])) {
+        idLen++;
+    }
+
+    /* If it's exactly 4 characters long, then it's a script and not a country. */
+    if (idLen == 4) {
+        int32_t i;
+        if (pEnd != NULL) {
+            *pEnd = localeID+idLen;
+        }
+        if (idLen >= 1) {
+            result.append((char)uprv_toupper(*(localeID++)), status);
+        }
+        for (i = 1; i < idLen; i++) {
+            result.append((char)uprv_tolower(*(localeID++)), status);
+        }
+    }
+
+    return result;
+}
+
+CharString U_EXPORT2
+ulocimp_getCountry(const char *localeID,
+                   const char **pEnd,
+                   UErrorCode &status) {
+    CharString result;
+    int32_t idLen=0;
+
+    /* copy the country as far as possible and count its length */
+    while(!_isTerminator(localeID[idLen]) && !_isIDSeparator(localeID[idLen])) {
+        result.append((char)uprv_toupper(localeID[idLen]), status);
+        idLen++;
+    }
+
+    /* the country should be either length 2 or 3 */
+    if (idLen == 2 || idLen == 3) {
+        /* convert 3 character code to 2 character code if possible *CWB*/
+        if(idLen==3) {
+            int32_t offset = _findIndex(COUNTRIES_3, result.data());
+            if(offset>=0) {
+                result.clear();
+                result.append(COUNTRIES[offset], status);
+            }
+        }
+        localeID+=idLen;
+    } else {
+        result.clear();
+    }
+
+    if(pEnd!=NULL) {
+        *pEnd=localeID;
+    }
+
+    return result;
+}
+
+/**
+ * @param needSeparator if true, then add leading '_' if any variants
+ * are added to 'variant'
+ */
+static void
+_getVariant(const char *localeID,
+            char prev,
+            ByteSink& sink,
+            UBool needSeparator) {
+    UBool hasVariant = FALSE;
+
+    /* get one or more variant tags and separate them with '_' */
+    if(_isIDSeparator(prev)) {
+        /* get a variant string after a '-' or '_' */
+        while(!_isTerminator(*localeID)) {
+            if (needSeparator) {
+                sink.Append("_", 1);
+                needSeparator = FALSE;
+            }
+            char c = (char)uprv_toupper(*localeID);
+            if (c == '-') c = '_';
+            sink.Append(&c, 1);
+            hasVariant = TRUE;
+            localeID++;
+        }
+    }
+
+    /* if there is no variant tag after a '-' or '_' then look for '@' */
+    if(!hasVariant) {
+        if(prev=='@') {
+            /* keep localeID */
+        } else if((localeID=locale_getKeywordsStart(localeID))!=NULL) {
+            ++localeID; /* point after the '@' */
+        } else {
+            return;
+        }
+        while(!_isTerminator(*localeID)) {
+            if (needSeparator) {
+                sink.Append("_", 1);
+                needSeparator = FALSE;
+            }
+            char c = (char)uprv_toupper(*localeID);
+            if (c == '-' || c == ',') c = '_';
+            sink.Append(&c, 1);
+            localeID++;
+        }
+    }
+}
+
+/* Keyword enumeration */
+
+typedef struct UKeywordsContext {
+    char* keywords;
+    char* current;
+} UKeywordsContext;
+
+U_CDECL_BEGIN
+
+static void U_CALLCONV
+uloc_kw_closeKeywords(UEnumeration *enumerator) {
+    uprv_free(((UKeywordsContext *)enumerator->context)->keywords);
+    uprv_free(enumerator->context);
+    uprv_free(enumerator);
+}
+
+static int32_t U_CALLCONV
+uloc_kw_countKeywords(UEnumeration *en, UErrorCode * /*status*/) {
+    char *kw = ((UKeywordsContext *)en->context)->keywords;
+    int32_t result = 0;
+    while(*kw) {
+        result++;
+        kw += uprv_strlen(kw)+1;
+    }
+    return result;
+}
+
+static const char * U_CALLCONV
+uloc_kw_nextKeyword(UEnumeration* en,
+                    int32_t* resultLength,
+                    UErrorCode* /*status*/) {
+    const char* result = ((UKeywordsContext *)en->context)->current;
+    int32_t len = 0;
+    if(*result) {
+        len = (int32_t)uprv_strlen(((UKeywordsContext *)en->context)->current);
+        ((UKeywordsContext *)en->context)->current += len+1;
+    } else {
+        result = NULL;
+    }
+    if (resultLength) {
+        *resultLength = len;
+    }
+    return result;
+}
+
+static void U_CALLCONV
+uloc_kw_resetKeywords(UEnumeration* en,
+                      UErrorCode* /*status*/) {
+    ((UKeywordsContext *)en->context)->current = ((UKeywordsContext *)en->context)->keywords;
+}
+
+U_CDECL_END
+
+
+static const UEnumeration gKeywordsEnum = {
+    NULL,
+    NULL,
+    uloc_kw_closeKeywords,
+    uloc_kw_countKeywords,
+    uenum_unextDefault,
+    uloc_kw_nextKeyword,
+    uloc_kw_resetKeywords
+};
+
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openKeywordList(const char *keywordList, int32_t keywordListSize, UErrorCode* status)
+{
+    LocalMemory<UKeywordsContext> myContext;
+    LocalMemory<UEnumeration> result;
+
+    if (U_FAILURE(*status)) {
+        return nullptr;
+    }
+    myContext.adoptInstead(static_cast<UKeywordsContext *>(uprv_malloc(sizeof(UKeywordsContext))));
+    result.adoptInstead(static_cast<UEnumeration *>(uprv_malloc(sizeof(UEnumeration))));
+    if (myContext.isNull() || result.isNull()) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+    uprv_memcpy(result.getAlias(), &gKeywordsEnum, sizeof(UEnumeration));
+    myContext->keywords = static_cast<char *>(uprv_malloc(keywordListSize+1));
+    if (myContext->keywords == nullptr) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return nullptr;
+    }
+    uprv_memcpy(myContext->keywords, keywordList, keywordListSize);
+    myContext->keywords[keywordListSize] = 0;
+    myContext->current = myContext->keywords;
+    result->context = myContext.orphan();
+    return result.orphan();
+}
+
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openKeywords(const char* localeID,
+                        UErrorCode* status)
+{
+    char tempBuffer[ULOC_FULLNAME_CAPACITY];
+    const char* tmpLocaleID;
+
+    if(status==NULL || U_FAILURE(*status)) {
+        return 0;
+    }
+
+    if (_hasBCP47Extension(localeID)) {
+        _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), status);
+    } else {
+        if (localeID==NULL) {
+           localeID=uloc_getDefault();
+        }
+        tmpLocaleID=localeID;
+    }
+
+    /* Skip the language */
+    ulocimp_getLanguage(tmpLocaleID, &tmpLocaleID, *status);
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    if(_isIDSeparator(*tmpLocaleID)) {
+        const char *scriptID;
+        /* Skip the script if available */
+        ulocimp_getScript(tmpLocaleID+1, &scriptID, *status);
+        if (U_FAILURE(*status)) {
+            return 0;
+        }
+        if(scriptID != tmpLocaleID+1) {
+            /* Found optional script */
+            tmpLocaleID = scriptID;
+        }
+        /* Skip the Country */
+        if (_isIDSeparator(*tmpLocaleID)) {
+            ulocimp_getCountry(tmpLocaleID+1, &tmpLocaleID, *status);
+            if (U_FAILURE(*status)) {
+                return 0;
+            }
+        }
+    }
+
+    /* keywords are located after '@' */
+    if((tmpLocaleID = locale_getKeywordsStart(tmpLocaleID)) != NULL) {
+        CharString keywords;
+        CharStringByteSink sink(&keywords);
+        ulocimp_getKeywords(tmpLocaleID+1, '@', sink, FALSE, status);
+        if (U_FAILURE(*status)) {
+            return NULL;
+        }
+        return uloc_openKeywordList(keywords.data(), keywords.length(), status);
+    }
+    return NULL;
+}
+
+
+/* bit-flags for 'options' parameter of _canonicalize */
+#define _ULOC_STRIP_KEYWORDS 0x2
+#define _ULOC_CANONICALIZE   0x1
+
+#define OPTION_SET(options, mask) ((options & mask) != 0)
+
+static const char i_default[] = {'i', '-', 'd', 'e', 'f', 'a', 'u', 'l', 't'};
+#define I_DEFAULT_LENGTH UPRV_LENGTHOF(i_default)
+
+/**
+ * Canonicalize the given localeID, to level 1 or to level 2,
+ * depending on the options.  To specify level 1, pass in options=0.
+ * To specify level 2, pass in options=_ULOC_CANONICALIZE.
+ *
+ * This is the code underlying uloc_getName and uloc_canonicalize.
+ */
+static void
+_canonicalize(const char* localeID,
+              ByteSink& sink,
+              uint32_t options,
+              UErrorCode* err) {
+    int32_t j, fieldCount=0, scriptSize=0, variantSize=0;
+    char tempBuffer[ULOC_FULLNAME_CAPACITY];
+    const char* origLocaleID;
+    const char* tmpLocaleID;
+    const char* keywordAssign = NULL;
+    const char* separatorIndicator = NULL;
+
+    if (U_FAILURE(*err)) {
+        return;
+    }
+
+    if (_hasBCP47Extension(localeID)) {
+        _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), err);
+    } else {
+        if (localeID==NULL) {
+           localeID=uloc_getDefault();
+        }
+        tmpLocaleID=localeID;
+    }
+
+    origLocaleID=tmpLocaleID;
+
+    /* get all pieces, one after another, and separate with '_' */
+    CharString tag = ulocimp_getLanguage(tmpLocaleID, &tmpLocaleID, *err);
+
+    if (tag.length() == I_DEFAULT_LENGTH &&
+            uprv_strncmp(origLocaleID, i_default, I_DEFAULT_LENGTH) == 0) {
+        tag.clear();
+        tag.append(uloc_getDefault(), *err);
+    } else if(_isIDSeparator(*tmpLocaleID)) {
+        const char *scriptID;
+
+        ++fieldCount;
+        tag.append('_', *err);
+
+        CharString script = ulocimp_getScript(tmpLocaleID+1, &scriptID, *err);
+        tag.append(script, *err);
+        scriptSize = script.length();
+        if(scriptSize > 0) {
+            /* Found optional script */
+            tmpLocaleID = scriptID;
+            ++fieldCount;
+            if (_isIDSeparator(*tmpLocaleID)) {
+                /* If there is something else, then we add the _ */
+                tag.append('_', *err);
+            }
+        }
+
+        if (_isIDSeparator(*tmpLocaleID)) {
+            const char *cntryID;
+
+            CharString country = ulocimp_getCountry(tmpLocaleID+1, &cntryID, *err);
+            tag.append(country, *err);
+            if (!country.isEmpty()) {
+                /* Found optional country */
+                tmpLocaleID = cntryID;
+            }
+            if(_isIDSeparator(*tmpLocaleID)) {
+                /* If there is something else, then we add the _  if we found country before. */
+                if (!_isIDSeparator(*(tmpLocaleID+1))) {
+                    ++fieldCount;
+                    tag.append('_', *err);
+                }
+
+                variantSize = -tag.length();
+                {
+                    CharStringByteSink s(&tag);
+                    _getVariant(tmpLocaleID+1, *tmpLocaleID, s, FALSE);
+                }
+                variantSize += tag.length();
+                if (variantSize > 0) {
+                    tmpLocaleID += variantSize + 1; /* skip '_' and variant */
+                }
+            }
+        }
+    }
+
+    /* Copy POSIX-style charset specifier, if any [mr.utf8] */
+    if (!OPTION_SET(options, _ULOC_CANONICALIZE) && *tmpLocaleID == '.') {
+        UBool done = FALSE;
+        do {
+            char c = *tmpLocaleID;
+            switch (c) {
+            case 0:
+            case '@':
+                done = TRUE;
+                break;
+            default:
+                tag.append(c, *err);
+                ++tmpLocaleID;
+                break;
+            }
+        } while (!done);
+    }
+
+    /* Scan ahead to next '@' and determine if it is followed by '=' and/or ';'
+       After this, tmpLocaleID either points to '@' or is NULL */
+    if ((tmpLocaleID=locale_getKeywordsStart(tmpLocaleID))!=NULL) {
+        keywordAssign = uprv_strchr(tmpLocaleID, '=');
+        separatorIndicator = uprv_strchr(tmpLocaleID, ';');
+    }
+
+    /* Copy POSIX-style variant, if any [mr@FOO] */
+    if (!OPTION_SET(options, _ULOC_CANONICALIZE) &&
+        tmpLocaleID != NULL && keywordAssign == NULL) {
+        for (;;) {
+            char c = *tmpLocaleID;
+            if (c == 0) {
+                break;
+            }
+            tag.append(c, *err);
+            ++tmpLocaleID;
+        }
+    }
+
+    if (OPTION_SET(options, _ULOC_CANONICALIZE)) {
+        /* Handle @FOO variant if @ is present and not followed by = */
+        if (tmpLocaleID!=NULL && keywordAssign==NULL) {
+            /* Add missing '_' if needed */
+            if (fieldCount < 2 || (fieldCount < 3 && scriptSize > 0)) {
+                do {
+                    tag.append('_', *err);
+                    ++fieldCount;
+                } while(fieldCount<2);
+            }
+
+            int32_t posixVariantSize = -tag.length();
+            {
+                CharStringByteSink s(&tag);
+                _getVariant(tmpLocaleID+1, '@', s, (UBool)(variantSize > 0));
+            }
+            posixVariantSize += tag.length();
+            if (posixVariantSize > 0) {
+                variantSize += posixVariantSize;
+            }
+        }
+
+        /* Look up the ID in the canonicalization map */
+        for (j=0; j<UPRV_LENGTHOF(CANONICALIZE_MAP); j++) {
+            StringPiece id(CANONICALIZE_MAP[j].id);
+            if (tag == id) {
+                if (id.empty() && tmpLocaleID != NULL) {
+                    break; /* Don't remap "" if keywords present */
+                }
+                tag.clear();
+                tag.append(CANONICALIZE_MAP[j].canonicalID, *err);
+                break;
+            }
+        }
+    }
+
+    sink.Append(tag.data(), tag.length());
+
+    if (!OPTION_SET(options, _ULOC_STRIP_KEYWORDS)) {
+        if (tmpLocaleID!=NULL && keywordAssign!=NULL &&
+            (!separatorIndicator || separatorIndicator > keywordAssign)) {
+            sink.Append("@", 1);
+            ++fieldCount;
+            ulocimp_getKeywords(tmpLocaleID+1, '@', sink, TRUE, err);
+        }
+    }
+}
+
+/* ### ID parsing API **************************************************/
+
+U_CAPI int32_t  U_EXPORT2
+uloc_getParent(const char*    localeID,
+               char* parent,
+               int32_t parentCapacity,
+               UErrorCode* err)
+{
+    const char *lastUnderscore;
+    int32_t i;
+
+    if (U_FAILURE(*err))
+        return 0;
+
+    if (localeID == NULL)
+        localeID = uloc_getDefault();
+
+    lastUnderscore=uprv_strrchr(localeID, '_');
+    if(lastUnderscore!=NULL) {
+        i=(int32_t)(lastUnderscore-localeID);
+    } else {
+        i=0;
+    }
+
+    if (i > 0) {
+        if (uprv_strnicmp(localeID, "und_", 4) == 0) {
+            localeID += 3;
+            i -= 3;
+            uprv_memmove(parent, localeID, uprv_min(i, parentCapacity));
+        } else if (parent != localeID) {
+            uprv_memcpy(parent, localeID, uprv_min(i, parentCapacity));
+        }
+    }
+
+    return u_terminateChars(parent, parentCapacity, i, err);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getLanguage(const char*    localeID,
+         char* language,
+         int32_t languageCapacity,
+         UErrorCode* err)
+{
+    /* uloc_getLanguage will return a 2 character iso-639 code if one exists. *CWB*/
+
+    if (err==NULL || U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(localeID==NULL) {
+        localeID=uloc_getDefault();
+    }
+
+    return ulocimp_getLanguage(localeID, NULL, *err).extract(language, languageCapacity, *err);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getScript(const char*    localeID,
+         char* script,
+         int32_t scriptCapacity,
+         UErrorCode* err)
+{
+    if(err==NULL || U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(localeID==NULL) {
+        localeID=uloc_getDefault();
+    }
+
+    /* skip the language */
+    ulocimp_getLanguage(localeID, &localeID, *err);
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(_isIDSeparator(*localeID)) {
+        return ulocimp_getScript(localeID+1, NULL, *err).extract(script, scriptCapacity, *err);
+    }
+    return u_terminateChars(script, scriptCapacity, 0, err);
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_getCountry(const char* localeID,
+            char* country,
+            int32_t countryCapacity,
+            UErrorCode* err)
+{
+    if(err==NULL || U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(localeID==NULL) {
+        localeID=uloc_getDefault();
+    }
+
+    /* Skip the language */
+    ulocimp_getLanguage(localeID, &localeID, *err);
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(_isIDSeparator(*localeID)) {
+        const char *scriptID;
+        /* Skip the script if available */
+        ulocimp_getScript(localeID+1, &scriptID, *err);
+        if (U_FAILURE(*err)) {
+            return 0;
+        }
+        if(scriptID != localeID+1) {
+            /* Found optional script */
+            localeID = scriptID;
+        }
+        if(_isIDSeparator(*localeID)) {
+            return ulocimp_getCountry(localeID+1, NULL, *err).extract(country, countryCapacity, *err);
+        }
+    }
+    return u_terminateChars(country, countryCapacity, 0, err);
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_getVariant(const char* localeID,
+                char* variant,
+                int32_t variantCapacity,
+                UErrorCode* err)
+{
+    char tempBuffer[ULOC_FULLNAME_CAPACITY];
+    const char* tmpLocaleID;
+    int32_t i=0;
+
+    if(err==NULL || U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if (_hasBCP47Extension(localeID)) {
+        _ConvertBCP47(tmpLocaleID, localeID, tempBuffer, sizeof(tempBuffer), err);
+    } else {
+        if (localeID==NULL) {
+           localeID=uloc_getDefault();
+        }
+        tmpLocaleID=localeID;
+    }
+
+    /* Skip the language */
+    ulocimp_getLanguage(tmpLocaleID, &tmpLocaleID, *err);
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    if(_isIDSeparator(*tmpLocaleID)) {
+        const char *scriptID;
+        /* Skip the script if available */
+        ulocimp_getScript(tmpLocaleID+1, &scriptID, *err);
+        if (U_FAILURE(*err)) {
+            return 0;
+        }
+        if(scriptID != tmpLocaleID+1) {
+            /* Found optional script */
+            tmpLocaleID = scriptID;
+        }
+        /* Skip the Country */
+        if (_isIDSeparator(*tmpLocaleID)) {
+            const char *cntryID;
+            ulocimp_getCountry(tmpLocaleID+1, &cntryID, *err);
+            if (U_FAILURE(*err)) {
+                return 0;
+            }
+            if (cntryID != tmpLocaleID+1) {
+                /* Found optional country */
+                tmpLocaleID = cntryID;
+            }
+            if(_isIDSeparator(*tmpLocaleID)) {
+                /* If there was no country ID, skip a possible extra IDSeparator */
+                if (tmpLocaleID != cntryID && _isIDSeparator(tmpLocaleID[1])) {
+                    tmpLocaleID++;
+                }
+
+                CheckedArrayByteSink sink(variant, variantCapacity);
+                _getVariant(tmpLocaleID+1, *tmpLocaleID, sink, FALSE);
+
+                i = sink.NumberOfBytesAppended();
+
+                if (U_FAILURE(*err)) {
+                    return i;
+                }
+
+                if (sink.Overflowed()) {
+                    *err = U_BUFFER_OVERFLOW_ERROR;
+                    return i;
+                }
+            }
+        }
+    }
+
+    return u_terminateChars(variant, variantCapacity, i, err);
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_getName(const char* localeID,
+             char* name,
+             int32_t nameCapacity,
+             UErrorCode* err)
+{
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    CheckedArrayByteSink sink(name, nameCapacity);
+    ulocimp_getName(localeID, sink, err);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*err)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(name, nameCapacity, reslen, err);
+    }
+
+    return reslen;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_getName(const char* localeID,
+                ByteSink& sink,
+                UErrorCode* err)
+{
+    _canonicalize(localeID, sink, 0, err);
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_getBaseName(const char* localeID,
+                 char* name,
+                 int32_t nameCapacity,
+                 UErrorCode* err)
+{
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    CheckedArrayByteSink sink(name, nameCapacity);
+    ulocimp_getBaseName(localeID, sink, err);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*err)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(name, nameCapacity, reslen, err);
+    }
+
+    return reslen;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_getBaseName(const char* localeID,
+                    ByteSink& sink,
+                    UErrorCode* err)
+{
+    _canonicalize(localeID, sink, _ULOC_STRIP_KEYWORDS, err);
+}
+
+U_CAPI int32_t  U_EXPORT2
+uloc_canonicalize(const char* localeID,
+                  char* name,
+                  int32_t nameCapacity,
+                  UErrorCode* err)
+{
+    if (U_FAILURE(*err)) {
+        return 0;
+    }
+
+    CheckedArrayByteSink sink(name, nameCapacity);
+    ulocimp_canonicalize(localeID, sink, err);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*err)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(name, nameCapacity, reslen, err);
+    }
+
+    return reslen;
+}
+
+U_CAPI void U_EXPORT2
+ulocimp_canonicalize(const char* localeID,
+                     ByteSink& sink,
+                     UErrorCode* err)
+{
+    _canonicalize(localeID, sink, _ULOC_CANONICALIZE, err);
+}
+
+U_CAPI const char*  U_EXPORT2
+uloc_getISO3Language(const char* localeID)
+{
+    int16_t offset;
+    char lang[ULOC_LANG_CAPACITY];
+    UErrorCode err = U_ZERO_ERROR;
+
+    if (localeID == NULL)
+    {
+        localeID = uloc_getDefault();
+    }
+    uloc_getLanguage(localeID, lang, ULOC_LANG_CAPACITY, &err);
+    if (U_FAILURE(err))
+        return "";
+    offset = _findIndex(LANGUAGES, lang);
+    if (offset < 0)
+        return "";
+    return LANGUAGES_3[offset];
+}
+
+U_CAPI const char*  U_EXPORT2
+uloc_getISO3Country(const char* localeID)
+{
+    int16_t offset;
+    char cntry[ULOC_LANG_CAPACITY];
+    UErrorCode err = U_ZERO_ERROR;
+
+    if (localeID == NULL)
+    {
+        localeID = uloc_getDefault();
+    }
+    uloc_getCountry(localeID, cntry, ULOC_LANG_CAPACITY, &err);
+    if (U_FAILURE(err))
+        return "";
+    offset = _findIndex(COUNTRIES, cntry);
+    if (offset < 0)
+        return "";
+
+    return COUNTRIES_3[offset];
+}
+
+U_CAPI uint32_t  U_EXPORT2
+uloc_getLCID(const char* localeID)
+{
+    UErrorCode status = U_ZERO_ERROR;
+    char       langID[ULOC_FULLNAME_CAPACITY];
+    uint32_t   lcid = 0;
+
+    /* Check for incomplete id. */
+    if (!localeID || uprv_strlen(localeID) < 2) {
+        return 0;
+    }
+
+    // First, attempt Windows platform lookup if available, but fall
+    // through to catch any special cases (ICU vs Windows name differences).
+    lcid = uprv_convertToLCIDPlatform(localeID, &status);
+    if (U_FAILURE(status)) {
+        return 0;
+    }
+    if (lcid > 0) {
+        // Windows found an LCID, return that
+        return lcid;
+    }
+
+    uloc_getLanguage(localeID, langID, sizeof(langID), &status);
+    if (U_FAILURE(status) || status == U_STRING_NOT_TERMINATED_WARNING) {
+        return 0;
+    }
+
+    if (uprv_strchr(localeID, '@')) {
+        // uprv_convertToLCID does not support keywords other than collation.
+        // Remove all keywords except collation.
+        int32_t len;
+        char tmpLocaleID[ULOC_FULLNAME_CAPACITY];
+
+        CharString collVal;
+        {
+            CharStringByteSink sink(&collVal);
+            ulocimp_getKeywordValue(localeID, "collation", sink, &status);
+        }
+
+        if (U_SUCCESS(status) && !collVal.isEmpty()) {
+            len = uloc_getBaseName(localeID, tmpLocaleID,
+                UPRV_LENGTHOF(tmpLocaleID) - 1, &status);
+
+            if (U_SUCCESS(status) && len > 0) {
+                tmpLocaleID[len] = 0;
+
+                len = uloc_setKeywordValue("collation", collVal.data(), tmpLocaleID,
+                    UPRV_LENGTHOF(tmpLocaleID) - len - 1, &status);
+
+                if (U_SUCCESS(status) && len > 0) {
+                    tmpLocaleID[len] = 0;
+                    return uprv_convertToLCID(langID, tmpLocaleID, &status);
+                }
+            }
+        }
+
+        // fall through - all keywords are simply ignored
+        status = U_ZERO_ERROR;
+    }
+
+    return uprv_convertToLCID(langID, localeID, &status);
+}
+
+U_CAPI int32_t U_EXPORT2
+uloc_getLocaleForLCID(uint32_t hostid, char *locale, int32_t localeCapacity,
+                UErrorCode *status)
+{
+    return uprv_convertToPosix(hostid, locale, localeCapacity, status);
+}
+
+/* ### Default locale **************************************************/
+
+U_CAPI const char*  U_EXPORT2
+uloc_getDefault()
+{
+    return locale_get_default();
+}
+
+U_CAPI void  U_EXPORT2
+uloc_setDefault(const char*   newDefaultLocale,
+             UErrorCode* err)
+{
+    if (U_FAILURE(*err))
+        return;
+    /* the error code isn't currently used for anything by this function*/
+
+    /* propagate change to C++ */
+    locale_set_default(newDefaultLocale);
+}
+
+/**
+ * Returns a list of all 2-letter language codes defined in ISO 639.  This is a pointer
+ * to an array of pointers to arrays of char.  All of these pointers are owned
+ * by ICU-- do not delete them, and do not write through them.  The array is
+ * terminated with a null pointer.
+ */
+U_CAPI const char* const*  U_EXPORT2
+uloc_getISOLanguages()
+{
+    return LANGUAGES;
+}
+
+/**
+ * Returns a list of all 2-letter country codes defined in ISO 639.  This is a
+ * pointer to an array of pointers to arrays of char.  All of these pointers are
+ * owned by ICU-- do not delete them, and do not write through them.  The array is
+ * terminated with a null pointer.
+ */
+U_CAPI const char* const*  U_EXPORT2
+uloc_getISOCountries()
+{
+    return COUNTRIES;
+}
+
+U_CAPI const char* U_EXPORT2
+uloc_toUnicodeLocaleKey(const char* keyword)
+{
+    const char* bcpKey = ulocimp_toBcpKey(keyword);
+    if (bcpKey == NULL && ultag_isUnicodeLocaleKey(keyword, -1)) {
+        // unknown keyword, but syntax is fine..
+        return keyword;
+    }
+    return bcpKey;
+}
+
+U_CAPI const char* U_EXPORT2
+uloc_toUnicodeLocaleType(const char* keyword, const char* value)
+{
+    const char* bcpType = ulocimp_toBcpType(keyword, value, NULL, NULL);
+    if (bcpType == NULL && ultag_isUnicodeLocaleType(value, -1)) {
+        // unknown keyword, but syntax is fine..
+        return value;
+    }
+    return bcpType;
+}
+
+static UBool
+isWellFormedLegacyKey(const char* legacyKey)
+{
+    const char* p = legacyKey;
+    while (*p) {
+        if (!UPRV_ISALPHANUM(*p)) {
+            return FALSE;
+        }
+        p++;
+    }
+    return TRUE;
+}
+
+static UBool
+isWellFormedLegacyType(const char* legacyType)
+{
+    const char* p = legacyType;
+    int32_t alphaNumLen = 0;
+    while (*p) {
+        if (*p == '_' || *p == '/' || *p == '-') {
+            if (alphaNumLen == 0) {
+                return FALSE;
+            }
+            alphaNumLen = 0;
+        } else if (UPRV_ISALPHANUM(*p)) {
+            alphaNumLen++;
+        } else {
+            return FALSE;
+        }
+        p++;
+    }
+    return (alphaNumLen != 0);
+}
+
+U_CAPI const char* U_EXPORT2
+uloc_toLegacyKey(const char* keyword)
+{
+    const char* legacyKey = ulocimp_toLegacyKey(keyword);
+    if (legacyKey == NULL) {
+        // Checks if the specified locale key is well-formed with the legacy locale syntax.
+        //
+        // Note:
+        //  LDML/CLDR provides some definition of keyword syntax in
+        //  * http://www.unicode.org/reports/tr35/#Unicode_locale_identifier and
+        //  * http://www.unicode.org/reports/tr35/#Old_Locale_Extension_Syntax
+        //  Keys can only consist of [0-9a-zA-Z].
+        if (isWellFormedLegacyKey(keyword)) {
+            return keyword;
+        }
+    }
+    return legacyKey;
+}
+
+U_CAPI const char* U_EXPORT2
+uloc_toLegacyType(const char* keyword, const char* value)
+{
+    const char* legacyType = ulocimp_toLegacyType(keyword, value, NULL, NULL);
+    if (legacyType == NULL) {
+        // Checks if the specified locale type is well-formed with the legacy locale syntax.
+        //
+        // Note:
+        //  LDML/CLDR provides some definition of keyword syntax in
+        //  * http://www.unicode.org/reports/tr35/#Unicode_locale_identifier and
+        //  * http://www.unicode.org/reports/tr35/#Old_Locale_Extension_Syntax
+        //  Values (types) can only consist of [0-9a-zA-Z], plus for legacy values
+        //  we allow [/_-+] in the middle (e.g. "Etc/GMT+1", "Asia/Tel_Aviv")
+        if (isWellFormedLegacyType(value)) {
+            return value;
+        }
+    }
+    return legacyType;
+}
+
+/*eof*/
diff --git a/thirdparty/icu4c/common/uloc_keytype.cpp b/thirdparty/icu4c/common/uloc_keytype.cpp
new file mode 100644
index 0000000000..019da058cf
--- /dev/null
+++ b/thirdparty/icu4c/common/uloc_keytype.cpp
@@ -0,0 +1,534 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2014-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+#include <algorithm>
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "uhash.h"
+#include "umutex.h"
+#include "uresimp.h"
+#include "uvector.h"
+#include "udataswp.h" /* for InvChar functions */
+
+static UHashtable* gLocExtKeyMap = NULL;
+static icu::UInitOnce gLocExtKeyMapInitOnce = U_INITONCE_INITIALIZER;
+
+// bit flags for special types
+typedef enum {
+    SPECIALTYPE_NONE = 0,
+    SPECIALTYPE_CODEPOINTS = 1,
+    SPECIALTYPE_REORDER_CODE = 2,
+    SPECIALTYPE_RG_KEY_VALUE = 4
+} SpecialType;
+
+struct LocExtKeyData : public icu::UMemory {
+    const char*     legacyId;
+    const char*     bcpId;
+    icu::LocalUHashtablePointer typeMap;
+    uint32_t        specialTypes;
+};
+
+struct LocExtType : public icu::UMemory {
+    const char*     legacyId;
+    const char*     bcpId;
+};
+
+static icu::MemoryPool<icu::CharString>* gKeyTypeStringPool = NULL;
+static icu::MemoryPool<LocExtKeyData>* gLocExtKeyDataEntries = NULL;
+static icu::MemoryPool<LocExtType>* gLocExtTypeEntries = NULL;
+
+U_CDECL_BEGIN
+
+static UBool U_CALLCONV
+uloc_key_type_cleanup(void) {
+    if (gLocExtKeyMap != NULL) {
+        uhash_close(gLocExtKeyMap);
+        gLocExtKeyMap = NULL;
+    }
+
+    delete gLocExtKeyDataEntries;
+    gLocExtKeyDataEntries = NULL;
+
+    delete gLocExtTypeEntries;
+    gLocExtTypeEntries = NULL;
+
+    delete gKeyTypeStringPool;
+    gKeyTypeStringPool = NULL;
+
+    gLocExtKeyMapInitOnce.reset();
+    return TRUE;
+}
+
+U_CDECL_END
+
+
+static void U_CALLCONV
+initFromResourceBundle(UErrorCode& sts) {
+    U_NAMESPACE_USE
+    ucln_common_registerCleanup(UCLN_COMMON_LOCALE_KEY_TYPE, uloc_key_type_cleanup);
+
+    gLocExtKeyMap = uhash_open(uhash_hashIChars, uhash_compareIChars, NULL, &sts);
+
+    LocalUResourceBundlePointer keyTypeDataRes(ures_openDirect(NULL, "keyTypeData", &sts));
+    LocalUResourceBundlePointer keyMapRes(ures_getByKey(keyTypeDataRes.getAlias(), "keyMap", NULL, &sts));
+    LocalUResourceBundlePointer typeMapRes(ures_getByKey(keyTypeDataRes.getAlias(), "typeMap", NULL, &sts));
+
+    if (U_FAILURE(sts)) {
+        return;
+    }
+
+    UErrorCode tmpSts = U_ZERO_ERROR;
+    LocalUResourceBundlePointer typeAliasRes(ures_getByKey(keyTypeDataRes.getAlias(), "typeAlias", NULL, &tmpSts));
+    tmpSts = U_ZERO_ERROR;
+    LocalUResourceBundlePointer bcpTypeAliasRes(ures_getByKey(keyTypeDataRes.getAlias(), "bcpTypeAlias", NULL, &tmpSts));
+
+    // initialize pools storing dynamically allocated objects
+    gKeyTypeStringPool = new icu::MemoryPool<icu::CharString>;
+    if (gKeyTypeStringPool == NULL) {
+        sts = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    gLocExtKeyDataEntries = new icu::MemoryPool<LocExtKeyData>;
+    if (gLocExtKeyDataEntries == NULL) {
+        sts = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    gLocExtTypeEntries = new icu::MemoryPool<LocExtType>;
+    if (gLocExtTypeEntries == NULL) {
+        sts = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    // iterate through keyMap resource
+    LocalUResourceBundlePointer keyMapEntry;
+
+    while (ures_hasNext(keyMapRes.getAlias())) {
+        keyMapEntry.adoptInstead(ures_getNextResource(keyMapRes.getAlias(), keyMapEntry.orphan(), &sts));
+        if (U_FAILURE(sts)) {
+            break;
+        }
+        const char* legacyKeyId = ures_getKey(keyMapEntry.getAlias());
+        UnicodeString uBcpKeyId = ures_getUnicodeString(keyMapEntry.getAlias(), &sts);
+        if (U_FAILURE(sts)) {
+            break;
+        }
+
+        // empty value indicates that BCP key is same with the legacy key.
+        const char* bcpKeyId = legacyKeyId;
+        if (!uBcpKeyId.isEmpty()) {
+            icu::CharString* bcpKeyIdBuf = gKeyTypeStringPool->create();
+            if (bcpKeyIdBuf == NULL) {
+                sts = U_MEMORY_ALLOCATION_ERROR;
+                break;
+            }
+            bcpKeyIdBuf->appendInvariantChars(uBcpKeyId, sts);
+            if (U_FAILURE(sts)) {
+                break;
+            }
+            bcpKeyId = bcpKeyIdBuf->data();
+        }
+
+        UBool isTZ = uprv_strcmp(legacyKeyId, "timezone") == 0;
+
+        UHashtable* typeDataMap = uhash_open(uhash_hashIChars, uhash_compareIChars, NULL, &sts);
+        if (U_FAILURE(sts)) {
+            break;
+        }
+        uint32_t specialTypes = SPECIALTYPE_NONE;
+
+        LocalUResourceBundlePointer typeAliasResByKey;
+        LocalUResourceBundlePointer bcpTypeAliasResByKey;
+
+        if (typeAliasRes.isValid()) {
+            tmpSts = U_ZERO_ERROR;
+            typeAliasResByKey.adoptInstead(ures_getByKey(typeAliasRes.getAlias(), legacyKeyId, NULL, &tmpSts));
+            if (U_FAILURE(tmpSts)) {
+                typeAliasResByKey.orphan();
+            }
+        }
+        if (bcpTypeAliasRes.isValid()) {
+            tmpSts = U_ZERO_ERROR;
+            bcpTypeAliasResByKey.adoptInstead(ures_getByKey(bcpTypeAliasRes.getAlias(), bcpKeyId, NULL, &tmpSts));
+            if (U_FAILURE(tmpSts)) {
+                bcpTypeAliasResByKey.orphan();
+            }
+        }
+
+        // look up type map for the key, and walk through the mapping data
+        tmpSts = U_ZERO_ERROR;
+        LocalUResourceBundlePointer typeMapResByKey(ures_getByKey(typeMapRes.getAlias(), legacyKeyId, NULL, &tmpSts));
+        if (U_FAILURE(tmpSts)) {
+            // type map for each key must exist
+            UPRV_UNREACHABLE;
+        } else {
+            LocalUResourceBundlePointer typeMapEntry;
+
+            while (ures_hasNext(typeMapResByKey.getAlias())) {
+                typeMapEntry.adoptInstead(ures_getNextResource(typeMapResByKey.getAlias(), typeMapEntry.orphan(), &sts));
+                if (U_FAILURE(sts)) {
+                    break;
+                }
+                const char* legacyTypeId = ures_getKey(typeMapEntry.getAlias());
+
+                // special types
+                if (uprv_strcmp(legacyTypeId, "CODEPOINTS") == 0) {
+                    specialTypes |= SPECIALTYPE_CODEPOINTS;
+                    continue;
+                }
+                if (uprv_strcmp(legacyTypeId, "REORDER_CODE") == 0) {
+                    specialTypes |= SPECIALTYPE_REORDER_CODE;
+                    continue;
+                }
+                if (uprv_strcmp(legacyTypeId, "RG_KEY_VALUE") == 0) {
+                    specialTypes |= SPECIALTYPE_RG_KEY_VALUE;
+                    continue;
+                }
+
+                if (isTZ) {
+                    // a timezone key uses a colon instead of a slash in the resource.
+                    // e.g. America:Los_Angeles
+                    if (uprv_strchr(legacyTypeId, ':') != NULL) {
+                        icu::CharString* legacyTypeIdBuf =
+                                gKeyTypeStringPool->create(legacyTypeId, sts);
+                        if (legacyTypeIdBuf == NULL) {
+                            sts = U_MEMORY_ALLOCATION_ERROR;
+                            break;
+                        }
+                        if (U_FAILURE(sts)) {
+                            break;
+                        }
+                        std::replace(
+                                legacyTypeIdBuf->data(),
+                                legacyTypeIdBuf->data() + legacyTypeIdBuf->length(),
+                                ':', '/');
+                        legacyTypeId = legacyTypeIdBuf->data();
+                    }
+                }
+
+                UnicodeString uBcpTypeId = ures_getUnicodeString(typeMapEntry.getAlias(), &sts);
+                if (U_FAILURE(sts)) {
+                    break;
+                }
+
+                // empty value indicates that BCP type is same with the legacy type.
+                const char* bcpTypeId = legacyTypeId;
+                if (!uBcpTypeId.isEmpty()) {
+                    icu::CharString* bcpTypeIdBuf = gKeyTypeStringPool->create();
+                    if (bcpTypeIdBuf == NULL) {
+                        sts = U_MEMORY_ALLOCATION_ERROR;
+                        break;
+                    }
+                    bcpTypeIdBuf->appendInvariantChars(uBcpTypeId, sts);
+                    if (U_FAILURE(sts)) {
+                        break;
+                    }
+                    bcpTypeId = bcpTypeIdBuf->data();
+                }
+
+                // Note: legacy type value should never be
+                // equivalent to bcp type value of a different
+                // type under the same key. So we use a single
+                // map for lookup.
+                LocExtType* t = gLocExtTypeEntries->create();
+                if (t == NULL) {
+                    sts = U_MEMORY_ALLOCATION_ERROR;
+                    break;
+                }
+                t->bcpId = bcpTypeId;
+                t->legacyId = legacyTypeId;
+
+                uhash_put(typeDataMap, (void*)legacyTypeId, t, &sts);
+                if (bcpTypeId != legacyTypeId) {
+                    // different type value
+                    uhash_put(typeDataMap, (void*)bcpTypeId, t, &sts);
+                }
+                if (U_FAILURE(sts)) {
+                    break;
+                }
+
+                // also put aliases in the map
+                if (typeAliasResByKey.isValid()) {
+                    LocalUResourceBundlePointer typeAliasDataEntry;
+
+                    ures_resetIterator(typeAliasResByKey.getAlias());
+                    while (ures_hasNext(typeAliasResByKey.getAlias()) && U_SUCCESS(sts)) {
+                        int32_t toLen;
+                        typeAliasDataEntry.adoptInstead(ures_getNextResource(typeAliasResByKey.getAlias(), typeAliasDataEntry.orphan(), &sts));
+                        const UChar* to = ures_getString(typeAliasDataEntry.getAlias(), &toLen, &sts);
+                        if (U_FAILURE(sts)) {
+                            break;
+                        }
+                        // check if this is an alias of canoncal legacy type
+                        if (uprv_compareInvWithUChar(NULL, legacyTypeId, -1, to, toLen) == 0) {
+                            const char* from = ures_getKey(typeAliasDataEntry.getAlias());
+                            if (isTZ) {
+                                // replace colon with slash if necessary
+                                if (uprv_strchr(from, ':') != NULL) {
+                                    icu::CharString* fromBuf =
+                                            gKeyTypeStringPool->create(from, sts);
+                                    if (fromBuf == NULL) {
+                                        sts = U_MEMORY_ALLOCATION_ERROR;
+                                        break;
+                                    }
+                                    if (U_FAILURE(sts)) {
+                                        break;
+                                    }
+                                    std::replace(
+                                            fromBuf->data(),
+                                            fromBuf->data() + fromBuf->length(),
+                                            ':', '/');
+                                    from = fromBuf->data();
+                                }
+                            }
+                            uhash_put(typeDataMap, (void*)from, t, &sts);
+                        }
+                    }
+                    if (U_FAILURE(sts)) {
+                        break;
+                    }
+                }
+
+                if (bcpTypeAliasResByKey.isValid()) {
+                    LocalUResourceBundlePointer bcpTypeAliasDataEntry;
+
+                    ures_resetIterator(bcpTypeAliasResByKey.getAlias());
+                    while (ures_hasNext(bcpTypeAliasResByKey.getAlias()) && U_SUCCESS(sts)) {
+                        int32_t toLen;
+                        bcpTypeAliasDataEntry.adoptInstead(ures_getNextResource(bcpTypeAliasResByKey.getAlias(), bcpTypeAliasDataEntry.orphan(), &sts));
+                        const UChar* to = ures_getString(bcpTypeAliasDataEntry.getAlias(), &toLen, &sts);
+                        if (U_FAILURE(sts)) {
+                            break;
+                        }
+                        // check if this is an alias of bcp type
+                        if (uprv_compareInvWithUChar(NULL, bcpTypeId, -1, to, toLen) == 0) {
+                            const char* from = ures_getKey(bcpTypeAliasDataEntry.getAlias());
+                            uhash_put(typeDataMap, (void*)from, t, &sts);
+                        }
+                    }
+                    if (U_FAILURE(sts)) {
+                        break;
+                    }
+                }
+            }
+        }
+        if (U_FAILURE(sts)) {
+            break;
+        }
+
+        LocExtKeyData* keyData = gLocExtKeyDataEntries->create();
+        if (keyData == NULL) {
+            sts = U_MEMORY_ALLOCATION_ERROR;
+            break;
+        }
+        keyData->bcpId = bcpKeyId;
+        keyData->legacyId = legacyKeyId;
+        keyData->specialTypes = specialTypes;
+        keyData->typeMap.adoptInstead(typeDataMap);
+
+        uhash_put(gLocExtKeyMap, (void*)legacyKeyId, keyData, &sts);
+        if (legacyKeyId != bcpKeyId) {
+            // different key value
+            uhash_put(gLocExtKeyMap, (void*)bcpKeyId, keyData, &sts);
+        }
+        if (U_FAILURE(sts)) {
+            break;
+        }
+    }
+}
+
+static UBool
+init() {
+    UErrorCode sts = U_ZERO_ERROR;
+    umtx_initOnce(gLocExtKeyMapInitOnce, &initFromResourceBundle, sts);
+    if (U_FAILURE(sts)) {
+        return FALSE;
+    }
+    return TRUE;
+}
+
+static UBool
+isSpecialTypeCodepoints(const char* val) {
+    int32_t subtagLen = 0;
+    const char* p = val;
+    while (*p) {
+        if (*p == '-') {
+            if (subtagLen < 4 || subtagLen > 6) {
+                return FALSE;
+            }
+            subtagLen = 0;
+        } else if ((*p >= '0' && *p <= '9') ||
+                    (*p >= 'A' && *p <= 'F') || // A-F/a-f are contiguous
+                    (*p >= 'a' && *p <= 'f')) { // also in EBCDIC
+            subtagLen++;
+        } else {
+            return FALSE;
+        }
+        p++;
+    }
+    return (subtagLen >= 4 && subtagLen <= 6);
+}
+
+static UBool
+isSpecialTypeReorderCode(const char* val) {
+    int32_t subtagLen = 0;
+    const char* p = val;
+    while (*p) {
+        if (*p == '-') {
+            if (subtagLen < 3 || subtagLen > 8) {
+                return FALSE;
+            }
+            subtagLen = 0;
+        } else if (uprv_isASCIILetter(*p)) {
+            subtagLen++;
+        } else {
+            return FALSE;
+        }
+        p++;
+    }
+    return (subtagLen >=3 && subtagLen <=8);
+}
+
+static UBool
+isSpecialTypeRgKeyValue(const char* val) {
+    int32_t subtagLen = 0;
+    const char* p = val;
+    while (*p) {
+        if ( (subtagLen < 2 && uprv_isASCIILetter(*p)) ||
+                    (subtagLen >= 2 && (*p == 'Z' || *p == 'z')) ) {
+            subtagLen++;
+        } else {
+            return FALSE;
+        }
+        p++;
+    }
+    return (subtagLen == 6);
+}
+
+U_CFUNC const char*
+ulocimp_toBcpKey(const char* key) {
+    if (!init()) {
+        return NULL;
+    }
+
+    LocExtKeyData* keyData = (LocExtKeyData*)uhash_get(gLocExtKeyMap, key);
+    if (keyData != NULL) {
+        return keyData->bcpId;
+    }
+    return NULL;
+}
+
+U_CFUNC const char*
+ulocimp_toLegacyKey(const char* key) {
+    if (!init()) {
+        return NULL;
+    }
+
+    LocExtKeyData* keyData = (LocExtKeyData*)uhash_get(gLocExtKeyMap, key);
+    if (keyData != NULL) {
+        return keyData->legacyId;
+    }
+    return NULL;
+}
+
+U_CFUNC const char*
+ulocimp_toBcpType(const char* key, const char* type, UBool* isKnownKey, UBool* isSpecialType) {
+    if (isKnownKey != NULL) {
+        *isKnownKey = FALSE;
+    }
+    if (isSpecialType != NULL) {
+        *isSpecialType = FALSE;
+    }
+
+    if (!init()) {
+        return NULL;
+    }
+
+    LocExtKeyData* keyData = (LocExtKeyData*)uhash_get(gLocExtKeyMap, key);
+    if (keyData != NULL) {
+        if (isKnownKey != NULL) {
+            *isKnownKey = TRUE;
+        }
+        LocExtType* t = (LocExtType*)uhash_get(keyData->typeMap.getAlias(), type);
+        if (t != NULL) {
+            return t->bcpId;
+        }
+        if (keyData->specialTypes != SPECIALTYPE_NONE) {
+            UBool matched = FALSE;
+            if (keyData->specialTypes & SPECIALTYPE_CODEPOINTS) {
+                matched = isSpecialTypeCodepoints(type);
+            }
+            if (!matched && keyData->specialTypes & SPECIALTYPE_REORDER_CODE) {
+                matched = isSpecialTypeReorderCode(type);
+            }
+            if (!matched && keyData->specialTypes & SPECIALTYPE_RG_KEY_VALUE) {
+                matched = isSpecialTypeRgKeyValue(type);
+            }
+            if (matched) {
+                if (isSpecialType != NULL) {
+                    *isSpecialType = TRUE;
+                }
+                return type;
+            }
+        }
+    }
+    return NULL;
+}
+
+
+U_CFUNC const char*
+ulocimp_toLegacyType(const char* key, const char* type, UBool* isKnownKey, UBool* isSpecialType) {
+    if (isKnownKey != NULL) {
+        *isKnownKey = FALSE;
+    }
+    if (isSpecialType != NULL) {
+        *isSpecialType = FALSE;
+    }
+
+    if (!init()) {
+        return NULL;
+    }
+
+    LocExtKeyData* keyData = (LocExtKeyData*)uhash_get(gLocExtKeyMap, key);
+    if (keyData != NULL) {
+        if (isKnownKey != NULL) {
+            *isKnownKey = TRUE;
+        }
+        LocExtType* t = (LocExtType*)uhash_get(keyData->typeMap.getAlias(), type);
+        if (t != NULL) {
+            return t->legacyId;
+        }
+        if (keyData->specialTypes != SPECIALTYPE_NONE) {
+            UBool matched = FALSE;
+            if (keyData->specialTypes & SPECIALTYPE_CODEPOINTS) {
+                matched = isSpecialTypeCodepoints(type);
+            }
+            if (!matched && keyData->specialTypes & SPECIALTYPE_REORDER_CODE) {
+                matched = isSpecialTypeReorderCode(type);
+            }
+            if (!matched && keyData->specialTypes & SPECIALTYPE_RG_KEY_VALUE) {
+                matched = isSpecialTypeRgKeyValue(type);
+            }
+            if (matched) {
+                if (isSpecialType != NULL) {
+                    *isSpecialType = TRUE;
+                }
+                return type;
+            }
+        }
+    }
+    return NULL;
+}
+
diff --git a/thirdparty/icu4c/common/uloc_tag.cpp b/thirdparty/icu4c/common/uloc_tag.cpp
new file mode 100644
index 0000000000..7f7fd9119e
--- /dev/null
+++ b/thirdparty/icu4c/common/uloc_tag.cpp
@@ -0,0 +1,2844 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2009-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "unicode/bytestream.h"
+#include "unicode/utypes.h"
+#include "unicode/ures.h"
+#include "unicode/localpointer.h"
+#include "unicode/putil.h"
+#include "unicode/uenum.h"
+#include "unicode/uloc.h"
+#include "ustr_imp.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "putilimp.h"
+#include "uinvchar.h"
+#include "ulocimp.h"
+#include "uassert.h"
+
+
+/* struct holding a single variant */
+typedef struct VariantListEntry {
+    const char              *variant;
+    struct VariantListEntry *next;
+} VariantListEntry;
+
+/* struct holding a single attribute value */
+struct AttributeListEntry : public icu::UMemory {
+    const char              *attribute;
+    struct AttributeListEntry *next;
+};
+
+/* struct holding a single extension */
+struct ExtensionListEntry : public icu::UMemory {
+    const char                  *key;
+    const char                  *value;
+    struct ExtensionListEntry   *next;
+};
+
+#define MAXEXTLANG 3
+typedef struct ULanguageTag {
+    char                *buf;   /* holding parsed subtags */
+    const char          *language;
+    const char          *extlang[MAXEXTLANG];
+    const char          *script;
+    const char          *region;
+    VariantListEntry    *variants;
+    ExtensionListEntry  *extensions;
+    const char          *privateuse;
+    const char          *legacy;
+} ULanguageTag;
+
+#define MINLEN 2
+#define SEP '-'
+#define PRIVATEUSE 'x'
+#define LDMLEXT 'u'
+
+#define LOCALE_SEP '_'
+#define LOCALE_EXT_SEP '@'
+#define LOCALE_KEYWORD_SEP ';'
+#define LOCALE_KEY_TYPE_SEP '='
+
+#define ISALPHA(c) uprv_isASCIILetter(c)
+#define ISNUMERIC(c) ((c)>='0' && (c)<='9')
+
+static const char EMPTY[] = "";
+static const char LANG_UND[] = "und";
+static const char PRIVATEUSE_KEY[] = "x";
+static const char _POSIX[] = "_POSIX";
+static const char POSIX_KEY[] = "va";
+static const char POSIX_VALUE[] = "posix";
+static const char LOCALE_ATTRIBUTE_KEY[] = "attribute";
+static const char PRIVUSE_VARIANT_PREFIX[] = "lvariant";
+static const char LOCALE_TYPE_YES[] = "yes";
+
+#define LANG_UND_LEN 3
+
+/*
+ Updated on 2018-09-12 from
+ https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry .
+
+ This table has 2 parts. The part for
+ legacy language tags (marked as “Type: grandfathered” in BCP 47)
+ is generated by the following scripts from the IANA language tag registry.
+
+ curl  https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry |\
+ egrep -A 7 'Type: grandfathered' | \
+ egrep 'Tag|Prefe' | grep -B1 'Preferred' | grep -v '^--' | \
+ awk -n '/Tag/ {printf("    \"%s\", ", $2);} /Preferred/ {printf("\"%s\",\n", $2);}' |\
+ tr 'A-Z' 'a-z'
+
+
+ The 2nd part is made of five ICU-specific entries. They're kept for
+ the backward compatibility for now, even though there are no preferred
+ values. They may have to be removed for the strict BCP 47 compliance.
+
+*/
+static const char* const LEGACY[] = {
+/*  legacy          preferred */
+    "art-lojban",   "jbo",
+    "en-gb-oed",    "en-gb-oxendict",
+    "i-ami",        "ami",
+    "i-bnn",        "bnn",
+    "i-hak",        "hak",
+    "i-klingon",    "tlh",
+    "i-lux",        "lb",
+    "i-navajo",     "nv",
+    "i-pwn",        "pwn",
+    "i-tao",        "tao",
+    "i-tay",        "tay",
+    "i-tsu",        "tsu",
+    "no-bok",       "nb",
+    "no-nyn",       "nn",
+    "sgn-be-fr",    "sfb",
+    "sgn-be-nl",    "vgt",
+    "sgn-ch-de",    "sgg",
+    "zh-guoyu",     "cmn",
+    "zh-hakka",     "hak",
+    "zh-min-nan",   "nan",
+    "zh-xiang",     "hsn",
+
+    // Legacy tags with no preferred value in the IANA
+    // registry. Kept for now for the backward compatibility
+    // because ICU has mapped them this way.
+    "cel-gaulish",  "xtg-x-cel-gaulish",
+    "i-default",    "en-x-i-default",
+    "i-enochian",   "und-x-i-enochian",
+    "i-mingo",      "see-x-i-mingo",
+    "zh-min",       "nan-x-zh-min",
+};
+
+/*
+ Updated on 2018-09-12 from
+ https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry .
+
+ The table lists redundant tags with preferred value in the IANA languate tag registry.
+ It's generated with the following command:
+
+ curl  https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry |\
+ grep 'Type: redundant' -A 5 | egrep '^(Tag:|Prefer)' | grep -B1 'Preferred' | \
+ awk -n '/Tag/ {printf("    \"%s\",       ", $2);} /Preferred/ {printf("\"%s\",\n", $2);}' | \
+ tr 'A-Z' 'a-z'
+
+ In addition, ja-latn-hepburn-heploc is mapped to ja-latn-alalc97 because
+ a variant tag 'hepburn-heploc' has the preferred subtag, 'alaic97'.
+*/
+
+static const char* const REDUNDANT[] = {
+//  redundant       preferred
+    "sgn-br",       "bzs",
+    "sgn-co",       "csn",
+    "sgn-de",       "gsg",
+    "sgn-dk",       "dsl",
+    "sgn-es",       "ssp",
+    "sgn-fr",       "fsl",
+    "sgn-gb",       "bfi",
+    "sgn-gr",       "gss",
+    "sgn-ie",       "isg",
+    "sgn-it",       "ise",
+    "sgn-jp",       "jsl",
+    "sgn-mx",       "mfs",
+    "sgn-ni",       "ncs",
+    "sgn-nl",       "dse",
+    "sgn-no",       "nsl",
+    "sgn-pt",       "psr",
+    "sgn-se",       "swl",
+    "sgn-us",       "ase",
+    "sgn-za",       "sfs",
+    "zh-cmn",       "cmn",
+    "zh-cmn-hans",  "cmn-hans",
+    "zh-cmn-hant",  "cmn-hant",
+    "zh-gan",       "gan",
+    "zh-wuu",       "wuu",
+    "zh-yue",       "yue",
+
+    // variant tag with preferred value
+    "ja-latn-hepburn-heploc", "ja-latn-alalc97",
+};
+
+/*
+  Updated on 2018-09-12 from
+  https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry .
+
+  grep 'Type: language' -A 7 language-subtag-registry  | egrep 'Subtag|Prefe' | \
+  grep -B1 'Preferred' | grep -v '^--' | \
+  awk -n '/Subtag/ {printf("    \"%s\",       ", $2);} /Preferred/ {printf("\"%s\",\n", $2);}'
+
+  Make sure that 2-letter language subtags come before 3-letter subtags.
+*/
+static const char DEPRECATEDLANGS[][4] = {
+/*  deprecated  new */
+    "in",       "id",
+    "iw",       "he",
+    "ji",       "yi",
+    "jw",       "jv",
+    "mo",       "ro",
+    "aam",       "aas",
+    "adp",       "dz",
+    "aue",       "ktz",
+    "ayx",       "nun",
+    "bgm",       "bcg",
+    "bjd",       "drl",
+    "ccq",       "rki",
+    "cjr",       "mom",
+    "cka",       "cmr",
+    "cmk",       "xch",
+    "coy",       "pij",
+    "cqu",       "quh",
+    "drh",       "khk",
+    "drw",       "prs",
+    "gav",       "dev",
+    "gfx",       "vaj",
+    "ggn",       "gvr",
+    "gti",       "nyc",
+    "guv",       "duz",
+    "hrr",       "jal",
+    "ibi",       "opa",
+    "ilw",       "gal",
+    "jeg",       "oyb",
+    "kgc",       "tdf",
+    "kgh",       "kml",
+    "koj",       "kwv",
+    "krm",       "bmf",
+    "ktr",       "dtp",
+    "kvs",       "gdj",
+    "kwq",       "yam",
+    "kxe",       "tvd",
+    "kzj",       "dtp",
+    "kzt",       "dtp",
+    "lii",       "raq",
+    "lmm",       "rmx",
+    "meg",       "cir",
+    "mst",       "mry",
+    "mwj",       "vaj",
+    "myt",       "mry",
+    "nad",       "xny",
+    "ncp",       "kdz",
+    "nnx",       "ngv",
+    "nts",       "pij",
+    "oun",       "vaj",
+    "pcr",       "adx",
+    "pmc",       "huw",
+    "pmu",       "phr",
+    "ppa",       "bfy",
+    "ppr",       "lcq",
+    "pry",       "prt",
+    "puz",       "pub",
+    "sca",       "hle",
+    "skk",       "oyb",
+    "tdu",       "dtp",
+    "thc",       "tpo",
+    "thx",       "oyb",
+    "tie",       "ras",
+    "tkk",       "twm",
+    "tlw",       "weo",
+    "tmp",       "tyj",
+    "tne",       "kak",
+    "tnf",       "prs",
+    "tsf",       "taj",
+    "uok",       "ema",
+    "xba",       "cax",
+    "xia",       "acn",
+    "xkh",       "waw",
+    "xsj",       "suj",
+    "ybd",       "rki",
+    "yma",       "lrr",
+    "ymt",       "mtm",
+    "yos",       "zom",
+    "yuu",       "yug",
+};
+
+/*
+  Updated on 2018-04-24 from
+
+  curl  https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry | \
+  grep 'Type: region' -A 7 | egrep 'Subtag|Prefe' | \
+  grep -B1 'Preferred' | \
+  awk -n '/Subtag/ {printf("    \"%s\",       ", $2);} /Preferred/ {printf("\"%s\",\n", $2);}'
+*/
+static const char DEPRECATEDREGIONS[][3] = {
+/*  deprecated  new */
+    "BU",       "MM",
+    "DD",       "DE",
+    "FX",       "FR",
+    "TP",       "TL",
+    "YD",       "YE",
+    "ZR",       "CD",
+};
+
+/*
+* -------------------------------------------------
+*
+* These ultag_ functions may be exposed as APIs later
+*
+* -------------------------------------------------
+*/
+
+static ULanguageTag*
+ultag_parse(const char* tag, int32_t tagLen, int32_t* parsedLen, UErrorCode* status);
+
+static void
+ultag_close(ULanguageTag* langtag);
+
+static const char*
+ultag_getLanguage(const ULanguageTag* langtag);
+
+#if 0
+static const char*
+ultag_getJDKLanguage(const ULanguageTag* langtag);
+#endif
+
+static const char*
+ultag_getExtlang(const ULanguageTag* langtag, int32_t idx);
+
+static int32_t
+ultag_getExtlangSize(const ULanguageTag* langtag);
+
+static const char*
+ultag_getScript(const ULanguageTag* langtag);
+
+static const char*
+ultag_getRegion(const ULanguageTag* langtag);
+
+static const char*
+ultag_getVariant(const ULanguageTag* langtag, int32_t idx);
+
+static int32_t
+ultag_getVariantsSize(const ULanguageTag* langtag);
+
+static const char*
+ultag_getExtensionKey(const ULanguageTag* langtag, int32_t idx);
+
+static const char*
+ultag_getExtensionValue(const ULanguageTag* langtag, int32_t idx);
+
+static int32_t
+ultag_getExtensionsSize(const ULanguageTag* langtag);
+
+static const char*
+ultag_getPrivateUse(const ULanguageTag* langtag);
+
+#if 0
+static const char*
+ultag_getLegacy(const ULanguageTag* langtag);
+#endif
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalULanguageTagPointer
+ * "Smart pointer" class, closes a ULanguageTag via ultag_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @internal
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalULanguageTagPointer, ULanguageTag, ultag_close);
+
+U_NAMESPACE_END
+
+/*
+* -------------------------------------------------
+*
+* Language subtag syntax validation functions
+*
+* -------------------------------------------------
+*/
+
+static UBool
+_isAlphaString(const char* s, int32_t len) {
+    int32_t i;
+    for (i = 0; i < len; i++) {
+        if (!ISALPHA(*(s + i))) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+static UBool
+_isNumericString(const char* s, int32_t len) {
+    int32_t i;
+    for (i = 0; i < len; i++) {
+        if (!ISNUMERIC(*(s + i))) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+static UBool
+_isAlphaNumericString(const char* s, int32_t len) {
+    int32_t i;
+    for (i = 0; i < len; i++) {
+        if (!ISALPHA(*(s + i)) && !ISNUMERIC(*(s + i))) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+static UBool
+_isAlphaNumericStringLimitedLength(const char* s, int32_t len, int32_t min, int32_t max) {
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len >= min && len <= max && _isAlphaNumericString(s, len)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC UBool
+ultag_isLanguageSubtag(const char* s, int32_t len) {
+    /*
+     * unicode_language_subtag = alpha{2,3} | alpha{5,8};
+     * NOTE: Per ICUTC 2019/01/23- accepting alpha 4
+     * See ICU-20372
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len >= 2 && len <= 8 && _isAlphaString(s, len)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isExtlangSubtag(const char* s, int32_t len) {
+    /*
+     * extlang       = 3ALPHA              ; selected ISO 639 codes
+     *                 *2("-" 3ALPHA)      ; permanently reserved
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 3 && _isAlphaString(s, len)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC UBool
+ultag_isScriptSubtag(const char* s, int32_t len) {
+    /*
+     * script        = 4ALPHA              ; ISO 15924 code
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 4 && _isAlphaString(s, len)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC UBool
+ultag_isRegionSubtag(const char* s, int32_t len) {
+    /*
+     * region        = 2ALPHA              ; ISO 3166-1 code
+     *               / 3DIGIT              ; UN M.49 code
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 2 && _isAlphaString(s, len)) {
+        return TRUE;
+    }
+    if (len == 3 && _isNumericString(s, len)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isVariantSubtag(const char* s, int32_t len) {
+    /*
+     * variant       = 5*8alphanum         ; registered variants
+     *               / (DIGIT 3alphanum)
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (_isAlphaNumericStringLimitedLength(s, len, 5, 8)) {
+        return TRUE;
+    }
+    if (len == 4 && ISNUMERIC(*s) && _isAlphaNumericString(s + 1, 3)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isSepListOf(UBool (*test)(const char*, int32_t), const char* s, int32_t len) {
+    const char *p = s;
+    const char *pSubtag = NULL;
+
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+
+    while ((p - s) < len) {
+        if (*p == SEP) {
+            if (pSubtag == NULL) {
+                return FALSE;
+            }
+            if (!test(pSubtag, (int32_t)(p - pSubtag))) {
+                return FALSE;
+            }
+            pSubtag = NULL;
+        } else if (pSubtag == NULL) {
+            pSubtag = p;
+        }
+        p++;
+    }
+    if (pSubtag == NULL) {
+        return FALSE;
+    }
+    return test(pSubtag, (int32_t)(p - pSubtag));
+}
+
+U_CFUNC UBool
+ultag_isVariantSubtags(const char* s, int32_t len) {
+    return _isSepListOf(&_isVariantSubtag, s, len);
+}
+
+// This is for the ICU-specific "lvariant" handling.
+static UBool
+_isPrivateuseVariantSubtag(const char* s, int32_t len) {
+    /*
+     * variant       = 1*8alphanum         ; registered variants
+     *               / (DIGIT 3alphanum)
+     */
+    return _isAlphaNumericStringLimitedLength(s, len , 1, 8);
+}
+
+static UBool
+_isExtensionSingleton(const char* s, int32_t len) {
+    /*
+     * extension     = singleton 1*("-" (2*8alphanum))
+     *
+     * singleton     = DIGIT               ; 0 - 9
+     *               / %x41-57             ; A - W
+     *               / %x59-5A             ; Y - Z
+     *               / %x61-77             ; a - w
+     *               / %x79-7A             ; y - z
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 1 && (ISALPHA(*s) || ISNUMERIC(*s)) && (uprv_tolower(*s) != PRIVATEUSE)) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isExtensionSubtag(const char* s, int32_t len) {
+    /*
+     * extension     = singleton 1*("-" (2*8alphanum))
+     */
+    return _isAlphaNumericStringLimitedLength(s, len, 2, 8);
+}
+
+U_CFUNC UBool
+ultag_isExtensionSubtags(const char* s, int32_t len) {
+    return _isSepListOf(&_isExtensionSubtag, s, len);
+}
+
+static UBool
+_isPrivateuseValueSubtag(const char* s, int32_t len) {
+    /*
+     * privateuse    = "x" 1*("-" (1*8alphanum))
+     */
+    return _isAlphaNumericStringLimitedLength(s, len, 1, 8);
+}
+
+U_CFUNC UBool
+ultag_isPrivateuseValueSubtags(const char* s, int32_t len) {
+    return _isSepListOf(&_isPrivateuseValueSubtag, s, len);
+}
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleAttribute(const char* s, int32_t len) {
+    /*
+     * attribute = alphanum{3,8} ;
+     */
+    return _isAlphaNumericStringLimitedLength(s, len , 3, 8);
+}
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleAttributes(const char* s, int32_t len) {
+    return _isSepListOf(&ultag_isUnicodeLocaleAttribute, s, len);
+}
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleKey(const char* s, int32_t len) {
+    /*
+     * key = alphanum alpha ;
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 2 && (ISALPHA(*s) || ISNUMERIC(*s)) && ISALPHA(s[1])) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC UBool
+_isUnicodeLocaleTypeSubtag(const char*s, int32_t len) {
+    /*
+     * alphanum{3,8}
+     */
+    return _isAlphaNumericStringLimitedLength(s, len , 3, 8);
+}
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleType(const char*s, int32_t len) {
+    /*
+     * type = alphanum{3,8} (sep alphanum{3,8})* ;
+     */
+    return _isSepListOf(&_isUnicodeLocaleTypeSubtag, s, len);
+}
+
+static UBool
+_isTKey(const char* s, int32_t len)
+{
+    /*
+     * tkey = alpha digit ;
+     */
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+    if (len == 2 && ISALPHA(*s) && ISNUMERIC(*(s + 1))) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isTValue(const char* s, int32_t len)
+{
+    /*
+     * tvalue = (sep alphanum{3,8})+ ;
+     */
+    return _isAlphaNumericStringLimitedLength(s, len , 3, 8);
+}
+
+static UBool
+_isTransformedExtensionSubtag(int32_t& state, const char* s, int32_t len)
+{
+    const int32_t kStart = 0;       // Start, wait for unicode_language_subtag, tkey or end
+    const int32_t kGotLanguage = 1; // Got unicode_language_subtag, wait for unicode_script_subtag,
+                                    // unicode_region_subtag, unicode_variant_subtag, tkey or end
+    const int32_t kGotScript = 2;   // Got unicode_script_subtag, wait for unicode_region_subtag,
+                                    // unicode_variant_subtag, tkey, or end
+    const int32_t kGotRegion = 3;   // Got unicode_region_subtag, wait for unicode_variant_subtag,
+                                    // tkey, or end.
+    const int32_t kGotVariant = 4;  // Got unicode_variant_subtag, wait for unicode_variant_subtag
+                                    // tkey or end.
+    const int32_t kGotTKey = -1;    // Got tkey, wait for tvalue. ERROR if stop here.
+    const int32_t kGotTValue = 6;   // Got tvalue, wait for tkey, tvalue or end
+
+    switch (state) {
+        case kStart:
+            if (ultag_isLanguageSubtag(s, len)) {
+                state = kGotLanguage;
+                return TRUE;
+            }
+            if (_isTKey(s, len)) {
+                state = kGotTKey;
+                return TRUE;
+            }
+            return FALSE;
+        case kGotLanguage:
+            if (ultag_isScriptSubtag(s, len)) {
+                state = kGotScript;
+                return TRUE;
+            }
+            U_FALLTHROUGH;
+        case kGotScript:
+            if (ultag_isRegionSubtag(s, len)) {
+                state = kGotRegion;
+                return TRUE;
+            }
+            U_FALLTHROUGH;
+        case kGotRegion:
+            U_FALLTHROUGH;
+        case kGotVariant:
+            if (_isVariantSubtag(s, len)) {
+                state = kGotVariant;
+                return TRUE;
+            }
+            if (_isTKey(s, len)) {
+                state = kGotTKey;
+                return TRUE;
+            }
+            return FALSE;
+        case kGotTKey:
+            if (_isTValue(s, len)) {
+                state = kGotTValue;
+                return TRUE;
+            }
+            return FALSE;
+        case kGotTValue:
+            if (_isTKey(s, len)) {
+                state = kGotTKey;
+                return TRUE;
+            }
+            if (_isTValue(s, len)) {
+                return TRUE;
+            }
+            return FALSE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isUnicodeExtensionSubtag(int32_t& state, const char* s, int32_t len)
+{
+    const int32_t kStart = 0;         // Start, wait for a key or attribute or end
+    const int32_t kGotKey = 1;        // Got a key, wait for type or key or end
+    const int32_t kGotType = 2;       // Got a type, wait for key or end
+
+    switch (state) {
+        case kStart:
+            if (ultag_isUnicodeLocaleKey(s, len)) {
+                state = kGotKey;
+                return TRUE;
+            }
+            if (ultag_isUnicodeLocaleAttribute(s, len)) {
+                return TRUE;
+            }
+            return FALSE;
+        case kGotKey:
+            if (ultag_isUnicodeLocaleKey(s, len)) {
+                return TRUE;
+            }
+            if (_isUnicodeLocaleTypeSubtag(s, len)) {
+                state = kGotType;
+                return TRUE;
+            }
+            return FALSE;
+        case kGotType:
+            if (ultag_isUnicodeLocaleKey(s, len)) {
+                state = kGotKey;
+                return TRUE;
+            }
+            if (_isUnicodeLocaleTypeSubtag(s, len)) {
+                return TRUE;
+            }
+            return FALSE;
+    }
+    return FALSE;
+}
+
+static UBool
+_isStatefulSepListOf(UBool (*test)(int32_t&, const char*, int32_t), const char* s, int32_t len)
+{
+    int32_t state = 0;
+    const char* p;
+    const char* start = s;
+    int32_t subtagLen = 0;
+
+    if (len < 0) {
+        len = (int32_t)uprv_strlen(s);
+    }
+
+    for (p = s; len > 0; p++, len--) {
+        if (*p == SEP) {
+            if (!test(state, start, subtagLen)) {
+                return FALSE;
+            }
+            subtagLen = 0;
+            start = p + 1;
+        } else {
+            subtagLen++;
+        }
+    }
+
+    if (test(state, start, subtagLen) && state >= 0) {
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC UBool
+ultag_isTransformedExtensionSubtags(const char* s, int32_t len)
+{
+    return _isStatefulSepListOf(&_isTransformedExtensionSubtag, s, len);
+}
+
+U_CFUNC UBool
+ultag_isUnicodeExtensionSubtags(const char* s, int32_t len) {
+    return _isStatefulSepListOf(&_isUnicodeExtensionSubtag, s, len);
+}
+
+
+/*
+* -------------------------------------------------
+*
+* Helper functions
+*
+* -------------------------------------------------
+*/
+
+static UBool
+_addVariantToList(VariantListEntry **first, VariantListEntry *var) {
+    UBool bAdded = TRUE;
+
+    if (*first == NULL) {
+        var->next = NULL;
+        *first = var;
+    } else {
+        VariantListEntry *prev, *cur;
+        int32_t cmp;
+
+        /* variants order should be preserved */
+        prev = NULL;
+        cur = *first;
+        while (TRUE) {
+            if (cur == NULL) {
+                prev->next = var;
+                var->next = NULL;
+                break;
+            }
+
+            /* Checking for duplicate variant */
+            cmp = uprv_compareInvCharsAsAscii(var->variant, cur->variant);
+            if (cmp == 0) {
+                /* duplicated variant */
+                bAdded = FALSE;
+                break;
+            }
+            prev = cur;
+            cur = cur->next;
+        }
+    }
+
+    return bAdded;
+}
+
+static UBool
+_addAttributeToList(AttributeListEntry **first, AttributeListEntry *attr) {
+    UBool bAdded = TRUE;
+
+    if (*first == NULL) {
+        attr->next = NULL;
+        *first = attr;
+    } else {
+        AttributeListEntry *prev, *cur;
+        int32_t cmp;
+
+        /* reorder variants in alphabetical order */
+        prev = NULL;
+        cur = *first;
+        while (TRUE) {
+            if (cur == NULL) {
+                prev->next = attr;
+                attr->next = NULL;
+                break;
+            }
+            cmp = uprv_compareInvCharsAsAscii(attr->attribute, cur->attribute);
+            if (cmp < 0) {
+                if (prev == NULL) {
+                    *first = attr;
+                } else {
+                    prev->next = attr;
+                }
+                attr->next = cur;
+                break;
+            }
+            if (cmp == 0) {
+                /* duplicated variant */
+                bAdded = FALSE;
+                break;
+            }
+            prev = cur;
+            cur = cur->next;
+        }
+    }
+
+    return bAdded;
+}
+
+
+static UBool
+_addExtensionToList(ExtensionListEntry **first, ExtensionListEntry *ext, UBool localeToBCP) {
+    UBool bAdded = TRUE;
+
+    if (*first == NULL) {
+        ext->next = NULL;
+        *first = ext;
+    } else {
+        ExtensionListEntry *prev, *cur;
+        int32_t cmp;
+
+        /* reorder variants in alphabetical order */
+        prev = NULL;
+        cur = *first;
+        while (TRUE) {
+            if (cur == NULL) {
+                prev->next = ext;
+                ext->next = NULL;
+                break;
+            }
+            if (localeToBCP) {
+                /* special handling for locale to bcp conversion */
+                int32_t len, curlen;
+
+                len = (int32_t)uprv_strlen(ext->key);
+                curlen = (int32_t)uprv_strlen(cur->key);
+
+                if (len == 1 && curlen == 1) {
+                    if (*(ext->key) == *(cur->key)) {
+                        cmp = 0;
+                    } else if (*(ext->key) == PRIVATEUSE) {
+                        cmp = 1;
+                    } else if (*(cur->key) == PRIVATEUSE) {
+                        cmp = -1;
+                    } else {
+                        cmp = *(ext->key) - *(cur->key);
+                    }
+                } else if (len == 1) {
+                    cmp = *(ext->key) - LDMLEXT; 
+                } else if (curlen == 1) {
+                    cmp = LDMLEXT - *(cur->key);
+                } else {
+                    cmp = uprv_compareInvCharsAsAscii(ext->key, cur->key);
+                    /* Both are u extension keys - we need special handling for 'attribute' */
+                    if (cmp != 0) {
+                        if (uprv_strcmp(cur->key, LOCALE_ATTRIBUTE_KEY) == 0) {
+                            cmp = 1;
+                        } else if (uprv_strcmp(ext->key, LOCALE_ATTRIBUTE_KEY) == 0) {
+                            cmp = -1;
+                        }
+                    }
+                }
+            } else {
+                cmp = uprv_compareInvCharsAsAscii(ext->key, cur->key);
+            }
+            if (cmp < 0) {
+                if (prev == NULL) {
+                    *first = ext;
+                } else {
+                    prev->next = ext;
+                }
+                ext->next = cur;
+                break;
+            }
+            if (cmp == 0) {
+                /* duplicated extension key */
+                bAdded = FALSE;
+                break;
+            }
+            prev = cur;
+            cur = cur->next;
+        }
+    }
+
+    return bAdded;
+}
+
+static void
+_initializeULanguageTag(ULanguageTag* langtag) {
+    int32_t i;
+
+    langtag->buf = NULL;
+
+    langtag->language = EMPTY;
+    for (i = 0; i < MAXEXTLANG; i++) {
+        langtag->extlang[i] = NULL;
+    }
+
+    langtag->script = EMPTY;
+    langtag->region = EMPTY;
+
+    langtag->variants = NULL;
+    langtag->extensions = NULL;
+
+    langtag->legacy = EMPTY;
+    langtag->privateuse = EMPTY;
+}
+
+static void
+_appendLanguageToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UErrorCode* status) {
+    char buf[ULOC_LANG_CAPACITY];
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    int32_t len, i;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    len = uloc_getLanguage(localeID, buf, sizeof(buf), &tmpStatus);
+    if (U_FAILURE(tmpStatus) || tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        len = 0;
+    }
+
+    /* Note: returned language code is in lower case letters */
+
+    if (len == 0) {
+        sink.Append(LANG_UND, LANG_UND_LEN);
+    } else if (!ultag_isLanguageSubtag(buf, len)) {
+            /* invalid language code */
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+        sink.Append(LANG_UND, LANG_UND_LEN);
+    } else {
+        /* resolve deprecated */
+        for (i = 0; i < UPRV_LENGTHOF(DEPRECATEDLANGS); i += 2) {
+            // 2-letter deprecated subtags are listede before 3-letter
+            // ones in DEPRECATEDLANGS[]. Get out of loop on coming
+            // across the 1st 3-letter subtag, if the input is a 2-letter code.
+            // to avoid continuing to try when there's no match.
+            if (uprv_strlen(buf) < uprv_strlen(DEPRECATEDLANGS[i])) break;
+            if (uprv_compareInvCharsAsAscii(buf, DEPRECATEDLANGS[i]) == 0) {
+                uprv_strcpy(buf, DEPRECATEDLANGS[i + 1]);
+                len = (int32_t)uprv_strlen(buf);
+                break;
+            }
+        }
+        sink.Append(buf, len);
+    }
+}
+
+static void
+_appendScriptToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UErrorCode* status) {
+    char buf[ULOC_SCRIPT_CAPACITY];
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    int32_t len;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    len = uloc_getScript(localeID, buf, sizeof(buf), &tmpStatus);
+    if (U_FAILURE(tmpStatus) || tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return;
+    }
+
+    if (len > 0) {
+        if (!ultag_isScriptSubtag(buf, len)) {
+            /* invalid script code */
+            if (strict) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+            }
+            return;
+        } else {
+            sink.Append("-", 1);
+            sink.Append(buf, len);
+        }
+    }
+}
+
+static void
+_appendRegionToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UErrorCode* status) {
+    char buf[ULOC_COUNTRY_CAPACITY];
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    int32_t len;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    len = uloc_getCountry(localeID, buf, sizeof(buf), &tmpStatus);
+    if (U_FAILURE(tmpStatus) || tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return;
+    }
+
+    if (len > 0) {
+        if (!ultag_isRegionSubtag(buf, len)) {
+            /* invalid region code */
+            if (strict) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+            }
+            return;
+        } else {
+            sink.Append("-", 1);
+            /* resolve deprecated */
+            for (int i = 0; i < UPRV_LENGTHOF(DEPRECATEDREGIONS); i += 2) {
+                if (uprv_compareInvCharsAsAscii(buf, DEPRECATEDREGIONS[i]) == 0) {
+                    uprv_strcpy(buf, DEPRECATEDREGIONS[i + 1]);
+                    len = (int32_t)uprv_strlen(buf);
+                    break;
+                }
+            }
+            sink.Append(buf, len);
+        }
+    }
+}
+
+static void _sortVariants(VariantListEntry* first) {
+    for (VariantListEntry* var1 = first; var1 != NULL; var1 = var1->next) {
+        for (VariantListEntry* var2 = var1->next; var2 != NULL; var2 = var2->next) {
+            // Swap var1->variant and var2->variant.
+            if (uprv_compareInvCharsAsAscii(var1->variant, var2->variant) > 0) {
+                const char* temp = var1->variant;
+                var1->variant = var2->variant;
+                var2->variant = temp;
+            }
+        }
+    }
+}
+
+static void
+_appendVariantsToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UBool *hadPosix, UErrorCode* status) {
+    char buf[ULOC_FULLNAME_CAPACITY];
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    int32_t len, i;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    len = uloc_getVariant(localeID, buf, sizeof(buf), &tmpStatus);
+    if (U_FAILURE(tmpStatus) || tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return;
+    }
+
+    if (len > 0) {
+        char *p, *pVar;
+        UBool bNext = TRUE;
+        VariantListEntry *var;
+        VariantListEntry *varFirst = NULL;
+
+        pVar = NULL;
+        p = buf;
+        while (bNext) {
+            if (*p == SEP || *p == LOCALE_SEP || *p == 0) {
+                if (*p == 0) {
+                    bNext = FALSE;
+                } else {
+                    *p = 0; /* terminate */
+                }
+                if (pVar == NULL) {
+                    if (strict) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        break;
+                    }
+                    /* ignore empty variant */
+                } else {
+                    /* ICU uses upper case letters for variants, but
+                       the canonical format is lowercase in BCP47 */
+                    for (i = 0; *(pVar + i) != 0; i++) {
+                        *(pVar + i) = uprv_tolower(*(pVar + i));
+                    }
+
+                    /* validate */
+                    if (_isVariantSubtag(pVar, -1)) {
+                        if (uprv_strcmp(pVar,POSIX_VALUE) || len != (int32_t)uprv_strlen(POSIX_VALUE)) {
+                            /* emit the variant to the list */
+                            var = (VariantListEntry*)uprv_malloc(sizeof(VariantListEntry));
+                            if (var == NULL) {
+                                *status = U_MEMORY_ALLOCATION_ERROR;
+                                break;
+                            }
+                            var->variant = pVar;
+                            if (!_addVariantToList(&varFirst, var)) {
+                                /* duplicated variant */
+                                uprv_free(var);
+                                if (strict) {
+                                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                                    break;
+                                }
+                            }
+                        } else {
+                            /* Special handling for POSIX variant, need to remember that we had it and then */
+                            /* treat it like an extension later. */
+                            *hadPosix = TRUE;
+                        }
+                    } else if (strict) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        break;
+                    } else if (_isPrivateuseValueSubtag(pVar, -1)) {
+                        /* Handle private use subtags separately */
+                        break;
+                    }
+                }
+                /* reset variant starting position */
+                pVar = NULL;
+            } else if (pVar == NULL) {
+                pVar = p;
+            }
+            p++;
+        }
+
+        if (U_SUCCESS(*status)) {
+            if (varFirst != NULL) {
+                int32_t varLen;
+
+                /* per UTS35, we should sort the variants */
+                _sortVariants(varFirst);
+
+                /* write out validated/normalized variants to the target */
+                var = varFirst;
+                while (var != NULL) {
+                    sink.Append("-", 1);
+                    varLen = (int32_t)uprv_strlen(var->variant);
+                    sink.Append(var->variant, varLen);
+                    var = var->next;
+                }
+            }
+        }
+
+        /* clean up */
+        var = varFirst;
+        while (var != NULL) {
+            VariantListEntry *tmpVar = var->next;
+            uprv_free(var);
+            var = tmpVar;
+        }
+
+        if (U_FAILURE(*status)) {
+            return;
+        }
+    }
+}
+
+static void
+_appendKeywordsToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UBool hadPosix, UErrorCode* status) {
+    char attrBuf[ULOC_KEYWORD_AND_VALUES_CAPACITY] = { 0 };
+    int32_t attrBufLength = 0;
+
+    icu::MemoryPool<AttributeListEntry> attrPool;
+    icu::MemoryPool<ExtensionListEntry> extPool;
+    icu::MemoryPool<icu::CharString> strPool;
+
+    icu::LocalUEnumerationPointer keywordEnum(uloc_openKeywords(localeID, status));
+    if (U_FAILURE(*status) && !hadPosix) {
+        return;
+    }
+    if (keywordEnum.isValid() || hadPosix) {
+        /* reorder extensions */
+        int32_t len;
+        const char *key;
+        ExtensionListEntry *firstExt = NULL;
+        ExtensionListEntry *ext;
+        AttributeListEntry *firstAttr = NULL;
+        AttributeListEntry *attr;
+        icu::MemoryPool<icu::CharString> extBufPool;
+        const char *bcpKey=nullptr, *bcpValue=nullptr;
+        UErrorCode tmpStatus = U_ZERO_ERROR;
+        int32_t keylen;
+        UBool isBcpUExt;
+
+        while (TRUE) {
+            key = uenum_next(keywordEnum.getAlias(), NULL, status);
+            if (key == NULL) {
+                break;
+            }
+
+            icu::CharString buf;
+            {
+                icu::CharStringByteSink sink(&buf);
+                ulocimp_getKeywordValue(localeID, key, sink, &tmpStatus);
+            }
+            len = buf.length();
+
+            if (U_FAILURE(tmpStatus)) {
+                if (tmpStatus == U_MEMORY_ALLOCATION_ERROR) {
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    break;
+                }
+                if (strict) {
+                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                    break;
+                }
+                /* ignore this keyword */
+                tmpStatus = U_ZERO_ERROR;
+                continue;
+            }
+
+            keylen = (int32_t)uprv_strlen(key);
+            isBcpUExt = (keylen > 1);
+
+            /* special keyword used for representing Unicode locale attributes */
+            if (uprv_strcmp(key, LOCALE_ATTRIBUTE_KEY) == 0) {
+                if (len > 0) {
+                    int32_t i = 0;
+                    while (TRUE) {
+                        attrBufLength = 0;
+                        for (; i < len; i++) {
+                            if (buf[i] != '-') {
+                                attrBuf[attrBufLength++] = buf[i];
+                            } else {
+                                i++;
+                                break;
+                            }
+                        }
+                        if (attrBufLength > 0) {
+                            attrBuf[attrBufLength] = 0;
+
+                        } else if (i >= len){
+                            break;
+                        }
+
+                        /* create AttributeListEntry */
+                        attr = attrPool.create();
+                        if (attr == NULL) {
+                            *status = U_MEMORY_ALLOCATION_ERROR;
+                            break;
+                        }
+                        icu::CharString* attrValue =
+                                strPool.create(attrBuf, attrBufLength, *status);
+                        if (attrValue == NULL) {
+                            *status = U_MEMORY_ALLOCATION_ERROR;
+                            break;
+                        }
+                        if (U_FAILURE(*status)) {
+                            break;
+                        }
+                        attr->attribute = attrValue->data();
+
+                        if (!_addAttributeToList(&firstAttr, attr)) {
+                            if (strict) {
+                                *status = U_ILLEGAL_ARGUMENT_ERROR;
+                                break;
+                            }
+                        }
+                    }
+                    /* for a place holder ExtensionListEntry */
+                    bcpKey = LOCALE_ATTRIBUTE_KEY;
+                    bcpValue = NULL;
+                }
+            } else if (isBcpUExt) {
+                bcpKey = uloc_toUnicodeLocaleKey(key);
+                if (bcpKey == NULL) {
+                    if (strict) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        break;
+                    }
+                    continue;
+                }
+
+                /* we've checked buf is null-terminated above */
+                bcpValue = uloc_toUnicodeLocaleType(key, buf.data());
+                if (bcpValue == NULL) {
+                    if (strict) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        break;
+                    }
+                    continue;
+                }
+                if (bcpValue == buf.data()) {
+                    /*
+                    When uloc_toUnicodeLocaleType(key, buf) returns the
+                    input value as is, the value is well-formed, but has
+                    no known mapping. This implementation normalizes the
+                    value to lower case
+                    */
+                    icu::CharString* extBuf = extBufPool.create(buf, tmpStatus);
+
+                    if (extBuf == nullptr) {
+                        *status = U_MEMORY_ALLOCATION_ERROR;
+                        break;
+                    }
+                    if (U_FAILURE(tmpStatus)) {
+                        *status = tmpStatus;
+                        break;
+                    }
+
+                    T_CString_toLowerCase(extBuf->data());
+                    bcpValue = extBuf->data();
+                }
+            } else {
+                if (*key == PRIVATEUSE) {
+                    if (!ultag_isPrivateuseValueSubtags(buf.data(), len)) {
+                        if (strict) {
+                            *status = U_ILLEGAL_ARGUMENT_ERROR;
+                            break;
+                        }
+                        continue;
+                    }
+                } else {
+                    if (!_isExtensionSingleton(key, keylen) || !ultag_isExtensionSubtags(buf.data(), len)) {
+                        if (strict) {
+                            *status = U_ILLEGAL_ARGUMENT_ERROR;
+                            break;
+                        }
+                        continue;
+                    }
+                }
+                bcpKey = key;
+                icu::CharString* extBuf =
+                    extBufPool.create(buf.data(), len, tmpStatus);
+                if (extBuf == nullptr) {
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    break;
+                }
+                if (U_FAILURE(tmpStatus)) {
+                    *status = tmpStatus;
+                    break;
+                }
+                bcpValue = extBuf->data();
+            }
+
+            /* create ExtensionListEntry */
+            ext = extPool.create();
+            if (ext == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                break;
+            }
+            ext->key = bcpKey;
+            ext->value = bcpValue;
+
+            if (!_addExtensionToList(&firstExt, ext, TRUE)) {
+                if (strict) {
+                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                    break;
+                }
+            }
+        }
+
+        /* Special handling for POSIX variant - add the keywords for POSIX */
+        if (hadPosix) {
+            /* create ExtensionListEntry for POSIX */
+            ext = extPool.create();
+            if (ext == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            ext->key = POSIX_KEY;
+            ext->value = POSIX_VALUE;
+
+            if (!_addExtensionToList(&firstExt, ext, TRUE)) {
+                // Silently ignore errors.
+            }
+        }
+
+        if (U_SUCCESS(*status) && (firstExt != NULL || firstAttr != NULL)) {
+            UBool startLDMLExtension = FALSE;
+            for (ext = firstExt; ext; ext = ext->next) {
+                if (!startLDMLExtension && uprv_strlen(ext->key) > 1) {
+                    /* first LDML u singlton extension */
+                   sink.Append("-u", 2);
+                   startLDMLExtension = TRUE;
+                }
+
+                /* write out the sorted BCP47 attributes, extensions and private use */
+                if (uprv_strcmp(ext->key, LOCALE_ATTRIBUTE_KEY) == 0) {
+                    /* write the value for the attributes */
+                    for (attr = firstAttr; attr; attr = attr->next) {
+                        sink.Append("-", 1);
+                        sink.Append(
+                                attr->attribute, static_cast<int32_t>(uprv_strlen(attr->attribute)));
+                    }
+                } else {
+                    sink.Append("-", 1);
+                    sink.Append(ext->key, static_cast<int32_t>(uprv_strlen(ext->key)));
+                    if (uprv_strcmp(ext->value, "true") != 0 &&
+                        uprv_strcmp(ext->value, "yes") != 0) {
+                      sink.Append("-", 1);
+                      sink.Append(ext->value, static_cast<int32_t>(uprv_strlen(ext->value)));
+                    }
+                }
+            }
+        }
+    }
+}
+
+/**
+ * Append keywords parsed from LDML extension value
+ * e.g. "u-ca-gregory-co-trad" -> {calendar = gregorian} {collation = traditional}
+ * Note: char* buf is used for storing keywords
+ */
+static void
+_appendLDMLExtensionAsKeywords(const char* ldmlext, ExtensionListEntry** appendTo, icu::MemoryPool<ExtensionListEntry>& extPool, icu::MemoryPool<icu::CharString>& kwdBuf, UBool *posixVariant, UErrorCode *status) {
+    const char *pTag;   /* beginning of current subtag */
+    const char *pKwds;  /* beginning of key-type pairs */
+    UBool variantExists = *posixVariant;
+
+    ExtensionListEntry *kwdFirst = NULL;    /* first LDML keyword */
+    ExtensionListEntry *kwd, *nextKwd;
+
+    int32_t len;
+
+    /* Reset the posixVariant value */
+    *posixVariant = FALSE;
+
+    pTag = ldmlext;
+    pKwds = NULL;
+
+    {
+        AttributeListEntry *attrFirst = NULL;   /* first attribute */
+        AttributeListEntry *attr, *nextAttr;
+
+        char attrBuf[ULOC_KEYWORD_AND_VALUES_CAPACITY];
+        int32_t attrBufIdx = 0;
+
+        icu::MemoryPool<AttributeListEntry> attrPool;
+
+        /* Iterate through u extension attributes */
+        while (*pTag) {
+            /* locate next separator char */
+            for (len = 0; *(pTag + len) && *(pTag + len) != SEP; len++);
+
+            if (ultag_isUnicodeLocaleKey(pTag, len)) {
+                pKwds = pTag;
+                break;
+            }
+
+            /* add this attribute to the list */
+            attr = attrPool.create();
+            if (attr == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+
+            if (len < (int32_t)sizeof(attrBuf) - attrBufIdx) {
+                uprv_memcpy(&attrBuf[attrBufIdx], pTag, len);
+                attrBuf[attrBufIdx + len] = 0;
+                attr->attribute = &attrBuf[attrBufIdx];
+                attrBufIdx += (len + 1);
+            } else {
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+                return;
+            }
+
+            // duplicate attribute is ignored, causes no error.
+            _addAttributeToList(&attrFirst, attr);
+
+            /* next tag */
+            pTag += len;
+            if (*pTag) {
+                /* next to the separator */
+                pTag++;
+            }
+        }
+
+        if (attrFirst) {
+            /* emit attributes as an LDML keyword, e.g. attribute=attr1-attr2 */
+
+            kwd = extPool.create();
+            if (kwd == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+
+            icu::CharString* value = kwdBuf.create();
+            if (value == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+
+            /* attribute subtags sorted in alphabetical order as type */
+            attr = attrFirst;
+            while (attr != NULL) {
+                nextAttr = attr->next;
+                if (attr != attrFirst) {
+                    value->append('-', *status);
+                }
+                value->append(attr->attribute, *status);
+                attr = nextAttr;
+            }
+            if (U_FAILURE(*status)) {
+                return;
+            }
+
+            kwd->key = LOCALE_ATTRIBUTE_KEY;
+            kwd->value = value->data();
+
+            if (!_addExtensionToList(&kwdFirst, kwd, FALSE)) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+                return;
+            }
+        }
+    }
+
+    if (pKwds) {
+        const char *pBcpKey = NULL;     /* u extenstion key subtag */
+        const char *pBcpType = NULL;    /* beginning of u extension type subtag(s) */
+        int32_t bcpKeyLen = 0;
+        int32_t bcpTypeLen = 0;
+        UBool isDone = FALSE;
+
+        pTag = pKwds;
+        /* BCP47 representation of LDML key/type pairs */
+        while (!isDone) {
+            const char *pNextBcpKey = NULL;
+            int32_t nextBcpKeyLen = 0;
+            UBool emitKeyword = FALSE;
+
+            if (*pTag) {
+                /* locate next separator char */
+                for (len = 0; *(pTag + len) && *(pTag + len) != SEP; len++);
+
+                if (ultag_isUnicodeLocaleKey(pTag, len)) {
+                    if (pBcpKey) {
+                        emitKeyword = TRUE;
+                        pNextBcpKey = pTag;
+                        nextBcpKeyLen = len;
+                    } else {
+                        pBcpKey = pTag;
+                        bcpKeyLen = len;
+                    }
+                } else {
+                    U_ASSERT(pBcpKey != NULL);
+                    /* within LDML type subtags */
+                    if (pBcpType) {
+                        bcpTypeLen += (len + 1);
+                    } else {
+                        pBcpType = pTag;
+                        bcpTypeLen = len;
+                    }
+                }
+
+                /* next tag */
+                pTag += len;
+                if (*pTag) {
+                    /* next to the separator */
+                    pTag++;
+                }
+            } else {
+                /* processing last one */
+                emitKeyword = TRUE;
+                isDone = TRUE;
+            }
+
+            if (emitKeyword) {
+                const char *pKey = NULL;    /* LDML key */
+                const char *pType = NULL;   /* LDML type */
+
+                char bcpKeyBuf[3];          /* BCP key length is always 2 for now */
+
+                U_ASSERT(pBcpKey != NULL);
+
+                if (bcpKeyLen >= (int32_t)sizeof(bcpKeyBuf)) {
+                    /* the BCP key is invalid */
+                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                    return;
+                }
+                U_ASSERT(bcpKeyLen <= 2);
+
+                uprv_strncpy(bcpKeyBuf, pBcpKey, bcpKeyLen);
+                bcpKeyBuf[bcpKeyLen] = 0;
+
+                /* u extension key to LDML key */
+                pKey = uloc_toLegacyKey(bcpKeyBuf);
+                if (pKey == NULL) {
+                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                    return;
+                }
+                if (pKey == bcpKeyBuf) {
+                    /*
+                    The key returned by toLegacyKey points to the input buffer.
+                    We normalize the result key to lower case.
+                    */
+                    T_CString_toLowerCase(bcpKeyBuf);
+                    icu::CharString* key = kwdBuf.create(bcpKeyBuf, bcpKeyLen, *status);
+                    if (key == NULL) {
+                        *status = U_MEMORY_ALLOCATION_ERROR;
+                        return;
+                    }
+                    if (U_FAILURE(*status)) {
+                        return;
+                    }
+                    pKey = key->data();
+                }
+
+                if (pBcpType) {
+                    char bcpTypeBuf[128];       /* practically long enough even considering multiple subtag type */
+                    if (bcpTypeLen >= (int32_t)sizeof(bcpTypeBuf)) {
+                        /* the BCP type is too long */
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        return;
+                    }
+
+                    uprv_strncpy(bcpTypeBuf, pBcpType, bcpTypeLen);
+                    bcpTypeBuf[bcpTypeLen] = 0;
+
+                    /* BCP type to locale type */
+                    pType = uloc_toLegacyType(pKey, bcpTypeBuf);
+                    if (pType == NULL) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        return;
+                    }
+                    if (pType == bcpTypeBuf) {
+                        /*
+                        The type returned by toLegacyType points to the input buffer.
+                        We normalize the result type to lower case.
+                        */
+                        /* normalize to lower case */
+                        T_CString_toLowerCase(bcpTypeBuf);
+                        icu::CharString* type = kwdBuf.create(bcpTypeBuf, bcpTypeLen, *status);
+                        if (type == NULL) {
+                            *status = U_MEMORY_ALLOCATION_ERROR;
+                            return;
+                        }
+                        if (U_FAILURE(*status)) {
+                            return;
+                        }
+                        pType = type->data();
+                    }
+                } else {
+                    /* typeless - default type value is "yes" */
+                    pType = LOCALE_TYPE_YES;
+                }
+
+                /* Special handling for u-va-posix, since we want to treat this as a variant, 
+                   not as a keyword */
+                if (!variantExists && !uprv_strcmp(pKey, POSIX_KEY) && !uprv_strcmp(pType, POSIX_VALUE) ) {
+                    *posixVariant = TRUE;
+                } else {
+                    /* create an ExtensionListEntry for this keyword */
+                    kwd = extPool.create();
+                    if (kwd == NULL) {
+                        *status = U_MEMORY_ALLOCATION_ERROR;
+                        return;
+                    }
+
+                    kwd->key = pKey;
+                    kwd->value = pType;
+
+                    if (!_addExtensionToList(&kwdFirst, kwd, FALSE)) {
+                        // duplicate keyword is allowed, Only the first
+                        // is honored.
+                    }
+                }
+
+                pBcpKey = pNextBcpKey;
+                bcpKeyLen = pNextBcpKey != NULL ? nextBcpKeyLen : 0;
+                pBcpType = NULL;
+                bcpTypeLen = 0;
+            }
+        }
+    }
+
+    kwd = kwdFirst;
+    while (kwd != NULL) {
+        nextKwd = kwd->next;
+        _addExtensionToList(appendTo, kwd, FALSE);
+        kwd = nextKwd;
+    }
+}
+
+
+static void
+_appendKeywords(ULanguageTag* langtag, icu::ByteSink& sink, UErrorCode* status) {
+    int32_t i, n;
+    int32_t len;
+    ExtensionListEntry *kwdFirst = NULL;
+    ExtensionListEntry *kwd;
+    const char *key, *type;
+    icu::MemoryPool<ExtensionListEntry> extPool;
+    icu::MemoryPool<icu::CharString> kwdBuf;
+    UBool posixVariant = FALSE;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    /* Determine if variants already exists */
+    if (ultag_getVariantsSize(langtag)) {
+        posixVariant = TRUE;
+    }
+
+    n = ultag_getExtensionsSize(langtag);
+
+    /* resolve locale keywords and reordering keys */
+    for (i = 0; i < n; i++) {
+        key = ultag_getExtensionKey(langtag, i);
+        type = ultag_getExtensionValue(langtag, i);
+        if (*key == LDMLEXT) {
+            _appendLDMLExtensionAsKeywords(type, &kwdFirst, extPool, kwdBuf, &posixVariant, status);
+            if (U_FAILURE(*status)) {
+                break;
+            }
+        } else {
+            kwd = extPool.create();
+            if (kwd == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                break;
+            }
+            kwd->key = key;
+            kwd->value = type;
+            if (!_addExtensionToList(&kwdFirst, kwd, FALSE)) {
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+                break;
+            }
+        }
+    }
+
+    if (U_SUCCESS(*status)) {
+        type = ultag_getPrivateUse(langtag);
+        if ((int32_t)uprv_strlen(type) > 0) {
+            /* add private use as a keyword */
+            kwd = extPool.create();
+            if (kwd == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+            } else {
+                kwd->key = PRIVATEUSE_KEY;
+                kwd->value = type;
+                if (!_addExtensionToList(&kwdFirst, kwd, FALSE)) {
+                    *status = U_ILLEGAL_ARGUMENT_ERROR;
+                }
+            }
+        }
+    }
+
+    /* If a POSIX variant was in the extensions, write it out before writing the keywords. */
+
+    if (U_SUCCESS(*status) && posixVariant) {
+        len = (int32_t) uprv_strlen(_POSIX);
+        sink.Append(_POSIX, len);
+    }
+
+    if (U_SUCCESS(*status) && kwdFirst != NULL) {
+        /* write out the sorted keywords */
+        UBool firstValue = TRUE;
+        kwd = kwdFirst;
+        do {
+            if (firstValue) {
+                sink.Append("@", 1);
+                firstValue = FALSE;
+            } else {
+                sink.Append(";", 1);
+            }
+
+            /* key */
+            len = (int32_t)uprv_strlen(kwd->key);
+            sink.Append(kwd->key, len);
+            sink.Append("=", 1);
+
+            /* type */
+            len = (int32_t)uprv_strlen(kwd->value);
+            sink.Append(kwd->value, len);
+
+            kwd = kwd->next;
+        } while (kwd);
+    }
+}
+
+static void
+_appendPrivateuseToLanguageTag(const char* localeID, icu::ByteSink& sink, UBool strict, UBool hadPosix, UErrorCode* status) {
+    (void)hadPosix;
+    char buf[ULOC_FULLNAME_CAPACITY];
+    char tmpAppend[ULOC_FULLNAME_CAPACITY];
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    int32_t len, i;
+    int32_t reslen = 0;
+    int32_t capacity = sizeof tmpAppend;
+
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    len = uloc_getVariant(localeID, buf, sizeof(buf), &tmpStatus);
+    if (U_FAILURE(tmpStatus) || tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+        if (strict) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return;
+    }
+
+    if (len > 0) {
+        char *p, *pPriv;
+        UBool bNext = TRUE;
+        UBool firstValue = TRUE;
+        UBool writeValue;
+
+        pPriv = NULL;
+        p = buf;
+        while (bNext) {
+            writeValue = FALSE;
+            if (*p == SEP || *p == LOCALE_SEP || *p == 0) {
+                if (*p == 0) {
+                    bNext = FALSE;
+                } else {
+                    *p = 0; /* terminate */
+                }
+                if (pPriv != NULL) {
+                    /* Private use in the canonical format is lowercase in BCP47 */
+                    for (i = 0; *(pPriv + i) != 0; i++) {
+                        *(pPriv + i) = uprv_tolower(*(pPriv + i));
+                    }
+
+                    /* validate */
+                    if (_isPrivateuseValueSubtag(pPriv, -1)) {
+                        if (firstValue) {
+                            if (!_isVariantSubtag(pPriv, -1)) {
+                                writeValue = TRUE;
+                            }
+                        } else {
+                            writeValue = TRUE;
+                        }
+                    } else if (strict) {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        break;
+                    } else {
+                        break;
+                    }
+
+                    if (writeValue) {
+                        if (reslen < capacity) {
+                            tmpAppend[reslen++] = SEP;
+                        }
+
+                        if (firstValue) {
+                            if (reslen < capacity) {
+                                tmpAppend[reslen++] = *PRIVATEUSE_KEY;
+                            }
+
+                            if (reslen < capacity) {
+                                tmpAppend[reslen++] = SEP;
+                            }
+
+                            len = (int32_t)uprv_strlen(PRIVUSE_VARIANT_PREFIX);
+                            if (reslen < capacity) {
+                                uprv_memcpy(tmpAppend + reslen, PRIVUSE_VARIANT_PREFIX, uprv_min(len, capacity - reslen));
+                            }
+                            reslen += len;
+
+                            if (reslen < capacity) {
+                                tmpAppend[reslen++] = SEP;
+                            }
+
+                            firstValue = FALSE;
+                        }
+
+                        len = (int32_t)uprv_strlen(pPriv);
+                        if (reslen < capacity) {
+                            uprv_memcpy(tmpAppend + reslen, pPriv, uprv_min(len, capacity - reslen));
+                        }
+                        reslen += len;
+                    }
+                }
+                /* reset private use starting position */
+                pPriv = NULL;
+            } else if (pPriv == NULL) {
+                pPriv = p;
+            }
+            p++;
+        }
+
+        if (U_FAILURE(*status)) {
+            return;
+        }
+    }
+
+    if (U_SUCCESS(*status)) {
+        len = reslen;
+        sink.Append(tmpAppend, len);
+    }
+}
+
+/*
+* -------------------------------------------------
+*
+* ultag_ functions
+*
+* -------------------------------------------------
+*/
+
+/* Bit flags used by the parser */
+#define LANG 0x0001
+#define EXTL 0x0002
+#define SCRT 0x0004
+#define REGN 0x0008
+#define VART 0x0010
+#define EXTS 0x0020
+#define EXTV 0x0040
+#define PRIV 0x0080
+
+/**
+ * Ticket #12705 - The optimizer in Visual Studio 2015 Update 3 has problems optimizing this function.
+ * As a work-around, optimization is disabled for this function on VS2015 and VS2017.
+ * This work-around should be removed once the following versions of Visual Studio are no
+ * longer supported: All versions of VS2015/VS2017, and versions of VS2019 below 16.4.
+ */
+#if defined(_MSC_VER) && (_MSC_VER >= 1900) && (_MSC_VER < 1924)
+#pragma optimize( "", off )
+#endif
+
+static ULanguageTag*
+ultag_parse(const char* tag, int32_t tagLen, int32_t* parsedLen, UErrorCode* status) {
+    char *tagBuf;
+    int16_t next;
+    char *pSubtag, *pNext, *pLastGoodPosition;
+    int32_t subtagLen;
+    int32_t extlangIdx;
+    ExtensionListEntry *pExtension;
+    char *pExtValueSubtag, *pExtValueSubtagEnd;
+    int32_t i;
+    UBool privateuseVar = FALSE;
+    int32_t legacyLen = 0;
+
+    if (parsedLen != NULL) {
+        *parsedLen = 0;
+    }
+
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    if (tagLen < 0) {
+        tagLen = (int32_t)uprv_strlen(tag);
+    }
+
+    /* copy the entire string */
+    tagBuf = (char*)uprv_malloc(tagLen + 1);
+    if (tagBuf == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memcpy(tagBuf, tag, tagLen);
+    *(tagBuf + tagLen) = 0;
+
+    /* create a ULanguageTag */
+    icu::LocalULanguageTagPointer t(
+            (ULanguageTag*)uprv_malloc(sizeof(ULanguageTag)));
+    if (t.isNull()) {
+        uprv_free(tagBuf);
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    _initializeULanguageTag(t.getAlias());
+    t->buf = tagBuf;
+
+    if (tagLen < MINLEN) {
+        /* the input tag is too short - return empty ULanguageTag */
+        return t.orphan();
+    }
+
+    size_t parsedLenDelta = 0;
+    // Legacy tag will be consider together. Legacy tag with intervening
+    // script and region such as art-DE-lojban or art-Latn-lojban won't be
+    // matched.
+    /* check if the tag is legacy */
+    for (i = 0; i < UPRV_LENGTHOF(LEGACY); i += 2) {
+        int32_t checkLegacyLen = static_cast<int32_t>(uprv_strlen(LEGACY[i]));
+        if (tagLen < checkLegacyLen) {
+            continue;
+        }
+        if (tagLen > checkLegacyLen && tagBuf[checkLegacyLen] != '-') {
+            // make sure next char is '-'.
+            continue;
+        }
+        if (uprv_strnicmp(LEGACY[i], tagBuf, checkLegacyLen) == 0) {
+            int32_t newTagLength;
+
+            legacyLen = checkLegacyLen;  /* back up for output parsedLen */
+            int32_t replacementLen = static_cast<int32_t>(uprv_strlen(LEGACY[i+1]));
+            newTagLength = replacementLen + tagLen - checkLegacyLen;
+            if (tagLen < newTagLength) {
+                uprv_free(tagBuf);
+                tagBuf = (char*)uprv_malloc(newTagLength + 1);
+                if (tagBuf == NULL) {
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    return NULL;
+                }
+                t->buf = tagBuf;
+                tagLen = newTagLength;
+            }
+            parsedLenDelta = checkLegacyLen - replacementLen;
+            uprv_strcpy(t->buf, LEGACY[i + 1]);
+            if (checkLegacyLen != tagLen) {
+                uprv_strcpy(t->buf + replacementLen, tag + checkLegacyLen);
+            }
+            break;
+        }
+    }
+
+    if (legacyLen == 0) {
+        for (i = 0; i < UPRV_LENGTHOF(REDUNDANT); i += 2) {
+            const char* redundantTag = REDUNDANT[i];
+            size_t redundantTagLen = uprv_strlen(redundantTag);
+            // The preferred tag for a redundant tag is always shorter than redundant
+            // tag. A redundant tag may or may not be followed by other subtags.
+            // (i.e. "zh-yue" or "zh-yue-u-co-pinyin").
+            if (uprv_strnicmp(redundantTag, tagBuf, static_cast<uint32_t>(redundantTagLen)) == 0) {
+                const char* redundantTagEnd = tagBuf + redundantTagLen;
+                if (*redundantTagEnd  == '\0' || *redundantTagEnd == SEP) {
+                    const char* preferredTag = REDUNDANT[i + 1];
+                    size_t preferredTagLen = uprv_strlen(preferredTag);
+                    uprv_strncpy(t->buf, preferredTag, preferredTagLen);
+                    if (*redundantTagEnd == SEP) {
+                        uprv_memmove(tagBuf + preferredTagLen,
+                                     redundantTagEnd,
+                                     tagLen - redundantTagLen + 1);
+                    } else {
+                        tagBuf[preferredTagLen] = '\0';
+                    }
+                    // parsedLen should be the length of the input
+                    // before redundantTag is replaced by preferredTag.
+                    // Save the delta to add it back later.
+                    parsedLenDelta = redundantTagLen - preferredTagLen;
+                    break;
+                }
+            }
+        }
+    }
+
+    /*
+     * langtag      =   language
+     *                  ["-" script]
+     *                  ["-" region]
+     *                  *("-" variant)
+     *                  *("-" extension)
+     *                  ["-" privateuse]
+     */
+
+    next = LANG | PRIV;
+    pNext = pLastGoodPosition = tagBuf;
+    extlangIdx = 0;
+    pExtension = NULL;
+    pExtValueSubtag = NULL;
+    pExtValueSubtagEnd = NULL;
+
+    while (pNext) {
+        char *pSep;
+
+        pSubtag = pNext;
+
+        /* locate next separator char */
+        pSep = pSubtag;
+        while (*pSep) {
+            if (*pSep == SEP) {
+                break;
+            }
+            pSep++;
+        }
+        if (*pSep == 0) {
+            /* last subtag */
+            pNext = NULL;
+        } else {
+            pNext = pSep + 1;
+        }
+        subtagLen = (int32_t)(pSep - pSubtag);
+
+        if (next & LANG) {
+            if (ultag_isLanguageSubtag(pSubtag, subtagLen)) {
+                *pSep = 0;  /* terminate */
+                // TODO: move deprecated language code handling here.
+                t->language = T_CString_toLowerCase(pSubtag);
+
+                pLastGoodPosition = pSep;
+                next = SCRT | REGN | VART | EXTS | PRIV;
+                if (subtagLen <= 3)
+                  next |= EXTL;
+                continue;
+            }
+        }
+        if (next & EXTL) {
+            if (_isExtlangSubtag(pSubtag, subtagLen)) {
+                *pSep = 0;
+                t->extlang[extlangIdx++] = T_CString_toLowerCase(pSubtag);
+
+                pLastGoodPosition = pSep;
+                if (extlangIdx < 3) {
+                    next = EXTL | SCRT | REGN | VART | EXTS | PRIV;
+                } else {
+                    next = SCRT | REGN | VART | EXTS | PRIV;
+                }
+                continue;
+            }
+        }
+        if (next & SCRT) {
+            if (ultag_isScriptSubtag(pSubtag, subtagLen)) {
+                char *p = pSubtag;
+
+                *pSep = 0;
+
+                /* to title case */
+                *p = uprv_toupper(*p);
+                p++;
+                for (; *p; p++) {
+                    *p = uprv_tolower(*p);
+                }
+
+                t->script = pSubtag;
+
+                pLastGoodPosition = pSep;
+                next = REGN | VART | EXTS | PRIV;
+                continue;
+            }
+        }
+        if (next & REGN) {
+            if (ultag_isRegionSubtag(pSubtag, subtagLen)) {
+                *pSep = 0;
+                // TODO: move deprecated region code handling here.
+                t->region = T_CString_toUpperCase(pSubtag);
+
+                pLastGoodPosition = pSep;
+                next = VART | EXTS | PRIV;
+                continue;
+            }
+        }
+        if (next & VART) {
+            if (_isVariantSubtag(pSubtag, subtagLen) ||
+               (privateuseVar && _isPrivateuseVariantSubtag(pSubtag, subtagLen))) {
+                VariantListEntry *var;
+                UBool isAdded;
+
+                var = (VariantListEntry*)uprv_malloc(sizeof(VariantListEntry));
+                if (var == NULL) {
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    return NULL;
+                }
+                *pSep = 0;
+                var->variant = T_CString_toUpperCase(pSubtag);
+                isAdded = _addVariantToList(&(t->variants), var);
+                if (!isAdded) {
+                    /* duplicated variant entry */
+                    uprv_free(var);
+                    break;
+                }
+                pLastGoodPosition = pSep;
+                next = VART | EXTS | PRIV;
+                continue;
+            }
+        }
+        if (next & EXTS) {
+            if (_isExtensionSingleton(pSubtag, subtagLen)) {
+                if (pExtension != NULL) {
+                    if (pExtValueSubtag == NULL || pExtValueSubtagEnd == NULL) {
+                        /* the previous extension is incomplete */
+                        uprv_free(pExtension);
+                        pExtension = NULL;
+                        break;
+                    }
+
+                    /* terminate the previous extension value */
+                    *pExtValueSubtagEnd = 0;
+                    pExtension->value = T_CString_toLowerCase(pExtValueSubtag);
+
+                    /* insert the extension to the list */
+                    if (_addExtensionToList(&(t->extensions), pExtension, FALSE)) {
+                        pLastGoodPosition = pExtValueSubtagEnd;
+                    } else {
+                        /* stop parsing here */
+                        uprv_free(pExtension);
+                        pExtension = NULL;
+                        break;
+                    }
+                }
+
+                /* create a new extension */
+                pExtension = (ExtensionListEntry*)uprv_malloc(sizeof(ExtensionListEntry));
+                if (pExtension == NULL) {
+                    *status = U_MEMORY_ALLOCATION_ERROR;
+                    return NULL;
+                }
+                *pSep = 0;
+                pExtension->key = T_CString_toLowerCase(pSubtag);
+                pExtension->value = NULL;   /* will be set later */
+
+                /*
+                 * reset the start and the end location of extension value
+                 * subtags for this extension
+                 */
+                pExtValueSubtag = NULL;
+                pExtValueSubtagEnd = NULL;
+
+                next = EXTV;
+                continue;
+            }
+        }
+        if (next & EXTV) {
+            if (_isExtensionSubtag(pSubtag, subtagLen)) {
+                if (pExtValueSubtag == NULL) {
+                    /* if the start postion of this extension's value is not yet,
+                        this one is the first value subtag */
+                    pExtValueSubtag = pSubtag;
+                }
+
+                /* Mark the end of this subtag */
+                pExtValueSubtagEnd = pSep;
+                next = EXTS | EXTV | PRIV;
+
+                continue;
+            }
+        }
+        if (next & PRIV) {
+            if (uprv_tolower(*pSubtag) == PRIVATEUSE && subtagLen == 1) {
+                char *pPrivuseVal;
+
+                if (pExtension != NULL) {
+                    /* Process the last extension */
+                    if (pExtValueSubtag == NULL || pExtValueSubtagEnd == NULL) {
+                        /* the previous extension is incomplete */
+                        uprv_free(pExtension);
+                        pExtension = NULL;
+                        break;
+                    } else {
+                        /* terminate the previous extension value */
+                        *pExtValueSubtagEnd = 0;
+                        pExtension->value = T_CString_toLowerCase(pExtValueSubtag);
+
+                        /* insert the extension to the list */
+                        if (_addExtensionToList(&(t->extensions), pExtension, FALSE)) {
+                            pLastGoodPosition = pExtValueSubtagEnd;
+                            pExtension = NULL;
+                        } else {
+                        /* stop parsing here */
+                            uprv_free(pExtension);
+                            pExtension = NULL;
+                            break;
+                        }
+                    }
+                }
+
+                /* The rest of part will be private use value subtags */
+                if (pNext == NULL) {
+                    /* empty private use subtag */
+                    break;
+                }
+                /* back up the private use value start position */
+                pPrivuseVal = pNext;
+
+                /* validate private use value subtags */
+                while (pNext) {
+                    pSubtag = pNext;
+                    pSep = pSubtag;
+                    while (*pSep) {
+                        if (*pSep == SEP) {
+                            break;
+                        }
+                        pSep++;
+                    }
+                    if (*pSep == 0) {
+                        /* last subtag */
+                        pNext = NULL;
+                    } else {
+                        pNext = pSep + 1;
+                    }
+                    subtagLen = (int32_t)(pSep - pSubtag);
+
+                    if (uprv_strncmp(pSubtag, PRIVUSE_VARIANT_PREFIX, uprv_strlen(PRIVUSE_VARIANT_PREFIX)) == 0) {
+                        *pSep = 0;
+                        next = VART;
+                        privateuseVar = TRUE;
+                        break;
+                    } else if (_isPrivateuseValueSubtag(pSubtag, subtagLen)) {
+                        pLastGoodPosition = pSep;
+                    } else {
+                        break;
+                    }
+                }
+
+                if (next == VART) {
+                    continue;
+                }
+
+                if (pLastGoodPosition - pPrivuseVal > 0) {
+                    *pLastGoodPosition = 0;
+                    t->privateuse = T_CString_toLowerCase(pPrivuseVal);
+                }
+                /* No more subtags, exiting the parse loop */
+                break;
+            }
+            break;
+        }
+
+        /* If we fell through here, it means this subtag is illegal - quit parsing */
+        break;
+    }
+
+    if (pExtension != NULL) {
+        /* Process the last extension */
+        if (pExtValueSubtag == NULL || pExtValueSubtagEnd == NULL) {
+            /* the previous extension is incomplete */
+            uprv_free(pExtension);
+        } else {
+            /* terminate the previous extension value */
+            *pExtValueSubtagEnd = 0;
+            pExtension->value = T_CString_toLowerCase(pExtValueSubtag);
+            /* insert the extension to the list */
+            if (_addExtensionToList(&(t->extensions), pExtension, FALSE)) {
+                pLastGoodPosition = pExtValueSubtagEnd;
+            } else {
+                uprv_free(pExtension);
+            }
+        }
+    }
+
+    if (parsedLen != NULL) {
+        *parsedLen = (int32_t)(pLastGoodPosition - t->buf + parsedLenDelta);
+    }
+
+    return t.orphan();
+}
+
+// Ticket #12705 - Turn optimization back on.
+#if defined(_MSC_VER) && (_MSC_VER >= 1900) && (_MSC_VER < 1924)
+#pragma optimize( "", on )
+#endif
+
+static void
+ultag_close(ULanguageTag* langtag) {
+
+    if (langtag == NULL) {
+        return;
+    }
+
+    uprv_free(langtag->buf);
+
+    if (langtag->variants) {
+        VariantListEntry *curVar = langtag->variants;
+        while (curVar) {
+            VariantListEntry *nextVar = curVar->next;
+            uprv_free(curVar);
+            curVar = nextVar;
+        }
+    }
+
+    if (langtag->extensions) {
+        ExtensionListEntry *curExt = langtag->extensions;
+        while (curExt) {
+            ExtensionListEntry *nextExt = curExt->next;
+            uprv_free(curExt);
+            curExt = nextExt;
+        }
+    }
+
+    uprv_free(langtag);
+}
+
+static const char*
+ultag_getLanguage(const ULanguageTag* langtag) {
+    return langtag->language;
+}
+
+#if 0
+static const char*
+ultag_getJDKLanguage(const ULanguageTag* langtag) {
+    int32_t i;
+    for (i = 0; DEPRECATEDLANGS[i] != NULL; i += 2) {
+        if (uprv_compareInvCharsAsAscii(DEPRECATEDLANGS[i], langtag->language) == 0) {
+            return DEPRECATEDLANGS[i + 1];
+        }
+    }
+    return langtag->language;
+}
+#endif
+
+static const char*
+ultag_getExtlang(const ULanguageTag* langtag, int32_t idx) {
+    if (idx >= 0 && idx < MAXEXTLANG) {
+        return langtag->extlang[idx];
+    }
+    return NULL;
+}
+
+static int32_t
+ultag_getExtlangSize(const ULanguageTag* langtag) {
+    int32_t size = 0;
+    int32_t i;
+    for (i = 0; i < MAXEXTLANG; i++) {
+        if (langtag->extlang[i]) {
+            size++;
+        }
+    }
+    return size;
+}
+
+static const char*
+ultag_getScript(const ULanguageTag* langtag) {
+    return langtag->script;
+}
+
+static const char*
+ultag_getRegion(const ULanguageTag* langtag) {
+    return langtag->region;
+}
+
+static const char*
+ultag_getVariant(const ULanguageTag* langtag, int32_t idx) {
+    const char *var = NULL;
+    VariantListEntry *cur = langtag->variants;
+    int32_t i = 0;
+    while (cur) {
+        if (i == idx) {
+            var = cur->variant;
+            break;
+        }
+        cur = cur->next;
+        i++;
+    }
+    return var;
+}
+
+static int32_t
+ultag_getVariantsSize(const ULanguageTag* langtag) {
+    int32_t size = 0;
+    VariantListEntry *cur = langtag->variants;
+    while (TRUE) {
+        if (cur == NULL) {
+            break;
+        }
+        size++;
+        cur = cur->next;
+    }
+    return size;
+}
+
+static const char*
+ultag_getExtensionKey(const ULanguageTag* langtag, int32_t idx) {
+    const char *key = NULL;
+    ExtensionListEntry *cur = langtag->extensions;
+    int32_t i = 0;
+    while (cur) {
+        if (i == idx) {
+            key = cur->key;
+            break;
+        }
+        cur = cur->next;
+        i++;
+    }
+    return key;
+}
+
+static const char*
+ultag_getExtensionValue(const ULanguageTag* langtag, int32_t idx) {
+    const char *val = NULL;
+    ExtensionListEntry *cur = langtag->extensions;
+    int32_t i = 0;
+    while (cur) {
+        if (i == idx) {
+            val = cur->value;
+            break;
+        }
+        cur = cur->next;
+        i++;
+    }
+    return val;
+}
+
+static int32_t
+ultag_getExtensionsSize(const ULanguageTag* langtag) {
+    int32_t size = 0;
+    ExtensionListEntry *cur = langtag->extensions;
+    while (TRUE) {
+        if (cur == NULL) {
+            break;
+        }
+        size++;
+        cur = cur->next;
+    }
+    return size;
+}
+
+static const char*
+ultag_getPrivateUse(const ULanguageTag* langtag) {
+    return langtag->privateuse;
+}
+
+#if 0
+static const char*
+ultag_getLegacy(const ULanguageTag* langtag) {
+    return langtag->legacy;
+}
+#endif
+
+
+/*
+* -------------------------------------------------
+*
+* Locale/BCP47 conversion APIs, exposed as uloc_*
+*
+* -------------------------------------------------
+*/
+U_CAPI int32_t U_EXPORT2
+uloc_toLanguageTag(const char* localeID,
+                   char* langtag,
+                   int32_t langtagCapacity,
+                   UBool strict,
+                   UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    icu::CheckedArrayByteSink sink(langtag, langtagCapacity);
+    ulocimp_toLanguageTag(localeID, sink, strict, status);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*status)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(langtag, langtagCapacity, reslen, status);
+    }
+
+    return reslen;
+}
+
+
+U_CAPI void U_EXPORT2
+ulocimp_toLanguageTag(const char* localeID,
+                      icu::ByteSink& sink,
+                      UBool strict,
+                      UErrorCode* status) {
+    icu::CharString canonical;
+    int32_t reslen;
+    UErrorCode tmpStatus = U_ZERO_ERROR;
+    UBool hadPosix = FALSE;
+    const char* pKeywordStart;
+
+    /* Note: uloc_canonicalize returns "en_US_POSIX" for input locale ID "".  See #6835 */
+    int32_t resultCapacity = static_cast<int32_t>(uprv_strlen(localeID));
+    if (resultCapacity > 0) {
+        char* buffer;
+
+        for (;;) {
+            buffer = canonical.getAppendBuffer(
+                    /*minCapacity=*/resultCapacity,
+                    /*desiredCapacityHint=*/resultCapacity,
+                    resultCapacity,
+                    tmpStatus);
+
+            if (U_FAILURE(tmpStatus)) {
+                *status = tmpStatus;
+                return;
+            }
+
+            reslen =
+                uloc_canonicalize(localeID, buffer, resultCapacity, &tmpStatus);
+
+            if (tmpStatus != U_BUFFER_OVERFLOW_ERROR) {
+                break;
+            }
+
+            resultCapacity = reslen;
+            tmpStatus = U_ZERO_ERROR;
+        }
+
+        if (U_FAILURE(tmpStatus)) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return;
+        }
+
+        canonical.append(buffer, reslen, tmpStatus);
+        if (tmpStatus == U_STRING_NOT_TERMINATED_WARNING) {
+            tmpStatus = U_ZERO_ERROR;  // Terminators provided by CharString.
+        }
+
+        if (U_FAILURE(tmpStatus)) {
+            *status = tmpStatus;
+            return;
+        }
+    }
+
+    /* For handling special case - private use only tag */
+    pKeywordStart = locale_getKeywordsStart(canonical.data());
+    if (pKeywordStart == canonical.data()) {
+        int kwdCnt = 0;
+        UBool done = FALSE;
+
+        icu::LocalUEnumerationPointer kwdEnum(uloc_openKeywords(canonical.data(), &tmpStatus));
+        if (U_SUCCESS(tmpStatus)) {
+            kwdCnt = uenum_count(kwdEnum.getAlias(), &tmpStatus);
+            if (kwdCnt == 1) {
+                const char *key;
+                int32_t len = 0;
+
+                key = uenum_next(kwdEnum.getAlias(), &len, &tmpStatus);
+                if (len == 1 && *key == PRIVATEUSE) {
+                    icu::CharString buf;
+                    {
+                        icu::CharStringByteSink sink(&buf);
+                        ulocimp_getKeywordValue(localeID, key, sink, &tmpStatus);
+                    }
+                    if (U_SUCCESS(tmpStatus)) {
+                        if (ultag_isPrivateuseValueSubtags(buf.data(), buf.length())) {
+                            /* return private use only tag */
+                            static const char PREFIX[] = { PRIVATEUSE, SEP };
+                            sink.Append(PREFIX, sizeof(PREFIX));
+                            sink.Append(buf.data(), buf.length());
+                            done = TRUE;
+                        } else if (strict) {
+                            *status = U_ILLEGAL_ARGUMENT_ERROR;
+                            done = TRUE;
+                        }
+                        /* if not strict mode, then "und" will be returned */
+                    } else {
+                        *status = U_ILLEGAL_ARGUMENT_ERROR;
+                        done = TRUE;
+                    }
+                }
+            }
+            if (done) {
+                return;
+            }
+        }
+    }
+
+    _appendLanguageToLanguageTag(canonical.data(), sink, strict, status);
+    _appendScriptToLanguageTag(canonical.data(), sink, strict, status);
+    _appendRegionToLanguageTag(canonical.data(), sink, strict, status);
+    _appendVariantsToLanguageTag(canonical.data(), sink, strict, &hadPosix, status);
+    _appendKeywordsToLanguageTag(canonical.data(), sink, strict, hadPosix, status);
+    _appendPrivateuseToLanguageTag(canonical.data(), sink, strict, hadPosix, status);
+}
+
+
+U_CAPI int32_t U_EXPORT2
+uloc_forLanguageTag(const char* langtag,
+                    char* localeID,
+                    int32_t localeIDCapacity,
+                    int32_t* parsedLength,
+                    UErrorCode* status) {
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+
+    icu::CheckedArrayByteSink sink(localeID, localeIDCapacity);
+    ulocimp_forLanguageTag(langtag, -1, sink, parsedLength, status);
+
+    int32_t reslen = sink.NumberOfBytesAppended();
+
+    if (U_FAILURE(*status)) {
+        return reslen;
+    }
+
+    if (sink.Overflowed()) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+    } else {
+        u_terminateChars(localeID, localeIDCapacity, reslen, status);
+    }
+
+    return reslen;
+}
+
+
+U_CAPI void U_EXPORT2
+ulocimp_forLanguageTag(const char* langtag,
+                       int32_t tagLen,
+                       icu::ByteSink& sink,
+                       int32_t* parsedLength,
+                       UErrorCode* status) {
+    UBool isEmpty = TRUE;
+    const char *subtag, *p;
+    int32_t len;
+    int32_t i, n;
+    UBool noRegion = TRUE;
+
+    icu::LocalULanguageTagPointer lt(ultag_parse(langtag, tagLen, parsedLength, status));
+    if (U_FAILURE(*status)) {
+        return;
+    }
+
+    /* language */
+    subtag = ultag_getExtlangSize(lt.getAlias()) > 0 ? ultag_getExtlang(lt.getAlias(), 0) : ultag_getLanguage(lt.getAlias());
+    if (uprv_compareInvCharsAsAscii(subtag, LANG_UND) != 0) {
+        len = (int32_t)uprv_strlen(subtag);
+        if (len > 0) {
+            sink.Append(subtag, len);
+            isEmpty = FALSE;
+        }
+    }
+
+    /* script */
+    subtag = ultag_getScript(lt.getAlias());
+    len = (int32_t)uprv_strlen(subtag);
+    if (len > 0) {
+        sink.Append("_", 1);
+        isEmpty = FALSE;
+
+        /* write out the script in title case */
+        char c = uprv_toupper(*subtag);
+        sink.Append(&c, 1);
+        sink.Append(subtag + 1, len - 1);
+    }
+
+    /* region */
+    subtag = ultag_getRegion(lt.getAlias());
+    len = (int32_t)uprv_strlen(subtag);
+    if (len > 0) {
+        sink.Append("_", 1);
+        isEmpty = FALSE;
+
+        /* write out the region in upper case */
+        p = subtag;
+        while (*p) {
+            char c = uprv_toupper(*p);
+            sink.Append(&c, 1);
+            p++;
+        }
+        noRegion = FALSE;
+    }
+
+    /* variants */
+    _sortVariants(lt.getAlias()->variants);
+    n = ultag_getVariantsSize(lt.getAlias());
+    if (n > 0) {
+        if (noRegion) {
+            sink.Append("_", 1);
+            isEmpty = FALSE;
+        }
+
+        for (i = 0; i < n; i++) {
+            subtag = ultag_getVariant(lt.getAlias(), i);
+            sink.Append("_", 1);
+
+            /* write out the variant in upper case */
+            p = subtag;
+            while (*p) {
+                char c = uprv_toupper(*p);
+                sink.Append(&c, 1);
+                p++;
+            }
+        }
+    }
+
+    /* keywords */
+    n = ultag_getExtensionsSize(lt.getAlias());
+    subtag = ultag_getPrivateUse(lt.getAlias());
+    if (n > 0 || uprv_strlen(subtag) > 0) {
+        if (isEmpty && n > 0) {
+            /* need a language */
+            sink.Append(LANG_UND, LANG_UND_LEN);
+        }
+        _appendKeywords(lt.getAlias(), sink, status);
+    }
+}
diff --git a/thirdparty/icu4c/common/ulocimp.h b/thirdparty/icu4c/common/ulocimp.h
new file mode 100644
index 0000000000..5691fe9a77
--- /dev/null
+++ b/thirdparty/icu4c/common/ulocimp.h
@@ -0,0 +1,307 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2004-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef ULOCIMP_H
+#define ULOCIMP_H
+
+#include "unicode/bytestream.h"
+#include "unicode/uloc.h"
+
+#include "charstr.h"
+
+/**
+ * Create an iterator over the specified keywords list
+ * @param keywordList double-null terminated list. Will be copied.
+ * @param keywordListSize size in bytes of keywordList
+ * @param status err code
+ * @return enumeration (owned by caller) of the keyword list.
+ * @internal ICU 3.0
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openKeywordList(const char *keywordList, int32_t keywordListSize, UErrorCode* status);
+
+/**
+ * Look up a resource bundle table item with fallback on the table level.
+ * This is accessible so it can be called by C++ code.
+ */
+U_CAPI const UChar * U_EXPORT2
+uloc_getTableStringWithFallback(
+    const char *path,
+    const char *locale,
+    const char *tableKey,
+    const char *subTableKey,
+    const char *itemKey,
+    int32_t *pLength,
+    UErrorCode *pErrorCode);
+
+/*returns true if a is an ID separator false otherwise*/
+#define _isIDSeparator(a) (a == '_' || a == '-')
+
+U_CFUNC const char* 
+uloc_getCurrentCountryID(const char* oldID);
+
+U_CFUNC const char* 
+uloc_getCurrentLanguageID(const char* oldID);
+
+U_CFUNC void
+ulocimp_getKeywords(const char *localeID,
+             char prev,
+             icu::ByteSink& sink,
+             UBool valuesToo,
+             UErrorCode *status);
+
+icu::CharString U_EXPORT2
+ulocimp_getLanguage(const char *localeID,
+                    const char **pEnd,
+                    UErrorCode &status);
+
+icu::CharString U_EXPORT2
+ulocimp_getScript(const char *localeID,
+                  const char **pEnd,
+                  UErrorCode &status);
+
+icu::CharString U_EXPORT2
+ulocimp_getCountry(const char *localeID,
+                   const char **pEnd,
+                   UErrorCode &status);
+
+U_CAPI void U_EXPORT2
+ulocimp_getName(const char* localeID,
+                icu::ByteSink& sink,
+                UErrorCode* err);
+
+U_CAPI void U_EXPORT2
+ulocimp_getBaseName(const char* localeID,
+                    icu::ByteSink& sink,
+                    UErrorCode* err);
+
+U_CAPI void U_EXPORT2
+ulocimp_canonicalize(const char* localeID,
+                     icu::ByteSink& sink,
+                     UErrorCode* err);
+
+U_CAPI void U_EXPORT2
+ulocimp_getKeywordValue(const char* localeID,
+                        const char* keywordName,
+                        icu::ByteSink& sink,
+                        UErrorCode* status);
+
+/**
+ * Writes a well-formed language tag for this locale ID.
+ *
+ * **Note**: When `strict` is false, any locale fields which do not satisfy the
+ * BCP47 syntax requirement will be omitted from the result.  When `strict` is
+ * true, this function sets U_ILLEGAL_ARGUMENT_ERROR to the `err` if any locale
+ * fields do not satisfy the BCP47 syntax requirement.
+ *
+ * @param localeID  the input locale ID
+ * @param sink      the output sink receiving the BCP47 language
+ *                  tag for this Locale.
+ * @param strict    boolean value indicating if the function returns
+ *                  an error for an ill-formed input locale ID.
+ * @param err       error information if receiving the language
+ *                  tag failed.
+ * @return          The length of the BCP47 language tag.
+ *
+ * @internal ICU 64
+ */
+U_CAPI void U_EXPORT2
+ulocimp_toLanguageTag(const char* localeID,
+                      icu::ByteSink& sink,
+                      UBool strict,
+                      UErrorCode* err);
+
+/**
+ * Returns a locale ID for the specified BCP47 language tag string.
+ * If the specified language tag contains any ill-formed subtags,
+ * the first such subtag and all following subtags are ignored.
+ * <p>
+ * This implements the 'Language-Tag' production of BCP 47, and so
+ * supports legacy language tags (marked as “Type: grandfathered” in BCP 47)
+ * (regular and irregular) as well as private use language tags.
+ *
+ * Private use tags are represented as 'x-whatever',
+ * and legacy tags are converted to their canonical replacements where they exist.
+ *
+ * Note that a few legacy tags have no modern replacement;
+ * these will be converted using the fallback described in
+ * the first paragraph, so some information might be lost.
+ *
+ * @param langtag   the input BCP47 language tag.
+ * @param tagLen    the length of langtag, or -1 to call uprv_strlen().
+ * @param sink      the output sink receiving a locale ID for the
+ *                  specified BCP47 language tag.
+ * @param parsedLength  if not NULL, successfully parsed length
+ *                      for the input language tag is set.
+ * @param err       error information if receiving the locald ID
+ *                  failed.
+ * @internal ICU 63
+ */
+U_CAPI void U_EXPORT2
+ulocimp_forLanguageTag(const char* langtag,
+                       int32_t tagLen,
+                       icu::ByteSink& sink,
+                       int32_t* parsedLength,
+                       UErrorCode* err);
+
+/**
+ * Get the region to use for supplemental data lookup. Uses
+ * (1) any region specified by locale tag "rg"; if none then
+ * (2) any unicode_region_tag in the locale ID; if none then
+ * (3) if inferRegion is true, the region suggested by
+ * getLikelySubtags on the localeID.
+ * If no region is found, returns length 0.
+ * 
+ * @param localeID
+ *     The complete locale ID (with keywords) from which
+ *     to get the region to use for supplemental data.
+ * @param inferRegion
+ *     If true, will try to infer region from localeID if
+ *     no other region is found.
+ * @param region
+ *     Buffer in which to put the region ID found; should
+ *     have a capacity at least ULOC_COUNTRY_CAPACITY. 
+ * @param regionCapacity
+ *     The actual capacity of the region buffer.
+ * @param status
+ *     Pointer to in/out UErrorCode value for latest status.
+ * @return
+ *     The length of any region code found, or 0 if none.
+ * @internal ICU 57
+ */
+U_CAPI int32_t U_EXPORT2
+ulocimp_getRegionForSupplementalData(const char *localeID, UBool inferRegion,
+                                     char *region, int32_t regionCapacity, UErrorCode* status);
+
+/**
+ * Add the likely subtags for a provided locale ID, per the algorithm described
+ * in the following CLDR technical report:
+ *
+ *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+ *
+ * If localeID is already in the maximal form, or there is no data available
+ * for maximization, it will be copied to the output buffer.  For example,
+ * "und-Zzzz" cannot be maximized, since there is no reasonable maximization.
+ *
+ * Examples:
+ *
+ * "en" maximizes to "en_Latn_US"
+ *
+ * "de" maximizes to "de_Latn_US"
+ *
+ * "sr" maximizes to "sr_Cyrl_RS"
+ *
+ * "sh" maximizes to "sr_Latn_RS" (Note this will not reverse.)
+ *
+ * "zh_Hani" maximizes to "zh_Hans_CN" (Note this will not reverse.)
+ *
+ * @param localeID The locale to maximize
+ * @param sink The output sink receiving the maximized locale
+ * @param err Error information if maximizing the locale failed.  If the length
+ * of the localeID and the null-terminator is greater than the maximum allowed size,
+ * or the localeId is not well-formed, the error code is U_ILLEGAL_ARGUMENT_ERROR.
+ * @internal ICU 64
+ */
+U_CAPI void U_EXPORT2
+ulocimp_addLikelySubtags(const char* localeID,
+                         icu::ByteSink& sink,
+                         UErrorCode* err);
+
+/**
+ * Minimize the subtags for a provided locale ID, per the algorithm described
+ * in the following CLDR technical report:
+ *
+ *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+ *
+ * If localeID is already in the minimal form, or there is no data available
+ * for minimization, it will be copied to the output buffer.  Since the
+ * minimization algorithm relies on proper maximization, see the comments
+ * for ulocimp_addLikelySubtags for reasons why there might not be any data.
+ *
+ * Examples:
+ *
+ * "en_Latn_US" minimizes to "en"
+ *
+ * "de_Latn_US" minimizes to "de"
+ *
+ * "sr_Cyrl_RS" minimizes to "sr"
+ *
+ * "zh_Hant_TW" minimizes to "zh_TW" (The region is preferred to the
+ * script, and minimizing to "zh" would imply "zh_Hans_CN".)
+ *
+ * @param localeID The locale to minimize
+ * @param sink The output sink receiving the maximized locale
+ * @param err Error information if minimizing the locale failed.  If the length
+ * of the localeID and the null-terminator is greater than the maximum allowed size,
+ * or the localeId is not well-formed, the error code is U_ILLEGAL_ARGUMENT_ERROR.
+ * @internal ICU 64
+ */
+U_CAPI void U_EXPORT2
+ulocimp_minimizeSubtags(const char* localeID,
+                        icu::ByteSink& sink,
+                        UErrorCode* err);
+
+U_CAPI const char * U_EXPORT2
+locale_getKeywordsStart(const char *localeID);
+
+U_CFUNC UBool
+ultag_isExtensionSubtags(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isLanguageSubtag(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isPrivateuseValueSubtags(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isRegionSubtag(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isScriptSubtag(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isTransformedExtensionSubtags(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isUnicodeExtensionSubtags(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleAttribute(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleAttributes(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleKey(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isUnicodeLocaleType(const char* s, int32_t len);
+
+U_CFUNC UBool
+ultag_isVariantSubtags(const char* s, int32_t len);
+
+U_CFUNC const char*
+ulocimp_toBcpKey(const char* key);
+
+U_CFUNC const char*
+ulocimp_toLegacyKey(const char* key);
+
+U_CFUNC const char*
+ulocimp_toBcpType(const char* key, const char* type, UBool* isKnownKey, UBool* isSpecialType);
+
+U_CFUNC const char*
+ulocimp_toLegacyType(const char* key, const char* type, UBool* isKnownKey, UBool* isSpecialType);
+
+/* Function for testing purpose */
+U_CAPI const char* const* ulocimp_getKnownCanonicalizedLocaleForTest(int32_t* length);
+
+// Return true if the value is already canonicalized.
+U_CAPI bool ulocimp_isCanonicalizedLocaleForTest(const char* localeName);
+
+#endif
diff --git a/thirdparty/icu4c/common/umapfile.cpp b/thirdparty/icu4c/common/umapfile.cpp
new file mode 100644
index 0000000000..3e714876a4
--- /dev/null
+++ b/thirdparty/icu4c/common/umapfile.cpp
@@ -0,0 +1,530 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+
+/*----------------------------------------------------------------------------
+ *
+ *       Memory mapped file wrappers for use by the ICU Data Implementation
+ *       All of the platform-specific implementation for mapping data files
+ *         is here.  The rest of the ICU Data implementation uses only the
+ *         wrapper functions.
+ *
+ *----------------------------------------------------------------------------*/
+/* Defines _XOPEN_SOURCE for access to POSIX functions.
+ * Must be before any other #includes. */
+#include "uposixdefs.h"
+
+#include "unicode/putil.h"
+#include "unicode/ustring.h"
+#include "udatamem.h"
+#include "umapfile.h"
+
+/* memory-mapping base definitions ------------------------------------------ */
+
+#if MAP_IMPLEMENTATION==MAP_WIN32
+#ifndef WIN32_LEAN_AND_MEAN
+#   define WIN32_LEAN_AND_MEAN
+#endif
+#   define VC_EXTRALEAN
+#   define NOUSER
+#   define NOSERVICE
+#   define NOIME
+#   define NOMCX
+
+#   if U_PLATFORM_HAS_WINUWP_API == 1
+        // Some previous versions of the Windows 10 SDK don't expose various APIs for UWP applications
+        // to use, even though UWP apps are allowed to call and use them.  Temporarily change the
+        // WINAPI family partition below to Desktop, so that function declarations are visible for UWP.
+#       include <winapifamily.h>
+#       if !(WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP | WINAPI_PARTITION_SYSTEM))
+#           pragma push_macro("WINAPI_PARTITION_DESKTOP")
+#           undef WINAPI_PARTITION_DESKTOP
+#           define WINAPI_PARTITION_DESKTOP 1
+#           define CHANGED_WINAPI_PARTITION_DESKTOP_VALUE
+#       endif
+#   endif
+
+#   include <windows.h>
+
+#   if U_PLATFORM_HAS_WINUWP_API == 1 && defined(CHANGED_WINAPI_PARTITION_DESKTOP_VALUE)
+#       pragma pop_macro("WINAPI_PARTITION_DESKTOP")
+#   endif
+
+#   include "cmemory.h"
+
+typedef HANDLE MemoryMap;
+
+#   define IS_MAP(map) ((map)!=nullptr)
+
+#elif MAP_IMPLEMENTATION==MAP_POSIX || MAP_IMPLEMENTATION==MAP_390DLL
+    typedef size_t MemoryMap;
+
+#   define IS_MAP(map) ((map)!=0)
+
+#   include <unistd.h>
+#   include <sys/mman.h>
+#   include <sys/stat.h>
+#   include <fcntl.h>
+
+#   ifndef MAP_FAILED
+#       define MAP_FAILED ((void*)-1)
+#   endif
+
+#   if MAP_IMPLEMENTATION==MAP_390DLL
+        /*   No memory mapping for 390 batch mode.  Fake it using dll loading.  */
+#       include <dll.h>
+#       include "cstring.h"
+#       include "cmemory.h"
+#       include "unicode/udata.h"
+#       define LIB_PREFIX "lib"
+#       define LIB_SUFFIX ".dll"
+        /* This is inconvenient until we figure out what to do with U_ICUDATA_NAME in utypes.h */
+#       define U_ICUDATA_ENTRY_NAME "icudt" U_ICU_VERSION_SHORT U_LIB_SUFFIX_C_NAME_STRING "_dat"
+#   endif
+#elif MAP_IMPLEMENTATION==MAP_STDIO
+#   include <stdio.h>
+#   include "cmemory.h"
+
+    typedef void *MemoryMap;
+
+#   define IS_MAP(map) ((map)!=nullptr)
+#endif
+
+/*----------------------------------------------------------------------------*
+ *                                                                            *
+ *   Memory Mapped File support.  Platform dependent implementation of        *
+ *                           functions used by the rest of the implementation.*
+ *                                                                            *
+ *----------------------------------------------------------------------------*/
+#if MAP_IMPLEMENTATION==MAP_NONE
+    U_CFUNC UBool
+    uprv_mapFile(UDataMemory *pData, const char *path, UErrorCode *status) {
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+        UDataMemory_init(pData); /* Clear the output struct. */
+        return FALSE;            /* no file access */
+    }
+
+    U_CFUNC void uprv_unmapFile(UDataMemory *pData) {
+        /* nothing to do */
+    }
+#elif MAP_IMPLEMENTATION==MAP_WIN32
+    U_CFUNC UBool
+    uprv_mapFile(
+         UDataMemory *pData,    /* Fill in with info on the result doing the mapping. */
+                                /*   Output only; any original contents are cleared.  */
+         const char *path,      /* File path to be opened/mapped.                     */
+         UErrorCode *status     /* Error status, used to report out-of-memory errors. */
+         )
+    {
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+
+        HANDLE map = nullptr;
+        HANDLE file = INVALID_HANDLE_VALUE;
+
+        UDataMemory_init(pData); /* Clear the output struct.        */
+
+        /* open the input file */
+#if U_PLATFORM_HAS_WINUWP_API == 0
+        // Note: In the non-UWP code-path (ie: Win32), the value of the path variable might have come from 
+        // the CRT 'getenv' function, and would be therefore be encoded in the default ANSI code page.
+        // This means that we can't call the *W version of API below, whereas in the UWP code-path
+        // there is no 'getenv' call, and thus the string will be only UTF-8/Invariant characters.
+        file=CreateFileA(path, GENERIC_READ, FILE_SHARE_READ, nullptr,
+            OPEN_EXISTING,
+            FILE_ATTRIBUTE_NORMAL|FILE_FLAG_RANDOM_ACCESS, nullptr);
+#else
+        // Convert from UTF-8 string to UTF-16 string.
+        wchar_t utf16Path[MAX_PATH];
+        int32_t pathUtf16Len = 0;
+        u_strFromUTF8(reinterpret_cast<UChar*>(utf16Path), static_cast<int32_t>(UPRV_LENGTHOF(utf16Path)), &pathUtf16Len, path, -1, status);
+
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+        if (*status == U_STRING_NOT_TERMINATED_WARNING) {
+            // Report back an error instead of a warning.
+            *status = U_BUFFER_OVERFLOW_ERROR;
+            return FALSE;
+        }
+
+        file = CreateFileW(utf16Path, GENERIC_READ, FILE_SHARE_READ, nullptr,
+            OPEN_EXISTING,
+            FILE_ATTRIBUTE_NORMAL | FILE_FLAG_RANDOM_ACCESS, nullptr);
+#endif
+        if (file == INVALID_HANDLE_VALUE) {
+            // If we failed to open the file due to an out-of-memory error, then we want
+            // to report that error back to the caller.
+            if (HRESULT_FROM_WIN32(GetLastError()) == E_OUTOFMEMORY) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+            }
+            return FALSE;
+        }
+
+        // Note: We use NULL/nullptr for lpAttributes parameter below.
+        // This means our handle cannot be inherited and we will get the default security descriptor.
+        /* create an unnamed Windows file-mapping object for the specified file */
+        map = CreateFileMappingW(file, nullptr, PAGE_READONLY, 0, 0, nullptr);
+
+        CloseHandle(file);
+        if (map == nullptr) {
+            // If we failed to create the mapping due to an out-of-memory error, then 
+            // we want to report that error back to the caller.
+            if (HRESULT_FROM_WIN32(GetLastError()) == E_OUTOFMEMORY) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+            }
+            return FALSE;
+        }
+
+        /* map a view of the file into our address space */
+        pData->pHeader = reinterpret_cast<const DataHeader *>(MapViewOfFile(map, FILE_MAP_READ, 0, 0, 0));
+        if (pData->pHeader == nullptr) {
+            CloseHandle(map);
+            return FALSE;
+        }
+        pData->map = map;
+        return TRUE;
+    }
+
+    U_CFUNC void
+    uprv_unmapFile(UDataMemory *pData) {
+        if (pData != nullptr && pData->map != nullptr) {
+            UnmapViewOfFile(pData->pHeader);
+            CloseHandle(pData->map);
+            pData->pHeader = nullptr;
+            pData->map = nullptr;
+        }
+    }
+
+
+
+#elif MAP_IMPLEMENTATION==MAP_POSIX
+    U_CFUNC UBool
+    uprv_mapFile(UDataMemory *pData, const char *path, UErrorCode *status) {
+        int fd;
+        int length;
+        struct stat mystat;
+        void *data;
+
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+
+        UDataMemory_init(pData); /* Clear the output struct.        */
+
+        /* determine the length of the file */
+        if(stat(path, &mystat)!=0 || mystat.st_size<=0) {
+            return FALSE;
+        }
+        length=mystat.st_size;
+
+        /* open the file */
+        fd=open(path, O_RDONLY);
+        if(fd==-1) {
+            return FALSE;
+        }
+
+        /* get a view of the mapping */
+#if U_PLATFORM != U_PF_HPUX
+        data=mmap(0, length, PROT_READ, MAP_SHARED,  fd, 0);
+#else
+        data=mmap(0, length, PROT_READ, MAP_PRIVATE, fd, 0);
+#endif
+        close(fd); /* no longer needed */
+        if(data==MAP_FAILED) {
+            // Possibly check the errno value for ENOMEM, and report U_MEMORY_ALLOCATION_ERROR?
+            return FALSE;
+        }
+
+        pData->map = (char *)data + length;
+        pData->pHeader=(const DataHeader *)data;
+        pData->mapAddr = data;
+#if U_PLATFORM == U_PF_IPHONE
+        posix_madvise(data, length, POSIX_MADV_RANDOM);
+#endif
+        return TRUE;
+    }
+
+    U_CFUNC void
+    uprv_unmapFile(UDataMemory *pData) {
+        if(pData!=nullptr && pData->map!=nullptr) {
+            size_t dataLen = (char *)pData->map - (char *)pData->mapAddr;
+            if(munmap(pData->mapAddr, dataLen)==-1) {
+            }
+            pData->pHeader=nullptr;
+            pData->map=0;
+            pData->mapAddr=nullptr;
+        }
+    }
+
+
+
+#elif MAP_IMPLEMENTATION==MAP_STDIO
+    /* copy of the filestrm.c/T_FileStream_size() implementation */
+    static int32_t
+    umap_fsize(FILE *f) {
+        int32_t savedPos = ftell(f);
+        int32_t size = 0;
+
+        /*Changes by Bertrand A. D. doesn't affect the current position
+        goes to the end of the file before ftell*/
+        fseek(f, 0, SEEK_END);
+        size = (int32_t)ftell(f);
+        fseek(f, savedPos, SEEK_SET);
+        return size;
+    }
+
+    U_CFUNC UBool
+    uprv_mapFile(UDataMemory *pData, const char *path, UErrorCode *status) {
+        FILE *file;
+        int32_t fileLength;
+        void *p;
+
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+
+        UDataMemory_init(pData); /* Clear the output struct.        */
+        /* open the input file */
+        file=fopen(path, "rb");
+        if(file==nullptr) {
+            return FALSE;
+        }
+
+        /* get the file length */
+        fileLength=umap_fsize(file);
+        if(ferror(file) || fileLength<=20) {
+            fclose(file);
+            return FALSE;
+        }
+
+        /* allocate the memory to hold the file data */
+        p=uprv_malloc(fileLength);
+        if(p==nullptr) {
+            fclose(file);
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return FALSE;
+        }
+
+        /* read the file */
+        if(fileLength!=fread(p, 1, fileLength, file)) {
+            uprv_free(p);
+            fclose(file);
+            return FALSE;
+        }
+
+        fclose(file);
+        pData->map=p;
+        pData->pHeader=(const DataHeader *)p;
+        pData->mapAddr=p;
+        return TRUE;
+    }
+
+    U_CFUNC void
+    uprv_unmapFile(UDataMemory *pData) {
+        if(pData!=nullptr && pData->map!=nullptr) {
+            uprv_free(pData->map);
+            pData->map     = nullptr;
+            pData->mapAddr = nullptr;
+            pData->pHeader = nullptr;
+        }
+    }
+
+
+#elif MAP_IMPLEMENTATION==MAP_390DLL
+    /*  390 specific Library Loading.
+     *  This is the only platform left that dynamically loads an ICU Data Library.
+     *  All other platforms use .data files when dynamic loading is required, but
+     *  this turn out to be awkward to support in 390 batch mode.
+     *
+     *  The idea here is to hide the fact that 390 is using dll loading from the
+     *   rest of ICU, and make it look like there is file loading happening.
+     *
+     */
+
+    static char *strcpy_returnEnd(char *dest, const char *src)
+    {
+        while((*dest=*src)!=0) {
+            ++dest;
+            ++src;
+        }
+        return dest;
+    }
+    
+    /*------------------------------------------------------------------------------
+     *                                                                              
+     *  computeDirPath   given a user-supplied path of an item to be opened,             
+     *                         compute and return 
+     *                            - the full directory path to be used 
+     *                              when opening the file.
+     *                            - Pointer to null at end of above returned path    
+     *
+     *                       Parameters:
+     *                          path:        input path.  Buffer is not altered.
+     *                          pathBuffer:  Output buffer.  Any contents are overwritten.
+     *
+     *                       Returns:
+     *                          Pointer to null termination in returned pathBuffer.
+     *
+     *                    TODO:  This works the way ICU historically has, but the
+     *                           whole data fallback search path is so complicated that
+     *                           probably almost no one will ever really understand it,
+     *                           the potential for confusion is large.  (It's not just 
+     *                           this one function, but the whole scheme.)
+     *                            
+     *------------------------------------------------------------------------------*/
+    static char *uprv_computeDirPath(const char *path, char *pathBuffer)
+    {
+        char   *finalSlash;       /* Ptr to last dir separator in input path, or null if none. */
+        int32_t pathLen;          /* Length of the returned directory path                     */
+        
+        finalSlash = 0;
+        if (path != 0) {
+            finalSlash = uprv_strrchr(path, U_FILE_SEP_CHAR);
+        }
+        
+        *pathBuffer = 0;
+        if (finalSlash == 0) {
+        /* No user-supplied path.  
+            * Copy the ICU_DATA path to the path buffer and return that*/
+            const char *icuDataDir;
+            icuDataDir=u_getDataDirectory();
+            if(icuDataDir!=nullptr && *icuDataDir!=0) {
+                return strcpy_returnEnd(pathBuffer, icuDataDir);
+            } else {
+                /* there is no icuDataDir either.  Just return the empty pathBuffer. */
+                return pathBuffer;
+            }
+        } 
+        
+        /* User supplied path did contain a directory portion.
+        * Copy it to the output path buffer */
+        pathLen = (int32_t)(finalSlash - path + 1);
+        uprv_memcpy(pathBuffer, path, pathLen);
+        *(pathBuffer+pathLen) = 0;
+        return pathBuffer+pathLen;
+    }
+    
+
+#   define DATA_TYPE "dat"
+
+    U_CFUNC UBool uprv_mapFile(UDataMemory *pData, const char *path, UErrorCode *status) {
+        const char *inBasename;
+        char *basename;
+        char pathBuffer[1024];
+        const DataHeader *pHeader;
+        dllhandle *handle;
+        void *val=0;
+
+        if (U_FAILURE(*status)) {
+            return FALSE;
+        }
+
+        inBasename=uprv_strrchr(path, U_FILE_SEP_CHAR);
+        if(inBasename==nullptr) {
+            inBasename = path;
+        } else {
+            inBasename++;
+        }
+        basename=uprv_computeDirPath(path, pathBuffer);
+        if(uprv_strcmp(inBasename, U_ICUDATA_NAME".dat") != 0) {
+            /* must mmap file... for build */
+            int fd;
+            int length;
+            struct stat mystat;
+            void *data;
+            UDataMemory_init(pData); /* Clear the output struct. */
+
+            /* determine the length of the file */
+            if(stat(path, &mystat)!=0 || mystat.st_size<=0) {
+                return FALSE;
+            }
+            length=mystat.st_size;
+
+            /* open the file */
+            fd=open(path, O_RDONLY);
+            if(fd==-1) {
+                return FALSE;
+            }
+
+            /* get a view of the mapping */
+            data=mmap(0, length, PROT_READ, MAP_PRIVATE, fd, 0);
+            close(fd); /* no longer needed */
+            if(data==MAP_FAILED) {
+                // Possibly check the errorno value for ENOMEM, and report U_MEMORY_ALLOCATION_ERROR?
+                return FALSE;
+            }
+            pData->map = (char *)data + length;
+            pData->pHeader=(const DataHeader *)data;
+            pData->mapAddr = data;
+            return TRUE;
+        }
+
+#       ifdef OS390BATCH
+            /* ### hack: we still need to get u_getDataDirectory() fixed
+            for OS/390 (batch mode - always return "//"? )
+            and this here straightened out with LIB_PREFIX and LIB_SUFFIX (both empty?!)
+            This is probably due to the strange file system on OS/390.  It's more like
+            a database with short entry names than a typical file system. */
+            /* U_ICUDATA_NAME should always have the correct name */
+            /* BUT FOR BATCH MODE IT IS AN EXCEPTION BECAUSE */
+            /* THE FIRST THREE LETTERS ARE PREASSIGNED TO THE */
+            /* PROJECT!!!!! */
+            uprv_strcpy(pathBuffer, "IXMI" U_ICU_VERSION_SHORT "DA");
+#       else
+            /* set up the library name */
+            uprv_strcpy(basename, LIB_PREFIX U_LIBICUDATA_NAME U_ICU_VERSION_SHORT LIB_SUFFIX);
+#       endif
+
+#       ifdef UDATA_DEBUG
+             fprintf(stderr, "dllload: %s ", pathBuffer);
+#       endif
+
+        handle=dllload(pathBuffer);
+
+#       ifdef UDATA_DEBUG
+               fprintf(stderr, " -> %08X\n", handle );
+#       endif
+
+        if(handle != nullptr) {
+               /* we have a data DLL - what kind of lookup do we need here? */
+               /* try to find the Table of Contents */
+               UDataMemory_init(pData); /* Clear the output struct.        */
+               val=dllqueryvar((dllhandle*)handle, U_ICUDATA_ENTRY_NAME);
+               if(val == 0) {
+                    /* failed... so keep looking */
+                    return FALSE;
+               }
+#              ifdef UDATA_DEBUG
+                    fprintf(stderr, "dllqueryvar(%08X, %s) -> %08X\n", handle, U_ICUDATA_ENTRY_NAME, val);
+#              endif
+
+               pData->pHeader=(const DataHeader *)val;
+               return TRUE;
+         } else {
+               return FALSE; /* no handle */
+         }
+    }
+
+    U_CFUNC void uprv_unmapFile(UDataMemory *pData) {
+        if(pData!=nullptr && pData->map!=nullptr) {
+            uprv_free(pData->map);
+            pData->map     = nullptr;
+            pData->mapAddr = nullptr;
+            pData->pHeader = nullptr;
+        }   
+    }
+
+#else
+#   error MAP_IMPLEMENTATION is set incorrectly
+#endif
diff --git a/thirdparty/icu4c/common/umapfile.h b/thirdparty/icu4c/common/umapfile.h
new file mode 100644
index 0000000000..92bd567a2a
--- /dev/null
+++ b/thirdparty/icu4c/common/umapfile.h
@@ -0,0 +1,57 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************/
+
+/*----------------------------------------------------------------------------------
+ *
+ *       Memory mapped file wrappers for use by the ICU Data Implementation
+ *
+ *           Porting note:  The implementation of these functions is very platform specific.
+ *             Not all platforms can do real memory mapping.  Those that can't
+ *             still must implement these functions, getting the data into memory using
+ *             whatever means are available.
+ *
+ *            These functions are part of the ICU internal implementation, and
+ *            are not inteded to be used directly by applications.
+ *
+ *----------------------------------------------------------------------------------*/
+
+#ifndef __UMAPFILE_H__
+#define __UMAPFILE_H__
+
+#include "unicode/putil.h"
+#include "unicode/udata.h"
+#include "putilimp.h"
+
+U_CFUNC UBool uprv_mapFile(UDataMemory *pdm, const char *path, UErrorCode *status);
+U_CFUNC void  uprv_unmapFile(UDataMemory *pData);
+
+/* MAP_NONE: no memory mapping, no file access at all */
+#define MAP_NONE        0
+#define MAP_WIN32       1
+#define MAP_POSIX       2
+#define MAP_STDIO       3
+#define MAP_390DLL      4
+
+#if UCONFIG_NO_FILE_IO
+#   define MAP_IMPLEMENTATION MAP_NONE
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define MAP_IMPLEMENTATION MAP_WIN32
+#elif U_HAVE_MMAP || U_PLATFORM == U_PF_OS390
+#   if U_PLATFORM == U_PF_OS390 && defined (OS390_STUBDATA)
+        /*   No memory mapping for 390 batch mode.  Fake it using dll loading.  */
+#       define MAP_IMPLEMENTATION MAP_390DLL
+#   else
+#       define MAP_IMPLEMENTATION MAP_POSIX
+#   endif
+#else /* unknown platform, no memory map implementation: use stdio.h and uprv_malloc() instead */
+#   define MAP_IMPLEMENTATION MAP_STDIO
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/umath.cpp b/thirdparty/icu4c/common/umath.cpp
new file mode 100644
index 0000000000..7cf4b31749
--- /dev/null
+++ b/thirdparty/icu4c/common/umath.cpp
@@ -0,0 +1,26 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+* This file contains platform independent math.
+*/
+
+#include "putilimp.h"
+
+U_CAPI int32_t U_EXPORT2
+uprv_max(int32_t x, int32_t y)
+{
+    return (x > y ? x : y);
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_min(int32_t x, int32_t y)
+{
+    return (x > y ? y : x);
+}
+
diff --git a/thirdparty/icu4c/common/umutablecptrie.cpp b/thirdparty/icu4c/common/umutablecptrie.cpp
new file mode 100644
index 0000000000..cdbe27080b
--- /dev/null
+++ b/thirdparty/icu4c/common/umutablecptrie.cpp
@@ -0,0 +1,1852 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// umutablecptrie.cpp (inspired by utrie2_builder.cpp)
+// created: 2017dec29 Markus W. Scherer
+
+// #define UCPTRIE_DEBUG
+#ifdef UCPTRIE_DEBUG
+#   include <stdio.h>
+#endif
+
+#include "unicode/utypes.h"
+#include "unicode/ucptrie.h"
+#include "unicode/umutablecptrie.h"
+#include "unicode/uobject.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "ucptrie_impl.h"
+
+// ICU-20235 In case Microsoft math.h has defined this, undefine it.
+#ifdef OVERFLOW
+#undef OVERFLOW
+#endif
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+constexpr int32_t MAX_UNICODE = 0x10ffff;
+
+constexpr int32_t UNICODE_LIMIT = 0x110000;
+constexpr int32_t BMP_LIMIT = 0x10000;
+constexpr int32_t ASCII_LIMIT = 0x80;
+
+constexpr int32_t I_LIMIT = UNICODE_LIMIT >> UCPTRIE_SHIFT_3;
+constexpr int32_t BMP_I_LIMIT = BMP_LIMIT >> UCPTRIE_SHIFT_3;
+constexpr int32_t ASCII_I_LIMIT = ASCII_LIMIT >> UCPTRIE_SHIFT_3;
+
+constexpr int32_t SMALL_DATA_BLOCKS_PER_BMP_BLOCK = (1 << (UCPTRIE_FAST_SHIFT - UCPTRIE_SHIFT_3));
+
+// Flag values for data blocks.
+constexpr uint8_t ALL_SAME = 0;
+constexpr uint8_t MIXED = 1;
+constexpr uint8_t SAME_AS = 2;
+
+/** Start with allocation of 16k data entries. */
+constexpr int32_t INITIAL_DATA_LENGTH = ((int32_t)1 << 14);
+
+/** Grow about 8x each time. */
+constexpr int32_t MEDIUM_DATA_LENGTH = ((int32_t)1 << 17);
+
+/**
+ * Maximum length of the build-time data array.
+ * One entry per 0x110000 code points.
+ */
+constexpr int32_t MAX_DATA_LENGTH = UNICODE_LIMIT;
+
+// Flag values for index-3 blocks while compacting/building.
+constexpr uint8_t I3_NULL = 0;
+constexpr uint8_t I3_BMP = 1;
+constexpr uint8_t I3_16 = 2;
+constexpr uint8_t I3_18 = 3;
+
+constexpr int32_t INDEX_3_18BIT_BLOCK_LENGTH = UCPTRIE_INDEX_3_BLOCK_LENGTH + UCPTRIE_INDEX_3_BLOCK_LENGTH / 8;
+
+class AllSameBlocks;
+class MixedBlocks;
+
+class MutableCodePointTrie : public UMemory {
+public:
+    MutableCodePointTrie(uint32_t initialValue, uint32_t errorValue, UErrorCode &errorCode);
+    MutableCodePointTrie(const MutableCodePointTrie &other, UErrorCode &errorCode);
+    MutableCodePointTrie(const MutableCodePointTrie &other) = delete;
+    ~MutableCodePointTrie();
+
+    MutableCodePointTrie &operator=(const MutableCodePointTrie &other) = delete;
+
+    static MutableCodePointTrie *fromUCPMap(const UCPMap *map, UErrorCode &errorCode);
+    static MutableCodePointTrie *fromUCPTrie(const UCPTrie *trie, UErrorCode &errorCode);
+
+    uint32_t get(UChar32 c) const;
+    int32_t getRange(UChar32 start, UCPMapValueFilter *filter, const void *context,
+                     uint32_t *pValue) const;
+
+    void set(UChar32 c, uint32_t value, UErrorCode &errorCode);
+    void setRange(UChar32 start, UChar32 end, uint32_t value, UErrorCode &errorCode);
+
+    UCPTrie *build(UCPTrieType type, UCPTrieValueWidth valueWidth, UErrorCode &errorCode);
+
+private:
+    void clear();
+
+    bool ensureHighStart(UChar32 c);
+    int32_t allocDataBlock(int32_t blockLength);
+    int32_t getDataBlock(int32_t i);
+
+    void maskValues(uint32_t mask);
+    UChar32 findHighStart() const;
+    int32_t compactWholeDataBlocks(int32_t fastILimit, AllSameBlocks &allSameBlocks);
+    int32_t compactData(
+            int32_t fastILimit, uint32_t *newData, int32_t newDataCapacity,
+            int32_t dataNullIndex, MixedBlocks &mixedBlocks, UErrorCode &errorCode);
+    int32_t compactIndex(int32_t fastILimit, MixedBlocks &mixedBlocks, UErrorCode &errorCode);
+    int32_t compactTrie(int32_t fastILimit, UErrorCode &errorCode);
+
+    uint32_t *index = nullptr;
+    int32_t indexCapacity = 0;
+    int32_t index3NullOffset = -1;
+    uint32_t *data = nullptr;
+    int32_t dataCapacity = 0;
+    int32_t dataLength = 0;
+    int32_t dataNullOffset = -1;
+
+    uint32_t origInitialValue;
+    uint32_t initialValue;
+    uint32_t errorValue;
+    UChar32 highStart;
+    uint32_t highValue;
+#ifdef UCPTRIE_DEBUG
+public:
+    const char *name;
+#endif
+private:
+    /** Temporary array while building the final data. */
+    uint16_t *index16 = nullptr;
+    uint8_t flags[UNICODE_LIMIT >> UCPTRIE_SHIFT_3];
+};
+
+MutableCodePointTrie::MutableCodePointTrie(uint32_t iniValue, uint32_t errValue, UErrorCode &errorCode) :
+        origInitialValue(iniValue), initialValue(iniValue), errorValue(errValue),
+        highStart(0), highValue(initialValue)
+#ifdef UCPTRIE_DEBUG
+        , name("open")
+#endif
+        {
+    if (U_FAILURE(errorCode)) { return; }
+    index = (uint32_t *)uprv_malloc(BMP_I_LIMIT * 4);
+    data = (uint32_t *)uprv_malloc(INITIAL_DATA_LENGTH * 4);
+    if (index == nullptr || data == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    indexCapacity = BMP_I_LIMIT;
+    dataCapacity = INITIAL_DATA_LENGTH;
+}
+
+MutableCodePointTrie::MutableCodePointTrie(const MutableCodePointTrie &other, UErrorCode &errorCode) :
+        index3NullOffset(other.index3NullOffset),
+        dataNullOffset(other.dataNullOffset),
+        origInitialValue(other.origInitialValue), initialValue(other.initialValue),
+        errorValue(other.errorValue),
+        highStart(other.highStart), highValue(other.highValue)
+#ifdef UCPTRIE_DEBUG
+        , name("mutable clone")
+#endif
+        {
+    if (U_FAILURE(errorCode)) { return; }
+    int32_t iCapacity = highStart <= BMP_LIMIT ? BMP_I_LIMIT : I_LIMIT;
+    index = (uint32_t *)uprv_malloc(iCapacity * 4);
+    data = (uint32_t *)uprv_malloc(other.dataCapacity * 4);
+    if (index == nullptr || data == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    indexCapacity = iCapacity;
+    dataCapacity = other.dataCapacity;
+
+    int32_t iLimit = highStart >> UCPTRIE_SHIFT_3;
+    uprv_memcpy(flags, other.flags, iLimit);
+    uprv_memcpy(index, other.index, iLimit * 4);
+    uprv_memcpy(data, other.data, (size_t)other.dataLength * 4);
+    dataLength = other.dataLength;
+    U_ASSERT(other.index16 == nullptr);
+}
+
+MutableCodePointTrie::~MutableCodePointTrie() {
+    uprv_free(index);
+    uprv_free(data);
+    uprv_free(index16);
+}
+
+MutableCodePointTrie *MutableCodePointTrie::fromUCPMap(const UCPMap *map, UErrorCode &errorCode) {
+    // Use the highValue as the initialValue to reduce the highStart.
+    uint32_t errorValue = ucpmap_get(map, -1);
+    uint32_t initialValue = ucpmap_get(map, 0x10ffff);
+    LocalPointer<MutableCodePointTrie> mutableTrie(
+        new MutableCodePointTrie(initialValue, errorValue, errorCode),
+        errorCode);
+    if (U_FAILURE(errorCode)) {
+        return nullptr;
+    }
+    UChar32 start = 0, end;
+    uint32_t value;
+    while ((end = ucpmap_getRange(map, start, UCPMAP_RANGE_NORMAL, 0,
+                                  nullptr, nullptr, &value)) >= 0) {
+        if (value != initialValue) {
+            if (start == end) {
+                mutableTrie->set(start, value, errorCode);
+            } else {
+                mutableTrie->setRange(start, end, value, errorCode);
+            }
+        }
+        start = end + 1;
+    }
+    if (U_SUCCESS(errorCode)) {
+        return mutableTrie.orphan();
+    } else {
+        return nullptr;
+    }
+}
+
+MutableCodePointTrie *MutableCodePointTrie::fromUCPTrie(const UCPTrie *trie, UErrorCode &errorCode) {
+    // Use the highValue as the initialValue to reduce the highStart.
+    uint32_t errorValue;
+    uint32_t initialValue;
+    switch (trie->valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        errorValue = trie->data.ptr16[trie->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET];
+        initialValue = trie->data.ptr16[trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET];
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        errorValue = trie->data.ptr32[trie->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET];
+        initialValue = trie->data.ptr32[trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET];
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        errorValue = trie->data.ptr8[trie->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET];
+        initialValue = trie->data.ptr8[trie->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET];
+        break;
+    default:
+        // Unreachable if the trie is properly initialized.
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    LocalPointer<MutableCodePointTrie> mutableTrie(
+        new MutableCodePointTrie(initialValue, errorValue, errorCode),
+        errorCode);
+    if (U_FAILURE(errorCode)) {
+        return nullptr;
+    }
+    UChar32 start = 0, end;
+    uint32_t value;
+    while ((end = ucptrie_getRange(trie, start, UCPMAP_RANGE_NORMAL, 0,
+                                   nullptr, nullptr, &value)) >= 0) {
+        if (value != initialValue) {
+            if (start == end) {
+                mutableTrie->set(start, value, errorCode);
+            } else {
+                mutableTrie->setRange(start, end, value, errorCode);
+            }
+        }
+        start = end + 1;
+    }
+    if (U_SUCCESS(errorCode)) {
+        return mutableTrie.orphan();
+    } else {
+        return nullptr;
+    }
+}
+
+void MutableCodePointTrie::clear() {
+    index3NullOffset = dataNullOffset = -1;
+    dataLength = 0;
+    highValue = initialValue = origInitialValue;
+    highStart = 0;
+    uprv_free(index16);
+    index16 = nullptr;
+}
+
+uint32_t MutableCodePointTrie::get(UChar32 c) const {
+    if ((uint32_t)c > MAX_UNICODE) {
+        return errorValue;
+    }
+    if (c >= highStart) {
+        return highValue;
+    }
+    int32_t i = c >> UCPTRIE_SHIFT_3;
+    if (flags[i] == ALL_SAME) {
+        return index[i];
+    } else {
+        return data[index[i] + (c & UCPTRIE_SMALL_DATA_MASK)];
+    }
+}
+
+inline uint32_t maybeFilterValue(uint32_t value, uint32_t initialValue, uint32_t nullValue,
+                                 UCPMapValueFilter *filter, const void *context) {
+    if (value == initialValue) {
+        value = nullValue;
+    } else if (filter != nullptr) {
+        value = filter(context, value);
+    }
+    return value;
+}
+
+UChar32 MutableCodePointTrie::getRange(
+        UChar32 start, UCPMapValueFilter *filter, const void *context,
+        uint32_t *pValue) const {
+    if ((uint32_t)start > MAX_UNICODE) {
+        return U_SENTINEL;
+    }
+    if (start >= highStart) {
+        if (pValue != nullptr) {
+            uint32_t value = highValue;
+            if (filter != nullptr) { value = filter(context, value); }
+            *pValue = value;
+        }
+        return MAX_UNICODE;
+    }
+    uint32_t nullValue = initialValue;
+    if (filter != nullptr) { nullValue = filter(context, nullValue); }
+    UChar32 c = start;
+    uint32_t trieValue, value;
+    bool haveValue = false;
+    int32_t i = c >> UCPTRIE_SHIFT_3;
+    do {
+        if (flags[i] == ALL_SAME) {
+            uint32_t trieValue2 = index[i];
+            if (haveValue) {
+                if (trieValue2 != trieValue) {
+                    if (filter == nullptr ||
+                            maybeFilterValue(trieValue2, initialValue, nullValue,
+                                             filter, context) != value) {
+                        return c - 1;
+                    }
+                    trieValue = trieValue2;  // may or may not help
+                }
+            } else {
+                trieValue = trieValue2;
+                value = maybeFilterValue(trieValue2, initialValue, nullValue, filter, context);
+                if (pValue != nullptr) { *pValue = value; }
+                haveValue = true;
+            }
+            c = (c + UCPTRIE_SMALL_DATA_BLOCK_LENGTH) & ~UCPTRIE_SMALL_DATA_MASK;
+        } else /* MIXED */ {
+            int32_t di = index[i] + (c & UCPTRIE_SMALL_DATA_MASK);
+            uint32_t trieValue2 = data[di];
+            if (haveValue) {
+                if (trieValue2 != trieValue) {
+                    if (filter == nullptr ||
+                            maybeFilterValue(trieValue2, initialValue, nullValue,
+                                             filter, context) != value) {
+                        return c - 1;
+                    }
+                    trieValue = trieValue2;  // may or may not help
+                }
+            } else {
+                trieValue = trieValue2;
+                value = maybeFilterValue(trieValue2, initialValue, nullValue, filter, context);
+                if (pValue != nullptr) { *pValue = value; }
+                haveValue = true;
+            }
+            while ((++c & UCPTRIE_SMALL_DATA_MASK) != 0) {
+                trieValue2 = data[++di];
+                if (trieValue2 != trieValue) {
+                    if (filter == nullptr ||
+                            maybeFilterValue(trieValue2, initialValue, nullValue,
+                                             filter, context) != value) {
+                        return c - 1;
+                    }
+                }
+                trieValue = trieValue2;  // may or may not help
+            }
+        }
+        ++i;
+    } while (c < highStart);
+    U_ASSERT(haveValue);
+    if (maybeFilterValue(highValue, initialValue, nullValue,
+                         filter, context) != value) {
+        return c - 1;
+    } else {
+        return MAX_UNICODE;
+    }
+}
+
+void
+writeBlock(uint32_t *block, uint32_t value) {
+    uint32_t *limit = block + UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+    while (block < limit) {
+        *block++ = value;
+    }
+}
+
+bool MutableCodePointTrie::ensureHighStart(UChar32 c) {
+    if (c >= highStart) {
+        // Round up to a UCPTRIE_CP_PER_INDEX_2_ENTRY boundary to simplify compaction.
+        c = (c + UCPTRIE_CP_PER_INDEX_2_ENTRY) & ~(UCPTRIE_CP_PER_INDEX_2_ENTRY - 1);
+        int32_t i = highStart >> UCPTRIE_SHIFT_3;
+        int32_t iLimit = c >> UCPTRIE_SHIFT_3;
+        if (iLimit > indexCapacity) {
+            uint32_t *newIndex = (uint32_t *)uprv_malloc(I_LIMIT * 4);
+            if (newIndex == nullptr) { return false; }
+            uprv_memcpy(newIndex, index, i * 4);
+            uprv_free(index);
+            index = newIndex;
+            indexCapacity = I_LIMIT;
+        }
+        do {
+            flags[i] = ALL_SAME;
+            index[i] = initialValue;
+        } while(++i < iLimit);
+        highStart = c;
+    }
+    return true;
+}
+
+int32_t MutableCodePointTrie::allocDataBlock(int32_t blockLength) {
+    int32_t newBlock = dataLength;
+    int32_t newTop = newBlock + blockLength;
+    if (newTop > dataCapacity) {
+        int32_t capacity;
+        if (dataCapacity < MEDIUM_DATA_LENGTH) {
+            capacity = MEDIUM_DATA_LENGTH;
+        } else if (dataCapacity < MAX_DATA_LENGTH) {
+            capacity = MAX_DATA_LENGTH;
+        } else {
+            // Should never occur.
+            // Either MAX_DATA_LENGTH is incorrect,
+            // or the code writes more values than should be possible.
+            return -1;
+        }
+        uint32_t *newData = (uint32_t *)uprv_malloc(capacity * 4);
+        if (newData == nullptr) {
+            return -1;
+        }
+        uprv_memcpy(newData, data, (size_t)dataLength * 4);
+        uprv_free(data);
+        data = newData;
+        dataCapacity = capacity;
+    }
+    dataLength = newTop;
+    return newBlock;
+}
+
+/**
+ * No error checking for illegal arguments.
+ *
+ * @return -1 if no new data block available (out of memory in data array)
+ * @internal
+ */
+int32_t MutableCodePointTrie::getDataBlock(int32_t i) {
+    if (flags[i] == MIXED) {
+        return index[i];
+    }
+    if (i < BMP_I_LIMIT) {
+        int32_t newBlock = allocDataBlock(UCPTRIE_FAST_DATA_BLOCK_LENGTH);
+        if (newBlock < 0) { return newBlock; }
+        int32_t iStart = i & ~(SMALL_DATA_BLOCKS_PER_BMP_BLOCK -1);
+        int32_t iLimit = iStart + SMALL_DATA_BLOCKS_PER_BMP_BLOCK;
+        do {
+            U_ASSERT(flags[iStart] == ALL_SAME);
+            writeBlock(data + newBlock, index[iStart]);
+            flags[iStart] = MIXED;
+            index[iStart++] = newBlock;
+            newBlock += UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+        } while (iStart < iLimit);
+        return index[i];
+    } else {
+        int32_t newBlock = allocDataBlock(UCPTRIE_SMALL_DATA_BLOCK_LENGTH);
+        if (newBlock < 0) { return newBlock; }
+        writeBlock(data + newBlock, index[i]);
+        flags[i] = MIXED;
+        index[i] = newBlock;
+        return newBlock;
+    }
+}
+
+void MutableCodePointTrie::set(UChar32 c, uint32_t value, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if ((uint32_t)c > MAX_UNICODE) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    int32_t block;
+    if (!ensureHighStart(c) || (block = getDataBlock(c >> UCPTRIE_SHIFT_3)) < 0) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    data[block + (c & UCPTRIE_SMALL_DATA_MASK)] = value;
+}
+
+void
+fillBlock(uint32_t *block, UChar32 start, UChar32 limit, uint32_t value) {
+    uint32_t *pLimit = block + limit;
+    block += start;
+    while (block < pLimit) {
+        *block++ = value;
+    }
+}
+
+void MutableCodePointTrie::setRange(UChar32 start, UChar32 end, uint32_t value, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return;
+    }
+    if ((uint32_t)start > MAX_UNICODE || (uint32_t)end > MAX_UNICODE || start > end) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if (!ensureHighStart(end)) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    UChar32 limit = end + 1;
+    if (start & UCPTRIE_SMALL_DATA_MASK) {
+        // Set partial block at [start..following block boundary[.
+        int32_t block = getDataBlock(start >> UCPTRIE_SHIFT_3);
+        if (block < 0) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+
+        UChar32 nextStart = (start + UCPTRIE_SMALL_DATA_MASK) & ~UCPTRIE_SMALL_DATA_MASK;
+        if (nextStart <= limit) {
+            fillBlock(data + block, start & UCPTRIE_SMALL_DATA_MASK, UCPTRIE_SMALL_DATA_BLOCK_LENGTH,
+                      value);
+            start = nextStart;
+        } else {
+            fillBlock(data + block, start & UCPTRIE_SMALL_DATA_MASK, limit & UCPTRIE_SMALL_DATA_MASK,
+                      value);
+            return;
+        }
+    }
+
+    // Number of positions in the last, partial block.
+    int32_t rest = limit & UCPTRIE_SMALL_DATA_MASK;
+
+    // Round down limit to a block boundary.
+    limit &= ~UCPTRIE_SMALL_DATA_MASK;
+
+    // Iterate over all-value blocks.
+    while (start < limit) {
+        int32_t i = start >> UCPTRIE_SHIFT_3;
+        if (flags[i] == ALL_SAME) {
+            index[i] = value;
+        } else /* MIXED */ {
+            fillBlock(data + index[i], 0, UCPTRIE_SMALL_DATA_BLOCK_LENGTH, value);
+        }
+        start += UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+    }
+
+    if (rest > 0) {
+        // Set partial block at [last block boundary..limit[.
+        int32_t block = getDataBlock(start >> UCPTRIE_SHIFT_3);
+        if (block < 0) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+
+        fillBlock(data + block, 0, rest, value);
+    }
+}
+
+/* compaction --------------------------------------------------------------- */
+
+void MutableCodePointTrie::maskValues(uint32_t mask) {
+    initialValue &= mask;
+    errorValue &= mask;
+    highValue &= mask;
+    int32_t iLimit = highStart >> UCPTRIE_SHIFT_3;
+    for (int32_t i = 0; i < iLimit; ++i) {
+        if (flags[i] == ALL_SAME) {
+            index[i] &= mask;
+        }
+    }
+    for (int32_t i = 0; i < dataLength; ++i) {
+        data[i] &= mask;
+    }
+}
+
+template<typename UIntA, typename UIntB>
+bool equalBlocks(const UIntA *s, const UIntB *t, int32_t length) {
+    while (length > 0 && *s == *t) {
+        ++s;
+        ++t;
+        --length;
+    }
+    return length == 0;
+}
+
+bool allValuesSameAs(const uint32_t *p, int32_t length, uint32_t value) {
+    const uint32_t *pLimit = p + length;
+    while (p < pLimit && *p == value) { ++p; }
+    return p == pLimit;
+}
+
+/** Search for an identical block. */
+int32_t findSameBlock(const uint16_t *p, int32_t pStart, int32_t length,
+                      const uint16_t *q, int32_t qStart, int32_t blockLength) {
+    // Ensure that we do not even partially get past length.
+    length -= blockLength;
+
+    q += qStart;
+    while (pStart <= length) {
+        if (equalBlocks(p + pStart, q, blockLength)) {
+            return pStart;
+        }
+        ++pStart;
+    }
+    return -1;
+}
+
+int32_t findAllSameBlock(const uint32_t *p, int32_t start, int32_t limit,
+                         uint32_t value, int32_t blockLength) {
+    // Ensure that we do not even partially get past limit.
+    limit -= blockLength;
+
+    for (int32_t block = start; block <= limit; ++block) {
+        if (p[block] == value) {
+            for (int32_t i = 1;; ++i) {
+                if (i == blockLength) {
+                    return block;
+                }
+                if (p[block + i] != value) {
+                    block += i;
+                    break;
+                }
+            }
+        }
+    }
+    return -1;
+}
+
+/**
+ * Look for maximum overlap of the beginning of the other block
+ * with the previous, adjacent block.
+ */
+template<typename UIntA, typename UIntB>
+int32_t getOverlap(const UIntA *p, int32_t length,
+                   const UIntB *q, int32_t qStart, int32_t blockLength) {
+    int32_t overlap = blockLength - 1;
+    U_ASSERT(overlap <= length);
+    q += qStart;
+    while (overlap > 0 && !equalBlocks(p + (length - overlap), q, overlap)) {
+        --overlap;
+    }
+    return overlap;
+}
+
+int32_t getAllSameOverlap(const uint32_t *p, int32_t length, uint32_t value,
+                          int32_t blockLength) {
+    int32_t min = length - (blockLength - 1);
+    int32_t i = length;
+    while (min < i && p[i - 1] == value) { --i; }
+    return length - i;
+}
+
+bool isStartOfSomeFastBlock(uint32_t dataOffset, const uint32_t index[], int32_t fastILimit) {
+    for (int32_t i = 0; i < fastILimit; i += SMALL_DATA_BLOCKS_PER_BMP_BLOCK) {
+        if (index[i] == dataOffset) {
+            return true;
+        }
+    }
+    return false;
+}
+
+/**
+ * Finds the start of the last range in the trie by enumerating backward.
+ * Indexes for code points higher than this will be omitted.
+ */
+UChar32 MutableCodePointTrie::findHighStart() const {
+    int32_t i = highStart >> UCPTRIE_SHIFT_3;
+    while (i > 0) {
+        bool match;
+        if (flags[--i] == ALL_SAME) {
+            match = index[i] == highValue;
+        } else /* MIXED */ {
+            const uint32_t *p = data + index[i];
+            for (int32_t j = 0;; ++j) {
+                if (j == UCPTRIE_SMALL_DATA_BLOCK_LENGTH) {
+                    match = true;
+                    break;
+                }
+                if (p[j] != highValue) {
+                    match = false;
+                    break;
+                }
+            }
+        }
+        if (!match) {
+            return (i + 1) << UCPTRIE_SHIFT_3;
+        }
+    }
+    return 0;
+}
+
+class AllSameBlocks {
+public:
+    static constexpr int32_t NEW_UNIQUE = -1;
+    static constexpr int32_t OVERFLOW = -2;
+
+    AllSameBlocks() : length(0), mostRecent(-1) {}
+
+    int32_t findOrAdd(int32_t index, int32_t count, uint32_t value) {
+        if (mostRecent >= 0 && values[mostRecent] == value) {
+            refCounts[mostRecent] += count;
+            return indexes[mostRecent];
+        }
+        for (int32_t i = 0; i < length; ++i) {
+            if (values[i] == value) {
+                mostRecent = i;
+                refCounts[i] += count;
+                return indexes[i];
+            }
+        }
+        if (length == CAPACITY) {
+            return OVERFLOW;
+        }
+        mostRecent = length;
+        indexes[length] = index;
+        values[length] = value;
+        refCounts[length++] = count;
+        return NEW_UNIQUE;
+    }
+
+    /** Replaces the block which has the lowest reference count. */
+    void add(int32_t index, int32_t count, uint32_t value) {
+        U_ASSERT(length == CAPACITY);
+        int32_t least = -1;
+        int32_t leastCount = I_LIMIT;
+        for (int32_t i = 0; i < length; ++i) {
+            U_ASSERT(values[i] != value);
+            if (refCounts[i] < leastCount) {
+                least = i;
+                leastCount = refCounts[i];
+            }
+        }
+        U_ASSERT(least >= 0);
+        mostRecent = least;
+        indexes[least] = index;
+        values[least] = value;
+        refCounts[least] = count;
+    }
+
+    int32_t findMostUsed() const {
+        if (length == 0) { return -1; }
+        int32_t max = -1;
+        int32_t maxCount = 0;
+        for (int32_t i = 0; i < length; ++i) {
+            if (refCounts[i] > maxCount) {
+                max = i;
+                maxCount = refCounts[i];
+            }
+        }
+        return indexes[max];
+    }
+
+private:
+    static constexpr int32_t CAPACITY = 32;
+
+    int32_t length;
+    int32_t mostRecent;
+
+    int32_t indexes[CAPACITY];
+    uint32_t values[CAPACITY];
+    int32_t refCounts[CAPACITY];
+};
+
+// Custom hash table for mixed-value blocks to be found anywhere in the
+// compacted data or index so far.
+class MixedBlocks {
+public:
+    MixedBlocks() {}
+    ~MixedBlocks() {
+        uprv_free(table);
+    }
+
+    bool init(int32_t maxLength, int32_t newBlockLength) {
+        // We store actual data indexes + 1 to reserve 0 for empty entries.
+        int32_t maxDataIndex = maxLength - newBlockLength + 1;
+        int32_t newLength;
+        if (maxDataIndex <= 0xfff) {  // 4k
+            newLength = 6007;
+            shift = 12;
+            mask = 0xfff;
+        } else if (maxDataIndex <= 0x7fff) {  // 32k
+            newLength = 50021;
+            shift = 15;
+            mask = 0x7fff;
+        } else if (maxDataIndex <= 0x1ffff) {  // 128k
+            newLength = 200003;
+            shift = 17;
+            mask = 0x1ffff;
+        } else {
+            // maxDataIndex up to around MAX_DATA_LENGTH, ca. 1.1M
+            newLength = 1500007;
+            shift = 21;
+            mask = 0x1fffff;
+        }
+        if (newLength > capacity) {
+            uprv_free(table);
+            table = (uint32_t *)uprv_malloc(newLength * 4);
+            if (table == nullptr) {
+                return false;
+            }
+            capacity = newLength;
+        }
+        length = newLength;
+        uprv_memset(table, 0, length * 4);
+
+        blockLength = newBlockLength;
+        return true;
+    }
+
+    template<typename UInt>
+    void extend(const UInt *data, int32_t minStart, int32_t prevDataLength, int32_t newDataLength) {
+        int32_t start = prevDataLength - blockLength;
+        if (start >= minStart) {
+            ++start;  // Skip the last block that we added last time.
+        } else {
+            start = minStart;  // Begin with the first full block.
+        }
+        for (int32_t end = newDataLength - blockLength; start <= end; ++start) {
+            uint32_t hashCode = makeHashCode(data, start);
+            addEntry(data, start, hashCode, start);
+        }
+    }
+
+    template<typename UIntA, typename UIntB>
+    int32_t findBlock(const UIntA *data, const UIntB *blockData, int32_t blockStart) const {
+        uint32_t hashCode = makeHashCode(blockData, blockStart);
+        int32_t entryIndex = findEntry(data, blockData, blockStart, hashCode);
+        if (entryIndex >= 0) {
+            return (table[entryIndex] & mask) - 1;
+        } else {
+            return -1;
+        }
+    }
+
+    int32_t findAllSameBlock(const uint32_t *data, uint32_t blockValue) const {
+        uint32_t hashCode = makeHashCode(blockValue);
+        int32_t entryIndex = findEntry(data, blockValue, hashCode);
+        if (entryIndex >= 0) {
+            return (table[entryIndex] & mask) - 1;
+        } else {
+            return -1;
+        }
+    }
+
+private:
+    template<typename UInt>
+    uint32_t makeHashCode(const UInt *blockData, int32_t blockStart) const {
+        int32_t blockLimit = blockStart + blockLength;
+        uint32_t hashCode = blockData[blockStart++];
+        do {
+            hashCode = 37 * hashCode + blockData[blockStart++];
+        } while (blockStart < blockLimit);
+        return hashCode;
+    }
+
+    uint32_t makeHashCode(uint32_t blockValue) const {
+        uint32_t hashCode = blockValue;
+        for (int32_t i = 1; i < blockLength; ++i) {
+            hashCode = 37 * hashCode + blockValue;
+        }
+        return hashCode;
+    }
+
+    template<typename UInt>
+    void addEntry(const UInt *data, int32_t blockStart, uint32_t hashCode, int32_t dataIndex) {
+        U_ASSERT(0 <= dataIndex && dataIndex < (int32_t)mask);
+        int32_t entryIndex = findEntry(data, data, blockStart, hashCode);
+        if (entryIndex < 0) {
+            table[~entryIndex] = (hashCode << shift) | (dataIndex + 1);
+        }
+    }
+
+    template<typename UIntA, typename UIntB>
+    int32_t findEntry(const UIntA *data, const UIntB *blockData, int32_t blockStart,
+                      uint32_t hashCode) const {
+        uint32_t shiftedHashCode = hashCode << shift;
+        int32_t initialEntryIndex = (hashCode % (length - 1)) + 1;  // 1..length-1
+        for (int32_t entryIndex = initialEntryIndex;;) {
+            uint32_t entry = table[entryIndex];
+            if (entry == 0) {
+                return ~entryIndex;
+            }
+            if ((entry & ~mask) == shiftedHashCode) {
+                int32_t dataIndex = (entry & mask) - 1;
+                if (equalBlocks(data + dataIndex, blockData + blockStart, blockLength)) {
+                    return entryIndex;
+                }
+            }
+            entryIndex = nextIndex(initialEntryIndex, entryIndex);
+        }
+    }
+
+    int32_t findEntry(const uint32_t *data, uint32_t blockValue, uint32_t hashCode) const {
+        uint32_t shiftedHashCode = hashCode << shift;
+        int32_t initialEntryIndex = (hashCode % (length - 1)) + 1;  // 1..length-1
+        for (int32_t entryIndex = initialEntryIndex;;) {
+            uint32_t entry = table[entryIndex];
+            if (entry == 0) {
+                return ~entryIndex;
+            }
+            if ((entry & ~mask) == shiftedHashCode) {
+                int32_t dataIndex = (entry & mask) - 1;
+                if (allValuesSameAs(data + dataIndex, blockLength, blockValue)) {
+                    return entryIndex;
+                }
+            }
+            entryIndex = nextIndex(initialEntryIndex, entryIndex);
+        }
+    }
+
+    inline int32_t nextIndex(int32_t initialEntryIndex, int32_t entryIndex) const {
+        // U_ASSERT(0 < initialEntryIndex && initialEntryIndex < length);
+        return (entryIndex + initialEntryIndex) % length;
+    }
+
+    // Hash table.
+    // The length is a prime number, larger than the maximum data length.
+    // The "shift" lower bits store a data index + 1.
+    // The remaining upper bits store a partial hashCode of the block data values.
+    uint32_t *table = nullptr;
+    int32_t capacity = 0;
+    int32_t length = 0;
+    int32_t shift = 0;
+    uint32_t mask = 0;
+
+    int32_t blockLength = 0;
+};
+
+int32_t MutableCodePointTrie::compactWholeDataBlocks(int32_t fastILimit, AllSameBlocks &allSameBlocks) {
+#ifdef UCPTRIE_DEBUG
+    bool overflow = false;
+#endif
+
+    // ASCII data will be stored as a linear table, even if the following code
+    // does not yet count it that way.
+    int32_t newDataCapacity = ASCII_LIMIT;
+    // Add room for a small data null block in case it would match the start of
+    // a fast data block where dataNullOffset must not be set in that case.
+    newDataCapacity += UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+    // Add room for special values (errorValue, highValue) and padding.
+    newDataCapacity += 4;
+    int32_t iLimit = highStart >> UCPTRIE_SHIFT_3;
+    int32_t blockLength = UCPTRIE_FAST_DATA_BLOCK_LENGTH;
+    int32_t inc = SMALL_DATA_BLOCKS_PER_BMP_BLOCK;
+    for (int32_t i = 0; i < iLimit; i += inc) {
+        if (i == fastILimit) {
+            blockLength = UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+            inc = 1;
+        }
+        uint32_t value = index[i];
+        if (flags[i] == MIXED) {
+            // Really mixed?
+            const uint32_t *p = data + value;
+            value = *p;
+            if (allValuesSameAs(p + 1, blockLength - 1, value)) {
+                flags[i] = ALL_SAME;
+                index[i] = value;
+                // Fall through to ALL_SAME handling.
+            } else {
+                newDataCapacity += blockLength;
+                continue;
+            }
+        } else {
+            U_ASSERT(flags[i] == ALL_SAME);
+            if (inc > 1) {
+                // Do all of the fast-range data block's ALL_SAME parts have the same value?
+                bool allSame = true;
+                int32_t next_i = i + inc;
+                for (int32_t j = i + 1; j < next_i; ++j) {
+                    U_ASSERT(flags[j] == ALL_SAME);
+                    if (index[j] != value) {
+                        allSame = false;
+                        break;
+                    }
+                }
+                if (!allSame) {
+                    // Turn it into a MIXED block.
+                    if (getDataBlock(i) < 0) {
+                        return -1;
+                    }
+                    newDataCapacity += blockLength;
+                    continue;
+                }
+            }
+        }
+        // Is there another ALL_SAME block with the same value?
+        int32_t other = allSameBlocks.findOrAdd(i, inc, value);
+        if (other == AllSameBlocks::OVERFLOW) {
+            // The fixed-size array overflowed. Slow check for a duplicate block.
+#ifdef UCPTRIE_DEBUG
+            if (!overflow) {
+                puts("UCPTrie AllSameBlocks overflow");
+                overflow = true;
+            }
+#endif
+            int32_t jInc = SMALL_DATA_BLOCKS_PER_BMP_BLOCK;
+            for (int32_t j = 0;; j += jInc) {
+                if (j == i) {
+                    allSameBlocks.add(i, inc, value);
+                    break;
+                }
+                if (j == fastILimit) {
+                    jInc = 1;
+                }
+                if (flags[j] == ALL_SAME && index[j] == value) {
+                    allSameBlocks.add(j, jInc + inc, value);
+                    other = j;
+                    break;
+                    // We could keep counting blocks with the same value
+                    // before we add the first one, which may improve compaction in rare cases,
+                    // but it would make it slower.
+                }
+            }
+        }
+        if (other >= 0) {
+            flags[i] = SAME_AS;
+            index[i] = other;
+        } else {
+            // New unique same-value block.
+            newDataCapacity += blockLength;
+        }
+    }
+    return newDataCapacity;
+}
+
+#ifdef UCPTRIE_DEBUG
+#   define DEBUG_DO(expr) expr
+#else
+#   define DEBUG_DO(expr)
+#endif
+
+#ifdef UCPTRIE_DEBUG
+// Braille symbols: U+28xx = UTF-8 E2 A0 80..E2 A3 BF
+int32_t appendValue(char s[], int32_t length, uint32_t value) {
+    value ^= value >> 16;
+    value ^= value >> 8;
+    s[length] = 0xE2;
+    s[length + 1] = (char)(0xA0 + ((value >> 6) & 3));
+    s[length + 2] = (char)(0x80 + (value & 0x3F));
+    return length + 3;
+}
+
+void printBlock(const uint32_t *block, int32_t blockLength, uint32_t value,
+                UChar32 start, int32_t overlap, uint32_t initialValue) {
+    char s[UCPTRIE_FAST_DATA_BLOCK_LENGTH * 3 + 3];
+    int32_t length = 0;
+    int32_t i;
+    for (i = 0; i < overlap; ++i) {
+        length = appendValue(s, length, 0);  // Braille blank
+    }
+    s[length++] = '|';
+    for (; i < blockLength; ++i) {
+        if (block != nullptr) {
+            value = block[i];
+        }
+        if (value == initialValue) {
+            value = 0x40;  // Braille lower left dot
+        }
+        length = appendValue(s, length, value);
+    }
+    s[length] = 0;
+    start += overlap;
+    if (start <= 0xffff) {
+        printf("    %04lX  %s|\n", (long)start, s);
+    } else if (start <= 0xfffff) {
+        printf("   %5lX  %s|\n", (long)start, s);
+    } else {
+        printf("  %6lX  %s|\n", (long)start, s);
+    }
+}
+#endif
+
+/**
+ * Compacts a build-time trie.
+ *
+ * The compaction
+ * - removes blocks that are identical with earlier ones
+ * - overlaps each new non-duplicate block as much as possible with the previously-written one
+ * - works with fast-range data blocks whose length is a multiple of that of
+ *   higher-code-point data blocks
+ *
+ * It does not try to find an optimal order of writing, deduplicating, and overlapping blocks.
+ */
+int32_t MutableCodePointTrie::compactData(
+        int32_t fastILimit, uint32_t *newData, int32_t newDataCapacity,
+        int32_t dataNullIndex, MixedBlocks &mixedBlocks, UErrorCode &errorCode) {
+#ifdef UCPTRIE_DEBUG
+    int32_t countSame=0, sumOverlaps=0;
+    bool printData = dataLength == 29088 /* line.brk */ ||
+        // dataLength == 30048 /* CanonIterData */ ||
+        dataLength == 50400 /* zh.txt~stroke */;
+#endif
+
+    // The linear ASCII data has been copied into newData already.
+    int32_t newDataLength = 0;
+    for (int32_t i = 0; newDataLength < ASCII_LIMIT;
+            newDataLength += UCPTRIE_FAST_DATA_BLOCK_LENGTH, i += SMALL_DATA_BLOCKS_PER_BMP_BLOCK) {
+        index[i] = newDataLength;
+#ifdef UCPTRIE_DEBUG
+        if (printData) {
+            printBlock(newData + newDataLength, UCPTRIE_FAST_DATA_BLOCK_LENGTH, 0, newDataLength, 0, initialValue);
+        }
+#endif
+    }
+
+    int32_t blockLength = UCPTRIE_FAST_DATA_BLOCK_LENGTH;
+    if (!mixedBlocks.init(newDataCapacity, blockLength)) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    mixedBlocks.extend(newData, 0, 0, newDataLength);
+
+    int32_t iLimit = highStart >> UCPTRIE_SHIFT_3;
+    int32_t inc = SMALL_DATA_BLOCKS_PER_BMP_BLOCK;
+    int32_t fastLength = 0;
+    for (int32_t i = ASCII_I_LIMIT; i < iLimit; i += inc) {
+        if (i == fastILimit) {
+            blockLength = UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+            inc = 1;
+            fastLength = newDataLength;
+            if (!mixedBlocks.init(newDataCapacity, blockLength)) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+                return 0;
+            }
+            mixedBlocks.extend(newData, 0, 0, newDataLength);
+        }
+        if (flags[i] == ALL_SAME) {
+            uint32_t value = index[i];
+            // Find an earlier part of the data array of length blockLength
+            // that is filled with this value.
+            int32_t n = mixedBlocks.findAllSameBlock(newData, value);
+            // If we find a match, and the current block is the data null block,
+            // and it is not a fast block but matches the start of a fast block,
+            // then we need to continue looking.
+            // This is because this small block is shorter than the fast block,
+            // and not all of the rest of the fast block is filled with this value.
+            // Otherwise trie.getRange() would detect that the fast block starts at
+            // dataNullOffset and assume incorrectly that it is filled with the null value.
+            while (n >= 0 && i == dataNullIndex && i >= fastILimit && n < fastLength &&
+                    isStartOfSomeFastBlock(n, index, fastILimit)) {
+                n = findAllSameBlock(newData, n + 1, newDataLength, value, blockLength);
+            }
+            if (n >= 0) {
+                DEBUG_DO(++countSame);
+                index[i] = n;
+            } else {
+                n = getAllSameOverlap(newData, newDataLength, value, blockLength);
+                DEBUG_DO(sumOverlaps += n);
+#ifdef UCPTRIE_DEBUG
+                if (printData) {
+                    printBlock(nullptr, blockLength, value, i << UCPTRIE_SHIFT_3, n, initialValue);
+                }
+#endif
+                index[i] = newDataLength - n;
+                int32_t prevDataLength = newDataLength;
+                while (n < blockLength) {
+                    newData[newDataLength++] = value;
+                    ++n;
+                }
+                mixedBlocks.extend(newData, 0, prevDataLength, newDataLength);
+            }
+        } else if (flags[i] == MIXED) {
+            const uint32_t *block = data + index[i];
+            int32_t n = mixedBlocks.findBlock(newData, block, 0);
+            if (n >= 0) {
+                DEBUG_DO(++countSame);
+                index[i] = n;
+            } else {
+                n = getOverlap(newData, newDataLength, block, 0, blockLength);
+                DEBUG_DO(sumOverlaps += n);
+#ifdef UCPTRIE_DEBUG
+                if (printData) {
+                    printBlock(block, blockLength, 0, i << UCPTRIE_SHIFT_3, n, initialValue);
+                }
+#endif
+                index[i] = newDataLength - n;
+                int32_t prevDataLength = newDataLength;
+                while (n < blockLength) {
+                    newData[newDataLength++] = block[n++];
+                }
+                mixedBlocks.extend(newData, 0, prevDataLength, newDataLength);
+            }
+        } else /* SAME_AS */ {
+            uint32_t j = index[i];
+            index[i] = index[j];
+        }
+    }
+
+#ifdef UCPTRIE_DEBUG
+    /* we saved some space */
+    printf("compacting UCPTrie: count of 32-bit data words %lu->%lu  countSame=%ld  sumOverlaps=%ld\n",
+            (long)dataLength, (long)newDataLength, (long)countSame, (long)sumOverlaps);
+#endif
+    return newDataLength;
+}
+
+int32_t MutableCodePointTrie::compactIndex(int32_t fastILimit, MixedBlocks &mixedBlocks,
+                                           UErrorCode &errorCode) {
+    int32_t fastIndexLength = fastILimit >> (UCPTRIE_FAST_SHIFT - UCPTRIE_SHIFT_3);
+    if ((highStart >> UCPTRIE_FAST_SHIFT) <= fastIndexLength) {
+        // Only the linear fast index, no multi-stage index tables.
+        index3NullOffset = UCPTRIE_NO_INDEX3_NULL_OFFSET;
+        return fastIndexLength;
+    }
+
+    // Condense the fast index table.
+    // Also, does it contain an index-3 block with all dataNullOffset?
+    uint16_t fastIndex[UCPTRIE_BMP_INDEX_LENGTH];  // fastIndexLength
+    int32_t i3FirstNull = -1;
+    for (int32_t i = 0, j = 0; i < fastILimit; ++j) {
+        uint32_t i3 = index[i];
+        fastIndex[j] = (uint16_t)i3;
+        if (i3 == (uint32_t)dataNullOffset) {
+            if (i3FirstNull < 0) {
+                i3FirstNull = j;
+            } else if (index3NullOffset < 0 &&
+                    (j - i3FirstNull + 1) == UCPTRIE_INDEX_3_BLOCK_LENGTH) {
+                index3NullOffset = i3FirstNull;
+            }
+        } else {
+            i3FirstNull = -1;
+        }
+        // Set the index entries that compactData() skipped.
+        // Needed when the multi-stage index covers the fast index range as well.
+        int32_t iNext = i + SMALL_DATA_BLOCKS_PER_BMP_BLOCK;
+        while (++i < iNext) {
+            i3 += UCPTRIE_SMALL_DATA_BLOCK_LENGTH;
+            index[i] = i3;
+        }
+    }
+
+    if (!mixedBlocks.init(fastIndexLength, UCPTRIE_INDEX_3_BLOCK_LENGTH)) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    mixedBlocks.extend(fastIndex, 0, 0, fastIndexLength);
+
+    // Examine index-3 blocks. For each determine one of:
+    // - same as the index-3 null block
+    // - same as a fast-index block
+    // - 16-bit indexes
+    // - 18-bit indexes
+    // We store this in the first flags entry for the index-3 block.
+    //
+    // Also determine an upper limit for the index-3 table length.
+    int32_t index3Capacity = 0;
+    i3FirstNull = index3NullOffset;
+    bool hasLongI3Blocks = false;
+    // If the fast index covers the whole BMP, then
+    // the multi-stage index is only for supplementary code points.
+    // Otherwise, the multi-stage index covers all of Unicode.
+    int32_t iStart = fastILimit < BMP_I_LIMIT ? 0 : BMP_I_LIMIT;
+    int32_t iLimit = highStart >> UCPTRIE_SHIFT_3;
+    for (int32_t i = iStart; i < iLimit;) {
+        int32_t j = i;
+        int32_t jLimit = i + UCPTRIE_INDEX_3_BLOCK_LENGTH;
+        uint32_t oredI3 = 0;
+        bool isNull = true;
+        do {
+            uint32_t i3 = index[j];
+            oredI3 |= i3;
+            if (i3 != (uint32_t)dataNullOffset) {
+                isNull = false;
+            }
+        } while (++j < jLimit);
+        if (isNull) {
+            flags[i] = I3_NULL;
+            if (i3FirstNull < 0) {
+                if (oredI3 <= 0xffff) {
+                    index3Capacity += UCPTRIE_INDEX_3_BLOCK_LENGTH;
+                } else {
+                    index3Capacity += INDEX_3_18BIT_BLOCK_LENGTH;
+                    hasLongI3Blocks = true;
+                }
+                i3FirstNull = 0;
+            }
+        } else {
+            if (oredI3 <= 0xffff) {
+                int32_t n = mixedBlocks.findBlock(fastIndex, index, i);
+                if (n >= 0) {
+                    flags[i] = I3_BMP;
+                    index[i] = n;
+                } else {
+                    flags[i] = I3_16;
+                    index3Capacity += UCPTRIE_INDEX_3_BLOCK_LENGTH;
+                }
+            } else {
+                flags[i] = I3_18;
+                index3Capacity += INDEX_3_18BIT_BLOCK_LENGTH;
+                hasLongI3Blocks = true;
+            }
+        }
+        i = j;
+    }
+
+    int32_t index2Capacity = (iLimit - iStart) >> UCPTRIE_SHIFT_2_3;
+
+    // Length of the index-1 table, rounded up.
+    int32_t index1Length = (index2Capacity + UCPTRIE_INDEX_2_MASK) >> UCPTRIE_SHIFT_1_2;
+
+    // Index table: Fast index, index-1, index-3, index-2.
+    // +1 for possible index table padding.
+    int32_t index16Capacity = fastIndexLength + index1Length + index3Capacity + index2Capacity + 1;
+    index16 = (uint16_t *)uprv_malloc(index16Capacity * 2);
+    if (index16 == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    uprv_memcpy(index16, fastIndex, fastIndexLength * 2);
+
+    if (!mixedBlocks.init(index16Capacity, UCPTRIE_INDEX_3_BLOCK_LENGTH)) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    MixedBlocks longI3Blocks;
+    if (hasLongI3Blocks) {
+        if (!longI3Blocks.init(index16Capacity, INDEX_3_18BIT_BLOCK_LENGTH)) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+            return 0;
+        }
+    }
+
+    // Compact the index-3 table and write an uncompacted version of the index-2 table.
+    uint16_t index2[UNICODE_LIMIT >> UCPTRIE_SHIFT_2];  // index2Capacity
+    int32_t i2Length = 0;
+    i3FirstNull = index3NullOffset;
+    int32_t index3Start = fastIndexLength + index1Length;
+    int32_t indexLength = index3Start;
+    for (int32_t i = iStart; i < iLimit; i += UCPTRIE_INDEX_3_BLOCK_LENGTH) {
+        int32_t i3;
+        uint8_t f = flags[i];
+        if (f == I3_NULL && i3FirstNull < 0) {
+            // First index-3 null block. Write & overlap it like a normal block, then remember it.
+            f = dataNullOffset <= 0xffff ? I3_16 : I3_18;
+            i3FirstNull = 0;
+        }
+        if (f == I3_NULL) {
+            i3 = index3NullOffset;
+        } else if (f == I3_BMP) {
+            i3 = index[i];
+        } else if (f == I3_16) {
+            int32_t n = mixedBlocks.findBlock(index16, index, i);
+            if (n >= 0) {
+                i3 = n;
+            } else {
+                if (indexLength == index3Start) {
+                    // No overlap at the boundary between the index-1 and index-3 tables.
+                    n = 0;
+                } else {
+                    n = getOverlap(index16, indexLength,
+                                   index, i, UCPTRIE_INDEX_3_BLOCK_LENGTH);
+                }
+                i3 = indexLength - n;
+                int32_t prevIndexLength = indexLength;
+                while (n < UCPTRIE_INDEX_3_BLOCK_LENGTH) {
+                    index16[indexLength++] = index[i + n++];
+                }
+                mixedBlocks.extend(index16, index3Start, prevIndexLength, indexLength);
+                if (hasLongI3Blocks) {
+                    longI3Blocks.extend(index16, index3Start, prevIndexLength, indexLength);
+                }
+            }
+        } else {
+            U_ASSERT(f == I3_18);
+            U_ASSERT(hasLongI3Blocks);
+            // Encode an index-3 block that contains one or more data indexes exceeding 16 bits.
+            int32_t j = i;
+            int32_t jLimit = i + UCPTRIE_INDEX_3_BLOCK_LENGTH;
+            int32_t k = indexLength;
+            do {
+                ++k;
+                uint32_t v = index[j++];
+                uint32_t upperBits = (v & 0x30000) >> 2;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 4;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 6;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 8;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 10;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 12;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 14;
+                index16[k++] = v;
+                v = index[j++];
+                upperBits |= (v & 0x30000) >> 16;
+                index16[k++] = v;
+                index16[k - 9] = upperBits;
+            } while (j < jLimit);
+            int32_t n = longI3Blocks.findBlock(index16, index16, indexLength);
+            if (n >= 0) {
+                i3 = n | 0x8000;
+            } else {
+                if (indexLength == index3Start) {
+                    // No overlap at the boundary between the index-1 and index-3 tables.
+                    n = 0;
+                } else {
+                    n = getOverlap(index16, indexLength,
+                                   index16, indexLength, INDEX_3_18BIT_BLOCK_LENGTH);
+                }
+                i3 = (indexLength - n) | 0x8000;
+                int32_t prevIndexLength = indexLength;
+                if (n > 0) {
+                    int32_t start = indexLength;
+                    while (n < INDEX_3_18BIT_BLOCK_LENGTH) {
+                        index16[indexLength++] = index16[start + n++];
+                    }
+                } else {
+                    indexLength += INDEX_3_18BIT_BLOCK_LENGTH;
+                }
+                mixedBlocks.extend(index16, index3Start, prevIndexLength, indexLength);
+                if (hasLongI3Blocks) {
+                    longI3Blocks.extend(index16, index3Start, prevIndexLength, indexLength);
+                }
+            }
+        }
+        if (index3NullOffset < 0 && i3FirstNull >= 0) {
+            index3NullOffset = i3;
+        }
+        // Set the index-2 table entry.
+        index2[i2Length++] = i3;
+    }
+    U_ASSERT(i2Length == index2Capacity);
+    U_ASSERT(indexLength <= index3Start + index3Capacity);
+
+    if (index3NullOffset < 0) {
+        index3NullOffset = UCPTRIE_NO_INDEX3_NULL_OFFSET;
+    }
+    if (indexLength >= (UCPTRIE_NO_INDEX3_NULL_OFFSET + UCPTRIE_INDEX_3_BLOCK_LENGTH)) {
+        // The index-3 offsets exceed 15 bits, or
+        // the last one cannot be distinguished from the no-null-block value.
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    // Compact the index-2 table and write the index-1 table.
+    static_assert(UCPTRIE_INDEX_2_BLOCK_LENGTH == UCPTRIE_INDEX_3_BLOCK_LENGTH,
+                  "must re-init mixedBlocks");
+    int32_t blockLength = UCPTRIE_INDEX_2_BLOCK_LENGTH;
+    int32_t i1 = fastIndexLength;
+    for (int32_t i = 0; i < i2Length; i += blockLength) {
+        int32_t n;
+        if ((i2Length - i) >= blockLength) {
+            // normal block
+            U_ASSERT(blockLength == UCPTRIE_INDEX_2_BLOCK_LENGTH);
+            n = mixedBlocks.findBlock(index16, index2, i);
+        } else {
+            // highStart is inside the last index-2 block. Shorten it.
+            blockLength = i2Length - i;
+            n = findSameBlock(index16, index3Start, indexLength,
+                              index2, i, blockLength);
+        }
+        int32_t i2;
+        if (n >= 0) {
+            i2 = n;
+        } else {
+            if (indexLength == index3Start) {
+                // No overlap at the boundary between the index-1 and index-3/2 tables.
+                n = 0;
+            } else {
+                n = getOverlap(index16, indexLength, index2, i, blockLength);
+            }
+            i2 = indexLength - n;
+            int32_t prevIndexLength = indexLength;
+            while (n < blockLength) {
+                index16[indexLength++] = index2[i + n++];
+            }
+            mixedBlocks.extend(index16, index3Start, prevIndexLength, indexLength);
+        }
+        // Set the index-1 table entry.
+        index16[i1++] = i2;
+    }
+    U_ASSERT(i1 == index3Start);
+    U_ASSERT(indexLength <= index16Capacity);
+
+#ifdef UCPTRIE_DEBUG
+    /* we saved some space */
+    printf("compacting UCPTrie: count of 16-bit index words %lu->%lu\n",
+            (long)iLimit, (long)indexLength);
+#endif
+
+    return indexLength;
+}
+
+int32_t MutableCodePointTrie::compactTrie(int32_t fastILimit, UErrorCode &errorCode) {
+    // Find the real highStart and round it up.
+    U_ASSERT((highStart & (UCPTRIE_CP_PER_INDEX_2_ENTRY - 1)) == 0);
+    highValue = get(MAX_UNICODE);
+    int32_t realHighStart = findHighStart();
+    realHighStart = (realHighStart + (UCPTRIE_CP_PER_INDEX_2_ENTRY - 1)) &
+        ~(UCPTRIE_CP_PER_INDEX_2_ENTRY - 1);
+    if (realHighStart == UNICODE_LIMIT) {
+        highValue = initialValue;
+    }
+
+#ifdef UCPTRIE_DEBUG
+    printf("UCPTrie: highStart U+%06lx  highValue 0x%lx  initialValue 0x%lx\n",
+            (long)realHighStart, (long)highValue, (long)initialValue);
+#endif
+
+    // We always store indexes and data values for the fast range.
+    // Pin highStart to the top of that range while building.
+    UChar32 fastLimit = fastILimit << UCPTRIE_SHIFT_3;
+    if (realHighStart < fastLimit) {
+        for (int32_t i = (realHighStart >> UCPTRIE_SHIFT_3); i < fastILimit; ++i) {
+            flags[i] = ALL_SAME;
+            index[i] = highValue;
+        }
+        highStart = fastLimit;
+    } else {
+        highStart = realHighStart;
+    }
+
+    uint32_t asciiData[ASCII_LIMIT];
+    for (int32_t i = 0; i < ASCII_LIMIT; ++i) {
+        asciiData[i] = get(i);
+    }
+
+    // First we look for which data blocks have the same value repeated over the whole block,
+    // deduplicate such blocks, find a good null data block (for faster enumeration),
+    // and get an upper bound for the necessary data array length.
+    AllSameBlocks allSameBlocks;
+    int32_t newDataCapacity = compactWholeDataBlocks(fastILimit, allSameBlocks);
+    if (newDataCapacity < 0) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    uint32_t *newData = (uint32_t *)uprv_malloc(newDataCapacity * 4);
+    if (newData == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    uprv_memcpy(newData, asciiData, sizeof(asciiData));
+
+    int32_t dataNullIndex = allSameBlocks.findMostUsed();
+
+    MixedBlocks mixedBlocks;
+    int32_t newDataLength = compactData(fastILimit, newData, newDataCapacity,
+                                        dataNullIndex, mixedBlocks, errorCode);
+    if (U_FAILURE(errorCode)) { return 0; }
+    U_ASSERT(newDataLength <= newDataCapacity);
+    uprv_free(data);
+    data = newData;
+    dataCapacity = newDataCapacity;
+    dataLength = newDataLength;
+    if (dataLength > (0x3ffff + UCPTRIE_SMALL_DATA_BLOCK_LENGTH)) {
+        // The offset of the last data block is too high to be stored in the index table.
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    if (dataNullIndex >= 0) {
+        dataNullOffset = index[dataNullIndex];
+#ifdef UCPTRIE_DEBUG
+        if (data[dataNullOffset] != initialValue) {
+            printf("UCPTrie initialValue %lx -> more common nullValue %lx\n",
+                   (long)initialValue, (long)data[dataNullOffset]);
+        }
+#endif
+        initialValue = data[dataNullOffset];
+    } else {
+        dataNullOffset = UCPTRIE_NO_DATA_NULL_OFFSET;
+    }
+
+    int32_t indexLength = compactIndex(fastILimit, mixedBlocks, errorCode);
+    highStart = realHighStart;
+    return indexLength;
+}
+
+UCPTrie *MutableCodePointTrie::build(UCPTrieType type, UCPTrieValueWidth valueWidth, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) {
+        return nullptr;
+    }
+    if (type < UCPTRIE_TYPE_FAST || UCPTRIE_TYPE_SMALL < type ||
+            valueWidth < UCPTRIE_VALUE_BITS_16 || UCPTRIE_VALUE_BITS_8 < valueWidth) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+
+    // The mutable trie always stores 32-bit values.
+    // When we build a UCPTrie for a smaller value width, we first mask off unused bits
+    // before compacting the data.
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_32:
+        break;
+    case UCPTRIE_VALUE_BITS_16:
+        maskValues(0xffff);
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        maskValues(0xff);
+        break;
+    default:
+        break;
+    }
+
+    UChar32 fastLimit = type == UCPTRIE_TYPE_FAST ? BMP_LIMIT : UCPTRIE_SMALL_LIMIT;
+    int32_t indexLength = compactTrie(fastLimit >> UCPTRIE_SHIFT_3, errorCode);
+    if (U_FAILURE(errorCode)) {
+        clear();
+        return nullptr;
+    }
+
+    // Ensure data table alignment: The index length must be even for uint32_t data.
+    if (valueWidth == UCPTRIE_VALUE_BITS_32 && (indexLength & 1) != 0) {
+        index16[indexLength++] = 0xffee;  // arbitrary value
+    }
+
+    // Make the total trie structure length a multiple of 4 bytes by padding the data table,
+    // and store special values as the last two data values.
+    int32_t length = indexLength * 2;
+    if (valueWidth == UCPTRIE_VALUE_BITS_16) {
+        if (((indexLength ^ dataLength) & 1) != 0) {
+            // padding
+            data[dataLength++] = errorValue;
+        }
+        if (data[dataLength - 1] != errorValue || data[dataLength - 2] != highValue) {
+            data[dataLength++] = highValue;
+            data[dataLength++] = errorValue;
+        }
+        length += dataLength * 2;
+    } else if (valueWidth == UCPTRIE_VALUE_BITS_32) {
+        // 32-bit data words never need padding to a multiple of 4 bytes.
+        if (data[dataLength - 1] != errorValue || data[dataLength - 2] != highValue) {
+            if (data[dataLength - 1] != highValue) {
+                data[dataLength++] = highValue;
+            }
+            data[dataLength++] = errorValue;
+        }
+        length += dataLength * 4;
+    } else {
+        int32_t and3 = (length + dataLength) & 3;
+        if (and3 == 0 && data[dataLength - 1] == errorValue && data[dataLength - 2] == highValue) {
+            // all set
+        } else if(and3 == 3 && data[dataLength - 1] == highValue) {
+            data[dataLength++] = errorValue;
+        } else {
+            while (and3 != 2) {
+                data[dataLength++] = highValue;
+                and3 = (and3 + 1) & 3;
+            }
+            data[dataLength++] = highValue;
+            data[dataLength++] = errorValue;
+        }
+        length += dataLength;
+    }
+
+    // Calculate the total length of the UCPTrie as a single memory block.
+    length += sizeof(UCPTrie);
+    U_ASSERT((length & 3) == 0);
+
+    uint8_t *bytes = (uint8_t *)uprv_malloc(length);
+    if (bytes == nullptr) {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+        clear();
+        return nullptr;
+    }
+    UCPTrie *trie = reinterpret_cast<UCPTrie *>(bytes);
+    uprv_memset(trie, 0, sizeof(UCPTrie));
+    trie->indexLength = indexLength;
+    trie->dataLength = dataLength;
+
+    trie->highStart = highStart;
+    // Round up shifted12HighStart to a multiple of 0x1000 for easy testing from UTF-8 lead bytes.
+    // Runtime code needs to then test for the real highStart as well.
+    trie->shifted12HighStart = (highStart + 0xfff) >> 12;
+    trie->type = type;
+    trie->valueWidth = valueWidth;
+
+    trie->index3NullOffset = index3NullOffset;
+    trie->dataNullOffset = dataNullOffset;
+    trie->nullValue = initialValue;
+
+    bytes += sizeof(UCPTrie);
+
+    // Fill the index and data arrays.
+    uint16_t *dest16 = (uint16_t *)bytes;
+    trie->index = dest16;
+
+    if (highStart <= fastLimit) {
+        // Condense only the fast index from the mutable-trie index.
+        for (int32_t i = 0, j = 0; j < indexLength; i += SMALL_DATA_BLOCKS_PER_BMP_BLOCK, ++j) {
+            *dest16++ = (uint16_t)index[i];  // dest16[j]
+        }
+    } else {
+        uprv_memcpy(dest16, index16, indexLength * 2);
+        dest16 += indexLength;
+    }
+    bytes += indexLength * 2;
+
+    // Write the data array.
+    const uint32_t *p = data;
+    switch (valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        // Write 16-bit data values.
+        trie->data.ptr16 = dest16;
+        for (int32_t i = dataLength; i > 0; --i) {
+            *dest16++ = (uint16_t)*p++;
+        }
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        // Write 32-bit data values.
+        trie->data.ptr32 = (uint32_t *)bytes;
+        uprv_memcpy(bytes, p, (size_t)dataLength * 4);
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        // Write 8-bit data values.
+        trie->data.ptr8 = bytes;
+        for (int32_t i = dataLength; i > 0; --i) {
+            *bytes++ = (uint8_t)*p++;
+        }
+        break;
+    default:
+        // Will not occur, valueWidth checked at the beginning.
+        break;
+    }
+
+#ifdef UCPTRIE_DEBUG
+    trie->name = name;
+
+    ucptrie_printLengths(trie, "");
+#endif
+
+    clear();
+    return trie;
+}
+
+}  // namespace
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_open(uint32_t initialValue, uint32_t errorValue, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    LocalPointer<MutableCodePointTrie> trie(
+        new MutableCodePointTrie(initialValue, errorValue, *pErrorCode), *pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    return reinterpret_cast<UMutableCPTrie *>(trie.orphan());
+}
+
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_clone(const UMutableCPTrie *other, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    if (other == nullptr) {
+        return nullptr;
+    }
+    LocalPointer<MutableCodePointTrie> clone(
+        new MutableCodePointTrie(*reinterpret_cast<const MutableCodePointTrie *>(other), *pErrorCode), *pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    return reinterpret_cast<UMutableCPTrie *>(clone.orphan());
+}
+
+U_CAPI void U_EXPORT2
+umutablecptrie_close(UMutableCPTrie *trie) {
+    delete reinterpret_cast<MutableCodePointTrie *>(trie);
+}
+
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_fromUCPMap(const UCPMap *map, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    if (map == nullptr) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    return reinterpret_cast<UMutableCPTrie *>(MutableCodePointTrie::fromUCPMap(map, *pErrorCode));
+}
+
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_fromUCPTrie(const UCPTrie *trie, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    if (trie == nullptr) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    return reinterpret_cast<UMutableCPTrie *>(MutableCodePointTrie::fromUCPTrie(trie, *pErrorCode));
+}
+
+U_CAPI uint32_t U_EXPORT2
+umutablecptrie_get(const UMutableCPTrie *trie, UChar32 c) {
+    return reinterpret_cast<const MutableCodePointTrie *>(trie)->get(c);
+}
+
+namespace {
+
+UChar32 getRange(const void *trie, UChar32 start,
+                 UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    return reinterpret_cast<const MutableCodePointTrie *>(trie)->
+        getRange(start, filter, context, pValue);
+}
+
+}  // namespace
+
+U_CAPI UChar32 U_EXPORT2
+umutablecptrie_getRange(const UMutableCPTrie *trie, UChar32 start,
+                        UCPMapRangeOption option, uint32_t surrogateValue,
+                        UCPMapValueFilter *filter, const void *context, uint32_t *pValue) {
+    return ucptrie_internalGetRange(getRange, trie, start,
+                                    option, surrogateValue,
+                                    filter, context, pValue);
+}
+
+U_CAPI void U_EXPORT2
+umutablecptrie_set(UMutableCPTrie *trie, UChar32 c, uint32_t value, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    reinterpret_cast<MutableCodePointTrie *>(trie)->set(c, value, *pErrorCode);
+}
+
+U_CAPI void U_EXPORT2
+umutablecptrie_setRange(UMutableCPTrie *trie, UChar32 start, UChar32 end,
+                   uint32_t value, UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    reinterpret_cast<MutableCodePointTrie *>(trie)->setRange(start, end, value, *pErrorCode);
+}
+
+/* Compact and internally serialize the trie. */
+U_CAPI UCPTrie * U_EXPORT2
+umutablecptrie_buildImmutable(UMutableCPTrie *trie, UCPTrieType type, UCPTrieValueWidth valueWidth,
+                              UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return nullptr;
+    }
+    return reinterpret_cast<MutableCodePointTrie *>(trie)->build(type, valueWidth, *pErrorCode);
+}
+
+#ifdef UCPTRIE_DEBUG
+U_CFUNC void umutablecptrie_setName(UMutableCPTrie *trie, const char *name) {
+    reinterpret_cast<MutableCodePointTrie *>(trie)->name = name;
+}
+#endif
diff --git a/thirdparty/icu4c/common/umutex.cpp b/thirdparty/icu4c/common/umutex.cpp
new file mode 100644
index 0000000000..ccbee9960a
--- /dev/null
+++ b/thirdparty/icu4c/common/umutex.cpp
@@ -0,0 +1,204 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File umutex.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/02/97    aliu        Creation.
+*   04/07/99    srl         updated
+*   05/13/99    stephen     Changed to umutex (from cmutex).
+*   11/22/99    aliu        Make non-global mutex autoinitialize [j151]
+******************************************************************************
+*/
+
+#include "umutex.h"
+
+#include "unicode/utypes.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+
+#if defined(U_USER_MUTEX_CPP)
+// Support for including an alternate implementation of mutexes has been withdrawn.
+// See issue ICU-20185.
+#error U_USER_MUTEX_CPP not supported
+#endif
+
+
+/*************************************************************************************************
+ *
+ *  ICU Mutex wrappers.
+ *
+ *************************************************************************************************/
+
+namespace {
+std::mutex *initMutex;
+std::condition_variable *initCondition;
+
+// The ICU global mutex.
+// Used when ICU implementation code passes nullptr for the mutex pointer.
+UMutex globalMutex;
+
+std::once_flag initFlag;
+std::once_flag *pInitFlag = &initFlag;
+
+}  // Anonymous namespace
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV umtx_cleanup() {
+    initMutex->~mutex();
+    initCondition->~condition_variable();
+    UMutex::cleanup();
+
+    // Reset the once_flag, by destructing it and creating a fresh one in its place.
+    // Do not use this trick anywhere else in ICU; use umtx_initOnce, not std::call_once().
+    pInitFlag->~once_flag();
+    pInitFlag = new(&initFlag) std::once_flag();
+    return true;
+}
+
+static void U_CALLCONV umtx_init() {
+    initMutex = STATIC_NEW(std::mutex);
+    initCondition = STATIC_NEW(std::condition_variable);
+    ucln_common_registerCleanup(UCLN_COMMON_MUTEX, umtx_cleanup);
+}
+U_CDECL_END
+
+
+std::mutex *UMutex::getMutex() {
+    std::mutex *retPtr = fMutex.load(std::memory_order_acquire);
+    if (retPtr == nullptr) {
+        std::call_once(*pInitFlag, umtx_init);
+        std::lock_guard<std::mutex> guard(*initMutex);
+        retPtr = fMutex.load(std::memory_order_acquire);
+        if (retPtr == nullptr) {
+            fMutex = new(fStorage) std::mutex();
+            retPtr = fMutex;
+            fListLink = gListHead;
+            gListHead = this;
+        }
+    }
+    U_ASSERT(retPtr != nullptr);
+    return retPtr;
+}
+
+UMutex *UMutex::gListHead = nullptr;
+
+void UMutex::cleanup() {
+    UMutex *next = nullptr;
+    for (UMutex *m = gListHead; m != nullptr; m = next) {
+        (*m->fMutex).~mutex();
+        m->fMutex = nullptr;
+        next = m->fListLink;
+        m->fListLink = nullptr;
+    }
+    gListHead = nullptr;
+}
+
+
+U_CAPI void  U_EXPORT2
+umtx_lock(UMutex *mutex) {
+    if (mutex == nullptr) {
+        mutex = &globalMutex;
+    }
+    mutex->lock();
+}
+
+
+U_CAPI void  U_EXPORT2
+umtx_unlock(UMutex* mutex)
+{
+    if (mutex == nullptr) {
+        mutex = &globalMutex;
+    }
+    mutex->unlock();
+}
+
+
+/*************************************************************************************************
+ *
+ *  UInitOnce Implementation
+ *
+ *************************************************************************************************/
+
+// This function is called when a test of a UInitOnce::fState reveals that
+//   initialization has not completed, that we either need to call the init
+//   function on this thread, or wait for some other thread to complete.
+//
+// The actual call to the init function is made inline by template code
+//   that knows the C++ types involved. This function returns true if
+//   the caller needs to call the Init function.
+//
+U_COMMON_API UBool U_EXPORT2
+umtx_initImplPreInit(UInitOnce &uio) {
+    std::call_once(*pInitFlag, umtx_init);
+    std::unique_lock<std::mutex> lock(*initMutex);
+    if (umtx_loadAcquire(uio.fState) == 0) {
+        umtx_storeRelease(uio.fState, 1);
+        return true;      // Caller will next call the init function.
+    } else {
+        while (umtx_loadAcquire(uio.fState) == 1) {
+            // Another thread is currently running the initialization.
+            // Wait until it completes.
+            initCondition->wait(lock);
+        }
+        U_ASSERT(uio.fState == 2);
+        return false;
+    }
+}
+
+
+// This function is called by the thread that ran an initialization function,
+// just after completing the function.
+//   Some threads may be waiting on the condition, requiring the broadcast wakeup.
+//   Some threads may be racing to test the fState variable outside of the mutex,
+//   requiring the use of store/release when changing its value.
+
+U_COMMON_API void U_EXPORT2
+umtx_initImplPostInit(UInitOnce &uio) {
+    {
+        std::unique_lock<std::mutex> lock(*initMutex);
+        umtx_storeRelease(uio.fState, 2);
+    }
+    initCondition->notify_all();
+}
+
+U_NAMESPACE_END
+
+/*************************************************************************************************
+ *
+ *  Deprecated functions for setting user mutexes.
+ *
+ *************************************************************************************************/
+
+U_DEPRECATED void U_EXPORT2
+u_setMutexFunctions(const void * /*context */, UMtxInitFn *, UMtxFn *,
+                    UMtxFn *,  UMtxFn *, UErrorCode *status) {
+    if (U_SUCCESS(*status)) {
+        *status = U_UNSUPPORTED_ERROR;
+    }
+    return;
+}
+
+
+
+U_DEPRECATED void U_EXPORT2
+u_setAtomicIncDecFunctions(const void * /*context */, UMtxAtomicFn *, UMtxAtomicFn *,
+                           UErrorCode *status) {
+    if (U_SUCCESS(*status)) {
+        *status = U_UNSUPPORTED_ERROR;
+    }
+    return;
+}
diff --git a/thirdparty/icu4c/common/umutex.h b/thirdparty/icu4c/common/umutex.h
new file mode 100644
index 0000000000..8d76b3f3e6
--- /dev/null
+++ b/thirdparty/icu4c/common/umutex.h
@@ -0,0 +1,277 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File UMUTEX.H
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/02/97  aliu        Creation.
+*   04/07/99  srl         rewrite - C interface, multiple mutices
+*   05/13/99  stephen     Changed to umutex (from cmutex)
+******************************************************************************
+*/
+
+#ifndef UMUTEX_H
+#define UMUTEX_H
+
+#include <atomic>
+#include <condition_variable>
+#include <mutex>
+#include <type_traits>
+
+#include "unicode/utypes.h"
+#include "unicode/uclean.h"
+#include "unicode/uobject.h"
+
+#include "putilimp.h"
+
+#if defined(U_USER_ATOMICS_H) || defined(U_USER_MUTEX_H)
+// Support for including an alternate implementation of atomic & mutex operations has been withdrawn.
+// See issue ICU-20185.
+#error U_USER_ATOMICS and U_USER_MUTEX_H are not supported
+#endif
+
+// Export an explicit template instantiation of std::atomic<int32_t>. 
+// When building DLLs for Windows this is required as it is used as a data member of the exported SharedObject class.
+// See digitlst.h, pluralaffix.h, datefmt.h, and others for similar examples.
+//
+// Similar story for std::atomic<std::mutex *>, and the exported UMutex class.
+#if U_PF_WINDOWS <= U_PLATFORM && U_PLATFORM <= U_PF_CYGWIN && !defined(U_IN_DOXYGEN)
+#if defined(__clang__) || defined(_MSC_VER)
+  #if defined(__clang__)
+    // Suppress the warning that the explicit instantiation after explicit specialization has no effect.
+    #pragma clang diagnostic push
+    #pragma clang diagnostic ignored "-Winstantiation-after-specialization"
+  #endif
+template struct U_COMMON_API std::atomic<int32_t>;
+template struct U_COMMON_API std::atomic<std::mutex *>;
+  #if defined(__clang__)
+    #pragma clang diagnostic pop
+  #endif
+#elif defined(__GNUC__)
+// For GCC this class is already exported/visible, so no need for U_COMMON_API.
+template struct std::atomic<int32_t>;
+template struct std::atomic<std::mutex *>;
+#endif
+#endif
+
+
+U_NAMESPACE_BEGIN
+
+/****************************************************************************
+ *
+ *   Low Level Atomic Operations, ICU wrappers for.
+ *
+ ****************************************************************************/
+
+typedef std::atomic<int32_t> u_atomic_int32_t;
+#define ATOMIC_INT32_T_INITIALIZER(val) ATOMIC_VAR_INIT(val)
+
+inline int32_t umtx_loadAcquire(u_atomic_int32_t &var) {
+    return var.load(std::memory_order_acquire);
+}
+
+inline void umtx_storeRelease(u_atomic_int32_t &var, int32_t val) {
+    var.store(val, std::memory_order_release);
+}
+
+inline int32_t umtx_atomic_inc(u_atomic_int32_t *var) {
+    return var->fetch_add(1) + 1;
+}
+
+inline int32_t umtx_atomic_dec(u_atomic_int32_t *var) {
+    return var->fetch_sub(1) - 1;
+}
+
+
+/*************************************************************************************************
+ *
+ *  UInitOnce Definitions.
+ *
+ *************************************************************************************************/
+
+struct UInitOnce {
+    u_atomic_int32_t   fState;
+    UErrorCode       fErrCode;
+    void reset() {fState = 0;}
+    UBool isReset() {return umtx_loadAcquire(fState) == 0;}
+// Note: isReset() is used by service registration code.
+//                 Thread safety of this usage needs review.
+};
+
+#define U_INITONCE_INITIALIZER {ATOMIC_INT32_T_INITIALIZER(0), U_ZERO_ERROR}
+
+
+U_COMMON_API UBool U_EXPORT2 umtx_initImplPreInit(UInitOnce &);
+U_COMMON_API void  U_EXPORT2 umtx_initImplPostInit(UInitOnce &);
+
+template<class T> void umtx_initOnce(UInitOnce &uio, T *obj, void (U_CALLCONV T::*fp)()) {
+    if (umtx_loadAcquire(uio.fState) == 2) {
+        return;
+    }
+    if (umtx_initImplPreInit(uio)) {
+        (obj->*fp)();
+        umtx_initImplPostInit(uio);
+    }
+}
+
+
+// umtx_initOnce variant for plain functions, or static class functions.
+//               No context parameter.
+inline void umtx_initOnce(UInitOnce &uio, void (U_CALLCONV *fp)()) {
+    if (umtx_loadAcquire(uio.fState) == 2) {
+        return;
+    }
+    if (umtx_initImplPreInit(uio)) {
+        (*fp)();
+        umtx_initImplPostInit(uio);
+    }
+}
+
+// umtx_initOnce variant for plain functions, or static class functions.
+//               With ErrorCode, No context parameter.
+inline void umtx_initOnce(UInitOnce &uio, void (U_CALLCONV *fp)(UErrorCode &), UErrorCode &errCode) {
+    if (U_FAILURE(errCode)) {
+        return;
+    }
+    if (umtx_loadAcquire(uio.fState) != 2 && umtx_initImplPreInit(uio)) {
+        // We run the initialization.
+        (*fp)(errCode);
+        uio.fErrCode = errCode;
+        umtx_initImplPostInit(uio);
+    } else {
+        // Someone else already ran the initialization.
+        if (U_FAILURE(uio.fErrCode)) {
+            errCode = uio.fErrCode;
+        }
+    }
+}
+
+// umtx_initOnce variant for plain functions, or static class functions,
+//               with a context parameter.
+template<class T> void umtx_initOnce(UInitOnce &uio, void (U_CALLCONV *fp)(T), T context) {
+    if (umtx_loadAcquire(uio.fState) == 2) {
+        return;
+    }
+    if (umtx_initImplPreInit(uio)) {
+        (*fp)(context);
+        umtx_initImplPostInit(uio);
+    }
+}
+
+// umtx_initOnce variant for plain functions, or static class functions,
+//               with a context parameter and an error code.
+template<class T> void umtx_initOnce(UInitOnce &uio, void (U_CALLCONV *fp)(T, UErrorCode &), T context, UErrorCode &errCode) {
+    if (U_FAILURE(errCode)) {
+        return;
+    }
+    if (umtx_loadAcquire(uio.fState) != 2 && umtx_initImplPreInit(uio)) {
+        // We run the initialization.
+        (*fp)(context, errCode);
+        uio.fErrCode = errCode;
+        umtx_initImplPostInit(uio);
+    } else {
+        // Someone else already ran the initialization.
+        if (U_FAILURE(uio.fErrCode)) {
+            errCode = uio.fErrCode;
+        }
+    }
+}
+
+// UMutex should be constexpr-constructible, so that no initialization code
+// is run during startup.
+// This works on all C++ libraries except MS VS before VS2019.
+#if (defined(_CPPLIB_VER) && !defined(_MSVC_STL_VERSION)) || \
+    (defined(_MSVC_STL_VERSION) && _MSVC_STL_VERSION < 142)
+    // (VS std lib older than VS2017) || (VS std lib version < VS2019)
+#   define UMUTEX_CONSTEXPR
+#else
+#   define UMUTEX_CONSTEXPR constexpr
+#endif
+
+/**
+ * UMutex - ICU Mutex class.
+ *
+ * This is the preferred Mutex class for use within ICU implementation code.
+ * It is a thin wrapper over C++ std::mutex, with these additions:
+ *    - Static instances are safe, not triggering static construction or destruction,
+ *      and the associated order of construction or destruction issues.
+ *    - Plumbed into u_cleanup() for destructing the underlying std::mutex,
+ *      which frees any OS level resources they may be holding.
+ *
+ * Limitations:
+ *    - Static or global instances only. Cannot be heap allocated. Cannot appear as a
+ *      member of another class.
+ *    - No condition variables or other advanced features. If needed, you will need to use
+ *      std::mutex and std::condition_variable directly. For an example, see unifiedcache.cpp
+ *
+ * Typical Usage:
+ *    static UMutex myMutex;
+ *
+ *    {
+ *       Mutex lock(myMutex);
+ *       ...    // Do stuff that is protected by myMutex;
+ *    }         // myMutex is released when lock goes out of scope.
+ */
+
+class U_COMMON_API UMutex {
+public:
+    UMUTEX_CONSTEXPR UMutex() {}
+    ~UMutex() = default;
+
+    UMutex(const UMutex &other) = delete;
+    UMutex &operator =(const UMutex &other) = delete;
+    void *operator new(size_t) = delete;
+
+    // requirements for C++ BasicLockable, allows UMutex to work with std::lock_guard
+    void lock() {
+        std::mutex *m = fMutex.load(std::memory_order_acquire);
+        if (m == nullptr) { m = getMutex(); }
+        m->lock();
+    }
+    void unlock() { fMutex.load(std::memory_order_relaxed)->unlock(); }
+
+    static void cleanup();
+
+private:
+    alignas(std::mutex) char fStorage[sizeof(std::mutex)] {};
+    std::atomic<std::mutex *> fMutex { nullptr };
+
+    /** All initialized UMutexes are kept in a linked list, so that they can be found,
+     * and the underlying std::mutex destructed, by u_cleanup().
+     */
+    UMutex *fListLink { nullptr };
+    static UMutex *gListHead;
+
+    /** Out-of-line function to lazily initialize a UMutex on first use.
+     * Initial fast check is inline, in lock().  The returned value may never
+     * be nullptr.
+     */
+    std::mutex *getMutex();
+};
+
+
+/* Lock a mutex.
+ * @param mutex The given mutex to be locked.  Pass NULL to specify
+ *              the global ICU mutex.  Recursive locks are an error
+ *              and may cause a deadlock on some platforms.
+ */
+U_CAPI void U_EXPORT2 umtx_lock(UMutex* mutex);
+
+/* Unlock a mutex.
+ * @param mutex The given mutex to be unlocked.  Pass NULL to specify
+ *              the global ICU mutex.
+ */
+U_CAPI void U_EXPORT2 umtx_unlock (UMutex* mutex);
+
+
+U_NAMESPACE_END
+
+#endif /* UMUTEX_H */
+/*eof*/
diff --git a/thirdparty/icu4c/common/unames.cpp b/thirdparty/icu4c/common/unames.cpp
new file mode 100644
index 0000000000..5776058f95
--- /dev/null
+++ b/thirdparty/icu4c/common/unames.cpp
@@ -0,0 +1,2108 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  unames.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999oct04
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/uchar.h"
+#include "unicode/udata.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "uassert.h"
+#include "ustr_imp.h"
+#include "umutex.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "ucln_cmn.h"
+#include "udataswp.h"
+#include "uprops.h"
+
+U_NAMESPACE_BEGIN
+
+/* prototypes ------------------------------------------------------------- */
+
+static const char DATA_NAME[] = "unames";
+static const char DATA_TYPE[] = "icu";
+
+#define GROUP_SHIFT 5
+#define LINES_PER_GROUP (1L<<GROUP_SHIFT)
+#define GROUP_MASK (LINES_PER_GROUP-1)
+
+/*
+ * This struct was replaced by explicitly accessing equivalent
+ * fields from triples of uint16_t.
+ * The Group struct was padded to 8 bytes on compilers for early ARM CPUs,
+ * which broke the assumption that sizeof(Group)==6 and that the ++ operator
+ * would advance by 6 bytes (3 uint16_t).
+ *
+ * We can't just change the data structure because it's loaded from a data file,
+ * and we don't want to make it less compact, so we changed the access code.
+ *
+ * For details see ICU tickets 6331 and 6008.
+typedef struct {
+    uint16_t groupMSB,
+             offsetHigh, offsetLow; / * avoid padding * /
+} Group;
+ */
+enum {
+    GROUP_MSB,
+    GROUP_OFFSET_HIGH,
+    GROUP_OFFSET_LOW,
+    GROUP_LENGTH
+};
+
+/*
+ * Get the 32-bit group offset.
+ * @param group (const uint16_t *) pointer to a Group triple of uint16_t
+ * @return group offset (int32_t)
+ */
+#define GET_GROUP_OFFSET(group) ((int32_t)(group)[GROUP_OFFSET_HIGH]<<16|(group)[GROUP_OFFSET_LOW])
+
+#define NEXT_GROUP(group) ((group)+GROUP_LENGTH)
+#define PREV_GROUP(group) ((group)-GROUP_LENGTH)
+
+typedef struct {
+    uint32_t start, end;
+    uint8_t type, variant;
+    uint16_t size;
+} AlgorithmicRange;
+
+typedef struct {
+    uint32_t tokenStringOffset, groupsOffset, groupStringOffset, algNamesOffset;
+} UCharNames;
+
+/*
+ * Get the groups table from a UCharNames struct.
+ * The groups table consists of one uint16_t groupCount followed by
+ * groupCount groups. Each group is a triple of uint16_t, see GROUP_LENGTH
+ * and the comment for the old struct Group above.
+ *
+ * @param names (const UCharNames *) pointer to the UCharNames indexes
+ * @return (const uint16_t *) pointer to the groups table
+ */
+#define GET_GROUPS(names) (const uint16_t *)((const char *)names+names->groupsOffset)
+
+typedef struct {
+    const char *otherName;
+    UChar32 code;
+} FindName;
+
+#define DO_FIND_NAME NULL
+
+static UDataMemory *uCharNamesData=NULL;
+static UCharNames *uCharNames=NULL;
+static icu::UInitOnce gCharNamesInitOnce = U_INITONCE_INITIALIZER;
+
+/*
+ * Maximum length of character names (regular & 1.0).
+ */
+static int32_t gMaxNameLength=0;
+
+/*
+ * Set of chars used in character names (regular & 1.0).
+ * Chars are platform-dependent (can be EBCDIC).
+ */
+static uint32_t gNameSet[8]={ 0 };
+
+#define U_NONCHARACTER_CODE_POINT U_CHAR_CATEGORY_COUNT
+#define U_LEAD_SURROGATE U_CHAR_CATEGORY_COUNT + 1
+#define U_TRAIL_SURROGATE U_CHAR_CATEGORY_COUNT + 2
+
+#define U_CHAR_EXTENDED_CATEGORY_COUNT (U_CHAR_CATEGORY_COUNT + 3)
+
+static const char * const charCatNames[U_CHAR_EXTENDED_CATEGORY_COUNT] = {
+    "unassigned",
+    "uppercase letter",
+    "lowercase letter",
+    "titlecase letter",
+    "modifier letter",
+    "other letter",
+    "non spacing mark",
+    "enclosing mark",
+    "combining spacing mark",
+    "decimal digit number",
+    "letter number",
+    "other number",
+    "space separator",
+    "line separator",
+    "paragraph separator",
+    "control",
+    "format",
+    "private use area",
+    "surrogate",
+    "dash punctuation",   
+    "start punctuation",
+    "end punctuation",
+    "connector punctuation",
+    "other punctuation",
+    "math symbol",
+    "currency symbol",
+    "modifier symbol",
+    "other symbol",
+    "initial punctuation",
+    "final punctuation",
+    "noncharacter",
+    "lead surrogate",
+    "trail surrogate"
+};
+
+/* implementation ----------------------------------------------------------- */
+
+static UBool U_CALLCONV unames_cleanup(void)
+{
+    if(uCharNamesData) {
+        udata_close(uCharNamesData);
+        uCharNamesData = NULL;
+    }
+    if(uCharNames) {
+        uCharNames = NULL;
+    }
+    gCharNamesInitOnce.reset();
+    gMaxNameLength=0;
+    return TRUE;
+}
+
+static UBool U_CALLCONV
+isAcceptable(void * /*context*/,
+             const char * /*type*/, const char * /*name*/,
+             const UDataInfo *pInfo) {
+    return (UBool)(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->dataFormat[0]==0x75 &&   /* dataFormat="unam" */
+        pInfo->dataFormat[1]==0x6e &&
+        pInfo->dataFormat[2]==0x61 &&
+        pInfo->dataFormat[3]==0x6d &&
+        pInfo->formatVersion[0]==1);
+}
+
+static void U_CALLCONV
+loadCharNames(UErrorCode &status) {
+    U_ASSERT(uCharNamesData == NULL);
+    U_ASSERT(uCharNames == NULL);
+
+    uCharNamesData = udata_openChoice(NULL, DATA_TYPE, DATA_NAME, isAcceptable, NULL, &status);
+    if(U_FAILURE(status)) {
+        uCharNamesData = NULL;
+    } else {
+        uCharNames = (UCharNames *)udata_getMemory(uCharNamesData);
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_UNAMES, unames_cleanup);
+}
+
+
+static UBool
+isDataLoaded(UErrorCode *pErrorCode) {
+    umtx_initOnce(gCharNamesInitOnce, &loadCharNames, *pErrorCode);
+    return U_SUCCESS(*pErrorCode);
+}
+
+#define WRITE_CHAR(buffer, bufferLength, bufferPos, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((bufferLength)>0) { \
+        *(buffer)++=c; \
+        --(bufferLength); \
+    } \
+    ++(bufferPos); \
+} UPRV_BLOCK_MACRO_END
+
+#define U_ISO_COMMENT U_CHAR_NAME_CHOICE_COUNT
+
+/*
+ * Important: expandName() and compareName() are almost the same -
+ * apply fixes to both.
+ *
+ * UnicodeData.txt uses ';' as a field separator, so no
+ * field can contain ';' as part of its contents.
+ * In unames.dat, it is marked as token[';']==-1 only if the
+ * semicolon is used in the data file - which is iff we
+ * have Unicode 1.0 names or ISO comments or aliases.
+ * So, it will be token[';']==-1 if we store U1.0 names/ISO comments/aliases
+ * although we know that it will never be part of a name.
+ */
+static uint16_t
+expandName(UCharNames *names,
+           const uint8_t *name, uint16_t nameLength, UCharNameChoice nameChoice,
+           char *buffer, uint16_t bufferLength) {
+    uint16_t *tokens=(uint16_t *)names+8;
+    uint16_t token, tokenCount=*tokens++, bufferPos=0;
+    uint8_t *tokenStrings=(uint8_t *)names+names->tokenStringOffset;
+    uint8_t c;
+
+    if(nameChoice!=U_UNICODE_CHAR_NAME && nameChoice!=U_EXTENDED_CHAR_NAME) {
+        /*
+         * skip the modern name if it is not requested _and_
+         * if the semicolon byte value is a character, not a token number
+         */
+        if((uint8_t)';'>=tokenCount || tokens[(uint8_t)';']==(uint16_t)(-1)) {
+            int fieldIndex= nameChoice==U_ISO_COMMENT ? 2 : nameChoice;
+            do {
+                while(nameLength>0) {
+                    --nameLength;
+                    if(*name++==';') {
+                        break;
+                    }
+                }
+            } while(--fieldIndex>0);
+        } else {
+            /*
+             * the semicolon byte value is a token number, therefore
+             * only modern names are stored in unames.dat and there is no
+             * such requested alternate name here
+             */
+            nameLength=0;
+        }
+    }
+
+    /* write each letter directly, and write a token word per token */
+    while(nameLength>0) {
+        --nameLength;
+        c=*name++;
+
+        if(c>=tokenCount) {
+            if(c!=';') {
+                /* implicit letter */
+                WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+            } else {
+                /* finished */
+                break;
+            }
+        } else {
+            token=tokens[c];
+            if(token==(uint16_t)(-2)) {
+                /* this is a lead byte for a double-byte token */
+                token=tokens[c<<8|*name++];
+                --nameLength;
+            }
+            if(token==(uint16_t)(-1)) {
+                if(c!=';') {
+                    /* explicit letter */
+                    WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+                } else {
+                    /* stop, but skip the semicolon if we are seeking
+                       extended names and there was no 2.0 name but there
+                       is a 1.0 name. */
+                    if(!bufferPos && nameChoice == U_EXTENDED_CHAR_NAME) {
+                        if ((uint8_t)';'>=tokenCount || tokens[(uint8_t)';']==(uint16_t)(-1)) {
+                            continue;
+                        }
+                    }
+                    /* finished */
+                    break;
+                }
+            } else {
+                /* write token word */
+                uint8_t *tokenString=tokenStrings+token;
+                while((c=*tokenString++)!=0) {
+                    WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+                }
+            }
+        }
+    }
+
+    /* zero-terminate */
+    if(bufferLength>0) {
+        *buffer=0;
+    }
+
+    return bufferPos;
+}
+
+/*
+ * compareName() is almost the same as expandName() except that it compares
+ * the currently expanded name to an input name.
+ * It returns the match/no match result as soon as possible.
+ */
+static UBool
+compareName(UCharNames *names,
+            const uint8_t *name, uint16_t nameLength, UCharNameChoice nameChoice,
+            const char *otherName) {
+    uint16_t *tokens=(uint16_t *)names+8;
+    uint16_t token, tokenCount=*tokens++;
+    uint8_t *tokenStrings=(uint8_t *)names+names->tokenStringOffset;
+    uint8_t c;
+    const char *origOtherName = otherName;
+
+    if(nameChoice!=U_UNICODE_CHAR_NAME && nameChoice!=U_EXTENDED_CHAR_NAME) {
+        /*
+         * skip the modern name if it is not requested _and_
+         * if the semicolon byte value is a character, not a token number
+         */
+        if((uint8_t)';'>=tokenCount || tokens[(uint8_t)';']==(uint16_t)(-1)) {
+            int fieldIndex= nameChoice==U_ISO_COMMENT ? 2 : nameChoice;
+            do {
+                while(nameLength>0) {
+                    --nameLength;
+                    if(*name++==';') {
+                        break;
+                    }
+                }
+            } while(--fieldIndex>0);
+        } else {
+            /*
+             * the semicolon byte value is a token number, therefore
+             * only modern names are stored in unames.dat and there is no
+             * such requested alternate name here
+             */
+            nameLength=0;
+        }
+    }
+
+    /* compare each letter directly, and compare a token word per token */
+    while(nameLength>0) {
+        --nameLength;
+        c=*name++;
+
+        if(c>=tokenCount) {
+            if(c!=';') {
+                /* implicit letter */
+                if((char)c!=*otherName++) {
+                    return FALSE;
+                }
+            } else {
+                /* finished */
+                break;
+            }
+        } else {
+            token=tokens[c];
+            if(token==(uint16_t)(-2)) {
+                /* this is a lead byte for a double-byte token */
+                token=tokens[c<<8|*name++];
+                --nameLength;
+            }
+            if(token==(uint16_t)(-1)) {
+                if(c!=';') {
+                    /* explicit letter */
+                    if((char)c!=*otherName++) {
+                        return FALSE;
+                    }
+                } else {
+                    /* stop, but skip the semicolon if we are seeking
+                       extended names and there was no 2.0 name but there
+                       is a 1.0 name. */
+                    if(otherName == origOtherName && nameChoice == U_EXTENDED_CHAR_NAME) {
+                        if ((uint8_t)';'>=tokenCount || tokens[(uint8_t)';']==(uint16_t)(-1)) {
+                            continue;
+                        }
+                    }
+                    /* finished */
+                    break;
+                }
+            } else {
+                /* write token word */
+                uint8_t *tokenString=tokenStrings+token;
+                while((c=*tokenString++)!=0) {
+                    if((char)c!=*otherName++) {
+                        return FALSE;
+                    }
+                }
+            }
+        }
+    }
+
+    /* complete match? */
+    return (UBool)(*otherName==0);
+}
+
+static uint8_t getCharCat(UChar32 cp) {
+    uint8_t cat;
+
+    if (U_IS_UNICODE_NONCHAR(cp)) {
+        return U_NONCHARACTER_CODE_POINT;
+    }
+
+    if ((cat = u_charType(cp)) == U_SURROGATE) {
+        cat = U_IS_LEAD(cp) ? U_LEAD_SURROGATE : U_TRAIL_SURROGATE;
+    }
+
+    return cat;
+}
+
+static const char *getCharCatName(UChar32 cp) {
+    uint8_t cat = getCharCat(cp);
+
+    /* Return unknown if the table of names above is not up to
+       date. */
+
+    if (cat >= UPRV_LENGTHOF(charCatNames)) {
+        return "unknown";
+    } else {
+        return charCatNames[cat];
+    }
+}
+
+static uint16_t getExtName(uint32_t code, char *buffer, uint16_t bufferLength) {
+    const char *catname = getCharCatName(code);
+    uint16_t length = 0;
+
+    UChar32 cp;
+    int ndigits, i;
+    
+    WRITE_CHAR(buffer, bufferLength, length, '<');
+    while (catname[length - 1]) {
+        WRITE_CHAR(buffer, bufferLength, length, catname[length - 1]);
+    }
+    WRITE_CHAR(buffer, bufferLength, length, '-');
+    for (cp = code, ndigits = 0; cp; ++ndigits, cp >>= 4)
+        ;
+    if (ndigits < 4)
+        ndigits = 4;
+    for (cp = code, i = ndigits; (cp || i > 0) && bufferLength; cp >>= 4, bufferLength--) {
+        uint8_t v = (uint8_t)(cp & 0xf);
+        buffer[--i] = (v < 10 ? '0' + v : 'A' + v - 10);
+    }
+    buffer += ndigits;
+    length += static_cast<uint16_t>(ndigits);
+    WRITE_CHAR(buffer, bufferLength, length, '>');
+
+    return length;
+}
+
+/*
+ * getGroup() does a binary search for the group that contains the
+ * Unicode code point "code".
+ * The return value is always a valid Group* that may contain "code"
+ * or else is the highest group before "code".
+ * If the lowest group is after "code", then that one is returned.
+ */
+static const uint16_t *
+getGroup(UCharNames *names, uint32_t code) {
+    const uint16_t *groups=GET_GROUPS(names);
+    uint16_t groupMSB=(uint16_t)(code>>GROUP_SHIFT),
+             start=0,
+             limit=*groups++,
+             number;
+
+    /* binary search for the group of names that contains the one for code */
+    while(start<limit-1) {
+        number=(uint16_t)((start+limit)/2);
+        if(groupMSB<groups[number*GROUP_LENGTH+GROUP_MSB]) {
+            limit=number;
+        } else {
+            start=number;
+        }
+    }
+
+    /* return this regardless of whether it is an exact match */
+    return groups+start*GROUP_LENGTH;
+}
+
+/*
+ * expandGroupLengths() reads a block of compressed lengths of 32 strings and
+ * expands them into offsets and lengths for each string.
+ * Lengths are stored with a variable-width encoding in consecutive nibbles:
+ * If a nibble<0xc, then it is the length itself (0=empty string).
+ * If a nibble>=0xc, then it forms a length value with the following nibble.
+ * Calculation see below.
+ * The offsets and lengths arrays must be at least 33 (one more) long because
+ * there is no check here at the end if the last nibble is still used.
+ */
+static const uint8_t *
+expandGroupLengths(const uint8_t *s,
+                   uint16_t offsets[LINES_PER_GROUP+1], uint16_t lengths[LINES_PER_GROUP+1]) {
+    /* read the lengths of the 32 strings in this group and get each string's offset */
+    uint16_t i=0, offset=0, length=0;
+    uint8_t lengthByte;
+
+    /* all 32 lengths must be read to get the offset of the first group string */
+    while(i<LINES_PER_GROUP) {
+        lengthByte=*s++;
+
+        /* read even nibble - MSBs of lengthByte */
+        if(length>=12) {
+            /* double-nibble length spread across two bytes */
+            length=(uint16_t)(((length&0x3)<<4|lengthByte>>4)+12);
+            lengthByte&=0xf;
+        } else if((lengthByte /* &0xf0 */)>=0xc0) {
+            /* double-nibble length spread across this one byte */
+            length=(uint16_t)((lengthByte&0x3f)+12);
+        } else {
+            /* single-nibble length in MSBs */
+            length=(uint16_t)(lengthByte>>4);
+            lengthByte&=0xf;
+        }
+
+        *offsets++=offset;
+        *lengths++=length;
+
+        offset+=length;
+        ++i;
+
+        /* read odd nibble - LSBs of lengthByte */
+        if((lengthByte&0xf0)==0) {
+            /* this nibble was not consumed for a double-nibble length above */
+            length=lengthByte;
+            if(length<12) {
+                /* single-nibble length in LSBs */
+                *offsets++=offset;
+                *lengths++=length;
+
+                offset+=length;
+                ++i;
+            }
+        } else {
+            length=0;   /* prevent double-nibble detection in the next iteration */
+        }
+    }
+
+    /* now, s is at the first group string */
+    return s;
+}
+
+static uint16_t
+expandGroupName(UCharNames *names, const uint16_t *group,
+                uint16_t lineNumber, UCharNameChoice nameChoice,
+                char *buffer, uint16_t bufferLength) {
+    uint16_t offsets[LINES_PER_GROUP+2], lengths[LINES_PER_GROUP+2];
+    const uint8_t *s=(uint8_t *)names+names->groupStringOffset+GET_GROUP_OFFSET(group);
+    s=expandGroupLengths(s, offsets, lengths);
+    return expandName(names, s+offsets[lineNumber], lengths[lineNumber], nameChoice,
+                      buffer, bufferLength);
+}
+
+static uint16_t
+getName(UCharNames *names, uint32_t code, UCharNameChoice nameChoice,
+        char *buffer, uint16_t bufferLength) {
+    const uint16_t *group=getGroup(names, code);
+    if((uint16_t)(code>>GROUP_SHIFT)==group[GROUP_MSB]) {
+        return expandGroupName(names, group, (uint16_t)(code&GROUP_MASK), nameChoice,
+                               buffer, bufferLength);
+    } else {
+        /* group not found */
+        /* zero-terminate */
+        if(bufferLength>0) {
+            *buffer=0;
+        }
+        return 0;
+    }
+}
+
+/*
+ * enumGroupNames() enumerates all the names in a 32-group
+ * and either calls the enumerator function or finds a given input name.
+ */
+static UBool
+enumGroupNames(UCharNames *names, const uint16_t *group,
+               UChar32 start, UChar32 end,
+               UEnumCharNamesFn *fn, void *context,
+               UCharNameChoice nameChoice) {
+    uint16_t offsets[LINES_PER_GROUP+2], lengths[LINES_PER_GROUP+2];
+    const uint8_t *s=(uint8_t *)names+names->groupStringOffset+GET_GROUP_OFFSET(group);
+
+    s=expandGroupLengths(s, offsets, lengths);
+    if(fn!=DO_FIND_NAME) {
+        char buffer[200];
+        uint16_t length;
+
+        while(start<=end) {
+            length=expandName(names, s+offsets[start&GROUP_MASK], lengths[start&GROUP_MASK], nameChoice, buffer, sizeof(buffer));
+            if (!length && nameChoice == U_EXTENDED_CHAR_NAME) {
+                buffer[length = getExtName(start, buffer, sizeof(buffer))] = 0;
+            }
+            /* here, we assume that the buffer is large enough */
+            if(length>0) {
+                if(!fn(context, start, nameChoice, buffer, length)) {
+                    return FALSE;
+                }
+            }
+            ++start;
+        }
+    } else {
+        const char *otherName=((FindName *)context)->otherName;
+        while(start<=end) {
+            if(compareName(names, s+offsets[start&GROUP_MASK], lengths[start&GROUP_MASK], nameChoice, otherName)) {
+                ((FindName *)context)->code=start;
+                return FALSE;
+            }
+            ++start;
+        }
+    }
+    return TRUE;
+}
+
+/*
+ * enumExtNames enumerate extended names.
+ * It only needs to do it if it is called with a real function and not
+ * with the dummy DO_FIND_NAME, because u_charFromName() does a check
+ * for extended names by itself.
+ */ 
+static UBool
+enumExtNames(UChar32 start, UChar32 end,
+             UEnumCharNamesFn *fn, void *context)
+{
+    if(fn!=DO_FIND_NAME) {
+        char buffer[200];
+        uint16_t length;
+        
+        while(start<=end) {
+            buffer[length = getExtName(start, buffer, sizeof(buffer))] = 0;
+            /* here, we assume that the buffer is large enough */
+            if(length>0) {
+                if(!fn(context, start, U_EXTENDED_CHAR_NAME, buffer, length)) {
+                    return FALSE;
+                }
+            }
+            ++start;
+        }
+    }
+
+    return TRUE;
+}
+
+static UBool
+enumNames(UCharNames *names,
+          UChar32 start, UChar32 limit,
+          UEnumCharNamesFn *fn, void *context,
+          UCharNameChoice nameChoice) {
+    uint16_t startGroupMSB, endGroupMSB, groupCount;
+    const uint16_t *group, *groupLimit;
+
+    startGroupMSB=(uint16_t)(start>>GROUP_SHIFT);
+    endGroupMSB=(uint16_t)((limit-1)>>GROUP_SHIFT);
+
+    /* find the group that contains start, or the highest before it */
+    group=getGroup(names, start);
+
+    if(startGroupMSB<group[GROUP_MSB] && nameChoice==U_EXTENDED_CHAR_NAME) {
+        /* enumerate synthetic names between start and the group start */
+        UChar32 extLimit=((UChar32)group[GROUP_MSB]<<GROUP_SHIFT);
+        if(extLimit>limit) {
+            extLimit=limit;
+        }
+        if(!enumExtNames(start, extLimit-1, fn, context)) {
+            return FALSE;
+        }
+        start=extLimit;
+    }
+
+    if(startGroupMSB==endGroupMSB) {
+        if(startGroupMSB==group[GROUP_MSB]) {
+            /* if start and limit-1 are in the same group, then enumerate only in that one */
+            return enumGroupNames(names, group, start, limit-1, fn, context, nameChoice);
+        }
+    } else {
+        const uint16_t *groups=GET_GROUPS(names);
+        groupCount=*groups++;
+        groupLimit=groups+groupCount*GROUP_LENGTH;
+
+        if(startGroupMSB==group[GROUP_MSB]) {
+            /* enumerate characters in the partial start group */
+            if((start&GROUP_MASK)!=0) {
+                if(!enumGroupNames(names, group,
+                                   start, ((UChar32)startGroupMSB<<GROUP_SHIFT)+LINES_PER_GROUP-1,
+                                   fn, context, nameChoice)) {
+                    return FALSE;
+                }
+                group=NEXT_GROUP(group); /* continue with the next group */
+            }
+        } else if(startGroupMSB>group[GROUP_MSB]) {
+            /* make sure that we start enumerating with the first group after start */
+            const uint16_t *nextGroup=NEXT_GROUP(group);
+            if (nextGroup < groupLimit && nextGroup[GROUP_MSB] > startGroupMSB && nameChoice == U_EXTENDED_CHAR_NAME) {
+                UChar32 end = nextGroup[GROUP_MSB] << GROUP_SHIFT;
+                if (end > limit) {
+                    end = limit;
+                }
+                if (!enumExtNames(start, end - 1, fn, context)) {
+                    return FALSE;
+                }
+            }
+            group=nextGroup;
+        }
+
+        /* enumerate entire groups between the start- and end-groups */
+        while(group<groupLimit && group[GROUP_MSB]<endGroupMSB) {
+            const uint16_t *nextGroup;
+            start=(UChar32)group[GROUP_MSB]<<GROUP_SHIFT;
+            if(!enumGroupNames(names, group, start, start+LINES_PER_GROUP-1, fn, context, nameChoice)) {
+                return FALSE;
+            }
+            nextGroup=NEXT_GROUP(group);
+            if (nextGroup < groupLimit && nextGroup[GROUP_MSB] > group[GROUP_MSB] + 1 && nameChoice == U_EXTENDED_CHAR_NAME) {
+                UChar32 end = nextGroup[GROUP_MSB] << GROUP_SHIFT;
+                if (end > limit) {
+                    end = limit;
+                }
+                if (!enumExtNames((group[GROUP_MSB] + 1) << GROUP_SHIFT, end - 1, fn, context)) {
+                    return FALSE;
+                }
+            }
+            group=nextGroup;
+        }
+
+        /* enumerate within the end group (group[GROUP_MSB]==endGroupMSB) */
+        if(group<groupLimit && group[GROUP_MSB]==endGroupMSB) {
+            return enumGroupNames(names, group, (limit-1)&~GROUP_MASK, limit-1, fn, context, nameChoice);
+        } else if (nameChoice == U_EXTENDED_CHAR_NAME && group == groupLimit) {
+            UChar32 next = (PREV_GROUP(group)[GROUP_MSB] + 1) << GROUP_SHIFT;
+            if (next > start) {
+                start = next;
+            }
+        } else {
+            return TRUE;
+        }
+    }
+
+    /* we have not found a group, which means everything is made of
+       extended names. */
+    if (nameChoice == U_EXTENDED_CHAR_NAME) {
+        if (limit > UCHAR_MAX_VALUE + 1) {
+            limit = UCHAR_MAX_VALUE + 1;
+        }
+        return enumExtNames(start, limit - 1, fn, context);
+    }
+    
+    return TRUE;
+}
+
+static uint16_t
+writeFactorSuffix(const uint16_t *factors, uint16_t count,
+                  const char *s, /* suffix elements */
+                  uint32_t code,
+                  uint16_t indexes[8], /* output fields from here */
+                  const char *elementBases[8], const char *elements[8],
+                  char *buffer, uint16_t bufferLength) {
+    uint16_t i, factor, bufferPos=0;
+    char c;
+
+    /* write elements according to the factors */
+
+    /*
+     * the factorized elements are determined by modulo arithmetic
+     * with the factors of this algorithm
+     *
+     * note that for fewer operations, count is decremented here
+     */
+    --count;
+    for(i=count; i>0; --i) {
+        factor=factors[i];
+        indexes[i]=(uint16_t)(code%factor);
+        code/=factor;
+    }
+    /*
+     * we don't need to calculate the last modulus because start<=code<=end
+     * guarantees here that code<=factors[0]
+     */
+    indexes[0]=(uint16_t)code;
+
+    /* write each element */
+    for(;;) {
+        if(elementBases!=NULL) {
+            *elementBases++=s;
+        }
+
+        /* skip indexes[i] strings */
+        factor=indexes[i];
+        while(factor>0) {
+            while(*s++!=0) {}
+            --factor;
+        }
+        if(elements!=NULL) {
+            *elements++=s;
+        }
+
+        /* write element */
+        while((c=*s++)!=0) {
+            WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+        }
+
+        /* we do not need to perform the rest of this loop for i==count - break here */
+        if(i>=count) {
+            break;
+        }
+
+        /* skip the rest of the strings for this factors[i] */
+        factor=(uint16_t)(factors[i]-indexes[i]-1);
+        while(factor>0) {
+            while(*s++!=0) {}
+            --factor;
+        }
+
+        ++i;
+    }
+
+    /* zero-terminate */
+    if(bufferLength>0) {
+        *buffer=0;
+    }
+
+    return bufferPos;
+}
+
+/*
+ * Important:
+ * Parts of findAlgName() are almost the same as some of getAlgName().
+ * Fixes must be applied to both.
+ */
+static uint16_t
+getAlgName(AlgorithmicRange *range, uint32_t code, UCharNameChoice nameChoice,
+        char *buffer, uint16_t bufferLength) {
+    uint16_t bufferPos=0;
+
+    /* Only the normative character name can be algorithmic. */
+    if(nameChoice!=U_UNICODE_CHAR_NAME && nameChoice!=U_EXTENDED_CHAR_NAME) {
+        /* zero-terminate */
+        if(bufferLength>0) {
+            *buffer=0;
+        }
+        return 0;
+    }
+
+    switch(range->type) {
+    case 0: {
+        /* name = prefix hex-digits */
+        const char *s=(const char *)(range+1);
+        char c;
+
+        uint16_t i, count;
+
+        /* copy prefix */
+        while((c=*s++)!=0) {
+            WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+        }
+
+        /* write hexadecimal code point value */
+        count=range->variant;
+
+        /* zero-terminate */
+        if(count<bufferLength) {
+            buffer[count]=0;
+        }
+
+        for(i=count; i>0;) {
+            if(--i<bufferLength) {
+                c=(char)(code&0xf);
+                if(c<10) {
+                    c+='0';
+                } else {
+                    c+='A'-10;
+                }
+                buffer[i]=c;
+            }
+            code>>=4;
+        }
+
+        bufferPos+=count;
+        break;
+    }
+    case 1: {
+        /* name = prefix factorized-elements */
+        uint16_t indexes[8];
+        const uint16_t *factors=(const uint16_t *)(range+1);
+        uint16_t count=range->variant;
+        const char *s=(const char *)(factors+count);
+        char c;
+
+        /* copy prefix */
+        while((c=*s++)!=0) {
+            WRITE_CHAR(buffer, bufferLength, bufferPos, c);
+        }
+
+        bufferPos+=writeFactorSuffix(factors, count,
+                                     s, code-range->start, indexes, NULL, NULL, buffer, bufferLength);
+        break;
+    }
+    default:
+        /* undefined type */
+        /* zero-terminate */
+        if(bufferLength>0) {
+            *buffer=0;
+        }
+        break;
+    }
+
+    return bufferPos;
+}
+
+/*
+ * Important: enumAlgNames() and findAlgName() are almost the same.
+ * Any fix must be applied to both.
+ */
+static UBool
+enumAlgNames(AlgorithmicRange *range,
+             UChar32 start, UChar32 limit,
+             UEnumCharNamesFn *fn, void *context,
+             UCharNameChoice nameChoice) {
+    char buffer[200];
+    uint16_t length;
+
+    if(nameChoice!=U_UNICODE_CHAR_NAME && nameChoice!=U_EXTENDED_CHAR_NAME) {
+        return TRUE;
+    }
+
+    switch(range->type) {
+    case 0: {
+        char *s, *end;
+        char c;
+
+        /* get the full name of the start character */
+        length=getAlgName(range, (uint32_t)start, nameChoice, buffer, sizeof(buffer));
+        if(length<=0) {
+            return TRUE;
+        }
+
+        /* call the enumerator function with this first character */
+        if(!fn(context, start, nameChoice, buffer, length)) {
+            return FALSE;
+        }
+
+        /* go to the end of the name; all these names have the same length */
+        end=buffer;
+        while(*end!=0) {
+            ++end;
+        }
+
+        /* enumerate the rest of the names */
+        while(++start<limit) {
+            /* increment the hexadecimal number on a character-basis */
+            s=end;
+            for (;;) {
+                c=*--s;
+                if(('0'<=c && c<'9') || ('A'<=c && c<'F')) {
+                    *s=(char)(c+1);
+                    break;
+                } else if(c=='9') {
+                    *s='A';
+                    break;
+                } else if(c=='F') {
+                    *s='0';
+                }
+            }
+
+            if(!fn(context, start, nameChoice, buffer, length)) {
+                return FALSE;
+            }
+        }
+        break;
+    }
+    case 1: {
+        uint16_t indexes[8];
+        const char *elementBases[8], *elements[8];
+        const uint16_t *factors=(const uint16_t *)(range+1);
+        uint16_t count=range->variant;
+        const char *s=(const char *)(factors+count);
+        char *suffix, *t;
+        uint16_t prefixLength, i, idx;
+
+        char c;
+
+        /* name = prefix factorized-elements */
+
+        /* copy prefix */
+        suffix=buffer;
+        prefixLength=0;
+        while((c=*s++)!=0) {
+            *suffix++=c;
+            ++prefixLength;
+        }
+
+        /* append the suffix of the start character */
+        length=(uint16_t)(prefixLength+writeFactorSuffix(factors, count,
+                                              s, (uint32_t)start-range->start,
+                                              indexes, elementBases, elements,
+                                              suffix, (uint16_t)(sizeof(buffer)-prefixLength)));
+
+        /* call the enumerator function with this first character */
+        if(!fn(context, start, nameChoice, buffer, length)) {
+            return FALSE;
+        }
+
+        /* enumerate the rest of the names */
+        while(++start<limit) {
+            /* increment the indexes in lexical order bound by the factors */
+            i=count;
+            for (;;) {
+                idx=(uint16_t)(indexes[--i]+1);
+                if(idx<factors[i]) {
+                    /* skip one index and its element string */
+                    indexes[i]=idx;
+                    s=elements[i];
+                    while(*s++!=0) {
+                    }
+                    elements[i]=s;
+                    break;
+                } else {
+                    /* reset this index to 0 and its element string to the first one */
+                    indexes[i]=0;
+                    elements[i]=elementBases[i];
+                }
+            }
+
+            /* to make matters a little easier, just append all elements to the suffix */
+            t=suffix;
+            length=prefixLength;
+            for(i=0; i<count; ++i) {
+                s=elements[i];
+                while((c=*s++)!=0) {
+                    *t++=c;
+                    ++length;
+                }
+            }
+            /* zero-terminate */
+            *t=0;
+
+            if(!fn(context, start, nameChoice, buffer, length)) {
+                return FALSE;
+            }
+        }
+        break;
+    }
+    default:
+        /* undefined type */
+        break;
+    }
+
+    return TRUE;
+}
+
+/*
+ * findAlgName() is almost the same as enumAlgNames() except that it
+ * returns the code point for a name if it fits into the range.
+ * It returns 0xffff otherwise.
+ */
+static UChar32
+findAlgName(AlgorithmicRange *range, UCharNameChoice nameChoice, const char *otherName) {
+    UChar32 code;
+
+    if(nameChoice!=U_UNICODE_CHAR_NAME && nameChoice!=U_EXTENDED_CHAR_NAME) {
+        return 0xffff;
+    }
+
+    switch(range->type) {
+    case 0: {
+        /* name = prefix hex-digits */
+        const char *s=(const char *)(range+1);
+        char c;
+
+        uint16_t i, count;
+
+        /* compare prefix */
+        while((c=*s++)!=0) {
+            if((char)c!=*otherName++) {
+                return 0xffff;
+            }
+        }
+
+        /* read hexadecimal code point value */
+        count=range->variant;
+        code=0;
+        for(i=0; i<count; ++i) {
+            c=*otherName++;
+            if('0'<=c && c<='9') {
+                code=(code<<4)|(c-'0');
+            } else if('A'<=c && c<='F') {
+                code=(code<<4)|(c-'A'+10);
+            } else {
+                return 0xffff;
+            }
+        }
+
+        /* does it fit into the range? */
+        if(*otherName==0 && range->start<=(uint32_t)code && (uint32_t)code<=range->end) {
+            return code;
+        }
+        break;
+    }
+    case 1: {
+        char buffer[64];
+        uint16_t indexes[8];
+        const char *elementBases[8], *elements[8];
+        const uint16_t *factors=(const uint16_t *)(range+1);
+        uint16_t count=range->variant;
+        const char *s=(const char *)(factors+count), *t;
+        UChar32 start, limit;
+        uint16_t i, idx;
+
+        char c;
+
+        /* name = prefix factorized-elements */
+
+        /* compare prefix */
+        while((c=*s++)!=0) {
+            if((char)c!=*otherName++) {
+                return 0xffff;
+            }
+        }
+
+        start=(UChar32)range->start;
+        limit=(UChar32)(range->end+1);
+
+        /* initialize the suffix elements for enumeration; indexes should all be set to 0 */
+        writeFactorSuffix(factors, count, s, 0,
+                          indexes, elementBases, elements, buffer, sizeof(buffer));
+
+        /* compare the first suffix */
+        if(0==uprv_strcmp(otherName, buffer)) {
+            return start;
+        }
+
+        /* enumerate and compare the rest of the suffixes */
+        while(++start<limit) {
+            /* increment the indexes in lexical order bound by the factors */
+            i=count;
+            for (;;) {
+                idx=(uint16_t)(indexes[--i]+1);
+                if(idx<factors[i]) {
+                    /* skip one index and its element string */
+                    indexes[i]=idx;
+                    s=elements[i];
+                    while(*s++!=0) {}
+                    elements[i]=s;
+                    break;
+                } else {
+                    /* reset this index to 0 and its element string to the first one */
+                    indexes[i]=0;
+                    elements[i]=elementBases[i];
+                }
+            }
+
+            /* to make matters a little easier, just compare all elements of the suffix */
+            t=otherName;
+            for(i=0; i<count; ++i) {
+                s=elements[i];
+                while((c=*s++)!=0) {
+                    if(c!=*t++) {
+                        s=""; /* does not match */
+                        i=99;
+                    }
+                }
+            }
+            if(i<99 && *t==0) {
+                return start;
+            }
+        }
+        break;
+    }
+    default:
+        /* undefined type */
+        break;
+    }
+
+    return 0xffff;
+}
+
+/* sets of name characters, maximum name lengths ---------------------------- */
+
+#define SET_ADD(set, c) ((set)[(uint8_t)c>>5]|=((uint32_t)1<<((uint8_t)c&0x1f)))
+#define SET_CONTAINS(set, c) (((set)[(uint8_t)c>>5]&((uint32_t)1<<((uint8_t)c&0x1f)))!=0)
+
+static int32_t
+calcStringSetLength(uint32_t set[8], const char *s) {
+    int32_t length=0;
+    char c;
+
+    while((c=*s++)!=0) {
+        SET_ADD(set, c);
+        ++length;
+    }
+    return length;
+}
+
+static int32_t
+calcAlgNameSetsLengths(int32_t maxNameLength) {
+    AlgorithmicRange *range;
+    uint32_t *p;
+    uint32_t rangeCount;
+    int32_t length;
+
+    /* enumerate algorithmic ranges */
+    p=(uint32_t *)((uint8_t *)uCharNames+uCharNames->algNamesOffset);
+    rangeCount=*p;
+    range=(AlgorithmicRange *)(p+1);
+    while(rangeCount>0) {
+        switch(range->type) {
+        case 0:
+            /* name = prefix + (range->variant times) hex-digits */
+            /* prefix */
+            length=calcStringSetLength(gNameSet, (const char *)(range+1))+range->variant;
+            if(length>maxNameLength) {
+                maxNameLength=length;
+            }
+            break;
+        case 1: {
+            /* name = prefix factorized-elements */
+            const uint16_t *factors=(const uint16_t *)(range+1);
+            const char *s;
+            int32_t i, count=range->variant, factor, factorLength, maxFactorLength;
+
+            /* prefix length */
+            s=(const char *)(factors+count);
+            length=calcStringSetLength(gNameSet, s);
+            s+=length+1; /* start of factor suffixes */
+
+            /* get the set and maximum factor suffix length for each factor */
+            for(i=0; i<count; ++i) {
+                maxFactorLength=0;
+                for(factor=factors[i]; factor>0; --factor) {
+                    factorLength=calcStringSetLength(gNameSet, s);
+                    s+=factorLength+1;
+                    if(factorLength>maxFactorLength) {
+                        maxFactorLength=factorLength;
+                    }
+                }
+                length+=maxFactorLength;
+            }
+
+            if(length>maxNameLength) {
+                maxNameLength=length;
+            }
+            break;
+        }
+        default:
+            /* unknown type */
+            break;
+        }
+
+        range=(AlgorithmicRange *)((uint8_t *)range+range->size);
+        --rangeCount;
+    }
+    return maxNameLength;
+}
+
+static int32_t
+calcExtNameSetsLengths(int32_t maxNameLength) {
+    int32_t i, length;
+
+    for(i=0; i<UPRV_LENGTHOF(charCatNames); ++i) {
+        /*
+         * for each category, count the length of the category name
+         * plus 9=
+         * 2 for <>
+         * 1 for -
+         * 6 for most hex digits per code point
+         */
+        length=9+calcStringSetLength(gNameSet, charCatNames[i]);
+        if(length>maxNameLength) {
+            maxNameLength=length;
+        }
+    }
+    return maxNameLength;
+}
+
+static int32_t
+calcNameSetLength(const uint16_t *tokens, uint16_t tokenCount, const uint8_t *tokenStrings, int8_t *tokenLengths,
+                  uint32_t set[8],
+                  const uint8_t **pLine, const uint8_t *lineLimit) {
+    const uint8_t *line=*pLine;
+    int32_t length=0, tokenLength;
+    uint16_t c, token;
+
+    while(line!=lineLimit && (c=*line++)!=(uint8_t)';') {
+        if(c>=tokenCount) {
+            /* implicit letter */
+            SET_ADD(set, c);
+            ++length;
+        } else {
+            token=tokens[c];
+            if(token==(uint16_t)(-2)) {
+                /* this is a lead byte for a double-byte token */
+                c=c<<8|*line++;
+                token=tokens[c];
+            }
+            if(token==(uint16_t)(-1)) {
+                /* explicit letter */
+                SET_ADD(set, c);
+                ++length;
+            } else {
+                /* count token word */
+                if(tokenLengths!=NULL) {
+                    /* use cached token length */
+                    tokenLength=tokenLengths[c];
+                    if(tokenLength==0) {
+                        tokenLength=calcStringSetLength(set, (const char *)tokenStrings+token);
+                        tokenLengths[c]=(int8_t)tokenLength;
+                    }
+                } else {
+                    tokenLength=calcStringSetLength(set, (const char *)tokenStrings+token);
+                }
+                length+=tokenLength;
+            }
+        }
+    }
+
+    *pLine=line;
+    return length;
+}
+
+static void
+calcGroupNameSetsLengths(int32_t maxNameLength) {
+    uint16_t offsets[LINES_PER_GROUP+2], lengths[LINES_PER_GROUP+2];
+
+    uint16_t *tokens=(uint16_t *)uCharNames+8;
+    uint16_t tokenCount=*tokens++;
+    uint8_t *tokenStrings=(uint8_t *)uCharNames+uCharNames->tokenStringOffset;
+
+    int8_t *tokenLengths;
+
+    const uint16_t *group;
+    const uint8_t *s, *line, *lineLimit;
+
+    int32_t groupCount, lineNumber, length;
+
+    tokenLengths=(int8_t *)uprv_malloc(tokenCount);
+    if(tokenLengths!=NULL) {
+        uprv_memset(tokenLengths, 0, tokenCount);
+    }
+
+    group=GET_GROUPS(uCharNames);
+    groupCount=*group++;
+
+    /* enumerate all groups */
+    while(groupCount>0) {
+        s=(uint8_t *)uCharNames+uCharNames->groupStringOffset+GET_GROUP_OFFSET(group);
+        s=expandGroupLengths(s, offsets, lengths);
+
+        /* enumerate all lines in each group */
+        for(lineNumber=0; lineNumber<LINES_PER_GROUP; ++lineNumber) {
+            line=s+offsets[lineNumber];
+            length=lengths[lineNumber];
+            if(length==0) {
+                continue;
+            }
+
+            lineLimit=line+length;
+
+            /* read regular name */
+            length=calcNameSetLength(tokens, tokenCount, tokenStrings, tokenLengths, gNameSet, &line, lineLimit);
+            if(length>maxNameLength) {
+                maxNameLength=length;
+            }
+            if(line==lineLimit) {
+                continue;
+            }
+
+            /* read Unicode 1.0 name */
+            length=calcNameSetLength(tokens, tokenCount, tokenStrings, tokenLengths, gNameSet, &line, lineLimit);
+            if(length>maxNameLength) {
+                maxNameLength=length;
+            }
+            if(line==lineLimit) {
+                continue;
+            }
+
+            /* read ISO comment */
+            /*length=calcNameSetLength(tokens, tokenCount, tokenStrings, tokenLengths, gISOCommentSet, &line, lineLimit);*/
+        }
+
+        group=NEXT_GROUP(group);
+        --groupCount;
+    }
+
+    if(tokenLengths!=NULL) {
+        uprv_free(tokenLengths);
+    }
+
+    /* set gMax... - name length last for threading */
+    gMaxNameLength=maxNameLength;
+}
+
+static UBool
+calcNameSetsLengths(UErrorCode *pErrorCode) {
+    static const char extChars[]="0123456789ABCDEF<>-";
+    int32_t i, maxNameLength;
+
+    if(gMaxNameLength!=0) {
+        return TRUE;
+    }
+
+    if(!isDataLoaded(pErrorCode)) {
+        return FALSE;
+    }
+
+    /* set hex digits, used in various names, and <>-, used in extended names */
+    for(i=0; i<(int32_t)sizeof(extChars)-1; ++i) {
+        SET_ADD(gNameSet, extChars[i]);
+    }
+
+    /* set sets and lengths from algorithmic names */
+    maxNameLength=calcAlgNameSetsLengths(0);
+
+    /* set sets and lengths from extended names */
+    maxNameLength=calcExtNameSetsLengths(maxNameLength);
+
+    /* set sets and lengths from group names, set global maximum values */
+    calcGroupNameSetsLengths(maxNameLength);
+
+    return TRUE;
+}
+
+U_NAMESPACE_END
+
+/* public API --------------------------------------------------------------- */
+
+U_NAMESPACE_USE
+
+U_CAPI int32_t U_EXPORT2
+u_charName(UChar32 code, UCharNameChoice nameChoice,
+           char *buffer, int32_t bufferLength,
+           UErrorCode *pErrorCode) {
+     AlgorithmicRange *algRange;
+    uint32_t *p;
+    uint32_t i;
+    int32_t length;
+
+    /* check the argument values */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    } else if(nameChoice>=U_CHAR_NAME_CHOICE_COUNT ||
+              bufferLength<0 || (bufferLength>0 && buffer==NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if((uint32_t)code>UCHAR_MAX_VALUE || !isDataLoaded(pErrorCode)) {
+        return u_terminateChars(buffer, bufferLength, 0, pErrorCode);
+    }
+
+    length=0;
+
+    /* try algorithmic names first */
+    p=(uint32_t *)((uint8_t *)uCharNames+uCharNames->algNamesOffset);
+    i=*p;
+    algRange=(AlgorithmicRange *)(p+1);
+    while(i>0) {
+        if(algRange->start<=(uint32_t)code && (uint32_t)code<=algRange->end) {
+            length=getAlgName(algRange, (uint32_t)code, nameChoice, buffer, (uint16_t)bufferLength);
+            break;
+        }
+        algRange=(AlgorithmicRange *)((uint8_t *)algRange+algRange->size);
+        --i;
+    }
+
+    if(i==0) {
+        if (nameChoice == U_EXTENDED_CHAR_NAME) {
+            length = getName(uCharNames, (uint32_t )code, U_EXTENDED_CHAR_NAME, buffer, (uint16_t) bufferLength);
+            if (!length) {
+                /* extended character name */
+                length = getExtName((uint32_t) code, buffer, (uint16_t) bufferLength);
+            }
+        } else {
+            /* normal character name */
+            length=getName(uCharNames, (uint32_t)code, nameChoice, buffer, (uint16_t)bufferLength);
+        }
+    }
+
+    return u_terminateChars(buffer, bufferLength, length, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_getISOComment(UChar32 /*c*/,
+                char *dest, int32_t destCapacity,
+                UErrorCode *pErrorCode) {
+    /* check the argument values */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    } else if(destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    return u_terminateChars(dest, destCapacity, 0, pErrorCode);
+}
+
+U_CAPI UChar32 U_EXPORT2
+u_charFromName(UCharNameChoice nameChoice,
+               const char *name,
+               UErrorCode *pErrorCode) {
+    char upper[120] = {0};
+    char lower[120] = {0};
+    FindName findName;
+    AlgorithmicRange *algRange;
+    uint32_t *p;
+    uint32_t i;
+    UChar32 cp = 0;
+    char c0;
+    static constexpr UChar32 error = 0xffff;     /* Undefined, but use this for backwards compatibility. */
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return error;
+    }
+
+    if(nameChoice>=U_CHAR_NAME_CHOICE_COUNT || name==NULL || *name==0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return error;
+    }
+
+    if(!isDataLoaded(pErrorCode)) {
+        return error;
+    }
+
+    /* construct the uppercase and lowercase of the name first */
+    for(i=0; i<sizeof(upper); ++i) {
+        if((c0=*name++)!=0) {
+            upper[i]=uprv_toupper(c0);
+            lower[i]=uprv_tolower(c0);
+        } else {
+            upper[i]=lower[i]=0;
+            break;
+        }
+    }
+    if(i==sizeof(upper)) {
+        /* name too long, there is no such character */
+        *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+        return error;
+    }
+    // i==strlen(name)==strlen(lower)==strlen(upper)
+
+    /* try extended names first */
+    if (lower[0] == '<') {
+        if (nameChoice == U_EXTENDED_CHAR_NAME && lower[--i] == '>') {
+            // Parse a string like "<category-HHHH>" where HHHH is a hex code point.
+            uint32_t limit = i;
+            while (i >= 3 && lower[--i] != '-') {}
+
+            // There should be 1 to 8 hex digits.
+            int32_t hexLength = limit - (i + 1);
+            if (i >= 2 && lower[i] == '-' && 1 <= hexLength && hexLength <= 8) {
+                uint32_t cIdx;
+
+                lower[i] = 0;
+
+                for (++i; i < limit; ++i) {
+                    if (lower[i] >= '0' && lower[i] <= '9') {
+                        cp = (cp << 4) + lower[i] - '0';
+                    } else if (lower[i] >= 'a' && lower[i] <= 'f') {
+                        cp = (cp << 4) + lower[i] - 'a' + 10;
+                    } else {
+                        *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+                        return error;
+                    }
+                    // Prevent signed-integer overflow and out-of-range code points.
+                    if (cp > UCHAR_MAX_VALUE) {
+                        *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+                        return error;
+                    }
+                }
+
+                /* Now validate the category name.
+                   We could use a binary search, or a trie, if
+                   we really wanted to. */
+                uint8_t cat = getCharCat(cp);
+                for (lower[i] = 0, cIdx = 0; cIdx < UPRV_LENGTHOF(charCatNames); ++cIdx) {
+
+                    if (!uprv_strcmp(lower + 1, charCatNames[cIdx])) {
+                        if (cat == cIdx) {
+                            return cp;
+                        }
+                        break;
+                    }
+                }
+            }
+        }
+
+        *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+        return error;
+    }
+
+    /* try algorithmic names now */
+    p=(uint32_t *)((uint8_t *)uCharNames+uCharNames->algNamesOffset);
+    i=*p;
+    algRange=(AlgorithmicRange *)(p+1);
+    while(i>0) {
+        if((cp=findAlgName(algRange, nameChoice, upper))!=0xffff) {
+            return cp;
+        }
+        algRange=(AlgorithmicRange *)((uint8_t *)algRange+algRange->size);
+        --i;
+    }
+
+    /* normal character name */
+    findName.otherName=upper;
+    findName.code=error;
+    enumNames(uCharNames, 0, UCHAR_MAX_VALUE + 1, DO_FIND_NAME, &findName, nameChoice);
+    if (findName.code == error) {
+         *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+    }
+    return findName.code;
+}
+
+U_CAPI void U_EXPORT2
+u_enumCharNames(UChar32 start, UChar32 limit,
+                UEnumCharNamesFn *fn,
+                void *context,
+                UCharNameChoice nameChoice,
+                UErrorCode *pErrorCode) {
+    AlgorithmicRange *algRange;
+    uint32_t *p;
+    uint32_t i;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    if(nameChoice>=U_CHAR_NAME_CHOICE_COUNT || fn==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if((uint32_t) limit > UCHAR_MAX_VALUE + 1) {
+        limit = UCHAR_MAX_VALUE + 1;
+    }
+    if((uint32_t)start>=(uint32_t)limit) {
+        return;
+    }
+
+    if(!isDataLoaded(pErrorCode)) {
+        return;
+    }
+
+    /* interleave the data-driven ones with the algorithmic ones */
+    /* iterate over all algorithmic ranges; assume that they are in ascending order */
+    p=(uint32_t *)((uint8_t *)uCharNames+uCharNames->algNamesOffset);
+    i=*p;
+    algRange=(AlgorithmicRange *)(p+1);
+    while(i>0) {
+        /* enumerate the character names before the current algorithmic range */
+        /* here: start<limit */
+        if((uint32_t)start<algRange->start) {
+            if((uint32_t)limit<=algRange->start) {
+                enumNames(uCharNames, start, limit, fn, context, nameChoice);
+                return;
+            }
+            if(!enumNames(uCharNames, start, (UChar32)algRange->start, fn, context, nameChoice)) {
+                return;
+            }
+            start=(UChar32)algRange->start;
+        }
+        /* enumerate the character names in the current algorithmic range */
+        /* here: algRange->start<=start<limit */
+        if((uint32_t)start<=algRange->end) {
+            if((uint32_t)limit<=(algRange->end+1)) {
+                enumAlgNames(algRange, start, limit, fn, context, nameChoice);
+                return;
+            }
+            if(!enumAlgNames(algRange, start, (UChar32)algRange->end+1, fn, context, nameChoice)) {
+                return;
+            }
+            start=(UChar32)algRange->end+1;
+        }
+        /* continue to the next algorithmic range (here: start<limit) */
+        algRange=(AlgorithmicRange *)((uint8_t *)algRange+algRange->size);
+        --i;
+    }
+    /* enumerate the character names after the last algorithmic range */
+    enumNames(uCharNames, start, limit, fn, context, nameChoice);
+}
+
+U_CAPI int32_t U_EXPORT2
+uprv_getMaxCharNameLength() {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    if(calcNameSetsLengths(&errorCode)) {
+        return gMaxNameLength;
+    } else {
+        return 0;
+    }
+}
+
+/**
+ * Converts the char set cset into a Unicode set uset.
+ * @param cset Set of 256 bit flags corresponding to a set of chars.
+ * @param uset USet to receive characters. Existing contents are deleted.
+ */
+static void
+charSetToUSet(uint32_t cset[8], const USetAdder *sa) {
+    UChar us[256];
+    char cs[256];
+
+    int32_t i, length;
+    UErrorCode errorCode;
+
+    errorCode=U_ZERO_ERROR;
+
+    if(!calcNameSetsLengths(&errorCode)) {
+        return;
+    }
+
+    /* build a char string with all chars that are used in character names */
+    length=0;
+    for(i=0; i<256; ++i) {
+        if(SET_CONTAINS(cset, i)) {
+            cs[length++]=(char)i;
+        }
+    }
+
+    /* convert the char string to a UChar string */
+    u_charsToUChars(cs, us, length);
+
+    /* add each UChar to the USet */
+    for(i=0; i<length; ++i) {
+        if(us[i]!=0 || cs[i]==0) { /* non-invariant chars become (UChar)0 */
+            sa->add(sa->set, us[i]);
+        }
+    }
+}
+
+/**
+ * Fills set with characters that are used in Unicode character names.
+ * @param set USet to receive characters.
+ */
+U_CAPI void U_EXPORT2
+uprv_getCharNameCharacters(const USetAdder *sa) {
+    charSetToUSet(gNameSet, sa);
+}
+
+/* data swapping ------------------------------------------------------------ */
+
+/*
+ * The token table contains non-negative entries for token bytes,
+ * and -1 for bytes that represent themselves in the data file's charset.
+ * -2 entries are used for lead bytes.
+ *
+ * Direct bytes (-1 entries) must be translated from the input charset family
+ * to the output charset family.
+ * makeTokenMap() writes a permutation mapping for this.
+ * Use it once for single-/lead-byte tokens and once more for all trail byte
+ * tokens. (';' is an unused trail byte marked with -1.)
+ */
+static void
+makeTokenMap(const UDataSwapper *ds,
+             int16_t tokens[], uint16_t tokenCount,
+             uint8_t map[256],
+             UErrorCode *pErrorCode) {
+    UBool usedOutChar[256];
+    uint16_t i, j;
+    uint8_t c1, c2;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    if(ds->inCharset==ds->outCharset) {
+        /* Same charset family: identity permutation */
+        for(i=0; i<256; ++i) {
+            map[i]=(uint8_t)i;
+        }
+    } else {
+        uprv_memset(map, 0, 256);
+        uprv_memset(usedOutChar, 0, 256);
+
+        if(tokenCount>256) {
+            tokenCount=256;
+        }
+
+        /* set the direct bytes (byte 0 always maps to itself) */
+        for(i=1; i<tokenCount; ++i) {
+            if(tokens[i]==-1) {
+                /* convert the direct byte character */
+                c1=(uint8_t)i;
+                ds->swapInvChars(ds, &c1, 1, &c2, pErrorCode);
+                if(U_FAILURE(*pErrorCode)) {
+                    udata_printError(ds, "unames/makeTokenMap() finds variant character 0x%02x used (input charset family %d)\n",
+                                     i, ds->inCharset);
+                    return;
+                }
+
+                /* enter the converted character into the map and mark it used */
+                map[c1]=c2;
+                usedOutChar[c2]=TRUE;
+            }
+        }
+
+        /* set the mappings for the rest of the permutation */
+        for(i=j=1; i<tokenCount; ++i) {
+            /* set mappings that were not set for direct bytes */
+            if(map[i]==0) {
+                /* set an output byte value that was not used as an output byte above */
+                while(usedOutChar[j]) {
+                    ++j;
+                }
+                map[i]=(uint8_t)j++;
+            }
+        }
+
+        /*
+         * leave mappings at tokenCount and above unset if tokenCount<256
+         * because they won't be used
+         */
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uchar_swapNames(const UDataSwapper *ds,
+                const void *inData, int32_t length, void *outData,
+                UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    uint32_t tokenStringOffset, groupsOffset, groupStringOffset, algNamesOffset,
+             offset, i, count, stringsCount;
+
+    const AlgorithmicRange *inRange;
+    AlgorithmicRange *outRange;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x75 &&   /* dataFormat="unam" */
+        pInfo->dataFormat[1]==0x6e &&
+        pInfo->dataFormat[2]==0x61 &&
+        pInfo->dataFormat[3]==0x6d &&
+        pInfo->formatVersion[0]==1
+    )) {
+        udata_printError(ds, "uchar_swapNames(): data format %02x.%02x.%02x.%02x (format version %02x) is not recognized as unames.icu\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData+headerSize;
+    outBytes=(uint8_t *)outData+headerSize;
+    if(length<0) {
+        algNamesOffset=ds->readUInt32(((const uint32_t *)inBytes)[3]);
+    } else {
+        length-=headerSize;
+        if( length<20 ||
+            (uint32_t)length<(algNamesOffset=ds->readUInt32(((const uint32_t *)inBytes)[3]))
+        ) {
+            udata_printError(ds, "uchar_swapNames(): too few bytes (%d after header) for unames.icu\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    if(length<0) {
+        /* preflighting: iterate through algorithmic ranges */
+        offset=algNamesOffset;
+        count=ds->readUInt32(*((const uint32_t *)(inBytes+offset)));
+        offset+=4;
+
+        for(i=0; i<count; ++i) {
+            inRange=(const AlgorithmicRange *)(inBytes+offset);
+            offset+=ds->readUInt16(inRange->size);
+        }
+    } else {
+        /* swap data */
+        const uint16_t *p;
+        uint16_t *q, *temp;
+
+        int16_t tokens[512];
+        uint16_t tokenCount;
+
+        uint8_t map[256], trailMap[256];
+
+        /* copy the data for inaccessible bytes */
+        if(inBytes!=outBytes) {
+            uprv_memcpy(outBytes, inBytes, length);
+        }
+
+        /* the initial 4 offsets first */
+        tokenStringOffset=ds->readUInt32(((const uint32_t *)inBytes)[0]);
+        groupsOffset=ds->readUInt32(((const uint32_t *)inBytes)[1]);
+        groupStringOffset=ds->readUInt32(((const uint32_t *)inBytes)[2]);
+        ds->swapArray32(ds, inBytes, 16, outBytes, pErrorCode);
+
+        /*
+         * now the tokens table
+         * it needs to be permutated along with the compressed name strings
+         */
+        p=(const uint16_t *)(inBytes+16);
+        q=(uint16_t *)(outBytes+16);
+
+        /* read and swap the tokenCount */
+        tokenCount=ds->readUInt16(*p);
+        ds->swapArray16(ds, p, 2, q, pErrorCode);
+        ++p;
+        ++q;
+
+        /* read the first 512 tokens and make the token maps */
+        if(tokenCount<=512) {
+            count=tokenCount;
+        } else {
+            count=512;
+        }
+        for(i=0; i<count; ++i) {
+            tokens[i]=udata_readInt16(ds, p[i]);
+        }
+        for(; i<512; ++i) {
+            tokens[i]=0; /* fill the rest of the tokens array if tokenCount<512 */
+        }
+        makeTokenMap(ds, tokens, tokenCount, map, pErrorCode);
+        makeTokenMap(ds, tokens+256, (uint16_t)(tokenCount>256 ? tokenCount-256 : 0), trailMap, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+
+        /*
+         * swap and permutate the tokens
+         * go through a temporary array to support in-place swapping
+         */
+        temp=(uint16_t *)uprv_malloc(tokenCount*2);
+        if(temp==NULL) {
+            udata_printError(ds, "out of memory swapping %u unames.icu tokens\n",
+                             tokenCount);
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return 0;
+        }
+
+        /* swap and permutate single-/lead-byte tokens */
+        for(i=0; i<tokenCount && i<256; ++i) {
+            ds->swapArray16(ds, p+i, 2, temp+map[i], pErrorCode);
+        }
+
+        /* swap and permutate trail-byte tokens */
+        for(; i<tokenCount; ++i) {
+            ds->swapArray16(ds, p+i, 2, temp+(i&0xffffff00)+trailMap[i&0xff], pErrorCode);
+        }
+
+        /* copy the result into the output and free the temporary array */
+        uprv_memcpy(q, temp, tokenCount*2);
+        uprv_free(temp);
+
+        /*
+         * swap the token strings but not a possible padding byte after
+         * the terminating NUL of the last string
+         */
+        udata_swapInvStringBlock(ds, inBytes+tokenStringOffset, (int32_t)(groupsOffset-tokenStringOffset),
+                                    outBytes+tokenStringOffset, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            udata_printError(ds, "uchar_swapNames(token strings) failed\n");
+            return 0;
+        }
+
+        /* swap the group table */
+        count=ds->readUInt16(*((const uint16_t *)(inBytes+groupsOffset)));
+        ds->swapArray16(ds, inBytes+groupsOffset, (int32_t)((1+count*3)*2),
+                           outBytes+groupsOffset, pErrorCode);
+
+        /*
+         * swap the group strings
+         * swap the string bytes but not the nibble-encoded string lengths
+         */
+        if(ds->inCharset!=ds->outCharset) {
+            uint16_t offsets[LINES_PER_GROUP+1], lengths[LINES_PER_GROUP+1];
+
+            const uint8_t *inStrings, *nextInStrings;
+            uint8_t *outStrings;
+
+            uint8_t c;
+
+            inStrings=inBytes+groupStringOffset;
+            outStrings=outBytes+groupStringOffset;
+
+            stringsCount=algNamesOffset-groupStringOffset;
+
+            /* iterate through string groups until only a few padding bytes are left */
+            while(stringsCount>32) {
+                nextInStrings=expandGroupLengths(inStrings, offsets, lengths);
+
+                /* move past the length bytes */
+                stringsCount-=(uint32_t)(nextInStrings-inStrings);
+                outStrings+=nextInStrings-inStrings;
+                inStrings=nextInStrings;
+
+                count=offsets[31]+lengths[31]; /* total number of string bytes in this group */
+                stringsCount-=count;
+
+                /* swap the string bytes using map[] and trailMap[] */
+                while(count>0) {
+                    c=*inStrings++;
+                    *outStrings++=map[c];
+                    if(tokens[c]!=-2) {
+                        --count;
+                    } else {
+                        /* token lead byte: swap the trail byte, too */
+                        *outStrings++=trailMap[*inStrings++];
+                        count-=2;
+                    }
+                }
+            }
+        }
+
+        /* swap the algorithmic ranges */
+        offset=algNamesOffset;
+        count=ds->readUInt32(*((const uint32_t *)(inBytes+offset)));
+        ds->swapArray32(ds, inBytes+offset, 4, outBytes+offset, pErrorCode);
+        offset+=4;
+
+        for(i=0; i<count; ++i) {
+            if(offset>(uint32_t)length) {
+                udata_printError(ds, "uchar_swapNames(): too few bytes (%d after header) for unames.icu algorithmic range %u\n",
+                                 length, i);
+                *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+
+            inRange=(const AlgorithmicRange *)(inBytes+offset);
+            outRange=(AlgorithmicRange *)(outBytes+offset);
+            offset+=ds->readUInt16(inRange->size);
+
+            ds->swapArray32(ds, inRange, 8, outRange, pErrorCode);
+            ds->swapArray16(ds, &inRange->size, 2, &outRange->size, pErrorCode);
+            switch(inRange->type) {
+            case 0:
+                /* swap prefix string */
+                ds->swapInvChars(ds, inRange+1, (int32_t)uprv_strlen((const char *)(inRange+1)),
+                                    outRange+1, pErrorCode);
+                if(U_FAILURE(*pErrorCode)) {
+                    udata_printError(ds, "uchar_swapNames(prefix string of algorithmic range %u) failed\n",
+                                     i);
+                    return 0;
+                }
+                break;
+            case 1:
+                {
+                    /* swap factors and the prefix and factor strings */
+                    uint32_t factorsCount;
+
+                    factorsCount=inRange->variant;
+                    p=(const uint16_t *)(inRange+1);
+                    q=(uint16_t *)(outRange+1);
+                    ds->swapArray16(ds, p, (int32_t)(factorsCount*2), q, pErrorCode);
+
+                    /* swap the strings, up to the last terminating NUL */
+                    p+=factorsCount;
+                    q+=factorsCount;
+                    stringsCount=(uint32_t)((inBytes+offset)-(const uint8_t *)p);
+                    while(stringsCount>0 && ((const uint8_t *)p)[stringsCount-1]!=0) {
+                        --stringsCount;
+                    }
+                    ds->swapInvChars(ds, p, (int32_t)stringsCount, q, pErrorCode);
+                }
+                break;
+            default:
+                udata_printError(ds, "uchar_swapNames(): unknown type %u of algorithmic range %u\n",
+                                 inRange->type, i);
+                *pErrorCode=U_UNSUPPORTED_ERROR;
+                return 0;
+            }
+        }
+    }
+
+    return headerSize+(int32_t)offset;
+}
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/unicode/appendable.h b/thirdparty/icu4c/common/unicode/appendable.h
new file mode 100644
index 0000000000..fc99254de1
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/appendable.h
@@ -0,0 +1,239 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  appendable.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010dec07
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __APPENDABLE_H__
+#define __APPENDABLE_H__
+
+/**
+ * \file
+ * \brief C++ API: Appendable class: Sink for Unicode code points and 16-bit code units (char16_ts).
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+
+/**
+ * Base class for objects to which Unicode characters and strings can be appended.
+ * Combines elements of Java Appendable and ICU4C ByteSink.
+ *
+ * This class can be used in APIs where it does not matter whether the actual destination is
+ * a UnicodeString, a char16_t[] array, a UnicodeSet, or any other object
+ * that receives and processes characters and/or strings.
+ *
+ * Implementation classes must implement at least appendCodeUnit(char16_t).
+ * The base class provides default implementations for the other methods.
+ *
+ * The methods do not take UErrorCode parameters.
+ * If an error occurs (e.g., out-of-memory),
+ * in addition to returning false from failing operations,
+ * the implementation must prevent unexpected behavior (e.g., crashes)
+ * from further calls and should make the error condition available separately
+ * (e.g., store a UErrorCode, make/keep a UnicodeString bogus).
+ * @stable ICU 4.8
+ */
+class U_COMMON_API Appendable : public UObject {
+public:
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    ~Appendable();
+
+    /**
+     * Appends a 16-bit code unit.
+     * @param c code unit
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendCodeUnit(char16_t c) = 0;
+
+    /**
+     * Appends a code point.
+     * The default implementation calls appendCodeUnit(char16_t) once or twice.
+     * @param c code point 0..0x10ffff
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendCodePoint(UChar32 c);
+
+    /**
+     * Appends a string.
+     * The default implementation calls appendCodeUnit(char16_t) for each code unit.
+     * @param s string, must not be NULL if length!=0
+     * @param length string length, or -1 if NUL-terminated
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendString(const char16_t *s, int32_t length);
+
+    /**
+     * Tells the object that the caller is going to append roughly
+     * appendCapacity char16_ts. A subclass might use this to pre-allocate
+     * a larger buffer if necessary.
+     * The default implementation does nothing. (It always returns true.)
+     * @param appendCapacity estimated number of char16_ts that will be appended
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool reserveAppendCapacity(int32_t appendCapacity);
+
+    /**
+     * Returns a writable buffer for appending and writes the buffer's capacity to
+     * *resultCapacity. Guarantees *resultCapacity>=minCapacity.
+     * May return a pointer to the caller-owned scratch buffer which must have
+     * scratchCapacity>=minCapacity.
+     * The returned buffer is only valid until the next operation
+     * on this Appendable.
+     *
+     * After writing at most *resultCapacity char16_ts, call appendString() with the
+     * pointer returned from this function and the number of char16_ts written.
+     * Many appendString() implementations will avoid copying char16_ts if this function
+     * returned an internal buffer.
+     *
+     * Partial usage example:
+     * \code
+     *  int32_t capacity;
+     *  char16_t* buffer = app.getAppendBuffer(..., &capacity);
+     *  ... Write n char16_ts into buffer, with n <= capacity.
+     *  app.appendString(buffer, n);
+     * \endcode
+     * In many implementations, that call to append will avoid copying char16_ts.
+     *
+     * If the Appendable allocates or reallocates an internal buffer, it should use
+     * the desiredCapacityHint if appropriate.
+     * If a caller cannot provide a reasonable guess at the desired capacity,
+     * it should pass desiredCapacityHint=0.
+     *
+     * If a non-scratch buffer is returned, the caller may only pass
+     * a prefix to it to appendString().
+     * That is, it is not correct to pass an interior pointer to appendString().
+     *
+     * The default implementation always returns the scratch buffer.
+     *
+     * @param minCapacity required minimum capacity of the returned buffer;
+     *                    must be non-negative
+     * @param desiredCapacityHint desired capacity of the returned buffer;
+     *                            must be non-negative
+     * @param scratch default caller-owned buffer
+     * @param scratchCapacity capacity of the scratch buffer
+     * @param resultCapacity pointer to an integer which will be set to the
+     *                       capacity of the returned buffer
+     * @return a buffer with *resultCapacity>=minCapacity
+     * @stable ICU 4.8
+     */
+    virtual char16_t *getAppendBuffer(int32_t minCapacity,
+                                   int32_t desiredCapacityHint,
+                                   char16_t *scratch, int32_t scratchCapacity,
+                                   int32_t *resultCapacity);
+};
+
+/**
+ * An Appendable implementation which writes to a UnicodeString.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API UnicodeStringAppendable : public Appendable {
+public:
+    /**
+     * Aliases the UnicodeString (keeps its reference) for writing.
+     * @param s The UnicodeString to which this Appendable will write.
+     * @stable ICU 4.8
+     */
+    explicit UnicodeStringAppendable(UnicodeString &s) : str(s) {}
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    ~UnicodeStringAppendable();
+
+    /**
+     * Appends a 16-bit code unit to the string.
+     * @param c code unit
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendCodeUnit(char16_t c);
+
+    /**
+     * Appends a code point to the string.
+     * @param c code point 0..0x10ffff
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendCodePoint(UChar32 c);
+
+    /**
+     * Appends a string to the UnicodeString.
+     * @param s string, must not be NULL if length!=0
+     * @param length string length, or -1 if NUL-terminated
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool appendString(const char16_t *s, int32_t length);
+
+    /**
+     * Tells the UnicodeString that the caller is going to append roughly
+     * appendCapacity char16_ts.
+     * @param appendCapacity estimated number of char16_ts that will be appended
+     * @return true if the operation succeeded
+     * @stable ICU 4.8
+     */
+    virtual UBool reserveAppendCapacity(int32_t appendCapacity);
+
+    /**
+     * Returns a writable buffer for appending and writes the buffer's capacity to
+     * *resultCapacity. Guarantees *resultCapacity>=minCapacity.
+     * May return a pointer to the caller-owned scratch buffer which must have
+     * scratchCapacity>=minCapacity.
+     * The returned buffer is only valid until the next write operation
+     * on the UnicodeString.
+     *
+     * For details see Appendable::getAppendBuffer().
+     *
+     * @param minCapacity required minimum capacity of the returned buffer;
+     *                    must be non-negative
+     * @param desiredCapacityHint desired capacity of the returned buffer;
+     *                            must be non-negative
+     * @param scratch default caller-owned buffer
+     * @param scratchCapacity capacity of the scratch buffer
+     * @param resultCapacity pointer to an integer which will be set to the
+     *                       capacity of the returned buffer
+     * @return a buffer with *resultCapacity>=minCapacity
+     * @stable ICU 4.8
+     */
+    virtual char16_t *getAppendBuffer(int32_t minCapacity,
+                                   int32_t desiredCapacityHint,
+                                   char16_t *scratch, int32_t scratchCapacity,
+                                   int32_t *resultCapacity);
+
+private:
+    UnicodeString &str;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __APPENDABLE_H__
diff --git a/thirdparty/icu4c/common/unicode/brkiter.h b/thirdparty/icu4c/common/unicode/brkiter.h
new file mode 100644
index 0000000000..9bba5fcccc
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/brkiter.h
@@ -0,0 +1,670 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*
+* File brkiter.h
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   02/18/97    aliu        Added typedef for TextCount.  Made DONE const.
+*   05/07/97    aliu        Fixed DLL declaration.
+*   07/09/97    jfitz       Renamed BreakIterator and interface synced with JDK
+*   08/11/98    helena      Sync-up JDK1.2.
+*   01/13/2000  helena      Added UErrorCode parameter to createXXXInstance methods.
+********************************************************************************
+*/
+
+#ifndef BRKITER_H
+#define BRKITER_H
+
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief C++ API: Break Iterator.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#if UCONFIG_NO_BREAK_ITERATION
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Allow the declaration of APIs with pointers to BreakIterator
+ * even when break iteration is removed from the build.
+ */
+class BreakIterator;
+
+U_NAMESPACE_END
+
+#else
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+#include "unicode/chariter.h"
+#include "unicode/locid.h"
+#include "unicode/ubrk.h"
+#include "unicode/strenum.h"
+#include "unicode/utext.h"
+#include "unicode/umisc.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * The BreakIterator class implements methods for finding the location
+ * of boundaries in text. BreakIterator is an abstract base class.
+ * Instances of BreakIterator maintain a current position and scan over
+ * text returning the index of characters where boundaries occur.
+ * <p>
+ * Line boundary analysis determines where a text string can be broken
+ * when line-wrapping. The mechanism correctly handles punctuation and
+ * hyphenated words.
+ * <p>
+ * Sentence boundary analysis allows selection with correct
+ * interpretation of periods within numbers and abbreviations, and
+ * trailing punctuation marks such as quotation marks and parentheses.
+ * <p>
+ * Word boundary analysis is used by search and replace functions, as
+ * well as within text editing applications that allow the user to
+ * select words with a double click. Word selection provides correct
+ * interpretation of punctuation marks within and following
+ * words. Characters that are not part of a word, such as symbols or
+ * punctuation marks, have word-breaks on both sides.
+ * <p>
+ * Character boundary analysis allows users to interact with
+ * characters as they expect to, for example, when moving the cursor
+ * through a text string. Character boundary analysis provides correct
+ * navigation of through character strings, regardless of how the
+ * character is stored.  For example, an accented character might be
+ * stored as a base character and a diacritical mark. What users
+ * consider to be a character can differ between languages.
+ * <p>
+ * The text boundary positions are found according to the rules
+ * described in Unicode Standard Annex #29, Text Boundaries, and
+ * Unicode Standard Annex #14, Line Breaking Properties.  These
+ * are available at http://www.unicode.org/reports/tr14/ and
+ * http://www.unicode.org/reports/tr29/.
+ * <p>
+ * In addition to the C++ API defined in this header file, a
+ * plain C API with equivalent functionality is defined in the
+ * file ubrk.h
+ * <p>
+ * Code snippets illustrating the use of the Break Iterator APIs
+ * are available in the ICU User Guide,
+ * http://icu-project.org/userguide/boundaryAnalysis.html
+ * and in the sample program icu/source/samples/break/break.cpp
+ *
+ */
+class U_COMMON_API BreakIterator : public UObject {
+public:
+    /**
+     *  destructor
+     *  @stable ICU 2.0
+     */
+    virtual ~BreakIterator();
+
+    /**
+     * Return true if another object is semantically equal to this
+     * one. The other object should be an instance of the same subclass of
+     * BreakIterator. Objects of different subclasses are considered
+     * unequal.
+     * <P>
+     * Return true if this BreakIterator is at the same position in the
+     * same text, and is the same class and type (word, line, etc.) of
+     * BreakIterator, as the argument.  Text is considered the same if
+     * it contains the same characters, it need not be the same
+     * object, and styles are not considered.
+     * @stable ICU 2.0
+     */
+    virtual UBool operator==(const BreakIterator&) const = 0;
+
+    /**
+     * Returns the complement of the result of operator==
+     * @param rhs The BreakIterator to be compared for inequality
+     * @return the complement of the result of operator==
+     * @stable ICU 2.0
+     */
+    UBool operator!=(const BreakIterator& rhs) const { return !operator==(rhs); }
+
+    /**
+     * Return a polymorphic copy of this object.  This is an abstract
+     * method which subclasses implement.
+     * @stable ICU 2.0
+     */
+    virtual BreakIterator* clone() const = 0;
+
+    /**
+     * Return a polymorphic class ID for this object. Different subclasses
+     * will return distinct unequal values.
+     * @stable ICU 2.0
+     */
+    virtual UClassID getDynamicClassID(void) const = 0;
+
+    /**
+     * Return a CharacterIterator over the text being analyzed.
+     * @stable ICU 2.0
+     */
+    virtual CharacterIterator& getText(void) const = 0;
+
+
+    /**
+      *  Get a UText for the text being analyzed.
+      *  The returned UText is a shallow clone of the UText used internally
+      *  by the break iterator implementation.  It can safely be used to
+      *  access the text without impacting any break iterator operations,
+      *  but the underlying text itself must not be altered.
+      *
+      * @param fillIn A UText to be filled in.  If NULL, a new UText will be
+      *           allocated to hold the result.
+      * @param status receives any error codes.
+      * @return   The current UText for this break iterator.  If an input
+      *           UText was provided, it will always be returned.
+      * @stable ICU 3.4
+      */
+     virtual UText *getUText(UText *fillIn, UErrorCode &status) const = 0;
+
+    /**
+     * Change the text over which this operates. The text boundary is
+     * reset to the start.
+     *
+     * The BreakIterator will retain a reference to the supplied string.
+     * The caller must not modify or delete the text while the BreakIterator
+     * retains the reference.
+     *
+     * @param text The UnicodeString used to change the text.
+     * @stable ICU 2.0
+     */
+    virtual void  setText(const UnicodeString &text) = 0;
+
+    /**
+     * Reset the break iterator to operate over the text represented by
+     * the UText.  The iterator position is reset to the start.
+     *
+     * This function makes a shallow clone of the supplied UText.  This means
+     * that the caller is free to immediately close or otherwise reuse the
+     * Utext that was passed as a parameter, but that the underlying text itself
+     * must not be altered while being referenced by the break iterator.
+     *
+     * All index positions returned by break iterator functions are
+     * native indices from the UText. For example, when breaking UTF-8
+     * encoded text, the break positions returned by next(), previous(), etc.
+     * will be UTF-8 string indices, not UTF-16 positions.
+     *
+     * @param text The UText used to change the text.
+     * @param status receives any error codes.
+     * @stable ICU 3.4
+     */
+    virtual void  setText(UText *text, UErrorCode &status) = 0;
+
+    /**
+     * Change the text over which this operates. The text boundary is
+     * reset to the start.
+     * Note that setText(UText *) provides similar functionality to this function,
+     * and is more efficient.
+     * @param it The CharacterIterator used to change the text.
+     * @stable ICU 2.0
+     */
+    virtual void  adoptText(CharacterIterator* it) = 0;
+
+    enum {
+        /**
+         * DONE is returned by previous() and next() after all valid
+         * boundaries have been returned.
+         * @stable ICU 2.0
+         */
+        DONE = (int32_t)-1
+    };
+
+    /**
+     * Sets the current iteration position to the beginning of the text, position zero.
+     * @return The offset of the beginning of the text, zero.
+     * @stable ICU 2.0
+     */
+    virtual int32_t first(void) = 0;
+
+    /**
+     * Set the iterator position to the index immediately BEYOND the last character in the text being scanned.
+     * @return The index immediately BEYOND the last character in the text being scanned.
+     * @stable ICU 2.0
+     */
+    virtual int32_t last(void) = 0;
+
+    /**
+     * Set the iterator position to the boundary preceding the current boundary.
+     * @return The character index of the previous text boundary or DONE if all
+     * boundaries have been returned.
+     * @stable ICU 2.0
+     */
+    virtual int32_t previous(void) = 0;
+
+    /**
+     * Advance the iterator to the boundary following the current boundary.
+     * @return The character index of the next text boundary or DONE if all
+     * boundaries have been returned.
+     * @stable ICU 2.0
+     */
+    virtual int32_t next(void) = 0;
+
+    /**
+     * Return character index of the current iterator position within the text.
+     * @return The boundary most recently returned.
+     * @stable ICU 2.0
+     */
+    virtual int32_t current(void) const = 0;
+
+    /**
+     * Advance the iterator to the first boundary following the specified offset.
+     * The value returned is always greater than the offset or
+     * the value BreakIterator.DONE
+     * @param offset the offset to begin scanning.
+     * @return The first boundary after the specified offset.
+     * @stable ICU 2.0
+     */
+    virtual int32_t following(int32_t offset) = 0;
+
+    /**
+     * Set the iterator position to the first boundary preceding the specified offset.
+     * The value returned is always smaller than the offset or
+     * the value BreakIterator.DONE
+     * @param offset the offset to begin scanning.
+     * @return The first boundary before the specified offset.
+     * @stable ICU 2.0
+     */
+    virtual int32_t preceding(int32_t offset) = 0;
+
+    /**
+     * Return true if the specified position is a boundary position.
+     * As a side effect, the current position of the iterator is set
+     * to the first boundary position at or following the specified offset.
+     * @param offset the offset to check.
+     * @return True if "offset" is a boundary position.
+     * @stable ICU 2.0
+     */
+    virtual UBool isBoundary(int32_t offset) = 0;
+
+    /**
+     * Set the iterator position to the nth boundary from the current boundary
+     * @param n the number of boundaries to move by.  A value of 0
+     * does nothing.  Negative values move to previous boundaries
+     * and positive values move to later boundaries.
+     * @return The new iterator position, or
+     * DONE if there are fewer than |n| boundaries in the specified direction.
+     * @stable ICU 2.0
+     */
+    virtual int32_t next(int32_t n) = 0;
+
+   /**
+     * For RuleBasedBreakIterators, return the status tag from the break rule
+     * that determined the boundary at the current iteration position.
+     * <p>
+     * For break iterator types that do not support a rule status,
+     * a default value of 0 is returned.
+     * <p>
+     * @return the status from the break rule that determined the boundary at
+     *         the current iteration position.
+     * @see RuleBaseBreakIterator::getRuleStatus()
+     * @see UWordBreak
+     * @stable ICU 52
+     */
+    virtual int32_t getRuleStatus() const;
+
+   /**
+    * For RuleBasedBreakIterators, get the status (tag) values from the break rule(s)
+    * that determined the boundary at the current iteration position.
+    * <p>
+    * For break iterator types that do not support rule status,
+    * no values are returned.
+    * <p>
+    * The returned status value(s) are stored into an array provided by the caller.
+    * The values are stored in sorted (ascending) order.
+    * If the capacity of the output array is insufficient to hold the data,
+    *  the output will be truncated to the available length, and a
+    *  U_BUFFER_OVERFLOW_ERROR will be signaled.
+    * <p>
+    * @see RuleBaseBreakIterator::getRuleStatusVec
+    *
+    * @param fillInVec an array to be filled in with the status values.
+    * @param capacity  the length of the supplied vector.  A length of zero causes
+    *                  the function to return the number of status values, in the
+    *                  normal way, without attempting to store any values.
+    * @param status    receives error codes.
+    * @return          The number of rule status values from rules that determined
+    *                  the boundary at the current iteration position.
+    *                  In the event of a U_BUFFER_OVERFLOW_ERROR, the return value
+    *                  is the total number of status values that were available,
+    *                  not the reduced number that were actually returned.
+    * @see getRuleStatus
+    * @stable ICU 52
+    */
+    virtual int32_t getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status);
+
+    /**
+     * Create BreakIterator for word-breaks using the given locale.
+     * Returns an instance of a BreakIterator implementing word breaks.
+     * WordBreak is useful for word selection (ex. double click)
+     * @param where the locale.
+     * @param status the error code
+     * @return A BreakIterator for word-breaks.  The UErrorCode& status
+     * parameter is used to return status information to the user.
+     * To check whether the construction succeeded or not, you should check
+     * the value of U_SUCCESS(err).  If you wish more detailed information, you
+     * can check for informational error results which still indicate success.
+     * U_USING_FALLBACK_WARNING indicates that a fall back locale was used.  For
+     * example, 'de_CH' was requested, but nothing was found there, so 'de' was
+     * used.  U_USING_DEFAULT_WARNING indicates that the default locale data was
+     * used; neither the requested locale nor any of its fall back locales
+     * could be found.
+     * The caller owns the returned object and is responsible for deleting it.
+     * @stable ICU 2.0
+     */
+    static BreakIterator* U_EXPORT2
+    createWordInstance(const Locale& where, UErrorCode& status);
+
+    /**
+     * Create BreakIterator for line-breaks using specified locale.
+     * Returns an instance of a BreakIterator implementing line breaks. Line
+     * breaks are logically possible line breaks, actual line breaks are
+     * usually determined based on display width.
+     * LineBreak is useful for word wrapping text.
+     * @param where the locale.
+     * @param status The error code.
+     * @return A BreakIterator for line-breaks.  The UErrorCode& status
+     * parameter is used to return status information to the user.
+     * To check whether the construction succeeded or not, you should check
+     * the value of U_SUCCESS(err).  If you wish more detailed information, you
+     * can check for informational error results which still indicate success.
+     * U_USING_FALLBACK_WARNING indicates that a fall back locale was used.  For
+     * example, 'de_CH' was requested, but nothing was found there, so 'de' was
+     * used.  U_USING_DEFAULT_WARNING indicates that the default locale data was
+     * used; neither the requested locale nor any of its fall back locales
+     * could be found.
+     * The caller owns the returned object and is responsible for deleting it.
+     * @stable ICU 2.0
+     */
+    static BreakIterator* U_EXPORT2
+    createLineInstance(const Locale& where, UErrorCode& status);
+
+    /**
+     * Create BreakIterator for character-breaks using specified locale
+     * Returns an instance of a BreakIterator implementing character breaks.
+     * Character breaks are boundaries of combining character sequences.
+     * @param where the locale.
+     * @param status The error code.
+     * @return A BreakIterator for character-breaks.  The UErrorCode& status
+     * parameter is used to return status information to the user.
+     * To check whether the construction succeeded or not, you should check
+     * the value of U_SUCCESS(err).  If you wish more detailed information, you
+     * can check for informational error results which still indicate success.
+     * U_USING_FALLBACK_WARNING indicates that a fall back locale was used.  For
+     * example, 'de_CH' was requested, but nothing was found there, so 'de' was
+     * used.  U_USING_DEFAULT_WARNING indicates that the default locale data was
+     * used; neither the requested locale nor any of its fall back locales
+     * could be found.
+     * The caller owns the returned object and is responsible for deleting it.
+     * @stable ICU 2.0
+     */
+    static BreakIterator* U_EXPORT2
+    createCharacterInstance(const Locale& where, UErrorCode& status);
+
+    /**
+     * Create BreakIterator for sentence-breaks using specified locale
+     * Returns an instance of a BreakIterator implementing sentence breaks.
+     * @param where the locale.
+     * @param status The error code.
+     * @return A BreakIterator for sentence-breaks.  The UErrorCode& status
+     * parameter is used to return status information to the user.
+     * To check whether the construction succeeded or not, you should check
+     * the value of U_SUCCESS(err).  If you wish more detailed information, you
+     * can check for informational error results which still indicate success.
+     * U_USING_FALLBACK_WARNING indicates that a fall back locale was used.  For
+     * example, 'de_CH' was requested, but nothing was found there, so 'de' was
+     * used.  U_USING_DEFAULT_WARNING indicates that the default locale data was
+     * used; neither the requested locale nor any of its fall back locales
+     * could be found.
+     * The caller owns the returned object and is responsible for deleting it.
+     * @stable ICU 2.0
+     */
+    static BreakIterator* U_EXPORT2
+    createSentenceInstance(const Locale& where, UErrorCode& status);
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Create BreakIterator for title-casing breaks using the specified locale
+     * Returns an instance of a BreakIterator implementing title breaks.
+     * The iterator returned locates title boundaries as described for
+     * Unicode 3.2 only. For Unicode 4.0 and above title boundary iteration,
+     * please use a word boundary iterator. See {@link #createWordInstance }.
+     *
+     * @param where the locale.
+     * @param status The error code.
+     * @return A BreakIterator for title-breaks.  The UErrorCode& status
+     * parameter is used to return status information to the user.
+     * To check whether the construction succeeded or not, you should check
+     * the value of U_SUCCESS(err).  If you wish more detailed information, you
+     * can check for informational error results which still indicate success.
+     * U_USING_FALLBACK_WARNING indicates that a fall back locale was used.  For
+     * example, 'de_CH' was requested, but nothing was found there, so 'de' was
+     * used.  U_USING_DEFAULT_WARNING indicates that the default locale data was
+     * used; neither the requested locale nor any of its fall back locales
+     * could be found.
+     * The caller owns the returned object and is responsible for deleting it.
+     * @deprecated ICU 64 Use createWordInstance instead.
+     */
+    static BreakIterator* U_EXPORT2
+    createTitleInstance(const Locale& where, UErrorCode& status);
+#endif /* U_HIDE_DEPRECATED_API */
+
+    /**
+     * Get the set of Locales for which TextBoundaries are installed.
+     * <p><b>Note:</b> this will not return locales added through the register
+     * call. To see the registered locales too, use the getAvailableLocales
+     * function that returns a StringEnumeration object </p>
+     * @param count the output parameter of number of elements in the locale list
+     * @return available locales
+     * @stable ICU 2.0
+     */
+    static const Locale* U_EXPORT2 getAvailableLocales(int32_t& count);
+
+    /**
+     * Get name of the object for the desired Locale, in the desired language.
+     * @param objectLocale must be from getAvailableLocales.
+     * @param displayLocale specifies the desired locale for output.
+     * @param name the fill-in parameter of the return value
+     * Uses best match.
+     * @return user-displayable name
+     * @stable ICU 2.0
+     */
+    static UnicodeString& U_EXPORT2 getDisplayName(const Locale& objectLocale,
+                                         const Locale& displayLocale,
+                                         UnicodeString& name);
+
+    /**
+     * Get name of the object for the desired Locale, in the language of the
+     * default locale.
+     * @param objectLocale must be from getMatchingLocales
+     * @param name the fill-in parameter of the return value
+     * @return user-displayable name
+     * @stable ICU 2.0
+     */
+    static UnicodeString& U_EXPORT2 getDisplayName(const Locale& objectLocale,
+                                         UnicodeString& name);
+
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+    /**
+     * Deprecated functionality. Use clone() instead.
+     *
+     * Thread safe client-buffer-based cloning operation
+     *    Do NOT call delete on a safeclone, since 'new' is not used to create it.
+     * @param stackBuffer user allocated space for the new clone. If NULL new memory will be allocated.
+     * If buffer is not large enough, new memory will be allocated.
+     * @param BufferSize reference to size of allocated space.
+     * If BufferSize == 0, a sufficient size for use in cloning will
+     * be returned ('pre-flighting')
+     * If BufferSize is not enough for a stack-based safe clone,
+     * new memory will be allocated.
+     * @param status to indicate whether the operation went on smoothly or there were errors
+     *  An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any allocations were
+     *  necessary.
+     * @return pointer to the new clone
+     *
+     * @deprecated ICU 52. Use clone() instead.
+     */
+    virtual BreakIterator *  createBufferClone(void *stackBuffer,
+                                               int32_t &BufferSize,
+                                               UErrorCode &status) = 0;
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+#ifndef U_HIDE_DEPRECATED_API
+
+    /**
+     *   Determine whether the BreakIterator was created in user memory by
+     *   createBufferClone(), and thus should not be deleted.  Such objects
+     *   must be closed by an explicit call to the destructor (not delete).
+     * @deprecated ICU 52. Always delete the BreakIterator.
+     */
+    inline UBool isBufferClone(void);
+
+#endif /* U_HIDE_DEPRECATED_API */
+
+#if !UCONFIG_NO_SERVICE
+    /**
+     * Register a new break iterator of the indicated kind, to use in the given locale.
+     * The break iterator will be adopted.  Clones of the iterator will be returned
+     * if a request for a break iterator of the given kind matches or falls back to
+     * this locale.
+     * Because ICU may choose to cache BreakIterators internally, this must
+     * be called at application startup, prior to any calls to
+     * BreakIterator::createXXXInstance to avoid undefined behavior.
+     * @param toAdopt the BreakIterator instance to be adopted
+     * @param locale the Locale for which this instance is to be registered
+     * @param kind the type of iterator for which this instance is to be registered
+     * @param status the in/out status code, no special meanings are assigned
+     * @return a registry key that can be used to unregister this instance
+     * @stable ICU 2.4
+     */
+    static URegistryKey U_EXPORT2 registerInstance(BreakIterator* toAdopt,
+                                        const Locale& locale,
+                                        UBreakIteratorType kind,
+                                        UErrorCode& status);
+
+    /**
+     * Unregister a previously-registered BreakIterator using the key returned from the
+     * register call.  Key becomes invalid after a successful call and should not be used again.
+     * The BreakIterator corresponding to the key will be deleted.
+     * Because ICU may choose to cache BreakIterators internally, this should
+     * be called during application shutdown, after all calls to
+     * BreakIterator::createXXXInstance to avoid undefined behavior.
+     * @param key the registry key returned by a previous call to registerInstance
+     * @param status the in/out status code, no special meanings are assigned
+     * @return true if the iterator for the key was successfully unregistered
+     * @stable ICU 2.4
+     */
+    static UBool U_EXPORT2 unregister(URegistryKey key, UErrorCode& status);
+
+    /**
+     * Return a StringEnumeration over the locales available at the time of the call,
+     * including registered locales.
+     * @return a StringEnumeration over the locales available at the time of the call
+     * @stable ICU 2.4
+     */
+    static StringEnumeration* U_EXPORT2 getAvailableLocales(void);
+#endif
+
+    /**
+     * Returns the locale for this break iterator. Two flavors are available: valid and
+     * actual locale.
+     * @stable ICU 2.8
+     */
+    Locale getLocale(ULocDataLocaleType type, UErrorCode& status) const;
+
+#ifndef U_HIDE_INTERNAL_API
+    /** Get the locale for this break iterator object. You can choose between valid and actual locale.
+     *  @param type type of the locale we're looking for (valid or actual)
+     *  @param status error code for the operation
+     *  @return the locale
+     *  @internal
+     */
+    const char *getLocaleID(ULocDataLocaleType type, UErrorCode& status) const;
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /**
+     *  Set the subject text string upon which the break iterator is operating
+     *  without changing any other aspect of the matching state.
+     *  The new and previous text strings must have the same content.
+     *
+     *  This function is intended for use in environments where ICU is operating on
+     *  strings that may move around in memory.  It provides a mechanism for notifying
+     *  ICU that the string has been relocated, and providing a new UText to access the
+     *  string in its new position.
+     *
+     *  Note that the break iterator implementation never copies the underlying text
+     *  of a string being processed, but always operates directly on the original text
+     *  provided by the user. Refreshing simply drops the references to the old text
+     *  and replaces them with references to the new.
+     *
+     *  Caution:  this function is normally used only by very specialized,
+     *  system-level code.  One example use case is with garbage collection that moves
+     *  the text in memory.
+     *
+     * @param input      The new (moved) text string.
+     * @param status     Receives errors detected by this function.
+     * @return           *this
+     *
+     * @stable ICU 49
+     */
+    virtual BreakIterator &refreshInputText(UText *input, UErrorCode &status) = 0;
+
+ private:
+    static BreakIterator* buildInstance(const Locale& loc, const char *type, UErrorCode& status);
+    static BreakIterator* createInstance(const Locale& loc, int32_t kind, UErrorCode& status);
+    static BreakIterator* makeInstance(const Locale& loc, int32_t kind, UErrorCode& status);
+
+    friend class ICUBreakIteratorFactory;
+    friend class ICUBreakIteratorService;
+
+protected:
+    // Do not enclose protected default/copy constructors with #ifndef U_HIDE_INTERNAL_API
+    // or else the compiler will create a public ones.
+    /** @internal */
+    BreakIterator();
+    /** @internal */
+    BreakIterator (const BreakIterator &other);
+#ifndef U_HIDE_INTERNAL_API
+    /** @internal */
+    BreakIterator (const Locale& valid, const Locale &actual);
+    /** @internal. Assignment Operator, used by RuleBasedBreakIterator. */
+    BreakIterator &operator = (const BreakIterator &other);
+#endif  /* U_HIDE_INTERNAL_API */
+
+private:
+
+    /** @internal (private) */
+    char actualLocale[ULOC_FULLNAME_CAPACITY];
+    char validLocale[ULOC_FULLNAME_CAPACITY];
+};
+
+#ifndef U_HIDE_DEPRECATED_API
+
+inline UBool BreakIterator::isBufferClone()
+{
+    return false;
+}
+
+#endif /* U_HIDE_DEPRECATED_API */
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif // BRKITER_H
+//eof
diff --git a/thirdparty/icu4c/common/unicode/bytestream.h b/thirdparty/icu4c/common/unicode/bytestream.h
new file mode 100644
index 0000000000..044f7a77e7
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/bytestream.h
@@ -0,0 +1,309 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+// Copyright (C) 2009-2012, International Business Machines
+// Corporation and others. All Rights Reserved.
+//
+// Copyright 2007 Google Inc. All Rights Reserved.
+// Author: sanjay@google.com (Sanjay Ghemawat)
+//
+// Abstract interface that consumes a sequence of bytes (ByteSink).
+//
+// Used so that we can write a single piece of code that can operate
+// on a variety of output string types.
+//
+// Various implementations of this interface are provided:
+//   ByteSink:
+//      CheckedArrayByteSink    Write to a flat array, with bounds checking
+//      StringByteSink          Write to an STL string
+
+// This code is a contribution of Google code, and the style used here is
+// a compromise between the original Google code and the ICU coding guidelines.
+// For example, data types are ICU-ified (size_t,int->int32_t),
+// and API comments doxygen-ified, but function names and behavior are
+// as in the original, if possible.
+// Assertion-style error handling, not available in ICU, was changed to
+// parameter "pinning" similar to UnicodeString.
+//
+// In addition, this is only a partial port of the original Google code,
+// limited to what was needed so far. The (nearly) complete original code
+// is in the ICU svn repository at icuhtml/trunk/design/strings/contrib
+// (see ICU ticket 6765, r25517).
+
+#ifndef __BYTESTREAM_H__
+#define __BYTESTREAM_H__
+
+/**
+ * \file
+ * \brief C++ API: Interface for writing bytes, and implementation classes.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+#include "unicode/std_string.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * A ByteSink can be filled with bytes.
+ * @stable ICU 4.2
+ */
+class U_COMMON_API ByteSink : public UMemory {
+public:
+  /**
+   * Default constructor.
+   * @stable ICU 4.2
+   */
+  ByteSink() { }
+  /**
+   * Virtual destructor.
+   * @stable ICU 4.2
+   */
+  virtual ~ByteSink();
+
+  /**
+   * Append "bytes[0,n-1]" to this.
+   * @param bytes the pointer to the bytes
+   * @param n the number of bytes; must be non-negative
+   * @stable ICU 4.2
+   */
+  virtual void Append(const char* bytes, int32_t n) = 0;
+
+#ifndef U_HIDE_DRAFT_API
+  /**
+   * Appends n bytes to this. Same as Append().
+   * Call AppendU8() with u8"string literals" which are const char * in C++11
+   * but const char8_t * in C++20.
+   * If the compiler does support char8_t as a distinct type,
+   * then an AppendU8() overload for that is defined and will be chosen.
+   *
+   * @param bytes the pointer to the bytes
+   * @param n the number of bytes; must be non-negative
+   * @draft ICU 67
+   */
+  inline void AppendU8(const char* bytes, int32_t n) {
+    Append(bytes, n);
+  }
+
+#if defined(__cpp_char8_t) || defined(U_IN_DOXYGEN)
+  /**
+   * Appends n bytes to this. Same as Append() but for a const char8_t * pointer.
+   * Call AppendU8() with u8"string literals" which are const char * in C++11
+   * but const char8_t * in C++20.
+   * If the compiler does support char8_t as a distinct type,
+   * then this AppendU8() overload for that is defined and will be chosen.
+   *
+   * @param bytes the pointer to the bytes
+   * @param n the number of bytes; must be non-negative
+   * @draft ICU 67
+   */
+  inline void AppendU8(const char8_t* bytes, int32_t n) {
+    Append(reinterpret_cast<const char*>(bytes), n);
+  }
+#endif
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Returns a writable buffer for appending and writes the buffer's capacity to
+   * *result_capacity. Guarantees *result_capacity>=min_capacity.
+   * May return a pointer to the caller-owned scratch buffer which must have
+   * scratch_capacity>=min_capacity.
+   * The returned buffer is only valid until the next operation
+   * on this ByteSink.
+   *
+   * After writing at most *result_capacity bytes, call Append() with the
+   * pointer returned from this function and the number of bytes written.
+   * Many Append() implementations will avoid copying bytes if this function
+   * returned an internal buffer.
+   *
+   * Partial usage example:
+   *  int32_t capacity;
+   *  char* buffer = sink->GetAppendBuffer(..., &capacity);
+   *  ... Write n bytes into buffer, with n <= capacity.
+   *  sink->Append(buffer, n);
+   * In many implementations, that call to Append will avoid copying bytes.
+   *
+   * If the ByteSink allocates or reallocates an internal buffer, it should use
+   * the desired_capacity_hint if appropriate.
+   * If a caller cannot provide a reasonable guess at the desired capacity,
+   * it should pass desired_capacity_hint=0.
+   *
+   * If a non-scratch buffer is returned, the caller may only pass
+   * a prefix to it to Append().
+   * That is, it is not correct to pass an interior pointer to Append().
+   *
+   * The default implementation always returns the scratch buffer.
+   *
+   * @param min_capacity required minimum capacity of the returned buffer;
+   *                     must be non-negative
+   * @param desired_capacity_hint desired capacity of the returned buffer;
+   *                              must be non-negative
+   * @param scratch default caller-owned buffer
+   * @param scratch_capacity capacity of the scratch buffer
+   * @param result_capacity pointer to an integer which will be set to the
+   *                        capacity of the returned buffer
+   * @return a buffer with *result_capacity>=min_capacity
+   * @stable ICU 4.2
+   */
+  virtual char* GetAppendBuffer(int32_t min_capacity,
+                                int32_t desired_capacity_hint,
+                                char* scratch, int32_t scratch_capacity,
+                                int32_t* result_capacity);
+
+  /**
+   * Flush internal buffers.
+   * Some byte sinks use internal buffers or provide buffering
+   * and require calling Flush() at the end of the stream.
+   * The ByteSink should be ready for further Append() calls after Flush().
+   * The default implementation of Flush() does nothing.
+   * @stable ICU 4.2
+   */
+  virtual void Flush();
+
+private:
+  ByteSink(const ByteSink &) = delete;
+  ByteSink &operator=(const ByteSink &) = delete;
+};
+
+// -------------------------------------------------------------
+// Some standard implementations
+
+/** 
+ * Implementation of ByteSink that writes to a flat byte array,
+ * with bounds-checking:
+ * This sink will not write more than capacity bytes to outbuf.
+ * If more than capacity bytes are Append()ed, then excess bytes are ignored,
+ * and Overflowed() will return true.
+ * Overflow does not cause a runtime error.
+ * @stable ICU 4.2
+ */
+class U_COMMON_API CheckedArrayByteSink : public ByteSink {
+public:
+  /**
+   * Constructs a ByteSink that will write to outbuf[0..capacity-1].
+   * @param outbuf buffer to write to
+   * @param capacity size of the buffer
+   * @stable ICU 4.2
+   */
+  CheckedArrayByteSink(char* outbuf, int32_t capacity);
+  /**
+   * Destructor.
+   * @stable ICU 4.2
+   */
+  virtual ~CheckedArrayByteSink();
+  /**
+   * Returns the sink to its original state, without modifying the buffer.
+   * Useful for reusing both the buffer and the sink for multiple streams.
+   * Resets the state to NumberOfBytesWritten()=NumberOfBytesAppended()=0
+   * and Overflowed()=false.
+   * @return *this
+   * @stable ICU 4.6
+   */
+  virtual CheckedArrayByteSink& Reset();
+  /**
+   * Append "bytes[0,n-1]" to this.
+   * @param bytes the pointer to the bytes
+   * @param n the number of bytes; must be non-negative
+   * @stable ICU 4.2
+   */
+  virtual void Append(const char* bytes, int32_t n);
+  /**
+   * Returns a writable buffer for appending and writes the buffer's capacity to
+   * *result_capacity. For details see the base class documentation.
+   * @param min_capacity required minimum capacity of the returned buffer;
+   *                     must be non-negative
+   * @param desired_capacity_hint desired capacity of the returned buffer;
+   *                              must be non-negative
+   * @param scratch default caller-owned buffer
+   * @param scratch_capacity capacity of the scratch buffer
+   * @param result_capacity pointer to an integer which will be set to the
+   *                        capacity of the returned buffer
+   * @return a buffer with *result_capacity>=min_capacity
+   * @stable ICU 4.2
+   */
+  virtual char* GetAppendBuffer(int32_t min_capacity,
+                                int32_t desired_capacity_hint,
+                                char* scratch, int32_t scratch_capacity,
+                                int32_t* result_capacity);
+  /**
+   * Returns the number of bytes actually written to the sink.
+   * @return number of bytes written to the buffer
+   * @stable ICU 4.2
+   */
+  int32_t NumberOfBytesWritten() const { return size_; }
+  /**
+   * Returns true if any bytes were discarded, i.e., if there was an
+   * attempt to write more than 'capacity' bytes.
+   * @return true if more than 'capacity' bytes were Append()ed
+   * @stable ICU 4.2
+   */
+  UBool Overflowed() const { return overflowed_; }
+  /**
+   * Returns the number of bytes appended to the sink.
+   * If Overflowed() then NumberOfBytesAppended()>NumberOfBytesWritten()
+   * else they return the same number.
+   * @return number of bytes written to the buffer
+   * @stable ICU 4.6
+   */
+  int32_t NumberOfBytesAppended() const { return appended_; }
+private:
+  char* outbuf_;
+  const int32_t capacity_;
+  int32_t size_;
+  int32_t appended_;
+  UBool overflowed_;
+
+  CheckedArrayByteSink() = delete;
+  CheckedArrayByteSink(const CheckedArrayByteSink &) = delete;
+  CheckedArrayByteSink &operator=(const CheckedArrayByteSink &) = delete;
+};
+
+/** 
+ * Implementation of ByteSink that writes to a "string".
+ * The StringClass is usually instantiated with a std::string.
+ * @stable ICU 4.2
+ */
+template<typename StringClass>
+class StringByteSink : public ByteSink {
+ public:
+  /**
+   * Constructs a ByteSink that will append bytes to the dest string.
+   * @param dest pointer to string object to append to
+   * @stable ICU 4.2
+   */
+  StringByteSink(StringClass* dest) : dest_(dest) { }
+  /**
+   * Constructs a ByteSink that reserves append capacity and will append bytes to the dest string.
+   * 
+   * @param dest pointer to string object to append to
+   * @param initialAppendCapacity capacity beyond dest->length() to be reserve()d
+   * @stable ICU 60
+   */
+  StringByteSink(StringClass* dest, int32_t initialAppendCapacity) : dest_(dest) {
+    if (initialAppendCapacity > 0 &&
+        (uint32_t)initialAppendCapacity > (dest->capacity() - dest->length())) {
+      dest->reserve(dest->length() + initialAppendCapacity);
+    }
+  }
+  /**
+   * Append "bytes[0,n-1]" to this.
+   * @param data the pointer to the bytes
+   * @param n the number of bytes; must be non-negative
+   * @stable ICU 4.2
+   */
+  virtual void Append(const char* data, int32_t n) { dest_->append(data, n); }
+ private:
+  StringClass* dest_;
+
+  StringByteSink() = delete;
+  StringByteSink(const StringByteSink &) = delete;
+  StringByteSink &operator=(const StringByteSink &) = delete;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __BYTESTREAM_H__
diff --git a/thirdparty/icu4c/common/unicode/bytestrie.h b/thirdparty/icu4c/common/unicode/bytestrie.h
new file mode 100644
index 0000000000..85f802df42
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/bytestrie.h
@@ -0,0 +1,565 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  bytestrie.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010sep25
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __BYTESTRIE_H__
+#define __BYTESTRIE_H__
+
+/**
+ * \file
+ * \brief C++ API: Trie for mapping byte sequences to integer values.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+#include "unicode/ustringtrie.h"
+
+U_NAMESPACE_BEGIN
+
+class ByteSink;
+class BytesTrieBuilder;
+class CharString;
+class UVector32;
+
+/**
+ * Light-weight, non-const reader class for a BytesTrie.
+ * Traverses a byte-serialized data structure with minimal state,
+ * for mapping byte sequences to non-negative integer values.
+ *
+ * This class owns the serialized trie data only if it was constructed by
+ * the builder's build() method.
+ * The public constructor and the copy constructor only alias the data (only copy the pointer).
+ * There is no assignment operator.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API BytesTrie : public UMemory {
+public:
+    /**
+     * Constructs a BytesTrie reader instance.
+     *
+     * The trieBytes must contain a copy of a byte sequence from the BytesTrieBuilder,
+     * starting with the first byte of that sequence.
+     * The BytesTrie object will not read more bytes than
+     * the BytesTrieBuilder generated in the corresponding build() call.
+     *
+     * The array is not copied/cloned and must not be modified while
+     * the BytesTrie object is in use.
+     *
+     * @param trieBytes The byte array that contains the serialized trie.
+     * @stable ICU 4.8
+     */
+    BytesTrie(const void *trieBytes)
+            : ownedArray_(NULL), bytes_(static_cast<const uint8_t *>(trieBytes)),
+              pos_(bytes_), remainingMatchLength_(-1) {}
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    ~BytesTrie();
+
+    /**
+     * Copy constructor, copies the other trie reader object and its state,
+     * but not the byte array which will be shared. (Shallow copy.)
+     * @param other Another BytesTrie object.
+     * @stable ICU 4.8
+     */
+    BytesTrie(const BytesTrie &other)
+            : ownedArray_(NULL), bytes_(other.bytes_),
+              pos_(other.pos_), remainingMatchLength_(other.remainingMatchLength_) {}
+
+    /**
+     * Resets this trie to its initial state.
+     * @return *this
+     * @stable ICU 4.8
+     */
+    BytesTrie &reset() {
+        pos_=bytes_;
+        remainingMatchLength_=-1;
+        return *this;
+    }
+
+    /**
+     * Returns the state of this trie as a 64-bit integer.
+     * The state value is never 0.
+     *
+     * @return opaque state value
+     * @see resetToState64
+     * @stable ICU 65
+     */
+    uint64_t getState64() const {
+        return (static_cast<uint64_t>(remainingMatchLength_ + 2) << kState64RemainingShift) |
+            (uint64_t)(pos_ - bytes_);
+    }
+
+    /**
+     * Resets this trie to the saved state.
+     * Unlike resetToState(State), the 64-bit state value
+     * must be from getState64() from the same trie object or
+     * from one initialized the exact same way.
+     * Because of no validation, this method is faster.
+     *
+     * @param state The opaque trie state value from getState64().
+     * @return *this
+     * @see getState64
+     * @see resetToState
+     * @see reset
+     * @stable ICU 65
+     */
+    BytesTrie &resetToState64(uint64_t state) {
+        remainingMatchLength_ = static_cast<int32_t>(state >> kState64RemainingShift) - 2;
+        pos_ = bytes_ + (state & kState64PosMask);
+        return *this;
+    }
+
+    /**
+     * BytesTrie state object, for saving a trie's current state
+     * and resetting the trie back to this state later.
+     * @stable ICU 4.8
+     */
+    class State : public UMemory {
+    public:
+        /**
+         * Constructs an empty State.
+         * @stable ICU 4.8
+         */
+        State() { bytes=NULL; }
+    private:
+        friend class BytesTrie;
+
+        const uint8_t *bytes;
+        const uint8_t *pos;
+        int32_t remainingMatchLength;
+    };
+
+    /**
+     * Saves the state of this trie.
+     * @param state The State object to hold the trie's state.
+     * @return *this
+     * @see resetToState
+     * @stable ICU 4.8
+     */
+    const BytesTrie &saveState(State &state) const {
+        state.bytes=bytes_;
+        state.pos=pos_;
+        state.remainingMatchLength=remainingMatchLength_;
+        return *this;
+    }
+
+    /**
+     * Resets this trie to the saved state.
+     * If the state object contains no state, or the state of a different trie,
+     * then this trie remains unchanged.
+     * @param state The State object which holds a saved trie state.
+     * @return *this
+     * @see saveState
+     * @see reset
+     * @stable ICU 4.8
+     */
+    BytesTrie &resetToState(const State &state) {
+        if(bytes_==state.bytes && bytes_!=NULL) {
+            pos_=state.pos;
+            remainingMatchLength_=state.remainingMatchLength;
+        }
+        return *this;
+    }
+
+    /**
+     * Determines whether the byte sequence so far matches, whether it has a value,
+     * and whether another input byte can continue a matching byte sequence.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult current() const;
+
+    /**
+     * Traverses the trie from the initial state for this input byte.
+     * Equivalent to reset().next(inByte).
+     * @param inByte Input byte value. Values -0x100..-1 are treated like 0..0xff.
+     *               Values below -0x100 and above 0xff will never match.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    inline UStringTrieResult first(int32_t inByte) {
+        remainingMatchLength_=-1;
+        if(inByte<0) {
+            inByte+=0x100;
+        }
+        return nextImpl(bytes_, inByte);
+    }
+
+    /**
+     * Traverses the trie from the current state for this input byte.
+     * @param inByte Input byte value. Values -0x100..-1 are treated like 0..0xff.
+     *               Values below -0x100 and above 0xff will never match.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult next(int32_t inByte);
+
+    /**
+     * Traverses the trie from the current state for this byte sequence.
+     * Equivalent to
+     * \code
+     * Result result=current();
+     * for(each c in s)
+     *   if(!USTRINGTRIE_HAS_NEXT(result)) return USTRINGTRIE_NO_MATCH;
+     *   result=next(c);
+     * return result;
+     * \endcode
+     * @param s A string or byte sequence. Can be NULL if length is 0.
+     * @param length The length of the byte sequence. Can be -1 if NUL-terminated.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult next(const char *s, int32_t length);
+
+    /**
+     * Returns a matching byte sequence's value if called immediately after
+     * current()/first()/next() returned USTRINGTRIE_INTERMEDIATE_VALUE or USTRINGTRIE_FINAL_VALUE.
+     * getValue() can be called multiple times.
+     *
+     * Do not call getValue() after USTRINGTRIE_NO_MATCH or USTRINGTRIE_NO_VALUE!
+     * @return The value for the byte sequence so far.
+     * @stable ICU 4.8
+     */
+    inline int32_t getValue() const {
+        const uint8_t *pos=pos_;
+        int32_t leadByte=*pos++;
+        // U_ASSERT(leadByte>=kMinValueLead);
+        return readValue(pos, leadByte>>1);
+    }
+
+    /**
+     * Determines whether all byte sequences reachable from the current state
+     * map to the same value.
+     * @param uniqueValue Receives the unique value, if this function returns true.
+     *                    (output-only)
+     * @return true if all byte sequences reachable from the current state
+     *         map to the same value.
+     * @stable ICU 4.8
+     */
+    inline UBool hasUniqueValue(int32_t &uniqueValue) const {
+        const uint8_t *pos=pos_;
+        // Skip the rest of a pending linear-match node.
+        return pos!=NULL && findUniqueValue(pos+remainingMatchLength_+1, false, uniqueValue);
+    }
+
+    /**
+     * Finds each byte which continues the byte sequence from the current state.
+     * That is, each byte b for which it would be next(b)!=USTRINGTRIE_NO_MATCH now.
+     * @param out Each next byte is appended to this object.
+     *            (Only uses the out.Append(s, length) method.)
+     * @return the number of bytes which continue the byte sequence from here
+     * @stable ICU 4.8
+     */
+    int32_t getNextBytes(ByteSink &out) const;
+
+    /**
+     * Iterator for all of the (byte sequence, value) pairs in a BytesTrie.
+     * @stable ICU 4.8
+     */
+    class U_COMMON_API Iterator : public UMemory {
+    public:
+        /**
+         * Iterates from the root of a byte-serialized BytesTrie.
+         * @param trieBytes The trie bytes.
+         * @param maxStringLength If 0, the iterator returns full strings/byte sequences.
+         *                        Otherwise, the iterator returns strings with this maximum length.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @stable ICU 4.8
+         */
+        Iterator(const void *trieBytes, int32_t maxStringLength, UErrorCode &errorCode);
+
+        /**
+         * Iterates from the current state of the specified BytesTrie.
+         * @param trie The trie whose state will be copied for iteration.
+         * @param maxStringLength If 0, the iterator returns full strings/byte sequences.
+         *                        Otherwise, the iterator returns strings with this maximum length.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @stable ICU 4.8
+         */
+        Iterator(const BytesTrie &trie, int32_t maxStringLength, UErrorCode &errorCode);
+
+        /**
+         * Destructor.
+         * @stable ICU 4.8
+         */
+        ~Iterator();
+
+        /**
+         * Resets this iterator to its initial state.
+         * @return *this
+         * @stable ICU 4.8
+         */
+        Iterator &reset();
+
+        /**
+         * @return true if there are more elements.
+         * @stable ICU 4.8
+         */
+        UBool hasNext() const;
+
+        /**
+         * Finds the next (byte sequence, value) pair if there is one.
+         *
+         * If the byte sequence is truncated to the maximum length and does not
+         * have a real value, then the value is set to -1.
+         * In this case, this "not a real value" is indistinguishable from
+         * a real value of -1.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @return true if there is another element.
+         * @stable ICU 4.8
+         */
+        UBool next(UErrorCode &errorCode);
+
+        /**
+         * @return The NUL-terminated byte sequence for the last successful next().
+         * @stable ICU 4.8
+         */
+        StringPiece getString() const;
+        /**
+         * @return The value for the last successful next().
+         * @stable ICU 4.8
+         */
+        int32_t getValue() const { return value_; }
+
+    private:
+        UBool truncateAndStop();
+
+        const uint8_t *branchNext(const uint8_t *pos, int32_t length, UErrorCode &errorCode);
+
+        const uint8_t *bytes_;
+        const uint8_t *pos_;
+        const uint8_t *initialPos_;
+        int32_t remainingMatchLength_;
+        int32_t initialRemainingMatchLength_;
+
+        CharString *str_;
+        int32_t maxLength_;
+        int32_t value_;
+
+        // The stack stores pairs of integers for backtracking to another
+        // outbound edge of a branch node.
+        // The first integer is an offset from bytes_.
+        // The second integer has the str_->length() from before the node in bits 15..0,
+        // and the remaining branch length in bits 24..16. (Bits 31..25 are unused.)
+        // (We could store the remaining branch length minus 1 in bits 23..16 and not use bits 31..24,
+        // but the code looks more confusing that way.)
+        UVector32 *stack_;
+    };
+
+private:
+    friend class BytesTrieBuilder;
+
+    /**
+     * Constructs a BytesTrie reader instance.
+     * Unlike the public constructor which just aliases an array,
+     * this constructor adopts the builder's array.
+     * This constructor is only called by the builder.
+     */
+    BytesTrie(void *adoptBytes, const void *trieBytes)
+            : ownedArray_(static_cast<uint8_t *>(adoptBytes)),
+              bytes_(static_cast<const uint8_t *>(trieBytes)),
+              pos_(bytes_), remainingMatchLength_(-1) {}
+
+    // No assignment operator.
+    BytesTrie &operator=(const BytesTrie &other);
+
+    inline void stop() {
+        pos_=NULL;
+    }
+
+    // Reads a compact 32-bit integer.
+    // pos is already after the leadByte, and the lead byte is already shifted right by 1.
+    static int32_t readValue(const uint8_t *pos, int32_t leadByte);
+    static inline const uint8_t *skipValue(const uint8_t *pos, int32_t leadByte) {
+        // U_ASSERT(leadByte>=kMinValueLead);
+        if(leadByte>=(kMinTwoByteValueLead<<1)) {
+            if(leadByte<(kMinThreeByteValueLead<<1)) {
+                ++pos;
+            } else if(leadByte<(kFourByteValueLead<<1)) {
+                pos+=2;
+            } else {
+                pos+=3+((leadByte>>1)&1);
+            }
+        }
+        return pos;
+    }
+    static inline const uint8_t *skipValue(const uint8_t *pos) {
+        int32_t leadByte=*pos++;
+        return skipValue(pos, leadByte);
+    }
+
+    // Reads a jump delta and jumps.
+    static const uint8_t *jumpByDelta(const uint8_t *pos);
+
+    static inline const uint8_t *skipDelta(const uint8_t *pos) {
+        int32_t delta=*pos++;
+        if(delta>=kMinTwoByteDeltaLead) {
+            if(delta<kMinThreeByteDeltaLead) {
+                ++pos;
+            } else if(delta<kFourByteDeltaLead) {
+                pos+=2;
+            } else {
+                pos+=3+(delta&1);
+            }
+        }
+        return pos;
+    }
+
+    static inline UStringTrieResult valueResult(int32_t node) {
+        return (UStringTrieResult)(USTRINGTRIE_INTERMEDIATE_VALUE-(node&kValueIsFinal));
+    }
+
+    // Handles a branch node for both next(byte) and next(string).
+    UStringTrieResult branchNext(const uint8_t *pos, int32_t length, int32_t inByte);
+
+    // Requires remainingLength_<0.
+    UStringTrieResult nextImpl(const uint8_t *pos, int32_t inByte);
+
+    // Helper functions for hasUniqueValue().
+    // Recursively finds a unique value (or whether there is not a unique one)
+    // from a branch.
+    static const uint8_t *findUniqueValueFromBranch(const uint8_t *pos, int32_t length,
+                                                    UBool haveUniqueValue, int32_t &uniqueValue);
+    // Recursively finds a unique value (or whether there is not a unique one)
+    // starting from a position on a node lead byte.
+    static UBool findUniqueValue(const uint8_t *pos, UBool haveUniqueValue, int32_t &uniqueValue);
+
+    // Helper functions for getNextBytes().
+    // getNextBytes() when pos is on a branch node.
+    static void getNextBranchBytes(const uint8_t *pos, int32_t length, ByteSink &out);
+    static void append(ByteSink &out, int c);
+
+    // BytesTrie data structure
+    //
+    // The trie consists of a series of byte-serialized nodes for incremental
+    // string/byte sequence matching. The root node is at the beginning of the trie data.
+    //
+    // Types of nodes are distinguished by their node lead byte ranges.
+    // After each node, except a final-value node, another node follows to
+    // encode match values or continue matching further bytes.
+    //
+    // Node types:
+    //  - Value node: Stores a 32-bit integer in a compact, variable-length format.
+    //    The value is for the string/byte sequence so far.
+    //    One node bit indicates whether the value is final or whether
+    //    matching continues with the next node.
+    //  - Linear-match node: Matches a number of bytes.
+    //  - Branch node: Branches to other nodes according to the current input byte.
+    //    The node byte is the length of the branch (number of bytes to select from)
+    //    minus 1. It is followed by a sub-node:
+    //    - If the length is at most kMaxBranchLinearSubNodeLength, then
+    //      there are length-1 (key, value) pairs and then one more comparison byte.
+    //      If one of the key bytes matches, then the value is either a final value for
+    //      the string/byte sequence so far, or a "jump" delta to the next node.
+    //      If the last byte matches, then matching continues with the next node.
+    //      (Values have the same encoding as value nodes.)
+    //    - If the length is greater than kMaxBranchLinearSubNodeLength, then
+    //      there is one byte and one "jump" delta.
+    //      If the input byte is less than the sub-node byte, then "jump" by delta to
+    //      the next sub-node which will have a length of length/2.
+    //      (The delta has its own compact encoding.)
+    //      Otherwise, skip the "jump" delta to the next sub-node
+    //      which will have a length of length-length/2.
+
+    // Node lead byte values.
+
+    // 00..0f: Branch node. If node!=0 then the length is node+1, otherwise
+    // the length is one more than the next byte.
+
+    // For a branch sub-node with at most this many entries, we drop down
+    // to a linear search.
+    static const int32_t kMaxBranchLinearSubNodeLength=5;
+
+    // 10..1f: Linear-match node, match 1..16 bytes and continue reading the next node.
+    static const int32_t kMinLinearMatch=0x10;
+    static const int32_t kMaxLinearMatchLength=0x10;
+
+    // 20..ff: Variable-length value node.
+    // If odd, the value is final. (Otherwise, intermediate value or jump delta.)
+    // Then shift-right by 1 bit.
+    // The remaining lead byte value indicates the number of following bytes (0..4)
+    // and contains the value's top bits.
+    static const int32_t kMinValueLead=kMinLinearMatch+kMaxLinearMatchLength;  // 0x20
+    // It is a final value if bit 0 is set.
+    static const int32_t kValueIsFinal=1;
+
+    // Compact value: After testing bit 0, shift right by 1 and then use the following thresholds.
+    static const int32_t kMinOneByteValueLead=kMinValueLead/2;  // 0x10
+    static const int32_t kMaxOneByteValue=0x40;  // At least 6 bits in the first byte.
+
+    static const int32_t kMinTwoByteValueLead=kMinOneByteValueLead+kMaxOneByteValue+1;  // 0x51
+    static const int32_t kMaxTwoByteValue=0x1aff;
+
+    static const int32_t kMinThreeByteValueLead=kMinTwoByteValueLead+(kMaxTwoByteValue>>8)+1;  // 0x6c
+    static const int32_t kFourByteValueLead=0x7e;
+
+    // A little more than Unicode code points. (0x11ffff)
+    static const int32_t kMaxThreeByteValue=((kFourByteValueLead-kMinThreeByteValueLead)<<16)-1;
+
+    static const int32_t kFiveByteValueLead=0x7f;
+
+    // Compact delta integers.
+    static const int32_t kMaxOneByteDelta=0xbf;
+    static const int32_t kMinTwoByteDeltaLead=kMaxOneByteDelta+1;  // 0xc0
+    static const int32_t kMinThreeByteDeltaLead=0xf0;
+    static const int32_t kFourByteDeltaLead=0xfe;
+    static const int32_t kFiveByteDeltaLead=0xff;
+
+    static const int32_t kMaxTwoByteDelta=((kMinThreeByteDeltaLead-kMinTwoByteDeltaLead)<<8)-1;  // 0x2fff
+    static const int32_t kMaxThreeByteDelta=((kFourByteDeltaLead-kMinThreeByteDeltaLead)<<16)-1;  // 0xdffff
+
+    // For getState64():
+    // The remainingMatchLength_ is -1..14=(kMaxLinearMatchLength=0x10)-2
+    // so we need at least 5 bits for that.
+    // We add 2 to store it as a positive value 1..16=kMaxLinearMatchLength.
+    static constexpr int32_t kState64RemainingShift = 59;
+    static constexpr uint64_t kState64PosMask = (UINT64_C(1) << kState64RemainingShift) - 1;
+
+    uint8_t *ownedArray_;
+
+    // Fixed value referencing the BytesTrie bytes.
+    const uint8_t *bytes_;
+
+    // Iterator variables.
+
+    // Pointer to next trie byte to read. NULL if no more matches.
+    const uint8_t *pos_;
+    // Remaining length of a linear-match node, minus 1. Negative if not in such a node.
+    int32_t remainingMatchLength_;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __BYTESTRIE_H__
diff --git a/thirdparty/icu4c/common/unicode/bytestriebuilder.h b/thirdparty/icu4c/common/unicode/bytestriebuilder.h
new file mode 100644
index 0000000000..cae16e48b4
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/bytestriebuilder.h
@@ -0,0 +1,188 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  bytestriebuilder.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010sep25
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C++ API: Builder for icu::BytesTrie
+ */
+
+#ifndef __BYTESTRIEBUILDER_H__
+#define __BYTESTRIEBUILDER_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/bytestrie.h"
+#include "unicode/stringpiece.h"
+#include "unicode/stringtriebuilder.h"
+
+U_NAMESPACE_BEGIN
+
+class BytesTrieElement;
+class CharString;
+/**
+ * Builder class for BytesTrie.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API BytesTrieBuilder : public StringTrieBuilder {
+public:
+    /**
+     * Constructs an empty builder.
+     * @param errorCode Standard ICU error code.
+     * @stable ICU 4.8
+     */
+    BytesTrieBuilder(UErrorCode &errorCode);
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    virtual ~BytesTrieBuilder();
+
+    /**
+     * Adds a (byte sequence, value) pair.
+     * The byte sequence must be unique.
+     * The bytes will be copied; the builder does not keep
+     * a reference to the input StringPiece or its data().
+     * @param s The input byte sequence.
+     * @param value The value associated with this byte sequence.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @stable ICU 4.8
+     */
+    BytesTrieBuilder &add(StringPiece s, int32_t value, UErrorCode &errorCode);
+
+    /**
+     * Builds a BytesTrie for the add()ed data.
+     * Once built, no further data can be add()ed until clear() is called.
+     *
+     * A BytesTrie cannot be empty. At least one (byte sequence, value) pair
+     * must have been add()ed.
+     *
+     * This method passes ownership of the builder's internal result array to the new trie object.
+     * Another call to any build() variant will re-serialize the trie.
+     * After clear() has been called, a new array will be used as well.
+     * @param buildOption Build option, see UStringTrieBuildOption.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return A new BytesTrie for the add()ed data.
+     * @stable ICU 4.8
+     */
+    BytesTrie *build(UStringTrieBuildOption buildOption, UErrorCode &errorCode);
+
+    /**
+     * Builds a BytesTrie for the add()ed data and byte-serializes it.
+     * Once built, no further data can be add()ed until clear() is called.
+     *
+     * A BytesTrie cannot be empty. At least one (byte sequence, value) pair
+     * must have been add()ed.
+     *
+     * Multiple calls to buildStringPiece() return StringPieces referring to the
+     * builder's same byte array, without rebuilding.
+     * If buildStringPiece() is called after build(), the trie will be
+     * re-serialized into a new array (because build() passes on ownership).
+     * If build() is called after buildStringPiece(), the trie object returned
+     * by build() will become the owner of the underlying string for the
+     * previously returned StringPiece.
+     * After clear() has been called, a new array will be used as well.
+     * @param buildOption Build option, see UStringTrieBuildOption.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return A StringPiece which refers to the byte-serialized BytesTrie for the add()ed data.
+     * @stable ICU 4.8
+     */
+    StringPiece buildStringPiece(UStringTrieBuildOption buildOption, UErrorCode &errorCode);
+
+    /**
+     * Removes all (byte sequence, value) pairs.
+     * New data can then be add()ed and a new trie can be built.
+     * @return *this
+     * @stable ICU 4.8
+     */
+    BytesTrieBuilder &clear();
+
+private:
+    BytesTrieBuilder(const BytesTrieBuilder &other);  // no copy constructor
+    BytesTrieBuilder &operator=(const BytesTrieBuilder &other);  // no assignment operator
+
+    void buildBytes(UStringTrieBuildOption buildOption, UErrorCode &errorCode);
+
+    virtual int32_t getElementStringLength(int32_t i) const;
+    virtual char16_t getElementUnit(int32_t i, int32_t byteIndex) const;
+    virtual int32_t getElementValue(int32_t i) const;
+
+    virtual int32_t getLimitOfLinearMatch(int32_t first, int32_t last, int32_t byteIndex) const;
+
+    virtual int32_t countElementUnits(int32_t start, int32_t limit, int32_t byteIndex) const;
+    virtual int32_t skipElementsBySomeUnits(int32_t i, int32_t byteIndex, int32_t count) const;
+    virtual int32_t indexOfElementWithNextUnit(int32_t i, int32_t byteIndex, char16_t byte) const;
+
+    virtual UBool matchNodesCanHaveValues() const { return false; }
+
+    virtual int32_t getMaxBranchLinearSubNodeLength() const { return BytesTrie::kMaxBranchLinearSubNodeLength; }
+    virtual int32_t getMinLinearMatch() const { return BytesTrie::kMinLinearMatch; }
+    virtual int32_t getMaxLinearMatchLength() const { return BytesTrie::kMaxLinearMatchLength; }
+
+    /**
+     * @internal (private)
+     */
+    class BTLinearMatchNode : public LinearMatchNode {
+    public:
+        BTLinearMatchNode(const char *units, int32_t len, Node *nextNode);
+        virtual UBool operator==(const Node &other) const;
+        virtual void write(StringTrieBuilder &builder);
+    private:
+        const char *s;
+    };
+    
+    virtual Node *createLinearMatchNode(int32_t i, int32_t byteIndex, int32_t length,
+                                        Node *nextNode) const;
+
+    UBool ensureCapacity(int32_t length);
+    virtual int32_t write(int32_t byte);
+    int32_t write(const char *b, int32_t length);
+    virtual int32_t writeElementUnits(int32_t i, int32_t byteIndex, int32_t length);
+    virtual int32_t writeValueAndFinal(int32_t i, UBool isFinal);
+    virtual int32_t writeValueAndType(UBool hasValue, int32_t value, int32_t node);
+    virtual int32_t writeDeltaTo(int32_t jumpTarget);
+
+    CharString *strings;  // Pointer not object so we need not #include internal charstr.h.
+    BytesTrieElement *elements;
+    int32_t elementsCapacity;
+    int32_t elementsLength;
+
+    // Byte serialization of the trie.
+    // Grows from the back: bytesLength measures from the end of the buffer!
+    char *bytes;
+    int32_t bytesCapacity;
+    int32_t bytesLength;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __BYTESTRIEBUILDER_H__
diff --git a/thirdparty/icu4c/common/unicode/caniter.h b/thirdparty/icu4c/common/unicode/caniter.h
new file mode 100644
index 0000000000..4ed2b74b10
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/caniter.h
@@ -0,0 +1,214 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ * Copyright (C) 1996-2014, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ *******************************************************************************
+ */
+
+#ifndef CANITER_H
+#define CANITER_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+
+/**
+ * \file
+ * \brief C++ API: Canonical Iterator
+ */
+ 
+/** Should permutation skip characters with combining class zero
+ *  Should be either true or false. This is a compile time option
+ *  @stable ICU 2.4
+ */
+#ifndef CANITER_SKIP_ZEROES
+#define CANITER_SKIP_ZEROES true
+#endif
+
+U_NAMESPACE_BEGIN
+
+class Hashtable;
+class Normalizer2;
+class Normalizer2Impl;
+
+/**
+ * This class allows one to iterate through all the strings that are canonically equivalent to a given
+ * string. For example, here are some sample results:
+Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+1: \\u0041\\u030A\\u0064\\u0307\\u0327
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+2: \\u0041\\u030A\\u0064\\u0327\\u0307
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+3: \\u0041\\u030A\\u1E0B\\u0327
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+4: \\u0041\\u030A\\u1E11\\u0307
+ = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+5: \\u00C5\\u0064\\u0307\\u0327
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+6: \\u00C5\\u0064\\u0327\\u0307
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+7: \\u00C5\\u1E0B\\u0327
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+8: \\u00C5\\u1E11\\u0307
+ = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+9: \\u212B\\u0064\\u0307\\u0327
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
+10: \\u212B\\u0064\\u0327\\u0307
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
+11: \\u212B\\u1E0B\\u0327
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
+12: \\u212B\\u1E11\\u0307
+ = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
+ *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
+ * since it has not been optimized for that situation.
+ * Note, CanonicalIterator is not intended to be subclassed.
+ * @author M. Davis
+ * @author C++ port by V. Weinstein
+ * @stable ICU 2.4
+ */
+class U_COMMON_API CanonicalIterator U_FINAL : public UObject {
+public:
+    /**
+     * Construct a CanonicalIterator object
+     * @param source    string to get results for
+     * @param status    Fill-in parameter which receives the status of this operation.
+     * @stable ICU 2.4
+     */
+    CanonicalIterator(const UnicodeString &source, UErrorCode &status);
+
+    /** Destructor
+     *  Cleans pieces
+     * @stable ICU 2.4
+     */
+    virtual ~CanonicalIterator();
+
+    /**
+     * Gets the NFD form of the current source we are iterating over.
+     * @return gets the source: NOTE: it is the NFD form of source
+     * @stable ICU 2.4
+     */
+    UnicodeString getSource();
+
+    /**
+     * Resets the iterator so that one can start again from the beginning.
+     * @stable ICU 2.4
+     */
+    void reset();
+
+    /**
+     * Get the next canonically equivalent string.
+     * <br><b>Warning: The strings are not guaranteed to be in any particular order.</b>
+     * @return the next string that is canonically equivalent. A bogus string is returned when
+     * the iteration is done.
+     * @stable ICU 2.4
+     */
+    UnicodeString next();
+
+    /**
+     * Set a new source for this iterator. Allows object reuse.
+     * @param newSource     the source string to iterate against. This allows the same iterator to be used
+     *                     while changing the source string, saving object creation.
+     * @param status        Fill-in parameter which receives the status of this operation.
+     * @stable ICU 2.4
+     */
+    void setSource(const UnicodeString &newSource, UErrorCode &status);
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Dumb recursive implementation of permutation.
+     * TODO: optimize
+     * @param source     the string to find permutations for
+     * @param skipZeros  determine if skip zeros
+     * @param result     the results in a set.
+     * @param status       Fill-in parameter which receives the status of this operation.
+     * @internal
+     */
+    static void U_EXPORT2 permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.2
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     *
+     * @stable ICU 2.2
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    // ===================== PRIVATES ==============================
+    // private default constructor
+    CanonicalIterator();
+
+
+    /**
+     * Copy constructor. Private for now.
+     * @internal (private)
+     */
+    CanonicalIterator(const CanonicalIterator& other);
+
+    /**
+     * Assignment operator. Private for now.
+     * @internal (private)
+     */
+    CanonicalIterator& operator=(const CanonicalIterator& other);
+
+    // fields
+    UnicodeString source;
+    UBool done;
+
+    // 2 dimensional array holds the pieces of the string with
+    // their different canonically equivalent representations
+    UnicodeString **pieces;
+    int32_t pieces_length;
+    int32_t *pieces_lengths;
+
+    // current is used in iterating to combine pieces
+    int32_t *current;
+    int32_t current_length;
+
+    // transient fields
+    UnicodeString buffer;
+
+    const Normalizer2 &nfd;
+    const Normalizer2Impl &nfcImpl;
+
+    // we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
+    UnicodeString *getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status); //private String[] getEquivalents(String segment)
+
+    //Set getEquivalents2(String segment);
+    Hashtable *getEquivalents2(Hashtable *fillinResult, const char16_t *segment, int32_t segLen, UErrorCode &status);
+    //Hashtable *getEquivalents2(const UnicodeString &segment, int32_t segLen, UErrorCode &status);
+
+    /**
+     * See if the decomposition of cp2 is at segment starting at segmentPos
+     * (with canonical rearrangment!)
+     * If so, take the remainder, and return the equivalents
+     */
+    //Set extract(int comp, String segment, int segmentPos, StringBuffer buffer);
+    Hashtable *extract(Hashtable *fillinResult, UChar32 comp, const char16_t *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
+    //Hashtable *extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status);
+
+    void cleanPieces();
+
+};
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/casemap.h b/thirdparty/icu4c/common/unicode/casemap.h
new file mode 100644
index 0000000000..53af84fa74
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/casemap.h
@@ -0,0 +1,497 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// casemap.h
+// created: 2017jan12 Markus W. Scherer
+
+#ifndef __CASEMAP_H__
+#define __CASEMAP_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: Low-level C++ case mapping functions.
+ */
+
+U_NAMESPACE_BEGIN
+
+class BreakIterator;
+class ByteSink;
+class Edits;
+
+/**
+ * Low-level C++ case mapping functions.
+ *
+ * @stable ICU 59
+ */
+class U_COMMON_API CaseMap U_FINAL : public UMemory {
+public:
+    /**
+     * Lowercases a UTF-16 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of char16_ts). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see u_strToLower
+     * @stable ICU 59
+     */
+     static int32_t toLower(
+            const char *locale, uint32_t options,
+            const char16_t *src, int32_t srcLength,
+            char16_t *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+    /**
+     * Uppercases a UTF-16 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of char16_ts). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see u_strToUpper
+     * @stable ICU 59
+     */
+    static int32_t toUpper(
+            const char *locale, uint32_t options,
+            const char16_t *src, int32_t srcLength,
+            char16_t *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Titlecases a UTF-16 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * Titlecasing uses a break iterator to find the first characters of words
+     * that are to be titlecased. It titlecases those characters and lowercases
+     * all others. (This can be modified with options bits.)
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET,
+     *                  U_TITLECASE_NO_LOWERCASE,
+     *                  U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED,
+     *                  U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
+     * @param iter      A break iterator to find the first characters of words that are to be titlecased.
+     *                  It is set to the source string (setText())
+     *                  and used one or more times for iteration (first() and next()).
+     *                  If NULL, then a word break iterator for the locale is used
+     *                  (or something equivalent).
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of char16_ts). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see u_strToTitle
+     * @see ucasemap_toTitle
+     * @stable ICU 59
+     */
+    static int32_t toTitle(
+            const char *locale, uint32_t options, BreakIterator *iter,
+            const char16_t *src, int32_t srcLength,
+            char16_t *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+#endif  // UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Case-folds a UTF-16 string and optionally records edits.
+     *
+     * Case folding is locale-independent and not context-sensitive,
+     * but there is an option for whether to include or exclude mappings for dotted I
+     * and dotless i that are marked with 'T' in CaseFolding.txt.
+     *
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET,
+     *                  U_FOLD_CASE_DEFAULT, U_FOLD_CASE_EXCLUDE_SPECIAL_I.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of char16_ts). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see u_strFoldCase
+     * @stable ICU 59
+     */
+    static int32_t fold(
+            uint32_t options,
+            const char16_t *src, int32_t srcLength,
+            char16_t *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+    /**
+     * Lowercases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param sink      A ByteSink to which the result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     *
+     * @see ucasemap_utf8ToLower
+     * @stable ICU 60
+     */
+    static void utf8ToLower(
+            const char *locale, uint32_t options,
+            StringPiece src, ByteSink &sink, Edits *edits,
+            UErrorCode &errorCode);
+
+    /**
+     * Uppercases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param sink      A ByteSink to which the result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     *
+     * @see ucasemap_utf8ToUpper
+     * @stable ICU 60
+     */
+    static void utf8ToUpper(
+            const char *locale, uint32_t options,
+            StringPiece src, ByteSink &sink, Edits *edits,
+            UErrorCode &errorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Titlecases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     *
+     * Titlecasing uses a break iterator to find the first characters of words
+     * that are to be titlecased. It titlecases those characters and lowercases
+     * all others. (This can be modified with options bits.)
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET,
+     *                  U_TITLECASE_NO_LOWERCASE,
+     *                  U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED,
+     *                  U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
+     * @param iter      A break iterator to find the first characters of words that are to be titlecased.
+     *                  It is set to the source string (setUText())
+     *                  and used one or more times for iteration (first() and next()).
+     *                  If NULL, then a word break iterator for the locale is used
+     *                  (or something equivalent).
+     * @param src       The original string.
+     * @param sink      A ByteSink to which the result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     *
+     * @see ucasemap_utf8ToTitle
+     * @stable ICU 60
+     */
+    static void utf8ToTitle(
+            const char *locale, uint32_t options, BreakIterator *iter,
+            StringPiece src, ByteSink &sink, Edits *edits,
+            UErrorCode &errorCode);
+
+#endif  // UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Case-folds a UTF-8 string and optionally records edits.
+     *
+     * Case folding is locale-independent and not context-sensitive,
+     * but there is an option for whether to include or exclude mappings for dotted I
+     * and dotless i that are marked with 'T' in CaseFolding.txt.
+     *
+     * The result may be longer or shorter than the original.
+     *
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param sink      A ByteSink to which the result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     *
+     * @see ucasemap_utf8FoldCase
+     * @stable ICU 60
+     */
+    static void utf8Fold(
+            uint32_t options,
+            StringPiece src, ByteSink &sink, Edits *edits,
+            UErrorCode &errorCode);
+
+    /**
+     * Lowercases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see ucasemap_utf8ToLower
+     * @stable ICU 59
+     */
+    static int32_t utf8ToLower(
+            const char *locale, uint32_t options,
+            const char *src, int32_t srcLength,
+            char *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+    /**
+     * Uppercases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see ucasemap_utf8ToUpper
+     * @stable ICU 59
+     */
+    static int32_t utf8ToUpper(
+            const char *locale, uint32_t options,
+            const char *src, int32_t srcLength,
+            char *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Titlecases a UTF-8 string and optionally records edits.
+     * Casing is locale-dependent and context-sensitive.
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * Titlecasing uses a break iterator to find the first characters of words
+     * that are to be titlecased. It titlecases those characters and lowercases
+     * all others. (This can be modified with options bits.)
+     *
+     * @param locale    The locale ID. ("" = root locale, NULL = default locale.)
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET,
+     *                  U_TITLECASE_NO_LOWERCASE,
+     *                  U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED,
+     *                  U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
+     * @param iter      A break iterator to find the first characters of words that are to be titlecased.
+     *                  It is set to the source string (setUText())
+     *                  and used one or more times for iteration (first() and next()).
+     *                  If NULL, then a word break iterator for the locale is used
+     *                  (or something equivalent).
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see ucasemap_utf8ToTitle
+     * @stable ICU 59
+     */
+    static int32_t utf8ToTitle(
+            const char *locale, uint32_t options, BreakIterator *iter,
+            const char *src, int32_t srcLength,
+            char *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+#endif  // UCONFIG_NO_BREAK_ITERATION
+
+    /**
+     * Case-folds a UTF-8 string and optionally records edits.
+     *
+     * Case folding is locale-independent and not context-sensitive,
+     * but there is an option for whether to include or exclude mappings for dotted I
+     * and dotless i that are marked with 'T' in CaseFolding.txt.
+     *
+     * The result may be longer or shorter than the original.
+     * The source string and the destination buffer must not overlap.
+     *
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET,
+     *                  U_FOLD_CASE_DEFAULT, U_FOLD_CASE_EXCLUDE_SPECIAL_I.
+     * @param src       The original string.
+     * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+     * @param dest      A buffer for the result string. The result will be NUL-terminated if
+     *                  the buffer is large enough.
+     *                  The contents is undefined in case of failure.
+     * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+     *                  dest may be NULL and the function will only return the length of the result
+     *                  without writing any of the result string.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be NULL.
+     * @param errorCode Reference to an in/out error code value
+     *                  which must not indicate a failure before the function call.
+     * @return The length of the result string, if successful.
+     *         When the result would be longer than destCapacity,
+     *         the full length is returned and a U_BUFFER_OVERFLOW_ERROR is set.
+     *
+     * @see ucasemap_utf8FoldCase
+     * @stable ICU 59
+     */
+    static int32_t utf8Fold(
+            uint32_t options,
+            const char *src, int32_t srcLength,
+            char *dest, int32_t destCapacity, Edits *edits,
+            UErrorCode &errorCode);
+
+private:
+    CaseMap() = delete;
+    CaseMap(const CaseMap &other) = delete;
+    CaseMap &operator=(const CaseMap &other) = delete;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __CASEMAP_H__
diff --git a/thirdparty/icu4c/common/unicode/char16ptr.h b/thirdparty/icu4c/common/unicode/char16ptr.h
new file mode 100644
index 0000000000..c8a9ae6c35
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/char16ptr.h
@@ -0,0 +1,313 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// char16ptr.h
+// created: 2017feb28 Markus W. Scherer
+
+#ifndef __CHAR16PTR_H__
+#define __CHAR16PTR_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include <cstddef>
+
+/**
+ * \file
+ * \brief C++ API: char16_t pointer wrappers with
+ *        implicit conversion from bit-compatible raw pointer types.
+ *        Also conversion functions from char16_t * to UChar * and OldUChar *.
+ */
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \def U_ALIASING_BARRIER
+ * Barrier for pointer anti-aliasing optimizations even across function boundaries.
+ * @internal
+ */
+#ifdef U_ALIASING_BARRIER
+    // Use the predefined value.
+#elif (defined(__clang__) || defined(__GNUC__)) && U_PLATFORM != U_PF_BROWSER_NATIVE_CLIENT
+#   define U_ALIASING_BARRIER(ptr) asm volatile("" : : "rm"(ptr) : "memory")
+#elif defined(U_IN_DOXYGEN)
+#   define U_ALIASING_BARRIER(ptr)
+#endif
+
+/**
+ * char16_t * wrapper with implicit conversion from distinct but bit-compatible pointer types.
+ * @stable ICU 59
+ */
+class U_COMMON_API Char16Ptr U_FINAL {
+public:
+    /**
+     * Copies the pointer.
+     * @param p pointer
+     * @stable ICU 59
+     */
+    inline Char16Ptr(char16_t *p);
+#if !U_CHAR16_IS_TYPEDEF
+    /**
+     * Converts the pointer to char16_t *.
+     * @param p pointer to be converted
+     * @stable ICU 59
+     */
+    inline Char16Ptr(uint16_t *p);
+#endif
+#if U_SIZEOF_WCHAR_T==2 || defined(U_IN_DOXYGEN)
+    /**
+     * Converts the pointer to char16_t *.
+     * (Only defined if U_SIZEOF_WCHAR_T==2.)
+     * @param p pointer to be converted
+     * @stable ICU 59
+     */
+    inline Char16Ptr(wchar_t *p);
+#endif
+    /**
+     * nullptr constructor.
+     * @param p nullptr
+     * @stable ICU 59
+     */
+    inline Char16Ptr(std::nullptr_t p);
+    /**
+     * Destructor.
+     * @stable ICU 59
+     */
+    inline ~Char16Ptr();
+
+    /**
+     * Pointer access.
+     * @return the wrapped pointer
+     * @stable ICU 59
+     */
+    inline char16_t *get() const;
+    /**
+     * char16_t pointer access via type conversion (e.g., static_cast).
+     * @return the wrapped pointer
+     * @stable ICU 59
+     */
+    inline operator char16_t *() const { return get(); }
+
+private:
+    Char16Ptr() = delete;
+
+#ifdef U_ALIASING_BARRIER
+    template<typename T> static char16_t *cast(T *t) {
+        U_ALIASING_BARRIER(t);
+        return reinterpret_cast<char16_t *>(t);
+    }
+
+    char16_t *p_;
+#else
+    union {
+        char16_t *cp;
+        uint16_t *up;
+        wchar_t *wp;
+    } u_;
+#endif
+};
+
+/// \cond
+#ifdef U_ALIASING_BARRIER
+
+Char16Ptr::Char16Ptr(char16_t *p) : p_(p) {}
+#if !U_CHAR16_IS_TYPEDEF
+Char16Ptr::Char16Ptr(uint16_t *p) : p_(cast(p)) {}
+#endif
+#if U_SIZEOF_WCHAR_T==2
+Char16Ptr::Char16Ptr(wchar_t *p) : p_(cast(p)) {}
+#endif
+Char16Ptr::Char16Ptr(std::nullptr_t p) : p_(p) {}
+Char16Ptr::~Char16Ptr() {
+    U_ALIASING_BARRIER(p_);
+}
+
+char16_t *Char16Ptr::get() const { return p_; }
+
+#else
+
+Char16Ptr::Char16Ptr(char16_t *p) { u_.cp = p; }
+#if !U_CHAR16_IS_TYPEDEF
+Char16Ptr::Char16Ptr(uint16_t *p) { u_.up = p; }
+#endif
+#if U_SIZEOF_WCHAR_T==2
+Char16Ptr::Char16Ptr(wchar_t *p) { u_.wp = p; }
+#endif
+Char16Ptr::Char16Ptr(std::nullptr_t p) { u_.cp = p; }
+Char16Ptr::~Char16Ptr() {}
+
+char16_t *Char16Ptr::get() const { return u_.cp; }
+
+#endif
+/// \endcond
+
+/**
+ * const char16_t * wrapper with implicit conversion from distinct but bit-compatible pointer types.
+ * @stable ICU 59
+ */
+class U_COMMON_API ConstChar16Ptr U_FINAL {
+public:
+    /**
+     * Copies the pointer.
+     * @param p pointer
+     * @stable ICU 59
+     */
+    inline ConstChar16Ptr(const char16_t *p);
+#if !U_CHAR16_IS_TYPEDEF
+    /**
+     * Converts the pointer to char16_t *.
+     * @param p pointer to be converted
+     * @stable ICU 59
+     */
+    inline ConstChar16Ptr(const uint16_t *p);
+#endif
+#if U_SIZEOF_WCHAR_T==2 || defined(U_IN_DOXYGEN)
+    /**
+     * Converts the pointer to char16_t *.
+     * (Only defined if U_SIZEOF_WCHAR_T==2.)
+     * @param p pointer to be converted
+     * @stable ICU 59
+     */
+    inline ConstChar16Ptr(const wchar_t *p);
+#endif
+    /**
+     * nullptr constructor.
+     * @param p nullptr
+     * @stable ICU 59
+     */
+    inline ConstChar16Ptr(const std::nullptr_t p);
+
+    /**
+     * Destructor.
+     * @stable ICU 59
+     */
+    inline ~ConstChar16Ptr();
+
+    /**
+     * Pointer access.
+     * @return the wrapped pointer
+     * @stable ICU 59
+     */
+    inline const char16_t *get() const;
+    /**
+     * char16_t pointer access via type conversion (e.g., static_cast).
+     * @return the wrapped pointer
+     * @stable ICU 59
+     */
+    inline operator const char16_t *() const { return get(); }
+
+private:
+    ConstChar16Ptr() = delete;
+
+#ifdef U_ALIASING_BARRIER
+    template<typename T> static const char16_t *cast(const T *t) {
+        U_ALIASING_BARRIER(t);
+        return reinterpret_cast<const char16_t *>(t);
+    }
+
+    const char16_t *p_;
+#else
+    union {
+        const char16_t *cp;
+        const uint16_t *up;
+        const wchar_t *wp;
+    } u_;
+#endif
+};
+
+/// \cond
+#ifdef U_ALIASING_BARRIER
+
+ConstChar16Ptr::ConstChar16Ptr(const char16_t *p) : p_(p) {}
+#if !U_CHAR16_IS_TYPEDEF
+ConstChar16Ptr::ConstChar16Ptr(const uint16_t *p) : p_(cast(p)) {}
+#endif
+#if U_SIZEOF_WCHAR_T==2
+ConstChar16Ptr::ConstChar16Ptr(const wchar_t *p) : p_(cast(p)) {}
+#endif
+ConstChar16Ptr::ConstChar16Ptr(const std::nullptr_t p) : p_(p) {}
+ConstChar16Ptr::~ConstChar16Ptr() {
+    U_ALIASING_BARRIER(p_);
+}
+
+const char16_t *ConstChar16Ptr::get() const { return p_; }
+
+#else
+
+ConstChar16Ptr::ConstChar16Ptr(const char16_t *p) { u_.cp = p; }
+#if !U_CHAR16_IS_TYPEDEF
+ConstChar16Ptr::ConstChar16Ptr(const uint16_t *p) { u_.up = p; }
+#endif
+#if U_SIZEOF_WCHAR_T==2
+ConstChar16Ptr::ConstChar16Ptr(const wchar_t *p) { u_.wp = p; }
+#endif
+ConstChar16Ptr::ConstChar16Ptr(const std::nullptr_t p) { u_.cp = p; }
+ConstChar16Ptr::~ConstChar16Ptr() {}
+
+const char16_t *ConstChar16Ptr::get() const { return u_.cp; }
+
+#endif
+/// \endcond
+
+/**
+ * Converts from const char16_t * to const UChar *.
+ * Includes an aliasing barrier if available.
+ * @param p pointer
+ * @return p as const UChar *
+ * @stable ICU 59
+ */
+inline const UChar *toUCharPtr(const char16_t *p) {
+#ifdef U_ALIASING_BARRIER
+    U_ALIASING_BARRIER(p);
+#endif
+    return reinterpret_cast<const UChar *>(p);
+}
+
+/**
+ * Converts from char16_t * to UChar *.
+ * Includes an aliasing barrier if available.
+ * @param p pointer
+ * @return p as UChar *
+ * @stable ICU 59
+ */
+inline UChar *toUCharPtr(char16_t *p) {
+#ifdef U_ALIASING_BARRIER
+    U_ALIASING_BARRIER(p);
+#endif
+    return reinterpret_cast<UChar *>(p);
+}
+
+/**
+ * Converts from const char16_t * to const OldUChar *.
+ * Includes an aliasing barrier if available.
+ * @param p pointer
+ * @return p as const OldUChar *
+ * @stable ICU 59
+ */
+inline const OldUChar *toOldUCharPtr(const char16_t *p) {
+#ifdef U_ALIASING_BARRIER
+    U_ALIASING_BARRIER(p);
+#endif
+    return reinterpret_cast<const OldUChar *>(p);
+}
+
+/**
+ * Converts from char16_t * to OldUChar *.
+ * Includes an aliasing barrier if available.
+ * @param p pointer
+ * @return p as OldUChar *
+ * @stable ICU 59
+ */
+inline OldUChar *toOldUCharPtr(char16_t *p) {
+#ifdef U_ALIASING_BARRIER
+    U_ALIASING_BARRIER(p);
+#endif
+    return reinterpret_cast<OldUChar *>(p);
+}
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __CHAR16PTR_H__
diff --git a/thirdparty/icu4c/common/unicode/chariter.h b/thirdparty/icu4c/common/unicode/chariter.h
new file mode 100644
index 0000000000..96dc5db7c9
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/chariter.h
@@ -0,0 +1,734 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************
+*
+*   Copyright (C) 1997-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+********************************************************************
+*/
+
+#ifndef CHARITER_H
+#define CHARITER_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+/**
+ * \file
+ * \brief C++ API: Character Iterator
+ */
+ 
+U_NAMESPACE_BEGIN
+/**
+ * Abstract class that defines an API for forward-only iteration
+ * on text objects.
+ * This is a minimal interface for iteration without random access
+ * or backwards iteration. It is especially useful for wrapping
+ * streams with converters into an object for collation or
+ * normalization.
+ *
+ * <p>Characters can be accessed in two ways: as code units or as
+ * code points.
+ * Unicode code points are 21-bit integers and are the scalar values
+ * of Unicode characters. ICU uses the type UChar32 for them.
+ * Unicode code units are the storage units of a given
+ * Unicode/UCS Transformation Format (a character encoding scheme).
+ * With UTF-16, all code points can be represented with either one
+ * or two code units ("surrogates").
+ * String storage is typically based on code units, while properties
+ * of characters are typically determined using code point values.
+ * Some processes may be designed to work with sequences of code units,
+ * or it may be known that all characters that are important to an
+ * algorithm can be represented with single code units.
+ * Other processes will need to use the code point access functions.</p>
+ *
+ * <p>ForwardCharacterIterator provides nextPostInc() to access
+ * a code unit and advance an internal position into the text object,
+ * similar to a <code>return text[position++]</code>.<br>
+ * It provides next32PostInc() to access a code point and advance an internal
+ * position.</p>
+ *
+ * <p>next32PostInc() assumes that the current position is that of
+ * the beginning of a code point, i.e., of its first code unit.
+ * After next32PostInc(), this will be true again.
+ * In general, access to code units and code points in the same
+ * iteration loop should not be mixed. In UTF-16, if the current position
+ * is on a second code unit (Low Surrogate), then only that code unit
+ * is returned even by next32PostInc().</p>
+ *
+ * <p>For iteration with either function, there are two ways to
+ * check for the end of the iteration. When there are no more
+ * characters in the text object:
+ * <ul>
+ * <li>The hasNext() function returns false.</li>
+ * <li>nextPostInc() and next32PostInc() return DONE
+ *     when one attempts to read beyond the end of the text object.</li>
+ * </ul>
+ *
+ * Example:
+ * \code 
+ * void function1(ForwardCharacterIterator &it) {
+ *     UChar32 c;
+ *     while(it.hasNext()) {
+ *         c=it.next32PostInc();
+ *         // use c
+ *     }
+ * }
+ *
+ * void function1(ForwardCharacterIterator &it) {
+ *     char16_t c;
+ *     while((c=it.nextPostInc())!=ForwardCharacterIterator::DONE) {
+ *         // use c
+ *      }
+ *  }
+ * \endcode
+ * </p>
+ *
+ * @stable ICU 2.0
+ */
+class U_COMMON_API ForwardCharacterIterator : public UObject {
+public:
+    /**
+     * Value returned by most of ForwardCharacterIterator's functions
+     * when the iterator has reached the limits of its iteration.
+     * @stable ICU 2.0
+     */
+    enum { DONE = 0xffff };
+    
+    /**
+     * Destructor.  
+     * @stable ICU 2.0
+     */
+    virtual ~ForwardCharacterIterator();
+    
+    /**
+     * Returns true when both iterators refer to the same
+     * character in the same character-storage object.  
+     * @param that The ForwardCharacterIterator to be compared for equality
+     * @return true when both iterators refer to the same
+     * character in the same character-storage object
+     * @stable ICU 2.0
+     */
+    virtual UBool operator==(const ForwardCharacterIterator& that) const = 0;
+    
+    /**
+     * Returns true when the iterators refer to different
+     * text-storage objects, or to different characters in the
+     * same text-storage object.  
+     * @param that The ForwardCharacterIterator to be compared for inequality
+     * @return true when the iterators refer to different
+     * text-storage objects, or to different characters in the
+     * same text-storage object
+     * @stable ICU 2.0
+     */
+    inline UBool operator!=(const ForwardCharacterIterator& that) const;
+    
+    /**
+     * Generates a hash code for this iterator.  
+     * @return the hash code.
+     * @stable ICU 2.0
+     */
+    virtual int32_t hashCode(void) const = 0;
+    
+    /**
+     * Returns a UClassID for this ForwardCharacterIterator ("poor man's
+     * RTTI").<P> Despite the fact that this function is public,
+     * DO NOT CONSIDER IT PART OF CHARACTERITERATOR'S API! 
+     * @return a UClassID for this ForwardCharacterIterator 
+     * @stable ICU 2.0
+     */
+    virtual UClassID getDynamicClassID(void) const = 0;
+    
+    /**
+     * Gets the current code unit for returning and advances to the next code unit
+     * in the iteration range
+     * (toward endIndex()).  If there are
+     * no more code units to return, returns DONE.
+     * @return the current code unit.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         nextPostInc(void) = 0;
+    
+    /**
+     * Gets the current code point for returning and advances to the next code point
+     * in the iteration range
+     * (toward endIndex()).  If there are
+     * no more code points to return, returns DONE.
+     * @return the current code point.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       next32PostInc(void) = 0;
+    
+    /**
+     * Returns false if there are no more code units or code points
+     * at or after the current position in the iteration range.
+     * This is used with nextPostInc() or next32PostInc() in forward
+     * iteration.
+     * @returns false if there are no more code units or code points
+     * at or after the current position in the iteration range.
+     * @stable ICU 2.0
+     */
+    virtual UBool        hasNext() = 0;
+    
+protected:
+    /** Default constructor to be overridden in the implementing class. @stable ICU 2.0*/
+    ForwardCharacterIterator();
+    
+    /** Copy constructor to be overridden in the implementing class. @stable ICU 2.0*/
+    ForwardCharacterIterator(const ForwardCharacterIterator &other);
+    
+    /**
+     * Assignment operator to be overridden in the implementing class.
+     * @stable ICU 2.0
+     */
+    ForwardCharacterIterator &operator=(const ForwardCharacterIterator&) { return *this; }
+};
+
+/**
+ * Abstract class that defines an API for iteration
+ * on text objects.
+ * This is an interface for forward and backward iteration
+ * and random access into a text object.
+ *
+ * <p>The API provides backward compatibility to the Java and older ICU
+ * CharacterIterator classes but extends them significantly:
+ * <ol>
+ * <li>CharacterIterator is now a subclass of ForwardCharacterIterator.</li>
+ * <li>While the old API functions provided forward iteration with
+ *     "pre-increment" semantics, the new one also provides functions
+ *     with "post-increment" semantics. They are more efficient and should
+ *     be the preferred iterator functions for new implementations.
+ *     The backward iteration always had "pre-decrement" semantics, which
+ *     are efficient.</li>
+ * <li>Just like ForwardCharacterIterator, it provides access to
+ *     both code units and code points. Code point access versions are available
+ *     for the old and the new iteration semantics.</li>
+ * <li>There are new functions for setting and moving the current position
+ *     without returning a character, for efficiency.</li>
+ * </ol>
+ *
+ * See ForwardCharacterIterator for examples for using the new forward iteration
+ * functions. For backward iteration, there is also a hasPrevious() function
+ * that can be used analogously to hasNext().
+ * The old functions work as before and are shown below.</p>
+ *
+ * <p>Examples for some of the new functions:</p>
+ *
+ * Forward iteration with hasNext():
+ * \code
+ * void forward1(CharacterIterator &it) {
+ *     UChar32 c;
+ *     for(it.setToStart(); it.hasNext();) {
+ *         c=it.next32PostInc();
+ *         // use c
+ *     }
+ *  }
+ * \endcode
+ * Forward iteration more similar to loops with the old forward iteration,
+ * showing a way to convert simple for() loops:
+ * \code
+ * void forward2(CharacterIterator &it) {
+ *     char16_t c;
+ *     for(c=it.firstPostInc(); c!=CharacterIterator::DONE; c=it.nextPostInc()) {
+ *          // use c
+ *      }
+ * }
+ * \endcode
+ * Backward iteration with setToEnd() and hasPrevious():
+ * \code
+ *  void backward1(CharacterIterator &it) {
+ *      UChar32 c;
+ *      for(it.setToEnd(); it.hasPrevious();) {
+ *         c=it.previous32();
+ *          // use c
+ *      }
+ *  }
+ * \endcode
+ * Backward iteration with a more traditional for() loop:
+ * \code
+ * void backward2(CharacterIterator &it) {
+ *     char16_t c;
+ *     for(c=it.last(); c!=CharacterIterator::DONE; c=it.previous()) {
+ *         // use c
+ *      }
+ *  }
+ * \endcode
+ *
+ * Example for random access:
+ * \code
+ *  void random(CharacterIterator &it) {
+ *      // set to the third code point from the beginning
+ *      it.move32(3, CharacterIterator::kStart);
+ *      // get a code point from here without moving the position
+ *      UChar32 c=it.current32();
+ *      // get the position
+ *      int32_t pos=it.getIndex();
+ *      // get the previous code unit
+ *      char16_t u=it.previous();
+ *      // move back one more code unit
+ *      it.move(-1, CharacterIterator::kCurrent);
+ *      // set the position back to where it was
+ *      // and read the same code point c and move beyond it
+ *      it.setIndex(pos);
+ *      if(c!=it.next32PostInc()) {
+ *          exit(1); // CharacterIterator inconsistent
+ *      }
+ *  }
+ * \endcode
+ *
+ * <p>Examples, especially for the old API:</p>
+ *
+ * Function processing characters, in this example simple output
+ * <pre>
+ * \code
+ *  void processChar( char16_t c )
+ *  {
+ *      cout << " " << c;
+ *  }
+ * \endcode
+ * </pre>
+ * Traverse the text from start to finish
+ * <pre> 
+ * \code
+ *  void traverseForward(CharacterIterator& iter)
+ *  {
+ *      for(char16_t c = iter.first(); c != CharacterIterator.DONE; c = iter.next()) {
+ *          processChar(c);
+ *      }
+ *  }
+ * \endcode
+ * </pre>
+ * Traverse the text backwards, from end to start
+ * <pre>
+ * \code
+ *  void traverseBackward(CharacterIterator& iter)
+ *  {
+ *      for(char16_t c = iter.last(); c != CharacterIterator.DONE; c = iter.previous()) {
+ *          processChar(c);
+ *      }
+ *  }
+ * \endcode
+ * </pre>
+ * Traverse both forward and backward from a given position in the text. 
+ * Calls to notBoundary() in this example represents some additional stopping criteria.
+ * <pre>
+ * \code
+ * void traverseOut(CharacterIterator& iter, int32_t pos)
+ * {
+ *      char16_t c;
+ *      for (c = iter.setIndex(pos);
+ *      c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c));
+ *          c = iter.next()) {}
+ *      int32_t end = iter.getIndex();
+ *      for (c = iter.setIndex(pos);
+ *          c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c));
+ *          c = iter.previous()) {}
+ *      int32_t start = iter.getIndex() + 1;
+ *  
+ *      cout << "start: " << start << " end: " << end << endl;
+ *      for (c = iter.setIndex(start); iter.getIndex() < end; c = iter.next() ) {
+ *          processChar(c);
+ *     }
+ *  }
+ * \endcode
+ * </pre>
+ * Creating a StringCharacterIterator and calling the test functions
+ * <pre>
+ * \code
+ *  void CharacterIterator_Example( void )
+ *   {
+ *       cout << endl << "===== CharacterIterator_Example: =====" << endl;
+ *       UnicodeString text("Ein kleiner Satz.");
+ *       StringCharacterIterator iterator(text);
+ *       cout << "----- traverseForward: -----------" << endl;
+ *       traverseForward( iterator );
+ *       cout << endl << endl << "----- traverseBackward: ----------" << endl;
+ *       traverseBackward( iterator );
+ *       cout << endl << endl << "----- traverseOut: ---------------" << endl;
+ *       traverseOut( iterator, 7 );
+ *       cout << endl << endl << "-----" << endl;
+ *   }
+ * \endcode
+ * </pre>
+ *
+ * @stable ICU 2.0
+ */
+class U_COMMON_API CharacterIterator : public ForwardCharacterIterator {
+public:
+    /**
+     * Origin enumeration for the move() and move32() functions.
+     * @stable ICU 2.0
+     */
+    enum EOrigin { kStart, kCurrent, kEnd };
+
+    /**
+     * Destructor.
+     * @stable ICU 2.0
+     */
+    virtual ~CharacterIterator();
+
+    /**
+     * Returns a pointer to a new CharacterIterator of the same
+     * concrete class as this one, and referring to the same
+     * character in the same text-storage object as this one.  The
+     * caller is responsible for deleting the new clone.  
+     * @return a pointer to a new CharacterIterator
+     * @stable ICU 2.0
+     */
+    virtual CharacterIterator* clone() const = 0;
+
+    /**
+     * Sets the iterator to refer to the first code unit in its
+     * iteration range, and returns that code unit.
+     * This can be used to begin an iteration with next().
+     * @return the first code unit in its iteration range.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         first(void) = 0;
+
+    /**
+     * Sets the iterator to refer to the first code unit in its
+     * iteration range, returns that code unit, and moves the position
+     * to the second code unit. This is an alternative to setToStart()
+     * for forward iteration with nextPostInc().
+     * @return the first code unit in its iteration range.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         firstPostInc(void);
+
+    /**
+     * Sets the iterator to refer to the first code point in its
+     * iteration range, and returns that code unit,
+     * This can be used to begin an iteration with next32().
+     * Note that an iteration with next32PostInc(), beginning with,
+     * e.g., setToStart() or firstPostInc(), is more efficient.
+     * @return the first code point in its iteration range.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       first32(void) = 0;
+
+    /**
+     * Sets the iterator to refer to the first code point in its
+     * iteration range, returns that code point, and moves the position
+     * to the second code point. This is an alternative to setToStart()
+     * for forward iteration with next32PostInc().
+     * @return the first code point in its iteration range.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       first32PostInc(void);
+
+    /**
+     * Sets the iterator to refer to the first code unit or code point in its
+     * iteration range. This can be used to begin a forward
+     * iteration with nextPostInc() or next32PostInc().
+     * @return the start position of the iteration range
+     * @stable ICU 2.0
+     */
+    inline int32_t    setToStart();
+
+    /**
+     * Sets the iterator to refer to the last code unit in its
+     * iteration range, and returns that code unit.
+     * This can be used to begin an iteration with previous().
+     * @return the last code unit.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         last(void) = 0;
+        
+    /**
+     * Sets the iterator to refer to the last code point in its
+     * iteration range, and returns that code unit.
+     * This can be used to begin an iteration with previous32().
+     * @return the last code point.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       last32(void) = 0;
+
+    /**
+     * Sets the iterator to the end of its iteration range, just behind
+     * the last code unit or code point. This can be used to begin a backward
+     * iteration with previous() or previous32().
+     * @return the end position of the iteration range
+     * @stable ICU 2.0
+     */
+    inline int32_t    setToEnd();
+
+    /**
+     * Sets the iterator to refer to the "position"-th code unit
+     * in the text-storage object the iterator refers to, and
+     * returns that code unit.  
+     * @param position the "position"-th code unit in the text-storage object
+     * @return the "position"-th code unit.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         setIndex(int32_t position) = 0;
+
+    /**
+     * Sets the iterator to refer to the beginning of the code point
+     * that contains the "position"-th code unit
+     * in the text-storage object the iterator refers to, and
+     * returns that code point.
+     * The current position is adjusted to the beginning of the code point
+     * (its first code unit).
+     * @param position the "position"-th code unit in the text-storage object
+     * @return the "position"-th code point.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       setIndex32(int32_t position) = 0;
+
+    /**
+     * Returns the code unit the iterator currently refers to. 
+     * @return the current code unit. 
+     * @stable ICU 2.0
+     */
+    virtual char16_t         current(void) const = 0;
+        
+    /**
+     * Returns the code point the iterator currently refers to.  
+     * @return the current code point.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       current32(void) const = 0;
+        
+    /**
+     * Advances to the next code unit in the iteration range
+     * (toward endIndex()), and returns that code unit.  If there are
+     * no more code units to return, returns DONE.
+     * @return the next code unit.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         next(void) = 0;
+        
+    /**
+     * Advances to the next code point in the iteration range
+     * (toward endIndex()), and returns that code point.  If there are
+     * no more code points to return, returns DONE.
+     * Note that iteration with "pre-increment" semantics is less
+     * efficient than iteration with "post-increment" semantics
+     * that is provided by next32PostInc().
+     * @return the next code point.
+     * @stable ICU 2.0
+     */
+    virtual UChar32       next32(void) = 0;
+        
+    /**
+     * Advances to the previous code unit in the iteration range
+     * (toward startIndex()), and returns that code unit.  If there are
+     * no more code units to return, returns DONE.  
+     * @return the previous code unit.
+     * @stable ICU 2.0
+     */
+    virtual char16_t         previous(void) = 0;
+
+    /**
+     * Advances to the previous code point in the iteration range
+     * (toward startIndex()), and returns that code point.  If there are
+     * no more code points to return, returns DONE. 
+     * @return the previous code point. 
+     * @stable ICU 2.0
+     */
+    virtual UChar32       previous32(void) = 0;
+
+    /**
+     * Returns false if there are no more code units or code points
+     * before the current position in the iteration range.
+     * This is used with previous() or previous32() in backward
+     * iteration.
+     * @return false if there are no more code units or code points
+     * before the current position in the iteration range, return true otherwise.
+     * @stable ICU 2.0
+     */
+    virtual UBool        hasPrevious() = 0;
+
+    /**
+     * Returns the numeric index in the underlying text-storage
+     * object of the character returned by first().  Since it's
+     * possible to create an iterator that iterates across only
+     * part of a text-storage object, this number isn't
+     * necessarily 0.  
+     * @returns the numeric index in the underlying text-storage
+     * object of the character returned by first().
+     * @stable ICU 2.0
+     */
+    inline int32_t       startIndex(void) const;
+        
+    /**
+     * Returns the numeric index in the underlying text-storage
+     * object of the position immediately BEYOND the character
+     * returned by last().  
+     * @return the numeric index in the underlying text-storage
+     * object of the position immediately BEYOND the character
+     * returned by last().
+     * @stable ICU 2.0
+     */
+    inline int32_t       endIndex(void) const;
+        
+    /**
+     * Returns the numeric index in the underlying text-storage
+     * object of the character the iterator currently refers to
+     * (i.e., the character returned by current()).  
+     * @return the numeric index in the text-storage object of 
+     * the character the iterator currently refers to
+     * @stable ICU 2.0
+     */
+    inline int32_t       getIndex(void) const;
+
+    /**
+     * Returns the length of the entire text in the underlying
+     * text-storage object.
+     * @return the length of the entire text in the text-storage object
+     * @stable ICU 2.0
+     */
+    inline int32_t           getLength() const;
+
+    /**
+     * Moves the current position relative to the start or end of the
+     * iteration range, or relative to the current position itself.
+     * The movement is expressed in numbers of code units forward
+     * or backward by specifying a positive or negative delta.
+     * @param delta the position relative to origin. A positive delta means forward;
+     * a negative delta means backward.
+     * @param origin Origin enumeration {kStart, kCurrent, kEnd}
+     * @return the new position
+     * @stable ICU 2.0
+     */
+    virtual int32_t      move(int32_t delta, EOrigin origin) = 0;
+
+    /**
+     * Moves the current position relative to the start or end of the
+     * iteration range, or relative to the current position itself.
+     * The movement is expressed in numbers of code points forward
+     * or backward by specifying a positive or negative delta.
+     * @param delta the position relative to origin. A positive delta means forward;
+     * a negative delta means backward.
+     * @param origin Origin enumeration {kStart, kCurrent, kEnd}
+     * @return the new position
+     * @stable ICU 2.0
+     */
+#ifdef move32
+     // One of the system headers right now is sometimes defining a conflicting macro we don't use
+#undef move32
+#endif
+    virtual int32_t      move32(int32_t delta, EOrigin origin) = 0;
+
+    /**
+     * Copies the text under iteration into the UnicodeString
+     * referred to by "result".  
+     * @param result Receives a copy of the text under iteration.  
+     * @stable ICU 2.0
+     */
+    virtual void            getText(UnicodeString&  result) = 0;
+
+protected:
+    /**
+     * Empty constructor.
+     * @stable ICU 2.0
+     */
+    CharacterIterator();
+
+    /**
+     * Constructor, just setting the length field in this base class.
+     * @stable ICU 2.0
+     */
+    CharacterIterator(int32_t length);
+
+    /**
+     * Constructor, just setting the length and position fields in this base class.
+     * @stable ICU 2.0
+     */
+    CharacterIterator(int32_t length, int32_t position);
+
+    /**
+     * Constructor, just setting the length, start, end, and position fields in this base class.
+     * @stable ICU 2.0
+     */
+    CharacterIterator(int32_t length, int32_t textBegin, int32_t textEnd, int32_t position);
+  
+    /**
+     * Copy constructor.
+     *
+     * @param that The CharacterIterator to be copied
+     * @stable ICU 2.0
+     */
+    CharacterIterator(const CharacterIterator &that);
+
+    /**
+     * Assignment operator.  Sets this CharacterIterator to have the same behavior,
+     * as the one passed in.
+     * @param that The CharacterIterator passed in.
+     * @return the newly set CharacterIterator.
+     * @stable ICU 2.0
+     */
+    CharacterIterator &operator=(const CharacterIterator &that);
+
+    /**
+     * Base class text length field.
+     * Necessary this for correct getText() and hashCode().
+     * @stable ICU 2.0
+     */
+    int32_t textLength;
+
+    /**
+     * Base class field for the current position.
+     * @stable ICU 2.0
+     */
+    int32_t  pos;
+
+    /**
+     * Base class field for the start of the iteration range.
+     * @stable ICU 2.0
+     */
+    int32_t  begin;
+
+    /**
+     * Base class field for the end of the iteration range.
+     * @stable ICU 2.0
+     */
+    int32_t  end;
+};
+
+inline UBool
+ForwardCharacterIterator::operator!=(const ForwardCharacterIterator& that) const {
+    return !operator==(that);
+}
+
+inline int32_t
+CharacterIterator::setToStart() {
+    return move(0, kStart);
+}
+
+inline int32_t
+CharacterIterator::setToEnd() {
+    return move(0, kEnd);
+}
+
+inline int32_t
+CharacterIterator::startIndex(void) const {
+    return begin;
+}
+
+inline int32_t
+CharacterIterator::endIndex(void) const {
+    return end;
+}
+
+inline int32_t
+CharacterIterator::getIndex(void) const {
+    return pos;
+}
+
+inline int32_t
+CharacterIterator::getLength(void) const {
+    return textLength;
+}
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/dbbi.h b/thirdparty/icu4c/common/unicode/dbbi.h
new file mode 100644
index 0000000000..3de9cc3814
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/dbbi.h
@@ -0,0 +1,48 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2006,2013 IBM Corp. All rights reserved.
+**********************************************************************
+*   Date        Name        Description
+*   12/1/99    rgillam     Complete port from Java.
+*   01/13/2000 helena      Added UErrorCode to ctors.
+**********************************************************************
+*/
+
+#ifndef DBBI_H
+#define DBBI_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/rbbi.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/**
+ * \file
+ * \brief C++ API: Dictionary Based Break Iterator
+ */
+ 
+U_NAMESPACE_BEGIN
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * An obsolete subclass of RuleBasedBreakIterator. Handling of dictionary-
+ * based break iteration has been folded into the base class. This class
+ * is deprecated as of ICU 3.6.
+ * @deprecated ICU 3.6
+ */
+typedef RuleBasedBreakIterator DictionaryBasedBreakIterator;
+
+#endif  /* U_HIDE_DEPRECATED_API */
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/docmain.h b/thirdparty/icu4c/common/unicode/docmain.h
new file mode 100644
index 0000000000..b7984ada03
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/docmain.h
@@ -0,0 +1,232 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/********************************************************************
+ * COPYRIGHT: 
+ * Copyright (c) 1997-2012, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ *
+ *  FILE NAME: DOCMAIN.h
+ *
+ *   Date          Name        Description
+ *   12/11/2000    Ram        Creation.
+ */
+
+/**
+ * \file
+ * \brief (Non API- contains Doxygen definitions)
+ *
+ * This file contains documentation for Doxygen and doesnot have
+ * any significance with respect to C or C++ API
+ */
+
+/*! \mainpage
+ *
+ * \section API API Reference Usage
+ * 
+ * <h3>C++ Programmers:</h3>
+ * <p>Use <a href="hierarchy.html">Class Hierarchy</a> or <a href="classes.html"> Alphabetical List </a>
+ * or <a href="annotated.html"> Compound List</a>
+ * to find the class you are interested in. For example, to find BreakIterator,
+ * you can go to the <a href="classes.html"> Alphabetical List</a>, then click on
+ * "BreakIterator". Once you are at the class, you will find an inheritance
+ * chart, a list of the public members, a detailed description of the class,
+ * then detailed member descriptions.</p>
+ * 
+ * <h3>C Programmers:</h3>
+ * <p>Use <a href="#Module">Module List</a> or <a href="globals_u.html">File Members</a>
+ * to find a list of all the functions and constants.
+ * For example, to find BreakIterator functions you would click on
+ * <a href="files.html"> File List</a>,
+ * then find "ubrk.h" and click on it. You will find descriptions of Defines,
+ * Typedefs, Enumerations, and Functions, with detailed descriptions below.
+ * If you want to find a specific function, such as ubrk_next(), then click
+ * first on <a href="globals.html"> File Members</a>, then use your browser
+ * Find dialog to search for "ubrk_next()".</p>
+ *
+ *
+ * <h3>API References for Previous Releases</h3>
+ * <p>The API References for each release of ICU are also available as
+ * a zip file from the ICU 
+ * <a href="http://site.icu-project.org/download">download page</a>.</p>
+ *
+ * <hr>
+ *
+ * <h2>Architecture (User's Guide)</h2>
+ * <ul>
+ *   <li><a href="https://unicode-org.github.io/icu/userguide/">Introduction</a></li>
+ *   <li><a href="https://unicode-org.github.io/icu/userguide/i18n">Internationalization</a></li>
+ *   <li><a href="https://unicode-org.github.io/icu/userguide/design">Locale Model, Multithreading, Error Handling, etc.</a></li>
+ *   <li><a href="https://unicode-org.github.io/icu/userguide/conversion">Conversion</a></li>
+ * </ul>
+ *
+ * <hr>
+ *\htmlonly <h2><a NAME="Module">Module List</a></h2> \endhtmlonly
+ * <table border="1" cols="3" align="center">
+ *   <tr>
+ *     <td><strong>Module Name</strong></td>
+ *     <td><strong>C</strong></td>
+ *     <td><strong>C++</strong></td>
+ *   </tr>
+ *   <tr>
+ *     <td>Basic Types and Constants</td>
+ *     <td>utypes.h</td>
+ *     <td>utypes.h</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Strings and Character Iteration</td>
+ *     <td>ustring.h, utf8.h, utf16.h, UText, UCharIterator</td>
+ *     <td>icu::UnicodeString, icu::CharacterIterator, icu::Appendable, icu::StringPiece,icu::ByteSink</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Unicode Character<br/>Properties and Names</td>
+ *     <td>uchar.h, uscript.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Sets of Unicode Code Points and Strings</td>
+ *     <td>uset.h</td>
+ *     <td>icu::UnicodeSet</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Maps from Unicode Code Points to Integer Values</td>
+ *     <td>ucptrie.h, umutablecptrie.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Maps from Strings to Integer Values</td>
+ *     <td>(no C API)</td>
+ *     <td>icu::BytesTrie, icu::UCharsTrie</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Codepage Conversion</td>
+ *     <td>ucnv.h, ucnvsel.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Codepage Detection</td>
+ *     <td>ucsdet.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Unicode Text Compression</td>
+ *     <td>ucnv.h<br/>(encoding name "SCSU" or "BOCU-1")</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Locales </td>
+ *     <td>uloc.h</a></td>
+ *     <td>icu::Locale, icu::LocaleBuilder, icu::LocaleMatcher</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Resource Bundles</td>
+ *     <td>ures.h</td>
+ *     <td>icu::ResourceBundle</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Normalization</td>
+ *     <td>unorm2.h</td>
+ *     <td>icu::Normalizer2</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Calendars</td>
+ *     <td>ucal.h</td>
+ *     <td>icu::Calendar</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Date and Time Formatting</td>
+ *     <td>udat.h</td>
+ *     <td>icu::DateFormat</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Message Formatting</td>
+ *     <td>umsg.h</td>
+ *     <td>icu::MessageFormat</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Number Formatting<br/>(includes currency and unit formatting)</td>
+ *     <td>unumberformatter.h, unum.h</td>
+ *     <td>icu::number::NumberFormatter (ICU 60+) or icu::NumberFormat (older versions)</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Number Range Formatting<br />(includes currency and unit ranges)</td>
+ *     <td>unumberrangeformatter.h</td>
+ *     <td>icu::number::NumberRangeFormatter</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Number Spellout<br/>(Rule Based Number Formatting)</td>
+ *     <td>unum.h<br/>(use UNUM_SPELLOUT)</td>
+ *     <td>icu::RuleBasedNumberFormat</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Text Transformation<br/>(Transliteration)</td>
+ *     <td>utrans.h</td>
+ *     <td>icu::Transliterator</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Bidirectional Algorithm</td>
+ *     <td>ubidi.h, ubiditransform.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Arabic Shaping</td>
+ *     <td>ushape.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Collation</td>
+ *     <td>ucol.h</td>
+ *     <td>icu::Collator</td>
+ *   </tr>
+ *   <tr>
+ *     <td>String Searching</td>
+ *     <td>usearch.h</td>
+ *     <td>icu::StringSearch</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Index Characters/<br/>Bucketing for Sorted Lists</td>
+ *     <td>(no C API)</td>
+ *     <td>icu::AlphabeticIndex</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Text Boundary Analysis<br/>(Break Iteration)</td>
+ *     <td>ubrk.h</td>
+ *     <td>icu::BreakIterator</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Regular Expressions</td>
+ *     <td>uregex.h</td>
+ *     <td>icu::RegexPattern, icu::RegexMatcher</td>
+ *   </tr>
+ *   <tr>
+ *     <td>StringPrep</td>
+ *     <td>usprep.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>International Domain Names in Applications:<br/>
+ *         UTS #46 in C/C++, IDNA2003 only via C API</td>
+ *     <td>uidna.h</td>
+ *     <td>idna.h</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Identifier Spoofing & Confusability</td>
+ *     <td>uspoof.h</td>
+ *     <td>C API</td>
+ *   <tr>
+ *     <td>Universal Time Scale</td>
+ *     <td>utmscale.h</td>
+ *     <td>C API</td>
+ *   </tr>
+ *   <tr>
+ *     <td>Paragraph Layout / Complex Text Layout</td>
+ *     <td>playout.h</td>
+ *     <td>icu::ParagraphLayout</td>
+ *   </tr>
+ *   <tr>
+ *     <td>ICU I/O</td>
+ *     <td>ustdio.h</td>
+ *     <td>ustream.h</td>
+ *   </tr>
+ * </table>
+ * <i>This main page is generated from docmain.h</i>
+ */
diff --git a/thirdparty/icu4c/common/unicode/dtintrv.h b/thirdparty/icu4c/common/unicode/dtintrv.h
new file mode 100644
index 0000000000..4f4b6bf7f4
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/dtintrv.h
@@ -0,0 +1,164 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 2008-2009, International Business Machines Corporation and
+* others. All Rights Reserved.
+*******************************************************************************
+*
+* File DTINTRV.H 
+*
+*******************************************************************************
+*/
+
+#ifndef __DTINTRV_H__
+#define __DTINTRV_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: Date Interval data type
+ */
+
+U_NAMESPACE_BEGIN
+
+
+/**
+ * This class represents a date interval.
+ * It is a pair of UDate representing from UDate 1 to UDate 2.
+ * @stable ICU 4.0
+**/
+class U_COMMON_API DateInterval : public UObject {
+public:
+
+    /** 
+     * Construct a DateInterval given a from date and a to date.
+     * @param fromDate  The from date in date interval.
+     * @param toDate    The to date in date interval.
+     * @stable ICU 4.0
+     */
+    DateInterval(UDate fromDate, UDate toDate);
+
+    /**
+     * destructor
+     * @stable ICU 4.0
+     */
+    virtual ~DateInterval();
+ 
+    /** 
+     * Get the from date.
+     * @return  the from date in dateInterval.
+     * @stable ICU 4.0
+     */
+    inline UDate getFromDate() const;
+
+    /** 
+     * Get the to date.
+     * @return  the to date in dateInterval.
+     * @stable ICU 4.0
+     */
+    inline UDate getToDate() const;
+
+
+    /**
+     * Return the class ID for this class. This is useful only for comparing to
+     * a return value from getDynamicClassID(). For example:
+     * <pre>
+     * .   Base* polymorphic_pointer = createPolymorphicObject();
+     * .   if (polymorphic_pointer->getDynamicClassID() ==
+     * .       derived::getStaticClassID()) ...
+     * </pre>
+     * @return          The class ID for all objects of this class.
+     * @stable ICU 4.0
+     */
+    static UClassID U_EXPORT2 getStaticClassID(void);
+
+    /**
+     * Returns a unique class ID POLYMORPHICALLY. Pure virtual override. This
+     * method is to implement a simple version of RTTI, since not all C++
+     * compilers support genuine RTTI. Polymorphic operator==() and clone()
+     * methods call this method.
+     *
+     * @return          The class ID for this object. All objects of a
+     *                  given class have the same class ID.  Objects of
+     *                  other classes have different class IDs.
+     * @stable ICU 4.0
+     */
+    virtual UClassID getDynamicClassID(void) const;
+
+    
+    /**
+     * Copy constructor.
+     * @stable ICU 4.0
+     */
+    DateInterval(const DateInterval& other);
+
+    /**
+     * Default assignment operator
+     * @stable ICU 4.0
+     */
+    DateInterval& operator=(const DateInterval&);
+
+    /**
+     * Equality operator.
+     * @return true if the two DateIntervals are the same
+     * @stable ICU 4.0
+     */
+    virtual UBool operator==(const DateInterval& other) const;
+
+    /**
+     * Non-equality operator
+     * @return true if the two DateIntervals are not the same
+     * @stable ICU 4.0
+     */
+    inline UBool operator!=(const DateInterval& other) const;
+
+
+    /**
+     * clone this object. 
+     * The caller owns the result and should delete it when done.
+     * @return a cloned DateInterval
+     * @stable ICU 4.0
+     */
+     virtual DateInterval* clone() const;
+
+private:
+    /** 
+     * Default constructor, not implemented.
+     */
+    DateInterval();
+
+    UDate fromDate;
+    UDate toDate;
+
+} ;// end class DateInterval
+
+
+inline UDate 
+DateInterval::getFromDate() const { 
+    return fromDate; 
+}
+
+
+inline UDate 
+DateInterval::getToDate() const { 
+    return toDate; 
+}
+
+
+inline UBool 
+DateInterval::operator!=(const DateInterval& other) const { 
+    return ( !operator==(other) );
+}
+
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/edits.h b/thirdparty/icu4c/common/unicode/edits.h
new file mode 100644
index 0000000000..bfa07fa676
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/edits.h
@@ -0,0 +1,531 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// edits.h
+// created: 2016dec30 Markus W. Scherer
+
+#ifndef __EDITS_H__
+#define __EDITS_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: C++ class Edits for low-level string transformations on styled text.
+ */
+
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+
+/**
+ * Records lengths of string edits but not replacement text. Supports replacements, insertions, deletions
+ * in linear progression. Does not support moving/reordering of text.
+ *
+ * There are two types of edits: <em>change edits</em> and <em>no-change edits</em>. Add edits to
+ * instances of this class using {@link #addReplace(int32_t, int32_t)} (for change edits) and
+ * {@link #addUnchanged(int32_t)} (for no-change edits). Change edits are retained with full granularity,
+ * whereas adjacent no-change edits are always merged together. In no-change edits, there is a one-to-one
+ * mapping between code points in the source and destination strings.
+ *
+ * After all edits have been added, instances of this class should be considered immutable, and an
+ * {@link Edits::Iterator} can be used for queries.
+ *
+ * There are four flavors of Edits::Iterator:
+ *
+ * <ul>
+ * <li>{@link #getFineIterator()} retains full granularity of change edits.
+ * <li>{@link #getFineChangesIterator()} retains full granularity of change edits, and when calling
+ * next() on the iterator, skips over no-change edits (unchanged regions).
+ * <li>{@link #getCoarseIterator()} treats adjacent change edits as a single edit. (Adjacent no-change
+ * edits are automatically merged during the construction phase.)
+ * <li>{@link #getCoarseChangesIterator()} treats adjacent change edits as a single edit, and when
+ * calling next() on the iterator, skips over no-change edits (unchanged regions).
+ * </ul>
+ *
+ * For example, consider the string "abcßDeF", which case-folds to "abcssdef". This string has the
+ * following fine edits:
+ * <ul>
+ * <li>abc ⇨ abc (no-change)
+ * <li>ß ⇨ ss (change)
+ * <li>D ⇨ d (change)
+ * <li>e ⇨ e (no-change)
+ * <li>F ⇨ f (change)
+ * </ul>
+ * and the following coarse edits (note how adjacent change edits get merged together):
+ * <ul>
+ * <li>abc ⇨ abc (no-change)
+ * <li>ßD ⇨ ssd (change)
+ * <li>e ⇨ e (no-change)
+ * <li>F ⇨ f (change)
+ * </ul>
+ *
+ * The "fine changes" and "coarse changes" iterators will step through only the change edits when their
+ * `Edits::Iterator::next()` methods are called. They are identical to the non-change iterators when
+ * their `Edits::Iterator::findSourceIndex()` or `Edits::Iterator::findDestinationIndex()`
+ * methods are used to walk through the string.
+ *
+ * For examples of how to use this class, see the test `TestCaseMapEditsIteratorDocs` in
+ * UCharacterCaseTest.java.
+ *
+ * An Edits object tracks a separate UErrorCode, but ICU string transformation functions
+ * (e.g., case mapping functions) merge any such errors into their API's UErrorCode.
+ *
+ * @stable ICU 59
+ */
+class U_COMMON_API Edits U_FINAL : public UMemory {
+public:
+    /**
+     * Constructs an empty object.
+     * @stable ICU 59
+     */
+    Edits() :
+            array(stackArray), capacity(STACK_CAPACITY), length(0), delta(0), numChanges(0),
+            errorCode_(U_ZERO_ERROR) {}
+    /**
+     * Copy constructor.
+     * @param other source edits
+     * @stable ICU 60
+     */
+    Edits(const Edits &other) :
+            array(stackArray), capacity(STACK_CAPACITY), length(other.length),
+            delta(other.delta), numChanges(other.numChanges),
+            errorCode_(other.errorCode_) {
+        copyArray(other);
+    }
+    /**
+     * Move constructor, might leave src empty.
+     * This object will have the same contents that the source object had.
+     * @param src source edits
+     * @stable ICU 60
+     */
+    Edits(Edits &&src) U_NOEXCEPT :
+            array(stackArray), capacity(STACK_CAPACITY), length(src.length),
+            delta(src.delta), numChanges(src.numChanges),
+            errorCode_(src.errorCode_) {
+        moveArray(src);
+    }
+
+    /**
+     * Destructor.
+     * @stable ICU 59
+     */
+    ~Edits();
+
+    /**
+     * Assignment operator.
+     * @param other source edits
+     * @return *this
+     * @stable ICU 60
+     */
+    Edits &operator=(const Edits &other);
+
+    /**
+     * Move assignment operator, might leave src empty.
+     * This object will have the same contents that the source object had.
+     * The behavior is undefined if *this and src are the same object.
+     * @param src source edits
+     * @return *this
+     * @stable ICU 60
+     */
+    Edits &operator=(Edits &&src) U_NOEXCEPT;
+
+    /**
+     * Resets the data but may not release memory.
+     * @stable ICU 59
+     */
+    void reset() U_NOEXCEPT;
+
+    /**
+     * Adds a no-change edit: a record for an unchanged segment of text.
+     * Normally called from inside ICU string transformation functions, not user code.
+     * @stable ICU 59
+     */
+    void addUnchanged(int32_t unchangedLength);
+    /**
+     * Adds a change edit: a record for a text replacement/insertion/deletion.
+     * Normally called from inside ICU string transformation functions, not user code.
+     * @stable ICU 59
+     */
+    void addReplace(int32_t oldLength, int32_t newLength);
+    /**
+     * Sets the UErrorCode if an error occurred while recording edits.
+     * Preserves older error codes in the outErrorCode.
+     * Normally called from inside ICU string transformation functions, not user code.
+     * @param outErrorCode Set to an error code if it does not contain one already
+     *                  and an error occurred while recording edits.
+     *                  Otherwise unchanged.
+     * @return true if U_FAILURE(outErrorCode)
+     * @stable ICU 59
+     */
+    UBool copyErrorTo(UErrorCode &outErrorCode) const;
+
+    /**
+     * How much longer is the new text compared with the old text?
+     * @return new length minus old length
+     * @stable ICU 59
+     */
+    int32_t lengthDelta() const { return delta; }
+    /**
+     * @return true if there are any change edits
+     * @stable ICU 59
+     */
+    UBool hasChanges() const { return numChanges != 0; }
+
+    /**
+     * @return the number of change edits
+     * @stable ICU 60
+     */
+    int32_t numberOfChanges() const { return numChanges; }
+
+    /**
+     * Access to the list of edits.
+     *
+     * At any moment in time, an instance of this class points to a single edit: a "window" into a span
+     * of the source string and the corresponding span of the destination string. The source string span
+     * starts at {@link #sourceIndex()} and runs for {@link #oldLength()} chars; the destination string
+     * span starts at {@link #destinationIndex()} and runs for {@link #newLength()} chars.
+     *
+     * The iterator can be moved between edits using the `next()`, `findSourceIndex(int32_t, UErrorCode &)`,
+     * and `findDestinationIndex(int32_t, UErrorCode &)` methods.
+     * Calling any of these methods mutates the iterator to make it point to the corresponding edit.
+     *
+     * For more information, see the documentation for {@link Edits}.
+     *
+     * @see getCoarseIterator
+     * @see getFineIterator
+     * @stable ICU 59
+     */
+    struct U_COMMON_API Iterator U_FINAL : public UMemory {
+        /**
+         * Default constructor, empty iterator.
+         * @stable ICU 60
+         */
+        Iterator() :
+                array(nullptr), index(0), length(0),
+                remaining(0), onlyChanges_(false), coarse(false),
+                dir(0), changed(false), oldLength_(0), newLength_(0),
+                srcIndex(0), replIndex(0), destIndex(0) {}
+        /**
+         * Copy constructor.
+         * @stable ICU 59
+         */
+        Iterator(const Iterator &other) = default;
+        /**
+         * Assignment operator.
+         * @stable ICU 59
+         */
+        Iterator &operator=(const Iterator &other) = default;
+
+        /**
+         * Advances the iterator to the next edit.
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return true if there is another edit
+         * @stable ICU 59
+         */
+        UBool next(UErrorCode &errorCode) { return next(onlyChanges_, errorCode); }
+
+        /**
+         * Moves the iterator to the edit that contains the source index.
+         * The source index may be found in a no-change edit
+         * even if normal iteration would skip no-change edits.
+         * Normal iteration can continue from a found edit.
+         *
+         * The iterator state before this search logically does not matter.
+         * (It may affect the performance of the search.)
+         *
+         * The iterator state after this search is undefined
+         * if the source index is out of bounds for the source string.
+         *
+         * @param i source index
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return true if the edit for the source index was found
+         * @stable ICU 59
+         */
+        UBool findSourceIndex(int32_t i, UErrorCode &errorCode) {
+            return findIndex(i, true, errorCode) == 0;
+        }
+
+        /**
+         * Moves the iterator to the edit that contains the destination index.
+         * The destination index may be found in a no-change edit
+         * even if normal iteration would skip no-change edits.
+         * Normal iteration can continue from a found edit.
+         *
+         * The iterator state before this search logically does not matter.
+         * (It may affect the performance of the search.)
+         *
+         * The iterator state after this search is undefined
+         * if the source index is out of bounds for the source string.
+         *
+         * @param i destination index
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return true if the edit for the destination index was found
+         * @stable ICU 60
+         */
+        UBool findDestinationIndex(int32_t i, UErrorCode &errorCode) {
+            return findIndex(i, false, errorCode) == 0;
+        }
+
+        /**
+         * Computes the destination index corresponding to the given source index.
+         * If the source index is inside a change edit (not at its start),
+         * then the destination index at the end of that edit is returned,
+         * since there is no information about index mapping inside a change edit.
+         *
+         * (This means that indexes to the start and middle of an edit,
+         * for example around a grapheme cluster, are mapped to indexes
+         * encompassing the entire edit.
+         * The alternative, mapping an interior index to the start,
+         * would map such an interval to an empty one.)
+         *
+         * This operation will usually but not always modify this object.
+         * The iterator state after this search is undefined.
+         *
+         * @param i source index
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return destination index; undefined if i is not 0..string length
+         * @stable ICU 60
+         */
+        int32_t destinationIndexFromSourceIndex(int32_t i, UErrorCode &errorCode);
+
+        /**
+         * Computes the source index corresponding to the given destination index.
+         * If the destination index is inside a change edit (not at its start),
+         * then the source index at the end of that edit is returned,
+         * since there is no information about index mapping inside a change edit.
+         *
+         * (This means that indexes to the start and middle of an edit,
+         * for example around a grapheme cluster, are mapped to indexes
+         * encompassing the entire edit.
+         * The alternative, mapping an interior index to the start,
+         * would map such an interval to an empty one.)
+         *
+         * This operation will usually but not always modify this object.
+         * The iterator state after this search is undefined.
+         *
+         * @param i destination index
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return source index; undefined if i is not 0..string length
+         * @stable ICU 60
+         */
+        int32_t sourceIndexFromDestinationIndex(int32_t i, UErrorCode &errorCode);
+
+        /**
+         * Returns whether the edit currently represented by the iterator is a change edit.
+         *
+         * @return true if this edit replaces oldLength() units with newLength() different ones.
+         *         false if oldLength units remain unchanged.
+         * @stable ICU 59
+         */
+        UBool hasChange() const { return changed; }
+
+        /**
+         * The length of the current span in the source string, which starts at {@link #sourceIndex}.
+         *
+         * @return the number of units in the original string which are replaced or remain unchanged.
+         * @stable ICU 59
+         */
+        int32_t oldLength() const { return oldLength_; }
+
+        /**
+         * The length of the current span in the destination string, which starts at
+         * {@link #destinationIndex}, or in the replacement string, which starts at
+         * {@link #replacementIndex}.
+         *
+         * @return the number of units in the modified string, if hasChange() is true.
+         *         Same as oldLength if hasChange() is false.
+         * @stable ICU 59
+         */
+        int32_t newLength() const { return newLength_; }
+
+        /**
+         * The start index of the current span in the source string; the span has length
+         * {@link #oldLength}.
+         *
+         * @return the current index into the source string
+         * @stable ICU 59
+         */
+        int32_t sourceIndex() const { return srcIndex; }
+
+        /**
+         * The start index of the current span in the replacement string; the span has length
+         * {@link #newLength}. Well-defined only if the current edit is a change edit.
+         *
+         * The *replacement string* is the concatenation of all substrings of the destination
+         * string corresponding to change edits.
+         *
+         * This method is intended to be used together with operations that write only replacement
+         * characters (e.g. operations specifying the \ref U_OMIT_UNCHANGED_TEXT option).
+         * The source string can then be modified in-place.
+         *
+         * @return the current index into the replacement-characters-only string,
+         *         not counting unchanged spans
+         * @stable ICU 59
+         */
+        int32_t replacementIndex() const {
+            // TODO: Throw an exception if we aren't in a change edit?
+            return replIndex;
+        }
+
+        /**
+         * The start index of the current span in the destination string; the span has length
+         * {@link #newLength}.
+         *
+         * @return the current index into the full destination string
+         * @stable ICU 59
+         */
+        int32_t destinationIndex() const { return destIndex; }
+
+#ifndef U_HIDE_INTERNAL_API
+        /**
+         * A string representation of the current edit represented by the iterator for debugging. You
+         * should not depend on the contents of the return string.
+         * @internal
+         */
+        UnicodeString& toString(UnicodeString& appendTo) const;
+#endif  // U_HIDE_INTERNAL_API
+
+    private:
+        friend class Edits;
+
+        Iterator(const uint16_t *a, int32_t len, UBool oc, UBool crs);
+
+        int32_t readLength(int32_t head);
+        void updateNextIndexes();
+        void updatePreviousIndexes();
+        UBool noNext();
+        UBool next(UBool onlyChanges, UErrorCode &errorCode);
+        UBool previous(UErrorCode &errorCode);
+        /** @return -1: error or i<0; 0: found; 1: i>=string length */
+        int32_t findIndex(int32_t i, UBool findSource, UErrorCode &errorCode);
+
+        const uint16_t *array;
+        int32_t index, length;
+        // 0 if we are not within compressed equal-length changes.
+        // Otherwise the number of remaining changes, including the current one.
+        int32_t remaining;
+        UBool onlyChanges_, coarse;
+
+        int8_t dir;  // iteration direction: back(<0), initial(0), forward(>0)
+        UBool changed;
+        int32_t oldLength_, newLength_;
+        int32_t srcIndex, replIndex, destIndex;
+    };
+
+    /**
+     * Returns an Iterator for coarse-grained change edits
+     * (adjacent change edits are treated as one).
+     * Can be used to perform simple string updates.
+     * Skips no-change edits.
+     * @return an Iterator that merges adjacent changes.
+     * @stable ICU 59
+     */
+    Iterator getCoarseChangesIterator() const {
+        return Iterator(array, length, true, true);
+    }
+
+    /**
+     * Returns an Iterator for coarse-grained change and no-change edits
+     * (adjacent change edits are treated as one).
+     * Can be used to perform simple string updates.
+     * Adjacent change edits are treated as one edit.
+     * @return an Iterator that merges adjacent changes.
+     * @stable ICU 59
+     */
+    Iterator getCoarseIterator() const {
+        return Iterator(array, length, false, true);
+    }
+
+    /**
+     * Returns an Iterator for fine-grained change edits
+     * (full granularity of change edits is retained).
+     * Can be used for modifying styled text.
+     * Skips no-change edits.
+     * @return an Iterator that separates adjacent changes.
+     * @stable ICU 59
+     */
+    Iterator getFineChangesIterator() const {
+        return Iterator(array, length, true, false);
+    }
+
+    /**
+     * Returns an Iterator for fine-grained change and no-change edits
+     * (full granularity of change edits is retained).
+     * Can be used for modifying styled text.
+     * @return an Iterator that separates adjacent changes.
+     * @stable ICU 59
+     */
+    Iterator getFineIterator() const {
+        return Iterator(array, length, false, false);
+    }
+
+    /**
+     * Merges the two input Edits and appends the result to this object.
+     *
+     * Consider two string transformations (for example, normalization and case mapping)
+     * where each records Edits in addition to writing an output string.<br>
+     * Edits ab reflect how substrings of input string a
+     * map to substrings of intermediate string b.<br>
+     * Edits bc reflect how substrings of intermediate string b
+     * map to substrings of output string c.<br>
+     * This function merges ab and bc such that the additional edits
+     * recorded in this object reflect how substrings of input string a
+     * map to substrings of output string c.
+     *
+     * If unrelated Edits are passed in where the output string of the first
+     * has a different length than the input string of the second,
+     * then a U_ILLEGAL_ARGUMENT_ERROR is reported.
+     *
+     * @param ab reflects how substrings of input string a
+     *     map to substrings of intermediate string b.
+     * @param bc reflects how substrings of intermediate string b
+     *     map to substrings of output string c.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return *this, with the merged edits appended
+     * @stable ICU 60
+     */
+    Edits &mergeAndAppend(const Edits &ab, const Edits &bc, UErrorCode &errorCode);
+
+private:
+    void releaseArray() U_NOEXCEPT;
+    Edits &copyArray(const Edits &other);
+    Edits &moveArray(Edits &src) U_NOEXCEPT;
+
+    void setLastUnit(int32_t last) { array[length - 1] = (uint16_t)last; }
+    int32_t lastUnit() const { return length > 0 ? array[length - 1] : 0xffff; }
+
+    void append(int32_t r);
+    UBool growArray();
+
+    static const int32_t STACK_CAPACITY = 100;
+    uint16_t *array;
+    int32_t capacity;
+    int32_t length;
+    int32_t delta;
+    int32_t numChanges;
+    UErrorCode errorCode_;
+    uint16_t stackArray[STACK_CAPACITY];
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __EDITS_H__
diff --git a/thirdparty/icu4c/common/unicode/enumset.h b/thirdparty/icu4c/common/unicode/enumset.h
new file mode 100644
index 0000000000..bde8c455c0
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/enumset.h
@@ -0,0 +1,69 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2012,2014 International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*/
+
+/**
+ * \file
+ * \brief C++: internal template EnumSet<>
+ */
+
+#ifndef ENUMSET_H
+#define ENUMSET_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/* Can't use #ifndef U_HIDE_INTERNAL_API for the entire EnumSet class, needed in .h file declarations */
+/**
+ * enum bitset for boolean fields. Similar to Java EnumSet<>. 
+ * Needs to range check. Used for private instance variables.
+ * @internal
+ * \cond
+ */
+template<typename T, uint32_t minValue, uint32_t limitValue>
+class EnumSet {
+public:
+    inline EnumSet() : fBools(0) {}
+    inline EnumSet(const EnumSet<T,minValue,limitValue>& other) : fBools(other.fBools) {}
+    inline ~EnumSet() {}
+#ifndef U_HIDE_INTERNAL_API
+    inline void clear() { fBools=0; }
+    inline void add(T toAdd) { set(toAdd, 1); }
+    inline void remove(T toRemove) { set(toRemove, 0); }
+    inline int32_t contains(T toCheck) const { return get(toCheck); }
+    inline void set(T toSet, int32_t v) { fBools=(fBools&(~flag(toSet)))|(v?(flag(toSet)):0); }
+    inline int32_t get(T toCheck) const { return (fBools & flag(toCheck))?1:0; }
+    inline UBool isValidEnum(T toCheck) const {  return (toCheck>=minValue&&toCheck<limitValue); }
+    inline UBool isValidValue(int32_t v) const { return (v==0||v==1); }
+    inline const EnumSet<T,minValue,limitValue>& operator=(const EnumSet<T,minValue,limitValue>& other) {
+        fBools = other.fBools;
+        return *this;
+    }
+  
+    inline uint32_t getAll() const {
+        return fBools; 
+    }
+#endif /* U_HIDE_INTERNAL_API */
+
+private:
+    inline uint32_t flag(T toCheck) const { return (1<<(toCheck-minValue)); }
+private:
+    uint32_t fBools;
+};
+
+/** \endcond */
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+#endif /* ENUMSET_H */
diff --git a/thirdparty/icu4c/common/unicode/errorcode.h b/thirdparty/icu4c/common/unicode/errorcode.h
new file mode 100644
index 0000000000..fe7b518323
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/errorcode.h
@@ -0,0 +1,144 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  errorcode.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009mar10
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __ERRORCODE_H__
+#define __ERRORCODE_H__
+
+/**
+ * \file 
+ * \brief C++ API: ErrorCode class intended to make it easier to use
+ *                 ICU C and C++ APIs from C++ user code.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Wrapper class for UErrorCode, with conversion operators for direct use
+ * in ICU C and C++ APIs.
+ * Intended to be used as a base class, where a subclass overrides
+ * the handleFailure() function so that it throws an exception,
+ * does an assert(), logs an error, etc.
+ * This is not an abstract base class. This class can be used and instantiated
+ * by itself, although it will be more useful when subclassed.
+ *
+ * Features:
+ * - The constructor initializes the internal UErrorCode to U_ZERO_ERROR,
+ *   removing one common source of errors.
+ * - Same use in C APIs taking a UErrorCode * (pointer)
+ *   and C++ taking UErrorCode & (reference) via conversion operators.
+ * - Possible automatic checking for success when it goes out of scope.
+ *
+ * Note: For automatic checking for success in the destructor, a subclass
+ * must implement such logic in its own destructor because the base class
+ * destructor cannot call a subclass function (like handleFailure()).
+ * The ErrorCode base class destructor does nothing.
+ *
+ * Note also: While it is possible for a destructor to throw an exception,
+ * it is generally unsafe to do so. This means that in a subclass the destructor
+ * and the handleFailure() function may need to take different actions.
+ *
+ * Sample code:
+ * \code
+ *   class IcuErrorCode: public icu::ErrorCode {
+ *   public:
+ *     virtual ~IcuErrorCode() {  // should be defined in .cpp as "key function"
+ *       // Safe because our handleFailure() does not throw exceptions.
+ *       if(isFailure()) { handleFailure(); }
+ *     }
+ *   protected:
+ *     virtual void handleFailure() const {
+ *       log_failure(u_errorName(errorCode));
+ *       exit(errorCode);
+ *     }
+ *   };
+ *   IcuErrorCode error_code;
+ *   UConverter *cnv = ucnv_open("Shift-JIS", error_code);
+ *   length = ucnv_fromUChars(dest, capacity, src, length, error_code);
+ *   ucnv_close(cnv);
+ *   // IcuErrorCode destructor checks for success.
+ * \endcode
+ *
+ * @stable ICU 4.2
+ */
+class U_COMMON_API ErrorCode: public UMemory {
+public:
+    /**
+     * Default constructor. Initializes its UErrorCode to U_ZERO_ERROR.
+     * @stable ICU 4.2
+     */
+    ErrorCode() : errorCode(U_ZERO_ERROR) {}
+    /** Destructor, does nothing. See class documentation for details. @stable ICU 4.2 */
+    virtual ~ErrorCode();
+    /** Conversion operator, returns a reference. @stable ICU 4.2 */
+    operator UErrorCode & () { return errorCode; }
+    /** Conversion operator, returns a pointer. @stable ICU 4.2 */
+    operator UErrorCode * () { return &errorCode; }
+    /** Tests for U_SUCCESS(). @stable ICU 4.2 */
+    UBool isSuccess() const { return U_SUCCESS(errorCode); }
+    /** Tests for U_FAILURE(). @stable ICU 4.2 */
+    UBool isFailure() const { return U_FAILURE(errorCode); }
+    /** Returns the UErrorCode value. @stable ICU 4.2 */
+    UErrorCode get() const { return errorCode; }
+    /** Sets the UErrorCode value. @stable ICU 4.2 */
+    void set(UErrorCode value) { errorCode=value; }
+    /** Returns the UErrorCode value and resets it to U_ZERO_ERROR. @stable ICU 4.2 */
+    UErrorCode reset();
+    /**
+     * Asserts isSuccess().
+     * In other words, this method checks for a failure code,
+     * and the base class handles it like this:
+     * \code
+     *   if(isFailure()) { handleFailure(); }
+     * \endcode
+     * @stable ICU 4.4
+     */
+    void assertSuccess() const;
+    /**
+     * Return a string for the UErrorCode value.
+     * The string will be the same as the name of the error code constant
+     * in the UErrorCode enum.
+     * @stable ICU 4.4
+     */
+    const char* errorName() const;
+
+protected:
+    /**
+     * Internal UErrorCode, accessible to subclasses.
+     * @stable ICU 4.2
+     */
+    UErrorCode errorCode;
+    /**
+     * Called by assertSuccess() if isFailure() is true.
+     * A subclass should override this function to deal with a failure code:
+     * Throw an exception, log an error, terminate the program, or similar.
+     * @stable ICU 4.2
+     */
+    virtual void handleFailure() const {}
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __ERRORCODE_H__
diff --git a/thirdparty/icu4c/common/unicode/filteredbrk.h b/thirdparty/icu4c/common/unicode/filteredbrk.h
new file mode 100644
index 0000000000..8b07e39ae1
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/filteredbrk.h
@@ -0,0 +1,152 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 1997-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*/
+
+#ifndef FILTEREDBRK_H
+#define FILTEREDBRK_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/brkiter.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION && !UCONFIG_NO_FILTERED_BREAK_ITERATION
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \file
+ * \brief C++ API: FilteredBreakIteratorBuilder
+ */
+
+/**
+ * The BreakIteratorFilter is used to modify the behavior of a BreakIterator
+ *  by constructing a new BreakIterator which suppresses certain segment boundaries.
+ *  See  http://www.unicode.org/reports/tr35/tr35-general.html#Segmentation_Exceptions .
+ *  For example, a typical English Sentence Break Iterator would break on the space
+ *  in the string "Mr. Smith" (resulting in two segments),
+ *  but with "Mr." as an exception, a filtered break iterator
+ *  would consider the string "Mr. Smith" to be a single segment.
+ *
+ * @stable ICU 56
+ */
+class U_COMMON_API FilteredBreakIteratorBuilder : public UObject {
+ public:
+  /**
+   *  destructor.
+   * @stable ICU 56
+   */
+  virtual ~FilteredBreakIteratorBuilder();
+
+  /**
+   * Construct a FilteredBreakIteratorBuilder based on rules in a locale.
+   * The rules are taken from CLDR exception data for the locale,
+   *  see http://www.unicode.org/reports/tr35/tr35-general.html#Segmentation_Exceptions
+   *  This is the equivalent of calling createInstance(UErrorCode&)
+   *    and then repeatedly calling addNoBreakAfter(...) with the contents
+   *    of the CLDR exception data.
+   * @param where the locale.
+   * @param status The error code.
+   * @return the new builder
+   * @stable ICU 56
+   */
+  static FilteredBreakIteratorBuilder *createInstance(const Locale& where, UErrorCode& status);
+
+#ifndef U_HIDE_DEPRECATED_API
+  /**
+   * This function has been deprecated in favor of createEmptyInstance, which has
+   * identical behavior.
+   * @param status The error code.
+   * @return the new builder
+   * @deprecated ICU 60 use createEmptyInstance instead
+   * @see createEmptyInstance()
+   */
+  static FilteredBreakIteratorBuilder *createInstance(UErrorCode &status);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+  /**
+   * Construct an empty FilteredBreakIteratorBuilder.
+   * In this state, it will not suppress any segment boundaries.
+   * @param status The error code.
+   * @return the new builder
+   * @stable ICU 60
+   */
+  static FilteredBreakIteratorBuilder *createEmptyInstance(UErrorCode &status);
+
+  /**
+   * Suppress a certain string from being the end of a segment.
+   * For example, suppressing "Mr.", then segments ending in "Mr." will not be returned
+   * by the iterator.
+   * @param string the string to suppress, such as "Mr."
+   * @param status error code
+   * @return returns true if the string was not present and now added,
+   * false if the call was a no-op because the string was already being suppressed.
+   * @stable ICU 56
+   */
+  virtual UBool suppressBreakAfter(const UnicodeString& string, UErrorCode& status) = 0;
+
+  /**
+   * Stop suppressing a certain string from being the end of the segment.
+   * This function does not create any new segment boundaries, but only serves to un-do
+   * the effect of earlier calls to suppressBreakAfter, or to un-do the effect of
+   * locale data which may be suppressing certain strings.
+   * @param string the exception to remove
+   * @param status error code
+   * @return returns true if the string was present and now removed,
+   * false if the call was a no-op because the string was not being suppressed.
+   * @stable ICU 56
+   */
+  virtual UBool unsuppressBreakAfter(const UnicodeString& string, UErrorCode& status) = 0;
+
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+  /**
+   * This function has been deprecated in favor of wrapIteratorWithFilter()
+   * The behavior is identical.
+   * @param adoptBreakIterator the break iterator to adopt
+   * @param status error code
+   * @return the new BreakIterator, owned by the caller.
+   * @deprecated ICU 60 use wrapIteratorWithFilter() instead
+   * @see wrapBreakIteratorWithFilter()
+   */
+  virtual BreakIterator *build(BreakIterator* adoptBreakIterator, UErrorCode& status) = 0;
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+  /**
+   * Wrap (adopt) an existing break iterator in a new filtered instance.
+   * The resulting BreakIterator is owned by the caller.
+   * The BreakIteratorFilter may be destroyed before the BreakIterator is destroyed.
+   * Note that the adoptBreakIterator is adopted by the new BreakIterator
+   * and should no longer be used by the caller.
+   * The FilteredBreakIteratorBuilder may be reused.
+   * This function is an alias for build()
+   * @param adoptBreakIterator the break iterator to adopt
+   * @param status error code
+   * @return the new BreakIterator, owned by the caller.
+   * @stable ICU 60
+   */
+  inline BreakIterator *wrapIteratorWithFilter(BreakIterator* adoptBreakIterator, UErrorCode& status) {
+    return build(adoptBreakIterator, status);
+  }
+
+ protected:
+  /**
+   * For subclass use
+   * @stable ICU 56
+   */
+  FilteredBreakIteratorBuilder();
+};
+
+
+U_NAMESPACE_END
+
+#endif // #if !UCONFIG_NO_BREAK_ITERATION && !UCONFIG_NO_FILTERED_BREAK_ITERATION
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif // #ifndef FILTEREDBRK_H
diff --git a/thirdparty/icu4c/common/unicode/icudataver.h b/thirdparty/icu4c/common/unicode/icudataver.h
new file mode 100644
index 0000000000..f218ed8ebc
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/icudataver.h
@@ -0,0 +1,43 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2009-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*/
+
+
+/**
+ * \file
+ * \brief C API: access to ICU Data Version number
+ */
+
+#ifndef __ICU_DATA_VER_H__
+#define __ICU_DATA_VER_H__
+
+#include "unicode/utypes.h"
+
+/**
+ * @stable ICU 49
+ */
+#define U_ICU_VERSION_BUNDLE "icuver"
+
+/**
+ * @stable ICU 49
+ */
+#define U_ICU_DATA_KEY "DataVersion"
+
+/**
+ * Retrieves the data version from icuver and stores it in dataVersionFillin.
+ * 
+ * @param dataVersionFillin icuver data version information to be filled in if not-null
+ * @param status stores the error code from the calls to resource bundle
+ * 
+ * @stable ICU 49
+ */
+U_CAPI void U_EXPORT2 u_getDataVersion(UVersionInfo dataVersionFillin, UErrorCode *status);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/icuplug.h b/thirdparty/icu4c/common/unicode/icuplug.h
new file mode 100644
index 0000000000..52f810da57
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/icuplug.h
@@ -0,0 +1,388 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2009-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : icuplug.h
+*
+*   Date         Name        Description
+*   10/29/2009   sl          New.
+******************************************************************************
+*/
+
+/**
+ * \file
+ * \brief C API: ICU Plugin API 
+ *
+ * <h2>C API: ICU Plugin API</h2>
+ *
+ * <p>C API allowing run-time loadable modules that extend or modify ICU functionality.</p>
+ *
+ * <h3>Loading and Configuration</h3>
+ *
+ * <p>At ICU startup time, the environment variable "ICU_PLUGINS" will be 
+ * queried for a directory name.  If it is not set, the preprocessor symbol 
+ * "DEFAULT_ICU_PLUGINS" will be checked for a default value.</p>
+ *
+ * <p>Within the above-named directory, the file  "icuplugins##.txt" will be 
+ * opened, if present, where ## is the major+minor number of the currently 
+ * running ICU (such as, 44 for ICU 4.4, thus icuplugins44.txt)</p>
+ *
+ * <p>The configuration file has this format:</p>
+ *
+ * <ul>
+ * <li>Hash (#) begins a comment line</li>
+ * 
+ * <li>Non-comment lines have two or three components:
+ * LIBRARYNAME     ENTRYPOINT     [ CONFIGURATION .. ]</li>
+ *
+ * <li>Tabs or spaces separate the three items.</li>
+ *
+ * <li>LIBRARYNAME is the name of a shared library, either a short name if 
+ * it is on the loader path,  or a full pathname.</li>
+ *
+ * <li>ENTRYPOINT is the short (undecorated) symbol name of the plugin's 
+ * entrypoint, as above.</li>
+ *
+ * <li>CONFIGURATION is the entire rest of the line . It's passed as-is to 
+ * the plugin.</li>
+ * </ul>
+ *
+ * <p>An example configuration file is, in its entirety:</p>
+ *
+ * \code
+ * # this is icuplugins44.txt
+ * testplug.dll    myPlugin        hello=world
+ * \endcode
+ * <p>Plugins are categorized as "high" or "low" level.  Low level are those 
+ * which must be run BEFORE high level plugins, and before any operations 
+ * which cause ICU to be 'initialized'.  If a plugin is low level but 
+ * causes ICU to allocate memory or become initialized, that plugin is said 
+ * to cause a 'level change'. </p>
+ *
+ * <p>At load time, ICU first queries all plugins to determine their level, 
+ * then loads all 'low' plugins first, and then loads all 'high' plugins.  
+ * Plugins are otherwise loaded in the order listed in the configuration file.</p>
+ * 
+ * <h3>Implementing a Plugin</h3>
+ * \code
+ * U_CAPI UPlugTokenReturn U_EXPORT2 
+ * myPlugin (UPlugData *plug, UPlugReason reason, UErrorCode *status) {
+ *   if(reason==UPLUG_REASON_QUERY) {
+ *      uplug_setPlugName(plug, "Simple Plugin");
+ *      uplug_setPlugLevel(plug, UPLUG_LEVEL_HIGH);
+ *    } else if(reason==UPLUG_REASON_LOAD) {
+ *       ... Set up some ICU things here.... 
+ *    } else if(reason==UPLUG_REASON_UNLOAD) {
+ *       ... unload, clean up ...
+ *    }
+ *   return UPLUG_TOKEN;
+ *  }
+ * \endcode
+ *
+ * <p>The UPlugData*  is an opaque pointer to the plugin-specific data, and is 
+ * used in all other API calls.</p>
+ *
+ * <p>The API contract is:</p>
+ * <ol><li>The plugin MUST always return UPLUG_TOKEN as a return value- to 
+ * indicate that it is a valid plugin.</li>
+ *
+ * <li>When the 'reason' parameter is set to UPLUG_REASON_QUERY,  the 
+ * plugin MUST call uplug_setPlugLevel() to indicate whether it is a high 
+ * level or low level plugin.</li>
+ *
+ * <li>When the 'reason' parameter is UPLUG_REASON_QUERY, the plugin 
+ * SHOULD call uplug_setPlugName to indicate a human readable plugin name.</li></ol>
+ * 
+ *
+ * \internal ICU 4.4 Technology Preview
+ */
+
+
+#ifndef ICUPLUG_H
+#define ICUPLUG_H
+
+#include "unicode/utypes.h"
+
+
+#if UCONFIG_ENABLE_PLUGINS || defined(U_IN_DOXYGEN)
+
+
+
+/* === Basic types === */
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * @{
+ * Opaque structure passed to/from a plugin. 
+ * use the APIs to access it.
+ * @internal ICU 4.4 Technology Preview
+ */
+
+struct UPlugData;
+typedef struct UPlugData UPlugData;
+
+/** @} */
+
+/**
+ * Random Token to identify a valid ICU plugin. Plugins must return this 
+ * from the entrypoint.
+ * @internal ICU 4.4 Technology Preview
+ */
+#define UPLUG_TOKEN 0x54762486
+
+/**
+ * Max width of names, symbols, and configuration strings
+ * @internal ICU 4.4 Technology Preview
+ */
+#define UPLUG_NAME_MAX              100
+
+
+/**
+ * Return value from a plugin entrypoint. 
+ * Must always be set to UPLUG_TOKEN
+ * @see UPLUG_TOKEN
+ * @internal ICU 4.4 Technology Preview
+ */
+typedef uint32_t UPlugTokenReturn;
+
+/**
+ * Reason code for the entrypoint's call
+ * @internal ICU 4.4 Technology Preview
+ */
+typedef enum {
+    UPLUG_REASON_QUERY = 0,     /**< The plugin is being queried for info. **/
+    UPLUG_REASON_LOAD = 1,     /**< The plugin is being loaded. **/
+    UPLUG_REASON_UNLOAD = 2,   /**< The plugin is being unloaded. **/
+    /**
+     * Number of known reasons.
+     * @internal The numeric value may change over time, see ICU ticket #12420.
+     */
+    UPLUG_REASON_COUNT
+} UPlugReason;
+
+
+/**
+ * Level of plugin loading
+ *     INITIAL:  UNKNOWN
+ *       QUERY:   INVALID ->  { LOW | HIGH }
+ *     ERR -> INVALID
+ * @internal ICU 4.4 Technology Preview
+ */
+typedef enum {
+    UPLUG_LEVEL_INVALID = 0,     /**< The plugin is invalid, hasn't called uplug_setLevel, or can't load. **/
+    UPLUG_LEVEL_UNKNOWN = 1,     /**< The plugin is waiting to be installed. **/
+    UPLUG_LEVEL_LOW     = 2,     /**< The plugin must be called before u_init completes **/
+    UPLUG_LEVEL_HIGH    = 3,     /**< The plugin can run at any time. **/
+    /**
+     * Number of known levels.
+     * @internal The numeric value may change over time, see ICU ticket #12420.
+     */
+    UPLUG_LEVEL_COUNT
+} UPlugLevel;
+
+/**
+ * Entrypoint for an ICU plugin.
+ * @param plug the UPlugData handle. 
+ * @param status the plugin's extended status code.
+ * @return A valid plugin must return UPLUG_TOKEN
+ * @internal ICU 4.4 Technology Preview
+ */
+typedef UPlugTokenReturn (U_EXPORT2 UPlugEntrypoint) (
+                  UPlugData *plug,
+                  UPlugReason reason,
+                  UErrorCode *status);
+
+/* === Needed for Implementing === */
+
+/**
+ * Request that this plugin not be unloaded at cleanup time.
+ * This is appropriate for plugins which cannot be cleaned up.
+ * @see u_cleanup()
+ * @param plug plugin
+ * @param dontUnload  set true if this plugin can't be unloaded
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void U_EXPORT2 
+uplug_setPlugNoUnload(UPlugData *plug, UBool dontUnload);
+
+/**
+ * Set the level of this plugin.
+ * @param plug plugin data handle
+ * @param level the level of this plugin
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void U_EXPORT2
+uplug_setPlugLevel(UPlugData *plug, UPlugLevel level);
+
+/**
+ * Get the level of this plugin.
+ * @param plug plugin data handle
+ * @return the level of this plugin
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UPlugLevel U_EXPORT2
+uplug_getPlugLevel(UPlugData *plug);
+
+/**
+ * Get the lowest level of plug which can currently load.
+ * For example, if UPLUG_LEVEL_LOW is returned, then low level plugins may load
+ * if UPLUG_LEVEL_HIGH is returned, then only high level plugins may load.
+ * @return the lowest level of plug which can currently load
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UPlugLevel U_EXPORT2
+uplug_getCurrentLevel(void);
+
+
+/**
+ * Get plug load status
+ * @return The error code of this plugin's load attempt.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UErrorCode U_EXPORT2
+uplug_getPlugLoadStatus(UPlugData *plug); 
+
+/**
+ * Set the human-readable name of this plugin.
+ * @param plug plugin data handle
+ * @param name the name of this plugin. The first UPLUG_NAME_MAX characters willi be copied into a new buffer.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void U_EXPORT2
+uplug_setPlugName(UPlugData *plug, const char *name);
+
+/**
+ * Get the human-readable name of this plugin.
+ * @param plug plugin data handle
+ * @return the name of this plugin
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI const char * U_EXPORT2
+uplug_getPlugName(UPlugData *plug);
+
+/**
+ * Return the symbol name for this plugin, if known.
+ * @param plug plugin data handle
+ * @return the symbol name, or NULL
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI const char * U_EXPORT2
+uplug_getSymbolName(UPlugData *plug);
+
+/**
+ * Return the library name for this plugin, if known.
+ * @param plug plugin data handle
+ * @param status error code
+ * @return the library name, or NULL
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI const char * U_EXPORT2
+uplug_getLibraryName(UPlugData *plug, UErrorCode *status);
+
+/**
+ * Return the library used for this plugin, if known.
+ * Plugins could use this to load data out of their 
+ * @param plug plugin data handle
+ * @return the library, or NULL
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void * U_EXPORT2
+uplug_getLibrary(UPlugData *plug);
+
+/**
+ * Return the plugin-specific context data.
+ * @param plug plugin data handle
+ * @return the context, or NULL if not set
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void * U_EXPORT2
+uplug_getContext(UPlugData *plug);
+
+/**
+ * Set the plugin-specific context data.
+ * @param plug plugin data handle
+ * @param context new context to set
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void U_EXPORT2
+uplug_setContext(UPlugData *plug, void *context);
+
+
+/**
+ * Get the configuration string, if available.
+ * The string is in the platform default codepage.
+ * @param plug plugin data handle
+ * @return configuration string, or else null.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI const char * U_EXPORT2
+uplug_getConfiguration(UPlugData *plug);
+
+/**
+ * Return all currently installed plugins, from newest to oldest
+ * Usage Example:
+ * \code
+ *    UPlugData *plug = NULL;
+ *    while(plug=uplug_nextPlug(plug)) {
+ *        ... do something with 'plug' ...
+ *    }
+ * \endcode
+ * Not thread safe- do not call while plugs are added or removed.
+ * @param prior pass in 'NULL' to get the first (most recent) plug, 
+ *  otherwise pass the value returned on a prior call to uplug_nextPlug
+ * @return the next oldest plugin, or NULL if no more.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UPlugData* U_EXPORT2
+uplug_nextPlug(UPlugData *prior);
+
+/**
+ * Inject a plugin as if it were loaded from a library.
+ * This is useful for testing plugins. 
+ * Note that it will have a 'NULL' library pointer associated
+ * with it, and therefore no llibrary will be closed at cleanup time.
+ * Low level plugins may not be able to load, as ordering can't be enforced.
+ * @param entrypoint entrypoint to install
+ * @param config user specified configuration string, if available, or NULL.
+ * @param status error result
+ * @return the new UPlugData associated with this plugin, or NULL if error.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UPlugData* U_EXPORT2
+uplug_loadPlugFromEntrypoint(UPlugEntrypoint *entrypoint, const char *config, UErrorCode *status);
+
+
+/**
+ * Inject a plugin from a library, as if the information came from a config file.
+ * Low level plugins may not be able to load, and ordering can't be enforced.
+ * @param libName DLL name to load
+ * @param sym symbol of plugin (UPlugEntrypoint function)
+ * @param config configuration string, or NULL
+ * @param status error result
+ * @return the new UPlugData associated with this plugin, or NULL if error.
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI UPlugData* U_EXPORT2
+uplug_loadPlugFromLibrary(const char *libName, const char *sym, const char *config, UErrorCode *status);
+
+/**
+ * Remove a plugin. 
+ * Will request the plugin to be unloaded, and close the library if needed
+ * @param plug plugin handle to close
+ * @param status error result
+ * @internal ICU 4.4 Technology Preview
+ */
+U_CAPI void U_EXPORT2
+uplug_removePlug(UPlugData *plug, UErrorCode *status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+#endif /* UCONFIG_ENABLE_PLUGINS */
+
+#endif /* _ICUPLUG */
+
diff --git a/thirdparty/icu4c/common/unicode/idna.h b/thirdparty/icu4c/common/unicode/idna.h
new file mode 100644
index 0000000000..1305dc6048
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/idna.h
@@ -0,0 +1,330 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  idna.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010mar05
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __IDNA_H__
+#define __IDNA_H__
+
+/**
+ * \file
+ * \brief C++ API: Internationalizing Domain Names in Applications (IDNA)
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/bytestream.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uidna.h"
+#include "unicode/unistr.h"
+
+U_NAMESPACE_BEGIN
+
+class IDNAInfo;
+
+/**
+ * Abstract base class for IDNA processing.
+ * See http://www.unicode.org/reports/tr46/
+ * and http://www.ietf.org/rfc/rfc3490.txt
+ *
+ * The IDNA class is not intended for public subclassing.
+ *
+ * This C++ API currently only implements UTS #46.
+ * The uidna.h C API implements both UTS #46 (functions using UIDNA service object)
+ * and IDNA2003 (functions that do not use a service object).
+ * @stable ICU 4.6
+ */
+class U_COMMON_API IDNA : public UObject {
+public:
+    /**
+     * Destructor.
+     * @stable ICU 4.6
+     */
+    ~IDNA();
+
+    /**
+     * Returns an IDNA instance which implements UTS #46.
+     * Returns an unmodifiable instance, owned by the caller.
+     * Cache it for multiple operations, and delete it when done.
+     * The instance is thread-safe, that is, it can be used concurrently.
+     *
+     * UTS #46 defines Unicode IDNA Compatibility Processing,
+     * updated to the latest version of Unicode and compatible with both
+     * IDNA2003 and IDNA2008.
+     *
+     * The worker functions use transitional processing, including deviation mappings,
+     * unless UIDNA_NONTRANSITIONAL_TO_ASCII or UIDNA_NONTRANSITIONAL_TO_UNICODE
+     * is used in which case the deviation characters are passed through without change.
+     *
+     * Disallowed characters are mapped to U+FFFD.
+     *
+     * For available options see the uidna.h header.
+     * Operations with the UTS #46 instance do not support the
+     * UIDNA_ALLOW_UNASSIGNED option.
+     *
+     * By default, the UTS #46 implementation allows all ASCII characters (as valid or mapped).
+     * When the UIDNA_USE_STD3_RULES option is used, ASCII characters other than
+     * letters, digits, hyphen (LDH) and dot/full stop are disallowed and mapped to U+FFFD.
+     *
+     * @param options Bit set to modify the processing and error checking.
+     *                See option bit set values in uidna.h.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the UTS #46 IDNA instance, if successful
+     * @stable ICU 4.6
+     */
+    static IDNA *
+    createUTS46Instance(uint32_t options, UErrorCode &errorCode);
+
+    /**
+     * Converts a single domain name label into its ASCII form for DNS lookup.
+     * If any processing step fails, then info.hasErrors() will be true and
+     * the result might not be an ASCII string.
+     * The label might be modified according to the types of errors.
+     * Labels with severe errors will be left in (or turned into) their Unicode form.
+     *
+     * The UErrorCode indicates an error only in exceptional cases,
+     * such as a U_MEMORY_ALLOCATION_ERROR.
+     *
+     * @param label Input domain name label
+     * @param dest Destination string object
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual UnicodeString &
+    labelToASCII(const UnicodeString &label, UnicodeString &dest,
+                 IDNAInfo &info, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Converts a single domain name label into its Unicode form for human-readable display.
+     * If any processing step fails, then info.hasErrors() will be true.
+     * The label might be modified according to the types of errors.
+     *
+     * The UErrorCode indicates an error only in exceptional cases,
+     * such as a U_MEMORY_ALLOCATION_ERROR.
+     *
+     * @param label Input domain name label
+     * @param dest Destination string object
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual UnicodeString &
+    labelToUnicode(const UnicodeString &label, UnicodeString &dest,
+                   IDNAInfo &info, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Converts a whole domain name into its ASCII form for DNS lookup.
+     * If any processing step fails, then info.hasErrors() will be true and
+     * the result might not be an ASCII string.
+     * The domain name might be modified according to the types of errors.
+     * Labels with severe errors will be left in (or turned into) their Unicode form.
+     *
+     * The UErrorCode indicates an error only in exceptional cases,
+     * such as a U_MEMORY_ALLOCATION_ERROR.
+     *
+     * @param name Input domain name
+     * @param dest Destination string object
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual UnicodeString &
+    nameToASCII(const UnicodeString &name, UnicodeString &dest,
+                IDNAInfo &info, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Converts a whole domain name into its Unicode form for human-readable display.
+     * If any processing step fails, then info.hasErrors() will be true.
+     * The domain name might be modified according to the types of errors.
+     *
+     * The UErrorCode indicates an error only in exceptional cases,
+     * such as a U_MEMORY_ALLOCATION_ERROR.
+     *
+     * @param name Input domain name
+     * @param dest Destination string object
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual UnicodeString &
+    nameToUnicode(const UnicodeString &name, UnicodeString &dest,
+                  IDNAInfo &info, UErrorCode &errorCode) const = 0;
+
+    // UTF-8 versions of the processing methods ---------------------------- ***
+
+    /**
+     * Converts a single domain name label into its ASCII form for DNS lookup.
+     * UTF-8 version of labelToASCII(), same behavior.
+     *
+     * @param label Input domain name label
+     * @param dest Destination byte sink; Flush()ed if successful
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual void
+    labelToASCII_UTF8(StringPiece label, ByteSink &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const;
+
+    /**
+     * Converts a single domain name label into its Unicode form for human-readable display.
+     * UTF-8 version of labelToUnicode(), same behavior.
+     *
+     * @param label Input domain name label
+     * @param dest Destination byte sink; Flush()ed if successful
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual void
+    labelToUnicodeUTF8(StringPiece label, ByteSink &dest,
+                       IDNAInfo &info, UErrorCode &errorCode) const;
+
+    /**
+     * Converts a whole domain name into its ASCII form for DNS lookup.
+     * UTF-8 version of nameToASCII(), same behavior.
+     *
+     * @param name Input domain name
+     * @param dest Destination byte sink; Flush()ed if successful
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual void
+    nameToASCII_UTF8(StringPiece name, ByteSink &dest,
+                     IDNAInfo &info, UErrorCode &errorCode) const;
+
+    /**
+     * Converts a whole domain name into its Unicode form for human-readable display.
+     * UTF-8 version of nameToUnicode(), same behavior.
+     *
+     * @param name Input domain name
+     * @param dest Destination byte sink; Flush()ed if successful
+     * @param info Output container of IDNA processing details.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.6
+     */
+    virtual void
+    nameToUnicodeUTF8(StringPiece name, ByteSink &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const;
+};
+
+class UTS46;
+
+/**
+ * Output container for IDNA processing errors.
+ * The IDNAInfo class is not suitable for subclassing.
+ * @stable ICU 4.6
+ */
+class U_COMMON_API IDNAInfo : public UMemory {
+public:
+    /**
+     * Constructor for stack allocation.
+     * @stable ICU 4.6
+     */
+    IDNAInfo() : errors(0), labelErrors(0), isTransDiff(false), isBiDi(false), isOkBiDi(true) {}
+    /**
+     * Were there IDNA processing errors?
+     * @return true if there were processing errors
+     * @stable ICU 4.6
+     */
+    UBool hasErrors() const { return errors!=0; }
+    /**
+     * Returns a bit set indicating IDNA processing errors.
+     * See UIDNA_ERROR_... constants in uidna.h.
+     * @return bit set of processing errors
+     * @stable ICU 4.6
+     */
+    uint32_t getErrors() const { return errors; }
+    /**
+     * Returns true if transitional and nontransitional processing produce different results.
+     * This is the case when the input label or domain name contains
+     * one or more deviation characters outside a Punycode label (see UTS #46).
+     * <ul>
+     * <li>With nontransitional processing, such characters are
+     * copied to the destination string.
+     * <li>With transitional processing, such characters are
+     * mapped (sharp s/sigma) or removed (joiner/nonjoiner).
+     * </ul>
+     * @return true if transitional and nontransitional processing produce different results
+     * @stable ICU 4.6
+     */
+    UBool isTransitionalDifferent() const { return isTransDiff; }
+
+private:
+    friend class UTS46;
+
+    IDNAInfo(const IDNAInfo &other);  // no copying
+    IDNAInfo &operator=(const IDNAInfo &other);  // no copying
+
+    void reset() {
+        errors=labelErrors=0;
+        isTransDiff=false;
+        isBiDi=false;
+        isOkBiDi=true;
+    }
+
+    uint32_t errors, labelErrors;
+    UBool isTransDiff;
+    UBool isBiDi;
+    UBool isOkBiDi;
+};
+
+U_NAMESPACE_END
+
+#endif  // UCONFIG_NO_IDNA
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __IDNA_H__
diff --git a/thirdparty/icu4c/common/unicode/localebuilder.h b/thirdparty/icu4c/common/unicode/localebuilder.h
new file mode 100644
index 0000000000..27a894de10
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/localebuilder.h
@@ -0,0 +1,311 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+#ifndef __LOCALEBUILDER_H__
+#define __LOCALEBUILDER_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/locid.h"
+#include "unicode/localematcher.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: Builder API for Locale
+ */
+
+U_NAMESPACE_BEGIN
+class CharString;
+
+/**
+ * <code>LocaleBuilder</code> is used to build instances of <code>Locale</code>
+ * from values configured by the setters.  Unlike the <code>Locale</code>
+ * constructors, the <code>LocaleBuilder</code> checks if a value configured by a
+ * setter satisfies the syntax requirements defined by the <code>Locale</code>
+ * class.  A <code>Locale</code> object created by a <code>LocaleBuilder</code> is
+ * well-formed and can be transformed to a well-formed IETF BCP 47 language tag
+ * without losing information.
+ *
+ * <p>The following example shows how to create a <code>Locale</code> object
+ * with the <code>LocaleBuilder</code>.
+ * <blockquote>
+ * <pre>
+ *     UErrorCode status = U_ZERO_ERROR;
+ *     Locale aLocale = LocaleBuilder()
+ *                          .setLanguage("sr")
+ *                          .setScript("Latn")
+ *                          .setRegion("RS")
+ *                          .build(status);
+ *     if (U_SUCCESS(status)) {
+ *       // ...
+ *     }
+ * </pre>
+ * </blockquote>
+ *
+ * <p>LocaleBuilders can be reused; <code>clear()</code> resets all
+ * fields to their default values.
+ *
+ * <p>LocaleBuilder tracks errors in an internal UErrorCode. For all setters,
+ * except setLanguageTag and setLocale, LocaleBuilder will return immediately
+ * if the internal UErrorCode is in error state.
+ * To reset internal state and error code, call clear method.
+ * The setLanguageTag and setLocale method will first clear the internal
+ * UErrorCode, then track the error of the validation of the input parameter
+ * into the internal UErrorCode.
+ *
+ * @stable ICU 64
+ */
+class U_COMMON_API LocaleBuilder : public UObject {
+public:
+    /**
+     * Constructs an empty LocaleBuilder. The default value of all
+     * fields, extensions, and private use information is the
+     * empty string.
+     *
+     * @stable ICU 64
+     */
+    LocaleBuilder();
+
+    /**
+     * Destructor
+     * @stable ICU 64
+     */
+    virtual ~LocaleBuilder();
+
+    /**
+     * Resets the <code>LocaleBuilder</code> to match the provided
+     * <code>locale</code>.  Existing state is discarded.
+     *
+     * <p>All fields of the locale must be well-formed.
+     * <p>This method clears the internal UErrorCode.
+     *
+     * @param locale the locale
+     * @return This builder.
+     *
+     * @stable ICU 64
+     */
+    LocaleBuilder& setLocale(const Locale& locale);
+
+    /**
+     * Resets the LocaleBuilder to match the provided
+     * [Unicode Locale Identifier](http://www.unicode.org/reports/tr35/tr35.html#unicode_locale_id) .
+     * Discards the existing state.
+     * The empty string causes the builder to be reset, like {@link #clear}.
+     * Legacy language tags (marked as “Type: grandfathered” in BCP 47)
+     * are converted to their canonical form before being processed.
+     * Otherwise, the <code>language tag</code> must be well-formed,
+     * or else the build() method will later report an U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p>This method clears the internal UErrorCode.
+     *
+     * @param tag the language tag, defined as
+     *   [unicode_locale_id](http://www.unicode.org/reports/tr35/tr35.html#unicode_locale_id).
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setLanguageTag(StringPiece tag);
+
+    /**
+     * Sets the language.  If <code>language</code> is the empty string, the
+     * language in this <code>LocaleBuilder</code> is removed. Otherwise, the
+     * <code>language</code> must be well-formed, or else the build() method will
+     * later report an U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p>The syntax of language value is defined as
+     * [unicode_language_subtag](http://www.unicode.org/reports/tr35/tr35.html#unicode_language_subtag).
+     *
+     * @param language the language
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setLanguage(StringPiece language);
+
+    /**
+     * Sets the script. If <code>script</code> is the empty string, the script in
+     * this <code>LocaleBuilder</code> is removed.
+     * Otherwise, the <code>script</code> must be well-formed, or else the build()
+     * method will later report an U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p>The script value is a four-letter script code as
+     * [unicode_script_subtag](http://www.unicode.org/reports/tr35/tr35.html#unicode_script_subtag)
+     * defined by ISO 15924
+     *
+     * @param script the script
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setScript(StringPiece script);
+
+    /**
+     * Sets the region.  If region is the empty string, the region in this
+     * <code>LocaleBuilder</code> is removed. Otherwise, the <code>region</code>
+     * must be well-formed, or else the build() method will later report an
+     * U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p>The region value is defined by
+     *  [unicode_region_subtag](http://www.unicode.org/reports/tr35/tr35.html#unicode_region_subtag)
+     * as a two-letter ISO 3166 code or a three-digit UN M.49 area code.
+     *
+     * <p>The region value in the <code>Locale</code> created by the
+     * <code>LocaleBuilder</code> is always normalized to upper case.
+     *
+     * @param region the region
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setRegion(StringPiece region);
+
+    /**
+     * Sets the variant.  If variant is the empty string, the variant in this
+     * <code>LocaleBuilder</code> is removed.  Otherwise, the <code>variant</code>
+     * must be well-formed, or else the build() method will later report an
+     * U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p><b>Note:</b> This method checks if <code>variant</code>
+     * satisfies the
+     * [unicode_variant_subtag](http://www.unicode.org/reports/tr35/tr35.html#unicode_variant_subtag)
+     * syntax requirements, and normalizes the value to lowercase letters. However,
+     * the <code>Locale</code> class does not impose any syntactic
+     * restriction on variant. To set an ill-formed variant, use a Locale constructor.
+     * If there are multiple unicode_variant_subtag, the caller must concatenate
+     * them with '-' as separator (ex: "foobar-fibar").
+     *
+     * @param variant the variant
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setVariant(StringPiece variant);
+
+    /**
+     * Sets the extension for the given key. If the value is the empty string,
+     * the extension is removed.  Otherwise, the <code>key</code> and
+     * <code>value</code> must be well-formed, or else the build() method will
+     * later report an U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p><b>Note:</b> The key ('u') is used for the Unicode locale extension.
+     * Setting a value for this key replaces any existing Unicode locale key/type
+     * pairs with those defined in the extension.
+     *
+     * <p><b>Note:</b> The key ('x') is used for the private use code. To be
+     * well-formed, the value for this key needs only to have subtags of one to
+     * eight alphanumeric characters, not two to eight as in the general case.
+     *
+     * @param key the extension key
+     * @param value the extension value
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setExtension(char key, StringPiece value);
+
+    /**
+     * Sets the Unicode locale keyword type for the given key. If the type
+     * StringPiece is constructed with a nullptr, the keyword is removed.
+     * If the type is the empty string, the keyword is set without type subtags.
+     * Otherwise, the key and type must be well-formed, or else the build()
+     * method will later report an U_ILLEGAL_ARGUMENT_ERROR.
+     *
+     * <p>Keys and types are converted to lower case.
+     *
+     * <p><b>Note</b>:Setting the 'u' extension via {@link #setExtension}
+     * replaces all Unicode locale keywords with those defined in the
+     * extension.
+     *
+     * @param key the Unicode locale key
+     * @param type the Unicode locale type
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& setUnicodeLocaleKeyword(
+        StringPiece key, StringPiece type);
+
+    /**
+     * Adds a unicode locale attribute, if not already present, otherwise
+     * has no effect.  The attribute must not be empty string and must be
+     * well-formed or U_ILLEGAL_ARGUMENT_ERROR will be set to status
+     * during the build() call.
+     *
+     * @param attribute the attribute
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& addUnicodeLocaleAttribute(StringPiece attribute);
+
+    /**
+     * Removes a unicode locale attribute, if present, otherwise has no
+     * effect.  The attribute must not be empty string and must be well-formed
+     * or U_ILLEGAL_ARGUMENT_ERROR will be set to status during the build() call.
+     *
+     * <p>Attribute comparison for removal is case-insensitive.
+     *
+     * @param attribute the attribute
+     * @return This builder.
+     * @stable ICU 64
+     */
+    LocaleBuilder& removeUnicodeLocaleAttribute(StringPiece attribute);
+
+    /**
+     * Resets the builder to its initial, empty state.
+     * <p>This method clears the internal UErrorCode.
+     *
+     * @return this builder
+     * @stable ICU 64
+     */
+    LocaleBuilder& clear();
+
+    /**
+     * Resets the extensions to their initial, empty state.
+     * Language, script, region and variant are unchanged.
+     *
+     * @return this builder
+     * @stable ICU 64
+     */
+    LocaleBuilder& clearExtensions();
+
+    /**
+     * Returns an instance of <code>Locale</code> created from the fields set
+     * on this builder.
+     * If any set methods or during the build() call require memory allocation
+     * but fail U_MEMORY_ALLOCATION_ERROR will be set to status.
+     * If any of the fields set by the setters are not well-formed, the status
+     * will be set to U_ILLEGAL_ARGUMENT_ERROR. The state of the builder will
+     * not change after the build() call and the caller is free to keep using
+     * the same builder to build more locales.
+     *
+     * @return a new Locale
+     * @stable ICU 64
+     */
+    Locale build(UErrorCode& status);
+
+    /**
+     * Sets the UErrorCode if an error occurred while recording sets.
+     * Preserves older error codes in the outErrorCode.
+     * @param outErrorCode Set to an error code that occurred while setting subtags.
+     *                  Unchanged if there is no such error or if outErrorCode
+     *                  already contained an error.
+     * @return true if U_FAILURE(outErrorCode)
+     * @stable ICU 65
+     */
+    UBool copyErrorTo(UErrorCode &outErrorCode) const;
+
+private:
+    friend class LocaleMatcher::Result;
+
+    void copyExtensionsFrom(const Locale& src, UErrorCode& errorCode);
+
+    UErrorCode status_;
+    char language_[9];
+    char script_[5];
+    char region_[4];
+    CharString *variant_;  // Pointer not object so we need not #include internal charstr.h.
+    icu::Locale *extensions_;  // Pointer not object. Storage for all other fields.
+
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __LOCALEBUILDER_H__
diff --git a/thirdparty/icu4c/common/unicode/localematcher.h b/thirdparty/icu4c/common/unicode/localematcher.h
new file mode 100644
index 0000000000..63a68b0b7f
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/localematcher.h
@@ -0,0 +1,720 @@
+// © 2019 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// localematcher.h
+// created: 2019may08 Markus W. Scherer
+
+#ifndef __LOCALEMATCHER_H__
+#define __LOCALEMATCHER_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/locid.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: Locale matcher: User's desired locales vs. application's supported locales.
+ */
+
+/**
+ * Builder option for whether the language subtag or the script subtag is most important.
+ *
+ * @see LocaleMatcher::Builder#setFavorSubtag(ULocMatchFavorSubtag)
+ * @stable ICU 65
+ */
+enum ULocMatchFavorSubtag {
+    /**
+     * Language differences are most important, then script differences, then region differences.
+     * (This is the default behavior.)
+     *
+     * @stable ICU 65
+     */
+    ULOCMATCH_FAVOR_LANGUAGE,
+    /**
+     * Makes script differences matter relatively more than language differences.
+     *
+     * @stable ICU 65
+     */
+    ULOCMATCH_FAVOR_SCRIPT
+};
+#ifndef U_IN_DOXYGEN
+typedef enum ULocMatchFavorSubtag ULocMatchFavorSubtag;
+#endif
+
+/**
+ * Builder option for whether all desired locales are treated equally or
+ * earlier ones are preferred.
+ *
+ * @see LocaleMatcher::Builder#setDemotionPerDesiredLocale(ULocMatchDemotion)
+ * @stable ICU 65
+ */
+enum ULocMatchDemotion {
+    /**
+     * All desired locales are treated equally.
+     *
+     * @stable ICU 65
+     */
+    ULOCMATCH_DEMOTION_NONE,
+    /**
+     * Earlier desired locales are preferred.
+     *
+     * <p>From each desired locale to the next,
+     * the distance to any supported locale is increased by an additional amount
+     * which is at least as large as most region mismatches.
+     * A later desired locale has to have a better match with some supported locale
+     * due to more than merely having the same region subtag.
+     *
+     * <p>For example: <code>Supported={en, sv}  desired=[en-GB, sv]</code>
+     * yields <code>Result(en-GB, en)</code> because
+     * with the demotion of sv its perfect match is no better than
+     * the region distance between the earlier desired locale en-GB and en=en-US.
+     *
+     * <p>Notes:
+     * <ul>
+     *   <li>In some cases, language and/or script differences can be as small as
+     *       the typical region difference. (Example: sr-Latn vs. sr-Cyrl)
+     *   <li>It is possible for certain region differences to be larger than usual,
+     *       and larger than the demotion.
+     *       (As of CLDR 35 there is no such case, but
+     *        this is possible in future versions of the data.)
+     * </ul>
+     *
+     * @stable ICU 65
+     */
+    ULOCMATCH_DEMOTION_REGION
+};
+#ifndef U_IN_DOXYGEN
+typedef enum ULocMatchDemotion ULocMatchDemotion;
+#endif
+
+#ifndef U_FORCE_HIDE_DRAFT_API
+
+/**
+ * Builder option for whether to include or ignore one-way (fallback) match data.
+ * The LocaleMatcher uses CLDR languageMatch data which includes fallback (oneway=true) entries.
+ * Sometimes it is desirable to ignore those.
+ *
+ * <p>For example, consider a web application with the UI in a given language,
+ * with a link to another, related web app.
+ * The link should include the UI language, and the target server may also use
+ * the client’s Accept-Language header data.
+ * The target server has its own list of supported languages.
+ * One may want to favor UI language consistency, that is,
+ * if there is a decent match for the original UI language, we want to use it,
+ * but not if it is merely a fallback.
+ *
+ * @see LocaleMatcher::Builder#setDirection(ULocMatchDirection)
+ * @draft ICU 67
+ */
+enum ULocMatchDirection {
+    /**
+     * Locale matching includes one-way matches such as Breton→French. (default)
+     *
+     * @draft ICU 67
+     */
+    ULOCMATCH_DIRECTION_WITH_ONE_WAY,
+    /**
+     * Locale matching limited to two-way matches including e.g. Danish↔Norwegian
+     * but ignoring one-way matches.
+     *
+     * @draft ICU 67
+     */
+    ULOCMATCH_DIRECTION_ONLY_TWO_WAY
+};
+#ifndef U_IN_DOXYGEN
+typedef enum ULocMatchDirection ULocMatchDirection;
+#endif
+
+#endif  // U_FORCE_HIDE_DRAFT_API
+
+struct UHashtable;
+
+U_NAMESPACE_BEGIN
+
+struct LSR;
+
+class LocaleDistance;
+class LocaleLsrIterator;
+class UVector;
+class XLikelySubtags;
+
+/**
+ * Immutable class that picks the best match between a user's desired locales and
+ * an application's supported locales.
+ * Movable but not copyable.
+ *
+ * <p>Example:
+ * <pre>
+ * UErrorCode errorCode = U_ZERO_ERROR;
+ * LocaleMatcher matcher = LocaleMatcher::Builder().setSupportedLocales("fr, en-GB, en").build(errorCode);
+ * Locale *bestSupported = matcher.getBestLocale(Locale.US, errorCode);  // "en"
+ * </pre>
+ *
+ * <p>A matcher takes into account when languages are close to one another,
+ * such as Danish and Norwegian,
+ * and when regional variants are close, like en-GB and en-AU as opposed to en-US.
+ *
+ * <p>If there are multiple supported locales with the same (language, script, region)
+ * likely subtags, then the current implementation returns the first of those locales.
+ * It ignores variant subtags (except for pseudolocale variants) and extensions.
+ * This may change in future versions.
+ *
+ * <p>For example, the current implementation does not distinguish between
+ * de, de-DE, de-Latn, de-1901, de-u-co-phonebk.
+ *
+ * <p>If you prefer one equivalent locale over another, then provide only the preferred one,
+ * or place it earlier in the list of supported locales.
+ *
+ * <p>Otherwise, the order of supported locales may have no effect on the best-match results.
+ * The current implementation compares each desired locale with supported locales
+ * in the following order:
+ * 1. Default locale, if supported;
+ * 2. CLDR "paradigm locales" like en-GB and es-419;
+ * 3. other supported locales.
+ * This may change in future versions.
+ *
+ * <p>Often a product will just need one matcher instance, built with the languages
+ * that it supports. However, it may want multiple instances with different
+ * default languages based on additional information, such as the domain.
+ *
+ * <p>This class is not intended for public subclassing.
+ *
+ * @stable ICU 65
+ */
+class U_COMMON_API LocaleMatcher : public UMemory {
+public:
+    /**
+     * Data for the best-matching pair of a desired and a supported locale.
+     * Movable but not copyable.
+     *
+     * @stable ICU 65
+     */
+    class U_COMMON_API Result : public UMemory {
+    public:
+        /**
+         * Move constructor; might modify the source.
+         * This object will have the same contents that the source object had.
+         *
+         * @param src Result to move contents from.
+         * @stable ICU 65
+         */
+        Result(Result &&src) U_NOEXCEPT;
+
+        /**
+         * Destructor.
+         *
+         * @stable ICU 65
+         */
+        ~Result();
+
+        /**
+         * Move assignment; might modify the source.
+         * This object will have the same contents that the source object had.
+         *
+         * @param src Result to move contents from.
+         * @stable ICU 65
+         */
+        Result &operator=(Result &&src) U_NOEXCEPT;
+
+        /**
+         * Returns the best-matching desired locale.
+         * nullptr if the list of desired locales is empty or if none matched well enough.
+         *
+         * @return the best-matching desired locale, or nullptr.
+         * @stable ICU 65
+         */
+        inline const Locale *getDesiredLocale() const { return desiredLocale; }
+
+        /**
+         * Returns the best-matching supported locale.
+         * If none matched well enough, this is the default locale.
+         * The default locale is nullptr if Builder::setNoDefaultLocale() was called,
+         * or if the list of supported locales is empty and no explicit default locale is set.
+         *
+         * @return the best-matching supported locale, or nullptr.
+         * @stable ICU 65
+         */
+        inline const Locale *getSupportedLocale() const { return supportedLocale; }
+
+        /**
+         * Returns the index of the best-matching desired locale in the input Iterable order.
+         * -1 if the list of desired locales is empty or if none matched well enough.
+         *
+         * @return the index of the best-matching desired locale, or -1.
+         * @stable ICU 65
+         */
+        inline int32_t getDesiredIndex() const { return desiredIndex; }
+
+        /**
+         * Returns the index of the best-matching supported locale in the
+         * constructor’s or builder’s input order (“set” Collection plus “added” locales).
+         * If the matcher was built from a locale list string, then the iteration order is that
+         * of a LocalePriorityList built from the same string.
+         * -1 if the list of supported locales is empty or if none matched well enough.
+         *
+         * @return the index of the best-matching supported locale, or -1.
+         * @stable ICU 65
+         */
+        inline int32_t getSupportedIndex() const { return supportedIndex; }
+
+        /**
+         * Takes the best-matching supported locale and adds relevant fields of the
+         * best-matching desired locale, such as the -t- and -u- extensions.
+         * May replace some fields of the supported locale.
+         * The result is the locale that should be used for date and number formatting, collation, etc.
+         * Returns the root locale if getSupportedLocale() returns nullptr.
+         *
+         * <p>Example: desired=ar-SA-u-nu-latn, supported=ar-EG, resolved locale=ar-SA-u-nu-latn
+         *
+         * @return a locale combining the best-matching desired and supported locales.
+         * @stable ICU 65
+         */
+        Locale makeResolvedLocale(UErrorCode &errorCode) const;
+
+    private:
+        Result(const Locale *desired, const Locale *supported,
+               int32_t desIndex, int32_t suppIndex, UBool owned) :
+                desiredLocale(desired), supportedLocale(supported),
+                desiredIndex(desIndex), supportedIndex(suppIndex),
+                desiredIsOwned(owned) {}
+
+        Result(const Result &other) = delete;
+        Result &operator=(const Result &other) = delete;
+
+        const Locale *desiredLocale;
+        const Locale *supportedLocale;
+        int32_t desiredIndex;
+        int32_t supportedIndex;
+        UBool desiredIsOwned;
+
+        friend class LocaleMatcher;
+    };
+
+    /**
+     * LocaleMatcher builder.
+     * Movable but not copyable.
+     *
+     * @stable ICU 65
+     */
+    class U_COMMON_API Builder : public UMemory {
+    public:
+        /**
+         * Constructs a builder used in chaining parameters for building a LocaleMatcher.
+         *
+         * @return a new Builder object
+         * @stable ICU 65
+         */
+        Builder() {}
+
+        /**
+         * Move constructor; might modify the source.
+         * This builder will have the same contents that the source builder had.
+         *
+         * @param src Builder to move contents from.
+         * @stable ICU 65
+         */
+        Builder(Builder &&src) U_NOEXCEPT;
+
+        /**
+         * Destructor.
+         *
+         * @stable ICU 65
+         */
+        ~Builder();
+
+        /**
+         * Move assignment; might modify the source.
+         * This builder will have the same contents that the source builder had.
+         *
+         * @param src Builder to move contents from.
+         * @stable ICU 65
+         */
+        Builder &operator=(Builder &&src) U_NOEXCEPT;
+
+        /**
+         * Parses an Accept-Language string
+         * (<a href="https://tools.ietf.org/html/rfc2616#section-14.4">RFC 2616 Section 14.4</a>),
+         * such as "af, en, fr;q=0.9", and sets the supported locales accordingly.
+         * Allows whitespace in more places but does not allow "*".
+         * Clears any previously set/added supported locales first.
+         *
+         * @param locales the Accept-Language string of locales to set
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &setSupportedLocalesFromListString(StringPiece locales);
+
+        /**
+         * Copies the supported locales, preserving iteration order.
+         * Clears any previously set/added supported locales first.
+         * Duplicates are allowed, and are not removed.
+         *
+         * @param locales the list of locale
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &setSupportedLocales(Locale::Iterator &locales);
+
+        /**
+         * Copies the supported locales from the begin/end range, preserving iteration order.
+         * Clears any previously set/added supported locales first.
+         * Duplicates are allowed, and are not removed.
+         *
+         * Each of the iterator parameter values must be an
+         * input iterator whose value is convertible to const Locale &.
+         *
+         * @param begin Start of range.
+         * @param end Exclusive end of range.
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        template<typename Iter>
+        Builder &setSupportedLocales(Iter begin, Iter end) {
+            if (U_FAILURE(errorCode_)) { return *this; }
+            clearSupportedLocales();
+            while (begin != end) {
+                addSupportedLocale(*begin++);
+            }
+            return *this;
+        }
+
+        /**
+         * Copies the supported locales from the begin/end range, preserving iteration order.
+         * Calls the converter to convert each *begin to a Locale or const Locale &.
+         * Clears any previously set/added supported locales first.
+         * Duplicates are allowed, and are not removed.
+         *
+         * Each of the iterator parameter values must be an
+         * input iterator whose value is convertible to const Locale &.
+         *
+         * @param begin Start of range.
+         * @param end Exclusive end of range.
+         * @param converter Converter from *begin to const Locale & or compatible.
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        template<typename Iter, typename Conv>
+        Builder &setSupportedLocalesViaConverter(Iter begin, Iter end, Conv converter) {
+            if (U_FAILURE(errorCode_)) { return *this; }
+            clearSupportedLocales();
+            while (begin != end) {
+                addSupportedLocale(converter(*begin++));
+            }
+            return *this;
+        }
+
+        /**
+         * Adds another supported locale.
+         * Duplicates are allowed, and are not removed.
+         *
+         * @param locale another locale
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &addSupportedLocale(const Locale &locale);
+
+#ifndef U_HIDE_DRAFT_API
+        /**
+         * Sets no default locale.
+         * There will be no explicit or implicit default locale.
+         * If there is no good match, then the matcher will return nullptr for the
+         * best supported locale.
+         *
+         * @draft ICU 68
+         */
+        Builder &setNoDefaultLocale();
+#endif  // U_HIDE_DRAFT_API
+
+        /**
+         * Sets the default locale; if nullptr, or if it is not set explicitly,
+         * then the first supported locale is used as the default locale.
+         * There is no default locale at all (nullptr will be returned instead)
+         * if setNoDefaultLocale() is called.
+         *
+         * @param defaultLocale the default locale (will be copied)
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &setDefaultLocale(const Locale *defaultLocale);
+
+        /**
+         * If ULOCMATCH_FAVOR_SCRIPT, then the language differences are smaller than script
+         * differences.
+         * This is used in situations (such as maps) where
+         * it is better to fall back to the same script than a similar language.
+         *
+         * @param subtag the subtag to favor
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &setFavorSubtag(ULocMatchFavorSubtag subtag);
+
+        /**
+         * Option for whether all desired locales are treated equally or
+         * earlier ones are preferred (this is the default).
+         *
+         * @param demotion the demotion per desired locale to set.
+         * @return this Builder object
+         * @stable ICU 65
+         */
+        Builder &setDemotionPerDesiredLocale(ULocMatchDemotion demotion);
+
+#ifndef U_HIDE_DRAFT_API
+        /**
+         * Option for whether to include or ignore one-way (fallback) match data.
+         * By default, they are included.
+         *
+         * @param direction the match direction to set.
+         * @return this Builder object
+         * @draft ICU 67
+         */
+        Builder &setDirection(ULocMatchDirection direction) {
+            if (U_SUCCESS(errorCode_)) {
+                direction_ = direction;
+            }
+            return *this;
+        }
+#endif  // U_HIDE_DRAFT_API
+
+#ifndef U_HIDE_DRAFT_API
+        /**
+         * Sets the maximum distance for an acceptable match.
+         * The matcher will return a match for a pair of locales only if
+         * they match at least as well as the pair given here.
+         *
+         * For example, setMaxDistance(en-US, en-GB) limits matches to ones where the
+         * (desired, support) locales have a distance no greater than a region subtag difference.
+         * This is much stricter than the CLDR default.
+         *
+         * The details of locale matching are subject to changes in
+         * CLDR data and in the algorithm.
+         * Specifying a maximum distance in relative terms via a sample pair of locales
+         * insulates from changes that affect all distance metrics similarly,
+         * but some changes will necessarily affect relative distances between
+         * different pairs of locales.
+         *
+         * @param desired the desired locale for distance comparison.
+         * @param supported the supported locale for distance comparison.
+         * @return this Builder object
+         * @draft ICU 68
+         */
+        Builder &setMaxDistance(const Locale &desired, const Locale &supported);
+#endif  // U_HIDE_DRAFT_API
+
+        /**
+         * Sets the UErrorCode if an error occurred while setting parameters.
+         * Preserves older error codes in the outErrorCode.
+         *
+         * @param outErrorCode Set to an error code if it does not contain one already
+         *                  and an error occurred while setting parameters.
+         *                  Otherwise unchanged.
+         * @return true if U_FAILURE(outErrorCode)
+         * @stable ICU 65
+         */
+        UBool copyErrorTo(UErrorCode &outErrorCode) const;
+
+        /**
+         * Builds and returns a new locale matcher.
+         * This builder can continue to be used.
+         *
+         * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+         *                  or else the function returns immediately. Check for U_FAILURE()
+         *                  on output or use with function chaining. (See User Guide for details.)
+         * @return LocaleMatcher
+         * @stable ICU 65
+         */
+        LocaleMatcher build(UErrorCode &errorCode) const;
+
+    private:
+        friend class LocaleMatcher;
+
+        Builder(const Builder &other) = delete;
+        Builder &operator=(const Builder &other) = delete;
+
+        void clearSupportedLocales();
+        bool ensureSupportedLocaleVector();
+
+        UErrorCode errorCode_ = U_ZERO_ERROR;
+        UVector *supportedLocales_ = nullptr;
+        int32_t thresholdDistance_ = -1;
+        ULocMatchDemotion demotion_ = ULOCMATCH_DEMOTION_REGION;
+        Locale *defaultLocale_ = nullptr;
+        bool withDefault_ = true;
+        ULocMatchFavorSubtag favor_ = ULOCMATCH_FAVOR_LANGUAGE;
+        ULocMatchDirection direction_ = ULOCMATCH_DIRECTION_WITH_ONE_WAY;
+        Locale *maxDistanceDesired_ = nullptr;
+        Locale *maxDistanceSupported_ = nullptr;
+    };
+
+    // FYI No public LocaleMatcher constructors in C++; use the Builder.
+
+    /**
+     * Move copy constructor; might modify the source.
+     * This matcher will have the same settings that the source matcher had.
+     * @param src source matcher
+     * @stable ICU 65
+     */
+    LocaleMatcher(LocaleMatcher &&src) U_NOEXCEPT;
+
+    /**
+     * Destructor.
+     * @stable ICU 65
+     */
+    ~LocaleMatcher();
+
+    /**
+     * Move assignment operator; might modify the source.
+     * This matcher will have the same settings that the source matcher had.
+     * The behavior is undefined if *this and src are the same object.
+     * @param src source matcher
+     * @return *this
+     * @stable ICU 65
+     */
+    LocaleMatcher &operator=(LocaleMatcher &&src) U_NOEXCEPT;
+
+    /**
+     * Returns the supported locale which best matches the desired locale.
+     *
+     * @param desiredLocale Typically a user's language.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return the best-matching supported locale.
+     * @stable ICU 65
+     */
+    const Locale *getBestMatch(const Locale &desiredLocale, UErrorCode &errorCode) const;
+
+    /**
+     * Returns the supported locale which best matches one of the desired locales.
+     *
+     * @param desiredLocales Typically a user's languages, in order of preference (descending).
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return the best-matching supported locale.
+     * @stable ICU 65
+     */
+    const Locale *getBestMatch(Locale::Iterator &desiredLocales, UErrorCode &errorCode) const;
+
+    /**
+     * Parses an Accept-Language string
+     * (<a href="https://tools.ietf.org/html/rfc2616#section-14.4">RFC 2616 Section 14.4</a>),
+     * such as "af, en, fr;q=0.9",
+     * and returns the supported locale which best matches one of the desired locales.
+     * Allows whitespace in more places but does not allow "*".
+     *
+     * @param desiredLocaleList Typically a user's languages, as an Accept-Language string.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return the best-matching supported locale.
+     * @stable ICU 65
+     */
+    const Locale *getBestMatchForListString(StringPiece desiredLocaleList, UErrorCode &errorCode) const;
+
+    /**
+     * Returns the best match between the desired locale and the supported locales.
+     * If the result's desired locale is not nullptr, then it is the address of the input locale.
+     * It has not been cloned.
+     *
+     * @param desiredLocale Typically a user's language.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return the best-matching pair of the desired and a supported locale.
+     * @stable ICU 65
+     */
+    Result getBestMatchResult(const Locale &desiredLocale, UErrorCode &errorCode) const;
+
+    /**
+     * Returns the best match between the desired and supported locales.
+     * If the result's desired locale is not nullptr, then it is a clone of
+     * the best-matching desired locale. The Result object owns the clone.
+     *
+     * @param desiredLocales Typically a user's languages, in order of preference (descending).
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return the best-matching pair of a desired and a supported locale.
+     * @stable ICU 65
+     */
+    Result getBestMatchResult(Locale::Iterator &desiredLocales, UErrorCode &errorCode) const;
+
+#ifndef U_HIDE_DRAFT_API
+    /**
+     * Returns true if the pair of locales matches acceptably.
+     * This is influenced by Builder options such as setDirection(), setFavorSubtag(),
+     * and setMaxDistance().
+     *
+     * @param desired The desired locale.
+     * @param supported The supported locale.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return true if the pair of locales matches acceptably.
+     * @draft ICU 68
+     */
+    UBool isMatch(const Locale &desired, const Locale &supported, UErrorCode &errorCode) const;
+#endif  // U_HIDE_DRAFT_API
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Returns a fraction between 0 and 1, where 1 means that the languages are a
+     * perfect match, and 0 means that they are completely different.
+     *
+     * <p>This is mostly an implementation detail, and the precise values may change over time.
+     * The implementation may use either the maximized forms or the others ones, or both.
+     * The implementation may or may not rely on the forms to be consistent with each other.
+     *
+     * <p>Callers should construct and use a matcher rather than match pairs of locales directly.
+     *
+     * @param desired Desired locale.
+     * @param supported Supported locale.
+     * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test,
+     *                  or else the function returns immediately. Check for U_FAILURE()
+     *                  on output or use with function chaining. (See User Guide for details.)
+     * @return value between 0 and 1, inclusive.
+     * @internal (has a known user)
+     */
+    double internalMatch(const Locale &desired, const Locale &supported, UErrorCode &errorCode) const;
+#endif  // U_HIDE_INTERNAL_API
+
+private:
+    LocaleMatcher(const Builder &builder, UErrorCode &errorCode);
+    LocaleMatcher(const LocaleMatcher &other) = delete;
+    LocaleMatcher &operator=(const LocaleMatcher &other) = delete;
+
+    int32_t putIfAbsent(const LSR &lsr, int32_t i, int32_t suppLength, UErrorCode &errorCode);
+
+    int32_t getBestSuppIndex(LSR desiredLSR, LocaleLsrIterator *remainingIter, UErrorCode &errorCode) const;
+
+    const XLikelySubtags &likelySubtags;
+    const LocaleDistance &localeDistance;
+    int32_t thresholdDistance;
+    int32_t demotionPerDesiredLocale;
+    ULocMatchFavorSubtag favorSubtag;
+    ULocMatchDirection direction;
+
+    // These are in input order.
+    const Locale ** supportedLocales;
+    LSR *lsrs;
+    int32_t supportedLocalesLength;
+    // These are in preference order: 1. Default locale 2. paradigm locales 3. others.
+    UHashtable *supportedLsrToIndex;  // Map<LSR, Integer> stores index+1 because 0 is "not found"
+    // Array versions of the supportedLsrToIndex keys and values.
+    // The distance lookup loops over the supportedLSRs and returns the index of the best match.
+    const LSR **supportedLSRs;
+    int32_t *supportedIndexes;
+    int32_t supportedLSRsLength;
+    Locale *ownedDefaultLocale;
+    const Locale *defaultLocale;
+};
+
+U_NAMESPACE_END
+
+#endif  // U_SHOW_CPLUSPLUS_API
+#endif  // __LOCALEMATCHER_H__
diff --git a/thirdparty/icu4c/common/unicode/localpointer.h b/thirdparty/icu4c/common/unicode/localpointer.h
new file mode 100644
index 0000000000..2a65f2d382
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/localpointer.h
@@ -0,0 +1,595 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  localpointer.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009nov13
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __LOCALPOINTER_H__
+#define __LOCALPOINTER_H__
+
+/**
+ * \file
+ * \brief C++ API: "Smart pointers" for use with and in ICU4C C++ code.
+ *
+ * These classes are inspired by
+ * - std::auto_ptr
+ * - boost::scoped_ptr & boost::scoped_array
+ * - Taligent Safe Pointers (TOnlyPointerTo)
+ *
+ * but none of those provide for all of the goals for ICU smart pointers:
+ * - Smart pointer owns the object and releases it when it goes out of scope.
+ * - No transfer of ownership via copy/assignment to reduce misuse. Simpler & more robust.
+ * - ICU-compatible: No exceptions.
+ * - Need to be able to orphan/release the pointer and its ownership.
+ * - Need variants for normal C++ object pointers, C++ arrays, and ICU C service objects.
+ *
+ * For details see http://site.icu-project.org/design/cpp/scoped_ptr
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include <memory>
+
+U_NAMESPACE_BEGIN
+
+/**
+ * "Smart pointer" base class; do not use directly: use LocalPointer etc.
+ *
+ * Base class for smart pointer classes that do not throw exceptions.
+ *
+ * Do not use this base class directly, since it does not delete its pointer.
+ * A subclass must implement methods that delete the pointer:
+ * Destructor and adoptInstead().
+ *
+ * There is no operator T *() provided because the programmer must decide
+ * whether to use getAlias() (without transfer of ownership) or orphan()
+ * (with transfer of ownership and NULLing of the pointer).
+ *
+ * @see LocalPointer
+ * @see LocalArray
+ * @see U_DEFINE_LOCAL_OPEN_POINTER
+ * @stable ICU 4.4
+ */
+template<typename T>
+class LocalPointerBase {
+public:
+    // No heap allocation. Use only on the stack.
+    static void* U_EXPORT2 operator new(size_t) = delete;
+    static void* U_EXPORT2 operator new[](size_t) = delete;
+#if U_HAVE_PLACEMENT_NEW
+    static void* U_EXPORT2 operator new(size_t, void*) = delete;
+#endif
+
+    /**
+     * Constructor takes ownership.
+     * @param p simple pointer to an object that is adopted
+     * @stable ICU 4.4
+     */
+    explicit LocalPointerBase(T *p=NULL) : ptr(p) {}
+    /**
+     * Destructor deletes the object it owns.
+     * Subclass must override: Base class does nothing.
+     * @stable ICU 4.4
+     */
+    ~LocalPointerBase() { /* delete ptr; */ }
+    /**
+     * NULL check.
+     * @return true if ==NULL
+     * @stable ICU 4.4
+     */
+    UBool isNull() const { return ptr==NULL; }
+    /**
+     * NULL check.
+     * @return true if !=NULL
+     * @stable ICU 4.4
+     */
+    UBool isValid() const { return ptr!=NULL; }
+    /**
+     * Comparison with a simple pointer, so that existing code
+     * with ==NULL need not be changed.
+     * @param other simple pointer for comparison
+     * @return true if this pointer value equals other
+     * @stable ICU 4.4
+     */
+    bool operator==(const T *other) const { return ptr==other; }
+    /**
+     * Comparison with a simple pointer, so that existing code
+     * with !=NULL need not be changed.
+     * @param other simple pointer for comparison
+     * @return true if this pointer value differs from other
+     * @stable ICU 4.4
+     */
+    bool operator!=(const T *other) const { return ptr!=other; }
+    /**
+     * Access without ownership change.
+     * @return the pointer value
+     * @stable ICU 4.4
+     */
+    T *getAlias() const { return ptr; }
+    /**
+     * Access without ownership change.
+     * @return the pointer value as a reference
+     * @stable ICU 4.4
+     */
+    T &operator*() const { return *ptr; }
+    /**
+     * Access without ownership change.
+     * @return the pointer value
+     * @stable ICU 4.4
+     */
+    T *operator->() const { return ptr; }
+    /**
+     * Gives up ownership; the internal pointer becomes NULL.
+     * @return the pointer value;
+     *         caller becomes responsible for deleting the object
+     * @stable ICU 4.4
+     */
+    T *orphan() {
+        T *p=ptr;
+        ptr=NULL;
+        return p;
+    }
+    /**
+     * Deletes the object it owns,
+     * and adopts (takes ownership of) the one passed in.
+     * Subclass must override: Base class does not delete the object.
+     * @param p simple pointer to an object that is adopted
+     * @stable ICU 4.4
+     */
+    void adoptInstead(T *p) {
+        // delete ptr;
+        ptr=p;
+    }
+protected:
+    /**
+     * Actual pointer.
+     * @internal
+     */
+    T *ptr;
+private:
+    // No comparison operators with other LocalPointerBases.
+    bool operator==(const LocalPointerBase<T> &other);
+    bool operator!=(const LocalPointerBase<T> &other);
+    // No ownership sharing: No copy constructor, no assignment operator.
+    LocalPointerBase(const LocalPointerBase<T> &other);
+    void operator=(const LocalPointerBase<T> &other);
+};
+
+/**
+ * "Smart pointer" class, deletes objects via the standard C++ delete operator.
+ * For most methods see the LocalPointerBase base class.
+ *
+ * Usage example:
+ * \code
+ * LocalPointer<UnicodeString> s(new UnicodeString((UChar32)0x50005));
+ * int32_t length=s->length();  // 2
+ * char16_t lead=s->charAt(0);  // 0xd900
+ * if(some condition) { return; }  // no need to explicitly delete the pointer
+ * s.adoptInstead(new UnicodeString((char16_t)0xfffc));
+ * length=s->length();  // 1
+ * // no need to explicitly delete the pointer
+ * \endcode
+ *
+ * @see LocalPointerBase
+ * @stable ICU 4.4
+ */
+template<typename T>
+class LocalPointer : public LocalPointerBase<T> {
+public:
+    using LocalPointerBase<T>::operator*;
+    using LocalPointerBase<T>::operator->;
+    /**
+     * Constructor takes ownership.
+     * @param p simple pointer to an object that is adopted
+     * @stable ICU 4.4
+     */
+    explicit LocalPointer(T *p=NULL) : LocalPointerBase<T>(p) {}
+    /**
+     * Constructor takes ownership and reports an error if NULL.
+     *
+     * This constructor is intended to be used with other-class constructors
+     * that may report a failure UErrorCode,
+     * so that callers need to check only for U_FAILURE(errorCode)
+     * and not also separately for isNull().
+     *
+     * @param p simple pointer to an object that is adopted
+     * @param errorCode in/out UErrorCode, set to U_MEMORY_ALLOCATION_ERROR
+     *     if p==NULL and no other failure code had been set
+     * @stable ICU 55
+     */
+    LocalPointer(T *p, UErrorCode &errorCode) : LocalPointerBase<T>(p) {
+        if(p==NULL && U_SUCCESS(errorCode)) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+    /**
+     * Move constructor, leaves src with isNull().
+     * @param src source smart pointer
+     * @stable ICU 56
+     */
+    LocalPointer(LocalPointer<T> &&src) U_NOEXCEPT : LocalPointerBase<T>(src.ptr) {
+        src.ptr=NULL;
+    }
+
+    /**
+     * Constructs a LocalPointer from a C++11 std::unique_ptr.
+     * The LocalPointer steals the object owned by the std::unique_ptr.
+     *
+     * This constructor works via move semantics. If your std::unique_ptr is
+     * in a local variable, you must use std::move.
+     *
+     * @param p The std::unique_ptr from which the pointer will be stolen.
+     * @stable ICU 64
+     */
+    explicit LocalPointer(std::unique_ptr<T> &&p)
+        : LocalPointerBase<T>(p.release()) {}
+
+    /**
+     * Destructor deletes the object it owns.
+     * @stable ICU 4.4
+     */
+    ~LocalPointer() {
+        delete LocalPointerBase<T>::ptr;
+    }
+    /**
+     * Move assignment operator, leaves src with isNull().
+     * The behavior is undefined if *this and src are the same object.
+     * @param src source smart pointer
+     * @return *this
+     * @stable ICU 56
+     */
+    LocalPointer<T> &operator=(LocalPointer<T> &&src) U_NOEXCEPT {
+        delete LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=src.ptr;
+        src.ptr=NULL;
+        return *this;
+    }
+
+    /**
+     * Move-assign from an std::unique_ptr to this LocalPointer.
+     * Steals the pointer from the std::unique_ptr.
+     *
+     * @param p The std::unique_ptr from which the pointer will be stolen.
+     * @return *this
+     * @stable ICU 64
+     */
+    LocalPointer<T> &operator=(std::unique_ptr<T> &&p) U_NOEXCEPT {
+        adoptInstead(p.release());
+        return *this;
+    }
+
+    /**
+     * Swap pointers.
+     * @param other other smart pointer
+     * @stable ICU 56
+     */
+    void swap(LocalPointer<T> &other) U_NOEXCEPT {
+        T *temp=LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=other.ptr;
+        other.ptr=temp;
+    }
+    /**
+     * Non-member LocalPointer swap function.
+     * @param p1 will get p2's pointer
+     * @param p2 will get p1's pointer
+     * @stable ICU 56
+     */
+    friend inline void swap(LocalPointer<T> &p1, LocalPointer<T> &p2) U_NOEXCEPT {
+        p1.swap(p2);
+    }
+    /**
+     * Deletes the object it owns,
+     * and adopts (takes ownership of) the one passed in.
+     * @param p simple pointer to an object that is adopted
+     * @stable ICU 4.4
+     */
+    void adoptInstead(T *p) {
+        delete LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=p;
+    }
+    /**
+     * Deletes the object it owns,
+     * and adopts (takes ownership of) the one passed in.
+     *
+     * If U_FAILURE(errorCode), then the current object is retained and the new one deleted.
+     *
+     * If U_SUCCESS(errorCode) but the input pointer is NULL,
+     * then U_MEMORY_ALLOCATION_ERROR is set,
+     * the current object is deleted, and NULL is set.
+     *
+     * @param p simple pointer to an object that is adopted
+     * @param errorCode in/out UErrorCode, set to U_MEMORY_ALLOCATION_ERROR
+     *     if p==NULL and no other failure code had been set
+     * @stable ICU 55
+     */
+    void adoptInsteadAndCheckErrorCode(T *p, UErrorCode &errorCode) {
+        if(U_SUCCESS(errorCode)) {
+            delete LocalPointerBase<T>::ptr;
+            LocalPointerBase<T>::ptr=p;
+            if(p==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+            }
+        } else {
+            delete p;
+        }
+    }
+
+    /**
+     * Conversion operator to a C++11 std::unique_ptr.
+     * Disowns the object and gives it to the returned std::unique_ptr.
+     *
+     * This operator works via move semantics. If your LocalPointer is
+     * in a local variable, you must use std::move.
+     *
+     * @return An std::unique_ptr owning the pointer previously owned by this
+     *         icu::LocalPointer.
+     * @stable ICU 64
+     */
+    operator std::unique_ptr<T> () && {
+        return std::unique_ptr<T>(LocalPointerBase<T>::orphan());
+    }
+};
+
+/**
+ * "Smart pointer" class, deletes objects via the C++ array delete[] operator.
+ * For most methods see the LocalPointerBase base class.
+ * Adds operator[] for array item access.
+ *
+ * Usage example:
+ * \code
+ * LocalArray<UnicodeString> a(new UnicodeString[2]);
+ * a[0].append((char16_t)0x61);
+ * if(some condition) { return; }  // no need to explicitly delete the array
+ * a.adoptInstead(new UnicodeString[4]);
+ * a[3].append((char16_t)0x62).append((char16_t)0x63).reverse();
+ * // no need to explicitly delete the array
+ * \endcode
+ *
+ * @see LocalPointerBase
+ * @stable ICU 4.4
+ */
+template<typename T>
+class LocalArray : public LocalPointerBase<T> {
+public:
+    using LocalPointerBase<T>::operator*;
+    using LocalPointerBase<T>::operator->;
+    /**
+     * Constructor takes ownership.
+     * @param p simple pointer to an array of T objects that is adopted
+     * @stable ICU 4.4
+     */
+    explicit LocalArray(T *p=NULL) : LocalPointerBase<T>(p) {}
+    /**
+     * Constructor takes ownership and reports an error if NULL.
+     *
+     * This constructor is intended to be used with other-class constructors
+     * that may report a failure UErrorCode,
+     * so that callers need to check only for U_FAILURE(errorCode)
+     * and not also separately for isNull().
+     *
+     * @param p simple pointer to an array of T objects that is adopted
+     * @param errorCode in/out UErrorCode, set to U_MEMORY_ALLOCATION_ERROR
+     *     if p==NULL and no other failure code had been set
+     * @stable ICU 56
+     */
+    LocalArray(T *p, UErrorCode &errorCode) : LocalPointerBase<T>(p) {
+        if(p==NULL && U_SUCCESS(errorCode)) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+    /**
+     * Move constructor, leaves src with isNull().
+     * @param src source smart pointer
+     * @stable ICU 56
+     */
+    LocalArray(LocalArray<T> &&src) U_NOEXCEPT : LocalPointerBase<T>(src.ptr) {
+        src.ptr=NULL;
+    }
+
+    /**
+     * Constructs a LocalArray from a C++11 std::unique_ptr of an array type.
+     * The LocalPointer steals the array owned by the std::unique_ptr.
+     *
+     * This constructor works via move semantics. If your std::unique_ptr is
+     * in a local variable, you must use std::move.
+     *
+     * @param p The std::unique_ptr from which the array will be stolen.
+     * @stable ICU 64
+     */
+    explicit LocalArray(std::unique_ptr<T[]> &&p)
+        : LocalPointerBase<T>(p.release()) {}
+
+    /**
+     * Destructor deletes the array it owns.
+     * @stable ICU 4.4
+     */
+    ~LocalArray() {
+        delete[] LocalPointerBase<T>::ptr;
+    }
+    /**
+     * Move assignment operator, leaves src with isNull().
+     * The behavior is undefined if *this and src are the same object.
+     * @param src source smart pointer
+     * @return *this
+     * @stable ICU 56
+     */
+    LocalArray<T> &operator=(LocalArray<T> &&src) U_NOEXCEPT {
+        delete[] LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=src.ptr;
+        src.ptr=NULL;
+        return *this;
+    }
+
+    /**
+     * Move-assign from an std::unique_ptr to this LocalPointer.
+     * Steals the array from the std::unique_ptr.
+     *
+     * @param p The std::unique_ptr from which the array will be stolen.
+     * @return *this
+     * @stable ICU 64
+     */
+    LocalArray<T> &operator=(std::unique_ptr<T[]> &&p) U_NOEXCEPT {
+        adoptInstead(p.release());
+        return *this;
+    }
+
+    /**
+     * Swap pointers.
+     * @param other other smart pointer
+     * @stable ICU 56
+     */
+    void swap(LocalArray<T> &other) U_NOEXCEPT {
+        T *temp=LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=other.ptr;
+        other.ptr=temp;
+    }
+    /**
+     * Non-member LocalArray swap function.
+     * @param p1 will get p2's pointer
+     * @param p2 will get p1's pointer
+     * @stable ICU 56
+     */
+    friend inline void swap(LocalArray<T> &p1, LocalArray<T> &p2) U_NOEXCEPT {
+        p1.swap(p2);
+    }
+    /**
+     * Deletes the array it owns,
+     * and adopts (takes ownership of) the one passed in.
+     * @param p simple pointer to an array of T objects that is adopted
+     * @stable ICU 4.4
+     */
+    void adoptInstead(T *p) {
+        delete[] LocalPointerBase<T>::ptr;
+        LocalPointerBase<T>::ptr=p;
+    }
+    /**
+     * Deletes the array it owns,
+     * and adopts (takes ownership of) the one passed in.
+     *
+     * If U_FAILURE(errorCode), then the current array is retained and the new one deleted.
+     *
+     * If U_SUCCESS(errorCode) but the input pointer is NULL,
+     * then U_MEMORY_ALLOCATION_ERROR is set,
+     * the current array is deleted, and NULL is set.
+     *
+     * @param p simple pointer to an array of T objects that is adopted
+     * @param errorCode in/out UErrorCode, set to U_MEMORY_ALLOCATION_ERROR
+     *     if p==NULL and no other failure code had been set
+     * @stable ICU 56
+     */
+    void adoptInsteadAndCheckErrorCode(T *p, UErrorCode &errorCode) {
+        if(U_SUCCESS(errorCode)) {
+            delete[] LocalPointerBase<T>::ptr;
+            LocalPointerBase<T>::ptr=p;
+            if(p==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+            }
+        } else {
+            delete[] p;
+        }
+    }
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     * @stable ICU 4.4
+     */
+    T &operator[](ptrdiff_t i) const { return LocalPointerBase<T>::ptr[i]; }
+
+    /**
+     * Conversion operator to a C++11 std::unique_ptr.
+     * Disowns the object and gives it to the returned std::unique_ptr.
+     *
+     * This operator works via move semantics. If your LocalPointer is
+     * in a local variable, you must use std::move.
+     *
+     * @return An std::unique_ptr owning the pointer previously owned by this
+     *         icu::LocalPointer.
+     * @stable ICU 64
+     */
+    operator std::unique_ptr<T[]> () && {
+        return std::unique_ptr<T[]>(LocalPointerBase<T>::orphan());
+    }
+};
+
+/**
+ * \def U_DEFINE_LOCAL_OPEN_POINTER
+ * "Smart pointer" definition macro, deletes objects via the closeFunction.
+ * Defines a subclass of LocalPointerBase which works just
+ * like LocalPointer<Type> except that this subclass will use the closeFunction
+ * rather than the C++ delete operator.
+ *
+ * Usage example:
+ * \code
+ * LocalUCaseMapPointer csm(ucasemap_open(localeID, options, &errorCode));
+ * utf8OutLength=ucasemap_utf8ToLower(csm.getAlias(),
+ *     utf8Out, (int32_t)sizeof(utf8Out),
+ *     utf8In, utf8InLength, &errorCode);
+ * if(U_FAILURE(errorCode)) { return; }  // no need to explicitly delete the UCaseMap
+ * \endcode
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+#define U_DEFINE_LOCAL_OPEN_POINTER(LocalPointerClassName, Type, closeFunction) \
+    class LocalPointerClassName : public LocalPointerBase<Type> { \
+    public: \
+        using LocalPointerBase<Type>::operator*; \
+        using LocalPointerBase<Type>::operator->; \
+        explicit LocalPointerClassName(Type *p=NULL) : LocalPointerBase<Type>(p) {} \
+        LocalPointerClassName(LocalPointerClassName &&src) U_NOEXCEPT \
+                : LocalPointerBase<Type>(src.ptr) { \
+            src.ptr=NULL; \
+        } \
+        /* TODO: Be agnostic of the deleter function signature from the user-provided std::unique_ptr? */ \
+        explicit LocalPointerClassName(std::unique_ptr<Type, decltype(&closeFunction)> &&p) \
+                : LocalPointerBase<Type>(p.release()) {} \
+        ~LocalPointerClassName() { if (ptr != NULL) { closeFunction(ptr); } } \
+        LocalPointerClassName &operator=(LocalPointerClassName &&src) U_NOEXCEPT { \
+            if (ptr != NULL) { closeFunction(ptr); } \
+            LocalPointerBase<Type>::ptr=src.ptr; \
+            src.ptr=NULL; \
+            return *this; \
+        } \
+        /* TODO: Be agnostic of the deleter function signature from the user-provided std::unique_ptr? */ \
+        LocalPointerClassName &operator=(std::unique_ptr<Type, decltype(&closeFunction)> &&p) { \
+            adoptInstead(p.release()); \
+            return *this; \
+        } \
+        void swap(LocalPointerClassName &other) U_NOEXCEPT { \
+            Type *temp=LocalPointerBase<Type>::ptr; \
+            LocalPointerBase<Type>::ptr=other.ptr; \
+            other.ptr=temp; \
+        } \
+        friend inline void swap(LocalPointerClassName &p1, LocalPointerClassName &p2) U_NOEXCEPT { \
+            p1.swap(p2); \
+        } \
+        void adoptInstead(Type *p) { \
+            if (ptr != NULL) { closeFunction(ptr); } \
+            ptr=p; \
+        } \
+        operator std::unique_ptr<Type, decltype(&closeFunction)> () && { \
+            return std::unique_ptr<Type, decltype(&closeFunction)>(LocalPointerBase<Type>::orphan(), closeFunction); \
+        } \
+    }
+
+U_NAMESPACE_END
+
+#endif  /* U_SHOW_CPLUSPLUS_API */
+#endif  /* __LOCALPOINTER_H__ */
diff --git a/thirdparty/icu4c/common/unicode/locdspnm.h b/thirdparty/icu4c/common/unicode/locdspnm.h
new file mode 100644
index 0000000000..4f06f85704
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/locdspnm.h
@@ -0,0 +1,211 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2010-2016, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*/
+
+#ifndef LOCDSPNM_H
+#define LOCDSPNM_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C++ API: Provides display names of Locale and its components.
+ */
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/locid.h"
+#include "unicode/strenum.h"
+#include "unicode/uscript.h"
+#include "unicode/uldnames.h"
+#include "unicode/udisplaycontext.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Returns display names of Locales and components of Locales. For
+ * more information on language, script, region, variant, key, and
+ * values, see Locale.
+ * @stable ICU 4.4
+ */
+class U_COMMON_API LocaleDisplayNames : public UObject {
+public:
+    /**
+     * Destructor.
+     * @stable ICU 4.4
+     */
+    virtual ~LocaleDisplayNames();
+
+    /**
+     * Convenience overload of
+     * {@link #createInstance(const Locale& locale, UDialectHandling dialectHandling)}
+     * that specifies STANDARD dialect handling.
+     * @param locale the display locale
+     * @return a LocaleDisplayNames instance
+     * @stable ICU 4.4
+     */
+    inline static LocaleDisplayNames* U_EXPORT2 createInstance(const Locale& locale);
+
+    /**
+     * Returns an instance of LocaleDisplayNames that returns names
+     * formatted for the provided locale, using the provided
+     * dialectHandling.
+     *
+     * @param locale the display locale
+     * @param dialectHandling how to select names for locales
+     * @return a LocaleDisplayNames instance
+     * @stable ICU 4.4
+     */
+    static LocaleDisplayNames* U_EXPORT2 createInstance(const Locale& locale,
+                            UDialectHandling dialectHandling);
+
+    /**
+     * Returns an instance of LocaleDisplayNames that returns names formatted
+     * for the provided locale, using the provided UDisplayContext settings.
+     *
+     * @param locale the display locale
+     * @param contexts List of one or more context settings (e.g. for dialect
+     *               handling, capitalization, etc.
+     * @param length Number of items in the contexts list
+     * @return a LocaleDisplayNames instance
+     * @stable ICU 51
+     */
+    static LocaleDisplayNames* U_EXPORT2 createInstance(const Locale& locale,
+                            UDisplayContext *contexts, int32_t length);
+
+    // getters for state
+    /**
+     * Returns the locale used to determine the display names. This is
+     * not necessarily the same locale passed to {@link #createInstance}.
+     * @return the display locale
+     * @stable ICU 4.4
+     */
+    virtual const Locale& getLocale() const = 0;
+
+    /**
+     * Returns the dialect handling used in the display names.
+     * @return the dialect handling enum
+     * @stable ICU 4.4
+     */
+    virtual UDialectHandling getDialectHandling() const = 0;
+
+    /**
+     * Returns the UDisplayContext value for the specified UDisplayContextType.
+     * @param type the UDisplayContextType whose value to return
+     * @return the UDisplayContext for the specified type.
+     * @stable ICU 51
+     */
+    virtual UDisplayContext getContext(UDisplayContextType type) const = 0;
+
+    // names for entire locales
+    /**
+     * Returns the display name of the provided locale.
+     * @param locale the locale whose display name to return
+     * @param result receives the locale's display name
+     * @return the display name of the provided locale
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& localeDisplayName(const Locale& locale,
+                         UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided locale id.
+     * @param localeId the id of the locale whose display name to return
+     * @param result receives the locale's display name
+     * @return the display name of the provided locale
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& localeDisplayName(const char* localeId,
+                         UnicodeString& result) const = 0;
+
+    // names for components of a locale id
+    /**
+     * Returns the display name of the provided language code.
+     * @param lang the language code
+     * @param result receives the language code's display name
+     * @return the display name of the provided language code
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& languageDisplayName(const char* lang,
+                           UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided script code.
+     * @param script the script code
+     * @param result receives the script code's display name
+     * @return the display name of the provided script code
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& scriptDisplayName(const char* script,
+                         UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided script code.
+     * @param scriptCode the script code number
+     * @param result receives the script code's display name
+     * @return the display name of the provided script code
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& scriptDisplayName(UScriptCode scriptCode,
+                         UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided region code.
+     * @param region the region code
+     * @param result receives the region code's display name
+     * @return the display name of the provided region code
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& regionDisplayName(const char* region,
+                         UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided variant.
+     * @param variant the variant string
+     * @param result receives the variant's display name
+     * @return the display name of the provided variant
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& variantDisplayName(const char* variant,
+                          UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided locale key.
+     * @param key the locale key name
+     * @param result receives the locale key's display name
+     * @return the display name of the provided locale key
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& keyDisplayName(const char* key,
+                      UnicodeString& result) const = 0;
+
+    /**
+     * Returns the display name of the provided value (used with the provided key).
+     * @param key the locale key name
+     * @param value the locale key's value
+     * @param result receives the value's display name
+     * @return the display name of the provided value
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString& keyValueDisplayName(const char* key, const char* value,
+                           UnicodeString& result) const = 0;
+};
+
+inline LocaleDisplayNames* LocaleDisplayNames::createInstance(const Locale& locale) {
+  return LocaleDisplayNames::createInstance(locale, ULDN_STANDARD_NAMES);
+}
+
+U_NAMESPACE_END
+
+#endif
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/locid.h b/thirdparty/icu4c/common/unicode/locid.h
new file mode 100644
index 0000000000..ba858d702a
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/locid.h
@@ -0,0 +1,1274 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1996-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File locid.h
+*
+* Created by: Helena Shih
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   02/11/97    aliu        Changed gLocPath to fgLocPath and added methods to
+*                           get and set it.
+*   04/02/97    aliu        Made operator!= inline; fixed return value of getName().
+*   04/15/97    aliu        Cleanup for AIX/Win32.
+*   04/24/97    aliu        Numerous changes per code review.
+*   08/18/98    stephen     Added tokenizeString(),changed getDisplayName()
+*   09/08/98    stephen     Moved definition of kEmptyString for Mac Port
+*   11/09/99    weiv        Added const char * getName() const;
+*   04/12/00    srl         removing unicodestring api's and cached hash code
+*   08/10/01    grhoten     Change the static Locales to accessor functions
+******************************************************************************
+*/
+
+#ifndef LOCID_H
+#define LOCID_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/bytestream.h"
+#include "unicode/localpointer.h"
+#include "unicode/strenum.h"
+#include "unicode/stringpiece.h"
+#include "unicode/uobject.h"
+#include "unicode/putil.h"
+#include "unicode/uloc.h"
+
+/**
+ * \file
+ * \brief C++ API: Locale ID object.
+ */
+
+U_NAMESPACE_BEGIN
+
+// Forward Declarations
+void U_CALLCONV locale_available_init(); /**< @internal */
+
+class StringEnumeration;
+class UnicodeString;
+
+/**
+ * A <code>Locale</code> object represents a specific geographical, political,
+ * or cultural region. An operation that requires a <code>Locale</code> to perform
+ * its task is called <em>locale-sensitive</em> and uses the <code>Locale</code>
+ * to tailor information for the user. For example, displaying a number
+ * is a locale-sensitive operation--the number should be formatted
+ * according to the customs/conventions of the user's native country,
+ * region, or culture.
+ *
+ * The Locale class is not suitable for subclassing.
+ *
+ * <P>
+ * You can create a <code>Locale</code> object using the constructor in
+ * this class:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ *       Locale( const   char*  language,
+ *               const   char*  country,
+ *               const   char*  variant);
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * The first argument to the constructors is a valid <STRONG>ISO
+ * Language Code.</STRONG> These codes are the lower-case two-letter
+ * codes as defined by ISO-639.
+ * You can find a full list of these codes at:
+ * <BR><a href ="http://www.loc.gov/standards/iso639-2/">
+ * http://www.loc.gov/standards/iso639-2/</a>
+ *
+ * <P>
+ * The second argument to the constructors is a valid <STRONG>ISO Country
+ * Code.</STRONG> These codes are the upper-case two-letter codes
+ * as defined by ISO-3166.
+ * You can find a full list of these codes at a number of sites, such as:
+ * <BR><a href="http://www.iso.org/iso/en/prods-services/iso3166ma/index.html">
+ * http://www.iso.org/iso/en/prods-services/iso3166ma/index.html</a>
+ *
+ * <P>
+ * The third constructor requires a third argument--the <STRONG>Variant.</STRONG>
+ * The Variant codes are vendor and browser-specific.
+ * For example, use REVISED for a language's revised script orthography, and POSIX for POSIX.
+ * Where there are two variants, separate them with an underscore, and
+ * put the most important one first. For
+ * example, a Traditional Spanish collation might be referenced, with
+ * "ES", "ES", "Traditional_POSIX".
+ *
+ * <P>
+ * Because a <code>Locale</code> object is just an identifier for a region,
+ * no validity check is performed when you construct a <code>Locale</code>.
+ * If you want to see whether particular resources are available for the
+ * <code>Locale</code> you construct, you must query those resources. For
+ * example, ask the <code>NumberFormat</code> for the locales it supports
+ * using its <code>getAvailableLocales</code> method.
+ * <BR><STRONG>Note:</STRONG> When you ask for a resource for a particular
+ * locale, you get back the best available match, not necessarily
+ * precisely what you asked for. For more information, look at
+ * <code>ResourceBundle</code>.
+ *
+ * <P>
+ * The <code>Locale</code> class provides a number of convenient constants
+ * that you can use to create <code>Locale</code> objects for commonly used
+ * locales. For example, the following refers to a <code>Locale</code> object
+ * for the United States:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ *       Locale::getUS()
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ *
+ * <P>
+ * Once you've created a <code>Locale</code> you can query it for information about
+ * itself. Use <code>getCountry</code> to get the ISO Country Code and
+ * <code>getLanguage</code> to get the ISO Language Code. You can
+ * use <code>getDisplayCountry</code> to get the
+ * name of the country suitable for displaying to the user. Similarly,
+ * you can use <code>getDisplayLanguage</code> to get the name of
+ * the language suitable for displaying to the user. Interestingly,
+ * the <code>getDisplayXXX</code> methods are themselves locale-sensitive
+ * and have two versions: one that uses the default locale and one
+ * that takes a locale as an argument and displays the name or country in
+ * a language appropriate to that locale.
+ *
+ * <P>
+ * ICU provides a number of classes that perform locale-sensitive
+ * operations. For example, the <code>NumberFormat</code> class formats
+ * numbers, currency, or percentages in a locale-sensitive manner. Classes
+ * such as <code>NumberFormat</code> have a number of convenience methods
+ * for creating a default object of that type. For example, the
+ * <code>NumberFormat</code> class provides these three convenience methods
+ * for creating a default <code>NumberFormat</code> object:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ *     UErrorCode success = U_ZERO_ERROR;
+ *     Locale myLocale;
+ *     NumberFormat *nf;
+ *
+ *     nf = NumberFormat::createInstance( success );          delete nf;
+ *     nf = NumberFormat::createCurrencyInstance( success );  delete nf;
+ *     nf = NumberFormat::createPercentInstance( success );   delete nf;
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * Each of these methods has two variants; one with an explicit locale
+ * and one without; the latter using the default locale.
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ *     nf = NumberFormat::createInstance( myLocale, success );          delete nf;
+ *     nf = NumberFormat::createCurrencyInstance( myLocale, success );  delete nf;
+ *     nf = NumberFormat::createPercentInstance( myLocale, success );   delete nf;
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * A <code>Locale</code> is the mechanism for identifying the kind of object
+ * (<code>NumberFormat</code>) that you would like to get. The locale is
+ * <STRONG>just</STRONG> a mechanism for identifying objects,
+ * <STRONG>not</STRONG> a container for the objects themselves.
+ *
+ * <P>
+ * Each class that performs locale-sensitive operations allows you
+ * to get all the available objects of that type. You can sift
+ * through these objects by language, country, or variant,
+ * and use the display names to present a menu to the user.
+ * For example, you can create a menu of all the collation objects
+ * suitable for a given language. Such classes implement these
+ * three class methods:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ *       static Locale* getAvailableLocales(int32_t& numLocales)
+ *       static UnicodeString& getDisplayName(const Locale&  objectLocale,
+ *                                            const Locale&  displayLocale,
+ *                                            UnicodeString& displayName)
+ *       static UnicodeString& getDisplayName(const Locale&  objectLocale,
+ *                                            UnicodeString& displayName)
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ *
+ * @stable ICU 2.0
+ * @see ResourceBundle
+ */
+class U_COMMON_API Locale : public UObject {
+public:
+    /** Useful constant for the Root locale. @stable ICU 4.4 */
+    static const Locale &U_EXPORT2 getRoot(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getEnglish(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getFrench(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getGerman(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getItalian(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getJapanese(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getKorean(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getChinese(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getSimplifiedChinese(void);
+    /** Useful constant for this language. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getTraditionalChinese(void);
+
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getFrance(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getGermany(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getItaly(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getJapan(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getKorea(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getChina(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getPRC(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getTaiwan(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getUK(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getUS(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getCanada(void);
+    /** Useful constant for this country/region. @stable ICU 2.0 */
+    static const Locale &U_EXPORT2 getCanadaFrench(void);
+
+
+    /**
+     * Construct a default locale object, a Locale for the default locale ID.
+     *
+     * @see getDefault
+     * @see uloc_getDefault
+     * @stable ICU 2.0
+     */
+    Locale();
+
+    /**
+     * Construct a locale from language, country, variant.
+     * If an error occurs, then the constructed object will be "bogus"
+     * (isBogus() will return true).
+     *
+     * @param language Lowercase two-letter or three-letter ISO-639 code.
+     *  This parameter can instead be an ICU style C locale (e.g. "en_US"),
+     *  but the other parameters must not be used.
+     *  This parameter can be NULL; if so,
+     *  the locale is initialized to match the current default locale.
+     *  (This is the same as using the default constructor.)
+     *  Please note: The Java Locale class does NOT accept the form
+     *  'new Locale("en_US")' but only 'new Locale("en","US")'
+     *
+     * @param country  Uppercase two-letter ISO-3166 code. (optional)
+     * @param variant  Uppercase vendor and browser specific code. See class
+     *                 description. (optional)
+     * @param keywordsAndValues A string consisting of keyword/values pairs, such as
+     *                 "collation=phonebook;currency=euro"
+     *
+     * @see getDefault
+     * @see uloc_getDefault
+     * @stable ICU 2.0
+     */
+    Locale( const   char * language,
+            const   char * country  = 0,
+            const   char * variant  = 0,
+            const   char * keywordsAndValues = 0);
+
+    /**
+     * Initializes a Locale object from another Locale object.
+     *
+     * @param other The Locale object being copied in.
+     * @stable ICU 2.0
+     */
+    Locale(const    Locale& other);
+
+    /**
+     * Move constructor; might leave source in bogus state.
+     * This locale will have the same contents that the source locale had.
+     *
+     * @param other The Locale object being moved in.
+     * @stable ICU 63
+     */
+    Locale(Locale&& other) U_NOEXCEPT;
+
+    /**
+     * Destructor
+     * @stable ICU 2.0
+     */
+    virtual ~Locale() ;
+
+    /**
+     * Replaces the entire contents of *this with the specified value.
+     *
+     * @param other The Locale object being copied in.
+     * @return      *this
+     * @stable ICU 2.0
+     */
+    Locale& operator=(const Locale& other);
+
+    /**
+     * Move assignment operator; might leave source in bogus state.
+     * This locale will have the same contents that the source locale had.
+     * The behavior is undefined if *this and the source are the same object.
+     *
+     * @param other The Locale object being moved in.
+     * @return      *this
+     * @stable ICU 63
+     */
+    Locale& operator=(Locale&& other) U_NOEXCEPT;
+
+    /**
+     * Checks if two locale keys are the same.
+     *
+     * @param other The locale key object to be compared with this.
+     * @return      True if the two locale keys are the same, false otherwise.
+     * @stable ICU 2.0
+     */
+    UBool   operator==(const    Locale&     other) const;
+
+    /**
+     * Checks if two locale keys are not the same.
+     *
+     * @param other The locale key object to be compared with this.
+     * @return      True if the two locale keys are not the same, false
+     *              otherwise.
+     * @stable ICU 2.0
+     */
+    inline UBool   operator!=(const    Locale&     other) const;
+
+    /**
+     * Clone this object.
+     * Clones can be used concurrently in multiple threads.
+     * If an error occurs, then NULL is returned.
+     * The caller must delete the clone.
+     *
+     * @return a clone of this object
+     *
+     * @see getDynamicClassID
+     * @stable ICU 2.8
+     */
+    Locale *clone() const;
+
+#ifndef U_HIDE_SYSTEM_API
+    /**
+     * Common methods of getting the current default Locale. Used for the
+     * presentation: menus, dialogs, etc. Generally set once when your applet or
+     * application is initialized, then never reset. (If you do reset the
+     * default locale, you probably want to reload your GUI, so that the change
+     * is reflected in your interface.)
+     *
+     * More advanced programs will allow users to use different locales for
+     * different fields, e.g. in a spreadsheet.
+     *
+     * Note that the initial setting will match the host system.
+     * @return a reference to the Locale object for the default locale ID
+     * @system
+     * @stable ICU 2.0
+     */
+    static const Locale& U_EXPORT2 getDefault(void);
+
+    /**
+     * Sets the default. Normally set once at the beginning of a process,
+     * then never reset.
+     * setDefault() only changes ICU's default locale ID, <strong>not</strong>
+     * the default locale ID of the runtime environment.
+     *
+     * @param newLocale Locale to set to.  If NULL, set to the value obtained
+     *                  from the runtime environment.
+     * @param success The error code.
+     * @system
+     * @stable ICU 2.0
+     */
+    static void U_EXPORT2 setDefault(const Locale& newLocale,
+                                     UErrorCode&   success);
+#endif  /* U_HIDE_SYSTEM_API */
+
+    /**
+     * Returns a Locale for the specified BCP47 language tag string.
+     * If the specified language tag contains any ill-formed subtags,
+     * the first such subtag and all following subtags are ignored.
+     * <p>
+     * This implements the 'Language-Tag' production of BCP 47, and so
+     * supports legacy language tags (marked as “Type: grandfathered” in BCP 47)
+     * (regular and irregular) as well as private use language tags.
+     *
+     * Private use tags are represented as 'x-whatever',
+     * and legacy tags are converted to their canonical replacements where they exist.
+     *
+     * Note that a few legacy tags have no modern replacement;
+     * these will be converted using the fallback described in
+     * the first paragraph, so some information might be lost.
+     *
+     * @param tag     the input BCP47 language tag.
+     * @param status  error information if creating the Locale failed.
+     * @return        the Locale for the specified BCP47 language tag.
+     * @stable ICU 63
+     */
+    static Locale U_EXPORT2 forLanguageTag(StringPiece tag, UErrorCode& status);
+
+    /**
+     * Returns a well-formed language tag for this Locale.
+     * <p>
+     * <b>Note</b>: Any locale fields which do not satisfy the BCP47 syntax
+     * requirement will be silently omitted from the result.
+     *
+     * If this function fails, partial output may have been written to the sink.
+     *
+     * @param sink    the output sink receiving the BCP47 language
+     *                tag for this Locale.
+     * @param status  error information if creating the language tag failed.
+     * @stable ICU 63
+     */
+    void toLanguageTag(ByteSink& sink, UErrorCode& status) const;
+
+    /**
+     * Returns a well-formed language tag for this Locale.
+     * <p>
+     * <b>Note</b>: Any locale fields which do not satisfy the BCP47 syntax
+     * requirement will be silently omitted from the result.
+     *
+     * @param status  error information if creating the language tag failed.
+     * @return        the BCP47 language tag for this Locale.
+     * @stable ICU 63
+     */
+    template<typename StringClass>
+    inline StringClass toLanguageTag(UErrorCode& status) const;
+
+    /**
+     * Creates a locale which has had minimal canonicalization
+     * as per uloc_getName().
+     * @param name The name to create from.  If name is null,
+     *  the default Locale is used.
+     * @return new locale object
+     * @stable ICU 2.0
+     * @see uloc_getName
+     */
+    static Locale U_EXPORT2 createFromName(const char *name);
+
+    /**
+     * Creates a locale from the given string after canonicalizing
+     * the string according to CLDR by calling uloc_canonicalize().
+     * @param name the locale ID to create from.  Must not be NULL.
+     * @return a new locale object corresponding to the given name
+     * @stable ICU 3.0
+     * @see uloc_canonicalize
+     */
+    static Locale U_EXPORT2 createCanonical(const char* name);
+
+    /**
+     * Returns the locale's ISO-639 language code.
+     * @return      An alias to the code
+     * @stable ICU 2.0
+     */
+    inline const char *  getLanguage( ) const;
+
+    /**
+     * Returns the locale's ISO-15924 abbreviation script code.
+     * @return      An alias to the code
+     * @see uscript_getShortName
+     * @see uscript_getCode
+     * @stable ICU 2.8
+     */
+    inline const char *  getScript( ) const;
+
+    /**
+     * Returns the locale's ISO-3166 country code.
+     * @return      An alias to the code
+     * @stable ICU 2.0
+     */
+    inline const char *  getCountry( ) const;
+
+    /**
+     * Returns the locale's variant code.
+     * @return      An alias to the code
+     * @stable ICU 2.0
+     */
+    inline const char *  getVariant( ) const;
+
+    /**
+     * Returns the programmatic name of the entire locale, with the language,
+     * country and variant separated by underbars. If a field is missing, up
+     * to two leading underbars will occur. Example: "en", "de_DE", "en_US_WIN",
+     * "de__POSIX", "fr__MAC", "__MAC", "_MT", "_FR_EURO"
+     * @return      A pointer to "name".
+     * @stable ICU 2.0
+     */
+    inline const char * getName() const;
+
+    /**
+     * Returns the programmatic name of the entire locale as getName() would return,
+     * but without keywords.
+     * @return      A pointer to "name".
+     * @see getName
+     * @stable ICU 2.8
+     */
+    const char * getBaseName() const;
+
+    /**
+     * Add the likely subtags for this Locale, per the algorithm described
+     * in the following CLDR technical report:
+     *
+     *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+     *
+     * If this Locale is already in the maximal form, or not valid, or there is
+     * no data available for maximization, the Locale will be unchanged.
+     *
+     * For example, "und-Zzzz" cannot be maximized, since there is no
+     * reasonable maximization.
+     *
+     * Examples:
+     *
+     * "en" maximizes to "en_Latn_US"
+     *
+     * "de" maximizes to "de_Latn_US"
+     *
+     * "sr" maximizes to "sr_Cyrl_RS"
+     *
+     * "sh" maximizes to "sr_Latn_RS" (Note this will not reverse.)
+     *
+     * "zh_Hani" maximizes to "zh_Hans_CN" (Note this will not reverse.)
+     *
+     * @param status  error information if maximizing this Locale failed.
+     *                If this Locale is not well-formed, the error code is
+     *                U_ILLEGAL_ARGUMENT_ERROR.
+     * @stable ICU 63
+     */
+    void addLikelySubtags(UErrorCode& status);
+
+    /**
+     * Minimize the subtags for this Locale, per the algorithm described
+     * in the following CLDR technical report:
+     *
+     *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+     *
+     * If this Locale is already in the minimal form, or not valid, or there is
+     * no data available for minimization, the Locale will be unchanged.
+     *
+     * Since the minimization algorithm relies on proper maximization, see the
+     * comments for addLikelySubtags for reasons why there might not be any
+     * data.
+     *
+     * Examples:
+     *
+     * "en_Latn_US" minimizes to "en"
+     *
+     * "de_Latn_US" minimizes to "de"
+     *
+     * "sr_Cyrl_RS" minimizes to "sr"
+     *
+     * "zh_Hant_TW" minimizes to "zh_TW" (The region is preferred to the
+     * script, and minimizing to "zh" would imply "zh_Hans_CN".)
+     *
+     * @param status  error information if maximizing this Locale failed.
+     *                If this Locale is not well-formed, the error code is
+     *                U_ILLEGAL_ARGUMENT_ERROR.
+     * @stable ICU 63
+     */
+    void minimizeSubtags(UErrorCode& status);
+
+#ifndef U_HIDE_DRAFT_API
+    /**
+     * Canonicalize the locale ID of this object according to CLDR.
+     * @param status the status code
+     * @draft ICU 67
+     * @see createCanonical
+     */
+    void canonicalize(UErrorCode& status);
+#endif  // U_HIDE_DRAFT_API
+
+    /**
+     * Gets the list of keywords for the specified locale.
+     *
+     * @param status the status code
+     * @return pointer to StringEnumeration class, or NULL if there are no keywords. 
+     * Client must dispose of it by calling delete.
+     * @see getKeywords
+     * @stable ICU 2.8
+     */
+    StringEnumeration * createKeywords(UErrorCode &status) const;
+
+    /**
+     * Gets the list of Unicode keywords for the specified locale.
+     *
+     * @param status the status code
+     * @return pointer to StringEnumeration class, or NULL if there are no keywords.
+     * Client must dispose of it by calling delete.
+     * @see getUnicodeKeywords
+     * @stable ICU 63
+     */
+    StringEnumeration * createUnicodeKeywords(UErrorCode &status) const;
+
+    /**
+     * Gets the set of keywords for this Locale.
+     *
+     * A wrapper to call createKeywords() and write the resulting
+     * keywords as standard strings (or compatible objects) into any kind of
+     * container that can be written to by an STL style output iterator.
+     *
+     * @param iterator  an STL style output iterator to write the keywords to.
+     * @param status    error information if creating set of keywords failed.
+     * @stable ICU 63
+     */
+    template<typename StringClass, typename OutputIterator>
+    inline void getKeywords(OutputIterator iterator, UErrorCode& status) const;
+
+    /**
+     * Gets the set of Unicode keywords for this Locale.
+     *
+     * A wrapper to call createUnicodeKeywords() and write the resulting
+     * keywords as standard strings (or compatible objects) into any kind of
+     * container that can be written to by an STL style output iterator.
+     *
+     * @param iterator  an STL style output iterator to write the keywords to.
+     * @param status    error information if creating set of keywords failed.
+     * @stable ICU 63
+     */
+    template<typename StringClass, typename OutputIterator>
+    inline void getUnicodeKeywords(OutputIterator iterator, UErrorCode& status) const;
+
+    /**
+     * Gets the value for a keyword.
+     *
+     * This uses legacy keyword=value pairs, like "collation=phonebook".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName name of the keyword for which we want the value. Case insensitive.
+     * @param buffer The buffer to receive the keyword value.
+     * @param bufferCapacity The capacity of receiving buffer
+     * @param status Returns any error information while performing this operation.
+     * @return the length of the keyword value
+     *
+     * @stable ICU 2.8
+     */
+    int32_t getKeywordValue(const char* keywordName, char *buffer, int32_t bufferCapacity, UErrorCode &status) const;
+
+    /**
+     * Gets the value for a keyword.
+     *
+     * This uses legacy keyword=value pairs, like "collation=phonebook".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName  name of the keyword for which we want the value.
+     * @param sink         the sink to receive the keyword value.
+     * @param status       error information if getting the value failed.
+     * @stable ICU 63
+     */
+    void getKeywordValue(StringPiece keywordName, ByteSink& sink, UErrorCode& status) const;
+
+    /**
+     * Gets the value for a keyword.
+     *
+     * This uses legacy keyword=value pairs, like "collation=phonebook".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName  name of the keyword for which we want the value.
+     * @param status       error information if getting the value failed.
+     * @return             the keyword value.
+     * @stable ICU 63
+     */
+    template<typename StringClass>
+    inline StringClass getKeywordValue(StringPiece keywordName, UErrorCode& status) const;
+
+    /**
+     * Gets the Unicode value for a Unicode keyword.
+     *
+     * This uses Unicode key-value pairs, like "co-phonebk".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName  name of the keyword for which we want the value.
+     * @param sink         the sink to receive the keyword value.
+     * @param status       error information if getting the value failed.
+     * @stable ICU 63
+     */
+    void getUnicodeKeywordValue(StringPiece keywordName, ByteSink& sink, UErrorCode& status) const;
+
+    /**
+     * Gets the Unicode value for a Unicode keyword.
+     *
+     * This uses Unicode key-value pairs, like "co-phonebk".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName  name of the keyword for which we want the value.
+     * @param status       error information if getting the value failed.
+     * @return             the keyword value.
+     * @stable ICU 63
+     */
+    template<typename StringClass>
+    inline StringClass getUnicodeKeywordValue(StringPiece keywordName, UErrorCode& status) const;
+
+    /**
+     * Sets or removes the value for a keyword.
+     *
+     * For removing all keywords, use getBaseName(),
+     * and construct a new Locale if it differs from getName().
+     *
+     * This uses legacy keyword=value pairs, like "collation=phonebook".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName name of the keyword to be set. Case insensitive.
+     * @param keywordValue value of the keyword to be set. If 0-length or
+     *  NULL, will result in the keyword being removed. No error is given if
+     *  that keyword does not exist.
+     * @param status Returns any error information while performing this operation.
+     *
+     * @stable ICU 49
+     */
+    void setKeywordValue(const char* keywordName, const char* keywordValue, UErrorCode &status);
+
+    /**
+     * Sets or removes the value for a keyword.
+     *
+     * For removing all keywords, use getBaseName(),
+     * and construct a new Locale if it differs from getName().
+     *
+     * This uses legacy keyword=value pairs, like "collation=phonebook".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName name of the keyword to be set.
+     * @param keywordValue value of the keyword to be set. If 0-length or
+     *  NULL, will result in the keyword being removed. No error is given if
+     *  that keyword does not exist.
+     * @param status Returns any error information while performing this operation.
+     * @stable ICU 63
+     */
+    void setKeywordValue(StringPiece keywordName, StringPiece keywordValue, UErrorCode& status);
+
+    /**
+     * Sets or removes the Unicode value for a Unicode keyword.
+     *
+     * For removing all keywords, use getBaseName(),
+     * and construct a new Locale if it differs from getName().
+     *
+     * This uses Unicode key-value pairs, like "co-phonebk".
+     *
+     * ICU4C doesn't do automatic conversion between legacy and Unicode
+     * keywords and values in getters and setters (as opposed to ICU4J).
+     *
+     * @param keywordName name of the keyword to be set.
+     * @param keywordValue value of the keyword to be set. If 0-length or
+     *  NULL, will result in the keyword being removed. No error is given if
+     *  that keyword does not exist.
+     * @param status Returns any error information while performing this operation.
+     * @stable ICU 63
+     */
+    void setUnicodeKeywordValue(StringPiece keywordName, StringPiece keywordValue, UErrorCode& status);
+
+    /**
+     * returns the locale's three-letter language code, as specified
+     * in ISO draft standard ISO-639-2.
+     * @return      An alias to the code, or an empty string
+     * @stable ICU 2.0
+     */
+    const char * getISO3Language() const;
+
+    /**
+     * Fills in "name" with the locale's three-letter ISO-3166 country code.
+     * @return      An alias to the code, or an empty string
+     * @stable ICU 2.0
+     */
+    const char * getISO3Country() const;
+
+    /**
+     * Returns the Windows LCID value corresponding to this locale.
+     * This value is stored in the resource data for the locale as a one-to-four-digit
+     * hexadecimal number.  If the resource is missing, in the wrong format, or
+     * there is no Windows LCID value that corresponds to this locale, returns 0.
+     * @stable ICU 2.0
+     */
+    uint32_t        getLCID(void) const;
+
+    /**
+     * Returns whether this locale's script is written right-to-left.
+     * If there is no script subtag, then the likely script is used, see uloc_addLikelySubtags().
+     * If no likely script is known, then false is returned.
+     *
+     * A script is right-to-left according to the CLDR script metadata
+     * which corresponds to whether the script's letters have Bidi_Class=R or AL.
+     *
+     * Returns true for "ar" and "en-Hebr", false for "zh" and "fa-Cyrl".
+     *
+     * @return true if the locale's script is written right-to-left
+     * @stable ICU 54
+     */
+    UBool isRightToLeft() const;
+
+    /**
+     * Fills in "dispLang" with the name of this locale's language in a format suitable for
+     * user display in the default locale.  For example, if the locale's language code is
+     * "fr" and the default locale's language code is "en", this function would set
+     * dispLang to "French".
+     * @param dispLang  Receives the language's display name.
+     * @return          A reference to "dispLang".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayLanguage(UnicodeString&   dispLang) const;
+
+    /**
+     * Fills in "dispLang" with the name of this locale's language in a format suitable for
+     * user display in the locale specified by "displayLocale".  For example, if the locale's
+     * language code is "en" and displayLocale's language code is "fr", this function would set
+     * dispLang to "Anglais".
+     * @param displayLocale  Specifies the locale to be used to display the name.  In other words,
+     *                  if the locale's language code is "en", passing Locale::getFrench() for
+     *                  displayLocale would result in "Anglais", while passing Locale::getGerman()
+     *                  for displayLocale would result in "Englisch".
+     * @param dispLang  Receives the language's display name.
+     * @return          A reference to "dispLang".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayLanguage( const   Locale&         displayLocale,
+                                                UnicodeString&  dispLang) const;
+
+    /**
+     * Fills in "dispScript" with the name of this locale's script in a format suitable
+     * for user display in the default locale.  For example, if the locale's script code
+     * is "LATN" and the default locale's language code is "en", this function would set
+     * dispScript to "Latin".
+     * @param dispScript    Receives the scripts's display name.
+     * @return              A reference to "dispScript".
+     * @stable ICU 2.8
+     */
+    UnicodeString&  getDisplayScript(          UnicodeString& dispScript) const;
+
+    /**
+     * Fills in "dispScript" with the name of this locale's country in a format suitable
+     * for user display in the locale specified by "displayLocale".  For example, if the locale's
+     * script code is "LATN" and displayLocale's language code is "en", this function would set
+     * dispScript to "Latin".
+     * @param displayLocale      Specifies the locale to be used to display the name.  In other
+     *                      words, if the locale's script code is "LATN", passing
+     *                      Locale::getFrench() for displayLocale would result in "", while
+     *                      passing Locale::getGerman() for displayLocale would result in
+     *                      "".
+     * @param dispScript    Receives the scripts's display name.
+     * @return              A reference to "dispScript".
+     * @stable ICU 2.8
+     */
+    UnicodeString&  getDisplayScript(  const   Locale&         displayLocale,
+                                               UnicodeString&  dispScript) const;
+
+    /**
+     * Fills in "dispCountry" with the name of this locale's country in a format suitable
+     * for user display in the default locale.  For example, if the locale's country code
+     * is "FR" and the default locale's language code is "en", this function would set
+     * dispCountry to "France".
+     * @param dispCountry   Receives the country's display name.
+     * @return              A reference to "dispCountry".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayCountry(          UnicodeString& dispCountry) const;
+
+    /**
+     * Fills in "dispCountry" with the name of this locale's country in a format suitable
+     * for user display in the locale specified by "displayLocale".  For example, if the locale's
+     * country code is "US" and displayLocale's language code is "fr", this function would set
+     * dispCountry to "&Eacute;tats-Unis".
+     * @param displayLocale      Specifies the locale to be used to display the name.  In other
+     *                      words, if the locale's country code is "US", passing
+     *                      Locale::getFrench() for displayLocale would result in "&Eacute;tats-Unis", while
+     *                      passing Locale::getGerman() for displayLocale would result in
+     *                      "Vereinigte Staaten".
+     * @param dispCountry   Receives the country's display name.
+     * @return              A reference to "dispCountry".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayCountry(  const   Locale&         displayLocale,
+                                                UnicodeString&  dispCountry) const;
+
+    /**
+     * Fills in "dispVar" with the name of this locale's variant code in a format suitable
+     * for user display in the default locale.
+     * @param dispVar   Receives the variant's name.
+     * @return          A reference to "dispVar".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayVariant(      UnicodeString& dispVar) const;
+
+    /**
+     * Fills in "dispVar" with the name of this locale's variant code in a format
+     * suitable for user display in the locale specified by "displayLocale".
+     * @param displayLocale  Specifies the locale to be used to display the name.
+     * @param dispVar   Receives the variant's display name.
+     * @return          A reference to "dispVar".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayVariant(  const   Locale&         displayLocale,
+                                                UnicodeString&  dispVar) const;
+
+    /**
+     * Fills in "name" with the name of this locale in a format suitable for user display
+     * in the default locale.  This function uses getDisplayLanguage(), getDisplayCountry(),
+     * and getDisplayVariant() to do its work, and outputs the display name in the format
+     * "language (country[,variant])".  For example, if the default locale is en_US, then
+     * fr_FR's display name would be "French (France)", and es_MX_Traditional's display name
+     * would be "Spanish (Mexico,Traditional)".
+     * @param name  Receives the locale's display name.
+     * @return      A reference to "name".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayName(         UnicodeString&  name) const;
+
+    /**
+     * Fills in "name" with the name of this locale in a format suitable for user display
+     * in the locale specified by "displayLocale".  This function uses getDisplayLanguage(),
+     * getDisplayCountry(), and getDisplayVariant() to do its work, and outputs the display
+     * name in the format "language (country[,variant])".  For example, if displayLocale is
+     * fr_FR, then en_US's display name would be "Anglais (&Eacute;tats-Unis)", and no_NO_NY's
+     * display name would be "norv&eacute;gien (Norv&egrave;ge,NY)".
+     * @param displayLocale  Specifies the locale to be used to display the name.
+     * @param name      Receives the locale's display name.
+     * @return          A reference to "name".
+     * @stable ICU 2.0
+     */
+    UnicodeString&  getDisplayName( const   Locale&         displayLocale,
+                                            UnicodeString&  name) const;
+
+    /**
+     * Generates a hash code for the locale.
+     * @stable ICU 2.0
+     */
+    int32_t         hashCode(void) const;
+
+    /**
+     * Sets the locale to bogus
+     * A bogus locale represents a non-existing locale associated
+     * with services that can be instantiated from non-locale data
+     * in addition to locale (for example, collation can be
+     * instantiated from a locale and from a rule set).
+     * @stable ICU 2.1
+     */
+    void setToBogus();
+
+    /**
+     * Gets the bogus state. Locale object can be bogus if it doesn't exist
+     * @return false if it is a real locale, true if it is a bogus locale
+     * @stable ICU 2.1
+     */
+    inline UBool isBogus(void) const;
+
+    /**
+     * Returns a list of all installed locales.
+     * @param count Receives the number of locales in the list.
+     * @return      A pointer to an array of Locale objects.  This array is the list
+     *              of all locales with installed resource files.  The called does NOT
+     *              get ownership of this list, and must NOT delete it.
+     * @stable ICU 2.0
+     */
+    static const Locale* U_EXPORT2 getAvailableLocales(int32_t& count);
+
+    /**
+     * Gets a list of all available 2-letter country codes defined in ISO 3166.  This is a
+     * pointer to an array of pointers to arrays of char.  All of these pointers are
+     * owned by ICU-- do not delete them, and do not write through them.  The array is
+     * terminated with a null pointer.
+     * @return a list of all available country codes
+     * @stable ICU 2.0
+     */
+    static const char* const* U_EXPORT2 getISOCountries();
+
+    /**
+     * Gets a list of all available language codes defined in ISO 639.  This is a pointer
+     * to an array of pointers to arrays of char.  All of these pointers are owned
+     * by ICU-- do not delete them, and do not write through them.  The array is
+     * terminated with a null pointer.
+     * @return a list of all available language codes
+     * @stable ICU 2.0
+     */
+    static const char* const* U_EXPORT2 getISOLanguages();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.2
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     *
+     * @stable ICU 2.2
+     */
+    virtual UClassID getDynamicClassID() const;
+
+    /**
+     * A Locale iterator interface similar to a Java Iterator<Locale>.
+     * @stable ICU 65
+     */
+    class U_COMMON_API Iterator /* not : public UObject because this is an interface/mixin class */ {
+    public:
+        /** @stable ICU 65 */
+        virtual ~Iterator();
+
+        /**
+         * @return true if next() can be called again.
+         * @stable ICU 65
+         */
+        virtual UBool hasNext() const = 0;
+
+        /**
+         * @return the next locale.
+         * @stable ICU 65
+         */
+        virtual const Locale &next() = 0;
+    };
+
+    /**
+     * A generic Locale iterator implementation over Locale input iterators.
+     * @stable ICU 65
+     */
+    template<typename Iter>
+    class RangeIterator : public Iterator, public UMemory {
+    public:
+        /**
+         * Constructs an iterator from a begin/end range.
+         * Each of the iterator parameter values must be an
+         * input iterator whose value is convertible to const Locale &.
+         *
+         * @param begin Start of range.
+         * @param end Exclusive end of range.
+         * @stable ICU 65
+         */
+        RangeIterator(Iter begin, Iter end) : it_(begin), end_(end) {}
+
+        /**
+         * @return true if next() can be called again.
+         * @stable ICU 65
+         */
+        UBool hasNext() const override { return it_ != end_; }
+
+        /**
+         * @return the next locale.
+         * @stable ICU 65
+         */
+        const Locale &next() override { return *it_++; }
+
+    private:
+        Iter it_;
+        const Iter end_;
+    };
+
+    /**
+     * A generic Locale iterator implementation over Locale input iterators.
+     * Calls the converter to convert each *begin to a const Locale &.
+     * @stable ICU 65
+     */
+    template<typename Iter, typename Conv>
+    class ConvertingIterator : public Iterator, public UMemory {
+    public:
+        /**
+         * Constructs an iterator from a begin/end range.
+         * Each of the iterator parameter values must be an
+         * input iterator whose value the converter converts to const Locale &.
+         *
+         * @param begin Start of range.
+         * @param end Exclusive end of range.
+         * @param converter Converter from *begin to const Locale & or compatible.
+         * @stable ICU 65
+         */
+        ConvertingIterator(Iter begin, Iter end, Conv converter) :
+                it_(begin), end_(end), converter_(converter) {}
+
+        /**
+         * @return true if next() can be called again.
+         * @stable ICU 65
+         */
+        UBool hasNext() const override { return it_ != end_; }
+
+        /**
+         * @return the next locale.
+         * @stable ICU 65
+         */
+        const Locale &next() override { return converter_(*it_++); }
+
+    private:
+        Iter it_;
+        const Iter end_;
+        Conv converter_;
+    };
+
+protected: /* only protected for testing purposes. DO NOT USE. */
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Set this from a single POSIX style locale string.
+     * @internal
+     */
+    void setFromPOSIXID(const char *posixID);
+#endif  /* U_HIDE_INTERNAL_API */
+
+private:
+    /**
+     * Initialize the locale object with a new name.
+     * Was deprecated - used in implementation - moved internal
+     *
+     * @param cLocaleID The new locale name.
+     * @param canonicalize whether to call uloc_canonicalize on cLocaleID
+     */
+    Locale& init(const char* cLocaleID, UBool canonicalize);
+
+    /*
+     * Internal constructor to allow construction of a locale object with
+     *   NO side effects.   (Default constructor tries to get
+     *   the default locale.)
+     */
+    enum ELocaleType {
+        eBOGUS
+    };
+    Locale(ELocaleType);
+
+    /**
+     * Initialize the locale cache for commonly used locales
+     */
+    static Locale *getLocaleCache(void);
+
+    char language[ULOC_LANG_CAPACITY];
+    char script[ULOC_SCRIPT_CAPACITY];
+    char country[ULOC_COUNTRY_CAPACITY];
+    int32_t variantBegin;
+    char* fullName;
+    char fullNameBuffer[ULOC_FULLNAME_CAPACITY];
+    // name without keywords
+    char* baseName;
+    void initBaseName(UErrorCode& status);
+
+    UBool fIsBogus;
+
+    static const Locale &getLocale(int locid);
+
+    /**
+     * A friend to allow the default locale to be set by either the C or C++ API.
+     * @internal (private)
+     */
+    friend Locale *locale_set_default_internal(const char *, UErrorCode& status);
+
+    /**
+     * @internal (private)
+     */
+    friend void U_CALLCONV locale_available_init();
+};
+
+inline UBool
+Locale::operator!=(const    Locale&     other) const
+{
+    return !operator==(other);
+}
+
+template<typename StringClass> inline StringClass
+Locale::toLanguageTag(UErrorCode& status) const
+{
+    StringClass result;
+    StringByteSink<StringClass> sink(&result);
+    toLanguageTag(sink, status);
+    return result;
+}
+
+inline const char *
+Locale::getCountry() const
+{
+    return country;
+}
+
+inline const char *
+Locale::getLanguage() const
+{
+    return language;
+}
+
+inline const char *
+Locale::getScript() const
+{
+    return script;
+}
+
+inline const char *
+Locale::getVariant() const
+{
+    return &baseName[variantBegin];
+}
+
+inline const char *
+Locale::getName() const
+{
+    return fullName;
+}
+
+template<typename StringClass, typename OutputIterator> inline void
+Locale::getKeywords(OutputIterator iterator, UErrorCode& status) const
+{
+    LocalPointer<StringEnumeration> keys(createKeywords(status));
+    if (U_FAILURE(status) || keys.isNull()) {
+        return;
+    }
+    for (;;) {
+        int32_t resultLength;
+        const char* buffer = keys->next(&resultLength, status);
+        if (U_FAILURE(status) || buffer == nullptr) {
+            return;
+        }
+        *iterator++ = StringClass(buffer, resultLength);
+    }
+}
+
+template<typename StringClass, typename OutputIterator> inline void
+Locale::getUnicodeKeywords(OutputIterator iterator, UErrorCode& status) const
+{
+    LocalPointer<StringEnumeration> keys(createUnicodeKeywords(status));
+    if (U_FAILURE(status) || keys.isNull()) {
+        return;
+    }
+    for (;;) {
+        int32_t resultLength;
+        const char* buffer = keys->next(&resultLength, status);
+        if (U_FAILURE(status) || buffer == nullptr) {
+            return;
+        }
+        *iterator++ = StringClass(buffer, resultLength);
+    }
+}
+
+template<typename StringClass> inline StringClass
+Locale::getKeywordValue(StringPiece keywordName, UErrorCode& status) const
+{
+    StringClass result;
+    StringByteSink<StringClass> sink(&result);
+    getKeywordValue(keywordName, sink, status);
+    return result;
+}
+
+template<typename StringClass> inline StringClass
+Locale::getUnicodeKeywordValue(StringPiece keywordName, UErrorCode& status) const
+{
+    StringClass result;
+    StringByteSink<StringClass> sink(&result);
+    getUnicodeKeywordValue(keywordName, sink, status);
+    return result;
+}
+
+inline UBool
+Locale::isBogus(void) const {
+    return fIsBogus;
+}
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/messagepattern.h b/thirdparty/icu4c/common/unicode/messagepattern.h
new file mode 100644
index 0000000000..98e7b70b1f
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/messagepattern.h
@@ -0,0 +1,949 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  messagepattern.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011mar14
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __MESSAGEPATTERN_H__
+#define __MESSAGEPATTERN_H__
+
+/**
+ * \file
+ * \brief C++ API: MessagePattern class: Parses and represents ICU MessageFormat patterns.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#if !UCONFIG_NO_FORMATTING
+
+#include "unicode/parseerr.h"
+#include "unicode/unistr.h"
+
+/**
+ * Mode for when an apostrophe starts quoted literal text for MessageFormat output.
+ * The default is DOUBLE_OPTIONAL unless overridden via uconfig.h
+ * (UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE).
+ * <p>
+ * A pair of adjacent apostrophes always results in a single apostrophe in the output,
+ * even when the pair is between two single, text-quoting apostrophes.
+ * <p>
+ * The following table shows examples of desired MessageFormat.format() output
+ * with the pattern strings that yield that output.
+ * <p>
+ * <table>
+ *   <tr>
+ *     <th>Desired output</th>
+ *     <th>DOUBLE_OPTIONAL</th>
+ *     <th>DOUBLE_REQUIRED</th>
+ *   </tr>
+ *   <tr>
+ *     <td>I see {many}</td>
+ *     <td>I see '{many}'</td>
+ *     <td>(same)</td>
+ *   </tr>
+ *   <tr>
+ *     <td>I said {'Wow!'}</td>
+ *     <td>I said '{''Wow!''}'</td>
+ *     <td>(same)</td>
+ *   </tr>
+ *   <tr>
+ *     <td>I don't know</td>
+ *     <td>I don't know OR<br> I don''t know</td>
+ *     <td>I don''t know</td>
+ *   </tr>
+ * </table>
+ * @stable ICU 4.8
+ * @see UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE
+ */
+enum UMessagePatternApostropheMode {
+    /**
+     * A literal apostrophe is represented by
+     * either a single or a double apostrophe pattern character.
+     * Within a MessageFormat pattern, a single apostrophe only starts quoted literal text
+     * if it immediately precedes a curly brace {},
+     * or a pipe symbol | if inside a choice format,
+     * or a pound symbol # if inside a plural format.
+     * <p>
+     * This is the default behavior starting with ICU 4.8.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_APOS_DOUBLE_OPTIONAL,
+    /**
+     * A literal apostrophe must be represented by
+     * a double apostrophe pattern character.
+     * A single apostrophe always starts quoted literal text.
+     * <p>
+     * This is the behavior of ICU 4.6 and earlier, and of the JDK.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_APOS_DOUBLE_REQUIRED
+};
+/**
+ * @stable ICU 4.8
+ */
+typedef enum UMessagePatternApostropheMode UMessagePatternApostropheMode;
+
+/**
+ * MessagePattern::Part type constants.
+ * @stable ICU 4.8
+ */
+enum UMessagePatternPartType {
+    /**
+     * Start of a message pattern (main or nested).
+     * The length is 0 for the top-level message
+     * and for a choice argument sub-message, otherwise 1 for the '{'.
+     * The value indicates the nesting level, starting with 0 for the main message.
+     * <p>
+     * There is always a later MSG_LIMIT part.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_MSG_START,
+    /**
+     * End of a message pattern (main or nested).
+     * The length is 0 for the top-level message and
+     * the last sub-message of a choice argument,
+     * otherwise 1 for the '}' or (in a choice argument style) the '|'.
+     * The value indicates the nesting level, starting with 0 for the main message.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_MSG_LIMIT,
+    /**
+     * Indicates a substring of the pattern string which is to be skipped when formatting.
+     * For example, an apostrophe that begins or ends quoted text
+     * would be indicated with such a part.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_SKIP_SYNTAX,
+    /**
+     * Indicates that a syntax character needs to be inserted for auto-quoting.
+     * The length is 0.
+     * The value is the character code of the insertion character. (U+0027=APOSTROPHE)
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_INSERT_CHAR,
+    /**
+     * Indicates a syntactic (non-escaped) # symbol in a plural variant.
+     * When formatting, replace this part's substring with the
+     * (value-offset) for the plural argument value.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_REPLACE_NUMBER,
+    /**
+     * Start of an argument.
+     * The length is 1 for the '{'.
+     * The value is the ordinal value of the ArgType. Use getArgType().
+     * <p>
+     * This part is followed by either an ARG_NUMBER or ARG_NAME,
+     * followed by optional argument sub-parts (see UMessagePatternArgType constants)
+     * and finally an ARG_LIMIT part.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_START,
+    /**
+     * End of an argument.
+     * The length is 1 for the '}'.
+     * The value is the ordinal value of the ArgType. Use getArgType().
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_LIMIT,
+    /**
+     * The argument number, provided by the value.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_NUMBER,
+    /**
+     * The argument name.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_NAME,
+    /**
+     * The argument type.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_TYPE,
+    /**
+     * The argument style text.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_STYLE,
+    /**
+     * A selector substring in a "complex" argument style.
+     * The value is undefined and currently always 0.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_SELECTOR,
+    /**
+     * An integer value, for example the offset or an explicit selector value
+     * in a PluralFormat style.
+     * The part value is the integer value.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_INT,
+    /**
+     * A numeric value, for example the offset or an explicit selector value
+     * in a PluralFormat style.
+     * The part value is an index into an internal array of numeric values;
+     * use getNumericValue().
+     * @stable ICU 4.8
+     */
+    UMSGPAT_PART_TYPE_ARG_DOUBLE
+};
+/**
+ * @stable ICU 4.8
+ */
+typedef enum UMessagePatternPartType UMessagePatternPartType;
+
+/**
+ * Argument type constants.
+ * Returned by Part.getArgType() for ARG_START and ARG_LIMIT parts.
+ *
+ * Messages nested inside an argument are each delimited by MSG_START and MSG_LIMIT,
+ * with a nesting level one greater than the surrounding message.
+ * @stable ICU 4.8
+ */
+enum UMessagePatternArgType {
+    /**
+     * The argument has no specified type.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_TYPE_NONE,
+    /**
+     * The argument has a "simple" type which is provided by the ARG_TYPE part.
+     * An ARG_STYLE part might follow that.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_TYPE_SIMPLE,
+    /**
+     * The argument is a ChoiceFormat with one or more
+     * ((ARG_INT | ARG_DOUBLE), ARG_SELECTOR, message) tuples.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_TYPE_CHOICE,
+    /**
+     * The argument is a cardinal-number PluralFormat with an optional ARG_INT or ARG_DOUBLE offset
+     * (e.g., offset:1)
+     * and one or more (ARG_SELECTOR [explicit-value] message) tuples.
+     * If the selector has an explicit value (e.g., =2), then
+     * that value is provided by the ARG_INT or ARG_DOUBLE part preceding the message.
+     * Otherwise the message immediately follows the ARG_SELECTOR.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_TYPE_PLURAL,
+    /**
+     * The argument is a SelectFormat with one or more (ARG_SELECTOR, message) pairs.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_TYPE_SELECT,
+    /**
+     * The argument is an ordinal-number PluralFormat
+     * with the same style parts sequence and semantics as UMSGPAT_ARG_TYPE_PLURAL.
+     * @stable ICU 50
+     */
+    UMSGPAT_ARG_TYPE_SELECTORDINAL
+};
+/**
+ * @stable ICU 4.8
+ */
+typedef enum UMessagePatternArgType UMessagePatternArgType;
+
+/**
+ * \def UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE
+ * Returns true if the argument type has a plural style part sequence and semantics,
+ * for example UMSGPAT_ARG_TYPE_PLURAL and UMSGPAT_ARG_TYPE_SELECTORDINAL.
+ * @stable ICU 50
+ */
+#define UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(argType) \
+    ((argType)==UMSGPAT_ARG_TYPE_PLURAL || (argType)==UMSGPAT_ARG_TYPE_SELECTORDINAL)
+
+enum {
+    /**
+     * Return value from MessagePattern.validateArgumentName() for when
+     * the string is a valid "pattern identifier" but not a number.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_NAME_NOT_NUMBER=-1,
+
+    /**
+     * Return value from MessagePattern.validateArgumentName() for when
+     * the string is invalid.
+     * It might not be a valid "pattern identifier",
+     * or it have only ASCII digits but there is a leading zero or the number is too large.
+     * @stable ICU 4.8
+     */
+    UMSGPAT_ARG_NAME_NOT_VALID=-2
+};
+
+/**
+ * Special value that is returned by getNumericValue(Part) when no
+ * numeric value is defined for a part.
+ * @see MessagePattern.getNumericValue()
+ * @stable ICU 4.8
+ */
+#define UMSGPAT_NO_NUMERIC_VALUE ((double)(-123456789))
+
+U_NAMESPACE_BEGIN
+
+class MessagePatternDoubleList;
+class MessagePatternPartsList;
+
+/**
+ * Parses and represents ICU MessageFormat patterns.
+ * Also handles patterns for ChoiceFormat, PluralFormat and SelectFormat.
+ * Used in the implementations of those classes as well as in tools
+ * for message validation, translation and format conversion.
+ * <p>
+ * The parser handles all syntax relevant for identifying message arguments.
+ * This includes "complex" arguments whose style strings contain
+ * nested MessageFormat pattern substrings.
+ * For "simple" arguments (with no nested MessageFormat pattern substrings),
+ * the argument style is not parsed any further.
+ * <p>
+ * The parser handles named and numbered message arguments and allows both in one message.
+ * <p>
+ * Once a pattern has been parsed successfully, iterate through the parsed data
+ * with countParts(), getPart() and related methods.
+ * <p>
+ * The data logically represents a parse tree, but is stored and accessed
+ * as a list of "parts" for fast and simple parsing and to minimize object allocations.
+ * Arguments and nested messages are best handled via recursion.
+ * For every _START "part", MessagePattern.getLimitPartIndex() efficiently returns
+ * the index of the corresponding _LIMIT "part".
+ * <p>
+ * List of "parts":
+ * <pre>
+ * message = MSG_START (SKIP_SYNTAX | INSERT_CHAR | REPLACE_NUMBER | argument)* MSG_LIMIT
+ * argument = noneArg | simpleArg | complexArg
+ * complexArg = choiceArg | pluralArg | selectArg
+ *
+ * noneArg = ARG_START.NONE (ARG_NAME | ARG_NUMBER) ARG_LIMIT.NONE
+ * simpleArg = ARG_START.SIMPLE (ARG_NAME | ARG_NUMBER) ARG_TYPE [ARG_STYLE] ARG_LIMIT.SIMPLE
+ * choiceArg = ARG_START.CHOICE (ARG_NAME | ARG_NUMBER) choiceStyle ARG_LIMIT.CHOICE
+ * pluralArg = ARG_START.PLURAL (ARG_NAME | ARG_NUMBER) pluralStyle ARG_LIMIT.PLURAL
+ * selectArg = ARG_START.SELECT (ARG_NAME | ARG_NUMBER) selectStyle ARG_LIMIT.SELECT
+ *
+ * choiceStyle = ((ARG_INT | ARG_DOUBLE) ARG_SELECTOR message)+
+ * pluralStyle = [ARG_INT | ARG_DOUBLE] (ARG_SELECTOR [ARG_INT | ARG_DOUBLE] message)+
+ * selectStyle = (ARG_SELECTOR message)+
+ * </pre>
+ * <ul>
+ *   <li>Literal output text is not represented directly by "parts" but accessed
+ *       between parts of a message, from one part's getLimit() to the next part's getIndex().
+ *   <li><code>ARG_START.CHOICE</code> stands for an ARG_START Part with ArgType CHOICE.
+ *   <li>In the choiceStyle, the ARG_SELECTOR has the '<', the '#' or
+ *       the less-than-or-equal-to sign (U+2264).
+ *   <li>In the pluralStyle, the first, optional numeric Part has the "offset:" value.
+ *       The optional numeric Part between each (ARG_SELECTOR, message) pair
+ *       is the value of an explicit-number selector like "=2",
+ *       otherwise the selector is a non-numeric identifier.
+ *   <li>The REPLACE_NUMBER Part can occur only in an immediate sub-message of the pluralStyle.
+ * </ul>
+ * <p>
+ * This class is not intended for public subclassing.
+ *
+ * @stable ICU 4.8
+ */
+class U_COMMON_API MessagePattern : public UObject {
+public:
+    /**
+     * Constructs an empty MessagePattern with default UMessagePatternApostropheMode.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @stable ICU 4.8
+     */
+    MessagePattern(UErrorCode &errorCode);
+
+    /**
+     * Constructs an empty MessagePattern.
+     * @param mode Explicit UMessagePatternApostropheMode.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @stable ICU 4.8
+     */
+    MessagePattern(UMessagePatternApostropheMode mode, UErrorCode &errorCode);
+
+    /**
+     * Constructs a MessagePattern with default UMessagePatternApostropheMode and
+     * parses the MessageFormat pattern string.
+     * @param pattern a MessageFormat pattern string
+     * @param parseError Struct to receive information on the position
+     *                   of an error within the pattern.
+     *                   Can be NULL.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * TODO: turn @throws into UErrorCode specifics?
+     * @throws IllegalArgumentException for syntax errors in the pattern string
+     * @throws IndexOutOfBoundsException if certain limits are exceeded
+     *         (e.g., argument number too high, argument name too long, etc.)
+     * @throws NumberFormatException if a number could not be parsed
+     * @stable ICU 4.8
+     */
+    MessagePattern(const UnicodeString &pattern, UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Copy constructor.
+     * @param other Object to copy.
+     * @stable ICU 4.8
+     */
+    MessagePattern(const MessagePattern &other);
+
+    /**
+     * Assignment operator.
+     * @param other Object to copy.
+     * @return *this=other
+     * @stable ICU 4.8
+     */
+    MessagePattern &operator=(const MessagePattern &other);
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    virtual ~MessagePattern();
+
+    /**
+     * Parses a MessageFormat pattern string.
+     * @param pattern a MessageFormat pattern string
+     * @param parseError Struct to receive information on the position
+     *                   of an error within the pattern.
+     *                   Can be NULL.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @throws IllegalArgumentException for syntax errors in the pattern string
+     * @throws IndexOutOfBoundsException if certain limits are exceeded
+     *         (e.g., argument number too high, argument name too long, etc.)
+     * @throws NumberFormatException if a number could not be parsed
+     * @stable ICU 4.8
+     */
+    MessagePattern &parse(const UnicodeString &pattern,
+                          UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Parses a ChoiceFormat pattern string.
+     * @param pattern a ChoiceFormat pattern string
+     * @param parseError Struct to receive information on the position
+     *                   of an error within the pattern.
+     *                   Can be NULL.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @throws IllegalArgumentException for syntax errors in the pattern string
+     * @throws IndexOutOfBoundsException if certain limits are exceeded
+     *         (e.g., argument number too high, argument name too long, etc.)
+     * @throws NumberFormatException if a number could not be parsed
+     * @stable ICU 4.8
+     */
+    MessagePattern &parseChoiceStyle(const UnicodeString &pattern,
+                                     UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Parses a PluralFormat pattern string.
+     * @param pattern a PluralFormat pattern string
+     * @param parseError Struct to receive information on the position
+     *                   of an error within the pattern.
+     *                   Can be NULL.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @throws IllegalArgumentException for syntax errors in the pattern string
+     * @throws IndexOutOfBoundsException if certain limits are exceeded
+     *         (e.g., argument number too high, argument name too long, etc.)
+     * @throws NumberFormatException if a number could not be parsed
+     * @stable ICU 4.8
+     */
+    MessagePattern &parsePluralStyle(const UnicodeString &pattern,
+                                     UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Parses a SelectFormat pattern string.
+     * @param pattern a SelectFormat pattern string
+     * @param parseError Struct to receive information on the position
+     *                   of an error within the pattern.
+     *                   Can be NULL.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @throws IllegalArgumentException for syntax errors in the pattern string
+     * @throws IndexOutOfBoundsException if certain limits are exceeded
+     *         (e.g., argument number too high, argument name too long, etc.)
+     * @throws NumberFormatException if a number could not be parsed
+     * @stable ICU 4.8
+     */
+    MessagePattern &parseSelectStyle(const UnicodeString &pattern,
+                                     UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Clears this MessagePattern.
+     * countParts() will return 0.
+     * @stable ICU 4.8
+     */
+    void clear();
+
+    /**
+     * Clears this MessagePattern and sets the UMessagePatternApostropheMode.
+     * countParts() will return 0.
+     * @param mode The new UMessagePatternApostropheMode.
+     * @stable ICU 4.8
+     */
+    void clearPatternAndSetApostropheMode(UMessagePatternApostropheMode mode) {
+        clear();
+        aposMode=mode;
+    }
+
+    /**
+     * @param other another object to compare with.
+     * @return true if this object is equivalent to the other one.
+     * @stable ICU 4.8
+     */
+    UBool operator==(const MessagePattern &other) const;
+
+    /**
+     * @param other another object to compare with.
+     * @return false if this object is equivalent to the other one.
+     * @stable ICU 4.8
+     */
+    inline UBool operator!=(const MessagePattern &other) const {
+        return !operator==(other);
+    }
+
+    /**
+     * @return A hash code for this object.
+     * @stable ICU 4.8
+     */
+    int32_t hashCode() const;
+
+    /**
+     * @return this instance's UMessagePatternApostropheMode.
+     * @stable ICU 4.8
+     */
+    UMessagePatternApostropheMode getApostropheMode() const {
+        return aposMode;
+    }
+
+    // Java has package-private jdkAposMode() here.
+    // In C++, this is declared in the MessageImpl class.
+
+    /**
+     * @return the parsed pattern string (null if none was parsed).
+     * @stable ICU 4.8
+     */
+    const UnicodeString &getPatternString() const {
+        return msg;
+    }
+
+    /**
+     * Does the parsed pattern have named arguments like {first_name}?
+     * @return true if the parsed pattern has at least one named argument.
+     * @stable ICU 4.8
+     */
+    UBool hasNamedArguments() const {
+        return hasArgNames;
+    }
+
+    /**
+     * Does the parsed pattern have numbered arguments like {2}?
+     * @return true if the parsed pattern has at least one numbered argument.
+     * @stable ICU 4.8
+     */
+    UBool hasNumberedArguments() const {
+        return hasArgNumbers;
+    }
+
+    /**
+     * Validates and parses an argument name or argument number string.
+     * An argument name must be a "pattern identifier", that is, it must contain
+     * no Unicode Pattern_Syntax or Pattern_White_Space characters.
+     * If it only contains ASCII digits, then it must be a small integer with no leading zero.
+     * @param name Input string.
+     * @return &gt;=0 if the name is a valid number,
+     *         ARG_NAME_NOT_NUMBER (-1) if it is a "pattern identifier" but not all ASCII digits,
+     *         ARG_NAME_NOT_VALID (-2) if it is neither.
+     * @stable ICU 4.8
+     */
+    static int32_t validateArgumentName(const UnicodeString &name);
+
+    /**
+     * Returns a version of the parsed pattern string where each ASCII apostrophe
+     * is doubled (escaped) if it is not already, and if it is not interpreted as quoting syntax.
+     * <p>
+     * For example, this turns "I don't '{know}' {gender,select,female{h''er}other{h'im}}."
+     * into "I don''t '{know}' {gender,select,female{h''er}other{h''im}}."
+     * @return the deep-auto-quoted version of the parsed pattern string.
+     * @see MessageFormat.autoQuoteApostrophe()
+     * @stable ICU 4.8
+     */
+    UnicodeString autoQuoteApostropheDeep() const;
+
+    class Part;
+
+    /**
+     * Returns the number of "parts" created by parsing the pattern string.
+     * Returns 0 if no pattern has been parsed or clear() was called.
+     * @return the number of pattern parts.
+     * @stable ICU 4.8
+     */
+    int32_t countParts() const {
+        return partsLength;
+    }
+
+    /**
+     * Gets the i-th pattern "part".
+     * @param i The index of the Part data. (0..countParts()-1)
+     * @return the i-th pattern "part".
+     * @stable ICU 4.8
+     */
+    const Part &getPart(int32_t i) const {
+        return parts[i];
+    }
+
+    /**
+     * Returns the UMessagePatternPartType of the i-th pattern "part".
+     * Convenience method for getPart(i).getType().
+     * @param i The index of the Part data. (0..countParts()-1)
+     * @return The UMessagePatternPartType of the i-th Part.
+     * @stable ICU 4.8
+     */
+    UMessagePatternPartType getPartType(int32_t i) const {
+        return getPart(i).type;
+    }
+
+    /**
+     * Returns the pattern index of the specified pattern "part".
+     * Convenience method for getPart(partIndex).getIndex().
+     * @param partIndex The index of the Part data. (0..countParts()-1)
+     * @return The pattern index of this Part.
+     * @stable ICU 4.8
+     */
+    int32_t getPatternIndex(int32_t partIndex) const {
+        return getPart(partIndex).index;
+    }
+
+    /**
+     * Returns the substring of the pattern string indicated by the Part.
+     * Convenience method for getPatternString().substring(part.getIndex(), part.getLimit()).
+     * @param part a part of this MessagePattern.
+     * @return the substring associated with part.
+     * @stable ICU 4.8
+     */
+    UnicodeString getSubstring(const Part &part) const {
+        return msg.tempSubString(part.index, part.length);
+    }
+
+    /**
+     * Compares the part's substring with the input string s.
+     * @param part a part of this MessagePattern.
+     * @param s a string.
+     * @return true if getSubstring(part).equals(s).
+     * @stable ICU 4.8
+     */
+    UBool partSubstringMatches(const Part &part, const UnicodeString &s) const {
+        return 0==msg.compare(part.index, part.length, s);
+    }
+
+    /**
+     * Returns the numeric value associated with an ARG_INT or ARG_DOUBLE.
+     * @param part a part of this MessagePattern.
+     * @return the part's numeric value, or UMSGPAT_NO_NUMERIC_VALUE if this is not a numeric part.
+     * @stable ICU 4.8
+     */
+    double getNumericValue(const Part &part) const;
+
+    /**
+     * Returns the "offset:" value of a PluralFormat argument, or 0 if none is specified.
+     * @param pluralStart the index of the first PluralFormat argument style part. (0..countParts()-1)
+     * @return the "offset:" value.
+     * @stable ICU 4.8
+     */
+    double getPluralOffset(int32_t pluralStart) const;
+
+    /**
+     * Returns the index of the ARG|MSG_LIMIT part corresponding to the ARG|MSG_START at start.
+     * @param start The index of some Part data (0..countParts()-1);
+     *        this Part should be of Type ARG_START or MSG_START.
+     * @return The first i>start where getPart(i).getType()==ARG|MSG_LIMIT at the same nesting level,
+     *         or start itself if getPartType(msgStart)!=ARG|MSG_START.
+     * @stable ICU 4.8
+     */
+    int32_t getLimitPartIndex(int32_t start) const {
+        int32_t limit=getPart(start).limitPartIndex;
+        if(limit<start) {
+            return start;
+        }
+        return limit;
+    }
+
+    /**
+     * A message pattern "part", representing a pattern parsing event.
+     * There is a part for the start and end of a message or argument,
+     * for quoting and escaping of and with ASCII apostrophes,
+     * and for syntax elements of "complex" arguments.
+     * @stable ICU 4.8
+     */
+    class Part : public UMemory {
+    public:
+        /**
+         * Default constructor, do not use.
+         * @internal
+         */
+        Part() {}
+
+        /**
+         * Returns the type of this part.
+         * @return the part type.
+         * @stable ICU 4.8
+         */
+        UMessagePatternPartType getType() const {
+            return type;
+        }
+
+        /**
+         * Returns the pattern string index associated with this Part.
+         * @return this part's pattern string index.
+         * @stable ICU 4.8
+         */
+        int32_t getIndex() const {
+            return index;
+        }
+
+        /**
+         * Returns the length of the pattern substring associated with this Part.
+         * This is 0 for some parts.
+         * @return this part's pattern substring length.
+         * @stable ICU 4.8
+         */
+        int32_t getLength() const {
+            return length;
+        }
+
+        /**
+         * Returns the pattern string limit (exclusive-end) index associated with this Part.
+         * Convenience method for getIndex()+getLength().
+         * @return this part's pattern string limit index, same as getIndex()+getLength().
+         * @stable ICU 4.8
+         */
+        int32_t getLimit() const {
+            return index+length;
+        }
+
+        /**
+         * Returns a value associated with this part.
+         * See the documentation of each part type for details.
+         * @return the part value.
+         * @stable ICU 4.8
+         */
+        int32_t getValue() const {
+            return value;
+        }
+
+        /**
+         * Returns the argument type if this part is of type ARG_START or ARG_LIMIT,
+         * otherwise UMSGPAT_ARG_TYPE_NONE.
+         * @return the argument type for this part.
+         * @stable ICU 4.8
+         */
+        UMessagePatternArgType getArgType() const {
+            UMessagePatternPartType msgType=getType();
+            if(msgType ==UMSGPAT_PART_TYPE_ARG_START || msgType ==UMSGPAT_PART_TYPE_ARG_LIMIT) {
+                return (UMessagePatternArgType)value;
+            } else {
+                return UMSGPAT_ARG_TYPE_NONE;
+            }
+        }
+
+        /**
+         * Indicates whether the Part type has a numeric value.
+         * If so, then that numeric value can be retrieved via MessagePattern.getNumericValue().
+         * @param type The Part type to be tested.
+         * @return true if the Part type has a numeric value.
+         * @stable ICU 4.8
+         */
+        static UBool hasNumericValue(UMessagePatternPartType type) {
+            return type==UMSGPAT_PART_TYPE_ARG_INT || type==UMSGPAT_PART_TYPE_ARG_DOUBLE;
+        }
+
+        /**
+         * @param other another object to compare with.
+         * @return true if this object is equivalent to the other one.
+         * @stable ICU 4.8
+         */
+        UBool operator==(const Part &other) const;
+
+        /**
+         * @param other another object to compare with.
+         * @return false if this object is equivalent to the other one.
+         * @stable ICU 4.8
+         */
+        inline UBool operator!=(const Part &other) const {
+            return !operator==(other);
+        }
+
+        /**
+         * @return A hash code for this object.
+         * @stable ICU 4.8
+         */
+        int32_t hashCode() const {
+            return ((type*37+index)*37+length)*37+value;
+        }
+
+    private:
+        friend class MessagePattern;
+
+        static const int32_t MAX_LENGTH=0xffff;
+        static const int32_t MAX_VALUE=0x7fff;
+
+        // Some fields are not final because they are modified during pattern parsing.
+        // After pattern parsing, the parts are effectively immutable.
+        UMessagePatternPartType type;
+        int32_t index;
+        uint16_t length;
+        int16_t value;
+        int32_t limitPartIndex;
+    };
+
+private:
+    void preParse(const UnicodeString &pattern, UParseError *parseError, UErrorCode &errorCode);
+
+    void postParse();
+
+    int32_t parseMessage(int32_t index, int32_t msgStartLength,
+                         int32_t nestingLevel, UMessagePatternArgType parentType,
+                         UParseError *parseError, UErrorCode &errorCode);
+
+    int32_t parseArg(int32_t index, int32_t argStartLength, int32_t nestingLevel,
+                     UParseError *parseError, UErrorCode &errorCode);
+
+    int32_t parseSimpleStyle(int32_t index, UParseError *parseError, UErrorCode &errorCode);
+
+    int32_t parseChoiceStyle(int32_t index, int32_t nestingLevel,
+                             UParseError *parseError, UErrorCode &errorCode);
+
+    int32_t parsePluralOrSelectStyle(UMessagePatternArgType argType, int32_t index, int32_t nestingLevel,
+                                     UParseError *parseError, UErrorCode &errorCode);
+
+    /**
+     * Validates and parses an argument name or argument number string.
+     * This internal method assumes that the input substring is a "pattern identifier".
+     * @return &gt;=0 if the name is a valid number,
+     *         ARG_NAME_NOT_NUMBER (-1) if it is a "pattern identifier" but not all ASCII digits,
+     *         ARG_NAME_NOT_VALID (-2) if it is neither.
+     * @see #validateArgumentName(String)
+     */
+    static int32_t parseArgNumber(const UnicodeString &s, int32_t start, int32_t limit);
+
+    int32_t parseArgNumber(int32_t start, int32_t limit) {
+        return parseArgNumber(msg, start, limit);
+    }
+
+    /**
+     * Parses a number from the specified message substring.
+     * @param start start index into the message string
+     * @param limit limit index into the message string, must be start<limit
+     * @param allowInfinity true if U+221E is allowed (for ChoiceFormat)
+     * @param parseError
+     * @param errorCode
+     */
+    void parseDouble(int32_t start, int32_t limit, UBool allowInfinity,
+                     UParseError *parseError, UErrorCode &errorCode);
+
+    // Java has package-private appendReducedApostrophes() here.
+    // In C++, this is declared in the MessageImpl class.
+
+    int32_t skipWhiteSpace(int32_t index);
+
+    int32_t skipIdentifier(int32_t index);
+
+    /**
+     * Skips a sequence of characters that could occur in a double value.
+     * Does not fully parse or validate the value.
+     */
+    int32_t skipDouble(int32_t index);
+
+    static UBool isArgTypeChar(UChar32 c);
+
+    UBool isChoice(int32_t index);
+
+    UBool isPlural(int32_t index);
+
+    UBool isSelect(int32_t index);
+
+    UBool isOrdinal(int32_t index);
+
+    /**
+     * @return true if we are inside a MessageFormat (sub-)pattern,
+     *         as opposed to inside a top-level choice/plural/select pattern.
+     */
+    UBool inMessageFormatPattern(int32_t nestingLevel);
+
+    /**
+     * @return true if we are in a MessageFormat sub-pattern
+     *         of a top-level ChoiceFormat pattern.
+     */
+    UBool inTopLevelChoiceMessage(int32_t nestingLevel, UMessagePatternArgType parentType);
+
+    void addPart(UMessagePatternPartType type, int32_t index, int32_t length,
+                 int32_t value, UErrorCode &errorCode);
+
+    void addLimitPart(int32_t start,
+                      UMessagePatternPartType type, int32_t index, int32_t length,
+                      int32_t value, UErrorCode &errorCode);
+
+    void addArgDoublePart(double numericValue, int32_t start, int32_t length, UErrorCode &errorCode);
+
+    void setParseError(UParseError *parseError, int32_t index);
+
+    UBool init(UErrorCode &errorCode);
+    UBool copyStorage(const MessagePattern &other, UErrorCode &errorCode);
+
+    UMessagePatternApostropheMode aposMode;
+    UnicodeString msg;
+    // ArrayList<Part> parts=new ArrayList<Part>();
+    MessagePatternPartsList *partsList;
+    Part *parts;
+    int32_t partsLength;
+    // ArrayList<Double> numericValues;
+    MessagePatternDoubleList *numericValuesList;
+    double *numericValues;
+    int32_t numericValuesLength;
+    UBool hasArgNames;
+    UBool hasArgNumbers;
+    UBool needsAutoQuoting;
+};
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_FORMATTING
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __MESSAGEPATTERN_H__
diff --git a/thirdparty/icu4c/common/unicode/normalizer2.h b/thirdparty/icu4c/common/unicode/normalizer2.h
new file mode 100644
index 0000000000..5eb1d95caf
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/normalizer2.h
@@ -0,0 +1,779 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  normalizer2.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009nov22
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __NORMALIZER2_H__
+#define __NORMALIZER2_H__
+
+/**
+ * \file
+ * \brief C++ API: New API for Unicode Normalization.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/stringpiece.h"
+#include "unicode/uniset.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm2.h"
+
+U_NAMESPACE_BEGIN
+
+class ByteSink;
+
+/**
+ * Unicode normalization functionality for standard Unicode normalization or
+ * for using custom mapping tables.
+ * All instances of this class are unmodifiable/immutable.
+ * Instances returned by getInstance() are singletons that must not be deleted by the caller.
+ * The Normalizer2 class is not intended for public subclassing.
+ *
+ * The primary functions are to produce a normalized string and to detect whether
+ * a string is already normalized.
+ * The most commonly used normalization forms are those defined in
+ * http://www.unicode.org/unicode/reports/tr15/
+ * However, this API supports additional normalization forms for specialized purposes.
+ * For example, NFKC_Casefold is provided via getInstance("nfkc_cf", COMPOSE)
+ * and can be used in implementations of UTS #46.
+ *
+ * Not only are the standard compose and decompose modes supplied,
+ * but additional modes are provided as documented in the Mode enum.
+ *
+ * Some of the functions in this class identify normalization boundaries.
+ * At a normalization boundary, the portions of the string
+ * before it and starting from it do not interact and can be handled independently.
+ *
+ * The spanQuickCheckYes() stops at a normalization boundary.
+ * When the goal is a normalized string, then the text before the boundary
+ * can be copied, and the remainder can be processed with normalizeSecondAndAppend().
+ *
+ * The hasBoundaryBefore(), hasBoundaryAfter() and isInert() functions test whether
+ * a character is guaranteed to be at a normalization boundary,
+ * regardless of context.
+ * This is used for moving from one normalization boundary to the next
+ * or preceding boundary, and for performing iterative normalization.
+ *
+ * Iterative normalization is useful when only a small portion of a
+ * longer string needs to be processed.
+ * For example, in ICU, iterative normalization is used by the NormalizationTransliterator
+ * (to avoid replacing already-normalized text) and ucol_nextSortKeyPart()
+ * (to process only the substring for which sort key bytes are computed).
+ *
+ * The set of normalization boundaries returned by these functions may not be
+ * complete: There may be more boundaries that could be returned.
+ * Different functions may return different boundaries.
+ * @stable ICU 4.4
+ */
+class U_COMMON_API Normalizer2 : public UObject {
+public:
+    /**
+     * Destructor.
+     * @stable ICU 4.4
+     */
+    ~Normalizer2();
+
+    /**
+     * Returns a Normalizer2 instance for Unicode NFC normalization.
+     * Same as getInstance(NULL, "nfc", UNORM2_COMPOSE, errorCode).
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 49
+     */
+    static const Normalizer2 *
+    getNFCInstance(UErrorCode &errorCode);
+
+    /**
+     * Returns a Normalizer2 instance for Unicode NFD normalization.
+     * Same as getInstance(NULL, "nfc", UNORM2_DECOMPOSE, errorCode).
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 49
+     */
+    static const Normalizer2 *
+    getNFDInstance(UErrorCode &errorCode);
+
+    /**
+     * Returns a Normalizer2 instance for Unicode NFKC normalization.
+     * Same as getInstance(NULL, "nfkc", UNORM2_COMPOSE, errorCode).
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 49
+     */
+    static const Normalizer2 *
+    getNFKCInstance(UErrorCode &errorCode);
+
+    /**
+     * Returns a Normalizer2 instance for Unicode NFKD normalization.
+     * Same as getInstance(NULL, "nfkc", UNORM2_DECOMPOSE, errorCode).
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 49
+     */
+    static const Normalizer2 *
+    getNFKDInstance(UErrorCode &errorCode);
+
+    /**
+     * Returns a Normalizer2 instance for Unicode NFKC_Casefold normalization.
+     * Same as getInstance(NULL, "nfkc_cf", UNORM2_COMPOSE, errorCode).
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 49
+     */
+    static const Normalizer2 *
+    getNFKCCasefoldInstance(UErrorCode &errorCode);
+
+    /**
+     * Returns a Normalizer2 instance which uses the specified data file
+     * (packageName/name similar to ucnv_openPackage() and ures_open()/ResourceBundle)
+     * and which composes or decomposes text according to the specified mode.
+     * Returns an unmodifiable singleton instance. Do not delete it.
+     *
+     * Use packageName=NULL for data files that are part of ICU's own data.
+     * Use name="nfc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFC/NFD.
+     * Use name="nfkc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFKC/NFKD.
+     * Use name="nfkc_cf" and UNORM2_COMPOSE for Unicode standard NFKC_CF=NFKC_Casefold.
+     *
+     * @param packageName NULL for ICU built-in data, otherwise application data package name
+     * @param name "nfc" or "nfkc" or "nfkc_cf" or name of custom data file
+     * @param mode normalization mode (compose or decompose etc.)
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return the requested Normalizer2, if successful
+     * @stable ICU 4.4
+     */
+    static const Normalizer2 *
+    getInstance(const char *packageName,
+                const char *name,
+                UNormalization2Mode mode,
+                UErrorCode &errorCode);
+
+    /**
+     * Returns the normalized form of the source string.
+     * @param src source string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return normalized src
+     * @stable ICU 4.4
+     */
+    UnicodeString
+    normalize(const UnicodeString &src, UErrorCode &errorCode) const {
+        UnicodeString result;
+        normalize(src, result, errorCode);
+        return result;
+    }
+    /**
+     * Writes the normalized form of the source string to the destination string
+     * (replacing its contents) and returns the destination string.
+     * The source and destination strings must be different objects.
+     * @param src source string
+     * @param dest destination string; its contents is replaced with normalized src
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    normalize(const UnicodeString &src,
+              UnicodeString &dest,
+              UErrorCode &errorCode) const = 0;
+
+    /**
+     * Normalizes a UTF-8 string and optionally records how source substrings
+     * relate to changed and unchanged result substrings.
+     *
+     * Currently implemented completely only for "compose" modes,
+     * such as for NFC, NFKC, and NFKC_Casefold
+     * (UNORM2_COMPOSE and UNORM2_COMPOSE_CONTIGUOUS).
+     * Otherwise currently converts to & from UTF-16 and does not support edits.
+     *
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       Source UTF-8 string.
+     * @param sink      A ByteSink to which the normalized UTF-8 result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be nullptr.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @stable ICU 60
+     */
+    virtual void
+    normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink,
+                  Edits *edits, UErrorCode &errorCode) const;
+
+    /**
+     * Appends the normalized form of the second string to the first string
+     * (merging them at the boundary) and returns the first string.
+     * The result is normalized if the first string was normalized.
+     * The first and second strings must be different objects.
+     * @param first string, should be normalized
+     * @param second string, will be normalized
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return first
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UErrorCode &errorCode) const = 0;
+    /**
+     * Appends the second string to the first string
+     * (merging them at the boundary) and returns the first string.
+     * The result is normalized if both the strings were normalized.
+     * The first and second strings must be different objects.
+     * @param first string, should be normalized
+     * @param second string, should be normalized
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return first
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    append(UnicodeString &first,
+           const UnicodeString &second,
+           UErrorCode &errorCode) const = 0;
+
+    /**
+     * Gets the decomposition mapping of c.
+     * Roughly equivalent to normalizing the String form of c
+     * on a UNORM2_DECOMPOSE Normalizer2 instance, but much faster, and except that this function
+     * returns false and does not write a string
+     * if c does not have a decomposition mapping in this instance's data.
+     * This function is independent of the mode of the Normalizer2.
+     * @param c code point
+     * @param decomposition String object which will be set to c's
+     *                      decomposition mapping, if there is one.
+     * @return true if c has a decomposition, otherwise false
+     * @stable ICU 4.6
+     */
+    virtual UBool
+    getDecomposition(UChar32 c, UnicodeString &decomposition) const = 0;
+
+    /**
+     * Gets the raw decomposition mapping of c.
+     *
+     * This is similar to the getDecomposition() method but returns the
+     * raw decomposition mapping as specified in UnicodeData.txt or
+     * (for custom data) in the mapping files processed by the gennorm2 tool.
+     * By contrast, getDecomposition() returns the processed,
+     * recursively-decomposed version of this mapping.
+     *
+     * When used on a standard NFKC Normalizer2 instance,
+     * getRawDecomposition() returns the Unicode Decomposition_Mapping (dm) property.
+     *
+     * When used on a standard NFC Normalizer2 instance,
+     * it returns the Decomposition_Mapping only if the Decomposition_Type (dt) is Canonical (Can);
+     * in this case, the result contains either one or two code points (=1..4 char16_ts).
+     *
+     * This function is independent of the mode of the Normalizer2.
+     * The default implementation returns false.
+     * @param c code point
+     * @param decomposition String object which will be set to c's
+     *                      raw decomposition mapping, if there is one.
+     * @return true if c has a decomposition, otherwise false
+     * @stable ICU 49
+     */
+    virtual UBool
+    getRawDecomposition(UChar32 c, UnicodeString &decomposition) const;
+
+    /**
+     * Performs pairwise composition of a & b and returns the composite if there is one.
+     *
+     * Returns a composite code point c only if c has a two-way mapping to a+b.
+     * In standard Unicode normalization, this means that
+     * c has a canonical decomposition to a+b
+     * and c does not have the Full_Composition_Exclusion property.
+     *
+     * This function is independent of the mode of the Normalizer2.
+     * The default implementation returns a negative value.
+     * @param a A (normalization starter) code point.
+     * @param b Another code point.
+     * @return The non-negative composite code point if there is one; otherwise a negative value.
+     * @stable ICU 49
+     */
+    virtual UChar32
+    composePair(UChar32 a, UChar32 b) const;
+
+    /**
+     * Gets the combining class of c.
+     * The default implementation returns 0
+     * but all standard implementations return the Unicode Canonical_Combining_Class value.
+     * @param c code point
+     * @return c's combining class
+     * @stable ICU 49
+     */
+    virtual uint8_t
+    getCombiningClass(UChar32 c) const;
+
+    /**
+     * Tests if the string is normalized.
+     * Internally, in cases where the quickCheck() method would return "maybe"
+     * (which is only possible for the two COMPOSE modes) this method
+     * resolves to "yes" or "no" to provide a definitive result,
+     * at the cost of doing more work in those cases.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return true if s is normalized
+     * @stable ICU 4.4
+     */
+    virtual UBool
+    isNormalized(const UnicodeString &s, UErrorCode &errorCode) const = 0;
+    /**
+     * Tests if the UTF-8 string is normalized.
+     * Internally, in cases where the quickCheck() method would return "maybe"
+     * (which is only possible for the two COMPOSE modes) this method
+     * resolves to "yes" or "no" to provide a definitive result,
+     * at the cost of doing more work in those cases.
+     *
+     * This works for all normalization modes,
+     * but it is currently optimized for UTF-8 only for "compose" modes,
+     * such as for NFC, NFKC, and NFKC_Casefold
+     * (UNORM2_COMPOSE and UNORM2_COMPOSE_CONTIGUOUS).
+     * For other modes it currently converts to UTF-16 and calls isNormalized().
+     *
+     * @param s UTF-8 input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return true if s is normalized
+     * @stable ICU 60
+     */
+    virtual UBool
+    isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const;
+
+
+    /**
+     * Tests if the string is normalized.
+     * For the two COMPOSE modes, the result could be "maybe" in cases that
+     * would take a little more work to resolve definitively.
+     * Use spanQuickCheckYes() and normalizeSecondAndAppend() for a faster
+     * combination of quick check + normalization, to avoid
+     * re-checking the "yes" prefix.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return UNormalizationCheckResult
+     * @stable ICU 4.4
+     */
+    virtual UNormalizationCheckResult
+    quickCheck(const UnicodeString &s, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Returns the end of the normalized substring of the input string.
+     * In other words, with <code>end=spanQuickCheckYes(s, ec);</code>
+     * the substring <code>UnicodeString(s, 0, end)</code>
+     * will pass the quick check with a "yes" result.
+     *
+     * The returned end index is usually one or more characters before the
+     * "no" or "maybe" character: The end index is at a normalization boundary.
+     * (See the class documentation for more about normalization boundaries.)
+     *
+     * When the goal is a normalized string and most input strings are expected
+     * to be normalized already, then call this method,
+     * and if it returns a prefix shorter than the input string,
+     * copy that prefix and use normalizeSecondAndAppend() for the remainder.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return "yes" span end index
+     * @stable ICU 4.4
+     */
+    virtual int32_t
+    spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const = 0;
+
+    /**
+     * Tests if the character always has a normalization boundary before it,
+     * regardless of context.
+     * If true, then the character does not normalization-interact with
+     * preceding characters.
+     * In other words, a string containing this character can be normalized
+     * by processing portions before this character and starting from this
+     * character independently.
+     * This is used for iterative normalization. See the class documentation for details.
+     * @param c character to test
+     * @return true if c has a normalization boundary before it
+     * @stable ICU 4.4
+     */
+    virtual UBool hasBoundaryBefore(UChar32 c) const = 0;
+
+    /**
+     * Tests if the character always has a normalization boundary after it,
+     * regardless of context.
+     * If true, then the character does not normalization-interact with
+     * following characters.
+     * In other words, a string containing this character can be normalized
+     * by processing portions up to this character and after this
+     * character independently.
+     * This is used for iterative normalization. See the class documentation for details.
+     * Note that this operation may be significantly slower than hasBoundaryBefore().
+     * @param c character to test
+     * @return true if c has a normalization boundary after it
+     * @stable ICU 4.4
+     */
+    virtual UBool hasBoundaryAfter(UChar32 c) const = 0;
+
+    /**
+     * Tests if the character is normalization-inert.
+     * If true, then the character does not change, nor normalization-interact with
+     * preceding or following characters.
+     * In other words, a string containing this character can be normalized
+     * by processing portions before this character and after this
+     * character independently.
+     * This is used for iterative normalization. See the class documentation for details.
+     * Note that this operation may be significantly slower than hasBoundaryBefore().
+     * @param c character to test
+     * @return true if c is normalization-inert
+     * @stable ICU 4.4
+     */
+    virtual UBool isInert(UChar32 c) const = 0;
+};
+
+/**
+ * Normalization filtered by a UnicodeSet.
+ * Normalizes portions of the text contained in the filter set and leaves
+ * portions not contained in the filter set unchanged.
+ * Filtering is done via UnicodeSet::span(..., USET_SPAN_SIMPLE).
+ * Not-in-the-filter text is treated as "is normalized" and "quick check yes".
+ * This class implements all of (and only) the Normalizer2 API.
+ * An instance of this class is unmodifiable/immutable but is constructed and
+ * must be destructed by the owner.
+ * @stable ICU 4.4
+ */
+class U_COMMON_API FilteredNormalizer2 : public Normalizer2 {
+public:
+    /**
+     * Constructs a filtered normalizer wrapping any Normalizer2 instance
+     * and a filter set.
+     * Both are aliased and must not be modified or deleted while this object
+     * is used.
+     * The filter set should be frozen; otherwise the performance will suffer greatly.
+     * @param n2 wrapped Normalizer2 instance
+     * @param filterSet UnicodeSet which determines the characters to be normalized
+     * @stable ICU 4.4
+     */
+    FilteredNormalizer2(const Normalizer2 &n2, const UnicodeSet &filterSet) :
+            norm2(n2), set(filterSet) {}
+
+    /**
+     * Destructor.
+     * @stable ICU 4.4
+     */
+    ~FilteredNormalizer2();
+
+    /**
+     * Writes the normalized form of the source string to the destination string
+     * (replacing its contents) and returns the destination string.
+     * The source and destination strings must be different objects.
+     * @param src source string
+     * @param dest destination string; its contents is replaced with normalized src
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return dest
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    normalize(const UnicodeString &src,
+              UnicodeString &dest,
+              UErrorCode &errorCode) const U_OVERRIDE;
+
+    /**
+     * Normalizes a UTF-8 string and optionally records how source substrings
+     * relate to changed and unchanged result substrings.
+     *
+     * Currently implemented completely only for "compose" modes,
+     * such as for NFC, NFKC, and NFKC_Casefold
+     * (UNORM2_COMPOSE and UNORM2_COMPOSE_CONTIGUOUS).
+     * Otherwise currently converts to & from UTF-16 and does not support edits.
+     *
+     * @param options   Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET.
+     * @param src       Source UTF-8 string.
+     * @param sink      A ByteSink to which the normalized UTF-8 result string is written.
+     *                  sink.Flush() is called at the end.
+     * @param edits     Records edits for index mapping, working with styled text,
+     *                  and getting only changes (if any).
+     *                  The Edits contents is undefined if any error occurs.
+     *                  This function calls edits->reset() first unless
+     *                  options includes U_EDITS_NO_RESET. edits can be nullptr.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @stable ICU 60
+     */
+    virtual void
+    normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink,
+                  Edits *edits, UErrorCode &errorCode) const U_OVERRIDE;
+
+    /**
+     * Appends the normalized form of the second string to the first string
+     * (merging them at the boundary) and returns the first string.
+     * The result is normalized if the first string was normalized.
+     * The first and second strings must be different objects.
+     * @param first string, should be normalized
+     * @param second string, will be normalized
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return first
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UErrorCode &errorCode) const U_OVERRIDE;
+    /**
+     * Appends the second string to the first string
+     * (merging them at the boundary) and returns the first string.
+     * The result is normalized if both the strings were normalized.
+     * The first and second strings must be different objects.
+     * @param first string, should be normalized
+     * @param second string, should be normalized
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return first
+     * @stable ICU 4.4
+     */
+    virtual UnicodeString &
+    append(UnicodeString &first,
+           const UnicodeString &second,
+           UErrorCode &errorCode) const U_OVERRIDE;
+
+    /**
+     * Gets the decomposition mapping of c.
+     * For details see the base class documentation.
+     *
+     * This function is independent of the mode of the Normalizer2.
+     * @param c code point
+     * @param decomposition String object which will be set to c's
+     *                      decomposition mapping, if there is one.
+     * @return true if c has a decomposition, otherwise false
+     * @stable ICU 4.6
+     */
+    virtual UBool
+    getDecomposition(UChar32 c, UnicodeString &decomposition) const U_OVERRIDE;
+
+    /**
+     * Gets the raw decomposition mapping of c.
+     * For details see the base class documentation.
+     *
+     * This function is independent of the mode of the Normalizer2.
+     * @param c code point
+     * @param decomposition String object which will be set to c's
+     *                      raw decomposition mapping, if there is one.
+     * @return true if c has a decomposition, otherwise false
+     * @stable ICU 49
+     */
+    virtual UBool
+    getRawDecomposition(UChar32 c, UnicodeString &decomposition) const U_OVERRIDE;
+
+    /**
+     * Performs pairwise composition of a & b and returns the composite if there is one.
+     * For details see the base class documentation.
+     *
+     * This function is independent of the mode of the Normalizer2.
+     * @param a A (normalization starter) code point.
+     * @param b Another code point.
+     * @return The non-negative composite code point if there is one; otherwise a negative value.
+     * @stable ICU 49
+     */
+    virtual UChar32
+    composePair(UChar32 a, UChar32 b) const U_OVERRIDE;
+
+    /**
+     * Gets the combining class of c.
+     * The default implementation returns 0
+     * but all standard implementations return the Unicode Canonical_Combining_Class value.
+     * @param c code point
+     * @return c's combining class
+     * @stable ICU 49
+     */
+    virtual uint8_t
+    getCombiningClass(UChar32 c) const U_OVERRIDE;
+
+    /**
+     * Tests if the string is normalized.
+     * For details see the Normalizer2 base class documentation.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return true if s is normalized
+     * @stable ICU 4.4
+     */
+    virtual UBool
+    isNormalized(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE;
+    /**
+     * Tests if the UTF-8 string is normalized.
+     * Internally, in cases where the quickCheck() method would return "maybe"
+     * (which is only possible for the two COMPOSE modes) this method
+     * resolves to "yes" or "no" to provide a definitive result,
+     * at the cost of doing more work in those cases.
+     *
+     * This works for all normalization modes,
+     * but it is currently optimized for UTF-8 only for "compose" modes,
+     * such as for NFC, NFKC, and NFKC_Casefold
+     * (UNORM2_COMPOSE and UNORM2_COMPOSE_CONTIGUOUS).
+     * For other modes it currently converts to UTF-16 and calls isNormalized().
+     *
+     * @param s UTF-8 input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return true if s is normalized
+     * @stable ICU 60
+     */
+    virtual UBool
+    isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const U_OVERRIDE;
+    /**
+     * Tests if the string is normalized.
+     * For details see the Normalizer2 base class documentation.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return UNormalizationCheckResult
+     * @stable ICU 4.4
+     */
+    virtual UNormalizationCheckResult
+    quickCheck(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE;
+    /**
+     * Returns the end of the normalized substring of the input string.
+     * For details see the Normalizer2 base class documentation.
+     * @param s input string
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return "yes" span end index
+     * @stable ICU 4.4
+     */
+    virtual int32_t
+    spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE;
+
+    /**
+     * Tests if the character always has a normalization boundary before it,
+     * regardless of context.
+     * For details see the Normalizer2 base class documentation.
+     * @param c character to test
+     * @return true if c has a normalization boundary before it
+     * @stable ICU 4.4
+     */
+    virtual UBool hasBoundaryBefore(UChar32 c) const U_OVERRIDE;
+
+    /**
+     * Tests if the character always has a normalization boundary after it,
+     * regardless of context.
+     * For details see the Normalizer2 base class documentation.
+     * @param c character to test
+     * @return true if c has a normalization boundary after it
+     * @stable ICU 4.4
+     */
+    virtual UBool hasBoundaryAfter(UChar32 c) const U_OVERRIDE;
+
+    /**
+     * Tests if the character is normalization-inert.
+     * For details see the Normalizer2 base class documentation.
+     * @param c character to test
+     * @return true if c is normalization-inert
+     * @stable ICU 4.4
+     */
+    virtual UBool isInert(UChar32 c) const U_OVERRIDE;
+private:
+    UnicodeString &
+    normalize(const UnicodeString &src,
+              UnicodeString &dest,
+              USetSpanCondition spanCondition,
+              UErrorCode &errorCode) const;
+
+    void
+    normalizeUTF8(uint32_t options, const char *src, int32_t length,
+                  ByteSink &sink, Edits *edits,
+                  USetSpanCondition spanCondition,
+                  UErrorCode &errorCode) const;
+
+    UnicodeString &
+    normalizeSecondAndAppend(UnicodeString &first,
+                             const UnicodeString &second,
+                             UBool doNormalize,
+                             UErrorCode &errorCode) const;
+
+    const Normalizer2 &norm2;
+    const UnicodeSet &set;
+};
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_NORMALIZATION
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __NORMALIZER2_H__
diff --git a/thirdparty/icu4c/common/unicode/normlzr.h b/thirdparty/icu4c/common/unicode/normlzr.h
new file mode 100644
index 0000000000..3352983cdc
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/normlzr.h
@@ -0,0 +1,816 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ ********************************************************************
+ * COPYRIGHT:
+ * Copyright (c) 1996-2015, International Business Machines Corporation and
+ * others. All Rights Reserved.
+ ********************************************************************
+ */
+
+#ifndef NORMLZR_H
+#define NORMLZR_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file 
+ * \brief C++ API: Unicode Normalization
+ */
+ 
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/chariter.h"
+#include "unicode/normalizer2.h"
+#include "unicode/unistr.h"
+#include "unicode/unorm.h"
+#include "unicode/uobject.h"
+
+U_NAMESPACE_BEGIN
+/**
+ * Old Unicode normalization API.
+ *
+ * This API has been replaced by the Normalizer2 class and is only available
+ * for backward compatibility. This class simply delegates to the Normalizer2 class.
+ * There is one exception: The new API does not provide a replacement for Normalizer::compare().
+ *
+ * The Normalizer class supports the standard normalization forms described in
+ * <a href="http://www.unicode.org/unicode/reports/tr15/" target="unicode">
+ * Unicode Standard Annex #15: Unicode Normalization Forms</a>.
+ *
+ * The Normalizer class consists of two parts:
+ * - static functions that normalize strings or test if strings are normalized
+ * - a Normalizer object is an iterator that takes any kind of text and
+ *   provides iteration over its normalized form
+ *
+ * The Normalizer class is not suitable for subclassing.
+ *
+ * For basic information about normalization forms and details about the C API
+ * please see the documentation in unorm.h.
+ *
+ * The iterator API with the Normalizer constructors and the non-static functions
+ * use a CharacterIterator as input. It is possible to pass a string which
+ * is then internally wrapped in a CharacterIterator.
+ * The input text is not normalized all at once, but incrementally where needed
+ * (providing efficient random access).
+ * This allows to pass in a large text but spend only a small amount of time
+ * normalizing a small part of that text.
+ * However, if the entire text is normalized, then the iterator will be
+ * slower than normalizing the entire text at once and iterating over the result.
+ * A possible use of the Normalizer iterator is also to report an index into the
+ * original text that is close to where the normalized characters come from.
+ *
+ * <em>Important:</em> The iterator API was cleaned up significantly for ICU 2.0.
+ * The earlier implementation reported the getIndex() inconsistently,
+ * and previous() could not be used after setIndex(), next(), first(), and current().
+ *
+ * Normalizer allows to start normalizing from anywhere in the input text by
+ * calling setIndexOnly(), first(), or last().
+ * Without calling any of these, the iterator will start at the beginning of the text.
+ *
+ * At any time, next() returns the next normalized code point (UChar32),
+ * with post-increment semantics (like CharacterIterator::next32PostInc()).
+ * previous() returns the previous normalized code point (UChar32),
+ * with pre-decrement semantics (like CharacterIterator::previous32()).
+ *
+ * current() returns the current code point
+ * (respectively the one at the newly set index) without moving
+ * the getIndex(). Note that if the text at the current position
+ * needs to be normalized, then these functions will do that.
+ * (This is why current() is not const.)
+ * It is more efficient to call setIndexOnly() instead, which does not
+ * normalize.
+ *
+ * getIndex() always refers to the position in the input text where the normalized
+ * code points are returned from. It does not always change with each returned
+ * code point.
+ * The code point that is returned from any of the functions
+ * corresponds to text at or after getIndex(), according to the
+ * function's iteration semantics (post-increment or pre-decrement).
+ *
+ * next() returns a code point from at or after the getIndex()
+ * from before the next() call. After the next() call, the getIndex()
+ * might have moved to where the next code point will be returned from
+ * (from a next() or current() call).
+ * This is semantically equivalent to array access with array[index++]
+ * (post-increment semantics).
+ *
+ * previous() returns a code point from at or after the getIndex()
+ * from after the previous() call.
+ * This is semantically equivalent to array access with array[--index]
+ * (pre-decrement semantics).
+ *
+ * Internally, the Normalizer iterator normalizes a small piece of text
+ * starting at the getIndex() and ending at a following "safe" index.
+ * The normalized results is stored in an internal string buffer, and
+ * the code points are iterated from there.
+ * With multiple iteration calls, this is repeated until the next piece
+ * of text needs to be normalized, and the getIndex() needs to be moved.
+ *
+ * The following "safe" index, the internal buffer, and the secondary
+ * iteration index into that buffer are not exposed on the API.
+ * This also means that it is currently not practical to return to
+ * a particular, arbitrary position in the text because one would need to
+ * know, and be able to set, in addition to the getIndex(), at least also the
+ * current index into the internal buffer.
+ * It is currently only possible to observe when getIndex() changes
+ * (with careful consideration of the iteration semantics),
+ * at which time the internal index will be 0.
+ * For example, if getIndex() is different after next() than before it,
+ * then the internal index is 0 and one can return to this getIndex()
+ * later with setIndexOnly().
+ *
+ * Note: While the setIndex() and getIndex() refer to indices in the
+ * underlying Unicode input text, the next() and previous() methods
+ * iterate through characters in the normalized output.
+ * This means that there is not necessarily a one-to-one correspondence
+ * between characters returned by next() and previous() and the indices
+ * passed to and returned from setIndex() and getIndex().
+ * It is for this reason that Normalizer does not implement the CharacterIterator interface.
+ *
+ * @author Laura Werner, Mark Davis, Markus Scherer
+ * @stable ICU 2.0
+ */
+class U_COMMON_API Normalizer : public UObject {
+public:
+#ifndef U_HIDE_DEPRECATED_API
+  /**
+   * If DONE is returned from an iteration function that returns a code point,
+   * then there are no more normalization results available.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  enum {
+      DONE=0xffff
+  };
+
+  // Constructors
+
+  /**
+   * Creates a new <code>Normalizer</code> object for iterating over the
+   * normalized form of a given string.
+   * <p>
+   * @param str   The string to be normalized.  The normalization
+   *              will start at the beginning of the string.
+   *
+   * @param mode  The normalization mode.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  Normalizer(const UnicodeString& str, UNormalizationMode mode);
+
+  /**
+   * Creates a new <code>Normalizer</code> object for iterating over the
+   * normalized form of a given string.
+   * <p>
+   * @param str   The string to be normalized.  The normalization
+   *              will start at the beginning of the string.
+   *
+   * @param length Length of the string, or -1 if NUL-terminated.
+   * @param mode  The normalization mode.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  Normalizer(ConstChar16Ptr str, int32_t length, UNormalizationMode mode);
+
+  /**
+   * Creates a new <code>Normalizer</code> object for iterating over the
+   * normalized form of the given text.
+   * <p>
+   * @param iter  The input text to be normalized.  The normalization
+   *              will start at the beginning of the string.
+   *
+   * @param mode  The normalization mode.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  Normalizer(const CharacterIterator& iter, UNormalizationMode mode);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+  /**
+   * Copy constructor.
+   * @param copy The object to be copied.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  Normalizer(const Normalizer& copy);
+
+  /**
+   * Destructor
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  virtual ~Normalizer();
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+  //-------------------------------------------------------------------------
+  // Static utility methods
+  //-------------------------------------------------------------------------
+
+#ifndef U_HIDE_DEPRECATED_API
+  /**
+   * Normalizes a <code>UnicodeString</code> according to the specified normalization mode.
+   * This is a wrapper for unorm_normalize(), using UnicodeString's.
+   *
+   * The <code>options</code> parameter specifies which optional
+   * <code>Normalizer</code> features are to be enabled for this operation.
+   *
+   * @param source    the input string to be normalized.
+   * @param mode      the normalization mode
+   * @param options   the optional features to be enabled (0 for no options)
+   * @param result    The normalized string (on output).
+   * @param status    The error code.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static void U_EXPORT2 normalize(const UnicodeString& source,
+                        UNormalizationMode mode, int32_t options,
+                        UnicodeString& result,
+                        UErrorCode &status);
+
+  /**
+   * Compose a <code>UnicodeString</code>.
+   * This is equivalent to normalize() with mode UNORM_NFC or UNORM_NFKC.
+   * This is a wrapper for unorm_normalize(), using UnicodeString's.
+   *
+   * The <code>options</code> parameter specifies which optional
+   * <code>Normalizer</code> features are to be enabled for this operation.
+   *
+   * @param source    the string to be composed.
+   * @param compat    Perform compatibility decomposition before composition.
+   *                  If this argument is <code>false</code>, only canonical
+   *                  decomposition will be performed.
+   * @param options   the optional features to be enabled (0 for no options)
+   * @param result    The composed string (on output).
+   * @param status    The error code.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static void U_EXPORT2 compose(const UnicodeString& source,
+                      UBool compat, int32_t options,
+                      UnicodeString& result,
+                      UErrorCode &status);
+
+  /**
+   * Static method to decompose a <code>UnicodeString</code>.
+   * This is equivalent to normalize() with mode UNORM_NFD or UNORM_NFKD.
+   * This is a wrapper for unorm_normalize(), using UnicodeString's.
+   *
+   * The <code>options</code> parameter specifies which optional
+   * <code>Normalizer</code> features are to be enabled for this operation.
+   *
+   * @param source    the string to be decomposed.
+   * @param compat    Perform compatibility decomposition.
+   *                  If this argument is <code>false</code>, only canonical
+   *                  decomposition will be performed.
+   * @param options   the optional features to be enabled (0 for no options)
+   * @param result    The decomposed string (on output).
+   * @param status    The error code.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static void U_EXPORT2 decompose(const UnicodeString& source,
+                        UBool compat, int32_t options,
+                        UnicodeString& result,
+                        UErrorCode &status);
+
+  /**
+   * Performing quick check on a string, to quickly determine if the string is
+   * in a particular normalization format.
+   * This is a wrapper for unorm_quickCheck(), using a UnicodeString.
+   *
+   * Three types of result can be returned UNORM_YES, UNORM_NO or
+   * UNORM_MAYBE. Result UNORM_YES indicates that the argument
+   * string is in the desired normalized format, UNORM_NO determines that
+   * argument string is not in the desired normalized format. A
+   * UNORM_MAYBE result indicates that a more thorough check is required,
+   * the user may have to put the string in its normalized form and compare the
+   * results.
+   * @param source       string for determining if it is in a normalized format
+   * @param mode         normalization format
+   * @param status A reference to a UErrorCode to receive any errors
+   * @return UNORM_YES, UNORM_NO or UNORM_MAYBE
+   *
+   * @see isNormalized
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static inline UNormalizationCheckResult
+  quickCheck(const UnicodeString &source, UNormalizationMode mode, UErrorCode &status);
+
+  /**
+   * Performing quick check on a string; same as the other version of quickCheck
+   * but takes an extra options parameter like most normalization functions.
+   *
+   * @param source       string for determining if it is in a normalized format
+   * @param mode         normalization format
+   * @param options      the optional features to be enabled (0 for no options)
+   * @param status A reference to a UErrorCode to receive any errors
+   * @return UNORM_YES, UNORM_NO or UNORM_MAYBE
+   *
+   * @see isNormalized
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static UNormalizationCheckResult
+  quickCheck(const UnicodeString &source, UNormalizationMode mode, int32_t options, UErrorCode &status);
+
+  /**
+   * Test if a string is in a given normalization form.
+   * This is semantically equivalent to source.equals(normalize(source, mode)) .
+   *
+   * Unlike unorm_quickCheck(), this function returns a definitive result,
+   * never a "maybe".
+   * For NFD, NFKD, and FCD, both functions work exactly the same.
+   * For NFC and NFKC where quickCheck may return "maybe", this function will
+   * perform further tests to arrive at a true/false result.
+   *
+   * @param src        String that is to be tested if it is in a normalization format.
+   * @param mode       Which normalization form to test for.
+   * @param errorCode  ICU error code in/out parameter.
+   *                   Must fulfill U_SUCCESS before the function call.
+   * @return Boolean value indicating whether the source string is in the
+   *         "mode" normalization form.
+   *
+   * @see quickCheck
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static inline UBool
+  isNormalized(const UnicodeString &src, UNormalizationMode mode, UErrorCode &errorCode);
+
+  /**
+   * Test if a string is in a given normalization form; same as the other version of isNormalized
+   * but takes an extra options parameter like most normalization functions.
+   *
+   * @param src        String that is to be tested if it is in a normalization format.
+   * @param mode       Which normalization form to test for.
+   * @param options      the optional features to be enabled (0 for no options)
+   * @param errorCode  ICU error code in/out parameter.
+   *                   Must fulfill U_SUCCESS before the function call.
+   * @return Boolean value indicating whether the source string is in the
+   *         "mode" normalization form.
+   *
+   * @see quickCheck
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static UBool
+  isNormalized(const UnicodeString &src, UNormalizationMode mode, int32_t options, UErrorCode &errorCode);
+
+  /**
+   * Concatenate normalized strings, making sure that the result is normalized as well.
+   *
+   * If both the left and the right strings are in
+   * the normalization form according to "mode/options",
+   * then the result will be
+   *
+   * \code
+   *     dest=normalize(left+right, mode, options)
+   * \endcode
+   *
+   * For details see unorm_concatenate in unorm.h.
+   *
+   * @param left Left source string.
+   * @param right Right source string.
+   * @param result The output string.
+   * @param mode The normalization mode.
+   * @param options A bit set of normalization options.
+   * @param errorCode ICU error code in/out parameter.
+   *                   Must fulfill U_SUCCESS before the function call.
+   * @return result
+   *
+   * @see unorm_concatenate
+   * @see normalize
+   * @see unorm_next
+   * @see unorm_previous
+   *
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static UnicodeString &
+  U_EXPORT2 concatenate(const UnicodeString &left, const UnicodeString &right,
+              UnicodeString &result,
+              UNormalizationMode mode, int32_t options,
+              UErrorCode &errorCode);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+  /**
+   * Compare two strings for canonical equivalence.
+   * Further options include case-insensitive comparison and
+   * code point order (as opposed to code unit order).
+   *
+   * Canonical equivalence between two strings is defined as their normalized
+   * forms (NFD or NFC) being identical.
+   * This function compares strings incrementally instead of normalizing
+   * (and optionally case-folding) both strings entirely,
+   * improving performance significantly.
+   *
+   * Bulk normalization is only necessary if the strings do not fulfill the FCD
+   * conditions. Only in this case, and only if the strings are relatively long,
+   * is memory allocated temporarily.
+   * For FCD strings and short non-FCD strings there is no memory allocation.
+   *
+   * Semantically, this is equivalent to
+   *   strcmp[CodePointOrder](NFD(foldCase(s1)), NFD(foldCase(s2)))
+   * where code point order and foldCase are all optional.
+   *
+   * UAX 21 2.5 Caseless Matching specifies that for a canonical caseless match
+   * the case folding must be performed first, then the normalization.
+   *
+   * @param s1 First source string.
+   * @param s2 Second source string.
+   *
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Case-sensitive comparison in code unit order, and the input strings
+   *     are quick-checked for FCD.
+   *
+   *   - UNORM_INPUT_IS_FCD
+   *     Set if the caller knows that both s1 and s2 fulfill the FCD conditions.
+   *     If not set, the function will quickCheck for FCD
+   *     and normalize if necessary.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_COMPARE_IGNORE_CASE
+   *     Set to compare strings case-insensitively using case folding,
+   *     instead of case-sensitively.
+   *     If set, then the following case folding options are used.
+   *
+   *   - Options as used with case-insensitive comparisons, currently:
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *    (see u_strCaseCompare for details)
+   *
+   *   - regular normalization options shifted left by UNORM_COMPARE_NORM_OPTIONS_SHIFT
+   *
+   * @param errorCode ICU error code in/out parameter.
+   *                  Must fulfill U_SUCCESS before the function call.
+   * @return <0 or 0 or >0 as usual for string comparisons
+   *
+   * @see unorm_compare
+   * @see normalize
+   * @see UNORM_FCD
+   * @see u_strCompare
+   * @see u_strCaseCompare
+   *
+   * @stable ICU 2.2
+   */
+  static inline int32_t
+  compare(const UnicodeString &s1, const UnicodeString &s2,
+          uint32_t options,
+          UErrorCode &errorCode);
+
+#ifndef U_HIDE_DEPRECATED_API
+  //-------------------------------------------------------------------------
+  // Iteration API
+  //-------------------------------------------------------------------------
+
+  /**
+   * Return the current character in the normalized text.
+   * current() may need to normalize some text at getIndex().
+   * The getIndex() is not changed.
+   *
+   * @return the current normalized code point
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UChar32              current(void);
+
+  /**
+   * Return the first character in the normalized text.
+   * This is equivalent to setIndexOnly(startIndex()) followed by next().
+   * (Post-increment semantics.)
+   *
+   * @return the first normalized code point
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UChar32              first(void);
+
+  /**
+   * Return the last character in the normalized text.
+   * This is equivalent to setIndexOnly(endIndex()) followed by previous().
+   * (Pre-decrement semantics.)
+   *
+   * @return the last normalized code point
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UChar32              last(void);
+
+  /**
+   * Return the next character in the normalized text.
+   * (Post-increment semantics.)
+   * If the end of the text has already been reached, DONE is returned.
+   * The DONE value could be confused with a U+FFFF non-character code point
+   * in the text. If this is possible, you can test getIndex()<endIndex()
+   * before calling next(), or (getIndex()<endIndex() || last()!=DONE)
+   * after calling next(). (Calling last() will change the iterator state!)
+   *
+   * The C API unorm_next() is more efficient and does not have this ambiguity.
+   *
+   * @return the next normalized code point
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UChar32              next(void);
+
+  /**
+   * Return the previous character in the normalized text and decrement.
+   * (Pre-decrement semantics.)
+   * If the beginning of the text has already been reached, DONE is returned.
+   * The DONE value could be confused with a U+FFFF non-character code point
+   * in the text. If this is possible, you can test
+   * (getIndex()>startIndex() || first()!=DONE). (Calling first() will change
+   * the iterator state!)
+   *
+   * The C API unorm_previous() is more efficient and does not have this ambiguity.
+   *
+   * @return the previous normalized code point
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UChar32              previous(void);
+
+  /**
+   * Set the iteration position in the input text that is being normalized,
+   * without any immediate normalization.
+   * After setIndexOnly(), getIndex() will return the same index that is
+   * specified here.
+   *
+   * @param index the desired index in the input text.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void                 setIndexOnly(int32_t index);
+
+  /**
+   * Reset the index to the beginning of the text.
+   * This is equivalent to setIndexOnly(startIndex)).
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void                reset(void);
+
+  /**
+   * Retrieve the current iteration position in the input text that is
+   * being normalized.
+   *
+   * A following call to next() will return a normalized code point from
+   * the input text at or after this index.
+   *
+   * After a call to previous(), getIndex() will point at or before the
+   * position in the input text where the normalized code point
+   * was returned from with previous().
+   *
+   * @return the current index in the input text
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  int32_t            getIndex(void) const;
+
+  /**
+   * Retrieve the index of the start of the input text. This is the begin index
+   * of the <code>CharacterIterator</code> or the start (i.e. index 0) of the string
+   * over which this <code>Normalizer</code> is iterating.
+   *
+   * @return the smallest index in the input text where the Normalizer operates
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  int32_t            startIndex(void) const;
+
+  /**
+   * Retrieve the index of the end of the input text. This is the end index
+   * of the <code>CharacterIterator</code> or the length of the string
+   * over which this <code>Normalizer</code> is iterating.
+   * This end index is exclusive, i.e., the Normalizer operates only on characters
+   * before this index.
+   *
+   * @return the first index in the input text where the Normalizer does not operate
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  int32_t            endIndex(void) const;
+
+  /**
+   * Returns true when both iterators refer to the same character in the same
+   * input text.
+   *
+   * @param that a Normalizer object to compare this one to
+   * @return comparison result
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UBool        operator==(const Normalizer& that) const;
+
+  /**
+   * Returns false when both iterators refer to the same character in the same
+   * input text.
+   *
+   * @param that a Normalizer object to compare this one to
+   * @return comparison result
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  inline UBool        operator!=(const Normalizer& that) const;
+
+  /**
+   * Returns a pointer to a new Normalizer that is a clone of this one.
+   * The caller is responsible for deleting the new clone.
+   * @return a pointer to a new Normalizer
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  Normalizer*        clone() const;
+
+  /**
+   * Generates a hash code for this iterator.
+   *
+   * @return the hash code
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  int32_t                hashCode(void) const;
+
+  //-------------------------------------------------------------------------
+  // Property access methods
+  //-------------------------------------------------------------------------
+
+  /**
+   * Set the normalization mode for this object.
+   * <p>
+   * <b>Note:</b>If the normalization mode is changed while iterating
+   * over a string, calls to {@link #next() } and {@link #previous() } may
+   * return previously buffers characters in the old normalization mode
+   * until the iteration is able to re-sync at the next base character.
+   * It is safest to call {@link #setIndexOnly }, {@link #reset() },
+   * {@link #setText }, {@link #first() },
+   * {@link #last() }, etc. after calling <code>setMode</code>.
+   * <p>
+   * @param newMode the new mode for this <code>Normalizer</code>.
+   * @see #getUMode
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void setMode(UNormalizationMode newMode);
+
+  /**
+   * Return the normalization mode for this object.
+   *
+   * This is an unusual name because there used to be a getMode() that
+   * returned a different type.
+   *
+   * @return the mode for this <code>Normalizer</code>
+   * @see #setMode
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UNormalizationMode getUMode(void) const;
+
+  /**
+   * Set options that affect this <code>Normalizer</code>'s operation.
+   * Options do not change the basic composition or decomposition operation
+   * that is being performed, but they control whether
+   * certain optional portions of the operation are done.
+   * Currently the only available option is obsolete.
+   *
+   * It is possible to specify multiple options that are all turned on or off.
+   *
+   * @param   option  the option(s) whose value is/are to be set.
+   * @param   value   the new setting for the option.  Use <code>true</code> to
+   *                  turn the option(s) on and <code>false</code> to turn it/them off.
+   *
+   * @see #getOption
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void setOption(int32_t option,
+         UBool value);
+
+  /**
+   * Determine whether an option is turned on or off.
+   * If multiple options are specified, then the result is true if any
+   * of them are set.
+   * <p>
+   * @param option the option(s) that are to be checked
+   * @return true if any of the option(s) are set
+   * @see #setOption
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  UBool getOption(int32_t option) const;
+
+  /**
+   * Set the input text over which this <code>Normalizer</code> will iterate.
+   * The iteration position is set to the beginning.
+   *
+   * @param newText a string that replaces the current input text
+   * @param status a UErrorCode
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void setText(const UnicodeString& newText,
+           UErrorCode &status);
+
+  /**
+   * Set the input text over which this <code>Normalizer</code> will iterate.
+   * The iteration position is set to the beginning.
+   *
+   * @param newText a CharacterIterator object that replaces the current input text
+   * @param status a UErrorCode
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void setText(const CharacterIterator& newText,
+           UErrorCode &status);
+
+  /**
+   * Set the input text over which this <code>Normalizer</code> will iterate.
+   * The iteration position is set to the beginning.
+   *
+   * @param newText a string that replaces the current input text
+   * @param length the length of the string, or -1 if NUL-terminated
+   * @param status a UErrorCode
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void setText(ConstChar16Ptr newText,
+                    int32_t length,
+            UErrorCode &status);
+  /**
+   * Copies the input text into the UnicodeString argument.
+   *
+   * @param result Receives a copy of the text under iteration.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  void            getText(UnicodeString&  result);
+
+  /**
+   * ICU "poor man's RTTI", returns a UClassID for this class.
+   * @returns a UClassID for this class.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  static UClassID U_EXPORT2 getStaticClassID();
+#endif  /* U_HIDE_DEPRECATED_API */
+
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+  /**
+   * ICU "poor man's RTTI", returns a UClassID for the actual class.
+   * @return a UClassID for the actual class.
+   * @deprecated ICU 56 Use Normalizer2 instead.
+   */
+  virtual UClassID getDynamicClassID() const;
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+private:
+  //-------------------------------------------------------------------------
+  // Private functions
+  //-------------------------------------------------------------------------
+
+  Normalizer(); // default constructor not implemented
+  Normalizer &operator=(const Normalizer &that); // assignment operator not implemented
+
+  // Private utility methods for iteration
+  // For documentation, see the source code
+  UBool nextNormalize();
+  UBool previousNormalize();
+
+  void    init();
+  void    clearBuffer(void);
+
+  //-------------------------------------------------------------------------
+  // Private data
+  //-------------------------------------------------------------------------
+
+  FilteredNormalizer2*fFilteredNorm2;  // owned if not NULL
+  const Normalizer2  *fNorm2;  // not owned; may be equal to fFilteredNorm2
+  UNormalizationMode  fUMode;  // deprecated
+  int32_t             fOptions;
+
+  // The input text and our position in it
+  CharacterIterator  *text;
+
+  // The normalization buffer is the result of normalization
+  // of the source in [currentIndex..nextIndex[ .
+  int32_t         currentIndex, nextIndex;
+
+  // A buffer for holding intermediate results
+  UnicodeString       buffer;
+  int32_t         bufferPos;
+};
+
+//-------------------------------------------------------------------------
+// Inline implementations
+//-------------------------------------------------------------------------
+
+#ifndef U_HIDE_DEPRECATED_API
+inline UBool
+Normalizer::operator!= (const Normalizer& other) const
+{ return ! operator==(other); }
+
+inline UNormalizationCheckResult
+Normalizer::quickCheck(const UnicodeString& source,
+                       UNormalizationMode mode,
+                       UErrorCode &status) {
+    return quickCheck(source, mode, 0, status);
+}
+
+inline UBool
+Normalizer::isNormalized(const UnicodeString& source,
+                         UNormalizationMode mode,
+                         UErrorCode &status) {
+    return isNormalized(source, mode, 0, status);
+}
+#endif  /* U_HIDE_DEPRECATED_API */
+
+inline int32_t
+Normalizer::compare(const UnicodeString &s1, const UnicodeString &s2,
+                    uint32_t options,
+                    UErrorCode &errorCode) {
+  // all argument checking is done in unorm_compare
+  return unorm_compare(toUCharPtr(s1.getBuffer()), s1.length(),
+                       toUCharPtr(s2.getBuffer()), s2.length(),
+                       options,
+                       &errorCode);
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
+
+#endif // NORMLZR_H
+
+#endif /* U_SHOW_CPLUSPLUS_API */
diff --git a/thirdparty/icu4c/common/unicode/parseerr.h b/thirdparty/icu4c/common/unicode/parseerr.h
new file mode 100644
index 0000000000..c23cc273b8
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/parseerr.h
@@ -0,0 +1,94 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2005, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   03/14/00    aliu        Creation.
+*   06/27/00    aliu        Change from C++ class to C struct
+**********************************************************************
+*/
+#ifndef PARSEERR_H
+#define PARSEERR_H
+
+#include "unicode/utypes.h"
+
+
+/**
+ * \file
+ * \brief C API: Parse Error Information
+ */
+/**
+ * The capacity of the context strings in UParseError.
+ * @stable ICU 2.0
+ */ 
+enum { U_PARSE_CONTEXT_LEN = 16 };
+
+/**
+ * A UParseError struct is used to returned detailed information about
+ * parsing errors.  It is used by ICU parsing engines that parse long
+ * rules, patterns, or programs, where the text being parsed is long
+ * enough that more information than a UErrorCode is needed to
+ * localize the error.
+ *
+ * <p>The line, offset, and context fields are optional; parsing
+ * engines may choose not to use to use them.
+ *
+ * <p>The preContext and postContext strings include some part of the
+ * context surrounding the error.  If the source text is "let for=7"
+ * and "for" is the error (e.g., because it is a reserved word), then
+ * some examples of what a parser might produce are the following:
+ *
+ * <pre>
+ * preContext   postContext
+ * ""           ""            The parser does not support context
+ * "let "       "=7"          Pre- and post-context only
+ * "let "       "for=7"       Pre- and post-context and error text
+ * ""           "for"         Error text only
+ * </pre>
+ *
+ * <p>Examples of engines which use UParseError (or may use it in the
+ * future) are Transliterator, RuleBasedBreakIterator, and
+ * RegexPattern.
+ * 
+ * @stable ICU 2.0
+ */
+typedef struct UParseError {
+
+    /**
+     * The line on which the error occurred.  If the parser uses this
+     * field, it sets it to the line number of the source text line on
+     * which the error appears, which will be a value >= 1.  If the
+     * parse does not support line numbers, the value will be <= 0.
+     * @stable ICU 2.0
+     */
+    int32_t        line;
+
+    /**
+     * The character offset to the error.  If the line field is >= 1,
+     * then this is the offset from the start of the line.  Otherwise,
+     * this is the offset from the start of the text.  If the parser
+     * does not support this field, it will have a value < 0.
+     * @stable ICU 2.0
+     */
+    int32_t        offset;
+
+    /**
+     * Textual context before the error.  Null-terminated.  The empty
+     * string if not supported by parser.
+     * @stable ICU 2.0   
+     */
+    UChar          preContext[U_PARSE_CONTEXT_LEN];
+
+    /**
+     * The error itself and/or textual context after the error.
+     * Null-terminated.  The empty string if not supported by parser.
+     * @stable ICU 2.0   
+     */
+    UChar          postContext[U_PARSE_CONTEXT_LEN];
+
+} UParseError;
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/parsepos.h b/thirdparty/icu4c/common/unicode/parsepos.h
new file mode 100644
index 0000000000..260ed4cbb8
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/parsepos.h
@@ -0,0 +1,237 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+* Copyright (C) 1997-2005, International Business Machines Corporation and others. All Rights Reserved.
+*******************************************************************************
+*
+* File PARSEPOS.H
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   07/09/97    helena      Converted from java.
+*   07/17/98    stephen     Added errorIndex support.
+*   05/11/99    stephen     Cleaned up.
+*******************************************************************************
+*/
+
+#ifndef PARSEPOS_H
+#define PARSEPOS_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+ 
+U_NAMESPACE_BEGIN
+
+/**
+ * \file
+ * \brief C++ API: Canonical Iterator
+ */
+/** 
+ * <code>ParsePosition</code> is a simple class used by <code>Format</code>
+ * and its subclasses to keep track of the current position during parsing.
+ * The <code>parseObject</code> method in the various <code>Format</code>
+ * classes requires a <code>ParsePosition</code> object as an argument.
+ *
+ * <p>
+ * By design, as you parse through a string with different formats,
+ * you can use the same <code>ParsePosition</code>, since the index parameter
+ * records the current position.
+ *
+ * The ParsePosition class is not suitable for subclassing.
+ *
+ * @version     1.3 10/30/97
+ * @author      Mark Davis, Helena Shih
+ * @see         java.text.Format
+ */
+
+class U_COMMON_API ParsePosition : public UObject {
+public:
+    /**
+     * Default constructor, the index starts with 0 as default.
+     * @stable ICU 2.0
+     */
+    ParsePosition()
+        : UObject(),
+        index(0),
+        errorIndex(-1)
+      {}
+
+    /**
+     * Create a new ParsePosition with the given initial index.
+     * @param newIndex the new text offset.
+     * @stable ICU 2.0
+     */
+    ParsePosition(int32_t newIndex)
+        : UObject(),
+        index(newIndex),
+        errorIndex(-1)
+      {}
+
+    /**
+     * Copy constructor
+     * @param copy the object to be copied from.
+     * @stable ICU 2.0
+     */
+    ParsePosition(const ParsePosition& copy)
+        : UObject(copy),
+        index(copy.index),
+        errorIndex(copy.errorIndex)
+      {}
+
+    /**
+     * Destructor
+     * @stable ICU 2.0
+     */
+    virtual ~ParsePosition();
+
+    /**
+     * Assignment operator
+     * @stable ICU 2.0
+     */
+    inline ParsePosition&      operator=(const ParsePosition& copy);
+
+    /**
+     * Equality operator.
+     * @return true if the two parse positions are equal, false otherwise.
+     * @stable ICU 2.0
+     */
+    inline UBool              operator==(const ParsePosition& that) const;
+
+    /**
+     * Equality operator.
+     * @return true if the two parse positions are not equal, false otherwise.
+     * @stable ICU 2.0
+     */
+    inline UBool              operator!=(const ParsePosition& that) const;
+
+    /**
+     * Clone this object.
+     * Clones can be used concurrently in multiple threads.
+     * If an error occurs, then NULL is returned.
+     * The caller must delete the clone.
+     *
+     * @return a clone of this object
+     *
+     * @see getDynamicClassID
+     * @stable ICU 2.8
+     */
+    ParsePosition *clone() const;
+
+    /**
+     * Retrieve the current parse position.  On input to a parse method, this
+     * is the index of the character at which parsing will begin; on output, it
+     * is the index of the character following the last character parsed.
+     * @return the current index.
+     * @stable ICU 2.0
+     */
+    inline int32_t getIndex(void) const;
+
+    /**
+     * Set the current parse position.
+     * @param index the new index.
+     * @stable ICU 2.0
+     */
+    inline void setIndex(int32_t index);
+
+    /**
+     * Set the index at which a parse error occurred.  Formatters
+     * should set this before returning an error code from their
+     * parseObject method.  The default value is -1 if this is not
+     * set.
+     * @stable ICU 2.0
+     */
+    inline void setErrorIndex(int32_t ei);
+
+    /**
+     * Retrieve the index at which an error occurred, or -1 if the
+     * error index has not been set.
+     * @stable ICU 2.0
+     */
+    inline int32_t getErrorIndex(void) const;
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.2
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     *
+     * @stable ICU 2.2
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    /**
+     * Input: the place you start parsing.
+     * <br>Output: position where the parse stopped.
+     * This is designed to be used serially,
+     * with each call setting index up for the next one.
+     */
+    int32_t index;
+
+    /**
+     * The index at which a parse error occurred.
+     */
+    int32_t errorIndex;
+
+};
+
+inline ParsePosition&
+ParsePosition::operator=(const ParsePosition& copy)
+{
+  index = copy.index;
+  errorIndex = copy.errorIndex;
+  return *this;
+}
+
+inline UBool
+ParsePosition::operator==(const ParsePosition& copy) const
+{
+  if(index != copy.index || errorIndex != copy.errorIndex)
+  return false;
+  else
+  return true;
+}
+
+inline UBool
+ParsePosition::operator!=(const ParsePosition& copy) const
+{
+  return !operator==(copy);
+}
+
+inline int32_t
+ParsePosition::getIndex() const
+{
+  return index;
+}
+
+inline void
+ParsePosition::setIndex(int32_t offset)
+{
+  this->index = offset;
+}
+
+inline int32_t
+ParsePosition::getErrorIndex() const
+{
+  return errorIndex;
+}
+
+inline void
+ParsePosition::setErrorIndex(int32_t ei)
+{
+  this->errorIndex = ei;
+}
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/platform.h b/thirdparty/icu4c/common/unicode/platform.h
new file mode 100644
index 0000000000..2bb2f8b318
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/platform.h
@@ -0,0 +1,885 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : platform.h
+*
+*   Date        Name        Description
+*   05/13/98    nos         Creation (content moved here from ptypes.h).
+*   03/02/99    stephen     Added AS400 support.
+*   03/30/99    stephen     Added Linux support.
+*   04/13/99    stephen     Reworked for autoconf.
+******************************************************************************
+*/
+
+#ifndef _PLATFORM_H
+#define _PLATFORM_H
+
+#include "unicode/uconfig.h"
+#include "unicode/uvernum.h"
+
+/**
+ * \file
+ * \brief Basic types for the platform.
+ *
+ * This file used to be generated by autoconf/configure.
+ * Starting with ICU 49, platform.h is a normal source file,
+ * to simplify cross-compiling and working with non-autoconf/make build systems.
+ *
+ * When a value in this file does not work on a platform, then please
+ * try to derive it from the U_PLATFORM value
+ * (for which we might need a new value constant in rare cases)
+ * and/or from other macros that are predefined by the compiler
+ * or defined in standard (POSIX or platform or compiler) headers.
+ *
+ * As a temporary workaround, you can add an explicit \#define for some macros
+ * before it is first tested, or add an equivalent -D macro definition
+ * to the compiler's command line.
+ *
+ * Note: Some compilers provide ways to show the predefined macros.
+ * For example, with gcc you can compile an empty .c file and have the compiler
+ * print the predefined macros with
+ * \code
+ * gcc -E -dM -x c /dev/null | sort
+ * \endcode
+ * (You can provide an actual empty .c file rather than /dev/null.
+ * <code>-x c++</code> is for C++.)
+ */
+
+/**
+ * Define some things so that they can be documented.
+ * @internal
+ */
+#ifdef U_IN_DOXYGEN
+/*
+ * Problem: "platform.h:335: warning: documentation for unknown define U_HAVE_STD_STRING found." means that U_HAVE_STD_STRING is not documented.
+ * Solution: #define any defines for non @internal API here, so that they are visible in the docs.  If you just set PREDEFINED in Doxyfile.in,  they won't be documented.
+ */
+
+/* None for now. */
+#endif
+
+/**
+ * \def U_PLATFORM
+ * The U_PLATFORM macro defines the platform we're on.
+ *
+ * We used to define one different, value-less macro per platform.
+ * That made it hard to know the set of relevant platforms and macros,
+ * and hard to deal with variants of platforms.
+ *
+ * Starting with ICU 49, we define platforms as numeric macros,
+ * with ranges of values for related platforms and their variants.
+ * The U_PLATFORM macro is set to one of these values.
+ *
+ * Historical note from the Solaris Wikipedia article:
+ * AT&T and Sun collaborated on a project to merge the most popular Unix variants
+ * on the market at that time: BSD, System V, and Xenix.
+ * This became Unix System V Release 4 (SVR4).
+ *
+ * @internal
+ */
+
+/** Unknown platform. @internal */
+#define U_PF_UNKNOWN 0
+/** Windows @internal */
+#define U_PF_WINDOWS 1000
+/** MinGW. Windows, calls to Win32 API, but using GNU gcc and binutils. @internal */
+#define U_PF_MINGW 1800
+/**
+ * Cygwin. Windows, calls to cygwin1.dll for Posix functions,
+ * using MSVC or GNU gcc and binutils.
+ * @internal
+ */
+#define U_PF_CYGWIN 1900
+/* Reserve 2000 for U_PF_UNIX? */
+/** HP-UX is based on UNIX System V. @internal */
+#define U_PF_HPUX 2100
+/** Solaris is a Unix operating system based on SVR4. @internal */
+#define U_PF_SOLARIS 2600
+/** BSD is a UNIX operating system derivative. @internal */
+#define U_PF_BSD 3000
+/** AIX is based on UNIX System V Releases and 4.3 BSD. @internal */
+#define U_PF_AIX 3100
+/** IRIX is based on UNIX System V with BSD extensions. @internal */
+#define U_PF_IRIX 3200
+/**
+ * Darwin is a POSIX-compliant operating system, composed of code developed by Apple,
+ * as well as code derived from NeXTSTEP, BSD, and other projects,
+ * built around the Mach kernel.
+ * Darwin forms the core set of components upon which Mac OS X, Apple TV, and iOS are based.
+ * (Original description modified from WikiPedia.)
+ * @internal
+ */
+#define U_PF_DARWIN 3500
+/** iPhone OS (iOS) is a derivative of Mac OS X. @internal */
+#define U_PF_IPHONE 3550
+/** QNX is a commercial Unix-like real-time operating system related to BSD. @internal */
+#define U_PF_QNX 3700
+/** Linux is a Unix-like operating system. @internal */
+#define U_PF_LINUX 4000
+/**
+ * Native Client is pretty close to Linux.
+ * See https://developer.chrome.com/native-client and
+ *  http://www.chromium.org/nativeclient
+ *  @internal
+ */
+#define U_PF_BROWSER_NATIVE_CLIENT 4020
+/** Android is based on Linux. @internal */
+#define U_PF_ANDROID 4050
+/** Fuchsia is a POSIX-ish platform. @internal */
+#define U_PF_FUCHSIA 4100
+/* Maximum value for Linux-based platform is 4499 */
+/**
+ * Emscripten is a C++ transpiler for the Web that can target asm.js or
+ * WebAssembly. It provides some POSIX-compatible wrappers and stubs and
+ * some Linux-like functionality, but is not fully compatible with
+ * either.
+ * @internal
+ */
+#define U_PF_EMSCRIPTEN 5010
+/** z/OS is the successor to OS/390 which was the successor to MVS. @internal */
+#define U_PF_OS390 9000
+/** "IBM i" is the current name of what used to be i5/OS and earlier OS/400. @internal */
+#define U_PF_OS400 9400
+
+#ifdef U_PLATFORM
+    /* Use the predefined value. */
+#elif defined(__MINGW32__)
+#   define U_PLATFORM U_PF_MINGW
+#elif defined(__CYGWIN__)
+#   define U_PLATFORM U_PF_CYGWIN
+#elif defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+#   define U_PLATFORM U_PF_WINDOWS
+#elif defined(__ANDROID__)
+#   define U_PLATFORM U_PF_ANDROID
+    /* Android wchar_t support depends on the API level. */
+#   include <android/api-level.h>
+#elif defined(__pnacl__) || defined(__native_client__)
+#   define U_PLATFORM U_PF_BROWSER_NATIVE_CLIENT
+#elif defined(__Fuchsia__)
+#   define U_PLATFORM U_PF_FUCHSIA
+#elif defined(linux) || defined(__linux__) || defined(__linux)
+#   define U_PLATFORM U_PF_LINUX
+#elif defined(__APPLE__) && defined(__MACH__)
+#   include <TargetConditionals.h>
+#   if defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE  /* variant of TARGET_OS_MAC */
+#       define U_PLATFORM U_PF_IPHONE
+#   else
+#       define U_PLATFORM U_PF_DARWIN
+#   endif
+#elif defined(BSD) || defined(__FreeBSD__) || defined(__FreeBSD_kernel__) || defined(__NetBSD__) || defined(__OpenBSD__) || defined(__MirBSD__)
+#   if defined(__FreeBSD__)
+#       include <sys/endian.h>
+#   endif
+#   define U_PLATFORM U_PF_BSD
+#elif defined(sun) || defined(__sun)
+    /* Check defined(__SVR4) || defined(__svr4__) to distinguish Solaris from SunOS? */
+#   define U_PLATFORM U_PF_SOLARIS
+#   if defined(__GNUC__)
+        /* Solaris/GCC needs this header file to get the proper endianness. Normally, this
+         * header file is included with stddef.h but on Solairs/GCC, the GCC version of stddef.h
+         *  is included which does not include this header file.
+         */
+#       include <sys/isa_defs.h>
+#   endif
+#elif defined(_AIX) || defined(__TOS_AIX__)
+#   define U_PLATFORM U_PF_AIX
+#elif defined(_hpux) || defined(hpux) || defined(__hpux)
+#   define U_PLATFORM U_PF_HPUX
+#elif defined(sgi) || defined(__sgi)
+#   define U_PLATFORM U_PF_IRIX
+#elif defined(__QNX__) || defined(__QNXNTO__)
+#   define U_PLATFORM U_PF_QNX
+#elif defined(__TOS_MVS__)
+#   define U_PLATFORM U_PF_OS390
+#elif defined(__OS400__) || defined(__TOS_OS400__)
+#   define U_PLATFORM U_PF_OS400
+#elif defined(__EMSCRIPTEN__)
+#   define U_PLATFORM U_PF_EMSCRIPTEN
+#else
+#   define U_PLATFORM U_PF_UNKNOWN
+#endif
+
+/**
+ * \def CYGWINMSVC
+ * Defined if this is Windows with Cygwin, but using MSVC rather than gcc.
+ * Otherwise undefined.
+ * @internal
+ */
+/* Commented out because this is already set in mh-cygwin-msvc
+#if U_PLATFORM == U_PF_CYGWIN && defined(_MSC_VER)
+#   define CYGWINMSVC
+#endif
+*/
+#ifdef U_IN_DOXYGEN
+#   define CYGWINMSVC
+#endif
+
+/**
+ * \def U_PLATFORM_USES_ONLY_WIN32_API
+ * Defines whether the platform uses only the Win32 API.
+ * Set to 1 for Windows/MSVC and MinGW but not Cygwin.
+ * @internal
+ */
+#ifdef U_PLATFORM_USES_ONLY_WIN32_API
+    /* Use the predefined value. */
+#elif (U_PF_WINDOWS <= U_PLATFORM && U_PLATFORM <= U_PF_MINGW) || defined(CYGWINMSVC)
+#   define U_PLATFORM_USES_ONLY_WIN32_API 1
+#else
+    /* Cygwin implements POSIX. */
+#   define U_PLATFORM_USES_ONLY_WIN32_API 0
+#endif
+
+/**
+ * \def U_PLATFORM_HAS_WIN32_API
+ * Defines whether the Win32 API is available on the platform.
+ * Set to 1 for Windows/MSVC, MinGW and Cygwin.
+ * @internal
+ */
+#ifdef U_PLATFORM_HAS_WIN32_API
+    /* Use the predefined value. */
+#elif U_PF_WINDOWS <= U_PLATFORM && U_PLATFORM <= U_PF_CYGWIN
+#   define U_PLATFORM_HAS_WIN32_API 1
+#else
+#   define U_PLATFORM_HAS_WIN32_API 0
+#endif
+
+/**
+ * \def U_PLATFORM_HAS_WINUWP_API
+ * Defines whether target is intended for Universal Windows Platform API
+ * Set to 1 for Windows10 Release Solution Configuration
+ * @internal
+ */
+#ifdef U_PLATFORM_HAS_WINUWP_API
+    /* Use the predefined value. */
+#else
+#   define U_PLATFORM_HAS_WINUWP_API 0
+#endif
+
+/**
+ * \def U_PLATFORM_IMPLEMENTS_POSIX
+ * Defines whether the platform implements (most of) the POSIX API.
+ * Set to 1 for Cygwin and most other platforms.
+ * @internal
+ */
+#ifdef U_PLATFORM_IMPLEMENTS_POSIX
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_PLATFORM_IMPLEMENTS_POSIX 0
+#else
+#   define U_PLATFORM_IMPLEMENTS_POSIX 1
+#endif
+
+/**
+ * \def U_PLATFORM_IS_LINUX_BASED
+ * Defines whether the platform is Linux or one of its derivatives.
+ * @internal
+ */
+#ifdef U_PLATFORM_IS_LINUX_BASED
+    /* Use the predefined value. */
+#elif U_PF_LINUX <= U_PLATFORM && U_PLATFORM <= 4499
+#   define U_PLATFORM_IS_LINUX_BASED 1
+#else
+#   define U_PLATFORM_IS_LINUX_BASED 0
+#endif
+
+/**
+ * \def U_PLATFORM_IS_DARWIN_BASED
+ * Defines whether the platform is Darwin or one of its derivatives.
+ * @internal
+ */
+#ifdef U_PLATFORM_IS_DARWIN_BASED
+    /* Use the predefined value. */
+#elif U_PF_DARWIN <= U_PLATFORM && U_PLATFORM <= U_PF_IPHONE
+#   define U_PLATFORM_IS_DARWIN_BASED 1
+#else
+#   define U_PLATFORM_IS_DARWIN_BASED 0
+#endif
+
+/**
+ * \def U_HAVE_STDINT_H
+ * Defines whether stdint.h is available. It is a C99 standard header.
+ * We used to include inttypes.h which includes stdint.h but we usually do not need
+ * the additional definitions from inttypes.h.
+ * @internal
+ */
+#ifdef U_HAVE_STDINT_H
+    /* Use the predefined value. */
+#elif U_PLATFORM_USES_ONLY_WIN32_API
+#   if defined(__BORLANDC__) || U_PLATFORM == U_PF_MINGW || (defined(_MSC_VER) && _MSC_VER>=1600)
+        /* Windows Visual Studio 9 and below do not have stdint.h & inttypes.h, but VS 2010 adds them. */
+#       define U_HAVE_STDINT_H 1
+#   else
+#       define U_HAVE_STDINT_H 0
+#   endif
+#elif U_PLATFORM == U_PF_SOLARIS
+    /* Solaris has inttypes.h but not stdint.h. */
+#   define U_HAVE_STDINT_H 0
+#elif U_PLATFORM == U_PF_AIX && !defined(_AIX51) && defined(_POWER)
+    /* PPC AIX <= 4.3 has inttypes.h but not stdint.h. */
+#   define U_HAVE_STDINT_H 0
+#else
+#   define U_HAVE_STDINT_H 1
+#endif
+
+/**
+ * \def U_HAVE_INTTYPES_H
+ * Defines whether inttypes.h is available. It is a C99 standard header.
+ * We include inttypes.h where it is available but stdint.h is not.
+ * @internal
+ */
+#ifdef U_HAVE_INTTYPES_H
+    /* Use the predefined value. */
+#elif U_PLATFORM == U_PF_SOLARIS
+    /* Solaris has inttypes.h but not stdint.h. */
+#   define U_HAVE_INTTYPES_H 1
+#elif U_PLATFORM == U_PF_AIX && !defined(_AIX51) && defined(_POWER)
+    /* PPC AIX <= 4.3 has inttypes.h but not stdint.h. */
+#   define U_HAVE_INTTYPES_H 1
+#else
+    /* Most platforms have both inttypes.h and stdint.h, or neither. */
+#   define U_HAVE_INTTYPES_H U_HAVE_STDINT_H
+#endif
+
+/*===========================================================================*/
+/** @{ Compiler and environment features                                     */
+/*===========================================================================*/
+
+/**
+ * \def U_GCC_MAJOR_MINOR
+ * Indicates whether the compiler is gcc (test for != 0),
+ * and if so, contains its major (times 100) and minor version numbers.
+ * If the compiler is not gcc, then U_GCC_MAJOR_MINOR == 0.
+ *
+ * For example, for testing for whether we have gcc, and whether it's 4.6 or higher,
+ * use "#if U_GCC_MAJOR_MINOR >= 406".
+ * @internal
+ */
+#ifdef __GNUC__
+#   define U_GCC_MAJOR_MINOR (__GNUC__ * 100 + __GNUC_MINOR__)
+#else
+#   define U_GCC_MAJOR_MINOR 0
+#endif
+
+/**
+ * \def U_IS_BIG_ENDIAN
+ * Determines the endianness of the platform.
+ * @internal
+ */
+#ifdef U_IS_BIG_ENDIAN
+    /* Use the predefined value. */
+#elif defined(BYTE_ORDER) && defined(BIG_ENDIAN)
+#   define U_IS_BIG_ENDIAN (BYTE_ORDER == BIG_ENDIAN)
+#elif defined(__BYTE_ORDER__) && defined(__ORDER_BIG_ENDIAN__)
+    /* gcc */
+#   define U_IS_BIG_ENDIAN (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
+#elif defined(__BIG_ENDIAN__) || defined(_BIG_ENDIAN)
+#   define U_IS_BIG_ENDIAN 1
+#elif defined(__LITTLE_ENDIAN__) || defined(_LITTLE_ENDIAN)
+#   define U_IS_BIG_ENDIAN 0
+#elif U_PLATFORM == U_PF_OS390 || U_PLATFORM == U_PF_OS400 || defined(__s390__) || defined(__s390x__)
+    /* These platforms do not appear to predefine any endianness macros. */
+#   define U_IS_BIG_ENDIAN 1
+#elif defined(_PA_RISC1_0) || defined(_PA_RISC1_1) || defined(_PA_RISC2_0)
+    /* HPPA do not appear to predefine any endianness macros. */
+#   define U_IS_BIG_ENDIAN 1
+#elif defined(sparc) || defined(__sparc) || defined(__sparc__)
+    /* Some sparc based systems (e.g. Linux) do not predefine any endianness macros. */
+#   define U_IS_BIG_ENDIAN 1
+#else
+#   define U_IS_BIG_ENDIAN 0
+#endif
+
+/**
+ * \def U_HAVE_PLACEMENT_NEW
+ * Determines whether to override placement new and delete for STL.
+ * @stable ICU 2.6
+ */
+#ifdef U_HAVE_PLACEMENT_NEW
+    /* Use the predefined value. */
+#elif defined(__BORLANDC__)
+#   define U_HAVE_PLACEMENT_NEW 0
+#else
+#   define U_HAVE_PLACEMENT_NEW 1
+#endif
+
+/**
+ * \def U_HAVE_DEBUG_LOCATION_NEW 
+ * Define this to define the MFC debug version of the operator new.
+ *
+ * @stable ICU 3.4
+ */
+#ifdef U_HAVE_DEBUG_LOCATION_NEW
+    /* Use the predefined value. */
+#elif defined(_MSC_VER)
+#   define U_HAVE_DEBUG_LOCATION_NEW 1
+#else
+#   define U_HAVE_DEBUG_LOCATION_NEW 0
+#endif
+
+/* Compatibility with compilers other than clang: http://clang.llvm.org/docs/LanguageExtensions.html */
+#ifdef __has_attribute
+#   define UPRV_HAS_ATTRIBUTE(x) __has_attribute(x)
+#else
+#   define UPRV_HAS_ATTRIBUTE(x) 0
+#endif
+#ifdef __has_cpp_attribute
+#   define UPRV_HAS_CPP_ATTRIBUTE(x) __has_cpp_attribute(x)
+#else
+#   define UPRV_HAS_CPP_ATTRIBUTE(x) 0
+#endif
+#ifdef __has_declspec_attribute
+#   define UPRV_HAS_DECLSPEC_ATTRIBUTE(x) __has_declspec_attribute(x)
+#else
+#   define UPRV_HAS_DECLSPEC_ATTRIBUTE(x) 0
+#endif
+#ifdef __has_builtin
+#   define UPRV_HAS_BUILTIN(x) __has_builtin(x)
+#else
+#   define UPRV_HAS_BUILTIN(x) 0
+#endif
+#ifdef __has_feature
+#   define UPRV_HAS_FEATURE(x) __has_feature(x)
+#else
+#   define UPRV_HAS_FEATURE(x) 0
+#endif
+#ifdef __has_extension
+#   define UPRV_HAS_EXTENSION(x) __has_extension(x)
+#else
+#   define UPRV_HAS_EXTENSION(x) 0
+#endif
+#ifdef __has_warning
+#   define UPRV_HAS_WARNING(x) __has_warning(x)
+#else
+#   define UPRV_HAS_WARNING(x) 0
+#endif
+
+/**
+ * \def U_MALLOC_ATTR
+ * Attribute to mark functions as malloc-like
+ * @internal
+ */
+#if defined(__GNUC__) && __GNUC__>=3
+#    define U_MALLOC_ATTR __attribute__ ((__malloc__))
+#else
+#    define U_MALLOC_ATTR
+#endif
+
+/**
+ * \def U_ALLOC_SIZE_ATTR
+ * Attribute to specify the size of the allocated buffer for malloc-like functions
+ * @internal
+ */
+#if (defined(__GNUC__) &&                                            \
+        (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) || \
+        UPRV_HAS_ATTRIBUTE(alloc_size)
+#   define U_ALLOC_SIZE_ATTR(X) __attribute__ ((alloc_size(X)))
+#   define U_ALLOC_SIZE_ATTR2(X,Y) __attribute__ ((alloc_size(X,Y)))
+#else
+#   define U_ALLOC_SIZE_ATTR(X)
+#   define U_ALLOC_SIZE_ATTR2(X,Y)
+#endif
+
+/**
+ * \def U_CPLUSPLUS_VERSION
+ * 0 if no C++; 1, 11, 14, ... if C++.
+ * Support for specific features cannot always be determined by the C++ version alone.
+ * @internal
+ */
+#ifdef U_CPLUSPLUS_VERSION
+#   if U_CPLUSPLUS_VERSION != 0 && !defined(__cplusplus)
+#       undef U_CPLUSPLUS_VERSION
+#       define U_CPLUSPLUS_VERSION 0
+#   endif
+    /* Otherwise use the predefined value. */
+#elif !defined(__cplusplus)
+#   define U_CPLUSPLUS_VERSION 0
+#elif __cplusplus >= 201402L || (defined(_MSVC_LANG) && _MSVC_LANG >= 201402L)
+#   define U_CPLUSPLUS_VERSION 14
+#elif __cplusplus >= 201103L || (defined(_MSVC_LANG) && _MSVC_LANG >= 201103L)
+#   define U_CPLUSPLUS_VERSION 11
+#else
+    // C++98 or C++03
+#   define U_CPLUSPLUS_VERSION 1
+#endif
+
+#if (U_PLATFORM == U_PF_AIX || U_PLATFORM == U_PF_OS390) && defined(__cplusplus) &&(U_CPLUSPLUS_VERSION < 11)
+// add in std::nullptr_t
+namespace std {
+  typedef decltype(nullptr) nullptr_t;
+};
+#endif
+
+/**
+ * \def U_NOEXCEPT
+ * "noexcept" if supported, otherwise empty.
+ * Some code, especially STL containers, uses move semantics of objects only
+ * if the move constructor and the move operator are declared as not throwing exceptions.
+ * @internal
+ */
+#ifdef U_NOEXCEPT
+    /* Use the predefined value. */
+#else
+#   define U_NOEXCEPT noexcept
+#endif
+
+/**
+ * \def U_FALLTHROUGH
+ * Annotate intentional fall-through between switch labels.
+ * http://clang.llvm.org/docs/AttributeReference.html#fallthrough-clang-fallthrough
+ * @internal
+ */
+#ifndef __cplusplus
+    // Not for C.
+#elif defined(U_FALLTHROUGH)
+    // Use the predefined value.
+#elif defined(__clang__)
+    // Test for compiler vs. feature separately.
+    // Other compilers might choke on the feature test.
+#    if UPRV_HAS_CPP_ATTRIBUTE(clang::fallthrough) || \
+             (UPRV_HAS_FEATURE(cxx_attributes) &&     \
+             UPRV_HAS_WARNING("-Wimplicit-fallthrough"))
+#       define U_FALLTHROUGH [[clang::fallthrough]]
+#   endif
+#elif defined(__GNUC__) && (__GNUC__ >= 7)
+#   define U_FALLTHROUGH __attribute__((fallthrough))
+#endif
+
+#ifndef U_FALLTHROUGH
+#   define U_FALLTHROUGH
+#endif
+
+/** @} */
+
+/*===========================================================================*/
+/** @{ Character data types                                                  */
+/*===========================================================================*/
+
+/**
+ * U_CHARSET_FAMILY is equal to this value when the platform is an ASCII based platform.
+ * @stable ICU 2.0
+ */
+#define U_ASCII_FAMILY 0
+
+/**
+ * U_CHARSET_FAMILY is equal to this value when the platform is an EBCDIC based platform.
+ * @stable ICU 2.0
+ */
+#define U_EBCDIC_FAMILY 1
+
+/**
+ * \def U_CHARSET_FAMILY
+ *
+ * <p>These definitions allow to specify the encoding of text
+ * in the char data type as defined by the platform and the compiler.
+ * It is enough to determine the code point values of "invariant characters",
+ * which are the ones shared by all encodings that are in use
+ * on a given platform.</p>
+ *
+ * <p>Those "invariant characters" should be all the uppercase and lowercase
+ * latin letters, the digits, the space, and "basic punctuation".
+ * Also, '\\n', '\\r', '\\t' should be available.</p>
+ *
+ * <p>The list of "invariant characters" is:<br>
+ * \code
+ *    A-Z  a-z  0-9  SPACE  "  %  &amp;  '  (  )  *  +  ,  -  .  /  :  ;  <  =  >  ?  _
+ * \endcode
+ * <br>
+ * (52 letters + 10 numbers + 20 punc/sym/space = 82 total)</p>
+ *
+ * <p>This matches the IBM Syntactic Character Set (CS 640).</p>
+ *
+ * <p>In other words, all the graphic characters in 7-bit ASCII should
+ * be safely accessible except the following:</p>
+ *
+ * \code
+ *    '\' <backslash>
+ *    '[' <left bracket>
+ *    ']' <right bracket>
+ *    '{' <left brace>
+ *    '}' <right brace>
+ *    '^' <circumflex>
+ *    '~' <tilde>
+ *    '!' <exclamation mark>
+ *    '#' <number sign>
+ *    '|' <vertical line>
+ *    '$' <dollar sign>
+ *    '@' <commercial at>
+ *    '`' <grave accent>
+ * \endcode
+ * @stable ICU 2.0
+ */
+#ifdef U_CHARSET_FAMILY
+    /* Use the predefined value. */
+#elif U_PLATFORM == U_PF_OS390 && (!defined(__CHARSET_LIB) || !__CHARSET_LIB)
+#   define U_CHARSET_FAMILY U_EBCDIC_FAMILY
+#elif U_PLATFORM == U_PF_OS400 && !defined(__UTF32__)
+#   define U_CHARSET_FAMILY U_EBCDIC_FAMILY
+#else
+#   define U_CHARSET_FAMILY U_ASCII_FAMILY
+#endif
+
+/**
+ * \def U_CHARSET_IS_UTF8
+ *
+ * Hardcode the default charset to UTF-8.
+ *
+ * If this is set to 1, then
+ * - ICU will assume that all non-invariant char*, StringPiece, std::string etc.
+ *   contain UTF-8 text, regardless of what the system API uses
+ * - some ICU code will use fast functions like u_strFromUTF8()
+ *   rather than the more general and more heavy-weight conversion API (ucnv.h)
+ * - ucnv_getDefaultName() always returns "UTF-8"
+ * - ucnv_setDefaultName() is disabled and will not change the default charset
+ * - static builds of ICU are smaller
+ * - more functionality is available with the UCONFIG_NO_CONVERSION build-time
+ *   configuration option (see unicode/uconfig.h)
+ * - the UCONFIG_NO_CONVERSION build option in uconfig.h is more usable
+ *
+ * @stable ICU 4.2
+ * @see UCONFIG_NO_CONVERSION
+ */
+#ifdef U_CHARSET_IS_UTF8
+    /* Use the predefined value. */
+#elif U_PLATFORM_IS_LINUX_BASED || U_PLATFORM_IS_DARWIN_BASED || \
+        U_PLATFORM == U_PF_EMSCRIPTEN
+#   define U_CHARSET_IS_UTF8 1
+#else
+#   define U_CHARSET_IS_UTF8 0
+#endif
+
+/** @} */
+
+/*===========================================================================*/
+/** @{ Information about wchar support                                       */
+/*===========================================================================*/
+
+/**
+ * \def U_HAVE_WCHAR_H
+ * Indicates whether <wchar.h> is available (1) or not (0). Set to 1 by default.
+ *
+ * @stable ICU 2.0
+ */
+#ifdef U_HAVE_WCHAR_H
+    /* Use the predefined value. */
+#elif U_PLATFORM == U_PF_ANDROID && __ANDROID_API__ < 9
+    /*
+     * Android before Gingerbread (Android 2.3, API level 9) did not support wchar_t.
+     * The type and header existed, but the library functions did not work as expected.
+     * The size of wchar_t was 1 but L"xyz" string literals had 32-bit units anyway.
+     */
+#   define U_HAVE_WCHAR_H 0
+#else
+#   define U_HAVE_WCHAR_H 1
+#endif
+
+/**
+ * \def U_SIZEOF_WCHAR_T
+ * U_SIZEOF_WCHAR_T==sizeof(wchar_t)
+ *
+ * @stable ICU 2.0
+ */
+#ifdef U_SIZEOF_WCHAR_T
+    /* Use the predefined value. */
+#elif (U_PLATFORM == U_PF_ANDROID && __ANDROID_API__ < 9)
+    /*
+     * Classic Mac OS and Mac OS X before 10.3 (Panther) did not support wchar_t or wstring.
+     * Newer Mac OS X has size 4.
+     */
+#   define U_SIZEOF_WCHAR_T 1
+#elif U_PLATFORM_HAS_WIN32_API || U_PLATFORM == U_PF_CYGWIN
+#   define U_SIZEOF_WCHAR_T 2
+#elif U_PLATFORM == U_PF_AIX
+    /*
+     * AIX 6.1 information, section "Wide character data representation":
+     * "... the wchar_t datatype is 32-bit in the 64-bit environment and
+     * 16-bit in the 32-bit environment."
+     * and
+     * "All locales use Unicode for their wide character code values (process code),
+     * except the IBM-eucTW codeset."
+     */
+#   ifdef __64BIT__
+#       define U_SIZEOF_WCHAR_T 4
+#   else
+#       define U_SIZEOF_WCHAR_T 2
+#   endif
+#elif U_PLATFORM == U_PF_OS390
+    /*
+     * z/OS V1R11 information center, section "LP64 | ILP32":
+     * "In 31-bit mode, the size of long and pointers is 4 bytes and the size of wchar_t is 2 bytes.
+     * Under LP64, the size of long and pointer is 8 bytes and the size of wchar_t is 4 bytes."
+     */
+#   ifdef _LP64
+#       define U_SIZEOF_WCHAR_T 4
+#   else
+#       define U_SIZEOF_WCHAR_T 2
+#   endif
+#elif U_PLATFORM == U_PF_OS400
+#   if defined(__UTF32__)
+        /*
+         * LOCALETYPE(*LOCALEUTF) is specified.
+         * Wide-character strings are in UTF-32,
+         * narrow-character strings are in UTF-8.
+         */
+#       define U_SIZEOF_WCHAR_T 4
+#   elif defined(__UCS2__)
+        /*
+         * LOCALETYPE(*LOCALEUCS2) is specified.
+         * Wide-character strings are in UCS-2,
+         * narrow-character strings are in EBCDIC.
+         */
+#       define U_SIZEOF_WCHAR_T 2
+#   else
+        /*
+         * LOCALETYPE(*CLD) or LOCALETYPE(*LOCALE) is specified.
+         * Wide-character strings are in 16-bit EBCDIC,
+         * narrow-character strings are in EBCDIC.
+         */
+#       define U_SIZEOF_WCHAR_T 2
+#   endif
+#else
+#   define U_SIZEOF_WCHAR_T 4
+#endif
+
+#ifndef U_HAVE_WCSCPY
+#define U_HAVE_WCSCPY U_HAVE_WCHAR_H
+#endif
+
+/** @} */
+
+/**
+ * \def U_HAVE_CHAR16_T
+ * Defines whether the char16_t type is available for UTF-16
+ * and u"abc" UTF-16 string literals are supported.
+ * This is a new standard type and standard string literal syntax in C++0x
+ * but has been available in some compilers before.
+ * @internal
+ */
+#ifdef U_HAVE_CHAR16_T
+    /* Use the predefined value. */
+#else
+    /*
+     * Notes:
+     * Visual Studio 2010 (_MSC_VER==1600) defines char16_t as a typedef
+     * and does not support u"abc" string literals.
+     * Visual Studio 2015 (_MSC_VER>=1900) and above adds support for
+     * both char16_t and u"abc" string literals.
+     * gcc 4.4 defines the __CHAR16_TYPE__ macro to a usable type but
+     * does not support u"abc" string literals.
+     * C++11 and C11 require support for UTF-16 literals
+     * TODO: Fix for plain C. Doesn't work on Mac.
+     */
+#   if U_CPLUSPLUS_VERSION >= 11 || (defined(__STDC_VERSION__) && __STDC_VERSION__ >= 201112L)
+#       define U_HAVE_CHAR16_T 1
+#   else
+#       define U_HAVE_CHAR16_T 0
+#   endif
+#endif
+
+/**
+ * @{
+ * \def U_DECLARE_UTF16
+ * Do not use this macro because it is not defined on all platforms.
+ * Use the UNICODE_STRING or U_STRING_DECL macros instead.
+ * @internal
+ */
+#ifdef U_DECLARE_UTF16
+    /* Use the predefined value. */
+#elif U_HAVE_CHAR16_T \
+    || (defined(__xlC__) && defined(__IBM_UTF_LITERAL) && U_SIZEOF_WCHAR_T != 2) \
+    || (defined(__HP_aCC) && __HP_aCC >= 035000) \
+    || (defined(__HP_cc) && __HP_cc >= 111106) \
+    || (defined(U_IN_DOXYGEN))
+#   define U_DECLARE_UTF16(string) u ## string
+#elif U_SIZEOF_WCHAR_T == 2 \
+    && (U_CHARSET_FAMILY == 0 || (U_PF_OS390 <= U_PLATFORM && U_PLATFORM <= U_PF_OS400 && defined(__UCS2__)))
+#   define U_DECLARE_UTF16(string) L ## string
+#else
+    /* Leave U_DECLARE_UTF16 undefined. See unistr.h. */
+#endif
+
+/** @} */
+
+/*===========================================================================*/
+/** @{ Symbol import-export control                                          */
+/*===========================================================================*/
+
+#ifdef U_EXPORT
+    /* Use the predefined value. */
+#elif defined(U_STATIC_IMPLEMENTATION)
+#   define U_EXPORT
+#elif defined(_MSC_VER) || (UPRV_HAS_DECLSPEC_ATTRIBUTE(dllexport) && \
+                            UPRV_HAS_DECLSPEC_ATTRIBUTE(dllimport))
+#   define U_EXPORT __declspec(dllexport)
+#elif defined(__GNUC__)
+#   define U_EXPORT __attribute__((visibility("default")))
+#elif (defined(__SUNPRO_CC) && __SUNPRO_CC >= 0x550) \
+   || (defined(__SUNPRO_C) && __SUNPRO_C >= 0x550) 
+#   define U_EXPORT __global
+/*#elif defined(__HP_aCC) || defined(__HP_cc)
+#   define U_EXPORT __declspec(dllexport)*/
+#else
+#   define U_EXPORT
+#endif
+
+/* U_CALLCONV is related to U_EXPORT2 */
+#ifdef U_EXPORT2
+    /* Use the predefined value. */
+#elif defined(_MSC_VER)
+#   define U_EXPORT2 __cdecl
+#else
+#   define U_EXPORT2
+#endif
+
+#ifdef U_IMPORT
+    /* Use the predefined value. */
+#elif defined(_MSC_VER) || (UPRV_HAS_DECLSPEC_ATTRIBUTE(dllexport) && \
+                            UPRV_HAS_DECLSPEC_ATTRIBUTE(dllimport))
+    /* Windows needs to export/import data. */
+#   define U_IMPORT __declspec(dllimport)
+#else
+#   define U_IMPORT 
+#endif
+
+/**
+ * \def U_CALLCONV
+ * Similar to U_CDECL_BEGIN/U_CDECL_END, this qualifier is necessary
+ * in callback function typedefs to make sure that the calling convention
+ * is compatible.
+ *
+ * This is only used for non-ICU-API functions.
+ * When a function is a public ICU API,
+ * you must use the U_CAPI and U_EXPORT2 qualifiers.
+ *
+ * Please note, you need to use U_CALLCONV after the *.
+ *
+ * NO : "static const char U_CALLCONV *func( . . . )"
+ * YES: "static const char* U_CALLCONV func( . . . )"
+ *
+ * @stable ICU 2.0
+ */
+#if U_PLATFORM == U_PF_OS390 && defined(__cplusplus)
+#    define U_CALLCONV __cdecl
+#else
+#    define U_CALLCONV U_EXPORT2
+#endif
+
+/**
+ * \def U_CALLCONV_FPTR
+ * Similar to U_CALLCONV, but only used on function pointers.
+ * @internal
+ */
+#if U_PLATFORM == U_PF_OS390 && defined(__cplusplus)
+#    define U_CALLCONV_FPTR U_CALLCONV
+#else
+#    define U_CALLCONV_FPTR
+#endif
+/* @} */
+
+#endif  // _PLATFORM_H
diff --git a/thirdparty/icu4c/common/unicode/ptypes.h b/thirdparty/icu4c/common/unicode/ptypes.h
new file mode 100644
index 0000000000..70324ffee3
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ptypes.h
@@ -0,0 +1,130 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : ptypes.h
+*
+*   Date        Name        Description
+*   05/13/98    nos         Creation (content moved here from ptypes.h).
+*   03/02/99    stephen     Added AS400 support.
+*   03/30/99    stephen     Added Linux support.
+*   04/13/99    stephen     Reworked for autoconf.
+*   09/18/08    srl         Moved basic types back to ptypes.h from platform.h
+******************************************************************************
+*/
+
+/**
+ * \file
+ * \brief C API: Definitions of integer types of various widths
+ */
+
+#ifndef _PTYPES_H
+#define _PTYPES_H
+
+/**
+ * \def __STDC_LIMIT_MACROS
+ * According to the Linux stdint.h, the ISO C99 standard specifies that in C++ implementations
+ * macros like INT32_MIN and UINTPTR_MAX should only be defined if explicitly requested.
+ * We need to define __STDC_LIMIT_MACROS before including stdint.h in C++ code
+ * that uses such limit macros.
+ * @internal
+ */
+#ifndef __STDC_LIMIT_MACROS
+#define __STDC_LIMIT_MACROS
+#endif
+
+/* NULL, size_t, wchar_t */
+#include <stddef.h>
+
+/*
+ * If all compilers provided all of the C99 headers and types,
+ * we would just unconditionally #include <stdint.h> here
+ * and not need any of the stuff after including platform.h.
+ */
+
+/* Find out if we have stdint.h etc. */
+#include "unicode/platform.h"
+
+/*===========================================================================*/
+/* Generic data types                                                        */
+/*===========================================================================*/
+
+/* If your platform does not have the <stdint.h> header, you may
+   need to edit the typedefs in the #else section below.
+   Use #if...#else...#endif with predefined compiler macros if possible. */
+#if U_HAVE_STDINT_H
+
+/*
+ * We mostly need <stdint.h> (which defines the standard integer types) but not <inttypes.h>.
+ * <inttypes.h> includes <stdint.h> and adds the printf/scanf helpers PRId32, SCNx16 etc.
+ * which we almost never use, plus stuff like imaxabs() which we never use.
+ */
+#include <stdint.h>
+
+#if U_PLATFORM == U_PF_OS390
+/* The features header is needed to get (u)int64_t sometimes. */
+#include <features.h>
+/* z/OS has <stdint.h>, but some versions are missing uint8_t (APAR PK62248). */
+#if !defined(__uint8_t)
+#define __uint8_t 1
+typedef unsigned char uint8_t;
+#endif
+#endif /* U_PLATFORM == U_PF_OS390 */
+
+#elif U_HAVE_INTTYPES_H
+
+#   include <inttypes.h>
+
+#else /* neither U_HAVE_STDINT_H nor U_HAVE_INTTYPES_H */
+
+/// \cond
+#if ! U_HAVE_INT8_T
+typedef signed char int8_t;
+#endif
+
+#if ! U_HAVE_UINT8_T
+typedef unsigned char uint8_t;
+#endif
+
+#if ! U_HAVE_INT16_T
+typedef signed short int16_t;
+#endif
+
+#if ! U_HAVE_UINT16_T
+typedef unsigned short uint16_t;
+#endif
+
+#if ! U_HAVE_INT32_T
+typedef signed int int32_t;
+#endif
+
+#if ! U_HAVE_UINT32_T
+typedef unsigned int uint32_t;
+#endif
+
+#if ! U_HAVE_INT64_T
+#ifdef _MSC_VER
+    typedef signed __int64 int64_t;
+#else
+    typedef signed long long int64_t;
+#endif
+#endif
+
+#if ! U_HAVE_UINT64_T
+#ifdef _MSC_VER
+    typedef unsigned __int64 uint64_t;
+#else
+    typedef unsigned long long uint64_t;
+#endif
+#endif
+/// \endcond
+
+#endif /* U_HAVE_STDINT_H / U_HAVE_INTTYPES_H */
+
+#endif /* _PTYPES_H */
diff --git a/thirdparty/icu4c/common/unicode/putil.h b/thirdparty/icu4c/common/unicode/putil.h
new file mode 100644
index 0000000000..500c21252f
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/putil.h
@@ -0,0 +1,183 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : putil.h
+*
+*   Date        Name        Description
+*   05/14/98    nos         Creation (content moved here from utypes.h).
+*   06/17/99    erm         Added IEEE_754
+*   07/22/98    stephen     Added IEEEremainder, max, min, trunc
+*   08/13/98    stephen     Added isNegativeInfinity, isPositiveInfinity
+*   08/24/98    stephen     Added longBitsFromDouble
+*   03/02/99    stephen     Removed openFile().  Added AS400 support.
+*   04/15/99    stephen     Converted to C
+*   11/15/99    helena      Integrated S/390 changes for IEEE support.
+*   01/11/00    helena      Added u_getVersion.
+******************************************************************************
+*/
+
+#ifndef PUTIL_H
+#define PUTIL_H
+
+#include "unicode/utypes.h"
+ /**
+  * \file
+  * \brief C API: Platform Utilities
+  */
+
+/*==========================================================================*/
+/* Platform utilities                                                       */
+/*==========================================================================*/
+
+/**
+ * Platform utilities isolates the platform dependencies of the
+ * library.  For each platform which this code is ported to, these
+ * functions may have to be re-implemented.
+ */
+
+/**
+ * Return the ICU data directory. 
+ * The data directory is where common format ICU data files (.dat files)
+ *   are loaded from.  Note that normal use of the built-in ICU
+ *   facilities does not require loading of an external data file;
+ *   unless you are adding custom data to ICU, the data directory
+ *   does not need to be set.
+ *
+ * The data directory is determined as follows:
+ *    If u_setDataDirectory() has been called, that is it, otherwise
+ *    if the ICU_DATA environment variable is set, use that, otherwise
+ *    If a data directory was specified at ICU build time
+ *      <code>
+ * \code
+ *        #define ICU_DATA_DIR "path" 
+ * \endcode
+ * </code> use that,
+ *    otherwise no data directory is available.
+ *
+ * @return the data directory, or an empty string ("") if no data directory has
+ *         been specified.
+ *   
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2 u_getDataDirectory(void);
+
+
+/** 
+ * Set the ICU data directory. 
+ * The data directory is where common format ICU data files (.dat files)
+ *   are loaded from.  Note that normal use of the built-in ICU
+ *   facilities does not require loading of an external data file;
+ *   unless you are adding custom data to ICU, the data directory
+ *   does not need to be set.
+ *
+ * This function should be called at most once in a process, before the
+ * first ICU operation (e.g., u_init()) that will require the loading of an
+ * ICU data file.
+ * This function is not thread-safe. Use it before calling ICU APIs from
+ * multiple threads.
+ *
+ * @param directory The directory to be set.
+ *
+ * @see u_init
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 u_setDataDirectory(const char *directory);
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+  * Return the time zone files override directory, or an empty string if
+  * no directory was specified. Certain time zone resources will be preferentially
+  * loaded from individual files in this directory.
+  *
+  * @return the time zone data override directory.
+  * @internal
+  */ 
+U_CAPI const char * U_EXPORT2 u_getTimeZoneFilesDirectory(UErrorCode *status);
+
+/**
+  * Set the time zone files override directory.
+  * This function is not thread safe; it must not be called concurrently with
+  *   u_getTimeZoneFilesDirectory() or any other use of ICU time zone functions.
+  * This function should only be called before using any ICU service that
+  *   will access the time zone data.
+  * @internal
+  */
+U_CAPI void U_EXPORT2 u_setTimeZoneFilesDirectory(const char *path, UErrorCode *status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+
+/**
+ * @{
+ * Filesystem file and path separator characters.
+ * Example: '/' and ':' on Unix, '\\' and ';' on Windows.
+ * @stable ICU 2.0
+ */
+#if U_PLATFORM_USES_ONLY_WIN32_API
+#   define U_FILE_SEP_CHAR '\\'
+#   define U_FILE_ALT_SEP_CHAR '/'
+#   define U_PATH_SEP_CHAR ';'
+#   define U_FILE_SEP_STRING "\\"
+#   define U_FILE_ALT_SEP_STRING "/"
+#   define U_PATH_SEP_STRING ";"
+#else
+#   define U_FILE_SEP_CHAR '/'
+#   define U_FILE_ALT_SEP_CHAR '/'
+#   define U_PATH_SEP_CHAR ':'
+#   define U_FILE_SEP_STRING "/"
+#   define U_FILE_ALT_SEP_STRING "/"
+#   define U_PATH_SEP_STRING ":"
+#endif
+
+/** @} */
+
+/**
+ * Convert char characters to UChar characters.
+ * This utility function is useful only for "invariant characters"
+ * that are encoded in the platform default encoding.
+ * They are a small, constant subset of the encoding and include
+ * just the latin letters, digits, and some punctuation.
+ * For details, see U_CHARSET_FAMILY.
+ *
+ * @param cs Input string, points to <code>length</code>
+ *           character bytes from a subset of the platform encoding.
+ * @param us Output string, points to memory for <code>length</code>
+ *           Unicode characters.
+ * @param length The number of characters to convert; this may
+ *               include the terminating <code>NUL</code>.
+ *
+ * @see U_CHARSET_FAMILY
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+u_charsToUChars(const char *cs, UChar *us, int32_t length);
+
+/**
+ * Convert UChar characters to char characters.
+ * This utility function is useful only for "invariant characters"
+ * that can be encoded in the platform default encoding.
+ * They are a small, constant subset of the encoding and include
+ * just the latin letters, digits, and some punctuation.
+ * For details, see U_CHARSET_FAMILY.
+ *
+ * @param us Input string, points to <code>length</code>
+ *           Unicode characters that can be encoded with the
+ *           codepage-invariant subset of the platform encoding.
+ * @param cs Output string, points to memory for <code>length</code>
+ *           character bytes.
+ * @param length The number of characters to convert; this may
+ *               include the terminating <code>NUL</code>.
+ *
+ * @see U_CHARSET_FAMILY
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+u_UCharsToChars(const UChar *us, char *cs, int32_t length);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/rbbi.h b/thirdparty/icu4c/common/unicode/rbbi.h
new file mode 100644
index 0000000000..65117f616c
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/rbbi.h
@@ -0,0 +1,732 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+*   Copyright (C) 1999-2016 International Business Machines Corporation   *
+*   and others. All rights reserved.                                      *
+***************************************************************************
+
+**********************************************************************
+*   Date        Name        Description
+*   10/22/99    alan        Creation.
+*   11/11/99    rgillam     Complete port from Java.
+**********************************************************************
+*/
+
+#ifndef RBBI_H
+#define RBBI_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C++ API: Rule Based Break Iterator
+ */
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/brkiter.h"
+#include "unicode/udata.h"
+#include "unicode/parseerr.h"
+#include "unicode/schriter.h"
+
+struct UCPTrie;
+
+U_NAMESPACE_BEGIN
+
+/** @internal */
+class  LanguageBreakEngine;
+struct RBBIDataHeader;
+class  RBBIDataWrapper;
+class  UnhandledEngine;
+class  UStack;
+
+/**
+ *
+ * A subclass of BreakIterator whose behavior is specified using a list of rules.
+ * <p>Instances of this class are most commonly created by the factory methods of
+ *  BreakIterator::createWordInstance(), BreakIterator::createLineInstance(), etc.,
+ *  and then used via the abstract API in class BreakIterator</p>
+ *
+ * <p>See the ICU User Guide for information on Break Iterator Rules.</p>
+ *
+ * <p>This class is not intended to be subclassed.</p>
+ */
+class U_COMMON_API RuleBasedBreakIterator /*U_FINAL*/ : public BreakIterator {
+
+private:
+    /**
+     * The UText through which this BreakIterator accesses the text
+     * @internal (private)
+     */
+    UText  fText;
+
+#ifndef U_HIDE_INTERNAL_API
+public:
+#endif /* U_HIDE_INTERNAL_API */
+    /**
+     * The rule data for this BreakIterator instance.
+     * Not for general use; Public only for testing purposes.
+     * @internal
+     */
+    RBBIDataWrapper    *fData;
+private:
+
+    /**
+      * The current  position of the iterator. Pinned, 0 < fPosition <= text.length.
+      * Never has the value UBRK_DONE (-1).
+      */
+    int32_t         fPosition;
+
+    /**
+      * TODO:
+      */
+    int32_t         fRuleStatusIndex;
+
+    /**
+     *   Cache of previously determined boundary positions.
+     */
+    class BreakCache;
+    BreakCache         *fBreakCache;
+
+    /**
+     *  Cache of boundary positions within a region of text that has been
+     *  sub-divided by dictionary based breaking.
+     */
+    class DictionaryCache;
+    DictionaryCache *fDictionaryCache;
+
+    /**
+     *
+     * If present, UStack of LanguageBreakEngine objects that might handle
+     * dictionary characters. Searched from top to bottom to find an object to
+     * handle a given character.
+     * @internal (private)
+     */
+    UStack              *fLanguageBreakEngines;
+
+    /**
+     *
+     * If present, the special LanguageBreakEngine used for handling
+     * characters that are in the dictionary set, but not handled by any
+     * LanguageBreakEngine.
+     * @internal (private)
+     */
+    UnhandledEngine     *fUnhandledBreakEngine;
+
+    /**
+     * Counter for the number of characters encountered with the "dictionary"
+     *   flag set.
+     * @internal (private)
+     */
+    uint32_t            fDictionaryCharCount;
+
+    /**
+     *   A character iterator that refers to the same text as the UText, above.
+     *   Only included for compatibility with old API, which was based on CharacterIterators.
+     *   Value may be adopted from outside, or one of fSCharIter or fDCharIter, below.
+     */
+    CharacterIterator  *fCharIter;
+
+    /**
+     *   When the input text is provided by a UnicodeString, this will point to
+     *    a characterIterator that wraps that data.  Needed only for the
+     *    implementation of getText(), a backwards compatibility issue.
+     */
+    StringCharacterIterator fSCharIter;
+
+    /**
+      * True when iteration has run off the end, and iterator functions should return UBRK_DONE.
+      */
+    UBool           fDone;
+
+    /**
+     *  Array of look-ahead tentative results.
+     */
+    int32_t *fLookAheadMatches;
+
+    //=======================================================================
+    // constructors
+    //=======================================================================
+
+    /**
+     * Constructor from a flattened set of RBBI data in malloced memory.
+     *             RulesBasedBreakIterators built from a custom set of rules
+     *             are created via this constructor; the rules are compiled
+     *             into memory, then the break iterator is constructed here.
+     *
+     *             The break iterator adopts the memory, and will
+     *             free it when done.
+     * @internal (private)
+     */
+    RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode &status);
+
+    /** @internal */
+    friend class RBBIRuleBuilder;
+    /** @internal */
+    friend class BreakIterator;
+
+public:
+
+    /** Default constructor.  Creates an empty shell of an iterator, with no
+     *  rules or text to iterate over.   Object can subsequently be assigned to.
+     *  @stable ICU 2.2
+     */
+    RuleBasedBreakIterator();
+
+    /**
+     * Copy constructor.  Will produce a break iterator with the same behavior,
+     * and which iterates over the same text, as the one passed in.
+     * @param that The RuleBasedBreakIterator passed to be copied
+     * @stable ICU 2.0
+     */
+    RuleBasedBreakIterator(const RuleBasedBreakIterator& that);
+
+    /**
+     * Construct a RuleBasedBreakIterator from a set of rules supplied as a string.
+     * @param rules The break rules to be used.
+     * @param parseError  In the event of a syntax error in the rules, provides the location
+     *                    within the rules of the problem.
+     * @param status Information on any errors encountered.
+     * @stable ICU 2.2
+     */
+    RuleBasedBreakIterator( const UnicodeString    &rules,
+                             UParseError           &parseError,
+                             UErrorCode            &status);
+
+    /**
+     * Construct a RuleBasedBreakIterator from a set of precompiled binary rules.
+     * Binary rules are obtained from RulesBasedBreakIterator::getBinaryRules().
+     * Construction of a break iterator in this way is substantially faster than
+     * construction from source rules.
+     *
+     * Ownership of the storage containing the compiled rules remains with the
+     * caller of this function.  The compiled rules must not be  modified or
+     * deleted during the life of the break iterator.
+     *
+     * The compiled rules are not compatible across different major versions of ICU.
+     * The compiled rules are compatible only between machines with the same
+     * byte ordering (little or big endian) and the same base character set family
+     * (ASCII or EBCDIC).
+     *
+     * @see #getBinaryRules
+     * @param compiledRules A pointer to the compiled break rules to be used.
+     * @param ruleLength The length of the compiled break rules, in bytes.  This
+     *   corresponds to the length value produced by getBinaryRules().
+     * @param status Information on any errors encountered, including invalid
+     *   binary rules.
+     * @stable ICU 4.8
+     */
+    RuleBasedBreakIterator(const uint8_t *compiledRules,
+                           uint32_t       ruleLength,
+                           UErrorCode    &status);
+
+    /**
+     * This constructor uses the udata interface to create a BreakIterator
+     * whose internal tables live in a memory-mapped file.  "image" is an
+     * ICU UDataMemory handle for the pre-compiled break iterator tables.
+     * @param image handle to the memory image for the break iterator data.
+     *        Ownership of the UDataMemory handle passes to the Break Iterator,
+     *        which will be responsible for closing it when it is no longer needed.
+     * @param status Information on any errors encountered.
+     * @see udata_open
+     * @see #getBinaryRules
+     * @stable ICU 2.8
+     */
+    RuleBasedBreakIterator(UDataMemory* image, UErrorCode &status);
+
+    /**
+     * Destructor
+     *  @stable ICU 2.0
+     */
+    virtual ~RuleBasedBreakIterator();
+
+    /**
+     * Assignment operator.  Sets this iterator to have the same behavior,
+     * and iterate over the same text, as the one passed in.
+     * @param that The RuleBasedBreakItertor passed in
+     * @return the newly created RuleBasedBreakIterator
+     *  @stable ICU 2.0
+     */
+    RuleBasedBreakIterator& operator=(const RuleBasedBreakIterator& that);
+
+    /**
+     * Equality operator.  Returns true if both BreakIterators are of the
+     * same class, have the same behavior, and iterate over the same text.
+     * @param that The BreakIterator to be compared for equality
+     * @return true if both BreakIterators are of the
+     * same class, have the same behavior, and iterate over the same text.
+     *  @stable ICU 2.0
+     */
+    virtual UBool operator==(const BreakIterator& that) const;
+
+    /**
+     * Not-equal operator.  If operator== returns true, this returns false,
+     * and vice versa.
+     * @param that The BreakIterator to be compared for inequality
+     * @return true if both BreakIterators are not same.
+     *  @stable ICU 2.0
+     */
+    inline UBool operator!=(const BreakIterator& that) const;
+
+    /**
+     * Returns a newly-constructed RuleBasedBreakIterator with the same
+     * behavior, and iterating over the same text, as this one.
+     * Differs from the copy constructor in that it is polymorphic, and
+     * will correctly clone (copy) a derived class.
+     * clone() is thread safe.  Multiple threads may simultaneously
+     * clone the same source break iterator.
+     * @return a newly-constructed RuleBasedBreakIterator
+     * @stable ICU 2.0
+     */
+    virtual RuleBasedBreakIterator* clone() const;
+
+    /**
+     * Compute a hash code for this BreakIterator
+     * @return A hash code
+     *  @stable ICU 2.0
+     */
+    virtual int32_t hashCode(void) const;
+
+    /**
+     * Returns the description used to create this iterator
+     * @return the description used to create this iterator
+     *  @stable ICU 2.0
+     */
+    virtual const UnicodeString& getRules(void) const;
+
+    //=======================================================================
+    // BreakIterator overrides
+    //=======================================================================
+
+    /**
+     * <p>
+     * Return a CharacterIterator over the text being analyzed.
+     * The returned character iterator is owned by the break iterator, and must
+     * not be deleted by the caller.  Repeated calls to this function may
+     * return the same CharacterIterator.
+     * </p>
+     * <p>
+     * The returned character iterator must not be used concurrently with
+     * the break iterator.  If concurrent operation is needed, clone the
+     * returned character iterator first and operate on the clone.
+     * </p>
+     * <p>
+     * When the break iterator is operating on text supplied via a UText,
+     * this function will fail.  Lacking any way to signal failures, it
+     * returns an CharacterIterator containing no text.
+     * The function getUText() provides similar functionality,
+     * is reliable, and is more efficient.
+     * </p>
+     *
+     * TODO:  deprecate this function?
+     *
+     * @return An iterator over the text being analyzed.
+     * @stable ICU 2.0
+     */
+    virtual  CharacterIterator& getText(void) const;
+
+
+    /**
+      *  Get a UText for the text being analyzed.
+      *  The returned UText is a shallow clone of the UText used internally
+      *  by the break iterator implementation.  It can safely be used to
+      *  access the text without impacting any break iterator operations,
+      *  but the underlying text itself must not be altered.
+      *
+      * @param fillIn A UText to be filled in.  If NULL, a new UText will be
+      *           allocated to hold the result.
+      * @param status receives any error codes.
+      * @return   The current UText for this break iterator.  If an input
+      *           UText was provided, it will always be returned.
+      * @stable ICU 3.4
+      */
+     virtual UText *getUText(UText *fillIn, UErrorCode &status) const;
+
+    /**
+     * Set the iterator to analyze a new piece of text.  This function resets
+     * the current iteration position to the beginning of the text.
+     * @param newText An iterator over the text to analyze.  The BreakIterator
+     * takes ownership of the character iterator.  The caller MUST NOT delete it!
+     *  @stable ICU 2.0
+     */
+    virtual void adoptText(CharacterIterator* newText);
+
+    /**
+     * Set the iterator to analyze a new piece of text.  This function resets
+     * the current iteration position to the beginning of the text.
+     *
+     * The BreakIterator will retain a reference to the supplied string.
+     * The caller must not modify or delete the text while the BreakIterator
+     * retains the reference.
+     *
+     * @param newText The text to analyze.
+     *  @stable ICU 2.0
+     */
+    virtual void setText(const UnicodeString& newText);
+
+    /**
+     * Reset the break iterator to operate over the text represented by
+     * the UText.  The iterator position is reset to the start.
+     *
+     * This function makes a shallow clone of the supplied UText.  This means
+     * that the caller is free to immediately close or otherwise reuse the
+     * Utext that was passed as a parameter, but that the underlying text itself
+     * must not be altered while being referenced by the break iterator.
+     *
+     * @param text    The UText used to change the text.
+     * @param status  Receives any error codes.
+     * @stable ICU 3.4
+     */
+    virtual void  setText(UText *text, UErrorCode &status);
+
+    /**
+     * Sets the current iteration position to the beginning of the text, position zero.
+     * @return The offset of the beginning of the text, zero.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t first(void);
+
+    /**
+     * Sets the current iteration position to the end of the text.
+     * @return The text's past-the-end offset.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t last(void);
+
+    /**
+     * Advances the iterator either forward or backward the specified number of steps.
+     * Negative values move backward, and positive values move forward.  This is
+     * equivalent to repeatedly calling next() or previous().
+     * @param n The number of steps to move.  The sign indicates the direction
+     * (negative is backwards, and positive is forwards).
+     * @return The character offset of the boundary position n boundaries away from
+     * the current one.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t next(int32_t n);
+
+    /**
+     * Advances the iterator to the next boundary position.
+     * @return The position of the first boundary after this one.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t next(void);
+
+    /**
+     * Moves the iterator backwards, to the last boundary preceding this one.
+     * @return The position of the last boundary position preceding this one.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t previous(void);
+
+    /**
+     * Sets the iterator to refer to the first boundary position following
+     * the specified position.
+     * @param offset The position from which to begin searching for a break position.
+     * @return The position of the first break after the current position.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t following(int32_t offset);
+
+    /**
+     * Sets the iterator to refer to the last boundary position before the
+     * specified position.
+     * @param offset The position to begin searching for a break from.
+     * @return The position of the last boundary before the starting position.
+     *  @stable ICU 2.0
+     */
+    virtual int32_t preceding(int32_t offset);
+
+    /**
+     * Returns true if the specified position is a boundary position.  As a side
+     * effect, leaves the iterator pointing to the first boundary position at
+     * or after "offset".
+     * @param offset the offset to check.
+     * @return True if "offset" is a boundary position.
+     *  @stable ICU 2.0
+     */
+    virtual UBool isBoundary(int32_t offset);
+
+    /**
+     * Returns the current iteration position. Note that UBRK_DONE is never
+     * returned from this function; if iteration has run to the end of a
+     * string, current() will return the length of the string while
+     * next() will return UBRK_DONE).
+     * @return The current iteration position.
+     * @stable ICU 2.0
+     */
+    virtual int32_t current(void) const;
+
+
+    /**
+     * Return the status tag from the break rule that determined the boundary at
+     * the current iteration position.  For break rules that do not specify a
+     * status, a default value of 0 is returned.  If more than one break rule
+     * would cause a boundary to be located at some position in the text,
+     * the numerically largest of the applicable status values is returned.
+     * <p>
+     * Of the standard types of ICU break iterators, only word break and
+     * line break provide status values.  The values are defined in
+     * the header file ubrk.h.  For Word breaks, the status allows distinguishing between words
+     * that contain alphabetic letters, "words" that appear to be numbers,
+     * punctuation and spaces, words containing ideographic characters, and
+     * more.  For Line Break, the status distinguishes between hard (mandatory) breaks
+     * and soft (potential) break positions.
+     * <p>
+     * <code>getRuleStatus()</code> can be called after obtaining a boundary
+     * position from <code>next()</code>, <code>previous()</code>, or
+     * any other break iterator functions that returns a boundary position.
+     * <p>
+     * Note that <code>getRuleStatus()</code> returns the value corresponding to
+     * <code>current()</code> index even after <code>next()</code> has returned DONE.
+     * <p>
+     * When creating custom break rules, one is free to define whatever
+     * status values may be convenient for the application.
+     * <p>
+     * @return the status from the break rule that determined the boundary
+     * at the current iteration position.
+     *
+     * @see UWordBreak
+     * @stable ICU 2.2
+     */
+    virtual int32_t getRuleStatus() const;
+
+   /**
+    * Get the status (tag) values from the break rule(s) that determined the boundary
+    * at the current iteration position.
+    * <p>
+    * The returned status value(s) are stored into an array provided by the caller.
+    * The values are stored in sorted (ascending) order.
+    * If the capacity of the output array is insufficient to hold the data,
+    *  the output will be truncated to the available length, and a
+    *  U_BUFFER_OVERFLOW_ERROR will be signaled.
+    *
+    * @param fillInVec an array to be filled in with the status values.
+    * @param capacity  the length of the supplied vector.  A length of zero causes
+    *                  the function to return the number of status values, in the
+    *                  normal way, without attempting to store any values.
+    * @param status    receives error codes.
+    * @return          The number of rule status values from the rules that determined
+    *                  the boundary at the current iteration position.
+    *                  In the event of a U_BUFFER_OVERFLOW_ERROR, the return value
+    *                  is the total number of status values that were available,
+    *                  not the reduced number that were actually returned.
+    * @see getRuleStatus
+    * @stable ICU 3.0
+    */
+    virtual int32_t getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status);
+
+    /**
+     * Returns a unique class ID POLYMORPHICALLY.  Pure virtual override.
+     * This method is to implement a simple version of RTTI, since not all
+     * C++ compilers support genuine RTTI.  Polymorphic operator==() and
+     * clone() methods call this method.
+     *
+     * @return          The class ID for this object. All objects of a
+     *                  given class have the same class ID.  Objects of
+     *                  other classes have different class IDs.
+     * @stable ICU 2.0
+     */
+    virtual UClassID getDynamicClassID(void) const;
+
+    /**
+     * Returns the class ID for this class.  This is useful only for
+     * comparing to a return value from getDynamicClassID().  For example:
+     *
+     *      Base* polymorphic_pointer = createPolymorphicObject();
+     *      if (polymorphic_pointer->getDynamicClassID() ==
+     *          Derived::getStaticClassID()) ...
+     *
+     * @return          The class ID for all objects of this class.
+     * @stable ICU 2.0
+     */
+    static UClassID U_EXPORT2 getStaticClassID(void);
+
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+    /**
+     * Deprecated functionality. Use clone() instead.
+     *
+     * Create a clone (copy) of this break iterator in memory provided
+     *  by the caller.  The idea is to increase performance by avoiding
+     *  a storage allocation.  Use of this function is NOT RECOMMENDED.
+     *  Performance gains are minimal, and correct buffer management is
+     *  tricky.  Use clone() instead.
+     *
+     * @param stackBuffer  The pointer to the memory into which the cloned object
+     *                     should be placed.  If NULL,  allocate heap memory
+     *                     for the cloned object.
+     * @param BufferSize   The size of the buffer.  If zero, return the required
+     *                     buffer size, but do not clone the object.  If the
+     *                     size was too small (but not zero), allocate heap
+     *                     storage for the cloned object.
+     *
+     * @param status       Error status.  U_SAFECLONE_ALLOCATED_WARNING will be
+     *                     returned if the provided buffer was too small, and
+     *                     the clone was therefore put on the heap.
+     *
+     * @return  Pointer to the clone object.  This may differ from the stackBuffer
+     *          address if the byte alignment of the stack buffer was not suitable
+     *          or if the stackBuffer was too small to hold the clone.
+     * @deprecated ICU 52. Use clone() instead.
+     */
+    virtual RuleBasedBreakIterator *createBufferClone(void *stackBuffer,
+                                                      int32_t &BufferSize,
+                                                      UErrorCode &status);
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+    /**
+     * Return the binary form of compiled break rules,
+     * which can then be used to create a new break iterator at some
+     * time in the future.  Creating a break iterator from pre-compiled rules
+     * is much faster than building one from the source form of the
+     * break rules.
+     *
+     * The binary data can only be used with the same version of ICU
+     *  and on the same platform type (processor endian-ness)
+     *
+     * @param length Returns the length of the binary data.  (Out parameter.)
+     *
+     * @return   A pointer to the binary (compiled) rule data.  The storage
+     *           belongs to the RulesBasedBreakIterator object, not the
+     *           caller, and must not be modified or deleted.
+     * @stable ICU 4.8
+     */
+    virtual const uint8_t *getBinaryRules(uint32_t &length);
+
+    /**
+     *  Set the subject text string upon which the break iterator is operating
+     *  without changing any other aspect of the matching state.
+     *  The new and previous text strings must have the same content.
+     *
+     *  This function is intended for use in environments where ICU is operating on
+     *  strings that may move around in memory.  It provides a mechanism for notifying
+     *  ICU that the string has been relocated, and providing a new UText to access the
+     *  string in its new position.
+     *
+     *  Note that the break iterator implementation never copies the underlying text
+     *  of a string being processed, but always operates directly on the original text
+     *  provided by the user. Refreshing simply drops the references to the old text
+     *  and replaces them with references to the new.
+     *
+     *  Caution:  this function is normally used only by very specialized,
+     *  system-level code.  One example use case is with garbage collection that moves
+     *  the text in memory.
+     *
+     * @param input      The new (moved) text string.
+     * @param status     Receives errors detected by this function.
+     * @return           *this
+     *
+     * @stable ICU 49
+     */
+    virtual RuleBasedBreakIterator &refreshInputText(UText *input, UErrorCode &status);
+
+
+private:
+    //=======================================================================
+    // implementation
+    //=======================================================================
+    /**
+     * Dumps caches and performs other actions associated with a complete change
+     * in text or iteration position.
+     * @internal (private)
+     */
+    void reset(void);
+
+    /**
+      * Common initialization function, used by constructors and bufferClone.
+      * @internal (private)
+      */
+    void init(UErrorCode &status);
+
+    /**
+     * Iterate backwards from an arbitrary position in the input text using the
+     * synthesized Safe Reverse rules.
+     * This locates a "Safe Position" from which the forward break rules
+     * will operate correctly. A Safe Position is not necessarily a boundary itself.
+     *
+     * @param fromPosition the position in the input text to begin the iteration.
+     * @internal (private)
+     */
+    int32_t handleSafePrevious(int32_t fromPosition);
+
+    /**
+     * Find a rule-based boundary by running the state machine.
+     * Input
+     *    fPosition, the position in the text to begin from.
+     * Output
+     *    fPosition:           the boundary following the starting position.
+     *    fDictionaryCharCount the number of dictionary characters encountered.
+     *                         If > 0, the segment will be further subdivided
+     *    fRuleStatusIndex     Info from the state table indicating which rules caused the boundary.
+     *
+     * @internal (private)
+     */
+    int32_t handleNext();
+
+    /*
+     * Templatized version of handleNext() and handleSafePrevious().
+     *
+     * There will be exactly four instantiations, two each for 8 and 16 bit tables,
+     * two each for 8 and 16 bit trie.
+     * Having separate instantiations for the table types keeps conditional tests of
+     * the table type out of the inner loops, at the expense of replicated code.
+     *
+     * The template parameter for the Trie access function is a value, not a type.
+     * Doing it this way, the compiler will inline the Trie function in the
+     * expanded functions. (Both the 8 and 16 bit access functions have the same type
+     * signature)
+     */
+
+    typedef uint16_t (*PTrieFunc)(const UCPTrie *, UChar32);
+
+    template<typename RowType, PTrieFunc trieFunc>
+    int32_t handleSafePrevious(int32_t fromPosition);
+
+    template<typename RowType, PTrieFunc trieFunc>
+    int32_t handleNext();
+
+
+    /**
+     * This function returns the appropriate LanguageBreakEngine for a
+     * given character c.
+     * @param c         A character in the dictionary set
+     * @internal (private)
+     */
+    const LanguageBreakEngine *getLanguageBreakEngine(UChar32 c);
+
+  public:
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     *   Debugging function only.
+     *   @internal
+     */
+     void dumpCache();
+
+    /**
+     * Debugging function only.
+     * @internal
+     */
+    void dumpTables();
+#endif  /* U_HIDE_INTERNAL_API */
+};
+
+//------------------------------------------------------------------------------
+//
+//   Inline Functions Definitions ...
+//
+//------------------------------------------------------------------------------
+
+inline UBool RuleBasedBreakIterator::operator!=(const BreakIterator& that) const {
+    return !operator==(that);
+}
+
+U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/rep.h b/thirdparty/icu4c/common/unicode/rep.h
new file mode 100644
index 0000000000..6dd4530647
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/rep.h
@@ -0,0 +1,266 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**************************************************************************
+* Copyright (C) 1999-2012, International Business Machines Corporation and
+* others. All Rights Reserved.
+**************************************************************************
+*   Date        Name        Description
+*   11/17/99    aliu        Creation.  Ported from java.  Modified to
+*                           match current UnicodeString API.  Forced
+*                           to use name "handleReplaceBetween" because
+*                           of existing methods in UnicodeString.
+**************************************************************************
+*/
+
+#ifndef REP_H
+#define REP_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file 
+ * \brief C++ API: Replaceable String
+ */
+ 
+U_NAMESPACE_BEGIN
+
+class UnicodeString;
+
+/**
+ * <code>Replaceable</code> is an abstract base class representing a
+ * string of characters that supports the replacement of a range of
+ * itself with a new string of characters.  It is used by APIs that
+ * change a piece of text while retaining metadata.  Metadata is data
+ * other than the Unicode characters returned by char32At().  One
+ * example of metadata is style attributes; another is an edit
+ * history, marking each character with an author and revision number.
+ *
+ * <p>An implicit aspect of the <code>Replaceable</code> API is that
+ * during a replace operation, new characters take on the metadata of
+ * the old characters.  For example, if the string "the <b>bold</b>
+ * font" has range (4, 8) replaced with "strong", then it becomes "the
+ * <b>strong</b> font".
+ *
+ * <p><code>Replaceable</code> specifies ranges using a start
+ * offset and a limit offset.  The range of characters thus specified
+ * includes the characters at offset start..limit-1.  That is, the
+ * start offset is inclusive, and the limit offset is exclusive.
+ *
+ * <p><code>Replaceable</code> also includes API to access characters
+ * in the string: <code>length()</code>, <code>charAt()</code>,
+ * <code>char32At()</code>, and <code>extractBetween()</code>.
+ *
+ * <p>For a subclass to support metadata, typical behavior of
+ * <code>replace()</code> is the following:
+ * <ul>
+ *   <li>Set the metadata of the new text to the metadata of the first
+ *   character replaced</li>
+ *   <li>If no characters are replaced, use the metadata of the
+ *   previous character</li>
+ *   <li>If there is no previous character (i.e. start == 0), use the
+ *   following character</li>
+ *   <li>If there is no following character (i.e. the replaceable was
+ *   empty), use default metadata.<br>
+ *   <li>If the code point U+FFFF is seen, it should be interpreted as
+ *   a special marker having no metadata<li>
+ *   </li>
+ * </ul>
+ * If this is not the behavior, the subclass should document any differences.
+ * @author Alan Liu
+ * @stable ICU 2.0
+ */
+class U_COMMON_API Replaceable : public UObject {
+
+public:
+    /**
+     * Destructor.
+     * @stable ICU 2.0
+     */
+    virtual ~Replaceable();
+
+    /**
+     * Returns the number of 16-bit code units in the text.
+     * @return number of 16-bit code units in text
+     * @stable ICU 1.8
+     */ 
+    inline int32_t length() const;
+
+    /**
+     * Returns the 16-bit code unit at the given offset into the text.
+     * @param offset an integer between 0 and <code>length()</code>-1
+     * inclusive
+     * @return 16-bit code unit of text at given offset
+     * @stable ICU 1.8
+     */
+    inline char16_t charAt(int32_t offset) const;
+
+    /**
+     * Returns the 32-bit code point at the given 16-bit offset into
+     * the text.  This assumes the text is stored as 16-bit code units
+     * with surrogate pairs intermixed.  If the offset of a leading or
+     * trailing code unit of a surrogate pair is given, return the
+     * code point of the surrogate pair.
+     *
+     * @param offset an integer between 0 and <code>length()</code>-1
+     * inclusive
+     * @return 32-bit code point of text at given offset
+     * @stable ICU 1.8
+     */
+    inline UChar32 char32At(int32_t offset) const;
+
+    /**
+     * Copies characters in the range [<tt>start</tt>, <tt>limit</tt>) 
+     * into the UnicodeString <tt>target</tt>.
+     * @param start offset of first character which will be copied
+     * @param limit offset immediately following the last character to
+     * be copied
+     * @param target UnicodeString into which to copy characters.
+     * @return A reference to <TT>target</TT>
+     * @stable ICU 2.1
+     */
+    virtual void extractBetween(int32_t start,
+                                int32_t limit,
+                                UnicodeString& target) const = 0;
+
+    /**
+     * Replaces a substring of this object with the given text.  If the
+     * characters being replaced have metadata, the new characters
+     * that replace them should be given the same metadata.
+     *
+     * <p>Subclasses must ensure that if the text between start and
+     * limit is equal to the replacement text, that replace has no
+     * effect. That is, any metadata
+     * should be unaffected. In addition, subclasses are encouraged to
+     * check for initial and trailing identical characters, and make a
+     * smaller replacement if possible. This will preserve as much
+     * metadata as possible.
+     * @param start the beginning index, inclusive; <code>0 <= start
+     * <= limit</code>.
+     * @param limit the ending index, exclusive; <code>start <= limit
+     * <= length()</code>.
+     * @param text the text to replace characters <code>start</code>
+     * to <code>limit - 1</code> 
+     * @stable ICU 2.0
+     */
+    virtual void handleReplaceBetween(int32_t start,
+                                      int32_t limit,
+                                      const UnicodeString& text) = 0;
+    // Note: All other methods in this class take the names of
+    // existing UnicodeString methods.  This method is the exception.
+    // It is named differently because all replace methods of
+    // UnicodeString return a UnicodeString&.  The 'between' is
+    // required in order to conform to the UnicodeString naming
+    // convention; API taking start/length are named <operation>, and
+    // those taking start/limit are named <operationBetween>.  The
+    // 'handle' is added because 'replaceBetween' and
+    // 'doReplaceBetween' are already taken.
+
+    /**
+     * Copies a substring of this object, retaining metadata.
+     * This method is used to duplicate or reorder substrings.
+     * The destination index must not overlap the source range.
+     * 
+     * @param start the beginning index, inclusive; <code>0 <= start <=
+     * limit</code>.
+     * @param limit the ending index, exclusive; <code>start <= limit <=
+     * length()</code>.
+     * @param dest the destination index.  The characters from
+     * <code>start..limit-1</code> will be copied to <code>dest</code>.
+     * Implementations of this method may assume that <code>dest <= start ||
+     * dest >= limit</code>.
+     * @stable ICU 2.0
+     */
+    virtual void copy(int32_t start, int32_t limit, int32_t dest) = 0;
+
+    /**
+     * Returns true if this object contains metadata.  If a
+     * Replaceable object has metadata, calls to the Replaceable API
+     * must be made so as to preserve metadata.  If it does not, calls
+     * to the Replaceable API may be optimized to improve performance.
+     * The default implementation returns true.
+     * @return true if this object contains metadata
+     * @stable ICU 2.2
+     */
+    virtual UBool hasMetaData() const;
+
+    /**
+     * Clone this object, an instance of a subclass of Replaceable.
+     * Clones can be used concurrently in multiple threads.
+     * If a subclass does not implement clone(), or if an error occurs,
+     * then NULL is returned.
+     * The caller must delete the clone.
+     *
+     * @return a clone of this object
+     *
+     * @see getDynamicClassID
+     * @stable ICU 2.6
+     */
+    virtual Replaceable *clone() const;
+
+protected:
+
+    /**
+     * Default constructor.
+     * @stable ICU 2.4
+     */
+    inline Replaceable();
+
+    /*
+     * Assignment operator not declared. The compiler will provide one
+     * which does nothing since this class does not contain any data members.
+     * API/code coverage may show the assignment operator as present and
+     * untested - ignore.
+     * Subclasses need this assignment operator if they use compiler-provided
+     * assignment operators of their own. An alternative to not declaring one
+     * here would be to declare and empty-implement a protected or public one.
+    Replaceable &Replaceable::operator=(const Replaceable &);
+     */
+
+    /**
+     * Virtual version of length().
+     * @stable ICU 2.4
+     */ 
+    virtual int32_t getLength() const = 0;
+
+    /**
+     * Virtual version of charAt().
+     * @stable ICU 2.4
+     */
+    virtual char16_t getCharAt(int32_t offset) const = 0;
+
+    /**
+     * Virtual version of char32At().
+     * @stable ICU 2.4
+     */
+    virtual UChar32 getChar32At(int32_t offset) const = 0;
+};
+
+inline Replaceable::Replaceable() {}
+
+inline int32_t
+Replaceable::length() const {
+    return getLength();
+}
+
+inline char16_t
+Replaceable::charAt(int32_t offset) const {
+    return getCharAt(offset);
+}
+
+inline UChar32
+Replaceable::char32At(int32_t offset) const {
+    return getChar32At(offset);
+}
+
+// There is no rep.cpp, see unistr.cpp for Replaceable function implementations.
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/resbund.h b/thirdparty/icu4c/common/unicode/resbund.h
new file mode 100644
index 0000000000..37738e277b
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/resbund.h
@@ -0,0 +1,498 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1996-2013, International Business Machines Corporation
+*   and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File resbund.h
+*
+*   CREATED BY
+*       Richard Gillam
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   2/5/97      aliu        Added scanForLocaleInFile.  Added
+*                           constructor which attempts to read resource bundle
+*                           from a specific file, without searching other files.
+*   2/11/97     aliu        Added UErrorCode return values to constructors.  Fixed
+*                           infinite loops in scanForFile and scanForLocale.
+*                           Modified getRawResourceData to not delete storage
+*                           in localeData and resourceData which it doesn't own.
+*                           Added Mac compatibility #ifdefs for tellp() and
+*                           ios::nocreate.
+*   2/18/97     helena      Updated with 100% documentation coverage.
+*   3/13/97     aliu        Rewrote to load in entire resource bundle and store
+*                           it as a Hashtable of ResourceBundleData objects.
+*                           Added state table to govern parsing of files.
+*                           Modified to load locale index out of new file
+*                           distinct from default.txt.
+*   3/25/97     aliu        Modified to support 2-d arrays, needed for timezone
+*                           data. Added support for custom file suffixes.  Again,
+*                           needed to support timezone data.
+*   4/7/97      aliu        Cleaned up.
+* 03/02/99      stephen     Removed dependency on FILE*.
+* 03/29/99      helena      Merged Bertrand and Stephen's changes.
+* 06/11/99      stephen     Removed parsing of .txt files.
+*                           Reworked to use new binary format.
+*                           Cleaned up.
+* 06/14/99      stephen     Removed methods taking a filename suffix.
+* 11/09/99      weiv        Added getLocale(), fRealLocale, removed fRealLocaleID
+******************************************************************************
+*/
+
+#ifndef RESBUND_H
+#define RESBUND_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+#include "unicode/ures.h"
+#include "unicode/unistr.h"
+#include "unicode/locid.h"
+
+/**
+ * \file 
+ * \brief C++ API: Resource Bundle
+ */
+ 
+U_NAMESPACE_BEGIN
+
+/**
+ * A class representing a collection of resource information pertaining to a given
+ * locale. A resource bundle provides a way of accessing locale- specfic information in
+ * a data file. You create a resource bundle that manages the resources for a given
+ * locale and then ask it for individual resources.
+ * <P>
+ * Resource bundles in ICU4C are currently defined using text files which conform to the following
+ * <a href="http://source.icu-project.org/repos/icu/icuhtml/trunk/design/bnf_rb.txt">BNF definition</a>.
+ * More on resource bundle concepts and syntax can be found in the
+ * <a href="http://icu-project.org/userguide/ResourceManagement.html">Users Guide</a>.
+ * <P>
+ *
+ * The ResourceBundle class is not suitable for subclassing.
+ *
+ * @stable ICU 2.0
+ */
+class U_COMMON_API ResourceBundle : public UObject {
+public:
+    /**
+     * Constructor
+     *
+     * @param packageName   The packageName and locale together point to an ICU udata object, 
+     *                      as defined by <code> udata_open( packageName, "res", locale, err) </code> 
+     *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+     *                      a package registered with udata_setAppData(). Using a full file or directory
+     *                      pathname for packageName is deprecated.
+     * @param locale  This is the locale this resource bundle is for. To get resources
+     *                for the French locale, for example, you would create a
+     *                ResourceBundle passing Locale::FRENCH for the "locale" parameter,
+     *                and all subsequent calls to that resource bundle will return
+     *                resources that pertain to the French locale. If the caller doesn't
+     *                pass a locale parameter, the default locale for the system (as
+     *                returned by Locale::getDefault()) will be used.
+     * @param err     The Error Code.
+     * The UErrorCode& err parameter is used to return status information to the user. To
+     * check whether the construction succeeded or not, you should check the value of
+     * U_SUCCESS(err). If you wish more detailed information, you can check for
+     * informational error results which still indicate success. U_USING_FALLBACK_WARNING
+     * indicates that a fall back locale was used. For example, 'de_CH' was requested,
+     * but nothing was found there, so 'de' was used. U_USING_DEFAULT_WARNING indicates that
+     * the default locale data was used; neither the requested locale nor any of its
+     * fall back locales could be found.
+     * @stable ICU 2.0
+     */
+    ResourceBundle(const UnicodeString&    packageName,
+                   const Locale&           locale,
+                   UErrorCode&              err);
+
+    /**
+     * Construct a resource bundle for the default bundle in the specified package.
+     *
+     * @param packageName   The packageName and locale together point to an ICU udata object, 
+     *                      as defined by <code> udata_open( packageName, "res", locale, err) </code> 
+     *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+     *                      a package registered with udata_setAppData(). Using a full file or directory
+     *                      pathname for packageName is deprecated.
+     * @param err A UErrorCode value
+     * @stable ICU 2.0
+     */
+    ResourceBundle(const UnicodeString&    packageName,
+                   UErrorCode&              err);
+
+    /**
+     * Construct a resource bundle for the ICU default bundle.
+     *
+     * @param err A UErrorCode value
+     * @stable ICU 2.0
+     */
+    ResourceBundle(UErrorCode &err);
+
+    /**
+     * Standard constructor, constructs a resource bundle for the locale-specific
+     * bundle in the specified package.
+     *
+     * @param packageName   The packageName and locale together point to an ICU udata object, 
+     *                      as defined by <code> udata_open( packageName, "res", locale, err) </code> 
+     *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+     *                      a package registered with udata_setAppData(). Using a full file or directory
+     *                      pathname for packageName is deprecated.
+     *                      NULL is used to refer to ICU data.
+     * @param locale The locale for which to open a resource bundle.
+     * @param err A UErrorCode value
+     * @stable ICU 2.0
+     */
+    ResourceBundle(const char* packageName,
+                   const Locale& locale,
+                   UErrorCode& err);
+
+    /**
+     * Copy constructor.
+     *
+     * @param original The resource bundle to copy.
+     * @stable ICU 2.0
+     */
+    ResourceBundle(const ResourceBundle &original);
+
+    /**
+     * Constructor from a C UResourceBundle. The resource bundle is
+     * copied and not adopted. ures_close will still need to be used on the
+     * original resource bundle.
+     *
+     * @param res A pointer to the C resource bundle.
+     * @param status A UErrorCode value.
+     * @stable ICU 2.0
+     */
+    ResourceBundle(UResourceBundle *res,
+                   UErrorCode &status);
+
+    /**
+     * Assignment operator.
+     *
+     * @param other The resource bundle to copy.
+     * @stable ICU 2.0
+     */
+    ResourceBundle&
+      operator=(const ResourceBundle& other);
+
+    /** Destructor.
+     * @stable ICU 2.0
+     */
+    virtual ~ResourceBundle();
+
+    /**
+     * Clone this object.
+     * Clones can be used concurrently in multiple threads.
+     * If an error occurs, then NULL is returned.
+     * The caller must delete the clone.
+     *
+     * @return a clone of this object
+     *
+     * @see getDynamicClassID
+     * @stable ICU 2.8
+     */
+    ResourceBundle *clone() const;
+
+    /**
+     * Returns the size of a resource. Size for scalar types is always 1, and for vector/table types is
+     * the number of child resources.
+     * @warning Integer array is treated as a scalar type. There are no
+     *          APIs to access individual members of an integer array. It
+     *          is always returned as a whole.
+     *
+     * @return number of resources in a given resource.
+     * @stable ICU 2.0
+     */
+    int32_t
+      getSize(void) const;
+
+    /**
+     * returns a string from a string resource type
+     *
+     * @param status  fills in the outgoing error code
+     *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+     *                could be a warning
+     *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+     * @return a pointer to a zero-terminated char16_t array which lives in a memory mapped/DLL file.
+     * @stable ICU 2.0
+     */
+    UnicodeString
+      getString(UErrorCode& status) const;
+
+    /**
+     * returns a binary data from a resource. Can be used at most primitive resource types (binaries,
+     * strings, ints)
+     *
+     * @param len     fills in the length of resulting byte chunk
+     * @param status  fills in the outgoing error code
+     *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+     *                could be a warning
+     *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+     * @return a pointer to a chunk of unsigned bytes which live in a memory mapped/DLL file.
+     * @stable ICU 2.0
+     */
+    const uint8_t*
+      getBinary(int32_t& len, UErrorCode& status) const;
+
+
+    /**
+     * returns an integer vector from a resource.
+     *
+     * @param len     fills in the length of resulting integer vector
+     * @param status  fills in the outgoing error code
+     *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+     *                could be a warning
+     *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+     * @return a pointer to a vector of integers that lives in a memory mapped/DLL file.
+     * @stable ICU 2.0
+     */
+    const int32_t*
+      getIntVector(int32_t& len, UErrorCode& status) const;
+
+    /**
+     * returns an unsigned integer from a resource.
+     * This integer is originally 28 bits.
+     *
+     * @param status  fills in the outgoing error code
+     *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+     *                could be a warning
+     *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+     * @return an unsigned integer value
+     * @stable ICU 2.0
+     */
+    uint32_t
+      getUInt(UErrorCode& status) const;
+
+    /**
+     * returns a signed integer from a resource.
+     * This integer is originally 28 bit and the sign gets propagated.
+     *
+     * @param status  fills in the outgoing error code
+     *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+     *                could be a warning
+     *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+     * @return a signed integer value
+     * @stable ICU 2.0
+     */
+    int32_t
+      getInt(UErrorCode& status) const;
+
+    /**
+     * Checks whether the resource has another element to iterate over.
+     *
+     * @return true if there are more elements, false if there is no more elements
+     * @stable ICU 2.0
+     */
+    UBool
+      hasNext(void) const;
+
+    /**
+     * Resets the internal context of a resource so that iteration starts from the first element.
+     *
+     * @stable ICU 2.0
+     */
+    void
+      resetIterator(void);
+
+    /**
+     * Returns the key associated with this resource. Not all the resources have a key - only
+     * those that are members of a table.
+     *
+     * @return a key associated to this resource, or NULL if it doesn't have a key
+     * @stable ICU 2.0
+     */
+    const char*
+      getKey(void) const;
+
+    /**
+     * Gets the locale ID of the resource bundle as a string.
+     * Same as getLocale().getName() .
+     *
+     * @return the locale ID of the resource bundle as a string
+     * @stable ICU 2.0
+     */
+    const char*
+      getName(void) const;
+
+
+    /**
+     * Returns the type of a resource. Available types are defined in enum UResType
+     *
+     * @return type of the given resource.
+     * @stable ICU 2.0
+     */
+    UResType
+      getType(void) const;
+
+    /**
+     * Returns the next resource in a given resource or NULL if there are no more resources
+     *
+     * @param status            fills in the outgoing error code
+     * @return                  ResourceBundle object.
+     * @stable ICU 2.0
+     */
+    ResourceBundle
+      getNext(UErrorCode& status);
+
+    /**
+     * Returns the next string in a resource or NULL if there are no more resources
+     * to iterate over.
+     *
+     * @param status            fills in the outgoing error code
+     * @return an UnicodeString object.
+     * @stable ICU 2.0
+     */
+    UnicodeString
+      getNextString(UErrorCode& status);
+
+    /**
+     * Returns the next string in a resource or NULL if there are no more resources
+     * to iterate over.
+     *
+     * @param key               fill in for key associated with this string
+     * @param status            fills in the outgoing error code
+     * @return an UnicodeString object.
+     * @stable ICU 2.0
+     */
+    UnicodeString
+      getNextString(const char ** key,
+                    UErrorCode& status);
+
+    /**
+     * Returns the resource in a resource at the specified index.
+     *
+     * @param index             an index to the wanted resource.
+     * @param status            fills in the outgoing error code
+     * @return                  ResourceBundle object. If there is an error, resource is invalid.
+     * @stable ICU 2.0
+     */
+    ResourceBundle
+      get(int32_t index,
+          UErrorCode& status) const;
+
+    /**
+     * Returns the string in a given resource at the specified index.
+     *
+     * @param index             an index to the wanted string.
+     * @param status            fills in the outgoing error code
+     * @return                  an UnicodeString object. If there is an error, string is bogus
+     * @stable ICU 2.0
+     */
+    UnicodeString
+      getStringEx(int32_t index,
+                  UErrorCode& status) const;
+
+    /**
+     * Returns a resource in a resource that has a given key. This procedure works only with table
+     * resources.
+     *
+     * @param key               a key associated with the wanted resource
+     * @param status            fills in the outgoing error code.
+     * @return                  ResourceBundle object. If there is an error, resource is invalid.
+     * @stable ICU 2.0
+     */
+    ResourceBundle
+      get(const char* key,
+          UErrorCode& status) const;
+
+    /**
+     * Returns a string in a resource that has a given key. This procedure works only with table
+     * resources.
+     *
+     * @param key               a key associated with the wanted string
+     * @param status            fills in the outgoing error code
+     * @return                  an UnicodeString object. If there is an error, string is bogus
+     * @stable ICU 2.0
+     */
+    UnicodeString
+      getStringEx(const char* key,
+                  UErrorCode& status) const;
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Return the version number associated with this ResourceBundle as a string. Please
+     * use getVersion, as this method is going to be deprecated.
+     *
+     * @return  A version number string as specified in the resource bundle or its parent.
+     *          The caller does not own this string.
+     * @see getVersion
+     * @deprecated ICU 2.8 Use getVersion instead.
+     */
+    const char*
+      getVersionNumber(void) const;
+#endif  /* U_HIDE_DEPRECATED_API */
+
+    /**
+     * Return the version number associated with this ResourceBundle as a UVersionInfo array.
+     *
+     * @param versionInfo A UVersionInfo array that is filled with the version number
+     *                    as specified in the resource bundle or its parent.
+     * @stable ICU 2.0
+     */
+    void
+      getVersion(UVersionInfo versionInfo) const;
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Return the Locale associated with this ResourceBundle.
+     *
+     * @return a Locale object
+     * @deprecated ICU 2.8 Use getLocale(ULocDataLocaleType type, UErrorCode &status) overload instead.
+     */
+    const Locale&
+      getLocale(void) const;
+#endif  /* U_HIDE_DEPRECATED_API */
+
+    /**
+     * Return the Locale associated with this ResourceBundle.
+     * @param type You can choose between requested, valid and actual
+     *             locale. For description see the definition of
+     *             ULocDataLocaleType in uloc.h
+     * @param status just for catching illegal arguments
+     *
+     * @return a Locale object
+     * @stable ICU 2.8
+     */
+    const Locale
+      getLocale(ULocDataLocaleType type, UErrorCode &status) const;
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * This API implements multilevel fallback
+     * @internal
+     */
+    ResourceBundle
+        getWithFallback(const char* key, UErrorCode& status);
+#endif  /* U_HIDE_INTERNAL_API */
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     *
+     * @stable ICU 2.2
+     */
+    virtual UClassID getDynamicClassID() const;
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.2
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+private:
+    ResourceBundle(); // default constructor not implemented
+
+    UResourceBundle *fResource;
+    void constructForLocale(const UnicodeString& path, const Locale& locale, UErrorCode& error);
+    Locale *fLocale;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/schriter.h b/thirdparty/icu4c/common/unicode/schriter.h
new file mode 100644
index 0000000000..1ca5b70fca
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/schriter.h
@@ -0,0 +1,195 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1998-2005, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File schriter.h
+*
+* Modification History:
+*
+*   Date        Name        Description
+*  05/05/99     stephen     Cleaned up.
+******************************************************************************
+*/
+
+#ifndef SCHRITER_H
+#define SCHRITER_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/chariter.h"
+#include "unicode/uchriter.h"
+
+/**
+ * \file 
+ * \brief C++ API: String Character Iterator
+ */
+ 
+U_NAMESPACE_BEGIN
+/**
+ * A concrete subclass of CharacterIterator that iterates over the
+ * characters (code units or code points) in a UnicodeString.
+ * It's possible not only to create an
+ * iterator that iterates over an entire UnicodeString, but also to
+ * create one that iterates over only a subrange of a UnicodeString
+ * (iterators over different subranges of the same UnicodeString don't
+ * compare equal).
+ * @see CharacterIterator
+ * @see ForwardCharacterIterator
+ * @stable ICU 2.0
+ */
+class U_COMMON_API StringCharacterIterator : public UCharCharacterIterator {
+public:
+  /**
+   * Create an iterator over the UnicodeString referred to by "textStr".
+   * The UnicodeString object is copied.
+   * The iteration range is the whole string, and the starting position is 0.
+   * @param textStr The unicode string used to create an iterator
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator(const UnicodeString& textStr);
+
+  /**
+   * Create an iterator over the UnicodeString referred to by "textStr".
+   * The iteration range is the whole string, and the starting
+   * position is specified by "textPos".  If "textPos" is outside the valid
+   * iteration range, the behavior of this object is undefined.
+   * @param textStr The unicode string used to create an iterator
+   * @param textPos The starting position of the iteration
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator(const UnicodeString&    textStr,
+              int32_t              textPos);
+
+  /**
+   * Create an iterator over the UnicodeString referred to by "textStr".
+   * The UnicodeString object is copied.
+   * The iteration range begins with the code unit specified by
+   * "textBegin" and ends with the code unit BEFORE the code unit specified
+   * by "textEnd".  The starting position is specified by "textPos".  If
+   * "textBegin" and "textEnd" don't form a valid range on "text" (i.e.,
+   * textBegin >= textEnd or either is negative or greater than text.size()),
+   * or "textPos" is outside the range defined by "textBegin" and "textEnd",
+   * the behavior of this iterator is undefined.
+   * @param textStr    The unicode string used to create the StringCharacterIterator
+   * @param textBegin  The begin position of the iteration range
+   * @param textEnd    The end position of the iteration range
+   * @param textPos    The starting position of the iteration
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator(const UnicodeString&    textStr,
+              int32_t              textBegin,
+              int32_t              textEnd,
+              int32_t              textPos);
+
+  /**
+   * Copy constructor.  The new iterator iterates over the same range
+   * of the same string as "that", and its initial position is the
+   * same as "that"'s current position.
+   * The UnicodeString object in "that" is copied.
+   * @param that The StringCharacterIterator to be copied
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator(const StringCharacterIterator&  that);
+
+  /**
+   * Destructor.
+   * @stable ICU 2.0
+   */
+  virtual ~StringCharacterIterator();
+
+  /**
+   * Assignment operator.  *this is altered to iterate over the same
+   * range of the same string as "that", and refers to the same
+   * character within that string as "that" does.
+   * @param that The object to be copied.
+   * @return the newly created object.
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator&
+  operator=(const StringCharacterIterator&    that);
+
+  /**
+   * Returns true if the iterators iterate over the same range of the
+   * same string and are pointing at the same character.
+   * @param that The ForwardCharacterIterator to be compared for equality
+   * @return true if the iterators iterate over the same range of the
+   * same string and are pointing at the same character.
+   * @stable ICU 2.0
+   */
+  virtual UBool          operator==(const ForwardCharacterIterator& that) const;
+
+  /**
+   * Returns a new StringCharacterIterator referring to the same
+   * character in the same range of the same string as this one.  The
+   * caller must delete the new iterator.
+   * @return the newly cloned object.
+   * @stable ICU 2.0
+   */
+  virtual StringCharacterIterator* clone() const;
+
+  /**
+   * Sets the iterator to iterate over the provided string.
+   * @param newText The string to be iterated over
+   * @stable ICU 2.0
+   */
+  void setText(const UnicodeString& newText);
+
+  /**
+   * Copies the UnicodeString under iteration into the UnicodeString
+   * referred to by "result".  Even if this iterator iterates across
+   * only a part of this string, the whole string is copied.
+   * @param result Receives a copy of the text under iteration.
+   * @stable ICU 2.0
+   */
+  virtual void            getText(UnicodeString& result);
+
+  /**
+   * Return a class ID for this object (not really public)
+   * @return a class ID for this object.
+   * @stable ICU 2.0
+   */
+  virtual UClassID         getDynamicClassID(void) const;
+
+  /**
+   * Return a class ID for this class (not really public)
+   * @return a class ID for this class
+   * @stable ICU 2.0
+   */
+  static UClassID   U_EXPORT2 getStaticClassID(void);
+
+protected:
+  /**
+   * Default constructor, iteration over empty string.
+   * @stable ICU 2.0
+   */
+  StringCharacterIterator();
+
+  /**
+   * Sets the iterator to iterate over the provided string.
+   * @param newText The string to be iterated over
+   * @param newTextLength The length of the String
+   * @stable ICU 2.0
+   */
+  void setText(const char16_t* newText, int32_t newTextLength);
+
+  /**
+   * Copy of the iterated string object.
+   * @stable ICU 2.0
+   */
+  UnicodeString            text;
+
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/simpleformatter.h b/thirdparty/icu4c/common/unicode/simpleformatter.h
new file mode 100644
index 0000000000..6d9c04ace2
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/simpleformatter.h
@@ -0,0 +1,341 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2014-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+******************************************************************************
+* simpleformatter.h
+*/
+
+#ifndef __SIMPLEFORMATTER_H__
+#define __SIMPLEFORMATTER_H__
+
+/**
+ * \file
+ * \brief C++ API: Simple formatter, minimal subset of MessageFormat.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/unistr.h"
+
+U_NAMESPACE_BEGIN
+
+// Forward declaration:
+namespace number {
+namespace impl {
+class SimpleModifier;
+}
+}
+
+/**
+ * Formats simple patterns like "{1} was born in {0}".
+ * Minimal subset of MessageFormat; fast, simple, minimal dependencies.
+ * Supports only numbered arguments with no type nor style parameters,
+ * and formats only string values.
+ * Quoting via ASCII apostrophe compatible with ICU MessageFormat default behavior.
+ *
+ * Factory methods set error codes for syntax errors
+ * and for too few or too many arguments/placeholders.
+ *
+ * SimpleFormatter objects are thread-safe except for assignment and applying new patterns.
+ *
+ * Example:
+ * <pre>
+ * UErrorCode errorCode = U_ZERO_ERROR;
+ * SimpleFormatter fmt("{1} '{born}' in {0}", errorCode);
+ * UnicodeString result;
+ *
+ * // Output: "paul {born} in england"
+ * fmt.format("england", "paul", result, errorCode);
+ * </pre>
+ *
+ * This class is not intended for public subclassing.
+ *
+ * @see MessageFormat
+ * @see UMessagePatternApostropheMode
+ * @stable ICU 57
+ */
+class U_COMMON_API SimpleFormatter U_FINAL : public UMemory {
+public:
+    /**
+     * Default constructor.
+     * @stable ICU 57
+     */
+    SimpleFormatter() : compiledPattern((char16_t)0) {}
+
+    /**
+     * Constructs a formatter from the pattern string.
+     *
+     * @param pattern The pattern string.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     *                  Set to U_ILLEGAL_ARGUMENT_ERROR for bad argument syntax.
+     * @stable ICU 57
+     */
+    SimpleFormatter(const UnicodeString& pattern, UErrorCode &errorCode) {
+        applyPattern(pattern, errorCode);
+    }
+
+    /**
+     * Constructs a formatter from the pattern string.
+     * The number of arguments checked against the given limits is the
+     * highest argument number plus one, not the number of occurrences of arguments.
+     *
+     * @param pattern The pattern string.
+     * @param min The pattern must have at least this many arguments.
+     * @param max The pattern must have at most this many arguments.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     *                  Set to U_ILLEGAL_ARGUMENT_ERROR for bad argument syntax and
+     *                  too few or too many arguments.
+     * @stable ICU 57
+     */
+    SimpleFormatter(const UnicodeString& pattern, int32_t min, int32_t max,
+                    UErrorCode &errorCode) {
+        applyPatternMinMaxArguments(pattern, min, max, errorCode);
+    }
+
+    /**
+     * Copy constructor.
+     * @stable ICU 57
+     */
+    SimpleFormatter(const SimpleFormatter& other)
+            : compiledPattern(other.compiledPattern) {}
+
+    /**
+     * Assignment operator.
+     * @stable ICU 57
+     */
+    SimpleFormatter &operator=(const SimpleFormatter& other);
+
+    /**
+     * Destructor.
+     * @stable ICU 57
+     */
+    ~SimpleFormatter();
+
+    /**
+     * Changes this object according to the new pattern.
+     *
+     * @param pattern The pattern string.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     *                  Set to U_ILLEGAL_ARGUMENT_ERROR for bad argument syntax.
+     * @return true if U_SUCCESS(errorCode).
+     * @stable ICU 57
+     */
+    UBool applyPattern(const UnicodeString &pattern, UErrorCode &errorCode) {
+        return applyPatternMinMaxArguments(pattern, 0, INT32_MAX, errorCode);
+    }
+
+    /**
+     * Changes this object according to the new pattern.
+     * The number of arguments checked against the given limits is the
+     * highest argument number plus one, not the number of occurrences of arguments.
+     *
+     * @param pattern The pattern string.
+     * @param min The pattern must have at least this many arguments.
+     * @param max The pattern must have at most this many arguments.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     *                  Set to U_ILLEGAL_ARGUMENT_ERROR for bad argument syntax and
+     *                  too few or too many arguments.
+     * @return true if U_SUCCESS(errorCode).
+     * @stable ICU 57
+     */
+    UBool applyPatternMinMaxArguments(const UnicodeString &pattern,
+                                      int32_t min, int32_t max, UErrorCode &errorCode);
+
+    /**
+     * @return The max argument number + 1.
+     * @stable ICU 57
+     */
+    int32_t getArgumentLimit() const {
+        return getArgumentLimit(compiledPattern.getBuffer(), compiledPattern.length());
+    }
+
+    /**
+     * Formats the given value, appending to the appendTo builder.
+     * The argument value must not be the same object as appendTo.
+     * getArgumentLimit() must be at most 1.
+     *
+     * @param value0 Value for argument {0}.
+     * @param appendTo Gets the formatted pattern and value appended.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     * @return appendTo
+     * @stable ICU 57
+     */
+    UnicodeString &format(
+            const UnicodeString &value0,
+            UnicodeString &appendTo, UErrorCode &errorCode) const;
+
+    /**
+     * Formats the given values, appending to the appendTo builder.
+     * An argument value must not be the same object as appendTo.
+     * getArgumentLimit() must be at most 2.
+     *
+     * @param value0 Value for argument {0}.
+     * @param value1 Value for argument {1}.
+     * @param appendTo Gets the formatted pattern and values appended.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     * @return appendTo
+     * @stable ICU 57
+     */
+    UnicodeString &format(
+            const UnicodeString &value0,
+            const UnicodeString &value1,
+            UnicodeString &appendTo, UErrorCode &errorCode) const;
+
+    /**
+     * Formats the given values, appending to the appendTo builder.
+     * An argument value must not be the same object as appendTo.
+     * getArgumentLimit() must be at most 3.
+     *
+     * @param value0 Value for argument {0}.
+     * @param value1 Value for argument {1}.
+     * @param value2 Value for argument {2}.
+     * @param appendTo Gets the formatted pattern and values appended.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     * @return appendTo
+     * @stable ICU 57
+     */
+    UnicodeString &format(
+            const UnicodeString &value0,
+            const UnicodeString &value1,
+            const UnicodeString &value2,
+            UnicodeString &appendTo, UErrorCode &errorCode) const;
+
+    /**
+     * Formats the given values, appending to the appendTo string.
+     *
+     * @param values The argument values.
+     *               An argument value must not be the same object as appendTo.
+     *               Can be NULL if valuesLength==getArgumentLimit()==0.
+     * @param valuesLength The length of the values array.
+     *                     Must be at least getArgumentLimit().
+     * @param appendTo Gets the formatted pattern and values appended.
+     * @param offsets offsets[i] receives the offset of where
+     *                values[i] replaced pattern argument {i}.
+     *                Can be shorter or longer than values. Can be NULL if offsetsLength==0.
+     *                If there is no {i} in the pattern, then offsets[i] is set to -1.
+     * @param offsetsLength The length of the offsets array.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     * @return appendTo
+     * @stable ICU 57
+     */
+    UnicodeString &formatAndAppend(
+            const UnicodeString *const *values, int32_t valuesLength,
+            UnicodeString &appendTo,
+            int32_t *offsets, int32_t offsetsLength, UErrorCode &errorCode) const;
+
+    /**
+     * Formats the given values, replacing the contents of the result string.
+     * May optimize by actually appending to the result if it is the same object
+     * as the value corresponding to the initial argument in the pattern.
+     *
+     * @param values The argument values.
+     *               An argument value may be the same object as result.
+     *               Can be NULL if valuesLength==getArgumentLimit()==0.
+     * @param valuesLength The length of the values array.
+     *                     Must be at least getArgumentLimit().
+     * @param result Gets its contents replaced by the formatted pattern and values.
+     * @param offsets offsets[i] receives the offset of where
+     *                values[i] replaced pattern argument {i}.
+     *                Can be shorter or longer than values. Can be NULL if offsetsLength==0.
+     *                If there is no {i} in the pattern, then offsets[i] is set to -1.
+     * @param offsetsLength The length of the offsets array.
+     * @param errorCode ICU error code in/out parameter.
+     *                  Must fulfill U_SUCCESS before the function call.
+     * @return result
+     * @stable ICU 57
+     */
+    UnicodeString &formatAndReplace(
+            const UnicodeString *const *values, int32_t valuesLength,
+            UnicodeString &result,
+            int32_t *offsets, int32_t offsetsLength, UErrorCode &errorCode) const;
+
+    /**
+     * Returns the pattern text with none of the arguments.
+     * Like formatting with all-empty string values.
+     * @stable ICU 57
+     */
+    UnicodeString getTextWithNoArguments() const {
+        return getTextWithNoArguments(
+            compiledPattern.getBuffer(),
+            compiledPattern.length(),
+            nullptr,
+            0);
+    }
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Returns the pattern text with none of the arguments.
+     * Like formatting with all-empty string values.
+     *
+     * TODO(ICU-20406): Replace this with an Iterator interface.
+     *
+     * @param offsets offsets[i] receives the offset of where {i} was located
+     *                before it was replaced by an empty string.
+     *                For example, "a{0}b{1}" produces offset 1 for i=0 and 2 for i=1.
+     *                Can be nullptr if offsetsLength==0.
+     *                If there is no {i} in the pattern, then offsets[i] is set to -1.
+     * @param offsetsLength The length of the offsets array.
+     *
+     * @internal
+     */
+    UnicodeString getTextWithNoArguments(int32_t *offsets, int32_t offsetsLength) const {
+        return getTextWithNoArguments(
+            compiledPattern.getBuffer(),
+            compiledPattern.length(),
+            offsets,
+            offsetsLength);
+    }
+#endif // U_HIDE_INTERNAL_API
+
+private:
+    /**
+     * Binary representation of the compiled pattern.
+     * Index 0: One more than the highest argument number.
+     * Followed by zero or more arguments or literal-text segments.
+     *
+     * An argument is stored as its number, less than ARG_NUM_LIMIT.
+     * A literal-text segment is stored as its length (at least 1) offset by ARG_NUM_LIMIT,
+     * followed by that many chars.
+     */
+    UnicodeString compiledPattern;
+
+    static inline int32_t getArgumentLimit(const char16_t *compiledPattern,
+                                              int32_t compiledPatternLength) {
+        return compiledPatternLength == 0 ? 0 : compiledPattern[0];
+    }
+
+    static UnicodeString getTextWithNoArguments(
+        const char16_t *compiledPattern,
+        int32_t compiledPatternLength,
+        int32_t *offsets,
+        int32_t offsetsLength);
+
+    static UnicodeString &format(
+            const char16_t *compiledPattern, int32_t compiledPatternLength,
+            const UnicodeString *const *values,
+            UnicodeString &result, const UnicodeString *resultCopy, UBool forbidResultAsValue,
+            int32_t *offsets, int32_t offsetsLength,
+            UErrorCode &errorCode);
+
+    // Give access to internals to SimpleModifier for number formatting
+    friend class number::impl::SimpleModifier;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __SIMPLEFORMATTER_H__
diff --git a/thirdparty/icu4c/common/unicode/std_string.h b/thirdparty/icu4c/common/unicode/std_string.h
new file mode 100644
index 0000000000..bf87230167
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/std_string.h
@@ -0,0 +1,41 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  std_string.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009feb19
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __STD_STRING_H__
+#define __STD_STRING_H__
+
+/**
+ * \file
+ * \brief C++ API: Central ICU header for including the C++ standard &lt;string&gt;
+ *                 header and for related definitions.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+// Workaround for a libstdc++ bug before libstdc++4.6 (2011).
+// https://bugs.llvm.org/show_bug.cgi?id=13364
+#if defined(__GLIBCXX__)
+namespace std { class type_info; }
+#endif
+#include <string>
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __STD_STRING_H__
diff --git a/thirdparty/icu4c/common/unicode/strenum.h b/thirdparty/icu4c/common/unicode/strenum.h
new file mode 100644
index 0000000000..df72b4b7e8
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/strenum.h
@@ -0,0 +1,281 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*/
+
+#ifndef STRENUM_H
+#define STRENUM_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+
+/**
+ * \file 
+ * \brief C++ API: String Enumeration
+ */
+ 
+U_NAMESPACE_BEGIN
+
+/**
+ * Base class for 'pure' C++ implementations of uenum api.  Adds a
+ * method that returns the next UnicodeString since in C++ this can
+ * be a common storage format for strings.
+ *
+ * <p>The model is that the enumeration is over strings maintained by
+ * a 'service.'  At any point, the service might change, invalidating
+ * the enumerator (though this is expected to be rare).  The iterator
+ * returns an error if this has occurred.  Lack of the error is no
+ * guarantee that the service didn't change immediately after the
+ * call, so the returned string still might not be 'valid' on
+ * subsequent use.</p>
+ *
+ * <p>Strings may take the form of const char*, const char16_t*, or const
+ * UnicodeString*.  The type you get is determine by the variant of
+ * 'next' that you call.  In general the StringEnumeration is
+ * optimized for one of these types, but all StringEnumerations can
+ * return all types.  Returned strings are each terminated with a NUL.
+ * Depending on the service data, they might also include embedded NUL
+ * characters, so API is provided to optionally return the true
+ * length, counting the embedded NULs but not counting the terminating
+ * NUL.</p>
+ *
+ * <p>The pointers returned by next, unext, and snext become invalid
+ * upon any subsequent call to the enumeration's destructor, next,
+ * unext, snext, or reset.</p>
+ *
+ * ICU 2.8 adds some default implementations and helper functions
+ * for subclasses.
+ *
+ * @stable ICU 2.4 
+ */
+class U_COMMON_API StringEnumeration : public UObject { 
+public:
+    /**
+     * Destructor.
+     * @stable ICU 2.4
+     */
+    virtual ~StringEnumeration();
+
+    /**
+     * Clone this object, an instance of a subclass of StringEnumeration.
+     * Clones can be used concurrently in multiple threads.
+     * If a subclass does not implement clone(), or if an error occurs,
+     * then NULL is returned.
+     * The caller must delete the clone.
+     *
+     * @return a clone of this object
+     *
+     * @see getDynamicClassID
+     * @stable ICU 2.8
+     */
+    virtual StringEnumeration *clone() const;
+
+    /**
+     * <p>Return the number of elements that the iterator traverses.  If
+     * the iterator is out of sync with its service, status is set to
+     * U_ENUM_OUT_OF_SYNC_ERROR, and the return value is zero.</p>
+     *
+     * <p>The return value will not change except possibly as a result of
+     * a subsequent call to reset, or if the iterator becomes out of sync.</p>
+     *
+     * <p>This is a convenience function. It can end up being very
+     * expensive as all the items might have to be pre-fetched
+     * (depending on the storage format of the data being
+     * traversed).</p>
+     *
+     * @param status the error code.
+     * @return number of elements in the iterator.
+     *
+     * @stable ICU 2.4 */
+    virtual int32_t count(UErrorCode& status) const = 0;
+
+    /**
+     * <p>Returns the next element as a NUL-terminated char*.  If there
+     * are no more elements, returns NULL.  If the resultLength pointer
+     * is not NULL, the length of the string (not counting the
+     * terminating NUL) is returned at that address.  If an error
+     * status is returned, the value at resultLength is undefined.</p>
+     *
+     * <p>The returned pointer is owned by this iterator and must not be
+     * deleted by the caller.  The pointer is valid until the next call
+     * to next, unext, snext, reset, or the enumerator's destructor.</p>
+     *
+     * <p>If the iterator is out of sync with its service, status is set
+     * to U_ENUM_OUT_OF_SYNC_ERROR and NULL is returned.</p>
+     *
+     * <p>If the native service string is a char16_t* string, it is
+     * converted to char* with the invariant converter.  If the
+     * conversion fails (because a character cannot be converted) then
+     * status is set to U_INVARIANT_CONVERSION_ERROR and the return
+     * value is undefined (though not NULL).</p>
+     *
+     * Starting with ICU 2.8, the default implementation calls snext()
+     * and handles the conversion.
+     * Either next() or snext() must be implemented differently by a subclass.
+     *
+     * @param status the error code.
+     * @param resultLength a pointer to receive the length, can be NULL.
+     * @return a pointer to the string, or NULL.
+     *
+     * @stable ICU 2.4 
+     */
+    virtual const char* next(int32_t *resultLength, UErrorCode& status);
+
+    /**
+     * <p>Returns the next element as a NUL-terminated char16_t*.  If there
+     * are no more elements, returns NULL.  If the resultLength pointer
+     * is not NULL, the length of the string (not counting the
+     * terminating NUL) is returned at that address.  If an error
+     * status is returned, the value at resultLength is undefined.</p>
+     *
+     * <p>The returned pointer is owned by this iterator and must not be
+     * deleted by the caller.  The pointer is valid until the next call
+     * to next, unext, snext, reset, or the enumerator's destructor.</p>
+     *
+     * <p>If the iterator is out of sync with its service, status is set
+     * to U_ENUM_OUT_OF_SYNC_ERROR and NULL is returned.</p>
+     *
+     * Starting with ICU 2.8, the default implementation calls snext()
+     * and handles the conversion.
+     *
+     * @param status the error code.
+     * @param resultLength a ponter to receive the length, can be NULL.
+     * @return a pointer to the string, or NULL.
+     *
+     * @stable ICU 2.4 
+     */
+    virtual const char16_t* unext(int32_t *resultLength, UErrorCode& status);
+
+    /**
+     * <p>Returns the next element a UnicodeString*.  If there are no
+     * more elements, returns NULL.</p>
+     *
+     * <p>The returned pointer is owned by this iterator and must not be
+     * deleted by the caller.  The pointer is valid until the next call
+     * to next, unext, snext, reset, or the enumerator's destructor.</p>
+     *
+     * <p>If the iterator is out of sync with its service, status is set
+     * to U_ENUM_OUT_OF_SYNC_ERROR and NULL is returned.</p>
+     *
+     * Starting with ICU 2.8, the default implementation calls next()
+     * and handles the conversion.
+     * Either next() or snext() must be implemented differently by a subclass.
+     *
+     * @param status the error code.
+     * @return a pointer to the string, or NULL.
+     *
+     * @stable ICU 2.4 
+     */
+    virtual const UnicodeString* snext(UErrorCode& status);
+
+    /**
+     * <p>Resets the iterator.  This re-establishes sync with the
+     * service and rewinds the iterator to start at the first
+     * element.</p>
+     *
+     * <p>Previous pointers returned by next, unext, or snext become
+     * invalid, and the value returned by count might change.</p>
+     *
+     * @param status the error code.
+     *
+     * @stable ICU 2.4 
+     */
+    virtual void reset(UErrorCode& status) = 0;
+
+    /**
+     * Compares this enumeration to other to check if both are equal
+     *
+     * @param that The other string enumeration to compare this object to
+     * @return true if the enumerations are equal. false if not.
+     * @stable ICU 3.6 
+     */
+    virtual UBool operator==(const StringEnumeration& that)const;
+    /**
+     * Compares this enumeration to other to check if both are not equal
+     *
+     * @param that The other string enumeration to compare this object to
+     * @return true if the enumerations are equal. false if not.
+     * @stable ICU 3.6 
+     */
+    virtual UBool operator!=(const StringEnumeration& that)const;
+
+protected:
+    /**
+     * UnicodeString field for use with default implementations and subclasses.
+     * @stable ICU 2.8
+     */
+    UnicodeString unistr;
+    /**
+     * char * default buffer for use with default implementations and subclasses.
+     * @stable ICU 2.8
+     */
+    char charsBuffer[32];
+    /**
+     * char * buffer for use with default implementations and subclasses.
+     * Allocated in constructor and in ensureCharsCapacity().
+     * @stable ICU 2.8
+     */
+    char *chars;
+    /**
+     * Capacity of chars, for use with default implementations and subclasses.
+     * @stable ICU 2.8
+     */
+    int32_t charsCapacity;
+
+    /**
+     * Default constructor for use with default implementations and subclasses.
+     * @stable ICU 2.8
+     */
+    StringEnumeration();
+
+    /**
+     * Ensures that chars is at least as large as the requested capacity.
+     * For use with default implementations and subclasses.
+     *
+     * @param capacity Requested capacity.
+     * @param status ICU in/out error code.
+     * @stable ICU 2.8
+     */
+    void ensureCharsCapacity(int32_t capacity, UErrorCode &status);
+
+    /**
+     * Converts s to Unicode and sets unistr to the result.
+     * For use with default implementations and subclasses,
+     * especially for implementations of snext() in terms of next().
+     * This is provided with a helper function instead of a default implementation
+     * of snext() to avoid potential infinite loops between next() and snext().
+     *
+     * For example:
+     * \code
+     * const UnicodeString* snext(UErrorCode& status) {
+     *   int32_t resultLength=0;
+     *   const char *s=next(&resultLength, status);
+     *   return setChars(s, resultLength, status);
+     * }
+     * \endcode
+     *
+     * @param s String to be converted to Unicode.
+     * @param length Length of the string.
+     * @param status ICU in/out error code.
+     * @return A pointer to unistr.
+     * @stable ICU 2.8
+     */
+    UnicodeString *setChars(const char *s, int32_t length, UErrorCode &status);
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+/* STRENUM_H */
+#endif
diff --git a/thirdparty/icu4c/common/unicode/stringoptions.h b/thirdparty/icu4c/common/unicode/stringoptions.h
new file mode 100644
index 0000000000..7b9f70944f
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/stringoptions.h
@@ -0,0 +1,190 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// stringoptions.h
+// created: 2017jun08 Markus W. Scherer
+
+#ifndef __STRINGOPTIONS_H__
+#define __STRINGOPTIONS_H__
+
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief C API: Bit set option bit constants for various string and character processing functions.
+ */
+
+/**
+ * Option value for case folding: Use default mappings defined in CaseFolding.txt.
+ *
+ * @stable ICU 2.0
+ */
+#define U_FOLD_CASE_DEFAULT 0
+
+/**
+ * Option value for case folding:
+ *
+ * Use the modified set of mappings provided in CaseFolding.txt to handle dotted I
+ * and dotless i appropriately for Turkic languages (tr, az).
+ *
+ * Before Unicode 3.2, CaseFolding.txt contains mappings marked with 'I' that
+ * are to be included for default mappings and
+ * excluded for the Turkic-specific mappings.
+ *
+ * Unicode 3.2 CaseFolding.txt instead contains mappings marked with 'T' that
+ * are to be excluded for default mappings and
+ * included for the Turkic-specific mappings.
+ *
+ * @stable ICU 2.0
+ */
+#define U_FOLD_CASE_EXCLUDE_SPECIAL_I 1
+
+/**
+ * Titlecase the string as a whole rather than each word.
+ * (Titlecase only the character at index 0, possibly adjusted.)
+ * Option bits value for titlecasing APIs that take an options bit set.
+ *
+ * It is an error to specify multiple titlecasing iterator options together,
+ * including both an options bit and an explicit BreakIterator.
+ *
+ * @see U_TITLECASE_ADJUST_TO_CASED
+ * @stable ICU 60
+ */
+#define U_TITLECASE_WHOLE_STRING 0x20
+
+/**
+ * Titlecase sentences rather than words.
+ * (Titlecase only the first character of each sentence, possibly adjusted.)
+ * Option bits value for titlecasing APIs that take an options bit set.
+ *
+ * It is an error to specify multiple titlecasing iterator options together,
+ * including both an options bit and an explicit BreakIterator.
+ *
+ * @see U_TITLECASE_ADJUST_TO_CASED
+ * @stable ICU 60
+ */
+#define U_TITLECASE_SENTENCES 0x40
+
+/**
+ * Do not lowercase non-initial parts of words when titlecasing.
+ * Option bit for titlecasing APIs that take an options bit set.
+ *
+ * By default, titlecasing will titlecase the character at each
+ * (possibly adjusted) BreakIterator index and
+ * lowercase all other characters up to the next iterator index.
+ * With this option, the other characters will not be modified.
+ *
+ * @see U_TITLECASE_ADJUST_TO_CASED
+ * @see UnicodeString::toTitle
+ * @see CaseMap::toTitle
+ * @see ucasemap_setOptions
+ * @see ucasemap_toTitle
+ * @see ucasemap_utf8ToTitle
+ * @stable ICU 3.8
+ */
+#define U_TITLECASE_NO_LOWERCASE 0x100
+
+/**
+ * Do not adjust the titlecasing BreakIterator indexes;
+ * titlecase exactly the characters at breaks from the iterator.
+ * Option bit for titlecasing APIs that take an options bit set.
+ *
+ * By default, titlecasing will take each break iterator index,
+ * adjust it to the next relevant character (see U_TITLECASE_ADJUST_TO_CASED),
+ * and titlecase that one.
+ *
+ * Other characters are lowercased.
+ *
+ * It is an error to specify multiple titlecasing adjustment options together.
+ *
+ * @see U_TITLECASE_ADJUST_TO_CASED
+ * @see U_TITLECASE_NO_LOWERCASE
+ * @see UnicodeString::toTitle
+ * @see CaseMap::toTitle
+ * @see ucasemap_setOptions
+ * @see ucasemap_toTitle
+ * @see ucasemap_utf8ToTitle
+ * @stable ICU 3.8
+ */
+#define U_TITLECASE_NO_BREAK_ADJUSTMENT 0x200
+
+/**
+ * Adjust each titlecasing BreakIterator index to the next cased character.
+ * (See the Unicode Standard, chapter 3, Default Case Conversion, R3 toTitlecase(X).)
+ * Option bit for titlecasing APIs that take an options bit set.
+ *
+ * This used to be the default index adjustment in ICU.
+ * Since ICU 60, the default index adjustment is to the next character that is
+ * a letter, number, symbol, or private use code point.
+ * (Uncased modifier letters are skipped.)
+ * The difference in behavior is small for word titlecasing,
+ * but the new adjustment is much better for whole-string and sentence titlecasing:
+ * It yields "49ers" and "«丰(abc)»" instead of "49Ers" and "«丰(Abc)»".
+ *
+ * It is an error to specify multiple titlecasing adjustment options together.
+ *
+ * @see U_TITLECASE_NO_BREAK_ADJUSTMENT
+ * @stable ICU 60
+ */
+#define U_TITLECASE_ADJUST_TO_CASED 0x400
+
+/**
+ * Option for string transformation functions to not first reset the Edits object.
+ * Used for example in some case-mapping and normalization functions.
+ *
+ * @see CaseMap
+ * @see Edits
+ * @see Normalizer2
+ * @stable ICU 60
+ */
+#define U_EDITS_NO_RESET 0x2000
+
+/**
+ * Omit unchanged text when recording how source substrings
+ * relate to changed and unchanged result substrings.
+ * Used for example in some case-mapping and normalization functions.
+ *
+ * @see CaseMap
+ * @see Edits
+ * @see Normalizer2
+ * @stable ICU 60
+ */
+#define U_OMIT_UNCHANGED_TEXT 0x4000
+
+/**
+ * Option bit for u_strCaseCompare, u_strcasecmp, unorm_compare, etc:
+ * Compare strings in code point order instead of code unit order.
+ * @stable ICU 2.2
+ */
+#define U_COMPARE_CODE_POINT_ORDER  0x8000
+
+/**
+ * Option bit for unorm_compare:
+ * Perform case-insensitive comparison.
+ * @stable ICU 2.2
+ */
+#define U_COMPARE_IGNORE_CASE       0x10000
+
+/**
+ * Option bit for unorm_compare:
+ * Both input strings are assumed to fulfill FCD conditions.
+ * @stable ICU 2.2
+ */
+#define UNORM_INPUT_IS_FCD          0x20000
+
+// Related definitions elsewhere.
+// Options that are not meaningful in the same functions
+// can share the same bits.
+//
+// Public:
+// unicode/unorm.h #define UNORM_COMPARE_NORM_OPTIONS_SHIFT 20
+//
+// Internal: (may change or be removed)
+// ucase.h #define _STRCASECMP_OPTIONS_MASK 0xffff
+// ucase.h #define _FOLD_CASE_OPTIONS_MASK 7
+// ucasemap_imp.h #define U_TITLECASE_ITERATOR_MASK 0xe0
+// ucasemap_imp.h #define U_TITLECASE_ADJUSTMENT_MASK 0x600
+// ustr_imp.h #define _STRNCMP_STYLE 0x1000
+// unormcmp.cpp #define _COMPARE_EQUIV 0x80000
+
+#endif  // __STRINGOPTIONS_H__
diff --git a/thirdparty/icu4c/common/unicode/stringpiece.h b/thirdparty/icu4c/common/unicode/stringpiece.h
new file mode 100644
index 0000000000..7d7d871e1f
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/stringpiece.h
@@ -0,0 +1,353 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+// Copyright (C) 2009-2013, International Business Machines
+// Corporation and others. All Rights Reserved.
+//
+// Copyright 2001 and onwards Google Inc.
+// Author: Sanjay Ghemawat
+
+// This code is a contribution of Google code, and the style used here is
+// a compromise between the original Google code and the ICU coding guidelines.
+// For example, data types are ICU-ified (size_t,int->int32_t),
+// and API comments doxygen-ified, but function names and behavior are
+// as in the original, if possible.
+// Assertion-style error handling, not available in ICU, was changed to
+// parameter "pinning" similar to UnicodeString.
+//
+// In addition, this is only a partial port of the original Google code,
+// limited to what was needed so far. The (nearly) complete original code
+// is in the ICU svn repository at icuhtml/trunk/design/strings/contrib
+// (see ICU ticket 6765, r25517).
+
+#ifndef __STRINGPIECE_H__
+#define __STRINGPIECE_H__
+
+/**
+ * \file 
+ * \brief C++ API: StringPiece: Read-only byte string wrapper class.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include <cstddef>
+#include <type_traits>
+
+#include "unicode/uobject.h"
+#include "unicode/std_string.h"
+
+// Arghh!  I wish C++ literals were "string".
+
+U_NAMESPACE_BEGIN
+
+/**
+ * A string-like object that points to a sized piece of memory.
+ *
+ * We provide non-explicit singleton constructors so users can pass
+ * in a "const char*" or a "string" wherever a "StringPiece" is
+ * expected.
+ *
+ * Functions or methods may use StringPiece parameters to accept either a
+ * "const char*" or a "string" value that will be implicitly converted to a
+ * StringPiece.
+ *
+ * Systematic usage of StringPiece is encouraged as it will reduce unnecessary
+ * conversions from "const char*" to "string" and back again.
+ *
+ * @stable ICU 4.2
+ */
+class U_COMMON_API StringPiece : public UMemory {
+ private:
+  const char*   ptr_;
+  int32_t       length_;
+
+ public:
+  /**
+   * Default constructor, creates an empty StringPiece.
+   * @stable ICU 4.2
+   */
+  StringPiece() : ptr_(nullptr), length_(0) { }
+
+  /**
+   * Constructs from a NUL-terminated const char * pointer.
+   * @param str a NUL-terminated const char * pointer
+   * @stable ICU 4.2
+   */
+  StringPiece(const char* str);
+#ifndef U_HIDE_DRAFT_API
+#if defined(__cpp_char8_t) || defined(U_IN_DOXYGEN)
+  /**
+   * Constructs from a NUL-terminated const char8_t * pointer.
+   * @param str a NUL-terminated const char8_t * pointer
+   * @draft ICU 67
+   */
+  StringPiece(const char8_t* str) : StringPiece(reinterpret_cast<const char*>(str)) {}
+#endif
+  /**
+   * Constructs an empty StringPiece.
+   * Needed for type disambiguation from multiple other overloads.
+   * @param p nullptr
+   * @draft ICU 67
+   */
+  StringPiece(std::nullptr_t p) : ptr_(p), length_(0) {}
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Constructs from a std::string.
+   * @stable ICU 4.2
+   */
+  StringPiece(const std::string& str)
+    : ptr_(str.data()), length_(static_cast<int32_t>(str.size())) { }
+#ifndef U_HIDE_DRAFT_API
+#if defined(__cpp_lib_char8_t) || defined(U_IN_DOXYGEN)
+  /**
+   * Constructs from a std::u8string.
+   * @draft ICU 67
+   */
+  StringPiece(const std::u8string& str)
+    : ptr_(reinterpret_cast<const char*>(str.data())),
+      length_(static_cast<int32_t>(str.size())) { }
+#endif
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Constructs from some other implementation of a string piece class, from any
+   * C++ record type that has these two methods:
+   *
+   * \code{.cpp}
+   *
+   *   struct OtherStringPieceClass {
+   *     const char* data();  // or const char8_t*
+   *     size_t size();
+   *   };
+   *
+   * \endcode
+   *
+   * The other string piece class will typically be std::string_view from C++17
+   * or absl::string_view from Abseil.
+   *
+   * Starting with C++20, data() may also return a const char8_t* pointer,
+   * as from std::u8string_view.
+   *
+   * @param str the other string piece
+   * @stable ICU 65
+   */
+  template <typename T,
+            typename = typename std::enable_if<
+                (std::is_same<decltype(T().data()), const char*>::value
+#if defined(__cpp_char8_t)
+                    || std::is_same<decltype(T().data()), const char8_t*>::value
+#endif
+                ) &&
+                std::is_same<decltype(T().size()), size_t>::value>::type>
+  StringPiece(T str)
+      : ptr_(reinterpret_cast<const char*>(str.data())),
+        length_(static_cast<int32_t>(str.size())) {}
+
+  /**
+   * Constructs from a const char * pointer and a specified length.
+   * @param offset a const char * pointer (need not be terminated)
+   * @param len the length of the string; must be non-negative
+   * @stable ICU 4.2
+   */
+  StringPiece(const char* offset, int32_t len) : ptr_(offset), length_(len) { }
+#ifndef U_HIDE_DRAFT_API
+#if defined(__cpp_char8_t) || defined(U_IN_DOXYGEN)
+  /**
+   * Constructs from a const char8_t * pointer and a specified length.
+   * @param str a const char8_t * pointer (need not be terminated)
+   * @param len the length of the string; must be non-negative
+   * @draft ICU 67
+   */
+  StringPiece(const char8_t* str, int32_t len) :
+      StringPiece(reinterpret_cast<const char*>(str), len) {}
+#endif
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Substring of another StringPiece.
+   * @param x the other StringPiece
+   * @param pos start position in x; must be non-negative and <= x.length().
+   * @stable ICU 4.2
+   */
+  StringPiece(const StringPiece& x, int32_t pos);
+  /**
+   * Substring of another StringPiece.
+   * @param x the other StringPiece
+   * @param pos start position in x; must be non-negative and <= x.length().
+   * @param len length of the substring;
+   *            must be non-negative and will be pinned to at most x.length() - pos.
+   * @stable ICU 4.2
+   */
+  StringPiece(const StringPiece& x, int32_t pos, int32_t len);
+
+  /**
+   * Returns the string pointer. May be nullptr if it is empty.
+   *
+   * data() may return a pointer to a buffer with embedded NULs, and the
+   * returned buffer may or may not be null terminated.  Therefore it is
+   * typically a mistake to pass data() to a routine that expects a NUL
+   * terminated string.
+   * @return the string pointer
+   * @stable ICU 4.2
+   */
+  const char* data() const { return ptr_; }
+  /**
+   * Returns the string length. Same as length().
+   * @return the string length
+   * @stable ICU 4.2
+   */
+  int32_t size() const { return length_; }
+  /**
+   * Returns the string length. Same as size().
+   * @return the string length
+   * @stable ICU 4.2
+   */
+  int32_t length() const { return length_; }
+  /**
+   * Returns whether the string is empty.
+   * @return true if the string is empty
+   * @stable ICU 4.2
+   */
+  UBool empty() const { return length_ == 0; }
+
+  /**
+   * Sets to an empty string.
+   * @stable ICU 4.2
+   */
+  void clear() { ptr_ = nullptr; length_ = 0; }
+
+  /**
+   * Reset the stringpiece to refer to new data.
+   * @param xdata pointer the new string data.  Need not be nul terminated.
+   * @param len the length of the new data
+   * @stable ICU 4.8
+   */
+  void set(const char* xdata, int32_t len) { ptr_ = xdata; length_ = len; }
+
+  /**
+   * Reset the stringpiece to refer to new data.
+   * @param str a pointer to a NUL-terminated string. 
+   * @stable ICU 4.8
+   */
+  void set(const char* str);
+
+#ifndef U_HIDE_DRAFT_API
+#if defined(__cpp_char8_t) || defined(U_IN_DOXYGEN)
+  /**
+   * Resets the stringpiece to refer to new data.
+   * @param xdata pointer the new string data. Need not be NUL-terminated.
+   * @param len the length of the new data
+   * @draft ICU 67
+   */
+  inline void set(const char8_t* xdata, int32_t len) {
+      set(reinterpret_cast<const char*>(xdata), len);
+  }
+
+  /**
+   * Resets the stringpiece to refer to new data.
+   * @param str a pointer to a NUL-terminated string.
+   * @draft ICU 67
+   */
+  inline void set(const char8_t* str) {
+      set(reinterpret_cast<const char*>(str));
+  }
+#endif
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Removes the first n string units.
+   * @param n prefix length, must be non-negative and <=length()
+   * @stable ICU 4.2
+   */
+  void remove_prefix(int32_t n) {
+    if (n >= 0) {
+      if (n > length_) {
+        n = length_;
+      }
+      ptr_ += n;
+      length_ -= n;
+    }
+  }
+
+  /**
+   * Removes the last n string units.
+   * @param n suffix length, must be non-negative and <=length()
+   * @stable ICU 4.2
+   */
+  void remove_suffix(int32_t n) {
+    if (n >= 0) {
+      if (n <= length_) {
+        length_ -= n;
+      } else {
+        length_ = 0;
+      }
+    }
+  }
+
+#ifndef U_HIDE_DRAFT_API
+  /**
+   * Searches the StringPiece for the given search string (needle);
+   * @param needle The string for which to search.
+   * @param offset Where to start searching within this string (haystack).
+   * @return The offset of needle in haystack, or -1 if not found.
+   * @draft ICU 67
+   */
+  int32_t find(StringPiece needle, int32_t offset);
+
+  /**
+   * Compares this StringPiece with the other StringPiece, with semantics
+   * similar to std::string::compare().
+   * @param other The string to compare to.
+   * @return below zero if this < other; above zero if this > other; 0 if this == other.
+   * @draft ICU 67
+   */
+  int32_t compare(StringPiece other);
+#endif  // U_HIDE_DRAFT_API
+
+  /**
+   * Maximum integer, used as a default value for substring methods.
+   * @stable ICU 4.2
+   */
+  static const int32_t npos; // = 0x7fffffff;
+
+  /**
+   * Returns a substring of this StringPiece.
+   * @param pos start position; must be non-negative and <= length().
+   * @param len length of the substring;
+   *            must be non-negative and will be pinned to at most length() - pos.
+   * @return the substring StringPiece
+   * @stable ICU 4.2
+   */
+  StringPiece substr(int32_t pos, int32_t len = npos) const {
+    return StringPiece(*this, pos, len);
+  }
+};
+
+/**
+ * Global operator == for StringPiece
+ * @param x The first StringPiece to compare.
+ * @param y The second StringPiece to compare.
+ * @return true if the string data is equal
+ * @stable ICU 4.8
+ */
+U_EXPORT UBool U_EXPORT2 
+operator==(const StringPiece& x, const StringPiece& y);
+
+/**
+ * Global operator != for StringPiece
+ * @param x The first StringPiece to compare.
+ * @param y The second StringPiece to compare.
+ * @return true if the string data is not equal
+ * @stable ICU 4.8
+ */
+inline UBool operator!=(const StringPiece& x, const StringPiece& y) {
+  return !(x == y);
+}
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __STRINGPIECE_H__
diff --git a/thirdparty/icu4c/common/unicode/stringtriebuilder.h b/thirdparty/icu4c/common/unicode/stringtriebuilder.h
new file mode 100644
index 0000000000..fe471bbbf9
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/stringtriebuilder.h
@@ -0,0 +1,426 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012,2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  stringtriebuilder.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010dec24
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __STRINGTRIEBUILDER_H__
+#define __STRINGTRIEBUILDER_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file
+ * \brief C++ API: Builder API for trie builders
+ */
+
+// Forward declaration.
+/// \cond
+struct UHashtable;
+typedef struct UHashtable UHashtable;
+/// \endcond
+
+/**
+ * Build options for BytesTrieBuilder and CharsTrieBuilder.
+ * @stable ICU 4.8
+ */
+enum UStringTrieBuildOption {
+    /**
+     * Builds a trie quickly.
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_BUILD_FAST,
+    /**
+     * Builds a trie more slowly, attempting to generate
+     * a shorter but equivalent serialization.
+     * This build option also uses more memory.
+     *
+     * This option can be effective when many integer values are the same
+     * and string/byte sequence suffixes can be shared.
+     * Runtime speed is not expected to improve.
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_BUILD_SMALL
+};
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Base class for string trie builder classes.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API StringTrieBuilder : public UObject {
+public:
+#ifndef U_HIDE_INTERNAL_API
+    /** @internal */
+    static int32_t hashNode(const void *node);
+    /** @internal */
+    static UBool equalNodes(const void *left, const void *right);
+#endif  /* U_HIDE_INTERNAL_API */
+
+protected:
+    // Do not enclose the protected default constructor with #ifndef U_HIDE_INTERNAL_API
+    // or else the compiler will create a public default constructor.
+    /** @internal */
+    StringTrieBuilder();
+    /** @internal */
+    virtual ~StringTrieBuilder();
+
+#ifndef U_HIDE_INTERNAL_API
+    /** @internal */
+    void createCompactBuilder(int32_t sizeGuess, UErrorCode &errorCode);
+    /** @internal */
+    void deleteCompactBuilder();
+
+    /** @internal */
+    void build(UStringTrieBuildOption buildOption, int32_t elementsLength, UErrorCode &errorCode);
+
+    /** @internal */
+    int32_t writeNode(int32_t start, int32_t limit, int32_t unitIndex);
+    /** @internal */
+    int32_t writeBranchSubNode(int32_t start, int32_t limit, int32_t unitIndex, int32_t length);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    class Node;
+
+#ifndef U_HIDE_INTERNAL_API
+    /** @internal */
+    Node *makeNode(int32_t start, int32_t limit, int32_t unitIndex, UErrorCode &errorCode);
+    /** @internal */
+    Node *makeBranchSubNode(int32_t start, int32_t limit, int32_t unitIndex,
+                            int32_t length, UErrorCode &errorCode);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /** @internal */
+    virtual int32_t getElementStringLength(int32_t i) const = 0;
+    /** @internal */
+    virtual char16_t getElementUnit(int32_t i, int32_t unitIndex) const = 0;
+    /** @internal */
+    virtual int32_t getElementValue(int32_t i) const = 0;
+
+    // Finds the first unit index after this one where
+    // the first and last element have different units again.
+    /** @internal */
+    virtual int32_t getLimitOfLinearMatch(int32_t first, int32_t last, int32_t unitIndex) const = 0;
+
+    // Number of different units at unitIndex.
+    /** @internal */
+    virtual int32_t countElementUnits(int32_t start, int32_t limit, int32_t unitIndex) const = 0;
+    /** @internal */
+    virtual int32_t skipElementsBySomeUnits(int32_t i, int32_t unitIndex, int32_t count) const = 0;
+    /** @internal */
+    virtual int32_t indexOfElementWithNextUnit(int32_t i, int32_t unitIndex, char16_t unit) const = 0;
+
+    /** @internal */
+    virtual UBool matchNodesCanHaveValues() const = 0;
+
+    /** @internal */
+    virtual int32_t getMaxBranchLinearSubNodeLength() const = 0;
+    /** @internal */
+    virtual int32_t getMinLinearMatch() const = 0;
+    /** @internal */
+    virtual int32_t getMaxLinearMatchLength() const = 0;
+
+#ifndef U_HIDE_INTERNAL_API
+    // max(BytesTrie::kMaxBranchLinearSubNodeLength, UCharsTrie::kMaxBranchLinearSubNodeLength).
+    /** @internal */
+    static const int32_t kMaxBranchLinearSubNodeLength=5;
+
+    // Maximum number of nested split-branch levels for a branch on all 2^16 possible char16_t units.
+    // log2(2^16/kMaxBranchLinearSubNodeLength) rounded up.
+    /** @internal */
+    static const int32_t kMaxSplitBranchLevels=14;
+
+    /**
+     * Makes sure that there is only one unique node registered that is
+     * equivalent to newNode.
+     * @param newNode Input node. The builder takes ownership.
+     * @param errorCode ICU in/out UErrorCode.
+                        Set to U_MEMORY_ALLOCATION_ERROR if it was success but newNode==NULL.
+     * @return newNode if it is the first of its kind, or
+     *         an equivalent node if newNode is a duplicate.
+     * @internal
+     */
+    Node *registerNode(Node *newNode, UErrorCode &errorCode);
+    /**
+     * Makes sure that there is only one unique FinalValueNode registered
+     * with this value.
+     * Avoids creating a node if the value is a duplicate.
+     * @param value A final value.
+     * @param errorCode ICU in/out UErrorCode.
+                        Set to U_MEMORY_ALLOCATION_ERROR if it was success but newNode==NULL.
+     * @return A FinalValueNode with the given value.
+     * @internal
+     */
+    Node *registerFinalValue(int32_t value, UErrorCode &errorCode);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /*
+     * C++ note:
+     * registerNode() and registerFinalValue() take ownership of their input nodes,
+     * and only return owned nodes.
+     * If they see a failure UErrorCode, they will delete the input node.
+     * If they get a NULL pointer, they will record a U_MEMORY_ALLOCATION_ERROR.
+     * If there is a failure, they return NULL.
+     *
+     * NULL Node pointers can be safely passed into other Nodes because
+     * they call the static Node::hashCode() which checks for a NULL pointer first.
+     *
+     * Therefore, as long as builder functions register a new node,
+     * they need to check for failures only before explicitly dereferencing
+     * a Node pointer, or before setting a new UErrorCode.
+     */
+
+    // Hash set of nodes, maps from nodes to integer 1.
+    /** @internal */
+    UHashtable *nodes;
+
+    // Do not conditionalize the following with #ifndef U_HIDE_INTERNAL_API,
+    // it is needed for layout of other objects.
+    /**
+     * @internal
+     * \cond
+     */
+    class Node : public UObject {
+    public:
+        Node(int32_t initialHash) : hash(initialHash), offset(0) {}
+        inline int32_t hashCode() const { return hash; }
+        // Handles node==NULL.
+        static inline int32_t hashCode(const Node *node) { return node==NULL ? 0 : node->hashCode(); }
+        // Base class operator==() compares the actual class types.
+        virtual UBool operator==(const Node &other) const;
+        inline UBool operator!=(const Node &other) const { return !operator==(other); }
+        /**
+         * Traverses the Node graph and numbers branch edges, with rightmost edges first.
+         * This is to avoid writing a duplicate node twice.
+         *
+         * Branch nodes in this trie data structure are not symmetric.
+         * Most branch edges "jump" to other nodes but the rightmost branch edges
+         * just continue without a jump.
+         * Therefore, write() must write the rightmost branch edge last
+         * (trie units are written backwards), and must write it at that point even if
+         * it is a duplicate of a node previously written elsewhere.
+         *
+         * This function visits and marks right branch edges first.
+         * Edges are numbered with increasingly negative values because we share the
+         * offset field which gets positive values when nodes are written.
+         * A branch edge also remembers the first number for any of its edges.
+         *
+         * When a further-left branch edge has a number in the range of the rightmost
+         * edge's numbers, then it will be written as part of the required right edge
+         * and we can avoid writing it first.
+         *
+         * After root.markRightEdgesFirst(-1) the offsets of all nodes are negative
+         * edge numbers.
+         *
+         * @param edgeNumber The first edge number for this node and its sub-nodes.
+         * @return An edge number that is at least the maximum-negative
+         *         of the input edge number and the numbers of this node and all of its sub-nodes.
+         */
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+        // write() must set the offset to a positive value.
+        virtual void write(StringTrieBuilder &builder) = 0;
+        // See markRightEdgesFirst.
+        inline void writeUnlessInsideRightEdge(int32_t firstRight, int32_t lastRight,
+                                               StringTrieBuilder &builder) {
+            // Note: Edge numbers are negative, lastRight<=firstRight.
+            // If offset>0 then this node and its sub-nodes have been written already
+            // and we need not write them again.
+            // If this node is part of the unwritten right branch edge,
+            // then we wait until that is written.
+            if(offset<0 && (offset<lastRight || firstRight<offset)) {
+                write(builder);
+            }
+        }
+        inline int32_t getOffset() const { return offset; }
+    protected:
+        int32_t hash;
+        int32_t offset;
+    };
+
+#ifndef U_HIDE_INTERNAL_API
+    // This class should not be overridden because
+    // registerFinalValue() compares a stack-allocated FinalValueNode
+    // (stack-allocated so that we don't unnecessarily create lots of duplicate nodes)
+    // with the input node, and the
+    // !Node::operator==(other) used inside FinalValueNode::operator==(other)
+    // will be false if the typeid's are different.
+    /** @internal */
+    class FinalValueNode : public Node {
+    public:
+        FinalValueNode(int32_t v) : Node(0x111111u*37u+v), value(v) {}
+        virtual UBool operator==(const Node &other) const;
+        virtual void write(StringTrieBuilder &builder);
+    protected:
+        int32_t value;
+    };
+#endif  /* U_HIDE_INTERNAL_API */
+
+    // Do not conditionalize the following with #ifndef U_HIDE_INTERNAL_API,
+    // it is needed for layout of other objects.
+    /**
+     * @internal 
+     */
+    class ValueNode : public Node {
+    public:
+        ValueNode(int32_t initialHash) : Node(initialHash), hasValue(false), value(0) {}
+        virtual UBool operator==(const Node &other) const;
+        void setValue(int32_t v) {
+            hasValue=true;
+            value=v;
+            hash=hash*37u+v;
+        }
+    protected:
+        UBool hasValue;
+        int32_t value;
+    };
+
+#ifndef U_HIDE_INTERNAL_API
+    /** 
+     * @internal 
+     */
+    class IntermediateValueNode : public ValueNode {
+    public:
+        IntermediateValueNode(int32_t v, Node *nextNode)
+                : ValueNode(0x222222u*37u+hashCode(nextNode)), next(nextNode) { setValue(v); }
+        virtual UBool operator==(const Node &other) const;
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+        virtual void write(StringTrieBuilder &builder);
+    protected:
+        Node *next;
+    };
+#endif  /* U_HIDE_INTERNAL_API */
+
+    // Do not conditionalize the following with #ifndef U_HIDE_INTERNAL_API,
+    // it is needed for layout of other objects.
+    /**
+     * @internal 
+     */
+    class LinearMatchNode : public ValueNode {
+    public:
+        LinearMatchNode(int32_t len, Node *nextNode)
+                : ValueNode((0x333333u*37u+len)*37u+hashCode(nextNode)),
+                  length(len), next(nextNode) {}
+        virtual UBool operator==(const Node &other) const;
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+    protected:
+        int32_t length;
+        Node *next;
+    };
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * @internal 
+     */
+    class BranchNode : public Node {
+    public:
+        BranchNode(int32_t initialHash) : Node(initialHash) {}
+    protected:
+        int32_t firstEdgeNumber;
+    };
+
+    /**
+     * @internal 
+     */
+    class ListBranchNode : public BranchNode {
+    public:
+        ListBranchNode() : BranchNode(0x444444), length(0) {}
+        virtual UBool operator==(const Node &other) const;
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+        virtual void write(StringTrieBuilder &builder);
+        // Adds a unit with a final value.
+        void add(int32_t c, int32_t value) {
+            units[length]=(char16_t)c;
+            equal[length]=NULL;
+            values[length]=value;
+            ++length;
+            hash=(hash*37u+c)*37u+value;
+        }
+        // Adds a unit which leads to another match node.
+        void add(int32_t c, Node *node) {
+            units[length]=(char16_t)c;
+            equal[length]=node;
+            values[length]=0;
+            ++length;
+            hash=(hash*37u+c)*37u+hashCode(node);
+        }
+    protected:
+        Node *equal[kMaxBranchLinearSubNodeLength];  // NULL means "has final value".
+        int32_t length;
+        int32_t values[kMaxBranchLinearSubNodeLength];
+        char16_t units[kMaxBranchLinearSubNodeLength];
+    };
+
+    /**
+     * @internal 
+     */
+    class SplitBranchNode : public BranchNode {
+    public:
+        SplitBranchNode(char16_t middleUnit, Node *lessThanNode, Node *greaterOrEqualNode)
+                : BranchNode(((0x555555u*37u+middleUnit)*37u+
+                              hashCode(lessThanNode))*37u+hashCode(greaterOrEqualNode)),
+                  unit(middleUnit), lessThan(lessThanNode), greaterOrEqual(greaterOrEqualNode) {}
+        virtual UBool operator==(const Node &other) const;
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+        virtual void write(StringTrieBuilder &builder);
+    protected:
+        char16_t unit;
+        Node *lessThan;
+        Node *greaterOrEqual;
+    };
+
+    // Branch head node, for writing the actual node lead unit.
+    /** @internal */
+    class BranchHeadNode : public ValueNode {
+    public:
+        BranchHeadNode(int32_t len, Node *subNode)
+                : ValueNode((0x666666u*37u+len)*37u+hashCode(subNode)),
+                  length(len), next(subNode) {}
+        virtual UBool operator==(const Node &other) const;
+        virtual int32_t markRightEdgesFirst(int32_t edgeNumber);
+        virtual void write(StringTrieBuilder &builder);
+    protected:
+        int32_t length;
+        Node *next;  // A branch sub-node.
+    };
+
+#endif  /* U_HIDE_INTERNAL_API */
+    /// \endcond
+
+    /** @internal */
+    virtual Node *createLinearMatchNode(int32_t i, int32_t unitIndex, int32_t length,
+                                        Node *nextNode) const = 0;
+
+    /** @internal */
+    virtual int32_t write(int32_t unit) = 0;
+    /** @internal */
+    virtual int32_t writeElementUnits(int32_t i, int32_t unitIndex, int32_t length) = 0;
+    /** @internal */
+    virtual int32_t writeValueAndFinal(int32_t i, UBool isFinal) = 0;
+    /** @internal */
+    virtual int32_t writeValueAndType(UBool hasValue, int32_t value, int32_t node) = 0;
+    /** @internal */
+    virtual int32_t writeDeltaTo(int32_t jumpTarget) = 0;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __STRINGTRIEBUILDER_H__
diff --git a/thirdparty/icu4c/common/unicode/symtable.h b/thirdparty/icu4c/common/unicode/symtable.h
new file mode 100644
index 0000000000..b64d877f97
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/symtable.h
@@ -0,0 +1,119 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2000-2005, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   02/04/00    aliu        Creation.
+**********************************************************************
+*/
+#ifndef SYMTABLE_H
+#define SYMTABLE_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file 
+ * \brief C++ API: An interface that defines both lookup protocol and parsing of
+ * symbolic names.
+ */
+ 
+U_NAMESPACE_BEGIN
+
+class ParsePosition;
+class UnicodeFunctor;
+class UnicodeSet;
+class UnicodeString;
+
+/**
+ * An interface that defines both lookup protocol and parsing of
+ * symbolic names.
+ *
+ * <p>A symbol table maintains two kinds of mappings.  The first is
+ * between symbolic names and their values.  For example, if the
+ * variable with the name "start" is set to the value "alpha"
+ * (perhaps, though not necessarily, through an expression such as
+ * "$start=alpha"), then the call lookup("start") will return the
+ * char[] array ['a', 'l', 'p', 'h', 'a'].
+ *
+ * <p>The second kind of mapping is between character values and
+ * UnicodeMatcher objects.  This is used by RuleBasedTransliterator,
+ * which uses characters in the private use area to represent objects
+ * such as UnicodeSets.  If U+E015 is mapped to the UnicodeSet [a-z],
+ * then lookupMatcher(0xE015) will return the UnicodeSet [a-z].
+ *
+ * <p>Finally, a symbol table defines parsing behavior for symbolic
+ * names.  All symbolic names start with the SYMBOL_REF character.
+ * When a parser encounters this character, it calls parseReference()
+ * with the position immediately following the SYMBOL_REF.  The symbol
+ * table parses the name, if there is one, and returns it.
+ *
+ * @stable ICU 2.8
+ */
+class U_COMMON_API SymbolTable /* not : public UObject because this is an interface/mixin class */ {
+public:
+
+    /**
+     * The character preceding a symbol reference name.
+     * @stable ICU 2.8
+     */
+    enum { SYMBOL_REF = 0x0024 /*$*/ };
+
+    /**
+     * Destructor.
+     * @stable ICU 2.8
+     */
+    virtual ~SymbolTable();
+
+    /**
+     * Lookup the characters associated with this string and return it.
+     * Return <tt>NULL</tt> if no such name exists.  The resultant
+     * string may have length zero.
+     * @param s the symbolic name to lookup
+     * @return a string containing the name's value, or <tt>NULL</tt> if
+     * there is no mapping for s.
+     * @stable ICU 2.8
+     */
+    virtual const UnicodeString* lookup(const UnicodeString& s) const = 0;
+
+    /**
+     * Lookup the UnicodeMatcher associated with the given character, and
+     * return it.  Return <tt>NULL</tt> if not found.
+     * @param ch a 32-bit code point from 0 to 0x10FFFF inclusive.
+     * @return the UnicodeMatcher object represented by the given
+     * character, or NULL if there is no mapping for ch.
+     * @stable ICU 2.8
+     */
+    virtual const UnicodeFunctor* lookupMatcher(UChar32 ch) const = 0;
+
+    /**
+     * Parse a symbol reference name from the given string, starting
+     * at the given position.  If no valid symbol reference name is
+     * found, return the empty string and leave pos unchanged.  That is, if the
+     * character at pos cannot start a name, or if pos is at or after
+     * text.length(), then return an empty string.  This indicates an
+     * isolated SYMBOL_REF character.
+     * @param text the text to parse for the name
+     * @param pos on entry, the index of the first character to parse.
+     * This is the character following the SYMBOL_REF character.  On
+     * exit, the index after the last parsed character.  If the parse
+     * failed, pos is unchanged on exit.
+     * @param limit the index after the last character to be parsed.
+     * @return the parsed name, or an empty string if there is no
+     * valid symbolic name at the given position.
+     * @stable ICU 2.8
+     */
+    virtual UnicodeString parseReference(const UnicodeString& text,
+                                         ParsePosition& pos, int32_t limit) const = 0;
+};
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ubidi.h b/thirdparty/icu4c/common/unicode/ubidi.h
new file mode 100644
index 0000000000..63d0e45cb7
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ubidi.h
@@ -0,0 +1,2210 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ubidi.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999jul27
+*   created by: Markus W. Scherer, updated by Matitiahu Allouche
+*/
+
+#ifndef UBIDI_H
+#define UBIDI_H
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ *\file
+ * \brief C API: Bidi algorithm
+ *
+ * <h2>Bidi algorithm for ICU</h2>
+ *
+ * This is an implementation of the Unicode Bidirectional Algorithm.
+ * The algorithm is defined in the
+ * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>.<p>
+ *
+ * Note: Libraries that perform a bidirectional algorithm and
+ * reorder strings accordingly are sometimes called "Storage Layout Engines".
+ * ICU's Bidi and shaping (u_shapeArabic()) APIs can be used at the core of such
+ * "Storage Layout Engines".
+ *
+ * <h3>General remarks about the API:</h3>
+ *
+ * In functions with an error code parameter,
+ * the <code>pErrorCode</code> pointer must be valid
+ * and the value that it points to must not indicate a failure before
+ * the function call. Otherwise, the function returns immediately.
+ * After the function call, the value indicates success or failure.<p>
+ *
+ * The &quot;limit&quot; of a sequence of characters is the position just after their
+ * last character, i.e., one more than that position.<p>
+ *
+ * Some of the API functions provide access to &quot;runs&quot;.
+ * Such a &quot;run&quot; is defined as a sequence of characters
+ * that are at the same embedding level
+ * after performing the Bidi algorithm.<p>
+ *
+ * @author Markus W. Scherer
+ * @version 1.0
+ *
+ *
+ * <h4> Sample code for the ICU Bidi API </h4>
+ *
+ * <h5>Rendering a paragraph with the ICU Bidi API</h5>
+ *
+ * This is (hypothetical) sample code that illustrates
+ * how the ICU Bidi API could be used to render a paragraph of text.
+ * Rendering code depends highly on the graphics system,
+ * therefore this sample code must make a lot of assumptions,
+ * which may or may not match any existing graphics system's properties.
+ *
+ * <p>The basic assumptions are:</p>
+ * <ul>
+ * <li>Rendering is done from left to right on a horizontal line.</li>
+ * <li>A run of single-style, unidirectional text can be rendered at once.</li>
+ * <li>Such a run of text is passed to the graphics system with
+ *     characters (code units) in logical order.</li>
+ * <li>The line-breaking algorithm is very complicated
+ *     and Locale-dependent -
+ *     and therefore its implementation omitted from this sample code.</li>
+ * </ul>
+ *
+ * <pre>
+ * \code
+ *#include "unicode/ubidi.h"
+ *
+ *typedef enum {
+ *     styleNormal=0, styleSelected=1,
+ *     styleBold=2, styleItalics=4,
+ *     styleSuper=8, styleSub=16
+ *} Style;
+ *
+ *typedef struct { int32_t limit; Style style; } StyleRun;
+ *
+ *int getTextWidth(const UChar *text, int32_t start, int32_t limit,
+ *                  const StyleRun *styleRuns, int styleRunCount);
+ *
+ * // set *pLimit and *pStyleRunLimit for a line
+ * // from text[start] and from styleRuns[styleRunStart]
+ * // using ubidi_getLogicalRun(para, ...)
+ *void getLineBreak(const UChar *text, int32_t start, int32_t *pLimit,
+ *                  UBiDi *para,
+ *                  const StyleRun *styleRuns, int styleRunStart, int *pStyleRunLimit,
+ *                  int *pLineWidth);
+ *
+ * // render runs on a line sequentially, always from left to right
+ *
+ * // prepare rendering a new line
+ * void startLine(UBiDiDirection textDirection, int lineWidth);
+ *
+ * // render a run of text and advance to the right by the run width
+ * // the text[start..limit-1] is always in logical order
+ * void renderRun(const UChar *text, int32_t start, int32_t limit,
+ *               UBiDiDirection textDirection, Style style);
+ *
+ * // We could compute a cross-product
+ * // from the style runs with the directional runs
+ * // and then reorder it.
+ * // Instead, here we iterate over each run type
+ * // and render the intersections -
+ * // with shortcuts in simple (and common) cases.
+ * // renderParagraph() is the main function.
+ *
+ * // render a directional run with
+ * // (possibly) multiple style runs intersecting with it
+ * void renderDirectionalRun(const UChar *text,
+ *                           int32_t start, int32_t limit,
+ *                           UBiDiDirection direction,
+ *                           const StyleRun *styleRuns, int styleRunCount) {
+ *     int i;
+ *
+ *     // iterate over style runs
+ *     if(direction==UBIDI_LTR) {
+ *         int styleLimit;
+ *
+ *         for(i=0; i<styleRunCount; ++i) {
+ *             styleLimit=styleRun[i].limit;
+ *             if(start<styleLimit) {
+ *                 if(styleLimit>limit) { styleLimit=limit; }
+ *                 renderRun(text, start, styleLimit,
+ *                           direction, styleRun[i].style);
+ *                 if(styleLimit==limit) { break; }
+ *                 start=styleLimit;
+ *             }
+ *         }
+ *     } else {
+ *         int styleStart;
+ *
+ *         for(i=styleRunCount-1; i>=0; --i) {
+ *             if(i>0) {
+ *                 styleStart=styleRun[i-1].limit;
+ *             } else {
+ *                 styleStart=0;
+ *             }
+ *             if(limit>=styleStart) {
+ *                 if(styleStart<start) { styleStart=start; }
+ *                 renderRun(text, styleStart, limit,
+ *                           direction, styleRun[i].style);
+ *                 if(styleStart==start) { break; }
+ *                 limit=styleStart;
+ *             }
+ *         }
+ *     }
+ * }
+ *
+ * // the line object represents text[start..limit-1]
+ * void renderLine(UBiDi *line, const UChar *text,
+ *                 int32_t start, int32_t limit,
+ *                 const StyleRun *styleRuns, int styleRunCount) {
+ *     UBiDiDirection direction=ubidi_getDirection(line);
+ *     if(direction!=UBIDI_MIXED) {
+ *         // unidirectional
+ *         if(styleRunCount<=1) {
+ *             renderRun(text, start, limit, direction, styleRuns[0].style);
+ *         } else {
+ *             renderDirectionalRun(text, start, limit,
+ *                                  direction, styleRuns, styleRunCount);
+ *         }
+ *     } else {
+ *         // mixed-directional
+ *         int32_t count, i, length;
+ *         UBiDiLevel level;
+ *
+ *         count=ubidi_countRuns(para, pErrorCode);
+ *         if(U_SUCCESS(*pErrorCode)) {
+ *             if(styleRunCount<=1) {
+ *                 Style style=styleRuns[0].style;
+ *
+ *                 // iterate over directional runs
+ *                for(i=0; i<count; ++i) {
+ *                    direction=ubidi_getVisualRun(para, i, &start, &length);
+ *                     renderRun(text, start, start+length, direction, style);
+ *                }
+ *             } else {
+ *                 int32_t j;
+ *
+ *                 // iterate over both directional and style runs
+ *                 for(i=0; i<count; ++i) {
+ *                     direction=ubidi_getVisualRun(line, i, &start, &length);
+ *                     renderDirectionalRun(text, start, start+length,
+ *                                          direction, styleRuns, styleRunCount);
+ *                 }
+ *             }
+ *         }
+ *     }
+ * }
+ *
+ *void renderParagraph(const UChar *text, int32_t length,
+ *                     UBiDiDirection textDirection,
+ *                      const StyleRun *styleRuns, int styleRunCount,
+ *                      int lineWidth,
+ *                      UErrorCode *pErrorCode) {
+ *     UBiDi *para;
+ *
+ *     if(pErrorCode==NULL || U_FAILURE(*pErrorCode) || length<=0) {
+ *         return;
+ *     }
+ *
+ *     para=ubidi_openSized(length, 0, pErrorCode);
+ *     if(para==NULL) { return; }
+ *
+ *     ubidi_setPara(para, text, length,
+ *                   textDirection ? UBIDI_DEFAULT_RTL : UBIDI_DEFAULT_LTR,
+ *                   NULL, pErrorCode);
+ *     if(U_SUCCESS(*pErrorCode)) {
+ *         UBiDiLevel paraLevel=1&ubidi_getParaLevel(para);
+ *         StyleRun styleRun={ length, styleNormal };
+ *         int width;
+ *
+ *         if(styleRuns==NULL || styleRunCount<=0) {
+ *            styleRunCount=1;
+ *             styleRuns=&styleRun;
+ *         }
+ *
+ *        // assume styleRuns[styleRunCount-1].limit>=length
+ *
+ *         width=getTextWidth(text, 0, length, styleRuns, styleRunCount);
+ *         if(width<=lineWidth) {
+ *             // everything fits onto one line
+ *
+ *            // prepare rendering a new line from either left or right
+ *             startLine(paraLevel, width);
+ *
+ *             renderLine(para, text, 0, length,
+ *                        styleRuns, styleRunCount);
+ *         } else {
+ *             UBiDi *line;
+ *
+ *             // we need to render several lines
+ *             line=ubidi_openSized(length, 0, pErrorCode);
+ *             if(line!=NULL) {
+ *                 int32_t start=0, limit;
+ *                 int styleRunStart=0, styleRunLimit;
+ *
+ *                 for(;;) {
+ *                     limit=length;
+ *                     styleRunLimit=styleRunCount;
+ *                     getLineBreak(text, start, &limit, para,
+ *                                  styleRuns, styleRunStart, &styleRunLimit,
+ *                                 &width);
+ *                     ubidi_setLine(para, start, limit, line, pErrorCode);
+ *                     if(U_SUCCESS(*pErrorCode)) {
+ *                         // prepare rendering a new line
+ *                         // from either left or right
+ *                         startLine(paraLevel, width);
+ *
+ *                         renderLine(line, text, start, limit,
+ *                                    styleRuns+styleRunStart,
+ *                                    styleRunLimit-styleRunStart);
+ *                     }
+ *                     if(limit==length) { break; }
+ *                     start=limit;
+ *                     styleRunStart=styleRunLimit-1;
+ *                     if(start>=styleRuns[styleRunStart].limit) {
+ *                         ++styleRunStart;
+ *                     }
+ *                 }
+ *
+ *                 ubidi_close(line);
+ *             }
+ *        }
+ *    }
+ *
+ *     ubidi_close(para);
+ *}
+ *\endcode
+ * </pre>
+ */
+
+/*DOCXX_TAG*/
+/*@{*/
+
+/**
+ * UBiDiLevel is the type of the level values in this
+ * Bidi implementation.
+ * It holds an embedding level and indicates the visual direction
+ * by its bit&nbsp;0 (even/odd value).<p>
+ *
+ * It can also hold non-level values for the
+ * <code>paraLevel</code> and <code>embeddingLevels</code>
+ * arguments of <code>ubidi_setPara()</code>; there:
+ * <ul>
+ * <li>bit&nbsp;7 of an <code>embeddingLevels[]</code>
+ * value indicates whether the using application is
+ * specifying the level of a character to <i>override</i> whatever the
+ * Bidi implementation would resolve it to.</li>
+ * <li><code>paraLevel</code> can be set to the
+ * pseudo-level values <code>UBIDI_DEFAULT_LTR</code>
+ * and <code>UBIDI_DEFAULT_RTL</code>.</li>
+ * </ul>
+ *
+ * @see ubidi_setPara
+ *
+ * <p>The related constants are not real, valid level values.
+ * <code>UBIDI_DEFAULT_XXX</code> can be used to specify
+ * a default for the paragraph level for
+ * when the <code>ubidi_setPara()</code> function
+ * shall determine it but there is no
+ * strongly typed character in the input.<p>
+ *
+ * Note that the value for <code>UBIDI_DEFAULT_LTR</code> is even
+ * and the one for <code>UBIDI_DEFAULT_RTL</code> is odd,
+ * just like with normal LTR and RTL level values -
+ * these special values are designed that way. Also, the implementation
+ * assumes that UBIDI_MAX_EXPLICIT_LEVEL is odd.
+ *
+ * Note: The numeric values of the related constants will not change:
+ * They are tied to the use of 7-bit byte values (plus the override bit)
+ * and of the UBiDiLevel=uint8_t data type in this API.
+ *
+ * @see UBIDI_DEFAULT_LTR
+ * @see UBIDI_DEFAULT_RTL
+ * @see UBIDI_LEVEL_OVERRIDE
+ * @see UBIDI_MAX_EXPLICIT_LEVEL
+ * @stable ICU 2.0
+ */
+typedef uint8_t UBiDiLevel;
+
+/** Paragraph level setting.<p>
+ *
+ * Constant indicating that the base direction depends on the first strong
+ * directional character in the text according to the Unicode Bidirectional
+ * Algorithm. If no strong directional character is present,
+ * then set the paragraph level to 0 (left-to-right).<p>
+ *
+ * If this value is used in conjunction with reordering modes
+ * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
+ * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
+ * is assumed to be visual LTR, and the text after reordering is required
+ * to be the corresponding logical string with appropriate contextual
+ * direction. The direction of the result string will be RTL if either
+ * the righmost or leftmost strong character of the source text is RTL
+ * or Arabic Letter, the direction will be LTR otherwise.<p>
+ *
+ * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
+ * be added at the beginning of the result string to ensure round trip
+ * (that the result string, when reordered back to visual, will produce
+ * the original source text).
+ * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
+ * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
+ * @stable ICU 2.0
+ */
+#define UBIDI_DEFAULT_LTR 0xfe
+
+/** Paragraph level setting.<p>
+ *
+ * Constant indicating that the base direction depends on the first strong
+ * directional character in the text according to the Unicode Bidirectional
+ * Algorithm. If no strong directional character is present,
+ * then set the paragraph level to 1 (right-to-left).<p>
+ *
+ * If this value is used in conjunction with reordering modes
+ * <code>UBIDI_REORDER_INVERSE_LIKE_DIRECT</code> or
+ * <code>UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the text to reorder
+ * is assumed to be visual LTR, and the text after reordering is required
+ * to be the corresponding logical string with appropriate contextual
+ * direction. The direction of the result string will be RTL if either
+ * the righmost or leftmost strong character of the source text is RTL
+ * or Arabic Letter, or if the text contains no strong character;
+ * the direction will be LTR otherwise.<p>
+ *
+ * If reordering option <code>UBIDI_OPTION_INSERT_MARKS</code> is set, an RLM may
+ * be added at the beginning of the result string to ensure round trip
+ * (that the result string, when reordered back to visual, will produce
+ * the original source text).
+ * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
+ * @see UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL
+ * @stable ICU 2.0
+ */
+#define UBIDI_DEFAULT_RTL 0xff
+
+/**
+ * Maximum explicit embedding level.
+ * Same as the max_depth value in the
+ * <a href="http://www.unicode.org/reports/tr9/#BD2">Unicode Bidirectional Algorithm</a>.
+ * (The maximum resolved level can be up to <code>UBIDI_MAX_EXPLICIT_LEVEL+1</code>).
+ * @stable ICU 2.0
+ */
+#define UBIDI_MAX_EXPLICIT_LEVEL 125
+
+/** Bit flag for level input.
+ *  Overrides directional properties.
+ * @stable ICU 2.0
+ */
+#define UBIDI_LEVEL_OVERRIDE 0x80
+
+/**
+ * Special value which can be returned by the mapping functions when a logical
+ * index has no corresponding visual index or vice-versa. This may happen
+ * for the logical-to-visual mapping of a Bidi control when option
+ * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is specified. This can also happen
+ * for the visual-to-logical mapping of a Bidi mark (LRM or RLM) inserted
+ * by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
+ * @see ubidi_getVisualIndex
+ * @see ubidi_getVisualMap
+ * @see ubidi_getLogicalIndex
+ * @see ubidi_getLogicalMap
+ * @stable ICU 3.6
+ */
+#define UBIDI_MAP_NOWHERE   (-1)
+
+/**
+ * <code>UBiDiDirection</code> values indicate the text direction.
+ * @stable ICU 2.0
+ */
+enum UBiDiDirection {
+  /** Left-to-right text. This is a 0 value.
+   * <ul>
+   * <li>As return value for <code>ubidi_getDirection()</code>, it means
+   *     that the source string contains no right-to-left characters, or
+   *     that the source string is empty and the paragraph level is even.
+   * <li> As return value for <code>ubidi_getBaseDirection()</code>, it
+   *      means that the first strong character of the source string has
+   *      a left-to-right direction.
+   * </ul>
+   * @stable ICU 2.0
+   */
+  UBIDI_LTR,
+  /** Right-to-left text. This is a 1 value.
+   * <ul>
+   * <li>As return value for <code>ubidi_getDirection()</code>, it means
+   *     that the source string contains no left-to-right characters, or
+   *     that the source string is empty and the paragraph level is odd.
+   * <li> As return value for <code>ubidi_getBaseDirection()</code>, it
+   *      means that the first strong character of the source string has
+   *      a right-to-left direction.
+   * </ul>
+   * @stable ICU 2.0
+   */
+  UBIDI_RTL,
+  /** Mixed-directional text.
+   * <p>As return value for <code>ubidi_getDirection()</code>, it means
+   *    that the source string contains both left-to-right and
+   *    right-to-left characters.
+   * @stable ICU 2.0
+   */
+  UBIDI_MIXED,
+  /** No strongly directional text.
+   * <p>As return value for <code>ubidi_getBaseDirection()</code>, it means
+   *    that the source string is missing or empty, or contains neither left-to-right
+   *    nor right-to-left characters.
+   * @stable ICU 4.6
+   */
+  UBIDI_NEUTRAL
+};
+
+/** @stable ICU 2.0 */
+typedef enum UBiDiDirection UBiDiDirection;
+
+/**
+ * Forward declaration of the <code>UBiDi</code> structure for the declaration of
+ * the API functions. Its fields are implementation-specific.<p>
+ * This structure holds information about a paragraph (or multiple paragraphs)
+ * of text with Bidi-algorithm-related details, or about one line of
+ * such a paragraph.<p>
+ * Reordering can be done on a line, or on one or more paragraphs which are
+ * then interpreted each as one single line.
+ * @stable ICU 2.0
+ */
+struct UBiDi;
+
+/** @stable ICU 2.0 */
+typedef struct UBiDi UBiDi;
+
+/**
+ * Allocate a <code>UBiDi</code> structure.
+ * Such an object is initially empty. It is assigned
+ * the Bidi properties of a piece of text containing one or more paragraphs
+ * by <code>ubidi_setPara()</code>
+ * or the Bidi properties of a line within a paragraph by
+ * <code>ubidi_setLine()</code>.<p>
+ * This object can be reused for as long as it is not deallocated
+ * by calling <code>ubidi_close()</code>.<p>
+ * <code>ubidi_setPara()</code> and <code>ubidi_setLine()</code> will allocate
+ * additional memory for internal structures as necessary.
+ *
+ * @return An empty <code>UBiDi</code> object.
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDi * U_EXPORT2
+ubidi_open(void);
+
+/**
+ * Allocate a <code>UBiDi</code> structure with preallocated memory
+ * for internal structures.
+ * This function provides a <code>UBiDi</code> object like <code>ubidi_open()</code>
+ * with no arguments, but it also preallocates memory for internal structures
+ * according to the sizings supplied by the caller.<p>
+ * Subsequent functions will not allocate any more memory, and are thus
+ * guaranteed not to fail because of lack of memory.<p>
+ * The preallocation can be limited to some of the internal memory
+ * by setting some values to 0 here. That means that if, e.g.,
+ * <code>maxRunCount</code> cannot be reasonably predetermined and should not
+ * be set to <code>maxLength</code> (the only failproof value) to avoid
+ * wasting memory, then <code>maxRunCount</code> could be set to 0 here
+ * and the internal structures that are associated with it will be allocated
+ * on demand, just like with <code>ubidi_open()</code>.
+ *
+ * @param maxLength is the maximum text or line length that internal memory
+ *        will be preallocated for. An attempt to associate this object with a
+ *        longer text will fail, unless this value is 0, which leaves the allocation
+ *        up to the implementation.
+ *
+ * @param maxRunCount is the maximum anticipated number of same-level runs
+ *        that internal memory will be preallocated for. An attempt to access
+ *        visual runs on an object that was not preallocated for as many runs
+ *        as the text was actually resolved to will fail,
+ *        unless this value is 0, which leaves the allocation up to the implementation.<br><br>
+ *        The number of runs depends on the actual text and maybe anywhere between
+ *        1 and <code>maxLength</code>. It is typically small.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return An empty <code>UBiDi</code> object with preallocated memory.
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDi * U_EXPORT2
+ubidi_openSized(int32_t maxLength, int32_t maxRunCount, UErrorCode *pErrorCode);
+
+/**
+ * <code>ubidi_close()</code> must be called to free the memory
+ * associated with a UBiDi object.<p>
+ *
+ * <strong>Important: </strong>
+ * A parent <code>UBiDi</code> object must not be destroyed or reused if
+ * it still has children.
+ * If a <code>UBiDi</code> object has become the <i>child</i>
+ * of another one (its <i>parent</i>) by calling
+ * <code>ubidi_setLine()</code>, then the child object must
+ * be destroyed (closed) or reused (by calling
+ * <code>ubidi_setPara()</code> or <code>ubidi_setLine()</code>)
+ * before the parent object.
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ *
+ * @see ubidi_setPara
+ * @see ubidi_setLine
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_close(UBiDi *pBiDi);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUBiDiPointer
+ * "Smart pointer" class, closes a UBiDi via ubidi_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUBiDiPointer, UBiDi, ubidi_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Modify the operation of the Bidi algorithm such that it
+ * approximates an "inverse Bidi" algorithm. This function
+ * must be called before <code>ubidi_setPara()</code>.
+ *
+ * <p>The normal operation of the Bidi algorithm as described
+ * in the Unicode Technical Report is to take text stored in logical
+ * (keyboard, typing) order and to determine the reordering of it for visual
+ * rendering.
+ * Some legacy systems store text in visual order, and for operations
+ * with standard, Unicode-based algorithms, the text needs to be transformed
+ * to logical order. This is effectively the inverse algorithm of the
+ * described Bidi algorithm. Note that there is no standard algorithm for
+ * this "inverse Bidi" and that the current implementation provides only an
+ * approximation of "inverse Bidi".</p>
+ *
+ * <p>With <code>isInverse</code> set to <code>true</code>,
+ * this function changes the behavior of some of the subsequent functions
+ * in a way that they can be used for the inverse Bidi algorithm.
+ * Specifically, runs of text with numeric characters will be treated in a
+ * special way and may need to be surrounded with LRM characters when they are
+ * written in reordered sequence.</p>
+ *
+ * <p>Output runs should be retrieved using <code>ubidi_getVisualRun()</code>.
+ * Since the actual input for "inverse Bidi" is visually ordered text and
+ * <code>ubidi_getVisualRun()</code> gets the reordered runs, these are actually
+ * the runs of the logically ordered output.</p>
+ *
+ * <p>Calling this function with argument <code>isInverse</code> set to
+ * <code>true</code> is equivalent to calling
+ * <code>ubidi_setReorderingMode</code> with argument
+ * <code>reorderingMode</code>
+ * set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
+ * Calling this function with argument <code>isInverse</code> set to
+ * <code>false</code> is equivalent to calling
+ * <code>ubidi_setReorderingMode</code> with argument
+ * <code>reorderingMode</code>
+ * set to <code>#UBIDI_REORDER_DEFAULT</code>.
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ *
+ * @param isInverse specifies "forward" or "inverse" Bidi operation.
+ *
+ * @see ubidi_setPara
+ * @see ubidi_writeReordered
+ * @see ubidi_setReorderingMode
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_setInverse(UBiDi *pBiDi, UBool isInverse);
+
+/**
+ * Is this Bidi object set to perform the inverse Bidi algorithm?
+ * <p>Note: calling this function after setting the reordering mode with
+ * <code>ubidi_setReorderingMode</code> will return <code>true</code> if the
+ * reordering mode was set to <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>,
+ * <code>false</code> for all other values.</p>
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @return true if the Bidi object is set to perform the inverse Bidi algorithm
+ * by handling numbers as L.
+ *
+ * @see ubidi_setInverse
+ * @see ubidi_setReorderingMode
+ * @stable ICU 2.0
+ */
+
+U_CAPI UBool U_EXPORT2
+ubidi_isInverse(UBiDi *pBiDi);
+
+/**
+ * Specify whether block separators must be allocated level zero,
+ * so that successive paragraphs will progress from left to right.
+ * This function must be called before <code>ubidi_setPara()</code>.
+ * Paragraph separators (B) may appear in the text.  Setting them to level zero
+ * means that all paragraph separators (including one possibly appearing
+ * in the last text position) are kept in the reordered text after the text
+ * that they follow in the source text.
+ * When this feature is not enabled, a paragraph separator at the last
+ * position of the text before reordering will go to the first position
+ * of the reordered text when the paragraph level is odd.
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ *
+ * @param orderParagraphsLTR specifies whether paragraph separators (B) must
+ * receive level 0, so that successive paragraphs progress from left to right.
+ *
+ * @see ubidi_setPara
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ubidi_orderParagraphsLTR(UBiDi *pBiDi, UBool orderParagraphsLTR);
+
+/**
+ * Is this Bidi object set to allocate level 0 to block separators so that
+ * successive paragraphs progress from left to right?
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @return true if the Bidi object is set to allocate level 0 to block
+ *         separators.
+ *
+ * @see ubidi_orderParagraphsLTR
+ * @stable ICU 3.4
+ */
+U_CAPI UBool U_EXPORT2
+ubidi_isOrderParagraphsLTR(UBiDi *pBiDi);
+
+/**
+ * <code>UBiDiReorderingMode</code> values indicate which variant of the Bidi
+ * algorithm to use.
+ *
+ * @see ubidi_setReorderingMode
+ * @stable ICU 3.6
+ */
+typedef enum UBiDiReorderingMode {
+    /** Regular Logical to Visual Bidi algorithm according to Unicode.
+      * This is a 0 value.
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_DEFAULT = 0,
+    /** Logical to Visual algorithm which handles numbers in a way which
+      * mimics the behavior of Windows XP.
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_NUMBERS_SPECIAL,
+    /** Logical to Visual algorithm grouping numbers with adjacent R characters
+      * (reversible algorithm).
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_GROUP_NUMBERS_WITH_R,
+    /** Reorder runs only to transform a Logical LTR string to the Logical RTL
+      * string with the same display, or vice-versa.<br>
+      * If this mode is set together with option
+      * <code>#UBIDI_OPTION_INSERT_MARKS</code>, some Bidi controls in the source
+      * text may be removed and other controls may be added to produce the
+      * minimum combination which has the required display.
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_RUNS_ONLY,
+    /** Visual to Logical algorithm which handles numbers like L
+      * (same algorithm as selected by <code>ubidi_setInverse(true)</code>.
+      * @see ubidi_setInverse
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_INVERSE_NUMBERS_AS_L,
+    /** Visual to Logical algorithm equivalent to the regular Logical to Visual
+      * algorithm.
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_INVERSE_LIKE_DIRECT,
+    /** Inverse Bidi (Visual to Logical) algorithm for the
+      * <code>UBIDI_REORDER_NUMBERS_SPECIAL</code> Bidi algorithm.
+      * @stable ICU 3.6 */
+    UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Number of values for reordering mode.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UBIDI_REORDER_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UBiDiReorderingMode;
+
+/**
+ * Modify the operation of the Bidi algorithm such that it implements some
+ * variant to the basic Bidi algorithm or approximates an "inverse Bidi"
+ * algorithm, depending on different values of the "reordering mode".
+ * This function must be called before <code>ubidi_setPara()</code>, and stays
+ * in effect until called again with a different argument.
+ *
+ * <p>The normal operation of the Bidi algorithm as described
+ * in the Unicode Standard Annex #9 is to take text stored in logical
+ * (keyboard, typing) order and to determine how to reorder it for visual
+ * rendering.</p>
+ *
+ * <p>With the reordering mode set to a value other than
+ * <code>#UBIDI_REORDER_DEFAULT</code>, this function changes the behavior of
+ * some of the subsequent functions in a way such that they implement an
+ * inverse Bidi algorithm or some other algorithm variants.</p>
+ *
+ * <p>Some legacy systems store text in visual order, and for operations
+ * with standard, Unicode-based algorithms, the text needs to be transformed
+ * into logical order. This is effectively the inverse algorithm of the
+ * described Bidi algorithm. Note that there is no standard algorithm for
+ * this "inverse Bidi", so a number of variants are implemented here.</p>
+ *
+ * <p>In other cases, it may be desirable to emulate some variant of the
+ * Logical to Visual algorithm (e.g. one used in MS Windows), or perform a
+ * Logical to Logical transformation.</p>
+ *
+ * <ul>
+ * <li>When the reordering mode is set to <code>#UBIDI_REORDER_DEFAULT</code>,
+ * the standard Bidi Logical to Visual algorithm is applied.</li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_NUMBERS_SPECIAL</code>,
+ * the algorithm used to perform Bidi transformations when calling
+ * <code>ubidi_setPara</code> should approximate the algorithm used in
+ * Microsoft Windows XP rather than strictly conform to the Unicode Bidi
+ * algorithm.
+ * <br>
+ * The differences between the basic algorithm and the algorithm addressed
+ * by this option are as follows:
+ * <ul>
+ *   <li>Within text at an even embedding level, the sequence "123AB"
+ *   (where AB represent R or AL letters) is transformed to "123BA" by the
+ *   Unicode algorithm and to "BA123" by the Windows algorithm.</li>
+ *   <li>Arabic-Indic numbers (AN) are handled by the Windows algorithm just
+ *   like regular numbers (EN).</li>
+ * </ul></li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_GROUP_NUMBERS_WITH_R</code>,
+ * numbers located between LTR text and RTL text are associated with the RTL
+ * text. For instance, an LTR paragraph with content "abc 123 DEF" (where
+ * upper case letters represent RTL characters) will be transformed to
+ * "abc FED 123" (and not "abc 123 FED"), "DEF 123 abc" will be transformed
+ * to "123 FED abc" and "123 FED abc" will be transformed to "DEF 123 abc".
+ * This makes the algorithm reversible and makes it useful when round trip
+ * (from visual to logical and back to visual) must be achieved without
+ * adding LRM characters. However, this is a variation from the standard
+ * Unicode Bidi algorithm.<br>
+ * The source text should not contain Bidi control characters other than LRM
+ * or RLM.</li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_RUNS_ONLY</code>,
+ * a "Logical to Logical" transformation must be performed:
+ * <ul>
+ * <li>If the default text level of the source text (argument <code>paraLevel</code>
+ * in <code>ubidi_setPara</code>) is even, the source text will be handled as
+ * LTR logical text and will be transformed to the RTL logical text which has
+ * the same LTR visual display.</li>
+ * <li>If the default level of the source text is odd, the source text
+ * will be handled as RTL logical text and will be transformed to the
+ * LTR logical text which has the same LTR visual display.</li>
+ * </ul>
+ * This mode may be needed when logical text which is basically Arabic or
+ * Hebrew, with possible included numbers or phrases in English, has to be
+ * displayed as if it had an even embedding level (this can happen if the
+ * displaying application treats all text as if it was basically LTR).
+ * <br>
+ * This mode may also be needed in the reverse case, when logical text which is
+ * basically English, with possible included phrases in Arabic or Hebrew, has to
+ * be displayed as if it had an odd embedding level.
+ * <br>
+ * Both cases could be handled by adding LRE or RLE at the head of the text,
+ * if the display subsystem supports these formatting controls. If it does not,
+ * the problem may be handled by transforming the source text in this mode
+ * before displaying it, so that it will be displayed properly.<br>
+ * The source text should not contain Bidi control characters other than LRM
+ * or RLM.</li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>, an "inverse Bidi" algorithm
+ * is applied.
+ * Runs of text with numeric characters will be treated like LTR letters and
+ * may need to be surrounded with LRM characters when they are written in
+ * reordered sequence (the option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> can
+ * be used with function <code>ubidi_writeReordered</code> to this end. This
+ * mode is equivalent to calling <code>ubidi_setInverse()</code> with
+ * argument <code>isInverse</code> set to <code>true</code>.</li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code>, the "direct" Logical to Visual
+ * Bidi algorithm is used as an approximation of an "inverse Bidi" algorithm.
+ * This mode is similar to mode <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>
+ * but is closer to the regular Bidi algorithm.
+ * <br>
+ * For example, an LTR paragraph with the content "FED 123 456 CBA" (where
+ * upper case represents RTL characters) will be transformed to
+ * "ABC 456 123 DEF", as opposed to "DEF 123 456 ABC"
+ * with mode <code>UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.<br>
+ * When used in conjunction with option
+ * <code>#UBIDI_OPTION_INSERT_MARKS</code>, this mode generally
+ * adds Bidi marks to the output significantly more sparingly than mode
+ * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> with option
+ * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to
+ * <code>ubidi_writeReordered</code>.</li>
+ *
+ * <li>When the reordering mode is set to
+ * <code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code>, the Logical to Visual
+ * Bidi algorithm used in Windows XP is used as an approximation of an "inverse Bidi" algorithm.
+ * <br>
+ * For example, an LTR paragraph with the content "abc FED123" (where
+ * upper case represents RTL characters) will be transformed to "abc 123DEF."</li>
+ * </ul>
+ *
+ * <p>In all the reordering modes specifying an "inverse Bidi" algorithm
+ * (i.e. those with a name starting with <code>UBIDI_REORDER_INVERSE</code>),
+ * output runs should be retrieved using
+ * <code>ubidi_getVisualRun()</code>, and the output text with
+ * <code>ubidi_writeReordered()</code>. The caller should keep in mind that in
+ * "inverse Bidi" modes the input is actually visually ordered text and
+ * reordered output returned by <code>ubidi_getVisualRun()</code> or
+ * <code>ubidi_writeReordered()</code> are actually runs or character string
+ * of logically ordered output.<br>
+ * For all the "inverse Bidi" modes, the source text should not contain
+ * Bidi control characters other than LRM or RLM.</p>
+ *
+ * <p>Note that option <code>#UBIDI_OUTPUT_REVERSE</code> of
+ * <code>ubidi_writeReordered</code> has no useful meaning and should not be
+ * used in conjunction with any value of the reordering mode specifying
+ * "inverse Bidi" or with value <code>UBIDI_REORDER_RUNS_ONLY</code>.
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @param reorderingMode specifies the required variant of the Bidi algorithm.
+ *
+ * @see UBiDiReorderingMode
+ * @see ubidi_setInverse
+ * @see ubidi_setPara
+ * @see ubidi_writeReordered
+ * @stable ICU 3.6
+ */
+U_CAPI void U_EXPORT2
+ubidi_setReorderingMode(UBiDi *pBiDi, UBiDiReorderingMode reorderingMode);
+
+/**
+ * What is the requested reordering mode for a given Bidi object?
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @return the current reordering mode of the Bidi object
+ * @see ubidi_setReorderingMode
+ * @stable ICU 3.6
+ */
+U_CAPI UBiDiReorderingMode U_EXPORT2
+ubidi_getReorderingMode(UBiDi *pBiDi);
+
+/**
+ * <code>UBiDiReorderingOption</code> values indicate which options are
+ * specified to affect the Bidi algorithm.
+ *
+ * @see ubidi_setReorderingOptions
+ * @stable ICU 3.6
+ */
+typedef enum UBiDiReorderingOption {
+    /**
+     * option value for <code>ubidi_setReorderingOptions</code>:
+     * disable all the options which can be set with this function
+     * @see ubidi_setReorderingOptions
+     * @stable ICU 3.6
+     */
+    UBIDI_OPTION_DEFAULT = 0,
+
+    /**
+     * option bit for <code>ubidi_setReorderingOptions</code>:
+     * insert Bidi marks (LRM or RLM) when needed to ensure correct result of
+     * a reordering to a Logical order
+     *
+     * <p>This option must be set or reset before calling
+     * <code>ubidi_setPara</code>.</p>
+     *
+     * <p>This option is significant only with reordering modes which generate
+     * a result with Logical order, specifically:</p>
+     * <ul>
+     *   <li><code>#UBIDI_REORDER_RUNS_ONLY</code></li>
+     *   <li><code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code></li>
+     *   <li><code>#UBIDI_REORDER_INVERSE_LIKE_DIRECT</code></li>
+     *   <li><code>#UBIDI_REORDER_INVERSE_FOR_NUMBERS_SPECIAL</code></li>
+     * </ul>
+     *
+     * <p>If this option is set in conjunction with reordering mode
+     * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> or with calling
+     * <code>ubidi_setInverse(true)</code>, it implies
+     * option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>
+     * in calls to function <code>ubidi_writeReordered()</code>.</p>
+     *
+     * <p>For other reordering modes, a minimum number of LRM or RLM characters
+     * will be added to the source text after reordering it so as to ensure
+     * round trip, i.e. when applying the inverse reordering mode on the
+     * resulting logical text with removal of Bidi marks
+     * (option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> set before calling
+     * <code>ubidi_setPara()</code> or option <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
+     * in <code>ubidi_writeReordered</code>), the result will be identical to the
+     * source text in the first transformation.
+     *
+     * <p>This option will be ignored if specified together with option
+     * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>. It inhibits option
+     * <code>UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to function
+     * <code>ubidi_writeReordered()</code> and it implies option
+     * <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls to function
+     * <code>ubidi_writeReordered()</code> if the reordering mode is
+     * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code>.</p>
+     *
+     * @see ubidi_setReorderingMode
+     * @see ubidi_setReorderingOptions
+     * @stable ICU 3.6
+     */
+    UBIDI_OPTION_INSERT_MARKS = 1,
+
+    /**
+     * option bit for <code>ubidi_setReorderingOptions</code>:
+     * remove Bidi control characters
+     *
+     * <p>This option must be set or reset before calling
+     * <code>ubidi_setPara</code>.</p>
+     *
+     * <p>This option nullifies option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
+     * It inhibits option <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code> in calls
+     * to function <code>ubidi_writeReordered()</code> and it implies option
+     * <code>#UBIDI_REMOVE_BIDI_CONTROLS</code> in calls to that function.</p>
+     *
+     * @see ubidi_setReorderingMode
+     * @see ubidi_setReorderingOptions
+     * @stable ICU 3.6
+     */
+    UBIDI_OPTION_REMOVE_CONTROLS = 2,
+
+    /**
+     * option bit for <code>ubidi_setReorderingOptions</code>:
+     * process the output as part of a stream to be continued
+     *
+     * <p>This option must be set or reset before calling
+     * <code>ubidi_setPara</code>.</p>
+     *
+     * <p>This option specifies that the caller is interested in processing large
+     * text object in parts.
+     * The results of the successive calls are expected to be concatenated by the
+     * caller. Only the call for the last part will have this option bit off.</p>
+     *
+     * <p>When this option bit is on, <code>ubidi_setPara()</code> may process
+     * less than the full source text in order to truncate the text at a meaningful
+     * boundary. The caller should call <code>ubidi_getProcessedLength()</code>
+     * immediately after calling <code>ubidi_setPara()</code> in order to
+     * determine how much of the source text has been processed.
+     * Source text beyond that length should be resubmitted in following calls to
+     * <code>ubidi_setPara</code>. The processed length may be less than
+     * the length of the source text if a character preceding the last character of
+     * the source text constitutes a reasonable boundary (like a block separator)
+     * for text to be continued.<br>
+     * If the last character of the source text constitutes a reasonable
+     * boundary, the whole text will be processed at once.<br>
+     * If nowhere in the source text there exists
+     * such a reasonable boundary, the processed length will be zero.<br>
+     * The caller should check for such an occurrence and do one of the following:
+     * <ul><li>submit a larger amount of text with a better chance to include
+     *         a reasonable boundary.</li>
+     *     <li>resubmit the same text after turning off option
+     *         <code>UBIDI_OPTION_STREAMING</code>.</li></ul>
+     * In all cases, this option should be turned off before processing the last
+     * part of the text.</p>
+     *
+     * <p>When the <code>UBIDI_OPTION_STREAMING</code> option is used,
+     * it is recommended to call <code>ubidi_orderParagraphsLTR()</code> with
+     * argument <code>orderParagraphsLTR</code> set to <code>true</code> before
+     * calling <code>ubidi_setPara</code> so that later paragraphs may be
+     * concatenated to previous paragraphs on the right.</p>
+     *
+     * @see ubidi_setReorderingMode
+     * @see ubidi_setReorderingOptions
+     * @see ubidi_getProcessedLength
+     * @see ubidi_orderParagraphsLTR
+     * @stable ICU 3.6
+     */
+    UBIDI_OPTION_STREAMING = 4
+} UBiDiReorderingOption;
+
+/**
+ * Specify which of the reordering options
+ * should be applied during Bidi transformations.
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @param reorderingOptions is a combination of zero or more of the following
+ * options:
+ * <code>#UBIDI_OPTION_DEFAULT</code>, <code>#UBIDI_OPTION_INSERT_MARKS</code>,
+ * <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>, <code>#UBIDI_OPTION_STREAMING</code>.
+ *
+ * @see ubidi_getReorderingOptions
+ * @stable ICU 3.6
+ */
+U_CAPI void U_EXPORT2
+ubidi_setReorderingOptions(UBiDi *pBiDi, uint32_t reorderingOptions);
+
+/**
+ * What are the reordering options applied to a given Bidi object?
+ *
+ * @param pBiDi is a <code>UBiDi</code> object.
+ * @return the current reordering options of the Bidi object
+ * @see ubidi_setReorderingOptions
+ * @stable ICU 3.6
+ */
+U_CAPI uint32_t U_EXPORT2
+ubidi_getReorderingOptions(UBiDi *pBiDi);
+
+/**
+ * Set the context before a call to ubidi_setPara().<p>
+ *
+ * ubidi_setPara() computes the left-right directionality for a given piece
+ * of text which is supplied as one of its arguments. Sometimes this piece
+ * of text (the "main text") should be considered in context, because text
+ * appearing before ("prologue") and/or after ("epilogue") the main text
+ * may affect the result of this computation.<p>
+ *
+ * This function specifies the prologue and/or the epilogue for the next
+ * call to ubidi_setPara(). The characters specified as prologue and
+ * epilogue should not be modified by the calling program until the call
+ * to ubidi_setPara() has returned. If successive calls to ubidi_setPara()
+ * all need specification of a context, ubidi_setContext() must be called
+ * before each call to ubidi_setPara(). In other words, a context is not
+ * "remembered" after the following successful call to ubidi_setPara().<p>
+ *
+ * If a call to ubidi_setPara() specifies UBIDI_DEFAULT_LTR or
+ * UBIDI_DEFAULT_RTL as paraLevel and is preceded by a call to
+ * ubidi_setContext() which specifies a prologue, the paragraph level will
+ * be computed taking in consideration the text in the prologue.<p>
+ *
+ * When ubidi_setPara() is called without a previous call to
+ * ubidi_setContext, the main text is handled as if preceded and followed
+ * by strong directional characters at the current paragraph level.
+ * Calling ubidi_setContext() with specification of a prologue will change
+ * this behavior by handling the main text as if preceded by the last
+ * strong character appearing in the prologue, if any.
+ * Calling ubidi_setContext() with specification of an epilogue will change
+ * the behavior of ubidi_setPara() by handling the main text as if followed
+ * by the first strong character or digit appearing in the epilogue, if any.<p>
+ *
+ * Note 1: if <code>ubidi_setContext</code> is called repeatedly without
+ *         calling <code>ubidi_setPara</code>, the earlier calls have no effect,
+ *         only the last call will be remembered for the next call to
+ *         <code>ubidi_setPara</code>.<p>
+ *
+ * Note 2: calling <code>ubidi_setContext(pBiDi, NULL, 0, NULL, 0, &errorCode)</code>
+ *         cancels any previous setting of non-empty prologue or epilogue.
+ *         The next call to <code>ubidi_setPara()</code> will process no
+ *         prologue or epilogue.<p>
+ *
+ * Note 3: users must be aware that even after setting the context
+ *         before a call to ubidi_setPara() to perform e.g. a logical to visual
+ *         transformation, the resulting string may not be identical to what it
+ *         would have been if all the text, including prologue and epilogue, had
+ *         been processed together.<br>
+ * Example (upper case letters represent RTL characters):<br>
+ * &nbsp;&nbsp;prologue = "<code>abc DE</code>"<br>
+ * &nbsp;&nbsp;epilogue = none<br>
+ * &nbsp;&nbsp;main text = "<code>FGH xyz</code>"<br>
+ * &nbsp;&nbsp;paraLevel = UBIDI_LTR<br>
+ * &nbsp;&nbsp;display without prologue = "<code>HGF xyz</code>"
+ *             ("HGF" is adjacent to "xyz")<br>
+ * &nbsp;&nbsp;display with prologue = "<code>abc HGFED xyz</code>"
+ *             ("HGF" is not adjacent to "xyz")<br>
+ *
+ * @param pBiDi is a paragraph <code>UBiDi</code> object.
+ *
+ * @param prologue is a pointer to the text which precedes the text that
+ *        will be specified in a coming call to ubidi_setPara().
+ *        If there is no prologue to consider, then <code>proLength</code>
+ *        must be zero and this pointer can be NULL.
+ *
+ * @param proLength is the length of the prologue; if <code>proLength==-1</code>
+ *        then the prologue must be zero-terminated.
+ *        Otherwise proLength must be >= 0. If <code>proLength==0</code>, it means
+ *        that there is no prologue to consider.
+ *
+ * @param epilogue is a pointer to the text which follows the text that
+ *        will be specified in a coming call to ubidi_setPara().
+ *        If there is no epilogue to consider, then <code>epiLength</code>
+ *        must be zero and this pointer can be NULL.
+ *
+ * @param epiLength is the length of the epilogue; if <code>epiLength==-1</code>
+ *        then the epilogue must be zero-terminated.
+ *        Otherwise epiLength must be >= 0. If <code>epiLength==0</code>, it means
+ *        that there is no epilogue to consider.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @see ubidi_setPara
+ * @stable ICU 4.8
+ */
+U_CAPI void U_EXPORT2
+ubidi_setContext(UBiDi *pBiDi,
+                 const UChar *prologue, int32_t proLength,
+                 const UChar *epilogue, int32_t epiLength,
+                 UErrorCode *pErrorCode);
+
+/**
+ * Perform the Unicode Bidi algorithm. It is defined in the
+ * <a href="http://www.unicode.org/unicode/reports/tr9/">Unicode Standard Annex #9</a>,
+ * version 13,
+ * also described in The Unicode Standard, Version 4.0 .<p>
+ *
+ * This function takes a piece of plain text containing one or more paragraphs,
+ * with or without externally specified embedding levels from <i>styled</i>
+ * text and computes the left-right-directionality of each character.<p>
+ *
+ * If the entire text is all of the same directionality, then
+ * the function may not perform all the steps described by the algorithm,
+ * i.e., some levels may not be the same as if all steps were performed.
+ * This is not relevant for unidirectional text.<br>
+ * For example, in pure LTR text with numbers the numbers would get
+ * a resolved level of 2 higher than the surrounding text according to
+ * the algorithm. This implementation may set all resolved levels to
+ * the same value in such a case.<p>
+ *
+ * The text can be composed of multiple paragraphs. Occurrence of a block
+ * separator in the text terminates a paragraph, and whatever comes next starts
+ * a new paragraph. The exception to this rule is when a Carriage Return (CR)
+ * is followed by a Line Feed (LF). Both CR and LF are block separators, but
+ * in that case, the pair of characters is considered as terminating the
+ * preceding paragraph, and a new paragraph will be started by a character
+ * coming after the LF.
+ *
+ * @param pBiDi A <code>UBiDi</code> object allocated with <code>ubidi_open()</code>
+ *        which will be set to contain the reordering information,
+ *        especially the resolved levels for all the characters in <code>text</code>.
+ *
+ * @param text is a pointer to the text that the Bidi algorithm will be performed on.
+ *        This pointer is stored in the UBiDi object and can be retrieved
+ *        with <code>ubidi_getText()</code>.<br>
+ *        <strong>Note:</strong> the text must be (at least) <code>length</code> long.
+ *
+ * @param length is the length of the text; if <code>length==-1</code> then
+ *        the text must be zero-terminated.
+ *
+ * @param paraLevel specifies the default level for the text;
+ *        it is typically 0 (LTR) or 1 (RTL).
+ *        If the function shall determine the paragraph level from the text,
+ *        then <code>paraLevel</code> can be set to
+ *        either <code>#UBIDI_DEFAULT_LTR</code>
+ *        or <code>#UBIDI_DEFAULT_RTL</code>; if the text contains multiple
+ *        paragraphs, the paragraph level shall be determined separately for
+ *        each paragraph; if a paragraph does not include any strongly typed
+ *        character, then the desired default is used (0 for LTR or 1 for RTL).
+ *        Any other value between 0 and <code>#UBIDI_MAX_EXPLICIT_LEVEL</code>
+ *        is also valid, with odd levels indicating RTL.
+ *
+ * @param embeddingLevels (in) may be used to preset the embedding and override levels,
+ *        ignoring characters like LRE and PDF in the text.
+ *        A level overrides the directional property of its corresponding
+ *        (same index) character if the level has the
+ *        <code>#UBIDI_LEVEL_OVERRIDE</code> bit set.<br><br>
+ *        Aside from that bit, it must be
+ *        <code>paraLevel<=embeddingLevels[]<=UBIDI_MAX_EXPLICIT_LEVEL</code>,
+ *        except that level 0 is always allowed.
+ *        Level 0 for a paragraph separator prevents reordering of paragraphs;
+ *        this only works reliably if <code>#UBIDI_LEVEL_OVERRIDE</code>
+ *        is also set for paragraph separators.
+ *        Level 0 for other characters is treated as a wildcard
+ *        and is lifted up to the resolved level of the surrounding paragraph.<br><br>
+ *        <strong>Caution: </strong>A copy of this pointer, not of the levels,
+ *        will be stored in the <code>UBiDi</code> object;
+ *        the <code>embeddingLevels</code> array must not be
+ *        deallocated before the <code>UBiDi</code> structure is destroyed or reused,
+ *        and the <code>embeddingLevels</code>
+ *        should not be modified to avoid unexpected results on subsequent Bidi operations.
+ *        However, the <code>ubidi_setPara()</code> and
+ *        <code>ubidi_setLine()</code> functions may modify some or all of the levels.<br><br>
+ *        After the <code>UBiDi</code> object is reused or destroyed, the caller
+ *        must take care of the deallocation of the <code>embeddingLevels</code> array.<br><br>
+ *        <strong>Note:</strong> the <code>embeddingLevels</code> array must be
+ *        at least <code>length</code> long.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_setPara(UBiDi *pBiDi, const UChar *text, int32_t length,
+              UBiDiLevel paraLevel, UBiDiLevel *embeddingLevels,
+              UErrorCode *pErrorCode);
+
+/**
+ * <code>ubidi_setLine()</code> sets a <code>UBiDi</code> to
+ * contain the reordering information, especially the resolved levels,
+ * for all the characters in a line of text. This line of text is
+ * specified by referring to a <code>UBiDi</code> object representing
+ * this information for a piece of text containing one or more paragraphs,
+ * and by specifying a range of indexes in this text.<p>
+ * In the new line object, the indexes will range from 0 to <code>limit-start-1</code>.<p>
+ *
+ * This is used after calling <code>ubidi_setPara()</code>
+ * for a piece of text, and after line-breaking on that text.
+ * It is not necessary if each paragraph is treated as a single line.<p>
+ *
+ * After line-breaking, rules (L1) and (L2) for the treatment of
+ * trailing WS and for reordering are performed on
+ * a <code>UBiDi</code> object that represents a line.<p>
+ *
+ * <strong>Important: </strong><code>pLineBiDi</code> shares data with
+ * <code>pParaBiDi</code>.
+ * You must destroy or reuse <code>pLineBiDi</code> before <code>pParaBiDi</code>.
+ * In other words, you must destroy or reuse the <code>UBiDi</code> object for a line
+ * before the object for its parent paragraph.<p>
+ *
+ * The text pointer that was stored in <code>pParaBiDi</code> is also copied,
+ * and <code>start</code> is added to it so that it points to the beginning of the
+ * line for this object.
+ *
+ * @param pParaBiDi is the parent paragraph object. It must have been set
+ * by a successful call to ubidi_setPara.
+ *
+ * @param start is the line's first index into the text.
+ *
+ * @param limit is just behind the line's last index into the text
+ *        (its last index +1).<br>
+ *        It must be <code>0<=start<limit<=</code>containing paragraph limit.
+ *        If the specified line crosses a paragraph boundary, the function
+ *        will terminate with error code U_ILLEGAL_ARGUMENT_ERROR.
+ *
+ * @param pLineBiDi is the object that will now represent a line of the text.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @see ubidi_setPara
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_setLine(const UBiDi *pParaBiDi,
+              int32_t start, int32_t limit,
+              UBiDi *pLineBiDi,
+              UErrorCode *pErrorCode);
+
+/**
+ * Get the directionality of the text.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @return a value of <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>
+ *         or <code>UBIDI_MIXED</code>
+ *         that indicates if the entire text
+ *         represented by this object is unidirectional,
+ *         and which direction, or if it is mixed-directional.
+ * Note -  The value <code>UBIDI_NEUTRAL</code> is never returned from this method.
+ *
+ * @see UBiDiDirection
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getDirection(const UBiDi *pBiDi);
+
+/**
+ * Gets the base direction of the text provided according
+ * to the Unicode Bidirectional Algorithm. The base direction
+ * is derived from the first character in the string with bidirectional
+ * character type L, R, or AL. If the first such character has type L,
+ * <code>UBIDI_LTR</code> is returned. If the first such character has
+ * type R or AL, <code>UBIDI_RTL</code> is returned. If the string does
+ * not contain any character of these types, then
+ * <code>UBIDI_NEUTRAL</code> is returned.
+ *
+ * This is a lightweight function for use when only the base direction
+ * is needed and no further bidi processing of the text is needed.
+ *
+ * @param text is a pointer to the text whose base
+ *             direction is needed.
+ * Note: the text must be (at least) @c length long.
+ *
+ * @param length is the length of the text;
+ *               if <code>length==-1</code> then the text
+ *               must be zero-terminated.
+ *
+ * @return  <code>UBIDI_LTR</code>, <code>UBIDI_RTL</code>,
+ *          <code>UBIDI_NEUTRAL</code>
+ *
+ * @see UBiDiDirection
+ * @stable ICU 4.6
+ */
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getBaseDirection(const UChar *text,  int32_t length );
+
+/**
+ * Get the pointer to the text.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @return The pointer to the text that the UBiDi object was created for.
+ *
+ * @see ubidi_setPara
+ * @see ubidi_setLine
+ * @stable ICU 2.0
+ */
+U_CAPI const UChar * U_EXPORT2
+ubidi_getText(const UBiDi *pBiDi);
+
+/**
+ * Get the length of the text.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @return The length of the text that the UBiDi object was created for.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getLength(const UBiDi *pBiDi);
+
+/**
+ * Get the paragraph level of the text.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @return The paragraph level. If there are multiple paragraphs, their
+ *         level may vary if the required paraLevel is UBIDI_DEFAULT_LTR or
+ *         UBIDI_DEFAULT_RTL.  In that case, the level of the first paragraph
+ *         is returned.
+ *
+ * @see UBiDiLevel
+ * @see ubidi_getParagraph
+ * @see ubidi_getParagraphByIndex
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDiLevel U_EXPORT2
+ubidi_getParaLevel(const UBiDi *pBiDi);
+
+/**
+ * Get the number of paragraphs.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @return The number of paragraphs.
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_countParagraphs(UBiDi *pBiDi);
+
+/**
+ * Get a paragraph, given a position within the text.
+ * This function returns information about a paragraph.<br>
+ * Note: if the paragraph index is known, it is more efficient to
+ * retrieve the paragraph information using ubidi_getParagraphByIndex().<p>
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param charIndex is the index of a character within the text, in the
+ *        range <code>[0..ubidi_getProcessedLength(pBiDi)-1]</code>.
+ *
+ * @param pParaStart will receive the index of the first character of the
+ *        paragraph in the text.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pParaLimit will receive the limit of the paragraph.
+ *        The l-value that you point to here may be the
+ *        same expression (variable) as the one for
+ *        <code>charIndex</code>.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pParaLevel will receive the level of the paragraph.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The index of the paragraph containing the specified position.
+ *
+ * @see ubidi_getProcessedLength
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getParagraph(const UBiDi *pBiDi, int32_t charIndex, int32_t *pParaStart,
+                   int32_t *pParaLimit, UBiDiLevel *pParaLevel,
+                   UErrorCode *pErrorCode);
+
+/**
+ * Get a paragraph, given the index of this paragraph.
+ *
+ * This function returns information about a paragraph.<p>
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @param paraIndex is the number of the paragraph, in the
+ *        range <code>[0..ubidi_countParagraphs(pBiDi)-1]</code>.
+ *
+ * @param pParaStart will receive the index of the first character of the
+ *        paragraph in the text.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pParaLimit will receive the limit of the paragraph.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pParaLevel will receive the level of the paragraph.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ubidi_getParagraphByIndex(const UBiDi *pBiDi, int32_t paraIndex,
+                          int32_t *pParaStart, int32_t *pParaLimit,
+                          UBiDiLevel *pParaLevel, UErrorCode *pErrorCode);
+
+/**
+ * Get the level for one character.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param charIndex the index of a character. It must be in the range
+ *         [0..ubidi_getProcessedLength(pBiDi)].
+ *
+ * @return The level for the character at charIndex (0 if charIndex is not
+ *         in the valid range).
+ *
+ * @see UBiDiLevel
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDiLevel U_EXPORT2
+ubidi_getLevelAt(const UBiDi *pBiDi, int32_t charIndex);
+
+/**
+ * Get an array of levels for each character.<p>
+ *
+ * Note that this function may allocate memory under some
+ * circumstances, unlike <code>ubidi_getLevelAt()</code>.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object, whose
+ *        text length must be strictly positive.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The levels array for the text,
+ *         or <code>NULL</code> if an error occurs.
+ *
+ * @see UBiDiLevel
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI const UBiDiLevel * U_EXPORT2
+ubidi_getLevels(UBiDi *pBiDi, UErrorCode *pErrorCode);
+
+/**
+ * Get a logical run.
+ * This function returns information about a run and is used
+ * to retrieve runs in logical order.<p>
+ * This is especially useful for line-breaking on a paragraph.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param logicalPosition is a logical position within the source text.
+ *
+ * @param pLogicalLimit will receive the limit of the corresponding run.
+ *        The l-value that you point to here may be the
+ *        same expression (variable) as the one for
+ *        <code>logicalPosition</code>.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @param pLevel will receive the level of the corresponding run.
+ *        This pointer can be <code>NULL</code> if this
+ *        value is not necessary.
+ *
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_getLogicalRun(const UBiDi *pBiDi, int32_t logicalPosition,
+                    int32_t *pLogicalLimit, UBiDiLevel *pLevel);
+
+/**
+ * Get the number of runs.
+ * This function may invoke the actual reordering on the
+ * <code>UBiDi</code> object, after <code>ubidi_setPara()</code>
+ * may have resolved only the levels of the text. Therefore,
+ * <code>ubidi_countRuns()</code> may have to allocate memory,
+ * and may fail doing so.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The number of runs.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_countRuns(UBiDi *pBiDi, UErrorCode *pErrorCode);
+
+/**
+ * Get one run's logical start, length, and directionality,
+ * which can be 0 for LTR or 1 for RTL.
+ * In an RTL run, the character at the logical start is
+ * visually on the right of the displayed run.
+ * The length is the number of characters in the run.<p>
+ * <code>ubidi_countRuns()</code> should be called
+ * before the runs are retrieved.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param runIndex is the number of the run in visual order, in the
+ *        range <code>[0..ubidi_countRuns(pBiDi)-1]</code>.
+ *
+ * @param pLogicalStart is the first logical character index in the text.
+ *        The pointer may be <code>NULL</code> if this index is not needed.
+ *
+ * @param pLength is the number of characters (at least one) in the run.
+ *        The pointer may be <code>NULL</code> if this is not needed.
+ *
+ * @return the directionality of the run,
+ *         <code>UBIDI_LTR==0</code> or <code>UBIDI_RTL==1</code>,
+ *         never <code>UBIDI_MIXED</code>,
+ *         never <code>UBIDI_NEUTRAL</code>.
+ *
+ * @see ubidi_countRuns
+ *
+ * Example:
+ * <pre>
+ * \code
+ * int32_t i, count=ubidi_countRuns(pBiDi),
+ *         logicalStart, visualIndex=0, length;
+ * for(i=0; i<count; ++i) {
+ *    if(UBIDI_LTR==ubidi_getVisualRun(pBiDi, i, &logicalStart, &length)) {
+ *         do { // LTR
+ *             show_char(text[logicalStart++], visualIndex++);
+ *         } while(--length>0);
+ *     } else {
+ *         logicalStart+=length;  // logicalLimit
+ *         do { // RTL
+ *             show_char(text[--logicalStart], visualIndex++);
+ *         } while(--length>0);
+ *     }
+ * }
+ *\endcode
+ * </pre>
+ *
+ * Note that in right-to-left runs, code like this places
+ * second surrogates before first ones (which is generally a bad idea)
+ * and combining characters before base characters.
+ * <p>
+ * Use of <code>ubidi_writeReordered()</code>, optionally with the
+ * <code>#UBIDI_KEEP_BASE_COMBINING</code> option, can be considered in order
+ * to avoid these issues.
+ * @stable ICU 2.0
+ */
+U_CAPI UBiDiDirection U_EXPORT2
+ubidi_getVisualRun(UBiDi *pBiDi, int32_t runIndex,
+                   int32_t *pLogicalStart, int32_t *pLength);
+
+/**
+ * Get the visual position from a logical text position.
+ * If such a mapping is used many times on the same
+ * <code>UBiDi</code> object, then calling
+ * <code>ubidi_getLogicalMap()</code> is more efficient.<p>
+ *
+ * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
+ * visual position because the corresponding text character is a Bidi control
+ * removed from output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ * <p>
+ * When the visual output is altered by using options of
+ * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
+ * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
+ * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual position returned may not
+ * be correct. It is advised to use, when possible, reordering options
+ * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ * <p>
+ * Note that in right-to-left runs, this mapping places
+ * second surrogates before first ones (which is generally a bad idea)
+ * and combining characters before base characters.
+ * Use of <code>ubidi_writeReordered()</code>, optionally with the
+ * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead
+ * of using the mapping, in order to avoid these issues.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param logicalIndex is the index of a character in the text.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The visual position of this character.
+ *
+ * @see ubidi_getLogicalMap
+ * @see ubidi_getLogicalIndex
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getVisualIndex(UBiDi *pBiDi, int32_t logicalIndex, UErrorCode *pErrorCode);
+
+/**
+ * Get the logical text position from a visual position.
+ * If such a mapping is used many times on the same
+ * <code>UBiDi</code> object, then calling
+ * <code>ubidi_getVisualMap()</code> is more efficient.<p>
+ *
+ * The value returned may be <code>#UBIDI_MAP_NOWHERE</code> if there is no
+ * logical position because the corresponding text character is a Bidi mark
+ * inserted in the output by option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
+ * <p>
+ * This is the inverse function to <code>ubidi_getVisualIndex()</code>.
+ * <p>
+ * When the visual output is altered by using options of
+ * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
+ * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
+ * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical position returned may not
+ * be correct. It is advised to use, when possible, reordering options
+ * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param visualIndex is the visual position of a character.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The index of this character in the text.
+ *
+ * @see ubidi_getVisualMap
+ * @see ubidi_getVisualIndex
+ * @see ubidi_getResultLength
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getLogicalIndex(UBiDi *pBiDi, int32_t visualIndex, UErrorCode *pErrorCode);
+
+/**
+ * Get a logical-to-visual index map (array) for the characters in the UBiDi
+ * (paragraph or line) object.
+ * <p>
+ * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
+ * corresponding text characters are Bidi controls removed from the visual
+ * output by the option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ * <p>
+ * When the visual output is altered by using options of
+ * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
+ * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
+ * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the visual positions returned may not
+ * be correct. It is advised to use, when possible, reordering options
+ * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ * <p>
+ * Note that in right-to-left runs, this mapping places
+ * second surrogates before first ones (which is generally a bad idea)
+ * and combining characters before base characters.
+ * Use of <code>ubidi_writeReordered()</code>, optionally with the
+ * <code>#UBIDI_KEEP_BASE_COMBINING</code> option can be considered instead
+ * of using the mapping, in order to avoid these issues.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param indexMap is a pointer to an array of <code>ubidi_getProcessedLength()</code>
+ *        indexes which will reflect the reordering of the characters.
+ *        If option <code>#UBIDI_OPTION_INSERT_MARKS</code> is set, the number
+ *        of elements allocated in <code>indexMap</code> must be no less than
+ *        <code>ubidi_getResultLength()</code>.
+ *        The array does not need to be initialized.<br><br>
+ *        The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @see ubidi_getVisualMap
+ * @see ubidi_getVisualIndex
+ * @see ubidi_getProcessedLength
+ * @see ubidi_getResultLength
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_getLogicalMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
+
+/**
+ * Get a visual-to-logical index map (array) for the characters in the UBiDi
+ * (paragraph or line) object.
+ * <p>
+ * Some values in the map may be <code>#UBIDI_MAP_NOWHERE</code> if the
+ * corresponding text characters are Bidi marks inserted in the visual output
+ * by the option <code>#UBIDI_OPTION_INSERT_MARKS</code>.
+ * <p>
+ * When the visual output is altered by using options of
+ * <code>ubidi_writeReordered()</code> such as <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
+ * <code>UBIDI_KEEP_BASE_COMBINING</code>, <code>UBIDI_OUTPUT_REVERSE</code>,
+ * <code>UBIDI_REMOVE_BIDI_CONTROLS</code>, the logical positions returned may not
+ * be correct. It is advised to use, when possible, reordering options
+ * such as <code>UBIDI_OPTION_INSERT_MARKS</code> and <code>UBIDI_OPTION_REMOVE_CONTROLS</code>.
+ *
+ * @param pBiDi is the paragraph or line <code>UBiDi</code> object.
+ *
+ * @param indexMap is a pointer to an array of <code>ubidi_getResultLength()</code>
+ *        indexes which will reflect the reordering of the characters.
+ *        If option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> is set, the number
+ *        of elements allocated in <code>indexMap</code> must be no less than
+ *        <code>ubidi_getProcessedLength()</code>.
+ *        The array does not need to be initialized.<br><br>
+ *        The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @see ubidi_getLogicalMap
+ * @see ubidi_getLogicalIndex
+ * @see ubidi_getProcessedLength
+ * @see ubidi_getResultLength
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_getVisualMap(UBiDi *pBiDi, int32_t *indexMap, UErrorCode *pErrorCode);
+
+/**
+ * This is a convenience function that does not use a UBiDi object.
+ * It is intended to be used for when an application has determined the levels
+ * of objects (character sequences) and just needs to have them reordered (L2).
+ * This is equivalent to using <code>ubidi_getLogicalMap()</code> on a
+ * <code>UBiDi</code> object.
+ *
+ * @param levels is an array with <code>length</code> levels that have been determined by
+ *        the application.
+ *
+ * @param length is the number of levels in the array, or, semantically,
+ *        the number of objects to be reordered.
+ *        It must be <code>length>0</code>.
+ *
+ * @param indexMap is a pointer to an array of <code>length</code>
+ *        indexes which will reflect the reordering of the characters.
+ *        The array does not need to be initialized.<p>
+ *        The index map will result in <code>indexMap[logicalIndex]==visualIndex</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_reorderLogical(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
+
+/**
+ * This is a convenience function that does not use a UBiDi object.
+ * It is intended to be used for when an application has determined the levels
+ * of objects (character sequences) and just needs to have them reordered (L2).
+ * This is equivalent to using <code>ubidi_getVisualMap()</code> on a
+ * <code>UBiDi</code> object.
+ *
+ * @param levels is an array with <code>length</code> levels that have been determined by
+ *        the application.
+ *
+ * @param length is the number of levels in the array, or, semantically,
+ *        the number of objects to be reordered.
+ *        It must be <code>length>0</code>.
+ *
+ * @param indexMap is a pointer to an array of <code>length</code>
+ *        indexes which will reflect the reordering of the characters.
+ *        The array does not need to be initialized.<p>
+ *        The index map will result in <code>indexMap[visualIndex]==logicalIndex</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_reorderVisual(const UBiDiLevel *levels, int32_t length, int32_t *indexMap);
+
+/**
+ * Invert an index map.
+ * The index mapping of the first map is inverted and written to
+ * the second one.
+ *
+ * @param srcMap is an array with <code>length</code> elements
+ *        which defines the original mapping from a source array containing
+ *        <code>length</code> elements to a destination array.
+ *        Some elements of the source array may have no mapping in the
+ *        destination array. In that case, their value will be
+ *        the special value <code>UBIDI_MAP_NOWHERE</code>.
+ *        All elements must be >=0 or equal to <code>UBIDI_MAP_NOWHERE</code>.
+ *        Some elements may have a value >= <code>length</code>, if the
+ *        destination array has more elements than the source array.
+ *        There must be no duplicate indexes (two or more elements with the
+ *        same value except <code>UBIDI_MAP_NOWHERE</code>).
+ *
+ * @param destMap is an array with a number of elements equal to 1 + the highest
+ *        value in <code>srcMap</code>.
+ *        <code>destMap</code> will be filled with the inverse mapping.
+ *        If element with index i in <code>srcMap</code> has a value k different
+ *        from <code>UBIDI_MAP_NOWHERE</code>, this means that element i of
+ *        the source array maps to element k in the destination array.
+ *        The inverse map will have value i in its k-th element.
+ *        For all elements of the destination array which do not map to
+ *        an element in the source array, the corresponding element in the
+ *        inverse map will have a value equal to <code>UBIDI_MAP_NOWHERE</code>.
+ *
+ * @param length is the length of each array.
+ * @see UBIDI_MAP_NOWHERE
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubidi_invertMap(const int32_t *srcMap, int32_t *destMap, int32_t length);
+
+/** option flags for ubidi_writeReordered() */
+
+/**
+ * option bit for ubidi_writeReordered():
+ * keep combining characters after their base characters in RTL runs
+ *
+ * @see ubidi_writeReordered
+ * @stable ICU 2.0
+ */
+#define UBIDI_KEEP_BASE_COMBINING       1
+
+/**
+ * option bit for ubidi_writeReordered():
+ * replace characters with the "mirrored" property in RTL runs
+ * by their mirror-image mappings
+ *
+ * @see ubidi_writeReordered
+ * @stable ICU 2.0
+ */
+#define UBIDI_DO_MIRRORING              2
+
+/**
+ * option bit for ubidi_writeReordered():
+ * surround the run with LRMs if necessary;
+ * this is part of the approximate "inverse Bidi" algorithm
+ *
+ * <p>This option does not imply corresponding adjustment of the index
+ * mappings.</p>
+ *
+ * @see ubidi_setInverse
+ * @see ubidi_writeReordered
+ * @stable ICU 2.0
+ */
+#define UBIDI_INSERT_LRM_FOR_NUMERIC    4
+
+/**
+ * option bit for ubidi_writeReordered():
+ * remove Bidi control characters
+ * (this does not affect #UBIDI_INSERT_LRM_FOR_NUMERIC)
+ *
+ * <p>This option does not imply corresponding adjustment of the index
+ * mappings.</p>
+ *
+ * @see ubidi_writeReordered
+ * @stable ICU 2.0
+ */
+#define UBIDI_REMOVE_BIDI_CONTROLS      8
+
+/**
+ * option bit for ubidi_writeReordered():
+ * write the output in reverse order
+ *
+ * <p>This has the same effect as calling <code>ubidi_writeReordered()</code>
+ * first without this option, and then calling
+ * <code>ubidi_writeReverse()</code> without mirroring.
+ * Doing this in the same step is faster and avoids a temporary buffer.
+ * An example for using this option is output to a character terminal that
+ * is designed for RTL scripts and stores text in reverse order.</p>
+ *
+ * @see ubidi_writeReordered
+ * @stable ICU 2.0
+ */
+#define UBIDI_OUTPUT_REVERSE            16
+
+/**
+ * Get the length of the source text processed by the last call to
+ * <code>ubidi_setPara()</code>. This length may be different from the length
+ * of the source text if option <code>#UBIDI_OPTION_STREAMING</code>
+ * has been set.
+ * <br>
+ * Note that whenever the length of the text affects the execution or the
+ * result of a function, it is the processed length which must be considered,
+ * except for <code>ubidi_setPara</code> (which receives unprocessed source
+ * text) and <code>ubidi_getLength</code> (which returns the original length
+ * of the source text).<br>
+ * In particular, the processed length is the one to consider in the following
+ * cases:
+ * <ul>
+ * <li>maximum value of the <code>limit</code> argument of
+ * <code>ubidi_setLine</code></li>
+ * <li>maximum value of the <code>charIndex</code> argument of
+ * <code>ubidi_getParagraph</code></li>
+ * <li>maximum value of the <code>charIndex</code> argument of
+ * <code>ubidi_getLevelAt</code></li>
+ * <li>number of elements in the array returned by <code>ubidi_getLevels</code></li>
+ * <li>maximum value of the <code>logicalStart</code> argument of
+ * <code>ubidi_getLogicalRun</code></li>
+ * <li>maximum value of the <code>logicalIndex</code> argument of
+ * <code>ubidi_getVisualIndex</code></li>
+ * <li>number of elements filled in the <code>*indexMap</code> argument of
+ * <code>ubidi_getLogicalMap</code></li>
+ * <li>length of text processed by <code>ubidi_writeReordered</code></li>
+ * </ul>
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @return The length of the part of the source text processed by
+ *         the last call to <code>ubidi_setPara</code>.
+ * @see ubidi_setPara
+ * @see UBIDI_OPTION_STREAMING
+ * @stable ICU 3.6
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getProcessedLength(const UBiDi *pBiDi);
+
+/**
+ * Get the length of the reordered text resulting from the last call to
+ * <code>ubidi_setPara()</code>. This length may be different from the length
+ * of the source text if option <code>#UBIDI_OPTION_INSERT_MARKS</code>
+ * or option <code>#UBIDI_OPTION_REMOVE_CONTROLS</code> has been set.
+ * <br>
+ * This resulting length is the one to consider in the following cases:
+ * <ul>
+ * <li>maximum value of the <code>visualIndex</code> argument of
+ * <code>ubidi_getLogicalIndex</code></li>
+ * <li>number of elements of the <code>*indexMap</code> argument of
+ * <code>ubidi_getVisualMap</code></li>
+ * </ul>
+ * Note that this length stays identical to the source text length if
+ * Bidi marks are inserted or removed using option bits of
+ * <code>ubidi_writeReordered</code>, or if option
+ * <code>#UBIDI_REORDER_INVERSE_NUMBERS_AS_L</code> has been set.
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @return The length of the reordered text resulting from
+ *         the last call to <code>ubidi_setPara</code>.
+ * @see ubidi_setPara
+ * @see UBIDI_OPTION_INSERT_MARKS
+ * @see UBIDI_OPTION_REMOVE_CONTROLS
+ * @stable ICU 3.6
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_getResultLength(const UBiDi *pBiDi);
+
+U_CDECL_BEGIN
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Value returned by <code>UBiDiClassCallback</code> callbacks when
+ * there is no need to override the standard Bidi class for a given code point.
+ *
+ * This constant is deprecated; use u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1 instead.
+ *
+ * @see UBiDiClassCallback
+ * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+ */
+#define U_BIDI_CLASS_DEFAULT  U_CHAR_DIRECTION_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+
+/**
+ * Callback type declaration for overriding default Bidi class values with
+ * custom ones.
+ * <p>Usually, the function pointer will be propagated to a <code>UBiDi</code>
+ * object by calling the <code>ubidi_setClassCallback()</code> function;
+ * then the callback will be invoked by the UBA implementation any time the
+ * class of a character is to be determined.</p>
+ *
+ * @param context is a pointer to the callback private data.
+ *
+ * @param c       is the code point to get a Bidi class for.
+ *
+ * @return The directional property / Bidi class for the given code point
+ *         <code>c</code> if the default class has been overridden, or
+ *         <code>u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1</code>
+ *         if the standard Bidi class value for <code>c</code> is to be used.
+ * @see ubidi_setClassCallback
+ * @see ubidi_getClassCallback
+ * @stable ICU 3.6
+ */
+typedef UCharDirection U_CALLCONV
+UBiDiClassCallback(const void *context, UChar32 c);
+
+U_CDECL_END
+
+/**
+ * Retrieve the Bidi class for a given code point.
+ * <p>If a <code>#UBiDiClassCallback</code> callback is defined and returns a
+ * value other than <code>u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS)+1</code>,
+ * that value is used; otherwise the default class determination mechanism is invoked.</p>
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @param c     is the code point whose Bidi class must be retrieved.
+ *
+ * @return The Bidi class for character <code>c</code> based
+ *         on the given <code>pBiDi</code> instance.
+ * @see UBiDiClassCallback
+ * @stable ICU 3.6
+ */
+U_CAPI UCharDirection U_EXPORT2
+ubidi_getCustomizedClass(UBiDi *pBiDi, UChar32 c);
+
+/**
+ * Set the callback function and callback data used by the UBA
+ * implementation for Bidi class determination.
+ * <p>This may be useful for assigning Bidi classes to PUA characters, or
+ * for special application needs. For instance, an application may want to
+ * handle all spaces like L or R characters (according to the base direction)
+ * when creating the visual ordering of logical lines which are part of a report
+ * organized in columns: there should not be interaction between adjacent
+ * cells.<p>
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @param newFn is the new callback function pointer.
+ *
+ * @param newContext is the new callback context pointer. This can be NULL.
+ *
+ * @param oldFn fillin: Returns the old callback function pointer. This can be
+ *                      NULL.
+ *
+ * @param oldContext fillin: Returns the old callback's context. This can be
+ *                           NULL.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @see ubidi_getClassCallback
+ * @stable ICU 3.6
+ */
+U_CAPI void U_EXPORT2
+ubidi_setClassCallback(UBiDi *pBiDi, UBiDiClassCallback *newFn,
+                       const void *newContext, UBiDiClassCallback **oldFn,
+                       const void **oldContext, UErrorCode *pErrorCode);
+
+/**
+ * Get the current callback function used for Bidi class determination.
+ *
+ * @param pBiDi is the paragraph <code>UBiDi</code> object.
+ *
+ * @param fn fillin: Returns the callback function pointer.
+ *
+ * @param context fillin: Returns the callback's private context.
+ *
+ * @see ubidi_setClassCallback
+ * @stable ICU 3.6
+ */
+U_CAPI void U_EXPORT2
+ubidi_getClassCallback(UBiDi *pBiDi, UBiDiClassCallback **fn, const void **context);
+
+/**
+ * Take a <code>UBiDi</code> object containing the reordering
+ * information for a piece of text (one or more paragraphs) set by
+ * <code>ubidi_setPara()</code> or for a line of text set by
+ * <code>ubidi_setLine()</code> and write a reordered string to the
+ * destination buffer.
+ *
+ * This function preserves the integrity of characters with multiple
+ * code units and (optionally) combining characters.
+ * Characters in RTL runs can be replaced by mirror-image characters
+ * in the destination buffer. Note that "real" mirroring has
+ * to be done in a rendering engine by glyph selection
+ * and that for many "mirrored" characters there are no
+ * Unicode characters as mirror-image equivalents.
+ * There are also options to insert or remove Bidi control
+ * characters; see the description of the <code>destSize</code>
+ * and <code>options</code> parameters and of the option bit flags.
+ *
+ * @param pBiDi A pointer to a <code>UBiDi</code> object that
+ *              is set by <code>ubidi_setPara()</code> or
+ *              <code>ubidi_setLine()</code> and contains the reordering
+ *              information for the text that it was defined for,
+ *              as well as a pointer to that text.<br><br>
+ *              The text was aliased (only the pointer was stored
+ *              without copying the contents) and must not have been modified
+ *              since the <code>ubidi_setPara()</code> call.
+ *
+ * @param dest A pointer to where the reordered text is to be copied.
+ *             The source text and <code>dest[destSize]</code>
+ *             must not overlap.
+ *
+ * @param destSize The size of the <code>dest</code> buffer,
+ *                 in number of UChars.
+ *                 If the <code>UBIDI_INSERT_LRM_FOR_NUMERIC</code>
+ *                 option is set, then the destination length could be
+ *                 as large as
+ *                 <code>ubidi_getLength(pBiDi)+2*ubidi_countRuns(pBiDi)</code>.
+ *                 If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
+ *                 is set, then the destination length may be less than
+ *                 <code>ubidi_getLength(pBiDi)</code>.
+ *                 If none of these options is set, then the destination length
+ *                 will be exactly <code>ubidi_getProcessedLength(pBiDi)</code>.
+ *
+ * @param options A bit set of options for the reordering that control
+ *                how the reordered text is written.
+ *                The options include mirroring the characters on a code
+ *                point basis and inserting LRM characters, which is used
+ *                especially for transforming visually stored text
+ *                to logically stored text (although this is still an
+ *                imperfect implementation of an "inverse Bidi" algorithm
+ *                because it uses the "forward Bidi" algorithm at its core).
+ *                The available options are:
+ *                <code>#UBIDI_DO_MIRRORING</code>,
+ *                <code>#UBIDI_INSERT_LRM_FOR_NUMERIC</code>,
+ *                <code>#UBIDI_KEEP_BASE_COMBINING</code>,
+ *                <code>#UBIDI_OUTPUT_REVERSE</code>,
+ *                <code>#UBIDI_REMOVE_BIDI_CONTROLS</code>
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The length of the output string.
+ *
+ * @see ubidi_getProcessedLength
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_writeReordered(UBiDi *pBiDi,
+                     UChar *dest, int32_t destSize,
+                     uint16_t options,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Reverse a Right-To-Left run of Unicode text.
+ *
+ * This function preserves the integrity of characters with multiple
+ * code units and (optionally) combining characters.
+ * Characters can be replaced by mirror-image characters
+ * in the destination buffer. Note that "real" mirroring has
+ * to be done in a rendering engine by glyph selection
+ * and that for many "mirrored" characters there are no
+ * Unicode characters as mirror-image equivalents.
+ * There are also options to insert or remove Bidi control
+ * characters.
+ *
+ * This function is the implementation for reversing RTL runs as part
+ * of <code>ubidi_writeReordered()</code>. For detailed descriptions
+ * of the parameters, see there.
+ * Since no Bidi controls are inserted here, the output string length
+ * will never exceed <code>srcLength</code>.
+ *
+ * @see ubidi_writeReordered
+ *
+ * @param src A pointer to the RTL run text.
+ *
+ * @param srcLength The length of the RTL run.
+ *
+ * @param dest A pointer to where the reordered text is to be copied.
+ *             <code>src[srcLength]</code> and <code>dest[destSize]</code>
+ *             must not overlap.
+ *
+ * @param destSize The size of the <code>dest</code> buffer,
+ *                 in number of UChars.
+ *                 If the <code>UBIDI_REMOVE_BIDI_CONTROLS</code> option
+ *                 is set, then the destination length may be less than
+ *                 <code>srcLength</code>.
+ *                 If this option is not set, then the destination length
+ *                 will be exactly <code>srcLength</code>.
+ *
+ * @param options A bit set of options for the reordering that control
+ *                how the reordered text is written.
+ *                See the <code>options</code> parameter in <code>ubidi_writeReordered()</code>.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value.
+ *
+ * @return The length of the output string.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubidi_writeReverse(const UChar *src, int32_t srcLength,
+                   UChar *dest, int32_t destSize,
+                   uint16_t options,
+                   UErrorCode *pErrorCode);
+
+/*#define BIDI_SAMPLE_CODE*/
+/*@}*/
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ubiditransform.h b/thirdparty/icu4c/common/unicode/ubiditransform.h
new file mode 100644
index 0000000000..2dd7564010
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ubiditransform.h
@@ -0,0 +1,326 @@
+/*
+******************************************************************************
+*
+* © 2016 and later: Unicode, Inc. and others.
+* License & terms of use: http://www.unicode.org/copyright.html
+*
+******************************************************************************
+*   file name:  ubiditransform.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2016jul24
+*   created by: Lina Kemmel
+*
+*/
+
+#ifndef UBIDITRANSFORM_H
+#define UBIDITRANSFORM_H
+
+#include "unicode/utypes.h"
+#include "unicode/ubidi.h"
+#include "unicode/uchar.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief Bidi Transformations
+ */
+
+/**
+ * `UBiDiOrder` indicates the order of text.
+ *
+ * This bidi transformation engine supports all possible combinations (4 in
+ * total) of input and output text order:
+ *
+ *   - <logical input, visual output>: unless the output direction is RTL, this
+ *     corresponds to a normal operation of the Bidi algorithm as described in the
+ *     Unicode Technical Report and implemented by `UBiDi` when the
+ *     reordering mode is set to `UBIDI_REORDER_DEFAULT`. Visual RTL
+ *     mode is not supported by `UBiDi` and is accomplished through
+ *     reversing a visual LTR string,
+ *
+ *   - <visual input, logical output>: unless the input direction is RTL, this
+ *     corresponds to an "inverse bidi algorithm" in `UBiDi` with the
+ *     reordering mode set to `UBIDI_REORDER_INVERSE_LIKE_DIRECT`.
+ *     Visual RTL mode is not not supported by `UBiDi` and is
+ *     accomplished through reversing a visual LTR string,
+ *
+ *   - <logical input, logical output>: if the input and output base directions
+ *     mismatch, this corresponds to the `UBiDi` implementation with the
+ *     reordering mode set to `UBIDI_REORDER_RUNS_ONLY`; and if the
+ *     input and output base directions are identical, the transformation engine
+ *     will only handle character mirroring and Arabic shaping operations without
+ *     reordering,
+ *
+ *   - <visual input, visual output>: this reordering mode is not supported by
+ *     the `UBiDi` engine; it implies character mirroring, Arabic
+ *     shaping, and - if the input/output base directions mismatch -  string
+ *     reverse operations.
+ * @see ubidi_setInverse
+ * @see ubidi_setReorderingMode
+ * @see UBIDI_REORDER_DEFAULT
+ * @see UBIDI_REORDER_INVERSE_LIKE_DIRECT
+ * @see UBIDI_REORDER_RUNS_ONLY
+ * @stable ICU 58
+ */
+typedef enum {
+    /** 0: Constant indicating a logical order.
+      * This is the default for input text.
+      * @stable ICU 58
+      */
+    UBIDI_LOGICAL = 0,
+    /** 1: Constant indicating a visual order.
+      * This is a default for output text.
+      * @stable ICU 58
+      */
+    UBIDI_VISUAL
+} UBiDiOrder;
+
+/**
+ * <code>UBiDiMirroring</code> indicates whether or not characters with the
+ * "mirrored" property in RTL runs should be replaced with their mirror-image
+ * counterparts.
+ * @see UBIDI_DO_MIRRORING
+ * @see ubidi_setReorderingOptions
+ * @see ubidi_writeReordered
+ * @see ubidi_writeReverse
+ * @stable ICU 58
+ */
+typedef enum {
+    /** 0: Constant indicating that character mirroring should not be
+      * performed.
+      * This is the default.
+      * @stable ICU 58
+      */
+    UBIDI_MIRRORING_OFF = 0,
+    /** 1: Constant indicating that character mirroring should be performed.
+      * This corresponds to calling <code>ubidi_writeReordered</code> or
+      * <code>ubidi_writeReverse</code> with the
+      * <code>UBIDI_DO_MIRRORING</code> option bit set.
+      * @stable ICU 58
+      */
+    UBIDI_MIRRORING_ON
+} UBiDiMirroring;
+
+/**
+ * Forward declaration of the <code>UBiDiTransform</code> structure that stores
+ * information used by the layout transformation engine.
+ * @stable ICU 58
+ */
+typedef struct UBiDiTransform UBiDiTransform;
+
+/**
+ * Performs transformation of text from the bidi layout defined by the input
+ * ordering scheme to the bidi layout defined by the output ordering scheme,
+ * and applies character mirroring and Arabic shaping operations.<p>
+ * In terms of <code>UBiDi</code>, such a transformation implies:
+ * <ul>
+ * <li>calling <code>ubidi_setReorderingMode</code> as needed (when the
+ * reordering mode is other than normal),</li>
+ * <li>calling <code>ubidi_setInverse</code> as needed (when text should be
+ * transformed from a visual to a logical form),</li>
+ * <li>resolving embedding levels of each character in the input text by
+ * calling <code>ubidi_setPara</code>,</li>
+ * <li>reordering the characters based on the computed embedding levels, also
+ * performing character mirroring as needed, and streaming the result to the
+ * output, by calling <code>ubidi_writeReordered</code>,</li>
+ * <li>performing Arabic digit and letter shaping on the output text by calling
+ * <code>u_shapeArabic</code>.</li>
+ * </ul>
+ * An "ordering scheme" encompasses the base direction and the order of text,
+ * and these characteristics must be defined by the caller for both input and
+ * output explicitly .<p>
+ * There are 36 possible combinations of <input, output> ordering schemes,
+ * which are partially supported by <code>UBiDi</code> already. Examples of the
+ * currently supported combinations:
+ * <ul>
+ * <li><Logical LTR, Visual LTR>: this is equivalent to calling
+ * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_LTR</code>,</li>
+ * <li><Logical RTL, Visual LTR>: this is equivalent to calling
+ * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_RTL</code>,</li>
+ * <li><Logical Default ("Auto") LTR, Visual LTR>: this is equivalent to
+ * calling <code>ubidi_setPara</code> with
+ * <code>paraLevel == UBIDI_DEFAULT_LTR</code>,</li>
+ * <li><Logical Default ("Auto") RTL, Visual LTR>: this is equivalent to
+ * calling <code>ubidi_setPara</code> with
+ * <code>paraLevel == UBIDI_DEFAULT_RTL</code>,</li>
+ * <li><Visual LTR, Logical LTR>: this is equivalent to
+ * calling <code>ubidi_setInverse(UBiDi*, true)</code> and then
+ * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_LTR</code>,</li>
+ * <li><Visual LTR, Logical RTL>: this is equivalent to
+ * calling <code>ubidi_setInverse(UBiDi*, true)</code> and then
+ * <code>ubidi_setPara</code> with <code>paraLevel == UBIDI_RTL</code>.</li>
+ * </ul>
+ * All combinations that involve the Visual RTL scheme are unsupported by
+ * <code>UBiDi</code>, for instance:
+ * <ul>
+ * <li><Logical LTR, Visual RTL>,</li>
+ * <li><Visual RTL, Logical RTL>.</li>
+ * </ul>
+ * <p>Example of usage of the transformation engine:<br>
+ * <pre>
+ * \code
+ * UChar text1[] = {'a', 'b', 'c', 0x0625, '1', 0};
+ * UChar text2[] = {'a', 'b', 'c', 0x0625, '1', 0};
+ * UErrorCode errorCode = U_ZERO_ERROR;
+ * // Run a transformation.
+ * ubiditransform_transform(pBidiTransform,
+ *          text1, -1, text2, -1,
+ *          UBIDI_LTR, UBIDI_VISUAL,
+ *          UBIDI_RTL, UBIDI_LOGICAL,
+ *          UBIDI_MIRRORING_OFF,
+ *          U_SHAPE_DIGITS_AN2EN | U_SHAPE_DIGIT_TYPE_AN_EXTENDED,
+ *          &errorCode);
+ * // Do something with text2.
+ *  text2[4] = '2';
+ * // Run a reverse transformation.
+ * ubiditransform_transform(pBidiTransform,
+ *          text2, -1, text1, -1,
+ *          UBIDI_RTL, UBIDI_LOGICAL,
+ *          UBIDI_LTR, UBIDI_VISUAL,
+ *          UBIDI_MIRRORING_OFF,
+ *          U_SHAPE_DIGITS_EN2AN | U_SHAPE_DIGIT_TYPE_AN_EXTENDED,
+ *          &errorCode);
+ *\endcode
+ * </pre>
+ * </p>
+ *
+ * @param pBiDiTransform A pointer to a <code>UBiDiTransform</code> object
+ *        allocated with <code>ubiditransform_open()</code> or
+ *        <code>NULL</code>.<p>
+ *        This object serves for one-time setup to amortize initialization
+ *        overheads. Use of this object is not thread-safe. All other threads
+ *        should allocate a new <code>UBiDiTransform</code> object by calling
+ *        <code>ubiditransform_open()</code> before using it. Alternatively,
+ *        a caller can set this parameter to <code>NULL</code>, in which case
+ *        the object will be allocated by the engine on the fly.</p>
+ * @param src A pointer to the text that the Bidi layout transformations will
+ *        be performed on.
+ *        <p><strong>Note:</strong> the text must be (at least)
+ *        <code>srcLength</code> long.</p>
+ * @param srcLength The length of the text, in number of UChars. If
+ *        <code>length == -1</code> then the text must be zero-terminated.
+ * @param dest A pointer to where the processed text is to be copied.
+ * @param destSize The size of the <code>dest</code> buffer, in number of
+ *        UChars. If the <code>U_SHAPE_LETTERS_UNSHAPE</code> option is set,
+ *        then the destination length could be as large as
+ *        <code>srcLength * 2</code>. Otherwise, the destination length will
+ *        not exceed <code>srcLength</code>. If the caller reserves the last
+ *        position for zero-termination, it should be excluded from
+ *        <code>destSize</code>.
+ *        <p><code>destSize == -1</code> is allowed and makes sense when
+ *        <code>dest</code> was holds some meaningful value, e.g. that of
+ *        <code>src</code>. In this case <code>dest</code> must be
+ *        zero-terminated.</p>
+ * @param inParaLevel A base embedding level of the input as defined in
+ *        <code>ubidi_setPara</code> documentation for the
+ *        <code>paraLevel</code> parameter.
+ * @param inOrder An order of the input, which can be one of the
+ *        <code>UBiDiOrder</code> values.
+ * @param outParaLevel A base embedding level of the output as defined in
+ *        <code>ubidi_setPara</code> documentation for the
+ *        <code>paraLevel</code> parameter.
+ * @param outOrder An order of the output, which can be one of the
+ *        <code>UBiDiOrder</code> values.
+ * @param doMirroring Indicates whether or not to perform character mirroring,
+ *        and can accept one of the <code>UBiDiMirroring</code> values.
+ * @param shapingOptions Arabic digit and letter shaping options defined in the
+ *        ushape.h documentation.
+ *        <p><strong>Note:</strong> Direction indicator options are computed by
+ *        the transformation engine based on the effective ordering schemes, so
+ *        user-defined direction indicators will be ignored.</p>
+ * @param pErrorCode A pointer to an error code value.
+ *
+ * @return The destination length, i.e. the number of UChars written to
+ *         <code>dest</code>. If the transformation fails, the return value
+ *         will be 0 (and the error code will be written to
+ *         <code>pErrorCode</code>).
+ *
+ * @see UBiDiLevel
+ * @see UBiDiOrder
+ * @see UBiDiMirroring
+ * @see ubidi_setPara
+ * @see u_shapeArabic
+ * @stable ICU 58
+ */
+U_CAPI uint32_t U_EXPORT2
+ubiditransform_transform(UBiDiTransform *pBiDiTransform,
+            const UChar *src, int32_t srcLength,
+            UChar *dest, int32_t destSize,
+            UBiDiLevel inParaLevel, UBiDiOrder inOrder,
+            UBiDiLevel outParaLevel, UBiDiOrder outOrder,
+            UBiDiMirroring doMirroring, uint32_t shapingOptions,
+            UErrorCode *pErrorCode);
+
+/**
+ * Allocates a <code>UBiDiTransform</code> object. This object can be reused,
+ * e.g. with different ordering schemes, mirroring or shaping options.<p>
+ * <strong>Note:</strong>The object can only be reused in the same thread.
+ * All other threads should allocate a new <code>UBiDiTransform</code> object
+ * before using it.<p>
+ * Example of usage:<p>
+ * <pre>
+ * \code
+ * UErrorCode errorCode = U_ZERO_ERROR;
+ * // Open a new UBiDiTransform.
+ * UBiDiTransform* transform = ubiditransform_open(&errorCode);
+ * // Run a transformation.
+ * ubiditransform_transform(transform,
+ *          text1, -1, text2, -1,
+ *          UBIDI_RTL, UBIDI_LOGICAL,
+ *          UBIDI_LTR, UBIDI_VISUAL,
+ *          UBIDI_MIRRORING_ON,
+ *          U_SHAPE_DIGITS_EN2AN,
+ *          &errorCode);
+ * // Do something with the output text and invoke another transformation using
+ * //   that text as input.
+ * ubiditransform_transform(transform,
+ *          text2, -1, text3, -1,
+ *          UBIDI_LTR, UBIDI_VISUAL,
+ *          UBIDI_RTL, UBIDI_VISUAL,
+ *          UBIDI_MIRRORING_ON,
+ *          0, &errorCode);
+ *\endcode
+ * </pre>
+ * <p>
+ * The <code>UBiDiTransform</code> object must be deallocated by calling
+ * <code>ubiditransform_close()</code>.
+ *
+ * @return An empty <code>UBiDiTransform</code> object.
+ * @stable ICU 58
+ */
+U_CAPI UBiDiTransform* U_EXPORT2
+ubiditransform_open(UErrorCode *pErrorCode);
+
+/**
+ * Deallocates the given <code>UBiDiTransform</code> object.
+ * @stable ICU 58
+ */
+U_CAPI void U_EXPORT2
+ubiditransform_close(UBiDiTransform *pBidiTransform);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUBiDiTransformPointer
+ * "Smart pointer" class, closes a UBiDiTransform via ubiditransform_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 58
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUBiDiTransformPointer, UBiDiTransform, ubiditransform_close);
+
+U_NAMESPACE_END
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ubrk.h b/thirdparty/icu4c/common/unicode/ubrk.h
new file mode 100644
index 0000000000..37189a8598
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ubrk.h
@@ -0,0 +1,631 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1996-2015, International Business Machines Corporation and others.
+* All Rights Reserved.
+******************************************************************************
+*/
+
+#ifndef UBRK_H
+#define UBRK_H
+
+#include "unicode/utypes.h"
+#include "unicode/uloc.h"
+#include "unicode/utext.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * A text-break iterator.
+ *  For usage in C programs.
+ */
+#ifndef UBRK_TYPEDEF_UBREAK_ITERATOR
+#   define UBRK_TYPEDEF_UBREAK_ITERATOR
+    /**
+     *  Opaque type representing an ICU Break iterator object.
+     *  @stable ICU 2.0
+     */
+    typedef struct UBreakIterator UBreakIterator;
+#endif
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/parseerr.h"
+
+/**
+ * \file
+ * \brief C API: BreakIterator
+ *
+ * <h2> BreakIterator C API </h2>
+ *
+ * The BreakIterator C API defines  methods for finding the location
+ * of boundaries in text. Pointer to a UBreakIterator maintain a
+ * current position and scan over text returning the index of characters
+ * where boundaries occur.
+ * <p>
+ * Line boundary analysis determines where a text string can be broken
+ * when line-wrapping. The mechanism correctly handles punctuation and
+ * hyphenated words.
+ * <p>
+ * Note: The locale keyword "lb" can be used to modify line break
+ * behavior according to the CSS level 3 line-break options, see
+ * <http://dev.w3.org/csswg/css-text/#line-breaking>. For example:
+ * "ja@lb=strict", "zh@lb=loose".
+ * <p>
+ * Sentence boundary analysis allows selection with correct
+ * interpretation of periods within numbers and abbreviations, and
+ * trailing punctuation marks such as quotation marks and parentheses.
+ * <p>
+ * Note: The locale keyword "ss" can be used to enable use of
+ * segmentation suppression data (preventing breaks in English after
+ * abbreviations such as "Mr." or "Est.", for example), as follows:
+ * "en@ss=standard".
+ * <p>
+ * Word boundary analysis is used by search and replace functions, as
+ * well as within text editing applications that allow the user to
+ * select words with a double click. Word selection provides correct
+ * interpretation of punctuation marks within and following
+ * words. Characters that are not part of a word, such as symbols or
+ * punctuation marks, have word-breaks on both sides.
+ * <p>
+ * Character boundary analysis identifies the boundaries of
+ * "Extended Grapheme Clusters", which are groupings of codepoints
+ * that should be treated as character-like units for many text operations.
+ * Please see Unicode Standard Annex #29, Unicode Text Segmentation,
+ * http://www.unicode.org/reports/tr29/ for additional information
+ * on grapheme clusters and guidelines on their use.
+ * <p>
+ * Title boundary analysis locates all positions,
+ * typically starts of words, that should be set to Title Case
+ * when title casing the text.
+ * <p>
+ * The text boundary positions are found according to the rules
+ * described in Unicode Standard Annex #29, Text Boundaries, and
+ * Unicode Standard Annex #14, Line Breaking Properties.  These
+ * are available at http://www.unicode.org/reports/tr14/ and
+ * http://www.unicode.org/reports/tr29/.
+ * <p>
+ * In addition to the plain C API defined in this header file, an
+ * object oriented C++ API with equivalent functionality is defined in the
+ * file brkiter.h.
+ * <p>
+ * Code snippets illustrating the use of the Break Iterator APIs
+ * are available in the ICU User Guide,
+ * http://icu-project.org/userguide/boundaryAnalysis.html
+ * and in the sample program icu/source/samples/break/break.cpp
+ */
+
+/** The possible types of text boundaries.  @stable ICU 2.0 */
+typedef enum UBreakIteratorType {
+  /** Character breaks  @stable ICU 2.0 */
+  UBRK_CHARACTER = 0,
+  /** Word breaks @stable ICU 2.0 */
+  UBRK_WORD = 1,
+  /** Line breaks @stable ICU 2.0 */
+  UBRK_LINE = 2,
+  /** Sentence breaks @stable ICU 2.0 */
+  UBRK_SENTENCE = 3,
+
+#ifndef U_HIDE_DEPRECATED_API
+  /**
+   * Title Case breaks
+   * The iterator created using this type locates title boundaries as described for
+   * Unicode 3.2 only. For Unicode 4.0 and above title boundary iteration,
+   * please use Word Boundary iterator.
+   *
+   * @deprecated ICU 2.8 Use the word break iterator for titlecasing for Unicode 4 and later.
+   */
+  UBRK_TITLE = 4,
+    /**
+     * One more than the highest normal UBreakIteratorType value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UBRK_COUNT = 5
+#endif  // U_HIDE_DEPRECATED_API
+} UBreakIteratorType;
+
+/** Value indicating all text boundaries have been returned.
+ *  @stable ICU 2.0
+ */
+#define UBRK_DONE ((int32_t) -1)
+
+
+/**
+ *  Enum constants for the word break tags returned by
+ *  getRuleStatus().  A range of values is defined for each category of
+ *  word, to allow for further subdivisions of a category in future releases.
+ *  Applications should check for tag values falling within the range, rather
+ *  than for single individual values.
+ *
+ * The numeric values of all of these constants are stable (will not change).
+ *
+ * @stable ICU 2.2
+*/
+typedef enum UWordBreak {
+    /** Tag value for "words" that do not fit into any of other categories.
+     *  Includes spaces and most punctuation. */
+    UBRK_WORD_NONE           = 0,
+    /** Upper bound for tags for uncategorized words. */
+    UBRK_WORD_NONE_LIMIT     = 100,
+    /** Tag value for words that appear to be numbers, lower limit.    */
+    UBRK_WORD_NUMBER         = 100,
+    /** Tag value for words that appear to be numbers, upper limit.    */
+    UBRK_WORD_NUMBER_LIMIT   = 200,
+    /** Tag value for words that contain letters, excluding
+     *  hiragana, katakana or ideographic characters, lower limit.    */
+    UBRK_WORD_LETTER         = 200,
+    /** Tag value for words containing letters, upper limit  */
+    UBRK_WORD_LETTER_LIMIT   = 300,
+    /** Tag value for words containing kana characters, lower limit */
+    UBRK_WORD_KANA           = 300,
+    /** Tag value for words containing kana characters, upper limit */
+    UBRK_WORD_KANA_LIMIT     = 400,
+    /** Tag value for words containing ideographic characters, lower limit */
+    UBRK_WORD_IDEO           = 400,
+    /** Tag value for words containing ideographic characters, upper limit */
+    UBRK_WORD_IDEO_LIMIT     = 500
+} UWordBreak;
+
+/**
+ *  Enum constants for the line break tags returned by getRuleStatus().
+ *  A range of values is defined for each category of
+ *  word, to allow for further subdivisions of a category in future releases.
+ *  Applications should check for tag values falling within the range, rather
+ *  than for single individual values.
+ *
+ * The numeric values of all of these constants are stable (will not change).
+ *
+ * @stable ICU 2.8
+*/
+typedef enum ULineBreakTag {
+    /** Tag value for soft line breaks, positions at which a line break
+      *  is acceptable but not required                */
+    UBRK_LINE_SOFT            = 0,
+    /** Upper bound for soft line breaks.              */
+    UBRK_LINE_SOFT_LIMIT      = 100,
+    /** Tag value for a hard, or mandatory line break  */
+    UBRK_LINE_HARD            = 100,
+    /** Upper bound for hard line breaks.              */
+    UBRK_LINE_HARD_LIMIT      = 200
+} ULineBreakTag;
+
+
+
+/**
+ *  Enum constants for the sentence break tags returned by getRuleStatus().
+ *  A range of values is defined for each category of
+ *  sentence, to allow for further subdivisions of a category in future releases.
+ *  Applications should check for tag values falling within the range, rather
+ *  than for single individual values.
+ *
+ * The numeric values of all of these constants are stable (will not change).
+ *
+ * @stable ICU 2.8
+*/
+typedef enum USentenceBreakTag {
+    /** Tag value for for sentences  ending with a sentence terminator
+      * ('.', '?', '!', etc.) character, possibly followed by a
+      * hard separator (CR, LF, PS, etc.)
+      */
+    UBRK_SENTENCE_TERM       = 0,
+    /** Upper bound for tags for sentences ended by sentence terminators.    */
+    UBRK_SENTENCE_TERM_LIMIT = 100,
+    /** Tag value for for sentences that do not contain an ending
+      * sentence terminator ('.', '?', '!', etc.) character, but
+      * are ended only by a hard separator (CR, LF, PS, etc.) or end of input.
+      */
+    UBRK_SENTENCE_SEP        = 100,
+    /** Upper bound for tags for sentences ended by a separator.              */
+    UBRK_SENTENCE_SEP_LIMIT  = 200
+    /** Tag value for a hard, or mandatory line break  */
+} USentenceBreakTag;
+
+
+/**
+ * Open a new UBreakIterator for locating text boundaries for a specified locale.
+ * A UBreakIterator may be used for detecting character, line, word,
+ * and sentence breaks in text.
+ * @param type The type of UBreakIterator to open: one of UBRK_CHARACTER, UBRK_WORD,
+ * UBRK_LINE, UBRK_SENTENCE
+ * @param locale The locale specifying the text-breaking conventions. Note that
+ * locale keys such as "lb" and "ss" may be used to modify text break behavior,
+ * see general discussion of BreakIterator C API.
+ * @param text The text to be iterated over. May be null, in which case ubrk_setText() is
+ *        used to specify the text to be iterated.
+ * @param textLength The number of characters in text, or -1 if null-terminated.
+ * @param status A UErrorCode to receive any errors.
+ * @return A UBreakIterator for the specified locale.
+ * @see ubrk_openRules
+ * @stable ICU 2.0
+ */
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_open(UBreakIteratorType type,
+      const char *locale,
+      const UChar *text,
+      int32_t textLength,
+      UErrorCode *status);
+
+/**
+ * Open a new UBreakIterator for locating text boundaries using specified breaking rules.
+ * The rule syntax is ... (TBD)
+ * @param rules A set of rules specifying the text breaking conventions.
+ * @param rulesLength The number of characters in rules, or -1 if null-terminated.
+ * @param text The text to be iterated over.  May be null, in which case ubrk_setText() is
+ *        used to specify the text to be iterated.
+ * @param textLength The number of characters in text, or -1 if null-terminated.
+ * @param parseErr   Receives position and context information for any syntax errors
+ *                   detected while parsing the rules.
+ * @param status A UErrorCode to receive any errors.
+ * @return A UBreakIterator for the specified rules.
+ * @see ubrk_open
+ * @stable ICU 2.2
+ */
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_openRules(const UChar     *rules,
+               int32_t         rulesLength,
+               const UChar     *text,
+               int32_t          textLength,
+               UParseError     *parseErr,
+               UErrorCode      *status);
+
+/**
+ * Open a new UBreakIterator for locating text boundaries using precompiled binary rules.
+ * Opening a UBreakIterator this way is substantially faster than using ubrk_openRules.
+ * Binary rules may be obtained using ubrk_getBinaryRules. The compiled rules are not
+ * compatible across different major versions of ICU, nor across platforms of different
+ * endianness or different base character set family (ASCII vs EBCDIC).
+ * @param binaryRules A set of compiled binary rules specifying the text breaking
+ *                    conventions. Ownership of the storage containing the compiled
+ *                    rules remains with the caller of this function. The compiled
+ *                    rules must not be modified or deleted during the life of the
+ *                    break iterator.
+ * @param rulesLength The length of binaryRules in bytes; must be >= 0.
+ * @param text        The text to be iterated over.  May be null, in which case
+ *                    ubrk_setText() is used to specify the text to be iterated.
+ * @param textLength  The number of characters in text, or -1 if null-terminated.
+ * @param status      Pointer to UErrorCode to receive any errors.
+ * @return            UBreakIterator for the specified rules.
+ * @see ubrk_getBinaryRules
+ * @stable ICU 59
+ */
+U_CAPI UBreakIterator* U_EXPORT2
+ubrk_openBinaryRules(const uint8_t *binaryRules, int32_t rulesLength,
+                     const UChar *  text, int32_t textLength,
+                     UErrorCode *   status);
+
+/**
+ * Thread safe cloning operation
+ * @param bi iterator to be cloned
+ * @param stackBuffer <em>Deprecated functionality as of ICU 52, use NULL.</em><br>
+ *  user allocated space for the new clone. If NULL new memory will be allocated.
+ *  If buffer is not large enough, new memory will be allocated.
+ *  Clients can use the U_BRK_SAFECLONE_BUFFERSIZE.
+ * @param pBufferSize <em>Deprecated functionality as of ICU 52, use NULL or 1.</em><br>
+ *  pointer to size of allocated space.
+ *  If *pBufferSize == 0, a sufficient size for use in cloning will
+ *  be returned ('pre-flighting')
+ *  If *pBufferSize is not enough for a stack-based safe clone,
+ *  new memory will be allocated.
+ * @param status to indicate whether the operation went on smoothly or there were errors
+ *  An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any allocations were necessary.
+ * @return pointer to the new clone
+ * @stable ICU 2.0
+ */
+U_CAPI UBreakIterator * U_EXPORT2
+ubrk_safeClone(
+          const UBreakIterator *bi,
+          void *stackBuffer,
+          int32_t *pBufferSize,
+          UErrorCode *status);
+
+#ifndef U_HIDE_DEPRECATED_API
+
+/**
+  * A recommended size (in bytes) for the memory buffer to be passed to ubrk_saveClone().
+  * @deprecated ICU 52. Do not rely on ubrk_safeClone() cloning into any provided buffer.
+  */
+#define U_BRK_SAFECLONE_BUFFERSIZE 1
+
+#endif /* U_HIDE_DEPRECATED_API */
+
+/**
+* Close a UBreakIterator.
+* Once closed, a UBreakIterator may no longer be used.
+* @param bi The break iterator to close.
+ * @stable ICU 2.0
+*/
+U_CAPI void U_EXPORT2
+ubrk_close(UBreakIterator *bi);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUBreakIteratorPointer
+ * "Smart pointer" class, closes a UBreakIterator via ubrk_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUBreakIteratorPointer, UBreakIterator, ubrk_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Sets an existing iterator to point to a new piece of text.
+ * The break iterator retains a pointer to the supplied text.
+ * The caller must not modify or delete the text while the BreakIterator
+ * retains the reference.
+ *
+ * @param bi The iterator to use
+ * @param text The text to be set
+ * @param textLength The length of the text
+ * @param status The error code
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ubrk_setText(UBreakIterator* bi,
+             const UChar*    text,
+             int32_t         textLength,
+             UErrorCode*     status);
+
+
+/**
+ * Sets an existing iterator to point to a new piece of text.
+ *
+ * All index positions returned by break iterator functions are
+ * native indices from the UText. For example, when breaking UTF-8
+ * encoded text, the break positions returned by \ref ubrk_next, \ref ubrk_previous, etc.
+ * will be UTF-8 string indices, not UTF-16 positions.
+ *
+ * @param bi The iterator to use
+ * @param text The text to be set.
+ *             This function makes a shallow clone of the supplied UText.  This means
+ *             that the caller is free to immediately close or otherwise reuse the
+ *             UText that was passed as a parameter, but that the underlying text itself
+ *             must not be altered while being referenced by the break iterator.
+ * @param status The error code
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ubrk_setUText(UBreakIterator* bi,
+             UText*          text,
+             UErrorCode*     status);
+
+
+
+/**
+ * Determine the most recently-returned text boundary.
+ *
+ * @param bi The break iterator to use.
+ * @return The character index most recently returned by \ref ubrk_next, \ref ubrk_previous,
+ * \ref ubrk_first, or \ref ubrk_last.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_current(const UBreakIterator *bi);
+
+/**
+ * Advance the iterator to the boundary following the current boundary.
+ *
+ * @param bi The break iterator to use.
+ * @return The character index of the next text boundary, or UBRK_DONE
+ * if all text boundaries have been returned.
+ * @see ubrk_previous
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_next(UBreakIterator *bi);
+
+/**
+ * Set the iterator position to the boundary preceding the current boundary.
+ *
+ * @param bi The break iterator to use.
+ * @return The character index of the preceding text boundary, or UBRK_DONE
+ * if all text boundaries have been returned.
+ * @see ubrk_next
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_previous(UBreakIterator *bi);
+
+/**
+ * Set the iterator position to zero, the start of the text being scanned.
+ * @param bi The break iterator to use.
+ * @return The new iterator position (zero).
+ * @see ubrk_last
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_first(UBreakIterator *bi);
+
+/**
+ * Set the iterator position to the index immediately <EM>beyond</EM> the last character in the text being scanned.
+ * This is not the same as the last character.
+ * @param bi The break iterator to use.
+ * @return The character offset immediately <EM>beyond</EM> the last character in the
+ * text being scanned.
+ * @see ubrk_first
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_last(UBreakIterator *bi);
+
+/**
+ * Set the iterator position to the first boundary preceding the specified offset.
+ * The new position is always smaller than offset, or UBRK_DONE.
+ * @param bi The break iterator to use.
+ * @param offset The offset to begin scanning.
+ * @return The text boundary preceding offset, or UBRK_DONE.
+ * @see ubrk_following
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_preceding(UBreakIterator *bi,
+           int32_t offset);
+
+/**
+ * Advance the iterator to the first boundary following the specified offset.
+ * The value returned is always greater than offset, or UBRK_DONE.
+ * @param bi The break iterator to use.
+ * @param offset The offset to begin scanning.
+ * @return The text boundary following offset, or UBRK_DONE.
+ * @see ubrk_preceding
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_following(UBreakIterator *bi,
+           int32_t offset);
+
+/**
+* Get a locale for which text breaking information is available.
+* A UBreakIterator in a locale returned by this function will perform the correct
+* text breaking for the locale.
+* @param index The index of the desired locale.
+* @return A locale for which number text breaking information is available, or 0 if none.
+* @see ubrk_countAvailable
+* @stable ICU 2.0
+*/
+U_CAPI const char* U_EXPORT2
+ubrk_getAvailable(int32_t index);
+
+/**
+* Determine how many locales have text breaking information available.
+* This function is most useful as determining the loop ending condition for
+* calls to \ref ubrk_getAvailable.
+* @return The number of locales for which text breaking information is available.
+* @see ubrk_getAvailable
+* @stable ICU 2.0
+*/
+U_CAPI int32_t U_EXPORT2
+ubrk_countAvailable(void);
+
+
+/**
+* Returns true if the specified position is a boundary position.  As a side
+* effect, leaves the iterator pointing to the first boundary position at
+* or after "offset".
+* @param bi The break iterator to use.
+* @param offset the offset to check.
+* @return True if "offset" is a boundary position.
+* @stable ICU 2.0
+*/
+U_CAPI  UBool U_EXPORT2
+ubrk_isBoundary(UBreakIterator *bi, int32_t offset);
+
+/**
+ * Return the status from the break rule that determined the most recently
+ * returned break position.  The values appear in the rule source
+ * within brackets, {123}, for example.  For rules that do not specify a
+ * status, a default value of 0 is returned.
+ * <p>
+ * For word break iterators, the possible values are defined in enum UWordBreak.
+ * @stable ICU 2.2
+ */
+U_CAPI  int32_t U_EXPORT2
+ubrk_getRuleStatus(UBreakIterator *bi);
+
+/**
+ * Get the statuses from the break rules that determined the most recently
+ * returned break position.  The values appear in the rule source
+ * within brackets, {123}, for example.  The default status value for rules
+ * that do not explicitly provide one is zero.
+ * <p>
+ * For word break iterators, the possible values are defined in enum UWordBreak.
+ * @param bi        The break iterator to use
+ * @param fillInVec an array to be filled in with the status values.
+ * @param capacity  the length of the supplied vector.  A length of zero causes
+ *                  the function to return the number of status values, in the
+ *                  normal way, without attempting to store any values.
+ * @param status    receives error codes.
+ * @return          The number of rule status values from rules that determined
+ *                  the most recent boundary returned by the break iterator.
+ * @stable ICU 3.0
+ */
+U_CAPI  int32_t U_EXPORT2
+ubrk_getRuleStatusVec(UBreakIterator *bi, int32_t *fillInVec, int32_t capacity, UErrorCode *status);
+
+/**
+ * Return the locale of the break iterator. You can choose between the valid and
+ * the actual locale.
+ * @param bi break iterator
+ * @param type locale type (valid or actual)
+ * @param status error code
+ * @return locale string
+ * @stable ICU 2.8
+ */
+U_CAPI const char* U_EXPORT2
+ubrk_getLocaleByType(const UBreakIterator *bi, ULocDataLocaleType type, UErrorCode* status);
+
+/**
+  *  Set the subject text string upon which the break iterator is operating
+  *  without changing any other aspect of the state.
+  *  The new and previous text strings must have the same content.
+  *
+  *  This function is intended for use in environments where ICU is operating on
+  *  strings that may move around in memory.  It provides a mechanism for notifying
+  *  ICU that the string has been relocated, and providing a new UText to access the
+  *  string in its new position.
+  *
+  *  Note that the break iterator never copies the underlying text
+  *  of a string being processed, but always operates directly on the original text
+  *  provided by the user. Refreshing simply drops the references to the old text
+  *  and replaces them with references to the new.
+  *
+  *  Caution:  this function is normally used only by very specialized
+  *            system-level code.   One example use case is with garbage collection
+  *            that moves the text in memory.
+  *
+  * @param bi         The break iterator.
+  * @param text       The new (moved) text string.
+  * @param status     Receives errors detected by this function.
+  *
+  * @stable ICU 49
+  */
+U_CAPI void U_EXPORT2
+ubrk_refreshUText(UBreakIterator *bi,
+                       UText          *text,
+                       UErrorCode     *status);
+
+
+/**
+ * Get a compiled binary version of the rules specifying the behavior of a UBreakIterator.
+ * The binary rules may be used with ubrk_openBinaryRules to open a new UBreakIterator
+ * more quickly than using ubrk_openRules. The compiled rules are not compatible across
+ * different major versions of ICU, nor across platforms of different endianness or
+ * different base character set family (ASCII vs EBCDIC). Supports preflighting (with
+ * binaryRules=NULL and rulesCapacity=0) to get the rules length without copying them to
+ * the binaryRules buffer. However, whether preflighting or not, if the actual length
+ * is greater than INT32_MAX, then the function returns 0 and sets *status to
+ * U_INDEX_OUTOFBOUNDS_ERROR.
+
+ * @param bi            The break iterator to use.
+ * @param binaryRules   Buffer to receive the compiled binary rules; set to NULL for
+ *                      preflighting.
+ * @param rulesCapacity Capacity (in bytes) of the binaryRules buffer; set to 0 for
+ *                      preflighting. Must be >= 0.
+ * @param status        Pointer to UErrorCode to receive any errors, such as
+ *                      U_BUFFER_OVERFLOW_ERROR, U_INDEX_OUTOFBOUNDS_ERROR, or
+ *                      U_ILLEGAL_ARGUMENT_ERROR.
+ * @return              The actual byte length of the binary rules, if <= INT32_MAX;
+ *                      otherwise 0. If not preflighting and this is larger than
+ *                      rulesCapacity, *status will be set to an error.
+ * @see ubrk_openBinaryRules
+ * @stable ICU 59
+ */
+U_CAPI int32_t U_EXPORT2
+ubrk_getBinaryRules(UBreakIterator *bi,
+                    uint8_t *       binaryRules, int32_t rulesCapacity,
+                    UErrorCode *    status);
+
+#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucasemap.h b/thirdparty/icu4c/common/unicode/ucasemap.h
new file mode 100644
index 0000000000..d1c1b483ab
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucasemap.h
@@ -0,0 +1,388 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2005-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ucasemap.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2005may06
+*   created by: Markus W. Scherer
+*
+*   Case mapping service object and functions using it.
+*/
+
+#ifndef __UCASEMAP_H__
+#define __UCASEMAP_H__
+
+#include "unicode/utypes.h"
+#include "unicode/stringoptions.h"
+#include "unicode/ustring.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C API: Unicode case mapping functions using a UCaseMap service object.
+ *
+ * The service object takes care of memory allocations, data loading, and setup
+ * for the attributes, as usual.
+ *
+ * Currently, the functionality provided here does not overlap with uchar.h
+ * and ustring.h, except for ucasemap_toTitle().
+ *
+ * ucasemap_utf8XYZ() functions operate directly on UTF-8 strings.
+ */
+
+/**
+ * UCaseMap is an opaque service object for newer ICU case mapping functions.
+ * Older functions did not use a service object.
+ * @stable ICU 3.4
+ */
+struct UCaseMap;
+typedef struct UCaseMap UCaseMap; /**< C typedef for struct UCaseMap. @stable ICU 3.4 */
+
+/**
+ * Open a UCaseMap service object for a locale and a set of options.
+ * The locale ID and options are preprocessed so that functions using the
+ * service object need not process them in each call.
+ *
+ * @param locale ICU locale ID, used for language-dependent
+ *               upper-/lower-/title-casing according to the Unicode standard.
+ *               Usual semantics: ""=root, NULL=default locale, etc.
+ * @param options Options bit set, used for case folding and string comparisons.
+ *                Same flags as for u_foldCase(), u_strFoldCase(),
+ *                u_strCaseCompare(), etc.
+ *                Use 0 or U_FOLD_CASE_DEFAULT for default behavior.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ * @return Pointer to a UCaseMap service object, if successful.
+ *
+ * @see U_FOLD_CASE_DEFAULT
+ * @see U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ * @see U_TITLECASE_NO_LOWERCASE
+ * @see U_TITLECASE_NO_BREAK_ADJUSTMENT
+ * @stable ICU 3.4
+ */
+U_CAPI UCaseMap * U_EXPORT2
+ucasemap_open(const char *locale, uint32_t options, UErrorCode *pErrorCode);
+
+/**
+ * Close a UCaseMap service object.
+ * @param csm Object to be closed.
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ucasemap_close(UCaseMap *csm);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUCaseMapPointer
+ * "Smart pointer" class, closes a UCaseMap via ucasemap_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUCaseMapPointer, UCaseMap, ucasemap_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Get the locale ID that is used for language-dependent case mappings.
+ * @param csm UCaseMap service object.
+ * @return locale ID
+ * @stable ICU 3.4
+ */
+U_CAPI const char * U_EXPORT2
+ucasemap_getLocale(const UCaseMap *csm);
+
+/**
+ * Get the options bit set that is used for case folding and string comparisons.
+ * @param csm UCaseMap service object.
+ * @return options bit set
+ * @stable ICU 3.4
+ */
+U_CAPI uint32_t U_EXPORT2
+ucasemap_getOptions(const UCaseMap *csm);
+
+/**
+ * Set the locale ID that is used for language-dependent case mappings.
+ *
+ * @param csm UCaseMap service object.
+ * @param locale Locale ID, see ucasemap_open().
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ *
+ * @see ucasemap_open
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode);
+
+/**
+ * Set the options bit set that is used for case folding and string comparisons.
+ *
+ * @param csm UCaseMap service object.
+ * @param options Options bit set, see ucasemap_open().
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ *
+ * @see ucasemap_open
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+ucasemap_setOptions(UCaseMap *csm, uint32_t options, UErrorCode *pErrorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/**
+ * Get the break iterator that is used for titlecasing.
+ * Do not modify the returned break iterator.
+ * @param csm UCaseMap service object.
+ * @return titlecasing break iterator
+ * @stable ICU 3.8
+ */
+U_CAPI const UBreakIterator * U_EXPORT2
+ucasemap_getBreakIterator(const UCaseMap *csm);
+
+/**
+ * Set the break iterator that is used for titlecasing.
+ * The UCaseMap service object releases a previously set break iterator
+ * and "adopts" this new one, taking ownership of it.
+ * It will be released in a subsequent call to ucasemap_setBreakIterator()
+ * or ucasemap_close().
+ *
+ * Break iterator operations are not thread-safe. Therefore, titlecasing
+ * functions use non-const UCaseMap objects. It is not possible to titlecase
+ * strings concurrently using the same UCaseMap.
+ *
+ * @param csm UCaseMap service object.
+ * @param iterToAdopt Break iterator to be adopted for titlecasing.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ *
+ * @see ucasemap_toTitle
+ * @see ucasemap_utf8ToTitle
+ * @stable ICU 3.8
+ */
+U_CAPI void U_EXPORT2
+ucasemap_setBreakIterator(UCaseMap *csm, UBreakIterator *iterToAdopt, UErrorCode *pErrorCode);
+
+/**
+ * Titlecase a UTF-16 string. This function is almost a duplicate of u_strToTitle(),
+ * except that it takes ucasemap_setOptions() into account and has performance
+ * advantages from being able to use a UCaseMap object for multiple case mapping
+ * operations, saving setup time.
+ *
+ * Casing is locale-dependent and context-sensitive.
+ * Titlecasing uses a break iterator to find the first characters of words
+ * that are to be titlecased. It titlecases those characters and lowercases
+ * all others. (This can be modified with ucasemap_setOptions().)
+ *
+ * Note: This function takes a non-const UCaseMap pointer because it will
+ * open a default break iterator if no break iterator was set yet,
+ * and effectively call ucasemap_setBreakIterator();
+ * also because the break iterator is stateful and will be modified during
+ * the iteration.
+ *
+ * The titlecase break iterator can be provided to customize for arbitrary
+ * styles, using rules and dictionaries beyond the standard iterators.
+ * The standard titlecase iterator for the root locale implements the
+ * algorithm of Unicode TR 21.
+ *
+ * This function uses only the setText(), first() and next() methods of the
+ * provided break iterator.
+ *
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer must not overlap.
+ *
+ * @param csm       UCaseMap service object. This pointer is non-const!
+ *                  See the note above for details.
+ * @param dest      A buffer for the result string. The result will be NUL-terminated if
+ *                  the buffer is large enough.
+ *                  The contents is undefined in case of failure.
+ * @param destCapacity The size of the buffer (number of UChars). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string.
+ * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ *
+ * @see u_strToTitle
+ * @stable ICU 3.8
+ */
+U_CAPI int32_t U_EXPORT2
+ucasemap_toTitle(UCaseMap *csm,
+                 UChar *dest, int32_t destCapacity,
+                 const UChar *src, int32_t srcLength,
+                 UErrorCode *pErrorCode);
+
+#endif  // UCONFIG_NO_BREAK_ITERATION
+
+/**
+ * Lowercase the characters in a UTF-8 string.
+ * Casing is locale-dependent and context-sensitive.
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer must not overlap.
+ *
+ * @param csm       UCaseMap service object.
+ * @param dest      A buffer for the result string. The result will be NUL-terminated if
+ *                  the buffer is large enough.
+ *                  The contents is undefined in case of failure.
+ * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string.
+ * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ *
+ * @see u_strToLower
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToLower(const UCaseMap *csm,
+                     char *dest, int32_t destCapacity,
+                     const char *src, int32_t srcLength,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Uppercase the characters in a UTF-8 string.
+ * Casing is locale-dependent and context-sensitive.
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer must not overlap.
+ *
+ * @param csm       UCaseMap service object.
+ * @param dest      A buffer for the result string. The result will be NUL-terminated if
+ *                  the buffer is large enough.
+ *                  The contents is undefined in case of failure.
+ * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string.
+ * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ *
+ * @see u_strToUpper
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToUpper(const UCaseMap *csm,
+                     char *dest, int32_t destCapacity,
+                     const char *src, int32_t srcLength,
+                     UErrorCode *pErrorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/**
+ * Titlecase a UTF-8 string.
+ * Casing is locale-dependent and context-sensitive.
+ * Titlecasing uses a break iterator to find the first characters of words
+ * that are to be titlecased. It titlecases those characters and lowercases
+ * all others. (This can be modified with ucasemap_setOptions().)
+ *
+ * Note: This function takes a non-const UCaseMap pointer because it will
+ * open a default break iterator if no break iterator was set yet,
+ * and effectively call ucasemap_setBreakIterator();
+ * also because the break iterator is stateful and will be modified during
+ * the iteration.
+ *
+ * The titlecase break iterator can be provided to customize for arbitrary
+ * styles, using rules and dictionaries beyond the standard iterators.
+ * The standard titlecase iterator for the root locale implements the
+ * algorithm of Unicode TR 21.
+ *
+ * This function uses only the setUText(), first(), next() and close() methods of the
+ * provided break iterator.
+ *
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer must not overlap.
+ *
+ * @param csm       UCaseMap service object. This pointer is non-const!
+ *                  See the note above for details.
+ * @param dest      A buffer for the result string. The result will be NUL-terminated if
+ *                  the buffer is large enough.
+ *                  The contents is undefined in case of failure.
+ * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string.
+ * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ *
+ * @see u_strToTitle
+ * @see U_TITLECASE_NO_LOWERCASE
+ * @see U_TITLECASE_NO_BREAK_ADJUSTMENT
+ * @stable ICU 3.8
+ */
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8ToTitle(UCaseMap *csm,
+                    char *dest, int32_t destCapacity,
+                    const char *src, int32_t srcLength,
+                    UErrorCode *pErrorCode);
+
+#endif
+
+/**
+ * Case-folds the characters in a UTF-8 string.
+ *
+ * Case-folding is locale-independent and not context-sensitive,
+ * but there is an option for whether to include or exclude mappings for dotted I
+ * and dotless i that are marked with 'T' in CaseFolding.txt.
+ *
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer must not overlap.
+ *
+ * @param csm       UCaseMap service object.
+ * @param dest      A buffer for the result string. The result will be NUL-terminated if
+ *                  the buffer is large enough.
+ *                  The contents is undefined in case of failure.
+ * @param destCapacity The size of the buffer (number of bytes). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string.
+ * @param srcLength The length of the original string. If -1, then src must be NUL-terminated.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ *
+ * @see u_strFoldCase
+ * @see ucasemap_setOptions
+ * @see U_FOLD_CASE_DEFAULT
+ * @see U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ * @stable ICU 3.8
+ */
+U_CAPI int32_t U_EXPORT2
+ucasemap_utf8FoldCase(const UCaseMap *csm,
+                      char *dest, int32_t destCapacity,
+                      const char *src, int32_t srcLength,
+                      UErrorCode *pErrorCode);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucat.h b/thirdparty/icu4c/common/unicode/ucat.h
new file mode 100644
index 0000000000..93850348ff
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucat.h
@@ -0,0 +1,160 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2003-2004, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: March 19 2003
+* Since: ICU 2.6
+**********************************************************************
+*/
+#ifndef UCAT_H
+#define UCAT_H
+
+#include "unicode/utypes.h"
+#include "unicode/ures.h"
+
+/**
+ * \file
+ * \brief C API: Message Catalog Wrappers
+ *
+ * This C API provides look-alike functions that deliberately resemble
+ * the POSIX catopen, catclose, and catgets functions.  The underlying
+ * implementation is in terms of ICU resource bundles, rather than
+ * POSIX message catalogs.
+ *
+ * The ICU resource bundles obey standard ICU inheritance policies.
+ * To facilitate this, sets and messages are flattened into one tier.
+ * This is done by creating resource bundle keys of the form
+ * &lt;set_num&gt;%&lt;msg_num&gt; where set_num is the set number and msg_num is
+ * the message number, formatted as decimal strings.
+ *
+ * Example:  Consider a message catalog containing two sets:
+ *
+ * Set 1: Message 4  = "Good morning."
+ *        Message 5  = "Good afternoon."
+ *        Message 7  = "Good evening."
+ *        Message 8  = "Good night."
+ * Set 4: Message 14 = "Please "
+ *        Message 19 = "Thank you."
+ *        Message 20 = "Sincerely,"
+ *
+ * The ICU resource bundle source file would, assuming it is named
+ * "greet.txt", would look like this:
+ *
+ * greet
+ * {
+ *     1%4  { "Good morning." }
+ *     1%5  { "Good afternoon." }
+ *     1%7  { "Good evening." }
+ *     1%8  { "Good night." }
+ * 
+ *     4%14 { "Please " }
+ *     4%19 { "Thank you." }
+ *     4%20 { "Sincerely," }
+ * }
+ *
+ * The catgets function is commonly used in combination with functions
+ * like printf and strftime.  ICU components like message format can
+ * be used instead, although they use a different format syntax.
+ * There is an ICU package, icuio, that provides some of
+ * the POSIX-style formatting API.
+ */
+
+U_CDECL_BEGIN
+
+/**
+ * An ICU message catalog descriptor, analogous to nl_catd.
+ * 
+ * @stable ICU 2.6
+ */
+typedef UResourceBundle* u_nl_catd;
+
+/**
+ * Open and return an ICU message catalog descriptor. The descriptor
+ * may be passed to u_catgets() to retrieve localized strings.
+ *
+ * @param name string containing the full path pointing to the
+ * directory where the resources reside followed by the package name
+ * e.g. "/usr/resource/my_app/resources/guimessages" on a Unix system.
+ * If NULL, ICU default data files will be used.
+ *
+ * Unlike POSIX, environment variables are not interpolated within the
+ * name.
+ *
+ * @param locale the locale for which we want to open the resource. If
+ * NULL, the default ICU locale will be used (see uloc_getDefault). If
+ * strlen(locale) == 0, the root locale will be used.
+ *
+ * @param ec input/output error code. Upon output,
+ * U_USING_FALLBACK_WARNING indicates that a fallback locale was
+ * used. For example, 'de_CH' was requested, but nothing was found
+ * there, so 'de' was used. U_USING_DEFAULT_WARNING indicates that the
+ * default locale data or root locale data was used; neither the
+ * requested locale nor any of its fallback locales were found.
+ *
+ * @return a message catalog descriptor that may be passed to
+ * u_catgets(). If the ec parameter indicates success, then the caller
+ * is responsible for calling u_catclose() to close the message
+ * catalog. If the ec parameter indicates failure, then NULL will be
+ * returned.
+ * 
+ * @stable ICU 2.6
+ */
+U_CAPI u_nl_catd U_EXPORT2
+u_catopen(const char* name, const char* locale, UErrorCode* ec);
+
+/**
+ * Close an ICU message catalog, given its descriptor.
+ *
+ * @param catd a message catalog descriptor to be closed. May be NULL,
+ * in which case no action is taken.
+ * 
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+u_catclose(u_nl_catd catd);
+
+/**
+ * Retrieve a localized string from an ICU message catalog.
+ *
+ * @param catd a message catalog descriptor returned by u_catopen.
+ *
+ * @param set_num the message catalog set number. Sets need not be
+ * numbered consecutively.
+ *
+ * @param msg_num the message catalog message number within the
+ * set. Messages need not be numbered consecutively.
+ *
+ * @param s the default string. This is returned if the string
+ * specified by the set_num and msg_num is not found. It must be
+ * zero-terminated.
+ *
+ * @param len fill-in parameter to receive the length of the result.
+ * May be NULL, in which case it is ignored.
+ *
+ * @param ec input/output error code. May be U_USING_FALLBACK_WARNING
+ * or U_USING_DEFAULT_WARNING. U_MISSING_RESOURCE_ERROR indicates that
+ * the set_num/msg_num tuple does not specify a valid message string
+ * in this catalog.
+ *
+ * @return a pointer to a zero-terminated UChar array which lives in
+ * an internal buffer area, typically a memory mapped/DLL file. The
+ * caller must NOT delete this pointer. If the call is unsuccessful
+ * for any reason, then s is returned.  This includes the situation in
+ * which ec indicates a failing error code upon entry to this
+ * function.
+ * 
+ * @stable ICU 2.6
+ */
+U_CAPI const UChar* U_EXPORT2
+u_catgets(u_nl_catd catd, int32_t set_num, int32_t msg_num,
+          const UChar* s,
+          int32_t* len, UErrorCode* ec);
+
+U_CDECL_END
+
+#endif /*UCAT_H*/
+/*eof*/
diff --git a/thirdparty/icu4c/common/unicode/uchar.h b/thirdparty/icu4c/common/unicode/uchar.h
new file mode 100644
index 0000000000..1e0f82e706
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uchar.h
@@ -0,0 +1,4056 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File UCHAR.H
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/02/97    aliu        Creation.
+*   03/29/99    helena      Updated for C APIs.
+*   4/15/99     Madhu       Updated for C Implementation and Javadoc
+*   5/20/99     Madhu       Added the function u_getVersion()
+*   8/19/1999   srl         Upgraded scripts to Unicode 3.0
+*   8/27/1999   schererm    UCharDirection constants: U_...
+*   11/11/1999  weiv        added u_isalnum(), cleaned comments
+*   01/11/2000  helena      Renamed u_getVersion to u_getUnicodeVersion().
+******************************************************************************
+*/
+
+#ifndef UCHAR_H
+#define UCHAR_H
+
+#include "unicode/utypes.h"
+#include "unicode/stringoptions.h"
+#include "unicode/ucpmap.h"
+
+#if !defined(USET_DEFINED) && !defined(U_IN_DOXYGEN)
+
+#define USET_DEFINED
+
+/**
+ * USet is the C API type corresponding to C++ class UnicodeSet.
+ * It is forward-declared here to avoid including unicode/uset.h file if related
+ * APIs are not used.
+ *
+ * @see ucnv_getUnicodeSet
+ * @stable ICU 2.4
+ */
+typedef struct USet USet;
+
+#endif
+
+
+U_CDECL_BEGIN
+
+/*==========================================================================*/
+/* Unicode version number                                                   */
+/*==========================================================================*/
+/**
+ * Unicode version number, default for the current ICU version.
+ * The actual Unicode Character Database (UCD) data is stored in uprops.dat
+ * and may be generated from UCD files from a different Unicode version.
+ * Call u_getUnicodeVersion to get the actual Unicode version of the data.
+ *
+ * @see u_getUnicodeVersion
+ * @stable ICU 2.0
+ */
+#define U_UNICODE_VERSION "13.0"
+
+/**
+ * \file
+ * \brief C API: Unicode Properties
+ *
+ * This C API provides low-level access to the Unicode Character Database.
+ * In addition to raw property values, some convenience functions calculate
+ * derived properties, for example for Java-style programming.
+ *
+ * Unicode assigns each code point (not just assigned character) values for
+ * many properties.
+ * Most of them are simple boolean flags, or constants from a small enumerated list.
+ * For some properties, values are strings or other relatively more complex types.
+ *
+ * For more information see
+ * "About the Unicode Character Database" (http://www.unicode.org/ucd/)
+ * and the ICU User Guide chapter on Properties (http://icu-project.org/userguide/properties.html).
+ *
+ * Many properties are accessible via generic functions that take a UProperty selector.
+ * - u_hasBinaryProperty() returns a binary value (true/false) per property and code point.
+ * - u_getIntPropertyValue() returns an integer value per property and code point.
+ *   For each supported enumerated or catalog property, there is
+ *   an enum type for all of the property's values, and
+ *   u_getIntPropertyValue() returns the numeric values of those constants.
+ * - u_getBinaryPropertySet() returns a set for each ICU-supported binary property with
+ *   all code points for which the property is true.
+ * - u_getIntPropertyMap() returns a map for each
+ *   ICU-supported enumerated/catalog/int-valued property which
+ *   maps all Unicode code points to their values for that property.
+ *
+ * Many functions are designed to match java.lang.Character functions.
+ * See the individual function documentation,
+ * and see the JDK 1.4 java.lang.Character documentation
+ * at http://java.sun.com/j2se/1.4/docs/api/java/lang/Character.html
+ *
+ * There are also functions that provide easy migration from C/POSIX functions
+ * like isblank(). Their use is generally discouraged because the C/POSIX
+ * standards do not define their semantics beyond the ASCII range, which means
+ * that different implementations exhibit very different behavior.
+ * Instead, Unicode properties should be used directly.
+ *
+ * There are also only a few, broad C/POSIX character classes, and they tend
+ * to be used for conflicting purposes. For example, the "isalpha()" class
+ * is sometimes used to determine word boundaries, while a more sophisticated
+ * approach would at least distinguish initial letters from continuation
+ * characters (the latter including combining marks).
+ * (In ICU, BreakIterator is the most sophisticated API for word boundaries.)
+ * Another example: There is no "istitle()" class for titlecase characters.
+ *
+ * ICU 3.4 and later provides API access for all twelve C/POSIX character classes.
+ * ICU implements them according to the Standard Recommendations in
+ * Annex C: Compatibility Properties of UTS #18 Unicode Regular Expressions
+ * (http://www.unicode.org/reports/tr18/#Compatibility_Properties).
+ *
+ * API access for C/POSIX character classes is as follows:
+ * - alpha:     u_isUAlphabetic(c) or u_hasBinaryProperty(c, UCHAR_ALPHABETIC)
+ * - lower:     u_isULowercase(c) or u_hasBinaryProperty(c, UCHAR_LOWERCASE)
+ * - upper:     u_isUUppercase(c) or u_hasBinaryProperty(c, UCHAR_UPPERCASE)
+ * - punct:     u_ispunct(c)
+ * - digit:     u_isdigit(c) or u_charType(c)==U_DECIMAL_DIGIT_NUMBER
+ * - xdigit:    u_isxdigit(c) or u_hasBinaryProperty(c, UCHAR_POSIX_XDIGIT)
+ * - alnum:     u_hasBinaryProperty(c, UCHAR_POSIX_ALNUM)
+ * - space:     u_isUWhiteSpace(c) or u_hasBinaryProperty(c, UCHAR_WHITE_SPACE)
+ * - blank:     u_isblank(c) or u_hasBinaryProperty(c, UCHAR_POSIX_BLANK)
+ * - cntrl:     u_charType(c)==U_CONTROL_CHAR
+ * - graph:     u_hasBinaryProperty(c, UCHAR_POSIX_GRAPH)
+ * - print:     u_hasBinaryProperty(c, UCHAR_POSIX_PRINT)
+ *
+ * Note: Some of the u_isxyz() functions in uchar.h predate, and do not match,
+ * the Standard Recommendations in UTS #18. Instead, they match Java
+ * functions according to their API documentation.
+ *
+ * \htmlonly
+ * The C/POSIX character classes are also available in UnicodeSet patterns,
+ * using patterns like [:graph:] or \p{graph}.
+ * \endhtmlonly
+ *
+ * Note: There are several ICU whitespace functions.
+ * Comparison:
+ * - u_isUWhiteSpace=UCHAR_WHITE_SPACE: Unicode White_Space property;
+ *       most of general categories "Z" (separators) + most whitespace ISO controls
+ *       (including no-break spaces, but excluding IS1..IS4)
+ * - u_isWhitespace: Java isWhitespace; Z + whitespace ISO controls but excluding no-break spaces
+ * - u_isJavaSpaceChar: Java isSpaceChar; just Z (including no-break spaces)
+ * - u_isspace: Z + whitespace ISO controls (including no-break spaces)
+ * - u_isblank: "horizontal spaces" = TAB + Zs
+ */
+
+/**
+ * Constants.
+ */
+
+/** The lowest Unicode code point value. Code points are non-negative. @stable ICU 2.0 */
+#define UCHAR_MIN_VALUE 0
+
+/**
+ * The highest Unicode code point value (scalar value) according to
+ * The Unicode Standard. This is a 21-bit value (20.1 bits, rounded up).
+ * For a single character, UChar32 is a simple type that can hold any code point value.
+ *
+ * @see UChar32
+ * @stable ICU 2.0
+ */
+#define UCHAR_MAX_VALUE 0x10ffff
+
+/**
+ * Get a single-bit bit set (a flag) from a bit number 0..31.
+ * @stable ICU 2.1
+ */
+#define U_MASK(x) ((uint32_t)1<<(x))
+
+/**
+ * Selection constants for Unicode properties.
+ * These constants are used in functions like u_hasBinaryProperty to select
+ * one of the Unicode properties.
+ *
+ * The properties APIs are intended to reflect Unicode properties as defined
+ * in the Unicode Character Database (UCD) and Unicode Technical Reports (UTR).
+ *
+ * For details about the properties see
+ * UAX #44: Unicode Character Database (http://www.unicode.org/reports/tr44/).
+ *
+ * Important: If ICU is built with UCD files from Unicode versions below, e.g., 3.2,
+ * then properties marked with "new in Unicode 3.2" are not or not fully available.
+ * Check u_getUnicodeVersion to be sure.
+ *
+ * @see u_hasBinaryProperty
+ * @see u_getIntPropertyValue
+ * @see u_getUnicodeVersion
+ * @stable ICU 2.1
+ */
+typedef enum UProperty {
+    /*
+     * Note: UProperty constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     UCHAR_<Unicode property name>=<integer>,
+     */
+
+    /*  Note: Place UCHAR_ALPHABETIC before UCHAR_BINARY_START so that
+    debuggers display UCHAR_ALPHABETIC as the symbolic name for 0,
+    rather than UCHAR_BINARY_START.  Likewise for other *_START
+    identifiers. */
+
+    /** Binary property Alphabetic. Same as u_isUAlphabetic, different from u_isalpha.
+        Lu+Ll+Lt+Lm+Lo+Nl+Other_Alphabetic @stable ICU 2.1 */
+    UCHAR_ALPHABETIC=0,
+    /** First constant for binary Unicode properties. @stable ICU 2.1 */
+    UCHAR_BINARY_START=UCHAR_ALPHABETIC,
+    /** Binary property ASCII_Hex_Digit. 0-9 A-F a-f @stable ICU 2.1 */
+    UCHAR_ASCII_HEX_DIGIT=1,
+    /** Binary property Bidi_Control.
+        Format controls which have specific functions
+        in the Bidi Algorithm. @stable ICU 2.1 */
+    UCHAR_BIDI_CONTROL=2,
+    /** Binary property Bidi_Mirrored.
+        Characters that may change display in RTL text.
+        Same as u_isMirrored.
+        See Bidi Algorithm, UTR 9. @stable ICU 2.1 */
+    UCHAR_BIDI_MIRRORED=3,
+    /** Binary property Dash. Variations of dashes. @stable ICU 2.1 */
+    UCHAR_DASH=4,
+    /** Binary property Default_Ignorable_Code_Point (new in Unicode 3.2).
+        Ignorable in most processing.
+        <2060..206F, FFF0..FFFB, E0000..E0FFF>+Other_Default_Ignorable_Code_Point+(Cf+Cc+Cs-White_Space) @stable ICU 2.1 */
+    UCHAR_DEFAULT_IGNORABLE_CODE_POINT=5,
+    /** Binary property Deprecated (new in Unicode 3.2).
+        The usage of deprecated characters is strongly discouraged. @stable ICU 2.1 */
+    UCHAR_DEPRECATED=6,
+    /** Binary property Diacritic. Characters that linguistically modify
+        the meaning of another character to which they apply. @stable ICU 2.1 */
+    UCHAR_DIACRITIC=7,
+    /** Binary property Extender.
+        Extend the value or shape of a preceding alphabetic character,
+        e.g., length and iteration marks. @stable ICU 2.1 */
+    UCHAR_EXTENDER=8,
+    /** Binary property Full_Composition_Exclusion.
+        CompositionExclusions.txt+Singleton Decompositions+
+        Non-Starter Decompositions. @stable ICU 2.1 */
+    UCHAR_FULL_COMPOSITION_EXCLUSION=9,
+    /** Binary property Grapheme_Base (new in Unicode 3.2).
+        For programmatic determination of grapheme cluster boundaries.
+        [0..10FFFF]-Cc-Cf-Cs-Co-Cn-Zl-Zp-Grapheme_Link-Grapheme_Extend-CGJ @stable ICU 2.1 */
+    UCHAR_GRAPHEME_BASE=10,
+    /** Binary property Grapheme_Extend (new in Unicode 3.2).
+        For programmatic determination of grapheme cluster boundaries.
+        Me+Mn+Mc+Other_Grapheme_Extend-Grapheme_Link-CGJ @stable ICU 2.1 */
+    UCHAR_GRAPHEME_EXTEND=11,
+    /** Binary property Grapheme_Link (new in Unicode 3.2).
+        For programmatic determination of grapheme cluster boundaries. @stable ICU 2.1 */
+    UCHAR_GRAPHEME_LINK=12,
+    /** Binary property Hex_Digit.
+        Characters commonly used for hexadecimal numbers. @stable ICU 2.1 */
+    UCHAR_HEX_DIGIT=13,
+    /** Binary property Hyphen. Dashes used to mark connections
+        between pieces of words, plus the Katakana middle dot. @stable ICU 2.1 */
+    UCHAR_HYPHEN=14,
+    /** Binary property ID_Continue.
+        Characters that can continue an identifier.
+        DerivedCoreProperties.txt also says "NOTE: Cf characters should be filtered out."
+        ID_Start+Mn+Mc+Nd+Pc @stable ICU 2.1 */
+    UCHAR_ID_CONTINUE=15,
+    /** Binary property ID_Start.
+        Characters that can start an identifier.
+        Lu+Ll+Lt+Lm+Lo+Nl @stable ICU 2.1 */
+    UCHAR_ID_START=16,
+    /** Binary property Ideographic.
+        CJKV ideographs. @stable ICU 2.1 */
+    UCHAR_IDEOGRAPHIC=17,
+    /** Binary property IDS_Binary_Operator (new in Unicode 3.2).
+        For programmatic determination of
+        Ideographic Description Sequences. @stable ICU 2.1 */
+    UCHAR_IDS_BINARY_OPERATOR=18,
+    /** Binary property IDS_Trinary_Operator (new in Unicode 3.2).
+        For programmatic determination of
+        Ideographic Description Sequences. @stable ICU 2.1 */
+    UCHAR_IDS_TRINARY_OPERATOR=19,
+    /** Binary property Join_Control.
+        Format controls for cursive joining and ligation. @stable ICU 2.1 */
+    UCHAR_JOIN_CONTROL=20,
+    /** Binary property Logical_Order_Exception (new in Unicode 3.2).
+        Characters that do not use logical order and
+        require special handling in most processing. @stable ICU 2.1 */
+    UCHAR_LOGICAL_ORDER_EXCEPTION=21,
+    /** Binary property Lowercase. Same as u_isULowercase, different from u_islower.
+        Ll+Other_Lowercase @stable ICU 2.1 */
+    UCHAR_LOWERCASE=22,
+    /** Binary property Math. Sm+Other_Math @stable ICU 2.1 */
+    UCHAR_MATH=23,
+    /** Binary property Noncharacter_Code_Point.
+        Code points that are explicitly defined as illegal
+        for the encoding of characters. @stable ICU 2.1 */
+    UCHAR_NONCHARACTER_CODE_POINT=24,
+    /** Binary property Quotation_Mark. @stable ICU 2.1 */
+    UCHAR_QUOTATION_MARK=25,
+    /** Binary property Radical (new in Unicode 3.2).
+        For programmatic determination of
+        Ideographic Description Sequences. @stable ICU 2.1 */
+    UCHAR_RADICAL=26,
+    /** Binary property Soft_Dotted (new in Unicode 3.2).
+        Characters with a "soft dot", like i or j.
+        An accent placed on these characters causes
+        the dot to disappear. @stable ICU 2.1 */
+    UCHAR_SOFT_DOTTED=27,
+    /** Binary property Terminal_Punctuation.
+        Punctuation characters that generally mark
+        the end of textual units. @stable ICU 2.1 */
+    UCHAR_TERMINAL_PUNCTUATION=28,
+    /** Binary property Unified_Ideograph (new in Unicode 3.2).
+        For programmatic determination of
+        Ideographic Description Sequences. @stable ICU 2.1 */
+    UCHAR_UNIFIED_IDEOGRAPH=29,
+    /** Binary property Uppercase. Same as u_isUUppercase, different from u_isupper.
+        Lu+Other_Uppercase @stable ICU 2.1 */
+    UCHAR_UPPERCASE=30,
+    /** Binary property White_Space.
+        Same as u_isUWhiteSpace, different from u_isspace and u_isWhitespace.
+        Space characters+TAB+CR+LF-ZWSP-ZWNBSP @stable ICU 2.1 */
+    UCHAR_WHITE_SPACE=31,
+    /** Binary property XID_Continue.
+        ID_Continue modified to allow closure under
+        normalization forms NFKC and NFKD. @stable ICU 2.1 */
+    UCHAR_XID_CONTINUE=32,
+    /** Binary property XID_Start. ID_Start modified to allow
+        closure under normalization forms NFKC and NFKD. @stable ICU 2.1 */
+    UCHAR_XID_START=33,
+    /** Binary property Case_Sensitive. Either the source of a case
+        mapping or _in_ the target of a case mapping. Not the same as
+        the general category Cased_Letter. @stable ICU 2.6 */
+   UCHAR_CASE_SENSITIVE=34,
+    /** Binary property STerm (new in Unicode 4.0.1).
+        Sentence Terminal. Used in UAX #29: Text Boundaries
+        (http://www.unicode.org/reports/tr29/)
+        @stable ICU 3.0 */
+    UCHAR_S_TERM=35,
+    /** Binary property Variation_Selector (new in Unicode 4.0.1).
+        Indicates all those characters that qualify as Variation Selectors.
+        For details on the behavior of these characters,
+        see StandardizedVariants.html and 15.6 Variation Selectors.
+        @stable ICU 3.0 */
+    UCHAR_VARIATION_SELECTOR=36,
+    /** Binary property NFD_Inert.
+        ICU-specific property for characters that are inert under NFD,
+        i.e., they do not interact with adjacent characters.
+        See the documentation for the Normalizer2 class and the
+        Normalizer2::isInert() method.
+        @stable ICU 3.0 */
+    UCHAR_NFD_INERT=37,
+    /** Binary property NFKD_Inert.
+        ICU-specific property for characters that are inert under NFKD,
+        i.e., they do not interact with adjacent characters.
+        See the documentation for the Normalizer2 class and the
+        Normalizer2::isInert() method.
+        @stable ICU 3.0 */
+    UCHAR_NFKD_INERT=38,
+    /** Binary property NFC_Inert.
+        ICU-specific property for characters that are inert under NFC,
+        i.e., they do not interact with adjacent characters.
+        See the documentation for the Normalizer2 class and the
+        Normalizer2::isInert() method.
+        @stable ICU 3.0 */
+    UCHAR_NFC_INERT=39,
+    /** Binary property NFKC_Inert.
+        ICU-specific property for characters that are inert under NFKC,
+        i.e., they do not interact with adjacent characters.
+        See the documentation for the Normalizer2 class and the
+        Normalizer2::isInert() method.
+        @stable ICU 3.0 */
+    UCHAR_NFKC_INERT=40,
+    /** Binary Property Segment_Starter.
+        ICU-specific property for characters that are starters in terms of
+        Unicode normalization and combining character sequences.
+        They have ccc=0 and do not occur in non-initial position of the
+        canonical decomposition of any character
+        (like a-umlaut in NFD and a Jamo T in an NFD(Hangul LVT)).
+        ICU uses this property for segmenting a string for generating a set of
+        canonically equivalent strings, e.g. for canonical closure while
+        processing collation tailoring rules.
+        @stable ICU 3.0 */
+    UCHAR_SEGMENT_STARTER=41,
+    /** Binary property Pattern_Syntax (new in Unicode 4.1).
+        See UAX #31 Identifier and Pattern Syntax
+        (http://www.unicode.org/reports/tr31/)
+        @stable ICU 3.4 */
+    UCHAR_PATTERN_SYNTAX=42,
+    /** Binary property Pattern_White_Space (new in Unicode 4.1).
+        See UAX #31 Identifier and Pattern Syntax
+        (http://www.unicode.org/reports/tr31/)
+        @stable ICU 3.4 */
+    UCHAR_PATTERN_WHITE_SPACE=43,
+    /** Binary property alnum (a C/POSIX character class).
+        Implemented according to the UTS #18 Annex C Standard Recommendation.
+        See the uchar.h file documentation.
+        @stable ICU 3.4 */
+    UCHAR_POSIX_ALNUM=44,
+    /** Binary property blank (a C/POSIX character class).
+        Implemented according to the UTS #18 Annex C Standard Recommendation.
+        See the uchar.h file documentation.
+        @stable ICU 3.4 */
+    UCHAR_POSIX_BLANK=45,
+    /** Binary property graph (a C/POSIX character class).
+        Implemented according to the UTS #18 Annex C Standard Recommendation.
+        See the uchar.h file documentation.
+        @stable ICU 3.4 */
+    UCHAR_POSIX_GRAPH=46,
+    /** Binary property print (a C/POSIX character class).
+        Implemented according to the UTS #18 Annex C Standard Recommendation.
+        See the uchar.h file documentation.
+        @stable ICU 3.4 */
+    UCHAR_POSIX_PRINT=47,
+    /** Binary property xdigit (a C/POSIX character class).
+        Implemented according to the UTS #18 Annex C Standard Recommendation.
+        See the uchar.h file documentation.
+        @stable ICU 3.4 */
+    UCHAR_POSIX_XDIGIT=48,
+    /** Binary property Cased. For Lowercase, Uppercase and Titlecase characters. @stable ICU 4.4 */
+    UCHAR_CASED=49,
+    /** Binary property Case_Ignorable. Used in context-sensitive case mappings. @stable ICU 4.4 */
+    UCHAR_CASE_IGNORABLE=50,
+    /** Binary property Changes_When_Lowercased. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_LOWERCASED=51,
+    /** Binary property Changes_When_Uppercased. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_UPPERCASED=52,
+    /** Binary property Changes_When_Titlecased. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_TITLECASED=53,
+    /** Binary property Changes_When_Casefolded. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_CASEFOLDED=54,
+    /** Binary property Changes_When_Casemapped. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_CASEMAPPED=55,
+    /** Binary property Changes_When_NFKC_Casefolded. @stable ICU 4.4 */
+    UCHAR_CHANGES_WHEN_NFKC_CASEFOLDED=56,
+    /**
+     * Binary property Emoji.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 57
+     */
+    UCHAR_EMOJI=57,
+    /**
+     * Binary property Emoji_Presentation.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 57
+     */
+    UCHAR_EMOJI_PRESENTATION=58,
+    /**
+     * Binary property Emoji_Modifier.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 57
+     */
+    UCHAR_EMOJI_MODIFIER=59,
+    /**
+     * Binary property Emoji_Modifier_Base.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 57
+     */
+    UCHAR_EMOJI_MODIFIER_BASE=60,
+    /**
+     * Binary property Emoji_Component.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 60
+     */
+    UCHAR_EMOJI_COMPONENT=61,
+    /**
+     * Binary property Regional_Indicator.
+     * @stable ICU 60
+     */
+    UCHAR_REGIONAL_INDICATOR=62,
+    /**
+     * Binary property Prepended_Concatenation_Mark.
+     * @stable ICU 60
+     */
+    UCHAR_PREPENDED_CONCATENATION_MARK=63,
+    /**
+     * Binary property Extended_Pictographic.
+     * See http://www.unicode.org/reports/tr51/#Emoji_Properties
+     *
+     * @stable ICU 62
+     */
+    UCHAR_EXTENDED_PICTOGRAPHIC=64,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for binary Unicode properties.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_BINARY_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** Enumerated property Bidi_Class.
+        Same as u_charDirection, returns UCharDirection values. @stable ICU 2.2 */
+    UCHAR_BIDI_CLASS=0x1000,
+    /** First constant for enumerated/integer Unicode properties. @stable ICU 2.2 */
+    UCHAR_INT_START=UCHAR_BIDI_CLASS,
+    /** Enumerated property Block.
+        Same as ublock_getCode, returns UBlockCode values. @stable ICU 2.2 */
+    UCHAR_BLOCK=0x1001,
+    /** Enumerated property Canonical_Combining_Class.
+        Same as u_getCombiningClass, returns 8-bit numeric values. @stable ICU 2.2 */
+    UCHAR_CANONICAL_COMBINING_CLASS=0x1002,
+    /** Enumerated property Decomposition_Type.
+        Returns UDecompositionType values. @stable ICU 2.2 */
+    UCHAR_DECOMPOSITION_TYPE=0x1003,
+    /** Enumerated property East_Asian_Width.
+        See http://www.unicode.org/reports/tr11/
+        Returns UEastAsianWidth values. @stable ICU 2.2 */
+    UCHAR_EAST_ASIAN_WIDTH=0x1004,
+    /** Enumerated property General_Category.
+        Same as u_charType, returns UCharCategory values. @stable ICU 2.2 */
+    UCHAR_GENERAL_CATEGORY=0x1005,
+    /** Enumerated property Joining_Group.
+        Returns UJoiningGroup values. @stable ICU 2.2 */
+    UCHAR_JOINING_GROUP=0x1006,
+    /** Enumerated property Joining_Type.
+        Returns UJoiningType values. @stable ICU 2.2 */
+    UCHAR_JOINING_TYPE=0x1007,
+    /** Enumerated property Line_Break.
+        Returns ULineBreak values. @stable ICU 2.2 */
+    UCHAR_LINE_BREAK=0x1008,
+    /** Enumerated property Numeric_Type.
+        Returns UNumericType values. @stable ICU 2.2 */
+    UCHAR_NUMERIC_TYPE=0x1009,
+    /** Enumerated property Script.
+        Same as uscript_getScript, returns UScriptCode values. @stable ICU 2.2 */
+    UCHAR_SCRIPT=0x100A,
+    /** Enumerated property Hangul_Syllable_Type, new in Unicode 4.
+        Returns UHangulSyllableType values. @stable ICU 2.6 */
+    UCHAR_HANGUL_SYLLABLE_TYPE=0x100B,
+    /** Enumerated property NFD_Quick_Check.
+        Returns UNormalizationCheckResult values. @stable ICU 3.0 */
+    UCHAR_NFD_QUICK_CHECK=0x100C,
+    /** Enumerated property NFKD_Quick_Check.
+        Returns UNormalizationCheckResult values. @stable ICU 3.0 */
+    UCHAR_NFKD_QUICK_CHECK=0x100D,
+    /** Enumerated property NFC_Quick_Check.
+        Returns UNormalizationCheckResult values. @stable ICU 3.0 */
+    UCHAR_NFC_QUICK_CHECK=0x100E,
+    /** Enumerated property NFKC_Quick_Check.
+        Returns UNormalizationCheckResult values. @stable ICU 3.0 */
+    UCHAR_NFKC_QUICK_CHECK=0x100F,
+    /** Enumerated property Lead_Canonical_Combining_Class.
+        ICU-specific property for the ccc of the first code point
+        of the decomposition, or lccc(c)=ccc(NFD(c)[0]).
+        Useful for checking for canonically ordered text;
+        see UNORM_FCD and http://www.unicode.org/notes/tn5/#FCD .
+        Returns 8-bit numeric values like UCHAR_CANONICAL_COMBINING_CLASS. @stable ICU 3.0 */
+    UCHAR_LEAD_CANONICAL_COMBINING_CLASS=0x1010,
+    /** Enumerated property Trail_Canonical_Combining_Class.
+        ICU-specific property for the ccc of the last code point
+        of the decomposition, or tccc(c)=ccc(NFD(c)[last]).
+        Useful for checking for canonically ordered text;
+        see UNORM_FCD and http://www.unicode.org/notes/tn5/#FCD .
+        Returns 8-bit numeric values like UCHAR_CANONICAL_COMBINING_CLASS. @stable ICU 3.0 */
+    UCHAR_TRAIL_CANONICAL_COMBINING_CLASS=0x1011,
+    /** Enumerated property Grapheme_Cluster_Break (new in Unicode 4.1).
+        Used in UAX #29: Text Boundaries
+        (http://www.unicode.org/reports/tr29/)
+        Returns UGraphemeClusterBreak values. @stable ICU 3.4 */
+    UCHAR_GRAPHEME_CLUSTER_BREAK=0x1012,
+    /** Enumerated property Sentence_Break (new in Unicode 4.1).
+        Used in UAX #29: Text Boundaries
+        (http://www.unicode.org/reports/tr29/)
+        Returns USentenceBreak values. @stable ICU 3.4 */
+    UCHAR_SENTENCE_BREAK=0x1013,
+    /** Enumerated property Word_Break (new in Unicode 4.1).
+        Used in UAX #29: Text Boundaries
+        (http://www.unicode.org/reports/tr29/)
+        Returns UWordBreakValues values. @stable ICU 3.4 */
+    UCHAR_WORD_BREAK=0x1014,
+    /** Enumerated property Bidi_Paired_Bracket_Type (new in Unicode 6.3).
+        Used in UAX #9: Unicode Bidirectional Algorithm
+        (http://www.unicode.org/reports/tr9/)
+        Returns UBidiPairedBracketType values. @stable ICU 52 */
+    UCHAR_BIDI_PAIRED_BRACKET_TYPE=0x1015,
+    /**
+     * Enumerated property Indic_Positional_Category.
+     * New in Unicode 6.0 as provisional property Indic_Matra_Category;
+     * renamed and changed to informative in Unicode 8.0.
+     * See http://www.unicode.org/reports/tr44/#IndicPositionalCategory.txt
+     * @stable ICU 63
+     */
+    UCHAR_INDIC_POSITIONAL_CATEGORY=0x1016,
+    /**
+     * Enumerated property Indic_Syllabic_Category.
+     * New in Unicode 6.0 as provisional; informative since Unicode 8.0.
+     * See http://www.unicode.org/reports/tr44/#IndicSyllabicCategory.txt
+     * @stable ICU 63
+     */
+    UCHAR_INDIC_SYLLABIC_CATEGORY=0x1017,
+    /**
+     * Enumerated property Vertical_Orientation.
+     * Used for UAX #50 Unicode Vertical Text Layout (https://www.unicode.org/reports/tr50/).
+     * New as a UCD property in Unicode 10.0.
+     * @stable ICU 63
+     */
+    UCHAR_VERTICAL_ORIENTATION=0x1018,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for enumerated/integer Unicode properties.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_INT_LIMIT=0x1019,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** Bitmask property General_Category_Mask.
+        This is the General_Category property returned as a bit mask.
+        When used in u_getIntPropertyValue(c), same as U_MASK(u_charType(c)),
+        returns bit masks for UCharCategory values where exactly one bit is set.
+        When used with u_getPropertyValueName() and u_getPropertyValueEnum(),
+        a multi-bit mask is used for sets of categories like "Letters".
+        Mask values should be cast to uint32_t.
+        @stable ICU 2.4 */
+    UCHAR_GENERAL_CATEGORY_MASK=0x2000,
+    /** First constant for bit-mask Unicode properties. @stable ICU 2.4 */
+    UCHAR_MASK_START=UCHAR_GENERAL_CATEGORY_MASK,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for bit-mask Unicode properties.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_MASK_LIMIT=0x2001,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** Double property Numeric_Value.
+        Corresponds to u_getNumericValue. @stable ICU 2.4 */
+    UCHAR_NUMERIC_VALUE=0x3000,
+    /** First constant for double Unicode properties. @stable ICU 2.4 */
+    UCHAR_DOUBLE_START=UCHAR_NUMERIC_VALUE,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for double Unicode properties.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_DOUBLE_LIMIT=0x3001,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** String property Age.
+        Corresponds to u_charAge. @stable ICU 2.4 */
+    UCHAR_AGE=0x4000,
+    /** First constant for string Unicode properties. @stable ICU 2.4 */
+    UCHAR_STRING_START=UCHAR_AGE,
+    /** String property Bidi_Mirroring_Glyph.
+        Corresponds to u_charMirror. @stable ICU 2.4 */
+    UCHAR_BIDI_MIRRORING_GLYPH=0x4001,
+    /** String property Case_Folding.
+        Corresponds to u_strFoldCase in ustring.h. @stable ICU 2.4 */
+    UCHAR_CASE_FOLDING=0x4002,
+#ifndef U_HIDE_DEPRECATED_API
+    /** Deprecated string property ISO_Comment.
+        Corresponds to u_getISOComment. @deprecated ICU 49 */
+    UCHAR_ISO_COMMENT=0x4003,
+#endif  /* U_HIDE_DEPRECATED_API */
+    /** String property Lowercase_Mapping.
+        Corresponds to u_strToLower in ustring.h. @stable ICU 2.4 */
+    UCHAR_LOWERCASE_MAPPING=0x4004,
+    /** String property Name.
+        Corresponds to u_charName. @stable ICU 2.4 */
+    UCHAR_NAME=0x4005,
+    /** String property Simple_Case_Folding.
+        Corresponds to u_foldCase. @stable ICU 2.4 */
+    UCHAR_SIMPLE_CASE_FOLDING=0x4006,
+    /** String property Simple_Lowercase_Mapping.
+        Corresponds to u_tolower. @stable ICU 2.4 */
+    UCHAR_SIMPLE_LOWERCASE_MAPPING=0x4007,
+    /** String property Simple_Titlecase_Mapping.
+        Corresponds to u_totitle. @stable ICU 2.4 */
+    UCHAR_SIMPLE_TITLECASE_MAPPING=0x4008,
+    /** String property Simple_Uppercase_Mapping.
+        Corresponds to u_toupper. @stable ICU 2.4 */
+    UCHAR_SIMPLE_UPPERCASE_MAPPING=0x4009,
+    /** String property Titlecase_Mapping.
+        Corresponds to u_strToTitle in ustring.h. @stable ICU 2.4 */
+    UCHAR_TITLECASE_MAPPING=0x400A,
+#ifndef U_HIDE_DEPRECATED_API
+    /** String property Unicode_1_Name.
+        This property is of little practical value.
+        Beginning with ICU 49, ICU APIs return an empty string for this property.
+        Corresponds to u_charName(U_UNICODE_10_CHAR_NAME). @deprecated ICU 49 */
+    UCHAR_UNICODE_1_NAME=0x400B,
+#endif  /* U_HIDE_DEPRECATED_API */
+    /** String property Uppercase_Mapping.
+        Corresponds to u_strToUpper in ustring.h. @stable ICU 2.4 */
+    UCHAR_UPPERCASE_MAPPING=0x400C,
+    /** String property Bidi_Paired_Bracket (new in Unicode 6.3).
+        Corresponds to u_getBidiPairedBracket. @stable ICU 52 */
+    UCHAR_BIDI_PAIRED_BRACKET=0x400D,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for string Unicode properties.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_STRING_LIMIT=0x400E,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** Miscellaneous property Script_Extensions (new in Unicode 6.0).
+        Some characters are commonly used in multiple scripts.
+        For more information, see UAX #24: http://www.unicode.org/reports/tr24/.
+        Corresponds to uscript_hasScript and uscript_getScriptExtensions in uscript.h.
+        @stable ICU 4.6 */
+    UCHAR_SCRIPT_EXTENSIONS=0x7000,
+    /** First constant for Unicode properties with unusual value types. @stable ICU 4.6 */
+    UCHAR_OTHER_PROPERTY_START=UCHAR_SCRIPT_EXTENSIONS,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last constant for Unicode properties with unusual value types.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCHAR_OTHER_PROPERTY_LIMIT=0x7001,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** Represents a nonexistent or invalid property or property value. @stable ICU 2.4 */
+    UCHAR_INVALID_CODE = -1
+} UProperty;
+
+/**
+ * Data for enumerated Unicode general category types.
+ * See http://www.unicode.org/Public/UNIDATA/UnicodeData.html .
+ * @stable ICU 2.0
+ */
+typedef enum UCharCategory
+{
+    /*
+     * Note: UCharCategory constants and their API comments are parsed by preparseucd.py.
+     * It matches pairs of lines like
+     *     / ** <Unicode 2-letter General_Category value> comment... * /
+     *     U_<[A-Z_]+> = <integer>,
+     */
+
+    /** Non-category for unassigned and non-character code points. @stable ICU 2.0 */
+    U_UNASSIGNED              = 0,
+    /** Cn "Other, Not Assigned (no characters in [UnicodeData.txt] have this property)" (same as U_UNASSIGNED!) @stable ICU 2.0 */
+    U_GENERAL_OTHER_TYPES     = 0,
+    /** Lu @stable ICU 2.0 */
+    U_UPPERCASE_LETTER        = 1,
+    /** Ll @stable ICU 2.0 */
+    U_LOWERCASE_LETTER        = 2,
+    /** Lt @stable ICU 2.0 */
+    U_TITLECASE_LETTER        = 3,
+    /** Lm @stable ICU 2.0 */
+    U_MODIFIER_LETTER         = 4,
+    /** Lo @stable ICU 2.0 */
+    U_OTHER_LETTER            = 5,
+    /** Mn @stable ICU 2.0 */
+    U_NON_SPACING_MARK        = 6,
+    /** Me @stable ICU 2.0 */
+    U_ENCLOSING_MARK          = 7,
+    /** Mc @stable ICU 2.0 */
+    U_COMBINING_SPACING_MARK  = 8,
+    /** Nd @stable ICU 2.0 */
+    U_DECIMAL_DIGIT_NUMBER    = 9,
+    /** Nl @stable ICU 2.0 */
+    U_LETTER_NUMBER           = 10,
+    /** No @stable ICU 2.0 */
+    U_OTHER_NUMBER            = 11,
+    /** Zs @stable ICU 2.0 */
+    U_SPACE_SEPARATOR         = 12,
+    /** Zl @stable ICU 2.0 */
+    U_LINE_SEPARATOR          = 13,
+    /** Zp @stable ICU 2.0 */
+    U_PARAGRAPH_SEPARATOR     = 14,
+    /** Cc @stable ICU 2.0 */
+    U_CONTROL_CHAR            = 15,
+    /** Cf @stable ICU 2.0 */
+    U_FORMAT_CHAR             = 16,
+    /** Co @stable ICU 2.0 */
+    U_PRIVATE_USE_CHAR        = 17,
+    /** Cs @stable ICU 2.0 */
+    U_SURROGATE               = 18,
+    /** Pd @stable ICU 2.0 */
+    U_DASH_PUNCTUATION        = 19,
+    /** Ps @stable ICU 2.0 */
+    U_START_PUNCTUATION       = 20,
+    /** Pe @stable ICU 2.0 */
+    U_END_PUNCTUATION         = 21,
+    /** Pc @stable ICU 2.0 */
+    U_CONNECTOR_PUNCTUATION   = 22,
+    /** Po @stable ICU 2.0 */
+    U_OTHER_PUNCTUATION       = 23,
+    /** Sm @stable ICU 2.0 */
+    U_MATH_SYMBOL             = 24,
+    /** Sc @stable ICU 2.0 */
+    U_CURRENCY_SYMBOL         = 25,
+    /** Sk @stable ICU 2.0 */
+    U_MODIFIER_SYMBOL         = 26,
+    /** So @stable ICU 2.0 */
+    U_OTHER_SYMBOL            = 27,
+    /** Pi @stable ICU 2.0 */
+    U_INITIAL_PUNCTUATION     = 28,
+    /** Pf @stable ICU 2.0 */
+    U_FINAL_PUNCTUATION       = 29,
+    /**
+     * One higher than the last enum UCharCategory constant.
+     * This numeric value is stable (will not change), see
+     * http://www.unicode.org/policies/stability_policy.html#Property_Value
+     *
+     * @stable ICU 2.0
+     */
+    U_CHAR_CATEGORY_COUNT
+} UCharCategory;
+
+/**
+ * U_GC_XX_MASK constants are bit flags corresponding to Unicode
+ * general category values.
+ * For each category, the nth bit is set if the numeric value of the
+ * corresponding UCharCategory constant is n.
+ *
+ * There are also some U_GC_Y_MASK constants for groups of general categories
+ * like L for all letter categories.
+ *
+ * @see u_charType
+ * @see U_GET_GC_MASK
+ * @see UCharCategory
+ * @stable ICU 2.1
+ */
+#define U_GC_CN_MASK    U_MASK(U_GENERAL_OTHER_TYPES)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_LU_MASK    U_MASK(U_UPPERCASE_LETTER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_LL_MASK    U_MASK(U_LOWERCASE_LETTER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_LT_MASK    U_MASK(U_TITLECASE_LETTER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_LM_MASK    U_MASK(U_MODIFIER_LETTER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_LO_MASK    U_MASK(U_OTHER_LETTER)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_MN_MASK    U_MASK(U_NON_SPACING_MARK)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_ME_MASK    U_MASK(U_ENCLOSING_MARK)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_MC_MASK    U_MASK(U_COMBINING_SPACING_MARK)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_ND_MASK    U_MASK(U_DECIMAL_DIGIT_NUMBER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_NL_MASK    U_MASK(U_LETTER_NUMBER)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_NO_MASK    U_MASK(U_OTHER_NUMBER)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_ZS_MASK    U_MASK(U_SPACE_SEPARATOR)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_ZL_MASK    U_MASK(U_LINE_SEPARATOR)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_ZP_MASK    U_MASK(U_PARAGRAPH_SEPARATOR)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_CC_MASK    U_MASK(U_CONTROL_CHAR)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_CF_MASK    U_MASK(U_FORMAT_CHAR)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_CO_MASK    U_MASK(U_PRIVATE_USE_CHAR)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_CS_MASK    U_MASK(U_SURROGATE)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PD_MASK    U_MASK(U_DASH_PUNCTUATION)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PS_MASK    U_MASK(U_START_PUNCTUATION)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PE_MASK    U_MASK(U_END_PUNCTUATION)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PC_MASK    U_MASK(U_CONNECTOR_PUNCTUATION)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PO_MASK    U_MASK(U_OTHER_PUNCTUATION)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_SM_MASK    U_MASK(U_MATH_SYMBOL)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_SC_MASK    U_MASK(U_CURRENCY_SYMBOL)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_SK_MASK    U_MASK(U_MODIFIER_SYMBOL)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_SO_MASK    U_MASK(U_OTHER_SYMBOL)
+
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PI_MASK    U_MASK(U_INITIAL_PUNCTUATION)
+/** Mask constant for a UCharCategory. @stable ICU 2.1 */
+#define U_GC_PF_MASK    U_MASK(U_FINAL_PUNCTUATION)
+
+
+/** Mask constant for multiple UCharCategory bits (L Letters). @stable ICU 2.1 */
+#define U_GC_L_MASK \
+            (U_GC_LU_MASK|U_GC_LL_MASK|U_GC_LT_MASK|U_GC_LM_MASK|U_GC_LO_MASK)
+
+/** Mask constant for multiple UCharCategory bits (LC Cased Letters). @stable ICU 2.1 */
+#define U_GC_LC_MASK \
+            (U_GC_LU_MASK|U_GC_LL_MASK|U_GC_LT_MASK)
+
+/** Mask constant for multiple UCharCategory bits (M Marks). @stable ICU 2.1 */
+#define U_GC_M_MASK (U_GC_MN_MASK|U_GC_ME_MASK|U_GC_MC_MASK)
+
+/** Mask constant for multiple UCharCategory bits (N Numbers). @stable ICU 2.1 */
+#define U_GC_N_MASK (U_GC_ND_MASK|U_GC_NL_MASK|U_GC_NO_MASK)
+
+/** Mask constant for multiple UCharCategory bits (Z Separators). @stable ICU 2.1 */
+#define U_GC_Z_MASK (U_GC_ZS_MASK|U_GC_ZL_MASK|U_GC_ZP_MASK)
+
+/** Mask constant for multiple UCharCategory bits (C Others). @stable ICU 2.1 */
+#define U_GC_C_MASK \
+            (U_GC_CN_MASK|U_GC_CC_MASK|U_GC_CF_MASK|U_GC_CO_MASK|U_GC_CS_MASK)
+
+/** Mask constant for multiple UCharCategory bits (P Punctuation). @stable ICU 2.1 */
+#define U_GC_P_MASK \
+            (U_GC_PD_MASK|U_GC_PS_MASK|U_GC_PE_MASK|U_GC_PC_MASK|U_GC_PO_MASK| \
+             U_GC_PI_MASK|U_GC_PF_MASK)
+
+/** Mask constant for multiple UCharCategory bits (S Symbols). @stable ICU 2.1 */
+#define U_GC_S_MASK (U_GC_SM_MASK|U_GC_SC_MASK|U_GC_SK_MASK|U_GC_SO_MASK)
+
+/**
+ * This specifies the language directional property of a character set.
+ * @stable ICU 2.0
+ */
+typedef enum UCharDirection {
+    /*
+     * Note: UCharDirection constants and their API comments are parsed by preparseucd.py.
+     * It matches pairs of lines like
+     *     / ** <Unicode 1..3-letter Bidi_Class value> comment... * /
+     *     U_<[A-Z_]+> = <integer>,
+     */
+
+    /** L @stable ICU 2.0 */
+    U_LEFT_TO_RIGHT               = 0,
+    /** R @stable ICU 2.0 */
+    U_RIGHT_TO_LEFT               = 1,
+    /** EN @stable ICU 2.0 */
+    U_EUROPEAN_NUMBER             = 2,
+    /** ES @stable ICU 2.0 */
+    U_EUROPEAN_NUMBER_SEPARATOR   = 3,
+    /** ET @stable ICU 2.0 */
+    U_EUROPEAN_NUMBER_TERMINATOR  = 4,
+    /** AN @stable ICU 2.0 */
+    U_ARABIC_NUMBER               = 5,
+    /** CS @stable ICU 2.0 */
+    U_COMMON_NUMBER_SEPARATOR     = 6,
+    /** B @stable ICU 2.0 */
+    U_BLOCK_SEPARATOR             = 7,
+    /** S @stable ICU 2.0 */
+    U_SEGMENT_SEPARATOR           = 8,
+    /** WS @stable ICU 2.0 */
+    U_WHITE_SPACE_NEUTRAL         = 9,
+    /** ON @stable ICU 2.0 */
+    U_OTHER_NEUTRAL               = 10,
+    /** LRE @stable ICU 2.0 */
+    U_LEFT_TO_RIGHT_EMBEDDING     = 11,
+    /** LRO @stable ICU 2.0 */
+    U_LEFT_TO_RIGHT_OVERRIDE      = 12,
+    /** AL @stable ICU 2.0 */
+    U_RIGHT_TO_LEFT_ARABIC        = 13,
+    /** RLE @stable ICU 2.0 */
+    U_RIGHT_TO_LEFT_EMBEDDING     = 14,
+    /** RLO @stable ICU 2.0 */
+    U_RIGHT_TO_LEFT_OVERRIDE      = 15,
+    /** PDF @stable ICU 2.0 */
+    U_POP_DIRECTIONAL_FORMAT      = 16,
+    /** NSM @stable ICU 2.0 */
+    U_DIR_NON_SPACING_MARK        = 17,
+    /** BN @stable ICU 2.0 */
+    U_BOUNDARY_NEUTRAL            = 18,
+    /** FSI @stable ICU 52 */
+    U_FIRST_STRONG_ISOLATE        = 19,
+    /** LRI @stable ICU 52 */
+    U_LEFT_TO_RIGHT_ISOLATE       = 20,
+    /** RLI @stable ICU 52 */
+    U_RIGHT_TO_LEFT_ISOLATE       = 21,
+    /** PDI @stable ICU 52 */
+    U_POP_DIRECTIONAL_ISOLATE     = 22,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest UCharDirection value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_BIDI_CLASS).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_CHAR_DIRECTION_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UCharDirection;
+
+/**
+ * Bidi Paired Bracket Type constants.
+ *
+ * @see UCHAR_BIDI_PAIRED_BRACKET_TYPE
+ * @stable ICU 52
+ */
+typedef enum UBidiPairedBracketType {
+    /*
+     * Note: UBidiPairedBracketType constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_BPT_<Unicode Bidi_Paired_Bracket_Type value name>
+     */
+
+    /** Not a paired bracket. @stable ICU 52 */
+    U_BPT_NONE,
+    /** Open paired bracket. @stable ICU 52 */
+    U_BPT_OPEN,
+    /** Close paired bracket. @stable ICU 52 */
+    U_BPT_CLOSE,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UBidiPairedBracketType value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_BIDI_PAIRED_BRACKET_TYPE).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_BPT_COUNT /* 3 */
+#endif  // U_HIDE_DEPRECATED_API
+} UBidiPairedBracketType;
+
+/**
+ * Constants for Unicode blocks, see the Unicode Data file Blocks.txt
+ * @stable ICU 2.0
+ */
+enum UBlockCode {
+    /*
+     * Note: UBlockCode constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     UBLOCK_<Unicode Block value name> = <integer>,
+     */
+
+    /** New No_Block value in Unicode 4. @stable ICU 2.6 */
+    UBLOCK_NO_BLOCK = 0, /*[none]*/ /* Special range indicating No_Block */
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BASIC_LATIN = 1, /*[0000]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LATIN_1_SUPPLEMENT=2, /*[0080]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LATIN_EXTENDED_A =3, /*[0100]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LATIN_EXTENDED_B =4, /*[0180]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_IPA_EXTENSIONS =5, /*[0250]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SPACING_MODIFIER_LETTERS =6, /*[02B0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_COMBINING_DIACRITICAL_MARKS =7, /*[0300]*/
+
+    /**
+     * Unicode 3.2 renames this block to "Greek and Coptic".
+     * @stable ICU 2.0
+     */
+    UBLOCK_GREEK =8, /*[0370]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CYRILLIC =9, /*[0400]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ARMENIAN =10, /*[0530]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HEBREW =11, /*[0590]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ARABIC =12, /*[0600]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SYRIAC =13, /*[0700]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_THAANA =14, /*[0780]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_DEVANAGARI =15, /*[0900]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BENGALI =16, /*[0980]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GURMUKHI =17, /*[0A00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GUJARATI =18, /*[0A80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ORIYA =19, /*[0B00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_TAMIL =20, /*[0B80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_TELUGU =21, /*[0C00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_KANNADA =22, /*[0C80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MALAYALAM =23, /*[0D00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SINHALA =24, /*[0D80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_THAI =25, /*[0E00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LAO =26, /*[0E80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_TIBETAN =27, /*[0F00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MYANMAR =28, /*[1000]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GEORGIAN =29, /*[10A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HANGUL_JAMO =30, /*[1100]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ETHIOPIC =31, /*[1200]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CHEROKEE =32, /*[13A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_UNIFIED_CANADIAN_ABORIGINAL_SYLLABICS =33, /*[1400]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_OGHAM =34, /*[1680]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_RUNIC =35, /*[16A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_KHMER =36, /*[1780]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MONGOLIAN =37, /*[1800]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LATIN_EXTENDED_ADDITIONAL =38, /*[1E00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GREEK_EXTENDED =39, /*[1F00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GENERAL_PUNCTUATION =40, /*[2000]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SUPERSCRIPTS_AND_SUBSCRIPTS =41, /*[2070]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CURRENCY_SYMBOLS =42, /*[20A0]*/
+
+    /**
+     * Unicode 3.2 renames this block to "Combining Diacritical Marks for Symbols".
+     * @stable ICU 2.0
+     */
+    UBLOCK_COMBINING_MARKS_FOR_SYMBOLS =43, /*[20D0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LETTERLIKE_SYMBOLS =44, /*[2100]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_NUMBER_FORMS =45, /*[2150]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ARROWS =46, /*[2190]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MATHEMATICAL_OPERATORS =47, /*[2200]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MISCELLANEOUS_TECHNICAL =48, /*[2300]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CONTROL_PICTURES =49, /*[2400]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_OPTICAL_CHARACTER_RECOGNITION =50, /*[2440]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ENCLOSED_ALPHANUMERICS =51, /*[2460]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BOX_DRAWING =52, /*[2500]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BLOCK_ELEMENTS =53, /*[2580]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_GEOMETRIC_SHAPES =54, /*[25A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_MISCELLANEOUS_SYMBOLS =55, /*[2600]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_DINGBATS =56, /*[2700]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BRAILLE_PATTERNS =57, /*[2800]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_RADICALS_SUPPLEMENT =58, /*[2E80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_KANGXI_RADICALS =59, /*[2F00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_IDEOGRAPHIC_DESCRIPTION_CHARACTERS =60, /*[2FF0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_SYMBOLS_AND_PUNCTUATION =61, /*[3000]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HIRAGANA =62, /*[3040]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_KATAKANA =63, /*[30A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BOPOMOFO =64, /*[3100]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HANGUL_COMPATIBILITY_JAMO =65, /*[3130]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_KANBUN =66, /*[3190]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_BOPOMOFO_EXTENDED =67, /*[31A0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ENCLOSED_CJK_LETTERS_AND_MONTHS =68, /*[3200]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_COMPATIBILITY =69, /*[3300]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_A =70, /*[3400]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS =71, /*[4E00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_YI_SYLLABLES =72, /*[A000]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_YI_RADICALS =73, /*[A490]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HANGUL_SYLLABLES =74, /*[AC00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HIGH_SURROGATES =75, /*[D800]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HIGH_PRIVATE_USE_SURROGATES =76, /*[DB80]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_LOW_SURROGATES =77, /*[DC00]*/
+
+    /**
+     * Same as UBLOCK_PRIVATE_USE.
+     * Until Unicode 3.1.1, the corresponding block name was "Private Use",
+     * and multiple code point ranges had this block.
+     * Unicode 3.2 renames the block for the BMP PUA to "Private Use Area" and
+     * adds separate blocks for the supplementary PUAs.
+     *
+     * @stable ICU 2.0
+     */
+    UBLOCK_PRIVATE_USE_AREA =78, /*[E000]*/
+    /**
+     * Same as UBLOCK_PRIVATE_USE_AREA.
+     * Until Unicode 3.1.1, the corresponding block name was "Private Use",
+     * and multiple code point ranges had this block.
+     * Unicode 3.2 renames the block for the BMP PUA to "Private Use Area" and
+     * adds separate blocks for the supplementary PUAs.
+     *
+     * @stable ICU 2.0
+     */
+    UBLOCK_PRIVATE_USE = UBLOCK_PRIVATE_USE_AREA,
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_COMPATIBILITY_IDEOGRAPHS =79, /*[F900]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ALPHABETIC_PRESENTATION_FORMS =80, /*[FB00]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ARABIC_PRESENTATION_FORMS_A =81, /*[FB50]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_COMBINING_HALF_MARKS =82, /*[FE20]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_COMPATIBILITY_FORMS =83, /*[FE30]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SMALL_FORM_VARIANTS =84, /*[FE50]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_ARABIC_PRESENTATION_FORMS_B =85, /*[FE70]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_SPECIALS =86, /*[FFF0]*/
+
+    /** @stable ICU 2.0 */
+    UBLOCK_HALFWIDTH_AND_FULLWIDTH_FORMS =87, /*[FF00]*/
+
+    /* New blocks in Unicode 3.1 */
+
+    /** @stable ICU 2.0 */
+    UBLOCK_OLD_ITALIC = 88, /*[10300]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_GOTHIC = 89, /*[10330]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_DESERET = 90, /*[10400]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_BYZANTINE_MUSICAL_SYMBOLS = 91, /*[1D000]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_MUSICAL_SYMBOLS = 92, /*[1D100]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_MATHEMATICAL_ALPHANUMERIC_SYMBOLS = 93, /*[1D400]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_B  = 94, /*[20000]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_CJK_COMPATIBILITY_IDEOGRAPHS_SUPPLEMENT = 95, /*[2F800]*/
+    /** @stable ICU 2.0 */
+    UBLOCK_TAGS = 96, /*[E0000]*/
+
+    /* New blocks in Unicode 3.2 */
+
+    /** @stable ICU 3.0  */
+    UBLOCK_CYRILLIC_SUPPLEMENT = 97, /*[0500]*/
+    /**
+     * Unicode 4.0.1 renames the "Cyrillic Supplementary" block to "Cyrillic Supplement".
+     * @stable ICU 2.2
+     */
+    UBLOCK_CYRILLIC_SUPPLEMENTARY = UBLOCK_CYRILLIC_SUPPLEMENT,
+    /** @stable ICU 2.2 */
+    UBLOCK_TAGALOG = 98, /*[1700]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_HANUNOO = 99, /*[1720]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_BUHID = 100, /*[1740]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_TAGBANWA = 101, /*[1760]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_MISCELLANEOUS_MATHEMATICAL_SYMBOLS_A = 102, /*[27C0]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_SUPPLEMENTAL_ARROWS_A = 103, /*[27F0]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_SUPPLEMENTAL_ARROWS_B = 104, /*[2900]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_MISCELLANEOUS_MATHEMATICAL_SYMBOLS_B = 105, /*[2980]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_SUPPLEMENTAL_MATHEMATICAL_OPERATORS = 106, /*[2A00]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_KATAKANA_PHONETIC_EXTENSIONS = 107, /*[31F0]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_VARIATION_SELECTORS = 108, /*[FE00]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_SUPPLEMENTARY_PRIVATE_USE_AREA_A = 109, /*[F0000]*/
+    /** @stable ICU 2.2 */
+    UBLOCK_SUPPLEMENTARY_PRIVATE_USE_AREA_B = 110, /*[100000]*/
+
+    /* New blocks in Unicode 4 */
+
+    /** @stable ICU 2.6 */
+    UBLOCK_LIMBU = 111, /*[1900]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_TAI_LE = 112, /*[1950]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_KHMER_SYMBOLS = 113, /*[19E0]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_PHONETIC_EXTENSIONS = 114, /*[1D00]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_MISCELLANEOUS_SYMBOLS_AND_ARROWS = 115, /*[2B00]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_YIJING_HEXAGRAM_SYMBOLS = 116, /*[4DC0]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_LINEAR_B_SYLLABARY = 117, /*[10000]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_LINEAR_B_IDEOGRAMS = 118, /*[10080]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_AEGEAN_NUMBERS = 119, /*[10100]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_UGARITIC = 120, /*[10380]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_SHAVIAN = 121, /*[10450]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_OSMANYA = 122, /*[10480]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_CYPRIOT_SYLLABARY = 123, /*[10800]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_TAI_XUAN_JING_SYMBOLS = 124, /*[1D300]*/
+    /** @stable ICU 2.6 */
+    UBLOCK_VARIATION_SELECTORS_SUPPLEMENT = 125, /*[E0100]*/
+
+    /* New blocks in Unicode 4.1 */
+
+    /** @stable ICU 3.4 */
+    UBLOCK_ANCIENT_GREEK_MUSICAL_NOTATION = 126, /*[1D200]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_ANCIENT_GREEK_NUMBERS = 127, /*[10140]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_ARABIC_SUPPLEMENT = 128, /*[0750]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_BUGINESE = 129, /*[1A00]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_CJK_STROKES = 130, /*[31C0]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_COMBINING_DIACRITICAL_MARKS_SUPPLEMENT = 131, /*[1DC0]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_COPTIC = 132, /*[2C80]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_ETHIOPIC_EXTENDED = 133, /*[2D80]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_ETHIOPIC_SUPPLEMENT = 134, /*[1380]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_GEORGIAN_SUPPLEMENT = 135, /*[2D00]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_GLAGOLITIC = 136, /*[2C00]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_KHAROSHTHI = 137, /*[10A00]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_MODIFIER_TONE_LETTERS = 138, /*[A700]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_NEW_TAI_LUE = 139, /*[1980]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_OLD_PERSIAN = 140, /*[103A0]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_PHONETIC_EXTENSIONS_SUPPLEMENT = 141, /*[1D80]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_SUPPLEMENTAL_PUNCTUATION = 142, /*[2E00]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_SYLOTI_NAGRI = 143, /*[A800]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_TIFINAGH = 144, /*[2D30]*/
+    /** @stable ICU 3.4 */
+    UBLOCK_VERTICAL_FORMS = 145, /*[FE10]*/
+
+    /* New blocks in Unicode 5.0 */
+
+    /** @stable ICU 3.6 */
+    UBLOCK_NKO = 146, /*[07C0]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_BALINESE = 147, /*[1B00]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_LATIN_EXTENDED_C = 148, /*[2C60]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_LATIN_EXTENDED_D = 149, /*[A720]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_PHAGS_PA = 150, /*[A840]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_PHOENICIAN = 151, /*[10900]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_CUNEIFORM = 152, /*[12000]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_CUNEIFORM_NUMBERS_AND_PUNCTUATION = 153, /*[12400]*/
+    /** @stable ICU 3.6 */
+    UBLOCK_COUNTING_ROD_NUMERALS = 154, /*[1D360]*/
+
+    /* New blocks in Unicode 5.1 */
+
+    /** @stable ICU 4.0 */
+    UBLOCK_SUNDANESE = 155, /*[1B80]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_LEPCHA = 156, /*[1C00]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_OL_CHIKI = 157, /*[1C50]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_CYRILLIC_EXTENDED_A = 158, /*[2DE0]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_VAI = 159, /*[A500]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_CYRILLIC_EXTENDED_B = 160, /*[A640]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_SAURASHTRA = 161, /*[A880]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_KAYAH_LI = 162, /*[A900]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_REJANG = 163, /*[A930]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_CHAM = 164, /*[AA00]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_ANCIENT_SYMBOLS = 165, /*[10190]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_PHAISTOS_DISC = 166, /*[101D0]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_LYCIAN = 167, /*[10280]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_CARIAN = 168, /*[102A0]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_LYDIAN = 169, /*[10920]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_MAHJONG_TILES = 170, /*[1F000]*/
+    /** @stable ICU 4.0 */
+    UBLOCK_DOMINO_TILES = 171, /*[1F030]*/
+
+    /* New blocks in Unicode 5.2 */
+
+    /** @stable ICU 4.4 */
+    UBLOCK_SAMARITAN = 172, /*[0800]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_UNIFIED_CANADIAN_ABORIGINAL_SYLLABICS_EXTENDED = 173, /*[18B0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_TAI_THAM = 174, /*[1A20]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_VEDIC_EXTENSIONS = 175, /*[1CD0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_LISU = 176, /*[A4D0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_BAMUM = 177, /*[A6A0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_COMMON_INDIC_NUMBER_FORMS = 178, /*[A830]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_DEVANAGARI_EXTENDED = 179, /*[A8E0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_HANGUL_JAMO_EXTENDED_A = 180, /*[A960]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_JAVANESE = 181, /*[A980]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_MYANMAR_EXTENDED_A = 182, /*[AA60]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_TAI_VIET = 183, /*[AA80]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_MEETEI_MAYEK = 184, /*[ABC0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_HANGUL_JAMO_EXTENDED_B = 185, /*[D7B0]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_IMPERIAL_ARAMAIC = 186, /*[10840]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_OLD_SOUTH_ARABIAN = 187, /*[10A60]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_AVESTAN = 188, /*[10B00]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_INSCRIPTIONAL_PARTHIAN = 189, /*[10B40]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_INSCRIPTIONAL_PAHLAVI = 190, /*[10B60]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_OLD_TURKIC = 191, /*[10C00]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_RUMI_NUMERAL_SYMBOLS = 192, /*[10E60]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_KAITHI = 193, /*[11080]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_EGYPTIAN_HIEROGLYPHS = 194, /*[13000]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_ENCLOSED_ALPHANUMERIC_SUPPLEMENT = 195, /*[1F100]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_ENCLOSED_IDEOGRAPHIC_SUPPLEMENT = 196, /*[1F200]*/
+    /** @stable ICU 4.4 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_C = 197, /*[2A700]*/
+
+    /* New blocks in Unicode 6.0 */
+
+    /** @stable ICU 4.6 */
+    UBLOCK_MANDAIC = 198, /*[0840]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_BATAK = 199, /*[1BC0]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_ETHIOPIC_EXTENDED_A = 200, /*[AB00]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_BRAHMI = 201, /*[11000]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_BAMUM_SUPPLEMENT = 202, /*[16800]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_KANA_SUPPLEMENT = 203, /*[1B000]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_PLAYING_CARDS = 204, /*[1F0A0]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_MISCELLANEOUS_SYMBOLS_AND_PICTOGRAPHS = 205, /*[1F300]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_EMOTICONS = 206, /*[1F600]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_TRANSPORT_AND_MAP_SYMBOLS = 207, /*[1F680]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_ALCHEMICAL_SYMBOLS = 208, /*[1F700]*/
+    /** @stable ICU 4.6 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_D = 209, /*[2B740]*/
+
+    /* New blocks in Unicode 6.1 */
+
+    /** @stable ICU 49 */
+    UBLOCK_ARABIC_EXTENDED_A = 210, /*[08A0]*/
+    /** @stable ICU 49 */
+    UBLOCK_ARABIC_MATHEMATICAL_ALPHABETIC_SYMBOLS = 211, /*[1EE00]*/
+    /** @stable ICU 49 */
+    UBLOCK_CHAKMA = 212, /*[11100]*/
+    /** @stable ICU 49 */
+    UBLOCK_MEETEI_MAYEK_EXTENSIONS = 213, /*[AAE0]*/
+    /** @stable ICU 49 */
+    UBLOCK_MEROITIC_CURSIVE = 214, /*[109A0]*/
+    /** @stable ICU 49 */
+    UBLOCK_MEROITIC_HIEROGLYPHS = 215, /*[10980]*/
+    /** @stable ICU 49 */
+    UBLOCK_MIAO = 216, /*[16F00]*/
+    /** @stable ICU 49 */
+    UBLOCK_SHARADA = 217, /*[11180]*/
+    /** @stable ICU 49 */
+    UBLOCK_SORA_SOMPENG = 218, /*[110D0]*/
+    /** @stable ICU 49 */
+    UBLOCK_SUNDANESE_SUPPLEMENT = 219, /*[1CC0]*/
+    /** @stable ICU 49 */
+    UBLOCK_TAKRI = 220, /*[11680]*/
+
+    /* New blocks in Unicode 7.0 */
+
+    /** @stable ICU 54 */
+    UBLOCK_BASSA_VAH = 221, /*[16AD0]*/
+    /** @stable ICU 54 */
+    UBLOCK_CAUCASIAN_ALBANIAN = 222, /*[10530]*/
+    /** @stable ICU 54 */
+    UBLOCK_COPTIC_EPACT_NUMBERS = 223, /*[102E0]*/
+    /** @stable ICU 54 */
+    UBLOCK_COMBINING_DIACRITICAL_MARKS_EXTENDED = 224, /*[1AB0]*/
+    /** @stable ICU 54 */
+    UBLOCK_DUPLOYAN = 225, /*[1BC00]*/
+    /** @stable ICU 54 */
+    UBLOCK_ELBASAN = 226, /*[10500]*/
+    /** @stable ICU 54 */
+    UBLOCK_GEOMETRIC_SHAPES_EXTENDED = 227, /*[1F780]*/
+    /** @stable ICU 54 */
+    UBLOCK_GRANTHA = 228, /*[11300]*/
+    /** @stable ICU 54 */
+    UBLOCK_KHOJKI = 229, /*[11200]*/
+    /** @stable ICU 54 */
+    UBLOCK_KHUDAWADI = 230, /*[112B0]*/
+    /** @stable ICU 54 */
+    UBLOCK_LATIN_EXTENDED_E = 231, /*[AB30]*/
+    /** @stable ICU 54 */
+    UBLOCK_LINEAR_A = 232, /*[10600]*/
+    /** @stable ICU 54 */
+    UBLOCK_MAHAJANI = 233, /*[11150]*/
+    /** @stable ICU 54 */
+    UBLOCK_MANICHAEAN = 234, /*[10AC0]*/
+    /** @stable ICU 54 */
+    UBLOCK_MENDE_KIKAKUI = 235, /*[1E800]*/
+    /** @stable ICU 54 */
+    UBLOCK_MODI = 236, /*[11600]*/
+    /** @stable ICU 54 */
+    UBLOCK_MRO = 237, /*[16A40]*/
+    /** @stable ICU 54 */
+    UBLOCK_MYANMAR_EXTENDED_B = 238, /*[A9E0]*/
+    /** @stable ICU 54 */
+    UBLOCK_NABATAEAN = 239, /*[10880]*/
+    /** @stable ICU 54 */
+    UBLOCK_OLD_NORTH_ARABIAN = 240, /*[10A80]*/
+    /** @stable ICU 54 */
+    UBLOCK_OLD_PERMIC = 241, /*[10350]*/
+    /** @stable ICU 54 */
+    UBLOCK_ORNAMENTAL_DINGBATS = 242, /*[1F650]*/
+    /** @stable ICU 54 */
+    UBLOCK_PAHAWH_HMONG = 243, /*[16B00]*/
+    /** @stable ICU 54 */
+    UBLOCK_PALMYRENE = 244, /*[10860]*/
+    /** @stable ICU 54 */
+    UBLOCK_PAU_CIN_HAU = 245, /*[11AC0]*/
+    /** @stable ICU 54 */
+    UBLOCK_PSALTER_PAHLAVI = 246, /*[10B80]*/
+    /** @stable ICU 54 */
+    UBLOCK_SHORTHAND_FORMAT_CONTROLS = 247, /*[1BCA0]*/
+    /** @stable ICU 54 */
+    UBLOCK_SIDDHAM = 248, /*[11580]*/
+    /** @stable ICU 54 */
+    UBLOCK_SINHALA_ARCHAIC_NUMBERS = 249, /*[111E0]*/
+    /** @stable ICU 54 */
+    UBLOCK_SUPPLEMENTAL_ARROWS_C = 250, /*[1F800]*/
+    /** @stable ICU 54 */
+    UBLOCK_TIRHUTA = 251, /*[11480]*/
+    /** @stable ICU 54 */
+    UBLOCK_WARANG_CITI = 252, /*[118A0]*/
+
+    /* New blocks in Unicode 8.0 */
+
+    /** @stable ICU 56 */
+    UBLOCK_AHOM = 253, /*[11700]*/
+    /** @stable ICU 56 */
+    UBLOCK_ANATOLIAN_HIEROGLYPHS = 254, /*[14400]*/
+    /** @stable ICU 56 */
+    UBLOCK_CHEROKEE_SUPPLEMENT = 255, /*[AB70]*/
+    /** @stable ICU 56 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_E = 256, /*[2B820]*/
+    /** @stable ICU 56 */
+    UBLOCK_EARLY_DYNASTIC_CUNEIFORM = 257, /*[12480]*/
+    /** @stable ICU 56 */
+    UBLOCK_HATRAN = 258, /*[108E0]*/
+    /** @stable ICU 56 */
+    UBLOCK_MULTANI = 259, /*[11280]*/
+    /** @stable ICU 56 */
+    UBLOCK_OLD_HUNGARIAN = 260, /*[10C80]*/
+    /** @stable ICU 56 */
+    UBLOCK_SUPPLEMENTAL_SYMBOLS_AND_PICTOGRAPHS = 261, /*[1F900]*/
+    /** @stable ICU 56 */
+    UBLOCK_SUTTON_SIGNWRITING = 262, /*[1D800]*/
+
+    /* New blocks in Unicode 9.0 */
+
+    /** @stable ICU 58 */
+    UBLOCK_ADLAM = 263, /*[1E900]*/
+    /** @stable ICU 58 */
+    UBLOCK_BHAIKSUKI = 264, /*[11C00]*/
+    /** @stable ICU 58 */
+    UBLOCK_CYRILLIC_EXTENDED_C = 265, /*[1C80]*/
+    /** @stable ICU 58 */
+    UBLOCK_GLAGOLITIC_SUPPLEMENT = 266, /*[1E000]*/
+    /** @stable ICU 58 */
+    UBLOCK_IDEOGRAPHIC_SYMBOLS_AND_PUNCTUATION = 267, /*[16FE0]*/
+    /** @stable ICU 58 */
+    UBLOCK_MARCHEN = 268, /*[11C70]*/
+    /** @stable ICU 58 */
+    UBLOCK_MONGOLIAN_SUPPLEMENT = 269, /*[11660]*/
+    /** @stable ICU 58 */
+    UBLOCK_NEWA = 270, /*[11400]*/
+    /** @stable ICU 58 */
+    UBLOCK_OSAGE = 271, /*[104B0]*/
+    /** @stable ICU 58 */
+    UBLOCK_TANGUT = 272, /*[17000]*/
+    /** @stable ICU 58 */
+    UBLOCK_TANGUT_COMPONENTS = 273, /*[18800]*/
+
+    // New blocks in Unicode 10.0
+
+    /** @stable ICU 60 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_F = 274, /*[2CEB0]*/
+    /** @stable ICU 60 */
+    UBLOCK_KANA_EXTENDED_A = 275, /*[1B100]*/
+    /** @stable ICU 60 */
+    UBLOCK_MASARAM_GONDI = 276, /*[11D00]*/
+    /** @stable ICU 60 */
+    UBLOCK_NUSHU = 277, /*[1B170]*/
+    /** @stable ICU 60 */
+    UBLOCK_SOYOMBO = 278, /*[11A50]*/
+    /** @stable ICU 60 */
+    UBLOCK_SYRIAC_SUPPLEMENT = 279, /*[0860]*/
+    /** @stable ICU 60 */
+    UBLOCK_ZANABAZAR_SQUARE = 280, /*[11A00]*/
+
+    // New blocks in Unicode 11.0
+
+    /** @stable ICU 62 */
+    UBLOCK_CHESS_SYMBOLS = 281, /*[1FA00]*/
+    /** @stable ICU 62 */
+    UBLOCK_DOGRA = 282, /*[11800]*/
+    /** @stable ICU 62 */
+    UBLOCK_GEORGIAN_EXTENDED = 283, /*[1C90]*/
+    /** @stable ICU 62 */
+    UBLOCK_GUNJALA_GONDI = 284, /*[11D60]*/
+    /** @stable ICU 62 */
+    UBLOCK_HANIFI_ROHINGYA = 285, /*[10D00]*/
+    /** @stable ICU 62 */
+    UBLOCK_INDIC_SIYAQ_NUMBERS = 286, /*[1EC70]*/
+    /** @stable ICU 62 */
+    UBLOCK_MAKASAR = 287, /*[11EE0]*/
+    /** @stable ICU 62 */
+    UBLOCK_MAYAN_NUMERALS = 288, /*[1D2E0]*/
+    /** @stable ICU 62 */
+    UBLOCK_MEDEFAIDRIN = 289, /*[16E40]*/
+    /** @stable ICU 62 */
+    UBLOCK_OLD_SOGDIAN = 290, /*[10F00]*/
+    /** @stable ICU 62 */
+    UBLOCK_SOGDIAN = 291, /*[10F30]*/
+
+    // New blocks in Unicode 12.0
+
+    /** @stable ICU 64 */
+    UBLOCK_EGYPTIAN_HIEROGLYPH_FORMAT_CONTROLS = 292, /*[13430]*/
+    /** @stable ICU 64 */
+    UBLOCK_ELYMAIC = 293, /*[10FE0]*/
+    /** @stable ICU 64 */
+    UBLOCK_NANDINAGARI = 294, /*[119A0]*/
+    /** @stable ICU 64 */
+    UBLOCK_NYIAKENG_PUACHUE_HMONG = 295, /*[1E100]*/
+    /** @stable ICU 64 */
+    UBLOCK_OTTOMAN_SIYAQ_NUMBERS = 296, /*[1ED00]*/
+    /** @stable ICU 64 */
+    UBLOCK_SMALL_KANA_EXTENSION = 297, /*[1B130]*/
+    /** @stable ICU 64 */
+    UBLOCK_SYMBOLS_AND_PICTOGRAPHS_EXTENDED_A = 298, /*[1FA70]*/
+    /** @stable ICU 64 */
+    UBLOCK_TAMIL_SUPPLEMENT = 299, /*[11FC0]*/
+    /** @stable ICU 64 */
+    UBLOCK_WANCHO = 300, /*[1E2C0]*/
+
+    // New blocks in Unicode 13.0
+
+    /** @stable ICU 66 */
+    UBLOCK_CHORASMIAN = 301, /*[10FB0]*/
+    /** @stable ICU 66 */
+    UBLOCK_CJK_UNIFIED_IDEOGRAPHS_EXTENSION_G = 302, /*[30000]*/
+    /** @stable ICU 66 */
+    UBLOCK_DIVES_AKURU = 303, /*[11900]*/
+    /** @stable ICU 66 */
+    UBLOCK_KHITAN_SMALL_SCRIPT = 304, /*[18B00]*/
+    /** @stable ICU 66 */
+    UBLOCK_LISU_SUPPLEMENT = 305, /*[11FB0]*/
+    /** @stable ICU 66 */
+    UBLOCK_SYMBOLS_FOR_LEGACY_COMPUTING = 306, /*[1FB00]*/
+    /** @stable ICU 66 */
+    UBLOCK_TANGUT_SUPPLEMENT = 307, /*[18D00]*/
+    /** @stable ICU 66 */
+    UBLOCK_YEZIDI = 308, /*[10E80]*/
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UBlockCode value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_BLOCK).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UBLOCK_COUNT = 309,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /** @stable ICU 2.0 */
+    UBLOCK_INVALID_CODE=-1
+};
+
+/** @stable ICU 2.0 */
+typedef enum UBlockCode UBlockCode;
+
+/**
+ * East Asian Width constants.
+ *
+ * @see UCHAR_EAST_ASIAN_WIDTH
+ * @see u_getIntPropertyValue
+ * @stable ICU 2.2
+ */
+typedef enum UEastAsianWidth {
+    /*
+     * Note: UEastAsianWidth constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_EA_<Unicode East_Asian_Width value name>
+     */
+
+    U_EA_NEUTRAL,   /*[N]*/
+    U_EA_AMBIGUOUS, /*[A]*/
+    U_EA_HALFWIDTH, /*[H]*/
+    U_EA_FULLWIDTH, /*[F]*/
+    U_EA_NARROW,    /*[Na]*/
+    U_EA_WIDE,      /*[W]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UEastAsianWidth value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_EAST_ASIAN_WIDTH).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_EA_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UEastAsianWidth;
+
+/**
+ * Selector constants for u_charName().
+ * u_charName() returns the "modern" name of a
+ * Unicode character; or the name that was defined in
+ * Unicode version 1.0, before the Unicode standard merged
+ * with ISO-10646; or an "extended" name that gives each
+ * Unicode code point a unique name.
+ *
+ * @see u_charName
+ * @stable ICU 2.0
+ */
+typedef enum UCharNameChoice {
+    /** Unicode character name (Name property). @stable ICU 2.0 */
+    U_UNICODE_CHAR_NAME,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * The Unicode_1_Name property value which is of little practical value.
+     * Beginning with ICU 49, ICU APIs return an empty string for this name choice.
+     * @deprecated ICU 49
+     */
+    U_UNICODE_10_CHAR_NAME,
+#endif  /* U_HIDE_DEPRECATED_API */
+    /** Standard or synthetic character name. @stable ICU 2.0 */
+    U_EXTENDED_CHAR_NAME = U_UNICODE_CHAR_NAME+2,
+    /** Corrected name from NameAliases.txt. @stable ICU 4.4 */
+    U_CHAR_NAME_ALIAS,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UCharNameChoice value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_CHAR_NAME_CHOICE_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UCharNameChoice;
+
+/**
+ * Selector constants for u_getPropertyName() and
+ * u_getPropertyValueName().  These selectors are used to choose which
+ * name is returned for a given property or value.  All properties and
+ * values have a long name.  Most have a short name, but some do not.
+ * Unicode allows for additional names, beyond the long and short
+ * name, which would be indicated by U_LONG_PROPERTY_NAME + i, where
+ * i=1, 2,...
+ *
+ * @see u_getPropertyName()
+ * @see u_getPropertyValueName()
+ * @stable ICU 2.4
+ */
+typedef enum UPropertyNameChoice {
+    U_SHORT_PROPERTY_NAME,
+    U_LONG_PROPERTY_NAME,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UPropertyNameChoice value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_PROPERTY_NAME_CHOICE_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UPropertyNameChoice;
+
+/**
+ * Decomposition Type constants.
+ *
+ * @see UCHAR_DECOMPOSITION_TYPE
+ * @stable ICU 2.2
+ */
+typedef enum UDecompositionType {
+    /*
+     * Note: UDecompositionType constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_DT_<Unicode Decomposition_Type value name>
+     */
+
+    U_DT_NONE,              /*[none]*/
+    U_DT_CANONICAL,         /*[can]*/
+    U_DT_COMPAT,            /*[com]*/
+    U_DT_CIRCLE,            /*[enc]*/
+    U_DT_FINAL,             /*[fin]*/
+    U_DT_FONT,              /*[font]*/
+    U_DT_FRACTION,          /*[fra]*/
+    U_DT_INITIAL,           /*[init]*/
+    U_DT_ISOLATED,          /*[iso]*/
+    U_DT_MEDIAL,            /*[med]*/
+    U_DT_NARROW,            /*[nar]*/
+    U_DT_NOBREAK,           /*[nb]*/
+    U_DT_SMALL,             /*[sml]*/
+    U_DT_SQUARE,            /*[sqr]*/
+    U_DT_SUB,               /*[sub]*/
+    U_DT_SUPER,             /*[sup]*/
+    U_DT_VERTICAL,          /*[vert]*/
+    U_DT_WIDE,              /*[wide]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UDecompositionType value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_DECOMPOSITION_TYPE).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_DT_COUNT /* 18 */
+#endif  // U_HIDE_DEPRECATED_API
+} UDecompositionType;
+
+/**
+ * Joining Type constants.
+ *
+ * @see UCHAR_JOINING_TYPE
+ * @stable ICU 2.2
+ */
+typedef enum UJoiningType {
+    /*
+     * Note: UJoiningType constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_JT_<Unicode Joining_Type value name>
+     */
+
+    U_JT_NON_JOINING,       /*[U]*/
+    U_JT_JOIN_CAUSING,      /*[C]*/
+    U_JT_DUAL_JOINING,      /*[D]*/
+    U_JT_LEFT_JOINING,      /*[L]*/
+    U_JT_RIGHT_JOINING,     /*[R]*/
+    U_JT_TRANSPARENT,       /*[T]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UJoiningType value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_JOINING_TYPE).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_JT_COUNT /* 6 */
+#endif  // U_HIDE_DEPRECATED_API
+} UJoiningType;
+
+/**
+ * Joining Group constants.
+ *
+ * @see UCHAR_JOINING_GROUP
+ * @stable ICU 2.2
+ */
+typedef enum UJoiningGroup {
+    /*
+     * Note: UJoiningGroup constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_JG_<Unicode Joining_Group value name>
+     */
+
+    U_JG_NO_JOINING_GROUP,
+    U_JG_AIN,
+    U_JG_ALAPH,
+    U_JG_ALEF,
+    U_JG_BEH,
+    U_JG_BETH,
+    U_JG_DAL,
+    U_JG_DALATH_RISH,
+    U_JG_E,
+    U_JG_FEH,
+    U_JG_FINAL_SEMKATH,
+    U_JG_GAF,
+    U_JG_GAMAL,
+    U_JG_HAH,
+    U_JG_TEH_MARBUTA_GOAL,  /**< @stable ICU 4.6 */
+    U_JG_HAMZA_ON_HEH_GOAL=U_JG_TEH_MARBUTA_GOAL,
+    U_JG_HE,
+    U_JG_HEH,
+    U_JG_HEH_GOAL,
+    U_JG_HETH,
+    U_JG_KAF,
+    U_JG_KAPH,
+    U_JG_KNOTTED_HEH,
+    U_JG_LAM,
+    U_JG_LAMADH,
+    U_JG_MEEM,
+    U_JG_MIM,
+    U_JG_NOON,
+    U_JG_NUN,
+    U_JG_PE,
+    U_JG_QAF,
+    U_JG_QAPH,
+    U_JG_REH,
+    U_JG_REVERSED_PE,
+    U_JG_SAD,
+    U_JG_SADHE,
+    U_JG_SEEN,
+    U_JG_SEMKATH,
+    U_JG_SHIN,
+    U_JG_SWASH_KAF,
+    U_JG_SYRIAC_WAW,
+    U_JG_TAH,
+    U_JG_TAW,
+    U_JG_TEH_MARBUTA,
+    U_JG_TETH,
+    U_JG_WAW,
+    U_JG_YEH,
+    U_JG_YEH_BARREE,
+    U_JG_YEH_WITH_TAIL,
+    U_JG_YUDH,
+    U_JG_YUDH_HE,
+    U_JG_ZAIN,
+    U_JG_FE,        /**< @stable ICU 2.6 */
+    U_JG_KHAPH,     /**< @stable ICU 2.6 */
+    U_JG_ZHAIN,     /**< @stable ICU 2.6 */
+    U_JG_BURUSHASKI_YEH_BARREE, /**< @stable ICU 4.0 */
+    U_JG_FARSI_YEH, /**< @stable ICU 4.4 */
+    U_JG_NYA,       /**< @stable ICU 4.4 */
+    U_JG_ROHINGYA_YEH,  /**< @stable ICU 49 */
+    U_JG_MANICHAEAN_ALEPH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_AYIN,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_BETH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_DALETH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_DHAMEDH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_FIVE,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_GIMEL,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_HETH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_HUNDRED,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_KAPH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_LAMEDH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_MEM,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_NUN,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_ONE,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_PE,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_QOPH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_RESH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_SADHE,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_SAMEKH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_TAW,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_TEN,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_TETH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_THAMEDH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_TWENTY,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_WAW,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_YODH,  /**< @stable ICU 54 */
+    U_JG_MANICHAEAN_ZAYIN,  /**< @stable ICU 54 */
+    U_JG_STRAIGHT_WAW,  /**< @stable ICU 54 */
+    U_JG_AFRICAN_FEH,  /**< @stable ICU 58 */
+    U_JG_AFRICAN_NOON,  /**< @stable ICU 58 */
+    U_JG_AFRICAN_QAF,  /**< @stable ICU 58 */
+
+    U_JG_MALAYALAM_BHA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_JA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_LLA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_LLLA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NGA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NNA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NNNA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_NYA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_RA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_SSA,  /**< @stable ICU 60 */
+    U_JG_MALAYALAM_TTA,  /**< @stable ICU 60 */
+
+    U_JG_HANIFI_ROHINGYA_KINNA_YA,  /**< @stable ICU 62 */
+    U_JG_HANIFI_ROHINGYA_PA,  /**< @stable ICU 62 */
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UJoiningGroup value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_JOINING_GROUP).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_JG_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UJoiningGroup;
+
+/**
+ * Grapheme Cluster Break constants.
+ *
+ * @see UCHAR_GRAPHEME_CLUSTER_BREAK
+ * @stable ICU 3.4
+ */
+typedef enum UGraphemeClusterBreak {
+    /*
+     * Note: UGraphemeClusterBreak constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_GCB_<Unicode Grapheme_Cluster_Break value name>
+     */
+
+    U_GCB_OTHER = 0,            /*[XX]*/
+    U_GCB_CONTROL = 1,          /*[CN]*/
+    U_GCB_CR = 2,               /*[CR]*/
+    U_GCB_EXTEND = 3,           /*[EX]*/
+    U_GCB_L = 4,                /*[L]*/
+    U_GCB_LF = 5,               /*[LF]*/
+    U_GCB_LV = 6,               /*[LV]*/
+    U_GCB_LVT = 7,              /*[LVT]*/
+    U_GCB_T = 8,                /*[T]*/
+    U_GCB_V = 9,                /*[V]*/
+    /** @stable ICU 4.0 */
+    U_GCB_SPACING_MARK = 10,    /*[SM]*/ /* from here on: new in Unicode 5.1/ICU 4.0 */
+    /** @stable ICU 4.0 */
+    U_GCB_PREPEND = 11,         /*[PP]*/
+    /** @stable ICU 50 */
+    U_GCB_REGIONAL_INDICATOR = 12,  /*[RI]*/ /* new in Unicode 6.2/ICU 50 */
+    /** @stable ICU 58 */
+    U_GCB_E_BASE = 13,          /*[EB]*/ /* from here on: new in Unicode 9.0/ICU 58 */
+    /** @stable ICU 58 */
+    U_GCB_E_BASE_GAZ = 14,      /*[EBG]*/
+    /** @stable ICU 58 */
+    U_GCB_E_MODIFIER = 15,      /*[EM]*/
+    /** @stable ICU 58 */
+    U_GCB_GLUE_AFTER_ZWJ = 16,  /*[GAZ]*/
+    /** @stable ICU 58 */
+    U_GCB_ZWJ = 17,             /*[ZWJ]*/
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UGraphemeClusterBreak value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_GRAPHEME_CLUSTER_BREAK).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_GCB_COUNT = 18
+#endif  // U_HIDE_DEPRECATED_API
+} UGraphemeClusterBreak;
+
+/**
+ * Word Break constants.
+ * (UWordBreak is a pre-existing enum type in ubrk.h for word break status tags.)
+ *
+ * @see UCHAR_WORD_BREAK
+ * @stable ICU 3.4
+ */
+typedef enum UWordBreakValues {
+    /*
+     * Note: UWordBreakValues constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_WB_<Unicode Word_Break value name>
+     */
+
+    U_WB_OTHER = 0,             /*[XX]*/
+    U_WB_ALETTER = 1,           /*[LE]*/
+    U_WB_FORMAT = 2,            /*[FO]*/
+    U_WB_KATAKANA = 3,          /*[KA]*/
+    U_WB_MIDLETTER = 4,         /*[ML]*/
+    U_WB_MIDNUM = 5,            /*[MN]*/
+    U_WB_NUMERIC = 6,           /*[NU]*/
+    U_WB_EXTENDNUMLET = 7,      /*[EX]*/
+    /** @stable ICU 4.0 */
+    U_WB_CR = 8,                /*[CR]*/ /* from here on: new in Unicode 5.1/ICU 4.0 */
+    /** @stable ICU 4.0 */
+    U_WB_EXTEND = 9,            /*[Extend]*/
+    /** @stable ICU 4.0 */
+    U_WB_LF = 10,               /*[LF]*/
+    /** @stable ICU 4.0 */
+    U_WB_MIDNUMLET =11,         /*[MB]*/
+    /** @stable ICU 4.0 */
+    U_WB_NEWLINE =12,           /*[NL]*/
+    /** @stable ICU 50 */
+    U_WB_REGIONAL_INDICATOR = 13,   /*[RI]*/ /* new in Unicode 6.2/ICU 50 */
+    /** @stable ICU 52 */
+    U_WB_HEBREW_LETTER = 14,    /*[HL]*/ /* from here on: new in Unicode 6.3/ICU 52 */
+    /** @stable ICU 52 */
+    U_WB_SINGLE_QUOTE = 15,     /*[SQ]*/
+    /** @stable ICU 52 */
+    U_WB_DOUBLE_QUOTE = 16,     /*[DQ]*/
+    /** @stable ICU 58 */
+    U_WB_E_BASE = 17,           /*[EB]*/ /* from here on: new in Unicode 9.0/ICU 58 */
+    /** @stable ICU 58 */
+    U_WB_E_BASE_GAZ = 18,       /*[EBG]*/
+    /** @stable ICU 58 */
+    U_WB_E_MODIFIER = 19,       /*[EM]*/
+    /** @stable ICU 58 */
+    U_WB_GLUE_AFTER_ZWJ = 20,   /*[GAZ]*/
+    /** @stable ICU 58 */
+    U_WB_ZWJ = 21,              /*[ZWJ]*/
+    /** @stable ICU 62 */
+    U_WB_WSEGSPACE = 22,        /*[WSEGSPACE]*/
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UWordBreakValues value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_WORD_BREAK).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_WB_COUNT = 23
+#endif  // U_HIDE_DEPRECATED_API
+} UWordBreakValues;
+
+/**
+ * Sentence Break constants.
+ *
+ * @see UCHAR_SENTENCE_BREAK
+ * @stable ICU 3.4
+ */
+typedef enum USentenceBreak {
+    /*
+     * Note: USentenceBreak constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_SB_<Unicode Sentence_Break value name>
+     */
+
+    U_SB_OTHER = 0,             /*[XX]*/
+    U_SB_ATERM = 1,             /*[AT]*/
+    U_SB_CLOSE = 2,             /*[CL]*/
+    U_SB_FORMAT = 3,            /*[FO]*/
+    U_SB_LOWER = 4,             /*[LO]*/
+    U_SB_NUMERIC = 5,           /*[NU]*/
+    U_SB_OLETTER = 6,           /*[LE]*/
+    U_SB_SEP = 7,               /*[SE]*/
+    U_SB_SP = 8,                /*[SP]*/
+    U_SB_STERM = 9,             /*[ST]*/
+    U_SB_UPPER = 10,            /*[UP]*/
+    U_SB_CR = 11,               /*[CR]*/ /* from here on: new in Unicode 5.1/ICU 4.0 */
+    U_SB_EXTEND = 12,           /*[EX]*/
+    U_SB_LF = 13,               /*[LF]*/
+    U_SB_SCONTINUE = 14,        /*[SC]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal USentenceBreak value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_SENTENCE_BREAK).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_SB_COUNT = 15
+#endif  // U_HIDE_DEPRECATED_API
+} USentenceBreak;
+
+/**
+ * Line Break constants.
+ *
+ * @see UCHAR_LINE_BREAK
+ * @stable ICU 2.2
+ */
+typedef enum ULineBreak {
+    /*
+     * Note: ULineBreak constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_LB_<Unicode Line_Break value name>
+     */
+
+    U_LB_UNKNOWN = 0,           /*[XX]*/
+    U_LB_AMBIGUOUS = 1,         /*[AI]*/
+    U_LB_ALPHABETIC = 2,        /*[AL]*/
+    U_LB_BREAK_BOTH = 3,        /*[B2]*/
+    U_LB_BREAK_AFTER = 4,       /*[BA]*/
+    U_LB_BREAK_BEFORE = 5,      /*[BB]*/
+    U_LB_MANDATORY_BREAK = 6,   /*[BK]*/
+    U_LB_CONTINGENT_BREAK = 7,  /*[CB]*/
+    U_LB_CLOSE_PUNCTUATION = 8, /*[CL]*/
+    U_LB_COMBINING_MARK = 9,    /*[CM]*/
+    U_LB_CARRIAGE_RETURN = 10,   /*[CR]*/
+    U_LB_EXCLAMATION = 11,       /*[EX]*/
+    U_LB_GLUE = 12,              /*[GL]*/
+    U_LB_HYPHEN = 13,            /*[HY]*/
+    U_LB_IDEOGRAPHIC = 14,       /*[ID]*/
+    /** Renamed from the misspelled "inseperable" in Unicode 4.0.1/ICU 3.0 @stable ICU 3.0 */
+    U_LB_INSEPARABLE = 15,       /*[IN]*/
+    U_LB_INSEPERABLE = U_LB_INSEPARABLE,
+    U_LB_INFIX_NUMERIC = 16,     /*[IS]*/
+    U_LB_LINE_FEED = 17,         /*[LF]*/
+    U_LB_NONSTARTER = 18,        /*[NS]*/
+    U_LB_NUMERIC = 19,           /*[NU]*/
+    U_LB_OPEN_PUNCTUATION = 20,  /*[OP]*/
+    U_LB_POSTFIX_NUMERIC = 21,   /*[PO]*/
+    U_LB_PREFIX_NUMERIC = 22,    /*[PR]*/
+    U_LB_QUOTATION = 23,         /*[QU]*/
+    U_LB_COMPLEX_CONTEXT = 24,   /*[SA]*/
+    U_LB_SURROGATE = 25,         /*[SG]*/
+    U_LB_SPACE = 26,             /*[SP]*/
+    U_LB_BREAK_SYMBOLS = 27,     /*[SY]*/
+    U_LB_ZWSPACE = 28,           /*[ZW]*/
+    /** @stable ICU 2.6 */
+    U_LB_NEXT_LINE = 29,         /*[NL]*/ /* from here on: new in Unicode 4/ICU 2.6 */
+    /** @stable ICU 2.6 */
+    U_LB_WORD_JOINER = 30,       /*[WJ]*/
+    /** @stable ICU 3.4 */
+    U_LB_H2 = 31,                /*[H2]*/ /* from here on: new in Unicode 4.1/ICU 3.4 */
+    /** @stable ICU 3.4 */
+    U_LB_H3 = 32,                /*[H3]*/
+    /** @stable ICU 3.4 */
+    U_LB_JL = 33,                /*[JL]*/
+    /** @stable ICU 3.4 */
+    U_LB_JT = 34,                /*[JT]*/
+    /** @stable ICU 3.4 */
+    U_LB_JV = 35,                /*[JV]*/
+    /** @stable ICU 4.4 */
+    U_LB_CLOSE_PARENTHESIS = 36, /*[CP]*/ /* new in Unicode 5.2/ICU 4.4 */
+    /** @stable ICU 49 */
+    U_LB_CONDITIONAL_JAPANESE_STARTER = 37,/*[CJ]*/ /* new in Unicode 6.1/ICU 49 */
+    /** @stable ICU 49 */
+    U_LB_HEBREW_LETTER = 38,     /*[HL]*/ /* new in Unicode 6.1/ICU 49 */
+    /** @stable ICU 50 */
+    U_LB_REGIONAL_INDICATOR = 39,/*[RI]*/ /* new in Unicode 6.2/ICU 50 */
+    /** @stable ICU 58 */
+    U_LB_E_BASE = 40,            /*[EB]*/ /* from here on: new in Unicode 9.0/ICU 58 */
+    /** @stable ICU 58 */
+    U_LB_E_MODIFIER = 41,        /*[EM]*/
+    /** @stable ICU 58 */
+    U_LB_ZWJ = 42,               /*[ZWJ]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal ULineBreak value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_LINE_BREAK).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_LB_COUNT = 43
+#endif  // U_HIDE_DEPRECATED_API
+} ULineBreak;
+
+/**
+ * Numeric Type constants.
+ *
+ * @see UCHAR_NUMERIC_TYPE
+ * @stable ICU 2.2
+ */
+typedef enum UNumericType {
+    /*
+     * Note: UNumericType constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_NT_<Unicode Numeric_Type value name>
+     */
+
+    U_NT_NONE,              /*[None]*/
+    U_NT_DECIMAL,           /*[de]*/
+    U_NT_DIGIT,             /*[di]*/
+    U_NT_NUMERIC,           /*[nu]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UNumericType value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_NUMERIC_TYPE).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_NT_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UNumericType;
+
+/**
+ * Hangul Syllable Type constants.
+ *
+ * @see UCHAR_HANGUL_SYLLABLE_TYPE
+ * @stable ICU 2.6
+ */
+typedef enum UHangulSyllableType {
+    /*
+     * Note: UHangulSyllableType constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_HST_<Unicode Hangul_Syllable_Type value name>
+     */
+
+    U_HST_NOT_APPLICABLE,   /*[NA]*/
+    U_HST_LEADING_JAMO,     /*[L]*/
+    U_HST_VOWEL_JAMO,       /*[V]*/
+    U_HST_TRAILING_JAMO,    /*[T]*/
+    U_HST_LV_SYLLABLE,      /*[LV]*/
+    U_HST_LVT_SYLLABLE,     /*[LVT]*/
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UHangulSyllableType value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_HANGUL_SYLLABLE_TYPE).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_HST_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UHangulSyllableType;
+
+/**
+ * Indic Positional Category constants.
+ *
+ * @see UCHAR_INDIC_POSITIONAL_CATEGORY
+ * @stable ICU 63
+ */
+typedef enum UIndicPositionalCategory {
+    /*
+     * Note: UIndicPositionalCategory constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_INPC_<Unicode Indic_Positional_Category value name>
+     */
+
+    /** @stable ICU 63 */
+    U_INPC_NA,
+    /** @stable ICU 63 */
+    U_INPC_BOTTOM,
+    /** @stable ICU 63 */
+    U_INPC_BOTTOM_AND_LEFT,
+    /** @stable ICU 63 */
+    U_INPC_BOTTOM_AND_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_LEFT,
+    /** @stable ICU 63 */
+    U_INPC_LEFT_AND_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_OVERSTRUCK,
+    /** @stable ICU 63 */
+    U_INPC_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_TOP,
+    /** @stable ICU 63 */
+    U_INPC_TOP_AND_BOTTOM,
+    /** @stable ICU 63 */
+    U_INPC_TOP_AND_BOTTOM_AND_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_TOP_AND_LEFT,
+    /** @stable ICU 63 */
+    U_INPC_TOP_AND_LEFT_AND_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_TOP_AND_RIGHT,
+    /** @stable ICU 63 */
+    U_INPC_VISUAL_ORDER_LEFT,
+    /** @stable ICU 66 */
+    U_INPC_TOP_AND_BOTTOM_AND_LEFT,
+} UIndicPositionalCategory;
+
+/**
+ * Indic Syllabic Category constants.
+ *
+ * @see UCHAR_INDIC_SYLLABIC_CATEGORY
+ * @stable ICU 63
+ */
+typedef enum UIndicSyllabicCategory {
+    /*
+     * Note: UIndicSyllabicCategory constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_INSC_<Unicode Indic_Syllabic_Category value name>
+     */
+
+    /** @stable ICU 63 */
+    U_INSC_OTHER,
+    /** @stable ICU 63 */
+    U_INSC_AVAGRAHA,
+    /** @stable ICU 63 */
+    U_INSC_BINDU,
+    /** @stable ICU 63 */
+    U_INSC_BRAHMI_JOINING_NUMBER,
+    /** @stable ICU 63 */
+    U_INSC_CANTILLATION_MARK,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_DEAD,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_FINAL,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_HEAD_LETTER,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_INITIAL_POSTFIXED,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_KILLER,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_MEDIAL,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_PLACEHOLDER,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_PRECEDING_REPHA,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_PREFIXED,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_SUBJOINED,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_SUCCEEDING_REPHA,
+    /** @stable ICU 63 */
+    U_INSC_CONSONANT_WITH_STACKER,
+    /** @stable ICU 63 */
+    U_INSC_GEMINATION_MARK,
+    /** @stable ICU 63 */
+    U_INSC_INVISIBLE_STACKER,
+    /** @stable ICU 63 */
+    U_INSC_JOINER,
+    /** @stable ICU 63 */
+    U_INSC_MODIFYING_LETTER,
+    /** @stable ICU 63 */
+    U_INSC_NON_JOINER,
+    /** @stable ICU 63 */
+    U_INSC_NUKTA,
+    /** @stable ICU 63 */
+    U_INSC_NUMBER,
+    /** @stable ICU 63 */
+    U_INSC_NUMBER_JOINER,
+    /** @stable ICU 63 */
+    U_INSC_PURE_KILLER,
+    /** @stable ICU 63 */
+    U_INSC_REGISTER_SHIFTER,
+    /** @stable ICU 63 */
+    U_INSC_SYLLABLE_MODIFIER,
+    /** @stable ICU 63 */
+    U_INSC_TONE_LETTER,
+    /** @stable ICU 63 */
+    U_INSC_TONE_MARK,
+    /** @stable ICU 63 */
+    U_INSC_VIRAMA,
+    /** @stable ICU 63 */
+    U_INSC_VISARGA,
+    /** @stable ICU 63 */
+    U_INSC_VOWEL,
+    /** @stable ICU 63 */
+    U_INSC_VOWEL_DEPENDENT,
+    /** @stable ICU 63 */
+    U_INSC_VOWEL_INDEPENDENT,
+} UIndicSyllabicCategory;
+
+/**
+ * Vertical Orientation constants.
+ *
+ * @see UCHAR_VERTICAL_ORIENTATION
+ * @stable ICU 63
+ */
+typedef enum UVerticalOrientation {
+    /*
+     * Note: UVerticalOrientation constants are parsed by preparseucd.py.
+     * It matches lines like
+     *     U_VO_<Unicode Vertical_Orientation value name>
+     */
+
+    /** @stable ICU 63 */
+    U_VO_ROTATED,
+    /** @stable ICU 63 */
+    U_VO_TRANSFORMED_ROTATED,
+    /** @stable ICU 63 */
+    U_VO_TRANSFORMED_UPRIGHT,
+    /** @stable ICU 63 */
+    U_VO_UPRIGHT,
+} UVerticalOrientation;
+
+/**
+ * Check a binary Unicode property for a code point.
+ *
+ * Unicode, especially in version 3.2, defines many more properties than the
+ * original set in UnicodeData.txt.
+ *
+ * The properties APIs are intended to reflect Unicode properties as defined
+ * in the Unicode Character Database (UCD) and Unicode Technical Reports (UTR).
+ * For details about the properties see http://www.unicode.org/ucd/ .
+ * For names of Unicode properties see the UCD file PropertyAliases.txt.
+ *
+ * Important: If ICU is built with UCD files from Unicode versions below 3.2,
+ * then properties marked with "new in Unicode 3.2" are not or not fully available.
+ *
+ * @param c Code point to test.
+ * @param which UProperty selector constant, identifies which binary property to check.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT.
+ * @return true or false according to the binary Unicode property value for c.
+ *         Also false if 'which' is out of bounds or if the Unicode version
+ *         does not have data for the property at all, or not for this code point.
+ *
+ * @see UProperty
+ * @see u_getBinaryPropertySet
+ * @see u_getIntPropertyValue
+ * @see u_getUnicodeVersion
+ * @stable ICU 2.1
+ */
+U_CAPI UBool U_EXPORT2
+u_hasBinaryProperty(UChar32 c, UProperty which);
+
+/**
+ * Returns a frozen USet for a binary property.
+ * The library retains ownership over the returned object.
+ * Sets an error code if the property number is not one for a binary property.
+ *
+ * The returned set contains all code points for which the property is true.
+ *
+ * @param property UCHAR_BINARY_START..UCHAR_BINARY_LIMIT-1
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the property as a set
+ * @see UProperty
+ * @see u_hasBinaryProperty
+ * @see Unicode::fromUSet
+ * @stable ICU 63
+ */
+U_CAPI const USet * U_EXPORT2
+u_getBinaryPropertySet(UProperty property, UErrorCode *pErrorCode);
+
+/**
+ * Check if a code point has the Alphabetic Unicode property.
+ * Same as u_hasBinaryProperty(c, UCHAR_ALPHABETIC).
+ * This is different from u_isalpha!
+ * @param c Code point to test
+ * @return true if the code point has the Alphabetic Unicode property, false otherwise
+ *
+ * @see UCHAR_ALPHABETIC
+ * @see u_isalpha
+ * @see u_hasBinaryProperty
+ * @stable ICU 2.1
+ */
+U_CAPI UBool U_EXPORT2
+u_isUAlphabetic(UChar32 c);
+
+/**
+ * Check if a code point has the Lowercase Unicode property.
+ * Same as u_hasBinaryProperty(c, UCHAR_LOWERCASE).
+ * This is different from u_islower!
+ * @param c Code point to test
+ * @return true if the code point has the Lowercase Unicode property, false otherwise
+ *
+ * @see UCHAR_LOWERCASE
+ * @see u_islower
+ * @see u_hasBinaryProperty
+ * @stable ICU 2.1
+ */
+U_CAPI UBool U_EXPORT2
+u_isULowercase(UChar32 c);
+
+/**
+ * Check if a code point has the Uppercase Unicode property.
+ * Same as u_hasBinaryProperty(c, UCHAR_UPPERCASE).
+ * This is different from u_isupper!
+ * @param c Code point to test
+ * @return true if the code point has the Uppercase Unicode property, false otherwise
+ *
+ * @see UCHAR_UPPERCASE
+ * @see u_isupper
+ * @see u_hasBinaryProperty
+ * @stable ICU 2.1
+ */
+U_CAPI UBool U_EXPORT2
+u_isUUppercase(UChar32 c);
+
+/**
+ * Check if a code point has the White_Space Unicode property.
+ * Same as u_hasBinaryProperty(c, UCHAR_WHITE_SPACE).
+ * This is different from both u_isspace and u_isWhitespace!
+ *
+ * Note: There are several ICU whitespace functions; please see the uchar.h
+ * file documentation for a detailed comparison.
+ *
+ * @param c Code point to test
+ * @return true if the code point has the White_Space Unicode property, false otherwise.
+ *
+ * @see UCHAR_WHITE_SPACE
+ * @see u_isWhitespace
+ * @see u_isspace
+ * @see u_isJavaSpaceChar
+ * @see u_hasBinaryProperty
+ * @stable ICU 2.1
+ */
+U_CAPI UBool U_EXPORT2
+u_isUWhiteSpace(UChar32 c);
+
+/**
+ * Get the property value for an enumerated or integer Unicode property for a code point.
+ * Also returns binary and mask property values.
+ *
+ * Unicode, especially in version 3.2, defines many more properties than the
+ * original set in UnicodeData.txt.
+ *
+ * The properties APIs are intended to reflect Unicode properties as defined
+ * in the Unicode Character Database (UCD) and Unicode Technical Reports (UTR).
+ * For details about the properties see http://www.unicode.org/ .
+ * For names of Unicode properties see the UCD file PropertyAliases.txt.
+ *
+ * Sample usage:
+ * UEastAsianWidth ea=(UEastAsianWidth)u_getIntPropertyValue(c, UCHAR_EAST_ASIAN_WIDTH);
+ * UBool b=(UBool)u_getIntPropertyValue(c, UCHAR_IDEOGRAPHIC);
+ *
+ * @param c Code point to test.
+ * @param which UProperty selector constant, identifies which property to check.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT
+ *        or UCHAR_INT_START<=which<UCHAR_INT_LIMIT
+ *        or UCHAR_MASK_START<=which<UCHAR_MASK_LIMIT.
+ * @return Numeric value that is directly the property value or,
+ *         for enumerated properties, corresponds to the numeric value of the enumerated
+ *         constant of the respective property value enumeration type
+ *         (cast to enum type if necessary).
+ *         Returns 0 or 1 (for false/true) for binary Unicode properties.
+ *         Returns a bit-mask for mask properties.
+ *         Returns 0 if 'which' is out of bounds or if the Unicode version
+ *         does not have data for the property at all, or not for this code point.
+ *
+ * @see UProperty
+ * @see u_hasBinaryProperty
+ * @see u_getIntPropertyMinValue
+ * @see u_getIntPropertyMaxValue
+ * @see u_getIntPropertyMap
+ * @see u_getUnicodeVersion
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyValue(UChar32 c, UProperty which);
+
+/**
+ * Get the minimum value for an enumerated/integer/binary Unicode property.
+ * Can be used together with u_getIntPropertyMaxValue
+ * to allocate arrays of UnicodeSet or similar.
+ *
+ * @param which UProperty selector constant, identifies which binary property to check.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT
+ *        or UCHAR_INT_START<=which<UCHAR_INT_LIMIT.
+ * @return Minimum value returned by u_getIntPropertyValue for a Unicode property.
+ *         0 if the property selector is out of range.
+ *
+ * @see UProperty
+ * @see u_hasBinaryProperty
+ * @see u_getUnicodeVersion
+ * @see u_getIntPropertyMaxValue
+ * @see u_getIntPropertyValue
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyMinValue(UProperty which);
+
+/**
+ * Get the maximum value for an enumerated/integer/binary Unicode property.
+ * Can be used together with u_getIntPropertyMinValue
+ * to allocate arrays of UnicodeSet or similar.
+ *
+ * Examples for min/max values (for Unicode 3.2):
+ *
+ * - UCHAR_BIDI_CLASS:    0/18 (U_LEFT_TO_RIGHT/U_BOUNDARY_NEUTRAL)
+ * - UCHAR_SCRIPT:        0/45 (USCRIPT_COMMON/USCRIPT_TAGBANWA)
+ * - UCHAR_IDEOGRAPHIC:   0/1  (false/true)
+ *
+ * For undefined UProperty constant values, min/max values will be 0/-1.
+ *
+ * @param which UProperty selector constant, identifies which binary property to check.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT
+ *        or UCHAR_INT_START<=which<UCHAR_INT_LIMIT.
+ * @return Maximum value returned by u_getIntPropertyValue for a Unicode property.
+ *         <=0 if the property selector is out of range.
+ *
+ * @see UProperty
+ * @see u_hasBinaryProperty
+ * @see u_getUnicodeVersion
+ * @see u_getIntPropertyMaxValue
+ * @see u_getIntPropertyValue
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyMaxValue(UProperty which);
+
+/**
+ * Returns an immutable UCPMap for an enumerated/catalog/int-valued property.
+ * The library retains ownership over the returned object.
+ * Sets an error code if the property number is not one for an "int property".
+ *
+ * The returned object maps all Unicode code points to their values for that property.
+ * For documentation of the integer values see u_getIntPropertyValue().
+ *
+ * @param property UCHAR_INT_START..UCHAR_INT_LIMIT-1
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the property as a map
+ * @see UProperty
+ * @see u_getIntPropertyValue
+ * @stable ICU 63
+ */
+U_CAPI const UCPMap * U_EXPORT2
+u_getIntPropertyMap(UProperty property, UErrorCode *pErrorCode);
+
+/**
+ * Get the numeric value for a Unicode code point as defined in the
+ * Unicode Character Database.
+ *
+ * A "double" return type is necessary because
+ * some numeric values are fractions, negative, or too large for int32_t.
+ *
+ * For characters without any numeric values in the Unicode Character Database,
+ * this function will return U_NO_NUMERIC_VALUE.
+ * Note: This is different from the Unicode Standard which specifies NaN as the default value.
+ * (NaN is not available on all platforms.)
+ *
+ * Similar to java.lang.Character.getNumericValue(), but u_getNumericValue()
+ * also supports negative values, large values, and fractions,
+ * while Java's getNumericValue() returns values 10..35 for ASCII letters.
+ *
+ * @param c Code point to get the numeric value for.
+ * @return Numeric value of c, or U_NO_NUMERIC_VALUE if none is defined.
+ *
+ * @see U_NO_NUMERIC_VALUE
+ * @stable ICU 2.2
+ */
+U_CAPI double U_EXPORT2
+u_getNumericValue(UChar32 c);
+
+/**
+ * Special value that is returned by u_getNumericValue when
+ * no numeric value is defined for a code point.
+ *
+ * @see u_getNumericValue
+ * @stable ICU 2.2
+ */
+#define U_NO_NUMERIC_VALUE ((double)-123456789.)
+
+/**
+ * Determines whether the specified code point has the general category "Ll"
+ * (lowercase letter).
+ *
+ * Same as java.lang.Character.isLowerCase().
+ *
+ * This misses some characters that are also lowercase but
+ * have a different general category value.
+ * In order to include those, use UCHAR_LOWERCASE.
+ *
+ * In addition to being equivalent to a Java function, this also serves
+ * as a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is an Ll lowercase letter
+ *
+ * @see UCHAR_LOWERCASE
+ * @see u_isupper
+ * @see u_istitle
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_islower(UChar32 c);
+
+/**
+ * Determines whether the specified code point has the general category "Lu"
+ * (uppercase letter).
+ *
+ * Same as java.lang.Character.isUpperCase().
+ *
+ * This misses some characters that are also uppercase but
+ * have a different general category value.
+ * In order to include those, use UCHAR_UPPERCASE.
+ *
+ * In addition to being equivalent to a Java function, this also serves
+ * as a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is an Lu uppercase letter
+ *
+ * @see UCHAR_UPPERCASE
+ * @see u_islower
+ * @see u_istitle
+ * @see u_tolower
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isupper(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a titlecase letter.
+ * True for general category "Lt" (titlecase letter).
+ *
+ * Same as java.lang.Character.isTitleCase().
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is an Lt titlecase letter
+ *
+ * @see u_isupper
+ * @see u_islower
+ * @see u_totitle
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_istitle(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a digit character according to Java.
+ * True for characters with general category "Nd" (decimal digit numbers).
+ * Beginning with Unicode 4, this is the same as
+ * testing for the Numeric_Type of Decimal.
+ *
+ * Same as java.lang.Character.isDigit().
+ *
+ * In addition to being equivalent to a Java function, this also serves
+ * as a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a digit character according to Character.isDigit()
+ *
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isdigit(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a letter character.
+ * True for general categories "L" (letters).
+ *
+ * Same as java.lang.Character.isLetter().
+ *
+ * In addition to being equivalent to a Java function, this also serves
+ * as a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a letter character
+ *
+ * @see u_isdigit
+ * @see u_isalnum
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isalpha(UChar32 c);
+
+/**
+ * Determines whether the specified code point is an alphanumeric character
+ * (letter or digit) according to Java.
+ * True for characters with general categories
+ * "L" (letters) and "Nd" (decimal digit numbers).
+ *
+ * Same as java.lang.Character.isLetterOrDigit().
+ *
+ * In addition to being equivalent to a Java function, this also serves
+ * as a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is an alphanumeric character according to Character.isLetterOrDigit()
+ *
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isalnum(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a hexadecimal digit.
+ * This is equivalent to u_digit(c, 16)>=0.
+ * True for characters with general category "Nd" (decimal digit numbers)
+ * as well as Latin letters a-f and A-F in both ASCII and Fullwidth ASCII.
+ * (That is, for letters with code points
+ * 0041..0046, 0061..0066, FF21..FF26, FF41..FF46.)
+ *
+ * In order to narrow the definition of hexadecimal digits to only ASCII
+ * characters, use (c<=0x7f && u_isxdigit(c)).
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a hexadecimal digit
+ *
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_isxdigit(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a punctuation character.
+ * True for characters with general categories "P" (punctuation).
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a punctuation character
+ *
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_ispunct(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a "graphic" character
+ * (printable, excluding spaces).
+ * true for all characters except those with general categories
+ * "Cc" (control codes), "Cf" (format controls), "Cs" (surrogates),
+ * "Cn" (unassigned), and "Z" (separators).
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a "graphic" character
+ *
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_isgraph(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a "blank" or "horizontal space",
+ * a character that visibly separates words on a line.
+ * The following are equivalent definitions:
+ *
+ * true for Unicode White_Space characters except for "vertical space controls"
+ * where "vertical space controls" are the following characters:
+ * U+000A (LF) U+000B (VT) U+000C (FF) U+000D (CR) U+0085 (NEL) U+2028 (LS) U+2029 (PS)
+ *
+ * same as
+ *
+ * true for U+0009 (TAB) and characters with general category "Zs" (space separators).
+ *
+ * Note: There are several ICU whitespace functions; please see the uchar.h
+ * file documentation for a detailed comparison.
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a "blank"
+ *
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_isblank(UChar32 c);
+
+/**
+ * Determines whether the specified code point is "defined",
+ * which usually means that it is assigned a character.
+ * True for general categories other than "Cn" (other, not assigned),
+ * i.e., true for all code points mentioned in UnicodeData.txt.
+ *
+ * Note that non-character code points (e.g., U+FDD0) are not "defined"
+ * (they are Cn), but surrogate code points are "defined" (Cs).
+ *
+ * Same as java.lang.Character.isDefined().
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is assigned a character
+ *
+ * @see u_isdigit
+ * @see u_isalpha
+ * @see u_isalnum
+ * @see u_isupper
+ * @see u_islower
+ * @see u_istitle
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isdefined(UChar32 c);
+
+/**
+ * Determines if the specified character is a space character or not.
+ *
+ * Note: There are several ICU whitespace functions; please see the uchar.h
+ * file documentation for a detailed comparison.
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c    the character to be tested
+ * @return  true if the character is a space character; false otherwise.
+ *
+ * @see u_isJavaSpaceChar
+ * @see u_isWhitespace
+ * @see u_isUWhiteSpace
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isspace(UChar32 c);
+
+/**
+ * Determine if the specified code point is a space character according to Java.
+ * True for characters with general categories "Z" (separators),
+ * which does not include control codes (e.g., TAB or Line Feed).
+ *
+ * Same as java.lang.Character.isSpaceChar().
+ *
+ * Note: There are several ICU whitespace functions; please see the uchar.h
+ * file documentation for a detailed comparison.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a space character according to Character.isSpaceChar()
+ *
+ * @see u_isspace
+ * @see u_isWhitespace
+ * @see u_isUWhiteSpace
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_isJavaSpaceChar(UChar32 c);
+
+/**
+ * Determines if the specified code point is a whitespace character according to Java/ICU.
+ * A character is considered to be a Java whitespace character if and only
+ * if it satisfies one of the following criteria:
+ *
+ * - It is a Unicode Separator character (categories "Z" = "Zs" or "Zl" or "Zp"), but is not
+ *      also a non-breaking space (U+00A0 NBSP or U+2007 Figure Space or U+202F Narrow NBSP).
+ * - It is U+0009 HORIZONTAL TABULATION.
+ * - It is U+000A LINE FEED.
+ * - It is U+000B VERTICAL TABULATION.
+ * - It is U+000C FORM FEED.
+ * - It is U+000D CARRIAGE RETURN.
+ * - It is U+001C FILE SEPARATOR.
+ * - It is U+001D GROUP SEPARATOR.
+ * - It is U+001E RECORD SEPARATOR.
+ * - It is U+001F UNIT SEPARATOR.
+ *
+ * This API tries to sync with the semantics of Java's
+ * java.lang.Character.isWhitespace(), but it may not return
+ * the exact same results because of the Unicode version
+ * difference.
+ *
+ * Note: Unicode 4.0.1 changed U+200B ZERO WIDTH SPACE from a Space Separator (Zs)
+ * to a Format Control (Cf). Since then, isWhitespace(0x200b) returns false.
+ * See http://www.unicode.org/versions/Unicode4.0.1/
+ *
+ * Note: There are several ICU whitespace functions; please see the uchar.h
+ * file documentation for a detailed comparison.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a whitespace character according to Java/ICU
+ *
+ * @see u_isspace
+ * @see u_isJavaSpaceChar
+ * @see u_isUWhiteSpace
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isWhitespace(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a control character
+ * (as defined by this function).
+ * A control character is one of the following:
+ * - ISO 8-bit control character (U+0000..U+001f and U+007f..U+009f)
+ * - U_CONTROL_CHAR (Cc)
+ * - U_FORMAT_CHAR (Cf)
+ * - U_LINE_SEPARATOR (Zl)
+ * - U_PARAGRAPH_SEPARATOR (Zp)
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a control character
+ *
+ * @see UCHAR_DEFAULT_IGNORABLE_CODE_POINT
+ * @see u_isprint
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_iscntrl(UChar32 c);
+
+/**
+ * Determines whether the specified code point is an ISO control code.
+ * True for U+0000..U+001f and U+007f..U+009f (general category "Cc").
+ *
+ * Same as java.lang.Character.isISOControl().
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is an ISO control code
+ *
+ * @see u_iscntrl
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+u_isISOControl(UChar32 c);
+
+/**
+ * Determines whether the specified code point is a printable character.
+ * True for general categories <em>other</em> than "C" (controls).
+ *
+ * This is a C/POSIX migration function.
+ * See the comments about C/POSIX character classification functions in the
+ * documentation at the top of this header file.
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a printable character
+ *
+ * @see UCHAR_DEFAULT_IGNORABLE_CODE_POINT
+ * @see u_iscntrl
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isprint(UChar32 c);
+
+/**
+ * Non-standard: Determines whether the specified code point is a base character.
+ * True for general categories "L" (letters), "N" (numbers),
+ * "Mc" (spacing combining marks), and "Me" (enclosing marks).
+ *
+ * Note that this is different from the Unicode Standard definition in
+ * chapter 3.6, conformance clause D51 “Base character”,
+ * which defines base characters as the code points with general categories
+ * Letter (L), Number (N), Punctuation (P), Symbol (S), or Space Separator (Zs).
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is a base character according to this function
+ *
+ * @see u_isalpha
+ * @see u_isdigit
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isbase(UChar32 c);
+
+/**
+ * Returns the bidirectional category value for the code point,
+ * which is used in the Unicode bidirectional algorithm
+ * (UAX #9 http://www.unicode.org/reports/tr9/).
+ * Note that some <em>unassigned</em> code points have bidi values
+ * of R or AL because they are in blocks that are reserved
+ * for Right-To-Left scripts.
+ *
+ * Same as java.lang.Character.getDirectionality()
+ *
+ * @param c the code point to be tested
+ * @return the bidirectional category (UCharDirection) value
+ *
+ * @see UCharDirection
+ * @stable ICU 2.0
+ */
+U_CAPI UCharDirection U_EXPORT2
+u_charDirection(UChar32 c);
+
+/**
+ * Determines whether the code point has the Bidi_Mirrored property.
+ * This property is set for characters that are commonly used in
+ * Right-To-Left contexts and need to be displayed with a "mirrored"
+ * glyph.
+ *
+ * Same as java.lang.Character.isMirrored().
+ * Same as UCHAR_BIDI_MIRRORED
+ *
+ * @param c the code point to be tested
+ * @return true if the character has the Bidi_Mirrored property
+ *
+ * @see UCHAR_BIDI_MIRRORED
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isMirrored(UChar32 c);
+
+/**
+ * Maps the specified character to a "mirror-image" character.
+ * For characters with the Bidi_Mirrored property, implementations
+ * sometimes need a "poor man's" mapping to another Unicode
+ * character (code point) such that the default glyph may serve
+ * as the mirror-image of the default glyph of the specified
+ * character. This is useful for text conversion to and from
+ * codepages with visual order, and for displays without glyph
+ * selection capabilities.
+ *
+ * @param c the code point to be mapped
+ * @return another Unicode code point that may serve as a mirror-image
+ *         substitute, or c itself if there is no such mapping or c
+ *         does not have the Bidi_Mirrored property
+ *
+ * @see UCHAR_BIDI_MIRRORED
+ * @see u_isMirrored
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_charMirror(UChar32 c);
+
+/**
+ * Maps the specified character to its paired bracket character.
+ * For Bidi_Paired_Bracket_Type!=None, this is the same as u_charMirror().
+ * Otherwise c itself is returned.
+ * See http://www.unicode.org/reports/tr9/
+ *
+ * @param c the code point to be mapped
+ * @return the paired bracket code point,
+ *         or c itself if there is no such mapping
+ *         (Bidi_Paired_Bracket_Type=None)
+ *
+ * @see UCHAR_BIDI_PAIRED_BRACKET
+ * @see UCHAR_BIDI_PAIRED_BRACKET_TYPE
+ * @see u_charMirror
+ * @stable ICU 52
+ */
+U_CAPI UChar32 U_EXPORT2
+u_getBidiPairedBracket(UChar32 c);
+
+/**
+ * Returns the general category value for the code point.
+ *
+ * Same as java.lang.Character.getType().
+ *
+ * @param c the code point to be tested
+ * @return the general category (UCharCategory) value
+ *
+ * @see UCharCategory
+ * @stable ICU 2.0
+ */
+U_CAPI int8_t U_EXPORT2
+u_charType(UChar32 c);
+
+/**
+ * Get a single-bit bit set for the general category of a character.
+ * This bit set can be compared bitwise with U_GC_SM_MASK, U_GC_L_MASK, etc.
+ * Same as U_MASK(u_charType(c)).
+ *
+ * @param c the code point to be tested
+ * @return a single-bit mask corresponding to the general category (UCharCategory) value
+ *
+ * @see u_charType
+ * @see UCharCategory
+ * @see U_GC_CN_MASK
+ * @stable ICU 2.1
+ */
+#define U_GET_GC_MASK(c) U_MASK(u_charType(c))
+
+/**
+ * Callback from u_enumCharTypes(), is called for each contiguous range
+ * of code points c (where start<=c<limit)
+ * with the same Unicode general category ("character type").
+ *
+ * The callback function can stop the enumeration by returning false.
+ *
+ * @param context an opaque pointer, as passed into utrie_enum()
+ * @param start the first code point in a contiguous range with value
+ * @param limit one past the last code point in a contiguous range with value
+ * @param type the general category for all code points in [start..limit[
+ * @return false to stop the enumeration
+ *
+ * @stable ICU 2.1
+ * @see UCharCategory
+ * @see u_enumCharTypes
+ */
+typedef UBool U_CALLCONV
+UCharEnumTypeRange(const void *context, UChar32 start, UChar32 limit, UCharCategory type);
+
+/**
+ * Enumerate efficiently all code points with their Unicode general categories.
+ *
+ * This is useful for building data structures (e.g., UnicodeSet's),
+ * for enumerating all assigned code points (type!=U_UNASSIGNED), etc.
+ *
+ * For each contiguous range of code points with a given general category ("character type"),
+ * the UCharEnumTypeRange function is called.
+ * Adjacent ranges have different types.
+ * The Unicode Standard guarantees that the numeric value of the type is 0..31.
+ *
+ * @param enumRange a pointer to a function that is called for each contiguous range
+ *                  of code points with the same general category
+ * @param context an opaque pointer that is passed on to the callback function
+ *
+ * @stable ICU 2.1
+ * @see UCharCategory
+ * @see UCharEnumTypeRange
+ */
+U_CAPI void U_EXPORT2
+u_enumCharTypes(UCharEnumTypeRange *enumRange, const void *context);
+
+#if !UCONFIG_NO_NORMALIZATION
+
+/**
+ * Returns the combining class of the code point as specified in UnicodeData.txt.
+ *
+ * @param c the code point of the character
+ * @return the combining class of the character
+ * @stable ICU 2.0
+ */
+U_CAPI uint8_t U_EXPORT2
+u_getCombiningClass(UChar32 c);
+
+#endif
+
+/**
+ * Returns the decimal digit value of a decimal digit character.
+ * Such characters have the general category "Nd" (decimal digit numbers)
+ * and a Numeric_Type of Decimal.
+ *
+ * Unlike ICU releases before 2.6, no digit values are returned for any
+ * Han characters because Han number characters are often used with a special
+ * Chinese-style number format (with characters for powers of 10 in between)
+ * instead of in decimal-positional notation.
+ * Unicode 4 explicitly assigns Han number characters the Numeric_Type
+ * Numeric instead of Decimal.
+ * See Jitterbug 1483 for more details.
+ *
+ * Use u_getIntPropertyValue(c, UCHAR_NUMERIC_TYPE) and u_getNumericValue()
+ * for complete numeric Unicode properties.
+ *
+ * @param c the code point for which to get the decimal digit value
+ * @return the decimal digit value of c,
+ *         or -1 if c is not a decimal digit character
+ *
+ * @see u_getNumericValue
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_charDigitValue(UChar32 c);
+
+/**
+ * Returns the Unicode allocation block that contains the character.
+ *
+ * @param c the code point to be tested
+ * @return the block value (UBlockCode) for c
+ *
+ * @see UBlockCode
+ * @stable ICU 2.0
+ */
+U_CAPI UBlockCode U_EXPORT2
+ublock_getCode(UChar32 c);
+
+/**
+ * Retrieve the name of a Unicode character.
+ * Depending on <code>nameChoice</code>, the character name written
+ * into the buffer is the "modern" name or the name that was defined
+ * in Unicode version 1.0.
+ * The name contains only "invariant" characters
+ * like A-Z, 0-9, space, and '-'.
+ * Unicode 1.0 names are only retrieved if they are different from the modern
+ * names and if the data file contains the data for them. gennames may or may
+ * not be called with a command line option to include 1.0 names in unames.dat.
+ *
+ * @param code The character (code point) for which to get the name.
+ *             It must be <code>0<=code<=0x10ffff</code>.
+ * @param nameChoice Selector for which name to get.
+ * @param buffer Destination address for copying the name.
+ *               The name will always be zero-terminated.
+ *               If there is no name, then the buffer will be set to the empty string.
+ * @param bufferLength <code>==sizeof(buffer)</code>
+ * @param pErrorCode Pointer to a UErrorCode variable;
+ *        check for <code>U_SUCCESS()</code> after <code>u_charName()</code>
+ *        returns.
+ * @return The length of the name, or 0 if there is no name for this character.
+ *         If the bufferLength is less than or equal to the length, then the buffer
+ *         contains the truncated name and the returned length indicates the full
+ *         length of the name.
+ *         The length does not include the zero-termination.
+ *
+ * @see UCharNameChoice
+ * @see u_charFromName
+ * @see u_enumCharNames
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_charName(UChar32 code, UCharNameChoice nameChoice,
+           char *buffer, int32_t bufferLength,
+           UErrorCode *pErrorCode);
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Returns an empty string.
+ * Used to return the ISO 10646 comment for a character.
+ * The Unicode ISO_Comment property is deprecated and has no values.
+ *
+ * @param c The character (code point) for which to get the ISO comment.
+ *             It must be <code>0<=c<=0x10ffff</code>.
+ * @param dest Destination address for copying the comment.
+ *             The comment will be zero-terminated if possible.
+ *             If there is no comment, then the buffer will be set to the empty string.
+ * @param destCapacity <code>==sizeof(dest)</code>
+ * @param pErrorCode Pointer to a UErrorCode variable;
+ *        check for <code>U_SUCCESS()</code> after <code>u_getISOComment()</code>
+ *        returns.
+ * @return 0
+ *
+ * @deprecated ICU 49
+ */
+U_DEPRECATED int32_t U_EXPORT2
+u_getISOComment(UChar32 c,
+                char *dest, int32_t destCapacity,
+                UErrorCode *pErrorCode);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+/**
+ * Find a Unicode character by its name and return its code point value.
+ * The name is matched exactly and completely.
+ * If the name does not correspond to a code point, <i>pErrorCode</i>
+ * is set to <code>U_INVALID_CHAR_FOUND</code>.
+ * A Unicode 1.0 name is matched only if it differs from the modern name.
+ * Unicode names are all uppercase. Extended names are lowercase followed
+ * by an uppercase hexadecimal number, and within angle brackets.
+ *
+ * @param nameChoice Selector for which name to match.
+ * @param name The name to match.
+ * @param pErrorCode Pointer to a UErrorCode variable
+ * @return The Unicode value of the code point with the given name,
+ *         or an undefined value if there is no such code point.
+ *
+ * @see UCharNameChoice
+ * @see u_charName
+ * @see u_enumCharNames
+ * @stable ICU 1.7
+ */
+U_CAPI UChar32 U_EXPORT2
+u_charFromName(UCharNameChoice nameChoice,
+               const char *name,
+               UErrorCode *pErrorCode);
+
+/**
+ * Type of a callback function for u_enumCharNames() that gets called
+ * for each Unicode character with the code point value and
+ * the character name.
+ * If such a function returns false, then the enumeration is stopped.
+ *
+ * @param context The context pointer that was passed to u_enumCharNames().
+ * @param code The Unicode code point for the character with this name.
+ * @param nameChoice Selector for which kind of names is enumerated.
+ * @param name The character's name, zero-terminated.
+ * @param length The length of the name.
+ * @return true if the enumeration should continue, false to stop it.
+ *
+ * @see UCharNameChoice
+ * @see u_enumCharNames
+ * @stable ICU 1.7
+ */
+typedef UBool U_CALLCONV UEnumCharNamesFn(void *context,
+                               UChar32 code,
+                               UCharNameChoice nameChoice,
+                               const char *name,
+                               int32_t length);
+
+/**
+ * Enumerate all assigned Unicode characters between the start and limit
+ * code points (start inclusive, limit exclusive) and call a function
+ * for each, passing the code point value and the character name.
+ * For Unicode 1.0 names, only those are enumerated that differ from the
+ * modern names.
+ *
+ * @param start The first code point in the enumeration range.
+ * @param limit One more than the last code point in the enumeration range
+ *              (the first one after the range).
+ * @param fn The function that is to be called for each character name.
+ * @param context An arbitrary pointer that is passed to the function.
+ * @param nameChoice Selector for which kind of names to enumerate.
+ * @param pErrorCode Pointer to a UErrorCode variable
+ *
+ * @see UCharNameChoice
+ * @see UEnumCharNamesFn
+ * @see u_charName
+ * @see u_charFromName
+ * @stable ICU 1.7
+ */
+U_CAPI void U_EXPORT2
+u_enumCharNames(UChar32 start, UChar32 limit,
+                UEnumCharNamesFn *fn,
+                void *context,
+                UCharNameChoice nameChoice,
+                UErrorCode *pErrorCode);
+
+/**
+ * Return the Unicode name for a given property, as given in the
+ * Unicode database file PropertyAliases.txt.
+ *
+ * In addition, this function maps the property
+ * UCHAR_GENERAL_CATEGORY_MASK to the synthetic names "gcm" /
+ * "General_Category_Mask".  These names are not in
+ * PropertyAliases.txt.
+ *
+ * @param property UProperty selector other than UCHAR_INVALID_CODE.
+ *         If out of range, NULL is returned.
+ *
+ * @param nameChoice selector for which name to get.  If out of range,
+ *         NULL is returned.  All properties have a long name.  Most
+ *         have a short name, but some do not.  Unicode allows for
+ *         additional names; if present these will be returned by
+ *         U_LONG_PROPERTY_NAME + i, where i=1, 2,...
+ *
+ * @return a pointer to the name, or NULL if either the
+ *         property or the nameChoice is out of range.  If a given
+ *         nameChoice returns NULL, then all larger values of
+ *         nameChoice will return NULL, with one exception: if NULL is
+ *         returned for U_SHORT_PROPERTY_NAME, then
+ *         U_LONG_PROPERTY_NAME (and higher) may still return a
+ *         non-NULL value.  The returned pointer is valid until
+ *         u_cleanup() is called.
+ *
+ * @see UProperty
+ * @see UPropertyNameChoice
+ * @stable ICU 2.4
+ */
+U_CAPI const char* U_EXPORT2
+u_getPropertyName(UProperty property,
+                  UPropertyNameChoice nameChoice);
+
+/**
+ * Return the UProperty enum for a given property name, as specified
+ * in the Unicode database file PropertyAliases.txt.  Short, long, and
+ * any other variants are recognized.
+ *
+ * In addition, this function maps the synthetic names "gcm" /
+ * "General_Category_Mask" to the property
+ * UCHAR_GENERAL_CATEGORY_MASK.  These names are not in
+ * PropertyAliases.txt.
+ *
+ * @param alias the property name to be matched.  The name is compared
+ *         using "loose matching" as described in PropertyAliases.txt.
+ *
+ * @return a UProperty enum, or UCHAR_INVALID_CODE if the given name
+ *         does not match any property.
+ *
+ * @see UProperty
+ * @stable ICU 2.4
+ */
+U_CAPI UProperty U_EXPORT2
+u_getPropertyEnum(const char* alias);
+
+/**
+ * Return the Unicode name for a given property value, as given in the
+ * Unicode database file PropertyValueAliases.txt.
+ *
+ * Note: Some of the names in PropertyValueAliases.txt can only be
+ * retrieved using UCHAR_GENERAL_CATEGORY_MASK, not
+ * UCHAR_GENERAL_CATEGORY.  These include: "C" / "Other", "L" /
+ * "Letter", "LC" / "Cased_Letter", "M" / "Mark", "N" / "Number", "P"
+ * / "Punctuation", "S" / "Symbol", and "Z" / "Separator".
+ *
+ * @param property UProperty selector constant.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT
+ *        or UCHAR_INT_START<=which<UCHAR_INT_LIMIT
+ *        or UCHAR_MASK_START<=which<UCHAR_MASK_LIMIT.
+ *        If out of range, NULL is returned.
+ *
+ * @param value selector for a value for the given property.  If out
+ *         of range, NULL is returned.  In general, valid values range
+ *         from 0 up to some maximum.  There are a few exceptions:
+ *         (1.) UCHAR_BLOCK values begin at the non-zero value
+ *         UBLOCK_BASIC_LATIN.  (2.)  UCHAR_CANONICAL_COMBINING_CLASS
+ *         values are not contiguous and range from 0..240.  (3.)
+ *         UCHAR_GENERAL_CATEGORY_MASK values are not values of
+ *         UCharCategory, but rather mask values produced by
+ *         U_GET_GC_MASK().  This allows grouped categories such as
+ *         [:L:] to be represented.  Mask values range
+ *         non-contiguously from 1..U_GC_P_MASK.
+ *
+ * @param nameChoice selector for which name to get.  If out of range,
+ *         NULL is returned.  All values have a long name.  Most have
+ *         a short name, but some do not.  Unicode allows for
+ *         additional names; if present these will be returned by
+ *         U_LONG_PROPERTY_NAME + i, where i=1, 2,...
+
+ * @return a pointer to the name, or NULL if either the
+ *         property or the nameChoice is out of range.  If a given
+ *         nameChoice returns NULL, then all larger values of
+ *         nameChoice will return NULL, with one exception: if NULL is
+ *         returned for U_SHORT_PROPERTY_NAME, then
+ *         U_LONG_PROPERTY_NAME (and higher) may still return a
+ *         non-NULL value.  The returned pointer is valid until
+ *         u_cleanup() is called.
+ *
+ * @see UProperty
+ * @see UPropertyNameChoice
+ * @stable ICU 2.4
+ */
+U_CAPI const char* U_EXPORT2
+u_getPropertyValueName(UProperty property,
+                       int32_t value,
+                       UPropertyNameChoice nameChoice);
+
+/**
+ * Return the property value integer for a given value name, as
+ * specified in the Unicode database file PropertyValueAliases.txt.
+ * Short, long, and any other variants are recognized.
+ *
+ * Note: Some of the names in PropertyValueAliases.txt will only be
+ * recognized with UCHAR_GENERAL_CATEGORY_MASK, not
+ * UCHAR_GENERAL_CATEGORY.  These include: "C" / "Other", "L" /
+ * "Letter", "LC" / "Cased_Letter", "M" / "Mark", "N" / "Number", "P"
+ * / "Punctuation", "S" / "Symbol", and "Z" / "Separator".
+ *
+ * @param property UProperty selector constant.
+ *        Must be UCHAR_BINARY_START<=which<UCHAR_BINARY_LIMIT
+ *        or UCHAR_INT_START<=which<UCHAR_INT_LIMIT
+ *        or UCHAR_MASK_START<=which<UCHAR_MASK_LIMIT.
+ *        If out of range, UCHAR_INVALID_CODE is returned.
+ *
+ * @param alias the value name to be matched.  The name is compared
+ *         using "loose matching" as described in
+ *         PropertyValueAliases.txt.
+ *
+ * @return a value integer or UCHAR_INVALID_CODE if the given name
+ *         does not match any value of the given property, or if the
+ *         property is invalid.  Note: UCHAR_GENERAL_CATEGORY_MASK values
+ *         are not values of UCharCategory, but rather mask values
+ *         produced by U_GET_GC_MASK().  This allows grouped
+ *         categories such as [:L:] to be represented.
+ *
+ * @see UProperty
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+u_getPropertyValueEnum(UProperty property,
+                       const char* alias);
+
+/**
+ * Determines if the specified character is permissible as the
+ * first character in an identifier according to Unicode
+ * (The Unicode Standard, Version 3.0, chapter 5.16 Identifiers).
+ * True for characters with general categories "L" (letters) and "Nl" (letter numbers).
+ *
+ * Same as java.lang.Character.isUnicodeIdentifierStart().
+ * Same as UCHAR_ID_START
+ *
+ * @param c the code point to be tested
+ * @return true if the code point may start an identifier
+ *
+ * @see UCHAR_ID_START
+ * @see u_isalpha
+ * @see u_isIDPart
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isIDStart(UChar32 c);
+
+/**
+ * Determines if the specified character is permissible
+ * in an identifier according to Java.
+ * True for characters with general categories "L" (letters),
+ * "Nl" (letter numbers), "Nd" (decimal digits),
+ * "Mc" and "Mn" (combining marks), "Pc" (connecting punctuation), and
+ * u_isIDIgnorable(c).
+ *
+ * Same as java.lang.Character.isUnicodeIdentifierPart().
+ * Almost the same as Unicode's ID_Continue (UCHAR_ID_CONTINUE)
+ * except that Unicode recommends to ignore Cf which is less than
+ * u_isIDIgnorable(c).
+ *
+ * @param c the code point to be tested
+ * @return true if the code point may occur in an identifier according to Java
+ *
+ * @see UCHAR_ID_CONTINUE
+ * @see u_isIDStart
+ * @see u_isIDIgnorable
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isIDPart(UChar32 c);
+
+/**
+ * Determines if the specified character should be regarded
+ * as an ignorable character in an identifier,
+ * according to Java.
+ * True for characters with general category "Cf" (format controls) as well as
+ * non-whitespace ISO controls
+ * (U+0000..U+0008, U+000E..U+001B, U+007F..U+009F).
+ *
+ * Same as java.lang.Character.isIdentifierIgnorable().
+ *
+ * Note that Unicode just recommends to ignore Cf (format controls).
+ *
+ * @param c the code point to be tested
+ * @return true if the code point is ignorable in identifiers according to Java
+ *
+ * @see UCHAR_DEFAULT_IGNORABLE_CODE_POINT
+ * @see u_isIDStart
+ * @see u_isIDPart
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isIDIgnorable(UChar32 c);
+
+/**
+ * Determines if the specified character is permissible as the
+ * first character in a Java identifier.
+ * In addition to u_isIDStart(c), true for characters with
+ * general categories "Sc" (currency symbols) and "Pc" (connecting punctuation).
+ *
+ * Same as java.lang.Character.isJavaIdentifierStart().
+ *
+ * @param c the code point to be tested
+ * @return true if the code point may start a Java identifier
+ *
+ * @see     u_isJavaIDPart
+ * @see     u_isalpha
+ * @see     u_isIDStart
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isJavaIDStart(UChar32 c);
+
+/**
+ * Determines if the specified character is permissible
+ * in a Java identifier.
+ * In addition to u_isIDPart(c), true for characters with
+ * general category "Sc" (currency symbols).
+ *
+ * Same as java.lang.Character.isJavaIdentifierPart().
+ *
+ * @param c the code point to be tested
+ * @return true if the code point may occur in a Java identifier
+ *
+ * @see     u_isIDIgnorable
+ * @see     u_isJavaIDStart
+ * @see     u_isalpha
+ * @see     u_isdigit
+ * @see     u_isIDPart
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+u_isJavaIDPart(UChar32 c);
+
+/**
+ * The given character is mapped to its lowercase equivalent according to
+ * UnicodeData.txt; if the character has no lowercase equivalent, the character
+ * itself is returned.
+ *
+ * Same as java.lang.Character.toLowerCase().
+ *
+ * This function only returns the simple, single-code point case mapping.
+ * Full case mappings should be used whenever possible because they produce
+ * better results by working on whole strings.
+ * They take into account the string context and the language and can map
+ * to a result string with a different length as appropriate.
+ * Full case mappings are applied by the string case mapping functions,
+ * see ustring.h and the UnicodeString class.
+ * See also the User Guide chapter on C/POSIX migration:
+ * http://icu-project.org/userguide/posix.html#case_mappings
+ *
+ * @param c the code point to be mapped
+ * @return the Simple_Lowercase_Mapping of the code point, if any;
+ *         otherwise the code point itself.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_tolower(UChar32 c);
+
+/**
+ * The given character is mapped to its uppercase equivalent according to UnicodeData.txt;
+ * if the character has no uppercase equivalent, the character itself is
+ * returned.
+ *
+ * Same as java.lang.Character.toUpperCase().
+ *
+ * This function only returns the simple, single-code point case mapping.
+ * Full case mappings should be used whenever possible because they produce
+ * better results by working on whole strings.
+ * They take into account the string context and the language and can map
+ * to a result string with a different length as appropriate.
+ * Full case mappings are applied by the string case mapping functions,
+ * see ustring.h and the UnicodeString class.
+ * See also the User Guide chapter on C/POSIX migration:
+ * http://icu-project.org/userguide/posix.html#case_mappings
+ *
+ * @param c the code point to be mapped
+ * @return the Simple_Uppercase_Mapping of the code point, if any;
+ *         otherwise the code point itself.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_toupper(UChar32 c);
+
+/**
+ * The given character is mapped to its titlecase equivalent
+ * according to UnicodeData.txt;
+ * if none is defined, the character itself is returned.
+ *
+ * Same as java.lang.Character.toTitleCase().
+ *
+ * This function only returns the simple, single-code point case mapping.
+ * Full case mappings should be used whenever possible because they produce
+ * better results by working on whole strings.
+ * They take into account the string context and the language and can map
+ * to a result string with a different length as appropriate.
+ * Full case mappings are applied by the string case mapping functions,
+ * see ustring.h and the UnicodeString class.
+ * See also the User Guide chapter on C/POSIX migration:
+ * http://icu-project.org/userguide/posix.html#case_mappings
+ *
+ * @param c the code point to be mapped
+ * @return the Simple_Titlecase_Mapping of the code point, if any;
+ *         otherwise the code point itself.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_totitle(UChar32 c);
+
+/**
+ * The given character is mapped to its case folding equivalent according to
+ * UnicodeData.txt and CaseFolding.txt;
+ * if the character has no case folding equivalent, the character
+ * itself is returned.
+ *
+ * This function only returns the simple, single-code point case mapping.
+ * Full case mappings should be used whenever possible because they produce
+ * better results by working on whole strings.
+ * They take into account the string context and the language and can map
+ * to a result string with a different length as appropriate.
+ * Full case mappings are applied by the string case mapping functions,
+ * see ustring.h and the UnicodeString class.
+ * See also the User Guide chapter on C/POSIX migration:
+ * http://icu-project.org/userguide/posix.html#case_mappings
+ *
+ * @param c the code point to be mapped
+ * @param options Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ * @return the Simple_Case_Folding of the code point, if any;
+ *         otherwise the code point itself.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_foldCase(UChar32 c, uint32_t options);
+
+/**
+ * Returns the decimal digit value of the code point in the
+ * specified radix.
+ *
+ * If the radix is not in the range <code>2<=radix<=36</code> or if the
+ * value of <code>c</code> is not a valid digit in the specified
+ * radix, <code>-1</code> is returned. A character is a valid digit
+ * if at least one of the following is true:
+ * <ul>
+ * <li>The character has a decimal digit value.
+ *     Such characters have the general category "Nd" (decimal digit numbers)
+ *     and a Numeric_Type of Decimal.
+ *     In this case the value is the character's decimal digit value.</li>
+ * <li>The character is one of the uppercase Latin letters
+ *     <code>'A'</code> through <code>'Z'</code>.
+ *     In this case the value is <code>c-'A'+10</code>.</li>
+ * <li>The character is one of the lowercase Latin letters
+ *     <code>'a'</code> through <code>'z'</code>.
+ *     In this case the value is <code>ch-'a'+10</code>.</li>
+ * <li>Latin letters from both the ASCII range (0061..007A, 0041..005A)
+ *     as well as from the Fullwidth ASCII range (FF41..FF5A, FF21..FF3A)
+ *     are recognized.</li>
+ * </ul>
+ *
+ * Same as java.lang.Character.digit().
+ *
+ * @param   ch      the code point to be tested.
+ * @param   radix   the radix.
+ * @return  the numeric value represented by the character in the
+ *          specified radix,
+ *          or -1 if there is no value or if the value exceeds the radix.
+ *
+ * @see     UCHAR_NUMERIC_TYPE
+ * @see     u_forDigit
+ * @see     u_charDigitValue
+ * @see     u_isdigit
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_digit(UChar32 ch, int8_t radix);
+
+/**
+ * Determines the character representation for a specific digit in
+ * the specified radix. If the value of <code>radix</code> is not a
+ * valid radix, or the value of <code>digit</code> is not a valid
+ * digit in the specified radix, the null character
+ * (<code>U+0000</code>) is returned.
+ * <p>
+ * The <code>radix</code> argument is valid if it is greater than or
+ * equal to 2 and less than or equal to 36.
+ * The <code>digit</code> argument is valid if
+ * <code>0 <= digit < radix</code>.
+ * <p>
+ * If the digit is less than 10, then
+ * <code>'0' + digit</code> is returned. Otherwise, the value
+ * <code>'a' + digit - 10</code> is returned.
+ *
+ * Same as java.lang.Character.forDigit().
+ *
+ * @param   digit   the number to convert to a character.
+ * @param   radix   the radix.
+ * @return  the <code>char</code> representation of the specified digit
+ *          in the specified radix.
+ *
+ * @see     u_digit
+ * @see     u_charDigitValue
+ * @see     u_isdigit
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_forDigit(int32_t digit, int8_t radix);
+
+/**
+ * Get the "age" of the code point.
+ * The "age" is the Unicode version when the code point was first
+ * designated (as a non-character or for Private Use)
+ * or assigned a character.
+ * This can be useful to avoid emitting code points to receiving
+ * processes that do not accept newer characters.
+ * The data is from the UCD file DerivedAge.txt.
+ *
+ * @param c The code point.
+ * @param versionArray The Unicode version number array, to be filled in.
+ *
+ * @stable ICU 2.1
+ */
+U_CAPI void U_EXPORT2
+u_charAge(UChar32 c, UVersionInfo versionArray);
+
+/**
+ * Gets the Unicode version information.
+ * The version array is filled in with the version information
+ * for the Unicode standard that is currently used by ICU.
+ * For example, Unicode version 3.1.1 is represented as an array with
+ * the values { 3, 1, 1, 0 }.
+ *
+ * @param versionArray an output array that will be filled in with
+ *                     the Unicode version number
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+u_getUnicodeVersion(UVersionInfo versionArray);
+
+#if !UCONFIG_NO_NORMALIZATION
+/**
+ * Get the FC_NFKC_Closure property string for a character.
+ * See Unicode Standard Annex #15 for details, search for "FC_NFKC_Closure"
+ * or for "FNC": http://www.unicode.org/reports/tr15/
+ *
+ * @param c The character (code point) for which to get the FC_NFKC_Closure string.
+ *             It must be <code>0<=c<=0x10ffff</code>.
+ * @param dest Destination address for copying the string.
+ *             The string will be zero-terminated if possible.
+ *             If there is no FC_NFKC_Closure string,
+ *             then the buffer will be set to the empty string.
+ * @param destCapacity <code>==sizeof(dest)</code>
+ * @param pErrorCode Pointer to a UErrorCode variable.
+ * @return The length of the string, or 0 if there is no FC_NFKC_Closure string for this character.
+ *         If the destCapacity is less than or equal to the length, then the buffer
+ *         contains the truncated name and the returned length indicates the full
+ *         length of the name.
+ *         The length does not include the zero-termination.
+ *
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_getFC_NFKC_Closure(UChar32 c, UChar *dest, int32_t destCapacity, UErrorCode *pErrorCode);
+
+#endif
+
+
+U_CDECL_END
+
+#endif /*_UCHAR*/
+/*eof*/
diff --git a/thirdparty/icu4c/common/unicode/ucharstrie.h b/thirdparty/icu4c/common/unicode/ucharstrie.h
new file mode 100644
index 0000000000..b6f9e3e075
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucharstrie.h
@@ -0,0 +1,623 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucharstrie.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov14
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UCHARSTRIE_H__
+#define __UCHARSTRIE_H__
+
+/**
+ * \file
+ * \brief C++ API: Trie for mapping Unicode strings (or 16-bit-unit sequences)
+ *                 to integer values.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+#include "unicode/ustringtrie.h"
+
+U_NAMESPACE_BEGIN
+
+class Appendable;
+class UCharsTrieBuilder;
+class UVector32;
+
+/**
+ * Light-weight, non-const reader class for a UCharsTrie.
+ * Traverses a char16_t-serialized data structure with minimal state,
+ * for mapping strings (16-bit-unit sequences) to non-negative integer values.
+ *
+ * This class owns the serialized trie data only if it was constructed by
+ * the builder's build() method.
+ * The public constructor and the copy constructor only alias the data (only copy the pointer).
+ * There is no assignment operator.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API UCharsTrie : public UMemory {
+public:
+    /**
+     * Constructs a UCharsTrie reader instance.
+     *
+     * The trieUChars must contain a copy of a char16_t sequence from the UCharsTrieBuilder,
+     * starting with the first char16_t of that sequence.
+     * The UCharsTrie object will not read more char16_ts than
+     * the UCharsTrieBuilder generated in the corresponding build() call.
+     *
+     * The array is not copied/cloned and must not be modified while
+     * the UCharsTrie object is in use.
+     *
+     * @param trieUChars The char16_t array that contains the serialized trie.
+     * @stable ICU 4.8
+     */
+    UCharsTrie(ConstChar16Ptr trieUChars)
+            : ownedArray_(NULL), uchars_(trieUChars),
+              pos_(uchars_), remainingMatchLength_(-1) {}
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    ~UCharsTrie();
+
+    /**
+     * Copy constructor, copies the other trie reader object and its state,
+     * but not the char16_t array which will be shared. (Shallow copy.)
+     * @param other Another UCharsTrie object.
+     * @stable ICU 4.8
+     */
+    UCharsTrie(const UCharsTrie &other)
+            : ownedArray_(NULL), uchars_(other.uchars_),
+              pos_(other.pos_), remainingMatchLength_(other.remainingMatchLength_) {}
+
+    /**
+     * Resets this trie to its initial state.
+     * @return *this
+     * @stable ICU 4.8
+     */
+    UCharsTrie &reset() {
+        pos_=uchars_;
+        remainingMatchLength_=-1;
+        return *this;
+    }
+
+    /**
+     * Returns the state of this trie as a 64-bit integer.
+     * The state value is never 0.
+     *
+     * @return opaque state value
+     * @see resetToState64
+     * @stable ICU 65
+     */
+    uint64_t getState64() const {
+        return (static_cast<uint64_t>(remainingMatchLength_ + 2) << kState64RemainingShift) |
+            (uint64_t)(pos_ - uchars_);
+    }
+
+    /**
+     * Resets this trie to the saved state.
+     * Unlike resetToState(State), the 64-bit state value
+     * must be from getState64() from the same trie object or
+     * from one initialized the exact same way.
+     * Because of no validation, this method is faster.
+     *
+     * @param state The opaque trie state value from getState64().
+     * @return *this
+     * @see getState64
+     * @see resetToState
+     * @see reset
+     * @stable ICU 65
+     */
+    UCharsTrie &resetToState64(uint64_t state) {
+        remainingMatchLength_ = static_cast<int32_t>(state >> kState64RemainingShift) - 2;
+        pos_ = uchars_ + (state & kState64PosMask);
+        return *this;
+    }
+
+    /**
+     * UCharsTrie state object, for saving a trie's current state
+     * and resetting the trie back to this state later.
+     * @stable ICU 4.8
+     */
+    class State : public UMemory {
+    public:
+        /**
+         * Constructs an empty State.
+         * @stable ICU 4.8
+         */
+        State() { uchars=NULL; }
+    private:
+        friend class UCharsTrie;
+
+        const char16_t *uchars;
+        const char16_t *pos;
+        int32_t remainingMatchLength;
+    };
+
+    /**
+     * Saves the state of this trie.
+     * @param state The State object to hold the trie's state.
+     * @return *this
+     * @see resetToState
+     * @stable ICU 4.8
+     */
+    const UCharsTrie &saveState(State &state) const {
+        state.uchars=uchars_;
+        state.pos=pos_;
+        state.remainingMatchLength=remainingMatchLength_;
+        return *this;
+    }
+
+    /**
+     * Resets this trie to the saved state.
+     * If the state object contains no state, or the state of a different trie,
+     * then this trie remains unchanged.
+     * @param state The State object which holds a saved trie state.
+     * @return *this
+     * @see saveState
+     * @see reset
+     * @stable ICU 4.8
+     */
+    UCharsTrie &resetToState(const State &state) {
+        if(uchars_==state.uchars && uchars_!=NULL) {
+            pos_=state.pos;
+            remainingMatchLength_=state.remainingMatchLength;
+        }
+        return *this;
+    }
+
+    /**
+     * Determines whether the string so far matches, whether it has a value,
+     * and whether another input char16_t can continue a matching string.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult current() const;
+
+    /**
+     * Traverses the trie from the initial state for this input char16_t.
+     * Equivalent to reset().next(uchar).
+     * @param uchar Input char value. Values below 0 and above 0xffff will never match.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    inline UStringTrieResult first(int32_t uchar) {
+        remainingMatchLength_=-1;
+        return nextImpl(uchars_, uchar);
+    }
+
+    /**
+     * Traverses the trie from the initial state for the
+     * one or two UTF-16 code units for this input code point.
+     * Equivalent to reset().nextForCodePoint(cp).
+     * @param cp A Unicode code point 0..0x10ffff.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult firstForCodePoint(UChar32 cp);
+
+    /**
+     * Traverses the trie from the current state for this input char16_t.
+     * @param uchar Input char value. Values below 0 and above 0xffff will never match.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult next(int32_t uchar);
+
+    /**
+     * Traverses the trie from the current state for the
+     * one or two UTF-16 code units for this input code point.
+     * @param cp A Unicode code point 0..0x10ffff.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult nextForCodePoint(UChar32 cp);
+
+    /**
+     * Traverses the trie from the current state for this string.
+     * Equivalent to
+     * \code
+     * Result result=current();
+     * for(each c in s)
+     *   if(!USTRINGTRIE_HAS_NEXT(result)) return USTRINGTRIE_NO_MATCH;
+     *   result=next(c);
+     * return result;
+     * \endcode
+     * @param s A string. Can be NULL if length is 0.
+     * @param length The length of the string. Can be -1 if NUL-terminated.
+     * @return The match/value Result.
+     * @stable ICU 4.8
+     */
+    UStringTrieResult next(ConstChar16Ptr s, int32_t length);
+
+    /**
+     * Returns a matching string's value if called immediately after
+     * current()/first()/next() returned USTRINGTRIE_INTERMEDIATE_VALUE or USTRINGTRIE_FINAL_VALUE.
+     * getValue() can be called multiple times.
+     *
+     * Do not call getValue() after USTRINGTRIE_NO_MATCH or USTRINGTRIE_NO_VALUE!
+     * @return The value for the string so far.
+     * @stable ICU 4.8
+     */
+    inline int32_t getValue() const {
+        const char16_t *pos=pos_;
+        int32_t leadUnit=*pos++;
+        // U_ASSERT(leadUnit>=kMinValueLead);
+        return leadUnit&kValueIsFinal ?
+            readValue(pos, leadUnit&0x7fff) : readNodeValue(pos, leadUnit);
+    }
+
+    /**
+     * Determines whether all strings reachable from the current state
+     * map to the same value.
+     * @param uniqueValue Receives the unique value, if this function returns true.
+     *                    (output-only)
+     * @return true if all strings reachable from the current state
+     *         map to the same value.
+     * @stable ICU 4.8
+     */
+    inline UBool hasUniqueValue(int32_t &uniqueValue) const {
+        const char16_t *pos=pos_;
+        // Skip the rest of a pending linear-match node.
+        return pos!=NULL && findUniqueValue(pos+remainingMatchLength_+1, false, uniqueValue);
+    }
+
+    /**
+     * Finds each char16_t which continues the string from the current state.
+     * That is, each char16_t c for which it would be next(c)!=USTRINGTRIE_NO_MATCH now.
+     * @param out Each next char16_t is appended to this object.
+     * @return the number of char16_ts which continue the string from here
+     * @stable ICU 4.8
+     */
+    int32_t getNextUChars(Appendable &out) const;
+
+    /**
+     * Iterator for all of the (string, value) pairs in a UCharsTrie.
+     * @stable ICU 4.8
+     */
+    class U_COMMON_API Iterator : public UMemory {
+    public:
+        /**
+         * Iterates from the root of a char16_t-serialized UCharsTrie.
+         * @param trieUChars The trie char16_ts.
+         * @param maxStringLength If 0, the iterator returns full strings.
+         *                        Otherwise, the iterator returns strings with this maximum length.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @stable ICU 4.8
+         */
+        Iterator(ConstChar16Ptr trieUChars, int32_t maxStringLength, UErrorCode &errorCode);
+
+        /**
+         * Iterates from the current state of the specified UCharsTrie.
+         * @param trie The trie whose state will be copied for iteration.
+         * @param maxStringLength If 0, the iterator returns full strings.
+         *                        Otherwise, the iterator returns strings with this maximum length.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @stable ICU 4.8
+         */
+        Iterator(const UCharsTrie &trie, int32_t maxStringLength, UErrorCode &errorCode);
+
+        /**
+         * Destructor.
+         * @stable ICU 4.8
+         */
+        ~Iterator();
+
+        /**
+         * Resets this iterator to its initial state.
+         * @return *this
+         * @stable ICU 4.8
+         */
+        Iterator &reset();
+
+        /**
+         * @return true if there are more elements.
+         * @stable ICU 4.8
+         */
+        UBool hasNext() const;
+
+        /**
+         * Finds the next (string, value) pair if there is one.
+         *
+         * If the string is truncated to the maximum length and does not
+         * have a real value, then the value is set to -1.
+         * In this case, this "not a real value" is indistinguishable from
+         * a real value of -1.
+         * @param errorCode Standard ICU error code. Its input value must
+         *                  pass the U_SUCCESS() test, or else the function returns
+         *                  immediately. Check for U_FAILURE() on output or use with
+         *                  function chaining. (See User Guide for details.)
+         * @return true if there is another element.
+         * @stable ICU 4.8
+         */
+        UBool next(UErrorCode &errorCode);
+
+        /**
+         * @return The string for the last successful next().
+         * @stable ICU 4.8
+         */
+        const UnicodeString &getString() const { return str_; }
+        /**
+         * @return The value for the last successful next().
+         * @stable ICU 4.8
+         */
+        int32_t getValue() const { return value_; }
+
+    private:
+        UBool truncateAndStop() {
+            pos_=NULL;
+            value_=-1;  // no real value for str
+            return true;
+        }
+
+        const char16_t *branchNext(const char16_t *pos, int32_t length, UErrorCode &errorCode);
+
+        const char16_t *uchars_;
+        const char16_t *pos_;
+        const char16_t *initialPos_;
+        int32_t remainingMatchLength_;
+        int32_t initialRemainingMatchLength_;
+        UBool skipValue_;  // Skip intermediate value which was already delivered.
+
+        UnicodeString str_;
+        int32_t maxLength_;
+        int32_t value_;
+
+        // The stack stores pairs of integers for backtracking to another
+        // outbound edge of a branch node.
+        // The first integer is an offset from uchars_.
+        // The second integer has the str_.length() from before the node in bits 15..0,
+        // and the remaining branch length in bits 31..16.
+        // (We could store the remaining branch length minus 1 in bits 30..16 and not use the sign bit,
+        // but the code looks more confusing that way.)
+        UVector32 *stack_;
+    };
+
+private:
+    friend class UCharsTrieBuilder;
+
+    /**
+     * Constructs a UCharsTrie reader instance.
+     * Unlike the public constructor which just aliases an array,
+     * this constructor adopts the builder's array.
+     * This constructor is only called by the builder.
+     */
+    UCharsTrie(char16_t *adoptUChars, const char16_t *trieUChars)
+            : ownedArray_(adoptUChars), uchars_(trieUChars),
+              pos_(uchars_), remainingMatchLength_(-1) {}
+
+    // No assignment operator.
+    UCharsTrie &operator=(const UCharsTrie &other);
+
+    inline void stop() {
+        pos_=NULL;
+    }
+
+    // Reads a compact 32-bit integer.
+    // pos is already after the leadUnit, and the lead unit has bit 15 reset.
+    static inline int32_t readValue(const char16_t *pos, int32_t leadUnit) {
+        int32_t value;
+        if(leadUnit<kMinTwoUnitValueLead) {
+            value=leadUnit;
+        } else if(leadUnit<kThreeUnitValueLead) {
+            value=((leadUnit-kMinTwoUnitValueLead)<<16)|*pos;
+        } else {
+            value=(pos[0]<<16)|pos[1];
+        }
+        return value;
+    }
+    static inline const char16_t *skipValue(const char16_t *pos, int32_t leadUnit) {
+        if(leadUnit>=kMinTwoUnitValueLead) {
+            if(leadUnit<kThreeUnitValueLead) {
+                ++pos;
+            } else {
+                pos+=2;
+            }
+        }
+        return pos;
+    }
+    static inline const char16_t *skipValue(const char16_t *pos) {
+        int32_t leadUnit=*pos++;
+        return skipValue(pos, leadUnit&0x7fff);
+    }
+
+    static inline int32_t readNodeValue(const char16_t *pos, int32_t leadUnit) {
+        // U_ASSERT(kMinValueLead<=leadUnit && leadUnit<kValueIsFinal);
+        int32_t value;
+        if(leadUnit<kMinTwoUnitNodeValueLead) {
+            value=(leadUnit>>6)-1;
+        } else if(leadUnit<kThreeUnitNodeValueLead) {
+            value=(((leadUnit&0x7fc0)-kMinTwoUnitNodeValueLead)<<10)|*pos;
+        } else {
+            value=(pos[0]<<16)|pos[1];
+        }
+        return value;
+    }
+    static inline const char16_t *skipNodeValue(const char16_t *pos, int32_t leadUnit) {
+        // U_ASSERT(kMinValueLead<=leadUnit && leadUnit<kValueIsFinal);
+        if(leadUnit>=kMinTwoUnitNodeValueLead) {
+            if(leadUnit<kThreeUnitNodeValueLead) {
+                ++pos;
+            } else {
+                pos+=2;
+            }
+        }
+        return pos;
+    }
+
+    static inline const char16_t *jumpByDelta(const char16_t *pos) {
+        int32_t delta=*pos++;
+        if(delta>=kMinTwoUnitDeltaLead) {
+            if(delta==kThreeUnitDeltaLead) {
+                delta=(pos[0]<<16)|pos[1];
+                pos+=2;
+            } else {
+                delta=((delta-kMinTwoUnitDeltaLead)<<16)|*pos++;
+            }
+        }
+        return pos+delta;
+    }
+
+    static const char16_t *skipDelta(const char16_t *pos) {
+        int32_t delta=*pos++;
+        if(delta>=kMinTwoUnitDeltaLead) {
+            if(delta==kThreeUnitDeltaLead) {
+                pos+=2;
+            } else {
+                ++pos;
+            }
+        }
+        return pos;
+    }
+
+    static inline UStringTrieResult valueResult(int32_t node) {
+        return (UStringTrieResult)(USTRINGTRIE_INTERMEDIATE_VALUE-(node>>15));
+    }
+
+    // Handles a branch node for both next(uchar) and next(string).
+    UStringTrieResult branchNext(const char16_t *pos, int32_t length, int32_t uchar);
+
+    // Requires remainingLength_<0.
+    UStringTrieResult nextImpl(const char16_t *pos, int32_t uchar);
+
+    // Helper functions for hasUniqueValue().
+    // Recursively finds a unique value (or whether there is not a unique one)
+    // from a branch.
+    static const char16_t *findUniqueValueFromBranch(const char16_t *pos, int32_t length,
+                                                  UBool haveUniqueValue, int32_t &uniqueValue);
+    // Recursively finds a unique value (or whether there is not a unique one)
+    // starting from a position on a node lead unit.
+    static UBool findUniqueValue(const char16_t *pos, UBool haveUniqueValue, int32_t &uniqueValue);
+
+    // Helper functions for getNextUChars().
+    // getNextUChars() when pos is on a branch node.
+    static void getNextBranchUChars(const char16_t *pos, int32_t length, Appendable &out);
+
+    // UCharsTrie data structure
+    //
+    // The trie consists of a series of char16_t-serialized nodes for incremental
+    // Unicode string/char16_t sequence matching. (char16_t=16-bit unsigned integer)
+    // The root node is at the beginning of the trie data.
+    //
+    // Types of nodes are distinguished by their node lead unit ranges.
+    // After each node, except a final-value node, another node follows to
+    // encode match values or continue matching further units.
+    //
+    // Node types:
+    //  - Final-value node: Stores a 32-bit integer in a compact, variable-length format.
+    //    The value is for the string/char16_t sequence so far.
+    //  - Match node, optionally with an intermediate value in a different compact format.
+    //    The value, if present, is for the string/char16_t sequence so far.
+    //
+    //  Aside from the value, which uses the node lead unit's high bits:
+    //
+    //  - Linear-match node: Matches a number of units.
+    //  - Branch node: Branches to other nodes according to the current input unit.
+    //    The node unit is the length of the branch (number of units to select from)
+    //    minus 1. It is followed by a sub-node:
+    //    - If the length is at most kMaxBranchLinearSubNodeLength, then
+    //      there are length-1 (key, value) pairs and then one more comparison unit.
+    //      If one of the key units matches, then the value is either a final value for
+    //      the string so far, or a "jump" delta to the next node.
+    //      If the last unit matches, then matching continues with the next node.
+    //      (Values have the same encoding as final-value nodes.)
+    //    - If the length is greater than kMaxBranchLinearSubNodeLength, then
+    //      there is one unit and one "jump" delta.
+    //      If the input unit is less than the sub-node unit, then "jump" by delta to
+    //      the next sub-node which will have a length of length/2.
+    //      (The delta has its own compact encoding.)
+    //      Otherwise, skip the "jump" delta to the next sub-node
+    //      which will have a length of length-length/2.
+
+    // Match-node lead unit values, after masking off intermediate-value bits:
+
+    // 0000..002f: Branch node. If node!=0 then the length is node+1, otherwise
+    // the length is one more than the next unit.
+
+    // For a branch sub-node with at most this many entries, we drop down
+    // to a linear search.
+    static const int32_t kMaxBranchLinearSubNodeLength=5;
+
+    // 0030..003f: Linear-match node, match 1..16 units and continue reading the next node.
+    static const int32_t kMinLinearMatch=0x30;
+    static const int32_t kMaxLinearMatchLength=0x10;
+
+    // Match-node lead unit bits 14..6 for the optional intermediate value.
+    // If these bits are 0, then there is no intermediate value.
+    // Otherwise, see the *NodeValue* constants below.
+    static const int32_t kMinValueLead=kMinLinearMatch+kMaxLinearMatchLength;  // 0x0040
+    static const int32_t kNodeTypeMask=kMinValueLead-1;  // 0x003f
+
+    // A final-value node has bit 15 set.
+    static const int32_t kValueIsFinal=0x8000;
+
+    // Compact value: After testing and masking off bit 15, use the following thresholds.
+    static const int32_t kMaxOneUnitValue=0x3fff;
+
+    static const int32_t kMinTwoUnitValueLead=kMaxOneUnitValue+1;  // 0x4000
+    static const int32_t kThreeUnitValueLead=0x7fff;
+
+    static const int32_t kMaxTwoUnitValue=((kThreeUnitValueLead-kMinTwoUnitValueLead)<<16)-1;  // 0x3ffeffff
+
+    // Compact intermediate-value integer, lead unit shared with a branch or linear-match node.
+    static const int32_t kMaxOneUnitNodeValue=0xff;
+    static const int32_t kMinTwoUnitNodeValueLead=kMinValueLead+((kMaxOneUnitNodeValue+1)<<6);  // 0x4040
+    static const int32_t kThreeUnitNodeValueLead=0x7fc0;
+
+    static const int32_t kMaxTwoUnitNodeValue=
+        ((kThreeUnitNodeValueLead-kMinTwoUnitNodeValueLead)<<10)-1;  // 0xfdffff
+
+    // Compact delta integers.
+    static const int32_t kMaxOneUnitDelta=0xfbff;
+    static const int32_t kMinTwoUnitDeltaLead=kMaxOneUnitDelta+1;  // 0xfc00
+    static const int32_t kThreeUnitDeltaLead=0xffff;
+
+    static const int32_t kMaxTwoUnitDelta=((kThreeUnitDeltaLead-kMinTwoUnitDeltaLead)<<16)-1;  // 0x03feffff
+
+    // For getState64():
+    // The remainingMatchLength_ is -1..14=(kMaxLinearMatchLength=0x10)-2
+    // so we need at least 5 bits for that.
+    // We add 2 to store it as a positive value 1..16=kMaxLinearMatchLength.
+    static constexpr int32_t kState64RemainingShift = 59;
+    static constexpr uint64_t kState64PosMask = (UINT64_C(1) << kState64RemainingShift) - 1;
+
+    char16_t *ownedArray_;
+
+    // Fixed value referencing the UCharsTrie words.
+    const char16_t *uchars_;
+
+    // Iterator variables.
+
+    // Pointer to next trie unit to read. NULL if no more matches.
+    const char16_t *pos_;
+    // Remaining length of a linear-match node, minus 1. Negative if not in such a node.
+    int32_t remainingMatchLength_;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __UCHARSTRIE_H__
diff --git a/thirdparty/icu4c/common/unicode/ucharstriebuilder.h b/thirdparty/icu4c/common/unicode/ucharstriebuilder.h
new file mode 100644
index 0000000000..15657702f9
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucharstriebuilder.h
@@ -0,0 +1,193 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ucharstriebuilder.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010nov14
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UCHARSTRIEBUILDER_H__
+#define __UCHARSTRIEBUILDER_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/stringtriebuilder.h"
+#include "unicode/ucharstrie.h"
+#include "unicode/unistr.h"
+
+/**
+ * \file
+ * \brief C++ API: Builder for icu::UCharsTrie
+ */
+
+U_NAMESPACE_BEGIN
+
+class UCharsTrieElement;
+
+/**
+ * Builder class for UCharsTrie.
+ *
+ * This class is not intended for public subclassing.
+ * @stable ICU 4.8
+ */
+class U_COMMON_API UCharsTrieBuilder : public StringTrieBuilder {
+public:
+    /**
+     * Constructs an empty builder.
+     * @param errorCode Standard ICU error code.
+     * @stable ICU 4.8
+     */
+    UCharsTrieBuilder(UErrorCode &errorCode);
+
+    /**
+     * Destructor.
+     * @stable ICU 4.8
+     */
+    virtual ~UCharsTrieBuilder();
+
+    /**
+     * Adds a (string, value) pair.
+     * The string must be unique.
+     * The string contents will be copied; the builder does not keep
+     * a reference to the input UnicodeString or its buffer.
+     * @param s The input string.
+     * @param value The value associated with this string.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return *this
+     * @stable ICU 4.8
+     */
+    UCharsTrieBuilder &add(const UnicodeString &s, int32_t value, UErrorCode &errorCode);
+
+    /**
+     * Builds a UCharsTrie for the add()ed data.
+     * Once built, no further data can be add()ed until clear() is called.
+     *
+     * A UCharsTrie cannot be empty. At least one (string, value) pair
+     * must have been add()ed.
+     *
+     * This method passes ownership of the builder's internal result array to the new trie object.
+     * Another call to any build() variant will re-serialize the trie.
+     * After clear() has been called, a new array will be used as well.
+     * @param buildOption Build option, see UStringTrieBuildOption.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return A new UCharsTrie for the add()ed data.
+     * @stable ICU 4.8
+     */
+    UCharsTrie *build(UStringTrieBuildOption buildOption, UErrorCode &errorCode);
+
+    /**
+     * Builds a UCharsTrie for the add()ed data and char16_t-serializes it.
+     * Once built, no further data can be add()ed until clear() is called.
+     *
+     * A UCharsTrie cannot be empty. At least one (string, value) pair
+     * must have been add()ed.
+     *
+     * Multiple calls to buildUnicodeString() set the UnicodeStrings to the
+     * builder's same char16_t array, without rebuilding.
+     * If buildUnicodeString() is called after build(), the trie will be
+     * re-serialized into a new array (because build() passes on ownership).
+     * If build() is called after buildUnicodeString(), the trie object returned
+     * by build() will become the owner of the underlying data for the
+     * previously returned UnicodeString.
+     * After clear() has been called, a new array will be used as well.
+     * @param buildOption Build option, see UStringTrieBuildOption.
+     * @param result A UnicodeString which will be set to the char16_t-serialized
+     *               UCharsTrie for the add()ed data.
+     * @param errorCode Standard ICU error code. Its input value must
+     *                  pass the U_SUCCESS() test, or else the function returns
+     *                  immediately. Check for U_FAILURE() on output or use with
+     *                  function chaining. (See User Guide for details.)
+     * @return result
+     * @stable ICU 4.8
+     */
+    UnicodeString &buildUnicodeString(UStringTrieBuildOption buildOption, UnicodeString &result,
+                                      UErrorCode &errorCode);
+
+    /**
+     * Removes all (string, value) pairs.
+     * New data can then be add()ed and a new trie can be built.
+     * @return *this
+     * @stable ICU 4.8
+     */
+    UCharsTrieBuilder &clear() {
+        strings.remove();
+        elementsLength=0;
+        ucharsLength=0;
+        return *this;
+    }
+
+private:
+    UCharsTrieBuilder(const UCharsTrieBuilder &other);  // no copy constructor
+    UCharsTrieBuilder &operator=(const UCharsTrieBuilder &other);  // no assignment operator
+
+    void buildUChars(UStringTrieBuildOption buildOption, UErrorCode &errorCode);
+
+    virtual int32_t getElementStringLength(int32_t i) const;
+    virtual char16_t getElementUnit(int32_t i, int32_t unitIndex) const;
+    virtual int32_t getElementValue(int32_t i) const;
+
+    virtual int32_t getLimitOfLinearMatch(int32_t first, int32_t last, int32_t unitIndex) const;
+
+    virtual int32_t countElementUnits(int32_t start, int32_t limit, int32_t unitIndex) const;
+    virtual int32_t skipElementsBySomeUnits(int32_t i, int32_t unitIndex, int32_t count) const;
+    virtual int32_t indexOfElementWithNextUnit(int32_t i, int32_t unitIndex, char16_t unit) const;
+
+    virtual UBool matchNodesCanHaveValues() const { return true; }
+
+    virtual int32_t getMaxBranchLinearSubNodeLength() const { return UCharsTrie::kMaxBranchLinearSubNodeLength; }
+    virtual int32_t getMinLinearMatch() const { return UCharsTrie::kMinLinearMatch; }
+    virtual int32_t getMaxLinearMatchLength() const { return UCharsTrie::kMaxLinearMatchLength; }
+
+    class UCTLinearMatchNode : public LinearMatchNode {
+    public:
+        UCTLinearMatchNode(const char16_t *units, int32_t len, Node *nextNode);
+        virtual UBool operator==(const Node &other) const;
+        virtual void write(StringTrieBuilder &builder);
+    private:
+        const char16_t *s;
+    };
+
+    virtual Node *createLinearMatchNode(int32_t i, int32_t unitIndex, int32_t length,
+                                        Node *nextNode) const;
+
+    UBool ensureCapacity(int32_t length);
+    virtual int32_t write(int32_t unit);
+    int32_t write(const char16_t *s, int32_t length);
+    virtual int32_t writeElementUnits(int32_t i, int32_t unitIndex, int32_t length);
+    virtual int32_t writeValueAndFinal(int32_t i, UBool isFinal);
+    virtual int32_t writeValueAndType(UBool hasValue, int32_t value, int32_t node);
+    virtual int32_t writeDeltaTo(int32_t jumpTarget);
+
+    UnicodeString strings;
+    UCharsTrieElement *elements;
+    int32_t elementsCapacity;
+    int32_t elementsLength;
+
+    // char16_t serialization of the trie.
+    // Grows from the back: ucharsLength measures from the end of the buffer!
+    char16_t *uchars;
+    int32_t ucharsCapacity;
+    int32_t ucharsLength;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif  // __UCHARSTRIEBUILDER_H__
diff --git a/thirdparty/icu4c/common/unicode/uchriter.h b/thirdparty/icu4c/common/unicode/uchriter.h
new file mode 100644
index 0000000000..f5083561a8
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uchriter.h
@@ -0,0 +1,393 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1998-2005, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef UCHRITER_H
+#define UCHRITER_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/chariter.h"
+
+/**
+ * \file 
+ * \brief C++ API: char16_t Character Iterator
+ */
+ 
+U_NAMESPACE_BEGIN
+
+/**
+ * A concrete subclass of CharacterIterator that iterates over the
+ * characters (code units or code points) in a char16_t array.
+ * It's possible not only to create an
+ * iterator that iterates over an entire char16_t array, but also to
+ * create one that iterates over only a subrange of a char16_t array
+ * (iterators over different subranges of the same char16_t array don't
+ * compare equal).
+ * @see CharacterIterator
+ * @see ForwardCharacterIterator
+ * @stable ICU 2.0
+ */
+class U_COMMON_API UCharCharacterIterator : public CharacterIterator {
+public:
+  /**
+   * Create an iterator over the char16_t array referred to by "textPtr".
+   * The iteration range is 0 to <code>length-1</code>.
+   * text is only aliased, not adopted (the
+   * destructor will not delete it).
+   * @param textPtr The char16_t array to be iterated over
+   * @param length The length of the char16_t array
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator(ConstChar16Ptr textPtr, int32_t length);
+
+  /**
+   * Create an iterator over the char16_t array referred to by "textPtr".
+   * The iteration range is 0 to <code>length-1</code>.
+   * text is only aliased, not adopted (the
+   * destructor will not delete it).
+   * The starting
+   * position is specified by "position". If "position" is outside the valid
+   * iteration range, the behavior of this object is undefined.
+   * @param textPtr The char16_t array to be iteratd over
+   * @param length The length of the char16_t array
+   * @param position The starting position of the iteration
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator(ConstChar16Ptr textPtr, int32_t length,
+                         int32_t position);
+
+  /**
+   * Create an iterator over the char16_t array referred to by "textPtr".
+   * The iteration range is 0 to <code>end-1</code>.
+   * text is only aliased, not adopted (the
+   * destructor will not delete it).
+   * The starting
+   * position is specified by "position". If begin and end do not
+   * form a valid iteration range or "position" is outside the valid
+   * iteration range, the behavior of this object is undefined.
+   * @param textPtr The char16_t array to be iterated over
+   * @param length The length of the char16_t array
+   * @param textBegin  The begin position of the iteration range
+   * @param textEnd    The end position of the iteration range
+   * @param position    The starting position of the iteration
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator(ConstChar16Ptr textPtr, int32_t length,
+                         int32_t textBegin,
+                         int32_t textEnd,
+                         int32_t position);
+
+  /**
+   * Copy constructor.  The new iterator iterates over the same range
+   * of the same string as "that", and its initial position is the
+   * same as "that"'s current position.
+   * @param that The UCharCharacterIterator to be copied
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator(const UCharCharacterIterator&  that);
+
+  /**
+   * Destructor.
+   * @stable ICU 2.0
+   */
+  virtual ~UCharCharacterIterator();
+
+  /**
+   * Assignment operator.  *this is altered to iterate over the sane
+   * range of the same string as "that", and refers to the same
+   * character within that string as "that" does.
+   * @param that The object to be copied
+   * @return the newly created object
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator&
+  operator=(const UCharCharacterIterator&    that);
+
+  /**
+   * Returns true if the iterators iterate over the same range of the
+   * same string and are pointing at the same character.
+   * @param that The ForwardCharacterIterator used to be compared for equality
+   * @return true if the iterators iterate over the same range of the
+   * same string and are pointing at the same character.
+   * @stable ICU 2.0
+   */
+  virtual UBool          operator==(const ForwardCharacterIterator& that) const;
+
+  /**
+   * Generates a hash code for this iterator.
+   * @return the hash code.
+   * @stable ICU 2.0
+   */
+  virtual int32_t         hashCode(void) const;
+
+  /**
+   * Returns a new UCharCharacterIterator referring to the same
+   * character in the same range of the same string as this one.  The
+   * caller must delete the new iterator.
+   * @return the CharacterIterator newly created
+   * @stable ICU 2.0
+   */
+  virtual UCharCharacterIterator* clone() const;
+
+  /**
+   * Sets the iterator to refer to the first code unit in its
+   * iteration range, and returns that code unit.
+   * This can be used to begin an iteration with next().
+   * @return the first code unit in its iteration range.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         first(void);
+
+  /**
+   * Sets the iterator to refer to the first code unit in its
+   * iteration range, returns that code unit, and moves the position
+   * to the second code unit. This is an alternative to setToStart()
+   * for forward iteration with nextPostInc().
+   * @return the first code unit in its iteration range
+   * @stable ICU 2.0
+   */
+  virtual char16_t         firstPostInc(void);
+
+  /**
+   * Sets the iterator to refer to the first code point in its
+   * iteration range, and returns that code unit,
+   * This can be used to begin an iteration with next32().
+   * Note that an iteration with next32PostInc(), beginning with,
+   * e.g., setToStart() or firstPostInc(), is more efficient.
+   * @return the first code point in its iteration range
+   * @stable ICU 2.0
+   */
+  virtual UChar32       first32(void);
+
+  /**
+   * Sets the iterator to refer to the first code point in its
+   * iteration range, returns that code point, and moves the position
+   * to the second code point. This is an alternative to setToStart()
+   * for forward iteration with next32PostInc().
+   * @return the first code point in its iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       first32PostInc(void);
+
+  /**
+   * Sets the iterator to refer to the last code unit in its
+   * iteration range, and returns that code unit.
+   * This can be used to begin an iteration with previous().
+   * @return the last code unit in its iteration range.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         last(void);
+
+  /**
+   * Sets the iterator to refer to the last code point in its
+   * iteration range, and returns that code unit.
+   * This can be used to begin an iteration with previous32().
+   * @return the last code point in its iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       last32(void);
+
+  /**
+   * Sets the iterator to refer to the "position"-th code unit
+   * in the text-storage object the iterator refers to, and
+   * returns that code unit.
+   * @param position the position within the text-storage object
+   * @return the code unit
+   * @stable ICU 2.0
+   */
+  virtual char16_t         setIndex(int32_t position);
+
+  /**
+   * Sets the iterator to refer to the beginning of the code point
+   * that contains the "position"-th code unit
+   * in the text-storage object the iterator refers to, and
+   * returns that code point.
+   * The current position is adjusted to the beginning of the code point
+   * (its first code unit).
+   * @param position the position within the text-storage object
+   * @return the code unit
+   * @stable ICU 2.0
+   */
+  virtual UChar32       setIndex32(int32_t position);
+
+  /**
+   * Returns the code unit the iterator currently refers to.
+   * @return the code unit the iterator currently refers to.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         current(void) const;
+
+  /**
+   * Returns the code point the iterator currently refers to.
+   * @return the code point the iterator currently refers to.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       current32(void) const;
+
+  /**
+   * Advances to the next code unit in the iteration range (toward
+   * endIndex()), and returns that code unit.  If there are no more
+   * code units to return, returns DONE.
+   * @return the next code unit in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         next(void);
+
+  /**
+   * Gets the current code unit for returning and advances to the next code unit
+   * in the iteration range
+   * (toward endIndex()).  If there are
+   * no more code units to return, returns DONE.
+   * @return the current code unit.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         nextPostInc(void);
+
+  /**
+   * Advances to the next code point in the iteration range (toward
+   * endIndex()), and returns that code point.  If there are no more
+   * code points to return, returns DONE.
+   * Note that iteration with "pre-increment" semantics is less
+   * efficient than iteration with "post-increment" semantics
+   * that is provided by next32PostInc().
+   * @return the next code point in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       next32(void);
+
+  /**
+   * Gets the current code point for returning and advances to the next code point
+   * in the iteration range
+   * (toward endIndex()).  If there are
+   * no more code points to return, returns DONE.
+   * @return the current point.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       next32PostInc(void);
+
+  /**
+   * Returns false if there are no more code units or code points
+   * at or after the current position in the iteration range.
+   * This is used with nextPostInc() or next32PostInc() in forward
+   * iteration.
+   * @return false if there are no more code units or code points
+   * at or after the current position in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UBool        hasNext();
+
+  /**
+   * Advances to the previous code unit in the iteration range (toward
+   * startIndex()), and returns that code unit.  If there are no more
+   * code units to return, returns DONE.
+   * @return the previous code unit in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual char16_t         previous(void);
+
+  /**
+   * Advances to the previous code point in the iteration range (toward
+   * startIndex()), and returns that code point.  If there are no more
+   * code points to return, returns DONE.
+   * @return the previous code point in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UChar32       previous32(void);
+
+  /**
+   * Returns false if there are no more code units or code points
+   * before the current position in the iteration range.
+   * This is used with previous() or previous32() in backward
+   * iteration.
+   * @return false if there are no more code units or code points
+   * before the current position in the iteration range.
+   * @stable ICU 2.0
+   */
+  virtual UBool        hasPrevious();
+
+  /**
+   * Moves the current position relative to the start or end of the
+   * iteration range, or relative to the current position itself.
+   * The movement is expressed in numbers of code units forward
+   * or backward by specifying a positive or negative delta.
+   * @param delta the position relative to origin. A positive delta means forward;
+   * a negative delta means backward.
+   * @param origin Origin enumeration {kStart, kCurrent, kEnd}
+   * @return the new position
+   * @stable ICU 2.0
+   */
+  virtual int32_t      move(int32_t delta, EOrigin origin);
+
+  /**
+   * Moves the current position relative to the start or end of the
+   * iteration range, or relative to the current position itself.
+   * The movement is expressed in numbers of code points forward
+   * or backward by specifying a positive or negative delta.
+   * @param delta the position relative to origin. A positive delta means forward;
+   * a negative delta means backward.
+   * @param origin Origin enumeration {kStart, kCurrent, kEnd}
+   * @return the new position
+   * @stable ICU 2.0
+   */
+#ifdef move32
+   // One of the system headers right now is sometimes defining a conflicting macro we don't use
+#undef move32
+#endif
+  virtual int32_t      move32(int32_t delta, EOrigin origin);
+
+  /**
+   * Sets the iterator to iterate over a new range of text
+   * @stable ICU 2.0
+   */
+  void setText(ConstChar16Ptr newText, int32_t newTextLength);
+
+  /**
+   * Copies the char16_t array under iteration into the UnicodeString
+   * referred to by "result".  Even if this iterator iterates across
+   * only a part of this string, the whole string is copied.
+   * @param result Receives a copy of the text under iteration.
+   * @stable ICU 2.0
+   */
+  virtual void            getText(UnicodeString& result);
+
+  /**
+   * Return a class ID for this class (not really public)
+   * @return a class ID for this class
+   * @stable ICU 2.0
+   */
+  static UClassID         U_EXPORT2 getStaticClassID(void);
+
+  /**
+   * Return a class ID for this object (not really public)
+   * @return a class ID for this object.
+   * @stable ICU 2.0
+   */
+  virtual UClassID        getDynamicClassID(void) const;
+
+protected:
+  /**
+   * Protected constructor
+   * @stable ICU 2.0
+   */
+  UCharCharacterIterator();
+  /**
+   * Protected member text
+   * @stable ICU 2.0
+   */
+  const char16_t*            text;
+
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uclean.h b/thirdparty/icu4c/common/unicode/uclean.h
new file mode 100644
index 0000000000..c2d920a16e
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uclean.h
@@ -0,0 +1,262 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2001-2014, International Business Machines
+*                Corporation and others. All Rights Reserved.
+******************************************************************************
+*   file name:  uclean.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001July05
+*   created by: George Rhoten
+*/
+
+#ifndef __UCLEAN_H__
+#define __UCLEAN_H__
+
+#include "unicode/utypes.h"
+/**
+ * \file
+ * \brief C API: Initialize and clean up ICU
+ */
+ 
+/**
+ *  Initialize ICU.
+ *
+ *  Use of this function is optional.  It is OK to simply use ICU
+ *  services and functions without first having initialized
+ *  ICU by calling u_init().
+ *
+ *  u_init() will attempt to load some part of ICU's data, and is
+ *  useful as a test for configuration or installation problems that
+ *  leave the ICU data inaccessible.  A successful invocation of u_init()
+ *  does not, however, guarantee that all ICU data is accessible.
+ *
+ *  Multiple calls to u_init() cause no harm, aside from the small amount
+ *  of time required.
+ *
+ *  In old versions of ICU, u_init() was required in multi-threaded applications
+ *  to ensure the thread safety of ICU.  u_init() is no longer needed for this purpose.
+ *
+ * @param status An ICU UErrorCode parameter. It must not be <code>NULL</code>.
+ *    An Error will be returned if some required part of ICU data can not
+ *    be loaded or initialized.
+ *    The function returns immediately if the input error code indicates a
+ *    failure, as usual.
+ *
+ * @stable ICU 2.6
+ */  
+U_CAPI void U_EXPORT2 
+u_init(UErrorCode *status);
+
+#ifndef U_HIDE_SYSTEM_API
+/**
+ * Clean up the system resources, such as allocated memory or open files,
+ * used in all ICU libraries. This will free/delete all memory owned by the
+ * ICU libraries, and return them to their original load state. All open ICU
+ * items (collators, resource bundles, converters, etc.) must be closed before
+ * calling this function, otherwise ICU may not free its allocated memory
+ * (e.g. close your converters and resource bundles before calling this
+ * function). Generally, this function should be called once just before
+ * an application exits. For applications that dynamically load and unload
+ * the ICU libraries (relatively uncommon), u_cleanup() should be called
+ * just before the library unload.
+ * <p>
+ * u_cleanup() also clears any ICU heap functions, mutex functions or
+ * trace functions that may have been set for the process.  
+ * This has the effect of restoring ICU to its initial condition, before
+ * any of these override functions were installed.  Refer to
+ * u_setMemoryFunctions(), u_setMutexFunctions and 
+ * utrace_setFunctions().  If ICU is to be reinitialized after
+ * calling u_cleanup(), these runtime override functions will need to
+ * be set up again if they are still required.
+ * <p>
+ * u_cleanup() is not thread safe.  All other threads should stop using ICU
+ * before calling this function.
+ * <p>
+ * Any open ICU items will be left in an undefined state by u_cleanup(),
+ * and any subsequent attempt to use such an item will give unpredictable
+ * results.
+ * <p>
+ * After calling u_cleanup(), an application may continue to use ICU by
+ * calling u_init().  An application must invoke u_init() first from one single
+ * thread before allowing other threads call u_init().  All threads existing
+ * at the time of the first thread's call to u_init() must also call
+ * u_init() themselves before continuing with other ICU operations.  
+ * <p>
+ * The use of u_cleanup() just before an application terminates is optional,
+ * but it should be called only once for performance reasons. The primary
+ * benefit is to eliminate reports of memory or resource leaks originating
+ * in ICU code from the results generated by heap analysis tools.
+ * <p>
+ * <strong>Use this function with great care!</strong>
+ * </p>
+ *
+ * @stable ICU 2.0
+ * @system
+ */
+U_CAPI void U_EXPORT2 
+u_cleanup(void);
+
+U_CDECL_BEGIN
+/**
+  *  Pointer type for a user supplied memory allocation function.
+  *  @param context user supplied value, obtained from u_setMemoryFunctions().
+  *  @param size    The number of bytes to be allocated
+  *  @return        Pointer to the newly allocated memory, or NULL if the allocation failed.
+  *  @stable ICU 2.8
+  *  @system
+  */
+typedef void *U_CALLCONV UMemAllocFn(const void *context, size_t size);
+/**
+  *  Pointer type for a user supplied memory re-allocation function.
+  *  @param context user supplied value, obtained from u_setMemoryFunctions().
+  *  @param size    The number of bytes to be allocated
+  *  @return        Pointer to the newly allocated memory, or NULL if the allocation failed.
+  *  @stable ICU 2.8
+  *  @system
+  */
+typedef void *U_CALLCONV UMemReallocFn(const void *context, void *mem, size_t size);
+/**
+  *  Pointer type for a user supplied memory free  function.  Behavior should be
+  *  similar the standard C library free().
+  *  @param context user supplied value, obtained from u_setMemoryFunctions().
+  *  @param mem     Pointer to the memory block to be resized
+  *  @param size    The new size for the block
+  *  @return        Pointer to the resized memory block, or NULL if the resizing failed.
+  *  @stable ICU 2.8
+  *  @system
+  */
+typedef void  U_CALLCONV UMemFreeFn (const void *context, void *mem);
+
+/**
+ *  Set the functions that ICU will use for memory allocation.
+ *  Use of this function is optional; by default (without this function), ICU will
+ *  use the standard C library malloc() and free() functions.
+ *  This function can only be used when ICU is in an initial, unused state, before
+ *  u_init() has been called.
+ *  @param context This pointer value will be saved, and then (later) passed as
+ *                 a parameter to the memory functions each time they
+ *                 are called.
+ *  @param a       Pointer to a user-supplied malloc function.
+ *  @param r       Pointer to a user-supplied realloc function.
+ *  @param f       Pointer to a user-supplied free function.
+ *  @param status  Receives error values.
+ *  @stable ICU 2.8
+ *  @system
+ */  
+U_CAPI void U_EXPORT2 
+u_setMemoryFunctions(const void *context, UMemAllocFn * U_CALLCONV_FPTR a, UMemReallocFn * U_CALLCONV_FPTR r, UMemFreeFn * U_CALLCONV_FPTR f, 
+                    UErrorCode *status);
+
+U_CDECL_END
+
+#ifndef U_HIDE_DEPRECATED_API
+/*********************************************************************************
+ *
+ * Deprecated Functions
+ *
+ *    The following functions for user supplied mutexes are no longer supported.
+ *    Any attempt to use them will return a U_UNSUPPORTED_ERROR.
+ *
+ **********************************************************************************/
+
+/**
+  * An opaque pointer type that represents an ICU mutex.
+  * For user-implemented mutexes, the value will typically point to a
+  *  struct or object that implements the mutex.
+  * @deprecated ICU 52. This type is no longer supported.
+  * @system
+  */
+typedef void *UMTX;
+
+U_CDECL_BEGIN
+/**
+  *  Function Pointer type for a user supplied mutex initialization function.
+  *  The user-supplied function will be called by ICU whenever ICU needs to create a
+  *  new mutex.  The function implementation should create a mutex, and store a pointer
+  *  to something that uniquely identifies the mutex into the UMTX that is supplied
+  *  as a parameter.
+  *  @param context user supplied value, obtained from u_setMutexFunctions().
+  *  @param mutex   Receives a pointer that identifies the new mutex.
+  *                 The mutex init function must set the UMTX to a non-null value.   
+  *                 Subsequent calls by ICU to lock, unlock, or destroy a mutex will 
+  *                 identify the mutex by the UMTX value.
+  *  @param status  Error status.  Report errors back to ICU by setting this variable
+  *                 with an error code.
+  *  @deprecated ICU 52. This function is no longer supported.
+  *  @system
+  */
+typedef void U_CALLCONV UMtxInitFn (const void *context, UMTX  *mutex, UErrorCode* status);
+
+
+/**
+  *  Function Pointer type for a user supplied mutex functions.
+  *  One of the  user-supplied functions with this signature will be called by ICU
+  *  whenever ICU needs to lock, unlock, or destroy a mutex.
+  *  @param context user supplied value, obtained from u_setMutexFunctions().
+  *  @param mutex   specify the mutex on which to operate.
+  *  @deprecated ICU 52. This function is no longer supported.
+  *  @system
+  */
+typedef void U_CALLCONV UMtxFn   (const void *context, UMTX  *mutex);
+U_CDECL_END
+
+/**
+  *  Set the functions that ICU will use for mutex operations
+  *  Use of this function is optional; by default (without this function), ICU will
+  *  directly access system functions for mutex operations
+  *  This function can only be used when ICU is in an initial, unused state, before
+  *  u_init() has been called.
+  *  @param context This pointer value will be saved, and then (later) passed as
+  *                 a parameter to the user-supplied mutex functions each time they
+  *                 are called. 
+  *  @param init    Pointer to a mutex initialization function.  Must be non-null.
+  *  @param destroy Pointer to the mutex destroy function.  Must be non-null.
+  *  @param lock    pointer to the mutex lock function.  Must be non-null.
+  *  @param unlock  Pointer to the mutex unlock function.  Must be non-null.
+  *  @param status  Receives error values.
+  *  @deprecated ICU 52. This function is no longer supported.
+  *  @system
+  */  
+U_DEPRECATED void U_EXPORT2 
+u_setMutexFunctions(const void *context, UMtxInitFn *init, UMtxFn *destroy, UMtxFn *lock, UMtxFn *unlock,
+                    UErrorCode *status);
+
+
+/**
+  *  Pointer type for a user supplied atomic increment or decrement function.
+  *  @param context user supplied value, obtained from u_setAtomicIncDecFunctions().
+  *  @param p   Pointer to a 32 bit int to be incremented or decremented
+  *  @return    The value of the variable after the inc or dec operation.
+  *  @deprecated ICU 52. This function is no longer supported.
+  *  @system
+  */
+typedef int32_t U_CALLCONV UMtxAtomicFn(const void *context, int32_t *p);
+
+/**
+ *  Set the functions that ICU will use for atomic increment and decrement of int32_t values.
+ *  Use of this function is optional; by default (without this function), ICU will
+ *  use its own internal implementation of atomic increment/decrement.
+ *  This function can only be used when ICU is in an initial, unused state, before
+ *  u_init() has been called.
+ *  @param context This pointer value will be saved, and then (later) passed as
+ *                 a parameter to the increment and decrement functions each time they
+ *                 are called.  This function can only be called 
+ *  @param inc     Pointer to a function to do an atomic increment operation.  Must be non-null.
+ *  @param dec     Pointer to a function to do an atomic decrement operation.  Must be non-null.
+ *  @param status  Receives error values.
+ *  @deprecated ICU 52. This function is no longer supported.
+ *  @system
+ */  
+U_DEPRECATED void U_EXPORT2 
+u_setAtomicIncDecFunctions(const void *context, UMtxAtomicFn *inc, UMtxAtomicFn *dec,
+                    UErrorCode *status);
+
+#endif  /* U_HIDE_DEPRECATED_API */
+#endif  /* U_HIDE_SYSTEM_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucnv.h b/thirdparty/icu4c/common/unicode/ucnv.h
new file mode 100644
index 0000000000..58f271cfb5
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucnv.h
@@ -0,0 +1,2045 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+ *  ucnv.h:
+ *  External APIs for the ICU's codeset conversion library
+ *  Bertrand A. Damiba
+ *
+ * Modification History:
+ *
+ *   Date        Name        Description
+ *   04/04/99    helena      Fixed internal header inclusion.
+ *   05/11/00    helena      Added setFallback and usesFallback APIs.
+ *   06/29/2000  helena      Major rewrite of the callback APIs.
+ *   12/07/2000  srl         Update of documentation
+ */
+
+/**
+ * \file
+ * \brief C API: Character conversion
+ *
+ * <h2>Character Conversion C API</h2>
+ *
+ * <p>This API is used to convert codepage or character encoded data to and
+ * from UTF-16. You can open a converter with {@link ucnv_open() }. With that
+ * converter, you can get its properties, set options, convert your data and
+ * close the converter.</p>
+ *
+ * <p>Since many software programs recognize different converter names for
+ * different types of converters, there are other functions in this API to
+ * iterate over the converter aliases. The functions {@link ucnv_getAvailableName() },
+ * {@link ucnv_getAlias() } and {@link ucnv_getStandardName() } are some of the
+ * more frequently used alias functions to get this information.</p>
+ *
+ * <p>When a converter encounters an illegal, irregular, invalid or unmappable character
+ * its default behavior is to use a substitution character to replace the
+ * bad byte sequence. This behavior can be changed by using {@link ucnv_setFromUCallBack() }
+ * or {@link ucnv_setToUCallBack() } on the converter. The header ucnv_err.h defines
+ * many other callback actions that can be used instead of a character substitution.</p>
+ *
+ * <p>More information about this API can be found in our
+ * <a href="http://icu-project.org/userguide/conversion.html">User's
+ * Guide</a>.</p>
+ */
+
+#ifndef UCNV_H
+#define UCNV_H
+
+#include "unicode/ucnv_err.h"
+#include "unicode/uenum.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+#if !defined(USET_DEFINED) && !defined(U_IN_DOXYGEN)
+
+#define USET_DEFINED
+
+/**
+ * USet is the C API type corresponding to C++ class UnicodeSet.
+ * It is forward-declared here to avoid including unicode/uset.h file if related
+ * conversion APIs are not used.
+ *
+ * @see ucnv_getUnicodeSet
+ * @stable ICU 2.4
+ */
+typedef struct USet USet;
+
+#endif
+
+#if !UCONFIG_NO_CONVERSION
+
+U_CDECL_BEGIN
+
+/** Maximum length of a converter name including the terminating NULL @stable ICU 2.0 */
+#define UCNV_MAX_CONVERTER_NAME_LENGTH 60
+/** Maximum length of a converter name including path and terminating NULL @stable ICU 2.0 */
+#define UCNV_MAX_FULL_FILE_NAME_LENGTH (600+UCNV_MAX_CONVERTER_NAME_LENGTH)
+
+/** Shift in for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
+#define  UCNV_SI 0x0F
+/** Shift out for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
+#define  UCNV_SO 0x0E
+
+/**
+ * Enum for specifying basic types of converters
+ * @see ucnv_getType
+ * @stable ICU 2.0
+ */
+typedef enum {
+    /** @stable ICU 2.0 */
+    UCNV_UNSUPPORTED_CONVERTER = -1,
+    /** @stable ICU 2.0 */
+    UCNV_SBCS = 0,
+    /** @stable ICU 2.0 */
+    UCNV_DBCS = 1,
+    /** @stable ICU 2.0 */
+    UCNV_MBCS = 2,
+    /** @stable ICU 2.0 */
+    UCNV_LATIN_1 = 3,
+    /** @stable ICU 2.0 */
+    UCNV_UTF8 = 4,
+    /** @stable ICU 2.0 */
+    UCNV_UTF16_BigEndian = 5,
+    /** @stable ICU 2.0 */
+    UCNV_UTF16_LittleEndian = 6,
+    /** @stable ICU 2.0 */
+    UCNV_UTF32_BigEndian = 7,
+    /** @stable ICU 2.0 */
+    UCNV_UTF32_LittleEndian = 8,
+    /** @stable ICU 2.0 */
+    UCNV_EBCDIC_STATEFUL = 9,
+    /** @stable ICU 2.0 */
+    UCNV_ISO_2022 = 10,
+
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_1 = 11,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_2,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_3,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_4,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_5,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_6,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_8,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_11,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_16,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_17,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_18,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_19,
+    /** @stable ICU 2.0 */
+    UCNV_LMBCS_LAST = UCNV_LMBCS_19,
+    /** @stable ICU 2.0 */
+    UCNV_HZ,
+    /** @stable ICU 2.0 */
+    UCNV_SCSU,
+    /** @stable ICU 2.0 */
+    UCNV_ISCII,
+    /** @stable ICU 2.0 */
+    UCNV_US_ASCII,
+    /** @stable ICU 2.0 */
+    UCNV_UTF7,
+    /** @stable ICU 2.2 */
+    UCNV_BOCU1,
+    /** @stable ICU 2.2 */
+    UCNV_UTF16,
+    /** @stable ICU 2.2 */
+    UCNV_UTF32,
+    /** @stable ICU 2.2 */
+    UCNV_CESU8,
+    /** @stable ICU 2.4 */
+    UCNV_IMAP_MAILBOX,
+    /** @stable ICU 4.8 */
+    UCNV_COMPOUND_TEXT,
+
+    /* Number of converter types for which we have conversion routines. */
+    UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES
+} UConverterType;
+
+/**
+ * Enum for specifying which platform a converter ID refers to.
+ * The use of platform/CCSID is not recommended. See ucnv_openCCSID().
+ *
+ * @see ucnv_getPlatform
+ * @see ucnv_openCCSID
+ * @see ucnv_getCCSID
+ * @stable ICU 2.0
+ */
+typedef enum {
+    UCNV_UNKNOWN = -1,
+    UCNV_IBM = 0
+} UConverterPlatform;
+
+/**
+ * Function pointer for error callback in the codepage to unicode direction.
+ * Called when an error has occurred in conversion to unicode, or on open/close of the callback (see reason).
+ * @param context Pointer to the callback's private data
+ * @param args Information about the conversion in progress
+ * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param reason Defines the reason the callback was invoked
+ * @param pErrorCode    ICU error code in/out parameter.
+ *                      For converter callback functions, set to a conversion error
+ *                      before the call, and the callback may reset it to U_ZERO_ERROR.
+ * @see ucnv_setToUCallBack
+ * @see UConverterToUnicodeArgs
+ * @stable ICU 2.0
+ */
+typedef void (U_EXPORT2 *UConverterToUCallback) (
+                  const void* context,
+                  UConverterToUnicodeArgs *args,
+                  const char *codeUnits,
+                  int32_t length,
+                  UConverterCallbackReason reason,
+                  UErrorCode *pErrorCode);
+
+/**
+ * Function pointer for error callback in the unicode to codepage direction.
+ * Called when an error has occurred in conversion from unicode, or on open/close of the callback (see reason).
+ * @param context Pointer to the callback's private data
+ * @param args Information about the conversion in progress
+ * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
+ * @param reason Defines the reason the callback was invoked
+ * @param pErrorCode    ICU error code in/out parameter.
+ *                      For converter callback functions, set to a conversion error
+ *                      before the call, and the callback may reset it to U_ZERO_ERROR.
+ * @see ucnv_setFromUCallBack
+ * @stable ICU 2.0
+ */
+typedef void (U_EXPORT2 *UConverterFromUCallback) (
+                    const void* context,
+                    UConverterFromUnicodeArgs *args,
+                    const UChar* codeUnits,
+                    int32_t length,
+                    UChar32 codePoint,
+                    UConverterCallbackReason reason,
+                    UErrorCode *pErrorCode);
+
+U_CDECL_END
+
+/**
+ * Character that separates converter names from options and options from each other.
+ * @see ucnv_open
+ * @stable ICU 2.0
+ */
+#define UCNV_OPTION_SEP_CHAR ','
+
+/**
+ * String version of UCNV_OPTION_SEP_CHAR.
+ * @see ucnv_open
+ * @stable ICU 2.0
+ */
+#define UCNV_OPTION_SEP_STRING ","
+
+/**
+ * Character that separates a converter option from its value.
+ * @see ucnv_open
+ * @stable ICU 2.0
+ */
+#define UCNV_VALUE_SEP_CHAR '='
+
+/**
+ * String version of UCNV_VALUE_SEP_CHAR.
+ * @see ucnv_open
+ * @stable ICU 2.0
+ */
+#define UCNV_VALUE_SEP_STRING "="
+
+/**
+ * Converter option for specifying a locale.
+ * For example, ucnv_open("SCSU,locale=ja", &errorCode);
+ * See convrtrs.txt.
+ *
+ * @see ucnv_open
+ * @stable ICU 2.0
+ */
+#define UCNV_LOCALE_OPTION_STRING ",locale="
+
+/**
+ * Converter option for specifying a version selector (0..9) for some converters.
+ * For example,
+ * \code
+ *   ucnv_open("UTF-7,version=1", &errorCode);
+ * \endcode
+ * See convrtrs.txt.
+ *
+ * @see ucnv_open
+ * @stable ICU 2.4
+ */
+#define UCNV_VERSION_OPTION_STRING ",version="
+
+/**
+ * Converter option for EBCDIC SBCS or mixed-SBCS/DBCS (stateful) codepages.
+ * Swaps Unicode mappings for EBCDIC LF and NL codes, as used on
+ * S/390 (z/OS) Unix System Services (Open Edition).
+ * For example, ucnv_open("ibm-1047,swaplfnl", &errorCode);
+ * See convrtrs.txt.
+ *
+ * @see ucnv_open
+ * @stable ICU 2.4
+ */
+#define UCNV_SWAP_LFNL_OPTION_STRING ",swaplfnl"
+
+/**
+ * Do a fuzzy compare of two converter/alias names.
+ * The comparison is case-insensitive, ignores leading zeroes if they are not
+ * followed by further digits, and ignores all but letters and digits.
+ * Thus the strings "UTF-8", "utf_8", "u*T@f08" and "Utf 8" are exactly equivalent.
+ * See section 1.4, Charset Alias Matching in Unicode Technical Standard #22
+ * at http://www.unicode.org/reports/tr22/
+ *
+ * @param name1 a converter name or alias, zero-terminated
+ * @param name2 a converter name or alias, zero-terminated
+ * @return 0 if the names match, or a negative value if the name1
+ * lexically precedes name2, or a positive value if the name1
+ * lexically follows name2.
+ * @stable ICU 2.0
+ */
+U_CAPI int U_EXPORT2
+ucnv_compareNames(const char *name1, const char *name2);
+
+
+/**
+ * Creates a UConverter object with the name of a coded character set specified as a C string.
+ * The actual name will be resolved with the alias file
+ * using a case-insensitive string comparison that ignores
+ * leading zeroes and all non-alphanumeric characters.
+ * E.g., the names "UTF8", "utf-8", "u*T@f08" and "Utf 8" are all equivalent.
+ * (See also ucnv_compareNames().)
+ * If <code>NULL</code> is passed for the converter name, it will create one with the
+ * getDefaultName return value.
+ *
+ * <p>A converter name for ICU 1.5 and above may contain options
+ * like a locale specification to control the specific behavior of
+ * the newly instantiated converter.
+ * The meaning of the options depends on the particular converter.
+ * If an option is not defined for or recognized by a given converter, then it is ignored.</p>
+ *
+ * <p>Options are appended to the converter name string, with a
+ * <code>UCNV_OPTION_SEP_CHAR</code> between the name and the first option and
+ * also between adjacent options.</p>
+ *
+ * <p>If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.</p>
+ *
+ * <p>The conversion behavior and names can vary between platforms. ICU may
+ * convert some characters differently from other platforms. Details on this topic
+ * are in the <a href="http://icu-project.org/userguide/conversion.html">User's
+ * Guide</a>. Aliases starting with a "cp" prefix have no specific meaning
+ * other than its an alias starting with the letters "cp". Please do not
+ * associate any meaning to these aliases.</p>
+ *
+ * \snippet samples/ucnv/convsamp.cpp ucnv_open
+ *
+ * @param converterName Name of the coded character set table.
+ *          This may have options appended to the string.
+ *          IANA alias character set names, IBM CCSIDs starting with "ibm-",
+ *          Windows codepage numbers starting with "windows-" are frequently
+ *          used for this parameter. See ucnv_getAvailableName and
+ *          ucnv_getAlias for a complete list that is available.
+ *          If this parameter is NULL, the default converter will be used.
+ * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
+ * @return the created Unicode converter object, or <TT>NULL</TT> if an error occurred
+ * @see ucnv_openU
+ * @see ucnv_openCCSID
+ * @see ucnv_getAvailableName
+ * @see ucnv_getAlias
+ * @see ucnv_getDefaultName
+ * @see ucnv_close
+ * @see ucnv_compareNames
+ * @stable ICU 2.0
+ */
+U_CAPI UConverter* U_EXPORT2
+ucnv_open(const char *converterName, UErrorCode *err);
+
+
+/**
+ * Creates a Unicode converter with the names specified as unicode string.
+ * The name should be limited to the ASCII-7 alphanumerics range.
+ * The actual name will be resolved with the alias file
+ * using a case-insensitive string comparison that ignores
+ * leading zeroes and all non-alphanumeric characters.
+ * E.g., the names "UTF8", "utf-8", "u*T@f08" and "Utf 8" are all equivalent.
+ * (See also ucnv_compareNames().)
+ * If <TT>NULL</TT> is passed for the converter name, it will create
+ * one with the ucnv_getDefaultName() return value.
+ * If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
+ *
+ * <p>See ucnv_open for the complete details</p>
+ * @param name Name of the UConverter table in a zero terminated
+ *        Unicode string
+ * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR,
+ *        U_FILE_ACCESS_ERROR</TT>
+ * @return the created Unicode converter object, or <TT>NULL</TT> if an
+ *        error occurred
+ * @see ucnv_open
+ * @see ucnv_openCCSID
+ * @see ucnv_close
+ * @see ucnv_compareNames
+ * @stable ICU 2.0
+ */
+U_CAPI UConverter* U_EXPORT2
+ucnv_openU(const UChar *name,
+           UErrorCode *err);
+
+/**
+ * Creates a UConverter object from a CCSID number and platform pair.
+ * Note that the usefulness of this function is limited to platforms with numeric
+ * encoding IDs. Only IBM and Microsoft platforms use numeric (16-bit) identifiers for
+ * encodings.
+ *
+ * In addition, IBM CCSIDs and Unicode conversion tables are not 1:1 related.
+ * For many IBM CCSIDs there are multiple (up to six) Unicode conversion tables, and
+ * for some Unicode conversion tables there are multiple CCSIDs.
+ * Some "alternate" Unicode conversion tables are provided by the
+ * IBM CDRA conversion table registry.
+ * The most prominent example of a systematic modification of conversion tables that is
+ * not provided in the form of conversion table files in the repository is
+ * that S/390 Unix System Services swaps the codes for Line Feed and New Line in all
+ * EBCDIC codepages, which requires such a swap in the Unicode conversion tables as well.
+ *
+ * Only IBM default conversion tables are accessible with ucnv_openCCSID().
+ * ucnv_getCCSID() will return the same CCSID for all conversion tables that are associated
+ * with that CCSID.
+ *
+ * Currently, the only "platform" supported in the ICU converter API is UCNV_IBM.
+ *
+ * In summary, the use of CCSIDs and the associated API functions is not recommended.
+ *
+ * In order to open a converter with the default IBM CDRA Unicode conversion table,
+ * you can use this function or use the prefix "ibm-":
+ * \code
+ *     char name[20];
+ *     sprintf(name, "ibm-%hu", ccsid);
+ *     cnv=ucnv_open(name, &errorCode);
+ * \endcode
+ *
+ * In order to open a converter with the IBM S/390 Unix System Services variant
+ * of a Unicode/EBCDIC conversion table,
+ * you can use the prefix "ibm-" together with the option string UCNV_SWAP_LFNL_OPTION_STRING:
+ * \code
+ *     char name[20];
+ *     sprintf(name, "ibm-%hu" UCNV_SWAP_LFNL_OPTION_STRING, ccsid);
+ *     cnv=ucnv_open(name, &errorCode);
+ * \endcode
+ *
+ * In order to open a converter from a Microsoft codepage number, use the prefix "cp":
+ * \code
+ *     char name[20];
+ *     sprintf(name, "cp%hu", codepageID);
+ *     cnv=ucnv_open(name, &errorCode);
+ * \endcode
+ *
+ * If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
+ *
+ * @param codepage codepage number to create
+ * @param platform the platform in which the codepage number exists
+ * @param err error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
+ * @return the created Unicode converter object, or <TT>NULL</TT> if an error
+ *   occurred.
+ * @see ucnv_open
+ * @see ucnv_openU
+ * @see ucnv_close
+ * @see ucnv_getCCSID
+ * @see ucnv_getPlatform
+ * @see UConverterPlatform
+ * @stable ICU 2.0
+ */
+U_CAPI UConverter* U_EXPORT2
+ucnv_openCCSID(int32_t codepage,
+               UConverterPlatform platform,
+               UErrorCode * err);
+
+/**
+ * <p>Creates a UConverter object specified from a packageName and a converterName.</p>
+ *
+ * <p>The packageName and converterName must point to an ICU udata object, as defined by
+ *   <code> udata_open( packageName, "cnv", converterName, err) </code> or equivalent.
+ * Typically, packageName will refer to a (.dat) file, or to a package registered with
+ * udata_setAppData(). Using a full file or directory pathname for packageName is deprecated.</p>
+ *
+ * <p>The name will NOT be looked up in the alias mechanism, nor will the converter be
+ * stored in the converter cache or the alias table. The only way to open further converters
+ * is call this function multiple times, or use the ucnv_safeClone() function to clone a
+ * 'primary' converter.</p>
+ *
+ * <p>A future version of ICU may add alias table lookups and/or caching
+ * to this function.</p>
+ *
+ * <p>Example Use:
+ *      <code>cnv = ucnv_openPackage("myapp", "myconverter", &err);</code>
+ * </p>
+ *
+ * @param packageName name of the package (equivalent to 'path' in udata_open() call)
+ * @param converterName name of the data item to be used, without suffix.
+ * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
+ * @return the created Unicode converter object, or <TT>NULL</TT> if an error occurred
+ * @see udata_open
+ * @see ucnv_open
+ * @see ucnv_safeClone
+ * @see ucnv_close
+ * @stable ICU 2.2
+ */
+U_CAPI UConverter* U_EXPORT2
+ucnv_openPackage(const char *packageName, const char *converterName, UErrorCode *err);
+
+/**
+ * Thread safe converter cloning operation.
+ * For most efficient operation, pass in a stackBuffer (and a *pBufferSize)
+ * with at least U_CNV_SAFECLONE_BUFFERSIZE bytes of space.
+ * If the buffer size is sufficient, then the clone will use the stack buffer;
+ * otherwise, it will be allocated, and *pBufferSize will indicate
+ * the actual size. (This should not occur with U_CNV_SAFECLONE_BUFFERSIZE.)
+ *
+ * You must ucnv_close() the clone in any case.
+ *
+ * If *pBufferSize==0, (regardless of whether stackBuffer==NULL or not)
+ * then *pBufferSize will be changed to a sufficient size
+ * for cloning this converter,
+ * without actually cloning the converter ("pure pre-flighting").
+ *
+ * If *pBufferSize is greater than zero but not large enough for a stack-based
+ * clone, then the converter is cloned using newly allocated memory
+ * and *pBufferSize is changed to the necessary size.
+ *
+ * If the converter clone fits into the stack buffer but the stack buffer is not
+ * sufficiently aligned for the clone, then the clone will use an
+ * adjusted pointer and use an accordingly smaller buffer size.
+ *
+ * @param cnv converter to be cloned
+ * @param stackBuffer <em>Deprecated functionality as of ICU 52, use NULL.</em><br>
+ *  user allocated space for the new clone. If NULL new memory will be allocated.
+ *  If buffer is not large enough, new memory will be allocated.
+ *  Clients can use the U_CNV_SAFECLONE_BUFFERSIZE. This will probably be enough to avoid memory allocations.
+ * @param pBufferSize <em>Deprecated functionality as of ICU 52, use NULL or 1.</em><br>
+ *  pointer to size of allocated space.
+ * @param status to indicate whether the operation went on smoothly or there were errors
+ *  An informational status value, U_SAFECLONE_ALLOCATED_WARNING,
+ *  is used if any allocations were necessary.
+ *  However, it is better to check if *pBufferSize grew for checking for
+ *  allocations because warning codes can be overridden by subsequent
+ *  function calls.
+ * @return pointer to the new clone
+ * @stable ICU 2.0
+ */
+U_CAPI UConverter * U_EXPORT2
+ucnv_safeClone(const UConverter *cnv,
+               void             *stackBuffer,
+               int32_t          *pBufferSize,
+               UErrorCode       *status);
+
+#ifndef U_HIDE_DEPRECATED_API
+
+/**
+ * \def U_CNV_SAFECLONE_BUFFERSIZE
+ * Definition of a buffer size that is designed to be large enough for
+ * converters to be cloned with ucnv_safeClone().
+ * @deprecated ICU 52. Do not rely on ucnv_safeClone() cloning into any provided buffer.
+ */
+#define U_CNV_SAFECLONE_BUFFERSIZE  1024
+
+#endif /* U_HIDE_DEPRECATED_API */
+
+/**
+ * Deletes the unicode converter and releases resources associated
+ * with just this instance.
+ * Does not free up shared converter tables.
+ *
+ * @param converter the converter object to be deleted
+ * @see ucnv_open
+ * @see ucnv_openU
+ * @see ucnv_openCCSID
+ * @stable ICU 2.0
+ */
+U_CAPI void  U_EXPORT2
+ucnv_close(UConverter * converter);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUConverterPointer
+ * "Smart pointer" class, closes a UConverter via ucnv_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUConverterPointer, UConverter, ucnv_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Fills in the output parameter, subChars, with the substitution characters
+ * as multiple bytes.
+ * If ucnv_setSubstString() set a Unicode string because the converter is
+ * stateful, then subChars will be an empty string.
+ *
+ * @param converter the Unicode converter
+ * @param subChars the substitution characters
+ * @param len on input the capacity of subChars, on output the number
+ * of bytes copied to it
+ * @param  err the outgoing error status code.
+ * If the substitution character array is too small, an
+ * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
+ * @see ucnv_setSubstString
+ * @see ucnv_setSubstChars
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getSubstChars(const UConverter *converter,
+                   char *subChars,
+                   int8_t *len,
+                   UErrorCode *err);
+
+/**
+ * Sets the substitution chars when converting from unicode to a codepage. The
+ * substitution is specified as a string of 1-4 bytes, and may contain
+ * <TT>NULL</TT> bytes.
+ * The subChars must represent a single character. The caller needs to know the
+ * byte sequence of a valid character in the converter's charset.
+ * For some converters, for example some ISO 2022 variants, only single-byte
+ * substitution characters may be supported.
+ * The newer ucnv_setSubstString() function relaxes these limitations.
+ *
+ * @param converter the Unicode converter
+ * @param subChars the substitution character byte sequence we want set
+ * @param len the number of bytes in subChars
+ * @param err the error status code.  <TT>U_INDEX_OUTOFBOUNDS_ERROR </TT> if
+ * len is bigger than the maximum number of bytes allowed in subchars
+ * @see ucnv_setSubstString
+ * @see ucnv_getSubstChars
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_setSubstChars(UConverter *converter,
+                   const char *subChars,
+                   int8_t len,
+                   UErrorCode *err);
+
+/**
+ * Set a substitution string for converting from Unicode to a charset.
+ * The caller need not know the charset byte sequence for each charset.
+ *
+ * Unlike ucnv_setSubstChars() which is designed to set a charset byte sequence
+ * for a single character, this function takes a Unicode string with
+ * zero, one or more characters, and immediately verifies that the string can be
+ * converted to the charset.
+ * If not, or if the result is too long (more than 32 bytes as of ICU 3.6),
+ * then the function returns with an error accordingly.
+ *
+ * Also unlike ucnv_setSubstChars(), this function works for stateful charsets
+ * by converting on the fly at the point of substitution rather than setting
+ * a fixed byte sequence.
+ *
+ * @param cnv The UConverter object.
+ * @param s The Unicode string.
+ * @param length The number of UChars in s, or -1 for a NUL-terminated string.
+ * @param err Pointer to a standard ICU error code. Its input value must
+ *            pass the U_SUCCESS() test, or else the function returns
+ *            immediately. Check for U_FAILURE() on output or use with
+ *            function chaining. (See User Guide for details.)
+ *
+ * @see ucnv_setSubstChars
+ * @see ucnv_getSubstChars
+ * @stable ICU 3.6
+ */
+U_CAPI void U_EXPORT2
+ucnv_setSubstString(UConverter *cnv,
+                    const UChar *s,
+                    int32_t length,
+                    UErrorCode *err);
+
+/**
+ * Fills in the output parameter, errBytes, with the error characters from the
+ * last failing conversion.
+ *
+ * @param converter the Unicode converter
+ * @param errBytes the codepage bytes which were in error
+ * @param len on input the capacity of errBytes, on output the number of
+ *  bytes which were copied to it
+ * @param err the error status code.
+ * If the substitution character array is too small, an
+ * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getInvalidChars(const UConverter *converter,
+                     char *errBytes,
+                     int8_t *len,
+                     UErrorCode *err);
+
+/**
+ * Fills in the output parameter, errChars, with the error characters from the
+ * last failing conversion.
+ *
+ * @param converter the Unicode converter
+ * @param errUChars the UChars which were in error
+ * @param len on input the capacity of errUChars, on output the number of
+ *  UChars which were copied to it
+ * @param err the error status code.
+ * If the substitution character array is too small, an
+ * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getInvalidUChars(const UConverter *converter,
+                      UChar *errUChars,
+                      int8_t *len,
+                      UErrorCode *err);
+
+/**
+ * Resets the state of a converter to the default state. This is used
+ * in the case of an error, to restart a conversion from a known default state.
+ * It will also empty the internal output buffers.
+ * @param converter the Unicode converter
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_reset(UConverter *converter);
+
+/**
+ * Resets the to-Unicode part of a converter state to the default state.
+ * This is used in the case of an error to restart a conversion to
+ * Unicode to a known default state. It will also empty the internal
+ * output buffers used for the conversion to Unicode codepoints.
+ * @param converter the Unicode converter
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_resetToUnicode(UConverter *converter);
+
+/**
+ * Resets the from-Unicode part of a converter state to the default state.
+ * This is used in the case of an error to restart a conversion from
+ * Unicode to a known default state. It will also empty the internal output
+ * buffers used for the conversion from Unicode codepoints.
+ * @param converter the Unicode converter
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_resetFromUnicode(UConverter *converter);
+
+/**
+ * Returns the maximum number of bytes that are output per UChar in conversion
+ * from Unicode using this converter.
+ * The returned number can be used with UCNV_GET_MAX_BYTES_FOR_STRING
+ * to calculate the size of a target buffer for conversion from Unicode.
+ *
+ * Note: Before ICU 2.8, this function did not return reliable numbers for
+ * some stateful converters (EBCDIC_STATEFUL, ISO-2022) and LMBCS.
+ *
+ * This number may not be the same as the maximum number of bytes per
+ * "conversion unit". In other words, it may not be the intuitively expected
+ * number of bytes per character that would be published for a charset,
+ * and may not fulfill any other purpose than the allocation of an output
+ * buffer of guaranteed sufficient size for a given input length and converter.
+ *
+ * Examples for special cases that are taken into account:
+ * - Supplementary code points may convert to more bytes than BMP code points.
+ *   This function returns bytes per UChar (UTF-16 code unit), not per
+ *   Unicode code point, for efficient buffer allocation.
+ * - State-shifting output (SI/SO, escapes, etc.) from stateful converters.
+ * - When m input UChars are converted to n output bytes, then the maximum m/n
+ *   is taken into account.
+ *
+ * The number returned here does not take into account
+ * (see UCNV_GET_MAX_BYTES_FOR_STRING):
+ * - callbacks which output more than one charset character sequence per call,
+ *   like escape callbacks
+ * - initial and final non-character bytes that are output by some converters
+ *   (automatic BOMs, initial escape sequence, final SI, etc.)
+ *
+ * Examples for returned values:
+ * - SBCS charsets: 1
+ * - Shift-JIS: 2
+ * - UTF-16: 2 (2 per BMP, 4 per surrogate _pair_, BOM not counted)
+ * - UTF-8: 3 (3 per BMP, 4 per surrogate _pair_)
+ * - EBCDIC_STATEFUL (EBCDIC mixed SBCS/DBCS): 3 (SO + DBCS)
+ * - ISO-2022: 3 (always outputs UTF-8)
+ * - ISO-2022-JP: 6 (4-byte escape sequences + DBCS)
+ * - ISO-2022-CN: 8 (4-byte designator sequences + 2-byte SS2/SS3 + DBCS)
+ *
+ * @param converter The Unicode converter.
+ * @return The maximum number of bytes per UChar (16 bit code unit)
+ *    that are output by ucnv_fromUnicode(),
+ *    to be used together with UCNV_GET_MAX_BYTES_FOR_STRING
+ *    for buffer allocation.
+ *
+ * @see UCNV_GET_MAX_BYTES_FOR_STRING
+ * @see ucnv_getMinCharSize
+ * @stable ICU 2.0
+ */
+U_CAPI int8_t U_EXPORT2
+ucnv_getMaxCharSize(const UConverter *converter);
+
+/**
+ * Calculates the size of a buffer for conversion from Unicode to a charset.
+ * The calculated size is guaranteed to be sufficient for this conversion.
+ *
+ * It takes into account initial and final non-character bytes that are output
+ * by some converters.
+ * It does not take into account callbacks which output more than one charset
+ * character sequence per call, like escape callbacks.
+ * The default (substitution) callback only outputs one charset character sequence.
+ *
+ * @param length Number of UChars to be converted.
+ * @param maxCharSize Return value from ucnv_getMaxCharSize() for the converter
+ *                    that will be used.
+ * @return Size of a buffer that will be large enough to hold the output bytes of
+ *         converting length UChars with the converter that returned the maxCharSize.
+ *
+ * @see ucnv_getMaxCharSize
+ * @stable ICU 2.8
+ */
+#define UCNV_GET_MAX_BYTES_FOR_STRING(length, maxCharSize) \
+     (((int32_t)(length)+10)*(int32_t)(maxCharSize))
+
+/**
+ * Returns the minimum byte length (per codepoint) for characters in this codepage.
+ * This is usually either 1 or 2.
+ * @param converter the Unicode converter
+ * @return the minimum number of bytes per codepoint allowed by this particular converter
+ * @see ucnv_getMaxCharSize
+ * @stable ICU 2.0
+ */
+U_CAPI int8_t U_EXPORT2
+ucnv_getMinCharSize(const UConverter *converter);
+
+/**
+ * Returns the display name of the converter passed in based on the Locale
+ * passed in. If the locale contains no display name, the internal ASCII
+ * name will be filled in.
+ *
+ * @param converter the Unicode converter.
+ * @param displayLocale is the specific Locale we want to localized for
+ * @param displayName user provided buffer to be filled in
+ * @param displayNameCapacity size of displayName Buffer
+ * @param err error status code
+ * @return displayNameLength number of UChar needed in displayName
+ * @see ucnv_getName
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_getDisplayName(const UConverter *converter,
+                    const char *displayLocale,
+                    UChar *displayName,
+                    int32_t displayNameCapacity,
+                    UErrorCode *err);
+
+/**
+ * Gets the internal, canonical name of the converter (zero-terminated).
+ * The lifetime of the returned string will be that of the converter
+ * passed to this function.
+ * @param converter the Unicode converter
+ * @param err UErrorCode status
+ * @return the internal name of the converter
+ * @see ucnv_getDisplayName
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getName(const UConverter *converter, UErrorCode *err);
+
+/**
+ * Gets a codepage number associated with the converter. This is not guaranteed
+ * to be the one used to create the converter. Some converters do not represent
+ * platform registered codepages and return zero for the codepage number.
+ * The error code fill-in parameter indicates if the codepage number
+ * is available.
+ * Does not check if the converter is <TT>NULL</TT> or if converter's data
+ * table is <TT>NULL</TT>.
+ *
+ * Important: The use of CCSIDs is not recommended because it is limited
+ * to only two platforms in principle and only one (UCNV_IBM) in the current
+ * ICU converter API.
+ * Also, CCSIDs are insufficient to identify IBM Unicode conversion tables precisely.
+ * For more details see ucnv_openCCSID().
+ *
+ * @param converter the Unicode converter
+ * @param err the error status code.
+ * @return If any error occurs, -1 will be returned otherwise, the codepage number
+ * will be returned
+ * @see ucnv_openCCSID
+ * @see ucnv_getPlatform
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_getCCSID(const UConverter *converter,
+              UErrorCode *err);
+
+/**
+ * Gets a codepage platform associated with the converter. Currently,
+ * only <TT>UCNV_IBM</TT> will be returned.
+ * Does not test if the converter is <TT>NULL</TT> or if converter's data
+ * table is <TT>NULL</TT>.
+ * @param converter the Unicode converter
+ * @param err the error status code.
+ * @return The codepage platform
+ * @stable ICU 2.0
+ */
+U_CAPI UConverterPlatform U_EXPORT2
+ucnv_getPlatform(const UConverter *converter,
+                 UErrorCode *err);
+
+/**
+ * Gets the type of the converter
+ * e.g. SBCS, MBCS, DBCS, UTF8, UTF16_BE, UTF16_LE, ISO_2022,
+ * EBCDIC_STATEFUL, LATIN_1
+ * @param converter a valid, opened converter
+ * @return the type of the converter
+ * @stable ICU 2.0
+ */
+U_CAPI UConverterType U_EXPORT2
+ucnv_getType(const UConverter * converter);
+
+/**
+ * Gets the "starter" (lead) bytes for converters of type MBCS.
+ * Will fill in an <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if converter passed in
+ * is not MBCS. Fills in an array of type UBool, with the value of the byte
+ * as offset to the array. For example, if (starters[0x20] == true) at return,
+ * it means that the byte 0x20 is a starter byte in this converter.
+ * Context pointers are always owned by the caller.
+ *
+ * @param converter a valid, opened converter of type MBCS
+ * @param starters an array of size 256 to be filled in
+ * @param err error status, <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if the
+ * converter is not a type which can return starters.
+ * @see ucnv_getType
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getStarters(const UConverter* converter,
+                 UBool starters[256],
+                 UErrorCode* err);
+
+
+/**
+ * Selectors for Unicode sets that can be returned by ucnv_getUnicodeSet().
+ * @see ucnv_getUnicodeSet
+ * @stable ICU 2.6
+ */
+typedef enum UConverterUnicodeSet {
+    /** Select the set of roundtrippable Unicode code points. @stable ICU 2.6 */
+    UCNV_ROUNDTRIP_SET,
+    /** Select the set of Unicode code points with roundtrip or fallback mappings. @stable ICU 4.0 */
+    UCNV_ROUNDTRIP_AND_FALLBACK_SET,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Number of UConverterUnicodeSet selectors.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCNV_SET_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UConverterUnicodeSet;
+
+
+/**
+ * Returns the set of Unicode code points that can be converted by an ICU converter.
+ *
+ * Returns one of several kinds of set:
+ *
+ * 1. UCNV_ROUNDTRIP_SET
+ *
+ * The set of all Unicode code points that can be roundtrip-converted
+ * (converted without any data loss) with the converter (ucnv_fromUnicode()).
+ * This set will not include code points that have fallback mappings
+ * or are only the result of reverse fallback mappings.
+ * This set will also not include PUA code points with fallbacks, although
+ * ucnv_fromUnicode() will always uses those mappings despite ucnv_setFallback().
+ * See UTR #22 "Character Mapping Markup Language"
+ * at http://www.unicode.org/reports/tr22/
+ *
+ * This is useful for example for
+ * - checking that a string or document can be roundtrip-converted with a converter,
+ *   without/before actually performing the conversion
+ * - testing if a converter can be used for text for typical text for a certain locale,
+ *   by comparing its roundtrip set with the set of ExemplarCharacters from
+ *   ICU's locale data or other sources
+ *
+ * 2. UCNV_ROUNDTRIP_AND_FALLBACK_SET
+ *
+ * The set of all Unicode code points that can be converted with the converter (ucnv_fromUnicode())
+ * when fallbacks are turned on (see ucnv_setFallback()).
+ * This set includes all code points with roundtrips and fallbacks (but not reverse fallbacks).
+ *
+ * In the future, there may be more UConverterUnicodeSet choices to select
+ * sets with different properties.
+ *
+ * @param cnv The converter for which a set is requested.
+ * @param setFillIn A valid USet *. It will be cleared by this function before
+ *            the converter's specific set is filled into the USet.
+ * @param whichSet A UConverterUnicodeSet selector;
+ *              currently UCNV_ROUNDTRIP_SET is the only supported value.
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ *
+ * @see UConverterUnicodeSet
+ * @see uset_open
+ * @see uset_close
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+ucnv_getUnicodeSet(const UConverter *cnv,
+                   USet *setFillIn,
+                   UConverterUnicodeSet whichSet,
+                   UErrorCode *pErrorCode);
+
+/**
+ * Gets the current calback function used by the converter when an illegal
+ *  or invalid codepage sequence is found.
+ * Context pointers are always owned by the caller.
+ *
+ * @param converter the unicode converter
+ * @param action fillin: returns the callback function pointer
+ * @param context fillin: returns the callback's private void* context
+ * @see ucnv_setToUCallBack
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getToUCallBack (const UConverter * converter,
+                     UConverterToUCallback *action,
+                     const void **context);
+
+/**
+ * Gets the current callback function used by the converter when illegal
+ * or invalid Unicode sequence is found.
+ * Context pointers are always owned by the caller.
+ *
+ * @param converter the unicode converter
+ * @param action fillin: returns the callback function pointer
+ * @param context fillin: returns the callback's private void* context
+ * @see ucnv_setFromUCallBack
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getFromUCallBack (const UConverter * converter,
+                       UConverterFromUCallback *action,
+                       const void **context);
+
+/**
+ * Changes the callback function used by the converter when
+ * an illegal or invalid sequence is found.
+ * Context pointers are always owned by the caller.
+ * Predefined actions and contexts can be found in the ucnv_err.h header.
+ *
+ * @param converter the unicode converter
+ * @param newAction the new callback function
+ * @param newContext the new toUnicode callback context pointer. This can be NULL.
+ * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
+ * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
+ * @param err The error code status
+ * @see ucnv_getToUCallBack
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_setToUCallBack (UConverter * converter,
+                     UConverterToUCallback newAction,
+                     const void* newContext,
+                     UConverterToUCallback *oldAction,
+                     const void** oldContext,
+                     UErrorCode * err);
+
+/**
+ * Changes the current callback function used by the converter when
+ * an illegal or invalid sequence is found.
+ * Context pointers are always owned by the caller.
+ * Predefined actions and contexts can be found in the ucnv_err.h header.
+ *
+ * @param converter the unicode converter
+ * @param newAction the new callback function
+ * @param newContext the new fromUnicode callback context pointer. This can be NULL.
+ * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
+ * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
+ * @param err The error code status
+ * @see ucnv_getFromUCallBack
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_setFromUCallBack (UConverter * converter,
+                       UConverterFromUCallback newAction,
+                       const void *newContext,
+                       UConverterFromUCallback *oldAction,
+                       const void **oldContext,
+                       UErrorCode * err);
+
+/**
+ * Converts an array of unicode characters to an array of codepage
+ * characters. This function is optimized for converting a continuous
+ * stream of data in buffer-sized chunks, where the entire source and
+ * target does not fit in available buffers.
+ *
+ * The source pointer is an in/out parameter. It starts out pointing where the
+ * conversion is to begin, and ends up pointing after the last UChar consumed.
+ *
+ * Target similarly starts out pointer at the first available byte in the output
+ * buffer, and ends up pointing after the last byte written to the output.
+ *
+ * The converter always attempts to consume the entire source buffer, unless
+ * (1.) the target buffer is full, or (2.) a failing error is returned from the
+ * current callback function.  When a successful error status has been
+ * returned, it means that all of the source buffer has been
+ *  consumed. At that point, the caller should reset the source and
+ *  sourceLimit pointers to point to the next chunk.
+ *
+ * At the end of the stream (flush==true), the input is completely consumed
+ * when *source==sourceLimit and no error code is set.
+ * The converter object is then automatically reset by this function.
+ * (This means that a converter need not be reset explicitly between data
+ * streams if it finishes the previous stream without errors.)
+ *
+ * This is a <I>stateful</I> conversion. Additionally, even when all source data has
+ * been consumed, some data may be in the converters' internal state.
+ * Call this function repeatedly, updating the target pointers with
+ * the next empty chunk of target in case of a
+ * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source  pointers
+ *  with the next chunk of source when a successful error status is
+ * returned, until there are no more chunks of source data.
+ * @param converter the Unicode converter
+ * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
+ *  codepage characters to. Output : points to after the last codepage character copied
+ *  to <TT>target</TT>.
+ * @param targetLimit the pointer just after last of the <TT>target</TT> buffer
+ * @param source I/O parameter, pointer to pointer to the source Unicode character buffer.
+ * @param sourceLimit the pointer just after the last of the source buffer
+ * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
+ * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
+ * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
+ * For output data carried across calls, and other data without a specific source character
+ * (such as from escape sequences or callbacks)  -1 will be placed for offsets.
+ * @param flush set to <TT>true</TT> if the current source buffer is the last available
+ * chunk of the source, <TT>false</TT> otherwise. Note that if a failing status is returned,
+ * this function may have to be called multiple times with flush set to <TT>true</TT> until
+ * the source buffer is consumed.
+ * @param err the error status.  <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
+ * converter is <TT>NULL</TT>.
+ * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
+ * still data to be written to the target.
+ * @see ucnv_fromUChars
+ * @see ucnv_convert
+ * @see ucnv_getMinCharSize
+ * @see ucnv_setToUCallBack
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_fromUnicode (UConverter * converter,
+                  char **target,
+                  const char *targetLimit,
+                  const UChar ** source,
+                  const UChar * sourceLimit,
+                  int32_t* offsets,
+                  UBool flush,
+                  UErrorCode * err);
+
+/**
+ * Converts a buffer of codepage bytes into an array of unicode UChars
+ * characters. This function is optimized for converting a continuous
+ * stream of data in buffer-sized chunks, where the entire source and
+ * target does not fit in available buffers.
+ *
+ * The source pointer is an in/out parameter. It starts out pointing where the
+ * conversion is to begin, and ends up pointing after the last byte of source consumed.
+ *
+ * Target similarly starts out pointer at the first available UChar in the output
+ * buffer, and ends up pointing after the last UChar written to the output.
+ * It does NOT necessarily keep UChar sequences together.
+ *
+ * The converter always attempts to consume the entire source buffer, unless
+ * (1.) the target buffer is full, or (2.) a failing error is returned from the
+ * current callback function.  When a successful error status has been
+ * returned, it means that all of the source buffer has been
+ *  consumed. At that point, the caller should reset the source and
+ *  sourceLimit pointers to point to the next chunk.
+ *
+ * At the end of the stream (flush==true), the input is completely consumed
+ * when *source==sourceLimit and no error code is set
+ * The converter object is then automatically reset by this function.
+ * (This means that a converter need not be reset explicitly between data
+ * streams if it finishes the previous stream without errors.)
+ *
+ * This is a <I>stateful</I> conversion. Additionally, even when all source data has
+ * been consumed, some data may be in the converters' internal state.
+ * Call this function repeatedly, updating the target pointers with
+ * the next empty chunk of target in case of a
+ * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source  pointers
+ *  with the next chunk of source when a successful error status is
+ * returned, until there are no more chunks of source data.
+ * @param converter the Unicode converter
+ * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
+ *  UChars into. Output : points to after the last UChar copied.
+ * @param targetLimit the pointer just after the end of the <TT>target</TT> buffer
+ * @param source I/O parameter, pointer to pointer to the source codepage buffer.
+ * @param sourceLimit the pointer to the byte after the end of the source buffer
+ * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
+ * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
+ * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
+ * For output data carried across calls, and other data without a specific source character
+ * (such as from escape sequences or callbacks)  -1 will be placed for offsets.
+ * @param flush set to <TT>true</TT> if the current source buffer is the last available
+ * chunk of the source, <TT>false</TT> otherwise. Note that if a failing status is returned,
+ * this function may have to be called multiple times with flush set to <TT>true</TT> until
+ * the source buffer is consumed.
+ * @param err the error status.  <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
+ * converter is <TT>NULL</TT>.
+ * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
+ * still data to be written to the target.
+ * @see ucnv_fromUChars
+ * @see ucnv_convert
+ * @see ucnv_getMinCharSize
+ * @see ucnv_setFromUCallBack
+ * @see ucnv_getNextUChar
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_toUnicode(UConverter *converter,
+               UChar **target,
+               const UChar *targetLimit,
+               const char **source,
+               const char *sourceLimit,
+               int32_t *offsets,
+               UBool flush,
+               UErrorCode *err);
+
+/**
+ * Convert the Unicode string into a codepage string using an existing UConverter.
+ * The output string is NUL-terminated if possible.
+ *
+ * This function is a more convenient but less powerful version of ucnv_fromUnicode().
+ * It is only useful for whole strings, not for streaming conversion.
+ *
+ * The maximum output buffer capacity required (barring output from callbacks) will be
+ * UCNV_GET_MAX_BYTES_FOR_STRING(srcLength, ucnv_getMaxCharSize(cnv)).
+ *
+ * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called)
+ * @param src the input Unicode string
+ * @param srcLength the input string length, or -1 if NUL-terminated
+ * @param dest destination string buffer, can be NULL if destCapacity==0
+ * @param destCapacity the number of chars available at dest
+ * @param pErrorCode normal ICU error code;
+ *                  common error codes that may be set by this function include
+ *                  U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
+ *                  U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
+ * @return the length of the output string, not counting the terminating NUL;
+ *         if the length is greater than destCapacity, then the string will not fit
+ *         and a buffer of the indicated length would need to be passed in
+ * @see ucnv_fromUnicode
+ * @see ucnv_convert
+ * @see UCNV_GET_MAX_BYTES_FOR_STRING
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_fromUChars(UConverter *cnv,
+                char *dest, int32_t destCapacity,
+                const UChar *src, int32_t srcLength,
+                UErrorCode *pErrorCode);
+
+/**
+ * Convert the codepage string into a Unicode string using an existing UConverter.
+ * The output string is NUL-terminated if possible.
+ *
+ * This function is a more convenient but less powerful version of ucnv_toUnicode().
+ * It is only useful for whole strings, not for streaming conversion.
+ *
+ * The maximum output buffer capacity required (barring output from callbacks) will be
+ * 2*srcLength (each char may be converted into a surrogate pair).
+ *
+ * @param cnv the converter object to be used (ucnv_resetToUnicode() will be called)
+ * @param src the input codepage string
+ * @param srcLength the input string length, or -1 if NUL-terminated
+ * @param dest destination string buffer, can be NULL if destCapacity==0
+ * @param destCapacity the number of UChars available at dest
+ * @param pErrorCode normal ICU error code;
+ *                  common error codes that may be set by this function include
+ *                  U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
+ *                  U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
+ * @return the length of the output string, not counting the terminating NUL;
+ *         if the length is greater than destCapacity, then the string will not fit
+ *         and a buffer of the indicated length would need to be passed in
+ * @see ucnv_toUnicode
+ * @see ucnv_convert
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_toUChars(UConverter *cnv,
+              UChar *dest, int32_t destCapacity,
+              const char *src, int32_t srcLength,
+              UErrorCode *pErrorCode);
+
+/**
+ * Convert a codepage buffer into Unicode one character at a time.
+ * The input is completely consumed when the U_INDEX_OUTOFBOUNDS_ERROR is set.
+ *
+ * Advantage compared to ucnv_toUnicode() or ucnv_toUChars():
+ * - Faster for small amounts of data, for most converters, e.g.,
+ *   US-ASCII, ISO-8859-1, UTF-8/16/32, and most "normal" charsets.
+ *   (For complex converters, e.g., SCSU, UTF-7 and ISO 2022 variants,
+ *    it uses ucnv_toUnicode() internally.)
+ * - Convenient.
+ *
+ * Limitations compared to ucnv_toUnicode():
+ * - Always assumes flush=true.
+ *   This makes ucnv_getNextUChar() unsuitable for "streaming" conversion,
+ *   that is, for where the input is supplied in multiple buffers,
+ *   because ucnv_getNextUChar() will assume the end of the input at the end
+ *   of the first buffer.
+ * - Does not provide offset output.
+ *
+ * It is possible to "mix" ucnv_getNextUChar() and ucnv_toUnicode() because
+ * ucnv_getNextUChar() uses the current state of the converter
+ * (unlike ucnv_toUChars() which always resets first).
+ * However, if ucnv_getNextUChar() is called after ucnv_toUnicode()
+ * stopped in the middle of a character sequence (with flush=false),
+ * then ucnv_getNextUChar() will always use the slower ucnv_toUnicode()
+ * internally until the next character boundary.
+ * (This is new in ICU 2.6. In earlier releases, ucnv_getNextUChar() had to
+ * start at a character boundary.)
+ *
+ * Instead of using ucnv_getNextUChar(), it is recommended
+ * to convert using ucnv_toUnicode() or ucnv_toUChars()
+ * and then iterate over the text using U16_NEXT() or a UCharIterator (uiter.h)
+ * or a C++ CharacterIterator or similar.
+ * This allows streaming conversion and offset output, for example.
+ *
+ * <p>Handling of surrogate pairs and supplementary-plane code points:<br>
+ * There are two different kinds of codepages that provide mappings for surrogate characters:
+ * <ul>
+ *   <li>Codepages like UTF-8, UTF-32, and GB 18030 provide direct representations for Unicode
+ *       code points U+10000-U+10ffff as well as for single surrogates U+d800-U+dfff.
+ *       Each valid sequence will result in exactly one returned code point.
+ *       If a sequence results in a single surrogate, then that will be returned
+ *       by itself, even if a neighboring sequence encodes the matching surrogate.</li>
+ *   <li>Codepages like SCSU and LMBCS (and UTF-16) provide direct representations only for BMP code points
+ *       including surrogates. Code points in supplementary planes are represented with
+ *       two sequences, each encoding a surrogate.
+ *       For these codepages, matching pairs of surrogates will be combined into single
+ *       code points for returning from this function.
+ *       (Note that SCSU is actually a mix of these codepage types.)</li>
+ * </ul></p>
+ *
+ * @param converter an open UConverter
+ * @param source the address of a pointer to the codepage buffer, will be
+ *  updated to point after the bytes consumed in the conversion call.
+ * @param sourceLimit points to the end of the input buffer
+ * @param err fills in error status (see ucnv_toUnicode)
+ * <code>U_INDEX_OUTOFBOUNDS_ERROR</code> will be set if the input
+ * is empty or does not convert to any output (e.g.: pure state-change
+ * codes SI/SO, escape sequences for ISO 2022,
+ * or if the callback did not output anything, ...).
+ * This function will not set a <code>U_BUFFER_OVERFLOW_ERROR</code> because
+ *  the "buffer" is the return code. However, there might be subsequent output
+ *  stored in the converter object
+ * that will be returned in following calls to this function.
+ * @return a UChar32 resulting from the partial conversion of source
+ * @see ucnv_toUnicode
+ * @see ucnv_toUChars
+ * @see ucnv_convert
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+ucnv_getNextUChar(UConverter * converter,
+                  const char **source,
+                  const char * sourceLimit,
+                  UErrorCode * err);
+
+/**
+ * Convert from one external charset to another using two existing UConverters.
+ * Internally, two conversions - ucnv_toUnicode() and ucnv_fromUnicode() -
+ * are used, "pivoting" through 16-bit Unicode.
+ *
+ * Important: For streaming conversion (multiple function calls for successive
+ * parts of a text stream), the caller must provide a pivot buffer explicitly,
+ * and must preserve the pivot buffer and associated pointers from one
+ * call to another. (The buffer may be moved if its contents and the relative
+ * pointer positions are preserved.)
+ *
+ * There is a similar function, ucnv_convert(),
+ * which has the following limitations:
+ * - it takes charset names, not converter objects, so that
+ *   - two converters are opened for each call
+ *   - only single-string conversion is possible, not streaming operation
+ * - it does not provide enough information to find out,
+ *   in case of failure, whether the toUnicode or
+ *   the fromUnicode conversion failed
+ *
+ * By contrast, ucnv_convertEx()
+ * - takes UConverter parameters instead of charset names
+ * - fully exposes the pivot buffer for streaming conversion and complete error handling
+ *
+ * ucnv_convertEx() also provides further convenience:
+ * - an option to reset the converters at the beginning
+ *   (if reset==true, see parameters;
+ *    also sets *pivotTarget=*pivotSource=pivotStart)
+ * - allow NUL-terminated input
+ *   (only a single NUL byte, will not work for charsets with multi-byte NULs)
+ *   (if sourceLimit==NULL, see parameters)
+ * - terminate with a NUL on output
+ *   (only a single NUL byte, not useful for charsets with multi-byte NULs),
+ *   or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
+ *   the target buffer
+ * - the pivot buffer can be provided internally;
+ *   possible only for whole-string conversion, not streaming conversion;
+ *   in this case, the caller will not be able to get details about where an
+ *   error occurred
+ *   (if pivotStart==NULL, see below)
+ *
+ * The function returns when one of the following is true:
+ * - the entire source text has been converted successfully to the target buffer
+ * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
+ * - a conversion error occurred
+ *   (other U_FAILURE(), see description of pErrorCode)
+ *
+ * Limitation compared to the direct use of
+ * ucnv_fromUnicode() and ucnv_toUnicode():
+ * ucnv_convertEx() does not provide offset information.
+ *
+ * Limitation compared to ucnv_fromUChars() and ucnv_toUChars():
+ * ucnv_convertEx() does not support preflighting directly.
+ *
+ * Sample code for converting a single string from
+ * one external charset to UTF-8, ignoring the location of errors:
+ *
+ * \code
+ * int32_t
+ * myToUTF8(UConverter *cnv,
+ *          const char *s, int32_t length,
+ *          char *u8, int32_t capacity,
+ *          UErrorCode *pErrorCode) {
+ *     UConverter *utf8Cnv;
+ *     char *target;
+ *
+ *     if(U_FAILURE(*pErrorCode)) {
+ *         return 0;
+ *     }
+ *
+ *     utf8Cnv=myGetCachedUTF8Converter(pErrorCode);
+ *     if(U_FAILURE(*pErrorCode)) {
+ *         return 0;
+ *     }
+ *
+ *     if(length<0) {
+ *         length=strlen(s);
+ *     }
+ *     target=u8;
+ *     ucnv_convertEx(utf8Cnv, cnv,
+ *                    &target, u8+capacity,
+ *                    &s, s+length,
+ *                    NULL, NULL, NULL, NULL,
+ *                    true, true,
+ *                    pErrorCode);
+ *
+ *     myReleaseCachedUTF8Converter(utf8Cnv);
+ *
+ *     // return the output string length, but without preflighting
+ *     return (int32_t)(target-u8);
+ * }
+ * \endcode
+ *
+ * @param targetCnv     Output converter, used to convert from the UTF-16 pivot
+ *                      to the target using ucnv_fromUnicode().
+ * @param sourceCnv     Input converter, used to convert from the source to
+ *                      the UTF-16 pivot using ucnv_toUnicode().
+ * @param target        I/O parameter, same as for ucnv_fromUChars().
+ *                      Input: *target points to the beginning of the target buffer.
+ *                      Output: *target points to the first unit after the last char written.
+ * @param targetLimit   Pointer to the first unit after the target buffer.
+ * @param source        I/O parameter, same as for ucnv_toUChars().
+ *                      Input: *source points to the beginning of the source buffer.
+ *                      Output: *source points to the first unit after the last char read.
+ * @param sourceLimit   Pointer to the first unit after the source buffer.
+ * @param pivotStart    Pointer to the UTF-16 pivot buffer. If pivotStart==NULL,
+ *                      then an internal buffer is used and the other pivot
+ *                      arguments are ignored and can be NULL as well.
+ * @param pivotSource   I/O parameter, same as source in ucnv_fromUChars() for
+ *                      conversion from the pivot buffer to the target buffer.
+ * @param pivotTarget   I/O parameter, same as target in ucnv_toUChars() for
+ *                      conversion from the source buffer to the pivot buffer.
+ *                      It must be pivotStart<=*pivotSource<=*pivotTarget<=pivotLimit
+ *                      and pivotStart<pivotLimit (unless pivotStart==NULL).
+ * @param pivotLimit    Pointer to the first unit after the pivot buffer.
+ * @param reset         If true, then ucnv_resetToUnicode(sourceCnv) and
+ *                      ucnv_resetFromUnicode(targetCnv) are called, and the
+ *                      pivot pointers are reset (*pivotTarget=*pivotSource=pivotStart).
+ * @param flush         If true, indicates the end of the input.
+ *                      Passed directly to ucnv_toUnicode(), and carried over to
+ *                      ucnv_fromUnicode() when the source is empty as well.
+ * @param pErrorCode    ICU error code in/out parameter.
+ *                      Must fulfill U_SUCCESS before the function call.
+ *                      U_BUFFER_OVERFLOW_ERROR always refers to the target buffer
+ *                      because overflows into the pivot buffer are handled internally.
+ *                      Other conversion errors are from the source-to-pivot
+ *                      conversion if *pivotSource==pivotStart, otherwise from
+ *                      the pivot-to-target conversion.
+ *
+ * @see ucnv_convert
+ * @see ucnv_fromAlgorithmic
+ * @see ucnv_toAlgorithmic
+ * @see ucnv_fromUnicode
+ * @see ucnv_toUnicode
+ * @see ucnv_fromUChars
+ * @see ucnv_toUChars
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
+               char **target, const char *targetLimit,
+               const char **source, const char *sourceLimit,
+               UChar *pivotStart, UChar **pivotSource,
+               UChar **pivotTarget, const UChar *pivotLimit,
+               UBool reset, UBool flush,
+               UErrorCode *pErrorCode);
+
+/**
+ * Convert from one external charset to another.
+ * Internally, two converters are opened according to the name arguments,
+ * then the text is converted to and from the 16-bit Unicode "pivot"
+ * using ucnv_convertEx(), then the converters are closed again.
+ *
+ * This is a convenience function, not an efficient way to convert a lot of text:
+ * ucnv_convert()
+ * - takes charset names, not converter objects, so that
+ *   - two converters are opened for each call
+ *   - only single-string conversion is possible, not streaming operation
+ * - does not provide enough information to find out,
+ *   in case of failure, whether the toUnicode or
+ *   the fromUnicode conversion failed
+ * - allows NUL-terminated input
+ *   (only a single NUL byte, will not work for charsets with multi-byte NULs)
+ *   (if sourceLength==-1, see parameters)
+ * - terminate with a NUL on output
+ *   (only a single NUL byte, not useful for charsets with multi-byte NULs),
+ *   or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
+ *   the target buffer
+ * - a pivot buffer is provided internally
+ *
+ * The function returns when one of the following is true:
+ * - the entire source text has been converted successfully to the target buffer
+ *   and either the target buffer is terminated with a single NUL byte
+ *   or the error code is set to U_STRING_NOT_TERMINATED_WARNING
+ * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
+ *   and the full output string length is returned ("preflighting")
+ * - a conversion error occurred
+ *   (other U_FAILURE(), see description of pErrorCode)
+ *
+ * @param toConverterName   The name of the converter that is used to convert
+ *                          from the UTF-16 pivot buffer to the target.
+ * @param fromConverterName The name of the converter that is used to convert
+ *                          from the source to the UTF-16 pivot buffer.
+ * @param target            Pointer to the output buffer.
+ * @param targetCapacity    Capacity of the target, in bytes.
+ * @param source            Pointer to the input buffer.
+ * @param sourceLength      Length of the input text, in bytes, or -1 for NUL-terminated input.
+ * @param pErrorCode        ICU error code in/out parameter.
+ *                          Must fulfill U_SUCCESS before the function call.
+ * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
+ *         and a U_BUFFER_OVERFLOW_ERROR is set.
+ *
+ * @see ucnv_convertEx
+ * @see ucnv_fromAlgorithmic
+ * @see ucnv_toAlgorithmic
+ * @see ucnv_fromUnicode
+ * @see ucnv_toUnicode
+ * @see ucnv_fromUChars
+ * @see ucnv_toUChars
+ * @see ucnv_getNextUChar
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_convert(const char *toConverterName,
+             const char *fromConverterName,
+             char *target,
+             int32_t targetCapacity,
+             const char *source,
+             int32_t sourceLength,
+             UErrorCode *pErrorCode);
+
+/**
+ * Convert from one external charset to another.
+ * Internally, the text is converted to and from the 16-bit Unicode "pivot"
+ * using ucnv_convertEx(). ucnv_toAlgorithmic() works exactly like ucnv_convert()
+ * except that the two converters need not be looked up and opened completely.
+ *
+ * The source-to-pivot conversion uses the cnv converter parameter.
+ * The pivot-to-target conversion uses a purely algorithmic converter
+ * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
+ *
+ * Internally, the algorithmic converter is opened and closed for each
+ * function call, which is more efficient than using the public ucnv_open()
+ * but somewhat less efficient than only resetting an existing converter
+ * and using ucnv_convertEx().
+ *
+ * This function is more convenient than ucnv_convertEx() for single-string
+ * conversions, especially when "preflighting" is desired (returning the length
+ * of the complete output even if it does not fit into the target buffer;
+ * see the User Guide Strings chapter). See ucnv_convert() for details.
+ *
+ * @param algorithmicType   UConverterType constant identifying the desired target
+ *                          charset as a purely algorithmic converter.
+ *                          Those are converters for Unicode charsets like
+ *                          UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
+ *                          as well as US-ASCII and ISO-8859-1.
+ * @param cnv               The converter that is used to convert
+ *                          from the source to the UTF-16 pivot buffer.
+ * @param target            Pointer to the output buffer.
+ * @param targetCapacity    Capacity of the target, in bytes.
+ * @param source            Pointer to the input buffer.
+ * @param sourceLength      Length of the input text, in bytes
+ * @param pErrorCode        ICU error code in/out parameter.
+ *                          Must fulfill U_SUCCESS before the function call.
+ * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
+ *         and a U_BUFFER_OVERFLOW_ERROR is set.
+ *
+ * @see ucnv_fromAlgorithmic
+ * @see ucnv_convert
+ * @see ucnv_convertEx
+ * @see ucnv_fromUnicode
+ * @see ucnv_toUnicode
+ * @see ucnv_fromUChars
+ * @see ucnv_toUChars
+ * @stable ICU 2.6
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_toAlgorithmic(UConverterType algorithmicType,
+                   UConverter *cnv,
+                   char *target, int32_t targetCapacity,
+                   const char *source, int32_t sourceLength,
+                   UErrorCode *pErrorCode);
+
+/**
+ * Convert from one external charset to another.
+ * Internally, the text is converted to and from the 16-bit Unicode "pivot"
+ * using ucnv_convertEx(). ucnv_fromAlgorithmic() works exactly like ucnv_convert()
+ * except that the two converters need not be looked up and opened completely.
+ *
+ * The source-to-pivot conversion uses a purely algorithmic converter
+ * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
+ * The pivot-to-target conversion uses the cnv converter parameter.
+ *
+ * Internally, the algorithmic converter is opened and closed for each
+ * function call, which is more efficient than using the public ucnv_open()
+ * but somewhat less efficient than only resetting an existing converter
+ * and using ucnv_convertEx().
+ *
+ * This function is more convenient than ucnv_convertEx() for single-string
+ * conversions, especially when "preflighting" is desired (returning the length
+ * of the complete output even if it does not fit into the target buffer;
+ * see the User Guide Strings chapter). See ucnv_convert() for details.
+ *
+ * @param cnv               The converter that is used to convert
+ *                          from the UTF-16 pivot buffer to the target.
+ * @param algorithmicType   UConverterType constant identifying the desired source
+ *                          charset as a purely algorithmic converter.
+ *                          Those are converters for Unicode charsets like
+ *                          UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
+ *                          as well as US-ASCII and ISO-8859-1.
+ * @param target            Pointer to the output buffer.
+ * @param targetCapacity    Capacity of the target, in bytes.
+ * @param source            Pointer to the input buffer.
+ * @param sourceLength      Length of the input text, in bytes
+ * @param pErrorCode        ICU error code in/out parameter.
+ *                          Must fulfill U_SUCCESS before the function call.
+ * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
+ *         and a U_BUFFER_OVERFLOW_ERROR is set.
+ *
+ * @see ucnv_fromAlgorithmic
+ * @see ucnv_convert
+ * @see ucnv_convertEx
+ * @see ucnv_fromUnicode
+ * @see ucnv_toUnicode
+ * @see ucnv_fromUChars
+ * @see ucnv_toUChars
+ * @stable ICU 2.6
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_fromAlgorithmic(UConverter *cnv,
+                     UConverterType algorithmicType,
+                     char *target, int32_t targetCapacity,
+                     const char *source, int32_t sourceLength,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Frees up memory occupied by unused, cached converter shared data.
+ *
+ * @return the number of cached converters successfully deleted
+ * @see ucnv_close
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_flushCache(void);
+
+/**
+ * Returns the number of available converters, as per the alias file.
+ *
+ * @return the number of available converters
+ * @see ucnv_getAvailableName
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_countAvailable(void);
+
+/**
+ * Gets the canonical converter name of the specified converter from a list of
+ * all available converters contaied in the alias file. All converters
+ * in this list can be opened.
+ *
+ * @param n the index to a converter available on the system (in the range <TT>[0..ucnv_countAvaiable()]</TT>)
+ * @return a pointer a string (library owned), or <TT>NULL</TT> if the index is out of bounds.
+ * @see ucnv_countAvailable
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2
+ucnv_getAvailableName(int32_t n);
+
+/**
+ * Returns a UEnumeration to enumerate all of the canonical converter
+ * names, as per the alias file, regardless of the ability to open each
+ * converter.
+ *
+ * @return A UEnumeration object for getting all the recognized canonical
+ *   converter names.
+ * @see ucnv_getAvailableName
+ * @see uenum_close
+ * @see uenum_next
+ * @stable ICU 2.4
+ */
+U_CAPI UEnumeration * U_EXPORT2
+ucnv_openAllNames(UErrorCode *pErrorCode);
+
+/**
+ * Gives the number of aliases for a given converter or alias name.
+ * If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
+ * This method only enumerates the listed entries in the alias file.
+ * @param alias alias name
+ * @param pErrorCode error status
+ * @return number of names on alias list for given alias
+ * @stable ICU 2.0
+ */
+U_CAPI uint16_t U_EXPORT2
+ucnv_countAliases(const char *alias, UErrorCode *pErrorCode);
+
+/**
+ * Gives the name of the alias at given index of alias list.
+ * This method only enumerates the listed entries in the alias file.
+ * If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
+ * @param alias alias name
+ * @param n index in alias list
+ * @param pErrorCode result of operation
+ * @return returns the name of the alias at given index
+ * @see ucnv_countAliases
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode);
+
+/**
+ * Fill-up the list of alias names for the given alias.
+ * This method only enumerates the listed entries in the alias file.
+ * If the alias is ambiguous, then the preferred converter is used
+ * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
+ * @param alias alias name
+ * @param aliases fill-in list, aliases is a pointer to an array of
+ *        <code>ucnv_countAliases()</code> string-pointers
+ *        (<code>const char *</code>) that will be filled in.
+ *        The strings themselves are owned by the library.
+ * @param pErrorCode result of operation
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode);
+
+/**
+ * Return a new UEnumeration object for enumerating all the
+ * alias names for a given converter that are recognized by a standard.
+ * This method only enumerates the listed entries in the alias file.
+ * The convrtrs.txt file can be modified to change the results of
+ * this function.
+ * The first result in this list is the same result given by
+ * <code>ucnv_getStandardName</code>, which is the default alias for
+ * the specified standard name. The returned object must be closed with
+ * <code>uenum_close</code> when you are done with the object.
+ *
+ * @param convName original converter name
+ * @param standard name of the standard governing the names; MIME and IANA
+ *      are such standards
+ * @param pErrorCode The error code
+ * @return A UEnumeration object for getting all aliases that are recognized
+ *      by a standard. If any of the parameters are invalid, NULL
+ *      is returned.
+ * @see ucnv_getStandardName
+ * @see uenum_close
+ * @see uenum_next
+ * @stable ICU 2.2
+ */
+U_CAPI UEnumeration * U_EXPORT2
+ucnv_openStandardNames(const char *convName,
+                       const char *standard,
+                       UErrorCode *pErrorCode);
+
+/**
+ * Gives the number of standards associated to converter names.
+ * @return number of standards
+ * @stable ICU 2.0
+ */
+U_CAPI uint16_t U_EXPORT2
+ucnv_countStandards(void);
+
+/**
+ * Gives the name of the standard at given index of standard list.
+ * @param n index in standard list
+ * @param pErrorCode result of operation
+ * @return returns the name of the standard at given index. Owned by the library.
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getStandard(uint16_t n, UErrorCode *pErrorCode);
+
+/**
+ * Returns a standard name for a given converter name.
+ * <p>
+ * Example alias table:<br>
+ * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
+ * <p>
+ * Result of ucnv_getStandardName("conv", "STANDARD1") from example
+ * alias table:<br>
+ * <b>"alias2"</b>
+ *
+ * @param name original converter name
+ * @param standard name of the standard governing the names; MIME and IANA
+ *        are such standards
+ * @param pErrorCode result of operation
+ * @return returns the standard converter name;
+ *         if a standard converter name cannot be determined,
+ *         then <code>NULL</code> is returned. Owned by the library.
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getStandardName(const char *name, const char *standard, UErrorCode *pErrorCode);
+
+/**
+ * This function will return the internal canonical converter name of the
+ * tagged alias. This is the opposite of ucnv_openStandardNames, which
+ * returns the tagged alias given the canonical name.
+ * <p>
+ * Example alias table:<br>
+ * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
+ * <p>
+ * Result of ucnv_getStandardName("alias1", "STANDARD1") from example
+ * alias table:<br>
+ * <b>"conv"</b>
+ *
+ * @return returns the canonical converter name;
+ *         if a standard or alias name cannot be determined,
+ *         then <code>NULL</code> is returned. The returned string is
+ *         owned by the library.
+ * @see ucnv_getStandardName
+ * @stable ICU 2.4
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getCanonicalName(const char *alias, const char *standard, UErrorCode *pErrorCode);
+
+/**
+ * Returns the current default converter name. If you want to open
+ * a default converter, you do not need to use this function.
+ * It is faster if you pass a NULL argument to ucnv_open the
+ * default converter.
+ *
+ * If U_CHARSET_IS_UTF8 is defined to 1 in utypes.h then this function
+ * always returns "UTF-8".
+ *
+ * @return returns the current default converter name.
+ *         Storage owned by the library
+ * @see ucnv_setDefaultName
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ucnv_getDefaultName(void);
+
+#ifndef U_HIDE_SYSTEM_API
+/**
+ * This function is not thread safe. DO NOT call this function when ANY ICU
+ * function is being used from more than one thread! This function sets the
+ * current default converter name. If this function needs to be called, it
+ * should be called during application initialization. Most of the time, the
+ * results from ucnv_getDefaultName() or ucnv_open with a NULL string argument
+ * is sufficient for your application.
+ *
+ * If U_CHARSET_IS_UTF8 is defined to 1 in utypes.h then this function
+ * does nothing.
+ *
+ * @param name the converter name to be the default (must be known by ICU).
+ * @see ucnv_getDefaultName
+ * @system
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_setDefaultName(const char *name);
+#endif  /* U_HIDE_SYSTEM_API */
+
+/**
+ * Fixes the backslash character mismapping.  For example, in SJIS, the backslash
+ * character in the ASCII portion is also used to represent the yen currency sign.
+ * When mapping from Unicode character 0x005C, it's unclear whether to map the
+ * character back to yen or backslash in SJIS.  This function will take the input
+ * buffer and replace all the yen sign characters with backslash.  This is necessary
+ * when the user tries to open a file with the input buffer on Windows.
+ * This function will test the converter to see whether such mapping is
+ * required.  You can sometimes avoid using this function by using the correct version
+ * of Shift-JIS.
+ *
+ * @param cnv The converter representing the target codepage.
+ * @param source the input buffer to be fixed
+ * @param sourceLen the length of the input buffer
+ * @see ucnv_isAmbiguous
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_fixFileSeparator(const UConverter *cnv, UChar *source, int32_t sourceLen);
+
+/**
+ * Determines if the converter contains ambiguous mappings of the same
+ * character or not.
+ * @param cnv the converter to be tested
+ * @return true if the converter contains ambiguous mapping of the same
+ * character, false otherwise.
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+ucnv_isAmbiguous(const UConverter *cnv);
+
+/**
+ * Sets the converter to use fallback mappings or not.
+ * Regardless of this flag, the converter will always use
+ * fallbacks from Unicode Private Use code points, as well as
+ * reverse fallbacks (to Unicode).
+ * For details see ".ucm File Format"
+ * in the Conversion Data chapter of the ICU User Guide:
+ * http://www.icu-project.org/userguide/conversion-data.html#ucmformat
+ *
+ * @param cnv The converter to set the fallback mapping usage on.
+ * @param usesFallback true if the user wants the converter to take advantage of the fallback
+ * mapping, false otherwise.
+ * @stable ICU 2.0
+ * @see ucnv_usesFallback
+ */
+U_CAPI void U_EXPORT2
+ucnv_setFallback(UConverter *cnv, UBool usesFallback);
+
+/**
+ * Determines if the converter uses fallback mappings or not.
+ * This flag has restrictions, see ucnv_setFallback().
+ *
+ * @param cnv The converter to be tested
+ * @return true if the converter uses fallback, false otherwise.
+ * @stable ICU 2.0
+ * @see ucnv_setFallback
+ */
+U_CAPI UBool U_EXPORT2
+ucnv_usesFallback(const UConverter *cnv);
+
+/**
+ * Detects Unicode signature byte sequences at the start of the byte stream
+ * and returns the charset name of the indicated Unicode charset.
+ * NULL is returned when no Unicode signature is recognized.
+ * The number of bytes in the signature is output as well.
+ *
+ * The caller can ucnv_open() a converter using the charset name.
+ * The first code unit (UChar) from the start of the stream will be U+FEFF
+ * (the Unicode BOM/signature character) and can usually be ignored.
+ *
+ * For most Unicode charsets it is also possible to ignore the indicated
+ * number of initial stream bytes and start converting after them.
+ * However, there are stateful Unicode charsets (UTF-7 and BOCU-1) for which
+ * this will not work. Therefore, it is best to ignore the first output UChar
+ * instead of the input signature bytes.
+ * <p>
+ * Usage:
+ * \snippet samples/ucnv/convsamp.cpp ucnv_detectUnicodeSignature
+ *
+ * @param source            The source string in which the signature should be detected.
+ * @param sourceLength      Length of the input string, or -1 if terminated with a NUL byte.
+ * @param signatureLength   A pointer to int32_t to receive the number of bytes that make up the signature
+ *                          of the detected UTF. 0 if not detected.
+ *                          Can be a NULL pointer.
+ * @param pErrorCode        ICU error code in/out parameter.
+ *                          Must fulfill U_SUCCESS before the function call.
+ * @return The name of the encoding detected. NULL if encoding is not detected.
+ * @stable ICU 2.4
+ */
+U_CAPI const char* U_EXPORT2
+ucnv_detectUnicodeSignature(const char* source,
+                            int32_t sourceLength,
+                            int32_t *signatureLength,
+                            UErrorCode *pErrorCode);
+
+/**
+ * Returns the number of UChars held in the converter's internal state
+ * because more input is needed for completing the conversion. This function is
+ * useful for mapping semantics of ICU's converter interface to those of iconv,
+ * and this information is not needed for normal conversion.
+ * @param cnv       The converter in which the input is held
+ * @param status    ICU error code in/out parameter.
+ *                  Must fulfill U_SUCCESS before the function call.
+ * @return The number of UChars in the state. -1 if an error is encountered.
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_fromUCountPending(const UConverter* cnv, UErrorCode* status);
+
+/**
+ * Returns the number of chars held in the converter's internal state
+ * because more input is needed for completing the conversion. This function is
+ * useful for mapping semantics of ICU's converter interface to those of iconv,
+ * and this information is not needed for normal conversion.
+ * @param cnv       The converter in which the input is held as internal state
+ * @param status    ICU error code in/out parameter.
+ *                  Must fulfill U_SUCCESS before the function call.
+ * @return The number of chars in the state. -1 if an error is encountered.
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+ucnv_toUCountPending(const UConverter* cnv, UErrorCode* status);
+
+/**
+ * Returns whether or not the charset of the converter has a fixed number of bytes
+ * per charset character.
+ * An example of this are converters that are of the type UCNV_SBCS or UCNV_DBCS.
+ * Another example is UTF-32 which is always 4 bytes per character.
+ * A Unicode code point may be represented by more than one UTF-8 or UTF-16 code unit
+ * but a UTF-32 converter encodes each code point with 4 bytes.
+ * Note: This method is not intended to be used to determine whether the charset has a
+ * fixed ratio of bytes to Unicode codes <i>units</i> for any particular Unicode encoding form.
+ * false is returned with the UErrorCode if error occurs or cnv is NULL.
+ * @param cnv       The converter to be tested
+ * @param status    ICU error code in/out paramter
+ * @return true if the converter is fixed-width
+ * @stable ICU 4.8
+ */
+U_CAPI UBool U_EXPORT2
+ucnv_isFixedWidth(UConverter *cnv, UErrorCode *status);
+
+#endif
+
+#endif
+/*_UCNV*/
diff --git a/thirdparty/icu4c/common/unicode/ucnv_cb.h b/thirdparty/icu4c/common/unicode/ucnv_cb.h
new file mode 100644
index 0000000000..41845d1bca
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucnv_cb.h
@@ -0,0 +1,164 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2000-2004, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+ *  ucnv_cb.h:
+ *  External APIs for the ICU's codeset conversion library
+ *  Helena Shih
+ * 
+ * Modification History:
+ *
+ *   Date        Name        Description
+ */
+
+/**
+ * \file 
+ * \brief C UConverter functions to aid the writers of callbacks
+ *
+ * <h2> Callback API for UConverter </h2>
+ * 
+ * These functions are provided here for the convenience of the callback
+ * writer. If you are just looking for callback functions to use, please
+ * see ucnv_err.h.  DO NOT call these functions directly when you are 
+ * working with converters, unless your code has been called as a callback
+ * via ucnv_setFromUCallback or ucnv_setToUCallback !!
+ * 
+ * A note about error codes and overflow.  Unlike other ICU functions,
+ * these functions do not expect the error status to be U_ZERO_ERROR.
+ * Callbacks must be much more careful about their error codes.
+ * The error codes used here are in/out parameters, which should be passed
+ * back in the callback's error parameter.
+ * 
+ * For example, if you call ucnv_cbfromUWriteBytes to write data out 
+ * to the output codepage, it may return U_BUFFER_OVERFLOW_ERROR if 
+ * the data did not fit in the target. But this isn't a failing error, 
+ * in fact, ucnv_cbfromUWriteBytes may be called AGAIN with the error
+ * status still U_BUFFER_OVERFLOW_ERROR to attempt to write further bytes,
+ * which will also go into the internal overflow buffers.
+ * 
+ * Concerning offsets, the 'offset' parameters here are relative to the start
+ * of SOURCE.  For example, Suppose the string "ABCD" was being converted 
+ * from Unicode into a codepage which doesn't have a mapping for 'B'.
+ * 'A' will be written out correctly, but
+ * The FromU Callback will be called on an unassigned character for 'B'.
+ * At this point, this is the state of the world:
+ *    Target:    A [..]     [points after A]
+ *    Source:  A B [C] D    [points to C - B has been consumed]
+ *             0 1  2  3 
+ *    codePoint = "B"       [the unassigned codepoint] 
+ * 
+ * Now, suppose a callback wants to write the substitution character '?' to
+ * the target. It calls ucnv_cbFromUWriteBytes() to write the ?. 
+ * It should pass ZERO as the offset, because the offset as far as the 
+ * callback is concerned is relative to the SOURCE pointer [which points 
+ * before 'C'.]  If the callback goes into the args and consumes 'C' also,
+ * it would call FromUWriteBytes with an offset of 1 (and advance the source
+ * pointer).
+ *
+ */
+
+#ifndef UCNV_CB_H
+#define UCNV_CB_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ucnv.h"
+#include "unicode/ucnv_err.h"
+
+/**
+ * ONLY used by FromU callback functions.
+ * Writes out the specified byte output bytes to the target byte buffer or to converter internal buffers.
+ *
+ * @param args callback fromUnicode arguments
+ * @param source source bytes to write
+ * @param length length of bytes to write
+ * @param offsetIndex the relative offset index from callback.
+ * @param err error status. If <TT>U_BUFFER_OVERFLOW</TT> is returned, then U_BUFFER_OVERFLOW <STRONG>must</STRONG> 
+ * be returned to the user, because it means that not all data could be written into the target buffer, and some is 
+ * in the converter error buffer.
+ * @see ucnv_cbFromUWriteSub
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ucnv_cbFromUWriteBytes (UConverterFromUnicodeArgs *args,
+                        const char* source,
+                        int32_t length,
+                        int32_t offsetIndex,
+                        UErrorCode * err);
+
+/**
+ * ONLY used by FromU callback functions.  
+ * This function will write out the correct substitution character sequence 
+ * to the target.
+ *
+ * @param args callback fromUnicode arguments
+ * @param offsetIndex the relative offset index from the current source pointer to be used
+ * @param err error status. If <TT>U_BUFFER_OVERFLOW</TT> is returned, then U_BUFFER_OVERFLOW <STRONG>must</STRONG> 
+ * be returned to the user, because it means that not all data could be written into the target buffer, and some is 
+ * in the converter error buffer.
+ * @see ucnv_cbFromUWriteBytes
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 
+ucnv_cbFromUWriteSub (UConverterFromUnicodeArgs *args,
+                      int32_t offsetIndex,
+                      UErrorCode * err);
+
+/**
+ * ONLY used by fromU callback functions.  
+ * This function will write out the error character(s) to the target UChar buffer.
+ *
+ * @param args callback fromUnicode arguments
+ * @param source pointer to pointer to first UChar to write [on exit: 1 after last UChar processed]
+ * @param sourceLimit pointer after last UChar to write
+ * @param offsetIndex the relative offset index from callback which will be set
+ * @param err error status <TT>U_BUFFER_OVERFLOW</TT>
+ * @see ucnv_cbToUWriteSub
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 ucnv_cbFromUWriteUChars(UConverterFromUnicodeArgs *args,
+                             const UChar** source,
+                             const UChar*  sourceLimit,
+                             int32_t offsetIndex,
+                             UErrorCode * err);
+
+/**
+ * ONLY used by ToU callback functions.
+ *  This function will write out the specified characters to the target 
+ * UChar buffer.
+ *
+ * @param args callback toUnicode arguments
+ * @param source source string to write
+ * @param length the length of source string
+ * @param offsetIndex the relative offset index which will be written.
+ * @param err error status <TT>U_BUFFER_OVERFLOW</TT>
+ * @see ucnv_cbToUWriteSub
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 ucnv_cbToUWriteUChars (UConverterToUnicodeArgs *args,
+                                             const UChar* source,
+                                             int32_t length,
+                                             int32_t offsetIndex,
+                                             UErrorCode * err);
+
+/**
+ * ONLY used by ToU  callback functions.  
+ * This function will write out the Unicode substitution character (U+FFFD).
+ *
+ * @param args callback fromUnicode arguments
+ * @param offsetIndex the relative offset index from callback.
+ * @param err error status <TT>U_BUFFER_OVERFLOW</TT>
+ * @see ucnv_cbToUWriteUChars
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 ucnv_cbToUWriteSub (UConverterToUnicodeArgs *args,
+                       int32_t offsetIndex,
+                       UErrorCode * err);
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucnv_err.h b/thirdparty/icu4c/common/unicode/ucnv_err.h
new file mode 100644
index 0000000000..7209ba5f7b
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucnv_err.h
@@ -0,0 +1,465 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2009, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+ *
+ *
+ *   ucnv_err.h:
+ */
+
+/**
+ * \file
+ * \brief C UConverter predefined error callbacks
+ *
+ *  <h2>Error Behaviour Functions</h2>
+ *  Defines some error behaviour functions called by ucnv_{from,to}Unicode
+ *  These are provided as part of ICU and many are stable, but they
+ *  can also be considered only as an example of what can be done with
+ *  callbacks.  You may of course write your own.
+ *
+ *  If you want to write your own, you may also find the functions from
+ *  ucnv_cb.h useful when writing your own callbacks.
+ *
+ *  These functions, although public, should NEVER be called directly.
+ *  They should be used as parameters to the ucnv_setFromUCallback
+ *  and ucnv_setToUCallback functions, to set the behaviour of a converter
+ *  when it encounters ILLEGAL/UNMAPPED/INVALID sequences.
+ *
+ *  usage example:  'STOP' doesn't need any context, but newContext
+ *    could be set to something other than 'NULL' if needed. The available
+ *    contexts in this header can modify the default behavior of the callback.
+ *
+ *  \code
+ *  UErrorCode err = U_ZERO_ERROR;
+ *  UConverter *myConverter = ucnv_open("ibm-949", &err);
+ *  const void *oldContext;
+ *  UConverterFromUCallback oldAction;
+ *
+ *
+ *  if (U_SUCCESS(err))
+ *  {
+ *      ucnv_setFromUCallBack(myConverter,
+ *                       UCNV_FROM_U_CALLBACK_STOP,
+ *                       NULL,
+ *                       &oldAction,
+ *                       &oldContext,
+ *                       &status);
+ *  }
+ *  \endcode
+ *
+ *  The code above tells "myConverter" to stop when it encounters an
+ *  ILLEGAL/TRUNCATED/INVALID sequences when it is used to convert from
+ *  Unicode -> Codepage. The behavior from Codepage to Unicode is not changed,
+ *  and ucnv_setToUCallBack would need to be called in order to change
+ *  that behavior too.
+ *
+ *  Here is an example with a context:
+ *
+ *  \code
+ *  UErrorCode err = U_ZERO_ERROR;
+ *  UConverter *myConverter = ucnv_open("ibm-949", &err);
+ *  const void *oldContext;
+ *  UConverterFromUCallback oldAction;
+ *
+ *
+ *  if (U_SUCCESS(err))
+ *  {
+ *      ucnv_setToUCallBack(myConverter,
+ *                       UCNV_TO_U_CALLBACK_SUBSTITUTE,
+ *                       UCNV_SUB_STOP_ON_ILLEGAL,
+ *                       &oldAction,
+ *                       &oldContext,
+ *                       &status);
+ *  }
+ *  \endcode
+ *
+ *  The code above tells "myConverter" to stop when it encounters an
+ *  ILLEGAL/TRUNCATED/INVALID sequences when it is used to convert from
+ *  Codepage -> Unicode. Any unmapped and legal characters will be
+ *  substituted to be the default substitution character.
+ */
+
+#ifndef UCNV_ERR_H
+#define UCNV_ERR_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+/** Forward declaring the UConverter structure. @stable ICU 2.0 */
+struct UConverter;
+
+/** @stable ICU 2.0 */
+typedef struct UConverter UConverter;
+
+/**
+ * FROM_U, TO_U context options for sub callback
+ * @stable ICU 2.0
+ */
+#define UCNV_SUB_STOP_ON_ILLEGAL "i"
+
+/**
+ * FROM_U, TO_U context options for skip callback
+ * @stable ICU 2.0
+ */
+#define UCNV_SKIP_STOP_ON_ILLEGAL "i"
+
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to ICU (%UXXXX) 
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_ICU       NULL
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to JAVA (\\uXXXX)
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_JAVA      "J"
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to C (\\uXXXX \\UXXXXXXXX)
+ * TO_U_CALLBACK_ESCAPE option to escape the character value according to C (\\xXXXX)
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_C         "C"
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to XML Decimal escape \htmlonly(&amp;#DDDD;)\endhtmlonly
+ * TO_U_CALLBACK_ESCAPE context option to escape the character value according to XML Decimal escape \htmlonly(&amp;#DDDD;)\endhtmlonly
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_XML_DEC   "D"
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to XML Hex escape \htmlonly(&amp;#xXXXX;)\endhtmlonly
+ * TO_U_CALLBACK_ESCAPE context option to escape the character value according to XML Hex escape \htmlonly(&amp;#xXXXX;)\endhtmlonly
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_XML_HEX   "X"
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to Unicode (U+XXXXX)
+ * @stable ICU 2.0
+ */
+#define UCNV_ESCAPE_UNICODE   "U"
+
+/**
+ * FROM_U_CALLBACK_ESCAPE context option to escape the code unit according to CSS2 conventions (\\HH..H<space>, that is,
+ * a backslash, 1..6 hex digits, and a space)
+ * @stable ICU 4.0
+ */
+#define UCNV_ESCAPE_CSS2   "S"
+
+/** 
+ * The process condition code to be used with the callbacks.  
+ * Codes which are greater than UCNV_IRREGULAR should be 
+ * passed on to any chained callbacks.
+ * @stable ICU 2.0
+ */
+typedef enum {
+    UCNV_UNASSIGNED = 0,  /**< The code point is unassigned.
+                             The error code U_INVALID_CHAR_FOUND will be set. */
+    UCNV_ILLEGAL = 1,     /**< The code point is illegal. For example, 
+                             \\x81\\x2E is illegal in SJIS because \\x2E
+                             is not a valid trail byte for the \\x81 
+                             lead byte.
+                             Also, starting with Unicode 3.0.1, non-shortest byte sequences
+                             in UTF-8 (like \\xC1\\xA1 instead of \\x61 for U+0061)
+                             are also illegal, not just irregular.
+                             The error code U_ILLEGAL_CHAR_FOUND will be set. */
+    UCNV_IRREGULAR = 2,   /**< The codepoint is not a regular sequence in 
+                             the encoding. For example, \\xED\\xA0\\x80..\\xED\\xBF\\xBF
+                             are irregular UTF-8 byte sequences for single surrogate
+                             code points.
+                             The error code U_INVALID_CHAR_FOUND will be set. */
+    UCNV_RESET = 3,       /**< The callback is called with this reason when a
+                             'reset' has occurred. Callback should reset all
+                             state. */
+    UCNV_CLOSE = 4,        /**< Called when the converter is closed. The
+                             callback should release any allocated memory.*/
+    UCNV_CLONE = 5         /**< Called when ucnv_safeClone() is called on the
+                              converter. the pointer available as the
+                              'context' is an alias to the original converters'
+                              context pointer. If the context must be owned
+                              by the new converter, the callback must clone 
+                              the data and call ucnv_setFromUCallback 
+                              (or setToUCallback) with the correct pointer.
+                              @stable ICU 2.2
+                           */
+} UConverterCallbackReason;
+
+
+/**
+ * The structure for the fromUnicode callback function parameter.
+ * @stable ICU 2.0
+ */
+typedef struct {
+    uint16_t size;              /**< The size of this struct. @stable ICU 2.0 */
+    UBool flush;                /**< The internal state of converter will be reset and data flushed if set to true. @stable ICU 2.0    */
+    UConverter *converter;      /**< Pointer to the converter that is opened and to which this struct is passed as an argument. @stable ICU 2.0  */
+    const UChar *source;        /**< Pointer to the source source buffer. @stable ICU 2.0    */
+    const UChar *sourceLimit;   /**< Pointer to the limit (end + 1) of source buffer. @stable ICU 2.0    */
+    char *target;               /**< Pointer to the target buffer. @stable ICU 2.0    */
+    const char *targetLimit;    /**< Pointer to the limit (end + 1) of target buffer. @stable ICU 2.0     */
+    int32_t *offsets;           /**< Pointer to the buffer that receives the offsets. *offset = blah ; offset++;. @stable ICU 2.0  */
+} UConverterFromUnicodeArgs;
+
+
+/**
+ * The structure for the toUnicode callback function parameter.
+ * @stable ICU 2.0
+ */
+typedef struct {
+    uint16_t size;              /**< The size of this struct   @stable ICU 2.0 */
+    UBool flush;                /**< The internal state of converter will be reset and data flushed if set to true. @stable ICU 2.0   */
+    UConverter *converter;      /**< Pointer to the converter that is opened and to which this struct is passed as an argument. @stable ICU 2.0 */
+    const char *source;         /**< Pointer to the source source buffer. @stable ICU 2.0    */
+    const char *sourceLimit;    /**< Pointer to the limit (end + 1) of source buffer. @stable ICU 2.0    */
+    UChar *target;              /**< Pointer to the target buffer. @stable ICU 2.0    */
+    const UChar *targetLimit;   /**< Pointer to the limit (end + 1) of target buffer. @stable ICU 2.0     */
+    int32_t *offsets;           /**< Pointer to the buffer that receives the offsets. *offset = blah ; offset++;. @stable ICU 2.0  */
+} UConverterToUnicodeArgs;
+
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This From Unicode callback STOPS at the ILLEGAL_SEQUENCE,
+ * returning the error code back to the caller immediately.
+ *
+ * @param context Pointer to the callback's private data
+ * @param fromUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
+ * @param reason Defines the reason the callback was invoked
+ * @param err This should always be set to a failure status prior to calling.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_FROM_U_CALLBACK_STOP (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This To Unicode callback STOPS at the ILLEGAL_SEQUENCE,
+ * returning the error code back to the caller immediately.
+ *
+ * @param context Pointer to the callback's private data
+ * @param toUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param reason Defines the reason the callback was invoked
+ * @param err This should always be set to a failure status prior to calling.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_TO_U_CALLBACK_STOP (
+                  const void *context,
+                  UConverterToUnicodeArgs *toUArgs,
+                  const char* codeUnits,
+                  int32_t length,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This From Unicode callback skips any ILLEGAL_SEQUENCE, or
+ * skips only UNASSINGED_SEQUENCE depending on the context parameter
+ * simply ignoring those characters. 
+ *
+ * @param context  The function currently recognizes the callback options:
+ *                 UCNV_SKIP_STOP_ON_ILLEGAL: STOPS at the ILLEGAL_SEQUENCE,
+ *                      returning the error code back to the caller immediately.
+ *                 NULL: Skips any ILLEGAL_SEQUENCE
+ * @param fromUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_FROM_U_CALLBACK_SKIP (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This From Unicode callback will Substitute the ILLEGAL SEQUENCE, or 
+ * UNASSIGNED_SEQUENCE depending on context parameter, with the
+ * current substitution string for the converter. This is the default
+ * callback.
+ *
+ * @param context The function currently recognizes the callback options:
+ *                 UCNV_SUB_STOP_ON_ILLEGAL: STOPS at the ILLEGAL_SEQUENCE,
+ *                      returning the error code back to the caller immediately.
+ *                 NULL: Substitutes any ILLEGAL_SEQUENCE
+ * @param fromUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @see ucnv_setSubstChars
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_FROM_U_CALLBACK_SUBSTITUTE (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This From Unicode callback will Substitute the ILLEGAL SEQUENCE with the
+ * hexadecimal representation of the illegal codepoints
+ *
+ * @param context The function currently recognizes the callback options:
+ *        <ul>
+ *        <li>UCNV_ESCAPE_ICU: Substitues the  ILLEGAL SEQUENCE with the hexadecimal 
+ *          representation in the format  %UXXXX, e.g. "%uFFFE%u00AC%uC8FE"). 
+ *          In the Event the converter doesn't support the characters {%,U}[A-F][0-9], 
+ *          it will  substitute  the illegal sequence with the substitution characters.
+ *          Note that  codeUnit(32bit int eg: unit of a surrogate pair) is represented as
+ *          %UD84D%UDC56</li>
+ *        <li>UCNV_ESCAPE_JAVA: Substitues the  ILLEGAL SEQUENCE with the hexadecimal 
+ *          representation in the format  \\uXXXX, e.g. "\\uFFFE\\u00AC\\uC8FE"). 
+ *          In the Event the converter doesn't support the characters {\,u}[A-F][0-9], 
+ *          it will  substitute  the illegal sequence with the substitution characters.
+ *          Note that  codeUnit(32bit int eg: unit of a surrogate pair) is represented as
+ *          \\uD84D\\uDC56</li>
+ *        <li>UCNV_ESCAPE_C: Substitues the  ILLEGAL SEQUENCE with the hexadecimal 
+ *          representation in the format  \\uXXXX, e.g. "\\uFFFE\\u00AC\\uC8FE"). 
+ *          In the Event the converter doesn't support the characters {\,u,U}[A-F][0-9], 
+ *          it will  substitute  the illegal sequence with the substitution characters.
+ *          Note that  codeUnit(32bit int eg: unit of a surrogate pair) is represented as
+ *          \\U00023456</li>
+ *        <li>UCNV_ESCAPE_XML_DEC: Substitues the  ILLEGAL SEQUENCE with the decimal 
+ *          representation in the format \htmlonly&amp;#DDDDDDDD;, e.g. "&amp;#65534;&amp;#172;&amp;#51454;")\endhtmlonly. 
+ *          In the Event the converter doesn't support the characters {&amp;,#}[0-9], 
+ *          it will  substitute  the illegal sequence with the substitution characters.
+ *          Note that  codeUnit(32bit int eg: unit of a surrogate pair) is represented as
+ *          &amp;#144470; and Zero padding is ignored.</li>
+ *        <li>UCNV_ESCAPE_XML_HEX:Substitues the  ILLEGAL SEQUENCE with the decimal 
+ *          representation in the format \htmlonly&amp;#xXXXX; e.g. "&amp;#xFFFE;&amp;#x00AC;&amp;#xC8FE;")\endhtmlonly. 
+ *          In the Event the converter doesn't support the characters {&,#,x}[0-9], 
+ *          it will  substitute  the illegal sequence with the substitution characters.
+ *          Note that  codeUnit(32bit int eg: unit of a surrogate pair) is represented as
+ *          \htmlonly&amp;#x23456;\endhtmlonly</li>
+ *        </ul>
+ * @param fromUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_FROM_U_CALLBACK_ESCAPE (
+                  const void *context,
+                  UConverterFromUnicodeArgs *fromUArgs,
+                  const UChar* codeUnits,
+                  int32_t length,
+                  UChar32 codePoint,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This To Unicode callback skips any ILLEGAL_SEQUENCE, or
+ * skips only UNASSINGED_SEQUENCE depending on the context parameter
+ * simply ignoring those characters. 
+ *
+ * @param context  The function currently recognizes the callback options:
+ *                 UCNV_SKIP_STOP_ON_ILLEGAL: STOPS at the ILLEGAL_SEQUENCE,
+ *                      returning the error code back to the caller immediately.
+ *                 NULL: Skips any ILLEGAL_SEQUENCE
+ * @param toUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_TO_U_CALLBACK_SKIP (
+                  const void *context,
+                  UConverterToUnicodeArgs *toUArgs,
+                  const char* codeUnits,
+                  int32_t length,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This To Unicode callback will Substitute the ILLEGAL SEQUENCE,or 
+ * UNASSIGNED_SEQUENCE depending on context parameter,  with the
+ * Unicode substitution character, U+FFFD.
+ *
+ * @param context  The function currently recognizes the callback options:
+ *                 UCNV_SUB_STOP_ON_ILLEGAL: STOPS at the ILLEGAL_SEQUENCE,
+ *                      returning the error code back to the caller immediately.
+ *                 NULL: Substitutes any ILLEGAL_SEQUENCE
+ * @param toUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2 UCNV_TO_U_CALLBACK_SUBSTITUTE (
+                  const void *context,
+                  UConverterToUnicodeArgs *toUArgs,
+                  const char* codeUnits,
+                  int32_t length,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+/**
+ * DO NOT CALL THIS FUNCTION DIRECTLY!
+ * This To Unicode callback will Substitute the ILLEGAL SEQUENCE with the
+ * hexadecimal representation of the illegal bytes
+ *  (in the format  %XNN, e.g. "%XFF%X0A%XC8%X03").
+ *
+ * @param context This function currently recognizes the callback options:
+ *      UCNV_ESCAPE_ICU, UCNV_ESCAPE_JAVA, UCNV_ESCAPE_C, UCNV_ESCAPE_XML_DEC,
+ *      UCNV_ESCAPE_XML_HEX and UCNV_ESCAPE_UNICODE.
+ * @param toUArgs Information about the conversion in progress
+ * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
+ * @param length Size (in bytes) of the concerned codepage sequence
+ * @param reason Defines the reason the callback was invoked
+ * @param err Return value will be set to success if the callback was handled,
+ *      otherwise this value will be set to a failure status.
+ * @stable ICU 2.0
+ */
+
+U_CAPI void U_EXPORT2 UCNV_TO_U_CALLBACK_ESCAPE (
+                  const void *context,
+                  UConverterToUnicodeArgs *toUArgs,
+                  const char* codeUnits,
+                  int32_t length,
+                  UConverterCallbackReason reason,
+                  UErrorCode * err);
+
+#endif
+
+#endif
+
+/*UCNV_ERR_H*/ 
diff --git a/thirdparty/icu4c/common/unicode/ucnvsel.h b/thirdparty/icu4c/common/unicode/ucnvsel.h
new file mode 100644
index 0000000000..5e0a71cf35
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucnvsel.h
@@ -0,0 +1,192 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2008-2011, International Business Machines
+*   Corporation, Google and others.  All Rights Reserved.
+*
+*******************************************************************************
+*/
+/*
+ * Author : eldawy@google.com (Mohamed Eldawy)
+ * ucnvsel.h
+ *
+ * Purpose: To generate a list of encodings capable of handling
+ * a given Unicode text
+ *
+ * Started 09-April-2008
+ */
+
+#ifndef __ICU_UCNV_SEL_H__
+#define __ICU_UCNV_SEL_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/uset.h"
+#include "unicode/utf16.h"
+#include "unicode/uenum.h"
+#include "unicode/ucnv.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ *
+ * A converter selector is built with a set of encoding/charset names
+ * and given an input string returns the set of names of the
+ * corresponding converters which can convert the string.
+ *
+ * A converter selector can be serialized into a buffer and reopened
+ * from the serialized form.
+ */
+
+/**
+ * @{
+ * The selector data structure
+ */
+struct UConverterSelector;
+typedef struct UConverterSelector UConverterSelector;
+/** @} */
+
+/**
+ * Open a selector.
+ * If converterListSize is 0, build for all available converters.
+ * If excludedCodePoints is NULL, don't exclude any code points.
+ *
+ * @param converterList a pointer to encoding names needed to be involved. 
+ *                      Can be NULL if converterListSize==0.
+ *                      The list and the names will be cloned, and the caller
+ *                      retains ownership of the original.
+ * @param converterListSize number of encodings in above list.
+ *                          If 0, builds a selector for all available converters.
+ * @param excludedCodePoints a set of code points to be excluded from consideration.
+ *                           That is, excluded code points in a string do not change
+ *                           the selection result. (They might be handled by a callback.)
+ *                           Use NULL to exclude nothing.
+ * @param whichSet what converter set to use? Use this to determine whether
+ *                 to consider only roundtrip mappings or also fallbacks.
+ * @param status an in/out ICU UErrorCode
+ * @return the new selector
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI UConverterSelector* U_EXPORT2
+ucnvsel_open(const char* const*  converterList, int32_t converterListSize,
+             const USet* excludedCodePoints,
+             const UConverterUnicodeSet whichSet, UErrorCode* status);
+
+/**
+ * Closes a selector.
+ * If any Enumerations were returned by ucnv_select*, they become invalid.
+ * They can be closed before or after calling ucnv_closeSelector,
+ * but should never be used after the selector is closed.
+ *
+ * @see ucnv_selectForString
+ * @see ucnv_selectForUTF8
+ *
+ * @param sel selector to close
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI void U_EXPORT2
+ucnvsel_close(UConverterSelector *sel);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUConverterSelectorPointer
+ * "Smart pointer" class, closes a UConverterSelector via ucnvsel_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUConverterSelectorPointer, UConverterSelector, ucnvsel_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Open a selector from its serialized form.
+ * The buffer must remain valid and unchanged for the lifetime of the selector.
+ * This is much faster than creating a selector from scratch.
+ * Using a serialized form from a different machine (endianness/charset) is supported.
+ *
+ * @param buffer pointer to the serialized form of a converter selector;
+ *               must be 32-bit-aligned
+ * @param length the capacity of this buffer (can be equal to or larger than
+ *               the actual data length)
+ * @param status an in/out ICU UErrorCode
+ * @return the new selector
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI UConverterSelector* U_EXPORT2
+ucnvsel_openFromSerialized(const void* buffer, int32_t length, UErrorCode* status);
+
+/**
+ * Serialize a selector into a linear buffer.
+ * The serialized form is portable to different machines.
+ *
+ * @param sel selector to consider
+ * @param buffer pointer to 32-bit-aligned memory to be filled with the
+ *               serialized form of this converter selector
+ * @param bufferCapacity the capacity of this buffer
+ * @param status an in/out ICU UErrorCode
+ * @return the required buffer capacity to hold serialize data (even if the call fails
+ *         with a U_BUFFER_OVERFLOW_ERROR, it will return the required capacity)
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI int32_t U_EXPORT2
+ucnvsel_serialize(const UConverterSelector* sel,
+                  void* buffer, int32_t bufferCapacity, UErrorCode* status);
+
+/**
+ * Select converters that can map all characters in a UTF-16 string,
+ * ignoring the excluded code points.
+ *
+ * @param sel a selector
+ * @param s UTF-16 string
+ * @param length length of the string, or -1 if NUL-terminated
+ * @param status an in/out ICU UErrorCode
+ * @return an enumeration containing encoding names.
+ *         The returned encoding names and their order will be the same as
+ *         supplied when building the selector.
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI UEnumeration * U_EXPORT2
+ucnvsel_selectForString(const UConverterSelector* sel,
+                        const UChar *s, int32_t length, UErrorCode *status);
+
+/**
+ * Select converters that can map all characters in a UTF-8 string,
+ * ignoring the excluded code points.
+ *
+ * @param sel a selector
+ * @param s UTF-8 string
+ * @param length length of the string, or -1 if NUL-terminated
+ * @param status an in/out ICU UErrorCode
+ * @return an enumeration containing encoding names.
+ *         The returned encoding names and their order will be the same as
+ *         supplied when building the selector.
+ *
+ * @stable ICU 4.2
+ */
+U_CAPI UEnumeration * U_EXPORT2
+ucnvsel_selectForUTF8(const UConverterSelector* sel,
+                      const char *s, int32_t length, UErrorCode *status);
+
+#endif  /* !UCONFIG_NO_CONVERSION */
+
+#endif  /* __ICU_UCNV_SEL_H__ */
diff --git a/thirdparty/icu4c/common/unicode/uconfig.h b/thirdparty/icu4c/common/unicode/uconfig.h
new file mode 100644
index 0000000000..bbc232d1ed
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uconfig.h
@@ -0,0 +1,456 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  uconfig.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002sep19
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UCONFIG_H__
+#define __UCONFIG_H__
+
+
+/*!
+ * \file
+ * \brief User-configurable settings
+ *
+ * Miscellaneous switches:
+ *
+ * A number of macros affect a variety of minor aspects of ICU.
+ * Most of them used to be defined elsewhere (e.g., in utypes.h or platform.h)
+ * and moved here to make them easier to find.
+ *
+ * Switches for excluding parts of ICU library code modules:
+ *
+ * Changing these macros allows building partial, smaller libraries for special purposes.
+ * By default, all modules are built.
+ * The switches are fairly coarse, controlling large modules.
+ * Basic services cannot be turned off.
+ *
+ * Building with any of these options does not guarantee that the
+ * ICU build process will completely work. It is recommended that
+ * the ICU libraries and data be built using the normal build.
+ * At that time you should remove the data used by those services.
+ * After building the ICU data library, you should rebuild the ICU
+ * libraries with these switches customized to your needs.
+ *
+ * @stable ICU 2.4
+ */
+
+/**
+ * If this switch is defined, ICU will attempt to load a header file named "uconfig_local.h"
+ * prior to determining default settings for uconfig variables.
+ *
+ * @internal ICU 4.0
+ */
+#if defined(UCONFIG_USE_LOCAL)
+#include "uconfig_local.h"
+#endif
+
+/**
+ * \def U_DEBUG
+ * Determines whether to include debugging code.
+ * Automatically set on Windows, but most compilers do not have
+ * related predefined macros.
+ * @internal
+ */
+#ifdef U_DEBUG
+    /* Use the predefined value. */
+#elif defined(_DEBUG)
+    /*
+     * _DEBUG is defined by Visual Studio debug compilation.
+     * Do *not* test for its NDEBUG macro: It is an orthogonal macro
+     * which disables assert().
+     */
+#   define U_DEBUG 1
+# else
+#   define U_DEBUG 0
+#endif
+
+/**
+ * Determines whether to enable auto cleanup of libraries. 
+ * @internal
+ */
+#ifndef UCLN_NO_AUTO_CLEANUP
+#define UCLN_NO_AUTO_CLEANUP 1
+#endif
+
+/**
+ * \def U_DISABLE_RENAMING
+ * Determines whether to disable renaming or not.
+ * @internal
+ */
+#ifndef U_DISABLE_RENAMING
+#define U_DISABLE_RENAMING 0
+#endif
+
+/**
+ * \def U_NO_DEFAULT_INCLUDE_UTF_HEADERS
+ * Determines whether utypes.h includes utf.h, utf8.h, utf16.h and utf_old.h.
+ * utypes.h includes those headers if this macro is defined to 0.
+ * Otherwise, each those headers must be included explicitly when using one of their macros.
+ * Defaults to 0 for backward compatibility, except inside ICU.
+ * @stable ICU 49
+ */
+#ifdef U_NO_DEFAULT_INCLUDE_UTF_HEADERS
+    /* Use the predefined value. */
+#elif defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || \
+      defined(U_IO_IMPLEMENTATION) || defined(U_LAYOUT_IMPLEMENTATION) || defined(U_LAYOUTEX_IMPLEMENTATION) || \
+      defined(U_TOOLUTIL_IMPLEMENTATION)
+#   define U_NO_DEFAULT_INCLUDE_UTF_HEADERS 1
+#else
+#   define U_NO_DEFAULT_INCLUDE_UTF_HEADERS 0
+#endif
+
+/**
+ * \def U_OVERRIDE_CXX_ALLOCATION
+ * Determines whether to override new and delete.
+ * ICU is normally built such that all of its C++ classes, via their UMemory base,
+ * override operators new and delete to use its internal, customizable,
+ * non-exception-throwing memory allocation functions. (Default value 1 for this macro.)
+ *
+ * This is especially important when the application and its libraries use multiple heaps.
+ * For example, on Windows, this allows the ICU DLL to be used by
+ * applications that statically link the C Runtime library.
+ *
+ * @stable ICU 2.2
+ */
+#ifndef U_OVERRIDE_CXX_ALLOCATION
+#define U_OVERRIDE_CXX_ALLOCATION 1
+#endif
+
+/**
+ * \def U_ENABLE_TRACING
+ * Determines whether to enable tracing.
+ * @internal
+ */
+#ifndef U_ENABLE_TRACING
+#define U_ENABLE_TRACING 0
+#endif
+
+/**
+ * \def UCONFIG_ENABLE_PLUGINS
+ * Determines whether to enable ICU plugins.
+ * @internal
+ */
+#ifndef UCONFIG_ENABLE_PLUGINS
+#define UCONFIG_ENABLE_PLUGINS 0
+#endif
+
+/**
+ * \def U_ENABLE_DYLOAD
+ * Whether to enable Dynamic loading in ICU.
+ * @internal
+ */
+#ifndef U_ENABLE_DYLOAD
+#define U_ENABLE_DYLOAD 1
+#endif
+
+/**
+ * \def U_CHECK_DYLOAD
+ * Whether to test Dynamic loading as an OS capability.
+ * @internal
+ */
+#ifndef U_CHECK_DYLOAD
+#define U_CHECK_DYLOAD 1
+#endif
+
+/**
+ * \def U_DEFAULT_SHOW_DRAFT
+ * Do we allow ICU users to use the draft APIs by default?
+ * @internal
+ */
+#ifndef U_DEFAULT_SHOW_DRAFT
+#define U_DEFAULT_SHOW_DRAFT 1
+#endif
+
+/*===========================================================================*/
+/* Custom icu entry point renaming                                           */
+/*===========================================================================*/
+
+/**
+ * \def U_HAVE_LIB_SUFFIX
+ * 1 if a custom library suffix is set.
+ * @internal
+ */
+#ifdef U_HAVE_LIB_SUFFIX
+    /* Use the predefined value. */
+#elif defined(U_LIB_SUFFIX_C_NAME) || defined(U_IN_DOXYGEN)
+#   define U_HAVE_LIB_SUFFIX 1
+#endif
+
+/**
+ * \def U_LIB_SUFFIX_C_NAME_STRING
+ * Defines the library suffix as a string with C syntax.
+ * @internal
+ */
+#ifdef U_LIB_SUFFIX_C_NAME_STRING
+    /* Use the predefined value. */
+#elif defined(U_LIB_SUFFIX_C_NAME)
+#   define CONVERT_TO_STRING(s) #s
+#   define U_LIB_SUFFIX_C_NAME_STRING CONVERT_TO_STRING(U_LIB_SUFFIX_C_NAME)
+#else
+#   define U_LIB_SUFFIX_C_NAME_STRING ""
+#endif
+
+/* common/i18n library switches --------------------------------------------- */
+
+/**
+ * \def UCONFIG_ONLY_COLLATION
+ * This switch turns off modules that are not needed for collation.
+ *
+ * It does not turn off legacy conversion because that is necessary
+ * for ICU to work on EBCDIC platforms (for the default converter).
+ * If you want "only collation" and do not build for EBCDIC,
+ * then you can define UCONFIG_NO_CONVERSION or UCONFIG_NO_LEGACY_CONVERSION to 1 as well.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_ONLY_COLLATION
+#   define UCONFIG_ONLY_COLLATION 0
+#endif
+
+#if UCONFIG_ONLY_COLLATION
+    /* common library */
+#   define UCONFIG_NO_BREAK_ITERATION 1
+#   define UCONFIG_NO_IDNA 1
+
+    /* i18n library */
+#   if UCONFIG_NO_COLLATION
+#       error Contradictory collation switches in uconfig.h.
+#   endif
+#   define UCONFIG_NO_FORMATTING 1
+#   define UCONFIG_NO_TRANSLITERATION 1
+#   define UCONFIG_NO_REGULAR_EXPRESSIONS 1
+#endif
+
+/* common library switches -------------------------------------------------- */
+
+/**
+ * \def UCONFIG_NO_FILE_IO
+ * This switch turns off all file access in the common library
+ * where file access is only used for data loading.
+ * ICU data must then be provided in the form of a data DLL (or with an
+ * equivalent way to link to the data residing in an executable,
+ * as in building a combined library with both the common library's code and
+ * the data), or via udata_setCommonData().
+ * Application data must be provided via udata_setAppData() or by using
+ * "open" functions that take pointers to data, for example ucol_openBinary().
+ *
+ * File access is not used at all in the i18n library.
+ *
+ * File access cannot be turned off for the icuio library or for the ICU
+ * test suites and ICU tools.
+ *
+ * @stable ICU 3.6
+ */
+#ifndef UCONFIG_NO_FILE_IO
+#   define UCONFIG_NO_FILE_IO 0
+#endif
+
+#if UCONFIG_NO_FILE_IO && defined(U_TIMEZONE_FILES_DIR) 
+#   error Contradictory file io switches in uconfig.h.
+#endif
+
+/**
+ * \def UCONFIG_NO_CONVERSION
+ * ICU will not completely build (compiling the tools fails) with this
+ * switch turned on.
+ * This switch turns off all converters.
+ *
+ * You may want to use this together with U_CHARSET_IS_UTF8 defined to 1
+ * in utypes.h if char* strings in your environment are always in UTF-8.
+ *
+ * @stable ICU 3.2
+ * @see U_CHARSET_IS_UTF8
+ */
+#ifndef UCONFIG_NO_CONVERSION
+#   define UCONFIG_NO_CONVERSION 0
+#endif
+
+#if UCONFIG_NO_CONVERSION
+#   define UCONFIG_NO_LEGACY_CONVERSION 1
+#endif
+
+/**
+ * \def UCONFIG_ONLY_HTML_CONVERSION
+ * This switch turns off all of the converters NOT listed in
+ * the HTML encoding standard:
+ * http://www.w3.org/TR/encoding/#names-and-labels
+ *
+ * This is not possible on EBCDIC platforms
+ * because they need ibm-37 or ibm-1047 default converters.
+ *
+ * @stable ICU 55
+ */
+#ifndef UCONFIG_ONLY_HTML_CONVERSION
+#   define UCONFIG_ONLY_HTML_CONVERSION 0
+#endif
+
+/**
+ * \def UCONFIG_NO_LEGACY_CONVERSION
+ * This switch turns off all converters except for
+ * - Unicode charsets (UTF-7/8/16/32, CESU-8, SCSU, BOCU-1)
+ * - US-ASCII
+ * - ISO-8859-1
+ *
+ * Turning off legacy conversion is not possible on EBCDIC platforms
+ * because they need ibm-37 or ibm-1047 default converters.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_LEGACY_CONVERSION
+#   define UCONFIG_NO_LEGACY_CONVERSION 0
+#endif
+
+/**
+ * \def UCONFIG_NO_NORMALIZATION
+ * This switch turns off normalization.
+ * It implies turning off several other services as well, for example
+ * collation and IDNA.
+ *
+ * @stable ICU 2.6
+ */
+#ifndef UCONFIG_NO_NORMALIZATION
+#   define UCONFIG_NO_NORMALIZATION 0
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+    /* common library */
+    /* ICU 50 CJK dictionary BreakIterator uses normalization */
+#   define UCONFIG_NO_BREAK_ITERATION 1
+    /* IDNA (UTS #46) is implemented via normalization */
+#   define UCONFIG_NO_IDNA 1
+
+    /* i18n library */
+#   if UCONFIG_ONLY_COLLATION
+#       error Contradictory collation switches in uconfig.h.
+#   endif
+#   define UCONFIG_NO_COLLATION 1
+#   define UCONFIG_NO_TRANSLITERATION 1
+#endif
+
+/**
+ * \def UCONFIG_NO_BREAK_ITERATION
+ * This switch turns off break iteration.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_BREAK_ITERATION
+#   define UCONFIG_NO_BREAK_ITERATION 0
+#endif
+
+/**
+ * \def UCONFIG_NO_IDNA
+ * This switch turns off IDNA.
+ *
+ * @stable ICU 2.6
+ */
+#ifndef UCONFIG_NO_IDNA
+#   define UCONFIG_NO_IDNA 0
+#endif
+
+/**
+ * \def UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE
+ * Determines the default UMessagePatternApostropheMode.
+ * See the documentation for that enum.
+ *
+ * @stable ICU 4.8
+ */
+#ifndef UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE
+#   define UCONFIG_MSGPAT_DEFAULT_APOSTROPHE_MODE UMSGPAT_APOS_DOUBLE_OPTIONAL
+#endif
+
+/**
+ * \def UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+ * On platforms where U_PLATFORM_HAS_WIN32_API is true, this switch determines
+ * if the Windows platform APIs are used for LCID<->Locale Name conversions.
+ * Otherwise, only the built-in ICU tables are used.
+ * 
+ * @internal ICU 64
+ */
+#ifndef UCONFIG_USE_WINDOWS_LCID_MAPPING_API
+#   define UCONFIG_USE_WINDOWS_LCID_MAPPING_API 1
+#endif
+
+/* i18n library switches ---------------------------------------------------- */
+
+/**
+ * \def UCONFIG_NO_COLLATION
+ * This switch turns off collation and collation-based string search.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_COLLATION
+#   define UCONFIG_NO_COLLATION 0
+#endif
+
+/**
+ * \def UCONFIG_NO_FORMATTING
+ * This switch turns off formatting and calendar/timezone services.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_FORMATTING
+#   define UCONFIG_NO_FORMATTING 0
+#endif
+
+/**
+ * \def UCONFIG_NO_TRANSLITERATION
+ * This switch turns off transliteration.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_TRANSLITERATION
+#   define UCONFIG_NO_TRANSLITERATION 0
+#endif
+
+/**
+ * \def UCONFIG_NO_REGULAR_EXPRESSIONS
+ * This switch turns off regular expressions.
+ *
+ * @stable ICU 2.4
+ */
+#ifndef UCONFIG_NO_REGULAR_EXPRESSIONS
+#   define UCONFIG_NO_REGULAR_EXPRESSIONS 0
+#endif
+
+/**
+ * \def UCONFIG_NO_SERVICE
+ * This switch turns off service registration.
+ *
+ * @stable ICU 3.2
+ */
+#ifndef UCONFIG_NO_SERVICE
+#   define UCONFIG_NO_SERVICE 0
+#endif
+
+/**
+ * \def UCONFIG_HAVE_PARSEALLINPUT
+ * This switch turns on the "parse all input" attribute. Binary incompatible.
+ *
+ * @internal
+ */
+#ifndef UCONFIG_HAVE_PARSEALLINPUT
+#   define UCONFIG_HAVE_PARSEALLINPUT 1
+#endif
+
+/**
+ * \def UCONFIG_NO_FILTERED_BREAK_ITERATION
+ * This switch turns off filtered break iteration code.
+ *
+ * @internal
+ */
+#ifndef UCONFIG_NO_FILTERED_BREAK_ITERATION
+#   define UCONFIG_NO_FILTERED_BREAK_ITERATION 0
+#endif
+
+#endif  // __UCONFIG_H__
diff --git a/thirdparty/icu4c/common/unicode/ucpmap.h b/thirdparty/icu4c/common/unicode/ucpmap.h
new file mode 100644
index 0000000000..31e1365cac
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucpmap.h
@@ -0,0 +1,159 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ucpmap.h
+// created: 2018sep03 Markus W. Scherer
+
+#ifndef __UCPMAP_H__
+#define __UCPMAP_H__
+
+#include "unicode/utypes.h"
+
+U_CDECL_BEGIN
+
+/**
+ * \file
+ *
+ * This file defines an abstract map from Unicode code points to integer values.
+ *
+ * @see UCPMap
+ * @see UCPTrie
+ * @see UMutableCPTrie
+ */
+
+/**
+ * Abstract map from Unicode code points (U+0000..U+10FFFF) to integer values.
+ *
+ * @see UCPTrie
+ * @see UMutableCPTrie
+ * @stable ICU 63
+ */
+typedef struct UCPMap UCPMap;
+
+/**
+ * Selectors for how ucpmap_getRange() etc. should report value ranges overlapping with surrogates.
+ * Most users should use UCPMAP_RANGE_NORMAL.
+ *
+ * @see ucpmap_getRange
+ * @see ucptrie_getRange
+ * @see umutablecptrie_getRange
+ * @stable ICU 63
+ */
+enum UCPMapRangeOption {
+    /**
+     * ucpmap_getRange() enumerates all same-value ranges as stored in the map.
+     * Most users should use this option.
+     * @stable ICU 63
+     */
+    UCPMAP_RANGE_NORMAL,
+    /**
+     * ucpmap_getRange() enumerates all same-value ranges as stored in the map,
+     * except that lead surrogates (U+D800..U+DBFF) are treated as having the
+     * surrogateValue, which is passed to getRange() as a separate parameter.
+     * The surrogateValue is not transformed via filter().
+     * See U_IS_LEAD(c).
+     *
+     * Most users should use UCPMAP_RANGE_NORMAL instead.
+     *
+     * This option is useful for maps that map surrogate code *units* to
+     * special values optimized for UTF-16 string processing
+     * or for special error behavior for unpaired surrogates,
+     * but those values are not to be associated with the lead surrogate code *points*.
+     * @stable ICU 63
+     */
+    UCPMAP_RANGE_FIXED_LEAD_SURROGATES,
+    /**
+     * ucpmap_getRange() enumerates all same-value ranges as stored in the map,
+     * except that all surrogates (U+D800..U+DFFF) are treated as having the
+     * surrogateValue, which is passed to getRange() as a separate parameter.
+     * The surrogateValue is not transformed via filter().
+     * See U_IS_SURROGATE(c).
+     *
+     * Most users should use UCPMAP_RANGE_NORMAL instead.
+     *
+     * This option is useful for maps that map surrogate code *units* to
+     * special values optimized for UTF-16 string processing
+     * or for special error behavior for unpaired surrogates,
+     * but those values are not to be associated with the lead surrogate code *points*.
+     * @stable ICU 63
+     */
+    UCPMAP_RANGE_FIXED_ALL_SURROGATES
+};
+#ifndef U_IN_DOXYGEN
+typedef enum UCPMapRangeOption UCPMapRangeOption;
+#endif
+
+/**
+ * Returns the value for a code point as stored in the map, with range checking.
+ * Returns an implementation-defined error value if c is not in the range 0..U+10FFFF.
+ *
+ * @param map the map
+ * @param c the code point
+ * @return the map value,
+ *         or an implementation-defined error value if the code point is not in the range 0..U+10FFFF
+ * @stable ICU 63
+ */
+U_CAPI uint32_t U_EXPORT2
+ucpmap_get(const UCPMap *map, UChar32 c);
+
+/**
+ * Callback function type: Modifies a map value.
+ * Optionally called by ucpmap_getRange()/ucptrie_getRange()/umutablecptrie_getRange().
+ * The modified value will be returned by the getRange function.
+ *
+ * Can be used to ignore some of the value bits,
+ * make a filter for one of several values,
+ * return a value index computed from the map value, etc.
+ *
+ * @param context an opaque pointer, as passed into the getRange function
+ * @param value a value from the map
+ * @return the modified value
+ * @stable ICU 63
+ */
+typedef uint32_t U_CALLCONV
+UCPMapValueFilter(const void *context, uint32_t value);
+
+/**
+ * Returns the last code point such that all those from start to there have the same value.
+ * Can be used to efficiently iterate over all same-value ranges in a map.
+ * (This is normally faster than iterating over code points and get()ting each value,
+ * but much slower than a data structure that stores ranges directly.)
+ *
+ * If the UCPMapValueFilter function pointer is not NULL, then
+ * the value to be delivered is passed through that function, and the return value is the end
+ * of the range where all values are modified to the same actual value.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * Example:
+ * \code
+ * UChar32 start = 0, end;
+ * uint32_t value;
+ * while ((end = ucpmap_getRange(map, start, UCPMAP_RANGE_NORMAL, 0,
+ *                               NULL, NULL, &value)) >= 0) {
+ *     // Work with the range start..end and its value.
+ *     start = end + 1;
+ * }
+ * \endcode
+ *
+ * @param map the map
+ * @param start range start
+ * @param option defines whether surrogates are treated normally,
+ *               or as having the surrogateValue; usually UCPMAP_RANGE_NORMAL
+ * @param surrogateValue value for surrogates; ignored if option==UCPMAP_RANGE_NORMAL
+ * @param filter a pointer to a function that may modify the map data value,
+ *     or NULL if the values from the map are to be used unmodified
+ * @param context an opaque pointer that is passed on to the filter function
+ * @param pValue if not NULL, receives the value that every code point start..end has;
+ *     may have been modified by filter(context, map value)
+ *     if that function pointer is not NULL
+ * @return the range end code point, or -1 if start is not a valid code point
+ * @stable ICU 63
+ */
+U_CAPI UChar32 U_EXPORT2
+ucpmap_getRange(const UCPMap *map, UChar32 start,
+                UCPMapRangeOption option, uint32_t surrogateValue,
+                UCPMapValueFilter *filter, const void *context, uint32_t *pValue);
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucptrie.h b/thirdparty/icu4c/common/unicode/ucptrie.h
new file mode 100644
index 0000000000..b95491b183
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucptrie.h
@@ -0,0 +1,646 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// ucptrie.h (modified from utrie2.h)
+// created: 2017dec29 Markus W. Scherer
+
+#ifndef __UCPTRIE_H__
+#define __UCPTRIE_H__
+
+#include "unicode/utypes.h"
+#include "unicode/ucpmap.h"
+#include "unicode/utf8.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+U_CDECL_BEGIN
+
+/**
+ * \file
+ *
+ * This file defines an immutable Unicode code point trie.
+ *
+ * @see UCPTrie
+ * @see UMutableCPTrie
+ */
+
+#ifndef U_IN_DOXYGEN
+/** @internal */
+typedef union UCPTrieData {
+    /** @internal */
+    const void *ptr0;
+    /** @internal */
+    const uint16_t *ptr16;
+    /** @internal */
+    const uint32_t *ptr32;
+    /** @internal */
+    const uint8_t *ptr8;
+} UCPTrieData;
+#endif
+
+/**
+ * Immutable Unicode code point trie structure.
+ * Fast, reasonably compact, map from Unicode code points (U+0000..U+10FFFF) to integer values.
+ * For details see http://site.icu-project.org/design/struct/utrie
+ *
+ * Do not access UCPTrie fields directly; use public functions and macros.
+ * Functions are easy to use: They support all trie types and value widths.
+ *
+ * When performance is really important, macros provide faster access.
+ * Most macros are specific to either "fast" or "small" tries, see UCPTrieType.
+ * There are "fast" macros for special optimized use cases.
+ *
+ * The macros will return bogus values, or may crash, if used on the wrong type or value width.
+ *
+ * @see UMutableCPTrie
+ * @stable ICU 63
+ */
+struct UCPTrie {
+#ifndef U_IN_DOXYGEN
+    /** @internal */
+    const uint16_t *index;
+    /** @internal */
+    UCPTrieData data;
+
+    /** @internal */
+    int32_t indexLength;
+    /** @internal */
+    int32_t dataLength;
+    /** Start of the last range which ends at U+10FFFF. @internal */
+    UChar32 highStart;
+    /** highStart>>12 @internal */
+    uint16_t shifted12HighStart;
+
+    /** @internal */
+    int8_t type;  // UCPTrieType
+    /** @internal */
+    int8_t valueWidth;  // UCPTrieValueWidth
+
+    /** padding/reserved @internal */
+    uint32_t reserved32;
+    /** padding/reserved @internal */
+    uint16_t reserved16;
+
+    /**
+     * Internal index-3 null block offset.
+     * Set to an impossibly high value (e.g., 0xffff) if there is no dedicated index-3 null block.
+     * @internal
+     */
+    uint16_t index3NullOffset;
+    /**
+     * Internal data null block offset, not shifted.
+     * Set to an impossibly high value (e.g., 0xfffff) if there is no dedicated data null block.
+     * @internal
+     */
+    int32_t dataNullOffset;
+    /** @internal */
+    uint32_t nullValue;
+
+#ifdef UCPTRIE_DEBUG
+    /** @internal */
+    const char *name;
+#endif
+#endif
+};
+#ifndef U_IN_DOXYGEN
+typedef struct UCPTrie UCPTrie;
+#endif
+
+/**
+ * Selectors for the type of a UCPTrie.
+ * Different trade-offs for size vs. speed.
+ *
+ * @see umutablecptrie_buildImmutable
+ * @see ucptrie_openFromBinary
+ * @see ucptrie_getType
+ * @stable ICU 63
+ */
+enum UCPTrieType {
+    /**
+     * For ucptrie_openFromBinary() to accept any type.
+     * ucptrie_getType() will return the actual type.
+     * @stable ICU 63
+     */
+    UCPTRIE_TYPE_ANY = -1,
+    /**
+     * Fast/simple/larger BMP data structure. Use functions and "fast" macros.
+     * @stable ICU 63
+     */
+    UCPTRIE_TYPE_FAST,
+    /**
+     * Small/slower BMP data structure. Use functions and "small" macros.
+     * @stable ICU 63
+     */
+    UCPTRIE_TYPE_SMALL
+};
+#ifndef U_IN_DOXYGEN
+typedef enum UCPTrieType UCPTrieType;
+#endif
+
+/**
+ * Selectors for the number of bits in a UCPTrie data value.
+ *
+ * @see umutablecptrie_buildImmutable
+ * @see ucptrie_openFromBinary
+ * @see ucptrie_getValueWidth
+ * @stable ICU 63
+ */
+enum UCPTrieValueWidth {
+    /**
+     * For ucptrie_openFromBinary() to accept any data value width.
+     * ucptrie_getValueWidth() will return the actual data value width.
+     * @stable ICU 63
+     */
+    UCPTRIE_VALUE_BITS_ANY = -1,
+    /**
+     * The trie stores 16 bits per data value.
+     * It returns them as unsigned values 0..0xffff=65535.
+     * @stable ICU 63
+     */
+    UCPTRIE_VALUE_BITS_16,
+    /**
+     * The trie stores 32 bits per data value.
+     * @stable ICU 63
+     */
+    UCPTRIE_VALUE_BITS_32,
+    /**
+     * The trie stores 8 bits per data value.
+     * It returns them as unsigned values 0..0xff=255.
+     * @stable ICU 63
+     */
+    UCPTRIE_VALUE_BITS_8
+};
+#ifndef U_IN_DOXYGEN
+typedef enum UCPTrieValueWidth UCPTrieValueWidth;
+#endif
+
+/**
+ * Opens a trie from its binary form, stored in 32-bit-aligned memory.
+ * Inverse of ucptrie_toBinary().
+ *
+ * The memory must remain valid and unchanged as long as the trie is used.
+ * You must ucptrie_close() the trie once you are done using it.
+ *
+ * @param type selects the trie type; results in an
+ *             U_INVALID_FORMAT_ERROR if it does not match the binary data;
+ *             use UCPTRIE_TYPE_ANY to accept any type
+ * @param valueWidth selects the number of bits in a data value; results in an
+ *                  U_INVALID_FORMAT_ERROR if it does not match the binary data;
+ *                  use UCPTRIE_VALUE_BITS_ANY to accept any data value width
+ * @param data a pointer to 32-bit-aligned memory containing the binary data of a UCPTrie
+ * @param length the number of bytes available at data;
+ *               can be more than necessary
+ * @param pActualLength receives the actual number of bytes at data taken up by the trie data;
+ *                      can be NULL
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the trie
+ *
+ * @see umutablecptrie_open
+ * @see umutablecptrie_buildImmutable
+ * @see ucptrie_toBinary
+ * @stable ICU 63
+ */
+U_CAPI UCPTrie * U_EXPORT2
+ucptrie_openFromBinary(UCPTrieType type, UCPTrieValueWidth valueWidth,
+                       const void *data, int32_t length, int32_t *pActualLength,
+                       UErrorCode *pErrorCode);
+
+/**
+ * Closes a trie and releases associated memory.
+ *
+ * @param trie the trie
+ * @stable ICU 63
+ */
+U_CAPI void U_EXPORT2
+ucptrie_close(UCPTrie *trie);
+
+/**
+ * Returns the trie type.
+ *
+ * @param trie the trie
+ * @return the trie type
+ * @see ucptrie_openFromBinary
+ * @see UCPTRIE_TYPE_ANY
+ * @stable ICU 63
+ */
+U_CAPI UCPTrieType U_EXPORT2
+ucptrie_getType(const UCPTrie *trie);
+
+/**
+ * Returns the number of bits in a trie data value.
+ *
+ * @param trie the trie
+ * @return the number of bits in a trie data value
+ * @see ucptrie_openFromBinary
+ * @see UCPTRIE_VALUE_BITS_ANY
+ * @stable ICU 63
+ */
+U_CAPI UCPTrieValueWidth U_EXPORT2
+ucptrie_getValueWidth(const UCPTrie *trie);
+
+/**
+ * Returns the value for a code point as stored in the trie, with range checking.
+ * Returns the trie error value if c is not in the range 0..U+10FFFF.
+ *
+ * Easier to use than UCPTRIE_FAST_GET() and similar macros but slower.
+ * Easier to use because, unlike the macros, this function works on all UCPTrie
+ * objects, for all types and value widths.
+ *
+ * @param trie the trie
+ * @param c the code point
+ * @return the trie value,
+ *         or the trie error value if the code point is not in the range 0..U+10FFFF
+ * @stable ICU 63
+ */
+U_CAPI uint32_t U_EXPORT2
+ucptrie_get(const UCPTrie *trie, UChar32 c);
+
+/**
+ * Returns the last code point such that all those from start to there have the same value.
+ * Can be used to efficiently iterate over all same-value ranges in a trie.
+ * (This is normally faster than iterating over code points and get()ting each value,
+ * but much slower than a data structure that stores ranges directly.)
+ *
+ * If the UCPMapValueFilter function pointer is not NULL, then
+ * the value to be delivered is passed through that function, and the return value is the end
+ * of the range where all values are modified to the same actual value.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * Example:
+ * \code
+ * UChar32 start = 0, end;
+ * uint32_t value;
+ * while ((end = ucptrie_getRange(trie, start, UCPMAP_RANGE_NORMAL, 0,
+ *                                NULL, NULL, &value)) >= 0) {
+ *     // Work with the range start..end and its value.
+ *     start = end + 1;
+ * }
+ * \endcode
+ *
+ * @param trie the trie
+ * @param start range start
+ * @param option defines whether surrogates are treated normally,
+ *               or as having the surrogateValue; usually UCPMAP_RANGE_NORMAL
+ * @param surrogateValue value for surrogates; ignored if option==UCPMAP_RANGE_NORMAL
+ * @param filter a pointer to a function that may modify the trie data value,
+ *     or NULL if the values from the trie are to be used unmodified
+ * @param context an opaque pointer that is passed on to the filter function
+ * @param pValue if not NULL, receives the value that every code point start..end has;
+ *     may have been modified by filter(context, trie value)
+ *     if that function pointer is not NULL
+ * @return the range end code point, or -1 if start is not a valid code point
+ * @stable ICU 63
+ */
+U_CAPI UChar32 U_EXPORT2
+ucptrie_getRange(const UCPTrie *trie, UChar32 start,
+                 UCPMapRangeOption option, uint32_t surrogateValue,
+                 UCPMapValueFilter *filter, const void *context, uint32_t *pValue);
+
+/**
+ * Writes a memory-mappable form of the trie into 32-bit aligned memory.
+ * Inverse of ucptrie_openFromBinary().
+ *
+ * @param trie the trie
+ * @param data a pointer to 32-bit-aligned memory to be filled with the trie data;
+ *             can be NULL if capacity==0
+ * @param capacity the number of bytes available at data, or 0 for pure preflighting
+ * @param pErrorCode an in/out ICU UErrorCode;
+ *                   U_BUFFER_OVERFLOW_ERROR if the capacity is too small
+ * @return the number of bytes written or (if buffer overflow) needed for the trie
+ *
+ * @see ucptrie_openFromBinary()
+ * @stable ICU 63
+ */
+U_CAPI int32_t U_EXPORT2
+ucptrie_toBinary(const UCPTrie *trie, void *data, int32_t capacity, UErrorCode *pErrorCode);
+
+/**
+ * Macro parameter value for a trie with 16-bit data values.
+ * Use the name of this macro as a "dataAccess" parameter in other macros.
+ * Do not use this macro in any other way.
+ *
+ * @see UCPTRIE_VALUE_BITS_16
+ * @stable ICU 63
+ */
+#define UCPTRIE_16(trie, i) ((trie)->data.ptr16[i])
+
+/**
+ * Macro parameter value for a trie with 32-bit data values.
+ * Use the name of this macro as a "dataAccess" parameter in other macros.
+ * Do not use this macro in any other way.
+ *
+ * @see UCPTRIE_VALUE_BITS_32
+ * @stable ICU 63
+ */
+#define UCPTRIE_32(trie, i) ((trie)->data.ptr32[i])
+
+/**
+ * Macro parameter value for a trie with 8-bit data values.
+ * Use the name of this macro as a "dataAccess" parameter in other macros.
+ * Do not use this macro in any other way.
+ *
+ * @see UCPTRIE_VALUE_BITS_8
+ * @stable ICU 63
+ */
+#define UCPTRIE_8(trie, i) ((trie)->data.ptr8[i])
+
+/**
+ * Returns a trie value for a code point, with range checking.
+ * Returns the trie error value if c is not in the range 0..U+10FFFF.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param c (UChar32, in) the input code point
+ * @return The code point's trie value.
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_CP_INDEX(trie, 0xffff, c))
+
+/**
+ * Returns a 16-bit trie value for a code point, with range checking.
+ * Returns the trie error value if c is not in the range U+0000..U+10FFFF.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_SMALL
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param c (UChar32, in) the input code point
+ * @return The code point's trie value.
+ * @stable ICU 63
+ */
+#define UCPTRIE_SMALL_GET(trie, dataAccess, c) \
+    dataAccess(trie, _UCPTRIE_CP_INDEX(trie, UCPTRIE_SMALL_MAX, c))
+
+/**
+ * UTF-16: Reads the next code point (UChar32 c, out), post-increments src,
+ * and gets a value from the trie.
+ * Sets the trie error value if c is an unpaired surrogate.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated
+ * @param c (UChar32, out) variable for the code point
+ * @param result (out) variable for the trie lookup result
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_U16_NEXT(trie, dataAccess, src, limit, c, result) UPRV_BLOCK_MACRO_BEGIN { \
+    (c) = *(src)++; \
+    int32_t __index; \
+    if (!U16_IS_SURROGATE(c)) { \
+        __index = _UCPTRIE_FAST_INDEX(trie, c); \
+    } else { \
+        uint16_t __c2; \
+        if (U16_IS_SURROGATE_LEAD(c) && (src) != (limit) && U16_IS_TRAIL(__c2 = *(src))) { \
+            ++(src); \
+            (c) = U16_GET_SUPPLEMENTARY((c), __c2); \
+            __index = _UCPTRIE_SMALL_INDEX(trie, c); \
+        } else { \
+            __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \
+        } \
+    } \
+    (result) = dataAccess(trie, __index); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * UTF-16: Reads the previous code point (UChar32 c, out), pre-decrements src,
+ * and gets a value from the trie.
+ * Sets the trie error value if c is an unpaired surrogate.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param start (const UChar *, in) the start pointer for the text
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param c (UChar32, out) variable for the code point
+ * @param result (out) variable for the trie lookup result
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_U16_PREV(trie, dataAccess, start, src, c, result) UPRV_BLOCK_MACRO_BEGIN { \
+    (c) = *--(src); \
+    int32_t __index; \
+    if (!U16_IS_SURROGATE(c)) { \
+        __index = _UCPTRIE_FAST_INDEX(trie, c); \
+    } else { \
+        uint16_t __c2; \
+        if (U16_IS_SURROGATE_TRAIL(c) && (src) != (start) && U16_IS_LEAD(__c2 = *((src) - 1))) { \
+            --(src); \
+            (c) = U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            __index = _UCPTRIE_SMALL_INDEX(trie, c); \
+        } else { \
+            __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \
+        } \
+    } \
+    (result) = dataAccess(trie, __index); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * UTF-8: Post-increments src and gets a value from the trie.
+ * Sets the trie error value for an ill-formed byte sequence.
+ *
+ * Unlike UCPTRIE_FAST_U16_NEXT() this UTF-8 macro does not provide the code point
+ * because it would be more work to do so and is often not needed.
+ * If the trie value differs from the error value, then the byte sequence is well-formed,
+ * and the code point can be assembled without revalidation.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param src (const char *, in/out) the source text pointer
+ * @param limit (const char *, in) the limit pointer for the text (must not be NULL)
+ * @param result (out) variable for the trie lookup result
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_U8_NEXT(trie, dataAccess, src, limit, result) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __lead = (uint8_t)*(src)++; \
+    if (!U8_IS_SINGLE(__lead)) { \
+        uint8_t __t1, __t2, __t3; \
+        if ((src) != (limit) && \
+            (__lead >= 0xe0 ? \
+                __lead < 0xf0 ?  /* U+0800..U+FFFF except surrogates */ \
+                    U8_LEAD3_T1_BITS[__lead &= 0xf] & (1 << ((__t1 = *(src)) >> 5)) && \
+                    ++(src) != (limit) && (__t2 = *(src) - 0x80) <= 0x3f && \
+                    (__lead = ((int32_t)(trie)->index[(__lead << 6) + (__t1 & 0x3f)]) + __t2, 1) \
+                :  /* U+10000..U+10FFFF */ \
+                    (__lead -= 0xf0) <= 4 && \
+                    U8_LEAD4_T1_BITS[(__t1 = *(src)) >> 4] & (1 << __lead) && \
+                    (__lead = (__lead << 6) | (__t1 & 0x3f), ++(src) != (limit)) && \
+                    (__t2 = *(src) - 0x80) <= 0x3f && \
+                    ++(src) != (limit) && (__t3 = *(src) - 0x80) <= 0x3f && \
+                    (__lead = __lead >= (trie)->shifted12HighStart ? \
+                        (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \
+                        ucptrie_internalSmallU8Index((trie), __lead, __t2, __t3), 1) \
+            :  /* U+0080..U+07FF */ \
+                __lead >= 0xc2 && (__t1 = *(src) - 0x80) <= 0x3f && \
+                (__lead = (int32_t)(trie)->index[__lead & 0x1f] + __t1, 1))) { \
+            ++(src); \
+        } else { \
+            __lead = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET;  /* ill-formed*/ \
+        } \
+    } \
+    (result) = dataAccess(trie, __lead); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * UTF-8: Pre-decrements src and gets a value from the trie.
+ * Sets the trie error value for an ill-formed byte sequence.
+ *
+ * Unlike UCPTRIE_FAST_U16_PREV() this UTF-8 macro does not provide the code point
+ * because it would be more work to do so and is often not needed.
+ * If the trie value differs from the error value, then the byte sequence is well-formed,
+ * and the code point can be assembled without revalidation.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param start (const char *, in) the start pointer for the text
+ * @param src (const char *, in/out) the source text pointer
+ * @param result (out) variable for the trie lookup result
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_U8_PREV(trie, dataAccess, start, src, result) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __index = (uint8_t)*--(src); \
+    if (!U8_IS_SINGLE(__index)) { \
+        __index = ucptrie_internalU8PrevIndex((trie), __index, (const uint8_t *)(start), \
+                                              (const uint8_t *)(src)); \
+        (src) -= __index & 7; \
+        __index >>= 3; \
+    } \
+    (result) = dataAccess(trie, __index); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Returns a trie value for an ASCII code point, without range checking.
+ *
+ * @param trie (const UCPTrie *, in) the trie (of either fast or small type)
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param c (UChar32, in) the input code point; must be U+0000..U+007F
+ * @return The ASCII code point's trie value.
+ * @stable ICU 63
+ */
+#define UCPTRIE_ASCII_GET(trie, dataAccess, c) dataAccess(trie, c)
+
+/**
+ * Returns a trie value for a BMP code point (U+0000..U+FFFF), without range checking.
+ * Can be used to look up a value for a UTF-16 code unit if other parts of
+ * the string processing check for surrogates.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param c (UChar32, in) the input code point, must be U+0000..U+FFFF
+ * @return The BMP code point's trie value.
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_BMP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_FAST_INDEX(trie, c))
+
+/**
+ * Returns a trie value for a supplementary code point (U+10000..U+10FFFF),
+ * without range checking.
+ *
+ * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST
+ * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width
+ * @param c (UChar32, in) the input code point, must be U+10000..U+10FFFF
+ * @return The supplementary code point's trie value.
+ * @stable ICU 63
+ */
+#define UCPTRIE_FAST_SUPP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_SMALL_INDEX(trie, c))
+
+/* Internal definitions ----------------------------------------------------- */
+
+#ifndef U_IN_DOXYGEN
+
+/**
+ * Internal implementation constants.
+ * These are needed for the API macros, but users should not use these directly.
+ * @internal
+ */
+enum {
+    /** @internal */
+    UCPTRIE_FAST_SHIFT = 6,
+
+    /** Number of entries in a data block for code points below the fast limit. 64=0x40 @internal */
+    UCPTRIE_FAST_DATA_BLOCK_LENGTH = 1 << UCPTRIE_FAST_SHIFT,
+
+    /** Mask for getting the lower bits for the in-fast-data-block offset. @internal */
+    UCPTRIE_FAST_DATA_MASK = UCPTRIE_FAST_DATA_BLOCK_LENGTH - 1,
+
+    /** @internal */
+    UCPTRIE_SMALL_MAX = 0xfff,
+
+    /**
+     * Offset from dataLength (to be subtracted) for fetching the
+     * value returned for out-of-range code points and ill-formed UTF-8/16.
+     * @internal
+     */
+    UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET = 1,
+    /**
+     * Offset from dataLength (to be subtracted) for fetching the
+     * value returned for code points highStart..U+10FFFF.
+     * @internal
+     */
+    UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET = 2
+};
+
+/* Internal functions and macros -------------------------------------------- */
+// Do not conditionalize with #ifndef U_HIDE_INTERNAL_API, needed for public API
+
+/** @internal */
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalSmallIndex(const UCPTrie *trie, UChar32 c);
+
+/** @internal */
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalSmallU8Index(const UCPTrie *trie, int32_t lt1, uint8_t t2, uint8_t t3);
+
+/**
+ * Internal function for part of the UCPTRIE_FAST_U8_PREVxx() macro implementations.
+ * Do not call directly.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ucptrie_internalU8PrevIndex(const UCPTrie *trie, UChar32 c,
+                            const uint8_t *start, const uint8_t *src);
+
+/** Internal trie getter for a code point below the fast limit. Returns the data index. @internal */
+#define _UCPTRIE_FAST_INDEX(trie, c) \
+    ((int32_t)(trie)->index[(c) >> UCPTRIE_FAST_SHIFT] + ((c) & UCPTRIE_FAST_DATA_MASK))
+
+/** Internal trie getter for a code point at or above the fast limit. Returns the data index. @internal */
+#define _UCPTRIE_SMALL_INDEX(trie, c) \
+    ((c) >= (trie)->highStart ? \
+        (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \
+        ucptrie_internalSmallIndex(trie, c))
+
+/**
+ * Internal trie getter for a code point, with checking that c is in U+0000..10FFFF.
+ * Returns the data index.
+ * @internal
+ */
+#define _UCPTRIE_CP_INDEX(trie, fastMax, c) \
+    ((uint32_t)(c) <= (uint32_t)(fastMax) ? \
+        _UCPTRIE_FAST_INDEX(trie, c) : \
+        (uint32_t)(c) <= 0x10ffff ? \
+            _UCPTRIE_SMALL_INDEX(trie, c) : \
+            (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET)
+
+U_CDECL_END
+
+#endif  // U_IN_DOXYGEN
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUCPTriePointer
+ * "Smart pointer" class, closes a UCPTrie via ucptrie_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 63
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUCPTriePointer, UCPTrie, ucptrie_close);
+
+U_NAMESPACE_END
+
+#endif  // U_SHOW_CPLUSPLUS_API
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ucurr.h b/thirdparty/icu4c/common/unicode/ucurr.h
new file mode 100644
index 0000000000..35c2a39389
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ucurr.h
@@ -0,0 +1,468 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2016, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+#ifndef _UCURR_H_
+#define _UCURR_H_
+
+#include "unicode/utypes.h"
+#include "unicode/uenum.h"
+
+/**
+ * \file 
+ * \brief C API: Encapsulates information about a currency.
+ *
+ * The ucurr API encapsulates information about a currency, as defined by
+ * ISO 4217.  A currency is represented by a 3-character string
+ * containing its ISO 4217 code.  This API can return various data
+ * necessary the proper display of a currency:
+ *
+ * <ul><li>A display symbol, for a specific locale
+ * <li>The number of fraction digits to display
+ * <li>A rounding increment
+ * </ul>
+ *
+ * The <tt>DecimalFormat</tt> class uses these data to display
+ * currencies.
+ * @author Alan Liu
+ * @since ICU 2.2
+ */
+
+#if !UCONFIG_NO_FORMATTING
+
+/**
+ * Currency Usage used for Decimal Format
+ * @stable ICU 54
+ */
+enum UCurrencyUsage {
+    /**
+     * a setting to specify currency usage which determines currency digit
+     * and rounding for standard usage, for example: "50.00 NT$"
+     * used as DEFAULT value
+     * @stable ICU 54
+     */
+    UCURR_USAGE_STANDARD=0,
+    /**
+     * a setting to specify currency usage which determines currency digit
+     * and rounding for cash usage, for example: "50 NT$"
+     * @stable ICU 54
+     */
+    UCURR_USAGE_CASH=1,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One higher than the last enum UCurrencyUsage constant.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UCURR_USAGE_COUNT=2
+#endif  // U_HIDE_DEPRECATED_API
+};
+/** Currency Usage used for Decimal Format */
+typedef enum UCurrencyUsage UCurrencyUsage; 
+
+/**
+ * Finds a currency code for the given locale.
+ * @param locale the locale for which to retrieve a currency code. 
+ *               Currency can be specified by the "currency" keyword
+ *               in which case it overrides the default currency code
+ * @param buff   fill in buffer. Can be NULL for preflighting.
+ * @param buffCapacity capacity of the fill in buffer. Can be 0 for
+ *               preflighting. If it is non-zero, the buff parameter
+ *               must not be NULL.
+ * @param ec error code
+ * @return length of the currency string. It should always be 3. If 0,
+ *                currency couldn't be found or the input values are 
+ *                invalid. 
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+ucurr_forLocale(const char* locale,
+                UChar* buff,
+                int32_t buffCapacity,
+                UErrorCode* ec);
+
+/**
+ * Selector constants for ucurr_getName().
+ *
+ * @see ucurr_getName
+ * @stable ICU 2.6
+ */
+typedef enum UCurrNameStyle {
+    /**
+     * Selector for ucurr_getName indicating a symbolic name for a
+     * currency, such as "$" for USD.
+     * @stable ICU 2.6
+     */
+    UCURR_SYMBOL_NAME,
+
+    /**
+     * Selector for ucurr_getName indicating the long name for a
+     * currency, such as "US Dollar" for USD.
+     * @stable ICU 2.6
+     */
+    UCURR_LONG_NAME,
+
+    /**
+     * Selector for getName() indicating the narrow currency symbol.
+     * The narrow currency symbol is similar to the regular currency
+     * symbol, but it always takes the shortest form: for example,
+     * "$" instead of "US$" for USD in en-CA.
+     *
+     * @stable ICU 61
+     */
+    UCURR_NARROW_SYMBOL_NAME,
+
+#ifndef U_HIDE_DRAFT_API
+    /**
+     * Selector for getName() indicating the formal currency symbol.
+     * The formal currency symbol is similar to the regular currency
+     * symbol, but it always takes the form used in formal settings
+     * such as banking; for example, "NT$" instead of "$" for TWD in zh-TW.
+     *
+     * @draft ICU 68
+     */
+    UCURR_FORMAL_SYMBOL_NAME,
+
+    /**
+     * Selector for getName() indicating the variant currency symbol.
+     * The variant symbol for a currency is an alternative symbol
+     * that is not necessarily as widely used as the regular symbol.
+     *
+     * @draft ICU 68
+     */
+    UCURR_VARIANT_SYMBOL_NAME
+#endif  // U_HIDE_DRAFT_API
+    
+} UCurrNameStyle;
+
+#if !UCONFIG_NO_SERVICE
+/**
+ * @stable ICU 2.6
+ */
+typedef const void* UCurrRegistryKey;
+
+/**
+ * Register an (existing) ISO 4217 currency code for the given locale.
+ * Only the country code and the two variants EURO and PRE_EURO are
+ * recognized.
+ * @param isoCode the three-letter ISO 4217 currency code
+ * @param locale  the locale for which to register this currency code
+ * @param status the in/out status code
+ * @return a registry key that can be used to unregister this currency code, or NULL
+ * if there was an error.
+ * @stable ICU 2.6
+ */
+U_CAPI UCurrRegistryKey U_EXPORT2
+ucurr_register(const UChar* isoCode, 
+                   const char* locale,  
+                   UErrorCode* status);
+/**
+ * Unregister the previously-registered currency definitions using the
+ * URegistryKey returned from ucurr_register.  Key becomes invalid after
+ * a successful call and should not be used again.  Any currency 
+ * that might have been hidden by the original ucurr_register call is 
+ * restored.
+ * @param key the registry key returned by a previous call to ucurr_register
+ * @param status the in/out status code, no special meanings are assigned
+ * @return true if the currency for this key was successfully unregistered
+ * @stable ICU 2.6
+ */
+U_CAPI UBool U_EXPORT2
+ucurr_unregister(UCurrRegistryKey key, UErrorCode* status);
+#endif /* UCONFIG_NO_SERVICE */
+
+/**
+ * Returns the display name for the given currency in the
+ * given locale.  For example, the display name for the USD
+ * currency object in the en_US locale is "$".
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param locale locale in which to display currency
+ * @param nameStyle selector for which kind of name to return
+ * @param isChoiceFormat always set to false, or can be NULL;
+ *     display names are static strings;
+ *     since ICU 4.4, ChoiceFormat patterns are no longer supported
+ * @param len fill-in parameter to receive length of result
+ * @param ec error code
+ * @return pointer to display string of 'len' UChars.  If the resource
+ * data contains no entry for 'currency', then 'currency' itself is
+ * returned.
+ * @stable ICU 2.6
+ */
+U_CAPI const UChar* U_EXPORT2
+ucurr_getName(const UChar* currency,
+              const char* locale,
+              UCurrNameStyle nameStyle,
+              UBool* isChoiceFormat,
+              int32_t* len,
+              UErrorCode* ec);
+
+/**
+ * Returns the plural name for the given currency in the
+ * given locale.  For example, the plural name for the USD
+ * currency object in the en_US locale is "US dollar" or "US dollars".
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param locale locale in which to display currency
+ * @param isChoiceFormat always set to false, or can be NULL;
+ *     display names are static strings;
+ *     since ICU 4.4, ChoiceFormat patterns are no longer supported
+ * @param pluralCount plural count
+ * @param len fill-in parameter to receive length of result
+ * @param ec error code
+ * @return pointer to display string of 'len' UChars.  If the resource
+ * data contains no entry for 'currency', then 'currency' itself is
+ * returned.
+ * @stable ICU 4.2
+ */
+U_CAPI const UChar* U_EXPORT2
+ucurr_getPluralName(const UChar* currency,
+                    const char* locale,
+                    UBool* isChoiceFormat,
+                    const char* pluralCount,
+                    int32_t* len,
+                    UErrorCode* ec);
+
+/**
+ * Returns the number of the number of fraction digits that should
+ * be displayed for the given currency.
+ * This is equivalent to ucurr_getDefaultFractionDigitsForUsage(currency,UCURR_USAGE_STANDARD,ec);
+ *
+ * Important: The number of fraction digits for a given currency is NOT
+ * guaranteed to be constant across versions of ICU or CLDR. For example,
+ * do NOT use this value as a mechanism for deciding the magnitude used
+ * to store currency values in a database. You should use this value for
+ * display purposes only.
+ *
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param ec input-output error code
+ * @return a non-negative number of fraction digits to be
+ * displayed, or 0 if there is an error
+ * @stable ICU 3.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucurr_getDefaultFractionDigits(const UChar* currency,
+                               UErrorCode* ec);
+
+/**
+ * Returns the number of the number of fraction digits that should
+ * be displayed for the given currency with usage.
+ *
+ * Important: The number of fraction digits for a given currency is NOT
+ * guaranteed to be constant across versions of ICU or CLDR. For example,
+ * do NOT use this value as a mechanism for deciding the magnitude used
+ * to store currency values in a database. You should use this value for
+ * display purposes only.
+ *
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param usage enum usage for the currency
+ * @param ec input-output error code
+ * @return a non-negative number of fraction digits to be
+ * displayed, or 0 if there is an error
+ * @stable ICU 54
+ */
+U_CAPI int32_t U_EXPORT2
+ucurr_getDefaultFractionDigitsForUsage(const UChar* currency, 
+                                       const UCurrencyUsage usage,
+                                       UErrorCode* ec);
+
+/**
+ * Returns the rounding increment for the given currency, or 0.0 if no
+ * rounding is done by the currency.
+ * This is equivalent to ucurr_getRoundingIncrementForUsage(currency,UCURR_USAGE_STANDARD,ec);
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param ec input-output error code
+ * @return the non-negative rounding increment, or 0.0 if none,
+ * or 0.0 if there is an error
+ * @stable ICU 3.0
+ */
+U_CAPI double U_EXPORT2
+ucurr_getRoundingIncrement(const UChar* currency,
+                           UErrorCode* ec);
+
+/**
+ * Returns the rounding increment for the given currency, or 0.0 if no
+ * rounding is done by the currency given usage.
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @param usage enum usage for the currency
+ * @param ec input-output error code
+ * @return the non-negative rounding increment, or 0.0 if none,
+ * or 0.0 if there is an error
+ * @stable ICU 54
+ */
+U_CAPI double U_EXPORT2
+ucurr_getRoundingIncrementForUsage(const UChar* currency,
+                                   const UCurrencyUsage usage,
+                                   UErrorCode* ec);
+
+/**
+ * Selector constants for ucurr_openCurrencies().
+ *
+ * @see ucurr_openCurrencies
+ * @stable ICU 3.2
+ */
+typedef enum UCurrCurrencyType {
+    /**
+     * Select all ISO-4217 currency codes.
+     * @stable ICU 3.2
+     */
+    UCURR_ALL = INT32_MAX,
+    /**
+     * Select only ISO-4217 commonly used currency codes.
+     * These currencies can be found in common use, and they usually have
+     * bank notes or coins associated with the currency code.
+     * This does not include fund codes, precious metals and other
+     * various ISO-4217 codes limited to special financial products.
+     * @stable ICU 3.2
+     */
+    UCURR_COMMON = 1,
+    /**
+     * Select ISO-4217 uncommon currency codes.
+     * These codes respresent fund codes, precious metals and other
+     * various ISO-4217 codes limited to special financial products.
+     * A fund code is a monetary resource associated with a currency.
+     * @stable ICU 3.2
+     */
+    UCURR_UNCOMMON = 2,
+    /**
+     * Select only deprecated ISO-4217 codes.
+     * These codes are no longer in general public use.
+     * @stable ICU 3.2
+     */
+    UCURR_DEPRECATED = 4,
+    /**
+     * Select only non-deprecated ISO-4217 codes.
+     * These codes are in general public use.
+     * @stable ICU 3.2
+     */
+    UCURR_NON_DEPRECATED = 8
+} UCurrCurrencyType;
+
+/**
+ * Provides a UEnumeration object for listing ISO-4217 codes.
+ * @param currType You can use one of several UCurrCurrencyType values for this
+ *      variable. You can also | (or) them together to get a specific list of
+ *      currencies. Most people will want to use the (UCURR_COMMON|UCURR_NON_DEPRECATED) value to
+ *      get a list of current currencies.
+ * @param pErrorCode Error code
+ * @stable ICU 3.2
+ */
+U_CAPI UEnumeration * U_EXPORT2
+ucurr_openISOCurrencies(uint32_t currType, UErrorCode *pErrorCode);
+
+/**
+  * Queries if the given ISO 4217 3-letter code is available on the specified date range. 
+  * 
+  * Note: For checking availability of a currency on a specific date, specify the date on both 'from' and 'to' 
+  * 
+  * When 'from' is U_DATE_MIN and 'to' is U_DATE_MAX, this method checks if the specified currency is available any time. 
+  * If 'from' and 'to' are same UDate value, this method checks if the specified currency is available on that date.
+  * 
+  * @param isoCode 
+  *            The ISO 4217 3-letter code. 
+  * 
+  * @param from 
+  *            The lower bound of the date range, inclusive. When 'from' is U_DATE_MIN, check the availability 
+  *            of the currency any date before 'to' 
+  * 
+  * @param to 
+  *            The upper bound of the date range, inclusive. When 'to' is U_DATE_MAX, check the availability of 
+  *            the currency any date after 'from' 
+  * 
+  * @param errorCode 
+  *            ICU error code 
+   * 
+  * @return true if the given ISO 4217 3-letter code is supported on the specified date range. 
+  * 
+  * @stable ICU 4.8 
+  */ 
+U_CAPI UBool U_EXPORT2
+ucurr_isAvailable(const UChar* isoCode, 
+             UDate from, 
+             UDate to, 
+             UErrorCode* errorCode);
+
+/** 
+ * Finds the number of valid currency codes for the
+ * given locale and date.
+ * @param locale the locale for which to retrieve the
+ *               currency count.
+ * @param date   the date for which to retrieve the
+ *               currency count for the given locale.
+ * @param ec     error code
+ * @return       the number of currency codes for the
+ *               given locale and date.  If 0, currency
+ *               codes couldn't be found for the input
+ *               values are invalid.
+ * @stable ICU 4.0
+ */
+U_CAPI int32_t U_EXPORT2
+ucurr_countCurrencies(const char* locale, 
+                 UDate date, 
+                 UErrorCode* ec); 
+
+/** 
+ * Finds a currency code for the given locale and date 
+ * @param locale the locale for which to retrieve a currency code.  
+ *               Currency can be specified by the "currency" keyword 
+ *               in which case it overrides the default currency code 
+ * @param date   the date for which to retrieve a currency code for 
+ *               the given locale. 
+ * @param index  the index within the available list of currency codes
+ *               for the given locale on the given date.
+ * @param buff   fill in buffer. Can be NULL for preflighting. 
+ * @param buffCapacity capacity of the fill in buffer. Can be 0 for 
+ *               preflighting. If it is non-zero, the buff parameter 
+ *               must not be NULL. 
+ * @param ec     error code 
+ * @return       length of the currency string. It should always be 3. 
+ *               If 0, currency couldn't be found or the input values are  
+ *               invalid.  
+ * @stable ICU 4.0 
+ */ 
+U_CAPI int32_t U_EXPORT2 
+ucurr_forLocaleAndDate(const char* locale, 
+                UDate date, 
+                int32_t index,
+                UChar* buff, 
+                int32_t buffCapacity, 
+                UErrorCode* ec); 
+
+/**
+ * Given a key and a locale, returns an array of string values in a preferred
+ * order that would make a difference. These are all and only those values where
+ * the open (creation) of the service with the locale formed from the input locale
+ * plus input keyword and that value has different behavior than creation with the
+ * input locale alone.
+ * @param key           one of the keys supported by this service.  For now, only
+ *                      "currency" is supported.
+ * @param locale        the locale
+ * @param commonlyUsed  if set to true it will return only commonly used values
+ *                      with the given locale in preferred order.  Otherwise,
+ *                      it will return all the available values for the locale.
+ * @param status error status
+ * @return a string enumeration over keyword values for the given key and the locale.
+ * @stable ICU 4.2
+ */
+U_CAPI UEnumeration* U_EXPORT2
+ucurr_getKeywordValuesForLocale(const char* key,
+                                const char* locale,
+                                UBool commonlyUsed,
+                                UErrorCode* status);
+
+/**
+ * Returns the ISO 4217 numeric code for the currency.
+ * <p>Note: If the ISO 4217 numeric code is not assigned for the currency or
+ * the currency is unknown, this function returns 0.
+ *
+ * @param currency null-terminated 3-letter ISO 4217 code
+ * @return The ISO 4217 numeric code of the currency
+ * @stable ICU 49
+ */
+U_CAPI int32_t U_EXPORT2
+ucurr_getNumericCode(const UChar* currency);
+
+#endif /* #if !UCONFIG_NO_FORMATTING */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/udata.h b/thirdparty/icu4c/common/unicode/udata.h
new file mode 100644
index 0000000000..6caa849c42
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/udata.h
@@ -0,0 +1,440 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  udata.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999oct25
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UDATA_H__
+#define __UDATA_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+U_CDECL_BEGIN
+
+/**
+ * \file
+ * \brief C API: Data loading interface
+ *
+ * <h2>Information about data loading interface</h2>
+ *
+ * This API is used to find and efficiently load data for ICU and applications
+ * using ICU. It provides an abstract interface that specifies a data type and
+ * name to find and load the data. Normally this API is used by other ICU APIs
+ * to load required data out of the ICU data library, but it can be used to
+ * load data out of other places.
+ *
+ * See the User Guide Data Management chapter.
+ */
+ 
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * Character used to separate package names from tree names 
+ * @internal ICU 3.0
+ */
+#define U_TREE_SEPARATOR '-'
+
+/**
+ * String used to separate package names from tree names 
+ * @internal ICU 3.0
+ */
+#define U_TREE_SEPARATOR_STRING "-"
+
+/**
+ * Character used to separate parts of entry names
+ * @internal ICU 3.0
+ */
+#define U_TREE_ENTRY_SEP_CHAR '/'
+
+/**
+ * String used to separate parts of entry names
+ * @internal ICU 3.0
+ */
+#define U_TREE_ENTRY_SEP_STRING "/"
+
+/**
+ * Alias for standard ICU data 
+ * @internal ICU 3.0
+ */
+#define U_ICUDATA_ALIAS "ICUDATA"
+
+#endif /* U_HIDE_INTERNAL_API */
+
+/**
+ * UDataInfo contains the properties about the requested data.
+ * This is meta data.
+ *
+ * <p>This structure may grow in the future, indicated by the
+ * <code>size</code> field.</p>
+ *
+ * <p>ICU data must be at least 8-aligned, and should be 16-aligned.
+ * The UDataInfo struct begins 4 bytes after the start of the data item,
+ * so it is 4-aligned.
+ *
+ * <p>The platform data property fields help determine if a data
+ * file can be efficiently used on a given machine.
+ * The particular fields are of importance only if the data
+ * is affected by the properties - if there is integer data
+ * with word sizes > 1 byte, char* text, or UChar* text.</p>
+ *
+ * <p>The implementation for the <code>udata_open[Choice]()</code>
+ * functions may reject data based on the value in <code>isBigEndian</code>.
+ * No other field is used by the <code>udata</code> API implementation.</p>
+ *
+ * <p>The <code>dataFormat</code> may be used to identify
+ * the kind of data, e.g. a converter table.</p>
+ *
+ * <p>The <code>formatVersion</code> field should be used to
+ * make sure that the format can be interpreted.
+ * It may be a good idea to check only for the one or two highest
+ * of the version elements to allow the data memory to
+ * get more or somewhat rearranged contents, for as long
+ * as the using code can still interpret the older contents.</p>
+ *
+ * <p>The <code>dataVersion</code> field is intended to be a
+ * common place to store the source version of the data;
+ * for data from the Unicode character database, this could
+ * reflect the Unicode version.</p>
+ *
+ * @stable ICU 2.0
+ */
+typedef struct {
+    /** sizeof(UDataInfo)
+     *  @stable ICU 2.0 */
+    uint16_t size;
+
+    /** unused, set to 0 
+     *  @stable ICU 2.0*/
+    uint16_t reservedWord;
+
+    /* platform data properties */
+    /** 0 for little-endian machine, 1 for big-endian
+     *  @stable ICU 2.0 */
+    uint8_t isBigEndian;
+
+    /** see U_CHARSET_FAMILY values in utypes.h 
+     *  @stable ICU 2.0*/
+    uint8_t charsetFamily;
+
+    /** sizeof(UChar), one of { 1, 2, 4 } 
+     *  @stable ICU 2.0*/
+    uint8_t sizeofUChar;
+
+    /** unused, set to 0 
+     *  @stable ICU 2.0*/
+    uint8_t reservedByte;
+
+    /** data format identifier 
+     *  @stable ICU 2.0*/
+    uint8_t dataFormat[4];
+
+    /** versions: [0] major [1] minor [2] milli [3] micro 
+     *  @stable ICU 2.0*/
+    uint8_t formatVersion[4];
+
+    /** versions: [0] major [1] minor [2] milli [3] micro 
+     *  @stable ICU 2.0*/
+    uint8_t dataVersion[4];
+} UDataInfo;
+
+/* API for reading data -----------------------------------------------------*/
+
+/**
+ * Forward declaration of the data memory type.
+ * @stable ICU 2.0
+ */
+typedef struct UDataMemory UDataMemory;
+
+/**
+ * Callback function for udata_openChoice().
+ * @param context parameter passed into <code>udata_openChoice()</code>.
+ * @param type The type of the data as passed into <code>udata_openChoice()</code>.
+ *             It may be <code>NULL</code>.
+ * @param name The name of the data as passed into <code>udata_openChoice()</code>.
+ * @param pInfo A pointer to the <code>UDataInfo</code> structure
+ *              of data that has been loaded and will be returned
+ *              by <code>udata_openChoice()</code> if this function
+ *              returns <code>true</code>.
+ * @return true if the current data memory is acceptable
+ * @stable ICU 2.0
+ */
+typedef UBool U_CALLCONV
+UDataMemoryIsAcceptable(void *context,
+                        const char *type, const char *name,
+                        const UDataInfo *pInfo);
+
+
+/**
+ * Convenience function.
+ * This function works the same as <code>udata_openChoice</code>
+ * except that any data that matches the type and name
+ * is assumed to be acceptable.
+ * @param path Specifies an absolute path and/or a basename for the
+ *             finding of the data in the file system.
+ *             <code>NULL</code> for ICU data.
+ * @param type A string that specifies the type of data to be loaded.
+ *             For example, resource bundles are loaded with type "res",
+ *             conversion tables with type "cnv".
+ *             This may be <code>NULL</code> or empty.
+ * @param name A string that specifies the name of the data.
+ * @param pErrorCode An ICU UErrorCode parameter. It must not be <code>NULL</code>.
+ * @return A pointer (handle) to a data memory object, or <code>NULL</code>
+ *         if an error occurs. Call <code>udata_getMemory()</code>
+ *         to get a pointer to the actual data.
+ *
+ * @see udata_openChoice
+ * @stable ICU 2.0
+ */
+U_CAPI UDataMemory * U_EXPORT2
+udata_open(const char *path, const char *type, const char *name,
+           UErrorCode *pErrorCode);
+
+/**
+ * Data loading function.
+ * This function is used to find and load efficiently data for
+ * ICU and applications using ICU.
+ * It provides an abstract interface that allows to specify a data
+ * type and name to find and load the data.
+ *
+ * <p>The implementation depends on platform properties and user preferences
+ * and may involve loading shared libraries (DLLs), mapping
+ * files into memory, or fopen()/fread() files.
+ * It may also involve using static memory or database queries etc.
+ * Several or all data items may be combined into one entity
+ * (DLL, memory-mappable file).</p>
+ *
+ * <p>The data is always preceded by a header that includes
+ * a <code>UDataInfo</code> structure.
+ * The caller's <code>isAcceptable()</code> function is called to make
+ * sure that the data is useful. It may be called several times if it
+ * rejects the data and there is more than one location with data
+ * matching the type and name.</p>
+ *
+ * <p>If <code>path==NULL</code>, then ICU data is loaded.
+ * Otherwise, it is separated into a basename and a basename-less directory string.
+ * The basename is used as the data package name, and the directory is
+ * logically prepended to the ICU data directory string.</p>
+ *
+ * <p>For details about ICU data loading see the User Guide
+ * Data Management chapter. (http://icu-project.org/userguide/icudata.html)</p>
+ *
+ * @param path Specifies an absolute path and/or a basename for the
+ *             finding of the data in the file system.
+ *             <code>NULL</code> for ICU data.
+ * @param type A string that specifies the type of data to be loaded.
+ *             For example, resource bundles are loaded with type "res",
+ *             conversion tables with type "cnv".
+ *             This may be <code>NULL</code> or empty.
+ * @param name A string that specifies the name of the data.
+ * @param isAcceptable This function is called to verify that loaded data
+ *                     is useful for the client code. If it returns false
+ *                     for all data items, then <code>udata_openChoice()</code>
+ *                     will return with an error.
+ * @param context Arbitrary parameter to be passed into isAcceptable.
+ * @param pErrorCode An ICU UErrorCode parameter. It must not be <code>NULL</code>.
+ * @return A pointer (handle) to a data memory object, or <code>NULL</code>
+ *         if an error occurs. Call <code>udata_getMemory()</code>
+ *         to get a pointer to the actual data.
+ * @stable ICU 2.0
+ */
+U_CAPI UDataMemory * U_EXPORT2
+udata_openChoice(const char *path, const char *type, const char *name,
+                 UDataMemoryIsAcceptable *isAcceptable, void *context,
+                 UErrorCode *pErrorCode);
+
+/**
+ * Close the data memory.
+ * This function must be called to allow the system to
+ * release resources associated with this data memory.
+ * @param pData The pointer to data memory object
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+udata_close(UDataMemory *pData);
+
+/**
+ * Get the pointer to the actual data inside the data memory.
+ * The data is read-only.
+ *
+ * ICU data must be at least 8-aligned, and should be 16-aligned.
+ *
+ * @param pData The pointer to data memory object
+ * @stable ICU 2.0
+ */
+U_CAPI const void * U_EXPORT2
+udata_getMemory(UDataMemory *pData);
+
+/**
+ * Get the information from the data memory header.
+ * This allows to get access to the header containing
+ * platform data properties etc. which is not part of
+ * the data itself and can therefore not be accessed
+ * via the pointer that <code>udata_getMemory()</code> returns.
+ *
+ * @param pData pointer to the data memory object
+ * @param pInfo pointer to a UDataInfo object;
+ *              its <code>size</code> field must be set correctly,
+ *              typically to <code>sizeof(UDataInfo)</code>.
+ *
+ * <code>*pInfo</code> will be filled with the UDataInfo structure
+ * in the data memory object. If this structure is smaller than
+ * <code>pInfo->size</code>, then the <code>size</code> will be
+ * adjusted and only part of the structure will be filled.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+udata_getInfo(UDataMemory *pData, UDataInfo *pInfo);
+
+/**
+ * This function bypasses the normal ICU data loading process and
+ * allows you to force ICU's system data to come out of a user-specified
+ * area in memory.
+ *
+ * ICU data must be at least 8-aligned, and should be 16-aligned.
+ * See https://unicode-org.github.io/icu/userguide/icudata
+ *
+ * The format of this data is that of the icu common data file, as is
+ * generated by the pkgdata tool with mode=common or mode=dll.
+ * You can read in a whole common mode file and pass the address to the start of the
+ * data, or (with the appropriate link options) pass in the pointer to
+ * the data that has been loaded from a dll by the operating system,
+ * as shown in this code:
+ *
+ *       extern const char U_IMPORT U_ICUDATA_ENTRY_POINT [];
+ *        // U_ICUDATA_ENTRY_POINT is same as entry point specified to pkgdata tool
+ *       UErrorCode  status = U_ZERO_ERROR;
+ *
+ *       udata_setCommonData(&U_ICUDATA_ENTRY_POINT, &status);
+ *
+ * It is important that the declaration be as above. The entry point
+ * must not be declared as an extern void*.
+ *
+ * Starting with ICU 4.4, it is possible to set several data packages,
+ * one per call to this function.
+ * udata_open() will look for data in the multiple data packages in the order
+ * in which they were set.
+ * The position of the linked-in or default-name ICU .data package in the
+ * search list depends on when the first data item is loaded that is not contained
+ * in the already explicitly set packages.
+ * If data was loaded implicitly before the first call to this function
+ * (for example, via opening a converter, constructing a UnicodeString
+ * from default-codepage data, using formatting or collation APIs, etc.),
+ * then the default data will be first in the list.
+ *
+ * This function has no effect on application (non ICU) data.  See udata_setAppData()
+ * for similar functionality for application data.
+ *
+ * @param data pointer to ICU common data
+ * @param err outgoing error status <code>U_USING_DEFAULT_WARNING, U_UNSUPPORTED_ERROR</code>
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+udata_setCommonData(const void *data, UErrorCode *err);
+
+
+/**
+ * This function bypasses the normal ICU data loading process for application-specific
+ * data and allows you to force the it to come out of a user-specified
+ * pointer.
+ *
+ * ICU data must be at least 8-aligned, and should be 16-aligned.
+ * See https://unicode-org.github.io/icu/userguide/icudata
+ *
+ * The format of this data is that of the icu common data file, like 'icudt26l.dat'
+ * or the corresponding shared library (DLL) file.
+ * The application must read in or otherwise construct an image of the data and then
+ * pass the address of it to this function.
+ *
+ *
+ * Warning:  setAppData will set a U_USING_DEFAULT_WARNING code if
+ *           data with the specifed path that has already been opened, or
+ *           if setAppData with the same path has already been called.
+ *           Any such calls to setAppData will have no effect.
+ *
+ *
+ * @param packageName the package name by which the application will refer
+ *             to (open) this data
+ * @param data pointer to the data
+ * @param err outgoing error status <code>U_USING_DEFAULT_WARNING, U_UNSUPPORTED_ERROR</code>
+ * @see udata_setCommonData
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+udata_setAppData(const char *packageName, const void *data, UErrorCode *err);
+
+/**
+ * Possible settings for udata_setFileAccess()
+ * @see udata_setFileAccess
+ * @stable ICU 3.4
+ */
+typedef enum UDataFileAccess {
+    /** ICU looks for data in single files first, then in packages. (default) @stable ICU 3.4 */
+    UDATA_FILES_FIRST,
+    /** An alias for the default access mode. @stable ICU 3.4 */
+    UDATA_DEFAULT_ACCESS = UDATA_FILES_FIRST,
+    /** ICU only loads data from packages, not from single files. @stable ICU 3.4 */
+    UDATA_ONLY_PACKAGES,
+    /** ICU loads data from packages first, and only from single files
+        if the data cannot be found in a package. @stable ICU 3.4 */
+    UDATA_PACKAGES_FIRST,
+    /** ICU does not access the file system for data loading. @stable ICU 3.4 */
+    UDATA_NO_FILES,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Number of real UDataFileAccess values.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UDATA_FILE_ACCESS_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} UDataFileAccess;
+
+/**
+ * This function may be called to control how ICU loads data. It must be called
+ * before any ICU data is loaded, including application data loaded with 
+ * ures/ResourceBundle or udata APIs. This function is not multithread safe.  
+ * The results of calling it while other threads are loading data are undefined.
+ * @param access The type of file access to be used
+ * @param status Error code.
+ * @see UDataFileAccess
+ * @stable ICU 3.4 
+ */
+U_CAPI void U_EXPORT2
+udata_setFileAccess(UDataFileAccess access, UErrorCode *status);
+
+U_CDECL_END
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUDataMemoryPointer
+ * "Smart pointer" class, closes a UDataMemory via udata_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUDataMemoryPointer, UDataMemory, udata_close);
+
+U_NAMESPACE_END
+
+#endif  // U_SHOW_CPLUSPLUS_API
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/udisplaycontext.h b/thirdparty/icu4c/common/unicode/udisplaycontext.h
new file mode 100644
index 0000000000..6e14217980
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/udisplaycontext.h
@@ -0,0 +1,173 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*****************************************************************************************
+* Copyright (C) 2014-2016, International Business Machines
+* Corporation and others. All Rights Reserved.
+*****************************************************************************************
+*/
+
+#ifndef UDISPLAYCONTEXT_H
+#define UDISPLAYCONTEXT_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
+/**
+ * \file
+ * \brief C API: Display context types (enum values)
+ */
+
+/**
+ * Display context types, for getting values of a particular setting.
+ * Note, the specific numeric values are internal and may change.
+ * @stable ICU 51
+ */
+enum UDisplayContextType {
+    /**
+     * Type to retrieve the dialect handling setting, e.g.
+     * UDISPCTX_STANDARD_NAMES or UDISPCTX_DIALECT_NAMES.
+     * @stable ICU 51
+     */
+    UDISPCTX_TYPE_DIALECT_HANDLING = 0,
+    /**
+     * Type to retrieve the capitalization context setting, e.g.
+     * UDISPCTX_CAPITALIZATION_NONE, UDISPCTX_CAPITALIZATION_FOR_MIDDLE_OF_SENTENCE,
+     * UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE, etc.
+     * @stable ICU 51
+     */
+    UDISPCTX_TYPE_CAPITALIZATION = 1,
+    /**
+     * Type to retrieve the display length setting, e.g.
+     * UDISPCTX_LENGTH_FULL, UDISPCTX_LENGTH_SHORT.
+     * @stable ICU 54
+     */
+    UDISPCTX_TYPE_DISPLAY_LENGTH = 2,
+    /**
+     * Type to retrieve the substitute handling setting, e.g.
+     * UDISPCTX_SUBSTITUTE, UDISPCTX_NO_SUBSTITUTE.
+     * @stable ICU 58
+     */
+    UDISPCTX_TYPE_SUBSTITUTE_HANDLING = 3
+};
+/**
+*  @stable ICU 51
+*/
+typedef enum UDisplayContextType UDisplayContextType;
+
+/**
+ * Display context settings.
+ * Note, the specific numeric values are internal and may change.
+ * @stable ICU 51
+ */
+enum UDisplayContext {
+    /**
+     * ================================
+     * DIALECT_HANDLING can be set to one of UDISPCTX_STANDARD_NAMES or
+     * UDISPCTX_DIALECT_NAMES. Use UDisplayContextType UDISPCTX_TYPE_DIALECT_HANDLING
+     * to get the value.
+     */
+    /**
+     * A possible setting for DIALECT_HANDLING:
+     * use standard names when generating a locale name,
+     * e.g. en_GB displays as 'English (United Kingdom)'.
+     * @stable ICU 51
+     */
+    UDISPCTX_STANDARD_NAMES = (UDISPCTX_TYPE_DIALECT_HANDLING<<8) + 0,
+    /**
+     * A possible setting for DIALECT_HANDLING:
+     * use dialect names, when generating a locale name,
+     * e.g. en_GB displays as 'British English'.
+     * @stable ICU 51
+     */
+    UDISPCTX_DIALECT_NAMES = (UDISPCTX_TYPE_DIALECT_HANDLING<<8) + 1,
+    /**
+     * ================================
+     * CAPITALIZATION can be set to one of UDISPCTX_CAPITALIZATION_NONE,
+     * UDISPCTX_CAPITALIZATION_FOR_MIDDLE_OF_SENTENCE,
+     * UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE,
+     * UDISPCTX_CAPITALIZATION_FOR_UI_LIST_OR_MENU, or
+     * UDISPCTX_CAPITALIZATION_FOR_STANDALONE.
+     * Use UDisplayContextType UDISPCTX_TYPE_CAPITALIZATION to get the value.
+     */
+    /**
+     * The capitalization context to be used is unknown (this is the default value).
+     * @stable ICU 51
+     */
+    UDISPCTX_CAPITALIZATION_NONE = (UDISPCTX_TYPE_CAPITALIZATION<<8) + 0,
+    /**
+     * The capitalization context if a date, date symbol or display name is to be
+     * formatted with capitalization appropriate for the middle of a sentence.
+     * @stable ICU 51
+     */
+    UDISPCTX_CAPITALIZATION_FOR_MIDDLE_OF_SENTENCE = (UDISPCTX_TYPE_CAPITALIZATION<<8) + 1,
+    /**
+     * The capitalization context if a date, date symbol or display name is to be
+     * formatted with capitalization appropriate for the beginning of a sentence.
+     * @stable ICU 51
+     */
+    UDISPCTX_CAPITALIZATION_FOR_BEGINNING_OF_SENTENCE = (UDISPCTX_TYPE_CAPITALIZATION<<8) + 2,
+    /**
+     * The capitalization context if a date, date symbol or display name is to be
+     * formatted with capitalization appropriate for a user-interface list or menu item.
+     * @stable ICU 51
+     */
+    UDISPCTX_CAPITALIZATION_FOR_UI_LIST_OR_MENU = (UDISPCTX_TYPE_CAPITALIZATION<<8) + 3,
+    /**
+     * The capitalization context if a date, date symbol or display name is to be
+     * formatted with capitalization appropriate for stand-alone usage such as an
+     * isolated name on a calendar page.
+     * @stable ICU 51
+     */
+    UDISPCTX_CAPITALIZATION_FOR_STANDALONE = (UDISPCTX_TYPE_CAPITALIZATION<<8) + 4,
+    /**
+     * ================================
+     * DISPLAY_LENGTH can be set to one of UDISPCTX_LENGTH_FULL or
+     * UDISPCTX_LENGTH_SHORT. Use UDisplayContextType UDISPCTX_TYPE_DISPLAY_LENGTH
+     * to get the value.
+     */
+    /**
+     * A possible setting for DISPLAY_LENGTH:
+     * use full names when generating a locale name,
+     * e.g. "United States" for US.
+     * @stable ICU 54
+     */
+    UDISPCTX_LENGTH_FULL = (UDISPCTX_TYPE_DISPLAY_LENGTH<<8) + 0,
+    /**
+     * A possible setting for DISPLAY_LENGTH:
+     * use short names when generating a locale name,
+     * e.g. "U.S." for US.
+     * @stable ICU 54
+     */
+    UDISPCTX_LENGTH_SHORT = (UDISPCTX_TYPE_DISPLAY_LENGTH<<8) + 1,
+    /**
+     * ================================
+     * SUBSTITUTE_HANDLING can be set to one of UDISPCTX_SUBSTITUTE or
+     * UDISPCTX_NO_SUBSTITUTE. Use UDisplayContextType UDISPCTX_TYPE_SUBSTITUTE_HANDLING
+     * to get the value.
+     */
+    /**
+     * A possible setting for SUBSTITUTE_HANDLING:
+     * Returns a fallback value (e.g., the input code) when no data is available.
+     * This is the default value.
+     * @stable ICU 58
+     */
+    UDISPCTX_SUBSTITUTE = (UDISPCTX_TYPE_SUBSTITUTE_HANDLING<<8) + 0,
+    /**
+     * A possible setting for SUBSTITUTE_HANDLING:
+     * Returns a null value with error code set to U_ILLEGAL_ARGUMENT_ERROR when no
+     * data is available.
+     * @stable ICU 58
+     */
+    UDISPCTX_NO_SUBSTITUTE = (UDISPCTX_TYPE_SUBSTITUTE_HANDLING<<8) + 1
+
+};
+/**
+*  @stable ICU 51
+*/
+typedef enum UDisplayContext UDisplayContext;
+
+#endif /* #if !UCONFIG_NO_FORMATTING */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uenum.h b/thirdparty/icu4c/common/unicode/uenum.h
new file mode 100644
index 0000000000..d9c893e06d
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uenum.h
@@ -0,0 +1,209 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uenum.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2002jul08
+*   created by: Vladimir Weinstein
+*/
+
+#ifndef __UENUM_H
+#define __UENUM_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+
+U_NAMESPACE_BEGIN
+class StringEnumeration;
+U_NAMESPACE_END
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C API: String Enumeration 
+ */
+ 
+/**
+ * An enumeration object.
+ * For usage in C programs.
+ * @stable ICU 2.2
+ */
+struct UEnumeration;
+/** structure representing an enumeration object instance @stable ICU 2.2 */
+typedef struct UEnumeration UEnumeration;
+
+/**
+ * Disposes of resources in use by the iterator.  If en is NULL,
+ * does nothing.  After this call, any char* or UChar* pointer
+ * returned by uenum_unext() or uenum_next() is invalid.
+ * @param en UEnumeration structure pointer
+ * @stable ICU 2.2
+ */
+U_CAPI void U_EXPORT2
+uenum_close(UEnumeration* en);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUEnumerationPointer
+ * "Smart pointer" class, closes a UEnumeration via uenum_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUEnumerationPointer, UEnumeration, uenum_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Returns the number of elements that the iterator traverses.  If
+ * the iterator is out-of-sync with its service, status is set to
+ * U_ENUM_OUT_OF_SYNC_ERROR.
+ * This is a convenience function. It can end up being very
+ * expensive as all the items might have to be pre-fetched (depending
+ * on the type of data being traversed). Use with caution and only 
+ * when necessary.
+ * @param en UEnumeration structure pointer
+ * @param status error code, can be U_ENUM_OUT_OF_SYNC_ERROR if the
+ *               iterator is out of sync.
+ * @return number of elements in the iterator
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+uenum_count(UEnumeration* en, UErrorCode* status);
+
+/**
+ * Returns the next element in the iterator's list.  If there are
+ * no more elements, returns NULL.  If the iterator is out-of-sync
+ * with its service, status is set to U_ENUM_OUT_OF_SYNC_ERROR and
+ * NULL is returned.  If the native service string is a char* string,
+ * it is converted to UChar* with the invariant converter.
+ * The result is terminated by (UChar)0.
+ * @param en the iterator object
+ * @param resultLength pointer to receive the length of the result
+ *                     (not including the terminating \\0).
+ *                     If the pointer is NULL it is ignored.
+ * @param status the error code, set to U_ENUM_OUT_OF_SYNC_ERROR if
+ *               the iterator is out of sync with its service.
+ * @return a pointer to the string.  The string will be
+ *         zero-terminated.  The return pointer is owned by this iterator
+ *         and must not be deleted by the caller.  The pointer is valid
+ *         until the next call to any uenum_... method, including
+ *         uenum_next() or uenum_unext().  When all strings have been
+ *         traversed, returns NULL.
+ * @stable ICU 2.2
+ */
+U_CAPI const UChar* U_EXPORT2
+uenum_unext(UEnumeration* en,
+            int32_t* resultLength,
+            UErrorCode* status);
+
+/**
+ * Returns the next element in the iterator's list.  If there are
+ * no more elements, returns NULL.  If the iterator is out-of-sync
+ * with its service, status is set to U_ENUM_OUT_OF_SYNC_ERROR and
+ * NULL is returned.  If the native service string is a UChar*
+ * string, it is converted to char* with the invariant converter.
+ * The result is terminated by (char)0.  If the conversion fails
+ * (because a character cannot be converted) then status is set to
+ * U_INVARIANT_CONVERSION_ERROR and the return value is undefined
+ * (but non-NULL).
+ * @param en the iterator object
+ * @param resultLength pointer to receive the length of the result
+ *                     (not including the terminating \\0).
+ *                     If the pointer is NULL it is ignored.
+ * @param status the error code, set to U_ENUM_OUT_OF_SYNC_ERROR if
+ *               the iterator is out of sync with its service.  Set to
+ *               U_INVARIANT_CONVERSION_ERROR if the underlying native string is
+ *               UChar* and conversion to char* with the invariant converter
+ *               fails. This error pertains only to current string, so iteration
+ *               might be able to continue successfully.
+ * @return a pointer to the string.  The string will be
+ *         zero-terminated.  The return pointer is owned by this iterator
+ *         and must not be deleted by the caller.  The pointer is valid
+ *         until the next call to any uenum_... method, including
+ *         uenum_next() or uenum_unext().  When all strings have been
+ *         traversed, returns NULL.
+ * @stable ICU 2.2
+ */
+U_CAPI const char* U_EXPORT2
+uenum_next(UEnumeration* en,
+           int32_t* resultLength,
+           UErrorCode* status);
+
+/**
+ * Resets the iterator to the current list of service IDs.  This
+ * re-establishes sync with the service and rewinds the iterator
+ * to start at the first element.
+ * @param en the iterator object
+ * @param status the error code, set to U_ENUM_OUT_OF_SYNC_ERROR if
+ *               the iterator is out of sync with its service.  
+ * @stable ICU 2.2
+ */
+U_CAPI void U_EXPORT2
+uenum_reset(UEnumeration* en, UErrorCode* status);
+
+#if U_SHOW_CPLUSPLUS_API
+
+/**
+ * Given a StringEnumeration, wrap it in a UEnumeration.  The
+ * StringEnumeration is adopted; after this call, the caller must not
+ * delete it (regardless of error status).
+ * @param adopted the C++ StringEnumeration to be wrapped in a UEnumeration.
+ * @param ec the error code.
+ * @return a UEnumeration wrapping the adopted StringEnumeration.
+ * @stable ICU 4.2
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openFromStringEnumeration(icu::StringEnumeration* adopted, UErrorCode* ec);
+
+#endif
+
+/**
+ * Given an array of const UChar* strings, return a UEnumeration.  String pointers from 0..count-1 must not be null.
+ * Do not free or modify either the string array or the characters it points to until this object has been destroyed with uenum_close.
+ * \snippet test/cintltst/uenumtst.c uenum_openUCharStringsEnumeration
+ * @param strings array of const UChar* strings (each null terminated). All storage is owned by the caller.
+ * @param count length of the array
+ * @param ec error code
+ * @return the new UEnumeration object. Caller is responsible for calling uenum_close to free memory.
+ * @see uenum_close
+ * @stable ICU 50
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openUCharStringsEnumeration(const UChar* const strings[], int32_t count,
+                                 UErrorCode* ec);
+
+/**
+ * Given an array of const char* strings (invariant chars only), return a UEnumeration.  String pointers from 0..count-1 must not be null.
+ * Do not free or modify either the string array or the characters it points to until this object has been destroyed with uenum_close.
+ * \snippet test/cintltst/uenumtst.c uenum_openCharStringsEnumeration
+ * @param strings array of char* strings (each null terminated).  All storage is owned by the caller.
+ * @param count length of the array
+ * @param ec error code
+ * @return the new UEnumeration object. Caller is responsible for calling uenum_close to free memory
+ * @see uenum_close
+ * @stable ICU 50
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openCharStringsEnumeration(const char* const strings[], int32_t count,
+                                 UErrorCode* ec);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uidna.h b/thirdparty/icu4c/common/unicode/uidna.h
new file mode 100644
index 0000000000..24a81ceadd
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uidna.h
@@ -0,0 +1,776 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ *
+ *   Copyright (C) 2003-2014, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *******************************************************************************
+ *   file name:  uidna.h
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2003feb1
+ *   created by: Ram Viswanadha
+ */
+
+#ifndef __UIDNA_H__
+#define __UIDNA_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include <stdbool.h>
+#include "unicode/parseerr.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C API: Internationalizing Domain Names in Applications (IDNA)
+ *
+ * IDNA2008 is implemented according to UTS #46, see the IDNA C++ class in idna.h.
+ *
+ * The C API functions which do take a UIDNA * service object pointer
+ * implement UTS #46 and IDNA2008.
+ *
+ * IDNA2003 is obsolete.
+ * The C API functions which do not take a service object pointer
+ * implement IDNA2003. They are all deprecated.
+ */
+
+/*
+ * IDNA option bit set values.
+ */
+enum {
+    /**
+     * Default options value: None of the other options are set.
+     * For use in static worker and factory methods.
+     * @stable ICU 2.6
+     */
+    UIDNA_DEFAULT=0,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * Option to allow unassigned code points in domain names and labels.
+     * For use in static worker and factory methods.
+     * <p>This option is ignored by the UTS46 implementation.
+     * (UTS #46 disallows unassigned code points.)
+     * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+     */
+    UIDNA_ALLOW_UNASSIGNED=1,
+#endif  /* U_HIDE_DEPRECATED_API */
+    /**
+     * Option to check whether the input conforms to the STD3 ASCII rules,
+     * for example the restriction of labels to LDH characters
+     * (ASCII Letters, Digits and Hyphen-Minus).
+     * For use in static worker and factory methods.
+     * @stable ICU 2.6
+     */
+    UIDNA_USE_STD3_RULES=2,
+    /**
+     * IDNA option to check for whether the input conforms to the BiDi rules.
+     * For use in static worker and factory methods.
+     * <p>This option is ignored by the IDNA2003 implementation.
+     * (IDNA2003 always performs a BiDi check.)
+     * @stable ICU 4.6
+     */
+    UIDNA_CHECK_BIDI=4,
+    /**
+     * IDNA option to check for whether the input conforms to the CONTEXTJ rules.
+     * For use in static worker and factory methods.
+     * <p>This option is ignored by the IDNA2003 implementation.
+     * (The CONTEXTJ check is new in IDNA2008.)
+     * @stable ICU 4.6
+     */
+    UIDNA_CHECK_CONTEXTJ=8,
+    /**
+     * IDNA option for nontransitional processing in ToASCII().
+     * For use in static worker and factory methods.
+     * <p>By default, ToASCII() uses transitional processing.
+     * <p>This option is ignored by the IDNA2003 implementation.
+     * (This is only relevant for compatibility of newer IDNA implementations with IDNA2003.)
+     * @stable ICU 4.6
+     */
+    UIDNA_NONTRANSITIONAL_TO_ASCII=0x10,
+    /**
+     * IDNA option for nontransitional processing in ToUnicode().
+     * For use in static worker and factory methods.
+     * <p>By default, ToUnicode() uses transitional processing.
+     * <p>This option is ignored by the IDNA2003 implementation.
+     * (This is only relevant for compatibility of newer IDNA implementations with IDNA2003.)
+     * @stable ICU 4.6
+     */
+    UIDNA_NONTRANSITIONAL_TO_UNICODE=0x20,
+    /**
+     * IDNA option to check for whether the input conforms to the CONTEXTO rules.
+     * For use in static worker and factory methods.
+     * <p>This option is ignored by the IDNA2003 implementation.
+     * (The CONTEXTO check is new in IDNA2008.)
+     * <p>This is for use by registries for IDNA2008 conformance.
+     * UTS #46 does not require the CONTEXTO check.
+     * @stable ICU 49
+     */
+    UIDNA_CHECK_CONTEXTO=0x40
+};
+
+/**
+ * Opaque C service object type for the new IDNA API.
+ * @stable ICU 4.6
+ */
+struct UIDNA;
+typedef struct UIDNA UIDNA;  /**< C typedef for struct UIDNA. @stable ICU 4.6 */
+
+/**
+ * Returns a UIDNA instance which implements UTS #46.
+ * Returns an unmodifiable instance, owned by the caller.
+ * Cache it for multiple operations, and uidna_close() it when done.
+ * The instance is thread-safe, that is, it can be used concurrently.
+ *
+ * For details about the UTS #46 implementation see the IDNA C++ class in idna.h.
+ *
+ * @param options Bit set to modify the processing and error checking.
+ *                See option bit set values in uidna.h.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the UTS #46 UIDNA instance, if successful
+ * @stable ICU 4.6
+ */
+U_CAPI UIDNA * U_EXPORT2
+uidna_openUTS46(uint32_t options, UErrorCode *pErrorCode);
+
+/**
+ * Closes a UIDNA instance.
+ * @param idna UIDNA instance to be closed
+ * @stable ICU 4.6
+ */
+U_CAPI void U_EXPORT2
+uidna_close(UIDNA *idna);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUIDNAPointer
+ * "Smart pointer" class, closes a UIDNA via uidna_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.6
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUIDNAPointer, UIDNA, uidna_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Output container for IDNA processing errors.
+ * Initialize with UIDNA_INFO_INITIALIZER:
+ * \code
+ * UIDNAInfo info = UIDNA_INFO_INITIALIZER;
+ * int32_t length = uidna_nameToASCII(..., &info, &errorCode);
+ * if(U_SUCCESS(errorCode) && info.errors!=0) { ... }
+ * \endcode
+ * @stable ICU 4.6
+ */
+typedef struct UIDNAInfo {
+    /** sizeof(UIDNAInfo) @stable ICU 4.6 */
+    int16_t size;
+    /**
+     * Set to true if transitional and nontransitional processing produce different results.
+     * For details see C++ IDNAInfo::isTransitionalDifferent().
+     * @stable ICU 4.6
+     */
+    UBool isTransitionalDifferent;
+    UBool reservedB3;  /**< Reserved field, do not use. @internal */
+    /**
+     * Bit set indicating IDNA processing errors. 0 if no errors.
+     * See UIDNA_ERROR_... constants.
+     * @stable ICU 4.6
+     */
+    uint32_t errors;
+    int32_t reservedI2;  /**< Reserved field, do not use. @internal */
+    int32_t reservedI3;  /**< Reserved field, do not use. @internal */
+} UIDNAInfo;
+
+/**
+ * Static initializer for a UIDNAInfo struct.
+ * @stable ICU 4.6
+ */
+#define UIDNA_INFO_INITIALIZER { \
+    (int16_t)sizeof(UIDNAInfo), \
+    false, false, \
+    0, 0, 0 }
+
+/**
+ * Converts a single domain name label into its ASCII form for DNS lookup.
+ * If any processing step fails, then pInfo->errors will be non-zero and
+ * the result might not be an ASCII string.
+ * The label might be modified according to the types of errors.
+ * Labels with severe errors will be left in (or turned into) their Unicode form.
+ *
+ * The UErrorCode indicates an error only in exceptional cases,
+ * such as a U_MEMORY_ALLOCATION_ERROR.
+ *
+ * @param idna UIDNA instance
+ * @param label Input domain name label
+ * @param length Label length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_labelToASCII(const UIDNA *idna,
+                   const UChar *label, int32_t length,
+                   UChar *dest, int32_t capacity,
+                   UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a single domain name label into its Unicode form for human-readable display.
+ * If any processing step fails, then pInfo->errors will be non-zero.
+ * The label might be modified according to the types of errors.
+ *
+ * The UErrorCode indicates an error only in exceptional cases,
+ * such as a U_MEMORY_ALLOCATION_ERROR.
+ *
+ * @param idna UIDNA instance
+ * @param label Input domain name label
+ * @param length Label length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_labelToUnicode(const UIDNA *idna,
+                     const UChar *label, int32_t length,
+                     UChar *dest, int32_t capacity,
+                     UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a whole domain name into its ASCII form for DNS lookup.
+ * If any processing step fails, then pInfo->errors will be non-zero and
+ * the result might not be an ASCII string.
+ * The domain name might be modified according to the types of errors.
+ * Labels with severe errors will be left in (or turned into) their Unicode form.
+ *
+ * The UErrorCode indicates an error only in exceptional cases,
+ * such as a U_MEMORY_ALLOCATION_ERROR.
+ *
+ * @param idna UIDNA instance
+ * @param name Input domain name
+ * @param length Domain name length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_nameToASCII(const UIDNA *idna,
+                  const UChar *name, int32_t length,
+                  UChar *dest, int32_t capacity,
+                  UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a whole domain name into its Unicode form for human-readable display.
+ * If any processing step fails, then pInfo->errors will be non-zero.
+ * The domain name might be modified according to the types of errors.
+ *
+ * The UErrorCode indicates an error only in exceptional cases,
+ * such as a U_MEMORY_ALLOCATION_ERROR.
+ *
+ * @param idna UIDNA instance
+ * @param name Input domain name
+ * @param length Domain name length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_nameToUnicode(const UIDNA *idna,
+                    const UChar *name, int32_t length,
+                    UChar *dest, int32_t capacity,
+                    UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/* UTF-8 versions of the processing methods --------------------------------- */
+
+/**
+ * Converts a single domain name label into its ASCII form for DNS lookup.
+ * UTF-8 version of uidna_labelToASCII(), same behavior.
+ *
+ * @param idna UIDNA instance
+ * @param label Input domain name label
+ * @param length Label length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_labelToASCII_UTF8(const UIDNA *idna,
+                        const char *label, int32_t length,
+                        char *dest, int32_t capacity,
+                        UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a single domain name label into its Unicode form for human-readable display.
+ * UTF-8 version of uidna_labelToUnicode(), same behavior.
+ *
+ * @param idna UIDNA instance
+ * @param label Input domain name label
+ * @param length Label length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_labelToUnicodeUTF8(const UIDNA *idna,
+                         const char *label, int32_t length,
+                         char *dest, int32_t capacity,
+                         UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a whole domain name into its ASCII form for DNS lookup.
+ * UTF-8 version of uidna_nameToASCII(), same behavior.
+ *
+ * @param idna UIDNA instance
+ * @param name Input domain name
+ * @param length Domain name length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_nameToASCII_UTF8(const UIDNA *idna,
+                       const char *name, int32_t length,
+                       char *dest, int32_t capacity,
+                       UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/**
+ * Converts a whole domain name into its Unicode form for human-readable display.
+ * UTF-8 version of uidna_nameToUnicode(), same behavior.
+ *
+ * @param idna UIDNA instance
+ * @param name Input domain name
+ * @param length Domain name length, or -1 if NUL-terminated
+ * @param dest Destination string buffer
+ * @param capacity Destination buffer capacity
+ * @param pInfo Output container of IDNA processing details.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return destination string length
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+uidna_nameToUnicodeUTF8(const UIDNA *idna,
+                        const char *name, int32_t length,
+                        char *dest, int32_t capacity,
+                        UIDNAInfo *pInfo, UErrorCode *pErrorCode);
+
+/*
+ * IDNA error bit set values.
+ * When a domain name or label fails a processing step or does not meet the
+ * validity criteria, then one or more of these error bits are set.
+ */
+enum {
+    /**
+     * A non-final domain name label (or the whole domain name) is empty.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_EMPTY_LABEL=1,
+    /**
+     * A domain name label is longer than 63 bytes.
+     * (See STD13/RFC1034 3.1. Name space specifications and terminology.)
+     * This is only checked in ToASCII operations, and only if the output label is all-ASCII.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_LABEL_TOO_LONG=2,
+    /**
+     * A domain name is longer than 255 bytes in its storage form.
+     * (See STD13/RFC1034 3.1. Name space specifications and terminology.)
+     * This is only checked in ToASCII operations, and only if the output domain name is all-ASCII.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_DOMAIN_NAME_TOO_LONG=4,
+    /**
+     * A label starts with a hyphen-minus ('-').
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_LEADING_HYPHEN=8,
+    /**
+     * A label ends with a hyphen-minus ('-').
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_TRAILING_HYPHEN=0x10,
+    /**
+     * A label contains hyphen-minus ('-') in the third and fourth positions.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_HYPHEN_3_4=0x20,
+    /**
+     * A label starts with a combining mark.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_LEADING_COMBINING_MARK=0x40,
+    /**
+     * A label or domain name contains disallowed characters.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_DISALLOWED=0x80,
+    /**
+     * A label starts with "xn--" but does not contain valid Punycode.
+     * That is, an xn-- label failed Punycode decoding.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_PUNYCODE=0x100,
+    /**
+     * A label contains a dot=full stop.
+     * This can occur in an input string for a single-label function.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_LABEL_HAS_DOT=0x200,
+    /**
+     * An ACE label does not contain a valid label string.
+     * The label was successfully ACE (Punycode) decoded but the resulting
+     * string had severe validation errors. For example,
+     * it might contain characters that are not allowed in ACE labels,
+     * or it might not be normalized.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_INVALID_ACE_LABEL=0x400,
+    /**
+     * A label does not meet the IDNA BiDi requirements (for right-to-left characters).
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_BIDI=0x800,
+    /**
+     * A label does not meet the IDNA CONTEXTJ requirements.
+     * @stable ICU 4.6
+     */
+    UIDNA_ERROR_CONTEXTJ=0x1000,
+    /**
+     * A label does not meet the IDNA CONTEXTO requirements for punctuation characters.
+     * Some punctuation characters "Would otherwise have been DISALLOWED"
+     * but are allowed in certain contexts. (RFC 5892)
+     * @stable ICU 49
+     */
+    UIDNA_ERROR_CONTEXTO_PUNCTUATION=0x2000,
+    /**
+     * A label does not meet the IDNA CONTEXTO requirements for digits.
+     * Arabic-Indic Digits (U+066x) must not be mixed with Extended Arabic-Indic Digits (U+06Fx).
+     * @stable ICU 49
+     */
+    UIDNA_ERROR_CONTEXTO_DIGITS=0x4000
+};
+
+#ifndef U_HIDE_DEPRECATED_API
+
+/* IDNA2003 API ------------------------------------------------------------- */
+
+/**
+ * IDNA2003: This function implements the ToASCII operation as defined in the IDNA RFC.
+ * This operation is done on <b>single labels</b> before sending it to something that expects
+ * ASCII names. A label is an individual part of a domain name. Labels are usually
+ * separated by dots; e.g. "www.example.com" is composed of 3 labels "www","example", and "com".
+ *
+ * IDNA2003 API Overview:
+ *
+ * The uidna_ API implements the IDNA protocol as defined in the IDNA RFC
+ * (http://www.ietf.org/rfc/rfc3490.txt).
+ * The RFC defines 2 operations: ToASCII and ToUnicode. Domain name labels
+ * containing non-ASCII code points are processed by the
+ * ToASCII operation before passing it to resolver libraries. Domain names
+ * that are obtained from resolver libraries are processed by the
+ * ToUnicode operation before displaying the domain name to the user.
+ * IDNA requires that implementations process input strings with Nameprep
+ * (http://www.ietf.org/rfc/rfc3491.txt),
+ * which is a profile of Stringprep (http://www.ietf.org/rfc/rfc3454.txt),
+ * and then with Punycode (http://www.ietf.org/rfc/rfc3492.txt).
+ * Implementations of IDNA MUST fully implement Nameprep and Punycode;
+ * neither Nameprep nor Punycode are optional.
+ * The input and output of ToASCII and ToUnicode operations are Unicode
+ * and are designed to be chainable, i.e., applying ToASCII or ToUnicode operations
+ * multiple times to an input string will yield the same result as applying the operation
+ * once.
+ * ToUnicode(ToUnicode(ToUnicode...(ToUnicode(string)))) == ToUnicode(string) 
+ * ToASCII(ToASCII(ToASCII...(ToASCII(string))) == ToASCII(string).
+ *
+ * @param src               Input UChar array containing label in Unicode.
+ * @param srcLength         Number of UChars in src, or -1 if NUL-terminated.
+ * @param dest              Output UChar array with ASCII (ACE encoded) label.
+ * @param destCapacity      Size of dest.
+ * @param options           A bit set of options:
+ *
+ *  - UIDNA_DEFAULT             Use default options, i.e., do not process unassigned code points
+ *                              and do not use STD3 ASCII rules
+ *                              If unassigned code points are found the operation fails with 
+ *                              U_UNASSIGNED_ERROR error code.
+ *
+ *  - UIDNA_ALLOW_UNASSIGNED    Unassigned values can be converted to ASCII for query operations
+ *                              If this option is set, the unassigned code points are in the input 
+ *                              are treated as normal Unicode code points.
+ *
+ *  - UIDNA_USE_STD3_RULES      Use STD3 ASCII rules for host name syntax restrictions
+ *                              If this option is set and the input does not satisfy STD3 rules,  
+ *                              the operation will fail with U_IDNA_STD3_ASCII_RULES_ERROR
+ *
+ * @param parseError        Pointer to UParseError struct to receive information on position 
+ *                          of error if an error is encountered. Can be NULL.
+ * @param status            ICU in/out error code parameter.
+ *                          U_INVALID_CHAR_FOUND if src contains
+ *                          unmatched single surrogates.
+ *                          U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                          too many code points.
+ *                          U_BUFFER_OVERFLOW_ERROR if destCapacity is not enough
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+uidna_toASCII(const UChar* src, int32_t srcLength, 
+              UChar* dest, int32_t destCapacity,
+              int32_t options,
+              UParseError* parseError,
+              UErrorCode* status);
+
+
+/**
+ * IDNA2003: This function implements the ToUnicode operation as defined in the IDNA RFC.
+ * This operation is done on <b>single labels</b> before sending it to something that expects
+ * Unicode names. A label is an individual part of a domain name. Labels are usually
+ * separated by dots; for e.g. "www.example.com" is composed of 3 labels "www","example", and "com".
+ *
+ * @param src               Input UChar array containing ASCII (ACE encoded) label.
+ * @param srcLength         Number of UChars in src, or -1 if NUL-terminated.
+ * @param dest Output       Converted UChar array containing Unicode equivalent of label.
+ * @param destCapacity      Size of dest.
+ * @param options           A bit set of options:
+ *
+ *  - UIDNA_DEFAULT             Use default options, i.e., do not process unassigned code points
+ *                              and do not use STD3 ASCII rules
+ *                              If unassigned code points are found the operation fails with 
+ *                              U_UNASSIGNED_ERROR error code.
+ *
+ *  - UIDNA_ALLOW_UNASSIGNED      Unassigned values can be converted to ASCII for query operations
+ *                              If this option is set, the unassigned code points are in the input 
+ *                              are treated as normal Unicode code points. <b> Note: </b> This option is 
+ *                              required on toUnicode operation because the RFC mandates 
+ *                              verification of decoded ACE input by applying toASCII and comparing
+ *                              its output with source
+ *
+ *  - UIDNA_USE_STD3_RULES      Use STD3 ASCII rules for host name syntax restrictions
+ *                              If this option is set and the input does not satisfy STD3 rules,  
+ *                              the operation will fail with U_IDNA_STD3_ASCII_RULES_ERROR
+ *
+ * @param parseError        Pointer to UParseError struct to receive information on position 
+ *                          of error if an error is encountered. Can be NULL.
+ * @param status            ICU in/out error code parameter.
+ *                          U_INVALID_CHAR_FOUND if src contains
+ *                          unmatched single surrogates.
+ *                          U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                          too many code points.
+ *                          U_BUFFER_OVERFLOW_ERROR if destCapacity is not enough
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+uidna_toUnicode(const UChar* src, int32_t srcLength,
+                UChar* dest, int32_t destCapacity,
+                int32_t options,
+                UParseError* parseError,
+                UErrorCode* status);
+
+
+/**
+ * IDNA2003: Convenience function that implements the IDNToASCII operation as defined in the IDNA RFC.
+ * This operation is done on complete domain names, e.g: "www.example.com". 
+ * It is important to note that this operation can fail. If it fails, then the input 
+ * domain name cannot be used as an Internationalized Domain Name and the application
+ * should have methods defined to deal with the failure.
+ *
+ * <b>Note:</b> IDNA RFC specifies that a conformant application should divide a domain name
+ * into separate labels, decide whether to apply allowUnassigned and useSTD3ASCIIRules on each, 
+ * and then convert. This function does not offer that level of granularity. The options once  
+ * set will apply to all labels in the domain name
+ *
+ * @param src               Input UChar array containing IDN in Unicode.
+ * @param srcLength         Number of UChars in src, or -1 if NUL-terminated.
+ * @param dest              Output UChar array with ASCII (ACE encoded) IDN.
+ * @param destCapacity      Size of dest.
+ * @param options           A bit set of options:
+ *
+ *  - UIDNA_DEFAULT             Use default options, i.e., do not process unassigned code points
+ *                              and do not use STD3 ASCII rules
+ *                              If unassigned code points are found the operation fails with 
+ *                              U_UNASSIGNED_CODE_POINT_FOUND error code.
+ *
+ *  - UIDNA_ALLOW_UNASSIGNED    Unassigned values can be converted to ASCII for query operations
+ *                              If this option is set, the unassigned code points are in the input 
+ *                              are treated as normal Unicode code points.
+ *
+ *  - UIDNA_USE_STD3_RULES      Use STD3 ASCII rules for host name syntax restrictions
+ *                              If this option is set and the input does not satisfy STD3 rules,  
+ *                              the operation will fail with U_IDNA_STD3_ASCII_RULES_ERROR
+ *
+ * @param parseError        Pointer to UParseError struct to receive information on position 
+ *                          of error if an error is encountered. Can be NULL.
+ * @param status            ICU in/out error code parameter.
+ *                          U_INVALID_CHAR_FOUND if src contains
+ *                          unmatched single surrogates.
+ *                          U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                          too many code points.
+ *                          U_BUFFER_OVERFLOW_ERROR if destCapacity is not enough
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+uidna_IDNToASCII(  const UChar* src, int32_t srcLength,
+                   UChar* dest, int32_t destCapacity,
+                   int32_t options,
+                   UParseError* parseError,
+                   UErrorCode* status);
+
+/**
+ * IDNA2003: Convenience function that implements the IDNToUnicode operation as defined in the IDNA RFC.
+ * This operation is done on complete domain names, e.g: "www.example.com". 
+ *
+ * <b>Note:</b> IDNA RFC specifies that a conformant application should divide a domain name
+ * into separate labels, decide whether to apply allowUnassigned and useSTD3ASCIIRules on each, 
+ * and then convert. This function does not offer that level of granularity. The options once  
+ * set will apply to all labels in the domain name
+ *
+ * @param src               Input UChar array containing IDN in ASCII (ACE encoded) form.
+ * @param srcLength         Number of UChars in src, or -1 if NUL-terminated.
+ * @param dest Output       UChar array containing Unicode equivalent of source IDN.
+ * @param destCapacity      Size of dest.
+ * @param options           A bit set of options:
+ *
+ *  - UIDNA_DEFAULT             Use default options, i.e., do not process unassigned code points
+ *                              and do not use STD3 ASCII rules
+ *                              If unassigned code points are found the operation fails with 
+ *                              U_UNASSIGNED_CODE_POINT_FOUND error code.
+ *
+ *  - UIDNA_ALLOW_UNASSIGNED    Unassigned values can be converted to ASCII for query operations
+ *                              If this option is set, the unassigned code points are in the input 
+ *                              are treated as normal Unicode code points.
+ *
+ *  - UIDNA_USE_STD3_RULES      Use STD3 ASCII rules for host name syntax restrictions
+ *                              If this option is set and the input does not satisfy STD3 rules,  
+ *                              the operation will fail with U_IDNA_STD3_ASCII_RULES_ERROR
+ *
+ * @param parseError        Pointer to UParseError struct to receive information on position 
+ *                          of error if an error is encountered. Can be NULL.
+ * @param status            ICU in/out error code parameter.
+ *                          U_INVALID_CHAR_FOUND if src contains
+ *                          unmatched single surrogates.
+ *                          U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                          too many code points.
+ *                          U_BUFFER_OVERFLOW_ERROR if destCapacity is not enough
+ * @return The length of the result string, if successful - or in case of a buffer overflow,
+ *         in which case it will be greater than destCapacity.
+ * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+uidna_IDNToUnicode(  const UChar* src, int32_t srcLength,
+                     UChar* dest, int32_t destCapacity,
+                     int32_t options,
+                     UParseError* parseError,
+                     UErrorCode* status);
+
+/**
+ * IDNA2003: Compare two IDN strings for equivalence.
+ * This function splits the domain names into labels and compares them.
+ * According to IDN RFC, whenever two labels are compared, they are 
+ * considered equal if and only if their ASCII forms (obtained by 
+ * applying toASCII) match using an case-insensitive ASCII comparison.
+ * Two domain names are considered a match if and only if all labels 
+ * match regardless of whether label separators match.
+ *
+ * @param s1                First source string.
+ * @param length1           Length of first source string, or -1 if NUL-terminated.
+ *
+ * @param s2                Second source string.
+ * @param length2           Length of second source string, or -1 if NUL-terminated.
+ * @param options           A bit set of options:
+ *
+ *  - UIDNA_DEFAULT             Use default options, i.e., do not process unassigned code points
+ *                              and do not use STD3 ASCII rules
+ *                              If unassigned code points are found the operation fails with 
+ *                              U_UNASSIGNED_CODE_POINT_FOUND error code.
+ *
+ *  - UIDNA_ALLOW_UNASSIGNED    Unassigned values can be converted to ASCII for query operations
+ *                              If this option is set, the unassigned code points are in the input 
+ *                              are treated as normal Unicode code points.
+ *
+ *  - UIDNA_USE_STD3_RULES      Use STD3 ASCII rules for host name syntax restrictions
+ *                              If this option is set and the input does not satisfy STD3 rules,  
+ *                              the operation will fail with U_IDNA_STD3_ASCII_RULES_ERROR
+ *
+ * @param status            ICU error code in/out parameter.
+ *                          Must fulfill U_SUCCESS before the function call.
+ * @return <0 or 0 or >0 as usual for string comparisons
+ * @deprecated ICU 55 Use UTS #46 instead via uidna_openUTS46() or class IDNA.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+uidna_compare(  const UChar *s1, int32_t length1,
+                const UChar *s2, int32_t length2,
+                int32_t options,
+                UErrorCode* status);
+
+#endif  /* U_HIDE_DEPRECATED_API */
+
+#endif /* #if !UCONFIG_NO_IDNA */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uiter.h b/thirdparty/icu4c/common/unicode/uiter.h
new file mode 100644
index 0000000000..be232c774d
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uiter.h
@@ -0,0 +1,709 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2011 International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uiter.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jan18
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UITER_H__
+#define __UITER_H__
+
+/**
+ * \file
+ * \brief C API: Unicode Character Iteration
+ *
+ * @see UCharIterator
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+    U_NAMESPACE_BEGIN
+
+    class CharacterIterator;
+    class Replaceable;
+
+    U_NAMESPACE_END
+#endif
+
+U_CDECL_BEGIN
+
+struct UCharIterator;
+typedef struct UCharIterator UCharIterator; /**< C typedef for struct UCharIterator. @stable ICU 2.1 */
+
+/**
+ * Origin constants for UCharIterator.getIndex() and UCharIterator.move().
+ * @see UCharIteratorMove
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef enum UCharIteratorOrigin {
+    UITER_START, UITER_CURRENT, UITER_LIMIT, UITER_ZERO, UITER_LENGTH
+} UCharIteratorOrigin;
+
+/** Constants for UCharIterator. @stable ICU 2.6 */
+enum {
+    /**
+     * Constant value that may be returned by UCharIteratorMove
+     * indicating that the final UTF-16 index is not known, but that the move succeeded.
+     * This can occur when moving relative to limit or length, or
+     * when moving relative to the current index after a setState()
+     * when the current UTF-16 index is not known.
+     *
+     * It would be very inefficient to have to count from the beginning of the text
+     * just to get the current/limit/length index after moving relative to it.
+     * The actual index can be determined with getIndex(UITER_CURRENT)
+     * which will count the UChars if necessary.
+     *
+     * @stable ICU 2.6
+     */
+    UITER_UNKNOWN_INDEX=-2
+};
+
+
+/**
+ * Constant for UCharIterator getState() indicating an error or
+ * an unknown state.
+ * Returned by uiter_getState()/UCharIteratorGetState
+ * when an error occurs.
+ * Also, some UCharIterator implementations may not be able to return
+ * a valid state for each position. This will be clearly documented
+ * for each such iterator (none of the public ones here).
+ *
+ * @stable ICU 2.6
+ */
+#define UITER_NO_STATE ((uint32_t)0xffffffff)
+
+/**
+ * Function type declaration for UCharIterator.getIndex().
+ *
+ * Gets the current position, or the start or limit of the
+ * iteration range.
+ *
+ * This function may perform slowly for UITER_CURRENT after setState() was called,
+ * or for UITER_LENGTH, because an iterator implementation may have to count
+ * UChars if the underlying storage is not UTF-16.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @param origin get the 0, start, limit, length, or current index
+ * @return the requested index, or U_SENTINEL in an error condition
+ *
+ * @see UCharIteratorOrigin
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef int32_t U_CALLCONV
+UCharIteratorGetIndex(UCharIterator *iter, UCharIteratorOrigin origin);
+
+/**
+ * Function type declaration for UCharIterator.move().
+ *
+ * Use iter->move(iter, index, UITER_ZERO) like CharacterIterator::setIndex(index).
+ *
+ * Moves the current position relative to the start or limit of the
+ * iteration range, or relative to the current position itself.
+ * The movement is expressed in numbers of code units forward
+ * or backward by specifying a positive or negative delta.
+ * Out of bounds movement will be pinned to the start or limit.
+ *
+ * This function may perform slowly for moving relative to UITER_LENGTH
+ * because an iterator implementation may have to count the rest of the
+ * UChars if the native storage is not UTF-16.
+ *
+ * When moving relative to the limit or length, or
+ * relative to the current position after setState() was called,
+ * move() may return UITER_UNKNOWN_INDEX (-2) to avoid an inefficient
+ * determination of the actual UTF-16 index.
+ * The actual index can be determined with getIndex(UITER_CURRENT)
+ * which will count the UChars if necessary.
+ * See UITER_UNKNOWN_INDEX for details.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @param delta can be positive, zero, or negative
+ * @param origin move relative to the 0, start, limit, length, or current index
+ * @return the new index, or U_SENTINEL on an error condition,
+ *         or UITER_UNKNOWN_INDEX when the index is not known.
+ *
+ * @see UCharIteratorOrigin
+ * @see UCharIterator
+ * @see UITER_UNKNOWN_INDEX
+ * @stable ICU 2.1
+ */
+typedef int32_t U_CALLCONV
+UCharIteratorMove(UCharIterator *iter, int32_t delta, UCharIteratorOrigin origin);
+
+/**
+ * Function type declaration for UCharIterator.hasNext().
+ *
+ * Check if current() and next() can still
+ * return another code unit.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return boolean value for whether current() and next() can still return another code unit
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef UBool U_CALLCONV
+UCharIteratorHasNext(UCharIterator *iter);
+
+/**
+ * Function type declaration for UCharIterator.hasPrevious().
+ *
+ * Check if previous() can still return another code unit.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return boolean value for whether previous() can still return another code unit
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef UBool U_CALLCONV
+UCharIteratorHasPrevious(UCharIterator *iter);
+ 
+/**
+ * Function type declaration for UCharIterator.current().
+ *
+ * Return the code unit at the current position,
+ * or U_SENTINEL if there is none (index is at the limit).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the current code unit
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef UChar32 U_CALLCONV
+UCharIteratorCurrent(UCharIterator *iter);
+
+/**
+ * Function type declaration for UCharIterator.next().
+ *
+ * Return the code unit at the current index and increment
+ * the index (post-increment, like s[i++]),
+ * or return U_SENTINEL if there is none (index is at the limit).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the current code unit (and post-increment the current index)
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef UChar32 U_CALLCONV
+UCharIteratorNext(UCharIterator *iter);
+
+/**
+ * Function type declaration for UCharIterator.previous().
+ *
+ * Decrement the index and return the code unit from there
+ * (pre-decrement, like s[--i]),
+ * or return U_SENTINEL if there is none (index is at the start).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the previous code unit (after pre-decrementing the current index)
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef UChar32 U_CALLCONV
+UCharIteratorPrevious(UCharIterator *iter);
+
+/**
+ * Function type declaration for UCharIterator.reservedFn().
+ * Reserved for future use.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @param something some integer argument
+ * @return some integer
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+typedef int32_t U_CALLCONV
+UCharIteratorReserved(UCharIterator *iter, int32_t something);
+
+/**
+ * Function type declaration for UCharIterator.getState().
+ *
+ * Get the "state" of the iterator in the form of a single 32-bit word.
+ * It is recommended that the state value be calculated to be as small as
+ * is feasible. For strings with limited lengths, fewer than 32 bits may
+ * be sufficient.
+ *
+ * This is used together with setState()/UCharIteratorSetState
+ * to save and restore the iterator position more efficiently than with
+ * getIndex()/move().
+ *
+ * The iterator state is defined as a uint32_t value because it is designed
+ * for use in ucol_nextSortKeyPart() which provides 32 bits to store the state
+ * of the character iterator.
+ *
+ * With some UCharIterator implementations (e.g., UTF-8),
+ * getting and setting the UTF-16 index with existing functions
+ * (getIndex(UITER_CURRENT) followed by move(pos, UITER_ZERO)) is possible but
+ * relatively slow because the iterator has to "walk" from a known index
+ * to the requested one.
+ * This takes more time the farther it needs to go.
+ *
+ * An opaque state value allows an iterator implementation to provide
+ * an internal index (UTF-8: the source byte array index) for
+ * fast, constant-time restoration.
+ *
+ * After calling setState(), a getIndex(UITER_CURRENT) may be slow because
+ * the UTF-16 index may not be restored as well, but the iterator can deliver
+ * the correct text contents and move relative to the current position
+ * without performance degradation.
+ *
+ * Some UCharIterator implementations may not be able to return
+ * a valid state for each position, in which case they return UITER_NO_STATE instead.
+ * This will be clearly documented for each such iterator (none of the public ones here).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the state word
+ *
+ * @see UCharIterator
+ * @see UCharIteratorSetState
+ * @see UITER_NO_STATE
+ * @stable ICU 2.6
+ */
+typedef uint32_t U_CALLCONV
+UCharIteratorGetState(const UCharIterator *iter);
+
+/**
+ * Function type declaration for UCharIterator.setState().
+ *
+ * Restore the "state" of the iterator using a state word from a getState() call.
+ * The iterator object need not be the same one as for which getState() was called,
+ * but it must be of the same type (set up using the same uiter_setXYZ function)
+ * and it must iterate over the same string
+ * (binary identical regardless of memory address).
+ * For more about the state word see UCharIteratorGetState.
+ *
+ * After calling setState(), a getIndex(UITER_CURRENT) may be slow because
+ * the UTF-16 index may not be restored as well, but the iterator can deliver
+ * the correct text contents and move relative to the current position
+ * without performance degradation.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @param state the state word from a getState() call
+ *              on a same-type, same-string iterator
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ *
+ * @see UCharIterator
+ * @see UCharIteratorGetState
+ * @stable ICU 2.6
+ */
+typedef void U_CALLCONV
+UCharIteratorSetState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode);
+
+
+/**
+ * C API for code unit iteration.
+ * This can be used as a C wrapper around
+ * CharacterIterator, Replaceable, or implemented using simple strings, etc.
+ *
+ * There are two roles for using UCharIterator:
+ *
+ * A "provider" sets the necessary function pointers and controls the "protected"
+ * fields of the UCharIterator structure. A "provider" passes a UCharIterator
+ * into C APIs that need a UCharIterator as an abstract, flexible string interface.
+ *
+ * Implementations of such C APIs are "callers" of UCharIterator functions;
+ * they only use the "public" function pointers and never access the "protected"
+ * fields directly.
+ *
+ * The current() and next() functions only check the current index against the
+ * limit, and previous() only checks the current index against the start,
+ * to see if the iterator already reached the end of the iteration range.
+ *
+ * The assumption - in all iterators - is that the index is moved via the API,
+ * which means it won't go out of bounds, or the index is modified by
+ * user code that knows enough about the iterator implementation to set valid
+ * index values.
+ *
+ * UCharIterator functions return code unit values 0..0xffff,
+ * or U_SENTINEL if the iteration bounds are reached.
+ *
+ * @stable ICU 2.1
+ */
+struct UCharIterator {
+    /**
+     * (protected) Pointer to string or wrapped object or similar.
+     * Not used by caller.
+     * @stable ICU 2.1
+     */
+    const void *context;
+
+    /**
+     * (protected) Length of string or similar.
+     * Not used by caller.
+     * @stable ICU 2.1
+     */
+    int32_t length;
+
+    /**
+     * (protected) Start index or similar.
+     * Not used by caller.
+     * @stable ICU 2.1
+     */
+    int32_t start;
+
+    /**
+     * (protected) Current index or similar.
+     * Not used by caller.
+     * @stable ICU 2.1
+     */
+    int32_t index;
+
+    /**
+     * (protected) Limit index or similar.
+     * Not used by caller.
+     * @stable ICU 2.1
+     */
+    int32_t limit;
+
+    /**
+     * (protected) Used by UTF-8 iterators and possibly others.
+     * @stable ICU 2.1
+     */
+    int32_t reservedField;
+
+    /**
+     * (public) Returns the current position or the
+     * start or limit index of the iteration range.
+     *
+     * @see UCharIteratorGetIndex
+     * @stable ICU 2.1
+     */
+    UCharIteratorGetIndex *getIndex;
+
+    /**
+     * (public) Moves the current position relative to the start or limit of the
+     * iteration range, or relative to the current position itself.
+     * The movement is expressed in numbers of code units forward
+     * or backward by specifying a positive or negative delta.
+     *
+     * @see UCharIteratorMove
+     * @stable ICU 2.1
+     */
+    UCharIteratorMove *move;
+
+    /**
+     * (public) Check if current() and next() can still
+     * return another code unit.
+     *
+     * @see UCharIteratorHasNext
+     * @stable ICU 2.1
+     */
+    UCharIteratorHasNext *hasNext;
+
+    /**
+     * (public) Check if previous() can still return another code unit.
+     *
+     * @see UCharIteratorHasPrevious
+     * @stable ICU 2.1
+     */
+    UCharIteratorHasPrevious *hasPrevious;
+
+    /**
+     * (public) Return the code unit at the current position,
+     * or U_SENTINEL if there is none (index is at the limit).
+     *
+     * @see UCharIteratorCurrent
+     * @stable ICU 2.1
+     */
+    UCharIteratorCurrent *current;
+
+    /**
+     * (public) Return the code unit at the current index and increment
+     * the index (post-increment, like s[i++]),
+     * or return U_SENTINEL if there is none (index is at the limit).
+     *
+     * @see UCharIteratorNext
+     * @stable ICU 2.1
+     */
+    UCharIteratorNext *next;
+
+    /**
+     * (public) Decrement the index and return the code unit from there
+     * (pre-decrement, like s[--i]),
+     * or return U_SENTINEL if there is none (index is at the start).
+     *
+     * @see UCharIteratorPrevious
+     * @stable ICU 2.1
+     */
+    UCharIteratorPrevious *previous;
+
+    /**
+     * (public) Reserved for future use. Currently NULL.
+     *
+     * @see UCharIteratorReserved
+     * @stable ICU 2.1
+     */
+    UCharIteratorReserved *reservedFn;
+
+    /**
+     * (public) Return the state of the iterator, to be restored later with setState().
+     * This function pointer is NULL if the iterator does not implement it.
+     *
+     * @see UCharIteratorGet
+     * @stable ICU 2.6
+     */
+    UCharIteratorGetState *getState;
+
+    /**
+     * (public) Restore the iterator state from the state word from a call
+     * to getState().
+     * This function pointer is NULL if the iterator does not implement it.
+     *
+     * @see UCharIteratorSet
+     * @stable ICU 2.6
+     */
+    UCharIteratorSetState *setState;
+};
+
+/**
+ * Helper function for UCharIterator to get the code point
+ * at the current index.
+ *
+ * Return the code point that includes the code unit at the current position,
+ * or U_SENTINEL if there is none (index is at the limit).
+ * If the current code unit is a lead or trail surrogate,
+ * then the following or preceding surrogate is used to form
+ * the code point value.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the current code point
+ *
+ * @see UCharIterator
+ * @see U16_GET
+ * @see UnicodeString::char32At()
+ * @stable ICU 2.1
+ */
+U_CAPI UChar32 U_EXPORT2
+uiter_current32(UCharIterator *iter);
+
+/**
+ * Helper function for UCharIterator to get the next code point.
+ *
+ * Return the code point at the current index and increment
+ * the index (post-increment, like s[i++]),
+ * or return U_SENTINEL if there is none (index is at the limit).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the current code point (and post-increment the current index)
+ *
+ * @see UCharIterator
+ * @see U16_NEXT
+ * @stable ICU 2.1
+ */
+U_CAPI UChar32 U_EXPORT2
+uiter_next32(UCharIterator *iter);
+
+/**
+ * Helper function for UCharIterator to get the previous code point.
+ *
+ * Decrement the index and return the code point from there
+ * (pre-decrement, like s[--i]),
+ * or return U_SENTINEL if there is none (index is at the start).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the previous code point (after pre-decrementing the current index)
+ *
+ * @see UCharIterator
+ * @see U16_PREV
+ * @stable ICU 2.1
+ */
+U_CAPI UChar32 U_EXPORT2
+uiter_previous32(UCharIterator *iter);
+
+/**
+ * Get the "state" of the iterator in the form of a single 32-bit word.
+ * This is a convenience function that calls iter->getState(iter)
+ * if iter->getState is not NULL;
+ * if it is NULL or any other error occurs, then UITER_NO_STATE is returned.
+ *
+ * Some UCharIterator implementations may not be able to return
+ * a valid state for each position, in which case they return UITER_NO_STATE instead.
+ * This will be clearly documented for each such iterator (none of the public ones here).
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @return the state word
+ *
+ * @see UCharIterator
+ * @see UCharIteratorGetState
+ * @see UITER_NO_STATE
+ * @stable ICU 2.6
+ */
+U_CAPI uint32_t U_EXPORT2
+uiter_getState(const UCharIterator *iter);
+
+/**
+ * Restore the "state" of the iterator using a state word from a getState() call.
+ * This is a convenience function that calls iter->setState(iter, state, pErrorCode)
+ * if iter->setState is not NULL; if it is NULL, then U_UNSUPPORTED_ERROR is set.
+ *
+ * @param iter the UCharIterator structure ("this pointer")
+ * @param state the state word from a getState() call
+ *              on a same-type, same-string iterator
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                   which must not indicate a failure before the function call.
+ *
+ * @see UCharIterator
+ * @see UCharIteratorSetState
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+uiter_setState(UCharIterator *iter, uint32_t state, UErrorCode *pErrorCode);
+
+/**
+ * Set up a UCharIterator to iterate over a string.
+ *
+ * Sets the UCharIterator function pointers for iteration over the string s
+ * with iteration boundaries start=index=0 and length=limit=string length.
+ * The "provider" may set the start, index, and limit values at any time
+ * within the range 0..length.
+ * The length field will be ignored.
+ *
+ * The string pointer s is set into UCharIterator.context without copying
+ * or reallocating the string contents.
+ *
+ * getState() simply returns the current index.
+ * move() will always return the final index.
+ *
+ * @param iter UCharIterator structure to be set for iteration
+ * @param s String to iterate over
+ * @param length Length of s, or -1 if NUL-terminated
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+U_CAPI void U_EXPORT2
+uiter_setString(UCharIterator *iter, const UChar *s, int32_t length);
+
+/**
+ * Set up a UCharIterator to iterate over a UTF-16BE string
+ * (byte vector with a big-endian pair of bytes per UChar).
+ *
+ * Everything works just like with a normal UChar iterator (uiter_setString),
+ * except that UChars are assembled from byte pairs,
+ * and that the length argument here indicates an even number of bytes.
+ *
+ * getState() simply returns the current index.
+ * move() will always return the final index.
+ *
+ * @param iter UCharIterator structure to be set for iteration
+ * @param s UTF-16BE string to iterate over
+ * @param length Length of s as an even number of bytes, or -1 if NUL-terminated
+ *               (NUL means pair of 0 bytes at even index from s)
+ *
+ * @see UCharIterator
+ * @see uiter_setString
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+uiter_setUTF16BE(UCharIterator *iter, const char *s, int32_t length);
+
+/**
+ * Set up a UCharIterator to iterate over a UTF-8 string.
+ *
+ * Sets the UCharIterator function pointers for iteration over the UTF-8 string s
+ * with UTF-8 iteration boundaries 0 and length.
+ * The implementation counts the UTF-16 index on the fly and
+ * lazily evaluates the UTF-16 length of the text.
+ *
+ * The start field is used as the UTF-8 offset, the limit field as the UTF-8 length.
+ * When the reservedField is not 0, then it contains a supplementary code point
+ * and the UTF-16 index is between the two corresponding surrogates.
+ * At that point, the UTF-8 index is behind that code point.
+ *
+ * The UTF-8 string pointer s is set into UCharIterator.context without copying
+ * or reallocating the string contents.
+ *
+ * getState() returns a state value consisting of
+ * - the current UTF-8 source byte index (bits 31..1)
+ * - a flag (bit 0) that indicates whether the UChar position is in the middle
+ *   of a surrogate pair
+ *   (from a 4-byte UTF-8 sequence for the corresponding supplementary code point)
+ *
+ * getState() cannot also encode the UTF-16 index in the state value.
+ * move(relative to limit or length), or
+ * move(relative to current) after setState(), may return UITER_UNKNOWN_INDEX.
+ *
+ * @param iter UCharIterator structure to be set for iteration
+ * @param s UTF-8 string to iterate over
+ * @param length Length of s in bytes, or -1 if NUL-terminated
+ *
+ * @see UCharIterator
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+uiter_setUTF8(UCharIterator *iter, const char *s, int32_t length);
+
+#if U_SHOW_CPLUSPLUS_API
+
+/**
+ * Set up a UCharIterator to wrap around a C++ CharacterIterator.
+ *
+ * Sets the UCharIterator function pointers for iteration using the
+ * CharacterIterator charIter.
+ *
+ * The CharacterIterator pointer charIter is set into UCharIterator.context
+ * without copying or cloning the CharacterIterator object.
+ * The other "protected" UCharIterator fields are set to 0 and will be ignored.
+ * The iteration index and boundaries are controlled by the CharacterIterator.
+ *
+ * getState() simply returns the current index.
+ * move() will always return the final index.
+ *
+ * @param iter UCharIterator structure to be set for iteration
+ * @param charIter CharacterIterator to wrap
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+U_CAPI void U_EXPORT2
+uiter_setCharacterIterator(UCharIterator *iter, icu::CharacterIterator *charIter);
+
+/**
+ * Set up a UCharIterator to iterate over a C++ Replaceable.
+ *
+ * Sets the UCharIterator function pointers for iteration over the
+ * Replaceable rep with iteration boundaries start=index=0 and
+ * length=limit=rep->length().
+ * The "provider" may set the start, index, and limit values at any time
+ * within the range 0..length=rep->length().
+ * The length field will be ignored.
+ *
+ * The Replaceable pointer rep is set into UCharIterator.context without copying
+ * or cloning/reallocating the Replaceable object.
+ *
+ * getState() simply returns the current index.
+ * move() will always return the final index.
+ *
+ * @param iter UCharIterator structure to be set for iteration
+ * @param rep Replaceable to iterate over
+ *
+ * @see UCharIterator
+ * @stable ICU 2.1
+ */
+U_CAPI void U_EXPORT2
+uiter_setReplaceable(UCharIterator *iter, const icu::Replaceable *rep);
+
+#endif
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uldnames.h b/thirdparty/icu4c/common/unicode/uldnames.h
new file mode 100644
index 0000000000..47b047ece9
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uldnames.h
@@ -0,0 +1,307 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2016, International Business Machines Corporation and
+*   others.  All Rights Reserved.
+*******************************************************************************
+*/
+
+#ifndef __ULDNAMES_H__
+#define __ULDNAMES_H__
+
+/**
+ * \file
+ * \brief C API: Provides display names of Locale ids and their components.
+ */
+
+#include "unicode/utypes.h"
+#include "unicode/uscript.h"
+#include "unicode/udisplaycontext.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * Enum used in LocaleDisplayNames::createInstance.
+ * @stable ICU 4.4
+ */
+typedef enum {
+    /**
+     * Use standard names when generating a locale name,
+     * e.g. en_GB displays as 'English (United Kingdom)'.
+     * @stable ICU 4.4
+     */
+    ULDN_STANDARD_NAMES = 0,
+    /**
+     * Use dialect names, when generating a locale name,
+     * e.g. en_GB displays as 'British English'.
+     * @stable ICU 4.4
+     */
+    ULDN_DIALECT_NAMES
+} UDialectHandling;
+
+/**
+ * Opaque C service object type for the locale display names API
+ * @stable ICU 4.4
+ */
+struct ULocaleDisplayNames;
+
+/** 
+ * C typedef for struct ULocaleDisplayNames. 
+ * @stable ICU 4.4 
+ */
+typedef struct ULocaleDisplayNames ULocaleDisplayNames;  
+
+#if !UCONFIG_NO_FORMATTING
+
+/**
+ * Returns an instance of LocaleDisplayNames that returns names
+ * formatted for the provided locale, using the provided
+ * dialectHandling.  The usual value for dialectHandling is
+ * ULOC_STANDARD_NAMES.
+ *
+ * @param locale the display locale 
+ * @param dialectHandling how to select names for locales 
+ * @return a ULocaleDisplayNames instance 
+ * @param pErrorCode the status code
+ * @stable ICU 4.4
+ */
+U_CAPI ULocaleDisplayNames * U_EXPORT2
+uldn_open(const char * locale,
+          UDialectHandling dialectHandling,
+          UErrorCode *pErrorCode);
+
+/**
+ * Closes a ULocaleDisplayNames instance obtained from uldn_open().
+ * @param ldn the ULocaleDisplayNames instance to be closed
+ * @stable ICU 4.4
+ */
+U_CAPI void U_EXPORT2
+uldn_close(ULocaleDisplayNames *ldn);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalULocaleDisplayNamesPointer
+ * "Smart pointer" class, closes a ULocaleDisplayNames via uldn_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalULocaleDisplayNamesPointer, ULocaleDisplayNames, uldn_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/* getters for state */
+
+/**
+ * Returns the locale used to determine the display names. This is
+ * not necessarily the same locale passed to {@link #uldn_open}.
+ * @param ldn the LocaleDisplayNames instance
+ * @return the display locale 
+ * @stable ICU 4.4
+ */
+U_CAPI const char * U_EXPORT2
+uldn_getLocale(const ULocaleDisplayNames *ldn);
+
+/**
+ * Returns the dialect handling used in the display names.
+ * @param ldn the LocaleDisplayNames instance
+ * @return the dialect handling enum
+ * @stable ICU 4.4
+ */
+U_CAPI UDialectHandling U_EXPORT2
+uldn_getDialectHandling(const ULocaleDisplayNames *ldn);
+
+/* names for entire locales */
+
+/**
+ * Returns the display name of the provided locale.
+ * @param ldn the LocaleDisplayNames instance
+ * @param locale the locale whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_localeDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *locale,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode);
+
+/* names for components of a locale */
+
+/**
+ * Returns the display name of the provided language code.
+ * @param ldn the LocaleDisplayNames instance
+ * @param lang the language code whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_languageDisplayName(const ULocaleDisplayNames *ldn,
+                         const char *lang,
+                         UChar *result,
+                         int32_t maxResultSize,
+                         UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided script.
+ * @param ldn the LocaleDisplayNames instance
+ * @param script the script whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_scriptDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *script,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided script code.
+ * @param ldn the LocaleDisplayNames instance
+ * @param scriptCode the script code whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_scriptCodeDisplayName(const ULocaleDisplayNames *ldn,
+                           UScriptCode scriptCode,
+                           UChar *result,
+                           int32_t maxResultSize,
+                           UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided region code.
+ * @param ldn the LocaleDisplayNames instance
+ * @param region the region code whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_regionDisplayName(const ULocaleDisplayNames *ldn,
+                       const char *region,
+                       UChar *result,
+                       int32_t maxResultSize,
+                       UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided variant
+ * @param ldn the LocaleDisplayNames instance
+ * @param variant the variant whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_variantDisplayName(const ULocaleDisplayNames *ldn,
+                        const char *variant,
+                        UChar *result,
+                        int32_t maxResultSize,
+                        UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided locale key
+ * @param ldn the LocaleDisplayNames instance
+ * @param key the locale key whose display name to return
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_keyDisplayName(const ULocaleDisplayNames *ldn,
+                    const char *key,
+                    UChar *result,
+                    int32_t maxResultSize,
+                    UErrorCode *pErrorCode);
+
+/**
+ * Returns the display name of the provided value (used with the provided key).
+ * @param ldn the LocaleDisplayNames instance
+ * @param key the locale key
+ * @param value the locale key's value
+ * @param result receives the display name
+ * @param maxResultSize the size of the result buffer
+ * @param pErrorCode the status code
+ * @return the actual buffer size needed for the display name.  If it's
+ * greater than maxResultSize, the returned name will be truncated.
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+uldn_keyValueDisplayName(const ULocaleDisplayNames *ldn,
+                         const char *key,
+                         const char *value,
+                         UChar *result,
+                         int32_t maxResultSize,
+                         UErrorCode *pErrorCode);
+
+/**
+* Returns an instance of LocaleDisplayNames that returns names formatted
+* for the provided locale, using the provided UDisplayContext settings.
+*
+* @param locale The display locale 
+* @param contexts List of one or more context settings (e.g. for dialect
+*               handling, capitalization, etc.
+* @param length Number of items in the contexts list
+* @param pErrorCode Pointer to UErrorCode input/output status. If at entry this indicates
+*               a failure status, the function will do nothing; otherwise this will be
+*               updated with any new status from the function. 
+* @return a ULocaleDisplayNames instance 
+* @stable ICU 51
+*/
+U_CAPI ULocaleDisplayNames * U_EXPORT2
+uldn_openForContext(const char * locale, UDisplayContext *contexts,
+                    int32_t length, UErrorCode *pErrorCode);
+
+/**
+* Returns the UDisplayContext value for the specified UDisplayContextType.
+* @param ldn the ULocaleDisplayNames instance
+* @param type the UDisplayContextType whose value to return
+* @param pErrorCode Pointer to UErrorCode input/output status. If at entry this indicates
+*               a failure status, the function will do nothing; otherwise this will be
+*               updated with any new status from the function. 
+* @return the UDisplayContextValue for the specified type.
+* @stable ICU 51
+*/
+U_CAPI UDisplayContext U_EXPORT2
+uldn_getContext(const ULocaleDisplayNames *ldn, UDisplayContextType type,
+                UErrorCode *pErrorCode);
+
+#endif  /* !UCONFIG_NO_FORMATTING */
+#endif  /* __ULDNAMES_H__ */
diff --git a/thirdparty/icu4c/common/unicode/uloc.h b/thirdparty/icu4c/common/unicode/uloc.h
new file mode 100644
index 0000000000..3addb847e7
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uloc.h
@@ -0,0 +1,1393 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File ULOC.H
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/01/97    aliu        Creation.
+*   08/22/98    stephen     JDK 1.2 sync.
+*   12/08/98    rtg         New C API for Locale
+*   03/30/99    damiba      overhaul
+*   03/31/99    helena      Javadoc for uloc functions.
+*   04/15/99    Madhu       Updated Javadoc
+********************************************************************************
+*/
+
+#ifndef ULOC_H
+#define ULOC_H
+
+#include "unicode/utypes.h"
+#include "unicode/uenum.h"
+
+/**    
+ * \file
+ * \brief  C API: Locale 
+ *
+ * <h2> ULoc C API for Locale </h2>
+ * A <code>Locale</code> represents a specific geographical, political,
+ * or cultural region. An operation that requires a <code>Locale</code> to perform
+ * its task is called <em>locale-sensitive</em> and uses the <code>Locale</code>
+ * to tailor information for the user. For example, displaying a number
+ * is a locale-sensitive operation--the number should be formatted
+ * according to the customs/conventions of the user's native country,
+ * region, or culture.  In the C APIs, a locales is simply a const char string.
+ *
+ * <P>
+ * You create a <code>Locale</code> with one of the three options listed below.
+ * Each of the component is separated by '_' in the locale string.
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ * \code
+ *       newLanguage
+ * 
+ *       newLanguage + newCountry
+ * 
+ *       newLanguage + newCountry + newVariant
+ * \endcode
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * The first option is a valid <STRONG>ISO
+ * Language Code.</STRONG> These codes are the lower-case two-letter
+ * codes as defined by ISO-639.
+ * You can find a full list of these codes at a number of sites, such as:
+ * <BR><a href ="http://www.ics.uci.edu/pub/ietf/http/related/iso639.txt">
+ * http://www.ics.uci.edu/pub/ietf/http/related/iso639.txt</a>
+ *
+ * <P>
+ * The second option includes an additional <STRONG>ISO Country
+ * Code.</STRONG> These codes are the upper-case two-letter codes
+ * as defined by ISO-3166.
+ * You can find a full list of these codes at a number of sites, such as:
+ * <BR><a href="http://www.chemie.fu-berlin.de/diverse/doc/ISO_3166.html">
+ * http://www.chemie.fu-berlin.de/diverse/doc/ISO_3166.html</a>
+ *
+ * <P>
+ * The third option requires another additional information--the 
+ * <STRONG>Variant.</STRONG>
+ * The Variant codes are vendor and browser-specific.
+ * For example, use WIN for Windows, MAC for Macintosh, and POSIX for POSIX.
+ * Where there are two variants, separate them with an underscore, and
+ * put the most important one first. For
+ * example, a Traditional Spanish collation might be referenced, with
+ * "ES", "ES", "Traditional_WIN".
+ *
+ * <P>
+ * Because a <code>Locale</code> is just an identifier for a region,
+ * no validity check is performed when you specify a <code>Locale</code>.
+ * If you want to see whether particular resources are available for the
+ * <code>Locale</code> you asked for, you must query those resources. For
+ * example, ask the <code>UNumberFormat</code> for the locales it supports
+ * using its <code>getAvailable</code> method.
+ * <BR><STRONG>Note:</STRONG> When you ask for a resource for a particular
+ * locale, you get back the best available match, not necessarily
+ * precisely what you asked for. For more information, look at
+ * <code>UResourceBundle</code>.
+ *
+ * <P>
+ * The <code>Locale</code> provides a number of convenient constants
+ * that you can use to specify the commonly used
+ * locales. For example, the following refers to a locale
+ * for the United States:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ * \code
+ *       ULOC_US
+ * \endcode
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ *
+ * <P>
+ * Once you've specified a locale you can query it for information about
+ * itself. Use <code>uloc_getCountry</code> to get the ISO Country Code and
+ * <code>uloc_getLanguage</code> to get the ISO Language Code. You can
+ * use <code>uloc_getDisplayCountry</code> to get the
+ * name of the country suitable for displaying to the user. Similarly,
+ * you can use <code>uloc_getDisplayLanguage</code> to get the name of
+ * the language suitable for displaying to the user. Interestingly,
+ * the <code>uloc_getDisplayXXX</code> methods are themselves locale-sensitive
+ * and have two versions: one that uses the default locale and one
+ * that takes a locale as an argument and displays the name or country in
+ * a language appropriate to that locale.
+ *
+ * <P>
+ * The ICU provides a number of services that perform locale-sensitive
+ * operations. For example, the <code>unum_xxx</code> functions format
+ * numbers, currency, or percentages in a locale-sensitive manner. 
+ * </P>
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ * \code
+ *     UErrorCode success = U_ZERO_ERROR;
+ *     UNumberFormat *nf;
+ *     const char* myLocale = "fr_FR";
+ * 
+ *     nf = unum_open( UNUM_DEFAULT, NULL, success );          
+ *     unum_close(nf);
+ *     nf = unum_open( UNUM_CURRENCY, NULL, success );
+ *     unum_close(nf);
+ *     nf = unum_open( UNUM_PERCENT, NULL, success );   
+ *     unum_close(nf);
+ * \endcode
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * Each of these methods has two variants; one with an explicit locale
+ * and one without; the latter using the default locale.
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ * \code 
+ * 
+ *     nf = unum_open( UNUM_DEFAULT, myLocale, success );          
+ *     unum_close(nf);
+ *     nf = unum_open( UNUM_CURRENCY, myLocale, success );
+ *     unum_close(nf);
+ *     nf = unum_open( UNUM_PERCENT, myLocale, success );   
+ *     unum_close(nf);
+ * \endcode
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * A <code>Locale</code> is the mechanism for identifying the kind of services
+ * (<code>UNumberFormat</code>) that you would like to get. The locale is
+ * <STRONG>just</STRONG> a mechanism for identifying these services.
+ *
+ * <P>
+ * Each international service that performs locale-sensitive operations 
+ * allows you
+ * to get all the available objects of that type. You can sift
+ * through these objects by language, country, or variant,
+ * and use the display names to present a menu to the user.
+ * For example, you can create a menu of all the collation objects
+ * suitable for a given language. Such classes implement these
+ * three class methods:
+ * \htmlonly<blockquote>\endhtmlonly
+ * <pre>
+ * \code
+ *       const char* uloc_getAvailable(int32_t index);
+ *       int32_t uloc_countAvailable();
+ *       int32_t
+ *       uloc_getDisplayName(const char* localeID,
+ *                 const char* inLocaleID, 
+ *                 UChar* result,
+ *                 int32_t maxResultSize,
+ *                  UErrorCode* err);
+ * 
+ * \endcode
+ * </pre>
+ * \htmlonly</blockquote>\endhtmlonly
+ * <P>
+ * Concerning POSIX/RFC1766 Locale IDs, 
+ *  the getLanguage/getCountry/getVariant/getName functions do understand
+ * the POSIX type form of  language_COUNTRY.ENCODING\@VARIANT
+ * and if there is not an ICU-stype variant, uloc_getVariant() for example
+ * will return the one listed after the \@at sign. As well, the hyphen
+ * "-" is recognized as a country/variant separator similarly to RFC1766.
+ * So for example, "en-us" will be interpreted as en_US.  
+ * As a result, uloc_getName() is far from a no-op, and will have the
+ * effect of converting POSIX/RFC1766 IDs into ICU form, although it does
+ * NOT map any of the actual codes (i.e. russian->ru) in any way.
+ * Applications should call uloc_getName() at the point where a locale ID
+ * is coming from an external source (user entry, OS, web browser)
+ * and pass the resulting string to other ICU functions.  For example,
+ * don't use de-de\@EURO as an argument to resourcebundle.
+ *
+ * @see UResourceBundle
+ */
+
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_CHINESE            "zh"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_ENGLISH            "en"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_FRENCH             "fr"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_GERMAN             "de"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_ITALIAN            "it"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_JAPANESE           "ja"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_KOREAN             "ko"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_SIMPLIFIED_CHINESE "zh_CN"
+/** Useful constant for this language. @stable ICU 2.0 */
+#define ULOC_TRADITIONAL_CHINESE "zh_TW"
+
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_CANADA         "en_CA"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_CANADA_FRENCH  "fr_CA"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_CHINA          "zh_CN"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_PRC            "zh_CN"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_FRANCE         "fr_FR"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_GERMANY        "de_DE"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_ITALY          "it_IT"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_JAPAN          "ja_JP"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_KOREA          "ko_KR"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_TAIWAN         "zh_TW"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_UK             "en_GB"
+/** Useful constant for this country/region. @stable ICU 2.0 */
+#define ULOC_US             "en_US"
+
+/**
+ * Useful constant for the maximum size of the language part of a locale ID.
+ * (including the terminating NULL).
+ * @stable ICU 2.0
+ */
+#define ULOC_LANG_CAPACITY 12
+
+/**
+ * Useful constant for the maximum size of the country part of a locale ID
+ * (including the terminating NULL).
+ * @stable ICU 2.0
+ */
+#define ULOC_COUNTRY_CAPACITY 4
+/**
+ * Useful constant for the maximum size of the whole locale ID
+ * (including the terminating NULL and all keywords).
+ * @stable ICU 2.0
+ */
+#define ULOC_FULLNAME_CAPACITY 157
+
+/**
+ * Useful constant for the maximum size of the script part of a locale ID
+ * (including the terminating NULL).
+ * @stable ICU 2.8
+ */
+#define ULOC_SCRIPT_CAPACITY 6
+
+/**
+ * Useful constant for the maximum size of keywords in a locale
+ * @stable ICU 2.8
+ */
+#define ULOC_KEYWORDS_CAPACITY 96
+
+/**
+ * Useful constant for the maximum total size of keywords and their values in a locale
+ * @stable ICU 2.8
+ */
+#define ULOC_KEYWORD_AND_VALUES_CAPACITY 100
+
+/**
+ * Invariant character separating keywords from the locale string
+ * @stable ICU 2.8
+ */
+#define ULOC_KEYWORD_SEPARATOR '@'
+
+/**
+  * Unicode code point for '@' separating keywords from the locale string.
+  * @see ULOC_KEYWORD_SEPARATOR
+  * @stable ICU 4.6
+  */
+#define ULOC_KEYWORD_SEPARATOR_UNICODE 0x40
+
+/**
+ * Invariant character for assigning value to a keyword
+ * @stable ICU 2.8
+ */
+#define ULOC_KEYWORD_ASSIGN '='
+
+/**
+  * Unicode code point for '=' for assigning value to a keyword.
+  * @see ULOC_KEYWORD_ASSIGN
+  * @stable ICU 4.6 
+  */
+#define ULOC_KEYWORD_ASSIGN_UNICODE 0x3D
+
+/**
+ * Invariant character separating keywords
+ * @stable ICU 2.8
+ */
+#define ULOC_KEYWORD_ITEM_SEPARATOR ';'
+
+/**
+  * Unicode code point for ';' separating keywords
+  * @see ULOC_KEYWORD_ITEM_SEPARATOR
+  * @stable ICU 4.6
+  */
+#define ULOC_KEYWORD_ITEM_SEPARATOR_UNICODE 0x3B
+
+/**
+ * Constants for *_getLocale()
+ * Allow user to select whether she wants information on 
+ * requested, valid or actual locale.
+ * For example, a collator for "en_US_CALIFORNIA" was
+ * requested. In the current state of ICU (2.0), 
+ * the requested locale is "en_US_CALIFORNIA",
+ * the valid locale is "en_US" (most specific locale supported by ICU)
+ * and the actual locale is "root" (the collation data comes unmodified 
+ * from the UCA)
+ * The locale is considered supported by ICU if there is a core ICU bundle 
+ * for that locale (although it may be empty).
+ * @stable ICU 2.1
+ */
+typedef enum {
+  /** This is locale the data actually comes from 
+   * @stable ICU 2.1
+   */
+  ULOC_ACTUAL_LOCALE    = 0,
+  /** This is the most specific locale supported by ICU 
+   * @stable ICU 2.1
+   */
+  ULOC_VALID_LOCALE    = 1,
+
+#ifndef U_HIDE_DEPRECATED_API
+  /** This is the requested locale
+   *  @deprecated ICU 2.8 
+   */
+  ULOC_REQUESTED_LOCALE = 2,
+
+    /**
+     * One more than the highest normal ULocDataLocaleType value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    ULOC_DATA_LOCALE_TYPE_LIMIT = 3
+#endif  // U_HIDE_DEPRECATED_API
+} ULocDataLocaleType;
+
+#ifndef U_HIDE_SYSTEM_API
+/**
+ * Gets ICU's default locale.  
+ * The returned string is a snapshot in time, and will remain valid
+ *   and unchanged even when uloc_setDefault() is called.
+ *   The returned storage is owned by ICU, and must not be altered or deleted
+ *   by the caller.
+ *  
+ * @return the ICU default locale
+ * @system
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2
+uloc_getDefault(void);
+
+/**
+ * Sets ICU's default locale.  
+ *    By default (without calling this function), ICU's default locale will be based
+ *    on information obtained from the underlying system environment.
+ *    <p>
+ *    Changes to ICU's default locale do not propagate back to the
+ *    system environment.
+ *    <p>
+ *    Changes to ICU's default locale to not affect any ICU services that
+ *    may already be open based on the previous default locale value.
+ *
+ * @param localeID the new ICU default locale. A value of NULL will try to get
+ *                 the system's default locale.
+ * @param status the error information if the setting of default locale fails
+ * @system
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+uloc_setDefault(const char* localeID,
+        UErrorCode*       status);
+#endif  /* U_HIDE_SYSTEM_API */
+
+/**
+ * Gets the language code for the specified locale.
+ *
+ * @param localeID the locale to get the ISO language code with
+ * @param language the language code for localeID
+ * @param languageCapacity the size of the language buffer to store the  
+ * language code with
+ * @param err error information if retrieving the language code failed
+ * @return the actual buffer size needed for the language code.  If it's greater 
+ * than languageCapacity, the returned language code will be truncated.  
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getLanguage(const char*    localeID,
+         char* language,
+         int32_t languageCapacity,
+         UErrorCode* err);
+
+/**
+ * Gets the script code for the specified locale.
+ *
+ * @param localeID the locale to get the ISO language code with
+ * @param script the language code for localeID
+ * @param scriptCapacity the size of the language buffer to store the  
+ * language code with
+ * @param err error information if retrieving the language code failed
+ * @return the actual buffer size needed for the language code.  If it's greater 
+ * than scriptCapacity, the returned language code will be truncated.  
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getScript(const char*    localeID,
+         char* script,
+         int32_t scriptCapacity,
+         UErrorCode* err);
+
+/**
+ * Gets the  country code for the specified locale.
+ *
+ * @param localeID the locale to get the country code with
+ * @param country the country code for localeID
+ * @param countryCapacity the size of the country buffer to store the  
+ * country code with
+ * @param err error information if retrieving the country code failed
+ * @return the actual buffer size needed for the country code.  If it's greater 
+ * than countryCapacity, the returned country code will be truncated.  
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getCountry(const char*    localeID,
+        char* country,
+        int32_t countryCapacity,
+        UErrorCode* err);
+
+/**
+ * Gets the variant code for the specified locale.
+ *
+ * @param localeID the locale to get the variant code with
+ * @param variant the variant code for localeID
+ * @param variantCapacity the size of the variant buffer to store the 
+ * variant code with
+ * @param err error information if retrieving the variant code failed
+ * @return the actual buffer size needed for the variant code.  If it's greater 
+ * than variantCapacity, the returned variant code will be truncated.  
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getVariant(const char*    localeID,
+        char* variant,
+        int32_t variantCapacity,
+        UErrorCode* err);
+
+
+/**
+ * Gets the full name for the specified locale.
+ * Note: This has the effect of 'canonicalizing' the ICU locale ID to
+ * a certain extent. Upper and lower case are set as needed.
+ * It does NOT map aliased names in any way.
+ * See the top of this header file.
+ * This API supports preflighting.
+ *
+ * @param localeID the locale to get the full name with
+ * @param name fill in buffer for the name without keywords.
+ * @param nameCapacity capacity of the fill in buffer.
+ * @param err error information if retrieving the full name failed
+ * @return the actual buffer size needed for the full name.  If it's greater 
+ * than nameCapacity, the returned full name will be truncated.  
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getName(const char*    localeID,
+         char* name,
+         int32_t nameCapacity,
+         UErrorCode* err);
+
+/**
+ * Gets the full name for the specified locale.
+ * Note: This has the effect of 'canonicalizing' the string to
+ * a certain extent. Upper and lower case are set as needed,
+ * and if the components were in 'POSIX' format they are changed to
+ * ICU format.  It does NOT map aliased names in any way.
+ * See the top of this header file.
+ *
+ * @param localeID the locale to get the full name with
+ * @param name the full name for localeID
+ * @param nameCapacity the size of the name buffer to store the 
+ * full name with
+ * @param err error information if retrieving the full name failed
+ * @return the actual buffer size needed for the full name.  If it's greater 
+ * than nameCapacity, the returned full name will be truncated.  
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_canonicalize(const char*    localeID,
+         char* name,
+         int32_t nameCapacity,
+         UErrorCode* err);
+
+/**
+ * Gets the ISO language code for the specified locale.
+ *
+ * @param localeID the locale to get the ISO language code with
+ * @return language the ISO language code for localeID
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2
+uloc_getISO3Language(const char* localeID);
+
+
+/**
+ * Gets the ISO country code for the specified locale.
+ *
+ * @param localeID the locale to get the ISO country code with
+ * @return country the ISO country code for localeID
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2
+uloc_getISO3Country(const char* localeID);
+
+/**
+ * Gets the Win32 LCID value for the specified locale.
+ * If the ICU locale is not recognized by Windows, 0 will be returned.
+ *
+ * LCIDs were deprecated with Windows Vista and Microsoft recommends
+ * that developers use BCP47 style tags instead (uloc_toLanguageTag).
+ *
+ * @param localeID the locale to get the Win32 LCID value with
+ * @return country the Win32 LCID for localeID
+ * @stable ICU 2.0
+ */
+U_CAPI uint32_t U_EXPORT2
+uloc_getLCID(const char* localeID);
+
+/**
+ * Gets the language name suitable for display for the specified locale.
+ *
+ * @param locale the locale to get the ISO language code with
+ * @param displayLocale Specifies the locale to be used to display the name. In
+ *                 other words, if the locale's language code is "en", passing
+ *                 Locale::getFrench() for inLocale would result in "Anglais",
+ *                 while passing Locale::getGerman() for inLocale would result
+ *                 in "Englisch".
+ * @param language the displayable language code for localeID
+ * @param languageCapacity the size of the language buffer to store the
+ *                 displayable language code with.
+ * @param status error information if retrieving the displayable language code
+ *                 failed. U_USING_DEFAULT_WARNING indicates that no data was
+ *                 found from the locale resources and a case canonicalized
+ *                 language code is placed into language as fallback.
+ * @return the actual buffer size needed for the displayable language code. If
+ *                 it's greater than languageCapacity, the returned language
+ *                 code will be truncated.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayLanguage(const char* locale,
+            const char* displayLocale,
+            UChar* language,
+            int32_t languageCapacity,
+            UErrorCode* status);
+
+/**
+ * Gets the script name suitable for display for the specified locale.
+ *
+ * @param locale the locale to get the displayable script code with. NULL may be
+ *                 used to specify the default.
+ * @param displayLocale Specifies the locale to be used to display the name. In
+ *                 other words, if the locale's language code is "en", passing
+ *                 Locale::getFrench() for inLocale would result in "", while
+ *                 passing Locale::getGerman() for inLocale would result in "".
+ *                 NULL may be used to specify the default.
+ * @param script the displayable script for the localeID.
+ * @param scriptCapacity the size of the script buffer to store the displayable
+ *                 script code with.
+ * @param status error information if retrieving the displayable script code
+ *                 failed. U_USING_DEFAULT_WARNING indicates that no data was
+ *                 found from the locale resources and a case canonicalized
+ *                 script code is placed into script as fallback.
+ * @return the actual buffer size needed for the displayable script code. If
+ *                 it's greater than scriptCapacity, the returned displayable
+ *                 script code will be truncated.
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayScript(const char* locale,
+            const char* displayLocale,
+            UChar* script,
+            int32_t scriptCapacity,
+            UErrorCode* status);
+
+/**
+ * Gets the country name suitable for display for the specified locale.
+ * Warning: this is for the region part of a valid locale ID; it cannot just be
+ * the region code (like "FR"). To get the display name for a region alone, or
+ * for other options, use ULocaleDisplayNames instead.
+ *
+ * @param locale the locale to get the displayable country code with. NULL may
+ *                 be used to specify the default.
+ * @param displayLocale Specifies the locale to be used to display the name. In
+ *                 other words, if the locale's language code is "en", passing
+ *                 Locale::getFrench() for inLocale would result in "Anglais",
+ *                 while passing Locale::getGerman() for inLocale would result
+ *                 in "Englisch". NULL may be used to specify the default.
+ * @param country the displayable country code for localeID.
+ * @param countryCapacity the size of the country buffer to store the
+ *                 displayable country code with.
+ * @param status error information if retrieving the displayable country code
+ *                 failed. U_USING_DEFAULT_WARNING indicates that no data was
+ *                 found from the locale resources and a case canonicalized
+ *                 country code is placed into country as fallback.
+ * @return the actual buffer size needed for the displayable country code. If
+ *                 it's greater than countryCapacity, the returned displayable
+ *                 country code will be truncated.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayCountry(const char* locale,
+                       const char* displayLocale,
+                       UChar* country,
+                       int32_t countryCapacity,
+                       UErrorCode* status);
+
+
+/**
+ * Gets the variant name suitable for display for the specified locale.
+ *
+ * @param locale the locale to get the displayable variant code with. NULL may
+ *                 be used to specify the default.
+ * @param displayLocale Specifies the locale to be used to display the name. In
+ *                 other words, if the locale's language code is "en", passing
+ *                 Locale::getFrench() for inLocale would result in "Anglais",
+ *                 while passing Locale::getGerman() for inLocale would result
+ *                 in "Englisch". NULL may be used to specify the default.
+ * @param variant the displayable variant code for localeID.
+ * @param variantCapacity the size of the variant buffer to store the
+ *                 displayable variant code with.
+ * @param status error information if retrieving the displayable variant code
+ *                 failed. U_USING_DEFAULT_WARNING indicates that no data was
+ *                 found from the locale resources and a case canonicalized
+ *                 variant code is placed into variant as fallback.
+ * @return the actual buffer size needed for the displayable variant code. If
+ *                 it's greater than variantCapacity, the returned displayable
+ *                 variant code will be truncated.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayVariant(const char* locale,
+                       const char* displayLocale,
+                       UChar* variant,
+                       int32_t variantCapacity,
+                       UErrorCode* status);
+
+/**
+ * Gets the keyword name suitable for display for the specified locale. E.g:
+ * for the locale string de_DE\@collation=PHONEBOOK, this API gets the display
+ * string for the keyword collation.
+ * Usage:
+ * <code>
+ *    UErrorCode status = U_ZERO_ERROR;
+ *    const char* keyword =NULL;
+ *    int32_t keywordLen = 0;
+ *    int32_t keywordCount = 0;
+ *    UChar displayKeyword[256];
+ *    int32_t displayKeywordLen = 0;
+ *    UEnumeration* keywordEnum = uloc_openKeywords("de_DE@collation=PHONEBOOK;calendar=TRADITIONAL", &status);
+ *    for(keywordCount = uenum_count(keywordEnum, &status); keywordCount > 0 ; keywordCount--){
+ *          if(U_FAILURE(status)){
+ *              ...something went wrong so handle the error...
+ *              break;
+ *          }
+ *          // the uenum_next returns NUL terminated string
+ *          keyword = uenum_next(keywordEnum, &keywordLen, &status);
+ *          displayKeywordLen = uloc_getDisplayKeyword(keyword, "en_US", displayKeyword, 256);
+ *          ... do something interesting .....
+ *    }
+ *    uenum_close(keywordEnum);
+ * </code>
+ * @param keyword           The keyword whose display string needs to be returned.
+ * @param displayLocale     Specifies the locale to be used to display the name.  In other words,
+ *                          if the locale's language code is "en", passing Locale::getFrench() for
+ *                          inLocale would result in "Anglais", while passing Locale::getGerman()
+ *                          for inLocale would result in "Englisch". NULL may be used to specify the default.
+ * @param dest              the buffer to which the displayable keyword should be written.
+ * @param destCapacity      The size of the buffer (number of UChars). If it is 0, then
+ *                          dest may be NULL and the function will only return the length of the
+ *                          result without writing any of the result string (pre-flighting).
+ * @param status            error information if retrieving the displayable string failed.
+ *                          Should not be NULL and should not indicate failure on entry.
+ *                          U_USING_DEFAULT_WARNING indicates that no data was found from the locale
+ *                          resources and the keyword is placed into dest as fallback.
+ * @return the actual buffer size needed for the displayable variant code.
+ * @see #uloc_openKeywords
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayKeyword(const char* keyword,
+                       const char* displayLocale,
+                       UChar* dest,
+                       int32_t destCapacity,
+                       UErrorCode* status);
+/**
+ * Gets the value of the keyword suitable for display for the specified locale.
+ * E.g: for the locale string de_DE\@collation=PHONEBOOK, this API gets the display
+ * string for PHONEBOOK, in the display locale, when "collation" is specified as the keyword.
+ *
+ * @param locale            The locale to get the displayable variant code with. NULL may be used to specify the default.
+ * @param keyword           The keyword for whose value should be used.
+ * @param displayLocale     Specifies the locale to be used to display the name.  In other words,
+ *                          if the locale's language code is "en", passing Locale::getFrench() for
+ *                          inLocale would result in "Anglais", while passing Locale::getGerman()
+ *                          for inLocale would result in "Englisch". NULL may be used to specify the default.
+ * @param dest              the buffer to which the displayable keyword should be written.
+ * @param destCapacity      The size of the buffer (number of UChars). If it is 0, then
+ *                          dest may be NULL and the function will only return the length of the
+ *                          result without writing any of the result string (pre-flighting).
+ * @param status            error information if retrieving the displayable string failed.
+ *                          Should not be NULL and must not indicate failure on entry.
+ *                          U_USING_DEFAULT_WARNING indicates that no data was found from the locale
+ *                          resources and the value of the keyword is placed into dest as fallback.
+ * @return the actual buffer size needed for the displayable variant code.
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayKeywordValue(   const char* locale,
+                               const char* keyword,
+                               const char* displayLocale,
+                               UChar* dest,
+                               int32_t destCapacity,
+                               UErrorCode* status);
+/**
+ * Gets the full name suitable for display for the specified locale.
+ *
+ * @param localeID the locale to get the displayable name with. NULL may be used to specify the default.
+ * @param inLocaleID Specifies the locale to be used to display the name.  In other words,
+ *                   if the locale's language code is "en", passing Locale::getFrench() for
+ *                   inLocale would result in "Anglais", while passing Locale::getGerman()
+ *                   for inLocale would result in "Englisch". NULL may be used to specify the default.
+ * @param result the displayable name for localeID
+ * @param maxResultSize the size of the name buffer to store the 
+ * displayable full name with
+ * @param err error information if retrieving the displayable name failed
+ * @return the actual buffer size needed for the displayable name.  If it's greater 
+ * than maxResultSize, the returned displayable name will be truncated.  
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getDisplayName(const char* localeID,
+            const char* inLocaleID,
+            UChar* result,
+            int32_t maxResultSize,
+            UErrorCode* err);
+
+
+/**
+ * Gets the specified locale from a list of available locales.
+ *
+ * This method corresponds to uloc_openAvailableByType called with the
+ * ULOC_AVAILABLE_DEFAULT type argument.
+ *
+ * The return value is a pointer to an item of a locale name array. Both this
+ * array and the pointers it contains are owned by ICU and should not be
+ * deleted or written through by the caller. The locale name is terminated by
+ * a null pointer.
+ *
+ * @param n the specific locale name index of the available locale list;
+ *     should not exceed the number returned by uloc_countAvailable.
+ * @return a specified locale name of all available locales
+ * @stable ICU 2.0
+ */
+U_CAPI const char* U_EXPORT2
+uloc_getAvailable(int32_t n);
+
+/**
+ * Gets the size of the all available locale list.
+ *
+ * @return the size of the locale list
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2 uloc_countAvailable(void);
+
+/**
+ * Types for uloc_getAvailableByType and uloc_countAvailableByType.
+ *
+ * @stable ICU 65
+ */
+typedef enum ULocAvailableType {
+  /**
+   * Locales that return data when passed to ICU APIs,
+   * but not including legacy or alias locales.
+   *
+   * @stable ICU 65
+   */
+  ULOC_AVAILABLE_DEFAULT,
+
+  /**
+   * Legacy or alias locales that return data when passed to ICU APIs.
+   * Examples of supported legacy or alias locales:
+   *
+   * - iw (alias to he)
+   * - mo (alias to ro)
+   * - zh_CN (alias to zh_Hans_CN)
+   * - sr_BA (alias to sr_Cyrl_BA)
+   * - ars (alias to ar_SA)
+   *
+   * The locales in this set are disjoint from the ones in
+   * ULOC_AVAILABLE_DEFAULT. To get both sets at the same time, use
+   * ULOC_AVAILABLE_WITH_LEGACY_ALIASES.
+   *
+   * @stable ICU 65
+   */
+  ULOC_AVAILABLE_ONLY_LEGACY_ALIASES,
+
+  /**
+   * The union of the locales in ULOC_AVAILABLE_DEFAULT and
+   * ULOC_AVAILABLE_ONLY_LEGACY_ALIAS.
+   *
+   * @stable ICU 65
+   */
+  ULOC_AVAILABLE_WITH_LEGACY_ALIASES,
+
+#ifndef U_HIDE_INTERNAL_API
+  /**
+   * @internal
+   */
+  ULOC_AVAILABLE_COUNT
+#endif
+} ULocAvailableType;
+
+/**
+ * Gets a list of available locales according to the type argument, allowing
+ * the user to access different sets of supported locales in ICU.
+ *
+ * The returned UEnumeration must be closed by the caller.
+ *
+ * @param type Type choice from ULocAvailableType.
+ * @param status Set if an error occurred.
+ * @return a UEnumeration owned by the caller, or nullptr on failure.
+ * @stable ICU 65
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openAvailableByType(ULocAvailableType type, UErrorCode* status);
+
+/**
+ *
+ * Gets a list of all available 2-letter language codes defined in ISO 639,
+ * plus additional 3-letter codes determined to be useful for locale generation as
+ * defined by Unicode CLDR. This is a pointer
+ * to an array of pointers to arrays of char.  All of these pointers are owned
+ * by ICU-- do not delete them, and do not write through them.  The array is
+ * terminated with a null pointer.
+ * @return a list of all available language codes
+ * @stable ICU 2.0
+ */
+U_CAPI const char* const* U_EXPORT2
+uloc_getISOLanguages(void);
+
+/**
+ *
+ * Gets a list of all available 2-letter country codes defined in ISO 639.  This is a
+ * pointer to an array of pointers to arrays of char.  All of these pointers are
+ * owned by ICU-- do not delete them, and do not write through them.  The array is
+ * terminated with a null pointer.
+ * @return a list of all available country codes
+ * @stable ICU 2.0
+ */
+U_CAPI const char* const* U_EXPORT2
+uloc_getISOCountries(void);
+
+/**
+ * Truncate the locale ID string to get the parent locale ID.
+ * Copies the part of the string before the last underscore.
+ * The parent locale ID will be an empty string if there is no
+ * underscore, or if there is only one underscore at localeID[0].
+ *
+ * @param localeID Input locale ID string.
+ * @param parent   Output string buffer for the parent locale ID.
+ * @param parentCapacity Size of the output buffer.
+ * @param err A UErrorCode value.
+ * @return The length of the parent locale ID.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getParent(const char*    localeID,
+                 char* parent,
+                 int32_t parentCapacity,
+                 UErrorCode* err);
+
+
+
+
+/**
+ * Gets the full name for the specified locale, like uloc_getName(),
+ * but without keywords.
+ *
+ * Note: This has the effect of 'canonicalizing' the string to
+ * a certain extent. Upper and lower case are set as needed,
+ * and if the components were in 'POSIX' format they are changed to
+ * ICU format.  It does NOT map aliased names in any way.
+ * See the top of this header file.
+ *
+ * This API strips off the keyword part, so "de_DE\@collation=phonebook" 
+ * will become "de_DE". 
+ * This API supports preflighting.
+ *
+ * @param localeID the locale to get the full name with
+ * @param name fill in buffer for the name without keywords.
+ * @param nameCapacity capacity of the fill in buffer.
+ * @param err error information if retrieving the full name failed
+ * @return the actual buffer size needed for the full name.  If it's greater 
+ * than nameCapacity, the returned full name will be truncated.  
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getBaseName(const char*    localeID,
+         char* name,
+         int32_t nameCapacity,
+         UErrorCode* err);
+
+/**
+ * Gets an enumeration of keywords for the specified locale. Enumeration
+ * must get disposed of by the client using uenum_close function.
+ *
+ * @param localeID the locale to get the variant code with
+ * @param status error information if retrieving the keywords failed
+ * @return enumeration of keywords or NULL if there are no keywords.
+ * @stable ICU 2.8
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uloc_openKeywords(const char* localeID,
+                        UErrorCode* status);
+
+/**
+ * Get the value for a keyword. Locale name does not need to be normalized.
+ * 
+ * @param localeID locale name containing the keyword ("de_DE@currency=EURO;collation=PHONEBOOK")
+ * @param keywordName name of the keyword for which we want the value; must not be
+ *  NULL or empty, and must consist only of [A-Za-z0-9]. Case insensitive.
+ * @param buffer receiving buffer
+ * @param bufferCapacity capacity of receiving buffer
+ * @param status containing error code: e.g. buffer not big enough or ill-formed localeID
+ *  or keywordName parameters.
+ * @return the length of keyword value
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getKeywordValue(const char* localeID,
+                     const char* keywordName,
+                     char* buffer, int32_t bufferCapacity,
+                     UErrorCode* status);
+
+
+/**
+ * Sets or removes the value of the specified keyword.
+ *
+ * For removing all keywords, use uloc_getBaseName().
+ *
+ * NOTE: Unlike almost every other ICU function which takes a
+ * buffer, this function will NOT truncate the output text, and will
+ * not update the buffer with unterminated text setting a status of
+ * U_STRING_NOT_TERMINATED_WARNING. If a BUFFER_OVERFLOW_ERROR is received,
+ * it means a terminated version of the updated locale ID would not fit
+ * in the buffer, and the original buffer is untouched. This is done to
+ * prevent incorrect or possibly even malformed locales from being generated
+ * and used.
+ *
+ * @param keywordName name of the keyword to be set; must not be
+ *  NULL or empty, and must consist only of [A-Za-z0-9]. Case insensitive.
+ * @param keywordValue value of the keyword to be set. If 0-length or
+ *  NULL, will result in the keyword being removed; no error is given if 
+ *  that keyword does not exist. Otherwise, must consist only of
+ *  [A-Za-z0-9] and [/_+-].
+ * @param buffer input buffer containing well-formed locale ID to be
+ *  modified.
+ * @param bufferCapacity capacity of receiving buffer
+ * @param status containing error code: e.g. buffer not big enough
+ *  or ill-formed keywordName or keywordValue parameters, or ill-formed
+ *  locale ID in buffer on input.
+ * @return the length needed for the buffer
+ * @see uloc_getKeywordValue
+ * @stable ICU 3.2
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_setKeywordValue(const char* keywordName,
+                     const char* keywordValue,
+                     char* buffer, int32_t bufferCapacity,
+                     UErrorCode* status);
+
+/**
+ * Returns whether the locale's script is written right-to-left.
+ * If there is no script subtag, then the likely script is used, see uloc_addLikelySubtags().
+ * If no likely script is known, then false is returned.
+ *
+ * A script is right-to-left according to the CLDR script metadata
+ * which corresponds to whether the script's letters have Bidi_Class=R or AL.
+ *
+ * Returns true for "ar" and "en-Hebr", false for "zh" and "fa-Cyrl".
+ *
+ * @param locale input locale ID
+ * @return true if the locale's script is written right-to-left
+ * @stable ICU 54
+ */
+U_CAPI UBool U_EXPORT2
+uloc_isRightToLeft(const char *locale);
+
+/**
+ * enums for the  return value for the character and line orientation
+ * functions.
+ * @stable ICU 4.0
+ */
+typedef enum {
+  ULOC_LAYOUT_LTR   = 0,  /* left-to-right. */
+  ULOC_LAYOUT_RTL    = 1,  /* right-to-left. */
+  ULOC_LAYOUT_TTB    = 2,  /* top-to-bottom. */
+  ULOC_LAYOUT_BTT    = 3,   /* bottom-to-top. */
+  ULOC_LAYOUT_UNKNOWN
+} ULayoutType;
+
+/**
+ * Get the layout character orientation for the specified locale.
+ * 
+ * @param localeId locale name
+ * @param status Error status
+ * @return an enum indicating the layout orientation for characters.
+ * @stable ICU 4.0
+ */
+U_CAPI ULayoutType U_EXPORT2
+uloc_getCharacterOrientation(const char* localeId,
+                             UErrorCode *status);
+
+/**
+ * Get the layout line orientation for the specified locale.
+ * 
+ * @param localeId locale name
+ * @param status Error status
+ * @return an enum indicating the layout orientation for lines.
+ * @stable ICU 4.0
+ */
+U_CAPI ULayoutType U_EXPORT2
+uloc_getLineOrientation(const char* localeId,
+                        UErrorCode *status);
+
+/**
+ * Output values which uloc_acceptLanguage() writes to the 'outResult' parameter.
+ *
+ * @see uloc_acceptLanguageFromHTTP
+ * @see uloc_acceptLanguage
+ * @stable ICU 3.2
+ */
+typedef enum {
+    /**
+     * No exact match was found.
+     * @stable ICU 3.2
+     */
+    ULOC_ACCEPT_FAILED   = 0,
+    /**
+     * An exact match was found.
+     * @stable ICU 3.2
+     */
+    ULOC_ACCEPT_VALID    = 1,
+    /**
+     * A fallback was found. For example, the Accept-Language list includes 'ja_JP'
+     * and is matched with available locale 'ja'.
+     * @stable ICU 3.2
+     */
+    ULOC_ACCEPT_FALLBACK = 2   /*  */
+} UAcceptResult;
+
+/**
+ * Based on a HTTP header from a web browser and a list of available locales,
+ * determine an acceptable locale for the user.
+ *
+ * This is a thin wrapper over C++ class LocaleMatcher.
+ *
+ * @param result - buffer to accept the result locale
+ * @param resultAvailable the size of the result buffer.
+ * @param outResult - An out parameter that contains the fallback status
+ * @param httpAcceptLanguage - "Accept-Language:" header as per HTTP.
+ * @param availableLocales - list of available locales to match
+ * @param status ICU error code. Its input value must pass the U_SUCCESS() test,
+ *               or else the function returns immediately. Check for U_FAILURE()
+ *               on output or use with function chaining. (See User Guide for details.)
+ * @return length needed for the locale.
+ * @stable ICU 3.2
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_acceptLanguageFromHTTP(char *result, int32_t resultAvailable,
+                            UAcceptResult *outResult,
+                            const char *httpAcceptLanguage,
+                            UEnumeration* availableLocales,
+                            UErrorCode *status);
+
+/**
+ * Based on a list of available locales,
+ * determine an acceptable locale for the user.
+ *
+ * This is a thin wrapper over C++ class LocaleMatcher.
+ *
+ * @param result - buffer to accept the result locale
+ * @param resultAvailable the size of the result buffer.
+ * @param outResult - An out parameter that contains the fallback status
+ * @param acceptList - list of acceptable languages
+ * @param acceptListCount - count of acceptList items
+ * @param availableLocales - list of available locales to match
+ * @param status ICU error code. Its input value must pass the U_SUCCESS() test,
+ *               or else the function returns immediately. Check for U_FAILURE()
+ *               on output or use with function chaining. (See User Guide for details.)
+ * @return length needed for the locale.
+ * @stable ICU 3.2
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_acceptLanguage(char *result, int32_t resultAvailable, 
+                    UAcceptResult *outResult, const char **acceptList,
+                    int32_t acceptListCount,
+                    UEnumeration* availableLocales,
+                    UErrorCode *status);
+
+
+/**
+ * Gets the ICU locale ID for the specified Win32 LCID value.
+ *
+ * @param hostID the Win32 LCID to translate
+ * @param locale the output buffer for the ICU locale ID, which will be NUL-terminated
+ *  if there is room.
+ * @param localeCapacity the size of the output buffer
+ * @param status an error is returned if the LCID is unrecognized or the output buffer
+ *  is too small
+ * @return actual the actual size of the locale ID, not including NUL-termination 
+ * @stable ICU 3.8
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_getLocaleForLCID(uint32_t hostID, char *locale, int32_t localeCapacity,
+                    UErrorCode *status);
+
+
+/**
+ * Add the likely subtags for a provided locale ID, per the algorithm described
+ * in the following CLDR technical report:
+ *
+ *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+ *
+ * If localeID is already in the maximal form, or there is no data available
+ * for maximization, it will be copied to the output buffer.  For example,
+ * "und-Zzzz" cannot be maximized, since there is no reasonable maximization.
+ *
+ * Examples:
+ *
+ * "en" maximizes to "en_Latn_US"
+ *
+ * "de" maximizes to "de_Latn_US"
+ *
+ * "sr" maximizes to "sr_Cyrl_RS"
+ *
+ * "sh" maximizes to "sr_Latn_RS" (Note this will not reverse.)
+ *
+ * "zh_Hani" maximizes to "zh_Hans_CN" (Note this will not reverse.)
+ *
+ * @param localeID The locale to maximize
+ * @param maximizedLocaleID The maximized locale
+ * @param maximizedLocaleIDCapacity The capacity of the maximizedLocaleID buffer
+ * @param err Error information if maximizing the locale failed.  If the length
+ * of the localeID and the null-terminator is greater than the maximum allowed size,
+ * or the localeId is not well-formed, the error code is U_ILLEGAL_ARGUMENT_ERROR.
+ * @return The actual buffer size needed for the maximized locale.  If it's
+ * greater than maximizedLocaleIDCapacity, the returned ID will be truncated.
+ * On error, the return value is -1.
+ * @stable ICU 4.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_addLikelySubtags(const char*    localeID,
+         char* maximizedLocaleID,
+         int32_t maximizedLocaleIDCapacity,
+         UErrorCode* err);
+
+
+/**
+ * Minimize the subtags for a provided locale ID, per the algorithm described
+ * in the following CLDR technical report:
+ *
+ *   http://www.unicode.org/reports/tr35/#Likely_Subtags
+ *
+ * If localeID is already in the minimal form, or there is no data available
+ * for minimization, it will be copied to the output buffer.  Since the
+ * minimization algorithm relies on proper maximization, see the comments
+ * for uloc_addLikelySubtags for reasons why there might not be any data.
+ *
+ * Examples:
+ *
+ * "en_Latn_US" minimizes to "en"
+ *
+ * "de_Latn_US" minimizes to "de"
+ *
+ * "sr_Cyrl_RS" minimizes to "sr"
+ *
+ * "zh_Hant_TW" minimizes to "zh_TW" (The region is preferred to the
+ * script, and minimizing to "zh" would imply "zh_Hans_CN".)
+ *
+ * @param localeID The locale to minimize
+ * @param minimizedLocaleID The minimized locale
+ * @param minimizedLocaleIDCapacity The capacity of the minimizedLocaleID buffer
+ * @param err Error information if minimizing the locale failed.  If the length
+ * of the localeID and the null-terminator is greater than the maximum allowed size,
+ * or the localeId is not well-formed, the error code is U_ILLEGAL_ARGUMENT_ERROR.
+ * @return The actual buffer size needed for the minimized locale.  If it's
+ * greater than minimizedLocaleIDCapacity, the returned ID will be truncated.
+ * On error, the return value is -1.
+ * @stable ICU 4.0
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_minimizeSubtags(const char*    localeID,
+         char* minimizedLocaleID,
+         int32_t minimizedLocaleIDCapacity,
+         UErrorCode* err);
+
+/**
+ * Returns a locale ID for the specified BCP47 language tag string.
+ * If the specified language tag contains any ill-formed subtags,
+ * the first such subtag and all following subtags are ignored.
+ * <p>
+ * This implements the 'Language-Tag' production of BCP 47, and so
+ * supports legacy language tags (marked as “Type: grandfathered” in BCP 47)
+ * (regular and irregular) as well as private use language tags.
+ *
+ * Private use tags are represented as 'x-whatever',
+ * and legacy tags are converted to their canonical replacements where they exist.
+ *
+ * Note that a few legacy tags have no modern replacement;
+ * these will be converted using the fallback described in
+ * the first paragraph, so some information might be lost.
+ *
+ * @param langtag   the input BCP47 language tag.
+ * @param localeID  the output buffer receiving a locale ID for the
+ *                  specified BCP47 language tag.
+ * @param localeIDCapacity  the size of the locale ID output buffer.
+ * @param parsedLength  if not NULL, successfully parsed length
+ *                      for the input language tag is set.
+ * @param err       error information if receiving the locald ID
+ *                  failed.
+ * @return          the length of the locale ID.
+ * @stable ICU 4.2
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_forLanguageTag(const char* langtag,
+                    char* localeID,
+                    int32_t localeIDCapacity,
+                    int32_t* parsedLength,
+                    UErrorCode* err);
+
+/**
+ * Returns a well-formed language tag for this locale ID. 
+ * <p> 
+ * <b>Note</b>: When <code>strict</code> is false, any locale
+ * fields which do not satisfy the BCP47 syntax requirement will
+ * be omitted from the result.  When <code>strict</code> is
+ * true, this function sets U_ILLEGAL_ARGUMENT_ERROR to the
+ * <code>err</code> if any locale fields do not satisfy the
+ * BCP47 syntax requirement.
+ * @param localeID  the input locale ID
+ * @param langtag   the output buffer receiving BCP47 language
+ *                  tag for the locale ID.
+ * @param langtagCapacity   the size of the BCP47 language tag
+ *                          output buffer.
+ * @param strict    boolean value indicating if the function returns
+ *                  an error for an ill-formed input locale ID.
+ * @param err       error information if receiving the language
+ *                  tag failed.
+ * @return          The length of the BCP47 language tag.
+ * @stable ICU 4.2
+ */
+U_CAPI int32_t U_EXPORT2
+uloc_toLanguageTag(const char* localeID,
+                   char* langtag,
+                   int32_t langtagCapacity,
+                   UBool strict,
+                   UErrorCode* err);
+
+/**
+ * Converts the specified keyword (legacy key, or BCP 47 Unicode locale
+ * extension key) to the equivalent BCP 47 Unicode locale extension key.
+ * For example, BCP 47 Unicode locale extension key "co" is returned for
+ * the input keyword "collation".
+ * <p>
+ * When the specified keyword is unknown, but satisfies the BCP syntax,
+ * then the pointer to the input keyword itself will be returned.
+ * For example,
+ * <code>uloc_toUnicodeLocaleKey("ZZ")</code> returns "ZZ".
+ * 
+ * @param keyword       the input locale keyword (either legacy key
+ *                      such as "collation" or BCP 47 Unicode locale extension
+ *                      key such as "co").
+ * @return              the well-formed BCP 47 Unicode locale extension key,
+ *                      or NULL if the specified locale keyword cannot be
+ *                      mapped to a well-formed BCP 47 Unicode locale extension
+ *                      key. 
+ * @see uloc_toLegacyKey
+ * @stable ICU 54
+ */
+U_CAPI const char* U_EXPORT2
+uloc_toUnicodeLocaleKey(const char* keyword);
+
+/**
+ * Converts the specified keyword value (legacy type, or BCP 47
+ * Unicode locale extension type) to the well-formed BCP 47 Unicode locale
+ * extension type for the specified keyword (category). For example, BCP 47
+ * Unicode locale extension type "phonebk" is returned for the input
+ * keyword value "phonebook", with the keyword "collation" (or "co").
+ * <p>
+ * When the specified keyword is not recognized, but the specified value
+ * satisfies the syntax of the BCP 47 Unicode locale extension type,
+ * or when the specified keyword allows 'variable' type and the specified
+ * value satisfies the syntax,  then the pointer to the input type value itself
+ * will be returned.
+ * For example,
+ * <code>uloc_toUnicodeLocaleType("Foo", "Bar")</code> returns "Bar",
+ * <code>uloc_toUnicodeLocaleType("variableTop", "00A4")</code> returns "00A4".
+ * 
+ * @param keyword       the locale keyword (either legacy key such as
+ *                      "collation" or BCP 47 Unicode locale extension
+ *                      key such as "co").
+ * @param value         the locale keyword value (either legacy type
+ *                      such as "phonebook" or BCP 47 Unicode locale extension
+ *                      type such as "phonebk").
+ * @return              the well-formed BCP47 Unicode locale extension type,
+ *                      or NULL if the locale keyword value cannot be mapped to
+ *                      a well-formed BCP 47 Unicode locale extension type.
+ * @see uloc_toLegacyType
+ * @stable ICU 54
+ */
+U_CAPI const char* U_EXPORT2
+uloc_toUnicodeLocaleType(const char* keyword, const char* value);
+
+/**
+ * Converts the specified keyword (BCP 47 Unicode locale extension key, or
+ * legacy key) to the legacy key. For example, legacy key "collation" is
+ * returned for the input BCP 47 Unicode locale extension key "co".
+ * 
+ * @param keyword       the input locale keyword (either BCP 47 Unicode locale
+ *                      extension key or legacy key).
+ * @return              the well-formed legacy key, or NULL if the specified
+ *                      keyword cannot be mapped to a well-formed legacy key.
+ * @see toUnicodeLocaleKey
+ * @stable ICU 54
+ */
+U_CAPI const char* U_EXPORT2
+uloc_toLegacyKey(const char* keyword);
+
+/**
+ * Converts the specified keyword value (BCP 47 Unicode locale extension type,
+ * or legacy type or type alias) to the canonical legacy type. For example,
+ * the legacy type "phonebook" is returned for the input BCP 47 Unicode
+ * locale extension type "phonebk" with the keyword "collation" (or "co").
+ * <p>
+ * When the specified keyword is not recognized, but the specified value
+ * satisfies the syntax of legacy key, or when the specified keyword
+ * allows 'variable' type and the specified value satisfies the syntax,
+ * then the pointer to the input type value itself will be returned.
+ * For example,
+ * <code>uloc_toLegacyType("Foo", "Bar")</code> returns "Bar",
+ * <code>uloc_toLegacyType("vt", "00A4")</code> returns "00A4".
+ *
+ * @param keyword       the locale keyword (either legacy keyword such as
+ *                      "collation" or BCP 47 Unicode locale extension
+ *                      key such as "co").
+ * @param value         the locale keyword value (either BCP 47 Unicode locale
+ *                      extension type such as "phonebk" or legacy keyword value
+ *                      such as "phonebook").
+ * @return              the well-formed legacy type, or NULL if the specified
+ *                      keyword value cannot be mapped to a well-formed legacy
+ *                      type.
+ * @see toUnicodeLocaleType
+ * @stable ICU 54
+ */
+U_CAPI const char* U_EXPORT2
+uloc_toLegacyType(const char* keyword, const char* value);
+
+#endif /*_ULOC*/
diff --git a/thirdparty/icu4c/common/unicode/umachine.h b/thirdparty/icu4c/common/unicode/umachine.h
new file mode 100644
index 0000000000..09c887c80e
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/umachine.h
@@ -0,0 +1,491 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  umachine.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep13
+*   created by: Markus W. Scherer
+*
+*   This file defines basic types and constants for ICU to be
+*   platform-independent. umachine.h and utf.h are included into
+*   utypes.h to provide all the general definitions for ICU.
+*   All of these definitions used to be in utypes.h before
+*   the UTF-handling macros made this unmaintainable.
+*/
+
+#ifndef __UMACHINE_H__
+#define __UMACHINE_H__
+
+
+/**
+ * \file
+ * \brief Basic types and constants for UTF
+ *
+ * <h2> Basic types and constants for UTF </h2>
+ *   This file defines basic types and constants for utf.h to be
+ *   platform-independent. umachine.h and utf.h are included into
+ *   utypes.h to provide all the general definitions for ICU.
+ *   All of these definitions used to be in utypes.h before
+ *   the UTF-handling macros made this unmaintainable.
+ *
+ */
+/*==========================================================================*/
+/* Include platform-dependent definitions                                   */
+/* which are contained in the platform-specific file platform.h             */
+/*==========================================================================*/
+
+#include "unicode/ptypes.h" /* platform.h is included in ptypes.h */
+
+/*
+ * ANSI C headers:
+ * stddef.h defines wchar_t
+ */
+#include <stdbool.h>
+#include <stddef.h>
+
+/*==========================================================================*/
+/* For C wrappers, we use the symbol U_CAPI.                                */
+/* This works properly if the includer is C or C++.                         */
+/* Functions are declared   U_CAPI return-type U_EXPORT2 function-name()... */
+/*==========================================================================*/
+
+/**
+ * \def U_CFUNC
+ * This is used in a declaration of a library private ICU C function.
+ * @stable ICU 2.4
+ */
+
+/**
+ * \def U_CDECL_BEGIN
+ * This is used to begin a declaration of a library private ICU C API.
+ * @stable ICU 2.4
+ */
+
+/**
+ * \def U_CDECL_END
+ * This is used to end a declaration of a library private ICU C API
+ * @stable ICU 2.4
+ */
+
+#ifdef __cplusplus
+#   define U_CFUNC extern "C"
+#   define U_CDECL_BEGIN extern "C" {
+#   define U_CDECL_END   }
+#else
+#   define U_CFUNC extern
+#   define U_CDECL_BEGIN
+#   define U_CDECL_END
+#endif
+
+#ifndef U_ATTRIBUTE_DEPRECATED
+/**
+ * \def U_ATTRIBUTE_DEPRECATED
+ *  This is used for GCC specific attributes
+ * @internal
+ */
+#if U_GCC_MAJOR_MINOR >= 302
+#    define U_ATTRIBUTE_DEPRECATED __attribute__ ((deprecated))
+/**
+ * \def U_ATTRIBUTE_DEPRECATED
+ * This is used for Visual C++ specific attributes 
+ * @internal
+ */
+#elif defined(_MSC_VER) && (_MSC_VER >= 1400)
+#    define U_ATTRIBUTE_DEPRECATED __declspec(deprecated)
+#else
+#    define U_ATTRIBUTE_DEPRECATED
+#endif
+#endif
+
+/** This is used to declare a function as a public ICU C API @stable ICU 2.0*/
+#define U_CAPI U_CFUNC U_EXPORT
+/** Obsolete/same as U_CAPI; was used to declare a function as a stable public ICU C API*/
+#define U_STABLE U_CAPI
+/** Obsolete/same as U_CAPI; was used to declare a function as a draft public ICU C API  */
+#define U_DRAFT  U_CAPI
+/** This is used to declare a function as a deprecated public ICU C API  */
+#define U_DEPRECATED U_CAPI U_ATTRIBUTE_DEPRECATED
+/** Obsolete/same as U_CAPI; was used to declare a function as an obsolete public ICU C API  */
+#define U_OBSOLETE U_CAPI
+/** Obsolete/same as U_CAPI; was used to declare a function as an internal ICU C API  */
+#define U_INTERNAL U_CAPI
+
+/**
+ * \def U_OVERRIDE
+ * Defined to the C++11 "override" keyword if available.
+ * Denotes a class or member which is an override of the base class.
+ * May result in an error if it applied to something not an override.
+ * @internal
+ */
+#ifndef U_OVERRIDE
+#define U_OVERRIDE override
+#endif
+
+/**
+ * \def U_FINAL
+ * Defined to the C++11 "final" keyword if available.
+ * Denotes a class or member which may not be overridden in subclasses.
+ * May result in an error if subclasses attempt to override.
+ * @internal
+ */
+#if !defined(U_FINAL) || defined(U_IN_DOXYGEN)
+#define U_FINAL final
+#endif
+
+// Before ICU 65, function-like, multi-statement ICU macros were just defined as
+// series of statements wrapped in { } blocks and the caller could choose to
+// either treat them as if they were actual functions and end the invocation
+// with a trailing ; creating an empty statement after the block or else omit
+// this trailing ; using the knowledge that the macro would expand to { }.
+//
+// But doing so doesn't work well with macros that look like functions and
+// compiler warnings about empty statements (ICU-20601) and ICU 65 therefore
+// switches to the standard solution of wrapping such macros in do { } while.
+//
+// This will however break existing code that depends on being able to invoke
+// these macros without a trailing ; so to be able to remain compatible with
+// such code the wrapper is itself defined as macros so that it's possible to
+// build ICU 65 and later with the old macro behaviour, like this:
+//
+// export CPPFLAGS='-DUPRV_BLOCK_MACRO_BEGIN="" -DUPRV_BLOCK_MACRO_END=""'
+// runConfigureICU ...
+//
+
+/**
+ * \def UPRV_BLOCK_MACRO_BEGIN
+ * Defined as the "do" keyword by default.
+ * @internal
+ */
+#ifndef UPRV_BLOCK_MACRO_BEGIN
+#define UPRV_BLOCK_MACRO_BEGIN do
+#endif
+
+/**
+ * \def UPRV_BLOCK_MACRO_END
+ * Defined as "while (false)" by default.
+ * @internal
+ */
+#ifndef UPRV_BLOCK_MACRO_END
+#define UPRV_BLOCK_MACRO_END while (false)
+#endif
+
+/*==========================================================================*/
+/* limits for int32_t etc., like in POSIX inttypes.h                        */
+/*==========================================================================*/
+
+#ifndef INT8_MIN
+/** The smallest value an 8 bit signed integer can hold @stable ICU 2.0 */
+#   define INT8_MIN        ((int8_t)(-128))
+#endif
+#ifndef INT16_MIN
+/** The smallest value a 16 bit signed integer can hold @stable ICU 2.0 */
+#   define INT16_MIN       ((int16_t)(-32767-1))
+#endif
+#ifndef INT32_MIN
+/** The smallest value a 32 bit signed integer can hold @stable ICU 2.0 */
+#   define INT32_MIN       ((int32_t)(-2147483647-1))
+#endif
+
+#ifndef INT8_MAX
+/** The largest value an 8 bit signed integer can hold @stable ICU 2.0 */
+#   define INT8_MAX        ((int8_t)(127))
+#endif
+#ifndef INT16_MAX
+/** The largest value a 16 bit signed integer can hold @stable ICU 2.0 */
+#   define INT16_MAX       ((int16_t)(32767))
+#endif
+#ifndef INT32_MAX
+/** The largest value a 32 bit signed integer can hold @stable ICU 2.0 */
+#   define INT32_MAX       ((int32_t)(2147483647))
+#endif
+
+#ifndef UINT8_MAX
+/** The largest value an 8 bit unsigned integer can hold @stable ICU 2.0 */
+#   define UINT8_MAX       ((uint8_t)(255U))
+#endif
+#ifndef UINT16_MAX
+/** The largest value a 16 bit unsigned integer can hold @stable ICU 2.0 */
+#   define UINT16_MAX      ((uint16_t)(65535U))
+#endif
+#ifndef UINT32_MAX
+/** The largest value a 32 bit unsigned integer can hold @stable ICU 2.0 */
+#   define UINT32_MAX      ((uint32_t)(4294967295U))
+#endif
+
+#if defined(U_INT64_T_UNAVAILABLE)
+# error int64_t is required for decimal format and rule-based number format.
+#else
+# ifndef INT64_C
+/**
+ * Provides a platform independent way to specify a signed 64-bit integer constant.
+ * note: may be wrong for some 64 bit platforms - ensure your compiler provides INT64_C
+ * @stable ICU 2.8
+ */
+#   define INT64_C(c) c ## LL
+# endif
+# ifndef UINT64_C
+/**
+ * Provides a platform independent way to specify an unsigned 64-bit integer constant.
+ * note: may be wrong for some 64 bit platforms - ensure your compiler provides UINT64_C
+ * @stable ICU 2.8
+ */
+#   define UINT64_C(c) c ## ULL
+# endif
+# ifndef U_INT64_MIN
+/** The smallest value a 64 bit signed integer can hold @stable ICU 2.8 */
+#     define U_INT64_MIN       ((int64_t)(INT64_C(-9223372036854775807)-1))
+# endif
+# ifndef U_INT64_MAX
+/** The largest value a 64 bit signed integer can hold @stable ICU 2.8 */
+#     define U_INT64_MAX       ((int64_t)(INT64_C(9223372036854775807)))
+# endif
+# ifndef U_UINT64_MAX
+/** The largest value a 64 bit unsigned integer can hold @stable ICU 2.8 */
+#     define U_UINT64_MAX      ((uint64_t)(UINT64_C(18446744073709551615)))
+# endif
+#endif
+
+/*==========================================================================*/
+/* Boolean data type                                                        */
+/*==========================================================================*/
+
+/**
+ * The ICU boolean type, a signed-byte integer.
+ * ICU-specific for historical reasons: The C and C++ standards used to not define type bool.
+ * Also provides a fixed type definition, as opposed to
+ * type bool whose details (e.g., sizeof) may vary by compiler and between C and C++.
+ *
+ * @stable ICU 2.0
+ */
+typedef int8_t UBool;
+
+/**
+ * \def U_DEFINE_FALSE_AND_TRUE
+ * Normally turns off defining macros FALSE=0 & TRUE=1 in public ICU headers.
+ * These obsolete macros sometimes break compilation of other code that
+ * defines enum constants or similar with these names.
+ * C++ has long defined bool/false/true.
+ * C99 also added definitions for these, although as macros; see stdbool.h.
+ *
+ * You may transitionally define U_DEFINE_FALSE_AND_TRUE=1 if you need time to migrate code.
+ *
+ * @internal ICU 68
+ */
+#ifdef U_DEFINE_FALSE_AND_TRUE
+    // Use the predefined value.
+#elif defined(U_COMBINED_IMPLEMENTATION) || \
+        defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || \
+        defined(U_IO_IMPLEMENTATION) || defined(U_LAYOUTEX_IMPLEMENTATION) || \
+        defined(U_TOOLUTIL_IMPLEMENTATION)
+    // Inside ICU: Keep FALSE & TRUE available.
+#   define U_DEFINE_FALSE_AND_TRUE 1
+#else
+    // Outside ICU: Avoid collision with non-macro definitions of FALSE & TRUE.
+#   define U_DEFINE_FALSE_AND_TRUE 0
+#endif
+
+#if U_DEFINE_FALSE_AND_TRUE || defined(U_IN_DOXYGEN)
+#ifndef TRUE
+/**
+ * The TRUE value of a UBool.
+ *
+ * @deprecated ICU 68 Use standard "true" instead.
+ */
+#   define TRUE  1
+#endif
+#ifndef FALSE
+/**
+ * The FALSE value of a UBool.
+ *
+ * @deprecated ICU 68 Use standard "false" instead.
+ */
+#   define FALSE 0
+#endif
+#endif  // U_DEFINE_FALSE_AND_TRUE
+
+/*==========================================================================*/
+/* Unicode data types                                                       */
+/*==========================================================================*/
+
+/* wchar_t-related definitions -------------------------------------------- */
+
+/*
+ * \def U_WCHAR_IS_UTF16
+ * Defined if wchar_t uses UTF-16.
+ *
+ * @stable ICU 2.0
+ */
+/*
+ * \def U_WCHAR_IS_UTF32
+ * Defined if wchar_t uses UTF-32.
+ *
+ * @stable ICU 2.0
+ */
+#if !defined(U_WCHAR_IS_UTF16) && !defined(U_WCHAR_IS_UTF32)
+#   ifdef __STDC_ISO_10646__
+#       if (U_SIZEOF_WCHAR_T==2)
+#           define U_WCHAR_IS_UTF16
+#       elif (U_SIZEOF_WCHAR_T==4)
+#           define  U_WCHAR_IS_UTF32
+#       endif
+#   elif defined __UCS2__
+#       if (U_PF_OS390 <= U_PLATFORM && U_PLATFORM <= U_PF_OS400) && (U_SIZEOF_WCHAR_T==2)
+#           define U_WCHAR_IS_UTF16
+#       endif
+#   elif defined(__UCS4__) || (U_PLATFORM == U_PF_OS400 && defined(__UTF32__))
+#       if (U_SIZEOF_WCHAR_T==4)
+#           define U_WCHAR_IS_UTF32
+#       endif
+#   elif U_PLATFORM_IS_DARWIN_BASED || (U_SIZEOF_WCHAR_T==4 && U_PLATFORM_IS_LINUX_BASED)
+#       define U_WCHAR_IS_UTF32
+#   elif U_PLATFORM_HAS_WIN32_API
+#       define U_WCHAR_IS_UTF16
+#   endif
+#endif
+
+/* UChar and UChar32 definitions -------------------------------------------- */
+
+/** Number of bytes in a UChar. @stable ICU 2.0 */
+#define U_SIZEOF_UCHAR 2
+
+/**
+ * \def U_CHAR16_IS_TYPEDEF
+ * If 1, then char16_t is a typedef and not a real type (yet)
+ * @internal
+ */
+#if (U_PLATFORM == U_PF_AIX) && defined(__cplusplus) &&(U_CPLUSPLUS_VERSION < 11)
+// for AIX, uchar.h needs to be included
+# include <uchar.h>
+# define U_CHAR16_IS_TYPEDEF 1
+#elif defined(_MSC_VER) && (_MSC_VER < 1900)
+// Versions of Visual Studio/MSVC below 2015 do not support char16_t as a real type,
+// and instead use a typedef.  https://msdn.microsoft.com/library/bb531344.aspx
+# define U_CHAR16_IS_TYPEDEF 1
+#else
+# define U_CHAR16_IS_TYPEDEF 0
+#endif
+
+
+/**
+ * \var UChar
+ *
+ * The base type for UTF-16 code units and pointers.
+ * Unsigned 16-bit integer.
+ * Starting with ICU 59, C++ API uses char16_t directly, while C API continues to use UChar.
+ *
+ * UChar is configurable by defining the macro UCHAR_TYPE
+ * on the preprocessor or compiler command line:
+ * -DUCHAR_TYPE=uint16_t or -DUCHAR_TYPE=wchar_t (if U_SIZEOF_WCHAR_T==2) etc.
+ * (The UCHAR_TYPE can also be \#defined earlier in this file, for outside the ICU library code.)
+ * This is for transitional use from application code that uses uint16_t or wchar_t for UTF-16.
+ *
+ * The default is UChar=char16_t.
+ *
+ * C++11 defines char16_t as bit-compatible with uint16_t, but as a distinct type.
+ *
+ * In C, char16_t is a simple typedef of uint_least16_t.
+ * ICU requires uint_least16_t=uint16_t for data memory mapping.
+ * On macOS, char16_t is not available because the uchar.h standard header is missing.
+ *
+ * @stable ICU 4.4
+ */
+
+#if 1
+    // #if 1 is normal. UChar defaults to char16_t in C++.
+    // For configuration testing of UChar=uint16_t temporarily change this to #if 0.
+    // The intltest Makefile #defines UCHAR_TYPE=char16_t,
+    // so we only #define it to uint16_t if it is undefined so far.
+#elif !defined(UCHAR_TYPE)
+#   define UCHAR_TYPE uint16_t
+#endif
+
+#if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || \
+        defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
+    // Inside the ICU library code, never configurable.
+    typedef char16_t UChar;
+#elif defined(UCHAR_TYPE)
+    typedef UCHAR_TYPE UChar;
+#elif (U_CPLUSPLUS_VERSION >= 11)
+    typedef char16_t UChar;
+#else
+    typedef uint16_t UChar;
+#endif
+
+/**
+ * \var OldUChar
+ * Default ICU 58 definition of UChar.
+ * A base type for UTF-16 code units and pointers.
+ * Unsigned 16-bit integer.
+ *
+ * Define OldUChar to be wchar_t if that is 16 bits wide.
+ * If wchar_t is not 16 bits wide, then define UChar to be uint16_t.
+ *
+ * This makes the definition of OldUChar platform-dependent
+ * but allows direct string type compatibility with platforms with
+ * 16-bit wchar_t types.
+ *
+ * This is how UChar was defined in ICU 58, for transition convenience.
+ * Exception: ICU 58 UChar was defined to UCHAR_TYPE if that macro was defined.
+ * The current UChar responds to UCHAR_TYPE but OldUChar does not.
+ *
+ * @stable ICU 59
+ */
+#if U_SIZEOF_WCHAR_T==2
+    typedef wchar_t OldUChar;
+#elif defined(__CHAR16_TYPE__)
+    typedef __CHAR16_TYPE__ OldUChar;
+#else
+    typedef uint16_t OldUChar;
+#endif
+
+/**
+ * Define UChar32 as a type for single Unicode code points.
+ * UChar32 is a signed 32-bit integer (same as int32_t).
+ *
+ * The Unicode code point range is 0..0x10ffff.
+ * All other values (negative or >=0x110000) are illegal as Unicode code points.
+ * They may be used as sentinel values to indicate "done", "error"
+ * or similar non-code point conditions.
+ *
+ * Before ICU 2.4 (Jitterbug 2146), UChar32 was defined
+ * to be wchar_t if that is 32 bits wide (wchar_t may be signed or unsigned)
+ * or else to be uint32_t.
+ * That is, the definition of UChar32 was platform-dependent.
+ *
+ * @see U_SENTINEL
+ * @stable ICU 2.4
+ */
+typedef int32_t UChar32;
+
+/**
+ * This value is intended for sentinel values for APIs that
+ * (take or) return single code points (UChar32).
+ * It is outside of the Unicode code point range 0..0x10ffff.
+ * 
+ * For example, a "done" or "error" value in a new API
+ * could be indicated with U_SENTINEL.
+ *
+ * ICU APIs designed before ICU 2.4 usually define service-specific "done"
+ * values, mostly 0xffff.
+ * Those may need to be distinguished from
+ * actual U+ffff text contents by calling functions like
+ * CharacterIterator::hasNext() or UnicodeString::length().
+ *
+ * @return -1
+ * @see UChar32
+ * @stable ICU 2.4
+ */
+#define U_SENTINEL (-1)
+
+#include "unicode/urename.h"
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/umisc.h b/thirdparty/icu4c/common/unicode/umisc.h
new file mode 100644
index 0000000000..213290b9af
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/umisc.h
@@ -0,0 +1,62 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  umisc.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999oct15
+*   created by: Markus W. Scherer
+*/
+
+#ifndef UMISC_H
+#define UMISC_H
+
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief  C API:misc definitions 
+ *
+ *  This file contains miscellaneous definitions for the C APIs. 
+ */
+
+U_CDECL_BEGIN
+
+/** A struct representing a range of text containing a specific field 
+ *  @stable ICU 2.0
+ */
+typedef struct UFieldPosition {
+  /**
+   * The field 
+   * @stable ICU 2.0
+   */
+  int32_t field;
+  /**
+   * The start of the text range containing field 
+   * @stable ICU 2.0
+   */
+  int32_t beginIndex;
+  /** 
+   * The limit of the text range containing field 
+   * @stable ICU 2.0
+   */
+  int32_t endIndex;
+} UFieldPosition;
+
+#if !UCONFIG_NO_SERVICE
+/**
+ * Opaque type returned by registerInstance, registerFactory and unregister for service registration.
+ * @stable ICU 2.6
+ */
+typedef const void* URegistryKey;
+#endif
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/umutablecptrie.h b/thirdparty/icu4c/common/unicode/umutablecptrie.h
new file mode 100644
index 0000000000..5325d58147
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/umutablecptrie.h
@@ -0,0 +1,241 @@
+// © 2017 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// umutablecptrie.h (split out of ucptrie.h)
+// created: 2018jan24 Markus W. Scherer
+
+#ifndef __UMUTABLECPTRIE_H__
+#define __UMUTABLECPTRIE_H__
+
+#include "unicode/utypes.h"
+
+#include "unicode/ucpmap.h"
+#include "unicode/ucptrie.h"
+#include "unicode/utf8.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+U_CDECL_BEGIN
+
+/**
+ * \file
+ *
+ * This file defines a mutable Unicode code point trie.
+ *
+ * @see UCPTrie
+ * @see UMutableCPTrie
+ */
+
+/**
+ * Mutable Unicode code point trie.
+ * Fast map from Unicode code points (U+0000..U+10FFFF) to 32-bit integer values.
+ * For details see http://site.icu-project.org/design/struct/utrie
+ *
+ * Setting values (especially ranges) and lookup is fast.
+ * The mutable trie is only somewhat space-efficient.
+ * It builds a compacted, immutable UCPTrie.
+ *
+ * This trie can be modified while iterating over its contents.
+ * For example, it is possible to merge its values with those from another
+ * set of ranges (e.g., another mutable or immutable trie):
+ * Iterate over those source ranges; for each of them iterate over this trie;
+ * add the source value into the value of each trie range.
+ *
+ * @see UCPTrie
+ * @see umutablecptrie_buildImmutable
+ * @stable ICU 63
+ */
+typedef struct UMutableCPTrie UMutableCPTrie;
+
+/**
+ * Creates a mutable trie that initially maps each Unicode code point to the same value.
+ * It uses 32-bit data values until umutablecptrie_buildImmutable() is called.
+ * umutablecptrie_buildImmutable() takes a valueWidth parameter which
+ * determines the number of bits in the data value in the resulting UCPTrie.
+ * You must umutablecptrie_close() the trie once you are done using it.
+ *
+ * @param initialValue the initial value that is set for all code points
+ * @param errorValue the value for out-of-range code points and ill-formed UTF-8/16
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the trie
+ * @stable ICU 63
+ */
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_open(uint32_t initialValue, uint32_t errorValue, UErrorCode *pErrorCode);
+
+/**
+ * Clones a mutable trie.
+ * You must umutablecptrie_close() the clone once you are done using it.
+ *
+ * @param other the trie to clone
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the trie clone
+ * @stable ICU 63
+ */
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_clone(const UMutableCPTrie *other, UErrorCode *pErrorCode);
+
+/**
+ * Closes a mutable trie and releases associated memory.
+ *
+ * @param trie the trie
+ * @stable ICU 63
+ */
+U_CAPI void U_EXPORT2
+umutablecptrie_close(UMutableCPTrie *trie);
+
+/**
+ * Creates a mutable trie with the same contents as the UCPMap.
+ * You must umutablecptrie_close() the mutable trie once you are done using it.
+ *
+ * @param map the source map
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the mutable trie
+ * @stable ICU 63
+ */
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_fromUCPMap(const UCPMap *map, UErrorCode *pErrorCode);
+
+/**
+ * Creates a mutable trie with the same contents as the immutable one.
+ * You must umutablecptrie_close() the mutable trie once you are done using it.
+ *
+ * @param trie the immutable trie
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the mutable trie
+ * @stable ICU 63
+ */
+U_CAPI UMutableCPTrie * U_EXPORT2
+umutablecptrie_fromUCPTrie(const UCPTrie *trie, UErrorCode *pErrorCode);
+
+/**
+ * Returns the value for a code point as stored in the trie.
+ *
+ * @param trie the trie
+ * @param c the code point
+ * @return the value
+ * @stable ICU 63
+ */
+U_CAPI uint32_t U_EXPORT2
+umutablecptrie_get(const UMutableCPTrie *trie, UChar32 c);
+
+/**
+ * Returns the last code point such that all those from start to there have the same value.
+ * Can be used to efficiently iterate over all same-value ranges in a trie.
+ * (This is normally faster than iterating over code points and get()ting each value,
+ * but much slower than a data structure that stores ranges directly.)
+ *
+ * The trie can be modified between calls to this function.
+ *
+ * If the UCPMapValueFilter function pointer is not NULL, then
+ * the value to be delivered is passed through that function, and the return value is the end
+ * of the range where all values are modified to the same actual value.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * See the same-signature ucptrie_getRange() for a code sample.
+ *
+ * @param trie the trie
+ * @param start range start
+ * @param option defines whether surrogates are treated normally,
+ *               or as having the surrogateValue; usually UCPMAP_RANGE_NORMAL
+ * @param surrogateValue value for surrogates; ignored if option==UCPMAP_RANGE_NORMAL
+ * @param filter a pointer to a function that may modify the trie data value,
+ *     or NULL if the values from the trie are to be used unmodified
+ * @param context an opaque pointer that is passed on to the filter function
+ * @param pValue if not NULL, receives the value that every code point start..end has;
+ *     may have been modified by filter(context, trie value)
+ *     if that function pointer is not NULL
+ * @return the range end code point, or -1 if start is not a valid code point
+ * @stable ICU 63
+ */
+U_CAPI UChar32 U_EXPORT2
+umutablecptrie_getRange(const UMutableCPTrie *trie, UChar32 start,
+                        UCPMapRangeOption option, uint32_t surrogateValue,
+                        UCPMapValueFilter *filter, const void *context, uint32_t *pValue);
+
+/**
+ * Sets a value for a code point.
+ *
+ * @param trie the trie
+ * @param c the code point
+ * @param value the value
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @stable ICU 63
+ */
+U_CAPI void U_EXPORT2
+umutablecptrie_set(UMutableCPTrie *trie, UChar32 c, uint32_t value, UErrorCode *pErrorCode);
+
+/**
+ * Sets a value for each code point [start..end].
+ * Faster and more space-efficient than setting the value for each code point separately.
+ *
+ * @param trie the trie
+ * @param start the first code point to get the value
+ * @param end the last code point to get the value (inclusive)
+ * @param value the value
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @stable ICU 63
+ */
+U_CAPI void U_EXPORT2
+umutablecptrie_setRange(UMutableCPTrie *trie,
+                        UChar32 start, UChar32 end,
+                        uint32_t value, UErrorCode *pErrorCode);
+
+/**
+ * Compacts the data and builds an immutable UCPTrie according to the parameters.
+ * After this, the mutable trie will be empty.
+ *
+ * The mutable trie stores 32-bit values until buildImmutable() is called.
+ * If values shorter than 32 bits are to be stored in the immutable trie,
+ * then the upper bits are discarded.
+ * For example, when the mutable trie contains values 0x81, -0x7f, and 0xa581,
+ * and the value width is 8 bits, then each of these is stored as 0x81
+ * and the immutable trie will return that as an unsigned value.
+ * (Some implementations may want to make productive temporary use of the upper bits
+ * until buildImmutable() discards them.)
+ *
+ * Not every possible set of mappings can be built into a UCPTrie,
+ * because of limitations resulting from speed and space optimizations.
+ * Every Unicode assigned character can be mapped to a unique value.
+ * Typical data yields data structures far smaller than the limitations.
+ *
+ * It is possible to construct extremely unusual mappings that exceed the data structure limits.
+ * In such a case this function will fail with a U_INDEX_OUTOFBOUNDS_ERROR.
+ *
+ * @param trie the trie trie
+ * @param type selects the trie type
+ * @param valueWidth selects the number of bits in a trie data value; if smaller than 32 bits,
+ *                   then the values stored in the trie will be truncated first
+ * @param pErrorCode an in/out ICU UErrorCode
+ *
+ * @see umutablecptrie_fromUCPTrie
+ * @stable ICU 63
+ */
+U_CAPI UCPTrie * U_EXPORT2
+umutablecptrie_buildImmutable(UMutableCPTrie *trie, UCPTrieType type, UCPTrieValueWidth valueWidth,
+                              UErrorCode *pErrorCode);
+
+U_CDECL_END
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUMutableCPTriePointer
+ * "Smart pointer" class, closes a UMutableCPTrie via umutablecptrie_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 63
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUMutableCPTriePointer, UMutableCPTrie, umutablecptrie_close);
+
+U_NAMESPACE_END
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unifilt.h b/thirdparty/icu4c/common/unicode/unifilt.h
new file mode 100644
index 0000000000..420e1a1905
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unifilt.h
@@ -0,0 +1,136 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (C) 1999-2010, International Business Machines Corporation and others.
+* All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   11/17/99    aliu        Creation.
+**********************************************************************
+*/
+#ifndef UNIFILT_H
+#define UNIFILT_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/unifunct.h"
+#include "unicode/unimatch.h"
+
+/**
+ * \file 
+ * \brief C++ API: Unicode Filter
+ */
+
+U_NAMESPACE_BEGIN
+
+/**
+ * U_ETHER is used to represent character values for positions outside
+ * a range.  For example, transliterator uses this to represent
+ * characters outside the range contextStart..contextLimit-1.  This
+ * allows explicit matching by rules and UnicodeSets of text outside a
+ * defined range.
+ * @stable ICU 3.0
+ */
+#define U_ETHER ((char16_t)0xFFFF)
+
+/**
+ *
+ * <code>UnicodeFilter</code> defines a protocol for selecting a
+ * subset of the full range (U+0000 to U+10FFFF) of Unicode characters.
+ * Currently, filters are used in conjunction with classes like {@link
+ * Transliterator} to only process selected characters through a
+ * transformation.
+ *
+ * <p>Note: UnicodeFilter currently stubs out two pure virtual methods
+ * of its base class, UnicodeMatcher.  These methods are toPattern()
+ * and matchesIndexValue().  This is done so that filter classes that
+ * are not actually used as matchers -- specifically, those in the
+ * UnicodeFilterLogic component, and those in tests -- can continue to
+ * work without defining these methods.  As long as a filter is not
+ * used in an RBT during real transliteration, these methods will not
+ * be called.  However, this breaks the UnicodeMatcher base class
+ * protocol, and it is not a correct solution.
+ *
+ * <p>In the future we may revisit the UnicodeMatcher / UnicodeFilter
+ * hierarchy and either redesign it, or simply remove the stubs in
+ * UnicodeFilter and force subclasses to implement the full
+ * UnicodeMatcher protocol.
+ *
+ * @see UnicodeFilterLogic
+ * @stable ICU 2.0
+ */
+class U_COMMON_API UnicodeFilter : public UnicodeFunctor, public UnicodeMatcher {
+
+public:
+    /**
+     * Destructor
+     * @stable ICU 2.0
+     */
+    virtual ~UnicodeFilter();
+
+    /**
+     * Clones this object polymorphically.
+     * The caller owns the result and should delete it when done.
+     * @return clone, or nullptr if an error occurred
+     * @stable ICU 2.4
+     */
+    virtual UnicodeFilter* clone() const = 0;
+
+    /**
+     * Returns <tt>true</tt> for characters that are in the selected
+     * subset.  In other words, if a character is <b>to be
+     * filtered</b>, then <tt>contains()</tt> returns
+     * <b><tt>false</tt></b>.
+     * @stable ICU 2.0
+     */
+    virtual UBool contains(UChar32 c) const = 0;
+
+    /**
+     * UnicodeFunctor API.  Cast 'this' to a UnicodeMatcher* pointer
+     * and return the pointer.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeMatcher* toMatcher() const;
+
+    /**
+     * Implement UnicodeMatcher API.
+     * @stable ICU 2.4
+     */
+    virtual UMatchDegree matches(const Replaceable& text,
+                                 int32_t& offset,
+                                 int32_t limit,
+                                 UBool incremental);
+
+    /**
+     * UnicodeFunctor API.  Nothing to do.
+     * @stable ICU 2.4
+     */
+    virtual void setData(const TransliterationRuleData*);
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.2
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+protected:
+
+    /*
+     * Since this class has pure virtual functions,
+     * a constructor can't be used.
+     * @stable ICU 2.0
+     */
+/*    UnicodeFilter();*/
+};
+
+/*inline UnicodeFilter::UnicodeFilter() {}*/
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unifunct.h b/thirdparty/icu4c/common/unicode/unifunct.h
new file mode 100644
index 0000000000..7d31af7daf
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unifunct.h
@@ -0,0 +1,132 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2002-2005, International Business Machines Corporation
+*   and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   01/14/2002  aliu        Creation.
+**********************************************************************
+*/
+#ifndef UNIFUNCT_H
+#define UNIFUNCT_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+
+/**
+ * \file 
+ * \brief C++ API: Unicode Functor
+ */
+ 
+U_NAMESPACE_BEGIN
+
+class UnicodeMatcher;
+class UnicodeReplacer;
+class TransliterationRuleData;
+
+/**
+ * <code>UnicodeFunctor</code> is an abstract base class for objects
+ * that perform match and/or replace operations on Unicode strings.
+ * @author Alan Liu
+ * @stable ICU 2.4
+ */
+class U_COMMON_API UnicodeFunctor : public UObject {
+
+public:
+
+    /**
+     * Destructor
+     * @stable ICU 2.4
+     */
+    virtual ~UnicodeFunctor();
+
+    /**
+     * Return a copy of this object.  All UnicodeFunctor objects
+     * have to support cloning in order to allow classes using
+     * UnicodeFunctor to implement cloning.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeFunctor* clone() const = 0;
+
+    /**
+     * Cast 'this' to a UnicodeMatcher* pointer and return the
+     * pointer, or null if this is not a UnicodeMatcher*.  Subclasses
+     * that mix in UnicodeMatcher as a base class must override this.
+     * This protocol is required because a pointer to a UnicodeFunctor
+     * cannot be cast to a pointer to a UnicodeMatcher, since
+     * UnicodeMatcher is a mixin that does not derive from
+     * UnicodeFunctor.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeMatcher* toMatcher() const;
+
+    /**
+     * Cast 'this' to a UnicodeReplacer* pointer and return the
+     * pointer, or null if this is not a UnicodeReplacer*.  Subclasses
+     * that mix in UnicodeReplacer as a base class must override this.
+     * This protocol is required because a pointer to a UnicodeFunctor
+     * cannot be cast to a pointer to a UnicodeReplacer, since
+     * UnicodeReplacer is a mixin that does not derive from
+     * UnicodeFunctor.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeReplacer* toReplacer() const;
+
+    /**
+     * Return the class ID for this class.  This is useful only for
+     * comparing to a return value from getDynamicClassID().
+     * @return          The class ID for all objects of this class.
+     * @stable ICU 2.0
+     */
+    static UClassID U_EXPORT2 getStaticClassID(void);
+
+    /**
+     * Returns a unique class ID <b>polymorphically</b>.  This method
+     * is to implement a simple version of RTTI, since not all C++
+     * compilers support genuine RTTI.  Polymorphic operator==() and
+     * clone() methods call this method.
+     *
+     * <p>Concrete subclasses of UnicodeFunctor should use the macro
+     *    UOBJECT_DEFINE_RTTI_IMPLEMENTATION from uobject.h to
+     *    provide definitios getStaticClassID and getDynamicClassID.
+     *
+     * @return The class ID for this object. All objects of a given
+     * class have the same class ID.  Objects of other classes have
+     * different class IDs.
+     * @stable ICU 2.4
+     */
+    virtual UClassID getDynamicClassID(void) const = 0;
+
+    /**
+     * Set the data object associated with this functor.  The data
+     * object provides context for functor-to-standin mapping.  This
+     * method is required when assigning a functor to a different data
+     * object.  This function MAY GO AWAY later if the architecture is
+     * changed to pass data object pointers through the API.
+     * @internal ICU 2.1
+     */
+    virtual void setData(const TransliterationRuleData*) = 0;
+
+protected:
+
+    /**
+     * Since this class has pure virtual functions,
+     * a constructor can't be used.
+     * @stable ICU 2.0
+     */
+    /*UnicodeFunctor();*/
+
+};
+
+/*inline UnicodeFunctor::UnicodeFunctor() {}*/
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unimatch.h b/thirdparty/icu4c/common/unicode/unimatch.h
new file mode 100644
index 0000000000..302332f455
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unimatch.h
@@ -0,0 +1,168 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+* Copyright (C) 2001-2005, International Business Machines Corporation and others. All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   07/18/01    aliu        Creation.
+**********************************************************************
+*/
+#ifndef UNIMATCH_H
+#define UNIMATCH_H
+
+#include "unicode/utypes.h"
+
+/**
+ * \file 
+ * \brief C++ API: Unicode Matcher
+ */
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+class Replaceable;
+class UnicodeString;
+class UnicodeSet;
+
+/**
+ * Constants returned by <code>UnicodeMatcher::matches()</code>
+ * indicating the degree of match.
+ * @stable ICU 2.4
+ */
+enum UMatchDegree {
+    /**
+     * Constant returned by <code>matches()</code> indicating a
+     * mismatch between the text and this matcher.  The text contains
+     * a character which does not match, or the text does not contain
+     * all desired characters for a non-incremental match.
+     * @stable ICU 2.4
+     */
+    U_MISMATCH,
+    
+    /**
+     * Constant returned by <code>matches()</code> indicating a
+     * partial match between the text and this matcher.  This value is
+     * only returned for incremental match operations.  All characters
+     * of the text match, but more characters are required for a
+     * complete match.  Alternatively, for variable-length matchers,
+     * all characters of the text match, and if more characters were
+     * supplied at limit, they might also match.
+     * @stable ICU 2.4
+     */
+    U_PARTIAL_MATCH,
+    
+    /**
+     * Constant returned by <code>matches()</code> indicating a
+     * complete match between the text and this matcher.  For an
+     * incremental variable-length match, this value is returned if
+     * the given text matches, and it is known that additional
+     * characters would not alter the extent of the match.
+     * @stable ICU 2.4
+     */
+    U_MATCH
+};
+
+/**
+ * <code>UnicodeMatcher</code> defines a protocol for objects that can
+ * match a range of characters in a Replaceable string.
+ * @stable ICU 2.4
+ */
+class U_COMMON_API UnicodeMatcher /* not : public UObject because this is an interface/mixin class */ {
+
+public:
+    /**
+     * Destructor.
+     * @stable ICU 2.4
+     */
+    virtual ~UnicodeMatcher();
+
+    /**
+     * Return a UMatchDegree value indicating the degree of match for
+     * the given text at the given offset.  Zero, one, or more
+     * characters may be matched.
+     *
+     * Matching in the forward direction is indicated by limit >
+     * offset.  Characters from offset forwards to limit-1 will be
+     * considered for matching.
+     * 
+     * Matching in the reverse direction is indicated by limit <
+     * offset.  Characters from offset backwards to limit+1 will be
+     * considered for matching.
+     *
+     * If limit == offset then the only match possible is a zero
+     * character match (which subclasses may implement if desired).
+     *
+     * As a side effect, advance the offset parameter to the limit of
+     * the matched substring.  In the forward direction, this will be
+     * the index of the last matched character plus one.  In the
+     * reverse direction, this will be the index of the last matched
+     * character minus one.
+     *
+     * <p>Note:  This method is not const because some classes may
+     * modify their state as the result of a match.
+     *
+     * @param text the text to be matched
+     * @param offset on input, the index into text at which to begin
+     * matching.  On output, the limit of the matched text.  The
+     * number of matched characters is the output value of offset
+     * minus the input value.  Offset should always point to the
+     * HIGH SURROGATE (leading code unit) of a pair of surrogates,
+     * both on entry and upon return.
+     * @param limit the limit index of text to be matched.  Greater
+     * than offset for a forward direction match, less than offset for
+     * a backward direction match.  The last character to be
+     * considered for matching will be text.charAt(limit-1) in the
+     * forward direction or text.charAt(limit+1) in the backward
+     * direction.
+     * @param incremental if true, then assume further characters may
+     * be inserted at limit and check for partial matching.  Otherwise
+     * assume the text as given is complete.
+     * @return a match degree value indicating a full match, a partial
+     * match, or a mismatch.  If incremental is false then
+     * U_PARTIAL_MATCH should never be returned.
+     * @stable ICU 2.4
+     */
+    virtual UMatchDegree matches(const Replaceable& text,
+                                 int32_t& offset,
+                                 int32_t limit,
+                                 UBool incremental) = 0;
+
+    /**
+     * Returns a string representation of this matcher.  If the result of
+     * calling this function is passed to the appropriate parser, it
+     * will produce another matcher that is equal to this one.
+     * @param result the string to receive the pattern.  Previous
+     * contents will be deleted.
+     * @param escapeUnprintable if true then convert unprintable
+     * character to their hex escape representations, \\uxxxx or
+     * \\Uxxxxxxxx.  Unprintable characters are those other than
+     * U+000A, U+0020..U+007E.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeString& toPattern(UnicodeString& result,
+                                     UBool escapeUnprintable = false) const = 0;
+
+    /**
+     * Returns true if this matcher will match a character c, where c
+     * & 0xFF == v, at offset, in the forward direction (with limit >
+     * offset).  This is used by <tt>RuleBasedTransliterator</tt> for
+     * indexing.
+     * @stable ICU 2.4
+     */
+    virtual UBool matchesIndexValue(uint8_t v) const = 0;
+
+    /**
+     * Union the set of all characters that may be matched by this object
+     * into the given set.
+     * @param toUnionTo the set into which to union the source characters
+     * @stable ICU 2.4
+     */
+    virtual void addMatchSetTo(UnicodeSet& toUnionTo) const = 0;
+};
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uniset.h b/thirdparty/icu4c/common/unicode/uniset.h
new file mode 100644
index 0000000000..50b6360f3a
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uniset.h
@@ -0,0 +1,1744 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+***************************************************************************
+* Copyright (C) 1999-2016, International Business Machines Corporation
+* and others. All Rights Reserved.
+***************************************************************************
+*   Date        Name        Description
+*   10/20/99    alan        Creation.
+***************************************************************************
+*/
+
+#ifndef UNICODESET_H
+#define UNICODESET_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/ucpmap.h"
+#include "unicode/unifilt.h"
+#include "unicode/unistr.h"
+#include "unicode/uset.h"
+
+/**
+ * \file
+ * \brief C++ API: Unicode Set
+ */
+
+U_NAMESPACE_BEGIN
+
+// Forward Declarations.
+class BMPSet;
+class ParsePosition;
+class RBBIRuleScanner;
+class SymbolTable;
+class UnicodeSetStringSpan;
+class UVector;
+class RuleCharacterIterator;
+
+/**
+ * A mutable set of Unicode characters and multicharacter strings.  Objects of this class
+ * represent <em>character classes</em> used in regular expressions.
+ * A character specifies a subset of Unicode code points.  Legal
+ * code points are U+0000 to U+10FFFF, inclusive.
+ *
+ * <p>The UnicodeSet class is not designed to be subclassed.
+ *
+ * <p><code>UnicodeSet</code> supports two APIs. The first is the
+ * <em>operand</em> API that allows the caller to modify the value of
+ * a <code>UnicodeSet</code> object. It conforms to Java 2's
+ * <code>java.util.Set</code> interface, although
+ * <code>UnicodeSet</code> does not actually implement that
+ * interface. All methods of <code>Set</code> are supported, with the
+ * modification that they take a character range or single character
+ * instead of an <code>Object</code>, and they take a
+ * <code>UnicodeSet</code> instead of a <code>Collection</code>.  The
+ * operand API may be thought of in terms of boolean logic: a boolean
+ * OR is implemented by <code>add</code>, a boolean AND is implemented
+ * by <code>retain</code>, a boolean XOR is implemented by
+ * <code>complement</code> taking an argument, and a boolean NOT is
+ * implemented by <code>complement</code> with no argument.  In terms
+ * of traditional set theory function names, <code>add</code> is a
+ * union, <code>retain</code> is an intersection, <code>remove</code>
+ * is an asymmetric difference, and <code>complement</code> with no
+ * argument is a set complement with respect to the superset range
+ * <code>MIN_VALUE-MAX_VALUE</code>
+ *
+ * <p>The second API is the
+ * <code>applyPattern()</code>/<code>toPattern()</code> API from the
+ * <code>java.text.Format</code>-derived classes.  Unlike the
+ * methods that add characters, add categories, and control the logic
+ * of the set, the method <code>applyPattern()</code> sets all
+ * attributes of a <code>UnicodeSet</code> at once, based on a
+ * string pattern.
+ *
+ * <p><b>Pattern syntax</b></p>
+ *
+ * Patterns are accepted by the constructors and the
+ * <code>applyPattern()</code> methods and returned by the
+ * <code>toPattern()</code> method.  These patterns follow a syntax
+ * similar to that employed by version 8 regular expression character
+ * classes.  Here are some simple examples:
+ *
+ * \htmlonly<blockquote>\endhtmlonly
+ *   <table>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="left"><code>[]</code></td>
+ *       <td valign="top">No characters</td>
+ *     </tr><tr align="top">
+ *       <td nowrap valign="top" align="left"><code>[a]</code></td>
+ *       <td valign="top">The character 'a'</td>
+ *     </tr><tr align="top">
+ *       <td nowrap valign="top" align="left"><code>[ae]</code></td>
+ *       <td valign="top">The characters 'a' and 'e'</td>
+ *     </tr>
+ *     <tr>
+ *       <td nowrap valign="top" align="left"><code>[a-e]</code></td>
+ *       <td valign="top">The characters 'a' through 'e' inclusive, in Unicode code
+ *       point order</td>
+ *     </tr>
+ *     <tr>
+ *       <td nowrap valign="top" align="left"><code>[\\u4E01]</code></td>
+ *       <td valign="top">The character U+4E01</td>
+ *     </tr>
+ *     <tr>
+ *       <td nowrap valign="top" align="left"><code>[a{ab}{ac}]</code></td>
+ *       <td valign="top">The character 'a' and the multicharacter strings &quot;ab&quot; and
+ *       &quot;ac&quot;</td>
+ *     </tr>
+ *     <tr>
+ *       <td nowrap valign="top" align="left"><code>[\\p{Lu}]</code></td>
+ *       <td valign="top">All characters in the general category Uppercase Letter</td>
+ *     </tr>
+ *   </table>
+ * \htmlonly</blockquote>\endhtmlonly
+ *
+ * Any character may be preceded by a backslash in order to remove any special
+ * meaning.  White space characters, as defined by UCharacter.isWhitespace(), are
+ * ignored, unless they are escaped.
+ *
+ * <p>Property patterns specify a set of characters having a certain
+ * property as defined by the Unicode standard.  Both the POSIX-like
+ * "[:Lu:]" and the Perl-like syntax "\\p{Lu}" are recognized.  For a
+ * complete list of supported property patterns, see the User's Guide
+ * for UnicodeSet at
+ * <a href="http://icu-project.org/userguide/unicodeSet.html">
+ * http://icu-project.org/userguide/unicodeSet.html</a>.
+ * Actual determination of property data is defined by the underlying
+ * Unicode database as implemented by UCharacter.
+ *
+ * <p>Patterns specify individual characters, ranges of characters, and
+ * Unicode property sets.  When elements are concatenated, they
+ * specify their union.  To complement a set, place a '^' immediately
+ * after the opening '['.  Property patterns are inverted by modifying
+ * their delimiters; "[:^foo]" and "\\P{foo}".  In any other location,
+ * '^' has no special meaning.
+ *
+ * <p>Ranges are indicated by placing two a '-' between two
+ * characters, as in "a-z".  This specifies the range of all
+ * characters from the left to the right, in Unicode order.  If the
+ * left character is greater than or equal to the
+ * right character it is a syntax error.  If a '-' occurs as the first
+ * character after the opening '[' or '[^', or if it occurs as the
+ * last character before the closing ']', then it is taken as a
+ * literal.  Thus "[a\-b]", "[-ab]", and "[ab-]" all indicate the same
+ * set of three characters, 'a', 'b', and '-'.
+ *
+ * <p>Sets may be intersected using the '&' operator or the asymmetric
+ * set difference may be taken using the '-' operator, for example,
+ * "[[:L:]&[\\u0000-\\u0FFF]]" indicates the set of all Unicode letters
+ * with values less than 4096.  Operators ('&' and '|') have equal
+ * precedence and bind left-to-right.  Thus
+ * "[[:L:]-[a-z]-[\\u0100-\\u01FF]]" is equivalent to
+ * "[[[:L:]-[a-z]]-[\\u0100-\\u01FF]]".  This only really matters for
+ * difference; intersection is commutative.
+ *
+ * <table>
+ * <tr valign=top><td nowrap><code>[a]</code><td>The set containing 'a'
+ * <tr valign=top><td nowrap><code>[a-z]</code><td>The set containing 'a'
+ * through 'z' and all letters in between, in Unicode order
+ * <tr valign=top><td nowrap><code>[^a-z]</code><td>The set containing
+ * all characters but 'a' through 'z',
+ * that is, U+0000 through 'a'-1 and 'z'+1 through U+10FFFF
+ * <tr valign=top><td nowrap><code>[[<em>pat1</em>][<em>pat2</em>]]</code>
+ * <td>The union of sets specified by <em>pat1</em> and <em>pat2</em>
+ * <tr valign=top><td nowrap><code>[[<em>pat1</em>]&[<em>pat2</em>]]</code>
+ * <td>The intersection of sets specified by <em>pat1</em> and <em>pat2</em>
+ * <tr valign=top><td nowrap><code>[[<em>pat1</em>]-[<em>pat2</em>]]</code>
+ * <td>The asymmetric difference of sets specified by <em>pat1</em> and
+ * <em>pat2</em>
+ * <tr valign=top><td nowrap><code>[:Lu:] or \\p{Lu}</code>
+ * <td>The set of characters having the specified
+ * Unicode property; in
+ * this case, Unicode uppercase letters
+ * <tr valign=top><td nowrap><code>[:^Lu:] or \\P{Lu}</code>
+ * <td>The set of characters <em>not</em> having the given
+ * Unicode property
+ * </table>
+ *
+ * <p><b>Warning</b>: you cannot add an empty string ("") to a UnicodeSet.</p>
+ *
+ * <p><b>Formal syntax</b></p>
+ *
+ * \htmlonly<blockquote>\endhtmlonly
+ *   <table>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>pattern :=&nbsp; </code></td>
+ *       <td valign="top"><code>('[' '^'? item* ']') |
+ *       property</code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>item :=&nbsp; </code></td>
+ *       <td valign="top"><code>char | (char '-' char) | pattern-expr<br>
+ *       </code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>pattern-expr :=&nbsp; </code></td>
+ *       <td valign="top"><code>pattern | pattern-expr pattern |
+ *       pattern-expr op pattern<br>
+ *       </code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>op :=&nbsp; </code></td>
+ *       <td valign="top"><code>'&amp;' | '-'<br>
+ *       </code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>special :=&nbsp; </code></td>
+ *       <td valign="top"><code>'[' | ']' | '-'<br>
+ *       </code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>char :=&nbsp; </code></td>
+ *       <td valign="top"><em>any character that is not</em><code> special<br>
+ *       | ('\' </code><em>any character</em><code>)<br>
+ *       | ('\\u' hex hex hex hex)<br>
+ *       </code></td>
+ *     </tr>
+ *     <tr align="top">
+ *       <td nowrap valign="top" align="right"><code>hex :=&nbsp; </code></td>
+ *       <td valign="top"><em>any character for which
+ *       </em><code>Character.digit(c, 16)</code><em>
+ *       returns a non-negative result</em></td>
+ *     </tr>
+ *     <tr>
+ *       <td nowrap valign="top" align="right"><code>property :=&nbsp; </code></td>
+ *       <td valign="top"><em>a Unicode property set pattern</em></td>
+ *     </tr>
+ *   </table>
+ *   <br>
+ *   <table border="1">
+ *     <tr>
+ *       <td>Legend: <table>
+ *         <tr>
+ *           <td nowrap valign="top"><code>a := b</code></td>
+ *           <td width="20" valign="top">&nbsp; </td>
+ *           <td valign="top"><code>a</code> may be replaced by <code>b</code> </td>
+ *         </tr>
+ *         <tr>
+ *           <td nowrap valign="top"><code>a?</code></td>
+ *           <td valign="top"></td>
+ *           <td valign="top">zero or one instance of <code>a</code><br>
+ *           </td>
+ *         </tr>
+ *         <tr>
+ *           <td nowrap valign="top"><code>a*</code></td>
+ *           <td valign="top"></td>
+ *           <td valign="top">one or more instances of <code>a</code><br>
+ *           </td>
+ *         </tr>
+ *         <tr>
+ *           <td nowrap valign="top"><code>a | b</code></td>
+ *           <td valign="top"></td>
+ *           <td valign="top">either <code>a</code> or <code>b</code><br>
+ *           </td>
+ *         </tr>
+ *         <tr>
+ *           <td nowrap valign="top"><code>'a'</code></td>
+ *           <td valign="top"></td>
+ *           <td valign="top">the literal string between the quotes </td>
+ *         </tr>
+ *       </table>
+ *       </td>
+ *     </tr>
+ *   </table>
+ * \htmlonly</blockquote>\endhtmlonly
+ * 
+ * <p>Note:
+ *  - Most UnicodeSet methods do not take a UErrorCode parameter because
+ *   there are usually very few opportunities for failure other than a shortage
+ *   of memory, error codes in low-level C++ string methods would be inconvenient,
+ *   and the error code as the last parameter (ICU convention) would prevent
+ *   the use of default parameter values.
+ *   Instead, such methods set the UnicodeSet into a "bogus" state
+ *   (see isBogus()) if an error occurs.
+ *
+ * @author Alan Liu
+ * @stable ICU 2.0
+ */
+class U_COMMON_API UnicodeSet U_FINAL : public UnicodeFilter {
+private:
+    /**
+     * Enough for sets with few ranges.
+     * For example, White_Space has 10 ranges, list length 21.
+     */
+    static constexpr int32_t INITIAL_CAPACITY = 25;
+    // fFlags constant
+    static constexpr uint8_t kIsBogus = 1;  // This set is bogus (i.e. not valid)
+
+    UChar32* list = stackList; // MUST be terminated with HIGH
+    int32_t capacity = INITIAL_CAPACITY; // capacity of list
+    int32_t len = 1; // length of list used; 1 <= len <= capacity
+    uint8_t fFlags = 0;         // Bit flag (see constants above)
+
+    BMPSet *bmpSet = nullptr; // The set is frozen iff either bmpSet or stringSpan is not NULL.
+    UChar32* buffer = nullptr; // internal buffer, may be NULL
+    int32_t bufferCapacity = 0; // capacity of buffer
+
+    /**
+     * The pattern representation of this set.  This may not be the
+     * most economical pattern.  It is the pattern supplied to
+     * applyPattern(), with variables substituted and whitespace
+     * removed.  For sets constructed without applyPattern(), or
+     * modified using the non-pattern API, this string will be empty,
+     * indicating that toPattern() must generate a pattern
+     * representation from the inversion list.
+     */
+    char16_t *pat = nullptr;
+    int32_t patLen = 0;
+
+    UVector* strings = nullptr; // maintained in sorted order
+    UnicodeSetStringSpan *stringSpan = nullptr;
+
+    /**
+     * Initial list array.
+     * Avoids some heap allocations, and list is never nullptr.
+     * Increases the object size a bit.
+     */
+    UChar32 stackList[INITIAL_CAPACITY];
+
+public:
+    /**
+     * Determine if this object contains a valid set.
+     * A bogus set has no value. It is different from an empty set.
+     * It can be used to indicate that no set value is available.
+     *
+     * @return true if the set is bogus/invalid, false otherwise
+     * @see setToBogus()
+     * @stable ICU 4.0
+     */
+    inline UBool isBogus(void) const;
+
+    /**
+     * Make this UnicodeSet object invalid.
+     * The string will test true with isBogus().
+     *
+     * A bogus set has no value. It is different from an empty set.
+     * It can be used to indicate that no set value is available.
+     *
+     * This utility function is used throughout the UnicodeSet
+     * implementation to indicate that a UnicodeSet operation failed,
+     * and may be used in other functions,
+     * especially but not exclusively when such functions do not
+     * take a UErrorCode for simplicity.
+     *
+     * @see isBogus()
+     * @stable ICU 4.0
+     */
+    void setToBogus();
+
+public:
+
+    enum {
+        /**
+         * Minimum value that can be stored in a UnicodeSet.
+         * @stable ICU 2.4
+         */
+        MIN_VALUE = 0,
+
+        /**
+         * Maximum value that can be stored in a UnicodeSet.
+         * @stable ICU 2.4
+         */
+        MAX_VALUE = 0x10ffff
+    };
+
+    //----------------------------------------------------------------
+    // Constructors &c
+    //----------------------------------------------------------------
+
+public:
+
+    /**
+     * Constructs an empty set.
+     * @stable ICU 2.0
+     */
+    UnicodeSet();
+
+    /**
+     * Constructs a set containing the given range. If <code>end <
+     * start</code> then an empty set is created.
+     *
+     * @param start first character, inclusive, of range
+     * @param end last character, inclusive, of range
+     * @stable ICU 2.4
+     */
+    UnicodeSet(UChar32 start, UChar32 end);
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * @internal
+     */
+    enum ESerialization {
+      kSerialized  /* result of serialize() */
+    };
+
+    /**
+     * Constructs a set from the output of serialize().
+     *
+     * @param buffer the 16 bit array
+     * @param bufferLen the original length returned from serialize()
+     * @param serialization the value 'kSerialized'
+     * @param status error code
+     *
+     * @internal
+     */
+    UnicodeSet(const uint16_t buffer[], int32_t bufferLen,
+               ESerialization serialization, UErrorCode &status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /**
+     * Constructs a set from the given pattern.  See the class
+     * description for the syntax of the pattern language.
+     * @param pattern a string specifying what characters are in the set
+     * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
+     * contains a syntax error.
+     * @stable ICU 2.0
+     */
+    UnicodeSet(const UnicodeString& pattern,
+               UErrorCode& status);
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Constructs a set from the given pattern.  See the class
+     * description for the syntax of the pattern language.
+     * @param pattern a string specifying what characters are in the set
+     * @param options bitmask for options to apply to the pattern.
+     * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+     * @param symbols a symbol table mapping variable names to values
+     * and stand-in characters to UnicodeSets; may be NULL
+     * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
+     * contains a syntax error.
+     * @internal
+     */
+    UnicodeSet(const UnicodeString& pattern,
+               uint32_t options,
+               const SymbolTable* symbols,
+               UErrorCode& status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /**
+     * Constructs a set from the given pattern.  See the class description
+     * for the syntax of the pattern language.
+     * @param pattern a string specifying what characters are in the set
+     * @param pos on input, the position in pattern at which to start parsing.
+     * On output, the position after the last character parsed.
+     * @param options bitmask for options to apply to the pattern.
+     * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+     * @param symbols a symbol table mapping variable names to values
+     * and stand-in characters to UnicodeSets; may be NULL
+     * @param status input-output error code
+     * @stable ICU 2.8
+     */
+    UnicodeSet(const UnicodeString& pattern, ParsePosition& pos,
+               uint32_t options,
+               const SymbolTable* symbols,
+               UErrorCode& status);
+
+    /**
+     * Constructs a set that is identical to the given UnicodeSet.
+     * @stable ICU 2.0
+     */
+    UnicodeSet(const UnicodeSet& o);
+
+    /**
+     * Destructs the set.
+     * @stable ICU 2.0
+     */
+    virtual ~UnicodeSet();
+
+    /**
+     * Assigns this object to be a copy of another.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    UnicodeSet& operator=(const UnicodeSet& o);
+
+    /**
+     * Compares the specified object with this set for equality.  Returns
+     * <tt>true</tt> if the two sets
+     * have the same size, and every member of the specified set is
+     * contained in this set (or equivalently, every member of this set is
+     * contained in the specified set).
+     *
+     * @param o set to be compared for equality with this set.
+     * @return <tt>true</tt> if the specified set is equal to this set.
+     * @stable ICU 2.0
+     */
+    virtual UBool operator==(const UnicodeSet& o) const;
+
+    /**
+     * Compares the specified object with this set for equality.  Returns
+     * <tt>true</tt> if the specified set is not equal to this set.
+     * @stable ICU 2.0
+     */
+    inline UBool operator!=(const UnicodeSet& o) const;
+
+    /**
+     * Returns a copy of this object.  All UnicodeFunctor objects have
+     * to support cloning in order to allow classes using
+     * UnicodeFunctors, such as Transliterator, to implement cloning.
+     * If this set is frozen, then the clone will be frozen as well.
+     * Use cloneAsThawed() for a mutable clone of a frozen set.
+     * @see cloneAsThawed
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet* clone() const;
+
+    /**
+     * Returns the hash code value for this set.
+     *
+     * @return the hash code value for this set.
+     * @see Object#hashCode()
+     * @stable ICU 2.0
+     */
+    virtual int32_t hashCode(void) const;
+
+    /**
+     * Get a UnicodeSet pointer from a USet
+     *
+     * @param uset a USet (the ICU plain C type for UnicodeSet)
+     * @return the corresponding UnicodeSet pointer.
+     *
+     * @stable ICU 4.2
+     */
+    inline static UnicodeSet *fromUSet(USet *uset);
+
+    /**
+     * Get a UnicodeSet pointer from a const USet
+     *
+     * @param uset a const USet (the ICU plain C type for UnicodeSet)
+     * @return the corresponding UnicodeSet pointer.
+     *
+     * @stable ICU 4.2
+     */
+    inline static const UnicodeSet *fromUSet(const USet *uset);
+    
+    /**
+     * Produce a USet * pointer for this UnicodeSet.
+     * USet is the plain C type for UnicodeSet
+     *
+     * @return a USet pointer for this UnicodeSet
+     * @stable ICU 4.2
+     */
+    inline USet *toUSet();
+
+
+    /**
+     * Produce a const USet * pointer for this UnicodeSet.
+     * USet is the plain C type for UnicodeSet
+     *
+     * @return a const USet pointer for this UnicodeSet
+     * @stable ICU 4.2
+     */
+    inline const USet * toUSet() const;
+
+
+    //----------------------------------------------------------------
+    // Freezable API
+    //----------------------------------------------------------------
+
+    /**
+     * Determines whether the set has been frozen (made immutable) or not.
+     * See the ICU4J Freezable interface for details.
+     * @return true/false for whether the set has been frozen
+     * @see freeze
+     * @see cloneAsThawed
+     * @stable ICU 3.8
+     */
+    inline UBool isFrozen() const;
+
+    /**
+     * Freeze the set (make it immutable).
+     * Once frozen, it cannot be unfrozen and is therefore thread-safe
+     * until it is deleted.
+     * See the ICU4J Freezable interface for details.
+     * Freezing the set may also make some operations faster, for example
+     * contains() and span().
+     * A frozen set will not be modified. (It remains frozen.)
+     * @return this set.
+     * @see isFrozen
+     * @see cloneAsThawed
+     * @stable ICU 3.8
+     */
+    UnicodeSet *freeze();
+
+    /**
+     * Clone the set and make the clone mutable.
+     * See the ICU4J Freezable interface for details.
+     * @return the mutable clone
+     * @see freeze
+     * @see isFrozen
+     * @stable ICU 3.8
+     */
+    UnicodeSet *cloneAsThawed() const;
+
+    //----------------------------------------------------------------
+    // Public API
+    //----------------------------------------------------------------
+
+    /**
+     * Make this object represent the range `start - end`.
+     * If `end > start` then this object is set to an empty range.
+     * A frozen set will not be modified.
+     *
+     * @param start first character in the set, inclusive
+     * @param end last character in the set, inclusive
+     * @stable ICU 2.4
+     */
+    UnicodeSet& set(UChar32 start, UChar32 end);
+
+    /**
+     * Return true if the given position, in the given pattern, appears
+     * to be the start of a UnicodeSet pattern.
+     * @stable ICU 2.4
+     */
+    static UBool resemblesPattern(const UnicodeString& pattern,
+                                  int32_t pos);
+
+    /**
+     * Modifies this set to represent the set specified by the given
+     * pattern, ignoring Unicode Pattern_White_Space characters.
+     * See the class description for the syntax of the pattern language.
+     * A frozen set will not be modified.
+     * @param pattern a string specifying what characters are in the set
+     * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
+     * contains a syntax error.
+     * <em> Empties the set passed before applying the pattern.</em>
+     * @return a reference to this
+     * @stable ICU 2.0
+     */
+    UnicodeSet& applyPattern(const UnicodeString& pattern,
+                             UErrorCode& status);
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * Modifies this set to represent the set specified by the given
+     * pattern, optionally ignoring Unicode Pattern_White_Space characters.
+     * See the class description for the syntax of the pattern language.
+     * A frozen set will not be modified.
+     * @param pattern a string specifying what characters are in the set
+     * @param options bitmask for options to apply to the pattern.
+     * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+     * @param symbols a symbol table mapping variable names to
+     * values and stand-ins to UnicodeSets; may be NULL
+     * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
+     * contains a syntax error.
+     *<em> Empties the set passed before applying the pattern.</em>
+     * @return a reference to this
+     * @internal
+     */
+    UnicodeSet& applyPattern(const UnicodeString& pattern,
+                             uint32_t options,
+                             const SymbolTable* symbols,
+                             UErrorCode& status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+    /**
+     * Parses the given pattern, starting at the given position.  The
+     * character at pattern.charAt(pos.getIndex()) must be '[', or the
+     * parse fails.  Parsing continues until the corresponding closing
+     * ']'.  If a syntax error is encountered between the opening and
+     * closing brace, the parse fails.  Upon return from a successful
+     * parse, the ParsePosition is updated to point to the character
+     * following the closing ']', and a StringBuffer containing a
+     * pairs list for the parsed pattern is returned.  This method calls
+     * itself recursively to parse embedded subpatterns.
+     *<em> Empties the set passed before applying the pattern.</em>
+     * A frozen set will not be modified.
+     *
+     * @param pattern the string containing the pattern to be parsed.
+     * The portion of the string from pos.getIndex(), which must be a
+     * '[', to the corresponding closing ']', is parsed.
+     * @param pos upon entry, the position at which to being parsing.
+     * The character at pattern.charAt(pos.getIndex()) must be a '['.
+     * Upon return from a successful parse, pos.getIndex() is either
+     * the character after the closing ']' of the parsed pattern, or
+     * pattern.length() if the closing ']' is the last character of
+     * the pattern string.
+     * @param options bitmask for options to apply to the pattern.
+     * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+     * @param symbols a symbol table mapping variable names to
+     * values and stand-ins to UnicodeSets; may be NULL
+     * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
+     * contains a syntax error.
+     * @return a reference to this
+     * @stable ICU 2.8
+     */
+    UnicodeSet& applyPattern(const UnicodeString& pattern,
+                             ParsePosition& pos,
+                             uint32_t options,
+                             const SymbolTable* symbols,
+                             UErrorCode& status);
+
+    /**
+     * Returns a string representation of this set.  If the result of
+     * calling this function is passed to a UnicodeSet constructor, it
+     * will produce another set that is equal to this one.
+     * A frozen set will not be modified.
+     * @param result the string to receive the rules.  Previous
+     * contents will be deleted.
+     * @param escapeUnprintable if true then convert unprintable
+     * character to their hex escape representations, \\uxxxx or
+     * \\Uxxxxxxxx.  Unprintable characters are those other than
+     * U+000A, U+0020..U+007E.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeString& toPattern(UnicodeString& result,
+                                     UBool escapeUnprintable = false) const;
+
+    /**
+     * Modifies this set to contain those code points which have the given value
+     * for the given binary or enumerated property, as returned by
+     * u_getIntPropertyValue.  Prior contents of this set are lost.
+     * A frozen set will not be modified.
+     *
+     * @param prop a property in the range UCHAR_BIN_START..UCHAR_BIN_LIMIT-1
+     * or UCHAR_INT_START..UCHAR_INT_LIMIT-1
+     * or UCHAR_MASK_START..UCHAR_MASK_LIMIT-1.
+     *
+     * @param value a value in the range u_getIntPropertyMinValue(prop)..
+     * u_getIntPropertyMaxValue(prop), with one exception.  If prop is
+     * UCHAR_GENERAL_CATEGORY_MASK, then value should not be a UCharCategory, but
+     * rather a mask value produced by U_GET_GC_MASK().  This allows grouped
+     * categories such as [:L:] to be represented.
+     *
+     * @param ec error code input/output parameter
+     *
+     * @return a reference to this set
+     *
+     * @stable ICU 2.4
+     */
+    UnicodeSet& applyIntPropertyValue(UProperty prop,
+                                      int32_t value,
+                                      UErrorCode& ec);
+
+    /**
+     * Modifies this set to contain those code points which have the
+     * given value for the given property.  Prior contents of this
+     * set are lost.
+     * A frozen set will not be modified.
+     *
+     * @param prop a property alias, either short or long.  The name is matched
+     * loosely.  See PropertyAliases.txt for names and a description of loose
+     * matching.  If the value string is empty, then this string is interpreted
+     * as either a General_Category value alias, a Script value alias, a binary
+     * property alias, or a special ID.  Special IDs are matched loosely and
+     * correspond to the following sets:
+     *
+     * "ANY" = [\\u0000-\\U0010FFFF],
+     * "ASCII" = [\\u0000-\\u007F],
+     * "Assigned" = [:^Cn:].
+     *
+     * @param value a value alias, either short or long.  The name is matched
+     * loosely.  See PropertyValueAliases.txt for names and a description of
+     * loose matching.  In addition to aliases listed, numeric values and
+     * canonical combining classes may be expressed numerically, e.g., ("nv",
+     * "0.5") or ("ccc", "220").  The value string may also be empty.
+     *
+     * @param ec error code input/output parameter
+     *
+     * @return a reference to this set
+     *
+     * @stable ICU 2.4
+     */
+    UnicodeSet& applyPropertyAlias(const UnicodeString& prop,
+                                   const UnicodeString& value,
+                                   UErrorCode& ec);
+
+    /**
+     * Returns the number of elements in this set (its cardinality).
+     * Note than the elements of a set may include both individual
+     * codepoints and strings.
+     *
+     * @return the number of elements in this set (its cardinality).
+     * @stable ICU 2.0
+     */
+    virtual int32_t size(void) const;
+
+    /**
+     * Returns <tt>true</tt> if this set contains no elements.
+     *
+     * @return <tt>true</tt> if this set contains no elements.
+     * @stable ICU 2.0
+     */
+    virtual UBool isEmpty(void) const;
+
+    /**
+     * Returns true if this set contains the given character.
+     * This function works faster with a frozen set.
+     * @param c character to be checked for containment
+     * @return true if the test condition is met
+     * @stable ICU 2.0
+     */
+    virtual UBool contains(UChar32 c) const;
+
+    /**
+     * Returns true if this set contains every character
+     * of the given range.
+     * @param start first character, inclusive, of the range
+     * @param end last character, inclusive, of the range
+     * @return true if the test condition is met
+     * @stable ICU 2.0
+     */
+    virtual UBool contains(UChar32 start, UChar32 end) const;
+
+    /**
+     * Returns <tt>true</tt> if this set contains the given
+     * multicharacter string.
+     * @param s string to be checked for containment
+     * @return <tt>true</tt> if this set contains the specified string
+     * @stable ICU 2.4
+     */
+    UBool contains(const UnicodeString& s) const;
+
+    /**
+     * Returns true if this set contains all the characters and strings
+     * of the given set.
+     * @param c set to be checked for containment
+     * @return true if the test condition is met
+     * @stable ICU 2.4
+     */
+    virtual UBool containsAll(const UnicodeSet& c) const;
+
+    /**
+     * Returns true if this set contains all the characters
+     * of the given string.
+     * @param s string containing characters to be checked for containment
+     * @return true if the test condition is met
+     * @stable ICU 2.4
+     */
+    UBool containsAll(const UnicodeString& s) const;
+
+    /**
+     * Returns true if this set contains none of the characters
+     * of the given range.
+     * @param start first character, inclusive, of the range
+     * @param end last character, inclusive, of the range
+     * @return true if the test condition is met
+     * @stable ICU 2.4
+     */
+    UBool containsNone(UChar32 start, UChar32 end) const;
+
+    /**
+     * Returns true if this set contains none of the characters and strings
+     * of the given set.
+     * @param c set to be checked for containment
+     * @return true if the test condition is met
+     * @stable ICU 2.4
+     */
+    UBool containsNone(const UnicodeSet& c) const;
+
+    /**
+     * Returns true if this set contains none of the characters
+     * of the given string.
+     * @param s string containing characters to be checked for containment
+     * @return true if the test condition is met
+     * @stable ICU 2.4
+     */
+    UBool containsNone(const UnicodeString& s) const;
+
+    /**
+     * Returns true if this set contains one or more of the characters
+     * in the given range.
+     * @param start first character, inclusive, of the range
+     * @param end last character, inclusive, of the range
+     * @return true if the condition is met
+     * @stable ICU 2.4
+     */
+    inline UBool containsSome(UChar32 start, UChar32 end) const;
+
+    /**
+     * Returns true if this set contains one or more of the characters
+     * and strings of the given set.
+     * @param s The set to be checked for containment
+     * @return true if the condition is met
+     * @stable ICU 2.4
+     */
+    inline UBool containsSome(const UnicodeSet& s) const;
+
+    /**
+     * Returns true if this set contains one or more of the characters
+     * of the given string.
+     * @param s string containing characters to be checked for containment
+     * @return true if the condition is met
+     * @stable ICU 2.4
+     */
+    inline UBool containsSome(const UnicodeString& s) const;
+
+    /**
+     * Returns the length of the initial substring of the input string which
+     * consists only of characters and strings that are contained in this set
+     * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+     * or only of characters and strings that are not contained
+     * in this set (USET_SPAN_NOT_CONTAINED).
+     * See USetSpanCondition for details.
+     * Similar to the strspn() C library function.
+     * Unpaired surrogates are treated according to contains() of their surrogate code points.
+     * This function works faster with a frozen set and with a non-negative string length argument.
+     * @param s start of the string
+     * @param length of the string; can be -1 for NUL-terminated
+     * @param spanCondition specifies the containment condition
+     * @return the length of the initial substring according to the spanCondition;
+     *         0 if the start of the string does not fit the spanCondition
+     * @stable ICU 3.8
+     * @see USetSpanCondition
+     */
+    int32_t span(const char16_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    /**
+     * Returns the end of the substring of the input string according to the USetSpanCondition.
+     * Same as <code>start+span(s.getBuffer()+start, s.length()-start, spanCondition)</code>
+     * after pinning start to 0<=start<=s.length().
+     * @param s the string
+     * @param start the start index in the string for the span operation
+     * @param spanCondition specifies the containment condition
+     * @return the exclusive end of the substring according to the spanCondition;
+     *         the substring s.tempSubStringBetween(start, end) fulfills the spanCondition
+     * @stable ICU 4.4
+     * @see USetSpanCondition
+     */
+    inline int32_t span(const UnicodeString &s, int32_t start, USetSpanCondition spanCondition) const;
+
+    /**
+     * Returns the start of the trailing substring of the input string which
+     * consists only of characters and strings that are contained in this set
+     * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+     * or only of characters and strings that are not contained
+     * in this set (USET_SPAN_NOT_CONTAINED).
+     * See USetSpanCondition for details.
+     * Unpaired surrogates are treated according to contains() of their surrogate code points.
+     * This function works faster with a frozen set and with a non-negative string length argument.
+     * @param s start of the string
+     * @param length of the string; can be -1 for NUL-terminated
+     * @param spanCondition specifies the containment condition
+     * @return the start of the trailing substring according to the spanCondition;
+     *         the string length if the end of the string does not fit the spanCondition
+     * @stable ICU 3.8
+     * @see USetSpanCondition
+     */
+    int32_t spanBack(const char16_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    /**
+     * Returns the start of the substring of the input string according to the USetSpanCondition.
+     * Same as <code>spanBack(s.getBuffer(), limit, spanCondition)</code>
+     * after pinning limit to 0<=end<=s.length().
+     * @param s the string
+     * @param limit the exclusive-end index in the string for the span operation
+     *              (use s.length() or INT32_MAX for spanning back from the end of the string)
+     * @param spanCondition specifies the containment condition
+     * @return the start of the substring according to the spanCondition;
+     *         the substring s.tempSubStringBetween(start, limit) fulfills the spanCondition
+     * @stable ICU 4.4
+     * @see USetSpanCondition
+     */
+    inline int32_t spanBack(const UnicodeString &s, int32_t limit, USetSpanCondition spanCondition) const;
+
+    /**
+     * Returns the length of the initial substring of the input string which
+     * consists only of characters and strings that are contained in this set
+     * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+     * or only of characters and strings that are not contained
+     * in this set (USET_SPAN_NOT_CONTAINED).
+     * See USetSpanCondition for details.
+     * Similar to the strspn() C library function.
+     * Malformed byte sequences are treated according to contains(0xfffd).
+     * This function works faster with a frozen set and with a non-negative string length argument.
+     * @param s start of the string (UTF-8)
+     * @param length of the string; can be -1 for NUL-terminated
+     * @param spanCondition specifies the containment condition
+     * @return the length of the initial substring according to the spanCondition;
+     *         0 if the start of the string does not fit the spanCondition
+     * @stable ICU 3.8
+     * @see USetSpanCondition
+     */
+    int32_t spanUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    /**
+     * Returns the start of the trailing substring of the input string which
+     * consists only of characters and strings that are contained in this set
+     * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+     * or only of characters and strings that are not contained
+     * in this set (USET_SPAN_NOT_CONTAINED).
+     * See USetSpanCondition for details.
+     * Malformed byte sequences are treated according to contains(0xfffd).
+     * This function works faster with a frozen set and with a non-negative string length argument.
+     * @param s start of the string (UTF-8)
+     * @param length of the string; can be -1 for NUL-terminated
+     * @param spanCondition specifies the containment condition
+     * @return the start of the trailing substring according to the spanCondition;
+     *         the string length if the end of the string does not fit the spanCondition
+     * @stable ICU 3.8
+     * @see USetSpanCondition
+     */
+    int32_t spanBackUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    /**
+     * Implement UnicodeMatcher::matches()
+     * @stable ICU 2.4
+     */
+    virtual UMatchDegree matches(const Replaceable& text,
+                         int32_t& offset,
+                         int32_t limit,
+                         UBool incremental);
+
+private:
+    /**
+     * Returns the longest match for s in text at the given position.
+     * If limit > start then match forward from start+1 to limit
+     * matching all characters except s.charAt(0).  If limit < start,
+     * go backward starting from start-1 matching all characters
+     * except s.charAt(s.length()-1).  This method assumes that the
+     * first character, text.charAt(start), matches s, so it does not
+     * check it.
+     * @param text the text to match
+     * @param start the first character to match.  In the forward
+     * direction, text.charAt(start) is matched against s.charAt(0).
+     * In the reverse direction, it is matched against
+     * s.charAt(s.length()-1).
+     * @param limit the limit offset for matching, either last+1 in
+     * the forward direction, or last-1 in the reverse direction,
+     * where last is the index of the last character to match.
+     * @param s
+     * @return If part of s matches up to the limit, return |limit -
+     * start|.  If all of s matches before reaching the limit, return
+     * s.length().  If there is a mismatch between s and text, return
+     * 0
+     */
+    static int32_t matchRest(const Replaceable& text,
+                             int32_t start, int32_t limit,
+                             const UnicodeString& s);
+
+    /**
+     * Returns the smallest value i such that c < list[i].  Caller
+     * must ensure that c is a legal value or this method will enter
+     * an infinite loop.  This method performs a binary search.
+     * @param c a character in the range MIN_VALUE..MAX_VALUE
+     * inclusive
+     * @return the smallest integer i in the range 0..len-1,
+     * inclusive, such that c < list[i]
+     */
+    int32_t findCodePoint(UChar32 c) const;
+
+public:
+
+    /**
+     * Implementation of UnicodeMatcher API.  Union the set of all
+     * characters that may be matched by this object into the given
+     * set.
+     * @param toUnionTo the set into which to union the source characters
+     * @stable ICU 2.4
+     */
+    virtual void addMatchSetTo(UnicodeSet& toUnionTo) const;
+
+    /**
+     * Returns the index of the given character within this set, where
+     * the set is ordered by ascending code point.  If the character
+     * is not in this set, return -1.  The inverse of this method is
+     * <code>charAt()</code>.
+     * @return an index from 0..size()-1, or -1
+     * @stable ICU 2.4
+     */
+    int32_t indexOf(UChar32 c) const;
+
+    /**
+     * Returns the character at the given index within this set, where
+     * the set is ordered by ascending code point.  If the index is
+     * out of range, return (UChar32)-1.  The inverse of this method is
+     * <code>indexOf()</code>.
+     * @param index an index from 0..size()-1
+     * @return the character at the given index, or (UChar32)-1.
+     * @stable ICU 2.4
+     */
+    UChar32 charAt(int32_t index) const;
+
+    /**
+     * Adds the specified range to this set if it is not already
+     * present.  If this set already contains the specified range,
+     * the call leaves this set unchanged.  If <code>end > start</code>
+     * then an empty range is added, leaving the set unchanged.
+     * This is equivalent to a boolean logic OR, or a set UNION.
+     * A frozen set will not be modified.
+     *
+     * @param start first character, inclusive, of range to be added
+     * to this set.
+     * @param end last character, inclusive, of range to be added
+     * to this set.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& add(UChar32 start, UChar32 end);
+
+    /**
+     * Adds the specified character to this set if it is not already
+     * present.  If this set already contains the specified character,
+     * the call leaves this set unchanged.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    UnicodeSet& add(UChar32 c);
+
+    /**
+     * Adds the specified multicharacter to this set if it is not already
+     * present.  If this set already contains the multicharacter,
+     * the call leaves this set unchanged.
+     * Thus "ch" => {"ch"}
+     * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& add(const UnicodeString& s);
+
+ private:
+    /**
+     * @return a code point IF the string consists of a single one.
+     * otherwise returns -1.
+     * @param s string to test
+     */
+    static int32_t getSingleCP(const UnicodeString& s);
+
+    void _add(const UnicodeString& s);
+
+ public:
+    /**
+     * Adds each of the characters in this string to the set. Thus "ch" => {"c", "h"}
+     * If this set already any particular character, it has no effect on that character.
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& addAll(const UnicodeString& s);
+
+    /**
+     * Retains EACH of the characters in this string. Note: "ch" == {"c", "h"}
+     * If this set already any particular character, it has no effect on that character.
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& retainAll(const UnicodeString& s);
+
+    /**
+     * Complement EACH of the characters in this string. Note: "ch" == {"c", "h"}
+     * If this set already any particular character, it has no effect on that character.
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& complementAll(const UnicodeString& s);
+
+    /**
+     * Remove EACH of the characters in this string. Note: "ch" == {"c", "h"}
+     * If this set already any particular character, it has no effect on that character.
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& removeAll(const UnicodeString& s);
+
+    /**
+     * Makes a set from a multicharacter string. Thus "ch" => {"ch"}
+     * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+     * @param s the source string
+     * @return a newly created set containing the given string.
+     * The caller owns the return object and is responsible for deleting it.
+     * @stable ICU 2.4
+     */
+    static UnicodeSet* U_EXPORT2 createFrom(const UnicodeString& s);
+
+
+    /**
+     * Makes a set from each of the characters in the string. Thus "ch" => {"c", "h"}
+     * @param s the source string
+     * @return a newly created set containing the given characters
+     * The caller owns the return object and is responsible for deleting it.
+     * @stable ICU 2.4
+     */
+    static UnicodeSet* U_EXPORT2 createFromAll(const UnicodeString& s);
+
+    /**
+     * Retain only the elements in this set that are contained in the
+     * specified range.  If <code>end > start</code> then an empty range is
+     * retained, leaving the set empty.  This is equivalent to
+     * a boolean logic AND, or a set INTERSECTION.
+     * A frozen set will not be modified.
+     *
+     * @param start first character, inclusive, of range to be retained
+     * to this set.
+     * @param end last character, inclusive, of range to be retained
+     * to this set.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& retain(UChar32 start, UChar32 end);
+
+
+    /**
+     * Retain the specified character from this set if it is present.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    UnicodeSet& retain(UChar32 c);
+
+    /**
+     * Removes the specified range from this set if it is present.
+     * The set will not contain the specified range once the call
+     * returns.  If <code>end > start</code> then an empty range is
+     * removed, leaving the set unchanged.
+     * A frozen set will not be modified.
+     *
+     * @param start first character, inclusive, of range to be removed
+     * from this set.
+     * @param end last character, inclusive, of range to be removed
+     * from this set.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& remove(UChar32 start, UChar32 end);
+
+    /**
+     * Removes the specified character from this set if it is present.
+     * The set will not contain the specified range once the call
+     * returns.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    UnicodeSet& remove(UChar32 c);
+
+    /**
+     * Removes the specified string from this set if it is present.
+     * The set will not contain the specified character once the call
+     * returns.
+     * A frozen set will not be modified.
+     * @param s the source string
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& remove(const UnicodeString& s);
+
+    /**
+     * Inverts this set.  This operation modifies this set so that
+     * its value is its complement.  This is equivalent to
+     * <code>complement(MIN_VALUE, MAX_VALUE)</code>.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& complement(void);
+
+    /**
+     * Complements the specified range in this set.  Any character in
+     * the range will be removed if it is in this set, or will be
+     * added if it is not in this set.  If <code>end > start</code>
+     * then an empty range is complemented, leaving the set unchanged.
+     * This is equivalent to a boolean logic XOR.
+     * A frozen set will not be modified.
+     *
+     * @param start first character, inclusive, of range to be removed
+     * from this set.
+     * @param end last character, inclusive, of range to be removed
+     * from this set.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& complement(UChar32 start, UChar32 end);
+
+    /**
+     * Complements the specified character in this set.  The character
+     * will be removed if it is in this set, or will be added if it is
+     * not in this set.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    UnicodeSet& complement(UChar32 c);
+
+    /**
+     * Complement the specified string in this set.
+     * The set will not contain the specified string once the call
+     * returns.
+     * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+     * A frozen set will not be modified.
+     * @param s the string to complement
+     * @return this object, for chaining
+     * @stable ICU 2.4
+     */
+    UnicodeSet& complement(const UnicodeString& s);
+
+    /**
+     * Adds all of the elements in the specified set to this set if
+     * they're not already present.  This operation effectively
+     * modifies this set so that its value is the <i>union</i> of the two
+     * sets.  The behavior of this operation is unspecified if the specified
+     * collection is modified while the operation is in progress.
+     * A frozen set will not be modified.
+     *
+     * @param c set whose elements are to be added to this set.
+     * @see #add(UChar32, UChar32)
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& addAll(const UnicodeSet& c);
+
+    /**
+     * Retains only the elements in this set that are contained in the
+     * specified set.  In other words, removes from this set all of
+     * its elements that are not contained in the specified set.  This
+     * operation effectively modifies this set so that its value is
+     * the <i>intersection</i> of the two sets.
+     * A frozen set will not be modified.
+     *
+     * @param c set that defines which elements this set will retain.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& retainAll(const UnicodeSet& c);
+
+    /**
+     * Removes from this set all of its elements that are contained in the
+     * specified set.  This operation effectively modifies this
+     * set so that its value is the <i>asymmetric set difference</i> of
+     * the two sets.
+     * A frozen set will not be modified.
+     *
+     * @param c set that defines which elements will be removed from
+     *          this set.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& removeAll(const UnicodeSet& c);
+
+    /**
+     * Complements in this set all elements contained in the specified
+     * set.  Any character in the other set will be removed if it is
+     * in this set, or will be added if it is not in this set.
+     * A frozen set will not be modified.
+     *
+     * @param c set that defines which elements will be xor'ed from
+     *          this set.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeSet& complementAll(const UnicodeSet& c);
+
+    /**
+     * Removes all of the elements from this set.  This set will be
+     * empty after this call returns.
+     * A frozen set will not be modified.
+     * @stable ICU 2.0
+     */
+    virtual UnicodeSet& clear(void);
+
+    /**
+     * Close this set over the given attribute.  For the attribute
+     * USET_CASE, the result is to modify this set so that:
+     *
+     * 1. For each character or string 'a' in this set, all strings or
+     * characters 'b' such that foldCase(a) == foldCase(b) are added
+     * to this set.
+     *
+     * 2. For each string 'e' in the resulting set, if e !=
+     * foldCase(e), 'e' will be removed.
+     *
+     * Example: [aq\\u00DF{Bc}{bC}{Fi}] => [aAqQ\\u00DF\\uFB01{ss}{bc}{fi}]
+     *
+     * (Here foldCase(x) refers to the operation u_strFoldCase, and a
+     * == b denotes that the contents are the same, not pointer
+     * comparison.)
+     *
+     * A frozen set will not be modified.
+     *
+     * @param attribute bitmask for attributes to close over.
+     * Currently only the USET_CASE bit is supported.  Any undefined bits
+     * are ignored.
+     * @return a reference to this set.
+     * @stable ICU 4.2
+     */
+    UnicodeSet& closeOver(int32_t attribute);
+
+    /**
+     * Remove all strings from this set.
+     *
+     * @return a reference to this set.
+     * @stable ICU 4.2
+     */
+    virtual UnicodeSet &removeAllStrings();
+
+    /**
+     * Iteration method that returns the number of ranges contained in
+     * this set.
+     * @see #getRangeStart
+     * @see #getRangeEnd
+     * @stable ICU 2.4
+     */
+    virtual int32_t getRangeCount(void) const;
+
+    /**
+     * Iteration method that returns the first character in the
+     * specified range of this set.
+     * @see #getRangeCount
+     * @see #getRangeEnd
+     * @stable ICU 2.4
+     */
+    virtual UChar32 getRangeStart(int32_t index) const;
+
+    /**
+     * Iteration method that returns the last character in the
+     * specified range of this set.
+     * @see #getRangeStart
+     * @see #getRangeEnd
+     * @stable ICU 2.4
+     */
+    virtual UChar32 getRangeEnd(int32_t index) const;
+
+    /**
+     * Serializes this set into an array of 16-bit integers.  Serialization
+     * (currently) only records the characters in the set; multicharacter
+     * strings are ignored.
+     *
+     * The array has following format (each line is one 16-bit
+     * integer):
+     *
+     *  length     = (n+2*m) | (m!=0?0x8000:0)
+     *  bmpLength  = n; present if m!=0
+     *  bmp[0]
+     *  bmp[1]
+     *  ...
+     *  bmp[n-1]
+     *  supp-high[0]
+     *  supp-low[0]
+     *  supp-high[1]
+     *  supp-low[1]
+     *  ...
+     *  supp-high[m-1]
+     *  supp-low[m-1]
+     *
+     * The array starts with a header.  After the header are n bmp
+     * code points, then m supplementary code points.  Either n or m
+     * or both may be zero.  n+2*m is always <= 0x7FFF.
+     *
+     * If there are no supplementary characters (if m==0) then the
+     * header is one 16-bit integer, 'length', with value n.
+     *
+     * If there are supplementary characters (if m!=0) then the header
+     * is two 16-bit integers.  The first, 'length', has value
+     * (n+2*m)|0x8000.  The second, 'bmpLength', has value n.
+     *
+     * After the header the code points are stored in ascending order.
+     * Supplementary code points are stored as most significant 16
+     * bits followed by least significant 16 bits.
+     *
+     * @param dest pointer to buffer of destCapacity 16-bit integers.
+     * May be NULL only if destCapacity is zero.
+     * @param destCapacity size of dest, or zero.  Must not be negative.
+     * @param ec error code.  Will be set to U_INDEX_OUTOFBOUNDS_ERROR
+     * if n+2*m > 0x7FFF.  Will be set to U_BUFFER_OVERFLOW_ERROR if
+     * n+2*m+(m!=0?2:1) > destCapacity.
+     * @return the total length of the serialized format, including
+     * the header, that is, n+2*m+(m!=0?2:1), or 0 on error other
+     * than U_BUFFER_OVERFLOW_ERROR.
+     * @stable ICU 2.4
+     */
+    int32_t serialize(uint16_t *dest, int32_t destCapacity, UErrorCode& ec) const;
+
+    /**
+     * Reallocate this objects internal structures to take up the least
+     * possible space, without changing this object's value.
+     * A frozen set will not be modified.
+     * @stable ICU 2.4
+     */
+    virtual UnicodeSet& compact();
+
+    /**
+     * Return the class ID for this class.  This is useful only for
+     * comparing to a return value from getDynamicClassID().  For example:
+     * <pre>
+     * .      Base* polymorphic_pointer = createPolymorphicObject();
+     * .      if (polymorphic_pointer->getDynamicClassID() ==
+     * .          Derived::getStaticClassID()) ...
+     * </pre>
+     * @return          The class ID for all objects of this class.
+     * @stable ICU 2.0
+     */
+    static UClassID U_EXPORT2 getStaticClassID(void);
+
+    /**
+     * Implement UnicodeFunctor API.
+     *
+     * @return The class ID for this object. All objects of a given
+     * class have the same class ID.  Objects of other classes have
+     * different class IDs.
+     * @stable ICU 2.4
+     */
+    virtual UClassID getDynamicClassID(void) const;
+
+private:
+
+    // Private API for the USet API
+
+    friend class USetAccess;
+
+    const UnicodeString* getString(int32_t index) const;
+
+    //----------------------------------------------------------------
+    // RuleBasedTransliterator support
+    //----------------------------------------------------------------
+
+private:
+
+    /**
+     * Returns <tt>true</tt> if this set contains any character whose low byte
+     * is the given value.  This is used by <tt>RuleBasedTransliterator</tt> for
+     * indexing.
+     */
+    virtual UBool matchesIndexValue(uint8_t v) const;
+
+private:
+    friend class RBBIRuleScanner;
+
+    //----------------------------------------------------------------
+    // Implementation: Clone as thawed (see ICU4J Freezable)
+    //----------------------------------------------------------------
+
+    UnicodeSet(const UnicodeSet& o, UBool /* asThawed */);
+    UnicodeSet& copyFrom(const UnicodeSet& o, UBool asThawed);
+
+    //----------------------------------------------------------------
+    // Implementation: Pattern parsing
+    //----------------------------------------------------------------
+
+    void applyPatternIgnoreSpace(const UnicodeString& pattern,
+                                 ParsePosition& pos,
+                                 const SymbolTable* symbols,
+                                 UErrorCode& status);
+
+    void applyPattern(RuleCharacterIterator& chars,
+                      const SymbolTable* symbols,
+                      UnicodeString& rebuiltPat,
+                      uint32_t options,
+                      UnicodeSet& (UnicodeSet::*caseClosure)(int32_t attribute),
+                      int32_t depth,
+                      UErrorCode& ec);
+
+    //----------------------------------------------------------------
+    // Implementation: Utility methods
+    //----------------------------------------------------------------
+
+    static int32_t nextCapacity(int32_t minCapacity);
+
+    bool ensureCapacity(int32_t newLen);
+
+    bool ensureBufferCapacity(int32_t newLen);
+
+    void swapBuffers(void);
+
+    UBool allocateStrings(UErrorCode &status);
+    UBool hasStrings() const;
+    int32_t stringsSize() const;
+    UBool stringsContains(const UnicodeString &s) const;
+
+    UnicodeString& _toPattern(UnicodeString& result,
+                              UBool escapeUnprintable) const;
+
+    UnicodeString& _generatePattern(UnicodeString& result,
+                                    UBool escapeUnprintable) const;
+
+    static void _appendToPat(UnicodeString& buf, const UnicodeString& s, UBool escapeUnprintable);
+
+    static void _appendToPat(UnicodeString& buf, UChar32 c, UBool escapeUnprintable);
+
+    //----------------------------------------------------------------
+    // Implementation: Fundamental operators
+    //----------------------------------------------------------------
+
+    void exclusiveOr(const UChar32* other, int32_t otherLen, int8_t polarity);
+
+    void add(const UChar32* other, int32_t otherLen, int8_t polarity);
+
+    void retain(const UChar32* other, int32_t otherLen, int8_t polarity);
+
+    /**
+     * Return true if the given position, in the given pattern, appears
+     * to be the start of a property set pattern [:foo:], \\p{foo}, or
+     * \\P{foo}, or \\N{name}.
+     */
+    static UBool resemblesPropertyPattern(const UnicodeString& pattern,
+                                          int32_t pos);
+
+    static UBool resemblesPropertyPattern(RuleCharacterIterator& chars,
+                                          int32_t iterOpts);
+
+    /**
+     * Parse the given property pattern at the given parse position
+     * and set this UnicodeSet to the result.
+     *
+     * The original design document is out of date, but still useful.
+     * Ignore the property and value names:
+     * http://source.icu-project.org/repos/icu/icuhtml/trunk/design/unicodeset_properties.html
+     *
+     * Recognized syntax:
+     *
+     * [:foo:] [:^foo:] - white space not allowed within "[:" or ":]"
+     * \\p{foo} \\P{foo}  - white space not allowed within "\\p" or "\\P"
+     * \\N{name}         - white space not allowed within "\\N"
+     *
+     * Other than the above restrictions, Unicode Pattern_White_Space characters are ignored.
+     * Case is ignored except in "\\p" and "\\P" and "\\N".  In 'name' leading
+     * and trailing space is deleted, and internal runs of whitespace
+     * are collapsed to a single space.
+     *
+     * We support binary properties, enumerated properties, and the
+     * following non-enumerated properties:
+     *
+     *  Numeric_Value
+     *  Name
+     *  Unicode_1_Name
+     *
+     * @param pattern the pattern string
+     * @param ppos on entry, the position at which to begin parsing.
+     * This should be one of the locations marked '^':
+     *
+     *   [:blah:]     \\p{blah}     \\P{blah}     \\N{name}
+     *   ^       %    ^       %    ^       %    ^       %
+     *
+     * On return, the position after the last character parsed, that is,
+     * the locations marked '%'.  If the parse fails, ppos is returned
+     * unchanged.
+     * @param ec status
+     * @return a reference to this.
+     */
+    UnicodeSet& applyPropertyPattern(const UnicodeString& pattern,
+                                     ParsePosition& ppos,
+                                     UErrorCode &ec);
+
+    void applyPropertyPattern(RuleCharacterIterator& chars,
+                              UnicodeString& rebuiltPat,
+                              UErrorCode& ec);
+
+    static const UnicodeSet* getInclusions(int32_t src, UErrorCode &status);
+
+    /**
+     * A filter that returns true if the given code point should be
+     * included in the UnicodeSet being constructed.
+     */
+    typedef UBool (*Filter)(UChar32 codePoint, void* context);
+
+    /**
+     * Given a filter, set this UnicodeSet to the code points
+     * contained by that filter.  The filter MUST be
+     * property-conformant.  That is, if it returns value v for one
+     * code point, then it must return v for all affiliated code
+     * points, as defined by the inclusions list.  See
+     * getInclusions().
+     * src is a UPropertySource value.
+     */
+    void applyFilter(Filter filter,
+                     void* context,
+                     const UnicodeSet* inclusions,
+                     UErrorCode &status);
+
+    // UCPMap is now stable ICU 63
+    void applyIntPropertyValue(const UCPMap *map,
+                               UCPMapValueFilter *filter, const void *context,
+                               UErrorCode &errorCode);
+
+    /**
+     * Set the new pattern to cache.
+     */
+    void setPattern(const UnicodeString& newPat) {
+        setPattern(newPat.getBuffer(), newPat.length());
+    }
+    void setPattern(const char16_t *newPat, int32_t newPatLen);
+    /**
+     * Release existing cached pattern.
+     */
+    void releasePattern();
+
+    friend class UnicodeSetIterator;
+};
+
+
+
+inline UBool UnicodeSet::operator!=(const UnicodeSet& o) const {
+    return !operator==(o);
+}
+
+inline UBool UnicodeSet::isFrozen() const {
+    return (UBool)(bmpSet!=NULL || stringSpan!=NULL);
+}
+
+inline UBool UnicodeSet::containsSome(UChar32 start, UChar32 end) const {
+    return !containsNone(start, end);
+}
+
+inline UBool UnicodeSet::containsSome(const UnicodeSet& s) const {
+    return !containsNone(s);
+}
+
+inline UBool UnicodeSet::containsSome(const UnicodeString& s) const {
+    return !containsNone(s);
+}
+
+inline UBool UnicodeSet::isBogus() const {
+    return (UBool)(fFlags & kIsBogus);
+}
+
+inline UnicodeSet *UnicodeSet::fromUSet(USet *uset) {
+    return reinterpret_cast<UnicodeSet *>(uset);
+}
+
+inline const UnicodeSet *UnicodeSet::fromUSet(const USet *uset) {
+    return reinterpret_cast<const UnicodeSet *>(uset);
+}
+
+inline USet *UnicodeSet::toUSet() {
+    return reinterpret_cast<USet *>(this);
+}
+
+inline const USet *UnicodeSet::toUSet() const {
+    return reinterpret_cast<const USet *>(this);
+}
+
+inline int32_t UnicodeSet::span(const UnicodeString &s, int32_t start, USetSpanCondition spanCondition) const {
+    int32_t sLength=s.length();
+    if(start<0) {
+        start=0;
+    } else if(start>sLength) {
+        start=sLength;
+    }
+    return start+span(s.getBuffer()+start, sLength-start, spanCondition);
+}
+
+inline int32_t UnicodeSet::spanBack(const UnicodeString &s, int32_t limit, USetSpanCondition spanCondition) const {
+    int32_t sLength=s.length();
+    if(limit<0) {
+        limit=0;
+    } else if(limit>sLength) {
+        limit=sLength;
+    }
+    return spanBack(s.getBuffer(), limit, spanCondition);
+}
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unistr.h b/thirdparty/icu4c/common/unicode/unistr.h
new file mode 100644
index 0000000000..456389f265
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unistr.h
@@ -0,0 +1,4757 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1998-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File unistr.h
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   09/25/98    stephen     Creation.
+*   11/11/98    stephen     Changed per 11/9 code review.
+*   04/20/99    stephen     Overhauled per 4/16 code review.
+*   11/18/99    aliu        Made to inherit from Replaceable.  Added method
+*                           handleReplaceBetween(); other methods unchanged.
+*   06/25/01    grhoten     Remove dependency on iostream.
+******************************************************************************
+*/
+
+#ifndef UNISTR_H
+#define UNISTR_H
+
+/**
+ * \file
+ * \brief C++ API: Unicode String
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include <cstddef>
+#include "unicode/char16ptr.h"
+#include "unicode/rep.h"
+#include "unicode/std_string.h"
+#include "unicode/stringpiece.h"
+#include "unicode/bytestream.h"
+
+struct UConverter;          // unicode/ucnv.h
+
+#ifndef USTRING_H
+/**
+ * \ingroup ustring_ustrlen
+ */
+U_CAPI int32_t U_EXPORT2
+u_strlen(const UChar *s);
+#endif
+
+U_NAMESPACE_BEGIN
+
+#if !UCONFIG_NO_BREAK_ITERATION
+class BreakIterator;        // unicode/brkiter.h
+#endif
+class Edits;
+
+U_NAMESPACE_END
+
+// Not #ifndef U_HIDE_INTERNAL_API because UnicodeString needs the UStringCaseMapper.
+/**
+ * Internal string case mapping function type.
+ * All error checking must be done.
+ * src and dest must not overlap.
+ * @internal
+ */
+typedef int32_t U_CALLCONV
+UStringCaseMapper(int32_t caseLocale, uint32_t options,
+#if !UCONFIG_NO_BREAK_ITERATION
+                  icu::BreakIterator *iter,
+#endif
+                  char16_t *dest, int32_t destCapacity,
+                  const char16_t *src, int32_t srcLength,
+                  icu::Edits *edits,
+                  UErrorCode &errorCode);
+
+U_NAMESPACE_BEGIN
+
+class Locale;               // unicode/locid.h
+class StringCharacterIterator;
+class UnicodeStringAppendable;  // unicode/appendable.h
+
+/* The <iostream> include has been moved to unicode/ustream.h */
+
+/**
+ * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
+ * which constructs a Unicode string from an invariant-character char * string.
+ * About invariant characters see utypes.h.
+ * This constructor has no runtime dependency on conversion code and is
+ * therefore recommended over ones taking a charset name string
+ * (where the empty string "" indicates invariant-character conversion).
+ *
+ * @stable ICU 3.2
+ */
+#define US_INV icu::UnicodeString::kInvariant
+
+/**
+ * Unicode String literals in C++.
+ *
+ * Note: these macros are not recommended for new code.
+ * Prior to the availability of C++11 and u"unicode string literals",
+ * these macros were provided for portability and efficiency when
+ * initializing UnicodeStrings from literals.
+ *
+ * They work only for strings that contain "invariant characters", i.e.,
+ * only latin letters, digits, and some punctuation.
+ * See utypes.h for details.
+ *
+ * The string parameter must be a C string literal.
+ * The length of the string, not including the terminating
+ * `NUL`, must be specified as a constant.
+ * @stable ICU 2.0
+ */
+#if !U_CHAR16_IS_TYPEDEF
+# define UNICODE_STRING(cs, _length) icu::UnicodeString(true, u ## cs, _length)
+#else
+# define UNICODE_STRING(cs, _length) icu::UnicodeString(true, (const char16_t*)u ## cs, _length)
+#endif
+
+/**
+ * Unicode String literals in C++.
+ * Dependent on the platform properties, different UnicodeString
+ * constructors should be used to create a UnicodeString object from
+ * a string literal.
+ * The macros are defined for improved performance.
+ * They work only for strings that contain "invariant characters", i.e.,
+ * only latin letters, digits, and some punctuation.
+ * See utypes.h for details.
+ *
+ * The string parameter must be a C string literal.
+ * @stable ICU 2.0
+ */
+#define UNICODE_STRING_SIMPLE(cs) UNICODE_STRING(cs, -1)
+
+/**
+ * \def UNISTR_FROM_CHAR_EXPLICIT
+ * This can be defined to be empty or "explicit".
+ * If explicit, then the UnicodeString(char16_t) and UnicodeString(UChar32)
+ * constructors are marked as explicit, preventing their inadvertent use.
+ * @stable ICU 49
+ */
+#ifndef UNISTR_FROM_CHAR_EXPLICIT
+# if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
+    // Auto-"explicit" in ICU library code.
+#   define UNISTR_FROM_CHAR_EXPLICIT explicit
+# else
+    // Empty by default for source code compatibility.
+#   define UNISTR_FROM_CHAR_EXPLICIT
+# endif
+#endif
+
+/**
+ * \def UNISTR_FROM_STRING_EXPLICIT
+ * This can be defined to be empty or "explicit".
+ * If explicit, then the UnicodeString(const char *) and UnicodeString(const char16_t *)
+ * constructors are marked as explicit, preventing their inadvertent use.
+ *
+ * In particular, this helps prevent accidentally depending on ICU conversion code
+ * by passing a string literal into an API with a const UnicodeString & parameter.
+ * @stable ICU 49
+ */
+#ifndef UNISTR_FROM_STRING_EXPLICIT
+# if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
+    // Auto-"explicit" in ICU library code.
+#   define UNISTR_FROM_STRING_EXPLICIT explicit
+# else
+    // Empty by default for source code compatibility.
+#   define UNISTR_FROM_STRING_EXPLICIT
+# endif
+#endif
+
+/**
+ * \def UNISTR_OBJECT_SIZE
+ * Desired sizeof(UnicodeString) in bytes.
+ * It should be a multiple of sizeof(pointer) to avoid unusable space for padding.
+ * The object size may want to be a multiple of 16 bytes,
+ * which is a common granularity for heap allocation.
+ *
+ * Any space inside the object beyond sizeof(vtable pointer) + 2
+ * is available for storing short strings inside the object.
+ * The bigger the object, the longer a string that can be stored inside the object,
+ * without additional heap allocation.
+ *
+ * Depending on a platform's pointer size, pointer alignment requirements,
+ * and struct padding, the compiler will usually round up sizeof(UnicodeString)
+ * to 4 * sizeof(pointer) (or 3 * sizeof(pointer) for P128 data models),
+ * to hold the fields for heap-allocated strings.
+ * Such a minimum size also ensures that the object is easily large enough
+ * to hold at least 2 char16_ts, for one supplementary code point (U16_MAX_LENGTH).
+ *
+ * sizeof(UnicodeString) >= 48 should work for all known platforms.
+ *
+ * For example, on a 64-bit machine where sizeof(vtable pointer) is 8,
+ * sizeof(UnicodeString) = 64 would leave space for
+ * (64 - sizeof(vtable pointer) - 2) / U_SIZEOF_UCHAR = (64 - 8 - 2) / 2 = 27
+ * char16_ts stored inside the object.
+ *
+ * The minimum object size on a 64-bit machine would be
+ * 4 * sizeof(pointer) = 4 * 8 = 32 bytes,
+ * and the internal buffer would hold up to 11 char16_ts in that case.
+ *
+ * @see U16_MAX_LENGTH
+ * @stable ICU 56
+ */
+#ifndef UNISTR_OBJECT_SIZE
+# define UNISTR_OBJECT_SIZE 64
+#endif
+
+/**
+ * UnicodeString is a string class that stores Unicode characters directly and provides
+ * similar functionality as the Java String and StringBuffer/StringBuilder classes.
+ * It is a concrete implementation of the abstract class Replaceable (for transliteration).
+ *
+ * The UnicodeString equivalent of std::string’s clear() is remove().
+ *
+ * A UnicodeString may "alias" an external array of characters
+ * (that is, point to it, rather than own the array)
+ * whose lifetime must then at least match the lifetime of the aliasing object.
+ * This aliasing may be preserved when returning a UnicodeString by value,
+ * depending on the compiler and the function implementation,
+ * via Return Value Optimization (RVO) or the move assignment operator.
+ * (However, the copy assignment operator does not preserve aliasing.)
+ * For details see the description of storage models at the end of the class API docs
+ * and in the User Guide chapter linked from there.
+ *
+ * The UnicodeString class is not suitable for subclassing.
+ *
+ * For an overview of Unicode strings in C and C++ see the
+ * [User Guide Strings chapter](https://unicode-org.github.io/icu/userguide/strings#strings-in-cc).
+ *
+ * In ICU, a Unicode string consists of 16-bit Unicode *code units*.
+ * A Unicode character may be stored with either one code unit
+ * (the most common case) or with a matched pair of special code units
+ * ("surrogates"). The data type for code units is char16_t.
+ * For single-character handling, a Unicode character code *point* is a value
+ * in the range 0..0x10ffff. ICU uses the UChar32 type for code points.
+ *
+ * Indexes and offsets into and lengths of strings always count code units, not code points.
+ * This is the same as with multi-byte char* strings in traditional string handling.
+ * Operations on partial strings typically do not test for code point boundaries.
+ * If necessary, the user needs to take care of such boundaries by testing for the code unit
+ * values or by using functions like
+ * UnicodeString::getChar32Start() and UnicodeString::getChar32Limit()
+ * (or, in C, the equivalent macros U16_SET_CP_START() and U16_SET_CP_LIMIT(), see utf.h).
+ *
+ * UnicodeString methods are more lenient with regard to input parameter values
+ * than other ICU APIs. In particular:
+ * - If indexes are out of bounds for a UnicodeString object
+ *   (< 0 or > length()) then they are "pinned" to the nearest boundary.
+ * - If the buffer passed to an insert/append/replace operation is owned by the
+ *   target object, e.g., calling str.append(str), an extra copy may take place
+ *   to ensure safety.
+ * - If primitive string pointer values (e.g., const char16_t * or char *)
+ *   for input strings are NULL, then those input string parameters are treated
+ *   as if they pointed to an empty string.
+ *   However, this is *not* the case for char * parameters for charset names
+ *   or other IDs.
+ * - Most UnicodeString methods do not take a UErrorCode parameter because
+ *   there are usually very few opportunities for failure other than a shortage
+ *   of memory, error codes in low-level C++ string methods would be inconvenient,
+ *   and the error code as the last parameter (ICU convention) would prevent
+ *   the use of default parameter values.
+ *   Instead, such methods set the UnicodeString into a "bogus" state
+ *   (see isBogus()) if an error occurs.
+ *
+ * In string comparisons, two UnicodeString objects that are both "bogus"
+ * compare equal (to be transitive and prevent endless loops in sorting),
+ * and a "bogus" string compares less than any non-"bogus" one.
+ *
+ * Const UnicodeString methods are thread-safe. Multiple threads can use
+ * const methods on the same UnicodeString object simultaneously,
+ * but non-const methods must not be called concurrently (in multiple threads)
+ * with any other (const or non-const) methods.
+ *
+ * Similarly, const UnicodeString & parameters are thread-safe.
+ * One object may be passed in as such a parameter concurrently in multiple threads.
+ * This includes the const UnicodeString & parameters for
+ * copy construction, assignment, and cloning.
+ *
+ * UnicodeString uses several storage methods.
+ * String contents can be stored inside the UnicodeString object itself,
+ * in an allocated and shared buffer, or in an outside buffer that is "aliased".
+ * Most of this is done transparently, but careful aliasing in particular provides
+ * significant performance improvements.
+ * Also, the internal buffer is accessible via special functions.
+ * For details see the
+ * [User Guide Strings chapter](https://unicode-org.github.io/icu/userguide/strings#maximizing-performance-with-the-unicodestring-storage-model).
+ *
+ * @see utf.h
+ * @see CharacterIterator
+ * @stable ICU 2.0
+ */
+class U_COMMON_API UnicodeString : public Replaceable
+{
+public:
+
+  /**
+   * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
+   * which constructs a Unicode string from an invariant-character char * string.
+   * Use the macro US_INV instead of the full qualification for this value.
+   *
+   * @see US_INV
+   * @stable ICU 3.2
+   */
+  enum EInvariant {
+    /**
+     * @see EInvariant
+     * @stable ICU 3.2
+     */
+    kInvariant
+  };
+
+  //========================================
+  // Read-only operations
+  //========================================
+
+  /* Comparison - bitwise only - for international comparison use collation */
+
+  /**
+   * Equality operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return true if `text` contains the same characters as this one,
+   * false otherwise.
+   * @stable ICU 2.0
+   */
+  inline UBool operator== (const UnicodeString& text) const;
+
+  /**
+   * Inequality operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return false if `text` contains the same characters as this one,
+   * true otherwise.
+   * @stable ICU 2.0
+   */
+  inline UBool operator!= (const UnicodeString& text) const;
+
+  /**
+   * Greater than operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return true if the characters in this are bitwise
+   * greater than the characters in `text`, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool operator> (const UnicodeString& text) const;
+
+  /**
+   * Less than operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return true if the characters in this are bitwise
+   * less than the characters in `text`, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool operator< (const UnicodeString& text) const;
+
+  /**
+   * Greater than or equal operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return true if the characters in this are bitwise
+   * greater than or equal to the characters in `text`, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool operator>= (const UnicodeString& text) const;
+
+  /**
+   * Less than or equal operator. Performs only bitwise comparison.
+   * @param text The UnicodeString to compare to this one.
+   * @return true if the characters in this are bitwise
+   * less than or equal to the characters in `text`, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool operator<= (const UnicodeString& text) const;
+
+  /**
+   * Compare the characters bitwise in this UnicodeString to
+   * the characters in `text`.
+   * @param text The UnicodeString to compare to this one.
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `text`, -1 if the characters in
+   * this are bitwise less than the characters in `text`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `text`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compare(const UnicodeString& text) const;
+
+  /**
+   * Compare the characters bitwise in the range
+   * [`start`, `start + length`) with the characters
+   * in the **entire string** `text`.
+   * (The parameters "start" and "length" are not applied to the other text "text".)
+   * @param start the offset at which the compare operation begins
+   * @param length the number of characters of text to compare.
+   * @param text the other text to be compared against this string.
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `text`, -1 if the characters in
+   * this are bitwise less than the characters in `text`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `text`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compare(int32_t start,
+         int32_t length,
+         const UnicodeString& text) const;
+
+  /**
+   * Compare the characters bitwise in the range
+   * [`start`, `start + length`) with the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`).
+   * @param start the offset at which the compare operation begins
+   * @param length the number of characters in this to compare.
+   * @param srcText the text to be compared
+   * @param srcStart the offset into `srcText` to start comparison
+   * @param srcLength the number of characters in `src` to compare
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `srcText`, -1 if the characters in
+   * this are bitwise less than the characters in `srcText`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `srcText`.
+   * @stable ICU 2.0
+   */
+   inline int8_t compare(int32_t start,
+         int32_t length,
+         const UnicodeString& srcText,
+         int32_t srcStart,
+         int32_t srcLength) const;
+
+  /**
+   * Compare the characters bitwise in this UnicodeString with the first
+   * `srcLength` characters in `srcChars`.
+   * @param srcChars The characters to compare to this UnicodeString.
+   * @param srcLength the number of characters in `srcChars` to compare
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `srcChars`, -1 if the characters in
+   * this are bitwise less than the characters in `srcChars`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `srcChars`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compare(ConstChar16Ptr srcChars,
+         int32_t srcLength) const;
+
+  /**
+   * Compare the characters bitwise in the range
+   * [`start`, `start + length`) with the first
+   * `length` characters in `srcChars`
+   * @param start the offset at which the compare operation begins
+   * @param length the number of characters to compare.
+   * @param srcChars the characters to be compared
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `srcChars`, -1 if the characters in
+   * this are bitwise less than the characters in `srcChars`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `srcChars`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compare(int32_t start,
+         int32_t length,
+         const char16_t *srcChars) const;
+
+  /**
+   * Compare the characters bitwise in the range
+   * [`start`, `start + length`) with the characters
+   * in `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`).
+   * @param start the offset at which the compare operation begins
+   * @param length the number of characters in this to compare
+   * @param srcChars the characters to be compared
+   * @param srcStart the offset into `srcChars` to start comparison
+   * @param srcLength the number of characters in `srcChars` to compare
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `srcChars`, -1 if the characters in
+   * this are bitwise less than the characters in `srcChars`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `srcChars`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compare(int32_t start,
+         int32_t length,
+         const char16_t *srcChars,
+         int32_t srcStart,
+         int32_t srcLength) const;
+
+  /**
+   * Compare the characters bitwise in the range
+   * [`start`, `limit`) with the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcLimit`).
+   * @param start the offset at which the compare operation begins
+   * @param limit the offset immediately following the compare operation
+   * @param srcText the text to be compared
+   * @param srcStart the offset into `srcText` to start comparison
+   * @param srcLimit the offset into `srcText` to limit comparison
+   * @return The result of bitwise character comparison: 0 if this
+   * contains the same characters as `srcText`, -1 if the characters in
+   * this are bitwise less than the characters in `srcText`, +1 if the
+   * characters in this are bitwise greater than the characters
+   * in `srcText`.
+   * @stable ICU 2.0
+   */
+  inline int8_t compareBetween(int32_t start,
+            int32_t limit,
+            const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLimit) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param text Another string to compare this one to.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrder(const UnicodeString& text) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrder(int32_t start,
+                                      int32_t length,
+                                      const UnicodeString& srcText) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLength The number of code units from that string to compare.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+   inline int8_t compareCodePointOrder(int32_t start,
+                                       int32_t length,
+                                       const UnicodeString& srcText,
+                                       int32_t srcStart,
+                                       int32_t srcLength) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param srcChars A pointer to another string to compare this one to.
+   * @param srcLength The number of code units from that string to compare.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrder(ConstChar16Ptr srcChars,
+                                      int32_t srcLength) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcChars A pointer to another string to compare this one to.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrder(int32_t start,
+                                      int32_t length,
+                                      const char16_t *srcChars) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcChars A pointer to another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLength The number of code units from that string to compare.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrder(int32_t start,
+                                      int32_t length,
+                                      const char16_t *srcChars,
+                                      int32_t srcStart,
+                                      int32_t srcLength) const;
+
+  /**
+   * Compare two Unicode strings in code point order.
+   * The result may be different from the results of compare(), operator<, etc.
+   * if supplementary characters are present:
+   *
+   * In UTF-16, supplementary characters (with code points U+10000 and above) are
+   * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
+   * which means that they compare as less than some other BMP characters like U+feff.
+   * This function compares Unicode strings in code point order.
+   * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param limit The offset after the last code unit from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLimit The offset after the last code unit from that string to compare.
+   * @return a negative/zero/positive integer corresponding to whether
+   * this string is less than/equal to/greater than the second one
+   * in code point order
+   * @stable ICU 2.0
+   */
+  inline int8_t compareCodePointOrderBetween(int32_t start,
+                                             int32_t limit,
+                                             const UnicodeString& srcText,
+                                             int32_t srcStart,
+                                             int32_t srcLimit) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(text.foldCase(options)).
+   *
+   * @param text Another string to compare this one to.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(const UnicodeString& text, uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(int32_t start,
+         int32_t length,
+         const UnicodeString& srcText,
+         uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLength The number of code units from that string to compare.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(int32_t start,
+         int32_t length,
+         const UnicodeString& srcText,
+         int32_t srcStart,
+         int32_t srcLength,
+         uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
+   *
+   * @param srcChars A pointer to another string to compare this one to.
+   * @param srcLength The number of code units from that string to compare.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(ConstChar16Ptr srcChars,
+         int32_t srcLength,
+         uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcChars A pointer to another string to compare this one to.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(int32_t start,
+         int32_t length,
+         const char16_t *srcChars,
+         uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param length The number of code units from this string to compare.
+   * @param srcChars A pointer to another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLength The number of code units from that string to compare.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompare(int32_t start,
+         int32_t length,
+         const char16_t *srcChars,
+         int32_t srcStart,
+         int32_t srcLength,
+         uint32_t options) const;
+
+  /**
+   * Compare two strings case-insensitively using full case folding.
+   * This is equivalent to this->foldCase(options).compareBetween(text.foldCase(options)).
+   *
+   * @param start The start offset in this string at which the compare operation begins.
+   * @param limit The offset after the last code unit from this string to compare.
+   * @param srcText Another string to compare this one to.
+   * @param srcStart The start offset in that string at which the compare operation begins.
+   * @param srcLimit The offset after the last code unit from that string to compare.
+   * @param options A bit set of options:
+   *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+   *     Comparison in code unit order with default case folding.
+   *
+   *   - U_COMPARE_CODE_POINT_ORDER
+   *     Set to choose code point order instead of code unit order
+   *     (see u_strCompare for details).
+   *
+   *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   *
+   * @return A negative, zero, or positive integer indicating the comparison result.
+   * @stable ICU 2.0
+   */
+  inline int8_t caseCompareBetween(int32_t start,
+            int32_t limit,
+            const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLimit,
+            uint32_t options) const;
+
+  /**
+   * Determine if this starts with the characters in `text`
+   * @param text The text to match.
+   * @return true if this starts with the characters in `text`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool startsWith(const UnicodeString& text) const;
+
+  /**
+   * Determine if this starts with the characters in `srcText`
+   * in the range [`srcStart`, `srcStart + srcLength`).
+   * @param srcText The text to match.
+   * @param srcStart the offset into `srcText` to start matching
+   * @param srcLength the number of characters in `srcText` to match
+   * @return true if this starts with the characters in `text`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool startsWith(const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLength) const;
+
+  /**
+   * Determine if this starts with the characters in `srcChars`
+   * @param srcChars The characters to match.
+   * @param srcLength the number of characters in `srcChars`
+   * @return true if this starts with the characters in `srcChars`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool startsWith(ConstChar16Ptr srcChars,
+            int32_t srcLength) const;
+
+  /**
+   * Determine if this ends with the characters in `srcChars`
+   * in the range  [`srcStart`, `srcStart + srcLength`).
+   * @param srcChars The characters to match.
+   * @param srcStart the offset into `srcText` to start matching
+   * @param srcLength the number of characters in `srcChars` to match
+   * @return true if this ends with the characters in `srcChars`, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool startsWith(const char16_t *srcChars,
+            int32_t srcStart,
+            int32_t srcLength) const;
+
+  /**
+   * Determine if this ends with the characters in `text`
+   * @param text The text to match.
+   * @return true if this ends with the characters in `text`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool endsWith(const UnicodeString& text) const;
+
+  /**
+   * Determine if this ends with the characters in `srcText`
+   * in the range [`srcStart`, `srcStart + srcLength`).
+   * @param srcText The text to match.
+   * @param srcStart the offset into `srcText` to start matching
+   * @param srcLength the number of characters in `srcText` to match
+   * @return true if this ends with the characters in `text`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool endsWith(const UnicodeString& srcText,
+          int32_t srcStart,
+          int32_t srcLength) const;
+
+  /**
+   * Determine if this ends with the characters in `srcChars`
+   * @param srcChars The characters to match.
+   * @param srcLength the number of characters in `srcChars`
+   * @return true if this ends with the characters in `srcChars`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool endsWith(ConstChar16Ptr srcChars,
+          int32_t srcLength) const;
+
+  /**
+   * Determine if this ends with the characters in `srcChars`
+   * in the range  [`srcStart`, `srcStart + srcLength`).
+   * @param srcChars The characters to match.
+   * @param srcStart the offset into `srcText` to start matching
+   * @param srcLength the number of characters in `srcChars` to match
+   * @return true if this ends with the characters in `srcChars`,
+   * false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool endsWith(const char16_t *srcChars,
+          int32_t srcStart,
+          int32_t srcLength) const;
+
+
+  /* Searching - bitwise only */
+
+  /**
+   * Locate in this the first occurrence of the characters in `text`,
+   * using bitwise comparison.
+   * @param text The text to search for.
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(const UnicodeString& text) const;
+
+  /**
+   * Locate in this the first occurrence of the characters in `text`
+   * starting at offset `start`, using bitwise comparison.
+   * @param text The text to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(const UnicodeString& text,
+              int32_t start) const;
+
+  /**
+   * Locate in this the first occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `text`, using bitwise comparison.
+   * @param text The text to search for.
+   * @param start The offset at which searching will start.
+   * @param length The number of characters to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(const UnicodeString& text,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the first occurrence in the range
+   * [`start`, `start + length`) of the characters
+   *  in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`),
+   * using bitwise comparison.
+   * @param srcText The text to search for.
+   * @param srcStart the offset into `srcText` at which
+   * to start matching
+   * @param srcLength the number of characters in `srcText` to match
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the first occurrence of the characters in
+   * `srcChars`
+   * starting at offset `start`, using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcLength the number of characters in `srcChars` to match
+   * @param start the offset into this at which to start matching
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(const char16_t *srcChars,
+              int32_t srcLength,
+              int32_t start) const;
+
+  /**
+   * Locate in this the first occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `srcChars`, using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcLength the number of characters in `srcChars`
+   * @param start The offset at which searching will start.
+   * @param length The number of characters to search
+   * @return The offset into this of the start of `srcChars`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(ConstChar16Ptr srcChars,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the first occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`),
+   * using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcStart the offset into `srcChars` at which
+   * to start matching
+   * @param srcLength the number of characters in `srcChars` to match
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  int32_t indexOf(const char16_t *srcChars,
+              int32_t srcStart,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the first occurrence of the BMP code point `c`,
+   * using bitwise comparison.
+   * @param c The code unit to search for.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(char16_t c) const;
+
+  /**
+   * Locate in this the first occurrence of the code point `c`,
+   * using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(UChar32 c) const;
+
+  /**
+   * Locate in this the first occurrence of the BMP code point `c`,
+   * starting at offset `start`, using bitwise comparison.
+   * @param c The code unit to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(char16_t c,
+              int32_t start) const;
+
+  /**
+   * Locate in this the first occurrence of the code point `c`
+   * starting at offset `start`, using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(UChar32 c,
+              int32_t start) const;
+
+  /**
+   * Locate in this the first occurrence of the BMP code point `c`
+   * in the range [`start`, `start + length`),
+   * using bitwise comparison.
+   * @param c The code unit to search for.
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(char16_t c,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the first occurrence of the code point `c`
+   * in the range [`start`, `start + length`),
+   * using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t indexOf(UChar32 c,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence of the characters in `text`,
+   * using bitwise comparison.
+   * @param text The text to search for.
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(const UnicodeString& text) const;
+
+  /**
+   * Locate in this the last occurrence of the characters in `text`
+   * starting at offset `start`, using bitwise comparison.
+   * @param text The text to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(const UnicodeString& text,
+              int32_t start) const;
+
+  /**
+   * Locate in this the last occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `text`, using bitwise comparison.
+   * @param text The text to search for.
+   * @param start The offset at which searching will start.
+   * @param length The number of characters to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(const UnicodeString& text,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`),
+   * using bitwise comparison.
+   * @param srcText The text to search for.
+   * @param srcStart the offset into `srcText` at which
+   * to start matching
+   * @param srcLength the number of characters in `srcText` to match
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence of the characters in `srcChars`
+   * starting at offset `start`, using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcLength the number of characters in `srcChars` to match
+   * @param start the offset into this at which to start matching
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(const char16_t *srcChars,
+              int32_t srcLength,
+              int32_t start) const;
+
+  /**
+   * Locate in this the last occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `srcChars`, using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcLength the number of characters in `srcChars`
+   * @param start The offset at which searching will start.
+   * @param length The number of characters to search
+   * @return The offset into this of the start of `srcChars`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(ConstChar16Ptr srcChars,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence in the range
+   * [`start`, `start + length`) of the characters
+   * in `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`),
+   * using bitwise comparison.
+   * @param srcChars The text to search for.
+   * @param srcStart the offset into `srcChars` at which
+   * to start matching
+   * @param srcLength the number of characters in `srcChars` to match
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of the start of `text`,
+   * or -1 if not found.
+   * @stable ICU 2.0
+   */
+  int32_t lastIndexOf(const char16_t *srcChars,
+              int32_t srcStart,
+              int32_t srcLength,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence of the BMP code point `c`,
+   * using bitwise comparison.
+   * @param c The code unit to search for.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(char16_t c) const;
+
+  /**
+   * Locate in this the last occurrence of the code point `c`,
+   * using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(UChar32 c) const;
+
+  /**
+   * Locate in this the last occurrence of the BMP code point `c`
+   * starting at offset `start`, using bitwise comparison.
+   * @param c The code unit to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(char16_t c,
+              int32_t start) const;
+
+  /**
+   * Locate in this the last occurrence of the code point `c`
+   * starting at offset `start`, using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @param start The offset at which searching will start.
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(UChar32 c,
+              int32_t start) const;
+
+  /**
+   * Locate in this the last occurrence of the BMP code point `c`
+   * in the range [`start`, `start + length`),
+   * using bitwise comparison.
+   * @param c The code unit to search for.
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(char16_t c,
+              int32_t start,
+              int32_t length) const;
+
+  /**
+   * Locate in this the last occurrence of the code point `c`
+   * in the range [`start`, `start + length`),
+   * using bitwise comparison.
+   *
+   * @param c The code point to search for.
+   * @param start the offset into this at which to start matching
+   * @param length the number of characters in this to search
+   * @return The offset into this of `c`, or -1 if not found.
+   * @stable ICU 2.0
+   */
+  inline int32_t lastIndexOf(UChar32 c,
+              int32_t start,
+              int32_t length) const;
+
+
+  /* Character access */
+
+  /**
+   * Return the code unit at offset `offset`.
+   * If the offset is not valid (0..length()-1) then U+ffff is returned.
+   * @param offset a valid offset into the text
+   * @return the code unit at offset `offset`
+   *         or 0xffff if the offset is not valid for this string
+   * @stable ICU 2.0
+   */
+  inline char16_t charAt(int32_t offset) const;
+
+  /**
+   * Return the code unit at offset `offset`.
+   * If the offset is not valid (0..length()-1) then U+ffff is returned.
+   * @param offset a valid offset into the text
+   * @return the code unit at offset `offset`
+   * @stable ICU 2.0
+   */
+  inline char16_t operator[] (int32_t offset) const;
+
+  /**
+   * Return the code point that contains the code unit
+   * at offset `offset`.
+   * If the offset is not valid (0..length()-1) then U+ffff is returned.
+   * @param offset a valid offset into the text
+   * that indicates the text offset of any of the code units
+   * that will be assembled into a code point (21-bit value) and returned
+   * @return the code point of text at `offset`
+   *         or 0xffff if the offset is not valid for this string
+   * @stable ICU 2.0
+   */
+  UChar32 char32At(int32_t offset) const;
+
+  /**
+   * Adjust a random-access offset so that
+   * it points to the beginning of a Unicode character.
+   * The offset that is passed in points to
+   * any code unit of a code point,
+   * while the returned offset will point to the first code unit
+   * of the same code point.
+   * In UTF-16, if the input offset points to a second surrogate
+   * of a surrogate pair, then the returned offset will point
+   * to the first surrogate.
+   * @param offset a valid offset into one code point of the text
+   * @return offset of the first code unit of the same code point
+   * @see U16_SET_CP_START
+   * @stable ICU 2.0
+   */
+  int32_t getChar32Start(int32_t offset) const;
+
+  /**
+   * Adjust a random-access offset so that
+   * it points behind a Unicode character.
+   * The offset that is passed in points behind
+   * any code unit of a code point,
+   * while the returned offset will point behind the last code unit
+   * of the same code point.
+   * In UTF-16, if the input offset points behind the first surrogate
+   * (i.e., to the second surrogate)
+   * of a surrogate pair, then the returned offset will point
+   * behind the second surrogate (i.e., to the first surrogate).
+   * @param offset a valid offset after any code unit of a code point of the text
+   * @return offset of the first code unit after the same code point
+   * @see U16_SET_CP_LIMIT
+   * @stable ICU 2.0
+   */
+  int32_t getChar32Limit(int32_t offset) const;
+
+  /**
+   * Move the code unit index along the string by delta code points.
+   * Interpret the input index as a code unit-based offset into the string,
+   * move the index forward or backward by delta code points, and
+   * return the resulting index.
+   * The input index should point to the first code unit of a code point,
+   * if there is more than one.
+   *
+   * Both input and output indexes are code unit-based as for all
+   * string indexes/offsets in ICU (and other libraries, like MBCS char*).
+   * If delta<0 then the index is moved backward (toward the start of the string).
+   * If delta>0 then the index is moved forward (toward the end of the string).
+   *
+   * This behaves like CharacterIterator::move32(delta, kCurrent).
+   *
+   * Behavior for out-of-bounds indexes:
+   * `moveIndex32` pins the input index to 0..length(), i.e.,
+   * if the input index<0 then it is pinned to 0;
+   * if it is index>length() then it is pinned to length().
+   * Afterwards, the index is moved by `delta` code points
+   * forward or backward,
+   * but no further backward than to 0 and no further forward than to length().
+   * The resulting index return value will be in between 0 and length(), inclusively.
+   *
+   * Examples:
+   * \code
+   *     // s has code points 'a' U+10000 'b' U+10ffff U+2029
+   *     UnicodeString s(u"a\U00010000b\U0010ffff\u2029");
+   *
+   *     // initial index: position of U+10000
+   *     int32_t index=1;
+   *
+   *     // the following examples will all result in index==4, position of U+10ffff
+   *
+   *     // skip 2 code points from some position in the string
+   *     index=s.moveIndex32(index, 2); // skips U+10000 and 'b'
+   *
+   *     // go to the 3rd code point from the start of s (0-based)
+   *     index=s.moveIndex32(0, 3); // skips 'a', U+10000, and 'b'
+   *
+   *     // go to the next-to-last code point of s
+   *     index=s.moveIndex32(s.length(), -2); // backward-skips U+2029 and U+10ffff
+   * \endcode
+   *
+   * @param index input code unit index
+   * @param delta (signed) code point count to move the index forward or backward
+   *        in the string
+   * @return the resulting code unit index
+   * @stable ICU 2.0
+   */
+  int32_t moveIndex32(int32_t index, int32_t delta) const;
+
+  /* Substring extraction */
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + length`) into the array `dst`,
+   * beginning at `dstStart`.
+   * If the string aliases to `dst` itself as an external buffer,
+   * then extract() will not copy the contents.
+   *
+   * @param start offset of first character which will be copied into the array
+   * @param length the number of characters to extract
+   * @param dst array in which to copy characters.  The length of `dst`
+   * must be at least (`dstStart + length`).
+   * @param dstStart the offset in `dst` where the first character
+   * will be extracted
+   * @stable ICU 2.0
+   */
+  inline void extract(int32_t start,
+           int32_t length,
+           Char16Ptr dst,
+           int32_t dstStart = 0) const;
+
+  /**
+   * Copy the contents of the string into dest.
+   * This is a convenience function that
+   * checks if there is enough space in dest,
+   * extracts the entire string if possible,
+   * and NUL-terminates dest if possible.
+   *
+   * If the string fits into dest but cannot be NUL-terminated
+   * (length()==destCapacity) then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
+   * If the string itself does not fit into dest
+   * (length()>destCapacity) then the error code is set to U_BUFFER_OVERFLOW_ERROR.
+   *
+   * If the string aliases to `dest` itself as an external buffer,
+   * then extract() will not copy the contents.
+   *
+   * @param dest Destination string buffer.
+   * @param destCapacity Number of char16_ts available at dest.
+   * @param errorCode ICU error code.
+   * @return length()
+   * @stable ICU 2.0
+   */
+  int32_t
+  extract(Char16Ptr dest, int32_t destCapacity,
+          UErrorCode &errorCode) const;
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + length`) into the  UnicodeString
+   * `target`.
+   * @param start offset of first character which will be copied
+   * @param length the number of characters to extract
+   * @param target UnicodeString into which to copy characters.
+   * @stable ICU 2.0
+   */
+  inline void extract(int32_t start,
+           int32_t length,
+           UnicodeString& target) const;
+
+  /**
+   * Copy the characters in the range [`start`, `limit`)
+   * into the array `dst`, beginning at `dstStart`.
+   * @param start offset of first character which will be copied into the array
+   * @param limit offset immediately following the last character to be copied
+   * @param dst array in which to copy characters.  The length of `dst`
+   * must be at least (`dstStart + (limit - start)`).
+   * @param dstStart the offset in `dst` where the first character
+   * will be extracted
+   * @stable ICU 2.0
+   */
+  inline void extractBetween(int32_t start,
+              int32_t limit,
+              char16_t *dst,
+              int32_t dstStart = 0) const;
+
+  /**
+   * Copy the characters in the range [`start`, `limit`)
+   * into the UnicodeString `target`.  Replaceable API.
+   * @param start offset of first character which will be copied
+   * @param limit offset immediately following the last character to be copied
+   * @param target UnicodeString into which to copy characters.
+   * @stable ICU 2.0
+   */
+  virtual void extractBetween(int32_t start,
+              int32_t limit,
+              UnicodeString& target) const;
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + startLength`) into an array of characters.
+   * All characters must be invariant (see utypes.h).
+   * Use US_INV as the last, signature-distinguishing parameter.
+   *
+   * This function does not write any more than `targetCapacity`
+   * characters but returns the length of the entire output string
+   * so that one can allocate a larger buffer and call the function again
+   * if necessary.
+   * The output string is NUL-terminated if possible.
+   *
+   * @param start offset of first character which will be copied
+   * @param startLength the number of characters to extract
+   * @param target the target buffer for extraction, can be NULL
+   *               if targetLength is 0
+   * @param targetCapacity the length of the target buffer
+   * @param inv Signature-distinguishing paramater, use US_INV.
+   * @return the output string length, not including the terminating NUL
+   * @stable ICU 3.2
+   */
+  int32_t extract(int32_t start,
+           int32_t startLength,
+           char *target,
+           int32_t targetCapacity,
+           enum EInvariant inv) const;
+
+#if U_CHARSET_IS_UTF8 || !UCONFIG_NO_CONVERSION
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + length`) into an array of characters
+   * in the platform's default codepage.
+   * This function does not write any more than `targetLength`
+   * characters but returns the length of the entire output string
+   * so that one can allocate a larger buffer and call the function again
+   * if necessary.
+   * The output string is NUL-terminated if possible.
+   *
+   * @param start offset of first character which will be copied
+   * @param startLength the number of characters to extract
+   * @param target the target buffer for extraction
+   * @param targetLength the length of the target buffer
+   * If `target` is NULL, then the number of bytes required for
+   * `target` is returned.
+   * @return the output string length, not including the terminating NUL
+   * @stable ICU 2.0
+   */
+  int32_t extract(int32_t start,
+           int32_t startLength,
+           char *target,
+           uint32_t targetLength) const;
+
+#endif
+
+#if !UCONFIG_NO_CONVERSION
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + length`) into an array of characters
+   * in a specified codepage.
+   * The output string is NUL-terminated.
+   *
+   * Recommendation: For invariant-character strings use
+   * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
+   * because it avoids object code dependencies of UnicodeString on
+   * the conversion code.
+   *
+   * @param start offset of first character which will be copied
+   * @param startLength the number of characters to extract
+   * @param target the target buffer for extraction
+   * @param codepage the desired codepage for the characters.  0 has
+   * the special meaning of the default codepage
+   * If `codepage` is an empty string (`""`),
+   * then a simple conversion is performed on the codepage-invariant
+   * subset ("invariant characters") of the platform encoding. See utypes.h.
+   * If `target` is NULL, then the number of bytes required for
+   * `target` is returned. It is assumed that the target is big enough
+   * to fit all of the characters.
+   * @return the output string length, not including the terminating NUL
+   * @stable ICU 2.0
+   */
+  inline int32_t extract(int32_t start,
+                 int32_t startLength,
+                 char *target,
+                 const char *codepage = 0) const;
+
+  /**
+   * Copy the characters in the range
+   * [`start`, `start + length`) into an array of characters
+   * in a specified codepage.
+   * This function does not write any more than `targetLength`
+   * characters but returns the length of the entire output string
+   * so that one can allocate a larger buffer and call the function again
+   * if necessary.
+   * The output string is NUL-terminated if possible.
+   *
+   * Recommendation: For invariant-character strings use
+   * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
+   * because it avoids object code dependencies of UnicodeString on
+   * the conversion code.
+   *
+   * @param start offset of first character which will be copied
+   * @param startLength the number of characters to extract
+   * @param target the target buffer for extraction
+   * @param targetLength the length of the target buffer
+   * @param codepage the desired codepage for the characters.  0 has
+   * the special meaning of the default codepage
+   * If `codepage` is an empty string (`""`),
+   * then a simple conversion is performed on the codepage-invariant
+   * subset ("invariant characters") of the platform encoding. See utypes.h.
+   * If `target` is NULL, then the number of bytes required for
+   * `target` is returned.
+   * @return the output string length, not including the terminating NUL
+   * @stable ICU 2.0
+   */
+  int32_t extract(int32_t start,
+           int32_t startLength,
+           char *target,
+           uint32_t targetLength,
+           const char *codepage) const;
+
+  /**
+   * Convert the UnicodeString into a codepage string using an existing UConverter.
+   * The output string is NUL-terminated if possible.
+   *
+   * This function avoids the overhead of opening and closing a converter if
+   * multiple strings are extracted.
+   *
+   * @param dest destination string buffer, can be NULL if destCapacity==0
+   * @param destCapacity the number of chars available at dest
+   * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called),
+   *        or NULL for the default converter
+   * @param errorCode normal ICU error code
+   * @return the length of the output string, not counting the terminating NUL;
+   *         if the length is greater than destCapacity, then the string will not fit
+   *         and a buffer of the indicated length would need to be passed in
+   * @stable ICU 2.0
+   */
+  int32_t extract(char *dest, int32_t destCapacity,
+                  UConverter *cnv,
+                  UErrorCode &errorCode) const;
+
+#endif
+
+  /**
+   * Create a temporary substring for the specified range.
+   * Unlike the substring constructor and setTo() functions,
+   * the object returned here will be a read-only alias (using getBuffer())
+   * rather than copying the text.
+   * As a result, this substring operation is much faster but requires
+   * that the original string not be modified or deleted during the lifetime
+   * of the returned substring object.
+   * @param start offset of the first character visible in the substring
+   * @param length length of the substring
+   * @return a read-only alias UnicodeString object for the substring
+   * @stable ICU 4.4
+   */
+  UnicodeString tempSubString(int32_t start=0, int32_t length=INT32_MAX) const;
+
+  /**
+   * Create a temporary substring for the specified range.
+   * Same as tempSubString(start, length) except that the substring range
+   * is specified as a (start, limit) pair (with an exclusive limit index)
+   * rather than a (start, length) pair.
+   * @param start offset of the first character visible in the substring
+   * @param limit offset immediately following the last character visible in the substring
+   * @return a read-only alias UnicodeString object for the substring
+   * @stable ICU 4.4
+   */
+  inline UnicodeString tempSubStringBetween(int32_t start, int32_t limit=INT32_MAX) const;
+
+  /**
+   * Convert the UnicodeString to UTF-8 and write the result
+   * to a ByteSink. This is called by toUTF8String().
+   * Unpaired surrogates are replaced with U+FFFD.
+   * Calls u_strToUTF8WithSub().
+   *
+   * @param sink A ByteSink to which the UTF-8 version of the string is written.
+   *             sink.Flush() is called at the end.
+   * @stable ICU 4.2
+   * @see toUTF8String
+   */
+  void toUTF8(ByteSink &sink) const;
+
+  /**
+   * Convert the UnicodeString to UTF-8 and append the result
+   * to a standard string.
+   * Unpaired surrogates are replaced with U+FFFD.
+   * Calls toUTF8().
+   *
+   * @param result A standard string (or a compatible object)
+   *        to which the UTF-8 version of the string is appended.
+   * @return The string object.
+   * @stable ICU 4.2
+   * @see toUTF8
+   */
+  template<typename StringClass>
+  StringClass &toUTF8String(StringClass &result) const {
+    StringByteSink<StringClass> sbs(&result, length());
+    toUTF8(sbs);
+    return result;
+  }
+
+  /**
+   * Convert the UnicodeString to UTF-32.
+   * Unpaired surrogates are replaced with U+FFFD.
+   * Calls u_strToUTF32WithSub().
+   *
+   * @param utf32 destination string buffer, can be NULL if capacity==0
+   * @param capacity the number of UChar32s available at utf32
+   * @param errorCode Standard ICU error code. Its input value must
+   *                  pass the U_SUCCESS() test, or else the function returns
+   *                  immediately. Check for U_FAILURE() on output or use with
+   *                  function chaining. (See User Guide for details.)
+   * @return The length of the UTF-32 string.
+   * @see fromUTF32
+   * @stable ICU 4.2
+   */
+  int32_t toUTF32(UChar32 *utf32, int32_t capacity, UErrorCode &errorCode) const;
+
+  /* Length operations */
+
+  /**
+   * Return the length of the UnicodeString object.
+   * The length is the number of char16_t code units are in the UnicodeString.
+   * If you want the number of code points, please use countChar32().
+   * @return the length of the UnicodeString object
+   * @see countChar32
+   * @stable ICU 2.0
+   */
+  inline int32_t length(void) const;
+
+  /**
+   * Count Unicode code points in the length char16_t code units of the string.
+   * A code point may occupy either one or two char16_t code units.
+   * Counting code points involves reading all code units.
+   *
+   * This functions is basically the inverse of moveIndex32().
+   *
+   * @param start the index of the first code unit to check
+   * @param length the number of char16_t code units to check
+   * @return the number of code points in the specified code units
+   * @see length
+   * @stable ICU 2.0
+   */
+  int32_t
+  countChar32(int32_t start=0, int32_t length=INT32_MAX) const;
+
+  /**
+   * Check if the length char16_t code units of the string
+   * contain more Unicode code points than a certain number.
+   * This is more efficient than counting all code points in this part of the string
+   * and comparing that number with a threshold.
+   * This function may not need to scan the string at all if the length
+   * falls within a certain range, and
+   * never needs to count more than 'number+1' code points.
+   * Logically equivalent to (countChar32(start, length)>number).
+   * A Unicode code point may occupy either one or two char16_t code units.
+   *
+   * @param start the index of the first code unit to check (0 for the entire string)
+   * @param length the number of char16_t code units to check
+   *               (use INT32_MAX for the entire string; remember that start/length
+   *                values are pinned)
+   * @param number The number of code points in the (sub)string is compared against
+   *               the 'number' parameter.
+   * @return Boolean value for whether the string contains more Unicode code points
+   *         than 'number'. Same as (u_countChar32(s, length)>number).
+   * @see countChar32
+   * @see u_strHasMoreChar32Than
+   * @stable ICU 2.4
+   */
+  UBool
+  hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const;
+
+  /**
+   * Determine if this string is empty.
+   * @return true if this string contains 0 characters, false otherwise.
+   * @stable ICU 2.0
+   */
+  inline UBool isEmpty(void) const;
+
+  /**
+   * Return the capacity of the internal buffer of the UnicodeString object.
+   * This is useful together with the getBuffer functions.
+   * See there for details.
+   *
+   * @return the number of char16_ts available in the internal buffer
+   * @see getBuffer
+   * @stable ICU 2.0
+   */
+  inline int32_t getCapacity(void) const;
+
+  /* Other operations */
+
+  /**
+   * Generate a hash code for this object.
+   * @return The hash code of this UnicodeString.
+   * @stable ICU 2.0
+   */
+  inline int32_t hashCode(void) const;
+
+  /**
+   * Determine if this object contains a valid string.
+   * A bogus string has no value. It is different from an empty string,
+   * although in both cases isEmpty() returns true and length() returns 0.
+   * setToBogus() and isBogus() can be used to indicate that no string value is available.
+   * For a bogus string, getBuffer() and getTerminatedBuffer() return NULL, and
+   * length() returns 0.
+   *
+   * @return true if the string is bogus/invalid, false otherwise
+   * @see setToBogus()
+   * @stable ICU 2.0
+   */
+  inline UBool isBogus(void) const;
+
+
+  //========================================
+  // Write operations
+  //========================================
+
+  /* Assignment operations */
+
+  /**
+   * Assignment operator.  Replace the characters in this UnicodeString
+   * with the characters from `srcText`.
+   *
+   * Starting with ICU 2.4, the assignment operator and the copy constructor
+   * allocate a new buffer and copy the buffer contents even for readonly aliases.
+   * By contrast, the fastCopyFrom() function implements the old,
+   * more efficient but less safe behavior
+   * of making this string also a readonly alias to the same buffer.
+   *
+   * If the source object has an "open" buffer from getBuffer(minCapacity),
+   * then the copy is an empty string.
+   *
+   * @param srcText The text containing the characters to replace
+   * @return a reference to this
+   * @stable ICU 2.0
+   * @see fastCopyFrom
+   */
+  UnicodeString &operator=(const UnicodeString &srcText);
+
+  /**
+   * Almost the same as the assignment operator.
+   * Replace the characters in this UnicodeString
+   * with the characters from `srcText`.
+   *
+   * This function works the same as the assignment operator
+   * for all strings except for ones that are readonly aliases.
+   *
+   * Starting with ICU 2.4, the assignment operator and the copy constructor
+   * allocate a new buffer and copy the buffer contents even for readonly aliases.
+   * This function implements the old, more efficient but less safe behavior
+   * of making this string also a readonly alias to the same buffer.
+   *
+   * The fastCopyFrom function must be used only if it is known that the lifetime of
+   * this UnicodeString does not exceed the lifetime of the aliased buffer
+   * including its contents, for example for strings from resource bundles
+   * or aliases to string constants.
+   *
+   * If the source object has an "open" buffer from getBuffer(minCapacity),
+   * then the copy is an empty string.
+   *
+   * @param src The text containing the characters to replace.
+   * @return a reference to this
+   * @stable ICU 2.4
+   */
+  UnicodeString &fastCopyFrom(const UnicodeString &src);
+
+  /**
+   * Move assignment operator; might leave src in bogus state.
+   * This string will have the same contents and state that the source string had.
+   * The behavior is undefined if *this and src are the same object.
+   * @param src source string
+   * @return *this
+   * @stable ICU 56
+   */
+  UnicodeString &operator=(UnicodeString &&src) U_NOEXCEPT;
+
+  /**
+   * Swap strings.
+   * @param other other string
+   * @stable ICU 56
+   */
+  void swap(UnicodeString &other) U_NOEXCEPT;
+
+  /**
+   * Non-member UnicodeString swap function.
+   * @param s1 will get s2's contents and state
+   * @param s2 will get s1's contents and state
+   * @stable ICU 56
+   */
+  friend inline void U_EXPORT2
+  swap(UnicodeString &s1, UnicodeString &s2) U_NOEXCEPT {
+    s1.swap(s2);
+  }
+
+  /**
+   * Assignment operator.  Replace the characters in this UnicodeString
+   * with the code unit `ch`.
+   * @param ch the code unit to replace
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& operator= (char16_t ch);
+
+  /**
+   * Assignment operator.  Replace the characters in this UnicodeString
+   * with the code point `ch`.
+   * @param ch the code point to replace
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& operator= (UChar32 ch);
+
+  /**
+   * Set the text in the UnicodeString object to the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcText.length()`).
+   * `srcText` is not modified.
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcText` where new characters
+   * will be obtained
+   * @return a reference to this
+   * @stable ICU 2.2
+   */
+  inline UnicodeString& setTo(const UnicodeString& srcText,
+               int32_t srcStart);
+
+  /**
+   * Set the text in the UnicodeString object to the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`).
+   * `srcText` is not modified.
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcText` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcText` in the
+   * replace string.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& setTo(const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength);
+
+  /**
+   * Set the text in the UnicodeString object to the characters in
+   * `srcText`.
+   * `srcText` is not modified.
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& setTo(const UnicodeString& srcText);
+
+  /**
+   * Set the characters in the UnicodeString object to the characters
+   * in `srcChars`. `srcChars` is not modified.
+   * @param srcChars the source for the new characters
+   * @param srcLength the number of Unicode characters in srcChars.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& setTo(const char16_t *srcChars,
+               int32_t srcLength);
+
+  /**
+   * Set the characters in the UnicodeString object to the code unit
+   * `srcChar`.
+   * @param srcChar the code unit which becomes the UnicodeString's character
+   * content
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& setTo(char16_t srcChar);
+
+  /**
+   * Set the characters in the UnicodeString object to the code point
+   * `srcChar`.
+   * @param srcChar the code point which becomes the UnicodeString's character
+   * content
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& setTo(UChar32 srcChar);
+
+  /**
+   * Aliasing setTo() function, analogous to the readonly-aliasing char16_t* constructor.
+   * The text will be used for the UnicodeString object, but
+   * it will not be released when the UnicodeString is destroyed.
+   * This has copy-on-write semantics:
+   * When the string is modified, then the buffer is first copied into
+   * newly allocated memory.
+   * The aliased buffer is never modified.
+   *
+   * In an assignment to another UnicodeString, when using the copy constructor
+   * or the assignment operator, the text will be copied.
+   * When using fastCopyFrom(), the text will be aliased again,
+   * so that both strings then alias the same readonly-text.
+   *
+   * @param isTerminated specifies if `text` is `NUL`-terminated.
+   *                     This must be true if `textLength==-1`.
+   * @param text The characters to alias for the UnicodeString.
+   * @param textLength The number of Unicode characters in `text` to alias.
+   *                   If -1, then this constructor will determine the length
+   *                   by calling `u_strlen()`.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString &setTo(UBool isTerminated,
+                       ConstChar16Ptr text,
+                       int32_t textLength);
+
+  /**
+   * Aliasing setTo() function, analogous to the writable-aliasing char16_t* constructor.
+   * The text will be used for the UnicodeString object, but
+   * it will not be released when the UnicodeString is destroyed.
+   * This has write-through semantics:
+   * For as long as the capacity of the buffer is sufficient, write operations
+   * will directly affect the buffer. When more capacity is necessary, then
+   * a new buffer will be allocated and the contents copied as with regularly
+   * constructed strings.
+   * In an assignment to another UnicodeString, the buffer will be copied.
+   * The extract(Char16Ptr dst) function detects whether the dst pointer is the same
+   * as the string buffer itself and will in this case not copy the contents.
+   *
+   * @param buffer The characters to alias for the UnicodeString.
+   * @param buffLength The number of Unicode characters in `buffer` to alias.
+   * @param buffCapacity The size of `buffer` in char16_ts.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString &setTo(char16_t *buffer,
+                       int32_t buffLength,
+                       int32_t buffCapacity);
+
+  /**
+   * Make this UnicodeString object invalid.
+   * The string will test true with isBogus().
+   *
+   * A bogus string has no value. It is different from an empty string.
+   * It can be used to indicate that no string value is available.
+   * getBuffer() and getTerminatedBuffer() return NULL, and
+   * length() returns 0.
+   *
+   * This utility function is used throughout the UnicodeString
+   * implementation to indicate that a UnicodeString operation failed,
+   * and may be used in other functions,
+   * especially but not exclusively when such functions do not
+   * take a UErrorCode for simplicity.
+   *
+   * The following methods, and no others, will clear a string object's bogus flag:
+   * - remove()
+   * - remove(0, INT32_MAX)
+   * - truncate(0)
+   * - operator=() (assignment operator)
+   * - setTo(...)
+   *
+   * The simplest ways to turn a bogus string into an empty one
+   * is to use the remove() function.
+   * Examples for other functions that are equivalent to "set to empty string":
+   * \code
+   * if(s.isBogus()) {
+   *   s.remove();           // set to an empty string (remove all), or
+   *   s.remove(0, INT32_MAX); // set to an empty string (remove all), or
+   *   s.truncate(0);        // set to an empty string (complete truncation), or
+   *   s=UnicodeString();    // assign an empty string, or
+   *   s.setTo((UChar32)-1); // set to a pseudo code point that is out of range, or
+   *   s.setTo(u"", 0);      // set to an empty C Unicode string
+   * }
+   * \endcode
+   *
+   * @see isBogus()
+   * @stable ICU 2.0
+   */
+  void setToBogus();
+
+  /**
+   * Set the character at the specified offset to the specified character.
+   * @param offset A valid offset into the text of the character to set
+   * @param ch The new character
+   * @return A reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString& setCharAt(int32_t offset,
+               char16_t ch);
+
+
+  /* Append operations */
+
+  /**
+   * Append operator. Append the code unit `ch` to the UnicodeString
+   * object.
+   * @param ch the code unit to be appended
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+ inline  UnicodeString& operator+= (char16_t ch);
+
+  /**
+   * Append operator. Append the code point `ch` to the UnicodeString
+   * object.
+   * @param ch the code point to be appended
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+ inline  UnicodeString& operator+= (UChar32 ch);
+
+  /**
+   * Append operator. Append the characters in `srcText` to the
+   * UnicodeString object. `srcText` is not modified.
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& operator+= (const UnicodeString& srcText);
+
+  /**
+   * Append the characters
+   * in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`) to the
+   * UnicodeString object at offset `start`. `srcText`
+   * is not modified.
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcText` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcText` in
+   * the append string
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& append(const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLength);
+
+  /**
+   * Append the characters in `srcText` to the UnicodeString object.
+   * `srcText` is not modified.
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& append(const UnicodeString& srcText);
+
+  /**
+   * Append the characters in `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`) to the UnicodeString
+   * object at offset
+   * `start`. `srcChars` is not modified.
+   * @param srcChars the source for the new characters
+   * @param srcStart the offset into `srcChars` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcChars` in
+   *                  the append string; can be -1 if `srcChars` is NUL-terminated
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& append(const char16_t *srcChars,
+            int32_t srcStart,
+            int32_t srcLength);
+
+  /**
+   * Append the characters in `srcChars` to the UnicodeString object
+   * at offset `start`. `srcChars` is not modified.
+   * @param srcChars the source for the new characters
+   * @param srcLength the number of Unicode characters in `srcChars`;
+   *                  can be -1 if `srcChars` is NUL-terminated
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& append(ConstChar16Ptr srcChars,
+            int32_t srcLength);
+
+  /**
+   * Append the code unit `srcChar` to the UnicodeString object.
+   * @param srcChar the code unit to append
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& append(char16_t srcChar);
+
+  /**
+   * Append the code point `srcChar` to the UnicodeString object.
+   * @param srcChar the code point to append
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString& append(UChar32 srcChar);
+
+
+  /* Insert operations */
+
+  /**
+   * Insert the characters in `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`) into the UnicodeString
+   * object at offset `start`. `srcText` is not modified.
+   * @param start the offset where the insertion begins
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcText` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcText` in
+   * the insert string
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLength);
+
+  /**
+   * Insert the characters in `srcText` into the UnicodeString object
+   * at offset `start`. `srcText` is not modified.
+   * @param start the offset where the insertion begins
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            const UnicodeString& srcText);
+
+  /**
+   * Insert the characters in `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`) into the UnicodeString
+   *  object at offset `start`. `srcChars` is not modified.
+   * @param start the offset at which the insertion begins
+   * @param srcChars the source for the new characters
+   * @param srcStart the offset into `srcChars` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcChars`
+   * in the insert string
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            const char16_t *srcChars,
+            int32_t srcStart,
+            int32_t srcLength);
+
+  /**
+   * Insert the characters in `srcChars` into the UnicodeString object
+   * at offset `start`. `srcChars` is not modified.
+   * @param start the offset where the insertion begins
+   * @param srcChars the source for the new characters
+   * @param srcLength the number of Unicode characters in srcChars.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            ConstChar16Ptr srcChars,
+            int32_t srcLength);
+
+  /**
+   * Insert the code unit `srcChar` into the UnicodeString object at
+   * offset `start`.
+   * @param start the offset at which the insertion occurs
+   * @param srcChar the code unit to insert
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            char16_t srcChar);
+
+  /**
+   * Insert the code point `srcChar` into the UnicodeString object at
+   * offset `start`.
+   * @param start the offset at which the insertion occurs
+   * @param srcChar the code point to insert
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& insert(int32_t start,
+            UChar32 srcChar);
+
+
+  /* Replace operations */
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`) with the characters in
+   * `srcText` in the range
+   * [`srcStart`, `srcStart + srcLength`).
+   * `srcText` is not modified.
+   * @param start the offset at which the replace operation begins
+   * @param length the number of characters to replace. The character at
+   * `start + length` is not modified.
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcText` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcText` in
+   * the replace string
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replace(int32_t start,
+             int32_t length,
+             const UnicodeString& srcText,
+             int32_t srcStart,
+             int32_t srcLength);
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`)
+   * with the characters in `srcText`.  `srcText` is
+   *  not modified.
+   * @param start the offset at which the replace operation begins
+   * @param length the number of characters to replace. The character at
+   * `start + length` is not modified.
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replace(int32_t start,
+             int32_t length,
+             const UnicodeString& srcText);
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`) with the characters in
+   * `srcChars` in the range
+   * [`srcStart`, `srcStart + srcLength`). `srcChars`
+   * is not modified.
+   * @param start the offset at which the replace operation begins
+   * @param length the number of characters to replace.  The character at
+   * `start + length` is not modified.
+   * @param srcChars the source for the new characters
+   * @param srcStart the offset into `srcChars` where new characters
+   * will be obtained
+   * @param srcLength the number of characters in `srcChars`
+   * in the replace string
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replace(int32_t start,
+             int32_t length,
+             const char16_t *srcChars,
+             int32_t srcStart,
+             int32_t srcLength);
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`) with the characters in
+   * `srcChars`.  `srcChars` is not modified.
+   * @param start the offset at which the replace operation begins
+   * @param length number of characters to replace.  The character at
+   * `start + length` is not modified.
+   * @param srcChars the source for the new characters
+   * @param srcLength the number of Unicode characters in srcChars
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replace(int32_t start,
+             int32_t length,
+             ConstChar16Ptr srcChars,
+             int32_t srcLength);
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`) with the code unit
+   * `srcChar`.
+   * @param start the offset at which the replace operation begins
+   * @param length the number of characters to replace.  The character at
+   * `start + length` is not modified.
+   * @param srcChar the new code unit
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replace(int32_t start,
+             int32_t length,
+             char16_t srcChar);
+
+  /**
+   * Replace the characters in the range
+   * [`start`, `start + length`) with the code point
+   * `srcChar`.
+   * @param start the offset at which the replace operation begins
+   * @param length the number of characters to replace.  The character at
+   * `start + length` is not modified.
+   * @param srcChar the new code point
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString& replace(int32_t start, int32_t length, UChar32 srcChar);
+
+  /**
+   * Replace the characters in the range [`start`, `limit`)
+   * with the characters in `srcText`. `srcText` is not modified.
+   * @param start the offset at which the replace operation begins
+   * @param limit the offset immediately following the replace range
+   * @param srcText the source for the new characters
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replaceBetween(int32_t start,
+                int32_t limit,
+                const UnicodeString& srcText);
+
+  /**
+   * Replace the characters in the range [`start`, `limit`)
+   * with the characters in `srcText` in the range
+   * [`srcStart`, `srcLimit`). `srcText` is not modified.
+   * @param start the offset at which the replace operation begins
+   * @param limit the offset immediately following the replace range
+   * @param srcText the source for the new characters
+   * @param srcStart the offset into `srcChars` where new characters
+   * will be obtained
+   * @param srcLimit the offset immediately following the range to copy
+   * in `srcText`
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& replaceBetween(int32_t start,
+                int32_t limit,
+                const UnicodeString& srcText,
+                int32_t srcStart,
+                int32_t srcLimit);
+
+  /**
+   * Replace a substring of this object with the given text.
+   * @param start the beginning index, inclusive; `0 <= start <= limit`.
+   * @param limit the ending index, exclusive; `start <= limit <= length()`.
+   * @param text the text to replace characters `start` to `limit - 1`
+   * @stable ICU 2.0
+   */
+  virtual void handleReplaceBetween(int32_t start,
+                                    int32_t limit,
+                                    const UnicodeString& text);
+
+  /**
+   * Replaceable API
+   * @return true if it has MetaData
+   * @stable ICU 2.4
+   */
+  virtual UBool hasMetaData() const;
+
+  /**
+   * Copy a substring of this object, retaining attribute (out-of-band)
+   * information.  This method is used to duplicate or reorder substrings.
+   * The destination index must not overlap the source range.
+   *
+   * @param start the beginning index, inclusive; `0 <= start <= limit`.
+   * @param limit the ending index, exclusive; `start <= limit <= length()`.
+   * @param dest the destination index.  The characters from
+   *             `start..limit-1` will be copied to `dest`.
+   * Implementations of this method may assume that `dest <= start ||
+   * dest >= limit`.
+   * @stable ICU 2.0
+   */
+  virtual void copy(int32_t start, int32_t limit, int32_t dest);
+
+  /* Search and replace operations */
+
+  /**
+   * Replace all occurrences of characters in oldText with the characters
+   * in newText
+   * @param oldText the text containing the search text
+   * @param newText the text containing the replacement text
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& findAndReplace(const UnicodeString& oldText,
+                const UnicodeString& newText);
+
+  /**
+   * Replace all occurrences of characters in oldText with characters
+   * in newText
+   * in the range [`start`, `start + length`).
+   * @param start the start of the range in which replace will performed
+   * @param length the length of the range in which replace will be performed
+   * @param oldText the text containing the search text
+   * @param newText the text containing the replacement text
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& findAndReplace(int32_t start,
+                int32_t length,
+                const UnicodeString& oldText,
+                const UnicodeString& newText);
+
+  /**
+   * Replace all occurrences of characters in oldText in the range
+   * [`oldStart`, `oldStart + oldLength`) with the characters
+   * in newText in the range
+   * [`newStart`, `newStart + newLength`)
+   * in the range [`start`, `start + length`).
+   * @param start the start of the range in which replace will performed
+   * @param length the length of the range in which replace will be performed
+   * @param oldText the text containing the search text
+   * @param oldStart the start of the search range in `oldText`
+   * @param oldLength the length of the search range in `oldText`
+   * @param newText the text containing the replacement text
+   * @param newStart the start of the replacement range in `newText`
+   * @param newLength the length of the replacement range in `newText`
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString& findAndReplace(int32_t start,
+                int32_t length,
+                const UnicodeString& oldText,
+                int32_t oldStart,
+                int32_t oldLength,
+                const UnicodeString& newText,
+                int32_t newStart,
+                int32_t newLength);
+
+
+  /* Remove operations */
+
+  /**
+   * Removes all characters from the UnicodeString object and clears the bogus flag.
+   * This is the UnicodeString equivalent of std::string’s clear().
+   *
+   * @return a reference to this
+   * @see setToBogus
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& remove();
+
+  /**
+   * Remove the characters in the range
+   * [`start`, `start + length`) from the UnicodeString object.
+   * @param start the offset of the first character to remove
+   * @param length the number of characters to remove
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& remove(int32_t start,
+                               int32_t length = (int32_t)INT32_MAX);
+
+  /**
+   * Remove the characters in the range
+   * [`start`, `limit`) from the UnicodeString object.
+   * @param start the offset of the first character to remove
+   * @param limit the offset immediately following the range to remove
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& removeBetween(int32_t start,
+                                      int32_t limit = (int32_t)INT32_MAX);
+
+  /**
+   * Retain only the characters in the range
+   * [`start`, `limit`) from the UnicodeString object.
+   * Removes characters before `start` and at and after `limit`.
+   * @param start the offset of the first character to retain
+   * @param limit the offset immediately following the range to retain
+   * @return a reference to this
+   * @stable ICU 4.4
+   */
+  inline UnicodeString &retainBetween(int32_t start, int32_t limit = INT32_MAX);
+
+  /* Length operations */
+
+  /**
+   * Pad the start of this UnicodeString with the character `padChar`.
+   * If the length of this UnicodeString is less than targetLength,
+   * length() - targetLength copies of padChar will be added to the
+   * beginning of this UnicodeString.
+   * @param targetLength the desired length of the string
+   * @param padChar the character to use for padding. Defaults to
+   * space (U+0020)
+   * @return true if the text was padded, false otherwise.
+   * @stable ICU 2.0
+   */
+  UBool padLeading(int32_t targetLength,
+                    char16_t padChar = 0x0020);
+
+  /**
+   * Pad the end of this UnicodeString with the character `padChar`.
+   * If the length of this UnicodeString is less than targetLength,
+   * length() - targetLength copies of padChar will be added to the
+   * end of this UnicodeString.
+   * @param targetLength the desired length of the string
+   * @param padChar the character to use for padding. Defaults to
+   * space (U+0020)
+   * @return true if the text was padded, false otherwise.
+   * @stable ICU 2.0
+   */
+  UBool padTrailing(int32_t targetLength,
+                     char16_t padChar = 0x0020);
+
+  /**
+   * Truncate this UnicodeString to the `targetLength`.
+   * @param targetLength the desired length of this UnicodeString.
+   * @return true if the text was truncated, false otherwise
+   * @stable ICU 2.0
+   */
+  inline UBool truncate(int32_t targetLength);
+
+  /**
+   * Trims leading and trailing whitespace from this UnicodeString.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  UnicodeString& trim(void);
+
+
+  /* Miscellaneous operations */
+
+  /**
+   * Reverse this UnicodeString in place.
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& reverse(void);
+
+  /**
+   * Reverse the range [`start`, `start + length`) in
+   * this UnicodeString.
+   * @param start the start of the range to reverse
+   * @param length the number of characters to to reverse
+   * @return a reference to this
+   * @stable ICU 2.0
+   */
+  inline UnicodeString& reverse(int32_t start,
+             int32_t length);
+
+  /**
+   * Convert the characters in this to UPPER CASE following the conventions of
+   * the default locale.
+   * @return A reference to this.
+   * @stable ICU 2.0
+   */
+  UnicodeString& toUpper(void);
+
+  /**
+   * Convert the characters in this to UPPER CASE following the conventions of
+   * a specific locale.
+   * @param locale The locale containing the conventions to use.
+   * @return A reference to this.
+   * @stable ICU 2.0
+   */
+  UnicodeString& toUpper(const Locale& locale);
+
+  /**
+   * Convert the characters in this to lower case following the conventions of
+   * the default locale.
+   * @return A reference to this.
+   * @stable ICU 2.0
+   */
+  UnicodeString& toLower(void);
+
+  /**
+   * Convert the characters in this to lower case following the conventions of
+   * a specific locale.
+   * @param locale The locale containing the conventions to use.
+   * @return A reference to this.
+   * @stable ICU 2.0
+   */
+  UnicodeString& toLower(const Locale& locale);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+  /**
+   * Titlecase this string, convenience function using the default locale.
+   *
+   * Casing is locale-dependent and context-sensitive.
+   * Titlecasing uses a break iterator to find the first characters of words
+   * that are to be titlecased. It titlecases those characters and lowercases
+   * all others.
+   *
+   * The titlecase break iterator can be provided to customize for arbitrary
+   * styles, using rules and dictionaries beyond the standard iterators.
+   * It may be more efficient to always provide an iterator to avoid
+   * opening and closing one for each string.
+   * The standard titlecase iterator for the root locale implements the
+   * algorithm of Unicode TR 21.
+   *
+   * This function uses only the setText(), first() and next() methods of the
+   * provided break iterator.
+   *
+   * @param titleIter A break iterator to find the first characters of words
+   *                  that are to be titlecased.
+   *                  If none is provided (0), then a standard titlecase
+   *                  break iterator is opened.
+   *                  Otherwise the provided iterator is set to the string's text.
+   * @return A reference to this.
+   * @stable ICU 2.1
+   */
+  UnicodeString &toTitle(BreakIterator *titleIter);
+
+  /**
+   * Titlecase this string.
+   *
+   * Casing is locale-dependent and context-sensitive.
+   * Titlecasing uses a break iterator to find the first characters of words
+   * that are to be titlecased. It titlecases those characters and lowercases
+   * all others.
+   *
+   * The titlecase break iterator can be provided to customize for arbitrary
+   * styles, using rules and dictionaries beyond the standard iterators.
+   * It may be more efficient to always provide an iterator to avoid
+   * opening and closing one for each string.
+   * The standard titlecase iterator for the root locale implements the
+   * algorithm of Unicode TR 21.
+   *
+   * This function uses only the setText(), first() and next() methods of the
+   * provided break iterator.
+   *
+   * @param titleIter A break iterator to find the first characters of words
+   *                  that are to be titlecased.
+   *                  If none is provided (0), then a standard titlecase
+   *                  break iterator is opened.
+   *                  Otherwise the provided iterator is set to the string's text.
+   * @param locale    The locale to consider.
+   * @return A reference to this.
+   * @stable ICU 2.1
+   */
+  UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale);
+
+  /**
+   * Titlecase this string, with options.
+   *
+   * Casing is locale-dependent and context-sensitive.
+   * Titlecasing uses a break iterator to find the first characters of words
+   * that are to be titlecased. It titlecases those characters and lowercases
+   * all others. (This can be modified with options.)
+   *
+   * The titlecase break iterator can be provided to customize for arbitrary
+   * styles, using rules and dictionaries beyond the standard iterators.
+   * It may be more efficient to always provide an iterator to avoid
+   * opening and closing one for each string.
+   * The standard titlecase iterator for the root locale implements the
+   * algorithm of Unicode TR 21.
+   *
+   * This function uses only the setText(), first() and next() methods of the
+   * provided break iterator.
+   *
+   * @param titleIter A break iterator to find the first characters of words
+   *                  that are to be titlecased.
+   *                  If none is provided (0), then a standard titlecase
+   *                  break iterator is opened.
+   *                  Otherwise the provided iterator is set to the string's text.
+   * @param locale    The locale to consider.
+   * @param options   Options bit set, usually 0. See U_TITLECASE_NO_LOWERCASE,
+   *                  U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED,
+   *                  U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES.
+   * @param options Options bit set, see ucasemap_open().
+   * @return A reference to this.
+   * @stable ICU 3.8
+   */
+  UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale, uint32_t options);
+
+#endif
+
+  /**
+   * Case-folds the characters in this string.
+   *
+   * Case-folding is locale-independent and not context-sensitive,
+   * but there is an option for whether to include or exclude mappings for dotted I
+   * and dotless i that are marked with 'T' in CaseFolding.txt.
+   *
+   * The result may be longer or shorter than the original.
+   *
+   * @param options Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
+   * @return A reference to this.
+   * @stable ICU 2.0
+   */
+  UnicodeString &foldCase(uint32_t options=0 /*U_FOLD_CASE_DEFAULT*/);
+
+  //========================================
+  // Access to the internal buffer
+  //========================================
+
+  /**
+   * Get a read/write pointer to the internal buffer.
+   * The buffer is guaranteed to be large enough for at least minCapacity char16_ts,
+   * writable, and is still owned by the UnicodeString object.
+   * Calls to getBuffer(minCapacity) must not be nested, and
+   * must be matched with calls to releaseBuffer(newLength).
+   * If the string buffer was read-only or shared,
+   * then it will be reallocated and copied.
+   *
+   * An attempted nested call will return 0, and will not further modify the
+   * state of the UnicodeString object.
+   * It also returns 0 if the string is bogus.
+   *
+   * The actual capacity of the string buffer may be larger than minCapacity.
+   * getCapacity() returns the actual capacity.
+   * For many operations, the full capacity should be used to avoid reallocations.
+   *
+   * While the buffer is "open" between getBuffer(minCapacity)
+   * and releaseBuffer(newLength), the following applies:
+   * - The string length is set to 0.
+   * - Any read API call on the UnicodeString object will behave like on a 0-length string.
+   * - Any write API call on the UnicodeString object is disallowed and will have no effect.
+   * - You can read from and write to the returned buffer.
+   * - The previous string contents will still be in the buffer;
+   *   if you want to use it, then you need to call length() before getBuffer(minCapacity).
+   *   If the length() was greater than minCapacity, then any contents after minCapacity
+   *   may be lost.
+   *   The buffer contents is not NUL-terminated by getBuffer().
+   *   If length() < getCapacity() then you can terminate it by writing a NUL
+   *   at index length().
+   * - You must call releaseBuffer(newLength) before and in order to
+   *   return to normal UnicodeString operation.
+   *
+   * @param minCapacity the minimum number of char16_ts that are to be available
+   *        in the buffer, starting at the returned pointer;
+   *        default to the current string capacity if minCapacity==-1
+   * @return a writable pointer to the internal string buffer,
+   *         or nullptr if an error occurs (nested calls, out of memory)
+   *
+   * @see releaseBuffer
+   * @see getTerminatedBuffer()
+   * @stable ICU 2.0
+   */
+  char16_t *getBuffer(int32_t minCapacity);
+
+  /**
+   * Release a read/write buffer on a UnicodeString object with an
+   * "open" getBuffer(minCapacity).
+   * This function must be called in a matched pair with getBuffer(minCapacity).
+   * releaseBuffer(newLength) must be called if and only if a getBuffer(minCapacity) is "open".
+   *
+   * It will set the string length to newLength, at most to the current capacity.
+   * If newLength==-1 then it will set the length according to the
+   * first NUL in the buffer, or to the capacity if there is no NUL.
+   *
+   * After calling releaseBuffer(newLength) the UnicodeString is back to normal operation.
+   *
+   * @param newLength the new length of the UnicodeString object;
+   *        defaults to the current capacity if newLength is greater than that;
+   *        if newLength==-1, it defaults to u_strlen(buffer) but not more than
+   *        the current capacity of the string
+   *
+   * @see getBuffer(int32_t minCapacity)
+   * @stable ICU 2.0
+   */
+  void releaseBuffer(int32_t newLength=-1);
+
+  /**
+   * Get a read-only pointer to the internal buffer.
+   * This can be called at any time on a valid UnicodeString.
+   *
+   * It returns 0 if the string is bogus, or
+   * during an "open" getBuffer(minCapacity).
+   *
+   * It can be called as many times as desired.
+   * The pointer that it returns will remain valid until the UnicodeString object is modified,
+   * at which time the pointer is semantically invalidated and must not be used any more.
+   *
+   * The capacity of the buffer can be determined with getCapacity().
+   * The part after length() may or may not be initialized and valid,
+   * depending on the history of the UnicodeString object.
+   *
+   * The buffer contents is (probably) not NUL-terminated.
+   * You can check if it is with
+   * `(s.length() < s.getCapacity() && buffer[s.length()]==0)`.
+   * (See getTerminatedBuffer().)
+   *
+   * The buffer may reside in read-only memory. Its contents must not
+   * be modified.
+   *
+   * @return a read-only pointer to the internal string buffer,
+   *         or nullptr if the string is empty or bogus
+   *
+   * @see getBuffer(int32_t minCapacity)
+   * @see getTerminatedBuffer()
+   * @stable ICU 2.0
+   */
+  inline const char16_t *getBuffer() const;
+
+  /**
+   * Get a read-only pointer to the internal buffer,
+   * making sure that it is NUL-terminated.
+   * This can be called at any time on a valid UnicodeString.
+   *
+   * It returns 0 if the string is bogus, or
+   * during an "open" getBuffer(minCapacity), or if the buffer cannot
+   * be NUL-terminated (because memory allocation failed).
+   *
+   * It can be called as many times as desired.
+   * The pointer that it returns will remain valid until the UnicodeString object is modified,
+   * at which time the pointer is semantically invalidated and must not be used any more.
+   *
+   * The capacity of the buffer can be determined with getCapacity().
+   * The part after length()+1 may or may not be initialized and valid,
+   * depending on the history of the UnicodeString object.
+   *
+   * The buffer contents is guaranteed to be NUL-terminated.
+   * getTerminatedBuffer() may reallocate the buffer if a terminating NUL
+   * is written.
+   * For this reason, this function is not const, unlike getBuffer().
+   * Note that a UnicodeString may also contain NUL characters as part of its contents.
+   *
+   * The buffer may reside in read-only memory. Its contents must not
+   * be modified.
+   *
+   * @return a read-only pointer to the internal string buffer,
+   *         or 0 if the string is empty or bogus
+   *
+   * @see getBuffer(int32_t minCapacity)
+   * @see getBuffer()
+   * @stable ICU 2.2
+   */
+  const char16_t *getTerminatedBuffer();
+
+  //========================================
+  // Constructors
+  //========================================
+
+  /** Construct an empty UnicodeString.
+   * @stable ICU 2.0
+   */
+  inline UnicodeString();
+
+  /**
+   * Construct a UnicodeString with capacity to hold `capacity` char16_ts
+   * @param capacity the number of char16_ts this UnicodeString should hold
+   * before a resize is necessary; if count is greater than 0 and count
+   * code points c take up more space than capacity, then capacity is adjusted
+   * accordingly.
+   * @param c is used to initially fill the string
+   * @param count specifies how many code points c are to be written in the
+   *              string
+   * @stable ICU 2.0
+   */
+  UnicodeString(int32_t capacity, UChar32 c, int32_t count);
+
+  /**
+   * Single char16_t (code unit) constructor.
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_CHAR_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param ch the character to place in the UnicodeString
+   * @stable ICU 2.0
+   */
+  UNISTR_FROM_CHAR_EXPLICIT UnicodeString(char16_t ch);
+
+  /**
+   * Single UChar32 (code point) constructor.
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_CHAR_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param ch the character to place in the UnicodeString
+   * @stable ICU 2.0
+   */
+  UNISTR_FROM_CHAR_EXPLICIT UnicodeString(UChar32 ch);
+
+  /**
+   * char16_t* constructor.
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param text The characters to place in the UnicodeString.  `text`
+   * must be NULL (U+0000) terminated.
+   * @stable ICU 2.0
+   */
+  UNISTR_FROM_STRING_EXPLICIT UnicodeString(const char16_t *text);
+
+#if !U_CHAR16_IS_TYPEDEF
+  /**
+   * uint16_t * constructor.
+   * Delegates to UnicodeString(const char16_t *).
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param text NUL-terminated UTF-16 string
+   * @stable ICU 59
+   */
+  UNISTR_FROM_STRING_EXPLICIT UnicodeString(const uint16_t *text) :
+      UnicodeString(ConstChar16Ptr(text)) {}
+#endif
+
+#if U_SIZEOF_WCHAR_T==2 || defined(U_IN_DOXYGEN)
+  /**
+   * wchar_t * constructor.
+   * (Only defined if U_SIZEOF_WCHAR_T==2.)
+   * Delegates to UnicodeString(const char16_t *).
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param text NUL-terminated UTF-16 string
+   * @stable ICU 59
+   */
+  UNISTR_FROM_STRING_EXPLICIT UnicodeString(const wchar_t *text) :
+      UnicodeString(ConstChar16Ptr(text)) {}
+#endif
+
+  /**
+   * nullptr_t constructor.
+   * Effectively the same as the default constructor, makes an empty string object.
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param text nullptr
+   * @stable ICU 59
+   */
+  UNISTR_FROM_STRING_EXPLICIT inline UnicodeString(const std::nullptr_t text);
+
+  /**
+   * char16_t* constructor.
+   * @param text The characters to place in the UnicodeString.
+   * @param textLength The number of Unicode characters in `text`
+   * to copy.
+   * @stable ICU 2.0
+   */
+  UnicodeString(const char16_t *text,
+        int32_t textLength);
+
+#if !U_CHAR16_IS_TYPEDEF
+  /**
+   * uint16_t * constructor.
+   * Delegates to UnicodeString(const char16_t *, int32_t).
+   * @param text UTF-16 string
+   * @param textLength string length
+   * @stable ICU 59
+   */
+  UnicodeString(const uint16_t *text, int32_t textLength) :
+      UnicodeString(ConstChar16Ptr(text), textLength) {}
+#endif
+
+#if U_SIZEOF_WCHAR_T==2 || defined(U_IN_DOXYGEN)
+  /**
+   * wchar_t * constructor.
+   * (Only defined if U_SIZEOF_WCHAR_T==2.)
+   * Delegates to UnicodeString(const char16_t *, int32_t).
+   * @param text NUL-terminated UTF-16 string
+   * @param textLength string length
+   * @stable ICU 59
+   */
+  UnicodeString(const wchar_t *text, int32_t textLength) :
+      UnicodeString(ConstChar16Ptr(text), textLength) {}
+#endif
+
+  /**
+   * nullptr_t constructor.
+   * Effectively the same as the default constructor, makes an empty string object.
+   * @param text nullptr
+   * @param textLength ignored
+   * @stable ICU 59
+   */
+  inline UnicodeString(const std::nullptr_t text, int32_t textLength);
+
+  /**
+   * Readonly-aliasing char16_t* constructor.
+   * The text will be used for the UnicodeString object, but
+   * it will not be released when the UnicodeString is destroyed.
+   * This has copy-on-write semantics:
+   * When the string is modified, then the buffer is first copied into
+   * newly allocated memory.
+   * The aliased buffer is never modified.
+   *
+   * In an assignment to another UnicodeString, when using the copy constructor
+   * or the assignment operator, the text will be copied.
+   * When using fastCopyFrom(), the text will be aliased again,
+   * so that both strings then alias the same readonly-text.
+   *
+   * @param isTerminated specifies if `text` is `NUL`-terminated.
+   *                     This must be true if `textLength==-1`.
+   * @param text The characters to alias for the UnicodeString.
+   * @param textLength The number of Unicode characters in `text` to alias.
+   *                   If -1, then this constructor will determine the length
+   *                   by calling `u_strlen()`.
+   * @stable ICU 2.0
+   */
+  UnicodeString(UBool isTerminated,
+                ConstChar16Ptr text,
+                int32_t textLength);
+
+  /**
+   * Writable-aliasing char16_t* constructor.
+   * The text will be used for the UnicodeString object, but
+   * it will not be released when the UnicodeString is destroyed.
+   * This has write-through semantics:
+   * For as long as the capacity of the buffer is sufficient, write operations
+   * will directly affect the buffer. When more capacity is necessary, then
+   * a new buffer will be allocated and the contents copied as with regularly
+   * constructed strings.
+   * In an assignment to another UnicodeString, the buffer will be copied.
+   * The extract(Char16Ptr dst) function detects whether the dst pointer is the same
+   * as the string buffer itself and will in this case not copy the contents.
+   *
+   * @param buffer The characters to alias for the UnicodeString.
+   * @param buffLength The number of Unicode characters in `buffer` to alias.
+   * @param buffCapacity The size of `buffer` in char16_ts.
+   * @stable ICU 2.0
+   */
+  UnicodeString(char16_t *buffer, int32_t buffLength, int32_t buffCapacity);
+
+#if !U_CHAR16_IS_TYPEDEF
+  /**
+   * Writable-aliasing uint16_t * constructor.
+   * Delegates to UnicodeString(const char16_t *, int32_t, int32_t).
+   * @param buffer writable buffer of/for UTF-16 text
+   * @param buffLength length of the current buffer contents
+   * @param buffCapacity buffer capacity
+   * @stable ICU 59
+   */
+  UnicodeString(uint16_t *buffer, int32_t buffLength, int32_t buffCapacity) :
+      UnicodeString(Char16Ptr(buffer), buffLength, buffCapacity) {}
+#endif
+
+#if U_SIZEOF_WCHAR_T==2 || defined(U_IN_DOXYGEN)
+  /**
+   * Writable-aliasing wchar_t * constructor.
+   * (Only defined if U_SIZEOF_WCHAR_T==2.)
+   * Delegates to UnicodeString(const char16_t *, int32_t, int32_t).
+   * @param buffer writable buffer of/for UTF-16 text
+   * @param buffLength length of the current buffer contents
+   * @param buffCapacity buffer capacity
+   * @stable ICU 59
+   */
+  UnicodeString(wchar_t *buffer, int32_t buffLength, int32_t buffCapacity) :
+      UnicodeString(Char16Ptr(buffer), buffLength, buffCapacity) {}
+#endif
+
+  /**
+   * Writable-aliasing nullptr_t constructor.
+   * Effectively the same as the default constructor, makes an empty string object.
+   * @param buffer nullptr
+   * @param buffLength ignored
+   * @param buffCapacity ignored
+   * @stable ICU 59
+   */
+  inline UnicodeString(std::nullptr_t buffer, int32_t buffLength, int32_t buffCapacity);
+
+#if U_CHARSET_IS_UTF8 || !UCONFIG_NO_CONVERSION
+
+  /**
+   * char* constructor.
+   * Uses the default converter (and thus depends on the ICU conversion code)
+   * unless U_CHARSET_IS_UTF8 is set to 1.
+   *
+   * For ASCII (really "invariant character") strings it is more efficient to use
+   * the constructor that takes a US_INV (for its enum EInvariant).
+   * For ASCII (invariant-character) string literals, see UNICODE_STRING and
+   * UNICODE_STRING_SIMPLE.
+   *
+   * It is recommended to mark this constructor "explicit" by
+   * `-DUNISTR_FROM_STRING_EXPLICIT=explicit`
+   * on the compiler command line or similar.
+   * @param codepageData an array of bytes, null-terminated,
+   *                     in the platform's default codepage.
+   * @stable ICU 2.0
+   * @see UNICODE_STRING
+   * @see UNICODE_STRING_SIMPLE
+   */
+  UNISTR_FROM_STRING_EXPLICIT UnicodeString(const char *codepageData);
+
+  /**
+   * char* constructor.
+   * Uses the default converter (and thus depends on the ICU conversion code)
+   * unless U_CHARSET_IS_UTF8 is set to 1.
+   * @param codepageData an array of bytes in the platform's default codepage.
+   * @param dataLength The number of bytes in `codepageData`.
+   * @stable ICU 2.0
+   */
+  UnicodeString(const char *codepageData, int32_t dataLength);
+
+#endif
+
+#if !UCONFIG_NO_CONVERSION
+
+  /**
+   * char* constructor.
+   * @param codepageData an array of bytes, null-terminated
+   * @param codepage the encoding of `codepageData`.  The special
+   * value 0 for `codepage` indicates that the text is in the
+   * platform's default codepage.
+   *
+   * If `codepage` is an empty string (`""`),
+   * then a simple conversion is performed on the codepage-invariant
+   * subset ("invariant characters") of the platform encoding. See utypes.h.
+   * Recommendation: For invariant-character strings use the constructor
+   * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
+   * because it avoids object code dependencies of UnicodeString on
+   * the conversion code.
+   *
+   * @stable ICU 2.0
+   */
+  UnicodeString(const char *codepageData, const char *codepage);
+
+  /**
+   * char* constructor.
+   * @param codepageData an array of bytes.
+   * @param dataLength The number of bytes in `codepageData`.
+   * @param codepage the encoding of `codepageData`.  The special
+   * value 0 for `codepage` indicates that the text is in the
+   * platform's default codepage.
+   * If `codepage` is an empty string (`""`),
+   * then a simple conversion is performed on the codepage-invariant
+   * subset ("invariant characters") of the platform encoding. See utypes.h.
+   * Recommendation: For invariant-character strings use the constructor
+   * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
+   * because it avoids object code dependencies of UnicodeString on
+   * the conversion code.
+   *
+   * @stable ICU 2.0
+   */
+  UnicodeString(const char *codepageData, int32_t dataLength, const char *codepage);
+
+  /**
+   * char * / UConverter constructor.
+   * This constructor uses an existing UConverter object to
+   * convert the codepage string to Unicode and construct a UnicodeString
+   * from that.
+   *
+   * The converter is reset at first.
+   * If the error code indicates a failure before this constructor is called,
+   * or if an error occurs during conversion or construction,
+   * then the string will be bogus.
+   *
+   * This function avoids the overhead of opening and closing a converter if
+   * multiple strings are constructed.
+   *
+   * @param src input codepage string
+   * @param srcLength length of the input string, can be -1 for NUL-terminated strings
+   * @param cnv converter object (ucnv_resetToUnicode() will be called),
+   *        can be NULL for the default converter
+   * @param errorCode normal ICU error code
+   * @stable ICU 2.0
+   */
+  UnicodeString(
+        const char *src, int32_t srcLength,
+        UConverter *cnv,
+        UErrorCode &errorCode);
+
+#endif
+
+  /**
+   * Constructs a Unicode string from an invariant-character char * string.
+   * About invariant characters see utypes.h.
+   * This constructor has no runtime dependency on conversion code and is
+   * therefore recommended over ones taking a charset name string
+   * (where the empty string "" indicates invariant-character conversion).
+   *
+   * Use the macro US_INV as the third, signature-distinguishing parameter.
+   *
+   * For example:
+   * \code
+   *     void fn(const char *s) {
+   *       UnicodeString ustr(s, -1, US_INV);
+   *       // use ustr ...
+   *     }
+   * \endcode
+   * @param src String using only invariant characters.
+   * @param textLength Length of src, or -1 if NUL-terminated.
+   * @param inv Signature-distinguishing paramater, use US_INV.
+   *
+   * @see US_INV
+   * @stable ICU 3.2
+   */
+  UnicodeString(const char *src, int32_t textLength, enum EInvariant inv);
+
+
+  /**
+   * Copy constructor.
+   *
+   * Starting with ICU 2.4, the assignment operator and the copy constructor
+   * allocate a new buffer and copy the buffer contents even for readonly aliases.
+   * By contrast, the fastCopyFrom() function implements the old,
+   * more efficient but less safe behavior
+   * of making this string also a readonly alias to the same buffer.
+   *
+   * If the source object has an "open" buffer from getBuffer(minCapacity),
+   * then the copy is an empty string.
+   *
+   * @param that The UnicodeString object to copy.
+   * @stable ICU 2.0
+   * @see fastCopyFrom
+   */
+  UnicodeString(const UnicodeString& that);
+
+  /**
+   * Move constructor; might leave src in bogus state.
+   * This string will have the same contents and state that the source string had.
+   * @param src source string
+   * @stable ICU 56
+   */
+  UnicodeString(UnicodeString &&src) U_NOEXCEPT;
+
+  /**
+   * 'Substring' constructor from tail of source string.
+   * @param src The UnicodeString object to copy.
+   * @param srcStart The offset into `src` at which to start copying.
+   * @stable ICU 2.2
+   */
+  UnicodeString(const UnicodeString& src, int32_t srcStart);
+
+  /**
+   * 'Substring' constructor from subrange of source string.
+   * @param src The UnicodeString object to copy.
+   * @param srcStart The offset into `src` at which to start copying.
+   * @param srcLength The number of characters from `src` to copy.
+   * @stable ICU 2.2
+   */
+  UnicodeString(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
+
+  /**
+   * Clone this object, an instance of a subclass of Replaceable.
+   * Clones can be used concurrently in multiple threads.
+   * If a subclass does not implement clone(), or if an error occurs,
+   * then NULL is returned.
+   * The caller must delete the clone.
+   *
+   * @return a clone of this object
+   *
+   * @see Replaceable::clone
+   * @see getDynamicClassID
+   * @stable ICU 2.6
+   */
+  virtual UnicodeString *clone() const;
+
+  /** Destructor.
+   * @stable ICU 2.0
+   */
+  virtual ~UnicodeString();
+
+  /**
+   * Create a UnicodeString from a UTF-8 string.
+   * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
+   * Calls u_strFromUTF8WithSub().
+   *
+   * @param utf8 UTF-8 input string.
+   *             Note that a StringPiece can be implicitly constructed
+   *             from a std::string or a NUL-terminated const char * string.
+   * @return A UnicodeString with equivalent UTF-16 contents.
+   * @see toUTF8
+   * @see toUTF8String
+   * @stable ICU 4.2
+   */
+  static UnicodeString fromUTF8(StringPiece utf8);
+
+  /**
+   * Create a UnicodeString from a UTF-32 string.
+   * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
+   * Calls u_strFromUTF32WithSub().
+   *
+   * @param utf32 UTF-32 input string. Must not be NULL.
+   * @param length Length of the input string, or -1 if NUL-terminated.
+   * @return A UnicodeString with equivalent UTF-16 contents.
+   * @see toUTF32
+   * @stable ICU 4.2
+   */
+  static UnicodeString fromUTF32(const UChar32 *utf32, int32_t length);
+
+  /* Miscellaneous operations */
+
+  /**
+   * Unescape a string of characters and return a string containing
+   * the result.  The following escape sequences are recognized:
+   *
+   * \\uhhhh       4 hex digits; h in [0-9A-Fa-f]
+   * \\Uhhhhhhhh   8 hex digits
+   * \\xhh         1-2 hex digits
+   * \\ooo         1-3 octal digits; o in [0-7]
+   * \\cX          control-X; X is masked with 0x1F
+   *
+   * as well as the standard ANSI C escapes:
+   *
+   * \\a => U+0007, \\b => U+0008, \\t => U+0009, \\n => U+000A,
+   * \\v => U+000B, \\f => U+000C, \\r => U+000D, \\e => U+001B,
+   * \\" => U+0022, \\' => U+0027, \\? => U+003F, \\\\ => U+005C
+   *
+   * Anything else following a backslash is generically escaped.  For
+   * example, "[a\\-z]" returns "[a-z]".
+   *
+   * If an escape sequence is ill-formed, this method returns an empty
+   * string.  An example of an ill-formed sequence is "\\u" followed by
+   * fewer than 4 hex digits.
+   *
+   * This function is similar to u_unescape() but not identical to it.
+   * The latter takes a source char*, so it does escape recognition
+   * and also invariant conversion.
+   *
+   * @return a string with backslash escapes interpreted, or an
+   * empty string on error.
+   * @see UnicodeString#unescapeAt()
+   * @see u_unescape()
+   * @see u_unescapeAt()
+   * @stable ICU 2.0
+   */
+  UnicodeString unescape() const;
+
+  /**
+   * Unescape a single escape sequence and return the represented
+   * character.  See unescape() for a listing of the recognized escape
+   * sequences.  The character at offset-1 is assumed (without
+   * checking) to be a backslash.  If the escape sequence is
+   * ill-formed, or the offset is out of range, U_SENTINEL=-1 is
+   * returned.
+   *
+   * @param offset an input output parameter.  On input, it is the
+   * offset into this string where the escape sequence is located,
+   * after the initial backslash.  On output, it is advanced after the
+   * last character parsed.  On error, it is not advanced at all.
+   * @return the character represented by the escape sequence at
+   * offset, or U_SENTINEL=-1 on error.
+   * @see UnicodeString#unescape()
+   * @see u_unescape()
+   * @see u_unescapeAt()
+   * @stable ICU 2.0
+   */
+  UChar32 unescapeAt(int32_t &offset) const;
+
+  /**
+   * ICU "poor man's RTTI", returns a UClassID for this class.
+   *
+   * @stable ICU 2.2
+   */
+  static UClassID U_EXPORT2 getStaticClassID();
+
+  /**
+   * ICU "poor man's RTTI", returns a UClassID for the actual class.
+   *
+   * @stable ICU 2.2
+   */
+  virtual UClassID getDynamicClassID() const;
+
+  //========================================
+  // Implementation methods
+  //========================================
+
+protected:
+  /**
+   * Implement Replaceable::getLength() (see jitterbug 1027).
+   * @stable ICU 2.4
+   */
+  virtual int32_t getLength() const;
+
+  /**
+   * The change in Replaceable to use virtual getCharAt() allows
+   * UnicodeString::charAt() to be inline again (see jitterbug 709).
+   * @stable ICU 2.4
+   */
+  virtual char16_t getCharAt(int32_t offset) const;
+
+  /**
+   * The change in Replaceable to use virtual getChar32At() allows
+   * UnicodeString::char32At() to be inline again (see jitterbug 709).
+   * @stable ICU 2.4
+   */
+  virtual UChar32 getChar32At(int32_t offset) const;
+
+private:
+  // For char* constructors. Could be made public.
+  UnicodeString &setToUTF8(StringPiece utf8);
+  // For extract(char*).
+  // We could make a toUTF8(target, capacity, errorCode) public but not
+  // this version: New API will be cleaner if we make callers create substrings
+  // rather than having start+length on every method,
+  // and it should take a UErrorCode&.
+  int32_t
+  toUTF8(int32_t start, int32_t len,
+         char *target, int32_t capacity) const;
+
+  /**
+   * Internal string contents comparison, called by operator==.
+   * Requires: this & text not bogus and have same lengths.
+   */
+  UBool doEquals(const UnicodeString &text, int32_t len) const;
+
+  inline int8_t
+  doCompare(int32_t start,
+           int32_t length,
+           const UnicodeString& srcText,
+           int32_t srcStart,
+           int32_t srcLength) const;
+
+  int8_t doCompare(int32_t start,
+           int32_t length,
+           const char16_t *srcChars,
+           int32_t srcStart,
+           int32_t srcLength) const;
+
+  inline int8_t
+  doCompareCodePointOrder(int32_t start,
+                          int32_t length,
+                          const UnicodeString& srcText,
+                          int32_t srcStart,
+                          int32_t srcLength) const;
+
+  int8_t doCompareCodePointOrder(int32_t start,
+                                 int32_t length,
+                                 const char16_t *srcChars,
+                                 int32_t srcStart,
+                                 int32_t srcLength) const;
+
+  inline int8_t
+  doCaseCompare(int32_t start,
+                int32_t length,
+                const UnicodeString &srcText,
+                int32_t srcStart,
+                int32_t srcLength,
+                uint32_t options) const;
+
+  int8_t
+  doCaseCompare(int32_t start,
+                int32_t length,
+                const char16_t *srcChars,
+                int32_t srcStart,
+                int32_t srcLength,
+                uint32_t options) const;
+
+  int32_t doIndexOf(char16_t c,
+            int32_t start,
+            int32_t length) const;
+
+  int32_t doIndexOf(UChar32 c,
+                        int32_t start,
+                        int32_t length) const;
+
+  int32_t doLastIndexOf(char16_t c,
+                int32_t start,
+                int32_t length) const;
+
+  int32_t doLastIndexOf(UChar32 c,
+                            int32_t start,
+                            int32_t length) const;
+
+  void doExtract(int32_t start,
+         int32_t length,
+         char16_t *dst,
+         int32_t dstStart) const;
+
+  inline void doExtract(int32_t start,
+         int32_t length,
+         UnicodeString& target) const;
+
+  inline char16_t doCharAt(int32_t offset)  const;
+
+  UnicodeString& doReplace(int32_t start,
+               int32_t length,
+               const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength);
+
+  UnicodeString& doReplace(int32_t start,
+               int32_t length,
+               const char16_t *srcChars,
+               int32_t srcStart,
+               int32_t srcLength);
+
+  UnicodeString& doAppend(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
+  UnicodeString& doAppend(const char16_t *srcChars, int32_t srcStart, int32_t srcLength);
+
+  UnicodeString& doReverse(int32_t start,
+               int32_t length);
+
+  // calculate hash code
+  int32_t doHashCode(void) const;
+
+  // get pointer to start of array
+  // these do not check for kOpenGetBuffer, unlike the public getBuffer() function
+  inline char16_t* getArrayStart(void);
+  inline const char16_t* getArrayStart(void) const;
+
+  inline UBool hasShortLength() const;
+  inline int32_t getShortLength() const;
+
+  // A UnicodeString object (not necessarily its current buffer)
+  // is writable unless it isBogus() or it has an "open" getBuffer(minCapacity).
+  inline UBool isWritable() const;
+
+  // Is the current buffer writable?
+  inline UBool isBufferWritable() const;
+
+  // None of the following does releaseArray().
+  inline void setZeroLength();
+  inline void setShortLength(int32_t len);
+  inline void setLength(int32_t len);
+  inline void setToEmpty();
+  inline void setArray(char16_t *array, int32_t len, int32_t capacity); // sets length but not flags
+
+  // allocate the array; result may be the stack buffer
+  // sets refCount to 1 if appropriate
+  // sets fArray, fCapacity, and flags
+  // sets length to 0
+  // returns boolean for success or failure
+  UBool allocate(int32_t capacity);
+
+  // release the array if owned
+  void releaseArray(void);
+
+  // turn a bogus string into an empty one
+  void unBogus();
+
+  // implements assigment operator, copy constructor, and fastCopyFrom()
+  UnicodeString &copyFrom(const UnicodeString &src, UBool fastCopy=false);
+
+  // Copies just the fields without memory management.
+  void copyFieldsFrom(UnicodeString &src, UBool setSrcToBogus) U_NOEXCEPT;
+
+  // Pin start and limit to acceptable values.
+  inline void pinIndex(int32_t& start) const;
+  inline void pinIndices(int32_t& start,
+                         int32_t& length) const;
+
+#if !UCONFIG_NO_CONVERSION
+
+  /* Internal extract() using UConverter. */
+  int32_t doExtract(int32_t start, int32_t length,
+                    char *dest, int32_t destCapacity,
+                    UConverter *cnv,
+                    UErrorCode &errorCode) const;
+
+  /*
+   * Real constructor for converting from codepage data.
+   * It assumes that it is called with !fRefCounted.
+   *
+   * If `codepage==0`, then the default converter
+   * is used for the platform encoding.
+   * If `codepage` is an empty string (`""`),
+   * then a simple conversion is performed on the codepage-invariant
+   * subset ("invariant characters") of the platform encoding. See utypes.h.
+   */
+  void doCodepageCreate(const char *codepageData,
+                        int32_t dataLength,
+                        const char *codepage);
+
+  /*
+   * Worker function for creating a UnicodeString from
+   * a codepage string using a UConverter.
+   */
+  void
+  doCodepageCreate(const char *codepageData,
+                   int32_t dataLength,
+                   UConverter *converter,
+                   UErrorCode &status);
+
+#endif
+
+  /*
+   * This function is called when write access to the array
+   * is necessary.
+   *
+   * We need to make a copy of the array if
+   * the buffer is read-only, or
+   * the buffer is refCounted (shared), and refCount>1, or
+   * the buffer is too small.
+   *
+   * Return false if memory could not be allocated.
+   */
+  UBool cloneArrayIfNeeded(int32_t newCapacity = -1,
+                            int32_t growCapacity = -1,
+                            UBool doCopyArray = true,
+                            int32_t **pBufferToDelete = 0,
+                            UBool forceClone = false);
+
+  /**
+   * Common function for UnicodeString case mappings.
+   * The stringCaseMapper has the same type UStringCaseMapper
+   * as in ustr_imp.h for ustrcase_map().
+   */
+  UnicodeString &
+  caseMap(int32_t caseLocale, uint32_t options,
+#if !UCONFIG_NO_BREAK_ITERATION
+          BreakIterator *iter,
+#endif
+          UStringCaseMapper *stringCaseMapper);
+
+  // ref counting
+  void addRef(void);
+  int32_t removeRef(void);
+  int32_t refCount(void) const;
+
+  // constants
+  enum {
+    /**
+     * Size of stack buffer for short strings.
+     * Must be at least U16_MAX_LENGTH for the single-code point constructor to work.
+     * @see UNISTR_OBJECT_SIZE
+     */
+    US_STACKBUF_SIZE=(int32_t)(UNISTR_OBJECT_SIZE-sizeof(void *)-2)/U_SIZEOF_UCHAR,
+    kInvalidUChar=0xffff, // U+FFFF returned by charAt(invalid index)
+    kInvalidHashCode=0, // invalid hash code
+    kEmptyHashCode=1, // hash code for empty string
+
+    // bit flag values for fLengthAndFlags
+    kIsBogus=1,         // this string is bogus, i.e., not valid or NULL
+    kUsingStackBuffer=2,// using fUnion.fStackFields instead of fUnion.fFields
+    kRefCounted=4,      // there is a refCount field before the characters in fArray
+    kBufferIsReadonly=8,// do not write to this buffer
+    kOpenGetBuffer=16,  // getBuffer(minCapacity) was called (is "open"),
+                        // and releaseBuffer(newLength) must be called
+    kAllStorageFlags=0x1f,
+
+    kLengthShift=5,     // remaining 11 bits for non-negative short length, or negative if long
+    kLength1=1<<kLengthShift,
+    kMaxShortLength=0x3ff,  // max non-negative short length (leaves top bit 0)
+    kLengthIsLarge=0xffe0,  // short length < 0, real length is in fUnion.fFields.fLength
+
+    // combined values for convenience
+    kShortString=kUsingStackBuffer,
+    kLongString=kRefCounted,
+    kReadonlyAlias=kBufferIsReadonly,
+    kWritableAlias=0
+  };
+
+  friend class UnicodeStringAppendable;
+
+  union StackBufferOrFields;        // forward declaration necessary before friend declaration
+  friend union StackBufferOrFields; // make US_STACKBUF_SIZE visible inside fUnion
+
+  /*
+   * The following are all the class fields that are stored
+   * in each UnicodeString object.
+   * Note that UnicodeString has virtual functions,
+   * therefore there is an implicit vtable pointer
+   * as the first real field.
+   * The fields should be aligned such that no padding is necessary.
+   * On 32-bit machines, the size should be 32 bytes,
+   * on 64-bit machines (8-byte pointers), it should be 40 bytes.
+   *
+   * We use a hack to achieve this.
+   *
+   * With at least some compilers, each of the following is forced to
+   * a multiple of sizeof(pointer) [the largest field base unit here is a data pointer],
+   * rounded up with additional padding if the fields do not already fit that requirement:
+   * - sizeof(class UnicodeString)
+   * - offsetof(UnicodeString, fUnion)
+   * - sizeof(fUnion)
+   * - sizeof(fStackFields)
+   *
+   * We optimize for the longest possible internal buffer for short strings.
+   * fUnion.fStackFields begins with 2 bytes for storage flags
+   * and the length of relatively short strings,
+   * followed by the buffer for short string contents.
+   * There is no padding inside fStackFields.
+   *
+   * Heap-allocated and aliased strings use fUnion.fFields.
+   * Both fStackFields and fFields must begin with the same fields for flags and short length,
+   * that is, those must have the same memory offsets inside the object,
+   * because the flags must be inspected in order to decide which half of fUnion is being used.
+   * We assume that the compiler does not reorder the fields.
+   *
+   * (Padding at the end of fFields is ok:
+   * As long as it is no larger than fStackFields, it is not wasted space.)
+   *
+   * For some of the history of the UnicodeString class fields layout, see
+   * - ICU ticket #11551 "longer UnicodeString contents in stack buffer"
+   * - ICU ticket #11336 "UnicodeString: recombine stack buffer arrays"
+   * - ICU ticket #8322 "why is sizeof(UnicodeString)==48?"
+   */
+  // (implicit) *vtable;
+  union StackBufferOrFields {
+    // fStackFields is used iff (fLengthAndFlags&kUsingStackBuffer) else fFields is used.
+    // Each struct of the union must begin with fLengthAndFlags.
+    struct {
+      int16_t fLengthAndFlags;          // bit fields: see constants above
+      char16_t fBuffer[US_STACKBUF_SIZE];  // buffer for short strings
+    } fStackFields;
+    struct {
+      int16_t fLengthAndFlags;          // bit fields: see constants above
+      int32_t fLength;    // number of characters in fArray if >127; else undefined
+      int32_t fCapacity;  // capacity of fArray (in char16_ts)
+      // array pointer last to minimize padding for machines with P128 data model
+      // or pointer sizes that are not a power of 2
+      char16_t   *fArray;    // the Unicode data
+    } fFields;
+  } fUnion;
+};
+
+/**
+ * Create a new UnicodeString with the concatenation of two others.
+ *
+ * @param s1 The first string to be copied to the new one.
+ * @param s2 The second string to be copied to the new one, after s1.
+ * @return UnicodeString(s1).append(s2)
+ * @stable ICU 2.8
+ */
+U_COMMON_API UnicodeString U_EXPORT2
+operator+ (const UnicodeString &s1, const UnicodeString &s2);
+
+//========================================
+// Inline members
+//========================================
+
+//========================================
+// Privates
+//========================================
+
+inline void
+UnicodeString::pinIndex(int32_t& start) const
+{
+  // pin index
+  if(start < 0) {
+    start = 0;
+  } else if(start > length()) {
+    start = length();
+  }
+}
+
+inline void
+UnicodeString::pinIndices(int32_t& start,
+                          int32_t& _length) const
+{
+  // pin indices
+  int32_t len = length();
+  if(start < 0) {
+    start = 0;
+  } else if(start > len) {
+    start = len;
+  }
+  if(_length < 0) {
+    _length = 0;
+  } else if(_length > (len - start)) {
+    _length = (len - start);
+  }
+}
+
+inline char16_t*
+UnicodeString::getArrayStart() {
+  return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
+    fUnion.fStackFields.fBuffer : fUnion.fFields.fArray;
+}
+
+inline const char16_t*
+UnicodeString::getArrayStart() const {
+  return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
+    fUnion.fStackFields.fBuffer : fUnion.fFields.fArray;
+}
+
+//========================================
+// Default constructor
+//========================================
+
+inline
+UnicodeString::UnicodeString() {
+  fUnion.fStackFields.fLengthAndFlags=kShortString;
+}
+
+inline UnicodeString::UnicodeString(const std::nullptr_t /*text*/) {
+  fUnion.fStackFields.fLengthAndFlags=kShortString;
+}
+
+inline UnicodeString::UnicodeString(const std::nullptr_t /*text*/, int32_t /*length*/) {
+  fUnion.fStackFields.fLengthAndFlags=kShortString;
+}
+
+inline UnicodeString::UnicodeString(std::nullptr_t /*buffer*/, int32_t /*buffLength*/, int32_t /*buffCapacity*/) {
+  fUnion.fStackFields.fLengthAndFlags=kShortString;
+}
+
+//========================================
+// Read-only implementation methods
+//========================================
+inline UBool
+UnicodeString::hasShortLength() const {
+  return fUnion.fFields.fLengthAndFlags>=0;
+}
+
+inline int32_t
+UnicodeString::getShortLength() const {
+  // fLengthAndFlags must be non-negative -> short length >= 0
+  // and arithmetic or logical shift does not matter.
+  return fUnion.fFields.fLengthAndFlags>>kLengthShift;
+}
+
+inline int32_t
+UnicodeString::length() const {
+  return hasShortLength() ? getShortLength() : fUnion.fFields.fLength;
+}
+
+inline int32_t
+UnicodeString::getCapacity() const {
+  return (fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) ?
+    US_STACKBUF_SIZE : fUnion.fFields.fCapacity;
+}
+
+inline int32_t
+UnicodeString::hashCode() const
+{ return doHashCode(); }
+
+inline UBool
+UnicodeString::isBogus() const
+{ return (UBool)(fUnion.fFields.fLengthAndFlags & kIsBogus); }
+
+inline UBool
+UnicodeString::isWritable() const
+{ return (UBool)!(fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kIsBogus)); }
+
+inline UBool
+UnicodeString::isBufferWritable() const
+{
+  return (UBool)(
+      !(fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kIsBogus|kBufferIsReadonly)) &&
+      (!(fUnion.fFields.fLengthAndFlags&kRefCounted) || refCount()==1));
+}
+
+inline const char16_t *
+UnicodeString::getBuffer() const {
+  if(fUnion.fFields.fLengthAndFlags&(kIsBogus|kOpenGetBuffer)) {
+    return nullptr;
+  } else if(fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) {
+    return fUnion.fStackFields.fBuffer;
+  } else {
+    return fUnion.fFields.fArray;
+  }
+}
+
+//========================================
+// Read-only alias methods
+//========================================
+inline int8_t
+UnicodeString::doCompare(int32_t start,
+              int32_t thisLength,
+              const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength) const
+{
+  if(srcText.isBogus()) {
+    return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
+  } else {
+    srcText.pinIndices(srcStart, srcLength);
+    return doCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
+  }
+}
+
+inline UBool
+UnicodeString::operator== (const UnicodeString& text) const
+{
+  if(isBogus()) {
+    return text.isBogus();
+  } else {
+    int32_t len = length(), textLength = text.length();
+    return !text.isBogus() && len == textLength && doEquals(text, len);
+  }
+}
+
+inline UBool
+UnicodeString::operator!= (const UnicodeString& text) const
+{ return (! operator==(text)); }
+
+inline UBool
+UnicodeString::operator> (const UnicodeString& text) const
+{ return doCompare(0, length(), text, 0, text.length()) == 1; }
+
+inline UBool
+UnicodeString::operator< (const UnicodeString& text) const
+{ return doCompare(0, length(), text, 0, text.length()) == -1; }
+
+inline UBool
+UnicodeString::operator>= (const UnicodeString& text) const
+{ return doCompare(0, length(), text, 0, text.length()) != -1; }
+
+inline UBool
+UnicodeString::operator<= (const UnicodeString& text) const
+{ return doCompare(0, length(), text, 0, text.length()) != 1; }
+
+inline int8_t
+UnicodeString::compare(const UnicodeString& text) const
+{ return doCompare(0, length(), text, 0, text.length()); }
+
+inline int8_t
+UnicodeString::compare(int32_t start,
+               int32_t _length,
+               const UnicodeString& srcText) const
+{ return doCompare(start, _length, srcText, 0, srcText.length()); }
+
+inline int8_t
+UnicodeString::compare(ConstChar16Ptr srcChars,
+               int32_t srcLength) const
+{ return doCompare(0, length(), srcChars, 0, srcLength); }
+
+inline int8_t
+UnicodeString::compare(int32_t start,
+               int32_t _length,
+               const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength) const
+{ return doCompare(start, _length, srcText, srcStart, srcLength); }
+
+inline int8_t
+UnicodeString::compare(int32_t start,
+               int32_t _length,
+               const char16_t *srcChars) const
+{ return doCompare(start, _length, srcChars, 0, _length); }
+
+inline int8_t
+UnicodeString::compare(int32_t start,
+               int32_t _length,
+               const char16_t *srcChars,
+               int32_t srcStart,
+               int32_t srcLength) const
+{ return doCompare(start, _length, srcChars, srcStart, srcLength); }
+
+inline int8_t
+UnicodeString::compareBetween(int32_t start,
+                  int32_t limit,
+                  const UnicodeString& srcText,
+                  int32_t srcStart,
+                  int32_t srcLimit) const
+{ return doCompare(start, limit - start,
+           srcText, srcStart, srcLimit - srcStart); }
+
+inline int8_t
+UnicodeString::doCompareCodePointOrder(int32_t start,
+                                       int32_t thisLength,
+                                       const UnicodeString& srcText,
+                                       int32_t srcStart,
+                                       int32_t srcLength) const
+{
+  if(srcText.isBogus()) {
+    return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
+  } else {
+    srcText.pinIndices(srcStart, srcLength);
+    return doCompareCodePointOrder(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
+  }
+}
+
+inline int8_t
+UnicodeString::compareCodePointOrder(const UnicodeString& text) const
+{ return doCompareCodePointOrder(0, length(), text, 0, text.length()); }
+
+inline int8_t
+UnicodeString::compareCodePointOrder(int32_t start,
+                                     int32_t _length,
+                                     const UnicodeString& srcText) const
+{ return doCompareCodePointOrder(start, _length, srcText, 0, srcText.length()); }
+
+inline int8_t
+UnicodeString::compareCodePointOrder(ConstChar16Ptr srcChars,
+                                     int32_t srcLength) const
+{ return doCompareCodePointOrder(0, length(), srcChars, 0, srcLength); }
+
+inline int8_t
+UnicodeString::compareCodePointOrder(int32_t start,
+                                     int32_t _length,
+                                     const UnicodeString& srcText,
+                                     int32_t srcStart,
+                                     int32_t srcLength) const
+{ return doCompareCodePointOrder(start, _length, srcText, srcStart, srcLength); }
+
+inline int8_t
+UnicodeString::compareCodePointOrder(int32_t start,
+                                     int32_t _length,
+                                     const char16_t *srcChars) const
+{ return doCompareCodePointOrder(start, _length, srcChars, 0, _length); }
+
+inline int8_t
+UnicodeString::compareCodePointOrder(int32_t start,
+                                     int32_t _length,
+                                     const char16_t *srcChars,
+                                     int32_t srcStart,
+                                     int32_t srcLength) const
+{ return doCompareCodePointOrder(start, _length, srcChars, srcStart, srcLength); }
+
+inline int8_t
+UnicodeString::compareCodePointOrderBetween(int32_t start,
+                                            int32_t limit,
+                                            const UnicodeString& srcText,
+                                            int32_t srcStart,
+                                            int32_t srcLimit) const
+{ return doCompareCodePointOrder(start, limit - start,
+           srcText, srcStart, srcLimit - srcStart); }
+
+inline int8_t
+UnicodeString::doCaseCompare(int32_t start,
+                             int32_t thisLength,
+                             const UnicodeString &srcText,
+                             int32_t srcStart,
+                             int32_t srcLength,
+                             uint32_t options) const
+{
+  if(srcText.isBogus()) {
+    return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
+  } else {
+    srcText.pinIndices(srcStart, srcLength);
+    return doCaseCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength, options);
+  }
+}
+
+inline int8_t
+UnicodeString::caseCompare(const UnicodeString &text, uint32_t options) const {
+  return doCaseCompare(0, length(), text, 0, text.length(), options);
+}
+
+inline int8_t
+UnicodeString::caseCompare(int32_t start,
+                           int32_t _length,
+                           const UnicodeString &srcText,
+                           uint32_t options) const {
+  return doCaseCompare(start, _length, srcText, 0, srcText.length(), options);
+}
+
+inline int8_t
+UnicodeString::caseCompare(ConstChar16Ptr srcChars,
+                           int32_t srcLength,
+                           uint32_t options) const {
+  return doCaseCompare(0, length(), srcChars, 0, srcLength, options);
+}
+
+inline int8_t
+UnicodeString::caseCompare(int32_t start,
+                           int32_t _length,
+                           const UnicodeString &srcText,
+                           int32_t srcStart,
+                           int32_t srcLength,
+                           uint32_t options) const {
+  return doCaseCompare(start, _length, srcText, srcStart, srcLength, options);
+}
+
+inline int8_t
+UnicodeString::caseCompare(int32_t start,
+                           int32_t _length,
+                           const char16_t *srcChars,
+                           uint32_t options) const {
+  return doCaseCompare(start, _length, srcChars, 0, _length, options);
+}
+
+inline int8_t
+UnicodeString::caseCompare(int32_t start,
+                           int32_t _length,
+                           const char16_t *srcChars,
+                           int32_t srcStart,
+                           int32_t srcLength,
+                           uint32_t options) const {
+  return doCaseCompare(start, _length, srcChars, srcStart, srcLength, options);
+}
+
+inline int8_t
+UnicodeString::caseCompareBetween(int32_t start,
+                                  int32_t limit,
+                                  const UnicodeString &srcText,
+                                  int32_t srcStart,
+                                  int32_t srcLimit,
+                                  uint32_t options) const {
+  return doCaseCompare(start, limit - start, srcText, srcStart, srcLimit - srcStart, options);
+}
+
+inline int32_t
+UnicodeString::indexOf(const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength,
+               int32_t start,
+               int32_t _length) const
+{
+  if(!srcText.isBogus()) {
+    srcText.pinIndices(srcStart, srcLength);
+    if(srcLength > 0) {
+      return indexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
+    }
+  }
+  return -1;
+}
+
+inline int32_t
+UnicodeString::indexOf(const UnicodeString& text) const
+{ return indexOf(text, 0, text.length(), 0, length()); }
+
+inline int32_t
+UnicodeString::indexOf(const UnicodeString& text,
+               int32_t start) const {
+  pinIndex(start);
+  return indexOf(text, 0, text.length(), start, length() - start);
+}
+
+inline int32_t
+UnicodeString::indexOf(const UnicodeString& text,
+               int32_t start,
+               int32_t _length) const
+{ return indexOf(text, 0, text.length(), start, _length); }
+
+inline int32_t
+UnicodeString::indexOf(const char16_t *srcChars,
+               int32_t srcLength,
+               int32_t start) const {
+  pinIndex(start);
+  return indexOf(srcChars, 0, srcLength, start, length() - start);
+}
+
+inline int32_t
+UnicodeString::indexOf(ConstChar16Ptr srcChars,
+               int32_t srcLength,
+               int32_t start,
+               int32_t _length) const
+{ return indexOf(srcChars, 0, srcLength, start, _length); }
+
+inline int32_t
+UnicodeString::indexOf(char16_t c,
+               int32_t start,
+               int32_t _length) const
+{ return doIndexOf(c, start, _length); }
+
+inline int32_t
+UnicodeString::indexOf(UChar32 c,
+               int32_t start,
+               int32_t _length) const
+{ return doIndexOf(c, start, _length); }
+
+inline int32_t
+UnicodeString::indexOf(char16_t c) const
+{ return doIndexOf(c, 0, length()); }
+
+inline int32_t
+UnicodeString::indexOf(UChar32 c) const
+{ return indexOf(c, 0, length()); }
+
+inline int32_t
+UnicodeString::indexOf(char16_t c,
+               int32_t start) const {
+  pinIndex(start);
+  return doIndexOf(c, start, length() - start);
+}
+
+inline int32_t
+UnicodeString::indexOf(UChar32 c,
+               int32_t start) const {
+  pinIndex(start);
+  return indexOf(c, start, length() - start);
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(ConstChar16Ptr srcChars,
+               int32_t srcLength,
+               int32_t start,
+               int32_t _length) const
+{ return lastIndexOf(srcChars, 0, srcLength, start, _length); }
+
+inline int32_t
+UnicodeString::lastIndexOf(const char16_t *srcChars,
+               int32_t srcLength,
+               int32_t start) const {
+  pinIndex(start);
+  return lastIndexOf(srcChars, 0, srcLength, start, length() - start);
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength,
+               int32_t start,
+               int32_t _length) const
+{
+  if(!srcText.isBogus()) {
+    srcText.pinIndices(srcStart, srcLength);
+    if(srcLength > 0) {
+      return lastIndexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
+    }
+  }
+  return -1;
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(const UnicodeString& text,
+               int32_t start,
+               int32_t _length) const
+{ return lastIndexOf(text, 0, text.length(), start, _length); }
+
+inline int32_t
+UnicodeString::lastIndexOf(const UnicodeString& text,
+               int32_t start) const {
+  pinIndex(start);
+  return lastIndexOf(text, 0, text.length(), start, length() - start);
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(const UnicodeString& text) const
+{ return lastIndexOf(text, 0, text.length(), 0, length()); }
+
+inline int32_t
+UnicodeString::lastIndexOf(char16_t c,
+               int32_t start,
+               int32_t _length) const
+{ return doLastIndexOf(c, start, _length); }
+
+inline int32_t
+UnicodeString::lastIndexOf(UChar32 c,
+               int32_t start,
+               int32_t _length) const {
+  return doLastIndexOf(c, start, _length);
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(char16_t c) const
+{ return doLastIndexOf(c, 0, length()); }
+
+inline int32_t
+UnicodeString::lastIndexOf(UChar32 c) const {
+  return lastIndexOf(c, 0, length());
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(char16_t c,
+               int32_t start) const {
+  pinIndex(start);
+  return doLastIndexOf(c, start, length() - start);
+}
+
+inline int32_t
+UnicodeString::lastIndexOf(UChar32 c,
+               int32_t start) const {
+  pinIndex(start);
+  return lastIndexOf(c, start, length() - start);
+}
+
+inline UBool
+UnicodeString::startsWith(const UnicodeString& text) const
+{ return compare(0, text.length(), text, 0, text.length()) == 0; }
+
+inline UBool
+UnicodeString::startsWith(const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength) const
+{ return doCompare(0, srcLength, srcText, srcStart, srcLength) == 0; }
+
+inline UBool
+UnicodeString::startsWith(ConstChar16Ptr srcChars, int32_t srcLength) const {
+  if(srcLength < 0) {
+    srcLength = u_strlen(toUCharPtr(srcChars));
+  }
+  return doCompare(0, srcLength, srcChars, 0, srcLength) == 0;
+}
+
+inline UBool
+UnicodeString::startsWith(const char16_t *srcChars, int32_t srcStart, int32_t srcLength) const {
+  if(srcLength < 0) {
+    srcLength = u_strlen(toUCharPtr(srcChars));
+  }
+  return doCompare(0, srcLength, srcChars, srcStart, srcLength) == 0;
+}
+
+inline UBool
+UnicodeString::endsWith(const UnicodeString& text) const
+{ return doCompare(length() - text.length(), text.length(),
+           text, 0, text.length()) == 0; }
+
+inline UBool
+UnicodeString::endsWith(const UnicodeString& srcText,
+            int32_t srcStart,
+            int32_t srcLength) const {
+  srcText.pinIndices(srcStart, srcLength);
+  return doCompare(length() - srcLength, srcLength,
+                   srcText, srcStart, srcLength) == 0;
+}
+
+inline UBool
+UnicodeString::endsWith(ConstChar16Ptr srcChars,
+            int32_t srcLength) const {
+  if(srcLength < 0) {
+    srcLength = u_strlen(toUCharPtr(srcChars));
+  }
+  return doCompare(length() - srcLength, srcLength,
+                   srcChars, 0, srcLength) == 0;
+}
+
+inline UBool
+UnicodeString::endsWith(const char16_t *srcChars,
+            int32_t srcStart,
+            int32_t srcLength) const {
+  if(srcLength < 0) {
+    srcLength = u_strlen(toUCharPtr(srcChars + srcStart));
+  }
+  return doCompare(length() - srcLength, srcLength,
+                   srcChars, srcStart, srcLength) == 0;
+}
+
+//========================================
+// replace
+//========================================
+inline UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               const UnicodeString& srcText)
+{ return doReplace(start, _length, srcText, 0, srcText.length()); }
+
+inline UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               const UnicodeString& srcText,
+               int32_t srcStart,
+               int32_t srcLength)
+{ return doReplace(start, _length, srcText, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               ConstChar16Ptr srcChars,
+               int32_t srcLength)
+{ return doReplace(start, _length, srcChars, 0, srcLength); }
+
+inline UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               const char16_t *srcChars,
+               int32_t srcStart,
+               int32_t srcLength)
+{ return doReplace(start, _length, srcChars, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               char16_t srcChar)
+{ return doReplace(start, _length, &srcChar, 0, 1); }
+
+inline UnicodeString&
+UnicodeString::replaceBetween(int32_t start,
+                  int32_t limit,
+                  const UnicodeString& srcText)
+{ return doReplace(start, limit - start, srcText, 0, srcText.length()); }
+
+inline UnicodeString&
+UnicodeString::replaceBetween(int32_t start,
+                  int32_t limit,
+                  const UnicodeString& srcText,
+                  int32_t srcStart,
+                  int32_t srcLimit)
+{ return doReplace(start, limit - start, srcText, srcStart, srcLimit - srcStart); }
+
+inline UnicodeString&
+UnicodeString::findAndReplace(const UnicodeString& oldText,
+                  const UnicodeString& newText)
+{ return findAndReplace(0, length(), oldText, 0, oldText.length(),
+            newText, 0, newText.length()); }
+
+inline UnicodeString&
+UnicodeString::findAndReplace(int32_t start,
+                  int32_t _length,
+                  const UnicodeString& oldText,
+                  const UnicodeString& newText)
+{ return findAndReplace(start, _length, oldText, 0, oldText.length(),
+            newText, 0, newText.length()); }
+
+// ============================
+// extract
+// ============================
+inline void
+UnicodeString::doExtract(int32_t start,
+             int32_t _length,
+             UnicodeString& target) const
+{ target.replace(0, target.length(), *this, start, _length); }
+
+inline void
+UnicodeString::extract(int32_t start,
+               int32_t _length,
+               Char16Ptr target,
+               int32_t targetStart) const
+{ doExtract(start, _length, target, targetStart); }
+
+inline void
+UnicodeString::extract(int32_t start,
+               int32_t _length,
+               UnicodeString& target) const
+{ doExtract(start, _length, target); }
+
+#if !UCONFIG_NO_CONVERSION
+
+inline int32_t
+UnicodeString::extract(int32_t start,
+               int32_t _length,
+               char *dst,
+               const char *codepage) const
+
+{
+  // This dstSize value will be checked explicitly
+  return extract(start, _length, dst, dst!=0 ? 0xffffffff : 0, codepage);
+}
+
+#endif
+
+inline void
+UnicodeString::extractBetween(int32_t start,
+                  int32_t limit,
+                  char16_t *dst,
+                  int32_t dstStart) const {
+  pinIndex(start);
+  pinIndex(limit);
+  doExtract(start, limit - start, dst, dstStart);
+}
+
+inline UnicodeString
+UnicodeString::tempSubStringBetween(int32_t start, int32_t limit) const {
+    return tempSubString(start, limit - start);
+}
+
+inline char16_t
+UnicodeString::doCharAt(int32_t offset) const
+{
+  if((uint32_t)offset < (uint32_t)length()) {
+    return getArrayStart()[offset];
+  } else {
+    return kInvalidUChar;
+  }
+}
+
+inline char16_t
+UnicodeString::charAt(int32_t offset) const
+{ return doCharAt(offset); }
+
+inline char16_t
+UnicodeString::operator[] (int32_t offset) const
+{ return doCharAt(offset); }
+
+inline UBool
+UnicodeString::isEmpty() const {
+  // Arithmetic or logical right shift does not matter: only testing for 0.
+  return (fUnion.fFields.fLengthAndFlags>>kLengthShift) == 0;
+}
+
+//========================================
+// Write implementation methods
+//========================================
+inline void
+UnicodeString::setZeroLength() {
+  fUnion.fFields.fLengthAndFlags &= kAllStorageFlags;
+}
+
+inline void
+UnicodeString::setShortLength(int32_t len) {
+  // requires 0 <= len <= kMaxShortLength
+  fUnion.fFields.fLengthAndFlags =
+    (int16_t)((fUnion.fFields.fLengthAndFlags & kAllStorageFlags) | (len << kLengthShift));
+}
+
+inline void
+UnicodeString::setLength(int32_t len) {
+  if(len <= kMaxShortLength) {
+    setShortLength(len);
+  } else {
+    fUnion.fFields.fLengthAndFlags |= kLengthIsLarge;
+    fUnion.fFields.fLength = len;
+  }
+}
+
+inline void
+UnicodeString::setToEmpty() {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+}
+
+inline void
+UnicodeString::setArray(char16_t *array, int32_t len, int32_t capacity) {
+  setLength(len);
+  fUnion.fFields.fArray = array;
+  fUnion.fFields.fCapacity = capacity;
+}
+
+inline UnicodeString&
+UnicodeString::operator= (char16_t ch)
+{ return doReplace(0, length(), &ch, 0, 1); }
+
+inline UnicodeString&
+UnicodeString::operator= (UChar32 ch)
+{ return replace(0, length(), ch); }
+
+inline UnicodeString&
+UnicodeString::setTo(const UnicodeString& srcText,
+             int32_t srcStart,
+             int32_t srcLength)
+{
+  unBogus();
+  return doReplace(0, length(), srcText, srcStart, srcLength);
+}
+
+inline UnicodeString&
+UnicodeString::setTo(const UnicodeString& srcText,
+             int32_t srcStart)
+{
+  unBogus();
+  srcText.pinIndex(srcStart);
+  return doReplace(0, length(), srcText, srcStart, srcText.length() - srcStart);
+}
+
+inline UnicodeString&
+UnicodeString::setTo(const UnicodeString& srcText)
+{
+  return copyFrom(srcText);
+}
+
+inline UnicodeString&
+UnicodeString::setTo(const char16_t *srcChars,
+             int32_t srcLength)
+{
+  unBogus();
+  return doReplace(0, length(), srcChars, 0, srcLength);
+}
+
+inline UnicodeString&
+UnicodeString::setTo(char16_t srcChar)
+{
+  unBogus();
+  return doReplace(0, length(), &srcChar, 0, 1);
+}
+
+inline UnicodeString&
+UnicodeString::setTo(UChar32 srcChar)
+{
+  unBogus();
+  return replace(0, length(), srcChar);
+}
+
+inline UnicodeString&
+UnicodeString::append(const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength)
+{ return doAppend(srcText, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::append(const UnicodeString& srcText)
+{ return doAppend(srcText, 0, srcText.length()); }
+
+inline UnicodeString&
+UnicodeString::append(const char16_t *srcChars,
+              int32_t srcStart,
+              int32_t srcLength)
+{ return doAppend(srcChars, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::append(ConstChar16Ptr srcChars,
+              int32_t srcLength)
+{ return doAppend(srcChars, 0, srcLength); }
+
+inline UnicodeString&
+UnicodeString::append(char16_t srcChar)
+{ return doAppend(&srcChar, 0, 1); }
+
+inline UnicodeString&
+UnicodeString::operator+= (char16_t ch)
+{ return doAppend(&ch, 0, 1); }
+
+inline UnicodeString&
+UnicodeString::operator+= (UChar32 ch) {
+  return append(ch);
+}
+
+inline UnicodeString&
+UnicodeString::operator+= (const UnicodeString& srcText)
+{ return doAppend(srcText, 0, srcText.length()); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              const UnicodeString& srcText,
+              int32_t srcStart,
+              int32_t srcLength)
+{ return doReplace(start, 0, srcText, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              const UnicodeString& srcText)
+{ return doReplace(start, 0, srcText, 0, srcText.length()); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              const char16_t *srcChars,
+              int32_t srcStart,
+              int32_t srcLength)
+{ return doReplace(start, 0, srcChars, srcStart, srcLength); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              ConstChar16Ptr srcChars,
+              int32_t srcLength)
+{ return doReplace(start, 0, srcChars, 0, srcLength); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              char16_t srcChar)
+{ return doReplace(start, 0, &srcChar, 0, 1); }
+
+inline UnicodeString&
+UnicodeString::insert(int32_t start,
+              UChar32 srcChar)
+{ return replace(start, 0, srcChar); }
+
+
+inline UnicodeString&
+UnicodeString::remove()
+{
+  // remove() of a bogus string makes the string empty and non-bogus
+  if(isBogus()) {
+    setToEmpty();
+  } else {
+    setZeroLength();
+  }
+  return *this;
+}
+
+inline UnicodeString&
+UnicodeString::remove(int32_t start,
+             int32_t _length)
+{
+    if(start <= 0 && _length == INT32_MAX) {
+        // remove(guaranteed everything) of a bogus string makes the string empty and non-bogus
+        return remove();
+    }
+    return doReplace(start, _length, NULL, 0, 0);
+}
+
+inline UnicodeString&
+UnicodeString::removeBetween(int32_t start,
+                int32_t limit)
+{ return doReplace(start, limit - start, NULL, 0, 0); }
+
+inline UnicodeString &
+UnicodeString::retainBetween(int32_t start, int32_t limit) {
+  truncate(limit);
+  return doReplace(0, start, NULL, 0, 0);
+}
+
+inline UBool
+UnicodeString::truncate(int32_t targetLength)
+{
+  if(isBogus() && targetLength == 0) {
+    // truncate(0) of a bogus string makes the string empty and non-bogus
+    unBogus();
+    return false;
+  } else if((uint32_t)targetLength < (uint32_t)length()) {
+    setLength(targetLength);
+    return true;
+  } else {
+    return false;
+  }
+}
+
+inline UnicodeString&
+UnicodeString::reverse()
+{ return doReverse(0, length()); }
+
+inline UnicodeString&
+UnicodeString::reverse(int32_t start,
+               int32_t _length)
+{ return doReverse(start, _length); }
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unorm.h b/thirdparty/icu4c/common/unicode/unorm.h
new file mode 100644
index 0000000000..c3c57582d4
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unorm.h
@@ -0,0 +1,476 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (c) 1996-2016, International Business Machines Corporation
+*               and others. All Rights Reserved.
+*******************************************************************************
+* File unorm.h
+*
+* Created by: Vladimir Weinstein 12052000
+*
+* Modification history :
+*
+* Date        Name        Description
+* 02/01/01    synwee      Added normalization quickcheck enum and method.
+*/
+#ifndef UNORM_H
+#define UNORM_H
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/uiter.h"
+#include "unicode/unorm2.h"
+
+/**
+ * \file
+ * \brief C API: Unicode Normalization
+ *
+ * Old Unicode normalization API.
+ *
+ * This API has been replaced by the unorm2.h API and is only available
+ * for backward compatibility. The functions here simply delegate to the
+ * unorm2.h functions, for example unorm2_getInstance() and unorm2_normalize().
+ * There is one exception: The new API does not provide a replacement for unorm_compare().
+ * Its declaration has been moved to unorm2.h.
+ *
+ * <code>unorm_normalize</code> transforms Unicode text into an equivalent composed or
+ * decomposed form, allowing for easier sorting and searching of text.
+ * <code>unorm_normalize</code> supports the standard normalization forms described in
+ * <a href="http://www.unicode.org/unicode/reports/tr15/" target="unicode">
+ * Unicode Standard Annex #15: Unicode Normalization Forms</a>.
+ *
+ * Characters with accents or other adornments can be encoded in
+ * several different ways in Unicode.  For example, take the character A-acute.
+ * In Unicode, this can be encoded as a single character (the
+ * "composed" form):
+ *
+ * \code
+ *      00C1    LATIN CAPITAL LETTER A WITH ACUTE
+ * \endcode
+ *
+ * or as two separate characters (the "decomposed" form):
+ *
+ * \code
+ *      0041    LATIN CAPITAL LETTER A
+ *      0301    COMBINING ACUTE ACCENT
+ * \endcode
+ *
+ * To a user of your program, however, both of these sequences should be
+ * treated as the same "user-level" character "A with acute accent".  When you are searching or
+ * comparing text, you must ensure that these two sequences are treated
+ * equivalently.  In addition, you must handle characters with more than one
+ * accent.  Sometimes the order of a character's combining accents is
+ * significant, while in other cases accent sequences in different orders are
+ * really equivalent.
+ *
+ * Similarly, the string "ffi" can be encoded as three separate letters:
+ *
+ * \code
+ *      0066    LATIN SMALL LETTER F
+ *      0066    LATIN SMALL LETTER F
+ *      0069    LATIN SMALL LETTER I
+ * \endcode
+ *
+ * or as the single character
+ *
+ * \code
+ *      FB03    LATIN SMALL LIGATURE FFI
+ * \endcode
+ *
+ * The ffi ligature is not a distinct semantic character, and strictly speaking
+ * it shouldn't be in Unicode at all, but it was included for compatibility
+ * with existing character sets that already provided it.  The Unicode standard
+ * identifies such characters by giving them "compatibility" decompositions
+ * into the corresponding semantic characters.  When sorting and searching, you
+ * will often want to use these mappings.
+ *
+ * <code>unorm_normalize</code> helps solve these problems by transforming text into the
+ * canonical composed and decomposed forms as shown in the first example above.
+ * In addition, you can have it perform compatibility decompositions so that
+ * you can treat compatibility characters the same as their equivalents.
+ * Finally, <code>unorm_normalize</code> rearranges accents into the proper canonical
+ * order, so that you do not have to worry about accent rearrangement on your
+ * own.
+ *
+ * Form FCD, "Fast C or D", is also designed for collation.
+ * It allows to work on strings that are not necessarily normalized
+ * with an algorithm (like in collation) that works under "canonical closure", i.e., it treats precomposed
+ * characters and their decomposed equivalents the same.
+ *
+ * It is not a normalization form because it does not provide for uniqueness of representation. Multiple strings
+ * may be canonically equivalent (their NFDs are identical) and may all conform to FCD without being identical
+ * themselves.
+ *
+ * The form is defined such that the "raw decomposition", the recursive canonical decomposition of each character,
+ * results in a string that is canonically ordered. This means that precomposed characters are allowed for as long
+ * as their decompositions do not need canonical reordering.
+ *
+ * Its advantage for a process like collation is that all NFD and most NFC texts - and many unnormalized texts -
+ * already conform to FCD and do not need to be normalized (NFD) for such a process. The FCD quick check will
+ * return UNORM_YES for most strings in practice.
+ *
+ * unorm_normalize(UNORM_FCD) may be implemented with UNORM_NFD.
+ *
+ * For more details on FCD see the collation design document:
+ * http://source.icu-project.org/repos/icu/icuhtml/trunk/design/collation/ICU_collation_design.htm
+ *
+ * ICU collation performs either NFD or FCD normalization automatically if normalization
+ * is turned on for the collator object.
+ * Beyond collation and string search, normalized strings may be useful for string equivalence comparisons,
+ * transliteration/transcription, unique representations, etc.
+ *
+ * The W3C generally recommends to exchange texts in NFC.
+ * Note also that most legacy character encodings use only precomposed forms and often do not
+ * encode any combining marks by themselves. For conversion to such character encodings the
+ * Unicode text needs to be normalized to NFC.
+ * For more usage examples, see the Unicode Standard Annex.
+ */
+
+// Do not conditionalize the following enum with #ifndef U_HIDE_DEPRECATED_API,
+// it is needed for layout of Normalizer object.
+#ifndef U_FORCE_HIDE_DEPRECATED_API
+
+/**
+ * Constants for normalization modes.
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+typedef enum {
+  /** No decomposition/composition. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_NONE = 1,
+  /** Canonical decomposition. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_NFD = 2,
+  /** Compatibility decomposition. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_NFKD = 3,
+  /** Canonical decomposition followed by canonical composition. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_NFC = 4,
+  /** Default normalization. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_DEFAULT = UNORM_NFC,
+  /** Compatibility decomposition followed by canonical composition. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_NFKC =5,
+  /** "Fast C or D" form. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_FCD = 6,
+
+  /** One more than the highest normalization mode constant. @deprecated ICU 56 Use unorm2.h instead. */
+  UNORM_MODE_COUNT
+} UNormalizationMode;
+
+#endif  // U_FORCE_HIDE_DEPRECATED_API
+
+#ifndef U_HIDE_DEPRECATED_API
+
+/**
+ * Constants for options flags for normalization.
+ * Use 0 for default options,
+ * including normalization according to the Unicode version
+ * that is currently supported by ICU (see u_getUnicodeVersion).
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+enum {
+    /**
+     * Options bit set value to select Unicode 3.2 normalization
+     * (except NormalizationCorrections).
+     * At most one Unicode version can be selected at a time.
+     * @deprecated ICU 56 Use unorm2.h instead.
+     */
+    UNORM_UNICODE_3_2=0x20
+};
+
+/**
+ * Lowest-order bit number of unorm_compare() options bits corresponding to
+ * normalization options bits.
+ *
+ * The options parameter for unorm_compare() uses most bits for
+ * itself and for various comparison and folding flags.
+ * The most significant bits, however, are shifted down and passed on
+ * to the normalization implementation.
+ * (That is, from unorm_compare(..., options, ...),
+ * options>>UNORM_COMPARE_NORM_OPTIONS_SHIFT will be passed on to the
+ * internal normalization functions.)
+ *
+ * @see unorm_compare
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+#define UNORM_COMPARE_NORM_OPTIONS_SHIFT 20
+
+/**
+ * Normalize a string.
+ * The string will be normalized according the specified normalization mode
+ * and options.
+ * The source and result buffers must not be the same, nor overlap.
+ *
+ * @param source The string to normalize.
+ * @param sourceLength The length of source, or -1 if NUL-terminated.
+ * @param mode The normalization mode; one of UNORM_NONE,
+ *             UNORM_NFD, UNORM_NFC, UNORM_NFKC, UNORM_NFKD, UNORM_DEFAULT.
+ * @param options The normalization options, ORed together (0 for no options).
+ * @param result A pointer to a buffer to receive the result string.
+ *               The result string is NUL-terminated if possible.
+ * @param resultLength The maximum size of result.
+ * @param status A pointer to a UErrorCode to receive any errors.
+ * @return The total buffer size needed; if greater than resultLength,
+ *         the output was truncated, and the error code is set to U_BUFFER_OVERFLOW_ERROR.
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+unorm_normalize(const UChar *source, int32_t sourceLength,
+                UNormalizationMode mode, int32_t options,
+                UChar *result, int32_t resultLength,
+                UErrorCode *status);
+
+/**
+ * Performing quick check on a string, to quickly determine if the string is
+ * in a particular normalization format.
+ * Three types of result can be returned UNORM_YES, UNORM_NO or
+ * UNORM_MAYBE. Result UNORM_YES indicates that the argument
+ * string is in the desired normalized format, UNORM_NO determines that
+ * argument string is not in the desired normalized format. A
+ * UNORM_MAYBE result indicates that a more thorough check is required,
+ * the user may have to put the string in its normalized form and compare the
+ * results.
+ *
+ * @param source       string for determining if it is in a normalized format
+ * @param sourcelength length of source to test, or -1 if NUL-terminated
+ * @param mode         which normalization form to test for
+ * @param status       a pointer to a UErrorCode to receive any errors
+ * @return UNORM_YES, UNORM_NO or UNORM_MAYBE
+ *
+ * @see unorm_isNormalized
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED UNormalizationCheckResult U_EXPORT2
+unorm_quickCheck(const UChar *source, int32_t sourcelength,
+                 UNormalizationMode mode,
+                 UErrorCode *status);
+
+/**
+ * Performing quick check on a string; same as unorm_quickCheck but
+ * takes an extra options parameter like most normalization functions.
+ *
+ * @param src        String that is to be tested if it is in a normalization format.
+ * @param srcLength  Length of source to test, or -1 if NUL-terminated.
+ * @param mode       Which normalization form to test for.
+ * @param options    The normalization options, ORed together (0 for no options).
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return UNORM_YES, UNORM_NO or UNORM_MAYBE
+ *
+ * @see unorm_quickCheck
+ * @see unorm_isNormalized
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED UNormalizationCheckResult U_EXPORT2
+unorm_quickCheckWithOptions(const UChar *src, int32_t srcLength,
+                            UNormalizationMode mode, int32_t options,
+                            UErrorCode *pErrorCode);
+
+/**
+ * Test if a string is in a given normalization form.
+ * This is semantically equivalent to source.equals(normalize(source, mode)) .
+ *
+ * Unlike unorm_quickCheck(), this function returns a definitive result,
+ * never a "maybe".
+ * For NFD, NFKD, and FCD, both functions work exactly the same.
+ * For NFC and NFKC where quickCheck may return "maybe", this function will
+ * perform further tests to arrive at a true/false result.
+ *
+ * @param src        String that is to be tested if it is in a normalization format.
+ * @param srcLength  Length of source to test, or -1 if NUL-terminated.
+ * @param mode       Which normalization form to test for.
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return Boolean value indicating whether the source string is in the
+ *         "mode" normalization form.
+ *
+ * @see unorm_quickCheck
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED UBool U_EXPORT2
+unorm_isNormalized(const UChar *src, int32_t srcLength,
+                   UNormalizationMode mode,
+                   UErrorCode *pErrorCode);
+
+/**
+ * Test if a string is in a given normalization form; same as unorm_isNormalized but
+ * takes an extra options parameter like most normalization functions.
+ *
+ * @param src        String that is to be tested if it is in a normalization format.
+ * @param srcLength  Length of source to test, or -1 if NUL-terminated.
+ * @param mode       Which normalization form to test for.
+ * @param options    The normalization options, ORed together (0 for no options).
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return Boolean value indicating whether the source string is in the
+ *         "mode/options" normalization form.
+ *
+ * @see unorm_quickCheck
+ * @see unorm_isNormalized
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED UBool U_EXPORT2
+unorm_isNormalizedWithOptions(const UChar *src, int32_t srcLength,
+                              UNormalizationMode mode, int32_t options,
+                              UErrorCode *pErrorCode);
+
+/**
+ * Iterative normalization forward.
+ * This function (together with unorm_previous) is somewhat
+ * similar to the C++ Normalizer class (see its non-static functions).
+ *
+ * Iterative normalization is useful when only a small portion of a longer
+ * string/text needs to be processed.
+ *
+ * For example, the likelihood may be high that processing the first 10% of some
+ * text will be sufficient to find certain data.
+ * Another example: When one wants to concatenate two normalized strings and get a
+ * normalized result, it is much more efficient to normalize just a small part of
+ * the result around the concatenation place instead of re-normalizing everything.
+ *
+ * The input text is an instance of the C character iteration API UCharIterator.
+ * It may wrap around a simple string, a CharacterIterator, a Replaceable, or any
+ * other kind of text object.
+ *
+ * If a buffer overflow occurs, then the caller needs to reset the iterator to the
+ * old index and call the function again with a larger buffer - if the caller cares
+ * for the actual output.
+ * Regardless of the output buffer, the iterator will always be moved to the next
+ * normalization boundary.
+ *
+ * This function (like unorm_previous) serves two purposes:
+ *
+ * 1) To find the next boundary so that the normalization of the part of the text
+ * from the current position to that boundary does not affect and is not affected
+ * by the part of the text beyond that boundary.
+ *
+ * 2) To normalize the text up to the boundary.
+ *
+ * The second step is optional, per the doNormalize parameter.
+ * It is omitted for operations like string concatenation, where the two adjacent
+ * string ends need to be normalized together.
+ * In such a case, the output buffer will just contain a copy of the text up to the
+ * boundary.
+ *
+ * pNeededToNormalize is an output-only parameter. Its output value is only defined
+ * if normalization was requested (doNormalize) and successful (especially, no
+ * buffer overflow).
+ * It is useful for operations like a normalizing transliterator, where one would
+ * not want to replace a piece of text if it is not modified.
+ *
+ * If doNormalize==true and pNeededToNormalize!=NULL then *pNeeded... is set true
+ * if the normalization was necessary.
+ *
+ * If doNormalize==false then *pNeededToNormalize will be set to false.
+ *
+ * If the buffer overflows, then *pNeededToNormalize will be undefined;
+ * essentially, whenever U_FAILURE is true (like in buffer overflows), this result
+ * will be undefined.
+ *
+ * @param src The input text in the form of a C character iterator.
+ * @param dest The output buffer; can be NULL if destCapacity==0 for pure preflighting.
+ * @param destCapacity The number of UChars that fit into dest.
+ * @param mode The normalization mode.
+ * @param options The normalization options, ORed together (0 for no options).
+ * @param doNormalize Indicates if the source text up to the next boundary
+ *                    is to be normalized (true) or just copied (false).
+ * @param pNeededToNormalize Output flag indicating if the normalization resulted in
+ *                           different text from the input.
+ *                           Not defined if an error occurs including buffer overflow.
+ *                           Always false if !doNormalize.
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return Length of output (number of UChars) when successful or buffer overflow.
+ *
+ * @see unorm_previous
+ * @see unorm_normalize
+ *
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+unorm_next(UCharIterator *src,
+           UChar *dest, int32_t destCapacity,
+           UNormalizationMode mode, int32_t options,
+           UBool doNormalize, UBool *pNeededToNormalize,
+           UErrorCode *pErrorCode);
+
+/**
+ * Iterative normalization backward.
+ * This function (together with unorm_next) is somewhat
+ * similar to the C++ Normalizer class (see its non-static functions).
+ * For all details see unorm_next.
+ *
+ * @param src The input text in the form of a C character iterator.
+ * @param dest The output buffer; can be NULL if destCapacity==0 for pure preflighting.
+ * @param destCapacity The number of UChars that fit into dest.
+ * @param mode The normalization mode.
+ * @param options The normalization options, ORed together (0 for no options).
+ * @param doNormalize Indicates if the source text up to the next boundary
+ *                    is to be normalized (true) or just copied (false).
+ * @param pNeededToNormalize Output flag indicating if the normalization resulted in
+ *                           different text from the input.
+ *                           Not defined if an error occurs including buffer overflow.
+ *                           Always false if !doNormalize.
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return Length of output (number of UChars) when successful or buffer overflow.
+ *
+ * @see unorm_next
+ * @see unorm_normalize
+ *
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+unorm_previous(UCharIterator *src,
+               UChar *dest, int32_t destCapacity,
+               UNormalizationMode mode, int32_t options,
+               UBool doNormalize, UBool *pNeededToNormalize,
+               UErrorCode *pErrorCode);
+
+/**
+ * Concatenate normalized strings, making sure that the result is normalized as well.
+ *
+ * If both the left and the right strings are in
+ * the normalization form according to "mode/options",
+ * then the result will be
+ *
+ * \code
+ *     dest=normalize(left+right, mode, options)
+ * \endcode
+ *
+ * With the input strings already being normalized,
+ * this function will use unorm_next() and unorm_previous()
+ * to find the adjacent end pieces of the input strings.
+ * Only the concatenation of these end pieces will be normalized and
+ * then concatenated with the remaining parts of the input strings.
+ *
+ * It is allowed to have dest==left to avoid copying the entire left string.
+ *
+ * @param left Left source string, may be same as dest.
+ * @param leftLength Length of left source string, or -1 if NUL-terminated.
+ * @param right Right source string. Must not be the same as dest, nor overlap.
+ * @param rightLength Length of right source string, or -1 if NUL-terminated.
+ * @param dest The output buffer; can be NULL if destCapacity==0 for pure preflighting.
+ * @param destCapacity The number of UChars that fit into dest.
+ * @param mode The normalization mode.
+ * @param options The normalization options, ORed together (0 for no options).
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return Length of output (number of UChars) when successful or buffer overflow.
+ *
+ * @see unorm_normalize
+ * @see unorm_next
+ * @see unorm_previous
+ *
+ * @deprecated ICU 56 Use unorm2.h instead.
+ */
+U_DEPRECATED int32_t U_EXPORT2
+unorm_concatenate(const UChar *left, int32_t leftLength,
+                  const UChar *right, int32_t rightLength,
+                  UChar *dest, int32_t destCapacity,
+                  UNormalizationMode mode, int32_t options,
+                  UErrorCode *pErrorCode);
+
+#endif  /* U_HIDE_DEPRECATED_API */
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
+#endif
diff --git a/thirdparty/icu4c/common/unicode/unorm2.h b/thirdparty/icu4c/common/unicode/unorm2.h
new file mode 100644
index 0000000000..24417b7103
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/unorm2.h
@@ -0,0 +1,606 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2009-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unorm2.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2009dec15
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UNORM2_H__
+#define __UNORM2_H__
+
+/**
+ * \file
+ * \brief C API: New API for Unicode Normalization.
+ *
+ * Unicode normalization functionality for standard Unicode normalization or
+ * for using custom mapping tables.
+ * All instances of UNormalizer2 are unmodifiable/immutable.
+ * Instances returned by unorm2_getInstance() are singletons that must not be deleted by the caller.
+ * For more details see the Normalizer2 C++ class.
+ */
+
+#include "unicode/utypes.h"
+#include "unicode/stringoptions.h"
+#include "unicode/uset.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * Constants for normalization modes.
+ * For details about standard Unicode normalization forms
+ * and about the algorithms which are also used with custom mapping tables
+ * see http://www.unicode.org/unicode/reports/tr15/
+ * @stable ICU 4.4
+ */
+typedef enum {
+    /**
+     * Decomposition followed by composition.
+     * Same as standard NFC when using an "nfc" instance.
+     * Same as standard NFKC when using an "nfkc" instance.
+     * For details about standard Unicode normalization forms
+     * see http://www.unicode.org/unicode/reports/tr15/
+     * @stable ICU 4.4
+     */
+    UNORM2_COMPOSE,
+    /**
+     * Map, and reorder canonically.
+     * Same as standard NFD when using an "nfc" instance.
+     * Same as standard NFKD when using an "nfkc" instance.
+     * For details about standard Unicode normalization forms
+     * see http://www.unicode.org/unicode/reports/tr15/
+     * @stable ICU 4.4
+     */
+    UNORM2_DECOMPOSE,
+    /**
+     * "Fast C or D" form.
+     * If a string is in this form, then further decomposition <i>without reordering</i>
+     * would yield the same form as DECOMPOSE.
+     * Text in "Fast C or D" form can be processed efficiently with data tables
+     * that are "canonically closed", that is, that provide equivalent data for
+     * equivalent text, without having to be fully normalized.
+     * Not a standard Unicode normalization form.
+     * Not a unique form: Different FCD strings can be canonically equivalent.
+     * For details see http://www.unicode.org/notes/tn5/#FCD
+     * @stable ICU 4.4
+     */
+    UNORM2_FCD,
+    /**
+     * Compose only contiguously.
+     * Also known as "FCC" or "Fast C Contiguous".
+     * The result will often but not always be in NFC.
+     * The result will conform to FCD which is useful for processing.
+     * Not a standard Unicode normalization form.
+     * For details see http://www.unicode.org/notes/tn5/#FCC
+     * @stable ICU 4.4
+     */
+    UNORM2_COMPOSE_CONTIGUOUS
+} UNormalization2Mode;
+
+/**
+ * Result values for normalization quick check functions.
+ * For details see http://www.unicode.org/reports/tr15/#Detecting_Normalization_Forms
+ * @stable ICU 2.0
+ */
+typedef enum UNormalizationCheckResult {
+  /**
+   * The input string is not in the normalization form.
+   * @stable ICU 2.0
+   */
+  UNORM_NO,
+  /**
+   * The input string is in the normalization form.
+   * @stable ICU 2.0
+   */
+  UNORM_YES,
+  /**
+   * The input string may or may not be in the normalization form.
+   * This value is only returned for composition forms like NFC and FCC,
+   * when a backward-combining character is found for which the surrounding text
+   * would have to be analyzed further.
+   * @stable ICU 2.0
+   */
+  UNORM_MAYBE
+} UNormalizationCheckResult;
+
+/**
+ * Opaque C service object type for the new normalization API.
+ * @stable ICU 4.4
+ */
+struct UNormalizer2;
+typedef struct UNormalizer2 UNormalizer2;  /**< C typedef for struct UNormalizer2. @stable ICU 4.4 */
+
+#if !UCONFIG_NO_NORMALIZATION
+
+/**
+ * Returns a UNormalizer2 instance for Unicode NFC normalization.
+ * Same as unorm2_getInstance(NULL, "nfc", UNORM2_COMPOSE, pErrorCode).
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested Normalizer2, if successful
+ * @stable ICU 49
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFCInstance(UErrorCode *pErrorCode);
+
+/**
+ * Returns a UNormalizer2 instance for Unicode NFD normalization.
+ * Same as unorm2_getInstance(NULL, "nfc", UNORM2_DECOMPOSE, pErrorCode).
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested Normalizer2, if successful
+ * @stable ICU 49
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFDInstance(UErrorCode *pErrorCode);
+
+/**
+ * Returns a UNormalizer2 instance for Unicode NFKC normalization.
+ * Same as unorm2_getInstance(NULL, "nfkc", UNORM2_COMPOSE, pErrorCode).
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested Normalizer2, if successful
+ * @stable ICU 49
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKCInstance(UErrorCode *pErrorCode);
+
+/**
+ * Returns a UNormalizer2 instance for Unicode NFKD normalization.
+ * Same as unorm2_getInstance(NULL, "nfkc", UNORM2_DECOMPOSE, pErrorCode).
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested Normalizer2, if successful
+ * @stable ICU 49
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKDInstance(UErrorCode *pErrorCode);
+
+/**
+ * Returns a UNormalizer2 instance for Unicode NFKC_Casefold normalization.
+ * Same as unorm2_getInstance(NULL, "nfkc_cf", UNORM2_COMPOSE, pErrorCode).
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested Normalizer2, if successful
+ * @stable ICU 49
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getNFKCCasefoldInstance(UErrorCode *pErrorCode);
+
+/**
+ * Returns a UNormalizer2 instance which uses the specified data file
+ * (packageName/name similar to ucnv_openPackage() and ures_open()/ResourceBundle)
+ * and which composes or decomposes text according to the specified mode.
+ * Returns an unmodifiable singleton instance. Do not delete it.
+ *
+ * Use packageName=NULL for data files that are part of ICU's own data.
+ * Use name="nfc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFC/NFD.
+ * Use name="nfkc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFKC/NFKD.
+ * Use name="nfkc_cf" and UNORM2_COMPOSE for Unicode standard NFKC_CF=NFKC_Casefold.
+ *
+ * @param packageName NULL for ICU built-in data, otherwise application data package name
+ * @param name "nfc" or "nfkc" or "nfkc_cf" or name of custom data file
+ * @param mode normalization mode (compose or decompose etc.)
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return the requested UNormalizer2, if successful
+ * @stable ICU 4.4
+ */
+U_CAPI const UNormalizer2 * U_EXPORT2
+unorm2_getInstance(const char *packageName,
+                   const char *name,
+                   UNormalization2Mode mode,
+                   UErrorCode *pErrorCode);
+
+/**
+ * Constructs a filtered normalizer wrapping any UNormalizer2 instance
+ * and a filter set.
+ * Both are aliased and must not be modified or deleted while this object
+ * is used.
+ * The filter set should be frozen; otherwise the performance will suffer greatly.
+ * @param norm2 wrapped UNormalizer2 instance
+ * @param filterSet USet which determines the characters to be normalized
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return the requested UNormalizer2, if successful
+ * @stable ICU 4.4
+ */
+U_CAPI UNormalizer2 * U_EXPORT2
+unorm2_openFiltered(const UNormalizer2 *norm2, const USet *filterSet, UErrorCode *pErrorCode);
+
+/**
+ * Closes a UNormalizer2 instance from unorm2_openFiltered().
+ * Do not close instances from unorm2_getInstance()!
+ * @param norm2 UNormalizer2 instance to be closed
+ * @stable ICU 4.4
+ */
+U_CAPI void U_EXPORT2
+unorm2_close(UNormalizer2 *norm2);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUNormalizer2Pointer
+ * "Smart pointer" class, closes a UNormalizer2 via unorm2_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUNormalizer2Pointer, UNormalizer2, unorm2_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Writes the normalized form of the source string to the destination string
+ * (replacing its contents) and returns the length of the destination string.
+ * The source and destination strings must be different buffers.
+ * @param norm2 UNormalizer2 instance
+ * @param src source string
+ * @param length length of the source string, or -1 if NUL-terminated
+ * @param dest destination string; its contents is replaced with normalized src
+ * @param capacity number of UChars that can be written to dest
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return dest
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_normalize(const UNormalizer2 *norm2,
+                 const UChar *src, int32_t length,
+                 UChar *dest, int32_t capacity,
+                 UErrorCode *pErrorCode);
+/**
+ * Appends the normalized form of the second string to the first string
+ * (merging them at the boundary) and returns the length of the first string.
+ * The result is normalized if the first string was normalized.
+ * The first and second strings must be different buffers.
+ * @param norm2 UNormalizer2 instance
+ * @param first string, should be normalized
+ * @param firstLength length of the first string, or -1 if NUL-terminated
+ * @param firstCapacity number of UChars that can be written to first
+ * @param second string, will be normalized
+ * @param secondLength length of the source string, or -1 if NUL-terminated
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return first
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_normalizeSecondAndAppend(const UNormalizer2 *norm2,
+                                UChar *first, int32_t firstLength, int32_t firstCapacity,
+                                const UChar *second, int32_t secondLength,
+                                UErrorCode *pErrorCode);
+/**
+ * Appends the second string to the first string
+ * (merging them at the boundary) and returns the length of the first string.
+ * The result is normalized if both the strings were normalized.
+ * The first and second strings must be different buffers.
+ * @param norm2 UNormalizer2 instance
+ * @param first string, should be normalized
+ * @param firstLength length of the first string, or -1 if NUL-terminated
+ * @param firstCapacity number of UChars that can be written to first
+ * @param second string, should be normalized
+ * @param secondLength length of the source string, or -1 if NUL-terminated
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return first
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_append(const UNormalizer2 *norm2,
+              UChar *first, int32_t firstLength, int32_t firstCapacity,
+              const UChar *second, int32_t secondLength,
+              UErrorCode *pErrorCode);
+
+/**
+ * Gets the decomposition mapping of c.
+ * Roughly equivalent to normalizing the String form of c
+ * on a UNORM2_DECOMPOSE UNormalizer2 instance, but much faster, and except that this function
+ * returns a negative value and does not write a string
+ * if c does not have a decomposition mapping in this instance's data.
+ * This function is independent of the mode of the UNormalizer2.
+ * @param norm2 UNormalizer2 instance
+ * @param c code point
+ * @param decomposition String buffer which will be set to c's
+ *                      decomposition mapping, if there is one.
+ * @param capacity number of UChars that can be written to decomposition
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return the non-negative length of c's decomposition, if there is one; otherwise a negative value
+ * @stable ICU 4.6
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_getDecomposition(const UNormalizer2 *norm2,
+                        UChar32 c, UChar *decomposition, int32_t capacity,
+                        UErrorCode *pErrorCode);
+
+/**
+ * Gets the raw decomposition mapping of c.
+ *
+ * This is similar to the unorm2_getDecomposition() function but returns the
+ * raw decomposition mapping as specified in UnicodeData.txt or
+ * (for custom data) in the mapping files processed by the gennorm2 tool.
+ * By contrast, unorm2_getDecomposition() returns the processed,
+ * recursively-decomposed version of this mapping.
+ *
+ * When used on a standard NFKC Normalizer2 instance,
+ * unorm2_getRawDecomposition() returns the Unicode Decomposition_Mapping (dm) property.
+ *
+ * When used on a standard NFC Normalizer2 instance,
+ * it returns the Decomposition_Mapping only if the Decomposition_Type (dt) is Canonical (Can);
+ * in this case, the result contains either one or two code points (=1..4 UChars).
+ *
+ * This function is independent of the mode of the UNormalizer2.
+ * @param norm2 UNormalizer2 instance
+ * @param c code point
+ * @param decomposition String buffer which will be set to c's
+ *                      raw decomposition mapping, if there is one.
+ * @param capacity number of UChars that can be written to decomposition
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return the non-negative length of c's raw decomposition, if there is one; otherwise a negative value
+ * @stable ICU 49
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_getRawDecomposition(const UNormalizer2 *norm2,
+                           UChar32 c, UChar *decomposition, int32_t capacity,
+                           UErrorCode *pErrorCode);
+
+/**
+ * Performs pairwise composition of a & b and returns the composite if there is one.
+ *
+ * Returns a composite code point c only if c has a two-way mapping to a+b.
+ * In standard Unicode normalization, this means that
+ * c has a canonical decomposition to a+b
+ * and c does not have the Full_Composition_Exclusion property.
+ *
+ * This function is independent of the mode of the UNormalizer2.
+ * @param norm2 UNormalizer2 instance
+ * @param a A (normalization starter) code point.
+ * @param b Another code point.
+ * @return The non-negative composite code point if there is one; otherwise a negative value.
+ * @stable ICU 49
+ */
+U_CAPI UChar32 U_EXPORT2
+unorm2_composePair(const UNormalizer2 *norm2, UChar32 a, UChar32 b);
+
+/**
+ * Gets the combining class of c.
+ * The default implementation returns 0
+ * but all standard implementations return the Unicode Canonical_Combining_Class value.
+ * @param norm2 UNormalizer2 instance
+ * @param c code point
+ * @return c's combining class
+ * @stable ICU 49
+ */
+U_CAPI uint8_t U_EXPORT2
+unorm2_getCombiningClass(const UNormalizer2 *norm2, UChar32 c);
+
+/**
+ * Tests if the string is normalized.
+ * Internally, in cases where the quickCheck() method would return "maybe"
+ * (which is only possible for the two COMPOSE modes) this method
+ * resolves to "yes" or "no" to provide a definitive result,
+ * at the cost of doing more work in those cases.
+ * @param norm2 UNormalizer2 instance
+ * @param s input string
+ * @param length length of the string, or -1 if NUL-terminated
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return true if s is normalized
+ * @stable ICU 4.4
+ */
+U_CAPI UBool U_EXPORT2
+unorm2_isNormalized(const UNormalizer2 *norm2,
+                    const UChar *s, int32_t length,
+                    UErrorCode *pErrorCode);
+
+/**
+ * Tests if the string is normalized.
+ * For the two COMPOSE modes, the result could be "maybe" in cases that
+ * would take a little more work to resolve definitively.
+ * Use spanQuickCheckYes() and normalizeSecondAndAppend() for a faster
+ * combination of quick check + normalization, to avoid
+ * re-checking the "yes" prefix.
+ * @param norm2 UNormalizer2 instance
+ * @param s input string
+ * @param length length of the string, or -1 if NUL-terminated
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return UNormalizationCheckResult
+ * @stable ICU 4.4
+ */
+U_CAPI UNormalizationCheckResult U_EXPORT2
+unorm2_quickCheck(const UNormalizer2 *norm2,
+                  const UChar *s, int32_t length,
+                  UErrorCode *pErrorCode);
+
+/**
+ * Returns the end of the normalized substring of the input string.
+ * In other words, with <code>end=spanQuickCheckYes(s, ec);</code>
+ * the substring <code>UnicodeString(s, 0, end)</code>
+ * will pass the quick check with a "yes" result.
+ *
+ * The returned end index is usually one or more characters before the
+ * "no" or "maybe" character: The end index is at a normalization boundary.
+ * (See the class documentation for more about normalization boundaries.)
+ *
+ * When the goal is a normalized string and most input strings are expected
+ * to be normalized already, then call this method,
+ * and if it returns a prefix shorter than the input string,
+ * copy that prefix and use normalizeSecondAndAppend() for the remainder.
+ * @param norm2 UNormalizer2 instance
+ * @param s input string
+ * @param length length of the string, or -1 if NUL-terminated
+ * @param pErrorCode Standard ICU error code. Its input value must
+ *                   pass the U_SUCCESS() test, or else the function returns
+ *                   immediately. Check for U_FAILURE() on output or use with
+ *                   function chaining. (See User Guide for details.)
+ * @return "yes" span end index
+ * @stable ICU 4.4
+ */
+U_CAPI int32_t U_EXPORT2
+unorm2_spanQuickCheckYes(const UNormalizer2 *norm2,
+                         const UChar *s, int32_t length,
+                         UErrorCode *pErrorCode);
+
+/**
+ * Tests if the character always has a normalization boundary before it,
+ * regardless of context.
+ * For details see the Normalizer2 base class documentation.
+ * @param norm2 UNormalizer2 instance
+ * @param c character to test
+ * @return true if c has a normalization boundary before it
+ * @stable ICU 4.4
+ */
+U_CAPI UBool U_EXPORT2
+unorm2_hasBoundaryBefore(const UNormalizer2 *norm2, UChar32 c);
+
+/**
+ * Tests if the character always has a normalization boundary after it,
+ * regardless of context.
+ * For details see the Normalizer2 base class documentation.
+ * @param norm2 UNormalizer2 instance
+ * @param c character to test
+ * @return true if c has a normalization boundary after it
+ * @stable ICU 4.4
+ */
+U_CAPI UBool U_EXPORT2
+unorm2_hasBoundaryAfter(const UNormalizer2 *norm2, UChar32 c);
+
+/**
+ * Tests if the character is normalization-inert.
+ * For details see the Normalizer2 base class documentation.
+ * @param norm2 UNormalizer2 instance
+ * @param c character to test
+ * @return true if c is normalization-inert
+ * @stable ICU 4.4
+ */
+U_CAPI UBool U_EXPORT2
+unorm2_isInert(const UNormalizer2 *norm2, UChar32 c);
+
+/**
+ * Compares two strings for canonical equivalence.
+ * Further options include case-insensitive comparison and
+ * code point order (as opposed to code unit order).
+ *
+ * Canonical equivalence between two strings is defined as their normalized
+ * forms (NFD or NFC) being identical.
+ * This function compares strings incrementally instead of normalizing
+ * (and optionally case-folding) both strings entirely,
+ * improving performance significantly.
+ *
+ * Bulk normalization is only necessary if the strings do not fulfill the FCD
+ * conditions. Only in this case, and only if the strings are relatively long,
+ * is memory allocated temporarily.
+ * For FCD strings and short non-FCD strings there is no memory allocation.
+ *
+ * Semantically, this is equivalent to
+ *   strcmp[CodePointOrder](NFD(foldCase(NFD(s1))), NFD(foldCase(NFD(s2))))
+ * where code point order and foldCase are all optional.
+ *
+ * UAX 21 2.5 Caseless Matching specifies that for a canonical caseless match
+ * the case folding must be performed first, then the normalization.
+ *
+ * @param s1 First source string.
+ * @param length1 Length of first source string, or -1 if NUL-terminated.
+ *
+ * @param s2 Second source string.
+ * @param length2 Length of second source string, or -1 if NUL-terminated.
+ *
+ * @param options A bit set of options:
+ *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+ *     Case-sensitive comparison in code unit order, and the input strings
+ *     are quick-checked for FCD.
+ *
+ *   - UNORM_INPUT_IS_FCD
+ *     Set if the caller knows that both s1 and s2 fulfill the FCD conditions.
+ *     If not set, the function will quickCheck for FCD
+ *     and normalize if necessary.
+ *
+ *   - U_COMPARE_CODE_POINT_ORDER
+ *     Set to choose code point order instead of code unit order
+ *     (see u_strCompare for details).
+ *
+ *   - U_COMPARE_IGNORE_CASE
+ *     Set to compare strings case-insensitively using case folding,
+ *     instead of case-sensitively.
+ *     If set, then the following case folding options are used.
+ *
+ *   - Options as used with case-insensitive comparisons, currently:
+ *
+ *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ *    (see u_strCaseCompare for details)
+ *
+ *   - regular normalization options shifted left by UNORM_COMPARE_NORM_OPTIONS_SHIFT
+ *
+ * @param pErrorCode ICU error code in/out parameter.
+ *                   Must fulfill U_SUCCESS before the function call.
+ * @return <0 or 0 or >0 as usual for string comparisons
+ *
+ * @see unorm_normalize
+ * @see UNORM_FCD
+ * @see u_strCompare
+ * @see u_strCaseCompare
+ *
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+unorm_compare(const UChar *s1, int32_t length1,
+              const UChar *s2, int32_t length2,
+              uint32_t options,
+              UErrorCode *pErrorCode);
+
+#endif  /* !UCONFIG_NO_NORMALIZATION */
+#endif  /* __UNORM2_H__ */
diff --git a/thirdparty/icu4c/common/unicode/uobject.h b/thirdparty/icu4c/common/unicode/uobject.h
new file mode 100644
index 0000000000..eeb331ce97
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uobject.h
@@ -0,0 +1,324 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  uobject.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jun26
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UOBJECT_H__
+#define __UOBJECT_H__
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/platform.h"
+
+/**
+ * \file
+ * \brief C++ API: Common ICU base class UObject.
+ */
+
+/**
+ * \def U_NO_THROW
+ *         Since ICU 64, use U_NOEXCEPT instead.
+ *
+ *         Previously, define this to define the throw() specification so
+ *                 certain functions do not throw any exceptions
+ *
+ *         UMemory operator new methods should have the throw() specification 
+ *         appended to them, so that the compiler adds the additional NULL check 
+ *         before calling constructors. Without, if <code>operator new</code> returns NULL the 
+ *         constructor is still called, and if the constructor references member 
+ *         data, (which it typically does), the result is a segmentation violation.
+ *
+ * @stable ICU 4.2. Since ICU 64, Use U_NOEXCEPT instead. See ICU-20422.
+ */
+#ifndef U_NO_THROW
+#define U_NO_THROW U_NOEXCEPT
+#endif
+
+/*===========================================================================*/
+/* UClassID-based RTTI */
+/*===========================================================================*/
+
+/**
+ * UClassID is used to identify classes without using the compiler's RTTI.
+ * This was used before C++ compilers consistently supported RTTI.
+ * ICU 4.6 requires compiler RTTI to be turned on.
+ *
+ * Each class hierarchy which needs
+ * to implement polymorphic clone() or operator==() defines two methods,
+ * described in detail below.  UClassID values can be compared using
+ * operator==(). Nothing else should be done with them.
+ *
+ * \par
+ * In class hierarchies that implement "poor man's RTTI",
+ * each concrete subclass implements getDynamicClassID() in the same way:
+ *
+ * \code
+ *      class Derived {
+ *      public:
+ *          virtual UClassID getDynamicClassID() const
+ *            { return Derived::getStaticClassID(); }
+ *      }
+ * \endcode
+ *
+ * Each concrete class implements getStaticClassID() as well, which allows
+ * clients to test for a specific type.
+ *
+ * \code
+ *      class Derived {
+ *      public:
+ *          static UClassID U_EXPORT2 getStaticClassID();
+ *      private:
+ *          static char fgClassID;
+ *      }
+ *
+ *      // In Derived.cpp:
+ *      UClassID Derived::getStaticClassID()
+ *        { return (UClassID)&Derived::fgClassID; }
+ *      char Derived::fgClassID = 0; // Value is irrelevant
+ * \endcode
+ * @stable ICU 2.0
+ */
+typedef void* UClassID;
+
+U_NAMESPACE_BEGIN
+
+/**
+ * UMemory is the common ICU base class.
+ * All other ICU C++ classes are derived from UMemory (starting with ICU 2.4).
+ *
+ * This is primarily to make it possible and simple to override the
+ * C++ memory management by adding new/delete operators to this base class.
+ *
+ * To override ALL ICU memory management, including that from plain C code,
+ * replace the allocation functions declared in cmemory.h
+ *
+ * UMemory does not contain any virtual functions.
+ * Common "boilerplate" functions are defined in UObject.
+ *
+ * @stable ICU 2.4
+ */
+class U_COMMON_API UMemory {
+public:
+
+/* test versions for debugging shaper heap memory problems */
+#ifdef SHAPER_MEMORY_DEBUG  
+    static void * NewArray(int size, int count);
+    static void * GrowArray(void * array, int newSize );
+    static void   FreeArray(void * array );
+#endif
+
+#if U_OVERRIDE_CXX_ALLOCATION
+    /**
+     * Override for ICU4C C++ memory management.
+     * simple, non-class types are allocated using the macros in common/cmemory.h
+     * (uprv_malloc(), uprv_free(), uprv_realloc());
+     * they or something else could be used here to implement C++ new/delete
+     * for ICU4C C++ classes
+     * @stable ICU 2.4
+     */
+    static void * U_EXPORT2 operator new(size_t size) U_NOEXCEPT;
+
+    /**
+     * Override for ICU4C C++ memory management.
+     * See new().
+     * @stable ICU 2.4
+     */
+    static void * U_EXPORT2 operator new[](size_t size) U_NOEXCEPT;
+
+    /**
+     * Override for ICU4C C++ memory management.
+     * simple, non-class types are allocated using the macros in common/cmemory.h
+     * (uprv_malloc(), uprv_free(), uprv_realloc());
+     * they or something else could be used here to implement C++ new/delete
+     * for ICU4C C++ classes
+     * @stable ICU 2.4
+     */
+    static void U_EXPORT2 operator delete(void *p) U_NOEXCEPT;
+
+    /**
+     * Override for ICU4C C++ memory management.
+     * See delete().
+     * @stable ICU 2.4
+     */
+    static void U_EXPORT2 operator delete[](void *p) U_NOEXCEPT;
+
+#if U_HAVE_PLACEMENT_NEW
+    /**
+     * Override for ICU4C C++ memory management for STL.
+     * See new().
+     * @stable ICU 2.6
+     */
+    static inline void * U_EXPORT2 operator new(size_t, void *ptr) U_NOEXCEPT { return ptr; }
+
+    /**
+     * Override for ICU4C C++ memory management for STL.
+     * See delete().
+     * @stable ICU 2.6
+     */
+    static inline void U_EXPORT2 operator delete(void *, void *) U_NOEXCEPT {}
+#endif /* U_HAVE_PLACEMENT_NEW */
+#if U_HAVE_DEBUG_LOCATION_NEW
+    /**
+      * This method overrides the MFC debug version of the operator new
+      * 
+      * @param size   The requested memory size
+      * @param file   The file where the allocation was requested
+      * @param line   The line where the allocation was requested 
+      */ 
+    static void * U_EXPORT2 operator new(size_t size, const char* file, int line) U_NOEXCEPT;
+    /**
+      * This method provides a matching delete for the MFC debug new
+      * 
+      * @param p      The pointer to the allocated memory
+      * @param file   The file where the allocation was requested
+      * @param line   The line where the allocation was requested 
+      */ 
+    static void U_EXPORT2 operator delete(void* p, const char* file, int line) U_NOEXCEPT;
+#endif /* U_HAVE_DEBUG_LOCATION_NEW */
+#endif /* U_OVERRIDE_CXX_ALLOCATION */
+
+    /*
+     * Assignment operator not declared. The compiler will provide one
+     * which does nothing since this class does not contain any data members.
+     * API/code coverage may show the assignment operator as present and
+     * untested - ignore.
+     * Subclasses need this assignment operator if they use compiler-provided
+     * assignment operators of their own. An alternative to not declaring one
+     * here would be to declare and empty-implement a protected or public one.
+    UMemory &UMemory::operator=(const UMemory &);
+     */
+};
+
+/**
+ * UObject is the common ICU "boilerplate" class.
+ * UObject inherits UMemory (starting with ICU 2.4),
+ * and all other public ICU C++ classes
+ * are derived from UObject (starting with ICU 2.2).
+ *
+ * UObject contains common virtual functions, in particular a virtual destructor.
+ *
+ * The clone() function is not available in UObject because it is not
+ * implemented by all ICU classes.
+ * Many ICU services provide a clone() function for their class trees,
+ * defined on the service's C++ base class
+ * (which itself is a subclass of UObject).
+ *
+ * @stable ICU 2.2
+ */
+class U_COMMON_API UObject : public UMemory {
+public:
+    /**
+     * Destructor.
+     *
+     * @stable ICU 2.2
+     */
+    virtual ~UObject();
+
+    /**
+     * ICU4C "poor man's RTTI", returns a UClassID for the actual ICU class.
+     * The base class implementation returns a dummy value.
+     *
+     * Use compiler RTTI rather than ICU's "poor man's RTTI".
+     * Since ICU 4.6, new ICU C++ class hierarchies do not implement "poor man's RTTI".
+     *
+     * @stable ICU 2.2
+     */
+    virtual UClassID getDynamicClassID() const;
+
+protected:
+    // the following functions are protected to prevent instantiation and
+    // direct use of UObject itself
+
+    // default constructor
+    // inline UObject() {}
+
+    // copy constructor
+    // inline UObject(const UObject &other) {}
+
+#if 0
+    // TODO Sometime in the future. Implement operator==().
+    // (This comment inserted in 2.2)
+    // some or all of the following "boilerplate" functions may be made public
+    // in a future ICU4C release when all subclasses implement them
+
+    // assignment operator
+    // (not virtual, see "Taligent's Guide to Designing Programs" pp.73..74)
+    // commented out because the implementation is the same as a compiler's default
+    // UObject &operator=(const UObject &other) { return *this; }
+
+    // comparison operators
+    virtual inline UBool operator==(const UObject &other) const { return this==&other; }
+    inline UBool operator!=(const UObject &other) const { return !operator==(other); }
+
+    // clone() commented out from the base class:
+    // some compilers do not support co-variant return types
+    // (i.e., subclasses would have to return UObject * as well, instead of SubClass *)
+    // see also UObject class documentation.
+    // virtual UObject *clone() const;
+#endif
+
+    /*
+     * Assignment operator not declared. The compiler will provide one
+     * which does nothing since this class does not contain any data members.
+     * API/code coverage may show the assignment operator as present and
+     * untested - ignore.
+     * Subclasses need this assignment operator if they use compiler-provided
+     * assignment operators of their own. An alternative to not declaring one
+     * here would be to declare and empty-implement a protected or public one.
+    UObject &UObject::operator=(const UObject &);
+     */
+};
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * This is a simple macro to add ICU RTTI to an ICU object implementation.
+ * This does not go into the header. This should only be used in *.cpp files.
+ *
+ * @param myClass The name of the class that needs RTTI defined.
+ * @internal
+ */
+#define UOBJECT_DEFINE_RTTI_IMPLEMENTATION(myClass) \
+    UClassID U_EXPORT2 myClass::getStaticClassID() { \
+        static char classID = 0; \
+        return (UClassID)&classID; \
+    } \
+    UClassID myClass::getDynamicClassID() const \
+    { return myClass::getStaticClassID(); }
+
+
+/**
+ * This macro adds ICU RTTI to an ICU abstract class implementation.
+ * This macro should be invoked in *.cpp files.  The corresponding
+ * header should declare getStaticClassID.
+ *
+ * @param myClass The name of the class that needs RTTI defined.
+ * @internal
+ */
+#define UOBJECT_DEFINE_ABSTRACT_RTTI_IMPLEMENTATION(myClass) \
+    UClassID U_EXPORT2 myClass::getStaticClassID() { \
+        static char classID = 0; \
+        return (UClassID)&classID; \
+    }
+
+#endif  /* U_HIDE_INTERNAL_API */
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/urename.h b/thirdparty/icu4c/common/unicode/urename.h
new file mode 100644
index 0000000000..20232cd209
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/urename.h
@@ -0,0 +1,1922 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*
+*   file name:  urename.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   Created by: Perl script tools/genren.pl written by Vladimir Weinstein
+*
+*  Contains data for renaming ICU exports.
+*  Gets included by umachine.h
+*
+*  THIS FILE IS MACHINE-GENERATED, DON'T PLAY WITH IT IF YOU DON'T KNOW WHAT
+*  YOU ARE DOING, OTHERWISE VERY BAD THINGS WILL HAPPEN!
+*/
+
+#ifndef URENAME_H
+#define URENAME_H
+
+/* U_DISABLE_RENAMING can be defined in the following ways:
+ *   - when running configure, e.g.
+ *        runConfigureICU Linux --disable-renaming
+ *   - by changing the default setting of U_DISABLE_RENAMING in uconfig.h
+ */
+
+#include "unicode/uconfig.h"
+
+#if !U_DISABLE_RENAMING
+
+// Disable Renaming for Visual Studio's IntelliSense feature, so that 'Go-to-Definition' (F12) will work.
+#if !(defined(_MSC_VER) && defined(__INTELLISENSE__))
+
+/* We need the U_ICU_ENTRY_POINT_RENAME definition. There's a default one in unicode/uvernum.h we can use, but we will give
+   the platform a chance to define it first.
+   Normally (if utypes.h or umachine.h was included first) this will not be necessary as it will already be defined.
+ */
+
+#ifndef U_ICU_ENTRY_POINT_RENAME
+#include "unicode/umachine.h"
+#endif
+
+/* If we still don't have U_ICU_ENTRY_POINT_RENAME use the default. */
+#ifndef U_ICU_ENTRY_POINT_RENAME
+#include "unicode/uvernum.h"
+#endif
+
+/* Error out before the following defines cause very strange and unexpected code breakage */
+#ifndef U_ICU_ENTRY_POINT_RENAME
+#error U_ICU_ENTRY_POINT_RENAME is not defined - cannot continue. Consider defining U_DISABLE_RENAMING if renaming should not be used.
+#endif
+
+
+/* C exports renaming data */
+
+#define T_CString_int64ToString U_ICU_ENTRY_POINT_RENAME(T_CString_int64ToString)
+#define T_CString_integerToString U_ICU_ENTRY_POINT_RENAME(T_CString_integerToString)
+#define T_CString_stringToInteger U_ICU_ENTRY_POINT_RENAME(T_CString_stringToInteger)
+#define T_CString_toLowerCase U_ICU_ENTRY_POINT_RENAME(T_CString_toLowerCase)
+#define T_CString_toUpperCase U_ICU_ENTRY_POINT_RENAME(T_CString_toUpperCase)
+#define UCNV_FROM_U_CALLBACK_ESCAPE U_ICU_ENTRY_POINT_RENAME(UCNV_FROM_U_CALLBACK_ESCAPE)
+#define UCNV_FROM_U_CALLBACK_SKIP U_ICU_ENTRY_POINT_RENAME(UCNV_FROM_U_CALLBACK_SKIP)
+#define UCNV_FROM_U_CALLBACK_STOP U_ICU_ENTRY_POINT_RENAME(UCNV_FROM_U_CALLBACK_STOP)
+#define UCNV_FROM_U_CALLBACK_SUBSTITUTE U_ICU_ENTRY_POINT_RENAME(UCNV_FROM_U_CALLBACK_SUBSTITUTE)
+#define UCNV_TO_U_CALLBACK_ESCAPE U_ICU_ENTRY_POINT_RENAME(UCNV_TO_U_CALLBACK_ESCAPE)
+#define UCNV_TO_U_CALLBACK_SKIP U_ICU_ENTRY_POINT_RENAME(UCNV_TO_U_CALLBACK_SKIP)
+#define UCNV_TO_U_CALLBACK_STOP U_ICU_ENTRY_POINT_RENAME(UCNV_TO_U_CALLBACK_STOP)
+#define UCNV_TO_U_CALLBACK_SUBSTITUTE U_ICU_ENTRY_POINT_RENAME(UCNV_TO_U_CALLBACK_SUBSTITUTE)
+#define UDataMemory_createNewInstance U_ICU_ENTRY_POINT_RENAME(UDataMemory_createNewInstance)
+#define UDataMemory_init U_ICU_ENTRY_POINT_RENAME(UDataMemory_init)
+#define UDataMemory_isLoaded U_ICU_ENTRY_POINT_RENAME(UDataMemory_isLoaded)
+#define UDataMemory_normalizeDataPointer U_ICU_ENTRY_POINT_RENAME(UDataMemory_normalizeDataPointer)
+#define UDataMemory_setData U_ICU_ENTRY_POINT_RENAME(UDataMemory_setData)
+#define UDatamemory_assign U_ICU_ENTRY_POINT_RENAME(UDatamemory_assign)
+#define _ASCIIData U_ICU_ENTRY_POINT_RENAME(_ASCIIData)
+#define _Bocu1Data U_ICU_ENTRY_POINT_RENAME(_Bocu1Data)
+#define _CESU8Data U_ICU_ENTRY_POINT_RENAME(_CESU8Data)
+#define _CompoundTextData U_ICU_ENTRY_POINT_RENAME(_CompoundTextData)
+#define _HZData U_ICU_ENTRY_POINT_RENAME(_HZData)
+#define _IMAPData U_ICU_ENTRY_POINT_RENAME(_IMAPData)
+#define _ISCIIData U_ICU_ENTRY_POINT_RENAME(_ISCIIData)
+#define _ISO2022Data U_ICU_ENTRY_POINT_RENAME(_ISO2022Data)
+#define _LMBCSData1 U_ICU_ENTRY_POINT_RENAME(_LMBCSData1)
+#define _LMBCSData11 U_ICU_ENTRY_POINT_RENAME(_LMBCSData11)
+#define _LMBCSData16 U_ICU_ENTRY_POINT_RENAME(_LMBCSData16)
+#define _LMBCSData17 U_ICU_ENTRY_POINT_RENAME(_LMBCSData17)
+#define _LMBCSData18 U_ICU_ENTRY_POINT_RENAME(_LMBCSData18)
+#define _LMBCSData19 U_ICU_ENTRY_POINT_RENAME(_LMBCSData19)
+#define _LMBCSData2 U_ICU_ENTRY_POINT_RENAME(_LMBCSData2)
+#define _LMBCSData3 U_ICU_ENTRY_POINT_RENAME(_LMBCSData3)
+#define _LMBCSData4 U_ICU_ENTRY_POINT_RENAME(_LMBCSData4)
+#define _LMBCSData5 U_ICU_ENTRY_POINT_RENAME(_LMBCSData5)
+#define _LMBCSData6 U_ICU_ENTRY_POINT_RENAME(_LMBCSData6)
+#define _LMBCSData8 U_ICU_ENTRY_POINT_RENAME(_LMBCSData8)
+#define _Latin1Data U_ICU_ENTRY_POINT_RENAME(_Latin1Data)
+#define _MBCSData U_ICU_ENTRY_POINT_RENAME(_MBCSData)
+#define _SCSUData U_ICU_ENTRY_POINT_RENAME(_SCSUData)
+#define _UTF16BEData U_ICU_ENTRY_POINT_RENAME(_UTF16BEData)
+#define _UTF16Data U_ICU_ENTRY_POINT_RENAME(_UTF16Data)
+#define _UTF16LEData U_ICU_ENTRY_POINT_RENAME(_UTF16LEData)
+#define _UTF16v2Data U_ICU_ENTRY_POINT_RENAME(_UTF16v2Data)
+#define _UTF32BEData U_ICU_ENTRY_POINT_RENAME(_UTF32BEData)
+#define _UTF32Data U_ICU_ENTRY_POINT_RENAME(_UTF32Data)
+#define _UTF32LEData U_ICU_ENTRY_POINT_RENAME(_UTF32LEData)
+#define _UTF7Data U_ICU_ENTRY_POINT_RENAME(_UTF7Data)
+#define _UTF8Data U_ICU_ENTRY_POINT_RENAME(_UTF8Data)
+#define _isUnicodeLocaleTypeSubtag U_ICU_ENTRY_POINT_RENAME(_isUnicodeLocaleTypeSubtag)
+#define allowedHourFormatsCleanup U_ICU_ENTRY_POINT_RENAME(allowedHourFormatsCleanup)
+#define cmemory_cleanup U_ICU_ENTRY_POINT_RENAME(cmemory_cleanup)
+#define dayPeriodRulesCleanup U_ICU_ENTRY_POINT_RENAME(dayPeriodRulesCleanup)
+#define deleteAllowedHourFormats U_ICU_ENTRY_POINT_RENAME(deleteAllowedHourFormats)
+#define gTimeZoneFilesInitOnce U_ICU_ENTRY_POINT_RENAME(gTimeZoneFilesInitOnce)
+#define initNumsysNames U_ICU_ENTRY_POINT_RENAME(initNumsysNames)
+#define izrule_clone U_ICU_ENTRY_POINT_RENAME(izrule_clone)
+#define izrule_close U_ICU_ENTRY_POINT_RENAME(izrule_close)
+#define izrule_equals U_ICU_ENTRY_POINT_RENAME(izrule_equals)
+#define izrule_getDSTSavings U_ICU_ENTRY_POINT_RENAME(izrule_getDSTSavings)
+#define izrule_getDynamicClassID U_ICU_ENTRY_POINT_RENAME(izrule_getDynamicClassID)
+#define izrule_getFinalStart U_ICU_ENTRY_POINT_RENAME(izrule_getFinalStart)
+#define izrule_getFirstStart U_ICU_ENTRY_POINT_RENAME(izrule_getFirstStart)
+#define izrule_getName U_ICU_ENTRY_POINT_RENAME(izrule_getName)
+#define izrule_getNextStart U_ICU_ENTRY_POINT_RENAME(izrule_getNextStart)
+#define izrule_getPreviousStart U_ICU_ENTRY_POINT_RENAME(izrule_getPreviousStart)
+#define izrule_getRawOffset U_ICU_ENTRY_POINT_RENAME(izrule_getRawOffset)
+#define izrule_getStaticClassID U_ICU_ENTRY_POINT_RENAME(izrule_getStaticClassID)
+#define izrule_isEquivalentTo U_ICU_ENTRY_POINT_RENAME(izrule_isEquivalentTo)
+#define izrule_open U_ICU_ENTRY_POINT_RENAME(izrule_open)
+#define locale_getKeywordsStart U_ICU_ENTRY_POINT_RENAME(locale_getKeywordsStart)
+#define locale_get_default U_ICU_ENTRY_POINT_RENAME(locale_get_default)
+#define locale_set_default U_ICU_ENTRY_POINT_RENAME(locale_set_default)
+#define numSysCleanup U_ICU_ENTRY_POINT_RENAME(numSysCleanup)
+#define rbbi_cleanup U_ICU_ENTRY_POINT_RENAME(rbbi_cleanup)
+#define pl_addFontRun U_ICU_ENTRY_POINT_RENAME(pl_addFontRun)
+#define pl_addLocaleRun U_ICU_ENTRY_POINT_RENAME(pl_addLocaleRun)
+#define pl_addValueRun U_ICU_ENTRY_POINT_RENAME(pl_addValueRun)
+#define pl_close U_ICU_ENTRY_POINT_RENAME(pl_close)
+#define pl_closeFontRuns U_ICU_ENTRY_POINT_RENAME(pl_closeFontRuns)
+#define pl_closeLine U_ICU_ENTRY_POINT_RENAME(pl_closeLine)
+#define pl_closeLocaleRuns U_ICU_ENTRY_POINT_RENAME(pl_closeLocaleRuns)
+#define pl_closeValueRuns U_ICU_ENTRY_POINT_RENAME(pl_closeValueRuns)
+#define pl_countLineRuns U_ICU_ENTRY_POINT_RENAME(pl_countLineRuns)
+#define pl_create U_ICU_ENTRY_POINT_RENAME(pl_create)
+#define pl_getAscent U_ICU_ENTRY_POINT_RENAME(pl_getAscent)
+#define pl_getDescent U_ICU_ENTRY_POINT_RENAME(pl_getDescent)
+#define pl_getFontRunCount U_ICU_ENTRY_POINT_RENAME(pl_getFontRunCount)
+#define pl_getFontRunFont U_ICU_ENTRY_POINT_RENAME(pl_getFontRunFont)
+#define pl_getFontRunLastLimit U_ICU_ENTRY_POINT_RENAME(pl_getFontRunLastLimit)
+#define pl_getFontRunLimit U_ICU_ENTRY_POINT_RENAME(pl_getFontRunLimit)
+#define pl_getLeading U_ICU_ENTRY_POINT_RENAME(pl_getLeading)
+#define pl_getLineAscent U_ICU_ENTRY_POINT_RENAME(pl_getLineAscent)
+#define pl_getLineDescent U_ICU_ENTRY_POINT_RENAME(pl_getLineDescent)
+#define pl_getLineLeading U_ICU_ENTRY_POINT_RENAME(pl_getLineLeading)
+#define pl_getLineVisualRun U_ICU_ENTRY_POINT_RENAME(pl_getLineVisualRun)
+#define pl_getLineWidth U_ICU_ENTRY_POINT_RENAME(pl_getLineWidth)
+#define pl_getLocaleRunCount U_ICU_ENTRY_POINT_RENAME(pl_getLocaleRunCount)
+#define pl_getLocaleRunLastLimit U_ICU_ENTRY_POINT_RENAME(pl_getLocaleRunLastLimit)
+#define pl_getLocaleRunLimit U_ICU_ENTRY_POINT_RENAME(pl_getLocaleRunLimit)
+#define pl_getLocaleRunLocale U_ICU_ENTRY_POINT_RENAME(pl_getLocaleRunLocale)
+#define pl_getParagraphLevel U_ICU_ENTRY_POINT_RENAME(pl_getParagraphLevel)
+#define pl_getTextDirection U_ICU_ENTRY_POINT_RENAME(pl_getTextDirection)
+#define pl_getValueRunCount U_ICU_ENTRY_POINT_RENAME(pl_getValueRunCount)
+#define pl_getValueRunLastLimit U_ICU_ENTRY_POINT_RENAME(pl_getValueRunLastLimit)
+#define pl_getValueRunLimit U_ICU_ENTRY_POINT_RENAME(pl_getValueRunLimit)
+#define pl_getValueRunValue U_ICU_ENTRY_POINT_RENAME(pl_getValueRunValue)
+#define pl_getVisualRunAscent U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunAscent)
+#define pl_getVisualRunDescent U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunDescent)
+#define pl_getVisualRunDirection U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunDirection)
+#define pl_getVisualRunFont U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunFont)
+#define pl_getVisualRunGlyphCount U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunGlyphCount)
+#define pl_getVisualRunGlyphToCharMap U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunGlyphToCharMap)
+#define pl_getVisualRunGlyphs U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunGlyphs)
+#define pl_getVisualRunLeading U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunLeading)
+#define pl_getVisualRunPositions U_ICU_ENTRY_POINT_RENAME(pl_getVisualRunPositions)
+#define pl_isComplex U_ICU_ENTRY_POINT_RENAME(pl_isComplex)
+#define pl_nextLine U_ICU_ENTRY_POINT_RENAME(pl_nextLine)
+#define pl_openEmptyFontRuns U_ICU_ENTRY_POINT_RENAME(pl_openEmptyFontRuns)
+#define pl_openEmptyLocaleRuns U_ICU_ENTRY_POINT_RENAME(pl_openEmptyLocaleRuns)
+#define pl_openEmptyValueRuns U_ICU_ENTRY_POINT_RENAME(pl_openEmptyValueRuns)
+#define pl_openFontRuns U_ICU_ENTRY_POINT_RENAME(pl_openFontRuns)
+#define pl_openLocaleRuns U_ICU_ENTRY_POINT_RENAME(pl_openLocaleRuns)
+#define pl_openValueRuns U_ICU_ENTRY_POINT_RENAME(pl_openValueRuns)
+#define pl_reflow U_ICU_ENTRY_POINT_RENAME(pl_reflow)
+#define pl_resetFontRuns U_ICU_ENTRY_POINT_RENAME(pl_resetFontRuns)
+#define pl_resetLocaleRuns U_ICU_ENTRY_POINT_RENAME(pl_resetLocaleRuns)
+#define pl_resetValueRuns U_ICU_ENTRY_POINT_RENAME(pl_resetValueRuns)
+#define res_countArrayItems U_ICU_ENTRY_POINT_RENAME(res_countArrayItems)
+#define res_findResource U_ICU_ENTRY_POINT_RENAME(res_findResource)
+#define res_getAlias U_ICU_ENTRY_POINT_RENAME(res_getAlias)
+#define res_getArrayItem U_ICU_ENTRY_POINT_RENAME(res_getArrayItem)
+#define res_getBinaryNoTrace U_ICU_ENTRY_POINT_RENAME(res_getBinaryNoTrace)
+#define res_getIntVectorNoTrace U_ICU_ENTRY_POINT_RENAME(res_getIntVectorNoTrace)
+#define res_getPublicType U_ICU_ENTRY_POINT_RENAME(res_getPublicType)
+#define res_getResource U_ICU_ENTRY_POINT_RENAME(res_getResource)
+#define res_getStringNoTrace U_ICU_ENTRY_POINT_RENAME(res_getStringNoTrace)
+#define res_getTableItemByIndex U_ICU_ENTRY_POINT_RENAME(res_getTableItemByIndex)
+#define res_getTableItemByKey U_ICU_ENTRY_POINT_RENAME(res_getTableItemByKey)
+#define res_load U_ICU_ENTRY_POINT_RENAME(res_load)
+#define res_read U_ICU_ENTRY_POINT_RENAME(res_read)
+#define res_unload U_ICU_ENTRY_POINT_RENAME(res_unload)
+#define u_UCharsToChars U_ICU_ENTRY_POINT_RENAME(u_UCharsToChars)
+#define u_asciiToUpper U_ICU_ENTRY_POINT_RENAME(u_asciiToUpper)
+#define u_austrcpy U_ICU_ENTRY_POINT_RENAME(u_austrcpy)
+#define u_austrncpy U_ICU_ENTRY_POINT_RENAME(u_austrncpy)
+#define u_caseInsensitivePrefixMatch U_ICU_ENTRY_POINT_RENAME(u_caseInsensitivePrefixMatch)
+#define u_catclose U_ICU_ENTRY_POINT_RENAME(u_catclose)
+#define u_catgets U_ICU_ENTRY_POINT_RENAME(u_catgets)
+#define u_catopen U_ICU_ENTRY_POINT_RENAME(u_catopen)
+#define u_charAge U_ICU_ENTRY_POINT_RENAME(u_charAge)
+#define u_charDigitValue U_ICU_ENTRY_POINT_RENAME(u_charDigitValue)
+#define u_charDirection U_ICU_ENTRY_POINT_RENAME(u_charDirection)
+#define u_charFromName U_ICU_ENTRY_POINT_RENAME(u_charFromName)
+#define u_charMirror U_ICU_ENTRY_POINT_RENAME(u_charMirror)
+#define u_charName U_ICU_ENTRY_POINT_RENAME(u_charName)
+#define u_charType U_ICU_ENTRY_POINT_RENAME(u_charType)
+#define u_charsToUChars U_ICU_ENTRY_POINT_RENAME(u_charsToUChars)
+#define u_cleanup U_ICU_ENTRY_POINT_RENAME(u_cleanup)
+#define u_countChar32 U_ICU_ENTRY_POINT_RENAME(u_countChar32)
+#define u_digit U_ICU_ENTRY_POINT_RENAME(u_digit)
+#define u_enumCharNames U_ICU_ENTRY_POINT_RENAME(u_enumCharNames)
+#define u_enumCharTypes U_ICU_ENTRY_POINT_RENAME(u_enumCharTypes)
+#define u_errorName U_ICU_ENTRY_POINT_RENAME(u_errorName)
+#define u_fadopt U_ICU_ENTRY_POINT_RENAME(u_fadopt)
+#define u_fclose U_ICU_ENTRY_POINT_RENAME(u_fclose)
+#define u_feof U_ICU_ENTRY_POINT_RENAME(u_feof)
+#define u_fflush U_ICU_ENTRY_POINT_RENAME(u_fflush)
+#define u_fgetConverter U_ICU_ENTRY_POINT_RENAME(u_fgetConverter)
+#define u_fgetNumberFormat U_ICU_ENTRY_POINT_RENAME(u_fgetNumberFormat)
+#define u_fgetc U_ICU_ENTRY_POINT_RENAME(u_fgetc)
+#define u_fgetcodepage U_ICU_ENTRY_POINT_RENAME(u_fgetcodepage)
+#define u_fgetcx U_ICU_ENTRY_POINT_RENAME(u_fgetcx)
+#define u_fgetfile U_ICU_ENTRY_POINT_RENAME(u_fgetfile)
+#define u_fgetlocale U_ICU_ENTRY_POINT_RENAME(u_fgetlocale)
+#define u_fgets U_ICU_ENTRY_POINT_RENAME(u_fgets)
+#define u_file_read U_ICU_ENTRY_POINT_RENAME(u_file_read)
+#define u_file_write U_ICU_ENTRY_POINT_RENAME(u_file_write)
+#define u_file_write_flush U_ICU_ENTRY_POINT_RENAME(u_file_write_flush)
+#define u_finit U_ICU_ENTRY_POINT_RENAME(u_finit)
+#define u_flushDefaultConverter U_ICU_ENTRY_POINT_RENAME(u_flushDefaultConverter)
+#define u_foldCase U_ICU_ENTRY_POINT_RENAME(u_foldCase)
+#define u_fopen U_ICU_ENTRY_POINT_RENAME(u_fopen)
+#define u_fopen_u U_ICU_ENTRY_POINT_RENAME(u_fopen_u)
+#define u_forDigit U_ICU_ENTRY_POINT_RENAME(u_forDigit)
+#define u_formatMessage U_ICU_ENTRY_POINT_RENAME(u_formatMessage)
+#define u_formatMessageWithError U_ICU_ENTRY_POINT_RENAME(u_formatMessageWithError)
+#define u_fprintf U_ICU_ENTRY_POINT_RENAME(u_fprintf)
+#define u_fprintf_u U_ICU_ENTRY_POINT_RENAME(u_fprintf_u)
+#define u_fputc U_ICU_ENTRY_POINT_RENAME(u_fputc)
+#define u_fputs U_ICU_ENTRY_POINT_RENAME(u_fputs)
+#define u_frewind U_ICU_ENTRY_POINT_RENAME(u_frewind)
+#define u_fscanf U_ICU_ENTRY_POINT_RENAME(u_fscanf)
+#define u_fscanf_u U_ICU_ENTRY_POINT_RENAME(u_fscanf_u)
+#define u_fsetcodepage U_ICU_ENTRY_POINT_RENAME(u_fsetcodepage)
+#define u_fsetlocale U_ICU_ENTRY_POINT_RENAME(u_fsetlocale)
+#define u_fsettransliterator U_ICU_ENTRY_POINT_RENAME(u_fsettransliterator)
+#define u_fstropen U_ICU_ENTRY_POINT_RENAME(u_fstropen)
+#define u_fungetc U_ICU_ENTRY_POINT_RENAME(u_fungetc)
+#define u_getBidiPairedBracket U_ICU_ENTRY_POINT_RENAME(u_getBidiPairedBracket)
+#define u_getBinaryPropertySet U_ICU_ENTRY_POINT_RENAME(u_getBinaryPropertySet)
+#define u_getCombiningClass U_ICU_ENTRY_POINT_RENAME(u_getCombiningClass)
+#define u_getDataDirectory U_ICU_ENTRY_POINT_RENAME(u_getDataDirectory)
+#define u_getDataVersion U_ICU_ENTRY_POINT_RENAME(u_getDataVersion)
+#define u_getDefaultConverter U_ICU_ENTRY_POINT_RENAME(u_getDefaultConverter)
+#define u_getFC_NFKC_Closure U_ICU_ENTRY_POINT_RENAME(u_getFC_NFKC_Closure)
+#define u_getISOComment U_ICU_ENTRY_POINT_RENAME(u_getISOComment)
+#define u_getIntPropertyMap U_ICU_ENTRY_POINT_RENAME(u_getIntPropertyMap)
+#define u_getIntPropertyMaxValue U_ICU_ENTRY_POINT_RENAME(u_getIntPropertyMaxValue)
+#define u_getIntPropertyMinValue U_ICU_ENTRY_POINT_RENAME(u_getIntPropertyMinValue)
+#define u_getIntPropertyValue U_ICU_ENTRY_POINT_RENAME(u_getIntPropertyValue)
+#define u_getMainProperties U_ICU_ENTRY_POINT_RENAME(u_getMainProperties)
+#define u_getNumericValue U_ICU_ENTRY_POINT_RENAME(u_getNumericValue)
+#define u_getPropertyEnum U_ICU_ENTRY_POINT_RENAME(u_getPropertyEnum)
+#define u_getPropertyName U_ICU_ENTRY_POINT_RENAME(u_getPropertyName)
+#define u_getPropertyValueEnum U_ICU_ENTRY_POINT_RENAME(u_getPropertyValueEnum)
+#define u_getPropertyValueName U_ICU_ENTRY_POINT_RENAME(u_getPropertyValueName)
+#define u_getTimeZoneFilesDirectory U_ICU_ENTRY_POINT_RENAME(u_getTimeZoneFilesDirectory)
+#define u_getUnicodeProperties U_ICU_ENTRY_POINT_RENAME(u_getUnicodeProperties)
+#define u_getUnicodeVersion U_ICU_ENTRY_POINT_RENAME(u_getUnicodeVersion)
+#define u_getVersion U_ICU_ENTRY_POINT_RENAME(u_getVersion)
+#define u_get_stdout U_ICU_ENTRY_POINT_RENAME(u_get_stdout)
+#define u_hasBinaryProperty U_ICU_ENTRY_POINT_RENAME(u_hasBinaryProperty)
+#define u_init U_ICU_ENTRY_POINT_RENAME(u_init)
+#define u_isIDIgnorable U_ICU_ENTRY_POINT_RENAME(u_isIDIgnorable)
+#define u_isIDPart U_ICU_ENTRY_POINT_RENAME(u_isIDPart)
+#define u_isIDStart U_ICU_ENTRY_POINT_RENAME(u_isIDStart)
+#define u_isISOControl U_ICU_ENTRY_POINT_RENAME(u_isISOControl)
+#define u_isJavaIDPart U_ICU_ENTRY_POINT_RENAME(u_isJavaIDPart)
+#define u_isJavaIDStart U_ICU_ENTRY_POINT_RENAME(u_isJavaIDStart)
+#define u_isJavaSpaceChar U_ICU_ENTRY_POINT_RENAME(u_isJavaSpaceChar)
+#define u_isMirrored U_ICU_ENTRY_POINT_RENAME(u_isMirrored)
+#define u_isUAlphabetic U_ICU_ENTRY_POINT_RENAME(u_isUAlphabetic)
+#define u_isULowercase U_ICU_ENTRY_POINT_RENAME(u_isULowercase)
+#define u_isUUppercase U_ICU_ENTRY_POINT_RENAME(u_isUUppercase)
+#define u_isUWhiteSpace U_ICU_ENTRY_POINT_RENAME(u_isUWhiteSpace)
+#define u_isWhitespace U_ICU_ENTRY_POINT_RENAME(u_isWhitespace)
+#define u_isalnum U_ICU_ENTRY_POINT_RENAME(u_isalnum)
+#define u_isalnumPOSIX U_ICU_ENTRY_POINT_RENAME(u_isalnumPOSIX)
+#define u_isalpha U_ICU_ENTRY_POINT_RENAME(u_isalpha)
+#define u_isbase U_ICU_ENTRY_POINT_RENAME(u_isbase)
+#define u_isblank U_ICU_ENTRY_POINT_RENAME(u_isblank)
+#define u_iscntrl U_ICU_ENTRY_POINT_RENAME(u_iscntrl)
+#define u_isdefined U_ICU_ENTRY_POINT_RENAME(u_isdefined)
+#define u_isdigit U_ICU_ENTRY_POINT_RENAME(u_isdigit)
+#define u_isgraph U_ICU_ENTRY_POINT_RENAME(u_isgraph)
+#define u_isgraphPOSIX U_ICU_ENTRY_POINT_RENAME(u_isgraphPOSIX)
+#define u_islower U_ICU_ENTRY_POINT_RENAME(u_islower)
+#define u_isprint U_ICU_ENTRY_POINT_RENAME(u_isprint)
+#define u_isprintPOSIX U_ICU_ENTRY_POINT_RENAME(u_isprintPOSIX)
+#define u_ispunct U_ICU_ENTRY_POINT_RENAME(u_ispunct)
+#define u_isspace U_ICU_ENTRY_POINT_RENAME(u_isspace)
+#define u_istitle U_ICU_ENTRY_POINT_RENAME(u_istitle)
+#define u_isupper U_ICU_ENTRY_POINT_RENAME(u_isupper)
+#define u_isxdigit U_ICU_ENTRY_POINT_RENAME(u_isxdigit)
+#define u_locbund_close U_ICU_ENTRY_POINT_RENAME(u_locbund_close)
+#define u_locbund_getNumberFormat U_ICU_ENTRY_POINT_RENAME(u_locbund_getNumberFormat)
+#define u_locbund_init U_ICU_ENTRY_POINT_RENAME(u_locbund_init)
+#define u_memcasecmp U_ICU_ENTRY_POINT_RENAME(u_memcasecmp)
+#define u_memchr U_ICU_ENTRY_POINT_RENAME(u_memchr)
+#define u_memchr32 U_ICU_ENTRY_POINT_RENAME(u_memchr32)
+#define u_memcmp U_ICU_ENTRY_POINT_RENAME(u_memcmp)
+#define u_memcmpCodePointOrder U_ICU_ENTRY_POINT_RENAME(u_memcmpCodePointOrder)
+#define u_memcpy U_ICU_ENTRY_POINT_RENAME(u_memcpy)
+#define u_memmove U_ICU_ENTRY_POINT_RENAME(u_memmove)
+#define u_memrchr U_ICU_ENTRY_POINT_RENAME(u_memrchr)
+#define u_memrchr32 U_ICU_ENTRY_POINT_RENAME(u_memrchr32)
+#define u_memset U_ICU_ENTRY_POINT_RENAME(u_memset)
+#define u_parseMessage U_ICU_ENTRY_POINT_RENAME(u_parseMessage)
+#define u_parseMessageWithError U_ICU_ENTRY_POINT_RENAME(u_parseMessageWithError)
+#define u_printf U_ICU_ENTRY_POINT_RENAME(u_printf)
+#define u_printf_parse U_ICU_ENTRY_POINT_RENAME(u_printf_parse)
+#define u_printf_u U_ICU_ENTRY_POINT_RENAME(u_printf_u)
+#define u_releaseDefaultConverter U_ICU_ENTRY_POINT_RENAME(u_releaseDefaultConverter)
+#define u_scanf_parse U_ICU_ENTRY_POINT_RENAME(u_scanf_parse)
+#define u_setAtomicIncDecFunctions U_ICU_ENTRY_POINT_RENAME(u_setAtomicIncDecFunctions)
+#define u_setDataDirectory U_ICU_ENTRY_POINT_RENAME(u_setDataDirectory)
+#define u_setMemoryFunctions U_ICU_ENTRY_POINT_RENAME(u_setMemoryFunctions)
+#define u_setMutexFunctions U_ICU_ENTRY_POINT_RENAME(u_setMutexFunctions)
+#define u_setTimeZoneFilesDirectory U_ICU_ENTRY_POINT_RENAME(u_setTimeZoneFilesDirectory)
+#define u_shapeArabic U_ICU_ENTRY_POINT_RENAME(u_shapeArabic)
+#define u_snprintf U_ICU_ENTRY_POINT_RENAME(u_snprintf)
+#define u_snprintf_u U_ICU_ENTRY_POINT_RENAME(u_snprintf_u)
+#define u_sprintf U_ICU_ENTRY_POINT_RENAME(u_sprintf)
+#define u_sprintf_u U_ICU_ENTRY_POINT_RENAME(u_sprintf_u)
+#define u_sscanf U_ICU_ENTRY_POINT_RENAME(u_sscanf)
+#define u_sscanf_u U_ICU_ENTRY_POINT_RENAME(u_sscanf_u)
+#define u_strCaseCompare U_ICU_ENTRY_POINT_RENAME(u_strCaseCompare)
+#define u_strCompare U_ICU_ENTRY_POINT_RENAME(u_strCompare)
+#define u_strCompareIter U_ICU_ENTRY_POINT_RENAME(u_strCompareIter)
+#define u_strFindFirst U_ICU_ENTRY_POINT_RENAME(u_strFindFirst)
+#define u_strFindLast U_ICU_ENTRY_POINT_RENAME(u_strFindLast)
+#define u_strFoldCase U_ICU_ENTRY_POINT_RENAME(u_strFoldCase)
+#define u_strFromJavaModifiedUTF8WithSub U_ICU_ENTRY_POINT_RENAME(u_strFromJavaModifiedUTF8WithSub)
+#define u_strFromPunycode U_ICU_ENTRY_POINT_RENAME(u_strFromPunycode)
+#define u_strFromUTF32 U_ICU_ENTRY_POINT_RENAME(u_strFromUTF32)
+#define u_strFromUTF32WithSub U_ICU_ENTRY_POINT_RENAME(u_strFromUTF32WithSub)
+#define u_strFromUTF8 U_ICU_ENTRY_POINT_RENAME(u_strFromUTF8)
+#define u_strFromUTF8Lenient U_ICU_ENTRY_POINT_RENAME(u_strFromUTF8Lenient)
+#define u_strFromUTF8WithSub U_ICU_ENTRY_POINT_RENAME(u_strFromUTF8WithSub)
+#define u_strFromWCS U_ICU_ENTRY_POINT_RENAME(u_strFromWCS)
+#define u_strHasMoreChar32Than U_ICU_ENTRY_POINT_RENAME(u_strHasMoreChar32Than)
+#define u_strToJavaModifiedUTF8 U_ICU_ENTRY_POINT_RENAME(u_strToJavaModifiedUTF8)
+#define u_strToLower U_ICU_ENTRY_POINT_RENAME(u_strToLower)
+#define u_strToPunycode U_ICU_ENTRY_POINT_RENAME(u_strToPunycode)
+#define u_strToTitle U_ICU_ENTRY_POINT_RENAME(u_strToTitle)
+#define u_strToUTF32 U_ICU_ENTRY_POINT_RENAME(u_strToUTF32)
+#define u_strToUTF32WithSub U_ICU_ENTRY_POINT_RENAME(u_strToUTF32WithSub)
+#define u_strToUTF8 U_ICU_ENTRY_POINT_RENAME(u_strToUTF8)
+#define u_strToUTF8WithSub U_ICU_ENTRY_POINT_RENAME(u_strToUTF8WithSub)
+#define u_strToUpper U_ICU_ENTRY_POINT_RENAME(u_strToUpper)
+#define u_strToWCS U_ICU_ENTRY_POINT_RENAME(u_strToWCS)
+#define u_strcasecmp U_ICU_ENTRY_POINT_RENAME(u_strcasecmp)
+#define u_strcat U_ICU_ENTRY_POINT_RENAME(u_strcat)
+#define u_strchr U_ICU_ENTRY_POINT_RENAME(u_strchr)
+#define u_strchr32 U_ICU_ENTRY_POINT_RENAME(u_strchr32)
+#define u_strcmp U_ICU_ENTRY_POINT_RENAME(u_strcmp)
+#define u_strcmpCodePointOrder U_ICU_ENTRY_POINT_RENAME(u_strcmpCodePointOrder)
+#define u_strcmpFold U_ICU_ENTRY_POINT_RENAME(u_strcmpFold)
+#define u_strcpy U_ICU_ENTRY_POINT_RENAME(u_strcpy)
+#define u_strcspn U_ICU_ENTRY_POINT_RENAME(u_strcspn)
+#define u_strlen U_ICU_ENTRY_POINT_RENAME(u_strlen)
+#define u_strncasecmp U_ICU_ENTRY_POINT_RENAME(u_strncasecmp)
+#define u_strncat U_ICU_ENTRY_POINT_RENAME(u_strncat)
+#define u_strncmp U_ICU_ENTRY_POINT_RENAME(u_strncmp)
+#define u_strncmpCodePointOrder U_ICU_ENTRY_POINT_RENAME(u_strncmpCodePointOrder)
+#define u_strncpy U_ICU_ENTRY_POINT_RENAME(u_strncpy)
+#define u_strpbrk U_ICU_ENTRY_POINT_RENAME(u_strpbrk)
+#define u_strrchr U_ICU_ENTRY_POINT_RENAME(u_strrchr)
+#define u_strrchr32 U_ICU_ENTRY_POINT_RENAME(u_strrchr32)
+#define u_strrstr U_ICU_ENTRY_POINT_RENAME(u_strrstr)
+#define u_strspn U_ICU_ENTRY_POINT_RENAME(u_strspn)
+#define u_strstr U_ICU_ENTRY_POINT_RENAME(u_strstr)
+#define u_strtok_r U_ICU_ENTRY_POINT_RENAME(u_strtok_r)
+#define u_terminateChars U_ICU_ENTRY_POINT_RENAME(u_terminateChars)
+#define u_terminateUChar32s U_ICU_ENTRY_POINT_RENAME(u_terminateUChar32s)
+#define u_terminateUChars U_ICU_ENTRY_POINT_RENAME(u_terminateUChars)
+#define u_terminateWChars U_ICU_ENTRY_POINT_RENAME(u_terminateWChars)
+#define u_tolower U_ICU_ENTRY_POINT_RENAME(u_tolower)
+#define u_totitle U_ICU_ENTRY_POINT_RENAME(u_totitle)
+#define u_toupper U_ICU_ENTRY_POINT_RENAME(u_toupper)
+#define u_uastrcpy U_ICU_ENTRY_POINT_RENAME(u_uastrcpy)
+#define u_uastrncpy U_ICU_ENTRY_POINT_RENAME(u_uastrncpy)
+#define u_unescape U_ICU_ENTRY_POINT_RENAME(u_unescape)
+#define u_unescapeAt U_ICU_ENTRY_POINT_RENAME(u_unescapeAt)
+#define u_versionFromString U_ICU_ENTRY_POINT_RENAME(u_versionFromString)
+#define u_versionFromUString U_ICU_ENTRY_POINT_RENAME(u_versionFromUString)
+#define u_versionToString U_ICU_ENTRY_POINT_RENAME(u_versionToString)
+#define u_vformatMessage U_ICU_ENTRY_POINT_RENAME(u_vformatMessage)
+#define u_vformatMessageWithError U_ICU_ENTRY_POINT_RENAME(u_vformatMessageWithError)
+#define u_vfprintf U_ICU_ENTRY_POINT_RENAME(u_vfprintf)
+#define u_vfprintf_u U_ICU_ENTRY_POINT_RENAME(u_vfprintf_u)
+#define u_vfscanf U_ICU_ENTRY_POINT_RENAME(u_vfscanf)
+#define u_vfscanf_u U_ICU_ENTRY_POINT_RENAME(u_vfscanf_u)
+#define u_vparseMessage U_ICU_ENTRY_POINT_RENAME(u_vparseMessage)
+#define u_vparseMessageWithError U_ICU_ENTRY_POINT_RENAME(u_vparseMessageWithError)
+#define u_vsnprintf U_ICU_ENTRY_POINT_RENAME(u_vsnprintf)
+#define u_vsnprintf_u U_ICU_ENTRY_POINT_RENAME(u_vsnprintf_u)
+#define u_vsprintf U_ICU_ENTRY_POINT_RENAME(u_vsprintf)
+#define u_vsprintf_u U_ICU_ENTRY_POINT_RENAME(u_vsprintf_u)
+#define u_vsscanf U_ICU_ENTRY_POINT_RENAME(u_vsscanf)
+#define u_vsscanf_u U_ICU_ENTRY_POINT_RENAME(u_vsscanf_u)
+#define u_writeIdenticalLevelRun U_ICU_ENTRY_POINT_RENAME(u_writeIdenticalLevelRun)
+#define ubidi_addPropertyStarts U_ICU_ENTRY_POINT_RENAME(ubidi_addPropertyStarts)
+#define ubidi_close U_ICU_ENTRY_POINT_RENAME(ubidi_close)
+#define ubidi_countParagraphs U_ICU_ENTRY_POINT_RENAME(ubidi_countParagraphs)
+#define ubidi_countRuns U_ICU_ENTRY_POINT_RENAME(ubidi_countRuns)
+#define ubidi_getBaseDirection U_ICU_ENTRY_POINT_RENAME(ubidi_getBaseDirection)
+#define ubidi_getClass U_ICU_ENTRY_POINT_RENAME(ubidi_getClass)
+#define ubidi_getClassCallback U_ICU_ENTRY_POINT_RENAME(ubidi_getClassCallback)
+#define ubidi_getCustomizedClass U_ICU_ENTRY_POINT_RENAME(ubidi_getCustomizedClass)
+#define ubidi_getDirection U_ICU_ENTRY_POINT_RENAME(ubidi_getDirection)
+#define ubidi_getJoiningGroup U_ICU_ENTRY_POINT_RENAME(ubidi_getJoiningGroup)
+#define ubidi_getJoiningType U_ICU_ENTRY_POINT_RENAME(ubidi_getJoiningType)
+#define ubidi_getLength U_ICU_ENTRY_POINT_RENAME(ubidi_getLength)
+#define ubidi_getLevelAt U_ICU_ENTRY_POINT_RENAME(ubidi_getLevelAt)
+#define ubidi_getLevels U_ICU_ENTRY_POINT_RENAME(ubidi_getLevels)
+#define ubidi_getLogicalIndex U_ICU_ENTRY_POINT_RENAME(ubidi_getLogicalIndex)
+#define ubidi_getLogicalMap U_ICU_ENTRY_POINT_RENAME(ubidi_getLogicalMap)
+#define ubidi_getLogicalRun U_ICU_ENTRY_POINT_RENAME(ubidi_getLogicalRun)
+#define ubidi_getMaxValue U_ICU_ENTRY_POINT_RENAME(ubidi_getMaxValue)
+#define ubidi_getMemory U_ICU_ENTRY_POINT_RENAME(ubidi_getMemory)
+#define ubidi_getMirror U_ICU_ENTRY_POINT_RENAME(ubidi_getMirror)
+#define ubidi_getPairedBracket U_ICU_ENTRY_POINT_RENAME(ubidi_getPairedBracket)
+#define ubidi_getPairedBracketType U_ICU_ENTRY_POINT_RENAME(ubidi_getPairedBracketType)
+#define ubidi_getParaLevel U_ICU_ENTRY_POINT_RENAME(ubidi_getParaLevel)
+#define ubidi_getParaLevelAtIndex U_ICU_ENTRY_POINT_RENAME(ubidi_getParaLevelAtIndex)
+#define ubidi_getParagraph U_ICU_ENTRY_POINT_RENAME(ubidi_getParagraph)
+#define ubidi_getParagraphByIndex U_ICU_ENTRY_POINT_RENAME(ubidi_getParagraphByIndex)
+#define ubidi_getProcessedLength U_ICU_ENTRY_POINT_RENAME(ubidi_getProcessedLength)
+#define ubidi_getReorderingMode U_ICU_ENTRY_POINT_RENAME(ubidi_getReorderingMode)
+#define ubidi_getReorderingOptions U_ICU_ENTRY_POINT_RENAME(ubidi_getReorderingOptions)
+#define ubidi_getResultLength U_ICU_ENTRY_POINT_RENAME(ubidi_getResultLength)
+#define ubidi_getRuns U_ICU_ENTRY_POINT_RENAME(ubidi_getRuns)
+#define ubidi_getText U_ICU_ENTRY_POINT_RENAME(ubidi_getText)
+#define ubidi_getVisualIndex U_ICU_ENTRY_POINT_RENAME(ubidi_getVisualIndex)
+#define ubidi_getVisualMap U_ICU_ENTRY_POINT_RENAME(ubidi_getVisualMap)
+#define ubidi_getVisualRun U_ICU_ENTRY_POINT_RENAME(ubidi_getVisualRun)
+#define ubidi_invertMap U_ICU_ENTRY_POINT_RENAME(ubidi_invertMap)
+#define ubidi_isBidiControl U_ICU_ENTRY_POINT_RENAME(ubidi_isBidiControl)
+#define ubidi_isInverse U_ICU_ENTRY_POINT_RENAME(ubidi_isInverse)
+#define ubidi_isJoinControl U_ICU_ENTRY_POINT_RENAME(ubidi_isJoinControl)
+#define ubidi_isMirrored U_ICU_ENTRY_POINT_RENAME(ubidi_isMirrored)
+#define ubidi_isOrderParagraphsLTR U_ICU_ENTRY_POINT_RENAME(ubidi_isOrderParagraphsLTR)
+#define ubidi_open U_ICU_ENTRY_POINT_RENAME(ubidi_open)
+#define ubidi_openSized U_ICU_ENTRY_POINT_RENAME(ubidi_openSized)
+#define ubidi_orderParagraphsLTR U_ICU_ENTRY_POINT_RENAME(ubidi_orderParagraphsLTR)
+#define ubidi_reorderLogical U_ICU_ENTRY_POINT_RENAME(ubidi_reorderLogical)
+#define ubidi_reorderVisual U_ICU_ENTRY_POINT_RENAME(ubidi_reorderVisual)
+#define ubidi_setClassCallback U_ICU_ENTRY_POINT_RENAME(ubidi_setClassCallback)
+#define ubidi_setContext U_ICU_ENTRY_POINT_RENAME(ubidi_setContext)
+#define ubidi_setInverse U_ICU_ENTRY_POINT_RENAME(ubidi_setInverse)
+#define ubidi_setLine U_ICU_ENTRY_POINT_RENAME(ubidi_setLine)
+#define ubidi_setPara U_ICU_ENTRY_POINT_RENAME(ubidi_setPara)
+#define ubidi_setReorderingMode U_ICU_ENTRY_POINT_RENAME(ubidi_setReorderingMode)
+#define ubidi_setReorderingOptions U_ICU_ENTRY_POINT_RENAME(ubidi_setReorderingOptions)
+#define ubidi_writeReordered U_ICU_ENTRY_POINT_RENAME(ubidi_writeReordered)
+#define ubidi_writeReverse U_ICU_ENTRY_POINT_RENAME(ubidi_writeReverse)
+#define ubiditransform_close U_ICU_ENTRY_POINT_RENAME(ubiditransform_close)
+#define ubiditransform_open U_ICU_ENTRY_POINT_RENAME(ubiditransform_open)
+#define ubiditransform_transform U_ICU_ENTRY_POINT_RENAME(ubiditransform_transform)
+#define ublock_getCode U_ICU_ENTRY_POINT_RENAME(ublock_getCode)
+#define ubrk_close U_ICU_ENTRY_POINT_RENAME(ubrk_close)
+#define ubrk_countAvailable U_ICU_ENTRY_POINT_RENAME(ubrk_countAvailable)
+#define ubrk_current U_ICU_ENTRY_POINT_RENAME(ubrk_current)
+#define ubrk_first U_ICU_ENTRY_POINT_RENAME(ubrk_first)
+#define ubrk_following U_ICU_ENTRY_POINT_RENAME(ubrk_following)
+#define ubrk_getAvailable U_ICU_ENTRY_POINT_RENAME(ubrk_getAvailable)
+#define ubrk_getBinaryRules U_ICU_ENTRY_POINT_RENAME(ubrk_getBinaryRules)
+#define ubrk_getLocaleByType U_ICU_ENTRY_POINT_RENAME(ubrk_getLocaleByType)
+#define ubrk_getRuleStatus U_ICU_ENTRY_POINT_RENAME(ubrk_getRuleStatus)
+#define ubrk_getRuleStatusVec U_ICU_ENTRY_POINT_RENAME(ubrk_getRuleStatusVec)
+#define ubrk_isBoundary U_ICU_ENTRY_POINT_RENAME(ubrk_isBoundary)
+#define ubrk_last U_ICU_ENTRY_POINT_RENAME(ubrk_last)
+#define ubrk_next U_ICU_ENTRY_POINT_RENAME(ubrk_next)
+#define ubrk_open U_ICU_ENTRY_POINT_RENAME(ubrk_open)
+#define ubrk_openBinaryRules U_ICU_ENTRY_POINT_RENAME(ubrk_openBinaryRules)
+#define ubrk_openRules U_ICU_ENTRY_POINT_RENAME(ubrk_openRules)
+#define ubrk_preceding U_ICU_ENTRY_POINT_RENAME(ubrk_preceding)
+#define ubrk_previous U_ICU_ENTRY_POINT_RENAME(ubrk_previous)
+#define ubrk_refreshUText U_ICU_ENTRY_POINT_RENAME(ubrk_refreshUText)
+#define ubrk_safeClone U_ICU_ENTRY_POINT_RENAME(ubrk_safeClone)
+#define ubrk_setText U_ICU_ENTRY_POINT_RENAME(ubrk_setText)
+#define ubrk_setUText U_ICU_ENTRY_POINT_RENAME(ubrk_setUText)
+#define ubrk_swap U_ICU_ENTRY_POINT_RENAME(ubrk_swap)
+#define ucache_compareKeys U_ICU_ENTRY_POINT_RENAME(ucache_compareKeys)
+#define ucache_deleteKey U_ICU_ENTRY_POINT_RENAME(ucache_deleteKey)
+#define ucache_hashKeys U_ICU_ENTRY_POINT_RENAME(ucache_hashKeys)
+#define ucal_add U_ICU_ENTRY_POINT_RENAME(ucal_add)
+#define ucal_clear U_ICU_ENTRY_POINT_RENAME(ucal_clear)
+#define ucal_clearField U_ICU_ENTRY_POINT_RENAME(ucal_clearField)
+#define ucal_clone U_ICU_ENTRY_POINT_RENAME(ucal_clone)
+#define ucal_close U_ICU_ENTRY_POINT_RENAME(ucal_close)
+#define ucal_countAvailable U_ICU_ENTRY_POINT_RENAME(ucal_countAvailable)
+#define ucal_equivalentTo U_ICU_ENTRY_POINT_RENAME(ucal_equivalentTo)
+#define ucal_get U_ICU_ENTRY_POINT_RENAME(ucal_get)
+#define ucal_getAttribute U_ICU_ENTRY_POINT_RENAME(ucal_getAttribute)
+#define ucal_getAvailable U_ICU_ENTRY_POINT_RENAME(ucal_getAvailable)
+#define ucal_getCanonicalTimeZoneID U_ICU_ENTRY_POINT_RENAME(ucal_getCanonicalTimeZoneID)
+#define ucal_getDSTSavings U_ICU_ENTRY_POINT_RENAME(ucal_getDSTSavings)
+#define ucal_getDayOfWeekType U_ICU_ENTRY_POINT_RENAME(ucal_getDayOfWeekType)
+#define ucal_getDefaultTimeZone U_ICU_ENTRY_POINT_RENAME(ucal_getDefaultTimeZone)
+#define ucal_getFieldDifference U_ICU_ENTRY_POINT_RENAME(ucal_getFieldDifference)
+#define ucal_getGregorianChange U_ICU_ENTRY_POINT_RENAME(ucal_getGregorianChange)
+#define ucal_getHostTimeZone U_ICU_ENTRY_POINT_RENAME(ucal_getHostTimeZone)
+#define ucal_getKeywordValuesForLocale U_ICU_ENTRY_POINT_RENAME(ucal_getKeywordValuesForLocale)
+#define ucal_getLimit U_ICU_ENTRY_POINT_RENAME(ucal_getLimit)
+#define ucal_getLocaleByType U_ICU_ENTRY_POINT_RENAME(ucal_getLocaleByType)
+#define ucal_getMillis U_ICU_ENTRY_POINT_RENAME(ucal_getMillis)
+#define ucal_getNow U_ICU_ENTRY_POINT_RENAME(ucal_getNow)
+#define ucal_getTZDataVersion U_ICU_ENTRY_POINT_RENAME(ucal_getTZDataVersion)
+#define ucal_getTimeZoneDisplayName U_ICU_ENTRY_POINT_RENAME(ucal_getTimeZoneDisplayName)
+#define ucal_getTimeZoneID U_ICU_ENTRY_POINT_RENAME(ucal_getTimeZoneID)
+#define ucal_getTimeZoneIDForWindowsID U_ICU_ENTRY_POINT_RENAME(ucal_getTimeZoneIDForWindowsID)
+#define ucal_getTimeZoneTransitionDate U_ICU_ENTRY_POINT_RENAME(ucal_getTimeZoneTransitionDate)
+#define ucal_getType U_ICU_ENTRY_POINT_RENAME(ucal_getType)
+#define ucal_getWeekendTransition U_ICU_ENTRY_POINT_RENAME(ucal_getWeekendTransition)
+#define ucal_getWindowsTimeZoneID U_ICU_ENTRY_POINT_RENAME(ucal_getWindowsTimeZoneID)
+#define ucal_inDaylightTime U_ICU_ENTRY_POINT_RENAME(ucal_inDaylightTime)
+#define ucal_isSet U_ICU_ENTRY_POINT_RENAME(ucal_isSet)
+#define ucal_isWeekend U_ICU_ENTRY_POINT_RENAME(ucal_isWeekend)
+#define ucal_open U_ICU_ENTRY_POINT_RENAME(ucal_open)
+#define ucal_openCountryTimeZones U_ICU_ENTRY_POINT_RENAME(ucal_openCountryTimeZones)
+#define ucal_openTimeZoneIDEnumeration U_ICU_ENTRY_POINT_RENAME(ucal_openTimeZoneIDEnumeration)
+#define ucal_openTimeZones U_ICU_ENTRY_POINT_RENAME(ucal_openTimeZones)
+#define ucal_roll U_ICU_ENTRY_POINT_RENAME(ucal_roll)
+#define ucal_set U_ICU_ENTRY_POINT_RENAME(ucal_set)
+#define ucal_setAttribute U_ICU_ENTRY_POINT_RENAME(ucal_setAttribute)
+#define ucal_setDate U_ICU_ENTRY_POINT_RENAME(ucal_setDate)
+#define ucal_setDateTime U_ICU_ENTRY_POINT_RENAME(ucal_setDateTime)
+#define ucal_setDefaultTimeZone U_ICU_ENTRY_POINT_RENAME(ucal_setDefaultTimeZone)
+#define ucal_setGregorianChange U_ICU_ENTRY_POINT_RENAME(ucal_setGregorianChange)
+#define ucal_setMillis U_ICU_ENTRY_POINT_RENAME(ucal_setMillis)
+#define ucal_setTimeZone U_ICU_ENTRY_POINT_RENAME(ucal_setTimeZone)
+#define ucase_addCaseClosure U_ICU_ENTRY_POINT_RENAME(ucase_addCaseClosure)
+#define ucase_addPropertyStarts U_ICU_ENTRY_POINT_RENAME(ucase_addPropertyStarts)
+#define ucase_addStringCaseClosure U_ICU_ENTRY_POINT_RENAME(ucase_addStringCaseClosure)
+#define ucase_fold U_ICU_ENTRY_POINT_RENAME(ucase_fold)
+#define ucase_getCaseLocale U_ICU_ENTRY_POINT_RENAME(ucase_getCaseLocale)
+#define ucase_getTrie U_ICU_ENTRY_POINT_RENAME(ucase_getTrie)
+#define ucase_getType U_ICU_ENTRY_POINT_RENAME(ucase_getType)
+#define ucase_getTypeOrIgnorable U_ICU_ENTRY_POINT_RENAME(ucase_getTypeOrIgnorable)
+#define ucase_hasBinaryProperty U_ICU_ENTRY_POINT_RENAME(ucase_hasBinaryProperty)
+#define ucase_isCaseSensitive U_ICU_ENTRY_POINT_RENAME(ucase_isCaseSensitive)
+#define ucase_isSoftDotted U_ICU_ENTRY_POINT_RENAME(ucase_isSoftDotted)
+#define ucase_toFullFolding U_ICU_ENTRY_POINT_RENAME(ucase_toFullFolding)
+#define ucase_toFullLower U_ICU_ENTRY_POINT_RENAME(ucase_toFullLower)
+#define ucase_toFullTitle U_ICU_ENTRY_POINT_RENAME(ucase_toFullTitle)
+#define ucase_toFullUpper U_ICU_ENTRY_POINT_RENAME(ucase_toFullUpper)
+#define ucase_tolower U_ICU_ENTRY_POINT_RENAME(ucase_tolower)
+#define ucase_totitle U_ICU_ENTRY_POINT_RENAME(ucase_totitle)
+#define ucase_toupper U_ICU_ENTRY_POINT_RENAME(ucase_toupper)
+#define ucasemap_close U_ICU_ENTRY_POINT_RENAME(ucasemap_close)
+#define ucasemap_getBreakIterator U_ICU_ENTRY_POINT_RENAME(ucasemap_getBreakIterator)
+#define ucasemap_getLocale U_ICU_ENTRY_POINT_RENAME(ucasemap_getLocale)
+#define ucasemap_getOptions U_ICU_ENTRY_POINT_RENAME(ucasemap_getOptions)
+#define ucasemap_internalUTF8ToTitle U_ICU_ENTRY_POINT_RENAME(ucasemap_internalUTF8ToTitle)
+#define ucasemap_open U_ICU_ENTRY_POINT_RENAME(ucasemap_open)
+#define ucasemap_setBreakIterator U_ICU_ENTRY_POINT_RENAME(ucasemap_setBreakIterator)
+#define ucasemap_setLocale U_ICU_ENTRY_POINT_RENAME(ucasemap_setLocale)
+#define ucasemap_setOptions U_ICU_ENTRY_POINT_RENAME(ucasemap_setOptions)
+#define ucasemap_toTitle U_ICU_ENTRY_POINT_RENAME(ucasemap_toTitle)
+#define ucasemap_utf8FoldCase U_ICU_ENTRY_POINT_RENAME(ucasemap_utf8FoldCase)
+#define ucasemap_utf8ToLower U_ICU_ENTRY_POINT_RENAME(ucasemap_utf8ToLower)
+#define ucasemap_utf8ToTitle U_ICU_ENTRY_POINT_RENAME(ucasemap_utf8ToTitle)
+#define ucasemap_utf8ToUpper U_ICU_ENTRY_POINT_RENAME(ucasemap_utf8ToUpper)
+#define ucfpos_close U_ICU_ENTRY_POINT_RENAME(ucfpos_close)
+#define ucfpos_constrainCategory U_ICU_ENTRY_POINT_RENAME(ucfpos_constrainCategory)
+#define ucfpos_constrainField U_ICU_ENTRY_POINT_RENAME(ucfpos_constrainField)
+#define ucfpos_getCategory U_ICU_ENTRY_POINT_RENAME(ucfpos_getCategory)
+#define ucfpos_getField U_ICU_ENTRY_POINT_RENAME(ucfpos_getField)
+#define ucfpos_getIndexes U_ICU_ENTRY_POINT_RENAME(ucfpos_getIndexes)
+#define ucfpos_getInt64IterationContext U_ICU_ENTRY_POINT_RENAME(ucfpos_getInt64IterationContext)
+#define ucfpos_matchesField U_ICU_ENTRY_POINT_RENAME(ucfpos_matchesField)
+#define ucfpos_open U_ICU_ENTRY_POINT_RENAME(ucfpos_open)
+#define ucfpos_reset U_ICU_ENTRY_POINT_RENAME(ucfpos_reset)
+#define ucfpos_setInt64IterationContext U_ICU_ENTRY_POINT_RENAME(ucfpos_setInt64IterationContext)
+#define ucfpos_setState U_ICU_ENTRY_POINT_RENAME(ucfpos_setState)
+#define uchar_addPropertyStarts U_ICU_ENTRY_POINT_RENAME(uchar_addPropertyStarts)
+#define uchar_swapNames U_ICU_ENTRY_POINT_RENAME(uchar_swapNames)
+#define ucln_cleanupOne U_ICU_ENTRY_POINT_RENAME(ucln_cleanupOne)
+#define ucln_common_registerCleanup U_ICU_ENTRY_POINT_RENAME(ucln_common_registerCleanup)
+#define ucln_i18n_registerCleanup U_ICU_ENTRY_POINT_RENAME(ucln_i18n_registerCleanup)
+#define ucln_io_registerCleanup U_ICU_ENTRY_POINT_RENAME(ucln_io_registerCleanup)
+#define ucln_lib_cleanup U_ICU_ENTRY_POINT_RENAME(ucln_lib_cleanup)
+#define ucln_registerCleanup U_ICU_ENTRY_POINT_RENAME(ucln_registerCleanup)
+#define ucnv_MBCSFromUChar32 U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSFromUChar32)
+#define ucnv_MBCSFromUnicodeWithOffsets U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSFromUnicodeWithOffsets)
+#define ucnv_MBCSGetFilteredUnicodeSetForUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSGetFilteredUnicodeSetForUnicode)
+#define ucnv_MBCSGetType U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSGetType)
+#define ucnv_MBCSGetUnicodeSetForUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSGetUnicodeSetForUnicode)
+#define ucnv_MBCSIsLeadByte U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSIsLeadByte)
+#define ucnv_MBCSSimpleGetNextUChar U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSSimpleGetNextUChar)
+#define ucnv_MBCSToUnicodeWithOffsets U_ICU_ENTRY_POINT_RENAME(ucnv_MBCSToUnicodeWithOffsets)
+#define ucnv_bld_countAvailableConverters U_ICU_ENTRY_POINT_RENAME(ucnv_bld_countAvailableConverters)
+#define ucnv_bld_getAvailableConverter U_ICU_ENTRY_POINT_RENAME(ucnv_bld_getAvailableConverter)
+#define ucnv_canCreateConverter U_ICU_ENTRY_POINT_RENAME(ucnv_canCreateConverter)
+#define ucnv_cbFromUWriteBytes U_ICU_ENTRY_POINT_RENAME(ucnv_cbFromUWriteBytes)
+#define ucnv_cbFromUWriteSub U_ICU_ENTRY_POINT_RENAME(ucnv_cbFromUWriteSub)
+#define ucnv_cbFromUWriteUChars U_ICU_ENTRY_POINT_RENAME(ucnv_cbFromUWriteUChars)
+#define ucnv_cbToUWriteSub U_ICU_ENTRY_POINT_RENAME(ucnv_cbToUWriteSub)
+#define ucnv_cbToUWriteUChars U_ICU_ENTRY_POINT_RENAME(ucnv_cbToUWriteUChars)
+#define ucnv_close U_ICU_ENTRY_POINT_RENAME(ucnv_close)
+#define ucnv_compareNames U_ICU_ENTRY_POINT_RENAME(ucnv_compareNames)
+#define ucnv_convert U_ICU_ENTRY_POINT_RENAME(ucnv_convert)
+#define ucnv_convertEx U_ICU_ENTRY_POINT_RENAME(ucnv_convertEx)
+#define ucnv_countAliases U_ICU_ENTRY_POINT_RENAME(ucnv_countAliases)
+#define ucnv_countAvailable U_ICU_ENTRY_POINT_RENAME(ucnv_countAvailable)
+#define ucnv_countStandards U_ICU_ENTRY_POINT_RENAME(ucnv_countStandards)
+#define ucnv_createAlgorithmicConverter U_ICU_ENTRY_POINT_RENAME(ucnv_createAlgorithmicConverter)
+#define ucnv_createConverter U_ICU_ENTRY_POINT_RENAME(ucnv_createConverter)
+#define ucnv_createConverterFromPackage U_ICU_ENTRY_POINT_RENAME(ucnv_createConverterFromPackage)
+#define ucnv_createConverterFromSharedData U_ICU_ENTRY_POINT_RENAME(ucnv_createConverterFromSharedData)
+#define ucnv_detectUnicodeSignature U_ICU_ENTRY_POINT_RENAME(ucnv_detectUnicodeSignature)
+#define ucnv_enableCleanup U_ICU_ENTRY_POINT_RENAME(ucnv_enableCleanup)
+#define ucnv_extContinueMatchFromU U_ICU_ENTRY_POINT_RENAME(ucnv_extContinueMatchFromU)
+#define ucnv_extContinueMatchToU U_ICU_ENTRY_POINT_RENAME(ucnv_extContinueMatchToU)
+#define ucnv_extGetUnicodeSet U_ICU_ENTRY_POINT_RENAME(ucnv_extGetUnicodeSet)
+#define ucnv_extInitialMatchFromU U_ICU_ENTRY_POINT_RENAME(ucnv_extInitialMatchFromU)
+#define ucnv_extInitialMatchToU U_ICU_ENTRY_POINT_RENAME(ucnv_extInitialMatchToU)
+#define ucnv_extSimpleMatchFromU U_ICU_ENTRY_POINT_RENAME(ucnv_extSimpleMatchFromU)
+#define ucnv_extSimpleMatchToU U_ICU_ENTRY_POINT_RENAME(ucnv_extSimpleMatchToU)
+#define ucnv_fixFileSeparator U_ICU_ENTRY_POINT_RENAME(ucnv_fixFileSeparator)
+#define ucnv_flushCache U_ICU_ENTRY_POINT_RENAME(ucnv_flushCache)
+#define ucnv_fromAlgorithmic U_ICU_ENTRY_POINT_RENAME(ucnv_fromAlgorithmic)
+#define ucnv_fromUChars U_ICU_ENTRY_POINT_RENAME(ucnv_fromUChars)
+#define ucnv_fromUCountPending U_ICU_ENTRY_POINT_RENAME(ucnv_fromUCountPending)
+#define ucnv_fromUWriteBytes U_ICU_ENTRY_POINT_RENAME(ucnv_fromUWriteBytes)
+#define ucnv_fromUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_fromUnicode)
+#define ucnv_fromUnicode_UTF8 U_ICU_ENTRY_POINT_RENAME(ucnv_fromUnicode_UTF8)
+#define ucnv_fromUnicode_UTF8_OFFSETS_LOGIC U_ICU_ENTRY_POINT_RENAME(ucnv_fromUnicode_UTF8_OFFSETS_LOGIC)
+#define ucnv_getAlias U_ICU_ENTRY_POINT_RENAME(ucnv_getAlias)
+#define ucnv_getAliases U_ICU_ENTRY_POINT_RENAME(ucnv_getAliases)
+#define ucnv_getAvailableName U_ICU_ENTRY_POINT_RENAME(ucnv_getAvailableName)
+#define ucnv_getCCSID U_ICU_ENTRY_POINT_RENAME(ucnv_getCCSID)
+#define ucnv_getCanonicalName U_ICU_ENTRY_POINT_RENAME(ucnv_getCanonicalName)
+#define ucnv_getCompleteUnicodeSet U_ICU_ENTRY_POINT_RENAME(ucnv_getCompleteUnicodeSet)
+#define ucnv_getDefaultName U_ICU_ENTRY_POINT_RENAME(ucnv_getDefaultName)
+#define ucnv_getDisplayName U_ICU_ENTRY_POINT_RENAME(ucnv_getDisplayName)
+#define ucnv_getFromUCallBack U_ICU_ENTRY_POINT_RENAME(ucnv_getFromUCallBack)
+#define ucnv_getInvalidChars U_ICU_ENTRY_POINT_RENAME(ucnv_getInvalidChars)
+#define ucnv_getInvalidUChars U_ICU_ENTRY_POINT_RENAME(ucnv_getInvalidUChars)
+#define ucnv_getMaxCharSize U_ICU_ENTRY_POINT_RENAME(ucnv_getMaxCharSize)
+#define ucnv_getMinCharSize U_ICU_ENTRY_POINT_RENAME(ucnv_getMinCharSize)
+#define ucnv_getName U_ICU_ENTRY_POINT_RENAME(ucnv_getName)
+#define ucnv_getNextUChar U_ICU_ENTRY_POINT_RENAME(ucnv_getNextUChar)
+#define ucnv_getNonSurrogateUnicodeSet U_ICU_ENTRY_POINT_RENAME(ucnv_getNonSurrogateUnicodeSet)
+#define ucnv_getPlatform U_ICU_ENTRY_POINT_RENAME(ucnv_getPlatform)
+#define ucnv_getStandard U_ICU_ENTRY_POINT_RENAME(ucnv_getStandard)
+#define ucnv_getStandardName U_ICU_ENTRY_POINT_RENAME(ucnv_getStandardName)
+#define ucnv_getStarters U_ICU_ENTRY_POINT_RENAME(ucnv_getStarters)
+#define ucnv_getSubstChars U_ICU_ENTRY_POINT_RENAME(ucnv_getSubstChars)
+#define ucnv_getToUCallBack U_ICU_ENTRY_POINT_RENAME(ucnv_getToUCallBack)
+#define ucnv_getType U_ICU_ENTRY_POINT_RENAME(ucnv_getType)
+#define ucnv_getUnicodeSet U_ICU_ENTRY_POINT_RENAME(ucnv_getUnicodeSet)
+#define ucnv_incrementRefCount U_ICU_ENTRY_POINT_RENAME(ucnv_incrementRefCount)
+#define ucnv_io_countKnownConverters U_ICU_ENTRY_POINT_RENAME(ucnv_io_countKnownConverters)
+#define ucnv_io_getConverterName U_ICU_ENTRY_POINT_RENAME(ucnv_io_getConverterName)
+#define ucnv_io_stripASCIIForCompare U_ICU_ENTRY_POINT_RENAME(ucnv_io_stripASCIIForCompare)
+#define ucnv_io_stripEBCDICForCompare U_ICU_ENTRY_POINT_RENAME(ucnv_io_stripEBCDICForCompare)
+#define ucnv_isAmbiguous U_ICU_ENTRY_POINT_RENAME(ucnv_isAmbiguous)
+#define ucnv_isFixedWidth U_ICU_ENTRY_POINT_RENAME(ucnv_isFixedWidth)
+#define ucnv_load U_ICU_ENTRY_POINT_RENAME(ucnv_load)
+#define ucnv_loadSharedData U_ICU_ENTRY_POINT_RENAME(ucnv_loadSharedData)
+#define ucnv_open U_ICU_ENTRY_POINT_RENAME(ucnv_open)
+#define ucnv_openAllNames U_ICU_ENTRY_POINT_RENAME(ucnv_openAllNames)
+#define ucnv_openCCSID U_ICU_ENTRY_POINT_RENAME(ucnv_openCCSID)
+#define ucnv_openPackage U_ICU_ENTRY_POINT_RENAME(ucnv_openPackage)
+#define ucnv_openStandardNames U_ICU_ENTRY_POINT_RENAME(ucnv_openStandardNames)
+#define ucnv_openU U_ICU_ENTRY_POINT_RENAME(ucnv_openU)
+#define ucnv_reset U_ICU_ENTRY_POINT_RENAME(ucnv_reset)
+#define ucnv_resetFromUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_resetFromUnicode)
+#define ucnv_resetToUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_resetToUnicode)
+#define ucnv_safeClone U_ICU_ENTRY_POINT_RENAME(ucnv_safeClone)
+#define ucnv_setDefaultName U_ICU_ENTRY_POINT_RENAME(ucnv_setDefaultName)
+#define ucnv_setFallback U_ICU_ENTRY_POINT_RENAME(ucnv_setFallback)
+#define ucnv_setFromUCallBack U_ICU_ENTRY_POINT_RENAME(ucnv_setFromUCallBack)
+#define ucnv_setSubstChars U_ICU_ENTRY_POINT_RENAME(ucnv_setSubstChars)
+#define ucnv_setSubstString U_ICU_ENTRY_POINT_RENAME(ucnv_setSubstString)
+#define ucnv_setToUCallBack U_ICU_ENTRY_POINT_RENAME(ucnv_setToUCallBack)
+#define ucnv_swap U_ICU_ENTRY_POINT_RENAME(ucnv_swap)
+#define ucnv_swapAliases U_ICU_ENTRY_POINT_RENAME(ucnv_swapAliases)
+#define ucnv_toAlgorithmic U_ICU_ENTRY_POINT_RENAME(ucnv_toAlgorithmic)
+#define ucnv_toUChars U_ICU_ENTRY_POINT_RENAME(ucnv_toUChars)
+#define ucnv_toUCountPending U_ICU_ENTRY_POINT_RENAME(ucnv_toUCountPending)
+#define ucnv_toUWriteCodePoint U_ICU_ENTRY_POINT_RENAME(ucnv_toUWriteCodePoint)
+#define ucnv_toUWriteUChars U_ICU_ENTRY_POINT_RENAME(ucnv_toUWriteUChars)
+#define ucnv_toUnicode U_ICU_ENTRY_POINT_RENAME(ucnv_toUnicode)
+#define ucnv_unload U_ICU_ENTRY_POINT_RENAME(ucnv_unload)
+#define ucnv_unloadSharedDataIfReady U_ICU_ENTRY_POINT_RENAME(ucnv_unloadSharedDataIfReady)
+#define ucnv_usesFallback U_ICU_ENTRY_POINT_RENAME(ucnv_usesFallback)
+#define ucnvsel_close U_ICU_ENTRY_POINT_RENAME(ucnvsel_close)
+#define ucnvsel_open U_ICU_ENTRY_POINT_RENAME(ucnvsel_open)
+#define ucnvsel_openFromSerialized U_ICU_ENTRY_POINT_RENAME(ucnvsel_openFromSerialized)
+#define ucnvsel_selectForString U_ICU_ENTRY_POINT_RENAME(ucnvsel_selectForString)
+#define ucnvsel_selectForUTF8 U_ICU_ENTRY_POINT_RENAME(ucnvsel_selectForUTF8)
+#define ucnvsel_serialize U_ICU_ENTRY_POINT_RENAME(ucnvsel_serialize)
+#define ucol_cloneBinary U_ICU_ENTRY_POINT_RENAME(ucol_cloneBinary)
+#define ucol_close U_ICU_ENTRY_POINT_RENAME(ucol_close)
+#define ucol_closeElements U_ICU_ENTRY_POINT_RENAME(ucol_closeElements)
+#define ucol_countAvailable U_ICU_ENTRY_POINT_RENAME(ucol_countAvailable)
+#define ucol_equal U_ICU_ENTRY_POINT_RENAME(ucol_equal)
+#define ucol_equals U_ICU_ENTRY_POINT_RENAME(ucol_equals)
+#define ucol_getAttribute U_ICU_ENTRY_POINT_RENAME(ucol_getAttribute)
+#define ucol_getAvailable U_ICU_ENTRY_POINT_RENAME(ucol_getAvailable)
+#define ucol_getBound U_ICU_ENTRY_POINT_RENAME(ucol_getBound)
+#define ucol_getContractions U_ICU_ENTRY_POINT_RENAME(ucol_getContractions)
+#define ucol_getContractionsAndExpansions U_ICU_ENTRY_POINT_RENAME(ucol_getContractionsAndExpansions)
+#define ucol_getDisplayName U_ICU_ENTRY_POINT_RENAME(ucol_getDisplayName)
+#define ucol_getEquivalentReorderCodes U_ICU_ENTRY_POINT_RENAME(ucol_getEquivalentReorderCodes)
+#define ucol_getFunctionalEquivalent U_ICU_ENTRY_POINT_RENAME(ucol_getFunctionalEquivalent)
+#define ucol_getKeywordValues U_ICU_ENTRY_POINT_RENAME(ucol_getKeywordValues)
+#define ucol_getKeywordValuesForLocale U_ICU_ENTRY_POINT_RENAME(ucol_getKeywordValuesForLocale)
+#define ucol_getKeywords U_ICU_ENTRY_POINT_RENAME(ucol_getKeywords)
+#define ucol_getLocale U_ICU_ENTRY_POINT_RENAME(ucol_getLocale)
+#define ucol_getLocaleByType U_ICU_ENTRY_POINT_RENAME(ucol_getLocaleByType)
+#define ucol_getMaxExpansion U_ICU_ENTRY_POINT_RENAME(ucol_getMaxExpansion)
+#define ucol_getMaxVariable U_ICU_ENTRY_POINT_RENAME(ucol_getMaxVariable)
+#define ucol_getOffset U_ICU_ENTRY_POINT_RENAME(ucol_getOffset)
+#define ucol_getReorderCodes U_ICU_ENTRY_POINT_RENAME(ucol_getReorderCodes)
+#define ucol_getRules U_ICU_ENTRY_POINT_RENAME(ucol_getRules)
+#define ucol_getRulesEx U_ICU_ENTRY_POINT_RENAME(ucol_getRulesEx)
+#define ucol_getShortDefinitionString U_ICU_ENTRY_POINT_RENAME(ucol_getShortDefinitionString)
+#define ucol_getSortKey U_ICU_ENTRY_POINT_RENAME(ucol_getSortKey)
+#define ucol_getStrength U_ICU_ENTRY_POINT_RENAME(ucol_getStrength)
+#define ucol_getTailoredSet U_ICU_ENTRY_POINT_RENAME(ucol_getTailoredSet)
+#define ucol_getUCAVersion U_ICU_ENTRY_POINT_RENAME(ucol_getUCAVersion)
+#define ucol_getUnsafeSet U_ICU_ENTRY_POINT_RENAME(ucol_getUnsafeSet)
+#define ucol_getVariableTop U_ICU_ENTRY_POINT_RENAME(ucol_getVariableTop)
+#define ucol_getVersion U_ICU_ENTRY_POINT_RENAME(ucol_getVersion)
+#define ucol_greater U_ICU_ENTRY_POINT_RENAME(ucol_greater)
+#define ucol_greaterOrEqual U_ICU_ENTRY_POINT_RENAME(ucol_greaterOrEqual)
+#define ucol_keyHashCode U_ICU_ENTRY_POINT_RENAME(ucol_keyHashCode)
+#define ucol_looksLikeCollationBinary U_ICU_ENTRY_POINT_RENAME(ucol_looksLikeCollationBinary)
+#define ucol_mergeSortkeys U_ICU_ENTRY_POINT_RENAME(ucol_mergeSortkeys)
+#define ucol_next U_ICU_ENTRY_POINT_RENAME(ucol_next)
+#define ucol_nextSortKeyPart U_ICU_ENTRY_POINT_RENAME(ucol_nextSortKeyPart)
+#define ucol_normalizeShortDefinitionString U_ICU_ENTRY_POINT_RENAME(ucol_normalizeShortDefinitionString)
+#define ucol_open U_ICU_ENTRY_POINT_RENAME(ucol_open)
+#define ucol_openAvailableLocales U_ICU_ENTRY_POINT_RENAME(ucol_openAvailableLocales)
+#define ucol_openBinary U_ICU_ENTRY_POINT_RENAME(ucol_openBinary)
+#define ucol_openElements U_ICU_ENTRY_POINT_RENAME(ucol_openElements)
+#define ucol_openFromShortString U_ICU_ENTRY_POINT_RENAME(ucol_openFromShortString)
+#define ucol_openRules U_ICU_ENTRY_POINT_RENAME(ucol_openRules)
+#define ucol_prepareShortStringOpen U_ICU_ENTRY_POINT_RENAME(ucol_prepareShortStringOpen)
+#define ucol_previous U_ICU_ENTRY_POINT_RENAME(ucol_previous)
+#define ucol_primaryOrder U_ICU_ENTRY_POINT_RENAME(ucol_primaryOrder)
+#define ucol_reset U_ICU_ENTRY_POINT_RENAME(ucol_reset)
+#define ucol_restoreVariableTop U_ICU_ENTRY_POINT_RENAME(ucol_restoreVariableTop)
+#define ucol_safeClone U_ICU_ENTRY_POINT_RENAME(ucol_safeClone)
+#define ucol_secondaryOrder U_ICU_ENTRY_POINT_RENAME(ucol_secondaryOrder)
+#define ucol_setAttribute U_ICU_ENTRY_POINT_RENAME(ucol_setAttribute)
+#define ucol_setMaxVariable U_ICU_ENTRY_POINT_RENAME(ucol_setMaxVariable)
+#define ucol_setOffset U_ICU_ENTRY_POINT_RENAME(ucol_setOffset)
+#define ucol_setReorderCodes U_ICU_ENTRY_POINT_RENAME(ucol_setReorderCodes)
+#define ucol_setStrength U_ICU_ENTRY_POINT_RENAME(ucol_setStrength)
+#define ucol_setText U_ICU_ENTRY_POINT_RENAME(ucol_setText)
+#define ucol_setVariableTop U_ICU_ENTRY_POINT_RENAME(ucol_setVariableTop)
+#define ucol_strcoll U_ICU_ENTRY_POINT_RENAME(ucol_strcoll)
+#define ucol_strcollIter U_ICU_ENTRY_POINT_RENAME(ucol_strcollIter)
+#define ucol_strcollUTF8 U_ICU_ENTRY_POINT_RENAME(ucol_strcollUTF8)
+#define ucol_swap U_ICU_ENTRY_POINT_RENAME(ucol_swap)
+#define ucol_swapInverseUCA U_ICU_ENTRY_POINT_RENAME(ucol_swapInverseUCA)
+#define ucol_tertiaryOrder U_ICU_ENTRY_POINT_RENAME(ucol_tertiaryOrder)
+#define ucpmap_get U_ICU_ENTRY_POINT_RENAME(ucpmap_get)
+#define ucpmap_getRange U_ICU_ENTRY_POINT_RENAME(ucpmap_getRange)
+#define ucptrie_close U_ICU_ENTRY_POINT_RENAME(ucptrie_close)
+#define ucptrie_get U_ICU_ENTRY_POINT_RENAME(ucptrie_get)
+#define ucptrie_getRange U_ICU_ENTRY_POINT_RENAME(ucptrie_getRange)
+#define ucptrie_getType U_ICU_ENTRY_POINT_RENAME(ucptrie_getType)
+#define ucptrie_getValueWidth U_ICU_ENTRY_POINT_RENAME(ucptrie_getValueWidth)
+#define ucptrie_internalGetRange U_ICU_ENTRY_POINT_RENAME(ucptrie_internalGetRange)
+#define ucptrie_internalSmallIndex U_ICU_ENTRY_POINT_RENAME(ucptrie_internalSmallIndex)
+#define ucptrie_internalSmallU8Index U_ICU_ENTRY_POINT_RENAME(ucptrie_internalSmallU8Index)
+#define ucptrie_internalU8PrevIndex U_ICU_ENTRY_POINT_RENAME(ucptrie_internalU8PrevIndex)
+#define ucptrie_openFromBinary U_ICU_ENTRY_POINT_RENAME(ucptrie_openFromBinary)
+#define ucptrie_swap U_ICU_ENTRY_POINT_RENAME(ucptrie_swap)
+#define ucptrie_toBinary U_ICU_ENTRY_POINT_RENAME(ucptrie_toBinary)
+#define ucsdet_close U_ICU_ENTRY_POINT_RENAME(ucsdet_close)
+#define ucsdet_detect U_ICU_ENTRY_POINT_RENAME(ucsdet_detect)
+#define ucsdet_detectAll U_ICU_ENTRY_POINT_RENAME(ucsdet_detectAll)
+#define ucsdet_enableInputFilter U_ICU_ENTRY_POINT_RENAME(ucsdet_enableInputFilter)
+#define ucsdet_getAllDetectableCharsets U_ICU_ENTRY_POINT_RENAME(ucsdet_getAllDetectableCharsets)
+#define ucsdet_getConfidence U_ICU_ENTRY_POINT_RENAME(ucsdet_getConfidence)
+#define ucsdet_getDetectableCharsets U_ICU_ENTRY_POINT_RENAME(ucsdet_getDetectableCharsets)
+#define ucsdet_getLanguage U_ICU_ENTRY_POINT_RENAME(ucsdet_getLanguage)
+#define ucsdet_getName U_ICU_ENTRY_POINT_RENAME(ucsdet_getName)
+#define ucsdet_getUChars U_ICU_ENTRY_POINT_RENAME(ucsdet_getUChars)
+#define ucsdet_isInputFilterEnabled U_ICU_ENTRY_POINT_RENAME(ucsdet_isInputFilterEnabled)
+#define ucsdet_open U_ICU_ENTRY_POINT_RENAME(ucsdet_open)
+#define ucsdet_setDeclaredEncoding U_ICU_ENTRY_POINT_RENAME(ucsdet_setDeclaredEncoding)
+#define ucsdet_setDetectableCharset U_ICU_ENTRY_POINT_RENAME(ucsdet_setDetectableCharset)
+#define ucsdet_setText U_ICU_ENTRY_POINT_RENAME(ucsdet_setText)
+#define ucurr_countCurrencies U_ICU_ENTRY_POINT_RENAME(ucurr_countCurrencies)
+#define ucurr_forLocale U_ICU_ENTRY_POINT_RENAME(ucurr_forLocale)
+#define ucurr_forLocaleAndDate U_ICU_ENTRY_POINT_RENAME(ucurr_forLocaleAndDate)
+#define ucurr_getDefaultFractionDigits U_ICU_ENTRY_POINT_RENAME(ucurr_getDefaultFractionDigits)
+#define ucurr_getDefaultFractionDigitsForUsage U_ICU_ENTRY_POINT_RENAME(ucurr_getDefaultFractionDigitsForUsage)
+#define ucurr_getKeywordValuesForLocale U_ICU_ENTRY_POINT_RENAME(ucurr_getKeywordValuesForLocale)
+#define ucurr_getName U_ICU_ENTRY_POINT_RENAME(ucurr_getName)
+#define ucurr_getNumericCode U_ICU_ENTRY_POINT_RENAME(ucurr_getNumericCode)
+#define ucurr_getPluralName U_ICU_ENTRY_POINT_RENAME(ucurr_getPluralName)
+#define ucurr_getRoundingIncrement U_ICU_ENTRY_POINT_RENAME(ucurr_getRoundingIncrement)
+#define ucurr_getRoundingIncrementForUsage U_ICU_ENTRY_POINT_RENAME(ucurr_getRoundingIncrementForUsage)
+#define ucurr_isAvailable U_ICU_ENTRY_POINT_RENAME(ucurr_isAvailable)
+#define ucurr_openISOCurrencies U_ICU_ENTRY_POINT_RENAME(ucurr_openISOCurrencies)
+#define ucurr_register U_ICU_ENTRY_POINT_RENAME(ucurr_register)
+#define ucurr_unregister U_ICU_ENTRY_POINT_RENAME(ucurr_unregister)
+#define udat_adoptNumberFormat U_ICU_ENTRY_POINT_RENAME(udat_adoptNumberFormat)
+#define udat_adoptNumberFormatForFields U_ICU_ENTRY_POINT_RENAME(udat_adoptNumberFormatForFields)
+#define udat_applyPattern U_ICU_ENTRY_POINT_RENAME(udat_applyPattern)
+#define udat_applyPatternRelative U_ICU_ENTRY_POINT_RENAME(udat_applyPatternRelative)
+#define udat_clone U_ICU_ENTRY_POINT_RENAME(udat_clone)
+#define udat_close U_ICU_ENTRY_POINT_RENAME(udat_close)
+#define udat_countAvailable U_ICU_ENTRY_POINT_RENAME(udat_countAvailable)
+#define udat_countSymbols U_ICU_ENTRY_POINT_RENAME(udat_countSymbols)
+#define udat_format U_ICU_ENTRY_POINT_RENAME(udat_format)
+#define udat_formatCalendar U_ICU_ENTRY_POINT_RENAME(udat_formatCalendar)
+#define udat_formatCalendarForFields U_ICU_ENTRY_POINT_RENAME(udat_formatCalendarForFields)
+#define udat_formatForFields U_ICU_ENTRY_POINT_RENAME(udat_formatForFields)
+#define udat_get2DigitYearStart U_ICU_ENTRY_POINT_RENAME(udat_get2DigitYearStart)
+#define udat_getAvailable U_ICU_ENTRY_POINT_RENAME(udat_getAvailable)
+#define udat_getBooleanAttribute U_ICU_ENTRY_POINT_RENAME(udat_getBooleanAttribute)
+#define udat_getCalendar U_ICU_ENTRY_POINT_RENAME(udat_getCalendar)
+#define udat_getContext U_ICU_ENTRY_POINT_RENAME(udat_getContext)
+#define udat_getLocaleByType U_ICU_ENTRY_POINT_RENAME(udat_getLocaleByType)
+#define udat_getNumberFormat U_ICU_ENTRY_POINT_RENAME(udat_getNumberFormat)
+#define udat_getNumberFormatForField U_ICU_ENTRY_POINT_RENAME(udat_getNumberFormatForField)
+#define udat_getSymbols U_ICU_ENTRY_POINT_RENAME(udat_getSymbols)
+#define udat_isLenient U_ICU_ENTRY_POINT_RENAME(udat_isLenient)
+#define udat_open U_ICU_ENTRY_POINT_RENAME(udat_open)
+#define udat_parse U_ICU_ENTRY_POINT_RENAME(udat_parse)
+#define udat_parseCalendar U_ICU_ENTRY_POINT_RENAME(udat_parseCalendar)
+#define udat_registerOpener U_ICU_ENTRY_POINT_RENAME(udat_registerOpener)
+#define udat_set2DigitYearStart U_ICU_ENTRY_POINT_RENAME(udat_set2DigitYearStart)
+#define udat_setBooleanAttribute U_ICU_ENTRY_POINT_RENAME(udat_setBooleanAttribute)
+#define udat_setCalendar U_ICU_ENTRY_POINT_RENAME(udat_setCalendar)
+#define udat_setContext U_ICU_ENTRY_POINT_RENAME(udat_setContext)
+#define udat_setLenient U_ICU_ENTRY_POINT_RENAME(udat_setLenient)
+#define udat_setNumberFormat U_ICU_ENTRY_POINT_RENAME(udat_setNumberFormat)
+#define udat_setSymbols U_ICU_ENTRY_POINT_RENAME(udat_setSymbols)
+#define udat_toCalendarDateField U_ICU_ENTRY_POINT_RENAME(udat_toCalendarDateField)
+#define udat_toPattern U_ICU_ENTRY_POINT_RENAME(udat_toPattern)
+#define udat_toPatternRelativeDate U_ICU_ENTRY_POINT_RENAME(udat_toPatternRelativeDate)
+#define udat_toPatternRelativeTime U_ICU_ENTRY_POINT_RENAME(udat_toPatternRelativeTime)
+#define udat_unregisterOpener U_ICU_ENTRY_POINT_RENAME(udat_unregisterOpener)
+#define udata_checkCommonData U_ICU_ENTRY_POINT_RENAME(udata_checkCommonData)
+#define udata_close U_ICU_ENTRY_POINT_RENAME(udata_close)
+#define udata_closeSwapper U_ICU_ENTRY_POINT_RENAME(udata_closeSwapper)
+#define udata_getHeaderSize U_ICU_ENTRY_POINT_RENAME(udata_getHeaderSize)
+#define udata_getInfo U_ICU_ENTRY_POINT_RENAME(udata_getInfo)
+#define udata_getInfoSize U_ICU_ENTRY_POINT_RENAME(udata_getInfoSize)
+#define udata_getLength U_ICU_ENTRY_POINT_RENAME(udata_getLength)
+#define udata_getMemory U_ICU_ENTRY_POINT_RENAME(udata_getMemory)
+#define udata_getRawMemory U_ICU_ENTRY_POINT_RENAME(udata_getRawMemory)
+#define udata_open U_ICU_ENTRY_POINT_RENAME(udata_open)
+#define udata_openChoice U_ICU_ENTRY_POINT_RENAME(udata_openChoice)
+#define udata_openSwapper U_ICU_ENTRY_POINT_RENAME(udata_openSwapper)
+#define udata_openSwapperForInputData U_ICU_ENTRY_POINT_RENAME(udata_openSwapperForInputData)
+#define udata_printError U_ICU_ENTRY_POINT_RENAME(udata_printError)
+#define udata_readInt16 U_ICU_ENTRY_POINT_RENAME(udata_readInt16)
+#define udata_readInt32 U_ICU_ENTRY_POINT_RENAME(udata_readInt32)
+#define udata_setAppData U_ICU_ENTRY_POINT_RENAME(udata_setAppData)
+#define udata_setCommonData U_ICU_ENTRY_POINT_RENAME(udata_setCommonData)
+#define udata_setFileAccess U_ICU_ENTRY_POINT_RENAME(udata_setFileAccess)
+#define udata_swapDataHeader U_ICU_ENTRY_POINT_RENAME(udata_swapDataHeader)
+#define udata_swapInvStringBlock U_ICU_ENTRY_POINT_RENAME(udata_swapInvStringBlock)
+#define udatpg_addPattern U_ICU_ENTRY_POINT_RENAME(udatpg_addPattern)
+#define udatpg_clone U_ICU_ENTRY_POINT_RENAME(udatpg_clone)
+#define udatpg_close U_ICU_ENTRY_POINT_RENAME(udatpg_close)
+#define udatpg_getAppendItemFormat U_ICU_ENTRY_POINT_RENAME(udatpg_getAppendItemFormat)
+#define udatpg_getAppendItemName U_ICU_ENTRY_POINT_RENAME(udatpg_getAppendItemName)
+#define udatpg_getBaseSkeleton U_ICU_ENTRY_POINT_RENAME(udatpg_getBaseSkeleton)
+#define udatpg_getBestPattern U_ICU_ENTRY_POINT_RENAME(udatpg_getBestPattern)
+#define udatpg_getBestPatternWithOptions U_ICU_ENTRY_POINT_RENAME(udatpg_getBestPatternWithOptions)
+#define udatpg_getDateTimeFormat U_ICU_ENTRY_POINT_RENAME(udatpg_getDateTimeFormat)
+#define udatpg_getDecimal U_ICU_ENTRY_POINT_RENAME(udatpg_getDecimal)
+#define udatpg_getDefaultHourCycle U_ICU_ENTRY_POINT_RENAME(udatpg_getDefaultHourCycle)
+#define udatpg_getFieldDisplayName U_ICU_ENTRY_POINT_RENAME(udatpg_getFieldDisplayName)
+#define udatpg_getPatternForSkeleton U_ICU_ENTRY_POINT_RENAME(udatpg_getPatternForSkeleton)
+#define udatpg_getSkeleton U_ICU_ENTRY_POINT_RENAME(udatpg_getSkeleton)
+#define udatpg_open U_ICU_ENTRY_POINT_RENAME(udatpg_open)
+#define udatpg_openBaseSkeletons U_ICU_ENTRY_POINT_RENAME(udatpg_openBaseSkeletons)
+#define udatpg_openEmpty U_ICU_ENTRY_POINT_RENAME(udatpg_openEmpty)
+#define udatpg_openSkeletons U_ICU_ENTRY_POINT_RENAME(udatpg_openSkeletons)
+#define udatpg_replaceFieldTypes U_ICU_ENTRY_POINT_RENAME(udatpg_replaceFieldTypes)
+#define udatpg_replaceFieldTypesWithOptions U_ICU_ENTRY_POINT_RENAME(udatpg_replaceFieldTypesWithOptions)
+#define udatpg_setAppendItemFormat U_ICU_ENTRY_POINT_RENAME(udatpg_setAppendItemFormat)
+#define udatpg_setAppendItemName U_ICU_ENTRY_POINT_RENAME(udatpg_setAppendItemName)
+#define udatpg_setDateTimeFormat U_ICU_ENTRY_POINT_RENAME(udatpg_setDateTimeFormat)
+#define udatpg_setDecimal U_ICU_ENTRY_POINT_RENAME(udatpg_setDecimal)
+#define udict_swap U_ICU_ENTRY_POINT_RENAME(udict_swap)
+#define udtitvfmt_close U_ICU_ENTRY_POINT_RENAME(udtitvfmt_close)
+#define udtitvfmt_closeResult U_ICU_ENTRY_POINT_RENAME(udtitvfmt_closeResult)
+#define udtitvfmt_format U_ICU_ENTRY_POINT_RENAME(udtitvfmt_format)
+#define udtitvfmt_formatCalendarToResult U_ICU_ENTRY_POINT_RENAME(udtitvfmt_formatCalendarToResult)
+#define udtitvfmt_formatToResult U_ICU_ENTRY_POINT_RENAME(udtitvfmt_formatToResult)
+#define udtitvfmt_getContext U_ICU_ENTRY_POINT_RENAME(udtitvfmt_getContext)
+#define udtitvfmt_open U_ICU_ENTRY_POINT_RENAME(udtitvfmt_open)
+#define udtitvfmt_openResult U_ICU_ENTRY_POINT_RENAME(udtitvfmt_openResult)
+#define udtitvfmt_resultAsValue U_ICU_ENTRY_POINT_RENAME(udtitvfmt_resultAsValue)
+#define udtitvfmt_setContext U_ICU_ENTRY_POINT_RENAME(udtitvfmt_setContext)
+#define uenum_close U_ICU_ENTRY_POINT_RENAME(uenum_close)
+#define uenum_count U_ICU_ENTRY_POINT_RENAME(uenum_count)
+#define uenum_next U_ICU_ENTRY_POINT_RENAME(uenum_next)
+#define uenum_nextDefault U_ICU_ENTRY_POINT_RENAME(uenum_nextDefault)
+#define uenum_openCharStringsEnumeration U_ICU_ENTRY_POINT_RENAME(uenum_openCharStringsEnumeration)
+#define uenum_openFromStringEnumeration U_ICU_ENTRY_POINT_RENAME(uenum_openFromStringEnumeration)
+#define uenum_openUCharStringsEnumeration U_ICU_ENTRY_POINT_RENAME(uenum_openUCharStringsEnumeration)
+#define uenum_reset U_ICU_ENTRY_POINT_RENAME(uenum_reset)
+#define uenum_unext U_ICU_ENTRY_POINT_RENAME(uenum_unext)
+#define uenum_unextDefault U_ICU_ENTRY_POINT_RENAME(uenum_unextDefault)
+#define ufieldpositer_close U_ICU_ENTRY_POINT_RENAME(ufieldpositer_close)
+#define ufieldpositer_next U_ICU_ENTRY_POINT_RENAME(ufieldpositer_next)
+#define ufieldpositer_open U_ICU_ENTRY_POINT_RENAME(ufieldpositer_open)
+#define ufile_getch U_ICU_ENTRY_POINT_RENAME(ufile_getch)
+#define ufile_getch32 U_ICU_ENTRY_POINT_RENAME(ufile_getch32)
+#define ufmt_close U_ICU_ENTRY_POINT_RENAME(ufmt_close)
+#define ufmt_getArrayItemByIndex U_ICU_ENTRY_POINT_RENAME(ufmt_getArrayItemByIndex)
+#define ufmt_getArrayLength U_ICU_ENTRY_POINT_RENAME(ufmt_getArrayLength)
+#define ufmt_getDate U_ICU_ENTRY_POINT_RENAME(ufmt_getDate)
+#define ufmt_getDecNumChars U_ICU_ENTRY_POINT_RENAME(ufmt_getDecNumChars)
+#define ufmt_getDouble U_ICU_ENTRY_POINT_RENAME(ufmt_getDouble)
+#define ufmt_getInt64 U_ICU_ENTRY_POINT_RENAME(ufmt_getInt64)
+#define ufmt_getLong U_ICU_ENTRY_POINT_RENAME(ufmt_getLong)
+#define ufmt_getObject U_ICU_ENTRY_POINT_RENAME(ufmt_getObject)
+#define ufmt_getType U_ICU_ENTRY_POINT_RENAME(ufmt_getType)
+#define ufmt_getUChars U_ICU_ENTRY_POINT_RENAME(ufmt_getUChars)
+#define ufmt_isNumeric U_ICU_ENTRY_POINT_RENAME(ufmt_isNumeric)
+#define ufmt_open U_ICU_ENTRY_POINT_RENAME(ufmt_open)
+#define ufmtval_getString U_ICU_ENTRY_POINT_RENAME(ufmtval_getString)
+#define ufmtval_nextPosition U_ICU_ENTRY_POINT_RENAME(ufmtval_nextPosition)
+#define ugender_getInstance U_ICU_ENTRY_POINT_RENAME(ugender_getInstance)
+#define ugender_getListGender U_ICU_ENTRY_POINT_RENAME(ugender_getListGender)
+#define uhash_close U_ICU_ENTRY_POINT_RENAME(uhash_close)
+#define uhash_compareCaselessUnicodeString U_ICU_ENTRY_POINT_RENAME(uhash_compareCaselessUnicodeString)
+#define uhash_compareChars U_ICU_ENTRY_POINT_RENAME(uhash_compareChars)
+#define uhash_compareIChars U_ICU_ENTRY_POINT_RENAME(uhash_compareIChars)
+#define uhash_compareLong U_ICU_ENTRY_POINT_RENAME(uhash_compareLong)
+#define uhash_compareScriptSet U_ICU_ENTRY_POINT_RENAME(uhash_compareScriptSet)
+#define uhash_compareUChars U_ICU_ENTRY_POINT_RENAME(uhash_compareUChars)
+#define uhash_compareUnicodeString U_ICU_ENTRY_POINT_RENAME(uhash_compareUnicodeString)
+#define uhash_count U_ICU_ENTRY_POINT_RENAME(uhash_count)
+#define uhash_deleteHashtable U_ICU_ENTRY_POINT_RENAME(uhash_deleteHashtable)
+#define uhash_deleteScriptSet U_ICU_ENTRY_POINT_RENAME(uhash_deleteScriptSet)
+#define uhash_equals U_ICU_ENTRY_POINT_RENAME(uhash_equals)
+#define uhash_equalsScriptSet U_ICU_ENTRY_POINT_RENAME(uhash_equalsScriptSet)
+#define uhash_find U_ICU_ENTRY_POINT_RENAME(uhash_find)
+#define uhash_get U_ICU_ENTRY_POINT_RENAME(uhash_get)
+#define uhash_geti U_ICU_ENTRY_POINT_RENAME(uhash_geti)
+#define uhash_hashCaselessUnicodeString U_ICU_ENTRY_POINT_RENAME(uhash_hashCaselessUnicodeString)
+#define uhash_hashChars U_ICU_ENTRY_POINT_RENAME(uhash_hashChars)
+#define uhash_hashIChars U_ICU_ENTRY_POINT_RENAME(uhash_hashIChars)
+#define uhash_hashLong U_ICU_ENTRY_POINT_RENAME(uhash_hashLong)
+#define uhash_hashScriptSet U_ICU_ENTRY_POINT_RENAME(uhash_hashScriptSet)
+#define uhash_hashUChars U_ICU_ENTRY_POINT_RENAME(uhash_hashUChars)
+#define uhash_hashUnicodeString U_ICU_ENTRY_POINT_RENAME(uhash_hashUnicodeString)
+#define uhash_iget U_ICU_ENTRY_POINT_RENAME(uhash_iget)
+#define uhash_igeti U_ICU_ENTRY_POINT_RENAME(uhash_igeti)
+#define uhash_init U_ICU_ENTRY_POINT_RENAME(uhash_init)
+#define uhash_initSize U_ICU_ENTRY_POINT_RENAME(uhash_initSize)
+#define uhash_iput U_ICU_ENTRY_POINT_RENAME(uhash_iput)
+#define uhash_iputi U_ICU_ENTRY_POINT_RENAME(uhash_iputi)
+#define uhash_iremove U_ICU_ENTRY_POINT_RENAME(uhash_iremove)
+#define uhash_iremovei U_ICU_ENTRY_POINT_RENAME(uhash_iremovei)
+#define uhash_nextElement U_ICU_ENTRY_POINT_RENAME(uhash_nextElement)
+#define uhash_open U_ICU_ENTRY_POINT_RENAME(uhash_open)
+#define uhash_openSize U_ICU_ENTRY_POINT_RENAME(uhash_openSize)
+#define uhash_put U_ICU_ENTRY_POINT_RENAME(uhash_put)
+#define uhash_puti U_ICU_ENTRY_POINT_RENAME(uhash_puti)
+#define uhash_remove U_ICU_ENTRY_POINT_RENAME(uhash_remove)
+#define uhash_removeAll U_ICU_ENTRY_POINT_RENAME(uhash_removeAll)
+#define uhash_removeElement U_ICU_ENTRY_POINT_RENAME(uhash_removeElement)
+#define uhash_removei U_ICU_ENTRY_POINT_RENAME(uhash_removei)
+#define uhash_setKeyComparator U_ICU_ENTRY_POINT_RENAME(uhash_setKeyComparator)
+#define uhash_setKeyDeleter U_ICU_ENTRY_POINT_RENAME(uhash_setKeyDeleter)
+#define uhash_setKeyHasher U_ICU_ENTRY_POINT_RENAME(uhash_setKeyHasher)
+#define uhash_setResizePolicy U_ICU_ENTRY_POINT_RENAME(uhash_setResizePolicy)
+#define uhash_setValueComparator U_ICU_ENTRY_POINT_RENAME(uhash_setValueComparator)
+#define uhash_setValueDeleter U_ICU_ENTRY_POINT_RENAME(uhash_setValueDeleter)
+#define uidna_IDNToASCII U_ICU_ENTRY_POINT_RENAME(uidna_IDNToASCII)
+#define uidna_IDNToUnicode U_ICU_ENTRY_POINT_RENAME(uidna_IDNToUnicode)
+#define uidna_close U_ICU_ENTRY_POINT_RENAME(uidna_close)
+#define uidna_compare U_ICU_ENTRY_POINT_RENAME(uidna_compare)
+#define uidna_labelToASCII U_ICU_ENTRY_POINT_RENAME(uidna_labelToASCII)
+#define uidna_labelToASCII_UTF8 U_ICU_ENTRY_POINT_RENAME(uidna_labelToASCII_UTF8)
+#define uidna_labelToUnicode U_ICU_ENTRY_POINT_RENAME(uidna_labelToUnicode)
+#define uidna_labelToUnicodeUTF8 U_ICU_ENTRY_POINT_RENAME(uidna_labelToUnicodeUTF8)
+#define uidna_nameToASCII U_ICU_ENTRY_POINT_RENAME(uidna_nameToASCII)
+#define uidna_nameToASCII_UTF8 U_ICU_ENTRY_POINT_RENAME(uidna_nameToASCII_UTF8)
+#define uidna_nameToUnicode U_ICU_ENTRY_POINT_RENAME(uidna_nameToUnicode)
+#define uidna_nameToUnicodeUTF8 U_ICU_ENTRY_POINT_RENAME(uidna_nameToUnicodeUTF8)
+#define uidna_openUTS46 U_ICU_ENTRY_POINT_RENAME(uidna_openUTS46)
+#define uidna_toASCII U_ICU_ENTRY_POINT_RENAME(uidna_toASCII)
+#define uidna_toUnicode U_ICU_ENTRY_POINT_RENAME(uidna_toUnicode)
+#define uiter_current32 U_ICU_ENTRY_POINT_RENAME(uiter_current32)
+#define uiter_getState U_ICU_ENTRY_POINT_RENAME(uiter_getState)
+#define uiter_next32 U_ICU_ENTRY_POINT_RENAME(uiter_next32)
+#define uiter_previous32 U_ICU_ENTRY_POINT_RENAME(uiter_previous32)
+#define uiter_setCharacterIterator U_ICU_ENTRY_POINT_RENAME(uiter_setCharacterIterator)
+#define uiter_setReplaceable U_ICU_ENTRY_POINT_RENAME(uiter_setReplaceable)
+#define uiter_setState U_ICU_ENTRY_POINT_RENAME(uiter_setState)
+#define uiter_setString U_ICU_ENTRY_POINT_RENAME(uiter_setString)
+#define uiter_setUTF16BE U_ICU_ENTRY_POINT_RENAME(uiter_setUTF16BE)
+#define uiter_setUTF8 U_ICU_ENTRY_POINT_RENAME(uiter_setUTF8)
+#define uldn_close U_ICU_ENTRY_POINT_RENAME(uldn_close)
+#define uldn_getContext U_ICU_ENTRY_POINT_RENAME(uldn_getContext)
+#define uldn_getDialectHandling U_ICU_ENTRY_POINT_RENAME(uldn_getDialectHandling)
+#define uldn_getLocale U_ICU_ENTRY_POINT_RENAME(uldn_getLocale)
+#define uldn_keyDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_keyDisplayName)
+#define uldn_keyValueDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_keyValueDisplayName)
+#define uldn_languageDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_languageDisplayName)
+#define uldn_localeDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_localeDisplayName)
+#define uldn_open U_ICU_ENTRY_POINT_RENAME(uldn_open)
+#define uldn_openForContext U_ICU_ENTRY_POINT_RENAME(uldn_openForContext)
+#define uldn_regionDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_regionDisplayName)
+#define uldn_scriptCodeDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_scriptCodeDisplayName)
+#define uldn_scriptDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_scriptDisplayName)
+#define uldn_variantDisplayName U_ICU_ENTRY_POINT_RENAME(uldn_variantDisplayName)
+#define ulist_addItemBeginList U_ICU_ENTRY_POINT_RENAME(ulist_addItemBeginList)
+#define ulist_addItemEndList U_ICU_ENTRY_POINT_RENAME(ulist_addItemEndList)
+#define ulist_close_keyword_values_iterator U_ICU_ENTRY_POINT_RENAME(ulist_close_keyword_values_iterator)
+#define ulist_containsString U_ICU_ENTRY_POINT_RENAME(ulist_containsString)
+#define ulist_count_keyword_values U_ICU_ENTRY_POINT_RENAME(ulist_count_keyword_values)
+#define ulist_createEmptyList U_ICU_ENTRY_POINT_RENAME(ulist_createEmptyList)
+#define ulist_deleteList U_ICU_ENTRY_POINT_RENAME(ulist_deleteList)
+#define ulist_getListFromEnum U_ICU_ENTRY_POINT_RENAME(ulist_getListFromEnum)
+#define ulist_getListSize U_ICU_ENTRY_POINT_RENAME(ulist_getListSize)
+#define ulist_getNext U_ICU_ENTRY_POINT_RENAME(ulist_getNext)
+#define ulist_next_keyword_value U_ICU_ENTRY_POINT_RENAME(ulist_next_keyword_value)
+#define ulist_removeString U_ICU_ENTRY_POINT_RENAME(ulist_removeString)
+#define ulist_resetList U_ICU_ENTRY_POINT_RENAME(ulist_resetList)
+#define ulist_reset_keyword_values_iterator U_ICU_ENTRY_POINT_RENAME(ulist_reset_keyword_values_iterator)
+#define ulistfmt_close U_ICU_ENTRY_POINT_RENAME(ulistfmt_close)
+#define ulistfmt_closeResult U_ICU_ENTRY_POINT_RENAME(ulistfmt_closeResult)
+#define ulistfmt_format U_ICU_ENTRY_POINT_RENAME(ulistfmt_format)
+#define ulistfmt_formatStringsToResult U_ICU_ENTRY_POINT_RENAME(ulistfmt_formatStringsToResult)
+#define ulistfmt_open U_ICU_ENTRY_POINT_RENAME(ulistfmt_open)
+#define ulistfmt_openForType U_ICU_ENTRY_POINT_RENAME(ulistfmt_openForType)
+#define ulistfmt_openResult U_ICU_ENTRY_POINT_RENAME(ulistfmt_openResult)
+#define ulistfmt_resultAsValue U_ICU_ENTRY_POINT_RENAME(ulistfmt_resultAsValue)
+#define uloc_acceptLanguage U_ICU_ENTRY_POINT_RENAME(uloc_acceptLanguage)
+#define uloc_acceptLanguageFromHTTP U_ICU_ENTRY_POINT_RENAME(uloc_acceptLanguageFromHTTP)
+#define uloc_addLikelySubtags U_ICU_ENTRY_POINT_RENAME(uloc_addLikelySubtags)
+#define uloc_canonicalize U_ICU_ENTRY_POINT_RENAME(uloc_canonicalize)
+#define uloc_countAvailable U_ICU_ENTRY_POINT_RENAME(uloc_countAvailable)
+#define uloc_forLanguageTag U_ICU_ENTRY_POINT_RENAME(uloc_forLanguageTag)
+#define uloc_getAvailable U_ICU_ENTRY_POINT_RENAME(uloc_getAvailable)
+#define uloc_getBaseName U_ICU_ENTRY_POINT_RENAME(uloc_getBaseName)
+#define uloc_getCharacterOrientation U_ICU_ENTRY_POINT_RENAME(uloc_getCharacterOrientation)
+#define uloc_getCountry U_ICU_ENTRY_POINT_RENAME(uloc_getCountry)
+#define uloc_getCurrentCountryID U_ICU_ENTRY_POINT_RENAME(uloc_getCurrentCountryID)
+#define uloc_getCurrentLanguageID U_ICU_ENTRY_POINT_RENAME(uloc_getCurrentLanguageID)
+#define uloc_getDefault U_ICU_ENTRY_POINT_RENAME(uloc_getDefault)
+#define uloc_getDisplayCountry U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayCountry)
+#define uloc_getDisplayKeyword U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayKeyword)
+#define uloc_getDisplayKeywordValue U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayKeywordValue)
+#define uloc_getDisplayLanguage U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayLanguage)
+#define uloc_getDisplayName U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayName)
+#define uloc_getDisplayScript U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayScript)
+#define uloc_getDisplayVariant U_ICU_ENTRY_POINT_RENAME(uloc_getDisplayVariant)
+#define uloc_getISO3Country U_ICU_ENTRY_POINT_RENAME(uloc_getISO3Country)
+#define uloc_getISO3Language U_ICU_ENTRY_POINT_RENAME(uloc_getISO3Language)
+#define uloc_getISOCountries U_ICU_ENTRY_POINT_RENAME(uloc_getISOCountries)
+#define uloc_getISOLanguages U_ICU_ENTRY_POINT_RENAME(uloc_getISOLanguages)
+#define uloc_getKeywordValue U_ICU_ENTRY_POINT_RENAME(uloc_getKeywordValue)
+#define uloc_getLCID U_ICU_ENTRY_POINT_RENAME(uloc_getLCID)
+#define uloc_getLanguage U_ICU_ENTRY_POINT_RENAME(uloc_getLanguage)
+#define uloc_getLineOrientation U_ICU_ENTRY_POINT_RENAME(uloc_getLineOrientation)
+#define uloc_getLocaleForLCID U_ICU_ENTRY_POINT_RENAME(uloc_getLocaleForLCID)
+#define uloc_getName U_ICU_ENTRY_POINT_RENAME(uloc_getName)
+#define uloc_getParent U_ICU_ENTRY_POINT_RENAME(uloc_getParent)
+#define uloc_getScript U_ICU_ENTRY_POINT_RENAME(uloc_getScript)
+#define uloc_getTableStringWithFallback U_ICU_ENTRY_POINT_RENAME(uloc_getTableStringWithFallback)
+#define uloc_getVariant U_ICU_ENTRY_POINT_RENAME(uloc_getVariant)
+#define uloc_isRightToLeft U_ICU_ENTRY_POINT_RENAME(uloc_isRightToLeft)
+#define uloc_minimizeSubtags U_ICU_ENTRY_POINT_RENAME(uloc_minimizeSubtags)
+#define uloc_openAvailableByType U_ICU_ENTRY_POINT_RENAME(uloc_openAvailableByType)
+#define uloc_openKeywordList U_ICU_ENTRY_POINT_RENAME(uloc_openKeywordList)
+#define uloc_openKeywords U_ICU_ENTRY_POINT_RENAME(uloc_openKeywords)
+#define uloc_setDefault U_ICU_ENTRY_POINT_RENAME(uloc_setDefault)
+#define uloc_setKeywordValue U_ICU_ENTRY_POINT_RENAME(uloc_setKeywordValue)
+#define uloc_toLanguageTag U_ICU_ENTRY_POINT_RENAME(uloc_toLanguageTag)
+#define uloc_toLegacyKey U_ICU_ENTRY_POINT_RENAME(uloc_toLegacyKey)
+#define uloc_toLegacyType U_ICU_ENTRY_POINT_RENAME(uloc_toLegacyType)
+#define uloc_toUnicodeLocaleKey U_ICU_ENTRY_POINT_RENAME(uloc_toUnicodeLocaleKey)
+#define uloc_toUnicodeLocaleType U_ICU_ENTRY_POINT_RENAME(uloc_toUnicodeLocaleType)
+#define ulocdata_close U_ICU_ENTRY_POINT_RENAME(ulocdata_close)
+#define ulocdata_getCLDRVersion U_ICU_ENTRY_POINT_RENAME(ulocdata_getCLDRVersion)
+#define ulocdata_getDelimiter U_ICU_ENTRY_POINT_RENAME(ulocdata_getDelimiter)
+#define ulocdata_getExemplarSet U_ICU_ENTRY_POINT_RENAME(ulocdata_getExemplarSet)
+#define ulocdata_getLocaleDisplayPattern U_ICU_ENTRY_POINT_RENAME(ulocdata_getLocaleDisplayPattern)
+#define ulocdata_getLocaleSeparator U_ICU_ENTRY_POINT_RENAME(ulocdata_getLocaleSeparator)
+#define ulocdata_getMeasurementSystem U_ICU_ENTRY_POINT_RENAME(ulocdata_getMeasurementSystem)
+#define ulocdata_getNoSubstitute U_ICU_ENTRY_POINT_RENAME(ulocdata_getNoSubstitute)
+#define ulocdata_getPaperSize U_ICU_ENTRY_POINT_RENAME(ulocdata_getPaperSize)
+#define ulocdata_open U_ICU_ENTRY_POINT_RENAME(ulocdata_open)
+#define ulocdata_setNoSubstitute U_ICU_ENTRY_POINT_RENAME(ulocdata_setNoSubstitute)
+#define ulocimp_addLikelySubtags U_ICU_ENTRY_POINT_RENAME(ulocimp_addLikelySubtags)
+#define ulocimp_canonicalize U_ICU_ENTRY_POINT_RENAME(ulocimp_canonicalize)
+#define ulocimp_forLanguageTag U_ICU_ENTRY_POINT_RENAME(ulocimp_forLanguageTag)
+#define ulocimp_getBaseName U_ICU_ENTRY_POINT_RENAME(ulocimp_getBaseName)
+#define ulocimp_getCountry U_ICU_ENTRY_POINT_RENAME(ulocimp_getCountry)
+#define ulocimp_getKeywordValue U_ICU_ENTRY_POINT_RENAME(ulocimp_getKeywordValue)
+#define ulocimp_getKeywords U_ICU_ENTRY_POINT_RENAME(ulocimp_getKeywords)
+#define ulocimp_getKnownCanonicalizedLocaleForTest U_ICU_ENTRY_POINT_RENAME(ulocimp_getKnownCanonicalizedLocaleForTest)
+#define ulocimp_getLanguage U_ICU_ENTRY_POINT_RENAME(ulocimp_getLanguage)
+#define ulocimp_getName U_ICU_ENTRY_POINT_RENAME(ulocimp_getName)
+#define ulocimp_getRegionForSupplementalData U_ICU_ENTRY_POINT_RENAME(ulocimp_getRegionForSupplementalData)
+#define ulocimp_getScript U_ICU_ENTRY_POINT_RENAME(ulocimp_getScript)
+#define ulocimp_isCanonicalizedLocaleForTest U_ICU_ENTRY_POINT_RENAME(ulocimp_isCanonicalizedLocaleForTest)
+#define ulocimp_minimizeSubtags U_ICU_ENTRY_POINT_RENAME(ulocimp_minimizeSubtags)
+#define ulocimp_toBcpKey U_ICU_ENTRY_POINT_RENAME(ulocimp_toBcpKey)
+#define ulocimp_toBcpType U_ICU_ENTRY_POINT_RENAME(ulocimp_toBcpType)
+#define ulocimp_toLanguageTag U_ICU_ENTRY_POINT_RENAME(ulocimp_toLanguageTag)
+#define ulocimp_toLegacyKey U_ICU_ENTRY_POINT_RENAME(ulocimp_toLegacyKey)
+#define ulocimp_toLegacyType U_ICU_ENTRY_POINT_RENAME(ulocimp_toLegacyType)
+#define ultag_isExtensionSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isExtensionSubtags)
+#define ultag_isLanguageSubtag U_ICU_ENTRY_POINT_RENAME(ultag_isLanguageSubtag)
+#define ultag_isPrivateuseValueSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isPrivateuseValueSubtags)
+#define ultag_isRegionSubtag U_ICU_ENTRY_POINT_RENAME(ultag_isRegionSubtag)
+#define ultag_isScriptSubtag U_ICU_ENTRY_POINT_RENAME(ultag_isScriptSubtag)
+#define ultag_isTransformedExtensionSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isTransformedExtensionSubtags)
+#define ultag_isUnicodeExtensionSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isUnicodeExtensionSubtags)
+#define ultag_isUnicodeLocaleAttribute U_ICU_ENTRY_POINT_RENAME(ultag_isUnicodeLocaleAttribute)
+#define ultag_isUnicodeLocaleAttributes U_ICU_ENTRY_POINT_RENAME(ultag_isUnicodeLocaleAttributes)
+#define ultag_isUnicodeLocaleKey U_ICU_ENTRY_POINT_RENAME(ultag_isUnicodeLocaleKey)
+#define ultag_isUnicodeLocaleType U_ICU_ENTRY_POINT_RENAME(ultag_isUnicodeLocaleType)
+#define ultag_isVariantSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isVariantSubtags)
+#define umsg_applyPattern U_ICU_ENTRY_POINT_RENAME(umsg_applyPattern)
+#define umsg_autoQuoteApostrophe U_ICU_ENTRY_POINT_RENAME(umsg_autoQuoteApostrophe)
+#define umsg_clone U_ICU_ENTRY_POINT_RENAME(umsg_clone)
+#define umsg_close U_ICU_ENTRY_POINT_RENAME(umsg_close)
+#define umsg_format U_ICU_ENTRY_POINT_RENAME(umsg_format)
+#define umsg_getLocale U_ICU_ENTRY_POINT_RENAME(umsg_getLocale)
+#define umsg_open U_ICU_ENTRY_POINT_RENAME(umsg_open)
+#define umsg_parse U_ICU_ENTRY_POINT_RENAME(umsg_parse)
+#define umsg_setLocale U_ICU_ENTRY_POINT_RENAME(umsg_setLocale)
+#define umsg_toPattern U_ICU_ENTRY_POINT_RENAME(umsg_toPattern)
+#define umsg_vformat U_ICU_ENTRY_POINT_RENAME(umsg_vformat)
+#define umsg_vparse U_ICU_ENTRY_POINT_RENAME(umsg_vparse)
+#define umtx_lock U_ICU_ENTRY_POINT_RENAME(umtx_lock)
+#define umtx_unlock U_ICU_ENTRY_POINT_RENAME(umtx_unlock)
+#define umutablecptrie_buildImmutable U_ICU_ENTRY_POINT_RENAME(umutablecptrie_buildImmutable)
+#define umutablecptrie_clone U_ICU_ENTRY_POINT_RENAME(umutablecptrie_clone)
+#define umutablecptrie_close U_ICU_ENTRY_POINT_RENAME(umutablecptrie_close)
+#define umutablecptrie_fromUCPMap U_ICU_ENTRY_POINT_RENAME(umutablecptrie_fromUCPMap)
+#define umutablecptrie_fromUCPTrie U_ICU_ENTRY_POINT_RENAME(umutablecptrie_fromUCPTrie)
+#define umutablecptrie_get U_ICU_ENTRY_POINT_RENAME(umutablecptrie_get)
+#define umutablecptrie_getRange U_ICU_ENTRY_POINT_RENAME(umutablecptrie_getRange)
+#define umutablecptrie_open U_ICU_ENTRY_POINT_RENAME(umutablecptrie_open)
+#define umutablecptrie_set U_ICU_ENTRY_POINT_RENAME(umutablecptrie_set)
+#define umutablecptrie_setRange U_ICU_ENTRY_POINT_RENAME(umutablecptrie_setRange)
+#define uniset_getUnicode32Instance U_ICU_ENTRY_POINT_RENAME(uniset_getUnicode32Instance)
+#define unorm2_append U_ICU_ENTRY_POINT_RENAME(unorm2_append)
+#define unorm2_close U_ICU_ENTRY_POINT_RENAME(unorm2_close)
+#define unorm2_composePair U_ICU_ENTRY_POINT_RENAME(unorm2_composePair)
+#define unorm2_getCombiningClass U_ICU_ENTRY_POINT_RENAME(unorm2_getCombiningClass)
+#define unorm2_getDecomposition U_ICU_ENTRY_POINT_RENAME(unorm2_getDecomposition)
+#define unorm2_getInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getInstance)
+#define unorm2_getNFCInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getNFCInstance)
+#define unorm2_getNFDInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getNFDInstance)
+#define unorm2_getNFKCCasefoldInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getNFKCCasefoldInstance)
+#define unorm2_getNFKCInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getNFKCInstance)
+#define unorm2_getNFKDInstance U_ICU_ENTRY_POINT_RENAME(unorm2_getNFKDInstance)
+#define unorm2_getRawDecomposition U_ICU_ENTRY_POINT_RENAME(unorm2_getRawDecomposition)
+#define unorm2_hasBoundaryAfter U_ICU_ENTRY_POINT_RENAME(unorm2_hasBoundaryAfter)
+#define unorm2_hasBoundaryBefore U_ICU_ENTRY_POINT_RENAME(unorm2_hasBoundaryBefore)
+#define unorm2_isInert U_ICU_ENTRY_POINT_RENAME(unorm2_isInert)
+#define unorm2_isNormalized U_ICU_ENTRY_POINT_RENAME(unorm2_isNormalized)
+#define unorm2_normalize U_ICU_ENTRY_POINT_RENAME(unorm2_normalize)
+#define unorm2_normalizeSecondAndAppend U_ICU_ENTRY_POINT_RENAME(unorm2_normalizeSecondAndAppend)
+#define unorm2_openFiltered U_ICU_ENTRY_POINT_RENAME(unorm2_openFiltered)
+#define unorm2_quickCheck U_ICU_ENTRY_POINT_RENAME(unorm2_quickCheck)
+#define unorm2_spanQuickCheckYes U_ICU_ENTRY_POINT_RENAME(unorm2_spanQuickCheckYes)
+#define unorm2_swap U_ICU_ENTRY_POINT_RENAME(unorm2_swap)
+#define unorm_compare U_ICU_ENTRY_POINT_RENAME(unorm_compare)
+#define unorm_concatenate U_ICU_ENTRY_POINT_RENAME(unorm_concatenate)
+#define unorm_getFCD16 U_ICU_ENTRY_POINT_RENAME(unorm_getFCD16)
+#define unorm_getQuickCheck U_ICU_ENTRY_POINT_RENAME(unorm_getQuickCheck)
+#define unorm_isNormalized U_ICU_ENTRY_POINT_RENAME(unorm_isNormalized)
+#define unorm_isNormalizedWithOptions U_ICU_ENTRY_POINT_RENAME(unorm_isNormalizedWithOptions)
+#define unorm_next U_ICU_ENTRY_POINT_RENAME(unorm_next)
+#define unorm_normalize U_ICU_ENTRY_POINT_RENAME(unorm_normalize)
+#define unorm_previous U_ICU_ENTRY_POINT_RENAME(unorm_previous)
+#define unorm_quickCheck U_ICU_ENTRY_POINT_RENAME(unorm_quickCheck)
+#define unorm_quickCheckWithOptions U_ICU_ENTRY_POINT_RENAME(unorm_quickCheckWithOptions)
+#define unum_applyPattern U_ICU_ENTRY_POINT_RENAME(unum_applyPattern)
+#define unum_clone U_ICU_ENTRY_POINT_RENAME(unum_clone)
+#define unum_close U_ICU_ENTRY_POINT_RENAME(unum_close)
+#define unum_countAvailable U_ICU_ENTRY_POINT_RENAME(unum_countAvailable)
+#define unum_format U_ICU_ENTRY_POINT_RENAME(unum_format)
+#define unum_formatDecimal U_ICU_ENTRY_POINT_RENAME(unum_formatDecimal)
+#define unum_formatDouble U_ICU_ENTRY_POINT_RENAME(unum_formatDouble)
+#define unum_formatDoubleCurrency U_ICU_ENTRY_POINT_RENAME(unum_formatDoubleCurrency)
+#define unum_formatDoubleForFields U_ICU_ENTRY_POINT_RENAME(unum_formatDoubleForFields)
+#define unum_formatInt64 U_ICU_ENTRY_POINT_RENAME(unum_formatInt64)
+#define unum_formatUFormattable U_ICU_ENTRY_POINT_RENAME(unum_formatUFormattable)
+#define unum_getAttribute U_ICU_ENTRY_POINT_RENAME(unum_getAttribute)
+#define unum_getAvailable U_ICU_ENTRY_POINT_RENAME(unum_getAvailable)
+#define unum_getContext U_ICU_ENTRY_POINT_RENAME(unum_getContext)
+#define unum_getDoubleAttribute U_ICU_ENTRY_POINT_RENAME(unum_getDoubleAttribute)
+#define unum_getLocaleByType U_ICU_ENTRY_POINT_RENAME(unum_getLocaleByType)
+#define unum_getSymbol U_ICU_ENTRY_POINT_RENAME(unum_getSymbol)
+#define unum_getTextAttribute U_ICU_ENTRY_POINT_RENAME(unum_getTextAttribute)
+#define unum_open U_ICU_ENTRY_POINT_RENAME(unum_open)
+#define unum_parse U_ICU_ENTRY_POINT_RENAME(unum_parse)
+#define unum_parseDecimal U_ICU_ENTRY_POINT_RENAME(unum_parseDecimal)
+#define unum_parseDouble U_ICU_ENTRY_POINT_RENAME(unum_parseDouble)
+#define unum_parseDoubleCurrency U_ICU_ENTRY_POINT_RENAME(unum_parseDoubleCurrency)
+#define unum_parseInt64 U_ICU_ENTRY_POINT_RENAME(unum_parseInt64)
+#define unum_parseToUFormattable U_ICU_ENTRY_POINT_RENAME(unum_parseToUFormattable)
+#define unum_setAttribute U_ICU_ENTRY_POINT_RENAME(unum_setAttribute)
+#define unum_setContext U_ICU_ENTRY_POINT_RENAME(unum_setContext)
+#define unum_setDoubleAttribute U_ICU_ENTRY_POINT_RENAME(unum_setDoubleAttribute)
+#define unum_setSymbol U_ICU_ENTRY_POINT_RENAME(unum_setSymbol)
+#define unum_setTextAttribute U_ICU_ENTRY_POINT_RENAME(unum_setTextAttribute)
+#define unum_toPattern U_ICU_ENTRY_POINT_RENAME(unum_toPattern)
+#define unumf_close U_ICU_ENTRY_POINT_RENAME(unumf_close)
+#define unumf_closeResult U_ICU_ENTRY_POINT_RENAME(unumf_closeResult)
+#define unumf_formatDecimal U_ICU_ENTRY_POINT_RENAME(unumf_formatDecimal)
+#define unumf_formatDouble U_ICU_ENTRY_POINT_RENAME(unumf_formatDouble)
+#define unumf_formatInt U_ICU_ENTRY_POINT_RENAME(unumf_formatInt)
+#define unumf_openForSkeletonAndLocale U_ICU_ENTRY_POINT_RENAME(unumf_openForSkeletonAndLocale)
+#define unumf_openForSkeletonAndLocaleWithError U_ICU_ENTRY_POINT_RENAME(unumf_openForSkeletonAndLocaleWithError)
+#define unumf_openResult U_ICU_ENTRY_POINT_RENAME(unumf_openResult)
+#define unumf_resultAsValue U_ICU_ENTRY_POINT_RENAME(unumf_resultAsValue)
+#define unumf_resultGetAllFieldPositions U_ICU_ENTRY_POINT_RENAME(unumf_resultGetAllFieldPositions)
+#define unumf_resultNextFieldPosition U_ICU_ENTRY_POINT_RENAME(unumf_resultNextFieldPosition)
+#define unumf_resultToDecimalNumber U_ICU_ENTRY_POINT_RENAME(unumf_resultToDecimalNumber)
+#define unumf_resultToString U_ICU_ENTRY_POINT_RENAME(unumf_resultToString)
+#define unumrf_close U_ICU_ENTRY_POINT_RENAME(unumrf_close)
+#define unumrf_closeResult U_ICU_ENTRY_POINT_RENAME(unumrf_closeResult)
+#define unumrf_formatDecimalRange U_ICU_ENTRY_POINT_RENAME(unumrf_formatDecimalRange)
+#define unumrf_formatDoubleRange U_ICU_ENTRY_POINT_RENAME(unumrf_formatDoubleRange)
+#define unumrf_openForSkeletonWithCollapseAndIdentityFallback U_ICU_ENTRY_POINT_RENAME(unumrf_openForSkeletonWithCollapseAndIdentityFallback)
+#define unumrf_openResult U_ICU_ENTRY_POINT_RENAME(unumrf_openResult)
+#define unumrf_resultAsValue U_ICU_ENTRY_POINT_RENAME(unumrf_resultAsValue)
+#define unumrf_resultGetFirstDecimalNumber U_ICU_ENTRY_POINT_RENAME(unumrf_resultGetFirstDecimalNumber)
+#define unumrf_resultGetIdentityResult U_ICU_ENTRY_POINT_RENAME(unumrf_resultGetIdentityResult)
+#define unumrf_resultGetSecondDecimalNumber U_ICU_ENTRY_POINT_RENAME(unumrf_resultGetSecondDecimalNumber)
+#define unumsys_close U_ICU_ENTRY_POINT_RENAME(unumsys_close)
+#define unumsys_getDescription U_ICU_ENTRY_POINT_RENAME(unumsys_getDescription)
+#define unumsys_getName U_ICU_ENTRY_POINT_RENAME(unumsys_getName)
+#define unumsys_getRadix U_ICU_ENTRY_POINT_RENAME(unumsys_getRadix)
+#define unumsys_isAlgorithmic U_ICU_ENTRY_POINT_RENAME(unumsys_isAlgorithmic)
+#define unumsys_open U_ICU_ENTRY_POINT_RENAME(unumsys_open)
+#define unumsys_openAvailableNames U_ICU_ENTRY_POINT_RENAME(unumsys_openAvailableNames)
+#define unumsys_openByName U_ICU_ENTRY_POINT_RENAME(unumsys_openByName)
+#define uplrules_close U_ICU_ENTRY_POINT_RENAME(uplrules_close)
+#define uplrules_getKeywords U_ICU_ENTRY_POINT_RENAME(uplrules_getKeywords)
+#define uplrules_open U_ICU_ENTRY_POINT_RENAME(uplrules_open)
+#define uplrules_openForType U_ICU_ENTRY_POINT_RENAME(uplrules_openForType)
+#define uplrules_select U_ICU_ENTRY_POINT_RENAME(uplrules_select)
+#define uplrules_selectForRange U_ICU_ENTRY_POINT_RENAME(uplrules_selectForRange)
+#define uplrules_selectFormatted U_ICU_ENTRY_POINT_RENAME(uplrules_selectFormatted)
+#define uplrules_selectWithFormat U_ICU_ENTRY_POINT_RENAME(uplrules_selectWithFormat)
+#define uplug_closeLibrary U_ICU_ENTRY_POINT_RENAME(uplug_closeLibrary)
+#define uplug_findLibrary U_ICU_ENTRY_POINT_RENAME(uplug_findLibrary)
+#define uplug_getConfiguration U_ICU_ENTRY_POINT_RENAME(uplug_getConfiguration)
+#define uplug_getContext U_ICU_ENTRY_POINT_RENAME(uplug_getContext)
+#define uplug_getCurrentLevel U_ICU_ENTRY_POINT_RENAME(uplug_getCurrentLevel)
+#define uplug_getLibrary U_ICU_ENTRY_POINT_RENAME(uplug_getLibrary)
+#define uplug_getLibraryName U_ICU_ENTRY_POINT_RENAME(uplug_getLibraryName)
+#define uplug_getPlugInternal U_ICU_ENTRY_POINT_RENAME(uplug_getPlugInternal)
+#define uplug_getPlugLevel U_ICU_ENTRY_POINT_RENAME(uplug_getPlugLevel)
+#define uplug_getPlugLoadStatus U_ICU_ENTRY_POINT_RENAME(uplug_getPlugLoadStatus)
+#define uplug_getPlugName U_ICU_ENTRY_POINT_RENAME(uplug_getPlugName)
+#define uplug_getPluginFile U_ICU_ENTRY_POINT_RENAME(uplug_getPluginFile)
+#define uplug_getSymbolName U_ICU_ENTRY_POINT_RENAME(uplug_getSymbolName)
+#define uplug_init U_ICU_ENTRY_POINT_RENAME(uplug_init)
+#define uplug_loadPlugFromEntrypoint U_ICU_ENTRY_POINT_RENAME(uplug_loadPlugFromEntrypoint)
+#define uplug_loadPlugFromLibrary U_ICU_ENTRY_POINT_RENAME(uplug_loadPlugFromLibrary)
+#define uplug_nextPlug U_ICU_ENTRY_POINT_RENAME(uplug_nextPlug)
+#define uplug_openLibrary U_ICU_ENTRY_POINT_RENAME(uplug_openLibrary)
+#define uplug_removePlug U_ICU_ENTRY_POINT_RENAME(uplug_removePlug)
+#define uplug_setContext U_ICU_ENTRY_POINT_RENAME(uplug_setContext)
+#define uplug_setPlugLevel U_ICU_ENTRY_POINT_RENAME(uplug_setPlugLevel)
+#define uplug_setPlugName U_ICU_ENTRY_POINT_RENAME(uplug_setPlugName)
+#define uplug_setPlugNoUnload U_ICU_ENTRY_POINT_RENAME(uplug_setPlugNoUnload)
+#define uprops_addPropertyStarts U_ICU_ENTRY_POINT_RENAME(uprops_addPropertyStarts)
+#define uprops_getSource U_ICU_ENTRY_POINT_RENAME(uprops_getSource)
+#define upropsvec_addPropertyStarts U_ICU_ENTRY_POINT_RENAME(upropsvec_addPropertyStarts)
+#define uprv_add32_overflow U_ICU_ENTRY_POINT_RENAME(uprv_add32_overflow)
+#define uprv_aestrncpy U_ICU_ENTRY_POINT_RENAME(uprv_aestrncpy)
+#define uprv_asciiFromEbcdic U_ICU_ENTRY_POINT_RENAME(uprv_asciiFromEbcdic)
+#define uprv_asciitolower U_ICU_ENTRY_POINT_RENAME(uprv_asciitolower)
+#define uprv_calloc U_ICU_ENTRY_POINT_RENAME(uprv_calloc)
+#define uprv_ceil U_ICU_ENTRY_POINT_RENAME(uprv_ceil)
+#define uprv_compareASCIIPropertyNames U_ICU_ENTRY_POINT_RENAME(uprv_compareASCIIPropertyNames)
+#define uprv_compareEBCDICPropertyNames U_ICU_ENTRY_POINT_RENAME(uprv_compareEBCDICPropertyNames)
+#define uprv_compareInvAscii U_ICU_ENTRY_POINT_RENAME(uprv_compareInvAscii)
+#define uprv_compareInvEbcdic U_ICU_ENTRY_POINT_RENAME(uprv_compareInvEbcdic)
+#define uprv_compareInvEbcdicAsAscii U_ICU_ENTRY_POINT_RENAME(uprv_compareInvEbcdicAsAscii)
+#define uprv_convertToLCID U_ICU_ENTRY_POINT_RENAME(uprv_convertToLCID)
+#define uprv_convertToLCIDPlatform U_ICU_ENTRY_POINT_RENAME(uprv_convertToLCIDPlatform)
+#define uprv_convertToPosix U_ICU_ENTRY_POINT_RENAME(uprv_convertToPosix)
+#define uprv_copyAscii U_ICU_ENTRY_POINT_RENAME(uprv_copyAscii)
+#define uprv_copyEbcdic U_ICU_ENTRY_POINT_RENAME(uprv_copyEbcdic)
+#define uprv_decContextClearStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextClearStatus)
+#define uprv_decContextDefault U_ICU_ENTRY_POINT_RENAME(uprv_decContextDefault)
+#define uprv_decContextGetRounding U_ICU_ENTRY_POINT_RENAME(uprv_decContextGetRounding)
+#define uprv_decContextGetStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextGetStatus)
+#define uprv_decContextRestoreStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextRestoreStatus)
+#define uprv_decContextSaveStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextSaveStatus)
+#define uprv_decContextSetRounding U_ICU_ENTRY_POINT_RENAME(uprv_decContextSetRounding)
+#define uprv_decContextSetStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextSetStatus)
+#define uprv_decContextSetStatusFromString U_ICU_ENTRY_POINT_RENAME(uprv_decContextSetStatusFromString)
+#define uprv_decContextSetStatusFromStringQuiet U_ICU_ENTRY_POINT_RENAME(uprv_decContextSetStatusFromStringQuiet)
+#define uprv_decContextSetStatusQuiet U_ICU_ENTRY_POINT_RENAME(uprv_decContextSetStatusQuiet)
+#define uprv_decContextStatusToString U_ICU_ENTRY_POINT_RENAME(uprv_decContextStatusToString)
+#define uprv_decContextTestSavedStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextTestSavedStatus)
+#define uprv_decContextTestStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextTestStatus)
+#define uprv_decContextZeroStatus U_ICU_ENTRY_POINT_RENAME(uprv_decContextZeroStatus)
+#define uprv_decNumberAbs U_ICU_ENTRY_POINT_RENAME(uprv_decNumberAbs)
+#define uprv_decNumberAdd U_ICU_ENTRY_POINT_RENAME(uprv_decNumberAdd)
+#define uprv_decNumberAnd U_ICU_ENTRY_POINT_RENAME(uprv_decNumberAnd)
+#define uprv_decNumberClassToString U_ICU_ENTRY_POINT_RENAME(uprv_decNumberClassToString)
+#define uprv_decNumberCompare U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCompare)
+#define uprv_decNumberCompareSignal U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCompareSignal)
+#define uprv_decNumberCompareTotal U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCompareTotal)
+#define uprv_decNumberCompareTotalMag U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCompareTotalMag)
+#define uprv_decNumberCopy U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCopy)
+#define uprv_decNumberCopyAbs U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCopyAbs)
+#define uprv_decNumberCopyNegate U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCopyNegate)
+#define uprv_decNumberCopySign U_ICU_ENTRY_POINT_RENAME(uprv_decNumberCopySign)
+#define uprv_decNumberDivide U_ICU_ENTRY_POINT_RENAME(uprv_decNumberDivide)
+#define uprv_decNumberDivideInteger U_ICU_ENTRY_POINT_RENAME(uprv_decNumberDivideInteger)
+#define uprv_decNumberExp U_ICU_ENTRY_POINT_RENAME(uprv_decNumberExp)
+#define uprv_decNumberFMA U_ICU_ENTRY_POINT_RENAME(uprv_decNumberFMA)
+#define uprv_decNumberFromInt32 U_ICU_ENTRY_POINT_RENAME(uprv_decNumberFromInt32)
+#define uprv_decNumberFromString U_ICU_ENTRY_POINT_RENAME(uprv_decNumberFromString)
+#define uprv_decNumberFromUInt32 U_ICU_ENTRY_POINT_RENAME(uprv_decNumberFromUInt32)
+#define uprv_decNumberGetBCD U_ICU_ENTRY_POINT_RENAME(uprv_decNumberGetBCD)
+#define uprv_decNumberInvert U_ICU_ENTRY_POINT_RENAME(uprv_decNumberInvert)
+#define uprv_decNumberIsNormal U_ICU_ENTRY_POINT_RENAME(uprv_decNumberIsNormal)
+#define uprv_decNumberIsSubnormal U_ICU_ENTRY_POINT_RENAME(uprv_decNumberIsSubnormal)
+#define uprv_decNumberLn U_ICU_ENTRY_POINT_RENAME(uprv_decNumberLn)
+#define uprv_decNumberLog10 U_ICU_ENTRY_POINT_RENAME(uprv_decNumberLog10)
+#define uprv_decNumberLogB U_ICU_ENTRY_POINT_RENAME(uprv_decNumberLogB)
+#define uprv_decNumberMax U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMax)
+#define uprv_decNumberMaxMag U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMaxMag)
+#define uprv_decNumberMin U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMin)
+#define uprv_decNumberMinMag U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMinMag)
+#define uprv_decNumberMinus U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMinus)
+#define uprv_decNumberMultiply U_ICU_ENTRY_POINT_RENAME(uprv_decNumberMultiply)
+#define uprv_decNumberNextMinus U_ICU_ENTRY_POINT_RENAME(uprv_decNumberNextMinus)
+#define uprv_decNumberNextPlus U_ICU_ENTRY_POINT_RENAME(uprv_decNumberNextPlus)
+#define uprv_decNumberNextToward U_ICU_ENTRY_POINT_RENAME(uprv_decNumberNextToward)
+#define uprv_decNumberNormalize U_ICU_ENTRY_POINT_RENAME(uprv_decNumberNormalize)
+#define uprv_decNumberOr U_ICU_ENTRY_POINT_RENAME(uprv_decNumberOr)
+#define uprv_decNumberPlus U_ICU_ENTRY_POINT_RENAME(uprv_decNumberPlus)
+#define uprv_decNumberPower U_ICU_ENTRY_POINT_RENAME(uprv_decNumberPower)
+#define uprv_decNumberQuantize U_ICU_ENTRY_POINT_RENAME(uprv_decNumberQuantize)
+#define uprv_decNumberReduce U_ICU_ENTRY_POINT_RENAME(uprv_decNumberReduce)
+#define uprv_decNumberRemainder U_ICU_ENTRY_POINT_RENAME(uprv_decNumberRemainder)
+#define uprv_decNumberRemainderNear U_ICU_ENTRY_POINT_RENAME(uprv_decNumberRemainderNear)
+#define uprv_decNumberRescale U_ICU_ENTRY_POINT_RENAME(uprv_decNumberRescale)
+#define uprv_decNumberRotate U_ICU_ENTRY_POINT_RENAME(uprv_decNumberRotate)
+#define uprv_decNumberSameQuantum U_ICU_ENTRY_POINT_RENAME(uprv_decNumberSameQuantum)
+#define uprv_decNumberScaleB U_ICU_ENTRY_POINT_RENAME(uprv_decNumberScaleB)
+#define uprv_decNumberSetBCD U_ICU_ENTRY_POINT_RENAME(uprv_decNumberSetBCD)
+#define uprv_decNumberShift U_ICU_ENTRY_POINT_RENAME(uprv_decNumberShift)
+#define uprv_decNumberSquareRoot U_ICU_ENTRY_POINT_RENAME(uprv_decNumberSquareRoot)
+#define uprv_decNumberSubtract U_ICU_ENTRY_POINT_RENAME(uprv_decNumberSubtract)
+#define uprv_decNumberToEngString U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToEngString)
+#define uprv_decNumberToInt32 U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToInt32)
+#define uprv_decNumberToIntegralExact U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToIntegralExact)
+#define uprv_decNumberToIntegralValue U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToIntegralValue)
+#define uprv_decNumberToString U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToString)
+#define uprv_decNumberToUInt32 U_ICU_ENTRY_POINT_RENAME(uprv_decNumberToUInt32)
+#define uprv_decNumberTrim U_ICU_ENTRY_POINT_RENAME(uprv_decNumberTrim)
+#define uprv_decNumberVersion U_ICU_ENTRY_POINT_RENAME(uprv_decNumberVersion)
+#define uprv_decNumberXor U_ICU_ENTRY_POINT_RENAME(uprv_decNumberXor)
+#define uprv_decNumberZero U_ICU_ENTRY_POINT_RENAME(uprv_decNumberZero)
+#define uprv_deleteConditionalCE32 U_ICU_ENTRY_POINT_RENAME(uprv_deleteConditionalCE32)
+#define uprv_deleteUObject U_ICU_ENTRY_POINT_RENAME(uprv_deleteUObject)
+#define uprv_dl_close U_ICU_ENTRY_POINT_RENAME(uprv_dl_close)
+#define uprv_dl_open U_ICU_ENTRY_POINT_RENAME(uprv_dl_open)
+#define uprv_dlsym_func U_ICU_ENTRY_POINT_RENAME(uprv_dlsym_func)
+#define uprv_eastrncpy U_ICU_ENTRY_POINT_RENAME(uprv_eastrncpy)
+#define uprv_ebcdicFromAscii U_ICU_ENTRY_POINT_RENAME(uprv_ebcdicFromAscii)
+#define uprv_ebcdicToAscii U_ICU_ENTRY_POINT_RENAME(uprv_ebcdicToAscii)
+#define uprv_ebcdicToLowercaseAscii U_ICU_ENTRY_POINT_RENAME(uprv_ebcdicToLowercaseAscii)
+#define uprv_ebcdictolower U_ICU_ENTRY_POINT_RENAME(uprv_ebcdictolower)
+#define uprv_fabs U_ICU_ENTRY_POINT_RENAME(uprv_fabs)
+#define uprv_floor U_ICU_ENTRY_POINT_RENAME(uprv_floor)
+#define uprv_fmax U_ICU_ENTRY_POINT_RENAME(uprv_fmax)
+#define uprv_fmin U_ICU_ENTRY_POINT_RENAME(uprv_fmin)
+#define uprv_fmod U_ICU_ENTRY_POINT_RENAME(uprv_fmod)
+#define uprv_free U_ICU_ENTRY_POINT_RENAME(uprv_free)
+#define uprv_getCharNameCharacters U_ICU_ENTRY_POINT_RENAME(uprv_getCharNameCharacters)
+#define uprv_getDefaultLocaleID U_ICU_ENTRY_POINT_RENAME(uprv_getDefaultLocaleID)
+#define uprv_getInfinity U_ICU_ENTRY_POINT_RENAME(uprv_getInfinity)
+#define uprv_getMaxCharNameLength U_ICU_ENTRY_POINT_RENAME(uprv_getMaxCharNameLength)
+#define uprv_getMaxValues U_ICU_ENTRY_POINT_RENAME(uprv_getMaxValues)
+#define uprv_getNaN U_ICU_ENTRY_POINT_RENAME(uprv_getNaN)
+#define uprv_getRawUTCtime U_ICU_ENTRY_POINT_RENAME(uprv_getRawUTCtime)
+#define uprv_getStaticCurrencyName U_ICU_ENTRY_POINT_RENAME(uprv_getStaticCurrencyName)
+#define uprv_getUTCtime U_ICU_ENTRY_POINT_RENAME(uprv_getUTCtime)
+#define uprv_int32Comparator U_ICU_ENTRY_POINT_RENAME(uprv_int32Comparator)
+#define uprv_isASCIILetter U_ICU_ENTRY_POINT_RENAME(uprv_isASCIILetter)
+#define uprv_isEbcdicAtSign U_ICU_ENTRY_POINT_RENAME(uprv_isEbcdicAtSign)
+#define uprv_isInfinite U_ICU_ENTRY_POINT_RENAME(uprv_isInfinite)
+#define uprv_isInvariantString U_ICU_ENTRY_POINT_RENAME(uprv_isInvariantString)
+#define uprv_isInvariantUString U_ICU_ENTRY_POINT_RENAME(uprv_isInvariantUString)
+#define uprv_isNaN U_ICU_ENTRY_POINT_RENAME(uprv_isNaN)
+#define uprv_isNegativeInfinity U_ICU_ENTRY_POINT_RENAME(uprv_isNegativeInfinity)
+#define uprv_isPositiveInfinity U_ICU_ENTRY_POINT_RENAME(uprv_isPositiveInfinity)
+#define uprv_itou U_ICU_ENTRY_POINT_RENAME(uprv_itou)
+#define uprv_log U_ICU_ENTRY_POINT_RENAME(uprv_log)
+#define uprv_malloc U_ICU_ENTRY_POINT_RENAME(uprv_malloc)
+#define uprv_mapFile U_ICU_ENTRY_POINT_RENAME(uprv_mapFile)
+#define uprv_max U_ICU_ENTRY_POINT_RENAME(uprv_max)
+#define uprv_maxMantissa U_ICU_ENTRY_POINT_RENAME(uprv_maxMantissa)
+#define uprv_maximumPtr U_ICU_ENTRY_POINT_RENAME(uprv_maximumPtr)
+#define uprv_min U_ICU_ENTRY_POINT_RENAME(uprv_min)
+#define uprv_modf U_ICU_ENTRY_POINT_RENAME(uprv_modf)
+#define uprv_mul32_overflow U_ICU_ENTRY_POINT_RENAME(uprv_mul32_overflow)
+#define uprv_parseCurrency U_ICU_ENTRY_POINT_RENAME(uprv_parseCurrency)
+#define uprv_pathIsAbsolute U_ICU_ENTRY_POINT_RENAME(uprv_pathIsAbsolute)
+#define uprv_pow U_ICU_ENTRY_POINT_RENAME(uprv_pow)
+#define uprv_pow10 U_ICU_ENTRY_POINT_RENAME(uprv_pow10)
+#define uprv_realloc U_ICU_ENTRY_POINT_RENAME(uprv_realloc)
+#define uprv_round U_ICU_ENTRY_POINT_RENAME(uprv_round)
+#define uprv_sortArray U_ICU_ENTRY_POINT_RENAME(uprv_sortArray)
+#define uprv_stableBinarySearch U_ICU_ENTRY_POINT_RENAME(uprv_stableBinarySearch)
+#define uprv_strCompare U_ICU_ENTRY_POINT_RENAME(uprv_strCompare)
+#define uprv_strdup U_ICU_ENTRY_POINT_RENAME(uprv_strdup)
+#define uprv_stricmp U_ICU_ENTRY_POINT_RENAME(uprv_stricmp)
+#define uprv_strndup U_ICU_ENTRY_POINT_RENAME(uprv_strndup)
+#define uprv_strnicmp U_ICU_ENTRY_POINT_RENAME(uprv_strnicmp)
+#define uprv_syntaxError U_ICU_ENTRY_POINT_RENAME(uprv_syntaxError)
+#define uprv_timezone U_ICU_ENTRY_POINT_RENAME(uprv_timezone)
+#define uprv_toupper U_ICU_ENTRY_POINT_RENAME(uprv_toupper)
+#define uprv_trunc U_ICU_ENTRY_POINT_RENAME(uprv_trunc)
+#define uprv_tzname U_ICU_ENTRY_POINT_RENAME(uprv_tzname)
+#define uprv_tzname_clear_cache U_ICU_ENTRY_POINT_RENAME(uprv_tzname_clear_cache)
+#define uprv_tzset U_ICU_ENTRY_POINT_RENAME(uprv_tzset)
+#define uprv_uint16Comparator U_ICU_ENTRY_POINT_RENAME(uprv_uint16Comparator)
+#define uprv_uint32Comparator U_ICU_ENTRY_POINT_RENAME(uprv_uint32Comparator)
+#define uprv_unmapFile U_ICU_ENTRY_POINT_RENAME(uprv_unmapFile)
+#define upvec_cloneArray U_ICU_ENTRY_POINT_RENAME(upvec_cloneArray)
+#define upvec_close U_ICU_ENTRY_POINT_RENAME(upvec_close)
+#define upvec_compact U_ICU_ENTRY_POINT_RENAME(upvec_compact)
+#define upvec_compactToUTrie2Handler U_ICU_ENTRY_POINT_RENAME(upvec_compactToUTrie2Handler)
+#define upvec_compactToUTrie2WithRowIndexes U_ICU_ENTRY_POINT_RENAME(upvec_compactToUTrie2WithRowIndexes)
+#define upvec_getArray U_ICU_ENTRY_POINT_RENAME(upvec_getArray)
+#define upvec_getRow U_ICU_ENTRY_POINT_RENAME(upvec_getRow)
+#define upvec_getValue U_ICU_ENTRY_POINT_RENAME(upvec_getValue)
+#define upvec_open U_ICU_ENTRY_POINT_RENAME(upvec_open)
+#define upvec_setValue U_ICU_ENTRY_POINT_RENAME(upvec_setValue)
+#define uregex_appendReplacement U_ICU_ENTRY_POINT_RENAME(uregex_appendReplacement)
+#define uregex_appendReplacementUText U_ICU_ENTRY_POINT_RENAME(uregex_appendReplacementUText)
+#define uregex_appendTail U_ICU_ENTRY_POINT_RENAME(uregex_appendTail)
+#define uregex_appendTailUText U_ICU_ENTRY_POINT_RENAME(uregex_appendTailUText)
+#define uregex_clone U_ICU_ENTRY_POINT_RENAME(uregex_clone)
+#define uregex_close U_ICU_ENTRY_POINT_RENAME(uregex_close)
+#define uregex_end U_ICU_ENTRY_POINT_RENAME(uregex_end)
+#define uregex_end64 U_ICU_ENTRY_POINT_RENAME(uregex_end64)
+#define uregex_find U_ICU_ENTRY_POINT_RENAME(uregex_find)
+#define uregex_find64 U_ICU_ENTRY_POINT_RENAME(uregex_find64)
+#define uregex_findNext U_ICU_ENTRY_POINT_RENAME(uregex_findNext)
+#define uregex_flags U_ICU_ENTRY_POINT_RENAME(uregex_flags)
+#define uregex_getFindProgressCallback U_ICU_ENTRY_POINT_RENAME(uregex_getFindProgressCallback)
+#define uregex_getMatchCallback U_ICU_ENTRY_POINT_RENAME(uregex_getMatchCallback)
+#define uregex_getStackLimit U_ICU_ENTRY_POINT_RENAME(uregex_getStackLimit)
+#define uregex_getText U_ICU_ENTRY_POINT_RENAME(uregex_getText)
+#define uregex_getTimeLimit U_ICU_ENTRY_POINT_RENAME(uregex_getTimeLimit)
+#define uregex_getUText U_ICU_ENTRY_POINT_RENAME(uregex_getUText)
+#define uregex_group U_ICU_ENTRY_POINT_RENAME(uregex_group)
+#define uregex_groupCount U_ICU_ENTRY_POINT_RENAME(uregex_groupCount)
+#define uregex_groupNumberFromCName U_ICU_ENTRY_POINT_RENAME(uregex_groupNumberFromCName)
+#define uregex_groupNumberFromName U_ICU_ENTRY_POINT_RENAME(uregex_groupNumberFromName)
+#define uregex_groupUText U_ICU_ENTRY_POINT_RENAME(uregex_groupUText)
+#define uregex_hasAnchoringBounds U_ICU_ENTRY_POINT_RENAME(uregex_hasAnchoringBounds)
+#define uregex_hasTransparentBounds U_ICU_ENTRY_POINT_RENAME(uregex_hasTransparentBounds)
+#define uregex_hitEnd U_ICU_ENTRY_POINT_RENAME(uregex_hitEnd)
+#define uregex_lookingAt U_ICU_ENTRY_POINT_RENAME(uregex_lookingAt)
+#define uregex_lookingAt64 U_ICU_ENTRY_POINT_RENAME(uregex_lookingAt64)
+#define uregex_matches U_ICU_ENTRY_POINT_RENAME(uregex_matches)
+#define uregex_matches64 U_ICU_ENTRY_POINT_RENAME(uregex_matches64)
+#define uregex_open U_ICU_ENTRY_POINT_RENAME(uregex_open)
+#define uregex_openC U_ICU_ENTRY_POINT_RENAME(uregex_openC)
+#define uregex_openUText U_ICU_ENTRY_POINT_RENAME(uregex_openUText)
+#define uregex_pattern U_ICU_ENTRY_POINT_RENAME(uregex_pattern)
+#define uregex_patternUText U_ICU_ENTRY_POINT_RENAME(uregex_patternUText)
+#define uregex_refreshUText U_ICU_ENTRY_POINT_RENAME(uregex_refreshUText)
+#define uregex_regionEnd U_ICU_ENTRY_POINT_RENAME(uregex_regionEnd)
+#define uregex_regionEnd64 U_ICU_ENTRY_POINT_RENAME(uregex_regionEnd64)
+#define uregex_regionStart U_ICU_ENTRY_POINT_RENAME(uregex_regionStart)
+#define uregex_regionStart64 U_ICU_ENTRY_POINT_RENAME(uregex_regionStart64)
+#define uregex_replaceAll U_ICU_ENTRY_POINT_RENAME(uregex_replaceAll)
+#define uregex_replaceAllUText U_ICU_ENTRY_POINT_RENAME(uregex_replaceAllUText)
+#define uregex_replaceFirst U_ICU_ENTRY_POINT_RENAME(uregex_replaceFirst)
+#define uregex_replaceFirstUText U_ICU_ENTRY_POINT_RENAME(uregex_replaceFirstUText)
+#define uregex_requireEnd U_ICU_ENTRY_POINT_RENAME(uregex_requireEnd)
+#define uregex_reset U_ICU_ENTRY_POINT_RENAME(uregex_reset)
+#define uregex_reset64 U_ICU_ENTRY_POINT_RENAME(uregex_reset64)
+#define uregex_setFindProgressCallback U_ICU_ENTRY_POINT_RENAME(uregex_setFindProgressCallback)
+#define uregex_setMatchCallback U_ICU_ENTRY_POINT_RENAME(uregex_setMatchCallback)
+#define uregex_setRegion U_ICU_ENTRY_POINT_RENAME(uregex_setRegion)
+#define uregex_setRegion64 U_ICU_ENTRY_POINT_RENAME(uregex_setRegion64)
+#define uregex_setRegionAndStart U_ICU_ENTRY_POINT_RENAME(uregex_setRegionAndStart)
+#define uregex_setStackLimit U_ICU_ENTRY_POINT_RENAME(uregex_setStackLimit)
+#define uregex_setText U_ICU_ENTRY_POINT_RENAME(uregex_setText)
+#define uregex_setTimeLimit U_ICU_ENTRY_POINT_RENAME(uregex_setTimeLimit)
+#define uregex_setUText U_ICU_ENTRY_POINT_RENAME(uregex_setUText)
+#define uregex_split U_ICU_ENTRY_POINT_RENAME(uregex_split)
+#define uregex_splitUText U_ICU_ENTRY_POINT_RENAME(uregex_splitUText)
+#define uregex_start U_ICU_ENTRY_POINT_RENAME(uregex_start)
+#define uregex_start64 U_ICU_ENTRY_POINT_RENAME(uregex_start64)
+#define uregex_ucstr_unescape_charAt U_ICU_ENTRY_POINT_RENAME(uregex_ucstr_unescape_charAt)
+#define uregex_useAnchoringBounds U_ICU_ENTRY_POINT_RENAME(uregex_useAnchoringBounds)
+#define uregex_useTransparentBounds U_ICU_ENTRY_POINT_RENAME(uregex_useTransparentBounds)
+#define uregex_utext_unescape_charAt U_ICU_ENTRY_POINT_RENAME(uregex_utext_unescape_charAt)
+#define uregion_areEqual U_ICU_ENTRY_POINT_RENAME(uregion_areEqual)
+#define uregion_contains U_ICU_ENTRY_POINT_RENAME(uregion_contains)
+#define uregion_getAvailable U_ICU_ENTRY_POINT_RENAME(uregion_getAvailable)
+#define uregion_getContainedRegions U_ICU_ENTRY_POINT_RENAME(uregion_getContainedRegions)
+#define uregion_getContainedRegionsOfType U_ICU_ENTRY_POINT_RENAME(uregion_getContainedRegionsOfType)
+#define uregion_getContainingRegion U_ICU_ENTRY_POINT_RENAME(uregion_getContainingRegion)
+#define uregion_getContainingRegionOfType U_ICU_ENTRY_POINT_RENAME(uregion_getContainingRegionOfType)
+#define uregion_getNumericCode U_ICU_ENTRY_POINT_RENAME(uregion_getNumericCode)
+#define uregion_getPreferredValues U_ICU_ENTRY_POINT_RENAME(uregion_getPreferredValues)
+#define uregion_getRegionCode U_ICU_ENTRY_POINT_RENAME(uregion_getRegionCode)
+#define uregion_getRegionFromCode U_ICU_ENTRY_POINT_RENAME(uregion_getRegionFromCode)
+#define uregion_getRegionFromNumericCode U_ICU_ENTRY_POINT_RENAME(uregion_getRegionFromNumericCode)
+#define uregion_getType U_ICU_ENTRY_POINT_RENAME(uregion_getType)
+#define ureldatefmt_close U_ICU_ENTRY_POINT_RENAME(ureldatefmt_close)
+#define ureldatefmt_closeResult U_ICU_ENTRY_POINT_RENAME(ureldatefmt_closeResult)
+#define ureldatefmt_combineDateAndTime U_ICU_ENTRY_POINT_RENAME(ureldatefmt_combineDateAndTime)
+#define ureldatefmt_format U_ICU_ENTRY_POINT_RENAME(ureldatefmt_format)
+#define ureldatefmt_formatNumeric U_ICU_ENTRY_POINT_RENAME(ureldatefmt_formatNumeric)
+#define ureldatefmt_formatNumericToResult U_ICU_ENTRY_POINT_RENAME(ureldatefmt_formatNumericToResult)
+#define ureldatefmt_formatToResult U_ICU_ENTRY_POINT_RENAME(ureldatefmt_formatToResult)
+#define ureldatefmt_open U_ICU_ENTRY_POINT_RENAME(ureldatefmt_open)
+#define ureldatefmt_openResult U_ICU_ENTRY_POINT_RENAME(ureldatefmt_openResult)
+#define ureldatefmt_resultAsValue U_ICU_ENTRY_POINT_RENAME(ureldatefmt_resultAsValue)
+#define ures_close U_ICU_ENTRY_POINT_RENAME(ures_close)
+#define ures_copyResb U_ICU_ENTRY_POINT_RENAME(ures_copyResb)
+#define ures_countArrayItems U_ICU_ENTRY_POINT_RENAME(ures_countArrayItems)
+#define ures_findResource U_ICU_ENTRY_POINT_RENAME(ures_findResource)
+#define ures_findSubResource U_ICU_ENTRY_POINT_RENAME(ures_findSubResource)
+#define ures_getAllItemsWithFallback U_ICU_ENTRY_POINT_RENAME(ures_getAllItemsWithFallback)
+#define ures_getBinary U_ICU_ENTRY_POINT_RENAME(ures_getBinary)
+#define ures_getByIndex U_ICU_ENTRY_POINT_RENAME(ures_getByIndex)
+#define ures_getByKey U_ICU_ENTRY_POINT_RENAME(ures_getByKey)
+#define ures_getByKeyWithFallback U_ICU_ENTRY_POINT_RENAME(ures_getByKeyWithFallback)
+#define ures_getFunctionalEquivalent U_ICU_ENTRY_POINT_RENAME(ures_getFunctionalEquivalent)
+#define ures_getInt U_ICU_ENTRY_POINT_RENAME(ures_getInt)
+#define ures_getIntVector U_ICU_ENTRY_POINT_RENAME(ures_getIntVector)
+#define ures_getKey U_ICU_ENTRY_POINT_RENAME(ures_getKey)
+#define ures_getKeywordValues U_ICU_ENTRY_POINT_RENAME(ures_getKeywordValues)
+#define ures_getLocale U_ICU_ENTRY_POINT_RENAME(ures_getLocale)
+#define ures_getLocaleByType U_ICU_ENTRY_POINT_RENAME(ures_getLocaleByType)
+#define ures_getLocaleInternal U_ICU_ENTRY_POINT_RENAME(ures_getLocaleInternal)
+#define ures_getName U_ICU_ENTRY_POINT_RENAME(ures_getName)
+#define ures_getNextResource U_ICU_ENTRY_POINT_RENAME(ures_getNextResource)
+#define ures_getNextString U_ICU_ENTRY_POINT_RENAME(ures_getNextString)
+#define ures_getSize U_ICU_ENTRY_POINT_RENAME(ures_getSize)
+#define ures_getString U_ICU_ENTRY_POINT_RENAME(ures_getString)
+#define ures_getStringByIndex U_ICU_ENTRY_POINT_RENAME(ures_getStringByIndex)
+#define ures_getStringByKey U_ICU_ENTRY_POINT_RENAME(ures_getStringByKey)
+#define ures_getStringByKeyWithFallback U_ICU_ENTRY_POINT_RENAME(ures_getStringByKeyWithFallback)
+#define ures_getType U_ICU_ENTRY_POINT_RENAME(ures_getType)
+#define ures_getUInt U_ICU_ENTRY_POINT_RENAME(ures_getUInt)
+#define ures_getUTF8String U_ICU_ENTRY_POINT_RENAME(ures_getUTF8String)
+#define ures_getUTF8StringByIndex U_ICU_ENTRY_POINT_RENAME(ures_getUTF8StringByIndex)
+#define ures_getUTF8StringByKey U_ICU_ENTRY_POINT_RENAME(ures_getUTF8StringByKey)
+#define ures_getValueWithFallback U_ICU_ENTRY_POINT_RENAME(ures_getValueWithFallback)
+#define ures_getVersion U_ICU_ENTRY_POINT_RENAME(ures_getVersion)
+#define ures_getVersionByKey U_ICU_ENTRY_POINT_RENAME(ures_getVersionByKey)
+#define ures_getVersionNumber U_ICU_ENTRY_POINT_RENAME(ures_getVersionNumber)
+#define ures_getVersionNumberInternal U_ICU_ENTRY_POINT_RENAME(ures_getVersionNumberInternal)
+#define ures_hasNext U_ICU_ENTRY_POINT_RENAME(ures_hasNext)
+#define ures_initStackObject U_ICU_ENTRY_POINT_RENAME(ures_initStackObject)
+#define ures_open U_ICU_ENTRY_POINT_RENAME(ures_open)
+#define ures_openAvailableLocales U_ICU_ENTRY_POINT_RENAME(ures_openAvailableLocales)
+#define ures_openDirect U_ICU_ENTRY_POINT_RENAME(ures_openDirect)
+#define ures_openDirectFillIn U_ICU_ENTRY_POINT_RENAME(ures_openDirectFillIn)
+#define ures_openFillIn U_ICU_ENTRY_POINT_RENAME(ures_openFillIn)
+#define ures_openNoDefault U_ICU_ENTRY_POINT_RENAME(ures_openNoDefault)
+#define ures_openU U_ICU_ENTRY_POINT_RENAME(ures_openU)
+#define ures_resetIterator U_ICU_ENTRY_POINT_RENAME(ures_resetIterator)
+#define ures_swap U_ICU_ENTRY_POINT_RENAME(ures_swap)
+#define uscript_breaksBetweenLetters U_ICU_ENTRY_POINT_RENAME(uscript_breaksBetweenLetters)
+#define uscript_closeRun U_ICU_ENTRY_POINT_RENAME(uscript_closeRun)
+#define uscript_getCode U_ICU_ENTRY_POINT_RENAME(uscript_getCode)
+#define uscript_getName U_ICU_ENTRY_POINT_RENAME(uscript_getName)
+#define uscript_getSampleString U_ICU_ENTRY_POINT_RENAME(uscript_getSampleString)
+#define uscript_getSampleUnicodeString U_ICU_ENTRY_POINT_RENAME(uscript_getSampleUnicodeString)
+#define uscript_getScript U_ICU_ENTRY_POINT_RENAME(uscript_getScript)
+#define uscript_getScriptExtensions U_ICU_ENTRY_POINT_RENAME(uscript_getScriptExtensions)
+#define uscript_getShortName U_ICU_ENTRY_POINT_RENAME(uscript_getShortName)
+#define uscript_getUsage U_ICU_ENTRY_POINT_RENAME(uscript_getUsage)
+#define uscript_hasScript U_ICU_ENTRY_POINT_RENAME(uscript_hasScript)
+#define uscript_isCased U_ICU_ENTRY_POINT_RENAME(uscript_isCased)
+#define uscript_isRightToLeft U_ICU_ENTRY_POINT_RENAME(uscript_isRightToLeft)
+#define uscript_nextRun U_ICU_ENTRY_POINT_RENAME(uscript_nextRun)
+#define uscript_openRun U_ICU_ENTRY_POINT_RENAME(uscript_openRun)
+#define uscript_resetRun U_ICU_ENTRY_POINT_RENAME(uscript_resetRun)
+#define uscript_setRunText U_ICU_ENTRY_POINT_RENAME(uscript_setRunText)
+#define usearch_close U_ICU_ENTRY_POINT_RENAME(usearch_close)
+#define usearch_first U_ICU_ENTRY_POINT_RENAME(usearch_first)
+#define usearch_following U_ICU_ENTRY_POINT_RENAME(usearch_following)
+#define usearch_getAttribute U_ICU_ENTRY_POINT_RENAME(usearch_getAttribute)
+#define usearch_getBreakIterator U_ICU_ENTRY_POINT_RENAME(usearch_getBreakIterator)
+#define usearch_getCollator U_ICU_ENTRY_POINT_RENAME(usearch_getCollator)
+#define usearch_getMatchedLength U_ICU_ENTRY_POINT_RENAME(usearch_getMatchedLength)
+#define usearch_getMatchedStart U_ICU_ENTRY_POINT_RENAME(usearch_getMatchedStart)
+#define usearch_getMatchedText U_ICU_ENTRY_POINT_RENAME(usearch_getMatchedText)
+#define usearch_getOffset U_ICU_ENTRY_POINT_RENAME(usearch_getOffset)
+#define usearch_getPattern U_ICU_ENTRY_POINT_RENAME(usearch_getPattern)
+#define usearch_getText U_ICU_ENTRY_POINT_RENAME(usearch_getText)
+#define usearch_handleNextCanonical U_ICU_ENTRY_POINT_RENAME(usearch_handleNextCanonical)
+#define usearch_handleNextExact U_ICU_ENTRY_POINT_RENAME(usearch_handleNextExact)
+#define usearch_handlePreviousCanonical U_ICU_ENTRY_POINT_RENAME(usearch_handlePreviousCanonical)
+#define usearch_handlePreviousExact U_ICU_ENTRY_POINT_RENAME(usearch_handlePreviousExact)
+#define usearch_last U_ICU_ENTRY_POINT_RENAME(usearch_last)
+#define usearch_next U_ICU_ENTRY_POINT_RENAME(usearch_next)
+#define usearch_open U_ICU_ENTRY_POINT_RENAME(usearch_open)
+#define usearch_openFromCollator U_ICU_ENTRY_POINT_RENAME(usearch_openFromCollator)
+#define usearch_preceding U_ICU_ENTRY_POINT_RENAME(usearch_preceding)
+#define usearch_previous U_ICU_ENTRY_POINT_RENAME(usearch_previous)
+#define usearch_reset U_ICU_ENTRY_POINT_RENAME(usearch_reset)
+#define usearch_search U_ICU_ENTRY_POINT_RENAME(usearch_search)
+#define usearch_searchBackwards U_ICU_ENTRY_POINT_RENAME(usearch_searchBackwards)
+#define usearch_setAttribute U_ICU_ENTRY_POINT_RENAME(usearch_setAttribute)
+#define usearch_setBreakIterator U_ICU_ENTRY_POINT_RENAME(usearch_setBreakIterator)
+#define usearch_setCollator U_ICU_ENTRY_POINT_RENAME(usearch_setCollator)
+#define usearch_setOffset U_ICU_ENTRY_POINT_RENAME(usearch_setOffset)
+#define usearch_setPattern U_ICU_ENTRY_POINT_RENAME(usearch_setPattern)
+#define usearch_setText U_ICU_ENTRY_POINT_RENAME(usearch_setText)
+#define uset_add U_ICU_ENTRY_POINT_RENAME(uset_add)
+#define uset_addAll U_ICU_ENTRY_POINT_RENAME(uset_addAll)
+#define uset_addAllCodePoints U_ICU_ENTRY_POINT_RENAME(uset_addAllCodePoints)
+#define uset_addRange U_ICU_ENTRY_POINT_RENAME(uset_addRange)
+#define uset_addString U_ICU_ENTRY_POINT_RENAME(uset_addString)
+#define uset_applyIntPropertyValue U_ICU_ENTRY_POINT_RENAME(uset_applyIntPropertyValue)
+#define uset_applyPattern U_ICU_ENTRY_POINT_RENAME(uset_applyPattern)
+#define uset_applyPropertyAlias U_ICU_ENTRY_POINT_RENAME(uset_applyPropertyAlias)
+#define uset_charAt U_ICU_ENTRY_POINT_RENAME(uset_charAt)
+#define uset_clear U_ICU_ENTRY_POINT_RENAME(uset_clear)
+#define uset_clone U_ICU_ENTRY_POINT_RENAME(uset_clone)
+#define uset_cloneAsThawed U_ICU_ENTRY_POINT_RENAME(uset_cloneAsThawed)
+#define uset_close U_ICU_ENTRY_POINT_RENAME(uset_close)
+#define uset_closeOver U_ICU_ENTRY_POINT_RENAME(uset_closeOver)
+#define uset_compact U_ICU_ENTRY_POINT_RENAME(uset_compact)
+#define uset_complement U_ICU_ENTRY_POINT_RENAME(uset_complement)
+#define uset_complementAll U_ICU_ENTRY_POINT_RENAME(uset_complementAll)
+#define uset_contains U_ICU_ENTRY_POINT_RENAME(uset_contains)
+#define uset_containsAll U_ICU_ENTRY_POINT_RENAME(uset_containsAll)
+#define uset_containsAllCodePoints U_ICU_ENTRY_POINT_RENAME(uset_containsAllCodePoints)
+#define uset_containsNone U_ICU_ENTRY_POINT_RENAME(uset_containsNone)
+#define uset_containsRange U_ICU_ENTRY_POINT_RENAME(uset_containsRange)
+#define uset_containsSome U_ICU_ENTRY_POINT_RENAME(uset_containsSome)
+#define uset_containsString U_ICU_ENTRY_POINT_RENAME(uset_containsString)
+#define uset_equals U_ICU_ENTRY_POINT_RENAME(uset_equals)
+#define uset_freeze U_ICU_ENTRY_POINT_RENAME(uset_freeze)
+#define uset_getItem U_ICU_ENTRY_POINT_RENAME(uset_getItem)
+#define uset_getItemCount U_ICU_ENTRY_POINT_RENAME(uset_getItemCount)
+#define uset_getSerializedRange U_ICU_ENTRY_POINT_RENAME(uset_getSerializedRange)
+#define uset_getSerializedRangeCount U_ICU_ENTRY_POINT_RENAME(uset_getSerializedRangeCount)
+#define uset_getSerializedSet U_ICU_ENTRY_POINT_RENAME(uset_getSerializedSet)
+#define uset_indexOf U_ICU_ENTRY_POINT_RENAME(uset_indexOf)
+#define uset_isEmpty U_ICU_ENTRY_POINT_RENAME(uset_isEmpty)
+#define uset_isFrozen U_ICU_ENTRY_POINT_RENAME(uset_isFrozen)
+#define uset_open U_ICU_ENTRY_POINT_RENAME(uset_open)
+#define uset_openEmpty U_ICU_ENTRY_POINT_RENAME(uset_openEmpty)
+#define uset_openPattern U_ICU_ENTRY_POINT_RENAME(uset_openPattern)
+#define uset_openPatternOptions U_ICU_ENTRY_POINT_RENAME(uset_openPatternOptions)
+#define uset_remove U_ICU_ENTRY_POINT_RENAME(uset_remove)
+#define uset_removeAll U_ICU_ENTRY_POINT_RENAME(uset_removeAll)
+#define uset_removeAllStrings U_ICU_ENTRY_POINT_RENAME(uset_removeAllStrings)
+#define uset_removeRange U_ICU_ENTRY_POINT_RENAME(uset_removeRange)
+#define uset_removeString U_ICU_ENTRY_POINT_RENAME(uset_removeString)
+#define uset_resemblesPattern U_ICU_ENTRY_POINT_RENAME(uset_resemblesPattern)
+#define uset_retain U_ICU_ENTRY_POINT_RENAME(uset_retain)
+#define uset_retainAll U_ICU_ENTRY_POINT_RENAME(uset_retainAll)
+#define uset_serialize U_ICU_ENTRY_POINT_RENAME(uset_serialize)
+#define uset_serializedContains U_ICU_ENTRY_POINT_RENAME(uset_serializedContains)
+#define uset_set U_ICU_ENTRY_POINT_RENAME(uset_set)
+#define uset_setSerializedToOne U_ICU_ENTRY_POINT_RENAME(uset_setSerializedToOne)
+#define uset_size U_ICU_ENTRY_POINT_RENAME(uset_size)
+#define uset_span U_ICU_ENTRY_POINT_RENAME(uset_span)
+#define uset_spanBack U_ICU_ENTRY_POINT_RENAME(uset_spanBack)
+#define uset_spanBackUTF8 U_ICU_ENTRY_POINT_RENAME(uset_spanBackUTF8)
+#define uset_spanUTF8 U_ICU_ENTRY_POINT_RENAME(uset_spanUTF8)
+#define uset_toPattern U_ICU_ENTRY_POINT_RENAME(uset_toPattern)
+#define uspoof_areConfusable U_ICU_ENTRY_POINT_RENAME(uspoof_areConfusable)
+#define uspoof_areConfusableUTF8 U_ICU_ENTRY_POINT_RENAME(uspoof_areConfusableUTF8)
+#define uspoof_areConfusableUnicodeString U_ICU_ENTRY_POINT_RENAME(uspoof_areConfusableUnicodeString)
+#define uspoof_check U_ICU_ENTRY_POINT_RENAME(uspoof_check)
+#define uspoof_check2 U_ICU_ENTRY_POINT_RENAME(uspoof_check2)
+#define uspoof_check2UTF8 U_ICU_ENTRY_POINT_RENAME(uspoof_check2UTF8)
+#define uspoof_check2UnicodeString U_ICU_ENTRY_POINT_RENAME(uspoof_check2UnicodeString)
+#define uspoof_checkUTF8 U_ICU_ENTRY_POINT_RENAME(uspoof_checkUTF8)
+#define uspoof_checkUnicodeString U_ICU_ENTRY_POINT_RENAME(uspoof_checkUnicodeString)
+#define uspoof_clone U_ICU_ENTRY_POINT_RENAME(uspoof_clone)
+#define uspoof_close U_ICU_ENTRY_POINT_RENAME(uspoof_close)
+#define uspoof_closeCheckResult U_ICU_ENTRY_POINT_RENAME(uspoof_closeCheckResult)
+#define uspoof_getAllowedChars U_ICU_ENTRY_POINT_RENAME(uspoof_getAllowedChars)
+#define uspoof_getAllowedLocales U_ICU_ENTRY_POINT_RENAME(uspoof_getAllowedLocales)
+#define uspoof_getAllowedUnicodeSet U_ICU_ENTRY_POINT_RENAME(uspoof_getAllowedUnicodeSet)
+#define uspoof_getCheckResultChecks U_ICU_ENTRY_POINT_RENAME(uspoof_getCheckResultChecks)
+#define uspoof_getCheckResultNumerics U_ICU_ENTRY_POINT_RENAME(uspoof_getCheckResultNumerics)
+#define uspoof_getCheckResultRestrictionLevel U_ICU_ENTRY_POINT_RENAME(uspoof_getCheckResultRestrictionLevel)
+#define uspoof_getChecks U_ICU_ENTRY_POINT_RENAME(uspoof_getChecks)
+#define uspoof_getInclusionSet U_ICU_ENTRY_POINT_RENAME(uspoof_getInclusionSet)
+#define uspoof_getInclusionUnicodeSet U_ICU_ENTRY_POINT_RENAME(uspoof_getInclusionUnicodeSet)
+#define uspoof_getRecommendedSet U_ICU_ENTRY_POINT_RENAME(uspoof_getRecommendedSet)
+#define uspoof_getRecommendedUnicodeSet U_ICU_ENTRY_POINT_RENAME(uspoof_getRecommendedUnicodeSet)
+#define uspoof_getRestrictionLevel U_ICU_ENTRY_POINT_RENAME(uspoof_getRestrictionLevel)
+#define uspoof_getSkeleton U_ICU_ENTRY_POINT_RENAME(uspoof_getSkeleton)
+#define uspoof_getSkeletonUTF8 U_ICU_ENTRY_POINT_RENAME(uspoof_getSkeletonUTF8)
+#define uspoof_getSkeletonUnicodeString U_ICU_ENTRY_POINT_RENAME(uspoof_getSkeletonUnicodeString)
+#define uspoof_internalInitStatics U_ICU_ENTRY_POINT_RENAME(uspoof_internalInitStatics)
+#define uspoof_open U_ICU_ENTRY_POINT_RENAME(uspoof_open)
+#define uspoof_openCheckResult U_ICU_ENTRY_POINT_RENAME(uspoof_openCheckResult)
+#define uspoof_openFromSerialized U_ICU_ENTRY_POINT_RENAME(uspoof_openFromSerialized)
+#define uspoof_openFromSource U_ICU_ENTRY_POINT_RENAME(uspoof_openFromSource)
+#define uspoof_serialize U_ICU_ENTRY_POINT_RENAME(uspoof_serialize)
+#define uspoof_setAllowedChars U_ICU_ENTRY_POINT_RENAME(uspoof_setAllowedChars)
+#define uspoof_setAllowedLocales U_ICU_ENTRY_POINT_RENAME(uspoof_setAllowedLocales)
+#define uspoof_setAllowedUnicodeSet U_ICU_ENTRY_POINT_RENAME(uspoof_setAllowedUnicodeSet)
+#define uspoof_setChecks U_ICU_ENTRY_POINT_RENAME(uspoof_setChecks)
+#define uspoof_setRestrictionLevel U_ICU_ENTRY_POINT_RENAME(uspoof_setRestrictionLevel)
+#define uspoof_swap U_ICU_ENTRY_POINT_RENAME(uspoof_swap)
+#define usprep_close U_ICU_ENTRY_POINT_RENAME(usprep_close)
+#define usprep_open U_ICU_ENTRY_POINT_RENAME(usprep_open)
+#define usprep_openByType U_ICU_ENTRY_POINT_RENAME(usprep_openByType)
+#define usprep_prepare U_ICU_ENTRY_POINT_RENAME(usprep_prepare)
+#define usprep_swap U_ICU_ENTRY_POINT_RENAME(usprep_swap)
+#define ustr_hashCharsN U_ICU_ENTRY_POINT_RENAME(ustr_hashCharsN)
+#define ustr_hashICharsN U_ICU_ENTRY_POINT_RENAME(ustr_hashICharsN)
+#define ustr_hashUCharsN U_ICU_ENTRY_POINT_RENAME(ustr_hashUCharsN)
+#define ustrcase_getCaseLocale U_ICU_ENTRY_POINT_RENAME(ustrcase_getCaseLocale)
+#define ustrcase_getTitleBreakIterator U_ICU_ENTRY_POINT_RENAME(ustrcase_getTitleBreakIterator)
+#define ustrcase_internalFold U_ICU_ENTRY_POINT_RENAME(ustrcase_internalFold)
+#define ustrcase_internalToLower U_ICU_ENTRY_POINT_RENAME(ustrcase_internalToLower)
+#define ustrcase_internalToTitle U_ICU_ENTRY_POINT_RENAME(ustrcase_internalToTitle)
+#define ustrcase_internalToUpper U_ICU_ENTRY_POINT_RENAME(ustrcase_internalToUpper)
+#define ustrcase_map U_ICU_ENTRY_POINT_RENAME(ustrcase_map)
+#define ustrcase_mapWithOverlap U_ICU_ENTRY_POINT_RENAME(ustrcase_mapWithOverlap)
+#define utext_char32At U_ICU_ENTRY_POINT_RENAME(utext_char32At)
+#define utext_clone U_ICU_ENTRY_POINT_RENAME(utext_clone)
+#define utext_close U_ICU_ENTRY_POINT_RENAME(utext_close)
+#define utext_copy U_ICU_ENTRY_POINT_RENAME(utext_copy)
+#define utext_current32 U_ICU_ENTRY_POINT_RENAME(utext_current32)
+#define utext_equals U_ICU_ENTRY_POINT_RENAME(utext_equals)
+#define utext_extract U_ICU_ENTRY_POINT_RENAME(utext_extract)
+#define utext_freeze U_ICU_ENTRY_POINT_RENAME(utext_freeze)
+#define utext_getNativeIndex U_ICU_ENTRY_POINT_RENAME(utext_getNativeIndex)
+#define utext_getPreviousNativeIndex U_ICU_ENTRY_POINT_RENAME(utext_getPreviousNativeIndex)
+#define utext_hasMetaData U_ICU_ENTRY_POINT_RENAME(utext_hasMetaData)
+#define utext_isLengthExpensive U_ICU_ENTRY_POINT_RENAME(utext_isLengthExpensive)
+#define utext_isWritable U_ICU_ENTRY_POINT_RENAME(utext_isWritable)
+#define utext_moveIndex32 U_ICU_ENTRY_POINT_RENAME(utext_moveIndex32)
+#define utext_nativeLength U_ICU_ENTRY_POINT_RENAME(utext_nativeLength)
+#define utext_next32 U_ICU_ENTRY_POINT_RENAME(utext_next32)
+#define utext_next32From U_ICU_ENTRY_POINT_RENAME(utext_next32From)
+#define utext_openCharacterIterator U_ICU_ENTRY_POINT_RENAME(utext_openCharacterIterator)
+#define utext_openConstUnicodeString U_ICU_ENTRY_POINT_RENAME(utext_openConstUnicodeString)
+#define utext_openReplaceable U_ICU_ENTRY_POINT_RENAME(utext_openReplaceable)
+#define utext_openUChars U_ICU_ENTRY_POINT_RENAME(utext_openUChars)
+#define utext_openUTF8 U_ICU_ENTRY_POINT_RENAME(utext_openUTF8)
+#define utext_openUnicodeString U_ICU_ENTRY_POINT_RENAME(utext_openUnicodeString)
+#define utext_previous32 U_ICU_ENTRY_POINT_RENAME(utext_previous32)
+#define utext_previous32From U_ICU_ENTRY_POINT_RENAME(utext_previous32From)
+#define utext_replace U_ICU_ENTRY_POINT_RENAME(utext_replace)
+#define utext_setNativeIndex U_ICU_ENTRY_POINT_RENAME(utext_setNativeIndex)
+#define utext_setup U_ICU_ENTRY_POINT_RENAME(utext_setup)
+#define utf8_appendCharSafeBody U_ICU_ENTRY_POINT_RENAME(utf8_appendCharSafeBody)
+#define utf8_back1SafeBody U_ICU_ENTRY_POINT_RENAME(utf8_back1SafeBody)
+#define utf8_countTrailBytes U_ICU_ENTRY_POINT_RENAME(utf8_countTrailBytes)
+#define utf8_nextCharSafeBody U_ICU_ENTRY_POINT_RENAME(utf8_nextCharSafeBody)
+#define utf8_prevCharSafeBody U_ICU_ENTRY_POINT_RENAME(utf8_prevCharSafeBody)
+#define utmscale_fromInt64 U_ICU_ENTRY_POINT_RENAME(utmscale_fromInt64)
+#define utmscale_getTimeScaleValue U_ICU_ENTRY_POINT_RENAME(utmscale_getTimeScaleValue)
+#define utmscale_toInt64 U_ICU_ENTRY_POINT_RENAME(utmscale_toInt64)
+#define utrace_cleanup U_ICU_ENTRY_POINT_RENAME(utrace_cleanup)
+#define utrace_data U_ICU_ENTRY_POINT_RENAME(utrace_data)
+#define utrace_entry U_ICU_ENTRY_POINT_RENAME(utrace_entry)
+#define utrace_exit U_ICU_ENTRY_POINT_RENAME(utrace_exit)
+#define utrace_format U_ICU_ENTRY_POINT_RENAME(utrace_format)
+#define utrace_functionName U_ICU_ENTRY_POINT_RENAME(utrace_functionName)
+#define utrace_getFunctions U_ICU_ENTRY_POINT_RENAME(utrace_getFunctions)
+#define utrace_getLevel U_ICU_ENTRY_POINT_RENAME(utrace_getLevel)
+#define utrace_setFunctions U_ICU_ENTRY_POINT_RENAME(utrace_setFunctions)
+#define utrace_setLevel U_ICU_ENTRY_POINT_RENAME(utrace_setLevel)
+#define utrace_vformat U_ICU_ENTRY_POINT_RENAME(utrace_vformat)
+#define utrans_clone U_ICU_ENTRY_POINT_RENAME(utrans_clone)
+#define utrans_close U_ICU_ENTRY_POINT_RENAME(utrans_close)
+#define utrans_countAvailableIDs U_ICU_ENTRY_POINT_RENAME(utrans_countAvailableIDs)
+#define utrans_getAvailableID U_ICU_ENTRY_POINT_RENAME(utrans_getAvailableID)
+#define utrans_getID U_ICU_ENTRY_POINT_RENAME(utrans_getID)
+#define utrans_getSourceSet U_ICU_ENTRY_POINT_RENAME(utrans_getSourceSet)
+#define utrans_getUnicodeID U_ICU_ENTRY_POINT_RENAME(utrans_getUnicodeID)
+#define utrans_open U_ICU_ENTRY_POINT_RENAME(utrans_open)
+#define utrans_openIDs U_ICU_ENTRY_POINT_RENAME(utrans_openIDs)
+#define utrans_openInverse U_ICU_ENTRY_POINT_RENAME(utrans_openInverse)
+#define utrans_openU U_ICU_ENTRY_POINT_RENAME(utrans_openU)
+#define utrans_register U_ICU_ENTRY_POINT_RENAME(utrans_register)
+#define utrans_rep_caseContextIterator U_ICU_ENTRY_POINT_RENAME(utrans_rep_caseContextIterator)
+#define utrans_setFilter U_ICU_ENTRY_POINT_RENAME(utrans_setFilter)
+#define utrans_stripRules U_ICU_ENTRY_POINT_RENAME(utrans_stripRules)
+#define utrans_toRules U_ICU_ENTRY_POINT_RENAME(utrans_toRules)
+#define utrans_trans U_ICU_ENTRY_POINT_RENAME(utrans_trans)
+#define utrans_transIncremental U_ICU_ENTRY_POINT_RENAME(utrans_transIncremental)
+#define utrans_transIncrementalUChars U_ICU_ENTRY_POINT_RENAME(utrans_transIncrementalUChars)
+#define utrans_transUChars U_ICU_ENTRY_POINT_RENAME(utrans_transUChars)
+#define utrans_transliterator_cleanup U_ICU_ENTRY_POINT_RENAME(utrans_transliterator_cleanup)
+#define utrans_unregister U_ICU_ENTRY_POINT_RENAME(utrans_unregister)
+#define utrans_unregisterID U_ICU_ENTRY_POINT_RENAME(utrans_unregisterID)
+#define utrie2_clone U_ICU_ENTRY_POINT_RENAME(utrie2_clone)
+#define utrie2_cloneAsThawed U_ICU_ENTRY_POINT_RENAME(utrie2_cloneAsThawed)
+#define utrie2_close U_ICU_ENTRY_POINT_RENAME(utrie2_close)
+#define utrie2_enum U_ICU_ENTRY_POINT_RENAME(utrie2_enum)
+#define utrie2_enumForLeadSurrogate U_ICU_ENTRY_POINT_RENAME(utrie2_enumForLeadSurrogate)
+#define utrie2_freeze U_ICU_ENTRY_POINT_RENAME(utrie2_freeze)
+#define utrie2_fromUTrie U_ICU_ENTRY_POINT_RENAME(utrie2_fromUTrie)
+#define utrie2_get32 U_ICU_ENTRY_POINT_RENAME(utrie2_get32)
+#define utrie2_get32FromLeadSurrogateCodeUnit U_ICU_ENTRY_POINT_RENAME(utrie2_get32FromLeadSurrogateCodeUnit)
+#define utrie2_internalU8NextIndex U_ICU_ENTRY_POINT_RENAME(utrie2_internalU8NextIndex)
+#define utrie2_internalU8PrevIndex U_ICU_ENTRY_POINT_RENAME(utrie2_internalU8PrevIndex)
+#define utrie2_isFrozen U_ICU_ENTRY_POINT_RENAME(utrie2_isFrozen)
+#define utrie2_open U_ICU_ENTRY_POINT_RENAME(utrie2_open)
+#define utrie2_openDummy U_ICU_ENTRY_POINT_RENAME(utrie2_openDummy)
+#define utrie2_openFromSerialized U_ICU_ENTRY_POINT_RENAME(utrie2_openFromSerialized)
+#define utrie2_serialize U_ICU_ENTRY_POINT_RENAME(utrie2_serialize)
+#define utrie2_set32 U_ICU_ENTRY_POINT_RENAME(utrie2_set32)
+#define utrie2_set32ForLeadSurrogateCodeUnit U_ICU_ENTRY_POINT_RENAME(utrie2_set32ForLeadSurrogateCodeUnit)
+#define utrie2_setRange32 U_ICU_ENTRY_POINT_RENAME(utrie2_setRange32)
+#define utrie2_swap U_ICU_ENTRY_POINT_RENAME(utrie2_swap)
+#define utrie_clone U_ICU_ENTRY_POINT_RENAME(utrie_clone)
+#define utrie_close U_ICU_ENTRY_POINT_RENAME(utrie_close)
+#define utrie_defaultGetFoldingOffset U_ICU_ENTRY_POINT_RENAME(utrie_defaultGetFoldingOffset)
+#define utrie_enum U_ICU_ENTRY_POINT_RENAME(utrie_enum)
+#define utrie_get32 U_ICU_ENTRY_POINT_RENAME(utrie_get32)
+#define utrie_getData U_ICU_ENTRY_POINT_RENAME(utrie_getData)
+#define utrie_open U_ICU_ENTRY_POINT_RENAME(utrie_open)
+#define utrie_serialize U_ICU_ENTRY_POINT_RENAME(utrie_serialize)
+#define utrie_set32 U_ICU_ENTRY_POINT_RENAME(utrie_set32)
+#define utrie_setRange32 U_ICU_ENTRY_POINT_RENAME(utrie_setRange32)
+#define utrie_swap U_ICU_ENTRY_POINT_RENAME(utrie_swap)
+#define utrie_swapAnyVersion U_ICU_ENTRY_POINT_RENAME(utrie_swapAnyVersion)
+#define utrie_unserialize U_ICU_ENTRY_POINT_RENAME(utrie_unserialize)
+#define utrie_unserializeDummy U_ICU_ENTRY_POINT_RENAME(utrie_unserializeDummy)
+#define vzone_clone U_ICU_ENTRY_POINT_RENAME(vzone_clone)
+#define vzone_close U_ICU_ENTRY_POINT_RENAME(vzone_close)
+#define vzone_countTransitionRules U_ICU_ENTRY_POINT_RENAME(vzone_countTransitionRules)
+#define vzone_equals U_ICU_ENTRY_POINT_RENAME(vzone_equals)
+#define vzone_getDynamicClassID U_ICU_ENTRY_POINT_RENAME(vzone_getDynamicClassID)
+#define vzone_getLastModified U_ICU_ENTRY_POINT_RENAME(vzone_getLastModified)
+#define vzone_getNextTransition U_ICU_ENTRY_POINT_RENAME(vzone_getNextTransition)
+#define vzone_getOffset U_ICU_ENTRY_POINT_RENAME(vzone_getOffset)
+#define vzone_getOffset2 U_ICU_ENTRY_POINT_RENAME(vzone_getOffset2)
+#define vzone_getOffset3 U_ICU_ENTRY_POINT_RENAME(vzone_getOffset3)
+#define vzone_getPreviousTransition U_ICU_ENTRY_POINT_RENAME(vzone_getPreviousTransition)
+#define vzone_getRawOffset U_ICU_ENTRY_POINT_RENAME(vzone_getRawOffset)
+#define vzone_getStaticClassID U_ICU_ENTRY_POINT_RENAME(vzone_getStaticClassID)
+#define vzone_getTZURL U_ICU_ENTRY_POINT_RENAME(vzone_getTZURL)
+#define vzone_hasSameRules U_ICU_ENTRY_POINT_RENAME(vzone_hasSameRules)
+#define vzone_inDaylightTime U_ICU_ENTRY_POINT_RENAME(vzone_inDaylightTime)
+#define vzone_openData U_ICU_ENTRY_POINT_RENAME(vzone_openData)
+#define vzone_openID U_ICU_ENTRY_POINT_RENAME(vzone_openID)
+#define vzone_setLastModified U_ICU_ENTRY_POINT_RENAME(vzone_setLastModified)
+#define vzone_setRawOffset U_ICU_ENTRY_POINT_RENAME(vzone_setRawOffset)
+#define vzone_setTZURL U_ICU_ENTRY_POINT_RENAME(vzone_setTZURL)
+#define vzone_useDaylightTime U_ICU_ENTRY_POINT_RENAME(vzone_useDaylightTime)
+#define vzone_write U_ICU_ENTRY_POINT_RENAME(vzone_write)
+#define vzone_writeFromStart U_ICU_ENTRY_POINT_RENAME(vzone_writeFromStart)
+#define vzone_writeSimple U_ICU_ENTRY_POINT_RENAME(vzone_writeSimple)
+#define zrule_close U_ICU_ENTRY_POINT_RENAME(zrule_close)
+#define zrule_equals U_ICU_ENTRY_POINT_RENAME(zrule_equals)
+#define zrule_getDSTSavings U_ICU_ENTRY_POINT_RENAME(zrule_getDSTSavings)
+#define zrule_getName U_ICU_ENTRY_POINT_RENAME(zrule_getName)
+#define zrule_getRawOffset U_ICU_ENTRY_POINT_RENAME(zrule_getRawOffset)
+#define zrule_isEquivalentTo U_ICU_ENTRY_POINT_RENAME(zrule_isEquivalentTo)
+#define ztrans_adoptFrom U_ICU_ENTRY_POINT_RENAME(ztrans_adoptFrom)
+#define ztrans_adoptTo U_ICU_ENTRY_POINT_RENAME(ztrans_adoptTo)
+#define ztrans_clone U_ICU_ENTRY_POINT_RENAME(ztrans_clone)
+#define ztrans_close U_ICU_ENTRY_POINT_RENAME(ztrans_close)
+#define ztrans_equals U_ICU_ENTRY_POINT_RENAME(ztrans_equals)
+#define ztrans_getDynamicClassID U_ICU_ENTRY_POINT_RENAME(ztrans_getDynamicClassID)
+#define ztrans_getFrom U_ICU_ENTRY_POINT_RENAME(ztrans_getFrom)
+#define ztrans_getStaticClassID U_ICU_ENTRY_POINT_RENAME(ztrans_getStaticClassID)
+#define ztrans_getTime U_ICU_ENTRY_POINT_RENAME(ztrans_getTime)
+#define ztrans_getTo U_ICU_ENTRY_POINT_RENAME(ztrans_getTo)
+#define ztrans_open U_ICU_ENTRY_POINT_RENAME(ztrans_open)
+#define ztrans_openEmpty U_ICU_ENTRY_POINT_RENAME(ztrans_openEmpty)
+#define ztrans_setFrom U_ICU_ENTRY_POINT_RENAME(ztrans_setFrom)
+#define ztrans_setTime U_ICU_ENTRY_POINT_RENAME(ztrans_setTime)
+#define ztrans_setTo U_ICU_ENTRY_POINT_RENAME(ztrans_setTo)
+
+#endif /* !(defined(_MSC_VER) && defined(__INTELLISENSE__)) */
+#endif /* U_DISABLE_RENAMING */
+#endif /* URENAME_H */
+
diff --git a/thirdparty/icu4c/common/unicode/urep.h b/thirdparty/icu4c/common/unicode/urep.h
new file mode 100644
index 0000000000..932202ddb0
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/urep.h
@@ -0,0 +1,157 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*   Copyright (C) 1997-2010, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   06/23/00    aliu        Creation.
+******************************************************************************
+*/
+
+#ifndef __UREP_H
+#define __UREP_H
+
+#include "unicode/utypes.h"
+
+U_CDECL_BEGIN
+
+/********************************************************************
+ * General Notes
+ ********************************************************************
+ * TODO
+ * Add usage scenario
+ * Add test code
+ * Talk about pinning
+ * Talk about "can truncate result if out of memory"
+ */
+
+/********************************************************************
+ * Data Structures
+ ********************************************************************/
+/**
+ * \file
+ * \brief C API: Callbacks for UReplaceable
+ */
+/**
+ * An opaque replaceable text object.  This will be manipulated only
+ * through the caller-supplied UReplaceableFunctor struct.  Related
+ * to the C++ class Replaceable.
+ * This is currently only used in the Transliterator C API, see utrans.h .
+ * @stable ICU 2.0
+ */
+typedef void* UReplaceable;
+
+/**
+ * A set of function pointers that transliterators use to manipulate a
+ * UReplaceable.  The caller should supply the required functions to
+ * manipulate their text appropriately.  Related to the C++ class
+ * Replaceable.
+ * @stable ICU 2.0
+ */
+typedef struct UReplaceableCallbacks {
+
+    /**
+     * Function pointer that returns the number of UChar code units in
+     * this text.
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @return The length of the text.
+     * @stable ICU 2.0
+     */
+    int32_t (*length)(const UReplaceable* rep);
+
+    /**
+     * Function pointer that returns a UChar code units at the given
+     * offset into this text; 0 <= offset < n, where n is the value
+     * returned by (*length)(rep).  See unistr.h for a description of
+     * charAt() vs. char32At().
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @param offset The index at which to fetch the UChar (code unit).
+     * @return The UChar (code unit) at offset, or U+FFFF if the offset is out of bounds.
+     * @stable ICU 2.0
+     */
+    UChar   (*charAt)(const UReplaceable* rep,
+                      int32_t offset);
+
+    /**
+     * Function pointer that returns a UChar32 code point at the given
+     * offset into this text.  See unistr.h for a description of
+     * charAt() vs. char32At().
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @param offset The index at which to fetch the UChar32 (code point).
+     * @return The UChar32 (code point) at offset, or U+FFFF if the offset is out of bounds.
+     * @stable ICU 2.0
+     */
+    UChar32 (*char32At)(const UReplaceable* rep,
+                        int32_t offset);
+    
+    /**
+     * Function pointer that replaces text between start and limit in
+     * this text with the given text.  Attributes (out of band info)
+     * should be retained.
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @param start the starting index of the text to be replaced,
+     * inclusive.
+     * @param limit the ending index of the text to be replaced,
+     * exclusive.
+     * @param text the new text to replace the UChars from
+     * start..limit-1.
+     * @param textLength the number of UChars at text, or -1 if text
+     * is null-terminated.
+     * @stable ICU 2.0
+     */
+    void    (*replace)(UReplaceable* rep,
+                       int32_t start,
+                       int32_t limit,
+                       const UChar* text,
+                       int32_t textLength);
+    
+    /**
+     * Function pointer that copies the characters in the range
+     * [<tt>start</tt>, <tt>limit</tt>) into the array <tt>dst</tt>.
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @param start offset of first character which will be copied
+     * into the array
+     * @param limit offset immediately following the last character to
+     * be copied
+     * @param dst array in which to copy characters.  The length of
+     * <tt>dst</tt> must be at least <tt>(limit - start)</tt>.
+     * @stable ICU 2.1
+     */
+    void    (*extract)(UReplaceable* rep,
+                       int32_t start,
+                       int32_t limit,
+                       UChar* dst);
+
+    /**
+     * Function pointer that copies text between start and limit in
+     * this text to another index in the text.  Attributes (out of
+     * band info) should be retained.  After this call, there will be
+     * (at least) two copies of the characters originally located at
+     * start..limit-1.
+     *
+     * @param rep A pointer to "this" UReplaceable object.
+     * @param start the starting index of the text to be copied,
+     * inclusive.
+     * @param limit the ending index of the text to be copied,
+     * exclusive.
+     * @param dest the index at which the copy of the UChars should be
+     * inserted.
+     * @stable ICU 2.0
+     */
+    void    (*copy)(UReplaceable* rep,
+                    int32_t start,
+                    int32_t limit,
+                    int32_t dest);    
+
+} UReplaceableCallbacks;
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ures.h b/thirdparty/icu4c/common/unicode/ures.h
new file mode 100644
index 0000000000..fff84043e8
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ures.h
@@ -0,0 +1,911 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File URES.H (formerly CRESBUND.H)
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/01/97    aliu        Creation.
+*   02/22/99    damiba      overhaul.
+*   04/04/99    helena      Fixed internal header inclusion.
+*   04/15/99    Madhu       Updated Javadoc
+*   06/14/99    stephen     Removed functions taking a filename suffix.
+*   07/20/99    stephen     Language-independent typedef to void*
+*   11/09/99    weiv        Added ures_getLocale()
+*   06/24/02    weiv        Added support for resource sharing
+******************************************************************************
+*/
+
+#ifndef URES_H
+#define URES_H
+
+#include "unicode/utypes.h"
+#include "unicode/uloc.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ * \file
+ * \brief C API: Resource Bundle
+ *
+ * <h2>C API: Resource Bundle</h2>
+ *
+ * C API representing a collection of resource information pertaining to a given
+ * locale. A resource bundle provides a way of accessing locale- specific information in
+ * a data file. You create a resource bundle that manages the resources for a given
+ * locale and then ask it for individual resources.
+ * <P>
+ * Resource bundles in ICU4C are currently defined using text files which conform to the following
+ * <a href="http://source.icu-project.org/repos/icu/icuhtml/trunk/design/bnf_rb.txt">BNF definition</a>.
+ * More on resource bundle concepts and syntax can be found in the
+ * <a href="http://icu-project.org/userguide/ResourceManagement.html">Users Guide</a>.
+ * <P>
+ */
+
+/**
+ * UResourceBundle is an opaque type for handles for resource bundles in C APIs.
+ * @stable ICU 2.0
+ */
+struct UResourceBundle;
+
+/**
+ * @stable ICU 2.0
+ */
+typedef struct UResourceBundle UResourceBundle;
+
+/**
+ * Numeric constants for types of resource items.
+ * @see ures_getType
+ * @stable ICU 2.0
+ */
+typedef enum {
+    /** Resource type constant for "no resource". @stable ICU 2.6 */
+    URES_NONE=-1,
+
+    /** Resource type constant for 16-bit Unicode strings. @stable ICU 2.6 */
+    URES_STRING=0,
+
+    /** Resource type constant for binary data. @stable ICU 2.6 */
+    URES_BINARY=1,
+
+    /** Resource type constant for tables of key-value pairs. @stable ICU 2.6 */
+    URES_TABLE=2,
+
+    /**
+     * Resource type constant for aliases;
+     * internally stores a string which identifies the actual resource
+     * storing the data (can be in a different resource bundle).
+     * Resolved internally before delivering the actual resource through the API.
+     * @stable ICU 2.6
+     */
+    URES_ALIAS=3,
+
+    /**
+     * Resource type constant for a single 28-bit integer, interpreted as
+     * signed or unsigned by the ures_getInt() or ures_getUInt() function.
+     * @see ures_getInt
+     * @see ures_getUInt
+     * @stable ICU 2.6
+     */
+    URES_INT=7,
+
+    /** Resource type constant for arrays of resources. @stable ICU 2.6 */
+    URES_ARRAY=8,
+
+    /**
+     * Resource type constant for vectors of 32-bit integers.
+     * @see ures_getIntVector
+     * @stable ICU 2.6
+     */
+    URES_INT_VECTOR = 14,
+#ifndef U_HIDE_DEPRECATED_API
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_NONE=URES_NONE,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_STRING=URES_STRING,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_BINARY=URES_BINARY,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_TABLE=URES_TABLE,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_ALIAS=URES_ALIAS,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_INT=URES_INT,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_ARRAY=URES_ARRAY,
+    /** @deprecated ICU 2.6 Use the URES_ constant instead. */
+    RES_INT_VECTOR=URES_INT_VECTOR,
+    /** @deprecated ICU 2.6 Not used. */
+    RES_RESERVED=15,
+
+    /**
+     * One more than the highest normal UResType value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    URES_LIMIT = 16
+#endif  // U_HIDE_DEPRECATED_API
+} UResType;
+
+/*
+ * Functions to create and destroy resource bundles.
+ */
+
+/**
+ * Opens a UResourceBundle, from which users can extract strings by using
+ * their corresponding keys.
+ * Note that the caller is responsible of calling <TT>ures_close</TT> on each successfully
+ * opened resource bundle.
+ * @param packageName   The packageName and locale together point to an ICU udata object,
+ *                      as defined by <code> udata_open( packageName, "res", locale, err) </code>
+ *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+ *                      a package registered with udata_setAppData(). Using a full file or directory
+ *                      pathname for packageName is deprecated. If NULL, ICU data will be used.
+ * @param locale  specifies the locale for which we want to open the resource
+ *                if NULL, the default locale will be used. If strlen(locale) == 0
+ *                root locale will be used.
+ *
+ * @param status  fills in the outgoing error code.
+ * The UErrorCode err parameter is used to return status information to the user. To
+ * check whether the construction succeeded or not, you should check the value of
+ * U_SUCCESS(err). If you wish more detailed information, you can check for
+ * informational status results which still indicate success. U_USING_FALLBACK_WARNING
+ * indicates that a fall back locale was used. For example, 'de_CH' was requested,
+ * but nothing was found there, so 'de' was used. U_USING_DEFAULT_WARNING indicates that
+ * the default locale data or root locale data was used; neither the requested locale
+ * nor any of its fall back locales could be found. Please see the users guide for more
+ * information on this topic.
+ * @return      a newly allocated resource bundle.
+ * @see ures_close
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle*  U_EXPORT2
+ures_open(const char*    packageName,
+          const char*  locale,
+          UErrorCode*     status);
+
+
+/** This function does not care what kind of localeID is passed in. It simply opens a bundle with
+ *  that name. Fallback mechanism is disabled for the new bundle. If the requested bundle contains
+ *  an %%ALIAS directive, the results are undefined.
+ * @param packageName   The packageName and locale together point to an ICU udata object,
+ *                      as defined by <code> udata_open( packageName, "res", locale, err) </code>
+ *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+ *                      a package registered with udata_setAppData(). Using a full file or directory
+ *                      pathname for packageName is deprecated. If NULL, ICU data will be used.
+ * @param locale  specifies the locale for which we want to open the resource
+ *                if NULL, the default locale will be used. If strlen(locale) == 0
+ *                root locale will be used.
+ *
+ * @param status fills in the outgoing error code. Either U_ZERO_ERROR or U_MISSING_RESOURCE_ERROR
+ * @return      a newly allocated resource bundle or NULL if it doesn't exist.
+ * @see ures_close
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_openDirect(const char* packageName,
+                const char* locale,
+                UErrorCode* status);
+
+/**
+ * Same as ures_open() but takes a const UChar *path.
+ * This path will be converted to char * using the default converter,
+ * then ures_open() is called.
+ *
+ * @param packageName   The packageName and locale together point to an ICU udata object,
+ *                      as defined by <code> udata_open( packageName, "res", locale, err) </code>
+ *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+ *                      a package registered with udata_setAppData(). Using a full file or directory
+ *                      pathname for packageName is deprecated. If NULL, ICU data will be used.
+ * @param locale  specifies the locale for which we want to open the resource
+ *                if NULL, the default locale will be used. If strlen(locale) == 0
+ *                root locale will be used.
+ * @param status  fills in the outgoing error code.
+ * @return      a newly allocated resource bundle.
+ * @see ures_open
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_openU(const UChar* packageName,
+           const char* locale,
+           UErrorCode* status);
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Returns the number of strings/arrays in resource bundles.
+ * Better to use ures_getSize, as this function will be deprecated.
+ *
+ *@param resourceBundle resource bundle containing the desired strings
+ *@param resourceKey key tagging the resource
+ *@param err fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_FALLBACK_WARNING </TT>
+ *@return: for    <STRONG>Arrays</STRONG>: returns the number of resources in the array
+ *                <STRONG>Tables</STRONG>: returns the number of resources in the table
+ *                <STRONG>single string</STRONG>: returns 1
+ *@see ures_getSize
+ * @deprecated ICU 2.8 User ures_getSize instead
+ */
+U_DEPRECATED int32_t U_EXPORT2
+ures_countArrayItems(const UResourceBundle* resourceBundle,
+                     const char* resourceKey,
+                     UErrorCode* err);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+/**
+ * Close a resource bundle, all pointers returned from the various ures_getXXX calls
+ * on this particular bundle should be considered invalid henceforth.
+ *
+ * @param resourceBundle a pointer to a resourceBundle struct. Can be NULL.
+ * @see ures_open
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ures_close(UResourceBundle* resourceBundle);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUResourceBundlePointer
+ * "Smart pointer" class, closes a UResourceBundle via ures_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUResourceBundlePointer, UResourceBundle, ures_close);
+
+U_NAMESPACE_END
+
+#endif
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Return the version number associated with this ResourceBundle as a string. Please
+ * use ures_getVersion as this function is going to be deprecated.
+ *
+ * @param resourceBundle The resource bundle for which the version is checked.
+ * @return  A version number string as specified in the resource bundle or its parent.
+ *          The caller does not own this string.
+ * @see ures_getVersion
+ * @deprecated ICU 2.8 Use ures_getVersion instead.
+ */
+U_DEPRECATED const char* U_EXPORT2
+ures_getVersionNumber(const UResourceBundle*   resourceBundle);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+/**
+ * Return the version number associated with this ResourceBundle as an
+ * UVersionInfo array.
+ *
+ * @param resB The resource bundle for which the version is checked.
+ * @param versionInfo A UVersionInfo array that is filled with the version number
+ *                    as specified in the resource bundle or its parent.
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ures_getVersion(const UResourceBundle* resB,
+                UVersionInfo versionInfo);
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Return the name of the Locale associated with this ResourceBundle. This API allows
+ * you to query for the real locale of the resource. For example, if you requested
+ * "en_US_CALIFORNIA" and only "en_US" bundle exists, "en_US" will be returned.
+ * For subresources, the locale where this resource comes from will be returned.
+ * If fallback has occurred, getLocale will reflect this.
+ *
+ * @param resourceBundle resource bundle in question
+ * @param status just for catching illegal arguments
+ * @return  A Locale name
+ * @deprecated ICU 2.8 Use ures_getLocaleByType instead.
+ */
+U_DEPRECATED const char* U_EXPORT2
+ures_getLocale(const UResourceBundle* resourceBundle,
+               UErrorCode* status);
+#endif  /* U_HIDE_DEPRECATED_API */
+
+/**
+ * Return the name of the Locale associated with this ResourceBundle.
+ * You can choose between requested, valid and real locale.
+ *
+ * @param resourceBundle resource bundle in question
+ * @param type You can choose between requested, valid and actual
+ *             locale. For description see the definition of
+ *             ULocDataLocaleType in uloc.h
+ * @param status just for catching illegal arguments
+ * @return  A Locale name
+ * @stable ICU 2.8
+ */
+U_CAPI const char* U_EXPORT2
+ures_getLocaleByType(const UResourceBundle* resourceBundle,
+                     ULocDataLocaleType type,
+                     UErrorCode* status);
+
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * Same as ures_open() but uses the fill-in parameter instead of allocating a new bundle.
+ *
+ * TODO need to revisit usefulness of this function
+ *      and usage model for fillIn parameters without knowing sizeof(UResourceBundle)
+ * @param r The existing UResourceBundle to fill in. If NULL then status will be
+ *               set to U_ILLEGAL_ARGUMENT_ERROR.
+ * @param packageName   The packageName and locale together point to an ICU udata object,
+ *                      as defined by <code> udata_open( packageName, "res", locale, err) </code>
+ *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+ *                      a package registered with udata_setAppData(). Using a full file or directory
+ *                      pathname for packageName is deprecated. If NULL, ICU data will be used.
+ * @param localeID specifies the locale for which we want to open the resource
+ * @param status The error code.
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+ures_openFillIn(UResourceBundle *r,
+                const char* packageName,
+                const char* localeID,
+                UErrorCode* status);
+#endif  /* U_HIDE_INTERNAL_API */
+
+/**
+ * Returns a string from a string resource type
+ *
+ * @param resourceBundle a string resource
+ * @param len    fills in the length of resulting string
+ * @param status fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                Always check the value of status. Don't count on returning NULL.
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return a pointer to a zero-terminated UChar array which lives in a memory mapped/DLL file.
+ * @see ures_getBinary
+ * @see ures_getIntVector
+ * @see ures_getInt
+ * @see ures_getUInt
+ * @stable ICU 2.0
+ */
+U_CAPI const UChar* U_EXPORT2
+ures_getString(const UResourceBundle* resourceBundle,
+               int32_t* len,
+               UErrorCode* status);
+
+/**
+ * Returns a UTF-8 string from a string resource.
+ * The UTF-8 string may be returnable directly as a pointer, or
+ * it may need to be copied, or transformed from UTF-16 using u_strToUTF8()
+ * or equivalent.
+ *
+ * If forceCopy==true, then the string is always written to the dest buffer
+ * and dest is returned.
+ *
+ * If forceCopy==false, then the string is returned as a pointer if possible,
+ * without needing a dest buffer (it can be NULL). If the string needs to be
+ * copied or transformed, then it may be placed into dest at an arbitrary offset.
+ *
+ * If the string is to be written to dest, then U_BUFFER_OVERFLOW_ERROR and
+ * U_STRING_NOT_TERMINATED_WARNING are set if appropriate, as usual.
+ *
+ * If the string is transformed from UTF-16, then a conversion error may occur
+ * if an unpaired surrogate is encountered. If the function is successful, then
+ * the output UTF-8 string is always well-formed.
+ *
+ * @param resB Resource bundle.
+ * @param dest Destination buffer. Can be NULL only if capacity=*length==0.
+ * @param length Input: Capacity of destination buffer.
+ *               Output: Actual length of the UTF-8 string, not counting the
+ *               terminating NUL, even in case of U_BUFFER_OVERFLOW_ERROR.
+ *               Can be NULL, meaning capacity=0 and the string length is not
+ *               returned to the caller.
+ * @param forceCopy If true, then the output string will always be written to
+ *                  dest, with U_BUFFER_OVERFLOW_ERROR and
+ *                  U_STRING_NOT_TERMINATED_WARNING set if appropriate.
+ *                  If false, then the dest buffer may or may not contain a
+ *                  copy of the string. dest may or may not be modified.
+ *                  If a copy needs to be written, then the UErrorCode parameter
+ *                  indicates overflow etc. as usual.
+ * @param status Pointer to a standard ICU error code. Its input value must
+ *               pass the U_SUCCESS() test, or else the function returns
+ *               immediately. Check for U_FAILURE() on output or use with
+ *               function chaining. (See User Guide for details.)
+ * @return The pointer to the UTF-8 string. It may be dest, or at some offset
+ *         from dest (only if !forceCopy), or in unrelated memory.
+ *         Always NUL-terminated unless the string was written to dest and
+ *         length==capacity (in which case U_STRING_NOT_TERMINATED_WARNING is set).
+ *
+ * @see ures_getString
+ * @see u_strToUTF8
+ * @stable ICU 3.6
+ */
+U_CAPI const char * U_EXPORT2
+ures_getUTF8String(const UResourceBundle *resB,
+                   char *dest, int32_t *length,
+                   UBool forceCopy,
+                   UErrorCode *status);
+
+/**
+ * Returns a binary data from a binary resource.
+ *
+ * @param resourceBundle a string resource
+ * @param len    fills in the length of resulting byte chunk
+ * @param status fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                Always check the value of status. Don't count on returning NULL.
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return a pointer to a chunk of unsigned bytes which live in a memory mapped/DLL file.
+ * @see ures_getString
+ * @see ures_getIntVector
+ * @see ures_getInt
+ * @see ures_getUInt
+ * @stable ICU 2.0
+ */
+U_CAPI const uint8_t* U_EXPORT2
+ures_getBinary(const UResourceBundle* resourceBundle,
+               int32_t* len,
+               UErrorCode* status);
+
+/**
+ * Returns a 32 bit integer array from a resource.
+ *
+ * @param resourceBundle an int vector resource
+ * @param len    fills in the length of resulting byte chunk
+ * @param status fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                Always check the value of status. Don't count on returning NULL.
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return a pointer to a chunk of integers which live in a memory mapped/DLL file.
+ * @see ures_getBinary
+ * @see ures_getString
+ * @see ures_getInt
+ * @see ures_getUInt
+ * @stable ICU 2.0
+ */
+U_CAPI const int32_t* U_EXPORT2
+ures_getIntVector(const UResourceBundle* resourceBundle,
+                  int32_t* len,
+                  UErrorCode* status);
+
+/**
+ * Returns an unsigned integer from a resource.
+ * This integer is originally 28 bits.
+ *
+ * @param resourceBundle a string resource
+ * @param status fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return an integer value
+ * @see ures_getInt
+ * @see ures_getIntVector
+ * @see ures_getBinary
+ * @see ures_getString
+ * @stable ICU 2.0
+ */
+U_CAPI uint32_t U_EXPORT2
+ures_getUInt(const UResourceBundle* resourceBundle,
+             UErrorCode *status);
+
+/**
+ * Returns a signed integer from a resource.
+ * This integer is originally 28 bit and the sign gets propagated.
+ *
+ * @param resourceBundle a string resource
+ * @param status  fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return an integer value
+ * @see ures_getUInt
+ * @see ures_getIntVector
+ * @see ures_getBinary
+ * @see ures_getString
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ures_getInt(const UResourceBundle* resourceBundle,
+            UErrorCode *status);
+
+/**
+ * Returns the size of a resource. Size for scalar types is always 1,
+ * and for vector/table types is the number of child resources.
+ * @warning Integer array is treated as a scalar type. There are no
+ *          APIs to access individual members of an integer array. It
+ *          is always returned as a whole.
+ * @param resourceBundle a resource
+ * @return number of resources in a given resource.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+ures_getSize(const UResourceBundle *resourceBundle);
+
+/**
+ * Returns the type of a resource. Available types are defined in enum UResType
+ *
+ * @param resourceBundle a resource
+ * @return type of the given resource.
+ * @see UResType
+ * @stable ICU 2.0
+ */
+U_CAPI UResType U_EXPORT2
+ures_getType(const UResourceBundle *resourceBundle);
+
+/**
+ * Returns the key associated with a given resource. Not all the resources have a key - only
+ * those that are members of a table.
+ *
+ * @param resourceBundle a resource
+ * @return a key associated to this resource, or NULL if it doesn't have a key
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+ures_getKey(const UResourceBundle *resourceBundle);
+
+/* ITERATION API
+    This API provides means for iterating through a resource
+*/
+
+/**
+ * Resets the internal context of a resource so that iteration starts from the first element.
+ *
+ * @param resourceBundle a resource
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+ures_resetIterator(UResourceBundle *resourceBundle);
+
+/**
+ * Checks whether the given resource has another element to iterate over.
+ *
+ * @param resourceBundle a resource
+ * @return true if there are more elements, false if there is no more elements
+ * @stable ICU 2.0
+ */
+U_CAPI UBool U_EXPORT2
+ures_hasNext(const UResourceBundle *resourceBundle);
+
+/**
+ * Returns the next resource in a given resource or NULL if there are no more resources
+ * to iterate over. Features a fill-in parameter.
+ *
+ * @param resourceBundle    a resource
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be closed by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status            fills in the outgoing error code. You may still get a non NULL result even if an
+ *                          error occurred. Check status instead.
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must close it
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_getNextResource(UResourceBundle *resourceBundle,
+                     UResourceBundle *fillIn,
+                     UErrorCode *status);
+
+/**
+ * Returns the next string in a given resource or NULL if there are no more resources
+ * to iterate over.
+ *
+ * @param resourceBundle    a resource
+ * @param len               fill in length of the string
+ * @param key               fill in for key associated with this string. NULL if no key
+ * @param status            fills in the outgoing error code. If an error occurred, we may return NULL, but don't
+ *                          count on it. Check status instead!
+ * @return a pointer to a zero-terminated UChar array which lives in a memory mapped/DLL file.
+ * @stable ICU 2.0
+ */
+U_CAPI const UChar* U_EXPORT2
+ures_getNextString(UResourceBundle *resourceBundle,
+                   int32_t* len,
+                   const char ** key,
+                   UErrorCode *status);
+
+/**
+ * Returns the resource in a given resource at the specified index. Features a fill-in parameter.
+ *
+ * @param resourceBundle    the resource bundle from which to get a sub-resource
+ * @param indexR            an index to the wanted resource.
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be closed by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status            fills in the outgoing error code. Don't count on NULL being returned if an error has
+ *                          occurred. Check status instead.
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must close it
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_getByIndex(const UResourceBundle *resourceBundle,
+                int32_t indexR,
+                UResourceBundle *fillIn,
+                UErrorCode *status);
+
+/**
+ * Returns the string in a given resource at the specified index.
+ *
+ * @param resourceBundle    a resource
+ * @param indexS            an index to the wanted string.
+ * @param len               fill in length of the string
+ * @param status            fills in the outgoing error code. If an error occurred, we may return NULL, but don't
+ *                          count on it. Check status instead!
+ * @return                  a pointer to a zero-terminated UChar array which lives in a memory mapped/DLL file.
+ * @stable ICU 2.0
+ */
+U_CAPI const UChar* U_EXPORT2
+ures_getStringByIndex(const UResourceBundle *resourceBundle,
+                      int32_t indexS,
+                      int32_t* len,
+                      UErrorCode *status);
+
+/**
+ * Returns a UTF-8 string from a resource at the specified index.
+ * The UTF-8 string may be returnable directly as a pointer, or
+ * it may need to be copied, or transformed from UTF-16 using u_strToUTF8()
+ * or equivalent.
+ *
+ * If forceCopy==true, then the string is always written to the dest buffer
+ * and dest is returned.
+ *
+ * If forceCopy==false, then the string is returned as a pointer if possible,
+ * without needing a dest buffer (it can be NULL). If the string needs to be
+ * copied or transformed, then it may be placed into dest at an arbitrary offset.
+ *
+ * If the string is to be written to dest, then U_BUFFER_OVERFLOW_ERROR and
+ * U_STRING_NOT_TERMINATED_WARNING are set if appropriate, as usual.
+ *
+ * If the string is transformed from UTF-16, then a conversion error may occur
+ * if an unpaired surrogate is encountered. If the function is successful, then
+ * the output UTF-8 string is always well-formed.
+ *
+ * @param resB Resource bundle.
+ * @param stringIndex An index to the wanted string.
+ * @param dest Destination buffer. Can be NULL only if capacity=*length==0.
+ * @param pLength Input: Capacity of destination buffer.
+ *               Output: Actual length of the UTF-8 string, not counting the
+ *               terminating NUL, even in case of U_BUFFER_OVERFLOW_ERROR.
+ *               Can be NULL, meaning capacity=0 and the string length is not
+ *               returned to the caller.
+ * @param forceCopy If true, then the output string will always be written to
+ *                  dest, with U_BUFFER_OVERFLOW_ERROR and
+ *                  U_STRING_NOT_TERMINATED_WARNING set if appropriate.
+ *                  If false, then the dest buffer may or may not contain a
+ *                  copy of the string. dest may or may not be modified.
+ *                  If a copy needs to be written, then the UErrorCode parameter
+ *                  indicates overflow etc. as usual.
+ * @param status Pointer to a standard ICU error code. Its input value must
+ *               pass the U_SUCCESS() test, or else the function returns
+ *               immediately. Check for U_FAILURE() on output or use with
+ *               function chaining. (See User Guide for details.)
+ * @return The pointer to the UTF-8 string. It may be dest, or at some offset
+ *         from dest (only if !forceCopy), or in unrelated memory.
+ *         Always NUL-terminated unless the string was written to dest and
+ *         length==capacity (in which case U_STRING_NOT_TERMINATED_WARNING is set).
+ *
+ * @see ures_getStringByIndex
+ * @see u_strToUTF8
+ * @stable ICU 3.6
+ */
+U_CAPI const char * U_EXPORT2
+ures_getUTF8StringByIndex(const UResourceBundle *resB,
+                          int32_t stringIndex,
+                          char *dest, int32_t *pLength,
+                          UBool forceCopy,
+                          UErrorCode *status);
+
+/**
+ * Returns a resource in a given resource that has a given key. This procedure works only with table
+ * resources. Features a fill-in parameter.
+ *
+ * @param resourceBundle    a resource
+ * @param key               a key associated with the wanted resource
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be closed by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status            fills in the outgoing error code.
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must close it
+ * @stable ICU 2.0
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_getByKey(const UResourceBundle *resourceBundle,
+              const char* key,
+              UResourceBundle *fillIn,
+              UErrorCode *status);
+
+/**
+ * Returns a string in a given resource that has a given key. This procedure works only with table
+ * resources.
+ *
+ * @param resB              a resource
+ * @param key               a key associated with the wanted string
+ * @param len               fill in length of the string
+ * @param status            fills in the outgoing error code. If an error occurred, we may return NULL, but don't
+ *                          count on it. Check status instead!
+ * @return                  a pointer to a zero-terminated UChar array which lives in a memory mapped/DLL file.
+ * @stable ICU 2.0
+ */
+U_CAPI const UChar* U_EXPORT2
+ures_getStringByKey(const UResourceBundle *resB,
+                    const char* key,
+                    int32_t* len,
+                    UErrorCode *status);
+
+/**
+ * Returns a UTF-8 string from a resource and a key.
+ * This function works only with table resources.
+ *
+ * The UTF-8 string may be returnable directly as a pointer, or
+ * it may need to be copied, or transformed from UTF-16 using u_strToUTF8()
+ * or equivalent.
+ *
+ * If forceCopy==true, then the string is always written to the dest buffer
+ * and dest is returned.
+ *
+ * If forceCopy==false, then the string is returned as a pointer if possible,
+ * without needing a dest buffer (it can be NULL). If the string needs to be
+ * copied or transformed, then it may be placed into dest at an arbitrary offset.
+ *
+ * If the string is to be written to dest, then U_BUFFER_OVERFLOW_ERROR and
+ * U_STRING_NOT_TERMINATED_WARNING are set if appropriate, as usual.
+ *
+ * If the string is transformed from UTF-16, then a conversion error may occur
+ * if an unpaired surrogate is encountered. If the function is successful, then
+ * the output UTF-8 string is always well-formed.
+ *
+ * @param resB Resource bundle.
+ * @param key  A key associated with the wanted resource
+ * @param dest Destination buffer. Can be NULL only if capacity=*length==0.
+ * @param pLength Input: Capacity of destination buffer.
+ *               Output: Actual length of the UTF-8 string, not counting the
+ *               terminating NUL, even in case of U_BUFFER_OVERFLOW_ERROR.
+ *               Can be NULL, meaning capacity=0 and the string length is not
+ *               returned to the caller.
+ * @param forceCopy If true, then the output string will always be written to
+ *                  dest, with U_BUFFER_OVERFLOW_ERROR and
+ *                  U_STRING_NOT_TERMINATED_WARNING set if appropriate.
+ *                  If false, then the dest buffer may or may not contain a
+ *                  copy of the string. dest may or may not be modified.
+ *                  If a copy needs to be written, then the UErrorCode parameter
+ *                  indicates overflow etc. as usual.
+ * @param status Pointer to a standard ICU error code. Its input value must
+ *               pass the U_SUCCESS() test, or else the function returns
+ *               immediately. Check for U_FAILURE() on output or use with
+ *               function chaining. (See User Guide for details.)
+ * @return The pointer to the UTF-8 string. It may be dest, or at some offset
+ *         from dest (only if !forceCopy), or in unrelated memory.
+ *         Always NUL-terminated unless the string was written to dest and
+ *         length==capacity (in which case U_STRING_NOT_TERMINATED_WARNING is set).
+ *
+ * @see ures_getStringByKey
+ * @see u_strToUTF8
+ * @stable ICU 3.6
+ */
+U_CAPI const char * U_EXPORT2
+ures_getUTF8StringByKey(const UResourceBundle *resB,
+                        const char *key,
+                        char *dest, int32_t *pLength,
+                        UBool forceCopy,
+                        UErrorCode *status);
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/unistr.h"
+
+U_NAMESPACE_BEGIN
+/**
+ * Returns the string value from a string resource bundle.
+ *
+ * @param resB    a resource, should have type URES_STRING
+ * @param status: fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return The string value, or a bogus string if there is a failure UErrorCode.
+ * @stable ICU 2.0
+ */
+inline UnicodeString
+ures_getUnicodeString(const UResourceBundle *resB, UErrorCode* status) {
+    UnicodeString result;
+    int32_t len = 0;
+    const UChar *r = ures_getString(resB, &len, status);
+    if(U_SUCCESS(*status)) {
+        result.setTo(true, r, len);
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+/**
+ * Returns the next string in a resource, or an empty string if there are no more resources
+ * to iterate over.
+ * Use ures_getNextString() instead to distinguish between
+ * the end of the iteration and a real empty string value.
+ *
+ * @param resB              a resource
+ * @param key               fill in for key associated with this string
+ * @param status            fills in the outgoing error code
+ * @return The string value, or a bogus string if there is a failure UErrorCode.
+ * @stable ICU 2.0
+ */
+inline UnicodeString
+ures_getNextUnicodeString(UResourceBundle *resB, const char ** key, UErrorCode* status) {
+    UnicodeString result;
+    int32_t len = 0;
+    const UChar* r = ures_getNextString(resB, &len, key, status);
+    if(U_SUCCESS(*status)) {
+        result.setTo(true, r, len);
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+/**
+ * Returns the string in a given resource array or table at the specified index.
+ *
+ * @param resB              a resource
+ * @param indexS            an index to the wanted string.
+ * @param status            fills in the outgoing error code
+ * @return The string value, or a bogus string if there is a failure UErrorCode.
+ * @stable ICU 2.0
+ */
+inline UnicodeString
+ures_getUnicodeStringByIndex(const UResourceBundle *resB, int32_t indexS, UErrorCode* status) {
+    UnicodeString result;
+    int32_t len = 0;
+    const UChar* r = ures_getStringByIndex(resB, indexS, &len, status);
+    if(U_SUCCESS(*status)) {
+        result.setTo(true, r, len);
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+/**
+ * Returns a string in a resource that has a given key.
+ * This procedure works only with table resources.
+ *
+ * @param resB              a resource
+ * @param key               a key associated with the wanted string
+ * @param status            fills in the outgoing error code
+ * @return The string value, or a bogus string if there is a failure UErrorCode.
+ * @stable ICU 2.0
+ */
+inline UnicodeString
+ures_getUnicodeStringByKey(const UResourceBundle *resB, const char* key, UErrorCode* status) {
+    UnicodeString result;
+    int32_t len = 0;
+    const UChar* r = ures_getStringByKey(resB, key, &len, status);
+    if(U_SUCCESS(*status)) {
+        result.setTo(true, r, len);
+    } else {
+        result.setToBogus();
+    }
+    return result;
+}
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Create a string enumerator, owned by the caller, of all locales located within
+ * the specified resource tree.
+ * @param packageName name of the tree, such as (NULL) or U_ICUDATA_ALIAS or  or "ICUDATA-coll"
+ * This call is similar to uloc_getAvailable().
+ * @param status error code
+ * @stable ICU 3.2
+ */
+U_CAPI UEnumeration* U_EXPORT2
+ures_openAvailableLocales(const char *packageName, UErrorCode *status);
+
+
+#endif /*_URES*/
+/*eof*/
diff --git a/thirdparty/icu4c/common/unicode/uscript.h b/thirdparty/icu4c/common/unicode/uscript.h
new file mode 100644
index 0000000000..8448afda76
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uscript.h
@@ -0,0 +1,708 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ **********************************************************************
+ *   Copyright (C) 1997-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ **********************************************************************
+ *
+ * File USCRIPT.H
+ *
+ * Modification History:
+ *
+ *   Date        Name        Description
+ *   07/06/2001    Ram         Creation.
+ ******************************************************************************
+ */
+
+#ifndef USCRIPT_H
+#define USCRIPT_H
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief C API: Unicode Script Information
+ */
+
+/**
+ * Constants for ISO 15924 script codes.
+ *
+ * The current set of script code constants supports at least all scripts
+ * that are encoded in the version of Unicode which ICU currently supports.
+ * The names of the constants are usually derived from the
+ * Unicode script property value aliases.
+ * See UAX #24 Unicode Script Property (http://www.unicode.org/reports/tr24/)
+ * and http://www.unicode.org/Public/UCD/latest/ucd/PropertyValueAliases.txt .
+ *
+ * In addition, constants for many ISO 15924 script codes
+ * are included, for use with language tags, CLDR data, and similar.
+ * Some of those codes are not used in the Unicode Character Database (UCD).
+ * For example, there are no characters that have a UCD script property value of
+ * Hans or Hant. All Han ideographs have the Hani script property value in Unicode.
+ *
+ * Private-use codes Qaaa..Qabx are not included, except as used in the UCD or in CLDR.
+ *
+ * Starting with ICU 55, script codes are only added when their scripts
+ * have been or will certainly be encoded in Unicode,
+ * and have been assigned Unicode script property value aliases,
+ * to ensure that their script names are stable and match the names of the constants.
+ * Script codes like Latf and Aran that are not subject to separate encoding
+ * may be added at any time.
+ *
+ * @stable ICU 2.2
+ */
+typedef enum UScriptCode {
+    /*
+     * Note: UScriptCode constants and their ISO script code comments
+     * are parsed by preparseucd.py.
+     * It matches lines like
+     *     USCRIPT_<Unicode Script value name> = <integer>,  / * <ISO script code> * /
+     */
+
+      /** @stable ICU 2.2 */
+      USCRIPT_INVALID_CODE = -1,
+      /** @stable ICU 2.2 */
+      USCRIPT_COMMON       =  0,  /* Zyyy */
+      /** @stable ICU 2.2 */
+      USCRIPT_INHERITED    =  1,  /* Zinh */ /* "Code for inherited script", for non-spacing combining marks; also Qaai */
+      /** @stable ICU 2.2 */
+      USCRIPT_ARABIC       =  2,  /* Arab */
+      /** @stable ICU 2.2 */
+      USCRIPT_ARMENIAN     =  3,  /* Armn */
+      /** @stable ICU 2.2 */
+      USCRIPT_BENGALI      =  4,  /* Beng */
+      /** @stable ICU 2.2 */
+      USCRIPT_BOPOMOFO     =  5,  /* Bopo */
+      /** @stable ICU 2.2 */
+      USCRIPT_CHEROKEE     =  6,  /* Cher */
+      /** @stable ICU 2.2 */
+      USCRIPT_COPTIC       =  7,  /* Copt */
+      /** @stable ICU 2.2 */
+      USCRIPT_CYRILLIC     =  8,  /* Cyrl */
+      /** @stable ICU 2.2 */
+      USCRIPT_DESERET      =  9,  /* Dsrt */
+      /** @stable ICU 2.2 */
+      USCRIPT_DEVANAGARI   = 10,  /* Deva */
+      /** @stable ICU 2.2 */
+      USCRIPT_ETHIOPIC     = 11,  /* Ethi */
+      /** @stable ICU 2.2 */
+      USCRIPT_GEORGIAN     = 12,  /* Geor */
+      /** @stable ICU 2.2 */
+      USCRIPT_GOTHIC       = 13,  /* Goth */
+      /** @stable ICU 2.2 */
+      USCRIPT_GREEK        = 14,  /* Grek */
+      /** @stable ICU 2.2 */
+      USCRIPT_GUJARATI     = 15,  /* Gujr */
+      /** @stable ICU 2.2 */
+      USCRIPT_GURMUKHI     = 16,  /* Guru */
+      /** @stable ICU 2.2 */
+      USCRIPT_HAN          = 17,  /* Hani */
+      /** @stable ICU 2.2 */
+      USCRIPT_HANGUL       = 18,  /* Hang */
+      /** @stable ICU 2.2 */
+      USCRIPT_HEBREW       = 19,  /* Hebr */
+      /** @stable ICU 2.2 */
+      USCRIPT_HIRAGANA     = 20,  /* Hira */
+      /** @stable ICU 2.2 */
+      USCRIPT_KANNADA      = 21,  /* Knda */
+      /** @stable ICU 2.2 */
+      USCRIPT_KATAKANA     = 22,  /* Kana */
+      /** @stable ICU 2.2 */
+      USCRIPT_KHMER        = 23,  /* Khmr */
+      /** @stable ICU 2.2 */
+      USCRIPT_LAO          = 24,  /* Laoo */
+      /** @stable ICU 2.2 */
+      USCRIPT_LATIN        = 25,  /* Latn */
+      /** @stable ICU 2.2 */
+      USCRIPT_MALAYALAM    = 26,  /* Mlym */
+      /** @stable ICU 2.2 */
+      USCRIPT_MONGOLIAN    = 27,  /* Mong */
+      /** @stable ICU 2.2 */
+      USCRIPT_MYANMAR      = 28,  /* Mymr */
+      /** @stable ICU 2.2 */
+      USCRIPT_OGHAM        = 29,  /* Ogam */
+      /** @stable ICU 2.2 */
+      USCRIPT_OLD_ITALIC   = 30,  /* Ital */
+      /** @stable ICU 2.2 */
+      USCRIPT_ORIYA        = 31,  /* Orya */
+      /** @stable ICU 2.2 */
+      USCRIPT_RUNIC        = 32,  /* Runr */
+      /** @stable ICU 2.2 */
+      USCRIPT_SINHALA      = 33,  /* Sinh */
+      /** @stable ICU 2.2 */
+      USCRIPT_SYRIAC       = 34,  /* Syrc */
+      /** @stable ICU 2.2 */
+      USCRIPT_TAMIL        = 35,  /* Taml */
+      /** @stable ICU 2.2 */
+      USCRIPT_TELUGU       = 36,  /* Telu */
+      /** @stable ICU 2.2 */
+      USCRIPT_THAANA       = 37,  /* Thaa */
+      /** @stable ICU 2.2 */
+      USCRIPT_THAI         = 38,  /* Thai */
+      /** @stable ICU 2.2 */
+      USCRIPT_TIBETAN      = 39,  /* Tibt */
+      /** Canadian_Aboriginal script. @stable ICU 2.6 */
+      USCRIPT_CANADIAN_ABORIGINAL = 40,  /* Cans */
+      /** Canadian_Aboriginal script (alias). @stable ICU 2.2 */
+      USCRIPT_UCAS         = USCRIPT_CANADIAN_ABORIGINAL,
+      /** @stable ICU 2.2 */
+      USCRIPT_YI           = 41,  /* Yiii */
+      /* New scripts in Unicode 3.2 */
+      /** @stable ICU 2.2 */
+      USCRIPT_TAGALOG      = 42,  /* Tglg */
+      /** @stable ICU 2.2 */
+      USCRIPT_HANUNOO      = 43,  /* Hano */
+      /** @stable ICU 2.2 */
+      USCRIPT_BUHID        = 44,  /* Buhd */
+      /** @stable ICU 2.2 */
+      USCRIPT_TAGBANWA     = 45,  /* Tagb */
+
+      /* New scripts in Unicode 4 */
+      /** @stable ICU 2.6 */
+      USCRIPT_BRAILLE      = 46,  /* Brai */
+      /** @stable ICU 2.6 */
+      USCRIPT_CYPRIOT      = 47,  /* Cprt */
+      /** @stable ICU 2.6 */
+      USCRIPT_LIMBU        = 48,  /* Limb */
+      /** @stable ICU 2.6 */
+      USCRIPT_LINEAR_B     = 49,  /* Linb */
+      /** @stable ICU 2.6 */
+      USCRIPT_OSMANYA      = 50,  /* Osma */
+      /** @stable ICU 2.6 */
+      USCRIPT_SHAVIAN      = 51,  /* Shaw */
+      /** @stable ICU 2.6 */
+      USCRIPT_TAI_LE       = 52,  /* Tale */
+      /** @stable ICU 2.6 */
+      USCRIPT_UGARITIC     = 53,  /* Ugar */
+
+      /** New script code in Unicode 4.0.1 @stable ICU 3.0 */
+      USCRIPT_KATAKANA_OR_HIRAGANA = 54,/*Hrkt */
+
+      /* New scripts in Unicode 4.1 */
+      /** @stable ICU 3.4 */
+      USCRIPT_BUGINESE      = 55, /* Bugi */
+      /** @stable ICU 3.4 */
+      USCRIPT_GLAGOLITIC    = 56, /* Glag */
+      /** @stable ICU 3.4 */
+      USCRIPT_KHAROSHTHI    = 57, /* Khar */
+      /** @stable ICU 3.4 */
+      USCRIPT_SYLOTI_NAGRI  = 58, /* Sylo */
+      /** @stable ICU 3.4 */
+      USCRIPT_NEW_TAI_LUE   = 59, /* Talu */
+      /** @stable ICU 3.4 */
+      USCRIPT_TIFINAGH      = 60, /* Tfng */
+      /** @stable ICU 3.4 */
+      USCRIPT_OLD_PERSIAN   = 61, /* Xpeo */
+
+      /* New script codes from Unicode and ISO 15924 */
+      /** @stable ICU 3.6 */
+      USCRIPT_BALINESE                      = 62, /* Bali */
+      /** @stable ICU 3.6 */
+      USCRIPT_BATAK                         = 63, /* Batk */
+      /** @stable ICU 3.6 */
+      USCRIPT_BLISSYMBOLS                   = 64, /* Blis */
+      /** @stable ICU 3.6 */
+      USCRIPT_BRAHMI                        = 65, /* Brah */
+      /** @stable ICU 3.6 */
+      USCRIPT_CHAM                          = 66, /* Cham */
+      /** @stable ICU 3.6 */
+      USCRIPT_CIRTH                         = 67, /* Cirt */
+      /** @stable ICU 3.6 */
+      USCRIPT_OLD_CHURCH_SLAVONIC_CYRILLIC  = 68, /* Cyrs */
+      /** @stable ICU 3.6 */
+      USCRIPT_DEMOTIC_EGYPTIAN              = 69, /* Egyd */
+      /** @stable ICU 3.6 */
+      USCRIPT_HIERATIC_EGYPTIAN             = 70, /* Egyh */
+      /** @stable ICU 3.6 */
+      USCRIPT_EGYPTIAN_HIEROGLYPHS          = 71, /* Egyp */
+      /** @stable ICU 3.6 */
+      USCRIPT_KHUTSURI                      = 72, /* Geok */
+      /** @stable ICU 3.6 */
+      USCRIPT_SIMPLIFIED_HAN                = 73, /* Hans */
+      /** @stable ICU 3.6 */
+      USCRIPT_TRADITIONAL_HAN               = 74, /* Hant */
+      /** @stable ICU 3.6 */
+      USCRIPT_PAHAWH_HMONG                  = 75, /* Hmng */
+      /** @stable ICU 3.6 */
+      USCRIPT_OLD_HUNGARIAN                 = 76, /* Hung */
+      /** @stable ICU 3.6 */
+      USCRIPT_HARAPPAN_INDUS                = 77, /* Inds */
+      /** @stable ICU 3.6 */
+      USCRIPT_JAVANESE                      = 78, /* Java */
+      /** @stable ICU 3.6 */
+      USCRIPT_KAYAH_LI                      = 79, /* Kali */
+      /** @stable ICU 3.6 */
+      USCRIPT_LATIN_FRAKTUR                 = 80, /* Latf */
+      /** @stable ICU 3.6 */
+      USCRIPT_LATIN_GAELIC                  = 81, /* Latg */
+      /** @stable ICU 3.6 */
+      USCRIPT_LEPCHA                        = 82, /* Lepc */
+      /** @stable ICU 3.6 */
+      USCRIPT_LINEAR_A                      = 83, /* Lina */
+      /** @stable ICU 4.6 */
+      USCRIPT_MANDAIC                       = 84, /* Mand */
+      /** @stable ICU 3.6 */
+      USCRIPT_MANDAEAN                      = USCRIPT_MANDAIC,
+      /** @stable ICU 3.6 */
+      USCRIPT_MAYAN_HIEROGLYPHS             = 85, /* Maya */
+      /** @stable ICU 4.6 */
+      USCRIPT_MEROITIC_HIEROGLYPHS          = 86, /* Mero */
+      /** @stable ICU 3.6 */
+      USCRIPT_MEROITIC                      = USCRIPT_MEROITIC_HIEROGLYPHS,
+      /** @stable ICU 3.6 */
+      USCRIPT_NKO                           = 87, /* Nkoo */
+      /** @stable ICU 3.6 */
+      USCRIPT_ORKHON                        = 88, /* Orkh */
+      /** @stable ICU 3.6 */
+      USCRIPT_OLD_PERMIC                    = 89, /* Perm */
+      /** @stable ICU 3.6 */
+      USCRIPT_PHAGS_PA                      = 90, /* Phag */
+      /** @stable ICU 3.6 */
+      USCRIPT_PHOENICIAN                    = 91, /* Phnx */
+      /** @stable ICU 52 */
+      USCRIPT_MIAO                          = 92, /* Plrd */
+      /** @stable ICU 3.6 */
+      USCRIPT_PHONETIC_POLLARD              = USCRIPT_MIAO,
+      /** @stable ICU 3.6 */
+      USCRIPT_RONGORONGO                    = 93, /* Roro */
+      /** @stable ICU 3.6 */
+      USCRIPT_SARATI                        = 94, /* Sara */
+      /** @stable ICU 3.6 */
+      USCRIPT_ESTRANGELO_SYRIAC             = 95, /* Syre */
+      /** @stable ICU 3.6 */
+      USCRIPT_WESTERN_SYRIAC                = 96, /* Syrj */
+      /** @stable ICU 3.6 */
+      USCRIPT_EASTERN_SYRIAC                = 97, /* Syrn */
+      /** @stable ICU 3.6 */
+      USCRIPT_TENGWAR                       = 98, /* Teng */
+      /** @stable ICU 3.6 */
+      USCRIPT_VAI                           = 99, /* Vaii */
+      /** @stable ICU 3.6 */
+      USCRIPT_VISIBLE_SPEECH                = 100,/* Visp */
+      /** @stable ICU 3.6 */
+      USCRIPT_CUNEIFORM                     = 101,/* Xsux */
+      /** @stable ICU 3.6 */
+      USCRIPT_UNWRITTEN_LANGUAGES           = 102,/* Zxxx */
+      /** @stable ICU 3.6 */
+      USCRIPT_UNKNOWN                       = 103,/* Zzzz */ /* Unknown="Code for uncoded script", for unassigned code points */
+
+      /** @stable ICU 3.8 */
+      USCRIPT_CARIAN                        = 104,/* Cari */
+      /** @stable ICU 3.8 */
+      USCRIPT_JAPANESE                      = 105,/* Jpan */
+      /** @stable ICU 3.8 */
+      USCRIPT_LANNA                         = 106,/* Lana */
+      /** @stable ICU 3.8 */
+      USCRIPT_LYCIAN                        = 107,/* Lyci */
+      /** @stable ICU 3.8 */
+      USCRIPT_LYDIAN                        = 108,/* Lydi */
+      /** @stable ICU 3.8 */
+      USCRIPT_OL_CHIKI                      = 109,/* Olck */
+      /** @stable ICU 3.8 */
+      USCRIPT_REJANG                        = 110,/* Rjng */
+      /** @stable ICU 3.8 */
+      USCRIPT_SAURASHTRA                    = 111,/* Saur */
+      /** Sutton SignWriting @stable ICU 3.8 */
+      USCRIPT_SIGN_WRITING                  = 112,/* Sgnw */
+      /** @stable ICU 3.8 */
+      USCRIPT_SUNDANESE                     = 113,/* Sund */
+      /** @stable ICU 3.8 */
+      USCRIPT_MOON                          = 114,/* Moon */
+      /** @stable ICU 3.8 */
+      USCRIPT_MEITEI_MAYEK                  = 115,/* Mtei */
+
+      /** @stable ICU 4.0 */
+      USCRIPT_IMPERIAL_ARAMAIC              = 116,/* Armi */
+      /** @stable ICU 4.0 */
+      USCRIPT_AVESTAN                       = 117,/* Avst */
+      /** @stable ICU 4.0 */
+      USCRIPT_CHAKMA                        = 118,/* Cakm */
+      /** @stable ICU 4.0 */
+      USCRIPT_KOREAN                        = 119,/* Kore */
+      /** @stable ICU 4.0 */
+      USCRIPT_KAITHI                        = 120,/* Kthi */
+      /** @stable ICU 4.0 */
+      USCRIPT_MANICHAEAN                    = 121,/* Mani */
+      /** @stable ICU 4.0 */
+      USCRIPT_INSCRIPTIONAL_PAHLAVI         = 122,/* Phli */
+      /** @stable ICU 4.0 */
+      USCRIPT_PSALTER_PAHLAVI               = 123,/* Phlp */
+      /** @stable ICU 4.0 */
+      USCRIPT_BOOK_PAHLAVI                  = 124,/* Phlv */
+      /** @stable ICU 4.0 */
+      USCRIPT_INSCRIPTIONAL_PARTHIAN        = 125,/* Prti */
+      /** @stable ICU 4.0 */
+      USCRIPT_SAMARITAN                     = 126,/* Samr */
+      /** @stable ICU 4.0 */
+      USCRIPT_TAI_VIET                      = 127,/* Tavt */
+      /** @stable ICU 4.0 */
+      USCRIPT_MATHEMATICAL_NOTATION         = 128,/* Zmth */
+      /** @stable ICU 4.0 */
+      USCRIPT_SYMBOLS                       = 129,/* Zsym */
+
+      /** @stable ICU 4.4 */
+      USCRIPT_BAMUM                         = 130,/* Bamu */
+      /** @stable ICU 4.4 */
+      USCRIPT_LISU                          = 131,/* Lisu */
+      /** @stable ICU 4.4 */
+      USCRIPT_NAKHI_GEBA                    = 132,/* Nkgb */
+      /** @stable ICU 4.4 */
+      USCRIPT_OLD_SOUTH_ARABIAN             = 133,/* Sarb */
+
+      /** @stable ICU 4.6 */
+      USCRIPT_BASSA_VAH                     = 134,/* Bass */
+      /** @stable ICU 54 */
+      USCRIPT_DUPLOYAN                      = 135,/* Dupl */
+#ifndef U_HIDE_DEPRECATED_API
+      /** @deprecated ICU 54 Typo, use USCRIPT_DUPLOYAN */
+      USCRIPT_DUPLOYAN_SHORTAND             = USCRIPT_DUPLOYAN,
+#endif  /* U_HIDE_DEPRECATED_API */
+      /** @stable ICU 4.6 */
+      USCRIPT_ELBASAN                       = 136,/* Elba */
+      /** @stable ICU 4.6 */
+      USCRIPT_GRANTHA                       = 137,/* Gran */
+      /** @stable ICU 4.6 */
+      USCRIPT_KPELLE                        = 138,/* Kpel */
+      /** @stable ICU 4.6 */
+      USCRIPT_LOMA                          = 139,/* Loma */
+      /** Mende Kikakui @stable ICU 4.6 */
+      USCRIPT_MENDE                         = 140,/* Mend */
+      /** @stable ICU 4.6 */
+      USCRIPT_MEROITIC_CURSIVE              = 141,/* Merc */
+      /** @stable ICU 4.6 */
+      USCRIPT_OLD_NORTH_ARABIAN             = 142,/* Narb */
+      /** @stable ICU 4.6 */
+      USCRIPT_NABATAEAN                     = 143,/* Nbat */
+      /** @stable ICU 4.6 */
+      USCRIPT_PALMYRENE                     = 144,/* Palm */
+      /** @stable ICU 54 */
+      USCRIPT_KHUDAWADI                     = 145,/* Sind */
+      /** @stable ICU 4.6 */
+      USCRIPT_SINDHI                        = USCRIPT_KHUDAWADI,
+      /** @stable ICU 4.6 */
+      USCRIPT_WARANG_CITI                   = 146,/* Wara */
+
+      /** @stable ICU 4.8 */
+      USCRIPT_AFAKA                         = 147,/* Afak */
+      /** @stable ICU 4.8 */
+      USCRIPT_JURCHEN                       = 148,/* Jurc */
+      /** @stable ICU 4.8 */
+      USCRIPT_MRO                           = 149,/* Mroo */
+      /** @stable ICU 4.8 */
+      USCRIPT_NUSHU                         = 150,/* Nshu */
+      /** @stable ICU 4.8 */
+      USCRIPT_SHARADA                       = 151,/* Shrd */
+      /** @stable ICU 4.8 */
+      USCRIPT_SORA_SOMPENG                  = 152,/* Sora */
+      /** @stable ICU 4.8 */
+      USCRIPT_TAKRI                         = 153,/* Takr */
+      /** @stable ICU 4.8 */
+      USCRIPT_TANGUT                        = 154,/* Tang */
+      /** @stable ICU 4.8 */
+      USCRIPT_WOLEAI                        = 155,/* Wole */
+
+      /** @stable ICU 49 */
+      USCRIPT_ANATOLIAN_HIEROGLYPHS         = 156,/* Hluw */
+      /** @stable ICU 49 */
+      USCRIPT_KHOJKI                        = 157,/* Khoj */
+      /** @stable ICU 49 */
+      USCRIPT_TIRHUTA                       = 158,/* Tirh */
+
+      /** @stable ICU 52 */
+      USCRIPT_CAUCASIAN_ALBANIAN            = 159,/* Aghb */
+      /** @stable ICU 52 */
+      USCRIPT_MAHAJANI                      = 160,/* Mahj */
+
+      /** @stable ICU 54 */
+      USCRIPT_AHOM                          = 161,/* Ahom */
+      /** @stable ICU 54 */
+      USCRIPT_HATRAN                        = 162,/* Hatr */
+      /** @stable ICU 54 */
+      USCRIPT_MODI                          = 163,/* Modi */
+      /** @stable ICU 54 */
+      USCRIPT_MULTANI                       = 164,/* Mult */
+      /** @stable ICU 54 */
+      USCRIPT_PAU_CIN_HAU                   = 165,/* Pauc */
+      /** @stable ICU 54 */
+      USCRIPT_SIDDHAM                       = 166,/* Sidd */
+
+      /** @stable ICU 58 */
+      USCRIPT_ADLAM                         = 167,/* Adlm */
+      /** @stable ICU 58 */
+      USCRIPT_BHAIKSUKI                     = 168,/* Bhks */
+      /** @stable ICU 58 */
+      USCRIPT_MARCHEN                       = 169,/* Marc */
+      /** @stable ICU 58 */
+      USCRIPT_NEWA                          = 170,/* Newa */
+      /** @stable ICU 58 */
+      USCRIPT_OSAGE                         = 171,/* Osge */
+
+      /** @stable ICU 58 */
+      USCRIPT_HAN_WITH_BOPOMOFO             = 172,/* Hanb */
+      /** @stable ICU 58 */
+      USCRIPT_JAMO                          = 173,/* Jamo */
+      /** @stable ICU 58 */
+      USCRIPT_SYMBOLS_EMOJI                 = 174,/* Zsye */
+
+      /** @stable ICU 60 */
+      USCRIPT_MASARAM_GONDI                 = 175,/* Gonm */
+      /** @stable ICU 60 */
+      USCRIPT_SOYOMBO                       = 176,/* Soyo */
+      /** @stable ICU 60 */
+      USCRIPT_ZANABAZAR_SQUARE              = 177,/* Zanb */
+
+      /** @stable ICU 62 */
+      USCRIPT_DOGRA                         = 178,/* Dogr */
+      /** @stable ICU 62 */
+      USCRIPT_GUNJALA_GONDI                 = 179,/* Gong */
+      /** @stable ICU 62 */
+      USCRIPT_MAKASAR                       = 180,/* Maka */
+      /** @stable ICU 62 */
+      USCRIPT_MEDEFAIDRIN                   = 181,/* Medf */
+      /** @stable ICU 62 */
+      USCRIPT_HANIFI_ROHINGYA               = 182,/* Rohg */
+      /** @stable ICU 62 */
+      USCRIPT_SOGDIAN                       = 183,/* Sogd */
+      /** @stable ICU 62 */
+      USCRIPT_OLD_SOGDIAN                   = 184,/* Sogo */
+
+      /** @stable ICU 64 */
+      USCRIPT_ELYMAIC                       = 185,/* Elym */
+      /** @stable ICU 64 */
+      USCRIPT_NYIAKENG_PUACHUE_HMONG        = 186,/* Hmnp */
+      /** @stable ICU 64 */
+      USCRIPT_NANDINAGARI                   = 187,/* Nand */
+      /** @stable ICU 64 */
+      USCRIPT_WANCHO                        = 188,/* Wcho */
+
+      /** @stable ICU 66 */
+      USCRIPT_CHORASMIAN                    = 189,/* Chrs */
+      /** @stable ICU 66 */
+      USCRIPT_DIVES_AKURU                   = 190,/* Diak */
+      /** @stable ICU 66 */
+      USCRIPT_KHITAN_SMALL_SCRIPT           = 191,/* Kits */
+      /** @stable ICU 66 */
+      USCRIPT_YEZIDI                        = 192,/* Yezi */
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UScriptCode value.
+     * The highest value is available via u_getIntPropertyMaxValue(UCHAR_SCRIPT).
+     *
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    USCRIPT_CODE_LIMIT    = 193
+#endif  // U_HIDE_DEPRECATED_API
+} UScriptCode;
+
+/**
+ * Gets the script codes associated with the given locale or ISO 15924 abbreviation or name.
+ * Fills in USCRIPT_MALAYALAM given "Malayam" OR "Mlym".
+ * Fills in USCRIPT_LATIN given "en" OR "en_US"
+ * If the required capacity is greater than the capacity of the destination buffer,
+ * then the error code is set to U_BUFFER_OVERFLOW_ERROR and the required capacity is returned.
+ *
+ * <p>Note: To search by short or long script alias only, use
+ * u_getPropertyValueEnum(UCHAR_SCRIPT, alias) instead.  That does
+ * a fast lookup with no access of the locale data.
+ *
+ * @param nameOrAbbrOrLocale name of the script, as given in
+ * PropertyValueAliases.txt, or ISO 15924 code or locale
+ * @param fillIn the UScriptCode buffer to fill in the script code
+ * @param capacity the capacity (size) of UScriptCode buffer passed in.
+ * @param err the error status code.
+ * @return The number of script codes filled in the buffer passed in
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t  U_EXPORT2
+uscript_getCode(const char* nameOrAbbrOrLocale,UScriptCode* fillIn,int32_t capacity,UErrorCode *err);
+
+/**
+ * Returns the long Unicode script name, if there is one.
+ * Otherwise returns the 4-letter ISO 15924 script code.
+ * Returns "Malayam" given USCRIPT_MALAYALAM.
+ *
+ * @param scriptCode UScriptCode enum
+ * @return long script name as given in PropertyValueAliases.txt, or the 4-letter code,
+ * or NULL if scriptCode is invalid
+ * @stable ICU 2.4
+ */
+U_CAPI const char*  U_EXPORT2
+uscript_getName(UScriptCode scriptCode);
+
+/**
+ * Returns the 4-letter ISO 15924 script code,
+ * which is the same as the short Unicode script name if Unicode has names for the script.
+ * Returns "Mlym" given USCRIPT_MALAYALAM.
+ *
+ * @param scriptCode UScriptCode enum
+ * @return short script name (4-letter code), or NULL if scriptCode is invalid
+ * @stable ICU 2.4
+ */
+U_CAPI const char*  U_EXPORT2
+uscript_getShortName(UScriptCode scriptCode);
+
+/**
+ * Gets the script code associated with the given codepoint.
+ * Returns USCRIPT_MALAYALAM given 0x0D02
+ * @param codepoint UChar32 codepoint
+ * @param err the error status code.
+ * @return The UScriptCode, or 0 if codepoint is invalid
+ * @stable ICU 2.4
+ */
+U_CAPI UScriptCode  U_EXPORT2
+uscript_getScript(UChar32 codepoint, UErrorCode *err);
+
+/**
+ * Do the Script_Extensions of code point c contain script sc?
+ * If c does not have explicit Script_Extensions, then this tests whether
+ * c has the Script property value sc.
+ *
+ * Some characters are commonly used in multiple scripts.
+ * For more information, see UAX #24: http://www.unicode.org/reports/tr24/.
+ * @param c code point
+ * @param sc script code
+ * @return true if sc is in Script_Extensions(c)
+ * @stable ICU 49
+ */
+U_CAPI UBool U_EXPORT2
+uscript_hasScript(UChar32 c, UScriptCode sc);
+
+/**
+ * Writes code point c's Script_Extensions as a list of UScriptCode values
+ * to the output scripts array and returns the number of script codes.
+ * - If c does have Script_Extensions, then the Script property value
+ *   (normally Common or Inherited) is not included.
+ * - If c does not have Script_Extensions, then the one Script code is written to the output array.
+ * - If c is not a valid code point, then the one USCRIPT_UNKNOWN code is written.
+ * In other words, if the return value is 1,
+ * then the output array contains exactly c's single Script code.
+ * If the return value is n>=2, then the output array contains c's n Script_Extensions script codes.
+ *
+ * Some characters are commonly used in multiple scripts.
+ * For more information, see UAX #24: http://www.unicode.org/reports/tr24/.
+ *
+ * If there are more than capacity script codes to be written, then
+ * U_BUFFER_OVERFLOW_ERROR is set and the number of Script_Extensions is returned.
+ * (Usual ICU buffer handling behavior.)
+ *
+ * @param c code point
+ * @param scripts output script code array
+ * @param capacity capacity of the scripts array
+ * @param errorCode Standard ICU error code. Its input value must
+ *                  pass the U_SUCCESS() test, or else the function returns
+ *                  immediately. Check for U_FAILURE() on output or use with
+ *                  function chaining. (See User Guide for details.)
+ * @return number of script codes in c's Script_Extensions, or 1 for the single Script value,
+ *         written to scripts unless U_BUFFER_OVERFLOW_ERROR indicates insufficient capacity
+ * @stable ICU 49
+ */
+U_CAPI int32_t U_EXPORT2
+uscript_getScriptExtensions(UChar32 c,
+                            UScriptCode *scripts, int32_t capacity,
+                            UErrorCode *errorCode);
+
+/**
+ * Script usage constants.
+ * See UAX #31 Unicode Identifier and Pattern Syntax.
+ * http://www.unicode.org/reports/tr31/#Table_Candidate_Characters_for_Exclusion_from_Identifiers
+ *
+ * @stable ICU 51
+ */
+typedef enum UScriptUsage {
+    /** Not encoded in Unicode. @stable ICU 51 */
+    USCRIPT_USAGE_NOT_ENCODED,
+    /** Unknown script usage. @stable ICU 51 */
+    USCRIPT_USAGE_UNKNOWN,
+    /** Candidate for Exclusion from Identifiers. @stable ICU 51 */
+    USCRIPT_USAGE_EXCLUDED,
+    /** Limited Use script. @stable ICU 51 */
+    USCRIPT_USAGE_LIMITED_USE,
+    /** Aspirational Use script. @stable ICU 51 */
+    USCRIPT_USAGE_ASPIRATIONAL,
+    /** Recommended script. @stable ICU 51 */
+    USCRIPT_USAGE_RECOMMENDED
+} UScriptUsage;
+
+/**
+ * Writes the script sample character string.
+ * This string normally consists of one code point but might be longer.
+ * The string is empty if the script is not encoded.
+ *
+ * @param script script code
+ * @param dest output string array
+ * @param capacity number of UChars in the dest array
+ * @param pErrorCode standard ICU in/out error code, must pass U_SUCCESS() on input
+ * @return the string length, even if U_BUFFER_OVERFLOW_ERROR
+ * @stable ICU 51
+ */
+U_CAPI int32_t U_EXPORT2
+uscript_getSampleString(UScriptCode script, UChar *dest, int32_t capacity, UErrorCode *pErrorCode);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+class UnicodeString;
+U_NAMESPACE_END
+
+/**
+ * Returns the script sample character string.
+ * This string normally consists of one code point but might be longer.
+ * The string is empty if the script is not encoded.
+ *
+ * @param script script code
+ * @return the sample character string
+ * @stable ICU 51
+ */
+U_COMMON_API icu::UnicodeString U_EXPORT2
+uscript_getSampleUnicodeString(UScriptCode script);
+
+#endif
+
+/**
+ * Returns the script usage according to UAX #31 Unicode Identifier and Pattern Syntax.
+ * Returns USCRIPT_USAGE_NOT_ENCODED if the script is not encoded in Unicode.
+ *
+ * @param script script code
+ * @return script usage
+ * @see UScriptUsage
+ * @stable ICU 51
+ */
+U_CAPI UScriptUsage U_EXPORT2
+uscript_getUsage(UScriptCode script);
+
+/**
+ * Returns true if the script is written right-to-left.
+ * For example, Arab and Hebr.
+ *
+ * @param script script code
+ * @return true if the script is right-to-left
+ * @stable ICU 51
+ */
+U_CAPI UBool U_EXPORT2
+uscript_isRightToLeft(UScriptCode script);
+
+/**
+ * Returns true if the script allows line breaks between letters (excluding hyphenation).
+ * Such a script typically requires dictionary-based line breaking.
+ * For example, Hani and Thai.
+ *
+ * @param script script code
+ * @return true if the script allows line breaks between letters
+ * @stable ICU 51
+ */
+U_CAPI UBool U_EXPORT2
+uscript_breaksBetweenLetters(UScriptCode script);
+
+/**
+ * Returns true if in modern (or most recent) usage of the script case distinctions are customary.
+ * For example, Latn and Cyrl.
+ *
+ * @param script script code
+ * @return true if the script is cased
+ * @stable ICU 51
+ */
+U_CAPI UBool U_EXPORT2
+uscript_isCased(UScriptCode script);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uset.h b/thirdparty/icu4c/common/unicode/uset.h
new file mode 100644
index 0000000000..502ea8dc14
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uset.h
@@ -0,0 +1,1137 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uset.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002mar07
+*   created by: Markus W. Scherer
+*
+*   C version of UnicodeSet.
+*/
+
+
+/**
+ * \file
+ * \brief C API: Unicode Set
+ *
+ * <p>This is a C wrapper around the C++ UnicodeSet class.</p>
+ */
+
+#ifndef __USET_H__
+#define __USET_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+#ifndef USET_DEFINED
+
+#ifndef U_IN_DOXYGEN
+#define USET_DEFINED
+#endif
+/**
+ * USet is the C API type corresponding to C++ class UnicodeSet.
+ * Use the uset_* API to manipulate.  Create with
+ * uset_open*, and destroy with uset_close.
+ * @stable ICU 2.4
+ */
+typedef struct USet USet;
+#endif
+
+/**
+ * Bitmask values to be passed to uset_openPatternOptions() or
+ * uset_applyPattern() taking an option parameter.
+ * @stable ICU 2.4
+ */
+enum {
+    /**
+     * Ignore white space within patterns unless quoted or escaped.
+     * @stable ICU 2.4
+     */
+    USET_IGNORE_SPACE = 1,  
+
+    /**
+     * Enable case insensitive matching.  E.g., "[ab]" with this flag
+     * will match 'a', 'A', 'b', and 'B'.  "[^ab]" with this flag will
+     * match all except 'a', 'A', 'b', and 'B'. This performs a full
+     * closure over case mappings, e.g. U+017F for s.
+     *
+     * The resulting set is a superset of the input for the code points but
+     * not for the strings.
+     * It performs a case mapping closure of the code points and adds
+     * full case folding strings for the code points, and reduces strings of
+     * the original set to their full case folding equivalents.
+     *
+     * This is designed for case-insensitive matches, for example
+     * in regular expressions. The full code point case closure allows checking of
+     * an input character directly against the closure set.
+     * Strings are matched by comparing the case-folded form from the closure
+     * set with an incremental case folding of the string in question.
+     *
+     * The closure set will also contain single code points if the original
+     * set contained case-equivalent strings (like U+00DF for "ss" or "Ss" etc.).
+     * This is not necessary (that is, redundant) for the above matching method
+     * but results in the same closure sets regardless of whether the original
+     * set contained the code point or a string.
+     *
+     * @stable ICU 2.4
+     */
+    USET_CASE_INSENSITIVE = 2,  
+
+    /**
+     * Enable case insensitive matching.  E.g., "[ab]" with this flag
+     * will match 'a', 'A', 'b', and 'B'.  "[^ab]" with this flag will
+     * match all except 'a', 'A', 'b', and 'B'. This adds the lower-,
+     * title-, and uppercase mappings as well as the case folding
+     * of each existing element in the set.
+     * @stable ICU 3.2
+     */
+    USET_ADD_CASE_MAPPINGS = 4
+};
+
+/**
+ * Argument values for whether span() and similar functions continue while
+ * the current character is contained vs. not contained in the set.
+ *
+ * The functionality is straightforward for sets with only single code points,
+ * without strings (which is the common case):
+ * - USET_SPAN_CONTAINED and USET_SPAN_SIMPLE work the same.
+ * - USET_SPAN_CONTAINED and USET_SPAN_SIMPLE are inverses of USET_SPAN_NOT_CONTAINED.
+ * - span() and spanBack() partition any string the same way when
+ *   alternating between span(USET_SPAN_NOT_CONTAINED) and
+ *   span(either "contained" condition).
+ * - Using a complemented (inverted) set and the opposite span conditions
+ *   yields the same results.
+ *
+ * When a set contains multi-code point strings, then these statements may not
+ * be true, depending on the strings in the set (for example, whether they
+ * overlap with each other) and the string that is processed.
+ * For a set with strings:
+ * - The complement of the set contains the opposite set of code points,
+ *   but the same set of strings.
+ *   Therefore, complementing both the set and the span conditions
+ *   may yield different results.
+ * - When starting spans at different positions in a string
+ *   (span(s, ...) vs. span(s+1, ...)) the ends of the spans may be different
+ *   because a set string may start before the later position.
+ * - span(USET_SPAN_SIMPLE) may be shorter than
+ *   span(USET_SPAN_CONTAINED) because it will not recursively try
+ *   all possible paths.
+ *   For example, with a set which contains the three strings "xy", "xya" and "ax",
+ *   span("xyax", USET_SPAN_CONTAINED) will return 4 but
+ *   span("xyax", USET_SPAN_SIMPLE) will return 3.
+ *   span(USET_SPAN_SIMPLE) will never be longer than
+ *   span(USET_SPAN_CONTAINED).
+ * - With either "contained" condition, span() and spanBack() may partition
+ *   a string in different ways.
+ *   For example, with a set which contains the two strings "ab" and "ba",
+ *   and when processing the string "aba",
+ *   span() will yield contained/not-contained boundaries of { 0, 2, 3 }
+ *   while spanBack() will yield boundaries of { 0, 1, 3 }.
+ *
+ * Note: If it is important to get the same boundaries whether iterating forward
+ * or backward through a string, then either only span() should be used and
+ * the boundaries cached for backward operation, or an ICU BreakIterator
+ * could be used.
+ *
+ * Note: Unpaired surrogates are treated like surrogate code points.
+ * Similarly, set strings match only on code point boundaries,
+ * never in the middle of a surrogate pair.
+ * Illegal UTF-8 sequences are treated like U+FFFD.
+ * When processing UTF-8 strings, malformed set strings
+ * (strings with unpaired surrogates which cannot be converted to UTF-8)
+ * are ignored.
+ *
+ * @stable ICU 3.8
+ */
+typedef enum USetSpanCondition {
+    /**
+     * Continues a span() while there is no set element at the current position.
+     * Increments by one code point at a time.
+     * Stops before the first set element (character or string).
+     * (For code points only, this is like while contains(current)==false).
+     *
+     * When span() returns, the substring between where it started and the position
+     * it returned consists only of characters that are not in the set,
+     * and none of its strings overlap with the span.
+     *
+     * @stable ICU 3.8
+     */
+    USET_SPAN_NOT_CONTAINED = 0,
+    /**
+     * Spans the longest substring that is a concatenation of set elements (characters or strings).
+     * (For characters only, this is like while contains(current)==true).
+     *
+     * When span() returns, the substring between where it started and the position
+     * it returned consists only of set elements (characters or strings) that are in the set.
+     *
+     * If a set contains strings, then the span will be the longest substring for which there
+     * exists at least one non-overlapping concatenation of set elements (characters or strings).
+     * This is equivalent to a POSIX regular expression for <code>(OR of each set element)*</code>.
+     * (Java/ICU/Perl regex stops at the first match of an OR.)
+     *
+     * @stable ICU 3.8
+     */
+    USET_SPAN_CONTAINED = 1,
+    /**
+     * Continues a span() while there is a set element at the current position.
+     * Increments by the longest matching element at each position.
+     * (For characters only, this is like while contains(current)==true).
+     *
+     * When span() returns, the substring between where it started and the position
+     * it returned consists only of set elements (characters or strings) that are in the set.
+     *
+     * If a set only contains single characters, then this is the same
+     * as USET_SPAN_CONTAINED.
+     *
+     * If a set contains strings, then the span will be the longest substring
+     * with a match at each position with the longest single set element (character or string).
+     *
+     * Use this span condition together with other longest-match algorithms,
+     * such as ICU converters (ucnv_getUnicodeSet()).
+     *
+     * @stable ICU 3.8
+     */
+    USET_SPAN_SIMPLE = 2,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the last span condition.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    USET_SPAN_CONDITION_COUNT
+#endif  // U_HIDE_DEPRECATED_API
+} USetSpanCondition;
+
+enum {
+    /**
+     * Capacity of USerializedSet::staticArray.
+     * Enough for any single-code point set.
+     * Also provides padding for nice sizeof(USerializedSet).
+     * @stable ICU 2.4
+     */
+    USET_SERIALIZED_STATIC_ARRAY_CAPACITY=8
+};
+
+/**
+ * A serialized form of a Unicode set.  Limited manipulations are
+ * possible directly on a serialized set.  See below.
+ * @stable ICU 2.4
+ */
+typedef struct USerializedSet {
+    /**
+     * The serialized Unicode Set.
+     * @stable ICU 2.4
+     */
+    const uint16_t *array;
+    /**
+     * The length of the array that contains BMP characters.
+     * @stable ICU 2.4
+     */
+    int32_t bmpLength;
+    /**
+     * The total length of the array.
+     * @stable ICU 2.4
+     */
+    int32_t length;
+    /**
+     * A small buffer for the array to reduce memory allocations.
+     * @stable ICU 2.4
+     */
+    uint16_t staticArray[USET_SERIALIZED_STATIC_ARRAY_CAPACITY];
+} USerializedSet;
+
+/*********************************************************************
+ * USet API
+ *********************************************************************/
+
+/**
+ * Create an empty USet object.
+ * Equivalent to uset_open(1, 0).
+ * @return a newly created USet.  The caller must call uset_close() on
+ * it when done.
+ * @stable ICU 4.2
+ */
+U_CAPI USet* U_EXPORT2
+uset_openEmpty(void);
+
+/**
+ * Creates a USet object that contains the range of characters
+ * start..end, inclusive.  If <code>start > end</code> 
+ * then an empty set is created (same as using uset_openEmpty()).
+ * @param start first character of the range, inclusive
+ * @param end last character of the range, inclusive
+ * @return a newly created USet.  The caller must call uset_close() on
+ * it when done.
+ * @stable ICU 2.4
+ */
+U_CAPI USet* U_EXPORT2
+uset_open(UChar32 start, UChar32 end);
+
+/**
+ * Creates a set from the given pattern.  See the UnicodeSet class
+ * description for the syntax of the pattern language.
+ * @param pattern a string specifying what characters are in the set
+ * @param patternLength the length of the pattern, or -1 if null
+ * terminated
+ * @param ec the error code
+ * @stable ICU 2.4
+ */
+U_CAPI USet* U_EXPORT2
+uset_openPattern(const UChar* pattern, int32_t patternLength,
+                 UErrorCode* ec);
+
+/**
+ * Creates a set from the given pattern.  See the UnicodeSet class
+ * description for the syntax of the pattern language.
+ * @param pattern a string specifying what characters are in the set
+ * @param patternLength the length of the pattern, or -1 if null
+ * terminated
+ * @param options bitmask for options to apply to the pattern.
+ * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+ * @param ec the error code
+ * @stable ICU 2.4
+ */
+U_CAPI USet* U_EXPORT2
+uset_openPatternOptions(const UChar* pattern, int32_t patternLength,
+                 uint32_t options,
+                 UErrorCode* ec);
+
+/**
+ * Disposes of the storage used by a USet object.  This function should
+ * be called exactly once for objects returned by uset_open().
+ * @param set the object to dispose of
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_close(USet* set);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUSetPointer
+ * "Smart pointer" class, closes a USet via uset_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUSetPointer, USet, uset_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Returns a copy of this object.
+ * If this set is frozen, then the clone will be frozen as well.
+ * Use uset_cloneAsThawed() for a mutable clone of a frozen set.
+ * @param set the original set
+ * @return the newly allocated copy of the set
+ * @see uset_cloneAsThawed
+ * @stable ICU 3.8
+ */
+U_CAPI USet * U_EXPORT2
+uset_clone(const USet *set);
+
+/**
+ * Determines whether the set has been frozen (made immutable) or not.
+ * See the ICU4J Freezable interface for details.
+ * @param set the set
+ * @return true/false for whether the set has been frozen
+ * @see uset_freeze
+ * @see uset_cloneAsThawed
+ * @stable ICU 3.8
+ */
+U_CAPI UBool U_EXPORT2
+uset_isFrozen(const USet *set);
+
+/**
+ * Freeze the set (make it immutable).
+ * Once frozen, it cannot be unfrozen and is therefore thread-safe
+ * until it is deleted.
+ * See the ICU4J Freezable interface for details.
+ * Freezing the set may also make some operations faster, for example
+ * uset_contains() and uset_span().
+ * A frozen set will not be modified. (It remains frozen.)
+ * @param set the set
+ * @return the same set, now frozen
+ * @see uset_isFrozen
+ * @see uset_cloneAsThawed
+ * @stable ICU 3.8
+ */
+U_CAPI void U_EXPORT2
+uset_freeze(USet *set);
+
+/**
+ * Clone the set and make the clone mutable.
+ * See the ICU4J Freezable interface for details.
+ * @param set the set
+ * @return the mutable clone
+ * @see uset_freeze
+ * @see uset_isFrozen
+ * @see uset_clone
+ * @stable ICU 3.8
+ */
+U_CAPI USet * U_EXPORT2
+uset_cloneAsThawed(const USet *set);
+
+/**
+ * Causes the USet object to represent the range <code>start - end</code>.
+ * If <code>start > end</code> then this USet is set to an empty range.
+ * A frozen set will not be modified.
+ * @param set the object to set to the given range
+ * @param start first character in the set, inclusive
+ * @param end last character in the set, inclusive
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_set(USet* set,
+         UChar32 start, UChar32 end);
+
+/**
+ * Modifies the set to represent the set specified by the given
+ * pattern. See the UnicodeSet class description for the syntax of 
+ * the pattern language. See also the User Guide chapter about UnicodeSet.
+ * <em>Empties the set passed before applying the pattern.</em>
+ * A frozen set will not be modified.
+ * @param set               The set to which the pattern is to be applied. 
+ * @param pattern           A pointer to UChar string specifying what characters are in the set.
+ *                          The character at pattern[0] must be a '['.
+ * @param patternLength     The length of the UChar string. -1 if NUL terminated.
+ * @param options           A bitmask for options to apply to the pattern.
+ *                          Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
+ * @param status            Returns an error if the pattern cannot be parsed.
+ * @return                  Upon successful parse, the value is either
+ *                          the index of the character after the closing ']' 
+ *                          of the parsed pattern.
+ *                          If the status code indicates failure, then the return value 
+ *                          is the index of the error in the source.
+ *
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2 
+uset_applyPattern(USet *set,
+                  const UChar *pattern, int32_t patternLength,
+                  uint32_t options,
+                  UErrorCode *status);
+
+/**
+ * Modifies the set to contain those code points which have the given value
+ * for the given binary or enumerated property, as returned by
+ * u_getIntPropertyValue.  Prior contents of this set are lost.
+ * A frozen set will not be modified.
+ *
+ * @param set the object to contain the code points defined by the property
+ *
+ * @param prop a property in the range UCHAR_BIN_START..UCHAR_BIN_LIMIT-1
+ * or UCHAR_INT_START..UCHAR_INT_LIMIT-1
+ * or UCHAR_MASK_START..UCHAR_MASK_LIMIT-1.
+ *
+ * @param value a value in the range u_getIntPropertyMinValue(prop)..
+ * u_getIntPropertyMaxValue(prop), with one exception.  If prop is
+ * UCHAR_GENERAL_CATEGORY_MASK, then value should not be a UCharCategory, but
+ * rather a mask value produced by U_GET_GC_MASK().  This allows grouped
+ * categories such as [:L:] to be represented.
+ *
+ * @param ec error code input/output parameter
+ *
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_applyIntPropertyValue(USet* set,
+                           UProperty prop, int32_t value, UErrorCode* ec);
+
+/**
+ * Modifies the set to contain those code points which have the
+ * given value for the given property.  Prior contents of this
+ * set are lost.
+ * A frozen set will not be modified.
+ *
+ * @param set the object to contain the code points defined by the given
+ * property and value alias
+ *
+ * @param prop a string specifying a property alias, either short or long.
+ * The name is matched loosely.  See PropertyAliases.txt for names and a
+ * description of loose matching.  If the value string is empty, then this
+ * string is interpreted as either a General_Category value alias, a Script
+ * value alias, a binary property alias, or a special ID.  Special IDs are
+ * matched loosely and correspond to the following sets:
+ *
+ * "ANY" = [\\u0000-\\U0010FFFF],
+ * "ASCII" = [\\u0000-\\u007F],
+ * "Assigned" = [:^Cn:].
+ *
+ * @param propLength the length of the prop, or -1 if NULL
+ *
+ * @param value a string specifying a value alias, either short or long.
+ * The name is matched loosely.  See PropertyValueAliases.txt for names
+ * and a description of loose matching.  In addition to aliases listed,
+ * numeric values and canonical combining classes may be expressed
+ * numerically, e.g., ("nv", "0.5") or ("ccc", "220").  The value string
+ * may also be empty.
+ *
+ * @param valueLength the length of the value, or -1 if NULL
+ *
+ * @param ec error code input/output parameter
+ *
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_applyPropertyAlias(USet* set,
+                        const UChar *prop, int32_t propLength,
+                        const UChar *value, int32_t valueLength,
+                        UErrorCode* ec);
+
+/**
+ * Return true if the given position, in the given pattern, appears
+ * to be the start of a UnicodeSet pattern.
+ *
+ * @param pattern a string specifying the pattern
+ * @param patternLength the length of the pattern, or -1 if NULL
+ * @param pos the given position
+ * @stable ICU 3.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_resemblesPattern(const UChar *pattern, int32_t patternLength,
+                      int32_t pos);
+
+/**
+ * Returns a string representation of this set.  If the result of
+ * calling this function is passed to a uset_openPattern(), it
+ * will produce another set that is equal to this one.
+ * @param set the set
+ * @param result the string to receive the rules, may be NULL
+ * @param resultCapacity the capacity of result, may be 0 if result is NULL
+ * @param escapeUnprintable if true then convert unprintable
+ * character to their hex escape representations, \\uxxxx or
+ * \\Uxxxxxxxx.  Unprintable characters are those other than
+ * U+000A, U+0020..U+007E.
+ * @param ec error code.
+ * @return length of string, possibly larger than resultCapacity
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_toPattern(const USet* set,
+               UChar* result, int32_t resultCapacity,
+               UBool escapeUnprintable,
+               UErrorCode* ec);
+
+/**
+ * Adds the given character to the given USet.  After this call,
+ * uset_contains(set, c) will return true.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param c the character to add
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_add(USet* set, UChar32 c);
+
+/**
+ * Adds all of the elements in the specified set to this set if
+ * they're not already present.  This operation effectively
+ * modifies this set so that its value is the <i>union</i> of the two
+ * sets.  The behavior of this operation is unspecified if the specified
+ * collection is modified while the operation is in progress.
+ * A frozen set will not be modified.
+ *
+ * @param set the object to which to add the set
+ * @param additionalSet the source set whose elements are to be added to this set.
+ * @stable ICU 2.6
+ */
+U_CAPI void U_EXPORT2
+uset_addAll(USet* set, const USet *additionalSet);
+
+/**
+ * Adds the given range of characters to the given USet.  After this call,
+ * uset_contains(set, start, end) will return true.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param start the first character of the range to add, inclusive
+ * @param end the last character of the range to add, inclusive
+ * @stable ICU 2.2
+ */
+U_CAPI void U_EXPORT2
+uset_addRange(USet* set, UChar32 start, UChar32 end);
+
+/**
+ * Adds the given string to the given USet.  After this call,
+ * uset_containsString(set, str, strLen) will return true.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param str the string to add
+ * @param strLen the length of the string or -1 if null terminated.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_addString(USet* set, const UChar* str, int32_t strLen);
+
+/**
+ * Adds each of the characters in this string to the set. Thus "ch" => {"c", "h"}
+ * If this set already any particular character, it has no effect on that character.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param str the source string
+ * @param strLen the length of the string or -1 if null terminated.
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+uset_addAllCodePoints(USet* set, const UChar *str, int32_t strLen);
+
+/**
+ * Removes the given character from the given USet.  After this call,
+ * uset_contains(set, c) will return false.
+ * A frozen set will not be modified.
+ * @param set the object from which to remove the character
+ * @param c the character to remove
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_remove(USet* set, UChar32 c);
+
+/**
+ * Removes the given range of characters from the given USet.  After this call,
+ * uset_contains(set, start, end) will return false.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param start the first character of the range to remove, inclusive
+ * @param end the last character of the range to remove, inclusive
+ * @stable ICU 2.2
+ */
+U_CAPI void U_EXPORT2
+uset_removeRange(USet* set, UChar32 start, UChar32 end);
+
+/**
+ * Removes the given string to the given USet.  After this call,
+ * uset_containsString(set, str, strLen) will return false.
+ * A frozen set will not be modified.
+ * @param set the object to which to add the character
+ * @param str the string to remove
+ * @param strLen the length of the string or -1 if null terminated.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_removeString(USet* set, const UChar* str, int32_t strLen);
+
+/**
+ * Removes from this set all of its elements that are contained in the
+ * specified set.  This operation effectively modifies this
+ * set so that its value is the <i>asymmetric set difference</i> of
+ * the two sets.
+ * A frozen set will not be modified.
+ * @param set the object from which the elements are to be removed
+ * @param removeSet the object that defines which elements will be
+ * removed from this set
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_removeAll(USet* set, const USet* removeSet);
+
+/**
+ * Retain only the elements in this set that are contained in the
+ * specified range.  If <code>start > end</code> then an empty range is
+ * retained, leaving the set empty.  This is equivalent to
+ * a boolean logic AND, or a set INTERSECTION.
+ * A frozen set will not be modified.
+ *
+ * @param set the object for which to retain only the specified range
+ * @param start first character, inclusive, of range to be retained
+ * to this set.
+ * @param end last character, inclusive, of range to be retained
+ * to this set.
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_retain(USet* set, UChar32 start, UChar32 end);
+
+/**
+ * Retains only the elements in this set that are contained in the
+ * specified set.  In other words, removes from this set all of
+ * its elements that are not contained in the specified set.  This
+ * operation effectively modifies this set so that its value is
+ * the <i>intersection</i> of the two sets.
+ * A frozen set will not be modified.
+ *
+ * @param set the object on which to perform the retain
+ * @param retain set that defines which elements this set will retain
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_retainAll(USet* set, const USet* retain);
+
+/**
+ * Reallocate this objects internal structures to take up the least
+ * possible space, without changing this object's value.
+ * A frozen set will not be modified.
+ *
+ * @param set the object on which to perfrom the compact
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_compact(USet* set);
+
+/**
+ * Inverts this set.  This operation modifies this set so that
+ * its value is its complement.  This operation does not affect
+ * the multicharacter strings, if any.
+ * A frozen set will not be modified.
+ * @param set the set
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_complement(USet* set);
+
+/**
+ * Complements in this set all elements contained in the specified
+ * set.  Any character in the other set will be removed if it is
+ * in this set, or will be added if it is not in this set.
+ * A frozen set will not be modified.
+ *
+ * @param set the set with which to complement
+ * @param complement set that defines which elements will be xor'ed
+ * from this set.
+ * @stable ICU 3.2
+ */
+U_CAPI void U_EXPORT2
+uset_complementAll(USet* set, const USet* complement);
+
+/**
+ * Removes all of the elements from this set.  This set will be
+ * empty after this call returns.
+ * A frozen set will not be modified.
+ * @param set the set
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_clear(USet* set);
+
+/**
+ * Close this set over the given attribute.  For the attribute
+ * USET_CASE, the result is to modify this set so that:
+ *
+ * 1. For each character or string 'a' in this set, all strings or
+ * characters 'b' such that foldCase(a) == foldCase(b) are added
+ * to this set.
+ *
+ * 2. For each string 'e' in the resulting set, if e !=
+ * foldCase(e), 'e' will be removed.
+ *
+ * Example: [aq\\u00DF{Bc}{bC}{Fi}] => [aAqQ\\u00DF\\uFB01{ss}{bc}{fi}]
+ *
+ * (Here foldCase(x) refers to the operation u_strFoldCase, and a
+ * == b denotes that the contents are the same, not pointer
+ * comparison.)
+ *
+ * A frozen set will not be modified.
+ *
+ * @param set the set
+ *
+ * @param attributes bitmask for attributes to close over.
+ * Currently only the USET_CASE bit is supported.  Any undefined bits
+ * are ignored.
+ * @stable ICU 4.2
+ */
+U_CAPI void U_EXPORT2
+uset_closeOver(USet* set, int32_t attributes);
+
+/**
+ * Remove all strings from this set.
+ *
+ * @param set the set
+ * @stable ICU 4.2
+ */
+U_CAPI void U_EXPORT2
+uset_removeAllStrings(USet* set);
+
+/**
+ * Returns true if the given USet contains no characters and no
+ * strings.
+ * @param set the set
+ * @return true if set is empty
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_isEmpty(const USet* set);
+
+/**
+ * Returns true if the given USet contains the given character.
+ * This function works faster with a frozen set.
+ * @param set the set
+ * @param c The codepoint to check for within the set
+ * @return true if set contains c
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_contains(const USet* set, UChar32 c);
+
+/**
+ * Returns true if the given USet contains all characters c
+ * where start <= c && c <= end.
+ * @param set the set
+ * @param start the first character of the range to test, inclusive
+ * @param end the last character of the range to test, inclusive
+ * @return true if set contains the range
+ * @stable ICU 2.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsRange(const USet* set, UChar32 start, UChar32 end);
+
+/**
+ * Returns true if the given USet contains the given string.
+ * @param set the set
+ * @param str the string
+ * @param strLen the length of the string or -1 if null terminated.
+ * @return true if set contains str
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsString(const USet* set, const UChar* str, int32_t strLen);
+
+/**
+ * Returns the index of the given character within this set, where
+ * the set is ordered by ascending code point.  If the character
+ * is not in this set, return -1.  The inverse of this method is
+ * <code>charAt()</code>.
+ * @param set the set
+ * @param c the character to obtain the index for
+ * @return an index from 0..size()-1, or -1
+ * @stable ICU 3.2
+ */
+U_CAPI int32_t U_EXPORT2
+uset_indexOf(const USet* set, UChar32 c);
+
+/**
+ * Returns the character at the given index within this set, where
+ * the set is ordered by ascending code point.  If the index is
+ * out of range, return (UChar32)-1.  The inverse of this method is
+ * <code>indexOf()</code>.
+ * @param set the set
+ * @param charIndex an index from 0..size()-1 to obtain the char for
+ * @return the character at the given index, or (UChar32)-1.
+ * @stable ICU 3.2
+ */
+U_CAPI UChar32 U_EXPORT2
+uset_charAt(const USet* set, int32_t charIndex);
+
+/**
+ * Returns the number of characters and strings contained in the given
+ * USet.
+ * @param set the set
+ * @return a non-negative integer counting the characters and strings
+ * contained in set
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_size(const USet* set);
+
+/**
+ * Returns the number of items in this set.  An item is either a range
+ * of characters or a single multicharacter string.
+ * @param set the set
+ * @return a non-negative integer counting the character ranges
+ * and/or strings contained in set
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_getItemCount(const USet* set);
+
+/**
+ * Returns an item of this set.  An item is either a range of
+ * characters or a single multicharacter string.
+ * @param set the set
+ * @param itemIndex a non-negative integer in the range 0..
+ * uset_getItemCount(set)-1
+ * @param start pointer to variable to receive first character
+ * in range, inclusive
+ * @param end pointer to variable to receive last character in range,
+ * inclusive
+ * @param str buffer to receive the string, may be NULL
+ * @param strCapacity capacity of str, or 0 if str is NULL
+ * @param ec error code
+ * @return the length of the string (>= 2), or 0 if the item is a
+ * range, in which case it is the range *start..*end, or -1 if
+ * itemIndex is out of range
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_getItem(const USet* set, int32_t itemIndex,
+             UChar32* start, UChar32* end,
+             UChar* str, int32_t strCapacity,
+             UErrorCode* ec);
+
+/**
+ * Returns true if set1 contains all the characters and strings
+ * of set2. It answers the question, 'Is set1 a superset of set2?'
+ * @param set1 set to be checked for containment
+ * @param set2 set to be checked for containment
+ * @return true if the test condition is met
+ * @stable ICU 3.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsAll(const USet* set1, const USet* set2);
+
+/**
+ * Returns true if this set contains all the characters
+ * of the given string. This is does not check containment of grapheme
+ * clusters, like uset_containsString.
+ * @param set set of characters to be checked for containment
+ * @param str string containing codepoints to be checked for containment
+ * @param strLen the length of the string or -1 if null terminated.
+ * @return true if the test condition is met
+ * @stable ICU 3.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsAllCodePoints(const USet* set, const UChar *str, int32_t strLen);
+
+/**
+ * Returns true if set1 contains none of the characters and strings
+ * of set2. It answers the question, 'Is set1 a disjoint set of set2?'
+ * @param set1 set to be checked for containment
+ * @param set2 set to be checked for containment
+ * @return true if the test condition is met
+ * @stable ICU 3.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsNone(const USet* set1, const USet* set2);
+
+/**
+ * Returns true if set1 contains some of the characters and strings
+ * of set2. It answers the question, 'Does set1 and set2 have an intersection?'
+ * @param set1 set to be checked for containment
+ * @param set2 set to be checked for containment
+ * @return true if the test condition is met
+ * @stable ICU 3.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_containsSome(const USet* set1, const USet* set2);
+
+/**
+ * Returns the length of the initial substring of the input string which
+ * consists only of characters and strings that are contained in this set
+ * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+ * or only of characters and strings that are not contained
+ * in this set (USET_SPAN_NOT_CONTAINED).
+ * See USetSpanCondition for details.
+ * Similar to the strspn() C library function.
+ * Unpaired surrogates are treated according to contains() of their surrogate code points.
+ * This function works faster with a frozen set and with a non-negative string length argument.
+ * @param set the set
+ * @param s start of the string
+ * @param length of the string; can be -1 for NUL-terminated
+ * @param spanCondition specifies the containment condition
+ * @return the length of the initial substring according to the spanCondition;
+ *         0 if the start of the string does not fit the spanCondition
+ * @stable ICU 3.8
+ * @see USetSpanCondition
+ */
+U_CAPI int32_t U_EXPORT2
+uset_span(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition);
+
+/**
+ * Returns the start of the trailing substring of the input string which
+ * consists only of characters and strings that are contained in this set
+ * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+ * or only of characters and strings that are not contained
+ * in this set (USET_SPAN_NOT_CONTAINED).
+ * See USetSpanCondition for details.
+ * Unpaired surrogates are treated according to contains() of their surrogate code points.
+ * This function works faster with a frozen set and with a non-negative string length argument.
+ * @param set the set
+ * @param s start of the string
+ * @param length of the string; can be -1 for NUL-terminated
+ * @param spanCondition specifies the containment condition
+ * @return the start of the trailing substring according to the spanCondition;
+ *         the string length if the end of the string does not fit the spanCondition
+ * @stable ICU 3.8
+ * @see USetSpanCondition
+ */
+U_CAPI int32_t U_EXPORT2
+uset_spanBack(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition);
+
+/**
+ * Returns the length of the initial substring of the input string which
+ * consists only of characters and strings that are contained in this set
+ * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+ * or only of characters and strings that are not contained
+ * in this set (USET_SPAN_NOT_CONTAINED).
+ * See USetSpanCondition for details.
+ * Similar to the strspn() C library function.
+ * Malformed byte sequences are treated according to contains(0xfffd).
+ * This function works faster with a frozen set and with a non-negative string length argument.
+ * @param set the set
+ * @param s start of the string (UTF-8)
+ * @param length of the string; can be -1 for NUL-terminated
+ * @param spanCondition specifies the containment condition
+ * @return the length of the initial substring according to the spanCondition;
+ *         0 if the start of the string does not fit the spanCondition
+ * @stable ICU 3.8
+ * @see USetSpanCondition
+ */
+U_CAPI int32_t U_EXPORT2
+uset_spanUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition);
+
+/**
+ * Returns the start of the trailing substring of the input string which
+ * consists only of characters and strings that are contained in this set
+ * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
+ * or only of characters and strings that are not contained
+ * in this set (USET_SPAN_NOT_CONTAINED).
+ * See USetSpanCondition for details.
+ * Malformed byte sequences are treated according to contains(0xfffd).
+ * This function works faster with a frozen set and with a non-negative string length argument.
+ * @param set the set
+ * @param s start of the string (UTF-8)
+ * @param length of the string; can be -1 for NUL-terminated
+ * @param spanCondition specifies the containment condition
+ * @return the start of the trailing substring according to the spanCondition;
+ *         the string length if the end of the string does not fit the spanCondition
+ * @stable ICU 3.8
+ * @see USetSpanCondition
+ */
+U_CAPI int32_t U_EXPORT2
+uset_spanBackUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition);
+
+/**
+ * Returns true if set1 contains all of the characters and strings
+ * of set2, and vis versa. It answers the question, 'Is set1 equal to set2?'
+ * @param set1 set to be checked for containment
+ * @param set2 set to be checked for containment
+ * @return true if the test condition is met
+ * @stable ICU 3.2
+ */
+U_CAPI UBool U_EXPORT2
+uset_equals(const USet* set1, const USet* set2);
+
+/*********************************************************************
+ * Serialized set API
+ *********************************************************************/
+
+/**
+ * Serializes this set into an array of 16-bit integers.  Serialization
+ * (currently) only records the characters in the set; multicharacter
+ * strings are ignored.
+ *
+ * The array
+ * has following format (each line is one 16-bit integer):
+ *
+ *  length     = (n+2*m) | (m!=0?0x8000:0)
+ *  bmpLength  = n; present if m!=0
+ *  bmp[0]
+ *  bmp[1]
+ *  ...
+ *  bmp[n-1]
+ *  supp-high[0]
+ *  supp-low[0]
+ *  supp-high[1]
+ *  supp-low[1]
+ *  ...
+ *  supp-high[m-1]
+ *  supp-low[m-1]
+ *
+ * The array starts with a header.  After the header are n bmp
+ * code points, then m supplementary code points.  Either n or m
+ * or both may be zero.  n+2*m is always <= 0x7FFF.
+ *
+ * If there are no supplementary characters (if m==0) then the
+ * header is one 16-bit integer, 'length', with value n.
+ *
+ * If there are supplementary characters (if m!=0) then the header
+ * is two 16-bit integers.  The first, 'length', has value
+ * (n+2*m)|0x8000.  The second, 'bmpLength', has value n.
+ *
+ * After the header the code points are stored in ascending order.
+ * Supplementary code points are stored as most significant 16
+ * bits followed by least significant 16 bits.
+ *
+ * @param set the set
+ * @param dest pointer to buffer of destCapacity 16-bit integers.
+ * May be NULL only if destCapacity is zero.
+ * @param destCapacity size of dest, or zero.  Must not be negative.
+ * @param pErrorCode pointer to the error code.  Will be set to
+ * U_INDEX_OUTOFBOUNDS_ERROR if n+2*m > 0x7FFF.  Will be set to
+ * U_BUFFER_OVERFLOW_ERROR if n+2*m+(m!=0?2:1) > destCapacity.
+ * @return the total length of the serialized format, including
+ * the header, that is, n+2*m+(m!=0?2:1), or 0 on error other
+ * than U_BUFFER_OVERFLOW_ERROR.
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_serialize(const USet* set, uint16_t* dest, int32_t destCapacity, UErrorCode* pErrorCode);
+
+/**
+ * Given a serialized array, fill in the given serialized set object.
+ * @param fillSet pointer to result
+ * @param src pointer to start of array
+ * @param srcLength length of array
+ * @return true if the given array is valid, otherwise false
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_getSerializedSet(USerializedSet* fillSet, const uint16_t* src, int32_t srcLength);
+
+/**
+ * Set the USerializedSet to contain the given character (and nothing
+ * else).
+ * @param fillSet pointer to result
+ * @param c The codepoint to set
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+uset_setSerializedToOne(USerializedSet* fillSet, UChar32 c);
+
+/**
+ * Returns true if the given USerializedSet contains the given
+ * character.
+ * @param set the serialized set
+ * @param c The codepoint to check for within the set
+ * @return true if set contains c
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_serializedContains(const USerializedSet* set, UChar32 c);
+
+/**
+ * Returns the number of disjoint ranges of characters contained in
+ * the given serialized set.  Ignores any strings contained in the
+ * set.
+ * @param set the serialized set
+ * @return a non-negative integer counting the character ranges
+ * contained in set
+ * @stable ICU 2.4
+ */
+U_CAPI int32_t U_EXPORT2
+uset_getSerializedRangeCount(const USerializedSet* set);
+
+/**
+ * Returns a range of characters contained in the given serialized
+ * set.
+ * @param set the serialized set
+ * @param rangeIndex a non-negative integer in the range 0..
+ * uset_getSerializedRangeCount(set)-1
+ * @param pStart pointer to variable to receive first character
+ * in range, inclusive
+ * @param pEnd pointer to variable to receive last character in range,
+ * inclusive
+ * @return true if rangeIndex is valid, otherwise false
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+uset_getSerializedRange(const USerializedSet* set, int32_t rangeIndex,
+                        UChar32* pStart, UChar32* pEnd);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/usetiter.h b/thirdparty/icu4c/common/unicode/usetiter.h
new file mode 100644
index 0000000000..a817ef72b3
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/usetiter.h
@@ -0,0 +1,325 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+#ifndef USETITER_H
+#define USETITER_H
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+
+/**
+ * \file 
+ * \brief C++ API: UnicodeSetIterator iterates over the contents of a UnicodeSet.
+ */
+
+U_NAMESPACE_BEGIN
+
+class UnicodeSet;
+class UnicodeString;
+
+/**
+ *
+ * UnicodeSetIterator iterates over the contents of a UnicodeSet.  It
+ * iterates over either code points or code point ranges.  After all
+ * code points or ranges have been returned, it returns the
+ * multicharacter strings of the UnicodeSet, if any.
+ *
+ * This class is not intended to be subclassed.  Consider any fields
+ *  or methods declared as "protected" to be private.  The use of
+ *  protected in this class is an artifact of history.
+ *
+ * <p>To iterate over code points and strings, use a loop like this:
+ * <pre>
+ * UnicodeSetIterator it(set);
+ * while (it.next()) {
+ *     processItem(it.getString());
+ * }
+ * </pre>
+ * <p>Each item in the set is accessed as a string.  Set elements
+ *    consisting of single code points are returned as strings containing
+ *    just the one code point.
+ *
+ * <p>To iterate over code point ranges, instead of individual code points,
+ *    use a loop like this:
+ * <pre>
+ * UnicodeSetIterator it(set);
+ * while (it.nextRange()) {
+ *   if (it.isString()) {
+ *     processString(it.getString());
+ *   } else {
+ *     processCodepointRange(it.getCodepoint(), it.getCodepointEnd());
+ *   }
+ * }
+ * </pre>
+ * @author M. Davis
+ * @stable ICU 2.4
+ */
+class U_COMMON_API UnicodeSetIterator : public UObject {
+
+ protected:
+
+    /**
+     * Value of <tt>codepoint</tt> if the iterator points to a string.
+     * If <tt>codepoint == IS_STRING</tt>, then examine
+     * <tt>string</tt> for the current iteration result.
+     * @stable ICU 2.4
+     */
+    enum { IS_STRING = -1 };
+
+    /**
+     * Current code point, or the special value <tt>IS_STRING</tt>, if
+     * the iterator points to a string.
+     * @stable ICU 2.4
+     */
+    UChar32 codepoint;
+
+    /**
+     * When iterating over ranges using <tt>nextRange()</tt>,
+     * <tt>codepointEnd</tt> contains the inclusive end of the
+     * iteration range, if <tt>codepoint != IS_STRING</tt>.  If
+     * iterating over code points using <tt>next()</tt>, or if
+     * <tt>codepoint == IS_STRING</tt>, then the value of
+     * <tt>codepointEnd</tt> is undefined.
+     * @stable ICU 2.4
+     */
+    UChar32 codepointEnd;
+
+    /**
+     * If <tt>codepoint == IS_STRING</tt>, then <tt>string</tt> points
+     * to the current string.  If <tt>codepoint != IS_STRING</tt>, the
+     * value of <tt>string</tt> is undefined.
+     * @stable ICU 2.4
+     */
+    const UnicodeString* string;
+
+ public:
+
+    /**
+     * Create an iterator over the given set.  The iterator is valid
+     * only so long as <tt>set</tt> is valid.
+     * @param set set to iterate over
+     * @stable ICU 2.4
+     */
+    UnicodeSetIterator(const UnicodeSet& set);
+
+    /**
+     * Create an iterator over nothing.  <tt>next()</tt> and
+     * <tt>nextRange()</tt> return false. This is a convenience
+     * constructor allowing the target to be set later.
+     * @stable ICU 2.4
+     */
+    UnicodeSetIterator();
+
+    /**
+     * Destructor.
+     * @stable ICU 2.4
+     */
+    virtual ~UnicodeSetIterator();
+
+    /**
+     * Returns true if the current element is a string.  If so, the
+     * caller can retrieve it with <tt>getString()</tt>.  If this
+     * method returns false, the current element is a code point or
+     * code point range, depending on whether <tt>next()</tt> or
+     * <tt>nextRange()</tt> was called.
+     * Elements of types string and codepoint can both be retrieved
+     * with the function <tt>getString()</tt>.
+     * Elements of type codepoint can also be retrieved with
+     * <tt>getCodepoint()</tt>.
+     * For ranges, <tt>getCodepoint()</tt> returns the starting codepoint
+     * of the range, and <tt>getCodepointEnd()</tt> returns the end
+     * of the range.
+     * @stable ICU 2.4
+     */
+    inline UBool isString() const;
+
+    /**
+     * Returns the current code point, if <tt>isString()</tt> returned
+     * false.  Otherwise returns an undefined result.
+     * @stable ICU 2.4
+     */
+    inline UChar32 getCodepoint() const;
+
+    /**
+     * Returns the end of the current code point range, if
+     * <tt>isString()</tt> returned false and <tt>nextRange()</tt> was
+     * called.  Otherwise returns an undefined result.
+     * @stable ICU 2.4
+     */
+    inline UChar32 getCodepointEnd() const;
+
+    /**
+     * Returns the current string, if <tt>isString()</tt> returned
+     * true.  If the current iteration item is a code point, a UnicodeString
+     * containing that single code point is returned.
+     *
+     * Ownership of the returned string remains with the iterator.
+     * The string is guaranteed to remain valid only until the iterator is
+     *   advanced to the next item, or until the iterator is deleted.
+     * 
+     * @stable ICU 2.4
+     */
+    const UnicodeString& getString();
+
+    /**
+     * Advances the iteration position to the next element in the set, 
+     * which can be either a single code point or a string.  
+     * If there are no more elements in the set, return false.
+     *
+     * <p>
+     * If <tt>isString() == true</tt>, the value is a
+     * string, otherwise the value is a
+     * single code point.  Elements of either type can be retrieved
+     * with the function <tt>getString()</tt>, while elements of
+     * consisting of a single code point can be retrieved with
+     * <tt>getCodepoint()</tt>
+     *
+     * <p>The order of iteration is all code points in sorted order,
+     * followed by all strings sorted order.    Do not mix
+     * calls to <tt>next()</tt> and <tt>nextRange()</tt> without
+     * calling <tt>reset()</tt> between them.  The results of doing so
+     * are undefined.
+     *
+     * @return true if there was another element in the set.
+     * @stable ICU 2.4
+     */
+    UBool next();
+
+    /**
+     * Returns the next element in the set, either a code point range
+     * or a string.  If there are no more elements in the set, return
+     * false.  If <tt>isString() == true</tt>, the value is a
+     * string and can be accessed with <tt>getString()</tt>.  Otherwise the value is a
+     * range of one or more code points from <tt>getCodepoint()</tt> to
+     * <tt>getCodepointeEnd()</tt> inclusive.
+     *
+     * <p>The order of iteration is all code points ranges in sorted
+     * order, followed by all strings sorted order.  Ranges are
+     * disjoint and non-contiguous.  The value returned from <tt>getString()</tt>
+     * is undefined unless <tt>isString() == true</tt>.  Do not mix calls to
+     * <tt>next()</tt> and <tt>nextRange()</tt> without calling
+     * <tt>reset()</tt> between them.  The results of doing so are
+     * undefined.
+     *
+     * @return true if there was another element in the set.
+     * @stable ICU 2.4
+     */
+    UBool nextRange();
+
+    /**
+     * Sets this iterator to visit the elements of the given set and
+     * resets it to the start of that set.  The iterator is valid only
+     * so long as <tt>set</tt> is valid.
+     * @param set the set to iterate over.
+     * @stable ICU 2.4
+     */
+    void reset(const UnicodeSet& set);
+
+    /**
+     * Resets this iterator to the start of the set.
+     * @stable ICU 2.4
+     */
+    void reset();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     *
+     * @stable ICU 2.4
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     *
+     * @stable ICU 2.4
+     */
+    virtual UClassID getDynamicClassID() const;
+
+    // ======================= PRIVATES ===========================
+
+ protected:
+
+    // endElement and nextElements are really UChar32's, but we keep
+    // them as signed int32_t's so we can do comparisons with
+    // endElement set to -1.  Leave them as int32_t's.
+    /** The set
+     * @stable ICU 2.4
+     */
+    const UnicodeSet* set;
+    /** End range
+     * @stable ICU 2.4
+     */
+    int32_t endRange;
+    /** Range
+     * @stable ICU 2.4
+     */
+    int32_t range;
+    /** End element
+     * @stable ICU 2.4
+     */
+    int32_t endElement;
+    /** Next element
+     * @stable ICU 2.4
+     */
+    int32_t nextElement;
+    //UBool abbreviated;
+    /** Next string
+     * @stable ICU 2.4
+     */
+    int32_t nextString;
+    /** String count
+     * @stable ICU 2.4
+     */
+    int32_t stringCount;
+
+    /**
+     *  Points to the string to use when the caller asks for a
+     *  string and the current iteration item is a code point, not a string.
+     *  @internal
+     */
+    UnicodeString *cpString;
+
+    /** Copy constructor. Disallowed.
+     * @stable ICU 2.4
+     */
+    UnicodeSetIterator(const UnicodeSetIterator&); // disallow
+
+    /** Assignment operator. Disallowed.
+     * @stable ICU 2.4
+     */
+    UnicodeSetIterator& operator=(const UnicodeSetIterator&); // disallow
+
+    /** Load range
+     * @stable ICU 2.4
+     */
+    virtual void loadRange(int32_t range);
+
+};
+
+inline UBool UnicodeSetIterator::isString() const {
+    return codepoint == (UChar32)IS_STRING;
+}
+
+inline UChar32 UnicodeSetIterator::getCodepoint() const {
+    return codepoint;
+}
+
+inline UChar32 UnicodeSetIterator::getCodepointEnd() const {
+    return codepointEnd;
+}
+
+
+U_NAMESPACE_END
+
+#endif /* U_SHOW_CPLUSPLUS_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ushape.h b/thirdparty/icu4c/common/unicode/ushape.h
new file mode 100644
index 0000000000..fed4869abd
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ushape.h
@@ -0,0 +1,476 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2000-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  ushape.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2000jun29
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __USHAPE_H__
+#define __USHAPE_H__
+
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief C API:  Arabic shaping
+ * 
+ */
+
+/**
+ * Shape Arabic text on a character basis.
+ *
+ * <p>This function performs basic operations for "shaping" Arabic text. It is most
+ * useful for use with legacy data formats and legacy display technology
+ * (simple terminals). All operations are performed on Unicode characters.</p>
+ *
+ * <p>Text-based shaping means that some character code points in the text are
+ * replaced by others depending on the context. It transforms one kind of text
+ * into another. In comparison, modern displays for Arabic text select
+ * appropriate, context-dependent font glyphs for each text element, which means
+ * that they transform text into a glyph vector.</p>
+ *
+ * <p>Text transformations are necessary when modern display technology is not
+ * available or when text needs to be transformed to or from legacy formats that
+ * use "shaped" characters. Since the Arabic script is cursive, connecting
+ * adjacent letters to each other, computers select images for each letter based
+ * on the surrounding letters. This usually results in four images per Arabic
+ * letter: initial, middle, final, and isolated forms. In Unicode, on the other
+ * hand, letters are normally stored abstract, and a display system is expected
+ * to select the necessary glyphs. (This makes searching and other text
+ * processing easier because the same letter has only one code.) It is possible
+ * to mimic this with text transformations because there are characters in
+ * Unicode that are rendered as letters with a specific shape
+ * (or cursive connectivity). They were included for interoperability with
+ * legacy systems and codepages, and for unsophisticated display systems.</p>
+ *
+ * <p>A second kind of text transformations is supported for Arabic digits:
+ * For compatibility with legacy codepages that only include European digits,
+ * it is possible to replace one set of digits by another, changing the
+ * character code points. These operations can be performed for either
+ * Arabic-Indic Digits (U+0660...U+0669) or Eastern (Extended) Arabic-Indic
+ * digits (U+06f0...U+06f9).</p>
+ *
+ * <p>Some replacements may result in more or fewer characters (code points).
+ * By default, this means that the destination buffer may receive text with a
+ * length different from the source length. Some legacy systems rely on the
+ * length of the text to be constant. They expect extra spaces to be added
+ * or consumed either next to the affected character or at the end of the
+ * text.</p>
+ *
+ * <p>For details about the available operations, see the description of the
+ * <code>U_SHAPE_...</code> options.</p>
+ *
+ * @param source The input text.
+ *
+ * @param sourceLength The number of UChars in <code>source</code>.
+ *
+ * @param dest The destination buffer that will receive the results of the
+ *             requested operations. It may be <code>NULL</code> only if
+ *             <code>destSize</code> is 0. The source and destination must not
+ *             overlap.
+ *
+ * @param destSize The size (capacity) of the destination buffer in UChars.
+ *                 If <code>destSize</code> is 0, then no output is produced,
+ *                 but the necessary buffer size is returned ("preflighting").
+ *
+ * @param options This is a 32-bit set of flags that specify the operations
+ *                that are performed on the input text. If no error occurs,
+ *                then the result will always be written to the destination
+ *                buffer.
+ *
+ * @param pErrorCode must be a valid pointer to an error code value,
+ *        which must not indicate a failure before the function call.
+ *
+ * @return The number of UChars written to the destination buffer.
+ *         If an error occurred, then no output was written, or it may be
+ *         incomplete. If <code>U_BUFFER_OVERFLOW_ERROR</code> is set, then
+ *         the return value indicates the necessary destination buffer size.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_shapeArabic(const UChar *source, int32_t sourceLength,
+              UChar *dest, int32_t destSize,
+              uint32_t options,
+              UErrorCode *pErrorCode);
+
+/**
+ * Memory option: allow the result to have a different length than the source.
+ * Affects: LamAlef options
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_LENGTH_GROW_SHRINK              0
+
+/**
+ * Memory option: allow the result to have a different length than the source.
+ * Affects: LamAlef options
+ * This option is an alias to U_SHAPE_LENGTH_GROW_SHRINK
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_RESIZE                  0 
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces next to modified characters.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_LENGTH_FIXED_SPACES_NEAR        1
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces next to modified characters.
+ * Affects: LamAlef options
+ * This option is an alias to U_SHAPE_LENGTH_FIXED_SPACES_NEAR
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_NEAR                    1 
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces at the end of the text.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_LENGTH_FIXED_SPACES_AT_END      2
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces at the end of the text.
+ * Affects: LamAlef options
+ * This option is an alias to U_SHAPE_LENGTH_FIXED_SPACES_AT_END
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_END                     2 
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces at the beginning of the text.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_LENGTH_FIXED_SPACES_AT_BEGINNING 3
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * If more room is necessary, then try to consume spaces at the beginning of the text.
+ * Affects: LamAlef options
+ * This option is an alias to U_SHAPE_LENGTH_FIXED_SPACES_AT_BEGINNING
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_BEGIN                    3 
+
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping Mode: For each LAMALEF character found, expand LAMALEF using space at end.
+ *               If there is no space at end, use spaces at beginning of the buffer. If there
+ *               is no space at beginning of the buffer, use spaces at the near (i.e. the space
+ *               after the LAMALEF character).
+ *               If there are no spaces found, an error U_NO_SPACE_AVAILABLE (as defined in utypes.h) 
+ *               will be set in pErrorCode
+ *
+ * Deshaping Mode: Perform the same function as the flag equals U_SHAPE_LAMALEF_END. 
+ * Affects: LamAlef options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_AUTO                     0x10000 
+
+/** Bit mask for memory options. @stable ICU 2.0 */
+#define U_SHAPE_LENGTH_MASK                      0x10003 /* Changed old value 3 */
+
+
+/**
+ * Bit mask for LamAlef memory options.
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_LAMALEF_MASK                     0x10003 /* updated */
+
+/** Direction indicator: the source is in logical (keyboard) order. @stable ICU 2.0 */
+#define U_SHAPE_TEXT_DIRECTION_LOGICAL          0
+
+/**
+ * Direction indicator:
+ * the source is in visual RTL order,
+ * the rightmost displayed character stored first.
+ * This option is an alias to U_SHAPE_TEXT_DIRECTION_LOGICAL
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TEXT_DIRECTION_VISUAL_RTL       0
+
+/**
+ * Direction indicator:
+ * the source is in visual LTR order,
+ * the leftmost displayed character stored first.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_TEXT_DIRECTION_VISUAL_LTR       4
+
+/** Bit mask for direction indicators. @stable ICU 2.0 */
+#define U_SHAPE_TEXT_DIRECTION_MASK             4
+
+
+/** Letter shaping option: do not perform letter shaping. @stable ICU 2.0 */
+#define U_SHAPE_LETTERS_NOOP                    0
+
+/** Letter shaping option: replace abstract letter characters by "shaped" ones. @stable ICU 2.0 */
+#define U_SHAPE_LETTERS_SHAPE                   8
+
+/** Letter shaping option: replace "shaped" letter characters by abstract ones. @stable ICU 2.0 */
+#define U_SHAPE_LETTERS_UNSHAPE                 0x10
+
+/**
+ * Letter shaping option: replace abstract letter characters by "shaped" ones.
+ * The only difference with U_SHAPE_LETTERS_SHAPE is that Tashkeel letters
+ * are always "shaped" into the isolated form instead of the medial form
+ * (selecting code points from the Arabic Presentation Forms-B block).
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED 0x18
+
+
+/** Bit mask for letter shaping options. @stable ICU 2.0 */
+#define U_SHAPE_LETTERS_MASK                        0x18
+
+
+/** Digit shaping option: do not perform digit shaping. @stable ICU 2.0 */
+#define U_SHAPE_DIGITS_NOOP                     0
+
+/**
+ * Digit shaping option:
+ * Replace European digits (U+0030...) by Arabic-Indic digits.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_DIGITS_EN2AN                    0x20
+
+/**
+ * Digit shaping option:
+ * Replace Arabic-Indic digits by European digits (U+0030...).
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_DIGITS_AN2EN                    0x40
+
+/**
+ * Digit shaping option:
+ * Replace European digits (U+0030...) by Arabic-Indic digits if the most recent
+ * strongly directional character is an Arabic letter
+ * (<code>u_charDirection()</code> result <code>U_RIGHT_TO_LEFT_ARABIC</code> [AL]).<br>
+ * The direction of "preceding" depends on the direction indicator option.
+ * For the first characters, the preceding strongly directional character
+ * (initial state) is assumed to be not an Arabic letter
+ * (it is <code>U_LEFT_TO_RIGHT</code> [L] or <code>U_RIGHT_TO_LEFT</code> [R]).
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_DIGITS_ALEN2AN_INIT_LR          0x60
+
+/**
+ * Digit shaping option:
+ * Replace European digits (U+0030...) by Arabic-Indic digits if the most recent
+ * strongly directional character is an Arabic letter
+ * (<code>u_charDirection()</code> result <code>U_RIGHT_TO_LEFT_ARABIC</code> [AL]).<br>
+ * The direction of "preceding" depends on the direction indicator option.
+ * For the first characters, the preceding strongly directional character
+ * (initial state) is assumed to be an Arabic letter.
+ * @stable ICU 2.0
+ */
+#define U_SHAPE_DIGITS_ALEN2AN_INIT_AL          0x80
+
+/** Not a valid option value. May be replaced by a new option. @stable ICU 2.0 */
+#define U_SHAPE_DIGITS_RESERVED                 0xa0
+
+/** Bit mask for digit shaping options. @stable ICU 2.0 */
+#define U_SHAPE_DIGITS_MASK                     0xe0
+
+
+/** Digit type option: Use Arabic-Indic digits (U+0660...U+0669). @stable ICU 2.0 */
+#define U_SHAPE_DIGIT_TYPE_AN                   0
+
+/** Digit type option: Use Eastern (Extended) Arabic-Indic digits (U+06f0...U+06f9). @stable ICU 2.0 */
+#define U_SHAPE_DIGIT_TYPE_AN_EXTENDED          0x100
+
+/** Not a valid option value. May be replaced by a new option. @stable ICU 2.0 */
+#define U_SHAPE_DIGIT_TYPE_RESERVED             0x200
+
+/** Bit mask for digit type options. @stable ICU 2.0 */
+#define U_SHAPE_DIGIT_TYPE_MASK                 0x300 /* I need to change this from 0x3f00 to 0x300 */
+
+/** 
+ * Tashkeel aggregation option:
+ * Replaces any combination of U+0651 with one of
+ * U+064C, U+064D, U+064E, U+064F, U+0650 with
+ * U+FC5E, U+FC5F, U+FC60, U+FC61, U+FC62 consecutively.
+ * @stable ICU 3.6
+ */
+#define U_SHAPE_AGGREGATE_TASHKEEL              0x4000
+/** Tashkeel aggregation option: do not aggregate tashkeels. @stable ICU 3.6 */
+#define U_SHAPE_AGGREGATE_TASHKEEL_NOOP         0
+/** Bit mask for tashkeel aggregation. @stable ICU 3.6 */
+#define U_SHAPE_AGGREGATE_TASHKEEL_MASK         0x4000
+
+/** 
+ * Presentation form option:
+ * Don't replace Arabic Presentation Forms-A and Arabic Presentation Forms-B
+ * characters with 0+06xx characters, before shaping.
+ * @stable ICU 3.6
+ */
+#define U_SHAPE_PRESERVE_PRESENTATION           0x8000
+/** Presentation form option: 
+ * Replace Arabic Presentation Forms-A and Arabic Presentationo Forms-B with 
+ * their unshaped correspondants in range 0+06xx, before shaping.
+ * @stable ICU 3.6 
+ */
+#define U_SHAPE_PRESERVE_PRESENTATION_NOOP      0
+/** Bit mask for preserve presentation form. @stable ICU 3.6 */
+#define U_SHAPE_PRESERVE_PRESENTATION_MASK      0x8000
+
+/* Seen Tail option */ 
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping mode: The SEEN family character will expand into two characters using space near 
+ *               the SEEN family character(i.e. the space after the character).
+ *               If there are no spaces found, an error U_NO_SPACE_AVAILABLE (as defined in utypes.h) 
+ *               will be set in pErrorCode
+ *
+ * De-shaping mode: Any Seen character followed by Tail character will be
+ *                  replaced by one cell Seen and a space will replace the Tail.
+ * Affects: Seen options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_SEEN_TWOCELL_NEAR     0x200000
+
+/**
+ * Bit mask for Seen memory options. 
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_SEEN_MASK             0x700000
+
+/* YehHamza option */ 
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping mode: The YEHHAMZA character will expand into two characters using space near it 
+ *              (i.e. the space after the character
+ *               If there are no spaces found, an error U_NO_SPACE_AVAILABLE (as defined in utypes.h) 
+ *               will be set in pErrorCode
+ *
+ * De-shaping mode: Any Yeh (final or isolated) character followed by Hamza character will be
+ *                  replaced by one cell YehHamza and space will replace the Hamza.
+ * Affects: YehHamza options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_YEHHAMZA_TWOCELL_NEAR      0x1000000
+
+
+/**
+ * Bit mask for YehHamza memory options. 
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_YEHHAMZA_MASK              0x3800000
+
+/* New Tashkeel options */ 
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping mode: Tashkeel characters will be replaced by spaces. 
+ *               Spaces will be placed at beginning of the buffer
+ *
+ * De-shaping mode: N/A
+ * Affects: Tashkeel options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TASHKEEL_BEGIN                      0x40000
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping mode: Tashkeel characters will be replaced by spaces. 
+ *               Spaces will be placed at end of the buffer
+ *
+ * De-shaping mode: N/A
+ * Affects: Tashkeel options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TASHKEEL_END                        0x60000
+
+/**
+ * Memory option: allow the result to have a different length than the source.
+ * Shaping mode: Tashkeel characters will be removed, buffer length will shrink. 
+ * De-shaping mode: N/A 
+ *
+ * Affect: Tashkeel options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TASHKEEL_RESIZE                     0x80000
+
+/**
+ * Memory option: the result must have the same length as the source.
+ * Shaping mode: Tashkeel characters will be replaced by Tatweel if it is connected to adjacent
+ *               characters (i.e. shaped on Tatweel) or replaced by space if it is not connected.
+ *
+ * De-shaping mode: N/A
+ * Affects: YehHamza options
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TASHKEEL_REPLACE_BY_TATWEEL         0xC0000
+
+/** 
+ * Bit mask for Tashkeel replacement with Space or Tatweel memory options. 
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_TASHKEEL_MASK                       0xE0000
+
+
+/* Space location Control options */ 
+/**
+ * This option affect the meaning of BEGIN and END options. if this option is not used the default
+ * for BEGIN and END will be as following: 
+ * The Default (for both Visual LTR, Visual RTL and Logical Text)
+ *           1. BEGIN always refers to the start address of physical memory.
+ *           2. END always refers to the end address of physical memory.
+ *
+ * If this option is used it will swap the meaning of BEGIN and END only for Visual LTR text. 
+ *
+ * The effect on BEGIN and END Memory Options will be as following:
+ *    A. BEGIN For Visual LTR text: This will be the beginning (right side) of the visual text(
+ *       corresponding to the physical memory address end for Visual LTR text, Same as END in 
+ *       default behavior)
+ *    B. BEGIN For Logical text: Same as BEGIN in default behavior. 
+ *    C. END For Visual LTR text: This will be the end (left side) of the visual text (corresponding
+ *       to the physical memory address beginning for Visual LTR text, Same as BEGIN in default behavior.
+ *    D. END For Logical text: Same as END in default behavior). 
+ * Affects: All LamAlef BEGIN, END and AUTO options.
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END 0x4000000
+
+/**
+ * Bit mask for swapping BEGIN and END for Visual LTR text 
+ * @stable ICU 4.2
+ */
+#define U_SHAPE_SPACES_RELATIVE_TO_TEXT_MASK      0x4000000
+
+/**
+ * If this option is used, shaping will use the new Unicode code point for TAIL (i.e. 0xFE73). 
+ * If this option is not specified (Default), old unofficial Unicode TAIL code point is used (i.e. 0x200B)
+ * De-shaping will not use this option as it will always search for both the new Unicode code point for the 
+ * TAIL (i.e. 0xFE73) or the old unofficial Unicode TAIL code point (i.e. 0x200B) and de-shape the
+ * Seen-Family letter accordingly.
+ *
+ * Shaping Mode: Only shaping.
+ * De-shaping Mode: N/A.
+ * Affects: All Seen options
+ * @stable ICU 4.8
+ */
+#define U_SHAPE_TAIL_NEW_UNICODE        0x8000000
+
+/**
+ * Bit mask for new Unicode Tail option 
+ * @stable ICU 4.8
+ */
+#define U_SHAPE_TAIL_TYPE_MASK          0x8000000
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/usprep.h b/thirdparty/icu4c/common/unicode/usprep.h
new file mode 100644
index 0000000000..f8a0f58e0d
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/usprep.h
@@ -0,0 +1,274 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ *
+ *   Copyright (C) 2003-2014, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *******************************************************************************
+ *   file name:  usprep.h
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2003jul2
+ *   created by: Ram Viswanadha
+ */
+
+#ifndef __USPREP_H__
+#define __USPREP_H__
+
+/**
+ * \file 
+ * \brief C API: Implements the StringPrep algorithm.
+ */
+
+#include "unicode/utypes.h"
+
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#endif   // U_SHOW_CPLUSPLUS_API
+
+/**
+ *
+ * StringPrep API implements the StingPrep framework as described by RFC 3454.
+ * StringPrep prepares Unicode strings for use in network protocols.
+ * Profiles of StingPrep are set of rules and data according to with the
+ * Unicode Strings are prepared. Each profiles contains tables which describe
+ * how a code point should be treated. The tables are broadly classified into
+ * <ul>
+ *     <li> Unassigned Table: Contains code points that are unassigned 
+ *          in the Unicode Version supported by StringPrep. Currently 
+ *          RFC 3454 supports Unicode 3.2. </li>
+ *     <li> Prohibited Table: Contains code points that are prohibited from
+ *          the output of the StringPrep processing function. </li>
+ *     <li> Mapping Table: Contains code points that are deleted from the output or case mapped. </li>
+ * </ul>
+ * 
+ * The procedure for preparing Unicode strings:
+ * <ol>
+ *      <li> Map: For each character in the input, check if it has a mapping
+ *           and, if so, replace it with its mapping. </li>
+ *      <li> Normalize: Possibly normalize the result of step 1 using Unicode
+ *           normalization. </li>
+ *      <li> Prohibit: Check for any characters that are not allowed in the
+ *        output.  If any are found, return an error.</li>
+ *      <li> Check bidi: Possibly check for right-to-left characters, and if
+ *           any are found, make sure that the whole string satisfies the
+ *           requirements for bidirectional strings.  If the string does not
+ *           satisfy the requirements for bidirectional strings, return an
+ *           error.  </li>
+ * </ol>
+ * @author Ram Viswanadha
+ */
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/parseerr.h"
+
+/**
+ * The StringPrep profile
+ * @stable ICU 2.8
+ */
+typedef struct UStringPrepProfile UStringPrepProfile;
+
+
+/** 
+ * Option to prohibit processing of unassigned code points in the input
+ * 
+ * @see  usprep_prepare
+ * @stable ICU 2.8
+ */
+#define USPREP_DEFAULT 0x0000
+
+/** 
+ * Option to allow processing of unassigned code points in the input
+ * 
+ * @see  usprep_prepare
+ * @stable ICU 2.8
+ */
+#define USPREP_ALLOW_UNASSIGNED 0x0001
+
+/**
+ * enums for the standard stringprep profile types
+ * supported by usprep_openByType.
+ * @see usprep_openByType
+ * @stable ICU 4.2
+ */
+typedef enum UStringPrepProfileType {
+    /**
+     * RFC3491 Nameprep
+     * @stable ICU 4.2
+     */
+    USPREP_RFC3491_NAMEPREP,
+    /**
+     * RFC3530 nfs4_cs_prep
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3530_NFS4_CS_PREP,
+    /**
+     * RFC3530 nfs4_cs_prep with case insensitive option
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3530_NFS4_CS_PREP_CI,
+    /**
+     * RFC3530 nfs4_cis_prep
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3530_NFS4_CIS_PREP,
+    /**
+     * RFC3530 nfs4_mixed_prep for prefix
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3530_NFS4_MIXED_PREP_PREFIX,
+    /**
+     * RFC3530 nfs4_mixed_prep for suffix
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3530_NFS4_MIXED_PREP_SUFFIX,
+    /**
+     * RFC3722 iSCSI
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3722_ISCSI,
+    /**
+     * RFC3920 XMPP Nodeprep
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3920_NODEPREP,
+    /**
+     * RFC3920 XMPP Resourceprep
+     * @stable ICU 4.2
+     */
+	USPREP_RFC3920_RESOURCEPREP,
+    /**
+     * RFC4011 Policy MIB Stringprep
+     * @stable ICU 4.2
+     */
+	USPREP_RFC4011_MIB,
+    /**
+     * RFC4013 SASLprep
+     * @stable ICU 4.2
+     */
+    USPREP_RFC4013_SASLPREP,
+    /**
+     * RFC4505 trace
+     * @stable ICU 4.2
+     */
+	USPREP_RFC4505_TRACE,
+    /**
+     * RFC4518 LDAP
+     * @stable ICU 4.2
+     */
+	USPREP_RFC4518_LDAP,
+    /**
+     * RFC4518 LDAP for case ignore, numeric and stored prefix
+     * matching rules
+     * @stable ICU 4.2
+     */
+	USPREP_RFC4518_LDAP_CI
+} UStringPrepProfileType;
+
+/**
+ * Creates a StringPrep profile from the data file.
+ *
+ * @param path      string containing the full path pointing to the directory
+ *                  where the profile reside followed by the package name
+ *                  e.g. "/usr/resource/my_app/profiles/mydata" on a Unix system.
+ *                  if NULL, ICU default data files will be used.
+ * @param fileName  name of the profile file to be opened
+ * @param status    ICU error code in/out parameter. Must not be NULL.
+ *                  Must fulfill U_SUCCESS before the function call.
+ * @return Pointer to UStringPrepProfile that is opened. Should be closed by
+ * calling usprep_close()
+ * @see usprep_close()
+ * @stable ICU 2.8
+ */
+U_CAPI UStringPrepProfile* U_EXPORT2
+usprep_open(const char* path, 
+            const char* fileName,
+            UErrorCode* status);
+
+/**
+ * Creates a StringPrep profile for the specified profile type.
+ *
+ * @param type		The profile type
+ * @param status    ICU error code in/out parameter. Must not be NULL.
+ *                  Must fulfill U_SUCCESS before the function call.
+ * @return          Pointer to UStringPrepProfile that is opened. Should be closed by
+ *                  calling usprep_close()
+ * @see usprep_close()
+ * @stable ICU 4.2
+ */
+U_CAPI UStringPrepProfile* U_EXPORT2
+usprep_openByType(UStringPrepProfileType type,
+				  UErrorCode* status);
+
+/**
+ * Closes the profile
+ * @param profile The profile to close
+ * @stable ICU 2.8
+ */
+U_CAPI void U_EXPORT2
+usprep_close(UStringPrepProfile* profile);
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUStringPrepProfilePointer
+ * "Smart pointer" class, closes a UStringPrepProfile via usprep_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUStringPrepProfilePointer, UStringPrepProfile, usprep_close);
+
+U_NAMESPACE_END
+
+#endif
+
+/**
+ * Prepare the input buffer for use in applications with the given profile. This operation maps, normalizes(NFKC),
+ * checks for prohibited and BiDi characters in the order defined by RFC 3454
+ * depending on the options specified in the profile.
+ *
+ * @param prep          The profile to use 
+ * @param src           Pointer to UChar buffer containing the string to prepare
+ * @param srcLength     Number of characters in the source string
+ * @param dest          Pointer to the destination buffer to receive the output
+ * @param destCapacity  The capacity of destination array
+ * @param options       A bit set of options:
+ *
+ *  - USPREP_DEFAULT            Prohibit processing of unassigned code points in the input
+ *
+ *  - USPREP_ALLOW_UNASSIGNED   Treat the unassigned code points are in the input 
+ *                              as normal Unicode code points.
+ *
+ * @param parseError        Pointer to UParseError struct to receive information on position 
+ *                          of error if an error is encountered. Can be NULL.
+ * @param status            ICU in/out error code parameter.
+ *                          U_INVALID_CHAR_FOUND if src contains
+ *                          unmatched single surrogates.
+ *                          U_INDEX_OUTOFBOUNDS_ERROR if src contains
+ *                          too many code points.
+ *                          U_BUFFER_OVERFLOW_ERROR if destCapacity is not enough
+ * @return The number of UChars in the destination buffer
+ * @stable ICU 2.8
+ */
+
+U_CAPI int32_t U_EXPORT2
+usprep_prepare(   const UStringPrepProfile* prep,
+                  const UChar* src, int32_t srcLength, 
+                  UChar* dest, int32_t destCapacity,
+                  int32_t options,
+                  UParseError* parseError,
+                  UErrorCode* status );
+
+
+#endif /* #if !UCONFIG_NO_IDNA */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ustring.h b/thirdparty/icu4c/common/unicode/ustring.h
new file mode 100644
index 0000000000..10ea45ead1
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ustring.h
@@ -0,0 +1,1689 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1998-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File ustring.h
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   12/07/98    bertrand    Creation.
+******************************************************************************
+*/
+
+#ifndef USTRING_H
+#define USTRING_H
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/uiter.h"
+
+/**
+ * \def UBRK_TYPEDEF_UBREAK_ITERATOR
+ * @internal 
+ */
+
+#ifndef UBRK_TYPEDEF_UBREAK_ITERATOR
+#   define UBRK_TYPEDEF_UBREAK_ITERATOR
+/** Simple declaration for u_strToTitle() to avoid including unicode/ubrk.h. @stable ICU 2.1*/
+    typedef struct UBreakIterator UBreakIterator;
+#endif
+
+/**
+ * \file
+ * \brief C API: Unicode string handling functions
+ *
+ * These C API functions provide general Unicode string handling.
+ *
+ * Some functions are equivalent in name, signature, and behavior to the ANSI C <string.h>
+ * functions. (For example, they do not check for bad arguments like NULL string pointers.)
+ * In some cases, only the thread-safe variant of such a function is implemented here
+ * (see u_strtok_r()).
+ *
+ * Other functions provide more Unicode-specific functionality like locale-specific
+ * upper/lower-casing and string comparison in code point order.
+ *
+ * ICU uses 16-bit Unicode (UTF-16) in the form of arrays of UChar code units.
+ * UTF-16 encodes each Unicode code point with either one or two UChar code units.
+ * (This is the default form of Unicode, and a forward-compatible extension of the original,
+ * fixed-width form that was known as UCS-2. UTF-16 superseded UCS-2 with Unicode 2.0
+ * in 1996.)
+ *
+ * Some APIs accept a 32-bit UChar32 value for a single code point.
+ *
+ * ICU also handles 16-bit Unicode text with unpaired surrogates.
+ * Such text is not well-formed UTF-16.
+ * Code-point-related functions treat unpaired surrogates as surrogate code points,
+ * i.e., as separate units.
+ *
+ * Although UTF-16 is a variable-width encoding form (like some legacy multi-byte encodings),
+ * it is much more efficient even for random access because the code unit values
+ * for single-unit characters vs. lead units vs. trail units are completely disjoint.
+ * This means that it is easy to determine character (code point) boundaries from
+ * random offsets in the string.
+ *
+ * Unicode (UTF-16) string processing is optimized for the single-unit case.
+ * Although it is important to support supplementary characters
+ * (which use pairs of lead/trail code units called "surrogates"),
+ * their occurrence is rare. Almost all characters in modern use require only
+ * a single UChar code unit (i.e., their code point values are <=0xffff).
+ *
+ * For more details see the User Guide Strings chapter (http://icu-project.org/userguide/strings.html).
+ * For a discussion of the handling of unpaired surrogates see also
+ * Jitterbug 2145 and its icu mailing list proposal on 2002-sep-18.
+ */
+
+/**
+ * \defgroup ustring_ustrlen String Length
+ * \ingroup ustring_strlen
+ */
+/*@{*/
+/**
+ * Determine the length of an array of UChar.
+ *
+ * @param s The array of UChars, NULL (U+0000) terminated.
+ * @return The number of UChars in <code>chars</code>, minus the terminator.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strlen(const UChar *s);
+/*@}*/
+
+/**
+ * Count Unicode code points in the length UChar code units of the string.
+ * A code point may occupy either one or two UChar code units.
+ * Counting code points involves reading all code units.
+ *
+ * This functions is basically the inverse of the U16_FWD_N() macro (see utf.h).
+ *
+ * @param s The input string.
+ * @param length The number of UChar code units to be checked, or -1 to count all
+ *               code points before the first NUL (U+0000).
+ * @return The number of code points in the specified code units.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_countChar32(const UChar *s, int32_t length);
+
+/**
+ * Check if the string contains more Unicode code points than a certain number.
+ * This is more efficient than counting all code points in the entire string
+ * and comparing that number with a threshold.
+ * This function may not need to scan the string at all if the length is known
+ * (not -1 for NUL-termination) and falls within a certain range, and
+ * never needs to count more than 'number+1' code points.
+ * Logically equivalent to (u_countChar32(s, length)>number).
+ * A Unicode code point may occupy either one or two UChar code units.
+ *
+ * @param s The input string.
+ * @param length The length of the string, or -1 if it is NUL-terminated.
+ * @param number The number of code points in the string is compared against
+ *               the 'number' parameter.
+ * @return Boolean value for whether the string contains more Unicode code points
+ *         than 'number'. Same as (u_countChar32(s, length)>number).
+ * @stable ICU 2.4
+ */
+U_CAPI UBool U_EXPORT2
+u_strHasMoreChar32Than(const UChar *s, int32_t length, int32_t number);
+
+/**
+ * Concatenate two ustrings.  Appends a copy of <code>src</code>,
+ * including the null terminator, to <code>dst</code>. The initial copied
+ * character from <code>src</code> overwrites the null terminator in <code>dst</code>.
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_strcat(UChar     *dst, 
+    const UChar     *src);
+
+/**
+ * Concatenate two ustrings.  
+ * Appends at most <code>n</code> characters from <code>src</code> to <code>dst</code>.
+ * Adds a terminating NUL.
+ * If src is too long, then only <code>n-1</code> characters will be copied
+ * before the terminating NUL.
+ * If <code>n&lt;=0</code> then dst is not modified.
+ *
+ * @param dst The destination string.
+ * @param src The source string (can be NULL/invalid if n<=0).
+ * @param n The maximum number of characters to append; no-op if <=0.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_strncat(UChar     *dst, 
+     const UChar     *src, 
+     int32_t     n);
+
+/**
+ * Find the first occurrence of a substring in a string.
+ * The substring is found at code point boundaries.
+ * That means that if the substring begins with
+ * a trail surrogate or ends with a lead surrogate,
+ * then it is found only if these surrogates stand alone in the text.
+ * Otherwise, the substring edge units would be matched against
+ * halves of surrogate pairs.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param substring The substring to find (NUL-terminated).
+ * @return A pointer to the first occurrence of <code>substring</code> in <code>s</code>,
+ *         or <code>s</code> itself if the <code>substring</code> is empty,
+ *         or <code>NULL</code> if <code>substring</code> is not in <code>s</code>.
+ * @stable ICU 2.0
+ *
+ * @see u_strrstr
+ * @see u_strFindFirst
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strstr(const UChar *s, const UChar *substring);
+
+/**
+ * Find the first occurrence of a substring in a string.
+ * The substring is found at code point boundaries.
+ * That means that if the substring begins with
+ * a trail surrogate or ends with a lead surrogate,
+ * then it is found only if these surrogates stand alone in the text.
+ * Otherwise, the substring edge units would be matched against
+ * halves of surrogate pairs.
+ *
+ * @param s The string to search.
+ * @param length The length of s (number of UChars), or -1 if it is NUL-terminated.
+ * @param substring The substring to find (NUL-terminated).
+ * @param subLength The length of substring (number of UChars), or -1 if it is NUL-terminated.
+ * @return A pointer to the first occurrence of <code>substring</code> in <code>s</code>,
+ *         or <code>s</code> itself if the <code>substring</code> is empty,
+ *         or <code>NULL</code> if <code>substring</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strstr
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strFindFirst(const UChar *s, int32_t length, const UChar *substring, int32_t subLength);
+
+/**
+ * Find the first occurrence of a BMP code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param c The BMP code point to find.
+ * @return A pointer to the first occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.0
+ *
+ * @see u_strchr32
+ * @see u_memchr
+ * @see u_strstr
+ * @see u_strFindFirst
+ */
+U_CAPI UChar * U_EXPORT2
+u_strchr(const UChar *s, UChar c);
+
+/**
+ * Find the first occurrence of a code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param c The code point to find.
+ * @return A pointer to the first occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.0
+ *
+ * @see u_strchr
+ * @see u_memchr32
+ * @see u_strstr
+ * @see u_strFindFirst
+ */
+U_CAPI UChar * U_EXPORT2
+u_strchr32(const UChar *s, UChar32 c);
+
+/**
+ * Find the last occurrence of a substring in a string.
+ * The substring is found at code point boundaries.
+ * That means that if the substring begins with
+ * a trail surrogate or ends with a lead surrogate,
+ * then it is found only if these surrogates stand alone in the text.
+ * Otherwise, the substring edge units would be matched against
+ * halves of surrogate pairs.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param substring The substring to find (NUL-terminated).
+ * @return A pointer to the last occurrence of <code>substring</code> in <code>s</code>,
+ *         or <code>s</code> itself if the <code>substring</code> is empty,
+ *         or <code>NULL</code> if <code>substring</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strstr
+ * @see u_strFindFirst
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strrstr(const UChar *s, const UChar *substring);
+
+/**
+ * Find the last occurrence of a substring in a string.
+ * The substring is found at code point boundaries.
+ * That means that if the substring begins with
+ * a trail surrogate or ends with a lead surrogate,
+ * then it is found only if these surrogates stand alone in the text.
+ * Otherwise, the substring edge units would be matched against
+ * halves of surrogate pairs.
+ *
+ * @param s The string to search.
+ * @param length The length of s (number of UChars), or -1 if it is NUL-terminated.
+ * @param substring The substring to find (NUL-terminated).
+ * @param subLength The length of substring (number of UChars), or -1 if it is NUL-terminated.
+ * @return A pointer to the last occurrence of <code>substring</code> in <code>s</code>,
+ *         or <code>s</code> itself if the <code>substring</code> is empty,
+ *         or <code>NULL</code> if <code>substring</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strstr
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strFindLast(const UChar *s, int32_t length, const UChar *substring, int32_t subLength);
+
+/**
+ * Find the last occurrence of a BMP code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param c The BMP code point to find.
+ * @return A pointer to the last occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strrchr32
+ * @see u_memrchr
+ * @see u_strrstr
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strrchr(const UChar *s, UChar c);
+
+/**
+ * Find the last occurrence of a code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (NUL-terminated).
+ * @param c The code point to find.
+ * @return A pointer to the last occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strrchr
+ * @see u_memchr32
+ * @see u_strrstr
+ * @see u_strFindLast
+ */
+U_CAPI UChar * U_EXPORT2
+u_strrchr32(const UChar *s, UChar32 c);
+
+/**
+ * Locates the first occurrence in the string <code>string</code> of any of the characters
+ * in the string <code>matchSet</code>.
+ * Works just like C's strpbrk but with Unicode.
+ *
+ * @param string The string in which to search, NUL-terminated.
+ * @param matchSet A NUL-terminated string defining a set of code points
+ *                 for which to search in the text string.
+ * @return A pointer to the  character in <code>string</code> that matches one of the
+ *         characters in <code>matchSet</code>, or NULL if no such character is found.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar * U_EXPORT2
+u_strpbrk(const UChar *string, const UChar *matchSet);
+
+/**
+ * Returns the number of consecutive characters in <code>string</code>,
+ * beginning with the first, that do not occur somewhere in <code>matchSet</code>.
+ * Works just like C's strcspn but with Unicode.
+ *
+ * @param string The string in which to search, NUL-terminated.
+ * @param matchSet A NUL-terminated string defining a set of code points
+ *                 for which to search in the text string.
+ * @return The number of initial characters in <code>string</code> that do not
+ *         occur in <code>matchSet</code>.
+ * @see u_strspn
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strcspn(const UChar *string, const UChar *matchSet);
+
+/**
+ * Returns the number of consecutive characters in <code>string</code>,
+ * beginning with the first, that occur somewhere in <code>matchSet</code>.
+ * Works just like C's strspn but with Unicode.
+ *
+ * @param string The string in which to search, NUL-terminated.
+ * @param matchSet A NUL-terminated string defining a set of code points
+ *                 for which to search in the text string.
+ * @return The number of initial characters in <code>string</code> that do
+ *         occur in <code>matchSet</code>.
+ * @see u_strcspn
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strspn(const UChar *string, const UChar *matchSet);
+
+/**
+ * The string tokenizer API allows an application to break a string into
+ * tokens. Unlike strtok(), the saveState (the current pointer within the
+ * original string) is maintained in saveState. In the first call, the
+ * argument src is a pointer to the string. In subsequent calls to
+ * return successive tokens of that string, src must be specified as
+ * NULL. The value saveState is set by this function to maintain the
+ * function's position within the string, and on each subsequent call
+ * you must give this argument the same variable. This function does
+ * handle surrogate pairs. This function is similar to the strtok_r()
+ * the POSIX Threads Extension (1003.1c-1995) version.
+ *
+ * @param src String containing token(s). This string will be modified.
+ *            After the first call to u_strtok_r(), this argument must
+ *            be NULL to get to the next token.
+ * @param delim Set of delimiter characters (Unicode code points).
+ * @param saveState The current pointer within the original string,
+ *              which is set by this function. The saveState
+ *              parameter should the address of a local variable of type
+ *              UChar *. (i.e. defined "UChar *myLocalSaveState" and use
+ *              &myLocalSaveState for this parameter).
+ * @return A pointer to the next token found in src, or NULL
+ *         when there are no more tokens.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar * U_EXPORT2
+u_strtok_r(UChar    *src, 
+     const UChar    *delim,
+           UChar   **saveState);
+
+/**
+ * Compare two Unicode strings for bitwise equality (code unit order).
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @return 0 if <code>s1</code> and <code>s2</code> are bitwise equal; a negative
+ * value if <code>s1</code> is bitwise less than <code>s2,</code>; a positive
+ * value if <code>s1</code> is bitwise greater than <code>s2</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t  U_EXPORT2
+u_strcmp(const UChar     *s1, 
+         const UChar     *s2);
+
+/**
+ * Compare two Unicode strings in code point order.
+ * See u_strCompare for details.
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @return a negative/zero/positive integer corresponding to whether
+ * the first string is less than/equal to/greater than the second one
+ * in code point order
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strcmpCodePointOrder(const UChar *s1, const UChar *s2);
+
+/**
+ * Compare two Unicode strings (binary order).
+ *
+ * The comparison can be done in code unit order or in code point order.
+ * They differ only in UTF-16 when
+ * comparing supplementary code points (U+10000..U+10ffff)
+ * to BMP code points near the end of the BMP (i.e., U+e000..U+ffff).
+ * In code unit order, high BMP code points sort after supplementary code points
+ * because they are stored as pairs of surrogates which are at U+d800..U+dfff.
+ *
+ * This functions works with strings of different explicitly specified lengths
+ * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
+ * NUL-terminated strings are possible with length arguments of -1.
+ *
+ * @param s1 First source string.
+ * @param length1 Length of first source string, or -1 if NUL-terminated.
+ *
+ * @param s2 Second source string.
+ * @param length2 Length of second source string, or -1 if NUL-terminated.
+ *
+ * @param codePointOrder Choose between code unit order (false)
+ *                       and code point order (true).
+ *
+ * @return <0 or 0 or >0 as usual for string comparisons
+ *
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_strCompare(const UChar *s1, int32_t length1,
+             const UChar *s2, int32_t length2,
+             UBool codePointOrder);
+
+/**
+ * Compare two Unicode strings (binary order)
+ * as presented by UCharIterator objects.
+ * Works otherwise just like u_strCompare().
+ *
+ * Both iterators are reset to their start positions.
+ * When the function returns, it is undefined where the iterators
+ * have stopped.
+ *
+ * @param iter1 First source string iterator.
+ * @param iter2 Second source string iterator.
+ * @param codePointOrder Choose between code unit order (false)
+ *                       and code point order (true).
+ *
+ * @return <0 or 0 or >0 as usual for string comparisons
+ *
+ * @see u_strCompare
+ *
+ * @stable ICU 2.6
+ */
+U_CAPI int32_t U_EXPORT2
+u_strCompareIter(UCharIterator *iter1, UCharIterator *iter2, UBool codePointOrder);
+
+/**
+ * Compare two strings case-insensitively using full case folding.
+ * This is equivalent to
+ *   u_strCompare(u_strFoldCase(s1, options),
+ *                u_strFoldCase(s2, options),
+ *                (options&U_COMPARE_CODE_POINT_ORDER)!=0).
+ *
+ * The comparison can be done in UTF-16 code unit order or in code point order.
+ * They differ only when comparing supplementary code points (U+10000..U+10ffff)
+ * to BMP code points near the end of the BMP (i.e., U+e000..U+ffff).
+ * In code unit order, high BMP code points sort after supplementary code points
+ * because they are stored as pairs of surrogates which are at U+d800..U+dfff.
+ *
+ * This functions works with strings of different explicitly specified lengths
+ * unlike the ANSI C-like u_strcmp() and u_memcmp() etc.
+ * NUL-terminated strings are possible with length arguments of -1.
+ *
+ * @param s1 First source string.
+ * @param length1 Length of first source string, or -1 if NUL-terminated.
+ *
+ * @param s2 Second source string.
+ * @param length2 Length of second source string, or -1 if NUL-terminated.
+ *
+ * @param options A bit set of options:
+ *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+ *     Comparison in code unit order with default case folding.
+ *
+ *   - U_COMPARE_CODE_POINT_ORDER
+ *     Set to choose code point order instead of code unit order
+ *     (see u_strCompare for details).
+ *
+ *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ *
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ *
+ * @return <0 or 0 or >0 as usual for string comparisons
+ *
+ * @stable ICU 2.2
+ */
+U_CAPI int32_t U_EXPORT2
+u_strCaseCompare(const UChar *s1, int32_t length1,
+                 const UChar *s2, int32_t length2,
+                 uint32_t options,
+                 UErrorCode *pErrorCode);
+
+/**
+ * Compare two ustrings for bitwise equality. 
+ * Compares at most <code>n</code> characters.
+ *
+ * @param ucs1 A string to compare (can be NULL/invalid if n<=0).
+ * @param ucs2 A string to compare (can be NULL/invalid if n<=0).
+ * @param n The maximum number of characters to compare; always returns 0 if n<=0.
+ * @return 0 if <code>s1</code> and <code>s2</code> are bitwise equal; a negative
+ * value if <code>s1</code> is bitwise less than <code>s2</code>; a positive
+ * value if <code>s1</code> is bitwise greater than <code>s2</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strncmp(const UChar     *ucs1, 
+     const UChar     *ucs2, 
+     int32_t     n);
+
+/**
+ * Compare two Unicode strings in code point order.
+ * This is different in UTF-16 from u_strncmp() if supplementary characters are present.
+ * For details, see u_strCompare().
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @param n The maximum number of characters to compare.
+ * @return a negative/zero/positive integer corresponding to whether
+ * the first string is less than/equal to/greater than the second one
+ * in code point order
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strncmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t n);
+
+/**
+ * Compare two strings case-insensitively using full case folding.
+ * This is equivalent to u_strcmp(u_strFoldCase(s1, options), u_strFoldCase(s2, options)).
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @param options A bit set of options:
+ *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+ *     Comparison in code unit order with default case folding.
+ *
+ *   - U_COMPARE_CODE_POINT_ORDER
+ *     Set to choose code point order instead of code unit order
+ *     (see u_strCompare for details).
+ *
+ *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ *
+ * @return A negative, zero, or positive integer indicating the comparison result.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strcasecmp(const UChar *s1, const UChar *s2, uint32_t options);
+
+/**
+ * Compare two strings case-insensitively using full case folding.
+ * This is equivalent to u_strcmp(u_strFoldCase(s1, at most n, options),
+ * u_strFoldCase(s2, at most n, options)).
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @param n The maximum number of characters each string to case-fold and then compare.
+ * @param options A bit set of options:
+ *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+ *     Comparison in code unit order with default case folding.
+ *
+ *   - U_COMPARE_CODE_POINT_ORDER
+ *     Set to choose code point order instead of code unit order
+ *     (see u_strCompare for details).
+ *
+ *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ *
+ * @return A negative, zero, or positive integer indicating the comparison result.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strncasecmp(const UChar *s1, const UChar *s2, int32_t n, uint32_t options);
+
+/**
+ * Compare two strings case-insensitively using full case folding.
+ * This is equivalent to u_strcmp(u_strFoldCase(s1, n, options),
+ * u_strFoldCase(s2, n, options)).
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @param length The number of characters in each string to case-fold and then compare.
+ * @param options A bit set of options:
+ *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
+ *     Comparison in code unit order with default case folding.
+ *
+ *   - U_COMPARE_CODE_POINT_ORDER
+ *     Set to choose code point order instead of code unit order
+ *     (see u_strCompare for details).
+ *
+ *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ *
+ * @return A negative, zero, or positive integer indicating the comparison result.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_memcasecmp(const UChar *s1, const UChar *s2, int32_t length, uint32_t options);
+
+/**
+ * Copy a ustring. Adds a null terminator.
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_strcpy(UChar     *dst, 
+    const UChar     *src);
+
+/**
+ * Copy a ustring.
+ * Copies at most <code>n</code> characters.  The result will be null terminated
+ * if the length of <code>src</code> is less than <code>n</code>.
+ *
+ * @param dst The destination string.
+ * @param src The source string (can be NULL/invalid if n<=0).
+ * @param n The maximum number of characters to copy; no-op if <=0.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_strncpy(UChar     *dst, 
+     const UChar     *src, 
+     int32_t     n);
+
+#if !UCONFIG_NO_CONVERSION
+
+/**
+ * Copy a byte string encoded in the default codepage to a ustring.
+ * Adds a null terminator.
+ * Performs a host byte to UChar conversion
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2 u_uastrcpy(UChar *dst,
+               const char *src );
+
+/**
+ * Copy a byte string encoded in the default codepage to a ustring.
+ * Copies at most <code>n</code> characters.  The result will be null terminated
+ * if the length of <code>src</code> is less than <code>n</code>.
+ * Performs a host byte to UChar conversion
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @param n The maximum number of characters to copy.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2 u_uastrncpy(UChar *dst,
+            const char *src,
+            int32_t n);
+
+/**
+ * Copy ustring to a byte string encoded in the default codepage.
+ * Adds a null terminator.
+ * Performs a UChar to host byte conversion
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI char* U_EXPORT2 u_austrcpy(char *dst,
+            const UChar *src );
+
+/**
+ * Copy ustring to a byte string encoded in the default codepage.
+ * Copies at most <code>n</code> characters.  The result will be null terminated
+ * if the length of <code>src</code> is less than <code>n</code>.
+ * Performs a UChar to host byte conversion
+ *
+ * @param dst The destination string.
+ * @param src The source string.
+ * @param n The maximum number of characters to copy.
+ * @return A pointer to <code>dst</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI char* U_EXPORT2 u_austrncpy(char *dst,
+            const UChar *src,
+            int32_t n );
+
+#endif
+
+/**
+ * Synonym for memcpy(), but with UChars only.
+ * @param dest The destination string
+ * @param src The source string (can be NULL/invalid if count<=0)
+ * @param count The number of characters to copy; no-op if <=0
+ * @return A pointer to <code>dest</code>
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_memcpy(UChar *dest, const UChar *src, int32_t count);
+
+/**
+ * Synonym for memmove(), but with UChars only.
+ * @param dest The destination string
+ * @param src The source string (can be NULL/invalid if count<=0)
+ * @param count The number of characters to move; no-op if <=0
+ * @return A pointer to <code>dest</code>
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_memmove(UChar *dest, const UChar *src, int32_t count);
+
+/**
+ * Initialize <code>count</code> characters of <code>dest</code> to <code>c</code>.
+ *
+ * @param dest The destination string.
+ * @param c The character to initialize the string.
+ * @param count The maximum number of characters to set.
+ * @return A pointer to <code>dest</code>.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_memset(UChar *dest, UChar c, int32_t count);
+
+/**
+ * Compare the first <code>count</code> UChars of each buffer.
+ *
+ * @param buf1 The first string to compare.
+ * @param buf2 The second string to compare.
+ * @param count The maximum number of UChars to compare.
+ * @return When buf1 < buf2, a negative number is returned.
+ *      When buf1 == buf2, 0 is returned.
+ *      When buf1 > buf2, a positive number is returned.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_memcmp(const UChar *buf1, const UChar *buf2, int32_t count);
+
+/**
+ * Compare two Unicode strings in code point order.
+ * This is different in UTF-16 from u_memcmp() if supplementary characters are present.
+ * For details, see u_strCompare().
+ *
+ * @param s1 A string to compare.
+ * @param s2 A string to compare.
+ * @param count The maximum number of characters to compare.
+ * @return a negative/zero/positive integer corresponding to whether
+ * the first string is less than/equal to/greater than the second one
+ * in code point order
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_memcmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t count);
+
+/**
+ * Find the first occurrence of a BMP code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (contains <code>count</code> UChars).
+ * @param c The BMP code point to find.
+ * @param count The length of the string.
+ * @return A pointer to the first occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.0
+ *
+ * @see u_strchr
+ * @see u_memchr32
+ * @see u_strFindFirst
+ */
+U_CAPI UChar* U_EXPORT2
+u_memchr(const UChar *s, UChar c, int32_t count);
+
+/**
+ * Find the first occurrence of a code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (contains <code>count</code> UChars).
+ * @param c The code point to find.
+ * @param count The length of the string.
+ * @return A pointer to the first occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.0
+ *
+ * @see u_strchr32
+ * @see u_memchr
+ * @see u_strFindFirst
+ */
+U_CAPI UChar* U_EXPORT2
+u_memchr32(const UChar *s, UChar32 c, int32_t count);
+
+/**
+ * Find the last occurrence of a BMP code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (contains <code>count</code> UChars).
+ * @param c The BMP code point to find.
+ * @param count The length of the string.
+ * @return A pointer to the last occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strrchr
+ * @see u_memrchr32
+ * @see u_strFindLast
+ */
+U_CAPI UChar* U_EXPORT2
+u_memrchr(const UChar *s, UChar c, int32_t count);
+
+/**
+ * Find the last occurrence of a code point in a string.
+ * A surrogate code point is found only if its match in the text is not
+ * part of a surrogate pair.
+ * A NUL character is found at the string terminator.
+ *
+ * @param s The string to search (contains <code>count</code> UChars).
+ * @param c The code point to find.
+ * @param count The length of the string.
+ * @return A pointer to the last occurrence of <code>c</code> in <code>s</code>
+ *         or <code>NULL</code> if <code>c</code> is not in <code>s</code>.
+ * @stable ICU 2.4
+ *
+ * @see u_strrchr32
+ * @see u_memrchr
+ * @see u_strFindLast
+ */
+U_CAPI UChar* U_EXPORT2
+u_memrchr32(const UChar *s, UChar32 c, int32_t count);
+
+/**
+ * Unicode String literals in C.
+ * We need one macro to declare a variable for the string
+ * and to statically preinitialize it if possible,
+ * and a second macro to dynamically initialize such a string variable if necessary.
+ *
+ * The macros are defined for maximum performance.
+ * They work only for strings that contain "invariant characters", i.e.,
+ * only latin letters, digits, and some punctuation.
+ * See utypes.h for details.
+ *
+ * A pair of macros for a single string must be used with the same
+ * parameters.
+ * The string parameter must be a C string literal.
+ * The length of the string, not including the terminating
+ * `NUL`, must be specified as a constant.
+ * The U_STRING_DECL macro should be invoked exactly once for one
+ * such string variable before it is used.
+ *
+ * Usage:
+ *
+ *     U_STRING_DECL(ustringVar1, "Quick-Fox 2", 11);
+ *     U_STRING_DECL(ustringVar2, "jumps 5%", 8);
+ *     static UBool didInit=false;
+ *
+ *     int32_t function() {
+ *         if(!didInit) {
+ *             U_STRING_INIT(ustringVar1, "Quick-Fox 2", 11);
+ *             U_STRING_INIT(ustringVar2, "jumps 5%", 8);
+ *             didInit=true;
+ *         }
+ *         return u_strcmp(ustringVar1, ustringVar2);
+ *     }
+ * 
+ * Note that the macros will NOT consistently work if their argument is another #`define`.
+ * The following will not work on all platforms, don't use it.
+ * 
+ *     #define GLUCK "Mr. Gluck"
+ *     U_STRING_DECL(var, GLUCK, 9)
+ *     U_STRING_INIT(var, GLUCK, 9)
+ *
+ * Instead, use the string literal "Mr. Gluck"  as the argument to both macro
+ * calls.
+ *
+ *
+ * @stable ICU 2.0
+ */
+#if defined(U_DECLARE_UTF16)
+#   define U_STRING_DECL(var, cs, length) static const UChar *var=(const UChar *)U_DECLARE_UTF16(cs)
+    /**@stable ICU 2.0 */
+#   define U_STRING_INIT(var, cs, length)
+#elif U_SIZEOF_WCHAR_T==U_SIZEOF_UCHAR && (U_CHARSET_FAMILY==U_ASCII_FAMILY || (U_SIZEOF_UCHAR == 2 && defined(U_WCHAR_IS_UTF16)))
+#   define U_STRING_DECL(var, cs, length) static const UChar var[(length)+1]=L ## cs
+    /**@stable ICU 2.0 */
+#   define U_STRING_INIT(var, cs, length)
+#elif U_SIZEOF_UCHAR==1 && U_CHARSET_FAMILY==U_ASCII_FAMILY
+#   define U_STRING_DECL(var, cs, length) static const UChar var[(length)+1]=cs
+    /**@stable ICU 2.0 */
+#   define U_STRING_INIT(var, cs, length)
+#else
+#   define U_STRING_DECL(var, cs, length) static UChar var[(length)+1]
+    /**@stable ICU 2.0 */
+#   define U_STRING_INIT(var, cs, length) u_charsToUChars(cs, var, length+1)
+#endif
+
+/**
+ * Unescape a string of characters and write the resulting
+ * Unicode characters to the destination buffer.  The following escape
+ * sequences are recognized:
+ *
+ * \\uhhhh       4 hex digits; h in [0-9A-Fa-f]
+ * \\Uhhhhhhhh   8 hex digits
+ * \\xhh         1-2 hex digits
+ * \\x{h...}     1-8 hex digits
+ * \\ooo         1-3 octal digits; o in [0-7]
+ * \\cX          control-X; X is masked with 0x1F
+ *
+ * as well as the standard ANSI C escapes:
+ *
+ * \\a => U+0007, \\b => U+0008, \\t => U+0009, \\n => U+000A,
+ * \\v => U+000B, \\f => U+000C, \\r => U+000D, \\e => U+001B,
+ * \\&quot; => U+0022, \\' => U+0027, \\? => U+003F, \\\\ => U+005C
+ *
+ * Anything else following a backslash is generically escaped.  For
+ * example, "[a\\-z]" returns "[a-z]".
+ *
+ * If an escape sequence is ill-formed, this method returns an empty
+ * string.  An example of an ill-formed sequence is "\\u" followed by
+ * fewer than 4 hex digits.
+ *
+ * The above characters are recognized in the compiler's codepage,
+ * that is, they are coded as 'u', '\\', etc.  Characters that are
+ * not parts of escape sequences are converted using u_charsToUChars().
+ *
+ * This function is similar to UnicodeString::unescape() but not
+ * identical to it.  The latter takes a source UnicodeString, so it
+ * does escape recognition but no conversion.
+ *
+ * @param src a zero-terminated string of invariant characters
+ * @param dest pointer to buffer to receive converted and unescaped
+ * text and, if there is room, a zero terminator.  May be NULL for
+ * preflighting, in which case no UChars will be written, but the
+ * return value will still be valid.  On error, an empty string is
+ * stored here (if possible).
+ * @param destCapacity the number of UChars that may be written at
+ * dest.  Ignored if dest == NULL.
+ * @return the length of unescaped string.
+ * @see u_unescapeAt
+ * @see UnicodeString#unescape()
+ * @see UnicodeString#unescapeAt()
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_unescape(const char *src,
+           UChar *dest, int32_t destCapacity);
+
+U_CDECL_BEGIN
+/**
+ * Callback function for u_unescapeAt() that returns a character of
+ * the source text given an offset and a context pointer.  The context
+ * pointer will be whatever is passed into u_unescapeAt().
+ *
+ * @param offset pointer to the offset that will be passed to u_unescapeAt().
+ * @param context an opaque pointer passed directly into u_unescapeAt()
+ * @return the character represented by the escape sequence at
+ * offset
+ * @see u_unescapeAt
+ * @stable ICU 2.0
+ */
+typedef UChar (U_CALLCONV *UNESCAPE_CHAR_AT)(int32_t offset, void *context);
+U_CDECL_END
+
+/**
+ * Unescape a single sequence. The character at offset-1 is assumed
+ * (without checking) to be a backslash.  This method takes a callback
+ * pointer to a function that returns the UChar at a given offset.  By
+ * varying this callback, ICU functions are able to unescape char*
+ * strings, UnicodeString objects, and UFILE pointers.
+ *
+ * If offset is out of range, or if the escape sequence is ill-formed,
+ * (UChar32)0xFFFFFFFF is returned.  See documentation of u_unescape()
+ * for a list of recognized sequences.
+ *
+ * @param charAt callback function that returns a UChar of the source
+ * text given an offset and a context pointer.
+ * @param offset pointer to the offset that will be passed to charAt.
+ * The offset value will be updated upon return to point after the
+ * last parsed character of the escape sequence.  On error the offset
+ * is unchanged.
+ * @param length the number of characters in the source text.  The
+ * last character of the source text is considered to be at offset
+ * length-1.
+ * @param context an opaque pointer passed directly into charAt.
+ * @return the character represented by the escape sequence at
+ * offset, or (UChar32)0xFFFFFFFF on error.
+ * @see u_unescape()
+ * @see UnicodeString#unescape()
+ * @see UnicodeString#unescapeAt()
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32 U_EXPORT2
+u_unescapeAt(UNESCAPE_CHAR_AT charAt,
+             int32_t *offset,
+             int32_t length,
+             void *context);
+
+/**
+ * Uppercase the characters in a string.
+ * Casing is locale-dependent and context-sensitive.
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer are allowed to overlap.
+ *
+ * @param dest      A buffer for the result string. The result will be zero-terminated if
+ *                  the buffer is large enough.
+ * @param destCapacity The size of the buffer (number of UChars). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string
+ * @param srcLength The length of the original string. If -1, then src must be zero-terminated.
+ * @param locale    The locale to consider, or "" for the root locale or NULL for the default locale.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string. It may be greater than destCapacity. In that case,
+ *         only some of the result was written to the destination buffer.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strToUpper(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             const char *locale,
+             UErrorCode *pErrorCode);
+
+/**
+ * Lowercase the characters in a string.
+ * Casing is locale-dependent and context-sensitive.
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer are allowed to overlap.
+ *
+ * @param dest      A buffer for the result string. The result will be zero-terminated if
+ *                  the buffer is large enough.
+ * @param destCapacity The size of the buffer (number of UChars). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string
+ * @param srcLength The length of the original string. If -1, then src must be zero-terminated.
+ * @param locale    The locale to consider, or "" for the root locale or NULL for the default locale.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string. It may be greater than destCapacity. In that case,
+ *         only some of the result was written to the destination buffer.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strToLower(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             const char *locale,
+             UErrorCode *pErrorCode);
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+/**
+ * Titlecase a string.
+ * Casing is locale-dependent and context-sensitive.
+ * Titlecasing uses a break iterator to find the first characters of words
+ * that are to be titlecased. It titlecases those characters and lowercases
+ * all others.
+ *
+ * The titlecase break iterator can be provided to customize for arbitrary
+ * styles, using rules and dictionaries beyond the standard iterators.
+ * It may be more efficient to always provide an iterator to avoid
+ * opening and closing one for each string.
+ * The standard titlecase iterator for the root locale implements the
+ * algorithm of Unicode TR 21.
+ *
+ * This function uses only the setText(), first() and next() methods of the
+ * provided break iterator.
+ *
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer are allowed to overlap.
+ *
+ * @param dest      A buffer for the result string. The result will be zero-terminated if
+ *                  the buffer is large enough.
+ * @param destCapacity The size of the buffer (number of UChars). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string
+ * @param srcLength The length of the original string. If -1, then src must be zero-terminated.
+ * @param titleIter A break iterator to find the first characters of words
+ *                  that are to be titlecased.
+ *                  If none is provided (NULL), then a standard titlecase
+ *                  break iterator is opened.
+ * @param locale    The locale to consider, or "" for the root locale or NULL for the default locale.
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string. It may be greater than destCapacity. In that case,
+ *         only some of the result was written to the destination buffer.
+ * @stable ICU 2.1
+ */
+U_CAPI int32_t U_EXPORT2
+u_strToTitle(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             UBreakIterator *titleIter,
+             const char *locale,
+             UErrorCode *pErrorCode);
+
+#endif
+
+/**
+ * Case-folds the characters in a string.
+ *
+ * Case-folding is locale-independent and not context-sensitive,
+ * but there is an option for whether to include or exclude mappings for dotted I
+ * and dotless i that are marked with 'T' in CaseFolding.txt.
+ *
+ * The result may be longer or shorter than the original.
+ * The source string and the destination buffer are allowed to overlap.
+ *
+ * @param dest      A buffer for the result string. The result will be zero-terminated if
+ *                  the buffer is large enough.
+ * @param destCapacity The size of the buffer (number of UChars). If it is 0, then
+ *                  dest may be NULL and the function will only return the length of the result
+ *                  without writing any of the result string.
+ * @param src       The original string
+ * @param srcLength The length of the original string. If -1, then src must be zero-terminated.
+ * @param options   Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
+ * @param pErrorCode Must be a valid pointer to an error code value,
+ *                  which must not indicate a failure before the function call.
+ * @return The length of the result string. It may be greater than destCapacity. In that case,
+ *         only some of the result was written to the destination buffer.
+ * @stable ICU 2.0
+ */
+U_CAPI int32_t U_EXPORT2
+u_strFoldCase(UChar *dest, int32_t destCapacity,
+              const UChar *src, int32_t srcLength,
+              uint32_t options,
+              UErrorCode *pErrorCode);
+
+#if defined(U_WCHAR_IS_UTF16) || defined(U_WCHAR_IS_UTF32) || !UCONFIG_NO_CONVERSION
+/**
+ * Convert a UTF-16 string to a wchar_t string.
+ * If it is known at compile time that wchar_t strings are in UTF-16 or UTF-32, then
+ * this function simply calls the fast, dedicated function for that.
+ * Otherwise, two conversions UTF-16 -> default charset -> wchar_t* are performed.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of wchar_t's). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @stable ICU 2.0
+ */
+U_CAPI wchar_t* U_EXPORT2
+u_strToWCS(wchar_t *dest, 
+           int32_t destCapacity,
+           int32_t *pDestLength,
+           const UChar *src, 
+           int32_t srcLength,
+           UErrorCode *pErrorCode);
+/**
+ * Convert a wchar_t string to UTF-16.
+ * If it is known at compile time that wchar_t strings are in UTF-16 or UTF-32, then
+ * this function simply calls the fast, dedicated function for that.
+ * Otherwise, two conversions wchar_t* -> default charset -> UTF-16 are performed.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2
+u_strFromWCS(UChar   *dest,
+             int32_t destCapacity, 
+             int32_t *pDestLength,
+             const wchar_t *src,
+             int32_t srcLength,
+             UErrorCode *pErrorCode);
+#endif /* defined(U_WCHAR_IS_UTF16) || defined(U_WCHAR_IS_UTF32) || !UCONFIG_NO_CONVERSION */
+
+/**
+ * Convert a UTF-16 string to UTF-8.
+ * If the input string is not well-formed, then the U_INVALID_CHAR_FOUND error code is set.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of chars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @stable ICU 2.0
+ * @see u_strToUTF8WithSub
+ * @see u_strFromUTF8
+ */
+U_CAPI char* U_EXPORT2 
+u_strToUTF8(char *dest,           
+            int32_t destCapacity,
+            int32_t *pDestLength,
+            const UChar *src, 
+            int32_t srcLength,
+            UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-8 string to UTF-16.
+ * If the input string is not well-formed, then the U_INVALID_CHAR_FOUND error code is set.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @stable ICU 2.0
+ * @see u_strFromUTF8WithSub
+ * @see u_strFromUTF8Lenient
+ */
+U_CAPI UChar* U_EXPORT2
+u_strFromUTF8(UChar *dest,             
+              int32_t destCapacity,
+              int32_t *pDestLength,
+              const char *src, 
+              int32_t srcLength,
+              UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-16 string to UTF-8.
+ *
+ * Same as u_strToUTF8() except for the additional subchar which is output for
+ * illegal input sequences, instead of stopping with the U_INVALID_CHAR_FOUND error code.
+ * With subchar==U_SENTINEL, this function behaves exactly like u_strToUTF8().
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of chars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param subchar       The substitution character to use in place of an illegal input sequence,
+ *                      or U_SENTINEL if the function is to return with U_INVALID_CHAR_FOUND instead.
+ *                      A substitution character can be any valid Unicode code point (up to U+10FFFF)
+ *                      except for surrogate code points (U+D800..U+DFFF).
+ *                      The recommended value is U+FFFD "REPLACEMENT CHARACTER".
+ * @param pNumSubstitutions Output parameter receiving the number of substitutions if subchar>=0.
+ *                      Set to 0 if no substitutions occur or subchar<0.
+ *                      pNumSubstitutions can be NULL.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strToUTF8
+ * @see u_strFromUTF8WithSub
+ * @stable ICU 3.6
+ */
+U_CAPI char* U_EXPORT2
+u_strToUTF8WithSub(char *dest,
+            int32_t destCapacity,
+            int32_t *pDestLength,
+            const UChar *src,
+            int32_t srcLength,
+            UChar32 subchar, int32_t *pNumSubstitutions,
+            UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-8 string to UTF-16.
+ *
+ * Same as u_strFromUTF8() except for the additional subchar which is output for
+ * illegal input sequences, instead of stopping with the U_INVALID_CHAR_FOUND error code.
+ * With subchar==U_SENTINEL, this function behaves exactly like u_strFromUTF8().
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param subchar       The substitution character to use in place of an illegal input sequence,
+ *                      or U_SENTINEL if the function is to return with U_INVALID_CHAR_FOUND instead.
+ *                      A substitution character can be any valid Unicode code point (up to U+10FFFF)
+ *                      except for surrogate code points (U+D800..U+DFFF).
+ *                      The recommended value is U+FFFD "REPLACEMENT CHARACTER".
+ * @param pNumSubstitutions Output parameter receiving the number of substitutions if subchar>=0.
+ *                      Set to 0 if no substitutions occur or subchar<0.
+ *                      pNumSubstitutions can be NULL.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strFromUTF8
+ * @see u_strFromUTF8Lenient
+ * @see u_strToUTF8WithSub
+ * @stable ICU 3.6
+ */
+U_CAPI UChar* U_EXPORT2
+u_strFromUTF8WithSub(UChar *dest,
+              int32_t destCapacity,
+              int32_t *pDestLength,
+              const char *src,
+              int32_t srcLength,
+              UChar32 subchar, int32_t *pNumSubstitutions,
+              UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-8 string to UTF-16.
+ *
+ * Same as u_strFromUTF8() except that this function is designed to be very fast,
+ * which it achieves by being lenient about malformed UTF-8 sequences.
+ * This function is intended for use in environments where UTF-8 text is
+ * expected to be well-formed.
+ *
+ * Its semantics are:
+ * - Well-formed UTF-8 text is correctly converted to well-formed UTF-16 text.
+ * - The function will not read beyond the input string, nor write beyond
+ *   the destCapacity.
+ * - Malformed UTF-8 results in "garbage" 16-bit Unicode strings which may not
+ *   be well-formed UTF-16.
+ *   The function will resynchronize to valid code point boundaries
+ *   within a small number of code points after an illegal sequence.
+ * - Non-shortest forms are not detected and will result in "spoofing" output.
+ *
+ * For further performance improvement, if srcLength is given (>=0),
+ * then it must be destCapacity>=srcLength.
+ *
+ * There is no inverse u_strToUTF8Lenient() function because there is practically
+ * no performance gain from not checking that a UTF-16 string is well-formed.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ *                      Unlike for other ICU functions, if srcLength>=0 then it
+ *                      must be destCapacity>=srcLength.
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ *                      Unlike for other ICU functions, if srcLength>=0 but
+ *                      destCapacity<srcLength, then *pDestLength will be set to srcLength
+ *                      (and U_BUFFER_OVERFLOW_ERROR will be set)
+ *                      regardless of the actual result length.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strFromUTF8
+ * @see u_strFromUTF8WithSub
+ * @see u_strToUTF8WithSub
+ * @stable ICU 3.6
+ */
+U_CAPI UChar * U_EXPORT2
+u_strFromUTF8Lenient(UChar *dest,
+                     int32_t destCapacity,
+                     int32_t *pDestLength,
+                     const char *src,
+                     int32_t srcLength,
+                     UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-16 string to UTF-32.
+ * If the input string is not well-formed, then the U_INVALID_CHAR_FOUND error code is set.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChar32s). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @see u_strToUTF32WithSub
+ * @see u_strFromUTF32
+ * @stable ICU 2.0
+ */
+U_CAPI UChar32* U_EXPORT2 
+u_strToUTF32(UChar32 *dest, 
+             int32_t  destCapacity,
+             int32_t  *pDestLength,
+             const UChar *src, 
+             int32_t  srcLength,
+             UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-32 string to UTF-16.
+ * If the input string is not well-formed, then the U_INVALID_CHAR_FOUND error code is set.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Must be a valid pointer to an error code value,
+ *                      which must not indicate a failure before the function call.
+ * @return The pointer to destination buffer.
+ * @see u_strFromUTF32WithSub
+ * @see u_strToUTF32
+ * @stable ICU 2.0
+ */
+U_CAPI UChar* U_EXPORT2 
+u_strFromUTF32(UChar   *dest,
+               int32_t destCapacity, 
+               int32_t *pDestLength,
+               const UChar32 *src,
+               int32_t srcLength,
+               UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-16 string to UTF-32.
+ *
+ * Same as u_strToUTF32() except for the additional subchar which is output for
+ * illegal input sequences, instead of stopping with the U_INVALID_CHAR_FOUND error code.
+ * With subchar==U_SENTINEL, this function behaves exactly like u_strToUTF32().
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChar32s). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If
+ *                      pDestLength!=NULL then *pDestLength is always set to the
+ *                      number of output units corresponding to the transformation of
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param subchar       The substitution character to use in place of an illegal input sequence,
+ *                      or U_SENTINEL if the function is to return with U_INVALID_CHAR_FOUND instead.
+ *                      A substitution character can be any valid Unicode code point (up to U+10FFFF)
+ *                      except for surrogate code points (U+D800..U+DFFF).
+ *                      The recommended value is U+FFFD "REPLACEMENT CHARACTER".
+ * @param pNumSubstitutions Output parameter receiving the number of substitutions if subchar>=0.
+ *                      Set to 0 if no substitutions occur or subchar<0.
+ *                      pNumSubstitutions can be NULL.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strToUTF32
+ * @see u_strFromUTF32WithSub
+ * @stable ICU 4.2
+ */
+U_CAPI UChar32* U_EXPORT2
+u_strToUTF32WithSub(UChar32 *dest,
+             int32_t destCapacity,
+             int32_t *pDestLength,
+             const UChar *src,
+             int32_t srcLength,
+             UChar32 subchar, int32_t *pNumSubstitutions,
+             UErrorCode *pErrorCode);
+
+/**
+ * Convert a UTF-32 string to UTF-16.
+ *
+ * Same as u_strFromUTF32() except for the additional subchar which is output for
+ * illegal input sequences, instead of stopping with the U_INVALID_CHAR_FOUND error code.
+ * With subchar==U_SENTINEL, this function behaves exactly like u_strFromUTF32().
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If
+ *                      pDestLength!=NULL then *pDestLength is always set to the
+ *                      number of output units corresponding to the transformation of
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param subchar       The substitution character to use in place of an illegal input sequence,
+ *                      or U_SENTINEL if the function is to return with U_INVALID_CHAR_FOUND instead.
+ *                      A substitution character can be any valid Unicode code point (up to U+10FFFF)
+ *                      except for surrogate code points (U+D800..U+DFFF).
+ *                      The recommended value is U+FFFD "REPLACEMENT CHARACTER".
+ * @param pNumSubstitutions Output parameter receiving the number of substitutions if subchar>=0.
+ *                      Set to 0 if no substitutions occur or subchar<0.
+ *                      pNumSubstitutions can be NULL.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strFromUTF32
+ * @see u_strToUTF32WithSub
+ * @stable ICU 4.2
+ */
+U_CAPI UChar* U_EXPORT2
+u_strFromUTF32WithSub(UChar *dest,
+               int32_t destCapacity,
+               int32_t *pDestLength,
+               const UChar32 *src,
+               int32_t srcLength,
+               UChar32 subchar, int32_t *pNumSubstitutions,
+               UErrorCode *pErrorCode);
+
+/**
+ * Convert a 16-bit Unicode string to Java Modified UTF-8.
+ * See http://java.sun.com/javase/6/docs/api/java/io/DataInput.html#modified-utf-8
+ *
+ * This function behaves according to the documentation for Java DataOutput.writeUTF()
+ * except that it does not encode the output length in the destination buffer
+ * and does not have an output length restriction.
+ * See http://java.sun.com/javase/6/docs/api/java/io/DataOutput.html#writeUTF(java.lang.String)
+ *
+ * The input string need not be well-formed UTF-16.
+ * (Therefore there is no subchar parameter.)
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of chars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @stable ICU 4.4
+ * @see u_strToUTF8WithSub
+ * @see u_strFromJavaModifiedUTF8WithSub
+ */
+U_CAPI char* U_EXPORT2 
+u_strToJavaModifiedUTF8(
+        char *dest,
+        int32_t destCapacity,
+        int32_t *pDestLength,
+        const UChar *src, 
+        int32_t srcLength,
+        UErrorCode *pErrorCode);
+
+/**
+ * Convert a Java Modified UTF-8 string to a 16-bit Unicode string.
+ * If the input string is not well-formed and no substitution char is specified, 
+ * then the U_INVALID_CHAR_FOUND error code is set.
+ *
+ * This function behaves according to the documentation for Java DataInput.readUTF()
+ * except that it takes a length parameter rather than
+ * interpreting the first two input bytes as the length.
+ * See http://java.sun.com/javase/6/docs/api/java/io/DataInput.html#readUTF()
+ *
+ * The output string may not be well-formed UTF-16.
+ *
+ * @param dest          A buffer for the result string. The result will be zero-terminated if
+ *                      the buffer is large enough.
+ * @param destCapacity  The size of the buffer (number of UChars). If it is 0, then
+ *                      dest may be NULL and the function will only return the length of the 
+ *                      result without writing any of the result string (pre-flighting).
+ * @param pDestLength   A pointer to receive the number of units written to the destination. If 
+ *                      pDestLength!=NULL then *pDestLength is always set to the 
+ *                      number of output units corresponding to the transformation of 
+ *                      all the input units, even in case of a buffer overflow.
+ * @param src           The original source string
+ * @param srcLength     The length of the original string. If -1, then src must be zero-terminated.
+ * @param subchar       The substitution character to use in place of an illegal input sequence,
+ *                      or U_SENTINEL if the function is to return with U_INVALID_CHAR_FOUND instead.
+ *                      A substitution character can be any valid Unicode code point (up to U+10FFFF)
+ *                      except for surrogate code points (U+D800..U+DFFF).
+ *                      The recommended value is U+FFFD "REPLACEMENT CHARACTER".
+ * @param pNumSubstitutions Output parameter receiving the number of substitutions if subchar>=0.
+ *                      Set to 0 if no substitutions occur or subchar<0.
+ *                      pNumSubstitutions can be NULL.
+ * @param pErrorCode    Pointer to a standard ICU error code. Its input value must
+ *                      pass the U_SUCCESS() test, or else the function returns
+ *                      immediately. Check for U_FAILURE() on output or use with
+ *                      function chaining. (See User Guide for details.)
+ * @return The pointer to destination buffer.
+ * @see u_strFromUTF8WithSub
+ * @see u_strFromUTF8Lenient
+ * @see u_strToJavaModifiedUTF8
+ * @stable ICU 4.4
+ */
+U_CAPI UChar* U_EXPORT2
+u_strFromJavaModifiedUTF8WithSub(
+        UChar *dest,
+        int32_t destCapacity,
+        int32_t *pDestLength,
+        const char *src,
+        int32_t srcLength,
+        UChar32 subchar, int32_t *pNumSubstitutions,
+        UErrorCode *pErrorCode);
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/ustringtrie.h b/thirdparty/icu4c/common/unicode/ustringtrie.h
new file mode 100644
index 0000000000..fd85648225
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/ustringtrie.h
@@ -0,0 +1,97 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  udicttrie.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010dec17
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __USTRINGTRIE_H__
+#define __USTRINGTRIE_H__
+
+/**
+ * \file
+ * \brief C API: Helper definitions for dictionary trie APIs.
+ */
+
+#include "unicode/utypes.h"
+
+
+/**
+ * Return values for BytesTrie::next(), UCharsTrie::next() and similar methods.
+ * @see USTRINGTRIE_MATCHES
+ * @see USTRINGTRIE_HAS_VALUE
+ * @see USTRINGTRIE_HAS_NEXT
+ * @stable ICU 4.8
+ */
+enum UStringTrieResult {
+    /**
+     * The input unit(s) did not continue a matching string.
+     * Once current()/next() return USTRINGTRIE_NO_MATCH,
+     * all further calls to current()/next() will also return USTRINGTRIE_NO_MATCH,
+     * until the trie is reset to its original state or to a saved state.
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_NO_MATCH,
+    /**
+     * The input unit(s) continued a matching string
+     * but there is no value for the string so far.
+     * (It is a prefix of a longer string.)
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_NO_VALUE,
+    /**
+     * The input unit(s) continued a matching string
+     * and there is a value for the string so far.
+     * This value will be returned by getValue().
+     * No further input byte/unit can continue a matching string.
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_FINAL_VALUE,
+    /**
+     * The input unit(s) continued a matching string
+     * and there is a value for the string so far.
+     * This value will be returned by getValue().
+     * Another input byte/unit can continue a matching string.
+     * @stable ICU 4.8
+     */
+    USTRINGTRIE_INTERMEDIATE_VALUE
+};
+
+/**
+ * Same as (result!=USTRINGTRIE_NO_MATCH).
+ * @param result A result from BytesTrie::first(), UCharsTrie::next() etc.
+ * @return true if the input bytes/units so far are part of a matching string/byte sequence.
+ * @stable ICU 4.8
+ */
+#define USTRINGTRIE_MATCHES(result) ((result)!=USTRINGTRIE_NO_MATCH)
+
+/**
+ * Equivalent to (result==USTRINGTRIE_INTERMEDIATE_VALUE || result==USTRINGTRIE_FINAL_VALUE) but
+ * this macro evaluates result exactly once.
+ * @param result A result from BytesTrie::first(), UCharsTrie::next() etc.
+ * @return true if there is a value for the input bytes/units so far.
+ * @see BytesTrie::getValue
+ * @see UCharsTrie::getValue
+ * @stable ICU 4.8
+ */
+#define USTRINGTRIE_HAS_VALUE(result) ((result)>=USTRINGTRIE_FINAL_VALUE)
+
+/**
+ * Equivalent to (result==USTRINGTRIE_NO_VALUE || result==USTRINGTRIE_INTERMEDIATE_VALUE) but
+ * this macro evaluates result exactly once.
+ * @param result A result from BytesTrie::first(), UCharsTrie::next() etc.
+ * @return true if another input byte/unit can continue a matching string.
+ * @stable ICU 4.8
+ */
+#define USTRINGTRIE_HAS_NEXT(result) ((result)&1)
+
+#endif  /* __USTRINGTRIE_H__ */
diff --git a/thirdparty/icu4c/common/unicode/utext.h b/thirdparty/icu4c/common/unicode/utext.h
new file mode 100644
index 0000000000..c6d1e53a8c
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utext.h
@@ -0,0 +1,1603 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utext.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004oct06
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UTEXT_H__
+#define __UTEXT_H__
+
+/**
+ * \file
+ * \brief C API: Abstract Unicode Text API
+ *
+ * The Text Access API provides a means to allow text that is stored in alternative
+ * formats to work with ICU services.  ICU normally operates on text that is
+ * stored in UTF-16 format, in (UChar *) arrays for the C APIs or as type
+ * UnicodeString for C++ APIs.
+ *
+ * ICU Text Access allows other formats, such as UTF-8 or non-contiguous
+ * UTF-16 strings, to be placed in a UText wrapper and then passed to ICU services.
+ *
+ * There are three general classes of usage for UText:
+ *
+ *     Application Level Use.  This is the simplest usage - applications would
+ *     use one of the utext_open() functions on their input text, and pass
+ *     the resulting UText to the desired ICU service.
+ *
+ *     Second is usage in ICU Services, such as break iteration, that will need to
+ *     operate on input presented to them as a UText.  These implementations
+ *     will need to use the iteration and related UText functions to gain
+ *     access to the actual text.
+ *
+ *     The third class of UText users are "text providers."  These are the
+ *     UText implementations for the various text storage formats.  An application
+ *     or system with a unique text storage format can implement a set of
+ *     UText provider functions for that format, which will then allow
+ *     ICU services to operate on that format.
+ *
+ *
+ * <em>Iterating over text</em>
+ *
+ * Here is sample code for a forward iteration over the contents of a UText
+ *
+ * \code
+ *    UChar32  c;
+ *    UText    *ut = whatever();
+ *
+ *    for (c=utext_next32From(ut, 0); c>=0; c=utext_next32(ut)) {
+ *       // do whatever with the codepoint c here.
+ *    }
+ * \endcode
+ *
+ * And here is similar code to iterate in the reverse direction, from the end
+ * of the text towards the beginning.
+ *
+ * \code
+ *    UChar32  c;
+ *    UText    *ut = whatever();
+ *    int      textLength = utext_nativeLength(ut);
+ *    for (c=utext_previous32From(ut, textLength); c>=0; c=utext_previous32(ut)) {
+ *       // do whatever with the codepoint c here.
+ *    }
+ * \endcode
+ *
+ * <em>Characters and Indexing</em>
+ *
+ * Indexing into text by UText functions is nearly always in terms of the native
+ * indexing of the underlying text storage.  The storage format could be UTF-8
+ * or UTF-32, for example.  When coding to the UText access API, no assumptions
+ * can be made regarding the size of characters, or how far an index
+ * may move when iterating between characters.
+ *
+ * All indices supplied to UText functions are pinned to the length of the
+ * text.  An out-of-bounds index is not considered to be an error, but is
+ * adjusted to be in the range  0 <= index <= length of input text.
+ *
+ *
+ * When an index position is returned from a UText function, it will be
+ * a native index to the underlying text.  In the case of multi-unit characters,
+ * it will  always refer to the first position of the character,
+ * never to the interior.  This is essentially the same thing as saying that
+ * a returned index will always point to a boundary between characters.
+ *
+ * When a native index is supplied to a UText function, all indices that
+ * refer to any part of a multi-unit character representation are considered
+ * to be equivalent.  In the case of multi-unit characters, an incoming index
+ * will be logically normalized to refer to the start of the character.
+ * 
+ * It is possible to test whether a native index is on a code point boundary
+ * by doing a utext_setNativeIndex() followed by a utext_getNativeIndex().
+ * If the index is returned unchanged, it was on a code point boundary.  If
+ * an adjusted index is returned, the original index referred to the
+ * interior of a character.
+ *
+ * <em>Conventions for calling UText functions</em>
+ *
+ * Most UText access functions have as their first parameter a (UText *) pointer,
+ * which specifies the UText to be used.  Unless otherwise noted, the
+ * pointer must refer to a valid, open UText.  Attempting to
+ * use a closed UText or passing a NULL pointer is a programming error and
+ * will produce undefined results or NULL pointer exceptions.
+ * 
+ * The UText_Open family of functions can either open an existing (closed)
+ * UText, or heap allocate a new UText.  Here is sample code for creating
+ * a stack-allocated UText.
+ *
+ * \code
+ *    char     *s = whatever();  // A utf-8 string 
+ *    U_ErrorCode status = U_ZERO_ERROR;
+ *    UText    ut = UTEXT_INITIALIZER;
+ *    utext_openUTF8(ut, s, -1, &status);
+ *    if (U_FAILURE(status)) {
+ *        // error handling
+ *    } else {
+ *        // work with the UText
+ *    }
+ * \endcode
+ *
+ * Any existing UText passed to an open function _must_ have been initialized, 
+ * either by the UTEXT_INITIALIZER, or by having been originally heap-allocated
+ * by an open function.  Passing NULL will cause the open function to
+ * heap-allocate and fully initialize a new UText.
+ *
+ */
+
+
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+#if U_SHOW_CPLUSPLUS_API
+#include "unicode/localpointer.h"
+#include "unicode/rep.h"
+#include "unicode/unistr.h"
+#include "unicode/chariter.h"
+#endif
+
+
+U_CDECL_BEGIN
+
+struct UText;
+typedef struct UText UText; /**< C typedef for struct UText. @stable ICU 3.6 */
+
+
+/***************************************************************************************
+ *
+ *   C Functions for creating UText wrappers around various kinds of text strings.
+ *
+ ****************************************************************************************/
+
+
+/**
+  * Close function for UText instances.
+  * Cleans up, releases any resources being held by an open UText.
+  * <p>
+  *   If the UText was originally allocated by one of the utext_open functions,
+  *   the storage associated with the utext will also be freed.
+  *   If the UText storage originated with the application, as it would with
+  *   a local or static instance, the storage will not be deleted.
+  *
+  *   An open UText can be reset to refer to new string by using one of the utext_open()
+  *   functions without first closing the UText.  
+  *
+  * @param ut  The UText to be closed.
+  * @return    NULL if the UText struct was deleted by the close.  If the UText struct
+  *            was originally provided by the caller to the open function, it is
+  *            returned by this function, and may be safely used again in
+  *            a subsequent utext_open.
+  *
+  * @stable ICU 3.4
+  */
+U_CAPI UText * U_EXPORT2
+utext_close(UText *ut);
+
+/**
+ * Open a read-only UText implementation for UTF-8 strings.
+ * 
+ * \htmlonly
+ * Any invalid UTF-8 in the input will be handled in this way:
+ * a sequence of bytes that has the form of a truncated, but otherwise valid,
+ * UTF-8 sequence will be replaced by a single unicode replacement character, \uFFFD. 
+ * Any other illegal bytes will each be replaced by a \uFFFD.
+ * \endhtmlonly
+ * 
+ * @param ut     Pointer to a UText struct.  If NULL, a new UText will be created.
+ *               If non-NULL, must refer to an initialized UText struct, which will then
+ *               be reset to reference the specified UTF-8 string.
+ * @param s      A UTF-8 string.  Must not be NULL.
+ * @param length The length of the UTF-8 string in bytes, or -1 if the string is
+ *               zero terminated.
+ * @param status Errors are returned here.
+ * @return       A pointer to the UText.  If a pre-allocated UText was provided, it
+ *               will always be used and returned.
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openUTF8(UText *ut, const char *s, int64_t length, UErrorCode *status);
+
+
+/**
+ * Open a read-only UText for UChar * string.
+ * 
+ * @param ut     Pointer to a UText struct.  If NULL, a new UText will be created.
+ *               If non-NULL, must refer to an initialized UText struct, which will then
+ *               be reset to reference the specified UChar string.
+ * @param s      A UChar (UTF-16) string
+ * @param length The number of UChars in the input string, or -1 if the string is
+ *               zero terminated.
+ * @param status Errors are returned here.
+ * @return       A pointer to the UText.  If a pre-allocated UText was provided, it
+ *               will always be used and returned.
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openUChars(UText *ut, const UChar *s, int64_t length, UErrorCode *status);
+
+
+#if U_SHOW_CPLUSPLUS_API
+/**
+ * Open a writable UText for a non-const UnicodeString. 
+ * 
+ * @param ut      Pointer to a UText struct.  If NULL, a new UText will be created.
+ *                 If non-NULL, must refer to an initialized UText struct, which will then
+ *                 be reset to reference the specified input string.
+ * @param s       A UnicodeString.
+ * @param status Errors are returned here.
+ * @return        Pointer to the UText.  If a UText was supplied as input, this
+ *                 will always be used and returned.
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openUnicodeString(UText *ut, icu::UnicodeString *s, UErrorCode *status);
+
+
+/**
+ * Open a UText for a const UnicodeString.   The resulting UText will not be writable.
+ * 
+ * @param ut    Pointer to a UText struct.  If NULL, a new UText will be created.
+ *               If non-NULL, must refer to an initialized UText struct, which will then
+ *               be reset to reference the specified input string.
+ * @param s      A const UnicodeString to be wrapped.
+ * @param status Errors are returned here.
+ * @return       Pointer to the UText.  If a UText was supplied as input, this
+ *               will always be used and returned.
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openConstUnicodeString(UText *ut, const icu::UnicodeString *s, UErrorCode *status);
+
+
+/**
+ * Open a writable UText implementation for an ICU Replaceable object.
+ * @param ut    Pointer to a UText struct.  If NULL, a new UText will be created.
+ *               If non-NULL, must refer to an already existing UText, which will then
+ *               be reset to reference the specified replaceable text.
+ * @param rep    A Replaceable text object.
+ * @param status Errors are returned here.
+ * @return       Pointer to the UText.  If a UText was supplied as input, this
+ *               will always be used and returned.
+ * @see Replaceable
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openReplaceable(UText *ut, icu::Replaceable *rep, UErrorCode *status);
+
+/**
+ * Open a  UText implementation over an ICU CharacterIterator.
+ * @param ut    Pointer to a UText struct.  If NULL, a new UText will be created.
+ *               If non-NULL, must refer to an already existing UText, which will then
+ *               be reset to reference the specified replaceable text.
+ * @param ci     A Character Iterator.
+ * @param status Errors are returned here.
+ * @return       Pointer to the UText.  If a UText was supplied as input, this
+ *               will always be used and returned.
+ * @see Replaceable
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_openCharacterIterator(UText *ut, icu::CharacterIterator *ci, UErrorCode *status);
+
+#endif
+
+
+/**
+  *  Clone a UText.  This is much like opening a UText where the source text is itself
+  *  another UText.
+  *
+  *  A deep clone will copy both the UText data structures and the underlying text.
+  *  The original and cloned UText will operate completely independently; modifications
+  *  made to the text in one will not affect the other.  Text providers are not
+  *  required to support deep clones.  The user of clone() must check the status return
+  *  and be prepared to handle failures.
+  *
+  *  The standard UText implementations for UTF8, UChar *, UnicodeString and
+  *  Replaceable all support deep cloning.
+  *
+  *  The UText returned from a deep clone will be writable, assuming that the text
+  *  provider is able to support writing, even if the source UText had been made
+  *  non-writable by means of UText_freeze().
+  *
+  *  A shallow clone replicates only the UText data structures; it does not make
+  *  a copy of the underlying text.  Shallow clones can be used as an efficient way to 
+  *  have multiple iterators active in a single text string that is not being
+  *  modified.
+  *
+  *  A shallow clone operation will not fail, barring truly exceptional conditions such
+  *  as memory allocation failures.
+  *
+  *  Shallow UText clones should be avoided if the UText functions that modify the
+  *  text are expected to be used, either on the original or the cloned UText.
+  *  Any such modifications  can cause unpredictable behavior.  Read Only
+  *  shallow clones provide some protection against errors of this type by
+  *  disabling text modification via the cloned UText.
+  *
+  *  A shallow clone made with the readOnly parameter == false will preserve the 
+  *  utext_isWritable() state of the source object.  Note, however, that
+  *  write operations must be avoided while more than one UText exists that refer
+  *  to the same underlying text.
+  *
+  *  A UText and its clone may be safely concurrently accessed by separate threads.
+  *  This is true for read access only with shallow clones, and for both read and
+  *  write access with deep clones.
+  *  It is the responsibility of the Text Provider to ensure that this thread safety
+  *  constraint is met.
+  *
+  *  @param dest   A UText struct to be filled in with the result of the clone operation,
+  *                or NULL if the clone function should heap-allocate a new UText struct.
+  *                If non-NULL, must refer to an already existing UText, which will then
+  *                be reset to become the clone.
+  *  @param src    The UText to be cloned.
+  *  @param deep   true to request a deep clone, false for a shallow clone.
+  *  @param readOnly true to request that the cloned UText have read only access to the 
+  *                underlying text.  
+
+  *  @param status Errors are returned here.  For deep clones, U_UNSUPPORTED_ERROR
+  *                will be returned if the text provider is unable to clone the
+  *                original text.
+  *  @return       The newly created clone, or NULL if the clone operation failed.
+  *  @stable ICU 3.4
+  */
+U_CAPI UText * U_EXPORT2
+utext_clone(UText *dest, const UText *src, UBool deep, UBool readOnly, UErrorCode *status);
+
+
+/**
+  *  Compare two UText objects for equality.
+  *  UTexts are equal if they are iterating over the same text, and
+  *    have the same iteration position within the text.
+  *    If either or both of the parameters are NULL, the comparison is false.
+  *
+  *  @param a   The first of the two UTexts to compare.
+  *  @param b   The other UText to be compared.
+  *  @return    true if the two UTexts are equal.
+  *  @stable ICU 3.6
+  */
+U_CAPI UBool U_EXPORT2
+utext_equals(const UText *a, const UText *b);
+
+
+/*****************************************************************************
+ *
+ *   Functions to work with the text represented by a UText wrapper
+ *
+ *****************************************************************************/
+
+/**
+  * Get the length of the text.  Depending on the characteristics
+  * of the underlying text representation, this may be expensive.  
+  * @see  utext_isLengthExpensive()
+  *
+  *
+  * @param ut  the text to be accessed.
+  * @return the length of the text, expressed in native units.
+  *
+  * @stable ICU 3.4
+  */
+U_CAPI int64_t U_EXPORT2
+utext_nativeLength(UText *ut);
+
+/**
+ *  Return true if calculating the length of the text could be expensive.
+ *  Finding the length of NUL terminated strings is considered to be expensive.
+ *
+ *  Note that the value of this function may change
+ *  as the result of other operations on a UText.
+ *  Once the length of a string has been discovered, it will no longer
+ *  be expensive to report it.
+ *
+ * @param ut the text to be accessed.
+ * @return true if determining the length of the text could be time consuming.
+ * @stable ICU 3.4
+ */
+U_CAPI UBool U_EXPORT2
+utext_isLengthExpensive(const UText *ut);
+
+/**
+ * Returns the code point at the requested index,
+ * or U_SENTINEL (-1) if it is out of bounds.
+ *
+ * If the specified index points to the interior of a multi-unit
+ * character - one of the trail bytes of a UTF-8 sequence, for example -
+ * the complete code point will be returned.
+ *
+ * The iteration position will be set to the start of the returned code point.
+ *
+ * This function is roughly equivalent to the sequence
+ *    utext_setNativeIndex(index);
+ *    utext_current32();
+ * (There is a subtle difference if the index is out of bounds by being less than zero - 
+ * utext_setNativeIndex(negative value) sets the index to zero, after which utext_current()
+ * will return the char at zero.  utext_char32At(negative index), on the other hand, will
+ * return the U_SENTINEL value of -1.)
+ * 
+ * @param ut the text to be accessed
+ * @param nativeIndex the native index of the character to be accessed.  If the index points
+ *        to other than the first unit of a multi-unit character, it will be adjusted
+ *        to the start of the character.
+ * @return the code point at the specified index.
+ * @stable ICU 3.4
+ */
+U_CAPI UChar32 U_EXPORT2
+utext_char32At(UText *ut, int64_t nativeIndex);
+
+
+/**
+ *
+ * Get the code point at the current iteration position,
+ * or U_SENTINEL (-1) if the iteration has reached the end of
+ * the input text.
+ *
+ * @param ut the text to be accessed.
+ * @return the Unicode code point at the current iterator position.
+ * @stable ICU 3.4
+ */
+U_CAPI UChar32 U_EXPORT2
+utext_current32(UText *ut);
+
+
+/**
+ * Get the code point at the current iteration position of the UText, and
+ * advance the position to the first index following the character.
+ *
+ * If the position is at the end of the text (the index following
+ * the last character, which is also the length of the text), 
+ * return U_SENTINEL (-1) and do not advance the index. 
+ *
+ * This is a post-increment operation.
+ *
+ * An inline macro version of this function, UTEXT_NEXT32(), 
+ * is available for performance critical use.
+ *
+ * @param ut the text to be accessed.
+ * @return the Unicode code point at the iteration position.
+ * @see UTEXT_NEXT32
+ * @stable ICU 3.4
+ */
+U_CAPI UChar32 U_EXPORT2
+utext_next32(UText *ut);
+
+
+/**
+ *  Move the iterator position to the character (code point) whose
+ *  index precedes the current position, and return that character.
+ *  This is a pre-decrement operation.
+ *
+ *  If the initial position is at the start of the text (index of 0) 
+ *  return U_SENTINEL (-1), and leave the position unchanged.
+ *
+ *  An inline macro version of this function, UTEXT_PREVIOUS32(), 
+ *  is available for performance critical use.
+ *
+ *  @param ut the text to be accessed.
+ *  @return the previous UChar32 code point, or U_SENTINEL (-1) 
+ *          if the iteration has reached the start of the text.
+ *  @see UTEXT_PREVIOUS32
+ *  @stable ICU 3.4
+ */
+U_CAPI UChar32 U_EXPORT2
+utext_previous32(UText *ut);
+
+
+/**
+  * Set the iteration index and return the code point at that index. 
+  * Leave the iteration index at the start of the following code point.
+  *
+  * This function is the most efficient and convenient way to
+  * begin a forward iteration.  The results are identical to the those
+  * from the sequence
+  * \code
+  *    utext_setIndex();
+  *    utext_next32();
+  * \endcode
+  *
+  *  @param ut the text to be accessed.
+  *  @param nativeIndex Iteration index, in the native units of the text provider.
+  *  @return Code point which starts at or before index,
+  *         or U_SENTINEL (-1) if it is out of bounds.
+  * @stable ICU 3.4
+  */
+U_CAPI UChar32 U_EXPORT2
+utext_next32From(UText *ut, int64_t nativeIndex);
+
+
+
+/**
+  * Set the iteration index, and return the code point preceding the
+  * one specified by the initial index.  Leave the iteration position
+  * at the start of the returned code point.
+  *
+  * This function is the most efficient and convenient way to
+  * begin a backwards iteration.
+  *
+  * @param ut the text to be accessed.
+  * @param nativeIndex Iteration index in the native units of the text provider.
+  * @return Code point preceding the one at the initial index,
+  *         or U_SENTINEL (-1) if it is out of bounds.
+  *
+  * @stable ICU 3.4
+  */
+U_CAPI UChar32 U_EXPORT2
+utext_previous32From(UText *ut, int64_t nativeIndex);
+
+/**
+  * Get the current iterator position, which can range from 0 to 
+  * the length of the text.
+  * The position is a native index into the input text, in whatever format it
+  * may have (possibly UTF-8 for example), and may not always be the same as
+  * the corresponding UChar (UTF-16) index.
+  * The returned position will always be aligned to a code point boundary. 
+  *
+  * @param ut the text to be accessed.
+  * @return the current index position, in the native units of the text provider.
+  * @stable ICU 3.4
+  */
+U_CAPI int64_t U_EXPORT2
+utext_getNativeIndex(const UText *ut);
+
+/**
+ * Set the current iteration position to the nearest code point
+ * boundary at or preceding the specified index.
+ * The index is in the native units of the original input text.
+ * If the index is out of range, it will be pinned to be within
+ * the range of the input text.
+ * <p>
+ * It will usually be more efficient to begin an iteration
+ * using the functions utext_next32From() or utext_previous32From()
+ * rather than setIndex().
+ * <p>
+ * Moving the index position to an adjacent character is best done
+ * with utext_next32(), utext_previous32() or utext_moveIndex32().
+ * Attempting to do direct arithmetic on the index position is
+ * complicated by the fact that the size (in native units) of a
+ * character depends on the underlying representation of the character
+ * (UTF-8, UTF-16, UTF-32, arbitrary codepage), and is not
+ * easily knowable.
+ *
+ * @param ut the text to be accessed.
+ * @param nativeIndex the native unit index of the new iteration position.
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+utext_setNativeIndex(UText *ut, int64_t nativeIndex);
+
+/**
+ * Move the iterator position by delta code points.  The number of code points
+ * is a signed number; a negative delta will move the iterator backwards,
+ * towards the start of the text.
+ * <p>
+ * The index is moved by <code>delta</code> code points
+ * forward or backward, but no further backward than to 0 and
+ * no further forward than to utext_nativeLength().
+ * The resulting index value will be in between 0 and length, inclusive.
+ *
+ * @param ut the text to be accessed.
+ * @param delta the signed number of code points to move the iteration position.
+ * @return true if the position could be moved the requested number of positions while
+ *              staying within the range [0 - text length].
+ * @stable ICU 3.4
+ */
+U_CAPI UBool U_EXPORT2
+utext_moveIndex32(UText *ut, int32_t delta);
+
+/**
+ * Get the native index of the character preceding the current position.
+ * If the iteration position is already at the start of the text, zero
+ * is returned.
+ * The value returned is the same as that obtained from the following sequence,
+ * but without the side effect of changing the iteration position.
+ *   
+ * \code
+ *    UText  *ut = whatever;
+ *      ...
+ *    utext_previous(ut)
+ *    utext_getNativeIndex(ut);
+ * \endcode
+ *
+ * This function is most useful during forwards iteration, where it will get the
+ *   native index of the character most recently returned from utext_next().
+ *
+ * @param ut the text to be accessed
+ * @return the native index of the character preceding the current index position,
+ *         or zero if the current position is at the start of the text.
+ * @stable ICU 3.6
+ */
+U_CAPI int64_t U_EXPORT2
+utext_getPreviousNativeIndex(UText *ut); 
+
+
+/**
+ *
+ * Extract text from a UText into a UChar buffer.  The range of text to be extracted
+ * is specified in the native indices of the UText provider.  These may not necessarily
+ * be UTF-16 indices.
+ * <p>
+ * The size (number of 16 bit UChars) of the data to be extracted is returned.  The
+ * full number of UChars is returned, even when the extracted text is truncated
+ * because the specified buffer size is too small.
+ * <p>
+ * The extracted string will (if you are a user) / must (if you are a text provider)
+ * be NUL-terminated if there is sufficient space in the destination buffer.  This
+ * terminating NUL is not included in the returned length.
+ * <p>
+ * The iteration index is left at the position following the last extracted character.
+ *
+ * @param  ut    the UText from which to extract data.
+ * @param  nativeStart the native index of the first character to extract.\
+ *               If the specified index is out of range,
+ *               it will be pinned to be within 0 <= index <= textLength
+ * @param  nativeLimit the native string index of the position following the last
+ *               character to extract.  If the specified index is out of range,
+ *               it will be pinned to be within 0 <= index <= textLength.
+ *               nativeLimit must be >= nativeStart.
+ * @param  dest  the UChar (UTF-16) buffer into which the extracted text is placed
+ * @param  destCapacity  The size, in UChars, of the destination buffer.  May be zero
+ *               for precomputing the required size.
+ * @param  status receives any error status.
+ *         U_BUFFER_OVERFLOW_ERROR: the extracted text was truncated because the 
+ *         buffer was too small.  Returns number of UChars for preflighting.
+ * @return Number of UChars in the data to be extracted.  Does not include a trailing NUL.
+ *
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+utext_extract(UText *ut,
+             int64_t nativeStart, int64_t nativeLimit,
+             UChar *dest, int32_t destCapacity,
+             UErrorCode *status);
+
+
+
+/************************************************************************************
+ *
+ *  #define inline versions of selected performance-critical text access functions
+ *          Caution:  do not use auto increment++ or decrement-- expressions
+ *                    as parameters to these macros.
+ *
+ *          For most use, where there is no extreme performance constraint, the
+ *          normal, non-inline functions are a better choice.  The resulting code
+ *          will be smaller, and, if the need ever arises, easier to debug.
+ *
+ *          These are implemented as #defines rather than real functions
+ *          because there is no fully portable way to do inline functions in plain C.
+ *
+ ************************************************************************************/
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * inline version of utext_current32(), for performance-critical situations.
+ *
+ * Get the code point at the current iteration position of the UText.
+ * Returns U_SENTINEL (-1) if the position is at the end of the
+ * text.
+ *
+ * @internal ICU 4.4 technology preview
+ */
+#define UTEXT_CURRENT32(ut)  \
+    ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkOffset]<0xd800 ? \
+    ((ut)->chunkContents)[((ut)->chunkOffset)] : utext_current32(ut))
+#endif  /* U_HIDE_INTERNAL_API */
+
+/**
+ * inline version of utext_next32(), for performance-critical situations.
+ *
+ * Get the code point at the current iteration position of the UText, and
+ * advance the position to the first index following the character.
+ * This is a post-increment operation.
+ * Returns U_SENTINEL (-1) if the position is at the end of the
+ * text.
+ *
+ * @stable ICU 3.4
+ */
+#define UTEXT_NEXT32(ut)  \
+    ((ut)->chunkOffset < (ut)->chunkLength && ((ut)->chunkContents)[(ut)->chunkOffset]<0xd800 ? \
+    ((ut)->chunkContents)[((ut)->chunkOffset)++] : utext_next32(ut))
+
+/**
+ * inline version of utext_previous32(), for performance-critical situations.
+ *
+ *  Move the iterator position to the character (code point) whose
+ *  index precedes the current position, and return that character.
+ *  This is a pre-decrement operation.
+ *  Returns U_SENTINEL (-1) if the position is at the start of the  text.
+ *
+ * @stable ICU 3.4
+ */
+#define UTEXT_PREVIOUS32(ut)  \
+    ((ut)->chunkOffset > 0 && \
+     (ut)->chunkContents[(ut)->chunkOffset-1] < 0xd800 ? \
+          (ut)->chunkContents[--((ut)->chunkOffset)]  :  utext_previous32(ut))
+
+/**
+  *  inline version of utext_getNativeIndex(), for performance-critical situations.
+  *
+  * Get the current iterator position, which can range from 0 to 
+  * the length of the text.
+  * The position is a native index into the input text, in whatever format it
+  * may have (possibly UTF-8 for example), and may not always be the same as
+  * the corresponding UChar (UTF-16) index.
+  * The returned position will always be aligned to a code point boundary. 
+  *
+  * @stable ICU 3.6
+  */
+#define UTEXT_GETNATIVEINDEX(ut)                       \
+    ((ut)->chunkOffset <= (ut)->nativeIndexingLimit?   \
+        (ut)->chunkNativeStart+(ut)->chunkOffset :     \
+        (ut)->pFuncs->mapOffsetToNative(ut))    
+
+/**
+  *  inline version of utext_setNativeIndex(), for performance-critical situations.
+  *
+  * Set the current iteration position to the nearest code point
+  * boundary at or preceding the specified index.
+  * The index is in the native units of the original input text.
+  * If the index is out of range, it will be pinned to be within
+  * the range of the input text.
+  *
+  * @stable ICU 3.8
+  */
+#define UTEXT_SETNATIVEINDEX(ut, ix) UPRV_BLOCK_MACRO_BEGIN { \
+    int64_t __offset = (ix) - (ut)->chunkNativeStart; \
+    if (__offset>=0 && __offset<(int64_t)(ut)->nativeIndexingLimit && (ut)->chunkContents[__offset]<0xdc00) { \
+        (ut)->chunkOffset=(int32_t)__offset; \
+    } else { \
+        utext_setNativeIndex((ut), (ix)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+
+
+/************************************************************************************
+ *
+ *   Functions related to writing or modifying the text.
+ *   These will work only with modifiable UTexts.  Attempting to
+ *   modify a read-only UText will return an error status.
+ *
+ ************************************************************************************/
+
+
+/**
+ *  Return true if the text can be written (modified) with utext_replace() or
+ *  utext_copy().  For the text to be writable, the text provider must
+ *  be of a type that supports writing and the UText must not be frozen.
+ *
+ *  Attempting to modify text when utext_isWriteable() is false will fail -
+ *  the text will not be modified, and an error will be returned from the function
+ *  that attempted the modification.
+ *
+ * @param  ut   the UText to be tested.
+ * @return true if the text is modifiable.
+ *
+ * @see    utext_freeze()
+ * @see    utext_replace()
+ * @see    utext_copy()
+ * @stable ICU 3.4
+ *
+ */
+U_CAPI UBool U_EXPORT2
+utext_isWritable(const UText *ut);
+
+
+/**
+  * Test whether there is meta data associated with the text.
+  * @see Replaceable::hasMetaData()
+  *
+  * @param ut The UText to be tested
+  * @return true if the underlying text includes meta data.
+  * @stable ICU 3.4
+  */
+U_CAPI UBool U_EXPORT2
+utext_hasMetaData(const UText *ut);
+
+
+/**
+ * Replace a range of the original text with a replacement text.
+ *
+ * Leaves the current iteration position at the position following the
+ *  newly inserted replacement text.
+ *
+ * This function is only available on UText types that support writing,
+ * that is, ones where utext_isWritable() returns true.
+ *
+ * When using this function, there should be only a single UText opened onto the
+ * underlying native text string.  Behavior after a replace operation
+ * on a UText is undefined for any other additional UTexts that refer to the
+ * modified string.
+ *
+ * @param ut               the UText representing the text to be operated on.
+ * @param nativeStart      the native index of the start of the region to be replaced
+ * @param nativeLimit      the native index of the character following the region to be replaced.
+ * @param replacementText  pointer to the replacement text
+ * @param replacementLength length of the replacement text, or -1 if the text is NUL terminated.
+ * @param status           receives any error status.  Possible errors include
+ *                         U_NO_WRITE_PERMISSION
+ *
+ * @return The signed number of (native) storage units by which
+ *         the length of the text expanded or contracted.
+ *
+ * @stable ICU 3.4
+ */
+U_CAPI int32_t U_EXPORT2
+utext_replace(UText *ut,
+             int64_t nativeStart, int64_t nativeLimit,
+             const UChar *replacementText, int32_t replacementLength,
+             UErrorCode *status);
+
+
+
+/**
+ *
+ * Copy or move a substring from one position to another within the text,
+ * while retaining any metadata associated with the text.
+ * This function is used to duplicate or reorder substrings.
+ * The destination index must not overlap the source range.
+ *
+ * The text to be copied or moved is inserted at destIndex;
+ * it does not replace or overwrite any existing text.
+ *
+ * The iteration position is left following the newly inserted text
+ * at the destination position.
+ *
+ * This function is only available on UText types that support writing,
+ * that is, ones where utext_isWritable() returns true.
+ *
+ * When using this function, there should be only a single UText opened onto the
+ * underlying native text string.  Behavior after a copy operation
+ * on a UText is undefined in any other additional UTexts that refer to the
+ * modified string.
+ *
+ * @param ut           The UText representing the text to be operated on.
+ * @param nativeStart  The native index of the start of the region to be copied or moved
+ * @param nativeLimit  The native index of the character position following the region
+ *                     to be copied.
+ * @param destIndex    The native destination index to which the source substring is
+ *                     copied or moved.
+ * @param move         If true, then the substring is moved, not copied/duplicated.
+ * @param status       receives any error status.  Possible errors include U_NO_WRITE_PERMISSION
+ *                       
+ * @stable ICU 3.4
+ */
+U_CAPI void U_EXPORT2
+utext_copy(UText *ut,
+          int64_t nativeStart, int64_t nativeLimit,
+          int64_t destIndex,
+          UBool move,
+          UErrorCode *status);
+
+
+/**
+  *  <p>
+  *  Freeze a UText.  This prevents any modification to the underlying text itself
+  *  by means of functions operating on this UText.
+  *  </p>
+  *  <p>
+  *  Once frozen, a UText can not be unfrozen.  The intent is to ensure
+  *  that a the text underlying a frozen UText wrapper cannot be modified via that UText.
+  *  </p>
+  *  <p>
+  *  Caution:  freezing a UText will disable changes made via the specific
+  *   frozen UText wrapper only; it will not have any effect on the ability to
+  *   directly modify the text by bypassing the UText.  Any such backdoor modifications
+  *   are always an error while UText access is occurring because the underlying
+  *   text can get out of sync with UText's buffering.
+  *  </p>
+  *
+  *  @param ut  The UText to be frozen.
+  *  @see   utext_isWritable()
+  *  @stable ICU 3.6
+  */
+U_CAPI void U_EXPORT2
+utext_freeze(UText *ut);
+
+
+/**
+ * UText provider properties (bit field indexes).
+ *
+ * @see UText
+ * @stable ICU 3.4
+ */
+enum {
+    /**
+     * It is potentially time consuming for the provider to determine the length of the text.
+     * @stable ICU 3.4
+     */
+    UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE = 1,
+    /**
+     * Text chunks remain valid and usable until the text object is modified or
+     * deleted, not just until the next time the access() function is called
+     * (which is the default).
+     * @stable ICU 3.4
+     */
+    UTEXT_PROVIDER_STABLE_CHUNKS = 2,
+    /**
+     * The provider supports modifying the text via the replace() and copy()
+     * functions.
+     * @see Replaceable
+     * @stable ICU 3.4
+     */
+    UTEXT_PROVIDER_WRITABLE = 3,
+    /**
+     * There is meta data associated with the text.
+     * @see Replaceable::hasMetaData()
+     * @stable ICU 3.4
+     */ 
+    UTEXT_PROVIDER_HAS_META_DATA = 4,
+    /**
+     * Text provider owns the text storage.
+     *  Generally occurs as the result of a deep clone of the UText.
+     *  When closing the UText, the associated text must
+     *  also be closed/deleted/freed/ whatever is appropriate.
+     * @stable ICU 3.6
+     */
+     UTEXT_PROVIDER_OWNS_TEXT = 5
+};
+
+/**
+  * Function type declaration for UText.clone().
+  *
+  *  clone a UText.  Much like opening a UText where the source text is itself
+  *  another UText.
+  *
+  *  A deep clone will copy both the UText data structures and the underlying text.
+  *  The original and cloned UText will operate completely independently; modifications
+  *  made to the text in one will not effect the other.  Text providers are not
+  *  required to support deep clones.  The user of clone() must check the status return
+  *  and be prepared to handle failures.
+  *
+  *  A shallow clone replicates only the UText data structures; it does not make
+  *  a copy of the underlying text.  Shallow clones can be used as an efficient way to 
+  *  have multiple iterators active in a single text string that is not being
+  *  modified.
+  *
+  *  A shallow clone operation must not fail except for truly exceptional conditions such
+  *  as memory allocation failures.
+  *
+  *  A UText and its clone may be safely concurrently accessed by separate threads.
+  *  This is true for both shallow and deep clones.
+  *  It is the responsibility of the Text Provider to ensure that this thread safety
+  *  constraint is met.
+
+  *
+  *  @param dest   A UText struct to be filled in with the result of the clone operation,
+  *                or NULL if the clone function should heap-allocate a new UText struct.
+  *  @param src    The UText to be cloned.
+  *  @param deep   true to request a deep clone, false for a shallow clone.
+  *  @param status Errors are returned here.  For deep clones, U_UNSUPPORTED_ERROR
+  *                should be returned if the text provider is unable to clone the
+  *                original text.
+  *  @return       The newly created clone, or NULL if the clone operation failed.
+  *
+  * @stable ICU 3.4
+  */
+typedef UText * U_CALLCONV
+UTextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status);
+
+
+/**
+ * Function type declaration for UText.nativeLength().
+ *
+ * @param ut the UText to get the length of.
+ * @return the length, in the native units of the original text string.
+ * @see UText
+ * @stable ICU 3.4
+ */
+typedef int64_t U_CALLCONV
+UTextNativeLength(UText *ut);
+
+/**
+ * Function type declaration for UText.access().  Get the description of the text chunk
+ *  containing the text at a requested native index.  The UText's iteration
+ *  position will be left at the requested index.  If the index is out
+ *  of bounds, the iteration position will be left at the start or end
+ *  of the string, as appropriate.
+ *
+ *  Chunks must begin and end on code point boundaries.  A single code point
+ *  comprised of multiple storage units must never span a chunk boundary.
+ *
+ *
+ * @param ut          the UText being accessed.
+ * @param nativeIndex Requested index of the text to be accessed.
+ * @param forward     If true, then the returned chunk must contain text
+ *                    starting from the index, so that start<=index<limit.
+ *                    If false, then the returned chunk must contain text
+ *                    before the index, so that start<index<=limit.
+ * @return            True if the requested index could be accessed.  The chunk
+ *                    will contain the requested text.
+ *                    False value if a chunk cannot be accessed
+ *                    (the requested index is out of bounds).
+ *
+ * @see UText
+ * @stable ICU 3.4
+ */
+typedef UBool U_CALLCONV
+UTextAccess(UText *ut, int64_t nativeIndex, UBool forward);
+
+/**
+ * Function type declaration for UText.extract().
+ *
+ * Extract text from a UText into a UChar buffer.  The range of text to be extracted
+ * is specified in the native indices of the UText provider.  These may not necessarily
+ * be UTF-16 indices.
+ * <p>
+ * The size (number of 16 bit UChars) in the data to be extracted is returned.  The
+ * full amount is returned, even when the specified buffer size is smaller.
+ * <p>
+ * The extracted string will (if you are a user) / must (if you are a text provider)
+ * be NUL-terminated if there is sufficient space in the destination buffer.
+ *
+ * @param  ut            the UText from which to extract data.
+ * @param  nativeStart   the native index of the first character to extract.
+ * @param  nativeLimit   the native string index of the position following the last
+ *                       character to extract.
+ * @param  dest          the UChar (UTF-16) buffer into which the extracted text is placed
+ * @param  destCapacity  The size, in UChars, of the destination buffer.  May be zero
+ *                       for precomputing the required size.
+ * @param  status        receives any error status.
+ *                       If U_BUFFER_OVERFLOW_ERROR: Returns number of UChars for
+ *                       preflighting.
+ * @return Number of UChars in the data.  Does not include a trailing NUL.
+ *
+ * @stable ICU 3.4
+ */
+typedef int32_t U_CALLCONV
+UTextExtract(UText *ut,
+             int64_t nativeStart, int64_t nativeLimit,
+             UChar *dest, int32_t destCapacity,
+             UErrorCode *status);
+
+/**
+ * Function type declaration for UText.replace().
+ *
+ * Replace a range of the original text with a replacement text.
+ *
+ * Leaves the current iteration position at the position following the
+ *  newly inserted replacement text.
+ *
+ * This function need only be implemented on UText types that support writing.
+ *
+ * When using this function, there should be only a single UText opened onto the
+ * underlying native text string.  The function is responsible for updating the
+ * text chunk within the UText to reflect the updated iteration position,
+ * taking into account any changes to the underlying string's structure caused
+ * by the replace operation.
+ *
+ * @param ut               the UText representing the text to be operated on.
+ * @param nativeStart      the index of the start of the region to be replaced
+ * @param nativeLimit      the index of the character following the region to be replaced.
+ * @param replacementText  pointer to the replacement text
+ * @param replacmentLength length of the replacement text in UChars, or -1 if the text is NUL terminated.
+ * @param status           receives any error status.  Possible errors include
+ *                         U_NO_WRITE_PERMISSION
+ *
+ * @return The signed number of (native) storage units by which
+ *         the length of the text expanded or contracted.
+ *
+ * @stable ICU 3.4
+ */
+typedef int32_t U_CALLCONV
+UTextReplace(UText *ut,
+             int64_t nativeStart, int64_t nativeLimit,
+             const UChar *replacementText, int32_t replacmentLength,
+             UErrorCode *status);
+
+/**
+ * Function type declaration for UText.copy().
+ *
+ * Copy or move a substring from one position to another within the text,
+ * while retaining any metadata associated with the text.
+ * This function is used to duplicate or reorder substrings.
+ * The destination index must not overlap the source range.
+ *
+ * The text to be copied or moved is inserted at destIndex;
+ * it does not replace or overwrite any existing text.
+ *
+ * This function need only be implemented for UText types that support writing.
+ *
+ * When using this function, there should be only a single UText opened onto the
+ * underlying native text string.  The function is responsible for updating the
+ * text chunk within the UText to reflect the updated iteration position,
+ * taking into account any changes to the underlying string's structure caused
+ * by the replace operation.
+ *
+ * @param ut           The UText representing the text to be operated on.
+ * @param nativeStart  The index of the start of the region to be copied or moved
+ * @param nativeLimit  The index of the character following the region to be replaced.
+ * @param nativeDest   The destination index to which the source substring is copied or moved.
+ * @param move         If true, then the substring is moved, not copied/duplicated.
+ * @param status       receives any error status.  Possible errors include U_NO_WRITE_PERMISSION
+ *
+ * @stable ICU 3.4
+ */
+typedef void U_CALLCONV
+UTextCopy(UText *ut,
+          int64_t nativeStart, int64_t nativeLimit,
+          int64_t nativeDest,
+          UBool move,
+          UErrorCode *status);
+
+/**
+ * Function type declaration for UText.mapOffsetToNative().
+ * Map from the current UChar offset within the current text chunk to
+ *  the corresponding native index in the original source text.
+ *
+ * This is required only for text providers that do not use native UTF-16 indexes.
+ *
+ * @param ut     the UText.
+ * @return Absolute (native) index corresponding to chunkOffset in the current chunk.
+ *         The returned native index should always be to a code point boundary.
+ *
+ * @stable ICU 3.4
+ */
+typedef int64_t U_CALLCONV
+UTextMapOffsetToNative(const UText *ut);
+
+/**
+ * Function type declaration for UText.mapIndexToUTF16().
+ * Map from a native index to a UChar offset within a text chunk.
+ * Behavior is undefined if the native index does not fall within the
+ *   current chunk.
+ *
+ * This function is required only for text providers that do not use native UTF-16 indexes.
+ *
+ * @param ut          The UText containing the text chunk.
+ * @param nativeIndex Absolute (native) text index, chunk->start<=index<=chunk->limit.
+ * @return            Chunk-relative UTF-16 offset corresponding to the specified native
+ *                    index.
+ *
+ * @stable ICU 3.4
+ */
+typedef int32_t U_CALLCONV
+UTextMapNativeIndexToUTF16(const UText *ut, int64_t nativeIndex);
+
+
+/**
+ * Function type declaration for UText.utextClose().
+ *
+ * A Text Provider close function is only required for provider types that make
+ *  allocations in their open function (or other functions) that must be 
+ *  cleaned when the UText is closed.
+ *
+ * The allocation of the UText struct itself and any "extra" storage
+ * associated with the UText is handled by the common UText implementation
+ * and does not require provider specific cleanup in a close function.
+ *
+ * Most UText provider implementations do not need to implement this function.
+ *
+ * @param ut A UText object to be closed.
+ *
+ * @stable ICU 3.4
+ */
+typedef void U_CALLCONV
+UTextClose(UText *ut);
+
+
+/**
+  *   (public)  Function dispatch table for UText.
+  *             Conceptually very much like a C++ Virtual Function Table.
+  *             This struct defines the organization of the table.
+  *             Each text provider implementation must provide an
+  *              actual table that is initialized with the appropriate functions
+  *              for the type of text being handled.
+  *   @stable ICU 3.6
+  */
+struct UTextFuncs {
+    /**
+     *   (public)  Function table size, sizeof(UTextFuncs)
+     *             Intended for use should the table grow to accommodate added
+     *             functions in the future, to allow tests for older format
+     *             function tables that do not contain the extensions.
+     *
+     *             Fields are placed for optimal alignment on
+     *             32/64/128-bit-pointer machines, by normally grouping together
+     *             4 32-bit fields,
+     *             4 pointers,
+     *             2 64-bit fields
+     *             in sequence.
+     *   @stable ICU 3.6
+     */
+    int32_t       tableSize;
+
+    /**
+      *   (private)  Alignment padding.
+      *              Do not use, reserved for use by the UText framework only.
+      *   @internal
+      */
+    int32_t       reserved1, /** @internal */ reserved2, /** @internal */ reserved3;
+
+
+    /**
+     * (public) Function pointer for UTextClone
+     *
+     * @see UTextClone
+     * @stable ICU 3.6
+     */
+    UTextClone *clone;
+
+    /**
+     * (public) function pointer for UTextLength
+     * May be expensive to compute!
+     *
+     * @see UTextLength
+     * @stable ICU 3.6
+     */
+    UTextNativeLength *nativeLength;
+
+    /**
+     * (public) Function pointer for UTextAccess.
+     *
+     * @see UTextAccess
+     * @stable ICU 3.6
+     */
+    UTextAccess *access;
+
+    /**
+     * (public) Function pointer for UTextExtract.
+     *
+     * @see UTextExtract
+     * @stable ICU 3.6
+     */
+    UTextExtract *extract;
+
+    /**
+     * (public) Function pointer for UTextReplace.
+     *
+     * @see UTextReplace
+     * @stable ICU 3.6
+     */
+    UTextReplace *replace;
+
+    /**
+     * (public) Function pointer for UTextCopy.
+     *
+     * @see UTextCopy
+     * @stable ICU 3.6
+     */
+    UTextCopy *copy;
+
+    /**
+     * (public) Function pointer for UTextMapOffsetToNative.
+     *
+     * @see UTextMapOffsetToNative
+     * @stable ICU 3.6
+     */
+    UTextMapOffsetToNative *mapOffsetToNative;
+
+    /**
+     * (public) Function pointer for UTextMapNativeIndexToUTF16.
+     *
+     * @see UTextMapNativeIndexToUTF16
+     * @stable ICU 3.6
+     */
+    UTextMapNativeIndexToUTF16 *mapNativeIndexToUTF16;
+
+    /**
+     * (public) Function pointer for UTextClose.
+      *
+      * @see UTextClose
+      * @stable ICU 3.6
+      */
+    UTextClose  *close;
+
+    /**
+      * (private)  Spare function pointer
+      * @internal
+      */
+    UTextClose  *spare1;
+    
+    /**
+      * (private)  Spare function pointer
+      * @internal
+      */
+    UTextClose  *spare2;
+
+    /**
+      * (private)  Spare function pointer
+      * @internal
+      */
+    UTextClose  *spare3;
+
+};
+/**
+ * Function dispatch table for UText
+ * @see UTextFuncs
+ */
+typedef struct UTextFuncs UTextFuncs;
+
+ /**
+  *   UText struct.  Provides the interface between the generic UText access code
+  *                  and the UText provider code that works on specific kinds of
+  *                  text  (UTF-8, noncontiguous UTF-16, whatever.)
+  *
+  *                  Applications that are using predefined types of text providers
+  *                  to pass text data to ICU services will have no need to view the
+  *                  internals of the UText structs that they open.
+  *
+  * @stable ICU 3.6
+  */
+struct UText {
+    /**
+     *     (private)  Magic.  Used to help detect when UText functions are handed
+     *                        invalid or uninitialized UText structs.
+     *                        utext_openXYZ() functions take an initialized,
+     *                        but not necessarily open, UText struct as an
+     *                        optional fill-in parameter.  This magic field
+     *                        is used to check for that initialization.
+     *                        Text provider close functions must NOT clear
+     *                        the magic field because that would prevent
+     *                        reuse of the UText struct.
+     * @internal
+     */
+    uint32_t       magic;
+
+
+    /**
+     *     (private)  Flags for managing the allocation and freeing of
+     *                memory associated with this UText.
+     * @internal
+     */
+    int32_t        flags;
+
+
+    /**
+      *  Text provider properties.  This set of flags is maintained by the
+      *                             text provider implementation.
+      *  @stable ICU 3.4
+      */
+    int32_t         providerProperties;
+
+    /**
+     * (public) sizeOfStruct=sizeof(UText)
+     * Allows possible backward compatible extension.
+     *
+     * @stable ICU 3.4
+     */
+    int32_t         sizeOfStruct;
+    
+    /* ------ 16 byte alignment boundary -----------  */
+    
+
+    /**
+      *  (protected) Native index of the first character position following
+      *              the current chunk.
+      *  @stable ICU 3.6
+      */
+    int64_t         chunkNativeLimit;
+
+    /**
+     *   (protected)  Size in bytes of the extra space (pExtra).
+     *  @stable ICU 3.4
+     */
+    int32_t        extraSize;
+
+    /**
+      *    (protected) The highest chunk offset where native indexing and
+      *    chunk (UTF-16) indexing correspond.  For UTF-16 sources, value
+      *    will be equal to chunkLength.
+      *
+      *    @stable ICU 3.6
+      */
+    int32_t         nativeIndexingLimit;
+
+    /* ---- 16 byte alignment boundary------ */
+    
+    /**
+     *  (protected) Native index of the first character in the text chunk.
+     *  @stable ICU 3.6
+     */
+    int64_t         chunkNativeStart;
+
+    /**
+     *  (protected) Current iteration position within the text chunk (UTF-16 buffer).
+     *  This is the index to the character that will be returned by utext_next32().
+     *  @stable ICU 3.6
+     */
+    int32_t         chunkOffset;
+
+    /**
+     *  (protected) Length the text chunk (UTF-16 buffer), in UChars.
+     *  @stable ICU 3.6
+     */
+    int32_t         chunkLength;
+
+    /* ---- 16  byte alignment boundary-- */
+    
+
+    /**
+     *  (protected)  pointer to a chunk of text in UTF-16 format.
+     *  May refer either to original storage of the source of the text, or
+     *  if conversion was required, to a buffer owned by the UText.
+     *  @stable ICU 3.6
+     */
+    const UChar    *chunkContents;
+
+     /**
+      * (public)     Pointer to Dispatch table for accessing functions for this UText.
+      * @stable ICU 3.6
+      */
+    const UTextFuncs     *pFuncs;
+
+    /**
+     *  (protected)  Pointer to additional space requested by the
+     *               text provider during the utext_open operation.
+     * @stable ICU 3.4
+     */
+    void          *pExtra;
+
+    /**
+     * (protected) Pointer to string or text-containing object or similar.
+     * This is the source of the text that this UText is wrapping, in a format
+     *  that is known to the text provider functions.
+     * @stable ICU 3.4
+     */
+    const void   *context;
+
+    /* --- 16 byte alignment boundary--- */
+
+    /**
+     * (protected) Pointer fields available for use by the text provider.
+     * Not used by UText common code.
+     * @stable ICU 3.6
+     */
+    const void     *p; 
+    /**
+     * (protected) Pointer fields available for use by the text provider.
+     * Not used by UText common code.
+     * @stable ICU 3.6
+     */
+    const void     *q;
+     /**
+     * (protected) Pointer fields available for use by the text provider.
+     * Not used by UText common code.
+     * @stable ICU 3.6
+      */
+    const void     *r;
+
+    /**
+      *  Private field reserved for future use by the UText framework
+      *     itself.  This is not to be touched by the text providers.
+      * @internal ICU 3.4
+      */
+    void           *privP;
+
+
+    /* --- 16 byte alignment boundary--- */
+    
+
+    /**
+      * (protected) Integer field reserved for use by the text provider.
+      * Not used by the UText framework, or by the client (user) of the UText.
+      * @stable ICU 3.4
+      */
+    int64_t         a;
+
+    /**
+      * (protected) Integer field reserved for use by the text provider.
+      * Not used by the UText framework, or by the client (user) of the UText.
+      * @stable ICU 3.4
+      */
+    int32_t         b;
+
+    /**
+      * (protected) Integer field reserved for use by the text provider.
+      * Not used by the UText framework, or by the client (user) of the UText.
+      * @stable ICU 3.4
+      */
+    int32_t         c;
+
+    /*  ---- 16 byte alignment boundary---- */
+
+
+    /**
+      *  Private field reserved for future use by the UText framework
+      *     itself.  This is not to be touched by the text providers.
+      * @internal ICU 3.4
+      */
+    int64_t         privA;
+    /**
+      *  Private field reserved for future use by the UText framework
+      *     itself.  This is not to be touched by the text providers.
+      * @internal ICU 3.4
+      */
+    int32_t         privB;
+    /**
+      *  Private field reserved for future use by the UText framework
+      *     itself.  This is not to be touched by the text providers.
+      * @internal ICU 3.4
+      */
+    int32_t         privC;
+};
+
+
+/**
+ *  Common function for use by Text Provider implementations to allocate and/or initialize
+ *  a new UText struct.  To be called in the implementation of utext_open() functions.
+ *  If the supplied UText parameter is null, a new UText struct will be allocated on the heap.
+ *  If the supplied UText is already open, the provider's close function will be called
+ *  so that the struct can be reused by the open that is in progress.
+ *
+ * @param ut   pointer to a UText struct to be re-used, or null if a new UText
+ *             should be allocated.
+ * @param extraSpace The amount of additional space to be allocated as part
+ *             of this UText, for use by types of providers that require
+ *             additional storage.
+ * @param status Errors are returned here.
+ * @return pointer to the UText, allocated if necessary, with extra space set up if requested.
+ * @stable ICU 3.4
+ */
+U_CAPI UText * U_EXPORT2
+utext_setup(UText *ut, int32_t extraSpace, UErrorCode *status);
+
+// do not use #ifndef U_HIDE_INTERNAL_API around the following!
+/**
+  * @internal
+  *  Value used to help identify correctly initialized UText structs.
+  *  Note:  must be publicly visible so that UTEXT_INITIALIZER can access it.
+  */
+enum {
+    UTEXT_MAGIC = 0x345ad82c
+};
+
+/**
+ * initializer to be used with local (stack) instances of a UText
+ *  struct.  UText structs must be initialized before passing
+ *  them to one of the utext_open functions.
+ *
+ * @stable ICU 3.6
+ */
+#define UTEXT_INITIALIZER {                                        \
+                  UTEXT_MAGIC,          /* magic                */ \
+                  0,                    /* flags                */ \
+                  0,                    /* providerProps        */ \
+                  sizeof(UText),        /* sizeOfStruct         */ \
+                  0,                    /* chunkNativeLimit     */ \
+                  0,                    /* extraSize            */ \
+                  0,                    /* nativeIndexingLimit  */ \
+                  0,                    /* chunkNativeStart     */ \
+                  0,                    /* chunkOffset          */ \
+                  0,                    /* chunkLength          */ \
+                  NULL,                 /* chunkContents        */ \
+                  NULL,                 /* pFuncs               */ \
+                  NULL,                 /* pExtra               */ \
+                  NULL,                 /* context              */ \
+                  NULL, NULL, NULL,     /* p, q, r              */ \
+                  NULL,                 /* privP                */ \
+                  0, 0, 0,              /* a, b, c              */ \
+                  0, 0, 0               /* privA,B,C,           */ \
+                  }
+
+
+U_CDECL_END
+
+
+#if U_SHOW_CPLUSPLUS_API
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class LocalUTextPointer
+ * "Smart pointer" class, closes a UText via utext_close().
+ * For most methods see the LocalPointerBase base class.
+ *
+ * @see LocalPointerBase
+ * @see LocalPointer
+ * @stable ICU 4.4
+ */
+U_DEFINE_LOCAL_OPEN_POINTER(LocalUTextPointer, UText, utext_close);
+
+U_NAMESPACE_END
+
+#endif
+
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/utf.h b/thirdparty/icu4c/common/unicode/utf.h
new file mode 100644
index 0000000000..c9d5f5785c
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utf.h
@@ -0,0 +1,225 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep09
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: Code point macros
+ *
+ * This file defines macros for checking whether a code point is
+ * a surrogate or a non-character etc.
+ *
+ * If U_NO_DEFAULT_INCLUDE_UTF_HEADERS is 0 then utf.h is included by utypes.h
+ * and itself includes utf8.h and utf16.h after some
+ * common definitions.
+ * If U_NO_DEFAULT_INCLUDE_UTF_HEADERS is 1 then each of these headers must be
+ * included explicitly if their definitions are used.
+ *
+ * utf8.h and utf16.h define macros for efficiently getting code points
+ * in and out of UTF-8/16 strings.
+ * utf16.h macros have "U16_" prefixes.
+ * utf8.h defines similar macros with "U8_" prefixes for UTF-8 string handling.
+ *
+ * ICU mostly processes 16-bit Unicode strings.
+ * Most of the time, such strings are well-formed UTF-16.
+ * Single, unpaired surrogates must be handled as well, and are treated in ICU
+ * like regular code points where possible.
+ * (Pairs of surrogate code points are indistinguishable from supplementary
+ * code points encoded as pairs of supplementary code units.)
+ *
+ * In fact, almost all Unicode code points in normal text (>99%)
+ * are on the BMP (<=U+ffff) and even <=U+d7ff.
+ * ICU functions handle supplementary code points (U+10000..U+10ffff)
+ * but are optimized for the much more frequently occurring BMP code points.
+ *
+ * umachine.h defines UChar to be an unsigned 16-bit integer.
+ * Since ICU 59, ICU uses char16_t in C++, UChar only in C,
+ * and defines UChar=char16_t by default. See the UChar API docs for details.
+ *
+ * UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit
+ * Unicode code point (Unicode scalar value, 0..0x10ffff) and U_SENTINEL (-1).
+ * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as
+ * the definition of UChar. For details see the documentation for UChar32 itself.
+ *
+ * utf.h defines a small number of C macros for single Unicode code points.
+ * These are simple checks for surrogates and non-characters.
+ * For actual Unicode character properties see uchar.h.
+ *
+ * By default, string operations must be done with error checking in case
+ * a string is not well-formed UTF-16 or UTF-8.
+ *
+ * The U16_ macros detect if a surrogate code unit is unpaired
+ * (lead unit without trail unit or vice versa) and just return the unit itself
+ * as the code point.
+ *
+ * The U8_ macros detect illegal byte sequences and return a negative value.
+ * Starting with ICU 60, the observable length of a single illegal byte sequence
+ * skipped by one of these macros follows the Unicode 6+ recommendation
+ * which is consistent with the W3C Encoding Standard.
+ *
+ * There are ..._OR_FFFD versions of both U16_ and U8_ macros
+ * that return U+FFFD for illegal code unit sequences.
+ *
+ * The regular "safe" macros require that the initial, passed-in string index
+ * is within bounds. They only check the index when they read more than one
+ * code unit. This is usually done with code similar to the following loop:
+ * <pre>while(i<length) {
+ *   U16_NEXT(s, i, length, c);
+ *   // use c
+ * }</pre>
+ *
+ * When it is safe to assume that text is well-formed UTF-16
+ * (does not contain single, unpaired surrogates), then one can use
+ * U16_..._UNSAFE macros.
+ * These do not check for proper code unit sequences or truncated text and may
+ * yield wrong results or even cause a crash if they are used with "malformed"
+ * text.
+ * In practice, U16_..._UNSAFE macros will produce slightly less code but
+ * should not be faster because the processing is only different when a
+ * surrogate code unit is detected, which will be rare.
+ *
+ * Similarly for UTF-8, there are "safe" macros without a suffix,
+ * and U8_..._UNSAFE versions.
+ * The performance differences are much larger here because UTF-8 provides so
+ * many opportunities for malformed sequences.
+ * The unsafe UTF-8 macros are entirely implemented inside the macro definitions
+ * and are fast, while the safe UTF-8 macros call functions for some complicated cases.
+ *
+ * Unlike with UTF-16, malformed sequences cannot be expressed with distinct
+ * code point values (0..U+10ffff). They are indicated with negative values instead.
+ *
+ * For more information see the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * <em>Usage:</em>
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ *
+ * @stable ICU 2.4
+ */
+
+#ifndef __UTF_H__
+#define __UTF_H__
+
+#include "unicode/umachine.h"
+/* include the utfXX.h after the following definitions */
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Is this code point a Unicode noncharacter?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_UNICODE_NONCHAR(c) \
+    ((c)>=0xfdd0 && \
+     ((c)<=0xfdef || ((c)&0xfffe)==0xfffe) && (c)<=0x10ffff)
+
+/**
+ * Is c a Unicode code point value (0..U+10ffff)
+ * that can be assigned a character?
+ *
+ * Code points that are not characters include:
+ * - single surrogate code points (U+d800..U+dfff, 2048 code points)
+ * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points)
+ * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points)
+ * - the highest Unicode code point value is U+10ffff
+ *
+ * This means that all code points below U+d800 are character code points,
+ * and that boundary is tested first for performance.
+ *
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_UNICODE_CHAR(c) \
+    ((uint32_t)(c)<0xd800 || \
+        (0xdfff<(c) && (c)<=0x10ffff && !U_IS_UNICODE_NONCHAR(c)))
+
+/**
+ * Is this code point a BMP code point (U+0000..U+ffff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.8
+ */
+#define U_IS_BMP(c) ((uint32_t)(c)<=0xffff)
+
+/**
+ * Is this code point a supplementary code point (U+10000..U+10ffff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.8
+ */
+#define U_IS_SUPPLEMENTARY(c) ((uint32_t)((c)-0x10000)<=0xfffff)
+ 
+/**
+ * Is this code point a lead surrogate (U+d800..U+dbff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_LEAD(c) (((c)&0xfffffc00)==0xd800)
+
+/**
+ * Is this code point a trail surrogate (U+dc00..U+dfff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00)
+
+/**
+ * Is this code point a surrogate (U+d800..U+dfff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_SURROGATE(c) (((c)&0xfffff800)==0xd800)
+
+/**
+ * Assuming c is a surrogate code point (U_IS_SURROGATE(c)),
+ * is it a lead surrogate?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_SURROGATE_LEAD(c) (((c)&0x400)==0)
+
+/**
+ * Assuming c is a surrogate code point (U_IS_SURROGATE(c)),
+ * is it a trail surrogate?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 4.2
+ */
+#define U_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0)
+
+/* include the utfXX.h ------------------------------------------------------ */
+
+#if !U_NO_DEFAULT_INCLUDE_UTF_HEADERS
+
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+
+/* utf_old.h contains deprecated, pre-ICU 2.4 definitions */
+#include "unicode/utf_old.h"
+
+#endif  /* !U_NO_DEFAULT_INCLUDE_UTF_HEADERS */
+
+#endif  /* __UTF_H__ */
diff --git a/thirdparty/icu4c/common/unicode/utf16.h b/thirdparty/icu4c/common/unicode/utf16.h
new file mode 100644
index 0000000000..3902c60e95
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utf16.h
@@ -0,0 +1,734 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf16.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep09
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: 16-bit Unicode handling macros
+ * 
+ * This file defines macros to deal with 16-bit Unicode (UTF-16) code units and strings.
+ *
+ * For more information see utf.h and the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * <em>Usage:</em>
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ */
+
+#ifndef __UTF16_H__
+#define __UTF16_H__
+
+#include <stdbool.h>
+#include "unicode/umachine.h"
+#ifndef __UTF_H__
+#   include "unicode/utf.h"
+#endif
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Does this code unit alone encode a code point (BMP, not a surrogate)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SINGLE(c) !U_IS_SURROGATE(c)
+
+/**
+ * Is this code unit a lead surrogate (U+d800..U+dbff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_LEAD(c) (((c)&0xfffffc00)==0xd800)
+
+/**
+ * Is this code unit a trail surrogate (U+dc00..U+dfff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00)
+
+/**
+ * Is this code unit a surrogate (U+d800..U+dfff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SURROGATE(c) U_IS_SURROGATE(c)
+
+/**
+ * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)),
+ * is it a lead surrogate?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SURROGATE_LEAD(c) (((c)&0x400)==0)
+
+/**
+ * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)),
+ * is it a trail surrogate?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 4.2
+ */
+#define U16_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0)
+
+/**
+ * Helper constant for U16_GET_SUPPLEMENTARY.
+ * @internal
+ */
+#define U16_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000)
+
+/**
+ * Get a supplementary code point value (U+10000..U+10ffff)
+ * from its lead and trail surrogates.
+ * The result is undefined if the input values are not
+ * lead and trail surrogates.
+ *
+ * @param lead lead surrogate (U+d800..U+dbff)
+ * @param trail trail surrogate (U+dc00..U+dfff)
+ * @return supplementary code point (U+10000..U+10ffff)
+ * @stable ICU 2.4
+ */
+#define U16_GET_SUPPLEMENTARY(lead, trail) \
+    (((UChar32)(lead)<<10UL)+(UChar32)(trail)-U16_SURROGATE_OFFSET)
+
+
+/**
+ * Get the lead surrogate (0xd800..0xdbff) for a
+ * supplementary code point (0x10000..0x10ffff).
+ * @param supplementary 32-bit code point (U+10000..U+10ffff)
+ * @return lead surrogate (U+d800..U+dbff) for supplementary
+ * @stable ICU 2.4
+ */
+#define U16_LEAD(supplementary) (UChar)(((supplementary)>>10)+0xd7c0)
+
+/**
+ * Get the trail surrogate (0xdc00..0xdfff) for a
+ * supplementary code point (0x10000..0x10ffff).
+ * @param supplementary 32-bit code point (U+10000..U+10ffff)
+ * @return trail surrogate (U+dc00..U+dfff) for supplementary
+ * @stable ICU 2.4
+ */
+#define U16_TRAIL(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00)
+
+/**
+ * How many 16-bit code units are used to encode this Unicode code point? (1 or 2)
+ * The result is not defined if c is not a Unicode code point (U+0000..U+10ffff).
+ * @param c 32-bit code point
+ * @return 1 or 2
+ * @stable ICU 2.4
+ */
+#define U16_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2)
+
+/**
+ * The maximum number of 16-bit code units per Unicode code point (U+0000..U+10ffff).
+ * @return 2
+ * @stable ICU 2.4
+ */
+#define U16_MAX_LENGTH 2
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ * The result is undefined if the offset points to a single, unpaired surrogate.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_GET
+ * @stable ICU 2.4
+ */
+#define U16_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(U16_IS_SURROGATE(c)) { \
+        if(U16_IS_SURROGATE_LEAD(c)) { \
+            (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)+1]); \
+        } else { \
+            (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to that unpaired surrogate.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_GET_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_GET(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if(U16_IS_SURROGATE_LEAD(c)) { \
+            if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \
+                (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+            } \
+        } else { \
+            if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+                (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to U+FFFD.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT_OR_FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_GET_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_GET_OR_FFFD(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if(U16_IS_SURROGATE_LEAD(c)) { \
+            if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \
+                (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+            } else { \
+                (c)=0xfffd; \
+            } \
+        } else { \
+            if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+                (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } else { \
+                (c)=0xfffd; \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with forward iteration --------------------------------------- */
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset points to a single, unpaired lead surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_NEXT
+ * @stable ICU 2.4
+ */
+#define U16_NEXT_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(U16_IS_LEAD(c)) { \
+        (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate or
+ * to a single, unpaired lead surrogate, then c is set to that unpaired surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_NEXT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_NEXT(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(U16_IS_LEAD(c)) { \
+        uint16_t __c2; \
+        if((i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \
+            ++(i); \
+            (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate or
+ * to a single, unpaired lead surrogate, then c is set to U+FFFD.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_NEXT_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_NEXT_OR_FFFD(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if(U16_IS_SURROGATE_LEAD(c) && (i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \
+            ++(i); \
+            (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+        } else { \
+            (c)=0xfffd; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 or 2 code units.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Unsafe" macro, assumes a valid code point and sufficient space in the string.
+ * Otherwise, the result is undefined.
+ *
+ * @param s const UChar * string buffer
+ * @param i string offset
+ * @param c code point to append
+ * @see U16_APPEND
+ * @stable ICU 2.4
+ */
+#define U16_APPEND_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0xffff) { \
+        (s)[(i)++]=(uint16_t)(c); \
+    } else { \
+        (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
+        (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 or 2 code units.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Safe" macro, checks for a valid code point.
+ * If a surrogate pair is written, checks for sufficient space in the string.
+ * If the code point is not valid or a trail surrogate does not fit,
+ * then isError is set to true.
+ *
+ * @param s const UChar * string buffer
+ * @param i string offset, must be i<capacity
+ * @param capacity size of the string buffer
+ * @param c code point to append
+ * @param isError output UBool set to true if an error occurs, otherwise not modified
+ * @see U16_APPEND_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_APPEND(s, i, capacity, c, isError) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0xffff) { \
+        (s)[(i)++]=(uint16_t)(c); \
+    } else if((uint32_t)(c)<=0x10ffff && (i)+1<(capacity)) { \
+        (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
+        (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+    } else /* c>0x10ffff or not enough space */ { \
+        (isError)=true; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_FWD_1
+ * @stable ICU 2.4
+ */
+#define U16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_LEAD((s)[(i)++])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i<length
+ * @param length string length
+ * @see U16_FWD_1_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_FWD_1(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_LEAD((s)[(i)++]) && (i)!=(length) && U16_IS_TRAIL((s)[i])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U16_FWD_N
+ * @stable ICU 2.4
+ */
+#define U16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U16_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param i int32_t string offset, must be i<length
+ * @param length int32_t string length
+ * @param n number of code points to skip
+ * @see U16_FWD_N_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_FWD_N(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \
+        U16_FWD_1(s, i, length); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to the trail surrogate of a surrogate pair,
+ * then the offset is decremented.
+ * Otherwise, it is not modified.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_SET_CP_START
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[i])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to the trail surrogate of a surrogate pair,
+ * then the offset is decremented.
+ * Otherwise, it is not modified.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i
+ * @see U16_SET_CP_START_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with backward iteration -------------------------------------- */
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset is behind a single, unpaired trail surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_PREV
+ * @stable ICU 2.4
+ */
+#define U16_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(U16_IS_TRAIL(c)) { \
+        (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate or behind a single, unpaired
+ * trail surrogate, then c is set to that unpaired surrogate.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<i
+ * @param c output UChar32 variable
+ * @see U16_PREV_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_PREV(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(U16_IS_TRAIL(c)) { \
+        uint16_t __c2; \
+        if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate or behind a single, unpaired
+ * trail surrogate, then c is set to U+FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<i
+ * @param c output UChar32 variable
+ * @see U16_PREV_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_PREV_OR_FFFD(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if(U16_IS_SURROGATE_TRAIL(c) && (i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+        } else { \
+            (c)=0xfffd; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_BACK_1
+ * @stable ICU 2.4
+ */
+#define U16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[--(i)])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<i
+ * @see U16_BACK_1_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_BACK_1(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[--(i)]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U16_BACK_N
+ * @stable ICU 2.4
+ */
+#define U16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U16_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start start of string
+ * @param i string offset, must be start<i
+ * @param n number of code points to skip
+ * @see U16_BACK_N_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_BACK_N(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0 && (i)>(start)) { \
+        U16_BACK_1(s, start, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind the lead surrogate of a surrogate pair,
+ * then the offset is incremented.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_SET_CP_LIMIT
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_LEAD((s)[(i)-1])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind the lead surrogate of a surrogate pair,
+ * then the offset is incremented.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, start<=i<=length
+ * @param length int32_t string length
+ * @see U16_SET_CP_LIMIT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((start)<(i) && ((i)<(length) || (length)<0) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/utf32.h b/thirdparty/icu4c/common/unicode/utf32.h
new file mode 100644
index 0000000000..8822c4dd09
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utf32.h
@@ -0,0 +1,25 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2001, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf32.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep20
+*   created by: Markus W. Scherer
+*/
+/**
+ * \file
+ * \brief C API: UTF-32 macros
+ *
+ * This file is obsolete and its contents moved to utf_old.h.
+ * See utf_old.h and Jitterbug 2150 and its discussion on the ICU mailing list
+ * in September 2002.
+ */
diff --git a/thirdparty/icu4c/common/unicode/utf8.h b/thirdparty/icu4c/common/unicode/utf8.h
new file mode 100644
index 0000000000..5a07435fcf
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utf8.h
@@ -0,0 +1,882 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf8.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep13
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: 8-bit Unicode handling macros
+ * 
+ * This file defines macros to deal with 8-bit Unicode (UTF-8) code units (bytes) and strings.
+ *
+ * For more information see utf.h and the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * <em>Usage:</em>
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ */
+
+#ifndef __UTF8_H__
+#define __UTF8_H__
+
+#include <stdbool.h>
+#include "unicode/umachine.h"
+#ifndef __UTF_H__
+#   include "unicode/utf.h"
+#endif
+
+/* internal definitions ----------------------------------------------------- */
+
+/**
+ * Counts the trail bytes for a UTF-8 lead byte.
+ * Returns 0 for 0..0xc1 as well as for 0xf5..0xff.
+ * leadByte might be evaluated multiple times.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this file and thus must remain stable.
+ *
+ * @param leadByte The first byte of a UTF-8 sequence. Must be 0..0xff.
+ * @internal
+ */
+#define U8_COUNT_TRAIL_BYTES(leadByte) \
+    (U8_IS_LEAD(leadByte) ? \
+        ((uint8_t)(leadByte)>=0xe0)+((uint8_t)(leadByte)>=0xf0)+1 : 0)
+
+/**
+ * Counts the trail bytes for a UTF-8 lead byte of a valid UTF-8 sequence.
+ * Returns 0 for 0..0xc1. Undefined for 0xf5..0xff.
+ * leadByte might be evaluated multiple times.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this file and thus must remain stable.
+ *
+ * @param leadByte The first byte of a UTF-8 sequence. Must be 0..0xff.
+ * @internal
+ */
+#define U8_COUNT_TRAIL_BYTES_UNSAFE(leadByte) \
+    (((uint8_t)(leadByte)>=0xc2)+((uint8_t)(leadByte)>=0xe0)+((uint8_t)(leadByte)>=0xf0))
+
+/**
+ * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this file and thus must remain stable.
+ * @internal
+ */
+#define U8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1)
+
+/**
+ * Internal bit vector for 3-byte UTF-8 validity check, for use in U8_IS_VALID_LEAD3_AND_T1.
+ * Each bit indicates whether one lead byte + first trail byte pair starts a valid sequence.
+ * Lead byte E0..EF bits 3..0 are used as byte index,
+ * first trail byte bits 7..5 are used as bit index into that byte.
+ * @see U8_IS_VALID_LEAD3_AND_T1
+ * @internal
+ */
+#define U8_LEAD3_T1_BITS "\x20\x30\x30\x30\x30\x30\x30\x30\x30\x30\x30\x30\x30\x10\x30\x30"
+
+/**
+ * Internal 3-byte UTF-8 validity check.
+ * Non-zero if lead byte E0..EF and first trail byte 00..FF start a valid sequence.
+ * @internal
+ */
+#define U8_IS_VALID_LEAD3_AND_T1(lead, t1) (U8_LEAD3_T1_BITS[(lead)&0xf]&(1<<((uint8_t)(t1)>>5)))
+
+/**
+ * Internal bit vector for 4-byte UTF-8 validity check, for use in U8_IS_VALID_LEAD4_AND_T1.
+ * Each bit indicates whether one lead byte + first trail byte pair starts a valid sequence.
+ * First trail byte bits 7..4 are used as byte index,
+ * lead byte F0..F4 bits 2..0 are used as bit index into that byte.
+ * @see U8_IS_VALID_LEAD4_AND_T1
+ * @internal
+ */
+#define U8_LEAD4_T1_BITS "\x00\x00\x00\x00\x00\x00\x00\x00\x1E\x0F\x0F\x0F\x00\x00\x00\x00"
+
+/**
+ * Internal 4-byte UTF-8 validity check.
+ * Non-zero if lead byte F0..F4 and first trail byte 00..FF start a valid sequence.
+ * @internal
+ */
+#define U8_IS_VALID_LEAD4_AND_T1(lead, t1) (U8_LEAD4_T1_BITS[(uint8_t)(t1)>>4]&(1<<((lead)&7)))
+
+/**
+ * Function for handling "next code point" with error-checking.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this
+ * file and thus must remain stable, and should not be hidden when other internal
+ * functions are hidden (otherwise public macros would fail to compile).
+ * @internal
+ */
+U_CAPI UChar32 U_EXPORT2
+utf8_nextCharSafeBody(const uint8_t *s, int32_t *pi, int32_t length, UChar32 c, UBool strict);
+
+/**
+ * Function for handling "append code point" with error-checking.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this
+ * file and thus must remain stable, and should not be hidden when other internal
+ * functions are hidden (otherwise public macros would fail to compile).
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utf8_appendCharSafeBody(uint8_t *s, int32_t i, int32_t length, UChar32 c, UBool *pIsError);
+
+/**
+ * Function for handling "previous code point" with error-checking.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this
+ * file and thus must remain stable, and should not be hidden when other internal
+ * functions are hidden (otherwise public macros would fail to compile).
+ * @internal
+ */
+U_CAPI UChar32 U_EXPORT2
+utf8_prevCharSafeBody(const uint8_t *s, int32_t start, int32_t *pi, UChar32 c, UBool strict);
+
+/**
+ * Function for handling "skip backward one code point" with error-checking.
+ *
+ * This is internal since it is not meant to be called directly by external clients;
+ * however it is called by public macros in this
+ * file and thus must remain stable, and should not be hidden when other internal
+ * functions are hidden (otherwise public macros would fail to compile).
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utf8_back1SafeBody(const uint8_t *s, int32_t start, int32_t i);
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Does this code unit (byte) encode a code point by itself (US-ASCII 0..0x7f)?
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_SINGLE(c) (((c)&0x80)==0)
+
+/**
+ * Is this code unit (byte) a UTF-8 lead byte? (0xC2..0xF4)
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_LEAD(c) ((uint8_t)((c)-0xc2)<=0x32)
+// 0x32=0xf4-0xc2
+
+/**
+ * Is this code unit (byte) a UTF-8 trail byte? (0x80..0xBF)
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_TRAIL(c) ((int8_t)(c)<-0x40)
+
+/**
+ * How many code units (bytes) are used for the UTF-8 encoding
+ * of this Unicode code point?
+ * @param c 32-bit code point
+ * @return 1..4, or 0 if c is a surrogate or not a Unicode code point
+ * @stable ICU 2.4
+ */
+#define U8_LENGTH(c) \
+    ((uint32_t)(c)<=0x7f ? 1 : \
+        ((uint32_t)(c)<=0x7ff ? 2 : \
+            ((uint32_t)(c)<=0xd7ff ? 3 : \
+                ((uint32_t)(c)<=0xdfff || (uint32_t)(c)>0x10ffff ? 0 : \
+                    ((uint32_t)(c)<=0xffff ? 3 : 4)\
+                ) \
+            ) \
+        ) \
+    )
+
+/**
+ * The maximum number of UTF-8 code units (bytes) per Unicode code point (U+0000..U+10ffff).
+ * @return 4
+ * @stable ICU 2.4
+ */
+#define U8_MAX_LENGTH 4
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * The offset may point to either the lead byte or one of the trail bytes
+ * for a code point, in which case the macro will read all of the bytes
+ * for the code point.
+ * The result is undefined if the offset points to an illegal UTF-8
+ * byte sequence.
+ * Iteration through a string is more efficient with U8_NEXT_UNSAFE or U8_NEXT.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U8_GET
+ * @stable ICU 2.4
+ */
+#define U8_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _u8_get_unsafe_index=(int32_t)(i); \
+    U8_SET_CP_START_UNSAFE(s, _u8_get_unsafe_index); \
+    U8_NEXT_UNSAFE(s, _u8_get_unsafe_index, c); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * The offset may point to either the lead byte or one of the trail bytes
+ * for a code point, in which case the macro will read all of the bytes
+ * for the code point.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to an illegal UTF-8 byte sequence, then
+ * c is set to a negative value.
+ * Iteration through a string is more efficient with U8_NEXT_UNSAFE or U8_NEXT.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset
+ * @param i int32_t string offset, must be start<=i<length
+ * @param length int32_t string length
+ * @param c output UChar32 variable, set to <0 in case of an error
+ * @see U8_GET_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_GET(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _u8_get_index=(i); \
+    U8_SET_CP_START(s, start, _u8_get_index); \
+    U8_NEXT(s, _u8_get_index, length, c); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * The offset may point to either the lead byte or one of the trail bytes
+ * for a code point, in which case the macro will read all of the bytes
+ * for the code point.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to an illegal UTF-8 byte sequence, then
+ * c is set to U+FFFD.
+ * Iteration through a string is more efficient with U8_NEXT_UNSAFE or U8_NEXT_OR_FFFD.
+ *
+ * This macro does not distinguish between a real U+FFFD in the text
+ * and U+FFFD returned for an ill-formed sequence.
+ * Use U8_GET() if that distinction is important.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset
+ * @param i int32_t string offset, must be start<=i<length
+ * @param length int32_t string length
+ * @param c output UChar32 variable, set to U+FFFD in case of an error
+ * @see U8_GET
+ * @stable ICU 51
+ */
+#define U8_GET_OR_FFFD(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _u8_get_index=(i); \
+    U8_SET_CP_START(s, start, _u8_get_index); \
+    U8_NEXT_OR_FFFD(s, _u8_get_index, length, c); \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with forward iteration --------------------------------------- */
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * The offset may point to the lead byte of a multi-byte sequence,
+ * in which case the macro will read the whole sequence.
+ * The result is undefined if the offset points to a trail byte
+ * or an illegal UTF-8 sequence.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U8_NEXT
+ * @stable ICU 2.4
+ */
+#define U8_NEXT_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[(i)++]; \
+    if(!U8_IS_SINGLE(c)) { \
+        if((c)<0xe0) { \
+            (c)=(((c)&0x1f)<<6)|((s)[(i)++]&0x3f); \
+        } else if((c)<0xf0) { \
+            /* no need for (c&0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */ \
+            (c)=(UChar)(((c)<<12)|(((s)[i]&0x3f)<<6)|((s)[(i)+1]&0x3f)); \
+            (i)+=2; \
+        } else { \
+            (c)=(((c)&7)<<18)|(((s)[i]&0x3f)<<12)|(((s)[(i)+1]&0x3f)<<6)|((s)[(i)+2]&0x3f); \
+            (i)+=3; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead byte of a multi-byte sequence,
+ * in which case the macro will read the whole sequence.
+ * If the offset points to a trail byte or an illegal UTF-8 sequence, then
+ * c is set to a negative value.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i<length
+ * @param length int32_t string length
+ * @param c output UChar32 variable, set to <0 in case of an error
+ * @see U8_NEXT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_NEXT(s, i, length, c) U8_INTERNAL_NEXT_OR_SUB(s, i, length, c, U_SENTINEL)
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead byte of a multi-byte sequence,
+ * in which case the macro will read the whole sequence.
+ * If the offset points to a trail byte or an illegal UTF-8 sequence, then
+ * c is set to U+FFFD.
+ *
+ * This macro does not distinguish between a real U+FFFD in the text
+ * and U+FFFD returned for an ill-formed sequence.
+ * Use U8_NEXT() if that distinction is important.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i<length
+ * @param length int32_t string length
+ * @param c output UChar32 variable, set to U+FFFD in case of an error
+ * @see U8_NEXT
+ * @stable ICU 51
+ */
+#define U8_NEXT_OR_FFFD(s, i, length, c) U8_INTERNAL_NEXT_OR_SUB(s, i, length, c, 0xfffd)
+
+/** @internal */
+#define U8_INTERNAL_NEXT_OR_SUB(s, i, length, c, sub) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[(i)++]; \
+    if(!U8_IS_SINGLE(c)) { \
+        uint8_t __t = 0; \
+        if((i)!=(length) && \
+            /* fetch/validate/assemble all but last trail byte */ \
+            ((c)>=0xe0 ? \
+                ((c)<0xf0 ?  /* U+0800..U+FFFF except surrogates */ \
+                    U8_LEAD3_T1_BITS[(c)&=0xf]&(1<<((__t=(s)[i])>>5)) && \
+                    (__t&=0x3f, 1) \
+                :  /* U+10000..U+10FFFF */ \
+                    ((c)-=0xf0)<=4 && \
+                    U8_LEAD4_T1_BITS[(__t=(s)[i])>>4]&(1<<(c)) && \
+                    ((c)=((c)<<6)|(__t&0x3f), ++(i)!=(length)) && \
+                    (__t=(s)[i]-0x80)<=0x3f) && \
+                /* valid second-to-last trail byte */ \
+                ((c)=((c)<<6)|__t, ++(i)!=(length)) \
+            :  /* U+0080..U+07FF */ \
+                (c)>=0xc2 && ((c)&=0x1f, 1)) && \
+            /* last trail byte */ \
+            (__t=(s)[i]-0x80)<=0x3f && \
+            ((c)=((c)<<6)|__t, ++(i), 1)) { \
+        } else { \
+            (c)=(sub);  /* ill-formed*/ \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 to 4 bytes.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Unsafe" macro, assumes a valid code point and sufficient space in the string.
+ * Otherwise, the result is undefined.
+ *
+ * @param s const uint8_t * string buffer
+ * @param i string offset
+ * @param c code point to append
+ * @see U8_APPEND
+ * @stable ICU 2.4
+ */
+#define U8_APPEND_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    uint32_t __uc=(c); \
+    if(__uc<=0x7f) { \
+        (s)[(i)++]=(uint8_t)__uc; \
+    } else { \
+        if(__uc<=0x7ff) { \
+            (s)[(i)++]=(uint8_t)((__uc>>6)|0xc0); \
+        } else { \
+            if(__uc<=0xffff) { \
+                (s)[(i)++]=(uint8_t)((__uc>>12)|0xe0); \
+            } else { \
+                (s)[(i)++]=(uint8_t)((__uc>>18)|0xf0); \
+                (s)[(i)++]=(uint8_t)(((__uc>>12)&0x3f)|0x80); \
+            } \
+            (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        } \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 to 4 bytes.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Safe" macro, checks for a valid code point.
+ * If a non-ASCII code point is written, checks for sufficient space in the string.
+ * If the code point is not valid or trail bytes do not fit,
+ * then isError is set to true.
+ *
+ * @param s const uint8_t * string buffer
+ * @param i int32_t string offset, must be i<capacity
+ * @param capacity int32_t size of the string buffer
+ * @param c UChar32 code point to append
+ * @param isError output UBool set to true if an error occurs, otherwise not modified
+ * @see U8_APPEND_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_APPEND(s, i, capacity, c, isError) UPRV_BLOCK_MACRO_BEGIN { \
+    uint32_t __uc=(c); \
+    if(__uc<=0x7f) { \
+        (s)[(i)++]=(uint8_t)__uc; \
+    } else if(__uc<=0x7ff && (i)+1<(capacity)) { \
+        (s)[(i)++]=(uint8_t)((__uc>>6)|0xc0); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else if((__uc<=0xd7ff || (0xe000<=__uc && __uc<=0xffff)) && (i)+2<(capacity)) { \
+        (s)[(i)++]=(uint8_t)((__uc>>12)|0xe0); \
+        (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else if(0xffff<__uc && __uc<=0x10ffff && (i)+3<(capacity)) { \
+        (s)[(i)++]=(uint8_t)((__uc>>18)|0xf0); \
+        (s)[(i)++]=(uint8_t)(((__uc>>12)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else { \
+        (isError)=true; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_FWD_1
+ * @stable ICU 2.4
+ */
+#define U8_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)+=1+U8_COUNT_TRAIL_BYTES_UNSAFE((s)[i]); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i<length
+ * @param length int32_t string length
+ * @see U8_FWD_1_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_FWD_1(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    uint8_t __b=(s)[(i)++]; \
+    if(U8_IS_LEAD(__b) && (i)!=(length)) { \
+        uint8_t __t1=(s)[i]; \
+        if((0xe0<=__b && __b<0xf0)) { \
+            if(U8_IS_VALID_LEAD3_AND_T1(__b, __t1) && \
+                    ++(i)!=(length) && U8_IS_TRAIL((s)[i])) { \
+                ++(i); \
+            } \
+        } else if(__b<0xe0) { \
+            if(U8_IS_TRAIL(__t1)) { \
+                ++(i); \
+            } \
+        } else /* c>=0xf0 */ { \
+            if(U8_IS_VALID_LEAD4_AND_T1(__b, __t1) && \
+                    ++(i)!=(length) && U8_IS_TRAIL((s)[i]) && \
+                    ++(i)!=(length) && U8_IS_TRAIL((s)[i])) { \
+                ++(i); \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U8_FWD_N
+ * @stable ICU 2.4
+ */
+#define U8_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U8_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i<length
+ * @param length int32_t string length
+ * @param n number of code points to skip
+ * @see U8_FWD_N_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_FWD_N(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \
+        U8_FWD_1(s, i, length); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to a UTF-8 trail byte,
+ * then the offset is moved backward to the corresponding lead byte.
+ * Otherwise, it is not modified.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_SET_CP_START
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    while(U8_IS_TRAIL((s)[i])) { --(i); } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to a UTF-8 trail byte,
+ * then the offset is moved backward to the corresponding lead byte.
+ * Otherwise, it is not modified.
+ *
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ * Unlike U8_TRUNCATE_IF_INCOMPLETE(), this macro always reads s[i].
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<=i
+ * @see U8_SET_CP_START_UNSAFE
+ * @see U8_TRUNCATE_IF_INCOMPLETE
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U8_IS_TRAIL((s)[(i)])) { \
+        (i)=utf8_back1SafeBody(s, start, (i)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * If the string ends with a UTF-8 byte sequence that is valid so far
+ * but incomplete, then reduce the length of the string to end before
+ * the lead byte of that incomplete sequence.
+ * For example, if the string ends with E1 80, the length is reduced by 2.
+ *
+ * In all other cases (the string ends with a complete sequence, or it is not
+ * possible for any further trail byte to extend the trailing sequence)
+ * the length remains unchanged.
+ *
+ * Useful for processing text split across multiple buffers
+ * (save the incomplete sequence for later)
+ * and for optimizing iteration
+ * (check for string length only once per character).
+ *
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ * Unlike U8_SET_CP_START(), this macro never reads s[length].
+ *
+ * (In UTF-16, simply check for U16_IS_LEAD(last code unit).)
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param length int32_t string length (usually start<=length)
+ * @see U8_SET_CP_START
+ * @stable ICU 61
+ */
+#define U8_TRUNCATE_IF_INCOMPLETE(s, start, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((length)>(start)) { \
+        uint8_t __b1=s[(length)-1]; \
+        if(U8_IS_SINGLE(__b1)) { \
+            /* common ASCII character */ \
+        } else if(U8_IS_LEAD(__b1)) { \
+            --(length); \
+        } else if(U8_IS_TRAIL(__b1) && ((length)-2)>=(start)) { \
+            uint8_t __b2=s[(length)-2]; \
+            if(0xe0<=__b2 && __b2<=0xf4) { \
+                if(__b2<0xf0 ? U8_IS_VALID_LEAD3_AND_T1(__b2, __b1) : \
+                        U8_IS_VALID_LEAD4_AND_T1(__b2, __b1)) { \
+                    (length)-=2; \
+                } \
+            } else if(U8_IS_TRAIL(__b2) && ((length)-3)>=(start)) { \
+                uint8_t __b3=s[(length)-3]; \
+                if(0xf0<=__b3 && __b3<=0xf4 && U8_IS_VALID_LEAD4_AND_T1(__b3, __b2)) { \
+                    (length)-=3; \
+                } \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with backward iteration -------------------------------------- */
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a multi-byte sequence, then the macro will read
+ * the whole sequence.
+ * If the offset is behind a lead byte, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset is behind an illegal UTF-8 sequence.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U8_PREV
+ * @stable ICU 2.4
+ */
+#define U8_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[--(i)]; \
+    if(U8_IS_TRAIL(c)) { \
+        uint8_t __b, __count=1, __shift=6; \
+\
+        /* c is a trail byte */ \
+        (c)&=0x3f; \
+        for(;;) { \
+            __b=(s)[--(i)]; \
+            if(__b>=0xc0) { \
+                U8_MASK_LEAD_BYTE(__b, __count); \
+                (c)|=(UChar32)__b<<__shift; \
+                break; \
+            } else { \
+                (c)|=(UChar32)(__b&0x3f)<<__shift; \
+                ++__count; \
+                __shift+=6; \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a multi-byte sequence, then the macro will read
+ * the whole sequence.
+ * If the offset is behind a lead byte, then that itself
+ * will be returned as the code point.
+ * If the offset is behind an illegal UTF-8 sequence, then c is set to a negative value.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<i
+ * @param c output UChar32 variable, set to <0 in case of an error
+ * @see U8_PREV_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_PREV(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[--(i)]; \
+    if(!U8_IS_SINGLE(c)) { \
+        (c)=utf8_prevCharSafeBody((const uint8_t *)s, start, &(i), c, -1); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a multi-byte sequence, then the macro will read
+ * the whole sequence.
+ * If the offset is behind a lead byte, then that itself
+ * will be returned as the code point.
+ * If the offset is behind an illegal UTF-8 sequence, then c is set to U+FFFD.
+ *
+ * This macro does not distinguish between a real U+FFFD in the text
+ * and U+FFFD returned for an ill-formed sequence.
+ * Use U8_PREV() if that distinction is important.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<i
+ * @param c output UChar32 variable, set to U+FFFD in case of an error
+ * @see U8_PREV
+ * @stable ICU 51
+ */
+#define U8_PREV_OR_FFFD(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[--(i)]; \
+    if(!U8_IS_SINGLE(c)) { \
+        (c)=utf8_prevCharSafeBody((const uint8_t *)s, start, &(i), c, -3); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_BACK_1
+ * @stable ICU 2.4
+ */
+#define U8_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    while(U8_IS_TRAIL((s)[--(i)])) {} \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<i
+ * @see U8_BACK_1_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_BACK_1(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U8_IS_TRAIL((s)[--(i)])) { \
+        (i)=utf8_back1SafeBody(s, start, (i)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U8_BACK_N
+ * @stable ICU 2.4
+ */
+#define U8_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U8_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t index of the start of the string
+ * @param i int32_t string offset, must be start<i
+ * @param n number of code points to skip
+ * @see U8_BACK_N_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_BACK_N(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0 && (i)>(start)) { \
+        U8_BACK_1(s, start, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind a partial multi-byte sequence,
+ * then the offset is incremented to behind the whole sequence.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_SET_CP_LIMIT
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    U8_BACK_1_UNSAFE(s, i); \
+    U8_FWD_1_UNSAFE(s, i); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind a partial multi-byte sequence,
+ * then the offset is incremented to behind the whole sequence.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<=i<=length
+ * @param length int32_t string length
+ * @see U8_SET_CP_LIMIT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((start)<(i) && ((i)<(length) || (length)<0)) { \
+        U8_BACK_1(s, start, i); \
+        U8_FWD_1(s, i, length); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/utf_old.h b/thirdparty/icu4c/common/unicode/utf_old.h
new file mode 100644
index 0000000000..160f5ad0a9
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utf_old.h
@@ -0,0 +1,1201 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf_old.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002sep21
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: Deprecated macros for Unicode string handling
+ *
+ * The macros in utf_old.h are all deprecated and their use discouraged.
+ * Some of the design principles behind the set of UTF macros
+ * have changed or proved impractical.
+ * Almost all of the old "UTF macros" are at least renamed.
+ * If you are looking for a new equivalent to an old macro, please see the
+ * comment at the old one.
+ *
+ * Brief summary of reasons for deprecation:
+ * - Switch on UTF_SIZE (selection of UTF-8/16/32 default string processing)
+ *   was impractical.
+ * - Switch on UTF_SAFE etc. (selection of unsafe/safe/strict default string processing)
+ *   was of little use and impractical.
+ * - Whole classes of macros became obsolete outside of the UTF_SIZE/UTF_SAFE
+ *   selection framework: UTF32_ macros (all trivial)
+ *   and UTF_ default and intermediate macros (all aliases).
+ * - The selection framework also caused many macro aliases.
+ * - Change in Unicode standard: "irregular" sequences (3.0) became illegal (3.2).
+ * - Change of language in Unicode standard:
+ *   Growing distinction between internal x-bit Unicode strings and external UTF-x
+ *   forms, with the former more lenient.
+ *   Suggests renaming of UTF16_ macros to U16_.
+ * - The prefix "UTF_" without a width number confused some users.
+ * - "Safe" append macros needed the addition of an error indicator output.
+ * - "Safe" UTF-8 macros used legitimate (if rarely used) code point values
+ *   to indicate error conditions.
+ * - The use of the "_CHAR" infix for code point operations confused some users.
+ *
+ * More details:
+ *
+ * Until ICU 2.2, utf.h theoretically allowed to choose among UTF-8/16/32
+ * for string processing, and among unsafe/safe/strict default macros for that.
+ *
+ * It proved nearly impossible to write non-trivial, high-performance code
+ * that is UTF-generic.
+ * Unsafe default macros would be dangerous for default string processing,
+ * and the main reason for the "strict" versions disappeared:
+ * Between Unicode 3.0 and 3.2 all "irregular" UTF-8 sequences became illegal.
+ * The only other conditions that "strict" checked for were non-characters,
+ * which are valid during processing. Only during text input/output should they
+ * be checked, and at that time other well-formedness checks may be
+ * necessary or useful as well.
+ * This can still be done by using U16_NEXT and U_IS_UNICODE_NONCHAR
+ * or U_IS_UNICODE_CHAR.
+ *
+ * The old UTF8_..._SAFE macros also used some normal Unicode code points
+ * to indicate malformed sequences.
+ * The new UTF8_ macros without suffix use negative values instead.
+ *
+ * The entire contents of utf32.h was moved here without replacement
+ * because all those macros were trivial and
+ * were meaningful only in the framework of choosing the UTF size.
+ *
+ * See Jitterbug 2150 and its discussion on the ICU mailing list
+ * in September 2002.
+ *
+ * <hr>
+ *
+ * <em>Obsolete part</em> of pre-ICU 2.4 utf.h file documentation:
+ *
+ * <p>The original concept for these files was for ICU to allow
+ * in principle to set which UTF (UTF-8/16/32) is used internally
+ * by defining UTF_SIZE to either 8, 16, or 32. utf.h would then define the UChar type
+ * accordingly. UTF-16 was the default.</p>
+ *
+ * <p>This concept has been abandoned.
+ * A lot of the ICU source code assumes UChar strings are in UTF-16.
+ * This is especially true for low-level code like
+ * conversion, normalization, and collation.
+ * The utf.h header enforces the default of UTF-16.
+ * The UTF-8 and UTF-32 macros remain for now for completeness and backward compatibility.</p>
+ *
+ * <p>Accordingly, utf.h defines UChar to be an unsigned 16-bit integer. If this matches wchar_t, then
+ * UChar is defined to be exactly wchar_t, otherwise uint16_t.</p>
+ *
+ * <p>UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit
+ * Unicode code point (Unicode scalar value, 0..0x10ffff).
+ * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as
+ * the definition of UChar. For details see the documentation for UChar32 itself.</p>
+ *
+ * <p>utf.h also defines a number of C macros for handling single Unicode code points and
+ * for using UTF Unicode strings. It includes utf8.h, utf16.h, and utf32.h for the actual
+ * implementations of those macros and then aliases one set of them (for UTF-16) for general use.
+ * The UTF-specific macros have the UTF size in the macro name prefixes (UTF16_...), while
+ * the general alias macros always begin with UTF_...</p>
+ *
+ * <p>Many string operations can be done with or without error checking.
+ * Where such a distinction is useful, there are two versions of the macros, "unsafe" and "safe"
+ * ones with ..._UNSAFE and ..._SAFE suffixes. The unsafe macros are fast but may cause
+ * program failures if the strings are not well-formed. The safe macros have an additional, boolean
+ * parameter "strict". If strict is false, then only illegal sequences are detected.
+ * Otherwise, irregular sequences and non-characters are detected as well (like single surrogates).
+ * Safe macros return special error code points for illegal/irregular sequences:
+ * Typically, U+ffff, or values that would result in a code unit sequence of the same length
+ * as the erroneous input sequence.<br>
+ * Note that _UNSAFE macros have fewer parameters: They do not have the strictness parameter, and
+ * they do not have start/length parameters for boundary checking.</p>
+ *
+ * <p>Here, the macros are aliased in two steps:
+ * In the first step, the UTF-specific macros with UTF16_ prefix and _UNSAFE and _SAFE suffixes are
+ * aliased according to the UTF_SIZE to macros with UTF_ prefix and the same suffixes and signatures.
+ * Then, in a second step, the default, general alias macros are set to use either the unsafe or
+ * the safe/not strict (default) or the safe/strict macro;
+ * these general macros do not have a strictness parameter.</p>
+ *
+ * <p>It is possible to change the default choice for the general alias macros to be unsafe, safe/not strict or safe/strict.
+ * The default is safe/not strict. It is not recommended to select the unsafe macros as the basis for
+ * Unicode string handling in ICU! To select this, define UTF_SAFE, UTF_STRICT, or UTF_UNSAFE.</p>
+ *
+ * <p>For general use, one should use the default, general macros with UTF_ prefix and no _SAFE/_UNSAFE suffix.
+ * Only in some cases it may be necessary to control the choice of macro directly and use a less generic alias.
+ * For example, if it can be assumed that a string is well-formed and the index will stay within the bounds,
+ * then the _UNSAFE version may be used.
+ * If a UTF-8 string is to be processed, then the macros with UTF8_ prefixes need to be used.</p>
+ *
+ * <hr>
+ *
+ * Deprecated ICU 2.4. Use the macros in utf.h, utf16.h, utf8.h instead.
+ */
+
+#ifndef __UTF_OLD_H__
+#define __UTF_OLD_H__
+
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+
+/**
+ * \def U_HIDE_OBSOLETE_UTF_OLD_H
+ *
+ * Hides the obsolete definitions in unicode/utf_old.h.
+ * Recommended to be set to 1 at compile time to make sure
+ * the long-deprecated macros are no longer used.
+ *
+ * For reasons for the deprecation see the utf_old.h file comments.
+ *
+ * @internal
+ */
+#ifndef U_HIDE_OBSOLETE_UTF_OLD_H
+#   define U_HIDE_OBSOLETE_UTF_OLD_H 0
+#endif
+
+#if !defined(U_HIDE_DEPRECATED_API) && !U_HIDE_OBSOLETE_UTF_OLD_H
+
+/* Formerly utf.h, part 1 --------------------------------------------------- */
+
+#ifdef U_USE_UTF_DEPRECATES
+/**
+ * Unicode string and array offset and index type.
+ * ICU always counts Unicode code units (UChars) for
+ * string offsets, indexes, and lengths, not Unicode code points.
+ *
+ * @obsolete ICU 2.6. Use int32_t directly instead since this API will be removed in that release.
+ */
+typedef int32_t UTextOffset;
+#endif
+
+/** Number of bits in a Unicode string code unit - ICU uses 16-bit Unicode. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF_SIZE 16
+
+/**
+ * The default choice for general Unicode string macros is to use the ..._SAFE macro implementations
+ * with strict=false.
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF_SAFE
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#undef UTF_UNSAFE
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#undef UTF_STRICT
+
+/**
+ * UTF8_ERROR_VALUE_1 and UTF8_ERROR_VALUE_2 are special error values for UTF-8,
+ * which need 1 or 2 bytes in UTF-8:
+ * \code
+ * U+0015 = NAK = Negative Acknowledge, C0 control character
+ * U+009f = highest C1 control character
+ * \endcode
+ *
+ * These are used by UTF8_..._SAFE macros so that they can return an error value
+ * that needs the same number of code units (bytes) as were seen by
+ * a macro. They should be tested with UTF_IS_ERROR() or UTF_IS_VALID().
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF8_ERROR_VALUE_1 0x15
+
+/**
+ * See documentation on UTF8_ERROR_VALUE_1 for details.
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF8_ERROR_VALUE_2 0x9f
+
+/**
+ * Error value for all UTFs. This code point value will be set by macros with error
+ * checking if an error is detected.
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF_ERROR_VALUE 0xffff
+
+/**
+ * Is a given 32-bit code an error value
+ * as returned by one of the macros for any UTF?
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF_IS_ERROR(c) \
+    (((c)&0xfffe)==0xfffe || (c)==UTF8_ERROR_VALUE_1 || (c)==UTF8_ERROR_VALUE_2)
+
+/**
+ * This is a combined macro: Is c a valid Unicode value _and_ not an error code?
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF_IS_VALID(c) \
+    (UTF_IS_UNICODE_CHAR(c) && \
+     (c)!=UTF8_ERROR_VALUE_1 && (c)!=UTF8_ERROR_VALUE_2)
+
+/**
+ * Is this code unit or code point a surrogate (U+d800..U+dfff)?
+ * @deprecated ICU 2.4. Renamed to U_IS_SURROGATE and U16_IS_SURROGATE, see utf_old.h.
+ */
+#define UTF_IS_SURROGATE(uchar) (((uchar)&0xfffff800)==0xd800)
+
+/**
+ * Is a given 32-bit code point a Unicode noncharacter?
+ *
+ * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_NONCHAR, see utf_old.h.
+ */
+#define UTF_IS_UNICODE_NONCHAR(c) \
+    ((c)>=0xfdd0 && \
+     ((uint32_t)(c)<=0xfdef || ((c)&0xfffe)==0xfffe) && \
+     (uint32_t)(c)<=0x10ffff)
+
+/**
+ * Is a given 32-bit value a Unicode code point value (0..U+10ffff)
+ * that can be assigned a character?
+ *
+ * Code points that are not characters include:
+ * - single surrogate code points (U+d800..U+dfff, 2048 code points)
+ * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points)
+ * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points)
+ * - the highest Unicode code point value is U+10ffff
+ *
+ * This means that all code points below U+d800 are character code points,
+ * and that boundary is tested first for performance.
+ *
+ * @deprecated ICU 2.4. Renamed to U_IS_UNICODE_CHAR, see utf_old.h.
+ */
+#define UTF_IS_UNICODE_CHAR(c) \
+    ((uint32_t)(c)<0xd800 || \
+        ((uint32_t)(c)>0xdfff && \
+         (uint32_t)(c)<=0x10ffff && \
+         !UTF_IS_UNICODE_NONCHAR(c)))
+
+/* Formerly utf8.h ---------------------------------------------------------- */
+
+/**
+* \var utf8_countTrailBytes
+* Internal array with numbers of trail bytes for any given byte used in
+* lead byte position.
+*
+* This is internal since it is not meant to be called directly by external clients;
+* however it is called by public macros in this file and thus must remain stable,
+* and should not be hidden when other internal functions are hidden (otherwise
+* public macros would fail to compile).
+* @internal
+*/
+#ifdef U_UTF8_IMPL
+// No forward declaration if compiling utf_impl.cpp, which defines utf8_countTrailBytes.
+#elif defined(U_STATIC_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION)
+U_CFUNC const uint8_t utf8_countTrailBytes[];
+#else
+U_CFUNC U_IMPORT const uint8_t utf8_countTrailBytes[];    /* U_IMPORT2? */ /*U_IMPORT*/
+#endif
+
+/**
+ * Count the trail bytes for a UTF-8 lead byte.
+ * @deprecated ICU 2.4. Renamed to U8_COUNT_TRAIL_BYTES, see utf_old.h.
+ */
+#define UTF8_COUNT_TRAIL_BYTES(leadByte) (utf8_countTrailBytes[(uint8_t)leadByte])
+
+/**
+ * Mask a UTF-8 lead byte, leave only the lower bits that form part of the code point value.
+ * @deprecated ICU 2.4. Renamed to U8_MASK_LEAD_BYTE, see utf_old.h.
+ */
+#define UTF8_MASK_LEAD_BYTE(leadByte, countTrailBytes) ((leadByte)&=(1<<(6-(countTrailBytes)))-1)
+
+/** Is this this code point a single code unit (byte)? @deprecated ICU 2.4. Renamed to U8_IS_SINGLE, see utf_old.h. */
+#define UTF8_IS_SINGLE(uchar) (((uchar)&0x80)==0)
+/** Is this this code unit the lead code unit (byte) of a code point? @deprecated ICU 2.4. Renamed to U8_IS_LEAD, see utf_old.h. */
+#define UTF8_IS_LEAD(uchar) ((uint8_t)((uchar)-0xc0)<0x3e)
+/** Is this this code unit a trailing code unit (byte) of a code point? @deprecated ICU 2.4. Renamed to U8_IS_TRAIL, see utf_old.h. */
+#define UTF8_IS_TRAIL(uchar) (((uchar)&0xc0)==0x80)
+
+/** Does this scalar Unicode value need multiple code units for storage? @deprecated ICU 2.4. Use U8_LENGTH or test ((uint32_t)(c)>0x7f) instead, see utf_old.h. */
+#define UTF8_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0x7f)
+
+/**
+ * Given the lead character, how many bytes are taken by this code point.
+ * ICU does not deal with code points >0x10ffff
+ * unless necessary for advancing in the byte stream.
+ *
+ * These length macros take into account that for values >0x10ffff
+ * the UTF8_APPEND_CHAR_SAFE macros would write the error code point 0xffff
+ * with 3 bytes.
+ * Code point comparisons need to be in uint32_t because UChar32
+ * may be a signed type, and negative values must be recognized.
+ *
+ * @deprecated ICU 2.4. Use U8_LENGTH instead, see utf.h.
+ */
+#if 1
+#   define UTF8_CHAR_LENGTH(c) \
+        ((uint32_t)(c)<=0x7f ? 1 : \
+            ((uint32_t)(c)<=0x7ff ? 2 : \
+                ((uint32_t)((c)-0x10000)>0xfffff ? 3 : 4) \
+            ) \
+        )
+#else
+#   define UTF8_CHAR_LENGTH(c) \
+        ((uint32_t)(c)<=0x7f ? 1 : \
+            ((uint32_t)(c)<=0x7ff ? 2 : \
+                ((uint32_t)(c)<=0xffff ? 3 : \
+                    ((uint32_t)(c)<=0x10ffff ? 4 : \
+                        ((uint32_t)(c)<=0x3ffffff ? 5 : \
+                            ((uint32_t)(c)<=0x7fffffff ? 6 : 3) \
+                        ) \
+                    ) \
+                ) \
+            ) \
+        )
+#endif
+
+/** The maximum number of bytes per code point. @deprecated ICU 2.4. Renamed to U8_MAX_LENGTH, see utf_old.h. */
+#define UTF8_MAX_CHAR_LENGTH 4
+
+/** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF8_ARRAY_SIZE(size) ((5*(size))/2)
+
+/** @deprecated ICU 2.4. Renamed to U8_GET_UNSAFE, see utf_old.h. */
+#define UTF8_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _utf8_get_char_unsafe_index=(int32_t)(i); \
+    UTF8_SET_CHAR_START_UNSAFE(s, _utf8_get_char_unsafe_index); \
+    UTF8_NEXT_CHAR_UNSAFE(s, _utf8_get_char_unsafe_index, c); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U8_GET instead, see utf_old.h. */
+#define UTF8_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _utf8_get_char_safe_index=(int32_t)(i); \
+    UTF8_SET_CHAR_START_SAFE(s, start, _utf8_get_char_safe_index); \
+    UTF8_NEXT_CHAR_SAFE(s, _utf8_get_char_safe_index, length, c, strict); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_NEXT_UNSAFE, see utf_old.h. */
+#define UTF8_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if((uint8_t)((c)-0xc0)<0x35) { \
+        uint8_t __count=UTF8_COUNT_TRAIL_BYTES(c); \
+        UTF8_MASK_LEAD_BYTE(c, __count); \
+        switch(__count) { \
+        /* each following branch falls through to the next one */ \
+        case 3: \
+            (c)=((c)<<6)|((s)[(i)++]&0x3f); \
+        case 2: \
+            (c)=((c)<<6)|((s)[(i)++]&0x3f); \
+        case 1: \
+            (c)=((c)<<6)|((s)[(i)++]&0x3f); \
+        /* no other branches to optimize switch() */ \
+            break; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_APPEND_UNSAFE, see utf_old.h. */
+#define UTF8_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0x7f) { \
+        (s)[(i)++]=(uint8_t)(c); \
+    } else { \
+        if((uint32_t)(c)<=0x7ff) { \
+            (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0); \
+        } else { \
+            if((uint32_t)(c)<=0xffff) { \
+                (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0); \
+            } else { \
+                (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0); \
+                (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80); \
+            } \
+            (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80); \
+        } \
+        (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_FWD_1_UNSAFE, see utf_old.h. */
+#define UTF8_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)+=1+UTF8_COUNT_TRAIL_BYTES((s)[i]); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_FWD_N_UNSAFE, see utf_old.h. */
+#define UTF8_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        UTF8_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_SET_CP_START_UNSAFE, see utf_old.h. */
+#define UTF8_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    while(UTF8_IS_TRAIL((s)[i])) { --(i); } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U8_NEXT instead, see utf_old.h. */
+#define UTF8_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if((c)>=0x80) { \
+        if(UTF8_IS_LEAD(c)) { \
+            (c)=utf8_nextCharSafeBody(s, &(i), (int32_t)(length), c, strict); \
+        } else { \
+            (c)=UTF8_ERROR_VALUE_1; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U8_APPEND instead, see utf_old.h. */
+#define UTF8_APPEND_CHAR_SAFE(s, i, length, c)  UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0x7f) { \
+        (s)[(i)++]=(uint8_t)(c); \
+    } else { \
+        (i)=utf8_appendCharSafeBody(s, (int32_t)(i), (int32_t)(length), c, NULL); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_FWD_1, see utf_old.h. */
+#define UTF8_FWD_1_SAFE(s, i, length) U8_FWD_1(s, i, length)
+
+/** @deprecated ICU 2.4. Renamed to U8_FWD_N, see utf_old.h. */
+#define UTF8_FWD_N_SAFE(s, i, length, n) U8_FWD_N(s, i, length, n)
+
+/** @deprecated ICU 2.4. Renamed to U8_SET_CP_START, see utf_old.h. */
+#define UTF8_SET_CHAR_START_SAFE(s, start, i) U8_SET_CP_START(s, start, i)
+
+/** @deprecated ICU 2.4. Renamed to U8_PREV_UNSAFE, see utf_old.h. */
+#define UTF8_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(UTF8_IS_TRAIL(c)) { \
+        uint8_t __b, __count=1, __shift=6; \
+\
+        /* c is a trail byte */ \
+        (c)&=0x3f; \
+        for(;;) { \
+            __b=(s)[--(i)]; \
+            if(__b>=0xc0) { \
+                UTF8_MASK_LEAD_BYTE(__b, __count); \
+                (c)|=(UChar32)__b<<__shift; \
+                break; \
+            } else { \
+                (c)|=(UChar32)(__b&0x3f)<<__shift; \
+                ++__count; \
+                __shift+=6; \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_BACK_1_UNSAFE, see utf_old.h. */
+#define UTF8_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    while(UTF8_IS_TRAIL((s)[--(i)])) {} \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_BACK_N_UNSAFE, see utf_old.h. */
+#define UTF8_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        UTF8_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
+#define UTF8_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    UTF8_BACK_1_UNSAFE(s, i); \
+    UTF8_FWD_1_UNSAFE(s, i); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U8_PREV instead, see utf_old.h. */
+#define UTF8_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if((c)>=0x80) { \
+        if((c)<=0xbf) { \
+            (c)=utf8_prevCharSafeBody(s, start, &(i), c, strict); \
+        } else { \
+            (c)=UTF8_ERROR_VALUE_1; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U8_BACK_1, see utf_old.h. */
+#define UTF8_BACK_1_SAFE(s, start, i) U8_BACK_1(s, start, i)
+
+/** @deprecated ICU 2.4. Renamed to U8_BACK_N, see utf_old.h. */
+#define UTF8_BACK_N_SAFE(s, start, i, n) U8_BACK_N(s, start, i, n)
+
+/** @deprecated ICU 2.4. Renamed to U8_SET_CP_LIMIT, see utf_old.h. */
+#define UTF8_SET_CHAR_LIMIT_SAFE(s, start, i, length) U8_SET_CP_LIMIT(s, start, i, length)
+
+/* Formerly utf16.h --------------------------------------------------------- */
+
+/** Is uchar a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h. */
+#define UTF_IS_FIRST_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xd800)
+
+/** Is uchar a second/trail surrogate? @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h. */
+#define UTF_IS_SECOND_SURROGATE(uchar) (((uchar)&0xfffffc00)==0xdc00)
+
+/** Assuming c is a surrogate, is it a first/lead surrogate? @deprecated ICU 2.4. Renamed to U_IS_SURROGATE_LEAD and U16_IS_SURROGATE_LEAD, see utf_old.h. */
+#define UTF_IS_SURROGATE_FIRST(c) (((c)&0x400)==0)
+
+/** Helper constant for UTF16_GET_PAIR_VALUE. @deprecated ICU 2.4. Renamed to U16_SURROGATE_OFFSET, see utf_old.h. */
+#define UTF_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000)
+
+/** Get the UTF-32 value from the surrogate code units. @deprecated ICU 2.4. Renamed to U16_GET_SUPPLEMENTARY, see utf_old.h. */
+#define UTF16_GET_PAIR_VALUE(first, second) \
+    (((first)<<10UL)+(second)-UTF_SURROGATE_OFFSET)
+
+/** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
+#define UTF_FIRST_SURROGATE(supplementary) (UChar)(((supplementary)>>10)+0xd7c0)
+
+/** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
+#define UTF_SECOND_SURROGATE(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00)
+
+/** @deprecated ICU 2.4. Renamed to U16_LEAD, see utf_old.h. */
+#define UTF16_LEAD(supplementary) UTF_FIRST_SURROGATE(supplementary)
+
+/** @deprecated ICU 2.4. Renamed to U16_TRAIL, see utf_old.h. */
+#define UTF16_TRAIL(supplementary) UTF_SECOND_SURROGATE(supplementary)
+
+/** @deprecated ICU 2.4. Renamed to U16_IS_SINGLE, see utf_old.h. */
+#define UTF16_IS_SINGLE(uchar) !UTF_IS_SURROGATE(uchar)
+
+/** @deprecated ICU 2.4. Renamed to U16_IS_LEAD, see utf_old.h. */
+#define UTF16_IS_LEAD(uchar) UTF_IS_FIRST_SURROGATE(uchar)
+
+/** @deprecated ICU 2.4. Renamed to U16_IS_TRAIL, see utf_old.h. */
+#define UTF16_IS_TRAIL(uchar) UTF_IS_SECOND_SURROGATE(uchar)
+
+/** Does this scalar Unicode value need multiple code units for storage? @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead, see utf_old.h. */
+#define UTF16_NEED_MULTIPLE_UCHAR(c) ((uint32_t)(c)>0xffff)
+
+/** @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h. */
+#define UTF16_CHAR_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2)
+
+/** @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h. */
+#define UTF16_MAX_CHAR_LENGTH 2
+
+/** Average number of code units compared to UTF-16. @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF16_ARRAY_SIZE(size) (size)
+
+/**
+ * Get a single code point from an offset that points to any
+ * of the code units that belong to that code point.
+ * Assume 0<=i<length.
+ *
+ * This could be used for iteration together with
+ * UTF16_CHAR_LENGTH() and UTF_IS_ERROR(),
+ * but the use of UTF16_NEXT_CHAR[_UNSAFE]() and
+ * UTF16_PREV_CHAR[_UNSAFE]() is more efficient for that.
+ * @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h.
+ */
+#define UTF16_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(UTF_IS_SURROGATE(c)) { \
+        if(UTF_IS_SURROGATE_FIRST(c)) { \
+            (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)+1]); \
+        } else { \
+            (c)=UTF16_GET_PAIR_VALUE((s)[(i)-1], (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
+#define UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(UTF_IS_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if(UTF_IS_SURROGATE_FIRST(c)) { \
+            if((i)+1<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)+1])) { \
+                (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
+                /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
+            } else if(strict) {\
+                /* unmatched first surrogate */ \
+                (c)=UTF_ERROR_VALUE; \
+            } \
+        } else { \
+            if((i)-1>=(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
+                (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
+                /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
+            } else if(strict) {\
+                /* unmatched second surrogate */ \
+                (c)=UTF_ERROR_VALUE; \
+            } \
+        } \
+    } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
+#define UTF16_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(UTF_IS_FIRST_SURROGATE(c)) { \
+        (c)=UTF16_GET_PAIR_VALUE((c), (s)[(i)++]); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
+#define UTF16_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0xffff) { \
+        (s)[(i)++]=(uint16_t)(c); \
+    } else { \
+        (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
+        (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
+#define UTF16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTF_IS_FIRST_SURROGATE((s)[(i)++])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
+#define UTF16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        UTF16_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
+#define UTF16_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTF_IS_SECOND_SURROGATE((s)[i])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
+#define UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(UTF_IS_FIRST_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if((i)<(length) && UTF_IS_SECOND_SURROGATE(__c2=(s)[(i)])) { \
+            ++(i); \
+            (c)=UTF16_GET_PAIR_VALUE((c), __c2); \
+            /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
+        } else if(strict) {\
+            /* unmatched first surrogate */ \
+            (c)=UTF_ERROR_VALUE; \
+        } \
+    } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
+        /* unmatched second surrogate or other non-character */ \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
+#define UTF16_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0xffff) { \
+        (s)[(i)++]=(uint16_t)(c); \
+    } else if((uint32_t)(c)<=0x10ffff) { \
+        if((i)+1<(length)) { \
+            (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
+            (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+        } else /* not enough space */ { \
+            (s)[(i)++]=UTF_ERROR_VALUE; \
+        } \
+    } else /* c>0x10ffff, write error value */ { \
+        (s)[(i)++]=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
+#define UTF16_FWD_1_SAFE(s, i, length) U16_FWD_1(s, i, length)
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
+#define UTF16_FWD_N_SAFE(s, i, length, n) U16_FWD_N(s, i, length, n)
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
+#define UTF16_SET_CHAR_START_SAFE(s, start, i) U16_SET_CP_START(s, start, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
+#define UTF16_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(UTF_IS_SECOND_SURROGATE(c)) { \
+        (c)=UTF16_GET_PAIR_VALUE((s)[--(i)], (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
+#define UTF16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTF_IS_SECOND_SURROGATE((s)[--(i)])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
+#define UTF16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        UTF16_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
+#define UTF16_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTF_IS_FIRST_SURROGATE((s)[(i)-1])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
+#define UTF16_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(UTF_IS_SECOND_SURROGATE(c)) { \
+        uint16_t __c2; \
+        if((i)>(start) && UTF_IS_FIRST_SURROGATE(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=UTF16_GET_PAIR_VALUE(__c2, (c)); \
+            /* strict: ((c)&0xfffe)==0xfffe is caught by UTF_IS_ERROR() and UTF_IS_UNICODE_CHAR() */ \
+        } else if(strict) {\
+            /* unmatched second surrogate */ \
+            (c)=UTF_ERROR_VALUE; \
+        } \
+    } else if((strict) && !UTF_IS_UNICODE_CHAR(c)) { \
+        /* unmatched first surrogate or other non-character */ \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
+#define UTF16_BACK_1_SAFE(s, start, i) U16_BACK_1(s, start, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
+#define UTF16_BACK_N_SAFE(s, start, i, n) U16_BACK_N(s, start, i, n)
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
+#define UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
+
+/* Formerly utf32.h --------------------------------------------------------- */
+
+/*
+* Old documentation:
+*
+*   This file defines macros to deal with UTF-32 code units and code points.
+*   Signatures and semantics are the same as for the similarly named macros
+*   in utf16.h.
+*   utf32.h is included by utf.h after unicode/umachine.h</p>
+*   and some common definitions.
+*   <p><b>Usage:</b>  ICU coding guidelines for if() statements should be followed when using these macros.
+*                  Compound statements (curly braces {}) must be used  for if-else-while...
+*                  bodies and all macro statements should be terminated with semicolon.</p>
+*/
+
+/* internal definitions ----------------------------------------------------- */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_IS_SAFE(c, strict) \
+    (!(strict) ? \
+        (uint32_t)(c)<=0x10ffff : \
+        UTF_IS_UNICODE_CHAR(c))
+
+/*
+ * For the semantics of all of these macros, see utf16.h.
+ * The UTF-32 versions are trivial because any code point is
+ * encoded using exactly one code unit.
+ */
+
+/* single-code point definitions -------------------------------------------- */
+
+/* classes of code unit values */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_IS_SINGLE(uchar) 1
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_IS_LEAD(uchar) 0
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_IS_TRAIL(uchar) 0
+
+/* number of code units per code point */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_NEED_MULTIPLE_UCHAR(c) 0
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_CHAR_LENGTH(c) 1
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_MAX_CHAR_LENGTH 1
+
+/* average number of code units compared to UTF-16 */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_ARRAY_SIZE(size) (size)
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_GET_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_GET_CHAR_SAFE(s, start, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(!UTF32_IS_SAFE(c, strict)) { \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with forward iteration --------------------------------------- */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_NEXT_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_APPEND_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (s)[(i)++]=(c); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    ++(i); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)+=(n); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_SET_CHAR_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_NEXT_CHAR_SAFE(s, i, length, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(!UTF32_IS_SAFE(c, strict)) { \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_APPEND_CHAR_SAFE(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c)<=0x10ffff) { \
+        (s)[(i)++]=(c); \
+    } else /* c>0x10ffff, write 0xfffd */ { \
+        (s)[(i)++]=0xfffd; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_FWD_1_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    ++(i); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_FWD_N_SAFE(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \
+    if(((i)+=(n))>(length)) { \
+        (i)=(length); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_SET_CHAR_START_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with backward iteration -------------------------------------- */
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_PREV_CHAR_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    --(i); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)-=(n); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_SET_CHAR_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_PREV_CHAR_SAFE(s, start, i, c, strict) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(!UTF32_IS_SAFE(c, strict)) { \
+        (c)=UTF_ERROR_VALUE; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_BACK_1_SAFE(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    --(i); \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_BACK_N_SAFE(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)-=(n); \
+    if((i)<(start)) { \
+        (i)=(start); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** @deprecated ICU 2.4. Obsolete, see utf_old.h. */
+#define UTF32_SET_CHAR_LIMIT_SAFE(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+} UPRV_BLOCK_MACRO_END
+
+/* Formerly utf.h, part 2 --------------------------------------------------- */
+
+/**
+ * Estimate the number of code units for a string based on the number of UTF-16 code units.
+ *
+ * @deprecated ICU 2.4. Obsolete, see utf_old.h.
+ */
+#define UTF_ARRAY_SIZE(size) UTF16_ARRAY_SIZE(size)
+
+/** @deprecated ICU 2.4. Renamed to U16_GET_UNSAFE, see utf_old.h. */
+#define UTF_GET_CHAR_UNSAFE(s, i, c)                 UTF16_GET_CHAR_UNSAFE(s, i, c)
+
+/** @deprecated ICU 2.4. Use U16_GET instead, see utf_old.h. */
+#define UTF_GET_CHAR_SAFE(s, start, i, length, c, strict) UTF16_GET_CHAR_SAFE(s, start, i, length, c, strict)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_NEXT_UNSAFE, see utf_old.h. */
+#define UTF_NEXT_CHAR_UNSAFE(s, i, c)                UTF16_NEXT_CHAR_UNSAFE(s, i, c)
+
+/** @deprecated ICU 2.4. Use U16_NEXT instead, see utf_old.h. */
+#define UTF_NEXT_CHAR_SAFE(s, i, length, c, strict)  UTF16_NEXT_CHAR_SAFE(s, i, length, c, strict)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_APPEND_UNSAFE, see utf_old.h. */
+#define UTF_APPEND_CHAR_UNSAFE(s, i, c)              UTF16_APPEND_CHAR_UNSAFE(s, i, c)
+
+/** @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h. */
+#define UTF_APPEND_CHAR_SAFE(s, i, length, c)        UTF16_APPEND_CHAR_SAFE(s, i, length, c)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_1_UNSAFE, see utf_old.h. */
+#define UTF_FWD_1_UNSAFE(s, i)                       UTF16_FWD_1_UNSAFE(s, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h. */
+#define UTF_FWD_1_SAFE(s, i, length)                 UTF16_FWD_1_SAFE(s, i, length)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_N_UNSAFE, see utf_old.h. */
+#define UTF_FWD_N_UNSAFE(s, i, n)                    UTF16_FWD_N_UNSAFE(s, i, n)
+
+/** @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h. */
+#define UTF_FWD_N_SAFE(s, i, length, n)              UTF16_FWD_N_SAFE(s, i, length, n)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_START_UNSAFE, see utf_old.h. */
+#define UTF_SET_CHAR_START_UNSAFE(s, i)              UTF16_SET_CHAR_START_UNSAFE(s, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h. */
+#define UTF_SET_CHAR_START_SAFE(s, start, i)         UTF16_SET_CHAR_START_SAFE(s, start, i)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_PREV_UNSAFE, see utf_old.h. */
+#define UTF_PREV_CHAR_UNSAFE(s, i, c)                UTF16_PREV_CHAR_UNSAFE(s, i, c)
+
+/** @deprecated ICU 2.4. Use U16_PREV instead, see utf_old.h. */
+#define UTF_PREV_CHAR_SAFE(s, start, i, c, strict)   UTF16_PREV_CHAR_SAFE(s, start, i, c, strict)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_1_UNSAFE, see utf_old.h. */
+#define UTF_BACK_1_UNSAFE(s, i)                      UTF16_BACK_1_UNSAFE(s, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h. */
+#define UTF_BACK_1_SAFE(s, start, i)                 UTF16_BACK_1_SAFE(s, start, i)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_N_UNSAFE, see utf_old.h. */
+#define UTF_BACK_N_UNSAFE(s, i, n)                   UTF16_BACK_N_UNSAFE(s, i, n)
+
+/** @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h. */
+#define UTF_BACK_N_SAFE(s, start, i, n)              UTF16_BACK_N_SAFE(s, start, i, n)
+
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT_UNSAFE, see utf_old.h. */
+#define UTF_SET_CHAR_LIMIT_UNSAFE(s, i)              UTF16_SET_CHAR_LIMIT_UNSAFE(s, i)
+
+/** @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h. */
+#define UTF_SET_CHAR_LIMIT_SAFE(s, start, i, length) UTF16_SET_CHAR_LIMIT_SAFE(s, start, i, length)
+
+/* Define default macros (UTF-16 "safe") ------------------------------------ */
+
+/**
+ * Does this code unit alone encode a code point (BMP, not a surrogate)?
+ * Same as UTF16_IS_SINGLE.
+ * @deprecated ICU 2.4. Renamed to U_IS_SINGLE and U16_IS_SINGLE, see utf_old.h.
+ */
+#define UTF_IS_SINGLE(uchar) U16_IS_SINGLE(uchar)
+
+/**
+ * Is this code unit the first one of several (a lead surrogate)?
+ * Same as UTF16_IS_LEAD.
+ * @deprecated ICU 2.4. Renamed to U_IS_LEAD and U16_IS_LEAD, see utf_old.h.
+ */
+#define UTF_IS_LEAD(uchar) U16_IS_LEAD(uchar)
+
+/**
+ * Is this code unit one of several but not the first one (a trail surrogate)?
+ * Same as UTF16_IS_TRAIL.
+ * @deprecated ICU 2.4. Renamed to U_IS_TRAIL and U16_IS_TRAIL, see utf_old.h.
+ */
+#define UTF_IS_TRAIL(uchar) U16_IS_TRAIL(uchar)
+
+/**
+ * Does this code point require multiple code units (is it a supplementary code point)?
+ * Same as UTF16_NEED_MULTIPLE_UCHAR.
+ * @deprecated ICU 2.4. Use U16_LENGTH or test ((uint32_t)(c)>0xffff) instead.
+ */
+#define UTF_NEED_MULTIPLE_UCHAR(c) UTF16_NEED_MULTIPLE_UCHAR(c)
+
+/**
+ * How many code units are used to encode this code point (1 or 2)?
+ * Same as UTF16_CHAR_LENGTH.
+ * @deprecated ICU 2.4. Renamed to U16_LENGTH, see utf_old.h.
+ */
+#define UTF_CHAR_LENGTH(c) U16_LENGTH(c)
+
+/**
+ * How many code units are used at most for any Unicode code point (2)?
+ * Same as UTF16_MAX_CHAR_LENGTH.
+ * @deprecated ICU 2.4. Renamed to U16_MAX_LENGTH, see utf_old.h.
+ */
+#define UTF_MAX_CHAR_LENGTH U16_MAX_LENGTH
+
+/**
+ * Set c to the code point that contains the code unit i.
+ * i could point to the lead or the trail surrogate for the code point.
+ * i is not modified.
+ * Same as UTF16_GET_CHAR.
+ * \pre 0<=i<length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_GET, see utf_old.h.
+ */
+#define UTF_GET_CHAR(s, start, i, length, c) U16_GET(s, start, i, length, c)
+
+/**
+ * Set c to the code point that starts at code unit i
+ * and advance i to beyond the code units of this code point (post-increment).
+ * i must point to the first code unit of a code point.
+ * Otherwise c is set to the trail unit (surrogate) itself.
+ * Same as UTF16_NEXT_CHAR.
+ * \pre 0<=i<length
+ * \post 0<i<=length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_NEXT, see utf_old.h.
+ */
+#define UTF_NEXT_CHAR(s, i, length, c) U16_NEXT(s, i, length, c)
+
+/**
+ * Append the code units of code point c to the string at index i
+ * and advance i to beyond the new code units (post-increment).
+ * The code units beginning at index i will be overwritten.
+ * Same as UTF16_APPEND_CHAR.
+ * \pre 0<=c<=0x10ffff
+ * \pre 0<=i<length
+ * \post 0<i<=length
+ *
+ * @deprecated ICU 2.4. Use U16_APPEND instead, see utf_old.h.
+ */
+#define UTF_APPEND_CHAR(s, i, length, c) UTF16_APPEND_CHAR_SAFE(s, i, length, c)
+
+/**
+ * Advance i to beyond the code units of the code point that begins at i.
+ * I.e., advance i by one code point.
+ * Same as UTF16_FWD_1.
+ * \pre 0<=i<length
+ * \post 0<i<=length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_FWD_1, see utf_old.h.
+ */
+#define UTF_FWD_1(s, i, length) U16_FWD_1(s, i, length)
+
+/**
+ * Advance i to beyond the code units of the n code points where the first one begins at i.
+ * I.e., advance i by n code points.
+ * Same as UT16_FWD_N.
+ * \pre 0<=i<length
+ * \post 0<i<=length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_FWD_N, see utf_old.h.
+ */
+#define UTF_FWD_N(s, i, length, n) U16_FWD_N(s, i, length, n)
+
+/**
+ * Take the random-access index i and adjust it so that it points to the beginning
+ * of a code point.
+ * The input index points to any code unit of a code point and is moved to point to
+ * the first code unit of the same code point. i is never incremented.
+ * In other words, if i points to a trail surrogate that is preceded by a matching
+ * lead surrogate, then i is decremented. Otherwise it is not modified.
+ * This can be used to start an iteration with UTF_NEXT_CHAR() from a random index.
+ * Same as UTF16_SET_CHAR_START.
+ * \pre start<=i<length
+ * \post start<=i<length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_SET_CP_START, see utf_old.h.
+ */
+#define UTF_SET_CHAR_START(s, start, i) U16_SET_CP_START(s, start, i)
+
+/**
+ * Set c to the code point that has code units before i
+ * and move i backward (towards the beginning of the string)
+ * to the first code unit of this code point (pre-increment).
+ * i must point to the first code unit after the last unit of a code point (i==length is allowed).
+ * Same as UTF16_PREV_CHAR.
+ * \pre start<i<=length
+ * \post start<=i<length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_PREV, see utf_old.h.
+ */
+#define UTF_PREV_CHAR(s, start, i, c) U16_PREV(s, start, i, c)
+
+/**
+ * Move i backward (towards the beginning of the string)
+ * to the first code unit of the code point that has code units before i.
+ * I.e., move i backward by one code point.
+ * i must point to the first code unit after the last unit of a code point (i==length is allowed).
+ * Same as UTF16_BACK_1.
+ * \pre start<i<=length
+ * \post start<=i<length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_BACK_1, see utf_old.h.
+ */
+#define UTF_BACK_1(s, start, i) U16_BACK_1(s, start, i)
+
+/**
+ * Move i backward (towards the beginning of the string)
+ * to the first code unit of the n code points that have code units before i.
+ * I.e., move i backward by n code points.
+ * i must point to the first code unit after the last unit of a code point (i==length is allowed).
+ * Same as UTF16_BACK_N.
+ * \pre start<i<=length
+ * \post start<=i<length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_BACK_N, see utf_old.h.
+ */
+#define UTF_BACK_N(s, start, i, n) U16_BACK_N(s, start, i, n)
+
+/**
+ * Take the random-access index i and adjust it so that it points beyond
+ * a code point. The input index points beyond any code unit
+ * of a code point and is moved to point beyond the last code unit of the same
+ * code point. i is never decremented.
+ * In other words, if i points to a trail surrogate that is preceded by a matching
+ * lead surrogate, then i is incremented. Otherwise it is not modified.
+ * This can be used to start an iteration with UTF_PREV_CHAR() from a random index.
+ * Same as UTF16_SET_CHAR_LIMIT.
+ * \pre start<i<=length
+ * \post start<i<=length
+ *
+ * @deprecated ICU 2.4. Renamed to U16_SET_CP_LIMIT, see utf_old.h.
+ */
+#define UTF_SET_CHAR_LIMIT(s, start, i, length) U16_SET_CP_LIMIT(s, start, i, length)
+
+#endif  // !U_HIDE_DEPRECATED_API && !U_HIDE_OBSOLETE_UTF_OLD_H
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/utrace.h b/thirdparty/icu4c/common/unicode/utrace.h
new file mode 100644
index 0000000000..28c313c582
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utrace.h
@@ -0,0 +1,509 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2013, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utrace.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003aug06
+*   created by: Markus W. Scherer
+*
+*   Definitions for ICU tracing/logging.
+*
+*/
+
+#ifndef __UTRACE_H__
+#define __UTRACE_H__
+
+#include <stdarg.h>
+#include "unicode/utypes.h"
+
+/**
+ * \file
+ * \brief C API:  Definitions for ICU tracing/logging. 
+ *
+ * This provides API for debugging the internals of ICU without the use of
+ * a traditional debugger.
+ *
+ * By default, tracing is disabled in ICU. If you need to debug ICU with 
+ * tracing, please compile ICU with the --enable-tracing configure option.
+ */
+ 
+U_CDECL_BEGIN
+
+/**
+ * Trace severity levels.  Higher levels increase the verbosity of the trace output.
+ * @see utrace_setLevel
+ * @stable ICU 2.8
+ */
+typedef enum UTraceLevel {
+    /** Disable all tracing  @stable ICU 2.8*/
+    UTRACE_OFF=-1,
+    /** Trace error conditions only  @stable ICU 2.8*/
+    UTRACE_ERROR=0,
+    /** Trace errors and warnings  @stable ICU 2.8*/
+    UTRACE_WARNING=3,
+    /** Trace opens and closes of ICU services  @stable ICU 2.8*/
+    UTRACE_OPEN_CLOSE=5,
+    /** Trace an intermediate number of ICU operations  @stable ICU 2.8*/
+    UTRACE_INFO=7,
+    /** Trace the maximum number of ICU operations  @stable ICU 2.8*/
+    UTRACE_VERBOSE=9
+} UTraceLevel;
+
+/**
+ *  These are the ICU functions that will be traced when tracing is enabled.
+ *  @stable ICU 2.8
+ */
+typedef enum UTraceFunctionNumber {
+    UTRACE_FUNCTION_START=0,
+    UTRACE_U_INIT=UTRACE_FUNCTION_START,
+    UTRACE_U_CLEANUP,
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal collation trace location.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UTRACE_FUNCTION_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    UTRACE_CONVERSION_START=0x1000,
+    UTRACE_UCNV_OPEN=UTRACE_CONVERSION_START,
+    UTRACE_UCNV_OPEN_PACKAGE,
+    UTRACE_UCNV_OPEN_ALGORITHMIC,
+    UTRACE_UCNV_CLONE,
+    UTRACE_UCNV_CLOSE,
+    UTRACE_UCNV_FLUSH_CACHE,
+    UTRACE_UCNV_LOAD,
+    UTRACE_UCNV_UNLOAD,
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal collation trace location.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UTRACE_CONVERSION_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    UTRACE_COLLATION_START=0x2000,
+    UTRACE_UCOL_OPEN=UTRACE_COLLATION_START,
+    UTRACE_UCOL_CLOSE,
+    UTRACE_UCOL_STRCOLL,
+    UTRACE_UCOL_GET_SORTKEY,
+    UTRACE_UCOL_GETLOCALE,
+    UTRACE_UCOL_NEXTSORTKEYPART,
+    UTRACE_UCOL_STRCOLLITER,
+    UTRACE_UCOL_OPEN_FROM_SHORT_STRING,
+    UTRACE_UCOL_STRCOLLUTF8, /**< @stable ICU 50 */
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal collation trace location.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    UTRACE_COLLATION_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /**
+     * The lowest resource/data location.
+     * @stable ICU 65
+     */
+    UTRACE_UDATA_START=0x3000,
+
+    /**
+     * Indicates that a value was read from a resource bundle. Provides three
+     * C-style strings to UTraceData: type, file name, and resource path. The
+     * possible types are:
+     *
+     * - "string" (a string value was accessed)
+     * - "binary" (a binary value was accessed)
+     * - "intvector" (a integer vector value was accessed)
+     * - "int" (a signed integer value was accessed)
+     * - "uint" (a unsigned integer value was accessed)
+     * - "get" (a path was loaded, but the value was not accessed)
+     * - "getalias" (a path was loaded, and an alias was resolved)
+     *
+     * @stable ICU 65
+     */
+    UTRACE_UDATA_RESOURCE=UTRACE_UDATA_START,
+
+    /**
+     * Indicates that a resource bundle was opened.
+     *
+     * Provides one C-style string to UTraceData: file name.
+     * @stable ICU 65
+     */
+    UTRACE_UDATA_BUNDLE,
+
+    /**
+     * Indicates that a data file was opened, but not *.res files.
+     *
+     * Provides one C-style string to UTraceData: file name.
+     *
+     * @stable ICU 65
+     */
+    UTRACE_UDATA_DATA_FILE,
+
+    /**
+     * Indicates that a *.res file was opened.
+     *
+     * This differs from UTRACE_UDATA_BUNDLE because a res file is typically
+     * opened only once per application runtime, but the bundle corresponding
+     * to that res file may be opened many times.
+     *
+     * Provides one C-style string to UTraceData: file name.
+     *
+     * @stable ICU 65
+     */
+    UTRACE_UDATA_RES_FILE,
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * One more than the highest normal resource/data trace location.
+     * @internal The numeric value may change over time, see ICU ticket #12420.
+     */
+    UTRACE_RES_DATA_LIMIT,
+#endif  // U_HIDE_INTERNAL_API
+
+#ifndef U_HIDE_DRAFT_API
+    /**
+     * The lowest break iterator location.
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_START=0x4000,
+
+    /**
+     * Indicates that a character instance of break iterator was created.
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_CHARACTER = UTRACE_UBRK_START,
+
+    /**
+     * Indicates that a word instance of break iterator was created.
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_WORD,
+
+    /**
+     * Indicates that a line instance of break iterator was created.
+     *
+     * Provides one C-style string to UTraceData: the lb value ("",
+     * "loose", "strict", or "normal").
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_LINE,
+
+    /**
+     * Indicates that a sentence instance of break iterator was created.
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_SENTENCE,
+
+    /**
+     * Indicates that a title instance of break iterator was created.
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_TITLE,
+
+    /**
+     * Indicates that an internal dictionary break engine was created.
+     *
+     * Provides one C-style string to UTraceData: the script code of what
+     * the break engine cover ("Hani", "Khmr", "Laoo", "Mymr", or "Thai").
+     *
+     * @draft ICU 67
+     */
+    UTRACE_UBRK_CREATE_BREAK_ENGINE,
+
+#endif  // U_HIDE_DRAFT_API
+
+#ifndef U_HIDE_INTERNAL_API
+    /**
+     * One more than the highest normal break iterator trace location.
+     * @internal The numeric value may change over time, see ICU ticket #12420.
+     */
+    UTRACE_UBRK_LIMIT,
+#endif  // U_HIDE_INTERNAL_API
+
+} UTraceFunctionNumber;
+
+/**
+ * Setter for the trace level.
+ * @param traceLevel A UTraceLevel value.
+ * @stable ICU 2.8
+ */
+U_CAPI void U_EXPORT2
+utrace_setLevel(int32_t traceLevel);
+
+/**
+ * Getter for the trace level.
+ * @return The UTraceLevel value being used by ICU.
+ * @stable ICU 2.8
+ */
+U_CAPI int32_t U_EXPORT2
+utrace_getLevel(void);
+
+/* Trace function pointers types  ----------------------------- */
+
+/**
+  *  Type signature for the trace function to be called when entering a function.
+  *  @param context value supplied at the time the trace functions are set.
+  *  @param fnNumber Enum value indicating the ICU function being entered.
+  *  @stable ICU 2.8
+  */
+typedef void U_CALLCONV
+UTraceEntry(const void *context, int32_t fnNumber);
+
+/**
+  *  Type signature for the trace function to be called when exiting from a function.
+  *  @param context value supplied at the time the trace functions are set.
+  *  @param fnNumber Enum value indicating the ICU function being exited.
+  *  @param fmt     A formatting string that describes the number and types
+  *                 of arguments included with the variable args.  The fmt
+  *                 string has the same form as the utrace_vformat format
+  *                 string.
+  *  @param args    A variable arguments list.  Contents are described by
+  *                 the fmt parameter.
+  *  @see   utrace_vformat
+  *  @stable ICU 2.8
+  */
+typedef void U_CALLCONV
+UTraceExit(const void *context, int32_t fnNumber, 
+           const char *fmt, va_list args);
+
+/**
+  *  Type signature for the trace function to be called from within an ICU function
+  *  to display data or messages.
+  *  @param context  value supplied at the time the trace functions are set.
+  *  @param fnNumber Enum value indicating the ICU function being exited.
+  *  @param level    The current tracing level
+  *  @param fmt      A format string describing the tracing data that is supplied
+  *                  as variable args
+  *  @param args     The data being traced, passed as variable args.
+  *  @stable ICU 2.8
+  */
+typedef void U_CALLCONV
+UTraceData(const void *context, int32_t fnNumber, int32_t level,
+           const char *fmt, va_list args);
+
+/**
+  *  Set ICU Tracing functions.  Installs application-provided tracing
+  *  functions into ICU.  After doing this, subsequent ICU operations
+  *  will call back to the installed functions, providing a trace
+  *  of the use of ICU.  Passing a NULL pointer for a tracing function
+  *  is allowed, and inhibits tracing action at points where that function
+  *  would be called.
+  *  <p>
+  *  Tracing and Threads:  Tracing functions are global to a process, and
+  *  will be called in response to ICU operations performed by any
+  *  thread.  If tracing of an individual thread is desired, the
+  *  tracing functions must themselves filter by checking that the
+  *  current thread is the desired thread.
+  *
+  *  @param context an uninterpreted pointer.  Whatever is passed in
+  *                 here will in turn be passed to each of the tracing
+  *                 functions UTraceEntry, UTraceExit and UTraceData.
+  *                 ICU does not use or alter this pointer.
+  *  @param e       Callback function to be called on entry to a 
+  *                 a traced ICU function.
+  *  @param x       Callback function to be called on exit from a
+  *                 traced ICU function.
+  *  @param d       Callback function to be called from within a 
+  *                 traced ICU function, for the purpose of providing
+  *                 data to the trace.
+  *
+  *  @stable ICU 2.8
+  */
+U_CAPI void U_EXPORT2
+utrace_setFunctions(const void *context,
+                    UTraceEntry *e, UTraceExit *x, UTraceData *d);
+
+/**
+  * Get the currently installed ICU tracing functions.   Note that a null function
+  *   pointer will be returned if no trace function has been set.
+  *
+  * @param context  The currently installed tracing context.
+  * @param e        The currently installed UTraceEntry function.
+  * @param x        The currently installed UTraceExit function.
+  * @param d        The currently installed UTraceData function.
+  * @stable ICU 2.8
+  */
+U_CAPI void U_EXPORT2
+utrace_getFunctions(const void **context,
+                    UTraceEntry **e, UTraceExit **x, UTraceData **d);
+
+
+
+/*
+ *
+ * ICU trace format string syntax
+ *
+ * Format Strings are passed to UTraceData functions, and define the
+ * number and types of the trace data being passed on each call.
+ *
+ * The UTraceData function, which is supplied by the application,
+ * not by ICU, can either forward the trace data (passed via
+ * varargs) and the format string back to ICU for formatting into
+ * a displayable string, or it can interpret the format itself,
+ * and do as it wishes with the trace data.
+ *
+ *
+ * Goals for the format string
+ * - basic data output
+ * - easy to use for trace programmer
+ * - sufficient provision for data types for trace output readability
+ * - well-defined types and binary portable APIs
+ *
+ * Non-goals
+ * - printf compatibility
+ * - fancy formatting
+ * - argument reordering and other internationalization features
+ *
+ * ICU trace format strings contain plain text with argument inserts,
+ * much like standard printf format strings.
+ * Each insert begins with a '%', then optionally contains a 'v',
+ * then exactly one type character.
+ * Two '%' in a row represent a '%' instead of an insert.
+ * The trace format strings need not have \n at the end.
+ *
+ *
+ * Types
+ * -----
+ *
+ * Type characters:
+ * - c A char character in the default codepage.
+ * - s A NUL-terminated char * string in the default codepage.
+ * - S A UChar * string.  Requires two params, (ptr, length).  Length=-1 for nul term.
+ * - b A byte (8-bit integer).
+ * - h A 16-bit integer.  Also a 16 bit Unicode code unit.
+ * - d A 32-bit integer.  Also a 20 bit Unicode code point value. 
+ * - l A 64-bit integer.
+ * - p A data pointer.
+ *
+ * Vectors
+ * -------
+ *
+ * If the 'v' is not specified, then one item of the specified type
+ * is passed in.
+ * If the 'v' (for "vector") is specified, then a vector of items of the
+ * specified type is passed in, via a pointer to the first item
+ * and an int32_t value for the length of the vector.
+ * Length==-1 means zero or NUL termination.  Works for vectors of all types.
+ *
+ * Note:  %vS is a vector of (UChar *) strings.  The strings must
+ *        be nul terminated as there is no way to provide a
+ *        separate length parameter for each string.  The length
+ *        parameter (required for all vectors) is the number of
+ *        strings, not the length of the strings.
+ *
+ * Examples
+ * --------
+ *
+ * These examples show the parameters that will be passed to an application's
+ *   UTraceData() function for various formats.
+ *
+ * - the precise formatting is up to the application!
+ * - the examples use type casts for arguments only to _show_ the types of
+ *   arguments without needing variable declarations in the examples;
+ *   the type casts will not be necessary in actual code
+ *
+ * UTraceDataFunc(context, fnNumber, level,
+ *              "There is a character %c in the string %s.",   // Format String 
+ *              (char)c, (const char *)s);                     // varargs parameters
+ * ->   There is a character 0x42 'B' in the string "Bravo".
+ *
+ * UTraceDataFunc(context, fnNumber, level,
+ *              "Vector of bytes %vb vector of chars %vc",
+ *              (const uint8_t *)bytes, (int32_t)bytesLength,
+ *              (const char *)chars, (int32_t)charsLength);
+ * ->  Vector of bytes
+ *      42 63 64 3f [4]
+ *     vector of chars
+ *      "Bcd?"[4]
+ *
+ * UTraceDataFunc(context, fnNumber, level,
+ *              "An int32_t %d and a whole bunch of them %vd",
+ *              (int32_t)-5, (const int32_t *)ints, (int32_t)intsLength);
+ * ->   An int32_t 0xfffffffb and a whole bunch of them
+ *      fffffffb 00000005 0000010a [3]
+ *
+ */
+
+
+
+/**
+  *  Trace output Formatter.  An application's UTraceData tracing functions may call
+  *                 back to this function to format the trace output in a
+  *                 human readable form.  Note that a UTraceData function may choose
+  *                 to not format the data;  it could, for example, save it in
+  *                 in the raw form it was received (more compact), leaving
+  *                 formatting for a later trace analysis tool.
+  *  @param outBuf  pointer to a buffer to receive the formatted output.  Output
+  *                 will be nul terminated if there is space in the buffer -
+  *                 if the length of the requested output < the output buffer size.
+  *  @param capacity  Length of the output buffer.
+  *  @param indent  Number of spaces to indent the output.  Intended to allow
+  *                 data displayed from nested functions to be indented for readability.
+  *  @param fmt     Format specification for the data to output
+  *  @param args    Data to be formatted.
+  *  @return        Length of formatted output, including the terminating NUL.
+  *                 If buffer capacity is insufficient, the required capacity is returned. 
+  *  @stable ICU 2.8
+  */
+U_CAPI int32_t U_EXPORT2
+utrace_vformat(char *outBuf, int32_t capacity,
+              int32_t indent, const char *fmt,  va_list args);
+
+/**
+  *  Trace output Formatter.  An application's UTraceData tracing functions may call
+  *                 this function to format any additional trace data, beyond that
+  *                 provided by default, in human readable form with the same
+  *                 formatting conventions used by utrace_vformat().
+  *  @param outBuf  pointer to a buffer to receive the formatted output.  Output
+  *                 will be nul terminated if there is space in the buffer -
+  *                 if the length of the requested output < the output buffer size.
+  *  @param capacity  Length of the output buffer.
+  *  @param indent  Number of spaces to indent the output.  Intended to allow
+  *                 data displayed from nested functions to be indented for readability.
+  *  @param fmt     Format specification for the data to output
+  *  @param ...     Data to be formatted.
+  *  @return        Length of formatted output, including the terminating NUL.
+  *                 If buffer capacity is insufficient, the required capacity is returned. 
+  *  @stable ICU 2.8
+  */
+U_CAPI int32_t U_EXPORT2
+utrace_format(char *outBuf, int32_t capacity,
+              int32_t indent, const char *fmt,  ...);
+
+
+
+/* Trace function numbers --------------------------------------------------- */
+
+/**
+ * Get the name of a function from its trace function number.
+ *
+ * @param fnNumber The trace number for an ICU function.
+ * @return The name string for the function.
+ *
+ * @see UTraceFunctionNumber
+ * @stable ICU 2.8
+ */
+U_CAPI const char * U_EXPORT2
+utrace_functionName(int32_t fnNumber);
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/utypes.h b/thirdparty/icu4c/common/unicode/utypes.h
new file mode 100644
index 0000000000..7c4ea7ac28
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/utypes.h
@@ -0,0 +1,732 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1996-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+*  FILE NAME : UTYPES.H (formerly ptypes.h)
+*
+*   Date        Name        Description
+*   12/11/96    helena      Creation.
+*   02/27/97    aliu        Added typedefs for UClassID, int8, int16, int32,
+*                           uint8, uint16, and uint32.
+*   04/01/97    aliu        Added XP_CPLUSPLUS and modified to work under C as
+*                            well as C++.
+*                           Modified to use memcpy() for uprv_arrayCopy() fns.
+*   04/14/97    aliu        Added TPlatformUtilities.
+*   05/07/97    aliu        Added import/export specifiers (replacing the old
+*                           broken EXT_CLASS).  Added version number for our
+*                           code.  Cleaned up header.
+*    6/20/97    helena      Java class name change.
+*   08/11/98    stephen     UErrorCode changed from typedef to enum
+*   08/12/98    erm         Changed T_ANALYTIC_PACKAGE_VERSION to 3
+*   08/14/98    stephen     Added uprv_arrayCopy() for int8_t, int16_t, int32_t
+*   12/09/98    jfitz       Added BUFFER_OVERFLOW_ERROR (bug 1100066)
+*   04/20/99    stephen     Cleaned up & reworked for autoconf.
+*                           Renamed to utypes.h.
+*   05/05/99    stephen     Changed to use <inttypes.h>
+*   12/07/99    helena      Moved copyright notice string from ucnv_bld.h here.
+*******************************************************************************
+*/
+
+#ifndef UTYPES_H
+#define UTYPES_H
+
+
+#include "unicode/umachine.h"
+#include "unicode/uversion.h"
+#include "unicode/uconfig.h"
+#include <float.h>
+
+#if !U_NO_DEFAULT_INCLUDE_UTF_HEADERS
+#   include "unicode/utf.h"
+#endif
+
+/*!
+ * \file
+ * \brief Basic definitions for ICU, for both C and C++ APIs
+ *
+ * This file defines basic types, constants, and enumerations directly or
+ * indirectly by including other header files, especially utf.h for the
+ * basic character and string definitions and umachine.h for consistent
+ * integer and other types.
+ */
+
+
+/**
+ * \def U_SHOW_CPLUSPLUS_API
+ * @internal
+ */
+#ifdef __cplusplus
+#   ifndef U_SHOW_CPLUSPLUS_API
+#       define U_SHOW_CPLUSPLUS_API 1
+#   endif
+#else
+#   undef U_SHOW_CPLUSPLUS_API
+#   define U_SHOW_CPLUSPLUS_API 0
+#endif
+
+/** @{ API visibility control */
+
+/**
+ * \def U_HIDE_DRAFT_API
+ * Define this to 1 to request that draft API be "hidden"
+ * @internal
+ */
+/**
+ * \def U_HIDE_INTERNAL_API
+ * Define this to 1 to request that internal API be "hidden"
+ * @internal
+ */
+#if !U_DEFAULT_SHOW_DRAFT && !defined(U_SHOW_DRAFT_API)
+#define U_HIDE_DRAFT_API 1
+#endif
+#if !U_DEFAULT_SHOW_DRAFT && !defined(U_SHOW_INTERNAL_API)
+#define U_HIDE_INTERNAL_API 1
+#endif
+
+/** @} */
+
+/*===========================================================================*/
+/* ICUDATA naming scheme                                                     */
+/*===========================================================================*/
+
+/**
+ * \def U_ICUDATA_TYPE_LETTER
+ *
+ * This is a platform-dependent string containing one letter:
+ * - b for big-endian, ASCII-family platforms
+ * - l for little-endian, ASCII-family platforms
+ * - e for big-endian, EBCDIC-family platforms
+ * This letter is part of the common data file name.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_ICUDATA_TYPE_LITLETTER
+ * The non-string form of U_ICUDATA_TYPE_LETTER
+ * @stable ICU 2.0
+ */
+#if U_CHARSET_FAMILY
+#   if U_IS_BIG_ENDIAN
+   /* EBCDIC - should always be BE */
+#     define U_ICUDATA_TYPE_LETTER "e"
+#     define U_ICUDATA_TYPE_LITLETTER e
+#   else
+#     error "Don't know what to do with little endian EBCDIC!"
+#     define U_ICUDATA_TYPE_LETTER "x"
+#     define U_ICUDATA_TYPE_LITLETTER x
+#   endif
+#else
+#   if U_IS_BIG_ENDIAN
+      /* Big-endian ASCII */
+#     define U_ICUDATA_TYPE_LETTER "b"
+#     define U_ICUDATA_TYPE_LITLETTER b
+#   else
+      /* Little-endian ASCII */
+#     define U_ICUDATA_TYPE_LETTER "l"
+#     define U_ICUDATA_TYPE_LITLETTER l
+#   endif
+#endif
+
+/**
+ * A single string literal containing the icudata stub name. i.e. 'icudt18e' for
+ * ICU 1.8.x on EBCDIC, etc..
+ * @stable ICU 2.0
+ */
+#define U_ICUDATA_NAME    "icudt" U_ICU_VERSION_SHORT U_ICUDATA_TYPE_LETTER
+#ifndef U_HIDE_INTERNAL_API
+#define U_USRDATA_NAME    "usrdt" U_ICU_VERSION_SHORT U_ICUDATA_TYPE_LETTER  /**< @internal */
+#define U_USE_USRDATA     0  /**< @internal */
+#endif  /* U_HIDE_INTERNAL_API */
+
+/**
+ *  U_ICU_ENTRY_POINT is the name of the DLL entry point to the ICU data library.
+ *    Defined as a literal, not a string.
+ *    Tricky Preprocessor use - ## operator replaces macro parameters with the literal string
+ *                              from the corresponding macro invocation, _before_ other macro substitutions.
+ *                              Need a nested \#defines to get the actual version numbers rather than
+ *                              the literal text U_ICU_VERSION_MAJOR_NUM into the name.
+ *                              The net result will be something of the form
+ *                                  \#define U_ICU_ENTRY_POINT icudt19_dat
+ * @stable ICU 2.4
+ */
+#define U_ICUDATA_ENTRY_POINT  U_DEF2_ICUDATA_ENTRY_POINT(U_ICU_VERSION_MAJOR_NUM,U_LIB_SUFFIX_C_NAME)
+
+#ifndef U_HIDE_INTERNAL_API
+/**
+ * Do not use. Note that it's OK for the 2nd argument to be undefined (literal).
+ * @internal
+ */
+#define U_DEF2_ICUDATA_ENTRY_POINT(major,suff) U_DEF_ICUDATA_ENTRY_POINT(major,suff)
+
+/**
+ * Do not use.
+ * @internal
+ */
+#ifndef U_DEF_ICUDATA_ENTRY_POINT
+/* affected by symbol renaming. See platform.h */
+#ifndef U_LIB_SUFFIX_C_NAME
+#define U_DEF_ICUDATA_ENTRY_POINT(major, suff) icudt##major##_dat
+#else
+#define U_DEF_ICUDATA_ENTRY_POINT(major, suff) icudt##suff ## major##_dat
+#endif
+#endif
+#endif  /* U_HIDE_INTERNAL_API */
+
+/**
+ * \def NULL
+ * Define NULL if necessary, to nullptr for C++ and to ((void *)0) for C.
+ * @stable ICU 2.0
+ */
+#ifndef NULL
+#ifdef __cplusplus
+#define NULL    nullptr
+#else
+#define NULL    ((void *)0)
+#endif
+#endif
+
+/*===========================================================================*/
+/* Calendar/TimeZone data types                                              */
+/*===========================================================================*/
+
+/**
+ * Date and Time data type.
+ * This is a primitive data type that holds the date and time
+ * as the number of milliseconds since 1970-jan-01, 00:00 UTC.
+ * UTC leap seconds are ignored.
+ * @stable ICU 2.0
+ */
+typedef double UDate;
+
+/** The number of milliseconds per second @stable ICU 2.0 */
+#define U_MILLIS_PER_SECOND        (1000)
+/** The number of milliseconds per minute @stable ICU 2.0 */
+#define U_MILLIS_PER_MINUTE       (60000)
+/** The number of milliseconds per hour @stable ICU 2.0 */
+#define U_MILLIS_PER_HOUR       (3600000)
+/** The number of milliseconds per day @stable ICU 2.0 */
+#define U_MILLIS_PER_DAY       (86400000)
+
+/** 
+ * Maximum UDate value 
+ * @stable ICU 4.8 
+ */ 
+#define U_DATE_MAX DBL_MAX
+
+/**
+ * Minimum UDate value 
+ * @stable ICU 4.8 
+ */ 
+#define U_DATE_MIN -U_DATE_MAX
+
+/*===========================================================================*/
+/* Shared library/DLL import-export API control                              */
+/*===========================================================================*/
+
+/*
+ * Control of symbol import/export.
+ * ICU is separated into three libraries.
+ */
+
+/**
+ * \def U_COMBINED_IMPLEMENTATION
+ * Set to export library symbols from inside the ICU library
+ * when all of ICU is in a single library.
+ * This can be set as a compiler option while building ICU, and it
+ * needs to be the first one tested to override U_COMMON_API, U_I18N_API, etc.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_DATA_API
+ * Set to export library symbols from inside the stubdata library,
+ * and to import them from outside.
+ * @stable ICU 3.0
+ */
+
+/**
+ * \def U_COMMON_API
+ * Set to export library symbols from inside the common library,
+ * and to import them from outside.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_I18N_API
+ * Set to export library symbols from inside the i18n library,
+ * and to import them from outside.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_LAYOUT_API
+ * Set to export library symbols from inside the layout engine library,
+ * and to import them from outside.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_LAYOUTEX_API
+ * Set to export library symbols from inside the layout extensions library,
+ * and to import them from outside.
+ * @stable ICU 2.6
+ */
+
+/**
+ * \def U_IO_API
+ * Set to export library symbols from inside the ustdio library,
+ * and to import them from outside.
+ * @stable ICU 2.0
+ */
+
+/**
+ * \def U_TOOLUTIL_API
+ * Set to export library symbols from inside the toolutil library,
+ * and to import them from outside.
+ * @stable ICU 3.4
+ */
+
+#ifdef U_IN_DOXYGEN
+// This definition is required when generating the API docs.
+#define U_COMBINED_IMPLEMENTATION 1
+#endif
+
+#if defined(U_COMBINED_IMPLEMENTATION)
+#define U_DATA_API     U_EXPORT
+#define U_COMMON_API   U_EXPORT
+#define U_I18N_API     U_EXPORT
+#define U_LAYOUT_API   U_EXPORT
+#define U_LAYOUTEX_API U_EXPORT
+#define U_IO_API       U_EXPORT
+#define U_TOOLUTIL_API U_EXPORT
+#elif defined(U_STATIC_IMPLEMENTATION)
+#define U_DATA_API
+#define U_COMMON_API
+#define U_I18N_API
+#define U_LAYOUT_API
+#define U_LAYOUTEX_API
+#define U_IO_API
+#define U_TOOLUTIL_API
+#elif defined(U_COMMON_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_EXPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_IMPORT
+#elif defined(U_I18N_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_EXPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_IMPORT
+#elif defined(U_LAYOUT_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_EXPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_IMPORT
+#elif defined(U_LAYOUTEX_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_EXPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_IMPORT
+#elif defined(U_IO_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_EXPORT
+#define U_TOOLUTIL_API U_IMPORT
+#elif defined(U_TOOLUTIL_IMPLEMENTATION)
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_EXPORT
+#else
+#define U_DATA_API     U_IMPORT
+#define U_COMMON_API   U_IMPORT
+#define U_I18N_API     U_IMPORT
+#define U_LAYOUT_API   U_IMPORT
+#define U_LAYOUTEX_API U_IMPORT
+#define U_IO_API       U_IMPORT
+#define U_TOOLUTIL_API U_IMPORT
+#endif
+
+/**
+ * \def U_STANDARD_CPP_NAMESPACE
+ * Control of C++ Namespace
+ * @stable ICU 2.0
+ */
+#ifdef __cplusplus
+#define U_STANDARD_CPP_NAMESPACE        ::
+#else
+#define U_STANDARD_CPP_NAMESPACE
+#endif
+
+/*===========================================================================*/
+/* UErrorCode */
+/*===========================================================================*/
+
+/**
+ * Standard ICU4C error code type, a substitute for exceptions.
+ *
+ * Initialize the UErrorCode with U_ZERO_ERROR, and check for success or
+ * failure using U_SUCCESS() or U_FAILURE():
+ *
+ *     UErrorCode errorCode = U_ZERO_ERROR;
+ *     // call ICU API that needs an error code parameter.
+ *     if (U_FAILURE(errorCode)) {
+ *         // An error occurred. Handle it here.
+ *     }
+ *
+ * C++ code should use icu::ErrorCode, available in unicode/errorcode.h, or a
+ * suitable subclass.
+ *
+ * For more information, see:
+ * http://icu-project.org/userguide/conventions
+ *
+ * Note: By convention, ICU functions that take a reference (C++) or a pointer
+ * (C) to a UErrorCode first test:
+ *
+ *     if (U_FAILURE(errorCode)) { return immediately; }
+ *
+ * so that in a chain of such functions the first one that sets an error code
+ * causes the following ones to not perform any operations.
+ *
+ * @stable ICU 2.0
+ */
+typedef enum UErrorCode {
+    /* The ordering of U_ERROR_INFO_START Vs U_USING_FALLBACK_WARNING looks weird
+     * and is that way because VC++ debugger displays first encountered constant,
+     * which is not the what the code is used for
+     */
+
+    U_USING_FALLBACK_WARNING  = -128,   /**< A resource bundle lookup returned a fallback result (not an error) */
+
+    U_ERROR_WARNING_START     = -128,   /**< Start of information results (semantically successful) */
+
+    U_USING_DEFAULT_WARNING   = -127,   /**< A resource bundle lookup returned a result from the root locale (not an error) */
+
+    U_SAFECLONE_ALLOCATED_WARNING = -126, /**< A SafeClone operation required allocating memory (informational only) */
+
+    U_STATE_OLD_WARNING       = -125,   /**< ICU has to use compatibility layer to construct the service. Expect performance/memory usage degradation. Consider upgrading */
+
+    U_STRING_NOT_TERMINATED_WARNING = -124,/**< An output string could not be NUL-terminated because output length==destCapacity. */
+
+    U_SORT_KEY_TOO_SHORT_WARNING = -123, /**< Number of levels requested in getBound is higher than the number of levels in the sort key */
+
+    U_AMBIGUOUS_ALIAS_WARNING = -122,   /**< This converter alias can go to different converter implementations */
+
+    U_DIFFERENT_UCA_VERSION = -121,     /**< ucol_open encountered a mismatch between UCA version and collator image version, so the collator was constructed from rules. No impact to further function */
+    
+    U_PLUGIN_CHANGED_LEVEL_WARNING = -120, /**< A plugin caused a level change. May not be an error, but later plugins may not load. */
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal UErrorCode warning value.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_ERROR_WARNING_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    U_ZERO_ERROR              =  0,     /**< No error, no warning. */
+
+    U_ILLEGAL_ARGUMENT_ERROR  =  1,     /**< Start of codes indicating failure */
+    U_MISSING_RESOURCE_ERROR  =  2,     /**< The requested resource cannot be found */
+    U_INVALID_FORMAT_ERROR    =  3,     /**< Data format is not what is expected */
+    U_FILE_ACCESS_ERROR       =  4,     /**< The requested file cannot be found */
+    U_INTERNAL_PROGRAM_ERROR  =  5,     /**< Indicates a bug in the library code */
+    U_MESSAGE_PARSE_ERROR     =  6,     /**< Unable to parse a message (message format) */
+    U_MEMORY_ALLOCATION_ERROR =  7,     /**< Memory allocation error */
+    U_INDEX_OUTOFBOUNDS_ERROR =  8,     /**< Trying to access the index that is out of bounds */
+    U_PARSE_ERROR             =  9,     /**< Equivalent to Java ParseException */
+    U_INVALID_CHAR_FOUND      = 10,     /**< Character conversion: Unmappable input sequence. In other APIs: Invalid character. */
+    U_TRUNCATED_CHAR_FOUND    = 11,     /**< Character conversion: Incomplete input sequence. */
+    U_ILLEGAL_CHAR_FOUND      = 12,     /**< Character conversion: Illegal input sequence/combination of input units. */
+    U_INVALID_TABLE_FORMAT    = 13,     /**< Conversion table file found, but corrupted */
+    U_INVALID_TABLE_FILE      = 14,     /**< Conversion table file not found */
+    U_BUFFER_OVERFLOW_ERROR   = 15,     /**< A result would not fit in the supplied buffer */
+    U_UNSUPPORTED_ERROR       = 16,     /**< Requested operation not supported in current context */
+    U_RESOURCE_TYPE_MISMATCH  = 17,     /**< an operation is requested over a resource that does not support it */
+    U_ILLEGAL_ESCAPE_SEQUENCE = 18,     /**< ISO-2022 illegal escape sequence */
+    U_UNSUPPORTED_ESCAPE_SEQUENCE = 19, /**< ISO-2022 unsupported escape sequence */
+    U_NO_SPACE_AVAILABLE      = 20,     /**< No space available for in-buffer expansion for Arabic shaping */
+    U_CE_NOT_FOUND_ERROR      = 21,     /**< Currently used only while setting variable top, but can be used generally */
+    U_PRIMARY_TOO_LONG_ERROR  = 22,     /**< User tried to set variable top to a primary that is longer than two bytes */
+    U_STATE_TOO_OLD_ERROR     = 23,     /**< ICU cannot construct a service from this state, as it is no longer supported */
+    U_TOO_MANY_ALIASES_ERROR  = 24,     /**< There are too many aliases in the path to the requested resource.
+                                             It is very possible that a circular alias definition has occurred */
+    U_ENUM_OUT_OF_SYNC_ERROR  = 25,     /**< UEnumeration out of sync with underlying collection */
+    U_INVARIANT_CONVERSION_ERROR = 26,  /**< Unable to convert a UChar* string to char* with the invariant converter. */
+    U_INVALID_STATE_ERROR     = 27,     /**< Requested operation can not be completed with ICU in its current state */
+    U_COLLATOR_VERSION_MISMATCH = 28,   /**< Collator version is not compatible with the base version */
+    U_USELESS_COLLATOR_ERROR  = 29,     /**< Collator is options only and no base is specified */
+    U_NO_WRITE_PERMISSION     = 30,     /**< Attempt to modify read-only or constant data. */
+#ifndef U_HIDE_DRAFT_API
+    /**
+     * The input is impractically long for an operation.
+     * It is rejected because it may lead to problems such as excessive
+     * processing time, stack depth, or heap memory requirements.
+     *
+     * @draft ICU 68
+     */
+    U_INPUT_TOO_LONG_ERROR = 31,
+#endif  // U_HIDE_DRAFT_API
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest standard error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_STANDARD_ERROR_LIMIT = 32,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /*
+     * Error codes in the range 0x10000 0x10100 are reserved for Transliterator.
+     */
+    U_BAD_VARIABLE_DEFINITION=0x10000,/**< Missing '$' or duplicate variable name */
+    U_PARSE_ERROR_START = 0x10000,    /**< Start of Transliterator errors */
+    U_MALFORMED_RULE,                 /**< Elements of a rule are misplaced */
+    U_MALFORMED_SET,                  /**< A UnicodeSet pattern is invalid*/
+    U_MALFORMED_SYMBOL_REFERENCE,     /**< UNUSED as of ICU 2.4 */
+    U_MALFORMED_UNICODE_ESCAPE,       /**< A Unicode escape pattern is invalid*/
+    U_MALFORMED_VARIABLE_DEFINITION,  /**< A variable definition is invalid */
+    U_MALFORMED_VARIABLE_REFERENCE,   /**< A variable reference is invalid */
+    U_MISMATCHED_SEGMENT_DELIMITERS,  /**< UNUSED as of ICU 2.4 */
+    U_MISPLACED_ANCHOR_START,         /**< A start anchor appears at an illegal position */
+    U_MISPLACED_CURSOR_OFFSET,        /**< A cursor offset occurs at an illegal position */
+    U_MISPLACED_QUANTIFIER,           /**< A quantifier appears after a segment close delimiter */
+    U_MISSING_OPERATOR,               /**< A rule contains no operator */
+    U_MISSING_SEGMENT_CLOSE,          /**< UNUSED as of ICU 2.4 */
+    U_MULTIPLE_ANTE_CONTEXTS,         /**< More than one ante context */
+    U_MULTIPLE_CURSORS,               /**< More than one cursor */
+    U_MULTIPLE_POST_CONTEXTS,         /**< More than one post context */
+    U_TRAILING_BACKSLASH,             /**< A dangling backslash */
+    U_UNDEFINED_SEGMENT_REFERENCE,    /**< A segment reference does not correspond to a defined segment */
+    U_UNDEFINED_VARIABLE,             /**< A variable reference does not correspond to a defined variable */
+    U_UNQUOTED_SPECIAL,               /**< A special character was not quoted or escaped */
+    U_UNTERMINATED_QUOTE,             /**< A closing single quote is missing */
+    U_RULE_MASK_ERROR,                /**< A rule is hidden by an earlier more general rule */
+    U_MISPLACED_COMPOUND_FILTER,      /**< A compound filter is in an invalid location */
+    U_MULTIPLE_COMPOUND_FILTERS,      /**< More than one compound filter */
+    U_INVALID_RBT_SYNTAX,             /**< A "::id" rule was passed to the RuleBasedTransliterator parser */
+    U_INVALID_PROPERTY_PATTERN,       /**< UNUSED as of ICU 2.4 */
+    U_MALFORMED_PRAGMA,               /**< A 'use' pragma is invalid */
+    U_UNCLOSED_SEGMENT,               /**< A closing ')' is missing */
+    U_ILLEGAL_CHAR_IN_SEGMENT,        /**< UNUSED as of ICU 2.4 */
+    U_VARIABLE_RANGE_EXHAUSTED,       /**< Too many stand-ins generated for the given variable range */
+    U_VARIABLE_RANGE_OVERLAP,         /**< The variable range overlaps characters used in rules */
+    U_ILLEGAL_CHARACTER,              /**< A special character is outside its allowed context */
+    U_INTERNAL_TRANSLITERATOR_ERROR,  /**< Internal transliterator system error */
+    U_INVALID_ID,                     /**< A "::id" rule specifies an unknown transliterator */
+    U_INVALID_FUNCTION,               /**< A "&fn()" rule specifies an unknown transliterator */
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal Transliterator error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_PARSE_ERROR_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /*
+     * Error codes in the range 0x10100 0x10200 are reserved for the formatting API.
+     */
+    U_UNEXPECTED_TOKEN=0x10100,       /**< Syntax error in format pattern */
+    U_FMT_PARSE_ERROR_START=0x10100,  /**< Start of format library errors */
+    U_MULTIPLE_DECIMAL_SEPARATORS,    /**< More than one decimal separator in number pattern */
+    U_MULTIPLE_DECIMAL_SEPERATORS = U_MULTIPLE_DECIMAL_SEPARATORS, /**< Typo: kept for backward compatibility. Use U_MULTIPLE_DECIMAL_SEPARATORS */
+    U_MULTIPLE_EXPONENTIAL_SYMBOLS,   /**< More than one exponent symbol in number pattern */
+    U_MALFORMED_EXPONENTIAL_PATTERN,  /**< Grouping symbol in exponent pattern */
+    U_MULTIPLE_PERCENT_SYMBOLS,       /**< More than one percent symbol in number pattern */
+    U_MULTIPLE_PERMILL_SYMBOLS,       /**< More than one permill symbol in number pattern */
+    U_MULTIPLE_PAD_SPECIFIERS,        /**< More than one pad symbol in number pattern */
+    U_PATTERN_SYNTAX_ERROR,           /**< Syntax error in format pattern */
+    U_ILLEGAL_PAD_POSITION,           /**< Pad symbol misplaced in number pattern */
+    U_UNMATCHED_BRACES,               /**< Braces do not match in message pattern */
+    U_UNSUPPORTED_PROPERTY,           /**< UNUSED as of ICU 2.4 */
+    U_UNSUPPORTED_ATTRIBUTE,          /**< UNUSED as of ICU 2.4 */
+    U_ARGUMENT_TYPE_MISMATCH,         /**< Argument name and argument index mismatch in MessageFormat functions */
+    U_DUPLICATE_KEYWORD,              /**< Duplicate keyword in PluralFormat */
+    U_UNDEFINED_KEYWORD,              /**< Undefined Plural keyword */
+    U_DEFAULT_KEYWORD_MISSING,        /**< Missing DEFAULT rule in plural rules */
+    U_DECIMAL_NUMBER_SYNTAX_ERROR,    /**< Decimal number syntax error */
+    U_FORMAT_INEXACT_ERROR,           /**< Cannot format a number exactly and rounding mode is ROUND_UNNECESSARY @stable ICU 4.8 */
+    U_NUMBER_ARG_OUTOFBOUNDS_ERROR,   /**< The argument to a NumberFormatter helper method was out of bounds; the bounds are usually 0 to 999. @stable ICU 61 */
+    U_NUMBER_SKELETON_SYNTAX_ERROR,   /**< The number skeleton passed to C++ NumberFormatter or C UNumberFormatter was invalid or contained a syntax error. @stable ICU 62 */
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal formatting API error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_FMT_PARSE_ERROR_LIMIT = 0x10114,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /*
+     * Error codes in the range 0x10200 0x102ff are reserved for BreakIterator.
+     */
+    U_BRK_INTERNAL_ERROR=0x10200,          /**< An internal error (bug) was detected.             */
+    U_BRK_ERROR_START=0x10200,             /**< Start of codes indicating Break Iterator failures */
+    U_BRK_HEX_DIGITS_EXPECTED,             /**< Hex digits expected as part of a escaped char in a rule. */
+    U_BRK_SEMICOLON_EXPECTED,              /**< Missing ';' at the end of a RBBI rule.            */
+    U_BRK_RULE_SYNTAX,                     /**< Syntax error in RBBI rule.                        */
+    U_BRK_UNCLOSED_SET,                    /**< UnicodeSet writing an RBBI rule missing a closing ']'. */
+    U_BRK_ASSIGN_ERROR,                    /**< Syntax error in RBBI rule assignment statement.   */
+    U_BRK_VARIABLE_REDFINITION,            /**< RBBI rule $Variable redefined.                    */
+    U_BRK_MISMATCHED_PAREN,                /**< Mis-matched parentheses in an RBBI rule.          */
+    U_BRK_NEW_LINE_IN_QUOTED_STRING,       /**< Missing closing quote in an RBBI rule.            */
+    U_BRK_UNDEFINED_VARIABLE,              /**< Use of an undefined $Variable in an RBBI rule.    */
+    U_BRK_INIT_ERROR,                      /**< Initialization failure.  Probable missing ICU Data. */
+    U_BRK_RULE_EMPTY_SET,                  /**< Rule contains an empty Unicode Set.               */
+    U_BRK_UNRECOGNIZED_OPTION,             /**< !!option in RBBI rules not recognized.            */
+    U_BRK_MALFORMED_RULE_TAG,              /**< The {nnn} tag on a rule is malformed              */
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal BreakIterator error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_BRK_ERROR_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /*
+     * Error codes in the range 0x10300-0x103ff are reserved for regular expression related errors.
+     */
+    U_REGEX_INTERNAL_ERROR=0x10300,       /**< An internal error (bug) was detected.              */
+    U_REGEX_ERROR_START=0x10300,          /**< Start of codes indicating Regexp failures          */
+    U_REGEX_RULE_SYNTAX,                  /**< Syntax error in regexp pattern.                    */
+    U_REGEX_INVALID_STATE,                /**< RegexMatcher in invalid state for requested operation */
+    U_REGEX_BAD_ESCAPE_SEQUENCE,          /**< Unrecognized backslash escape sequence in pattern  */
+    U_REGEX_PROPERTY_SYNTAX,              /**< Incorrect Unicode property                         */
+    U_REGEX_UNIMPLEMENTED,                /**< Use of regexp feature that is not yet implemented. */
+    U_REGEX_MISMATCHED_PAREN,             /**< Incorrectly nested parentheses in regexp pattern.  */
+    U_REGEX_NUMBER_TOO_BIG,               /**< Decimal number is too large.                       */
+    U_REGEX_BAD_INTERVAL,                 /**< Error in {min,max} interval                        */
+    U_REGEX_MAX_LT_MIN,                   /**< In {min,max}, max is less than min.                */
+    U_REGEX_INVALID_BACK_REF,             /**< Back-reference to a non-existent capture group.    */
+    U_REGEX_INVALID_FLAG,                 /**< Invalid value for match mode flags.                */
+    U_REGEX_LOOK_BEHIND_LIMIT,            /**< Look-Behind pattern matches must have a bounded maximum length.    */
+    U_REGEX_SET_CONTAINS_STRING,          /**< Regexps cannot have UnicodeSets containing strings.*/
+#ifndef U_HIDE_DEPRECATED_API
+    U_REGEX_OCTAL_TOO_BIG,                /**< Octal character constants must be <= 0377. @deprecated ICU 54. This error cannot occur. */
+#endif  /* U_HIDE_DEPRECATED_API */
+    U_REGEX_MISSING_CLOSE_BRACKET=U_REGEX_SET_CONTAINS_STRING+2, /**< Missing closing bracket on a bracket expression. */
+    U_REGEX_INVALID_RANGE,                /**< In a character range [x-y], x is greater than y.   */
+    U_REGEX_STACK_OVERFLOW,               /**< Regular expression backtrack stack overflow.       */
+    U_REGEX_TIME_OUT,                     /**< Maximum allowed match time exceeded                */
+    U_REGEX_STOPPED_BY_CALLER,            /**< Matching operation aborted by user callback fn.    */
+    U_REGEX_PATTERN_TOO_BIG,              /**< Pattern exceeds limits on size or complexity. @stable ICU 55 */
+    U_REGEX_INVALID_CAPTURE_GROUP_NAME,   /**< Invalid capture group name. @stable ICU 55 */
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal regular expression error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_REGEX_ERROR_LIMIT=U_REGEX_STOPPED_BY_CALLER+3,
+#endif  // U_HIDE_DEPRECATED_API
+
+    /*
+     * Error codes in the range 0x10400-0x104ff are reserved for IDNA related error codes.
+     */
+    U_IDNA_PROHIBITED_ERROR=0x10400,
+    U_IDNA_ERROR_START=0x10400,
+    U_IDNA_UNASSIGNED_ERROR,
+    U_IDNA_CHECK_BIDI_ERROR,
+    U_IDNA_STD3_ASCII_RULES_ERROR,
+    U_IDNA_ACE_PREFIX_ERROR,
+    U_IDNA_VERIFICATION_ERROR,
+    U_IDNA_LABEL_TOO_LONG_ERROR,
+    U_IDNA_ZERO_LENGTH_LABEL_ERROR,
+    U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR,
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal IDNA error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_IDNA_ERROR_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+    /*
+     * Aliases for StringPrep
+     */
+    U_STRINGPREP_PROHIBITED_ERROR = U_IDNA_PROHIBITED_ERROR,
+    U_STRINGPREP_UNASSIGNED_ERROR = U_IDNA_UNASSIGNED_ERROR,
+    U_STRINGPREP_CHECK_BIDI_ERROR = U_IDNA_CHECK_BIDI_ERROR,
+    
+    /*
+     * Error codes in the range 0x10500-0x105ff are reserved for Plugin related error codes.
+     */
+    U_PLUGIN_ERROR_START=0x10500,         /**< Start of codes indicating plugin failures */
+    U_PLUGIN_TOO_HIGH=0x10500,            /**< The plugin's level is too high to be loaded right now. */
+    U_PLUGIN_DIDNT_SET_LEVEL,             /**< The plugin didn't call uplug_setPlugLevel in response to a QUERY */
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal plug-in error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_PLUGIN_ERROR_LIMIT,
+#endif  // U_HIDE_DEPRECATED_API
+
+#ifndef U_HIDE_DEPRECATED_API
+    /**
+     * One more than the highest normal error code.
+     * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
+     */
+    U_ERROR_LIMIT=U_PLUGIN_ERROR_LIMIT
+#endif  // U_HIDE_DEPRECATED_API
+} UErrorCode;
+
+/* Use the following to determine if an UErrorCode represents */
+/* operational success or failure. */
+
+#ifdef __cplusplus
+    /**
+     * Does the error code indicate success?
+     * @stable ICU 2.0
+     */
+    static
+    inline UBool U_SUCCESS(UErrorCode code) { return (UBool)(code<=U_ZERO_ERROR); }
+    /**
+     * Does the error code indicate a failure?
+     * @stable ICU 2.0
+     */
+    static
+    inline UBool U_FAILURE(UErrorCode code) { return (UBool)(code>U_ZERO_ERROR); }
+#else
+    /**
+     * Does the error code indicate success?
+     * @stable ICU 2.0
+     */
+#   define U_SUCCESS(x) ((x)<=U_ZERO_ERROR)
+    /**
+     * Does the error code indicate a failure?
+     * @stable ICU 2.0
+     */
+#   define U_FAILURE(x) ((x)>U_ZERO_ERROR)
+#endif
+
+/**
+ * Return a string for a UErrorCode value.
+ * The string will be the same as the name of the error code constant
+ * in the UErrorCode enum above.
+ * @stable ICU 2.0
+ */
+U_CAPI const char * U_EXPORT2
+u_errorName(UErrorCode code);
+
+
+#endif /* _UTYPES */
diff --git a/thirdparty/icu4c/common/unicode/uvernum.h b/thirdparty/icu4c/common/unicode/uvernum.h
new file mode 100644
index 0000000000..a4cbb9e0fe
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uvernum.h
@@ -0,0 +1,198 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*
+*   file name:  uvernum.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   Created by: Vladimir Weinstein
+*   Updated by: Steven R. Loomis
+*
+*/
+
+/**
+ * \file
+ * \brief C API: definitions of ICU version numbers
+ *
+ * This file is included by uversion.h and other files. This file contains only
+ * macros and definitions. The actual version numbers are defined here.
+ */
+
+ /*
+  * IMPORTANT: When updating version, the following things need to be done:
+  * source/common/unicode/uvernum.h - this file: update major, minor,
+  *        patchlevel, suffix, version, short version constants, namespace,
+  *                    renaming macro, and copyright
+  *
+  * The following files need to be updated as well, which can be done
+  *  by running the UNIX makefile target 'update-windows-makefiles' in icu/source.
+  *
+  *
+  * source/common/common_uwp.vcxproj
+  * source/common/common.vcxproj - update 'Output file name' on the link tab so
+  *                   that it contains the new major/minor combination
+  * source/i18n/i18n.vcxproj - same as for the common.vcxproj
+  * source/i18n/i18n_uwp.vcxproj - same as for the common_uwp.vcxproj
+  * source/layoutex/layoutex.vcproj - same
+  * source/stubdata/stubdata.vcproj - same as for the common.vcxproj
+  * source/io/io.vcproj - same as for the common.vcxproj
+  * source/data/makedata.mak - change U_ICUDATA_NAME so that it contains
+  *                            the new major/minor combination and the Unicode version.
+  */
+
+#ifndef UVERNUM_H
+#define UVERNUM_H
+
+/** The standard copyright notice that gets compiled into each library.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_COPYRIGHT_STRING \
+  " Copyright (C) 2016 and later: Unicode, Inc. and others. License & terms of use: http://www.unicode.org/copyright.html "
+
+/** The current ICU major version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION_MAJOR_NUM 68
+
+/** The current ICU minor version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.6
+ */
+#define U_ICU_VERSION_MINOR_NUM 1
+
+/** The current ICU patchlevel version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION_PATCHLEVEL_NUM 0
+
+/** The current ICU build level version as an integer.
+ *  This value is for use by ICU clients. It defaults to 0.
+ *  @stable ICU 4.0
+ */
+#ifndef U_ICU_VERSION_BUILDLEVEL_NUM
+#define U_ICU_VERSION_BUILDLEVEL_NUM 0
+#endif
+
+/** Glued version suffix for renamers
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.6
+ */
+#define U_ICU_VERSION_SUFFIX _68
+
+/**
+ * \def U_DEF2_ICU_ENTRY_POINT_RENAME
+ * @internal
+ */
+/**
+ * \def U_DEF_ICU_ENTRY_POINT_RENAME
+ * @internal
+ */
+/** Glued version suffix function for renamers
+ *  This value will change in the subsequent releases of ICU.
+ *  If a custom suffix (such as matching library suffixes) is desired, this can be modified.
+ *  Note that if present, platform.h may contain an earlier definition of this macro.
+ *  \def U_ICU_ENTRY_POINT_RENAME
+ *  @stable ICU 4.2
+ */
+/**
+ * Disable the version suffix. Use the custom suffix if exists.
+ * \def U_DISABLE_VERSION_SUFFIX
+ * @internal
+ */
+#ifndef U_DISABLE_VERSION_SUFFIX
+#define U_DISABLE_VERSION_SUFFIX 0
+#endif
+
+#ifndef U_ICU_ENTRY_POINT_RENAME
+#ifdef U_HAVE_LIB_SUFFIX
+#   if !U_DISABLE_VERSION_SUFFIX
+#       define U_DEF_ICU_ENTRY_POINT_RENAME(x,y,z) x ## y ##  z
+#       define U_DEF2_ICU_ENTRY_POINT_RENAME(x,y,z) U_DEF_ICU_ENTRY_POINT_RENAME(x,y,z)
+#       define U_ICU_ENTRY_POINT_RENAME(x)    U_DEF2_ICU_ENTRY_POINT_RENAME(x,U_ICU_VERSION_SUFFIX,U_LIB_SUFFIX_C_NAME)
+#   else
+#       define U_DEF_ICU_ENTRY_POINT_RENAME(x,y) x ## y
+#       define U_DEF2_ICU_ENTRY_POINT_RENAME(x,y) U_DEF_ICU_ENTRY_POINT_RENAME(x,y)
+#       define U_ICU_ENTRY_POINT_RENAME(x)    U_DEF2_ICU_ENTRY_POINT_RENAME(x,U_LIB_SUFFIX_C_NAME)
+#   endif
+#else
+#   if !U_DISABLE_VERSION_SUFFIX
+#       define U_DEF_ICU_ENTRY_POINT_RENAME(x,y) x ## y
+#       define U_DEF2_ICU_ENTRY_POINT_RENAME(x,y) U_DEF_ICU_ENTRY_POINT_RENAME(x,y)
+#       define U_ICU_ENTRY_POINT_RENAME(x)    U_DEF2_ICU_ENTRY_POINT_RENAME(x,U_ICU_VERSION_SUFFIX)
+#   else
+#       define U_ICU_ENTRY_POINT_RENAME(x)    x
+#   endif
+#endif
+#endif
+
+/** The current ICU library version as a dotted-decimal string. The patchlevel
+ *  only appears in this string if it non-zero.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION "68.1"
+
+/**
+ * The current ICU library major version number as a string, for library name suffixes.
+ * This value will change in subsequent releases of ICU.
+ *
+ * Until ICU 4.8, this was the combination of the single-digit major and minor ICU version numbers
+ * into one string without dots ("48").
+ * Since ICU 49, it is the double-digit major ICU version number.
+ * See https://unicode-org.github.io/icu/userguide/design#version-numbers-in-icu
+ *
+ * @stable ICU 2.6
+ */
+#define U_ICU_VERSION_SHORT "68"
+
+#ifndef U_HIDE_INTERNAL_API
+/** Data version in ICU4C.
+ * @internal ICU 4.4 Internal Use Only
+ **/
+#define U_ICU_DATA_VERSION "68.1"
+#endif  /* U_HIDE_INTERNAL_API */
+
+/*===========================================================================
+ * ICU collation framework version information
+ * Version info that can be obtained from a collator is affected by these
+ * numbers in a secret and magic way. Please use collator version as whole
+ *===========================================================================
+ */
+
+/**
+ * Collation runtime version (sort key generator, strcoll).
+ * If the version is different, sort keys for the same string could be different.
+ * This value may change in subsequent releases of ICU.
+ * @stable ICU 2.4
+ */
+#define UCOL_RUNTIME_VERSION 9
+
+/**
+ * Collation builder code version.
+ * When this is different, the same tailoring might result
+ * in assigning different collation elements to code points.
+ * This value may change in subsequent releases of ICU.
+ * @stable ICU 2.4
+ */
+#define UCOL_BUILDER_VERSION 9
+
+#ifndef U_HIDE_DEPRECATED_API
+/**
+ * Constant 1.
+ * This was intended to be the version of collation tailorings,
+ * but instead the tailoring data carries a version number.
+ * @deprecated ICU 54
+ */
+#define UCOL_TAILORINGS_VERSION 1
+#endif  /* U_HIDE_DEPRECATED_API */
+
+#endif
diff --git a/thirdparty/icu4c/common/unicode/uversion.h b/thirdparty/icu4c/common/unicode/uversion.h
new file mode 100644
index 0000000000..113568df8c
--- /dev/null
+++ b/thirdparty/icu4c/common/unicode/uversion.h
@@ -0,0 +1,187 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2000-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*
+*   file name:  uversion.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   Created by: Vladimir Weinstein
+*
+*  Gets included by utypes.h and Windows .rc files
+*/
+
+/**
+ * \file
+ * \brief C API: API for accessing ICU version numbers. 
+ */
+/*===========================================================================*/
+/* Main ICU version information                                              */
+/*===========================================================================*/
+
+#ifndef UVERSION_H
+#define UVERSION_H
+
+#include "unicode/umachine.h"
+
+/* Actual version info lives in uvernum.h */
+#include "unicode/uvernum.h"
+
+/** Maximum length of the copyright string.
+ *  @stable ICU 2.4
+ */
+#define U_COPYRIGHT_STRING_LENGTH  128
+
+/** An ICU version consists of up to 4 numbers from 0..255.
+ *  @stable ICU 2.4
+ */
+#define U_MAX_VERSION_LENGTH 4
+
+/** In a string, ICU version fields are delimited by dots.
+ *  @stable ICU 2.4
+ */
+#define U_VERSION_DELIMITER '.'
+
+/** The maximum length of an ICU version string.
+ *  @stable ICU 2.4
+ */
+#define U_MAX_VERSION_STRING_LENGTH 20
+
+/** The binary form of a version on ICU APIs is an array of 4 uint8_t.
+ *  To compare two versions, use memcmp(v1,v2,sizeof(UVersionInfo)).
+ *  @stable ICU 2.4
+ */
+typedef uint8_t UVersionInfo[U_MAX_VERSION_LENGTH];
+
+/*===========================================================================*/
+/* C++ namespace if supported. Versioned unless versioning is disabled.      */
+/*===========================================================================*/
+
+/* Define C++ namespace symbols. */
+#ifdef __cplusplus
+
+/**
+ * \def U_NAMESPACE_BEGIN
+ * This is used to begin a declaration of a public ICU C++ API within
+ * versioned-ICU-namespace block.
+ *
+ * @stable ICU 2.4
+ */
+
+/**
+ * \def U_NAMESPACE_END
+ * This is used to end a declaration of a public ICU C++ API.
+ * It ends the versioned-ICU-namespace block begun by U_NAMESPACE_BEGIN.
+ *
+ * @stable ICU 2.4
+ */
+
+/**
+ * \def U_NAMESPACE_USE
+ * This is used to specify that the rest of the code uses the
+ * public ICU C++ API namespace.
+ * @stable ICU 2.4
+ */
+
+/**
+ * \def U_NAMESPACE_QUALIFIER
+ * This is used to qualify that a function or class is part of
+ * the public ICU C++ API namespace.
+ *
+ * This macro is unnecessary since ICU 49 requires namespace support.
+ * You can just use "icu::" instead.
+ * @stable ICU 2.4
+ */
+
+#   if U_DISABLE_RENAMING
+#       define U_ICU_NAMESPACE icu
+        namespace U_ICU_NAMESPACE { }
+#   else
+#       define U_ICU_NAMESPACE U_ICU_ENTRY_POINT_RENAME(icu)
+        namespace U_ICU_NAMESPACE { }
+        namespace icu = U_ICU_NAMESPACE;
+#   endif
+
+#   define U_NAMESPACE_BEGIN namespace U_ICU_NAMESPACE {
+#   define U_NAMESPACE_END }
+#   define U_NAMESPACE_USE using namespace U_ICU_NAMESPACE;
+#   define U_NAMESPACE_QUALIFIER U_ICU_NAMESPACE::
+
+#   ifndef U_USING_ICU_NAMESPACE
+#       if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || \
+                defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION) || \
+                defined(U_LAYOUTEX_IMPLEMENTATION) || defined(U_TOOLUTIL_IMPLEMENTATION)
+#           define U_USING_ICU_NAMESPACE 0
+#       else
+#           define U_USING_ICU_NAMESPACE 0
+#       endif
+#   endif
+#   if U_USING_ICU_NAMESPACE
+        U_NAMESPACE_USE
+#   endif
+#endif /* __cplusplus */
+
+/*===========================================================================*/
+/* General version helper functions. Definitions in putil.c                  */
+/*===========================================================================*/
+
+/**
+ * Parse a string with dotted-decimal version information and
+ * fill in a UVersionInfo structure with the result.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The destination structure for the version information.
+ * @param versionString A string with dotted-decimal version information,
+ *                      with up to four non-negative number fields with
+ *                      values of up to 255 each.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+u_versionFromString(UVersionInfo versionArray, const char *versionString);
+
+/**
+ * Parse a Unicode string with dotted-decimal version information and
+ * fill in a UVersionInfo structure with the result.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The destination structure for the version information.
+ * @param versionString A Unicode string with dotted-decimal version
+ *                      information, with up to four non-negative number
+ *                      fields with values of up to 255 each.
+ * @stable ICU 4.2
+ */
+U_CAPI void U_EXPORT2
+u_versionFromUString(UVersionInfo versionArray, const UChar *versionString);
+
+
+/**
+ * Write a string with dotted-decimal version information according
+ * to the input UVersionInfo.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The version information to be written as a string.
+ * @param versionString A string buffer that will be filled in with
+ *                      a string corresponding to the numeric version
+ *                      information in versionArray.
+ *                      The buffer size must be at least U_MAX_VERSION_STRING_LENGTH.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+u_versionToString(const UVersionInfo versionArray, char *versionString);
+
+/**
+ * Gets the ICU release version.  The version array stores the version information
+ * for ICU.  For example, release "1.3.31.2" is then represented as 0x01031F02.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray the version # information, the result will be filled in
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+u_getVersion(UVersionInfo versionArray);
+#endif
diff --git a/thirdparty/icu4c/common/unifiedcache.cpp b/thirdparty/icu4c/common/unifiedcache.cpp
new file mode 100644
index 0000000000..493ab79f6d
--- /dev/null
+++ b/thirdparty/icu4c/common/unifiedcache.cpp
@@ -0,0 +1,522 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File unifiedcache.cpp
+******************************************************************************
+*/
+
+#include "unifiedcache.h"
+
+#include <algorithm>      // For std::max()
+#include <mutex>
+
+#include "uassert.h"
+#include "uhash.h"
+#include "ucln_cmn.h"
+
+static icu::UnifiedCache *gCache = NULL;
+static std::mutex *gCacheMutex = nullptr;
+static std::condition_variable *gInProgressValueAddedCond;
+static icu::UInitOnce gCacheInitOnce = U_INITONCE_INITIALIZER;
+
+static const int32_t MAX_EVICT_ITERATIONS = 10;
+static const int32_t DEFAULT_MAX_UNUSED = 1000;
+static const int32_t DEFAULT_PERCENTAGE_OF_IN_USE = 100;
+
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV unifiedcache_cleanup() {
+    gCacheInitOnce.reset();
+    delete gCache;
+    gCache = nullptr;
+    gCacheMutex->~mutex();
+    gCacheMutex = nullptr;
+    gInProgressValueAddedCond->~condition_variable();
+    gInProgressValueAddedCond = nullptr;
+    return TRUE;
+}
+U_CDECL_END
+
+
+U_NAMESPACE_BEGIN
+
+U_CAPI int32_t U_EXPORT2
+ucache_hashKeys(const UHashTok key) {
+    const CacheKeyBase *ckey = (const CacheKeyBase *) key.pointer;
+    return ckey->hashCode();
+}
+
+U_CAPI UBool U_EXPORT2
+ucache_compareKeys(const UHashTok key1, const UHashTok key2) {
+    const CacheKeyBase *p1 = (const CacheKeyBase *) key1.pointer;
+    const CacheKeyBase *p2 = (const CacheKeyBase *) key2.pointer;
+    return *p1 == *p2;
+}
+
+U_CAPI void U_EXPORT2
+ucache_deleteKey(void *obj) {
+    CacheKeyBase *p = (CacheKeyBase *) obj;
+    delete p;
+}
+
+CacheKeyBase::~CacheKeyBase() {
+}
+
+static void U_CALLCONV cacheInit(UErrorCode &status) {
+    U_ASSERT(gCache == NULL);
+    ucln_common_registerCleanup(
+            UCLN_COMMON_UNIFIED_CACHE, unifiedcache_cleanup);
+
+    gCacheMutex = STATIC_NEW(std::mutex);
+    gInProgressValueAddedCond = STATIC_NEW(std::condition_variable);
+    gCache = new UnifiedCache(status);
+    if (gCache == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+    if (U_FAILURE(status)) {
+        delete gCache;
+        gCache = NULL;
+        return;
+    }
+}
+
+UnifiedCache *UnifiedCache::getInstance(UErrorCode &status) {
+    umtx_initOnce(gCacheInitOnce, &cacheInit, status);
+    if (U_FAILURE(status)) {
+        return NULL;
+    }
+    U_ASSERT(gCache != NULL);
+    return gCache;
+}
+
+UnifiedCache::UnifiedCache(UErrorCode &status) :
+        fHashtable(NULL),
+        fEvictPos(UHASH_FIRST),
+        fNumValuesTotal(0),
+        fNumValuesInUse(0),
+        fMaxUnused(DEFAULT_MAX_UNUSED),
+        fMaxPercentageOfInUse(DEFAULT_PERCENTAGE_OF_IN_USE),
+        fAutoEvictedCount(0),
+        fNoValue(nullptr) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    fNoValue = new SharedObject();
+    if (fNoValue == nullptr) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    fNoValue->softRefCount = 1;  // Add fake references to prevent fNoValue from being deleted
+    fNoValue->hardRefCount = 1;  // when other references to it are removed.
+    fNoValue->cachePtr = this;
+
+    fHashtable = uhash_open(
+            &ucache_hashKeys,
+            &ucache_compareKeys,
+            NULL,
+            &status);
+    if (U_FAILURE(status)) {
+        return;
+    }
+    uhash_setKeyDeleter(fHashtable, &ucache_deleteKey);
+}
+
+void UnifiedCache::setEvictionPolicy(
+        int32_t count, int32_t percentageOfInUseItems, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (count < 0 || percentageOfInUseItems < 0) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    fMaxUnused = count;
+    fMaxPercentageOfInUse = percentageOfInUseItems;
+}
+
+int32_t UnifiedCache::unusedCount() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    return uhash_count(fHashtable) - fNumValuesInUse;
+}
+
+int64_t UnifiedCache::autoEvictedCount() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    return fAutoEvictedCount;
+}
+
+int32_t UnifiedCache::keyCount() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    return uhash_count(fHashtable);
+}
+
+void UnifiedCache::flush() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+
+    // Use a loop in case cache items that are flushed held hard references to
+    // other cache items making those additional cache items eligible for
+    // flushing.
+    while (_flush(FALSE));
+}
+
+void UnifiedCache::handleUnreferencedObject() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    --fNumValuesInUse;
+    _runEvictionSlice();
+}
+
+#ifdef UNIFIED_CACHE_DEBUG
+#include <stdio.h>
+
+void UnifiedCache::dump() {
+    UErrorCode status = U_ZERO_ERROR;
+    const UnifiedCache *cache = getInstance(status);
+    if (U_FAILURE(status)) {
+        fprintf(stderr, "Unified Cache: Error fetching cache.\n");
+        return;
+    }
+    cache->dumpContents();
+}
+
+void UnifiedCache::dumpContents() const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    _dumpContents();
+}
+
+// Dumps content of cache.
+// On entry, gCacheMutex must be held.
+// On exit, cache contents dumped to stderr.
+void UnifiedCache::_dumpContents() const {
+    int32_t pos = UHASH_FIRST;
+    const UHashElement *element = uhash_nextElement(fHashtable, &pos);
+    char buffer[256];
+    int32_t cnt = 0;
+    for (; element != NULL; element = uhash_nextElement(fHashtable, &pos)) {
+        const SharedObject *sharedObject =
+                (const SharedObject *) element->value.pointer;
+        const CacheKeyBase *key =
+                (const CacheKeyBase *) element->key.pointer;
+        if (sharedObject->hasHardReferences()) {
+            ++cnt;
+            fprintf(
+                    stderr,
+                    "Unified Cache: Key '%s', error %d, value %p, total refcount %d, soft refcount %d\n",
+                    key->writeDescription(buffer, 256),
+                    key->creationStatus,
+                    sharedObject == fNoValue ? NULL :sharedObject,
+                    sharedObject->getRefCount(),
+                    sharedObject->getSoftRefCount());
+        }
+    }
+    fprintf(stderr, "Unified Cache: %d out of a total of %d still have hard references\n", cnt, uhash_count(fHashtable));
+}
+#endif
+
+UnifiedCache::~UnifiedCache() {
+    // Try our best to clean up first.
+    flush();
+    {
+        // Now all that should be left in the cache are entries that refer to
+        // each other and entries with hard references from outside the cache.
+        // Nothing we can do about these so proceed to wipe out the cache.
+        std::lock_guard<std::mutex> lock(*gCacheMutex);
+        _flush(TRUE);
+    }
+    uhash_close(fHashtable);
+    fHashtable = nullptr;
+    delete fNoValue;
+    fNoValue = nullptr;
+}
+
+const UHashElement *
+UnifiedCache::_nextElement() const {
+    const UHashElement *element = uhash_nextElement(fHashtable, &fEvictPos);
+    if (element == NULL) {
+        fEvictPos = UHASH_FIRST;
+        return uhash_nextElement(fHashtable, &fEvictPos);
+    }
+    return element;
+}
+
+UBool UnifiedCache::_flush(UBool all) const {
+    UBool result = FALSE;
+    int32_t origSize = uhash_count(fHashtable);
+    for (int32_t i = 0; i < origSize; ++i) {
+        const UHashElement *element = _nextElement();
+        if (element == nullptr) {
+            break;
+        }
+        if (all || _isEvictable(element)) {
+            const SharedObject *sharedObject =
+                    (const SharedObject *) element->value.pointer;
+            U_ASSERT(sharedObject->cachePtr == this);
+            uhash_removeElement(fHashtable, element);
+            removeSoftRef(sharedObject);    // Deletes the sharedObject when softRefCount goes to zero.
+            result = TRUE;
+        }
+    }
+    return result;
+}
+
+int32_t UnifiedCache::_computeCountOfItemsToEvict() const {
+    int32_t totalItems = uhash_count(fHashtable);
+    int32_t evictableItems = totalItems - fNumValuesInUse;
+
+    int32_t unusedLimitByPercentage = fNumValuesInUse * fMaxPercentageOfInUse / 100;
+    int32_t unusedLimit = std::max(unusedLimitByPercentage, fMaxUnused);
+    int32_t countOfItemsToEvict = std::max(0, evictableItems - unusedLimit);
+    return countOfItemsToEvict;
+}
+
+void UnifiedCache::_runEvictionSlice() const {
+    int32_t maxItemsToEvict = _computeCountOfItemsToEvict();
+    if (maxItemsToEvict <= 0) {
+        return;
+    }
+    for (int32_t i = 0; i < MAX_EVICT_ITERATIONS; ++i) {
+        const UHashElement *element = _nextElement();
+        if (element == nullptr) {
+            break;
+        }
+        if (_isEvictable(element)) {
+            const SharedObject *sharedObject =
+                    (const SharedObject *) element->value.pointer;
+            uhash_removeElement(fHashtable, element);
+            removeSoftRef(sharedObject);   // Deletes sharedObject when SoftRefCount goes to zero.
+            ++fAutoEvictedCount;
+            if (--maxItemsToEvict == 0) {
+                break;
+            }
+        }
+    }
+}
+
+void UnifiedCache::_putNew(
+        const CacheKeyBase &key,
+        const SharedObject *value,
+        const UErrorCode creationStatus,
+        UErrorCode &status) const {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    CacheKeyBase *keyToAdopt = key.clone();
+    if (keyToAdopt == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    keyToAdopt->fCreationStatus = creationStatus;
+    if (value->softRefCount == 0) {
+        _registerPrimary(keyToAdopt, value);
+    }
+    void *oldValue = uhash_put(fHashtable, keyToAdopt, (void *) value, &status);
+    U_ASSERT(oldValue == nullptr);
+    (void)oldValue;
+    if (U_SUCCESS(status)) {
+        value->softRefCount++;
+    }
+}
+
+void UnifiedCache::_putIfAbsentAndGet(
+        const CacheKeyBase &key,
+        const SharedObject *&value,
+        UErrorCode &status) const {
+    std::lock_guard<std::mutex> lock(*gCacheMutex);
+    const UHashElement *element = uhash_find(fHashtable, &key);
+    if (element != NULL && !_inProgress(element)) {
+        _fetch(element, value, status);
+        return;
+    }
+    if (element == NULL) {
+        UErrorCode putError = U_ZERO_ERROR;
+        // best-effort basis only.
+        _putNew(key, value, status, putError);
+    } else {
+        _put(element, value, status);
+    }
+    // Run an eviction slice. This will run even if we added a primary entry
+    // which doesn't increase the unused count, but that is still o.k
+    _runEvictionSlice();
+}
+
+
+UBool UnifiedCache::_poll(
+        const CacheKeyBase &key,
+        const SharedObject *&value,
+        UErrorCode &status) const {
+    U_ASSERT(value == NULL);
+    U_ASSERT(status == U_ZERO_ERROR);
+    std::unique_lock<std::mutex> lock(*gCacheMutex);
+    const UHashElement *element = uhash_find(fHashtable, &key);
+
+    // If the hash table contains an inProgress placeholder entry for this key,
+    // this means that another thread is currently constructing the value object.
+    // Loop, waiting for that construction to complete.
+     while (element != NULL && _inProgress(element)) {
+         gInProgressValueAddedCond->wait(lock);
+         element = uhash_find(fHashtable, &key);
+    }
+
+    // If the hash table contains an entry for the key,
+    // fetch out the contents and return them.
+    if (element != NULL) {
+         _fetch(element, value, status);
+        return TRUE;
+    }
+
+    // The hash table contained nothing for this key.
+    // Insert an inProgress place holder value.
+    // Our caller will create the final value and update the hash table.
+    _putNew(key, fNoValue, U_ZERO_ERROR, status);
+    return FALSE;
+}
+
+void UnifiedCache::_get(
+        const CacheKeyBase &key,
+        const SharedObject *&value,
+        const void *creationContext,
+        UErrorCode &status) const {
+    U_ASSERT(value == NULL);
+    U_ASSERT(status == U_ZERO_ERROR);
+    if (_poll(key, value, status)) {
+        if (value == fNoValue) {
+            SharedObject::clearPtr(value);
+        }
+        return;
+    }
+    if (U_FAILURE(status)) {
+        return;
+    }
+    value = key.createObject(creationContext, status);
+    U_ASSERT(value == NULL || value->hasHardReferences());
+    U_ASSERT(value != NULL || status != U_ZERO_ERROR);
+    if (value == NULL) {
+        SharedObject::copyPtr(fNoValue, value);
+    }
+    _putIfAbsentAndGet(key, value, status);
+    if (value == fNoValue) {
+        SharedObject::clearPtr(value);
+    }
+}
+
+void UnifiedCache::_registerPrimary(
+            const CacheKeyBase *theKey, const SharedObject *value) const {
+    theKey->fIsPrimary = true;
+    value->cachePtr = this;
+    ++fNumValuesTotal;
+    ++fNumValuesInUse;
+}
+
+void UnifiedCache::_put(
+        const UHashElement *element,
+        const SharedObject *value,
+        const UErrorCode status) const {
+    U_ASSERT(_inProgress(element));
+    const CacheKeyBase *theKey = (const CacheKeyBase *) element->key.pointer;
+    const SharedObject *oldValue = (const SharedObject *) element->value.pointer;
+    theKey->fCreationStatus = status;
+    if (value->softRefCount == 0) {
+        _registerPrimary(theKey, value);
+    }
+    value->softRefCount++;
+    UHashElement *ptr = const_cast<UHashElement *>(element);
+    ptr->value.pointer = (void *) value;
+    U_ASSERT(oldValue == fNoValue);
+    removeSoftRef(oldValue);
+
+    // Tell waiting threads that we replace in-progress status with
+    // an error.
+    gInProgressValueAddedCond->notify_all();
+}
+
+void UnifiedCache::_fetch(
+        const UHashElement *element,
+        const SharedObject *&value,
+        UErrorCode &status) const {
+    const CacheKeyBase *theKey = (const CacheKeyBase *) element->key.pointer;
+    status = theKey->fCreationStatus;
+
+    // Since we have the cache lock, calling regular SharedObject add/removeRef
+    // could cause us to deadlock on ourselves since they may need to lock
+    // the cache mutex.
+    removeHardRef(value);
+    value = static_cast<const SharedObject *>(element->value.pointer);
+    addHardRef(value);
+}
+
+
+UBool UnifiedCache::_inProgress(const UHashElement* element) const {
+    UErrorCode status = U_ZERO_ERROR;
+    const SharedObject * value = NULL;
+    _fetch(element, value, status);
+    UBool result = _inProgress(value, status);
+    removeHardRef(value);
+    return result;
+}
+
+UBool UnifiedCache::_inProgress(
+        const SharedObject* theValue, UErrorCode creationStatus) const {
+    return (theValue == fNoValue && creationStatus == U_ZERO_ERROR);
+}
+
+UBool UnifiedCache::_isEvictable(const UHashElement *element) const
+{
+    const CacheKeyBase *theKey = (const CacheKeyBase *) element->key.pointer;
+    const SharedObject *theValue =
+            (const SharedObject *) element->value.pointer;
+
+    // Entries that are under construction are never evictable
+    if (_inProgress(theValue, theKey->fCreationStatus)) {
+        return FALSE;
+    }
+
+    // We can evict entries that are either not a primary or have just
+    // one reference (The one reference being from the cache itself).
+    return (!theKey->fIsPrimary || (theValue->softRefCount == 1 && theValue->noHardReferences()));
+}
+
+void UnifiedCache::removeSoftRef(const SharedObject *value) const {
+    U_ASSERT(value->cachePtr == this);
+    U_ASSERT(value->softRefCount > 0);
+    if (--value->softRefCount == 0) {
+        --fNumValuesTotal;
+        if (value->noHardReferences()) {
+            delete value;
+        } else {
+            // This path only happens from flush(all). Which only happens from the
+            // UnifiedCache destructor.  Nulling out value.cacheptr changes the behavior
+            // of value.removeRef(), causing the deletion to be done there.
+            value->cachePtr = nullptr;
+        }
+    }
+}
+
+int32_t UnifiedCache::removeHardRef(const SharedObject *value) const {
+    int refCount = 0;
+    if (value) {
+        refCount = umtx_atomic_dec(&value->hardRefCount);
+        U_ASSERT(refCount >= 0);
+        if (refCount == 0) {
+            --fNumValuesInUse;
+        }
+    }
+    return refCount;
+}
+
+int32_t UnifiedCache::addHardRef(const SharedObject *value) const {
+    int refCount = 0;
+    if (value) {
+        refCount = umtx_atomic_inc(&value->hardRefCount);
+        U_ASSERT(refCount >= 1);
+        if (refCount == 1) {
+            fNumValuesInUse++;
+        }
+    }
+    return refCount;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/unifiedcache.h b/thirdparty/icu4c/common/unifiedcache.h
new file mode 100644
index 0000000000..a31998db20
--- /dev/null
+++ b/thirdparty/icu4c/common/unifiedcache.h
@@ -0,0 +1,556 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File UNIFIEDCACHE.H - The ICU Unified cache.
+******************************************************************************
+*/
+
+#ifndef __UNIFIED_CACHE_H__
+#define __UNIFIED_CACHE_H__
+
+#include "utypeinfo.h"  // for 'typeid' to work
+
+#include "unicode/uobject.h"
+#include "unicode/locid.h"
+#include "sharedobject.h"
+#include "unicode/unistr.h"
+#include "cstring.h"
+#include "ustr_imp.h"
+
+struct UHashtable;
+struct UHashElement;
+
+U_NAMESPACE_BEGIN
+
+class UnifiedCache;
+
+/**
+ * A base class for all cache keys.
+ */
+class U_COMMON_API CacheKeyBase : public UObject {
+ public:
+   CacheKeyBase() : fCreationStatus(U_ZERO_ERROR), fIsPrimary(false) {}
+
+   /**
+    * Copy constructor. Needed to support cloning.
+    */
+   CacheKeyBase(const CacheKeyBase &other) 
+           : UObject(other), fCreationStatus(other.fCreationStatus), fIsPrimary(false) { }
+   virtual ~CacheKeyBase();
+
+   /**
+    * Returns the hash code for this object.
+    */
+   virtual int32_t hashCode() const = 0;
+
+   /**
+    * Clones this object polymorphically. Caller owns returned value.
+    */
+   virtual CacheKeyBase *clone() const = 0;
+
+   /**
+    * Equality operator.
+    */
+   virtual UBool operator == (const CacheKeyBase &other) const = 0;
+
+   /**
+    * Create a new object for this key. Called by cache on cache miss.
+    * createObject must add a reference to the object it returns. Note
+    * that getting an object from the cache and returning it without calling
+    * removeRef on it satisfies this requirement. It can also return NULL
+    * and set status to an error.
+    *
+    * @param creationContext the context in which the object is being
+    *                        created. May be NULL.
+    * @param status          Implementations can return a failure here.
+    *                        In addition, implementations may return a
+    *                        non NULL object and set a warning status.
+    */
+   virtual const SharedObject *createObject(
+           const void *creationContext, UErrorCode &status) const = 0;
+
+   /**
+    * Writes a description of this key to buffer and returns buffer. Written
+    * description is NULL terminated.
+    */
+   virtual char *writeDescription(char *buffer, int32_t bufSize) const = 0;
+
+   /**
+    * Inequality operator.
+    */
+   UBool operator != (const CacheKeyBase &other) const {
+       return !(*this == other);
+   }
+ private:
+   mutable UErrorCode fCreationStatus;
+   mutable UBool fIsPrimary;
+   friend class UnifiedCache;
+};
+
+
+
+/**
+ * Templated version of CacheKeyBase. 
+ * A key of type LocaleCacheKey<T> maps to a value of type T.
+ */
+template<typename T>
+class CacheKey : public CacheKeyBase {
+ public:
+   virtual ~CacheKey() { }
+   /**
+    * The template parameter, T, determines the hash code returned.
+    */
+   virtual int32_t hashCode() const {
+       const char *s = typeid(T).name();
+       return ustr_hashCharsN(s, static_cast<int32_t>(uprv_strlen(s)));
+   }
+
+   /**
+    * Use the value type, T,  as the description.
+    */
+   virtual char *writeDescription(char *buffer, int32_t bufLen) const {
+       const char *s = typeid(T).name();
+       uprv_strncpy(buffer, s, bufLen);
+       buffer[bufLen - 1] = 0;
+       return buffer;
+   }
+
+   /**
+    * Two objects are equal if they are of the same type.
+    */
+   virtual UBool operator == (const CacheKeyBase &other) const {
+       return typeid(*this) == typeid(other);
+   }
+};
+
+/**
+ * Cache key based on locale.
+ * A key of type LocaleCacheKey<T> maps to a value of type T.
+ */
+template<typename T>
+class LocaleCacheKey : public CacheKey<T> {
+ protected:
+   Locale   fLoc;
+ public:
+   LocaleCacheKey(const Locale &loc) : fLoc(loc) {}
+   LocaleCacheKey(const LocaleCacheKey<T> &other)
+           : CacheKey<T>(other), fLoc(other.fLoc) { }
+   virtual ~LocaleCacheKey() { }
+   virtual int32_t hashCode() const {
+       return (int32_t)(37u * (uint32_t)CacheKey<T>::hashCode() + (uint32_t)fLoc.hashCode());
+   }
+   virtual UBool operator == (const CacheKeyBase &other) const {
+       // reflexive
+       if (this == &other) {
+           return true;
+       }
+       if (!CacheKey<T>::operator == (other)) {
+           return false;
+       }
+       // We know this and other are of same class because operator== on
+       // CacheKey returned true.
+       const LocaleCacheKey<T> *fOther =
+               static_cast<const LocaleCacheKey<T> *>(&other);
+       return fLoc == fOther->fLoc;
+   }
+   virtual CacheKeyBase *clone() const {
+       return new LocaleCacheKey<T>(*this);
+   }
+   virtual const T *createObject(
+           const void *creationContext, UErrorCode &status) const;
+   /**
+    * Use the locale id as the description.
+    */
+   virtual char *writeDescription(char *buffer, int32_t bufLen) const {
+       const char *s = fLoc.getName();
+       uprv_strncpy(buffer, s, bufLen);
+       buffer[bufLen - 1] = 0;
+       return buffer;
+   }
+
+};
+
+/**
+ * The unified cache. A singleton type.
+ * Design doc here:
+ * https://docs.google.com/document/d/1RwGQJs4N4tawNbf809iYDRCvXoMKqDJihxzYt1ysmd8/edit?usp=sharing
+ */
+class U_COMMON_API UnifiedCache : public UnifiedCacheBase {
+ public:
+   /**
+    * @internal
+    * Do not call directly. Instead use UnifiedCache::getInstance() as
+    * there should be only one UnifiedCache in an application.
+    */
+   UnifiedCache(UErrorCode &status);
+
+   /**
+    * Return a pointer to the global cache instance.
+    */
+   static UnifiedCache *getInstance(UErrorCode &status);
+
+   /**
+    * Fetches a value from the cache by key. Equivalent to
+    * get(key, NULL, ptr, status);
+    */
+   template<typename T>
+   void get(
+           const CacheKey<T>& key,
+           const T *&ptr,
+           UErrorCode &status) const {
+       get(key, NULL, ptr, status);
+   }
+
+   /**
+    * Fetches value from the cache by key.
+    *
+    * @param key             the cache key.
+    * @param creationContext passed verbatim to createObject method of key
+    * @param ptr             On entry, ptr must be NULL or be included if
+    *                        the reference count of the object it points
+    *                        to. On exit, ptr points to the fetched object
+    *                        from the cache or is left unchanged on
+    *                        failure. Caller must call removeRef on ptr
+    *                        if set to a non NULL value.
+    * @param status          Any error returned here. May be set to a
+    *                        warning value even if ptr is set.
+    */
+   template<typename T>
+   void get(
+           const CacheKey<T>& key,
+           const void *creationContext,
+           const T *&ptr,
+           UErrorCode &status) const {
+       if (U_FAILURE(status)) {
+           return;
+       }
+       UErrorCode creationStatus = U_ZERO_ERROR;
+       const SharedObject *value = NULL;
+       _get(key, value, creationContext, creationStatus);
+       const T *tvalue = (const T *) value;
+       if (U_SUCCESS(creationStatus)) {
+           SharedObject::copyPtr(tvalue, ptr);
+       }
+       SharedObject::clearPtr(tvalue);
+       // Take care not to overwrite a warning status passed in with
+       // another warning or U_ZERO_ERROR.
+       if (status == U_ZERO_ERROR || U_FAILURE(creationStatus)) {
+           status = creationStatus;
+       }
+   }
+
+#ifdef UNIFIED_CACHE_DEBUG
+   /**
+    * Dumps the contents of this cache to standard error. Used for testing of
+    * cache only.
+    */
+   void dumpContents() const;
+#endif
+
+   /**
+    * Convenience method to get a value of type T from cache for a
+    * particular locale with creationContext == NULL.
+    * @param loc    the locale
+    * @param ptr    On entry, must be NULL or included in the ref count
+    *               of the object to which it points.
+    *               On exit, fetched value stored here or is left
+    *               unchanged on failure. Caller must call removeRef on
+    *               ptr if set to a non NULL value.
+    * @param status Any error returned here. May be set to a
+    *               warning value even if ptr is set.
+    */
+   template<typename T>
+   static void getByLocale(
+           const Locale &loc, const T *&ptr, UErrorCode &status) {
+       const UnifiedCache *cache = getInstance(status);
+       if (U_FAILURE(status)) {
+           return;
+       }
+       cache->get(LocaleCacheKey<T>(loc), ptr, status);
+   }
+
+#ifdef UNIFIED_CACHE_DEBUG
+   /**
+    * Dumps the cache contents to stderr. For testing only.
+    */
+   static void dump();
+#endif
+
+   /**
+    * Returns the number of keys in this cache. For testing only.
+    */
+   int32_t keyCount() const;
+
+   /**
+    * Removes any values from cache that are not referenced outside
+    * the cache.
+    */
+   void flush() const;
+
+   /**
+    * Configures at what point evcition of unused entries will begin.
+    * Eviction is triggered whenever the number of evictable keys exeeds
+    * BOTH count AND (number of in-use items) * (percentageOfInUseItems / 100).
+    * Once the number of unused entries drops below one of these,
+    * eviction ceases. Because eviction happens incrementally,
+    * the actual unused entry count may exceed both these numbers
+    * from time to time.
+    *
+    * A cache entry is defined as unused if it is not essential to guarantee
+    * that for a given key X, the cache returns the same reference to the
+    * same value as long as the client already holds a reference to that
+    * value.
+    *
+    * If this method is never called, the default settings are 1000 and 100%.
+    *
+    * Although this method is thread-safe, it is designed to be called at
+    * application startup. If it is called in the middle of execution, it
+    * will have no immediate effect on the cache. However over time, the
+    * cache will perform eviction slices in an attempt to honor the new
+    * settings.
+    *
+    * If a client already holds references to many different unique values
+    * in the cache such that the number of those unique values far exeeds
+    * "count" then the cache may not be able to maintain this maximum.
+    * However, if this happens, the cache still guarantees that the number of
+    * unused entries will remain only a small percentage of the total cache
+    * size.
+    *
+    * If the parameters passed are negative, setEvctionPolicy sets status to
+    * U_ILLEGAL_ARGUMENT_ERROR.
+    */
+   void setEvictionPolicy(
+           int32_t count, int32_t percentageOfInUseItems, UErrorCode &status);
+
+
+   /**
+    * Returns how many entries have been auto evicted during the lifetime
+    * of this cache. This only includes auto evicted entries, not
+    * entries evicted because of a call to flush().
+    */
+   int64_t autoEvictedCount() const;
+
+   /**
+    * Returns the unused entry count in this cache. For testing only,
+    * Regular clients will not need this.
+    */
+   int32_t unusedCount() const;
+
+   virtual void handleUnreferencedObject() const;
+   virtual ~UnifiedCache();
+   
+ private:
+   UHashtable *fHashtable;
+   mutable int32_t fEvictPos;
+   mutable int32_t fNumValuesTotal;
+   mutable int32_t fNumValuesInUse;
+   int32_t fMaxUnused;
+   int32_t fMaxPercentageOfInUse;
+   mutable int64_t fAutoEvictedCount;
+   SharedObject *fNoValue;
+   
+   UnifiedCache(const UnifiedCache &other);
+   UnifiedCache &operator=(const UnifiedCache &other);
+   
+   /**
+    * Flushes the contents of the cache. If cache values hold references to other
+    * cache values then _flush should be called in a loop until it returns false.
+    * 
+    * On entry, gCacheMutex must be held.
+    * On exit, those values with are evictable are flushed.
+    * 
+    *  @param all if false flush evictable items only, which are those with no external
+    *                    references, plus those that can be safely recreated.<br>
+    *            if true, flush all elements. Any values (sharedObjects) with remaining
+    *                     hard (external) references are not deleted, but are detached from
+    *                     the cache, so that a subsequent removeRefs can delete them.
+    *                     _flush is not thread safe when all is true.
+    *   @return true if any value in cache was flushed or false otherwise.
+    */
+   UBool _flush(UBool all) const;
+   
+   /**
+    * Gets value out of cache.
+    * On entry. gCacheMutex must not be held. value must be NULL. status
+    * must be U_ZERO_ERROR.
+    * On exit. value and status set to what is in cache at key or on cache
+    * miss the key's createObject() is called and value and status are set to
+    * the result of that. In this latter case, best effort is made to add the
+    * value and status to the cache. If createObject() fails to create a value,
+    * fNoValue is stored in cache, and value is set to NULL. Caller must call
+    * removeRef on value if non NULL.
+    */
+   void _get(
+           const CacheKeyBase &key,
+           const SharedObject *&value,
+           const void *creationContext,
+           UErrorCode &status) const;
+
+    /**
+     * Attempts to fetch value and status for key from cache.
+     * On entry, gCacheMutex must not be held value must be NULL and status must
+     * be U_ZERO_ERROR.
+     * On exit, either returns false (In this
+     * case caller should try to create the object) or returns true with value
+     * pointing to the fetched value and status set to fetched status. When
+     * false is returned status may be set to failure if an in progress hash
+     * entry could not be made but value will remain unchanged. When true is
+     * returned, caller must call removeRef() on value.
+     */
+    UBool _poll(
+            const CacheKeyBase &key,
+            const SharedObject *&value,
+            UErrorCode &status) const;
+    
+    /**
+     * Places a new value and creationStatus in the cache for the given key.
+     * On entry, gCacheMutex must be held. key must not exist in the cache. 
+     * On exit, value and creation status placed under key. Soft reference added
+     * to value on successful add. On error sets status.
+     */
+    void _putNew(
+        const CacheKeyBase &key,
+        const SharedObject *value,
+        const UErrorCode creationStatus,
+        UErrorCode &status) const;
+           
+    /**
+     * Places value and status at key if there is no value at key or if cache
+     * entry for key is in progress. Otherwise, it leaves the current value and
+     * status there.
+     * 
+     * On entry. gCacheMutex must not be held. Value must be
+     * included in the reference count of the object to which it points.
+     * 
+     * On exit, value and status are changed to what was already in the cache if
+     * something was there and not in progress. Otherwise, value and status are left
+     * unchanged in which case they are placed in the cache on a best-effort basis.
+     * Caller must call removeRef() on value.
+     */
+   void _putIfAbsentAndGet(
+           const CacheKeyBase &key,
+           const SharedObject *&value,
+           UErrorCode &status) const;
+
+    /**
+     * Returns the next element in the cache round robin style.
+     * Returns nullptr if the cache is empty.
+     * On entry, gCacheMutex must be held.
+     */
+    const UHashElement *_nextElement() const;
+   
+   /**
+    * Return the number of cache items that would need to be evicted
+    * to bring usage into conformance with eviction policy.
+    * 
+    * An item corresponds to an entry in the hash table, a hash table element.
+    * 
+    * On entry, gCacheMutex must be held.
+    */
+   int32_t _computeCountOfItemsToEvict() const;
+   
+   /**
+    * Run an eviction slice.
+    * On entry, gCacheMutex must be held.
+    * _runEvictionSlice runs a slice of the evict pipeline by examining the next
+    * 10 entries in the cache round robin style evicting them if they are eligible.
+    */
+   void _runEvictionSlice() const;
+ 
+   /**
+    * Register a primary cache entry. A primary key is the first key to create
+    * a given  SharedObject value. Subsequent keys whose create function
+    * produce referneces to an already existing SharedObject are not primary -
+    * they can be evicted and subsequently recreated.
+    * 
+    * On entry, gCacheMutex must be held.
+    * On exit, items in use count incremented, entry is marked as a primary
+    * entry, and value registered with cache so that subsequent calls to
+    * addRef() and removeRef() on it correctly interact with the cache.
+    */
+   void _registerPrimary(const CacheKeyBase *theKey, const SharedObject *value) const;
+        
+   /**
+    * Store a value and creation error status in given hash entry.
+    * On entry, gCacheMutex must be held. Hash entry element must be in progress.
+    * value must be non NULL.
+    * On Exit, soft reference added to value. value and status stored in hash
+    * entry. Soft reference removed from previous stored value. Waiting
+    * threads notified.
+    */
+   void _put(
+           const UHashElement *element,
+           const SharedObject *value,
+           const UErrorCode status) const;
+    /**
+     * Remove a soft reference, and delete the SharedObject if no references remain.
+     * To be used from within the UnifiedCache implementation only.
+     * gCacheMutex must be held by caller.
+     * @param value the SharedObject to be acted on.
+     */
+   void removeSoftRef(const SharedObject *value) const;
+   
+   /**
+    * Increment the hard reference count of the given SharedObject.
+    * gCacheMutex must be held by the caller.
+    * Update numValuesEvictable on transitions between zero and one reference.
+    * 
+    * @param value The SharedObject to be referenced.
+    * @return the hard reference count after the addition.
+    */
+   int32_t addHardRef(const SharedObject *value) const;
+   
+  /**
+    * Decrement the hard reference count of the given SharedObject.
+    * gCacheMutex must be held by the caller.
+    * Update numValuesEvictable on transitions between one and zero reference.
+    * 
+    * @param value The SharedObject to be referenced.
+    * @return the hard reference count after the removal.
+    */
+   int32_t removeHardRef(const SharedObject *value) const;
+
+   
+#ifdef UNIFIED_CACHE_DEBUG
+   void _dumpContents() const;
+#endif
+   
+   /**
+    *  Fetch value and error code from a particular hash entry.
+    *  On entry, gCacheMutex must be held. value must be either NULL or must be
+    *  included in the ref count of the object to which it points.
+    *  On exit, value and status set to what is in the hash entry. Caller must
+    *  eventually call removeRef on value.
+    *  If hash entry is in progress, value will be set to gNoValue and status will
+    *  be set to U_ZERO_ERROR.
+    */
+   void _fetch(const UHashElement *element, const SharedObject *&value,
+                       UErrorCode &status) const;
+                       
+    /**
+     * Determine if given hash entry is in progress.
+     * On entry, gCacheMutex must be held.
+     */
+   UBool _inProgress(const UHashElement *element) const;
+   
+   /**
+    * Determine if given hash entry is in progress.
+    * On entry, gCacheMutex must be held.
+    */
+   UBool _inProgress(const SharedObject *theValue, UErrorCode creationStatus) const;
+   
+   /**
+    * Determine if given hash entry is eligible for eviction.
+    * On entry, gCacheMutex must be held.
+    */
+   UBool _isEvictable(const UHashElement *element) const;
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unifilt.cpp b/thirdparty/icu4c/common/unifilt.cpp
new file mode 100644
index 0000000000..4ab0d9b5f9
--- /dev/null
+++ b/thirdparty/icu4c/common/unifilt.cpp
@@ -0,0 +1,71 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2001-2012, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   07/18/01    aliu        Creation.
+**********************************************************************
+*/
+
+#include "unicode/unifilt.h"
+#include "unicode/rep.h"
+#include "unicode/utf16.h"
+
+U_NAMESPACE_BEGIN
+UOBJECT_DEFINE_ABSTRACT_RTTI_IMPLEMENTATION(UnicodeFilter)
+
+
+/* Define this here due to the lack of another file.
+   It can't be defined in the header */
+UnicodeMatcher::~UnicodeMatcher() {}
+
+UnicodeFilter::~UnicodeFilter() {}
+
+/**
+ * UnicodeFunctor API.
+ *   Note that UnicodeMatcher is a base class of UnicodeFilter.
+ */
+UnicodeMatcher* UnicodeFilter::toMatcher() const {
+  return const_cast<UnicodeFilter *>(this);
+}
+
+void UnicodeFilter::setData(const TransliterationRuleData*) {}
+
+/**
+ * Default implementation of UnicodeMatcher::matches() for Unicode
+ * filters.  Matches a single code point at offset (either one or
+ * two 16-bit code units).
+ */
+UMatchDegree UnicodeFilter::matches(const Replaceable& text,
+                                    int32_t& offset,
+                                    int32_t limit,
+                                    UBool incremental) {
+    UChar32 c;
+    if (offset < limit &&
+        contains(c = text.char32At(offset))) {
+        offset += U16_LENGTH(c);
+        return U_MATCH;
+    }
+    if (offset > limit &&
+        contains(c = text.char32At(offset))) {
+        // Backup offset by 1, unless the preceding character is a
+        // surrogate pair -- then backup by 2 (keep offset pointing at
+        // the lead surrogate).
+        --offset;
+        if (offset >= 0) {
+            offset -= U16_LENGTH(text.char32At(offset)) - 1;
+        }
+        return U_MATCH;
+    }
+    if (incremental && offset == limit) {
+        return U_PARTIAL_MATCH;
+    }
+    return U_MISMATCH;
+}
+
+U_NAMESPACE_END
+
+//eof
diff --git a/thirdparty/icu4c/common/unifunct.cpp b/thirdparty/icu4c/common/unifunct.cpp
new file mode 100644
index 0000000000..f3995b298d
--- /dev/null
+++ b/thirdparty/icu4c/common/unifunct.cpp
@@ -0,0 +1,28 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2004, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "unicode/unifunct.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_ABSTRACT_RTTI_IMPLEMENTATION(UnicodeFunctor)
+
+UnicodeFunctor::~UnicodeFunctor() {}
+
+UnicodeMatcher* UnicodeFunctor::toMatcher() const {
+    return 0;
+}
+
+UnicodeReplacer* UnicodeFunctor::toReplacer() const {
+    return 0;
+}
+
+U_NAMESPACE_END
+
+//eof
diff --git a/thirdparty/icu4c/common/uniquecharstr.h b/thirdparty/icu4c/common/uniquecharstr.h
new file mode 100644
index 0000000000..10cc924f7f
--- /dev/null
+++ b/thirdparty/icu4c/common/uniquecharstr.h
@@ -0,0 +1,98 @@
+// © 2020 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// uniquecharstr.h
+// created: 2020sep01 Frank Yung-Fong Tang
+
+#ifndef __UNIQUECHARSTR_H__
+#define __UNIQUECHARSTR_H__
+
+#include "charstr.h"
+#include "uassert.h"
+#include "uhash.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Stores NUL-terminated strings with duplicate elimination.
+ * Checks for unique UTF-16 string pointers and converts to invariant characters.
+ *
+ * Intended to be stack-allocated. Add strings, get a unique number for each,
+ * freeze the object, get a char * pointer for each string,
+ * call orphanCharStrings() to capture the string storage, and let this object go out of scope.
+ */
+class UniqueCharStrings {
+public:
+    UniqueCharStrings(UErrorCode &errorCode) : strings(nullptr) {
+        // Note: We hash on string contents but store stable char16_t * pointers.
+        // If the strings are stored in resource bundles which should be built with
+        // duplicate elimination, then we should be able to hash on just the pointer values.
+        uhash_init(&map, uhash_hashUChars, uhash_compareUChars, uhash_compareLong, &errorCode);
+        if (U_FAILURE(errorCode)) { return; }
+        strings = new CharString();
+        if (strings == nullptr) {
+            errorCode = U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+    ~UniqueCharStrings() {
+        uhash_close(&map);
+        delete strings;
+    }
+
+    /** Returns/orphans the CharString that contains all strings. */
+    CharString *orphanCharStrings() {
+        CharString *result = strings;
+        strings = nullptr;
+        return result;
+    }
+
+    /**
+     * Adds a string and returns a unique number for it.
+     * The string's buffer contents must not change, nor move around in memory,
+     * while this UniqueCharStrings is in use.
+     * The string contents must be NUL-terminated exactly at s.length().
+     *
+     * Best used with read-only-alias UnicodeString objects that point to
+     * stable storage, such as strings returned by resource bundle functions.
+     */
+    int32_t add(const UnicodeString &s, UErrorCode &errorCode) {
+        if (U_FAILURE(errorCode)) { return 0; }
+        if (isFrozen) {
+            errorCode = U_NO_WRITE_PERMISSION;
+            return 0;
+        }
+        // The string points into the resource bundle.
+        const char16_t *p = s.getBuffer();
+        int32_t oldIndex = uhash_geti(&map, p);
+        if (oldIndex != 0) {  // found duplicate
+            return oldIndex;
+        }
+        // Explicit NUL terminator for the previous string.
+        // The strings object is also terminated with one implicit NUL.
+        strings->append(0, errorCode);
+        int32_t newIndex = strings->length();
+        strings->appendInvariantChars(s, errorCode);
+        uhash_puti(&map, const_cast<char16_t *>(p), newIndex, &errorCode);
+        return newIndex;
+    }
+
+    void freeze() { isFrozen = true; }
+
+    /**
+     * Returns a string pointer for its unique number, if this object is frozen.
+     * Otherwise nullptr.
+     */
+    const char *get(int32_t i) const {
+        U_ASSERT(isFrozen);
+        return isFrozen && i > 0 ? strings->data() + i : nullptr;
+    }
+
+private:
+    UHashtable map;
+    CharString *strings;
+    bool isFrozen = false;
+};
+
+U_NAMESPACE_END
+
+#endif  // __UNIQUECHARSTR_H__
diff --git a/thirdparty/icu4c/common/uniset.cpp b/thirdparty/icu4c/common/uniset.cpp
new file mode 100644
index 0000000000..b73d612f24
--- /dev/null
+++ b/thirdparty/icu4c/common/uniset.cpp
@@ -0,0 +1,2356 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   10/20/99    alan        Creation.
+**********************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/parsepos.h"
+#include "unicode/symtable.h"
+#include "unicode/uniset.h"
+#include "unicode/ustring.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "ruleiter.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "patternprops.h"
+#include "uelement.h"
+#include "util.h"
+#include "uvector.h"
+#include "charstr.h"
+#include "ustrfmt.h"
+#include "uassert.h"
+#include "bmpset.h"
+#include "unisetspan.h"
+
+// Define UChar constants using hex for EBCDIC compatibility
+// Used #define to reduce private static exports and memory access time.
+#define SET_OPEN        ((UChar)0x005B) /*[*/
+#define SET_CLOSE       ((UChar)0x005D) /*]*/
+#define HYPHEN          ((UChar)0x002D) /*-*/
+#define COMPLEMENT      ((UChar)0x005E) /*^*/
+#define COLON           ((UChar)0x003A) /*:*/
+#define BACKSLASH       ((UChar)0x005C) /*\*/
+#define INTERSECTION    ((UChar)0x0026) /*&*/
+#define UPPER_U         ((UChar)0x0055) /*U*/
+#define LOWER_U         ((UChar)0x0075) /*u*/
+#define OPEN_BRACE      ((UChar)123)    /*{*/
+#define CLOSE_BRACE     ((UChar)125)    /*}*/
+#define UPPER_P         ((UChar)0x0050) /*P*/
+#define LOWER_P         ((UChar)0x0070) /*p*/
+#define UPPER_N         ((UChar)78)     /*N*/
+#define EQUALS          ((UChar)0x003D) /*=*/
+
+// HIGH_VALUE > all valid values. 110000 for codepoints
+#define UNICODESET_HIGH 0x0110000
+
+// LOW <= all valid values. ZERO for codepoints
+#define UNICODESET_LOW 0x000000
+
+/** Max list [0, 1, 2, ..., max code point, HIGH] */
+constexpr int32_t MAX_LENGTH = UNICODESET_HIGH + 1;
+
+U_NAMESPACE_BEGIN
+
+SymbolTable::~SymbolTable() {}
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UnicodeSet)
+
+/**
+ * Modify the given UChar32 variable so that it is in range, by
+ * pinning values < UNICODESET_LOW to UNICODESET_LOW, and
+ * pinning values > UNICODESET_HIGH-1 to UNICODESET_HIGH-1.
+ * It modifies its argument in-place and also returns it.
+ */
+static inline UChar32 pinCodePoint(UChar32& c) {
+    if (c < UNICODESET_LOW) {
+        c = UNICODESET_LOW;
+    } else if (c > (UNICODESET_HIGH-1)) {
+        c = (UNICODESET_HIGH-1);
+    }
+    return c;
+}
+
+//----------------------------------------------------------------
+// Debugging
+//----------------------------------------------------------------
+
+// DO NOT DELETE THIS CODE.  This code is used to debug memory leaks.
+// To enable the debugging, define the symbol DEBUG_MEM in the line
+// below.  This will result in text being sent to stdout that looks
+// like this:
+//   DEBUG UnicodeSet: ct 0x00A39B20; 397 [\u0A81-\u0A83\u0A85-
+//   DEBUG UnicodeSet: dt 0x00A39B20; 396 [\u0A81-\u0A83\u0A85-
+// Each line lists a construction (ct) or destruction (dt) event, the
+// object address, the number of outstanding objects after the event,
+// and the pattern of the object in question.
+
+// #define DEBUG_MEM
+
+#ifdef DEBUG_MEM
+#include <stdio.h>
+static int32_t _dbgCount = 0;
+
+static inline void _dbgct(UnicodeSet* set) {
+    UnicodeString str;
+    set->toPattern(str, TRUE);
+    char buf[40];
+    str.extract(0, 39, buf, "");
+    printf("DEBUG UnicodeSet: ct 0x%08X; %d %s\n", set, ++_dbgCount, buf);
+}
+
+static inline void _dbgdt(UnicodeSet* set) {
+    UnicodeString str;
+    set->toPattern(str, TRUE);
+    char buf[40];
+    str.extract(0, 39, buf, "");
+    printf("DEBUG UnicodeSet: dt 0x%08X; %d %s\n", set, --_dbgCount, buf);
+}
+
+#else
+
+#define _dbgct(set)
+#define _dbgdt(set)
+
+#endif
+
+//----------------------------------------------------------------
+// UnicodeString in UVector support
+//----------------------------------------------------------------
+
+static void U_CALLCONV cloneUnicodeString(UElement *dst, UElement *src) {
+    dst->pointer = new UnicodeString(*(UnicodeString*)src->pointer);
+}
+
+static int8_t U_CALLCONV compareUnicodeString(UElement t1, UElement t2) {
+    const UnicodeString &a = *(const UnicodeString*)t1.pointer;
+    const UnicodeString &b = *(const UnicodeString*)t2.pointer;
+    return a.compare(b);
+}
+
+UBool UnicodeSet::hasStrings() const {
+    return strings != nullptr && !strings->isEmpty();
+}
+
+int32_t UnicodeSet::stringsSize() const {
+    return strings == nullptr ? 0 : strings->size();
+}
+
+UBool UnicodeSet::stringsContains(const UnicodeString &s) const {
+    return strings != nullptr && strings->contains((void*) &s);
+}
+
+//----------------------------------------------------------------
+// Constructors &c
+//----------------------------------------------------------------
+
+/**
+ * Constructs an empty set.
+ */
+UnicodeSet::UnicodeSet() {
+    list[0] = UNICODESET_HIGH;
+    _dbgct(this);
+}
+
+/**
+ * Constructs a set containing the given range. If <code>end >
+ * start</code> then an empty set is created.
+ *
+ * @param start first character, inclusive, of range
+ * @param end last character, inclusive, of range
+ */
+UnicodeSet::UnicodeSet(UChar32 start, UChar32 end) {
+    list[0] = UNICODESET_HIGH;
+    add(start, end);
+    _dbgct(this);
+}
+
+/**
+ * Constructs a set that is identical to the given UnicodeSet.
+ */
+UnicodeSet::UnicodeSet(const UnicodeSet& o) : UnicodeFilter(o) {
+    *this = o;
+    _dbgct(this);
+}
+
+// Copy-construct as thawed.
+UnicodeSet::UnicodeSet(const UnicodeSet& o, UBool /* asThawed */) : UnicodeFilter(o) {
+    if (ensureCapacity(o.len)) {
+        // *this = o except for bmpSet and stringSpan
+        len = o.len;
+        uprv_memcpy(list, o.list, (size_t)len*sizeof(UChar32));
+        if (o.hasStrings()) {
+            UErrorCode status = U_ZERO_ERROR;
+            if (!allocateStrings(status) ||
+                    (strings->assign(*o.strings, cloneUnicodeString, status), U_FAILURE(status))) {
+                setToBogus();
+                return;
+            }
+        }
+        if (o.pat) {
+            setPattern(o.pat, o.patLen);
+        }
+        _dbgct(this);
+    }
+}
+
+/**
+ * Destructs the set.
+ */
+UnicodeSet::~UnicodeSet() {
+    _dbgdt(this); // first!
+    if (list != stackList) {
+        uprv_free(list);
+    }
+    delete bmpSet;
+    if (buffer != stackList) {
+        uprv_free(buffer);
+    }
+    delete strings;
+    delete stringSpan;
+    releasePattern();
+}
+
+/**
+ * Assigns this object to be a copy of another.
+ */
+UnicodeSet& UnicodeSet::operator=(const UnicodeSet& o) {
+    return copyFrom(o, FALSE);
+}
+
+UnicodeSet& UnicodeSet::copyFrom(const UnicodeSet& o, UBool asThawed) {
+    if (this == &o) {
+        return *this;
+    }
+    if (isFrozen()) {
+        return *this;
+    }
+    if (o.isBogus()) {
+        setToBogus();
+        return *this;
+    }
+    if (!ensureCapacity(o.len)) {
+        // ensureCapacity will mark the UnicodeSet as Bogus if OOM failure happens.
+        return *this;
+    }
+    len = o.len;
+    uprv_memcpy(list, o.list, (size_t)len*sizeof(UChar32));
+    if (o.bmpSet != nullptr && !asThawed) {
+        bmpSet = new BMPSet(*o.bmpSet, list, len);
+        if (bmpSet == NULL) { // Check for memory allocation error.
+            setToBogus();
+            return *this;
+        }
+    }
+    if (o.hasStrings()) {
+        UErrorCode status = U_ZERO_ERROR;
+        if ((strings == nullptr && !allocateStrings(status)) ||
+                (strings->assign(*o.strings, cloneUnicodeString, status), U_FAILURE(status))) {
+            setToBogus();
+            return *this;
+        }
+    } else if (hasStrings()) {
+        strings->removeAllElements();
+    }
+    if (o.stringSpan != nullptr && !asThawed) {
+        stringSpan = new UnicodeSetStringSpan(*o.stringSpan, *strings);
+        if (stringSpan == NULL) { // Check for memory allocation error.
+            setToBogus();
+            return *this;
+        }
+    }
+    releasePattern();
+    if (o.pat) {
+        setPattern(o.pat, o.patLen);
+    }
+    return *this;
+}
+
+/**
+ * Returns a copy of this object.  All UnicodeMatcher objects have
+ * to support cloning in order to allow classes using
+ * UnicodeMatchers, such as Transliterator, to implement cloning.
+ */
+UnicodeSet* UnicodeSet::clone() const {
+    return new UnicodeSet(*this);
+}
+
+UnicodeSet *UnicodeSet::cloneAsThawed() const {
+    return new UnicodeSet(*this, TRUE);
+}
+
+/**
+ * Compares the specified object with this set for equality.  Returns
+ * <tt>true</tt> if the two sets
+ * have the same size, and every member of the specified set is
+ * contained in this set (or equivalently, every member of this set is
+ * contained in the specified set).
+ *
+ * @param o set to be compared for equality with this set.
+ * @return <tt>true</tt> if the specified set is equal to this set.
+ */
+UBool UnicodeSet::operator==(const UnicodeSet& o) const {
+    if (len != o.len) return FALSE;
+    for (int32_t i = 0; i < len; ++i) {
+        if (list[i] != o.list[i]) return FALSE;
+    }
+    if (hasStrings() != o.hasStrings()) { return FALSE; }
+    if (hasStrings() && *strings != *o.strings) return FALSE;
+    return TRUE;
+}
+
+/**
+ * Returns the hash code value for this set.
+ *
+ * @return the hash code value for this set.
+ * @see Object#hashCode()
+ */
+int32_t UnicodeSet::hashCode(void) const {
+    uint32_t result = static_cast<uint32_t>(len);
+    for (int32_t i = 0; i < len; ++i) {
+        result *= 1000003u;
+        result += list[i];
+    }
+    return static_cast<int32_t>(result);
+}
+
+//----------------------------------------------------------------
+// Public API
+//----------------------------------------------------------------
+
+/**
+ * Returns the number of elements in this set (its cardinality),
+ * Note than the elements of a set may include both individual
+ * codepoints and strings.
+ *
+ * @return the number of elements in this set (its cardinality).
+ */
+int32_t UnicodeSet::size(void) const {
+    int32_t n = 0;
+    int32_t count = getRangeCount();
+    for (int32_t i = 0; i < count; ++i) {
+        n += getRangeEnd(i) - getRangeStart(i) + 1;
+    }
+    return n + stringsSize();
+}
+
+/**
+ * Returns <tt>true</tt> if this set contains no elements.
+ *
+ * @return <tt>true</tt> if this set contains no elements.
+ */
+UBool UnicodeSet::isEmpty(void) const {
+    return len == 1 && !hasStrings();
+}
+
+/**
+ * Returns true if this set contains the given character.
+ * @param c character to be checked for containment
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::contains(UChar32 c) const {
+    // Set i to the index of the start item greater than ch
+    // We know we will terminate without length test!
+    // LATER: for large sets, add binary search
+    //int32_t i = -1;
+    //for (;;) {
+    //    if (c < list[++i]) break;
+    //}
+    if (bmpSet != NULL) {
+        return bmpSet->contains(c);
+    }
+    if (stringSpan != NULL) {
+        return stringSpan->contains(c);
+    }
+    if (c >= UNICODESET_HIGH) { // Don't need to check LOW bound
+        return FALSE;
+    }
+    int32_t i = findCodePoint(c);
+    return (UBool)(i & 1); // return true if odd
+}
+
+/**
+ * Returns the smallest value i such that c < list[i].  Caller
+ * must ensure that c is a legal value or this method will enter
+ * an infinite loop.  This method performs a binary search.
+ * @param c a character in the range MIN_VALUE..MAX_VALUE
+ * inclusive
+ * @return the smallest integer i in the range 0..len-1,
+ * inclusive, such that c < list[i]
+ */
+int32_t UnicodeSet::findCodePoint(UChar32 c) const {
+    /* Examples:
+                                       findCodePoint(c)
+       set              list[]         c=0 1 3 4 7 8
+       ===              ==============   ===========
+       []               [110000]         0 0 0 0 0 0
+       [\u0000-\u0003]  [0, 4, 110000]   1 1 1 2 2 2
+       [\u0004-\u0007]  [4, 8, 110000]   0 0 0 1 1 2
+       [:Any:]          [0, 110000]      1 1 1 1 1 1
+     */
+
+    // Return the smallest i such that c < list[i].  Assume
+    // list[len - 1] == HIGH and that c is legal (0..HIGH-1).
+    if (c < list[0])
+        return 0;
+    // High runner test.  c is often after the last range, so an
+    // initial check for this condition pays off.
+    int32_t lo = 0;
+    int32_t hi = len - 1;
+    if (lo >= hi || c >= list[hi-1])
+        return hi;
+    // invariant: c >= list[lo]
+    // invariant: c < list[hi]
+    for (;;) {
+        int32_t i = (lo + hi) >> 1;
+        if (i == lo) {
+            break; // Found!
+        } else if (c < list[i]) {
+            hi = i;
+        } else {
+            lo = i;
+        }
+    }
+    return hi;
+}
+
+/**
+ * Returns true if this set contains every character
+ * of the given range.
+ * @param start first character, inclusive, of the range
+ * @param end last character, inclusive, of the range
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::contains(UChar32 start, UChar32 end) const {
+    //int32_t i = -1;
+    //for (;;) {
+    //    if (start < list[++i]) break;
+    //}
+    int32_t i = findCodePoint(start);
+    return ((i & 1) != 0 && end < list[i]);
+}
+
+/**
+ * Returns <tt>true</tt> if this set contains the given
+ * multicharacter string.
+ * @param s string to be checked for containment
+ * @return <tt>true</tt> if this set contains the specified string
+ */
+UBool UnicodeSet::contains(const UnicodeString& s) const {
+    if (s.length() == 0) return FALSE;
+    int32_t cp = getSingleCP(s);
+    if (cp < 0) {
+        return stringsContains(s);
+    } else {
+        return contains((UChar32) cp);
+    }
+}
+
+/**
+ * Returns true if this set contains all the characters and strings
+ * of the given set.
+ * @param c set to be checked for containment
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::containsAll(const UnicodeSet& c) const {
+    // The specified set is a subset if all of its pairs are contained in
+    // this set.  It's possible to code this more efficiently in terms of
+    // direct manipulation of the inversion lists if the need arises.
+    int32_t n = c.getRangeCount();
+    for (int i=0; i<n; ++i) {
+        if (!contains(c.getRangeStart(i), c.getRangeEnd(i))) {
+            return FALSE;
+        }
+    }
+    return !c.hasStrings() || (strings != nullptr && strings->containsAll(*c.strings));
+}
+
+/**
+ * Returns true if this set contains all the characters
+ * of the given string.
+ * @param s string containing characters to be checked for containment
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::containsAll(const UnicodeString& s) const {
+    return (UBool)(span(s.getBuffer(), s.length(), USET_SPAN_CONTAINED) ==
+                   s.length());
+}
+
+/**
+ * Returns true if this set contains none of the characters
+ * of the given range.
+ * @param start first character, inclusive, of the range
+ * @param end last character, inclusive, of the range
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::containsNone(UChar32 start, UChar32 end) const {
+    //int32_t i = -1;
+    //for (;;) {
+    //    if (start < list[++i]) break;
+    //}
+    int32_t i = findCodePoint(start);
+    return ((i & 1) == 0 && end < list[i]);
+}
+
+/**
+ * Returns true if this set contains none of the characters and strings
+ * of the given set.
+ * @param c set to be checked for containment
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::containsNone(const UnicodeSet& c) const {
+    // The specified set is a subset if all of its pairs are contained in
+    // this set.  It's possible to code this more efficiently in terms of
+    // direct manipulation of the inversion lists if the need arises.
+    int32_t n = c.getRangeCount();
+    for (int32_t i=0; i<n; ++i) {
+        if (!containsNone(c.getRangeStart(i), c.getRangeEnd(i))) {
+            return FALSE;
+        }
+    }
+    return strings == nullptr || !c.hasStrings() || strings->containsNone(*c.strings);
+}
+
+/**
+ * Returns true if this set contains none of the characters
+ * of the given string.
+ * @param s string containing characters to be checked for containment
+ * @return true if the test condition is met
+ */
+UBool UnicodeSet::containsNone(const UnicodeString& s) const {
+    return (UBool)(span(s.getBuffer(), s.length(), USET_SPAN_NOT_CONTAINED) ==
+                   s.length());
+}
+
+/**
+ * Returns <tt>true</tt> if this set contains any character whose low byte
+ * is the given value.  This is used by <tt>RuleBasedTransliterator</tt> for
+ * indexing.
+ */
+UBool UnicodeSet::matchesIndexValue(uint8_t v) const {
+    /* The index value v, in the range [0,255], is contained in this set if
+     * it is contained in any pair of this set.  Pairs either have the high
+     * bytes equal, or unequal.  If the high bytes are equal, then we have
+     * aaxx..aayy, where aa is the high byte.  Then v is contained if xx <=
+     * v <= yy.  If the high bytes are unequal we have aaxx..bbyy, bb>aa.
+     * Then v is contained if xx <= v || v <= yy.  (This is identical to the
+     * time zone month containment logic.)
+     */
+    int32_t i;
+    int32_t rangeCount=getRangeCount();
+    for (i=0; i<rangeCount; ++i) {
+        UChar32 low = getRangeStart(i);
+        UChar32 high = getRangeEnd(i);
+        if ((low & ~0xFF) == (high & ~0xFF)) {
+            if ((low & 0xFF) <= v && v <= (high & 0xFF)) {
+                return TRUE;
+            }
+        } else if ((low & 0xFF) <= v || v <= (high & 0xFF)) {
+            return TRUE;
+        }
+    }
+    if (hasStrings()) {
+        for (i=0; i<strings->size(); ++i) {
+            const UnicodeString& s = *(const UnicodeString*)strings->elementAt(i);
+            //if (s.length() == 0) {
+            //    // Empty strings match everything
+            //    return TRUE;
+            //}
+            // assert(s.length() != 0); // We enforce this elsewhere
+            UChar32 c = s.char32At(0);
+            if ((c & 0xFF) == v) {
+                return TRUE;
+            }
+        }
+    }
+    return FALSE;
+}
+
+/**
+ * Implementation of UnicodeMatcher::matches().  Always matches the
+ * longest possible multichar string.
+ */
+UMatchDegree UnicodeSet::matches(const Replaceable& text,
+                                 int32_t& offset,
+                                 int32_t limit,
+                                 UBool incremental) {
+    if (offset == limit) {
+        // Strings, if any, have length != 0, so we don't worry
+        // about them here.  If we ever allow zero-length strings
+        // we much check for them here.
+        if (contains(U_ETHER)) {
+            return incremental ? U_PARTIAL_MATCH : U_MATCH;
+        } else {
+            return U_MISMATCH;
+        }
+    } else {
+        if (hasStrings()) { // try strings first
+
+            // might separate forward and backward loops later
+            // for now they are combined
+
+            // TODO Improve efficiency of this, at least in the forward
+            // direction, if not in both.  In the forward direction we
+            // can assume the strings are sorted.
+
+            int32_t i;
+            UBool forward = offset < limit;
+
+            // firstChar is the leftmost char to match in the
+            // forward direction or the rightmost char to match in
+            // the reverse direction.
+            UChar firstChar = text.charAt(offset);
+
+            // If there are multiple strings that can match we
+            // return the longest match.
+            int32_t highWaterLength = 0;
+
+            for (i=0; i<strings->size(); ++i) {
+                const UnicodeString& trial = *(const UnicodeString*)strings->elementAt(i);
+
+                //if (trial.length() == 0) {
+                //    return U_MATCH; // null-string always matches
+                //}
+                // assert(trial.length() != 0); // We ensure this elsewhere
+
+                UChar c = trial.charAt(forward ? 0 : trial.length() - 1);
+
+                // Strings are sorted, so we can optimize in the
+                // forward direction.
+                if (forward && c > firstChar) break;
+                if (c != firstChar) continue;
+
+                int32_t matchLen = matchRest(text, offset, limit, trial);
+
+                if (incremental) {
+                    int32_t maxLen = forward ? limit-offset : offset-limit;
+                    if (matchLen == maxLen) {
+                        // We have successfully matched but only up to limit.
+                        return U_PARTIAL_MATCH;
+                    }
+                }
+
+                if (matchLen == trial.length()) {
+                    // We have successfully matched the whole string.
+                    if (matchLen > highWaterLength) {
+                        highWaterLength = matchLen;
+                    }
+                    // In the forward direction we know strings
+                    // are sorted so we can bail early.
+                    if (forward && matchLen < highWaterLength) {
+                        break;
+                    }
+                    continue;
+                }
+            }
+
+            // We've checked all strings without a partial match.
+            // If we have full matches, return the longest one.
+            if (highWaterLength != 0) {
+                offset += forward ? highWaterLength : -highWaterLength;
+                return U_MATCH;
+            }
+        }
+        return UnicodeFilter::matches(text, offset, limit, incremental);
+    }
+}
+
+/**
+ * Returns the longest match for s in text at the given position.
+ * If limit > start then match forward from start+1 to limit
+ * matching all characters except s.charAt(0).  If limit < start,
+ * go backward starting from start-1 matching all characters
+ * except s.charAt(s.length()-1).  This method assumes that the
+ * first character, text.charAt(start), matches s, so it does not
+ * check it.
+ * @param text the text to match
+ * @param start the first character to match.  In the forward
+ * direction, text.charAt(start) is matched against s.charAt(0).
+ * In the reverse direction, it is matched against
+ * s.charAt(s.length()-1).
+ * @param limit the limit offset for matching, either last+1 in
+ * the forward direction, or last-1 in the reverse direction,
+ * where last is the index of the last character to match.
+ * @return If part of s matches up to the limit, return |limit -
+ * start|.  If all of s matches before reaching the limit, return
+ * s.length().  If there is a mismatch between s and text, return
+ * 0
+ */
+int32_t UnicodeSet::matchRest(const Replaceable& text,
+                              int32_t start, int32_t limit,
+                              const UnicodeString& s) {
+    int32_t i;
+    int32_t maxLen;
+    int32_t slen = s.length();
+    if (start < limit) {
+        maxLen = limit - start;
+        if (maxLen > slen) maxLen = slen;
+        for (i = 1; i < maxLen; ++i) {
+            if (text.charAt(start + i) != s.charAt(i)) return 0;
+        }
+    } else {
+        maxLen = start - limit;
+        if (maxLen > slen) maxLen = slen;
+        --slen; // <=> slen = s.length() - 1;
+        for (i = 1; i < maxLen; ++i) {
+            if (text.charAt(start - i) != s.charAt(slen - i)) return 0;
+        }
+    }
+    return maxLen;
+}
+
+/**
+ * Implement of UnicodeMatcher
+ */
+void UnicodeSet::addMatchSetTo(UnicodeSet& toUnionTo) const {
+    toUnionTo.addAll(*this);
+}
+
+/**
+ * Returns the index of the given character within this set, where
+ * the set is ordered by ascending code point.  If the character
+ * is not in this set, return -1.  The inverse of this method is
+ * <code>charAt()</code>.
+ * @return an index from 0..size()-1, or -1
+ */
+int32_t UnicodeSet::indexOf(UChar32 c) const {
+    if (c < MIN_VALUE || c > MAX_VALUE) {
+        return -1;
+    }
+    int32_t i = 0;
+    int32_t n = 0;
+    for (;;) {
+        UChar32 start = list[i++];
+        if (c < start) {
+            return -1;
+        }
+        UChar32 limit = list[i++];
+        if (c < limit) {
+            return n + c - start;
+        }
+        n += limit - start;
+    }
+}
+
+/**
+ * Returns the character at the given index within this set, where
+ * the set is ordered by ascending code point.  If the index is
+ * out of range, return (UChar32)-1.  The inverse of this method is
+ * <code>indexOf()</code>.
+ * @param index an index from 0..size()-1
+ * @return the character at the given index, or (UChar32)-1.
+ */
+UChar32 UnicodeSet::charAt(int32_t index) const {
+    if (index >= 0) {
+        // len2 is the largest even integer <= len, that is, it is len
+        // for even values and len-1 for odd values.  With odd values
+        // the last entry is UNICODESET_HIGH.
+        int32_t len2 = len & ~1;
+        for (int32_t i=0; i < len2;) {
+            UChar32 start = list[i++];
+            int32_t count = list[i++] - start;
+            if (index < count) {
+                return (UChar32)(start + index);
+            }
+            index -= count;
+        }
+    }
+    return (UChar32)-1;
+}
+
+/**
+ * Make this object represent the range <code>start - end</code>.
+ * If <code>end > start</code> then this object is set to an
+ * an empty range.
+ *
+ * @param start first character in the set, inclusive
+ * @rparam end last character in the set, inclusive
+ */
+UnicodeSet& UnicodeSet::set(UChar32 start, UChar32 end) {
+    clear();
+    complement(start, end);
+    return *this;
+}
+
+/**
+ * Adds the specified range to this set if it is not already
+ * present.  If this set already contains the specified range,
+ * the call leaves this set unchanged.  If <code>end > start</code>
+ * then an empty range is added, leaving the set unchanged.
+ *
+ * @param start first character, inclusive, of range to be added
+ * to this set.
+ * @param end last character, inclusive, of range to be added
+ * to this set.
+ */
+UnicodeSet& UnicodeSet::add(UChar32 start, UChar32 end) {
+    if (pinCodePoint(start) < pinCodePoint(end)) {
+        UChar32 limit = end + 1;
+        // Fast path for adding a new range after the last one.
+        // Odd list length: [..., lastStart, lastLimit, HIGH]
+        if ((len & 1) != 0) {
+            // If the list is empty, set lastLimit low enough to not be adjacent to 0.
+            UChar32 lastLimit = len == 1 ? -2 : list[len - 2];
+            if (lastLimit <= start && !isFrozen() && !isBogus()) {
+                if (lastLimit == start) {
+                    // Extend the last range.
+                    list[len - 2] = limit;
+                    if (limit == UNICODESET_HIGH) {
+                        --len;
+                    }
+                } else {
+                    list[len - 1] = start;
+                    if (limit < UNICODESET_HIGH) {
+                        if (ensureCapacity(len + 2)) {
+                            list[len++] = limit;
+                            list[len++] = UNICODESET_HIGH;
+                        }
+                    } else {  // limit == UNICODESET_HIGH
+                        if (ensureCapacity(len + 1)) {
+                            list[len++] = UNICODESET_HIGH;
+                        }
+                    }
+                }
+                releasePattern();
+                return *this;
+            }
+        }
+        // This is slow. Could be much faster using findCodePoint(start)
+        // and modifying the list, dealing with adjacent & overlapping ranges.
+        UChar32 range[3] = { start, limit, UNICODESET_HIGH };
+        add(range, 2, 0);
+    } else if (start == end) {
+        add(start);
+    }
+    return *this;
+}
+
+// #define DEBUG_US_ADD
+
+#ifdef DEBUG_US_ADD
+#include <stdio.h>
+void dump(UChar32 c) {
+    if (c <= 0xFF) {
+        printf("%c", (char)c);
+    } else {
+        printf("U+%04X", c);
+    }
+}
+void dump(const UChar32* list, int32_t len) {
+    printf("[");
+    for (int32_t i=0; i<len; ++i) {
+        if (i != 0) printf(", ");
+        dump(list[i]);
+    }
+    printf("]");
+}
+#endif
+
+/**
+ * Adds the specified character to this set if it is not already
+ * present.  If this set already contains the specified character,
+ * the call leaves this set unchanged.
+ */
+UnicodeSet& UnicodeSet::add(UChar32 c) {
+    // find smallest i such that c < list[i]
+    // if odd, then it is IN the set
+    // if even, then it is OUT of the set
+    int32_t i = findCodePoint(pinCodePoint(c));
+
+    // already in set?
+    if ((i & 1) != 0  || isFrozen() || isBogus()) return *this;
+
+    // HIGH is 0x110000
+    // assert(list[len-1] == HIGH);
+
+    // empty = [HIGH]
+    // [start_0, limit_0, start_1, limit_1, HIGH]
+
+    // [..., start_k-1, limit_k-1, start_k, limit_k, ..., HIGH]
+    //                             ^
+    //                             list[i]
+
+    // i == 0 means c is before the first range
+
+#ifdef DEBUG_US_ADD
+    printf("Add of ");
+    dump(c);
+    printf(" found at %d", i);
+    printf(": ");
+    dump(list, len);
+    printf(" => ");
+#endif
+
+    if (c == list[i]-1) {
+        // c is before start of next range
+        list[i] = c;
+        // if we touched the HIGH mark, then add a new one
+        if (c == (UNICODESET_HIGH - 1)) {
+            if (!ensureCapacity(len+1)) {
+                // ensureCapacity will mark the object as Bogus if OOM failure happens.
+                return *this;
+            }
+            list[len++] = UNICODESET_HIGH;
+        }
+        if (i > 0 && c == list[i-1]) {
+            // collapse adjacent ranges
+
+            // [..., start_k-1, c, c, limit_k, ..., HIGH]
+            //                     ^
+            //                     list[i]
+
+            //for (int32_t k=i-1; k<len-2; ++k) {
+            //    list[k] = list[k+2];
+            //}
+            UChar32* dst = list + i - 1;
+            UChar32* src = dst + 2;
+            UChar32* srclimit = list + len;
+            while (src < srclimit) *(dst++) = *(src++);
+
+            len -= 2;
+        }
+    }
+
+    else if (i > 0 && c == list[i-1]) {
+        // c is after end of prior range
+        list[i-1]++;
+        // no need to check for collapse here
+    }
+
+    else {
+        // At this point we know the new char is not adjacent to
+        // any existing ranges, and it is not 10FFFF.
+
+
+        // [..., start_k-1, limit_k-1, start_k, limit_k, ..., HIGH]
+        //                             ^
+        //                             list[i]
+
+        // [..., start_k-1, limit_k-1, c, c+1, start_k, limit_k, ..., HIGH]
+        //                             ^
+        //                             list[i]
+
+        if (!ensureCapacity(len+2)) {
+            // ensureCapacity will mark the object as Bogus if OOM failure happens.
+            return *this;
+        }
+
+        UChar32 *p = list + i;
+        uprv_memmove(p + 2, p, (len - i) * sizeof(*p));
+        list[i] = c;
+        list[i+1] = c+1;
+        len += 2;
+    }
+
+#ifdef DEBUG_US_ADD
+    dump(list, len);
+    printf("\n");
+
+    for (i=1; i<len; ++i) {
+        if (list[i] <= list[i-1]) {
+            // Corrupt array!
+            printf("ERROR: list has been corrupted\n");
+            exit(1);
+        }
+    }
+#endif
+
+    releasePattern();
+    return *this;
+}
+
+/**
+ * Adds the specified multicharacter to this set if it is not already
+ * present.  If this set already contains the multicharacter,
+ * the call leaves this set unchanged.
+ * Thus "ch" => {"ch"}
+ * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+ * @param s the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::add(const UnicodeString& s) {
+    if (s.length() == 0 || isFrozen() || isBogus()) return *this;
+    int32_t cp = getSingleCP(s);
+    if (cp < 0) {
+        if (!stringsContains(s)) {
+            _add(s);
+            releasePattern();
+        }
+    } else {
+        add((UChar32)cp);
+    }
+    return *this;
+}
+
+/**
+ * Adds the given string, in order, to 'strings'.  The given string
+ * must have been checked by the caller to not be empty and to not
+ * already be in 'strings'.
+ */
+void UnicodeSet::_add(const UnicodeString& s) {
+    if (isFrozen() || isBogus()) {
+        return;
+    }
+    UErrorCode ec = U_ZERO_ERROR;
+    if (strings == nullptr && !allocateStrings(ec)) {
+        setToBogus();
+        return;
+    }
+    UnicodeString* t = new UnicodeString(s);
+    if (t == NULL) { // Check for memory allocation error.
+        setToBogus();
+        return;
+    }
+    strings->sortedInsert(t, compareUnicodeString, ec);
+    if (U_FAILURE(ec)) {
+        setToBogus();
+        delete t;
+    }
+}
+
+/**
+ * @return a code point IF the string consists of a single one.
+ * otherwise returns -1.
+ * @param string to test
+ */
+int32_t UnicodeSet::getSingleCP(const UnicodeString& s) {
+    //if (s.length() < 1) {
+    //    throw new IllegalArgumentException("Can't use zero-length strings in UnicodeSet");
+    //}
+    if (s.length() > 2) return -1;
+    if (s.length() == 1) return s.charAt(0);
+
+    // at this point, len = 2
+    UChar32 cp = s.char32At(0);
+    if (cp > 0xFFFF) { // is surrogate pair
+        return cp;
+    }
+    return -1;
+}
+
+/**
+ * Adds each of the characters in this string to the set. Thus "ch" => {"c", "h"}
+ * If this set already any particular character, it has no effect on that character.
+ * @param the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::addAll(const UnicodeString& s) {
+    UChar32 cp;
+    for (int32_t i = 0; i < s.length(); i += U16_LENGTH(cp)) {
+        cp = s.char32At(i);
+        add(cp);
+    }
+    return *this;
+}
+
+/**
+ * Retains EACH of the characters in this string. Note: "ch" == {"c", "h"}
+ * If this set already any particular character, it has no effect on that character.
+ * @param the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::retainAll(const UnicodeString& s) {
+    UnicodeSet set;
+    set.addAll(s);
+    retainAll(set);
+    return *this;
+}
+
+/**
+ * Complement EACH of the characters in this string. Note: "ch" == {"c", "h"}
+ * If this set already any particular character, it has no effect on that character.
+ * @param the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::complementAll(const UnicodeString& s) {
+    UnicodeSet set;
+    set.addAll(s);
+    complementAll(set);
+    return *this;
+}
+
+/**
+ * Remove EACH of the characters in this string. Note: "ch" == {"c", "h"}
+ * If this set already any particular character, it has no effect on that character.
+ * @param the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::removeAll(const UnicodeString& s) {
+    UnicodeSet set;
+    set.addAll(s);
+    removeAll(set);
+    return *this;
+}
+
+UnicodeSet& UnicodeSet::removeAllStrings() {
+    if (!isFrozen() && hasStrings()) {
+        strings->removeAllElements();
+        releasePattern();
+    }
+    return *this;
+}
+
+
+/**
+ * Makes a set from a multicharacter string. Thus "ch" => {"ch"}
+ * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+ * @param the source string
+ * @return a newly created set containing the given string
+ */
+UnicodeSet* U_EXPORT2 UnicodeSet::createFrom(const UnicodeString& s) {
+    UnicodeSet *set = new UnicodeSet();
+    if (set != NULL) { // Check for memory allocation error.
+        set->add(s);
+    }
+    return set;
+}
+
+
+/**
+ * Makes a set from each of the characters in the string. Thus "ch" => {"c", "h"}
+ * @param the source string
+ * @return a newly created set containing the given characters
+ */
+UnicodeSet* U_EXPORT2 UnicodeSet::createFromAll(const UnicodeString& s) {
+    UnicodeSet *set = new UnicodeSet();
+    if (set != NULL) { // Check for memory allocation error.
+        set->addAll(s);
+    }
+    return set;
+}
+
+/**
+ * Retain only the elements in this set that are contained in the
+ * specified range.  If <code>end > start</code> then an empty range is
+ * retained, leaving the set empty.
+ *
+ * @param start first character, inclusive, of range to be retained
+ * to this set.
+ * @param end last character, inclusive, of range to be retained
+ * to this set.
+ */
+UnicodeSet& UnicodeSet::retain(UChar32 start, UChar32 end) {
+    if (pinCodePoint(start) <= pinCodePoint(end)) {
+        UChar32 range[3] = { start, end+1, UNICODESET_HIGH };
+        retain(range, 2, 0);
+    } else {
+        clear();
+    }
+    return *this;
+}
+
+UnicodeSet& UnicodeSet::retain(UChar32 c) {
+    return retain(c, c);
+}
+
+/**
+ * Removes the specified range from this set if it is present.
+ * The set will not contain the specified range once the call
+ * returns.  If <code>end > start</code> then an empty range is
+ * removed, leaving the set unchanged.
+ *
+ * @param start first character, inclusive, of range to be removed
+ * from this set.
+ * @param end last character, inclusive, of range to be removed
+ * from this set.
+ */
+UnicodeSet& UnicodeSet::remove(UChar32 start, UChar32 end) {
+    if (pinCodePoint(start) <= pinCodePoint(end)) {
+        UChar32 range[3] = { start, end+1, UNICODESET_HIGH };
+        retain(range, 2, 2);
+    }
+    return *this;
+}
+
+/**
+ * Removes the specified character from this set if it is present.
+ * The set will not contain the specified range once the call
+ * returns.
+ */
+UnicodeSet& UnicodeSet::remove(UChar32 c) {
+    return remove(c, c);
+}
+
+/**
+ * Removes the specified string from this set if it is present.
+ * The set will not contain the specified character once the call
+ * returns.
+ * @param the source string
+ * @return the modified set, for chaining
+ */
+UnicodeSet& UnicodeSet::remove(const UnicodeString& s) {
+    if (s.length() == 0 || isFrozen() || isBogus()) return *this;
+    int32_t cp = getSingleCP(s);
+    if (cp < 0) {
+        if (strings != nullptr && strings->removeElement((void*) &s)) {
+            releasePattern();
+        }
+    } else {
+        remove((UChar32)cp, (UChar32)cp);
+    }
+    return *this;
+}
+
+/**
+ * Complements the specified range in this set.  Any character in
+ * the range will be removed if it is in this set, or will be
+ * added if it is not in this set.  If <code>end > start</code>
+ * then an empty range is xor'ed, leaving the set unchanged.
+ *
+ * @param start first character, inclusive, of range to be removed
+ * from this set.
+ * @param end last character, inclusive, of range to be removed
+ * from this set.
+ */
+UnicodeSet& UnicodeSet::complement(UChar32 start, UChar32 end) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    if (pinCodePoint(start) <= pinCodePoint(end)) {
+        UChar32 range[3] = { start, end+1, UNICODESET_HIGH };
+        exclusiveOr(range, 2, 0);
+    }
+    releasePattern();
+    return *this;
+}
+
+UnicodeSet& UnicodeSet::complement(UChar32 c) {
+    return complement(c, c);
+}
+
+/**
+ * This is equivalent to
+ * <code>complement(MIN_VALUE, MAX_VALUE)</code>.
+ */
+UnicodeSet& UnicodeSet::complement(void) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    if (list[0] == UNICODESET_LOW) {
+        uprv_memmove(list, list + 1, (size_t)(len-1)*sizeof(UChar32));
+        --len;
+    } else {
+        if (!ensureCapacity(len+1)) {
+            return *this;
+        }
+        uprv_memmove(list + 1, list, (size_t)len*sizeof(UChar32));
+        list[0] = UNICODESET_LOW;
+        ++len;
+    }
+    releasePattern();
+    return *this;
+}
+
+/**
+ * Complement the specified string in this set.
+ * The set will not contain the specified string once the call
+ * returns.
+ * <br><b>Warning: you cannot add an empty string ("") to a UnicodeSet.</b>
+ * @param s the string to complement
+ * @return this object, for chaining
+ */
+UnicodeSet& UnicodeSet::complement(const UnicodeString& s) {
+    if (s.length() == 0 || isFrozen() || isBogus()) return *this;
+    int32_t cp = getSingleCP(s);
+    if (cp < 0) {
+        if (stringsContains(s)) {
+            strings->removeElement((void*) &s);
+        } else {
+            _add(s);
+        }
+        releasePattern();
+    } else {
+        complement((UChar32)cp, (UChar32)cp);
+    }
+    return *this;
+}
+
+/**
+ * Adds all of the elements in the specified set to this set if
+ * they're not already present.  This operation effectively
+ * modifies this set so that its value is the <i>union</i> of the two
+ * sets.  The behavior of this operation is unspecified if the specified
+ * collection is modified while the operation is in progress.
+ *
+ * @param c set whose elements are to be added to this set.
+ * @see #add(char, char)
+ */
+UnicodeSet& UnicodeSet::addAll(const UnicodeSet& c) {
+    if ( c.len>0 && c.list!=NULL ) {
+        add(c.list, c.len, 0);
+    }
+
+    // Add strings in order
+    if ( c.strings!=NULL ) {
+        for (int32_t i=0; i<c.strings->size(); ++i) {
+            const UnicodeString* s = (const UnicodeString*)c.strings->elementAt(i);
+            if (!stringsContains(*s)) {
+                _add(*s);
+            }
+        }
+    }
+    return *this;
+}
+
+/**
+ * Retains only the elements in this set that are contained in the
+ * specified set.  In other words, removes from this set all of
+ * its elements that are not contained in the specified set.  This
+ * operation effectively modifies this set so that its value is
+ * the <i>intersection</i> of the two sets.
+ *
+ * @param c set that defines which elements this set will retain.
+ */
+UnicodeSet& UnicodeSet::retainAll(const UnicodeSet& c) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    retain(c.list, c.len, 0);
+    if (hasStrings()) {
+        if (!c.hasStrings()) {
+            strings->removeAllElements();
+        } else {
+            strings->retainAll(*c.strings);
+        }
+    }
+    return *this;
+}
+
+/**
+ * Removes from this set all of its elements that are contained in the
+ * specified set.  This operation effectively modifies this
+ * set so that its value is the <i>asymmetric set difference</i> of
+ * the two sets.
+ *
+ * @param c set that defines which elements will be removed from
+ *          this set.
+ */
+UnicodeSet& UnicodeSet::removeAll(const UnicodeSet& c) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    retain(c.list, c.len, 2);
+    if (hasStrings() && c.hasStrings()) {
+        strings->removeAll(*c.strings);
+    }
+    return *this;
+}
+
+/**
+ * Complements in this set all elements contained in the specified
+ * set.  Any character in the other set will be removed if it is
+ * in this set, or will be added if it is not in this set.
+ *
+ * @param c set that defines which elements will be xor'ed from
+ *          this set.
+ */
+UnicodeSet& UnicodeSet::complementAll(const UnicodeSet& c) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    exclusiveOr(c.list, c.len, 0);
+
+    if (c.strings != nullptr) {
+        for (int32_t i=0; i<c.strings->size(); ++i) {
+            void* e = c.strings->elementAt(i);
+            if (strings == nullptr || !strings->removeElement(e)) {
+                _add(*(const UnicodeString*)e);
+            }
+        }
+    }
+    return *this;
+}
+
+/**
+ * Removes all of the elements from this set.  This set will be
+ * empty after this call returns.
+ */
+UnicodeSet& UnicodeSet::clear(void) {
+    if (isFrozen()) {
+        return *this;
+    }
+    list[0] = UNICODESET_HIGH;
+    len = 1;
+    releasePattern();
+    if (strings != NULL) {
+        strings->removeAllElements();
+    }
+    // Remove bogus
+    fFlags = 0;
+    return *this;
+}
+
+/**
+ * Iteration method that returns the number of ranges contained in
+ * this set.
+ * @see #getRangeStart
+ * @see #getRangeEnd
+ */
+int32_t UnicodeSet::getRangeCount() const {
+    return len/2;
+}
+
+/**
+ * Iteration method that returns the first character in the
+ * specified range of this set.
+ * @see #getRangeCount
+ * @see #getRangeEnd
+ */
+UChar32 UnicodeSet::getRangeStart(int32_t index) const {
+    return list[index*2];
+}
+
+/**
+ * Iteration method that returns the last character in the
+ * specified range of this set.
+ * @see #getRangeStart
+ * @see #getRangeEnd
+ */
+UChar32 UnicodeSet::getRangeEnd(int32_t index) const {
+    return list[index*2 + 1] - 1;
+}
+
+const UnicodeString* UnicodeSet::getString(int32_t index) const {
+    return (const UnicodeString*) strings->elementAt(index);
+}
+
+/**
+ * Reallocate this objects internal structures to take up the least
+ * possible space, without changing this object's value.
+ */
+UnicodeSet& UnicodeSet::compact() {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    // Delete buffer first to defragment memory less.
+    if (buffer != stackList) {
+        uprv_free(buffer);
+        buffer = NULL;
+        bufferCapacity = 0;
+    }
+    if (list == stackList) {
+        // pass
+    } else if (len <= INITIAL_CAPACITY) {
+        uprv_memcpy(stackList, list, len * sizeof(UChar32));
+        uprv_free(list);
+        list = stackList;
+        capacity = INITIAL_CAPACITY;
+    } else if ((len + 7) < capacity) {
+        // If we have more than a little unused capacity, shrink it to len.
+        UChar32* temp = (UChar32*) uprv_realloc(list, sizeof(UChar32) * len);
+        if (temp) {
+            list = temp;
+            capacity = len;
+        }
+        // else what the heck happened?! We allocated less memory!
+        // Oh well. We'll keep our original array.
+    }
+    if (strings != nullptr && strings->isEmpty()) {
+        delete strings;
+        strings = nullptr;
+    }
+    return *this;
+}
+
+#ifdef DEBUG_SERIALIZE
+#include <stdio.h>
+#endif
+
+/**
+ * Deserialize constructor.
+ */
+UnicodeSet::UnicodeSet(const uint16_t data[], int32_t dataLen, ESerialization serialization,
+                       UErrorCode &ec) {
+
+  if(U_FAILURE(ec)) {
+    setToBogus();
+    return;
+  }
+
+  if( (serialization != kSerialized)
+      || (data==NULL)
+      || (dataLen < 1)) {
+    ec = U_ILLEGAL_ARGUMENT_ERROR;
+    setToBogus();
+    return;
+  }
+
+  // bmp?
+  int32_t headerSize = ((data[0]&0x8000)) ?2:1;
+  int32_t bmpLength = (headerSize==1)?data[0]:data[1];
+
+  int32_t newLength = (((data[0]&0x7FFF)-bmpLength)/2)+bmpLength;
+#ifdef DEBUG_SERIALIZE
+  printf("dataLen %d headerSize %d bmpLen %d len %d. data[0]=%X/%X/%X/%X\n", dataLen,headerSize,bmpLength,newLength, data[0],data[1],data[2],data[3]);
+#endif
+  if(!ensureCapacity(newLength + 1)) {  // +1 for HIGH
+    return;
+  }
+  // copy bmp
+  int32_t i;
+  for(i = 0; i< bmpLength;i++) {
+    list[i] = data[i+headerSize];
+#ifdef DEBUG_SERIALIZE
+    printf("<<16@%d[%d] %X\n", i+headerSize, i, list[i]);
+#endif
+  }
+  // copy smp
+  for(i=bmpLength;i<newLength;i++) {
+    list[i] = ((UChar32)data[headerSize+bmpLength+(i-bmpLength)*2+0] << 16) +
+              ((UChar32)data[headerSize+bmpLength+(i-bmpLength)*2+1]);
+#ifdef DEBUG_SERIALIZE
+    printf("<<32@%d+[%d] %lX\n", headerSize+bmpLength+i, i, list[i]);
+#endif
+  }
+  U_ASSERT(i == newLength);
+  if (i == 0 || list[i - 1] != UNICODESET_HIGH) {
+    list[i++] = UNICODESET_HIGH;
+  }
+  len = i;
+}
+
+
+int32_t UnicodeSet::serialize(uint16_t *dest, int32_t destCapacity, UErrorCode& ec) const {
+    int32_t bmpLength, length, destLength;
+
+    if (U_FAILURE(ec)) {
+        return 0;
+    }
+
+    if (destCapacity<0 || (destCapacity>0 && dest==NULL)) {
+        ec=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* count necessary 16-bit units */
+    length=this->len-1; // Subtract 1 to ignore final UNICODESET_HIGH
+    // assert(length>=0);
+    if (length==0) {
+        /* empty set */
+        if (destCapacity>0) {
+            *dest=0;
+        } else {
+            ec=U_BUFFER_OVERFLOW_ERROR;
+        }
+        return 1;
+    }
+    /* now length>0 */
+
+    if (this->list[length-1]<=0xffff) {
+        /* all BMP */
+        bmpLength=length;
+    } else if (this->list[0]>=0x10000) {
+        /* all supplementary */
+        bmpLength=0;
+        length*=2;
+    } else {
+        /* some BMP, some supplementary */
+        for (bmpLength=0; bmpLength<length && this->list[bmpLength]<=0xffff; ++bmpLength) {}
+        length=bmpLength+2*(length-bmpLength);
+    }
+#ifdef DEBUG_SERIALIZE
+    printf(">> bmpLength%d length%d len%d\n", bmpLength, length, len);
+#endif
+    /* length: number of 16-bit array units */
+    if (length>0x7fff) {
+        /* there are only 15 bits for the length in the first serialized word */
+        ec=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    /*
+     * total serialized length:
+     * number of 16-bit array units (length) +
+     * 1 length unit (always) +
+     * 1 bmpLength unit (if there are supplementary values)
+     */
+    destLength=length+((length>bmpLength)?2:1);
+    if (destLength<=destCapacity) {
+        const UChar32 *p;
+        int32_t i;
+
+#ifdef DEBUG_SERIALIZE
+        printf("writeHdr\n");
+#endif
+        *dest=(uint16_t)length;
+        if (length>bmpLength) {
+            *dest|=0x8000;
+            *++dest=(uint16_t)bmpLength;
+        }
+        ++dest;
+
+        /* write the BMP part of the array */
+        p=this->list;
+        for (i=0; i<bmpLength; ++i) {
+#ifdef DEBUG_SERIALIZE
+          printf("writebmp: %x\n", (int)*p);
+#endif
+            *dest++=(uint16_t)*p++;
+        }
+
+        /* write the supplementary part of the array */
+        for (; i<length; i+=2) {
+#ifdef DEBUG_SERIALIZE
+          printf("write32: %x\n", (int)*p);
+#endif
+            *dest++=(uint16_t)(*p>>16);
+            *dest++=(uint16_t)*p++;
+        }
+    } else {
+        ec=U_BUFFER_OVERFLOW_ERROR;
+    }
+    return destLength;
+}
+
+//----------------------------------------------------------------
+// Implementation: Utility methods
+//----------------------------------------------------------------
+
+/**
+ * Allocate our strings vector and return TRUE if successful.
+ */
+UBool UnicodeSet::allocateStrings(UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return FALSE;
+    }
+    strings = new UVector(uprv_deleteUObject,
+                          uhash_compareUnicodeString, 1, status);
+    if (strings == NULL) { // Check for memory allocation error.
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    if (U_FAILURE(status)) {
+        delete strings;
+        strings = NULL;
+        return FALSE;
+    } 
+    return TRUE;
+}
+
+int32_t UnicodeSet::nextCapacity(int32_t minCapacity) {
+    // Grow exponentially to reduce the frequency of allocations.
+    if (minCapacity < INITIAL_CAPACITY) {
+        return minCapacity + INITIAL_CAPACITY;
+    } else if (minCapacity <= 2500) {
+        return 5 * minCapacity;
+    } else {
+        int32_t newCapacity = 2 * minCapacity;
+        if (newCapacity > MAX_LENGTH) {
+            newCapacity = MAX_LENGTH;
+        }
+        return newCapacity;
+    }
+}
+
+bool UnicodeSet::ensureCapacity(int32_t newLen) {
+    if (newLen > MAX_LENGTH) {
+        newLen = MAX_LENGTH;
+    }
+    if (newLen <= capacity) {
+        return true;
+    }
+    int32_t newCapacity = nextCapacity(newLen);
+    UChar32* temp = (UChar32*) uprv_malloc(newCapacity * sizeof(UChar32));
+    if (temp == NULL) {
+        setToBogus(); // set the object to bogus state if an OOM failure occurred.
+        return false;
+    }
+    // Copy only the actual contents.
+    uprv_memcpy(temp, list, len * sizeof(UChar32));
+    if (list != stackList) {
+        uprv_free(list);
+    }
+    list = temp;
+    capacity = newCapacity;
+    return true;
+}
+
+bool UnicodeSet::ensureBufferCapacity(int32_t newLen) {
+    if (newLen > MAX_LENGTH) {
+        newLen = MAX_LENGTH;
+    }
+    if (newLen <= bufferCapacity) {
+        return true;
+    }
+    int32_t newCapacity = nextCapacity(newLen);
+    UChar32* temp = (UChar32*) uprv_malloc(newCapacity * sizeof(UChar32));
+    if (temp == NULL) {
+        setToBogus();
+        return false;
+    }
+    // The buffer has no contents to be copied.
+    // It is always filled from scratch after this call.
+    if (buffer != stackList) {
+        uprv_free(buffer);
+    }
+    buffer = temp;
+    bufferCapacity = newCapacity;
+    return true;
+}
+
+/**
+ * Swap list and buffer.
+ */
+void UnicodeSet::swapBuffers(void) {
+    // swap list and buffer
+    UChar32* temp = list;
+    list = buffer;
+    buffer = temp;
+
+    int32_t c = capacity;
+    capacity = bufferCapacity;
+    bufferCapacity = c;
+}
+
+void UnicodeSet::setToBogus() {
+    clear(); // Remove everything in the set.
+    fFlags = kIsBogus;
+}
+
+//----------------------------------------------------------------
+// Implementation: Fundamental operators
+//----------------------------------------------------------------
+
+static inline UChar32 max(UChar32 a, UChar32 b) {
+    return (a > b) ? a : b;
+}
+
+// polarity = 0, 3 is normal: x xor y
+// polarity = 1, 2: x xor ~y == x === y
+
+void UnicodeSet::exclusiveOr(const UChar32* other, int32_t otherLen, int8_t polarity) {
+    if (isFrozen() || isBogus()) {
+        return;
+    }
+    if (!ensureBufferCapacity(len + otherLen)) {
+        return;
+    }
+
+    int32_t i = 0, j = 0, k = 0;
+    UChar32 a = list[i++];
+    UChar32 b;
+    if (polarity == 1 || polarity == 2) {
+        b = UNICODESET_LOW;
+        if (other[j] == UNICODESET_LOW) { // skip base if already LOW
+            ++j;
+            b = other[j];
+        }
+    } else {
+        b = other[j++];
+    }
+    // simplest of all the routines
+    // sort the values, discarding identicals!
+    for (;;) {
+        if (a < b) {
+            buffer[k++] = a;
+            a = list[i++];
+        } else if (b < a) {
+            buffer[k++] = b;
+            b = other[j++];
+        } else if (a != UNICODESET_HIGH) { // at this point, a == b
+            // discard both values!
+            a = list[i++];
+            b = other[j++];
+        } else { // DONE!
+            buffer[k++] = UNICODESET_HIGH;
+            len = k;
+            break;
+        }
+    }
+    swapBuffers();
+    releasePattern();
+}
+
+// polarity = 0 is normal: x union y
+// polarity = 2: x union ~y
+// polarity = 1: ~x union y
+// polarity = 3: ~x union ~y
+
+void UnicodeSet::add(const UChar32* other, int32_t otherLen, int8_t polarity) {
+    if (isFrozen() || isBogus() || other==NULL) {
+        return;
+    }
+    if (!ensureBufferCapacity(len + otherLen)) {
+        return;
+    }
+
+    int32_t i = 0, j = 0, k = 0;
+    UChar32 a = list[i++];
+    UChar32 b = other[j++];
+    // change from xor is that we have to check overlapping pairs
+    // polarity bit 1 means a is second, bit 2 means b is.
+    for (;;) {
+        switch (polarity) {
+          case 0: // both first; take lower if unequal
+            if (a < b) { // take a
+                // Back up over overlapping ranges in buffer[]
+                if (k > 0 && a <= buffer[k-1]) {
+                    // Pick latter end value in buffer[] vs. list[]
+                    a = max(list[i], buffer[--k]);
+                } else {
+                    // No overlap
+                    buffer[k++] = a;
+                    a = list[i];
+                }
+                i++; // Common if/else code factored out
+                polarity ^= 1;
+            } else if (b < a) { // take b
+                if (k > 0 && b <= buffer[k-1]) {
+                    b = max(other[j], buffer[--k]);
+                } else {
+                    buffer[k++] = b;
+                    b = other[j];
+                }
+                j++;
+                polarity ^= 2;
+            } else { // a == b, take a, drop b
+                if (a == UNICODESET_HIGH) goto loop_end;
+                // This is symmetrical; it doesn't matter if
+                // we backtrack with a or b. - liu
+                if (k > 0 && a <= buffer[k-1]) {
+                    a = max(list[i], buffer[--k]);
+                } else {
+                    // No overlap
+                    buffer[k++] = a;
+                    a = list[i];
+                }
+                i++;
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+          case 3: // both second; take higher if unequal, and drop other
+            if (b <= a) { // take a
+                if (a == UNICODESET_HIGH) goto loop_end;
+                buffer[k++] = a;
+            } else { // take b
+                if (b == UNICODESET_HIGH) goto loop_end;
+                buffer[k++] = b;
+            }
+            a = list[i++];
+            polarity ^= 1;   // factored common code
+            b = other[j++];
+            polarity ^= 2;
+            break;
+          case 1: // a second, b first; if b < a, overlap
+            if (a < b) { // no overlap, take a
+                buffer[k++] = a; a = list[i++]; polarity ^= 1;
+            } else if (b < a) { // OVERLAP, drop b
+                b = other[j++];
+                polarity ^= 2;
+            } else { // a == b, drop both!
+                if (a == UNICODESET_HIGH) goto loop_end;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+          case 2: // a first, b second; if a < b, overlap
+            if (b < a) { // no overlap, take b
+                buffer[k++] = b;
+                b = other[j++];
+                polarity ^= 2;
+            } else  if (a < b) { // OVERLAP, drop a
+                a = list[i++];
+                polarity ^= 1;
+            } else { // a == b, drop both!
+                if (a == UNICODESET_HIGH) goto loop_end;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+        }
+    }
+ loop_end:
+    buffer[k++] = UNICODESET_HIGH;    // terminate
+    len = k;
+    swapBuffers();
+    releasePattern();
+}
+
+// polarity = 0 is normal: x intersect y
+// polarity = 2: x intersect ~y == set-minus
+// polarity = 1: ~x intersect y
+// polarity = 3: ~x intersect ~y
+
+void UnicodeSet::retain(const UChar32* other, int32_t otherLen, int8_t polarity) {
+    if (isFrozen() || isBogus()) {
+        return;
+    }
+    if (!ensureBufferCapacity(len + otherLen)) {
+        return;
+    }
+
+    int32_t i = 0, j = 0, k = 0;
+    UChar32 a = list[i++];
+    UChar32 b = other[j++];
+    // change from xor is that we have to check overlapping pairs
+    // polarity bit 1 means a is second, bit 2 means b is.
+    for (;;) {
+        switch (polarity) {
+          case 0: // both first; drop the smaller
+            if (a < b) { // drop a
+                a = list[i++];
+                polarity ^= 1;
+            } else if (b < a) { // drop b
+                b = other[j++];
+                polarity ^= 2;
+            } else { // a == b, take one, drop other
+                if (a == UNICODESET_HIGH) goto loop_end;
+                buffer[k++] = a;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+          case 3: // both second; take lower if unequal
+            if (a < b) { // take a
+                buffer[k++] = a;
+                a = list[i++];
+                polarity ^= 1;
+            } else if (b < a) { // take b
+                buffer[k++] = b;
+                b = other[j++];
+                polarity ^= 2;
+            } else { // a == b, take one, drop other
+                if (a == UNICODESET_HIGH) goto loop_end;
+                buffer[k++] = a;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+          case 1: // a second, b first;
+            if (a < b) { // NO OVERLAP, drop a
+                a = list[i++];
+                polarity ^= 1;
+            } else if (b < a) { // OVERLAP, take b
+                buffer[k++] = b;
+                b = other[j++];
+                polarity ^= 2;
+            } else { // a == b, drop both!
+                if (a == UNICODESET_HIGH) goto loop_end;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+          case 2: // a first, b second; if a < b, overlap
+            if (b < a) { // no overlap, drop b
+                b = other[j++];
+                polarity ^= 2;
+            } else  if (a < b) { // OVERLAP, take a
+                buffer[k++] = a;
+                a = list[i++];
+                polarity ^= 1;
+            } else { // a == b, drop both!
+                if (a == UNICODESET_HIGH) goto loop_end;
+                a = list[i++];
+                polarity ^= 1;
+                b = other[j++];
+                polarity ^= 2;
+            }
+            break;
+        }
+    }
+ loop_end:
+    buffer[k++] = UNICODESET_HIGH;    // terminate
+    len = k;
+    swapBuffers();
+    releasePattern();
+}
+
+/**
+ * Append the <code>toPattern()</code> representation of a
+ * string to the given <code>StringBuffer</code>.
+ */
+void UnicodeSet::_appendToPat(UnicodeString& buf, const UnicodeString& s, UBool
+escapeUnprintable) {
+    UChar32 cp;
+    for (int32_t i = 0; i < s.length(); i += U16_LENGTH(cp)) {
+        _appendToPat(buf, cp = s.char32At(i), escapeUnprintable);
+    }
+}
+
+/**
+ * Append the <code>toPattern()</code> representation of a
+ * character to the given <code>StringBuffer</code>.
+ */
+void UnicodeSet::_appendToPat(UnicodeString& buf, UChar32 c, UBool
+escapeUnprintable) {
+    if (escapeUnprintable && ICU_Utility::isUnprintable(c)) {
+        // Use hex escape notation (\uxxxx or \Uxxxxxxxx) for anything
+        // unprintable
+        if (ICU_Utility::escapeUnprintable(buf, c)) {
+            return;
+        }
+    }
+    // Okay to let ':' pass through
+    switch (c) {
+    case SET_OPEN:
+    case SET_CLOSE:
+    case HYPHEN:
+    case COMPLEMENT:
+    case INTERSECTION:
+    case BACKSLASH:
+    case OPEN_BRACE:
+    case CLOSE_BRACE:
+    case COLON:
+    case SymbolTable::SYMBOL_REF:
+        buf.append(BACKSLASH);
+        break;
+    default:
+        // Escape whitespace
+        if (PatternProps::isWhiteSpace(c)) {
+            buf.append(BACKSLASH);
+        }
+        break;
+    }
+    buf.append(c);
+}
+
+/**
+ * Append a string representation of this set to result.  This will be
+ * a cleaned version of the string passed to applyPattern(), if there
+ * is one.  Otherwise it will be generated.
+ */
+UnicodeString& UnicodeSet::_toPattern(UnicodeString& result,
+                                      UBool escapeUnprintable) const
+{
+    if (pat != NULL) {
+        int32_t i;
+        int32_t backslashCount = 0;
+        for (i=0; i<patLen; ) {
+            UChar32 c;
+            U16_NEXT(pat, i, patLen, c);
+            if (escapeUnprintable && ICU_Utility::isUnprintable(c)) {
+                // If the unprintable character is preceded by an odd
+                // number of backslashes, then it has been escaped.
+                // Before unescaping it, we delete the final
+                // backslash.
+                if ((backslashCount % 2) == 1) {
+                    result.truncate(result.length() - 1);
+                }
+                ICU_Utility::escapeUnprintable(result, c);
+                backslashCount = 0;
+            } else {
+                result.append(c);
+                if (c == BACKSLASH) {
+                    ++backslashCount;
+                } else {
+                    backslashCount = 0;
+                }
+            }
+        }
+        return result;
+    }
+
+    return _generatePattern(result, escapeUnprintable);
+}
+
+/**
+ * Returns a string representation of this set.  If the result of
+ * calling this function is passed to a UnicodeSet constructor, it
+ * will produce another set that is equal to this one.
+ */
+UnicodeString& UnicodeSet::toPattern(UnicodeString& result,
+                                     UBool escapeUnprintable) const
+{
+    result.truncate(0);
+    return _toPattern(result, escapeUnprintable);
+}
+
+/**
+ * Generate and append a string representation of this set to result.
+ * This does not use this.pat, the cleaned up copy of the string
+ * passed to applyPattern().
+ */
+UnicodeString& UnicodeSet::_generatePattern(UnicodeString& result,
+                                            UBool escapeUnprintable) const
+{
+    result.append(SET_OPEN);
+
+//  // Check against the predefined categories.  We implicitly build
+//  // up ALL category sets the first time toPattern() is called.
+//  for (int8_t cat=0; cat<Unicode::GENERAL_TYPES_COUNT; ++cat) {
+//      if (*this == getCategorySet(cat)) {
+//          result.append(COLON);
+//          result.append(CATEGORY_NAMES, cat*2, 2);
+//          return result.append(CATEGORY_CLOSE);
+//      }
+//  }
+
+    int32_t count = getRangeCount();
+
+    // If the set contains at least 2 intervals and includes both
+    // MIN_VALUE and MAX_VALUE, then the inverse representation will
+    // be more economical.
+    if (count > 1 &&
+        getRangeStart(0) == MIN_VALUE &&
+        getRangeEnd(count-1) == MAX_VALUE) {
+
+        // Emit the inverse
+        result.append(COMPLEMENT);
+
+        for (int32_t i = 1; i < count; ++i) {
+            UChar32 start = getRangeEnd(i-1)+1;
+            UChar32 end = getRangeStart(i)-1;
+            _appendToPat(result, start, escapeUnprintable);
+            if (start != end) {
+                if ((start+1) != end) {
+                    result.append(HYPHEN);
+                }
+                _appendToPat(result, end, escapeUnprintable);
+            }
+        }
+    }
+
+    // Default; emit the ranges as pairs
+    else {
+        for (int32_t i = 0; i < count; ++i) {
+            UChar32 start = getRangeStart(i);
+            UChar32 end = getRangeEnd(i);
+            _appendToPat(result, start, escapeUnprintable);
+            if (start != end) {
+                if ((start+1) != end) {
+                    result.append(HYPHEN);
+                }
+                _appendToPat(result, end, escapeUnprintable);
+            }
+        }
+    }
+
+    if (strings != nullptr) {
+        for (int32_t i = 0; i<strings->size(); ++i) {
+            result.append(OPEN_BRACE);
+            _appendToPat(result,
+                         *(const UnicodeString*) strings->elementAt(i),
+                         escapeUnprintable);
+            result.append(CLOSE_BRACE);
+        }
+    }
+    return result.append(SET_CLOSE);
+}
+
+/**
+* Release existing cached pattern
+*/
+void UnicodeSet::releasePattern() {
+    if (pat) {
+        uprv_free(pat);
+        pat = NULL;
+        patLen = 0;
+    }
+}
+
+/**
+* Set the new pattern to cache.
+*/
+void UnicodeSet::setPattern(const char16_t *newPat, int32_t newPatLen) {
+    releasePattern();
+    pat = (UChar *)uprv_malloc((newPatLen + 1) * sizeof(UChar));
+    if (pat) {
+        patLen = newPatLen;
+        u_memcpy(pat, newPat, patLen);
+        pat[patLen] = 0;
+    }
+    // else we don't care if malloc failed. This was just a nice cache.
+    // We can regenerate an equivalent pattern later when requested.
+}
+
+UnicodeSet *UnicodeSet::freeze() {
+    if(!isFrozen() && !isBogus()) {
+        compact();
+
+        // Optimize contains() and span() and similar functions.
+        if (hasStrings()) {
+            stringSpan = new UnicodeSetStringSpan(*this, *strings, UnicodeSetStringSpan::ALL);
+            if (stringSpan == nullptr) {
+                setToBogus();
+                return this;
+            } else if (!stringSpan->needsStringSpanUTF16()) {
+                // All strings are irrelevant for span() etc. because
+                // all of each string's code points are contained in this set.
+                // Do not check needsStringSpanUTF8() because UTF-8 has at most as
+                // many relevant strings as UTF-16.
+                // (Thus needsStringSpanUTF8() implies needsStringSpanUTF16().)
+                delete stringSpan;
+                stringSpan = NULL;
+            }
+        }
+        if (stringSpan == NULL) {
+            // No span-relevant strings: Optimize for code point spans.
+            bmpSet=new BMPSet(list, len);
+            if (bmpSet == NULL) { // Check for memory allocation error.
+                setToBogus();
+            }
+        }
+    }
+    return this;
+}
+
+int32_t UnicodeSet::span(const UChar *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(length>0 && bmpSet!=NULL) {
+        return (int32_t)(bmpSet->span(s, s+length, spanCondition)-s);
+    }
+    if(length<0) {
+        length=u_strlen(s);
+    }
+    if(length==0) {
+        return 0;
+    }
+    if(stringSpan!=NULL) {
+        return stringSpan->span(s, length, spanCondition);
+    } else if(hasStrings()) {
+        uint32_t which= spanCondition==USET_SPAN_NOT_CONTAINED ?
+                            UnicodeSetStringSpan::FWD_UTF16_NOT_CONTAINED :
+                            UnicodeSetStringSpan::FWD_UTF16_CONTAINED;
+        UnicodeSetStringSpan strSpan(*this, *strings, which);
+        if(strSpan.needsStringSpanUTF16()) {
+            return strSpan.span(s, length, spanCondition);
+        }
+    }
+
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    UChar32 c;
+    int32_t start=0, prev=0;
+    do {
+        U16_NEXT(s, start, length, c);
+        if(spanCondition!=contains(c)) {
+            break;
+        }
+    } while((prev=start)<length);
+    return prev;
+}
+
+int32_t UnicodeSet::spanBack(const UChar *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(length>0 && bmpSet!=NULL) {
+        return (int32_t)(bmpSet->spanBack(s, s+length, spanCondition)-s);
+    }
+    if(length<0) {
+        length=u_strlen(s);
+    }
+    if(length==0) {
+        return 0;
+    }
+    if(stringSpan!=NULL) {
+        return stringSpan->spanBack(s, length, spanCondition);
+    } else if(hasStrings()) {
+        uint32_t which= spanCondition==USET_SPAN_NOT_CONTAINED ?
+                            UnicodeSetStringSpan::BACK_UTF16_NOT_CONTAINED :
+                            UnicodeSetStringSpan::BACK_UTF16_CONTAINED;
+        UnicodeSetStringSpan strSpan(*this, *strings, which);
+        if(strSpan.needsStringSpanUTF16()) {
+            return strSpan.spanBack(s, length, spanCondition);
+        }
+    }
+
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    UChar32 c;
+    int32_t prev=length;
+    do {
+        U16_PREV(s, 0, length, c);
+        if(spanCondition!=contains(c)) {
+            break;
+        }
+    } while((prev=length)>0);
+    return prev;
+}
+
+int32_t UnicodeSet::spanUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(length>0 && bmpSet!=NULL) {
+        const uint8_t *s0=(const uint8_t *)s;
+        return (int32_t)(bmpSet->spanUTF8(s0, length, spanCondition)-s0);
+    }
+    if(length<0) {
+        length=(int32_t)uprv_strlen(s);
+    }
+    if(length==0) {
+        return 0;
+    }
+    if(stringSpan!=NULL) {
+        return stringSpan->spanUTF8((const uint8_t *)s, length, spanCondition);
+    } else if(hasStrings()) {
+        uint32_t which= spanCondition==USET_SPAN_NOT_CONTAINED ?
+                            UnicodeSetStringSpan::FWD_UTF8_NOT_CONTAINED :
+                            UnicodeSetStringSpan::FWD_UTF8_CONTAINED;
+        UnicodeSetStringSpan strSpan(*this, *strings, which);
+        if(strSpan.needsStringSpanUTF8()) {
+            return strSpan.spanUTF8((const uint8_t *)s, length, spanCondition);
+        }
+    }
+
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    UChar32 c;
+    int32_t start=0, prev=0;
+    do {
+        U8_NEXT_OR_FFFD(s, start, length, c);
+        if(spanCondition!=contains(c)) {
+            break;
+        }
+    } while((prev=start)<length);
+    return prev;
+}
+
+int32_t UnicodeSet::spanBackUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(length>0 && bmpSet!=NULL) {
+        const uint8_t *s0=(const uint8_t *)s;
+        return bmpSet->spanBackUTF8(s0, length, spanCondition);
+    }
+    if(length<0) {
+        length=(int32_t)uprv_strlen(s);
+    }
+    if(length==0) {
+        return 0;
+    }
+    if(stringSpan!=NULL) {
+        return stringSpan->spanBackUTF8((const uint8_t *)s, length, spanCondition);
+    } else if(hasStrings()) {
+        uint32_t which= spanCondition==USET_SPAN_NOT_CONTAINED ?
+                            UnicodeSetStringSpan::BACK_UTF8_NOT_CONTAINED :
+                            UnicodeSetStringSpan::BACK_UTF8_CONTAINED;
+        UnicodeSetStringSpan strSpan(*this, *strings, which);
+        if(strSpan.needsStringSpanUTF8()) {
+            return strSpan.spanBackUTF8((const uint8_t *)s, length, spanCondition);
+        }
+    }
+
+    if(spanCondition!=USET_SPAN_NOT_CONTAINED) {
+        spanCondition=USET_SPAN_CONTAINED;  // Pin to 0/1 values.
+    }
+
+    UChar32 c;
+    int32_t prev=length;
+    do {
+        U8_PREV_OR_FFFD(s, 0, length, c);
+        if(spanCondition!=contains(c)) {
+            break;
+        }
+    } while((prev=length)>0);
+    return prev;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/uniset_closure.cpp b/thirdparty/icu4c/common/uniset_closure.cpp
new file mode 100644
index 0000000000..882231ba1a
--- /dev/null
+++ b/thirdparty/icu4c/common/uniset_closure.cpp
@@ -0,0 +1,250 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uniset_closure.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011may30
+*   created by: Markus W. Scherer
+*
+*   UnicodeSet::closeOver() and related methods moved here from uniset_props.cpp
+*   to simplify dependencies.
+*   In particular, this depends on the BreakIterator, but the BreakIterator
+*   code also builds UnicodeSets from patterns and needs uniset_props.
+*/
+
+#include "unicode/brkiter.h"
+#include "unicode/locid.h"
+#include "unicode/parsepos.h"
+#include "unicode/uniset.h"
+#include "cmemory.h"
+#include "ruleiter.h"
+#include "ucase.h"
+#include "util.h"
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+// TODO memory debugging provided inside uniset.cpp
+// could be made available here but probably obsolete with use of modern
+// memory leak checker tools
+#define _dbgct(me)
+
+//----------------------------------------------------------------
+// Constructors &c
+//----------------------------------------------------------------
+
+UnicodeSet::UnicodeSet(const UnicodeString& pattern,
+                       uint32_t options,
+                       const SymbolTable* symbols,
+                       UErrorCode& status) {
+    applyPattern(pattern, options, symbols, status);
+    _dbgct(this);
+}
+
+UnicodeSet::UnicodeSet(const UnicodeString& pattern, ParsePosition& pos,
+                       uint32_t options,
+                       const SymbolTable* symbols,
+                       UErrorCode& status) {
+    applyPattern(pattern, pos, options, symbols, status);
+    _dbgct(this);
+}
+
+//----------------------------------------------------------------
+// Public API
+//----------------------------------------------------------------
+
+UnicodeSet& UnicodeSet::applyPattern(const UnicodeString& pattern,
+                                     uint32_t options,
+                                     const SymbolTable* symbols,
+                                     UErrorCode& status) {
+    ParsePosition pos(0);
+    applyPattern(pattern, pos, options, symbols, status);
+    if (U_FAILURE(status)) return *this;
+
+    int32_t i = pos.getIndex();
+
+    if (options & USET_IGNORE_SPACE) {
+        // Skip over trailing whitespace
+        ICU_Utility::skipWhitespace(pattern, i, TRUE);
+    }
+
+    if (i != pattern.length()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return *this;
+}
+
+UnicodeSet& UnicodeSet::applyPattern(const UnicodeString& pattern,
+                              ParsePosition& pos,
+                              uint32_t options,
+                              const SymbolTable* symbols,
+                              UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return *this;
+    }
+    if (isFrozen()) {
+        status = U_NO_WRITE_PERMISSION;
+        return *this;
+    }
+    // Need to build the pattern in a temporary string because
+    // _applyPattern calls add() etc., which set pat to empty.
+    UnicodeString rebuiltPat;
+    RuleCharacterIterator chars(pattern, symbols, pos);
+    applyPattern(chars, symbols, rebuiltPat, options, &UnicodeSet::closeOver, 0, status);
+    if (U_FAILURE(status)) return *this;
+    if (chars.inVariable()) {
+        // syntaxError(chars, "Extra chars in variable value");
+        status = U_MALFORMED_SET;
+        return *this;
+    }
+    setPattern(rebuiltPat);
+    return *this;
+}
+
+// USetAdder implementation
+// Does not use uset.h to reduce code dependencies
+static void U_CALLCONV
+_set_add(USet *set, UChar32 c) {
+    ((UnicodeSet *)set)->add(c);
+}
+
+static void U_CALLCONV
+_set_addRange(USet *set, UChar32 start, UChar32 end) {
+    ((UnicodeSet *)set)->add(start, end);
+}
+
+static void U_CALLCONV
+_set_addString(USet *set, const UChar *str, int32_t length) {
+    ((UnicodeSet *)set)->add(UnicodeString((UBool)(length<0), str, length));
+}
+
+//----------------------------------------------------------------
+// Case folding API
+//----------------------------------------------------------------
+
+// add the result of a full case mapping to the set
+// use str as a temporary string to avoid constructing one
+static inline void
+addCaseMapping(UnicodeSet &set, int32_t result, const UChar *full, UnicodeString &str) {
+    if(result >= 0) {
+        if(result > UCASE_MAX_STRING_LENGTH) {
+            // add a single-code point case mapping
+            set.add(result);
+        } else {
+            // add a string case mapping from full with length result
+            str.setTo((UBool)FALSE, full, result);
+            set.add(str);
+        }
+    }
+    // result < 0: the code point mapped to itself, no need to add it
+    // see ucase.h
+}
+
+UnicodeSet& UnicodeSet::closeOver(int32_t attribute) {
+    if (isFrozen() || isBogus()) {
+        return *this;
+    }
+    if (attribute & (USET_CASE_INSENSITIVE | USET_ADD_CASE_MAPPINGS)) {
+        {
+            UnicodeSet foldSet(*this);
+            UnicodeString str;
+            USetAdder sa = {
+                foldSet.toUSet(),
+                _set_add,
+                _set_addRange,
+                _set_addString,
+                NULL, // don't need remove()
+                NULL // don't need removeRange()
+            };
+
+            // start with input set to guarantee inclusion
+            // USET_CASE: remove strings because the strings will actually be reduced (folded);
+            //            therefore, start with no strings and add only those needed
+            if ((attribute & USET_CASE_INSENSITIVE) && foldSet.hasStrings()) {
+                foldSet.strings->removeAllElements();
+            }
+
+            int32_t n = getRangeCount();
+            UChar32 result;
+            const UChar *full;
+
+            for (int32_t i=0; i<n; ++i) {
+                UChar32 start = getRangeStart(i);
+                UChar32 end   = getRangeEnd(i);
+
+                if (attribute & USET_CASE_INSENSITIVE) {
+                    // full case closure
+                    for (UChar32 cp=start; cp<=end; ++cp) {
+                        ucase_addCaseClosure(cp, &sa);
+                    }
+                } else {
+                    // add case mappings
+                    // (does not add long s for regular s, or Kelvin for k, for example)
+                    for (UChar32 cp=start; cp<=end; ++cp) {
+                        result = ucase_toFullLower(cp, NULL, NULL, &full, UCASE_LOC_ROOT);
+                        addCaseMapping(foldSet, result, full, str);
+
+                        result = ucase_toFullTitle(cp, NULL, NULL, &full, UCASE_LOC_ROOT);
+                        addCaseMapping(foldSet, result, full, str);
+
+                        result = ucase_toFullUpper(cp, NULL, NULL, &full, UCASE_LOC_ROOT);
+                        addCaseMapping(foldSet, result, full, str);
+
+                        result = ucase_toFullFolding(cp, &full, 0);
+                        addCaseMapping(foldSet, result, full, str);
+                    }
+                }
+            }
+            if (hasStrings()) {
+                if (attribute & USET_CASE_INSENSITIVE) {
+                    for (int32_t j=0; j<strings->size(); ++j) {
+                        str = *(const UnicodeString *) strings->elementAt(j);
+                        str.foldCase();
+                        if(!ucase_addStringCaseClosure(str.getBuffer(), str.length(), &sa)) {
+                            foldSet.add(str); // does not map to code points: add the folded string itself
+                        }
+                    }
+                } else {
+                    Locale root("");
+#if !UCONFIG_NO_BREAK_ITERATION
+                    UErrorCode status = U_ZERO_ERROR;
+                    BreakIterator *bi = BreakIterator::createWordInstance(root, status);
+                    if (U_SUCCESS(status)) {
+#endif
+                        const UnicodeString *pStr;
+
+                        for (int32_t j=0; j<strings->size(); ++j) {
+                            pStr = (const UnicodeString *) strings->elementAt(j);
+                            (str = *pStr).toLower(root);
+                            foldSet.add(str);
+#if !UCONFIG_NO_BREAK_ITERATION
+                            (str = *pStr).toTitle(bi, root);
+                            foldSet.add(str);
+#endif
+                            (str = *pStr).toUpper(root);
+                            foldSet.add(str);
+                            (str = *pStr).foldCase();
+                            foldSet.add(str);
+                        }
+#if !UCONFIG_NO_BREAK_ITERATION
+                    }
+                    delete bi;
+#endif
+                }
+            }
+            *this = foldSet;
+        }
+    }
+    return *this;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/uniset_props.cpp b/thirdparty/icu4c/common/uniset_props.cpp
new file mode 100644
index 0000000000..37277fcb75
--- /dev/null
+++ b/thirdparty/icu4c/common/uniset_props.cpp
@@ -0,0 +1,1174 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uniset_props.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug25
+*   created by: Markus W. Scherer
+*
+*   Character property dependent functions moved here from uniset.cpp
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+#include "unicode/parsepos.h"
+#include "unicode/uchar.h"
+#include "unicode/uscript.h"
+#include "unicode/symtable.h"
+#include "unicode/uset.h"
+#include "unicode/locid.h"
+#include "unicode/brkiter.h"
+#include "uset_imp.h"
+#include "ruleiter.h"
+#include "cmemory.h"
+#include "ucln_cmn.h"
+#include "util.h"
+#include "uvector.h"
+#include "uprops.h"
+#include "propname.h"
+#include "normalizer2impl.h"
+#include "uinvchar.h"
+#include "uprops.h"
+#include "charstr.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "umutex.h"
+#include "uassert.h"
+#include "hash.h"
+
+U_NAMESPACE_USE
+
+// Define UChar constants using hex for EBCDIC compatibility
+// Used #define to reduce private static exports and memory access time.
+#define SET_OPEN        ((UChar)0x005B) /*[*/
+#define SET_CLOSE       ((UChar)0x005D) /*]*/
+#define HYPHEN          ((UChar)0x002D) /*-*/
+#define COMPLEMENT      ((UChar)0x005E) /*^*/
+#define COLON           ((UChar)0x003A) /*:*/
+#define BACKSLASH       ((UChar)0x005C) /*\*/
+#define INTERSECTION    ((UChar)0x0026) /*&*/
+#define UPPER_U         ((UChar)0x0055) /*U*/
+#define LOWER_U         ((UChar)0x0075) /*u*/
+#define OPEN_BRACE      ((UChar)123)    /*{*/
+#define CLOSE_BRACE     ((UChar)125)    /*}*/
+#define UPPER_P         ((UChar)0x0050) /*P*/
+#define LOWER_P         ((UChar)0x0070) /*p*/
+#define UPPER_N         ((UChar)78)     /*N*/
+#define EQUALS          ((UChar)0x003D) /*=*/
+
+//static const UChar POSIX_OPEN[]  = { SET_OPEN,COLON,0 };  // "[:"
+static const UChar POSIX_CLOSE[] = { COLON,SET_CLOSE,0 };  // ":]"
+//static const UChar PERL_OPEN[]   = { BACKSLASH,LOWER_P,0 }; // "\\p"
+//static const UChar PERL_CLOSE[]  = { CLOSE_BRACE,0 };    // "}"
+//static const UChar NAME_OPEN[]   = { BACKSLASH,UPPER_N,0 };  // "\\N"
+static const UChar HYPHEN_RIGHT_BRACE[] = {HYPHEN,SET_CLOSE,0}; /*-]*/
+
+// Special property set IDs
+static const char ANY[]   = "ANY";   // [\u0000-\U0010FFFF]
+static const char ASCII[] = "ASCII"; // [\u0000-\u007F]
+static const char ASSIGNED[] = "Assigned"; // [:^Cn:]
+
+// Unicode name property alias
+#define NAME_PROP "na"
+#define NAME_PROP_LENGTH 2
+
+/**
+ * Delimiter string used in patterns to close a category reference:
+ * ":]".  Example: "[:Lu:]".
+ */
+//static const UChar CATEGORY_CLOSE[] = {COLON, SET_CLOSE, 0x0000}; /* ":]" */
+
+// Cached sets ------------------------------------------------------------- ***
+
+U_CDECL_BEGIN
+static UBool U_CALLCONV uset_cleanup();
+
+static UnicodeSet *uni32Singleton;
+static icu::UInitOnce uni32InitOnce = U_INITONCE_INITIALIZER;
+
+/**
+ * Cleanup function for UnicodeSet
+ */
+static UBool U_CALLCONV uset_cleanup(void) {
+    delete uni32Singleton;
+    uni32Singleton = NULL;
+    uni32InitOnce.reset();
+    return TRUE;
+}
+
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+// Cache some sets for other services -------------------------------------- ***
+void U_CALLCONV createUni32Set(UErrorCode &errorCode) {
+    U_ASSERT(uni32Singleton == NULL);
+    uni32Singleton = new UnicodeSet(UNICODE_STRING_SIMPLE("[:age=3.2:]"), errorCode);
+    if(uni32Singleton==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        uni32Singleton->freeze();
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_USET, uset_cleanup);
+}
+
+
+U_CFUNC UnicodeSet *
+uniset_getUnicode32Instance(UErrorCode &errorCode) {
+    umtx_initOnce(uni32InitOnce, &createUni32Set, errorCode);
+    return uni32Singleton;
+}
+
+// helper functions for matching of pattern syntax pieces ------------------ ***
+// these functions are parallel to the PERL_OPEN etc. strings above
+
+// using these functions is not only faster than UnicodeString::compare() and
+// caseCompare(), but they also make UnicodeSet work for simple patterns when
+// no Unicode properties data is available - when caseCompare() fails
+
+static inline UBool
+isPerlOpen(const UnicodeString &pattern, int32_t pos) {
+    UChar c;
+    return pattern.charAt(pos)==BACKSLASH && ((c=pattern.charAt(pos+1))==LOWER_P || c==UPPER_P);
+}
+
+/*static inline UBool
+isPerlClose(const UnicodeString &pattern, int32_t pos) {
+    return pattern.charAt(pos)==CLOSE_BRACE;
+}*/
+
+static inline UBool
+isNameOpen(const UnicodeString &pattern, int32_t pos) {
+    return pattern.charAt(pos)==BACKSLASH && pattern.charAt(pos+1)==UPPER_N;
+}
+
+static inline UBool
+isPOSIXOpen(const UnicodeString &pattern, int32_t pos) {
+    return pattern.charAt(pos)==SET_OPEN && pattern.charAt(pos+1)==COLON;
+}
+
+/*static inline UBool
+isPOSIXClose(const UnicodeString &pattern, int32_t pos) {
+    return pattern.charAt(pos)==COLON && pattern.charAt(pos+1)==SET_CLOSE;
+}*/
+
+// TODO memory debugging provided inside uniset.cpp
+// could be made available here but probably obsolete with use of modern
+// memory leak checker tools
+#define _dbgct(me)
+
+}  // namespace
+
+//----------------------------------------------------------------
+// Constructors &c
+//----------------------------------------------------------------
+
+/**
+ * Constructs a set from the given pattern, optionally ignoring
+ * white space.  See the class description for the syntax of the
+ * pattern language.
+ * @param pattern a string specifying what characters are in the set
+ */
+UnicodeSet::UnicodeSet(const UnicodeString& pattern,
+                       UErrorCode& status) {
+    applyPattern(pattern, status);
+    _dbgct(this);
+}
+
+//----------------------------------------------------------------
+// Public API
+//----------------------------------------------------------------
+
+UnicodeSet& UnicodeSet::applyPattern(const UnicodeString& pattern,
+                                     UErrorCode& status) {
+    // Equivalent to
+    //   return applyPattern(pattern, USET_IGNORE_SPACE, NULL, status);
+    // but without dependency on closeOver().
+    ParsePosition pos(0);
+    applyPatternIgnoreSpace(pattern, pos, NULL, status);
+    if (U_FAILURE(status)) return *this;
+
+    int32_t i = pos.getIndex();
+    // Skip over trailing whitespace
+    ICU_Utility::skipWhitespace(pattern, i, TRUE);
+    if (i != pattern.length()) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return *this;
+}
+
+void
+UnicodeSet::applyPatternIgnoreSpace(const UnicodeString& pattern,
+                                    ParsePosition& pos,
+                                    const SymbolTable* symbols,
+                                    UErrorCode& status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    if (isFrozen()) {
+        status = U_NO_WRITE_PERMISSION;
+        return;
+    }
+    // Need to build the pattern in a temporary string because
+    // _applyPattern calls add() etc., which set pat to empty.
+    UnicodeString rebuiltPat;
+    RuleCharacterIterator chars(pattern, symbols, pos);
+    applyPattern(chars, symbols, rebuiltPat, USET_IGNORE_SPACE, NULL, 0, status);
+    if (U_FAILURE(status)) return;
+    if (chars.inVariable()) {
+        // syntaxError(chars, "Extra chars in variable value");
+        status = U_MALFORMED_SET;
+        return;
+    }
+    setPattern(rebuiltPat);
+}
+
+/**
+ * Return true if the given position, in the given pattern, appears
+ * to be the start of a UnicodeSet pattern.
+ */
+UBool UnicodeSet::resemblesPattern(const UnicodeString& pattern, int32_t pos) {
+    return ((pos+1) < pattern.length() &&
+            pattern.charAt(pos) == (UChar)91/*[*/) ||
+        resemblesPropertyPattern(pattern, pos);
+}
+
+//----------------------------------------------------------------
+// Implementation: Pattern parsing
+//----------------------------------------------------------------
+
+namespace {
+
+/**
+ * A small all-inline class to manage a UnicodeSet pointer.  Add
+ * operator->() etc. as needed.
+ */
+class UnicodeSetPointer {
+    UnicodeSet* p;
+public:
+    inline UnicodeSetPointer() : p(0) {}
+    inline ~UnicodeSetPointer() { delete p; }
+    inline UnicodeSet* pointer() { return p; }
+    inline UBool allocate() {
+        if (p == 0) {
+            p = new UnicodeSet();
+        }
+        return p != 0;
+    }
+};
+
+constexpr int32_t MAX_DEPTH = 100;
+
+}  // namespace
+
+/**
+ * Parse the pattern from the given RuleCharacterIterator.  The
+ * iterator is advanced over the parsed pattern.
+ * @param chars iterator over the pattern characters.  Upon return
+ * it will be advanced to the first character after the parsed
+ * pattern, or the end of the iteration if all characters are
+ * parsed.
+ * @param symbols symbol table to use to parse and dereference
+ * variables, or null if none.
+ * @param rebuiltPat the pattern that was parsed, rebuilt or
+ * copied from the input pattern, as appropriate.
+ * @param options a bit mask of zero or more of the following:
+ * IGNORE_SPACE, CASE.
+ */
+void UnicodeSet::applyPattern(RuleCharacterIterator& chars,
+                              const SymbolTable* symbols,
+                              UnicodeString& rebuiltPat,
+                              uint32_t options,
+                              UnicodeSet& (UnicodeSet::*caseClosure)(int32_t attribute),
+                              int32_t depth,
+                              UErrorCode& ec) {
+    if (U_FAILURE(ec)) return;
+    if (depth > MAX_DEPTH) {
+        ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    // Syntax characters: [ ] ^ - & { }
+
+    // Recognized special forms for chars, sets: c-c s-s s&s
+
+    int32_t opts = RuleCharacterIterator::PARSE_VARIABLES |
+                   RuleCharacterIterator::PARSE_ESCAPES;
+    if ((options & USET_IGNORE_SPACE) != 0) {
+        opts |= RuleCharacterIterator::SKIP_WHITESPACE;
+    }
+
+    UnicodeString patLocal, buf;
+    UBool usePat = FALSE;
+    UnicodeSetPointer scratch;
+    RuleCharacterIterator::Pos backup;
+
+    // mode: 0=before [, 1=between [...], 2=after ]
+    // lastItem: 0=none, 1=char, 2=set
+    int8_t lastItem = 0, mode = 0;
+    UChar32 lastChar = 0;
+    UChar op = 0;
+
+    UBool invert = FALSE;
+
+    clear();
+
+    while (mode != 2 && !chars.atEnd()) {
+        U_ASSERT((lastItem == 0 && op == 0) ||
+                 (lastItem == 1 && (op == 0 || op == HYPHEN /*'-'*/)) ||
+                 (lastItem == 2 && (op == 0 || op == HYPHEN /*'-'*/ ||
+                                    op == INTERSECTION /*'&'*/)));
+
+        UChar32 c = 0;
+        UBool literal = FALSE;
+        UnicodeSet* nested = 0; // alias - do not delete
+
+        // -------- Check for property pattern
+
+        // setMode: 0=none, 1=unicodeset, 2=propertypat, 3=preparsed
+        int8_t setMode = 0;
+        if (resemblesPropertyPattern(chars, opts)) {
+            setMode = 2;
+        }
+
+        // -------- Parse '[' of opening delimiter OR nested set.
+        // If there is a nested set, use `setMode' to define how
+        // the set should be parsed.  If the '[' is part of the
+        // opening delimiter for this pattern, parse special
+        // strings "[", "[^", "[-", and "[^-".  Check for stand-in
+        // characters representing a nested set in the symbol
+        // table.
+
+        else {
+            // Prepare to backup if necessary
+            chars.getPos(backup);
+            c = chars.next(opts, literal, ec);
+            if (U_FAILURE(ec)) return;
+
+            if (c == 0x5B /*'['*/ && !literal) {
+                if (mode == 1) {
+                    chars.setPos(backup); // backup
+                    setMode = 1;
+                } else {
+                    // Handle opening '[' delimiter
+                    mode = 1;
+                    patLocal.append((UChar) 0x5B /*'['*/);
+                    chars.getPos(backup); // prepare to backup
+                    c = chars.next(opts, literal, ec); 
+                    if (U_FAILURE(ec)) return;
+                    if (c == 0x5E /*'^'*/ && !literal) {
+                        invert = TRUE;
+                        patLocal.append((UChar) 0x5E /*'^'*/);
+                        chars.getPos(backup); // prepare to backup
+                        c = chars.next(opts, literal, ec);
+                        if (U_FAILURE(ec)) return;
+                    }
+                    // Fall through to handle special leading '-';
+                    // otherwise restart loop for nested [], \p{}, etc.
+                    if (c == HYPHEN /*'-'*/) {
+                        literal = TRUE;
+                        // Fall through to handle literal '-' below
+                    } else {
+                        chars.setPos(backup); // backup
+                        continue;
+                    }
+                }
+            } else if (symbols != 0) {
+                const UnicodeFunctor *m = symbols->lookupMatcher(c);
+                if (m != 0) {
+                    const UnicodeSet *ms = dynamic_cast<const UnicodeSet *>(m);
+                    if (ms == NULL) {
+                        ec = U_MALFORMED_SET;
+                        return;
+                    }
+                    // casting away const, but `nested' won't be modified
+                    // (important not to modify stored set)
+                    nested = const_cast<UnicodeSet*>(ms);
+                    setMode = 3;
+                }
+            }
+        }
+
+        // -------- Handle a nested set.  This either is inline in
+        // the pattern or represented by a stand-in that has
+        // previously been parsed and was looked up in the symbol
+        // table.
+
+        if (setMode != 0) {
+            if (lastItem == 1) {
+                if (op != 0) {
+                    // syntaxError(chars, "Char expected after operator");
+                    ec = U_MALFORMED_SET;
+                    return;
+                }
+                add(lastChar, lastChar);
+                _appendToPat(patLocal, lastChar, FALSE);
+                lastItem = 0;
+                op = 0;
+            }
+
+            if (op == HYPHEN /*'-'*/ || op == INTERSECTION /*'&'*/) {
+                patLocal.append(op);
+            }
+
+            if (nested == 0) {
+                // lazy allocation
+                if (!scratch.allocate()) {
+                    ec = U_MEMORY_ALLOCATION_ERROR;
+                    return;
+                }
+                nested = scratch.pointer();
+            }
+            switch (setMode) {
+            case 1:
+                nested->applyPattern(chars, symbols, patLocal, options, caseClosure, depth + 1, ec);
+                break;
+            case 2:
+                chars.skipIgnored(opts);
+                nested->applyPropertyPattern(chars, patLocal, ec);
+                if (U_FAILURE(ec)) return;
+                break;
+            case 3: // `nested' already parsed
+                nested->_toPattern(patLocal, FALSE);
+                break;
+            }
+
+            usePat = TRUE;
+
+            if (mode == 0) {
+                // Entire pattern is a category; leave parse loop
+                *this = *nested;
+                mode = 2;
+                break;
+            }
+
+            switch (op) {
+            case HYPHEN: /*'-'*/
+                removeAll(*nested);
+                break;
+            case INTERSECTION: /*'&'*/
+                retainAll(*nested);
+                break;
+            case 0:
+                addAll(*nested);
+                break;
+            }
+
+            op = 0;
+            lastItem = 2;
+
+            continue;
+        }
+
+        if (mode == 0) {
+            // syntaxError(chars, "Missing '['");
+            ec = U_MALFORMED_SET;
+            return;
+        }
+
+        // -------- Parse special (syntax) characters.  If the
+        // current character is not special, or if it is escaped,
+        // then fall through and handle it below.
+
+        if (!literal) {
+            switch (c) {
+            case 0x5D /*']'*/:
+                if (lastItem == 1) {
+                    add(lastChar, lastChar);
+                    _appendToPat(patLocal, lastChar, FALSE);
+                }
+                // Treat final trailing '-' as a literal
+                if (op == HYPHEN /*'-'*/) {
+                    add(op, op);
+                    patLocal.append(op);
+                } else if (op == INTERSECTION /*'&'*/) {
+                    // syntaxError(chars, "Trailing '&'");
+                    ec = U_MALFORMED_SET;
+                    return;
+                }
+                patLocal.append((UChar) 0x5D /*']'*/);
+                mode = 2;
+                continue;
+            case HYPHEN /*'-'*/:
+                if (op == 0) {
+                    if (lastItem != 0) {
+                        op = (UChar) c;
+                        continue;
+                    } else {
+                        // Treat final trailing '-' as a literal
+                        add(c, c);
+                        c = chars.next(opts, literal, ec);
+                        if (U_FAILURE(ec)) return;
+                        if (c == 0x5D /*']'*/ && !literal) {
+                            patLocal.append(HYPHEN_RIGHT_BRACE, 2);
+                            mode = 2;
+                            continue;
+                        }
+                    }
+                }
+                // syntaxError(chars, "'-' not after char or set");
+                ec = U_MALFORMED_SET;
+                return;
+            case INTERSECTION /*'&'*/:
+                if (lastItem == 2 && op == 0) {
+                    op = (UChar) c;
+                    continue;
+                }
+                // syntaxError(chars, "'&' not after set");
+                ec = U_MALFORMED_SET;
+                return;
+            case 0x5E /*'^'*/:
+                // syntaxError(chars, "'^' not after '['");
+                ec = U_MALFORMED_SET;
+                return;
+            case 0x7B /*'{'*/:
+                if (op != 0) {
+                    // syntaxError(chars, "Missing operand after operator");
+                    ec = U_MALFORMED_SET;
+                    return;
+                }
+                if (lastItem == 1) {
+                    add(lastChar, lastChar);
+                    _appendToPat(patLocal, lastChar, FALSE);
+                }
+                lastItem = 0;
+                buf.truncate(0);
+                {
+                    UBool ok = FALSE;
+                    while (!chars.atEnd()) {
+                        c = chars.next(opts, literal, ec);
+                        if (U_FAILURE(ec)) return;
+                        if (c == 0x7D /*'}'*/ && !literal) {
+                            ok = TRUE;
+                            break;
+                        }
+                        buf.append(c);
+                    }
+                    if (buf.length() < 1 || !ok) {
+                        // syntaxError(chars, "Invalid multicharacter string");
+                        ec = U_MALFORMED_SET;
+                        return;
+                    }
+                }
+                // We have new string. Add it to set and continue;
+                // we don't need to drop through to the further
+                // processing
+                add(buf);
+                patLocal.append((UChar) 0x7B /*'{'*/);
+                _appendToPat(patLocal, buf, FALSE);
+                patLocal.append((UChar) 0x7D /*'}'*/);
+                continue;
+            case SymbolTable::SYMBOL_REF:
+                //         symbols  nosymbols
+                // [a-$]   error    error (ambiguous)
+                // [a$]    anchor   anchor
+                // [a-$x]  var "x"* literal '$'
+                // [a-$.]  error    literal '$'
+                // *We won't get here in the case of var "x"
+                {
+                    chars.getPos(backup);
+                    c = chars.next(opts, literal, ec);
+                    if (U_FAILURE(ec)) return;
+                    UBool anchor = (c == 0x5D /*']'*/ && !literal);
+                    if (symbols == 0 && !anchor) {
+                        c = SymbolTable::SYMBOL_REF;
+                        chars.setPos(backup);
+                        break; // literal '$'
+                    }
+                    if (anchor && op == 0) {
+                        if (lastItem == 1) {
+                            add(lastChar, lastChar);
+                            _appendToPat(patLocal, lastChar, FALSE);
+                        }
+                        add(U_ETHER);
+                        usePat = TRUE;
+                        patLocal.append((UChar) SymbolTable::SYMBOL_REF);
+                        patLocal.append((UChar) 0x5D /*']'*/);
+                        mode = 2;
+                        continue;
+                    }
+                    // syntaxError(chars, "Unquoted '$'");
+                    ec = U_MALFORMED_SET;
+                    return;
+                }
+            default:
+                break;
+            }
+        }
+
+        // -------- Parse literal characters.  This includes both
+        // escaped chars ("\u4E01") and non-syntax characters
+        // ("a").
+
+        switch (lastItem) {
+        case 0:
+            lastItem = 1;
+            lastChar = c;
+            break;
+        case 1:
+            if (op == HYPHEN /*'-'*/) {
+                if (lastChar >= c) {
+                    // Don't allow redundant (a-a) or empty (b-a) ranges;
+                    // these are most likely typos.
+                    // syntaxError(chars, "Invalid range");
+                    ec = U_MALFORMED_SET;
+                    return;
+                }
+                add(lastChar, c);
+                _appendToPat(patLocal, lastChar, FALSE);
+                patLocal.append(op);
+                _appendToPat(patLocal, c, FALSE);
+                lastItem = 0;
+                op = 0;
+            } else {
+                add(lastChar, lastChar);
+                _appendToPat(patLocal, lastChar, FALSE);
+                lastChar = c;
+            }
+            break;
+        case 2:
+            if (op != 0) {
+                // syntaxError(chars, "Set expected after operator");
+                ec = U_MALFORMED_SET;
+                return;
+            }
+            lastChar = c;
+            lastItem = 1;
+            break;
+        }
+    }
+
+    if (mode != 2) {
+        // syntaxError(chars, "Missing ']'");
+        ec = U_MALFORMED_SET;
+        return;
+    }
+
+    chars.skipIgnored(opts);
+
+    /**
+     * Handle global flags (invert, case insensitivity).  If this
+     * pattern should be compiled case-insensitive, then we need
+     * to close over case BEFORE COMPLEMENTING.  This makes
+     * patterns like /[^abc]/i work.
+     */
+    if ((options & USET_CASE_INSENSITIVE) != 0) {
+        (this->*caseClosure)(USET_CASE_INSENSITIVE);
+    }
+    else if ((options & USET_ADD_CASE_MAPPINGS) != 0) {
+        (this->*caseClosure)(USET_ADD_CASE_MAPPINGS);
+    }
+    if (invert) {
+        complement();
+    }
+
+    // Use the rebuilt pattern (patLocal) only if necessary.  Prefer the
+    // generated pattern.
+    if (usePat) {
+        rebuiltPat.append(patLocal);
+    } else {
+        _generatePattern(rebuiltPat, FALSE);
+    }
+    if (isBogus() && U_SUCCESS(ec)) {
+        // We likely ran out of memory. AHHH!
+        ec = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+//----------------------------------------------------------------
+// Property set implementation
+//----------------------------------------------------------------
+
+namespace {
+
+static UBool numericValueFilter(UChar32 ch, void* context) {
+    return u_getNumericValue(ch) == *(double*)context;
+}
+
+static UBool generalCategoryMaskFilter(UChar32 ch, void* context) {
+    int32_t value = *(int32_t*)context;
+    return (U_GET_GC_MASK((UChar32) ch) & value) != 0;
+}
+
+static UBool versionFilter(UChar32 ch, void* context) {
+    static const UVersionInfo none = { 0, 0, 0, 0 };
+    UVersionInfo v;
+    u_charAge(ch, v);
+    UVersionInfo* version = (UVersionInfo*)context;
+    return uprv_memcmp(&v, &none, sizeof(v)) > 0 && uprv_memcmp(&v, version, sizeof(v)) <= 0;
+}
+
+typedef struct {
+    UProperty prop;
+    int32_t value;
+} IntPropertyContext;
+
+static UBool intPropertyFilter(UChar32 ch, void* context) {
+    IntPropertyContext* c = (IntPropertyContext*)context;
+    return u_getIntPropertyValue((UChar32) ch, c->prop) == c->value;
+}
+
+static UBool scriptExtensionsFilter(UChar32 ch, void* context) {
+    return uscript_hasScript(ch, *(UScriptCode*)context);
+}
+
+}  // namespace
+
+/**
+ * Generic filter-based scanning code for UCD property UnicodeSets.
+ */
+void UnicodeSet::applyFilter(UnicodeSet::Filter filter,
+                             void* context,
+                             const UnicodeSet* inclusions,
+                             UErrorCode &status) {
+    if (U_FAILURE(status)) return;
+
+    // Logically, walk through all Unicode characters, noting the start
+    // and end of each range for which filter.contain(c) is
+    // true.  Add each range to a set.
+    //
+    // To improve performance, use an inclusions set which
+    // encodes information about character ranges that are known
+    // to have identical properties.
+    // inclusions contains the first characters of
+    // same-value ranges for the given property.
+
+    clear();
+
+    UChar32 startHasProperty = -1;
+    int32_t limitRange = inclusions->getRangeCount();
+
+    for (int j=0; j<limitRange; ++j) {
+        // get current range
+        UChar32 start = inclusions->getRangeStart(j);
+        UChar32 end = inclusions->getRangeEnd(j);
+
+        // for all the code points in the range, process
+        for (UChar32 ch = start; ch <= end; ++ch) {
+            // only add to this UnicodeSet on inflection points --
+            // where the hasProperty value changes to false
+            if ((*filter)(ch, context)) {
+                if (startHasProperty < 0) {
+                    startHasProperty = ch;
+                }
+            } else if (startHasProperty >= 0) {
+                add(startHasProperty, ch-1);
+                startHasProperty = -1;
+            }
+        }
+    }
+    if (startHasProperty >= 0) {
+        add((UChar32)startHasProperty, (UChar32)0x10FFFF);
+    }
+    if (isBogus() && U_SUCCESS(status)) {
+        // We likely ran out of memory. AHHH!
+        status = U_MEMORY_ALLOCATION_ERROR;
+    }
+}
+
+namespace {
+
+static UBool mungeCharName(char* dst, const char* src, int32_t dstCapacity) {
+    /* Note: we use ' ' in compiler code page */
+    int32_t j = 0;
+    char ch;
+    --dstCapacity; /* make room for term. zero */
+    while ((ch = *src++) != 0) {
+        if (ch == ' ' && (j==0 || (j>0 && dst[j-1]==' '))) {
+            continue;
+        }
+        if (j >= dstCapacity) return FALSE;
+        dst[j++] = ch;
+    }
+    if (j > 0 && dst[j-1] == ' ') --j;
+    dst[j] = 0;
+    return TRUE;
+}
+
+}  // namespace
+
+//----------------------------------------------------------------
+// Property set API
+//----------------------------------------------------------------
+
+#define FAIL(ec) UPRV_BLOCK_MACRO_BEGIN { \
+    ec=U_ILLEGAL_ARGUMENT_ERROR; \
+    return *this; \
+} UPRV_BLOCK_MACRO_END
+
+UnicodeSet&
+UnicodeSet::applyIntPropertyValue(UProperty prop, int32_t value, UErrorCode& ec) {
+    if (U_FAILURE(ec) || isFrozen()) { return *this; }
+    if (prop == UCHAR_GENERAL_CATEGORY_MASK) {
+        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
+        applyFilter(generalCategoryMaskFilter, &value, inclusions, ec);
+    } else if (prop == UCHAR_SCRIPT_EXTENSIONS) {
+        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
+        UScriptCode script = (UScriptCode)value;
+        applyFilter(scriptExtensionsFilter, &script, inclusions, ec);
+    } else if (0 <= prop && prop < UCHAR_BINARY_LIMIT) {
+        if (value == 0 || value == 1) {
+            const USet *set = u_getBinaryPropertySet(prop, &ec);
+            if (U_FAILURE(ec)) { return *this; }
+            copyFrom(*UnicodeSet::fromUSet(set), TRUE);
+            if (value == 0) {
+                complement();
+            }
+        } else {
+            clear();
+        }
+    } else if (UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT) {
+        const UnicodeSet* inclusions = CharacterProperties::getInclusionsForProperty(prop, ec);
+        IntPropertyContext c = {prop, value};
+        applyFilter(intPropertyFilter, &c, inclusions, ec);
+    } else {
+        ec = U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    return *this;
+}
+
+UnicodeSet&
+UnicodeSet::applyPropertyAlias(const UnicodeString& prop,
+                               const UnicodeString& value,
+                               UErrorCode& ec) {
+    if (U_FAILURE(ec) || isFrozen()) return *this;
+
+    // prop and value used to be converted to char * using the default
+    // converter instead of the invariant conversion.
+    // This should not be necessary because all Unicode property and value
+    // names use only invariant characters.
+    // If there are any variant characters, then we won't find them anyway.
+    // Checking first avoids assertion failures in the conversion.
+    if( !uprv_isInvariantUString(prop.getBuffer(), prop.length()) ||
+        !uprv_isInvariantUString(value.getBuffer(), value.length())
+    ) {
+        FAIL(ec);
+    }
+    CharString pname, vname;
+    pname.appendInvariantChars(prop, ec);
+    vname.appendInvariantChars(value, ec);
+    if (U_FAILURE(ec)) return *this;
+
+    UProperty p;
+    int32_t v;
+    UBool invert = FALSE;
+
+    if (value.length() > 0) {
+        p = u_getPropertyEnum(pname.data());
+        if (p == UCHAR_INVALID_CODE) FAIL(ec);
+
+        // Treat gc as gcm
+        if (p == UCHAR_GENERAL_CATEGORY) {
+            p = UCHAR_GENERAL_CATEGORY_MASK;
+        }
+
+        if ((p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) ||
+            (p >= UCHAR_INT_START && p < UCHAR_INT_LIMIT) ||
+            (p >= UCHAR_MASK_START && p < UCHAR_MASK_LIMIT)) {
+            v = u_getPropertyValueEnum(p, vname.data());
+            if (v == UCHAR_INVALID_CODE) {
+                // Handle numeric CCC
+                if (p == UCHAR_CANONICAL_COMBINING_CLASS ||
+                    p == UCHAR_TRAIL_CANONICAL_COMBINING_CLASS ||
+                    p == UCHAR_LEAD_CANONICAL_COMBINING_CLASS) {
+                    char* end;
+                    double val = uprv_strtod(vname.data(), &end);
+                    // Anything between 0 and 255 is valid even if unused.
+                    // Cast double->int only after range check.
+                    // We catch NaN here because comparing it with both 0 and 255 will be false
+                    // (as are all comparisons with NaN).
+                    if (*end != 0 || !(0 <= val && val <= 255) ||
+                            (v = (int32_t)val) != val) {
+                        // non-integral value or outside 0..255, or trailing junk
+                        FAIL(ec);
+                    }
+                } else {
+                    FAIL(ec);
+                }
+            }
+        }
+
+        else {
+
+            switch (p) {
+            case UCHAR_NUMERIC_VALUE:
+                {
+                    char* end;
+                    double val = uprv_strtod(vname.data(), &end);
+                    if (*end != 0) {
+                        FAIL(ec);
+                    }
+                    applyFilter(numericValueFilter, &val,
+                                CharacterProperties::getInclusionsForProperty(p, ec), ec);
+                    return *this;
+                }
+            case UCHAR_NAME:
+                {
+                    // Must munge name, since u_charFromName() does not do
+                    // 'loose' matching.
+                    char buf[128]; // it suffices that this be > uprv_getMaxCharNameLength
+                    if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
+                    UChar32 ch = u_charFromName(U_EXTENDED_CHAR_NAME, buf, &ec);
+                    if (U_SUCCESS(ec)) {
+                        clear();
+                        add(ch);
+                        return *this;
+                    } else {
+                        FAIL(ec);
+                    }
+                }
+            case UCHAR_UNICODE_1_NAME:
+                // ICU 49 deprecates the Unicode_1_Name property APIs.
+                FAIL(ec);
+            case UCHAR_AGE:
+                {
+                    // Must munge name, since u_versionFromString() does not do
+                    // 'loose' matching.
+                    char buf[128];
+                    if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
+                    UVersionInfo version;
+                    u_versionFromString(version, buf);
+                    applyFilter(versionFilter, &version,
+                                CharacterProperties::getInclusionsForProperty(p, ec), ec);
+                    return *this;
+                }
+            case UCHAR_SCRIPT_EXTENSIONS:
+                v = u_getPropertyValueEnum(UCHAR_SCRIPT, vname.data());
+                if (v == UCHAR_INVALID_CODE) {
+                    FAIL(ec);
+                }
+                // fall through to calling applyIntPropertyValue()
+                break;
+            default:
+                // p is a non-binary, non-enumerated property that we
+                // don't support (yet).
+                FAIL(ec);
+            }
+        }
+    }
+
+    else {
+        // value is empty.  Interpret as General Category, Script, or
+        // Binary property.
+        p = UCHAR_GENERAL_CATEGORY_MASK;
+        v = u_getPropertyValueEnum(p, pname.data());
+        if (v == UCHAR_INVALID_CODE) {
+            p = UCHAR_SCRIPT;
+            v = u_getPropertyValueEnum(p, pname.data());
+            if (v == UCHAR_INVALID_CODE) {
+                p = u_getPropertyEnum(pname.data());
+                if (p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) {
+                    v = 1;
+                } else if (0 == uprv_comparePropertyNames(ANY, pname.data())) {
+                    set(MIN_VALUE, MAX_VALUE);
+                    return *this;
+                } else if (0 == uprv_comparePropertyNames(ASCII, pname.data())) {
+                    set(0, 0x7F);
+                    return *this;
+                } else if (0 == uprv_comparePropertyNames(ASSIGNED, pname.data())) {
+                    // [:Assigned:]=[:^Cn:]
+                    p = UCHAR_GENERAL_CATEGORY_MASK;
+                    v = U_GC_CN_MASK;
+                    invert = TRUE;
+                } else {
+                    FAIL(ec);
+                }
+            }
+        }
+    }
+
+    applyIntPropertyValue(p, v, ec);
+    if(invert) {
+        complement();
+    }
+
+    if (isBogus() && U_SUCCESS(ec)) {
+        // We likely ran out of memory. AHHH!
+        ec = U_MEMORY_ALLOCATION_ERROR;
+    }
+    return *this;
+}
+
+//----------------------------------------------------------------
+// Property set patterns
+//----------------------------------------------------------------
+
+/**
+ * Return true if the given position, in the given pattern, appears
+ * to be the start of a property set pattern.
+ */
+UBool UnicodeSet::resemblesPropertyPattern(const UnicodeString& pattern,
+                                           int32_t pos) {
+    // Patterns are at least 5 characters long
+    if ((pos+5) > pattern.length()) {
+        return FALSE;
+    }
+
+    // Look for an opening [:, [:^, \p, or \P
+    return isPOSIXOpen(pattern, pos) || isPerlOpen(pattern, pos) || isNameOpen(pattern, pos);
+}
+
+/**
+ * Return true if the given iterator appears to point at a
+ * property pattern.  Regardless of the result, return with the
+ * iterator unchanged.
+ * @param chars iterator over the pattern characters.  Upon return
+ * it will be unchanged.
+ * @param iterOpts RuleCharacterIterator options
+ */
+UBool UnicodeSet::resemblesPropertyPattern(RuleCharacterIterator& chars,
+                                           int32_t iterOpts) {
+    // NOTE: literal will always be FALSE, because we don't parse escapes.
+    UBool result = FALSE, literal;
+    UErrorCode ec = U_ZERO_ERROR;
+    iterOpts &= ~RuleCharacterIterator::PARSE_ESCAPES;
+    RuleCharacterIterator::Pos pos;
+    chars.getPos(pos);
+    UChar32 c = chars.next(iterOpts, literal, ec);
+    if (c == 0x5B /*'['*/ || c == 0x5C /*'\\'*/) {
+        UChar32 d = chars.next(iterOpts & ~RuleCharacterIterator::SKIP_WHITESPACE,
+                               literal, ec);
+        result = (c == 0x5B /*'['*/) ? (d == 0x3A /*':'*/) :
+                 (d == 0x4E /*'N'*/ || d == 0x70 /*'p'*/ || d == 0x50 /*'P'*/);
+    }
+    chars.setPos(pos);
+    return result && U_SUCCESS(ec);
+}
+
+/**
+ * Parse the given property pattern at the given parse position.
+ */
+UnicodeSet& UnicodeSet::applyPropertyPattern(const UnicodeString& pattern,
+                                             ParsePosition& ppos,
+                                             UErrorCode &ec) {
+    int32_t pos = ppos.getIndex();
+
+    UBool posix = FALSE; // true for [:pat:], false for \p{pat} \P{pat} \N{pat}
+    UBool isName = FALSE; // true for \N{pat}, o/w false
+    UBool invert = FALSE;
+
+    if (U_FAILURE(ec)) return *this;
+
+    // Minimum length is 5 characters, e.g. \p{L}
+    if ((pos+5) > pattern.length()) {
+        FAIL(ec);
+    }
+
+    // On entry, ppos should point to one of the following locations:
+    // Look for an opening [:, [:^, \p, or \P
+    if (isPOSIXOpen(pattern, pos)) {
+        posix = TRUE;
+        pos += 2;
+        pos = ICU_Utility::skipWhitespace(pattern, pos);
+        if (pos < pattern.length() && pattern.charAt(pos) == COMPLEMENT) {
+            ++pos;
+            invert = TRUE;
+        }
+    } else if (isPerlOpen(pattern, pos) || isNameOpen(pattern, pos)) {
+        UChar c = pattern.charAt(pos+1);
+        invert = (c == UPPER_P);
+        isName = (c == UPPER_N);
+        pos += 2;
+        pos = ICU_Utility::skipWhitespace(pattern, pos);
+        if (pos == pattern.length() || pattern.charAt(pos++) != OPEN_BRACE) {
+            // Syntax error; "\p" or "\P" not followed by "{"
+            FAIL(ec);
+        }
+    } else {
+        // Open delimiter not seen
+        FAIL(ec);
+    }
+
+    // Look for the matching close delimiter, either :] or }
+    int32_t close;
+    if (posix) {
+      close = pattern.indexOf(POSIX_CLOSE, 2, pos);
+    } else {
+      close = pattern.indexOf(CLOSE_BRACE, pos);
+    }
+    if (close < 0) {
+        // Syntax error; close delimiter missing
+        FAIL(ec);
+    }
+
+    // Look for an '=' sign.  If this is present, we will parse a
+    // medium \p{gc=Cf} or long \p{GeneralCategory=Format}
+    // pattern.
+    int32_t equals = pattern.indexOf(EQUALS, pos);
+    UnicodeString propName, valueName;
+    if (equals >= 0 && equals < close && !isName) {
+        // Equals seen; parse medium/long pattern
+        pattern.extractBetween(pos, equals, propName);
+        pattern.extractBetween(equals+1, close, valueName);
+    }
+
+    else {
+        // Handle case where no '=' is seen, and \N{}
+        pattern.extractBetween(pos, close, propName);
+            
+        // Handle \N{name}
+        if (isName) {
+            // This is a little inefficient since it means we have to
+            // parse NAME_PROP back to UCHAR_NAME even though we already
+            // know it's UCHAR_NAME.  If we refactor the API to
+            // support args of (UProperty, char*) then we can remove
+            // NAME_PROP and make this a little more efficient.
+            valueName = propName;
+            propName = UnicodeString(NAME_PROP, NAME_PROP_LENGTH, US_INV);
+        }
+    }
+
+    applyPropertyAlias(propName, valueName, ec);
+
+    if (U_SUCCESS(ec)) {
+        if (invert) {
+            complement();
+        }
+            
+        // Move to the limit position after the close delimiter if the
+        // parse succeeded.
+        ppos.setIndex(close + (posix ? 2 : 1));
+    }
+
+    return *this;
+}
+
+/**
+ * Parse a property pattern.
+ * @param chars iterator over the pattern characters.  Upon return
+ * it will be advanced to the first character after the parsed
+ * pattern, or the end of the iteration if all characters are
+ * parsed.
+ * @param rebuiltPat the pattern that was parsed, rebuilt or
+ * copied from the input pattern, as appropriate.
+ */
+void UnicodeSet::applyPropertyPattern(RuleCharacterIterator& chars,
+                                      UnicodeString& rebuiltPat,
+                                      UErrorCode& ec) {
+    if (U_FAILURE(ec)) return;
+    UnicodeString pattern;
+    chars.lookahead(pattern);
+    ParsePosition pos(0);
+    applyPropertyPattern(pattern, pos, ec);
+    if (U_FAILURE(ec)) return;
+    if (pos.getIndex() == 0) {
+        // syntaxError(chars, "Invalid property pattern");
+        ec = U_MALFORMED_SET;
+        return;
+    }
+    chars.jumpahead(pos.getIndex());
+    rebuiltPat.append(pattern, 0, pos.getIndex());
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/unisetspan.cpp b/thirdparty/icu4c/common/unisetspan.cpp
new file mode 100644
index 0000000000..68e44d91ee
--- /dev/null
+++ b/thirdparty/icu4c/common/unisetspan.cpp
@@ -0,0 +1,1509 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2007-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  unisetspan.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2007mar01
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+#include "unicode/ustring.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "uvector.h"
+#include "unisetspan.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * List of offsets from the current position from where to try matching
+ * a code point or a string.
+ * Store offsets rather than indexes to simplify the code and use the same list
+ * for both increments (in span()) and decrements (in spanBack()).
+ *
+ * Assumption: The maximum offset is limited, and the offsets that are stored
+ * at any one time are relatively dense, that is, there are normally no gaps of
+ * hundreds or thousands of offset values.
+ *
+ * The implementation uses a circular buffer of byte flags,
+ * each indicating whether the corresponding offset is in the list.
+ * This avoids inserting into a sorted list of offsets (or absolute indexes) and
+ * physically moving part of the list.
+ *
+ * Note: In principle, the caller should setMaxLength() to the maximum of the
+ * max string length and U16_LENGTH/U8_LENGTH to account for
+ * "long" single code points.
+ * However, this implementation uses at least a staticList with more than
+ * U8_LENGTH entries anyway.
+ *
+ * Note: If maxLength were guaranteed to be no more than 32 or 64,
+ * the list could be stored as bit flags in a single integer.
+ * Rather than handling a circular buffer with a start list index,
+ * the integer would simply be shifted when lower offsets are removed.
+ * UnicodeSet does not have a limit on the lengths of strings.
+ */
+class OffsetList {  // Only ever stack-allocated, does not need to inherit UMemory.
+public:
+    OffsetList() : list(staticList), capacity(0), length(0), start(0) {}
+
+    ~OffsetList() {
+        if(list!=staticList) {
+            uprv_free(list);
+        }
+    }
+
+    // Call exactly once if the list is to be used.
+    void setMaxLength(int32_t maxLength) {
+        if(maxLength<=(int32_t)sizeof(staticList)) {
+            capacity=(int32_t)sizeof(staticList);
+        } else {
+            UBool *l=(UBool *)uprv_malloc(maxLength);
+            if(l!=NULL) {
+                list=l;
+                capacity=maxLength;
+            }
+        }
+        uprv_memset(list, 0, capacity);
+    }
+
+    void clear() {
+        uprv_memset(list, 0, capacity);
+        start=length=0;
+    }
+
+    UBool isEmpty() const {
+        return (UBool)(length==0);
+    }
+
+    // Reduce all stored offsets by delta, used when the current position
+    // moves by delta.
+    // There must not be any offsets lower than delta.
+    // If there is an offset equal to delta, it is removed.
+    // delta=[1..maxLength]
+    void shift(int32_t delta) {
+        int32_t i=start+delta;
+        if(i>=capacity) {
+            i-=capacity;
+        }
+        if(list[i]) {
+            list[i]=FALSE;
+            --length;
+        }
+        start=i;
+    }
+
+    // Add an offset. The list must not contain it yet.
+    // offset=[1..maxLength]
+    void addOffset(int32_t offset) {
+        int32_t i=start+offset;
+        if(i>=capacity) {
+            i-=capacity;
+        }
+        list[i]=TRUE;
+        ++length;
+    }
+
+    // offset=[1..maxLength]
+    UBool containsOffset(int32_t offset) const {
+        int32_t i=start+offset;
+        if(i>=capacity) {
+            i-=capacity;
+        }
+        return list[i];
+    }
+
+    // Find the lowest stored offset from a non-empty list, remove it,
+    // and reduce all other offsets by this minimum.
+    // Returns [1..maxLength].
+    int32_t popMinimum() {
+        // Look for the next offset in list[start+1..capacity-1].
+        int32_t i=start, result;
+        while(++i<capacity) {
+            if(list[i]) {
+                list[i]=FALSE;
+                --length;
+                result=i-start;
+                start=i;
+                return result;
+            }
+        }
+        // i==capacity
+
+        // Wrap around and look for the next offset in list[0..start].
+        // Since the list is not empty, there will be one.
+        result=capacity-start;
+        i=0;
+        while(!list[i]) {
+            ++i;
+        }
+        list[i]=FALSE;
+        --length;
+        start=i;
+        return result+=i;
+    }
+
+private:
+    UBool *list;
+    int32_t capacity;
+    int32_t length;
+    int32_t start;
+
+    UBool staticList[16];
+};
+
+// Get the number of UTF-8 bytes for a UTF-16 (sub)string.
+static int32_t
+getUTF8Length(const UChar *s, int32_t length) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length8=0;
+    u_strToUTF8(NULL, 0, &length8, s, length, &errorCode);
+    if(U_SUCCESS(errorCode) || errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        return length8;
+    } else {
+        // The string contains an unpaired surrogate.
+        // Ignore this string.
+        return 0;
+    }
+}
+
+// Append the UTF-8 version of the string to t and return the appended UTF-8 length.
+static int32_t
+appendUTF8(const UChar *s, int32_t length, uint8_t *t, int32_t capacity) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t length8=0;
+    u_strToUTF8((char *)t, capacity, &length8, s, length, &errorCode);
+    if(U_SUCCESS(errorCode)) {
+        return length8;
+    } else {
+        // The string contains an unpaired surrogate.
+        // Ignore this string.
+        return 0;
+    }
+}
+
+static inline uint8_t
+makeSpanLengthByte(int32_t spanLength) {
+    // 0xfe==UnicodeSetStringSpan::LONG_SPAN
+    return spanLength<0xfe ? (uint8_t)spanLength : (uint8_t)0xfe;
+}
+
+// Construct for all variants of span(), or only for any one variant.
+// Initialize as little as possible, for single use.
+UnicodeSetStringSpan::UnicodeSetStringSpan(const UnicodeSet &set,
+                                           const UVector &setStrings,
+                                           uint32_t which)
+        : spanSet(0, 0x10ffff), pSpanNotSet(NULL), strings(setStrings),
+          utf8Lengths(NULL), spanLengths(NULL), utf8(NULL),
+          utf8Length(0),
+          maxLength16(0), maxLength8(0),
+          all((UBool)(which==ALL)) {
+    spanSet.retainAll(set);
+    if(which&NOT_CONTAINED) {
+        // Default to the same sets.
+        // addToSpanNotSet() will create a separate set if necessary.
+        pSpanNotSet=&spanSet;
+    }
+
+    // Determine if the strings even need to be taken into account at all for span() etc.
+    // If any string is relevant, then all strings need to be used for
+    // span(longest match) but only the relevant ones for span(while contained).
+    // TODO: Possible optimization: Distinguish CONTAINED vs. LONGEST_MATCH
+    //   and do not store UTF-8 strings if !thisRelevant and CONTAINED.
+    //   (Only store irrelevant UTF-8 strings for LONGEST_MATCH where they are relevant after all.)
+    // Also count the lengths of the UTF-8 versions of the strings for memory allocation.
+    int32_t stringsLength=strings.size();
+
+    int32_t i, spanLength;
+    UBool someRelevant=FALSE;
+    for(i=0; i<stringsLength; ++i) {
+        const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+        const UChar *s16=string.getBuffer();
+        int32_t length16=string.length();
+        UBool thisRelevant;
+        spanLength=spanSet.span(s16, length16, USET_SPAN_CONTAINED);
+        if(spanLength<length16) {  // Relevant string.
+            someRelevant=thisRelevant=TRUE;
+        } else {
+            thisRelevant=FALSE;
+        }
+        if((which&UTF16) && length16>maxLength16) {
+            maxLength16=length16;
+        }
+        if((which&UTF8) && (thisRelevant || (which&CONTAINED))) {
+            int32_t length8=getUTF8Length(s16, length16);
+            utf8Length+=length8;
+            if(length8>maxLength8) {
+                maxLength8=length8;
+            }
+        }
+    }
+    if(!someRelevant) {
+        maxLength16=maxLength8=0;
+        return;
+    }
+
+    // Freeze after checking for the need to use strings at all because freezing
+    // a set takes some time and memory which are wasted if there are no relevant strings.
+    if(all) {
+        spanSet.freeze();
+    }
+
+    uint8_t *spanBackLengths;
+    uint8_t *spanUTF8Lengths;
+    uint8_t *spanBackUTF8Lengths;
+
+    // Allocate a block of meta data.
+    int32_t allocSize;
+    if(all) {
+        // UTF-8 lengths, 4 sets of span lengths, UTF-8 strings.
+        allocSize=stringsLength*(4+1+1+1+1)+utf8Length;
+    } else {
+        allocSize=stringsLength;  // One set of span lengths.
+        if(which&UTF8) {
+            // UTF-8 lengths and UTF-8 strings.
+            allocSize+=stringsLength*4+utf8Length;
+        }
+    }
+    if(allocSize<=(int32_t)sizeof(staticLengths)) {
+        utf8Lengths=staticLengths;
+    } else {
+        utf8Lengths=(int32_t *)uprv_malloc(allocSize);
+        if(utf8Lengths==NULL) {
+            maxLength16=maxLength8=0;  // Prevent usage by making needsStringSpanUTF16/8() return FALSE.
+            return;  // Out of memory.
+        }
+    }
+
+    if(all) {
+        // Store span lengths for all span() variants.
+        spanLengths=(uint8_t *)(utf8Lengths+stringsLength);
+        spanBackLengths=spanLengths+stringsLength;
+        spanUTF8Lengths=spanBackLengths+stringsLength;
+        spanBackUTF8Lengths=spanUTF8Lengths+stringsLength;
+        utf8=spanBackUTF8Lengths+stringsLength;
+    } else {
+        // Store span lengths for only one span() variant.
+        if(which&UTF8) {
+            spanLengths=(uint8_t *)(utf8Lengths+stringsLength);
+            utf8=spanLengths+stringsLength;
+        } else {
+            spanLengths=(uint8_t *)utf8Lengths;
+        }
+        spanBackLengths=spanUTF8Lengths=spanBackUTF8Lengths=spanLengths;
+    }
+
+    // Set the meta data and pSpanNotSet and write the UTF-8 strings.
+    int32_t utf8Count=0;  // Count UTF-8 bytes written so far.
+
+    for(i=0; i<stringsLength; ++i) {
+        const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+        const UChar *s16=string.getBuffer();
+        int32_t length16=string.length();
+        spanLength=spanSet.span(s16, length16, USET_SPAN_CONTAINED);
+        if(spanLength<length16) {  // Relevant string.
+            if(which&UTF16) {
+                if(which&CONTAINED) {
+                    if(which&FWD) {
+                        spanLengths[i]=makeSpanLengthByte(spanLength);
+                    }
+                    if(which&BACK) {
+                        spanLength=length16-spanSet.spanBack(s16, length16, USET_SPAN_CONTAINED);
+                        spanBackLengths[i]=makeSpanLengthByte(spanLength);
+                    }
+                } else /* not CONTAINED, not all, but NOT_CONTAINED */ {
+                    spanLengths[i]=spanBackLengths[i]=0;  // Only store a relevant/irrelevant flag.
+                }
+            }
+            if(which&UTF8) {
+                uint8_t *s8=utf8+utf8Count;
+                int32_t length8=appendUTF8(s16, length16, s8, utf8Length-utf8Count);
+                utf8Count+=utf8Lengths[i]=length8;
+                if(length8==0) {  // Irrelevant for UTF-8 because not representable in UTF-8.
+                    spanUTF8Lengths[i]=spanBackUTF8Lengths[i]=(uint8_t)ALL_CP_CONTAINED;
+                } else {  // Relevant for UTF-8.
+                    if(which&CONTAINED) {
+                        if(which&FWD) {
+                            spanLength=spanSet.spanUTF8((const char *)s8, length8, USET_SPAN_CONTAINED);
+                            spanUTF8Lengths[i]=makeSpanLengthByte(spanLength);
+                        }
+                        if(which&BACK) {
+                            spanLength=length8-spanSet.spanBackUTF8((const char *)s8, length8, USET_SPAN_CONTAINED);
+                            spanBackUTF8Lengths[i]=makeSpanLengthByte(spanLength);
+                        }
+                    } else /* not CONTAINED, not all, but NOT_CONTAINED */ {
+                        spanUTF8Lengths[i]=spanBackUTF8Lengths[i]=0;  // Only store a relevant/irrelevant flag.
+                    }
+                }
+            }
+            if(which&NOT_CONTAINED) {
+                // Add string start and end code points to the spanNotSet so that
+                // a span(while not contained) stops before any string.
+                UChar32 c;
+                if(which&FWD) {
+                    int32_t len=0;
+                    U16_NEXT(s16, len, length16, c);
+                    addToSpanNotSet(c);
+                }
+                if(which&BACK) {
+                    int32_t len=length16;
+                    U16_PREV(s16, 0, len, c);
+                    addToSpanNotSet(c);
+                }
+            }
+        } else {  // Irrelevant string.
+            if(which&UTF8) {
+                if(which&CONTAINED) {  // Only necessary for LONGEST_MATCH.
+                    uint8_t *s8=utf8+utf8Count;
+                    int32_t length8=appendUTF8(s16, length16, s8, utf8Length-utf8Count);
+                    utf8Count+=utf8Lengths[i]=length8;
+                } else {
+                    utf8Lengths[i]=0;
+                }
+            }
+            if(all) {
+                spanLengths[i]=spanBackLengths[i]=
+                    spanUTF8Lengths[i]=spanBackUTF8Lengths[i]=
+                        (uint8_t)ALL_CP_CONTAINED;
+            } else {
+                // All spanXYZLengths pointers contain the same address.
+                spanLengths[i]=(uint8_t)ALL_CP_CONTAINED;
+            }
+        }
+    }
+
+    // Finish.
+    if(all) {
+        pSpanNotSet->freeze();
+    }
+}
+
+// Copy constructor. Assumes which==ALL for a frozen set.
+UnicodeSetStringSpan::UnicodeSetStringSpan(const UnicodeSetStringSpan &otherStringSpan,
+                                           const UVector &newParentSetStrings)
+        : spanSet(otherStringSpan.spanSet), pSpanNotSet(NULL), strings(newParentSetStrings),
+          utf8Lengths(NULL), spanLengths(NULL), utf8(NULL),
+          utf8Length(otherStringSpan.utf8Length),
+          maxLength16(otherStringSpan.maxLength16), maxLength8(otherStringSpan.maxLength8),
+          all(TRUE) {
+    if(otherStringSpan.pSpanNotSet==&otherStringSpan.spanSet) {
+        pSpanNotSet=&spanSet;
+    } else {
+        pSpanNotSet=otherStringSpan.pSpanNotSet->clone();
+    }
+
+    // Allocate a block of meta data.
+    // UTF-8 lengths, 4 sets of span lengths, UTF-8 strings.
+    int32_t stringsLength=strings.size();
+    int32_t allocSize=stringsLength*(4+1+1+1+1)+utf8Length;
+    if(allocSize<=(int32_t)sizeof(staticLengths)) {
+        utf8Lengths=staticLengths;
+    } else {
+        utf8Lengths=(int32_t *)uprv_malloc(allocSize);
+        if(utf8Lengths==NULL) {
+            maxLength16=maxLength8=0;  // Prevent usage by making needsStringSpanUTF16/8() return FALSE.
+            return;  // Out of memory.
+        }
+    }
+
+    spanLengths=(uint8_t *)(utf8Lengths+stringsLength);
+    utf8=spanLengths+stringsLength*4;
+    uprv_memcpy(utf8Lengths, otherStringSpan.utf8Lengths, allocSize);
+}
+
+UnicodeSetStringSpan::~UnicodeSetStringSpan() {
+    if(pSpanNotSet!=NULL && pSpanNotSet!=&spanSet) {
+        delete pSpanNotSet;
+    }
+    if(utf8Lengths!=NULL && utf8Lengths!=staticLengths) {
+        uprv_free(utf8Lengths);
+    }
+}
+
+void UnicodeSetStringSpan::addToSpanNotSet(UChar32 c) {
+    if(pSpanNotSet==NULL || pSpanNotSet==&spanSet) {
+        if(spanSet.contains(c)) {
+            return;  // Nothing to do.
+        }
+        UnicodeSet *newSet=spanSet.cloneAsThawed();
+        if(newSet==NULL) {
+            return;  // Out of memory.
+        } else {
+            pSpanNotSet=newSet;
+        }
+    }
+    pSpanNotSet->add(c);
+}
+
+// Compare strings without any argument checks. Requires length>0.
+static inline UBool
+matches16(const UChar *s, const UChar *t, int32_t length) {
+    do {
+        if(*s++!=*t++) {
+            return FALSE;
+        }
+    } while(--length>0);
+    return TRUE;
+}
+
+static inline UBool
+matches8(const uint8_t *s, const uint8_t *t, int32_t length) {
+    do {
+        if(*s++!=*t++) {
+            return FALSE;
+        }
+    } while(--length>0);
+    return TRUE;
+}
+
+// Compare 16-bit Unicode strings (which may be malformed UTF-16)
+// at code point boundaries.
+// That is, each edge of a match must not be in the middle of a surrogate pair.
+static inline UBool
+matches16CPB(const UChar *s, int32_t start, int32_t limit, const UChar *t, int32_t length) {
+    s+=start;
+    limit-=start;
+    return matches16(s, t, length) &&
+           !(0<start && U16_IS_LEAD(s[-1]) && U16_IS_TRAIL(s[0])) &&
+           !(length<limit && U16_IS_LEAD(s[length-1]) && U16_IS_TRAIL(s[length]));
+}
+
+// Does the set contain the next code point?
+// If so, return its length; otherwise return its negative length.
+static inline int32_t
+spanOne(const UnicodeSet &set, const UChar *s, int32_t length) {
+    UChar c=*s, c2;
+    if(c>=0xd800 && c<=0xdbff && length>=2 && U16_IS_TRAIL(c2=s[1])) {
+        return set.contains(U16_GET_SUPPLEMENTARY(c, c2)) ? 2 : -2;
+    }
+    return set.contains(c) ? 1 : -1;
+}
+
+static inline int32_t
+spanOneBack(const UnicodeSet &set, const UChar *s, int32_t length) {
+    UChar c=s[length-1], c2;
+    if(c>=0xdc00 && c<=0xdfff && length>=2 && U16_IS_LEAD(c2=s[length-2])) {
+        return set.contains(U16_GET_SUPPLEMENTARY(c2, c)) ? 2 : -2;
+    }
+    return set.contains(c) ? 1 : -1;
+}
+
+static inline int32_t
+spanOneUTF8(const UnicodeSet &set, const uint8_t *s, int32_t length) {
+    UChar32 c=*s;
+    if(U8_IS_SINGLE(c)) {
+        return set.contains(c) ? 1 : -1;
+    }
+    // Take advantage of non-ASCII fastpaths in U8_NEXT_OR_FFFD().
+    int32_t i=0;
+    U8_NEXT_OR_FFFD(s, i, length, c);
+    return set.contains(c) ? i : -i;
+}
+
+static inline int32_t
+spanOneBackUTF8(const UnicodeSet &set, const uint8_t *s, int32_t length) {
+    UChar32 c=s[length-1];
+    if(U8_IS_SINGLE(c)) {
+        return set.contains(c) ? 1 : -1;
+    }
+    int32_t i=length-1;
+    c=utf8_prevCharSafeBody(s, 0, &i, c, -3);
+    length-=i;
+    return set.contains(c) ? length : -length;
+}
+
+/*
+ * Note: In span() when spanLength==0 (after a string match, or at the beginning
+ * after an empty code point span) and in spanNot() and spanNotUTF8(),
+ * string matching could use a binary search
+ * because all string matches are done from the same start index.
+ *
+ * For UTF-8, this would require a comparison function that returns UTF-16 order.
+ *
+ * This optimization should not be necessary for normal UnicodeSets because
+ * most sets have no strings, and most sets with strings have
+ * very few very short strings.
+ * For cases with many strings, it might be better to use a different API
+ * and implementation with a DFA (state machine).
+ */
+
+/*
+ * Algorithm for span(USET_SPAN_CONTAINED)
+ *
+ * Theoretical algorithm:
+ * - Iterate through the string, and at each code point boundary:
+ *   + If the code point there is in the set, then remember to continue after it.
+ *   + If a set string matches at the current position, then remember to continue after it.
+ *   + Either recursively span for each code point or string match,
+ *     or recursively span for all but the shortest one and
+ *     iteratively continue the span with the shortest local match.
+ *   + Remember the longest recursive span (the farthest end point).
+ *   + If there is no match at the current position, neither for the code point there
+ *     nor for any set string, then stop and return the longest recursive span length.
+ *
+ * Optimized implementation:
+ *
+ * (We assume that most sets will have very few very short strings.
+ * A span using a string-less set is extremely fast.)
+ *
+ * Create and cache a spanSet which contains all of the single code points
+ * of the original set but none of its strings.
+ *
+ * - Start with spanLength=spanSet.span(USET_SPAN_CONTAINED).
+ * - Loop:
+ *   + Try to match each set string at the end of the spanLength.
+ *     ~ Set strings that start with set-contained code points must be matched
+ *       with a partial overlap because the recursive algorithm would have tried
+ *       to match them at every position.
+ *     ~ Set strings that entirely consist of set-contained code points
+ *       are irrelevant for span(USET_SPAN_CONTAINED) because the
+ *       recursive algorithm would continue after them anyway
+ *       and find the longest recursive match from their end.
+ *     ~ Rather than recursing, note each end point of a set string match.
+ *   + If no set string matched after spanSet.span(), then return
+ *     with where the spanSet.span() ended.
+ *   + If at least one set string matched after spanSet.span(), then
+ *     pop the shortest string match end point and continue
+ *     the loop, trying to match all set strings from there.
+ *   + If at least one more set string matched after a previous string match,
+ *     then test if the code point after the previous string match is also
+ *     contained in the set.
+ *     Continue the loop with the shortest end point of either this code point
+ *     or a matching set string.
+ *   + If no more set string matched after a previous string match,
+ *     then try another spanLength=spanSet.span(USET_SPAN_CONTAINED).
+ *     Stop if spanLength==0, otherwise continue the loop.
+ *
+ * By noting each end point of a set string match,
+ * the function visits each string position at most once and finishes
+ * in linear time.
+ *
+ * The recursive algorithm may visit the same string position many times
+ * if multiple paths lead to it and finishes in exponential time.
+ */
+
+/*
+ * Algorithm for span(USET_SPAN_SIMPLE)
+ *
+ * Theoretical algorithm:
+ * - Iterate through the string, and at each code point boundary:
+ *   + If the code point there is in the set, then remember to continue after it.
+ *   + If a set string matches at the current position, then remember to continue after it.
+ *   + Continue from the farthest match position and ignore all others.
+ *   + If there is no match at the current position,
+ *     then stop and return the current position.
+ *
+ * Optimized implementation:
+ *
+ * (Same assumption and spanSet as above.)
+ *
+ * - Start with spanLength=spanSet.span(USET_SPAN_CONTAINED).
+ * - Loop:
+ *   + Try to match each set string at the end of the spanLength.
+ *     ~ Set strings that start with set-contained code points must be matched
+ *       with a partial overlap because the standard algorithm would have tried
+ *       to match them earlier.
+ *     ~ Set strings that entirely consist of set-contained code points
+ *       must be matched with a full overlap because the longest-match algorithm
+ *       would hide set string matches that end earlier.
+ *       Such set strings need not be matched earlier inside the code point span
+ *       because the standard algorithm would then have continued after
+ *       the set string match anyway.
+ *     ~ Remember the longest set string match (farthest end point) from the earliest
+ *       starting point.
+ *   + If no set string matched after spanSet.span(), then return
+ *     with where the spanSet.span() ended.
+ *   + If at least one set string matched, then continue the loop after the
+ *     longest match from the earliest position.
+ *   + If no more set string matched after a previous string match,
+ *     then try another spanLength=spanSet.span(USET_SPAN_CONTAINED).
+ *     Stop if spanLength==0, otherwise continue the loop.
+ */
+
+int32_t UnicodeSetStringSpan::span(const UChar *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+        return spanNot(s, length);
+    }
+    int32_t spanLength=spanSet.span(s, length, USET_SPAN_CONTAINED);
+    if(spanLength==length) {
+        return length;
+    }
+
+    // Consider strings; they may overlap with the span.
+    OffsetList offsets;
+    if(spanCondition==USET_SPAN_CONTAINED) {
+        // Use offset list to try all possibilities.
+        offsets.setMaxLength(maxLength16);
+    }
+    int32_t pos=spanLength, rest=length-pos;
+    int32_t i, stringsLength=strings.size();
+    for(;;) {
+        if(spanCondition==USET_SPAN_CONTAINED) {
+            for(i=0; i<stringsLength; ++i) {
+                int32_t overlap=spanLengths[i];
+                if(overlap==ALL_CP_CONTAINED) {
+                    continue;  // Irrelevant string.
+                }
+                const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+                const UChar *s16=string.getBuffer();
+                int32_t length16=string.length();
+
+                // Try to match this string at pos-overlap..pos.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length16;
+                    // While contained: No point matching fully inside the code point span.
+                    U16_BACK_1(s16, 0, overlap);  // Length of the string minus the last code point.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t inc=length16-overlap;  // Keep overlap+inc==length16.
+                for(;;) {
+                    if(inc>rest) {
+                        break;
+                    }
+                    // Try to match if the increment is not listed already.
+                    if(!offsets.containsOffset(inc) && matches16CPB(s, pos-overlap, length, s16, length16)) {
+                        if(inc==rest) {
+                            return length;  // Reached the end of the string.
+                        }
+                        offsets.addOffset(inc);
+                    }
+                    if(overlap==0) {
+                        break;
+                    }
+                    --overlap;
+                    ++inc;
+                }
+            }
+        } else /* USET_SPAN_SIMPLE */ {
+            int32_t maxInc=0, maxOverlap=0;
+            for(i=0; i<stringsLength; ++i) {
+                int32_t overlap=spanLengths[i];
+                // For longest match, we do need to try to match even an all-contained string
+                // to find the match from the earliest start.
+
+                const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+                const UChar *s16=string.getBuffer();
+                int32_t length16=string.length();
+
+                // Try to match this string at pos-overlap..pos.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length16;
+                    // Longest match: Need to match fully inside the code point span
+                    // to find the match from the earliest start.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t inc=length16-overlap;  // Keep overlap+inc==length16.
+                for(;;) {
+                    if(inc>rest || overlap<maxOverlap) {
+                        break;
+                    }
+                    // Try to match if the string is longer or starts earlier.
+                    if( (overlap>maxOverlap || /* redundant overlap==maxOverlap && */ inc>maxInc) &&
+                        matches16CPB(s, pos-overlap, length, s16, length16)
+                    ) {
+                        maxInc=inc;  // Longest match from earliest start.
+                        maxOverlap=overlap;
+                        break;
+                    }
+                    --overlap;
+                    ++inc;
+                }
+            }
+
+            if(maxInc!=0 || maxOverlap!=0) {
+                // Longest-match algorithm, and there was a string match.
+                // Simply continue after it.
+                pos+=maxInc;
+                rest-=maxInc;
+                if(rest==0) {
+                    return length;  // Reached the end of the string.
+                }
+                spanLength=0;  // Match strings from after a string match.
+                continue;
+            }
+        }
+        // Finished trying to match all strings at pos.
+
+        if(spanLength!=0 || pos==0) {
+            // The position is after an unlimited code point span (spanLength!=0),
+            // not after a string match.
+            // The only position where spanLength==0 after a span is pos==0.
+            // Otherwise, an unlimited code point span is only tried again when no
+            // strings match, and if such a non-initial span fails we stop.
+            if(offsets.isEmpty()) {
+                return pos;  // No strings matched after a span.
+            }
+            // Match strings from after the next string match.
+        } else {
+            // The position is after a string match (or a single code point).
+            if(offsets.isEmpty()) {
+                // No more strings matched after a previous string match.
+                // Try another code point span from after the last string match.
+                spanLength=spanSet.span(s+pos, rest, USET_SPAN_CONTAINED);
+                if( spanLength==rest || // Reached the end of the string, or
+                    spanLength==0       // neither strings nor span progressed.
+                ) {
+                    return pos+spanLength;
+                }
+                pos+=spanLength;
+                rest-=spanLength;
+                continue;  // spanLength>0: Match strings from after a span.
+            } else {
+                // Try to match only one code point from after a string match if some
+                // string matched beyond it, so that we try all possible positions
+                // and don't overshoot.
+                spanLength=spanOne(spanSet, s+pos, rest);
+                if(spanLength>0) {
+                    if(spanLength==rest) {
+                        return length;  // Reached the end of the string.
+                    }
+                    // Match strings after this code point.
+                    // There cannot be any increments below it because UnicodeSet strings
+                    // contain multiple code points.
+                    pos+=spanLength;
+                    rest-=spanLength;
+                    offsets.shift(spanLength);
+                    spanLength=0;
+                    continue;  // Match strings from after a single code point.
+                }
+                // Match strings from after the next string match.
+            }
+        }
+        int32_t minOffset=offsets.popMinimum();
+        pos+=minOffset;
+        rest-=minOffset;
+        spanLength=0;  // Match strings from after a string match.
+    }
+}
+
+int32_t UnicodeSetStringSpan::spanBack(const UChar *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+        return spanNotBack(s, length);
+    }
+    int32_t pos=spanSet.spanBack(s, length, USET_SPAN_CONTAINED);
+    if(pos==0) {
+        return 0;
+    }
+    int32_t spanLength=length-pos;
+
+    // Consider strings; they may overlap with the span.
+    OffsetList offsets;
+    if(spanCondition==USET_SPAN_CONTAINED) {
+        // Use offset list to try all possibilities.
+        offsets.setMaxLength(maxLength16);
+    }
+    int32_t i, stringsLength=strings.size();
+    uint8_t *spanBackLengths=spanLengths;
+    if(all) {
+        spanBackLengths+=stringsLength;
+    }
+    for(;;) {
+        if(spanCondition==USET_SPAN_CONTAINED) {
+            for(i=0; i<stringsLength; ++i) {
+                int32_t overlap=spanBackLengths[i];
+                if(overlap==ALL_CP_CONTAINED) {
+                    continue;  // Irrelevant string.
+                }
+                const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+                const UChar *s16=string.getBuffer();
+                int32_t length16=string.length();
+
+                // Try to match this string at pos-(length16-overlap)..pos-length16.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length16;
+                    // While contained: No point matching fully inside the code point span.
+                    int32_t len1=0;
+                    U16_FWD_1(s16, len1, overlap);
+                    overlap-=len1;  // Length of the string minus the first code point.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t dec=length16-overlap;  // Keep dec+overlap==length16.
+                for(;;) {
+                    if(dec>pos) {
+                        break;
+                    }
+                    // Try to match if the decrement is not listed already.
+                    if(!offsets.containsOffset(dec) && matches16CPB(s, pos-dec, length, s16, length16)) {
+                        if(dec==pos) {
+                            return 0;  // Reached the start of the string.
+                        }
+                        offsets.addOffset(dec);
+                    }
+                    if(overlap==0) {
+                        break;
+                    }
+                    --overlap;
+                    ++dec;
+                }
+            }
+        } else /* USET_SPAN_SIMPLE */ {
+            int32_t maxDec=0, maxOverlap=0;
+            for(i=0; i<stringsLength; ++i) {
+                int32_t overlap=spanBackLengths[i];
+                // For longest match, we do need to try to match even an all-contained string
+                // to find the match from the latest end.
+
+                const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+                const UChar *s16=string.getBuffer();
+                int32_t length16=string.length();
+
+                // Try to match this string at pos-(length16-overlap)..pos-length16.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length16;
+                    // Longest match: Need to match fully inside the code point span
+                    // to find the match from the latest end.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t dec=length16-overlap;  // Keep dec+overlap==length16.
+                for(;;) {
+                    if(dec>pos || overlap<maxOverlap) {
+                        break;
+                    }
+                    // Try to match if the string is longer or ends later.
+                    if( (overlap>maxOverlap || /* redundant overlap==maxOverlap && */ dec>maxDec) &&
+                        matches16CPB(s, pos-dec, length, s16, length16)
+                    ) {
+                        maxDec=dec;  // Longest match from latest end.
+                        maxOverlap=overlap;
+                        break;
+                    }
+                    --overlap;
+                    ++dec;
+                }
+            }
+
+            if(maxDec!=0 || maxOverlap!=0) {
+                // Longest-match algorithm, and there was a string match.
+                // Simply continue before it.
+                pos-=maxDec;
+                if(pos==0) {
+                    return 0;  // Reached the start of the string.
+                }
+                spanLength=0;  // Match strings from before a string match.
+                continue;
+            }
+        }
+        // Finished trying to match all strings at pos.
+
+        if(spanLength!=0 || pos==length) {
+            // The position is before an unlimited code point span (spanLength!=0),
+            // not before a string match.
+            // The only position where spanLength==0 before a span is pos==length.
+            // Otherwise, an unlimited code point span is only tried again when no
+            // strings match, and if such a non-initial span fails we stop.
+            if(offsets.isEmpty()) {
+                return pos;  // No strings matched before a span.
+            }
+            // Match strings from before the next string match.
+        } else {
+            // The position is before a string match (or a single code point).
+            if(offsets.isEmpty()) {
+                // No more strings matched before a previous string match.
+                // Try another code point span from before the last string match.
+                int32_t oldPos=pos;
+                pos=spanSet.spanBack(s, oldPos, USET_SPAN_CONTAINED);
+                spanLength=oldPos-pos;
+                if( pos==0 ||           // Reached the start of the string, or
+                    spanLength==0       // neither strings nor span progressed.
+                ) {
+                    return pos;
+                }
+                continue;  // spanLength>0: Match strings from before a span.
+            } else {
+                // Try to match only one code point from before a string match if some
+                // string matched beyond it, so that we try all possible positions
+                // and don't overshoot.
+                spanLength=spanOneBack(spanSet, s, pos);
+                if(spanLength>0) {
+                    if(spanLength==pos) {
+                        return 0;  // Reached the start of the string.
+                    }
+                    // Match strings before this code point.
+                    // There cannot be any decrements below it because UnicodeSet strings
+                    // contain multiple code points.
+                    pos-=spanLength;
+                    offsets.shift(spanLength);
+                    spanLength=0;
+                    continue;  // Match strings from before a single code point.
+                }
+                // Match strings from before the next string match.
+            }
+        }
+        pos-=offsets.popMinimum();
+        spanLength=0;  // Match strings from before a string match.
+    }
+}
+
+int32_t UnicodeSetStringSpan::spanUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+        return spanNotUTF8(s, length);
+    }
+    int32_t spanLength=spanSet.spanUTF8((const char *)s, length, USET_SPAN_CONTAINED);
+    if(spanLength==length) {
+        return length;
+    }
+
+    // Consider strings; they may overlap with the span.
+    OffsetList offsets;
+    if(spanCondition==USET_SPAN_CONTAINED) {
+        // Use offset list to try all possibilities.
+        offsets.setMaxLength(maxLength8);
+    }
+    int32_t pos=spanLength, rest=length-pos;
+    int32_t i, stringsLength=strings.size();
+    uint8_t *spanUTF8Lengths=spanLengths;
+    if(all) {
+        spanUTF8Lengths+=2*stringsLength;
+    }
+    for(;;) {
+        const uint8_t *s8=utf8;
+        int32_t length8;
+        if(spanCondition==USET_SPAN_CONTAINED) {
+            for(i=0; i<stringsLength; ++i) {
+                length8=utf8Lengths[i];
+                if(length8==0) {
+                    continue;  // String not representable in UTF-8.
+                }
+                int32_t overlap=spanUTF8Lengths[i];
+                if(overlap==ALL_CP_CONTAINED) {
+                    s8+=length8;
+                    continue;  // Irrelevant string.
+                }
+
+                // Try to match this string at pos-overlap..pos.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length8;
+                    // While contained: No point matching fully inside the code point span.
+                    U8_BACK_1(s8, 0, overlap);  // Length of the string minus the last code point.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t inc=length8-overlap;  // Keep overlap+inc==length8.
+                for(;;) {
+                    if(inc>rest) {
+                        break;
+                    }
+                    // Try to match if the increment is not listed already.
+                    // Match at code point boundaries. (The UTF-8 strings were converted
+                    // from UTF-16 and are guaranteed to be well-formed.)
+                    if(!U8_IS_TRAIL(s[pos-overlap]) &&
+                            !offsets.containsOffset(inc) &&
+                            matches8(s+pos-overlap, s8, length8)) {
+                        if(inc==rest) {
+                            return length;  // Reached the end of the string.
+                        }
+                        offsets.addOffset(inc);
+                    }
+                    if(overlap==0) {
+                        break;
+                    }
+                    --overlap;
+                    ++inc;
+                }
+                s8+=length8;
+            }
+        } else /* USET_SPAN_SIMPLE */ {
+            int32_t maxInc=0, maxOverlap=0;
+            for(i=0; i<stringsLength; ++i) {
+                length8=utf8Lengths[i];
+                if(length8==0) {
+                    continue;  // String not representable in UTF-8.
+                }
+                int32_t overlap=spanUTF8Lengths[i];
+                // For longest match, we do need to try to match even an all-contained string
+                // to find the match from the earliest start.
+
+                // Try to match this string at pos-overlap..pos.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length8;
+                    // Longest match: Need to match fully inside the code point span
+                    // to find the match from the earliest start.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t inc=length8-overlap;  // Keep overlap+inc==length8.
+                for(;;) {
+                    if(inc>rest || overlap<maxOverlap) {
+                        break;
+                    }
+                    // Try to match if the string is longer or starts earlier.
+                    // Match at code point boundaries. (The UTF-8 strings were converted
+                    // from UTF-16 and are guaranteed to be well-formed.)
+                    if(!U8_IS_TRAIL(s[pos-overlap]) &&
+                            (overlap>maxOverlap ||
+                                /* redundant overlap==maxOverlap && */ inc>maxInc) &&
+                            matches8(s+pos-overlap, s8, length8)) {
+                        maxInc=inc;  // Longest match from earliest start.
+                        maxOverlap=overlap;
+                        break;
+                    }
+                    --overlap;
+                    ++inc;
+                }
+                s8+=length8;
+            }
+
+            if(maxInc!=0 || maxOverlap!=0) {
+                // Longest-match algorithm, and there was a string match.
+                // Simply continue after it.
+                pos+=maxInc;
+                rest-=maxInc;
+                if(rest==0) {
+                    return length;  // Reached the end of the string.
+                }
+                spanLength=0;  // Match strings from after a string match.
+                continue;
+            }
+        }
+        // Finished trying to match all strings at pos.
+
+        if(spanLength!=0 || pos==0) {
+            // The position is after an unlimited code point span (spanLength!=0),
+            // not after a string match.
+            // The only position where spanLength==0 after a span is pos==0.
+            // Otherwise, an unlimited code point span is only tried again when no
+            // strings match, and if such a non-initial span fails we stop.
+            if(offsets.isEmpty()) {
+                return pos;  // No strings matched after a span.
+            }
+            // Match strings from after the next string match.
+        } else {
+            // The position is after a string match (or a single code point).
+            if(offsets.isEmpty()) {
+                // No more strings matched after a previous string match.
+                // Try another code point span from after the last string match.
+                spanLength=spanSet.spanUTF8((const char *)s+pos, rest, USET_SPAN_CONTAINED);
+                if( spanLength==rest || // Reached the end of the string, or
+                    spanLength==0       // neither strings nor span progressed.
+                ) {
+                    return pos+spanLength;
+                }
+                pos+=spanLength;
+                rest-=spanLength;
+                continue;  // spanLength>0: Match strings from after a span.
+            } else {
+                // Try to match only one code point from after a string match if some
+                // string matched beyond it, so that we try all possible positions
+                // and don't overshoot.
+                spanLength=spanOneUTF8(spanSet, s+pos, rest);
+                if(spanLength>0) {
+                    if(spanLength==rest) {
+                        return length;  // Reached the end of the string.
+                    }
+                    // Match strings after this code point.
+                    // There cannot be any increments below it because UnicodeSet strings
+                    // contain multiple code points.
+                    pos+=spanLength;
+                    rest-=spanLength;
+                    offsets.shift(spanLength);
+                    spanLength=0;
+                    continue;  // Match strings from after a single code point.
+                }
+                // Match strings from after the next string match.
+            }
+        }
+        int32_t minOffset=offsets.popMinimum();
+        pos+=minOffset;
+        rest-=minOffset;
+        spanLength=0;  // Match strings from after a string match.
+    }
+}
+
+int32_t UnicodeSetStringSpan::spanBackUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const {
+    if(spanCondition==USET_SPAN_NOT_CONTAINED) {
+        return spanNotBackUTF8(s, length);
+    }
+    int32_t pos=spanSet.spanBackUTF8((const char *)s, length, USET_SPAN_CONTAINED);
+    if(pos==0) {
+        return 0;
+    }
+    int32_t spanLength=length-pos;
+
+    // Consider strings; they may overlap with the span.
+    OffsetList offsets;
+    if(spanCondition==USET_SPAN_CONTAINED) {
+        // Use offset list to try all possibilities.
+        offsets.setMaxLength(maxLength8);
+    }
+    int32_t i, stringsLength=strings.size();
+    uint8_t *spanBackUTF8Lengths=spanLengths;
+    if(all) {
+        spanBackUTF8Lengths+=3*stringsLength;
+    }
+    for(;;) {
+        const uint8_t *s8=utf8;
+        int32_t length8;
+        if(spanCondition==USET_SPAN_CONTAINED) {
+            for(i=0; i<stringsLength; ++i) {
+                length8=utf8Lengths[i];
+                if(length8==0) {
+                    continue;  // String not representable in UTF-8.
+                }
+                int32_t overlap=spanBackUTF8Lengths[i];
+                if(overlap==ALL_CP_CONTAINED) {
+                    s8+=length8;
+                    continue;  // Irrelevant string.
+                }
+
+                // Try to match this string at pos-(length8-overlap)..pos-length8.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length8;
+                    // While contained: No point matching fully inside the code point span.
+                    int32_t len1=0;
+                    U8_FWD_1(s8, len1, overlap);
+                    overlap-=len1;  // Length of the string minus the first code point.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t dec=length8-overlap;  // Keep dec+overlap==length8.
+                for(;;) {
+                    if(dec>pos) {
+                        break;
+                    }
+                    // Try to match if the decrement is not listed already.
+                    // Match at code point boundaries. (The UTF-8 strings were converted
+                    // from UTF-16 and are guaranteed to be well-formed.)
+                    if( !U8_IS_TRAIL(s[pos-dec]) &&
+                        !offsets.containsOffset(dec) &&
+                        matches8(s+pos-dec, s8, length8)
+                    ) {
+                        if(dec==pos) {
+                            return 0;  // Reached the start of the string.
+                        }
+                        offsets.addOffset(dec);
+                    }
+                    if(overlap==0) {
+                        break;
+                    }
+                    --overlap;
+                    ++dec;
+                }
+                s8+=length8;
+            }
+        } else /* USET_SPAN_SIMPLE */ {
+            int32_t maxDec=0, maxOverlap=0;
+            for(i=0; i<stringsLength; ++i) {
+                length8=utf8Lengths[i];
+                if(length8==0) {
+                    continue;  // String not representable in UTF-8.
+                }
+                int32_t overlap=spanBackUTF8Lengths[i];
+                // For longest match, we do need to try to match even an all-contained string
+                // to find the match from the latest end.
+
+                // Try to match this string at pos-(length8-overlap)..pos-length8.
+                if(overlap>=LONG_SPAN) {
+                    overlap=length8;
+                    // Longest match: Need to match fully inside the code point span
+                    // to find the match from the latest end.
+                }
+                if(overlap>spanLength) {
+                    overlap=spanLength;
+                }
+                int32_t dec=length8-overlap;  // Keep dec+overlap==length8.
+                for(;;) {
+                    if(dec>pos || overlap<maxOverlap) {
+                        break;
+                    }
+                    // Try to match if the string is longer or ends later.
+                    // Match at code point boundaries. (The UTF-8 strings were converted
+                    // from UTF-16 and are guaranteed to be well-formed.)
+                    if( !U8_IS_TRAIL(s[pos-dec]) &&
+                        (overlap>maxOverlap || /* redundant overlap==maxOverlap && */ dec>maxDec) &&
+                        matches8(s+pos-dec, s8, length8)
+                    ) {
+                        maxDec=dec;  // Longest match from latest end.
+                        maxOverlap=overlap;
+                        break;
+                    }
+                    --overlap;
+                    ++dec;
+                }
+                s8+=length8;
+            }
+
+            if(maxDec!=0 || maxOverlap!=0) {
+                // Longest-match algorithm, and there was a string match.
+                // Simply continue before it.
+                pos-=maxDec;
+                if(pos==0) {
+                    return 0;  // Reached the start of the string.
+                }
+                spanLength=0;  // Match strings from before a string match.
+                continue;
+            }
+        }
+        // Finished trying to match all strings at pos.
+
+        if(spanLength!=0 || pos==length) {
+            // The position is before an unlimited code point span (spanLength!=0),
+            // not before a string match.
+            // The only position where spanLength==0 before a span is pos==length.
+            // Otherwise, an unlimited code point span is only tried again when no
+            // strings match, and if such a non-initial span fails we stop.
+            if(offsets.isEmpty()) {
+                return pos;  // No strings matched before a span.
+            }
+            // Match strings from before the next string match.
+        } else {
+            // The position is before a string match (or a single code point).
+            if(offsets.isEmpty()) {
+                // No more strings matched before a previous string match.
+                // Try another code point span from before the last string match.
+                int32_t oldPos=pos;
+                pos=spanSet.spanBackUTF8((const char *)s, oldPos, USET_SPAN_CONTAINED);
+                spanLength=oldPos-pos;
+                if( pos==0 ||           // Reached the start of the string, or
+                    spanLength==0       // neither strings nor span progressed.
+                ) {
+                    return pos;
+                }
+                continue;  // spanLength>0: Match strings from before a span.
+            } else {
+                // Try to match only one code point from before a string match if some
+                // string matched beyond it, so that we try all possible positions
+                // and don't overshoot.
+                spanLength=spanOneBackUTF8(spanSet, s, pos);
+                if(spanLength>0) {
+                    if(spanLength==pos) {
+                        return 0;  // Reached the start of the string.
+                    }
+                    // Match strings before this code point.
+                    // There cannot be any decrements below it because UnicodeSet strings
+                    // contain multiple code points.
+                    pos-=spanLength;
+                    offsets.shift(spanLength);
+                    spanLength=0;
+                    continue;  // Match strings from before a single code point.
+                }
+                // Match strings from before the next string match.
+            }
+        }
+        pos-=offsets.popMinimum();
+        spanLength=0;  // Match strings from before a string match.
+    }
+}
+
+/*
+ * Algorithm for spanNot()==span(USET_SPAN_NOT_CONTAINED)
+ *
+ * Theoretical algorithm:
+ * - Iterate through the string, and at each code point boundary:
+ *   + If the code point there is in the set, then return with the current position.
+ *   + If a set string matches at the current position, then return with the current position.
+ *
+ * Optimized implementation:
+ *
+ * (Same assumption as for span() above.)
+ *
+ * Create and cache a spanNotSet which contains all of the single code points
+ * of the original set but none of its strings.
+ * For each set string add its initial code point to the spanNotSet.
+ * (Also add its final code point for spanNotBack().)
+ *
+ * - Loop:
+ *   + Do spanLength=spanNotSet.span(USET_SPAN_NOT_CONTAINED).
+ *   + If the current code point is in the original set, then
+ *     return the current position.
+ *   + If any set string matches at the current position, then
+ *     return the current position.
+ *   + If there is no match at the current position, neither for the code point there
+ *     nor for any set string, then skip this code point and continue the loop.
+ *     This happens for set-string-initial code points that were added to spanNotSet
+ *     when there is not actually a match for such a set string.
+ */
+
+int32_t UnicodeSetStringSpan::spanNot(const UChar *s, int32_t length) const {
+    int32_t pos=0, rest=length;
+    int32_t i, stringsLength=strings.size();
+    do {
+        // Span until we find a code point from the set,
+        // or a code point that starts or ends some string.
+        i=pSpanNotSet->span(s+pos, rest, USET_SPAN_NOT_CONTAINED);
+        if(i==rest) {
+            return length;  // Reached the end of the string.
+        }
+        pos+=i;
+        rest-=i;
+
+        // Check whether the current code point is in the original set,
+        // without the string starts and ends.
+        int32_t cpLength=spanOne(spanSet, s+pos, rest);
+        if(cpLength>0) {
+            return pos;  // There is a set element at pos.
+        }
+
+        // Try to match the strings at pos.
+        for(i=0; i<stringsLength; ++i) {
+            if(spanLengths[i]==ALL_CP_CONTAINED) {
+                continue;  // Irrelevant string.
+            }
+            const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+            const UChar *s16=string.getBuffer();
+            int32_t length16=string.length();
+            if(length16<=rest && matches16CPB(s, pos, length, s16, length16)) {
+                return pos;  // There is a set element at pos.
+            }
+        }
+
+        // The span(while not contained) ended on a string start/end which is
+        // not in the original set. Skip this code point and continue.
+        // cpLength<0
+        pos-=cpLength;
+        rest+=cpLength;
+    } while(rest!=0);
+    return length;  // Reached the end of the string.
+}
+
+int32_t UnicodeSetStringSpan::spanNotBack(const UChar *s, int32_t length) const {
+    int32_t pos=length;
+    int32_t i, stringsLength=strings.size();
+    do {
+        // Span until we find a code point from the set,
+        // or a code point that starts or ends some string.
+        pos=pSpanNotSet->spanBack(s, pos, USET_SPAN_NOT_CONTAINED);
+        if(pos==0) {
+            return 0;  // Reached the start of the string.
+        }
+
+        // Check whether the current code point is in the original set,
+        // without the string starts and ends.
+        int32_t cpLength=spanOneBack(spanSet, s, pos);
+        if(cpLength>0) {
+            return pos;  // There is a set element at pos.
+        }
+
+        // Try to match the strings at pos.
+        for(i=0; i<stringsLength; ++i) {
+            // Use spanLengths rather than a spanBackLengths pointer because
+            // it is easier and we only need to know whether the string is irrelevant
+            // which is the same in either array.
+            if(spanLengths[i]==ALL_CP_CONTAINED) {
+                continue;  // Irrelevant string.
+            }
+            const UnicodeString &string=*(const UnicodeString *)strings.elementAt(i);
+            const UChar *s16=string.getBuffer();
+            int32_t length16=string.length();
+            if(length16<=pos && matches16CPB(s, pos-length16, length, s16, length16)) {
+                return pos;  // There is a set element at pos.
+            }
+        }
+
+        // The span(while not contained) ended on a string start/end which is
+        // not in the original set. Skip this code point and continue.
+        // cpLength<0
+        pos+=cpLength;
+    } while(pos!=0);
+    return 0;  // Reached the start of the string.
+}
+
+int32_t UnicodeSetStringSpan::spanNotUTF8(const uint8_t *s, int32_t length) const {
+    int32_t pos=0, rest=length;
+    int32_t i, stringsLength=strings.size();
+    uint8_t *spanUTF8Lengths=spanLengths;
+    if(all) {
+        spanUTF8Lengths+=2*stringsLength;
+    }
+    do {
+        // Span until we find a code point from the set,
+        // or a code point that starts or ends some string.
+        i=pSpanNotSet->spanUTF8((const char *)s+pos, rest, USET_SPAN_NOT_CONTAINED);
+        if(i==rest) {
+            return length;  // Reached the end of the string.
+        }
+        pos+=i;
+        rest-=i;
+
+        // Check whether the current code point is in the original set,
+        // without the string starts and ends.
+        int32_t cpLength=spanOneUTF8(spanSet, s+pos, rest);
+        if(cpLength>0) {
+            return pos;  // There is a set element at pos.
+        }
+
+        // Try to match the strings at pos.
+        const uint8_t *s8=utf8;
+        int32_t length8;
+        for(i=0; i<stringsLength; ++i) {
+            length8=utf8Lengths[i];
+            // ALL_CP_CONTAINED: Irrelevant string.
+            if(length8!=0 && spanUTF8Lengths[i]!=ALL_CP_CONTAINED && length8<=rest && matches8(s+pos, s8, length8)) {
+                return pos;  // There is a set element at pos.
+            }
+            s8+=length8;
+        }
+
+        // The span(while not contained) ended on a string start/end which is
+        // not in the original set. Skip this code point and continue.
+        // cpLength<0
+        pos-=cpLength;
+        rest+=cpLength;
+    } while(rest!=0);
+    return length;  // Reached the end of the string.
+}
+
+int32_t UnicodeSetStringSpan::spanNotBackUTF8(const uint8_t *s, int32_t length) const {
+    int32_t pos=length;
+    int32_t i, stringsLength=strings.size();
+    uint8_t *spanBackUTF8Lengths=spanLengths;
+    if(all) {
+        spanBackUTF8Lengths+=3*stringsLength;
+    }
+    do {
+        // Span until we find a code point from the set,
+        // or a code point that starts or ends some string.
+        pos=pSpanNotSet->spanBackUTF8((const char *)s, pos, USET_SPAN_NOT_CONTAINED);
+        if(pos==0) {
+            return 0;  // Reached the start of the string.
+        }
+
+        // Check whether the current code point is in the original set,
+        // without the string starts and ends.
+        int32_t cpLength=spanOneBackUTF8(spanSet, s, pos);
+        if(cpLength>0) {
+            return pos;  // There is a set element at pos.
+        }
+
+        // Try to match the strings at pos.
+        const uint8_t *s8=utf8;
+        int32_t length8;
+        for(i=0; i<stringsLength; ++i) {
+            length8=utf8Lengths[i];
+            // ALL_CP_CONTAINED: Irrelevant string.
+            if(length8!=0 && spanBackUTF8Lengths[i]!=ALL_CP_CONTAINED && length8<=pos && matches8(s+pos-length8, s8, length8)) {
+                return pos;  // There is a set element at pos.
+            }
+            s8+=length8;
+        }
+
+        // The span(while not contained) ended on a string start/end which is
+        // not in the original set. Skip this code point and continue.
+        // cpLength<0
+        pos+=cpLength;
+    } while(pos!=0);
+    return 0;  // Reached the start of the string.
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/unisetspan.h b/thirdparty/icu4c/common/unisetspan.h
new file mode 100644
index 0000000000..9a1307a907
--- /dev/null
+++ b/thirdparty/icu4c/common/unisetspan.h
@@ -0,0 +1,157 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2007, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  unisetspan.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2007mar01
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UNISETSPAN_H__
+#define __UNISETSPAN_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uniset.h"
+
+U_NAMESPACE_BEGIN
+
+/*
+ * Implement span() etc. for a set with strings.
+ * Avoid recursion because of its exponential complexity.
+ * Instead, try multiple paths at once and track them with an IndexList.
+ */
+class UnicodeSetStringSpan : public UMemory {
+public:
+    /*
+     * Which span() variant will be used?
+     * The object is either built for one variant and used once,
+     * or built for all and may be used many times.
+     */
+    enum {
+        FWD             = 0x20,
+        BACK            = 0x10,
+        UTF16           = 8,
+        UTF8            = 4,
+        CONTAINED       = 2,
+        NOT_CONTAINED   = 1,
+
+        ALL             = 0x3f,
+
+        FWD_UTF16_CONTAINED     = FWD  | UTF16 |     CONTAINED,
+        FWD_UTF16_NOT_CONTAINED = FWD  | UTF16 | NOT_CONTAINED,
+        FWD_UTF8_CONTAINED      = FWD  | UTF8  |     CONTAINED,
+        FWD_UTF8_NOT_CONTAINED  = FWD  | UTF8  | NOT_CONTAINED,
+        BACK_UTF16_CONTAINED    = BACK | UTF16 |     CONTAINED,
+        BACK_UTF16_NOT_CONTAINED= BACK | UTF16 | NOT_CONTAINED,
+        BACK_UTF8_CONTAINED     = BACK | UTF8  |     CONTAINED,
+        BACK_UTF8_NOT_CONTAINED = BACK | UTF8  | NOT_CONTAINED
+    };
+
+    UnicodeSetStringSpan(const UnicodeSet &set, const UVector &setStrings, uint32_t which);
+
+    // Copy constructor. Assumes which==ALL for a frozen set.
+    UnicodeSetStringSpan(const UnicodeSetStringSpan &otherStringSpan, const UVector &newParentSetStrings);
+
+    ~UnicodeSetStringSpan();
+
+    /*
+     * Do the strings need to be checked in span() etc.?
+     * @return true if strings need to be checked (call span() here),
+     *         false if not (use a BMPSet for best performance).
+     */
+    inline UBool needsStringSpanUTF16();
+    inline UBool needsStringSpanUTF8();
+
+    // For fast UnicodeSet::contains(c).
+    inline UBool contains(UChar32 c) const;
+
+    int32_t span(const UChar *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    int32_t spanBack(const UChar *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    int32_t spanUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+    int32_t spanBackUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const;
+
+private:
+    // Special spanLength byte values.
+    enum {
+        // The spanLength is >=0xfe.
+        LONG_SPAN=0xfe,
+        // All code points in the string are contained in the parent set.
+        ALL_CP_CONTAINED=0xff
+    };
+
+    // Add a starting or ending string character to the spanNotSet
+    // so that a character span ends before any string.
+    void addToSpanNotSet(UChar32 c);
+
+    int32_t spanNot(const UChar *s, int32_t length) const;
+    int32_t spanNotBack(const UChar *s, int32_t length) const;
+    int32_t spanNotUTF8(const uint8_t *s, int32_t length) const;
+    int32_t spanNotBackUTF8(const uint8_t *s, int32_t length) const;
+
+    // Set for span(). Same as parent but without strings.
+    UnicodeSet spanSet;
+
+    // Set for span(not contained).
+    // Same as spanSet, plus characters that start or end strings.
+    UnicodeSet *pSpanNotSet;
+
+    // The strings of the parent set.
+    const UVector &strings;
+
+    // Pointer to the UTF-8 string lengths.
+    // Also pointer to further allocated storage for meta data and
+    // UTF-8 string contents as necessary.
+    int32_t *utf8Lengths;
+
+    // Pointer to the part of the (utf8Lengths) memory block that stores
+    // the lengths of span(), spanBack() etc. for each string.
+    uint8_t *spanLengths;
+
+    // Pointer to the part of the (utf8Lengths) memory block that stores
+    // the UTF-8 versions of the parent set's strings.
+    uint8_t *utf8;
+
+    // Number of bytes for all UTF-8 versions of strings together.
+    int32_t utf8Length;
+
+    // Maximum lengths of relevant strings.
+    int32_t maxLength16;
+    int32_t maxLength8;
+
+    // Set up for all variants of span()?
+    UBool all;
+
+    // Memory for small numbers and lengths of strings.
+    // For example, for 8 strings:
+    // 8 UTF-8 lengths, 8*4 bytes span lengths, 8*2 3-byte UTF-8 characters
+    // = 112 bytes = int32_t[28].
+    int32_t staticLengths[32];
+};
+
+UBool UnicodeSetStringSpan::needsStringSpanUTF16() {
+    return (UBool)(maxLength16!=0);
+}
+
+UBool UnicodeSetStringSpan::needsStringSpanUTF8() {
+    return (UBool)(maxLength8!=0);
+}
+
+UBool UnicodeSetStringSpan::contains(UChar32 c) const {
+    return spanSet.contains(c);
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unistr.cpp b/thirdparty/icu4c/common/unistr.cpp
new file mode 100644
index 0000000000..077b4d6ef2
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr.cpp
@@ -0,0 +1,1982 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1999-2016, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File unistr.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   09/25/98    stephen     Creation.
+*   04/20/99    stephen     Overhauled per 4/16 code review.
+*   07/09/99    stephen     Renamed {hi,lo},{byte,word} to icu_X for HP/UX
+*   11/18/99    aliu        Added handleReplaceBetween() to make inherit from
+*                           Replaceable.
+*   06/25/01    grhoten     Removed the dependency on iostream
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/appendable.h"
+#include "unicode/putil.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "unicode/ustring.h"
+#include "unicode/unistr.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "uelement.h"
+#include "ustr_imp.h"
+#include "umutex.h"
+#include "uassert.h"
+
+#if 0
+
+#include <iostream>
+using namespace std;
+
+//DEBUGGING
+void
+print(const UnicodeString& s,
+      const char *name)
+{
+  UChar c;
+  cout << name << ":|";
+  for(int i = 0; i < s.length(); ++i) {
+    c = s[i];
+    if(c>= 0x007E || c < 0x0020)
+      cout << "[0x" << hex << s[i] << "]";
+    else
+      cout << (char) s[i];
+  }
+  cout << '|' << endl;
+}
+
+void
+print(const UChar *s,
+      int32_t len,
+      const char *name)
+{
+  UChar c;
+  cout << name << ":|";
+  for(int i = 0; i < len; ++i) {
+    c = s[i];
+    if(c>= 0x007E || c < 0x0020)
+      cout << "[0x" << hex << s[i] << "]";
+    else
+      cout << (char) s[i];
+  }
+  cout << '|' << endl;
+}
+// END DEBUGGING
+#endif
+
+// Local function definitions for now
+
+// need to copy areas that may overlap
+static
+inline void
+us_arrayCopy(const UChar *src, int32_t srcStart,
+         UChar *dst, int32_t dstStart, int32_t count)
+{
+  if(count>0) {
+    uprv_memmove(dst+dstStart, src+srcStart, (size_t)count*sizeof(*src));
+  }
+}
+
+// u_unescapeAt() callback to get a UChar from a UnicodeString
+U_CDECL_BEGIN
+static UChar U_CALLCONV
+UnicodeString_charAt(int32_t offset, void *context) {
+    return ((icu::UnicodeString*) context)->charAt(offset);
+}
+U_CDECL_END
+
+U_NAMESPACE_BEGIN
+
+/* The Replaceable virtual destructor can't be defined in the header
+   due to how AIX works with multiple definitions of virtual functions.
+*/
+Replaceable::~Replaceable() {}
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UnicodeString)
+
+UnicodeString U_EXPORT2
+operator+ (const UnicodeString &s1, const UnicodeString &s2) {
+    return
+        UnicodeString(s1.length()+s2.length()+1, (UChar32)0, 0).
+            append(s1).
+                append(s2);
+}
+
+//========================================
+// Reference Counting functions, put at top of file so that optimizing compilers
+//                               have a chance to automatically inline.
+//========================================
+
+void
+UnicodeString::addRef() {
+  umtx_atomic_inc((u_atomic_int32_t *)fUnion.fFields.fArray - 1);
+}
+
+int32_t
+UnicodeString::removeRef() {
+  return umtx_atomic_dec((u_atomic_int32_t *)fUnion.fFields.fArray - 1);
+}
+
+int32_t
+UnicodeString::refCount() const {
+  return umtx_loadAcquire(*((u_atomic_int32_t *)fUnion.fFields.fArray - 1));
+}
+
+void
+UnicodeString::releaseArray() {
+  if((fUnion.fFields.fLengthAndFlags & kRefCounted) && removeRef() == 0) {
+    uprv_free((int32_t *)fUnion.fFields.fArray - 1);
+  }
+}
+
+
+
+//========================================
+// Constructors
+//========================================
+
+// The default constructor is inline in unistr.h.
+
+UnicodeString::UnicodeString(int32_t capacity, UChar32 c, int32_t count) {
+  fUnion.fFields.fLengthAndFlags = 0;
+  if(count <= 0 || (uint32_t)c > 0x10ffff) {
+    // just allocate and do not do anything else
+    allocate(capacity);
+  } else if(c <= 0xffff) {
+    int32_t length = count;
+    if(capacity < length) {
+      capacity = length;
+    }
+    if(allocate(capacity)) {
+      UChar *array = getArrayStart();
+      UChar unit = (UChar)c;
+      for(int32_t i = 0; i < length; ++i) {
+        array[i] = unit;
+      }
+      setLength(length);
+    }
+  } else {  // supplementary code point, write surrogate pairs
+    if(count > (INT32_MAX / 2)) {
+      // We would get more than 2G UChars.
+      allocate(capacity);
+      return;
+    }
+    int32_t length = count * 2;
+    if(capacity < length) {
+      capacity = length;
+    }
+    if(allocate(capacity)) {
+      UChar *array = getArrayStart();
+      UChar lead = U16_LEAD(c);
+      UChar trail = U16_TRAIL(c);
+      for(int32_t i = 0; i < length; i += 2) {
+        array[i] = lead;
+        array[i + 1] = trail;
+      }
+      setLength(length);
+    }
+  }
+}
+
+UnicodeString::UnicodeString(UChar ch) {
+  fUnion.fFields.fLengthAndFlags = kLength1 | kShortString;
+  fUnion.fStackFields.fBuffer[0] = ch;
+}
+
+UnicodeString::UnicodeString(UChar32 ch) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  int32_t i = 0;
+  UBool isError = FALSE;
+  U16_APPEND(fUnion.fStackFields.fBuffer, i, US_STACKBUF_SIZE, ch, isError);
+  // We test isError so that the compiler does not complain that we don't.
+  // If isError then i==0 which is what we want anyway.
+  if(!isError) {
+    setShortLength(i);
+  }
+}
+
+UnicodeString::UnicodeString(const UChar *text) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  doAppend(text, 0, -1);
+}
+
+UnicodeString::UnicodeString(const UChar *text,
+                             int32_t textLength) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  doAppend(text, 0, textLength);
+}
+
+UnicodeString::UnicodeString(UBool isTerminated,
+                             ConstChar16Ptr textPtr,
+                             int32_t textLength) {
+  fUnion.fFields.fLengthAndFlags = kReadonlyAlias;
+  const UChar *text = textPtr;
+  if(text == NULL) {
+    // treat as an empty string, do not alias
+    setToEmpty();
+  } else if(textLength < -1 ||
+            (textLength == -1 && !isTerminated) ||
+            (textLength >= 0 && isTerminated && text[textLength] != 0)
+  ) {
+    setToBogus();
+  } else {
+    if(textLength == -1) {
+      // text is terminated, or else it would have failed the above test
+      textLength = u_strlen(text);
+    }
+    setArray(const_cast<UChar *>(text), textLength,
+             isTerminated ? textLength + 1 : textLength);
+  }
+}
+
+UnicodeString::UnicodeString(UChar *buff,
+                             int32_t buffLength,
+                             int32_t buffCapacity) {
+  fUnion.fFields.fLengthAndFlags = kWritableAlias;
+  if(buff == NULL) {
+    // treat as an empty string, do not alias
+    setToEmpty();
+  } else if(buffLength < -1 || buffCapacity < 0 || buffLength > buffCapacity) {
+    setToBogus();
+  } else {
+    if(buffLength == -1) {
+      // fLength = u_strlen(buff); but do not look beyond buffCapacity
+      const UChar *p = buff, *limit = buff + buffCapacity;
+      while(p != limit && *p != 0) {
+        ++p;
+      }
+      buffLength = (int32_t)(p - buff);
+    }
+    setArray(buff, buffLength, buffCapacity);
+  }
+}
+
+UnicodeString::UnicodeString(const char *src, int32_t length, EInvariant) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  if(src==NULL) {
+    // treat as an empty string
+  } else {
+    if(length<0) {
+      length=(int32_t)uprv_strlen(src);
+    }
+    if(cloneArrayIfNeeded(length, length, FALSE)) {
+      u_charsToUChars(src, getArrayStart(), length);
+      setLength(length);
+    } else {
+      setToBogus();
+    }
+  }
+}
+
+#if U_CHARSET_IS_UTF8
+
+UnicodeString::UnicodeString(const char *codepageData) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  if(codepageData != 0) {
+    setToUTF8(codepageData);
+  }
+}
+
+UnicodeString::UnicodeString(const char *codepageData, int32_t dataLength) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  // if there's nothing to convert, do nothing
+  if(codepageData == 0 || dataLength == 0 || dataLength < -1) {
+    return;
+  }
+  if(dataLength == -1) {
+    dataLength = (int32_t)uprv_strlen(codepageData);
+  }
+  setToUTF8(StringPiece(codepageData, dataLength));
+}
+
+// else see unistr_cnv.cpp
+#endif
+
+UnicodeString::UnicodeString(const UnicodeString& that) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  copyFrom(that);
+}
+
+UnicodeString::UnicodeString(UnicodeString &&src) U_NOEXCEPT {
+  copyFieldsFrom(src, TRUE);
+}
+
+UnicodeString::UnicodeString(const UnicodeString& that,
+                             int32_t srcStart) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  setTo(that, srcStart);
+}
+
+UnicodeString::UnicodeString(const UnicodeString& that,
+                             int32_t srcStart,
+                             int32_t srcLength) {
+  fUnion.fFields.fLengthAndFlags = kShortString;
+  setTo(that, srcStart, srcLength);
+}
+
+// Replaceable base class clone() default implementation, does not clone
+Replaceable *
+Replaceable::clone() const {
+  return NULL;
+}
+
+// UnicodeString overrides clone() with a real implementation
+UnicodeString *
+UnicodeString::clone() const {
+  return new UnicodeString(*this);
+}
+
+//========================================
+// array allocation
+//========================================
+
+namespace {
+
+const int32_t kGrowSize = 128;
+
+// The number of bytes for one int32_t reference counter and capacity UChars
+// must fit into a 32-bit size_t (at least when on a 32-bit platform).
+// We also add one for the NUL terminator, to avoid reallocation in getTerminatedBuffer(),
+// and round up to a multiple of 16 bytes.
+// This means that capacity must be at most (0xfffffff0 - 4) / 2 - 1 = 0x7ffffff5.
+// (With more complicated checks we could go up to 0x7ffffffd without rounding up,
+// but that does not seem worth it.)
+const int32_t kMaxCapacity = 0x7ffffff5;
+
+int32_t getGrowCapacity(int32_t newLength) {
+  int32_t growSize = (newLength >> 2) + kGrowSize;
+  if(growSize <= (kMaxCapacity - newLength)) {
+    return newLength + growSize;
+  } else {
+    return kMaxCapacity;
+  }
+}
+
+}  // namespace
+
+UBool
+UnicodeString::allocate(int32_t capacity) {
+  if(capacity <= US_STACKBUF_SIZE) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    return TRUE;
+  }
+  if(capacity <= kMaxCapacity) {
+    ++capacity;  // for the NUL
+    // Switch to size_t which is unsigned so that we can allocate up to 4GB.
+    // Reference counter + UChars.
+    size_t numBytes = sizeof(int32_t) + (size_t)capacity * U_SIZEOF_UCHAR;
+    // Round up to a multiple of 16.
+    numBytes = (numBytes + 15) & ~15;
+    int32_t *array = (int32_t *) uprv_malloc(numBytes);
+    if(array != NULL) {
+      // set initial refCount and point behind the refCount
+      *array++ = 1;
+      numBytes -= sizeof(int32_t);
+
+      // have fArray point to the first UChar
+      fUnion.fFields.fArray = (UChar *)array;
+      fUnion.fFields.fCapacity = (int32_t)(numBytes / U_SIZEOF_UCHAR);
+      fUnion.fFields.fLengthAndFlags = kLongString;
+      return TRUE;
+    }
+  }
+  fUnion.fFields.fLengthAndFlags = kIsBogus;
+  fUnion.fFields.fArray = 0;
+  fUnion.fFields.fCapacity = 0;
+  return FALSE;
+}
+
+//========================================
+// Destructor
+//========================================
+
+#ifdef UNISTR_COUNT_FINAL_STRING_LENGTHS
+static u_atomic_int32_t finalLengthCounts[0x400];  // UnicodeString::kMaxShortLength+1
+static u_atomic_int32_t beyondCount(0);
+
+U_CAPI void unistr_printLengths() {
+  int32_t i;
+  for(i = 0; i <= 59; ++i) {
+    printf("%2d,  %9d\n", i, (int32_t)finalLengthCounts[i]);
+  }
+  int32_t beyond = beyondCount;
+  for(; i < UPRV_LENGTHOF(finalLengthCounts); ++i) {
+    beyond += finalLengthCounts[i];
+  }
+  printf(">59, %9d\n", beyond);
+}
+#endif
+
+UnicodeString::~UnicodeString()
+{
+#ifdef UNISTR_COUNT_FINAL_STRING_LENGTHS
+  // Count lengths of strings at the end of their lifetime.
+  // Useful for discussion of a desirable stack buffer size.
+  // Count the contents length, not the optional NUL terminator nor further capacity.
+  // Ignore open-buffer strings and strings which alias external storage.
+  if((fUnion.fFields.fLengthAndFlags&(kOpenGetBuffer|kReadonlyAlias|kWritableAlias)) == 0) {
+    if(hasShortLength()) {
+      umtx_atomic_inc(finalLengthCounts + getShortLength());
+    } else {
+      umtx_atomic_inc(&beyondCount);
+    }
+  }
+#endif
+
+  releaseArray();
+}
+
+//========================================
+// Factory methods
+//========================================
+
+UnicodeString UnicodeString::fromUTF8(StringPiece utf8) {
+  UnicodeString result;
+  result.setToUTF8(utf8);
+  return result;
+}
+
+UnicodeString UnicodeString::fromUTF32(const UChar32 *utf32, int32_t length) {
+  UnicodeString result;
+  int32_t capacity;
+  // Most UTF-32 strings will be BMP-only and result in a same-length
+  // UTF-16 string. We overestimate the capacity just slightly,
+  // just in case there are a few supplementary characters.
+  if(length <= US_STACKBUF_SIZE) {
+    capacity = US_STACKBUF_SIZE;
+  } else {
+    capacity = length + (length >> 4) + 4;
+  }
+  do {
+    UChar *utf16 = result.getBuffer(capacity);
+    int32_t length16;
+    UErrorCode errorCode = U_ZERO_ERROR;
+    u_strFromUTF32WithSub(utf16, result.getCapacity(), &length16,
+        utf32, length,
+        0xfffd,  // Substitution character.
+        NULL,    // Don't care about number of substitutions.
+        &errorCode);
+    result.releaseBuffer(length16);
+    if(errorCode == U_BUFFER_OVERFLOW_ERROR) {
+      capacity = length16 + 1;  // +1 for the terminating NUL.
+      continue;
+    } else if(U_FAILURE(errorCode)) {
+      result.setToBogus();
+    }
+    break;
+  } while(TRUE);
+  return result;
+}
+
+//========================================
+// Assignment
+//========================================
+
+UnicodeString &
+UnicodeString::operator=(const UnicodeString &src) {
+  return copyFrom(src);
+}
+
+UnicodeString &
+UnicodeString::fastCopyFrom(const UnicodeString &src) {
+  return copyFrom(src, TRUE);
+}
+
+UnicodeString &
+UnicodeString::copyFrom(const UnicodeString &src, UBool fastCopy) {
+  // if assigning to ourselves, do nothing
+  if(this == &src) {
+    return *this;
+  }
+
+  // is the right side bogus?
+  if(src.isBogus()) {
+    setToBogus();
+    return *this;
+  }
+
+  // delete the current contents
+  releaseArray();
+
+  if(src.isEmpty()) {
+    // empty string - use the stack buffer
+    setToEmpty();
+    return *this;
+  }
+
+  // fLength>0 and not an "open" src.getBuffer(minCapacity)
+  fUnion.fFields.fLengthAndFlags = src.fUnion.fFields.fLengthAndFlags;
+  switch(src.fUnion.fFields.fLengthAndFlags & kAllStorageFlags) {
+  case kShortString:
+    // short string using the stack buffer, do the same
+    uprv_memcpy(fUnion.fStackFields.fBuffer, src.fUnion.fStackFields.fBuffer,
+                getShortLength() * U_SIZEOF_UCHAR);
+    break;
+  case kLongString:
+    // src uses a refCounted string buffer, use that buffer with refCount
+    // src is const, use a cast - we don't actually change it
+    ((UnicodeString &)src).addRef();
+    // copy all fields, share the reference-counted buffer
+    fUnion.fFields.fArray = src.fUnion.fFields.fArray;
+    fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
+    if(!hasShortLength()) {
+      fUnion.fFields.fLength = src.fUnion.fFields.fLength;
+    }
+    break;
+  case kReadonlyAlias:
+    if(fastCopy) {
+      // src is a readonly alias, do the same
+      // -> maintain the readonly alias as such
+      fUnion.fFields.fArray = src.fUnion.fFields.fArray;
+      fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
+      if(!hasShortLength()) {
+        fUnion.fFields.fLength = src.fUnion.fFields.fLength;
+      }
+      break;
+    }
+    // else if(!fastCopy) fall through to case kWritableAlias
+    // -> allocate a new buffer and copy the contents
+    U_FALLTHROUGH;
+  case kWritableAlias: {
+    // src is a writable alias; we make a copy of that instead
+    int32_t srcLength = src.length();
+    if(allocate(srcLength)) {
+      u_memcpy(getArrayStart(), src.getArrayStart(), srcLength);
+      setLength(srcLength);
+      break;
+    }
+    // if there is not enough memory, then fall through to setting to bogus
+    U_FALLTHROUGH;
+  }
+  default:
+    // if src is bogus, set ourselves to bogus
+    // do not call setToBogus() here because fArray and flags are not consistent here
+    fUnion.fFields.fLengthAndFlags = kIsBogus;
+    fUnion.fFields.fArray = 0;
+    fUnion.fFields.fCapacity = 0;
+    break;
+  }
+
+  return *this;
+}
+
+UnicodeString &UnicodeString::operator=(UnicodeString &&src) U_NOEXCEPT {
+  // No explicit check for self move assignment, consistent with standard library.
+  // Self move assignment causes no crash nor leak but might make the object bogus.
+  releaseArray();
+  copyFieldsFrom(src, TRUE);
+  return *this;
+}
+
+// Same as move assignment except without memory management.
+void UnicodeString::copyFieldsFrom(UnicodeString &src, UBool setSrcToBogus) U_NOEXCEPT {
+  int16_t lengthAndFlags = fUnion.fFields.fLengthAndFlags = src.fUnion.fFields.fLengthAndFlags;
+  if(lengthAndFlags & kUsingStackBuffer) {
+    // Short string using the stack buffer, copy the contents.
+    // Check for self assignment to prevent "overlap in memcpy" warnings,
+    // although it should be harmless to copy a buffer to itself exactly.
+    if(this != &src) {
+      uprv_memcpy(fUnion.fStackFields.fBuffer, src.fUnion.fStackFields.fBuffer,
+                  getShortLength() * U_SIZEOF_UCHAR);
+    }
+  } else {
+    // In all other cases, copy all fields.
+    fUnion.fFields.fArray = src.fUnion.fFields.fArray;
+    fUnion.fFields.fCapacity = src.fUnion.fFields.fCapacity;
+    if(!hasShortLength()) {
+      fUnion.fFields.fLength = src.fUnion.fFields.fLength;
+    }
+    if(setSrcToBogus) {
+      // Set src to bogus without releasing any memory.
+      src.fUnion.fFields.fLengthAndFlags = kIsBogus;
+      src.fUnion.fFields.fArray = NULL;
+      src.fUnion.fFields.fCapacity = 0;
+    }
+  }
+}
+
+void UnicodeString::swap(UnicodeString &other) U_NOEXCEPT {
+  UnicodeString temp;  // Empty short string: Known not to need releaseArray().
+  // Copy fields without resetting source values in between.
+  temp.copyFieldsFrom(*this, FALSE);
+  this->copyFieldsFrom(other, FALSE);
+  other.copyFieldsFrom(temp, FALSE);
+  // Set temp to an empty string so that other's memory is not released twice.
+  temp.fUnion.fFields.fLengthAndFlags = kShortString;
+}
+
+//========================================
+// Miscellaneous operations
+//========================================
+
+UnicodeString UnicodeString::unescape() const {
+    UnicodeString result(length(), (UChar32)0, (int32_t)0); // construct with capacity
+    if (result.isBogus()) {
+        return result;
+    }
+    const UChar *array = getBuffer();
+    int32_t len = length();
+    int32_t prev = 0;
+    for (int32_t i=0;;) {
+        if (i == len) {
+            result.append(array, prev, len - prev);
+            break;
+        }
+        if (array[i++] == 0x5C /*'\\'*/) {
+            result.append(array, prev, (i - 1) - prev);
+            UChar32 c = unescapeAt(i); // advances i
+            if (c < 0) {
+                result.remove(); // return empty string
+                break; // invalid escape sequence
+            }
+            result.append(c);
+            prev = i;
+        }
+    }
+    return result;
+}
+
+UChar32 UnicodeString::unescapeAt(int32_t &offset) const {
+    return u_unescapeAt(UnicodeString_charAt, &offset, length(), (void*)this);
+}
+
+//========================================
+// Read-only implementation
+//========================================
+UBool
+UnicodeString::doEquals(const UnicodeString &text, int32_t len) const {
+  // Requires: this & text not bogus and have same lengths.
+  // Byte-wise comparison works for equality regardless of endianness.
+  return uprv_memcmp(getArrayStart(), text.getArrayStart(), len * U_SIZEOF_UCHAR) == 0;
+}
+
+int8_t
+UnicodeString::doCompare( int32_t start,
+              int32_t length,
+              const UChar *srcChars,
+              int32_t srcStart,
+              int32_t srcLength) const
+{
+  // compare illegal string values
+  if(isBogus()) {
+    return -1;
+  }
+  
+  // pin indices to legal values
+  pinIndices(start, length);
+
+  if(srcChars == NULL) {
+    // treat const UChar *srcChars==NULL as an empty string
+    return length == 0 ? 0 : 1;
+  }
+
+  // get the correct pointer
+  const UChar *chars = getArrayStart();
+
+  chars += start;
+  srcChars += srcStart;
+
+  int32_t minLength;
+  int8_t lengthResult;
+
+  // get the srcLength if necessary
+  if(srcLength < 0) {
+    srcLength = u_strlen(srcChars + srcStart);
+  }
+
+  // are we comparing different lengths?
+  if(length != srcLength) {
+    if(length < srcLength) {
+      minLength = length;
+      lengthResult = -1;
+    } else {
+      minLength = srcLength;
+      lengthResult = 1;
+    }
+  } else {
+    minLength = length;
+    lengthResult = 0;
+  }
+
+  /*
+   * note that uprv_memcmp() returns an int but we return an int8_t;
+   * we need to take care not to truncate the result -
+   * one way to do this is to right-shift the value to
+   * move the sign bit into the lower 8 bits and making sure that this
+   * does not become 0 itself
+   */
+
+  if(minLength > 0 && chars != srcChars) {
+    int32_t result;
+
+#   if U_IS_BIG_ENDIAN 
+      // big-endian: byte comparison works
+      result = uprv_memcmp(chars, srcChars, minLength * sizeof(UChar));
+      if(result != 0) {
+        return (int8_t)(result >> 15 | 1);
+      }
+#   else
+      // little-endian: compare UChar units
+      do {
+        result = ((int32_t)*(chars++) - (int32_t)*(srcChars++));
+        if(result != 0) {
+          return (int8_t)(result >> 15 | 1);
+        }
+      } while(--minLength > 0);
+#   endif
+  }
+  return lengthResult;
+}
+
+/* String compare in code point order - doCompare() compares in code unit order. */
+int8_t
+UnicodeString::doCompareCodePointOrder(int32_t start,
+                                       int32_t length,
+                                       const UChar *srcChars,
+                                       int32_t srcStart,
+                                       int32_t srcLength) const
+{
+  // compare illegal string values
+  // treat const UChar *srcChars==NULL as an empty string
+  if(isBogus()) {
+    return -1;
+  }
+
+  // pin indices to legal values
+  pinIndices(start, length);
+
+  if(srcChars == NULL) {
+    srcStart = srcLength = 0;
+  }
+
+  int32_t diff = uprv_strCompare(getArrayStart() + start, length, (srcChars!=NULL)?(srcChars + srcStart):NULL, srcLength, FALSE, TRUE);
+  /* translate the 32-bit result into an 8-bit one */
+  if(diff!=0) {
+    return (int8_t)(diff >> 15 | 1);
+  } else {
+    return 0;
+  }
+}
+
+int32_t
+UnicodeString::getLength() const {
+    return length();
+}
+
+UChar
+UnicodeString::getCharAt(int32_t offset) const {
+  return charAt(offset);
+}
+
+UChar32
+UnicodeString::getChar32At(int32_t offset) const {
+  return char32At(offset);
+}
+
+UChar32
+UnicodeString::char32At(int32_t offset) const
+{
+  int32_t len = length();
+  if((uint32_t)offset < (uint32_t)len) {
+    const UChar *array = getArrayStart();
+    UChar32 c;
+    U16_GET(array, 0, offset, len, c);
+    return c;
+  } else {
+    return kInvalidUChar;
+  }
+}
+
+int32_t
+UnicodeString::getChar32Start(int32_t offset) const {
+  if((uint32_t)offset < (uint32_t)length()) {
+    const UChar *array = getArrayStart();
+    U16_SET_CP_START(array, 0, offset);
+    return offset;
+  } else {
+    return 0;
+  }
+}
+
+int32_t
+UnicodeString::getChar32Limit(int32_t offset) const {
+  int32_t len = length();
+  if((uint32_t)offset < (uint32_t)len) {
+    const UChar *array = getArrayStart();
+    U16_SET_CP_LIMIT(array, 0, offset, len);
+    return offset;
+  } else {
+    return len;
+  }
+}
+
+int32_t
+UnicodeString::countChar32(int32_t start, int32_t length) const {
+  pinIndices(start, length);
+  // if(isBogus()) then fArray==0 and start==0 - u_countChar32() checks for NULL
+  return u_countChar32(getArrayStart()+start, length);
+}
+
+UBool
+UnicodeString::hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const {
+  pinIndices(start, length);
+  // if(isBogus()) then fArray==0 and start==0 - u_strHasMoreChar32Than() checks for NULL
+  return u_strHasMoreChar32Than(getArrayStart()+start, length, number);
+}
+
+int32_t
+UnicodeString::moveIndex32(int32_t index, int32_t delta) const {
+  // pin index
+  int32_t len = length();
+  if(index<0) {
+    index=0;
+  } else if(index>len) {
+    index=len;
+  }
+
+  const UChar *array = getArrayStart();
+  if(delta>0) {
+    U16_FWD_N(array, index, len, delta);
+  } else {
+    U16_BACK_N(array, 0, index, -delta);
+  }
+
+  return index;
+}
+
+void
+UnicodeString::doExtract(int32_t start,
+             int32_t length,
+             UChar *dst,
+             int32_t dstStart) const
+{
+  // pin indices to legal values
+  pinIndices(start, length);
+
+  // do not copy anything if we alias dst itself
+  const UChar *array = getArrayStart();
+  if(array + start != dst + dstStart) {
+    us_arrayCopy(array, start, dst, dstStart, length);
+  }
+}
+
+int32_t
+UnicodeString::extract(Char16Ptr dest, int32_t destCapacity,
+                       UErrorCode &errorCode) const {
+  int32_t len = length();
+  if(U_SUCCESS(errorCode)) {
+    if(isBogus() || destCapacity<0 || (destCapacity>0 && dest==0)) {
+      errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    } else {
+      const UChar *array = getArrayStart();
+      if(len>0 && len<=destCapacity && array!=dest) {
+        u_memcpy(dest, array, len);
+      }
+      return u_terminateUChars(dest, destCapacity, len, &errorCode);
+    }
+  }
+
+  return len;
+}
+
+int32_t
+UnicodeString::extract(int32_t start,
+                       int32_t length,
+                       char *target,
+                       int32_t targetCapacity,
+                       enum EInvariant) const
+{
+  // if the arguments are illegal, then do nothing
+  if(targetCapacity < 0 || (targetCapacity > 0 && target == NULL)) {
+    return 0;
+  }
+
+  // pin the indices to legal values
+  pinIndices(start, length);
+
+  if(length <= targetCapacity) {
+    u_UCharsToChars(getArrayStart() + start, target, length);
+  }
+  UErrorCode status = U_ZERO_ERROR;
+  return u_terminateChars(target, targetCapacity, length, &status);
+}
+
+UnicodeString
+UnicodeString::tempSubString(int32_t start, int32_t len) const {
+  pinIndices(start, len);
+  const UChar *array = getBuffer();  // not getArrayStart() to check kIsBogus & kOpenGetBuffer
+  if(array==NULL) {
+    array=fUnion.fStackFields.fBuffer;  // anything not NULL because that would make an empty string
+    len=-2;  // bogus result string
+  }
+  return UnicodeString(FALSE, array + start, len);
+}
+
+int32_t
+UnicodeString::toUTF8(int32_t start, int32_t len,
+                      char *target, int32_t capacity) const {
+  pinIndices(start, len);
+  int32_t length8;
+  UErrorCode errorCode = U_ZERO_ERROR;
+  u_strToUTF8WithSub(target, capacity, &length8,
+                     getBuffer() + start, len,
+                     0xFFFD,  // Standard substitution character.
+                     NULL,    // Don't care about number of substitutions.
+                     &errorCode);
+  return length8;
+}
+
+#if U_CHARSET_IS_UTF8
+
+int32_t
+UnicodeString::extract(int32_t start, int32_t len,
+                       char *target, uint32_t dstSize) const {
+  // if the arguments are illegal, then do nothing
+  if(/*dstSize < 0 || */(dstSize > 0 && target == 0)) {
+    return 0;
+  }
+  return toUTF8(start, len, target, dstSize <= 0x7fffffff ? (int32_t)dstSize : 0x7fffffff);
+}
+
+// else see unistr_cnv.cpp
+#endif
+
+void 
+UnicodeString::extractBetween(int32_t start,
+                  int32_t limit,
+                  UnicodeString& target) const {
+  pinIndex(start);
+  pinIndex(limit);
+  doExtract(start, limit - start, target);
+}
+
+// When converting from UTF-16 to UTF-8, the result will have at most 3 times
+// as many bytes as the source has UChars.
+// The "worst cases" are writing systems like Indic, Thai and CJK with
+// 3:1 bytes:UChars.
+void
+UnicodeString::toUTF8(ByteSink &sink) const {
+  int32_t length16 = length();
+  if(length16 != 0) {
+    char stackBuffer[1024];
+    int32_t capacity = (int32_t)sizeof(stackBuffer);
+    UBool utf8IsOwned = FALSE;
+    char *utf8 = sink.GetAppendBuffer(length16 < capacity ? length16 : capacity,
+                                      3*length16,
+                                      stackBuffer, capacity,
+                                      &capacity);
+    int32_t length8 = 0;
+    UErrorCode errorCode = U_ZERO_ERROR;
+    u_strToUTF8WithSub(utf8, capacity, &length8,
+                       getBuffer(), length16,
+                       0xFFFD,  // Standard substitution character.
+                       NULL,    // Don't care about number of substitutions.
+                       &errorCode);
+    if(errorCode == U_BUFFER_OVERFLOW_ERROR) {
+      utf8 = (char *)uprv_malloc(length8);
+      if(utf8 != NULL) {
+        utf8IsOwned = TRUE;
+        errorCode = U_ZERO_ERROR;
+        u_strToUTF8WithSub(utf8, length8, &length8,
+                           getBuffer(), length16,
+                           0xFFFD,  // Standard substitution character.
+                           NULL,    // Don't care about number of substitutions.
+                           &errorCode);
+      } else {
+        errorCode = U_MEMORY_ALLOCATION_ERROR;
+      }
+    }
+    if(U_SUCCESS(errorCode)) {
+      sink.Append(utf8, length8);
+      sink.Flush();
+    }
+    if(utf8IsOwned) {
+      uprv_free(utf8);
+    }
+  }
+}
+
+int32_t
+UnicodeString::toUTF32(UChar32 *utf32, int32_t capacity, UErrorCode &errorCode) const {
+  int32_t length32=0;
+  if(U_SUCCESS(errorCode)) {
+    // getBuffer() and u_strToUTF32WithSub() check for illegal arguments.
+    u_strToUTF32WithSub(utf32, capacity, &length32,
+        getBuffer(), length(),
+        0xfffd,  // Substitution character.
+        NULL,    // Don't care about number of substitutions.
+        &errorCode);
+  }
+  return length32;
+}
+
+int32_t 
+UnicodeString::indexOf(const UChar *srcChars,
+               int32_t srcStart,
+               int32_t srcLength,
+               int32_t start,
+               int32_t length) const
+{
+  if(isBogus() || srcChars == 0 || srcStart < 0 || srcLength == 0) {
+    return -1;
+  }
+
+  // UnicodeString does not find empty substrings
+  if(srcLength < 0 && srcChars[srcStart] == 0) {
+    return -1;
+  }
+
+  // get the indices within bounds
+  pinIndices(start, length);
+
+  // find the first occurrence of the substring
+  const UChar *array = getArrayStart();
+  const UChar *match = u_strFindFirst(array + start, length, srcChars + srcStart, srcLength);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+int32_t
+UnicodeString::doIndexOf(UChar c,
+             int32_t start,
+             int32_t length) const
+{
+  // pin indices
+  pinIndices(start, length);
+
+  // find the first occurrence of c
+  const UChar *array = getArrayStart();
+  const UChar *match = u_memchr(array + start, c, length);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+int32_t
+UnicodeString::doIndexOf(UChar32 c,
+                         int32_t start,
+                         int32_t length) const {
+  // pin indices
+  pinIndices(start, length);
+
+  // find the first occurrence of c
+  const UChar *array = getArrayStart();
+  const UChar *match = u_memchr32(array + start, c, length);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+int32_t 
+UnicodeString::lastIndexOf(const UChar *srcChars,
+               int32_t srcStart,
+               int32_t srcLength,
+               int32_t start,
+               int32_t length) const
+{
+  if(isBogus() || srcChars == 0 || srcStart < 0 || srcLength == 0) {
+    return -1;
+  }
+
+  // UnicodeString does not find empty substrings
+  if(srcLength < 0 && srcChars[srcStart] == 0) {
+    return -1;
+  }
+
+  // get the indices within bounds
+  pinIndices(start, length);
+
+  // find the last occurrence of the substring
+  const UChar *array = getArrayStart();
+  const UChar *match = u_strFindLast(array + start, length, srcChars + srcStart, srcLength);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+int32_t
+UnicodeString::doLastIndexOf(UChar c,
+                 int32_t start,
+                 int32_t length) const
+{
+  if(isBogus()) {
+    return -1;
+  }
+
+  // pin indices
+  pinIndices(start, length);
+
+  // find the last occurrence of c
+  const UChar *array = getArrayStart();
+  const UChar *match = u_memrchr(array + start, c, length);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+int32_t
+UnicodeString::doLastIndexOf(UChar32 c,
+                             int32_t start,
+                             int32_t length) const {
+  // pin indices
+  pinIndices(start, length);
+
+  // find the last occurrence of c
+  const UChar *array = getArrayStart();
+  const UChar *match = u_memrchr32(array + start, c, length);
+  if(match == NULL) {
+    return -1;
+  } else {
+    return (int32_t)(match - array);
+  }
+}
+
+//========================================
+// Write implementation
+//========================================
+
+UnicodeString& 
+UnicodeString::findAndReplace(int32_t start,
+                  int32_t length,
+                  const UnicodeString& oldText,
+                  int32_t oldStart,
+                  int32_t oldLength,
+                  const UnicodeString& newText,
+                  int32_t newStart,
+                  int32_t newLength)
+{
+  if(isBogus() || oldText.isBogus() || newText.isBogus()) {
+    return *this;
+  }
+
+  pinIndices(start, length);
+  oldText.pinIndices(oldStart, oldLength);
+  newText.pinIndices(newStart, newLength);
+
+  if(oldLength == 0) {
+    return *this;
+  }
+
+  while(length > 0 && length >= oldLength) {
+    int32_t pos = indexOf(oldText, oldStart, oldLength, start, length);
+    if(pos < 0) {
+      // no more oldText's here: done
+      break;
+    } else {
+      // we found oldText, replace it by newText and go beyond it
+      replace(pos, oldLength, newText, newStart, newLength);
+      length -= pos + oldLength - start;
+      start = pos + newLength;
+    }
+  }
+
+  return *this;
+}
+
+
+void
+UnicodeString::setToBogus()
+{
+  releaseArray();
+
+  fUnion.fFields.fLengthAndFlags = kIsBogus;
+  fUnion.fFields.fArray = 0;
+  fUnion.fFields.fCapacity = 0;
+}
+
+// turn a bogus string into an empty one
+void
+UnicodeString::unBogus() {
+  if(fUnion.fFields.fLengthAndFlags & kIsBogus) {
+    setToEmpty();
+  }
+}
+
+const char16_t *
+UnicodeString::getTerminatedBuffer() {
+  if(!isWritable()) {
+    return nullptr;
+  }
+  UChar *array = getArrayStart();
+  int32_t len = length();
+  if(len < getCapacity()) {
+    if(fUnion.fFields.fLengthAndFlags & kBufferIsReadonly) {
+      // If len<capacity on a read-only alias, then array[len] is
+      // either the original NUL (if constructed with (TRUE, s, length))
+      // or one of the original string contents characters (if later truncated),
+      // therefore we can assume that array[len] is initialized memory.
+      if(array[len] == 0) {
+        return array;
+      }
+    } else if(((fUnion.fFields.fLengthAndFlags & kRefCounted) == 0 || refCount() == 1)) {
+      // kRefCounted: Do not write the NUL if the buffer is shared.
+      // That is mostly safe, except when the length of one copy was modified
+      // without copy-on-write, e.g., via truncate(newLength) or remove(void).
+      // Then the NUL would be written into the middle of another copy's string.
+
+      // Otherwise, the buffer is fully writable and it is anyway safe to write the NUL.
+      // Do not test if there is a NUL already because it might be uninitialized memory.
+      // (That would be safe, but tools like valgrind & Purify would complain.)
+      array[len] = 0;
+      return array;
+    }
+  }
+  if(len<INT32_MAX && cloneArrayIfNeeded(len+1)) {
+    array = getArrayStart();
+    array[len] = 0;
+    return array;
+  } else {
+    return nullptr;
+  }
+}
+
+// setTo() analogous to the readonly-aliasing constructor with the same signature
+UnicodeString &
+UnicodeString::setTo(UBool isTerminated,
+                     ConstChar16Ptr textPtr,
+                     int32_t textLength)
+{
+  if(fUnion.fFields.fLengthAndFlags & kOpenGetBuffer) {
+    // do not modify a string that has an "open" getBuffer(minCapacity)
+    return *this;
+  }
+
+  const UChar *text = textPtr;
+  if(text == NULL) {
+    // treat as an empty string, do not alias
+    releaseArray();
+    setToEmpty();
+    return *this;
+  }
+
+  if( textLength < -1 ||
+      (textLength == -1 && !isTerminated) ||
+      (textLength >= 0 && isTerminated && text[textLength] != 0)
+  ) {
+    setToBogus();
+    return *this;
+  }
+
+  releaseArray();
+
+  if(textLength == -1) {
+    // text is terminated, or else it would have failed the above test
+    textLength = u_strlen(text);
+  }
+  fUnion.fFields.fLengthAndFlags = kReadonlyAlias;
+  setArray((UChar *)text, textLength, isTerminated ? textLength + 1 : textLength);
+  return *this;
+}
+
+// setTo() analogous to the writable-aliasing constructor with the same signature
+UnicodeString &
+UnicodeString::setTo(UChar *buffer,
+                     int32_t buffLength,
+                     int32_t buffCapacity) {
+  if(fUnion.fFields.fLengthAndFlags & kOpenGetBuffer) {
+    // do not modify a string that has an "open" getBuffer(minCapacity)
+    return *this;
+  }
+
+  if(buffer == NULL) {
+    // treat as an empty string, do not alias
+    releaseArray();
+    setToEmpty();
+    return *this;
+  }
+
+  if(buffLength < -1 || buffCapacity < 0 || buffLength > buffCapacity) {
+    setToBogus();
+    return *this;
+  } else if(buffLength == -1) {
+    // buffLength = u_strlen(buff); but do not look beyond buffCapacity
+    const UChar *p = buffer, *limit = buffer + buffCapacity;
+    while(p != limit && *p != 0) {
+      ++p;
+    }
+    buffLength = (int32_t)(p - buffer);
+  }
+
+  releaseArray();
+
+  fUnion.fFields.fLengthAndFlags = kWritableAlias;
+  setArray(buffer, buffLength, buffCapacity);
+  return *this;
+}
+
+UnicodeString &UnicodeString::setToUTF8(StringPiece utf8) {
+  unBogus();
+  int32_t length = utf8.length();
+  int32_t capacity;
+  // The UTF-16 string will be at most as long as the UTF-8 string.
+  if(length <= US_STACKBUF_SIZE) {
+    capacity = US_STACKBUF_SIZE;
+  } else {
+    capacity = length + 1;  // +1 for the terminating NUL.
+  }
+  UChar *utf16 = getBuffer(capacity);
+  int32_t length16;
+  UErrorCode errorCode = U_ZERO_ERROR;
+  u_strFromUTF8WithSub(utf16, getCapacity(), &length16,
+      utf8.data(), length,
+      0xfffd,  // Substitution character.
+      NULL,    // Don't care about number of substitutions.
+      &errorCode);
+  releaseBuffer(length16);
+  if(U_FAILURE(errorCode)) {
+    setToBogus();
+  }
+  return *this;
+}
+
+UnicodeString&
+UnicodeString::setCharAt(int32_t offset,
+             UChar c)
+{
+  int32_t len = length();
+  if(cloneArrayIfNeeded() && len > 0) {
+    if(offset < 0) {
+      offset = 0;
+    } else if(offset >= len) {
+      offset = len - 1;
+    }
+
+    getArrayStart()[offset] = c;
+  }
+  return *this;
+}
+
+UnicodeString&
+UnicodeString::replace(int32_t start,
+               int32_t _length,
+               UChar32 srcChar) {
+  UChar buffer[U16_MAX_LENGTH];
+  int32_t count = 0;
+  UBool isError = FALSE;
+  U16_APPEND(buffer, count, U16_MAX_LENGTH, srcChar, isError);
+  // We test isError so that the compiler does not complain that we don't.
+  // If isError (srcChar is not a valid code point) then count==0 which means
+  // we remove the source segment rather than replacing it with srcChar.
+  return doReplace(start, _length, buffer, 0, isError ? 0 : count);
+}
+
+UnicodeString&
+UnicodeString::append(UChar32 srcChar) {
+  UChar buffer[U16_MAX_LENGTH];
+  int32_t _length = 0;
+  UBool isError = FALSE;
+  U16_APPEND(buffer, _length, U16_MAX_LENGTH, srcChar, isError);
+  // We test isError so that the compiler does not complain that we don't.
+  // If isError then _length==0 which turns the doAppend() into a no-op anyway.
+  return isError ? *this : doAppend(buffer, 0, _length);
+}
+
+UnicodeString&
+UnicodeString::doReplace( int32_t start,
+              int32_t length,
+              const UnicodeString& src,
+              int32_t srcStart,
+              int32_t srcLength)
+{
+  // pin the indices to legal values
+  src.pinIndices(srcStart, srcLength);
+
+  // get the characters from src
+  // and replace the range in ourselves with them
+  return doReplace(start, length, src.getArrayStart(), srcStart, srcLength);
+}
+
+UnicodeString&
+UnicodeString::doReplace(int32_t start,
+             int32_t length,
+             const UChar *srcChars,
+             int32_t srcStart,
+             int32_t srcLength)
+{
+  if(!isWritable()) {
+    return *this;
+  }
+
+  int32_t oldLength = this->length();
+
+  // optimize (read-only alias).remove(0, start) and .remove(start, end)
+  if((fUnion.fFields.fLengthAndFlags&kBufferIsReadonly) && srcLength == 0) {
+    if(start == 0) {
+      // remove prefix by adjusting the array pointer
+      pinIndex(length);
+      fUnion.fFields.fArray += length;
+      fUnion.fFields.fCapacity -= length;
+      setLength(oldLength - length);
+      return *this;
+    } else {
+      pinIndex(start);
+      if(length >= (oldLength - start)) {
+        // remove suffix by reducing the length (like truncate())
+        setLength(start);
+        fUnion.fFields.fCapacity = start;  // not NUL-terminated any more
+        return *this;
+      }
+    }
+  }
+
+  if(start == oldLength) {
+    return doAppend(srcChars, srcStart, srcLength);
+  }
+
+  if(srcChars == 0) {
+    srcLength = 0;
+  } else {
+    // Perform all remaining operations relative to srcChars + srcStart.
+    // From this point forward, do not use srcStart.
+    srcChars += srcStart;
+    if (srcLength < 0) {
+      // get the srcLength if necessary
+      srcLength = u_strlen(srcChars);
+    }
+  }
+
+  // pin the indices to legal values
+  pinIndices(start, length);
+
+  // Calculate the size of the string after the replace.
+  // Avoid int32_t overflow.
+  int32_t newLength = oldLength - length;
+  if(srcLength > (INT32_MAX - newLength)) {
+    setToBogus();
+    return *this;
+  }
+  newLength += srcLength;
+
+  // Check for insertion into ourself
+  const UChar *oldArray = getArrayStart();
+  if (isBufferWritable() &&
+      oldArray < srcChars + srcLength &&
+      srcChars < oldArray + oldLength) {
+    // Copy into a new UnicodeString and start over
+    UnicodeString copy(srcChars, srcLength);
+    if (copy.isBogus()) {
+      setToBogus();
+      return *this;
+    }
+    return doReplace(start, length, copy.getArrayStart(), 0, srcLength);
+  }
+
+  // cloneArrayIfNeeded(doCopyArray=FALSE) may change fArray but will not copy the current contents;
+  // therefore we need to keep the current fArray
+  UChar oldStackBuffer[US_STACKBUF_SIZE];
+  if((fUnion.fFields.fLengthAndFlags&kUsingStackBuffer) && (newLength > US_STACKBUF_SIZE)) {
+    // copy the stack buffer contents because it will be overwritten with
+    // fUnion.fFields values
+    u_memcpy(oldStackBuffer, oldArray, oldLength);
+    oldArray = oldStackBuffer;
+  }
+
+  // clone our array and allocate a bigger array if needed
+  int32_t *bufferToDelete = 0;
+  if(!cloneArrayIfNeeded(newLength, getGrowCapacity(newLength),
+                         FALSE, &bufferToDelete)
+  ) {
+    return *this;
+  }
+
+  // now do the replace
+
+  UChar *newArray = getArrayStart();
+  if(newArray != oldArray) {
+    // if fArray changed, then we need to copy everything except what will change
+    us_arrayCopy(oldArray, 0, newArray, 0, start);
+    us_arrayCopy(oldArray, start + length,
+                 newArray, start + srcLength,
+                 oldLength - (start + length));
+  } else if(length != srcLength) {
+    // fArray did not change; copy only the portion that isn't changing, leaving a hole
+    us_arrayCopy(oldArray, start + length,
+                 newArray, start + srcLength,
+                 oldLength - (start + length));
+  }
+
+  // now fill in the hole with the new string
+  us_arrayCopy(srcChars, 0, newArray, start, srcLength);
+
+  setLength(newLength);
+
+  // delayed delete in case srcChars == fArray when we started, and
+  // to keep oldArray alive for the above operations
+  if (bufferToDelete) {
+    uprv_free(bufferToDelete);
+  }
+
+  return *this;
+}
+
+// Versions of doReplace() only for append() variants.
+// doReplace() and doAppend() optimize for different cases.
+
+UnicodeString&
+UnicodeString::doAppend(const UnicodeString& src, int32_t srcStart, int32_t srcLength) {
+  if(srcLength == 0) {
+    return *this;
+  }
+
+  // pin the indices to legal values
+  src.pinIndices(srcStart, srcLength);
+  return doAppend(src.getArrayStart(), srcStart, srcLength);
+}
+
+UnicodeString&
+UnicodeString::doAppend(const UChar *srcChars, int32_t srcStart, int32_t srcLength) {
+  if(!isWritable() || srcLength == 0 || srcChars == NULL) {
+    return *this;
+  }
+
+  // Perform all remaining operations relative to srcChars + srcStart.
+  // From this point forward, do not use srcStart.
+  srcChars += srcStart;
+
+  if(srcLength < 0) {
+    // get the srcLength if necessary
+    if((srcLength = u_strlen(srcChars)) == 0) {
+      return *this;
+    }
+  }
+
+  int32_t oldLength = length();
+  int32_t newLength;
+  if (uprv_add32_overflow(oldLength, srcLength, &newLength)) {
+    setToBogus();
+    return *this;
+  }
+
+  // Check for append onto ourself
+  const UChar* oldArray = getArrayStart();
+  if (isBufferWritable() &&
+      oldArray < srcChars + srcLength &&
+      srcChars < oldArray + oldLength) {
+    // Copy into a new UnicodeString and start over
+    UnicodeString copy(srcChars, srcLength);
+    if (copy.isBogus()) {
+      setToBogus();
+      return *this;
+    }
+    return doAppend(copy.getArrayStart(), 0, srcLength);
+  }
+
+  // optimize append() onto a large-enough, owned string
+  if((newLength <= getCapacity() && isBufferWritable()) ||
+      cloneArrayIfNeeded(newLength, getGrowCapacity(newLength))) {
+    UChar *newArray = getArrayStart();
+    // Do not copy characters when
+    //   UChar *buffer=str.getAppendBuffer(...);
+    // is followed by
+    //   str.append(buffer, length);
+    // or
+    //   str.appendString(buffer, length)
+    // or similar.
+    if(srcChars != newArray + oldLength) {
+      us_arrayCopy(srcChars, 0, newArray, oldLength, srcLength);
+    }
+    setLength(newLength);
+  }
+  return *this;
+}
+
+/**
+ * Replaceable API
+ */
+void
+UnicodeString::handleReplaceBetween(int32_t start,
+                                    int32_t limit,
+                                    const UnicodeString& text) {
+    replaceBetween(start, limit, text);
+}
+
+/**
+ * Replaceable API
+ */
+void 
+UnicodeString::copy(int32_t start, int32_t limit, int32_t dest) {
+    if (limit <= start) {
+        return; // Nothing to do; avoid bogus malloc call
+    }
+    UChar* text = (UChar*) uprv_malloc( sizeof(UChar) * (limit - start) );
+    // Check to make sure text is not null.
+    if (text != NULL) {
+	    extractBetween(start, limit, text, 0);
+	    insert(dest, text, 0, limit - start);    
+	    uprv_free(text);
+    }
+}
+
+/**
+ * Replaceable API
+ *
+ * NOTE: This is for the Replaceable class.  There is no rep.cpp,
+ * so we implement this function here.
+ */
+UBool Replaceable::hasMetaData() const {
+    return TRUE;
+}
+
+/**
+ * Replaceable API
+ */
+UBool UnicodeString::hasMetaData() const {
+    return FALSE;
+}
+
+UnicodeString&
+UnicodeString::doReverse(int32_t start, int32_t length) {
+  if(length <= 1 || !cloneArrayIfNeeded()) {
+    return *this;
+  }
+
+  // pin the indices to legal values
+  pinIndices(start, length);
+  if(length <= 1) {  // pinIndices() might have shrunk the length
+    return *this;
+  }
+
+  UChar *left = getArrayStart() + start;
+  UChar *right = left + length - 1;  // -1 for inclusive boundary (length>=2)
+  UChar swap;
+  UBool hasSupplementary = FALSE;
+
+  // Before the loop we know left<right because length>=2.
+  do {
+    hasSupplementary |= (UBool)U16_IS_LEAD(swap = *left);
+    hasSupplementary |= (UBool)U16_IS_LEAD(*left++ = *right);
+    *right-- = swap;
+  } while(left < right);
+  // Make sure to test the middle code unit of an odd-length string.
+  // Redundant if the length is even.
+  hasSupplementary |= (UBool)U16_IS_LEAD(*left);
+
+  /* if there are supplementary code points in the reversed range, then re-swap their surrogates */
+  if(hasSupplementary) {
+    UChar swap2;
+
+    left = getArrayStart() + start;
+    right = left + length - 1; // -1 so that we can look at *(left+1) if left<right
+    while(left < right) {
+      if(U16_IS_TRAIL(swap = *left) && U16_IS_LEAD(swap2 = *(left + 1))) {
+        *left++ = swap2;
+        *left++ = swap;
+      } else {
+        ++left;
+      }
+    }
+  }
+
+  return *this;
+}
+
+UBool 
+UnicodeString::padLeading(int32_t targetLength,
+                          UChar padChar)
+{
+  int32_t oldLength = length();
+  if(oldLength >= targetLength || !cloneArrayIfNeeded(targetLength)) {
+    return FALSE;
+  } else {
+    // move contents up by padding width
+    UChar *array = getArrayStart();
+    int32_t start = targetLength - oldLength;
+    us_arrayCopy(array, 0, array, start, oldLength);
+
+    // fill in padding character
+    while(--start >= 0) {
+      array[start] = padChar;
+    }
+    setLength(targetLength);
+    return TRUE;
+  }
+}
+
+UBool 
+UnicodeString::padTrailing(int32_t targetLength,
+                           UChar padChar)
+{
+  int32_t oldLength = length();
+  if(oldLength >= targetLength || !cloneArrayIfNeeded(targetLength)) {
+    return FALSE;
+  } else {
+    // fill in padding character
+    UChar *array = getArrayStart();
+    int32_t length = targetLength;
+    while(--length >= oldLength) {
+      array[length] = padChar;
+    }
+    setLength(targetLength);
+    return TRUE;
+  }
+}
+
+//========================================
+// Hashing
+//========================================
+int32_t
+UnicodeString::doHashCode() const
+{
+    /* Delegate hash computation to uhash.  This makes UnicodeString
+     * hashing consistent with UChar* hashing.  */
+    int32_t hashCode = ustr_hashUCharsN(getArrayStart(), length());
+    if (hashCode == kInvalidHashCode) {
+        hashCode = kEmptyHashCode;
+    }
+    return hashCode;
+}
+
+//========================================
+// External Buffer
+//========================================
+
+char16_t *
+UnicodeString::getBuffer(int32_t minCapacity) {
+  if(minCapacity>=-1 && cloneArrayIfNeeded(minCapacity)) {
+    fUnion.fFields.fLengthAndFlags|=kOpenGetBuffer;
+    setZeroLength();
+    return getArrayStart();
+  } else {
+    return nullptr;
+  }
+}
+
+void
+UnicodeString::releaseBuffer(int32_t newLength) {
+  if(fUnion.fFields.fLengthAndFlags&kOpenGetBuffer && newLength>=-1) {
+    // set the new fLength
+    int32_t capacity=getCapacity();
+    if(newLength==-1) {
+      // the new length is the string length, capped by fCapacity
+      const UChar *array=getArrayStart(), *p=array, *limit=array+capacity;
+      while(p<limit && *p!=0) {
+        ++p;
+      }
+      newLength=(int32_t)(p-array);
+    } else if(newLength>capacity) {
+      newLength=capacity;
+    }
+    setLength(newLength);
+    fUnion.fFields.fLengthAndFlags&=~kOpenGetBuffer;
+  }
+}
+
+//========================================
+// Miscellaneous
+//========================================
+UBool
+UnicodeString::cloneArrayIfNeeded(int32_t newCapacity,
+                                  int32_t growCapacity,
+                                  UBool doCopyArray,
+                                  int32_t **pBufferToDelete,
+                                  UBool forceClone) {
+  // default parameters need to be static, therefore
+  // the defaults are -1 to have convenience defaults
+  if(newCapacity == -1) {
+    newCapacity = getCapacity();
+  }
+
+  // while a getBuffer(minCapacity) is "open",
+  // prevent any modifications of the string by returning FALSE here
+  // if the string is bogus, then only an assignment or similar can revive it
+  if(!isWritable()) {
+    return FALSE;
+  }
+
+  /*
+   * We need to make a copy of the array if
+   * the buffer is read-only, or
+   * the buffer is refCounted (shared), and refCount>1, or
+   * the buffer is too small.
+   * Return FALSE if memory could not be allocated.
+   */
+  if(forceClone ||
+     fUnion.fFields.fLengthAndFlags & kBufferIsReadonly ||
+     (fUnion.fFields.fLengthAndFlags & kRefCounted && refCount() > 1) ||
+     newCapacity > getCapacity()
+  ) {
+    // check growCapacity for default value and use of the stack buffer
+    if(growCapacity < 0) {
+      growCapacity = newCapacity;
+    } else if(newCapacity <= US_STACKBUF_SIZE && growCapacity > US_STACKBUF_SIZE) {
+      growCapacity = US_STACKBUF_SIZE;
+    }
+
+    // save old values
+    UChar oldStackBuffer[US_STACKBUF_SIZE];
+    UChar *oldArray;
+    int32_t oldLength = length();
+    int16_t flags = fUnion.fFields.fLengthAndFlags;
+
+    if(flags&kUsingStackBuffer) {
+      U_ASSERT(!(flags&kRefCounted)); /* kRefCounted and kUsingStackBuffer are mutally exclusive */
+      if(doCopyArray && growCapacity > US_STACKBUF_SIZE) {
+        // copy the stack buffer contents because it will be overwritten with
+        // fUnion.fFields values
+        us_arrayCopy(fUnion.fStackFields.fBuffer, 0, oldStackBuffer, 0, oldLength);
+        oldArray = oldStackBuffer;
+      } else {
+        oldArray = NULL; // no need to copy from the stack buffer to itself
+      }
+    } else {
+      oldArray = fUnion.fFields.fArray;
+      U_ASSERT(oldArray!=NULL); /* when stack buffer is not used, oldArray must have a non-NULL reference */
+    }
+
+    // allocate a new array
+    if(allocate(growCapacity) ||
+       (newCapacity < growCapacity && allocate(newCapacity))
+    ) {
+      if(doCopyArray) {
+        // copy the contents
+        // do not copy more than what fits - it may be smaller than before
+        int32_t minLength = oldLength;
+        newCapacity = getCapacity();
+        if(newCapacity < minLength) {
+          minLength = newCapacity;
+        }
+        if(oldArray != NULL) {
+          us_arrayCopy(oldArray, 0, getArrayStart(), 0, minLength);
+        }
+        setLength(minLength);
+      } else {
+        setZeroLength();
+      }
+
+      // release the old array
+      if(flags & kRefCounted) {
+        // the array is refCounted; decrement and release if 0
+        u_atomic_int32_t *pRefCount = ((u_atomic_int32_t *)oldArray - 1);
+        if(umtx_atomic_dec(pRefCount) == 0) {
+          if(pBufferToDelete == 0) {
+              // Note: cast to (void *) is needed with MSVC, where u_atomic_int32_t
+              // is defined as volatile. (Volatile has useful non-standard behavior
+              //   with this compiler.)
+            uprv_free((void *)pRefCount);
+          } else {
+            // the caller requested to delete it himself
+            *pBufferToDelete = (int32_t *)pRefCount;
+          }
+        }
+      }
+    } else {
+      // not enough memory for growCapacity and not even for the smaller newCapacity
+      // reset the old values for setToBogus() to release the array
+      if(!(flags&kUsingStackBuffer)) {
+        fUnion.fFields.fArray = oldArray;
+      }
+      fUnion.fFields.fLengthAndFlags = flags;
+      setToBogus();
+      return FALSE;
+    }
+  }
+  return TRUE;
+}
+
+// UnicodeStringAppendable ------------------------------------------------- ***
+
+UnicodeStringAppendable::~UnicodeStringAppendable() {}
+
+UBool
+UnicodeStringAppendable::appendCodeUnit(UChar c) {
+  return str.doAppend(&c, 0, 1).isWritable();
+}
+
+UBool
+UnicodeStringAppendable::appendCodePoint(UChar32 c) {
+  UChar buffer[U16_MAX_LENGTH];
+  int32_t cLength = 0;
+  UBool isError = FALSE;
+  U16_APPEND(buffer, cLength, U16_MAX_LENGTH, c, isError);
+  return !isError && str.doAppend(buffer, 0, cLength).isWritable();
+}
+
+UBool
+UnicodeStringAppendable::appendString(const UChar *s, int32_t length) {
+  return str.doAppend(s, 0, length).isWritable();
+}
+
+UBool
+UnicodeStringAppendable::reserveAppendCapacity(int32_t appendCapacity) {
+  return str.cloneArrayIfNeeded(str.length() + appendCapacity);
+}
+
+UChar *
+UnicodeStringAppendable::getAppendBuffer(int32_t minCapacity,
+                                         int32_t desiredCapacityHint,
+                                         UChar *scratch, int32_t scratchCapacity,
+                                         int32_t *resultCapacity) {
+  if(minCapacity < 1 || scratchCapacity < minCapacity) {
+    *resultCapacity = 0;
+    return NULL;
+  }
+  int32_t oldLength = str.length();
+  if(minCapacity <= (kMaxCapacity - oldLength) &&
+      desiredCapacityHint <= (kMaxCapacity - oldLength) &&
+      str.cloneArrayIfNeeded(oldLength + minCapacity, oldLength + desiredCapacityHint)) {
+    *resultCapacity = str.getCapacity() - oldLength;
+    return str.getArrayStart() + oldLength;
+  }
+  *resultCapacity = scratchCapacity;
+  return scratch;
+}
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI int32_t U_EXPORT2
+uhash_hashUnicodeString(const UElement key) {
+    const UnicodeString *str = (const UnicodeString*) key.pointer;
+    return (str == NULL) ? 0 : str->hashCode();
+}
+
+// Moved here from uhash_us.cpp so that using a UVector of UnicodeString*
+// does not depend on hashtable code.
+U_CAPI UBool U_EXPORT2
+uhash_compareUnicodeString(const UElement key1, const UElement key2) {
+    const UnicodeString *str1 = (const UnicodeString*) key1.pointer;
+    const UnicodeString *str2 = (const UnicodeString*) key2.pointer;
+    if (str1 == str2) {
+        return TRUE;
+    }
+    if (str1 == NULL || str2 == NULL) {
+        return FALSE;
+    }
+    return *str1 == *str2;
+}
+
+#ifdef U_STATIC_IMPLEMENTATION
+/*
+This should never be called. It is defined here to make sure that the
+virtual vector deleting destructor is defined within unistr.cpp.
+The vector deleting destructor is already a part of UObject,
+but defining it here makes sure that it is included with this object file.
+This makes sure that static library dependencies are kept to a minimum.
+*/
+static void uprv_UnicodeStringDummy(void) {
+    delete [] (new UnicodeString[2]);
+}
+#endif
diff --git a/thirdparty/icu4c/common/unistr_case.cpp b/thirdparty/icu4c/common/unistr_case.cpp
new file mode 100644
index 0000000000..2138d60c01
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr_case.cpp
@@ -0,0 +1,250 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unistr_case.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2004aug19
+*   created by: Markus W. Scherer
+*
+*   Case-mapping functions moved here from unistr.cpp
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/brkiter.h"
+#include "unicode/casemap.h"
+#include "unicode/edits.h"
+#include "unicode/putil.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "unicode/ustring.h"
+#include "unicode/unistr.h"
+#include "unicode/uchar.h"
+#include "uassert.h"
+#include "ucasemap_imp.h"
+#include "uelement.h"
+
+U_NAMESPACE_BEGIN
+
+//========================================
+// Read-only implementation
+//========================================
+
+int8_t
+UnicodeString::doCaseCompare(int32_t start,
+                             int32_t length,
+                             const UChar *srcChars,
+                             int32_t srcStart,
+                             int32_t srcLength,
+                             uint32_t options) const
+{
+  // compare illegal string values
+  // treat const UChar *srcChars==NULL as an empty string
+  if(isBogus()) {
+    return -1;
+  }
+
+  // pin indices to legal values
+  pinIndices(start, length);
+
+  if(srcChars == NULL) {
+    srcStart = srcLength = 0;
+  }
+
+  // get the correct pointer
+  const UChar *chars = getArrayStart();
+
+  chars += start;
+  if(srcStart!=0) {
+    srcChars += srcStart;
+  }
+
+  if(chars != srcChars) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    int32_t result=u_strcmpFold(chars, length, srcChars, srcLength,
+                                options|U_COMPARE_IGNORE_CASE, &errorCode);
+    if(result!=0) {
+      return (int8_t)(result >> 24 | 1);
+    }
+  } else {
+    // get the srcLength if necessary
+    if(srcLength < 0) {
+      srcLength = u_strlen(srcChars + srcStart);
+    }
+    if(length != srcLength) {
+      return (int8_t)((length - srcLength) >> 24 | 1);
+    }
+  }
+  return 0;
+}
+
+//========================================
+// Write implementation
+//========================================
+
+UnicodeString &
+UnicodeString::caseMap(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                       UStringCaseMapper *stringCaseMapper) {
+  if(isEmpty() || !isWritable()) {
+    // nothing to do
+    return *this;
+  }
+
+  UChar oldBuffer[2 * US_STACKBUF_SIZE];
+  UChar *oldArray;
+  int32_t oldLength = length();
+  int32_t newLength;
+  UBool writable = isBufferWritable();
+  UErrorCode errorCode = U_ZERO_ERROR;
+
+#if !UCONFIG_NO_BREAK_ITERATION
+  // Read-only alias to the original string contents for the titlecasing BreakIterator.
+  // We cannot set the iterator simply to *this because *this is being modified.
+  UnicodeString oldString;
+#endif
+
+  // Try to avoid heap-allocating a new character array for this string.
+  if (writable ? oldLength <= UPRV_LENGTHOF(oldBuffer) : oldLength < US_STACKBUF_SIZE) {
+    // Short string: Copy the contents into a temporary buffer and
+    // case-map back into the current array, or into the stack buffer.
+    UChar *buffer = getArrayStart();
+    int32_t capacity;
+    oldArray = oldBuffer;
+    u_memcpy(oldBuffer, buffer, oldLength);
+    if (writable) {
+      capacity = getCapacity();
+    } else {
+      // Switch from the read-only alias or shared heap buffer to the stack buffer.
+      if (!cloneArrayIfNeeded(US_STACKBUF_SIZE, US_STACKBUF_SIZE, /* doCopyArray= */ FALSE)) {
+        return *this;
+      }
+      U_ASSERT(fUnion.fFields.fLengthAndFlags & kUsingStackBuffer);
+      buffer = fUnion.fStackFields.fBuffer;
+      capacity = US_STACKBUF_SIZE;
+    }
+#if !UCONFIG_NO_BREAK_ITERATION
+    if (iter != nullptr) {
+      oldString.setTo(FALSE, oldArray, oldLength);
+      iter->setText(oldString);
+    }
+#endif
+    newLength = stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                                 buffer, capacity,
+                                 oldArray, oldLength, NULL, errorCode);
+    if (U_SUCCESS(errorCode)) {
+      setLength(newLength);
+      return *this;
+    } else if (errorCode == U_BUFFER_OVERFLOW_ERROR) {
+      // common overflow handling below
+    } else {
+      setToBogus();
+      return *this;
+    }
+  } else {
+    // Longer string or read-only buffer:
+    // Collect only changes and then apply them to this string.
+    // Case mapping often changes only small parts of a string,
+    // and often does not change its length.
+    oldArray = getArrayStart();
+    Edits edits;
+    UChar replacementChars[200];
+#if !UCONFIG_NO_BREAK_ITERATION
+    if (iter != nullptr) {
+      oldString.setTo(FALSE, oldArray, oldLength);
+      iter->setText(oldString);
+    }
+#endif
+    stringCaseMapper(caseLocale, options | U_OMIT_UNCHANGED_TEXT, UCASEMAP_BREAK_ITERATOR
+                     replacementChars, UPRV_LENGTHOF(replacementChars),
+                     oldArray, oldLength, &edits, errorCode);
+    if (U_SUCCESS(errorCode)) {
+      // Grow the buffer at most once, not for multiple doReplace() calls.
+      newLength = oldLength + edits.lengthDelta();
+      if (newLength > oldLength && !cloneArrayIfNeeded(newLength, newLength)) {
+        return *this;
+      }
+      for (Edits::Iterator ei = edits.getCoarseChangesIterator(); ei.next(errorCode);) {
+        doReplace(ei.destinationIndex(), ei.oldLength(),
+                  replacementChars, ei.replacementIndex(), ei.newLength());
+      }
+      if (U_FAILURE(errorCode)) {
+        setToBogus();
+      }
+      return *this;
+    } else if (errorCode == U_BUFFER_OVERFLOW_ERROR) {
+      // common overflow handling below
+      newLength = oldLength + edits.lengthDelta();
+    } else {
+      setToBogus();
+      return *this;
+    }
+  }
+
+  // Handle buffer overflow, newLength is known.
+  // We need to allocate a new buffer for the internal string case mapping function.
+  // This is very similar to how doReplace() keeps the old array pointer
+  // and deletes the old array itself after it is done.
+  // In addition, we are forcing cloneArrayIfNeeded() to always allocate a new array.
+  int32_t *bufferToDelete = 0;
+  if (!cloneArrayIfNeeded(newLength, newLength, FALSE, &bufferToDelete, TRUE)) {
+    return *this;
+  }
+  errorCode = U_ZERO_ERROR;
+  // No need to iter->setText() again: The case mapper restarts via iter->first().
+  newLength = stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                               getArrayStart(), getCapacity(),
+                               oldArray, oldLength, NULL, errorCode);
+  if (bufferToDelete) {
+    uprv_free(bufferToDelete);
+  }
+  if (U_SUCCESS(errorCode)) {
+    setLength(newLength);
+  } else {
+    setToBogus();
+  }
+  return *this;
+}
+
+UnicodeString &
+UnicodeString::foldCase(uint32_t options) {
+  return caseMap(UCASE_LOC_ROOT, options, UCASEMAP_BREAK_ITERATOR_NULL ustrcase_internalFold);
+}
+
+U_NAMESPACE_END
+
+// Defined here to reduce dependencies on break iterator
+U_CAPI int32_t U_EXPORT2
+uhash_hashCaselessUnicodeString(const UElement key) {
+    U_NAMESPACE_USE
+    const UnicodeString *str = (const UnicodeString*) key.pointer;
+    if (str == NULL) {
+        return 0;
+    }
+    // Inefficient; a better way would be to have a hash function in
+    // UnicodeString that does case folding on the fly.
+    UnicodeString copy(*str);
+    return copy.foldCase().hashCode();
+}
+
+// Defined here to reduce dependencies on break iterator
+U_CAPI UBool U_EXPORT2
+uhash_compareCaselessUnicodeString(const UElement key1, const UElement key2) {
+    U_NAMESPACE_USE
+    const UnicodeString *str1 = (const UnicodeString*) key1.pointer;
+    const UnicodeString *str2 = (const UnicodeString*) key2.pointer;
+    if (str1 == str2) {
+        return TRUE;
+    }
+    if (str1 == NULL || str2 == NULL) {
+        return FALSE;
+    }
+    return str1->caseCompare(*str2, U_FOLD_CASE_DEFAULT) == 0;
+}
diff --git a/thirdparty/icu4c/common/unistr_case_locale.cpp b/thirdparty/icu4c/common/unistr_case_locale.cpp
new file mode 100644
index 0000000000..f0f3048d06
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr_case_locale.cpp
@@ -0,0 +1,56 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  unistr_case_locale.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011may31
+*   created by: Markus W. Scherer
+*
+*   Locale-sensitive case mapping functions (ones that call uloc_getDefault())
+*   were moved here to break dependency cycles among parts of the common library.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/locid.h"
+#include "unicode/ucasemap.h"
+#include "unicode/unistr.h"
+#include "ucasemap_imp.h"
+
+U_NAMESPACE_BEGIN
+
+//========================================
+// Write implementation
+//========================================
+
+UnicodeString &
+UnicodeString::toLower() {
+  return caseMap(ustrcase_getCaseLocale(NULL), 0,
+                 UCASEMAP_BREAK_ITERATOR_NULL ustrcase_internalToLower);
+}
+
+UnicodeString &
+UnicodeString::toLower(const Locale &locale) {
+  return caseMap(ustrcase_getCaseLocale(locale.getBaseName()), 0,
+                 UCASEMAP_BREAK_ITERATOR_NULL ustrcase_internalToLower);
+}
+
+UnicodeString &
+UnicodeString::toUpper() {
+  return caseMap(ustrcase_getCaseLocale(NULL), 0,
+                 UCASEMAP_BREAK_ITERATOR_NULL ustrcase_internalToUpper);
+}
+
+UnicodeString &
+UnicodeString::toUpper(const Locale &locale) {
+  return caseMap(ustrcase_getCaseLocale(locale.getBaseName()), 0,
+                 UCASEMAP_BREAK_ITERATOR_NULL ustrcase_internalToUpper);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/unistr_cnv.cpp b/thirdparty/icu4c/common/unistr_cnv.cpp
new file mode 100644
index 0000000000..64d3c16801
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr_cnv.cpp
@@ -0,0 +1,417 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unistr_cnv.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2004aug19
+*   created by: Markus W. Scherer
+*
+*   Character conversion functions moved here from unistr.cpp
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/putil.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "unicode/ustring.h"
+#include "unicode/unistr.h"
+#include "unicode/ucnv.h"
+#include "ucnv_imp.h"
+#include "putilimp.h"
+#include "ustr_cnv.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_BEGIN
+
+//========================================
+// Constructors
+//========================================
+
+#if !U_CHARSET_IS_UTF8
+
+UnicodeString::UnicodeString(const char *codepageData) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    if(codepageData != 0) {
+        doCodepageCreate(codepageData, (int32_t)uprv_strlen(codepageData), 0);
+    }
+}
+
+UnicodeString::UnicodeString(const char *codepageData,
+                             int32_t dataLength) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    if(codepageData != 0) {
+        doCodepageCreate(codepageData, dataLength, 0);
+    }
+}
+
+// else see unistr.cpp
+#endif
+
+UnicodeString::UnicodeString(const char *codepageData,
+                             const char *codepage) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    if(codepageData != 0) {
+        doCodepageCreate(codepageData, (int32_t)uprv_strlen(codepageData), codepage);
+    }
+}
+
+UnicodeString::UnicodeString(const char *codepageData,
+                             int32_t dataLength,
+                             const char *codepage) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    if(codepageData != 0) {
+        doCodepageCreate(codepageData, dataLength, codepage);
+    }
+}
+
+UnicodeString::UnicodeString(const char *src, int32_t srcLength,
+                             UConverter *cnv,
+                             UErrorCode &errorCode) {
+    fUnion.fFields.fLengthAndFlags = kShortString;
+    if(U_SUCCESS(errorCode)) {
+        // check arguments
+        if(src==NULL) {
+            // treat as an empty string, do nothing more
+        } else if(srcLength<-1) {
+            errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        } else {
+            // get input length
+            if(srcLength==-1) {
+                srcLength=(int32_t)uprv_strlen(src);
+            }
+            if(srcLength>0) {
+                if(cnv!=0) {
+                    // use the provided converter
+                    ucnv_resetToUnicode(cnv);
+                    doCodepageCreate(src, srcLength, cnv, errorCode);
+                } else {
+                    // use the default converter
+                    cnv=u_getDefaultConverter(&errorCode);
+                    doCodepageCreate(src, srcLength, cnv, errorCode);
+                    u_releaseDefaultConverter(cnv);
+                }
+            }
+        }
+
+        if(U_FAILURE(errorCode)) {
+            setToBogus();
+        }
+    }
+}
+
+//========================================
+// Codeset conversion
+//========================================
+
+#if !U_CHARSET_IS_UTF8
+
+int32_t
+UnicodeString::extract(int32_t start,
+                       int32_t length,
+                       char *target,
+                       uint32_t dstSize) const {
+    return extract(start, length, target, dstSize, 0);
+}
+
+// else see unistr.cpp
+#endif
+
+int32_t
+UnicodeString::extract(int32_t start,
+                       int32_t length,
+                       char *target,
+                       uint32_t dstSize,
+                       const char *codepage) const
+{
+    // if the arguments are illegal, then do nothing
+    if(/*dstSize < 0 || */(dstSize > 0 && target == 0)) {
+        return 0;
+    }
+
+    // pin the indices to legal values
+    pinIndices(start, length);
+
+    // We need to cast dstSize to int32_t for all subsequent code.
+    // I don't know why the API was defined with uint32_t but we are stuck with it.
+    // Also, dstSize==0xffffffff means "unlimited" but if we use target+dstSize
+    // as a limit in some functions, it may wrap around and yield a pointer
+    // that compares less-than target.
+    int32_t capacity;
+    if(dstSize < 0x7fffffff) {
+        // Assume that the capacity is real and a limit pointer won't wrap around.
+        capacity = (int32_t)dstSize;
+    } else {
+        // Pin the capacity so that a limit pointer does not wrap around.
+        char *targetLimit = (char *)U_MAX_PTR(target);
+        // U_MAX_PTR(target) returns a targetLimit that is at most 0x7fffffff
+        // greater than target and does not wrap around the top of the address space.
+        capacity = (int32_t)(targetLimit - target);
+    }
+
+    // create the converter
+    UConverter *converter;
+    UErrorCode status = U_ZERO_ERROR;
+
+    // just write the NUL if the string length is 0
+    if(length == 0) {
+        return u_terminateChars(target, capacity, 0, &status);
+    }
+
+    // if the codepage is the default, use our cache
+    // if it is an empty string, then use the "invariant character" conversion
+    if (codepage == 0) {
+        const char *defaultName = ucnv_getDefaultName();
+        if(UCNV_FAST_IS_UTF8(defaultName)) {
+            return toUTF8(start, length, target, capacity);
+        }
+        converter = u_getDefaultConverter(&status);
+    } else if (*codepage == 0) {
+        // use the "invariant characters" conversion
+        int32_t destLength;
+        if(length <= capacity) {
+            destLength = length;
+        } else {
+            destLength = capacity;
+        }
+        u_UCharsToChars(getArrayStart() + start, target, destLength);
+        return u_terminateChars(target, capacity, length, &status);
+    } else {
+        converter = ucnv_open(codepage, &status);
+    }
+
+    length = doExtract(start, length, target, capacity, converter, status);
+
+    // close the converter
+    if (codepage == 0) {
+        u_releaseDefaultConverter(converter);
+    } else {
+        ucnv_close(converter);
+    }
+
+    return length;
+}
+
+int32_t
+UnicodeString::extract(char *dest, int32_t destCapacity,
+                       UConverter *cnv,
+                       UErrorCode &errorCode) const
+{
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+
+    if(isBogus() || destCapacity<0 || (destCapacity>0 && dest==0)) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    // nothing to do?
+    if(isEmpty()) {
+        return u_terminateChars(dest, destCapacity, 0, &errorCode);
+    }
+
+    // get the converter
+    UBool isDefaultConverter;
+    if(cnv==0) {
+        isDefaultConverter=TRUE;
+        cnv=u_getDefaultConverter(&errorCode);
+        if(U_FAILURE(errorCode)) {
+            return 0;
+        }
+    } else {
+        isDefaultConverter=FALSE;
+        ucnv_resetFromUnicode(cnv);
+    }
+
+    // convert
+    int32_t len=doExtract(0, length(), dest, destCapacity, cnv, errorCode);
+
+    // release the converter
+    if(isDefaultConverter) {
+        u_releaseDefaultConverter(cnv);
+    }
+
+    return len;
+}
+
+int32_t
+UnicodeString::doExtract(int32_t start, int32_t length,
+                         char *dest, int32_t destCapacity,
+                         UConverter *cnv,
+                         UErrorCode &errorCode) const
+{
+    if(U_FAILURE(errorCode)) {
+        if(destCapacity!=0) {
+            *dest=0;
+        }
+        return 0;
+    }
+
+    const UChar *src=getArrayStart()+start, *srcLimit=src+length;
+    char *originalDest=dest;
+    const char *destLimit;
+
+    if(destCapacity==0) {
+        destLimit=dest=0;
+    } else if(destCapacity==-1) {
+        // Pin the limit to U_MAX_PTR if the "magic" destCapacity is used.
+        destLimit=(char*)U_MAX_PTR(dest);
+        // for NUL-termination, translate into highest int32_t
+        destCapacity=0x7fffffff;
+    } else {
+        destLimit=dest+destCapacity;
+    }
+
+    // perform the conversion
+    ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, &errorCode);
+    length=(int32_t)(dest-originalDest);
+
+    // if an overflow occurs, then get the preflighting length
+    if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+        char buffer[1024];
+
+        destLimit=buffer+sizeof(buffer);
+        do {
+            dest=buffer;
+            errorCode=U_ZERO_ERROR;
+            ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, &errorCode);
+            length+=(int32_t)(dest-buffer);
+        } while(errorCode==U_BUFFER_OVERFLOW_ERROR);
+    }
+
+    return u_terminateChars(originalDest, destCapacity, length, &errorCode);
+}
+
+void
+UnicodeString::doCodepageCreate(const char *codepageData,
+                                int32_t dataLength,
+                                const char *codepage)
+{
+    // if there's nothing to convert, do nothing
+    if(codepageData == 0 || dataLength == 0 || dataLength < -1) {
+        return;
+    }
+    if(dataLength == -1) {
+        dataLength = (int32_t)uprv_strlen(codepageData);
+    }
+
+    UErrorCode status = U_ZERO_ERROR;
+
+    // create the converter
+    // if the codepage is the default, use our cache
+    // if it is an empty string, then use the "invariant character" conversion
+    UConverter *converter;
+    if (codepage == 0) {
+        const char *defaultName = ucnv_getDefaultName();
+        if(UCNV_FAST_IS_UTF8(defaultName)) {
+            setToUTF8(StringPiece(codepageData, dataLength));
+            return;
+        }
+        converter = u_getDefaultConverter(&status);
+    } else if(*codepage == 0) {
+        // use the "invariant characters" conversion
+        if(cloneArrayIfNeeded(dataLength, dataLength, FALSE)) {
+            u_charsToUChars(codepageData, getArrayStart(), dataLength);
+            setLength(dataLength);
+        } else {
+            setToBogus();
+        }
+        return;
+    } else {
+        converter = ucnv_open(codepage, &status);
+    }
+
+    // if we failed, set the appropriate flags and return
+    if(U_FAILURE(status)) {
+        setToBogus();
+        return;
+    }
+
+    // perform the conversion
+    doCodepageCreate(codepageData, dataLength, converter, status);
+    if(U_FAILURE(status)) {
+        setToBogus();
+    }
+
+    // close the converter
+    if(codepage == 0) {
+        u_releaseDefaultConverter(converter);
+    } else {
+        ucnv_close(converter);
+    }
+}
+
+void
+UnicodeString::doCodepageCreate(const char *codepageData,
+                                int32_t dataLength,
+                                UConverter *converter,
+                                UErrorCode &status)
+{
+    if(U_FAILURE(status)) {
+        return;
+    }
+
+    // set up the conversion parameters
+    const char *mySource     = codepageData;
+    const char *mySourceEnd  = mySource + dataLength;
+    UChar *array, *myTarget;
+
+    // estimate the size needed:
+    int32_t arraySize;
+    if(dataLength <= US_STACKBUF_SIZE) {
+        // try to use the stack buffer
+        arraySize = US_STACKBUF_SIZE;
+    } else {
+        // 1.25 UChar's per source byte should cover most cases
+        arraySize = dataLength + (dataLength >> 2);
+    }
+
+    // we do not care about the current contents
+    UBool doCopyArray = FALSE;
+    for(;;) {
+        if(!cloneArrayIfNeeded(arraySize, arraySize, doCopyArray)) {
+            setToBogus();
+            break;
+        }
+
+        // perform the conversion
+        array = getArrayStart();
+        myTarget = array + length();
+        ucnv_toUnicode(converter, &myTarget,  array + getCapacity(),
+            &mySource, mySourceEnd, 0, TRUE, &status);
+
+        // update the conversion parameters
+        setLength((int32_t)(myTarget - array));
+
+        // allocate more space and copy data, if needed
+        if(status == U_BUFFER_OVERFLOW_ERROR) {
+            // reset the error code
+            status = U_ZERO_ERROR;
+
+            // keep the previous conversion results
+            doCopyArray = TRUE;
+
+            // estimate the new size needed, larger than before
+            // try 2 UChar's per remaining source byte
+            arraySize = (int32_t)(length() + 2 * (mySourceEnd - mySource));
+        } else {
+            break;
+        }
+    }
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unistr_props.cpp b/thirdparty/icu4c/common/unistr_props.cpp
new file mode 100644
index 0000000000..4006475790
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr_props.cpp
@@ -0,0 +1,77 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unistr_props.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2004aug25
+*   created by: Markus W. Scherer
+*
+*   Character property dependent functions moved here from unistr.cpp
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+#include "unicode/unistr.h"
+#include "unicode/utf16.h"
+
+U_NAMESPACE_BEGIN
+
+UnicodeString& 
+UnicodeString::trim()
+{
+  if(isBogus()) {
+    return *this;
+  }
+
+  UChar *array = getArrayStart();
+  UChar32 c;
+  int32_t oldLength = this->length();
+  int32_t i = oldLength, length;
+
+  // first cut off trailing white space
+  for(;;) {
+    length = i;
+    if(i <= 0) {
+      break;
+    }
+    U16_PREV(array, 0, i, c);
+    if(!(c == 0x20 || u_isWhitespace(c))) {
+      break;
+    }
+  }
+  if(length < oldLength) {
+    setLength(length);
+  }
+
+  // find leading white space
+  int32_t start;
+  i = 0;
+  for(;;) {
+    start = i;
+    if(i >= length) {
+      break;
+    }
+    U16_NEXT(array, i, length, c);
+    if(!(c == 0x20 || u_isWhitespace(c))) {
+      break;
+    }
+  }
+
+  // move string forward over leading white space
+  if(start > 0) {
+    doReplace(0, start, 0, 0, 0);
+  }
+
+  return *this;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/unistr_titlecase_brkiter.cpp b/thirdparty/icu4c/common/unistr_titlecase_brkiter.cpp
new file mode 100644
index 0000000000..4969884b0d
--- /dev/null
+++ b/thirdparty/icu4c/common/unistr_titlecase_brkiter.cpp
@@ -0,0 +1,57 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  unistr_titlecase_brkiter.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:2
+*
+*   created on: 2011may30
+*   created by: Markus W. Scherer
+*
+*   Titlecasing functions that are based on BreakIterator
+*   were moved here to break dependency cycles among parts of the common library.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/brkiter.h"
+#include "unicode/locid.h"
+#include "unicode/ucasemap.h"
+#include "unicode/unistr.h"
+#include "ucasemap_imp.h"
+
+U_NAMESPACE_BEGIN
+
+UnicodeString &
+UnicodeString::toTitle(BreakIterator *iter) {
+    return toTitle(iter, Locale::getDefault(), 0);
+}
+
+UnicodeString &
+UnicodeString::toTitle(BreakIterator *iter, const Locale &locale) {
+    return toTitle(iter, locale, 0);
+}
+
+UnicodeString &
+UnicodeString::toTitle(BreakIterator *iter, const Locale &locale, uint32_t options) {
+    LocalPointer<BreakIterator> ownedIter;
+    UErrorCode errorCode = U_ZERO_ERROR;
+    iter = ustrcase_getTitleBreakIterator(&locale, "", options, iter, ownedIter, errorCode);
+    if (iter == nullptr) {
+        setToBogus();
+        return *this;
+    }
+    caseMap(ustrcase_getCaseLocale(locale.getBaseName()), options, iter, ustrcase_internalToTitle);
+    return *this;
+}
+
+U_NAMESPACE_END
+
+#endif  // !UCONFIG_NO_BREAK_ITERATION
diff --git a/thirdparty/icu4c/common/unistrappender.h b/thirdparty/icu4c/common/unistrappender.h
new file mode 100644
index 0000000000..75fcb9e775
--- /dev/null
+++ b/thirdparty/icu4c/common/unistrappender.h
@@ -0,0 +1,90 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File unistrappender.h
+******************************************************************************
+*/
+
+#ifndef __UNISTRAPPENDER_H__
+#define __UNISTRAPPENDER_H__
+
+#include "unicode/unistr.h"
+#include "unicode/uobject.h"
+#include "unicode/utf16.h"
+#include "unicode/utypes.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * An optimization for the slowness of calling UnicodeString::append()
+ * one character at a time in a loop. It stores appends in a buffer while
+ * never actually calling append on the unicode string unless the buffer
+ * fills up or is flushed.
+ * 
+ * proper usage:
+ * {
+ *     UnicodeStringAppender appender(astring);
+ *     for (int32_t i = 0; i < 100; ++i) {
+ *        appender.append((UChar) i);
+ *     }
+ *     // appender flushed automatically when it goes out of scope.
+ * }
+ */
+class UnicodeStringAppender : public UMemory {
+public:
+    
+    /**
+     * dest is the UnicodeString being appended to. It must always
+     * exist while this instance exists.
+     */
+    UnicodeStringAppender(UnicodeString &dest) : fDest(&dest), fIdx(0) { }
+
+    inline void append(UChar x) {
+        if (fIdx == UPRV_LENGTHOF(fBuffer)) {
+            fDest->append(fBuffer, 0, fIdx);
+            fIdx = 0;
+        }
+        fBuffer[fIdx++] = x;
+    }
+
+    inline void append(UChar32 x) {
+        if (fIdx >= UPRV_LENGTHOF(fBuffer) - 1) {
+            fDest->append(fBuffer, 0, fIdx);
+            fIdx = 0;
+        }
+        U16_APPEND_UNSAFE(fBuffer, fIdx, x);
+    }
+
+    /**
+     * Ensures that all appended characters have been written out to dest.
+     */
+    inline void flush() {
+        if (fIdx) {
+            fDest->append(fBuffer, 0, fIdx);
+        }
+        fIdx = 0;
+    }
+
+    /**
+     * flush the buffer when we go out of scope.
+     */
+    ~UnicodeStringAppender() {
+        flush();
+    }
+private:
+    UnicodeString *fDest;
+    int32_t fIdx;
+    UChar fBuffer[32];
+    UnicodeStringAppender(const UnicodeStringAppender &other);
+    UnicodeStringAppender &operator=(const UnicodeStringAppender &other);
+};
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/unorm.cpp b/thirdparty/icu4c/common/unorm.cpp
new file mode 100644
index 0000000000..2d9f46052f
--- /dev/null
+++ b/thirdparty/icu4c/common/unorm.cpp
@@ -0,0 +1,280 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (c) 1996-2014, International Business Machines
+* Corporation and others. All Rights Reserved.
+******************************************************************************
+* File unorm.cpp
+*
+* Created by: Vladimir Weinstein 12052000
+*
+* Modification history :
+*
+* Date        Name        Description
+* 02/01/01    synwee      Added normalization quickcheck enum and method.
+* 02/12/01    synwee      Commented out quickcheck util api has been approved
+*                         Added private method for doing FCD checks
+* 02/23/01    synwee      Modified quickcheck and checkFCE to run through 
+*                         string for codepoints < 0x300 for the normalization 
+*                         mode NFC.
+* 05/25/01+   Markus Scherer total rewrite, implement all normalization here
+*                         instead of just wrappers around normlzr.cpp,
+*                         load unorm.dat, support Unicode 3.1 with
+*                         supplementary code points, etc.
+* 2009-nov..2010-jan  Markus Scherer  total rewrite, new Normalizer2 API & code
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/udata.h"
+#include "unicode/ustring.h"
+#include "unicode/uiter.h"
+#include "unicode/unorm.h"
+#include "unicode/unorm2.h"
+#include "normalizer2impl.h"
+#include "unormimp.h"
+#include "uprops.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_USE
+
+/* quick check functions ---------------------------------------------------- */
+
+U_CAPI UNormalizationCheckResult U_EXPORT2
+unorm_quickCheck(const UChar *src,
+                 int32_t srcLength, 
+                 UNormalizationMode mode,
+                 UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    return unorm2_quickCheck((const UNormalizer2 *)n2, src, srcLength, pErrorCode);
+}
+
+U_CAPI UNormalizationCheckResult U_EXPORT2
+unorm_quickCheckWithOptions(const UChar *src, int32_t srcLength, 
+                            UNormalizationMode mode, int32_t options,
+                            UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    if(options&UNORM_UNICODE_3_2) {
+        FilteredNormalizer2 fn2(*n2, *uniset_getUnicode32Instance(*pErrorCode));
+        return unorm2_quickCheck(
+            reinterpret_cast<const UNormalizer2 *>(static_cast<Normalizer2 *>(&fn2)),
+            src, srcLength, pErrorCode);
+    } else {
+        return unorm2_quickCheck((const UNormalizer2 *)n2, src, srcLength, pErrorCode);
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+unorm_isNormalized(const UChar *src, int32_t srcLength,
+                   UNormalizationMode mode,
+                   UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    return unorm2_isNormalized((const UNormalizer2 *)n2, src, srcLength, pErrorCode);
+}
+
+U_CAPI UBool U_EXPORT2
+unorm_isNormalizedWithOptions(const UChar *src, int32_t srcLength,
+                              UNormalizationMode mode, int32_t options,
+                              UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    if(options&UNORM_UNICODE_3_2) {
+        FilteredNormalizer2 fn2(*n2, *uniset_getUnicode32Instance(*pErrorCode));
+        return unorm2_isNormalized(
+            reinterpret_cast<const UNormalizer2 *>(static_cast<Normalizer2 *>(&fn2)),
+            src, srcLength, pErrorCode);
+    } else {
+        return unorm2_isNormalized((const UNormalizer2 *)n2, src, srcLength, pErrorCode);
+    }
+}
+
+/* normalize() API ---------------------------------------------------------- */
+
+/** Public API for normalizing. */
+U_CAPI int32_t U_EXPORT2
+unorm_normalize(const UChar *src, int32_t srcLength,
+                UNormalizationMode mode, int32_t options,
+                UChar *dest, int32_t destCapacity,
+                UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    if(options&UNORM_UNICODE_3_2) {
+        FilteredNormalizer2 fn2(*n2, *uniset_getUnicode32Instance(*pErrorCode));
+        return unorm2_normalize(
+            reinterpret_cast<const UNormalizer2 *>(static_cast<Normalizer2 *>(&fn2)),
+            src, srcLength, dest, destCapacity, pErrorCode);
+    } else {
+        return unorm2_normalize((const UNormalizer2 *)n2,
+            src, srcLength, dest, destCapacity, pErrorCode);
+    }
+}
+
+
+/* iteration functions ------------------------------------------------------ */
+
+static int32_t
+_iterate(UCharIterator *src, UBool forward,
+              UChar *dest, int32_t destCapacity,
+              const Normalizer2 *n2,
+              UBool doNormalize, UBool *pNeededToNormalize,
+              UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0) || src==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(pNeededToNormalize!=NULL) {
+        *pNeededToNormalize=FALSE;
+    }
+    if(!(forward ? src->hasNext(src) : src->hasPrevious(src))) {
+        return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
+    }
+
+    UnicodeString buffer;
+    UChar32 c;
+    if(forward) {
+        /* get one character and ignore its properties */
+        buffer.append(uiter_next32(src));
+        /* get all following characters until we see a boundary */
+        while((c=uiter_next32(src))>=0) {
+            if(n2->hasBoundaryBefore(c)) {
+                /* back out the latest movement to stop at the boundary */
+                src->move(src, -U16_LENGTH(c), UITER_CURRENT);
+                break;
+            } else {
+                buffer.append(c);
+            }
+        }
+    } else {
+        while((c=uiter_previous32(src))>=0) {
+            /* always write this character to the front of the buffer */
+            buffer.insert(0, c);
+            /* stop if this just-copied character is a boundary */
+            if(n2->hasBoundaryBefore(c)) {
+                break;
+            }
+        }
+    }
+
+    UnicodeString destString(dest, 0, destCapacity);
+    if(buffer.length()>0 && doNormalize) {
+        n2->normalize(buffer, destString, *pErrorCode).extract(dest, destCapacity, *pErrorCode);
+        if(pNeededToNormalize!=NULL && U_SUCCESS(*pErrorCode)) {
+            *pNeededToNormalize= destString!=buffer;
+        }
+        return destString.length();
+    } else {
+        /* just copy the source characters */
+        return buffer.extract(dest, destCapacity, *pErrorCode);
+    }
+}
+
+static int32_t
+unorm_iterate(UCharIterator *src, UBool forward,
+              UChar *dest, int32_t destCapacity,
+              UNormalizationMode mode, int32_t options,
+              UBool doNormalize, UBool *pNeededToNormalize,
+              UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    if(options&UNORM_UNICODE_3_2) {
+        const UnicodeSet *uni32 = uniset_getUnicode32Instance(*pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+        FilteredNormalizer2 fn2(*n2, *uni32);
+        return _iterate(src, forward, dest, destCapacity,
+            &fn2, doNormalize, pNeededToNormalize, pErrorCode);
+    }
+    return _iterate(src, forward, dest, destCapacity,
+            n2, doNormalize, pNeededToNormalize, pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm_previous(UCharIterator *src,
+               UChar *dest, int32_t destCapacity,
+               UNormalizationMode mode, int32_t options,
+               UBool doNormalize, UBool *pNeededToNormalize,
+               UErrorCode *pErrorCode) {
+    return unorm_iterate(src, FALSE,
+                         dest, destCapacity,
+                         mode, options,
+                         doNormalize, pNeededToNormalize,
+                         pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm_next(UCharIterator *src,
+           UChar *dest, int32_t destCapacity,
+           UNormalizationMode mode, int32_t options,
+           UBool doNormalize, UBool *pNeededToNormalize,
+           UErrorCode *pErrorCode) {
+    return unorm_iterate(src, TRUE,
+                         dest, destCapacity,
+                         mode, options,
+                         doNormalize, pNeededToNormalize,
+                         pErrorCode);
+}
+
+/* Concatenation of normalized strings -------------------------------------- */
+
+static int32_t
+_concatenate(const UChar *left, int32_t leftLength,
+                  const UChar *right, int32_t rightLength,
+                  UChar *dest, int32_t destCapacity,
+                  const Normalizer2 *n2,
+                  UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0) ||
+        left==NULL || leftLength<-1 || right==NULL || rightLength<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* check for overlapping right and destination */
+    if( dest!=NULL &&
+        ((right>=dest && right<(dest+destCapacity)) ||
+         (rightLength>0 && dest>=right && dest<(right+rightLength)))
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* allow left==dest */
+    UnicodeString destString;
+    if(left==dest) {
+        destString.setTo(dest, leftLength, destCapacity);
+    } else {
+        destString.setTo(dest, 0, destCapacity);
+        destString.append(left, leftLength);
+    }
+    return n2->append(destString, UnicodeString(rightLength<0, right, rightLength), *pErrorCode).
+           extract(dest, destCapacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm_concatenate(const UChar *left, int32_t leftLength,
+                  const UChar *right, int32_t rightLength,
+                  UChar *dest, int32_t destCapacity,
+                  UNormalizationMode mode, int32_t options,
+                  UErrorCode *pErrorCode) {
+    const Normalizer2 *n2=Normalizer2Factory::getInstance(mode, *pErrorCode);
+    if(options&UNORM_UNICODE_3_2) {
+        const UnicodeSet *uni32 = uniset_getUnicode32Instance(*pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+        FilteredNormalizer2 fn2(*n2, *uni32);
+        return _concatenate(left, leftLength, right, rightLength,
+            dest, destCapacity, &fn2, pErrorCode);
+    }
+    return _concatenate(left, leftLength, right, rightLength,
+        dest, destCapacity, n2, pErrorCode);
+}
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
diff --git a/thirdparty/icu4c/common/unormcmp.cpp b/thirdparty/icu4c/common/unormcmp.cpp
new file mode 100644
index 0000000000..689b0b53b2
--- /dev/null
+++ b/thirdparty/icu4c/common/unormcmp.cpp
@@ -0,0 +1,640 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2001-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unormcmp.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004sep13
+*   created by: Markus W. Scherer
+*
+*   unorm_compare() function moved here from unorm.cpp for better modularization.
+*   Depends on both normalization and case folding.
+*   Allows unorm.cpp to not depend on any character properties code.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "unicode/unorm.h"
+#include "unicode/ustring.h"
+#include "cmemory.h"
+#include "normalizer2impl.h"
+#include "ucase.h"
+#include "uprops.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_USE
+
+/* compare canonically equivalent ------------------------------------------- */
+
+/*
+ * Compare two strings for canonical equivalence.
+ * Further options include case-insensitive comparison and
+ * code point order (as opposed to code unit order).
+ *
+ * In this function, canonical equivalence is optional as well.
+ * If canonical equivalence is tested, then both strings must fulfill
+ * the FCD check.
+ *
+ * Semantically, this is equivalent to
+ *   strcmp[CodePointOrder](NFD(foldCase(s1)), NFD(foldCase(s2)))
+ * where code point order, NFD and foldCase are all optional.
+ *
+ * String comparisons almost always yield results before processing both strings
+ * completely.
+ * They are generally more efficient working incrementally instead of
+ * performing the sub-processing (strlen, normalization, case-folding)
+ * on the entire strings first.
+ *
+ * It is also unnecessary to not normalize identical characters.
+ *
+ * This function works in principle as follows:
+ *
+ * loop {
+ *   get one code unit c1 from s1 (-1 if end of source)
+ *   get one code unit c2 from s2 (-1 if end of source)
+ *
+ *   if(either string finished) {
+ *     return result;
+ *   }
+ *   if(c1==c2) {
+ *     continue;
+ *   }
+ *
+ *   // c1!=c2
+ *   try to decompose/case-fold c1/c2, and continue if one does;
+ *
+ *   // still c1!=c2 and neither decomposes/case-folds, return result
+ *   return c1-c2;
+ * }
+ *
+ * When a character decomposes, then the pointer for that source changes to
+ * the decomposition, pushing the previous pointer onto a stack.
+ * When the end of the decomposition is reached, then the code unit reader
+ * pops the previous source from the stack.
+ * (Same for case-folding.)
+ *
+ * This is complicated further by operating on variable-width UTF-16.
+ * The top part of the loop works on code units, while lookups for decomposition
+ * and case-folding need code points.
+ * Code points are assembled after the equality/end-of-source part.
+ * The source pointer is only advanced beyond all code units when the code point
+ * actually decomposes/case-folds.
+ *
+ * If we were on a trail surrogate unit when assembling a code point,
+ * and the code point decomposes/case-folds, then the decomposition/folding
+ * result must be compared with the part of the other string that corresponds to
+ * this string's lead surrogate.
+ * Since we only assemble a code point when hitting a trail unit when the
+ * preceding lead units were identical, we back up the other string by one unit
+ * in such a case.
+ *
+ * The optional code point order comparison at the end works with
+ * the same fix-up as the other code point order comparison functions.
+ * See ustring.c and the comment near the end of this function.
+ *
+ * Assumption: A decomposition or case-folding result string never contains
+ * a single surrogate. This is a safe assumption in the Unicode Standard.
+ * Therefore, we do not need to check for surrogate pairs across
+ * decomposition/case-folding boundaries.
+ *
+ * Further assumptions (see verifications tstnorm.cpp):
+ * The API function checks for FCD first, while the core function
+ * first case-folds and then decomposes. This requires that case-folding does not
+ * un-FCD any strings.
+ *
+ * The API function may also NFD the input and turn off decomposition.
+ * This requires that case-folding does not un-NFD strings either.
+ *
+ * TODO If any of the above two assumptions is violated,
+ * then this entire code must be re-thought.
+ * If this happens, then a simple solution is to case-fold both strings up front
+ * and to turn off UNORM_INPUT_IS_FCD.
+ * We already do this when not both strings are in FCD because makeFCD
+ * would be a partial NFD before the case folding, which does not work.
+ * Note that all of this is only a problem when case-folding _and_
+ * canonical equivalence come together.
+ * (Comments in unorm_compare() are more up to date than this TODO.)
+ */
+
+/* stack element for previous-level source/decomposition pointers */
+struct CmpEquivLevel {
+    const UChar *start, *s, *limit;
+};
+typedef struct CmpEquivLevel CmpEquivLevel;
+
+/**
+ * Internal option for unorm_cmpEquivFold() for decomposing.
+ * If not set, just do strcasecmp().
+ */
+#define _COMPARE_EQUIV 0x80000
+
+/* internal function */
+static int32_t
+unorm_cmpEquivFold(const UChar *s1, int32_t length1,
+                   const UChar *s2, int32_t length2,
+                   uint32_t options,
+                   UErrorCode *pErrorCode) {
+    const Normalizer2Impl *nfcImpl;
+
+    /* current-level start/limit - s1/s2 as current */
+    const UChar *start1, *start2, *limit1, *limit2;
+
+    /* decomposition and case folding variables */
+    const UChar *p;
+    int32_t length;
+
+    /* stacks of previous-level start/current/limit */
+    CmpEquivLevel stack1[2], stack2[2];
+
+    /* buffers for algorithmic decompositions */
+    UChar decomp1[4], decomp2[4];
+
+    /* case folding buffers, only use current-level start/limit */
+    UChar fold1[UCASE_MAX_STRING_LENGTH+1], fold2[UCASE_MAX_STRING_LENGTH+1];
+
+    /* track which is the current level per string */
+    int32_t level1, level2;
+
+    /* current code units, and code points for lookups */
+    UChar32 c1, c2, cp1, cp2;
+
+    /* no argument error checking because this itself is not an API */
+
+    /*
+     * assume that at least one of the options _COMPARE_EQUIV and U_COMPARE_IGNORE_CASE is set
+     * otherwise this function must behave exactly as uprv_strCompare()
+     * not checking for that here makes testing this function easier
+     */
+
+    /* normalization/properties data loaded? */
+    if((options&_COMPARE_EQUIV)!=0) {
+        nfcImpl=Normalizer2Factory::getNFCImpl(*pErrorCode);
+    } else {
+        nfcImpl=NULL;
+    }
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* initialize */
+    start1=s1;
+    if(length1==-1) {
+        limit1=NULL;
+    } else {
+        limit1=s1+length1;
+    }
+
+    start2=s2;
+    if(length2==-1) {
+        limit2=NULL;
+    } else {
+        limit2=s2+length2;
+    }
+
+    level1=level2=0;
+    c1=c2=-1;
+
+    /* comparison loop */
+    for(;;) {
+        /*
+         * here a code unit value of -1 means "get another code unit"
+         * below it will mean "this source is finished"
+         */
+
+        if(c1<0) {
+            /* get next code unit from string 1, post-increment */
+            for(;;) {
+                if(s1==limit1 || ((c1=*s1)==0 && (limit1==NULL || (options&_STRNCMP_STYLE)))) {
+                    if(level1==0) {
+                        c1=-1;
+                        break;
+                    }
+                } else {
+                    ++s1;
+                    break;
+                }
+
+                /* reached end of level buffer, pop one level */
+                do {
+                    --level1;
+                    start1=stack1[level1].start;    /*Not uninitialized*/
+                } while(start1==NULL);
+                s1=stack1[level1].s;                /*Not uninitialized*/
+                limit1=stack1[level1].limit;        /*Not uninitialized*/
+            }
+        }
+
+        if(c2<0) {
+            /* get next code unit from string 2, post-increment */
+            for(;;) {
+                if(s2==limit2 || ((c2=*s2)==0 && (limit2==NULL || (options&_STRNCMP_STYLE)))) {
+                    if(level2==0) {
+                        c2=-1;
+                        break;
+                    }
+                } else {
+                    ++s2;
+                    break;
+                }
+
+                /* reached end of level buffer, pop one level */
+                do {
+                    --level2;
+                    start2=stack2[level2].start;    /*Not uninitialized*/
+                } while(start2==NULL);
+                s2=stack2[level2].s;                /*Not uninitialized*/
+                limit2=stack2[level2].limit;        /*Not uninitialized*/
+            }
+        }
+
+        /*
+         * compare c1 and c2
+         * either variable c1, c2 is -1 only if the corresponding string is finished
+         */
+        if(c1==c2) {
+            if(c1<0) {
+                return 0;   /* c1==c2==-1 indicating end of strings */
+            }
+            c1=c2=-1;       /* make us fetch new code units */
+            continue;
+        } else if(c1<0) {
+            return -1;      /* string 1 ends before string 2 */
+        } else if(c2<0) {
+            return 1;       /* string 2 ends before string 1 */
+        }
+        /* c1!=c2 && c1>=0 && c2>=0 */
+
+        /* get complete code points for c1, c2 for lookups if either is a surrogate */
+        cp1=c1;
+        if(U_IS_SURROGATE(c1)) {
+            UChar c;
+
+            if(U_IS_SURROGATE_LEAD(c1)) {
+                if(s1!=limit1 && U16_IS_TRAIL(c=*s1)) {
+                    /* advance ++s1; only below if cp1 decomposes/case-folds */
+                    cp1=U16_GET_SUPPLEMENTARY(c1, c);
+                }
+            } else /* isTrail(c1) */ {
+                if(start1<=(s1-2) && U16_IS_LEAD(c=*(s1-2))) {
+                    cp1=U16_GET_SUPPLEMENTARY(c, c1);
+                }
+            }
+        }
+
+        cp2=c2;
+        if(U_IS_SURROGATE(c2)) {
+            UChar c;
+
+            if(U_IS_SURROGATE_LEAD(c2)) {
+                if(s2!=limit2 && U16_IS_TRAIL(c=*s2)) {
+                    /* advance ++s2; only below if cp2 decomposes/case-folds */
+                    cp2=U16_GET_SUPPLEMENTARY(c2, c);
+                }
+            } else /* isTrail(c2) */ {
+                if(start2<=(s2-2) && U16_IS_LEAD(c=*(s2-2))) {
+                    cp2=U16_GET_SUPPLEMENTARY(c, c2);
+                }
+            }
+        }
+
+        /*
+         * go down one level for each string
+         * continue with the main loop as soon as there is a real change
+         */
+
+        if( level1==0 && (options&U_COMPARE_IGNORE_CASE) &&
+            (length=ucase_toFullFolding((UChar32)cp1, &p, options))>=0
+        ) {
+            /* cp1 case-folds to the code point "length" or to p[length] */
+            if(U_IS_SURROGATE(c1)) {
+                if(U_IS_SURROGATE_LEAD(c1)) {
+                    /* advance beyond source surrogate pair if it case-folds */
+                    ++s1;
+                } else /* isTrail(c1) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s2;
+                    c2=*(s2-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack1[0].start=start1;
+            stack1[0].s=s1;
+            stack1[0].limit=limit1;
+            ++level1;
+
+            /* copy the folding result to fold1[] */
+            if(length<=UCASE_MAX_STRING_LENGTH) {
+                u_memcpy(fold1, p, length);
+            } else {
+                int32_t i=0;
+                U16_APPEND_UNSAFE(fold1, i, length);
+                length=i;
+            }
+
+            /* set next level pointers to case folding */
+            start1=s1=fold1;
+            limit1=fold1+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c1=-1;
+            continue;
+        }
+
+        if( level2==0 && (options&U_COMPARE_IGNORE_CASE) &&
+            (length=ucase_toFullFolding((UChar32)cp2, &p, options))>=0
+        ) {
+            /* cp2 case-folds to the code point "length" or to p[length] */
+            if(U_IS_SURROGATE(c2)) {
+                if(U_IS_SURROGATE_LEAD(c2)) {
+                    /* advance beyond source surrogate pair if it case-folds */
+                    ++s2;
+                } else /* isTrail(c2) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s1;
+                    c1=*(s1-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack2[0].start=start2;
+            stack2[0].s=s2;
+            stack2[0].limit=limit2;
+            ++level2;
+
+            /* copy the folding result to fold2[] */
+            if(length<=UCASE_MAX_STRING_LENGTH) {
+                u_memcpy(fold2, p, length);
+            } else {
+                int32_t i=0;
+                U16_APPEND_UNSAFE(fold2, i, length);
+                length=i;
+            }
+
+            /* set next level pointers to case folding */
+            start2=s2=fold2;
+            limit2=fold2+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c2=-1;
+            continue;
+        }
+
+        if( level1<2 && (options&_COMPARE_EQUIV) &&
+            0!=(p=nfcImpl->getDecomposition((UChar32)cp1, decomp1, length))
+        ) {
+            /* cp1 decomposes into p[length] */
+            if(U_IS_SURROGATE(c1)) {
+                if(U_IS_SURROGATE_LEAD(c1)) {
+                    /* advance beyond source surrogate pair if it decomposes */
+                    ++s1;
+                } else /* isTrail(c1) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s2;
+                    c2=*(s2-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack1[level1].start=start1;
+            stack1[level1].s=s1;
+            stack1[level1].limit=limit1;
+            ++level1;
+
+            /* set empty intermediate level if skipped */
+            if(level1<2) {
+                stack1[level1++].start=NULL;
+            }
+
+            /* set next level pointers to decomposition */
+            start1=s1=p;
+            limit1=p+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c1=-1;
+            continue;
+        }
+
+        if( level2<2 && (options&_COMPARE_EQUIV) &&
+            0!=(p=nfcImpl->getDecomposition((UChar32)cp2, decomp2, length))
+        ) {
+            /* cp2 decomposes into p[length] */
+            if(U_IS_SURROGATE(c2)) {
+                if(U_IS_SURROGATE_LEAD(c2)) {
+                    /* advance beyond source surrogate pair if it decomposes */
+                    ++s2;
+                } else /* isTrail(c2) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s1;
+                    c1=*(s1-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack2[level2].start=start2;
+            stack2[level2].s=s2;
+            stack2[level2].limit=limit2;
+            ++level2;
+
+            /* set empty intermediate level if skipped */
+            if(level2<2) {
+                stack2[level2++].start=NULL;
+            }
+
+            /* set next level pointers to decomposition */
+            start2=s2=p;
+            limit2=p+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c2=-1;
+            continue;
+        }
+
+        /*
+         * no decomposition/case folding, max level for both sides:
+         * return difference result
+         *
+         * code point order comparison must not just return cp1-cp2
+         * because when single surrogates are present then the surrogate pairs
+         * that formed cp1 and cp2 may be from different string indexes
+         *
+         * example: { d800 d800 dc01 } vs. { d800 dc00 }, compare at second code units
+         * c1=d800 cp1=10001 c2=dc00 cp2=10000
+         * cp1-cp2>0 but c1-c2<0 and in fact in UTF-32 it is { d800 10001 } < { 10000 }
+         *
+         * therefore, use same fix-up as in ustring.c/uprv_strCompare()
+         * except: uprv_strCompare() fetches c=*s while this functions fetches c=*s++
+         * so we have slightly different pointer/start/limit comparisons here
+         */
+
+        if(c1>=0xd800 && c2>=0xd800 && (options&U_COMPARE_CODE_POINT_ORDER)) {
+            /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
+            if(
+                (c1<=0xdbff && s1!=limit1 && U16_IS_TRAIL(*s1)) ||
+                (U16_IS_TRAIL(c1) && start1!=(s1-1) && U16_IS_LEAD(*(s1-2)))
+            ) {
+                /* part of a surrogate pair, leave >=d800 */
+            } else {
+                /* BMP code point - may be surrogate code point - make <d800 */
+                c1-=0x2800;
+            }
+
+            if(
+                (c2<=0xdbff && s2!=limit2 && U16_IS_TRAIL(*s2)) ||
+                (U16_IS_TRAIL(c2) && start2!=(s2-1) && U16_IS_LEAD(*(s2-2)))
+            ) {
+                /* part of a surrogate pair, leave >=d800 */
+            } else {
+                /* BMP code point - may be surrogate code point - make <d800 */
+                c2-=0x2800;
+            }
+        }
+
+        return c1-c2;
+    }
+}
+
+static
+UBool _normalize(const Normalizer2 *n2, const UChar *s, int32_t length,
+                UnicodeString &normalized, UErrorCode *pErrorCode) {
+    UnicodeString str(length<0, s, length);
+
+    // check if s fulfill the conditions
+    int32_t spanQCYes=n2->spanQuickCheckYes(str, *pErrorCode);
+    if (U_FAILURE(*pErrorCode)) {
+        return FALSE;
+    }
+    /*
+     * ICU 2.4 had a further optimization:
+     * If both strings were not in FCD, then they were both NFD'ed,
+     * and the _COMPARE_EQUIV option was turned off.
+     * It is not entirely clear that this is valid with the current
+     * definition of the canonical caseless match.
+     * Therefore, ICU 2.6 removes that optimization.
+     */
+    if(spanQCYes<str.length()) {
+        UnicodeString unnormalized=str.tempSubString(spanQCYes);
+        normalized.setTo(FALSE, str.getBuffer(), spanQCYes);
+        n2->normalizeSecondAndAppend(normalized, unnormalized, *pErrorCode);
+        if (U_SUCCESS(*pErrorCode)) {
+            return TRUE;
+        }
+    }
+    return FALSE;
+}
+
+U_CAPI int32_t U_EXPORT2
+unorm_compare(const UChar *s1, int32_t length1,
+              const UChar *s2, int32_t length2,
+              uint32_t options,
+              UErrorCode *pErrorCode) {
+    /* argument checking */
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(s1==0 || length1<-1 || s2==0 || length2<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    UnicodeString fcd1, fcd2;
+    int32_t normOptions=(int32_t)(options>>UNORM_COMPARE_NORM_OPTIONS_SHIFT);
+    options|=_COMPARE_EQUIV;
+
+    /*
+     * UAX #21 Case Mappings, as fixed for Unicode version 4
+     * (see Jitterbug 2021), defines a canonical caseless match as
+     *
+     * A string X is a canonical caseless match
+     * for a string Y if and only if
+     * NFD(toCasefold(NFD(X))) = NFD(toCasefold(NFD(Y)))
+     *
+     * For better performance, we check for FCD (or let the caller tell us that
+     * both strings are in FCD) for the inner normalization.
+     * BasicNormalizerTest::FindFoldFCDExceptions() makes sure that
+     * case-folding preserves the FCD-ness of a string.
+     * The outer normalization is then only performed by unorm_cmpEquivFold()
+     * when there is a difference.
+     *
+     * Exception: When using the Turkic case-folding option, we do perform
+     * full NFD first. This is because in the Turkic case precomposed characters
+     * with 0049 capital I or 0069 small i fold differently whether they
+     * are first decomposed or not, so an FCD check - a check only for
+     * canonical order - is not sufficient.
+     */
+    if(!(options&UNORM_INPUT_IS_FCD) || (options&U_FOLD_CASE_EXCLUDE_SPECIAL_I)) {
+        const Normalizer2 *n2;
+        if(options&U_FOLD_CASE_EXCLUDE_SPECIAL_I) {
+            n2=Normalizer2::getNFDInstance(*pErrorCode);
+        } else {
+            n2=Normalizer2Factory::getFCDInstance(*pErrorCode);
+        }
+        if (U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+
+        if(normOptions&UNORM_UNICODE_3_2) {
+            const UnicodeSet *uni32=uniset_getUnicode32Instance(*pErrorCode);
+            FilteredNormalizer2 fn2(*n2, *uni32);
+            if(_normalize(&fn2, s1, length1, fcd1, pErrorCode)) {
+                s1=fcd1.getBuffer();
+                length1=fcd1.length();
+            }
+            if(_normalize(&fn2, s2, length2, fcd2, pErrorCode)) {
+                s2=fcd2.getBuffer();
+                length2=fcd2.length();
+            }
+        } else {
+            if(_normalize(n2, s1, length1, fcd1, pErrorCode)) {
+                s1=fcd1.getBuffer();
+                length1=fcd1.length();
+            }
+            if(_normalize(n2, s2, length2, fcd2, pErrorCode)) {
+                s2=fcd2.getBuffer();
+                length2=fcd2.length();
+            }
+        }
+    }
+
+    if(U_SUCCESS(*pErrorCode)) {
+        return unorm_cmpEquivFold(s1, length1, s2, length2, options, pErrorCode);
+    } else {
+        return 0;
+    }
+}
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
diff --git a/thirdparty/icu4c/common/unormimp.h b/thirdparty/icu4c/common/unormimp.h
new file mode 100644
index 0000000000..d2604adb4a
--- /dev/null
+++ b/thirdparty/icu4c/common/unormimp.h
@@ -0,0 +1,488 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2001-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  unormimp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001may25
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UNORMIMP_H__
+#define __UNORMIMP_H__
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_NORMALIZATION
+
+#include "udataswp.h"
+
+/*
+ * The 2001-2010 implementation of the normalization code loads its data from
+ * unorm.icu, which is generated with the gennorm tool.
+ * The format of that file is described at the end of this file.
+ */
+
+/* norm32 value constants */
+enum {
+    /* quick check flags 0..3 set mean "no" for their forms */
+    _NORM_QC_NFC=0x11,          /* no|maybe */
+    _NORM_QC_NFKC=0x22,         /* no|maybe */
+    _NORM_QC_NFD=4,             /* no */
+    _NORM_QC_NFKD=8,            /* no */
+
+    _NORM_QC_ANY_NO=0xf,
+
+    /* quick check flags 4..5 mean "maybe" for their forms; test flags>=_NORM_QC_MAYBE */
+    _NORM_QC_MAYBE=0x10,
+    _NORM_QC_ANY_MAYBE=0x30,
+
+    _NORM_QC_MASK=0x3f,
+
+    _NORM_COMBINES_FWD=0x40,
+    _NORM_COMBINES_BACK=0x80,
+    _NORM_COMBINES_ANY=0xc0,
+
+    _NORM_CC_SHIFT=8,           /* UnicodeData.txt combining class in bits 15..8 */
+    _NORM_CC_MASK=0xff00,
+
+    _NORM_EXTRA_SHIFT=16,               /* 16 bits for the index to UChars and other extra data */
+    _NORM_EXTRA_INDEX_TOP=0xfc00,       /* start of surrogate specials after shift */
+
+    _NORM_EXTRA_SURROGATE_MASK=0x3ff,
+    _NORM_EXTRA_SURROGATE_TOP=0x3f0,    /* hangul etc. */
+
+    _NORM_EXTRA_HANGUL=_NORM_EXTRA_SURROGATE_TOP,
+    _NORM_EXTRA_JAMO_L,
+    _NORM_EXTRA_JAMO_V,
+    _NORM_EXTRA_JAMO_T
+};
+
+/* norm32 value constants using >16 bits */
+#define _NORM_MIN_SPECIAL       0xfc000000
+#define _NORM_SURROGATES_TOP    0xfff00000
+#define _NORM_MIN_HANGUL        0xfff00000
+#define _NORM_MIN_JAMO_V        0xfff20000
+#define _NORM_JAMO_V_TOP        0xfff30000
+
+/* value constants for auxTrie */
+enum {
+    _NORM_AUX_COMP_EX_SHIFT=10,
+    _NORM_AUX_UNSAFE_SHIFT=11,
+    _NORM_AUX_NFC_SKIPPABLE_F_SHIFT=12
+};
+
+#define _NORM_AUX_MAX_FNC           ((int32_t)1<<_NORM_AUX_COMP_EX_SHIFT)
+
+#define _NORM_AUX_FNC_MASK          (uint32_t)(_NORM_AUX_MAX_FNC-1)
+#define _NORM_AUX_COMP_EX_MASK      ((uint32_t)1<<_NORM_AUX_COMP_EX_SHIFT)
+#define _NORM_AUX_UNSAFE_MASK       ((uint32_t)1<<_NORM_AUX_UNSAFE_SHIFT)
+#define _NORM_AUX_NFC_SKIP_F_MASK   ((uint32_t)1<<_NORM_AUX_NFC_SKIPPABLE_F_SHIFT)
+
+/* canonStartSets[0..31] contains indexes for what is in the array */
+enum {
+    _NORM_SET_INDEX_CANON_SETS_LENGTH,      /* number of uint16_t in canonical starter sets */
+    _NORM_SET_INDEX_CANON_BMP_TABLE_LENGTH, /* number of uint16_t in the BMP search table (contains pairs) */
+    _NORM_SET_INDEX_CANON_SUPP_TABLE_LENGTH,/* number of uint16_t in the supplementary search table (contains triplets) */
+
+    /* from formatVersion 2.3: */
+    _NORM_SET_INDEX_NX_CJK_COMPAT_OFFSET,   /* uint16_t offset from canonStartSets[0] to the
+                                               exclusion set for CJK compatibility characters */
+    _NORM_SET_INDEX_NX_UNICODE32_OFFSET,    /* uint16_t offset from canonStartSets[0] to the
+                                               exclusion set for Unicode 3.2 characters */
+    _NORM_SET_INDEX_NX_RESERVED_OFFSET,     /* uint16_t offset from canonStartSets[0] to the
+                                               end of the previous exclusion set */
+
+    _NORM_SET_INDEX_TOP=32                  /* changing this requires a new formatVersion */
+};
+
+/* more constants for canonical starter sets */
+
+/* 14 bit indexes to canonical USerializedSets */
+#define _NORM_MAX_CANON_SETS            0x4000
+
+/* single-code point BMP sets are encoded directly in the search table except if result=0x4000..0x7fff */
+#define _NORM_CANON_SET_BMP_MASK        0xc000
+#define _NORM_CANON_SET_BMP_IS_INDEX    0x4000
+
+/* indexes[] value names */
+enum {
+    _NORM_INDEX_TRIE_SIZE,              /* number of bytes in normalization trie */
+    _NORM_INDEX_UCHAR_COUNT,            /* number of UChars in extra data */
+
+    _NORM_INDEX_COMBINE_DATA_COUNT,     /* number of uint16_t words for combining data */
+    _NORM_INDEX_COMBINE_FWD_COUNT,      /* number of code points that combine forward */
+    _NORM_INDEX_COMBINE_BOTH_COUNT,     /* number of code points that combine forward and backward */
+    _NORM_INDEX_COMBINE_BACK_COUNT,     /* number of code points that combine backward */
+
+    _NORM_INDEX_MIN_NFC_NO_MAYBE,       /* first code point with quick check NFC NO/MAYBE */
+    _NORM_INDEX_MIN_NFKC_NO_MAYBE,      /* first code point with quick check NFKC NO/MAYBE */
+    _NORM_INDEX_MIN_NFD_NO_MAYBE,       /* first code point with quick check NFD NO/MAYBE */
+    _NORM_INDEX_MIN_NFKD_NO_MAYBE,      /* first code point with quick check NFKD NO/MAYBE */
+
+    _NORM_INDEX_FCD_TRIE_SIZE,          /* number of bytes in FCD trie */
+
+    _NORM_INDEX_AUX_TRIE_SIZE,          /* number of bytes in the auxiliary trie */
+    _NORM_INDEX_CANON_SET_COUNT,        /* number of uint16_t in the array of serialized USet */
+
+    _NORM_INDEX_TOP=32                  /* changing this requires a new formatVersion */
+};
+
+enum {
+    /* FCD check: everything below this code point is known to have a 0 lead combining class */
+    _NORM_MIN_WITH_LEAD_CC=0x300
+};
+
+enum {
+    /**
+     * Bit 7 of the length byte for a decomposition string in extra data is
+     * a flag indicating whether the decomposition string is
+     * preceded by a 16-bit word with the leading and trailing cc
+     * of the decomposition (like for A-umlaut);
+     * if not, then both cc's are zero (like for compatibility ideographs).
+     */
+    _NORM_DECOMP_FLAG_LENGTH_HAS_CC=0x80,
+    /**
+     * Bits 6..0 of the length byte contain the actual length.
+     */
+    _NORM_DECOMP_LENGTH_MASK=0x7f
+};
+
+/** Constants for options flags for normalization. */
+enum {
+    /** Options bit 0, do not decompose Hangul syllables. */
+    UNORM_NX_HANGUL=1,
+    /** Options bit 1, do not decompose CJK compatibility characters. */
+    UNORM_NX_CJK_COMPAT=2
+};
+
+/**
+ * Description of the format of unorm.icu version 2.3.
+ *
+ * Main change from version 1 to version 2:
+ * Use of new, common UTrie instead of normalization-specific tries.
+ * Change to version 2.1: add third/auxiliary trie with associated data.
+ * Change to version 2.2: add skippable (f) flag data (_NORM_AUX_NFC_SKIP_F_MASK).
+ * Change to version 2.3: add serialized sets for normalization exclusions
+ *                        stored inside canonStartSets[]
+ *
+ * For more details of how to use the data structures see the code
+ * in unorm.cpp (runtime normalization code) and
+ * in gennorm.c and gennorm/store.c (build-time data generation).
+ *
+ * For the serialized format of UTrie see utrie.c/UTrieHeader.
+ *
+ * - Overall partition
+ *
+ * unorm.dat customarily begins with a UDataInfo structure, see udata.h and .c.
+ * After that there are the following structures:
+ *
+ * int32_t indexes[_NORM_INDEX_TOP];            -- _NORM_INDEX_TOP=32, see enum in this file
+ *
+ * UTrie normTrie;                              -- size in bytes=indexes[_NORM_INDEX_TRIE_SIZE]
+ * 
+ * uint16_t extraData[extraDataTop];            -- extraDataTop=indexes[_NORM_INDEX_UCHAR_COUNT]
+ *                                                 extraData[0] contains the number of units for
+ *                                                 FC_NFKC_Closure (formatVersion>=2.1)
+ *
+ * uint16_t combiningTable[combiningTableTop];  -- combiningTableTop=indexes[_NORM_INDEX_COMBINE_DATA_COUNT]
+ *                                                 combiningTableTop may include one 16-bit padding unit
+ *                                                 to make sure that fcdTrie is 32-bit-aligned
+ *
+ * UTrie fcdTrie;                               -- size in bytes=indexes[_NORM_INDEX_FCD_TRIE_SIZE]
+ *
+ * UTrie auxTrie;                               -- size in bytes=indexes[_NORM_INDEX_AUX_TRIE_SIZE]
+ *
+ * uint16_t canonStartSets[canonStartSetsTop]   -- canonStartSetsTop=indexes[_NORM_INDEX_CANON_SET_COUNT]
+ *                                                 serialized USets and binary search tables, see below
+ *
+ *
+ * The indexes array contains lengths and sizes of the following arrays and structures
+ * as well as the following values:
+ *  indexes[_NORM_INDEX_COMBINE_FWD_COUNT]=combineFwdTop
+ *      -- one more than the highest combining index computed for forward-only-combining characters
+ *  indexes[_NORM_INDEX_COMBINE_BOTH_COUNT]=combineBothTop-combineFwdTop
+ *      -- number of combining indexes computed for both-ways-combining characters
+ *  indexes[_NORM_INDEX_COMBINE_BACK_COUNT]=combineBackTop-combineBothTop
+ *      -- number of combining indexes computed for backward-only-combining characters
+ *
+ *  indexes[_NORM_INDEX_MIN_NF*_NO_MAYBE] (where *={ C, D, KC, KD })
+ *      -- first code point with a quick check NF* value of NO/MAYBE
+ *
+ *
+ * - Tries
+ *
+ * The main structures are two UTrie tables ("compact arrays"),
+ * each with one index array and one data array.
+ * See utrie.h and utrie.c.
+ *
+ *
+ * - Tries in unorm.dat
+ *
+ * The first trie (normTrie above)
+ * provides data for the NF* quick checks and normalization.
+ * The second trie (fcdTrie above) provides data just for FCD checks.
+ *
+ *
+ * - norm32 data words from the first trie
+ *
+ * The norm32Table contains one 32-bit word "norm32" per code point.
+ * It contains the following bit fields:
+ * 31..16   extra data index, _NORM_EXTRA_SHIFT is used to shift this field down
+ *          if this index is <_NORM_EXTRA_INDEX_TOP then it is an index into
+ *              extraData[] where variable-length normalization data for this
+ *              code point is found
+ *          if this index is <_NORM_EXTRA_INDEX_TOP+_NORM_EXTRA_SURROGATE_TOP
+ *              then this is a norm32 for a leading surrogate, and the index
+ *              value is used together with the following trailing surrogate
+ *              code unit in the second trie access
+ *          if this index is >=_NORM_EXTRA_INDEX_TOP+_NORM_EXTRA_SURROGATE_TOP
+ *              then this is a norm32 for a "special" character,
+ *              i.e., the character is a Hangul syllable or a Jamo
+ *              see _NORM_EXTRA_HANGUL etc.
+ *          generally, instead of extracting this index from the norm32 and
+ *              comparing it with the above constants,
+ *              the normalization code compares the entire norm32 value
+ *              with _NORM_MIN_SPECIAL, _NORM_SURROGATES_TOP, _NORM_MIN_HANGUL etc.
+ *
+ * 15..8    combining class (cc) according to UnicodeData.txt
+ *
+ *  7..6    _NORM_COMBINES_ANY flags, used in composition to see if a character
+ *              combines with any following or preceding character(s)
+ *              at all
+ *     7    _NORM_COMBINES_BACK
+ *     6    _NORM_COMBINES_FWD
+ *
+ *  5..0    quick check flags, set for "no" or "maybe", with separate flags for
+ *              each normalization form
+ *              the higher bits are "maybe" flags; for NF*D there are no such flags
+ *              the lower bits are "no" flags for all forms, in the same order
+ *              as the "maybe" flags,
+ *              which is (MSB to LSB): NFKD NFD NFKC NFC
+ *  5..4    _NORM_QC_ANY_MAYBE
+ *  3..0    _NORM_QC_ANY_NO
+ *              see further related constants
+ *
+ *
+ * - Extra data per code point
+ *
+ * "Extra data" is referenced by the index in norm32.
+ * It is variable-length data. It is only present, and only those parts
+ * of it are, as needed for a given character.
+ * The norm32 extra data index is added to the beginning of extraData[]
+ * to get to a vector of 16-bit words with data at the following offsets:
+ *
+ * [-1]     Combining index for composition.
+ *              Stored only if norm32&_NORM_COMBINES_ANY .
+ * [0]      Lengths of the canonical and compatibility decomposition strings.
+ *              Stored only if there are decompositions, i.e.,
+ *              if norm32&(_NORM_QC_NFD|_NORM_QC_NFKD)
+ *          High byte: length of NFKD, or 0 if none
+ *          Low byte: length of NFD, or 0 if none
+ *          Each length byte also has another flag:
+ *              Bit 7 of a length byte is set if there are non-zero
+ *              combining classes (cc's) associated with the respective
+ *              decomposition. If this flag is set, then the decomposition
+ *              is preceded by a 16-bit word that contains the
+ *              leading and trailing cc's.
+ *              Bits 6..0 of a length byte are the length of the
+ *              decomposition string, not counting the cc word.
+ * [1..n]   NFD
+ * [n+1..]  NFKD
+ *
+ * Each of the two decompositions consists of up to two parts:
+ * - The 16-bit words with the leading and trailing cc's.
+ *   This is only stored if bit 7 of the corresponding length byte
+ *   is set. In this case, at least one of the cc's is not zero.
+ *   High byte: leading cc==cc of the first code point in the decomposition string
+ *   Low byte: trailing cc==cc of the last code point in the decomposition string
+ * - The decomposition string in UTF-16, with length code units.
+ *
+ *
+ * - Combining indexes and combiningTable[]
+ *
+ * Combining indexes are stored at the [-1] offset of the extra data
+ * if the character combines forward or backward with any other characters.
+ * They are used for (re)composition in NF*C.
+ * Values of combining indexes are arranged according to whether a character
+ * combines forward, backward, or both ways:
+ *    forward-only < both ways < backward-only
+ *
+ * The index values for forward-only and both-ways combining characters
+ * are indexes into the combiningTable[].
+ * The index values for backward-only combining characters are simply
+ * incremented from the preceding index values to be unique.
+ *
+ * In the combiningTable[], a variable-length list
+ * of variable-length (back-index, code point) pair entries is stored
+ * for each forward-combining character.
+ *
+ * These back-indexes are the combining indexes of both-ways or backward-only
+ * combining characters that the forward-combining character combines with.
+ *
+ * Each list is sorted in ascending order of back-indexes.
+ * Each list is terminated with the last back-index having bit 15 set.
+ *
+ * Each pair (back-index, code point) takes up either 2 or 3
+ * 16-bit words.
+ * The first word of a list entry is the back-index, with its bit 15 set if
+ * this is the last pair in the list.
+ *
+ * The second word contains flags in bits 15..13 that determine
+ * if there is a third word and how the combined character is encoded:
+ * 15   set if there is a third word in this list entry
+ * 14   set if the result is a supplementary character
+ * 13   set if the result itself combines forward
+ *
+ * According to these bits 15..14 of the second word,
+ * the result character is encoded as follows:
+ * 00 or 01 The result is <=0x1fff and stored in bits 12..0 of
+ *          the second word.
+ * 10       The result is 0x2000..0xffff and stored in the third word.
+ *          Bits 12..0 of the second word are not used.
+ * 11       The result is a supplementary character.
+ *          Bits 9..0 of the leading surrogate are in bits 9..0 of
+ *          the second word.
+ *          Add 0xd800 to these bits to get the complete surrogate.
+ *          Bits 12..10 of the second word are not used.
+ *          The trailing surrogate is stored in the third word.
+ *
+ *
+ * - FCD trie
+ *
+ * The FCD trie is very simple.
+ * It is a folded trie with 16-bit data words.
+ * In each word, the high byte contains the leading cc of the character,
+ * and the low byte contains the trailing cc of the character.
+ * These cc's are the cc's of the first and last code points in the
+ * canonical decomposition of the character.
+ *
+ * Since all 16 bits are used for cc's, lead surrogates must be tested
+ * by checking the code unit instead of the trie data.
+ * This is done only if the 16-bit data word is not zero.
+ * If the code unit is a leading surrogate and the data word is not zero,
+ * then instead of cc's it contains the offset for the second trie lookup.
+ *
+ *
+ * - Auxiliary trie and data
+ *
+ * The auxiliary 16-bit trie contains data for additional properties.
+ * Bits
+ * 15..13   reserved
+ *     12   not NFC_Skippable (f) (formatVersion>=2.2)
+ *     11   flag: not a safe starter for canonical closure
+ *     10   composition exclusion
+ *  9.. 0   index into extraData[] to FC_NFKC_Closure string
+ *          (not for lead surrogate),
+ *          or lead surrogate offset (for lead surrogate, if 9..0 not zero)
+ *
+ * - FC_NFKC_Closure strings in extraData[]
+ *
+ * Strings are either stored as a single code unit or as the length
+ * followed by that many units.
+ *   const UChar *s=extraData+(index from auxTrie data bits 9..0);
+ *   int32_t length;
+ *   if(*s<0xff00) {
+ *     // s points to the single-unit string
+ *     length=1;
+ *   } else {
+ *     length=*s&0xff;
+ *     ++s;
+ *   }
+ *
+ * Conditions for "NF* Skippable" from Mark Davis' com.ibm.text.UCD.NFSkippable:
+ * (used in NormalizerTransliterator)
+ *
+ * A skippable character is
+ * a) unassigned, or ALL of the following:
+ * b) of combining class 0.
+ * c) not decomposed by this normalization form.
+ * AND if NFC or NFKC,
+ * d) can never compose with a previous character.
+ * e) can never compose with a following character.
+ * f) can never change if another character is added.
+ *    Example: a-breve might satisfy all but f, but if you
+ *    add an ogonek it changes to a-ogonek + breve
+ *
+ * a)..e) must be tested from norm32.
+ * Since f) is more complicated, the (not-)NFC_Skippable flag (f) is built
+ * into the auxiliary trie.
+ * The same bit is used for NFC and NFKC; (c) differs for them.
+ * As usual, we build the "not skippable" flags so that unassigned
+ * code points get a 0 bit.
+ * This bit is only valid after (a)..(e) test false; test NFD_NO before (f) as well.
+ * Test Hangul LV syllables entirely in code.
+ *
+ *
+ * - structure inside canonStartSets[]
+ *
+ * This array maps from code points c to sets of code points (USerializedSet).
+ * The result sets are the code points whose canonical decompositions start
+ * with c.
+ *
+ * canonStartSets[] contains the following sub-arrays:
+ *
+ * indexes[_NORM_SET_INDEX_TOP]
+ *   - contains lengths of sub-arrays etc.
+ *
+ * startSets[indexes[_NORM_SET_INDEX_CANON_SETS_LENGTH]-_NORM_SET_INDEX_TOP]
+ *   - contains serialized sets (USerializedSet) of canonical starters for
+ *     enumerating canonically equivalent strings
+ *     indexes[_NORM_SET_INDEX_CANON_SETS_LENGTH] includes _NORM_SET_INDEX_TOP
+ *     for details about the structure see uset.c
+ *
+ * bmpTable[indexes[_NORM_SET_INDEX_CANON_BMP_TABLE_LENGTH]]
+ *   - a sorted search table for BMP code points whose results are
+ *     either indexes to USerializedSets or single code points for
+ *     single-code point sets;
+ *     each entry is a pair of { code point, result } with result=(binary) yy xxxxxx xxxxxxxx
+ *     if yy==01 then there is a USerializedSet at canonStartSets+x
+ *     else build a USerializedSet with result as the single code point
+ *
+ * suppTable[indexes[_NORM_SET_INDEX_CANON_SUPP_TABLE_LENGTH]]
+ *   - a sorted search table for supplementary code points whose results are
+ *     either indexes to USerializedSets or single code points for
+ *     single-code point sets;
+ *     each entry is a triplet of { high16(cp), low16(cp), result }
+ *     each code point's high-word may contain extra data in bits 15..5:
+ *     if the high word has bit 15 set, then build a set with a single code point
+ *     which is (((high16(cp)&0x1f00)<<8)|result;
+ *     else there is a USerializedSet at canonStartSets+result
+ *
+ * FormatVersion 2.3 adds 2 serialized sets for normalization exclusions.
+ * They are stored in the data file so that the runtime normalization code need
+ * not depend on other properties and their data and implementation files.
+ * The _NORM_SET_INDEX_NX_..._OFFSET offsets in the canonStartSets index table
+ * give the location for each set.
+ * There is no set stored for UNORM_NX_HANGUL because it's trivial to create
+ * without using properties.
+ *
+ * Set contents:
+ *
+ * _NORM_SET_INDEX_NX_CJK_COMPAT_OFFSET (for UNORM_NX_CJK_COMPAT)
+ *     [[:Ideographic:]&[:NFD_QC=No:]]
+ *     =[CJK Ideographs]&[has canonical decomposition]
+ *
+ * _NORM_SET_INDEX_NX_UNICODE32_OFFSET (for UNORM_UNICODE_3_2)
+ *     [:^Age=3.2:]
+ *     =set with all code points that were not designated by the specified Unicode version
+ *
+ * _NORM_SET_INDEX_NX_RESERVED_OFFSET
+ *     This is an offset that points to where the next, future set would start.
+ *     Currently it indicates where the previous set ends, and thus its length.
+ *     The name for this enum constant may in the future be applied to different
+ *     index slots. In order to get the limit of a set, use its index slot and
+ *     the immediately following one regardless of that one's enum name.
+ */
+
+#endif /* #if !UCONFIG_NO_NORMALIZATION */
+
+#endif
diff --git a/thirdparty/icu4c/common/uobject.cpp b/thirdparty/icu4c/common/uobject.cpp
new file mode 100644
index 0000000000..e222b2ce9b
--- /dev/null
+++ b/thirdparty/icu4c/common/uobject.cpp
@@ -0,0 +1,105 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2002-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  uobject.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002jun26
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/uobject.h"
+#include "cmemory.h"
+
+U_NAMESPACE_BEGIN
+
+#if U_OVERRIDE_CXX_ALLOCATION
+
+/*
+ * Default implementation of UMemory::new/delete
+ * using uprv_malloc() and uprv_free().
+ *
+ * For testing, this is used together with a list of imported symbols to verify
+ * that ICU is not using the global ::new and ::delete operators.
+ *
+ * These operators can be implemented like this or any other appropriate way
+ * when customizing ICU for certain environments.
+ * Whenever ICU is customized in binary incompatible ways please be sure
+ * to use library name suffixes to distinguish such libraries from
+ * the standard build.
+ *
+ * Instead of just modifying these C++ new/delete operators, it is usually best
+ * to modify the uprv_malloc()/uprv_free()/uprv_realloc() functions in cmemory.c.
+ *
+ * Memory test on Windows/MSVC 6:
+ * The global operators new and delete look as follows:
+ *   04F 00000000 UNDEF  notype ()    External     | ??2@YAPAXI@Z (void * __cdecl operator new(unsigned int))
+ *   03F 00000000 UNDEF  notype ()    External     | ??3@YAXPAX@Z (void __cdecl operator delete(void *))
+ *
+ * These lines are from output generated by the MSVC 6 tool dumpbin with
+ * dumpbin /symbols *.obj
+ *
+ * ??2@YAPAXI@Z and ??3@YAXPAX@Z are the linker symbols in the .obj
+ * files and are imported from msvcrtd.dll (in a debug build).
+ *
+ * Make sure that with the UMemory operators new and delete defined these two symbols
+ * do not appear in the dumpbin /symbols output for the ICU libraries!
+ *
+ * If such a symbol appears in the output then look in the preceding lines in the output
+ * for which file and function calls the global new or delete operator,
+ * and replace with uprv_malloc/uprv_free.
+ */
+
+void * U_EXPORT2 UMemory::operator new(size_t size) U_NOEXCEPT {
+    return uprv_malloc(size);
+}
+
+void U_EXPORT2 UMemory::operator delete(void *p) U_NOEXCEPT {
+    if(p!=NULL) {
+        uprv_free(p);
+    }
+}
+
+void * U_EXPORT2 UMemory::operator new[](size_t size) U_NOEXCEPT {
+    return uprv_malloc(size);
+}
+
+void U_EXPORT2 UMemory::operator delete[](void *p) U_NOEXCEPT {
+    if(p!=NULL) {
+        uprv_free(p);
+    }
+}
+
+#if U_HAVE_DEBUG_LOCATION_NEW
+void * U_EXPORT2 UMemory::operator new(size_t size, const char* /*file*/, int /*line*/) U_NOEXCEPT {
+    return UMemory::operator new(size);
+}
+
+void U_EXPORT2 UMemory::operator delete(void* p, const char* /*file*/, int /*line*/) U_NOEXCEPT {
+    UMemory::operator delete(p);
+}
+#endif /* U_HAVE_DEBUG_LOCATION_NEW */
+
+
+#endif
+
+UObject::~UObject() {}
+
+UClassID UObject::getDynamicClassID() const { return NULL; }
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI void U_EXPORT2
+uprv_deleteUObject(void *obj) {
+    delete static_cast<UObject *>(obj);
+}
diff --git a/thirdparty/icu4c/common/uposixdefs.h b/thirdparty/icu4c/common/uposixdefs.h
new file mode 100644
index 0000000000..23c3f6d466
--- /dev/null
+++ b/thirdparty/icu4c/common/uposixdefs.h
@@ -0,0 +1,77 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  uposixdefs.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011jul25
+*   created by: Markus W. Scherer
+*
+*   Common definitions for implementation files working with POSIX functions.
+*   *Important*: #include this file before any other header files!
+*/
+
+#ifndef __UPOSIXDEFS_H__
+#define __UPOSIXDEFS_H__
+
+/*
+ * Define _XOPEN_SOURCE for access to POSIX functions.
+ *
+ * We cannot use U_PLATFORM from platform.h/utypes.h because
+ * "The Open Group Base Specifications"
+ * chapter "2.2 The Compilation Environment" says:
+ * "In the compilation of an application that #defines a feature test macro
+ * specified by IEEE Std 1003.1-2001,
+ * no header defined by IEEE Std 1003.1-2001 shall be included prior to
+ * the definition of the feature test macro."
+ */
+#ifdef _XOPEN_SOURCE
+    /* Use the predefined value. */
+#else
+    /*
+     * Version 6.0:
+     * The Open Group Base Specifications Issue 6 (IEEE Std 1003.1, 2004 Edition)
+     * also known as
+     * SUSv3 = Open Group Single UNIX Specification, Version 3 (UNIX03)
+     *
+     * Note: This definition used to be in C source code (e.g., putil.c)
+     * and define _XOPEN_SOURCE to different values depending on __STDC_VERSION__.
+     * In C++ source code (e.g., putil.cpp), __STDC_VERSION__ is not defined at all.
+     */
+#   define _XOPEN_SOURCE 600
+#endif
+
+/*
+ * Make sure things like readlink and such functions work.
+ * Poorly upgraded Solaris machines can't have this defined.
+ * Cleanly installed Solaris can use this #define.
+ *
+ * z/OS needs this definition for timeval and to get usleep.
+ */
+#if !defined(_XOPEN_SOURCE_EXTENDED) && defined(__TOS_MVS__)
+#   define _XOPEN_SOURCE_EXTENDED 1
+#endif
+
+/**
+ * Solaris says:
+ *   "...it is invalid to compile an XPG6 or a POSIX.1-2001 application with anything other
+ *   than a c99 or later compiler."
+ * Apparently C++11 is not "or later". Work around this.
+ */
+#if defined(__cplusplus) && (defined(sun) || defined(__sun)) && !defined (_STDC_C99)
+#   define _STDC_C99
+#endif
+
+#if !defined _POSIX_C_SOURCE && \
+    defined(__APPLE__) && defined(__MACH__) && !defined(__clang__)
+// Needed to prevent EOWNERDEAD issues with GCC on Mac
+#define _POSIX_C_SOURCE 200809L
+#endif
+
+#endif  /* __UPOSIXDEFS_H__ */
diff --git a/thirdparty/icu4c/common/uprops.cpp b/thirdparty/icu4c/common/uprops.cpp
new file mode 100644
index 0000000000..1604ad9a17
--- /dev/null
+++ b/thirdparty/icu4c/common/uprops.cpp
@@ -0,0 +1,797 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uprops.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002feb24
+*   created by: Markus W. Scherer
+*
+*   Implementations for mostly non-core Unicode character properties
+*   stored in uprops.icu.
+*
+*   With the APIs implemented here, almost all properties files and
+*   their associated implementation files are used from this file,
+*   including those for normalization and case mappings.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+#include "unicode/ucptrie.h"
+#include "unicode/udata.h"
+#include "unicode/unorm2.h"
+#include "unicode/uscript.h"
+#include "unicode/ustring.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "normalizer2impl.h"
+#include "umutex.h"
+#include "ubidi_props.h"
+#include "uprops.h"
+#include "ucase.h"
+#include "ucln_cmn.h"
+#include "ulayout_props.h"
+#include "ustr_imp.h"
+
+U_NAMESPACE_USE
+
+// Unicode text layout properties data -----------------------------------------
+
+namespace {
+
+icu::UInitOnce gLayoutInitOnce = U_INITONCE_INITIALIZER;
+UDataMemory *gLayoutMemory = nullptr;
+
+UCPTrie *gInpcTrie = nullptr;  // Indic_Positional_Category
+UCPTrie *gInscTrie = nullptr;  // Indic_Syllabic_Category
+UCPTrie *gVoTrie = nullptr;  // Vertical_Orientation
+
+int32_t gMaxInpcValue = 0;
+int32_t gMaxInscValue = 0;
+int32_t gMaxVoValue = 0;
+
+UBool U_CALLCONV uprops_cleanup() {
+    udata_close(gLayoutMemory);
+    gLayoutMemory = nullptr;
+
+    ucptrie_close(gInpcTrie);
+    gInpcTrie = nullptr;
+    ucptrie_close(gInscTrie);
+    gInscTrie = nullptr;
+    ucptrie_close(gVoTrie);
+    gVoTrie = nullptr;
+
+    gMaxInpcValue = 0;
+    gMaxInscValue = 0;
+    gMaxVoValue = 0;
+
+    gLayoutInitOnce.reset();
+    return TRUE;
+}
+
+UBool U_CALLCONV
+ulayout_isAcceptable(void * /*context*/,
+                     const char * /* type */, const char * /*name*/,
+                     const UDataInfo *pInfo) {
+    return pInfo->size >= 20 &&
+        pInfo->isBigEndian == U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily == U_CHARSET_FAMILY &&
+        pInfo->dataFormat[0] == ULAYOUT_FMT_0 &&
+        pInfo->dataFormat[1] == ULAYOUT_FMT_1 &&
+        pInfo->dataFormat[2] == ULAYOUT_FMT_2 &&
+        pInfo->dataFormat[3] == ULAYOUT_FMT_3 &&
+        pInfo->formatVersion[0] == 1;
+}
+
+// UInitOnce singleton initialization function
+void U_CALLCONV ulayout_load(UErrorCode &errorCode) {
+    gLayoutMemory = udata_openChoice(
+        nullptr, ULAYOUT_DATA_TYPE, ULAYOUT_DATA_NAME,
+        ulayout_isAcceptable, nullptr, &errorCode);
+    if (U_FAILURE(errorCode)) { return; }
+
+    const uint8_t *inBytes = (const uint8_t *)udata_getMemory(gLayoutMemory);
+    const int32_t *inIndexes = (const int32_t *)inBytes;
+    int32_t indexesLength = inIndexes[ULAYOUT_IX_INDEXES_LENGTH];
+    if (indexesLength < 12) {
+        errorCode = U_INVALID_FORMAT_ERROR;  // Not enough indexes.
+        return;
+    }
+    int32_t offset = indexesLength * 4;
+    int32_t top = inIndexes[ULAYOUT_IX_INPC_TRIE_TOP];
+    int32_t trieSize = top - offset;
+    if (trieSize >= 16) {
+        gInpcTrie = ucptrie_openFromBinary(
+            UCPTRIE_TYPE_ANY, UCPTRIE_VALUE_BITS_ANY,
+            inBytes + offset, trieSize, nullptr, &errorCode);
+    }
+    offset = top;
+    top = inIndexes[ULAYOUT_IX_INSC_TRIE_TOP];
+    trieSize = top - offset;
+    if (trieSize >= 16) {
+        gInscTrie = ucptrie_openFromBinary(
+            UCPTRIE_TYPE_ANY, UCPTRIE_VALUE_BITS_ANY,
+            inBytes + offset, trieSize, nullptr, &errorCode);
+    }
+    offset = top;
+    top = inIndexes[ULAYOUT_IX_VO_TRIE_TOP];
+    trieSize = top - offset;
+    if (trieSize >= 16) {
+        gVoTrie = ucptrie_openFromBinary(
+            UCPTRIE_TYPE_ANY, UCPTRIE_VALUE_BITS_ANY,
+            inBytes + offset, trieSize, nullptr, &errorCode);
+    }
+
+    uint32_t maxValues = inIndexes[ULAYOUT_IX_MAX_VALUES];
+    gMaxInpcValue = maxValues >> ULAYOUT_MAX_INPC_SHIFT;
+    gMaxInscValue = (maxValues >> ULAYOUT_MAX_INSC_SHIFT) & 0xff;
+    gMaxVoValue = (maxValues >> ULAYOUT_MAX_VO_SHIFT) & 0xff;
+
+    ucln_common_registerCleanup(UCLN_COMMON_UPROPS, uprops_cleanup);
+}
+
+UBool ulayout_ensureData(UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return FALSE; }
+    umtx_initOnce(gLayoutInitOnce, &ulayout_load, errorCode);
+    return U_SUCCESS(errorCode);
+}
+
+UBool ulayout_ensureData() {
+    UErrorCode errorCode = U_ZERO_ERROR;
+    return ulayout_ensureData(errorCode);
+}
+
+}  // namespace
+
+/* general properties API functions ----------------------------------------- */
+
+struct BinaryProperty;
+
+typedef UBool BinaryPropertyContains(const BinaryProperty &prop, UChar32 c, UProperty which);
+
+struct BinaryProperty {
+    int32_t column;  // SRC_PROPSVEC column, or "source" if mask==0
+    uint32_t mask;
+    BinaryPropertyContains *contains;
+};
+
+static UBool defaultContains(const BinaryProperty &prop, UChar32 c, UProperty /*which*/) {
+    /* systematic, directly stored properties */
+    return (u_getUnicodeProperties(c, prop.column)&prop.mask)!=0;
+}
+
+static UBool caseBinaryPropertyContains(const BinaryProperty &/*prop*/, UChar32 c, UProperty which) {
+    return static_cast<UBool>(ucase_hasBinaryProperty(c, which));
+}
+
+static UBool isBidiControl(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return ubidi_isBidiControl(c);
+}
+
+static UBool isMirrored(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return ubidi_isMirrored(c);
+}
+
+static UBool isJoinControl(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return ubidi_isJoinControl(c);
+}
+
+#if UCONFIG_NO_NORMALIZATION
+static UBool hasFullCompositionExclusion(const BinaryProperty &, UChar32, UProperty) {
+    return FALSE;
+}
+#else
+static UBool hasFullCompositionExclusion(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    // By definition, Full_Composition_Exclusion is the same as NFC_QC=No.
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+    return U_SUCCESS(errorCode) && impl->isCompNo(impl->getNorm16(c));
+}
+#endif
+
+// UCHAR_NF*_INERT properties
+#if UCONFIG_NO_NORMALIZATION
+static UBool isNormInert(const BinaryProperty &, UChar32, UProperty) {
+    return FALSE;
+}
+#else
+static UBool isNormInert(const BinaryProperty &/*prop*/, UChar32 c, UProperty which) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2 *norm2=Normalizer2Factory::getInstance(
+        (UNormalizationMode)(which-UCHAR_NFD_INERT+UNORM_NFD), errorCode);
+    return U_SUCCESS(errorCode) && norm2->isInert(c);
+}
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+static UBool changesWhenCasefolded(const BinaryProperty &, UChar32, UProperty) {
+    return FALSE;
+}
+#else
+static UBool changesWhenCasefolded(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    UnicodeString nfd;
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2 *nfcNorm2=Normalizer2::getNFCInstance(errorCode);
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    if(nfcNorm2->getDecomposition(c, nfd)) {
+        /* c has a decomposition */
+        if(nfd.length()==1) {
+            c=nfd[0];  /* single BMP code point */
+        } else if(nfd.length()<=U16_MAX_LENGTH &&
+                  nfd.length()==U16_LENGTH(c=nfd.char32At(0))
+        ) {
+            /* single supplementary code point */
+        } else {
+            c=U_SENTINEL;
+        }
+    } else if(c<0) {
+        return FALSE;  /* protect against bad input */
+    }
+    if(c>=0) {
+        /* single code point */
+        const UChar *resultString;
+        return (UBool)(ucase_toFullFolding(c, &resultString, U_FOLD_CASE_DEFAULT)>=0);
+    } else {
+        /* guess some large but stack-friendly capacity */
+        UChar dest[2*UCASE_MAX_STRING_LENGTH];
+        int32_t destLength;
+        destLength=u_strFoldCase(dest, UPRV_LENGTHOF(dest),
+                                  nfd.getBuffer(), nfd.length(),
+                                  U_FOLD_CASE_DEFAULT, &errorCode);
+        return (UBool)(U_SUCCESS(errorCode) &&
+                       0!=u_strCompare(nfd.getBuffer(), nfd.length(),
+                                       dest, destLength, FALSE));
+    }
+}
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+static UBool changesWhenNFKC_Casefolded(const BinaryProperty &, UChar32, UProperty) {
+    return FALSE;
+}
+#else
+static UBool changesWhenNFKC_Casefolded(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2Impl *kcf=Normalizer2Factory::getNFKC_CFImpl(errorCode);
+    if(U_FAILURE(errorCode)) {
+        return FALSE;
+    }
+    UnicodeString src(c);
+    UnicodeString dest;
+    {
+        // The ReorderingBuffer must be in a block because its destructor
+        // needs to release dest's buffer before we look at its contents.
+        ReorderingBuffer buffer(*kcf, dest);
+        // Small destCapacity for NFKC_CF(c).
+        if(buffer.init(5, errorCode)) {
+            const UChar *srcArray=src.getBuffer();
+            kcf->compose(srcArray, srcArray+src.length(), FALSE,
+                          TRUE, buffer, errorCode);
+        }
+    }
+    return U_SUCCESS(errorCode) && dest!=src;
+}
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+static UBool isCanonSegmentStarter(const BinaryProperty &, UChar32, UProperty) {
+    return FALSE;
+}
+#else
+static UBool isCanonSegmentStarter(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
+    return
+        U_SUCCESS(errorCode) && impl->ensureCanonIterData(errorCode) &&
+        impl->isCanonSegmentStarter(c);
+}
+#endif
+
+static UBool isPOSIX_alnum(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_isalnumPOSIX(c);
+}
+
+static UBool isPOSIX_blank(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_isblank(c);
+}
+
+static UBool isPOSIX_graph(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_isgraphPOSIX(c);
+}
+
+static UBool isPOSIX_print(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_isprintPOSIX(c);
+}
+
+static UBool isPOSIX_xdigit(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_isxdigit(c);
+}
+
+static UBool isRegionalIndicator(const BinaryProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    // Property starts are a subset of lb=RI etc.
+    return 0x1F1E6<=c && c<=0x1F1FF;
+}
+
+static const BinaryProperty binProps[UCHAR_BINARY_LIMIT]={
+    /*
+     * column and mask values for binary properties from u_getUnicodeProperties().
+     * Must be in order of corresponding UProperty,
+     * and there must be exactly one entry per binary UProperty.
+     *
+     * Properties with mask==0 are handled in code.
+     * For them, column is the UPropertySource value.
+     */
+    { 1,                U_MASK(UPROPS_ALPHABETIC), defaultContains },
+    { 1,                U_MASK(UPROPS_ASCII_HEX_DIGIT), defaultContains },
+    { UPROPS_SRC_BIDI,  0, isBidiControl },
+    { UPROPS_SRC_BIDI,  0, isMirrored },
+    { 1,                U_MASK(UPROPS_DASH), defaultContains },
+    { 1,                U_MASK(UPROPS_DEFAULT_IGNORABLE_CODE_POINT), defaultContains },
+    { 1,                U_MASK(UPROPS_DEPRECATED), defaultContains },
+    { 1,                U_MASK(UPROPS_DIACRITIC), defaultContains },
+    { 1,                U_MASK(UPROPS_EXTENDER), defaultContains },
+    { UPROPS_SRC_NFC,   0, hasFullCompositionExclusion },
+    { 1,                U_MASK(UPROPS_GRAPHEME_BASE), defaultContains },
+    { 1,                U_MASK(UPROPS_GRAPHEME_EXTEND), defaultContains },
+    { 1,                U_MASK(UPROPS_GRAPHEME_LINK), defaultContains },
+    { 1,                U_MASK(UPROPS_HEX_DIGIT), defaultContains },
+    { 1,                U_MASK(UPROPS_HYPHEN), defaultContains },
+    { 1,                U_MASK(UPROPS_ID_CONTINUE), defaultContains },
+    { 1,                U_MASK(UPROPS_ID_START), defaultContains },
+    { 1,                U_MASK(UPROPS_IDEOGRAPHIC), defaultContains },
+    { 1,                U_MASK(UPROPS_IDS_BINARY_OPERATOR), defaultContains },
+    { 1,                U_MASK(UPROPS_IDS_TRINARY_OPERATOR), defaultContains },
+    { UPROPS_SRC_BIDI,  0, isJoinControl },
+    { 1,                U_MASK(UPROPS_LOGICAL_ORDER_EXCEPTION), defaultContains },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_LOWERCASE
+    { 1,                U_MASK(UPROPS_MATH), defaultContains },
+    { 1,                U_MASK(UPROPS_NONCHARACTER_CODE_POINT), defaultContains },
+    { 1,                U_MASK(UPROPS_QUOTATION_MARK), defaultContains },
+    { 1,                U_MASK(UPROPS_RADICAL), defaultContains },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_SOFT_DOTTED
+    { 1,                U_MASK(UPROPS_TERMINAL_PUNCTUATION), defaultContains },
+    { 1,                U_MASK(UPROPS_UNIFIED_IDEOGRAPH), defaultContains },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_UPPERCASE
+    { 1,                U_MASK(UPROPS_WHITE_SPACE), defaultContains },
+    { 1,                U_MASK(UPROPS_XID_CONTINUE), defaultContains },
+    { 1,                U_MASK(UPROPS_XID_START), defaultContains },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CASE_SENSITIVE
+    { 1,                U_MASK(UPROPS_S_TERM), defaultContains },
+    { 1,                U_MASK(UPROPS_VARIATION_SELECTOR), defaultContains },
+    { UPROPS_SRC_NFC,   0, isNormInert },  // UCHAR_NFD_INERT
+    { UPROPS_SRC_NFKC,  0, isNormInert },  // UCHAR_NFKD_INERT
+    { UPROPS_SRC_NFC,   0, isNormInert },  // UCHAR_NFC_INERT
+    { UPROPS_SRC_NFKC,  0, isNormInert },  // UCHAR_NFKC_INERT
+    { UPROPS_SRC_NFC_CANON_ITER, 0, isCanonSegmentStarter },
+    { 1,                U_MASK(UPROPS_PATTERN_SYNTAX), defaultContains },
+    { 1,                U_MASK(UPROPS_PATTERN_WHITE_SPACE), defaultContains },
+    { UPROPS_SRC_CHAR_AND_PROPSVEC,  0, isPOSIX_alnum },
+    { UPROPS_SRC_CHAR,  0, isPOSIX_blank },
+    { UPROPS_SRC_CHAR,  0, isPOSIX_graph },
+    { UPROPS_SRC_CHAR,  0, isPOSIX_print },
+    { UPROPS_SRC_CHAR,  0, isPOSIX_xdigit },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CASED
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CASE_IGNORABLE
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CHANGES_WHEN_LOWERCASED
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CHANGES_WHEN_UPPERCASED
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CHANGES_WHEN_TITLECASED
+    { UPROPS_SRC_CASE_AND_NORM,  0, changesWhenCasefolded },
+    { UPROPS_SRC_CASE,  0, caseBinaryPropertyContains },  // UCHAR_CHANGES_WHEN_CASEMAPPED
+    { UPROPS_SRC_NFKC_CF, 0, changesWhenNFKC_Casefolded },
+    { 2,                U_MASK(UPROPS_2_EMOJI), defaultContains },
+    { 2,                U_MASK(UPROPS_2_EMOJI_PRESENTATION), defaultContains },
+    { 2,                U_MASK(UPROPS_2_EMOJI_MODIFIER), defaultContains },
+    { 2,                U_MASK(UPROPS_2_EMOJI_MODIFIER_BASE), defaultContains },
+    { 2,                U_MASK(UPROPS_2_EMOJI_COMPONENT), defaultContains },
+    { 2,                0, isRegionalIndicator },
+    { 1,                U_MASK(UPROPS_PREPENDED_CONCATENATION_MARK), defaultContains },
+    { 2,                U_MASK(UPROPS_2_EXTENDED_PICTOGRAPHIC), defaultContains },
+};
+
+U_CAPI UBool U_EXPORT2
+u_hasBinaryProperty(UChar32 c, UProperty which) {
+    /* c is range-checked in the functions that are called from here */
+    if(which<UCHAR_BINARY_START || UCHAR_BINARY_LIMIT<=which) {
+        /* not a known binary property */
+        return FALSE;
+    } else {
+        const BinaryProperty &prop=binProps[which];
+        return prop.contains(prop, c, which);
+    }
+}
+
+struct IntProperty;
+
+typedef int32_t IntPropertyGetValue(const IntProperty &prop, UChar32 c, UProperty which);
+typedef int32_t IntPropertyGetMaxValue(const IntProperty &prop, UProperty which);
+
+struct IntProperty {
+    int32_t column;  // SRC_PROPSVEC column, or "source" if mask==0
+    uint32_t mask;
+    int32_t shift;  // =maxValue if getMaxValueFromShift() is used
+    IntPropertyGetValue *getValue;
+    IntPropertyGetMaxValue *getMaxValue;
+};
+
+static int32_t defaultGetValue(const IntProperty &prop, UChar32 c, UProperty /*which*/) {
+    /* systematic, directly stored properties */
+    return (int32_t)(u_getUnicodeProperties(c, prop.column)&prop.mask)>>prop.shift;
+}
+
+static int32_t defaultGetMaxValue(const IntProperty &prop, UProperty /*which*/) {
+    return (uprv_getMaxValues(prop.column)&prop.mask)>>prop.shift;
+}
+
+static int32_t getMaxValueFromShift(const IntProperty &prop, UProperty /*which*/) {
+    return prop.shift;
+}
+
+static int32_t getBiDiClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return (int32_t)u_charDirection(c);
+}
+
+static int32_t getBiDiPairedBracketType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return (int32_t)ubidi_getPairedBracketType(c);
+}
+
+static int32_t biDiGetMaxValue(const IntProperty &/*prop*/, UProperty which) {
+    return ubidi_getMaxValue(which);
+}
+
+#if UCONFIG_NO_NORMALIZATION
+static int32_t getCombiningClass(const IntProperty &, UChar32, UProperty) {
+    return 0;
+}
+#else
+static int32_t getCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return u_getCombiningClass(c);
+}
+#endif
+
+static int32_t getGeneralCategory(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return (int32_t)u_charType(c);
+}
+
+static int32_t getJoiningGroup(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return ubidi_getJoiningGroup(c);
+}
+
+static int32_t getJoiningType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return ubidi_getJoiningType(c);
+}
+
+static int32_t getNumericType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    int32_t ntv=(int32_t)GET_NUMERIC_TYPE_VALUE(u_getMainProperties(c));
+    return UPROPS_NTV_GET_TYPE(ntv);
+}
+
+static int32_t getScript(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    return (int32_t)uscript_getScript(c, &errorCode);
+}
+
+static int32_t scriptGetMaxValue(const IntProperty &/*prop*/, UProperty /*which*/) {
+    uint32_t scriptX=uprv_getMaxValues(0)&UPROPS_SCRIPT_X_MASK;
+    return uprops_mergeScriptCodeOrIndex(scriptX);
+}
+
+/*
+ * Map some of the Grapheme Cluster Break values to Hangul Syllable Types.
+ * Hangul_Syllable_Type is fully redundant with a subset of Grapheme_Cluster_Break.
+ */
+static const UHangulSyllableType gcbToHst[]={
+    U_HST_NOT_APPLICABLE,   /* U_GCB_OTHER */
+    U_HST_NOT_APPLICABLE,   /* U_GCB_CONTROL */
+    U_HST_NOT_APPLICABLE,   /* U_GCB_CR */
+    U_HST_NOT_APPLICABLE,   /* U_GCB_EXTEND */
+    U_HST_LEADING_JAMO,     /* U_GCB_L */
+    U_HST_NOT_APPLICABLE,   /* U_GCB_LF */
+    U_HST_LV_SYLLABLE,      /* U_GCB_LV */
+    U_HST_LVT_SYLLABLE,     /* U_GCB_LVT */
+    U_HST_TRAILING_JAMO,    /* U_GCB_T */
+    U_HST_VOWEL_JAMO        /* U_GCB_V */
+    /*
+     * Omit GCB values beyond what we need for hst.
+     * The code below checks for the array length.
+     */
+};
+
+static int32_t getHangulSyllableType(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    /* see comments on gcbToHst[] above */
+    int32_t gcb=(int32_t)(u_getUnicodeProperties(c, 2)&UPROPS_GCB_MASK)>>UPROPS_GCB_SHIFT;
+    if(gcb<UPRV_LENGTHOF(gcbToHst)) {
+        return gcbToHst[gcb];
+    } else {
+        return U_HST_NOT_APPLICABLE;
+    }
+}
+
+#if UCONFIG_NO_NORMALIZATION
+static int32_t getNormQuickCheck(const IntProperty &, UChar32, UProperty) {
+    return 0;
+}
+#else
+static int32_t getNormQuickCheck(const IntProperty &/*prop*/, UChar32 c, UProperty which) {
+    return (int32_t)unorm_getQuickCheck(c, (UNormalizationMode)(which-UCHAR_NFD_QUICK_CHECK+UNORM_NFD));
+}
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+static int32_t getLeadCombiningClass(const IntProperty &, UChar32, UProperty) {
+    return 0;
+}
+#else
+static int32_t getLeadCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return unorm_getFCD16(c)>>8;
+}
+#endif
+
+#if UCONFIG_NO_NORMALIZATION
+static int32_t getTrailCombiningClass(const IntProperty &, UChar32, UProperty) {
+    return 0;
+}
+#else
+static int32_t getTrailCombiningClass(const IntProperty &/*prop*/, UChar32 c, UProperty /*which*/) {
+    return unorm_getFCD16(c)&0xff;
+}
+#endif
+
+static int32_t getInPC(const IntProperty &, UChar32 c, UProperty) {
+    return ulayout_ensureData() && gInpcTrie != nullptr ? ucptrie_get(gInpcTrie, c) : 0;
+}
+
+static int32_t getInSC(const IntProperty &, UChar32 c, UProperty) {
+    return ulayout_ensureData() && gInscTrie != nullptr ? ucptrie_get(gInscTrie, c) : 0;
+}
+
+static int32_t getVo(const IntProperty &, UChar32 c, UProperty) {
+    return ulayout_ensureData() && gVoTrie != nullptr ? ucptrie_get(gVoTrie, c) : 0;
+}
+
+static int32_t layoutGetMaxValue(const IntProperty &/*prop*/, UProperty which) {
+    if (!ulayout_ensureData()) { return 0; }
+    switch (which) {
+    case UCHAR_INDIC_POSITIONAL_CATEGORY:
+        return gMaxInpcValue;
+    case UCHAR_INDIC_SYLLABIC_CATEGORY:
+        return gMaxInscValue;
+    case UCHAR_VERTICAL_ORIENTATION:
+        return gMaxVoValue;
+    default:
+        return 0;
+    }
+}
+
+static const IntProperty intProps[UCHAR_INT_LIMIT-UCHAR_INT_START]={
+    /*
+     * column, mask and shift values for int-value properties from u_getUnicodeProperties().
+     * Must be in order of corresponding UProperty,
+     * and there must be exactly one entry per int UProperty.
+     *
+     * Properties with mask==0 are handled in code.
+     * For them, column is the UPropertySource value.
+     */
+    { UPROPS_SRC_BIDI,  0, 0,                               getBiDiClass, biDiGetMaxValue },
+    { 0,                UPROPS_BLOCK_MASK, UPROPS_BLOCK_SHIFT, defaultGetValue, defaultGetMaxValue },
+    { UPROPS_SRC_NFC,   0, 0xff,                            getCombiningClass, getMaxValueFromShift },
+    { 2,                UPROPS_DT_MASK, 0,                  defaultGetValue, defaultGetMaxValue },
+    { 0,                UPROPS_EA_MASK, UPROPS_EA_SHIFT,    defaultGetValue, defaultGetMaxValue },
+    { UPROPS_SRC_CHAR,  0, (int32_t)U_CHAR_CATEGORY_COUNT-1,getGeneralCategory, getMaxValueFromShift },
+    { UPROPS_SRC_BIDI,  0, 0,                               getJoiningGroup, biDiGetMaxValue },
+    { UPROPS_SRC_BIDI,  0, 0,                               getJoiningType, biDiGetMaxValue },
+    { 2,                UPROPS_LB_MASK, UPROPS_LB_SHIFT,    defaultGetValue, defaultGetMaxValue },
+    { UPROPS_SRC_CHAR,  0, (int32_t)U_NT_COUNT-1,           getNumericType, getMaxValueFromShift },
+    { UPROPS_SRC_PROPSVEC, 0, 0,                            getScript, scriptGetMaxValue },
+    { UPROPS_SRC_PROPSVEC, 0, (int32_t)U_HST_COUNT-1,       getHangulSyllableType, getMaxValueFromShift },
+    // UCHAR_NFD_QUICK_CHECK: max=1=YES -- never "maybe", only "no" or "yes"
+    { UPROPS_SRC_NFC,   0, (int32_t)UNORM_YES,              getNormQuickCheck, getMaxValueFromShift },
+    // UCHAR_NFKD_QUICK_CHECK: max=1=YES -- never "maybe", only "no" or "yes"
+    { UPROPS_SRC_NFKC,  0, (int32_t)UNORM_YES,              getNormQuickCheck, getMaxValueFromShift },
+    // UCHAR_NFC_QUICK_CHECK: max=2=MAYBE
+    { UPROPS_SRC_NFC,   0, (int32_t)UNORM_MAYBE,            getNormQuickCheck, getMaxValueFromShift },
+    // UCHAR_NFKC_QUICK_CHECK: max=2=MAYBE
+    { UPROPS_SRC_NFKC,  0, (int32_t)UNORM_MAYBE,            getNormQuickCheck, getMaxValueFromShift },
+    { UPROPS_SRC_NFC,   0, 0xff,                            getLeadCombiningClass, getMaxValueFromShift },
+    { UPROPS_SRC_NFC,   0, 0xff,                            getTrailCombiningClass, getMaxValueFromShift },
+    { 2,                UPROPS_GCB_MASK, UPROPS_GCB_SHIFT,  defaultGetValue, defaultGetMaxValue },
+    { 2,                UPROPS_SB_MASK, UPROPS_SB_SHIFT,    defaultGetValue, defaultGetMaxValue },
+    { 2,                UPROPS_WB_MASK, UPROPS_WB_SHIFT,    defaultGetValue, defaultGetMaxValue },
+    { UPROPS_SRC_BIDI,  0, 0,                               getBiDiPairedBracketType, biDiGetMaxValue },
+    { UPROPS_SRC_INPC,  0, 0,                               getInPC, layoutGetMaxValue },
+    { UPROPS_SRC_INSC,  0, 0,                               getInSC, layoutGetMaxValue },
+    { UPROPS_SRC_VO,    0, 0,                               getVo, layoutGetMaxValue },
+};
+
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyValue(UChar32 c, UProperty which) {
+    if(which<UCHAR_INT_START) {
+        if(UCHAR_BINARY_START<=which && which<UCHAR_BINARY_LIMIT) {
+            const BinaryProperty &prop=binProps[which];
+            return prop.contains(prop, c, which);
+        }
+    } else if(which<UCHAR_INT_LIMIT) {
+        const IntProperty &prop=intProps[which-UCHAR_INT_START];
+        return prop.getValue(prop, c, which);
+    } else if(which==UCHAR_GENERAL_CATEGORY_MASK) {
+        return U_MASK(u_charType(c));
+    }
+    return 0;  // undefined
+}
+
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyMinValue(UProperty /*which*/) {
+    return 0; /* all binary/enum/int properties have a minimum value of 0 */
+}
+
+U_CAPI int32_t U_EXPORT2
+u_getIntPropertyMaxValue(UProperty which) {
+    if(which<UCHAR_INT_START) {
+        if(UCHAR_BINARY_START<=which && which<UCHAR_BINARY_LIMIT) {
+            return 1;  // maximum TRUE for all binary properties
+        }
+    } else if(which<UCHAR_INT_LIMIT) {
+        const IntProperty &prop=intProps[which-UCHAR_INT_START];
+        return prop.getMaxValue(prop, which);
+    }
+    return -1;  // undefined
+}
+
+U_CFUNC UPropertySource U_EXPORT2
+uprops_getSource(UProperty which) {
+    if(which<UCHAR_BINARY_START) {
+        return UPROPS_SRC_NONE; /* undefined */
+    } else if(which<UCHAR_BINARY_LIMIT) {
+        const BinaryProperty &prop=binProps[which];
+        if(prop.mask!=0) {
+            return UPROPS_SRC_PROPSVEC;
+        } else {
+            return (UPropertySource)prop.column;
+        }
+    } else if(which<UCHAR_INT_START) {
+        return UPROPS_SRC_NONE; /* undefined */
+    } else if(which<UCHAR_INT_LIMIT) {
+        const IntProperty &prop=intProps[which-UCHAR_INT_START];
+        if(prop.mask!=0) {
+            return UPROPS_SRC_PROPSVEC;
+        } else {
+            return (UPropertySource)prop.column;
+        }
+    } else if(which<UCHAR_STRING_START) {
+        switch(which) {
+        case UCHAR_GENERAL_CATEGORY_MASK:
+        case UCHAR_NUMERIC_VALUE:
+            return UPROPS_SRC_CHAR;
+
+        default:
+            return UPROPS_SRC_NONE;
+        }
+    } else if(which<UCHAR_STRING_LIMIT) {
+        switch(which) {
+        case UCHAR_AGE:
+            return UPROPS_SRC_PROPSVEC;
+
+        case UCHAR_BIDI_MIRRORING_GLYPH:
+            return UPROPS_SRC_BIDI;
+
+        case UCHAR_CASE_FOLDING:
+        case UCHAR_LOWERCASE_MAPPING:
+        case UCHAR_SIMPLE_CASE_FOLDING:
+        case UCHAR_SIMPLE_LOWERCASE_MAPPING:
+        case UCHAR_SIMPLE_TITLECASE_MAPPING:
+        case UCHAR_SIMPLE_UPPERCASE_MAPPING:
+        case UCHAR_TITLECASE_MAPPING:
+        case UCHAR_UPPERCASE_MAPPING:
+            return UPROPS_SRC_CASE;
+
+        case UCHAR_ISO_COMMENT:
+        case UCHAR_NAME:
+        case UCHAR_UNICODE_1_NAME:
+            return UPROPS_SRC_NAMES;
+
+        default:
+            return UPROPS_SRC_NONE;
+        }
+    } else {
+        switch(which) {
+        case UCHAR_SCRIPT_EXTENSIONS:
+            return UPROPS_SRC_PROPSVEC;
+        default:
+            return UPROPS_SRC_NONE; /* undefined */
+        }
+    }
+}
+
+U_CFUNC void U_EXPORT2
+uprops_addPropertyStarts(UPropertySource src, const USetAdder *sa, UErrorCode *pErrorCode) {
+    if (!ulayout_ensureData(*pErrorCode)) { return; }
+    const UCPTrie *trie;
+    switch (src) {
+    case UPROPS_SRC_INPC:
+        trie = gInpcTrie;
+        break;
+    case UPROPS_SRC_INSC:
+        trie = gInscTrie;
+        break;
+    case UPROPS_SRC_VO:
+        trie = gVoTrie;
+        break;
+    default:
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    if (trie == nullptr) {
+        *pErrorCode = U_MISSING_RESOURCE_ERROR;
+        return;
+    }
+
+    // Add the start code point of each same-value range of the trie.
+    UChar32 start = 0, end;
+    while ((end = ucptrie_getRange(trie, start, UCPMAP_RANGE_NORMAL, 0,
+                                   nullptr, nullptr, nullptr)) >= 0) {
+        sa->add(sa->set, start);
+        start = end + 1;
+    }
+}
+
+#if !UCONFIG_NO_NORMALIZATION
+
+U_CAPI int32_t U_EXPORT2
+u_getFC_NFKC_Closure(UChar32 c, UChar *dest, int32_t destCapacity, UErrorCode *pErrorCode) {
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    // Compute the FC_NFKC_Closure on the fly:
+    // We have the API for complete coverage of Unicode properties, although
+    // this value by itself is not useful via API.
+    // (What could be useful is a custom normalization table that combines
+    // case folding and NFKC.)
+    // For the derivation, see Unicode's DerivedNormalizationProps.txt.
+    const Normalizer2 *nfkc=Normalizer2::getNFKCInstance(*pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    // first: b = NFKC(Fold(a))
+    UnicodeString folded1String;
+    const UChar *folded1;
+    int32_t folded1Length=ucase_toFullFolding(c, &folded1, U_FOLD_CASE_DEFAULT);
+    if(folded1Length<0) {
+        const Normalizer2Impl *nfkcImpl=Normalizer2Factory::getImpl(nfkc);
+        if(nfkcImpl->getCompQuickCheck(nfkcImpl->getNorm16(c))!=UNORM_NO) {
+            return u_terminateUChars(dest, destCapacity, 0, pErrorCode);  // c does not change at all under CaseFolding+NFKC
+        }
+        folded1String.setTo(c);
+    } else {
+        if(folded1Length>UCASE_MAX_STRING_LENGTH) {
+            folded1String.setTo(folded1Length);
+        } else {
+            folded1String.setTo(FALSE, folded1, folded1Length);
+        }
+    }
+    UnicodeString kc1=nfkc->normalize(folded1String, *pErrorCode);
+    // second: c = NFKC(Fold(b))
+    UnicodeString folded2String(kc1);
+    UnicodeString kc2=nfkc->normalize(folded2String.foldCase(), *pErrorCode);
+    // if (c != b) add the mapping from a to c
+    if(U_FAILURE(*pErrorCode) || kc1==kc2) {
+        return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
+    } else {
+        return kc2.extract(dest, destCapacity, *pErrorCode);
+    }
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/uprops.h b/thirdparty/icu4c/common/uprops.h
new file mode 100644
index 0000000000..8bf929919f
--- /dev/null
+++ b/thirdparty/icu4c/common/uprops.h
@@ -0,0 +1,504 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uprops.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002feb24
+*   created by: Markus W. Scherer
+*
+*   Constants for mostly non-core Unicode character properties
+*   stored in uprops.icu.
+*/
+
+#ifndef __UPROPS_H__
+#define __UPROPS_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+#include "uset_imp.h"
+#include "udataswp.h"
+
+/* indexes[] entries */
+enum {
+    UPROPS_PROPS32_INDEX,
+    UPROPS_EXCEPTIONS_INDEX,
+    UPROPS_EXCEPTIONS_TOP_INDEX,
+
+    UPROPS_ADDITIONAL_TRIE_INDEX,
+    UPROPS_ADDITIONAL_VECTORS_INDEX,
+    UPROPS_ADDITIONAL_VECTORS_COLUMNS_INDEX,
+
+    UPROPS_SCRIPT_EXTENSIONS_INDEX,
+
+    UPROPS_RESERVED_INDEX_7,
+    UPROPS_RESERVED_INDEX_8,
+
+    /* size of the data file (number of 32-bit units after the header) */
+    UPROPS_DATA_TOP_INDEX,
+
+    /* maximum values for code values in vector word 0 */
+    UPROPS_MAX_VALUES_INDEX=10,
+    /* maximum values for code values in vector word 2 */
+    UPROPS_MAX_VALUES_2_INDEX,
+
+    UPROPS_INDEX_COUNT=16
+};
+
+/* definitions for the main properties words */
+enum {
+    /* general category shift==0                                0 (5 bits) */
+    /* reserved                                                 5 (1 bit) */
+    UPROPS_NUMERIC_TYPE_VALUE_SHIFT=6                       /*  6 (10 bits) */
+};
+
+#define GET_CATEGORY(props) ((props)&0x1f)
+#define CAT_MASK(props) U_MASK(GET_CATEGORY(props))
+
+#define GET_NUMERIC_TYPE_VALUE(props) ((props)>>UPROPS_NUMERIC_TYPE_VALUE_SHIFT)
+
+/* constants for the storage form of numeric types and values */
+enum {
+    /** No numeric value. */
+    UPROPS_NTV_NONE=0,
+    /** Decimal digits: nv=0..9 */
+    UPROPS_NTV_DECIMAL_START=1,
+    /** Other digits: nv=0..9 */
+    UPROPS_NTV_DIGIT_START=11,
+    /** Small integers: nv=0..154 */
+    UPROPS_NTV_NUMERIC_START=21,
+    /** Fractions: ((ntv>>4)-12) / ((ntv&0xf)+1) = -1..17 / 1..16 */
+    UPROPS_NTV_FRACTION_START=0xb0,
+    /**
+     * Large integers:
+     * ((ntv>>5)-14) * 10^((ntv&0x1f)+2) = (1..9)*(10^2..10^33)
+     * (only one significant decimal digit)
+     */
+    UPROPS_NTV_LARGE_START=0x1e0,
+    /**
+     * Sexagesimal numbers:
+     * ((ntv>>2)-0xbf) * 60^((ntv&3)+1) = (1..9)*(60^1..60^4)
+     */
+    UPROPS_NTV_BASE60_START=0x300,
+    /**
+     * Fraction-20 values:
+     * frac20 = ntv-0x324 = 0..0x17 -> 1|3|5|7 / 20|40|80|160|320|640
+     * numerator: num = 2*(frac20&3)+1
+     * denominator: den = 20<<(frac20>>2)
+     */
+    UPROPS_NTV_FRACTION20_START=UPROPS_NTV_BASE60_START+36,  // 0x300+9*4=0x324
+    /**
+     * Fraction-32 values:
+     * frac32 = ntv-0x34c = 0..15 -> 1|3|5|7 / 32|64|128|256
+     * numerator: num = 2*(frac32&3)+1
+     * denominator: den = 32<<(frac32>>2)
+     */
+    UPROPS_NTV_FRACTION32_START=UPROPS_NTV_FRACTION20_START+24,  // 0x324+6*4=0x34c
+    /** No numeric value (yet). */
+    UPROPS_NTV_RESERVED_START=UPROPS_NTV_FRACTION32_START+16,  // 0x34c+4*4=0x35c
+
+    UPROPS_NTV_MAX_SMALL_INT=UPROPS_NTV_FRACTION_START-UPROPS_NTV_NUMERIC_START-1
+};
+
+#define UPROPS_NTV_GET_TYPE(ntv) \
+    ((ntv==UPROPS_NTV_NONE) ? U_NT_NONE : \
+    (ntv<UPROPS_NTV_DIGIT_START) ?  U_NT_DECIMAL : \
+    (ntv<UPROPS_NTV_NUMERIC_START) ? U_NT_DIGIT : \
+    U_NT_NUMERIC)
+
+/* number of properties vector words */
+#define UPROPS_VECTOR_WORDS     3
+
+/*
+ * Properties in vector word 0
+ * Bits
+ * 31..24   DerivedAge version major/minor one nibble each
+ * 23..22   3..1: Bits 21..20 & 7..0 = Script_Extensions index
+ *             3: Script value from Script_Extensions
+ *             2: Script=Inherited
+ *             1: Script=Common
+ *             0: Script=bits 21..20 & 7..0
+ * 21..20   Bits 9..8 of the UScriptCode, or index to Script_Extensions
+ * 19..17   East Asian Width
+ * 16.. 8   UBlockCode
+ *  7.. 0   UScriptCode, or index to Script_Extensions
+ */
+
+/* derived age: one nibble each for major and minor version numbers */
+#define UPROPS_AGE_MASK         0xff000000
+#define UPROPS_AGE_SHIFT        24
+
+/* Script_Extensions: mask includes Script */
+#define UPROPS_SCRIPT_X_MASK    0x00f000ff
+#define UPROPS_SCRIPT_X_SHIFT   22
+
+// The UScriptCode or Script_Extensions index is split across two bit fields.
+// (Starting with Unicode 13/ICU 66/2019 due to more varied Script_Extensions.)
+// Shift the high bits right by 12 to assemble the full value.
+#define UPROPS_SCRIPT_HIGH_MASK    0x00300000
+#define UPROPS_SCRIPT_HIGH_SHIFT   12
+#define UPROPS_MAX_SCRIPT          0x3ff
+
+#define UPROPS_EA_MASK          0x000e0000
+#define UPROPS_EA_SHIFT         17
+
+#define UPROPS_BLOCK_MASK       0x0001ff00
+#define UPROPS_BLOCK_SHIFT      8
+
+#define UPROPS_SCRIPT_LOW_MASK  0x000000ff
+
+/* UPROPS_SCRIPT_X_WITH_COMMON must be the lowest value that involves Script_Extensions. */
+#define UPROPS_SCRIPT_X_WITH_COMMON     0x400000
+#define UPROPS_SCRIPT_X_WITH_INHERITED  0x800000
+#define UPROPS_SCRIPT_X_WITH_OTHER      0xc00000
+
+#ifdef __cplusplus
+
+namespace {
+
+inline uint32_t uprops_mergeScriptCodeOrIndex(uint32_t scriptX) {
+    return
+        ((scriptX & UPROPS_SCRIPT_HIGH_MASK) >> UPROPS_SCRIPT_HIGH_SHIFT) |
+        (scriptX & UPROPS_SCRIPT_LOW_MASK);
+}
+
+}  // namespace
+
+#endif  // __cplusplus
+
+/*
+ * Properties in vector word 1
+ * Each bit encodes one binary property.
+ * The following constants represent the bit number, use 1<<UPROPS_XYZ.
+ * UPROPS_BINARY_1_TOP<=32!
+ *
+ * Keep this list of property enums in sync with
+ * propListNames[] in icu/source/tools/genprops/props2.c!
+ *
+ * ICU 2.6/uprops format version 3.2 stores full properties instead of "Other_".
+ */
+enum {
+    UPROPS_WHITE_SPACE,
+    UPROPS_DASH,
+    UPROPS_HYPHEN,
+    UPROPS_QUOTATION_MARK,
+    UPROPS_TERMINAL_PUNCTUATION,
+    UPROPS_MATH,
+    UPROPS_HEX_DIGIT,
+    UPROPS_ASCII_HEX_DIGIT,
+    UPROPS_ALPHABETIC,
+    UPROPS_IDEOGRAPHIC,
+    UPROPS_DIACRITIC,
+    UPROPS_EXTENDER,
+    UPROPS_NONCHARACTER_CODE_POINT,
+    UPROPS_GRAPHEME_EXTEND,
+    UPROPS_GRAPHEME_LINK,
+    UPROPS_IDS_BINARY_OPERATOR,
+    UPROPS_IDS_TRINARY_OPERATOR,
+    UPROPS_RADICAL,
+    UPROPS_UNIFIED_IDEOGRAPH,
+    UPROPS_DEFAULT_IGNORABLE_CODE_POINT,
+    UPROPS_DEPRECATED,
+    UPROPS_LOGICAL_ORDER_EXCEPTION,
+    UPROPS_XID_START,
+    UPROPS_XID_CONTINUE,
+    UPROPS_ID_START,                            /* ICU 2.6, uprops format version 3.2 */
+    UPROPS_ID_CONTINUE,
+    UPROPS_GRAPHEME_BASE,
+    UPROPS_S_TERM,                              /* new in ICU 3.0 and Unicode 4.0.1 */
+    UPROPS_VARIATION_SELECTOR,
+    UPROPS_PATTERN_SYNTAX,                      /* new in ICU 3.4 and Unicode 4.1 */
+    UPROPS_PATTERN_WHITE_SPACE,
+    UPROPS_PREPENDED_CONCATENATION_MARK,        // new in ICU 60 and Unicode 10
+    UPROPS_BINARY_1_TOP                         /* ==32 - full! */
+};
+
+/*
+ * Properties in vector word 2
+ * Bits
+ * 31..26   http://www.unicode.org/reports/tr51/#Emoji_Properties
+ * 25..20   Line Break
+ * 19..15   Sentence Break
+ * 14..10   Word Break
+ *  9.. 5   Grapheme Cluster Break
+ *  4.. 0   Decomposition Type
+ */
+enum {
+    UPROPS_2_EXTENDED_PICTOGRAPHIC=26,
+    UPROPS_2_EMOJI_COMPONENT,
+    UPROPS_2_EMOJI,
+    UPROPS_2_EMOJI_PRESENTATION,
+    UPROPS_2_EMOJI_MODIFIER,
+    UPROPS_2_EMOJI_MODIFIER_BASE
+};
+
+#define UPROPS_LB_MASK          0x03f00000
+#define UPROPS_LB_SHIFT         20
+
+#define UPROPS_SB_MASK          0x000f8000
+#define UPROPS_SB_SHIFT         15
+
+#define UPROPS_WB_MASK          0x00007c00
+#define UPROPS_WB_SHIFT         10
+
+#define UPROPS_GCB_MASK         0x000003e0
+#define UPROPS_GCB_SHIFT        5
+
+#define UPROPS_DT_MASK          0x0000001f
+
+/**
+ * Gets the main properties value for a code point.
+ * Implemented in uchar.c for uprops.cpp.
+ */
+U_CFUNC uint32_t
+u_getMainProperties(UChar32 c);
+
+/**
+ * Get a properties vector word for a code point.
+ * Implemented in uchar.c for uprops.cpp.
+ * @return 0 if no data or illegal argument
+ */
+U_CFUNC uint32_t
+u_getUnicodeProperties(UChar32 c, int32_t column);
+
+/**
+ * Get the the maximum values for some enum/int properties.
+ * Use the same column numbers as for u_getUnicodeProperties().
+ * The returned value will contain maximum values stored in the same bit fields
+ * as where the enum values are stored in the u_getUnicodeProperties()
+ * return values for the same columns.
+ *
+ * Valid columns are those for properties words that contain enumerated values.
+ * (ICU 2.6: columns 0 and 2)
+ * For other column numbers, this function will return 0.
+ *
+ * @internal
+ */
+U_CFUNC int32_t
+uprv_getMaxValues(int32_t column);
+
+/**
+ * Checks if c is alphabetic, or a decimal digit; implements UCHAR_POSIX_ALNUM.
+ * @internal
+ */
+U_CFUNC UBool
+u_isalnumPOSIX(UChar32 c);
+
+/**
+ * Checks if c is in
+ * [^\p{space}\p{gc=Control}\p{gc=Surrogate}\p{gc=Unassigned}]
+ * with space=\p{Whitespace} and Control=Cc.
+ * Implements UCHAR_POSIX_GRAPH.
+ * @internal
+ */
+U_CFUNC UBool
+u_isgraphPOSIX(UChar32 c);
+
+/**
+ * Checks if c is in \p{graph}\p{blank} - \p{cntrl}.
+ * Implements UCHAR_POSIX_PRINT.
+ * @internal
+ */
+U_CFUNC UBool
+u_isprintPOSIX(UChar32 c);
+
+/** Turn a bit index into a bit flag. @internal */
+#define FLAG(n) ((uint32_t)1<<(n))
+
+/** Flags for general categories in the order of UCharCategory. @internal */
+#define _Cn     FLAG(U_GENERAL_OTHER_TYPES)
+#define _Lu     FLAG(U_UPPERCASE_LETTER)
+#define _Ll     FLAG(U_LOWERCASE_LETTER)
+#define _Lt     FLAG(U_TITLECASE_LETTER)
+#define _Lm     FLAG(U_MODIFIER_LETTER)
+/* #define _Lo     FLAG(U_OTHER_LETTER) -- conflicts with MS Visual Studio 9.0 xiosbase */
+#define _Mn     FLAG(U_NON_SPACING_MARK)
+#define _Me     FLAG(U_ENCLOSING_MARK)
+#define _Mc     FLAG(U_COMBINING_SPACING_MARK)
+#define _Nd     FLAG(U_DECIMAL_DIGIT_NUMBER)
+#define _Nl     FLAG(U_LETTER_NUMBER)
+#define _No     FLAG(U_OTHER_NUMBER)
+#define _Zs     FLAG(U_SPACE_SEPARATOR)
+#define _Zl     FLAG(U_LINE_SEPARATOR)
+#define _Zp     FLAG(U_PARAGRAPH_SEPARATOR)
+#define _Cc     FLAG(U_CONTROL_CHAR)
+#define _Cf     FLAG(U_FORMAT_CHAR)
+#define _Co     FLAG(U_PRIVATE_USE_CHAR)
+#define _Cs     FLAG(U_SURROGATE)
+#define _Pd     FLAG(U_DASH_PUNCTUATION)
+#define _Ps     FLAG(U_START_PUNCTUATION)
+/* #define _Pe     FLAG(U_END_PUNCTUATION) -- conflicts with MS Visual Studio 9.0 xlocnum */
+/* #define _Pc     FLAG(U_CONNECTOR_PUNCTUATION) -- conflicts with MS Visual Studio 9.0 streambuf */
+#define _Po     FLAG(U_OTHER_PUNCTUATION)
+#define _Sm     FLAG(U_MATH_SYMBOL)
+#define _Sc     FLAG(U_CURRENCY_SYMBOL)
+#define _Sk     FLAG(U_MODIFIER_SYMBOL)
+#define _So     FLAG(U_OTHER_SYMBOL)
+#define _Pi     FLAG(U_INITIAL_PUNCTUATION)
+/* #define _Pf     FLAG(U_FINAL_PUNCTUATION) -- conflicts with MS Visual Studio 9.0 streambuf */
+
+/** Some code points. @internal */
+enum {
+    TAB     =0x0009,
+    LF      =0x000a,
+    FF      =0x000c,
+    CR      =0x000d,
+    U_A     =0x0041,
+    U_F     =0x0046,
+    U_Z     =0x005a,
+    U_a     =0x0061,
+    U_f     =0x0066,
+    U_z     =0x007a,
+    DEL     =0x007f,
+    NL      =0x0085,
+    NBSP    =0x00a0,
+    CGJ     =0x034f,
+    FIGURESP=0x2007,
+    HAIRSP  =0x200a,
+    ZWNJ    =0x200c,
+    ZWJ     =0x200d,
+    RLM     =0x200f,
+    NNBSP   =0x202f,
+    WJ      =0x2060,
+    INHSWAP =0x206a,
+    NOMDIG  =0x206f,
+    U_FW_A  =0xff21,
+    U_FW_F  =0xff26,
+    U_FW_Z  =0xff3a,
+    U_FW_a  =0xff41,
+    U_FW_f  =0xff46,
+    U_FW_z  =0xff5a,
+    ZWNBSP  =0xfeff
+};
+
+/**
+ * Get the maximum length of a (regular/1.0/extended) character name.
+ * @return 0 if no character names available.
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_getMaxCharNameLength(void);
+
+/**
+ * Fills set with characters that are used in Unicode character names.
+ * Includes all characters that are used in regular/Unicode 1.0/extended names.
+ * Just empties the set if no character names are available.
+ * @param sa USetAdder to receive characters.
+ */
+U_CAPI void U_EXPORT2
+uprv_getCharNameCharacters(const USetAdder *sa);
+
+/**
+ * Constants for which data and implementation files provide which properties.
+ * Used by UnicodeSet for service-specific property enumeration.
+ * @internal
+ */
+enum UPropertySource {
+    /** No source, not a supported property. */
+    UPROPS_SRC_NONE,
+    /** From uchar.c/uprops.icu main trie */
+    UPROPS_SRC_CHAR,
+    /** From uchar.c/uprops.icu properties vectors trie */
+    UPROPS_SRC_PROPSVEC,
+    /** From unames.c/unames.icu */
+    UPROPS_SRC_NAMES,
+    /** From ucase.c/ucase.icu */
+    UPROPS_SRC_CASE,
+    /** From ubidi_props.c/ubidi.icu */
+    UPROPS_SRC_BIDI,
+    /** From uchar.c/uprops.icu main trie as well as properties vectors trie */
+    UPROPS_SRC_CHAR_AND_PROPSVEC,
+    /** From ucase.c/ucase.icu as well as unorm.cpp/unorm.icu */
+    UPROPS_SRC_CASE_AND_NORM,
+    /** From normalizer2impl.cpp/nfc.nrm */
+    UPROPS_SRC_NFC,
+    /** From normalizer2impl.cpp/nfkc.nrm */
+    UPROPS_SRC_NFKC,
+    /** From normalizer2impl.cpp/nfkc_cf.nrm */
+    UPROPS_SRC_NFKC_CF,
+    /** From normalizer2impl.cpp/nfc.nrm canonical iterator data */
+    UPROPS_SRC_NFC_CANON_ITER,
+    // Text layout properties.
+    UPROPS_SRC_INPC,
+    UPROPS_SRC_INSC,
+    UPROPS_SRC_VO,
+    /** One more than the highest UPropertySource (UPROPS_SRC_) constant. */
+    UPROPS_SRC_COUNT
+};
+typedef enum UPropertySource UPropertySource;
+
+/**
+ * @see UPropertySource
+ * @internal
+ */
+U_CFUNC UPropertySource U_EXPORT2
+uprops_getSource(UProperty which);
+
+/**
+ * Enumerate uprops.icu's main data trie and add the
+ * start of each range of same properties to the set.
+ * @internal
+ */
+U_CFUNC void U_EXPORT2
+uchar_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);
+
+/**
+ * Enumerate uprops.icu's properties vectors trie and add the
+ * start of each range of same properties to the set.
+ * @internal
+ */
+U_CFUNC void U_EXPORT2
+upropsvec_addPropertyStarts(const USetAdder *sa, UErrorCode *pErrorCode);
+
+U_CFUNC void U_EXPORT2
+uprops_addPropertyStarts(UPropertySource src, const USetAdder *sa, UErrorCode *pErrorCode);
+
+/**
+ * Return a set of characters for property enumeration.
+ * For each two consecutive characters (start, limit) in the set,
+ * all of the properties for start..limit-1 are all the same.
+ *
+ * @param sa USetAdder to receive result. Existing contents are lost.
+ * @internal
+ */
+/*U_CFUNC void U_EXPORT2
+uprv_getInclusions(const USetAdder *sa, UErrorCode *pErrorCode);
+*/
+
+/**
+ * Swap the ICU Unicode character names file. See uchar.c.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+uchar_swapNames(const UDataSwapper *ds,
+                const void *inData, int32_t length, void *outData,
+                UErrorCode *pErrorCode);
+
+#ifdef __cplusplus
+
+U_NAMESPACE_BEGIN
+
+class UnicodeSet;
+
+class CharacterProperties {
+public:
+    CharacterProperties() = delete;
+    static const UnicodeSet *getInclusionsForProperty(UProperty prop, UErrorCode &errorCode);
+};
+
+// implemented in uniset_props.cpp
+U_CFUNC UnicodeSet *
+uniset_getUnicode32Instance(UErrorCode &errorCode);
+
+U_NAMESPACE_END
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/ures_cnv.cpp b/thirdparty/icu4c/common/ures_cnv.cpp
new file mode 100644
index 0000000000..1aa58e753c
--- /dev/null
+++ b/thirdparty/icu4c/common/ures_cnv.cpp
@@ -0,0 +1,78 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1997-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ures_cnv.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug25
+*   created by: Markus W. Scherer
+*
+*   Character conversion functions moved here from uresbund.c
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/ustring.h"
+#include "unicode/ucnv.h"
+#include "unicode/ures.h"
+#include "uinvchar.h"
+#include "ustr_cnv.h"
+
+U_CAPI UResourceBundle * U_EXPORT2
+ures_openU(const UChar *myPath, 
+           const char *localeID, 
+           UErrorCode *status)
+{
+    char pathBuffer[1024];
+    int32_t length;
+    char *path = pathBuffer;
+
+    if(status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(myPath==NULL) {
+        path = NULL;
+    }
+    else {
+        length=u_strlen(myPath);
+        if(length>=(int32_t)sizeof(pathBuffer)) {
+            *status=U_ILLEGAL_ARGUMENT_ERROR;
+            return NULL;
+        } else if(uprv_isInvariantUString(myPath, length)) {
+            /*
+             * the invariant converter is sufficient for package and tree names
+             * and is more efficient
+             */
+            u_UCharsToChars(myPath, path, length+1); /* length+1 to include the NUL */
+        } else {
+#if !UCONFIG_NO_CONVERSION
+            /* use the default converter to support variant-character paths */
+            UConverter *cnv=u_getDefaultConverter(status);
+            length=ucnv_fromUChars(cnv, path, (int32_t)sizeof(pathBuffer), myPath, length, status);
+            u_releaseDefaultConverter(cnv);
+            if(U_FAILURE(*status)) {
+                return NULL;
+            }
+            if(length>=(int32_t)sizeof(pathBuffer)) {
+                /* not NUL-terminated - path too long */
+                *status=U_ILLEGAL_ARGUMENT_ERROR;
+                return NULL;
+            }
+#else
+            /* the default converter is not available */
+            *status=U_UNSUPPORTED_ERROR;
+            return NULL;
+#endif
+        }
+    }
+
+    return ures_open(path, localeID, status);
+}
diff --git a/thirdparty/icu4c/common/uresbund.cpp b/thirdparty/icu4c/common/uresbund.cpp
new file mode 100644
index 0000000000..2ece87897d
--- /dev/null
+++ b/thirdparty/icu4c/common/uresbund.cpp
@@ -0,0 +1,3090 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1997-2016, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*
+* File uresbund.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   04/01/97    aliu        Creation.
+*   06/14/99    stephen     Removed functions taking a filename suffix.
+*   07/20/99    stephen     Changed for UResourceBundle typedef'd to void*
+*   11/09/99    weiv            Added ures_getLocale()
+*   March 2000  weiv        Total overhaul - using data in DLLs
+*   06/20/2000  helena      OS/400 port changes; mostly typecast.
+*   06/24/02    weiv        Added support for resource sharing
+******************************************************************************
+*/
+
+#include "unicode/ures.h"
+#include "unicode/ustring.h"
+#include "unicode/ucnv.h"
+#include "charstr.h"
+#include "uresimp.h"
+#include "ustr_imp.h"
+#include "cwchar.h"
+#include "ucln_cmn.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "mutex.h"
+#include "uhash.h"
+#include "unicode/uenum.h"
+#include "uenumimp.h"
+#include "ulocimp.h"
+#include "umutex.h"
+#include "putilimp.h"
+#include "uassert.h"
+#include "uresdata.h"
+
+using namespace icu;
+
+/*
+Static cache for already opened resource bundles - mostly for keeping fallback info
+TODO: This cache should probably be removed when the deprecated code is
+      completely removed.
+*/
+static UHashtable *cache = NULL;
+static icu::UInitOnce gCacheInitOnce = U_INITONCE_INITIALIZER;
+
+static UMutex resbMutex;
+
+/* INTERNAL: hashes an entry  */
+static int32_t U_CALLCONV hashEntry(const UHashTok parm) {
+    UResourceDataEntry *b = (UResourceDataEntry *)parm.pointer;
+    UHashTok namekey, pathkey;
+    namekey.pointer = b->fName;
+    pathkey.pointer = b->fPath;
+    return uhash_hashChars(namekey)+37u*uhash_hashChars(pathkey);
+}
+
+/* INTERNAL: compares two entries */
+static UBool U_CALLCONV compareEntries(const UHashTok p1, const UHashTok p2) {
+    UResourceDataEntry *b1 = (UResourceDataEntry *)p1.pointer;
+    UResourceDataEntry *b2 = (UResourceDataEntry *)p2.pointer;
+    UHashTok name1, name2, path1, path2;
+    name1.pointer = b1->fName;
+    name2.pointer = b2->fName;
+    path1.pointer = b1->fPath;
+    path2.pointer = b2->fPath;
+    return (UBool)(uhash_compareChars(name1, name2) &&
+        uhash_compareChars(path1, path2));
+}
+
+
+/**
+ *  Internal function, gets parts of locale name according 
+ *  to the position of '_' character
+ */
+static UBool chopLocale(char *name) {
+    char *i = uprv_strrchr(name, '_');
+
+    if(i != NULL) {
+        *i = '\0';
+        return TRUE;
+    }
+
+    return FALSE;
+}
+
+/**
+ *  Internal function
+ */
+static void entryIncrease(UResourceDataEntry *entry) {
+    Mutex lock(&resbMutex);
+    entry->fCountExisting++;
+    while(entry->fParent != NULL) {
+      entry = entry->fParent;
+      entry->fCountExisting++;
+    }
+}
+
+/**
+ *  Internal function. Tries to find a resource in given Resource 
+ *  Bundle, as well as in its parents
+ */
+static const ResourceData *getFallbackData(const UResourceBundle* resBundle, const char* * resTag, UResourceDataEntry* *realData, Resource *res, UErrorCode *status) {
+    UResourceDataEntry *resB = resBundle->fData;
+    int32_t indexR = -1;
+    int32_t i = 0;
+    *res = RES_BOGUS;
+    if(resB != NULL) {
+        if(resB->fBogus == U_ZERO_ERROR) { /* if this resource is real, */
+            *res = res_getTableItemByKey(&(resB->fData), resB->fData.rootRes, &indexR, resTag); /* try to get data from there */
+            i++;
+        }
+        if(resBundle->fHasFallback == TRUE) {
+            while(*res == RES_BOGUS && resB->fParent != NULL) { /* Otherwise, we'll look in parents */
+                resB = resB->fParent;
+                if(resB->fBogus == U_ZERO_ERROR) {
+                    i++;
+                    *res = res_getTableItemByKey(&(resB->fData), resB->fData.rootRes, &indexR, resTag);
+                }
+            }
+        }
+
+        if(*res != RES_BOGUS) { /* If the resource is found in parents, we need to adjust the error */
+            if(i>1) {
+                if(uprv_strcmp(resB->fName, uloc_getDefault())==0 || uprv_strcmp(resB->fName, kRootLocaleName)==0) {
+                    *status = U_USING_DEFAULT_WARNING;
+                } else {
+                    *status = U_USING_FALLBACK_WARNING;
+                }
+            }
+            *realData = resB;
+            return (&(resB->fData));
+        } else { /* If resource is not found, we need to give an error */
+            *status = U_MISSING_RESOURCE_ERROR;
+            return NULL;
+        }
+    } else {
+            *status = U_MISSING_RESOURCE_ERROR;
+            return NULL;
+    }
+}
+
+static void
+free_entry(UResourceDataEntry *entry) {
+    UResourceDataEntry *alias;
+    res_unload(&(entry->fData));
+    if(entry->fName != NULL && entry->fName != entry->fNameBuffer) {
+        uprv_free(entry->fName);
+    }
+    if(entry->fPath != NULL) {
+        uprv_free(entry->fPath);
+    }
+    if(entry->fPool != NULL) {
+        --entry->fPool->fCountExisting;
+    }
+    alias = entry->fAlias;
+    if(alias != NULL) {
+        while(alias->fAlias != NULL) {
+            alias = alias->fAlias;
+        }
+        --alias->fCountExisting;
+    }
+    uprv_free(entry);
+}
+
+/* Works just like ucnv_flushCache() */
+static int32_t ures_flushCache()
+{
+    UResourceDataEntry *resB;
+    int32_t pos;
+    int32_t rbDeletedNum = 0;
+    const UHashElement *e;
+    UBool deletedMore;
+
+    /*if shared data hasn't even been lazy evaluated yet
+    * return 0
+    */
+    Mutex lock(&resbMutex);
+    if (cache == NULL) {
+        return 0;
+    }
+
+    do {
+        deletedMore = FALSE;
+        /*creates an enumeration to iterate through every element in the table */
+        pos = UHASH_FIRST;
+        while ((e = uhash_nextElement(cache, &pos)) != NULL)
+        {
+            resB = (UResourceDataEntry *) e->value.pointer;
+            /* Deletes only if reference counter == 0
+             * Don't worry about the children of this node.
+             * Those will eventually get deleted too, if not already.
+             * Don't worry about the parents of this node.
+             * Those will eventually get deleted too, if not already.
+             */
+            /* 04/05/2002 [weiv] fCountExisting should now be accurate. If it's not zero, that means that    */
+            /* some resource bundles are still open somewhere. */
+
+            if (resB->fCountExisting == 0) {
+                rbDeletedNum++;
+                deletedMore = TRUE;
+                uhash_removeElement(cache, e);
+                free_entry(resB);
+            }
+        }
+        /*
+         * Do it again to catch bundles (aliases, pool bundle) whose fCountExisting
+         * got decremented by free_entry().
+         */
+    } while(deletedMore);
+
+    return rbDeletedNum;
+}
+
+#ifdef URES_DEBUG
+#include <stdio.h>
+
+U_CAPI UBool U_EXPORT2 ures_dumpCacheContents(void) {
+  UBool cacheNotEmpty = FALSE;
+  int32_t pos = UHASH_FIRST;
+  const UHashElement *e;
+  UResourceDataEntry *resB;
+  
+    Mutex lock(&resbMutex);
+    if (cache == NULL) {
+      fprintf(stderr,"%s:%d: RB Cache is NULL.\n", __FILE__, __LINE__);
+      return FALSE;
+    }
+
+    while ((e = uhash_nextElement(cache, &pos)) != NULL) {
+      cacheNotEmpty=TRUE;
+      resB = (UResourceDataEntry *) e->value.pointer;
+      fprintf(stderr,"%s:%d: RB Cache: Entry @0x%p, refcount %d, name %s:%s.  Pool 0x%p, alias 0x%p, parent 0x%p\n",
+              __FILE__, __LINE__,
+              (void*)resB, resB->fCountExisting,
+              resB->fName?resB->fName:"NULL",
+              resB->fPath?resB->fPath:"NULL",
+              (void*)resB->fPool,
+              (void*)resB->fAlias,
+              (void*)resB->fParent);       
+    }
+    
+    fprintf(stderr,"%s:%d: RB Cache still contains %d items.\n", __FILE__, __LINE__, uhash_count(cache));
+    return cacheNotEmpty;
+}
+
+#endif
+
+static UBool U_CALLCONV ures_cleanup(void)
+{
+    if (cache != NULL) {
+        ures_flushCache();
+        uhash_close(cache);
+        cache = NULL;
+    }
+    gCacheInitOnce.reset();
+    return TRUE;
+}
+
+/** INTERNAL: Initializes the cache for resources */
+static void U_CALLCONV createCache(UErrorCode &status) {
+    U_ASSERT(cache == NULL);
+    cache = uhash_open(hashEntry, compareEntries, NULL, &status);
+    ucln_common_registerCleanup(UCLN_COMMON_URES, ures_cleanup);
+}
+     
+static void initCache(UErrorCode *status) {
+    umtx_initOnce(gCacheInitOnce, &createCache, *status);
+}
+
+/** INTERNAL: sets the name (locale) of the resource bundle to given name */
+
+static void setEntryName(UResourceDataEntry *res, const char *name, UErrorCode *status) {
+    int32_t len = (int32_t)uprv_strlen(name);
+    if(res->fName != NULL && res->fName != res->fNameBuffer) {
+        uprv_free(res->fName);
+    }
+    if (len < (int32_t)sizeof(res->fNameBuffer)) {
+        res->fName = res->fNameBuffer;
+    }
+    else {
+        res->fName = (char *)uprv_malloc(len+1);
+    }
+    if(res->fName == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        uprv_strcpy(res->fName, name);
+    }
+}
+
+static UResourceDataEntry *
+getPoolEntry(const char *path, UErrorCode *status);
+
+/**
+ *  INTERNAL: Inits and opens an entry from a data DLL.
+ *    CAUTION:  resbMutex must be locked when calling this function.
+ */
+static UResourceDataEntry *init_entry(const char *localeID, const char *path, UErrorCode *status) {
+    UResourceDataEntry *r = NULL;
+    UResourceDataEntry find;
+    /*int32_t hashValue;*/
+    const char *name;
+    char aliasName[100] = { 0 };
+    int32_t aliasLen = 0;
+    /*UBool isAlias = FALSE;*/
+    /*UHashTok hashkey; */
+
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    /* here we try to deduce the right locale name */
+    if(localeID == NULL) { /* if localeID is NULL, we're trying to open default locale */
+        name = uloc_getDefault();
+    } else if(*localeID == 0) { /* if localeID is "" then we try to open root locale */
+        name = kRootLocaleName;
+    } else { /* otherwise, we'll open what we're given */
+        name = localeID;
+    }
+
+    find.fName = (char *)name;
+    find.fPath = (char *)path;
+
+    /* calculate the hash value of the entry */
+    /*hashkey.pointer = (void *)&find;*/
+    /*hashValue = hashEntry(hashkey);*/
+
+    /* check to see if we already have this entry */
+    r = (UResourceDataEntry *)uhash_get(cache, &find);
+    if(r == NULL) {
+        /* if the entry is not yet in the hash table, we'll try to construct a new one */
+        r = (UResourceDataEntry *) uprv_malloc(sizeof(UResourceDataEntry));
+        if(r == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+
+        uprv_memset(r, 0, sizeof(UResourceDataEntry));
+        /*r->fHashKey = hashValue;*/
+
+        setEntryName(r, name, status);
+        if (U_FAILURE(*status)) {
+            uprv_free(r);
+            return NULL;
+        }
+
+        if(path != NULL) {
+            r->fPath = (char *)uprv_strdup(path);
+            if(r->fPath == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                uprv_free(r);
+                return NULL;
+            }
+        }
+
+        /* this is the actual loading */
+        res_load(&(r->fData), r->fPath, r->fName, status);
+
+        if (U_FAILURE(*status)) {
+            /* if we failed to load due to an out-of-memory error, exit early. */
+            if (*status == U_MEMORY_ALLOCATION_ERROR) {
+                uprv_free(r);
+                return NULL;
+            }
+            /* we have no such entry in dll, so it will always use fallback */
+            *status = U_USING_FALLBACK_WARNING;
+            r->fBogus = U_USING_FALLBACK_WARNING;
+        } else { /* if we have a regular entry */
+            Resource aliasres;
+            if (r->fData.usesPoolBundle) {
+                r->fPool = getPoolEntry(r->fPath, status);
+                if (U_SUCCESS(*status)) {
+                    const int32_t *poolIndexes = r->fPool->fData.pRoot + 1;
+                    if(r->fData.pRoot[1 + URES_INDEX_POOL_CHECKSUM] == poolIndexes[URES_INDEX_POOL_CHECKSUM]) {
+                        r->fData.poolBundleKeys = (const char *)(poolIndexes + (poolIndexes[URES_INDEX_LENGTH] & 0xff));
+                        r->fData.poolBundleStrings = r->fPool->fData.p16BitUnits;
+                    } else {
+                        r->fBogus = *status = U_INVALID_FORMAT_ERROR;
+                    }
+                } else {
+                    r->fBogus = *status;
+                }
+            }
+            if (U_SUCCESS(*status)) {
+                /* handle the alias by trying to get out the %%Alias tag.*/
+                /* We'll try to get alias string from the bundle */
+                aliasres = res_getResource(&(r->fData), "%%ALIAS");
+                if (aliasres != RES_BOGUS) {
+                    // No tracing: called during initial data loading
+                    const UChar *alias = res_getStringNoTrace(&(r->fData), aliasres, &aliasLen);
+                    if(alias != NULL && aliasLen > 0) { /* if there is actual alias - unload and load new data */
+                        u_UCharsToChars(alias, aliasName, aliasLen+1);
+                        r->fAlias = init_entry(aliasName, path, status);
+                    }
+                }
+            }
+        }
+
+        {
+            UResourceDataEntry *oldR = NULL;
+            if((oldR = (UResourceDataEntry *)uhash_get(cache, r)) == NULL) { /* if the data is not cached */
+                /* just insert it in the cache */
+                UErrorCode cacheStatus = U_ZERO_ERROR;
+                uhash_put(cache, (void *)r, r, &cacheStatus);
+                if (U_FAILURE(cacheStatus)) {
+                    *status = cacheStatus;
+                    free_entry(r);
+                    r = NULL;
+                }
+            } else {
+                /* somebody have already inserted it while we were working, discard newly opened data */
+                /* Also, we could get here IF we opened an alias */
+                free_entry(r);
+                r = oldR;
+            }
+        }
+
+    }
+    if(r != NULL) {
+        /* return the real bundle */
+        while(r->fAlias != NULL) {
+            r = r->fAlias;
+        }
+        r->fCountExisting++; /* we increase its reference count */
+        /* if the resource has a warning */
+        /* we don't want to overwrite a status with no error */
+        if(r->fBogus != U_ZERO_ERROR && U_SUCCESS(*status)) {
+             *status = r->fBogus; /* set the returning status */
+        }
+    }
+    return r;
+}
+
+static UResourceDataEntry *
+getPoolEntry(const char *path, UErrorCode *status) {
+    UResourceDataEntry *poolBundle = init_entry(kPoolBundleName, path, status);
+    if( U_SUCCESS(*status) &&
+        (poolBundle == NULL || poolBundle->fBogus != U_ZERO_ERROR || !poolBundle->fData.isPoolBundle)
+    ) {
+        *status = U_INVALID_FORMAT_ERROR;
+    }
+    return poolBundle;
+}
+
+/* INTERNAL: */
+/*   CAUTION:  resbMutex must be locked when calling this function! */
+static UResourceDataEntry *
+findFirstExisting(const char* path, char* name,
+                  UBool *isRoot, UBool *hasChopped, UBool *isDefault, UErrorCode* status) {
+    UResourceDataEntry *r = NULL;
+    UBool hasRealData = FALSE;
+    const char *defaultLoc = uloc_getDefault();
+    *hasChopped = TRUE; /* we're starting with a fresh name */
+
+    while(*hasChopped && !hasRealData) {
+        r = init_entry(name, path, status);
+        /* Null pointer test */
+        if (U_FAILURE(*status)) {
+            return NULL;
+        }
+        *isDefault = (UBool)(uprv_strncmp(name, defaultLoc, uprv_strlen(name)) == 0);
+        hasRealData = (UBool)(r->fBogus == U_ZERO_ERROR);
+        if(!hasRealData) {
+            /* this entry is not real. We will discard it. */
+            /* However, the parent line for this entry is  */
+            /* not to be used - as there might be parent   */
+            /* lines in cache from previous openings that  */
+            /* are not updated yet. */
+            r->fCountExisting--;
+            /*entryCloseInt(r);*/
+            r = NULL;
+            *status = U_USING_FALLBACK_WARNING;
+        } else {
+            uprv_strcpy(name, r->fName); /* this is needed for supporting aliases */
+        }
+
+        *isRoot = (UBool)(uprv_strcmp(name, kRootLocaleName) == 0);
+
+        /*Fallback data stuff*/
+        *hasChopped = chopLocale(name);
+        if (*hasChopped && *name == '\0') {
+            uprv_strcpy(name, "und");
+        }
+    }
+    return r;
+}
+
+static void ures_setIsStackObject( UResourceBundle* resB, UBool state) {
+    if(state) {
+        resB->fMagic1 = 0;
+        resB->fMagic2 = 0;
+    } else {
+        resB->fMagic1 = MAGIC1;
+        resB->fMagic2 = MAGIC2;
+    }
+}
+
+static UBool ures_isStackObject(const UResourceBundle* resB) {
+  return((resB->fMagic1 == MAGIC1 && resB->fMagic2 == MAGIC2)?FALSE:TRUE);
+}
+
+
+U_CFUNC void ures_initStackObject(UResourceBundle* resB) {
+  uprv_memset(resB, 0, sizeof(UResourceBundle));
+  ures_setIsStackObject(resB, TRUE);
+}
+
+U_NAMESPACE_BEGIN
+
+StackUResourceBundle::StackUResourceBundle() {
+    ures_initStackObject(&bundle);
+}
+
+StackUResourceBundle::~StackUResourceBundle() {
+    ures_close(&bundle);
+}
+
+U_NAMESPACE_END
+
+static UBool  // returns U_SUCCESS(*status)
+loadParentsExceptRoot(UResourceDataEntry *&t1,
+                      char name[], int32_t nameCapacity,
+                      UBool usingUSRData, char usrDataPath[], UErrorCode *status) {
+    if (U_FAILURE(*status)) { return FALSE; }
+    UBool hasChopped = TRUE;
+    while (hasChopped && t1->fParent == NULL && !t1->fData.noFallback &&
+            res_getResource(&t1->fData,"%%ParentIsRoot") == RES_BOGUS) {
+        Resource parentRes = res_getResource(&t1->fData, "%%Parent");
+        if (parentRes != RES_BOGUS) {  // An explicit parent was found.
+            int32_t parentLocaleLen = 0;
+            // No tracing: called during initial data loading
+            const UChar *parentLocaleName = res_getStringNoTrace(&(t1->fData), parentRes, &parentLocaleLen);
+            if(parentLocaleName != NULL && 0 < parentLocaleLen && parentLocaleLen < nameCapacity) {
+                u_UCharsToChars(parentLocaleName, name, parentLocaleLen + 1);
+                if (uprv_strcmp(name, kRootLocaleName) == 0) {
+                    return TRUE;
+                }
+            }
+        }
+        // Insert regular parents.
+        UErrorCode parentStatus = U_ZERO_ERROR;
+        UResourceDataEntry *t2 = init_entry(name, t1->fPath, &parentStatus);
+        if (U_FAILURE(parentStatus)) {
+            *status = parentStatus;
+            return FALSE;
+        }
+        UResourceDataEntry *u2 = NULL;
+        UErrorCode usrStatus = U_ZERO_ERROR;
+        if (usingUSRData) {  // This code inserts user override data into the inheritance chain.
+            u2 = init_entry(name, usrDataPath, &usrStatus);
+            // If we failed due to out-of-memory, report that to the caller and exit early.
+            if (usrStatus == U_MEMORY_ALLOCATION_ERROR) {
+                *status = usrStatus;
+                return FALSE;
+            }
+        }
+
+        if (usingUSRData && U_SUCCESS(usrStatus) && u2->fBogus == U_ZERO_ERROR) {
+            t1->fParent = u2;
+            u2->fParent = t2;
+        } else {
+            t1->fParent = t2;
+            if (usingUSRData) {
+                // The USR override data wasn't found, set it to be deleted.
+                u2->fCountExisting = 0;
+            }
+        }
+        t1 = t2;
+        hasChopped = chopLocale(name);
+    }
+    return TRUE;
+}
+
+static UBool  // returns U_SUCCESS(*status)
+insertRootBundle(UResourceDataEntry *&t1, UErrorCode *status) {
+    if (U_FAILURE(*status)) { return FALSE; }
+    UErrorCode parentStatus = U_ZERO_ERROR;
+    UResourceDataEntry *t2 = init_entry(kRootLocaleName, t1->fPath, &parentStatus);
+    if (U_FAILURE(parentStatus)) {
+        *status = parentStatus;
+        return FALSE;
+    }
+    t1->fParent = t2;
+    t1 = t2;
+    return TRUE;
+}
+
+enum UResOpenType {
+    /**
+     * Open a resource bundle for the locale;
+     * if there is not even a base language bundle, then fall back to the default locale;
+     * if there is no bundle for that either, then load the root bundle.
+     *
+     * This is the default bundle loading behavior.
+     */
+    URES_OPEN_LOCALE_DEFAULT_ROOT,
+    // TODO: ICU ticket #11271 "consistent default locale across locale trees"
+    // Add an option to look at the main locale tree for whether to
+    // fall back to root directly (if the locale has main data) or
+    // fall back to the default locale first (if the locale does not even have main data).
+    /**
+     * Open a resource bundle for the locale;
+     * if there is not even a base language bundle, then load the root bundle;
+     * never fall back to the default locale.
+     *
+     * This is used for algorithms that have good pan-Unicode default behavior,
+     * such as case mappings, collation, and segmentation (BreakIterator).
+     */
+    URES_OPEN_LOCALE_ROOT,
+    /**
+     * Open a resource bundle for the exact bundle name as requested;
+     * no fallbacks, do not load parent bundles.
+     *
+     * This is used for supplemental (non-locale) data.
+     */
+    URES_OPEN_DIRECT
+};
+typedef enum UResOpenType UResOpenType;
+
+static UResourceDataEntry *entryOpen(const char* path, const char* localeID,
+                                     UResOpenType openType, UErrorCode* status) {
+    U_ASSERT(openType != URES_OPEN_DIRECT);
+    UErrorCode intStatus = U_ZERO_ERROR;
+    UResourceDataEntry *r = NULL;
+    UResourceDataEntry *t1 = NULL;
+    UBool isDefault = FALSE;
+    UBool isRoot = FALSE;
+    UBool hasRealData = FALSE;
+    UBool hasChopped = TRUE;
+    UBool usingUSRData = U_USE_USRDATA && ( path == NULL || uprv_strncmp(path,U_ICUDATA_NAME,8) == 0);
+
+    char name[ULOC_FULLNAME_CAPACITY];
+    char usrDataPath[96];
+
+    initCache(status);
+
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    uprv_strncpy(name, localeID, sizeof(name) - 1);
+    name[sizeof(name) - 1] = 0;
+
+    if ( usingUSRData ) {
+        if ( path == NULL ) {
+            uprv_strcpy(usrDataPath, U_USRDATA_NAME);
+        } else {
+            uprv_strncpy(usrDataPath, path, sizeof(usrDataPath) - 1);
+            usrDataPath[0] = 'u';
+            usrDataPath[1] = 's';
+            usrDataPath[2] = 'r';
+            usrDataPath[sizeof(usrDataPath) - 1] = 0;
+        }
+    }
+ 
+    Mutex lock(&resbMutex);    // Lock resbMutex until the end of this function.
+
+    /* We're going to skip all the locales that do not have any data */
+    r = findFirstExisting(path, name, &isRoot, &hasChopped, &isDefault, &intStatus);
+
+    // If we failed due to out-of-memory, report the failure and exit early.
+    if (intStatus == U_MEMORY_ALLOCATION_ERROR) {
+        *status = intStatus;
+        goto finish;
+    }
+
+    if(r != NULL) { /* if there is one real locale, we can look for parents. */
+        t1 = r;
+        hasRealData = TRUE;
+        if ( usingUSRData ) {  /* This code inserts user override data into the inheritance chain */
+            UErrorCode usrStatus = U_ZERO_ERROR;
+            UResourceDataEntry *u1 = init_entry(t1->fName, usrDataPath, &usrStatus);
+            // If we failed due to out-of-memory, report the failure and exit early.
+            if (intStatus == U_MEMORY_ALLOCATION_ERROR) {
+                *status = intStatus;
+                goto finish;
+            }
+            if ( u1 != NULL ) {
+                if(u1->fBogus == U_ZERO_ERROR) {
+                    u1->fParent = t1;
+                    r = u1;
+                } else {
+                    /* the USR override data wasn't found, set it to be deleted */
+                    u1->fCountExisting = 0;
+                }
+            }
+        }
+        if (hasChopped && !isRoot) {
+            if (!loadParentsExceptRoot(t1, name, UPRV_LENGTHOF(name), usingUSRData, usrDataPath, status)) {
+                goto finish;
+            }
+        }
+    }
+
+    /* we could have reached this point without having any real data */
+    /* if that is the case, we need to chain in the default locale   */
+    if(r==NULL && openType == URES_OPEN_LOCALE_DEFAULT_ROOT && !isDefault && !isRoot) {
+        /* insert default locale */
+        uprv_strcpy(name, uloc_getDefault());
+        r = findFirstExisting(path, name, &isRoot, &hasChopped, &isDefault, &intStatus);
+        // If we failed due to out-of-memory, report the failure and exit early.
+        if (intStatus == U_MEMORY_ALLOCATION_ERROR) {
+            *status = intStatus;
+            goto finish;
+        }
+        intStatus = U_USING_DEFAULT_WARNING;
+        if(r != NULL) { /* the default locale exists */
+            t1 = r;
+            hasRealData = TRUE;
+            isDefault = TRUE;
+            // TODO: Why not if (usingUSRData) { ... } like in the non-default-locale code path?
+            if (hasChopped && !isRoot) {
+                if (!loadParentsExceptRoot(t1, name, UPRV_LENGTHOF(name), usingUSRData, usrDataPath, status)) {
+                    goto finish;
+                }
+            }
+        }
+    }
+
+    /* we could still have r == NULL at this point - maybe even default locale is not */
+    /* present */
+    if(r == NULL) {
+        uprv_strcpy(name, kRootLocaleName);
+        r = findFirstExisting(path, name, &isRoot, &hasChopped, &isDefault, &intStatus);
+        // If we failed due to out-of-memory, report the failure and exit early.
+        if (intStatus == U_MEMORY_ALLOCATION_ERROR) {
+            *status = intStatus;
+            goto finish;
+        }
+        if(r != NULL) {
+            t1 = r;
+            intStatus = U_USING_DEFAULT_WARNING;
+            hasRealData = TRUE;
+        } else { /* we don't even have the root locale */
+            *status = U_MISSING_RESOURCE_ERROR;
+            goto finish;
+        }
+    } else if(!isRoot && uprv_strcmp(t1->fName, kRootLocaleName) != 0 &&
+            t1->fParent == NULL && !r->fData.noFallback) {
+        if (!insertRootBundle(t1, status)) {
+            goto finish;
+        }
+        if(!hasRealData) {
+            r->fBogus = U_USING_DEFAULT_WARNING;
+        }
+    }
+
+    // TODO: Does this ever loop?
+    while(r != NULL && !isRoot && t1->fParent != NULL) {
+        t1->fParent->fCountExisting++;
+        t1 = t1->fParent;
+    }
+
+finish:
+    if(U_SUCCESS(*status)) {
+        if(intStatus != U_ZERO_ERROR) {
+            *status = intStatus;  
+        }
+        return r;
+    } else {
+        return NULL;
+    }
+}
+
+/**
+ * Version of entryOpen() and findFirstExisting() for ures_openDirect(),
+ * with no fallbacks.
+ * Parent and root locale bundles are loaded if
+ * the requested bundle does not have the "nofallback" flag.
+ */
+static UResourceDataEntry *
+entryOpenDirect(const char* path, const char* localeID, UErrorCode* status) {
+    initCache(status);
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    Mutex lock(&resbMutex);
+    // findFirstExisting() without fallbacks.
+    UResourceDataEntry *r = init_entry(localeID, path, status);
+    if(U_SUCCESS(*status)) {
+        if(r->fBogus != U_ZERO_ERROR) {
+            r->fCountExisting--;
+            r = NULL;
+        }
+    } else {
+        r = NULL;
+    }
+
+    // Some code depends on the ures_openDirect() bundle to have a parent bundle chain,
+    // unless it is marked with "nofallback".
+    UResourceDataEntry *t1 = r;
+    if(r != NULL && uprv_strcmp(localeID, kRootLocaleName) != 0 &&  // not root
+            r->fParent == NULL && !r->fData.noFallback &&
+            uprv_strlen(localeID) < ULOC_FULLNAME_CAPACITY) {
+        char name[ULOC_FULLNAME_CAPACITY];
+        uprv_strcpy(name, localeID);
+        if(!chopLocale(name) || uprv_strcmp(name, kRootLocaleName) == 0 ||
+                loadParentsExceptRoot(t1, name, UPRV_LENGTHOF(name), FALSE, NULL, status)) {
+            if(uprv_strcmp(t1->fName, kRootLocaleName) != 0 && t1->fParent == NULL) {
+                insertRootBundle(t1, status);
+            }
+        }
+        if(U_FAILURE(*status)) {
+            r = NULL;
+        }
+    }
+
+    if(r != NULL) {
+        // TODO: Does this ever loop?
+        while(t1->fParent != NULL) {
+            t1->fParent->fCountExisting++;
+            t1 = t1->fParent;
+        }
+    }
+    return r;
+}
+
+/**
+ * Functions to create and destroy resource bundles.
+ *     CAUTION:  resbMutex must be locked when calling this function.
+ */
+/* INTERNAL: */
+static void entryCloseInt(UResourceDataEntry *resB) {
+    UResourceDataEntry *p = resB;
+
+    while(resB != NULL) {
+        p = resB->fParent;
+        resB->fCountExisting--;
+
+        /* Entries are left in the cache. TODO: add ures_flushCache() to force a flush
+         of the cache. */
+/*
+        if(resB->fCountExisting <= 0) {
+            uhash_remove(cache, resB);
+            if(resB->fBogus == U_ZERO_ERROR) {
+                res_unload(&(resB->fData));
+            }
+            if(resB->fName != NULL) {
+                uprv_free(resB->fName);
+            }
+            if(resB->fPath != NULL) {
+                uprv_free(resB->fPath);
+            }
+            uprv_free(resB);
+        }
+*/
+
+        resB = p;
+    }
+}
+
+/** 
+ *  API: closes a resource bundle and cleans up.
+ */
+
+static void entryClose(UResourceDataEntry *resB) {
+  Mutex lock(&resbMutex);
+  entryCloseInt(resB);
+}
+
+/*
+U_CFUNC void ures_setResPath(UResourceBundle *resB, const char* toAdd) {
+  if(resB->fResPath == NULL) {
+    resB->fResPath = resB->fResBuf;
+    *(resB->fResPath) = 0;
+  } 
+  resB->fResPathLen = uprv_strlen(toAdd);
+  if(RES_BUFSIZE <= resB->fResPathLen+1) {
+    if(resB->fResPath == resB->fResBuf) {
+      resB->fResPath = (char *)uprv_malloc((resB->fResPathLen+1)*sizeof(char));
+    } else {
+      resB->fResPath = (char *)uprv_realloc(resB->fResPath, (resB->fResPathLen+1)*sizeof(char));
+    }
+  }
+  uprv_strcpy(resB->fResPath, toAdd);
+}
+*/
+static void ures_appendResPath(UResourceBundle *resB, const char* toAdd, int32_t lenToAdd, UErrorCode *status) {
+    int32_t resPathLenOrig = resB->fResPathLen;
+    if(resB->fResPath == NULL) {
+        resB->fResPath = resB->fResBuf;
+        *(resB->fResPath) = 0;
+        resB->fResPathLen = 0;
+    } 
+    resB->fResPathLen += lenToAdd;
+    if(RES_BUFSIZE <= resB->fResPathLen+1) {
+        if(resB->fResPath == resB->fResBuf) {
+            resB->fResPath = (char *)uprv_malloc((resB->fResPathLen+1)*sizeof(char));
+            /* Check that memory was allocated correctly. */
+            if (resB->fResPath == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            uprv_strcpy(resB->fResPath, resB->fResBuf);
+        } else {
+            char *temp = (char *)uprv_realloc(resB->fResPath, (resB->fResPathLen+1)*sizeof(char));
+            /* Check that memory was reallocated correctly. */
+            if (temp == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+            resB->fResPath = temp;
+        }
+    }
+    uprv_strcpy(resB->fResPath + resPathLenOrig, toAdd);
+}
+
+static void ures_freeResPath(UResourceBundle *resB) {
+    if (resB->fResPath && resB->fResPath != resB->fResBuf) {
+        uprv_free(resB->fResPath);
+    }
+    resB->fResPath = NULL;
+    resB->fResPathLen = 0;
+}
+
+static void
+ures_closeBundle(UResourceBundle* resB, UBool freeBundleObj)
+{
+    if(resB != NULL) {
+        if(resB->fData != NULL) {
+            entryClose(resB->fData);
+        }
+        if(resB->fVersion != NULL) {
+            uprv_free(resB->fVersion);
+        }
+        ures_freeResPath(resB);
+
+        if(ures_isStackObject(resB) == FALSE && freeBundleObj) {
+            uprv_free(resB);
+        }
+#if 0 /*U_DEBUG*/
+        else {
+            /* poison the data */
+            uprv_memset(resB, -1, sizeof(UResourceBundle));
+        }
+#endif
+    }
+}
+
+U_CAPI void  U_EXPORT2
+ures_close(UResourceBundle* resB)
+{
+    ures_closeBundle(resB, TRUE);
+}
+
+static UResourceBundle *init_resb_result(const ResourceData *rdata, Resource r, 
+                                         const char *key, int32_t idx, UResourceDataEntry *realData, 
+                                         const UResourceBundle *parent, int32_t noAlias,
+                                         UResourceBundle *resB, UErrorCode *status) 
+{
+    if(status == NULL || U_FAILURE(*status)) {
+        return resB;
+    }
+    if (parent == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    if(RES_GET_TYPE(r) == URES_ALIAS) { /* This is an alias, need to exchange with real data */
+        if(noAlias < URES_MAX_ALIAS_LEVEL) { 
+            int32_t len = 0;
+            const UChar *alias = res_getAlias(rdata, r, &len); 
+            if(len > 0) {
+                /* we have an alias, now let's cut it up */
+                char stackAlias[200];
+                char *chAlias = NULL, *path = NULL, *locale = NULL, *keyPath = NULL;
+                int32_t capacity;
+
+                /*
+                * Allocate enough space for both the char * version
+                * of the alias and parent->fResPath.
+                *
+                * We do this so that res_findResource() can modify the path,
+                * which allows us to remove redundant _res_findResource() variants
+                * in uresdata.c.
+                * res_findResource() now NUL-terminates each segment so that table keys
+                * can always be compared with strcmp() instead of strncmp().
+                * Saves code there and simplifies testing and code coverage.
+                *
+                * markus 2003oct17
+                */
+                ++len; /* count the terminating NUL */
+                if(parent->fResPath != NULL) {
+                    capacity = (int32_t)uprv_strlen(parent->fResPath) + 1;
+                } else {
+                    capacity = 0;
+                }
+                if(capacity < len) {
+                    capacity = len;
+                }
+                if(capacity <= (int32_t)sizeof(stackAlias)) {
+                    capacity = (int32_t)sizeof(stackAlias);
+                    chAlias = stackAlias;
+                } else {
+                    chAlias = (char *)uprv_malloc(capacity);
+                    /* test for NULL */
+                    if(chAlias == NULL) {
+                        *status = U_MEMORY_ALLOCATION_ERROR;
+                        return NULL;
+                    }
+                }
+                u_UCharsToChars(alias, chAlias, len);
+
+                if(*chAlias == RES_PATH_SEPARATOR) {
+                    /* there is a path included */
+                    locale = uprv_strchr(chAlias+1, RES_PATH_SEPARATOR);
+                    if(locale == NULL) {
+                        locale = uprv_strchr(chAlias, 0); /* avoid locale == NULL to make code below work */
+                    } else {
+                        *locale = 0;
+                        locale++;
+                    }
+                    path = chAlias+1;
+                    if(uprv_strcmp(path, "LOCALE") == 0) {
+                        /* this is an XPath alias, starting with "/LOCALE/" */
+                        /* it contains the path to a resource which should be looked up */
+                        /* starting in the requested locale */
+                        keyPath = locale; 
+                        locale = parent->fTopLevelData->fName; /* this is the requested locale's name */
+                        path = realData->fPath; /* we will be looking in the same package */
+                    } else {
+                        if(uprv_strcmp(path, "ICUDATA") == 0) { /* want ICU data */
+                            path = NULL;
+                        }
+                        keyPath = uprv_strchr(locale, RES_PATH_SEPARATOR);
+                        if(keyPath) {
+                            *keyPath = 0;
+                            keyPath++;
+                        }
+                    }
+                } else {
+                    /* no path, start with a locale */
+                    locale = chAlias;
+                    keyPath = uprv_strchr(locale, RES_PATH_SEPARATOR);
+                    if(keyPath) {
+                        *keyPath = 0;
+                        keyPath++;
+                    }
+                    path = realData->fPath;
+                }
+
+
+                {
+                    /* got almost everything, let's try to open */
+                    /* first, open the bundle with real data */
+                    UResourceBundle *result = resB;
+                    const char* temp = NULL;
+                    UErrorCode intStatus = U_ZERO_ERROR;
+                    UResourceBundle *mainRes = ures_openDirect(path, locale, &intStatus); 
+                    if(U_SUCCESS(intStatus)) {
+                        if(keyPath == NULL) {
+                            /* no key path. This means that we are going to 
+                            * to use the corresponding resource from
+                            * another bundle
+                            */
+                            /* first, we are going to get a corresponding parent 
+                            * resource to the one we are searching.
+                            */
+                            char *aKey = parent->fResPath;
+                            if(aKey) {
+                                uprv_strcpy(chAlias, aKey); /* allocated large enough above */
+                                aKey = chAlias;
+                                r = res_findResource(&(mainRes->fResData), mainRes->fRes, &aKey, &temp);
+                            } else {
+                                r = mainRes->fRes;
+                            }
+                            if(key) {
+                                /* we need to make keyPath from parent's fResPath and
+                                * current key, if there is a key associated
+                                */
+                                len = (int32_t)(uprv_strlen(key) + 1);
+                                if(len > capacity) {
+                                    capacity = len;
+                                    if(chAlias == stackAlias) {
+                                        chAlias = (char *)uprv_malloc(capacity);
+                                    } else {
+                                        chAlias = (char *)uprv_realloc(chAlias, capacity);
+                                    }
+                                    if(chAlias == NULL) {
+                                        ures_close(mainRes);
+                                        *status = U_MEMORY_ALLOCATION_ERROR;
+                                        return NULL;
+                                    }
+                                }
+                                uprv_memcpy(chAlias, key, len);
+                                aKey = chAlias;
+                                r = res_findResource(&(mainRes->fResData), r, &aKey, &temp);
+                            } else if(idx != -1) {
+                                /* if there is no key, but there is an index, try to get by the index */
+                                /* here we have either a table or an array, so get the element */
+                                int32_t type = RES_GET_TYPE(r);
+                                if(URES_IS_TABLE(type)) {
+                                    r = res_getTableItemByIndex(&(mainRes->fResData), r, idx, (const char **)&aKey);
+                                } else { /* array */
+                                    r = res_getArrayItem(&(mainRes->fResData), r, idx);
+                                }
+                            }
+                            if(r != RES_BOGUS) {
+                                result = init_resb_result(&(mainRes->fResData), r, temp, -1, mainRes->fData, mainRes, noAlias+1, resB, status);
+                            } else {
+                                *status = U_MISSING_RESOURCE_ERROR;
+                                result = resB;
+                            }
+                        } else {
+                            /* this one is a bit trickier. 
+                            * we start finding keys, but after we resolve one alias, the path might continue.
+                            * Consider: 
+                            *     aliastest:alias { "testtypes/anotheralias/Sequence" }
+                            *     anotheralias:alias { "/ICUDATA/sh/CollationElements" }
+                            * aliastest resource should finally have the sequence, not collation elements.
+                            */
+                            UResourceDataEntry *dataEntry = mainRes->fData;
+                            char stackPath[URES_MAX_BUFFER_SIZE];
+                            char *pathBuf = stackPath, *myPath = pathBuf;
+                            if(uprv_strlen(keyPath) >= UPRV_LENGTHOF(stackPath)) {
+                                pathBuf = (char *)uprv_malloc((uprv_strlen(keyPath)+1)*sizeof(char));
+                                if(pathBuf == NULL) {
+                                    *status = U_MEMORY_ALLOCATION_ERROR;
+                                    ures_close(mainRes);
+                                    return NULL;
+                                }
+                            }
+                            uprv_strcpy(pathBuf, keyPath);
+                            result = mainRes;
+                            /* now we have fallback following here */
+                            do {
+                                r = dataEntry->fData.rootRes;     
+                                /* this loop handles 'found' resources over several levels */
+                                while(*myPath && U_SUCCESS(*status)) {
+                                    r = res_findResource(&(dataEntry->fData), r, &myPath, &temp);
+                                    if(r != RES_BOGUS) { /* found a resource, but it might be an indirection */
+                                        resB = init_resb_result(&(dataEntry->fData), r, temp, -1, dataEntry, result, noAlias+1, resB, status);
+                                        result = resB;
+                                        if(result) {
+                                            r = result->fRes; /* switch to a new resource, possibly a new tree */
+                                            dataEntry = result->fData;
+                                        }
+                                    } else { /* no resource found, we don't really want to look anymore on this level */
+                                        break;
+                                    }
+                                }
+                                dataEntry = dataEntry->fParent;
+                                uprv_strcpy(pathBuf, keyPath);
+                                myPath = pathBuf;
+                            } while(r == RES_BOGUS && dataEntry != NULL);
+                            if(r == RES_BOGUS) {
+                                *status = U_MISSING_RESOURCE_ERROR;
+                                result = resB;
+                            }
+                            if(pathBuf != stackPath) {
+                                uprv_free(pathBuf);
+                            }
+                        }
+                    } else { /* we failed to open the resource we're aliasing to */
+                        *status = intStatus;
+                    }
+                    if(chAlias != stackAlias) {
+                        uprv_free(chAlias);
+                    }
+                    if(mainRes != result) {
+                        ures_close(mainRes);
+                    }
+                    ResourceTracer(resB).maybeTrace("getalias");
+                    return result;
+                }
+            } else {
+                /* bad alias, should be an error */ 
+                *status = U_ILLEGAL_ARGUMENT_ERROR;
+                return resB;
+            }
+        } else {
+            *status = U_TOO_MANY_ALIASES_ERROR;
+            return resB;
+        }
+    }
+    if(resB == NULL) {
+        resB = (UResourceBundle *)uprv_malloc(sizeof(UResourceBundle));
+        /* test for NULL */
+        if (resB == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+        ures_setIsStackObject(resB, FALSE);
+        resB->fResPath = NULL;
+        resB->fResPathLen = 0;
+    } else {
+        if(resB->fData != NULL) {
+            entryClose(resB->fData);
+        }
+        if(resB->fVersion != NULL) {
+            uprv_free(resB->fVersion);
+        }
+        /* 
+        weiv: if stack object was passed in, it doesn't really need to be reinited,
+        since the purpose of initing is to remove stack junk. However, at this point 
+        we would not do anything to an allocated object, so stack object should be
+        treated the same
+        */
+        /*
+        if(ures_isStackObject(resB) != FALSE) {
+        ures_initStackObject(resB);
+        }
+        */
+        if(parent != resB) {
+            ures_freeResPath(resB);
+        }
+    }
+    resB->fData = realData;
+    entryIncrease(resB->fData);
+    resB->fHasFallback = FALSE;
+    resB->fIsTopLevel = FALSE;
+    resB->fIndex = -1;
+    resB->fKey = key; 
+    /*resB->fParentRes = parent;*/
+    resB->fTopLevelData = parent->fTopLevelData;
+    if(parent->fResPath && parent != resB) {
+        ures_appendResPath(resB, parent->fResPath, parent->fResPathLen, status);
+    }
+    if(key != NULL) {
+        ures_appendResPath(resB, key, (int32_t)uprv_strlen(key), status);
+        if(resB->fResPath[resB->fResPathLen-1] != RES_PATH_SEPARATOR) {
+            ures_appendResPath(resB, RES_PATH_SEPARATOR_S, 1, status);
+        }
+    } else if(idx >= 0) {
+        char buf[256];
+        int32_t len = T_CString_integerToString(buf, idx, 10);
+        ures_appendResPath(resB, buf, len, status);
+        if(resB->fResPath[resB->fResPathLen-1] != RES_PATH_SEPARATOR) {
+            ures_appendResPath(resB, RES_PATH_SEPARATOR_S, 1, status);
+        }
+    }
+    /* Make sure that Purify doesn't complain about uninitialized memory copies. */
+    {
+        int32_t usedLen = ((resB->fResBuf == resB->fResPath) ? resB->fResPathLen : 0);
+        uprv_memset(resB->fResBuf + usedLen, 0, sizeof(resB->fResBuf) - usedLen);
+    }
+
+    resB->fVersion = NULL;
+    resB->fRes = r;
+    /*resB->fParent = parent->fRes;*/
+    uprv_memmove(&resB->fResData, rdata, sizeof(ResourceData));
+    resB->fSize = res_countArrayItems(&(resB->fResData), resB->fRes);
+    ResourceTracer(resB).trace("get");
+    return resB;
+}
+
+UResourceBundle *ures_copyResb(UResourceBundle *r, const UResourceBundle *original, UErrorCode *status) {
+    UBool isStackObject;
+    if(U_FAILURE(*status) || r == original) {
+        return r;
+    }
+    if(original != NULL) {
+        if(r == NULL) {
+            isStackObject = FALSE;
+            r = (UResourceBundle *)uprv_malloc(sizeof(UResourceBundle));
+            /* test for NULL */
+            if (r == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+                return NULL;
+            }
+        } else {
+            isStackObject = ures_isStackObject(r);
+            ures_closeBundle(r, FALSE);
+        }
+        uprv_memcpy(r, original, sizeof(UResourceBundle));
+        r->fResPath = NULL;
+        r->fResPathLen = 0;
+        if(original->fResPath) {
+            ures_appendResPath(r, original->fResPath, original->fResPathLen, status);
+        }
+        ures_setIsStackObject(r, isStackObject);
+        if(r->fData != NULL) {
+            entryIncrease(r->fData);
+        }
+    }
+    return r;
+}
+
+/**
+ * Functions to retrieve data from resource bundles.
+ */
+
+U_CAPI const UChar* U_EXPORT2 ures_getString(const UResourceBundle* resB, int32_t* len, UErrorCode* status) {
+    const UChar *s;
+    if (status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    s = res_getString({resB}, &(resB->fResData), resB->fRes, len);
+    if (s == NULL) {
+        *status = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return s;
+}
+
+static const char *
+ures_toUTF8String(const UChar *s16, int32_t length16,
+                  char *dest, int32_t *pLength,
+                  UBool forceCopy,
+                  UErrorCode *status) {
+    int32_t capacity;
+
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (pLength != NULL) {
+        capacity = *pLength;
+    } else {
+        capacity = 0;
+    }
+    if (capacity < 0 || (capacity > 0 && dest == NULL)) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if (length16 == 0) {
+        /* empty string, return as read-only pointer */
+        if (pLength != NULL) {
+            *pLength = 0;
+        }
+        if (forceCopy) {
+            u_terminateChars(dest, capacity, 0, status);
+            return dest;
+        } else {
+            return "";
+        }
+    } else {
+        /* We need to transform the string to the destination buffer. */
+        if (capacity < length16) {
+            /* No chance for the string to fit. Pure preflighting. */
+            return u_strToUTF8(NULL, 0, pLength, s16, length16, status);
+        }
+        if (!forceCopy && (length16 <= 0x2aaaaaaa)) {
+            /*
+             * We know the string will fit into dest because each UChar turns
+             * into at most three UTF-8 bytes. Fill the latter part of dest
+             * so that callers do not expect to use dest as a string pointer,
+             * hopefully leading to more robust code for when resource bundles
+             * may store UTF-8 natively.
+             * (In which case dest would not be used at all.)
+             *
+             * We do not do this if forceCopy=TRUE because then the caller
+             * expects the string to start exactly at dest.
+             *
+             * The test above for <= 0x2aaaaaaa prevents overflows.
+             * The +1 is for the NUL terminator.
+             */
+            int32_t maxLength = 3 * length16 + 1;
+            if (capacity > maxLength) {
+                dest += capacity - maxLength;
+                capacity = maxLength;
+            }
+        }
+        return u_strToUTF8(dest, capacity, pLength, s16, length16, status);
+    }
+}
+
+U_CAPI const char * U_EXPORT2
+ures_getUTF8String(const UResourceBundle *resB,
+                   char *dest, int32_t *pLength,
+                   UBool forceCopy,
+                   UErrorCode *status) {
+    int32_t length16;
+    const UChar *s16 = ures_getString(resB, &length16, status);
+    return ures_toUTF8String(s16, length16, dest, pLength, forceCopy, status);
+}
+
+U_CAPI const uint8_t* U_EXPORT2 ures_getBinary(const UResourceBundle* resB, int32_t* len, 
+                                               UErrorCode*               status) {
+  const uint8_t *p;
+  if (status==NULL || U_FAILURE(*status)) {
+    return NULL;
+  }
+  if(resB == NULL) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+  p = res_getBinary({resB}, &(resB->fResData), resB->fRes, len);
+  if (p == NULL) {
+    *status = U_RESOURCE_TYPE_MISMATCH;
+  }
+  return p;
+}
+
+U_CAPI const int32_t* U_EXPORT2 ures_getIntVector(const UResourceBundle* resB, int32_t* len, 
+                                                   UErrorCode*               status) {
+  const int32_t *p;
+  if (status==NULL || U_FAILURE(*status)) {
+    return NULL;
+  }
+  if(resB == NULL) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+  p = res_getIntVector({resB}, &(resB->fResData), resB->fRes, len);
+  if (p == NULL) {
+    *status = U_RESOURCE_TYPE_MISMATCH;
+  }
+  return p;
+}
+
+/* this function returns a signed integer */ 
+/* it performs sign extension */
+U_CAPI int32_t U_EXPORT2 ures_getInt(const UResourceBundle* resB, UErrorCode *status) {
+  if (status==NULL || U_FAILURE(*status)) {
+    return 0xffffffff;
+  }
+  if(resB == NULL) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0xffffffff;
+  }
+  if(RES_GET_TYPE(resB->fRes) != URES_INT) {
+    *status = U_RESOURCE_TYPE_MISMATCH;
+    return 0xffffffff;
+  }
+  return res_getInt({resB}, resB->fRes);
+}
+
+U_CAPI uint32_t U_EXPORT2 ures_getUInt(const UResourceBundle* resB, UErrorCode *status) {
+  if (status==NULL || U_FAILURE(*status)) {
+    return 0xffffffff;
+  }
+  if(resB == NULL) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return 0xffffffff;
+  }
+  if(RES_GET_TYPE(resB->fRes) != URES_INT) {
+    *status = U_RESOURCE_TYPE_MISMATCH;
+    return 0xffffffff;
+  }
+  return res_getUInt({resB}, resB->fRes);
+}
+
+U_CAPI UResType U_EXPORT2 ures_getType(const UResourceBundle *resB) {
+  if(resB == NULL) {
+    return URES_NONE;
+  }
+  return res_getPublicType(resB->fRes);
+}
+
+U_CAPI const char * U_EXPORT2 ures_getKey(const UResourceBundle *resB) {
+  //
+  // TODO: Trace ures_getKey? I guess not usually.
+  //
+  // We usually get the key string to decide whether we want the value, or to
+  // make a key-value pair. Tracing the value should suffice.
+  //
+  // However, I believe we have some data (e.g., in res_index) where the key
+  // strings are the data. Tracing the enclosing table should suffice.
+  //
+  if(resB == NULL) {
+    return NULL;
+  }
+  return(resB->fKey);
+}
+
+U_CAPI int32_t U_EXPORT2 ures_getSize(const UResourceBundle *resB) {
+  if(resB == NULL) {
+    return 0;
+  }
+  
+  return resB->fSize;
+}
+
+static const UChar* ures_getStringWithAlias(const UResourceBundle *resB, Resource r, int32_t sIndex, int32_t *len, UErrorCode *status) {
+  if(RES_GET_TYPE(r) == URES_ALIAS) {
+    const UChar* result = 0;
+    UResourceBundle *tempRes = ures_getByIndex(resB, sIndex, NULL, status);
+    result = ures_getString(tempRes, len, status);
+    ures_close(tempRes);
+    return result;
+  } else {
+    return res_getString({resB, sIndex}, &(resB->fResData), r, len); 
+  }
+}
+
+U_CAPI void U_EXPORT2 ures_resetIterator(UResourceBundle *resB){
+  if(resB == NULL) {
+    return;
+  }
+  resB->fIndex = -1;
+}
+
+U_CAPI UBool U_EXPORT2 ures_hasNext(const UResourceBundle *resB) {
+  if(resB == NULL) {
+    return FALSE;
+  }
+  return (UBool)(resB->fIndex < resB->fSize-1);
+}
+
+U_CAPI const UChar* U_EXPORT2 ures_getNextString(UResourceBundle *resB, int32_t* len, const char ** key, UErrorCode *status) {
+  Resource r = RES_BOGUS;
+  
+  if (status==NULL || U_FAILURE(*status)) {
+    return NULL;
+  }
+  if(resB == NULL) {
+    *status = U_ILLEGAL_ARGUMENT_ERROR;
+    return NULL;
+  }
+  
+  if(resB->fIndex == resB->fSize-1) {
+    *status = U_INDEX_OUTOFBOUNDS_ERROR;
+  } else {
+    resB->fIndex++;
+    switch(RES_GET_TYPE(resB->fRes)) {
+    case URES_STRING:
+    case URES_STRING_V2:
+      return res_getString({resB}, &(resB->fResData), resB->fRes, len);
+    case URES_TABLE:
+    case URES_TABLE16:
+    case URES_TABLE32:
+      r = res_getTableItemByIndex(&(resB->fResData), resB->fRes, resB->fIndex, key);
+      if(r == RES_BOGUS && resB->fHasFallback) {
+        /* TODO: do the fallback */
+      }
+      return ures_getStringWithAlias(resB, r, resB->fIndex, len, status);
+    case URES_ARRAY:
+    case URES_ARRAY16:
+      r = res_getArrayItem(&(resB->fResData), resB->fRes, resB->fIndex);
+      if(r == RES_BOGUS && resB->fHasFallback) {
+        /* TODO: do the fallback */
+      }
+      return ures_getStringWithAlias(resB, r, resB->fIndex, len, status);
+    case URES_ALIAS:
+      return ures_getStringWithAlias(resB, resB->fRes, resB->fIndex, len, status);
+    case URES_INT:
+    case URES_BINARY:
+    case URES_INT_VECTOR:
+        *status = U_RESOURCE_TYPE_MISMATCH;
+        U_FALLTHROUGH;
+    default:
+      return NULL;
+    }
+  }
+
+  return NULL;
+}
+
+U_CAPI UResourceBundle* U_EXPORT2 ures_getNextResource(UResourceBundle *resB, UResourceBundle *fillIn, UErrorCode *status) {
+    const char *key = NULL;
+    Resource r = RES_BOGUS;
+
+    if (status==NULL || U_FAILURE(*status)) {
+            /*return NULL;*/
+            return fillIn;
+    }
+    if(resB == NULL) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            /*return NULL;*/
+            return fillIn;
+    }
+
+    if(resB->fIndex == resB->fSize-1) {
+      *status = U_INDEX_OUTOFBOUNDS_ERROR;
+      /*return NULL;*/
+    } else {
+        resB->fIndex++;
+        switch(RES_GET_TYPE(resB->fRes)) {
+        case URES_INT:
+        case URES_BINARY:
+        case URES_STRING:
+        case URES_STRING_V2:
+        case URES_INT_VECTOR:
+            return ures_copyResb(fillIn, resB, status);
+        case URES_TABLE:
+        case URES_TABLE16:
+        case URES_TABLE32:
+            r = res_getTableItemByIndex(&(resB->fResData), resB->fRes, resB->fIndex, &key);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return init_resb_result(&(resB->fResData), r, key, resB->fIndex, resB->fData, resB, 0, fillIn, status);
+        case URES_ARRAY:
+        case URES_ARRAY16:
+            r = res_getArrayItem(&(resB->fResData), resB->fRes, resB->fIndex);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return init_resb_result(&(resB->fResData), r, key, resB->fIndex, resB->fData, resB, 0, fillIn, status);
+        default:
+            /*return NULL;*/
+            return fillIn;
+        }
+    }
+    /*return NULL;*/
+    return fillIn;
+}
+
+U_CAPI UResourceBundle* U_EXPORT2 ures_getByIndex(const UResourceBundle *resB, int32_t indexR, UResourceBundle *fillIn, UErrorCode *status) {
+    const char* key = NULL;
+    Resource r = RES_BOGUS;
+
+    if (status==NULL || U_FAILURE(*status)) {
+        /*return NULL;*/
+        return fillIn;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        /*return NULL;*/
+        return fillIn;
+    }
+
+    if(indexR >= 0 && resB->fSize > indexR) {
+        switch(RES_GET_TYPE(resB->fRes)) {
+        case URES_INT:
+        case URES_BINARY:
+        case URES_STRING:
+        case URES_STRING_V2:
+        case URES_INT_VECTOR:
+            return ures_copyResb(fillIn, resB, status);
+        case URES_TABLE:
+        case URES_TABLE16:
+        case URES_TABLE32:
+            r = res_getTableItemByIndex(&(resB->fResData), resB->fRes, indexR, &key);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return init_resb_result(&(resB->fResData), r, key, indexR, resB->fData, resB, 0, fillIn, status);
+        case URES_ARRAY:
+        case URES_ARRAY16:
+            r = res_getArrayItem(&(resB->fResData), resB->fRes, indexR);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return init_resb_result(&(resB->fResData), r, key, indexR, resB->fData, resB, 0, fillIn, status);
+        default:
+            /*return NULL;*/
+            return fillIn;
+        }
+    } else {
+        *status = U_MISSING_RESOURCE_ERROR;
+    }
+    /*return NULL;*/
+    return fillIn;
+}
+
+U_CAPI const UChar* U_EXPORT2 ures_getStringByIndex(const UResourceBundle *resB, int32_t indexS, int32_t* len, UErrorCode *status) {
+    const char* key = NULL;
+    Resource r = RES_BOGUS;
+
+    if (status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(indexS >= 0 && resB->fSize > indexS) {
+        switch(RES_GET_TYPE(resB->fRes)) {
+        case URES_STRING:
+        case URES_STRING_V2:
+            return res_getString({resB}, &(resB->fResData), resB->fRes, len);
+        case URES_TABLE:
+        case URES_TABLE16:
+        case URES_TABLE32:
+            r = res_getTableItemByIndex(&(resB->fResData), resB->fRes, indexS, &key);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return ures_getStringWithAlias(resB, r, indexS, len, status);
+        case URES_ARRAY:
+        case URES_ARRAY16:
+            r = res_getArrayItem(&(resB->fResData), resB->fRes, indexS);
+            if(r == RES_BOGUS && resB->fHasFallback) {
+                /* TODO: do the fallback */
+            }
+            return ures_getStringWithAlias(resB, r, indexS, len, status);
+        case URES_ALIAS:
+            return ures_getStringWithAlias(resB, resB->fRes, indexS, len, status);
+        case URES_INT:
+        case URES_BINARY:
+        case URES_INT_VECTOR:
+            *status = U_RESOURCE_TYPE_MISMATCH;
+            break;
+        default:
+          /* must not occur */
+          *status = U_INTERNAL_PROGRAM_ERROR;
+          break;
+        }
+    } else {
+        *status = U_MISSING_RESOURCE_ERROR;
+    }
+    return NULL;
+}
+
+U_CAPI const char * U_EXPORT2
+ures_getUTF8StringByIndex(const UResourceBundle *resB,
+                          int32_t idx,
+                          char *dest, int32_t *pLength,
+                          UBool forceCopy,
+                          UErrorCode *status) {
+    int32_t length16;
+    const UChar *s16 = ures_getStringByIndex(resB, idx, &length16, status);
+    return ures_toUTF8String(s16, length16, dest, pLength, forceCopy, status);
+}
+
+/*U_CAPI const char *ures_getResPath(UResourceBundle *resB) {
+  return resB->fResPath;
+}*/
+
+U_CAPI UResourceBundle* U_EXPORT2
+ures_findResource(const char* path, UResourceBundle *fillIn, UErrorCode *status) 
+{
+  UResourceBundle *first = NULL; 
+  UResourceBundle *result = fillIn;
+  char *packageName = NULL;
+  char *pathToResource = NULL, *save = NULL;
+  char *locale = NULL, *localeEnd = NULL;
+  int32_t length;
+
+  if(status == NULL || U_FAILURE(*status)) {
+    return result;
+  }
+
+  length = (int32_t)(uprv_strlen(path)+1);
+  save = pathToResource = (char *)uprv_malloc(length*sizeof(char));
+  /* test for NULL */
+  if(pathToResource == NULL) {
+    *status = U_MEMORY_ALLOCATION_ERROR;
+    return result;
+  }
+  uprv_memcpy(pathToResource, path, length);
+
+  locale = pathToResource;
+  if(*pathToResource == RES_PATH_SEPARATOR) { /* there is a path specification */
+    pathToResource++;
+    packageName = pathToResource;
+    pathToResource = uprv_strchr(pathToResource, RES_PATH_SEPARATOR);
+    if(pathToResource == NULL) {
+      *status = U_ILLEGAL_ARGUMENT_ERROR;
+    } else {
+      *pathToResource = 0;
+      locale = pathToResource+1;
+    }
+  }
+
+  localeEnd = uprv_strchr(locale, RES_PATH_SEPARATOR);
+  if(localeEnd != NULL) {
+    *localeEnd = 0;
+  }
+
+  first = ures_open(packageName, locale, status);
+
+  if(U_SUCCESS(*status)) {
+    if(localeEnd) {
+      result = ures_findSubResource(first, localeEnd+1, fillIn, status);
+    } else {
+      result = ures_copyResb(fillIn, first, status);
+    }
+    ures_close(first);
+  }
+  uprv_free(save);
+  return result;
+}
+
+U_CAPI UResourceBundle* U_EXPORT2
+ures_findSubResource(const UResourceBundle *resB, char* path, UResourceBundle *fillIn, UErrorCode *status) 
+{
+  Resource res = RES_BOGUS;
+  UResourceBundle *result = fillIn;
+  const char *key;
+
+  if(status == NULL || U_FAILURE(*status)) {
+    return result;
+  }
+
+  /* here we do looping and circular alias checking */
+  /* this loop is here because aliasing is resolved on this level, not on res level */
+  /* so, when we encounter an alias, it is not an aggregate resource, so we return */
+  do {
+    res = res_findResource(&(resB->fResData), resB->fRes, &path, &key); 
+    if(res != RES_BOGUS) {
+        result = init_resb_result(&(resB->fResData), res, key, -1, resB->fData, resB, 0, fillIn, status);
+        resB = result;
+    } else {
+        *status = U_MISSING_RESOURCE_ERROR;
+        break;
+    }
+  } while(*path); /* there is more stuff in the path */
+
+  return result;
+}
+U_CAPI const UChar* U_EXPORT2 
+ures_getStringByKeyWithFallback(const UResourceBundle *resB, 
+                                const char* inKey, 
+                                int32_t* len,
+                                UErrorCode *status) {
+
+    UResourceBundle stack;
+    const UChar* retVal = NULL;
+    ures_initStackObject(&stack);
+    ures_getByKeyWithFallback(resB, inKey, &stack, status);
+    int32_t length;
+    retVal = ures_getString(&stack, &length, status);
+    ures_close(&stack);
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (length == 3 && retVal[0] == EMPTY_SET && retVal[1] == EMPTY_SET && retVal[2] == EMPTY_SET ) {
+        retVal = NULL;
+        length = 0;
+        *status = U_MISSING_RESOURCE_ERROR;
+    }
+    if (len != NULL) {
+        *len = length;
+    }
+    return retVal;
+}
+
+/*
+  Like res_getTableItemByKey but accepts full paths like "NumberElements/latn/patternsShort".
+*/  
+static Resource getTableItemByKeyPath(const ResourceData *pResData, Resource table, const char *key) {
+  Resource resource = table;  /* The current resource */
+  icu::CharString path;
+  UErrorCode errorCode = U_ZERO_ERROR;
+  path.append(key, errorCode);
+  if (U_FAILURE(errorCode)) { return RES_BOGUS; }
+  char *pathPart = path.data();  /* Path from current resource to desired resource */
+  UResType type = (UResType)RES_GET_TYPE(resource);  /* the current resource type */
+  while (*pathPart && resource != RES_BOGUS && URES_IS_CONTAINER(type)) {
+    char *nextPathPart = uprv_strchr(pathPart, RES_PATH_SEPARATOR);
+    if (nextPathPart != NULL) {
+      *nextPathPart = 0;  /* Terminating null for this part of path. */
+      nextPathPart++;
+    } else {
+      nextPathPart = uprv_strchr(pathPart, 0);
+    }
+    int32_t t;
+    const char *pathP = pathPart;
+    resource = res_getTableItemByKey(pResData, resource, &t, &pathP);
+    type = (UResType)RES_GET_TYPE(resource);
+    pathPart = nextPathPart; 
+  }
+  if (*pathPart) {
+    return RES_BOGUS;
+  }
+  return resource;
+}
+
+U_CAPI UResourceBundle* U_EXPORT2 
+ures_getByKeyWithFallback(const UResourceBundle *resB, 
+                          const char* inKey, 
+                          UResourceBundle *fillIn, 
+                          UErrorCode *status) {
+    Resource res = RES_BOGUS, rootRes = RES_BOGUS;
+    /*UResourceDataEntry *realData = NULL;*/
+    UResourceBundle *helper = NULL;
+
+    if (status==NULL || U_FAILURE(*status)) {
+        return fillIn;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return fillIn;
+    }
+
+    int32_t type = RES_GET_TYPE(resB->fRes);
+    if(URES_IS_TABLE(type)) {
+        res = getTableItemByKeyPath(&(resB->fResData), resB->fRes, inKey);
+        const char* key = inKey;
+        if(res == RES_BOGUS) {
+            UResourceDataEntry *dataEntry = resB->fData;
+            CharString path;
+            char *myPath = NULL;
+            const char* resPath = resB->fResPath;
+            int32_t len = resB->fResPathLen;
+            while(res == RES_BOGUS && dataEntry->fParent != NULL) { /* Otherwise, we'll look in parents */
+                dataEntry = dataEntry->fParent;
+                rootRes = dataEntry->fData.rootRes;
+
+                if(dataEntry->fBogus == U_ZERO_ERROR) {
+                    path.clear();
+                    if (len > 0) {
+                        path.append(resPath, len, *status);
+                    }
+                    path.append(inKey, *status);
+                    if (U_FAILURE(*status)) {
+                        ures_close(helper);
+                        return fillIn;
+                    }
+                    myPath = path.data();
+                    key = inKey;
+                    do {
+                        res = res_findResource(&(dataEntry->fData), rootRes, &myPath, &key);
+                        if (RES_GET_TYPE(res) == URES_ALIAS && *myPath) {
+                            /* We hit an alias, but we didn't finish following the path. */
+                            helper = init_resb_result(&(dataEntry->fData), res, NULL, -1, dataEntry, resB, 0, helper, status); 
+                            /*helper = init_resb_result(&(dataEntry->fData), res, inKey, -1, dataEntry, resB, 0, helper, status);*/
+                            if(helper) {
+                              dataEntry = helper->fData;
+                              rootRes = helper->fRes;
+                              resPath = helper->fResPath;
+                              len = helper->fResPathLen;
+
+                            } else {
+                              break;
+                            }
+                        }
+                    } while(*myPath); /* Continue until the whole path is consumed */
+                }
+            }
+            /*const ResourceData *rd = getFallbackData(resB, &key, &realData, &res, status);*/
+            if(res != RES_BOGUS) {
+              /* check if resB->fResPath gives the right name here */
+                if(uprv_strcmp(dataEntry->fName, uloc_getDefault())==0 || uprv_strcmp(dataEntry->fName, kRootLocaleName)==0) {
+                    *status = U_USING_DEFAULT_WARNING;
+                } else {
+                    *status = U_USING_FALLBACK_WARNING;
+                }
+
+                fillIn = init_resb_result(&(dataEntry->fData), res, inKey, -1, dataEntry, resB, 0, fillIn, status);
+            } else {
+                *status = U_MISSING_RESOURCE_ERROR;
+            }
+        } else {
+            fillIn = init_resb_result(&(resB->fResData), res, key, -1, resB->fData, resB, 0, fillIn, status);
+        }
+    } 
+    else {
+        *status = U_RESOURCE_TYPE_MISMATCH;
+    }
+    ures_close(helper);
+    return fillIn;
+}
+
+namespace {
+
+void getAllItemsWithFallback(
+        const UResourceBundle *bundle, ResourceDataValue &value,
+        ResourceSink &sink,
+        UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return; }
+    // We recursively enumerate child-first,
+    // only storing parent items in the absence of child items.
+    // The sink needs to store a placeholder value for the no-fallback/no-inheritance marker
+    // to prevent a parent item from being stored.
+    //
+    // It would be possible to recursively enumerate parent-first,
+    // overriding parent items with child items.
+    // When the sink sees the no-fallback/no-inheritance marker,
+    // then it would remove the parent's item.
+    // We would deserialize parent values even though they are overridden in a child bundle.
+    value.setData(&bundle->fResData);
+    UResourceDataEntry *parentEntry = bundle->fData->fParent;
+    UBool hasParent = parentEntry != NULL && U_SUCCESS(parentEntry->fBogus);
+    value.setResource(bundle->fRes, ResourceTracer(bundle));
+    sink.put(bundle->fKey, value, !hasParent, errorCode);
+    if (hasParent) {
+        // We might try to query the sink whether
+        // any fallback from the parent bundle is still possible.
+
+        // Turn the parent UResourceDataEntry into a UResourceBundle,
+        // much like in ures_openWithType().
+        // TODO: See if we can refactor ures_getByKeyWithFallback()
+        // and pull out an inner function that takes and returns a UResourceDataEntry
+        // so that we need not create UResourceBundle objects.
+        UResourceBundle parentBundle;
+        ures_initStackObject(&parentBundle);
+        parentBundle.fTopLevelData = parentBundle.fData = parentEntry;
+        // TODO: What is the difference between bundle fData and fTopLevelData?
+        uprv_memcpy(&parentBundle.fResData, &parentEntry->fData, sizeof(ResourceData));
+        // TODO: Try to replace bundle.fResData with just using bundle.fData->fData.
+        parentBundle.fHasFallback = !parentBundle.fResData.noFallback;
+        parentBundle.fIsTopLevel = TRUE;
+        parentBundle.fRes = parentBundle.fResData.rootRes;
+        parentBundle.fSize = res_countArrayItems(&(parentBundle.fResData), parentBundle.fRes);
+        parentBundle.fIndex = -1;
+        entryIncrease(parentEntry);
+
+        // Look up the container item in the parent bundle.
+        UResourceBundle containerBundle;
+        ures_initStackObject(&containerBundle);
+        const UResourceBundle *rb;
+        UErrorCode pathErrorCode = U_ZERO_ERROR;  // Ignore if parents up to root do not have this path.
+        if (bundle->fResPath == NULL || *bundle->fResPath == 0) {
+            rb = &parentBundle;
+        } else {
+            rb = ures_getByKeyWithFallback(&parentBundle, bundle->fResPath,
+                                           &containerBundle, &pathErrorCode);
+        }
+        if (U_SUCCESS(pathErrorCode)) {
+            getAllItemsWithFallback(rb, value, sink, errorCode);
+        }
+        ures_close(&containerBundle);
+        ures_close(&parentBundle);
+    }
+}
+
+}  // namespace
+
+// Requires a ResourceDataValue fill-in, so that we need not cast from a ResourceValue.
+// Unfortunately, the caller must know which subclass to make and pass in.
+// Alternatively, we could make it as polymorphic as in Java by
+// returning a ResourceValue pointer (possibly wrapped into a LocalPointer)
+// that the caller then owns.
+//
+// Also requires a UResourceBundle fill-in, so that the value's ResourceTracer
+// can point to a non-local bundle.
+// Without tracing, the child bundle could be a function-local object.
+U_CAPI void U_EXPORT2
+ures_getValueWithFallback(const UResourceBundle *bundle, const char *path,
+                          UResourceBundle *tempFillIn,
+                          ResourceDataValue &value, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return; }
+    if (path == nullptr) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    const UResourceBundle *rb;
+    if (*path == 0) {
+        // empty path
+        rb = bundle;
+    } else {
+        rb = ures_getByKeyWithFallback(bundle, path, tempFillIn, &errorCode);
+        if (U_FAILURE(errorCode)) {
+            return;
+        }
+    }
+    value.setData(&rb->fResData);
+    value.setResource(rb->fRes, ResourceTracer(rb));
+}
+
+U_CAPI void U_EXPORT2
+ures_getAllItemsWithFallback(const UResourceBundle *bundle, const char *path,
+                             icu::ResourceSink &sink, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return; }
+    if (path == nullptr) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    StackUResourceBundle stackBundle;
+    const UResourceBundle *rb;
+    if (*path == 0) {
+        // empty path
+        rb = bundle;
+    } else {
+        rb = ures_getByKeyWithFallback(bundle, path, stackBundle.getAlias(), &errorCode);
+        if (U_FAILURE(errorCode)) {
+            return;
+        }
+    }
+    // Get all table items with fallback.
+    ResourceDataValue value;
+    getAllItemsWithFallback(rb, value, sink, errorCode);
+}
+
+U_CAPI UResourceBundle* U_EXPORT2 ures_getByKey(const UResourceBundle *resB, const char* inKey, UResourceBundle *fillIn, UErrorCode *status) {
+    Resource res = RES_BOGUS;
+    UResourceDataEntry *realData = NULL;
+    const char *key = inKey;
+
+    if (status==NULL || U_FAILURE(*status)) {
+        return fillIn;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return fillIn;
+    }
+
+    int32_t type = RES_GET_TYPE(resB->fRes);
+    if(URES_IS_TABLE(type)) {
+        int32_t t;
+        res = res_getTableItemByKey(&(resB->fResData), resB->fRes, &t, &key);
+        if(res == RES_BOGUS) {
+            key = inKey;
+            if(resB->fHasFallback == TRUE) {
+                const ResourceData *rd = getFallbackData(resB, &key, &realData, &res, status);
+                if(U_SUCCESS(*status)) {
+                  /* check if resB->fResPath gives the right name here */
+                    return init_resb_result(rd, res, key, -1, realData, resB, 0, fillIn, status);
+                } else {
+                    *status = U_MISSING_RESOURCE_ERROR;
+                }
+            } else {
+                *status = U_MISSING_RESOURCE_ERROR;
+            }
+        } else {
+            return init_resb_result(&(resB->fResData), res, key, -1, resB->fData, resB, 0, fillIn, status);
+        }
+    } 
+#if 0
+    /* this is a kind of TODO item. If we have an array with an index table, we could do this. */
+    /* not currently */
+    else if(RES_GET_TYPE(resB->fRes) == URES_ARRAY && resB->fHasFallback == TRUE) {
+        /* here should go a first attempt to locate the key using index table */
+        const ResourceData *rd = getFallbackData(resB, &key, &realData, &res, status);
+        if(U_SUCCESS(*status)) {
+            return init_resb_result(rd, res, key, realData, resB, fillIn, status);
+        } else {
+            *status = U_MISSING_RESOURCE_ERROR;
+        }
+    }
+#endif    
+    else {
+        *status = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return fillIn;
+}
+
+U_CAPI const UChar* U_EXPORT2 ures_getStringByKey(const UResourceBundle *resB, const char* inKey, int32_t* len, UErrorCode *status) {
+    Resource res = RES_BOGUS;
+    UResourceDataEntry *realData = NULL;
+    const char* key = inKey;
+
+    if (status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(resB == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    int32_t type = RES_GET_TYPE(resB->fRes);
+    if(URES_IS_TABLE(type)) {
+        int32_t t=0;
+
+        res = res_getTableItemByKey(&(resB->fResData), resB->fRes, &t, &key);
+
+        if(res == RES_BOGUS) {
+            key = inKey;
+            if(resB->fHasFallback == TRUE) {
+                const ResourceData *rd = getFallbackData(resB, &key, &realData, &res, status);
+                if(U_SUCCESS(*status)) {
+                    switch (RES_GET_TYPE(res)) {
+                    case URES_STRING:
+                    case URES_STRING_V2:
+                        return res_getString({resB, key}, rd, res, len);
+                    case URES_ALIAS:
+                      {
+                        const UChar* result = 0;
+                        UResourceBundle *tempRes = ures_getByKey(resB, inKey, NULL, status);
+                        result = ures_getString(tempRes, len, status);
+                        ures_close(tempRes);
+                        return result;
+                      }
+                    default:
+                        *status = U_RESOURCE_TYPE_MISMATCH;
+                    }
+                } else {
+                    *status = U_MISSING_RESOURCE_ERROR;
+                }
+            } else {
+                *status = U_MISSING_RESOURCE_ERROR;
+            }
+        } else {
+            switch (RES_GET_TYPE(res)) {
+            case URES_STRING:
+            case URES_STRING_V2:
+                return res_getString({resB, key}, &(resB->fResData), res, len);
+            case URES_ALIAS:
+              {
+                const UChar* result = 0;
+                UResourceBundle *tempRes = ures_getByKey(resB, inKey, NULL, status);
+                result = ures_getString(tempRes, len, status);
+                ures_close(tempRes);
+                return result;
+              }
+            default:
+                *status = U_RESOURCE_TYPE_MISMATCH;
+            }
+        }
+    } 
+#if 0 
+    /* this is a kind of TODO item. If we have an array with an index table, we could do this. */
+    /* not currently */   
+    else if(RES_GET_TYPE(resB->fRes) == URES_ARRAY && resB->fHasFallback == TRUE) {
+        /* here should go a first attempt to locate the key using index table */
+        const ResourceData *rd = getFallbackData(resB, &key, &realData, &res, status);
+        if(U_SUCCESS(*status)) {
+            // TODO: Tracing
+            return res_getString(rd, res, len);
+        } else {
+            *status = U_MISSING_RESOURCE_ERROR;
+        }
+    } 
+#endif    
+    else {
+        *status = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return NULL;
+}
+
+U_CAPI const char * U_EXPORT2
+ures_getUTF8StringByKey(const UResourceBundle *resB,
+                        const char *key,
+                        char *dest, int32_t *pLength,
+                        UBool forceCopy,
+                        UErrorCode *status) {
+    int32_t length16;
+    const UChar *s16 = ures_getStringByKey(resB, key, &length16, status);
+    return ures_toUTF8String(s16, length16, dest, pLength, forceCopy, status);
+}
+
+/* TODO: clean from here down */
+
+/**
+ *  INTERNAL: Get the name of the first real locale (not placeholder) 
+ *  that has resource bundle data.
+ */
+U_CAPI const char*  U_EXPORT2
+ures_getLocaleInternal(const UResourceBundle* resourceBundle, UErrorCode* status)
+{
+    if (status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (!resourceBundle) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    } else {
+      return resourceBundle->fData->fName;
+    }
+}
+
+U_CAPI const char* U_EXPORT2 
+ures_getLocale(const UResourceBundle* resourceBundle, 
+               UErrorCode* status)
+{
+  return ures_getLocaleInternal(resourceBundle, status);
+}
+
+
+U_CAPI const char* U_EXPORT2 
+ures_getLocaleByType(const UResourceBundle* resourceBundle, 
+                     ULocDataLocaleType type, 
+                     UErrorCode* status) {
+    if (status==NULL || U_FAILURE(*status)) {
+        return NULL;
+    }
+    if (!resourceBundle) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    } else {
+        switch(type) {
+        case ULOC_ACTUAL_LOCALE:
+            return resourceBundle->fData->fName;
+        case ULOC_VALID_LOCALE:
+            return resourceBundle->fTopLevelData->fName;
+        case ULOC_REQUESTED_LOCALE:
+        default:
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return NULL;
+        }
+    }
+}
+
+U_CFUNC const char* ures_getName(const UResourceBundle* resB) {
+  if(resB == NULL) {
+    return NULL;
+  }
+
+  return resB->fData->fName;
+}
+
+#ifdef URES_DEBUG
+U_CFUNC const char* ures_getPath(const UResourceBundle* resB) {
+  if(resB == NULL) {
+    return NULL;
+  }
+
+  return resB->fData->fPath;
+}
+#endif
+
+static UResourceBundle*
+ures_openWithType(UResourceBundle *r, const char* path, const char* localeID,
+                  UResOpenType openType, UErrorCode* status) {
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    UResourceDataEntry *entry;
+    if(openType != URES_OPEN_DIRECT) {
+        /* first "canonicalize" the locale ID */
+        char canonLocaleID[ULOC_FULLNAME_CAPACITY];
+        uloc_getBaseName(localeID, canonLocaleID, UPRV_LENGTHOF(canonLocaleID), status);
+        if(U_FAILURE(*status) || *status == U_STRING_NOT_TERMINATED_WARNING) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return NULL;
+        }
+        entry = entryOpen(path, canonLocaleID, openType, status);
+    } else {
+        entry = entryOpenDirect(path, localeID, status);
+    }
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(entry == NULL) {
+        *status = U_MISSING_RESOURCE_ERROR;
+        return NULL;
+    }
+
+    UBool isStackObject;
+    if(r == NULL) {
+        r = (UResourceBundle *)uprv_malloc(sizeof(UResourceBundle));
+        if(r == NULL) {
+            entryClose(entry);
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+        isStackObject = FALSE;
+    } else {  // fill-in
+        isStackObject = ures_isStackObject(r);
+        ures_closeBundle(r, FALSE);
+    }
+    uprv_memset(r, 0, sizeof(UResourceBundle));
+    ures_setIsStackObject(r, isStackObject);
+
+    r->fTopLevelData = r->fData = entry;
+    uprv_memcpy(&r->fResData, &entry->fData, sizeof(ResourceData));
+    r->fHasFallback = openType != URES_OPEN_DIRECT && !r->fResData.noFallback;
+    r->fIsTopLevel = TRUE;
+    r->fRes = r->fResData.rootRes;
+    r->fSize = res_countArrayItems(&(r->fResData), r->fRes);
+    r->fIndex = -1;
+
+    ResourceTracer(r).traceOpen();
+
+    return r;
+}
+
+U_CAPI UResourceBundle* U_EXPORT2
+ures_open(const char* path, const char* localeID, UErrorCode* status) {
+    return ures_openWithType(NULL, path, localeID, URES_OPEN_LOCALE_DEFAULT_ROOT, status);
+}
+
+U_CAPI UResourceBundle* U_EXPORT2
+ures_openNoDefault(const char* path, const char* localeID, UErrorCode* status) {
+    return ures_openWithType(NULL, path, localeID, URES_OPEN_LOCALE_ROOT, status);
+}
+
+/**
+ *  Opens a resource bundle without "canonicalizing" the locale name. No fallback will be performed 
+ *  or sought. However, alias substitution will happen!
+ */
+U_CAPI UResourceBundle*  U_EXPORT2
+ures_openDirect(const char* path, const char* localeID, UErrorCode* status) {
+    return ures_openWithType(NULL, path, localeID, URES_OPEN_DIRECT, status);
+}
+
+/**
+ *  Internal API: This function is used to open a resource bundle 
+ *  proper fallback chaining is executed while initialization. 
+ *  The result is stored in cache for later fallback search.
+ * 
+ * Same as ures_open(), but uses the fill-in parameter and does not allocate a new bundle.
+ */
+U_CAPI void U_EXPORT2
+ures_openFillIn(UResourceBundle *r, const char* path,
+                const char* localeID, UErrorCode* status) {
+    if(U_SUCCESS(*status) && r == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    ures_openWithType(r, path, localeID, URES_OPEN_LOCALE_DEFAULT_ROOT, status);
+}
+
+/**
+ * Same as ures_openDirect(), but uses the fill-in parameter and does not allocate a new bundle.
+ */
+U_CAPI void U_EXPORT2
+ures_openDirectFillIn(UResourceBundle *r, const char* path, const char* localeID, UErrorCode* status) {
+    if(U_SUCCESS(*status) && r == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    ures_openWithType(r, path, localeID, URES_OPEN_DIRECT, status);
+}
+
+/**
+ *  API: Counts members. For arrays and tables, returns number of resources.
+ *  For strings, returns 1.
+ */
+U_CAPI int32_t  U_EXPORT2
+ures_countArrayItems(const UResourceBundle* resourceBundle,
+                  const char* resourceKey,
+                  UErrorCode* status)
+{
+    UResourceBundle resData;
+    ures_initStackObject(&resData);
+    if (status==NULL || U_FAILURE(*status)) {
+        return 0;
+    }
+    if(resourceBundle == NULL) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    ures_getByKey(resourceBundle, resourceKey, &resData, status);
+    
+    if(resData.fResData.data != NULL) {
+        int32_t result = res_countArrayItems(&resData.fResData, resData.fRes);
+        ures_close(&resData);
+        return result;
+    } else {
+        *status = U_MISSING_RESOURCE_ERROR;
+        ures_close(&resData);
+        return 0;
+    }
+}
+
+/**
+ * Internal function.
+ * Return the version number associated with this ResourceBundle as a string.
+ *
+ * @param resourceBundle The resource bundle for which the version is checked.
+ * @return  A version number string as specified in the resource bundle or its parent.
+ *          The caller does not own this string.
+ * @see ures_getVersion
+ * @internal
+ */
+U_CAPI const char* U_EXPORT2 
+ures_getVersionNumberInternal(const UResourceBundle *resourceBundle)
+{
+    if (!resourceBundle) return NULL;
+
+    if(resourceBundle->fVersion == NULL) {
+
+        /* If the version ID has not been built yet, then do so.  Retrieve */
+        /* the minor version from the file. */
+        UErrorCode status = U_ZERO_ERROR;
+        int32_t minor_len = 0;
+        int32_t len;
+
+        const UChar* minor_version = ures_getStringByKey(resourceBundle, kVersionTag, &minor_len, &status);
+        
+        /* Determine the length of of the final version string.  This is */
+        /* the length of the major part + the length of the separator */
+        /* (==1) + the length of the minor part (+ 1 for the zero byte at */
+        /* the end). */
+
+        len = (minor_len > 0) ? minor_len : 1;
+    
+        /* Allocate the string, and build it up. */
+        /* + 1 for zero byte */
+
+
+        ((UResourceBundle *)resourceBundle)->fVersion = (char *)uprv_malloc(1 + len); 
+        /* Check for null pointer. */
+        if (((UResourceBundle *)resourceBundle)->fVersion == NULL) {
+            return NULL;
+        }
+       
+        if(minor_len > 0) {
+            u_UCharsToChars(minor_version, resourceBundle->fVersion , minor_len);
+            resourceBundle->fVersion[len] =  '\0';
+        }
+        else {
+            uprv_strcpy(resourceBundle->fVersion, kDefaultMinorVersion);
+        }
+    }
+
+    return resourceBundle->fVersion;
+}
+
+U_CAPI const char*  U_EXPORT2
+ures_getVersionNumber(const UResourceBundle*   resourceBundle)
+{
+    return ures_getVersionNumberInternal(resourceBundle);
+}
+
+U_CAPI void U_EXPORT2 ures_getVersion(const UResourceBundle* resB, UVersionInfo versionInfo) {
+    if (!resB) return;
+
+    u_versionFromString(versionInfo, ures_getVersionNumberInternal(resB));
+}
+
+/** Tree support functions *******************************/
+#define INDEX_LOCALE_NAME "res_index"
+#define INDEX_TAG         "InstalledLocales"
+#define DEFAULT_TAG       "default"
+
+#if defined(URES_TREE_DEBUG)
+#include <stdio.h>
+#endif
+
+typedef struct ULocalesContext {
+    UResourceBundle installed;
+    UResourceBundle curr;
+} ULocalesContext;
+
+static void U_CALLCONV
+ures_loc_closeLocales(UEnumeration *enumerator) {
+    ULocalesContext *ctx = (ULocalesContext *)enumerator->context;
+    ures_close(&ctx->curr);
+    ures_close(&ctx->installed);
+    uprv_free(ctx);
+    uprv_free(enumerator);
+}
+
+static int32_t U_CALLCONV
+ures_loc_countLocales(UEnumeration *en, UErrorCode * /*status*/) {
+    ULocalesContext *ctx = (ULocalesContext *)en->context;
+    return ures_getSize(&ctx->installed);
+}
+
+U_CDECL_BEGIN
+
+
+static const char * U_CALLCONV
+ures_loc_nextLocale(UEnumeration* en,
+                    int32_t* resultLength,
+                    UErrorCode* status) {
+    ULocalesContext *ctx = (ULocalesContext *)en->context;
+    UResourceBundle *res = &(ctx->installed);
+    UResourceBundle *k = NULL;
+    const char *result = NULL;
+    int32_t len = 0;
+    if(ures_hasNext(res) && (k = ures_getNextResource(res, &ctx->curr, status)) != 0) {
+        result = ures_getKey(k);
+        len = (int32_t)uprv_strlen(result);
+    }
+    if (resultLength) {
+        *resultLength = len;
+    }
+    return result;
+}
+
+static void U_CALLCONV 
+ures_loc_resetLocales(UEnumeration* en, 
+                      UErrorCode* /*status*/) {
+    UResourceBundle *res = &((ULocalesContext *)en->context)->installed;
+    ures_resetIterator(res);
+}
+
+U_CDECL_END
+
+static const UEnumeration gLocalesEnum = {
+    NULL,
+        NULL,
+        ures_loc_closeLocales,
+        ures_loc_countLocales,
+        uenum_unextDefault,
+        ures_loc_nextLocale,
+        ures_loc_resetLocales
+};
+
+
+U_CAPI UEnumeration* U_EXPORT2
+ures_openAvailableLocales(const char *path, UErrorCode *status)
+{
+    UResourceBundle *idx = NULL;
+    UEnumeration *en = NULL;
+    ULocalesContext *myContext = NULL;
+
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+    myContext = static_cast<ULocalesContext *>(uprv_malloc(sizeof(ULocalesContext)));
+    en =  (UEnumeration *)uprv_malloc(sizeof(UEnumeration));
+    if(!en || !myContext) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        uprv_free(en);
+        uprv_free(myContext);
+        return NULL;
+    }
+    uprv_memcpy(en, &gLocalesEnum, sizeof(UEnumeration));
+
+    ures_initStackObject(&myContext->installed);
+    ures_initStackObject(&myContext->curr);
+    idx = ures_openDirect(path, INDEX_LOCALE_NAME, status);
+    ures_getByKey(idx, INDEX_TAG, &myContext->installed, status);
+    if(U_SUCCESS(*status)) {
+#if defined(URES_TREE_DEBUG)
+        fprintf(stderr, "Got %s::%s::[%s] : %s\n", 
+            path, INDEX_LOCALE_NAME, INDEX_TAG, ures_getKey(&myContext->installed));
+#endif
+        en->context = myContext;
+    } else {
+#if defined(URES_TREE_DEBUG)
+        fprintf(stderr, "%s open failed - %s\n", path, u_errorName(*status));
+#endif
+        ures_close(&myContext->installed);
+        uprv_free(myContext);
+        uprv_free(en);
+        en = NULL;
+    }
+    
+    ures_close(idx);
+    
+    return en;
+}
+
+static UBool isLocaleInList(UEnumeration *locEnum, const char *locToSearch, UErrorCode *status) {
+    const char *loc;
+    while ((loc = uenum_next(locEnum, NULL, status)) != NULL) {
+        if (uprv_strcmp(loc, locToSearch) == 0) {
+            return TRUE;
+        }
+    }
+    return FALSE;
+}
+
+U_CAPI int32_t U_EXPORT2
+ures_getFunctionalEquivalent(char *result, int32_t resultCapacity,
+                             const char *path, const char *resName, const char *keyword, const char *locid,
+                             UBool *isAvailable, UBool omitDefault, UErrorCode *status)
+{
+    char kwVal[1024] = ""; /* value of keyword 'keyword' */
+    char defVal[1024] = ""; /* default value for given locale */
+    char defLoc[1024] = ""; /* default value for given locale */
+    char base[1024] = ""; /* base locale */
+    char found[1024] = "";
+    char parent[1024] = "";
+    char full[1024] = "";
+    UResourceBundle bund1, bund2;
+    UResourceBundle *res = NULL;
+    UErrorCode subStatus = U_ZERO_ERROR;
+    int32_t length = 0;
+    if(U_FAILURE(*status)) return 0;
+    uloc_getKeywordValue(locid, keyword, kwVal, 1024-1,&subStatus);
+    if(!uprv_strcmp(kwVal, DEFAULT_TAG)) {
+        kwVal[0]=0;
+    }
+    uloc_getBaseName(locid, base, 1024-1,&subStatus);
+#if defined(URES_TREE_DEBUG)
+    fprintf(stderr, "getFunctionalEquivalent: \"%s\" [%s=%s] in %s - %s\n", 
+            locid, keyword, kwVal, base, u_errorName(subStatus));
+#endif
+    ures_initStackObject(&bund1);
+    ures_initStackObject(&bund2);
+    
+    
+    uprv_strcpy(parent, base);
+    uprv_strcpy(found, base);
+
+    if(isAvailable) { 
+        UEnumeration *locEnum = ures_openAvailableLocales(path, &subStatus);
+        *isAvailable = TRUE;
+        if (U_SUCCESS(subStatus)) {
+            *isAvailable = isLocaleInList(locEnum, parent, &subStatus);
+        }
+        uenum_close(locEnum);
+    }
+
+    if(U_FAILURE(subStatus)) {
+        *status = subStatus;
+        return 0;
+    }
+    
+    do {
+        subStatus = U_ZERO_ERROR;
+        res = ures_open(path, parent, &subStatus);
+        if(((subStatus == U_USING_FALLBACK_WARNING) ||
+            (subStatus == U_USING_DEFAULT_WARNING)) && isAvailable)
+        {
+            *isAvailable = FALSE;
+        }
+        isAvailable = NULL; /* only want to set this the first time around */
+        
+#if defined(URES_TREE_DEBUG)
+        fprintf(stderr, "%s;%s -> %s [%s]\n", path?path:"ICUDATA", parent, u_errorName(subStatus), ures_getLocale(res, &subStatus));
+#endif
+        if(U_FAILURE(subStatus)) {
+            *status = subStatus;
+        } else if(subStatus == U_ZERO_ERROR) {
+            ures_getByKey(res,resName,&bund1, &subStatus);
+            if(subStatus == U_ZERO_ERROR) {
+                const UChar *defUstr;
+                int32_t defLen;
+                /* look for default item */
+#if defined(URES_TREE_DEBUG)
+                fprintf(stderr, "%s;%s : loaded default -> %s\n",
+                    path?path:"ICUDATA", parent, u_errorName(subStatus));
+#endif
+                defUstr = ures_getStringByKey(&bund1, DEFAULT_TAG, &defLen, &subStatus);
+                if(U_SUCCESS(subStatus) && defLen) {
+                    u_UCharsToChars(defUstr, defVal, u_strlen(defUstr));
+#if defined(URES_TREE_DEBUG)
+                    fprintf(stderr, "%s;%s -> default %s=%s,  %s\n", 
+                        path?path:"ICUDATA", parent, keyword, defVal, u_errorName(subStatus));
+#endif
+                    uprv_strcpy(defLoc, parent);
+                    if(kwVal[0]==0) {
+                        uprv_strcpy(kwVal, defVal);
+#if defined(URES_TREE_DEBUG)
+                        fprintf(stderr, "%s;%s -> kwVal =  %s\n", 
+                            path?path:"ICUDATA", parent, keyword, kwVal);
+#endif
+                    }
+                }
+            }
+        }
+        
+        subStatus = U_ZERO_ERROR;
+
+        if (res != NULL) {
+            uprv_strcpy(found, ures_getLocaleByType(res, ULOC_VALID_LOCALE, &subStatus));
+        }
+
+        uloc_getParent(found,parent,sizeof(parent),&subStatus);
+        ures_close(res);
+    } while(!defVal[0] && *found && uprv_strcmp(found, "root") != 0 && U_SUCCESS(*status));
+    
+    /* Now, see if we can find the kwVal collator.. start the search over.. */
+    uprv_strcpy(parent, base);
+    uprv_strcpy(found, base);
+    
+    do {
+        subStatus = U_ZERO_ERROR;
+        res = ures_open(path, parent, &subStatus);
+        if((subStatus == U_USING_FALLBACK_WARNING) && isAvailable) {
+            *isAvailable = FALSE;
+        }
+        isAvailable = NULL; /* only want to set this the first time around */
+        
+#if defined(URES_TREE_DEBUG)
+        fprintf(stderr, "%s;%s -> %s (looking for %s)\n", 
+            path?path:"ICUDATA", parent, u_errorName(subStatus), kwVal);
+#endif
+        if(U_FAILURE(subStatus)) {
+            *status = subStatus;
+        } else if(subStatus == U_ZERO_ERROR) {
+            ures_getByKey(res,resName,&bund1, &subStatus);
+#if defined(URES_TREE_DEBUG)
+/**/ fprintf(stderr,"@%d [%s] %s\n", __LINE__, resName, u_errorName(subStatus));
+#endif
+            if(subStatus == U_ZERO_ERROR) {
+                ures_getByKey(&bund1, kwVal, &bund2, &subStatus);
+#if defined(URES_TREE_DEBUG)
+/**/ fprintf(stderr,"@%d [%s] %s\n", __LINE__, kwVal, u_errorName(subStatus));
+#endif
+                if(subStatus == U_ZERO_ERROR) {
+#if defined(URES_TREE_DEBUG)
+                    fprintf(stderr, "%s;%s -> full0 %s=%s,  %s\n", 
+                        path?path:"ICUDATA", parent, keyword, kwVal, u_errorName(subStatus));
+#endif
+                    uprv_strcpy(full, parent);
+                    if(*full == 0) {
+                        uprv_strcpy(full, "root");
+                    }
+                        /* now, recalculate default kw if need be */
+                        if(uprv_strlen(defLoc) > uprv_strlen(full)) {
+                          const UChar *defUstr;
+                          int32_t defLen;
+                          /* look for default item */
+#if defined(URES_TREE_DEBUG)
+                            fprintf(stderr, "%s;%s -> recalculating Default0\n", 
+                                    path?path:"ICUDATA", full);
+#endif
+                          defUstr = ures_getStringByKey(&bund1, DEFAULT_TAG, &defLen, &subStatus);
+                          if(U_SUCCESS(subStatus) && defLen) {
+                            u_UCharsToChars(defUstr, defVal, u_strlen(defUstr));
+#if defined(URES_TREE_DEBUG)
+                            fprintf(stderr, "%s;%s -> default0 %s=%s,  %s\n", 
+                                    path?path:"ICUDATA", full, keyword, defVal, u_errorName(subStatus));
+#endif
+                            uprv_strcpy(defLoc, full);
+                          }
+                        } /* end of recalculate default KW */
+#if defined(URES_TREE_DEBUG)
+                        else {
+                          fprintf(stderr, "No trim0,  %s <= %s\n", defLoc, full);
+                        }
+#endif
+                } else {
+#if defined(URES_TREE_DEBUG)
+                    fprintf(stderr, "err=%s in %s looking for %s\n", 
+                        u_errorName(subStatus), parent, kwVal);
+#endif
+                }
+            }
+        }
+        
+        subStatus = U_ZERO_ERROR;
+        
+        uprv_strcpy(found, parent);
+        uloc_getParent(found,parent,1023,&subStatus);
+        ures_close(res);
+    } while(!full[0] && *found && U_SUCCESS(*status));
+    
+    if((full[0]==0) && uprv_strcmp(kwVal, defVal)) {
+#if defined(URES_TREE_DEBUG)
+        fprintf(stderr, "Failed to locate kw %s - try default %s\n", kwVal, defVal);
+#endif
+        uprv_strcpy(kwVal, defVal);
+        uprv_strcpy(parent, base);
+        uprv_strcpy(found, base);
+        
+        do { /* search for 'default' named item */
+            subStatus = U_ZERO_ERROR;
+            res = ures_open(path, parent, &subStatus);
+            if((subStatus == U_USING_FALLBACK_WARNING) && isAvailable) {
+                *isAvailable = FALSE;
+            }
+            isAvailable = NULL; /* only want to set this the first time around */
+            
+#if defined(URES_TREE_DEBUG)
+            fprintf(stderr, "%s;%s -> %s (looking for default %s)\n",
+                path?path:"ICUDATA", parent, u_errorName(subStatus), kwVal);
+#endif
+            if(U_FAILURE(subStatus)) {
+                *status = subStatus;
+            } else if(subStatus == U_ZERO_ERROR) {
+                ures_getByKey(res,resName,&bund1, &subStatus);
+                if(subStatus == U_ZERO_ERROR) {
+                    ures_getByKey(&bund1, kwVal, &bund2, &subStatus);
+                    if(subStatus == U_ZERO_ERROR) {
+#if defined(URES_TREE_DEBUG)
+                        fprintf(stderr, "%s;%s -> full1 %s=%s,  %s\n", path?path:"ICUDATA",
+                            parent, keyword, kwVal, u_errorName(subStatus));
+#endif
+                        uprv_strcpy(full, parent);
+                        if(*full == 0) {
+                            uprv_strcpy(full, "root");
+                        }
+                        
+                        /* now, recalculate default kw if need be */
+                        if(uprv_strlen(defLoc) > uprv_strlen(full)) {
+                          const UChar *defUstr;
+                          int32_t defLen;
+                          /* look for default item */
+#if defined(URES_TREE_DEBUG)
+                            fprintf(stderr, "%s;%s -> recalculating Default1\n", 
+                                    path?path:"ICUDATA", full);
+#endif
+                          defUstr = ures_getStringByKey(&bund1, DEFAULT_TAG, &defLen, &subStatus);
+                          if(U_SUCCESS(subStatus) && defLen) {
+                            u_UCharsToChars(defUstr, defVal, u_strlen(defUstr));
+#if defined(URES_TREE_DEBUG)
+                            fprintf(stderr, "%s;%s -> default %s=%s,  %s\n", 
+                                    path?path:"ICUDATA", full, keyword, defVal, u_errorName(subStatus));
+#endif
+                            uprv_strcpy(defLoc, full);
+                          }
+                        } /* end of recalculate default KW */
+#if defined(URES_TREE_DEBUG)
+                        else {
+                          fprintf(stderr, "No trim1,  %s <= %s\n", defLoc, full);
+                        }
+#endif
+                    }
+                }
+            }
+            subStatus = U_ZERO_ERROR;
+            
+            uprv_strcpy(found, parent);
+            uloc_getParent(found,parent,1023,&subStatus);
+            ures_close(res);
+        } while(!full[0] && *found && U_SUCCESS(*status));
+    }
+    
+    if(U_SUCCESS(*status)) {
+        if(!full[0]) {
+#if defined(URES_TREE_DEBUG)
+          fprintf(stderr, "Still could not load keyword %s=%s\n", keyword, kwVal);
+#endif
+          *status = U_MISSING_RESOURCE_ERROR;
+        } else if(omitDefault) {
+#if defined(URES_TREE_DEBUG)
+          fprintf(stderr,"Trim? full=%s, defLoc=%s, found=%s\n", full, defLoc, found);
+#endif        
+          if(uprv_strlen(defLoc) <= uprv_strlen(full)) {
+            /* found the keyword in a *child* of where the default tag was present. */
+            if(!uprv_strcmp(kwVal, defVal)) { /* if the requested kw is default, */
+              /* and the default is in or in an ancestor of the current locale */
+#if defined(URES_TREE_DEBUG)
+              fprintf(stderr, "Removing unneeded var %s=%s\n", keyword, kwVal);
+#endif
+              kwVal[0]=0;
+            }
+          }
+        }
+        uprv_strcpy(found, full);
+        if(kwVal[0]) {
+            uprv_strcat(found, "@");
+            uprv_strcat(found, keyword);
+            uprv_strcat(found, "=");
+            uprv_strcat(found, kwVal);
+        } else if(!omitDefault) {
+            uprv_strcat(found, "@");
+            uprv_strcat(found, keyword);
+            uprv_strcat(found, "=");
+            uprv_strcat(found, defVal);
+        }
+    }
+    /* we found the default locale - no need to repeat it.*/
+    
+    ures_close(&bund1);
+    ures_close(&bund2);
+    
+    length = (int32_t)uprv_strlen(found);
+
+    if(U_SUCCESS(*status)) {
+        int32_t copyLength = uprv_min(length, resultCapacity);
+        if(copyLength>0) {
+            uprv_strncpy(result, found, copyLength);
+        }
+        if(length == 0) {
+          *status = U_MISSING_RESOURCE_ERROR; 
+        }
+    } else {
+        length = 0;
+        result[0]=0;
+    }
+    return u_terminateChars(result, resultCapacity, length, status);
+}
+
+U_CAPI UEnumeration* U_EXPORT2
+ures_getKeywordValues(const char *path, const char *keyword, UErrorCode *status)
+{
+#define VALUES_BUF_SIZE 2048
+#define VALUES_LIST_SIZE 512
+    
+    char       valuesBuf[VALUES_BUF_SIZE];
+    int32_t    valuesIndex = 0;
+    const char *valuesList[VALUES_LIST_SIZE];
+    int32_t    valuesCount = 0;
+    
+    const char *locale;
+    int32_t     locLen;
+    
+    UEnumeration *locs = NULL;
+    
+    UResourceBundle    item;
+    UResourceBundle    subItem;
+    
+    ures_initStackObject(&item);
+    ures_initStackObject(&subItem);
+    locs = ures_openAvailableLocales(path, status);
+    
+    if(U_FAILURE(*status)) {
+        ures_close(&item);
+        ures_close(&subItem);
+        return NULL;
+    }
+    
+    valuesBuf[0]=0;
+    valuesBuf[1]=0;
+    
+    while((locale = uenum_next(locs, &locLen, status)) != 0) {
+        UResourceBundle   *bund = NULL;
+        UResourceBundle   *subPtr = NULL;
+        UErrorCode subStatus = U_ZERO_ERROR; /* don't fail if a bundle is unopenable */
+        bund = ures_openDirect(path, locale, &subStatus);
+        
+#if defined(URES_TREE_DEBUG)
+        if(!bund || U_FAILURE(subStatus)) {
+            fprintf(stderr, "%s-%s values: Can't open %s locale - skipping. (%s)\n", 
+                path?path:"<ICUDATA>", keyword, locale, u_errorName(subStatus));
+        }
+#endif
+        
+        ures_getByKey(bund, keyword, &item, &subStatus);
+        
+        if(!bund || U_FAILURE(subStatus)) {
+#if defined(URES_TREE_DEBUG)
+            fprintf(stderr, "%s-%s values: Can't find in %s - skipping. (%s)\n", 
+                path?path:"<ICUDATA>", keyword, locale, u_errorName(subStatus));
+#endif
+            ures_close(bund);
+            bund = NULL;
+            continue;
+        }
+        
+        while((subPtr = ures_getNextResource(&item,&subItem,&subStatus)) != 0
+            && U_SUCCESS(subStatus)) {
+            const char *k;
+            int32_t i;
+            k = ures_getKey(subPtr);
+
+#if defined(URES_TREE_DEBUG)
+            /* fprintf(stderr, "%s | %s | %s | %s\n", path?path:"<ICUDATA>", keyword, locale, k); */
+#endif
+            if(k == NULL || *k == 0 ||
+                    uprv_strcmp(k, DEFAULT_TAG) == 0 || uprv_strncmp(k, "private-", 8) == 0) {
+                // empty or "default" or unlisted type
+                continue;
+            }
+            for(i=0; i<valuesCount; i++) {
+                if(!uprv_strcmp(valuesList[i],k)) {
+                    k = NULL; /* found duplicate */
+                    break;
+                }
+            }
+            if(k != NULL) {
+                int32_t kLen = (int32_t)uprv_strlen(k);
+                if((valuesCount >= (VALUES_LIST_SIZE-1)) ||       /* no more space in list .. */
+                    ((valuesIndex+kLen+1+1) >= VALUES_BUF_SIZE)) { /* no more space in buffer (string + 2 nulls) */
+                    *status = U_ILLEGAL_ARGUMENT_ERROR; /* out of space.. */
+                } else {
+                    uprv_strcpy(valuesBuf+valuesIndex, k);
+                    valuesList[valuesCount++] = valuesBuf+valuesIndex;
+                    valuesIndex += kLen;
+#if defined(URES_TREE_DEBUG)
+                    fprintf(stderr, "%s | %s | %s | [%s]   (UNIQUE)\n",
+                        path?path:"<ICUDATA>", keyword, locale, k);
+#endif
+                    valuesBuf[valuesIndex++] = 0; /* terminate */
+                }
+            }
+        }
+        ures_close(bund);
+    }
+    valuesBuf[valuesIndex++] = 0; /* terminate */
+    
+    ures_close(&item);
+    ures_close(&subItem);
+    uenum_close(locs);
+#if defined(URES_TREE_DEBUG)
+    fprintf(stderr, "%s:  size %d, #%d\n", u_errorName(*status), 
+        valuesIndex, valuesCount);
+#endif
+    return uloc_openKeywordList(valuesBuf, valuesIndex, status);
+}
+#if 0
+/* This code isn't needed, and given the documentation warnings the implementation is suspect */
+U_CAPI UBool U_EXPORT2
+ures_equal(const UResourceBundle* res1, const UResourceBundle* res2){
+    if(res1==NULL || res2==NULL){
+        return res1==res2; /* pointer comparision */
+    }
+    if(res1->fKey==NULL||  res2->fKey==NULL){
+        return (res1->fKey==res2->fKey);
+    }else{
+        if(uprv_strcmp(res1->fKey, res2->fKey)!=0){
+            return FALSE;
+        }
+    }
+    if(uprv_strcmp(res1->fData->fName, res2->fData->fName)!=0){
+        return FALSE;
+    }
+    if(res1->fData->fPath == NULL||  res2->fData->fPath==NULL){
+        return (res1->fData->fPath == res2->fData->fPath);
+    }else{
+        if(uprv_strcmp(res1->fData->fPath, res2->fData->fPath)!=0){
+            return FALSE;
+        }
+    }
+    if(uprv_strcmp(res1->fData->fParent->fName, res2->fData->fParent->fName)!=0){
+        return FALSE;
+    }
+    if(uprv_strcmp(res1->fData->fParent->fPath, res2->fData->fParent->fPath)!=0){
+        return FALSE;
+    }
+    if(uprv_strncmp(res1->fResPath, res2->fResPath, res1->fResPathLen)!=0){
+        return FALSE;
+    }
+    if(res1->fRes != res2->fRes){
+        return FALSE;
+    }
+    return TRUE;
+}
+U_CAPI UResourceBundle* U_EXPORT2
+ures_clone(const UResourceBundle* res, UErrorCode* status){
+    UResourceBundle* bundle = NULL;
+    UResourceBundle* ret = NULL;
+    if(U_FAILURE(*status) || res == NULL){
+        return NULL;
+    }
+    bundle = ures_open(res->fData->fPath, res->fData->fName, status);
+    if(res->fResPath!=NULL){
+        ret = ures_findSubResource(bundle, res->fResPath, NULL, status);
+        ures_close(bundle);
+    }else{
+        ret = bundle;
+    }
+    return ret;
+}
+U_CAPI const UResourceBundle* U_EXPORT2
+ures_getParentBundle(const UResourceBundle* res){
+    if(res==NULL){
+        return NULL;
+    }
+    return res->fParentRes;
+}
+#endif
+
+U_CAPI void U_EXPORT2
+ures_getVersionByKey(const UResourceBundle* res, const char *key, UVersionInfo ver, UErrorCode *status) {
+  const UChar *str;
+  int32_t len;
+  str = ures_getStringByKey(res, key, &len, status);
+  if(U_SUCCESS(*status)) {
+    u_versionFromUString(ver, str);
+  } 
+}
+
+/* eof */
diff --git a/thirdparty/icu4c/common/uresdata.cpp b/thirdparty/icu4c/common/uresdata.cpp
new file mode 100644
index 0000000000..ae731e4544
--- /dev/null
+++ b/thirdparty/icu4c/common/uresdata.cpp
@@ -0,0 +1,1518 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+* Copyright (C) 1999-2016, International Business Machines Corporation
+*               and others. All Rights Reserved.
+*******************************************************************************
+*   file name:  uresdata.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999dec08
+*   created by: Markus W. Scherer
+* Modification History:
+*
+*   Date        Name        Description
+*   06/20/2000  helena      OS/400 port changes; mostly typecast.
+*   06/24/02    weiv        Added support for resource sharing
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/udata.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "resource.h"
+#include "uarrsort.h"
+#include "uassert.h"
+#include "ucol_swp.h"
+#include "udataswp.h"
+#include "uinvchar.h"
+#include "uresdata.h"
+#include "uresimp.h"
+#include "utracimp.h"
+
+/*
+ * Resource access helpers
+ */
+
+/* get a const char* pointer to the key with the keyOffset byte offset from pRoot */
+#define RES_GET_KEY16(pResData, keyOffset) \
+    ((keyOffset)<(pResData)->localKeyLimit ? \
+        (const char *)(pResData)->pRoot+(keyOffset) : \
+        (pResData)->poolBundleKeys+(keyOffset)-(pResData)->localKeyLimit)
+
+#define RES_GET_KEY32(pResData, keyOffset) \
+    ((keyOffset)>=0 ? \
+        (const char *)(pResData)->pRoot+(keyOffset) : \
+        (pResData)->poolBundleKeys+((keyOffset)&0x7fffffff))
+
+#define URESDATA_ITEM_NOT_FOUND -1
+
+/* empty resources, returned when the resource offset is 0 */
+static const uint16_t gEmpty16=0;
+
+static const struct {
+    int32_t length;
+    int32_t res;
+} gEmpty32={ 0, 0 };
+
+static const struct {
+    int32_t length;
+    UChar nul;
+    UChar pad;
+} gEmptyString={ 0, 0, 0 };
+
+/*
+ * All the type-access functions assume that
+ * the resource is of the expected type.
+ */
+
+static int32_t
+_res_findTableItem(const ResourceData *pResData, const uint16_t *keyOffsets, int32_t length,
+                   const char *key, const char **realKey) {
+    const char *tableKey;
+    int32_t mid, start, limit;
+    int result;
+
+    /* do a binary search for the key */
+    start=0;
+    limit=length;
+    while(start<limit) {
+        mid = (start + limit) / 2;
+        tableKey = RES_GET_KEY16(pResData, keyOffsets[mid]);
+        if (pResData->useNativeStrcmp) {
+            result = uprv_strcmp(key, tableKey);
+        } else {
+            result = uprv_compareInvCharsAsAscii(key, tableKey);
+        }
+        if (result < 0) {
+            limit = mid;
+        } else if (result > 0) {
+            start = mid + 1;
+        } else {
+            /* We found it! */
+            *realKey=tableKey;
+            return mid;
+        }
+    }
+    return URESDATA_ITEM_NOT_FOUND;  /* not found or table is empty. */
+}
+
+static int32_t
+_res_findTable32Item(const ResourceData *pResData, const int32_t *keyOffsets, int32_t length,
+                     const char *key, const char **realKey) {
+    const char *tableKey;
+    int32_t mid, start, limit;
+    int result;
+
+    /* do a binary search for the key */
+    start=0;
+    limit=length;
+    while(start<limit) {
+        mid = (start + limit) / 2;
+        tableKey = RES_GET_KEY32(pResData, keyOffsets[mid]);
+        if (pResData->useNativeStrcmp) {
+            result = uprv_strcmp(key, tableKey);
+        } else {
+            result = uprv_compareInvCharsAsAscii(key, tableKey);
+        }
+        if (result < 0) {
+            limit = mid;
+        } else if (result > 0) {
+            start = mid + 1;
+        } else {
+            /* We found it! */
+            *realKey=tableKey;
+            return mid;
+        }
+    }
+    return URESDATA_ITEM_NOT_FOUND;  /* not found or table is empty. */
+}
+
+/* helper for res_load() ---------------------------------------------------- */
+
+static UBool U_CALLCONV
+isAcceptable(void *context,
+             const char * /*type*/, const char * /*name*/,
+             const UDataInfo *pInfo) {
+    uprv_memcpy(context, pInfo->formatVersion, 4);
+    return (UBool)(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->sizeofUChar==U_SIZEOF_UCHAR &&
+        pInfo->dataFormat[0]==0x52 &&   /* dataFormat="ResB" */
+        pInfo->dataFormat[1]==0x65 &&
+        pInfo->dataFormat[2]==0x73 &&
+        pInfo->dataFormat[3]==0x42 &&
+        (1<=pInfo->formatVersion[0] && pInfo->formatVersion[0]<=3));
+}
+
+/* semi-public functions ---------------------------------------------------- */
+
+static void
+res_init(ResourceData *pResData,
+         UVersionInfo formatVersion, const void *inBytes, int32_t length,
+         UErrorCode *errorCode) {
+    UResType rootType;
+
+    /* get the root resource */
+    pResData->pRoot=(const int32_t *)inBytes;
+    pResData->rootRes=(Resource)*pResData->pRoot;
+    pResData->p16BitUnits=&gEmpty16;
+
+    /* formatVersion 1.1 must have a root item and at least 5 indexes */
+    if(length>=0 && (length/4)<((formatVersion[0]==1 && formatVersion[1]==0) ? 1 : 1+5)) {
+        *errorCode=U_INVALID_FORMAT_ERROR;
+        res_unload(pResData);
+        return;
+    }
+
+    /* currently, we accept only resources that have a Table as their roots */
+    rootType=(UResType)RES_GET_TYPE(pResData->rootRes);
+    if(!URES_IS_TABLE(rootType)) {
+        *errorCode=U_INVALID_FORMAT_ERROR;
+        res_unload(pResData);
+        return;
+    }
+
+    if(formatVersion[0]==1 && formatVersion[1]==0) {
+        pResData->localKeyLimit=0x10000;  /* greater than any 16-bit key string offset */
+    } else {
+        /* bundles with formatVersion 1.1 and later contain an indexes[] array */
+        const int32_t *indexes=pResData->pRoot+1;
+        int32_t indexLength=indexes[URES_INDEX_LENGTH]&0xff;
+        if(indexLength<=URES_INDEX_MAX_TABLE_LENGTH) {
+            *errorCode=U_INVALID_FORMAT_ERROR;
+            res_unload(pResData);
+            return;
+        }
+        if( length>=0 &&
+            (length<((1+indexLength)<<2) ||
+             length<(indexes[URES_INDEX_BUNDLE_TOP]<<2))
+        ) {
+            *errorCode=U_INVALID_FORMAT_ERROR;
+            res_unload(pResData);
+            return;
+        }
+        if(indexes[URES_INDEX_KEYS_TOP]>(1+indexLength)) {
+            pResData->localKeyLimit=indexes[URES_INDEX_KEYS_TOP]<<2;
+        }
+        if(formatVersion[0]>=3) {
+            // In formatVersion 1, the indexLength took up this whole int.
+            // In version 2, bits 31..8 were reserved and always 0.
+            // In version 3, they contain bits 23..0 of the poolStringIndexLimit.
+            // Bits 27..24 are in indexes[URES_INDEX_ATTRIBUTES] bits 15..12.
+            pResData->poolStringIndexLimit=(int32_t)((uint32_t)indexes[URES_INDEX_LENGTH]>>8);
+        }
+        if(indexLength>URES_INDEX_ATTRIBUTES) {
+            int32_t att=indexes[URES_INDEX_ATTRIBUTES];
+            pResData->noFallback=(UBool)(att&URES_ATT_NO_FALLBACK);
+            pResData->isPoolBundle=(UBool)((att&URES_ATT_IS_POOL_BUNDLE)!=0);
+            pResData->usesPoolBundle=(UBool)((att&URES_ATT_USES_POOL_BUNDLE)!=0);
+            pResData->poolStringIndexLimit|=(att&0xf000)<<12;  // bits 15..12 -> 27..24
+            pResData->poolStringIndex16Limit=(int32_t)((uint32_t)att>>16);
+        }
+        if((pResData->isPoolBundle || pResData->usesPoolBundle) && indexLength<=URES_INDEX_POOL_CHECKSUM) {
+            *errorCode=U_INVALID_FORMAT_ERROR;
+            res_unload(pResData);
+            return;
+        }
+        if( indexLength>URES_INDEX_16BIT_TOP &&
+            indexes[URES_INDEX_16BIT_TOP]>indexes[URES_INDEX_KEYS_TOP]
+        ) {
+            pResData->p16BitUnits=(const uint16_t *)(pResData->pRoot+indexes[URES_INDEX_KEYS_TOP]);
+        }
+    }
+
+    if(formatVersion[0]==1 || U_CHARSET_FAMILY==U_ASCII_FAMILY) {
+        /*
+         * formatVersion 1: compare key strings in native-charset order
+         * formatVersion 2 and up: compare key strings in ASCII order
+         */
+        pResData->useNativeStrcmp=TRUE;
+    }
+}
+
+U_CAPI void U_EXPORT2
+res_read(ResourceData *pResData,
+         const UDataInfo *pInfo, const void *inBytes, int32_t length,
+         UErrorCode *errorCode) {
+    UVersionInfo formatVersion;
+
+    uprv_memset(pResData, 0, sizeof(ResourceData));
+    if(U_FAILURE(*errorCode)) {
+        return;
+    }
+    if(!isAcceptable(formatVersion, NULL, NULL, pInfo)) {
+        *errorCode=U_INVALID_FORMAT_ERROR;
+        return;
+    }
+    res_init(pResData, formatVersion, inBytes, length, errorCode);
+}
+
+U_CFUNC void
+res_load(ResourceData *pResData,
+         const char *path, const char *name, UErrorCode *errorCode) {
+    UVersionInfo formatVersion;
+
+    uprv_memset(pResData, 0, sizeof(ResourceData));
+
+    /* load the ResourceBundle file */
+    pResData->data=udata_openChoice(path, "res", name, isAcceptable, formatVersion, errorCode);
+    if(U_FAILURE(*errorCode)) {
+        return;
+    }
+
+    /* get its memory and initialize *pResData */
+    res_init(pResData, formatVersion, udata_getMemory(pResData->data), -1, errorCode);
+}
+
+U_CFUNC void
+res_unload(ResourceData *pResData) {
+    if(pResData->data!=NULL) {
+        udata_close(pResData->data);
+        pResData->data=NULL;
+    }
+}
+
+static const int8_t gPublicTypes[URES_LIMIT] = {
+    URES_STRING,
+    URES_BINARY,
+    URES_TABLE,
+    URES_ALIAS,
+
+    URES_TABLE,     /* URES_TABLE32 */
+    URES_TABLE,     /* URES_TABLE16 */
+    URES_STRING,    /* URES_STRING_V2 */
+    URES_INT,
+
+    URES_ARRAY,
+    URES_ARRAY,     /* URES_ARRAY16 */
+    URES_NONE,
+    URES_NONE,
+
+    URES_NONE,
+    URES_NONE,
+    URES_INT_VECTOR,
+    URES_NONE
+};
+
+U_CAPI UResType U_EXPORT2
+res_getPublicType(Resource res) {
+    return (UResType)gPublicTypes[RES_GET_TYPE(res)];
+}
+
+U_CAPI const UChar * U_EXPORT2
+res_getStringNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength) {
+    const UChar *p;
+    uint32_t offset=RES_GET_OFFSET(res);
+    int32_t length;
+    if(RES_GET_TYPE(res)==URES_STRING_V2) {
+        int32_t first;
+        if((int32_t)offset<pResData->poolStringIndexLimit) {
+            p=(const UChar *)pResData->poolBundleStrings+offset;
+        } else {
+            p=(const UChar *)pResData->p16BitUnits+(offset-pResData->poolStringIndexLimit);
+        }
+        first=*p;
+        if(!U16_IS_TRAIL(first)) {
+            length=u_strlen(p);
+        } else if(first<0xdfef) {
+            length=first&0x3ff;
+            ++p;
+        } else if(first<0xdfff) {
+            length=((first-0xdfef)<<16)|p[1];
+            p+=2;
+        } else {
+            length=((int32_t)p[1]<<16)|p[2];
+            p+=3;
+        }
+    } else if(res==offset) /* RES_GET_TYPE(res)==URES_STRING */ {
+        const int32_t *p32= res==0 ? &gEmptyString.length : pResData->pRoot+res;
+        length=*p32++;
+        p=(const UChar *)p32;
+    } else {
+        p=NULL;
+        length=0;
+    }
+    if(pLength) {
+        *pLength=length;
+    }
+    return p;
+}
+
+namespace {
+
+/**
+ * CLDR string value (three empty-set symbols)=={2205, 2205, 2205}
+ * prevents fallback to the parent bundle.
+ * TODO: combine with other code that handles this marker, use EMPTY_SET constant.
+ * TODO: maybe move to uresbund.cpp?
+ */
+UBool isNoInheritanceMarker(const ResourceData *pResData, Resource res) {
+    uint32_t offset=RES_GET_OFFSET(res);
+    if (offset == 0) {
+        // empty string
+    } else if (res == offset) {
+        const int32_t *p32=pResData->pRoot+res;
+        int32_t length=*p32;
+        const UChar *p=(const UChar *)p32;
+        return length == 3 && p[2] == 0x2205 && p[3] == 0x2205 && p[4] == 0x2205;
+    } else if (RES_GET_TYPE(res) == URES_STRING_V2) {
+        const UChar *p;
+        if((int32_t)offset<pResData->poolStringIndexLimit) {
+            p=(const UChar *)pResData->poolBundleStrings+offset;
+        } else {
+            p=(const UChar *)pResData->p16BitUnits+(offset-pResData->poolStringIndexLimit);
+        }
+        int32_t first=*p;
+        if (first == 0x2205) {  // implicit length
+            return p[1] == 0x2205 && p[2] == 0x2205 && p[3] == 0;
+        } else if (first == 0xdc03) {  // explicit length 3 (should not occur)
+            return p[1] == 0x2205 && p[2] == 0x2205 && p[3] == 0x2205;
+        } else {
+            // Assume that the string has not been stored with more length units than necessary.
+            return FALSE;
+        }
+    }
+    return FALSE;
+}
+
+int32_t getStringArray(const ResourceData *pResData, const icu::ResourceArray &array,
+                       icu::UnicodeString *dest, int32_t capacity,
+                       UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(dest == NULL ? capacity != 0 : capacity < 0) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    int32_t length = array.getSize();
+    if(length == 0) {
+        return 0;
+    }
+    if(length > capacity) {
+        errorCode = U_BUFFER_OVERFLOW_ERROR;
+        return length;
+    }
+    for(int32_t i = 0; i < length; ++i) {
+        int32_t sLength;
+        // No tracing: handled by the caller
+        const UChar *s = res_getStringNoTrace(pResData, array.internalGetResource(pResData, i), &sLength);
+        if(s == NULL) {
+            errorCode = U_RESOURCE_TYPE_MISMATCH;
+            return 0;
+        }
+        dest[i].setTo(TRUE, s, sLength);
+    }
+    return length;
+}
+
+}  // namespace
+
+U_CAPI const UChar * U_EXPORT2
+res_getAlias(const ResourceData *pResData, Resource res, int32_t *pLength) {
+    const UChar *p;
+    uint32_t offset=RES_GET_OFFSET(res);
+    int32_t length;
+    if(RES_GET_TYPE(res)==URES_ALIAS) {
+        const int32_t *p32= offset==0 ? &gEmptyString.length : pResData->pRoot+offset;
+        length=*p32++;
+        p=(const UChar *)p32;
+    } else {
+        p=NULL;
+        length=0;
+    }
+    if(pLength) {
+        *pLength=length;
+    }
+    return p;
+}
+
+U_CAPI const uint8_t * U_EXPORT2
+res_getBinaryNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength) {
+    const uint8_t *p;
+    uint32_t offset=RES_GET_OFFSET(res);
+    int32_t length;
+    if(RES_GET_TYPE(res)==URES_BINARY) {
+        const int32_t *p32= offset==0 ? (const int32_t*)&gEmpty32 : pResData->pRoot+offset;
+        length=*p32++;
+        p=(const uint8_t *)p32;
+    } else {
+        p=NULL;
+        length=0;
+    }
+    if(pLength) {
+        *pLength=length;
+    }
+    return p;
+}
+
+
+U_CAPI const int32_t * U_EXPORT2
+res_getIntVectorNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength) {
+    const int32_t *p;
+    uint32_t offset=RES_GET_OFFSET(res);
+    int32_t length;
+    if(RES_GET_TYPE(res)==URES_INT_VECTOR) {
+        p= offset==0 ? (const int32_t *)&gEmpty32 : pResData->pRoot+offset;
+        length=*p++;
+    } else {
+        p=NULL;
+        length=0;
+    }
+    if(pLength) {
+        *pLength=length;
+    }
+    return p;
+}
+
+U_CAPI int32_t U_EXPORT2
+res_countArrayItems(const ResourceData *pResData, Resource res) {
+    uint32_t offset=RES_GET_OFFSET(res);
+    switch(RES_GET_TYPE(res)) {
+    case URES_STRING:
+    case URES_STRING_V2:
+    case URES_BINARY:
+    case URES_ALIAS:
+    case URES_INT:
+    case URES_INT_VECTOR:
+        return 1;
+    case URES_ARRAY:
+    case URES_TABLE32:
+        return offset==0 ? 0 : *(pResData->pRoot+offset);
+    case URES_TABLE:
+        return offset==0 ? 0 : *((const uint16_t *)(pResData->pRoot+offset));
+    case URES_ARRAY16:
+    case URES_TABLE16:
+        return pResData->p16BitUnits[offset];
+    default:
+        return 0;
+    }
+}
+
+U_NAMESPACE_BEGIN
+
+ResourceDataValue::~ResourceDataValue() {}
+
+UResType ResourceDataValue::getType() const {
+    return res_getPublicType(res);
+}
+
+const UChar *ResourceDataValue::getString(int32_t &length, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    const UChar *s = res_getString(fTraceInfo, &getData(), res, &length);
+    if(s == NULL) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return s;
+}
+
+const UChar *ResourceDataValue::getAliasString(int32_t &length, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    const UChar *s = res_getAlias(&getData(), res, &length);
+    if(s == NULL) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return s;
+}
+
+int32_t ResourceDataValue::getInt(UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(RES_GET_TYPE(res) != URES_INT) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return res_getInt(fTraceInfo, res);
+}
+
+uint32_t ResourceDataValue::getUInt(UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(RES_GET_TYPE(res) != URES_INT) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return res_getUInt(fTraceInfo, res);
+}
+
+const int32_t *ResourceDataValue::getIntVector(int32_t &length, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    const int32_t *iv = res_getIntVector(fTraceInfo, &getData(), res, &length);
+    if(iv == NULL) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return iv;
+}
+
+const uint8_t *ResourceDataValue::getBinary(int32_t &length, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return NULL;
+    }
+    const uint8_t *b = res_getBinary(fTraceInfo, &getData(), res, &length);
+    if(b == NULL) {
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+    }
+    return b;
+}
+
+ResourceArray ResourceDataValue::getArray(UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return ResourceArray();
+    }
+    const uint16_t *items16 = NULL;
+    const Resource *items32 = NULL;
+    uint32_t offset=RES_GET_OFFSET(res);
+    int32_t length = 0;
+    switch(RES_GET_TYPE(res)) {
+    case URES_ARRAY:
+        if (offset!=0) {  // empty if offset==0
+            items32 = (const Resource *)getData().pRoot+offset;
+            length = *items32++;
+        }
+        break;
+    case URES_ARRAY16:
+        items16 = getData().p16BitUnits+offset;
+        length = *items16++;
+        break;
+    default:
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+        return ResourceArray();
+    }
+    return ResourceArray(items16, items32, length, fTraceInfo);
+}
+
+ResourceTable ResourceDataValue::getTable(UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return ResourceTable();
+    }
+    const uint16_t *keys16 = NULL;
+    const int32_t *keys32 = NULL;
+    const uint16_t *items16 = NULL;
+    const Resource *items32 = NULL;
+    uint32_t offset = RES_GET_OFFSET(res);
+    int32_t length = 0;
+    switch(RES_GET_TYPE(res)) {
+    case URES_TABLE:
+        if (offset != 0) {  // empty if offset==0
+            keys16 = (const uint16_t *)(getData().pRoot+offset);
+            length = *keys16++;
+            items32 = (const Resource *)(keys16+length+(~length&1));
+        }
+        break;
+    case URES_TABLE16:
+        keys16 = getData().p16BitUnits+offset;
+        length = *keys16++;
+        items16 = keys16 + length;
+        break;
+    case URES_TABLE32:
+        if (offset != 0) {  // empty if offset==0
+            keys32 = getData().pRoot+offset;
+            length = *keys32++;
+            items32 = (const Resource *)keys32 + length;
+        }
+        break;
+    default:
+        errorCode = U_RESOURCE_TYPE_MISMATCH;
+        return ResourceTable();
+    }
+    return ResourceTable(keys16, keys32, items16, items32, length, fTraceInfo);
+}
+
+UBool ResourceDataValue::isNoInheritanceMarker() const {
+    return ::isNoInheritanceMarker(&getData(), res);
+}
+
+int32_t ResourceDataValue::getStringArray(UnicodeString *dest, int32_t capacity,
+                                          UErrorCode &errorCode) const {
+    return ::getStringArray(&getData(), getArray(errorCode), dest, capacity, errorCode);
+}
+
+int32_t ResourceDataValue::getStringArrayOrStringAsArray(UnicodeString *dest, int32_t capacity,
+                                                         UErrorCode &errorCode) const {
+    if(URES_IS_ARRAY(res)) {
+        return ::getStringArray(&getData(), getArray(errorCode), dest, capacity, errorCode);
+    }
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(dest == NULL ? capacity != 0 : capacity < 0) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    if(capacity < 1) {
+        errorCode = U_BUFFER_OVERFLOW_ERROR;
+        return 1;
+    }
+    int32_t sLength;
+    const UChar *s = res_getString(fTraceInfo, &getData(), res, &sLength);
+    if(s != NULL) {
+        dest[0].setTo(TRUE, s, sLength);
+        return 1;
+    }
+    errorCode = U_RESOURCE_TYPE_MISMATCH;
+    return 0;
+}
+
+UnicodeString ResourceDataValue::getStringOrFirstOfArray(UErrorCode &errorCode) const {
+    UnicodeString us;
+    if(U_FAILURE(errorCode)) {
+        return us;
+    }
+    int32_t sLength;
+    const UChar *s = res_getString(fTraceInfo, &getData(), res, &sLength);
+    if(s != NULL) {
+        us.setTo(TRUE, s, sLength);
+        return us;
+    }
+    ResourceArray array = getArray(errorCode);
+    if(U_FAILURE(errorCode)) {
+        return us;
+    }
+    if(array.getSize() > 0) {
+        // Tracing is already performed above (unimportant for trace that this is an array)
+        s = res_getStringNoTrace(&getData(), array.internalGetResource(&getData(), 0), &sLength);
+        if(s != NULL) {
+            us.setTo(TRUE, s, sLength);
+            return us;
+        }
+    }
+    errorCode = U_RESOURCE_TYPE_MISMATCH;
+    return us;
+}
+
+U_NAMESPACE_END
+
+static Resource
+makeResourceFrom16(const ResourceData *pResData, int32_t res16) {
+    if(res16<pResData->poolStringIndex16Limit) {
+        // Pool string, nothing to do.
+    } else {
+        // Local string, adjust the 16-bit offset to a regular one,
+        // with a larger pool string index limit.
+        res16=res16-pResData->poolStringIndex16Limit+pResData->poolStringIndexLimit;
+    }
+    return URES_MAKE_RESOURCE(URES_STRING_V2, res16);
+}
+
+U_CAPI Resource U_EXPORT2
+res_getTableItemByKey(const ResourceData *pResData, Resource table,
+                      int32_t *indexR, const char **key) {
+    uint32_t offset=RES_GET_OFFSET(table);
+    int32_t length;
+    int32_t idx;
+    if(key == NULL || *key == NULL) {
+        return RES_BOGUS;
+    }
+    switch(RES_GET_TYPE(table)) {
+    case URES_TABLE: {
+        if (offset!=0) { /* empty if offset==0 */
+            const uint16_t *p= (const uint16_t *)(pResData->pRoot+offset);
+            length=*p++;
+            *indexR=idx=_res_findTableItem(pResData, p, length, *key, key);
+            if(idx>=0) {
+                const Resource *p32=(const Resource *)(p+length+(~length&1));
+                return p32[idx];
+            }
+        }
+        break;
+    }
+    case URES_TABLE16: {
+        const uint16_t *p=pResData->p16BitUnits+offset;
+        length=*p++;
+        *indexR=idx=_res_findTableItem(pResData, p, length, *key, key);
+        if(idx>=0) {
+            return makeResourceFrom16(pResData, p[length+idx]);
+        }
+        break;
+    }
+    case URES_TABLE32: {
+        if (offset!=0) { /* empty if offset==0 */
+            const int32_t *p= pResData->pRoot+offset;
+            length=*p++;
+            *indexR=idx=_res_findTable32Item(pResData, p, length, *key, key);
+            if(idx>=0) {
+                return (Resource)p[length+idx];
+            }
+        }
+        break;
+    }
+    default:
+        break;
+    }
+    return RES_BOGUS;
+}
+
+U_CAPI Resource U_EXPORT2
+res_getTableItemByIndex(const ResourceData *pResData, Resource table,
+                        int32_t indexR, const char **key) {
+    uint32_t offset=RES_GET_OFFSET(table);
+    int32_t length;
+    if (indexR < 0) {
+        return RES_BOGUS;
+    }
+    switch(RES_GET_TYPE(table)) {
+    case URES_TABLE: {
+        if (offset != 0) { /* empty if offset==0 */
+            const uint16_t *p= (const uint16_t *)(pResData->pRoot+offset);
+            length=*p++;
+            if(indexR<length) {
+                const Resource *p32=(const Resource *)(p+length+(~length&1));
+                if(key!=NULL) {
+                    *key=RES_GET_KEY16(pResData, p[indexR]);
+                }
+                return p32[indexR];
+            }
+        }
+        break;
+    }
+    case URES_TABLE16: {
+        const uint16_t *p=pResData->p16BitUnits+offset;
+        length=*p++;
+        if(indexR<length) {
+            if(key!=NULL) {
+                *key=RES_GET_KEY16(pResData, p[indexR]);
+            }
+            return makeResourceFrom16(pResData, p[length+indexR]);
+        }
+        break;
+    }
+    case URES_TABLE32: {
+        if (offset != 0) { /* empty if offset==0 */
+            const int32_t *p= pResData->pRoot+offset;
+            length=*p++;
+            if(indexR<length) {
+                if(key!=NULL) {
+                    *key=RES_GET_KEY32(pResData, p[indexR]);
+                }
+                return (Resource)p[length+indexR];
+            }
+        }
+        break;
+    }
+    default:
+        break;
+    }
+    return RES_BOGUS;
+}
+
+U_CAPI Resource U_EXPORT2
+res_getResource(const ResourceData *pResData, const char *key) {
+    const char *realKey=key;
+    int32_t idx;
+    return res_getTableItemByKey(pResData, pResData->rootRes, &idx, &realKey);
+}
+
+
+UBool icu::ResourceTable::getKeyAndValue(int32_t i,
+                                         const char *&key, icu::ResourceValue &value) const {
+    if(0 <= i && i < length) {
+        icu::ResourceDataValue &rdValue = static_cast<icu::ResourceDataValue &>(value);
+        if (keys16 != nullptr) {
+            key = RES_GET_KEY16(&rdValue.getData(), keys16[i]);
+        } else {
+            key = RES_GET_KEY32(&rdValue.getData(), keys32[i]);
+        }
+        Resource res;
+        if (items16 != nullptr) {
+            res = makeResourceFrom16(&rdValue.getData(), items16[i]);
+        } else {
+            res = items32[i];
+        }
+        // Note: the ResourceTracer keeps a reference to the field of this
+        // ResourceTable. This is OK because the ResourceTable should remain
+        // alive for the duration that fields are being read from it
+        // (including nested fields).
+        rdValue.setResource(res, ResourceTracer(fTraceInfo, key));
+        return TRUE;
+    }
+    return FALSE;
+}
+
+UBool icu::ResourceTable::findValue(const char *key, ResourceValue &value) const {
+    icu::ResourceDataValue &rdValue = static_cast<icu::ResourceDataValue &>(value);
+    const char *realKey = nullptr;
+    int32_t i;
+    if (keys16 != nullptr) {
+        i = _res_findTableItem(&rdValue.getData(), keys16, length, key, &realKey);
+    } else {
+        i = _res_findTable32Item(&rdValue.getData(), keys32, length, key, &realKey);
+    }
+    if (i >= 0) {
+        Resource res;
+        if (items16 != nullptr) {
+            res = makeResourceFrom16(&rdValue.getData(), items16[i]);
+        } else {
+            res = items32[i];
+        }
+        // Same note about lifetime as in getKeyAndValue().
+        rdValue.setResource(res, ResourceTracer(fTraceInfo, key));
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CAPI Resource U_EXPORT2
+res_getArrayItem(const ResourceData *pResData, Resource array, int32_t indexR) {
+    uint32_t offset=RES_GET_OFFSET(array);
+    if (indexR < 0) {
+        return RES_BOGUS;
+    }
+    switch(RES_GET_TYPE(array)) {
+    case URES_ARRAY: {
+        if (offset!=0) { /* empty if offset==0 */
+            const int32_t *p= pResData->pRoot+offset;
+            if(indexR<*p) {
+                return (Resource)p[1+indexR];
+            }
+        }
+        break;
+    }
+    case URES_ARRAY16: {
+        const uint16_t *p=pResData->p16BitUnits+offset;
+        if(indexR<*p) {
+            return makeResourceFrom16(pResData, p[1+indexR]);
+        }
+        break;
+    }
+    default:
+        break;
+    }
+    return RES_BOGUS;
+}
+
+uint32_t icu::ResourceArray::internalGetResource(const ResourceData *pResData, int32_t i) const {
+    if (items16 != NULL) {
+        return makeResourceFrom16(pResData, items16[i]);
+    } else {
+        return items32[i];
+    }
+}
+
+UBool icu::ResourceArray::getValue(int32_t i, icu::ResourceValue &value) const {
+    if(0 <= i && i < length) {
+        icu::ResourceDataValue &rdValue = static_cast<icu::ResourceDataValue &>(value);
+        // Note: the ResourceTracer keeps a reference to the field of this
+        // ResourceArray. This is OK because the ResourceArray should remain
+        // alive for the duration that fields are being read from it
+        // (including nested fields).
+        rdValue.setResource(
+            internalGetResource(&rdValue.getData(), i),
+            ResourceTracer(fTraceInfo, i));
+        return TRUE;
+    }
+    return FALSE;
+}
+
+U_CFUNC Resource
+res_findResource(const ResourceData *pResData, Resource r, char** path, const char** key) {
+  char *pathP = *path, *nextSepP = *path;
+  char *closeIndex = NULL;
+  Resource t1 = r;
+  Resource t2;
+  int32_t indexR = 0;
+  UResType type = (UResType)RES_GET_TYPE(t1);
+
+  /* if you come in with an empty path, you'll be getting back the same resource */
+  if(!uprv_strlen(pathP)) {
+      return r;
+  }
+
+  /* one needs to have an aggregate resource in order to search in it */
+  if(!URES_IS_CONTAINER(type)) {
+      return RES_BOGUS;
+  }
+  
+  while(nextSepP && *pathP && t1 != RES_BOGUS && URES_IS_CONTAINER(type)) {
+    /* Iteration stops if: the path has been consumed, we found a non-existing
+     * resource (t1 == RES_BOGUS) or we found a scalar resource (including alias)
+     */
+    nextSepP = uprv_strchr(pathP, RES_PATH_SEPARATOR);
+    /* if there are more separators, terminate string 
+     * and set path to the remaining part of the string
+     */
+    if(nextSepP != NULL) {
+      if(nextSepP == pathP) {
+        // Empty key string.
+        return RES_BOGUS;
+      }
+      *nextSepP = 0; /* overwrite the separator with a NUL to terminate the key */
+      *path = nextSepP+1;
+    } else {
+      *path = uprv_strchr(pathP, 0);
+    }
+
+    /* if the resource is a table */
+    /* try the key based access */
+    if(URES_IS_TABLE(type)) {
+      *key = pathP;
+      t2 = res_getTableItemByKey(pResData, t1, &indexR, key);
+      if(t2 == RES_BOGUS) { 
+        /* if we fail to get the resource by key, maybe we got an index */
+        indexR = uprv_strtol(pathP, &closeIndex, 10);
+        if(indexR >= 0 && *closeIndex == 0 && (*pathP != '0' || closeIndex - pathP == 1)) {
+          /* if we indeed have an index, try to get the item by index */
+          t2 = res_getTableItemByIndex(pResData, t1, indexR, key);
+        } // else t2 is already RES_BOGUS
+      }
+    } else if(URES_IS_ARRAY(type)) {
+      indexR = uprv_strtol(pathP, &closeIndex, 10);
+      if(indexR >= 0 && *closeIndex == 0) {
+        t2 = res_getArrayItem(pResData, t1, indexR);
+      } else {
+        t2 = RES_BOGUS; /* have an array, but don't have a valid index */
+      }
+      *key = NULL;
+    } else { /* can't do much here, except setting t2 to bogus */
+      t2 = RES_BOGUS;
+    }
+    t1 = t2;
+    type = (UResType)RES_GET_TYPE(t1);
+    /* position pathP to next resource key/index */
+    pathP = *path;
+  }
+
+  return t1;
+}
+
+/* resource bundle swapping ------------------------------------------------- */
+
+/*
+ * Need to always enumerate the entire item tree,
+ * track the lowest address of any item to use as the limit for char keys[],
+ * track the highest address of any item to return the size of the data.
+ *
+ * We should have thought of storing those in the data...
+ * It is possible to extend the data structure by putting additional values
+ * in places that are inaccessible by ordinary enumeration of the item tree.
+ * For example, additional integers could be stored at the beginning or
+ * end of the key strings; this could be indicated by a minor version number,
+ * and the data swapping would have to know about these values.
+ *
+ * The data structure does not forbid keys to be shared, so we must swap
+ * all keys once instead of each key when it is referenced.
+ *
+ * These swapping functions assume that a resource bundle always has a length
+ * that is a multiple of 4 bytes.
+ * Currently, this is trivially true because genrb writes bundle tree leaves
+ * physically first, before their branches, so that the root table with its
+ * array of resource items (uint32_t values) is always last.
+ */
+
+/* definitions for table sorting ------------------------ */
+
+/*
+ * row of a temporary array
+ *
+ * gets platform-endian key string indexes and sorting indexes;
+ * after sorting this array by keys, the actual key/value arrays are permutated
+ * according to the sorting indexes
+ */
+typedef struct Row {
+    int32_t keyIndex, sortIndex;
+} Row;
+
+static int32_t U_CALLCONV
+ures_compareRows(const void *context, const void *left, const void *right) {
+    const char *keyChars=(const char *)context;
+    return (int32_t)uprv_strcmp(keyChars+((const Row *)left)->keyIndex,
+                                keyChars+((const Row *)right)->keyIndex);
+}
+
+typedef struct TempTable {
+    const char *keyChars;
+    Row *rows;
+    int32_t *resort;
+    uint32_t *resFlags;
+    int32_t localKeyLimit;
+    uint8_t majorFormatVersion;
+} TempTable;
+
+enum {
+    STACK_ROW_CAPACITY=200
+};
+
+/* The table item key string is not locally available. */
+static const char *const gUnknownKey="";
+
+/* resource table key for collation binaries: "%%CollationBin" */
+static const UChar gCollationBinKey[]={
+    0x25, 0x25,
+    0x43, 0x6f, 0x6c, 0x6c, 0x61, 0x74, 0x69, 0x6f, 0x6e,
+    0x42, 0x69, 0x6e,
+    0
+};
+
+/*
+ * swap one resource item
+ */
+static void
+ures_swapResource(const UDataSwapper *ds,
+                  const Resource *inBundle, Resource *outBundle,
+                  Resource res, /* caller swaps res itself */
+                  const char *key,
+                  TempTable *pTempTable,
+                  UErrorCode *pErrorCode) {
+    const Resource *p;
+    Resource *q;
+    int32_t offset, count;
+
+    switch(RES_GET_TYPE(res)) {
+    case URES_TABLE16:
+    case URES_STRING_V2:
+    case URES_INT:
+    case URES_ARRAY16:
+        /* integer, or points to 16-bit units, nothing to do here */
+        return;
+    default:
+        break;
+    }
+
+    /* all other types use an offset to point to their data */
+    offset=(int32_t)RES_GET_OFFSET(res);
+    if(offset==0) {
+        /* special offset indicating an empty item */
+        return;
+    }
+    if(pTempTable->resFlags[offset>>5]&((uint32_t)1<<(offset&0x1f))) {
+        /* we already swapped this resource item */
+        return;
+    } else {
+        /* mark it as swapped now */
+        pTempTable->resFlags[offset>>5]|=((uint32_t)1<<(offset&0x1f));
+    }
+
+    p=inBundle+offset;
+    q=outBundle+offset;
+
+    switch(RES_GET_TYPE(res)) {
+    case URES_ALIAS:
+        /* physically same value layout as string, fall through */
+        U_FALLTHROUGH;
+    case URES_STRING:
+        count=udata_readInt32(ds, (int32_t)*p);
+        /* swap length */
+        ds->swapArray32(ds, p, 4, q, pErrorCode);
+        /* swap each UChar (the terminating NUL would not change) */
+        ds->swapArray16(ds, p+1, 2*count, q+1, pErrorCode);
+        break;
+    case URES_BINARY:
+        count=udata_readInt32(ds, (int32_t)*p);
+        /* swap length */
+        ds->swapArray32(ds, p, 4, q, pErrorCode);
+        /* no need to swap or copy bytes - ures_swap() copied them all */
+
+        /* swap known formats */
+#if !UCONFIG_NO_COLLATION
+        if( key!=NULL &&  /* the binary is in a table */
+            (key!=gUnknownKey ?
+                /* its table key string is "%%CollationBin" */
+                0==ds->compareInvChars(ds, key, -1,
+                                       gCollationBinKey, UPRV_LENGTHOF(gCollationBinKey)-1) :
+                /* its table key string is unknown but it looks like a collation binary */
+                ucol_looksLikeCollationBinary(ds, p+1, count))
+        ) {
+            ucol_swap(ds, p+1, count, q+1, pErrorCode);
+        }
+#endif
+        break;
+    case URES_TABLE:
+    case URES_TABLE32:
+        {
+            const uint16_t *pKey16;
+            uint16_t *qKey16;
+
+            const int32_t *pKey32;
+            int32_t *qKey32;
+
+            Resource item;
+            int32_t i, oldIndex;
+
+            if(RES_GET_TYPE(res)==URES_TABLE) {
+                /* get table item count */
+                pKey16=(const uint16_t *)p;
+                qKey16=(uint16_t *)q;
+                count=ds->readUInt16(*pKey16);
+
+                pKey32=qKey32=NULL;
+
+                /* swap count */
+                ds->swapArray16(ds, pKey16++, 2, qKey16++, pErrorCode);
+
+                offset+=((1+count)+1)/2;
+            } else {
+                /* get table item count */
+                pKey32=(const int32_t *)p;
+                qKey32=(int32_t *)q;
+                count=udata_readInt32(ds, *pKey32);
+
+                pKey16=qKey16=NULL;
+
+                /* swap count */
+                ds->swapArray32(ds, pKey32++, 4, qKey32++, pErrorCode);
+
+                offset+=1+count;
+            }
+
+            if(count==0) {
+                break;
+            }
+
+            p=inBundle+offset; /* pointer to table resources */
+            q=outBundle+offset;
+
+            /* recurse */
+            for(i=0; i<count; ++i) {
+                const char *itemKey=gUnknownKey;
+                if(pKey16!=NULL) {
+                    int32_t keyOffset=ds->readUInt16(pKey16[i]);
+                    if(keyOffset<pTempTable->localKeyLimit) {
+                        itemKey=(const char *)outBundle+keyOffset;
+                    }
+                } else {
+                    int32_t keyOffset=udata_readInt32(ds, pKey32[i]);
+                    if(keyOffset>=0) {
+                        itemKey=(const char *)outBundle+keyOffset;
+                    }
+                }
+                item=ds->readUInt32(p[i]);
+                ures_swapResource(ds, inBundle, outBundle, item, itemKey, pTempTable, pErrorCode);
+                if(U_FAILURE(*pErrorCode)) {
+                    udata_printError(ds, "ures_swapResource(table res=%08x)[%d].recurse(%08x) failed\n",
+                                     res, i, item);
+                    return;
+                }
+            }
+
+            if(pTempTable->majorFormatVersion>1 || ds->inCharset==ds->outCharset) {
+                /* no need to sort, just swap the offset/value arrays */
+                if(pKey16!=NULL) {
+                    ds->swapArray16(ds, pKey16, count*2, qKey16, pErrorCode);
+                    ds->swapArray32(ds, p, count*4, q, pErrorCode);
+                } else {
+                    /* swap key offsets and items as one array */
+                    ds->swapArray32(ds, pKey32, count*2*4, qKey32, pErrorCode);
+                }
+                break;
+            }
+
+            /*
+             * We need to sort tables by outCharset key strings because they
+             * sort differently for different charset families.
+             * ures_swap() already set pTempTable->keyChars appropriately.
+             * First we set up a temporary table with the key indexes and
+             * sorting indexes and sort that.
+             * Then we permutate and copy/swap the actual values.
+             */
+            if(pKey16!=NULL) {
+                for(i=0; i<count; ++i) {
+                    pTempTable->rows[i].keyIndex=ds->readUInt16(pKey16[i]);
+                    pTempTable->rows[i].sortIndex=i;
+                }
+            } else {
+                for(i=0; i<count; ++i) {
+                    pTempTable->rows[i].keyIndex=udata_readInt32(ds, pKey32[i]);
+                    pTempTable->rows[i].sortIndex=i;
+                }
+            }
+            uprv_sortArray(pTempTable->rows, count, sizeof(Row),
+                           ures_compareRows, pTempTable->keyChars,
+                           FALSE, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                udata_printError(ds, "ures_swapResource(table res=%08x).uprv_sortArray(%d items) failed\n",
+                                 res, count);
+                return;
+            }
+
+            /*
+             * copy/swap/permutate items
+             *
+             * If we swap in-place, then the permutation must use another
+             * temporary array (pTempTable->resort)
+             * before the results are copied to the outBundle.
+             */
+            /* keys */
+            if(pKey16!=NULL) {
+                uint16_t *rKey16;
+
+                if(pKey16!=qKey16) {
+                    rKey16=qKey16;
+                } else {
+                    rKey16=(uint16_t *)pTempTable->resort;
+                }
+                for(i=0; i<count; ++i) {
+                    oldIndex=pTempTable->rows[i].sortIndex;
+                    ds->swapArray16(ds, pKey16+oldIndex, 2, rKey16+i, pErrorCode);
+                }
+                if(qKey16!=rKey16) {
+                    uprv_memcpy(qKey16, rKey16, 2*count);
+                }
+            } else {
+                int32_t *rKey32;
+
+                if(pKey32!=qKey32) {
+                    rKey32=qKey32;
+                } else {
+                    rKey32=pTempTable->resort;
+                }
+                for(i=0; i<count; ++i) {
+                    oldIndex=pTempTable->rows[i].sortIndex;
+                    ds->swapArray32(ds, pKey32+oldIndex, 4, rKey32+i, pErrorCode);
+                }
+                if(qKey32!=rKey32) {
+                    uprv_memcpy(qKey32, rKey32, 4*count);
+                }
+            }
+
+            /* resources */
+            {
+                Resource *r;
+
+
+                if(p!=q) {
+                    r=q;
+                } else {
+                    r=(Resource *)pTempTable->resort;
+                }
+                for(i=0; i<count; ++i) {
+                    oldIndex=pTempTable->rows[i].sortIndex;
+                    ds->swapArray32(ds, p+oldIndex, 4, r+i, pErrorCode);
+                }
+                if(q!=r) {
+                    uprv_memcpy(q, r, 4*count);
+                }
+            }
+        }
+        break;
+    case URES_ARRAY:
+        {
+            Resource item;
+            int32_t i;
+
+            count=udata_readInt32(ds, (int32_t)*p);
+            /* swap length */
+            ds->swapArray32(ds, p++, 4, q++, pErrorCode);
+
+            /* recurse */
+            for(i=0; i<count; ++i) {
+                item=ds->readUInt32(p[i]);
+                ures_swapResource(ds, inBundle, outBundle, item, NULL, pTempTable, pErrorCode);
+                if(U_FAILURE(*pErrorCode)) {
+                    udata_printError(ds, "ures_swapResource(array res=%08x)[%d].recurse(%08x) failed\n",
+                                     res, i, item);
+                    return;
+                }
+            }
+
+            /* swap items */
+            ds->swapArray32(ds, p, 4*count, q, pErrorCode);
+        }
+        break;
+    case URES_INT_VECTOR:
+        count=udata_readInt32(ds, (int32_t)*p);
+        /* swap length and each integer */
+        ds->swapArray32(ds, p, 4*(1+count), q, pErrorCode);
+        break;
+    default:
+        /* also catches RES_BOGUS */
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        break;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+ures_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    const Resource *inBundle;
+    Resource rootRes;
+    int32_t headerSize, maxTableLength;
+
+    Row rows[STACK_ROW_CAPACITY];
+    int32_t resort[STACK_ROW_CAPACITY];
+    TempTable tempTable;
+
+    const int32_t *inIndexes;
+
+    /* the following integers count Resource item offsets (4 bytes each), not bytes */
+    int32_t bundleLength, indexLength, keysBottom, keysTop, resBottom, top;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x52 &&   /* dataFormat="ResB" */
+        pInfo->dataFormat[1]==0x65 &&
+        pInfo->dataFormat[2]==0x73 &&
+        pInfo->dataFormat[3]==0x42 &&
+        /* formatVersion 1.1+ or 2.x or 3.x */
+        ((pInfo->formatVersion[0]==1 && pInfo->formatVersion[1]>=1) ||
+            pInfo->formatVersion[0]==2 || pInfo->formatVersion[0]==3)
+    )) {
+        udata_printError(ds, "ures_swap(): data format %02x.%02x.%02x.%02x (format version %02x.%02x) is not a resource bundle\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0], pInfo->formatVersion[1]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+    tempTable.majorFormatVersion=pInfo->formatVersion[0];
+
+    /* a resource bundle must contain at least one resource item */
+    if(length<0) {
+        bundleLength=-1;
+    } else {
+        bundleLength=(length-headerSize)/4;
+
+        /* formatVersion 1.1 must have a root item and at least 5 indexes */
+        if(bundleLength<(1+5)) {
+            udata_printError(ds, "ures_swap(): too few bytes (%d after header) for a resource bundle\n",
+                             length-headerSize);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    inBundle=(const Resource *)((const char *)inData+headerSize);
+    rootRes=ds->readUInt32(*inBundle);
+
+    /* formatVersion 1.1 adds the indexes[] array */
+    inIndexes=(const int32_t *)(inBundle+1);
+
+    indexLength=udata_readInt32(ds, inIndexes[URES_INDEX_LENGTH])&0xff;
+    if(indexLength<=URES_INDEX_MAX_TABLE_LENGTH) {
+        udata_printError(ds, "ures_swap(): too few indexes for a 1.1+ resource bundle\n");
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    keysBottom=1+indexLength;
+    keysTop=udata_readInt32(ds, inIndexes[URES_INDEX_KEYS_TOP]);
+    if(indexLength>URES_INDEX_16BIT_TOP) {
+        resBottom=udata_readInt32(ds, inIndexes[URES_INDEX_16BIT_TOP]);
+    } else {
+        resBottom=keysTop;
+    }
+    top=udata_readInt32(ds, inIndexes[URES_INDEX_BUNDLE_TOP]);
+    maxTableLength=udata_readInt32(ds, inIndexes[URES_INDEX_MAX_TABLE_LENGTH]);
+
+    if(0<=bundleLength && bundleLength<top) {
+        udata_printError(ds, "ures_swap(): resource top %d exceeds bundle length %d\n",
+                         top, bundleLength);
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    if(keysTop>(1+indexLength)) {
+        tempTable.localKeyLimit=keysTop<<2;
+    } else {
+        tempTable.localKeyLimit=0;
+    }
+
+    if(length>=0) {
+        Resource *outBundle=(Resource *)((char *)outData+headerSize);
+
+        /* track which resources we have already swapped */
+        uint32_t stackResFlags[STACK_ROW_CAPACITY];
+        int32_t resFlagsLength;
+
+        /*
+         * We need one bit per 4 resource bundle bytes so that we can track
+         * every possible Resource for whether we have swapped it already.
+         * Multiple Resource words can refer to the same bundle offsets
+         * for sharing identical values.
+         * We could optimize this by allocating only for locations above
+         * where Resource values are stored (above keys & strings).
+         */
+        resFlagsLength=(length+31)>>5;          /* number of bytes needed */
+        resFlagsLength=(resFlagsLength+3)&~3;   /* multiple of 4 bytes for uint32_t */
+        if(resFlagsLength<=(int32_t)sizeof(stackResFlags)) {
+            tempTable.resFlags=stackResFlags;
+        } else {
+            tempTable.resFlags=(uint32_t *)uprv_malloc(resFlagsLength);
+            if(tempTable.resFlags==NULL) {
+                udata_printError(ds, "ures_swap(): unable to allocate memory for tracking resources\n");
+                *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                return 0;
+            }
+        }
+        uprv_memset(tempTable.resFlags, 0, resFlagsLength);
+
+        /* copy the bundle for binary and inaccessible data */
+        if(inData!=outData) {
+            uprv_memcpy(outBundle, inBundle, 4*top);
+        }
+
+        /* swap the key strings, but not the padding bytes (0xaa) after the last string and its NUL */
+        udata_swapInvStringBlock(ds, inBundle+keysBottom, 4*(keysTop-keysBottom),
+                                    outBundle+keysBottom, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            udata_printError(ds, "ures_swap().udata_swapInvStringBlock(keys[%d]) failed\n", 4*(keysTop-keysBottom));
+            return 0;
+        }
+
+        /* swap the 16-bit units (strings, table16, array16) */
+        if(keysTop<resBottom) {
+            ds->swapArray16(ds, inBundle+keysTop, (resBottom-keysTop)*4, outBundle+keysTop, pErrorCode);
+            if(U_FAILURE(*pErrorCode)) {
+                udata_printError(ds, "ures_swap().swapArray16(16-bit units[%d]) failed\n", 2*(resBottom-keysTop));
+                return 0;
+            }
+        }
+
+        /* allocate the temporary table for sorting resource tables */
+        tempTable.keyChars=(const char *)outBundle; /* sort by outCharset */
+        if(tempTable.majorFormatVersion>1 || maxTableLength<=STACK_ROW_CAPACITY) {
+            tempTable.rows=rows;
+            tempTable.resort=resort;
+        } else {
+            tempTable.rows=(Row *)uprv_malloc(maxTableLength*sizeof(Row)+maxTableLength*4);
+            if(tempTable.rows==NULL) {
+                udata_printError(ds, "ures_swap(): unable to allocate memory for sorting tables (max length: %d)\n",
+                                 maxTableLength);
+                *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                if(tempTable.resFlags!=stackResFlags) {
+                    uprv_free(tempTable.resFlags);
+                }
+                return 0;
+            }
+            tempTable.resort=(int32_t *)(tempTable.rows+maxTableLength);
+        }
+
+        /* swap the resources */
+        ures_swapResource(ds, inBundle, outBundle, rootRes, NULL, &tempTable, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            udata_printError(ds, "ures_swapResource(root res=%08x) failed\n",
+                             rootRes);
+        }
+
+        if(tempTable.rows!=rows) {
+            uprv_free(tempTable.rows);
+        }
+        if(tempTable.resFlags!=stackResFlags) {
+            uprv_free(tempTable.resFlags);
+        }
+
+        /* swap the root resource and indexes */
+        ds->swapArray32(ds, inBundle, keysBottom*4, outBundle, pErrorCode);
+    }
+
+    return headerSize+4*top;
+}
diff --git a/thirdparty/icu4c/common/uresdata.h b/thirdparty/icu4c/common/uresdata.h
new file mode 100644
index 0000000000..7c2152e57b
--- /dev/null
+++ b/thirdparty/icu4c/common/uresdata.h
@@ -0,0 +1,565 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1999-2016, International Business Machines
+*                Corporation and others. All Rights Reserved.
+******************************************************************************
+*   file name:  uresdata.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999dec08
+*   created by: Markus W. Scherer
+*   06/24/02    weiv        Added support for resource sharing
+*/
+
+#ifndef __RESDATA_H__
+#define __RESDATA_H__
+
+#include "unicode/utypes.h"
+#include "unicode/udata.h"
+#include "unicode/ures.h"
+#include "putilimp.h"
+#include "udataswp.h"
+
+/**
+ * Numeric constants for internal-only types of resource items.
+ * These must use different numeric values than UResType constants
+ * because they are used together.
+ * Internal types are never returned by ures_getType().
+ */
+typedef enum {
+    /** Include a negative value so that the compiler uses the same int type as for UResType. */
+    URES_INTERNAL_NONE=-1,
+
+    /** Resource type constant for tables with 32-bit count, key offsets and values. */
+    URES_TABLE32=4,
+
+    /**
+     * Resource type constant for tables with 16-bit count, key offsets and values.
+     * All values are URES_STRING_V2 strings.
+     */
+    URES_TABLE16=5,
+
+    /** Resource type constant for 16-bit Unicode strings in formatVersion 2. */
+    URES_STRING_V2=6,
+
+    /**
+     * Resource type constant for arrays with 16-bit count and values.
+     * All values are URES_STRING_V2 strings.
+     */
+    URES_ARRAY16=9
+
+    /* Resource type 15 is not defined but effectively used by RES_BOGUS=0xffffffff. */
+} UResInternalType;
+
+/*
+ * A Resource is a 32-bit value that has 2 bit fields:
+ * 31..28   4-bit type, see enum below
+ * 27..0    28-bit four-byte-offset or value according to the type
+ */
+typedef uint32_t Resource;
+
+#define RES_BOGUS 0xffffffff
+#define RES_MAX_OFFSET 0x0fffffff
+
+#define RES_GET_TYPE(res) ((int32_t)((res)>>28UL))
+#define RES_GET_OFFSET(res) ((res)&0x0fffffff)
+#define RES_GET_POINTER(pRoot, res) ((pRoot)+RES_GET_OFFSET(res))
+
+/* get signed and unsigned integer values directly from the Resource handle
+ * NOTE: For proper logging, please use the res_getInt() constexpr
+ */
+#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
+#   define RES_GET_INT_NO_TRACE(res) (((int32_t)((res)<<4L))>>4L)
+#else
+#   define RES_GET_INT_NO_TRACE(res) (int32_t)(((res)&0x08000000) ? (res)|0xf0000000 : (res)&0x07ffffff)
+#endif
+
+#define RES_GET_UINT_NO_TRACE(res) ((res)&0x0fffffff)
+
+#define URES_IS_ARRAY(type) ((int32_t)(type)==URES_ARRAY || (int32_t)(type)==URES_ARRAY16)
+#define URES_IS_TABLE(type) ((int32_t)(type)==URES_TABLE || (int32_t)(type)==URES_TABLE16 || (int32_t)(type)==URES_TABLE32)
+#define URES_IS_CONTAINER(type) (URES_IS_TABLE(type) || URES_IS_ARRAY(type))
+
+#define URES_MAKE_RESOURCE(type, offset) (((Resource)(type)<<28)|(Resource)(offset))
+#define URES_MAKE_EMPTY_RESOURCE(type) ((Resource)(type)<<28)
+
+/* indexes[] value names; indexes are generally 32-bit (Resource) indexes */
+enum {
+    /**
+     * [0] contains the length of indexes[]
+     * which is at most URES_INDEX_TOP of the latest format version
+     *
+     * formatVersion==1: all bits contain the length of indexes[]
+     *   but the length is much less than 0xff;
+     * formatVersion>1:
+     *   only bits  7..0 contain the length of indexes[],
+     *        bits 31..8 are reserved and set to 0
+     * formatVersion>=3:
+     *        bits 31..8 poolStringIndexLimit bits 23..0
+     */
+    URES_INDEX_LENGTH,
+    /**
+     * [1] contains the top of the key strings,
+     *     same as the bottom of resources or UTF-16 strings, rounded up
+     */
+    URES_INDEX_KEYS_TOP,
+    /** [2] contains the top of all resources */
+    URES_INDEX_RESOURCES_TOP,
+    /**
+     * [3] contains the top of the bundle,
+     *     in case it were ever different from [2]
+     */
+    URES_INDEX_BUNDLE_TOP,
+    /** [4] max. length of any table */
+    URES_INDEX_MAX_TABLE_LENGTH,
+    /**
+     * [5] attributes bit set, see URES_ATT_* (new in formatVersion 1.2)
+     *
+     * formatVersion>=3:
+     *   bits 31..16 poolStringIndex16Limit
+     *   bits 15..12 poolStringIndexLimit bits 27..24
+     */
+    URES_INDEX_ATTRIBUTES,
+    /**
+     * [6] top of the 16-bit units (UTF-16 string v2 UChars, URES_TABLE16, URES_ARRAY16),
+     *     rounded up (new in formatVersion 2.0, ICU 4.4)
+     */
+    URES_INDEX_16BIT_TOP,
+    /** [7] checksum of the pool bundle (new in formatVersion 2.0, ICU 4.4) */
+    URES_INDEX_POOL_CHECKSUM,
+    URES_INDEX_TOP
+};
+
+/*
+ * Nofallback attribute, attribute bit 0 in indexes[URES_INDEX_ATTRIBUTES].
+ * New in formatVersion 1.2 (ICU 3.6).
+ *
+ * If set, then this resource bundle is a standalone bundle.
+ * If not set, then the bundle participates in locale fallback, eventually
+ * all the way to the root bundle.
+ * If indexes[] is missing or too short, then the attribute cannot be determined
+ * reliably. Dependency checking should ignore such bundles, and loading should
+ * use fallbacks.
+ */
+#define URES_ATT_NO_FALLBACK 1
+
+/*
+ * Attributes for bundles that are, or use, a pool bundle.
+ * A pool bundle provides key strings that are shared among several other bundles
+ * to reduce their total size.
+ * New in formatVersion 2 (ICU 4.4).
+ */
+#define URES_ATT_IS_POOL_BUNDLE 2
+#define URES_ATT_USES_POOL_BUNDLE 4
+
+/*
+ * File format for .res resource bundle files
+ *
+ * ICU 56: New in formatVersion 3 compared with 2: -------------
+ *
+ * Resource bundles can optionally use shared string-v2 values
+ * stored in the pool bundle.
+ * If so, then the indexes[] contain two new values
+ * in previously-unused bits of existing indexes[] slots:
+ * - poolStringIndexLimit:
+ *     String-v2 offsets (in 32-bit Resource words) below this limit
+ *     point to pool bundle string-v2 values.
+ * - poolStringIndex16Limit:
+ *     Resource16 string-v2 offsets below this limit
+ *     point to pool bundle string-v2 values.
+ * Guarantee: poolStringIndex16Limit <= poolStringIndexLimit
+ *
+ * The local bundle's poolStringIndexLimit is greater than
+ * any pool bundle string index used in the local bundle.
+ * The poolStringIndexLimit should not be greater than
+ * the maximum possible pool bundle string index.
+ *
+ * The maximum possible pool bundle string index is the index to the last non-NUL
+ * pool string character, due to suffix sharing.
+ *
+ * In the pool bundle, there is no structure that lists the strings.
+ * (The root resource is an empty Table.)
+ * If the strings need to be enumerated (as genrb --usePoolBundle does),
+ * then iterate through the pool bundle's 16-bit-units array from the beginning.
+ * Stop at the end of the array, or when an explicit or implicit string length
+ * would lead beyond the end of the array,
+ * or when an apparent string is not NUL-terminated.
+ * (Future genrb version might terminate the strings with
+ * what looks like a large explicit string length.)
+ *
+ * ICU 4.4: New in formatVersion 2 compared with 1.3: -------------
+ *
+ * Three new resource types -- String-v2, Table16 and Array16 -- have their
+ * values stored in a new array of 16-bit units between the table key strings
+ * and the start of the other resources.
+ *
+ * genrb eliminates duplicates among Unicode string-v2 values.
+ * Multiple Unicode strings may use the same offset and string data,
+ * or a short string may point to the suffix of a longer string. ("Suffix sharing")
+ * For example, one string "abc" may be reused for another string "bc" by pointing
+ * to the second character. (Short strings-v2 are NUL-terminated
+ * and not preceded by an explicit length value.)
+ *
+ * It is allowed for all resource types to share values.
+ * The swapper code (ures_swap()) has been modified so that it swaps each item
+ * exactly once.
+ *
+ * A resource bundle may use a special pool bundle. Some or all of the table key strings
+ * of the using-bundle are omitted, and the key string offsets for such key strings refer
+ * to offsets in the pool bundle.
+ * The using-bundle's and the pool-bundle's indexes[URES_INDEX_POOL_CHECKSUM] values
+ * must match.
+ * Two bits in indexes[URES_INDEX_ATTRIBUTES] indicate whether a resource bundle
+ * is or uses a pool bundle.
+ *
+ * Table key strings must be compared in ASCII order, even if they are not
+ * stored in ASCII.
+ *
+ * New in formatVersion 1.3 compared with 1.2: -------------
+ *
+ * genrb eliminates duplicates among key strings.
+ * Multiple table items may share one key string, or one item may point
+ * to the suffix of another's key string. ("Suffix sharing")
+ * For example, one key "abc" may be reused for another key "bc" by pointing
+ * to the second character. (Key strings are NUL-terminated.)
+ *
+ * -------------
+ *
+ * An ICU4C resource bundle file (.res) is a binary, memory-mappable file
+ * with nested, hierarchical data structures.
+ * It physically contains the following:
+ *
+ *   Resource root; -- 32-bit Resource item, root item for this bundle's tree;
+ *                     currently, the root item must be a table or table32 resource item
+ *   int32_t indexes[indexes[0]]; -- array of indexes for friendly
+ *                                   reading and swapping; see URES_INDEX_* above
+ *                                   new in formatVersion 1.1 (ICU 2.8)
+ *   char keys[]; -- characters for key strings
+ *                   (formatVersion 1.0: up to 65k of characters; 1.1: <2G)
+ *                   (minus the space for root and indexes[]),
+ *                   which consist of invariant characters (ASCII/EBCDIC) and are NUL-terminated;
+ *                   padded to multiple of 4 bytes for 4-alignment of the following data
+ *   uint16_t 16BitUnits[]; -- resources that are stored entirely as sequences of 16-bit units
+ *                             (new in formatVersion 2/ICU 4.4)
+ *                             data is indexed by the offset values in 16-bit resource types,
+ *                             with offset 0 pointing to the beginning of this array;
+ *                             there is a 0 at offset 0, for empty resources;
+ *                             padded to multiple of 4 bytes for 4-alignment of the following data
+ *   data; -- data directly and indirectly indexed by the root item;
+ *            the structure is determined by walking the tree
+ *
+ * Each resource bundle item has a 32-bit Resource handle (see typedef above)
+ * which contains the item type number in its upper 4 bits (31..28) and either
+ * an offset or a direct value in its lower 28 bits (27..0).
+ * The order of items is undefined and only determined by walking the tree.
+ * Leaves of the tree may be stored first or last or anywhere in between,
+ * and it is in theory possible to have unreferenced holes in the file.
+ *
+ * 16-bit-unit values:
+ * Starting with formatVersion 2/ICU 4.4, some resources are stored in a special
+ * array of 16-bit units. Each resource value is a sequence of 16-bit units,
+ * with no per-resource padding to a 4-byte boundary.
+ * 16-bit container types (Table16 and Array16) contain Resource16 values
+ * which are offsets to String-v2 resources in the same 16-bit-units array.
+ *
+ * Direct values:
+ * - Empty Unicode strings have an offset value of 0 in the Resource handle itself.
+ * - Starting with formatVersion 2/ICU 4.4, an offset value of 0 for
+ *   _any_ resource type indicates an empty value.
+ * - Integer values are 28-bit values stored in the Resource handle itself;
+ *   the interpretation of unsigned vs. signed integers is up to the application.
+ *
+ * All other types and values use 28-bit offsets to point to the item's data.
+ * The offset is an index to the first 32-bit word of the value, relative to the
+ * start of the resource data (i.e., the root item handle is at offset 0).
+ * To get byte offsets, the offset is multiplied by 4 (or shifted left by 2 bits).
+ * All resource item values are 4-aligned.
+ *
+ * New in formatVersion 2/ICU 4.4: Some types use offsets into the 16-bit-units array,
+ * indexing 16-bit units in that array.
+ *
+ * The structures (memory layouts) for the values for each item type are listed
+ * in the table below.
+ *
+ * Nested, hierarchical structures: -------------
+ *
+ * Table items contain key-value pairs where the keys are offsets to char * key strings.
+ * The values of these pairs are either Resource handles or
+ * offsets into the 16-bit-units array, depending on the table type.
+ *
+ * Array items are simple vectors of Resource handles,
+ * or of offsets into the 16-bit-units array, depending on the array type.
+ *
+ * Table key string offsets: -------
+ *
+ * Key string offsets are relative to the start of the resource data (of the root handle),
+ * i.e., the first string has an offset of 4+sizeof(indexes).
+ * (After the 4-byte root handle and after the indexes array.)
+ *
+ * If the resource bundle uses a pool bundle, then some key strings are stored
+ * in the pool bundle rather than in the local bundle itself.
+ * - In a Table or Table16, the 16-bit key string offset is local if it is
+ *   less than indexes[URES_INDEX_KEYS_TOP]<<2.
+ *   Otherwise, subtract indexes[URES_INDEX_KEYS_TOP]<<2 to get the offset into
+ *   the pool bundle key strings.
+ * - In a Table32, the 32-bit key string offset is local if it is non-negative.
+ *   Otherwise, reset bit 31 to get the pool key string offset.
+ *
+ * Unlike the local offset, the pool key offset is relative to
+ * the start of the key strings, not to the start of the bundle.
+ *
+ * An alias item is special (and new in ICU 2.4): --------------
+ *
+ * Its memory layout is just like for a UnicodeString, but at runtime it resolves to
+ * another resource bundle's item according to the path in the string.
+ * This is used to share items across bundles that are in different lookup/fallback
+ * chains (e.g., large collation data among zh_TW and zh_HK).
+ * This saves space (for large items) and maintenance effort (less duplication of data).
+ *
+ * --------------------------------------------------------------------------
+ *
+ * Resource types:
+ *
+ * Most resources have their values stored at four-byte offsets from the start
+ * of the resource data. These values are at least 4-aligned.
+ * Some resource values are stored directly in the offset field of the Resource itself.
+ * See UResType in unicode/ures.h for enumeration constants for Resource types.
+ *
+ * Some resources have their values stored as sequences of 16-bit units,
+ * at 2-byte offsets from the start of a contiguous 16-bit-unit array between
+ * the table key strings and the other resources. (new in formatVersion 2/ICU 4.4)
+ * At offset 0 of that array is a 16-bit zero value for empty 16-bit resources.
+ *
+ * Resource16 values in Table16 and Array16 are 16-bit offsets to String-v2
+ * resources, with the offsets relative to the start of the 16-bit-units array.
+ * Starting with formatVersion 3/ICU 56, if offset<poolStringIndex16Limit
+ * then use the pool bundle's 16-bit-units array,
+ * otherwise subtract that limit and use the local 16-bit-units array.
+ *
+ * Type Name            Memory layout of values
+ *                      (in parentheses: scalar, non-offset values)
+ *
+ * 0  Unicode String:   int32_t length, UChar[length], (UChar)0, (padding)
+ *                  or  (empty string ("") if offset==0)
+ * 1  Binary:           int32_t length, uint8_t[length], (padding)
+ *                      - the start of the bytes is 16-aligned -
+ * 2  Table:            uint16_t count, uint16_t keyStringOffsets[count], (uint16_t padding), Resource[count]
+ * 3  Alias:            (physically same value layout as string, new in ICU 2.4)
+ * 4  Table32:          int32_t count, int32_t keyStringOffsets[count], Resource[count]
+ *                      (new in formatVersion 1.1/ICU 2.8)
+ * 5  Table16:          uint16_t count, uint16_t keyStringOffsets[count], Resource16[count]
+ *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
+ * 6  Unicode String-v2:UChar[length], (UChar)0; length determined by the first UChar:
+ *                      - if first is not a trail surrogate, then the length is implicit
+ *                        and u_strlen() needs to be called
+ *                      - if first<0xdfef then length=first&0x3ff (and skip first)
+ *                      - if first<0xdfff then length=((first-0xdfef)<<16) | second UChar
+ *                      - if first==0xdfff then length=((second UChar)<<16) | third UChar
+ *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
+ *
+ *                      Starting with formatVersion 3/ICU 56, if offset<poolStringIndexLimit
+ *                      then use the pool bundle's 16-bit-units array,
+ *                      otherwise subtract that limit and use the local 16-bit-units array.
+ *                      (Note different limits for Resource16 vs. Resource.)
+ *
+ * 7  Integer:          (28-bit offset is integer value)
+ * 8  Array:            int32_t count, Resource[count]
+ * 9  Array16:          uint16_t count, Resource16[count]
+ *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
+ * 14 Integer Vector:   int32_t length, int32_t[length]
+ * 15 Reserved:         This value denotes special purpose resources and is for internal use.
+ *
+ * Note that there are 3 types with data vector values:
+ * - Vectors of 8-bit bytes stored as type Binary.
+ * - Vectors of 16-bit words stored as type Unicode String or Unicode String-v2
+ *                     (no value restrictions, all values 0..ffff allowed!).
+ * - Vectors of 32-bit words stored as type Integer Vector.
+ */
+
+/*
+ * Structure for a single, memory-mapped ResourceBundle.
+ */
+typedef struct ResourceData {
+    UDataMemory *data;
+    const int32_t *pRoot;
+    const uint16_t *p16BitUnits;
+    const char *poolBundleKeys;
+    Resource rootRes;
+    int32_t localKeyLimit;
+    const uint16_t *poolBundleStrings;
+    int32_t poolStringIndexLimit;
+    int32_t poolStringIndex16Limit;
+    UBool noFallback; /* see URES_ATT_NO_FALLBACK */
+    UBool isPoolBundle;
+    UBool usesPoolBundle;
+    UBool useNativeStrcmp;
+} ResourceData;
+
+/*
+ * Read a resource bundle from memory.
+ */
+U_CAPI void U_EXPORT2
+res_read(ResourceData *pResData,
+         const UDataInfo *pInfo, const void *inBytes, int32_t length,
+         UErrorCode *errorCode);
+
+/*
+ * Load a resource bundle file.
+ * The ResourceData structure must be allocated externally.
+ */
+U_CFUNC void
+res_load(ResourceData *pResData,
+         const char *path, const char *name, UErrorCode *errorCode);
+
+/*
+ * Release a resource bundle file.
+ * This does not release the ResourceData structure itself.
+ */
+U_CFUNC void
+res_unload(ResourceData *pResData);
+
+U_CAPI UResType U_EXPORT2
+res_getPublicType(Resource res);
+
+///////////////////////////////////////////////////////////////////////////
+// To enable tracing, use the inline versions of the res_get* functions. //
+///////////////////////////////////////////////////////////////////////////
+
+/*
+ * Return a pointer to a zero-terminated, const UChar* string
+ * and set its length in *pLength.
+ * Returns NULL if not found.
+ */
+U_CAPI const UChar * U_EXPORT2
+res_getStringNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength);
+
+U_CAPI const uint8_t * U_EXPORT2
+res_getBinaryNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength);
+
+U_CAPI const int32_t * U_EXPORT2
+res_getIntVectorNoTrace(const ResourceData *pResData, Resource res, int32_t *pLength);
+
+U_CAPI const UChar * U_EXPORT2
+res_getAlias(const ResourceData *pResData, Resource res, int32_t *pLength);
+
+U_CAPI Resource U_EXPORT2
+res_getResource(const ResourceData *pResData, const char *key);
+
+U_CAPI int32_t U_EXPORT2
+res_countArrayItems(const ResourceData *pResData, Resource res);
+
+U_CAPI Resource U_EXPORT2
+res_getArrayItem(const ResourceData *pResData, Resource array, int32_t indexS);
+
+U_CAPI Resource U_EXPORT2
+res_getTableItemByIndex(const ResourceData *pResData, Resource table, int32_t indexS, const char ** key);
+
+U_CAPI Resource U_EXPORT2
+res_getTableItemByKey(const ResourceData *pResData, Resource table, int32_t *indexS, const char* * key);
+
+/**
+ * Iterates over the path and stops when a scalar resource is found.
+ * Follows aliases.
+ * Modifies the contents of *path (replacing separators with NULs),
+ * and also moves *path forward while it finds items.
+ *
+ * @param path input: "CollationElements/Sequence" or "zoneStrings/3/2" etc.;
+ *             output: points to the part that has not yet been processed
+ */
+U_CFUNC Resource res_findResource(const ResourceData *pResData, Resource r,
+                                  char** path, const char** key);
+
+#ifdef __cplusplus
+
+#include "resource.h"
+#include "restrace.h"
+
+U_NAMESPACE_BEGIN
+
+inline const UChar* res_getString(const ResourceTracer& traceInfo,
+        const ResourceData *pResData, Resource res, int32_t *pLength) {
+    traceInfo.trace("string");
+    return res_getStringNoTrace(pResData, res, pLength);
+}
+
+inline const uint8_t* res_getBinary(const ResourceTracer& traceInfo,
+        const ResourceData *pResData, Resource res, int32_t *pLength) {
+    traceInfo.trace("binary");
+    return res_getBinaryNoTrace(pResData, res, pLength);
+}
+
+inline const int32_t* res_getIntVector(const ResourceTracer& traceInfo,
+        const ResourceData *pResData, Resource res, int32_t *pLength) {
+    traceInfo.trace("intvector");
+    return res_getIntVectorNoTrace(pResData, res, pLength);
+}
+
+inline int32_t res_getInt(const ResourceTracer& traceInfo, Resource res) {
+    traceInfo.trace("int");
+    return RES_GET_INT_NO_TRACE(res);
+}
+
+inline uint32_t res_getUInt(const ResourceTracer& traceInfo, Resource res) {
+    traceInfo.trace("uint");
+    return RES_GET_UINT_NO_TRACE(res);
+}
+
+class ResourceDataValue : public ResourceValue {
+public:
+    ResourceDataValue() :
+        res(static_cast<Resource>(URES_NONE)),
+        fTraceInfo() {}
+    virtual ~ResourceDataValue();
+
+    void setData(const ResourceData *data) {
+        resData = *data;
+    }
+
+    void setResource(Resource r, ResourceTracer&& traceInfo) {
+        res = r;
+        fTraceInfo = traceInfo;
+    }
+
+    const ResourceData &getData() const { return resData; }
+    virtual UResType getType() const;
+    virtual const UChar *getString(int32_t &length, UErrorCode &errorCode) const;
+    virtual const UChar *getAliasString(int32_t &length, UErrorCode &errorCode) const;
+    virtual int32_t getInt(UErrorCode &errorCode) const;
+    virtual uint32_t getUInt(UErrorCode &errorCode) const;
+    virtual const int32_t *getIntVector(int32_t &length, UErrorCode &errorCode) const;
+    virtual const uint8_t *getBinary(int32_t &length, UErrorCode &errorCode) const;
+    virtual ResourceArray getArray(UErrorCode &errorCode) const;
+    virtual ResourceTable getTable(UErrorCode &errorCode) const;
+    virtual UBool isNoInheritanceMarker() const;
+    virtual int32_t getStringArray(UnicodeString *dest, int32_t capacity,
+                                   UErrorCode &errorCode) const;
+    virtual int32_t getStringArrayOrStringAsArray(UnicodeString *dest, int32_t capacity,
+                                                  UErrorCode &errorCode) const;
+    virtual UnicodeString getStringOrFirstOfArray(UErrorCode &errorCode) const;
+
+private:
+    // TODO(ICU-20769): If UResourceBundle.fResData becomes a pointer,
+    // then remove this value field again and just store a pResData pointer.
+    ResourceData resData;
+    Resource res;
+    ResourceTracer fTraceInfo;
+};
+
+U_NAMESPACE_END
+
+#endif  /* __cplusplus */
+
+/**
+ * Swap an ICU resource bundle. See udataswp.h.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+ures_swap(const UDataSwapper *ds,
+          const void *inData, int32_t length, void *outData,
+          UErrorCode *pErrorCode);
+
+#endif
diff --git a/thirdparty/icu4c/common/uresimp.h b/thirdparty/icu4c/common/uresimp.h
new file mode 100644
index 0000000000..69d82566fe
--- /dev/null
+++ b/thirdparty/icu4c/common/uresimp.h
@@ -0,0 +1,364 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef URESIMP_H
+#define URESIMP_H
+
+#include "unicode/ures.h"
+#include "unicode/utypes.h"
+
+#include "uresdata.h"
+
+#define kRootLocaleName         "root"
+#define kPoolBundleName         "pool"
+
+/*
+ The default minor version and the version separator must be exactly one
+ character long.
+*/
+
+#define kDefaultMinorVersion    "0"
+#define kVersionSeparator       "."
+#define kVersionTag             "Version"
+
+#define MAGIC1 19700503
+#define MAGIC2 19641227
+
+#define URES_MAX_ALIAS_LEVEL 256
+#define URES_MAX_BUFFER_SIZE 256
+
+#define EMPTY_SET 0x2205
+
+struct UResourceDataEntry;
+typedef struct UResourceDataEntry UResourceDataEntry;
+
+/*
+ * Note: If we wanted to make this structure smaller, then we could try
+ * to use one UResourceDataEntry pointer for fAlias and fPool, with a separate
+ * flag to distinguish whether this struct is for a real bundle with a pool,
+ * or for an alias entry for which we won't use the pool after loading.
+ */
+struct UResourceDataEntry {
+    char *fName; /* name of the locale for bundle - still to decide whether it is original or fallback */
+    char *fPath; /* path to bundle - used for distinguishing between resources with the same name */
+    UResourceDataEntry *fParent; /*next resource in fallback chain*/
+    UResourceDataEntry *fAlias;
+    UResourceDataEntry *fPool;
+    ResourceData fData; /* data for low level access */
+    char fNameBuffer[3]; /* A small buffer of free space for fName. The free space is due to struct padding. */
+    uint32_t fCountExisting; /* how much is this resource used */
+    UErrorCode fBogus;
+    /* int32_t fHashKey;*/ /* for faster access in the hashtable */
+};
+
+#define RES_BUFSIZE 64
+#define RES_PATH_SEPARATOR   '/'
+#define RES_PATH_SEPARATOR_S   "/"
+
+struct UResourceBundle {
+    const char *fKey; /*tag*/
+    UResourceDataEntry *fData; /*for low-level access*/
+    char *fVersion;
+    UResourceDataEntry *fTopLevelData; /* for getting the valid locale */
+    char *fResPath; /* full path to the resource: "zh_TW/CollationElements/Sequence" */
+    // TODO(ICU-20769): Try to change the by-value fResData into a pointer,
+    // with the struct in only one place for each bundle.
+    // Also replace class ResourceDataValue.resData with a pResData pointer again.
+    ResourceData fResData;
+    char fResBuf[RES_BUFSIZE];
+    int32_t fResPathLen;
+    Resource fRes;
+    UBool fHasFallback;
+    UBool fIsTopLevel;
+    uint32_t fMagic1;   /* For determining if it's a stack object */
+    uint32_t fMagic2;   /* For determining if it's a stack object */
+    int32_t fIndex;
+    int32_t fSize;
+
+    /*const UResourceBundle *fParentRes;*/ /* needed to get the actual locale for a child resource */
+};
+
+U_CAPI void U_EXPORT2 ures_initStackObject(UResourceBundle* resB);
+
+#ifdef __cplusplus
+
+U_NAMESPACE_BEGIN
+
+/**
+ * \class StackUResourceBundle
+ * "Smart pointer" like class, closes a UResourceBundle via ures_close().
+ *
+ * This code:
+ *
+ *   StackUResourceBundle bundle;
+ *   foo(bundle.getAlias());
+ *
+ * Is equivalent to this code:
+ *
+ *   UResourceBundle bundle;
+ *   ures_initStackObject(&bundle);
+ *   foo(&bundle);
+ *   ures_close(&bundle);
+ *
+ * @see LocalUResourceBundlePointer
+ * @internal
+ */
+class U_COMMON_API StackUResourceBundle {
+public:
+    // No heap allocation. Use only on the stack.
+    static void* U_EXPORT2 operator new(size_t) U_NOEXCEPT = delete;
+    static void* U_EXPORT2 operator new[](size_t) U_NOEXCEPT = delete;
+#if U_HAVE_PLACEMENT_NEW
+    static void* U_EXPORT2 operator new(size_t, void*) U_NOEXCEPT = delete;
+#endif
+
+    StackUResourceBundle();
+    ~StackUResourceBundle();
+
+    UResourceBundle* getAlias() { return &bundle; }
+
+    UResourceBundle& ref() { return bundle; }
+    const UResourceBundle& ref() const { return bundle; }
+
+    StackUResourceBundle(const StackUResourceBundle&) = delete;
+    StackUResourceBundle& operator=(const StackUResourceBundle&) = delete;
+
+    StackUResourceBundle(StackUResourceBundle&&) = delete;
+    StackUResourceBundle& operator=(StackUResourceBundle&&) = delete;
+
+private:
+    UResourceBundle bundle;
+};
+
+U_NAMESPACE_END
+
+#endif  /* __cplusplus */
+
+/**
+ * Opens a resource bundle for the locale;
+ * if there is not even a base language bundle, then loads the root bundle;
+ * never falls back to the default locale.
+ *
+ * This is used for algorithms that have good pan-Unicode default behavior,
+ * such as case mappings, collation, and segmentation (BreakIterator).
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_openNoDefault(const char* path, const char* localeID, UErrorCode* status);
+
+/* Some getters used by the copy constructor */
+U_CFUNC const char* ures_getName(const UResourceBundle* resB);
+#ifdef URES_DEBUG
+U_CFUNC const char* ures_getPath(const UResourceBundle* resB);
+/**
+ * If anything was in the RB cache, dump it to the screen.
+ * @return true if there was anything into the cache
+ */
+U_CAPI UBool U_EXPORT2 ures_dumpCacheContents(void);
+#endif
+/*U_CFUNC void ures_appendResPath(UResourceBundle *resB, const char* toAdd, int32_t lenToAdd);*/
+/*U_CFUNC void ures_setResPath(UResourceBundle *resB, const char* toAdd);*/
+/*U_CFUNC void ures_freeResPath(UResourceBundle *resB);*/
+
+/* Candidates for export */
+U_CFUNC UResourceBundle *ures_copyResb(UResourceBundle *r, const UResourceBundle *original, UErrorCode *status);
+
+/**
+ * Returns a resource that can be located using the pathToResource argument. One needs optional package, locale
+ * and path inside the locale, for example: "/myData/en/zoneStrings/3". Keys and indexes are supported. Keys
+ * need to reference data in named structures, while indexes can reference both named and anonymous resources.
+ * Features a fill-in parameter. 
+ * 
+ * Note, this function does NOT have a syntax for specifying items within a tree.  May want to consider a
+ * syntax that delineates between package/tree and resource.  
+ *
+ * @param pathToResource    a path that will lead to the requested resource
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be deleted by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status            fills in the outgoing error code.
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must delete it
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_findResource(const char* pathToResource, 
+                  UResourceBundle *fillIn, UErrorCode *status); 
+
+/**
+ * Returns a sub resource that can be located using the pathToResource argument. One needs a path inside 
+ * the supplied resource, for example, if you have "en_US" resource bundle opened, you might ask for
+ * "zoneStrings/3". Keys and indexes are supported. Keys
+ * need to reference data in named structures, while indexes can reference both 
+ * named and anonymous resources.
+ * Features a fill-in parameter. 
+ *
+ * @param resourceBundle    a resource
+ * @param pathToResource    a path that will lead to the requested resource
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be deleted by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status            fills in the outgoing error code.
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must delete it
+ */
+U_CAPI UResourceBundle* U_EXPORT2
+ures_findSubResource(const UResourceBundle *resB, 
+                     char* pathToResource, 
+                     UResourceBundle *fillIn, UErrorCode *status);
+
+/**
+ * Returns a functionally equivalent locale (considering keywords) for the specified keyword.
+ * @param result fillin for the equivalent locale
+ * @param resultCapacity capacity of the fillin buffer
+ * @param path path to the tree, or NULL for ICU data
+ * @param resName top level resource. Example: "collations"
+ * @param keyword locale keyword. Example: "collation"
+ * @param locid The requested locale
+ * @param isAvailable If non-null, pointer to fillin parameter that indicates whether the 
+ * requested locale was available. The locale is defined as 'available' if it physically 
+ * exists within the specified tree.
+ * @param omitDefault if true, omit keyword and value if default. 'de_DE\@collation=standard' -> 'de_DE'
+ * @param status error code
+ * @return  the actual buffer size needed for the full locale.  If it's greater 
+ * than resultCapacity, the returned full name will be truncated and an error code will be returned.
+ */
+U_CAPI int32_t U_EXPORT2
+ures_getFunctionalEquivalent(char *result, int32_t resultCapacity, 
+                             const char *path, const char *resName, const char *keyword, const char *locid,
+                             UBool *isAvailable, UBool omitDefault, UErrorCode *status);
+
+/**
+ * Given a tree path and keyword, return a string enumeration of all possible values for that keyword.
+ * @param path path to the tree, or NULL for ICU data
+ * @param keyword a particular keyword to consider, must match a top level resource name 
+ * within the tree.
+ * @param status error code
+ */
+U_CAPI UEnumeration* U_EXPORT2
+ures_getKeywordValues(const char *path, const char *keyword, UErrorCode *status);
+
+
+/**
+ * Get a resource with multi-level fallback. Normally only the top level resources will
+ * fallback to its parent. This performs fallback on subresources. For example, when a table
+ * is defined in a resource bundle and a parent resource bundle, normally no fallback occurs
+ * on the sub-resources because the table is defined in the current resource bundle, but this
+ * function can perform fallback on the sub-resources of the table.
+ * @param resB              a resource
+ * @param inKey             a key associated with the requested resource
+ * @param fillIn            if NULL a new UResourceBundle struct is allocated and must be deleted by the caller.
+ *                          Alternatively, you can supply a struct to be filled by this function.
+ * @param status: fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error 
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must delete it
+ */
+U_CAPI UResourceBundle* U_EXPORT2 
+ures_getByKeyWithFallback(const UResourceBundle *resB, 
+                          const char* inKey, 
+                          UResourceBundle *fillIn, 
+                          UErrorCode *status);
+
+
+/**
+ * Get a String with multi-level fallback. Normally only the top level resources will
+ * fallback to its parent. This performs fallback on subresources. For example, when a table
+ * is defined in a resource bundle and a parent resource bundle, normally no fallback occurs
+ * on the sub-resources because the table is defined in the current resource bundle, but this
+ * function can perform fallback on the sub-resources of the table.
+ * @param resB              a resource
+ * @param inKey             a key associated with the requested resource
+ * @param status: fills in the outgoing error code
+ *                could be <TT>U_MISSING_RESOURCE_ERROR</TT> if the key is not found
+ *                could be a non-failing error 
+ *                e.g.: <TT>U_USING_FALLBACK_WARNING</TT>,<TT>U_USING_DEFAULT_WARNING </TT>
+ * @return                  a pointer to a UResourceBundle struct. If fill in param was NULL, caller must delete it
+ */
+U_CAPI const UChar* U_EXPORT2 
+ures_getStringByKeyWithFallback(const UResourceBundle *resB, 
+                          const char* inKey,  
+                          int32_t* len,
+                          UErrorCode *status);
+
+#ifdef __cplusplus
+
+U_CAPI void U_EXPORT2
+ures_getValueWithFallback(const UResourceBundle *bundle, const char *path,
+                          UResourceBundle *tempFillIn,
+                          icu::ResourceDataValue &value, UErrorCode &errorCode);
+
+U_CAPI void U_EXPORT2
+ures_getAllItemsWithFallback(const UResourceBundle *bundle, const char *path,
+                             icu::ResourceSink &sink, UErrorCode &errorCode);
+
+#endif  /* __cplusplus */
+
+/**
+ * Get a version number by key
+ * @param resB bundle containing version number
+ * @param key the key for the version number
+ * @param ver fillin for the version number
+ * @param status error code
+ */
+U_CAPI void U_EXPORT2
+ures_getVersionByKey(const UResourceBundle *resB,
+                     const char *key,
+                     UVersionInfo ver,
+                     UErrorCode *status);
+
+
+/**
+ * Internal function.
+ * Return the version number associated with this ResourceBundle as a string.
+ *
+ * @param resourceBundle The resource bundle for which the version is checked.
+ * @return  A version number string as specified in the resource bundle or its parent.
+ *          The caller does not own this string.
+ * @see ures_getVersion
+ */
+U_CAPI const char* U_EXPORT2 
+ures_getVersionNumberInternal(const UResourceBundle *resourceBundle);
+
+/**
+ * Return the name of the Locale associated with this ResourceBundle. This API allows
+ * you to query for the real locale of the resource. For example, if you requested 
+ * "en_US_CALIFORNIA" and only "en_US" bundle exists, "en_US" will be returned. 
+ * For subresources, the locale where this resource comes from will be returned.
+ * If fallback has occured, getLocale will reflect this.
+ *
+ * This internal version avoids deprecated-warnings in ICU code.
+ *
+ * @param resourceBundle resource bundle in question
+ * @param status just for catching illegal arguments
+ * @return  A Locale name
+ */
+U_CAPI const char* U_EXPORT2 
+ures_getLocaleInternal(const UResourceBundle* resourceBundle, 
+               UErrorCode* status);
+
+/**
+ * Same as ures_openDirect() but uses the fill-in parameter instead of allocating a new bundle.
+ * 
+ * @param r The existing UResourceBundle to fill in. If NULL then status will be
+ *          set to U_ILLEGAL_ARGUMENT_ERROR.
+ * @param packageName   The packageName and locale together point to an ICU udata object,
+ *                      as defined by <code> udata_open( packageName, "res", locale, err) </code>
+ *                      or equivalent.  Typically, packageName will refer to a (.dat) file, or to
+ *                      a package registered with udata_setAppData(). Using a full file or directory
+ *                      pathname for packageName is deprecated. If NULL, ICU data will be used.
+ * @param locale  specifies the locale for which we want to open the resource
+ *                if NULL, the default locale will be used. If strlen(locale) == 0
+ *                root locale will be used.
+ * @param status The error code.
+ * @see ures_openDirect
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+ures_openDirectFillIn(UResourceBundle *r,
+                      const char *packageName,
+                      const char *locale,
+                      UErrorCode *status);
+
+#endif /*URESIMP_H*/
diff --git a/thirdparty/icu4c/common/ureslocs.h b/thirdparty/icu4c/common/ureslocs.h
new file mode 100644
index 0000000000..f7c3344ef2
--- /dev/null
+++ b/thirdparty/icu4c/common/ureslocs.h
@@ -0,0 +1,27 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2009-2014 International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef __URESLOCS_H__
+#define __URESLOCS_H__
+
+#include "unicode/utypes.h"
+#include "unicode/udata.h"
+
+U_CDECL_BEGIN
+
+
+#define U_ICUDATA_LANG U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "lang"
+#define U_ICUDATA_REGION U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "region"
+#define U_ICUDATA_CURR U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "curr"
+#define U_ICUDATA_ZONE U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "zone"
+#define U_ICUDATA_UNIT U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "unit"
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/usc_impl.cpp b/thirdparty/icu4c/common/usc_impl.cpp
new file mode 100644
index 0000000000..111029b974
--- /dev/null
+++ b/thirdparty/icu4c/common/usc_impl.cpp
@@ -0,0 +1,361 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File USC_IMPL.C
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   07/08/2002  Eric Mader  Creation.
+******************************************************************************
+*/
+
+#include "unicode/uscript.h"
+#include "usc_impl.h"
+#include "cmemory.h"
+
+#define PAREN_STACK_DEPTH 32
+
+#define MOD(sp) ((sp) % PAREN_STACK_DEPTH)
+#define LIMIT_INC(sp) (((sp) < PAREN_STACK_DEPTH)? (sp) + 1 : PAREN_STACK_DEPTH)
+#define INC(sp,count) (MOD((sp) + (count)))
+#define INC1(sp) (INC(sp, 1))
+#define DEC(sp,count) (MOD((sp) + PAREN_STACK_DEPTH - (count)))
+#define DEC1(sp) (DEC(sp, 1))
+#define STACK_IS_EMPTY(scriptRun) ((scriptRun)->pushCount <= 0)
+#define STACK_IS_NOT_EMPTY(scriptRun) (! STACK_IS_EMPTY(scriptRun))
+#define TOP(scriptRun) ((scriptRun)->parenStack[(scriptRun)->parenSP])
+#define SYNC_FIXUP(scriptRun) ((scriptRun)->fixupCount = 0)
+
+struct ParenStackEntry
+{
+    int32_t pairIndex;
+    UScriptCode scriptCode;
+};
+
+struct UScriptRun
+{
+    int32_t textLength;
+    const UChar *textArray;
+
+    int32_t scriptStart;
+    int32_t scriptLimit;
+    UScriptCode scriptCode;
+
+    struct ParenStackEntry parenStack[PAREN_STACK_DEPTH];
+    int32_t parenSP;
+    int32_t pushCount;
+    int32_t fixupCount;
+};
+
+static int8_t highBit(int32_t value);
+
+static const UChar32 pairedChars[] = {
+    0x0028, 0x0029, /* ascii paired punctuation */
+    0x003c, 0x003e,
+    0x005b, 0x005d,
+    0x007b, 0x007d,
+    0x00ab, 0x00bb, /* guillemets */
+    0x2018, 0x2019, /* general punctuation */
+    0x201c, 0x201d,
+    0x2039, 0x203a,
+    0x3008, 0x3009, /* chinese paired punctuation */
+    0x300a, 0x300b,
+    0x300c, 0x300d,
+    0x300e, 0x300f,
+    0x3010, 0x3011,
+    0x3014, 0x3015,
+    0x3016, 0x3017,
+    0x3018, 0x3019,
+    0x301a, 0x301b
+};
+
+static void push(UScriptRun *scriptRun, int32_t pairIndex, UScriptCode scriptCode)
+{
+    scriptRun->pushCount  = LIMIT_INC(scriptRun->pushCount);
+    scriptRun->fixupCount = LIMIT_INC(scriptRun->fixupCount);
+    
+    scriptRun->parenSP = INC1(scriptRun->parenSP);
+    scriptRun->parenStack[scriptRun->parenSP].pairIndex  = pairIndex;
+    scriptRun->parenStack[scriptRun->parenSP].scriptCode = scriptCode;
+}
+
+static void pop(UScriptRun *scriptRun)
+{
+    if (STACK_IS_EMPTY(scriptRun)) {
+        return;
+    }
+    
+    if (scriptRun->fixupCount > 0) {
+        scriptRun->fixupCount -= 1;
+    }
+    
+    scriptRun->pushCount -= 1;
+    scriptRun->parenSP = DEC1(scriptRun->parenSP);
+    
+    /* If the stack is now empty, reset the stack
+       pointers to their initial values.
+     */
+    if (STACK_IS_EMPTY(scriptRun)) {
+        scriptRun->parenSP = -1;
+    }
+}
+
+static void fixup(UScriptRun *scriptRun, UScriptCode scriptCode)
+{
+    int32_t fixupSP = DEC(scriptRun->parenSP, scriptRun->fixupCount);
+    
+    while (scriptRun->fixupCount-- > 0) {
+        fixupSP = INC1(fixupSP);
+        scriptRun->parenStack[fixupSP].scriptCode = scriptCode;
+    }
+}
+
+static int8_t
+highBit(int32_t value)
+{
+    int8_t bit = 0;
+
+    if (value <= 0) {
+        return -32;
+    }
+
+    if (value >= 1 << 16) {
+        value >>= 16;
+        bit += 16;
+    }
+
+    if (value >= 1 << 8) {
+        value >>= 8;
+        bit += 8;
+    }
+
+    if (value >= 1 << 4) {
+        value >>= 4;
+        bit += 4;
+    }
+
+    if (value >= 1 << 2) {
+        value >>= 2;
+        bit += 2;
+    }
+
+    if (value >= 1 << 1) {
+        //value >>= 1;
+        bit += 1;
+    }
+
+    return bit;
+}
+
+static int32_t
+getPairIndex(UChar32 ch)
+{
+    int32_t pairedCharCount = UPRV_LENGTHOF(pairedChars);
+    int32_t pairedCharPower = 1 << highBit(pairedCharCount);
+    int32_t pairedCharExtra = pairedCharCount - pairedCharPower;
+
+    int32_t probe = pairedCharPower;
+    int32_t pairIndex = 0;
+
+    if (ch >= pairedChars[pairedCharExtra]) {
+        pairIndex = pairedCharExtra;
+    }
+
+    while (probe > (1 << 0)) {
+        probe >>= 1;
+
+        if (ch >= pairedChars[pairIndex + probe]) {
+            pairIndex += probe;
+        }
+    }
+
+    if (pairedChars[pairIndex] != ch) {
+        pairIndex = -1;
+    }
+
+    return pairIndex;
+}
+
+static UBool
+sameScript(UScriptCode scriptOne, UScriptCode scriptTwo)
+{
+    return scriptOne <= USCRIPT_INHERITED || scriptTwo <= USCRIPT_INHERITED || scriptOne == scriptTwo;
+}
+
+U_CAPI UScriptRun * U_EXPORT2
+uscript_openRun(const UChar *src, int32_t length, UErrorCode *pErrorCode)
+{
+    UScriptRun *result = NULL;
+
+    if (pErrorCode == NULL || U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+
+    result = (UScriptRun *)uprv_malloc(sizeof (UScriptRun));
+
+    if (result == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+
+    uscript_setRunText(result, src, length, pErrorCode);
+
+    /* Release the UScriptRun if uscript_setRunText() returns an error */
+    if (U_FAILURE(*pErrorCode)) {
+        uprv_free(result);
+        result = NULL;
+    }
+
+    return result;
+}
+
+U_CAPI void U_EXPORT2
+uscript_closeRun(UScriptRun *scriptRun)
+{
+    if (scriptRun != NULL) {
+        uprv_free(scriptRun);
+    }
+}
+
+U_CAPI void U_EXPORT2
+uscript_resetRun(UScriptRun *scriptRun)
+{
+    if (scriptRun != NULL) {
+        scriptRun->scriptStart = 0;
+        scriptRun->scriptLimit = 0;
+        scriptRun->scriptCode  = USCRIPT_INVALID_CODE;
+        scriptRun->parenSP     = -1;
+        scriptRun->pushCount   =  0;
+        scriptRun->fixupCount  =  0;
+    }
+}
+
+U_CAPI void U_EXPORT2
+uscript_setRunText(UScriptRun *scriptRun, const UChar *src, int32_t length, UErrorCode *pErrorCode)
+{
+    if (pErrorCode == NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    if (scriptRun == NULL || length < 0 || ((src == NULL) != (length == 0))) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+    scriptRun->textArray  = src;
+    scriptRun->textLength = length;
+
+    uscript_resetRun(scriptRun);
+}
+
+U_CAPI UBool U_EXPORT2
+uscript_nextRun(UScriptRun *scriptRun, int32_t *pRunStart, int32_t *pRunLimit, UScriptCode *pRunScript)
+{
+    UErrorCode error = U_ZERO_ERROR;
+
+    /* if we've fallen off the end of the text, we're done */
+    if (scriptRun == NULL || scriptRun->scriptLimit >= scriptRun->textLength) {
+        return FALSE;
+    }
+    
+    SYNC_FIXUP(scriptRun);
+    scriptRun->scriptCode = USCRIPT_COMMON;
+
+    for (scriptRun->scriptStart = scriptRun->scriptLimit; scriptRun->scriptLimit < scriptRun->textLength; scriptRun->scriptLimit += 1) {
+        UChar   high = scriptRun->textArray[scriptRun->scriptLimit];
+        UChar32 ch   = high;
+        UScriptCode sc;
+        int32_t pairIndex;
+
+        /*
+         * if the character is a high surrogate and it's not the last one
+         * in the text, see if it's followed by a low surrogate
+         */
+        if (high >= 0xD800 && high <= 0xDBFF && scriptRun->scriptLimit < scriptRun->textLength - 1) {
+            UChar low = scriptRun->textArray[scriptRun->scriptLimit + 1];
+
+            /*
+             * if it is followed by a low surrogate,
+             * consume it and form the full character
+             */
+            if (low >= 0xDC00 && low <= 0xDFFF) {
+                ch = (high - 0xD800) * 0x0400 + low - 0xDC00 + 0x10000;
+                scriptRun->scriptLimit += 1;
+            }
+        }
+
+        sc = uscript_getScript(ch, &error);
+        pairIndex = getPairIndex(ch);
+
+        /*
+         * Paired character handling:
+         *
+         * if it's an open character, push it onto the stack.
+         * if it's a close character, find the matching open on the
+         * stack, and use that script code. Any non-matching open
+         * characters above it on the stack will be poped.
+         */
+        if (pairIndex >= 0) {
+            if ((pairIndex & 1) == 0) {
+                push(scriptRun, pairIndex, scriptRun->scriptCode);
+            } else {
+                int32_t pi = pairIndex & ~1;
+
+                while (STACK_IS_NOT_EMPTY(scriptRun) && TOP(scriptRun).pairIndex != pi) {
+                    pop(scriptRun);
+                }
+
+                if (STACK_IS_NOT_EMPTY(scriptRun)) {
+                    sc = TOP(scriptRun).scriptCode;
+                }
+            }
+        }
+
+        if (sameScript(scriptRun->scriptCode, sc)) {
+            if (scriptRun->scriptCode <= USCRIPT_INHERITED && sc > USCRIPT_INHERITED) {
+                scriptRun->scriptCode = sc;
+
+                fixup(scriptRun, scriptRun->scriptCode);
+            }
+
+            /*
+             * if this character is a close paired character,
+             * pop the matching open character from the stack
+             */
+            if (pairIndex >= 0 && (pairIndex & 1) != 0) {
+                pop(scriptRun);
+            }
+        } else {
+            /*
+             * if the run broke on a surrogate pair,
+             * end it before the high surrogate
+             */
+            if (ch >= 0x10000) {
+                scriptRun->scriptLimit -= 1;
+            }
+
+            break;
+        }
+    }
+
+
+    if (pRunStart != NULL) {
+        *pRunStart = scriptRun->scriptStart;
+    }
+
+    if (pRunLimit != NULL) {
+        *pRunLimit = scriptRun->scriptLimit;
+    }
+
+    if (pRunScript != NULL) {
+        *pRunScript = scriptRun->scriptCode;
+    }
+
+    return TRUE;
+}
diff --git a/thirdparty/icu4c/common/usc_impl.h b/thirdparty/icu4c/common/usc_impl.h
new file mode 100644
index 0000000000..44899649d4
--- /dev/null
+++ b/thirdparty/icu4c/common/usc_impl.h
@@ -0,0 +1,139 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File USC_IMPL.H
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   07/08/2002  Eric Mader  Creation.
+******************************************************************************
+*/
+
+#ifndef USC_IMPL_H
+#define USC_IMPL_H
+#include "unicode/utypes.h"
+#include "unicode/uscript.h"
+
+/**
+ * <code>UScriptRun</code> is used to find runs of characters in
+ * the same script. It implements a simple iterator over an array
+ * of characters. The iterator will resolve script-neutral characters
+ * like punctuation into the script of the surrounding characters.
+ *
+ * The iterator will try to match paired punctuation. If it sees an
+ * opening punctuation character, it will remember the script that
+ * was assigned to that character, and assign the same script to the
+ * matching closing punctuation.
+ *
+ * Scripts are chosen based on the <code>UScriptCode</code> enumeration.
+ * No attempt is made to combine related scripts into a single run. In
+ * particular, Hiragana, Katakana, and Han characters will appear in seperate
+ * runs.
+
+ * Here is an example of how to iterate over script runs:
+ * <pre>
+ * \code
+ * void printScriptRuns(const UChar *text, int32_t length)
+ * {
+ *     UErrorCode error = U_ZERO_ERROR;
+ *     UScriptRun *scriptRun = uscript_openRun(text, testLength, &error);
+ *     int32_t start = 0, limit = 0;
+ *     UScriptCode code = USCRIPT_INVALID_CODE;
+ *
+ *     while (uscript_nextRun(&start, &limit, &code)) {
+ *         printf("Script '%s' from %d to %d.\n", uscript_getName(code), start, limit);
+ *     }
+ *
+ *     uscript_closeRun(scriptRun);
+ *  }
+ * </pre>
+ */
+struct UScriptRun;
+
+typedef struct UScriptRun UScriptRun;
+
+/**
+ * Create a <code>UScriptRun</code> object for iterating over the given text. This object must
+ * be freed using <code>uscript_closeRun()</code>. Note that this object does not copy the source text,
+ * only the pointer to it. You must make sure that the pointer remains valid until you call
+ * <code>uscript_closeRun()</code> or <code>uscript_setRunText()</code>.
+ *
+ * @param src is the address of the array of characters over which to iterate.
+ *        if <code>src == NULL</code> and <code>length == 0</code>,
+ *        an empty <code>UScriptRun</code> object will be returned.
+ *
+ * @param length is the number of characters over which to iterate.
+ *
+ * @param pErrorCode is a pointer to a valid <code>UErrorCode</code> value. If this value
+ *        indicates a failure on entry, the function will immediately return.
+ *        On exit the value will indicate the success of the operation.
+ *
+ * @return the address of <code>UScriptRun</code> object which will iterate over the text,
+ *         or <code>NULL</code> if the operation failed.
+ */
+U_CAPI UScriptRun * U_EXPORT2
+uscript_openRun(const UChar *src, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * Frees the given <code>UScriptRun</code> object and any storage associated with it.
+ * On return, scriptRun no longer points to a valid <code>UScriptRun</code> object.
+ *
+ * @param scriptRun is the <code>UScriptRun</code> object which will be freed.
+ */
+U_CAPI void U_EXPORT2
+uscript_closeRun(UScriptRun *scriptRun);
+
+/**
+ * Reset the <code>UScriptRun</code> object so that it will start iterating from
+ * the beginning.
+ *
+ * @param scriptRun is the address of the <code>UScriptRun</code> object to be reset.
+ */
+U_CAPI void U_EXPORT2
+uscript_resetRun(UScriptRun *scriptRun);
+
+/**
+ * Change the text over which the given <code>UScriptRun</code> object iterates.
+ *
+ * @param scriptRun is the <code>UScriptRun</code> object which will be changed.
+ *
+ * @param src is the address of the new array of characters over which to iterate.
+ *        If <code>src == NULL</code> and <code>length == 0</code>,
+ *        the <code>UScriptRun</code> object will become empty.
+ *
+ * @param length is the new number of characters over which to iterate
+ *
+ * @param pErrorCode is a pointer to a valid <code>UErrorCode</code> value. If this value
+ *        indicates a failure on entry, the function will immediately return.
+ *        On exit the value will indicate the success of the operation.
+ */
+U_CAPI void U_EXPORT2
+uscript_setRunText(UScriptRun *scriptRun, const UChar *src, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * Advance the <code>UScriptRun</code> object to the next script run, return the start and limit
+ * offsets, and the script of the run.
+ *
+ * @param scriptRun is the address of the <code>UScriptRun</code> object.
+ *
+ * @param pRunStart is a pointer to the variable to receive the starting offset of the next run.
+ *        This pointer can be <code>NULL</code> if the value is not needed.
+ *
+ * @param pRunLimit is a pointer to the variable to receive the limit offset of the next run.
+ *        This pointer can be <code>NULL</code> if the value is not needed.
+ *
+ * @param pRunScript is a pointer to the variable to receive the UScriptCode for the
+ *        script of the current run. This pointer can be <code>NULL</code> if the value is not needed.
+ *
+ * @return true if there was another script run.
+ */
+U_CAPI UBool U_EXPORT2
+uscript_nextRun(UScriptRun *scriptRun, int32_t *pRunStart, int32_t *pRunLimit, UScriptCode *pRunScript);
+
+#endif
diff --git a/thirdparty/icu4c/common/uscript.cpp b/thirdparty/icu4c/common/uscript.cpp
new file mode 100644
index 0000000000..f8bd7e7fdd
--- /dev/null
+++ b/thirdparty/icu4c/common/uscript.cpp
@@ -0,0 +1,149 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1997-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*
+* File USCRIPT.C
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   07/06/2001    Ram         Creation.
+******************************************************************************
+*/
+
+#include "unicode/uchar.h"
+#include "unicode/uscript.h"
+#include "unicode/uloc.h"
+#include "bytesinkutil.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "ulocimp.h"
+
+static const UScriptCode JAPANESE[3] = { USCRIPT_KATAKANA, USCRIPT_HIRAGANA, USCRIPT_HAN };
+static const UScriptCode KOREAN[2] = { USCRIPT_HANGUL, USCRIPT_HAN };
+static const UScriptCode HAN_BOPO[2] = { USCRIPT_HAN, USCRIPT_BOPOMOFO };
+
+static int32_t
+setCodes(const UScriptCode *src, int32_t length,
+         UScriptCode *dest, int32_t capacity, UErrorCode *err) {
+    int32_t i;
+    if(U_FAILURE(*err)) { return 0; }
+    if(length > capacity) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+        return length;
+    }
+    for(i = 0; i < length; ++i) {
+        dest[i] = src[i];
+    }
+    return length;
+}
+
+static int32_t
+setOneCode(UScriptCode script, UScriptCode *scripts, int32_t capacity, UErrorCode *err) {
+    if(U_FAILURE(*err)) { return 0; }
+    if(1 > capacity) {
+        *err = U_BUFFER_OVERFLOW_ERROR;
+        return 1;
+    }
+    scripts[0] = script;
+    return 1;
+}
+
+static int32_t
+getCodesFromLocale(const char *locale,
+                   UScriptCode *scripts, int32_t capacity, UErrorCode *err) {
+    UErrorCode internalErrorCode = U_ZERO_ERROR;
+    char lang[8] = {0};
+    char script[8] = {0};
+    int32_t scriptLength;
+    if(U_FAILURE(*err)) { return 0; }
+    // Multi-script languages, equivalent to the LocaleScript data
+    // that we used to load from locale resource bundles.
+    /*length = */ uloc_getLanguage(locale, lang, UPRV_LENGTHOF(lang), &internalErrorCode);
+    if(U_FAILURE(internalErrorCode) || internalErrorCode == U_STRING_NOT_TERMINATED_WARNING) {
+        return 0;
+    }
+    if(0 == uprv_strcmp(lang, "ja")) {
+        return setCodes(JAPANESE, UPRV_LENGTHOF(JAPANESE), scripts, capacity, err);
+    }
+    if(0 == uprv_strcmp(lang, "ko")) {
+        return setCodes(KOREAN, UPRV_LENGTHOF(KOREAN), scripts, capacity, err);
+    }
+    scriptLength = uloc_getScript(locale, script, UPRV_LENGTHOF(script), &internalErrorCode);
+    if(U_FAILURE(internalErrorCode) || internalErrorCode == U_STRING_NOT_TERMINATED_WARNING) {
+        return 0;
+    }
+    if(0 == uprv_strcmp(lang, "zh") && 0 == uprv_strcmp(script, "Hant")) {
+        return setCodes(HAN_BOPO, UPRV_LENGTHOF(HAN_BOPO), scripts, capacity, err);
+    }
+    // Explicit script code.
+    if(scriptLength != 0) {
+        UScriptCode scriptCode = (UScriptCode)u_getPropertyValueEnum(UCHAR_SCRIPT, script);
+        if(scriptCode != USCRIPT_INVALID_CODE) {
+            if(scriptCode == USCRIPT_SIMPLIFIED_HAN || scriptCode == USCRIPT_TRADITIONAL_HAN) {
+                scriptCode = USCRIPT_HAN;
+            }
+            return setOneCode(scriptCode, scripts, capacity, err);
+        }
+    }
+    return 0;
+}
+
+/* TODO: this is a bad API and should be deprecated, ticket #11141 */
+U_CAPI int32_t  U_EXPORT2
+uscript_getCode(const char* nameOrAbbrOrLocale,
+                UScriptCode* fillIn,
+                int32_t capacity,
+                UErrorCode* err){
+    UBool triedCode;
+    UErrorCode internalErrorCode;
+    int32_t length;
+
+    if(U_FAILURE(*err)) {
+        return 0;
+    }
+    if(nameOrAbbrOrLocale==NULL ||
+            (fillIn == NULL ? capacity != 0 : capacity < 0)) {
+        *err = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    triedCode = FALSE;
+    if(uprv_strchr(nameOrAbbrOrLocale, '-')==NULL && uprv_strchr(nameOrAbbrOrLocale, '_')==NULL ){
+        /* try long and abbreviated script names first */
+        UScriptCode code = (UScriptCode) u_getPropertyValueEnum(UCHAR_SCRIPT, nameOrAbbrOrLocale);
+        if(code!=USCRIPT_INVALID_CODE) {
+            return setOneCode(code, fillIn, capacity, err);
+        }
+        triedCode = TRUE;
+    }
+    internalErrorCode = U_ZERO_ERROR;
+    length = getCodesFromLocale(nameOrAbbrOrLocale, fillIn, capacity, err);
+    if(U_FAILURE(*err) || length != 0) {
+        return length;
+    }
+    icu::CharString likely;
+    {
+        icu::CharStringByteSink sink(&likely);
+        ulocimp_addLikelySubtags(nameOrAbbrOrLocale, sink, &internalErrorCode);
+    }
+    if(U_SUCCESS(internalErrorCode) && internalErrorCode != U_STRING_NOT_TERMINATED_WARNING) {
+        length = getCodesFromLocale(likely.data(), fillIn, capacity, err);
+        if(U_FAILURE(*err) || length != 0) {
+            return length;
+        }
+    }
+    if(!triedCode) {
+        /* still not found .. try long and abbreviated script names again */
+        UScriptCode code = (UScriptCode) u_getPropertyValueEnum(UCHAR_SCRIPT, nameOrAbbrOrLocale);
+        if(code!=USCRIPT_INVALID_CODE) {
+            return setOneCode(code, fillIn, capacity, err);
+        }
+    }
+    return 0;
+}
diff --git a/thirdparty/icu4c/common/uscript_props.cpp b/thirdparty/icu4c/common/uscript_props.cpp
new file mode 100644
index 0000000000..25d287b57a
--- /dev/null
+++ b/thirdparty/icu4c/common/uscript_props.cpp
@@ -0,0 +1,302 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2013-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  uscript_props.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2013feb16
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/unistr.h"
+#include "unicode/uscript.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+#include "cmemory.h"
+
+namespace {
+
+// Script metadata (script properties).
+// See http://unicode.org/cldr/trac/browser/trunk/common/properties/scriptMetadata.txt
+
+// 0 = NOT_ENCODED, no sample character, default false script properties.
+// Bits 20.. 0: sample character
+
+// Bits 23..21: usage
+const int32_t UNKNOWN = 1 << 21;
+const int32_t EXCLUSION = 2 << 21;
+const int32_t LIMITED_USE = 3 << 21;
+// st int32_t ASPIRATIONAL = 4 << 21; -- not used any more since Unicode 10
+const int32_t RECOMMENDED = 5 << 21;
+
+// Bits 31..24: Single-bit flags
+const int32_t RTL = 1 << 24;
+const int32_t LB_LETTERS = 1 << 25;
+const int32_t CASED = 1 << 26;
+
+const int32_t SCRIPT_PROPS[] = {
+    // Begin copy-paste output from
+    // tools/trunk/unicode/py/parsescriptmetadata.py
+    0x0040 | RECOMMENDED,  // Zyyy
+    0x0308 | RECOMMENDED,  // Zinh
+    0x0628 | RECOMMENDED | RTL,  // Arab
+    0x0531 | RECOMMENDED | CASED,  // Armn
+    0x0995 | RECOMMENDED,  // Beng
+    0x3105 | RECOMMENDED | LB_LETTERS,  // Bopo
+    0x13C4 | LIMITED_USE | CASED,  // Cher
+    0x03E2 | EXCLUSION | CASED,  // Copt
+    0x042F | RECOMMENDED | CASED,  // Cyrl
+    0x10414 | EXCLUSION | CASED,  // Dsrt
+    0x0905 | RECOMMENDED,  // Deva
+    0x12A0 | RECOMMENDED,  // Ethi
+    0x10D3 | RECOMMENDED,  // Geor
+    0x10330 | EXCLUSION,  // Goth
+    0x03A9 | RECOMMENDED | CASED,  // Grek
+    0x0A95 | RECOMMENDED,  // Gujr
+    0x0A15 | RECOMMENDED,  // Guru
+    0x5B57 | RECOMMENDED | LB_LETTERS,  // Hani
+    0xAC00 | RECOMMENDED,  // Hang
+    0x05D0 | RECOMMENDED | RTL,  // Hebr
+    0x304B | RECOMMENDED | LB_LETTERS,  // Hira
+    0x0C95 | RECOMMENDED,  // Knda
+    0x30AB | RECOMMENDED | LB_LETTERS,  // Kana
+    0x1780 | RECOMMENDED | LB_LETTERS,  // Khmr
+    0x0EA5 | RECOMMENDED | LB_LETTERS,  // Laoo
+    0x004C | RECOMMENDED | CASED,  // Latn
+    0x0D15 | RECOMMENDED,  // Mlym
+    0x1826 | EXCLUSION,  // Mong
+    0x1000 | RECOMMENDED | LB_LETTERS,  // Mymr
+    0x168F | EXCLUSION,  // Ogam
+    0x10300 | EXCLUSION,  // Ital
+    0x0B15 | RECOMMENDED,  // Orya
+    0x16A0 | EXCLUSION,  // Runr
+    0x0D85 | RECOMMENDED,  // Sinh
+    0x0710 | LIMITED_USE | RTL,  // Syrc
+    0x0B95 | RECOMMENDED,  // Taml
+    0x0C15 | RECOMMENDED,  // Telu
+    0x078C | RECOMMENDED | RTL,  // Thaa
+    0x0E17 | RECOMMENDED | LB_LETTERS,  // Thai
+    0x0F40 | RECOMMENDED,  // Tibt
+    0x14C0 | LIMITED_USE,  // Cans
+    0xA288 | LIMITED_USE | LB_LETTERS,  // Yiii
+    0x1703 | EXCLUSION,  // Tglg
+    0x1723 | EXCLUSION,  // Hano
+    0x1743 | EXCLUSION,  // Buhd
+    0x1763 | EXCLUSION,  // Tagb
+    0x280E | UNKNOWN,  // Brai
+    0x10800 | EXCLUSION | RTL,  // Cprt
+    0x1900 | LIMITED_USE,  // Limb
+    0x10000 | EXCLUSION,  // Linb
+    0x10480 | EXCLUSION,  // Osma
+    0x10450 | EXCLUSION,  // Shaw
+    0x1950 | LIMITED_USE | LB_LETTERS,  // Tale
+    0x10380 | EXCLUSION,  // Ugar
+    0,
+    0x1A00 | EXCLUSION,  // Bugi
+    0x2C00 | EXCLUSION | CASED,  // Glag
+    0x10A00 | EXCLUSION | RTL,  // Khar
+    0xA800 | LIMITED_USE,  // Sylo
+    0x1980 | LIMITED_USE | LB_LETTERS,  // Talu
+    0x2D30 | LIMITED_USE,  // Tfng
+    0x103A0 | EXCLUSION,  // Xpeo
+    0x1B05 | LIMITED_USE,  // Bali
+    0x1BC0 | LIMITED_USE,  // Batk
+    0,
+    0x11005 | EXCLUSION,  // Brah
+    0xAA00 | LIMITED_USE,  // Cham
+    0,
+    0,
+    0,
+    0,
+    0x13153 | EXCLUSION,  // Egyp
+    0,
+    0x5B57 | RECOMMENDED | LB_LETTERS,  // Hans
+    0x5B57 | RECOMMENDED | LB_LETTERS,  // Hant
+    0x16B1C | EXCLUSION,  // Hmng
+    0x10CA1 | EXCLUSION | RTL | CASED,  // Hung
+    0,
+    0xA984 | LIMITED_USE,  // Java
+    0xA90A | LIMITED_USE,  // Kali
+    0,
+    0,
+    0x1C00 | LIMITED_USE,  // Lepc
+    0x10647 | EXCLUSION,  // Lina
+    0x0840 | LIMITED_USE | RTL,  // Mand
+    0,
+    0x10980 | EXCLUSION | RTL,  // Mero
+    0x07CA | LIMITED_USE | RTL,  // Nkoo
+    0x10C00 | EXCLUSION | RTL,  // Orkh
+    0x1036B | EXCLUSION,  // Perm
+    0xA840 | EXCLUSION,  // Phag
+    0x10900 | EXCLUSION | RTL,  // Phnx
+    0x16F00 | LIMITED_USE,  // Plrd
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0xA549 | LIMITED_USE,  // Vaii
+    0,
+    0x12000 | EXCLUSION,  // Xsux
+    0,
+    0xFDD0 | UNKNOWN,  // Zzzz
+    0x102A0 | EXCLUSION,  // Cari
+    0x304B | RECOMMENDED | LB_LETTERS,  // Jpan
+    0x1A20 | LIMITED_USE | LB_LETTERS,  // Lana
+    0x10280 | EXCLUSION,  // Lyci
+    0x10920 | EXCLUSION | RTL,  // Lydi
+    0x1C5A | LIMITED_USE,  // Olck
+    0xA930 | EXCLUSION,  // Rjng
+    0xA882 | LIMITED_USE,  // Saur
+    0x1D850 | EXCLUSION,  // Sgnw
+    0x1B83 | LIMITED_USE,  // Sund
+    0,
+    0xABC0 | LIMITED_USE,  // Mtei
+    0x10840 | EXCLUSION | RTL,  // Armi
+    0x10B00 | EXCLUSION | RTL,  // Avst
+    0x11103 | LIMITED_USE,  // Cakm
+    0xAC00 | RECOMMENDED,  // Kore
+    0x11083 | EXCLUSION,  // Kthi
+    0x10AD8 | EXCLUSION | RTL,  // Mani
+    0x10B60 | EXCLUSION | RTL,  // Phli
+    0x10B8F | EXCLUSION | RTL,  // Phlp
+    0,
+    0x10B40 | EXCLUSION | RTL,  // Prti
+    0x0800 | EXCLUSION | RTL,  // Samr
+    0xAA80 | LIMITED_USE | LB_LETTERS,  // Tavt
+    0,
+    0,
+    0xA6A0 | LIMITED_USE,  // Bamu
+    0xA4D0 | LIMITED_USE,  // Lisu
+    0,
+    0x10A60 | EXCLUSION | RTL,  // Sarb
+    0x16AE6 | EXCLUSION,  // Bass
+    0x1BC20 | EXCLUSION,  // Dupl
+    0x10500 | EXCLUSION,  // Elba
+    0x11315 | EXCLUSION,  // Gran
+    0,
+    0,
+    0x1E802 | EXCLUSION | RTL,  // Mend
+    0x109A0 | EXCLUSION | RTL,  // Merc
+    0x10A95 | EXCLUSION | RTL,  // Narb
+    0x10896 | EXCLUSION | RTL,  // Nbat
+    0x10873 | EXCLUSION | RTL,  // Palm
+    0x112BE | EXCLUSION,  // Sind
+    0x118B4 | EXCLUSION | CASED,  // Wara
+    0,
+    0,
+    0x16A4F | EXCLUSION,  // Mroo
+    0x1B1C4 | EXCLUSION | LB_LETTERS,  // Nshu
+    0x11183 | EXCLUSION,  // Shrd
+    0x110D0 | EXCLUSION,  // Sora
+    0x11680 | EXCLUSION,  // Takr
+    0x18229 | EXCLUSION | LB_LETTERS,  // Tang
+    0,
+    0x14400 | EXCLUSION,  // Hluw
+    0x11208 | EXCLUSION,  // Khoj
+    0x11484 | EXCLUSION,  // Tirh
+    0x10537 | EXCLUSION,  // Aghb
+    0x11152 | EXCLUSION,  // Mahj
+    0x11717 | EXCLUSION | LB_LETTERS,  // Ahom
+    0x108F4 | EXCLUSION | RTL,  // Hatr
+    0x1160E | EXCLUSION,  // Modi
+    0x1128F | EXCLUSION,  // Mult
+    0x11AC0 | EXCLUSION,  // Pauc
+    0x1158E | EXCLUSION,  // Sidd
+    0x1E909 | LIMITED_USE | RTL | CASED,  // Adlm
+    0x11C0E | EXCLUSION,  // Bhks
+    0x11C72 | EXCLUSION,  // Marc
+    0x11412 | LIMITED_USE,  // Newa
+    0x104B5 | LIMITED_USE | CASED,  // Osge
+    0x5B57 | RECOMMENDED | LB_LETTERS,  // Hanb
+    0x1112 | RECOMMENDED,  // Jamo
+    0,
+    0x11D10 | EXCLUSION,  // Gonm
+    0x11A5C | EXCLUSION,  // Soyo
+    0x11A0B | EXCLUSION,  // Zanb
+    0x1180B | EXCLUSION,  // Dogr
+    0x11D71 | LIMITED_USE,  // Gong
+    0x11EE5 | EXCLUSION,  // Maka
+    0x16E40 | EXCLUSION | CASED,  // Medf
+    0x10D12 | LIMITED_USE | RTL,  // Rohg
+    0x10F42 | EXCLUSION | RTL,  // Sogd
+    0x10F19 | EXCLUSION | RTL,  // Sogo
+    0x10FF1 | EXCLUSION | RTL,  // Elym
+    0x1E108 | LIMITED_USE,  // Hmnp
+    0x119CE | EXCLUSION,  // Nand
+    0x1E2E1 | LIMITED_USE,  // Wcho
+    0x10FBF | EXCLUSION | RTL,  // Chrs
+    0x1190C | EXCLUSION,  // Diak
+    0x18C65 | EXCLUSION | LB_LETTERS,  // Kits
+    0x10E88 | EXCLUSION | RTL,  // Yezi
+    // End copy-paste from parsescriptmetadata.py
+};
+
+int32_t getScriptProps(UScriptCode script) {
+    if (0 <= script && script < UPRV_LENGTHOF(SCRIPT_PROPS)) {
+        return SCRIPT_PROPS[script];
+    } else {
+        return 0;
+    }
+}
+
+}  // namespace
+
+U_CAPI int32_t U_EXPORT2
+uscript_getSampleString(UScriptCode script, UChar *dest, int32_t capacity, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) { return 0; }
+    if(capacity < 0 || (capacity > 0 && dest == NULL)) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    int32_t sampleChar = getScriptProps(script) & 0x1fffff;
+    int32_t length;
+    if(sampleChar == 0) {
+        length = 0;
+    } else {
+        length = U16_LENGTH(sampleChar);
+        if(length <= capacity) {
+            int32_t i = 0;
+            U16_APPEND_UNSAFE(dest, i, sampleChar);
+        }
+    }
+    return u_terminateUChars(dest, capacity, length, pErrorCode);
+}
+
+U_COMMON_API icu::UnicodeString U_EXPORT2
+uscript_getSampleUnicodeString(UScriptCode script) {
+    icu::UnicodeString sample;
+    int32_t sampleChar = getScriptProps(script) & 0x1fffff;
+    if(sampleChar != 0) {
+        sample.append(sampleChar);
+    }
+    return sample;
+}
+
+U_CAPI UScriptUsage U_EXPORT2
+uscript_getUsage(UScriptCode script) {
+    return (UScriptUsage)((getScriptProps(script) >> 21) & 7);
+}
+
+U_CAPI UBool U_EXPORT2
+uscript_isRightToLeft(UScriptCode script) {
+    return (getScriptProps(script) & RTL) != 0;
+}
+
+U_CAPI UBool U_EXPORT2
+uscript_breaksBetweenLetters(UScriptCode script) {
+    return (getScriptProps(script) & LB_LETTERS) != 0;
+}
+
+U_CAPI UBool U_EXPORT2
+uscript_isCased(UScriptCode script) {
+    return (getScriptProps(script) & CASED) != 0;
+}
diff --git a/thirdparty/icu4c/common/uset.cpp b/thirdparty/icu4c/common/uset.cpp
new file mode 100644
index 0000000000..eae7981d52
--- /dev/null
+++ b/thirdparty/icu4c/common/uset.cpp
@@ -0,0 +1,641 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uset.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002mar07
+*   created by: Markus W. Scherer
+*
+*   There are functions to efficiently serialize a USet into an array of uint16_t
+*   and functions to use such a serialized form efficiently without
+*   instantiating a new USet.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "unicode/uset.h"
+#include "unicode/uniset.h"
+#include "cmemory.h"
+#include "unicode/ustring.h"
+#include "unicode/parsepos.h"
+
+U_NAMESPACE_USE
+
+U_CAPI USet* U_EXPORT2
+uset_openEmpty() {
+    return (USet*) new UnicodeSet();
+}
+
+U_CAPI USet* U_EXPORT2
+uset_open(UChar32 start, UChar32 end) {
+    return (USet*) new UnicodeSet(start, end);
+}
+
+U_CAPI void U_EXPORT2
+uset_close(USet* set) {
+    delete (UnicodeSet*) set;
+}
+
+U_CAPI USet * U_EXPORT2
+uset_clone(const USet *set) {
+    return (USet*) (((UnicodeSet*) set)->UnicodeSet::clone());
+}
+
+U_CAPI UBool U_EXPORT2
+uset_isFrozen(const USet *set) {
+    return ((UnicodeSet*) set)->UnicodeSet::isFrozen();
+}
+
+U_CAPI void U_EXPORT2
+uset_freeze(USet *set) {
+    ((UnicodeSet*) set)->UnicodeSet::freeze();
+}
+
+U_CAPI USet * U_EXPORT2
+uset_cloneAsThawed(const USet *set) {
+    return (USet*) (((UnicodeSet*) set)->UnicodeSet::cloneAsThawed());
+}
+
+U_CAPI void U_EXPORT2
+uset_set(USet* set,
+     UChar32 start, UChar32 end) {
+    ((UnicodeSet*) set)->UnicodeSet::set(start, end);
+}
+
+U_CAPI void U_EXPORT2
+uset_addAll(USet* set, const USet *additionalSet) {
+    ((UnicodeSet*) set)->UnicodeSet::addAll(*((const UnicodeSet*)additionalSet));
+}
+
+U_CAPI void U_EXPORT2
+uset_add(USet* set, UChar32 c) {
+    ((UnicodeSet*) set)->UnicodeSet::add(c);
+}
+
+U_CAPI void U_EXPORT2
+uset_addRange(USet* set, UChar32 start, UChar32 end) {
+    ((UnicodeSet*) set)->UnicodeSet::add(start, end);    
+}
+
+U_CAPI void U_EXPORT2
+uset_addString(USet* set, const UChar* str, int32_t strLen) {
+    // UnicodeString handles -1 for strLen
+    UnicodeString s(strLen<0, str, strLen);
+    ((UnicodeSet*) set)->UnicodeSet::add(s);
+}
+
+U_CAPI void U_EXPORT2
+uset_addAllCodePoints(USet* set, const UChar *str, int32_t strLen) {
+    // UnicodeString handles -1 for strLen
+    UnicodeString s(str, strLen);
+    ((UnicodeSet*) set)->UnicodeSet::addAll(s);
+}
+
+U_CAPI void U_EXPORT2
+uset_remove(USet* set, UChar32 c) {
+    ((UnicodeSet*) set)->UnicodeSet::remove(c);
+}
+
+U_CAPI void U_EXPORT2
+uset_removeRange(USet* set, UChar32 start, UChar32 end) {
+    ((UnicodeSet*) set)->UnicodeSet::remove(start, end);
+}
+
+U_CAPI void U_EXPORT2
+uset_removeString(USet* set, const UChar* str, int32_t strLen) {
+    UnicodeString s(strLen==-1, str, strLen);
+    ((UnicodeSet*) set)->UnicodeSet::remove(s);
+}
+
+U_CAPI void U_EXPORT2
+uset_removeAll(USet* set, const USet* remove) {
+    ((UnicodeSet*) set)->UnicodeSet::removeAll(*(const UnicodeSet*)remove);
+}
+
+U_CAPI void U_EXPORT2
+uset_retain(USet* set, UChar32 start, UChar32 end) {
+    ((UnicodeSet*) set)->UnicodeSet::retain(start, end);
+}
+
+U_CAPI void U_EXPORT2
+uset_retainAll(USet* set, const USet* retain) {
+    ((UnicodeSet*) set)->UnicodeSet::retainAll(*(const UnicodeSet*)retain);
+}
+
+U_CAPI void U_EXPORT2
+uset_compact(USet* set) {
+    ((UnicodeSet*) set)->UnicodeSet::compact();
+}
+
+U_CAPI void U_EXPORT2
+uset_complement(USet* set) {
+    ((UnicodeSet*) set)->UnicodeSet::complement();
+}
+
+U_CAPI void U_EXPORT2
+uset_complementAll(USet* set, const USet* complement) {
+    ((UnicodeSet*) set)->UnicodeSet::complementAll(*(const UnicodeSet*)complement);
+}
+
+U_CAPI void U_EXPORT2
+uset_clear(USet* set) {
+    ((UnicodeSet*) set)->UnicodeSet::clear();
+}
+
+U_CAPI void U_EXPORT2
+uset_removeAllStrings(USet* set) {
+    ((UnicodeSet*) set)->UnicodeSet::removeAllStrings();
+}
+
+U_CAPI UBool U_EXPORT2
+uset_isEmpty(const USet* set) {
+    return ((const UnicodeSet*) set)->UnicodeSet::isEmpty();
+}
+
+U_CAPI UBool U_EXPORT2
+uset_contains(const USet* set, UChar32 c) {
+    return ((const UnicodeSet*) set)->UnicodeSet::contains(c);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsRange(const USet* set, UChar32 start, UChar32 end) {
+    return ((const UnicodeSet*) set)->UnicodeSet::contains(start, end);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsString(const USet* set, const UChar* str, int32_t strLen) {
+    UnicodeString s(strLen==-1, str, strLen);
+    return ((const UnicodeSet*) set)->UnicodeSet::contains(s);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsAll(const USet* set1, const USet* set2) {
+    return ((const UnicodeSet*) set1)->UnicodeSet::containsAll(* (const UnicodeSet*) set2);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsAllCodePoints(const USet* set, const UChar *str, int32_t strLen) {
+    // Create a string alias, since nothing is being added to the set.
+    UnicodeString s(strLen==-1, str, strLen);
+    return ((const UnicodeSet*) set)->UnicodeSet::containsAll(s);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsNone(const USet* set1, const USet* set2) {
+    return ((const UnicodeSet*) set1)->UnicodeSet::containsNone(* (const UnicodeSet*) set2);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_containsSome(const USet* set1, const USet* set2) {
+    return ((const UnicodeSet*) set1)->UnicodeSet::containsSome(* (const UnicodeSet*) set2);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_span(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition) {
+    return ((UnicodeSet*) set)->UnicodeSet::span(s, length, spanCondition);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_spanBack(const USet *set, const UChar *s, int32_t length, USetSpanCondition spanCondition) {
+    return ((UnicodeSet*) set)->UnicodeSet::spanBack(s, length, spanCondition);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_spanUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition) {
+    return ((UnicodeSet*) set)->UnicodeSet::spanUTF8(s, length, spanCondition);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_spanBackUTF8(const USet *set, const char *s, int32_t length, USetSpanCondition spanCondition) {
+    return ((UnicodeSet*) set)->UnicodeSet::spanBackUTF8(s, length, spanCondition);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_equals(const USet* set1, const USet* set2) {
+    return *(const UnicodeSet*)set1 == *(const UnicodeSet*)set2;
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_indexOf(const USet* set, UChar32 c) {
+    return ((UnicodeSet*) set)->UnicodeSet::indexOf(c);
+}
+
+U_CAPI UChar32 U_EXPORT2
+uset_charAt(const USet* set, int32_t index) {
+    return ((UnicodeSet*) set)->UnicodeSet::charAt(index);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_size(const USet* set) {
+    return ((const UnicodeSet*) set)->UnicodeSet::size();
+}
+
+U_NAMESPACE_BEGIN
+/**
+ * This class only exists to provide access to the UnicodeSet private
+ * USet support API.  Declaring a class a friend is more portable than
+ * trying to declare extern "C" functions as friends.
+ */
+class USetAccess /* not : public UObject because all methods are static */ {
+public:
+    /* Try to have the compiler inline these*/
+    inline static int32_t getStringCount(const UnicodeSet& set) {
+        return set.stringsSize();
+    }
+    inline static const UnicodeString* getString(const UnicodeSet& set,
+                                                 int32_t i) {
+        return set.getString(i);
+    }
+private:
+    /* do not instantiate*/
+    USetAccess();
+};
+U_NAMESPACE_END
+
+U_CAPI int32_t U_EXPORT2
+uset_getItemCount(const USet* uset) {
+    const UnicodeSet& set = *(const UnicodeSet*)uset;
+    return set.getRangeCount() + USetAccess::getStringCount(set);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_getItem(const USet* uset, int32_t itemIndex,
+             UChar32* start, UChar32* end,
+             UChar* str, int32_t strCapacity,
+             UErrorCode* ec) {
+    if (U_FAILURE(*ec)) return 0;
+    const UnicodeSet& set = *(const UnicodeSet*)uset;
+    int32_t rangeCount;
+
+    if (itemIndex < 0) {
+        *ec = U_ILLEGAL_ARGUMENT_ERROR;
+        return -1;
+    } else if (itemIndex < (rangeCount = set.getRangeCount())) {
+        *start = set.getRangeStart(itemIndex);
+        *end = set.getRangeEnd(itemIndex);
+        return 0;
+    } else {
+        itemIndex -= rangeCount;
+        if (itemIndex < USetAccess::getStringCount(set)) {
+            const UnicodeString* s = USetAccess::getString(set, itemIndex);
+            return s->extract(str, strCapacity, *ec);
+        } else {
+            *ec = U_INDEX_OUTOFBOUNDS_ERROR;
+            return -1;
+        }
+    }
+}
+
+//U_CAPI int32_t U_EXPORT2
+//uset_getRangeCount(const USet* set) {
+//    return ((const UnicodeSet*) set)->getRangeCount();
+//}
+//
+//U_CAPI UBool U_EXPORT2
+//uset_getRange(const USet* set, int32_t rangeIndex,
+//              UChar32* pStart, UChar32* pEnd) {
+//    if ((uint32_t) rangeIndex >= (uint32_t) uset_getRangeCount(set)) {
+//        return FALSE;
+//    }
+//    const UnicodeSet* us = (const UnicodeSet*) set;
+//    *pStart = us->getRangeStart(rangeIndex);
+//    *pEnd = us->getRangeEnd(rangeIndex);
+//    return TRUE;
+//}
+
+/*
+ * Serialize a USet into 16-bit units.
+ * Store BMP code points as themselves with one 16-bit unit each.
+ *
+ * Important: the code points in the array are in ascending order,
+ * therefore all BMP code points precede all supplementary code points.
+ *
+ * Store each supplementary code point in 2 16-bit units,
+ * simply with higher-then-lower 16-bit halfs.
+ *
+ * Precede the entire list with the length.
+ * If there are supplementary code points, then set bit 15 in the length
+ * and add the bmpLength between it and the array.
+ *
+ * In other words:
+ * - all BMP:            (length=bmpLength) BMP, .., BMP
+ * - some supplementary: (length|0x8000) (bmpLength<length) BMP, .., BMP, supp-high, supp-low, ..
+ */
+U_CAPI int32_t U_EXPORT2
+uset_serialize(const USet* set, uint16_t* dest, int32_t destCapacity, UErrorCode* ec) {
+    if (ec==NULL || U_FAILURE(*ec)) {
+        return 0;
+    }
+
+    return ((const UnicodeSet*) set)->UnicodeSet::serialize(dest, destCapacity,* ec);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_getSerializedSet(USerializedSet* fillSet, const uint16_t* src, int32_t srcLength) {
+    int32_t length;
+
+    if(fillSet==NULL) {
+        return FALSE;
+    }
+    if(src==NULL || srcLength<=0) {
+        fillSet->length=fillSet->bmpLength=0;
+        return FALSE;
+    }
+
+    length=*src++;
+    if(length&0x8000) {
+        /* there are supplementary values */
+        length&=0x7fff;
+        if(srcLength<(2+length)) {
+            fillSet->length=fillSet->bmpLength=0;
+            return FALSE;
+        }
+        fillSet->bmpLength=*src++;
+    } else {
+        /* only BMP values */
+        if(srcLength<(1+length)) {
+            fillSet->length=fillSet->bmpLength=0;
+            return FALSE;
+        }
+        fillSet->bmpLength=length;
+    }
+    fillSet->array=src;
+    fillSet->length=length;
+    return TRUE;
+}
+
+U_CAPI void U_EXPORT2
+uset_setSerializedToOne(USerializedSet* fillSet, UChar32 c) {
+    if(fillSet==NULL || (uint32_t)c>0x10ffff) {
+        return;
+    }
+
+    fillSet->array=fillSet->staticArray;
+    if(c<0xffff) {
+        fillSet->bmpLength=fillSet->length=2;
+        fillSet->staticArray[0]=(uint16_t)c;
+        fillSet->staticArray[1]=(uint16_t)c+1;
+    } else if(c==0xffff) {
+        fillSet->bmpLength=1;
+        fillSet->length=3;
+        fillSet->staticArray[0]=0xffff;
+        fillSet->staticArray[1]=1;
+        fillSet->staticArray[2]=0;
+    } else if(c<0x10ffff) {
+        fillSet->bmpLength=0;
+        fillSet->length=4;
+        fillSet->staticArray[0]=(uint16_t)(c>>16);
+        fillSet->staticArray[1]=(uint16_t)c;
+        ++c;
+        fillSet->staticArray[2]=(uint16_t)(c>>16);
+        fillSet->staticArray[3]=(uint16_t)c;
+    } else /* c==0x10ffff */ {
+        fillSet->bmpLength=0;
+        fillSet->length=2;
+        fillSet->staticArray[0]=0x10;
+        fillSet->staticArray[1]=0xffff;
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+uset_serializedContains(const USerializedSet* set, UChar32 c) {
+    const uint16_t* array;
+
+    if(set==NULL || (uint32_t)c>0x10ffff) {
+        return FALSE;
+    }
+
+    array=set->array;
+    if(c<=0xffff) {
+        /* find c in the BMP part */
+        int32_t lo = 0;
+        int32_t hi = set->bmpLength-1;
+        if (c < array[0]) {
+            hi = 0;
+        } else if (c < array[hi]) {
+            for(;;) {
+                int32_t i = (lo + hi) >> 1;
+                if (i == lo) {
+                    break;  // Done!
+                } else if (c < array[i]) {
+                    hi = i;
+                } else {
+                    lo = i;
+                }
+            }
+        } else {
+            hi += 1;
+        }
+        return (UBool)(hi&1);
+    } else {
+        /* find c in the supplementary part */
+        uint16_t high=(uint16_t)(c>>16), low=(uint16_t)c;
+        int32_t base = set->bmpLength;
+        int32_t lo = 0;
+        int32_t hi = set->length - 2 - base;
+        if (high < array[base] || (high==array[base] && low<array[base+1])) {
+            hi = 0;
+        } else if (high < array[base+hi] || (high==array[base+hi] && low<array[base+hi+1])) {
+            for (;;) {
+                int32_t i = ((lo + hi) >> 1) & ~1;  // Guarantee even result
+                int32_t iabs = i + base;
+                if (i == lo) {
+                    break;  // Done!
+                } else if (high < array[iabs] || (high==array[iabs] && low<array[iabs+1])) {
+                    hi = i;
+                } else {
+                    lo = i;
+                }
+            }
+        } else {
+            hi += 2;
+        }
+        /* count pairs of 16-bit units even per BMP and check if the number of pairs is odd */
+        return (UBool)(((hi+(base<<1))&2)!=0);
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_getSerializedRangeCount(const USerializedSet* set) {
+    if(set==NULL) {
+        return 0;
+    }
+
+    return (set->bmpLength+(set->length-set->bmpLength)/2+1)/2;
+}
+
+U_CAPI UBool U_EXPORT2
+uset_getSerializedRange(const USerializedSet* set, int32_t rangeIndex,
+                        UChar32* pStart, UChar32* pEnd) {
+    const uint16_t* array;
+    int32_t bmpLength, length;
+
+    if(set==NULL || rangeIndex<0 || pStart==NULL || pEnd==NULL) {
+        return FALSE;
+    }
+
+    array=set->array;
+    length=set->length;
+    bmpLength=set->bmpLength;
+
+    rangeIndex*=2; /* address start/limit pairs */
+    if(rangeIndex<bmpLength) {
+        *pStart=array[rangeIndex++];
+        if(rangeIndex<bmpLength) {
+            *pEnd=array[rangeIndex]-1;
+        } else if(rangeIndex<length) {
+            *pEnd=((((int32_t)array[rangeIndex])<<16)|array[rangeIndex+1])-1;
+        } else {
+            *pEnd=0x10ffff;
+        }
+        return TRUE;
+    } else {
+        rangeIndex-=bmpLength;
+        rangeIndex*=2; /* address pairs of pairs of units */
+        length-=bmpLength;
+        if(rangeIndex<length) {
+            array+=bmpLength;
+            *pStart=(((int32_t)array[rangeIndex])<<16)|array[rangeIndex+1];
+            rangeIndex+=2;
+            if(rangeIndex<length) {
+                *pEnd=((((int32_t)array[rangeIndex])<<16)|array[rangeIndex+1])-1;
+            } else {
+                *pEnd=0x10ffff;
+            }
+            return TRUE;
+        } else {
+            return FALSE;
+        }
+    }
+}
+
+// TODO The old, internal uset.c had an efficient uset_containsOne function.
+// Returned the one and only code point, or else -1 or something.
+// Consider adding such a function to both C and C++ UnicodeSet/uset.
+// See tools/gennorm/store.c for usage, now usetContainsOne there.
+
+// TODO Investigate incorporating this code into UnicodeSet to improve
+// efficiency.
+// ---
+// #define USET_GROW_DELTA 20
+// 
+// static int32_t
+// findChar(const UChar32* array, int32_t length, UChar32 c) {
+//     int32_t i;
+// 
+//     /* check the last range limit first for more efficient appending */
+//     if(length>0) {
+//         if(c>=array[length-1]) {
+//             return length;
+//         }
+// 
+//         /* do not check the last range limit again in the loop below */
+//         --length;
+//     }
+// 
+//     for(i=0; i<length && c>=array[i]; ++i) {}
+//     return i;
+// }
+// 
+// static UBool
+// addRemove(USet* set, UChar32 c, int32_t doRemove) {
+//     int32_t i, length, more;
+// 
+//     if(set==NULL || (uint32_t)c>0x10ffff) {
+//         return FALSE;
+//     }
+// 
+//     length=set->length;
+//     i=findChar(set->array, length, c);
+//     if((i&1)^doRemove) {
+//         /* c is already in the set */
+//         return TRUE;
+//     }
+// 
+//     /* how many more array items do we need? */
+//     if(i<length && (c+1)==set->array[i]) {
+//         /* c is just before the following range, extend that in-place by one */
+//         set->array[i]=c;
+//         if(i>0) {
+//             --i;
+//             if(c==set->array[i]) {
+//                 /* the previous range collapsed, remove it */
+//                 set->length=length-=2;
+//                 if(i<length) {
+//                     uprv_memmove(set->array+i, set->array+i+2, (length-i)*4);
+//                 }
+//             }
+//         }
+//         return TRUE;
+//     } else if(i>0 && c==set->array[i-1]) {
+//         /* c is just after the previous range, extend that in-place by one */
+//         if(++c<=0x10ffff) {
+//             set->array[i-1]=c;
+//             if(i<length && c==set->array[i]) {
+//                 /* the following range collapsed, remove it */
+//                 --i;
+//                 set->length=length-=2;
+//                 if(i<length) {
+//                     uprv_memmove(set->array+i, set->array+i+2, (length-i)*4);
+//                 }
+//             }
+//         } else {
+//             /* extend the previous range (had limit 0x10ffff) to the end of Unicode */
+//             set->length=i-1;
+//         }
+//         return TRUE;
+//     } else if(i==length && c==0x10ffff) {
+//         /* insert one range limit c */
+//         more=1;
+//     } else {
+//         /* insert two range limits c, c+1 */
+//         more=2;
+//     }
+// 
+//     /* insert <more> range limits */
+//     if(length+more>set->capacity) {
+//         /* reallocate */
+//         int32_t newCapacity=set->capacity+set->capacity/2+USET_GROW_DELTA;
+//         UChar32* newArray=(UChar32* )uprv_malloc(newCapacity*4);
+//         if(newArray==NULL) {
+//             return FALSE;
+//         }
+//         set->capacity=newCapacity;
+//         uprv_memcpy(newArray, set->array, length*4);
+// 
+//         if(set->array!=set->staticBuffer) {
+//             uprv_free(set->array);
+//         }
+//         set->array=newArray;
+//     }
+// 
+//     if(i<length) {
+//         uprv_memmove(set->array+i+more, set->array+i, (length-i)*4);
+//     }
+//     set->array[i]=c;
+//     if(more==2) {
+//         set->array[i+1]=c+1;
+//     }
+//     set->length+=more;
+// 
+//     return TRUE;
+// }
+// 
+// U_CAPI UBool U_EXPORT2
+// uset_add(USet* set, UChar32 c) {
+//     return addRemove(set, c, 0);
+// }
+// 
+// U_CAPI void U_EXPORT2
+// uset_remove(USet* set, UChar32 c) {
+//     addRemove(set, c, 1);
+// }
diff --git a/thirdparty/icu4c/common/uset_imp.h b/thirdparty/icu4c/common/uset_imp.h
new file mode 100644
index 0000000000..7233b9303c
--- /dev/null
+++ b/thirdparty/icu4c/common/uset_imp.h
@@ -0,0 +1,62 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2004-2007, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uset_imp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004sep07
+*   created by: Markus W. Scherer
+*
+*   Internal USet definitions.
+*/
+
+#ifndef __USET_IMP_H__
+#define __USET_IMP_H__
+
+#include "unicode/utypes.h"
+#include "unicode/uset.h"
+
+U_CDECL_BEGIN
+
+typedef void U_CALLCONV
+USetAdd(USet *set, UChar32 c);
+
+typedef void U_CALLCONV
+USetAddRange(USet *set, UChar32 start, UChar32 end);
+
+typedef void U_CALLCONV
+USetAddString(USet *set, const UChar *str, int32_t length);
+
+typedef void U_CALLCONV
+USetRemove(USet *set, UChar32 c);
+
+typedef void U_CALLCONV
+USetRemoveRange(USet *set, UChar32 start, UChar32 end);
+
+/**
+ * Interface for adding items to a USet, to keep low-level code from
+ * statically depending on the USet implementation.
+ * Calls will look like sa->add(sa->set, c);
+ */
+struct USetAdder {
+    USet *set;
+    USetAdd *add;
+    USetAddRange *addRange;
+    USetAddString *addString;
+    USetRemove *remove;
+    USetRemoveRange *removeRange;
+};
+typedef struct USetAdder USetAdder;
+
+U_CDECL_END
+
+#endif
+
diff --git a/thirdparty/icu4c/common/uset_props.cpp b/thirdparty/icu4c/common/uset_props.cpp
new file mode 100644
index 0000000000..f08e760b10
--- /dev/null
+++ b/thirdparty/icu4c/common/uset_props.cpp
@@ -0,0 +1,143 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2002-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  uset_props.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug30
+*   created by: Markus W. Scherer
+*
+*   C wrappers around UnicodeSet functions that are implemented in
+*   uniset_props.cpp, split off for modularization.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "unicode/uset.h"
+#include "unicode/uniset.h"
+#include "cmemory.h"
+#include "unicode/ustring.h"
+#include "unicode/parsepos.h"
+
+U_NAMESPACE_USE
+
+U_CAPI USet* U_EXPORT2
+uset_openPattern(const UChar* pattern, int32_t patternLength,
+                 UErrorCode* ec)
+{
+    UnicodeString pat(patternLength==-1, pattern, patternLength);
+    UnicodeSet* set = new UnicodeSet(pat, *ec);
+    /* test for NULL */
+    if(set == 0) {
+        *ec = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+
+    if (U_FAILURE(*ec)) {
+        delete set;
+        set = NULL;
+    }
+    return (USet*) set;
+}
+
+U_CAPI USet* U_EXPORT2
+uset_openPatternOptions(const UChar* pattern, int32_t patternLength,
+                 uint32_t options,
+                 UErrorCode* ec)
+{
+    UnicodeString pat(patternLength==-1, pattern, patternLength);
+    UnicodeSet* set = new UnicodeSet(pat, options, NULL, *ec);
+    /* test for NULL */
+    if(set == 0) {
+        *ec = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+
+    if (U_FAILURE(*ec)) {
+        delete set;
+        set = NULL;
+    }
+    return (USet*) set;
+}
+
+
+U_CAPI int32_t U_EXPORT2 
+uset_applyPattern(USet *set,
+                  const UChar *pattern, int32_t patternLength,
+                  uint32_t options,
+                  UErrorCode *status){
+
+    // status code needs to be checked since we 
+    // dereference it
+    if(status == NULL || U_FAILURE(*status)){
+        return 0;
+    }
+
+    // check only the set paramenter
+    // if pattern is NULL or null terminate
+    // UnicodeString constructor takes care of it
+    if(set == NULL){
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    UnicodeString pat(pattern, patternLength);
+
+    ParsePosition pos;
+   
+    ((UnicodeSet*) set)->applyPattern(pat, pos, options, NULL, *status);
+    
+    return pos.getIndex();
+}
+
+U_CAPI void U_EXPORT2
+uset_applyIntPropertyValue(USet* set,
+               UProperty prop, int32_t value, UErrorCode* ec) {
+    ((UnicodeSet*) set)->applyIntPropertyValue(prop, value, *ec);
+}
+
+U_CAPI void U_EXPORT2
+uset_applyPropertyAlias(USet* set,
+                        const UChar *prop, int32_t propLength,
+                        const UChar *value, int32_t valueLength,
+            UErrorCode* ec) {
+
+    UnicodeString p(prop, propLength);
+    UnicodeString v(value, valueLength);
+
+    ((UnicodeSet*) set)->applyPropertyAlias(p, v, *ec);
+}
+
+U_CAPI UBool U_EXPORT2
+uset_resemblesPattern(const UChar *pattern, int32_t patternLength,
+                      int32_t pos) {
+
+    UnicodeString pat(pattern, patternLength);
+
+    return ((pos+1) < pat.length() &&
+            pat.charAt(pos) == (UChar)91/*[*/) ||
+            UnicodeSet::resemblesPattern(pat, pos);
+}
+
+U_CAPI int32_t U_EXPORT2
+uset_toPattern(const USet* set,
+               UChar* result, int32_t resultCapacity,
+               UBool escapeUnprintable,
+               UErrorCode* ec) {
+    UnicodeString pat;
+    ((const UnicodeSet*) set)->toPattern(pat, escapeUnprintable);
+    return pat.extract(result, resultCapacity, *ec);
+}
+
+U_CAPI void U_EXPORT2
+uset_closeOver(USet* set, int32_t attributes) {
+    ((UnicodeSet*) set)->UnicodeSet::closeOver(attributes);
+}
diff --git a/thirdparty/icu4c/common/usetiter.cpp b/thirdparty/icu4c/common/usetiter.cpp
new file mode 100644
index 0000000000..7915169049
--- /dev/null
+++ b/thirdparty/icu4c/common/usetiter.cpp
@@ -0,0 +1,152 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2006, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+#include "unicode/usetiter.h"
+#include "unicode/uniset.h"
+#include "unicode/unistr.h"
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UnicodeSetIterator)
+
+/**
+ * Create an iterator
+ * @param set set to iterate over
+ */
+UnicodeSetIterator::UnicodeSetIterator(const UnicodeSet& uSet) {
+    cpString  = NULL;
+    reset(uSet);
+}
+
+/**
+ * Create an iterator. Convenience for when the contents are to be set later.
+ */
+UnicodeSetIterator::UnicodeSetIterator() {
+    this->set = NULL;
+    cpString  = NULL;
+    reset();
+}
+
+UnicodeSetIterator::~UnicodeSetIterator() {
+    delete cpString;
+}
+
+/**
+ * Returns the next element in the set.
+ * @return true if there was another element in the set.
+ * if so, if codepoint == IS_STRING, the value is a string in the string field
+ * else the value is a single code point in the codepoint field.
+ * <br>You are guaranteed that the codepoints are in sorted order, and the strings are in sorted order,
+ * and that all code points are returned before any strings are returned.
+ * <br>Note also that the codepointEnd is undefined after calling this method.
+ */
+UBool UnicodeSetIterator::next() {
+    if (nextElement <= endElement) {
+        codepoint = codepointEnd = nextElement++;
+        string = NULL;
+        return TRUE;
+    }
+    if (range < endRange) {
+        loadRange(++range);
+        codepoint = codepointEnd = nextElement++;
+        string = NULL;
+        return TRUE;
+    }
+
+    if (nextString >= stringCount) return FALSE;
+    codepoint = (UChar32)IS_STRING; // signal that value is actually a string
+    string = (const UnicodeString*) set->strings->elementAt(nextString++);
+    return TRUE;
+}
+
+/**
+ * @return true if there was another element in the set.
+ * if so, if codepoint == IS_STRING, the value is a string in the string field
+ * else the value is a range of codepoints in the <codepoint, codepointEnd> fields.
+ * <br>Note that the codepoints are in sorted order, and the strings are in sorted order,
+ * and that all code points are returned before any strings are returned.
+ * <br>You are guaranteed that the ranges are in sorted order, and the strings are in sorted order,
+ * and that all ranges are returned before any strings are returned.
+ * <br>You are also guaranteed that ranges are disjoint and non-contiguous.
+ * <br>Note also that the codepointEnd is undefined after calling this method.
+ */
+UBool UnicodeSetIterator::nextRange() {
+    string = NULL;
+    if (nextElement <= endElement) {
+        codepointEnd = endElement;
+        codepoint = nextElement;
+        nextElement = endElement+1;
+        return TRUE;
+    }
+    if (range < endRange) {
+        loadRange(++range);
+        codepointEnd = endElement;
+        codepoint = nextElement;
+        nextElement = endElement+1;
+        return TRUE;
+    }
+
+    if (nextString >= stringCount) return FALSE;
+    codepoint = (UChar32)IS_STRING; // signal that value is actually a string
+    string = (const UnicodeString*) set->strings->elementAt(nextString++);
+    return TRUE;
+}
+
+/**
+ *@param set the set to iterate over. This allows reuse of the iterator.
+ */
+void UnicodeSetIterator::reset(const UnicodeSet& uSet) {
+    this->set = &uSet;
+    reset();
+}
+
+/**
+ * Resets to the start, to allow the iteration to start over again.
+ */
+void UnicodeSetIterator::reset() {
+    if (set == NULL) {
+        // Set up indices to empty iteration
+        endRange = -1;
+        stringCount = 0;
+    } else {
+        endRange = set->getRangeCount() - 1;
+        stringCount = set->stringsSize();
+    }
+    range = 0;
+    endElement = -1;
+    nextElement = 0;            
+    if (endRange >= 0) {
+        loadRange(range);
+    }
+    nextString = 0;
+    string = NULL;
+}
+
+void UnicodeSetIterator::loadRange(int32_t iRange) {
+    nextElement = set->getRangeStart(iRange);
+    endElement = set->getRangeEnd(iRange);
+}
+
+
+const UnicodeString& UnicodeSetIterator::getString()  {
+    if (string==NULL && codepoint!=(UChar32)IS_STRING) {
+       if (cpString == NULL) {
+          cpString = new UnicodeString();
+       }
+       if (cpString != NULL) {
+          cpString->setTo((UChar32)codepoint);
+       }
+       string = cpString;
+    }
+    return *string;
+}
+
+U_NAMESPACE_END
+
+//eof
diff --git a/thirdparty/icu4c/common/ushape.cpp b/thirdparty/icu4c/common/ushape.cpp
new file mode 100644
index 0000000000..ae13b5c118
--- /dev/null
+++ b/thirdparty/icu4c/common/ushape.cpp
@@ -0,0 +1,1728 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ ******************************************************************************
+ *
+ *   Copyright (C) 2000-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ ******************************************************************************
+ *   file name:  ushape.cpp
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2000jun29
+ *   created by: Markus W. Scherer
+ *
+ *   Arabic letter shaping implemented by Ayman Roshdy
+ */
+
+#include "unicode/utypes.h"
+#include "unicode/uchar.h"
+#include "unicode/ustring.h"
+#include "unicode/ushape.h"
+#include "cmemory.h"
+#include "putilimp.h"
+#include "ustr_imp.h"
+#include "ubidi_props.h"
+#include "uassert.h"
+
+/*
+ * This implementation is designed for 16-bit Unicode strings.
+ * The main assumption is that the Arabic characters and their
+ * presentation forms each fit into a single UChar.
+ * With UTF-8, they occupy 2 or 3 bytes, and more than the ASCII
+ * characters.
+ */
+
+/*
+ * ### TODO in general for letter shaping:
+ * - the letter shaping code is UTF-16-unaware; needs update
+ *   + especially invertBuffer()?!
+ * - needs to handle the "Arabic Tail" that is used in some legacy codepages
+ *   as a glyph fragment of wide-glyph letters
+ *   + IBM Unicode conversion tables map it to U+200B (ZWSP)
+ *   + IBM Egypt has proposed to encode the tail in Unicode among Arabic Presentation Forms
+ *   + Unicode 3.2 added U+FE73 ARABIC TAIL FRAGMENT
+ */
+
+/* definitions for Arabic letter shaping ------------------------------------ */
+
+#define IRRELEVANT 4
+#define LAMTYPE    16
+#define ALEFTYPE   32
+#define LINKR      1
+#define LINKL      2
+#define APRESENT   8
+#define SHADDA     64
+#define CSHADDA    128
+#define COMBINE    (SHADDA+CSHADDA)
+
+#define HAMZAFE_CHAR       0xfe80
+#define HAMZA06_CHAR       0x0621
+#define YEH_HAMZA_CHAR     0x0626
+#define YEH_HAMZAFE_CHAR   0xFE89
+#define LAMALEF_SPACE_SUB  0xFFFF
+#define TASHKEEL_SPACE_SUB 0xFFFE
+#define NEW_TAIL_CHAR      0xFE73
+#define OLD_TAIL_CHAR      0x200B
+#define LAM_CHAR           0x0644
+#define SPACE_CHAR         0x0020
+#define SHADDA_CHAR        0xFE7C
+#define TATWEEL_CHAR       0x0640
+#define SHADDA_TATWEEL_CHAR  0xFE7D
+#define SHADDA06_CHAR      0x0651
+
+#define SHAPE_MODE   0
+#define DESHAPE_MODE 1
+
+struct uShapeVariables {
+     UChar tailChar;
+     uint32_t uShapeLamalefBegin;
+     uint32_t uShapeLamalefEnd;
+     uint32_t uShapeTashkeelBegin;
+     uint32_t uShapeTashkeelEnd;
+     int spacesRelativeToTextBeginEnd;
+};
+
+static const uint8_t tailFamilyIsolatedFinal[] = {
+    /* FEB1 */ 1,
+    /* FEB2 */ 1,
+    /* FEB3 */ 0,
+    /* FEB4 */ 0,
+    /* FEB5 */ 1,
+    /* FEB6 */ 1,
+    /* FEB7 */ 0,
+    /* FEB8 */ 0,
+    /* FEB9 */ 1,
+    /* FEBA */ 1,
+    /* FEBB */ 0,
+    /* FEBC */ 0,
+    /* FEBD */ 1,
+    /* FEBE */ 1
+};
+
+static const uint8_t tashkeelMedial[] = {
+    /* FE70 */ 0,
+    /* FE71 */ 1,
+    /* FE72 */ 0,
+    /* FE73 */ 0,
+    /* FE74 */ 0,
+    /* FE75 */ 0,
+    /* FE76 */ 0,
+    /* FE77 */ 1,
+    /* FE78 */ 0,
+    /* FE79 */ 1,
+    /* FE7A */ 0,
+    /* FE7B */ 1,
+    /* FE7C */ 0,
+    /* FE7D */ 1,
+    /* FE7E */ 0,
+    /* FE7F */ 1
+};
+
+static const UChar yehHamzaToYeh[] =
+{
+/* isolated*/ 0xFEEF,
+/* final   */ 0xFEF0
+};
+
+static const uint8_t IrrelevantPos[] = {
+    0x0, 0x2, 0x4, 0x6,
+    0x8, 0xA, 0xC, 0xE
+};
+
+
+static const UChar convertLamAlef[] =
+{
+/*FEF5*/    0x0622,
+/*FEF6*/    0x0622,
+/*FEF7*/    0x0623,
+/*FEF8*/    0x0623,
+/*FEF9*/    0x0625,
+/*FEFA*/    0x0625,
+/*FEFB*/    0x0627,
+/*FEFC*/    0x0627
+};
+
+static const UChar araLink[178]=
+{
+  1           + 32 + 256 * 0x11,/*0x0622*/
+  1           + 32 + 256 * 0x13,/*0x0623*/
+  1                + 256 * 0x15,/*0x0624*/
+  1           + 32 + 256 * 0x17,/*0x0625*/
+  1 + 2            + 256 * 0x19,/*0x0626*/
+  1           + 32 + 256 * 0x1D,/*0x0627*/
+  1 + 2            + 256 * 0x1F,/*0x0628*/
+  1                + 256 * 0x23,/*0x0629*/
+  1 + 2            + 256 * 0x25,/*0x062A*/
+  1 + 2            + 256 * 0x29,/*0x062B*/
+  1 + 2            + 256 * 0x2D,/*0x062C*/
+  1 + 2            + 256 * 0x31,/*0x062D*/
+  1 + 2            + 256 * 0x35,/*0x062E*/
+  1                + 256 * 0x39,/*0x062F*/
+  1                + 256 * 0x3B,/*0x0630*/
+  1                + 256 * 0x3D,/*0x0631*/
+  1                + 256 * 0x3F,/*0x0632*/
+  1 + 2            + 256 * 0x41,/*0x0633*/
+  1 + 2            + 256 * 0x45,/*0x0634*/
+  1 + 2            + 256 * 0x49,/*0x0635*/
+  1 + 2            + 256 * 0x4D,/*0x0636*/
+  1 + 2            + 256 * 0x51,/*0x0637*/
+  1 + 2            + 256 * 0x55,/*0x0638*/
+  1 + 2            + 256 * 0x59,/*0x0639*/
+  1 + 2            + 256 * 0x5D,/*0x063A*/
+  0, 0, 0, 0, 0,                /*0x063B-0x063F*/
+  1 + 2,                        /*0x0640*/
+  1 + 2            + 256 * 0x61,/*0x0641*/
+  1 + 2            + 256 * 0x65,/*0x0642*/
+  1 + 2            + 256 * 0x69,/*0x0643*/
+  1 + 2       + 16 + 256 * 0x6D,/*0x0644*/
+  1 + 2            + 256 * 0x71,/*0x0645*/
+  1 + 2            + 256 * 0x75,/*0x0646*/
+  1 + 2            + 256 * 0x79,/*0x0647*/
+  1                + 256 * 0x7D,/*0x0648*/
+  1                + 256 * 0x7F,/*0x0649*/
+  1 + 2            + 256 * 0x81,/*0x064A*/
+         4         + 256 * 1,   /*0x064B*/
+         4 + 128   + 256 * 1,   /*0x064C*/
+         4 + 128   + 256 * 1,   /*0x064D*/
+         4 + 128   + 256 * 1,   /*0x064E*/
+         4 + 128   + 256 * 1,   /*0x064F*/
+         4 + 128   + 256 * 1,   /*0x0650*/
+         4 + 64    + 256 * 3,   /*0x0651*/
+         4         + 256 * 1,   /*0x0652*/
+         4         + 256 * 7,   /*0x0653*/
+         4         + 256 * 8,   /*0x0654*/
+         4         + 256 * 8,   /*0x0655*/
+         4         + 256 * 1,   /*0x0656*/
+  0, 0, 0, 0, 0,                /*0x0657-0x065B*/
+  1                + 256 * 0x85,/*0x065C*/
+  1                + 256 * 0x87,/*0x065D*/
+  1                + 256 * 0x89,/*0x065E*/
+  1                + 256 * 0x8B,/*0x065F*/
+  0, 0, 0, 0, 0,                /*0x0660-0x0664*/
+  0, 0, 0, 0, 0,                /*0x0665-0x0669*/
+  0, 0, 0, 0, 0, 0,             /*0x066A-0x066F*/
+         4         + 256 * 6,   /*0x0670*/
+  1        + 8     + 256 * 0x00,/*0x0671*/
+  1            + 32,            /*0x0672*/
+  1            + 32,            /*0x0673*/
+  0,                            /*0x0674*/
+  1            + 32,            /*0x0675*/
+  1, 1,                         /*0x0676-0x0677*/
+  1 + 2,                        /*0x0678*/
+  1 + 2 + 8        + 256 * 0x16,/*0x0679*/
+  1 + 2 + 8        + 256 * 0x0E,/*0x067A*/
+  1 + 2 + 8        + 256 * 0x02,/*0x067B*/
+  1+2, 1+2,                     /*0x67C-0x067D*/
+  1+2+8+256 * 0x06, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x067E-0x0683*/
+  1+2, 1+2, 1+2+8+256 * 0x2A, 1+2,           /*0x0684-0x0687*/
+  1     + 8        + 256 * 0x38,/*0x0688*/
+  1, 1, 1,                      /*0x0689-0x068B*/
+  1     + 8        + 256 * 0x34,/*0x068C*/
+  1     + 8        + 256 * 0x32,/*0x068D*/
+  1     + 8        + 256 * 0x36,/*0x068E*/
+  1, 1,                         /*0x068F-0x0690*/
+  1     + 8        + 256 * 0x3C,/*0x0691*/
+  1, 1, 1, 1, 1, 1, 1+8+256 * 0x3A, 1,       /*0x0692-0x0699*/
+  1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/
+  1+2, 1+2, 1+2, 1+2,           /*0x069A-0x06A3*/
+  1+2, 1+2, 1+2, 1+2, 1+2, 1+2+8+256 * 0x3E, /*0x06A4-0x06AD*/
+  1+2, 1+2, 1+2, 1+2,           /*0x06A4-0x06AD*/
+  1+2, 1+2+8+256 * 0x42, 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/
+  1+2, 1+2, 1+2, 1+2,           /*0x06AE-0x06B7*/
+  1+2, 1+2,                     /*0x06B8-0x06B9*/
+  1     + 8        + 256 * 0x4E,/*0x06BA*/
+  1 + 2 + 8        + 256 * 0x50,/*0x06BB*/
+  1+2, 1+2,                     /*0x06BC-0x06BD*/
+  1 + 2 + 8        + 256 * 0x5A,/*0x06BE*/
+  1+2,                          /*0x06BF*/
+  1     + 8        + 256 * 0x54,/*0x06C0*/
+  1 + 2 + 8        + 256 * 0x56,/*0x06C1*/
+  1, 1, 1,                      /*0x06C2-0x06C4*/
+  1     + 8        + 256 * 0x90,/*0x06C5*/
+  1     + 8        + 256 * 0x89,/*0x06C6*/
+  1     + 8        + 256 * 0x87,/*0x06C7*/
+  1     + 8        + 256 * 0x8B,/*0x06C8*/
+  1     + 8        + 256 * 0x92,/*0x06C9*/
+  1,                            /*0x06CA*/
+  1     + 8        + 256 * 0x8E,/*0x06CB*/
+  1 + 2 + 8        + 256 * 0xAC,/*0x06CC*/
+  1,                            /*0x06CD*/
+  1+2, 1+2,                     /*0x06CE-0x06CF*/
+  1 + 2 + 8        + 256 * 0x94,/*0x06D0*/
+  1+2,                          /*0x06D1*/
+  1     + 8        + 256 * 0x5E,/*0x06D2*/
+  1     + 8        + 256 * 0x60 /*0x06D3*/
+};
+
+static const uint8_t presALink[] = {
+/***********0*****1*****2*****3*****4*****5*****6*****7*****8*****9*****A*****B*****C*****D*****E*****F*/
+/*FB5*/    0,    1,    0,    0,    0,    0,    0,    1,    2,1 + 2,    0,    0,    0,    0,    0,    0,
+/*FB6*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FB7*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    1,    2,1 + 2,    0,    0,
+/*FB8*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    1,    0,    0,    0,    1,
+/*FB9*/    2,1 + 2,    0,    1,    2,1 + 2,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBA*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBB*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBC*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBD*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBE*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FBF*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    1,    2,1 + 2,
+/*FC0*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FC1*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FC2*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FC3*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FC4*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,
+/*FC5*/    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    0,    4,    4,
+/*FC6*/    4,    4,    4
+};
+
+static const uint8_t presBLink[]=
+{
+/***********0*****1*****2*****3*****4*****5*****6*****7*****8*****9*****A*****B*****C*****D*****E*****F*/
+/*FE7*/1 + 2,1 + 2,1 + 2,    0,1 + 2,    0,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,1 + 2,
+/*FE8*/    0,    0,    1,    0,    1,    0,    1,    0,    1,    0,    1,    2,1 + 2,    0,    1,    0,
+/*FE9*/    1,    2,1 + 2,    0,    1,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,
+/*FEA*/1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    0,    1,    0,    1,    0,
+/*FEB*/    1,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,
+/*FEC*/1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,
+/*FED*/1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,
+/*FEE*/1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    2,1 + 2,    0,    1,    0,
+/*FEF*/    1,    0,    1,    2,1 + 2,    0,    1,    0,    1,    0,    1,    0,    1,    0,    0,    0
+};
+
+static const UChar convertFBto06[] =
+{
+/***********0******1******2******3******4******5******6******7******8******9******A******B******C******D******E******F***/
+/*FB5*/   0x671, 0x671, 0x67B, 0x67B, 0x67B, 0x67B, 0x67E, 0x67E, 0x67E, 0x67E,     0,     0,     0,     0, 0x67A, 0x67A,
+/*FB6*/   0x67A, 0x67A,     0,     0,     0,     0, 0x679, 0x679, 0x679, 0x679,     0,     0,     0,     0,     0,     0,
+/*FB7*/       0,     0,     0,     0,     0,     0,     0,     0,     0,     0, 0x686, 0x686, 0x686, 0x686,     0,     0,
+/*FB8*/       0,     0, 0x68D, 0x68D, 0x68C, 0x68C, 0x68E, 0x68E, 0x688, 0x688, 0x698, 0x698, 0x691, 0x691, 0x6A9, 0x6A9,
+/*FB9*/   0x6A9, 0x6A9, 0x6AF, 0x6AF, 0x6AF, 0x6AF,     0,     0,     0,     0,     0,     0,     0,     0, 0x6BA, 0x6BA,
+/*FBA*/   0x6BB, 0x6BB, 0x6BB, 0x6BB, 0x6C0, 0x6C0, 0x6C1, 0x6C1, 0x6C1, 0x6C1, 0x6BE, 0x6BE, 0x6BE, 0x6BE, 0x6d2, 0x6D2,
+/*FBB*/   0x6D3, 0x6D3,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,
+/*FBC*/       0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,
+/*FBD*/       0,     0,     0,     0,     0,     0,     0, 0x6C7, 0x6C7, 0x6C6, 0x6C6, 0x6C8, 0x6C8,     0, 0x6CB, 0x6CB,
+/*FBE*/   0x6C5, 0x6C5, 0x6C9, 0x6C9, 0x6D0, 0x6D0, 0x6D0, 0x6D0,     0,     0,     0,     0,     0,     0,     0,     0,
+/*FBF*/       0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0,     0, 0x6CC, 0x6CC, 0x6CC, 0x6CC
+};
+
+static const UChar convertFEto06[] =
+{
+/***********0******1******2******3******4******5******6******7******8******9******A******B******C******D******E******F***/
+/*FE7*/   0x64B, 0x64B, 0x64C, 0x64C, 0x64D, 0x64D, 0x64E, 0x64E, 0x64F, 0x64F, 0x650, 0x650, 0x651, 0x651, 0x652, 0x652,
+/*FE8*/   0x621, 0x622, 0x622, 0x623, 0x623, 0x624, 0x624, 0x625, 0x625, 0x626, 0x626, 0x626, 0x626, 0x627, 0x627, 0x628,
+/*FE9*/   0x628, 0x628, 0x628, 0x629, 0x629, 0x62A, 0x62A, 0x62A, 0x62A, 0x62B, 0x62B, 0x62B, 0x62B, 0x62C, 0x62C, 0x62C,
+/*FEA*/   0x62C, 0x62D, 0x62D, 0x62D, 0x62D, 0x62E, 0x62E, 0x62E, 0x62E, 0x62F, 0x62F, 0x630, 0x630, 0x631, 0x631, 0x632,
+/*FEB*/   0x632, 0x633, 0x633, 0x633, 0x633, 0x634, 0x634, 0x634, 0x634, 0x635, 0x635, 0x635, 0x635, 0x636, 0x636, 0x636,
+/*FEC*/   0x636, 0x637, 0x637, 0x637, 0x637, 0x638, 0x638, 0x638, 0x638, 0x639, 0x639, 0x639, 0x639, 0x63A, 0x63A, 0x63A,
+/*FED*/   0x63A, 0x641, 0x641, 0x641, 0x641, 0x642, 0x642, 0x642, 0x642, 0x643, 0x643, 0x643, 0x643, 0x644, 0x644, 0x644,
+/*FEE*/   0x644, 0x645, 0x645, 0x645, 0x645, 0x646, 0x646, 0x646, 0x646, 0x647, 0x647, 0x647, 0x647, 0x648, 0x648, 0x649,
+/*FEF*/   0x649, 0x64A, 0x64A, 0x64A, 0x64A, 0x65C, 0x65C, 0x65D, 0x65D, 0x65E, 0x65E, 0x65F, 0x65F
+};
+
+static const uint8_t shapeTable[4][4][4]=
+{
+  { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,1} },
+  { {0,0,2,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} },
+  { {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,3} },
+  { {0,0,1,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} }
+};
+
+/*
+ * This function shapes European digits to Arabic-Indic digits
+ * in-place, writing over the input characters.
+ * Since we know that we are only looking for BMP code points,
+ * we can safely just work with code units (again, at least UTF-16).
+ */
+static void
+_shapeToArabicDigitsWithContext(UChar *s, int32_t length,
+                                UChar digitBase,
+                                UBool isLogical, UBool lastStrongWasAL) {
+    int32_t i;
+    UChar c;
+
+    digitBase-=0x30;
+
+    /* the iteration direction depends on the type of input */
+    if(isLogical) {
+        for(i=0; i<length; ++i) {
+            c=s[i];
+            switch(ubidi_getClass(c)) {
+            case U_LEFT_TO_RIGHT: /* L */
+            case U_RIGHT_TO_LEFT: /* R */
+                lastStrongWasAL=FALSE;
+                break;
+            case U_RIGHT_TO_LEFT_ARABIC: /* AL */
+                lastStrongWasAL=TRUE;
+                break;
+            case U_EUROPEAN_NUMBER: /* EN */
+                if(lastStrongWasAL && (uint32_t)(c-0x30)<10) {
+                    s[i]=(UChar)(digitBase+c); /* digitBase+(c-0x30) - digitBase was modified above */
+                }
+                break;
+            default :
+                break;
+            }
+        }
+    } else {
+        for(i=length; i>0; /* pre-decrement in the body */) {
+            c=s[--i];
+            switch(ubidi_getClass(c)) {
+            case U_LEFT_TO_RIGHT: /* L */
+            case U_RIGHT_TO_LEFT: /* R */
+                lastStrongWasAL=FALSE;
+                break;
+            case U_RIGHT_TO_LEFT_ARABIC: /* AL */
+                lastStrongWasAL=TRUE;
+                break;
+            case U_EUROPEAN_NUMBER: /* EN */
+                if(lastStrongWasAL && (uint32_t)(c-0x30)<10) {
+                    s[i]=(UChar)(digitBase+c); /* digitBase+(c-0x30) - digitBase was modified above */
+                }
+                break;
+            default :
+                break;
+            }
+        }
+    }
+}
+
+/*
+ *Name     : invertBuffer
+ *Function : This function inverts the buffer, it's used
+ *           in case the user specifies the buffer to be
+ *           U_SHAPE_TEXT_DIRECTION_LOGICAL
+ */
+static void
+invertBuffer(UChar *buffer, int32_t size, uint32_t /*options*/, int32_t lowlimit, int32_t highlimit) {
+    UChar temp;
+    int32_t i=0,j=0;
+    for(i=lowlimit,j=size-highlimit-1;i<j;i++,j--) {
+        temp = buffer[i];
+        buffer[i] = buffer[j];
+        buffer[j] = temp;
+    }
+}
+
+/*
+ *Name     : changeLamAlef
+ *Function : Converts the Alef characters into an equivalent
+ *           LamAlef location in the 0x06xx Range, this is an
+ *           intermediate stage in the operation of the program
+ *           later it'll be converted into the 0xFExx LamAlefs
+ *           in the shaping function.
+ */
+static inline UChar
+changeLamAlef(UChar ch) {
+    switch(ch) {
+    case 0x0622 :
+        return 0x065C;
+    case 0x0623 :
+        return 0x065D;
+    case 0x0625 :
+        return 0x065E;
+    case 0x0627 :
+        return 0x065F;
+    }
+    return 0;
+}
+
+/*
+ *Name     : getLink
+ *Function : Resolves the link between the characters as
+ *           Arabic characters have four forms :
+ *           Isolated, Initial, Middle and Final Form
+ */
+static UChar
+getLink(UChar ch) {
+    if(ch >= 0x0622 && ch <= 0x06D3) {
+        return(araLink[ch-0x0622]);
+    } else if(ch == 0x200D) {
+        return(3);
+    } else if(ch >= 0x206D && ch <= 0x206F) {
+        return(4);
+    }else if(ch >= 0xFB50 && ch <= 0xFC62) {
+        return(presALink[ch-0xFB50]);
+    } else if(ch >= 0xFE70 && ch <= 0xFEFC) {
+        return(presBLink[ch-0xFE70]);
+    }else {
+        return(0);
+    }
+}
+
+/*
+ *Name     : countSpaces
+ *Function : Counts the number of spaces
+ *           at each end of the logical buffer
+ */
+static void
+countSpaces(UChar *dest, int32_t size, uint32_t /*options*/, int32_t *spacesCountl, int32_t *spacesCountr) {
+    int32_t i = 0;
+    int32_t countl = 0,countr = 0;
+    while((dest[i] == SPACE_CHAR) && (countl < size)) {
+       countl++;
+       i++;
+    }
+    if (countl < size) {  /* the entire buffer is not all space */
+        while(dest[size-1] == SPACE_CHAR) {
+            countr++;
+            size--;
+        }
+    }
+    *spacesCountl = countl;
+    *spacesCountr = countr;
+}
+
+/*
+ *Name     : isTashkeelChar
+ *Function : Returns 1 for Tashkeel characters in 06 range else return 0
+ */
+static inline int32_t
+isTashkeelChar(UChar ch) {
+    return (int32_t)( ch>=0x064B && ch<= 0x0652 );
+}
+
+/*
+ *Name     : isTashkeelCharFE
+ *Function : Returns 1 for Tashkeel characters in FE range else return 0
+ */
+static inline int32_t
+isTashkeelCharFE(UChar ch) {
+    return (int32_t)( ch>=0xFE70 && ch<= 0xFE7F );
+}
+
+/*
+ *Name     : isAlefChar
+ *Function : Returns 1 for Alef characters else return 0
+ */
+static inline int32_t
+isAlefChar(UChar ch) {
+    return (int32_t)( (ch==0x0622)||(ch==0x0623)||(ch==0x0625)||(ch==0x0627) );
+}
+
+/*
+ *Name     : isLamAlefChar
+ *Function : Returns 1 for LamAlef characters else return 0
+ */
+static inline int32_t
+isLamAlefChar(UChar ch) {
+    return (int32_t)((ch>=0xFEF5)&&(ch<=0xFEFC) );
+}
+
+/*BIDI
+ *Name     : isTailChar
+ *Function : returns 1 if the character matches one of the tail characters (0xfe73 or 0x200b) otherwise returns 0
+ */
+
+static inline int32_t
+isTailChar(UChar ch) {
+    if(ch == OLD_TAIL_CHAR || ch == NEW_TAIL_CHAR){
+            return 1;
+    }else{
+            return 0;
+    }
+}
+
+/*BIDI
+ *Name     : isSeenTailFamilyChar
+ *Function : returns 1 if the character is a seen family isolated character
+ *           in the FE range otherwise returns 0
+ */
+
+static inline int32_t
+isSeenTailFamilyChar(UChar ch) {
+    if(ch >= 0xfeb1 && ch < 0xfebf){
+            return tailFamilyIsolatedFinal [ch - 0xFEB1];
+    }else{
+            return 0;
+    }
+}
+
+ /* Name     : isSeenFamilyChar
+  * Function : returns 1 if the character is a seen family character in the Unicode
+  *            06 range otherwise returns 0
+ */
+
+static inline int32_t
+isSeenFamilyChar(UChar  ch){
+    if(ch >= 0x633 && ch <= 0x636){
+        return 1;
+    }else {
+        return 0;
+    }
+}
+
+/*Start of BIDI*/
+/*
+ *Name     : isAlefMaksouraChar
+ *Function : returns 1 if the character is a Alef Maksoura Final or isolated
+ *           otherwise returns 0
+ */
+static inline int32_t
+isAlefMaksouraChar(UChar ch) {
+    return (int32_t)( (ch == 0xFEEF) || ( ch == 0xFEF0) || (ch == 0x0649));
+}
+
+/*
+ * Name     : isYehHamzaChar
+ * Function : returns 1 if the character is a yehHamza isolated or yehhamza
+ *            final is found otherwise returns 0
+ */
+static inline int32_t
+isYehHamzaChar(UChar ch) {
+    if((ch==0xFE89)||(ch==0xFE8A)){
+        return 1;
+    }else{
+        return 0;
+    }
+}
+
+ /*
+ * Name: isTashkeelOnTatweelChar
+ * Function: Checks if the Tashkeel Character is on Tatweel or not,if the
+ *           Tashkeel on tatweel (FE range), it returns 1 else if the
+ *           Tashkeel with shadda on tatweel (FC range)return 2 otherwise
+ *           returns 0
+ */
+static inline int32_t
+isTashkeelOnTatweelChar(UChar ch){
+    if(ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75 && ch != SHADDA_TATWEEL_CHAR)
+    {
+        return tashkeelMedial [ch - 0xFE70];
+    }else if( (ch >= 0xfcf2 && ch <= 0xfcf4) || (ch == SHADDA_TATWEEL_CHAR)) {
+        return 2;
+    }else{
+        return 0;
+    }
+}
+
+/*
+ * Name: isIsolatedTashkeelChar
+ * Function: Checks if the Tashkeel Character is in the isolated form
+ *           (i.e. Unicode FE range) returns 1 else if the Tashkeel
+ *           with shadda is in the isolated form (i.e. Unicode FC range)
+ *           returns 2 otherwise returns 0
+ */
+static inline int32_t
+isIsolatedTashkeelChar(UChar ch){
+    if(ch >= 0xfe70 && ch <= 0xfe7f && ch != NEW_TAIL_CHAR && ch != 0xFE75){
+        return (1 - tashkeelMedial [ch - 0xFE70]);
+    }else if(ch >= 0xfc5e && ch <= 0xfc63){
+        return 1;
+    }else{
+        return 0;
+    }
+}
+
+
+
+
+/*
+ *Name     : calculateSize
+ *Function : This function calculates the destSize to be used in preflighting
+ *           when the destSize is equal to 0
+ *           It is used also to calculate the new destsize in case the
+ *           destination buffer will be resized.
+ */
+
+static int32_t
+calculateSize(const UChar *source, int32_t sourceLength,
+int32_t destSize,uint32_t options) {
+    int32_t i = 0;
+
+    int lamAlefOption = 0;
+    int tashkeelOption = 0;
+
+    destSize = sourceLength;
+
+    if (((options&U_SHAPE_LETTERS_MASK) == U_SHAPE_LETTERS_SHAPE ||
+        ((options&U_SHAPE_LETTERS_MASK) == U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED )) &&
+        ((options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_RESIZE )){
+            lamAlefOption = 1;
+    }
+    if((options&U_SHAPE_LETTERS_MASK) == U_SHAPE_LETTERS_SHAPE &&
+       ((options&U_SHAPE_TASHKEEL_MASK) == U_SHAPE_TASHKEEL_RESIZE ) ){
+            tashkeelOption = 1;
+        }
+
+    if(lamAlefOption || tashkeelOption){
+        if((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_VISUAL_LTR) {
+            for(i=0;i<sourceLength;i++) {
+                if( ((isAlefChar(source[i]))&& (i<(sourceLength-1)) &&(source[i+1] == LAM_CHAR)) || (isTashkeelCharFE(source[i])) ) {
+                        destSize--;
+                    }
+                }
+            }else if((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL) {
+                for(i=0;i<sourceLength;i++) {
+                    if( ( (source[i] == LAM_CHAR) && (i<(sourceLength-1)) && (isAlefChar(source[i+1]))) || (isTashkeelCharFE(source[i])) ) {
+                        destSize--;
+                    }
+                }
+            }
+        }
+
+    if ((options&U_SHAPE_LETTERS_MASK) == U_SHAPE_LETTERS_UNSHAPE){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_RESIZE){
+            for(i=0;i<sourceLength;i++) {
+                if(isLamAlefChar(source[i]))
+                destSize++;
+            }
+        }
+    }
+
+    return destSize;
+}
+
+/*
+ *Name     : handleTashkeelWithTatweel
+ *Function : Replaces Tashkeel as following:
+ *            Case 1 :if the Tashkeel on tatweel, replace it with Tatweel.
+ *            Case 2 :if the Tashkeel aggregated with Shadda on Tatweel, replace
+ *                   it with Shadda on Tatweel.
+ *            Case 3: if the Tashkeel is isolated replace it with Space.
+ *
+ */
+static int32_t
+handleTashkeelWithTatweel(UChar *dest, int32_t sourceLength,
+             int32_t /*destSize*/, uint32_t /*options*/,
+             UErrorCode * /*pErrorCode*/) {
+                 int i;
+                 for(i = 0; i < sourceLength; i++){
+                     if((isTashkeelOnTatweelChar(dest[i]) == 1)){
+                         dest[i] = TATWEEL_CHAR;
+                    }else if((isTashkeelOnTatweelChar(dest[i]) == 2)){
+                         dest[i] = SHADDA_TATWEEL_CHAR;
+                    }else if(isIsolatedTashkeelChar(dest[i]) && dest[i] != SHADDA_CHAR){
+                         dest[i] = SPACE_CHAR;
+                    }
+                 }
+                 return sourceLength;
+}
+
+
+
+/*
+ *Name     : handleGeneratedSpaces
+ *Function : The shapeUnicode function converts Lam + Alef into LamAlef + space,
+ *           and Tashkeel to space.
+ *           handleGeneratedSpaces function puts these generated spaces
+ *           according to the options the user specifies. LamAlef and Tashkeel
+ *           spaces can be replaced at begin, at end, at near or decrease the
+ *           buffer size.
+ *
+ *           There is also Auto option for LamAlef and tashkeel, which will put
+ *           the spaces at end of the buffer (or end of text if the user used
+ *           the option U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END).
+ *
+ *           If the text type was visual_LTR and the option
+ *           U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected the END
+ *           option will place the space at the beginning of the buffer and
+ *           BEGIN will place the space at the end of the buffer.
+ */
+
+static int32_t
+handleGeneratedSpaces(UChar *dest, int32_t sourceLength,
+                    int32_t destSize,
+                    uint32_t options,
+                    UErrorCode *pErrorCode,struct uShapeVariables shapeVars ) {
+
+    int32_t i = 0, j = 0;
+    int32_t count = 0;
+    UChar *tempbuffer=NULL;
+
+    int lamAlefOption = 0;
+    int tashkeelOption = 0;
+    int shapingMode = SHAPE_MODE;
+
+    if (shapingMode == 0){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_RESIZE ){
+            lamAlefOption = 1;
+        }
+        if ( (options&U_SHAPE_TASHKEEL_MASK) == U_SHAPE_TASHKEEL_RESIZE ){
+            tashkeelOption = 1;
+        }
+    }
+
+    tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
+    /* Test for NULL */
+    if(tempbuffer == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+
+
+    if (lamAlefOption || tashkeelOption){
+        uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
+
+        i = j = 0; count = 0;
+        while(i < sourceLength) {
+            if ( (lamAlefOption && dest[i] == LAMALEF_SPACE_SUB) ||
+               (tashkeelOption && dest[i] == TASHKEEL_SPACE_SUB) ){
+                j--;
+                count++;
+            } else {
+                tempbuffer[j] = dest[i];
+            }
+            i++;
+            j++;
+        }
+
+        while(count >= 0) {
+            tempbuffer[i] = 0x0000;
+            i--;
+            count--;
+        }
+
+        u_memcpy(dest, tempbuffer, sourceLength);
+        destSize = u_strlen(dest);
+    }
+
+      lamAlefOption = 0;
+
+    if (shapingMode == 0){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_NEAR ){
+            lamAlefOption = 1;
+        }
+    }
+
+    if (lamAlefOption){
+        /* Lam+Alef is already shaped into LamAlef + FFFF */
+        i = 0;
+        while(i < sourceLength) {
+            if(lamAlefOption&&dest[i] == LAMALEF_SPACE_SUB){
+                dest[i] = SPACE_CHAR;
+            }
+            i++;
+        }
+        destSize = sourceLength;
+    }
+    lamAlefOption = 0;
+    tashkeelOption = 0;
+
+    if (shapingMode == 0) {
+        if ( ((options&U_SHAPE_LAMALEF_MASK) == shapeVars.uShapeLamalefBegin) ||
+              (((options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_AUTO )
+              && (shapeVars.spacesRelativeToTextBeginEnd==1)) ) {
+            lamAlefOption = 1;
+        }
+        if ( (options&U_SHAPE_TASHKEEL_MASK) == shapeVars.uShapeTashkeelBegin ) {
+            tashkeelOption = 1;
+        }
+    }
+
+    if(lamAlefOption || tashkeelOption){
+        uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
+        
+        i = j = sourceLength; count = 0;
+        
+        while(i >= 0) {
+            if ( (lamAlefOption && dest[i] == LAMALEF_SPACE_SUB) ||
+                 (tashkeelOption && dest[i] == TASHKEEL_SPACE_SUB) ){
+                j++;
+                count++;
+            }else {
+                tempbuffer[j] = dest[i];
+            }
+            i--;
+            j--;
+        }
+
+        for(i=0 ;i < count; i++){
+                tempbuffer[i] = SPACE_CHAR;
+        }
+
+        u_memcpy(dest, tempbuffer, sourceLength);
+        destSize = sourceLength;
+    }
+
+
+
+    lamAlefOption = 0;
+    tashkeelOption = 0;
+
+    if (shapingMode == 0) {
+        if ( ((options&U_SHAPE_LAMALEF_MASK) == shapeVars.uShapeLamalefEnd) ||
+              (((options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_AUTO )
+              && (shapeVars.spacesRelativeToTextBeginEnd==0)) ) {
+            lamAlefOption = 1;
+        }
+        if ( (options&U_SHAPE_TASHKEEL_MASK) == shapeVars.uShapeTashkeelEnd ){
+            tashkeelOption = 1;
+        }
+    }
+
+    if(lamAlefOption || tashkeelOption){
+        uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
+
+        i = j = 0; count = 0;
+        while(i < sourceLength) {
+            if ( (lamAlefOption && dest[i] == LAMALEF_SPACE_SUB) ||
+                 (tashkeelOption && dest[i] == TASHKEEL_SPACE_SUB) ){
+                j--;
+                count++;
+            }else {
+                tempbuffer[j] = dest[i];
+            }
+            i++;
+            j++;
+        }
+
+        while(count >= 0) {
+            tempbuffer[i] = SPACE_CHAR;
+            i--;
+            count--;
+        }
+
+        u_memcpy(dest, tempbuffer, sourceLength);
+        destSize = sourceLength;
+    }
+
+
+    if(tempbuffer){
+        uprv_free(tempbuffer);
+    }
+
+    return destSize;
+}
+
+/*
+ *Name     :expandCompositCharAtBegin
+ *Function :Expands the LamAlef character to Lam and Alef consuming the required
+ *         space from beginning of the buffer. If the text type was visual_LTR
+ *         and the option U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END was selected
+ *         the spaces will be located at end of buffer.
+ *         If there are no spaces to expand the LamAlef, an error
+ *         will be set to U_NO_SPACE_AVAILABLE as defined in utypes.h
+ */
+
+static int32_t
+expandCompositCharAtBegin(UChar *dest, int32_t sourceLength, int32_t destSize,UErrorCode *pErrorCode) {
+    int32_t      i = 0,j = 0;
+    int32_t      countl = 0;
+    UChar    *tempbuffer=NULL;
+
+    tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
+
+    /* Test for NULL */
+    if(tempbuffer == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+
+        uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
+
+        i = 0;
+        while(dest[i] == SPACE_CHAR) {
+            countl++;
+            i++;
+        }
+
+        i = j = sourceLength-1;
+
+        while(i >= 0 && j >= 0) {
+            if( countl>0 && isLamAlefChar(dest[i])) {
+                tempbuffer[j] = LAM_CHAR;
+                /* to ensure the array index is within the range */
+                U_ASSERT(dest[i] >= 0xFEF5u
+                    && dest[i]-0xFEF5u < UPRV_LENGTHOF(convertLamAlef));
+                tempbuffer[j-1] = convertLamAlef[ dest[i] - 0xFEF5 ];
+                j--;
+                countl--;
+            }else {
+                 if( countl == 0 && isLamAlefChar(dest[i]) ) {
+                     *pErrorCode=U_NO_SPACE_AVAILABLE;
+                     }
+                 tempbuffer[j] = dest[i];
+            }
+            i--;
+            j--;
+        }
+        u_memcpy(dest, tempbuffer, sourceLength);
+
+        uprv_free(tempbuffer);
+
+        destSize = sourceLength;
+        return destSize;
+}
+
+/*
+ *Name     : expandCompositCharAtEnd
+ *Function : Expands the LamAlef character to Lam and Alef consuming the
+ *           required space from end of the buffer. If the text type was
+ *           Visual LTR and the option U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END
+ *           was used, the spaces will be consumed from begin of buffer. If
+ *           there are no spaces to expand the LamAlef, an error
+ *           will be set to U_NO_SPACE_AVAILABLE as defined in utypes.h
+ */
+
+static int32_t
+expandCompositCharAtEnd(UChar *dest, int32_t sourceLength, int32_t destSize,UErrorCode *pErrorCode) {
+    int32_t      i = 0,j = 0;
+
+    int32_t      countr = 0;
+    int32_t  inpsize = sourceLength;
+
+    UChar    *tempbuffer=NULL;
+    tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
+
+    /* Test for NULL */
+    if(tempbuffer == NULL) {
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+         return 0;
+    }
+
+    uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
+
+    while(dest[inpsize-1] == SPACE_CHAR) {
+        countr++;
+        inpsize--;
+    }
+
+    i = sourceLength - countr - 1;
+    j = sourceLength - 1;
+
+    while(i >= 0 && j >= 0) {
+        if( countr>0 && isLamAlefChar(dest[i]) ) {
+            tempbuffer[j] = LAM_CHAR;
+            tempbuffer[j-1] = convertLamAlef[ dest[i] - 0xFEF5 ];
+            j--;
+            countr--;
+        }else {
+            if ((countr == 0) && isLamAlefChar(dest[i]) ) {
+                *pErrorCode=U_NO_SPACE_AVAILABLE;
+            }
+            tempbuffer[j] = dest[i];
+        }
+        i--;
+        j--;
+    }
+
+    if(countr > 0) {
+        u_memmove(tempbuffer, tempbuffer+countr, sourceLength);
+        if(u_strlen(tempbuffer) < sourceLength) {
+            for(i=sourceLength-1;i>=sourceLength-countr;i--) {
+                tempbuffer[i] = SPACE_CHAR;
+            }
+        }
+    }
+    u_memcpy(dest, tempbuffer, sourceLength);
+
+    uprv_free(tempbuffer);
+
+    destSize = sourceLength;
+    return destSize;
+}
+
+/*
+ *Name     : expandCompositCharAtNear
+ *Function : Expands the LamAlef character into Lam + Alef, YehHamza character
+ *           into Yeh + Hamza, SeenFamily character into SeenFamily character
+ *           + Tail, while consuming the space next to the character.
+ *           If there are no spaces next to the character, an error
+ *           will be set to U_NO_SPACE_AVAILABLE as defined in utypes.h
+ */
+
+static int32_t
+expandCompositCharAtNear(UChar *dest, int32_t sourceLength, int32_t destSize,UErrorCode *pErrorCode,
+                         int yehHamzaOption, int seenTailOption, int lamAlefOption, struct uShapeVariables shapeVars) {
+    int32_t      i = 0;
+
+
+    UChar    lamalefChar, yehhamzaChar;
+
+    for(i = 0 ;i<=sourceLength-1;i++) {
+            if (seenTailOption && isSeenTailFamilyChar(dest[i])) {
+                if ((i>0) && (dest[i-1] == SPACE_CHAR) ) {
+                    dest[i-1] = shapeVars.tailChar;
+                }else {
+                    *pErrorCode=U_NO_SPACE_AVAILABLE;
+                }
+            }else if(yehHamzaOption && (isYehHamzaChar(dest[i])) ) {
+                if ((i>0) && (dest[i-1] == SPACE_CHAR) ) {
+                    yehhamzaChar = dest[i];
+                    dest[i] = yehHamzaToYeh[yehhamzaChar - YEH_HAMZAFE_CHAR];
+                    dest[i-1] = HAMZAFE_CHAR;
+                }else {
+
+                    *pErrorCode=U_NO_SPACE_AVAILABLE;
+                }
+            }else if(lamAlefOption && isLamAlefChar(dest[i+1])) {
+                if(dest[i] == SPACE_CHAR){
+                    lamalefChar = dest[i+1];
+                    dest[i+1] = LAM_CHAR;
+                    dest[i] = convertLamAlef[ lamalefChar - 0xFEF5 ];
+                }else {
+                    *pErrorCode=U_NO_SPACE_AVAILABLE;
+                }
+            }
+       }
+       destSize = sourceLength;
+       return destSize;
+}
+ /*
+ * Name     : expandCompositChar
+ * Function : LamAlef, need special handling, since it expands from one
+ *            character into two characters while shaping or deshaping.
+ *            In order to expand it, near or far spaces according to the
+ *            options user specifies. Also buffer size can be increased.
+ *
+ *            For SeenFamily characters and YehHamza only the near option is
+ *            supported, while for LamAlef we can take spaces from begin, end,
+ *            near or even increase the buffer size.
+ *            There is also the Auto option for LamAlef only, which will first
+ *            search for a space at end, begin then near, respectively.
+ *            If there are no spaces to expand these characters, an error will be set to
+ *            U_NO_SPACE_AVAILABLE as defined in utypes.h
+ */
+
+static int32_t
+expandCompositChar(UChar *dest, int32_t sourceLength,
+              int32_t destSize,uint32_t options,
+              UErrorCode *pErrorCode, int shapingMode,struct uShapeVariables shapeVars) {
+
+    int32_t      i = 0,j = 0;
+
+    UChar    *tempbuffer=NULL;
+    int yehHamzaOption = 0;
+    int seenTailOption = 0;
+    int lamAlefOption = 0;
+
+    if (shapingMode == 1){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_AUTO){
+
+            if(shapeVars.spacesRelativeToTextBeginEnd == 0) {
+                destSize = expandCompositCharAtEnd(dest, sourceLength, destSize, pErrorCode);
+
+                if(*pErrorCode == U_NO_SPACE_AVAILABLE) {
+                    *pErrorCode = U_ZERO_ERROR;
+                    destSize = expandCompositCharAtBegin(dest, sourceLength, destSize, pErrorCode);
+                }
+            }else {
+                destSize = expandCompositCharAtBegin(dest, sourceLength, destSize, pErrorCode);
+
+                if(*pErrorCode == U_NO_SPACE_AVAILABLE) {
+                    *pErrorCode = U_ZERO_ERROR;
+                    destSize = expandCompositCharAtEnd(dest, sourceLength, destSize, pErrorCode);
+                }
+            }
+
+            if(*pErrorCode == U_NO_SPACE_AVAILABLE) {
+                *pErrorCode = U_ZERO_ERROR;
+                destSize = expandCompositCharAtNear(dest, sourceLength, destSize, pErrorCode, yehHamzaOption,
+                                                seenTailOption, 1,shapeVars);
+            }
+        }
+    }
+
+    if (shapingMode == 1){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == shapeVars.uShapeLamalefEnd){
+            destSize = expandCompositCharAtEnd(dest, sourceLength, destSize, pErrorCode);
+        }
+    }
+
+    if (shapingMode == 1){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == shapeVars.uShapeLamalefBegin){
+            destSize = expandCompositCharAtBegin(dest, sourceLength, destSize, pErrorCode);
+        }
+    }
+
+    if (shapingMode == 0){
+         if ((options&U_SHAPE_YEHHAMZA_MASK) == U_SHAPE_YEHHAMZA_TWOCELL_NEAR){
+             yehHamzaOption = 1;
+         }
+         if ((options&U_SHAPE_SEEN_MASK) == U_SHAPE_SEEN_TWOCELL_NEAR){
+            seenTailOption = 1;
+         }
+    }
+    if (shapingMode == 1) {
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_NEAR) {
+            lamAlefOption = 1;
+        }
+    }
+
+
+    if (yehHamzaOption || seenTailOption || lamAlefOption){
+        destSize = expandCompositCharAtNear(dest, sourceLength, destSize, pErrorCode, yehHamzaOption,
+                                            seenTailOption,lamAlefOption,shapeVars);
+    }
+
+
+    if (shapingMode == 1){
+        if ( (options&U_SHAPE_LAMALEF_MASK) == U_SHAPE_LAMALEF_RESIZE){
+            destSize = calculateSize(dest,sourceLength,destSize,options);
+            tempbuffer = (UChar *)uprv_malloc((destSize+1)*U_SIZEOF_UCHAR);
+
+            /* Test for NULL */
+            if(tempbuffer == NULL) {
+                *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                return 0;
+            }
+
+            uprv_memset(tempbuffer, 0, (destSize+1)*U_SIZEOF_UCHAR);
+
+            i = j = 0;
+            while(i < destSize && j < destSize) {
+                if(isLamAlefChar(dest[i]) ) {
+                    tempbuffer[j] = convertLamAlef[ dest[i] - 0xFEF5 ];
+                    tempbuffer[j+1] = LAM_CHAR;
+                    j++;
+                }else {
+                    tempbuffer[j] = dest[i];
+                }
+                i++;
+                j++;
+            }
+
+            u_memcpy(dest, tempbuffer, destSize);
+        }
+    }
+
+    if(tempbuffer) {
+        uprv_free(tempbuffer);
+    }
+    return destSize;
+}
+
+/*
+ *Name     : shapeUnicode
+ *Function : Converts an Arabic Unicode buffer in 06xx Range into a shaped
+ *           arabic Unicode buffer in FExx Range
+ */
+static int32_t
+shapeUnicode(UChar *dest, int32_t sourceLength,
+             int32_t destSize,uint32_t options,
+             UErrorCode *pErrorCode,
+             int tashkeelFlag, struct uShapeVariables shapeVars) {
+
+    int32_t          i, iend;
+    int32_t          step;
+    int32_t          lastPos,Nx, Nw;
+    unsigned int     Shape;
+    int32_t          lamalef_found = 0;
+    int32_t seenfamFound = 0, yehhamzaFound =0, tashkeelFound  = 0;
+    UChar            prevLink = 0, lastLink = 0, currLink, nextLink = 0;
+    UChar            wLamalef;
+
+    /*
+     * Converts the input buffer from FExx Range into 06xx Range
+     * to make sure that all characters are in the 06xx range
+     * even the lamalef is converted to the special region in
+     * the 06xx range
+     */
+    if ((options & U_SHAPE_PRESERVE_PRESENTATION_MASK)  == U_SHAPE_PRESERVE_PRESENTATION_NOOP) {
+        for (i = 0; i < sourceLength; i++) {
+            UChar inputChar  = dest[i];
+            if ( (inputChar >= 0xFB50) && (inputChar <= 0xFBFF)) {
+                UChar c = convertFBto06 [ (inputChar - 0xFB50) ];
+                if (c != 0)
+                    dest[i] = c;
+            } else if ( (inputChar >= 0xFE70) && (inputChar <= 0xFEFC)) {
+                dest[i] = convertFEto06 [ (inputChar - 0xFE70) ] ;
+            } else {
+                dest[i] = inputChar ;
+            }
+        }
+    }
+
+
+    /* sets the index to the end of the buffer, together with the step point to -1 */
+    i = sourceLength - 1;
+    iend = -1;
+    step = -1;
+
+    /*
+     * This function resolves the link between the characters .
+     * Arabic characters have four forms :
+     * Isolated Form, Initial Form, Middle Form and Final Form
+     */
+    currLink = getLink(dest[i]);
+
+    lastPos = i;
+    Nx = -2, Nw = 0;
+
+    while (i != iend) {
+        /* If high byte of currLink > 0 then more than one shape */
+        if ((currLink & 0xFF00) > 0 || (getLink(dest[i]) & IRRELEVANT) != 0) {
+            Nw = i + step;
+            while (Nx < 0) {         /* we need to know about next char */
+                if(Nw == iend) {
+                    nextLink = 0;
+                    Nx = 3000;
+                } else {
+                    nextLink = getLink(dest[Nw]);
+                    if((nextLink & IRRELEVANT) == 0) {
+                        Nx = Nw;
+                    } else {
+                        Nw = Nw + step;
+                    }
+                }
+            }
+
+            if ( ((currLink & ALEFTYPE) > 0)  &&  ((lastLink & LAMTYPE) > 0) ) {
+                lamalef_found = 1;
+                wLamalef = changeLamAlef(dest[i]); /*get from 0x065C-0x065f */
+                if ( wLamalef != 0) {
+                    dest[i] = LAMALEF_SPACE_SUB;            /* The default case is to drop the Alef and replace */
+                    dest[lastPos] =wLamalef;     /* it by LAMALEF_SPACE_SUB which is the last character in the  */
+                    i=lastPos;                   /* unicode private use area, this is done to make   */
+                }                                /* sure that removeLamAlefSpaces() handles only the */
+                lastLink = prevLink;             /* spaces generated during lamalef generation.      */
+                currLink = getLink(wLamalef);    /* LAMALEF_SPACE_SUB is added here and is replaced by spaces   */
+            }                                    /* in removeLamAlefSpaces()                         */
+
+            if ((i > 0) && (dest[i-1] == SPACE_CHAR)){
+                if ( isSeenFamilyChar(dest[i])) {
+                    seenfamFound = 1;
+                } else if (dest[i] == YEH_HAMZA_CHAR) {
+                    yehhamzaFound = 1;
+                }
+            }
+            else if(i==0){
+                if ( isSeenFamilyChar(dest[i])){
+                    seenfamFound = 1;
+                } else if (dest[i] == YEH_HAMZA_CHAR) {
+                    yehhamzaFound = 1;
+                }
+            }
+
+            /*
+             * get the proper shape according to link ability of neighbors
+             * and of character; depends on the order of the shapes
+             * (isolated, initial, middle, final) in the compatibility area
+             */
+            Shape = shapeTable[nextLink & (LINKR + LINKL)]
+                              [lastLink & (LINKR + LINKL)]
+                              [currLink & (LINKR + LINKL)];
+
+            if ((currLink & (LINKR+LINKL)) == 1) {
+                Shape &= 1;
+            } else if(isTashkeelChar(dest[i])) {
+                if( (lastLink & LINKL) && (nextLink & LINKR) && (tashkeelFlag == 1) &&
+                     dest[i] != 0x064C && dest[i] != 0x064D )
+                {
+                    Shape = 1;
+                    if( (nextLink&ALEFTYPE) == ALEFTYPE && (lastLink&LAMTYPE) == LAMTYPE ) {
+                        Shape = 0;
+                    }
+                } else if(tashkeelFlag == 2 && dest[i] == SHADDA06_CHAR){
+                    Shape = 1;
+                } else {
+                    Shape = 0;
+                }
+            }
+            if ((dest[i] ^ 0x0600) < 0x100) {
+                if ( isTashkeelChar(dest[i]) ){
+                    if (tashkeelFlag == 2  && dest[i] != SHADDA06_CHAR){
+                        dest[i] = TASHKEEL_SPACE_SUB;
+                        tashkeelFound  = 1;
+                    } else {
+                        /* to ensure the array index is within the range */
+                        U_ASSERT(dest[i] >= 0x064Bu
+                            && dest[i]-0x064Bu < UPRV_LENGTHOF(IrrelevantPos));
+                        dest[i] =  0xFE70 + IrrelevantPos[(dest[i] - 0x064B)] + static_cast<UChar>(Shape);
+                    }
+                }else if ((currLink & APRESENT) > 0) {
+                    dest[i] = (UChar)(0xFB50 + (currLink >> 8) + Shape);
+                }else if ((currLink >> 8) > 0 && (currLink & IRRELEVANT) == 0) {
+                    dest[i] = (UChar)(0xFE70 + (currLink >> 8) + Shape);
+                }
+            }
+        }
+
+        /* move one notch forward */
+        if ((currLink & IRRELEVANT) == 0) {
+            prevLink = lastLink;
+            lastLink = currLink;
+            lastPos = i;
+        }
+
+        i = i + step;
+        if (i == Nx) {
+            currLink = nextLink;
+            Nx = -2;
+        } else if(i != iend) {
+            currLink = getLink(dest[i]);
+        }
+    }
+    destSize = sourceLength;
+    if ( (lamalef_found != 0 ) || (tashkeelFound  != 0) ){
+        destSize = handleGeneratedSpaces(dest,sourceLength,destSize,options,pErrorCode, shapeVars);
+    }
+
+    if ( (seenfamFound != 0) || (yehhamzaFound != 0) ) {
+        destSize = expandCompositChar(dest, sourceLength,destSize,options,pErrorCode, SHAPE_MODE,shapeVars);
+    }
+    return destSize;
+}
+
+/*
+ *Name     : deShapeUnicode
+ *Function : Converts an Arabic Unicode buffer in FExx Range into unshaped
+ *           arabic Unicode buffer in 06xx Range
+ */
+static int32_t
+deShapeUnicode(UChar *dest, int32_t sourceLength,
+               int32_t destSize,uint32_t options,
+               UErrorCode *pErrorCode, struct uShapeVariables shapeVars) {
+    int32_t i = 0;
+    int32_t lamalef_found = 0;
+    int32_t yehHamzaComposeEnabled = 0;
+    int32_t seenComposeEnabled = 0;
+
+    yehHamzaComposeEnabled = ((options&U_SHAPE_YEHHAMZA_MASK) == U_SHAPE_YEHHAMZA_TWOCELL_NEAR) ? 1 : 0;
+    seenComposeEnabled = ((options&U_SHAPE_SEEN_MASK) == U_SHAPE_SEEN_TWOCELL_NEAR)? 1 : 0;
+
+    /*
+     *This for loop changes the buffer from the Unicode FE range to
+     *the Unicode 06 range
+     */
+
+    for(i = 0; i < sourceLength; i++) {
+        UChar  inputChar = dest[i];
+        if ( (inputChar >= 0xFB50) && (inputChar <= 0xFBFF)) { /* FBxx Arabic range */
+            UChar c = convertFBto06 [ (inputChar - 0xFB50) ];
+            if (c != 0)
+                dest[i] = c;
+        } else if( (yehHamzaComposeEnabled == 1) && ((inputChar == HAMZA06_CHAR) || (inputChar == HAMZAFE_CHAR))
+                && (i < (sourceLength - 1)) && isAlefMaksouraChar(dest[i+1] )) {
+                 dest[i] = SPACE_CHAR;
+                 dest[i+1] = YEH_HAMZA_CHAR;
+        } else if ( (seenComposeEnabled == 1) && (isTailChar(inputChar)) && (i< (sourceLength - 1))
+                        && (isSeenTailFamilyChar(dest[i+1])) ) {
+                dest[i] = SPACE_CHAR;
+        } else if (( inputChar >= 0xFE70) && (inputChar <= 0xFEF4 )) { /* FExx Arabic range */
+                dest[i] = convertFEto06 [ (inputChar - 0xFE70) ];
+        } else {
+            dest[i] = inputChar ;
+        }
+
+        if( isLamAlefChar(dest[i]) )
+            lamalef_found = 1;
+    }
+
+   destSize = sourceLength;
+   if (lamalef_found != 0){
+          destSize = expandCompositChar(dest,sourceLength,destSize,options,pErrorCode,DESHAPE_MODE, shapeVars);
+   }
+   return destSize;
+}
+
+/*
+ ****************************************
+ * u_shapeArabic
+ ****************************************
+ */
+
+U_CAPI int32_t U_EXPORT2
+u_shapeArabic(const UChar *source, int32_t sourceLength,
+              UChar *dest, int32_t destCapacity,
+              uint32_t options,
+              UErrorCode *pErrorCode) {
+
+    int32_t destLength;
+    struct  uShapeVariables shapeVars = { OLD_TAIL_CHAR,U_SHAPE_LAMALEF_BEGIN,U_SHAPE_LAMALEF_END,U_SHAPE_TASHKEEL_BEGIN,U_SHAPE_TASHKEEL_END,0};
+
+    /* usual error checking */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* make sure that no reserved options values are used; allow dest==NULL only for preflighting */
+    if( source==NULL || sourceLength<-1 || (dest==NULL && destCapacity!=0) || destCapacity<0 ||
+                (((options&U_SHAPE_TASHKEEL_MASK) > 0) &&
+                 ((options&U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED) == U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED) ) ||
+                (((options&U_SHAPE_TASHKEEL_MASK) > 0) &&
+                 ((options&U_SHAPE_LETTERS_MASK) == U_SHAPE_LETTERS_UNSHAPE)) ||
+                (options&U_SHAPE_DIGIT_TYPE_RESERVED)==U_SHAPE_DIGIT_TYPE_RESERVED ||
+                (options&U_SHAPE_DIGITS_MASK)==U_SHAPE_DIGITS_RESERVED ||
+                ((options&U_SHAPE_LAMALEF_MASK) != U_SHAPE_LAMALEF_RESIZE  &&
+                (options&U_SHAPE_AGGREGATE_TASHKEEL_MASK) != 0) ||
+                ((options&U_SHAPE_AGGREGATE_TASHKEEL_MASK) == U_SHAPE_AGGREGATE_TASHKEEL &&
+                (options&U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED) != U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED)
+    )
+    {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    /* Validate  lamalef options */
+    if(((options&U_SHAPE_LAMALEF_MASK) > 0)&&
+              !(((options & U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_BEGIN) ||
+                ((options & U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_END ) ||
+                ((options & U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_RESIZE )||
+                 ((options & U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_AUTO) ||
+                 ((options & U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_NEAR)))
+    {
+         *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    /* Validate  Tashkeel options */
+    if(((options&U_SHAPE_TASHKEEL_MASK) > 0)&&
+                   !(((options & U_SHAPE_TASHKEEL_MASK)==U_SHAPE_TASHKEEL_BEGIN) ||
+                     ((options & U_SHAPE_TASHKEEL_MASK)==U_SHAPE_TASHKEEL_END )
+                    ||((options & U_SHAPE_TASHKEEL_MASK)==U_SHAPE_TASHKEEL_RESIZE )||
+                    ((options & U_SHAPE_TASHKEEL_MASK)==U_SHAPE_TASHKEEL_REPLACE_BY_TATWEEL)))
+    {
+         *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    /* determine the source length */
+    if(sourceLength==-1) {
+        sourceLength=u_strlen(source);
+    }
+    if(sourceLength<=0) {
+        return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
+    }
+
+    /* check that source and destination do not overlap */
+    if( dest!=NULL &&
+        ((source<=dest && dest<source+sourceLength) ||
+         (dest<=source && source<dest+destCapacity))) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* Does Options contain the new Seen Tail Unicode code point option */
+    if ( (options&U_SHAPE_TAIL_TYPE_MASK) == U_SHAPE_TAIL_NEW_UNICODE){
+        shapeVars.tailChar = NEW_TAIL_CHAR;
+    }else {
+        shapeVars.tailChar = OLD_TAIL_CHAR;
+    }
+
+    if((options&U_SHAPE_LETTERS_MASK)!=U_SHAPE_LETTERS_NOOP) {
+        UChar buffer[300];
+        UChar *tempbuffer, *tempsource = NULL;
+        int32_t outputSize, spacesCountl=0, spacesCountr=0;
+
+        if((options&U_SHAPE_AGGREGATE_TASHKEEL_MASK)>0) {
+            int32_t logical_order = (options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_LOGICAL;
+            int32_t aggregate_tashkeel =
+                        (options&(U_SHAPE_AGGREGATE_TASHKEEL_MASK+U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED)) ==
+                        (U_SHAPE_AGGREGATE_TASHKEEL+U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED);
+            int step=logical_order?1:-1;
+            int j=logical_order?-1:2*sourceLength;
+            int i=logical_order?-1:sourceLength;
+            int end=logical_order?sourceLength:-1;
+            int aggregation_possible = 1;
+            UChar prev = 0;
+            UChar prevLink, currLink = 0;
+            int newSourceLength = 0;
+            tempsource = (UChar *)uprv_malloc(2*sourceLength*U_SIZEOF_UCHAR);
+            if(tempsource == NULL) {
+                *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                return 0;
+            }
+
+            while ((i+=step) != end) {
+                prevLink = currLink;
+                currLink = getLink(source[i]);
+                if (aggregate_tashkeel && ((prevLink|currLink)&COMBINE) == COMBINE && aggregation_possible) {
+                    aggregation_possible = 0;
+                    tempsource[j] = (prev<source[i]?prev:source[i])-0x064C+0xFC5E;
+                    currLink = getLink(tempsource[j]);
+                } else {
+                    aggregation_possible = 1;
+                    tempsource[j+=step] = source[i];
+                    prev = source[i];
+                    newSourceLength++;
+                }
+            }
+            source = tempsource+(logical_order?0:j);
+            sourceLength = newSourceLength;
+        }
+
+        /* calculate destination size */
+        /* TODO: do we ever need to do this pure preflighting? */
+        if(((options&U_SHAPE_LAMALEF_MASK)==U_SHAPE_LAMALEF_RESIZE) ||
+           ((options&U_SHAPE_TASHKEEL_MASK)==U_SHAPE_TASHKEEL_RESIZE)) {
+            outputSize=calculateSize(source,sourceLength,destCapacity,options);
+        } else {
+            outputSize=sourceLength;
+        }
+
+        if(outputSize>destCapacity) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+                if (tempsource != NULL) uprv_free(tempsource);
+            return outputSize;
+        }
+
+        /*
+         * need a temporary buffer of size max(outputSize, sourceLength)
+         * because at first we copy source->temp
+         */
+        if(sourceLength>outputSize) {
+            outputSize=sourceLength;
+        }
+
+        /* Start of Arabic letter shaping part */
+        if(outputSize<=UPRV_LENGTHOF(buffer)) {
+            outputSize=UPRV_LENGTHOF(buffer);
+            tempbuffer=buffer;
+        } else {
+            tempbuffer = (UChar *)uprv_malloc(outputSize*U_SIZEOF_UCHAR);
+
+            /*Test for NULL*/
+            if(tempbuffer == NULL) {
+                *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                if (tempsource != NULL) uprv_free(tempsource);
+                return 0;
+            }
+        }
+        u_memcpy(tempbuffer, source, sourceLength);
+        if (tempsource != NULL){
+            uprv_free(tempsource);
+        }
+
+        if(sourceLength<outputSize) {
+            uprv_memset(tempbuffer+sourceLength, 0, (outputSize-sourceLength)*U_SIZEOF_UCHAR);
+        }
+
+        if((options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_LOGICAL) {
+            countSpaces(tempbuffer,sourceLength,options,&spacesCountl,&spacesCountr);
+            invertBuffer(tempbuffer,sourceLength,options,spacesCountl,spacesCountr);
+        }
+
+        if((options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_VISUAL_LTR) {
+            if((options&U_SHAPE_SPACES_RELATIVE_TO_TEXT_MASK) == U_SHAPE_SPACES_RELATIVE_TO_TEXT_BEGIN_END) {
+                shapeVars.spacesRelativeToTextBeginEnd = 1;
+                shapeVars.uShapeLamalefBegin = U_SHAPE_LAMALEF_END;
+                shapeVars.uShapeLamalefEnd = U_SHAPE_LAMALEF_BEGIN;
+                shapeVars.uShapeTashkeelBegin = U_SHAPE_TASHKEEL_END;
+                shapeVars.uShapeTashkeelEnd = U_SHAPE_TASHKEEL_BEGIN;
+            }
+        }
+
+        switch(options&U_SHAPE_LETTERS_MASK) {
+        case U_SHAPE_LETTERS_SHAPE :
+             if( (options&U_SHAPE_TASHKEEL_MASK)> 0
+                 && ((options&U_SHAPE_TASHKEEL_MASK) !=U_SHAPE_TASHKEEL_REPLACE_BY_TATWEEL)) {
+                /* Call the shaping function with tashkeel flag == 2 for removal of tashkeel */
+                destLength = shapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,2,shapeVars);
+             }else {
+                /* default Call the shaping function with tashkeel flag == 1 */
+                destLength = shapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,1,shapeVars);
+
+                /*After shaping text check if user wants to remove tashkeel and replace it with tatweel*/
+                if( (options&U_SHAPE_TASHKEEL_MASK) == U_SHAPE_TASHKEEL_REPLACE_BY_TATWEEL){
+                  destLength = handleTashkeelWithTatweel(tempbuffer,destLength,destCapacity,options,pErrorCode);
+                }
+            }
+            break;
+        case U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED :
+            /* Call the shaping function with tashkeel flag == 0 */
+            destLength = shapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,0,shapeVars);
+            break;
+
+        case U_SHAPE_LETTERS_UNSHAPE :
+            /* Call the deshaping function */
+            destLength = deShapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,shapeVars);
+            break;
+        default :
+            /* will never occur because of validity checks above */
+            destLength = 0;
+            break;
+        }
+
+        /*
+         * TODO: (markus 2002aug01)
+         * For as long as we always preflight the outputSize above
+         * we should U_ASSERT(outputSize==destLength)
+         * except for the adjustment above before the tempbuffer allocation
+         */
+
+        if((options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_LOGICAL) {
+            countSpaces(tempbuffer,destLength,options,&spacesCountl,&spacesCountr);
+            invertBuffer(tempbuffer,destLength,options,spacesCountl,spacesCountr);
+        }
+        u_memcpy(dest, tempbuffer, uprv_min(destLength, destCapacity));
+
+        if(tempbuffer!=buffer) {
+            uprv_free(tempbuffer);
+        }
+
+        if(destLength>destCapacity) {
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return destLength;
+        }
+
+        /* End of Arabic letter shaping part */
+    } else {
+        /*
+         * No letter shaping:
+         * just make sure the destination is large enough and copy the string.
+         */
+        if(destCapacity<sourceLength) {
+            /* this catches preflighting, too */
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+            return sourceLength;
+        }
+        u_memcpy(dest, source, sourceLength);
+        destLength=sourceLength;
+    }
+
+    /*
+     * Perform number shaping.
+     * With UTF-16 or UTF-32, the length of the string is constant.
+     * The easiest way to do this is to operate on the destination and
+     * "shape" the digits in-place.
+     */
+    if((options&U_SHAPE_DIGITS_MASK)!=U_SHAPE_DIGITS_NOOP) {
+        UChar digitBase;
+        int32_t i;
+
+        /* select the requested digit group */
+        switch(options&U_SHAPE_DIGIT_TYPE_MASK) {
+        case U_SHAPE_DIGIT_TYPE_AN:
+            digitBase=0x660; /* Unicode: "Arabic-Indic digits" */
+            break;
+        case U_SHAPE_DIGIT_TYPE_AN_EXTENDED:
+            digitBase=0x6f0; /* Unicode: "Eastern Arabic-Indic digits (Persian and Urdu)" */
+            break;
+        default:
+            /* will never occur because of validity checks above */
+            digitBase=0;
+            break;
+        }
+
+        /* perform the requested operation */
+        switch(options&U_SHAPE_DIGITS_MASK) {
+        case U_SHAPE_DIGITS_EN2AN:
+            /* add (digitBase-'0') to each European (ASCII) digit code point */
+            digitBase-=0x30;
+            for(i=0; i<destLength; ++i) {
+                if(((uint32_t)dest[i]-0x30)<10) {
+                    dest[i]+=digitBase;
+                }
+            }
+            break;
+        case U_SHAPE_DIGITS_AN2EN:
+            /* subtract (digitBase-'0') from each Arabic digit code point */
+            for(i=0; i<destLength; ++i) {
+                if(((uint32_t)dest[i]-(uint32_t)digitBase)<10) {
+                    dest[i]-=digitBase-0x30;
+                }
+            }
+            break;
+        case U_SHAPE_DIGITS_ALEN2AN_INIT_LR:
+            _shapeToArabicDigitsWithContext(dest, destLength,
+                                            digitBase,
+                                            (UBool)((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL),
+                                            FALSE);
+            break;
+        case U_SHAPE_DIGITS_ALEN2AN_INIT_AL:
+            _shapeToArabicDigitsWithContext(dest, destLength,
+                                            digitBase,
+                                            (UBool)((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL),
+                                            TRUE);
+            break;
+        default:
+            /* will never occur because of validity checks above */
+            break;
+        }
+    }
+
+    return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
+}
diff --git a/thirdparty/icu4c/common/usprep.cpp b/thirdparty/icu4c/common/usprep.cpp
new file mode 100644
index 0000000000..8351a77370
--- /dev/null
+++ b/thirdparty/icu4c/common/usprep.cpp
@@ -0,0 +1,871 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ *******************************************************************************
+ *
+ *   Copyright (C) 2003-2016, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ *
+ *******************************************************************************
+ *   file name:  usprep.cpp
+ *   encoding:   UTF-8
+ *   tab size:   8 (not used)
+ *   indentation:4
+ *
+ *   created on: 2003jul2
+ *   created by: Ram Viswanadha
+ */
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/usprep.h"
+
+#include "unicode/normalizer2.h"
+#include "unicode/ustring.h"
+#include "unicode/uchar.h"
+#include "unicode/uversion.h"
+#include "umutex.h"
+#include "cmemory.h"
+#include "sprpimpl.h"
+#include "ustr_imp.h"
+#include "uhash.h"
+#include "cstring.h"
+#include "udataswp.h"
+#include "ucln_cmn.h"
+#include "ubidi_props.h"
+#include "uprops.h"
+
+U_NAMESPACE_USE
+
+U_CDECL_BEGIN
+
+/*
+Static cache for already opened StringPrep profiles
+*/
+static UHashtable *SHARED_DATA_HASHTABLE = NULL;
+static icu::UInitOnce gSharedDataInitOnce = U_INITONCE_INITIALIZER;
+
+static UMutex usprepMutex;
+/* format version of spp file */
+//static uint8_t formatVersion[4]={ 0, 0, 0, 0 };
+
+/* the Unicode version of the sprep data */
+static UVersionInfo dataVersion={ 0, 0, 0, 0 };
+
+/* Profile names must be aligned to UStringPrepProfileType */
+static const char * const PROFILE_NAMES[] = {
+    "rfc3491",      /* USPREP_RFC3491_NAMEPREP */
+    "rfc3530cs",    /* USPREP_RFC3530_NFS4_CS_PREP */
+    "rfc3530csci",  /* USPREP_RFC3530_NFS4_CS_PREP_CI */
+    "rfc3491",      /* USPREP_RFC3530_NSF4_CIS_PREP */
+    "rfc3530mixp",  /* USPREP_RFC3530_NSF4_MIXED_PREP_PREFIX */
+    "rfc3491",      /* USPREP_RFC3530_NSF4_MIXED_PREP_SUFFIX */
+    "rfc3722",      /* USPREP_RFC3722_ISCSI */
+    "rfc3920node",  /* USPREP_RFC3920_NODEPREP */
+    "rfc3920res",   /* USPREP_RFC3920_RESOURCEPREP */
+    "rfc4011",      /* USPREP_RFC4011_MIB */
+    "rfc4013",      /* USPREP_RFC4013_SASLPREP */
+    "rfc4505",      /* USPREP_RFC4505_TRACE */
+    "rfc4518",      /* USPREP_RFC4518_LDAP */
+    "rfc4518ci",    /* USPREP_RFC4518_LDAP_CI */
+};
+
+static UBool U_CALLCONV
+isSPrepAcceptable(void * /* context */,
+             const char * /* type */, 
+             const char * /* name */,
+             const UDataInfo *pInfo) {
+    if(
+        pInfo->size>=20 &&
+        pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
+        pInfo->charsetFamily==U_CHARSET_FAMILY &&
+        pInfo->dataFormat[0]==0x53 &&   /* dataFormat="SPRP" */
+        pInfo->dataFormat[1]==0x50 &&
+        pInfo->dataFormat[2]==0x52 &&
+        pInfo->dataFormat[3]==0x50 &&
+        pInfo->formatVersion[0]==3 &&
+        pInfo->formatVersion[2]==UTRIE_SHIFT &&
+        pInfo->formatVersion[3]==UTRIE_INDEX_SHIFT
+    ) {
+        //uprv_memcpy(formatVersion, pInfo->formatVersion, 4);
+        uprv_memcpy(dataVersion, pInfo->dataVersion, 4);
+        return TRUE;
+    } else {
+        return FALSE;
+    }
+}
+
+static int32_t U_CALLCONV
+getSPrepFoldingOffset(uint32_t data) {
+
+    return (int32_t)data;
+
+}
+
+/* hashes an entry  */
+static int32_t U_CALLCONV 
+hashEntry(const UHashTok parm) {
+    UStringPrepKey *b = (UStringPrepKey *)parm.pointer;
+    UHashTok namekey, pathkey;
+    namekey.pointer = b->name;
+    pathkey.pointer = b->path;
+    uint32_t unsignedHash = static_cast<uint32_t>(uhash_hashChars(namekey)) +
+            37u * static_cast<uint32_t>(uhash_hashChars(pathkey));
+    return static_cast<int32_t>(unsignedHash);
+}
+
+/* compares two entries */
+static UBool U_CALLCONV 
+compareEntries(const UHashTok p1, const UHashTok p2) {
+    UStringPrepKey *b1 = (UStringPrepKey *)p1.pointer;
+    UStringPrepKey *b2 = (UStringPrepKey *)p2.pointer;
+    UHashTok name1, name2, path1, path2;
+    name1.pointer = b1->name;
+    name2.pointer = b2->name;
+    path1.pointer = b1->path;
+    path2.pointer = b2->path;
+    return ((UBool)(uhash_compareChars(name1, name2) & 
+        uhash_compareChars(path1, path2)));
+}
+
+static void 
+usprep_unload(UStringPrepProfile* data){
+    udata_close(data->sprepData);
+}
+
+static int32_t 
+usprep_internal_flushCache(UBool noRefCount){
+    UStringPrepProfile *profile = NULL;
+    UStringPrepKey  *key  = NULL;
+    int32_t pos = UHASH_FIRST;
+    int32_t deletedNum = 0;
+    const UHashElement *e;
+
+    /*
+     * if shared data hasn't even been lazy evaluated yet
+     * return 0
+     */
+    umtx_lock(&usprepMutex);
+    if (SHARED_DATA_HASHTABLE == NULL) {
+        umtx_unlock(&usprepMutex);
+        return 0;
+    }
+
+    /*creates an enumeration to iterate through every element in the table */
+    while ((e = uhash_nextElement(SHARED_DATA_HASHTABLE, &pos)) != NULL)
+    {
+        profile = (UStringPrepProfile *) e->value.pointer;
+        key  = (UStringPrepKey *) e->key.pointer;
+
+        if ((noRefCount== FALSE && profile->refCount == 0) || 
+             noRefCount== TRUE) {
+            deletedNum++;
+            uhash_removeElement(SHARED_DATA_HASHTABLE, e);
+
+            /* unload the data */
+            usprep_unload(profile);
+
+            if(key->name != NULL) {
+                uprv_free(key->name);
+                key->name=NULL;
+            }
+            if(key->path != NULL) {
+                uprv_free(key->path);
+                key->path=NULL;
+            }
+            uprv_free(profile);
+            uprv_free(key);
+        }
+       
+    }
+    umtx_unlock(&usprepMutex);
+
+    return deletedNum;
+}
+
+/* Works just like ucnv_flushCache() 
+static int32_t 
+usprep_flushCache(){
+    return usprep_internal_flushCache(FALSE);
+}
+*/
+
+static UBool U_CALLCONV usprep_cleanup(void){
+    if (SHARED_DATA_HASHTABLE != NULL) {
+        usprep_internal_flushCache(TRUE);
+        if (SHARED_DATA_HASHTABLE != NULL && uhash_count(SHARED_DATA_HASHTABLE) == 0) {
+            uhash_close(SHARED_DATA_HASHTABLE);
+            SHARED_DATA_HASHTABLE = NULL;
+        }
+    }
+    gSharedDataInitOnce.reset();
+    return (SHARED_DATA_HASHTABLE == NULL);
+}
+U_CDECL_END
+
+
+/** Initializes the cache for resources */
+static void U_CALLCONV
+createCache(UErrorCode &status) {
+    SHARED_DATA_HASHTABLE = uhash_open(hashEntry, compareEntries, NULL, &status);
+    if (U_FAILURE(status)) {
+        SHARED_DATA_HASHTABLE = NULL;
+    }
+    ucln_common_registerCleanup(UCLN_COMMON_USPREP, usprep_cleanup);
+}
+
+static void 
+initCache(UErrorCode *status) {
+    umtx_initOnce(gSharedDataInitOnce, &createCache, *status);
+}
+
+static UBool U_CALLCONV
+loadData(UStringPrepProfile* profile, 
+         const char* path, 
+         const char* name, 
+         const char* type, 
+         UErrorCode* errorCode) {
+    /* load Unicode SPREP data from file */    
+    UTrie _sprepTrie={ 0,0,0,0,0,0,0 };
+    UDataMemory *dataMemory;
+    const int32_t *p=NULL;
+    const uint8_t *pb;
+    UVersionInfo normUnicodeVersion;
+    int32_t normUniVer, sprepUniVer, normCorrVer;
+
+    if(errorCode==NULL || U_FAILURE(*errorCode)) {
+        return 0;
+    }
+
+    /* open the data outside the mutex block */
+    //TODO: change the path
+    dataMemory=udata_openChoice(path, type, name, isSPrepAcceptable, NULL, errorCode);
+    if(U_FAILURE(*errorCode)) {
+        return FALSE;
+    }
+
+    p=(const int32_t *)udata_getMemory(dataMemory);
+    pb=(const uint8_t *)(p+_SPREP_INDEX_TOP);
+    utrie_unserialize(&_sprepTrie, pb, p[_SPREP_INDEX_TRIE_SIZE], errorCode);
+    _sprepTrie.getFoldingOffset=getSPrepFoldingOffset;
+
+
+    if(U_FAILURE(*errorCode)) {
+        udata_close(dataMemory);
+        return FALSE;
+    }
+
+    /* in the mutex block, set the data for this process */
+    umtx_lock(&usprepMutex);
+    if(profile->sprepData==NULL) {
+        profile->sprepData=dataMemory;
+        dataMemory=NULL;
+        uprv_memcpy(&profile->indexes, p, sizeof(profile->indexes));
+        uprv_memcpy(&profile->sprepTrie, &_sprepTrie, sizeof(UTrie));
+    } else {
+        p=(const int32_t *)udata_getMemory(profile->sprepData);
+    }
+    umtx_unlock(&usprepMutex);
+    /* initialize some variables */
+    profile->mappingData=(uint16_t *)((uint8_t *)(p+_SPREP_INDEX_TOP)+profile->indexes[_SPREP_INDEX_TRIE_SIZE]);
+    
+    u_getUnicodeVersion(normUnicodeVersion);
+    normUniVer = (normUnicodeVersion[0] << 24) + (normUnicodeVersion[1] << 16) + 
+                 (normUnicodeVersion[2] << 8 ) + (normUnicodeVersion[3]);
+    sprepUniVer = (dataVersion[0] << 24) + (dataVersion[1] << 16) + 
+                  (dataVersion[2] << 8 ) + (dataVersion[3]);
+    normCorrVer = profile->indexes[_SPREP_NORM_CORRECTNS_LAST_UNI_VERSION];
+    
+    if(U_FAILURE(*errorCode)){
+        udata_close(dataMemory);
+        return FALSE;
+    }
+    if( normUniVer < sprepUniVer && /* the Unicode version of SPREP file must be less than the Unicode Vesion of the normalization data */
+        normUniVer < normCorrVer && /* the Unicode version of the NormalizationCorrections.txt file should be less than the Unicode Vesion of the normalization data */
+        ((profile->indexes[_SPREP_OPTIONS] & _SPREP_NORMALIZATION_ON) > 0) /* normalization turned on*/
+      ){
+        *errorCode = U_INVALID_FORMAT_ERROR;
+        udata_close(dataMemory);
+        return FALSE;
+    }
+    profile->isDataLoaded = TRUE;
+
+    /* if a different thread set it first, then close the extra data */
+    if(dataMemory!=NULL) {
+        udata_close(dataMemory); /* NULL if it was set correctly */
+    }
+
+
+    return profile->isDataLoaded;
+}
+
+static UStringPrepProfile* 
+usprep_getProfile(const char* path, 
+                  const char* name,
+                  UErrorCode *status){
+
+    UStringPrepProfile* profile = NULL;
+
+    initCache(status);
+
+    if(U_FAILURE(*status)){
+        return NULL;
+    }
+
+    UStringPrepKey stackKey;
+    /* 
+     * const is cast way to save malloc, strcpy and free calls 
+     * we use the passed in pointers for fetching the data from the 
+     * hash table which is safe
+     */
+    stackKey.name = (char*) name;
+    stackKey.path = (char*) path;
+
+    /* fetch the data from the cache */
+    umtx_lock(&usprepMutex);
+    profile = (UStringPrepProfile*) (uhash_get(SHARED_DATA_HASHTABLE,&stackKey));
+    if(profile != NULL) {
+        profile->refCount++;
+    }
+    umtx_unlock(&usprepMutex);
+    
+    if(profile == NULL) {
+        /* else load the data and put the data in the cache */
+        LocalMemory<UStringPrepProfile> newProfile;
+        if(newProfile.allocateInsteadAndReset() == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        }
+
+        /* load the data */
+        if(!loadData(newProfile.getAlias(), path, name, _SPREP_DATA_TYPE, status) || U_FAILURE(*status) ){
+            return NULL;
+        }
+
+        /* get the options */
+        newProfile->doNFKC = (UBool)((newProfile->indexes[_SPREP_OPTIONS] & _SPREP_NORMALIZATION_ON) > 0);
+        newProfile->checkBiDi = (UBool)((newProfile->indexes[_SPREP_OPTIONS] & _SPREP_CHECK_BIDI_ON) > 0);
+
+        LocalMemory<UStringPrepKey> key;
+        LocalMemory<char> keyName;
+        LocalMemory<char> keyPath;
+        if( key.allocateInsteadAndReset() == NULL ||
+            keyName.allocateInsteadAndCopy(static_cast<int32_t>(uprv_strlen(name)+1)) == NULL ||
+            (path != NULL &&
+             keyPath.allocateInsteadAndCopy(static_cast<int32_t>(uprv_strlen(path)+1)) == NULL)
+         ) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            usprep_unload(newProfile.getAlias());
+            return NULL;
+        }
+
+        umtx_lock(&usprepMutex);
+        // If another thread already inserted the same key/value, refcount and cleanup our thread data
+        profile = (UStringPrepProfile*) (uhash_get(SHARED_DATA_HASHTABLE,&stackKey));
+        if(profile != NULL) {
+            profile->refCount++;
+            usprep_unload(newProfile.getAlias());
+        }
+        else {
+            /* initialize the key members */
+            key->name = keyName.orphan();
+            uprv_strcpy(key->name, name);
+            if(path != NULL){
+                key->path = keyPath.orphan();
+                uprv_strcpy(key->path, path);
+            }        
+            profile = newProfile.orphan();
+    
+            /* add the data object to the cache */
+            profile->refCount = 1;
+            uhash_put(SHARED_DATA_HASHTABLE, key.orphan(), profile, status);
+        }
+        umtx_unlock(&usprepMutex);
+    }
+
+    return profile;
+}
+
+U_CAPI UStringPrepProfile* U_EXPORT2
+usprep_open(const char* path, 
+            const char* name,
+            UErrorCode* status){
+
+    if(status == NULL || U_FAILURE(*status)){
+        return NULL;
+    }
+       
+    /* initialize the profile struct members */
+    return usprep_getProfile(path,name,status);
+}
+
+U_CAPI UStringPrepProfile* U_EXPORT2
+usprep_openByType(UStringPrepProfileType type,
+				  UErrorCode* status) {
+    if(status == NULL || U_FAILURE(*status)){
+        return NULL;
+    }
+    int32_t index = (int32_t)type;
+    if (index < 0 || index >= UPRV_LENGTHOF(PROFILE_NAMES)) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    return usprep_open(NULL, PROFILE_NAMES[index], status);
+}
+
+U_CAPI void U_EXPORT2
+usprep_close(UStringPrepProfile* profile){
+    if(profile==NULL){
+        return;
+    }
+
+    umtx_lock(&usprepMutex);
+    /* decrement the ref count*/
+    if(profile->refCount > 0){
+        profile->refCount--;
+    }
+    umtx_unlock(&usprepMutex);
+    
+}
+
+U_CFUNC void 
+uprv_syntaxError(const UChar* rules, 
+                 int32_t pos,
+                 int32_t rulesLen,
+                 UParseError* parseError){
+    if(parseError == NULL){
+        return;
+    }
+    parseError->offset = pos;
+    parseError->line = 0 ; // we are not using line numbers 
+    
+    // for pre-context
+    int32_t start = (pos < U_PARSE_CONTEXT_LEN)? 0 : (pos - (U_PARSE_CONTEXT_LEN-1));
+    int32_t limit = pos;
+    
+    u_memcpy(parseError->preContext,rules+start,limit-start);
+    //null terminate the buffer
+    parseError->preContext[limit-start] = 0;
+    
+    // for post-context; include error rules[pos]
+    start = pos;
+    limit = start + (U_PARSE_CONTEXT_LEN-1);
+    if (limit > rulesLen) {
+        limit = rulesLen;
+    }
+    if (start < rulesLen) {
+        u_memcpy(parseError->postContext,rules+start,limit-start);
+    }
+    //null terminate the buffer
+    parseError->postContext[limit-start]= 0;
+}
+
+
+static inline UStringPrepType
+getValues(uint16_t trieWord, int16_t& value, UBool& isIndex){
+
+    UStringPrepType type;
+    if(trieWord == 0){
+        /* 
+         * Initial value stored in the mapping table 
+         * just return USPREP_TYPE_LIMIT .. so that
+         * the source codepoint is copied to the destination
+         */
+        type = USPREP_TYPE_LIMIT;
+        isIndex =FALSE;
+        value = 0;
+    }else if(trieWord >= _SPREP_TYPE_THRESHOLD){
+        type = (UStringPrepType) (trieWord - _SPREP_TYPE_THRESHOLD);
+        isIndex =FALSE;
+        value = 0;
+    }else{
+        /* get the type */
+        type = USPREP_MAP;
+        /* ascertain if the value is index or delta */
+        if(trieWord & 0x02){
+            isIndex = TRUE;
+            value = trieWord  >> 2; //mask off the lower 2 bits and shift
+        }else{
+            isIndex = FALSE;
+            value = (int16_t)trieWord;
+            value =  (value >> 2);
+        }
+ 
+        if((trieWord>>2) == _SPREP_MAX_INDEX_VALUE){
+            type = USPREP_DELETE;
+            isIndex =FALSE;
+            value = 0;
+        }
+    }
+    return type;
+}
+
+// TODO: change to writing to UnicodeString not UChar *
+static int32_t 
+usprep_map(  const UStringPrepProfile* profile, 
+             const UChar* src, int32_t srcLength, 
+             UChar* dest, int32_t destCapacity,
+             int32_t options,
+             UParseError* parseError,
+             UErrorCode* status ){
+    
+    uint16_t result;
+    int32_t destIndex=0;
+    int32_t srcIndex;
+    UBool allowUnassigned = (UBool) ((options & USPREP_ALLOW_UNASSIGNED)>0);
+    UStringPrepType type;
+    int16_t value;
+    UBool isIndex;
+    const int32_t* indexes = profile->indexes;
+
+    // no error checking the caller check for error and arguments
+    // no string length check the caller finds out the string length
+
+    for(srcIndex=0;srcIndex<srcLength;){
+        UChar32 ch;
+
+        U16_NEXT(src,srcIndex,srcLength,ch);
+        
+        result=0;
+
+        UTRIE_GET16(&profile->sprepTrie,ch,result);
+        
+        type = getValues(result, value, isIndex);
+
+        // check if the source codepoint is unassigned
+        if(type == USPREP_UNASSIGNED && allowUnassigned == FALSE){
+
+            uprv_syntaxError(src,srcIndex-U16_LENGTH(ch), srcLength,parseError);
+            *status = U_STRINGPREP_UNASSIGNED_ERROR;
+            return 0;
+            
+        }else if(type == USPREP_MAP){
+            
+            int32_t index, length;
+
+            if(isIndex){
+                index = value;
+                if(index >= indexes[_SPREP_ONE_UCHAR_MAPPING_INDEX_START] &&
+                         index < indexes[_SPREP_TWO_UCHARS_MAPPING_INDEX_START]){
+                    length = 1;
+                }else if(index >= indexes[_SPREP_TWO_UCHARS_MAPPING_INDEX_START] &&
+                         index < indexes[_SPREP_THREE_UCHARS_MAPPING_INDEX_START]){
+                    length = 2;
+                }else if(index >= indexes[_SPREP_THREE_UCHARS_MAPPING_INDEX_START] &&
+                         index < indexes[_SPREP_FOUR_UCHARS_MAPPING_INDEX_START]){
+                    length = 3;
+                }else{
+                    length = profile->mappingData[index++];
+         
+                }
+
+                /* copy mapping to destination */
+                for(int32_t i=0; i< length; i++){
+                    if(destIndex < destCapacity  ){
+                        dest[destIndex] = profile->mappingData[index+i];
+                    }
+                    destIndex++; /* for pre-flighting */
+                }  
+                continue;
+            }else{
+                // subtract the delta to arrive at the code point
+                ch -= value;
+            }
+
+        }else if(type==USPREP_DELETE){
+             // just consume the codepoint and contine
+            continue;
+        }
+        //copy the code point into destination
+        if(ch <= 0xFFFF){
+            if(destIndex < destCapacity ){
+                dest[destIndex] = (UChar)ch;
+            }
+            destIndex++;
+        }else{
+            if(destIndex+1 < destCapacity ){
+                dest[destIndex]   = U16_LEAD(ch);
+                dest[destIndex+1] = U16_TRAIL(ch);
+            }
+            destIndex +=2;
+        }
+       
+    }
+        
+    return u_terminateUChars(dest, destCapacity, destIndex, status);
+}
+
+/*
+   1) Map -- For each character in the input, check if it has a mapping
+      and, if so, replace it with its mapping.  
+
+   2) Normalize -- Possibly normalize the result of step 1 using Unicode
+      normalization. 
+
+   3) Prohibit -- Check for any characters that are not allowed in the
+      output.  If any are found, return an error.  
+
+   4) Check bidi -- Possibly check for right-to-left characters, and if
+      any are found, make sure that the whole string satisfies the
+      requirements for bidirectional strings.  If the string does not
+      satisfy the requirements for bidirectional strings, return an
+      error.  
+      [Unicode3.2] defines several bidirectional categories; each character
+       has one bidirectional category assigned to it.  For the purposes of
+       the requirements below, an "RandALCat character" is a character that
+       has Unicode bidirectional categories "R" or "AL"; an "LCat character"
+       is a character that has Unicode bidirectional category "L".  Note
+
+
+       that there are many characters which fall in neither of the above
+       definitions; Latin digits (<U+0030> through <U+0039>) are examples of
+       this because they have bidirectional category "EN".
+
+       In any profile that specifies bidirectional character handling, all
+       three of the following requirements MUST be met:
+
+       1) The characters in section 5.8 MUST be prohibited.
+
+       2) If a string contains any RandALCat character, the string MUST NOT
+          contain any LCat character.
+
+       3) If a string contains any RandALCat character, a RandALCat
+          character MUST be the first character of the string, and a
+          RandALCat character MUST be the last character of the string.
+*/
+U_CAPI int32_t U_EXPORT2
+usprep_prepare(   const UStringPrepProfile* profile,
+                  const UChar* src, int32_t srcLength, 
+                  UChar* dest, int32_t destCapacity,
+                  int32_t options,
+                  UParseError* parseError,
+                  UErrorCode* status ){
+
+    // check error status
+    if(U_FAILURE(*status)){
+        return 0;
+    }
+
+    //check arguments
+    if(profile==NULL ||
+            (src==NULL ? srcLength!=0 : srcLength<-1) ||
+            (dest==NULL ? destCapacity!=0 : destCapacity<0)) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    //get the string length
+    if(srcLength < 0){
+        srcLength = u_strlen(src);
+    }
+    // map
+    UnicodeString s1;
+    UChar *b1 = s1.getBuffer(srcLength);
+    if(b1==NULL){
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    int32_t b1Len = usprep_map(profile, src, srcLength,
+                               b1, s1.getCapacity(), options, parseError, status);
+    s1.releaseBuffer(U_SUCCESS(*status) ? b1Len : 0);
+
+    if(*status == U_BUFFER_OVERFLOW_ERROR){
+        // redo processing of string
+        /* we do not have enough room so grow the buffer*/
+        b1 = s1.getBuffer(b1Len);
+        if(b1==NULL){
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return 0;
+        }
+
+        *status = U_ZERO_ERROR; // reset error
+        b1Len = usprep_map(profile, src, srcLength,
+                           b1, s1.getCapacity(), options, parseError, status);
+        s1.releaseBuffer(U_SUCCESS(*status) ? b1Len : 0);
+    }
+    if(U_FAILURE(*status)){
+        return 0;
+    }
+
+    // normalize
+    UnicodeString s2;
+    if(profile->doNFKC){
+        const Normalizer2 *n2 = Normalizer2::getNFKCInstance(*status);
+        FilteredNormalizer2 fn2(*n2, *uniset_getUnicode32Instance(*status));
+        if(U_FAILURE(*status)){
+            return 0;
+        }
+        fn2.normalize(s1, s2, *status);
+    }else{
+        s2.fastCopyFrom(s1);
+    }
+    if(U_FAILURE(*status)){
+        return 0;
+    }
+
+    // Prohibit and checkBiDi in one pass
+    const UChar *b2 = s2.getBuffer();
+    int32_t b2Len = s2.length();
+    UCharDirection direction=U_CHAR_DIRECTION_COUNT, firstCharDir=U_CHAR_DIRECTION_COUNT;
+    UBool leftToRight=FALSE, rightToLeft=FALSE;
+    int32_t rtlPos =-1, ltrPos =-1;
+
+    for(int32_t b2Index=0; b2Index<b2Len;){
+        UChar32 ch = 0;
+        U16_NEXT(b2, b2Index, b2Len, ch);
+
+        uint16_t result;
+        UTRIE_GET16(&profile->sprepTrie,ch,result);
+
+        int16_t value;
+        UBool isIndex;
+        UStringPrepType type = getValues(result, value, isIndex);
+
+        if( type == USPREP_PROHIBITED || 
+            ((result < _SPREP_TYPE_THRESHOLD) && (result & 0x01) /* first bit says it the code point is prohibited*/)
+           ){
+            *status = U_STRINGPREP_PROHIBITED_ERROR;
+            uprv_syntaxError(b2, b2Index-U16_LENGTH(ch), b2Len, parseError);
+            return 0;
+        }
+
+        if(profile->checkBiDi) {
+            direction = ubidi_getClass(ch);
+            if(firstCharDir == U_CHAR_DIRECTION_COUNT){
+                firstCharDir = direction;
+            }
+            if(direction == U_LEFT_TO_RIGHT){
+                leftToRight = TRUE;
+                ltrPos = b2Index-1;
+            }
+            if(direction == U_RIGHT_TO_LEFT || direction == U_RIGHT_TO_LEFT_ARABIC){
+                rightToLeft = TRUE;
+                rtlPos = b2Index-1;
+            }
+        }
+    }
+    if(profile->checkBiDi == TRUE){
+        // satisfy 2
+        if( leftToRight == TRUE && rightToLeft == TRUE){
+            *status = U_STRINGPREP_CHECK_BIDI_ERROR;
+            uprv_syntaxError(b2,(rtlPos>ltrPos) ? rtlPos : ltrPos, b2Len, parseError);
+            return 0;
+        }
+
+        //satisfy 3
+        if( rightToLeft == TRUE && 
+            !((firstCharDir == U_RIGHT_TO_LEFT || firstCharDir == U_RIGHT_TO_LEFT_ARABIC) &&
+              (direction == U_RIGHT_TO_LEFT || direction == U_RIGHT_TO_LEFT_ARABIC))
+           ){
+            *status = U_STRINGPREP_CHECK_BIDI_ERROR;
+            uprv_syntaxError(b2, rtlPos, b2Len, parseError);
+            return FALSE;
+        }
+    }
+    return s2.extract(dest, destCapacity, *status);
+}
+
+
+/* data swapping ------------------------------------------------------------ */
+
+U_CAPI int32_t U_EXPORT2
+usprep_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode) {
+    const UDataInfo *pInfo;
+    int32_t headerSize;
+
+    const uint8_t *inBytes;
+    uint8_t *outBytes;
+
+    const int32_t *inIndexes;
+    int32_t indexes[16];
+
+    int32_t i, offset, count, size;
+
+    /* udata_swapDataHeader checks the arguments */
+    headerSize=udata_swapDataHeader(ds, inData, length, outData, pErrorCode);
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* check data format and format version */
+    pInfo=(const UDataInfo *)((const char *)inData+4);
+    if(!(
+        pInfo->dataFormat[0]==0x53 &&   /* dataFormat="SPRP" */
+        pInfo->dataFormat[1]==0x50 &&
+        pInfo->dataFormat[2]==0x52 &&
+        pInfo->dataFormat[3]==0x50 &&
+        pInfo->formatVersion[0]==3
+    )) {
+        udata_printError(ds, "usprep_swap(): data format %02x.%02x.%02x.%02x (format version %02x) is not recognized as StringPrep .spp data\n",
+                         pInfo->dataFormat[0], pInfo->dataFormat[1],
+                         pInfo->dataFormat[2], pInfo->dataFormat[3],
+                         pInfo->formatVersion[0]);
+        *pErrorCode=U_UNSUPPORTED_ERROR;
+        return 0;
+    }
+
+    inBytes=(const uint8_t *)inData+headerSize;
+    outBytes=(uint8_t *)outData+headerSize;
+
+    inIndexes=(const int32_t *)inBytes;
+
+    if(length>=0) {
+        length-=headerSize;
+        if(length<16*4) {
+            udata_printError(ds, "usprep_swap(): too few bytes (%d after header) for StringPrep .spp data\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+    }
+
+    /* read the first 16 indexes (ICU 2.8/format version 3: _SPREP_INDEX_TOP==16, might grow) */
+    for(i=0; i<16; ++i) {
+        indexes[i]=udata_readInt32(ds, inIndexes[i]);
+    }
+
+    /* calculate the total length of the data */
+    size=
+        16*4+ /* size of indexes[] */
+        indexes[_SPREP_INDEX_TRIE_SIZE]+
+        indexes[_SPREP_INDEX_MAPPING_DATA_SIZE];
+
+    if(length>=0) {
+        if(length<size) {
+            udata_printError(ds, "usprep_swap(): too few bytes (%d after header) for all of StringPrep .spp data\n",
+                             length);
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        /* copy the data for inaccessible bytes */
+        if(inBytes!=outBytes) {
+            uprv_memcpy(outBytes, inBytes, size);
+        }
+
+        offset=0;
+
+        /* swap the int32_t indexes[] */
+        count=16*4;
+        ds->swapArray32(ds, inBytes, count, outBytes, pErrorCode);
+        offset+=count;
+
+        /* swap the UTrie */
+        count=indexes[_SPREP_INDEX_TRIE_SIZE];
+        utrie_swap(ds, inBytes+offset, count, outBytes+offset, pErrorCode);
+        offset+=count;
+
+        /* swap the uint16_t mappingTable[] */
+        count=indexes[_SPREP_INDEX_MAPPING_DATA_SIZE];
+        ds->swapArray16(ds, inBytes+offset, count, outBytes+offset, pErrorCode);
+        //offset+=count;
+    }
+
+    return headerSize+size;
+}
+
+#endif /* #if !UCONFIG_NO_IDNA */
diff --git a/thirdparty/icu4c/common/ustack.cpp b/thirdparty/icu4c/common/ustack.cpp
new file mode 100644
index 0000000000..fb314b0ebe
--- /dev/null
+++ b/thirdparty/icu4c/common/ustack.cpp
@@ -0,0 +1,63 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2003-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "uvector.h"
+
+U_NAMESPACE_BEGIN
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UStack)
+
+UStack::UStack(UErrorCode &status) :
+    UVector(status)
+{
+}
+
+UStack::UStack(int32_t initialCapacity, UErrorCode &status) :
+    UVector(initialCapacity, status)
+{
+}
+
+UStack::UStack(UObjectDeleter *d, UElementsAreEqual *c, UErrorCode &status) :
+    UVector(d, c, status)
+{
+}
+
+UStack::UStack(UObjectDeleter *d, UElementsAreEqual *c, int32_t initialCapacity, UErrorCode &status) :
+    UVector(d, c, initialCapacity, status)
+{
+}
+
+UStack::~UStack() {}
+
+void* UStack::pop(void) {
+    int32_t n = size() - 1;
+    void* result = 0;
+    if (n >= 0) {
+        result = elementAt(n);
+        removeElementAt(n);
+    }
+    return result;
+}
+
+int32_t UStack::popi(void) {
+    int32_t n = size() - 1;
+    int32_t result = 0;
+    if (n >= 0) {
+        result = elementAti(n);
+        removeElementAt(n);
+    }
+    return result;
+}
+
+int32_t UStack::search(void* obj) const {
+    int32_t i = indexOf(obj);
+    return (i >= 0) ? size() - i : i;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ustr_cnv.cpp b/thirdparty/icu4c/common/ustr_cnv.cpp
new file mode 100644
index 0000000000..9a25a9905a
--- /dev/null
+++ b/thirdparty/icu4c/common/ustr_cnv.cpp
@@ -0,0 +1,256 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1998-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ustr_cnv.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004aug24
+*   created by: Markus W. Scherer
+*
+*   Character conversion functions moved here from ustring.c
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+#include "unicode/ustring.h"
+#include "unicode/ucnv.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "umutex.h"
+#include "ustr_cnv.h"
+#include "ucnv_bld.h"
+
+/* mutexed access to a shared default converter ----------------------------- */
+
+static UConverter *gDefaultConverter = NULL;
+
+U_CAPI UConverter* U_EXPORT2
+u_getDefaultConverter(UErrorCode *status)
+{
+    UConverter *converter = NULL;
+    
+    if (gDefaultConverter != NULL) {
+        icu::umtx_lock(NULL);
+        
+        /* need to check to make sure it wasn't taken out from under us */
+        if (gDefaultConverter != NULL) {
+            converter = gDefaultConverter;
+            gDefaultConverter = NULL;
+        }
+        icu::umtx_unlock(NULL);
+    }
+
+    /* if the cache was empty, create a converter */
+    if(converter == NULL) {
+        converter = ucnv_open(NULL, status);
+        if(U_FAILURE(*status)) {
+            ucnv_close(converter);
+            converter = NULL;
+        }
+    }
+
+    return converter;
+}
+
+U_CAPI void U_EXPORT2
+u_releaseDefaultConverter(UConverter *converter)
+{
+    if(gDefaultConverter == NULL) {
+        if (converter != NULL) {
+            ucnv_reset(converter);
+        }
+        ucnv_enableCleanup();
+        icu::umtx_lock(NULL);
+        if(gDefaultConverter == NULL) {
+            gDefaultConverter = converter;
+            converter = NULL;
+        }
+        icu::umtx_unlock(NULL);
+    }
+
+    if(converter != NULL) {
+        ucnv_close(converter);
+    }
+}
+
+U_CAPI void U_EXPORT2
+u_flushDefaultConverter()
+{
+    UConverter *converter = NULL;
+    
+    if (gDefaultConverter != NULL) {
+        icu::umtx_lock(NULL);
+        
+        /* need to check to make sure it wasn't taken out from under us */
+        if (gDefaultConverter != NULL) {
+            converter = gDefaultConverter;
+            gDefaultConverter = NULL;
+        }
+        icu::umtx_unlock(NULL);
+    }
+
+    /* if the cache was populated, flush it */
+    if(converter != NULL) {
+         ucnv_close(converter);
+    }
+}
+
+
+/* conversions between char* and UChar* ------------------------------------- */
+
+/* maximum string length for u_uastrcpy() and u_austrcpy() implementations */
+#define MAX_STRLEN 0x0FFFFFFF
+
+/*
+ returns the minimum of (the length of the null-terminated string) and n.
+*/
+static int32_t u_astrnlen(const char *s1, int32_t n)
+{
+    int32_t len = 0;
+
+    if (s1)
+    {
+        while (n-- && *(s1++))
+        {
+            len++;
+        }
+    }
+    return len;
+}
+
+U_CAPI UChar*  U_EXPORT2
+u_uastrncpy(UChar *ucs1,
+           const char *s2,
+           int32_t n)
+{
+  UChar *target = ucs1;
+  UErrorCode err = U_ZERO_ERROR;
+  UConverter *cnv = u_getDefaultConverter(&err);
+  if(U_SUCCESS(err) && cnv != NULL) {
+    ucnv_reset(cnv);
+    ucnv_toUnicode(cnv,
+                   &target,
+                   ucs1+n,
+                   &s2,
+                   s2+u_astrnlen(s2, n),
+                   NULL,
+                   TRUE,
+                   &err);
+    ucnv_reset(cnv); /* be good citizens */
+    u_releaseDefaultConverter(cnv);
+    if(U_FAILURE(err) && (err != U_BUFFER_OVERFLOW_ERROR) ) {
+      *ucs1 = 0; /* failure */
+    }
+    if(target < (ucs1+n)) { /* U_BUFFER_OVERFLOW_ERROR isn't an err, just means no termination will happen. */
+      *target = 0;  /* terminate */
+    }
+  } else {
+    *ucs1 = 0;
+  }
+  return ucs1;
+}
+
+U_CAPI UChar*  U_EXPORT2
+u_uastrcpy(UChar *ucs1,
+          const char *s2 )
+{
+  UErrorCode err = U_ZERO_ERROR;
+  UConverter *cnv = u_getDefaultConverter(&err);
+  if(U_SUCCESS(err) && cnv != NULL) {
+    ucnv_toUChars(cnv,
+                    ucs1,
+                    MAX_STRLEN,
+                    s2,
+                    (int32_t)uprv_strlen(s2),
+                    &err);
+    u_releaseDefaultConverter(cnv);
+    if(U_FAILURE(err)) {
+      *ucs1 = 0;
+    }
+  } else {
+    *ucs1 = 0;
+  }
+  return ucs1;
+}
+
+/*
+ returns the minimum of (the length of the null-terminated string) and n.
+*/
+static int32_t u_ustrnlen(const UChar *ucs1, int32_t n)
+{
+    int32_t len = 0;
+
+    if (ucs1)
+    {
+        while (n-- && *(ucs1++))
+        {
+            len++;
+        }
+    }
+    return len;
+}
+
+U_CAPI char*  U_EXPORT2
+u_austrncpy(char *s1,
+        const UChar *ucs2,
+        int32_t n)
+{
+  char *target = s1;
+  UErrorCode err = U_ZERO_ERROR;
+  UConverter *cnv = u_getDefaultConverter(&err);
+  if(U_SUCCESS(err) && cnv != NULL) {
+    ucnv_reset(cnv);
+    ucnv_fromUnicode(cnv,
+                  &target,
+                  s1+n,
+                  &ucs2,
+                  ucs2+u_ustrnlen(ucs2, n),
+                  NULL,
+                  TRUE,
+                  &err);
+    ucnv_reset(cnv); /* be good citizens */
+    u_releaseDefaultConverter(cnv);
+    if(U_FAILURE(err) && (err != U_BUFFER_OVERFLOW_ERROR) ) {
+      *s1 = 0; /* failure */
+    }
+    if(target < (s1+n)) { /* U_BUFFER_OVERFLOW_ERROR isn't an err, just means no termination will happen. */
+      *target = 0;  /* terminate */
+    }
+  } else {
+    *s1 = 0;
+  }
+  return s1;
+}
+
+U_CAPI char*  U_EXPORT2
+u_austrcpy(char *s1,
+         const UChar *ucs2 )
+{
+  UErrorCode err = U_ZERO_ERROR;
+  UConverter *cnv = u_getDefaultConverter(&err);
+  if(U_SUCCESS(err) && cnv != NULL) {
+    int32_t len = ucnv_fromUChars(cnv,
+                  s1,
+                  MAX_STRLEN,
+                  ucs2,
+                  -1,
+                  &err);
+    u_releaseDefaultConverter(cnv);
+    s1[len] = 0;
+  } else {
+    *s1 = 0;
+  }
+  return s1;
+}
+
+#endif
diff --git a/thirdparty/icu4c/common/ustr_cnv.h b/thirdparty/icu4c/common/ustr_cnv.h
new file mode 100644
index 0000000000..861e3ebff0
--- /dev/null
+++ b/thirdparty/icu4c/common/ustr_cnv.h
@@ -0,0 +1,51 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 1999-2010, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ustr_cnv.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004Aug27
+*   created by: George Rhoten
+*/
+
+#ifndef USTR_CNV_IMP_H
+#define USTR_CNV_IMP_H
+
+#include "unicode/utypes.h"
+#include "unicode/ucnv.h"
+
+#if !UCONFIG_NO_CONVERSION
+
+/**
+ * Get the default converter. This is a commonly used converter
+ * that is used for the ustring and UnicodeString API.
+ * Remember to use the u_releaseDefaultConverter when you are done.
+ * @internal
+ */
+U_CAPI UConverter* U_EXPORT2
+u_getDefaultConverter(UErrorCode *status);
+
+
+/**
+ * Release the default converter to the converter cache.
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+u_releaseDefaultConverter(UConverter *converter);
+
+/**
+ * Flush the default converter, if cached. 
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+u_flushDefaultConverter(void);
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/ustr_imp.h b/thirdparty/icu4c/common/ustr_imp.h
new file mode 100644
index 0000000000..3c4b9cc2a5
--- /dev/null
+++ b/thirdparty/icu4c/common/ustr_imp.h
@@ -0,0 +1,155 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*  
+**********************************************************************
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   file name:  ustr_imp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001jan30
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __USTR_IMP_H__
+#define __USTR_IMP_H__
+
+#include "unicode/utypes.h"
+#include "unicode/utf8.h"
+
+/**
+ * Internal option for unorm_cmpEquivFold() for strncmp style.
+ * If set, checks for both string length and terminating NUL.
+ */
+#define _STRNCMP_STYLE 0x1000
+
+/**
+ * Compare two strings in code point order or code unit order.
+ * Works in strcmp style (both lengths -1),
+ * strncmp style (lengths equal and >=0, flag true),
+ * and memcmp/UnicodeString style (at least one length >=0).
+ */
+U_CFUNC int32_t U_EXPORT2
+uprv_strCompare(const UChar *s1, int32_t length1,
+                const UChar *s2, int32_t length2,
+                UBool strncmpStyle, UBool codePointOrder);
+
+U_CAPI int32_t U_EXPORT2 
+ustr_hashUCharsN(const UChar *str, int32_t length);
+
+U_CAPI int32_t U_EXPORT2 
+ustr_hashCharsN(const char *str, int32_t length);
+
+U_CAPI int32_t U_EXPORT2
+ustr_hashICharsN(const char *str, int32_t length);
+
+/**
+ * Convert an ASCII-range lowercase character to uppercase.
+ * 
+ * @param c A UChar.
+ * @return If UChar is a lowercase ASCII character, returns the uppercase version.
+ *         Otherwise, returns the input character.
+ */
+U_CAPI UChar U_EXPORT2
+u_asciiToUpper(UChar c);
+
+// TODO: Add u_asciiToLower if/when there is a need for it.
+
+/**
+ * NUL-terminate a UChar * string if possible.
+ * If length  < destCapacity then NUL-terminate.
+ * If length == destCapacity then do not terminate but set U_STRING_NOT_TERMINATED_WARNING.
+ * If length  > destCapacity then do not terminate but set U_BUFFER_OVERFLOW_ERROR.
+ *
+ * @param dest Destination buffer, can be NULL if destCapacity==0.
+ * @param destCapacity Number of UChars available at dest.
+ * @param length Number of UChars that were (to be) written to dest.
+ * @param pErrorCode ICU error code.
+ * @return length
+ */
+U_CAPI int32_t U_EXPORT2
+u_terminateUChars(UChar *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * NUL-terminate a char * string if possible.
+ * Same as u_terminateUChars() but for a different string type.
+ */
+U_CAPI int32_t U_EXPORT2
+u_terminateChars(char *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * NUL-terminate a UChar32 * string if possible.
+ * Same as u_terminateUChars() but for a different string type.
+ */
+U_CAPI int32_t U_EXPORT2
+u_terminateUChar32s(UChar32 *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * NUL-terminate a wchar_t * string if possible.
+ * Same as u_terminateUChars() but for a different string type.
+ */
+U_CAPI int32_t U_EXPORT2
+u_terminateWChars(wchar_t *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * Counts the bytes of any whole valid sequence for a UTF-8 lead byte.
+ * Returns 1 for ASCII 0..0x7f.
+ * Returns 0 for 0x80..0xc1 as well as for 0xf5..0xff.
+ * leadByte might be evaluated multiple times.
+ *
+ * @param leadByte The first byte of a UTF-8 sequence. Must be 0..0xff.
+ * @return 0..4
+ */
+#define U8_COUNT_BYTES(leadByte) \
+    (U8_IS_SINGLE(leadByte) ? 1 : U8_COUNT_BYTES_NON_ASCII(leadByte))
+
+/**
+ * Counts the bytes of any whole valid sequence for a UTF-8 lead byte.
+ * Returns 0 for 0x00..0xc1 as well as for 0xf5..0xff.
+ * leadByte might be evaluated multiple times.
+ *
+ * @param leadByte The first byte of a UTF-8 sequence. Must be 0..0xff.
+ * @return 0 or 2..4
+ */
+#define U8_COUNT_BYTES_NON_ASCII(leadByte) \
+    (U8_IS_LEAD(leadByte) ? ((uint8_t)(leadByte)>=0xe0)+((uint8_t)(leadByte)>=0xf0)+2 : 0)
+
+#ifdef __cplusplus
+
+U_NAMESPACE_BEGIN
+
+class UTF8 {
+public:
+    UTF8() = delete;  // all static
+
+    /**
+     * Is t a valid UTF-8 trail byte?
+     *
+     * @param prev Must be the preceding lead byte if i==1 and length>=3;
+     *             otherwise ignored.
+     * @param t The i-th byte following the lead byte.
+     * @param i The index (1..3) of byte t in the byte sequence. 0<i<length
+     * @param length The length (2..4) of the byte sequence according to the lead byte.
+     * @return true if t is a valid trail byte in this context.
+     */
+    static inline UBool isValidTrail(int32_t prev, uint8_t t, int32_t i, int32_t length) {
+        // The first trail byte after a 3- or 4-byte lead byte
+        // needs to be validated together with its lead byte.
+        if (length <= 2 || i > 1) {
+            return U8_IS_TRAIL(t);
+        } else if (length == 3) {
+            return U8_IS_VALID_LEAD3_AND_T1(prev, t);
+        } else {  // length == 4
+            return U8_IS_VALID_LEAD4_AND_T1(prev, t);
+        }
+    }
+};
+
+U_NAMESPACE_END
+
+#endif  // __cplusplus
+
+#endif
diff --git a/thirdparty/icu4c/common/ustr_titlecase_brkiter.cpp b/thirdparty/icu4c/common/ustr_titlecase_brkiter.cpp
new file mode 100644
index 0000000000..457905eb60
--- /dev/null
+++ b/thirdparty/icu4c/common/ustr_titlecase_brkiter.cpp
@@ -0,0 +1,237 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ustr_titlecase_brkiter.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011may30
+*   created by: Markus W. Scherer
+*
+*   Titlecasing functions that are based on BreakIterator
+*   were moved here to break dependency cycles among parts of the common library.
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+#include "unicode/brkiter.h"
+#include "unicode/casemap.h"
+#include "unicode/chariter.h"
+#include "unicode/localpointer.h"
+#include "unicode/ubrk.h"
+#include "unicode/ucasemap.h"
+#include "unicode/utext.h"
+#include "cmemory.h"
+#include "uassert.h"
+#include "ucase.h"
+#include "ucasemap_imp.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Whole-string BreakIterator.
+ * Titlecasing only calls setText(), first(), and next().
+ * We implement the rest only to satisfy the abstract interface.
+ */
+class WholeStringBreakIterator : public BreakIterator {
+public:
+    WholeStringBreakIterator() : BreakIterator(), length(0) {}
+    ~WholeStringBreakIterator() U_OVERRIDE;
+    UBool operator==(const BreakIterator&) const U_OVERRIDE;
+    WholeStringBreakIterator *clone() const U_OVERRIDE;
+    static UClassID U_EXPORT2 getStaticClassID();
+    UClassID getDynamicClassID() const U_OVERRIDE;
+    CharacterIterator &getText() const U_OVERRIDE;
+    UText *getUText(UText *fillIn, UErrorCode &errorCode) const U_OVERRIDE;
+    void  setText(const UnicodeString &text) U_OVERRIDE;
+    void  setText(UText *text, UErrorCode &errorCode) U_OVERRIDE;
+    void  adoptText(CharacterIterator* it) U_OVERRIDE;
+    int32_t first() U_OVERRIDE;
+    int32_t last() U_OVERRIDE;
+    int32_t previous() U_OVERRIDE;
+    int32_t next() U_OVERRIDE;
+    int32_t current() const U_OVERRIDE;
+    int32_t following(int32_t offset) U_OVERRIDE;
+    int32_t preceding(int32_t offset) U_OVERRIDE;
+    UBool isBoundary(int32_t offset) U_OVERRIDE;
+    int32_t next(int32_t n) U_OVERRIDE;
+    WholeStringBreakIterator *createBufferClone(void *stackBuffer, int32_t &BufferSize,
+                                                UErrorCode &errorCode) U_OVERRIDE;
+    WholeStringBreakIterator &refreshInputText(UText *input, UErrorCode &errorCode) U_OVERRIDE;
+
+private:
+    int32_t length;
+};
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(WholeStringBreakIterator)
+
+WholeStringBreakIterator::~WholeStringBreakIterator() {}
+UBool WholeStringBreakIterator::operator==(const BreakIterator&) const { return FALSE; }
+WholeStringBreakIterator *WholeStringBreakIterator::clone() const { return nullptr; }
+
+CharacterIterator &WholeStringBreakIterator::getText() const {
+    UPRV_UNREACHABLE;  // really should not be called
+}
+UText *WholeStringBreakIterator::getUText(UText * /*fillIn*/, UErrorCode &errorCode) const {
+    if (U_SUCCESS(errorCode)) {
+        errorCode = U_UNSUPPORTED_ERROR;
+    }
+    return nullptr;
+}
+
+void  WholeStringBreakIterator::setText(const UnicodeString &text) {
+    length = text.length();
+}
+void  WholeStringBreakIterator::setText(UText *text, UErrorCode &errorCode) {
+    if (U_SUCCESS(errorCode)) {
+        int64_t length64 = utext_nativeLength(text);
+        if (length64 <= INT32_MAX) {
+            length = (int32_t)length64;
+        } else {
+            errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        }
+    }
+}
+void  WholeStringBreakIterator::adoptText(CharacterIterator*) {
+    UPRV_UNREACHABLE;  // should not be called
+}
+
+int32_t WholeStringBreakIterator::first() { return 0; }
+int32_t WholeStringBreakIterator::last() { return length; }
+int32_t WholeStringBreakIterator::previous() { return 0; }
+int32_t WholeStringBreakIterator::next() { return length; }
+int32_t WholeStringBreakIterator::current() const { return 0; }
+int32_t WholeStringBreakIterator::following(int32_t /*offset*/) { return length; }
+int32_t WholeStringBreakIterator::preceding(int32_t /*offset*/) { return 0; }
+UBool WholeStringBreakIterator::isBoundary(int32_t /*offset*/) { return FALSE; }
+int32_t WholeStringBreakIterator::next(int32_t /*n*/) { return length; }
+
+WholeStringBreakIterator *WholeStringBreakIterator::createBufferClone(
+        void * /*stackBuffer*/, int32_t & /*BufferSize*/, UErrorCode &errorCode) {
+    if (U_SUCCESS(errorCode)) {
+        errorCode = U_UNSUPPORTED_ERROR;
+    }
+    return nullptr;
+}
+WholeStringBreakIterator &WholeStringBreakIterator::refreshInputText(
+        UText * /*input*/, UErrorCode &errorCode) {
+    if (U_SUCCESS(errorCode)) {
+        errorCode = U_UNSUPPORTED_ERROR;
+    }
+    return *this;
+}
+
+U_CFUNC
+BreakIterator *ustrcase_getTitleBreakIterator(
+        const Locale *locale, const char *locID, uint32_t options, BreakIterator *iter,
+        LocalPointer<BreakIterator> &ownedIter, UErrorCode &errorCode) {
+    if (U_FAILURE(errorCode)) { return nullptr; }
+    options &= U_TITLECASE_ITERATOR_MASK;
+    if (options != 0 && iter != nullptr) {
+        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return nullptr;
+    }
+    if (iter == nullptr) {
+        switch (options) {
+        case 0:
+            iter = BreakIterator::createWordInstance(
+                locale != nullptr ? *locale : Locale(locID), errorCode);
+            break;
+        case U_TITLECASE_WHOLE_STRING:
+            iter = new WholeStringBreakIterator();
+            if (iter == nullptr) {
+                errorCode = U_MEMORY_ALLOCATION_ERROR;
+            }
+            break;
+        case U_TITLECASE_SENTENCES:
+            iter = BreakIterator::createSentenceInstance(
+                locale != nullptr ? *locale : Locale(locID), errorCode);
+            break;
+        default:
+            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+            break;
+        }
+        ownedIter.adoptInstead(iter);
+    }
+    return iter;
+}
+
+int32_t CaseMap::toTitle(
+        const char *locale, uint32_t options, BreakIterator *iter,
+        const UChar *src, int32_t srcLength,
+        UChar *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    LocalPointer<BreakIterator> ownedIter;
+    iter = ustrcase_getTitleBreakIterator(nullptr, locale, options, iter, ownedIter, errorCode);
+    if(iter==NULL) {
+        return 0;
+    }
+    UnicodeString s(srcLength<0, src, srcLength);
+    iter->setText(s);
+    return ustrcase_map(
+        ustrcase_getCaseLocale(locale), options, iter,
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToTitle, edits, errorCode);
+}
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+U_CAPI int32_t U_EXPORT2
+u_strToTitle(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             UBreakIterator *titleIter,
+             const char *locale,
+             UErrorCode *pErrorCode) {
+    LocalPointer<BreakIterator> ownedIter;
+    BreakIterator *iter = ustrcase_getTitleBreakIterator(
+        nullptr, locale, 0, reinterpret_cast<BreakIterator *>(titleIter),
+        ownedIter, *pErrorCode);
+    if (iter == nullptr) {
+        return 0;
+    }
+    UnicodeString s(srcLength<0, src, srcLength);
+    iter->setText(s);
+    return ustrcase_mapWithOverlap(
+        ustrcase_getCaseLocale(locale), 0, iter,
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToTitle, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+ucasemap_toTitle(UCaseMap *csm,
+                 UChar *dest, int32_t destCapacity,
+                 const UChar *src, int32_t srcLength,
+                 UErrorCode *pErrorCode) {
+    if (U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if (csm->iter == NULL) {
+        LocalPointer<BreakIterator> ownedIter;
+        BreakIterator *iter = ustrcase_getTitleBreakIterator(
+            nullptr, csm->locale, csm->options, nullptr, ownedIter, *pErrorCode);
+        if (iter == nullptr) {
+            return 0;
+        }
+        csm->iter = ownedIter.orphan();
+    }
+    UnicodeString s(srcLength<0, src, srcLength);
+    csm->iter->setText(s);
+    return ustrcase_map(
+        csm->caseLocale, csm->options, csm->iter,
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToTitle, NULL, *pErrorCode);
+}
+
+#endif  // !UCONFIG_NO_BREAK_ITERATION
diff --git a/thirdparty/icu4c/common/ustr_wcs.cpp b/thirdparty/icu4c/common/ustr_wcs.cpp
new file mode 100644
index 0000000000..e9f278e969
--- /dev/null
+++ b/thirdparty/icu4c/common/ustr_wcs.cpp
@@ -0,0 +1,535 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2001-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ustr_wcs.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2004sep07
+*   created by: Markus W. Scherer
+*
+*   u_strToWCS() and u_strFromWCS() functions
+*   moved here from ustrtrns.c for better modularization.
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "cstring.h"
+#include "cwchar.h"
+#include "cmemory.h"
+#include "ustr_imp.h"
+#include "ustr_cnv.h"
+
+#if defined(U_WCHAR_IS_UTF16) || defined(U_WCHAR_IS_UTF32) || !UCONFIG_NO_CONVERSION
+
+#define _STACK_BUFFER_CAPACITY 1000
+#define _BUFFER_CAPACITY_MULTIPLIER 2
+
+#if !defined(U_WCHAR_IS_UTF16) && !defined(U_WCHAR_IS_UTF32)
+// TODO: We should use CharString for char buffers and UnicodeString for UChar buffers.
+// Then we could change this to work only with wchar_t buffers.
+static inline UBool 
+u_growAnyBufferFromStatic(void *context,
+                       void **pBuffer, int32_t *pCapacity, int32_t reqCapacity,
+                       int32_t length, int32_t size) {
+    // Use char* not void* to avoid the compiler's strict-aliasing assumptions
+    // and related warnings.
+    char *newBuffer=(char *)uprv_malloc(reqCapacity*size);
+    if(newBuffer!=NULL) {
+        if(length>0) {
+            uprv_memcpy(newBuffer, *pBuffer, (size_t)length*size);
+        }
+        *pCapacity=reqCapacity;
+    } else {
+        *pCapacity=0;
+    }
+
+    /* release the old pBuffer if it was not statically allocated */
+    if(*pBuffer!=(char *)context) {
+        uprv_free(*pBuffer);
+    }
+
+    *pBuffer=newBuffer;
+    return (UBool)(newBuffer!=NULL);
+}
+
+/* helper function */
+static wchar_t* 
+_strToWCS(wchar_t *dest, 
+           int32_t destCapacity,
+           int32_t *pDestLength,
+           const UChar *src, 
+           int32_t srcLength,
+           UErrorCode *pErrorCode){
+
+    char stackBuffer [_STACK_BUFFER_CAPACITY];
+    char* tempBuf = stackBuffer;
+    int32_t tempBufCapacity = _STACK_BUFFER_CAPACITY;
+    char* tempBufLimit = stackBuffer + tempBufCapacity;
+    UConverter* conv = NULL;
+    char* saveBuf = tempBuf;
+    wchar_t* intTarget=NULL;
+    int32_t intTargetCapacity=0;
+    int count=0,retVal=0;
+    
+    const UChar *pSrcLimit =NULL;
+    const UChar *pSrc = src;
+
+    conv = u_getDefaultConverter(pErrorCode);
+    
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+    
+    if(srcLength == -1){
+        srcLength = u_strlen(pSrc);
+    }
+    
+    pSrcLimit = pSrc + srcLength;
+
+    for(;;) {
+        /* reset the error state */
+        *pErrorCode = U_ZERO_ERROR;
+
+        /* convert to chars using default converter */
+        ucnv_fromUnicode(conv,&tempBuf,tempBufLimit,&pSrc,pSrcLimit,NULL,(UBool)(pSrc==pSrcLimit),pErrorCode);
+        count =(tempBuf - saveBuf);
+        
+        /* This should rarely occur */
+        if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR){
+            tempBuf = saveBuf;
+            
+            /* we dont have enough room on the stack grow the buffer */
+            int32_t newCapacity = 2 * srcLength;
+            if(newCapacity <= tempBufCapacity) {
+                newCapacity = _BUFFER_CAPACITY_MULTIPLIER * tempBufCapacity;
+            }
+            if(!u_growAnyBufferFromStatic(stackBuffer,(void**) &tempBuf, &tempBufCapacity,
+                    newCapacity, count, 1)) {
+                goto cleanup;
+            }
+          
+           saveBuf = tempBuf;
+           tempBufLimit = tempBuf + tempBufCapacity;
+           tempBuf = tempBuf + count;
+
+        } else {
+            break;
+        }
+    }
+
+    if(U_FAILURE(*pErrorCode)){
+        goto cleanup;
+    }
+
+    /* done with conversion null terminate the char buffer */
+    if(count>=tempBufCapacity){
+        tempBuf = saveBuf;
+        /* we dont have enough room on the stack grow the buffer */
+        if(!u_growAnyBufferFromStatic(stackBuffer,(void**) &tempBuf, &tempBufCapacity, 
+                count+1, count, 1)) {
+            goto cleanup;
+        }              
+       saveBuf = tempBuf;
+    }
+    
+    saveBuf[count]=0;
+      
+
+    /* allocate more space than required 
+     * here we assume that every char requires 
+     * no more than 2 wchar_ts
+     */
+    intTargetCapacity =  (count * _BUFFER_CAPACITY_MULTIPLIER + 1) /*for null termination */;
+    intTarget = (wchar_t*)uprv_malloc( intTargetCapacity * sizeof(wchar_t) );
+
+    if(intTarget){
+
+        int32_t nulLen = 0;
+        int32_t remaining = intTargetCapacity;
+        wchar_t* pIntTarget=intTarget;
+        tempBuf = saveBuf;
+        
+        /* now convert the mbs to wcs */
+        for(;;){
+            
+            /* we can call the system API since we are sure that
+             * there is atleast 1 null in the input
+             */
+            retVal = uprv_mbstowcs(pIntTarget,(tempBuf+nulLen),remaining);
+            
+            if(retVal==-1){
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                break;
+            }else if(retVal== remaining){/* should never occur */
+                int numWritten = (pIntTarget-intTarget);
+                u_growAnyBufferFromStatic(NULL,(void**) &intTarget,
+                                          &intTargetCapacity,
+                                          intTargetCapacity * _BUFFER_CAPACITY_MULTIPLIER,
+                                          numWritten,
+                                          sizeof(wchar_t));
+                pIntTarget = intTarget;
+                remaining=intTargetCapacity;
+
+                if(nulLen!=count){ /*there are embedded nulls*/
+                    pIntTarget+=numWritten;
+                    remaining-=numWritten;
+                }
+
+            }else{
+                int32_t nulVal;
+                /*scan for nulls */
+                /* we donot check for limit since tempBuf is null terminated */
+                while(tempBuf[nulLen++] != 0){
+                }
+                nulVal = (nulLen < srcLength) ? 1 : 0; 
+                pIntTarget = pIntTarget + retVal+nulVal;
+                remaining -=(retVal+nulVal);
+            
+                /* check if we have reached the source limit*/
+                if(nulLen>=(count)){
+                    break;
+                }
+            }
+        }
+        count = (int32_t)(pIntTarget-intTarget);
+       
+        if(0 < count && count <= destCapacity){
+            uprv_memcpy(dest, intTarget, (size_t)count*sizeof(wchar_t));
+        }  
+
+        if(pDestLength){
+            *pDestLength = count;
+        }
+
+        /* free the allocated memory */
+        uprv_free(intTarget);
+
+    }else{
+        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+    }
+cleanup:
+    /* are we still using stack buffer */
+    if(stackBuffer != saveBuf){
+        uprv_free(saveBuf);
+    }
+    u_terminateWChars(dest,destCapacity,count,pErrorCode);
+
+    u_releaseDefaultConverter(conv);
+
+    return dest;
+}
+#endif
+
+U_CAPI wchar_t* U_EXPORT2
+u_strToWCS(wchar_t *dest, 
+           int32_t destCapacity,
+           int32_t *pDestLength,
+           const UChar *src, 
+           int32_t srcLength,
+           UErrorCode *pErrorCode){
+
+    /* args check */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+        
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    
+#ifdef U_WCHAR_IS_UTF16
+    /* wchar_t is UTF-16 just do a memcpy */
+    if(srcLength == -1){
+        srcLength = u_strlen(src);
+    }
+    if(0 < srcLength && srcLength <= destCapacity){
+        u_memcpy((UChar *)dest, src, srcLength);
+    }
+    if(pDestLength){
+       *pDestLength = srcLength;
+    }
+
+    u_terminateUChars((UChar *)dest,destCapacity,srcLength,pErrorCode);
+
+    return dest;
+
+#elif defined U_WCHAR_IS_UTF32
+    
+    return (wchar_t*)u_strToUTF32((UChar32*)dest, destCapacity, pDestLength,
+                                  src, srcLength, pErrorCode);
+
+#else
+    
+    return _strToWCS(dest,destCapacity,pDestLength,src,srcLength, pErrorCode);
+    
+#endif
+
+}
+
+#if !defined(U_WCHAR_IS_UTF16) && !defined(U_WCHAR_IS_UTF32)
+/* helper function */
+static UChar* 
+_strFromWCS( UChar   *dest,
+             int32_t destCapacity, 
+             int32_t *pDestLength,
+             const wchar_t *src,
+             int32_t srcLength,
+             UErrorCode *pErrorCode)
+{
+    int32_t retVal =0, count =0 ;
+    UConverter* conv = NULL;
+    UChar* pTarget = NULL;
+    UChar* pTargetLimit = NULL;
+    UChar* target = NULL;
+    
+    UChar uStack [_STACK_BUFFER_CAPACITY];
+
+    wchar_t wStack[_STACK_BUFFER_CAPACITY];
+    wchar_t* pWStack = wStack;
+
+
+    char cStack[_STACK_BUFFER_CAPACITY];
+    int32_t cStackCap = _STACK_BUFFER_CAPACITY;
+    char* pCSrc=cStack;
+    char* pCSave=pCSrc;
+    char* pCSrcLimit=NULL;
+
+    const wchar_t* pSrc = src;
+    const wchar_t* pSrcLimit = NULL;
+
+    if(srcLength ==-1){
+        /* if the wchar_t source is null terminated we can safely
+         * assume that there are no embedded nulls, this is a fast
+         * path for null terminated strings.
+         */
+        for(;;){
+            /* convert wchars  to chars */
+            retVal = uprv_wcstombs(pCSrc,src, cStackCap);
+    
+            if(retVal == -1){
+                *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+                goto cleanup;
+            }else if(retVal >= (cStackCap-1)){
+                /* Should rarely occur */
+                u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
+                    cStackCap * _BUFFER_CAPACITY_MULTIPLIER, 0, sizeof(char));
+                pCSave = pCSrc;
+            }else{
+                /* converted every thing */
+                pCSrc = pCSrc+retVal;
+                break;
+            }
+        }
+        
+    }else{
+        /* here the source is not null terminated 
+         * so it may have nulls embeded and we need to
+         * do some extra processing 
+         */
+        int32_t remaining =cStackCap;
+        
+        pSrcLimit = src + srcLength;
+
+        for(;;){
+            int32_t nulLen = 0;
+
+            /* find nulls in the string */
+            while(nulLen<srcLength && pSrc[nulLen++]!=0){
+            }
+
+            if((pSrc+nulLen) < pSrcLimit){
+                /* check if we have enough room in pCSrc */
+                if(remaining < (nulLen * MB_CUR_MAX)){
+                    /* should rarely occur */
+                    int32_t len = (pCSrc-pCSave);
+                    pCSrc = pCSave;
+                    /* we do not have enough room so grow the buffer*/
+                    u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
+                           _BUFFER_CAPACITY_MULTIPLIER*cStackCap+(nulLen*MB_CUR_MAX),len,sizeof(char));
+
+                    pCSave = pCSrc;
+                    pCSrc = pCSave+len;
+                    remaining = cStackCap-(pCSrc - pCSave);
+                }
+
+                /* we have found a null  so convert the 
+                 * chunk from begining of non-null char to null
+                 */
+                retVal = uprv_wcstombs(pCSrc,pSrc,remaining);
+
+                if(retVal==-1){
+                    /* an error occurred bail out */
+                    *pErrorCode = U_ILLEGAL_CHAR_FOUND;
+                    goto cleanup;
+                }
+
+                pCSrc += retVal+1 /* already null terminated */;
+
+                pSrc += nulLen; /* skip past the null */
+                srcLength-=nulLen; /* decrement the srcLength */
+                remaining -= (pCSrc-pCSave);
+
+
+            }else{
+                /* the source is not null terminated and we are 
+                 * end of source so we copy the source to a temp buffer
+                 * null terminate it and convert wchar_ts to chars
+                 */
+                if(nulLen >= _STACK_BUFFER_CAPACITY){
+                    /* Should rarely occcur */
+                    /* allocate new buffer buffer */
+                    pWStack =(wchar_t*) uprv_malloc(sizeof(wchar_t) * (nulLen + 1));
+                    if(pWStack==NULL){
+                        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
+                        goto cleanup;
+                    }
+                }
+                if(nulLen>0){
+                    /* copy the contents to tempStack */
+                    uprv_memcpy(pWStack, pSrc, (size_t)nulLen*sizeof(wchar_t));
+                }
+            
+                /* null terminate the tempBuffer */
+                pWStack[nulLen] =0 ;
+            
+                if(remaining < (nulLen * MB_CUR_MAX)){
+                    /* Should rarely occur */
+                    int32_t len = (pCSrc-pCSave);
+                    pCSrc = pCSave;
+                    /* we do not have enough room so grow the buffer*/
+                    u_growAnyBufferFromStatic(cStack,(void**)&pCSrc,&cStackCap,
+                           cStackCap+(nulLen*MB_CUR_MAX),len,sizeof(char));
+
+                    pCSave = pCSrc;
+                    pCSrc = pCSave+len;
+                    remaining = cStackCap-(pCSrc - pCSave);
+                }
+                /* convert to chars */
+                retVal = uprv_wcstombs(pCSrc,pWStack,remaining);
+            
+                pCSrc += retVal;
+                pSrc  += nulLen;
+                srcLength-=nulLen; /* decrement the srcLength */
+                break;
+            }
+        }
+    }
+
+    /* OK..now we have converted from wchar_ts to chars now 
+     * convert chars to UChars 
+     */
+    pCSrcLimit = pCSrc;
+    pCSrc = pCSave;
+    pTarget = target= dest;
+    pTargetLimit = dest + destCapacity;    
+    
+    conv= u_getDefaultConverter(pErrorCode);
+    
+    if(U_FAILURE(*pErrorCode)|| conv==NULL){
+        goto cleanup;
+    }
+    
+    for(;;) {
+        
+        *pErrorCode = U_ZERO_ERROR;
+        
+        /* convert to stack buffer*/
+        ucnv_toUnicode(conv,&pTarget,pTargetLimit,(const char**)&pCSrc,pCSrcLimit,NULL,(UBool)(pCSrc==pCSrcLimit),pErrorCode);
+        
+        /* increment count to number written to stack */
+        count+= pTarget - target;
+        
+        if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR){
+            target = uStack;
+            pTarget = uStack;
+            pTargetLimit = uStack + _STACK_BUFFER_CAPACITY;
+        } else {
+            break;
+        }
+        
+    }
+    
+    if(pDestLength){
+        *pDestLength =count;
+    }
+
+    u_terminateUChars(dest,destCapacity,count,pErrorCode);
+    
+cleanup:
+ 
+    if(cStack != pCSave){
+        uprv_free(pCSave);
+    }
+
+    if(wStack != pWStack){
+        uprv_free(pWStack);
+    }
+    
+    u_releaseDefaultConverter(conv);
+
+    return dest;
+}
+#endif
+
+U_CAPI UChar* U_EXPORT2
+u_strFromWCS(UChar   *dest,
+             int32_t destCapacity, 
+             int32_t *pDestLength,
+             const wchar_t *src,
+             int32_t srcLength,
+             UErrorCode *pErrorCode)
+{
+
+    /* args check */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+#ifdef U_WCHAR_IS_UTF16
+    /* wchar_t is UTF-16 just do a memcpy */
+    if(srcLength == -1){
+        srcLength = u_strlen((const UChar *)src);
+    }
+    if(0 < srcLength && srcLength <= destCapacity){
+        u_memcpy(dest, (const UChar *)src, srcLength);
+    }
+    if(pDestLength){
+       *pDestLength = srcLength;
+    }
+
+    u_terminateUChars(dest,destCapacity,srcLength,pErrorCode);
+
+    return dest;
+
+#elif defined U_WCHAR_IS_UTF32
+    
+    return u_strFromUTF32(dest, destCapacity, pDestLength,
+                          (UChar32*)src, srcLength, pErrorCode);
+
+#else
+
+    return _strFromWCS(dest,destCapacity,pDestLength,src,srcLength,pErrorCode);  
+
+#endif
+
+}
+
+#endif /* #if !defined(U_WCHAR_IS_UTF16) && !defined(U_WCHAR_IS_UTF32) && !UCONFIG_NO_CONVERSION */
diff --git a/thirdparty/icu4c/common/ustrcase.cpp b/thirdparty/icu4c/common/ustrcase.cpp
new file mode 100644
index 0000000000..618e847c65
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrcase.cpp
@@ -0,0 +1,1818 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2001-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  ustrcase.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2002feb20
+*   created by: Markus W. Scherer
+*
+*   Implementation file for string casing C API functions.
+*   Uses functions from uchar.c for basic functionality that requires access
+*   to the Unicode Character Database (uprops.dat).
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/brkiter.h"
+#include "unicode/casemap.h"
+#include "unicode/edits.h"
+#include "unicode/stringoptions.h"
+#include "unicode/ustring.h"
+#include "unicode/ucasemap.h"
+#include "unicode/ubrk.h"
+#include "unicode/utf.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "ucase.h"
+#include "ucasemap_imp.h"
+#include "ustr_imp.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+namespace {
+
+int32_t checkOverflowAndEditsError(int32_t destIndex, int32_t destCapacity,
+                                   Edits *edits, UErrorCode &errorCode) {
+    if (U_SUCCESS(errorCode)) {
+        if (destIndex > destCapacity) {
+            errorCode = U_BUFFER_OVERFLOW_ERROR;
+        } else if (edits != NULL) {
+            edits->copyErrorTo(errorCode);
+        }
+    }
+    return destIndex;
+}
+
+/* Appends a full case mapping result, see UCASE_MAX_STRING_LENGTH. */
+inline int32_t
+appendResult(UChar *dest, int32_t destIndex, int32_t destCapacity,
+             int32_t result, const UChar *s,
+             int32_t cpLength, uint32_t options, icu::Edits *edits) {
+    UChar32 c;
+    int32_t length;
+
+    /* decode the result */
+    if(result<0) {
+        /* (not) original code point */
+        if(edits!=NULL) {
+            edits->addUnchanged(cpLength);
+        }
+        if(options & U_OMIT_UNCHANGED_TEXT) {
+            return destIndex;
+        }
+        c=~result;
+        if(destIndex<destCapacity && c<=0xffff) {  // BMP slightly-fastpath
+            dest[destIndex++]=(UChar)c;
+            return destIndex;
+        }
+        length=cpLength;
+    } else {
+        if(result<=UCASE_MAX_STRING_LENGTH) {
+            c=U_SENTINEL;
+            length=result;
+        } else if(destIndex<destCapacity && result<=0xffff) {  // BMP slightly-fastpath
+            dest[destIndex++]=(UChar)result;
+            if(edits!=NULL) {
+                edits->addReplace(cpLength, 1);
+            }
+            return destIndex;
+        } else {
+            c=result;
+            length=U16_LENGTH(c);
+        }
+        if(edits!=NULL) {
+            edits->addReplace(cpLength, length);
+        }
+    }
+    if(length>(INT32_MAX-destIndex)) {
+        return -1;  // integer overflow
+    }
+
+    if(destIndex<destCapacity) {
+        /* append the result */
+        if(c>=0) {
+            /* code point */
+            UBool isError=FALSE;
+            U16_APPEND(dest, destIndex, destCapacity, c, isError);
+            if(isError) {
+                /* overflow, nothing written */
+                destIndex+=length;
+            }
+        } else {
+            /* string */
+            if((destIndex+length)<=destCapacity) {
+                while(length>0) {
+                    dest[destIndex++]=*s++;
+                    --length;
+                }
+            } else {
+                /* overflow */
+                destIndex+=length;
+            }
+        }
+    } else {
+        /* preflight */
+        destIndex+=length;
+    }
+    return destIndex;
+}
+
+inline int32_t
+appendUChar(UChar *dest, int32_t destIndex, int32_t destCapacity, UChar c) {
+    if(destIndex<destCapacity) {
+        dest[destIndex]=c;
+    } else if(destIndex==INT32_MAX) {
+        return -1;  // integer overflow
+    }
+    return destIndex+1;
+}
+
+int32_t
+appendNonEmptyUnchanged(UChar *dest, int32_t destIndex, int32_t destCapacity,
+                        const UChar *s, int32_t length, uint32_t options, icu::Edits *edits) {
+    if(edits!=NULL) {
+        edits->addUnchanged(length);
+    }
+    if(options & U_OMIT_UNCHANGED_TEXT) {
+        return destIndex;
+    }
+    if(length>(INT32_MAX-destIndex)) {
+        return -1;  // integer overflow
+    }
+    if((destIndex+length)<=destCapacity) {
+        u_memcpy(dest+destIndex, s, length);
+    }
+    return destIndex + length;
+}
+
+inline int32_t
+appendUnchanged(UChar *dest, int32_t destIndex, int32_t destCapacity,
+                const UChar *s, int32_t length, uint32_t options, icu::Edits *edits) {
+    if (length <= 0) {
+        return destIndex;
+    }
+    return appendNonEmptyUnchanged(dest, destIndex, destCapacity, s, length, options, edits);
+}
+
+UChar32 U_CALLCONV
+utf16_caseContextIterator(void *context, int8_t dir) {
+    UCaseContext *csc=(UCaseContext *)context;
+    UChar32 c;
+
+    if(dir<0) {
+        /* reset for backward iteration */
+        csc->index=csc->cpStart;
+        csc->dir=dir;
+    } else if(dir>0) {
+        /* reset for forward iteration */
+        csc->index=csc->cpLimit;
+        csc->dir=dir;
+    } else {
+        /* continue current iteration direction */
+        dir=csc->dir;
+    }
+
+    if(dir<0) {
+        if(csc->start<csc->index) {
+            U16_PREV((const UChar *)csc->p, csc->start, csc->index, c);
+            return c;
+        }
+    } else {
+        if(csc->index<csc->limit) {
+            U16_NEXT((const UChar *)csc->p, csc->index, csc->limit, c);
+            return c;
+        }
+    }
+    return U_SENTINEL;
+}
+
+/**
+ * caseLocale >= 0: Lowercases [srcStart..srcLimit[ but takes context [0..srcLength[ into account.
+ * caseLocale < 0: Case-folds [srcStart..srcLimit[.
+ */
+int32_t toLower(int32_t caseLocale, uint32_t options,
+                UChar *dest, int32_t destCapacity,
+                const UChar *src, UCaseContext *csc, int32_t srcStart, int32_t srcLimit,
+                icu::Edits *edits, UErrorCode &errorCode) {
+    const int8_t *latinToLower;
+    if (caseLocale == UCASE_LOC_ROOT ||
+            (caseLocale >= 0 ?
+                !(caseLocale == UCASE_LOC_TURKISH || caseLocale == UCASE_LOC_LITHUANIAN) :
+                (options & _FOLD_CASE_OPTIONS_MASK) == U_FOLD_CASE_DEFAULT)) {
+        latinToLower = LatinCase::TO_LOWER_NORMAL;
+    } else {
+        latinToLower = LatinCase::TO_LOWER_TR_LT;
+    }
+    const UTrie2 *trie = ucase_getTrie();
+    int32_t destIndex = 0;
+    int32_t prev = srcStart;
+    int32_t srcIndex = srcStart;
+    for (;;) {
+        // fast path for simple cases
+        UChar lead = 0;
+        while (srcIndex < srcLimit) {
+            lead = src[srcIndex];
+            int32_t delta;
+            if (lead < LatinCase::LONG_S) {
+                int8_t d = latinToLower[lead];
+                if (d == LatinCase::EXC) { break; }
+                ++srcIndex;
+                if (d == 0) { continue; }
+                delta = d;
+            } else if (lead >= 0xd800) {
+                break;  // surrogate or higher
+            } else {
+                uint16_t props = UTRIE2_GET16_FROM_U16_SINGLE_LEAD(trie, lead);
+                if (UCASE_HAS_EXCEPTION(props)) { break; }
+                ++srcIndex;
+                if (!UCASE_IS_UPPER_OR_TITLE(props) || (delta = UCASE_GET_DELTA(props)) == 0) {
+                    continue;
+                }
+            }
+            lead += static_cast<UChar>(delta);
+            destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                        src + prev, srcIndex - 1 - prev, options, edits);
+            if (destIndex >= 0) {
+                destIndex = appendUChar(dest, destIndex, destCapacity, lead);
+                if (edits != nullptr) {
+                    edits->addReplace(1, 1);
+                }
+            }
+            if (destIndex < 0) {
+                errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+            prev = srcIndex;
+        }
+        if (srcIndex >= srcLimit) {
+            break;
+        }
+        // slow path
+        int32_t cpStart = srcIndex++;
+        UChar trail;
+        UChar32 c;
+        if (U16_IS_LEAD(lead) && srcIndex < srcLimit && U16_IS_TRAIL(trail = src[srcIndex])) {
+            c = U16_GET_SUPPLEMENTARY(lead, trail);
+            ++srcIndex;
+        } else {
+            c = lead;
+        }
+        const UChar *s;
+        if (caseLocale >= 0) {
+            csc->cpStart = cpStart;
+            csc->cpLimit = srcIndex;
+            c = ucase_toFullLower(c, utf16_caseContextIterator, csc, &s, caseLocale);
+        } else {
+            c = ucase_toFullFolding(c, &s, options);
+        }
+        if (c >= 0) {
+            destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                        src + prev, cpStart - prev, options, edits);
+            if (destIndex >= 0) {
+                destIndex = appendResult(dest, destIndex, destCapacity, c, s,
+                                         srcIndex - cpStart, options, edits);
+            }
+            if (destIndex < 0) {
+                errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+            prev = srcIndex;
+        }
+    }
+    destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                src + prev, srcIndex - prev, options, edits);
+    if (destIndex < 0) {
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    return destIndex;
+}
+
+int32_t toUpper(int32_t caseLocale, uint32_t options,
+                UChar *dest, int32_t destCapacity,
+                const UChar *src, UCaseContext *csc, int32_t srcLength,
+                icu::Edits *edits, UErrorCode &errorCode) {
+    const int8_t *latinToUpper;
+    if (caseLocale == UCASE_LOC_TURKISH) {
+        latinToUpper = LatinCase::TO_UPPER_TR;
+    } else {
+        latinToUpper = LatinCase::TO_UPPER_NORMAL;
+    }
+    const UTrie2 *trie = ucase_getTrie();
+    int32_t destIndex = 0;
+    int32_t prev = 0;
+    int32_t srcIndex = 0;
+    for (;;) {
+        // fast path for simple cases
+        UChar lead = 0;
+        while (srcIndex < srcLength) {
+            lead = src[srcIndex];
+            int32_t delta;
+            if (lead < LatinCase::LONG_S) {
+                int8_t d = latinToUpper[lead];
+                if (d == LatinCase::EXC) { break; }
+                ++srcIndex;
+                if (d == 0) { continue; }
+                delta = d;
+            } else if (lead >= 0xd800) {
+                break;  // surrogate or higher
+            } else {
+                uint16_t props = UTRIE2_GET16_FROM_U16_SINGLE_LEAD(trie, lead);
+                if (UCASE_HAS_EXCEPTION(props)) { break; }
+                ++srcIndex;
+                if (UCASE_GET_TYPE(props) != UCASE_LOWER || (delta = UCASE_GET_DELTA(props)) == 0) {
+                    continue;
+                }
+            }
+            lead += static_cast<UChar>(delta);
+            destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                        src + prev, srcIndex - 1 - prev, options, edits);
+            if (destIndex >= 0) {
+                destIndex = appendUChar(dest, destIndex, destCapacity, lead);
+                if (edits != nullptr) {
+                    edits->addReplace(1, 1);
+                }
+            }
+            if (destIndex < 0) {
+                errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+            prev = srcIndex;
+        }
+        if (srcIndex >= srcLength) {
+            break;
+        }
+        // slow path
+        int32_t cpStart;
+        csc->cpStart = cpStart = srcIndex++;
+        UChar trail;
+        UChar32 c;
+        if (U16_IS_LEAD(lead) && srcIndex < srcLength && U16_IS_TRAIL(trail = src[srcIndex])) {
+            c = U16_GET_SUPPLEMENTARY(lead, trail);
+            ++srcIndex;
+        } else {
+            c = lead;
+        }
+        csc->cpLimit = srcIndex;
+        const UChar *s;
+        c = ucase_toFullUpper(c, utf16_caseContextIterator, csc, &s, caseLocale);
+        if (c >= 0) {
+            destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                        src + prev, cpStart - prev, options, edits);
+            if (destIndex >= 0) {
+                destIndex = appendResult(dest, destIndex, destCapacity, c, s,
+                                         srcIndex - cpStart, options, edits);
+            }
+            if (destIndex < 0) {
+                errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+            prev = srcIndex;
+        }
+    }
+    destIndex = appendUnchanged(dest, destIndex, destCapacity,
+                                src + prev, srcIndex - prev, options, edits);
+    if (destIndex < 0) {
+        errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    return destIndex;
+}
+
+}  // namespace
+
+U_NAMESPACE_END
+
+U_NAMESPACE_USE
+
+#if !UCONFIG_NO_BREAK_ITERATION
+
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToTitle(int32_t caseLocale, uint32_t options, BreakIterator *iter,
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode) {
+    if (!ustrcase_checkTitleAdjustmentOptions(options, errorCode)) {
+        return 0;
+    }
+
+    /* set up local variables */
+    UCaseContext csc=UCASECONTEXT_INITIALIZER;
+    csc.p=(void *)src;
+    csc.limit=srcLength;
+    int32_t destIndex=0;
+    int32_t prev=0;
+    UBool isFirstIndex=TRUE;
+
+    /* titlecasing loop */
+    while(prev<srcLength) {
+        /* find next index where to titlecase */
+        int32_t index;
+        if(isFirstIndex) {
+            isFirstIndex=FALSE;
+            index=iter->first();
+        } else {
+            index=iter->next();
+        }
+        if(index==UBRK_DONE || index>srcLength) {
+            index=srcLength;
+        }
+
+        /*
+         * Segment [prev..index[ into 3 parts:
+         * a) skipped characters (copy as-is) [prev..titleStart[
+         * b) first letter (titlecase)              [titleStart..titleLimit[
+         * c) subsequent characters (lowercase)                 [titleLimit..index[
+         */
+        if(prev<index) {
+            // Find and copy skipped characters [prev..titleStart[
+            int32_t titleStart=prev;
+            int32_t titleLimit=prev;
+            UChar32 c;
+            U16_NEXT(src, titleLimit, index, c);
+            if ((options&U_TITLECASE_NO_BREAK_ADJUSTMENT)==0) {
+                // Adjust the titlecasing index to the next cased character,
+                // or to the next letter/number/symbol/private use.
+                // Stop with titleStart<titleLimit<=index
+                // if there is a character to be titlecased,
+                // or else stop with titleStart==titleLimit==index.
+                UBool toCased = (options&U_TITLECASE_ADJUST_TO_CASED) != 0;
+                while (toCased ? UCASE_NONE==ucase_getType(c) : !ustrcase_isLNS(c)) {
+                    titleStart=titleLimit;
+                    if(titleLimit==index) {
+                        break;
+                    }
+                    U16_NEXT(src, titleLimit, index, c);
+                }
+                if (prev < titleStart) {
+                    destIndex=appendUnchanged(dest, destIndex, destCapacity,
+                                              src+prev, titleStart-prev, options, edits);
+                    if(destIndex<0) {
+                        errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                        return 0;
+                    }
+                }
+            }
+
+            if(titleStart<titleLimit) {
+                /* titlecase c which is from [titleStart..titleLimit[ */
+                csc.cpStart=titleStart;
+                csc.cpLimit=titleLimit;
+                const UChar *s;
+                c=ucase_toFullTitle(c, utf16_caseContextIterator, &csc, &s, caseLocale);
+                destIndex=appendResult(dest, destIndex, destCapacity, c, s,
+                                       titleLimit-titleStart, options, edits);
+                if(destIndex<0) {
+                    errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                    return 0;
+                }
+
+                /* Special case Dutch IJ titlecasing */
+                if (titleStart+1 < index &&
+                        caseLocale == UCASE_LOC_DUTCH &&
+                        (src[titleStart] == 0x0049 || src[titleStart] == 0x0069)) {
+                    if (src[titleStart+1] == 0x006A) {
+                        destIndex=appendUChar(dest, destIndex, destCapacity, 0x004A);
+                        if(destIndex<0) {
+                            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                            return 0;
+                        }
+                        if(edits!=NULL) {
+                            edits->addReplace(1, 1);
+                        }
+                        titleLimit++;
+                    } else if (src[titleStart+1] == 0x004A) {
+                        // Keep the capital J from getting lowercased.
+                        destIndex=appendUnchanged(dest, destIndex, destCapacity,
+                                                  src+titleStart+1, 1, options, edits);
+                        if(destIndex<0) {
+                            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                            return 0;
+                        }
+                        titleLimit++;
+                    }
+                }
+
+                /* lowercase [titleLimit..index[ */
+                if(titleLimit<index) {
+                    if((options&U_TITLECASE_NO_LOWERCASE)==0) {
+                        /* Normal operation: Lowercase the rest of the word. */
+                        destIndex+=
+                            toLower(
+                                caseLocale, options,
+                                dest+destIndex, destCapacity-destIndex,
+                                src, &csc, titleLimit, index,
+                                edits, errorCode);
+                        if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+                            errorCode=U_ZERO_ERROR;
+                        }
+                        if(U_FAILURE(errorCode)) {
+                            return destIndex;
+                        }
+                    } else {
+                        /* Optionally just copy the rest of the word unchanged. */
+                        destIndex=appendUnchanged(dest, destIndex, destCapacity,
+                                                  src+titleLimit, index-titleLimit, options, edits);
+                        if(destIndex<0) {
+                            errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                            return 0;
+                        }
+                    }
+                }
+            }
+        }
+
+        prev=index;
+    }
+
+    return checkOverflowAndEditsError(destIndex, destCapacity, edits, errorCode);
+}
+
+#endif  // !UCONFIG_NO_BREAK_ITERATION
+
+U_NAMESPACE_BEGIN
+namespace GreekUpper {
+
+// Data generated by prototype code, see
+// http://site.icu-project.org/design/case/greek-upper
+// TODO: Move this data into ucase.icu.
+static const uint16_t data0370[] = {
+    // U+0370..03FF
+    0x0370,
+    0x0370,
+    0x0372,
+    0x0372,
+    0,
+    0,
+    0x0376,
+    0x0376,
+    0,
+    0,
+    0x037A,
+    0x03FD,
+    0x03FE,
+    0x03FF,
+    0,
+    0x037F,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x0391 | HAS_VOWEL,
+    0x0392,
+    0x0393,
+    0x0394,
+    0x0395 | HAS_VOWEL,
+    0x0396,
+    0x0397 | HAS_VOWEL,
+    0x0398,
+    0x0399 | HAS_VOWEL,
+    0x039A,
+    0x039B,
+    0x039C,
+    0x039D,
+    0x039E,
+    0x039F | HAS_VOWEL,
+    0x03A0,
+    0x03A1,
+    0,
+    0x03A3,
+    0x03A4,
+    0x03A5 | HAS_VOWEL,
+    0x03A6,
+    0x03A7,
+    0x03A8,
+    0x03A9 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_DIALYTIKA,
+    0x03A5 | HAS_VOWEL | HAS_DIALYTIKA,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x0391 | HAS_VOWEL,
+    0x0392,
+    0x0393,
+    0x0394,
+    0x0395 | HAS_VOWEL,
+    0x0396,
+    0x0397 | HAS_VOWEL,
+    0x0398,
+    0x0399 | HAS_VOWEL,
+    0x039A,
+    0x039B,
+    0x039C,
+    0x039D,
+    0x039E,
+    0x039F | HAS_VOWEL,
+    0x03A0,
+    0x03A1,
+    0x03A3,
+    0x03A3,
+    0x03A4,
+    0x03A5 | HAS_VOWEL,
+    0x03A6,
+    0x03A7,
+    0x03A8,
+    0x03A9 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_DIALYTIKA,
+    0x03A5 | HAS_VOWEL | HAS_DIALYTIKA,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03CF,
+    0x0392,
+    0x0398,
+    0x03D2,
+    0x03D2 | HAS_ACCENT,
+    0x03D2 | HAS_DIALYTIKA,
+    0x03A6,
+    0x03A0,
+    0x03CF,
+    0x03D8,
+    0x03D8,
+    0x03DA,
+    0x03DA,
+    0x03DC,
+    0x03DC,
+    0x03DE,
+    0x03DE,
+    0x03E0,
+    0x03E0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0x039A,
+    0x03A1,
+    0x03F9,
+    0x037F,
+    0x03F4,
+    0x0395 | HAS_VOWEL,
+    0,
+    0x03F7,
+    0x03F7,
+    0x03F9,
+    0x03FA,
+    0x03FA,
+    0x03FC,
+    0x03FD,
+    0x03FE,
+    0x03FF,
+};
+
+static const uint16_t data1F00[] = {
+    // U+1F00..1FFF
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL,
+    0x0395 | HAS_VOWEL,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0x0395 | HAS_VOWEL,
+    0x0395 | HAS_VOWEL,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0x0397 | HAS_VOWEL,
+    0x0397 | HAS_VOWEL,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL,
+    0x0397 | HAS_VOWEL,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL,
+    0x039F | HAS_VOWEL,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0x039F | HAS_VOWEL,
+    0x039F | HAS_VOWEL,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x03A5 | HAS_VOWEL,
+    0,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL,
+    0x03A9 | HAS_VOWEL,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL,
+    0x03A9 | HAS_VOWEL,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_ACCENT,
+    0x0391 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0,
+    0x0399 | HAS_VOWEL,
+    0,
+    0,
+    0,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0395 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_ACCENT,
+    0x0397 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0,
+    0,
+    0,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x0399 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0,
+    0,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0x0399 | HAS_VOWEL | HAS_ACCENT,
+    0,
+    0,
+    0,
+    0,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x03A1,
+    0x03A1,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT | HAS_DIALYTIKA,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A5 | HAS_VOWEL | HAS_ACCENT,
+    0x03A1,
+    0,
+    0,
+    0,
+    0,
+    0,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x039F | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_ACCENT,
+    0x03A9 | HAS_VOWEL | HAS_YPOGEGRAMMENI,
+    0,
+    0,
+    0,
+};
+
+// U+2126 Ohm sign
+static const uint16_t data2126 = 0x03A9 | HAS_VOWEL;
+
+uint32_t getLetterData(UChar32 c) {
+    if (c < 0x370 || 0x2126 < c || (0x3ff < c && c < 0x1f00)) {
+        return 0;
+    } else if (c <= 0x3ff) {
+        return data0370[c - 0x370];
+    } else if (c <= 0x1fff) {
+        return data1F00[c - 0x1f00];
+    } else if (c == 0x2126) {
+        return data2126;
+    } else {
+        return 0;
+    }
+}
+
+uint32_t getDiacriticData(UChar32 c) {
+    switch (c) {
+    case 0x0300:  // varia
+    case 0x0301:  // tonos = oxia
+    case 0x0342:  // perispomeni
+    case 0x0302:  // circumflex can look like perispomeni
+    case 0x0303:  // tilde can look like perispomeni
+    case 0x0311:  // inverted breve can look like perispomeni
+        return HAS_ACCENT;
+    case 0x0308:  // dialytika = diaeresis
+        return HAS_COMBINING_DIALYTIKA;
+    case 0x0344:  // dialytika tonos
+        return HAS_COMBINING_DIALYTIKA | HAS_ACCENT;
+    case 0x0345:  // ypogegrammeni = iota subscript
+        return HAS_YPOGEGRAMMENI;
+    case 0x0304:  // macron
+    case 0x0306:  // breve
+    case 0x0313:  // comma above
+    case 0x0314:  // reversed comma above
+    case 0x0343:  // koronis
+        return HAS_OTHER_GREEK_DIACRITIC;
+    default:
+        return 0;
+    }
+}
+
+UBool isFollowedByCasedLetter(const UChar *s, int32_t i, int32_t length) {
+    while (i < length) {
+        UChar32 c;
+        U16_NEXT(s, i, length, c);
+        int32_t type = ucase_getTypeOrIgnorable(c);
+        if ((type & UCASE_IGNORABLE) != 0) {
+            // Case-ignorable, continue with the loop.
+        } else if (type != UCASE_NONE) {
+            return TRUE;  // Followed by cased letter.
+        } else {
+            return FALSE;  // Uncased and not case-ignorable.
+        }
+    }
+    return FALSE;  // Not followed by cased letter.
+}
+
+/**
+ * Greek string uppercasing with a state machine.
+ * Probably simpler than a stateless function that has to figure out complex context-before
+ * for each character.
+ * TODO: Try to re-consolidate one way or another with the non-Greek function.
+ */
+int32_t toUpper(uint32_t options,
+                UChar *dest, int32_t destCapacity,
+                const UChar *src, int32_t srcLength,
+                Edits *edits,
+                UErrorCode &errorCode) {
+    int32_t destIndex=0;
+    uint32_t state = 0;
+    for (int32_t i = 0; i < srcLength;) {
+        int32_t nextIndex = i;
+        UChar32 c;
+        U16_NEXT(src, nextIndex, srcLength, c);
+        uint32_t nextState = 0;
+        int32_t type = ucase_getTypeOrIgnorable(c);
+        if ((type & UCASE_IGNORABLE) != 0) {
+            // c is case-ignorable
+            nextState |= (state & AFTER_CASED);
+        } else if (type != UCASE_NONE) {
+            // c is cased
+            nextState |= AFTER_CASED;
+        }
+        uint32_t data = getLetterData(c);
+        if (data > 0) {
+            uint32_t upper = data & UPPER_MASK;
+            // Add a dialytika to this iota or ypsilon vowel
+            // if we removed a tonos from the previous vowel,
+            // and that previous vowel did not also have (or gain) a dialytika.
+            // Adding one only to the final vowel in a longer sequence
+            // (which does not occur in normal writing) would require lookahead.
+            // Set the same flag as for preserving an existing dialytika.
+            if ((data & HAS_VOWEL) != 0 && (state & AFTER_VOWEL_WITH_ACCENT) != 0 &&
+                    (upper == 0x399 || upper == 0x3A5)) {
+                data |= HAS_DIALYTIKA;
+            }
+            int32_t numYpogegrammeni = 0;  // Map each one to a trailing, spacing, capital iota.
+            if ((data & HAS_YPOGEGRAMMENI) != 0) {
+                numYpogegrammeni = 1;
+            }
+            // Skip combining diacritics after this Greek letter.
+            while (nextIndex < srcLength) {
+                uint32_t diacriticData = getDiacriticData(src[nextIndex]);
+                if (diacriticData != 0) {
+                    data |= diacriticData;
+                    if ((diacriticData & HAS_YPOGEGRAMMENI) != 0) {
+                        ++numYpogegrammeni;
+                    }
+                    ++nextIndex;
+                } else {
+                    break;  // not a Greek diacritic
+                }
+            }
+            if ((data & HAS_VOWEL_AND_ACCENT_AND_DIALYTIKA) == HAS_VOWEL_AND_ACCENT) {
+                nextState |= AFTER_VOWEL_WITH_ACCENT;
+            }
+            // Map according to Greek rules.
+            UBool addTonos = FALSE;
+            if (upper == 0x397 &&
+                    (data & HAS_ACCENT) != 0 &&
+                    numYpogegrammeni == 0 &&
+                    (state & AFTER_CASED) == 0 &&
+                    !isFollowedByCasedLetter(src, nextIndex, srcLength)) {
+                // Keep disjunctive "or" with (only) a tonos.
+                // We use the same "word boundary" conditions as for the Final_Sigma test.
+                if (i == nextIndex) {
+                    upper = 0x389;  // Preserve the precomposed form.
+                } else {
+                    addTonos = TRUE;
+                }
+            } else if ((data & HAS_DIALYTIKA) != 0) {
+                // Preserve a vowel with dialytika in precomposed form if it exists.
+                if (upper == 0x399) {
+                    upper = 0x3AA;
+                    data &= ~HAS_EITHER_DIALYTIKA;
+                } else if (upper == 0x3A5) {
+                    upper = 0x3AB;
+                    data &= ~HAS_EITHER_DIALYTIKA;
+                }
+            }
+
+            UBool change;
+            if (edits == nullptr && (options & U_OMIT_UNCHANGED_TEXT) == 0) {
+                change = TRUE;  // common, simple usage
+            } else {
+                // Find out first whether we are changing the text.
+                change = src[i] != upper || numYpogegrammeni > 0;
+                int32_t i2 = i + 1;
+                if ((data & HAS_EITHER_DIALYTIKA) != 0) {
+                    change |= i2 >= nextIndex || src[i2] != 0x308;
+                    ++i2;
+                }
+                if (addTonos) {
+                    change |= i2 >= nextIndex || src[i2] != 0x301;
+                    ++i2;
+                }
+                int32_t oldLength = nextIndex - i;
+                int32_t newLength = (i2 - i) + numYpogegrammeni;
+                change |= oldLength != newLength;
+                if (change) {
+                    if (edits != NULL) {
+                        edits->addReplace(oldLength, newLength);
+                    }
+                } else {
+                    if (edits != NULL) {
+                        edits->addUnchanged(oldLength);
+                    }
+                    // Write unchanged text?
+                    change = (options & U_OMIT_UNCHANGED_TEXT) == 0;
+                }
+            }
+
+            if (change) {
+                destIndex=appendUChar(dest, destIndex, destCapacity, (UChar)upper);
+                if (destIndex >= 0 && (data & HAS_EITHER_DIALYTIKA) != 0) {
+                    destIndex=appendUChar(dest, destIndex, destCapacity, 0x308);  // restore or add a dialytika
+                }
+                if (destIndex >= 0 && addTonos) {
+                    destIndex=appendUChar(dest, destIndex, destCapacity, 0x301);
+                }
+                while (destIndex >= 0 && numYpogegrammeni > 0) {
+                    destIndex=appendUChar(dest, destIndex, destCapacity, 0x399);
+                    --numYpogegrammeni;
+                }
+                if(destIndex<0) {
+                    errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+                    return 0;
+                }
+            }
+        } else {
+            const UChar *s;
+            c=ucase_toFullUpper(c, NULL, NULL, &s, UCASE_LOC_GREEK);
+            destIndex = appendResult(dest, destIndex, destCapacity, c, s,
+                                     nextIndex - i, options, edits);
+            if (destIndex < 0) {
+                errorCode = U_INDEX_OUTOFBOUNDS_ERROR;
+                return 0;
+            }
+        }
+        i = nextIndex;
+        state = nextState;
+    }
+
+    return destIndex;
+}
+
+}  // namespace GreekUpper
+U_NAMESPACE_END
+
+/* functions available in the common library (for unistr_case.cpp) */
+
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToLower(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode) {
+    UCaseContext csc=UCASECONTEXT_INITIALIZER;
+    csc.p=(void *)src;
+    csc.limit=srcLength;
+    int32_t destIndex = toLower(
+        caseLocale, options,
+        dest, destCapacity,
+        src, &csc, 0, srcLength,
+        edits, errorCode);
+    return checkOverflowAndEditsError(destIndex, destCapacity, edits, errorCode);
+}
+
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalToUpper(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                         UChar *dest, int32_t destCapacity,
+                         const UChar *src, int32_t srcLength,
+                         icu::Edits *edits,
+                         UErrorCode &errorCode) {
+    int32_t destIndex;
+    if (caseLocale == UCASE_LOC_GREEK) {
+        destIndex = GreekUpper::toUpper(options, dest, destCapacity,
+                                        src, srcLength, edits, errorCode);
+    } else {
+        UCaseContext csc=UCASECONTEXT_INITIALIZER;
+        csc.p=(void *)src;
+        csc.limit=srcLength;
+        destIndex = toUpper(
+            caseLocale, options,
+            dest, destCapacity,
+            src, &csc, srcLength,
+            edits, errorCode);
+    }
+    return checkOverflowAndEditsError(destIndex, destCapacity, edits, errorCode);
+}
+
+U_CFUNC int32_t U_CALLCONV
+ustrcase_internalFold(int32_t /* caseLocale */, uint32_t options, UCASEMAP_BREAK_ITERATOR_UNUSED
+                      UChar *dest, int32_t destCapacity,
+                      const UChar *src, int32_t srcLength,
+                      icu::Edits *edits,
+                      UErrorCode &errorCode) {
+    int32_t destIndex = toLower(
+        -1, options,
+        dest, destCapacity,
+        src, nullptr, 0, srcLength,
+        edits, errorCode);
+    return checkOverflowAndEditsError(destIndex, destCapacity, edits, errorCode);
+}
+
+U_CFUNC int32_t
+ustrcase_map(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+             UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             UStringCaseMapper *stringCaseMapper,
+             icu::Edits *edits,
+             UErrorCode &errorCode) {
+    int32_t destLength;
+
+    /* check argument values */
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if( destCapacity<0 ||
+        (dest==NULL && destCapacity>0) ||
+        src==NULL ||
+        srcLength<-1
+    ) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* get the string length */
+    if(srcLength==-1) {
+        srcLength=u_strlen(src);
+    }
+
+    /* check for overlapping source and destination */
+    if( dest!=NULL &&
+        ((src>=dest && src<(dest+destCapacity)) ||
+         (dest>=src && dest<(src+srcLength)))
+    ) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if (edits != nullptr && (options & U_EDITS_NO_RESET) == 0) {
+        edits->reset();
+    }
+    destLength=stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                                dest, destCapacity, src, srcLength, edits, errorCode);
+    return u_terminateUChars(dest, destCapacity, destLength, &errorCode);
+}
+
+U_CFUNC int32_t
+ustrcase_mapWithOverlap(int32_t caseLocale, uint32_t options, UCASEMAP_BREAK_ITERATOR_PARAM
+                        UChar *dest, int32_t destCapacity,
+                        const UChar *src, int32_t srcLength,
+                        UStringCaseMapper *stringCaseMapper,
+                        UErrorCode &errorCode) {
+    UChar buffer[300];
+    UChar *temp;
+
+    int32_t destLength;
+
+    /* check argument values */
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if( destCapacity<0 ||
+        (dest==NULL && destCapacity>0) ||
+        src==NULL ||
+        srcLength<-1
+    ) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* get the string length */
+    if(srcLength==-1) {
+        srcLength=u_strlen(src);
+    }
+
+    /* check for overlapping source and destination */
+    if( dest!=NULL &&
+        ((src>=dest && src<(dest+destCapacity)) ||
+         (dest>=src && dest<(src+srcLength)))
+    ) {
+        /* overlap: provide a temporary destination buffer and later copy the result */
+        if(destCapacity<=UPRV_LENGTHOF(buffer)) {
+            /* the stack buffer is large enough */
+            temp=buffer;
+        } else {
+            /* allocate a buffer */
+            temp=(UChar *)uprv_malloc(destCapacity*U_SIZEOF_UCHAR);
+            if(temp==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+                return 0;
+            }
+        }
+    } else {
+        temp=dest;
+    }
+
+    destLength=stringCaseMapper(caseLocale, options, UCASEMAP_BREAK_ITERATOR
+                                temp, destCapacity, src, srcLength, NULL, errorCode);
+    if(temp!=dest) {
+        /* copy the result string to the destination buffer */
+        if (U_SUCCESS(errorCode) && 0 < destLength && destLength <= destCapacity) {
+            u_memmove(dest, temp, destLength);
+        }
+        if(temp!=buffer) {
+            uprv_free(temp);
+        }
+    }
+
+    return u_terminateUChars(dest, destCapacity, destLength, &errorCode);
+}
+
+/* public API functions */
+
+U_CAPI int32_t U_EXPORT2
+u_strFoldCase(UChar *dest, int32_t destCapacity,
+              const UChar *src, int32_t srcLength,
+              uint32_t options,
+              UErrorCode *pErrorCode) {
+    return ustrcase_mapWithOverlap(
+        UCASE_LOC_ROOT, options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalFold, *pErrorCode);
+}
+
+U_NAMESPACE_BEGIN
+
+int32_t CaseMap::fold(
+        uint32_t options,
+        const UChar *src, int32_t srcLength,
+        UChar *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ustrcase_map(
+        UCASE_LOC_ROOT, options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalFold, edits, errorCode);
+}
+
+U_NAMESPACE_END
+
+/* case-insensitive string comparisons -------------------------------------- */
+
+/*
+ * This function is a copy of unorm_cmpEquivFold() minus the parts for
+ * canonical equivalence.
+ * Keep the functions in sync, and see there for how this works.
+ * The duplication is for modularization:
+ * It makes caseless (but not canonical caseless) matches independent of
+ * the normalization code.
+ */
+
+/* stack element for previous-level source/decomposition pointers */
+struct CmpEquivLevel {
+    const UChar *start, *s, *limit;
+};
+typedef struct CmpEquivLevel CmpEquivLevel;
+
+/**
+ * Internal implementation code comparing string with case fold.
+ * This function is called from u_strcmpFold() and u_caseInsensitivePrefixMatch().
+ *
+ * @param s1            input string 1
+ * @param length1       length of string 1, or -1 (NULL terminated)
+ * @param s2            input string 2
+ * @param length2       length of string 2, or -1 (NULL terminated)
+ * @param options       compare options
+ * @param matchLen1     (output) length of partial prefix match in s1
+ * @param matchLen2     (output) length of partial prefix match in s2
+ * @param pErrorCode    receives error status
+ * @return The result of comparison
+ */
+static int32_t _cmpFold(
+            const UChar *s1, int32_t length1,
+            const UChar *s2, int32_t length2,
+            uint32_t options,
+            int32_t *matchLen1, int32_t *matchLen2,
+            UErrorCode *pErrorCode) {
+    int32_t cmpRes = 0;
+
+    /* current-level start/limit - s1/s2 as current */
+    const UChar *start1, *start2, *limit1, *limit2;
+
+    /* points to the original start address */
+    const UChar *org1, *org2;
+
+    /* points to the end of match + 1 */
+    const UChar *m1, *m2;
+
+    /* case folding variables */
+    const UChar *p;
+    int32_t length;
+
+    /* stacks of previous-level start/current/limit */
+    CmpEquivLevel stack1[2], stack2[2];
+
+    /* case folding buffers, only use current-level start/limit */
+    UChar fold1[UCASE_MAX_STRING_LENGTH+1], fold2[UCASE_MAX_STRING_LENGTH+1];
+
+    /* track which is the current level per string */
+    int32_t level1, level2;
+
+    /* current code units, and code points for lookups */
+    UChar32 c1, c2, cp1, cp2;
+
+    /* no argument error checking because this itself is not an API */
+
+    /*
+     * assume that at least the option U_COMPARE_IGNORE_CASE is set
+     * otherwise this function would have to behave exactly as uprv_strCompare()
+     */
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    /* initialize */
+    if(matchLen1) {
+        U_ASSERT(matchLen2 !=NULL);
+        *matchLen1=0;
+        *matchLen2=0;
+    }
+
+    start1=m1=org1=s1;
+    if(length1==-1) {
+        limit1=NULL;
+    } else {
+        limit1=s1+length1;
+    }
+
+    start2=m2=org2=s2;
+    if(length2==-1) {
+        limit2=NULL;
+    } else {
+        limit2=s2+length2;
+    }
+
+    level1=level2=0;
+    c1=c2=-1;
+
+    /* comparison loop */
+    for(;;) {
+        /*
+         * here a code unit value of -1 means "get another code unit"
+         * below it will mean "this source is finished"
+         */
+
+        if(c1<0) {
+            /* get next code unit from string 1, post-increment */
+            for(;;) {
+                if(s1==limit1 || ((c1=*s1)==0 && (limit1==NULL || (options&_STRNCMP_STYLE)))) {
+                    if(level1==0) {
+                        c1=-1;
+                        break;
+                    }
+                } else {
+                    ++s1;
+                    break;
+                }
+
+                /* reached end of level buffer, pop one level */
+                do {
+                    --level1;
+                    start1=stack1[level1].start;    /*Not uninitialized*/
+                } while(start1==NULL);
+                s1=stack1[level1].s;                /*Not uninitialized*/
+                limit1=stack1[level1].limit;        /*Not uninitialized*/
+            }
+        }
+
+        if(c2<0) {
+            /* get next code unit from string 2, post-increment */
+            for(;;) {
+                if(s2==limit2 || ((c2=*s2)==0 && (limit2==NULL || (options&_STRNCMP_STYLE)))) {
+                    if(level2==0) {
+                        c2=-1;
+                        break;
+                    }
+                } else {
+                    ++s2;
+                    break;
+                }
+
+                /* reached end of level buffer, pop one level */
+                do {
+                    --level2;
+                    start2=stack2[level2].start;    /*Not uninitialized*/
+                } while(start2==NULL);
+                s2=stack2[level2].s;                /*Not uninitialized*/
+                limit2=stack2[level2].limit;        /*Not uninitialized*/
+            }
+        }
+
+        /*
+         * compare c1 and c2
+         * either variable c1, c2 is -1 only if the corresponding string is finished
+         */
+        if(c1==c2) {
+            const UChar *next1, *next2;
+
+            if(c1<0) {
+                cmpRes=0;   /* c1==c2==-1 indicating end of strings */
+                break;
+            }
+
+            /*
+             * Note: Move the match positions in both strings at the same time
+             *      only when corresponding code point(s) in the original strings
+             *      are fully consumed. For example, when comparing s1="Fust" and
+             *      s2="Fu\u00dfball", s2[2] is folded into "ss", and s1[2] matches
+             *      the first code point in the case-folded data. But the second "s"
+             *      has no matching code point in s1, so this implementation returns
+             *      2 as the prefix match length ("Fu").
+             */
+            next1=next2=NULL;
+            if(level1==0) {
+                next1=s1;
+            } else if(s1==limit1) {
+                /* Note: This implementation only use a single level of stack.
+                 *      If this code needs to be changed to use multiple levels
+                 *      of stacks, the code above should check if the current
+                 *      code is at the end of all stacks.
+                 */
+                U_ASSERT(level1==1);
+
+                /* is s1 at the end of the current stack? */
+                next1=stack1[0].s;
+            }
+
+            if (next1!=NULL) {
+                if(level2==0) {
+                    next2=s2;
+                } else if(s2==limit2) {
+                    U_ASSERT(level2==1);
+
+                    /* is s2 at the end of the current stack? */
+                    next2=stack2[0].s;
+                }
+                if(next2!=NULL) {
+                    m1=next1;
+                    m2=next2;
+                }
+            }
+            c1=c2=-1;       /* make us fetch new code units */
+            continue;
+        } else if(c1<0) {
+            cmpRes=-1;      /* string 1 ends before string 2 */
+            break;
+        } else if(c2<0) {
+            cmpRes=1;       /* string 2 ends before string 1 */
+            break;
+        }
+        /* c1!=c2 && c1>=0 && c2>=0 */
+
+        /* get complete code points for c1, c2 for lookups if either is a surrogate */
+        cp1=c1;
+        if(U_IS_SURROGATE(c1)) {
+            UChar c;
+
+            if(U_IS_SURROGATE_LEAD(c1)) {
+                if(s1!=limit1 && U16_IS_TRAIL(c=*s1)) {
+                    /* advance ++s1; only below if cp1 decomposes/case-folds */
+                    cp1=U16_GET_SUPPLEMENTARY(c1, c);
+                }
+            } else /* isTrail(c1) */ {
+                if(start1<=(s1-2) && U16_IS_LEAD(c=*(s1-2))) {
+                    cp1=U16_GET_SUPPLEMENTARY(c, c1);
+                }
+            }
+        }
+
+        cp2=c2;
+        if(U_IS_SURROGATE(c2)) {
+            UChar c;
+
+            if(U_IS_SURROGATE_LEAD(c2)) {
+                if(s2!=limit2 && U16_IS_TRAIL(c=*s2)) {
+                    /* advance ++s2; only below if cp2 decomposes/case-folds */
+                    cp2=U16_GET_SUPPLEMENTARY(c2, c);
+                }
+            } else /* isTrail(c2) */ {
+                if(start2<=(s2-2) && U16_IS_LEAD(c=*(s2-2))) {
+                    cp2=U16_GET_SUPPLEMENTARY(c, c2);
+                }
+            }
+        }
+
+        /*
+         * go down one level for each string
+         * continue with the main loop as soon as there is a real change
+         */
+
+        if( level1==0 &&
+            (length=ucase_toFullFolding((UChar32)cp1, &p, options))>=0
+        ) {
+            /* cp1 case-folds to the code point "length" or to p[length] */
+            if(U_IS_SURROGATE(c1)) {
+                if(U_IS_SURROGATE_LEAD(c1)) {
+                    /* advance beyond source surrogate pair if it case-folds */
+                    ++s1;
+                } else /* isTrail(c1) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s2;
+                    --m2;
+                    c2=*(s2-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack1[0].start=start1;
+            stack1[0].s=s1;
+            stack1[0].limit=limit1;
+            ++level1;
+
+            /* copy the folding result to fold1[] */
+            if(length<=UCASE_MAX_STRING_LENGTH) {
+                u_memcpy(fold1, p, length);
+            } else {
+                int32_t i=0;
+                U16_APPEND_UNSAFE(fold1, i, length);
+                length=i;
+            }
+
+            /* set next level pointers to case folding */
+            start1=s1=fold1;
+            limit1=fold1+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c1=-1;
+            continue;
+        }
+
+        if( level2==0 &&
+            (length=ucase_toFullFolding((UChar32)cp2, &p, options))>=0
+        ) {
+            /* cp2 case-folds to the code point "length" or to p[length] */
+            if(U_IS_SURROGATE(c2)) {
+                if(U_IS_SURROGATE_LEAD(c2)) {
+                    /* advance beyond source surrogate pair if it case-folds */
+                    ++s2;
+                } else /* isTrail(c2) */ {
+                    /*
+                     * we got a supplementary code point when hitting its trail surrogate,
+                     * therefore the lead surrogate must have been the same as in the other string;
+                     * compare this decomposition with the lead surrogate in the other string
+                     * remember that this simulates bulk text replacement:
+                     * the decomposition would replace the entire code point
+                     */
+                    --s1;
+                    --m2;
+                    c1=*(s1-1);
+                }
+            }
+
+            /* push current level pointers */
+            stack2[0].start=start2;
+            stack2[0].s=s2;
+            stack2[0].limit=limit2;
+            ++level2;
+
+            /* copy the folding result to fold2[] */
+            if(length<=UCASE_MAX_STRING_LENGTH) {
+                u_memcpy(fold2, p, length);
+            } else {
+                int32_t i=0;
+                U16_APPEND_UNSAFE(fold2, i, length);
+                length=i;
+            }
+
+            /* set next level pointers to case folding */
+            start2=s2=fold2;
+            limit2=fold2+length;
+
+            /* get ready to read from decomposition, continue with loop */
+            c2=-1;
+            continue;
+        }
+
+        /*
+         * no decomposition/case folding, max level for both sides:
+         * return difference result
+         *
+         * code point order comparison must not just return cp1-cp2
+         * because when single surrogates are present then the surrogate pairs
+         * that formed cp1 and cp2 may be from different string indexes
+         *
+         * example: { d800 d800 dc01 } vs. { d800 dc00 }, compare at second code units
+         * c1=d800 cp1=10001 c2=dc00 cp2=10000
+         * cp1-cp2>0 but c1-c2<0 and in fact in UTF-32 it is { d800 10001 } < { 10000 }
+         *
+         * therefore, use same fix-up as in ustring.c/uprv_strCompare()
+         * except: uprv_strCompare() fetches c=*s while this functions fetches c=*s++
+         * so we have slightly different pointer/start/limit comparisons here
+         */
+
+        if(c1>=0xd800 && c2>=0xd800 && (options&U_COMPARE_CODE_POINT_ORDER)) {
+            /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
+            if(
+                (c1<=0xdbff && s1!=limit1 && U16_IS_TRAIL(*s1)) ||
+                (U16_IS_TRAIL(c1) && start1!=(s1-1) && U16_IS_LEAD(*(s1-2)))
+            ) {
+                /* part of a surrogate pair, leave >=d800 */
+            } else {
+                /* BMP code point - may be surrogate code point - make <d800 */
+                c1-=0x2800;
+            }
+
+            if(
+                (c2<=0xdbff && s2!=limit2 && U16_IS_TRAIL(*s2)) ||
+                (U16_IS_TRAIL(c2) && start2!=(s2-1) && U16_IS_LEAD(*(s2-2)))
+            ) {
+                /* part of a surrogate pair, leave >=d800 */
+            } else {
+                /* BMP code point - may be surrogate code point - make <d800 */
+                c2-=0x2800;
+            }
+        }
+
+        cmpRes=c1-c2;
+        break;
+    }
+
+    if(matchLen1) {
+        *matchLen1=static_cast<int32_t>(m1-org1);
+        *matchLen2=static_cast<int32_t>(m2-org2);
+    }
+    return cmpRes;
+}
+
+/* internal function */
+U_CFUNC int32_t
+u_strcmpFold(const UChar *s1, int32_t length1,
+             const UChar *s2, int32_t length2,
+             uint32_t options,
+             UErrorCode *pErrorCode) {
+    return _cmpFold(s1, length1, s2, length2, options, NULL, NULL, pErrorCode);
+}
+
+/* public API functions */
+
+U_CAPI int32_t U_EXPORT2
+u_strCaseCompare(const UChar *s1, int32_t length1,
+                 const UChar *s2, int32_t length2,
+                 uint32_t options,
+                 UErrorCode *pErrorCode) {
+    /* argument checking */
+    if(pErrorCode==0 || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(s1==NULL || length1<-1 || s2==NULL || length2<-1) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    return u_strcmpFold(s1, length1, s2, length2,
+                        options|U_COMPARE_IGNORE_CASE,
+                        pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_strcasecmp(const UChar *s1, const UChar *s2, uint32_t options) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    return u_strcmpFold(s1, -1, s2, -1,
+                        options|U_COMPARE_IGNORE_CASE,
+                        &errorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_memcasecmp(const UChar *s1, const UChar *s2, int32_t length, uint32_t options) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    return u_strcmpFold(s1, length, s2, length,
+                        options|U_COMPARE_IGNORE_CASE,
+                        &errorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_strncasecmp(const UChar *s1, const UChar *s2, int32_t n, uint32_t options) {
+    UErrorCode errorCode=U_ZERO_ERROR;
+    return u_strcmpFold(s1, n, s2, n,
+                        options|(U_COMPARE_IGNORE_CASE|_STRNCMP_STYLE),
+                        &errorCode);
+}
+
+/* internal API - detect length of shared prefix */
+U_CAPI void
+u_caseInsensitivePrefixMatch(const UChar *s1, int32_t length1,
+                             const UChar *s2, int32_t length2,
+                             uint32_t options,
+                             int32_t *matchLen1, int32_t *matchLen2,
+                             UErrorCode *pErrorCode) {
+    _cmpFold(s1, length1, s2, length2, options,
+        matchLen1, matchLen2, pErrorCode);
+}
diff --git a/thirdparty/icu4c/common/ustrcase_locale.cpp b/thirdparty/icu4c/common/ustrcase_locale.cpp
new file mode 100644
index 0000000000..2ecd24f03e
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrcase_locale.cpp
@@ -0,0 +1,94 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  ustrcase_locale.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2011may31
+*   created by: Markus W. Scherer
+*
+*   Locale-sensitive case mapping functions (ones that call uloc_getDefault())
+*   were moved here to break dependency cycles among parts of the common library.
+*/
+
+#include "unicode/utypes.h"
+#include "uassert.h"
+#include "unicode/brkiter.h"
+#include "unicode/casemap.h"
+#include "unicode/ucasemap.h"
+#include "unicode/uloc.h"
+#include "unicode/ustring.h"
+#include "ucase.h"
+#include "ucasemap_imp.h"
+
+U_CFUNC int32_t
+ustrcase_getCaseLocale(const char *locale) {
+    if (locale == NULL) {
+        locale = uloc_getDefault();
+    }
+    if (*locale == 0) {
+        return UCASE_LOC_ROOT;
+    } else {
+        return ucase_getCaseLocale(locale);
+    }
+}
+
+/* public API functions */
+
+U_CAPI int32_t U_EXPORT2
+u_strToLower(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             const char *locale,
+             UErrorCode *pErrorCode) {
+    return ustrcase_mapWithOverlap(
+        ustrcase_getCaseLocale(locale), 0, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToLower, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+u_strToUpper(UChar *dest, int32_t destCapacity,
+             const UChar *src, int32_t srcLength,
+             const char *locale,
+             UErrorCode *pErrorCode) {
+    return ustrcase_mapWithOverlap(
+        ustrcase_getCaseLocale(locale), 0, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToUpper, *pErrorCode);
+}
+
+U_NAMESPACE_BEGIN
+
+int32_t CaseMap::toLower(
+        const char *locale, uint32_t options,
+        const UChar *src, int32_t srcLength,
+        UChar *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ustrcase_map(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToLower, edits, errorCode);
+}
+
+int32_t CaseMap::toUpper(
+        const char *locale, uint32_t options,
+        const UChar *src, int32_t srcLength,
+        UChar *dest, int32_t destCapacity, Edits *edits,
+        UErrorCode &errorCode) {
+    return ustrcase_map(
+        ustrcase_getCaseLocale(locale), options, UCASEMAP_BREAK_ITERATOR_NULL
+        dest, destCapacity,
+        src, srcLength,
+        ustrcase_internalToUpper, edits, errorCode);
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/ustrenum.cpp b/thirdparty/icu4c/common/ustrenum.cpp
new file mode 100644
index 0000000000..ed23eaa232
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrenum.cpp
@@ -0,0 +1,398 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: November 11 2002
+* Since: ICU 2.4
+**********************************************************************
+*/
+#include "utypeinfo.h"  // for 'typeid' to work 
+
+#include "unicode/ustring.h"
+#include "unicode/strenum.h"
+#include "unicode/putil.h"
+#include "uenumimp.h"
+#include "ustrenum.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+// StringEnumeration implementation ---------------------------------------- ***
+
+StringEnumeration::StringEnumeration()
+    : chars(charsBuffer), charsCapacity(sizeof(charsBuffer)) {
+}
+
+StringEnumeration::~StringEnumeration() {
+    if (chars != NULL && chars != charsBuffer) {
+        uprv_free(chars);
+    }
+}
+
+// StringEnumeration base class clone() default implementation, does not clone
+StringEnumeration *
+StringEnumeration::clone() const {
+  return NULL;
+}
+
+const char *
+StringEnumeration::next(int32_t *resultLength, UErrorCode &status) {
+    const UnicodeString *s=snext(status);
+    if(U_SUCCESS(status) && s!=NULL) {
+        unistr=*s;
+        ensureCharsCapacity(unistr.length()+1, status);
+        if(U_SUCCESS(status)) {
+            if(resultLength!=NULL) {
+                *resultLength=unistr.length();
+            }
+            unistr.extract(0, INT32_MAX, chars, charsCapacity, US_INV);
+            return chars;
+        }
+    }
+
+    return NULL;
+}
+
+const UChar *
+StringEnumeration::unext(int32_t *resultLength, UErrorCode &status) {
+    const UnicodeString *s=snext(status);
+    if(U_SUCCESS(status) && s!=NULL) {
+        unistr=*s;
+        if(resultLength!=NULL) {
+            *resultLength=unistr.length();
+        }
+        return unistr.getTerminatedBuffer();
+    }
+
+    return NULL;
+}
+
+const UnicodeString *
+StringEnumeration::snext(UErrorCode &status) {
+    int32_t length;
+    const char *s=next(&length, status);
+    return setChars(s, length, status);
+}
+
+void
+StringEnumeration::ensureCharsCapacity(int32_t capacity, UErrorCode &status) {
+    if(U_SUCCESS(status) && capacity>charsCapacity) {
+        if(capacity<(charsCapacity+charsCapacity/2)) {
+            // avoid allocation thrashing
+            capacity=charsCapacity+charsCapacity/2;
+        }
+        if(chars!=charsBuffer) {
+            uprv_free(chars);
+        }
+        chars=(char *)uprv_malloc(capacity);
+        if(chars==NULL) {
+            chars=charsBuffer;
+            charsCapacity=sizeof(charsBuffer);
+            status=U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            charsCapacity=capacity;
+        }
+    }
+}
+
+UnicodeString *
+StringEnumeration::setChars(const char *s, int32_t length, UErrorCode &status) {
+    if(U_SUCCESS(status) && s!=NULL) {
+        if(length<0) {
+            length=(int32_t)uprv_strlen(s);
+        }
+
+        UChar *buffer=unistr.getBuffer(length+1);
+        if(buffer!=NULL) {
+            u_charsToUChars(s, buffer, length);
+            buffer[length]=0;
+            unistr.releaseBuffer(length);
+            return &unistr;
+        } else {
+            status=U_MEMORY_ALLOCATION_ERROR;
+        }
+    }
+
+    return NULL;
+}
+UBool 
+StringEnumeration::operator==(const StringEnumeration& that)const {
+    return typeid(*this) == typeid(that); 
+}
+
+UBool
+StringEnumeration::operator!=(const StringEnumeration& that)const {
+    return !operator==(that);
+}
+
+// UStringEnumeration implementation --------------------------------------- ***
+
+UStringEnumeration * U_EXPORT2
+UStringEnumeration::fromUEnumeration(
+        UEnumeration *uenumToAdopt, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        uenum_close(uenumToAdopt);
+        return NULL;
+    }
+    UStringEnumeration *result = new UStringEnumeration(uenumToAdopt);
+    if (result == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        uenum_close(uenumToAdopt);
+        return NULL;
+    }
+    return result;
+}
+
+UStringEnumeration::UStringEnumeration(UEnumeration* _uenum) :
+    uenum(_uenum) {
+    U_ASSERT(_uenum != 0);
+}
+
+UStringEnumeration::~UStringEnumeration() {
+    uenum_close(uenum);
+}
+
+int32_t UStringEnumeration::count(UErrorCode& status) const {
+    return uenum_count(uenum, &status);
+}
+
+const char *UStringEnumeration::next(int32_t *resultLength, UErrorCode &status) {
+    return uenum_next(uenum, resultLength, &status);
+}
+
+const UnicodeString* UStringEnumeration::snext(UErrorCode& status) {
+    int32_t length;
+    const UChar* str = uenum_unext(uenum, &length, &status);
+    if (str == 0 || U_FAILURE(status)) {
+        return 0;
+    }
+    return &unistr.setTo(str, length);
+}
+
+void UStringEnumeration::reset(UErrorCode& status) {
+    uenum_reset(uenum, &status);
+}
+
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UStringEnumeration)
+U_NAMESPACE_END
+
+// C wrapper --------------------------------------------------------------- ***
+
+#define THIS(en) ((icu::StringEnumeration*)(en->context))
+
+U_CDECL_BEGIN
+
+/**
+ * Wrapper API to make StringEnumeration look like UEnumeration.
+ */
+static void U_CALLCONV
+ustrenum_close(UEnumeration* en) {
+    delete THIS(en);
+    uprv_free(en);
+}
+
+/**
+ * Wrapper API to make StringEnumeration look like UEnumeration.
+ */
+static int32_t U_CALLCONV
+ustrenum_count(UEnumeration* en,
+               UErrorCode* ec)
+{
+    return THIS(en)->count(*ec);
+}
+
+/**
+ * Wrapper API to make StringEnumeration look like UEnumeration.
+ */
+static const UChar* U_CALLCONV
+ustrenum_unext(UEnumeration* en,
+               int32_t* resultLength,
+               UErrorCode* ec)
+{
+    return THIS(en)->unext(resultLength, *ec);
+}
+
+/**
+ * Wrapper API to make StringEnumeration look like UEnumeration.
+ */
+static const char* U_CALLCONV
+ustrenum_next(UEnumeration* en,
+              int32_t* resultLength,
+              UErrorCode* ec)
+{
+    return THIS(en)->next(resultLength, *ec);
+}
+
+/**
+ * Wrapper API to make StringEnumeration look like UEnumeration.
+ */
+static void U_CALLCONV
+ustrenum_reset(UEnumeration* en,
+               UErrorCode* ec)
+{
+    THIS(en)->reset(*ec);
+}
+
+/**
+ * Pseudo-vtable for UEnumeration wrapper around StringEnumeration.
+ * The StringEnumeration pointer will be stored in 'context'.
+ */
+static const UEnumeration USTRENUM_VT = {
+    NULL,
+    NULL, // store StringEnumeration pointer here
+    ustrenum_close,
+    ustrenum_count,
+    ustrenum_unext,
+    ustrenum_next,
+    ustrenum_reset
+};
+
+U_CDECL_END
+
+/**
+ * Given a StringEnumeration, wrap it in a UEnumeration.  The
+ * StringEnumeration is adopted; after this call, the caller must not
+ * delete it (regardless of error status).
+ */
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openFromStringEnumeration(icu::StringEnumeration* adopted, UErrorCode* ec) { 
+    UEnumeration* result = NULL;
+    if (U_SUCCESS(*ec) && adopted != NULL) {
+        result = (UEnumeration*) uprv_malloc(sizeof(UEnumeration));
+        if (result == NULL) {
+            *ec = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            uprv_memcpy(result, &USTRENUM_VT, sizeof(USTRENUM_VT));
+            result->context = adopted;
+        }
+    }
+    if (result == NULL) {
+        delete adopted;
+    }
+    return result;
+}
+
+// C wrapper --------------------------------------------------------------- ***
+
+U_CDECL_BEGIN
+
+typedef struct UCharStringEnumeration {
+    UEnumeration uenum;
+    int32_t index, count;
+} UCharStringEnumeration;
+
+static void U_CALLCONV
+ucharstrenum_close(UEnumeration* en) {
+    uprv_free(en);
+}
+
+static int32_t U_CALLCONV
+ucharstrenum_count(UEnumeration* en,
+                   UErrorCode* /*ec*/) {
+    return ((UCharStringEnumeration*)en)->count;
+}
+
+static const UChar* U_CALLCONV
+ucharstrenum_unext(UEnumeration* en,
+                  int32_t* resultLength,
+                  UErrorCode* /*ec*/) {
+    UCharStringEnumeration *e = (UCharStringEnumeration*) en;
+    if (e->index >= e->count) {
+        return NULL;
+    }
+    const UChar* result = ((const UChar**)e->uenum.context)[e->index++];
+    if (resultLength) {
+        *resultLength = (int32_t)u_strlen(result);
+    }
+    return result;
+}
+
+
+static const char* U_CALLCONV
+ucharstrenum_next(UEnumeration* en,
+                  int32_t* resultLength,
+                  UErrorCode* /*ec*/) {
+    UCharStringEnumeration *e = (UCharStringEnumeration*) en;
+    if (e->index >= e->count) {
+        return NULL;
+    }
+    const char* result = ((const char**)e->uenum.context)[e->index++];
+    if (resultLength) {
+        *resultLength = (int32_t)uprv_strlen(result);
+    }
+    return result;
+}
+
+static void U_CALLCONV
+ucharstrenum_reset(UEnumeration* en,
+                   UErrorCode* /*ec*/) {
+    ((UCharStringEnumeration*)en)->index = 0;
+}
+
+static const UEnumeration UCHARSTRENUM_VT = {
+    NULL,
+    NULL, // store StringEnumeration pointer here
+    ucharstrenum_close,
+    ucharstrenum_count,
+    uenum_unextDefault,
+    ucharstrenum_next,
+    ucharstrenum_reset
+};
+
+static const UEnumeration UCHARSTRENUM_U_VT = {
+    NULL,
+    NULL, // store StringEnumeration pointer here
+    ucharstrenum_close,
+    ucharstrenum_count,
+    ucharstrenum_unext,
+    uenum_nextDefault,
+    ucharstrenum_reset
+};
+
+U_CDECL_END
+
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openCharStringsEnumeration(const char* const strings[], int32_t count,
+                                 UErrorCode* ec) {
+    UCharStringEnumeration* result = NULL;
+    if (U_SUCCESS(*ec) && count >= 0 && (count == 0 || strings != 0)) {
+        result = (UCharStringEnumeration*) uprv_malloc(sizeof(UCharStringEnumeration));
+        if (result == NULL) {
+            *ec = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            U_ASSERT((char*)result==(char*)(&result->uenum));
+            uprv_memcpy(result, &UCHARSTRENUM_VT, sizeof(UCHARSTRENUM_VT));
+            result->uenum.context = (void*)strings;
+            result->index = 0;
+            result->count = count;
+        }
+    }
+    return (UEnumeration*) result;
+}
+
+U_CAPI UEnumeration* U_EXPORT2
+uenum_openUCharStringsEnumeration(const UChar* const strings[], int32_t count,
+                                 UErrorCode* ec) {
+    UCharStringEnumeration* result = NULL;
+    if (U_SUCCESS(*ec) && count >= 0 && (count == 0 || strings != 0)) {
+        result = (UCharStringEnumeration*) uprv_malloc(sizeof(UCharStringEnumeration));
+        if (result == NULL) {
+            *ec = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            U_ASSERT((char*)result==(char*)(&result->uenum));
+            uprv_memcpy(result, &UCHARSTRENUM_U_VT, sizeof(UCHARSTRENUM_U_VT));
+            result->uenum.context = (void*)strings;
+            result->index = 0;
+            result->count = count;
+        }
+    }
+    return (UEnumeration*) result;
+}
+
+
+// end C Wrapper
diff --git a/thirdparty/icu4c/common/ustrenum.h b/thirdparty/icu4c/common/ustrenum.h
new file mode 100644
index 0000000000..a82162e2bd
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrenum.h
@@ -0,0 +1,87 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+* Copyright (c) 2002-2014, International Business Machines
+* Corporation and others.  All Rights Reserved.
+**********************************************************************
+* Author: Alan Liu
+* Created: November 11 2002
+* Since: ICU 2.4
+**********************************************************************
+*/
+#ifndef _USTRENUM_H_
+#define _USTRENUM_H_
+
+#include "unicode/uenum.h"
+#include "unicode/strenum.h"
+
+//----------------------------------------------------------------------
+U_NAMESPACE_BEGIN
+
+/**
+ * A wrapper to make a UEnumeration into a StringEnumeration.  The
+ * wrapper adopts the UEnumeration is wraps.
+ */
+class U_COMMON_API UStringEnumeration : public StringEnumeration {
+
+public:
+    /**
+     * Constructor.  This constructor adopts its UEnumeration
+     * argument.
+     * @param uenum a UEnumeration object.  This object takes
+     * ownership of 'uenum' and will close it in its destructor.  The
+     * caller must not call uenum_close on 'uenum' after calling this
+     * constructor.
+     */
+    UStringEnumeration(UEnumeration* uenum);
+
+    /**
+     * Destructor.  This closes the UEnumeration passed in to the
+     * constructor.
+     */
+    virtual ~UStringEnumeration();
+
+    /**
+     * Return the number of elements that the iterator traverses.
+     * @param status the error code.
+     * @return number of elements in the iterator.
+     */
+    virtual int32_t count(UErrorCode& status) const;
+
+    virtual const char* next(int32_t *resultLength, UErrorCode& status);
+
+    /**
+     * Returns the next element a UnicodeString*.  If there are no
+     * more elements, returns NULL.
+     * @param status the error code.
+     * @return a pointer to the string, or NULL.
+     */
+    virtual const UnicodeString* snext(UErrorCode& status);
+
+    /**
+     * Resets the iterator.
+     * @param status the error code.
+     */
+    virtual void reset(UErrorCode& status);
+
+    /**
+     * ICU4C "poor man's RTTI", returns a UClassID for the actual ICU class.
+     */
+    virtual UClassID getDynamicClassID() const;
+
+    /**
+     * ICU4C "poor man's RTTI", returns a UClassID for this ICU class.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    static UStringEnumeration * U_EXPORT2 fromUEnumeration(
+            UEnumeration *enumToAdopt, UErrorCode &status);
+private:
+    UEnumeration *uenum; // owned
+};
+
+U_NAMESPACE_END
+
+#endif
+
diff --git a/thirdparty/icu4c/common/ustrfmt.cpp b/thirdparty/icu4c/common/ustrfmt.cpp
new file mode 100644
index 0000000000..1a9b15a59f
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrfmt.cpp
@@ -0,0 +1,59 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2001-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#include "cstring.h"
+#include "ustrfmt.h"
+
+
+/***
+ * Fills in a UChar* string with the radix-based representation of a
+ * uint32_t number padded with zeroes to minwidth.  The result
+ * will be null terminated if there is room.
+ *
+ * @param buffer UChar buffer to receive result
+ * @param capacity capacity of buffer
+ * @param i the unsigned number to be formatted
+ * @param radix the radix from 2..36
+ * @param minwidth the minimum width.  If the result is narrower than
+ *        this, '0's will be added on the left.  Must be <=
+ *        capacity.
+ * @return the length of the result, not including any terminating
+ *        null
+ */
+U_CAPI int32_t U_EXPORT2
+uprv_itou (UChar * buffer, int32_t capacity,
+           uint32_t i, uint32_t radix, int32_t minwidth)
+{
+    int32_t length = 0;
+    int digit;
+    int32_t j;
+    UChar temp;
+
+    do{
+        digit = (int)(i % radix);
+        buffer[length++]=(UChar)(digit<=9?(0x0030+digit):(0x0030+digit+7));
+        i=i/radix;
+    } while(i && length<capacity);
+
+    while (length < minwidth){
+        buffer[length++] = (UChar) 0x0030;/*zero padding */
+    }
+    /* null terminate the buffer */
+    if(length<capacity){
+        buffer[length] = (UChar) 0x0000;
+    }
+
+    /* Reverses the string */
+    for (j = 0; j < (length / 2); j++){
+        temp = buffer[(length-1) - j];
+        buffer[(length-1) - j] = buffer[j];
+        buffer[j] = temp;
+    }
+    return length;
+}
diff --git a/thirdparty/icu4c/common/ustrfmt.h b/thirdparty/icu4c/common/ustrfmt.h
new file mode 100644
index 0000000000..53eb0557e4
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrfmt.h
@@ -0,0 +1,19 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 2001-2006, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+#ifndef USTRFMT_H
+#define USTRFMT_H
+
+#include "unicode/utypes.h"
+
+U_CAPI int32_t U_EXPORT2
+uprv_itou (UChar * buffer, int32_t capacity, uint32_t i, uint32_t radix, int32_t minwidth);
+
+
+#endif
diff --git a/thirdparty/icu4c/common/ustring.cpp b/thirdparty/icu4c/common/ustring.cpp
new file mode 100644
index 0000000000..bba2d45c4e
--- /dev/null
+++ b/thirdparty/icu4c/common/ustring.cpp
@@ -0,0 +1,1537 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1998-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File ustring.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   12/07/98    bertrand    Creation.
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+#include "unicode/putil.h"
+#include "unicode/uchar.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "cstring.h"
+#include "cwchar.h"
+#include "cmemory.h"
+#include "ustr_imp.h"
+
+/* ANSI string.h - style functions ------------------------------------------ */
+
+/* U+ffff is the highest BMP code point, the highest one that fits into a 16-bit UChar */
+#define U_BMP_MAX 0xffff
+
+/* Forward binary string search functions ----------------------------------- */
+
+/*
+ * Test if a substring match inside a string is at code point boundaries.
+ * All pointers refer to the same buffer.
+ * The limit pointer may be NULL, all others must be real pointers.
+ */
+static inline UBool
+isMatchAtCPBoundary(const UChar *start, const UChar *match, const UChar *matchLimit, const UChar *limit) {
+    if(U16_IS_TRAIL(*match) && start!=match && U16_IS_LEAD(*(match-1))) {
+        /* the leading edge of the match is in the middle of a surrogate pair */
+        return FALSE;
+    }
+    if(U16_IS_LEAD(*(matchLimit-1)) && matchLimit!=limit && U16_IS_TRAIL(*matchLimit)) {
+        /* the trailing edge of the match is in the middle of a surrogate pair */
+        return FALSE;
+    }
+    return TRUE;
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strFindFirst(const UChar *s, int32_t length,
+               const UChar *sub, int32_t subLength) {
+    const UChar *start, *p, *q, *subLimit;
+    UChar c, cs, cq;
+
+    if(sub==NULL || subLength<-1) {
+        return (UChar *)s;
+    }
+    if(s==NULL || length<-1) {
+        return NULL;
+    }
+
+    start=s;
+
+    if(length<0 && subLength<0) {
+        /* both strings are NUL-terminated */
+        if((cs=*sub++)==0) {
+            return (UChar *)s;
+        }
+        if(*sub==0 && !U16_IS_SURROGATE(cs)) {
+            /* the substring consists of a single, non-surrogate BMP code point */
+            return u_strchr(s, cs);
+        }
+
+        while((c=*s++)!=0) {
+            if(c==cs) {
+                /* found first substring UChar, compare rest */
+                p=s;
+                q=sub;
+                for(;;) {
+                    if((cq=*q)==0) {
+                        if(isMatchAtCPBoundary(start, s-1, p, NULL)) {
+                            return (UChar *)(s-1); /* well-formed match */
+                        } else {
+                            break; /* no match because surrogate pair is split */
+                        }
+                    }
+                    if((c=*p)==0) {
+                        return NULL; /* no match, and none possible after s */
+                    }
+                    if(c!=cq) {
+                        break; /* no match */
+                    }
+                    ++p;
+                    ++q;
+                }
+            }
+        }
+
+        /* not found */
+        return NULL;
+    }
+
+    if(subLength<0) {
+        subLength=u_strlen(sub);
+    }
+    if(subLength==0) {
+        return (UChar *)s;
+    }
+
+    /* get sub[0] to search for it fast */
+    cs=*sub++;
+    --subLength;
+    subLimit=sub+subLength;
+
+    if(subLength==0 && !U16_IS_SURROGATE(cs)) {
+        /* the substring consists of a single, non-surrogate BMP code point */
+        return length<0 ? u_strchr(s, cs) : u_memchr(s, cs, length);
+    }
+
+    if(length<0) {
+        /* s is NUL-terminated */
+        while((c=*s++)!=0) {
+            if(c==cs) {
+                /* found first substring UChar, compare rest */
+                p=s;
+                q=sub;
+                for(;;) {
+                    if(q==subLimit) {
+                        if(isMatchAtCPBoundary(start, s-1, p, NULL)) {
+                            return (UChar *)(s-1); /* well-formed match */
+                        } else {
+                            break; /* no match because surrogate pair is split */
+                        }
+                    }
+                    if((c=*p)==0) {
+                        return NULL; /* no match, and none possible after s */
+                    }
+                    if(c!=*q) {
+                        break; /* no match */
+                    }
+                    ++p;
+                    ++q;
+                }
+            }
+        }
+    } else {
+        const UChar *limit, *preLimit;
+
+        /* subLength was decremented above */
+        if(length<=subLength) {
+            return NULL; /* s is shorter than sub */
+        }
+
+        limit=s+length;
+
+        /* the substring must start before preLimit */
+        preLimit=limit-subLength;
+
+        while(s!=preLimit) {
+            c=*s++;
+            if(c==cs) {
+                /* found first substring UChar, compare rest */
+                p=s;
+                q=sub;
+                for(;;) {
+                    if(q==subLimit) {
+                        if(isMatchAtCPBoundary(start, s-1, p, limit)) {
+                            return (UChar *)(s-1); /* well-formed match */
+                        } else {
+                            break; /* no match because surrogate pair is split */
+                        }
+                    }
+                    if(*p!=*q) {
+                        break; /* no match */
+                    }
+                    ++p;
+                    ++q;
+                }
+            }
+        }
+    }
+
+    /* not found */
+    return NULL;
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strstr(const UChar *s, const UChar *substring) {
+    return u_strFindFirst(s, -1, substring, -1);
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strchr(const UChar *s, UChar c) {
+    if(U16_IS_SURROGATE(c)) {
+        /* make sure to not find half of a surrogate pair */
+        return u_strFindFirst(s, -1, &c, 1);
+    } else {
+        UChar cs;
+
+        /* trivial search for a BMP code point */
+        for(;;) {
+            if((cs=*s)==c) {
+                return (UChar *)s;
+            }
+            if(cs==0) {
+                return NULL;
+            }
+            ++s;
+        }
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strchr32(const UChar *s, UChar32 c) {
+    if((uint32_t)c<=U_BMP_MAX) {
+        /* find BMP code point */
+        return u_strchr(s, (UChar)c);
+    } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
+        /* find supplementary code point as surrogate pair */
+        UChar cs, lead=U16_LEAD(c), trail=U16_TRAIL(c);
+
+        while((cs=*s++)!=0) {
+            if(cs==lead && *s==trail) {
+                return (UChar *)(s-1);
+            }
+        }
+        return NULL;
+    } else {
+        /* not a Unicode code point, not findable */
+        return NULL;
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memchr(const UChar *s, UChar c, int32_t count) {
+    if(count<=0) {
+        return NULL; /* no string */
+    } else if(U16_IS_SURROGATE(c)) {
+        /* make sure to not find half of a surrogate pair */
+        return u_strFindFirst(s, count, &c, 1);
+    } else {
+        /* trivial search for a BMP code point */
+        const UChar *limit=s+count;
+        do {
+            if(*s==c) {
+                return (UChar *)s;
+            }
+        } while(++s!=limit);
+        return NULL;
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memchr32(const UChar *s, UChar32 c, int32_t count) {
+    if((uint32_t)c<=U_BMP_MAX) {
+        /* find BMP code point */
+        return u_memchr(s, (UChar)c, count);
+    } else if(count<2) {
+        /* too short for a surrogate pair */
+        return NULL;
+    } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
+        /* find supplementary code point as surrogate pair */
+        const UChar *limit=s+count-1; /* -1 so that we do not need a separate check for the trail unit */
+        UChar lead=U16_LEAD(c), trail=U16_TRAIL(c);
+
+        do {
+            if(*s==lead && *(s+1)==trail) {
+                return (UChar *)s;
+            }
+        } while(++s!=limit);
+        return NULL;
+    } else {
+        /* not a Unicode code point, not findable */
+        return NULL;
+    }
+}
+
+/* Backward binary string search functions ---------------------------------- */
+
+U_CAPI UChar * U_EXPORT2
+u_strFindLast(const UChar *s, int32_t length,
+              const UChar *sub, int32_t subLength) {
+    const UChar *start, *limit, *p, *q, *subLimit;
+    UChar c, cs;
+
+    if(sub==NULL || subLength<-1) {
+        return (UChar *)s;
+    }
+    if(s==NULL || length<-1) {
+        return NULL;
+    }
+
+    /*
+     * This implementation is more lazy than the one for u_strFindFirst():
+     * There is no special search code for NUL-terminated strings.
+     * It does not seem to be worth it for searching substrings to
+     * search forward and find all matches like in u_strrchr() and similar.
+     * Therefore, we simply get both string lengths and search backward.
+     *
+     * markus 2002oct23
+     */
+
+    if(subLength<0) {
+        subLength=u_strlen(sub);
+    }
+    if(subLength==0) {
+        return (UChar *)s;
+    }
+
+    /* get sub[subLength-1] to search for it fast */
+    subLimit=sub+subLength;
+    cs=*(--subLimit);
+    --subLength;
+
+    if(subLength==0 && !U16_IS_SURROGATE(cs)) {
+        /* the substring consists of a single, non-surrogate BMP code point */
+        return length<0 ? u_strrchr(s, cs) : u_memrchr(s, cs, length);
+    }
+
+    if(length<0) {
+        length=u_strlen(s);
+    }
+
+    /* subLength was decremented above */
+    if(length<=subLength) {
+        return NULL; /* s is shorter than sub */
+    }
+
+    start=s;
+    limit=s+length;
+
+    /* the substring must start no later than s+subLength */
+    s+=subLength;
+
+    while(s!=limit) {
+        c=*(--limit);
+        if(c==cs) {
+            /* found last substring UChar, compare rest */
+            p=limit;
+            q=subLimit;
+            for(;;) {
+                if(q==sub) {
+                    if(isMatchAtCPBoundary(start, p, limit+1, start+length)) {
+                        return (UChar *)p; /* well-formed match */
+                    } else {
+                        break; /* no match because surrogate pair is split */
+                    }
+                }
+                if(*(--p)!=*(--q)) {
+                    break; /* no match */
+                }
+            }
+        }
+    }
+
+    /* not found */
+    return NULL;
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strrstr(const UChar *s, const UChar *substring) {
+    return u_strFindLast(s, -1, substring, -1);
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strrchr(const UChar *s, UChar c) {
+    if(U16_IS_SURROGATE(c)) {
+        /* make sure to not find half of a surrogate pair */
+        return u_strFindLast(s, -1, &c, 1);
+    } else {
+        const UChar *result=NULL;
+        UChar cs;
+
+        /* trivial search for a BMP code point */
+        for(;;) {
+            if((cs=*s)==c) {
+                result=s;
+            }
+            if(cs==0) {
+                return (UChar *)result;
+            }
+            ++s;
+        }
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strrchr32(const UChar *s, UChar32 c) {
+    if((uint32_t)c<=U_BMP_MAX) {
+        /* find BMP code point */
+        return u_strrchr(s, (UChar)c);
+    } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
+        /* find supplementary code point as surrogate pair */
+        const UChar *result=NULL;
+        UChar cs, lead=U16_LEAD(c), trail=U16_TRAIL(c);
+
+        while((cs=*s++)!=0) {
+            if(cs==lead && *s==trail) {
+                result=s-1;
+            }
+        }
+        return (UChar *)result;
+    } else {
+        /* not a Unicode code point, not findable */
+        return NULL;
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memrchr(const UChar *s, UChar c, int32_t count) {
+    if(count<=0) {
+        return NULL; /* no string */
+    } else if(U16_IS_SURROGATE(c)) {
+        /* make sure to not find half of a surrogate pair */
+        return u_strFindLast(s, count, &c, 1);
+    } else {
+        /* trivial search for a BMP code point */
+        const UChar *limit=s+count;
+        do {
+            if(*(--limit)==c) {
+                return (UChar *)limit;
+            }
+        } while(s!=limit);
+        return NULL;
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memrchr32(const UChar *s, UChar32 c, int32_t count) {
+    if((uint32_t)c<=U_BMP_MAX) {
+        /* find BMP code point */
+        return u_memrchr(s, (UChar)c, count);
+    } else if(count<2) {
+        /* too short for a surrogate pair */
+        return NULL;
+    } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
+        /* find supplementary code point as surrogate pair */
+        const UChar *limit=s+count-1;
+        UChar lead=U16_LEAD(c), trail=U16_TRAIL(c);
+
+        do {
+            if(*limit==trail && *(limit-1)==lead) {
+                return (UChar *)(limit-1);
+            }
+        } while(s!=--limit);
+        return NULL;
+    } else {
+        /* not a Unicode code point, not findable */
+        return NULL;
+    }
+}
+
+/* Tokenization functions --------------------------------------------------- */
+
+/*
+ * Match each code point in a string against each code point in the matchSet.
+ * Return the index of the first string code point that
+ * is (polarity==TRUE) or is not (FALSE) contained in the matchSet.
+ * Return -(string length)-1 if there is no such code point.
+ */
+static int32_t
+_matchFromSet(const UChar *string, const UChar *matchSet, UBool polarity) {
+    int32_t matchLen, matchBMPLen, strItr, matchItr;
+    UChar32 stringCh, matchCh;
+    UChar c, c2;
+
+    /* first part of matchSet contains only BMP code points */
+    matchBMPLen = 0;
+    while((c = matchSet[matchBMPLen]) != 0 && U16_IS_SINGLE(c)) {
+        ++matchBMPLen;
+    }
+
+    /* second part of matchSet contains BMP and supplementary code points */
+    matchLen = matchBMPLen;
+    while(matchSet[matchLen] != 0) {
+        ++matchLen;
+    }
+
+    for(strItr = 0; (c = string[strItr]) != 0;) {
+        ++strItr;
+        if(U16_IS_SINGLE(c)) {
+            if(polarity) {
+                for(matchItr = 0; matchItr < matchLen; ++matchItr) {
+                    if(c == matchSet[matchItr]) {
+                        return strItr - 1; /* one matches */
+                    }
+                }
+            } else {
+                for(matchItr = 0; matchItr < matchLen; ++matchItr) {
+                    if(c == matchSet[matchItr]) {
+                        goto endloop;
+                    }
+                }
+                return strItr - 1; /* none matches */
+            }
+        } else {
+            /*
+             * No need to check for string length before U16_IS_TRAIL
+             * because c2 could at worst be the terminating NUL.
+             */
+            if(U16_IS_SURROGATE_LEAD(c) && U16_IS_TRAIL(c2 = string[strItr])) {
+                ++strItr;
+                stringCh = U16_GET_SUPPLEMENTARY(c, c2);
+            } else {
+                stringCh = c; /* unpaired trail surrogate */
+            }
+
+            if(polarity) {
+                for(matchItr = matchBMPLen; matchItr < matchLen;) {
+                    U16_NEXT(matchSet, matchItr, matchLen, matchCh);
+                    if(stringCh == matchCh) {
+                        return strItr - U16_LENGTH(stringCh); /* one matches */
+                    }
+                }
+            } else {
+                for(matchItr = matchBMPLen; matchItr < matchLen;) {
+                    U16_NEXT(matchSet, matchItr, matchLen, matchCh);
+                    if(stringCh == matchCh) {
+                        goto endloop;
+                    }
+                }
+                return strItr - U16_LENGTH(stringCh); /* none matches */
+            }
+        }
+endloop:
+        /* wish C had continue with labels like Java... */;
+    }
+
+    /* Didn't find it. */
+    return -strItr-1;
+}
+
+/* Search for a codepoint in a string that matches one of the matchSet codepoints. */
+U_CAPI UChar * U_EXPORT2
+u_strpbrk(const UChar *string, const UChar *matchSet)
+{
+    int32_t idx = _matchFromSet(string, matchSet, TRUE);
+    if(idx >= 0) {
+        return (UChar *)string + idx;
+    } else {
+        return NULL;
+    }
+}
+
+/* Search for a codepoint in a string that matches one of the matchSet codepoints. */
+U_CAPI int32_t U_EXPORT2
+u_strcspn(const UChar *string, const UChar *matchSet)
+{
+    int32_t idx = _matchFromSet(string, matchSet, TRUE);
+    if(idx >= 0) {
+        return idx;
+    } else {
+        return -idx - 1; /* == u_strlen(string) */
+    }
+}
+
+/* Search for a codepoint in a string that does not match one of the matchSet codepoints. */
+U_CAPI int32_t U_EXPORT2
+u_strspn(const UChar *string, const UChar *matchSet)
+{
+    int32_t idx = _matchFromSet(string, matchSet, FALSE);
+    if(idx >= 0) {
+        return idx;
+    } else {
+        return -idx - 1; /* == u_strlen(string) */
+    }
+}
+
+/* ----- Text manipulation functions --- */
+
+U_CAPI UChar* U_EXPORT2
+u_strtok_r(UChar    *src, 
+     const UChar    *delim,
+           UChar   **saveState)
+{
+    UChar *tokSource;
+    UChar *nextToken;
+    uint32_t nonDelimIdx;
+
+    /* If saveState is NULL, the user messed up. */
+    if (src != NULL) {
+        tokSource = src;
+        *saveState = src; /* Set to "src" in case there are no delimiters */
+    }
+    else if (*saveState) {
+        tokSource = *saveState;
+    }
+    else {
+        /* src == NULL && *saveState == NULL */
+        /* This shouldn't happen. We already finished tokenizing. */
+        return NULL;
+    }
+
+    /* Skip initial delimiters */
+    nonDelimIdx = u_strspn(tokSource, delim);
+    tokSource = &tokSource[nonDelimIdx];
+
+    if (*tokSource) {
+        nextToken = u_strpbrk(tokSource, delim);
+        if (nextToken != NULL) {
+            /* Create a token */
+            *(nextToken++) = 0;
+            *saveState = nextToken;
+            return tokSource;
+        }
+        else if (*saveState) {
+            /* Return the last token */
+            *saveState = NULL;
+            return tokSource;
+        }
+    }
+    else {
+        /* No tokens were found. Only delimiters were left. */
+        *saveState = NULL;
+    }
+    return NULL;
+}
+
+/* Miscellaneous functions -------------------------------------------------- */
+
+U_CAPI UChar* U_EXPORT2
+u_strcat(UChar     *dst, 
+    const UChar     *src)
+{
+    UChar *anchor = dst;            /* save a pointer to start of dst */
+
+    while(*dst != 0) {              /* To end of first string          */
+        ++dst;
+    }
+    while((*(dst++) = *(src++)) != 0) {     /* copy string 2 over              */
+    }
+
+    return anchor;
+}
+
+U_CAPI UChar*  U_EXPORT2
+u_strncat(UChar     *dst, 
+     const UChar     *src, 
+     int32_t     n ) 
+{
+    if(n > 0) {
+        UChar *anchor = dst;            /* save a pointer to start of dst */
+
+        while(*dst != 0) {              /* To end of first string          */
+            ++dst;
+        }
+        while((*dst = *src) != 0) {     /* copy string 2 over              */
+            ++dst;
+            if(--n == 0) {
+                *dst = 0;
+                break;
+            }
+            ++src;
+        }
+
+        return anchor;
+    } else {
+        return dst;
+    }
+}
+
+/* ----- Text property functions --- */
+
+U_CAPI int32_t   U_EXPORT2
+u_strcmp(const UChar *s1, 
+    const UChar *s2) 
+{
+    UChar  c1, c2;
+
+    for(;;) {
+        c1=*s1++;
+        c2=*s2++;
+        if (c1 != c2 || c1 == 0) {
+            break;
+        }
+    }
+    return (int32_t)c1 - (int32_t)c2;
+}
+
+U_CFUNC int32_t U_EXPORT2
+uprv_strCompare(const UChar *s1, int32_t length1,
+                const UChar *s2, int32_t length2,
+                UBool strncmpStyle, UBool codePointOrder) {
+    const UChar *start1, *start2, *limit1, *limit2;
+    UChar c1, c2;
+
+    /* setup for fix-up */
+    start1=s1;
+    start2=s2;
+
+    /* compare identical prefixes - they do not need to be fixed up */
+    if(length1<0 && length2<0) {
+        /* strcmp style, both NUL-terminated */
+        if(s1==s2) {
+            return 0;
+        }
+
+        for(;;) {
+            c1=*s1;
+            c2=*s2;
+            if(c1!=c2) {
+                break;
+            }
+            if(c1==0) {
+                return 0;
+            }
+            ++s1;
+            ++s2;
+        }
+
+        /* setup for fix-up */
+        limit1=limit2=NULL;
+    } else if(strncmpStyle) {
+        /* special handling for strncmp, assume length1==length2>=0 but also check for NUL */
+        if(s1==s2) {
+            return 0;
+        }
+
+        limit1=start1+length1;
+
+        for(;;) {
+            /* both lengths are same, check only one limit */
+            if(s1==limit1) {
+                return 0;
+            }
+
+            c1=*s1;
+            c2=*s2;
+            if(c1!=c2) {
+                break;
+            }
+            if(c1==0) {
+                return 0;
+            }
+            ++s1;
+            ++s2;
+        }
+
+        /* setup for fix-up */
+        limit2=start2+length1; /* use length1 here, too, to enforce assumption */
+    } else {
+        /* memcmp/UnicodeString style, both length-specified */
+        int32_t lengthResult;
+
+        if(length1<0) {
+            length1=u_strlen(s1);
+        }
+        if(length2<0) {
+            length2=u_strlen(s2);
+        }
+
+        /* limit1=start1+min(lenght1, length2) */
+        if(length1<length2) {
+            lengthResult=-1;
+            limit1=start1+length1;
+        } else if(length1==length2) {
+            lengthResult=0;
+            limit1=start1+length1;
+        } else /* length1>length2 */ {
+            lengthResult=1;
+            limit1=start1+length2;
+        }
+
+        if(s1==s2) {
+            return lengthResult;
+        }
+
+        for(;;) {
+            /* check pseudo-limit */
+            if(s1==limit1) {
+                return lengthResult;
+            }
+
+            c1=*s1;
+            c2=*s2;
+            if(c1!=c2) {
+                break;
+            }
+            ++s1;
+            ++s2;
+        }
+
+        /* setup for fix-up */
+        limit1=start1+length1;
+        limit2=start2+length2;
+    }
+
+    /* if both values are in or above the surrogate range, fix them up */
+    if(c1>=0xd800 && c2>=0xd800 && codePointOrder) {
+        /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
+        if(
+            (c1<=0xdbff && (s1+1)!=limit1 && U16_IS_TRAIL(*(s1+1))) ||
+            (U16_IS_TRAIL(c1) && start1!=s1 && U16_IS_LEAD(*(s1-1)))
+        ) {
+            /* part of a surrogate pair, leave >=d800 */
+        } else {
+            /* BMP code point - may be surrogate code point - make <d800 */
+            c1-=0x2800;
+        }
+
+        if(
+            (c2<=0xdbff && (s2+1)!=limit2 && U16_IS_TRAIL(*(s2+1))) ||
+            (U16_IS_TRAIL(c2) && start2!=s2 && U16_IS_LEAD(*(s2-1)))
+        ) {
+            /* part of a surrogate pair, leave >=d800 */
+        } else {
+            /* BMP code point - may be surrogate code point - make <d800 */
+            c2-=0x2800;
+        }
+    }
+
+    /* now c1 and c2 are in the requested (code unit or code point) order */
+    return (int32_t)c1-(int32_t)c2;
+}
+
+/*
+ * Compare two strings as presented by UCharIterators.
+ * Use code unit or code point order.
+ * When the function returns, it is undefined where the iterators
+ * have stopped.
+ */
+U_CAPI int32_t U_EXPORT2
+u_strCompareIter(UCharIterator *iter1, UCharIterator *iter2, UBool codePointOrder) {
+    UChar32 c1, c2;
+
+    /* argument checking */
+    if(iter1==NULL || iter2==NULL) {
+        return 0; /* bad arguments */
+    }
+    if(iter1==iter2) {
+        return 0; /* identical iterators */
+    }
+
+    /* reset iterators to start? */
+    iter1->move(iter1, 0, UITER_START);
+    iter2->move(iter2, 0, UITER_START);
+
+    /* compare identical prefixes - they do not need to be fixed up */
+    for(;;) {
+        c1=iter1->next(iter1);
+        c2=iter2->next(iter2);
+        if(c1!=c2) {
+            break;
+        }
+        if(c1==-1) {
+            return 0;
+        }
+    }
+
+    /* if both values are in or above the surrogate range, fix them up */
+    if(c1>=0xd800 && c2>=0xd800 && codePointOrder) {
+        /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
+        if(
+            (c1<=0xdbff && U16_IS_TRAIL(iter1->current(iter1))) ||
+            (U16_IS_TRAIL(c1) && (iter1->previous(iter1), U16_IS_LEAD(iter1->previous(iter1))))
+        ) {
+            /* part of a surrogate pair, leave >=d800 */
+        } else {
+            /* BMP code point - may be surrogate code point - make <d800 */
+            c1-=0x2800;
+        }
+
+        if(
+            (c2<=0xdbff && U16_IS_TRAIL(iter2->current(iter2))) ||
+            (U16_IS_TRAIL(c2) && (iter2->previous(iter2), U16_IS_LEAD(iter2->previous(iter2))))
+        ) {
+            /* part of a surrogate pair, leave >=d800 */
+        } else {
+            /* BMP code point - may be surrogate code point - make <d800 */
+            c2-=0x2800;
+        }
+    }
+
+    /* now c1 and c2 are in the requested (code unit or code point) order */
+    return (int32_t)c1-(int32_t)c2;
+}
+
+#if 0
+/*
+ * u_strCompareIter() does not leave the iterators _on_ the different units.
+ * This is possible but would cost a few extra indirect function calls to back
+ * up if the last unit (c1 or c2 respectively) was >=0.
+ *
+ * Consistently leaving them _behind_ the different units is not an option
+ * because the current "unit" is the end of the string if that is reached,
+ * and in such a case the iterator does not move.
+ * For example, when comparing "ab" with "abc", both iterators rest _on_ the end
+ * of their strings. Calling previous() on each does not move them to where
+ * the comparison fails.
+ *
+ * So the simplest semantics is to not define where the iterators end up.
+ *
+ * The following fragment is part of what would need to be done for backing up.
+ */
+void fragment {
+        /* iff a surrogate is part of a surrogate pair, leave >=d800 */
+        if(c1<=0xdbff) {
+            if(!U16_IS_TRAIL(iter1->current(iter1))) {
+                /* lead surrogate code point - make <d800 */
+                c1-=0x2800;
+            }
+        } else if(c1<=0xdfff) {
+            int32_t idx=iter1->getIndex(iter1, UITER_CURRENT);
+            iter1->previous(iter1); /* ==c1 */
+            if(!U16_IS_LEAD(iter1->previous(iter1))) {
+                /* trail surrogate code point - make <d800 */
+                c1-=0x2800;
+            }
+            /* go back to behind where the difference is */
+            iter1->move(iter1, idx, UITER_ZERO);
+        } else /* 0xe000<=c1<=0xffff */ {
+            /* BMP code point - make <d800 */
+            c1-=0x2800;
+        }
+}
+#endif
+
+U_CAPI int32_t U_EXPORT2
+u_strCompare(const UChar *s1, int32_t length1,
+             const UChar *s2, int32_t length2,
+             UBool codePointOrder) {
+    /* argument checking */
+    if(s1==NULL || length1<-1 || s2==NULL || length2<-1) {
+        return 0;
+    }
+    return uprv_strCompare(s1, length1, s2, length2, FALSE, codePointOrder);
+}
+
+/* String compare in code point order - u_strcmp() compares in code unit order. */
+U_CAPI int32_t U_EXPORT2
+u_strcmpCodePointOrder(const UChar *s1, const UChar *s2) {
+    return uprv_strCompare(s1, -1, s2, -1, FALSE, TRUE);
+}
+
+U_CAPI int32_t   U_EXPORT2
+u_strncmp(const UChar     *s1, 
+     const UChar     *s2, 
+     int32_t     n) 
+{
+    if(n > 0) {
+        int32_t rc;
+        for(;;) {
+            rc = (int32_t)*s1 - (int32_t)*s2;
+            if(rc != 0 || *s1 == 0 || --n == 0) {
+                return rc;
+            }
+            ++s1;
+            ++s2;
+        }
+    } else {
+        return 0;
+    }
+}
+
+U_CAPI int32_t U_EXPORT2
+u_strncmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t n) {
+    return uprv_strCompare(s1, n, s2, n, TRUE, TRUE);
+}
+
+U_CAPI UChar* U_EXPORT2
+u_strcpy(UChar     *dst, 
+    const UChar     *src) 
+{
+    UChar *anchor = dst;            /* save a pointer to start of dst */
+
+    while((*(dst++) = *(src++)) != 0) {     /* copy string 2 over              */
+    }
+
+    return anchor;
+}
+
+U_CAPI UChar*  U_EXPORT2
+u_strncpy(UChar     *dst, 
+     const UChar     *src, 
+     int32_t     n) 
+{
+    UChar *anchor = dst;            /* save a pointer to start of dst */
+
+    /* copy string 2 over */
+    while(n > 0 && (*(dst++) = *(src++)) != 0) {
+        --n;
+    }
+
+    return anchor;
+}
+
+U_CAPI int32_t   U_EXPORT2
+u_strlen(const UChar *s) 
+{
+#if U_SIZEOF_WCHAR_T == U_SIZEOF_UCHAR
+    return (int32_t)uprv_wcslen((const wchar_t *)s);
+#else
+    const UChar *t = s;
+    while(*t != 0) {
+      ++t;
+    }
+    return t - s;
+#endif
+}
+
+U_CAPI int32_t U_EXPORT2
+u_countChar32(const UChar *s, int32_t length) {
+    int32_t count;
+
+    if(s==NULL || length<-1) {
+        return 0;
+    }
+
+    count=0;
+    if(length>=0) {
+        while(length>0) {
+            ++count;
+            if(U16_IS_LEAD(*s) && length>=2 && U16_IS_TRAIL(*(s+1))) {
+                s+=2;
+                length-=2;
+            } else {
+                ++s;
+                --length;
+            }
+        }
+    } else /* length==-1 */ {
+        UChar c;
+
+        for(;;) {
+            if((c=*s++)==0) {
+                break;
+            }
+            ++count;
+
+            /*
+             * sufficient to look ahead one because of UTF-16;
+             * safe to look ahead one because at worst that would be the terminating NUL
+             */
+            if(U16_IS_LEAD(c) && U16_IS_TRAIL(*s)) {
+                ++s;
+            }
+        }
+    }
+    return count;
+}
+
+U_CAPI UBool U_EXPORT2
+u_strHasMoreChar32Than(const UChar *s, int32_t length, int32_t number) {
+
+    if(number<0) {
+        return TRUE;
+    }
+    if(s==NULL || length<-1) {
+        return FALSE;
+    }
+
+    if(length==-1) {
+        /* s is NUL-terminated */
+        UChar c;
+
+        /* count code points until they exceed */
+        for(;;) {
+            if((c=*s++)==0) {
+                return FALSE;
+            }
+            if(number==0) {
+                return TRUE;
+            }
+            if(U16_IS_LEAD(c) && U16_IS_TRAIL(*s)) {
+                ++s;
+            }
+            --number;
+        }
+    } else {
+        /* length>=0 known */
+        const UChar *limit;
+        int32_t maxSupplementary;
+
+        /* s contains at least (length+1)/2 code points: <=2 UChars per cp */
+        if(((length+1)/2)>number) {
+            return TRUE;
+        }
+
+        /* check if s does not even contain enough UChars */
+        maxSupplementary=length-number;
+        if(maxSupplementary<=0) {
+            return FALSE;
+        }
+        /* there are maxSupplementary=length-number more UChars than asked-for code points */
+
+        /*
+         * count code points until they exceed and also check that there are
+         * no more than maxSupplementary supplementary code points (UChar pairs)
+         */
+        limit=s+length;
+        for(;;) {
+            if(s==limit) {
+                return FALSE;
+            }
+            if(number==0) {
+                return TRUE;
+            }
+            if(U16_IS_LEAD(*s++) && s!=limit && U16_IS_TRAIL(*s)) {
+                ++s;
+                if(--maxSupplementary<=0) {
+                    /* too many pairs - too few code points */
+                    return FALSE;
+                }
+            }
+            --number;
+        }
+    }
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memcpy(UChar *dest, const UChar *src, int32_t count) {
+    if(count > 0) {
+        uprv_memcpy(dest, src, (size_t)count*U_SIZEOF_UCHAR);
+    }
+    return dest;
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memmove(UChar *dest, const UChar *src, int32_t count) {
+    if(count > 0) {
+        uprv_memmove(dest, src, (size_t)count*U_SIZEOF_UCHAR);
+    }
+    return dest;
+}
+
+U_CAPI UChar * U_EXPORT2
+u_memset(UChar *dest, UChar c, int32_t count) {
+    if(count > 0) {
+        UChar *ptr = dest;
+        UChar *limit = dest + count;
+
+        while (ptr < limit) {
+            *(ptr++) = c;
+        }
+    }
+    return dest;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_memcmp(const UChar *buf1, const UChar *buf2, int32_t count) {
+    if(count > 0) {
+        const UChar *limit = buf1 + count;
+        int32_t result;
+
+        while (buf1 < limit) {
+            result = (int32_t)(uint16_t)*buf1 - (int32_t)(uint16_t)*buf2;
+            if (result != 0) {
+                return result;
+            }
+            buf1++;
+            buf2++;
+        }
+    }
+    return 0;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_memcmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t count) {
+    return uprv_strCompare(s1, count, s2, count, FALSE, TRUE);
+}
+
+/* u_unescape & support fns ------------------------------------------------- */
+
+/* This map must be in ASCENDING ORDER OF THE ESCAPE CODE */
+static const UChar UNESCAPE_MAP[] = {
+    /*"   0x22, 0x22 */
+    /*'   0x27, 0x27 */
+    /*?   0x3F, 0x3F */
+    /*\   0x5C, 0x5C */
+    /*a*/ 0x61, 0x07,
+    /*b*/ 0x62, 0x08,
+    /*e*/ 0x65, 0x1b,
+    /*f*/ 0x66, 0x0c,
+    /*n*/ 0x6E, 0x0a,
+    /*r*/ 0x72, 0x0d,
+    /*t*/ 0x74, 0x09,
+    /*v*/ 0x76, 0x0b
+};
+enum { UNESCAPE_MAP_LENGTH = UPRV_LENGTHOF(UNESCAPE_MAP) };
+
+/* Convert one octal digit to a numeric value 0..7, or -1 on failure */
+static int8_t _digit8(UChar c) {
+    if (c >= 0x0030 && c <= 0x0037) {
+        return (int8_t)(c - 0x0030);
+    }
+    return -1;
+}
+
+/* Convert one hex digit to a numeric value 0..F, or -1 on failure */
+static int8_t _digit16(UChar c) {
+    if (c >= 0x0030 && c <= 0x0039) {
+        return (int8_t)(c - 0x0030);
+    }
+    if (c >= 0x0041 && c <= 0x0046) {
+        return (int8_t)(c - (0x0041 - 10));
+    }
+    if (c >= 0x0061 && c <= 0x0066) {
+        return (int8_t)(c - (0x0061 - 10));
+    }
+    return -1;
+}
+
+/* Parse a single escape sequence.  Although this method deals in
+ * UChars, it does not use C++ or UnicodeString.  This allows it to
+ * be used from C contexts. */
+U_CAPI UChar32 U_EXPORT2
+u_unescapeAt(UNESCAPE_CHAR_AT charAt,
+             int32_t *offset,
+             int32_t length,
+             void *context) {
+
+    int32_t start = *offset;
+    UChar c;
+    UChar32 result = 0;
+    int8_t n = 0;
+    int8_t minDig = 0;
+    int8_t maxDig = 0;
+    int8_t bitsPerDigit = 4; 
+    int8_t dig;
+    int32_t i;
+    UBool braces = FALSE;
+
+    /* Check that offset is in range */
+    if (*offset < 0 || *offset >= length) {
+        goto err;
+    }
+
+    /* Fetch first UChar after '\\' */
+    c = charAt((*offset)++, context);
+
+    /* Convert hexadecimal and octal escapes */
+    switch (c) {
+    case 0x0075 /*'u'*/:
+        minDig = maxDig = 4;
+        break;
+    case 0x0055 /*'U'*/:
+        minDig = maxDig = 8;
+        break;
+    case 0x0078 /*'x'*/:
+        minDig = 1;
+        if (*offset < length && charAt(*offset, context) == 0x7B /*{*/) {
+            ++(*offset);
+            braces = TRUE;
+            maxDig = 8;
+        } else {
+            maxDig = 2;
+        }
+        break;
+    default:
+        dig = _digit8(c);
+        if (dig >= 0) {
+            minDig = 1;
+            maxDig = 3;
+            n = 1; /* Already have first octal digit */
+            bitsPerDigit = 3;
+            result = dig;
+        }
+        break;
+    }
+    if (minDig != 0) {
+        while (*offset < length && n < maxDig) {
+            c = charAt(*offset, context);
+            dig = (int8_t)((bitsPerDigit == 3) ? _digit8(c) : _digit16(c));
+            if (dig < 0) {
+                break;
+            }
+            result = (result << bitsPerDigit) | dig;
+            ++(*offset);
+            ++n;
+        }
+        if (n < minDig) {
+            goto err;
+        }
+        if (braces) {
+            if (c != 0x7D /*}*/) {
+                goto err;
+            }
+            ++(*offset);
+        }
+        if (result < 0 || result >= 0x110000) {
+            goto err;
+        }
+        /* If an escape sequence specifies a lead surrogate, see if
+         * there is a trail surrogate after it, either as an escape or
+         * as a literal.  If so, join them up into a supplementary.
+         */
+        if (*offset < length && U16_IS_LEAD(result)) {
+            int32_t ahead = *offset + 1;
+            c = charAt(*offset, context);
+            if (c == 0x5C /*'\\'*/ && ahead < length) {
+                // Calling u_unescapeAt recursively may cause a stack overflow if
+                // we have repeated surrogate lead after that. Limit the
+                // length to 5 ('u' and 4 hex) after ahead.
+                int32_t tailLimit = ahead + 5;
+                if (tailLimit > length) {
+                    tailLimit = length;
+                }
+                c = (UChar) u_unescapeAt(charAt, &ahead, tailLimit,
+                                         context);
+            }
+            if (U16_IS_TRAIL(c)) {
+                *offset = ahead;
+                result = U16_GET_SUPPLEMENTARY(result, c);
+            }
+        }
+        return result;
+    }
+
+    /* Convert C-style escapes in table */
+    for (i=0; i<UNESCAPE_MAP_LENGTH; i+=2) {
+        if (c == UNESCAPE_MAP[i]) {
+            return UNESCAPE_MAP[i+1];
+        } else if (c < UNESCAPE_MAP[i]) {
+            break;
+        }
+    }
+
+    /* Map \cX to control-X: X & 0x1F */
+    if (c == 0x0063 /*'c'*/ && *offset < length) {
+        c = charAt((*offset)++, context);
+        if (U16_IS_LEAD(c) && *offset < length) {
+            UChar c2 = charAt(*offset, context);
+            if (U16_IS_TRAIL(c2)) {
+                ++(*offset);
+                c = (UChar) U16_GET_SUPPLEMENTARY(c, c2); /* [sic] */
+            }
+        }
+        return 0x1F & c;
+    }
+
+    /* If no special forms are recognized, then consider
+     * the backslash to generically escape the next character.
+     * Deal with surrogate pairs. */
+    if (U16_IS_LEAD(c) && *offset < length) {
+        UChar c2 = charAt(*offset, context);
+        if (U16_IS_TRAIL(c2)) {
+            ++(*offset);
+            return U16_GET_SUPPLEMENTARY(c, c2);
+        }
+    }
+    return c;
+
+ err:
+    /* Invalid escape sequence */
+    *offset = start; /* Reset to initial value */
+    return (UChar32)0xFFFFFFFF;
+}
+
+/* u_unescapeAt() callback to return a UChar from a char* */
+static UChar U_CALLCONV
+_charPtr_charAt(int32_t offset, void *context) {
+    UChar c16;
+    /* It would be more efficient to access the invariant tables
+     * directly but there is no API for that. */
+    u_charsToUChars(((char*) context) + offset, &c16, 1);
+    return c16;
+}
+
+/* Append an escape-free segment of the text; used by u_unescape() */
+static void _appendUChars(UChar *dest, int32_t destCapacity,
+                          const char *src, int32_t srcLen) {
+    if (destCapacity < 0) {
+        destCapacity = 0;
+    }
+    if (srcLen > destCapacity) {
+        srcLen = destCapacity;
+    }
+    u_charsToUChars(src, dest, srcLen);
+}
+
+/* Do an invariant conversion of char* -> UChar*, with escape parsing */
+U_CAPI int32_t U_EXPORT2
+u_unescape(const char *src, UChar *dest, int32_t destCapacity) {
+    const char *segment = src;
+    int32_t i = 0;
+    char c;
+
+    while ((c=*src) != 0) {
+        /* '\\' intentionally written as compiler-specific
+         * character constant to correspond to compiler-specific
+         * char* constants. */
+        if (c == '\\') {
+            int32_t lenParsed = 0;
+            UChar32 c32;
+            if (src != segment) {
+                if (dest != NULL) {
+                    _appendUChars(dest + i, destCapacity - i,
+                                  segment, (int32_t)(src - segment));
+                }
+                i += (int32_t)(src - segment);
+            }
+            ++src; /* advance past '\\' */
+            c32 = (UChar32)u_unescapeAt(_charPtr_charAt, &lenParsed, (int32_t)uprv_strlen(src), (void*)src);
+            if (lenParsed == 0) {
+                goto err;
+            }
+            src += lenParsed; /* advance past escape seq. */
+            if (dest != NULL && U16_LENGTH(c32) <= (destCapacity - i)) {
+                U16_APPEND_UNSAFE(dest, i, c32);
+            } else {
+                i += U16_LENGTH(c32);
+            }
+            segment = src;
+        } else {
+            ++src;
+        }
+    }
+    if (src != segment) {
+        if (dest != NULL) {
+            _appendUChars(dest + i, destCapacity - i,
+                          segment, (int32_t)(src - segment));
+        }
+        i += (int32_t)(src - segment);
+    }
+    if (dest != NULL && i < destCapacity) {
+        dest[i] = 0;
+    }
+    return i;
+
+ err:
+    if (dest != NULL && destCapacity > 0) {
+        *dest = 0;
+    }
+    return 0;
+}
+
+/* NUL-termination of strings ----------------------------------------------- */
+
+/**
+ * NUL-terminate a string no matter what its type.
+ * Set warning and error codes accordingly.
+ */
+#define __TERMINATE_STRING(dest, destCapacity, length, pErrorCode) UPRV_BLOCK_MACRO_BEGIN { \
+    if(pErrorCode!=NULL && U_SUCCESS(*pErrorCode)) {                    \
+        /* not a public function, so no complete argument checking */   \
+                                                                        \
+        if(length<0) {                                                  \
+            /* assume that the caller handles this */                   \
+        } else if(length<destCapacity) {                                \
+            /* NUL-terminate the string, the NUL fits */                \
+            dest[length]=0;                                             \
+            /* unset the not-terminated warning but leave all others */ \
+            if(*pErrorCode==U_STRING_NOT_TERMINATED_WARNING) {          \
+                *pErrorCode=U_ZERO_ERROR;                               \
+            }                                                           \
+        } else if(length==destCapacity) {                               \
+            /* unable to NUL-terminate, but the string itself fit - set a warning code */ \
+            *pErrorCode=U_STRING_NOT_TERMINATED_WARNING;                \
+        } else /* length>destCapacity */ {                              \
+            /* even the string itself did not fit - set an error code */ \
+            *pErrorCode=U_BUFFER_OVERFLOW_ERROR;                        \
+        }                                                               \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+U_CAPI UChar U_EXPORT2
+u_asciiToUpper(UChar c) {
+    if (u'a' <= c && c <= u'z') {
+        c = c + u'A' - u'a';
+    }
+    return c;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_terminateUChars(UChar *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
+    __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
+    return length;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_terminateChars(char *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
+    __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
+    return length;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_terminateUChar32s(UChar32 *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
+    __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
+    return length;
+}
+
+U_CAPI int32_t U_EXPORT2
+u_terminateWChars(wchar_t *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
+    __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
+    return length;
+}
+
+// Compute the hash code for a string -------------------------------------- ***
+
+// Moved here from uhash.c so that UnicodeString::hashCode() does not depend
+// on UHashtable code.
+
+/*
+  Compute the hash by iterating sparsely over about 32 (up to 63)
+  characters spaced evenly through the string.  For each character,
+  multiply the previous hash value by a prime number and add the new
+  character in, like a linear congruential random number generator,
+  producing a pseudorandom deterministic value well distributed over
+  the output range. [LIU]
+*/
+
+#define STRING_HASH(TYPE, STR, STRLEN, DEREF) UPRV_BLOCK_MACRO_BEGIN { \
+    uint32_t hash = 0;                        \
+    const TYPE *p = (const TYPE*) STR;        \
+    if (p != NULL) {                          \
+        int32_t len = (int32_t)(STRLEN);      \
+        int32_t inc = ((len - 32) / 32) + 1;  \
+        const TYPE *limit = p + len;          \
+        while (p<limit) {                     \
+            hash = (hash * 37) + DEREF;       \
+            p += inc;                         \
+        }                                     \
+    }                                         \
+    return static_cast<int32_t>(hash);        \
+} UPRV_BLOCK_MACRO_END
+
+/* Used by UnicodeString to compute its hashcode - Not public API. */
+U_CAPI int32_t U_EXPORT2
+ustr_hashUCharsN(const UChar *str, int32_t length) {
+    STRING_HASH(UChar, str, length, *p);
+}
+
+U_CAPI int32_t U_EXPORT2
+ustr_hashCharsN(const char *str, int32_t length) {
+    STRING_HASH(uint8_t, str, length, *p);
+}
+
+U_CAPI int32_t U_EXPORT2
+ustr_hashICharsN(const char *str, int32_t length) {
+    STRING_HASH(char, str, length, (uint8_t)uprv_tolower(*p));
+}
diff --git a/thirdparty/icu4c/common/ustrtrns.cpp b/thirdparty/icu4c/common/ustrtrns.cpp
new file mode 100644
index 0000000000..5dc032c02f
--- /dev/null
+++ b/thirdparty/icu4c/common/ustrtrns.cpp
@@ -0,0 +1,1451 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+* File ustrtrns.cpp
+*
+* Modification History:
+*
+*   Date        Name        Description
+*   9/10/2001    Ram    Creation.
+******************************************************************************
+*/
+
+/*******************************************************************************
+ *
+ * u_strTo* and u_strFrom* APIs
+ * WCS functions moved to ustr_wcs.c for better modularization
+ *
+ *******************************************************************************
+ */
+
+
+#include "unicode/putil.h"
+#include "unicode/ustring.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cstring.h"
+#include "cmemory.h"
+#include "ustr_imp.h"
+#include "uassert.h"
+
+U_CAPI UChar* U_EXPORT2 
+u_strFromUTF32WithSub(UChar *dest,
+               int32_t destCapacity,
+               int32_t *pDestLength,
+               const UChar32 *src,
+               int32_t srcLength,
+               UChar32 subchar, int32_t *pNumSubstitutions,
+               UErrorCode *pErrorCode) {
+    const UChar32 *srcLimit;
+    UChar32 ch;
+    UChar *destLimit;
+    UChar *pDest;
+    int32_t reqLength;
+    int32_t numSubstitutions;
+
+    /* args check */
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0) ||
+        subchar > 0x10ffff || U_IS_SURROGATE(subchar)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(pNumSubstitutions != NULL) {
+        *pNumSubstitutions = 0;
+    }
+
+    pDest = dest;
+    destLimit = (dest!=NULL)?(dest + destCapacity):NULL;
+    reqLength = 0;
+    numSubstitutions = 0;
+
+    if(srcLength < 0) {
+        /* simple loop for conversion of a NUL-terminated BMP string */
+        while((ch=*src) != 0 &&
+              ((uint32_t)ch < 0xd800 || (0xe000 <= ch && ch <= 0xffff))) {
+            ++src;
+            if(pDest < destLimit) {
+                *pDest++ = (UChar)ch;
+            } else {
+                ++reqLength;
+            }
+        }
+        srcLimit = src;
+        if(ch != 0) {
+            /* "complicated" case, find the end of the remaining string */
+            while(*++srcLimit != 0) {}
+        }
+    } else {
+      srcLimit = (src!=NULL)?(src + srcLength):NULL;
+    }
+
+    /* convert with length */
+    while(src < srcLimit) {
+        ch = *src++;
+        do {
+            /* usually "loops" once; twice only for writing subchar */
+            if((uint32_t)ch < 0xd800 || (0xe000 <= ch && ch <= 0xffff)) {
+                if(pDest < destLimit) {
+                    *pDest++ = (UChar)ch;
+                } else {
+                    ++reqLength;
+                }
+                break;
+            } else if(0x10000 <= ch && ch <= 0x10ffff) {
+                if(pDest!=NULL && ((pDest + 2) <= destLimit)) {
+                    *pDest++ = U16_LEAD(ch);
+                    *pDest++ = U16_TRAIL(ch);
+                } else {
+                    reqLength += 2;
+                }
+                break;
+            } else if((ch = subchar) < 0) {
+                /* surrogate code point, or not a Unicode code point at all */
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                return NULL;
+            } else {
+                ++numSubstitutions;
+            }
+        } while(TRUE);
+    }
+
+    reqLength += (int32_t)(pDest - dest);
+    if(pDestLength) {
+        *pDestLength = reqLength;
+    }
+    if(pNumSubstitutions != NULL) {
+        *pNumSubstitutions = numSubstitutions;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChars(dest, destCapacity, reqLength, pErrorCode);
+    
+    return dest;
+}
+
+U_CAPI UChar* U_EXPORT2 
+u_strFromUTF32(UChar *dest,
+               int32_t destCapacity, 
+               int32_t *pDestLength,
+               const UChar32 *src,
+               int32_t srcLength,
+               UErrorCode *pErrorCode) {
+    return u_strFromUTF32WithSub(
+            dest, destCapacity, pDestLength,
+            src, srcLength,
+            U_SENTINEL, NULL,
+            pErrorCode);
+}
+
+U_CAPI UChar32* U_EXPORT2 
+u_strToUTF32WithSub(UChar32 *dest,
+             int32_t destCapacity,
+             int32_t *pDestLength,
+             const UChar *src,
+             int32_t srcLength,
+             UChar32 subchar, int32_t *pNumSubstitutions,
+             UErrorCode *pErrorCode) {
+    const UChar *srcLimit;
+    UChar32 ch;
+    UChar ch2;
+    UChar32 *destLimit;
+    UChar32 *pDest;
+    int32_t reqLength;
+    int32_t numSubstitutions;
+
+    /* args check */
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0) ||
+        subchar > 0x10ffff || U_IS_SURROGATE(subchar)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(pNumSubstitutions != NULL) {
+        *pNumSubstitutions = 0;
+    }
+
+    pDest = dest;
+    destLimit = (dest!=NULL)?(dest + destCapacity):NULL;
+    reqLength = 0;
+    numSubstitutions = 0;
+
+    if(srcLength < 0) {
+        /* simple loop for conversion of a NUL-terminated BMP string */
+        while((ch=*src) != 0 && !U16_IS_SURROGATE(ch)) {
+            ++src;
+            if(pDest < destLimit) {
+                *pDest++ = ch;
+            } else {
+                ++reqLength;
+            }
+        }
+        srcLimit = src;
+        if(ch != 0) {
+            /* "complicated" case, find the end of the remaining string */
+            while(*++srcLimit != 0) {}
+        }
+    } else {
+        srcLimit = (src!=NULL)?(src + srcLength):NULL;
+    }
+
+    /* convert with length */
+    while(src < srcLimit) {
+        ch = *src++;
+        if(!U16_IS_SURROGATE(ch)) {
+            /* write or count ch below */
+        } else if(U16_IS_SURROGATE_LEAD(ch) && src < srcLimit && U16_IS_TRAIL(ch2 = *src)) {
+            ++src;
+            ch = U16_GET_SUPPLEMENTARY(ch, ch2);
+        } else if((ch = subchar) < 0) {
+            /* unpaired surrogate */
+            *pErrorCode = U_INVALID_CHAR_FOUND;
+            return NULL;
+        } else {
+            ++numSubstitutions;
+        }
+        if(pDest < destLimit) {
+            *pDest++ = ch;
+        } else {
+            ++reqLength;
+        }
+    }
+
+    reqLength += (int32_t)(pDest - dest);
+    if(pDestLength) {
+        *pDestLength = reqLength;
+    }
+    if(pNumSubstitutions != NULL) {
+        *pNumSubstitutions = numSubstitutions;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChar32s(dest, destCapacity, reqLength, pErrorCode);
+
+    return dest;
+}
+
+U_CAPI UChar32* U_EXPORT2 
+u_strToUTF32(UChar32 *dest, 
+             int32_t destCapacity,
+             int32_t *pDestLength,
+             const UChar *src, 
+             int32_t srcLength,
+             UErrorCode *pErrorCode) {
+    return u_strToUTF32WithSub(
+            dest, destCapacity, pDestLength,
+            src, srcLength,
+            U_SENTINEL, NULL,
+            pErrorCode);
+}
+
+U_CAPI UChar* U_EXPORT2
+u_strFromUTF8WithSub(UChar *dest,
+              int32_t destCapacity,
+              int32_t *pDestLength,
+              const char* src,
+              int32_t srcLength,
+              UChar32 subchar, int32_t *pNumSubstitutions,
+              UErrorCode *pErrorCode){
+    /* args check */
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0) ||
+        subchar > 0x10ffff || U_IS_SURROGATE(subchar)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=0;
+    }
+    UChar *pDest = dest;
+    UChar *pDestLimit = dest+destCapacity;
+    int32_t reqLength = 0;
+    int32_t numSubstitutions=0;
+
+    /*
+     * Inline processing of UTF-8 byte sequences:
+     *
+     * Byte sequences for the most common characters are handled inline in
+     * the conversion loops. In order to reduce the path lengths for those
+     * characters, the tests are arranged in a kind of binary search.
+     * ASCII (<=0x7f) is checked first, followed by the dividing point
+     * between 2- and 3-byte sequences (0xe0).
+     * The 3-byte branch is tested first to speed up CJK text.
+     * The compiler should combine the subtractions for the two tests for 0xe0.
+     * Each branch then tests for the other end of its range.
+     */
+
+    if(srcLength < 0){
+        /*
+         * Transform a NUL-terminated string.
+         * The code explicitly checks for NULs only in the lead byte position.
+         * A NUL byte in the trail byte position fails the trail byte range check anyway.
+         */
+        int32_t i;
+        UChar32 c;
+        for(i = 0; (c = (uint8_t)src[i]) != 0 && (pDest < pDestLimit);) {
+            // modified copy of U8_NEXT()
+            ++i;
+            if(U8_IS_SINGLE(c)) {
+                *pDest++=(UChar)c;
+            } else {
+                uint8_t __t1, __t2;
+                if( /* handle U+0800..U+FFFF inline */
+                        (0xe0<=(c) && (c)<0xf0) &&
+                        U8_IS_VALID_LEAD3_AND_T1((c), src[i]) &&
+                        (__t2=src[(i)+1]-0x80)<=0x3f) {
+                    *pDest++ = (((c)&0xf)<<12)|((src[i]&0x3f)<<6)|__t2;
+                    i+=2;
+                } else if( /* handle U+0080..U+07FF inline */
+                        ((c)<0xe0 && (c)>=0xc2) &&
+                        (__t1=src[i]-0x80)<=0x3f) {
+                    *pDest++ = (((c)&0x1f)<<6)|__t1;
+                    ++(i);
+                } else {
+                    /* function call for "complicated" and error cases */
+                    (c)=utf8_nextCharSafeBody((const uint8_t *)src, &(i), -1, c, -1);
+                    if(c<0 && (++numSubstitutions, c = subchar) < 0) {
+                        *pErrorCode = U_INVALID_CHAR_FOUND;
+                        return NULL;
+                    } else if(c<=0xFFFF) {
+                        *(pDest++)=(UChar)c;
+                    } else {
+                        *(pDest++)=U16_LEAD(c);
+                        if(pDest<pDestLimit) {
+                            *(pDest++)=U16_TRAIL(c);
+                        } else {
+                            reqLength++;
+                            break;
+                        }
+                    }
+                }
+            }
+        }
+
+        /* Pre-flight the rest of the string. */
+        while((c = (uint8_t)src[i]) != 0) {
+            // modified copy of U8_NEXT()
+            ++i;
+            if(U8_IS_SINGLE(c)) {
+                ++reqLength;
+            } else {
+                uint8_t __t1, __t2;
+                if( /* handle U+0800..U+FFFF inline */
+                        (0xe0<=(c) && (c)<0xf0) &&
+                        U8_IS_VALID_LEAD3_AND_T1((c), src[i]) &&
+                        (__t2=src[(i)+1]-0x80)<=0x3f) {
+                    ++reqLength;
+                    i+=2;
+                } else if( /* handle U+0080..U+07FF inline */
+                        ((c)<0xe0 && (c)>=0xc2) &&
+                        (__t1=src[i]-0x80)<=0x3f) {
+                    ++reqLength;
+                    ++(i);
+                } else {
+                    /* function call for "complicated" and error cases */
+                    (c)=utf8_nextCharSafeBody((const uint8_t *)src, &(i), -1, c, -1);
+                    if(c<0 && (++numSubstitutions, c = subchar) < 0) {
+                        *pErrorCode = U_INVALID_CHAR_FOUND;
+                        return NULL;
+                    }
+                    reqLength += U16_LENGTH(c);
+                }
+            }
+        }
+    } else /* srcLength >= 0 */ {
+        /* Faster loop without ongoing checking for srcLength and pDestLimit. */
+        int32_t i = 0;
+        UChar32 c;
+        for(;;) {
+            /*
+             * Each iteration of the inner loop progresses by at most 3 UTF-8
+             * bytes and one UChar, for most characters.
+             * For supplementary code points (4 & 2), which are rare,
+             * there is an additional adjustment.
+             */
+            int32_t count = (int32_t)(pDestLimit - pDest);
+            int32_t count2 = (srcLength - i) / 3;
+            if(count > count2) {
+                count = count2; /* min(remaining dest, remaining src/3) */
+            }
+            if(count < 3) {
+                /*
+                 * Too much overhead if we get near the end of the string,
+                 * continue with the next loop.
+                 */
+                break;
+            }
+
+            do {
+                // modified copy of U8_NEXT()
+                c = (uint8_t)src[i++];
+                if(U8_IS_SINGLE(c)) {
+                    *pDest++=(UChar)c;
+                } else {
+                    uint8_t __t1, __t2;
+                    if( /* handle U+0800..U+FFFF inline */
+                            (0xe0<=(c) && (c)<0xf0) &&
+                            ((i)+1)<srcLength &&
+                            U8_IS_VALID_LEAD3_AND_T1((c), src[i]) &&
+                            (__t2=src[(i)+1]-0x80)<=0x3f) {
+                        *pDest++ = (((c)&0xf)<<12)|((src[i]&0x3f)<<6)|__t2;
+                        i+=2;
+                    } else if( /* handle U+0080..U+07FF inline */
+                            ((c)<0xe0 && (c)>=0xc2) &&
+                            ((i)!=srcLength) &&
+                            (__t1=src[i]-0x80)<=0x3f) {
+                        *pDest++ = (((c)&0x1f)<<6)|__t1;
+                        ++(i);
+                    } else {
+                        if(c >= 0xf0 || subchar > 0xffff) {
+                            // We may read up to four bytes and write up to two UChars,
+                            // which we didn't account for with computing count,
+                            // so we adjust it here.
+                            if(--count == 0) {
+                                --i;  // back out byte c
+                                break;
+                            }
+                        }
+
+                        /* function call for "complicated" and error cases */
+                        (c)=utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, c, -1);
+                        if(c<0 && (++numSubstitutions, c = subchar) < 0) {
+                            *pErrorCode = U_INVALID_CHAR_FOUND;
+                            return NULL;
+                        } else if(c<=0xFFFF) {
+                            *(pDest++)=(UChar)c;
+                        } else {
+                            *(pDest++)=U16_LEAD(c);
+                            *(pDest++)=U16_TRAIL(c);
+                        }
+                    }
+                }
+            } while(--count > 0);
+        }
+
+        while(i < srcLength && (pDest < pDestLimit)) {
+            // modified copy of U8_NEXT()
+            c = (uint8_t)src[i++];
+            if(U8_IS_SINGLE(c)) {
+                *pDest++=(UChar)c;
+            } else {
+                uint8_t __t1, __t2;
+                if( /* handle U+0800..U+FFFF inline */
+                        (0xe0<=(c) && (c)<0xf0) &&
+                        ((i)+1)<srcLength &&
+                        U8_IS_VALID_LEAD3_AND_T1((c), src[i]) &&
+                        (__t2=src[(i)+1]-0x80)<=0x3f) {
+                    *pDest++ = (((c)&0xf)<<12)|((src[i]&0x3f)<<6)|__t2;
+                    i+=2;
+                } else if( /* handle U+0080..U+07FF inline */
+                        ((c)<0xe0 && (c)>=0xc2) &&
+                        ((i)!=srcLength) &&
+                        (__t1=src[i]-0x80)<=0x3f) {
+                    *pDest++ = (((c)&0x1f)<<6)|__t1;
+                    ++(i);
+                } else {
+                    /* function call for "complicated" and error cases */
+                    (c)=utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, c, -1);
+                    if(c<0 && (++numSubstitutions, c = subchar) < 0) {
+                        *pErrorCode = U_INVALID_CHAR_FOUND;
+                        return NULL;
+                    } else if(c<=0xFFFF) {
+                        *(pDest++)=(UChar)c;
+                    } else {
+                        *(pDest++)=U16_LEAD(c);
+                        if(pDest<pDestLimit) {
+                            *(pDest++)=U16_TRAIL(c);
+                        } else {
+                            reqLength++;
+                            break;
+                        }
+                    }
+                }
+            }
+        }
+
+        /* Pre-flight the rest of the string. */
+        while(i < srcLength) {
+            // modified copy of U8_NEXT()
+            c = (uint8_t)src[i++];
+            if(U8_IS_SINGLE(c)) {
+                ++reqLength;
+            } else {
+                uint8_t __t1, __t2;
+                if( /* handle U+0800..U+FFFF inline */
+                        (0xe0<=(c) && (c)<0xf0) &&
+                        ((i)+1)<srcLength &&
+                        U8_IS_VALID_LEAD3_AND_T1((c), src[i]) &&
+                        (__t2=src[(i)+1]-0x80)<=0x3f) {
+                    ++reqLength;
+                    i+=2;
+                } else if( /* handle U+0080..U+07FF inline */
+                        ((c)<0xe0 && (c)>=0xc2) &&
+                        ((i)!=srcLength) &&
+                        (__t1=src[i]-0x80)<=0x3f) {
+                    ++reqLength;
+                    ++(i);
+                } else {
+                    /* function call for "complicated" and error cases */
+                    (c)=utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, c, -1);
+                    if(c<0 && (++numSubstitutions, c = subchar) < 0) {
+                        *pErrorCode = U_INVALID_CHAR_FOUND;
+                        return NULL;
+                    }
+                    reqLength += U16_LENGTH(c);
+                }
+            }
+        }
+    }
+
+    reqLength+=(int32_t)(pDest - dest);
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=numSubstitutions;
+    }
+
+    if(pDestLength){
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChars(dest,destCapacity,reqLength,pErrorCode);
+
+    return dest;
+}
+
+U_CAPI UChar* U_EXPORT2
+u_strFromUTF8(UChar *dest,
+              int32_t destCapacity,
+              int32_t *pDestLength,
+              const char* src,
+              int32_t srcLength,
+              UErrorCode *pErrorCode){
+    return u_strFromUTF8WithSub(
+            dest, destCapacity, pDestLength,
+            src, srcLength,
+            U_SENTINEL, NULL,
+            pErrorCode);
+}
+
+U_CAPI UChar * U_EXPORT2
+u_strFromUTF8Lenient(UChar *dest,
+                     int32_t destCapacity,
+                     int32_t *pDestLength,
+                     const char *src,
+                     int32_t srcLength,
+                     UErrorCode *pErrorCode) {
+    UChar *pDest = dest;
+    UChar32 ch;
+    int32_t reqLength = 0;
+    uint8_t* pSrc = (uint8_t*) src;
+
+    /* args check */
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+        
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(srcLength < 0) {
+        /* Transform a NUL-terminated string. */
+        UChar *pDestLimit = (dest!=NULL)?(dest+destCapacity):NULL;
+        uint8_t t1, t2, t3; /* trail bytes */
+
+        while(((ch = *pSrc) != 0) && (pDest < pDestLimit)) {
+            if(ch < 0xc0) {
+                /*
+                 * ASCII, or a trail byte in lead position which is treated like
+                 * a single-byte sequence for better character boundary
+                 * resynchronization after illegal sequences.
+                 */
+                *pDest++=(UChar)ch;
+                ++pSrc;
+                continue;
+            } else if(ch < 0xe0) { /* U+0080..U+07FF */
+                if((t1 = pSrc[1]) != 0) {
+                    /* 0x3080 = (0xc0 << 6) + 0x80 */
+                    *pDest++ = (UChar)((ch << 6) + t1 - 0x3080);
+                    pSrc += 2;
+                    continue;
+                }
+            } else if(ch < 0xf0) { /* U+0800..U+FFFF */
+                if((t1 = pSrc[1]) != 0 && (t2 = pSrc[2]) != 0) {
+                    /* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
+                    /* 0x2080 = (0x80 << 6) + 0x80 */
+                    *pDest++ = (UChar)((ch << 12) + (t1 << 6) + t2 - 0x2080);
+                    pSrc += 3;
+                    continue;
+                }
+            } else /* f0..f4 */ { /* U+10000..U+10FFFF */
+                if((t1 = pSrc[1]) != 0 && (t2 = pSrc[2]) != 0 && (t3 = pSrc[3]) != 0) {
+                    pSrc += 4;
+                    /* 0x3c82080 = (0xf0 << 18) + (0x80 << 12) + (0x80 << 6) + 0x80 */
+                    ch = (ch << 18) + (t1 << 12) + (t2 << 6) + t3 - 0x3c82080;
+                    *(pDest++) = U16_LEAD(ch);
+                    if(pDest < pDestLimit) {
+                        *(pDest++) = U16_TRAIL(ch);
+                    } else {
+                        reqLength = 1;
+                        break;
+                    }
+                    continue;
+                }
+            }
+
+            /* truncated character at the end */
+            *pDest++ = 0xfffd;
+            while(*++pSrc != 0) {}
+            break;
+        }
+
+        /* Pre-flight the rest of the string. */
+        while((ch = *pSrc) != 0) {
+            if(ch < 0xc0) {
+                /*
+                 * ASCII, or a trail byte in lead position which is treated like
+                 * a single-byte sequence for better character boundary
+                 * resynchronization after illegal sequences.
+                 */
+                ++reqLength;
+                ++pSrc;
+                continue;
+            } else if(ch < 0xe0) { /* U+0080..U+07FF */
+                if(pSrc[1] != 0) {
+                    ++reqLength;
+                    pSrc += 2;
+                    continue;
+                }
+            } else if(ch < 0xf0) { /* U+0800..U+FFFF */
+                if(pSrc[1] != 0 && pSrc[2] != 0) {
+                    ++reqLength;
+                    pSrc += 3;
+                    continue;
+                }
+            } else /* f0..f4 */ { /* U+10000..U+10FFFF */
+                if(pSrc[1] != 0 && pSrc[2] != 0 && pSrc[3] != 0) {
+                    reqLength += 2;
+                    pSrc += 4;
+                    continue;
+                }
+            }
+
+            /* truncated character at the end */
+            ++reqLength;
+            break;
+        }
+    } else /* srcLength >= 0 */ {
+      const uint8_t *pSrcLimit = (pSrc!=NULL)?(pSrc + srcLength):NULL;
+
+        /*
+         * This function requires that if srcLength is given, then it must be
+         * destCapatity >= srcLength so that we need not check for
+         * destination buffer overflow in the loop.
+         */
+        if(destCapacity < srcLength) {
+            if(pDestLength != NULL) {
+                *pDestLength = srcLength; /* this likely overestimates the true destLength! */
+            }
+            *pErrorCode = U_BUFFER_OVERFLOW_ERROR;
+            return NULL;
+        }
+
+        if((pSrcLimit - pSrc) >= 4) {
+            pSrcLimit -= 3; /* temporarily reduce pSrcLimit */
+
+            /* in this loop, we can always access at least 4 bytes, up to pSrc+3 */
+            do {
+                ch = *pSrc++;
+                if(ch < 0xc0) {
+                    /*
+                     * ASCII, or a trail byte in lead position which is treated like
+                     * a single-byte sequence for better character boundary
+                     * resynchronization after illegal sequences.
+                     */
+                    *pDest++=(UChar)ch;
+                } else if(ch < 0xe0) { /* U+0080..U+07FF */
+                    /* 0x3080 = (0xc0 << 6) + 0x80 */
+                    *pDest++ = (UChar)((ch << 6) + *pSrc++ - 0x3080);
+                } else if(ch < 0xf0) { /* U+0800..U+FFFF */
+                    /* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
+                    /* 0x2080 = (0x80 << 6) + 0x80 */
+                    ch = (ch << 12) + (*pSrc++ << 6);
+                    *pDest++ = (UChar)(ch + *pSrc++ - 0x2080);
+                } else /* f0..f4 */ { /* U+10000..U+10FFFF */
+                    /* 0x3c82080 = (0xf0 << 18) + (0x80 << 12) + (0x80 << 6) + 0x80 */
+                    ch = (ch << 18) + (*pSrc++ << 12);
+                    ch += *pSrc++ << 6;
+                    ch += *pSrc++ - 0x3c82080;
+                    *(pDest++) = U16_LEAD(ch);
+                    *(pDest++) = U16_TRAIL(ch);
+                }
+            } while(pSrc < pSrcLimit);
+
+            pSrcLimit += 3; /* restore original pSrcLimit */
+        }
+
+        while(pSrc < pSrcLimit) {
+            ch = *pSrc++;
+            if(ch < 0xc0) {
+                /*
+                 * ASCII, or a trail byte in lead position which is treated like
+                 * a single-byte sequence for better character boundary
+                 * resynchronization after illegal sequences.
+                 */
+                *pDest++=(UChar)ch;
+                continue;
+            } else if(ch < 0xe0) { /* U+0080..U+07FF */
+                if(pSrc < pSrcLimit) {
+                    /* 0x3080 = (0xc0 << 6) + 0x80 */
+                    *pDest++ = (UChar)((ch << 6) + *pSrc++ - 0x3080);
+                    continue;
+                }
+            } else if(ch < 0xf0) { /* U+0800..U+FFFF */
+                if((pSrcLimit - pSrc) >= 2) {
+                    /* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
+                    /* 0x2080 = (0x80 << 6) + 0x80 */
+                    ch = (ch << 12) + (*pSrc++ << 6);
+                    *pDest++ = (UChar)(ch + *pSrc++ - 0x2080);
+                    pSrc += 3;
+                    continue;
+                }
+            } else /* f0..f4 */ { /* U+10000..U+10FFFF */
+                if((pSrcLimit - pSrc) >= 3) {
+                    /* 0x3c82080 = (0xf0 << 18) + (0x80 << 12) + (0x80 << 6) + 0x80 */
+                    ch = (ch << 18) + (*pSrc++ << 12);
+                    ch += *pSrc++ << 6;
+                    ch += *pSrc++ - 0x3c82080;
+                    *(pDest++) = U16_LEAD(ch);
+                    *(pDest++) = U16_TRAIL(ch);
+                    pSrc += 4;
+                    continue;
+                }
+            }
+
+            /* truncated character at the end */
+            *pDest++ = 0xfffd;
+            break;
+        }
+    }
+
+    reqLength+=(int32_t)(pDest - dest);
+
+    if(pDestLength){
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChars(dest,destCapacity,reqLength,pErrorCode);
+
+    return dest;
+}
+
+static inline uint8_t *
+_appendUTF8(uint8_t *pDest, UChar32 c) {
+    /* it is 0<=c<=0x10ffff and not a surrogate if called by a validating function */
+    if((c)<=0x7f) {
+        *pDest++=(uint8_t)c;
+    } else if(c<=0x7ff) {
+        *pDest++=(uint8_t)((c>>6)|0xc0);
+        *pDest++=(uint8_t)((c&0x3f)|0x80);
+    } else if(c<=0xffff) {
+        *pDest++=(uint8_t)((c>>12)|0xe0);
+        *pDest++=(uint8_t)(((c>>6)&0x3f)|0x80);
+        *pDest++=(uint8_t)(((c)&0x3f)|0x80);
+    } else /* if((uint32_t)(c)<=0x10ffff) */ {
+        *pDest++=(uint8_t)(((c)>>18)|0xf0);
+        *pDest++=(uint8_t)((((c)>>12)&0x3f)|0x80);
+        *pDest++=(uint8_t)((((c)>>6)&0x3f)|0x80);
+        *pDest++=(uint8_t)(((c)&0x3f)|0x80);
+    }
+    return pDest;
+}
+
+   
+U_CAPI char* U_EXPORT2 
+u_strToUTF8WithSub(char *dest,
+            int32_t destCapacity,
+            int32_t *pDestLength,
+            const UChar *pSrc,
+            int32_t srcLength,
+            UChar32 subchar, int32_t *pNumSubstitutions,
+            UErrorCode *pErrorCode){
+    int32_t reqLength=0;
+    uint32_t ch=0,ch2=0;
+    uint8_t *pDest = (uint8_t *)dest;
+    uint8_t *pDestLimit = (pDest!=NULL)?(pDest + destCapacity):NULL;
+    int32_t numSubstitutions;
+
+    /* args check */
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+        
+    if( (pSrc==NULL && srcLength!=0) || srcLength < -1 ||
+        (destCapacity<0) || (dest == NULL && destCapacity > 0) ||
+        subchar > 0x10ffff || U_IS_SURROGATE(subchar)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=0;
+    }
+    numSubstitutions=0;
+
+    if(srcLength==-1) {
+        while((ch=*pSrc)!=0) {
+            ++pSrc;
+            if(ch <= 0x7f) {
+                if(pDest<pDestLimit) {
+                    *pDest++ = (uint8_t)ch;
+                } else {
+                    reqLength = 1;
+                    break;
+                }
+            } else if(ch <= 0x7ff) {
+                if((pDestLimit - pDest) >= 2) {
+                    *pDest++=(uint8_t)((ch>>6)|0xc0);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else {
+                    reqLength = 2;
+                    break;
+                }
+            } else if(ch <= 0xd7ff || ch >= 0xe000) {
+                if((pDestLimit - pDest) >= 3) {
+                    *pDest++=(uint8_t)((ch>>12)|0xe0);
+                    *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else {
+                    reqLength = 3;
+                    break;
+                }
+            } else /* ch is a surrogate */ {
+                int32_t length;
+
+                /*need not check for NUL because NUL fails U16_IS_TRAIL() anyway*/
+                if(U16_IS_SURROGATE_LEAD(ch) && U16_IS_TRAIL(ch2=*pSrc)) { 
+                    ++pSrc;
+                    ch=U16_GET_SUPPLEMENTARY(ch, ch2);
+                } else if(subchar>=0) {
+                    ch=subchar;
+                    ++numSubstitutions;
+                } else {
+                    /* Unicode 3.2 forbids surrogate code points in UTF-8 */
+                    *pErrorCode = U_INVALID_CHAR_FOUND;
+                    return NULL;
+                }
+
+                length = U8_LENGTH(ch);
+                if((pDestLimit - pDest) >= length) {
+                    /* convert and append*/
+                    pDest=_appendUTF8(pDest, ch);
+                } else {
+                    reqLength = length;
+                    break;
+                }
+            }
+        }
+        while((ch=*pSrc++)!=0) {
+            if(ch<=0x7f) {
+                ++reqLength;
+            } else if(ch<=0x7ff) {
+                reqLength+=2;
+            } else if(!U16_IS_SURROGATE(ch)) {
+                reqLength+=3;
+            } else if(U16_IS_SURROGATE_LEAD(ch) && U16_IS_TRAIL(ch2=*pSrc)) {
+                ++pSrc;
+                reqLength+=4;
+            } else if(subchar>=0) {
+                reqLength+=U8_LENGTH(subchar);
+                ++numSubstitutions;
+            } else {
+                /* Unicode 3.2 forbids surrogate code points in UTF-8 */
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                return NULL;
+            }
+        }
+    } else {
+        const UChar *pSrcLimit = (pSrc!=NULL)?(pSrc+srcLength):NULL;
+        int32_t count;
+
+        /* Faster loop without ongoing checking for pSrcLimit and pDestLimit. */
+        for(;;) {
+            /*
+             * Each iteration of the inner loop progresses by at most 3 UTF-8
+             * bytes and one UChar, for most characters.
+             * For supplementary code points (4 & 2), which are rare,
+             * there is an additional adjustment.
+             */
+            count = (int32_t)((pDestLimit - pDest) / 3);
+            srcLength = (int32_t)(pSrcLimit - pSrc);
+            if(count > srcLength) {
+                count = srcLength; /* min(remaining dest/3, remaining src) */
+            }
+            if(count < 3) {
+                /*
+                 * Too much overhead if we get near the end of the string,
+                 * continue with the next loop.
+                 */
+                break;
+            }
+            do {
+                ch=*pSrc++;
+                if(ch <= 0x7f) {
+                    *pDest++ = (uint8_t)ch;
+                } else if(ch <= 0x7ff) {
+                    *pDest++=(uint8_t)((ch>>6)|0xc0);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else if(ch <= 0xd7ff || ch >= 0xe000) {
+                    *pDest++=(uint8_t)((ch>>12)|0xe0);
+                    *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else /* ch is a surrogate */ {
+                    /*
+                     * We will read two UChars and probably output four bytes,
+                     * which we didn't account for with computing count,
+                     * so we adjust it here.
+                     */
+                    if(--count == 0) {
+                        --pSrc; /* undo ch=*pSrc++ for the lead surrogate */
+                        break;  /* recompute count */
+                    }
+
+                    if(U16_IS_SURROGATE_LEAD(ch) && U16_IS_TRAIL(ch2=*pSrc)) { 
+                        ++pSrc;
+                        ch=U16_GET_SUPPLEMENTARY(ch, ch2);
+
+                        /* writing 4 bytes per 2 UChars is ok */
+                        *pDest++=(uint8_t)((ch>>18)|0xf0);
+                        *pDest++=(uint8_t)(((ch>>12)&0x3f)|0x80);
+                        *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                        *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                    } else  {
+                        /* Unicode 3.2 forbids surrogate code points in UTF-8 */
+                        if(subchar>=0) {
+                            ch=subchar;
+                            ++numSubstitutions;
+                        } else {
+                            *pErrorCode = U_INVALID_CHAR_FOUND;
+                            return NULL;
+                        }
+
+                        /* convert and append*/
+                        pDest=_appendUTF8(pDest, ch);
+                    }
+                }
+            } while(--count > 0);
+        }
+
+        while(pSrc<pSrcLimit) {
+            ch=*pSrc++;
+            if(ch <= 0x7f) {
+                if(pDest<pDestLimit) {
+                    *pDest++ = (uint8_t)ch;
+                } else {
+                    reqLength = 1;
+                    break;
+                }
+            } else if(ch <= 0x7ff) {
+                if((pDestLimit - pDest) >= 2) {
+                    *pDest++=(uint8_t)((ch>>6)|0xc0);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else {
+                    reqLength = 2;
+                    break;
+                }
+            } else if(ch <= 0xd7ff || ch >= 0xe000) {
+                if((pDestLimit - pDest) >= 3) {
+                    *pDest++=(uint8_t)((ch>>12)|0xe0);
+                    *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                    *pDest++=(uint8_t)((ch&0x3f)|0x80);
+                } else {
+                    reqLength = 3;
+                    break;
+                }
+            } else /* ch is a surrogate */ {
+                int32_t length;
+
+                if(U16_IS_SURROGATE_LEAD(ch) && pSrc<pSrcLimit && U16_IS_TRAIL(ch2=*pSrc)) { 
+                    ++pSrc;
+                    ch=U16_GET_SUPPLEMENTARY(ch, ch2);
+                } else if(subchar>=0) {
+                    ch=subchar;
+                    ++numSubstitutions;
+                } else {
+                    /* Unicode 3.2 forbids surrogate code points in UTF-8 */
+                    *pErrorCode = U_INVALID_CHAR_FOUND;
+                    return NULL;
+                }
+
+                length = U8_LENGTH(ch);
+                if((pDestLimit - pDest) >= length) {
+                    /* convert and append*/
+                    pDest=_appendUTF8(pDest, ch);
+                } else {
+                    reqLength = length;
+                    break;
+                }
+            }
+        }
+        while(pSrc<pSrcLimit) {
+            ch=*pSrc++;
+            if(ch<=0x7f) {
+                ++reqLength;
+            } else if(ch<=0x7ff) {
+                reqLength+=2;
+            } else if(!U16_IS_SURROGATE(ch)) {
+                reqLength+=3;
+            } else if(U16_IS_SURROGATE_LEAD(ch) && pSrc<pSrcLimit && U16_IS_TRAIL(ch2=*pSrc)) {
+                ++pSrc;
+                reqLength+=4;
+            } else if(subchar>=0) {
+                reqLength+=U8_LENGTH(subchar);
+                ++numSubstitutions;
+            } else {
+                /* Unicode 3.2 forbids surrogate code points in UTF-8 */
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                return NULL;
+            }
+        }
+    }
+
+    reqLength+=(int32_t)(pDest - (uint8_t *)dest);
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=numSubstitutions;
+    }
+
+    if(pDestLength){
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateChars(dest, destCapacity, reqLength, pErrorCode);
+    return dest;
+}
+
+U_CAPI char* U_EXPORT2 
+u_strToUTF8(char *dest,
+            int32_t destCapacity,
+            int32_t *pDestLength,
+            const UChar *pSrc,
+            int32_t srcLength,
+            UErrorCode *pErrorCode){
+    return u_strToUTF8WithSub(
+            dest, destCapacity, pDestLength,
+            pSrc, srcLength,
+            U_SENTINEL, NULL,
+            pErrorCode);
+}
+
+U_CAPI UChar* U_EXPORT2
+u_strFromJavaModifiedUTF8WithSub(
+        UChar *dest,
+        int32_t destCapacity,
+        int32_t *pDestLength,
+        const char *src,
+        int32_t srcLength,
+        UChar32 subchar, int32_t *pNumSubstitutions,
+        UErrorCode *pErrorCode) {
+    /* args check */
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (dest==NULL && destCapacity!=0) || destCapacity<0 ||
+        subchar > 0x10ffff || U_IS_SURROGATE(subchar)
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=0;
+    }
+    UChar *pDest = dest;
+    UChar *pDestLimit = dest+destCapacity;
+    int32_t reqLength = 0;
+    int32_t numSubstitutions=0;
+
+    if(srcLength < 0) {
+        /*
+         * Transform a NUL-terminated ASCII string.
+         * Handle non-ASCII strings with slower code.
+         */
+        UChar32 c;
+        while(((c = (uint8_t)*src) != 0) && c <= 0x7f && (pDest < pDestLimit)) {
+            *pDest++=(UChar)c;
+            ++src;
+        }
+        if(c == 0) {
+            reqLength=(int32_t)(pDest - dest);
+            if(pDestLength) {
+                *pDestLength = reqLength;
+            }
+
+            /* Terminate the buffer */
+            u_terminateUChars(dest, destCapacity, reqLength, pErrorCode);
+            return dest;
+        }
+        srcLength = static_cast<int32_t>(uprv_strlen(src));
+    }
+
+    /* Faster loop without ongoing checking for srcLength and pDestLimit. */
+    UChar32 ch;
+    uint8_t t1, t2;
+    int32_t i = 0;
+    for(;;) {
+        int32_t count = (int32_t)(pDestLimit - pDest);
+        int32_t count2 = srcLength - i;
+        if(count >= count2 && srcLength > 0 && U8_IS_SINGLE(*src)) {
+            /* fast ASCII loop */
+            int32_t start = i;
+            uint8_t b;
+            while(i < srcLength && U8_IS_SINGLE(b = src[i])) {
+                *pDest++=b;
+                ++i;
+            }
+            int32_t delta = i - start;
+            count -= delta;
+            count2 -= delta;
+        }
+        /*
+         * Each iteration of the inner loop progresses by at most 3 UTF-8
+         * bytes and one UChar.
+         */
+        if(subchar > 0xFFFF) {
+            break;
+        }
+        count2 /= 3;
+        if(count > count2) {
+            count = count2; /* min(remaining dest, remaining src/3) */
+        }
+        if(count < 3) {
+            /*
+             * Too much overhead if we get near the end of the string,
+             * continue with the next loop.
+             */
+            break;
+        }
+        do {
+            ch = (uint8_t)src[i++];
+            if(U8_IS_SINGLE(ch)) {
+                *pDest++=(UChar)ch;
+            } else {
+                if(ch >= 0xe0) {
+                    if( /* handle U+0000..U+FFFF inline */
+                        ch <= 0xef &&
+                        (t1 = (uint8_t)(src[i] - 0x80)) <= 0x3f &&
+                        (t2 = (uint8_t)(src[i+1] - 0x80)) <= 0x3f
+                    ) {
+                        /* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
+                        *pDest++ = (UChar)((ch << 12) | (t1 << 6) | t2);
+                        i += 2;
+                        continue;
+                    }
+                } else {
+                    if( /* handle U+0000..U+07FF inline */
+                        ch >= 0xc0 &&
+                        (t1 = (uint8_t)(src[i] - 0x80)) <= 0x3f
+                    ) {
+                        *pDest++ = (UChar)(((ch & 0x1f) << 6) | t1);
+                        ++i;
+                        continue;
+                    }
+                }
+
+                if(subchar < 0) {
+                    *pErrorCode = U_INVALID_CHAR_FOUND;
+                    return NULL;
+                } else if(subchar > 0xffff && --count == 0) {
+                    /*
+                     * We need to write two UChars, adjusted count for that,
+                     * and ran out of space.
+                     */
+                    --i;  // back out byte ch
+                    break;
+                } else {
+                    /* function call for error cases */
+                    utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, ch, -1);
+                    ++numSubstitutions;
+                    *(pDest++)=(UChar)subchar;
+                }
+            }
+        } while(--count > 0);
+    }
+
+    while(i < srcLength && (pDest < pDestLimit)) {
+        ch = (uint8_t)src[i++];
+        if(U8_IS_SINGLE(ch)){
+            *pDest++=(UChar)ch;
+        } else {
+            if(ch >= 0xe0) {
+                if( /* handle U+0000..U+FFFF inline */
+                    ch <= 0xef &&
+                    (i+1) < srcLength &&
+                    (t1 = (uint8_t)(src[i] - 0x80)) <= 0x3f &&
+                    (t2 = (uint8_t)(src[i+1] - 0x80)) <= 0x3f
+                ) {
+                    /* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
+                    *pDest++ = (UChar)((ch << 12) | (t1 << 6) | t2);
+                    i += 2;
+                    continue;
+                }
+            } else {
+                if( /* handle U+0000..U+07FF inline */
+                    ch >= 0xc0 &&
+                    i < srcLength &&
+                    (t1 = (uint8_t)(src[i] - 0x80)) <= 0x3f
+                ) {
+                    *pDest++ = (UChar)(((ch & 0x1f) << 6) | t1);
+                    ++i;
+                    continue;
+                }
+            }
+
+            if(subchar < 0) {
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                return NULL;
+            } else {
+                /* function call for error cases */
+                utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, ch, -1);
+                ++numSubstitutions;
+                if(subchar<=0xFFFF) {
+                    *(pDest++)=(UChar)subchar;
+                } else {
+                    *(pDest++)=U16_LEAD(subchar);
+                    if(pDest<pDestLimit) {
+                        *(pDest++)=U16_TRAIL(subchar);
+                    } else {
+                        reqLength++;
+                        break;
+                    }
+                }
+            }
+        }
+    }
+
+    /* Pre-flight the rest of the string. */
+    while(i < srcLength) {
+        ch = (uint8_t)src[i++];
+        if(U8_IS_SINGLE(ch)) {
+            reqLength++;
+        } else {
+            if(ch >= 0xe0) {
+                if( /* handle U+0000..U+FFFF inline */
+                    ch <= 0xef &&
+                    (i+1) < srcLength &&
+                    (uint8_t)(src[i] - 0x80) <= 0x3f &&
+                    (uint8_t)(src[i+1] - 0x80) <= 0x3f
+                ) {
+                    reqLength++;
+                    i += 2;
+                    continue;
+                }
+            } else {
+                if( /* handle U+0000..U+07FF inline */
+                    ch >= 0xc0 &&
+                    i < srcLength &&
+                    (uint8_t)(src[i] - 0x80) <= 0x3f
+                ) {
+                    reqLength++;
+                    ++i;
+                    continue;
+                }
+            }
+
+            if(subchar < 0) {
+                *pErrorCode = U_INVALID_CHAR_FOUND;
+                return NULL;
+            } else {
+                /* function call for error cases */
+                utf8_nextCharSafeBody((const uint8_t *)src, &(i), srcLength, ch, -1);
+                ++numSubstitutions;
+                reqLength+=U16_LENGTH(ch);
+            }
+        }
+    }
+
+    if(pNumSubstitutions!=NULL) {
+        *pNumSubstitutions=numSubstitutions;
+    }
+
+    reqLength+=(int32_t)(pDest - dest);
+    if(pDestLength) {
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChars(dest, destCapacity, reqLength, pErrorCode);
+    return dest;
+}
+
+U_CAPI char* U_EXPORT2 
+u_strToJavaModifiedUTF8(
+        char *dest,
+        int32_t destCapacity,
+        int32_t *pDestLength,
+        const UChar *src, 
+        int32_t srcLength,
+        UErrorCode *pErrorCode) {
+    int32_t reqLength=0;
+    uint32_t ch=0;
+    uint8_t *pDest = (uint8_t *)dest;
+    uint8_t *pDestLimit = pDest + destCapacity;
+    const UChar *pSrcLimit;
+    int32_t count;
+
+    /* args check */
+    if(U_FAILURE(*pErrorCode)){
+        return NULL;
+    }
+    if( (src==NULL && srcLength!=0) || srcLength < -1 ||
+        (dest==NULL && destCapacity!=0) || destCapacity<0
+    ) {
+        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    if(srcLength==-1) {
+        /* Convert NUL-terminated ASCII, then find the string length. */
+        while((ch=*src)<=0x7f && ch != 0 && pDest<pDestLimit) {
+            *pDest++ = (uint8_t)ch;
+            ++src;
+        }
+        if(ch == 0) {
+            reqLength=(int32_t)(pDest - (uint8_t *)dest);
+            if(pDestLength) {
+                *pDestLength = reqLength;
+            }
+
+            /* Terminate the buffer */
+            u_terminateChars(dest, destCapacity, reqLength, pErrorCode);
+            return dest;
+        }
+        srcLength = u_strlen(src);
+    }
+
+    /* Faster loop without ongoing checking for pSrcLimit and pDestLimit. */
+    pSrcLimit = (src!=NULL)?(src+srcLength):NULL;
+    for(;;) {
+        count = (int32_t)(pDestLimit - pDest);
+        srcLength = (int32_t)(pSrcLimit - src);
+        if(count >= srcLength && srcLength > 0 && *src <= 0x7f) {
+            /* fast ASCII loop */
+            const UChar *prevSrc = src;
+            int32_t delta;
+            while(src < pSrcLimit && (ch = *src) <= 0x7f && ch != 0) {
+                *pDest++=(uint8_t)ch;
+                ++src;
+            }
+            delta = (int32_t)(src - prevSrc);
+            count -= delta;
+            srcLength -= delta;
+        }
+        /*
+         * Each iteration of the inner loop progresses by at most 3 UTF-8
+         * bytes and one UChar.
+         */
+        count /= 3;
+        if(count > srcLength) {
+            count = srcLength; /* min(remaining dest/3, remaining src) */
+        }
+        if(count < 3) {
+            /*
+             * Too much overhead if we get near the end of the string,
+             * continue with the next loop.
+             */
+            break;
+        }
+        do {
+            ch=*src++;
+            if(ch <= 0x7f && ch != 0) {
+                *pDest++ = (uint8_t)ch;
+            } else if(ch <= 0x7ff) {
+                *pDest++=(uint8_t)((ch>>6)|0xc0);
+                *pDest++=(uint8_t)((ch&0x3f)|0x80);
+            } else {
+                *pDest++=(uint8_t)((ch>>12)|0xe0);
+                *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                *pDest++=(uint8_t)((ch&0x3f)|0x80);
+            }
+        } while(--count > 0);
+    }
+
+    while(src<pSrcLimit) {
+        ch=*src++;
+        if(ch <= 0x7f && ch != 0) {
+            if(pDest<pDestLimit) {
+                *pDest++ = (uint8_t)ch;
+            } else {
+                reqLength = 1;
+                break;
+            }
+        } else if(ch <= 0x7ff) {
+            if((pDestLimit - pDest) >= 2) {
+                *pDest++=(uint8_t)((ch>>6)|0xc0);
+                *pDest++=(uint8_t)((ch&0x3f)|0x80);
+            } else {
+                reqLength = 2;
+                break;
+            }
+        } else {
+            if((pDestLimit - pDest) >= 3) {
+                *pDest++=(uint8_t)((ch>>12)|0xe0);
+                *pDest++=(uint8_t)(((ch>>6)&0x3f)|0x80);
+                *pDest++=(uint8_t)((ch&0x3f)|0x80);
+            } else {
+                reqLength = 3;
+                break;
+            }
+        }
+    }
+    while(src<pSrcLimit) {
+        ch=*src++;
+        if(ch <= 0x7f && ch != 0) {
+            ++reqLength;
+        } else if(ch<=0x7ff) {
+            reqLength+=2;
+        } else {
+            reqLength+=3;
+        }
+    }
+
+    reqLength+=(int32_t)(pDest - (uint8_t *)dest);
+    if(pDestLength){
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateChars(dest, destCapacity, reqLength, pErrorCode);
+    return dest;
+}
diff --git a/thirdparty/icu4c/common/utext.cpp b/thirdparty/icu4c/common/utext.cpp
new file mode 100644
index 0000000000..763b6684fb
--- /dev/null
+++ b/thirdparty/icu4c/common/utext.cpp
@@ -0,0 +1,2877 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2005-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utext.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2005apr12
+*   created by: Markus W. Scherer
+*/
+
+#include <cstddef>
+
+#include "unicode/utypes.h"
+#include "unicode/ustring.h"
+#include "unicode/unistr.h"
+#include "unicode/chariter.h"
+#include "unicode/utext.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "ustr_imp.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "putilimp.h"
+
+U_NAMESPACE_USE
+
+#define I32_FLAG(bitIndex) ((int32_t)1<<(bitIndex))
+
+
+static UBool
+utext_access(UText *ut, int64_t index, UBool forward) {
+    return ut->pFuncs->access(ut, index, forward);
+}
+
+
+
+U_CAPI UBool U_EXPORT2
+utext_moveIndex32(UText *ut, int32_t delta) {
+    UChar32  c;
+    if (delta > 0) {
+        do {
+            if(ut->chunkOffset>=ut->chunkLength && !utext_access(ut, ut->chunkNativeLimit, TRUE)) {
+                return FALSE;
+            }
+            c = ut->chunkContents[ut->chunkOffset];
+            if (U16_IS_SURROGATE(c)) {
+                c = utext_next32(ut);
+                if (c == U_SENTINEL) {
+                    return FALSE;
+                }
+            } else {
+                ut->chunkOffset++;
+            }
+        } while(--delta>0);
+
+    } else if (delta<0) {
+        do {
+            if(ut->chunkOffset<=0 && !utext_access(ut, ut->chunkNativeStart, FALSE)) {
+                return FALSE;
+            }
+            c = ut->chunkContents[ut->chunkOffset-1];
+            if (U16_IS_SURROGATE(c)) {
+                c = utext_previous32(ut);
+                if (c == U_SENTINEL) {
+                    return FALSE;
+                }
+            } else {
+                ut->chunkOffset--;
+            }
+        } while(++delta<0);
+    }
+
+    return TRUE;
+}
+
+
+U_CAPI int64_t U_EXPORT2
+utext_nativeLength(UText *ut) {
+    return ut->pFuncs->nativeLength(ut);
+}
+
+
+U_CAPI UBool U_EXPORT2
+utext_isLengthExpensive(const UText *ut) {
+    UBool r = (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE)) != 0;
+    return r;
+}
+
+
+U_CAPI int64_t U_EXPORT2
+utext_getNativeIndex(const UText *ut) {
+    if(ut->chunkOffset <= ut->nativeIndexingLimit) {
+        return ut->chunkNativeStart+ut->chunkOffset;
+    } else {
+        return ut->pFuncs->mapOffsetToNative(ut);
+    }
+}
+
+
+U_CAPI void U_EXPORT2
+utext_setNativeIndex(UText *ut, int64_t index) {
+    if(index<ut->chunkNativeStart || index>=ut->chunkNativeLimit) {
+        // The desired position is outside of the current chunk.
+        // Access the new position.  Assume a forward iteration from here,
+        // which will also be optimimum for a single random access.
+        // Reverse iterations may suffer slightly.
+        ut->pFuncs->access(ut, index, TRUE);
+    } else if((int32_t)(index - ut->chunkNativeStart) <= ut->nativeIndexingLimit) {
+        // utf-16 indexing.
+        ut->chunkOffset=(int32_t)(index-ut->chunkNativeStart);
+    } else {
+         ut->chunkOffset=ut->pFuncs->mapNativeIndexToUTF16(ut, index);
+    }
+    // The convention is that the index must always be on a code point boundary.
+    // Adjust the index position if it is in the middle of a surrogate pair.
+    if (ut->chunkOffset<ut->chunkLength) {
+        UChar c= ut->chunkContents[ut->chunkOffset];
+        if (U16_IS_TRAIL(c)) {
+            if (ut->chunkOffset==0) {
+                ut->pFuncs->access(ut, ut->chunkNativeStart, FALSE);
+            }
+            if (ut->chunkOffset>0) {
+                UChar lead = ut->chunkContents[ut->chunkOffset-1];
+                if (U16_IS_LEAD(lead)) {
+                    ut->chunkOffset--;
+                }
+            }
+        }
+    }
+}
+
+
+
+U_CAPI int64_t U_EXPORT2
+utext_getPreviousNativeIndex(UText *ut) {
+    //
+    //  Fast-path the common case.
+    //     Common means current position is not at the beginning of a chunk
+    //     and the preceding character is not supplementary.
+    //
+    int32_t i = ut->chunkOffset - 1;
+    int64_t result;
+    if (i >= 0) {
+        UChar c = ut->chunkContents[i];
+        if (U16_IS_TRAIL(c) == FALSE) {
+            if (i <= ut->nativeIndexingLimit) {
+                result = ut->chunkNativeStart + i;
+            } else {
+                ut->chunkOffset = i;
+                result = ut->pFuncs->mapOffsetToNative(ut);
+                ut->chunkOffset++;
+            }
+            return result;
+        }
+    }
+
+    // If at the start of text, simply return 0.
+    if (ut->chunkOffset==0 && ut->chunkNativeStart==0) {
+        return 0;
+    }
+
+    // Harder, less common cases.  We are at a chunk boundary, or on a surrogate.
+    //    Keep it simple, use other functions to handle the edges.
+    //
+    utext_previous32(ut);
+    result = UTEXT_GETNATIVEINDEX(ut);
+    utext_next32(ut);
+    return result;
+}
+
+
+//
+//  utext_current32.  Get the UChar32 at the current position.
+//                    UText iteration position is always on a code point boundary,
+//                    never on the trail half of a surrogate pair.
+//
+U_CAPI UChar32 U_EXPORT2
+utext_current32(UText *ut) {
+    UChar32  c;
+    if (ut->chunkOffset==ut->chunkLength) {
+        // Current position is just off the end of the chunk.
+        if (ut->pFuncs->access(ut, ut->chunkNativeLimit, TRUE) == FALSE) {
+            // Off the end of the text.
+            return U_SENTINEL;
+        }
+    }
+
+    c = ut->chunkContents[ut->chunkOffset];
+    if (U16_IS_LEAD(c) == FALSE) {
+        // Normal, non-supplementary case.
+        return c;
+    }
+
+    //
+    //  Possible supplementary char.
+    //
+    UChar32   trail = 0;
+    UChar32   supplementaryC = c;
+    if ((ut->chunkOffset+1) < ut->chunkLength) {
+        // The trail surrogate is in the same chunk.
+        trail = ut->chunkContents[ut->chunkOffset+1];
+    } else {
+        //  The trail surrogate is in a different chunk.
+        //     Because we must maintain the iteration position, we need to switch forward
+        //     into the new chunk, get the trail surrogate, then revert the chunk back to the
+        //     original one.
+        //     An edge case to be careful of:  the entire text may end with an unpaired
+        //        leading surrogate.  The attempt to access the trail will fail, but
+        //        the original position before the unpaired lead still needs to be restored.
+        int64_t  nativePosition = ut->chunkNativeLimit;
+        int32_t  originalOffset = ut->chunkOffset;
+        if (ut->pFuncs->access(ut, nativePosition, TRUE)) {
+            trail = ut->chunkContents[ut->chunkOffset];
+        }
+        UBool r = ut->pFuncs->access(ut, nativePosition, FALSE);  // reverse iteration flag loads preceding chunk
+        U_ASSERT(r==TRUE);
+        ut->chunkOffset = originalOffset;
+        if(!r) {
+            return U_SENTINEL;
+        }
+    }
+
+    if (U16_IS_TRAIL(trail)) {
+        supplementaryC = U16_GET_SUPPLEMENTARY(c, trail);
+    }
+    return supplementaryC;
+
+}
+
+
+U_CAPI UChar32 U_EXPORT2
+utext_char32At(UText *ut, int64_t nativeIndex) {
+    UChar32 c = U_SENTINEL;
+
+    // Fast path the common case.
+    if (nativeIndex>=ut->chunkNativeStart && nativeIndex < ut->chunkNativeStart + ut->nativeIndexingLimit) {
+        ut->chunkOffset = (int32_t)(nativeIndex - ut->chunkNativeStart);
+        c = ut->chunkContents[ut->chunkOffset];
+        if (U16_IS_SURROGATE(c) == FALSE) {
+            return c;
+        }
+    }
+
+
+    utext_setNativeIndex(ut, nativeIndex);
+    if (nativeIndex>=ut->chunkNativeStart && ut->chunkOffset<ut->chunkLength) {
+        c = ut->chunkContents[ut->chunkOffset];
+        if (U16_IS_SURROGATE(c)) {
+            // For surrogates, let current32() deal with the complications
+            //    of supplementaries that may span chunk boundaries.
+            c = utext_current32(ut);
+        }
+    }
+    return c;
+}
+
+
+U_CAPI UChar32 U_EXPORT2
+utext_next32(UText *ut) {
+    UChar32       c;
+
+    if (ut->chunkOffset >= ut->chunkLength) {
+        if (ut->pFuncs->access(ut, ut->chunkNativeLimit, TRUE) == FALSE) {
+            return U_SENTINEL;
+        }
+    }
+
+    c = ut->chunkContents[ut->chunkOffset++];
+    if (U16_IS_LEAD(c) == FALSE) {
+        // Normal case, not supplementary.
+        //   (A trail surrogate seen here is just returned as is, as a surrogate value.
+        //    It cannot be part of a pair.)
+        return c;
+    }
+
+    if (ut->chunkOffset >= ut->chunkLength) {
+        if (ut->pFuncs->access(ut, ut->chunkNativeLimit, TRUE) == FALSE) {
+            // c is an unpaired lead surrogate at the end of the text.
+            // return it as it is.
+            return c;
+        }
+    }
+    UChar32 trail = ut->chunkContents[ut->chunkOffset];
+    if (U16_IS_TRAIL(trail) == FALSE) {
+        // c was an unpaired lead surrogate, not at the end of the text.
+        // return it as it is (unpaired).  Iteration position is on the
+        // following character, possibly in the next chunk, where the
+        //  trail surrogate would have been if it had existed.
+        return c;
+    }
+
+    UChar32 supplementary = U16_GET_SUPPLEMENTARY(c, trail);
+    ut->chunkOffset++;   // move iteration position over the trail surrogate.
+    return supplementary;
+    }
+
+
+U_CAPI UChar32 U_EXPORT2
+utext_previous32(UText *ut) {
+    UChar32       c;
+
+    if (ut->chunkOffset <= 0) {
+        if (ut->pFuncs->access(ut, ut->chunkNativeStart, FALSE) == FALSE) {
+            return U_SENTINEL;
+        }
+    }
+    ut->chunkOffset--;
+    c = ut->chunkContents[ut->chunkOffset];
+    if (U16_IS_TRAIL(c) == FALSE) {
+        // Normal case, not supplementary.
+        //   (A lead surrogate seen here is just returned as is, as a surrogate value.
+        //    It cannot be part of a pair.)
+        return c;
+    }
+
+    if (ut->chunkOffset <= 0) {
+        if (ut->pFuncs->access(ut, ut->chunkNativeStart, FALSE) == FALSE) {
+            // c is an unpaired trail surrogate at the start of the text.
+            // return it as it is.
+            return c;
+        }
+    }
+
+    UChar32 lead = ut->chunkContents[ut->chunkOffset-1];
+    if (U16_IS_LEAD(lead) == FALSE) {
+        // c was an unpaired trail surrogate, not at the end of the text.
+        // return it as it is (unpaired).  Iteration position is at c
+        return c;
+    }
+
+    UChar32 supplementary = U16_GET_SUPPLEMENTARY(lead, c);
+    ut->chunkOffset--;   // move iteration position over the lead surrogate.
+    return supplementary;
+}
+
+
+
+U_CAPI UChar32 U_EXPORT2
+utext_next32From(UText *ut, int64_t index) {
+    UChar32       c      = U_SENTINEL;
+
+    if(index<ut->chunkNativeStart || index>=ut->chunkNativeLimit) {
+        // Desired position is outside of the current chunk.
+        if(!ut->pFuncs->access(ut, index, TRUE)) {
+            // no chunk available here
+            return U_SENTINEL;
+        }
+    } else if (index - ut->chunkNativeStart  <= (int64_t)ut->nativeIndexingLimit) {
+        // Desired position is in chunk, with direct 1:1 native to UTF16 indexing
+        ut->chunkOffset = (int32_t)(index - ut->chunkNativeStart);
+    } else {
+        // Desired position is in chunk, with non-UTF16 indexing.
+        ut->chunkOffset = ut->pFuncs->mapNativeIndexToUTF16(ut, index);
+    }
+
+    c = ut->chunkContents[ut->chunkOffset++];
+    if (U16_IS_SURROGATE(c)) {
+        // Surrogates.  Many edge cases.  Use other functions that already
+        //              deal with the problems.
+        utext_setNativeIndex(ut, index);
+        c = utext_next32(ut);
+    }
+    return c;
+}
+
+
+U_CAPI UChar32 U_EXPORT2
+utext_previous32From(UText *ut, int64_t index) {
+    //
+    //  Return the character preceding the specified index.
+    //  Leave the iteration position at the start of the character that was returned.
+    //
+    UChar32     cPrev;    // The character preceding cCurr, which is what we will return.
+
+    // Address the chunk containg the position preceding the incoming index
+    // A tricky edge case:
+    //   We try to test the requested native index against the chunkNativeStart to determine
+    //    whether the character preceding the one at the index is in the current chunk.
+    //    BUT, this test can fail with UTF-8 (or any other multibyte encoding), when the
+    //    requested index is on something other than the first position of the first char.
+    //
+    if(index<=ut->chunkNativeStart || index>ut->chunkNativeLimit) {
+        // Requested native index is outside of the current chunk.
+        if(!ut->pFuncs->access(ut, index, FALSE)) {
+            // no chunk available here
+            return U_SENTINEL;
+        }
+    } else if(index - ut->chunkNativeStart <= (int64_t)ut->nativeIndexingLimit) {
+        // Direct UTF-16 indexing.
+        ut->chunkOffset = (int32_t)(index - ut->chunkNativeStart);
+    } else {
+        ut->chunkOffset=ut->pFuncs->mapNativeIndexToUTF16(ut, index);
+        if (ut->chunkOffset==0 && !ut->pFuncs->access(ut, index, FALSE)) {
+            // no chunk available here
+            return U_SENTINEL;
+        }
+    }
+
+    //
+    // Simple case with no surrogates.
+    //
+    ut->chunkOffset--;
+    cPrev = ut->chunkContents[ut->chunkOffset];
+
+    if (U16_IS_SURROGATE(cPrev)) {
+        // Possible supplementary.  Many edge cases.
+        // Let other functions do the heavy lifting.
+        utext_setNativeIndex(ut, index);
+        cPrev = utext_previous32(ut);
+    }
+    return cPrev;
+}
+
+
+U_CAPI int32_t U_EXPORT2
+utext_extract(UText *ut,
+             int64_t start, int64_t limit,
+             UChar *dest, int32_t destCapacity,
+             UErrorCode *status) {
+                 return ut->pFuncs->extract(ut, start, limit, dest, destCapacity, status);
+             }
+
+
+
+U_CAPI UBool U_EXPORT2
+utext_equals(const UText *a, const UText *b) {
+    if (a==NULL || b==NULL ||
+        a->magic != UTEXT_MAGIC ||
+        b->magic != UTEXT_MAGIC) {
+            // Null or invalid arguments don't compare equal to anything.
+            return FALSE;
+    }
+
+    if (a->pFuncs != b->pFuncs) {
+        // Different types of text providers.
+        return FALSE;
+    }
+
+    if (a->context != b->context) {
+        // Different sources (different strings)
+        return FALSE;
+    }
+    if (utext_getNativeIndex(a) != utext_getNativeIndex(b)) {
+        // Different current position in the string.
+        return FALSE;
+    }
+
+    return TRUE;
+}
+
+U_CAPI UBool U_EXPORT2
+utext_isWritable(const UText *ut)
+{
+    UBool b = (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_WRITABLE)) != 0;
+    return b;
+}
+
+
+U_CAPI void U_EXPORT2
+utext_freeze(UText *ut) {
+    // Zero out the WRITABLE flag.
+    ut->providerProperties &= ~(I32_FLAG(UTEXT_PROVIDER_WRITABLE));
+}
+
+
+U_CAPI UBool U_EXPORT2
+utext_hasMetaData(const UText *ut)
+{
+    UBool b = (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_HAS_META_DATA)) != 0;
+    return b;
+}
+
+
+
+U_CAPI int32_t U_EXPORT2
+utext_replace(UText *ut,
+             int64_t nativeStart, int64_t nativeLimit,
+             const UChar *replacementText, int32_t replacementLength,
+             UErrorCode *status)
+{
+    if (U_FAILURE(*status)) {
+        return 0;
+    }
+    if ((ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_WRITABLE)) == 0) {
+        *status = U_NO_WRITE_PERMISSION;
+        return 0;
+    }
+    int32_t i = ut->pFuncs->replace(ut, nativeStart, nativeLimit, replacementText, replacementLength, status);
+    return i;
+}
+
+U_CAPI void U_EXPORT2
+utext_copy(UText *ut,
+          int64_t nativeStart, int64_t nativeLimit,
+          int64_t destIndex,
+          UBool move,
+          UErrorCode *status)
+{
+    if (U_FAILURE(*status)) {
+        return;
+    }
+    if ((ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_WRITABLE)) == 0) {
+        *status = U_NO_WRITE_PERMISSION;
+        return;
+    }
+    ut->pFuncs->copy(ut, nativeStart, nativeLimit, destIndex, move, status);
+}
+
+
+
+U_CAPI UText * U_EXPORT2
+utext_clone(UText *dest, const UText *src, UBool deep, UBool readOnly, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return dest;
+    }
+    UText *result = src->pFuncs->clone(dest, src, deep, status);
+    if (U_FAILURE(*status)) {
+        return result;
+    }
+    if (result == NULL) {
+        *status = U_MEMORY_ALLOCATION_ERROR;
+        return result;
+    }
+    if (readOnly) {
+        utext_freeze(result);
+    }
+    return result;
+}
+
+
+
+//------------------------------------------------------------------------------
+//
+//   UText common functions implementation
+//
+//------------------------------------------------------------------------------
+
+//
+//  UText.flags bit definitions
+//
+enum {
+    UTEXT_HEAP_ALLOCATED  = 1,      //  1 if ICU has allocated this UText struct on the heap.
+                                    //  0 if caller provided storage for the UText.
+
+    UTEXT_EXTRA_HEAP_ALLOCATED = 2, //  1 if ICU has allocated extra storage as a separate
+                                    //     heap block.
+                                    //  0 if there is no separate allocation.  Either no extra
+                                    //     storage was requested, or it is appended to the end
+                                    //     of the main UText storage.
+
+    UTEXT_OPEN = 4                  //  1 if this UText is currently open
+                                    //  0 if this UText is not open.
+};
+
+
+//
+//  Extended form of a UText.  The purpose is to aid in computing the total size required
+//    when a provider asks for a UText to be allocated with extra storage.
+
+struct ExtendedUText {
+    UText               ut;
+    std::max_align_t    extension;
+};
+
+static const UText emptyText = UTEXT_INITIALIZER;
+
+U_CAPI UText * U_EXPORT2
+utext_setup(UText *ut, int32_t extraSpace, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return ut;
+    }
+
+    if (ut == NULL) {
+        // We need to heap-allocate storage for the new UText
+        int32_t spaceRequired = sizeof(UText);
+        if (extraSpace > 0) {
+            spaceRequired = sizeof(ExtendedUText) + extraSpace - sizeof(std::max_align_t);
+        }
+        ut = (UText *)uprv_malloc(spaceRequired);
+        if (ut == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+            return NULL;
+        } else {
+            *ut = emptyText;
+            ut->flags |= UTEXT_HEAP_ALLOCATED;
+            if (spaceRequired>0) {
+                ut->extraSize = extraSpace;
+                ut->pExtra    = &((ExtendedUText *)ut)->extension;
+            }
+        }
+    } else {
+        // We have been supplied with an already existing UText.
+        // Verify that it really appears to be a UText.
+        if (ut->magic != UTEXT_MAGIC) {
+            *status = U_ILLEGAL_ARGUMENT_ERROR;
+            return ut;
+        }
+        // If the ut is already open and there's a provider supplied close
+        //   function, call it.
+        if ((ut->flags & UTEXT_OPEN) && ut->pFuncs->close != NULL)  {
+            ut->pFuncs->close(ut);
+        }
+        ut->flags &= ~UTEXT_OPEN;
+
+        // If extra space was requested by our caller, check whether
+        //   sufficient already exists, and allocate new if needed.
+        if (extraSpace > ut->extraSize) {
+            // Need more space.  If there is existing separately allocated space,
+            //   delete it first, then allocate new space.
+            if (ut->flags & UTEXT_EXTRA_HEAP_ALLOCATED) {
+                uprv_free(ut->pExtra);
+                ut->extraSize = 0;
+            }
+            ut->pExtra = uprv_malloc(extraSpace);
+            if (ut->pExtra == NULL) {
+                *status = U_MEMORY_ALLOCATION_ERROR;
+            } else {
+                ut->extraSize = extraSpace;
+                ut->flags |= UTEXT_EXTRA_HEAP_ALLOCATED;
+            }
+        }
+    }
+    if (U_SUCCESS(*status)) {
+        ut->flags |= UTEXT_OPEN;
+
+        // Initialize all remaining fields of the UText.
+        //
+        ut->context             = NULL;
+        ut->chunkContents       = NULL;
+        ut->p                   = NULL;
+        ut->q                   = NULL;
+        ut->r                   = NULL;
+        ut->a                   = 0;
+        ut->b                   = 0;
+        ut->c                   = 0;
+        ut->chunkOffset         = 0;
+        ut->chunkLength         = 0;
+        ut->chunkNativeStart    = 0;
+        ut->chunkNativeLimit    = 0;
+        ut->nativeIndexingLimit = 0;
+        ut->providerProperties  = 0;
+        ut->privA               = 0;
+        ut->privB               = 0;
+        ut->privC               = 0;
+        ut->privP               = NULL;
+        if (ut->pExtra!=NULL && ut->extraSize>0)
+            uprv_memset(ut->pExtra, 0, ut->extraSize);
+
+    }
+    return ut;
+}
+
+
+U_CAPI UText * U_EXPORT2
+utext_close(UText *ut) {
+    if (ut==NULL ||
+        ut->magic != UTEXT_MAGIC ||
+        (ut->flags & UTEXT_OPEN) == 0)
+    {
+        // The supplied ut is not an open UText.
+        // Do nothing.
+        return ut;
+    }
+
+    // If the provider gave us a close function, call it now.
+    // This will clean up anything allocated specifically by the provider.
+    if (ut->pFuncs->close != NULL) {
+        ut->pFuncs->close(ut);
+    }
+    ut->flags &= ~UTEXT_OPEN;
+
+    // If we (the framework) allocated the UText or subsidiary storage,
+    //   delete it.
+    if (ut->flags & UTEXT_EXTRA_HEAP_ALLOCATED) {
+        uprv_free(ut->pExtra);
+        ut->pExtra = NULL;
+        ut->flags &= ~UTEXT_EXTRA_HEAP_ALLOCATED;
+        ut->extraSize = 0;
+    }
+
+    // Zero out function table of the closed UText.  This is a defensive move,
+    //   inteded to cause applications that inadvertantly use a closed
+    //   utext to crash with null pointer errors.
+    ut->pFuncs        = NULL;
+
+    if (ut->flags & UTEXT_HEAP_ALLOCATED) {
+        // This UText was allocated by UText setup.  We need to free it.
+        // Clear magic, so we can detect if the user messes up and immediately
+        //  tries to reopen another UText using the deleted storage.
+        ut->magic = 0;
+        uprv_free(ut);
+        ut = NULL;
+    }
+    return ut;
+}
+
+
+
+
+//
+// invalidateChunk   Reset a chunk to have no contents, so that the next call
+//                   to access will cause new data to load.
+//                   This is needed when copy/move/replace operate directly on the
+//                   backing text, potentially putting it out of sync with the
+//                   contents in the chunk.
+//
+static void
+invalidateChunk(UText *ut) {
+    ut->chunkLength = 0;
+    ut->chunkNativeLimit = 0;
+    ut->chunkNativeStart = 0;
+    ut->chunkOffset = 0;
+    ut->nativeIndexingLimit = 0;
+}
+
+//
+// pinIndex        Do range pinning on a native index parameter.
+//                 64 bit pinning is done in place.
+//                 32 bit truncated result is returned as a convenience for
+//                        use in providers that don't need 64 bits.
+static int32_t
+pinIndex(int64_t &index, int64_t limit) {
+    if (index<0) {
+        index = 0;
+    } else if (index > limit) {
+        index = limit;
+    }
+    return (int32_t)index;
+}
+
+
+U_CDECL_BEGIN
+
+//
+// Pointer relocation function,
+//   a utility used by shallow clone.
+//   Adjust a pointer that refers to something within one UText (the source)
+//   to refer to the same relative offset within a another UText (the target)
+//
+static void adjustPointer(UText *dest, const void **destPtr, const UText *src) {
+    // convert all pointers to (char *) so that byte address arithmetic will work.
+    char  *dptr = (char *)*destPtr;
+    char  *dUText = (char *)dest;
+    char  *sUText = (char *)src;
+
+    if (dptr >= (char *)src->pExtra && dptr < ((char*)src->pExtra)+src->extraSize) {
+        // target ptr was to something within the src UText's pExtra storage.
+        //   relocate it into the target UText's pExtra region.
+        *destPtr = ((char *)dest->pExtra) + (dptr - (char *)src->pExtra);
+    } else if (dptr>=sUText && dptr < sUText+src->sizeOfStruct) {
+        // target ptr was pointing to somewhere within the source UText itself.
+        //   Move it to the same offset within the target UText.
+        *destPtr = dUText + (dptr-sUText);
+    }
+}
+
+
+//
+//  Clone.  This is a generic copy-the-utext-by-value clone function that can be
+//          used as-is with some utext types, and as a helper by other clones.
+//
+static UText * U_CALLCONV
+shallowTextClone(UText * dest, const UText * src, UErrorCode * status) {
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    int32_t  srcExtraSize = src->extraSize;
+
+    //
+    // Use the generic text_setup to allocate storage if required.
+    //
+    dest = utext_setup(dest, srcExtraSize, status);
+    if (U_FAILURE(*status)) {
+        return dest;
+    }
+
+    //
+    //  flags (how the UText was allocated) and the pointer to the
+    //   extra storage must retain the values in the cloned utext that
+    //   were set up by utext_setup.  Save them separately before
+    //   copying the whole struct.
+    //
+    void *destExtra = dest->pExtra;
+    int32_t flags   = dest->flags;
+
+
+    //
+    //  Copy the whole UText struct by value.
+    //  Any "Extra" storage is copied also.
+    //
+    int sizeToCopy = src->sizeOfStruct;
+    if (sizeToCopy > dest->sizeOfStruct) {
+        sizeToCopy = dest->sizeOfStruct;
+    }
+    uprv_memcpy(dest, src, sizeToCopy);
+    dest->pExtra = destExtra;
+    dest->flags  = flags;
+    if (srcExtraSize > 0) {
+        uprv_memcpy(dest->pExtra, src->pExtra, srcExtraSize);
+    }
+
+    //
+    // Relocate any pointers in the target that refer to the UText itself
+    //   to point to the cloned copy rather than the original source.
+    //
+    adjustPointer(dest, &dest->context, src);
+    adjustPointer(dest, &dest->p, src);
+    adjustPointer(dest, &dest->q, src);
+    adjustPointer(dest, &dest->r, src);
+    adjustPointer(dest, (const void **)&dest->chunkContents, src);
+
+    // The newly shallow-cloned UText does _not_ own the underlying storage for the text.
+    // (The source for the clone may or may not have owned the text.)
+
+    dest->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT);
+
+    return dest;
+}
+
+
+U_CDECL_END
+
+
+
+//------------------------------------------------------------------------------
+//
+//     UText implementation for UTF-8 char * strings (read-only)
+//     Limitation:  string length must be <= 0x7fffffff in length.
+//                  (length must for in an int32_t variable)
+//
+//         Use of UText data members:
+//              context    pointer to UTF-8 string
+//              utext.b    is the input string length (bytes).
+//              utext.c    Length scanned so far in string
+//                           (for optimizing finding length of zero terminated strings.)
+//              utext.p    pointer to the current buffer
+//              utext.q    pointer to the other buffer.
+//
+//------------------------------------------------------------------------------
+
+// Chunk size.
+//     Must be less than 85 (256/3), because of byte mapping from UChar indexes to native indexes.
+//     Worst case is three native bytes to one UChar.  (Supplemenaries are 4 native bytes
+//     to two UChars.)
+//     The longest illegal byte sequence treated as a single error (and converted to U+FFFD)
+//     is a three-byte sequence (truncated four-byte sequence).
+//
+enum { UTF8_TEXT_CHUNK_SIZE=32 };
+
+//
+// UTF8Buf  Two of these structs will be set up in the UText's extra allocated space.
+//          Each contains the UChar chunk buffer, the to and from native maps, and
+//          header info.
+//
+//     because backwards iteration fills the buffers starting at the end and
+//     working towards the front, the filled part of the buffers may not begin
+//     at the start of the available storage for the buffers.
+//
+//     Buffer size is one bigger than the specified UTF8_TEXT_CHUNK_SIZE to allow for
+//     the last character added being a supplementary, and thus requiring a surrogate
+//     pair.  Doing this is simpler than checking for the edge case.
+//
+
+struct UTF8Buf {
+    int32_t   bufNativeStart;                        // Native index of first char in UChar buf
+    int32_t   bufNativeLimit;                        // Native index following last char in buf.
+    int32_t   bufStartIdx;                           // First filled position in buf.
+    int32_t   bufLimitIdx;                           // Limit of filled range in buf.
+    int32_t   bufNILimit;                            // Limit of native indexing part of buf
+    int32_t   toUCharsMapStart;                      // Native index corresponding to
+                                                     //   mapToUChars[0].
+                                                     //   Set to bufNativeStart when filling forwards.
+                                                     //   Set to computed value when filling backwards.
+
+    UChar     buf[UTF8_TEXT_CHUNK_SIZE+4];           // The UChar buffer.  Requires one extra position beyond the
+                                                     //   the chunk size, to allow for surrogate at the end.
+                                                     //   Length must be identical to mapToNative array, below,
+                                                     //   because of the way indexing works when the array is
+                                                     //   filled backwards during a reverse iteration.  Thus,
+                                                     //   the additional extra size.
+    uint8_t   mapToNative[UTF8_TEXT_CHUNK_SIZE+4];   // map UChar index in buf to
+                                                     //  native offset from bufNativeStart.
+                                                     //  Requires two extra slots,
+                                                     //    one for a supplementary starting in the last normal position,
+                                                     //    and one for an entry for the buffer limit position.
+    uint8_t   mapToUChars[UTF8_TEXT_CHUNK_SIZE*3+6]; // Map native offset from bufNativeStart to
+                                                     //   correspoding offset in filled part of buf.
+    int32_t   align;
+};
+
+U_CDECL_BEGIN
+
+//
+//   utf8TextLength
+//
+//        Get the length of the string.  If we don't already know it,
+//              we'll need to scan for the trailing  nul.
+//
+static int64_t U_CALLCONV
+utf8TextLength(UText *ut) {
+    if (ut->b < 0) {
+        // Zero terminated string, and we haven't scanned to the end yet.
+        // Scan it now.
+        const char *r = (const char *)ut->context + ut->c;
+        while (*r != 0) {
+            r++;
+        }
+        if ((r - (const char *)ut->context) < 0x7fffffff) {
+            ut->b = (int32_t)(r - (const char *)ut->context);
+        } else {
+            // Actual string was bigger (more than 2 gig) than we
+            //   can handle.  Clip it to 2 GB.
+            ut->b = 0x7fffffff;
+        }
+        ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+    }
+    return ut->b;
+}
+
+
+
+
+
+
+static UBool U_CALLCONV
+utf8TextAccess(UText *ut, int64_t index, UBool forward) {
+    //
+    //  Apologies to those who are allergic to goto statements.
+    //    Consider each goto to a labelled block to be the equivalent of
+    //         call the named block as if it were a function();
+    //         return;
+    //
+    const uint8_t *s8=(const uint8_t *)ut->context;
+    UTF8Buf *u8b = NULL;
+    int32_t  length = ut->b;         // Length of original utf-8
+    int32_t  ix= (int32_t)index;     // Requested index, trimmed to 32 bits.
+    int32_t  mapIndex = 0;
+    if (index<0) {
+        ix=0;
+    } else if (index > 0x7fffffff) {
+        // Strings with 64 bit lengths not supported by this UTF-8 provider.
+        ix = 0x7fffffff;
+    }
+
+    // Pin requested index to the string length.
+    if (ix>length) {
+        if (length>=0) {
+            ix=length;
+        } else if (ix>=ut->c) {
+            // Zero terminated string, and requested index is beyond
+            //   the region that has already been scanned.
+            //   Scan up to either the end of the string or to the
+            //   requested position, whichever comes first.
+            while (ut->c<ix && s8[ut->c]!=0) {
+                ut->c++;
+            }
+            //  TODO:  support for null terminated string length > 32 bits.
+            if (s8[ut->c] == 0) {
+                // We just found the actual length of the string.
+                //  Trim the requested index back to that.
+                ix     = ut->c;
+                ut->b  = ut->c;
+                length = ut->c;
+                ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+            }
+        }
+    }
+
+    //
+    // Dispatch to the appropriate action for a forward iteration request.
+    //
+    if (forward) {
+        if (ix==ut->chunkNativeLimit) {
+            // Check for normal sequential iteration cases first.
+            if (ix==length) {
+                // Just reached end of string
+                // Don't swap buffers, but do set the
+                //   current buffer position.
+                ut->chunkOffset = ut->chunkLength;
+                return FALSE;
+            } else {
+                // End of current buffer.
+                //   check whether other buffer already has what we need.
+                UTF8Buf *altB = (UTF8Buf *)ut->q;
+                if (ix>=altB->bufNativeStart && ix<altB->bufNativeLimit) {
+                    goto swapBuffers;
+                }
+            }
+        }
+
+        // A random access.  Desired index could be in either or niether buf.
+        // For optimizing the order of testing, first check for the index
+        //    being in the other buffer.  This will be the case for uses that
+        //    move back and forth over a fairly limited range
+        {
+            u8b = (UTF8Buf *)ut->q;   // the alternate buffer
+            if (ix>=u8b->bufNativeStart && ix<u8b->bufNativeLimit) {
+                // Requested index is in the other buffer.
+                goto swapBuffers;
+            }
+            if (ix == length) {
+                // Requested index is end-of-string.
+                //   (this is the case of randomly seeking to the end.
+                //    The case of iterating off the end is handled earlier.)
+                if (ix == ut->chunkNativeLimit) {
+                    // Current buffer extends up to the end of the string.
+                    //   Leave it as the current buffer.
+                    ut->chunkOffset = ut->chunkLength;
+                    return FALSE;
+                }
+                if (ix == u8b->bufNativeLimit) {
+                    // Alternate buffer extends to the end of string.
+                    //   Swap it in as the current buffer.
+                    goto swapBuffersAndFail;
+                }
+
+                // Neither existing buffer extends to the end of the string.
+                goto makeStubBuffer;
+            }
+
+            if (ix<ut->chunkNativeStart || ix>=ut->chunkNativeLimit) {
+                // Requested index is in neither buffer.
+                goto fillForward;
+            }
+
+            // Requested index is in this buffer.
+            u8b = (UTF8Buf *)ut->p;   // the current buffer
+            mapIndex = ix - u8b->toUCharsMapStart;
+            U_ASSERT(mapIndex < (int32_t)sizeof(UTF8Buf::mapToUChars));
+            ut->chunkOffset = u8b->mapToUChars[mapIndex] - u8b->bufStartIdx;
+            return TRUE;
+
+        }
+    }
+
+
+    //
+    // Dispatch to the appropriate action for a
+    //   Backwards Diretion iteration request.
+    //
+    if (ix==ut->chunkNativeStart) {
+        // Check for normal sequential iteration cases first.
+        if (ix==0) {
+            // Just reached the start of string
+            // Don't swap buffers, but do set the
+            //   current buffer position.
+            ut->chunkOffset = 0;
+            return FALSE;
+        } else {
+            // Start of current buffer.
+            //   check whether other buffer already has what we need.
+            UTF8Buf *altB = (UTF8Buf *)ut->q;
+            if (ix>altB->bufNativeStart && ix<=altB->bufNativeLimit) {
+                goto swapBuffers;
+            }
+        }
+    }
+
+    // A random access.  Desired index could be in either or niether buf.
+    // For optimizing the order of testing,
+    //    Most likely case:  in the other buffer.
+    //    Second most likely: in neither buffer.
+    //    Unlikely, but must work:  in the current buffer.
+    u8b = (UTF8Buf *)ut->q;   // the alternate buffer
+    if (ix>u8b->bufNativeStart && ix<=u8b->bufNativeLimit) {
+        // Requested index is in the other buffer.
+        goto swapBuffers;
+    }
+    // Requested index is start-of-string.
+    //   (this is the case of randomly seeking to the start.
+    //    The case of iterating off the start is handled earlier.)
+    if (ix==0) {
+        if (u8b->bufNativeStart==0) {
+            // Alternate buffer contains the data for the start string.
+            // Make it be the current buffer.
+            goto swapBuffersAndFail;
+        } else {
+            // Request for data before the start of string,
+            //   neither buffer is usable.
+            //   set up a zero-length buffer.
+            goto makeStubBuffer;
+        }
+    }
+
+    if (ix<=ut->chunkNativeStart || ix>ut->chunkNativeLimit) {
+        // Requested index is in neither buffer.
+        goto fillReverse;
+    }
+
+    // Requested index is in this buffer.
+    //   Set the utf16 buffer index.
+    u8b = (UTF8Buf *)ut->p;
+    mapIndex = ix - u8b->toUCharsMapStart;
+    ut->chunkOffset = u8b->mapToUChars[mapIndex] - u8b->bufStartIdx;
+    if (ut->chunkOffset==0) {
+        // This occurs when the first character in the text is
+        //   a multi-byte UTF-8 char, and the requested index is to
+        //   one of the trailing bytes.  Because there is no preceding ,
+        //   character, this access fails.  We can't pick up on the
+        //   situation sooner because the requested index is not zero.
+        return FALSE;
+    } else {
+        return TRUE;
+    }
+
+
+
+swapBuffers:
+    //  The alternate buffer (ut->q) has the string data that was requested.
+    //  Swap the primary and alternate buffers, and set the
+    //   chunk index into the new primary buffer.
+    {
+        u8b   = (UTF8Buf *)ut->q;
+        ut->q = ut->p;
+        ut->p = u8b;
+        ut->chunkContents       = &u8b->buf[u8b->bufStartIdx];
+        ut->chunkLength         = u8b->bufLimitIdx - u8b->bufStartIdx;
+        ut->chunkNativeStart    = u8b->bufNativeStart;
+        ut->chunkNativeLimit    = u8b->bufNativeLimit;
+        ut->nativeIndexingLimit = u8b->bufNILimit;
+
+        // Index into the (now current) chunk
+        // Use the map to set the chunk index.  It's more trouble than it's worth
+        //    to check whether native indexing can be used.
+        U_ASSERT(ix>=u8b->bufNativeStart);
+        U_ASSERT(ix<=u8b->bufNativeLimit);
+        mapIndex = ix - u8b->toUCharsMapStart;
+        U_ASSERT(mapIndex>=0);
+        U_ASSERT(mapIndex<(int32_t)sizeof(u8b->mapToUChars));
+        ut->chunkOffset = u8b->mapToUChars[mapIndex] - u8b->bufStartIdx;
+
+        return TRUE;
+    }
+
+
+ swapBuffersAndFail:
+    // We got a request for either the start or end of the string,
+    //  with iteration continuing in the out-of-bounds direction.
+    // The alternate buffer already contains the data up to the
+    //  start/end.
+    // Swap the buffers, then return failure, indicating that we couldn't
+    //  make things correct for continuing the iteration in the requested
+    //  direction.  The position & buffer are correct should the
+    //  user decide to iterate in the opposite direction.
+    u8b   = (UTF8Buf *)ut->q;
+    ut->q = ut->p;
+    ut->p = u8b;
+    ut->chunkContents       = &u8b->buf[u8b->bufStartIdx];
+    ut->chunkLength         = u8b->bufLimitIdx - u8b->bufStartIdx;
+    ut->chunkNativeStart    = u8b->bufNativeStart;
+    ut->chunkNativeLimit    = u8b->bufNativeLimit;
+    ut->nativeIndexingLimit = u8b->bufNILimit;
+
+    // Index into the (now current) chunk
+    //  For this function  (swapBuffersAndFail), the requested index
+    //    will always be at either the start or end of the chunk.
+    if (ix==u8b->bufNativeLimit) {
+        ut->chunkOffset = ut->chunkLength;
+    } else  {
+        ut->chunkOffset = 0;
+        U_ASSERT(ix == u8b->bufNativeStart);
+    }
+    return FALSE;
+
+makeStubBuffer:
+    //   The user has done a seek/access past the start or end
+    //   of the string.  Rather than loading data that is likely
+    //   to never be used, just set up a zero-length buffer at
+    //   the position.
+    u8b = (UTF8Buf *)ut->q;
+    u8b->bufNativeStart   = ix;
+    u8b->bufNativeLimit   = ix;
+    u8b->bufStartIdx      = 0;
+    u8b->bufLimitIdx      = 0;
+    u8b->bufNILimit       = 0;
+    u8b->toUCharsMapStart = ix;
+    u8b->mapToNative[0]   = 0;
+    u8b->mapToUChars[0]   = 0;
+    goto swapBuffersAndFail;
+
+
+
+fillForward:
+    {
+        // Move the incoming index to a code point boundary.
+        U8_SET_CP_START(s8, 0, ix);
+
+        // Swap the UText buffers.
+        //  We want to fill what was previously the alternate buffer,
+        //  and make what was the current buffer be the new alternate.
+        UTF8Buf *u8b_swap = (UTF8Buf *)ut->q;
+        ut->q = ut->p;
+        ut->p = u8b_swap;
+
+        int32_t strLen = ut->b;
+        UBool   nulTerminated = FALSE;
+        if (strLen < 0) {
+            strLen = 0x7fffffff;
+            nulTerminated = TRUE;
+        }
+
+        UChar   *buf = u8b_swap->buf;
+        uint8_t *mapToNative  = u8b_swap->mapToNative;
+        uint8_t *mapToUChars  = u8b_swap->mapToUChars;
+        int32_t  destIx       = 0;
+        int32_t  srcIx        = ix;
+        UBool    seenNonAscii = FALSE;
+        UChar32  c = 0;
+
+        // Fill the chunk buffer and mapping arrays.
+        while (destIx<UTF8_TEXT_CHUNK_SIZE) {
+            c = s8[srcIx];
+            if (c>0 && c<0x80) {
+                // Special case ASCII range for speed.
+                //   zero is excluded to simplify bounds checking.
+                buf[destIx] = (UChar)c;
+                mapToNative[destIx]    = (uint8_t)(srcIx - ix);
+                mapToUChars[srcIx-ix]  = (uint8_t)destIx;
+                srcIx++;
+                destIx++;
+            } else {
+                // General case, handle everything.
+                if (seenNonAscii == FALSE) {
+                    seenNonAscii = TRUE;
+                    u8b_swap->bufNILimit = destIx;
+                }
+
+                int32_t  cIx      = srcIx;
+                int32_t  dIx      = destIx;
+                int32_t  dIxSaved = destIx;
+                U8_NEXT_OR_FFFD(s8, srcIx, strLen, c);
+                if (c==0 && nulTerminated) {
+                    srcIx--;
+                    break;
+                }
+
+                U16_APPEND_UNSAFE(buf, destIx, c);
+                do {
+                    mapToNative[dIx++] = (uint8_t)(cIx - ix);
+                } while (dIx < destIx);
+
+                do {
+                    mapToUChars[cIx++ - ix] = (uint8_t)dIxSaved;
+                } while (cIx < srcIx);
+            }
+            if (srcIx>=strLen) {
+                break;
+            }
+
+        }
+
+        //  store Native <--> Chunk Map entries for the end of the buffer.
+        //    There is no actual character here, but the index position is valid.
+        mapToNative[destIx]     = (uint8_t)(srcIx - ix);
+        mapToUChars[srcIx - ix] = (uint8_t)destIx;
+
+        //  fill in Buffer descriptor
+        u8b_swap->bufNativeStart     = ix;
+        u8b_swap->bufNativeLimit     = srcIx;
+        u8b_swap->bufStartIdx        = 0;
+        u8b_swap->bufLimitIdx        = destIx;
+        if (seenNonAscii == FALSE) {
+            u8b_swap->bufNILimit     = destIx;
+        }
+        u8b_swap->toUCharsMapStart   = u8b_swap->bufNativeStart;
+
+        // Set UText chunk to refer to this buffer.
+        ut->chunkContents       = buf;
+        ut->chunkOffset         = 0;
+        ut->chunkLength         = u8b_swap->bufLimitIdx;
+        ut->chunkNativeStart    = u8b_swap->bufNativeStart;
+        ut->chunkNativeLimit    = u8b_swap->bufNativeLimit;
+        ut->nativeIndexingLimit = u8b_swap->bufNILimit;
+
+        // For zero terminated strings, keep track of the maximum point
+        //   scanned so far.
+        if (nulTerminated && srcIx>ut->c) {
+            ut->c = srcIx;
+            if (c==0) {
+                // We scanned to the end.
+                //   Remember the actual length.
+                ut->b = srcIx;
+                ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+            }
+        }
+        return TRUE;
+    }
+
+
+fillReverse:
+    {
+        // Move the incoming index to a code point boundary.
+        // Can only do this if the incoming index is somewhere in the interior of the string.
+        //   If index is at the end, there is no character there to look at.
+        if (ix != ut->b) {
+            // Note: this function will only move the index back if it is on a trail byte
+            //       and there is a preceding lead byte and the sequence from the lead 
+            //       through this trail could be part of a valid UTF-8 sequence
+            //       Otherwise the index remains unchanged.
+            U8_SET_CP_START(s8, 0, ix);
+        }
+
+        // Swap the UText buffers.
+        //  We want to fill what was previously the alternate buffer,
+        //  and make what was the current buffer be the new alternate.
+        UTF8Buf *u8b_swap = (UTF8Buf *)ut->q;
+        ut->q = ut->p;
+        ut->p = u8b_swap;
+
+        UChar   *buf = u8b_swap->buf;
+        uint8_t *mapToNative = u8b_swap->mapToNative;
+        uint8_t *mapToUChars = u8b_swap->mapToUChars;
+        int32_t  toUCharsMapStart = ix - sizeof(UTF8Buf::mapToUChars) + 1;
+        // Note that toUCharsMapStart can be negative. Happens when the remaining
+        // text from current position to the beginning is less than the buffer size.
+        // + 1 because mapToUChars must have a slot at the end for the bufNativeLimit entry.
+        int32_t  destIx = UTF8_TEXT_CHUNK_SIZE+2;   // Start in the overflow region
+                                                    //   at end of buffer to leave room
+                                                    //   for a surrogate pair at the
+                                                    //   buffer start.
+        int32_t  srcIx  = ix;
+        int32_t  bufNILimit = destIx;
+        UChar32   c;
+
+        // Map to/from Native Indexes, fill in for the position at the end of
+        //   the buffer.
+        //
+        mapToNative[destIx] = (uint8_t)(srcIx - toUCharsMapStart);
+        mapToUChars[srcIx - toUCharsMapStart] = (uint8_t)destIx;
+
+        // Fill the chunk buffer
+        // Work backwards, filling from the end of the buffer towards the front.
+        //
+        while (destIx>2 && (srcIx - toUCharsMapStart > 5) && (srcIx > 0)) {
+            srcIx--;
+            destIx--;
+
+            // Get last byte of the UTF-8 character
+            c = s8[srcIx];
+            if (c<0x80) {
+                // Special case ASCII range for speed.
+                buf[destIx] = (UChar)c;
+                U_ASSERT(toUCharsMapStart <= srcIx);
+                mapToUChars[srcIx - toUCharsMapStart] = (uint8_t)destIx;
+                mapToNative[destIx] = (uint8_t)(srcIx - toUCharsMapStart);
+            } else {
+                // General case, handle everything non-ASCII.
+
+                int32_t  sIx      = srcIx;  // ix of last byte of multi-byte u8 char
+
+                // Get the full character from the UTF8 string.
+                //   use code derived from tbe macros in utf8.h
+                //   Leaves srcIx pointing at the first byte of the UTF-8 char.
+                //
+                c=utf8_prevCharSafeBody(s8, 0, &srcIx, c, -3);
+                // leaves srcIx at first byte of the multi-byte char.
+
+                // Store the character in UTF-16 buffer.
+                if (c<0x10000) {
+                    buf[destIx] = (UChar)c;
+                    mapToNative[destIx] = (uint8_t)(srcIx - toUCharsMapStart);
+                } else {
+                    buf[destIx]         = U16_TRAIL(c);
+                    mapToNative[destIx] = (uint8_t)(srcIx - toUCharsMapStart);
+                    buf[--destIx]       = U16_LEAD(c);
+                    mapToNative[destIx] = (uint8_t)(srcIx - toUCharsMapStart);
+                }
+
+                // Fill in the map from native indexes to UChars buf index.
+                do {
+                    mapToUChars[sIx-- - toUCharsMapStart] = (uint8_t)destIx;
+                } while (sIx >= srcIx);
+                U_ASSERT(toUCharsMapStart <= (srcIx+1));
+
+                // Set native indexing limit to be the current position.
+                //   We are processing a non-ascii, non-native-indexing char now;
+                //     the limit will be here if the rest of the chars to be
+                //     added to this buffer are ascii.
+                bufNILimit = destIx;
+            }
+        }
+        u8b_swap->bufNativeStart     = srcIx;
+        u8b_swap->bufNativeLimit     = ix;
+        u8b_swap->bufStartIdx        = destIx;
+        u8b_swap->bufLimitIdx        = UTF8_TEXT_CHUNK_SIZE+2;
+        u8b_swap->bufNILimit         = bufNILimit - u8b_swap->bufStartIdx;
+        u8b_swap->toUCharsMapStart   = toUCharsMapStart;
+
+        ut->chunkContents       = &buf[u8b_swap->bufStartIdx];
+        ut->chunkLength         = u8b_swap->bufLimitIdx - u8b_swap->bufStartIdx;
+        ut->chunkOffset         = ut->chunkLength;
+        ut->chunkNativeStart    = u8b_swap->bufNativeStart;
+        ut->chunkNativeLimit    = u8b_swap->bufNativeLimit;
+        ut->nativeIndexingLimit = u8b_swap->bufNILimit;
+        return TRUE;
+    }
+
+}
+
+
+
+//
+//  This is a slightly modified copy of u_strFromUTF8,
+//     Inserts a Replacement Char rather than failing on invalid UTF-8
+//     Removes unnecessary features.
+//
+static UChar*
+utext_strFromUTF8(UChar *dest,
+              int32_t destCapacity,
+              int32_t *pDestLength,
+              const char* src,
+              int32_t srcLength,        // required.  NUL terminated not supported.
+              UErrorCode *pErrorCode
+              )
+{
+
+    UChar *pDest = dest;
+    UChar *pDestLimit = (dest!=NULL)?(dest+destCapacity):NULL;
+    UChar32 ch=0;
+    int32_t index = 0;
+    int32_t reqLength = 0;
+    uint8_t* pSrc = (uint8_t*) src;
+
+
+    while((index < srcLength)&&(pDest<pDestLimit)){
+        ch = pSrc[index++];
+        if(ch <=0x7f){
+            *pDest++=(UChar)ch;
+        }else{
+            ch=utf8_nextCharSafeBody(pSrc, &index, srcLength, ch, -3);
+            if(U_IS_BMP(ch)){
+                *(pDest++)=(UChar)ch;
+            }else{
+                *(pDest++)=U16_LEAD(ch);
+                if(pDest<pDestLimit){
+                    *(pDest++)=U16_TRAIL(ch);
+                }else{
+                    reqLength++;
+                    break;
+                }
+            }
+        }
+    }
+    /* donot fill the dest buffer just count the UChars needed */
+    while(index < srcLength){
+        ch = pSrc[index++];
+        if(ch <= 0x7f){
+            reqLength++;
+        }else{
+            ch=utf8_nextCharSafeBody(pSrc, &index, srcLength, ch, -3);
+            reqLength+=U16_LENGTH(ch);
+        }
+    }
+
+    reqLength+=(int32_t)(pDest - dest);
+
+    if(pDestLength){
+        *pDestLength = reqLength;
+    }
+
+    /* Terminate the buffer */
+    u_terminateUChars(dest,destCapacity,reqLength,pErrorCode);
+
+    return dest;
+}
+
+
+
+static int32_t U_CALLCONV
+utf8TextExtract(UText *ut,
+                int64_t start, int64_t limit,
+                UChar *dest, int32_t destCapacity,
+                UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    int32_t  length  = ut->b;
+    int32_t  start32 = pinIndex(start, length);
+    int32_t  limit32 = pinIndex(limit, length);
+
+    if(start32>limit32) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+
+    // adjust the incoming indexes to land on code point boundaries if needed.
+    //    adjust by no more than three, because that is the largest number of trail bytes
+    //    in a well formed UTF8 character.
+    const uint8_t *buf = (const uint8_t *)ut->context;
+    int i;
+    if (start32 < ut->chunkNativeLimit) {
+        for (i=0; i<3; i++) {
+            if (U8_IS_SINGLE(buf[start32]) || U8_IS_LEAD(buf[start32]) || start32==0) {
+                break;
+            }
+            start32--;
+        }
+    }
+
+    if (limit32 < ut->chunkNativeLimit) {
+        for (i=0; i<3; i++) {
+            if (U8_IS_SINGLE(buf[limit32]) || U8_IS_LEAD(buf[limit32]) || limit32==0) {
+                break;
+            }
+            limit32--;
+        }
+    }
+
+    // Do the actual extract.
+    int32_t destLength=0;
+    utext_strFromUTF8(dest, destCapacity, &destLength,
+                    (const char *)ut->context+start32, limit32-start32,
+                    pErrorCode);
+    utf8TextAccess(ut, limit32, TRUE);
+    return destLength;
+}
+
+//
+// utf8TextMapOffsetToNative
+//
+// Map a chunk (UTF-16) offset to a native index.
+static int64_t U_CALLCONV
+utf8TextMapOffsetToNative(const UText *ut) {
+    //
+    UTF8Buf *u8b = (UTF8Buf *)ut->p;
+    U_ASSERT(ut->chunkOffset>ut->nativeIndexingLimit && ut->chunkOffset<=ut->chunkLength);
+    int32_t nativeOffset = u8b->mapToNative[ut->chunkOffset + u8b->bufStartIdx] + u8b->toUCharsMapStart;
+    U_ASSERT(nativeOffset >= ut->chunkNativeStart && nativeOffset <= ut->chunkNativeLimit);
+    return nativeOffset;
+}
+
+//
+// Map a native index to the corrsponding chunk offset
+//
+static int32_t U_CALLCONV
+utf8TextMapIndexToUTF16(const UText *ut, int64_t index64) {
+    U_ASSERT(index64 <= 0x7fffffff);
+    int32_t index = (int32_t)index64;
+    UTF8Buf *u8b = (UTF8Buf *)ut->p;
+    U_ASSERT(index>=ut->chunkNativeStart+ut->nativeIndexingLimit);
+    U_ASSERT(index<=ut->chunkNativeLimit);
+    int32_t mapIndex = index - u8b->toUCharsMapStart;
+    U_ASSERT(mapIndex < (int32_t)sizeof(UTF8Buf::mapToUChars));
+    int32_t offset = u8b->mapToUChars[mapIndex] - u8b->bufStartIdx;
+    U_ASSERT(offset>=0 && offset<=ut->chunkLength);
+    return offset;
+}
+
+static UText * U_CALLCONV
+utf8TextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status)
+{
+    // First do a generic shallow clone.  Does everything needed for the UText struct itself.
+    dest = shallowTextClone(dest, src, status);
+
+    // For deep clones, make a copy of the string.
+    //  The copied storage is owned by the newly created clone.
+    //
+    // TODO:  There is an isssue with using utext_nativeLength().
+    //        That function is non-const in cases where the input was NUL terminated
+    //          and the length has not yet been determined.
+    //        This function (clone()) is const.
+    //        There potentially a thread safety issue lurking here.
+    //
+    if (deep && U_SUCCESS(*status)) {
+        int32_t  len = (int32_t)utext_nativeLength((UText *)src);
+        char *copyStr = (char *)uprv_malloc(len+1);
+        if (copyStr == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            uprv_memcpy(copyStr, src->context, len+1);
+            dest->context = copyStr;
+            dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT);
+        }
+    }
+    return dest;
+}
+
+
+static void U_CALLCONV
+utf8TextClose(UText *ut) {
+    // Most of the work of close is done by the generic UText framework close.
+    // All that needs to be done here is to delete the UTF8 string if the UText
+    //  owns it.  This occurs if the UText was created by cloning.
+    if (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT)) {
+        char *s = (char *)ut->context;
+        uprv_free(s);
+        ut->context = NULL;
+    }
+}
+
+U_CDECL_END
+
+
+static const struct UTextFuncs utf8Funcs =
+{
+    sizeof(UTextFuncs),
+    0, 0, 0,             // Reserved alignment padding
+    utf8TextClone,
+    utf8TextLength,
+    utf8TextAccess,
+    utf8TextExtract,
+    NULL,                /* replace*/
+    NULL,                /* copy   */
+    utf8TextMapOffsetToNative,
+    utf8TextMapIndexToUTF16,
+    utf8TextClose,
+    NULL,                // spare 1
+    NULL,                // spare 2
+    NULL                 // spare 3
+};
+
+
+static const char gEmptyString[] = {0};
+
+U_CAPI UText * U_EXPORT2
+utext_openUTF8(UText *ut, const char *s, int64_t length, UErrorCode *status) {
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(s==NULL && length==0) {
+        s = gEmptyString;
+    }
+
+    if(s==NULL || length<-1 || length>INT32_MAX) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    ut = utext_setup(ut, sizeof(UTF8Buf) * 2, status);
+    if (U_FAILURE(*status)) {
+        return ut;
+    }
+
+    ut->pFuncs  = &utf8Funcs;
+    ut->context = s;
+    ut->b       = (int32_t)length;
+    ut->c       = (int32_t)length;
+    if (ut->c < 0) {
+        ut->c = 0;
+        ut->providerProperties |= I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+    }
+    ut->p = ut->pExtra;
+    ut->q = (char *)ut->pExtra + sizeof(UTF8Buf);
+    return ut;
+
+}
+
+
+
+
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//     UText implementation wrapper for Replaceable (read/write)
+//
+//         Use of UText data members:
+//            context    pointer to Replaceable.
+//            p          pointer to Replaceable if it is owned by the UText.
+//
+//------------------------------------------------------------------------------
+
+
+
+// minimum chunk size for this implementation: 3
+// to allow for possible trimming for code point boundaries
+enum { REP_TEXT_CHUNK_SIZE=10 };
+
+struct ReplExtra {
+    /*
+     * Chunk UChars.
+     * +1 to simplify filling with surrogate pair at the end.
+     */
+    UChar s[REP_TEXT_CHUNK_SIZE+1];
+};
+
+
+U_CDECL_BEGIN
+
+static UText * U_CALLCONV
+repTextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status) {
+    // First do a generic shallow clone.  Does everything needed for the UText struct itself.
+    dest = shallowTextClone(dest, src, status);
+
+    // For deep clones, make a copy of the Replaceable.
+    //  The copied Replaceable storage is owned by the newly created UText clone.
+    //  A non-NULL pointer in UText.p is the signal to the close() function to delete
+    //    it.
+    //
+    if (deep && U_SUCCESS(*status)) {
+        const Replaceable *replSrc = (const Replaceable *)src->context;
+        dest->context = replSrc->clone();
+        dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT);
+
+        // with deep clone, the copy is writable, even when the source is not.
+        dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_WRITABLE);
+    }
+    return dest;
+}
+
+
+static void U_CALLCONV
+repTextClose(UText *ut) {
+    // Most of the work of close is done by the generic UText framework close.
+    // All that needs to be done here is delete the Replaceable if the UText
+    //  owns it.  This occurs if the UText was created by cloning.
+    if (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT)) {
+        Replaceable *rep = (Replaceable *)ut->context;
+        delete rep;
+        ut->context = NULL;
+    }
+}
+
+
+static int64_t U_CALLCONV
+repTextLength(UText *ut) {
+    const Replaceable *replSrc = (const Replaceable *)ut->context;
+    int32_t  len = replSrc->length();
+    return len;
+}
+
+
+static UBool U_CALLCONV
+repTextAccess(UText *ut, int64_t index, UBool forward) {
+    const Replaceable *rep=(const Replaceable *)ut->context;
+    int32_t length=rep->length();   // Full length of the input text (bigger than a chunk)
+
+    // clip the requested index to the limits of the text.
+    int32_t index32 = pinIndex(index, length);
+    U_ASSERT(index<=INT32_MAX);
+
+
+    /*
+     * Compute start/limit boundaries around index, for a segment of text
+     * to be extracted.
+     * To allow for the possibility that our user gave an index to the trailing
+     * half of a surrogate pair, we must request one extra preceding UChar when
+     * going in the forward direction.  This will ensure that the buffer has the
+     * entire code point at the specified index.
+     */
+    if(forward) {
+
+        if (index32>=ut->chunkNativeStart && index32<ut->chunkNativeLimit) {
+            // Buffer already contains the requested position.
+            ut->chunkOffset = (int32_t)(index - ut->chunkNativeStart);
+            return TRUE;
+        }
+        if (index32>=length && ut->chunkNativeLimit==length) {
+            // Request for end of string, and buffer already extends up to it.
+            // Can't get the data, but don't change the buffer.
+            ut->chunkOffset = length - (int32_t)ut->chunkNativeStart;
+            return FALSE;
+        }
+
+        ut->chunkNativeLimit = index + REP_TEXT_CHUNK_SIZE - 1;
+        // Going forward, so we want to have the buffer with stuff at and beyond
+        //   the requested index.  The -1 gets us one code point before the
+        //   requested index also, to handle the case of the index being on
+        //   a trail surrogate of a surrogate pair.
+        if(ut->chunkNativeLimit > length) {
+            ut->chunkNativeLimit = length;
+        }
+        // unless buffer ran off end, start is index-1.
+        ut->chunkNativeStart = ut->chunkNativeLimit - REP_TEXT_CHUNK_SIZE;
+        if(ut->chunkNativeStart < 0) {
+            ut->chunkNativeStart = 0;
+        }
+    } else {
+        // Reverse iteration.  Fill buffer with data preceding the requested index.
+        if (index32>ut->chunkNativeStart && index32<=ut->chunkNativeLimit) {
+            // Requested position already in buffer.
+            ut->chunkOffset = index32 - (int32_t)ut->chunkNativeStart;
+            return TRUE;
+        }
+        if (index32==0 && ut->chunkNativeStart==0) {
+            // Request for start, buffer already begins at start.
+            //  No data, but keep the buffer as is.
+            ut->chunkOffset = 0;
+            return FALSE;
+        }
+
+        // Figure out the bounds of the chunk to extract for reverse iteration.
+        // Need to worry about chunk not splitting surrogate pairs, and while still
+        // containing the data we need.
+        // Fix by requesting a chunk that includes an extra UChar at the end.
+        // If this turns out to be a lead surrogate, we can lop it off and still have
+        //   the data we wanted.
+        ut->chunkNativeStart = index32 + 1 - REP_TEXT_CHUNK_SIZE;
+        if (ut->chunkNativeStart < 0) {
+            ut->chunkNativeStart = 0;
+        }
+
+        ut->chunkNativeLimit = index32 + 1;
+        if (ut->chunkNativeLimit > length) {
+            ut->chunkNativeLimit = length;
+        }
+    }
+
+    // Extract the new chunk of text from the Replaceable source.
+    ReplExtra *ex = (ReplExtra *)ut->pExtra;
+    // UnicodeString with its buffer a writable alias to the chunk buffer
+    UnicodeString buffer(ex->s, 0 /*buffer length*/, REP_TEXT_CHUNK_SIZE /*buffer capacity*/);
+    rep->extractBetween((int32_t)ut->chunkNativeStart, (int32_t)ut->chunkNativeLimit, buffer);
+
+    ut->chunkContents  = ex->s;
+    ut->chunkLength    = (int32_t)(ut->chunkNativeLimit - ut->chunkNativeStart);
+    ut->chunkOffset    = (int32_t)(index32 - ut->chunkNativeStart);
+
+    // Surrogate pairs from the input text must not span chunk boundaries.
+    // If end of chunk could be the start of a surrogate, trim it off.
+    if (ut->chunkNativeLimit < length &&
+        U16_IS_LEAD(ex->s[ut->chunkLength-1])) {
+            ut->chunkLength--;
+            ut->chunkNativeLimit--;
+            if (ut->chunkOffset > ut->chunkLength) {
+                ut->chunkOffset = ut->chunkLength;
+            }
+        }
+
+    // if the first UChar in the chunk could be the trailing half of a surrogate pair,
+    // trim it off.
+    if(ut->chunkNativeStart>0 && U16_IS_TRAIL(ex->s[0])) {
+        ++(ut->chunkContents);
+        ++(ut->chunkNativeStart);
+        --(ut->chunkLength);
+        --(ut->chunkOffset);
+    }
+
+    // adjust the index/chunkOffset to a code point boundary
+    U16_SET_CP_START(ut->chunkContents, 0, ut->chunkOffset);
+
+    // Use fast indexing for get/setNativeIndex()
+    ut->nativeIndexingLimit = ut->chunkLength;
+
+    return TRUE;
+}
+
+
+
+static int32_t U_CALLCONV
+repTextExtract(UText *ut,
+               int64_t start, int64_t limit,
+               UChar *dest, int32_t destCapacity,
+               UErrorCode *status) {
+    const Replaceable *rep=(const Replaceable *)ut->context;
+    int32_t  length=rep->length();
+
+    if(U_FAILURE(*status)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0)) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    if(start>limit) {
+        *status=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    int32_t  start32 = pinIndex(start, length);
+    int32_t  limit32 = pinIndex(limit, length);
+
+    // adjust start, limit if they point to trail half of surrogates
+    if (start32<length && U16_IS_TRAIL(rep->charAt(start32)) &&
+        U_IS_SUPPLEMENTARY(rep->char32At(start32))){
+            start32--;
+    }
+    if (limit32<length && U16_IS_TRAIL(rep->charAt(limit32)) &&
+        U_IS_SUPPLEMENTARY(rep->char32At(limit32))){
+            limit32--;
+    }
+
+    length=limit32-start32;
+    if(length>destCapacity) {
+        limit32 = start32 + destCapacity;
+    }
+    UnicodeString buffer(dest, 0, destCapacity); // writable alias
+    rep->extractBetween(start32, limit32, buffer);
+    repTextAccess(ut, limit32, TRUE);
+
+    return u_terminateUChars(dest, destCapacity, length, status);
+}
+
+static int32_t U_CALLCONV
+repTextReplace(UText *ut,
+               int64_t start, int64_t limit,
+               const UChar *src, int32_t length,
+               UErrorCode *status) {
+    Replaceable *rep=(Replaceable *)ut->context;
+    int32_t oldLength;
+
+    if(U_FAILURE(*status)) {
+        return 0;
+    }
+    if(src==NULL && length!=0) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    oldLength=rep->length(); // will subtract from new length
+    if(start>limit ) {
+        *status=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    int32_t start32 = pinIndex(start, oldLength);
+    int32_t limit32 = pinIndex(limit, oldLength);
+
+    // Snap start & limit to code point boundaries.
+    if (start32<oldLength && U16_IS_TRAIL(rep->charAt(start32)) &&
+        start32>0 && U16_IS_LEAD(rep->charAt(start32-1)))
+    {
+            start32--;
+    }
+    if (limit32<oldLength && U16_IS_LEAD(rep->charAt(limit32-1)) &&
+        U16_IS_TRAIL(rep->charAt(limit32)))
+    {
+            limit32++;
+    }
+
+    // Do the actual replace operation using methods of the Replaceable class
+    UnicodeString replStr((UBool)(length<0), src, length); // read-only alias
+    rep->handleReplaceBetween(start32, limit32, replStr);
+    int32_t newLength = rep->length();
+    int32_t lengthDelta = newLength - oldLength;
+
+    // Is the UText chunk buffer OK?
+    if (ut->chunkNativeLimit > start32) {
+        // this replace operation may have impacted the current chunk.
+        // invalidate it, which will force a reload on the next access.
+        invalidateChunk(ut);
+    }
+
+    // set the iteration position to the end of the newly inserted replacement text.
+    int32_t newIndexPos = limit32 + lengthDelta;
+    repTextAccess(ut, newIndexPos, TRUE);
+
+    return lengthDelta;
+}
+
+
+static void U_CALLCONV
+repTextCopy(UText *ut,
+                int64_t start, int64_t limit,
+                int64_t destIndex,
+                UBool move,
+                UErrorCode *status)
+{
+    Replaceable *rep=(Replaceable *)ut->context;
+    int32_t length=rep->length();
+
+    if(U_FAILURE(*status)) {
+        return;
+    }
+    if (start>limit || (start<destIndex && destIndex<limit))
+    {
+        *status=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+
+    int32_t start32     = pinIndex(start, length);
+    int32_t limit32     = pinIndex(limit, length);
+    int32_t destIndex32 = pinIndex(destIndex, length);
+
+    // TODO:  snap input parameters to code point boundaries.
+
+    if(move) {
+        // move: copy to destIndex, then replace original with nothing
+        int32_t segLength=limit32-start32;
+        rep->copy(start32, limit32, destIndex32);
+        if(destIndex32<start32) {
+            start32+=segLength;
+            limit32+=segLength;
+        }
+        rep->handleReplaceBetween(start32, limit32, UnicodeString());
+    } else {
+        // copy
+        rep->copy(start32, limit32, destIndex32);
+    }
+
+    // If the change to the text touched the region in the chunk buffer,
+    //  invalidate the buffer.
+    int32_t firstAffectedIndex = destIndex32;
+    if (move && start32<firstAffectedIndex) {
+        firstAffectedIndex = start32;
+    }
+    if (firstAffectedIndex < ut->chunkNativeLimit) {
+        // changes may have affected range covered by the chunk
+        invalidateChunk(ut);
+    }
+
+    // Put iteration position at the newly inserted (moved) block,
+    int32_t  nativeIterIndex = destIndex32 + limit32 - start32;
+    if (move && destIndex32>start32) {
+        // moved a block of text towards the end of the string.
+        nativeIterIndex = destIndex32;
+    }
+
+    // Set position, reload chunk if needed.
+    repTextAccess(ut, nativeIterIndex, TRUE);
+}
+
+static const struct UTextFuncs repFuncs =
+{
+    sizeof(UTextFuncs),
+    0, 0, 0,           // Reserved alignment padding
+    repTextClone,
+    repTextLength,
+    repTextAccess,
+    repTextExtract,
+    repTextReplace,
+    repTextCopy,
+    NULL,              // MapOffsetToNative,
+    NULL,              // MapIndexToUTF16,
+    repTextClose,
+    NULL,              // spare 1
+    NULL,              // spare 2
+    NULL               // spare 3
+};
+
+
+U_CAPI UText * U_EXPORT2
+utext_openReplaceable(UText *ut, Replaceable *rep, UErrorCode *status)
+{
+    if(U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(rep==NULL) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    ut = utext_setup(ut, sizeof(ReplExtra), status);
+    if(U_FAILURE(*status)) {
+        return ut;
+    }
+
+    ut->providerProperties = I32_FLAG(UTEXT_PROVIDER_WRITABLE);
+    if(rep->hasMetaData()) {
+        ut->providerProperties |=I32_FLAG(UTEXT_PROVIDER_HAS_META_DATA);
+    }
+
+    ut->pFuncs  = &repFuncs;
+    ut->context =  rep;
+    return ut;
+}
+
+U_CDECL_END
+
+
+
+
+
+
+
+
+//------------------------------------------------------------------------------
+//
+//     UText implementation for UnicodeString (read/write)  and
+//                    for const UnicodeString (read only)
+//             (same implementation, only the flags are different)
+//
+//         Use of UText data members:
+//            context    pointer to UnicodeString
+//            p          pointer to UnicodeString IF this UText owns the string
+//                       and it must be deleted on close().  NULL otherwise.
+//
+//------------------------------------------------------------------------------
+
+U_CDECL_BEGIN
+
+
+static UText * U_CALLCONV
+unistrTextClone(UText *dest, const UText *src, UBool deep, UErrorCode *status) {
+    // First do a generic shallow clone.  Does everything needed for the UText struct itself.
+    dest = shallowTextClone(dest, src, status);
+
+    // For deep clones, make a copy of the UnicodeSring.
+    //  The copied UnicodeString storage is owned by the newly created UText clone.
+    //  A non-NULL pointer in UText.p is the signal to the close() function to delete
+    //    the UText.
+    //
+    if (deep && U_SUCCESS(*status)) {
+        const UnicodeString *srcString = (const UnicodeString *)src->context;
+        dest->context = new UnicodeString(*srcString);
+        dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT);
+
+        // with deep clone, the copy is writable, even when the source is not.
+        dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_WRITABLE);
+    }
+    return dest;
+}
+
+static void U_CALLCONV
+unistrTextClose(UText *ut) {
+    // Most of the work of close is done by the generic UText framework close.
+    // All that needs to be done here is delete the UnicodeString if the UText
+    //  owns it.  This occurs if the UText was created by cloning.
+    if (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT)) {
+        UnicodeString *str = (UnicodeString *)ut->context;
+        delete str;
+        ut->context = NULL;
+    }
+}
+
+
+static int64_t U_CALLCONV
+unistrTextLength(UText *t) {
+    return ((const UnicodeString *)t->context)->length();
+}
+
+
+static UBool U_CALLCONV
+unistrTextAccess(UText *ut, int64_t index, UBool  forward) {
+    int32_t length  = ut->chunkLength;
+    ut->chunkOffset = pinIndex(index, length);
+
+    // Check whether request is at the start or end
+    UBool retVal = (forward && index<length) || (!forward && index>0);
+    return retVal;
+}
+
+
+
+static int32_t U_CALLCONV
+unistrTextExtract(UText *t,
+                  int64_t start, int64_t limit,
+                  UChar *dest, int32_t destCapacity,
+                  UErrorCode *pErrorCode) {
+    const UnicodeString *us=(const UnicodeString *)t->context;
+    int32_t length=us->length();
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    if(start<0 || start>limit) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    int32_t start32 = start<length ? us->getChar32Start((int32_t)start) : length;
+    int32_t limit32 = limit<length ? us->getChar32Start((int32_t)limit) : length;
+
+    length=limit32-start32;
+    if (destCapacity>0 && dest!=NULL) {
+        int32_t trimmedLength = length;
+        if(trimmedLength>destCapacity) {
+            trimmedLength=destCapacity;
+        }
+        us->extract(start32, trimmedLength, dest);
+        t->chunkOffset = start32+trimmedLength;
+    } else {
+        t->chunkOffset = start32;
+    }
+    u_terminateUChars(dest, destCapacity, length, pErrorCode);
+    return length;
+}
+
+static int32_t U_CALLCONV
+unistrTextReplace(UText *ut,
+                  int64_t start, int64_t limit,
+                  const UChar *src, int32_t length,
+                  UErrorCode *pErrorCode) {
+    UnicodeString *us=(UnicodeString *)ut->context;
+    int32_t oldLength;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(src==NULL && length!=0) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+    }
+    if(start>limit) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+    oldLength=us->length();
+    int32_t start32 = pinIndex(start, oldLength);
+    int32_t limit32 = pinIndex(limit, oldLength);
+    if (start32 < oldLength) {
+        start32 = us->getChar32Start(start32);
+    }
+    if (limit32 < oldLength) {
+        limit32 = us->getChar32Start(limit32);
+    }
+
+    // replace
+    us->replace(start32, limit32-start32, src, length);
+    int32_t newLength = us->length();
+
+    // Update the chunk description.
+    ut->chunkContents    = us->getBuffer();
+    ut->chunkLength      = newLength;
+    ut->chunkNativeLimit = newLength;
+    ut->nativeIndexingLimit = newLength;
+
+    // Set iteration position to the point just following the newly inserted text.
+    int32_t lengthDelta = newLength - oldLength;
+    ut->chunkOffset = limit32 + lengthDelta;
+
+    return lengthDelta;
+}
+
+static void U_CALLCONV
+unistrTextCopy(UText *ut,
+               int64_t start, int64_t limit,
+               int64_t destIndex,
+               UBool move,
+               UErrorCode *pErrorCode) {
+    UnicodeString *us=(UnicodeString *)ut->context;
+    int32_t length=us->length();
+
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    int32_t start32 = pinIndex(start, length);
+    int32_t limit32 = pinIndex(limit, length);
+    int32_t destIndex32 = pinIndex(destIndex, length);
+
+    if( start32>limit32 || (start32<destIndex32 && destIndex32<limit32)) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+
+    if(move) {
+        // move: copy to destIndex, then remove original
+        int32_t segLength=limit32-start32;
+        us->copy(start32, limit32, destIndex32);
+        if(destIndex32<start32) {
+            start32+=segLength;
+        }
+        us->remove(start32, segLength);
+    } else {
+        // copy
+        us->copy(start32, limit32, destIndex32);
+    }
+
+    // update chunk description, set iteration position.
+    ut->chunkContents = us->getBuffer();
+    if (move==FALSE) {
+        // copy operation, string length grows
+        ut->chunkLength += limit32-start32;
+        ut->chunkNativeLimit = ut->chunkLength;
+        ut->nativeIndexingLimit = ut->chunkLength;
+    }
+
+    // Iteration position to end of the newly inserted text.
+    ut->chunkOffset = destIndex32+limit32-start32;
+    if (move && destIndex32>start32) {
+        ut->chunkOffset = destIndex32;
+    }
+
+}
+
+static const struct UTextFuncs unistrFuncs =
+{
+    sizeof(UTextFuncs),
+    0, 0, 0,             // Reserved alignment padding
+    unistrTextClone,
+    unistrTextLength,
+    unistrTextAccess,
+    unistrTextExtract,
+    unistrTextReplace,
+    unistrTextCopy,
+    NULL,                // MapOffsetToNative,
+    NULL,                // MapIndexToUTF16,
+    unistrTextClose,
+    NULL,                // spare 1
+    NULL,                // spare 2
+    NULL                 // spare 3
+};
+
+
+
+U_CDECL_END
+
+
+U_CAPI UText * U_EXPORT2
+utext_openUnicodeString(UText *ut, UnicodeString *s, UErrorCode *status) {
+    ut = utext_openConstUnicodeString(ut, s, status);
+    if (U_SUCCESS(*status)) {
+        ut->providerProperties |= I32_FLAG(UTEXT_PROVIDER_WRITABLE);
+    }
+    return ut;
+}
+
+
+
+U_CAPI UText * U_EXPORT2
+utext_openConstUnicodeString(UText *ut, const UnicodeString *s, UErrorCode *status) {
+    if (U_SUCCESS(*status) && s->isBogus()) {
+        // The UnicodeString is bogus, but we still need to detach the UText
+        //   from whatever it was hooked to before, if anything.
+        utext_openUChars(ut, NULL, 0, status);
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return ut;
+    }
+    ut = utext_setup(ut, 0, status);
+    //    note:  use the standard (writable) function table for UnicodeString.
+    //           The flag settings disable writing, so having the functions in
+    //           the table is harmless.
+    if (U_SUCCESS(*status)) {
+        ut->pFuncs              = &unistrFuncs;
+        ut->context             = s;
+        ut->providerProperties  = I32_FLAG(UTEXT_PROVIDER_STABLE_CHUNKS);
+        ut->chunkContents       = s->getBuffer();
+        ut->chunkLength         = s->length();
+        ut->chunkNativeStart    = 0;
+        ut->chunkNativeLimit    = ut->chunkLength;
+        ut->nativeIndexingLimit = ut->chunkLength;
+    }
+    return ut;
+}
+
+//------------------------------------------------------------------------------
+//
+//     UText implementation for const UChar * strings
+//
+//         Use of UText data members:
+//            context    pointer to UnicodeString
+//            a          length.  -1 if not yet known.
+//
+//         TODO:  support 64 bit lengths.
+//
+//------------------------------------------------------------------------------
+
+U_CDECL_BEGIN
+
+
+static UText * U_CALLCONV
+ucstrTextClone(UText *dest, const UText * src, UBool deep, UErrorCode * status) {
+    // First do a generic shallow clone.
+    dest = shallowTextClone(dest, src, status);
+
+    // For deep clones, make a copy of the string.
+    //  The copied storage is owned by the newly created clone.
+    //  A non-NULL pointer in UText.p is the signal to the close() function to delete
+    //    it.
+    //
+    if (deep && U_SUCCESS(*status)) {
+        U_ASSERT(utext_nativeLength(dest) < INT32_MAX);
+        int32_t  len = (int32_t)utext_nativeLength(dest);
+
+        // The cloned string IS going to be NUL terminated, whether or not the original was.
+        const UChar *srcStr = (const UChar *)src->context;
+        UChar *copyStr = (UChar *)uprv_malloc((len+1) * sizeof(UChar));
+        if (copyStr == NULL) {
+            *status = U_MEMORY_ALLOCATION_ERROR;
+        } else {
+            int64_t i;
+            for (i=0; i<len; i++) {
+                copyStr[i] = srcStr[i];
+            }
+            copyStr[len] = 0;
+            dest->context = copyStr;
+            dest->providerProperties |= I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT);
+        }
+    }
+    return dest;
+}
+
+
+static void U_CALLCONV
+ucstrTextClose(UText *ut) {
+    // Most of the work of close is done by the generic UText framework close.
+    // All that needs to be done here is delete the string if the UText
+    //  owns it.  This occurs if the UText was created by cloning.
+    if (ut->providerProperties & I32_FLAG(UTEXT_PROVIDER_OWNS_TEXT)) {
+        UChar *s = (UChar *)ut->context;
+        uprv_free(s);
+        ut->context = NULL;
+    }
+}
+
+
+
+static int64_t U_CALLCONV
+ucstrTextLength(UText *ut) {
+    if (ut->a < 0) {
+        // null terminated, we don't yet know the length.  Scan for it.
+        //    Access is not convenient for doing this
+        //    because the current interation postion can't be changed.
+        const UChar  *str = (const UChar *)ut->context;
+        for (;;) {
+            if (str[ut->chunkNativeLimit] == 0) {
+                break;
+            }
+            ut->chunkNativeLimit++;
+        }
+        ut->a = ut->chunkNativeLimit;
+        ut->chunkLength = (int32_t)ut->chunkNativeLimit;
+        ut->nativeIndexingLimit = ut->chunkLength;
+        ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+    }
+    return ut->a;
+}
+
+
+static UBool U_CALLCONV
+ucstrTextAccess(UText *ut, int64_t index, UBool  forward) {
+    const UChar *str   = (const UChar *)ut->context;
+
+    // pin the requested index to the bounds of the string,
+    //  and set current iteration position.
+    if (index<0) {
+        index = 0;
+    } else if (index < ut->chunkNativeLimit) {
+        // The request data is within the chunk as it is known so far.
+        // Put index on a code point boundary.
+        U16_SET_CP_START(str, 0, index);
+    } else if (ut->a >= 0) {
+        // We know the length of this string, and the user is requesting something
+        // at or beyond the length.  Pin the requested index to the length.
+        index = ut->a;
+    } else {
+        // Null terminated string, length not yet known, and the requested index
+        //  is beyond where we have scanned so far.
+        //  Scan to 32 UChars beyond the requested index.  The strategy here is
+        //  to avoid fully scanning a long string when the caller only wants to
+        //  see a few characters at its beginning.
+        int32_t scanLimit = (int32_t)index + 32;
+        if ((index + 32)>INT32_MAX || (index + 32)<0 ) {   // note: int64 expression
+            scanLimit = INT32_MAX;
+        }
+
+        int32_t chunkLimit = (int32_t)ut->chunkNativeLimit;
+        for (; chunkLimit<scanLimit; chunkLimit++) {
+            if (str[chunkLimit] == 0) {
+                // We found the end of the string.  Remember it, pin the requested index to it,
+                //  and bail out of here.
+                ut->a = chunkLimit;
+                ut->chunkLength = chunkLimit;
+                ut->nativeIndexingLimit = chunkLimit;
+                if (index >= chunkLimit) {
+                    index = chunkLimit;
+                } else {
+                    U16_SET_CP_START(str, 0, index);
+                }
+
+                ut->chunkNativeLimit = chunkLimit;
+                ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+                goto breakout;
+            }
+        }
+        // We scanned through the next batch of UChars without finding the end.
+        U16_SET_CP_START(str, 0, index);
+        if (chunkLimit == INT32_MAX) {
+            // Scanned to the limit of a 32 bit length.
+            // Forceably trim the overlength string back so length fits in int32
+            //  TODO:  add support for 64 bit strings.
+            ut->a = chunkLimit;
+            ut->chunkLength = chunkLimit;
+            ut->nativeIndexingLimit = chunkLimit;
+            if (index > chunkLimit) {
+                index = chunkLimit;
+            }
+            ut->chunkNativeLimit = chunkLimit;
+            ut->providerProperties &= ~I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+        } else {
+            // The endpoint of a chunk must not be left in the middle of a surrogate pair.
+            // If the current end is on a lead surrogate, back the end up by one.
+            // It doesn't matter if the end char happens to be an unpaired surrogate,
+            //    and it's simpler not to worry about it.
+            if (U16_IS_LEAD(str[chunkLimit-1])) {
+                --chunkLimit;
+            }
+            // Null-terminated chunk with end still unknown.
+            // Update the chunk length to reflect what has been scanned thus far.
+            // That the full length is still unknown is (still) flagged by
+            //    ut->a being < 0.
+            ut->chunkNativeLimit = chunkLimit;
+            ut->nativeIndexingLimit = chunkLimit;
+            ut->chunkLength = chunkLimit;
+        }
+
+    }
+breakout:
+    U_ASSERT(index<=INT32_MAX);
+    ut->chunkOffset = (int32_t)index;
+
+    // Check whether request is at the start or end
+    UBool retVal = (forward && index<ut->chunkNativeLimit) || (!forward && index>0);
+    return retVal;
+}
+
+
+
+static int32_t U_CALLCONV
+ucstrTextExtract(UText *ut,
+                  int64_t start, int64_t limit,
+                  UChar *dest, int32_t destCapacity,
+                  UErrorCode *pErrorCode)
+{
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0) || start>limit) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    //const UChar *s=(const UChar *)ut->context;
+    int32_t si, di;
+
+    int32_t start32;
+    int32_t limit32;
+
+    // Access the start.  Does two things we need:
+    //   Pins 'start' to the length of the string, if it came in out-of-bounds.
+    //   Snaps 'start' to the beginning of a code point.
+    ucstrTextAccess(ut, start, TRUE);
+    const UChar *s=ut->chunkContents;
+    start32 = ut->chunkOffset;
+
+    int32_t strLength=(int32_t)ut->a;
+    if (strLength >= 0) {
+        limit32 = pinIndex(limit, strLength);
+    } else {
+        limit32 = pinIndex(limit, INT32_MAX);
+    }
+    di = 0;
+    for (si=start32; si<limit32; si++) {
+        if (strLength<0 && s[si]==0) {
+            // Just hit the end of a null-terminated string.
+            ut->a = si;               // set string length for this UText
+            ut->chunkNativeLimit    = si;
+            ut->chunkLength         = si;
+            ut->nativeIndexingLimit = si;
+            strLength               = si;
+            limit32                 = si;
+            break;
+        }
+        U_ASSERT(di>=0); /* to ensure di never exceeds INT32_MAX, which must not happen logically */
+        if (di<destCapacity) {
+            // only store if there is space.
+            dest[di] = s[si];
+        } else {
+            if (strLength>=0) {
+                // We have filled the destination buffer, and the string length is known.
+                //  Cut the loop short.  There is no need to scan string termination.
+                di = limit32 - start32;
+                si = limit32;
+                break;
+            }
+        }
+        di++;
+    }
+
+    // If the limit index points to a lead surrogate of a pair,
+    //   add the corresponding trail surrogate to the destination.
+    if (si>0 && U16_IS_LEAD(s[si-1]) &&
+            ((si<strLength || strLength<0)  && U16_IS_TRAIL(s[si])))
+    {
+        if (di<destCapacity) {
+            // store only if there is space in the output buffer.
+            dest[di++] = s[si];
+        }
+        si++;
+    }
+
+    // Put iteration position at the point just following the extracted text
+    if (si <= ut->chunkNativeLimit) {
+        ut->chunkOffset = si;
+    } else {
+        ucstrTextAccess(ut, si, TRUE);
+    }
+
+    // Add a terminating NUL if space in the buffer permits,
+    // and set the error status as required.
+    u_terminateUChars(dest, destCapacity, di, pErrorCode);
+    return di;
+}
+
+static const struct UTextFuncs ucstrFuncs =
+{
+    sizeof(UTextFuncs),
+    0, 0, 0,           // Reserved alignment padding
+    ucstrTextClone,
+    ucstrTextLength,
+    ucstrTextAccess,
+    ucstrTextExtract,
+    NULL,              // Replace
+    NULL,              // Copy
+    NULL,              // MapOffsetToNative,
+    NULL,              // MapIndexToUTF16,
+    ucstrTextClose,
+    NULL,              // spare 1
+    NULL,              // spare 2
+    NULL,              // spare 3
+};
+
+U_CDECL_END
+
+static const UChar gEmptyUString[] = {0};
+
+U_CAPI UText * U_EXPORT2
+utext_openUChars(UText *ut, const UChar *s, int64_t length, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+    if(s==NULL && length==0) {
+        s = gEmptyUString;
+    }
+    if (s==NULL || length < -1 || length>INT32_MAX) {
+        *status = U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    ut = utext_setup(ut, 0, status);
+    if (U_SUCCESS(*status)) {
+        ut->pFuncs               = &ucstrFuncs;
+        ut->context              = s;
+        ut->providerProperties   = I32_FLAG(UTEXT_PROVIDER_STABLE_CHUNKS);
+        if (length==-1) {
+            ut->providerProperties |= I32_FLAG(UTEXT_PROVIDER_LENGTH_IS_EXPENSIVE);
+        }
+        ut->a                    = length;
+        ut->chunkContents        = s;
+        ut->chunkNativeStart     = 0;
+        ut->chunkNativeLimit     = length>=0? length : 0;
+        ut->chunkLength          = (int32_t)ut->chunkNativeLimit;
+        ut->chunkOffset          = 0;
+        ut->nativeIndexingLimit  = ut->chunkLength;
+    }
+    return ut;
+}
+
+
+//------------------------------------------------------------------------------
+//
+//     UText implementation for text from ICU CharacterIterators
+//
+//         Use of UText data members:
+//            context    pointer to the CharacterIterator
+//            a          length of the full text.
+//            p          pointer to  buffer 1
+//            b          start index of local buffer 1 contents
+//            q          pointer to buffer 2
+//            c          start index of local buffer 2 contents
+//            r          pointer to the character iterator if the UText owns it.
+//                       Null otherwise.
+//
+//------------------------------------------------------------------------------
+#define CIBufSize 16
+
+U_CDECL_BEGIN
+static void U_CALLCONV
+charIterTextClose(UText *ut) {
+    // Most of the work of close is done by the generic UText framework close.
+    // All that needs to be done here is delete the CharacterIterator if the UText
+    //  owns it.  This occurs if the UText was created by cloning.
+    CharacterIterator *ci = (CharacterIterator *)ut->r;
+    delete ci;
+    ut->r = NULL;
+}
+
+static int64_t U_CALLCONV
+charIterTextLength(UText *ut) {
+    return (int32_t)ut->a;
+}
+
+static UBool U_CALLCONV
+charIterTextAccess(UText *ut, int64_t index, UBool  forward) {
+    CharacterIterator *ci   = (CharacterIterator *)ut->context;
+
+    int32_t clippedIndex = (int32_t)index;
+    if (clippedIndex<0) {
+        clippedIndex=0;
+    } else if (clippedIndex>=ut->a) {
+        clippedIndex=(int32_t)ut->a;
+    }
+    int32_t neededIndex = clippedIndex;
+    if (!forward && neededIndex>0) {
+        // reverse iteration, want the position just before what was asked for.
+        neededIndex--;
+    } else if (forward && neededIndex==ut->a && neededIndex>0) {
+        // Forward iteration, don't ask for something past the end of the text.
+        neededIndex--;
+    }
+
+    // Find the native index of the start of the buffer containing what we want.
+    neededIndex -= neededIndex % CIBufSize;
+
+    UChar *buf = NULL;
+    UBool  needChunkSetup = TRUE;
+    int    i;
+    if (ut->chunkNativeStart == neededIndex) {
+        // The buffer we want is already the current chunk.
+        needChunkSetup = FALSE;
+    } else if (ut->b == neededIndex) {
+        // The first buffer (buffer p) has what we need.
+        buf = (UChar *)ut->p;
+    } else if (ut->c == neededIndex) {
+        // The second buffer (buffer q) has what we need.
+        buf = (UChar *)ut->q;
+    } else {
+        // Neither buffer already has what we need.
+        // Load new data from the character iterator.
+        // Use the buf that is not the current buffer.
+        buf = (UChar *)ut->p;
+        if (ut->p == ut->chunkContents) {
+            buf = (UChar *)ut->q;
+        }
+        ci->setIndex(neededIndex);
+        for (i=0; i<CIBufSize; i++) {
+            buf[i] = ci->nextPostInc();
+            if (i+neededIndex > ut->a) {
+                break;
+            }
+        }
+    }
+
+    // We have a buffer with the data we need.
+    // Set it up as the current chunk, if it wasn't already.
+    if (needChunkSetup) {
+        ut->chunkContents = buf;
+        ut->chunkLength   = CIBufSize;
+        ut->chunkNativeStart = neededIndex;
+        ut->chunkNativeLimit = neededIndex + CIBufSize;
+        if (ut->chunkNativeLimit > ut->a) {
+            ut->chunkNativeLimit = ut->a;
+            ut->chunkLength  = (int32_t)(ut->chunkNativeLimit)-(int32_t)(ut->chunkNativeStart);
+        }
+        ut->nativeIndexingLimit = ut->chunkLength;
+        U_ASSERT(ut->chunkOffset>=0 && ut->chunkOffset<=CIBufSize);
+    }
+    ut->chunkOffset = clippedIndex - (int32_t)ut->chunkNativeStart;
+    UBool success = (forward? ut->chunkOffset<ut->chunkLength : ut->chunkOffset>0);
+    return success;
+}
+
+static UText * U_CALLCONV
+charIterTextClone(UText *dest, const UText *src, UBool deep, UErrorCode * status) {
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    if (deep) {
+        // There is no CharacterIterator API for cloning the underlying text storage.
+        *status = U_UNSUPPORTED_ERROR;
+        return NULL;
+    } else {
+        CharacterIterator *srcCI =(CharacterIterator *)src->context;
+        srcCI = srcCI->clone();
+        dest = utext_openCharacterIterator(dest, srcCI, status);
+        if (U_FAILURE(*status)) {
+            return dest;
+        }
+        // cast off const on getNativeIndex.
+        //   For CharacterIterator based UTexts, this is safe, the operation is const.
+        int64_t  ix = utext_getNativeIndex((UText *)src);
+        utext_setNativeIndex(dest, ix);
+        dest->r = srcCI;    // flags that this UText owns the CharacterIterator
+    }
+    return dest;
+}
+
+static int32_t U_CALLCONV
+charIterTextExtract(UText *ut,
+                  int64_t start, int64_t limit,
+                  UChar *dest, int32_t destCapacity,
+                  UErrorCode *status)
+{
+    if(U_FAILURE(*status)) {
+        return 0;
+    }
+    if(destCapacity<0 || (dest==NULL && destCapacity>0) || start>limit) {
+        *status=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    int32_t  length  = (int32_t)ut->a;
+    int32_t  start32 = pinIndex(start, length);
+    int32_t  limit32 = pinIndex(limit, length);
+    int32_t  desti   = 0;
+    int32_t  srci;
+    int32_t  copyLimit;
+
+    CharacterIterator *ci = (CharacterIterator *)ut->context;
+    ci->setIndex32(start32);   // Moves ix to lead of surrogate pair, if needed.
+    srci = ci->getIndex();
+    copyLimit = srci;
+    while (srci<limit32) {
+        UChar32 c = ci->next32PostInc();
+        int32_t  len = U16_LENGTH(c);
+        U_ASSERT(desti+len>0); /* to ensure desti+len never exceeds MAX_INT32, which must not happen logically */
+        if (desti+len <= destCapacity) {
+            U16_APPEND_UNSAFE(dest, desti, c);
+            copyLimit = srci+len;
+        } else {
+            desti += len;
+            *status = U_BUFFER_OVERFLOW_ERROR;
+        }
+        srci += len;
+    }
+
+    charIterTextAccess(ut, copyLimit, TRUE);
+
+    u_terminateUChars(dest, destCapacity, desti, status);
+    return desti;
+}
+
+static const struct UTextFuncs charIterFuncs =
+{
+    sizeof(UTextFuncs),
+    0, 0, 0,             // Reserved alignment padding
+    charIterTextClone,
+    charIterTextLength,
+    charIterTextAccess,
+    charIterTextExtract,
+    NULL,                // Replace
+    NULL,                // Copy
+    NULL,                // MapOffsetToNative,
+    NULL,                // MapIndexToUTF16,
+    charIterTextClose,
+    NULL,                // spare 1
+    NULL,                // spare 2
+    NULL                 // spare 3
+};
+U_CDECL_END
+
+
+U_CAPI UText * U_EXPORT2
+utext_openCharacterIterator(UText *ut, CharacterIterator *ci, UErrorCode *status) {
+    if (U_FAILURE(*status)) {
+        return NULL;
+    }
+
+    if (ci->startIndex() > 0) {
+        // No support for CharacterIterators that do not start indexing from zero.
+        *status = U_UNSUPPORTED_ERROR;
+        return NULL;
+    }
+
+    // Extra space in UText for 2 buffers of CIBufSize UChars each.
+    int32_t  extraSpace = 2 * CIBufSize * sizeof(UChar);
+    ut = utext_setup(ut, extraSpace, status);
+    if (U_SUCCESS(*status)) {
+        ut->pFuncs                = &charIterFuncs;
+        ut->context              = ci;
+        ut->providerProperties   = 0;
+        ut->a                    = ci->endIndex();        // Length of text
+        ut->p                    = ut->pExtra;            // First buffer
+        ut->b                    = -1;                    // Native index of first buffer contents
+        ut->q                    = (UChar*)ut->pExtra+CIBufSize;  // Second buffer
+        ut->c                    = -1;                    // Native index of second buffer contents
+
+        // Initialize current chunk contents to be empty.
+        //   First access will fault something in.
+        //   Note:  The initial nativeStart and chunkOffset must sum to zero
+        //          so that getNativeIndex() will correctly compute to zero
+        //          if no call to Access() has ever been made.  They can't be both
+        //          zero without Access() thinking that the chunk is valid.
+        ut->chunkContents        = (UChar *)ut->p;
+        ut->chunkNativeStart     = -1;
+        ut->chunkOffset          = 1;
+        ut->chunkNativeLimit     = 0;
+        ut->chunkLength          = 0;
+        ut->nativeIndexingLimit  = ut->chunkOffset;  // enables native indexing
+    }
+    return ut;
+}
diff --git a/thirdparty/icu4c/common/utf_impl.cpp b/thirdparty/icu4c/common/utf_impl.cpp
new file mode 100644
index 0000000000..9dd241a12b
--- /dev/null
+++ b/thirdparty/icu4c/common/utf_impl.cpp
@@ -0,0 +1,329 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1999-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utf_impl.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep13
+*   created by: Markus W. Scherer
+*
+*   This file provides implementation functions for macros in the utfXX.h
+*   that would otherwise be too long as macros.
+*/
+
+/* set import/export definitions */
+#ifndef U_UTF8_IMPL
+#   define U_UTF8_IMPL
+#endif
+
+#include "unicode/utypes.h"
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "uassert.h"
+
+/*
+ * Table of the number of utf8 trail bytes, indexed by the lead byte.
+ * Used by the deprecated macro UTF8_COUNT_TRAIL_BYTES, defined in utf_old.h
+ *
+ * The current macro, U8_COUNT_TRAIL_BYTES, does _not_ use this table.
+ *
+ * Note that this table cannot be removed, even if UTF8_COUNT_TRAIL_BYTES were
+ * changed to no longer use it. References to the table from expansions of UTF8_COUNT_TRAIL_BYTES
+ * may exist in old client code that must continue to run with newer icu library versions.
+ *
+ * This table could be replaced on many machines by
+ * a few lines of assembler code using an
+ * "index of first 0-bit from msb" instruction and
+ * one or two more integer instructions.
+ *
+ * For example, on an i386, do something like
+ * - MOV AL, leadByte
+ * - NOT AL         (8-bit, leave b15..b8==0..0, reverse only b7..b0)
+ * - MOV AH, 0
+ * - BSR BX, AX     (16-bit)
+ * - MOV AX, 6      (result)
+ * - JZ finish      (ZF==1 if leadByte==0xff)
+ * - SUB AX, BX (result)
+ * -finish:
+ * (BSR: Bit Scan Reverse, scans for a 1-bit, starting from the MSB)
+ */
+extern "C" U_EXPORT const uint8_t
+utf8_countTrailBytes[256]={
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+
+    // illegal C0 & C1
+    // 2-byte lead bytes C2..DF
+    0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+
+    // 3-byte lead bytes E0..EF
+    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+    // 4-byte lead bytes F0..F4
+    // illegal F5..FF
+    3, 3, 3, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+};
+
+static const UChar32
+utf8_errorValue[6]={
+    // Same values as UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_2, UTF_ERROR_VALUE,
+    // but without relying on the obsolete unicode/utf_old.h.
+    0x15, 0x9f, 0xffff,
+    0x10ffff
+};
+
+static UChar32
+errorValue(int32_t count, int8_t strict) {
+    if(strict>=0) {
+        return utf8_errorValue[count];
+    } else if(strict==-3) {
+        return 0xfffd;
+    } else {
+        return U_SENTINEL;
+    }
+}
+
+/*
+ * Handle the non-inline part of the U8_NEXT() and U8_NEXT_FFFD() macros
+ * and their obsolete sibling UTF8_NEXT_CHAR_SAFE().
+ *
+ * U8_NEXT() supports NUL-terminated strings indicated via length<0.
+ *
+ * The "strict" parameter controls the error behavior:
+ * <0  "Safe" behavior of U8_NEXT():
+ *     -1: All illegal byte sequences yield U_SENTINEL=-1.
+ *     -2: Same as -1, except for lenient treatment of surrogate code points as legal.
+ *         Some implementations use this for roundtripping of
+ *         Unicode 16-bit strings that are not well-formed UTF-16, that is, they
+ *         contain unpaired surrogates.
+ *     -3: All illegal byte sequences yield U+FFFD.
+ *  0  Obsolete "safe" behavior of UTF8_NEXT_CHAR_SAFE(..., FALSE):
+ *     All illegal byte sequences yield a positive code point such that this
+ *     result code point would be encoded with the same number of bytes as
+ *     the illegal sequence.
+ * >0  Obsolete "strict" behavior of UTF8_NEXT_CHAR_SAFE(..., TRUE):
+ *     Same as the obsolete "safe" behavior, but non-characters are also treated
+ *     like illegal sequences.
+ *
+ * Note that a UBool is the same as an int8_t.
+ */
+U_CAPI UChar32 U_EXPORT2
+utf8_nextCharSafeBody(const uint8_t *s, int32_t *pi, int32_t length, UChar32 c, UBool strict) {
+    // *pi is one after byte c.
+    int32_t i=*pi;
+    // length can be negative for NUL-terminated strings: Read and validate one byte at a time.
+    if(i==length || c>0xf4) {
+        // end of string, or not a lead byte
+    } else if(c>=0xf0) {
+        // Test for 4-byte sequences first because
+        // U8_NEXT() handles shorter valid sequences inline.
+        uint8_t t1=s[i], t2, t3;
+        c&=7;
+        if(U8_IS_VALID_LEAD4_AND_T1(c, t1) &&
+                ++i!=length && (t2=s[i]-0x80)<=0x3f &&
+                ++i!=length && (t3=s[i]-0x80)<=0x3f) {
+            ++i;
+            c=(c<<18)|((t1&0x3f)<<12)|(t2<<6)|t3;
+            // strict: forbid non-characters like U+fffe
+            if(strict<=0 || !U_IS_UNICODE_NONCHAR(c)) {
+                *pi=i;
+                return c;
+            }
+        }
+    } else if(c>=0xe0) {
+        c&=0xf;
+        if(strict!=-2) {
+            uint8_t t1=s[i], t2;
+            if(U8_IS_VALID_LEAD3_AND_T1(c, t1) &&
+                    ++i!=length && (t2=s[i]-0x80)<=0x3f) {
+                ++i;
+                c=(c<<12)|((t1&0x3f)<<6)|t2;
+                // strict: forbid non-characters like U+fffe
+                if(strict<=0 || !U_IS_UNICODE_NONCHAR(c)) {
+                    *pi=i;
+                    return c;
+                }
+            }
+        } else {
+            // strict=-2 -> lenient: allow surrogates
+            uint8_t t1=s[i]-0x80, t2;
+            if(t1<=0x3f && (c>0 || t1>=0x20) &&
+                    ++i!=length && (t2=s[i]-0x80)<=0x3f) {
+                *pi=i+1;
+                return (c<<12)|(t1<<6)|t2;
+            }
+        }
+    } else if(c>=0xc2) {
+        uint8_t t1=s[i]-0x80;
+        if(t1<=0x3f) {
+            *pi=i+1;
+            return ((c-0xc0)<<6)|t1;
+        }
+    }  // else 0x80<=c<0xc2 is not a lead byte
+
+    /* error handling */
+    c=errorValue(i-*pi, strict);
+    *pi=i;
+    return c;
+}
+
+U_CAPI int32_t U_EXPORT2
+utf8_appendCharSafeBody(uint8_t *s, int32_t i, int32_t length, UChar32 c, UBool *pIsError) {
+    if((uint32_t)(c)<=0x7ff) {
+        if((i)+1<(length)) {
+            (s)[(i)++]=(uint8_t)(((c)>>6)|0xc0);
+            (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80);
+            return i;
+        }
+    } else if((uint32_t)(c)<=0xffff) {
+        /* Starting with Unicode 3.2, surrogate code points must not be encoded in UTF-8. */
+        if((i)+2<(length) && !U_IS_SURROGATE(c)) {
+            (s)[(i)++]=(uint8_t)(((c)>>12)|0xe0);
+            (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80);
+            (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80);
+            return i;
+        }
+    } else if((uint32_t)(c)<=0x10ffff) {
+        if((i)+3<(length)) {
+            (s)[(i)++]=(uint8_t)(((c)>>18)|0xf0);
+            (s)[(i)++]=(uint8_t)((((c)>>12)&0x3f)|0x80);
+            (s)[(i)++]=(uint8_t)((((c)>>6)&0x3f)|0x80);
+            (s)[(i)++]=(uint8_t)(((c)&0x3f)|0x80);
+            return i;
+        }
+    }
+    /* c>0x10ffff or not enough space, write an error value */
+    if(pIsError!=NULL) {
+        *pIsError=TRUE;
+    } else {
+        length-=i;
+        if(length>0) {
+            int32_t offset;
+            if(length>3) {
+                length=3;
+            }
+            s+=i;
+            offset=0;
+            c=utf8_errorValue[length-1];
+            U8_APPEND_UNSAFE(s, offset, c);
+            i=i+offset;
+        }
+    }
+    return i;
+}
+
+U_CAPI UChar32 U_EXPORT2
+utf8_prevCharSafeBody(const uint8_t *s, int32_t start, int32_t *pi, UChar32 c, UBool strict) {
+    // *pi is the index of byte c.
+    int32_t i=*pi;
+    if(U8_IS_TRAIL(c) && i>start) {
+        uint8_t b1=s[--i];
+        if(U8_IS_LEAD(b1)) {
+            if(b1<0xe0) {
+                *pi=i;
+                return ((b1-0xc0)<<6)|(c&0x3f);
+            } else if(b1<0xf0 ? U8_IS_VALID_LEAD3_AND_T1(b1, c) : U8_IS_VALID_LEAD4_AND_T1(b1, c)) {
+                // Truncated 3- or 4-byte sequence.
+                *pi=i;
+                return errorValue(1, strict);
+            }
+        } else if(U8_IS_TRAIL(b1) && i>start) {
+            // Extract the value bits from the last trail byte.
+            c&=0x3f;
+            uint8_t b2=s[--i];
+            if(0xe0<=b2 && b2<=0xf4) {
+                if(b2<0xf0) {
+                    b2&=0xf;
+                    if(strict!=-2) {
+                        if(U8_IS_VALID_LEAD3_AND_T1(b2, b1)) {
+                            *pi=i;
+                            c=(b2<<12)|((b1&0x3f)<<6)|c;
+                            if(strict<=0 || !U_IS_UNICODE_NONCHAR(c)) {
+                                return c;
+                            } else {
+                                // strict: forbid non-characters like U+fffe
+                                return errorValue(2, strict);
+                            }
+                        }
+                    } else {
+                        // strict=-2 -> lenient: allow surrogates
+                        b1-=0x80;
+                        if((b2>0 || b1>=0x20)) {
+                            *pi=i;
+                            return (b2<<12)|(b1<<6)|c;
+                        }
+                    }
+                } else if(U8_IS_VALID_LEAD4_AND_T1(b2, b1)) {
+                    // Truncated 4-byte sequence.
+                    *pi=i;
+                    return errorValue(2, strict);
+                }
+            } else if(U8_IS_TRAIL(b2) && i>start) {
+                uint8_t b3=s[--i];
+                if(0xf0<=b3 && b3<=0xf4) {
+                    b3&=7;
+                    if(U8_IS_VALID_LEAD4_AND_T1(b3, b2)) {
+                        *pi=i;
+                        c=(b3<<18)|((b2&0x3f)<<12)|((b1&0x3f)<<6)|c;
+                        if(strict<=0 || !U_IS_UNICODE_NONCHAR(c)) {
+                            return c;
+                        } else {
+                            // strict: forbid non-characters like U+fffe
+                            return errorValue(3, strict);
+                        }
+                    }
+                }
+            }
+        }
+    }
+    return errorValue(0, strict);
+}
+
+U_CAPI int32_t U_EXPORT2
+utf8_back1SafeBody(const uint8_t *s, int32_t start, int32_t i) {
+    // Same as utf8_prevCharSafeBody(..., strict=-1) minus assembling code points.
+    int32_t orig_i=i;
+    uint8_t c=s[i];
+    if(U8_IS_TRAIL(c) && i>start) {
+        uint8_t b1=s[--i];
+        if(U8_IS_LEAD(b1)) {
+            if(b1<0xe0 ||
+                    (b1<0xf0 ? U8_IS_VALID_LEAD3_AND_T1(b1, c) : U8_IS_VALID_LEAD4_AND_T1(b1, c))) {
+                return i;
+            }
+        } else if(U8_IS_TRAIL(b1) && i>start) {
+            uint8_t b2=s[--i];
+            if(0xe0<=b2 && b2<=0xf4) {
+                if(b2<0xf0 ? U8_IS_VALID_LEAD3_AND_T1(b2, b1) : U8_IS_VALID_LEAD4_AND_T1(b2, b1)) {
+                    return i;
+                }
+            } else if(U8_IS_TRAIL(b2) && i>start) {
+                uint8_t b3=s[--i];
+                if(0xf0<=b3 && b3<=0xf4 && U8_IS_VALID_LEAD4_AND_T1(b3, b2)) {
+                    return i;
+                }
+            }
+        }
+    }
+    return orig_i;
+}
diff --git a/thirdparty/icu4c/common/util.cpp b/thirdparty/icu4c/common/util.cpp
new file mode 100644
index 0000000000..86e5c791ba
--- /dev/null
+++ b/thirdparty/icu4c/common/util.cpp
@@ -0,0 +1,421 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2001-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   11/19/2001  aliu        Creation.
+**********************************************************************
+*/
+
+#include "unicode/unimatch.h"
+#include "unicode/utf16.h"
+#include "patternprops.h"
+#include "util.h"
+
+// Define UChar constants using hex for EBCDIC compatibility
+
+static const UChar BACKSLASH  = 0x005C; /*\*/
+static const UChar UPPER_U    = 0x0055; /*U*/
+static const UChar LOWER_U    = 0x0075; /*u*/
+static const UChar APOSTROPHE = 0x0027; // '\''
+static const UChar SPACE      = 0x0020; // ' '
+
+// "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+static const UChar DIGITS[] = {
+    48,49,50,51,52,53,54,55,56,57,
+    65,66,67,68,69,70,71,72,73,74,
+    75,76,77,78,79,80,81,82,83,84,
+    85,86,87,88,89,90
+};
+
+U_NAMESPACE_BEGIN
+
+UnicodeString& ICU_Utility::appendNumber(UnicodeString& result, int32_t n,
+                                     int32_t radix, int32_t minDigits) {
+    if (radix < 2 || radix > 36) {
+        // Bogus radix
+        return result.append((UChar)63/*?*/);
+    }
+    // Handle negatives
+    if (n < 0) {
+        n = -n;
+        result.append((UChar)45/*-*/);
+    }
+    // First determine the number of digits
+    int32_t nn = n;
+    int32_t r = 1;
+    while (nn >= radix) {
+        nn /= radix;
+        r *= radix;
+        --minDigits;
+    }
+    // Now generate the digits
+    while (--minDigits > 0) {
+        result.append(DIGITS[0]);
+    }
+    while (r > 0) {
+        int32_t digit = n / r;
+        result.append(DIGITS[digit]);
+        n -= digit * r;
+        r /= radix;
+    }
+    return result;
+}
+
+/**
+ * Return true if the character is NOT printable ASCII.
+ */
+UBool ICU_Utility::isUnprintable(UChar32 c) {
+    return !(c >= 0x20 && c <= 0x7E);
+}
+
+/**
+ * Escape unprintable characters using \uxxxx notation for U+0000 to
+ * U+FFFF and \Uxxxxxxxx for U+10000 and above.  If the character is
+ * printable ASCII, then do nothing and return FALSE.  Otherwise,
+ * append the escaped notation and return TRUE.
+ */
+UBool ICU_Utility::escapeUnprintable(UnicodeString& result, UChar32 c) {
+    if (isUnprintable(c)) {
+        result.append(BACKSLASH);
+        if (c & ~0xFFFF) {
+            result.append(UPPER_U);
+            result.append(DIGITS[0xF&(c>>28)]);
+            result.append(DIGITS[0xF&(c>>24)]);
+            result.append(DIGITS[0xF&(c>>20)]);
+            result.append(DIGITS[0xF&(c>>16)]);
+        } else {
+            result.append(LOWER_U);
+        }
+        result.append(DIGITS[0xF&(c>>12)]);
+        result.append(DIGITS[0xF&(c>>8)]);
+        result.append(DIGITS[0xF&(c>>4)]);
+        result.append(DIGITS[0xF&c]);
+        return TRUE;
+    }
+    return FALSE;
+}
+
+/**
+ * Returns the index of a character, ignoring quoted text.
+ * For example, in the string "abc'hide'h", the 'h' in "hide" will not be
+ * found by a search for 'h'.
+ */
+// FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
+/*
+int32_t ICU_Utility::quotedIndexOf(const UnicodeString& text,
+                               int32_t start, int32_t limit,
+                               UChar charToFind) {
+    for (int32_t i=start; i<limit; ++i) {
+        UChar c = text.charAt(i);
+        if (c == BACKSLASH) {
+            ++i;
+        } else if (c == APOSTROPHE) {
+            while (++i < limit
+                   && text.charAt(i) != APOSTROPHE) {}
+        } else if (c == charToFind) {
+            return i;
+        }
+    }
+    return -1;
+}
+*/
+
+/**
+ * Skip over a sequence of zero or more white space characters at pos.
+ * @param advance if true, advance pos to the first non-white-space
+ * character at or after pos, or str.length(), if there is none.
+ * Otherwise leave pos unchanged.
+ * @return the index of the first non-white-space character at or
+ * after pos, or str.length(), if there is none.
+ */
+int32_t ICU_Utility::skipWhitespace(const UnicodeString& str, int32_t& pos,
+                                    UBool advance) {
+    int32_t p = pos;
+    const UChar* s = str.getBuffer();
+    p = (int32_t)(PatternProps::skipWhiteSpace(s + p, str.length() - p) - s);
+    if (advance) {
+        pos = p;
+    }
+    return p;
+}
+
+/**
+ * Skip over Pattern_White_Space in a Replaceable.
+ * Skipping may be done in the forward or
+ * reverse direction.  In either case, the leftmost index will be
+ * inclusive, and the rightmost index will be exclusive.  That is,
+ * given a range defined as [start, limit), the call
+ * skipWhitespace(text, start, limit) will advance start past leading
+ * whitespace, whereas the call skipWhitespace(text, limit, start),
+ * will back up limit past trailing whitespace.
+ * @param text the text to be analyzed
+ * @param pos either the start or limit of a range of 'text', to skip
+ * leading or trailing whitespace, respectively
+ * @param stop either the limit or start of a range of 'text', to skip
+ * leading or trailing whitespace, respectively
+ * @return the new start or limit, depending on what was passed in to
+ * 'pos'
+ */
+//?FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
+//?int32_t ICU_Utility::skipWhitespace(const Replaceable& text,
+//?                                    int32_t pos, int32_t stop) {
+//?    UChar32 c;
+//?    UBool isForward = (stop >= pos);
+//?
+//?    if (!isForward) {
+//?        --pos; // pos is a limit, so back up by one
+//?    }
+//?    
+//?    while (pos != stop &&
+//?           PatternProps::isWhiteSpace(c = text.char32At(pos))) {
+//?        if (isForward) {
+//?            pos += U16_LENGTH(c);
+//?        } else {
+//?            pos -= U16_LENGTH(c);
+//?        }
+//?    }
+//?
+//?    if (!isForward) {
+//?        ++pos; // make pos back into a limit
+//?    }
+//?
+//?    return pos;
+//?}
+
+/**
+ * Parse a single non-whitespace character 'ch', optionally
+ * preceded by whitespace.
+ * @param id the string to be parsed
+ * @param pos INPUT-OUTPUT parameter.  On input, pos[0] is the
+ * offset of the first character to be parsed.  On output, pos[0]
+ * is the index after the last parsed character.  If the parse
+ * fails, pos[0] will be unchanged.
+ * @param ch the non-whitespace character to be parsed.
+ * @return true if 'ch' is seen preceded by zero or more
+ * whitespace characters.
+ */
+UBool ICU_Utility::parseChar(const UnicodeString& id, int32_t& pos, UChar ch) {
+    int32_t start = pos;
+    skipWhitespace(id, pos, TRUE);
+    if (pos == id.length() ||
+        id.charAt(pos) != ch) {
+        pos = start;
+        return FALSE;
+    }
+    ++pos;
+    return TRUE;
+}
+
+/**
+ * Parse a pattern string within the given Replaceable and a parsing
+ * pattern.  Characters are matched literally and case-sensitively
+ * except for the following special characters:
+ *
+ * ~  zero or more Pattern_White_Space chars
+ *
+ * If end of pattern is reached with all matches along the way,
+ * pos is advanced to the first unparsed index and returned.
+ * Otherwise -1 is returned.
+ * @param pat pattern that controls parsing
+ * @param text text to be parsed, starting at index
+ * @param index offset to first character to parse
+ * @param limit offset after last character to parse
+ * @return index after last parsed character, or -1 on parse failure.
+ */
+int32_t ICU_Utility::parsePattern(const UnicodeString& pat,
+                                  const Replaceable& text,
+                                  int32_t index,
+                                  int32_t limit) {
+    int32_t ipat = 0;
+
+    // empty pattern matches immediately
+    if (ipat == pat.length()) {
+        return index;
+    }
+
+    UChar32 cpat = pat.char32At(ipat);
+
+    while (index < limit) {
+        UChar32 c = text.char32At(index);
+
+        // parse \s*
+        if (cpat == 126 /*~*/) {
+            if (PatternProps::isWhiteSpace(c)) {
+                index += U16_LENGTH(c);
+                continue;
+            } else {
+                if (++ipat == pat.length()) {
+                    return index; // success; c unparsed
+                }
+                // fall thru; process c again with next cpat
+            }
+        }
+
+        // parse literal
+        else if (c == cpat) {
+            index += U16_LENGTH(c);
+            ipat += U16_LENGTH(cpat);
+            if (ipat == pat.length()) {
+                return index; // success; c parsed
+            }
+            // fall thru; get next cpat
+        }
+
+        // match failure of literal
+        else {
+            return -1;
+        }
+
+        cpat = pat.char32At(ipat);
+    }
+
+    return -1; // text ended before end of pat
+}
+
+int32_t ICU_Utility::parseAsciiInteger(const UnicodeString& str, int32_t& pos) {
+    int32_t result = 0;
+    UChar c;
+    while (pos < str.length() && (c = str.charAt(pos)) >= u'0' && c <= u'9') {
+        result = result * 10 + (c - u'0');
+        pos++;
+    }
+    return result;
+}
+
+/**
+ * Append a character to a rule that is being built up.  To flush
+ * the quoteBuf to rule, make one final call with isLiteral == TRUE.
+ * If there is no final character, pass in (UChar32)-1 as c.
+ * @param rule the string to append the character to
+ * @param c the character to append, or (UChar32)-1 if none.
+ * @param isLiteral if true, then the given character should not be
+ * quoted or escaped.  Usually this means it is a syntactic element
+ * such as > or $
+ * @param escapeUnprintable if true, then unprintable characters
+ * should be escaped using \uxxxx or \Uxxxxxxxx.  These escapes will
+ * appear outside of quotes.
+ * @param quoteBuf a buffer which is used to build up quoted
+ * substrings.  The caller should initially supply an empty buffer,
+ * and thereafter should not modify the buffer.  The buffer should be
+ * cleared out by, at the end, calling this method with a literal
+ * character.
+ */
+void ICU_Utility::appendToRule(UnicodeString& rule,
+                               UChar32 c,
+                               UBool isLiteral,
+                               UBool escapeUnprintable,
+                               UnicodeString& quoteBuf) {
+    // If we are escaping unprintables, then escape them outside
+    // quotes.  \u and \U are not recognized within quotes.  The same
+    // logic applies to literals, but literals are never escaped.
+    if (isLiteral ||
+        (escapeUnprintable && ICU_Utility::isUnprintable(c))) {
+        if (quoteBuf.length() > 0) {
+            // We prefer backslash APOSTROPHE to double APOSTROPHE
+            // (more readable, less similar to ") so if there are
+            // double APOSTROPHEs at the ends, we pull them outside
+            // of the quote.
+
+            // If the first thing in the quoteBuf is APOSTROPHE
+            // (doubled) then pull it out.
+            while (quoteBuf.length() >= 2 &&
+                   quoteBuf.charAt(0) == APOSTROPHE &&
+                   quoteBuf.charAt(1) == APOSTROPHE) {
+                rule.append(BACKSLASH).append(APOSTROPHE);
+                quoteBuf.remove(0, 2);
+            }
+            // If the last thing in the quoteBuf is APOSTROPHE
+            // (doubled) then remove and count it and add it after.
+            int32_t trailingCount = 0;
+            while (quoteBuf.length() >= 2 &&
+                   quoteBuf.charAt(quoteBuf.length()-2) == APOSTROPHE &&
+                   quoteBuf.charAt(quoteBuf.length()-1) == APOSTROPHE) {
+                quoteBuf.truncate(quoteBuf.length()-2);
+                ++trailingCount;
+            }
+            if (quoteBuf.length() > 0) {
+                rule.append(APOSTROPHE);
+                rule.append(quoteBuf);
+                rule.append(APOSTROPHE);
+                quoteBuf.truncate(0);
+            }
+            while (trailingCount-- > 0) {
+                rule.append(BACKSLASH).append(APOSTROPHE);
+            }
+        }
+        if (c != (UChar32)-1) {
+            /* Since spaces are ignored during parsing, they are
+             * emitted only for readability.  We emit one here
+             * only if there isn't already one at the end of the
+             * rule.
+             */
+            if (c == SPACE) {
+                int32_t len = rule.length();
+                if (len > 0 && rule.charAt(len-1) != c) {
+                    rule.append(c);
+                }
+            } else if (!escapeUnprintable || !ICU_Utility::escapeUnprintable(rule, c)) {
+                rule.append(c);
+            }
+        }
+    }
+
+    // Escape ' and '\' and don't begin a quote just for them
+    else if (quoteBuf.length() == 0 &&
+             (c == APOSTROPHE || c == BACKSLASH)) {
+        rule.append(BACKSLASH);
+        rule.append(c);
+    }
+
+    // Specials (printable ascii that isn't [0-9a-zA-Z]) and
+    // whitespace need quoting.  Also append stuff to quotes if we are
+    // building up a quoted substring already.
+    else if (quoteBuf.length() > 0 ||
+             (c >= 0x0021 && c <= 0x007E &&
+              !((c >= 0x0030/*'0'*/ && c <= 0x0039/*'9'*/) ||
+                (c >= 0x0041/*'A'*/ && c <= 0x005A/*'Z'*/) ||
+                (c >= 0x0061/*'a'*/ && c <= 0x007A/*'z'*/))) ||
+             PatternProps::isWhiteSpace(c)) {
+        quoteBuf.append(c);
+        // Double ' within a quote
+        if (c == APOSTROPHE) {
+            quoteBuf.append(c);
+        }
+    }
+    
+    // Otherwise just append
+    else {
+        rule.append(c);
+    }
+}
+
+void ICU_Utility::appendToRule(UnicodeString& rule,
+                               const UnicodeString& text,
+                               UBool isLiteral,
+                               UBool escapeUnprintable,
+                               UnicodeString& quoteBuf) {
+    for (int32_t i=0; i<text.length(); ++i) {
+        appendToRule(rule, text[i], isLiteral, escapeUnprintable, quoteBuf);
+    }
+}
+
+/**
+ * Given a matcher reference, which may be null, append its
+ * pattern as a literal to the given rule.
+ */
+void ICU_Utility::appendToRule(UnicodeString& rule,
+                               const UnicodeMatcher* matcher,
+                               UBool escapeUnprintable,
+                               UnicodeString& quoteBuf) {
+    if (matcher != NULL) {
+        UnicodeString pat;
+        appendToRule(rule, matcher->toPattern(pat, escapeUnprintable),
+                     TRUE, escapeUnprintable, quoteBuf);
+    }
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/util.h b/thirdparty/icu4c/common/util.h
new file mode 100644
index 0000000000..9c3b76d9ed
--- /dev/null
+++ b/thirdparty/icu4c/common/util.h
@@ -0,0 +1,257 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+ **********************************************************************
+ *   Copyright (c) 2001-2011, International Business Machines
+ *   Corporation and others.  All Rights Reserved.
+ **********************************************************************
+ *   Date        Name        Description
+ *   11/19/2001  aliu        Creation.
+ **********************************************************************
+ */
+
+#ifndef ICU_UTIL_H
+#define ICU_UTIL_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "unicode/unistr.h"
+
+//--------------------------------------------------------------------
+// class ICU_Utility
+// i18n utility functions, scoped into the class ICU_Utility.
+//--------------------------------------------------------------------
+
+U_NAMESPACE_BEGIN
+
+class UnicodeMatcher;
+
+class U_COMMON_API ICU_Utility /* not : public UObject because all methods are static */ {
+ public:
+
+    /**
+     * Append a number to the given UnicodeString in the given radix.
+     * Standard digits '0'-'9' are used and letters 'A'-'Z' for
+     * radices 11 through 36.
+     * @param result the digits of the number are appended here
+     * @param n the number to be converted to digits; may be negative.
+     * If negative, a '-' is prepended to the digits.
+     * @param radix a radix from 2 to 36 inclusive.
+     * @param minDigits the minimum number of digits, not including
+     * any '-', to produce.  Values less than 2 have no effect.  One
+     * digit is always emitted regardless of this parameter.
+     * @return a reference to result
+     */
+    static UnicodeString& appendNumber(UnicodeString& result, int32_t n,
+                                       int32_t radix = 10,
+                                       int32_t minDigits = 1);
+
+    /** Returns a bogus UnicodeString by value. */
+    static inline UnicodeString makeBogusString() {
+        UnicodeString result;
+        result.setToBogus();
+        return result;
+    }
+
+    /**
+     * Return true if the character is NOT printable ASCII.
+     *
+     * This method should really be in UnicodeString (or similar).  For
+     * now, we implement it here and share it with friend classes.
+     */
+    static UBool isUnprintable(UChar32 c);
+
+    /**
+     * Escape unprintable characters using \uxxxx notation for U+0000 to
+     * U+FFFF and \Uxxxxxxxx for U+10000 and above.  If the character is
+     * printable ASCII, then do nothing and return false.  Otherwise,
+     * append the escaped notation and return true.
+     */
+    static UBool escapeUnprintable(UnicodeString& result, UChar32 c);
+
+    /**
+     * Returns the index of a character, ignoring quoted text.
+     * For example, in the string "abc'hide'h", the 'h' in "hide" will not be
+     * found by a search for 'h'.
+     * @param text text to be searched
+     * @param start the beginning index, inclusive; <code>0 <= start
+     * <= limit</code>.
+     * @param limit the ending index, exclusive; <code>start <= limit
+     * <= text.length()</code>.
+     * @param c character to search for
+     * @return Offset of the first instance of c, or -1 if not found.
+     */
+//?FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
+//    static int32_t quotedIndexOf(const UnicodeString& text,
+//                                 int32_t start, int32_t limit,
+//                                 UChar c);
+
+    /**
+     * Skip over a sequence of zero or more white space characters at pos.
+     * @param advance if true, advance pos to the first non-white-space
+     * character at or after pos, or str.length(), if there is none.
+     * Otherwise leave pos unchanged.
+     * @return the index of the first non-white-space character at or
+     * after pos, or str.length(), if there is none.
+     */
+    static int32_t skipWhitespace(const UnicodeString& str, int32_t& pos,
+                                  UBool advance = false);
+
+    /**
+     * Skip over Pattern_White_Space in a Replaceable.
+     * Skipping may be done in the forward or
+     * reverse direction.  In either case, the leftmost index will be
+     * inclusive, and the rightmost index will be exclusive.  That is,
+     * given a range defined as [start, limit), the call
+     * skipWhitespace(text, start, limit) will advance start past leading
+     * whitespace, whereas the call skipWhitespace(text, limit, start),
+     * will back up limit past trailing whitespace.
+     * @param text the text to be analyzed
+     * @param pos either the start or limit of a range of 'text', to skip
+     * leading or trailing whitespace, respectively
+     * @param stop either the limit or start of a range of 'text', to skip
+     * leading or trailing whitespace, respectively
+     * @return the new start or limit, depending on what was passed in to
+     * 'pos'
+     */
+//?FOR FUTURE USE.  DISABLE FOR NOW for coverage reasons.
+//?    static int32_t skipWhitespace(const Replaceable& text,
+//?                                  int32_t pos, int32_t stop);
+
+    /**
+     * Parse a single non-whitespace character 'ch', optionally
+     * preceded by whitespace.
+     * @param id the string to be parsed
+     * @param pos INPUT-OUTPUT parameter.  On input, pos[0] is the
+     * offset of the first character to be parsed.  On output, pos[0]
+     * is the index after the last parsed character.  If the parse
+     * fails, pos[0] will be unchanged.
+     * @param ch the non-whitespace character to be parsed.
+     * @return true if 'ch' is seen preceded by zero or more
+     * whitespace characters.
+     */
+    static UBool parseChar(const UnicodeString& id, int32_t& pos, UChar ch);
+
+    /**
+     * Parse a pattern string starting at offset pos.  Keywords are
+     * matched case-insensitively.  Spaces may be skipped and may be
+     * optional or required.  Integer values may be parsed, and if
+     * they are, they will be returned in the given array.  If
+     * successful, the offset of the next non-space character is
+     * returned.  On failure, -1 is returned.
+     * @param pattern must only contain lowercase characters, which
+     * will match their uppercase equivalents as well.  A space
+     * character matches one or more required spaces.  A '~' character
+     * matches zero or more optional spaces.  A '#' character matches
+     * an integer and stores it in parsedInts, which the caller must
+     * ensure has enough capacity.
+     * @param parsedInts array to receive parsed integers.  Caller
+     * must ensure that parsedInts.length is >= the number of '#'
+     * signs in 'pattern'.
+     * @return the position after the last character parsed, or -1 if
+     * the parse failed
+     */
+    static int32_t parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit,
+                                const UnicodeString& pattern, int32_t* parsedInts);
+        
+    /**
+     * Parse a pattern string within the given Replaceable and a parsing
+     * pattern.  Characters are matched literally and case-sensitively
+     * except for the following special characters:
+     *
+     * ~  zero or more Pattern_White_Space chars
+     *
+     * If end of pattern is reached with all matches along the way,
+     * pos is advanced to the first unparsed index and returned.
+     * Otherwise -1 is returned.
+     * @param pat pattern that controls parsing
+     * @param text text to be parsed, starting at index
+     * @param index offset to first character to parse
+     * @param limit offset after last character to parse
+     * @return index after last parsed character, or -1 on parse failure.
+     */
+    static int32_t parsePattern(const UnicodeString& pat,
+                                const Replaceable& text,
+                                int32_t index,
+                                int32_t limit);
+
+    /**
+     * Parse an integer at pos, either of the form \d+ or of the form
+     * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex,
+     * or octal format.
+     * @param pos INPUT-OUTPUT parameter.  On input, the index of the first
+     * character to parse.  On output, the index of the character after the
+     * last parsed character.
+     */
+    static int32_t parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit);
+
+    /**
+     * Parse an integer at pos using only ASCII digits.
+     * Base 10 only.
+     * @param pos INPUT-OUTPUT parameter.  On input, the index of the first
+     * character to parse.  On output, the index of the character after the
+     * last parsed character.
+     */
+    static int32_t parseAsciiInteger(const UnicodeString& str, int32_t& pos);
+
+    /**
+     * Parse a Unicode identifier from the given string at the given
+     * position.  Return the identifier, or an empty string if there
+     * is no identifier.
+     * @param str the string to parse
+     * @param pos INPUT-OUPUT parameter.  On INPUT, pos is the
+     * first character to examine.  It must be less than str.length(),
+     * and it must not point to a whitespace character.  That is, must
+     * have pos < str.length() and
+     * !UCharacter::isWhitespace(str.char32At(pos)).  On
+     * OUTPUT, the position after the last parsed character.
+     * @return the Unicode identifier, or an empty string if there is
+     * no valid identifier at pos.
+     */
+    static UnicodeString parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos);
+
+    /**
+     * Parse an unsigned 31-bit integer at the given offset.  Use
+     * UCharacter.digit() to parse individual characters into digits.
+     * @param text the text to be parsed
+     * @param pos INPUT-OUTPUT parameter.  On entry, pos is the
+     * offset within text at which to start parsing; it should point
+     * to a valid digit.  On exit, pos is the offset after the last
+     * parsed character.  If the parse failed, it will be unchanged on
+     * exit.  Must be >= 0 on entry.
+     * @param radix the radix in which to parse; must be >= 2 and <=
+     * 36.
+     * @return a non-negative parsed number, or -1 upon parse failure.
+     * Parse fails if there are no digits, that is, if pos does not
+     * point to a valid digit on entry, or if the number to be parsed
+     * does not fit into a 31-bit unsigned integer.
+     */
+    static int32_t parseNumber(const UnicodeString& text,
+                               int32_t& pos, int8_t radix);
+
+    static void appendToRule(UnicodeString& rule,
+                             UChar32 c,
+                             UBool isLiteral,
+                             UBool escapeUnprintable,
+                             UnicodeString& quoteBuf);
+    
+    static void appendToRule(UnicodeString& rule,
+                             const UnicodeString& text,
+                             UBool isLiteral,
+                             UBool escapeUnprintable,
+                             UnicodeString& quoteBuf);
+
+    static void appendToRule(UnicodeString& rule,
+                             const UnicodeMatcher* matcher,
+                             UBool escapeUnprintable,
+                             UnicodeString& quoteBuf);
+
+private:
+    // do not instantiate
+    ICU_Utility();
+};
+
+U_NAMESPACE_END
+
+#endif
+//eof
diff --git a/thirdparty/icu4c/common/util_props.cpp b/thirdparty/icu4c/common/util_props.cpp
new file mode 100644
index 0000000000..95a112bc91
--- /dev/null
+++ b/thirdparty/icu4c/common/util_props.cpp
@@ -0,0 +1,217 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (c) 2001-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   11/19/2001  aliu        Creation.
+**********************************************************************
+*/
+
+#include "unicode/uchar.h"
+#include "unicode/utf16.h"
+#include "patternprops.h"
+#include "util.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * Parse an integer at pos, either of the form \d+ or of the form
+ * 0x[0-9A-Fa-f]+ or 0[0-7]+, that is, in standard decimal, hex,
+ * or octal format.
+ * @param pos INPUT-OUTPUT parameter.  On input, the first
+ * character to parse.  On output, the character after the last
+ * parsed character.
+ */
+int32_t ICU_Utility::parseInteger(const UnicodeString& rule, int32_t& pos, int32_t limit) {
+    int32_t count = 0;
+    int32_t value = 0;
+    int32_t p = pos;
+    int8_t radix = 10;
+
+    if (p < limit && rule.charAt(p) == 48 /*0*/) {
+        if (p+1 < limit && (rule.charAt(p+1) == 0x78 /*x*/ || rule.charAt(p+1) == 0x58 /*X*/)) {
+            p += 2;
+            radix = 16;
+        }
+        else {
+            p++;
+            count = 1;
+            radix = 8;
+        }
+    }
+
+    while (p < limit) {
+        int32_t d = u_digit(rule.charAt(p++), radix);
+        if (d < 0) {
+            --p;
+            break;
+        }
+        ++count;
+        int32_t v = (value * radix) + d;
+        if (v <= value) {
+            // If there are too many input digits, at some point
+            // the value will go negative, e.g., if we have seen
+            // "0x8000000" already and there is another '0', when
+            // we parse the next 0 the value will go negative.
+            return 0;
+        }
+        value = v;
+    }
+    if (count > 0) {
+        pos = p;
+    }
+    return value;
+}
+
+/**
+ * Parse a pattern string starting at offset pos.  Keywords are
+ * matched case-insensitively.  Spaces may be skipped and may be
+ * optional or required.  Integer values may be parsed, and if
+ * they are, they will be returned in the given array.  If
+ * successful, the offset of the next non-space character is
+ * returned.  On failure, -1 is returned.
+ * @param pattern must only contain lowercase characters, which
+ * will match their uppercase equivalents as well.  A space
+ * character matches one or more required spaces.  A '~' character
+ * matches zero or more optional spaces.  A '#' character matches
+ * an integer and stores it in parsedInts, which the caller must
+ * ensure has enough capacity.
+ * @param parsedInts array to receive parsed integers.  Caller
+ * must ensure that parsedInts.length is >= the number of '#'
+ * signs in 'pattern'.
+ * @return the position after the last character parsed, or -1 if
+ * the parse failed
+ */
+int32_t ICU_Utility::parsePattern(const UnicodeString& rule, int32_t pos, int32_t limit,
+                              const UnicodeString& pattern, int32_t* parsedInts) {
+    // TODO Update this to handle surrogates
+    int32_t p;
+    int32_t intCount = 0; // number of integers parsed
+    for (int32_t i=0; i<pattern.length(); ++i) {
+        UChar cpat = pattern.charAt(i);
+        UChar c;
+        switch (cpat) {
+        case 32 /*' '*/:
+            if (pos >= limit) {
+                return -1;
+            }
+            c = rule.charAt(pos++);
+            if (!PatternProps::isWhiteSpace(c)) {
+                return -1;
+            }
+            // FALL THROUGH to skipWhitespace
+            U_FALLTHROUGH;
+        case 126 /*'~'*/:
+            pos = skipWhitespace(rule, pos);
+            break;
+        case 35 /*'#'*/:
+            p = pos;
+            parsedInts[intCount++] = parseInteger(rule, p, limit);
+            if (p == pos) {
+                // Syntax error; failed to parse integer
+                return -1;
+            }
+            pos = p;
+            break;
+        default:
+            if (pos >= limit) {
+                return -1;
+            }
+            c = (UChar) u_tolower(rule.charAt(pos++));
+            if (c != cpat) {
+                return -1;
+            }
+            break;
+        }
+    }
+    return pos;
+}
+
+/**
+ * Parse a Unicode identifier from the given string at the given
+ * position.  Return the identifier, or an empty string if there
+ * is no identifier.
+ * @param str the string to parse
+ * @param pos INPUT-OUPUT parameter.  On INPUT, pos is the
+ * first character to examine.  It must be less than str.length(),
+ * and it must not point to a whitespace character.  That is, must
+ * have pos < str.length().  On
+ * OUTPUT, the position after the last parsed character.
+ * @return the Unicode identifier, or an empty string if there is
+ * no valid identifier at pos.
+ */
+UnicodeString ICU_Utility::parseUnicodeIdentifier(const UnicodeString& str, int32_t& pos) {
+    // assert(pos < str.length());
+    UnicodeString buf;
+    int p = pos;
+    while (p < str.length()) {
+        UChar32 ch = str.char32At(p);
+        if (buf.length() == 0) {
+            if (u_isIDStart(ch)) {
+                buf.append(ch);
+            } else {
+                buf.truncate(0);
+                return buf;
+            }
+        } else {
+            if (u_isIDPart(ch)) {
+                buf.append(ch);
+            } else {
+                break;
+            }
+        }
+        p += U16_LENGTH(ch);
+    }
+    pos = p;
+    return buf;
+}
+
+/**
+ * Parse an unsigned 31-bit integer at the given offset.  Use
+ * UCharacter.digit() to parse individual characters into digits.
+ * @param text the text to be parsed
+ * @param pos INPUT-OUTPUT parameter.  On entry, pos[0] is the
+ * offset within text at which to start parsing; it should point
+ * to a valid digit.  On exit, pos[0] is the offset after the last
+ * parsed character.  If the parse failed, it will be unchanged on
+ * exit.  Must be >= 0 on entry.
+ * @param radix the radix in which to parse; must be >= 2 and <=
+ * 36.
+ * @return a non-negative parsed number, or -1 upon parse failure.
+ * Parse fails if there are no digits, that is, if pos[0] does not
+ * point to a valid digit on entry, or if the number to be parsed
+ * does not fit into a 31-bit unsigned integer.
+ */
+int32_t ICU_Utility::parseNumber(const UnicodeString& text,
+                                 int32_t& pos, int8_t radix) {
+    // assert(pos[0] >= 0);
+    // assert(radix >= 2);
+    // assert(radix <= 36);
+    int32_t n = 0;
+    int32_t p = pos;
+    while (p < text.length()) {
+        UChar32 ch = text.char32At(p);
+        int32_t d = u_digit(ch, radix);
+        if (d < 0) {
+            break;
+        }
+        n = radix*n + d;
+        // ASSUME that when a 32-bit integer overflows it becomes
+        // negative.  E.g., 214748364 * 10 + 8 => negative value.
+        if (n < 0) {
+            return -1;
+        }
+        ++p;
+    }
+    if (p == pos) {
+        return -1;
+    }
+    pos = p;
+    return n;
+}
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/utrace.cpp b/thirdparty/icu4c/common/utrace.cpp
new file mode 100644
index 0000000000..c981546594
--- /dev/null
+++ b/thirdparty/icu4c/common/utrace.cpp
@@ -0,0 +1,504 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2003-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  utrace.c
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*/
+
+#include "unicode/utrace.h"
+#include "utracimp.h"
+#include "cstring.h"
+#include "uassert.h"
+#include "ucln_cmn.h"
+
+
+static UTraceEntry     *pTraceEntryFunc = NULL;
+static UTraceExit      *pTraceExitFunc  = NULL;
+static UTraceData      *pTraceDataFunc  = NULL;
+static const void      *gTraceContext   = NULL;
+
+/**
+ * \var utrace_level
+ * Trace level variable. Negative for "off".
+ */
+static int32_t
+utrace_level = UTRACE_ERROR;
+
+U_CAPI void U_EXPORT2
+utrace_entry(int32_t fnNumber) {
+    if (pTraceEntryFunc != NULL) {
+        (*pTraceEntryFunc)(gTraceContext, fnNumber);
+    }
+}
+
+
+static const char gExitFmt[]             = "Returns.";
+static const char gExitFmtValue[]        = "Returns %d.";
+static const char gExitFmtStatus[]       = "Returns.  Status = %d.";
+static const char gExitFmtValueStatus[]  = "Returns %d.  Status = %d.";
+static const char gExitFmtPtrStatus[]    = "Returns %d.  Status = %p.";
+
+U_CAPI void U_EXPORT2
+utrace_exit(int32_t fnNumber, int32_t returnType, ...) {
+    if (pTraceExitFunc != NULL) {
+        va_list     args;
+        const char *fmt;
+
+        switch (returnType) {
+        case 0:
+            fmt = gExitFmt;
+            break;
+        case UTRACE_EXITV_I32:
+            fmt = gExitFmtValue;
+            break;
+        case UTRACE_EXITV_STATUS:
+            fmt = gExitFmtStatus;
+            break;
+        case UTRACE_EXITV_I32 | UTRACE_EXITV_STATUS:
+            fmt = gExitFmtValueStatus;
+            break;
+        case UTRACE_EXITV_PTR | UTRACE_EXITV_STATUS:
+            fmt = gExitFmtPtrStatus;
+            break;
+        default:
+            UPRV_UNREACHABLE;
+        }
+
+        va_start(args, returnType);
+        (*pTraceExitFunc)(gTraceContext, fnNumber, fmt, args);
+        va_end(args);
+    }
+}
+ 
+
+ 
+U_CAPI void U_EXPORT2 
+utrace_data(int32_t fnNumber, int32_t level, const char *fmt, ...) {
+    if (pTraceDataFunc != NULL) {
+           va_list args;
+           va_start(args, fmt ); 
+           (*pTraceDataFunc)(gTraceContext, fnNumber, level, fmt, args);
+           va_end(args);
+    }
+}
+
+
+static void outputChar(char c, char *outBuf, int32_t *outIx, int32_t capacity, int32_t indent) {
+    int32_t i;
+    /* Check whether a start of line indenting is needed.  Three cases:
+     *   1.  At the start of the first line  (output index == 0).
+     *   2.  At the start of subsequent lines  (preceeding char in buffer == '\n')
+     *   3.  When preflighting buffer len (buffer capacity is exceeded), when
+     *       a \n is output.  Ideally we wouldn't do the indent until the following char
+     *       is received, but that won't work because there's no place to remember that
+     *       the preceding char was \n.  Meaning that we may overstimate the
+     *       buffer size needed.  No harm done.
+     */
+    if (*outIx==0 ||   /* case 1. */
+        (c!='\n' && c!=0 && *outIx < capacity && outBuf[(*outIx)-1]=='\n') ||  /* case 2. */
+        (c=='\n' && *outIx>=capacity))    /* case 3 */
+    {
+        /* At the start of a line.  Indent. */
+        for(i=0; i<indent; i++) {
+            if (*outIx < capacity) {
+                outBuf[*outIx] = ' ';
+            }
+            (*outIx)++;
+        }
+    }
+
+    if (*outIx < capacity) {
+        outBuf[*outIx] = c;
+    }
+    if (c != 0) {
+        /* Nulls only appear as end-of-string terminators.  Move them to the output
+         *  buffer, but do not update the length of the buffer, so that any
+         *  following output will overwrite the null. */
+        (*outIx)++;
+    }
+}
+
+static void outputHexBytes(int64_t val, int32_t charsToOutput,
+                           char *outBuf, int32_t *outIx, int32_t capacity) {
+    static const char gHexChars[] = "0123456789abcdef";
+    int32_t shiftCount;
+    for  (shiftCount=(charsToOutput-1)*4; shiftCount >= 0; shiftCount-=4) {
+        char c = gHexChars[(val >> shiftCount) & 0xf];
+        outputChar(c, outBuf, outIx, capacity, 0);
+    }
+}
+
+/* Output a pointer value in hex.  Work with any size of pointer   */
+static void outputPtrBytes(void *val, char *outBuf, int32_t *outIx, int32_t capacity) {
+    uint32_t  i;
+    int32_t  incVal = 1;              /* +1 for big endian, -1 for little endian          */
+    char     *p     = (char *)&val;   /* point to current byte to output in the ptr val  */
+
+#if !U_IS_BIG_ENDIAN
+    /* Little Endian.  Move p to most significant end of the value      */
+    incVal = -1;
+    p += sizeof(void *) - 1;
+#endif
+
+    /* Loop through the bytes of the ptr as it sits in memory, from 
+     * most significant to least significant end                    */
+    for (i=0; i<sizeof(void *); i++) {
+        outputHexBytes(*p, 2, outBuf, outIx, capacity);
+        p += incVal;
+    }
+}
+
+static void outputString(const char *s, char *outBuf, int32_t *outIx, int32_t capacity, int32_t indent) {
+    int32_t i = 0;
+    char    c;
+    if (s==NULL) {
+        s = "*NULL*";
+    }
+    do {
+        c = s[i++];
+        outputChar(c, outBuf, outIx, capacity, indent);
+    } while (c != 0);
+}
+        
+
+
+static void outputUString(const UChar *s, int32_t len, 
+                          char *outBuf, int32_t *outIx, int32_t capacity, int32_t indent) {
+    int32_t i = 0;
+    UChar   c;
+    if (s==NULL) {
+        outputString(NULL, outBuf, outIx, capacity, indent);
+        return;
+    }
+
+    for (i=0; i<len || len==-1; i++) {
+        c = s[i];
+        outputHexBytes(c, 4, outBuf, outIx, capacity);
+        outputChar(' ', outBuf, outIx, capacity, indent);
+        if (len == -1 && c==0) {
+            break;
+        }
+    }
+}
+        
+U_CAPI int32_t U_EXPORT2
+utrace_vformat(char *outBuf, int32_t capacity, int32_t indent, const char *fmt, va_list args) {
+    int32_t   outIx  = 0;
+    int32_t   fmtIx  = 0;
+    char      fmtC;
+    char      c;
+    int32_t   intArg;
+    int64_t   longArg = 0;
+    char      *ptrArg;
+
+    /*   Loop runs once for each character in the format string.
+     */
+    for (;;) {
+        fmtC = fmt[fmtIx++];
+        if (fmtC != '%') {
+            /* Literal character, not part of a %sequence.  Just copy it to the output. */
+            outputChar(fmtC, outBuf, &outIx, capacity, indent);
+            if (fmtC == 0) {
+                /* We hit the null that terminates the format string.
+                 * This is the normal (and only) exit from the loop that
+                 * interprets the format
+                 */
+                break;
+            }
+            continue;
+        }
+
+        /* We encountered a '%'.  Pick up the following format char */
+        fmtC = fmt[fmtIx++];
+
+        switch (fmtC) {
+        case 'c':
+            /* single 8 bit char   */
+            c = (char)va_arg(args, int32_t);
+            outputChar(c, outBuf, &outIx, capacity, indent);
+            break;
+
+        case 's':
+            /* char * string, null terminated.  */
+            ptrArg = va_arg(args, char *);
+            outputString((const char *)ptrArg, outBuf, &outIx, capacity, indent);
+            break;
+
+        case 'S':
+            /* UChar * string, with length, len==-1 for null terminated. */
+            ptrArg = va_arg(args, char *);             /* Ptr    */
+            intArg =(int32_t)va_arg(args, int32_t);    /* Length */
+            outputUString((const UChar *)ptrArg, intArg, outBuf, &outIx, capacity, indent);
+            break;
+
+        case 'b':
+            /*  8 bit int  */
+            intArg = va_arg(args, int);
+            outputHexBytes(intArg, 2, outBuf, &outIx, capacity);
+            break;
+
+        case 'h':
+            /*  16 bit int  */
+            intArg = va_arg(args, int);
+            outputHexBytes(intArg, 4, outBuf, &outIx, capacity);
+            break;
+
+        case 'd':
+            /*  32 bit int  */
+            intArg = va_arg(args, int);
+            outputHexBytes(intArg, 8, outBuf, &outIx, capacity);
+            break;
+
+        case 'l':
+            /*  64 bit long  */
+            longArg = va_arg(args, int64_t);
+            outputHexBytes(longArg, 16, outBuf, &outIx, capacity);
+            break;
+            
+        case 'p':
+            /*  Pointers.   */
+            ptrArg = va_arg(args, char *);
+            outputPtrBytes(ptrArg, outBuf, &outIx, capacity);
+            break;
+
+        case 0:
+            /* Single '%' at end of fmt string.  Output as literal '%'.   
+             * Back up index into format string so that the terminating null will be
+             * re-fetched in the outer loop, causing it to terminate.
+             */
+            outputChar('%', outBuf, &outIx, capacity, indent);
+            fmtIx--;
+            break;
+
+        case 'v':
+            {
+                /* Vector of values, e.g. %vh */
+                char     vectorType;
+                int32_t  vectorLen;
+                const char   *i8Ptr;
+                int16_t  *i16Ptr;
+                int32_t  *i32Ptr;
+                int64_t  *i64Ptr;
+                void     **ptrPtr;
+                int32_t   charsToOutput = 0;
+                int32_t   i;
+                
+                vectorType = fmt[fmtIx];    /* b, h, d, l, p, etc. */
+                if (vectorType != 0) {
+                    fmtIx++;
+                }
+                i8Ptr = (const char *)va_arg(args, void*);
+                i16Ptr = (int16_t *)i8Ptr;
+                i32Ptr = (int32_t *)i8Ptr;
+                i64Ptr = (int64_t *)i8Ptr;
+                ptrPtr = (void **)i8Ptr;
+                vectorLen =(int32_t)va_arg(args, int32_t);
+                if (ptrPtr == NULL) {
+                    outputString("*NULL* ", outBuf, &outIx, capacity, indent);
+                } else {
+                    for (i=0; i<vectorLen || vectorLen==-1; i++) { 
+                        switch (vectorType) {
+                        case 'b':
+                            charsToOutput = 2;
+                            longArg = *i8Ptr++;
+                            break;
+                        case 'h':
+                            charsToOutput = 4;
+                            longArg = *i16Ptr++;
+                            break;
+                        case 'd':
+                            charsToOutput = 8;
+                            longArg = *i32Ptr++;
+                            break;
+                        case 'l':
+                            charsToOutput = 16;
+                            longArg = *i64Ptr++;
+                            break;
+                        case 'p':
+                            charsToOutput = 0;
+                            outputPtrBytes(*ptrPtr, outBuf, &outIx, capacity);
+                            longArg = *ptrPtr==NULL? 0: 1;    /* test for null terminated array. */
+                            ptrPtr++;
+                            break;
+                        case 'c':
+                            charsToOutput = 0;
+                            outputChar(*i8Ptr, outBuf, &outIx, capacity, indent);
+                            longArg = *i8Ptr;    /* for test for null terminated array. */
+                            i8Ptr++;
+                            break;
+                        case 's':
+                            charsToOutput = 0;
+                            outputString((const char *)*ptrPtr, outBuf, &outIx, capacity, indent);
+                            outputChar('\n', outBuf, &outIx, capacity, indent);
+                            longArg = *ptrPtr==NULL? 0: 1;   /* for test for null term. array. */
+                            ptrPtr++;
+                            break;
+
+                        case 'S':
+                            charsToOutput = 0;
+                            outputUString((const UChar *)*ptrPtr, -1, outBuf, &outIx, capacity, indent);
+                            outputChar('\n', outBuf, &outIx, capacity, indent);
+                            longArg = *ptrPtr==NULL? 0: 1;   /* for test for null term. array. */
+                            ptrPtr++;
+                            break;
+
+                            
+                        }
+                        if (charsToOutput > 0) {
+                            outputHexBytes(longArg, charsToOutput, outBuf, &outIx, capacity);
+                            outputChar(' ', outBuf, &outIx, capacity, indent);
+                        }
+                        if (vectorLen == -1 && longArg == 0) {
+                            break;
+                        }
+                    }
+                }
+                outputChar('[', outBuf, &outIx, capacity, indent);
+                outputHexBytes(vectorLen, 8, outBuf, &outIx, capacity);
+                outputChar(']', outBuf, &outIx, capacity, indent);
+            }
+            break;
+
+
+        default:
+            /* %. in format string, where . is some character not in the set
+             *    of recognized format chars.  Just output it as if % wasn't there.
+             *    (Covers "%%" outputing a single '%')
+             */
+             outputChar(fmtC, outBuf, &outIx, capacity, indent);
+        }
+    }
+    outputChar(0, outBuf, &outIx, capacity, indent);  /* Make sure that output is null terminated  */
+    return outIx + 1;     /* outIx + 1 because outIx does not increment when outputing final null. */
+}
+
+
+
+
+U_CAPI int32_t U_EXPORT2
+utrace_format(char *outBuf, int32_t capacity,
+                int32_t indent, const char *fmt,  ...) {
+    int32_t retVal;
+    va_list args;
+    va_start(args, fmt ); 
+    retVal = utrace_vformat(outBuf, capacity, indent, fmt, args);
+    va_end(args);
+    return retVal;
+}
+
+
+U_CAPI void U_EXPORT2
+utrace_setFunctions(const void *context,
+                    UTraceEntry *e, UTraceExit *x, UTraceData *d) {
+    pTraceEntryFunc = e;
+    pTraceExitFunc  = x;
+    pTraceDataFunc  = d;
+    gTraceContext   = context;
+}
+
+
+U_CAPI void U_EXPORT2
+utrace_getFunctions(const void **context,
+                    UTraceEntry **e, UTraceExit **x, UTraceData **d) {
+    *e = pTraceEntryFunc;
+    *x = pTraceExitFunc;
+    *d = pTraceDataFunc;
+    *context = gTraceContext;
+}
+
+U_CAPI void U_EXPORT2
+utrace_setLevel(int32_t level) {
+    if (level < UTRACE_OFF) {
+        level = UTRACE_OFF;
+    }
+    if (level > UTRACE_VERBOSE) {
+        level = UTRACE_VERBOSE;
+    }
+    utrace_level = level;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrace_getLevel() {
+    return utrace_level;
+}
+
+
+U_CFUNC UBool 
+utrace_cleanup() {
+    pTraceEntryFunc = NULL;
+    pTraceExitFunc  = NULL;
+    pTraceDataFunc  = NULL;
+    utrace_level    = UTRACE_OFF;
+    gTraceContext   = NULL;
+    return TRUE;
+}
+
+
+static const char * const
+trFnName[] = {
+    "u_init",
+    "u_cleanup",
+    NULL
+};
+
+
+static const char * const
+trConvNames[] = {
+    "ucnv_open",
+    "ucnv_openPackage",
+    "ucnv_openAlgorithmic",
+    "ucnv_clone",
+    "ucnv_close",
+    "ucnv_flushCache",
+    "ucnv_load",
+    "ucnv_unload",
+    NULL
+};
+
+    
+static const char * const
+trCollNames[] = {
+    "ucol_open",
+    "ucol_close",
+    "ucol_strcoll",
+    "ucol_getSortKey",
+    "ucol_getLocale",
+    "ucol_nextSortKeyPart",
+    "ucol_strcollIter",
+    "ucol_openFromShortString",
+    "ucol_strcollUTF8",
+    NULL
+};
+
+
+static const char* const
+trResDataNames[] = {
+    "resc",
+    "bundle-open",
+    "file-open",
+    "res-open",
+    NULL
+};
+
+                
+U_CAPI const char * U_EXPORT2
+utrace_functionName(int32_t fnNumber) {
+    if(UTRACE_FUNCTION_START <= fnNumber && fnNumber < UTRACE_FUNCTION_LIMIT) {
+        return trFnName[fnNumber];
+    } else if(UTRACE_CONVERSION_START <= fnNumber && fnNumber < UTRACE_CONVERSION_LIMIT) {
+        return trConvNames[fnNumber - UTRACE_CONVERSION_START];
+    } else if(UTRACE_COLLATION_START <= fnNumber && fnNumber < UTRACE_COLLATION_LIMIT){
+        return trCollNames[fnNumber - UTRACE_COLLATION_START];
+    } else if(UTRACE_UDATA_START <= fnNumber && fnNumber < UTRACE_RES_DATA_LIMIT){
+        return trResDataNames[fnNumber - UTRACE_UDATA_START];
+    } else {
+        return "[BOGUS Trace Function Number]";
+    }
+}
+
diff --git a/thirdparty/icu4c/common/utracimp.h b/thirdparty/icu4c/common/utracimp.h
new file mode 100644
index 0000000000..f32fe1db39
--- /dev/null
+++ b/thirdparty/icu4c/common/utracimp.h
@@ -0,0 +1,391 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 2003-2009, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utracimp.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2003aug06
+*   created by: Markus W. Scherer
+*
+*   Internal header for ICU tracing/logging.
+*
+*
+*   Various notes:
+*   - using a trace level variable to only call trace functions
+*     when the level is sufficient
+*   - using the same variable for tracing on/off to never make a function
+*     call when off
+*   - the function number is put into a local variable by the entry macro
+*     and used implicitly to avoid copy&paste/typing mistakes by the developer
+*   - the application must call utrace_setFunctions() and pass in
+*     implementations for the trace functions
+*   - ICU trace macros call ICU functions that route through the function
+*     pointers if they have been set;
+*     this avoids an indirection at the call site
+*     (which would cost more code for another check and for the indirection)
+*
+*   ### TODO Issues:
+*   - Verify that va_list is portable among compilers for the same platform.
+*     va_list should be portable because printf() would fail otherwise!
+*   - Should enum values like UTraceLevel be passed into int32_t-type arguments,
+*     or should enum types be used?
+*/
+
+#ifndef __UTRACIMP_H__
+#define __UTRACIMP_H__
+
+#include "unicode/utrace.h"
+#include <stdarg.h>
+
+U_CDECL_BEGIN
+
+/** 
+ *   Traced Function Exit return types.  
+ *   Flags indicating the number and types of varargs included in a call
+ *   to a UTraceExit function.
+ *   Bits 0-3:  The function return type.  First variable param.
+ *   Bit    4:  Flag for presence of U_ErrorCode status param.
+ *   @internal
+ */
+typedef enum UTraceExitVal {
+    /** The traced function returns no value  @internal */
+    UTRACE_EXITV_NONE   = 0,
+    /** The traced function returns an int32_t, or compatible, type.  @internal */
+    UTRACE_EXITV_I32    = 1,
+    /** The traced function returns a pointer  @internal */
+    UTRACE_EXITV_PTR    = 2,
+    /** The traced function returns a UBool  @internal */
+    UTRACE_EXITV_BOOL   = 3,
+    /** Mask to extract the return type values from a UTraceExitVal  @internal */
+    UTRACE_EXITV_MASK   = 0xf,
+    /** Bit indicating that the traced function includes a UErrorCode parameter  @internal */
+    UTRACE_EXITV_STATUS = 0x10
+} UTraceExitVal;
+
+/**
+ * Trace function for the entry point of a function.
+ * Do not use directly, use UTRACE_ENTRY instead.
+ * @param fnNumber The UTraceFunctionNumber for the current function.
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+utrace_entry(int32_t fnNumber);
+
+/**
+ * Trace function for each exit point of a function.
+ * Do not use directly, use UTRACE_EXIT* instead.
+ * @param fnNumber The UTraceFunctionNumber for the current function.
+ * @param returnType The type of the value returned by the function.
+ * @param errorCode The UErrorCode value at function exit. See UTRACE_EXIT.
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+utrace_exit(int32_t fnNumber, int32_t returnType, ...);
+
+
+/**
+ * Trace function used inside functions that have a UTRACE_ENTRY() statement.
+ * Do not use directly, use UTRACE_DATAX() macros instead.
+ *
+ * @param utraceFnNumber The number of the current function, from the local
+ *        variable of the same name.
+ * @param level The trace level for this message.
+ * @param fmt The trace format string.
+ *
+ * @internal
+ */
+U_CAPI void U_EXPORT2
+utrace_data(int32_t utraceFnNumber, int32_t level, const char *fmt, ...);
+
+U_CDECL_END
+
+#if U_ENABLE_TRACING
+
+/**
+ * Boolean expression to see if ICU tracing is turned on
+ * to at least the specified level.
+ * @internal
+ */
+#define UTRACE_LEVEL(level) (utrace_getLevel()>=(level))
+
+/**
+  *  Flag bit in utraceFnNumber, the local variable added to each function 
+  *  with tracing code to contains the function number.
+  *
+  *  Set the flag if the function's entry is traced, which will cause the
+  *  function's exit to also be traced.  utraceFnNumber is uncoditionally 
+  *  set at entry, whether or not the entry is traced, so that it will
+  *  always be available for error trace output.
+  *  @internal
+  */            
+#define UTRACE_TRACED_ENTRY 0x80000000
+
+/**
+ * Trace statement for the entry point of a function.
+ * Stores the function number in a local variable.
+ * In C code, must be placed immediately after the last variable declaration.
+ * Must be matched with UTRACE_EXIT() at all function exit points.
+ *
+ * Tracing should start with UTRACE_ENTRY after checking for
+ * U_FAILURE at function entry, so that if a function returns immediately
+ * because of a pre-existing error condition, it does not show up in the trace,
+ * consistent with ICU's error handling model.
+ *
+ * @param fnNumber The UTraceFunctionNumber for the current function.
+ * @internal
+ */
+#define UTRACE_ENTRY(fnNumber) \
+    int32_t utraceFnNumber=(fnNumber); \
+UPRV_BLOCK_MACRO_BEGIN { \
+    if(utrace_getLevel()>=UTRACE_INFO) { \
+        utrace_entry(fnNumber); \
+        utraceFnNumber |= UTRACE_TRACED_ENTRY; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+
+/**
+ * Trace statement for the entry point of open and close functions.
+ * Produces trace output at a less verbose setting than plain UTRACE_ENTRY
+ * Stores the function number in a local variable.
+ * In C code, must be placed immediately after the last variable declaration.
+ * Must be matched with UTRACE_EXIT() at all function exit points.
+ *
+ * @param fnNumber The UTraceFunctionNumber for the current function.
+ * @internal
+ */
+#define UTRACE_ENTRY_OC(fnNumber) \
+    int32_t utraceFnNumber=(fnNumber); \
+UPRV_BLOCK_MACRO_BEGIN { \
+    if(utrace_getLevel()>=UTRACE_OPEN_CLOSE) { \
+        utrace_entry(fnNumber); \
+        utraceFnNumber |= UTRACE_TRACED_ENTRY; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement for each exit point of a function that has a UTRACE_ENTRY()
+ * statement.
+ *
+ * @param errorCode The function's ICU UErrorCode value at function exit,
+ *                  or U_ZERO_ERROR if the function does not use a UErrorCode.
+ *                  0==U_ZERO_ERROR indicates success,
+ *                  positive values an error (see u_errorName()),
+ *                  negative values an informational status.
+ *
+ * @internal
+ */
+#define UTRACE_EXIT() UPRV_BLOCK_MACRO_BEGIN { \
+    if(utraceFnNumber & UTRACE_TRACED_ENTRY) { \
+        utrace_exit(utraceFnNumber & ~UTRACE_TRACED_ENTRY, UTRACE_EXITV_NONE); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement for each exit point of a function that has a UTRACE_ENTRY()
+ * statement, and that returns a value.
+ *
+ * @param val       The function's return value, int32_t or comatible type.
+ *
+ * @internal 
+ */
+#define UTRACE_EXIT_VALUE(val) UPRV_BLOCK_MACRO_BEGIN { \
+    if(utraceFnNumber & UTRACE_TRACED_ENTRY) { \
+        utrace_exit(utraceFnNumber & ~UTRACE_TRACED_ENTRY, UTRACE_EXITV_I32, val); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#define UTRACE_EXIT_STATUS(status) UPRV_BLOCK_MACRO_BEGIN { \
+    if(utraceFnNumber & UTRACE_TRACED_ENTRY) { \
+        utrace_exit(utraceFnNumber & ~UTRACE_TRACED_ENTRY, UTRACE_EXITV_STATUS, status); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#define UTRACE_EXIT_VALUE_STATUS(val, status) UPRV_BLOCK_MACRO_BEGIN { \
+    if(utraceFnNumber & UTRACE_TRACED_ENTRY) { \
+        utrace_exit(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (UTRACE_EXITV_I32 | UTRACE_EXITV_STATUS), val, status); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#define UTRACE_EXIT_PTR_STATUS(ptr, status) UPRV_BLOCK_MACRO_BEGIN { \
+    if(utraceFnNumber & UTRACE_TRACED_ENTRY) { \
+        utrace_exit(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (UTRACE_EXITV_PTR | UTRACE_EXITV_STATUS), ptr, status); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes no data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA0(level, fmt) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes one data argument.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA1(level, fmt, a) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY , (level), (fmt), (a)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes two data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA2(level, fmt, a, b) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY , (level), (fmt), (a), (b)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes three data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA3(level, fmt, a, b, c) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes four data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA4(level, fmt, a, b, c, d) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes five data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA5(level, fmt, a, b, c, d, e) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d), (e)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes six data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA6(level, fmt, a, b, c, d, e, f) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d), (e), (f)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes seven data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA7(level, fmt, a, b, c, d, e, f, g) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d), (e), (f), (g)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes eight data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA8(level, fmt, a, b, c, d, e, f, g, h) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d), (e), (f), (g), (h)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Trace statement used inside functions that have a UTRACE_ENTRY() statement.
+ * Takes nine data arguments.
+ * The number of arguments for this macro must match the number of inserts
+ * in the format string. Vector inserts count as two arguments.
+ * Calls utrace_data() if the level is high enough.
+ * @internal
+ */
+#define UTRACE_DATA9(level, fmt, a, b, c, d, e, f, g, h, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(UTRACE_LEVEL(level)) { \
+        utrace_data(utraceFnNumber & ~UTRACE_TRACED_ENTRY, (level), (fmt), (a), (b), (c), (d), (e), (f), (g), (h), (i)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#else
+
+/*
+ * When tracing is disabled, the following macros become empty
+ */
+
+#define UTRACE_LEVEL(level) 0
+#define UTRACE_ENTRY(fnNumber)
+#define UTRACE_ENTRY_OC(fnNumber)
+#define UTRACE_EXIT()
+#define UTRACE_EXIT_VALUE(val)
+#define UTRACE_EXIT_STATUS(status)
+#define UTRACE_EXIT_VALUE_STATUS(val, status)
+#define UTRACE_EXIT_PTR_STATUS(ptr, status)
+#define UTRACE_DATA0(level, fmt)
+#define UTRACE_DATA1(level, fmt, a)
+#define UTRACE_DATA2(level, fmt, a, b)
+#define UTRACE_DATA3(level, fmt, a, b, c)
+#define UTRACE_DATA4(level, fmt, a, b, c, d)
+#define UTRACE_DATA5(level, fmt, a, b, c, d, e)
+#define UTRACE_DATA6(level, fmt, a, b, c, d, e, f)
+#define UTRACE_DATA7(level, fmt, a, b, c, d, e, f, g)
+#define UTRACE_DATA8(level, fmt, a, b, c, d, e, f, g, h)
+#define UTRACE_DATA9(level, fmt, a, b, c, d, e, f, g, h, i)
+
+#endif
+
+#endif
diff --git a/thirdparty/icu4c/common/utrie.cpp b/thirdparty/icu4c/common/utrie.cpp
new file mode 100644
index 0000000000..ecf9b1cba7
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie.cpp
@@ -0,0 +1,1234 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001oct20
+*   created by: Markus W. Scherer
+*
+*   This is a common implementation of a "folded" trie.
+*   It is a kind of compressed, serializable table of 16- or 32-bit values associated with
+*   Unicode code points (0..0x10ffff).
+*/
+
+#ifdef UTRIE_DEBUG
+#   include <stdio.h>
+#endif
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "utrie.h"
+
+/* miscellaneous ------------------------------------------------------------ */
+
+#undef ABS
+#define ABS(x) ((x)>=0 ? (x) : -(x))
+
+static inline UBool
+equal_uint32(const uint32_t *s, const uint32_t *t, int32_t length) {
+    while(length>0 && *s==*t) {
+        ++s;
+        ++t;
+        --length;
+    }
+    return (UBool)(length==0);
+}
+
+/* Building a trie ----------------------------------------------------------*/
+
+U_CAPI UNewTrie * U_EXPORT2
+utrie_open(UNewTrie *fillIn,
+           uint32_t *aliasData, int32_t maxDataLength,
+           uint32_t initialValue, uint32_t leadUnitValue,
+           UBool latin1Linear) {
+    UNewTrie *trie;
+    int32_t i, j;
+
+    if( maxDataLength<UTRIE_DATA_BLOCK_LENGTH ||
+        (latin1Linear && maxDataLength<1024)
+    ) {
+        return NULL;
+    }
+
+    if(fillIn!=NULL) {
+        trie=fillIn;
+    } else {
+        trie=(UNewTrie *)uprv_malloc(sizeof(UNewTrie));
+        if(trie==NULL) {
+            return NULL;
+        }
+    }
+    uprv_memset(trie, 0, sizeof(UNewTrie));
+    trie->isAllocated= (UBool)(fillIn==NULL);
+
+    if(aliasData!=NULL) {
+        trie->data=aliasData;
+        trie->isDataAllocated=FALSE;
+    } else {
+        trie->data=(uint32_t *)uprv_malloc(maxDataLength*4);
+        if(trie->data==NULL) {
+            uprv_free(trie);
+            return NULL;
+        }
+        trie->isDataAllocated=TRUE;
+    }
+
+    /* preallocate and reset the first data block (block index 0) */
+    j=UTRIE_DATA_BLOCK_LENGTH;
+
+    if(latin1Linear) {
+        /* preallocate and reset the first block (number 0) and Latin-1 (U+0000..U+00ff) after that */
+        /* made sure above that maxDataLength>=1024 */
+
+        /* set indexes to point to consecutive data blocks */
+        i=0;
+        do {
+            /* do this at least for trie->index[0] even if that block is only partly used for Latin-1 */
+            trie->index[i++]=j;
+            j+=UTRIE_DATA_BLOCK_LENGTH;
+        } while(i<(256>>UTRIE_SHIFT));
+    }
+
+    /* reset the initially allocated blocks to the initial value */
+    trie->dataLength=j;
+    while(j>0) {
+        trie->data[--j]=initialValue;
+    }
+
+    trie->leadUnitValue=leadUnitValue;
+    trie->indexLength=UTRIE_MAX_INDEX_LENGTH;
+    trie->dataCapacity=maxDataLength;
+    trie->isLatin1Linear=latin1Linear;
+    trie->isCompacted=FALSE;
+    return trie;
+}
+
+U_CAPI UNewTrie * U_EXPORT2
+utrie_clone(UNewTrie *fillIn, const UNewTrie *other, uint32_t *aliasData, int32_t aliasDataCapacity) {
+    UNewTrie *trie;
+    UBool isDataAllocated;
+
+    /* do not clone if other is not valid or already compacted */
+    if(other==NULL || other->data==NULL || other->isCompacted) {
+        return NULL;
+    }
+
+    /* clone data */
+    if(aliasData!=NULL && aliasDataCapacity>=other->dataCapacity) {
+        isDataAllocated=FALSE;
+    } else {
+        aliasDataCapacity=other->dataCapacity;
+        aliasData=(uint32_t *)uprv_malloc(other->dataCapacity*4);
+        if(aliasData==NULL) {
+            return NULL;
+        }
+        isDataAllocated=TRUE;
+    }
+
+    trie=utrie_open(fillIn, aliasData, aliasDataCapacity,
+                    other->data[0], other->leadUnitValue,
+                    other->isLatin1Linear);
+    if(trie==NULL) {
+        uprv_free(aliasData);
+    } else {
+        uprv_memcpy(trie->index, other->index, sizeof(trie->index));
+        uprv_memcpy(trie->data, other->data, (size_t)other->dataLength*4);
+        trie->dataLength=other->dataLength;
+        trie->isDataAllocated=isDataAllocated;
+    }
+
+    return trie;
+}
+
+U_CAPI void U_EXPORT2
+utrie_close(UNewTrie *trie) {
+    if(trie!=NULL) {
+        if(trie->isDataAllocated) {
+            uprv_free(trie->data);
+            trie->data=NULL;
+        }
+        if(trie->isAllocated) {
+            uprv_free(trie);
+        }
+    }
+}
+
+U_CAPI uint32_t * U_EXPORT2
+utrie_getData(UNewTrie *trie, int32_t *pLength) {
+    if(trie==NULL || pLength==NULL) {
+        return NULL;
+    }
+
+    *pLength=trie->dataLength;
+    return trie->data;
+}
+
+static int32_t
+utrie_allocDataBlock(UNewTrie *trie) {
+    int32_t newBlock, newTop;
+
+    newBlock=trie->dataLength;
+    newTop=newBlock+UTRIE_DATA_BLOCK_LENGTH;
+    if(newTop>trie->dataCapacity) {
+        /* out of memory in the data array */
+        return -1;
+    }
+    trie->dataLength=newTop;
+    return newBlock;
+}
+
+/**
+ * No error checking for illegal arguments.
+ *
+ * @return -1 if no new data block available (out of memory in data array)
+ * @internal
+ */
+static int32_t
+utrie_getDataBlock(UNewTrie *trie, UChar32 c) {
+    int32_t indexValue, newBlock;
+
+    c>>=UTRIE_SHIFT;
+    indexValue=trie->index[c];
+    if(indexValue>0) {
+        return indexValue;
+    }
+
+    /* allocate a new data block */
+    newBlock=utrie_allocDataBlock(trie);
+    if(newBlock<0) {
+        /* out of memory in the data array */
+        return -1;
+    }
+    trie->index[c]=newBlock;
+
+    /* copy-on-write for a block from a setRange() */
+    uprv_memcpy(trie->data+newBlock, trie->data-indexValue, 4*UTRIE_DATA_BLOCK_LENGTH);
+    return newBlock;
+}
+
+/**
+ * @return TRUE if the value was successfully set
+ */
+U_CAPI UBool U_EXPORT2
+utrie_set32(UNewTrie *trie, UChar32 c, uint32_t value) {
+    int32_t block;
+
+    /* valid, uncompacted trie and valid c? */
+    if(trie==NULL || trie->isCompacted || (uint32_t)c>0x10ffff) {
+        return FALSE;
+    }
+
+    block=utrie_getDataBlock(trie, c);
+    if(block<0) {
+        return FALSE;
+    }
+
+    trie->data[block+(c&UTRIE_MASK)]=value;
+    return TRUE;
+}
+
+U_CAPI uint32_t U_EXPORT2
+utrie_get32(UNewTrie *trie, UChar32 c, UBool *pInBlockZero) {
+    int32_t block;
+
+    /* valid, uncompacted trie and valid c? */
+    if(trie==NULL || trie->isCompacted || (uint32_t)c>0x10ffff) {
+        if(pInBlockZero!=NULL) {
+            *pInBlockZero=TRUE;
+        }
+        return 0;
+    }
+
+    block=trie->index[c>>UTRIE_SHIFT];
+    if(pInBlockZero!=NULL) {
+        *pInBlockZero= (UBool)(block==0);
+    }
+
+    return trie->data[ABS(block)+(c&UTRIE_MASK)];
+}
+
+/**
+ * @internal
+ */
+static void
+utrie_fillBlock(uint32_t *block, UChar32 start, UChar32 limit,
+                uint32_t value, uint32_t initialValue, UBool overwrite) {
+    uint32_t *pLimit;
+
+    pLimit=block+limit;
+    block+=start;
+    if(overwrite) {
+        while(block<pLimit) {
+            *block++=value;
+        }
+    } else {
+        while(block<pLimit) {
+            if(*block==initialValue) {
+                *block=value;
+            }
+            ++block;
+        }
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+utrie_setRange32(UNewTrie *trie, UChar32 start, UChar32 limit, uint32_t value, UBool overwrite) {
+    /*
+     * repeat value in [start..limit[
+     * mark index values for repeat-data blocks by setting bit 31 of the index values
+     * fill around existing values if any, if(overwrite)
+     */
+    uint32_t initialValue;
+    int32_t block, rest, repeatBlock;
+
+    /* valid, uncompacted trie and valid indexes? */
+    if( trie==NULL || trie->isCompacted ||
+        (uint32_t)start>0x10ffff || (uint32_t)limit>0x110000 || start>limit
+    ) {
+        return FALSE;
+    }
+    if(start==limit) {
+        return TRUE; /* nothing to do */
+    }
+
+    initialValue=trie->data[0];
+    if(start&UTRIE_MASK) {
+        UChar32 nextStart;
+
+        /* set partial block at [start..following block boundary[ */
+        block=utrie_getDataBlock(trie, start);
+        if(block<0) {
+            return FALSE;
+        }
+
+        nextStart=(start+UTRIE_DATA_BLOCK_LENGTH)&~UTRIE_MASK;
+        if(nextStart<=limit) {
+            utrie_fillBlock(trie->data+block, start&UTRIE_MASK, UTRIE_DATA_BLOCK_LENGTH,
+                            value, initialValue, overwrite);
+            start=nextStart;
+        } else {
+            utrie_fillBlock(trie->data+block, start&UTRIE_MASK, limit&UTRIE_MASK,
+                            value, initialValue, overwrite);
+            return TRUE;
+        }
+    }
+
+    /* number of positions in the last, partial block */
+    rest=limit&UTRIE_MASK;
+
+    /* round down limit to a block boundary */
+    limit&=~UTRIE_MASK;
+
+    /* iterate over all-value blocks */
+    if(value==initialValue) {
+        repeatBlock=0;
+    } else {
+        repeatBlock=-1;
+    }
+    while(start<limit) {
+        /* get index value */
+        block=trie->index[start>>UTRIE_SHIFT];
+        if(block>0) {
+            /* already allocated, fill in value */
+            utrie_fillBlock(trie->data+block, 0, UTRIE_DATA_BLOCK_LENGTH, value, initialValue, overwrite);
+        } else if(trie->data[-block]!=value && (block==0 || overwrite)) {
+            /* set the repeatBlock instead of the current block 0 or range block */
+            if(repeatBlock>=0) {
+                trie->index[start>>UTRIE_SHIFT]=-repeatBlock;
+            } else {
+                /* create and set and fill the repeatBlock */
+                repeatBlock=utrie_getDataBlock(trie, start);
+                if(repeatBlock<0) {
+                    return FALSE;
+                }
+
+                /* set the negative block number to indicate that it is a repeat block */
+                trie->index[start>>UTRIE_SHIFT]=-repeatBlock;
+                utrie_fillBlock(trie->data+repeatBlock, 0, UTRIE_DATA_BLOCK_LENGTH, value, initialValue, TRUE);
+            }
+        }
+
+        start+=UTRIE_DATA_BLOCK_LENGTH;
+    }
+
+    if(rest>0) {
+        /* set partial block at [last block boundary..limit[ */
+        block=utrie_getDataBlock(trie, start);
+        if(block<0) {
+            return FALSE;
+        }
+
+        utrie_fillBlock(trie->data+block, 0, rest, value, initialValue, overwrite);
+    }
+
+    return TRUE;
+}
+
+static int32_t
+_findSameIndexBlock(const int32_t *idx, int32_t indexLength,
+                    int32_t otherBlock) {
+    int32_t block, i;
+
+    for(block=UTRIE_BMP_INDEX_LENGTH; block<indexLength; block+=UTRIE_SURROGATE_BLOCK_COUNT) {
+        for(i=0; i<UTRIE_SURROGATE_BLOCK_COUNT; ++i) {
+            if(idx[block+i]!=idx[otherBlock+i]) {
+                break;
+            }
+        }
+        if(i==UTRIE_SURROGATE_BLOCK_COUNT) {
+            return block;
+        }
+    }
+    return indexLength;
+}
+
+/*
+ * Fold the normalization data for supplementary code points into
+ * a compact area on top of the BMP-part of the trie index,
+ * with the lead surrogates indexing this compact area.
+ *
+ * Duplicate the index values for lead surrogates:
+ * From inside the BMP area, where some may be overridden with folded values,
+ * to just after the BMP area, where they can be retrieved for
+ * code point lookups.
+ */
+static void
+utrie_fold(UNewTrie *trie, UNewTrieGetFoldedValue *getFoldedValue, UErrorCode *pErrorCode) {
+    int32_t leadIndexes[UTRIE_SURROGATE_BLOCK_COUNT];
+    int32_t *idx;
+    uint32_t value;
+    UChar32 c;
+    int32_t indexLength, block;
+#ifdef UTRIE_DEBUG
+    int countLeadCUWithData=0;
+#endif
+
+    idx=trie->index;
+
+    /* copy the lead surrogate indexes into a temporary array */
+    uprv_memcpy(leadIndexes, idx+(0xd800>>UTRIE_SHIFT), 4*UTRIE_SURROGATE_BLOCK_COUNT);
+
+    /*
+     * set all values for lead surrogate code *units* to leadUnitValue
+     * so that, by default, runtime lookups will find no data for associated
+     * supplementary code points, unless there is data for such code points
+     * which will result in a non-zero folding value below that is set for
+     * the respective lead units
+     *
+     * the above saved the indexes for surrogate code *points*
+     * fill the indexes with simplified code from utrie_setRange32()
+     */
+    if(trie->leadUnitValue==trie->data[0]) {
+        block=0; /* leadUnitValue==initialValue, use all-initial-value block */
+    } else {
+        /* create and fill the repeatBlock */
+        block=utrie_allocDataBlock(trie);
+        if(block<0) {
+            /* data table overflow */
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+        utrie_fillBlock(trie->data+block, 0, UTRIE_DATA_BLOCK_LENGTH, trie->leadUnitValue, trie->data[0], TRUE);
+        block=-block; /* negative block number to indicate that it is a repeat block */
+    }
+    for(c=(0xd800>>UTRIE_SHIFT); c<(0xdc00>>UTRIE_SHIFT); ++c) {
+        trie->index[c]=block;
+    }
+
+    /*
+     * Fold significant index values into the area just after the BMP indexes.
+     * In case the first lead surrogate has significant data,
+     * its index block must be used first (in which case the folding is a no-op).
+     * Later all folded index blocks are moved up one to insert the copied
+     * lead surrogate indexes.
+     */
+    indexLength=UTRIE_BMP_INDEX_LENGTH;
+
+    /* search for any index (stage 1) entries for supplementary code points */
+    for(c=0x10000; c<0x110000;) {
+        if(idx[c>>UTRIE_SHIFT]!=0) {
+            /* there is data, treat the full block for a lead surrogate */
+            c&=~0x3ff;
+
+#ifdef UTRIE_DEBUG
+            ++countLeadCUWithData;
+            /* printf("supplementary data for lead surrogate U+%04lx\n", (long)(0xd7c0+(c>>10))); */
+#endif
+
+            /* is there an identical index block? */
+            block=_findSameIndexBlock(idx, indexLength, c>>UTRIE_SHIFT);
+
+            /*
+             * get a folded value for [c..c+0x400[ and,
+             * if different from the value for the lead surrogate code point,
+             * set it for the lead surrogate code unit
+             */
+            value=getFoldedValue(trie, c, block+UTRIE_SURROGATE_BLOCK_COUNT);
+            if(value!=utrie_get32(trie, U16_LEAD(c), NULL)) {
+                if(!utrie_set32(trie, U16_LEAD(c), value)) {
+                    /* data table overflow */
+                    *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return;
+                }
+
+                /* if we did not find an identical index block... */
+                if(block==indexLength) {
+                    /* move the actual index (stage 1) entries from the supplementary position to the new one */
+                    uprv_memmove(idx+indexLength,
+                                 idx+(c>>UTRIE_SHIFT),
+                                 4*UTRIE_SURROGATE_BLOCK_COUNT);
+                    indexLength+=UTRIE_SURROGATE_BLOCK_COUNT;
+                }
+            }
+            c+=0x400;
+        } else {
+            c+=UTRIE_DATA_BLOCK_LENGTH;
+        }
+    }
+#ifdef UTRIE_DEBUG
+    if(countLeadCUWithData>0) {
+        printf("supplementary data for %d lead surrogates\n", countLeadCUWithData);
+    }
+#endif
+
+    /*
+     * index array overflow?
+     * This is to guarantee that a folding offset is of the form
+     * UTRIE_BMP_INDEX_LENGTH+n*UTRIE_SURROGATE_BLOCK_COUNT with n=0..1023.
+     * If the index is too large, then n>=1024 and more than 10 bits are necessary.
+     *
+     * In fact, it can only ever become n==1024 with completely unfoldable data and
+     * the additional block of duplicated values for lead surrogates.
+     */
+    if(indexLength>=UTRIE_MAX_INDEX_LENGTH) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+
+    /*
+     * make space for the lead surrogate index block and
+     * insert it between the BMP indexes and the folded ones
+     */
+    uprv_memmove(idx+UTRIE_BMP_INDEX_LENGTH+UTRIE_SURROGATE_BLOCK_COUNT,
+                 idx+UTRIE_BMP_INDEX_LENGTH,
+                 4*(indexLength-UTRIE_BMP_INDEX_LENGTH));
+    uprv_memcpy(idx+UTRIE_BMP_INDEX_LENGTH,
+                leadIndexes,
+                4*UTRIE_SURROGATE_BLOCK_COUNT);
+    indexLength+=UTRIE_SURROGATE_BLOCK_COUNT;
+
+#ifdef UTRIE_DEBUG
+    printf("trie index count: BMP %ld  all Unicode %ld  folded %ld\n",
+           UTRIE_BMP_INDEX_LENGTH, (long)UTRIE_MAX_INDEX_LENGTH, indexLength);
+#endif
+
+    trie->indexLength=indexLength;
+}
+
+/*
+ * Set a value in the trie index map to indicate which data block
+ * is referenced and which one is not.
+ * utrie_compact() will remove data blocks that are not used at all.
+ * Set
+ * - 0 if it is used
+ * - -1 if it is not used
+ */
+static void
+_findUnusedBlocks(UNewTrie *trie) {
+    int32_t i;
+
+    /* fill the entire map with "not used" */
+    uprv_memset(trie->map, 0xff, (UTRIE_MAX_BUILD_TIME_DATA_LENGTH>>UTRIE_SHIFT)*4);
+
+    /* mark each block that _is_ used with 0 */
+    for(i=0; i<trie->indexLength; ++i) {
+        trie->map[ABS(trie->index[i])>>UTRIE_SHIFT]=0;
+    }
+
+    /* never move the all-initial-value block 0 */
+    trie->map[0]=0;
+}
+
+static int32_t
+_findSameDataBlock(const uint32_t *data, int32_t dataLength,
+                   int32_t otherBlock, int32_t step) {
+    int32_t block;
+
+    /* ensure that we do not even partially get past dataLength */
+    dataLength-=UTRIE_DATA_BLOCK_LENGTH;
+
+    for(block=0; block<=dataLength; block+=step) {
+        if(equal_uint32(data+block, data+otherBlock, UTRIE_DATA_BLOCK_LENGTH)) {
+            return block;
+        }
+    }
+    return -1;
+}
+
+/*
+ * Compact a folded build-time trie.
+ *
+ * The compaction
+ * - removes blocks that are identical with earlier ones
+ * - overlaps adjacent blocks as much as possible (if overlap==TRUE)
+ * - moves blocks in steps of the data granularity
+ * - moves and overlaps blocks that overlap with multiple values in the overlap region
+ *
+ * It does not
+ * - try to move and overlap blocks that are not already adjacent
+ */
+static void
+utrie_compact(UNewTrie *trie, UBool overlap, UErrorCode *pErrorCode) {
+    int32_t i, start, newStart, overlapStart;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return;
+    }
+
+    /* valid, uncompacted trie? */
+    if(trie==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    if(trie->isCompacted) {
+        return; /* nothing left to do */
+    }
+
+    /* compaction */
+
+    /* initialize the index map with "block is used/unused" flags */
+    _findUnusedBlocks(trie);
+
+    /* if Latin-1 is preallocated and linear, then do not compact Latin-1 data */
+    if(trie->isLatin1Linear && UTRIE_SHIFT<=8) {
+        overlapStart=UTRIE_DATA_BLOCK_LENGTH+256;
+    } else {
+        overlapStart=UTRIE_DATA_BLOCK_LENGTH;
+    }
+
+    newStart=UTRIE_DATA_BLOCK_LENGTH;
+    for(start=newStart; start<trie->dataLength;) {
+        /*
+         * start: index of first entry of current block
+         * newStart: index where the current block is to be moved
+         *           (right after current end of already-compacted data)
+         */
+
+        /* skip blocks that are not used */
+        if(trie->map[start>>UTRIE_SHIFT]<0) {
+            /* advance start to the next block */
+            start+=UTRIE_DATA_BLOCK_LENGTH;
+
+            /* leave newStart with the previous block! */
+            continue;
+        }
+
+        /* search for an identical block */
+        if( start>=overlapStart &&
+            (i=_findSameDataBlock(trie->data, newStart, start,
+                            overlap ? UTRIE_DATA_GRANULARITY : UTRIE_DATA_BLOCK_LENGTH))
+             >=0
+        ) {
+            /* found an identical block, set the other block's index value for the current block */
+            trie->map[start>>UTRIE_SHIFT]=i;
+
+            /* advance start to the next block */
+            start+=UTRIE_DATA_BLOCK_LENGTH;
+
+            /* leave newStart with the previous block! */
+            continue;
+        }
+
+        /* see if the beginning of this block can be overlapped with the end of the previous block */
+        if(overlap && start>=overlapStart) {
+            /* look for maximum overlap (modulo granularity) with the previous, adjacent block */
+            for(i=UTRIE_DATA_BLOCK_LENGTH-UTRIE_DATA_GRANULARITY;
+                i>0 && !equal_uint32(trie->data+(newStart-i), trie->data+start, i);
+                i-=UTRIE_DATA_GRANULARITY) {}
+        } else {
+            i=0;
+        }
+
+        if(i>0) {
+            /* some overlap */
+            trie->map[start>>UTRIE_SHIFT]=newStart-i;
+
+            /* move the non-overlapping indexes to their new positions */
+            start+=i;
+            for(i=UTRIE_DATA_BLOCK_LENGTH-i; i>0; --i) {
+                trie->data[newStart++]=trie->data[start++];
+            }
+        } else if(newStart<start) {
+            /* no overlap, just move the indexes to their new positions */
+            trie->map[start>>UTRIE_SHIFT]=newStart;
+            for(i=UTRIE_DATA_BLOCK_LENGTH; i>0; --i) {
+                trie->data[newStart++]=trie->data[start++];
+            }
+        } else /* no overlap && newStart==start */ {
+            trie->map[start>>UTRIE_SHIFT]=start;
+            newStart+=UTRIE_DATA_BLOCK_LENGTH;
+            start=newStart;
+        }
+    }
+
+    /* now adjust the index (stage 1) table */
+    for(i=0; i<trie->indexLength; ++i) {
+        trie->index[i]=trie->map[ABS(trie->index[i])>>UTRIE_SHIFT];
+    }
+
+#ifdef UTRIE_DEBUG
+    /* we saved some space */
+    printf("compacting trie: count of 32-bit words %lu->%lu\n",
+            (long)trie->dataLength, (long)newStart);
+#endif
+
+    trie->dataLength=newStart;
+}
+
+/* serialization ------------------------------------------------------------ */
+
+/*
+ * Default function for the folding value:
+ * Just store the offset (16 bits) if there is any non-initial-value entry.
+ *
+ * The offset parameter is never 0.
+ * Returning the offset itself is safe for UTRIE_SHIFT>=5 because
+ * for UTRIE_SHIFT==5 the maximum index length is UTRIE_MAX_INDEX_LENGTH==0x8800
+ * which fits into 16-bit trie values;
+ * for higher UTRIE_SHIFT, UTRIE_MAX_INDEX_LENGTH decreases.
+ *
+ * Theoretically, it would be safer for all possible UTRIE_SHIFT including
+ * those of 4 and lower to return offset>>UTRIE_SURROGATE_BLOCK_BITS
+ * which would always result in a value of 0x40..0x43f
+ * (start/end 1k blocks of supplementary Unicode code points).
+ * However, this would be uglier, and would not work for some existing
+ * binary data file formats.
+ *
+ * Also, we do not plan to change UTRIE_SHIFT because it would change binary
+ * data file formats, and we would probably not make it smaller because of
+ * the then even larger BMP index length even for empty tries.
+ */
+static uint32_t U_CALLCONV
+defaultGetFoldedValue(UNewTrie *trie, UChar32 start, int32_t offset) {
+    uint32_t value, initialValue;
+    UChar32 limit;
+    UBool inBlockZero;
+
+    initialValue=trie->data[0];
+    limit=start+0x400;
+    while(start<limit) {
+        value=utrie_get32(trie, start, &inBlockZero);
+        if(inBlockZero) {
+            start+=UTRIE_DATA_BLOCK_LENGTH;
+        } else if(value!=initialValue) {
+            return (uint32_t)offset;
+        } else {
+            ++start;
+        }
+    }
+    return 0;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie_serialize(UNewTrie *trie, void *dt, int32_t capacity,
+                UNewTrieGetFoldedValue *getFoldedValue,
+                UBool reduceTo16Bits,
+                UErrorCode *pErrorCode) {
+    UTrieHeader *header;
+    uint32_t *p;
+    uint16_t *dest16;
+    int32_t i, length;
+    uint8_t* data = NULL;
+
+    /* argument check */
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(trie==NULL || capacity<0 || (capacity>0 && dt==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+    if(getFoldedValue==NULL) {
+        getFoldedValue=defaultGetFoldedValue;
+    }
+
+    data = (uint8_t*)dt;
+    /* fold and compact if necessary, also checks that indexLength is within limits */
+    if(!trie->isCompacted) {
+        /* compact once without overlap to improve folding */
+        utrie_compact(trie, FALSE, pErrorCode);
+
+        /* fold the supplementary part of the index array */
+        utrie_fold(trie, getFoldedValue, pErrorCode);
+
+        /* compact again with overlap for minimum data array length */
+        utrie_compact(trie, TRUE, pErrorCode);
+
+        trie->isCompacted=TRUE;
+        if(U_FAILURE(*pErrorCode)) {
+            return 0;
+        }
+    }
+
+    /* is dataLength within limits? */
+    if( (reduceTo16Bits ? (trie->dataLength+trie->indexLength) : trie->dataLength) >= UTRIE_MAX_DATA_LENGTH) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+    }
+
+    length=sizeof(UTrieHeader)+2*trie->indexLength;
+    if(reduceTo16Bits) {
+        length+=2*trie->dataLength;
+    } else {
+        length+=4*trie->dataLength;
+    }
+
+    if(length>capacity) {
+        return length; /* preflighting */
+    }
+
+#ifdef UTRIE_DEBUG
+    printf("**UTrieLengths(serialize)** index:%6ld  data:%6ld  serialized:%6ld\n",
+           (long)trie->indexLength, (long)trie->dataLength, (long)length);
+#endif
+
+    /* set the header fields */
+    header=(UTrieHeader *)data;
+    data+=sizeof(UTrieHeader);
+
+    header->signature=0x54726965; /* "Trie" */
+    header->options=UTRIE_SHIFT | (UTRIE_INDEX_SHIFT<<UTRIE_OPTIONS_INDEX_SHIFT);
+
+    if(!reduceTo16Bits) {
+        header->options|=UTRIE_OPTIONS_DATA_IS_32_BIT;
+    }
+    if(trie->isLatin1Linear) {
+        header->options|=UTRIE_OPTIONS_LATIN1_IS_LINEAR;
+    }
+
+    header->indexLength=trie->indexLength;
+    header->dataLength=trie->dataLength;
+
+    /* write the index (stage 1) array and the 16/32-bit data (stage 2) array */
+    if(reduceTo16Bits) {
+        /* write 16-bit index values shifted right by UTRIE_INDEX_SHIFT, after adding indexLength */
+        p=(uint32_t *)trie->index;
+        dest16=(uint16_t *)data;
+        for(i=trie->indexLength; i>0; --i) {
+            *dest16++=(uint16_t)((*p++ + trie->indexLength)>>UTRIE_INDEX_SHIFT);
+        }
+
+        /* write 16-bit data values */
+        p=trie->data;
+        for(i=trie->dataLength; i>0; --i) {
+            *dest16++=(uint16_t)*p++;
+        }
+    } else {
+        /* write 16-bit index values shifted right by UTRIE_INDEX_SHIFT */
+        p=(uint32_t *)trie->index;
+        dest16=(uint16_t *)data;
+        for(i=trie->indexLength; i>0; --i) {
+            *dest16++=(uint16_t)(*p++ >> UTRIE_INDEX_SHIFT);
+        }
+
+        /* write 32-bit data values */
+        uprv_memcpy(dest16, trie->data, 4*(size_t)trie->dataLength);
+    }
+
+    return length;
+}
+
+/* inverse to defaultGetFoldedValue() */
+U_CAPI int32_t U_EXPORT2
+utrie_defaultGetFoldingOffset(uint32_t data) {
+    return (int32_t)data;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie_unserialize(UTrie *trie, const void *data, int32_t length, UErrorCode *pErrorCode) {
+    const UTrieHeader *header;
+    const uint16_t *p16;
+    uint32_t options;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return -1;
+    }
+
+    /* enough data for a trie header? */
+    if(length<(int32_t)sizeof(UTrieHeader)) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return -1;
+    }
+
+    /* check the signature */
+    header=(const UTrieHeader *)data;
+    if(header->signature!=0x54726965) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return -1;
+    }
+
+    /* get the options and check the shift values */
+    options=header->options;
+    if( (options&UTRIE_OPTIONS_SHIFT_MASK)!=UTRIE_SHIFT ||
+        ((options>>UTRIE_OPTIONS_INDEX_SHIFT)&UTRIE_OPTIONS_SHIFT_MASK)!=UTRIE_INDEX_SHIFT
+    ) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return -1;
+    }
+    trie->isLatin1Linear= (UBool)((options&UTRIE_OPTIONS_LATIN1_IS_LINEAR)!=0);
+
+    /* get the length values */
+    trie->indexLength=header->indexLength;
+    trie->dataLength=header->dataLength;
+
+    length-=(int32_t)sizeof(UTrieHeader);
+
+    /* enough data for the index? */
+    if(length<2*trie->indexLength) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return -1;
+    }
+    p16=(const uint16_t *)(header+1);
+    trie->index=p16;
+    p16+=trie->indexLength;
+    length-=2*trie->indexLength;
+
+    /* get the data */
+    if(options&UTRIE_OPTIONS_DATA_IS_32_BIT) {
+        if(length<4*trie->dataLength) {
+            *pErrorCode=U_INVALID_FORMAT_ERROR;
+            return -1;
+        }
+        trie->data32=(const uint32_t *)p16;
+        trie->initialValue=trie->data32[0];
+        length=(int32_t)sizeof(UTrieHeader)+2*trie->indexLength+4*trie->dataLength;
+    } else {
+        if(length<2*trie->dataLength) {
+            *pErrorCode=U_INVALID_FORMAT_ERROR;
+            return -1;
+        }
+
+        /* the "data16" data is used via the index pointer */
+        trie->data32=NULL;
+        trie->initialValue=trie->index[trie->indexLength];
+        length=(int32_t)sizeof(UTrieHeader)+2*trie->indexLength+2*trie->dataLength;
+    }
+
+    trie->getFoldingOffset=utrie_defaultGetFoldingOffset;
+
+    return length;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie_unserializeDummy(UTrie *trie,
+                       void *data, int32_t length,
+                       uint32_t initialValue, uint32_t leadUnitValue,
+                       UBool make16BitTrie,
+                       UErrorCode *pErrorCode) {
+    uint16_t *p16;
+    int32_t actualLength, latin1Length, i, limit;
+    uint16_t block;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return -1;
+    }
+
+    /* calculate the actual size of the dummy trie data */
+
+    /* max(Latin-1, block 0) */
+    latin1Length= 256; /*UTRIE_SHIFT<=8 ? 256 : UTRIE_DATA_BLOCK_LENGTH;*/
+
+    trie->indexLength=UTRIE_BMP_INDEX_LENGTH+UTRIE_SURROGATE_BLOCK_COUNT;
+    trie->dataLength=latin1Length;
+    if(leadUnitValue!=initialValue) {
+        trie->dataLength+=UTRIE_DATA_BLOCK_LENGTH;
+    }
+
+    actualLength=trie->indexLength*2;
+    if(make16BitTrie) {
+        actualLength+=trie->dataLength*2;
+    } else {
+        actualLength+=trie->dataLength*4;
+    }
+
+    /* enough space for the dummy trie? */
+    if(length<actualLength) {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+        return actualLength;
+    }
+
+    trie->isLatin1Linear=TRUE;
+    trie->initialValue=initialValue;
+
+    /* fill the index and data arrays */
+    p16=(uint16_t *)data;
+    trie->index=p16;
+
+    if(make16BitTrie) {
+        /* indexes to block 0 */
+        block=(uint16_t)(trie->indexLength>>UTRIE_INDEX_SHIFT);
+        limit=trie->indexLength;
+        for(i=0; i<limit; ++i) {
+            p16[i]=block;
+        }
+
+        if(leadUnitValue!=initialValue) {
+            /* indexes for lead surrogate code units to the block after Latin-1 */
+            block+=(uint16_t)(latin1Length>>UTRIE_INDEX_SHIFT);
+            i=0xd800>>UTRIE_SHIFT;
+            limit=0xdc00>>UTRIE_SHIFT;
+            for(; i<limit; ++i) {
+                p16[i]=block;
+            }
+        }
+
+        trie->data32=NULL;
+
+        /* Latin-1 data */
+        p16+=trie->indexLength;
+        for(i=0; i<latin1Length; ++i) {
+            p16[i]=(uint16_t)initialValue;
+        }
+
+        /* data for lead surrogate code units */
+        if(leadUnitValue!=initialValue) {
+            limit=latin1Length+UTRIE_DATA_BLOCK_LENGTH;
+            for(/* i=latin1Length */; i<limit; ++i) {
+                p16[i]=(uint16_t)leadUnitValue;
+            }
+        }
+    } else {
+        uint32_t *p32;
+
+        /* indexes to block 0 */
+        uprv_memset(p16, 0, trie->indexLength*2);
+
+        if(leadUnitValue!=initialValue) {
+            /* indexes for lead surrogate code units to the block after Latin-1 */
+            block=(uint16_t)(latin1Length>>UTRIE_INDEX_SHIFT);
+            i=0xd800>>UTRIE_SHIFT;
+            limit=0xdc00>>UTRIE_SHIFT;
+            for(; i<limit; ++i) {
+                p16[i]=block;
+            }
+        }
+
+        trie->data32=p32=(uint32_t *)(p16+trie->indexLength);
+
+        /* Latin-1 data */
+        for(i=0; i<latin1Length; ++i) {
+            p32[i]=initialValue;
+        }
+
+        /* data for lead surrogate code units */
+        if(leadUnitValue!=initialValue) {
+            limit=latin1Length+UTRIE_DATA_BLOCK_LENGTH;
+            for(/* i=latin1Length */; i<limit; ++i) {
+                p32[i]=leadUnitValue;
+            }
+        }
+    }
+
+    trie->getFoldingOffset=utrie_defaultGetFoldingOffset;
+
+    return actualLength;
+}
+
+/* enumeration -------------------------------------------------------------- */
+
+/* default UTrieEnumValue() returns the input value itself */
+static uint32_t U_CALLCONV
+enumSameValue(const void * /*context*/, uint32_t value) {
+    return value;
+}
+
+/**
+ * Enumerate all ranges of code points with the same relevant values.
+ * The values are transformed from the raw trie entries by the enumValue function.
+ */
+U_CAPI void U_EXPORT2
+utrie_enum(const UTrie *trie,
+           UTrieEnumValue *enumValue, UTrieEnumRange *enumRange, const void *context) {
+    const uint32_t *data32;
+    const uint16_t *idx;
+
+    uint32_t value, prevValue, initialValue;
+    UChar32 c, prev;
+    int32_t l, i, j, block, prevBlock, nullBlock, offset;
+
+    /* check arguments */
+    if(trie==NULL || trie->index==NULL || enumRange==NULL) {
+        return;
+    }
+    if(enumValue==NULL) {
+        enumValue=enumSameValue;
+    }
+
+    idx=trie->index;
+    data32=trie->data32;
+
+    /* get the enumeration value that corresponds to an initial-value trie data entry */
+    initialValue=enumValue(context, trie->initialValue);
+
+    if(data32==NULL) {
+        nullBlock=trie->indexLength;
+    } else {
+        nullBlock=0;
+    }
+
+    /* set variables for previous range */
+    prevBlock=nullBlock;
+    prev=0;
+    prevValue=initialValue;
+
+    /* enumerate BMP - the main loop enumerates data blocks */
+    for(i=0, c=0; c<=0xffff; ++i) {
+        if(c==0xd800) {
+            /* skip lead surrogate code _units_, go to lead surr. code _points_ */
+            i=UTRIE_BMP_INDEX_LENGTH;
+        } else if(c==0xdc00) {
+            /* go back to regular BMP code points */
+            i=c>>UTRIE_SHIFT;
+        }
+
+        block=idx[i]<<UTRIE_INDEX_SHIFT;
+        if(block==prevBlock) {
+            /* the block is the same as the previous one, and filled with value */
+            c+=UTRIE_DATA_BLOCK_LENGTH;
+        } else if(block==nullBlock) {
+            /* this is the all-initial-value block */
+            if(prevValue!=initialValue) {
+                if(prev<c) {
+                    if(!enumRange(context, prev, c, prevValue)) {
+                        return;
+                    }
+                }
+                prevBlock=nullBlock;
+                prev=c;
+                prevValue=initialValue;
+            }
+            c+=UTRIE_DATA_BLOCK_LENGTH;
+        } else {
+            prevBlock=block;
+            for(j=0; j<UTRIE_DATA_BLOCK_LENGTH; ++j) {
+                value=enumValue(context, data32!=NULL ? data32[block+j] : idx[block+j]);
+                if(value!=prevValue) {
+                    if(prev<c) {
+                        if(!enumRange(context, prev, c, prevValue)) {
+                            return;
+                        }
+                    }
+                    if(j>0) {
+                        /* the block is not filled with all the same value */
+                        prevBlock=-1;
+                    }
+                    prev=c;
+                    prevValue=value;
+                }
+                ++c;
+            }
+        }
+    }
+
+    /* enumerate supplementary code points */
+    for(l=0xd800; l<0xdc00;) {
+        /* lead surrogate access */
+        offset=idx[l>>UTRIE_SHIFT]<<UTRIE_INDEX_SHIFT;
+        if(offset==nullBlock) {
+            /* no entries for a whole block of lead surrogates */
+            if(prevValue!=initialValue) {
+                if(prev<c) {
+                    if(!enumRange(context, prev, c, prevValue)) {
+                        return;
+                    }
+                }
+                prevBlock=nullBlock;
+                prev=c;
+                prevValue=initialValue;
+            }
+
+            l+=UTRIE_DATA_BLOCK_LENGTH;
+            c+=UTRIE_DATA_BLOCK_LENGTH<<10;
+            continue;
+        }
+
+        value= data32!=NULL ? data32[offset+(l&UTRIE_MASK)] : idx[offset+(l&UTRIE_MASK)];
+
+        /* enumerate trail surrogates for this lead surrogate */
+        offset=trie->getFoldingOffset(value);
+        if(offset<=0) {
+            /* no data for this lead surrogate */
+            if(prevValue!=initialValue) {
+                if(prev<c) {
+                    if(!enumRange(context, prev, c, prevValue)) {
+                        return;
+                    }
+                }
+                prevBlock=nullBlock;
+                prev=c;
+                prevValue=initialValue;
+            }
+
+            /* nothing else to do for the supplementary code points for this lead surrogate */
+            c+=0x400;
+        } else {
+            /* enumerate code points for this lead surrogate */
+            i=offset;
+            offset+=UTRIE_SURROGATE_BLOCK_COUNT;
+            do {
+                /* copy of most of the body of the BMP loop */
+                block=idx[i]<<UTRIE_INDEX_SHIFT;
+                if(block==prevBlock) {
+                    /* the block is the same as the previous one, and filled with value */
+                    c+=UTRIE_DATA_BLOCK_LENGTH;
+                } else if(block==nullBlock) {
+                    /* this is the all-initial-value block */
+                    if(prevValue!=initialValue) {
+                        if(prev<c) {
+                            if(!enumRange(context, prev, c, prevValue)) {
+                                return;
+                            }
+                        }
+                        prevBlock=nullBlock;
+                        prev=c;
+                        prevValue=initialValue;
+                    }
+                    c+=UTRIE_DATA_BLOCK_LENGTH;
+                } else {
+                    prevBlock=block;
+                    for(j=0; j<UTRIE_DATA_BLOCK_LENGTH; ++j) {
+                        value=enumValue(context, data32!=NULL ? data32[block+j] : idx[block+j]);
+                        if(value!=prevValue) {
+                            if(prev<c) {
+                                if(!enumRange(context, prev, c, prevValue)) {
+                                    return;
+                                }
+                            }
+                            if(j>0) {
+                                /* the block is not filled with all the same value */
+                                prevBlock=-1;
+                            }
+                            prev=c;
+                            prevValue=value;
+                        }
+                        ++c;
+                    }
+                }
+            } while(++i<offset);
+        }
+
+        ++l;
+    }
+
+    /* deliver last range */
+    enumRange(context, prev, c, prevValue);
+}
diff --git a/thirdparty/icu4c/common/utrie.h b/thirdparty/icu4c/common/utrie.h
new file mode 100644
index 0000000000..2fd2c461ff
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie.h
@@ -0,0 +1,793 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2001nov08
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UTRIE_H__
+#define __UTRIE_H__
+
+#include "unicode/utypes.h"
+#include "unicode/utf16.h"
+
+U_CDECL_BEGIN
+
+/**
+ * \file
+ *
+ * This is a common implementation of a "folded" trie.
+ * It is a kind of compressed, serializable table of 16- or 32-bit values associated with
+ * Unicode code points (0..0x10ffff).
+ *
+ * This implementation is optimized for getting values while walking forward
+ * through a UTF-16 string.
+ * Therefore, the simplest and fastest access macros are the
+ * _FROM_LEAD() and _FROM_OFFSET_TRAIL() macros.
+ *
+ * The _FROM_BMP() macros are a little more complicated; they get values
+ * even for lead surrogate code _points_, while the _FROM_LEAD() macros
+ * get special "folded" values for lead surrogate code _units_ if
+ * there is relevant data associated with them.
+ * From such a folded value, an offset needs to be extracted to supply
+ * to the _FROM_OFFSET_TRAIL() macros.
+ *
+ * Most of the more complex (and more convenient) functions/macros call a callback function
+ * to get that offset from the folded value for a lead surrogate unit.
+ */
+
+/**
+ * Trie constants, defining shift widths, index array lengths, etc.
+ */
+enum {
+    /** Shift size for shifting right the input index. 1..9 */
+    UTRIE_SHIFT=5,
+
+    /** Number of data values in a stage 2 (data array) block. 2, 4, 8, .., 0x200 */
+    UTRIE_DATA_BLOCK_LENGTH=1<<UTRIE_SHIFT,
+
+    /** Mask for getting the lower bits from the input index. */
+    UTRIE_MASK=UTRIE_DATA_BLOCK_LENGTH-1,
+
+    /**
+     * Lead surrogate code points' index displacement in the index array.
+     * 0x10000-0xd800=0x2800
+     */
+    UTRIE_LEAD_INDEX_DISP=0x2800>>UTRIE_SHIFT,
+
+    /**
+     * Shift size for shifting left the index array values.
+     * Increases possible data size with 16-bit index values at the cost
+     * of compactability.
+     * This requires blocks of stage 2 data to be aligned by UTRIE_DATA_GRANULARITY.
+     * 0..UTRIE_SHIFT
+     */
+    UTRIE_INDEX_SHIFT=2,
+
+    /** The alignment size of a stage 2 data block. Also the granularity for compaction. */
+    UTRIE_DATA_GRANULARITY=1<<UTRIE_INDEX_SHIFT,
+
+    /** Number of bits of a trail surrogate that are used in index table lookups. */
+    UTRIE_SURROGATE_BLOCK_BITS=10-UTRIE_SHIFT,
+
+    /**
+     * Number of index (stage 1) entries per lead surrogate.
+     * Same as number of index entries for 1024 trail surrogates,
+     * ==0x400>>UTRIE_SHIFT
+     */
+    UTRIE_SURROGATE_BLOCK_COUNT=(1<<UTRIE_SURROGATE_BLOCK_BITS),
+
+    /** Length of the BMP portion of the index (stage 1) array. */
+    UTRIE_BMP_INDEX_LENGTH=0x10000>>UTRIE_SHIFT
+};
+
+/**
+ * Length of the index (stage 1) array before folding.
+ * Maximum number of Unicode code points (0x110000) shifted right by UTRIE_SHIFT.
+ */
+#define UTRIE_MAX_INDEX_LENGTH (0x110000>>UTRIE_SHIFT)
+
+/**
+ * Maximum length of the runtime data (stage 2) array.
+ * Limited by 16-bit index values that are left-shifted by UTRIE_INDEX_SHIFT.
+ */
+#define UTRIE_MAX_DATA_LENGTH (0x10000<<UTRIE_INDEX_SHIFT)
+
+/**
+ * Maximum length of the build-time data (stage 2) array.
+ * The maximum length is 0x110000+UTRIE_DATA_BLOCK_LENGTH+0x400.
+ * (Number of Unicode code points + one all-initial-value block +
+ *  possible duplicate entries for 1024 lead surrogates.)
+ */
+#define UTRIE_MAX_BUILD_TIME_DATA_LENGTH (0x110000+UTRIE_DATA_BLOCK_LENGTH+0x400)
+
+/**
+ * Number of bytes for a dummy trie.
+ * A dummy trie is an empty runtime trie, used when a real data trie cannot
+ * be loaded.
+ * The number of bytes works for Latin-1-linear tries with 32-bit data
+ * (worst case).
+ *
+ * Calculation:
+ *   BMP index + 1 index block for lead surrogate code points +
+ *   Latin-1-linear array + 1 data block for lead surrogate code points
+ *
+ * Latin-1: if(UTRIE_SHIFT<=8) { 256 } else { included in first data block }
+ *
+ * @see utrie_unserializeDummy
+ */
+#define UTRIE_DUMMY_SIZE ((UTRIE_BMP_INDEX_LENGTH+UTRIE_SURROGATE_BLOCK_COUNT)*2+(UTRIE_SHIFT<=8?256:UTRIE_DATA_BLOCK_LENGTH)*4+UTRIE_DATA_BLOCK_LENGTH*4)
+
+/**
+ * Runtime UTrie callback function.
+ * Extract from a lead surrogate's data the
+ * index array offset of the indexes for that lead surrogate.
+ *
+ * @param data data value for a surrogate from the trie, including the folding offset
+ * @return offset>=UTRIE_BMP_INDEX_LENGTH, or 0 if there is no data for the lead surrogate
+ */
+typedef int32_t U_CALLCONV
+UTrieGetFoldingOffset(uint32_t data);
+
+/**
+ * Run-time Trie structure.
+ *
+ * Either the data table is 16 bits wide and accessed via the index
+ * pointer, with each index item increased by indexLength;
+ * in this case, data32==NULL.
+ *
+ * Or the data table is 32 bits wide and accessed via the data32 pointer.
+ */
+struct UTrie {
+    const uint16_t *index;
+    const uint32_t *data32; /* NULL if 16b data is used via index */
+
+    /**
+     * This function is not used in _FROM_LEAD, _FROM_BMP, and _FROM_OFFSET_TRAIL macros.
+     * If convenience macros like _GET16 or _NEXT32 are used, this function must be set.
+     *
+     * utrie_unserialize() sets a default function which simply returns
+     * the lead surrogate's value itself - which is the inverse of the default
+     * folding function used by utrie_serialize().
+     *
+     * @see UTrieGetFoldingOffset
+     */
+    UTrieGetFoldingOffset *getFoldingOffset;
+
+    int32_t indexLength, dataLength;
+    uint32_t initialValue;
+    UBool isLatin1Linear;
+};
+
+#ifndef __UTRIE2_H__
+typedef struct UTrie UTrie;
+#endif
+
+/** Internal trie getter from an offset (0 if c16 is a BMP/lead units) and a 16-bit unit */
+#define _UTRIE_GET_RAW(trie, data, offset, c16) \
+    (trie)->data[ \
+        ((int32_t)((trie)->index[(offset)+((c16)>>UTRIE_SHIFT)])<<UTRIE_INDEX_SHIFT)+ \
+        ((c16)&UTRIE_MASK) \
+    ]
+
+/** Internal trie getter from a pair of surrogates */
+#define _UTRIE_GET_FROM_PAIR(trie, data, c, c2, result, resultType) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __offset; \
+\
+    /* get data for lead surrogate */ \
+    (result)=_UTRIE_GET_RAW((trie), data, 0, (c)); \
+    __offset=(trie)->getFoldingOffset(result); \
+\
+    /* get the real data from the folded lead/trail units */ \
+    if(__offset>0) { \
+        (result)=_UTRIE_GET_RAW((trie), data, __offset, (c2)&0x3ff); \
+    } else { \
+        (result)=(resultType)((trie)->initialValue); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal trie getter from a BMP code point, treating a lead surrogate as a normal code point */
+#define _UTRIE_GET_FROM_BMP(trie, data, c16) \
+    _UTRIE_GET_RAW(trie, data, 0xd800<=(c16) && (c16)<=0xdbff ? UTRIE_LEAD_INDEX_DISP : 0, c16)
+
+/**
+ * Internal trie getter from a code point.
+ * Could be faster(?) but longer with
+ *   if((c32)<=0xd7ff) { (result)=_UTRIE_GET_RAW(trie, data, 0, c32); }
+ */
+#define _UTRIE_GET(trie, data, c32, result, resultType) UPRV_BLOCK_MACRO_BEGIN { \
+    if((uint32_t)(c32)<=0xffff) { \
+        /* BMP code points */ \
+        (result)=_UTRIE_GET_FROM_BMP(trie, data, c32); \
+    } else if((uint32_t)(c32)<=0x10ffff) { \
+        /* supplementary code point */ \
+        UChar __lead16=U16_LEAD(c32); \
+        _UTRIE_GET_FROM_PAIR(trie, data, __lead16, c32, result, resultType); \
+    } else { \
+        /* out of range */ \
+        (result)=(resultType)((trie)->initialValue); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal next-post-increment: get the next code point (c, c2) and its data */
+#define _UTRIE_NEXT(trie, data, src, limit, c, c2, result, resultType) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=*(src)++; \
+    if(!U16_IS_LEAD(c)) { \
+        (c2)=0; \
+        (result)=_UTRIE_GET_RAW((trie), data, 0, (c)); \
+    } else if((src)!=(limit) && U16_IS_TRAIL((c2)=*(src))) { \
+        ++(src); \
+        _UTRIE_GET_FROM_PAIR((trie), data, (c), (c2), (result), resultType); \
+    } else { \
+        /* unpaired lead surrogate code point */ \
+        (c2)=0; \
+        (result)=_UTRIE_GET_RAW((trie), data, UTRIE_LEAD_INDEX_DISP, (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal previous: get the previous code point (c, c2) and its data */
+#define _UTRIE_PREVIOUS(trie, data, start, src, c, c2, result, resultType) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=*--(src); \
+    if(!U16_IS_SURROGATE(c)) { \
+        (c2)=0; \
+        (result)=_UTRIE_GET_RAW((trie), data, 0, (c)); \
+    } else if(!U16_IS_SURROGATE_LEAD(c)) { \
+        /* trail surrogate */ \
+        if((start)!=(src) && U16_IS_LEAD((c2)=*((src)-1))) { \
+            --(src); \
+            (result)=(c); (c)=(c2); (c2)=(UChar)(result); /* swap c, c2 */ \
+            _UTRIE_GET_FROM_PAIR((trie), data, (c), (c2), (result), resultType); \
+        } else { \
+            /* unpaired trail surrogate code point */ \
+            (c2)=0; \
+            (result)=_UTRIE_GET_RAW((trie), data, 0, (c)); \
+        } \
+    } else { \
+        /* unpaired lead surrogate code point */ \
+        (c2)=0; \
+        (result)=_UTRIE_GET_RAW((trie), data, UTRIE_LEAD_INDEX_DISP, (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* Public UTrie API ---------------------------------------------------------*/
+
+/**
+ * Get a pointer to the contiguous part of the data array
+ * for the Latin-1 range (U+0000..U+00ff).
+ * Must be used only if the Latin-1 range is in fact linear
+ * (trie->isLatin1Linear).
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @return (const uint16_t *) pointer to values for Latin-1 code points
+ */
+#define UTRIE_GET16_LATIN1(trie) ((trie)->index+(trie)->indexLength+UTRIE_DATA_BLOCK_LENGTH)
+
+/**
+ * Get a pointer to the contiguous part of the data array
+ * for the Latin-1 range (U+0000..U+00ff).
+ * Must be used only if the Latin-1 range is in fact linear
+ * (trie->isLatin1Linear).
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @return (const uint32_t *) pointer to values for Latin-1 code points
+ */
+#define UTRIE_GET32_LATIN1(trie) ((trie)->data32+UTRIE_DATA_BLOCK_LENGTH)
+
+/**
+ * Get a 16-bit trie value from a BMP code point (UChar, <=U+ffff).
+ * c16 may be a lead surrogate, which may have a value including a folding offset.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c16 (UChar, in) the input BMP code point
+ * @return (uint16_t) trie lookup result
+ */
+#define UTRIE_GET16_FROM_LEAD(trie, c16) _UTRIE_GET_RAW(trie, index, 0, c16)
+
+/**
+ * Get a 32-bit trie value from a BMP code point (UChar, <=U+ffff).
+ * c16 may be a lead surrogate, which may have a value including a folding offset.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c16 (UChar, in) the input BMP code point
+ * @return (uint32_t) trie lookup result
+ */
+#define UTRIE_GET32_FROM_LEAD(trie, c16) _UTRIE_GET_RAW(trie, data32, 0, c16)
+
+/**
+ * Get a 16-bit trie value from a BMP code point (UChar, <=U+ffff).
+ * Even lead surrogate code points are treated as normal code points,
+ * with unfolded values that may differ from _FROM_LEAD() macro results for them.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c16 (UChar, in) the input BMP code point
+ * @return (uint16_t) trie lookup result
+ */
+#define UTRIE_GET16_FROM_BMP(trie, c16) _UTRIE_GET_FROM_BMP(trie, index, c16)
+
+/**
+ * Get a 32-bit trie value from a BMP code point (UChar, <=U+ffff).
+ * Even lead surrogate code points are treated as normal code points,
+ * with unfolded values that may differ from _FROM_LEAD() macro results for them.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c16 (UChar, in) the input BMP code point
+ * @return (uint32_t) trie lookup result
+ */
+#define UTRIE_GET32_FROM_BMP(trie, c16) _UTRIE_GET_FROM_BMP(trie, data32, c16)
+
+/**
+ * Get a 16-bit trie value from a code point.
+ * Even lead surrogate code points are treated as normal code points,
+ * with unfolded values that may differ from _FROM_LEAD() macro results for them.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c32 (UChar32, in) the input code point
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE_GET16(trie, c32, result) _UTRIE_GET(trie, index, c32, result, uint16_t)
+
+/**
+ * Get a 32-bit trie value from a code point.
+ * Even lead surrogate code points are treated as normal code points,
+ * with unfolded values that may differ from _FROM_LEAD() macro results for them.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c32 (UChar32, in) the input code point
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE_GET32(trie, c32, result) _UTRIE_GET(trie, data32, c32, result, uint32_t)
+
+/**
+ * Get the next code point (c, c2), post-increment src,
+ * and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param limit (const UChar *, in) the limit pointer for the text, or NULL
+ * @param c (UChar, out) variable for the BMP or lead code unit
+ * @param c2 (UChar, out) variable for 0 or the trail code unit
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE_NEXT16(trie, src, limit, c, c2, result) _UTRIE_NEXT(trie, index, src, limit, c, c2, result, uint16_t)
+
+/**
+ * Get the next code point (c, c2), post-increment src,
+ * and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param limit (const UChar *, in) the limit pointer for the text, or NULL
+ * @param c (UChar, out) variable for the BMP or lead code unit
+ * @param c2 (UChar, out) variable for 0 or the trail code unit
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE_NEXT32(trie, src, limit, c, c2, result) _UTRIE_NEXT(trie, data32, src, limit, c, c2, result, uint32_t)
+
+/**
+ * Get the previous code point (c, c2), pre-decrement src,
+ * and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param start (const UChar *, in) the start pointer for the text, or NULL
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param c (UChar, out) variable for the BMP or lead code unit
+ * @param c2 (UChar, out) variable for 0 or the trail code unit
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE_PREVIOUS16(trie, start, src, c, c2, result) _UTRIE_PREVIOUS(trie, index, start, src, c, c2, result, uint16_t)
+
+/**
+ * Get the previous code point (c, c2), pre-decrement src,
+ * and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param start (const UChar *, in) the start pointer for the text, or NULL
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param c (UChar, out) variable for the BMP or lead code unit
+ * @param c2 (UChar, out) variable for 0 or the trail code unit
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE_PREVIOUS32(trie, start, src, c, c2, result) _UTRIE_PREVIOUS(trie, data32, start, src, c, c2, result, uint32_t)
+
+/**
+ * Get a 16-bit trie value from a pair of surrogates.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c (UChar, in) a lead surrogate
+ * @param c2 (UChar, in) a trail surrogate
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE_GET16_FROM_PAIR(trie, c, c2, result) _UTRIE_GET_FROM_PAIR(trie, index, c, c2, result, uint16_t)
+
+/**
+ * Get a 32-bit trie value from a pair of surrogates.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param c (UChar, in) a lead surrogate
+ * @param c2 (UChar, in) a trail surrogate
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE_GET32_FROM_PAIR(trie, c, c2, result) _UTRIE_GET_FROM_PAIR(trie, data32, c, c2, result, uint32_t)
+
+/**
+ * Get a 16-bit trie value from a folding offset (from the value of a lead surrogate)
+ * and a trail surrogate.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param offset (int32_t, in) the folding offset from the value of a lead surrogate
+ * @param c2 (UChar, in) a trail surrogate (only the 10 low bits are significant)
+ * @return (uint16_t) trie lookup result
+ */
+#define UTRIE_GET16_FROM_OFFSET_TRAIL(trie, offset, c2) _UTRIE_GET_RAW(trie, index, offset, (c2)&0x3ff)
+
+/**
+ * Get a 32-bit trie value from a folding offset (from the value of a lead surrogate)
+ * and a trail surrogate.
+ *
+ * @param trie (const UTrie *, in) a pointer to the runtime trie structure
+ * @param offset (int32_t, in) the folding offset from the value of a lead surrogate
+ * @param c2 (UChar, in) a trail surrogate (only the 10 low bits are significant)
+ * @return (uint32_t) trie lookup result
+ */
+#define UTRIE_GET32_FROM_OFFSET_TRAIL(trie, offset, c2) _UTRIE_GET_RAW(trie, data32, offset, (c2)&0x3ff)
+
+/* enumeration callback types */
+
+/**
+ * Callback from utrie_enum(), extracts a uint32_t value from a
+ * trie value. This value will be passed on to the UTrieEnumRange function.
+ *
+ * @param context an opaque pointer, as passed into utrie_enum()
+ * @param value a value from the trie
+ * @return the value that is to be passed on to the UTrieEnumRange function
+ */
+typedef uint32_t U_CALLCONV
+UTrieEnumValue(const void *context, uint32_t value);
+
+/**
+ * Callback from utrie_enum(), is called for each contiguous range
+ * of code points with the same value as retrieved from the trie and
+ * transformed by the UTrieEnumValue function.
+ *
+ * The callback function can stop the enumeration by returning false.
+ *
+ * @param context an opaque pointer, as passed into utrie_enum()
+ * @param start the first code point in a contiguous range with value
+ * @param limit one past the last code point in a contiguous range with value
+ * @param value the value that is set for all code points in [start..limit[
+ * @return false to stop the enumeration
+ */
+typedef UBool U_CALLCONV
+UTrieEnumRange(const void *context, UChar32 start, UChar32 limit, uint32_t value);
+
+/**
+ * Enumerate efficiently all values in a trie.
+ * For each entry in the trie, the value to be delivered is passed through
+ * the UTrieEnumValue function.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * For each contiguous range of code points with a given value,
+ * the UTrieEnumRange function is called.
+ *
+ * @param trie a pointer to the runtime trie structure
+ * @param enumValue a pointer to a function that may transform the trie entry value,
+ *                  or NULL if the values from the trie are to be used directly
+ * @param enumRange a pointer to a function that is called for each contiguous range
+ *                  of code points with the same value
+ * @param context an opaque pointer that is passed on to the callback functions
+ */
+U_CAPI void U_EXPORT2
+utrie_enum(const UTrie *trie,
+           UTrieEnumValue *enumValue, UTrieEnumRange *enumRange, const void *context);
+
+/**
+ * Unserialize a trie from 32-bit-aligned memory.
+ * Inverse of utrie_serialize().
+ * Fills the UTrie runtime trie structure with the settings for the trie data.
+ *
+ * @param trie a pointer to the runtime trie structure
+ * @param data a pointer to 32-bit-aligned memory containing trie data
+ * @param length the number of bytes available at data
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the number of bytes at data taken up by the trie data
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_unserialize(UTrie *trie, const void *data, int32_t length, UErrorCode *pErrorCode);
+
+/**
+ * "Unserialize" a dummy trie.
+ * A dummy trie is an empty runtime trie, used when a real data trie cannot
+ * be loaded.
+ *
+ * The input memory is filled so that the trie always returns the initialValue,
+ * or the leadUnitValue for lead surrogate code points.
+ * The Latin-1 part is always set up to be linear.
+ *
+ * @param trie a pointer to the runtime trie structure
+ * @param data a pointer to 32-bit-aligned memory to be filled with the dummy trie data
+ * @param length the number of bytes available at data (recommended to use UTRIE_DUMMY_SIZE)
+ * @param initialValue the initial value that is set for all code points
+ * @param leadUnitValue the value for lead surrogate code _units_ that do not
+ *                      have associated supplementary data
+ * @param pErrorCode an in/out ICU UErrorCode
+ *
+ * @see UTRIE_DUMMY_SIZE
+ * @see utrie_open
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_unserializeDummy(UTrie *trie,
+                       void *data, int32_t length,
+                       uint32_t initialValue, uint32_t leadUnitValue,
+                       UBool make16BitTrie,
+                       UErrorCode *pErrorCode);
+
+/**
+ * Default implementation for UTrie.getFoldingOffset, set automatically by
+ * utrie_unserialize().
+ * Simply returns the lead surrogate's value itself - which is the inverse
+ * of the default folding function used by utrie_serialize().
+ * Exported for static const UTrie structures.
+ *
+ * @see UTrieGetFoldingOffset
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_defaultGetFoldingOffset(uint32_t data);
+
+/* Building a trie ----------------------------------------------------------*/
+
+/**
+ * Build-time trie structure.
+ * Opaque definition, here only to make fillIn parameters possible
+ * for utrie_open() and utrie_clone().
+ */
+struct UNewTrie {
+    /**
+     * Index values at build-time are 32 bits wide for easier processing.
+     * Bit 31 is set if the data block is used by multiple index values (from utrie_setRange()).
+     */
+    int32_t index[UTRIE_MAX_INDEX_LENGTH+UTRIE_SURROGATE_BLOCK_COUNT];
+    uint32_t *data;
+
+    uint32_t leadUnitValue;
+    int32_t indexLength, dataCapacity, dataLength;
+    UBool isAllocated, isDataAllocated;
+    UBool isLatin1Linear, isCompacted;
+
+    /**
+     * Map of adjusted indexes, used in utrie_compact().
+     * Maps from original indexes to new ones.
+     */
+    int32_t map[UTRIE_MAX_BUILD_TIME_DATA_LENGTH>>UTRIE_SHIFT];
+};
+
+typedef struct UNewTrie UNewTrie;
+
+/**
+ * Build-time trie callback function, used with utrie_serialize().
+ * This function calculates a lead surrogate's value including a folding offset
+ * from the 1024 supplementary code points [start..start+1024[ .
+ * It is U+10000 <= start <= U+10fc00 and (start&0x3ff)==0.
+ *
+ * The folding offset is provided by the caller.
+ * It is offset=UTRIE_BMP_INDEX_LENGTH+n*UTRIE_SURROGATE_BLOCK_COUNT with n=0..1023.
+ * Instead of the offset itself, n can be stored in 10 bits -
+ * or fewer if it can be assumed that few lead surrogates have associated data.
+ *
+ * The returned value must be
+ * - not zero if and only if there is relevant data
+ *   for the corresponding 1024 supplementary code points
+ * - such that UTrie.getFoldingOffset(UNewTrieGetFoldedValue(..., offset))==offset
+ *
+ * @return a folded value, or 0 if there is no relevant data for the lead surrogate.
+ */
+typedef uint32_t U_CALLCONV
+UNewTrieGetFoldedValue(UNewTrie *trie, UChar32 start, int32_t offset);
+
+/**
+ * Open a build-time trie structure.
+ * The size of the build-time data array is specified to avoid allocating a large
+ * array in all cases. The array itself can also be passed in.
+ *
+ * Although the trie is never fully expanded to a linear array, especially when
+ * utrie_setRange32() is used, the data array could be large during build time.
+ * The maximum length is
+ * UTRIE_MAX_BUILD_TIME_DATA_LENGTH=0x110000+UTRIE_DATA_BLOCK_LENGTH+0x400.
+ * (Number of Unicode code points + one all-initial-value block +
+ *  possible duplicate entries for 1024 lead surrogates.)
+ * (UTRIE_DATA_BLOCK_LENGTH<=0x200 in all cases.)
+ *
+ * @param fillIn a pointer to a UNewTrie structure to be initialized (will not be released), or
+ *               NULL if one is to be allocated
+ * @param aliasData a pointer to a data array to be used (will not be released), or
+ *                  NULL if one is to be allocated
+ * @param maxDataLength the capacity of aliasData (if not NULL) or
+ *                      the length of the data array to be allocated
+ * @param initialValue the initial value that is set for all code points
+ * @param leadUnitValue the value for lead surrogate code _units_ that do not
+ *                      have associated supplementary data
+ * @param latin1Linear a flag indicating whether the Latin-1 range is to be allocated and
+ *                     kept in a linear, contiguous part of the data array
+ * @return a pointer to the initialized fillIn or the allocated and initialized new UNewTrie
+ */
+U_CAPI UNewTrie * U_EXPORT2
+utrie_open(UNewTrie *fillIn,
+           uint32_t *aliasData, int32_t maxDataLength,
+           uint32_t initialValue, uint32_t leadUnitValue,
+           UBool latin1Linear);
+
+/**
+ * Clone a build-time trie structure with all entries.
+ *
+ * @param fillIn like in utrie_open()
+ * @param other the build-time trie structure to clone
+ * @param aliasData like in utrie_open(),
+ *                  used if aliasDataLength>=(capacity of other's data array)
+ * @param aliasDataLength the length of aliasData
+ * @return a pointer to the initialized fillIn or the allocated and initialized new UNewTrie
+ */
+U_CAPI UNewTrie * U_EXPORT2
+utrie_clone(UNewTrie *fillIn, const UNewTrie *other, uint32_t *aliasData, int32_t aliasDataLength);
+
+/**
+ * Close a build-time trie structure, and release memory
+ * that was allocated by utrie_open() or utrie_clone().
+ *
+ * @param trie the build-time trie
+ */
+U_CAPI void U_EXPORT2
+utrie_close(UNewTrie *trie);
+
+/**
+ * Get the data array of a build-time trie.
+ * The data may be modified, but entries that are equal before
+ * must still be equal after modification.
+ *
+ * @param trie the build-time trie
+ * @param pLength (out) a pointer to a variable that receives the number
+ *                of entries in the data array
+ * @return the data array
+ */
+U_CAPI uint32_t * U_EXPORT2
+utrie_getData(UNewTrie *trie, int32_t *pLength);
+
+/**
+ * Set a value for a code point.
+ *
+ * @param trie the build-time trie
+ * @param c the code point
+ * @param value the value
+ * @return false if a failure occurred (illegal argument or data array overrun)
+ */
+U_CAPI UBool U_EXPORT2
+utrie_set32(UNewTrie *trie, UChar32 c, uint32_t value);
+
+/**
+ * Get a value from a code point as stored in the build-time trie.
+ *
+ * @param trie the build-time trie
+ * @param c the code point
+ * @param pInBlockZero if not NULL, then *pInBlockZero is set to true
+ *                     iff the value is retrieved from block 0;
+ *                     block 0 is the all-initial-value initial block
+ * @return the value
+ */
+U_CAPI uint32_t U_EXPORT2
+utrie_get32(UNewTrie *trie, UChar32 c, UBool *pInBlockZero);
+
+/**
+ * Set a value in a range of code points [start..limit[.
+ * All code points c with start<=c<limit will get the value if
+ * overwrite is true or if the old value is 0.
+ *
+ * @param trie the build-time trie
+ * @param start the first code point to get the value
+ * @param limit one past the last code point to get the value
+ * @param value the value
+ * @param overwrite flag for whether old non-initial values are to be overwritten
+ * @return false if a failure occurred (illegal argument or data array overrun)
+ */
+U_CAPI UBool U_EXPORT2
+utrie_setRange32(UNewTrie *trie, UChar32 start, UChar32 limit, uint32_t value, UBool overwrite);
+
+/**
+ * Compact the build-time trie after all values are set, and then
+ * serialize it into 32-bit aligned memory.
+ *
+ * After this, the trie can only be serizalized again and/or closed;
+ * no further values can be added.
+ *
+ * @see utrie_unserialize()
+ *
+ * @param trie the build-time trie
+ * @param data a pointer to 32-bit-aligned memory for the trie data
+ * @param capacity the number of bytes available at data
+ * @param getFoldedValue a callback function that calculates the value for
+ *                       a lead surrogate from all of its supplementary code points
+ *                       and the folding offset;
+ *                       if NULL, then a default function is used which returns just
+ *                       the input offset when there are any non-initial-value entries
+ * @param reduceTo16Bits flag for whether the values are to be reduced to a
+ *                       width of 16 bits for serialization and runtime
+ * @param pErrorCode a UErrorCode argument; among other possible error codes:
+ * - U_BUFFER_OVERFLOW_ERROR if the data storage block is too small for serialization
+ * - U_MEMORY_ALLOCATION_ERROR if the trie data array is too small
+ * - U_INDEX_OUTOFBOUNDS_ERROR if the index or data arrays are too long after compaction for serialization
+ *
+ * @return the number of bytes written for the trie
+ */
+U_CAPI int32_t U_EXPORT2
+utrie_serialize(UNewTrie *trie, void *data, int32_t capacity,
+                UNewTrieGetFoldedValue *getFoldedValue,
+                UBool reduceTo16Bits,
+                UErrorCode *pErrorCode);
+
+/* serialization ------------------------------------------------------------ */
+
+// UTrie signature values, in platform endianness and opposite endianness.
+// The UTrie signature ASCII byte values spell "Trie".
+#define UTRIE_SIG       0x54726965
+#define UTRIE_OE_SIG    0x65697254
+
+/**
+ * Trie data structure in serialized form:
+ *
+ * UTrieHeader header;
+ * uint16_t index[header.indexLength];
+ * uint16_t data[header.dataLength];
+ * @internal
+ */
+typedef struct UTrieHeader {
+    /** "Trie" in big-endian US-ASCII (0x54726965) */
+    uint32_t signature;
+
+    /**
+     * options bit field:
+     *     9    1=Latin-1 data is stored linearly at data+UTRIE_DATA_BLOCK_LENGTH
+     *     8    0=16-bit data, 1=32-bit data
+     *  7..4    UTRIE_INDEX_SHIFT   // 0..UTRIE_SHIFT
+     *  3..0    UTRIE_SHIFT         // 1..9
+     */
+    uint32_t options;
+
+    /** indexLength is a multiple of UTRIE_SURROGATE_BLOCK_COUNT */
+    int32_t indexLength;
+
+    /** dataLength>=UTRIE_DATA_BLOCK_LENGTH */
+    int32_t dataLength;
+} UTrieHeader;
+
+/**
+ * Constants for use with UTrieHeader.options.
+ * @internal
+ */
+enum {
+    /** Mask to get the UTRIE_SHIFT value from options. */
+    UTRIE_OPTIONS_SHIFT_MASK=0xf,
+
+    /** Shift options right this much to get the UTRIE_INDEX_SHIFT value. */
+    UTRIE_OPTIONS_INDEX_SHIFT=4,
+
+    /** If set, then the data (stage 2) array is 32 bits wide. */
+    UTRIE_OPTIONS_DATA_IS_32_BIT=0x100,
+
+    /**
+     * If set, then Latin-1 data (for U+0000..U+00ff) is stored in the data (stage 2) array
+     * as a simple, linear array at data+UTRIE_DATA_BLOCK_LENGTH.
+     */
+    UTRIE_OPTIONS_LATIN1_IS_LINEAR=0x200
+};
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/utrie2.cpp b/thirdparty/icu4c/common/utrie2.cpp
new file mode 100644
index 0000000000..24ef5782c9
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie2.cpp
@@ -0,0 +1,663 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie2.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2008aug16 (starting from a copy of utrie.c)
+*   created by: Markus W. Scherer
+*
+*   This is a common implementation of a Unicode trie.
+*   It is a kind of compressed, serializable table of 16- or 32-bit values associated with
+*   Unicode code points (0..0x10ffff).
+*   This is the second common version of a Unicode trie (hence the name UTrie2).
+*   See utrie2.h for a comparison.
+*
+*   This file contains only the runtime and enumeration code, for read-only access.
+*   See utrie2_builder.c for the builder code.
+*/
+#include "unicode/utypes.h"
+#ifdef UCPTRIE_DEBUG
+#include "unicode/umutablecptrie.h"
+#endif
+#include "unicode/utf.h"
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "utrie2.h"
+#include "utrie2_impl.h"
+#include "uassert.h"
+
+/* Public UTrie2 API implementation ----------------------------------------- */
+
+static uint32_t
+get32(const UNewTrie2 *trie, UChar32 c, UBool fromLSCP) {
+    int32_t i2, block;
+
+    if(c>=trie->highStart && (!U_IS_LEAD(c) || fromLSCP)) {
+        return trie->data[trie->dataLength-UTRIE2_DATA_GRANULARITY];
+    }
+
+    if(U_IS_LEAD(c) && fromLSCP) {
+        i2=(UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2))+
+            (c>>UTRIE2_SHIFT_2);
+    } else {
+        i2=trie->index1[c>>UTRIE2_SHIFT_1]+
+            ((c>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK);
+    }
+    block=trie->index2[i2];
+    return trie->data[block+(c&UTRIE2_DATA_MASK)];
+}
+
+U_CAPI uint32_t U_EXPORT2
+utrie2_get32(const UTrie2 *trie, UChar32 c) {
+    if(trie->data16!=NULL) {
+        return UTRIE2_GET16(trie, c);
+    } else if(trie->data32!=NULL) {
+        return UTRIE2_GET32(trie, c);
+    } else if((uint32_t)c>0x10ffff) {
+        return trie->errorValue;
+    } else {
+        return get32(trie->newTrie, c, TRUE);
+    }
+}
+
+U_CAPI uint32_t U_EXPORT2
+utrie2_get32FromLeadSurrogateCodeUnit(const UTrie2 *trie, UChar32 c) {
+    if(!U_IS_LEAD(c)) {
+        return trie->errorValue;
+    }
+    if(trie->data16!=NULL) {
+        return UTRIE2_GET16_FROM_U16_SINGLE_LEAD(trie, c);
+    } else if(trie->data32!=NULL) {
+        return UTRIE2_GET32_FROM_U16_SINGLE_LEAD(trie, c);
+    } else {
+        return get32(trie->newTrie, c, FALSE);
+    }
+}
+
+static inline int32_t
+u8Index(const UTrie2 *trie, UChar32 c, int32_t i) {
+    int32_t idx=
+        _UTRIE2_INDEX_FROM_CP(
+            trie,
+            trie->data32==NULL ? trie->indexLength : 0,
+            c);
+    return (idx<<3)|i;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie2_internalU8NextIndex(const UTrie2 *trie, UChar32 c,
+                           const uint8_t *src, const uint8_t *limit) {
+    int32_t i, length;
+    i=0;
+    /* support 64-bit pointers by avoiding cast of arbitrary difference */
+    if((limit-src)<=7) {
+        length=(int32_t)(limit-src);
+    } else {
+        length=7;
+    }
+    c=utf8_nextCharSafeBody(src, &i, length, c, -1);
+    return u8Index(trie, c, i);
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie2_internalU8PrevIndex(const UTrie2 *trie, UChar32 c,
+                           const uint8_t *start, const uint8_t *src) {
+    int32_t i, length;
+    /* support 64-bit pointers by avoiding cast of arbitrary difference */
+    if((src-start)<=7) {
+        i=length=(int32_t)(src-start);
+    } else {
+        i=length=7;
+        start=src-7;
+    }
+    c=utf8_prevCharSafeBody(start, 0, &i, c, -1);
+    i=length-i;  /* number of bytes read backward from src */
+    return u8Index(trie, c, i);
+}
+
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_openFromSerialized(UTrie2ValueBits valueBits,
+                          const void *data, int32_t length, int32_t *pActualLength,
+                          UErrorCode *pErrorCode) {
+    const UTrie2Header *header;
+    const uint16_t *p16;
+    int32_t actualLength;
+
+    UTrie2 tempTrie;
+    UTrie2 *trie;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( length<=0 || (U_POINTER_MASK_LSB(data, 3)!=0) ||
+        valueBits<0 || UTRIE2_COUNT_VALUE_BITS<=valueBits
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* enough data for a trie header? */
+    if(length<(int32_t)sizeof(UTrie2Header)) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    /* check the signature */
+    header=(const UTrie2Header *)data;
+    if(header->signature!=UTRIE2_SIG) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    /* get the options */
+    if(valueBits!=(UTrie2ValueBits)(header->options&UTRIE2_OPTIONS_VALUE_BITS_MASK)) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    /* get the length values and offsets */
+    uprv_memset(&tempTrie, 0, sizeof(tempTrie));
+    tempTrie.indexLength=header->indexLength;
+    tempTrie.dataLength=header->shiftedDataLength<<UTRIE2_INDEX_SHIFT;
+    tempTrie.index2NullOffset=header->index2NullOffset;
+    tempTrie.dataNullOffset=header->dataNullOffset;
+
+    tempTrie.highStart=header->shiftedHighStart<<UTRIE2_SHIFT_1;
+    tempTrie.highValueIndex=tempTrie.dataLength-UTRIE2_DATA_GRANULARITY;
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        tempTrie.highValueIndex+=tempTrie.indexLength;
+    }
+
+    /* calculate the actual length */
+    actualLength=(int32_t)sizeof(UTrie2Header)+tempTrie.indexLength*2;
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        actualLength+=tempTrie.dataLength*2;
+    } else {
+        actualLength+=tempTrie.dataLength*4;
+    }
+    if(length<actualLength) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR;  /* not enough bytes */
+        return 0;
+    }
+
+    /* allocate the trie */
+    trie=(UTrie2 *)uprv_malloc(sizeof(UTrie2));
+    if(trie==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    uprv_memcpy(trie, &tempTrie, sizeof(tempTrie));
+    trie->memory=(uint32_t *)data;
+    trie->length=actualLength;
+    trie->isMemoryOwned=FALSE;
+#ifdef UTRIE2_DEBUG
+    trie->name="fromSerialized";
+#endif
+
+    /* set the pointers to its index and data arrays */
+    p16=(const uint16_t *)(header+1);
+    trie->index=p16;
+    p16+=trie->indexLength;
+
+    /* get the data */
+    switch(valueBits) {
+    case UTRIE2_16_VALUE_BITS:
+        trie->data16=p16;
+        trie->data32=NULL;
+        trie->initialValue=trie->index[trie->dataNullOffset];
+        trie->errorValue=trie->data16[UTRIE2_BAD_UTF8_DATA_OFFSET];
+        break;
+    case UTRIE2_32_VALUE_BITS:
+        trie->data16=NULL;
+        trie->data32=(const uint32_t *)p16;
+        trie->initialValue=trie->data32[trie->dataNullOffset];
+        trie->errorValue=trie->data32[UTRIE2_BAD_UTF8_DATA_OFFSET];
+        break;
+    default:
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    if(pActualLength!=NULL) {
+        *pActualLength=actualLength;
+    }
+    return trie;
+}
+
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_openDummy(UTrie2ValueBits valueBits,
+                 uint32_t initialValue, uint32_t errorValue,
+                 UErrorCode *pErrorCode) {
+    UTrie2 *trie;
+    UTrie2Header *header;
+    uint32_t *p;
+    uint16_t *dest16;
+    int32_t indexLength, dataLength, length, i;
+    int32_t dataMove;  /* >0 if the data is moved to the end of the index array */
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if(valueBits<0 || UTRIE2_COUNT_VALUE_BITS<=valueBits) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* calculate the total length of the dummy trie data */
+    indexLength=UTRIE2_INDEX_1_OFFSET;
+    dataLength=UTRIE2_DATA_START_OFFSET+UTRIE2_DATA_GRANULARITY;
+    length=(int32_t)sizeof(UTrie2Header)+indexLength*2;
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        length+=dataLength*2;
+    } else {
+        length+=dataLength*4;
+    }
+
+    /* allocate the trie */
+    trie=(UTrie2 *)uprv_malloc(sizeof(UTrie2));
+    if(trie==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    uprv_memset(trie, 0, sizeof(UTrie2));
+    trie->memory=uprv_malloc(length);
+    if(trie->memory==NULL) {
+        uprv_free(trie);
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+    trie->length=length;
+    trie->isMemoryOwned=TRUE;
+
+    /* set the UTrie2 fields */
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        dataMove=indexLength;
+    } else {
+        dataMove=0;
+    }
+
+    trie->indexLength=indexLength;
+    trie->dataLength=dataLength;
+    trie->index2NullOffset=UTRIE2_INDEX_2_OFFSET;
+    trie->dataNullOffset=(uint16_t)dataMove;
+    trie->initialValue=initialValue;
+    trie->errorValue=errorValue;
+    trie->highStart=0;
+    trie->highValueIndex=dataMove+UTRIE2_DATA_START_OFFSET;
+#ifdef UTRIE2_DEBUG
+    trie->name="dummy";
+#endif
+
+    /* set the header fields */
+    header=(UTrie2Header *)trie->memory;
+
+    header->signature=UTRIE2_SIG; /* "Tri2" */
+    header->options=(uint16_t)valueBits;
+
+    header->indexLength=(uint16_t)indexLength;
+    header->shiftedDataLength=(uint16_t)(dataLength>>UTRIE2_INDEX_SHIFT);
+    header->index2NullOffset=(uint16_t)UTRIE2_INDEX_2_OFFSET;
+    header->dataNullOffset=(uint16_t)dataMove;
+    header->shiftedHighStart=0;
+
+    /* fill the index and data arrays */
+    dest16=(uint16_t *)(header+1);
+    trie->index=dest16;
+
+    /* write the index-2 array values shifted right by UTRIE2_INDEX_SHIFT */
+    for(i=0; i<UTRIE2_INDEX_2_BMP_LENGTH; ++i) {
+        *dest16++=(uint16_t)(dataMove>>UTRIE2_INDEX_SHIFT);  /* null data block */
+    }
+
+    /* write UTF-8 2-byte index-2 values, not right-shifted */
+    for(i=0; i<(0xc2-0xc0); ++i) {                                  /* C0..C1 */
+        *dest16++=(uint16_t)(dataMove+UTRIE2_BAD_UTF8_DATA_OFFSET);
+    }
+    for(; i<(0xe0-0xc0); ++i) {                                     /* C2..DF */
+        *dest16++=(uint16_t)dataMove;
+    }
+
+    /* write the 16/32-bit data array */
+    switch(valueBits) {
+    case UTRIE2_16_VALUE_BITS:
+        /* write 16-bit data values */
+        trie->data16=dest16;
+        trie->data32=NULL;
+        for(i=0; i<0x80; ++i) {
+            *dest16++=(uint16_t)initialValue;
+        }
+        for(; i<0xc0; ++i) {
+            *dest16++=(uint16_t)errorValue;
+        }
+        /* highValue and reserved values */
+        for(i=0; i<UTRIE2_DATA_GRANULARITY; ++i) {
+            *dest16++=(uint16_t)initialValue;
+        }
+        break;
+    case UTRIE2_32_VALUE_BITS:
+        /* write 32-bit data values */
+        p=(uint32_t *)dest16;
+        trie->data16=NULL;
+        trie->data32=p;
+        for(i=0; i<0x80; ++i) {
+            *p++=initialValue;
+        }
+        for(; i<0xc0; ++i) {
+            *p++=errorValue;
+        }
+        /* highValue and reserved values */
+        for(i=0; i<UTRIE2_DATA_GRANULARITY; ++i) {
+            *p++=initialValue;
+        }
+        break;
+    default:
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    return trie;
+}
+
+U_CAPI void U_EXPORT2
+utrie2_close(UTrie2 *trie) {
+    if(trie!=NULL) {
+        if(trie->isMemoryOwned) {
+            uprv_free(trie->memory);
+        }
+        if(trie->newTrie!=NULL) {
+            uprv_free(trie->newTrie->data);
+#ifdef UCPTRIE_DEBUG
+            umutablecptrie_close(trie->newTrie->t3);
+#endif
+            uprv_free(trie->newTrie);
+        }
+        uprv_free(trie);
+    }
+}
+
+U_CAPI UBool U_EXPORT2
+utrie2_isFrozen(const UTrie2 *trie) {
+    return (UBool)(trie->newTrie==NULL);
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie2_serialize(const UTrie2 *trie,
+                 void *data, int32_t capacity,
+                 UErrorCode *pErrorCode) {
+    /* argument check */
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+
+    if( trie==NULL || trie->memory==NULL || trie->newTrie!=NULL ||
+        capacity<0 || (capacity>0 && (data==NULL || (U_POINTER_MASK_LSB(data, 3)!=0)))
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    if(capacity>=trie->length) {
+        uprv_memcpy(data, trie->memory, trie->length);
+    } else {
+        *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
+    }
+    return trie->length;
+}
+
+/* enumeration -------------------------------------------------------------- */
+
+#define MIN_VALUE(a, b) ((a)<(b) ? (a) : (b))
+
+/* default UTrie2EnumValue() returns the input value itself */
+static uint32_t U_CALLCONV
+enumSameValue(const void * /*context*/, uint32_t value) {
+    return value;
+}
+
+/**
+ * Enumerate all ranges of code points with the same relevant values.
+ * The values are transformed from the raw trie entries by the enumValue function.
+ *
+ * Currently requires start<limit and both start and limit must be multiples
+ * of UTRIE2_DATA_BLOCK_LENGTH.
+ *
+ * Optimizations:
+ * - Skip a whole block if we know that it is filled with a single value,
+ *   and it is the same as we visited just before.
+ * - Handle the null block specially because we know a priori that it is filled
+ *   with a single value.
+ */
+static void
+enumEitherTrie(const UTrie2 *trie,
+               UChar32 start, UChar32 limit,
+               UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange, const void *context) {
+    const uint32_t *data32;
+    const uint16_t *idx;
+
+    uint32_t value, prevValue, initialValue;
+    UChar32 c, prev, highStart;
+    int32_t j, i2Block, prevI2Block, index2NullOffset, block, prevBlock, nullBlock;
+
+    if(enumRange==NULL) {
+        return;
+    }
+    if(enumValue==NULL) {
+        enumValue=enumSameValue;
+    }
+
+    if(trie->newTrie==NULL) {
+        /* frozen trie */
+        idx=trie->index;
+        U_ASSERT(idx!=NULL); /* the following code assumes trie->newTrie is not NULL when idx is NULL */
+        data32=trie->data32;
+
+        index2NullOffset=trie->index2NullOffset;
+        nullBlock=trie->dataNullOffset;
+    } else {
+        /* unfrozen, mutable trie */
+        idx=NULL;
+        data32=trie->newTrie->data;
+        U_ASSERT(data32!=NULL); /* the following code assumes idx is not NULL when data32 is NULL */
+
+        index2NullOffset=trie->newTrie->index2NullOffset;
+        nullBlock=trie->newTrie->dataNullOffset;
+    }
+
+    highStart=trie->highStart;
+
+    /* get the enumeration value that corresponds to an initial-value trie data entry */
+    initialValue=enumValue(context, trie->initialValue);
+
+    /* set variables for previous range */
+    prevI2Block=-1;
+    prevBlock=-1;
+    prev=start;
+    prevValue=0;
+
+    /* enumerate index-2 blocks */
+    for(c=start; c<limit && c<highStart;) {
+        /* Code point limit for iterating inside this i2Block. */
+        UChar32 tempLimit=c+UTRIE2_CP_PER_INDEX_1_ENTRY;
+        if(limit<tempLimit) {
+            tempLimit=limit;
+        }
+        if(c<=0xffff) {
+            if(!U_IS_SURROGATE(c)) {
+                i2Block=c>>UTRIE2_SHIFT_2;
+            } else if(U_IS_SURROGATE_LEAD(c)) {
+                /*
+                 * Enumerate values for lead surrogate code points, not code units:
+                 * This special block has half the normal length.
+                 */
+                i2Block=UTRIE2_LSCP_INDEX_2_OFFSET;
+                tempLimit=MIN_VALUE(0xdc00, limit);
+            } else {
+                /*
+                 * Switch back to the normal part of the index-2 table.
+                 * Enumerate the second half of the surrogates block.
+                 */
+                i2Block=0xd800>>UTRIE2_SHIFT_2;
+                tempLimit=MIN_VALUE(0xe000, limit);
+            }
+        } else {
+            /* supplementary code points */
+            if(idx!=NULL) {
+                i2Block=idx[(UTRIE2_INDEX_1_OFFSET-UTRIE2_OMITTED_BMP_INDEX_1_LENGTH)+
+                              (c>>UTRIE2_SHIFT_1)];
+            } else {
+                i2Block=trie->newTrie->index1[c>>UTRIE2_SHIFT_1];
+            }
+            if(i2Block==prevI2Block && (c-prev)>=UTRIE2_CP_PER_INDEX_1_ENTRY) {
+                /*
+                 * The index-2 block is the same as the previous one, and filled with prevValue.
+                 * Only possible for supplementary code points because the linear-BMP index-2
+                 * table creates unique i2Block values.
+                 */
+                c+=UTRIE2_CP_PER_INDEX_1_ENTRY;
+                continue;
+            }
+        }
+        prevI2Block=i2Block;
+        if(i2Block==index2NullOffset) {
+            /* this is the null index-2 block */
+            if(prevValue!=initialValue) {
+                if(prev<c && !enumRange(context, prev, c-1, prevValue)) {
+                    return;
+                }
+                prevBlock=nullBlock;
+                prev=c;
+                prevValue=initialValue;
+            }
+            c+=UTRIE2_CP_PER_INDEX_1_ENTRY;
+        } else {
+            /* enumerate data blocks for one index-2 block */
+            int32_t i2, i2Limit;
+            i2=(c>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK;
+            if((c>>UTRIE2_SHIFT_1)==(tempLimit>>UTRIE2_SHIFT_1)) {
+                i2Limit=(tempLimit>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK;
+            } else {
+                i2Limit=UTRIE2_INDEX_2_BLOCK_LENGTH;
+            }
+            for(; i2<i2Limit; ++i2) {
+                if(idx!=NULL) {
+                    block=(int32_t)idx[i2Block+i2]<<UTRIE2_INDEX_SHIFT;
+                } else {
+                    block=trie->newTrie->index2[i2Block+i2];
+                }
+                if(block==prevBlock && (c-prev)>=UTRIE2_DATA_BLOCK_LENGTH) {
+                    /* the block is the same as the previous one, and filled with prevValue */
+                    c+=UTRIE2_DATA_BLOCK_LENGTH;
+                    continue;
+                }
+                prevBlock=block;
+                if(block==nullBlock) {
+                    /* this is the null data block */
+                    if(prevValue!=initialValue) {
+                        if(prev<c && !enumRange(context, prev, c-1, prevValue)) {
+                            return;
+                        }
+                        prev=c;
+                        prevValue=initialValue;
+                    }
+                    c+=UTRIE2_DATA_BLOCK_LENGTH;
+                } else {
+                    for(j=0; j<UTRIE2_DATA_BLOCK_LENGTH; ++j) {
+                        value=enumValue(context, data32!=NULL ? data32[block+j] : idx[block+j]);
+                        if(value!=prevValue) {
+                            if(prev<c && !enumRange(context, prev, c-1, prevValue)) {
+                                return;
+                            }
+                            prev=c;
+                            prevValue=value;
+                        }
+                        ++c;
+                    }
+                }
+            }
+        }
+    }
+
+    if(c>limit) {
+        c=limit;  /* could be higher if in the index2NullOffset */
+    } else if(c<limit) {
+        /* c==highStart<limit */
+        uint32_t highValue;
+        if(idx!=NULL) {
+            highValue=
+                data32!=NULL ?
+                    data32[trie->highValueIndex] :
+                    idx[trie->highValueIndex];
+        } else {
+            highValue=trie->newTrie->data[trie->newTrie->dataLength-UTRIE2_DATA_GRANULARITY];
+        }
+        value=enumValue(context, highValue);
+        if(value!=prevValue) {
+            if(prev<c && !enumRange(context, prev, c-1, prevValue)) {
+                return;
+            }
+            prev=c;
+            prevValue=value;
+        }
+        c=limit;
+    }
+
+    /* deliver last range */
+    enumRange(context, prev, c-1, prevValue);
+}
+
+U_CAPI void U_EXPORT2
+utrie2_enum(const UTrie2 *trie,
+            UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange, const void *context) {
+    enumEitherTrie(trie, 0, 0x110000, enumValue, enumRange, context);
+}
+
+U_CAPI void U_EXPORT2
+utrie2_enumForLeadSurrogate(const UTrie2 *trie, UChar32 lead,
+                            UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange,
+                            const void *context) {
+    if(!U16_IS_LEAD(lead)) {
+        return;
+    }
+    lead=(lead-0xd7c0)<<10;   /* start code point */
+    enumEitherTrie(trie, lead, lead+0x400, enumValue, enumRange, context);
+}
+
+/* C++ convenience wrappers ------------------------------------------------- */
+
+U_NAMESPACE_BEGIN
+
+uint16_t BackwardUTrie2StringIterator::previous16() {
+    codePointLimit=codePointStart;
+    if(start>=codePointStart) {
+        codePoint=U_SENTINEL;
+        return static_cast<uint16_t>(trie->errorValue);
+    }
+    uint16_t result;
+    UTRIE2_U16_PREV16(trie, start, codePointStart, codePoint, result);
+    return result;
+}
+
+uint16_t ForwardUTrie2StringIterator::next16() {
+    codePointStart=codePointLimit;
+    if(codePointLimit==limit) {
+        codePoint=U_SENTINEL;
+        return static_cast<uint16_t>(trie->errorValue);
+    }
+    uint16_t result;
+    UTRIE2_U16_NEXT16(trie, codePointLimit, limit, codePoint, result);
+    return result;
+}
+
+U_NAMESPACE_END
diff --git a/thirdparty/icu4c/common/utrie2.h b/thirdparty/icu4c/common/utrie2.h
new file mode 100644
index 0000000000..d1e1e15a6e
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie2.h
@@ -0,0 +1,955 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie2.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2008aug16 (starting from a copy of utrie.h)
+*   created by: Markus W. Scherer
+*/
+
+#ifndef __UTRIE2_H__
+#define __UTRIE2_H__
+
+#include "unicode/utypes.h"
+#include "unicode/utf8.h"
+#include "putilimp.h"
+
+U_CDECL_BEGIN
+
+struct UTrie;  /* forward declaration */
+#ifndef __UTRIE_H__
+typedef struct UTrie UTrie;
+#endif
+
+/**
+ * \file
+ *
+ * This is a common implementation of a Unicode trie.
+ * It is a kind of compressed, serializable table of 16- or 32-bit values associated with
+ * Unicode code points (0..0x10ffff). (A map from code points to integers.)
+ *
+ * This is the second common version of a Unicode trie (hence the name UTrie2).
+ * Compared with UTrie version 1:
+ * - Still splitting BMP code points 11:5 bits for index and data table lookups.
+ * - Still separate data for lead surrogate code _units_ vs. code _points_,
+ *   but the lead surrogate code unit values are not required any more
+ *   for data lookup for supplementary code points.
+ * - The "folding" mechanism is removed. In UTrie version 1, this somewhat
+ *   hard-to-explain mechanism was meant to be used for optimized UTF-16
+ *   processing, with application-specific encoding of indexing bits
+ *   in the lead surrogate data for the associated supplementary code points.
+ * - For the last single-value code point range (ending with U+10ffff),
+ *   the starting code point ("highStart") and the value are stored.
+ * - For supplementary code points U+10000..highStart-1 a three-table lookup
+ *   (two index tables and one data table) is used. The first index
+ *   is truncated, omitting both the BMP portion and the high range.
+ * - There is a special small index for 2-byte UTF-8, and the initial data
+ *   entries are designed for fast 1/2-byte UTF-8 lookup.
+ *   Starting with ICU 60, C0 and C1 are not recognized as UTF-8 lead bytes any more at all,
+ *   and the associated 2-byte indexes are unused.
+ */
+
+/**
+ * Trie structure.
+ * Use only with public API macros and functions.
+ */
+struct UTrie2;
+typedef struct UTrie2 UTrie2;
+
+/* Public UTrie2 API functions: read-only access ---------------------------- */
+
+/**
+ * Selectors for the width of a UTrie2 data value.
+ */
+enum UTrie2ValueBits {
+    /** 16 bits per UTrie2 data value. */
+    UTRIE2_16_VALUE_BITS,
+    /** 32 bits per UTrie2 data value. */
+    UTRIE2_32_VALUE_BITS,
+    /** Number of selectors for the width of UTrie2 data values. */
+    UTRIE2_COUNT_VALUE_BITS
+};
+typedef enum UTrie2ValueBits UTrie2ValueBits;
+
+/**
+ * Open a frozen trie from its serialized from, stored in 32-bit-aligned memory.
+ * Inverse of utrie2_serialize().
+ * The memory must remain valid and unchanged as long as the trie is used.
+ * You must utrie2_close() the trie once you are done using it.
+ *
+ * @param valueBits selects the data entry size; results in an
+ *                  U_INVALID_FORMAT_ERROR if it does not match the serialized form
+ * @param data a pointer to 32-bit-aligned memory containing the serialized form of a UTrie2
+ * @param length the number of bytes available at data;
+ *               can be more than necessary
+ * @param pActualLength receives the actual number of bytes at data taken up by the trie data;
+ *                      can be NULL
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the unserialized trie
+ *
+ * @see utrie2_open
+ * @see utrie2_serialize
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_openFromSerialized(UTrie2ValueBits valueBits,
+                          const void *data, int32_t length, int32_t *pActualLength,
+                          UErrorCode *pErrorCode);
+
+/**
+ * Open a frozen, empty "dummy" trie.
+ * A dummy trie is an empty trie, used when a real data trie cannot
+ * be loaded. Equivalent to calling utrie2_open() and utrie2_freeze(),
+ * but without internally creating and compacting/serializing the
+ * builder data structure.
+ *
+ * The trie always returns the initialValue,
+ * or the errorValue for out-of-range code points and illegal UTF-8.
+ *
+ * You must utrie2_close() the trie once you are done using it.
+ *
+ * @param valueBits selects the data entry size
+ * @param initialValue the initial value that is set for all code points
+ * @param errorValue the value for out-of-range code points and illegal UTF-8
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return the dummy trie
+ *
+ * @see utrie2_openFromSerialized
+ * @see utrie2_open
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_openDummy(UTrie2ValueBits valueBits,
+                 uint32_t initialValue, uint32_t errorValue,
+                 UErrorCode *pErrorCode);
+
+/**
+ * Get a value from a code point as stored in the trie.
+ * Easier to use than UTRIE2_GET16() and UTRIE2_GET32() but slower.
+ * Easier to use because, unlike the macros, this function works on all UTrie2
+ * objects, frozen or not, holding 16-bit or 32-bit data values.
+ *
+ * @param trie the trie
+ * @param c the code point
+ * @return the value
+ */
+U_CAPI uint32_t U_EXPORT2
+utrie2_get32(const UTrie2 *trie, UChar32 c);
+
+/* enumeration callback types */
+
+/**
+ * Callback from utrie2_enum(), extracts a uint32_t value from a
+ * trie value. This value will be passed on to the UTrie2EnumRange function.
+ *
+ * @param context an opaque pointer, as passed into utrie2_enum()
+ * @param value a value from the trie
+ * @return the value that is to be passed on to the UTrie2EnumRange function
+ */
+typedef uint32_t U_CALLCONV
+UTrie2EnumValue(const void *context, uint32_t value);
+
+/**
+ * Callback from utrie2_enum(), is called for each contiguous range
+ * of code points with the same value as retrieved from the trie and
+ * transformed by the UTrie2EnumValue function.
+ *
+ * The callback function can stop the enumeration by returning false.
+ *
+ * @param context an opaque pointer, as passed into utrie2_enum()
+ * @param start the first code point in a contiguous range with value
+ * @param end the last code point in a contiguous range with value (inclusive)
+ * @param value the value that is set for all code points in [start..end]
+ * @return false to stop the enumeration
+ */
+typedef UBool U_CALLCONV
+UTrie2EnumRange(const void *context, UChar32 start, UChar32 end, uint32_t value);
+
+/**
+ * Enumerate efficiently all values in a trie.
+ * Do not modify the trie during the enumeration.
+ *
+ * For each entry in the trie, the value to be delivered is passed through
+ * the UTrie2EnumValue function.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * For each contiguous range of code points with a given (transformed) value,
+ * the UTrie2EnumRange function is called.
+ *
+ * @param trie a pointer to the trie
+ * @param enumValue a pointer to a function that may transform the trie entry value,
+ *                  or NULL if the values from the trie are to be used directly
+ * @param enumRange a pointer to a function that is called for each contiguous range
+ *                  of code points with the same (transformed) value
+ * @param context an opaque pointer that is passed on to the callback functions
+ */
+U_CAPI void U_EXPORT2
+utrie2_enum(const UTrie2 *trie,
+            UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange, const void *context);
+
+/* Building a trie ---------------------------------------------------------- */
+
+/**
+ * Open an empty, writable trie. At build time, 32-bit data values are used.
+ * utrie2_freeze() takes a valueBits parameter
+ * which determines the data value width in the serialized and frozen forms.
+ * You must utrie2_close() the trie once you are done using it.
+ *
+ * @param initialValue the initial value that is set for all code points
+ * @param errorValue the value for out-of-range code points and illegal UTF-8
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return a pointer to the allocated and initialized new trie
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_open(uint32_t initialValue, uint32_t errorValue, UErrorCode *pErrorCode);
+
+/**
+ * Clone a trie.
+ * You must utrie2_close() the clone once you are done using it.
+ *
+ * @param other the trie to clone
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return a pointer to the new trie clone
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_clone(const UTrie2 *other, UErrorCode *pErrorCode);
+
+/**
+ * Clone a trie. The clone will be mutable/writable even if the other trie
+ * is frozen. (See utrie2_freeze().)
+ * You must utrie2_close() the clone once you are done using it.
+ *
+ * @param other the trie to clone
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return a pointer to the new trie clone
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_cloneAsThawed(const UTrie2 *other, UErrorCode *pErrorCode);
+
+/**
+ * Close a trie and release associated memory.
+ *
+ * @param trie the trie
+ */
+U_CAPI void U_EXPORT2
+utrie2_close(UTrie2 *trie);
+
+/**
+ * Set a value for a code point.
+ *
+ * @param trie the unfrozen trie
+ * @param c the code point
+ * @param value the value
+ * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes:
+ * - U_NO_WRITE_PERMISSION if the trie is frozen
+ */
+U_CAPI void U_EXPORT2
+utrie2_set32(UTrie2 *trie, UChar32 c, uint32_t value, UErrorCode *pErrorCode);
+
+/**
+ * Set a value in a range of code points [start..end].
+ * All code points c with start<=c<=end will get the value if
+ * overwrite is true or if the old value is the initial value.
+ *
+ * @param trie the unfrozen trie
+ * @param start the first code point to get the value
+ * @param end the last code point to get the value (inclusive)
+ * @param value the value
+ * @param overwrite flag for whether old non-initial values are to be overwritten
+ * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes:
+ * - U_NO_WRITE_PERMISSION if the trie is frozen
+ */
+U_CAPI void U_EXPORT2
+utrie2_setRange32(UTrie2 *trie,
+                  UChar32 start, UChar32 end,
+                  uint32_t value, UBool overwrite,
+                  UErrorCode *pErrorCode);
+
+/**
+ * Freeze a trie. Make it immutable (read-only) and compact it,
+ * ready for serialization and for use with fast macros.
+ * Functions to set values will fail after serializing.
+ *
+ * A trie can be frozen only once. If this function is called again with different
+ * valueBits then it will set a U_ILLEGAL_ARGUMENT_ERROR.
+ *
+ * @param trie the trie
+ * @param valueBits selects the data entry size; if smaller than 32 bits, then
+ *                  the values stored in the trie will be truncated
+ * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes:
+ * - U_INDEX_OUTOFBOUNDS_ERROR if the compacted index or data arrays are too long
+ *                             for serialization
+ *                             (the trie will be immutable and usable,
+ *                             but not frozen and not usable with the fast macros)
+ *
+ * @see utrie2_cloneAsThawed
+ */
+U_CAPI void U_EXPORT2
+utrie2_freeze(UTrie2 *trie, UTrie2ValueBits valueBits, UErrorCode *pErrorCode);
+
+/**
+ * Test if the trie is frozen. (See utrie2_freeze().)
+ *
+ * @param trie the trie
+ * @return true if the trie is frozen, that is, immutable, ready for serialization
+ *         and for use with fast macros
+ */
+U_CAPI UBool U_EXPORT2
+utrie2_isFrozen(const UTrie2 *trie);
+
+/**
+ * Serialize a frozen trie into 32-bit aligned memory.
+ * If the trie is not frozen, then the function returns with a U_ILLEGAL_ARGUMENT_ERROR.
+ * A trie can be serialized multiple times.
+ *
+ * @param trie the frozen trie
+ * @param data a pointer to 32-bit-aligned memory to be filled with the trie data,
+ *             can be NULL if capacity==0
+ * @param capacity the number of bytes available at data,
+ *                 or 0 for preflighting
+ * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes:
+ * - U_BUFFER_OVERFLOW_ERROR if the data storage block is too small for serialization
+ * - U_ILLEGAL_ARGUMENT_ERROR if the trie is not frozen or the data and capacity
+ *                            parameters are bad
+ * @return the number of bytes written or needed for the trie
+ *
+ * @see utrie2_openFromSerialized()
+ */
+U_CAPI int32_t U_EXPORT2
+utrie2_serialize(const UTrie2 *trie,
+                 void *data, int32_t capacity,
+                 UErrorCode *pErrorCode);
+
+/* Public UTrie2 API: miscellaneous functions ------------------------------- */
+
+/**
+ * Build a UTrie2 (version 2) from a UTrie (version 1).
+ * Enumerates all values in the UTrie and builds a UTrie2 with the same values.
+ * The resulting UTrie2 will be frozen.
+ *
+ * @param trie1 the runtime UTrie structure to be enumerated
+ * @param errorValue the value for out-of-range code points and illegal UTF-8
+ * @param pErrorCode an in/out ICU UErrorCode
+ * @return The frozen UTrie2 with the same values as the UTrie.
+ */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_fromUTrie(const UTrie *trie1, uint32_t errorValue, UErrorCode *pErrorCode);
+
+/* Public UTrie2 API macros ------------------------------------------------- */
+
+/*
+ * These macros provide fast data lookup from a frozen trie.
+ * They will crash when used on an unfrozen trie.
+ */
+
+/**
+ * Return a 16-bit trie value from a code point, with range checking.
+ * Returns trie->errorValue if c is not in the range 0..U+10ffff.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code point
+ * @return (uint16_t) The code point's trie value.
+ */
+#define UTRIE2_GET16(trie, c) _UTRIE2_GET((trie), index, (trie)->indexLength, (c))
+
+/**
+ * Return a 32-bit trie value from a code point, with range checking.
+ * Returns trie->errorValue if c is not in the range 0..U+10ffff.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code point
+ * @return (uint32_t) The code point's trie value.
+ */
+#define UTRIE2_GET32(trie, c) _UTRIE2_GET((trie), data32, 0, (c))
+
+/**
+ * UTF-16: Get the next code point (UChar32 c, out), post-increment src,
+ * and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated
+ * @param c (UChar32, out) variable for the code point
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE2_U16_NEXT16(trie, src, limit, c, result) _UTRIE2_U16_NEXT(trie, index, src, limit, c, result)
+
+/**
+ * UTF-16: Get the next code point (UChar32 c, out), post-increment src,
+ * and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated
+ * @param c (UChar32, out) variable for the code point
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE2_U16_NEXT32(trie, src, limit, c, result) _UTRIE2_U16_NEXT(trie, data32, src, limit, c, result)
+
+/**
+ * UTF-16: Get the previous code point (UChar32 c, out), pre-decrement src,
+ * and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param start (const UChar *, in) the start pointer for the text
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param c (UChar32, out) variable for the code point
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE2_U16_PREV16(trie, start, src, c, result) _UTRIE2_U16_PREV(trie, index, start, src, c, result)
+
+/**
+ * UTF-16: Get the previous code point (UChar32 c, out), pre-decrement src,
+ * and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param start (const UChar *, in) the start pointer for the text
+ * @param src (const UChar *, in/out) the source text pointer
+ * @param c (UChar32, out) variable for the code point
+ * @param result (uint32_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE2_U16_PREV32(trie, start, src, c, result) _UTRIE2_U16_PREV(trie, data32, start, src, c, result)
+
+/**
+ * UTF-8: Post-increment src and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param src (const char *, in/out) the source text pointer
+ * @param limit (const char *, in) the limit pointer for the text (must not be NULL)
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE2_U8_NEXT16(trie, src, limit, result)\
+    _UTRIE2_U8_NEXT(trie, data16, index, src, limit, result)
+
+/**
+ * UTF-8: Post-increment src and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param src (const char *, in/out) the source text pointer
+ * @param limit (const char *, in) the limit pointer for the text (must not be NULL)
+ * @param result (uint16_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE2_U8_NEXT32(trie, src, limit, result) \
+    _UTRIE2_U8_NEXT(trie, data32, data32, src, limit, result)
+
+/**
+ * UTF-8: Pre-decrement src and get a 16-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param start (const char *, in) the start pointer for the text
+ * @param src (const char *, in/out) the source text pointer
+ * @param result (uint16_t, out) uint16_t variable for the trie lookup result
+ */
+#define UTRIE2_U8_PREV16(trie, start, src, result) \
+    _UTRIE2_U8_PREV(trie, data16, index, start, src, result)
+
+/**
+ * UTF-8: Pre-decrement src and get a 32-bit value from the trie.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param start (const char *, in) the start pointer for the text
+ * @param src (const char *, in/out) the source text pointer
+ * @param result (uint16_t, out) uint32_t variable for the trie lookup result
+ */
+#define UTRIE2_U8_PREV32(trie, start, src, result) \
+    _UTRIE2_U8_PREV(trie, data32, data32, start, src, result)
+
+/* Public UTrie2 API: optimized UTF-16 access ------------------------------- */
+
+/*
+ * The following functions and macros are used for highly optimized UTF-16
+ * text processing. The UTRIE2_U16_NEXTxy() macros do not depend on these.
+ *
+ * A UTrie2 stores separate values for lead surrogate code _units_ vs. code _points_.
+ * UTF-16 text processing can be optimized by detecting surrogate pairs and
+ * assembling supplementary code points only when there is non-trivial data
+ * available.
+ *
+ * At build-time, use utrie2_enumForLeadSurrogate() to see if there
+ * is non-trivial (non-initialValue) data for any of the supplementary
+ * code points associated with a lead surrogate.
+ * If so, then set a special (application-specific) value for the
+ * lead surrogate code _unit_, with utrie2_set32ForLeadSurrogateCodeUnit().
+ *
+ * At runtime, use UTRIE2_GET16_FROM_U16_SINGLE_LEAD() or
+ * UTRIE2_GET32_FROM_U16_SINGLE_LEAD() per code unit. If there is non-trivial
+ * data and the code unit is a lead surrogate, then check if a trail surrogate
+ * follows. If so, assemble the supplementary code point with
+ * U16_GET_SUPPLEMENTARY() and look up its value with UTRIE2_GET16_FROM_SUPP()
+ * or UTRIE2_GET32_FROM_SUPP(); otherwise reset the lead
+ * surrogate's value or do a code point lookup for it.
+ *
+ * If there is only trivial data for lead and trail surrogates, then processing
+ * can often skip them. For example, in normalization or case mapping
+ * all characters that do not have any mappings are simply copied as is.
+ */
+
+/**
+ * Get a value from a lead surrogate code unit as stored in the trie.
+ *
+ * @param trie the trie
+ * @param c the code unit (U+D800..U+DBFF)
+ * @return the value
+ */
+U_CAPI uint32_t U_EXPORT2
+utrie2_get32FromLeadSurrogateCodeUnit(const UTrie2 *trie, UChar32 c);
+
+/**
+ * Enumerate the trie values for the 1024=0x400 code points
+ * corresponding to a given lead surrogate.
+ * For example, for the lead surrogate U+D87E it will enumerate the values
+ * for [U+2F800..U+2FC00[.
+ * Used by data builder code that sets special lead surrogate code unit values
+ * for optimized UTF-16 string processing.
+ *
+ * Do not modify the trie during the enumeration.
+ *
+ * Except for the limited code point range, this functions just like utrie2_enum():
+ * For each entry in the trie, the value to be delivered is passed through
+ * the UTrie2EnumValue function.
+ * The value is unchanged if that function pointer is NULL.
+ *
+ * For each contiguous range of code points with a given (transformed) value,
+ * the UTrie2EnumRange function is called.
+ *
+ * @param trie a pointer to the trie
+ * @param enumValue a pointer to a function that may transform the trie entry value,
+ *                  or NULL if the values from the trie are to be used directly
+ * @param enumRange a pointer to a function that is called for each contiguous range
+ *                  of code points with the same (transformed) value
+ * @param context an opaque pointer that is passed on to the callback functions
+ */
+U_CAPI void U_EXPORT2
+utrie2_enumForLeadSurrogate(const UTrie2 *trie, UChar32 lead,
+                            UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange,
+                            const void *context);
+
+/**
+ * Set a value for a lead surrogate code unit.
+ *
+ * @param trie the unfrozen trie
+ * @param lead the lead surrogate code unit (U+D800..U+DBFF)
+ * @param value the value
+ * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes:
+ * - U_NO_WRITE_PERMISSION if the trie is frozen
+ */
+U_CAPI void U_EXPORT2
+utrie2_set32ForLeadSurrogateCodeUnit(UTrie2 *trie,
+                                     UChar32 lead, uint32_t value,
+                                     UErrorCode *pErrorCode);
+
+/**
+ * Return a 16-bit trie value from a UTF-16 single/lead code unit (<=U+ffff).
+ * Same as UTRIE2_GET16() if c is a BMP code point except for lead surrogates,
+ * but smaller and faster.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code unit, must be 0<=c<=U+ffff
+ * @return (uint16_t) The code unit's trie value.
+ */
+#define UTRIE2_GET16_FROM_U16_SINGLE_LEAD(trie, c) _UTRIE2_GET_FROM_U16_SINGLE_LEAD((trie), index, c)
+
+/**
+ * Return a 32-bit trie value from a UTF-16 single/lead code unit (<=U+ffff).
+ * Same as UTRIE2_GET32() if c is a BMP code point except for lead surrogates,
+ * but smaller and faster.
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code unit, must be 0<=c<=U+ffff
+ * @return (uint32_t) The code unit's trie value.
+ */
+#define UTRIE2_GET32_FROM_U16_SINGLE_LEAD(trie, c) _UTRIE2_GET_FROM_U16_SINGLE_LEAD((trie), data32, c)
+
+/**
+ * Return a 16-bit trie value from a supplementary code point (U+10000..U+10ffff).
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code point, must be U+10000<=c<=U+10ffff
+ * @return (uint16_t) The code point's trie value.
+ */
+#define UTRIE2_GET16_FROM_SUPP(trie, c) _UTRIE2_GET_FROM_SUPP((trie), index, c)
+
+/**
+ * Return a 32-bit trie value from a supplementary code point (U+10000..U+10ffff).
+ *
+ * @param trie (const UTrie2 *, in) a frozen trie
+ * @param c (UChar32, in) the input code point, must be U+10000<=c<=U+10ffff
+ * @return (uint32_t) The code point's trie value.
+ */
+#define UTRIE2_GET32_FROM_SUPP(trie, c) _UTRIE2_GET_FROM_SUPP((trie), data32, c)
+
+U_CDECL_END
+
+/* C++ convenience wrappers ------------------------------------------------- */
+
+#ifdef __cplusplus
+
+#include "unicode/utf.h"
+#include "mutex.h"
+
+U_NAMESPACE_BEGIN
+
+// Use the Forward/Backward subclasses below.
+class UTrie2StringIterator : public UMemory {
+public:
+    UTrie2StringIterator(const UTrie2 *t, const UChar *p) :
+        trie(t), codePointStart(p), codePointLimit(p), codePoint(U_SENTINEL) {}
+
+    const UTrie2 *trie;
+    const UChar *codePointStart, *codePointLimit;
+    UChar32 codePoint;
+};
+
+class BackwardUTrie2StringIterator : public UTrie2StringIterator {
+public:
+    BackwardUTrie2StringIterator(const UTrie2 *t, const UChar *s, const UChar *p) :
+        UTrie2StringIterator(t, p), start(s) {}
+
+    uint16_t previous16();
+
+    const UChar *start;
+};
+
+class ForwardUTrie2StringIterator : public UTrie2StringIterator {
+public:
+    // Iteration limit l can be NULL.
+    // In that case, the caller must detect c==0 and stop.
+    ForwardUTrie2StringIterator(const UTrie2 *t, const UChar *p, const UChar *l) :
+        UTrie2StringIterator(t, p), limit(l) {}
+
+    uint16_t next16();
+
+    const UChar *limit;
+};
+
+U_NAMESPACE_END
+
+#endif
+
+/* Internal definitions ----------------------------------------------------- */
+
+U_CDECL_BEGIN
+
+/** Build-time trie structure. */
+struct UNewTrie2;
+typedef struct UNewTrie2 UNewTrie2;
+
+/*
+ * Trie structure definition.
+ *
+ * Either the data table is 16 bits wide and accessed via the index
+ * pointer, with each index item increased by indexLength;
+ * in this case, data32==NULL, and data16 is used for direct ASCII access.
+ *
+ * Or the data table is 32 bits wide and accessed via the data32 pointer.
+ */
+struct UTrie2 {
+    /* protected: used by macros and functions for reading values */
+    const uint16_t *index;
+    const uint16_t *data16;     /* for fast UTF-8 ASCII access, if 16b data */
+    const uint32_t *data32;     /* NULL if 16b data is used via index */
+
+    int32_t indexLength, dataLength;
+    uint16_t index2NullOffset;  /* 0xffff if there is no dedicated index-2 null block */
+    uint16_t dataNullOffset;
+    uint32_t initialValue;
+    /** Value returned for out-of-range code points and illegal UTF-8. */
+    uint32_t errorValue;
+
+    /* Start of the last range which ends at U+10ffff, and its value. */
+    UChar32 highStart;
+    int32_t highValueIndex;
+
+    /* private: used by builder and unserialization functions */
+    void *memory;           /* serialized bytes; NULL if not frozen yet */
+    int32_t length;         /* number of serialized bytes at memory; 0 if not frozen yet */
+    UBool isMemoryOwned;    /* true if the trie owns the memory */
+    UBool padding1;
+    int16_t padding2;
+    UNewTrie2 *newTrie;     /* builder object; NULL when frozen */
+
+#ifdef UTRIE2_DEBUG
+    const char *name;
+#endif
+};
+
+/**
+ * Trie constants, defining shift widths, index array lengths, etc.
+ *
+ * These are needed for the runtime macros but users can treat these as
+ * implementation details and skip to the actual public API further below.
+ */
+enum {
+    /** Shift size for getting the index-1 table offset. */
+    UTRIE2_SHIFT_1=6+5,
+
+    /** Shift size for getting the index-2 table offset. */
+    UTRIE2_SHIFT_2=5,
+
+    /**
+     * Difference between the two shift sizes,
+     * for getting an index-1 offset from an index-2 offset. 6=11-5
+     */
+    UTRIE2_SHIFT_1_2=UTRIE2_SHIFT_1-UTRIE2_SHIFT_2,
+
+    /**
+     * Number of index-1 entries for the BMP. 32=0x20
+     * This part of the index-1 table is omitted from the serialized form.
+     */
+    UTRIE2_OMITTED_BMP_INDEX_1_LENGTH=0x10000>>UTRIE2_SHIFT_1,
+
+    /** Number of code points per index-1 table entry. 2048=0x800 */
+    UTRIE2_CP_PER_INDEX_1_ENTRY=1<<UTRIE2_SHIFT_1,
+
+    /** Number of entries in an index-2 block. 64=0x40 */
+    UTRIE2_INDEX_2_BLOCK_LENGTH=1<<UTRIE2_SHIFT_1_2,
+
+    /** Mask for getting the lower bits for the in-index-2-block offset. */
+    UTRIE2_INDEX_2_MASK=UTRIE2_INDEX_2_BLOCK_LENGTH-1,
+
+    /** Number of entries in a data block. 32=0x20 */
+    UTRIE2_DATA_BLOCK_LENGTH=1<<UTRIE2_SHIFT_2,
+
+    /** Mask for getting the lower bits for the in-data-block offset. */
+    UTRIE2_DATA_MASK=UTRIE2_DATA_BLOCK_LENGTH-1,
+
+    /**
+     * Shift size for shifting left the index array values.
+     * Increases possible data size with 16-bit index values at the cost
+     * of compactability.
+     * This requires data blocks to be aligned by UTRIE2_DATA_GRANULARITY.
+     */
+    UTRIE2_INDEX_SHIFT=2,
+
+    /** The alignment size of a data block. Also the granularity for compaction. */
+    UTRIE2_DATA_GRANULARITY=1<<UTRIE2_INDEX_SHIFT,
+
+    /* Fixed layout of the first part of the index array. ------------------- */
+
+    /**
+     * The BMP part of the index-2 table is fixed and linear and starts at offset 0.
+     * Length=2048=0x800=0x10000>>UTRIE2_SHIFT_2.
+     */
+    UTRIE2_INDEX_2_OFFSET=0,
+
+    /**
+     * The part of the index-2 table for U+D800..U+DBFF stores values for
+     * lead surrogate code _units_ not code _points_.
+     * Values for lead surrogate code _points_ are indexed with this portion of the table.
+     * Length=32=0x20=0x400>>UTRIE2_SHIFT_2. (There are 1024=0x400 lead surrogates.)
+     */
+    UTRIE2_LSCP_INDEX_2_OFFSET=0x10000>>UTRIE2_SHIFT_2,
+    UTRIE2_LSCP_INDEX_2_LENGTH=0x400>>UTRIE2_SHIFT_2,
+
+    /** Count the lengths of both BMP pieces. 2080=0x820 */
+    UTRIE2_INDEX_2_BMP_LENGTH=UTRIE2_LSCP_INDEX_2_OFFSET+UTRIE2_LSCP_INDEX_2_LENGTH,
+
+    /**
+     * The 2-byte UTF-8 version of the index-2 table follows at offset 2080=0x820.
+     * Length 32=0x20 for lead bytes C0..DF, regardless of UTRIE2_SHIFT_2.
+     */
+    UTRIE2_UTF8_2B_INDEX_2_OFFSET=UTRIE2_INDEX_2_BMP_LENGTH,
+    UTRIE2_UTF8_2B_INDEX_2_LENGTH=0x800>>6,  /* U+0800 is the first code point after 2-byte UTF-8 */
+
+    /**
+     * The index-1 table, only used for supplementary code points, at offset 2112=0x840.
+     * Variable length, for code points up to highStart, where the last single-value range starts.
+     * Maximum length 512=0x200=0x100000>>UTRIE2_SHIFT_1.
+     * (For 0x100000 supplementary code points U+10000..U+10ffff.)
+     *
+     * The part of the index-2 table for supplementary code points starts
+     * after this index-1 table.
+     *
+     * Both the index-1 table and the following part of the index-2 table
+     * are omitted completely if there is only BMP data.
+     */
+    UTRIE2_INDEX_1_OFFSET=UTRIE2_UTF8_2B_INDEX_2_OFFSET+UTRIE2_UTF8_2B_INDEX_2_LENGTH,
+    UTRIE2_MAX_INDEX_1_LENGTH=0x100000>>UTRIE2_SHIFT_1,
+
+    /*
+     * Fixed layout of the first part of the data array. -----------------------
+     * Starts with 4 blocks (128=0x80 entries) for ASCII.
+     */
+
+    /**
+     * The illegal-UTF-8 data block follows the ASCII block, at offset 128=0x80.
+     * Used with linear access for single bytes 0..0xbf for simple error handling.
+     * Length 64=0x40, not UTRIE2_DATA_BLOCK_LENGTH.
+     */
+    UTRIE2_BAD_UTF8_DATA_OFFSET=0x80,
+
+    /** The start of non-linear-ASCII data blocks, at offset 192=0xc0. */
+    UTRIE2_DATA_START_OFFSET=0xc0
+};
+
+/* Internal functions and macros -------------------------------------------- */
+
+/**
+ * Internal function for part of the UTRIE2_U8_NEXTxx() macro implementations.
+ * Do not call directly.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utrie2_internalU8NextIndex(const UTrie2 *trie, UChar32 c,
+                           const uint8_t *src, const uint8_t *limit);
+
+/**
+ * Internal function for part of the UTRIE2_U8_PREVxx() macro implementations.
+ * Do not call directly.
+ * @internal
+ */
+U_CAPI int32_t U_EXPORT2
+utrie2_internalU8PrevIndex(const UTrie2 *trie, UChar32 c,
+                           const uint8_t *start, const uint8_t *src);
+
+
+/** Internal low-level trie getter. Returns a data index. */
+#define _UTRIE2_INDEX_RAW(offset, trieIndex, c) \
+    (((int32_t)((trieIndex)[(offset)+((c)>>UTRIE2_SHIFT_2)]) \
+    <<UTRIE2_INDEX_SHIFT)+ \
+    ((c)&UTRIE2_DATA_MASK))
+
+/** Internal trie getter from a UTF-16 single/lead code unit. Returns the data index. */
+#define _UTRIE2_INDEX_FROM_U16_SINGLE_LEAD(trieIndex, c) _UTRIE2_INDEX_RAW(0, trieIndex, c)
+
+/** Internal trie getter from a lead surrogate code point (D800..DBFF). Returns the data index. */
+#define _UTRIE2_INDEX_FROM_LSCP(trieIndex, c) \
+    _UTRIE2_INDEX_RAW(UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2), trieIndex, c)
+
+/** Internal trie getter from a BMP code point. Returns the data index. */
+#define _UTRIE2_INDEX_FROM_BMP(trieIndex, c) \
+    _UTRIE2_INDEX_RAW(U_IS_LEAD(c) ? UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2) : 0, \
+                      trieIndex, c)
+
+/** Internal trie getter from a supplementary code point below highStart. Returns the data index. */
+#define _UTRIE2_INDEX_FROM_SUPP(trieIndex, c) \
+    (((int32_t)((trieIndex)[ \
+        (trieIndex)[(UTRIE2_INDEX_1_OFFSET-UTRIE2_OMITTED_BMP_INDEX_1_LENGTH)+ \
+                      ((c)>>UTRIE2_SHIFT_1)]+ \
+        (((c)>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK)]) \
+    <<UTRIE2_INDEX_SHIFT)+ \
+    ((c)&UTRIE2_DATA_MASK))
+
+/**
+ * Internal trie getter from a code point, with checking that c is in 0..10FFFF.
+ * Returns the data index.
+ */
+#define _UTRIE2_INDEX_FROM_CP(trie, asciiOffset, c) \
+    ((uint32_t)(c)<0xd800 ? \
+        _UTRIE2_INDEX_RAW(0, (trie)->index, c) : \
+        (uint32_t)(c)<=0xffff ? \
+            _UTRIE2_INDEX_RAW( \
+                (c)<=0xdbff ? UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2) : 0, \
+                (trie)->index, c) : \
+            (uint32_t)(c)>0x10ffff ? \
+                (asciiOffset)+UTRIE2_BAD_UTF8_DATA_OFFSET : \
+                (c)>=(trie)->highStart ? \
+                    (trie)->highValueIndex : \
+                    _UTRIE2_INDEX_FROM_SUPP((trie)->index, c))
+
+/** Internal trie getter from a UTF-16 single/lead code unit. Returns the data. */
+#define _UTRIE2_GET_FROM_U16_SINGLE_LEAD(trie, data, c) \
+    (trie)->data[_UTRIE2_INDEX_FROM_U16_SINGLE_LEAD((trie)->index, c)]
+
+/** Internal trie getter from a supplementary code point. Returns the data. */
+#define _UTRIE2_GET_FROM_SUPP(trie, data, c) \
+    (trie)->data[(c)>=(trie)->highStart ? (trie)->highValueIndex : \
+                 _UTRIE2_INDEX_FROM_SUPP((trie)->index, c)]
+
+/**
+ * Internal trie getter from a code point, with checking that c is in 0..10FFFF.
+ * Returns the data.
+ */
+#define _UTRIE2_GET(trie, data, asciiOffset, c) \
+    (trie)->data[_UTRIE2_INDEX_FROM_CP(trie, asciiOffset, c)]
+
+/** Internal next-post-increment: get the next code point (c) and its data. */
+#define _UTRIE2_U16_NEXT(trie, data, src, limit, c, result) UPRV_BLOCK_MACRO_BEGIN { \
+    { \
+        uint16_t __c2; \
+        (c)=*(src)++; \
+        if(!U16_IS_LEAD(c)) { \
+            (result)=_UTRIE2_GET_FROM_U16_SINGLE_LEAD(trie, data, c); \
+        } else if((src)==(limit) || !U16_IS_TRAIL(__c2=*(src))) { \
+            (result)=(trie)->data[_UTRIE2_INDEX_FROM_LSCP((trie)->index, c)]; \
+        } else { \
+            ++(src); \
+            (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+            (result)=_UTRIE2_GET_FROM_SUPP((trie), data, (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal pre-decrement-previous: get the previous code point (c) and its data */
+#define _UTRIE2_U16_PREV(trie, data, start, src, c, result) UPRV_BLOCK_MACRO_BEGIN { \
+    { \
+        uint16_t __c2; \
+        (c)=*--(src); \
+        if(!U16_IS_TRAIL(c) || (src)==(start) || !U16_IS_LEAD(__c2=*((src)-1))) { \
+            (result)=(trie)->data[_UTRIE2_INDEX_FROM_BMP((trie)->index, c)]; \
+        } else { \
+            --(src); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            (result)=_UTRIE2_GET_FROM_SUPP((trie), data, (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal UTF-8 next-post-increment: get the next code point's data. */
+#define _UTRIE2_U8_NEXT(trie, ascii, data, src, limit, result) UPRV_BLOCK_MACRO_BEGIN { \
+    uint8_t __lead=(uint8_t)*(src)++; \
+    if(U8_IS_SINGLE(__lead)) { \
+        (result)=(trie)->ascii[__lead]; \
+    } else { \
+        uint8_t __t1, __t2; \
+        if( /* handle U+0800..U+FFFF inline */ \
+            0xe0<=__lead && __lead<0xf0 && ((src)+1)<(limit) && \
+            U8_IS_VALID_LEAD3_AND_T1(__lead, __t1=(uint8_t)*(src)) && \
+            (__t2=(uint8_t)(*((src)+1)-0x80))<= 0x3f \
+        ) { \
+            (src)+=2; \
+            (result)=(trie)->data[ \
+                ((int32_t)((trie)->index[((__lead-0xe0)<<(12-UTRIE2_SHIFT_2))+ \
+                                         ((__t1&0x3f)<<(6-UTRIE2_SHIFT_2))+(__t2>>UTRIE2_SHIFT_2)]) \
+                <<UTRIE2_INDEX_SHIFT)+ \
+                (__t2&UTRIE2_DATA_MASK)]; \
+        } else if( /* handle U+0080..U+07FF inline */ \
+            __lead<0xe0 && __lead>=0xc2 && (src)<(limit) && \
+            (__t1=(uint8_t)(*(src)-0x80))<=0x3f \
+        ) { \
+            ++(src); \
+            (result)=(trie)->data[ \
+                (trie)->index[(UTRIE2_UTF8_2B_INDEX_2_OFFSET-0xc0)+__lead]+ \
+                __t1]; \
+        } else { \
+            int32_t __index=utrie2_internalU8NextIndex((trie), __lead, (const uint8_t *)(src), \
+                                                                       (const uint8_t *)(limit)); \
+            (src)+=__index&7; \
+            (result)=(trie)->data[__index>>3]; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/** Internal UTF-8 pre-decrement-previous: get the previous code point's data. */
+#define _UTRIE2_U8_PREV(trie, ascii, data, start, src, result) UPRV_BLOCK_MACRO_BEGIN { \
+    uint8_t __b=(uint8_t)*--(src); \
+    if(U8_IS_SINGLE(__b)) { \
+        (result)=(trie)->ascii[__b]; \
+    } else { \
+        int32_t __index=utrie2_internalU8PrevIndex((trie), __b, (const uint8_t *)(start), \
+                                                                (const uint8_t *)(src)); \
+        (src)-=__index&7; \
+        (result)=(trie)->data[__index>>3]; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+U_CDECL_END
+
+#endif
diff --git a/thirdparty/icu4c/common/utrie2_builder.cpp b/thirdparty/icu4c/common/utrie2_builder.cpp
new file mode 100644
index 0000000000..8de824cc3d
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie2_builder.cpp
@@ -0,0 +1,1483 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie2_builder.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2008sep26 (split off from utrie2.c)
+*   created by: Markus W. Scherer
+*
+*   This is a common implementation of a Unicode trie.
+*   It is a kind of compressed, serializable table of 16- or 32-bit values associated with
+*   Unicode code points (0..0x10ffff).
+*   This is the second common version of a Unicode trie (hence the name UTrie2).
+*   See utrie2.h for a comparison.
+*
+*   This file contains only the builder code.
+*   See utrie2.c for the runtime and enumeration code.
+*/
+// #define UTRIE2_DEBUG
+#ifdef UTRIE2_DEBUG
+#   include <stdio.h>
+#endif
+// #define UCPTRIE_DEBUG
+
+#include "unicode/utypes.h"
+#ifdef UCPTRIE_DEBUG
+#include "unicode/ucptrie.h"
+#include "unicode/umutablecptrie.h"
+#include "ucptrie_impl.h"
+#endif
+#include "cmemory.h"
+#include "utrie2.h"
+#include "utrie2_impl.h"
+
+#include "utrie.h"  // for utrie2_fromUTrie()
+
+/* Implementation notes ----------------------------------------------------- */
+
+/*
+ * The UTRIE2_SHIFT_1, UTRIE2_SHIFT_2, UTRIE2_INDEX_SHIFT and other values
+ * have been chosen to minimize trie sizes overall.
+ * Most of the code is flexible enough to work with a range of values,
+ * within certain limits.
+ *
+ * Exception: Support for separate values for lead surrogate code _units_
+ * vs. code _points_ was added after the constants were fixed,
+ * and has not been tested nor particularly designed for different constant values.
+ * (Especially the utrie2_enum() code that jumps to the special LSCP index-2
+ * part and back.)
+ *
+ * Requires UTRIE2_SHIFT_2<=6. Otherwise 0xc0 which is the top of the ASCII-linear data
+ * including the bad-UTF-8-data block is not a multiple of UTRIE2_DATA_BLOCK_LENGTH
+ * and map[block>>UTRIE2_SHIFT_2] (used in reference counting and compaction
+ * remapping) stops working.
+ *
+ * Requires UTRIE2_SHIFT_1>=10 because utrie2_enumForLeadSurrogate()
+ * assumes that a single index-2 block is used for 0x400 code points
+ * corresponding to one lead surrogate.
+ *
+ * Requires UTRIE2_SHIFT_1<=16. Otherwise one single index-2 block contains
+ * more than one Unicode plane, and the split of the index-2 table into a BMP
+ * part and a supplementary part, with a gap in between, would not work.
+ *
+ * Requires UTRIE2_INDEX_SHIFT>=1 not because of the code but because
+ * there is data with more than 64k distinct values,
+ * for example for Unihan collation with a separate collation weight per
+ * Han character.
+ */
+
+/* Building a trie ----------------------------------------------------------*/
+
+enum {
+    /** The null index-2 block, following the gap in the index-2 table. */
+    UNEWTRIE2_INDEX_2_NULL_OFFSET=UNEWTRIE2_INDEX_GAP_OFFSET+UNEWTRIE2_INDEX_GAP_LENGTH,
+
+    /** The start of allocated index-2 blocks. */
+    UNEWTRIE2_INDEX_2_START_OFFSET=UNEWTRIE2_INDEX_2_NULL_OFFSET+UTRIE2_INDEX_2_BLOCK_LENGTH,
+
+    /**
+     * The null data block.
+     * Length 64=0x40 even if UTRIE2_DATA_BLOCK_LENGTH is smaller,
+     * to work with 6-bit trail bytes from 2-byte UTF-8.
+     */
+    UNEWTRIE2_DATA_NULL_OFFSET=UTRIE2_DATA_START_OFFSET,
+
+    /** The start of allocated data blocks. */
+    UNEWTRIE2_DATA_START_OFFSET=UNEWTRIE2_DATA_NULL_OFFSET+0x40,
+
+    /**
+     * The start of data blocks for U+0800 and above.
+     * Below, compaction uses a block length of 64 for 2-byte UTF-8.
+     * From here on, compaction uses UTRIE2_DATA_BLOCK_LENGTH.
+     * Data values for 0x780 code points beyond ASCII.
+     */
+    UNEWTRIE2_DATA_0800_OFFSET=UNEWTRIE2_DATA_START_OFFSET+0x780
+};
+
+/* Start with allocation of 16k data entries. */
+#define UNEWTRIE2_INITIAL_DATA_LENGTH ((int32_t)1<<14)
+
+/* Grow about 8x each time. */
+#define UNEWTRIE2_MEDIUM_DATA_LENGTH ((int32_t)1<<17)
+
+static int32_t
+allocIndex2Block(UNewTrie2 *trie);
+
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_open(uint32_t initialValue, uint32_t errorValue, UErrorCode *pErrorCode) {
+    UTrie2 *trie;
+    UNewTrie2 *newTrie;
+    uint32_t *data;
+    int32_t i, j;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+
+    trie=(UTrie2 *)uprv_malloc(sizeof(UTrie2));
+    newTrie=(UNewTrie2 *)uprv_malloc(sizeof(UNewTrie2));
+    data=(uint32_t *)uprv_malloc(UNEWTRIE2_INITIAL_DATA_LENGTH*4);
+    if(trie==NULL || newTrie==NULL || data==NULL) {
+        uprv_free(trie);
+        uprv_free(newTrie);
+        uprv_free(data);
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return 0;
+    }
+
+    uprv_memset(trie, 0, sizeof(UTrie2));
+    trie->initialValue=initialValue;
+    trie->errorValue=errorValue;
+    trie->highStart=0x110000;
+    trie->newTrie=newTrie;
+#ifdef UTRIE2_DEBUG
+    trie->name="open";
+#endif
+
+    newTrie->data=data;
+#ifdef UCPTRIE_DEBUG
+    newTrie->t3=umutablecptrie_open(initialValue, errorValue, pErrorCode);
+#endif
+    newTrie->dataCapacity=UNEWTRIE2_INITIAL_DATA_LENGTH;
+    newTrie->initialValue=initialValue;
+    newTrie->errorValue=errorValue;
+    newTrie->highStart=0x110000;
+    newTrie->firstFreeBlock=0;  /* no free block in the list */
+    newTrie->isCompacted=FALSE;
+
+    /*
+     * preallocate and reset
+     * - ASCII
+     * - the bad-UTF-8-data block
+     * - the null data block
+     */
+    for(i=0; i<0x80; ++i) {
+        newTrie->data[i]=initialValue;
+    }
+    for(; i<0xc0; ++i) {
+        newTrie->data[i]=errorValue;
+    }
+    for(i=UNEWTRIE2_DATA_NULL_OFFSET; i<UNEWTRIE2_DATA_START_OFFSET; ++i) {
+        newTrie->data[i]=initialValue;
+    }
+    newTrie->dataNullOffset=UNEWTRIE2_DATA_NULL_OFFSET;
+    newTrie->dataLength=UNEWTRIE2_DATA_START_OFFSET;
+
+    /* set the index-2 indexes for the 2=0x80>>UTRIE2_SHIFT_2 ASCII data blocks */
+    for(i=0, j=0; j<0x80; ++i, j+=UTRIE2_DATA_BLOCK_LENGTH) {
+        newTrie->index2[i]=j;
+        newTrie->map[i]=1;
+    }
+    /* reference counts for the bad-UTF-8-data block */
+    for(; j<0xc0; ++i, j+=UTRIE2_DATA_BLOCK_LENGTH) {
+        newTrie->map[i]=0;
+    }
+    /*
+     * Reference counts for the null data block: all blocks except for the ASCII blocks.
+     * Plus 1 so that we don't drop this block during compaction.
+     * Plus as many as needed for lead surrogate code points.
+     */
+    /* i==newTrie->dataNullOffset */
+    newTrie->map[i++]=
+        (0x110000>>UTRIE2_SHIFT_2)-
+        (0x80>>UTRIE2_SHIFT_2)+
+        1+
+        UTRIE2_LSCP_INDEX_2_LENGTH;
+    j+=UTRIE2_DATA_BLOCK_LENGTH;
+    for(; j<UNEWTRIE2_DATA_START_OFFSET; ++i, j+=UTRIE2_DATA_BLOCK_LENGTH) {
+        newTrie->map[i]=0;
+    }
+
+    /*
+     * set the remaining indexes in the BMP index-2 block
+     * to the null data block
+     */
+    for(i=0x80>>UTRIE2_SHIFT_2; i<UTRIE2_INDEX_2_BMP_LENGTH; ++i) {
+        newTrie->index2[i]=UNEWTRIE2_DATA_NULL_OFFSET;
+    }
+
+    /*
+     * Fill the index gap with impossible values so that compaction
+     * does not overlap other index-2 blocks with the gap.
+     */
+    for(i=0; i<UNEWTRIE2_INDEX_GAP_LENGTH; ++i) {
+        newTrie->index2[UNEWTRIE2_INDEX_GAP_OFFSET+i]=-1;
+    }
+
+    /* set the indexes in the null index-2 block */
+    for(i=0; i<UTRIE2_INDEX_2_BLOCK_LENGTH; ++i) {
+        newTrie->index2[UNEWTRIE2_INDEX_2_NULL_OFFSET+i]=UNEWTRIE2_DATA_NULL_OFFSET;
+    }
+    newTrie->index2NullOffset=UNEWTRIE2_INDEX_2_NULL_OFFSET;
+    newTrie->index2Length=UNEWTRIE2_INDEX_2_START_OFFSET;
+
+    /* set the index-1 indexes for the linear index-2 block */
+    for(i=0, j=0;
+        i<UTRIE2_OMITTED_BMP_INDEX_1_LENGTH;
+        ++i, j+=UTRIE2_INDEX_2_BLOCK_LENGTH
+    ) {
+        newTrie->index1[i]=j;
+    }
+
+    /* set the remaining index-1 indexes to the null index-2 block */
+    for(; i<UNEWTRIE2_INDEX_1_LENGTH; ++i) {
+        newTrie->index1[i]=UNEWTRIE2_INDEX_2_NULL_OFFSET;
+    }
+
+    /*
+     * Preallocate and reset data for U+0080..U+07ff,
+     * for 2-byte UTF-8 which will be compacted in 64-blocks
+     * even if UTRIE2_DATA_BLOCK_LENGTH is smaller.
+     */
+    for(i=0x80; i<0x800; i+=UTRIE2_DATA_BLOCK_LENGTH) {
+        utrie2_set32(trie, i, initialValue, pErrorCode);
+    }
+
+    return trie;
+}
+
+static UNewTrie2 *
+cloneBuilder(const UNewTrie2 *other) {
+    UNewTrie2 *trie;
+
+    trie=(UNewTrie2 *)uprv_malloc(sizeof(UNewTrie2));
+    if(trie==NULL) {
+        return NULL;
+    }
+
+    trie->data=(uint32_t *)uprv_malloc(other->dataCapacity*4);
+    if(trie->data==NULL) {
+        uprv_free(trie);
+        return NULL;
+    }
+#ifdef UCPTRIE_DEBUG
+    if(other->t3==nullptr) {
+        trie->t3=nullptr;
+    } else {
+        UErrorCode errorCode=U_ZERO_ERROR;
+        trie->t3=umutablecptrie_clone(other->t3, &errorCode);
+    }
+#endif
+    trie->dataCapacity=other->dataCapacity;
+
+    /* clone data */
+    uprv_memcpy(trie->index1, other->index1, sizeof(trie->index1));
+    uprv_memcpy(trie->index2, other->index2, (size_t)other->index2Length*4);
+    trie->index2NullOffset=other->index2NullOffset;
+    trie->index2Length=other->index2Length;
+
+    uprv_memcpy(trie->data, other->data, (size_t)other->dataLength*4);
+    trie->dataNullOffset=other->dataNullOffset;
+    trie->dataLength=other->dataLength;
+
+    /* reference counters */
+    if(other->isCompacted) {
+        trie->firstFreeBlock=0;
+    } else {
+        uprv_memcpy(trie->map, other->map, ((size_t)other->dataLength>>UTRIE2_SHIFT_2)*4);
+        trie->firstFreeBlock=other->firstFreeBlock;
+    }
+
+    trie->initialValue=other->initialValue;
+    trie->errorValue=other->errorValue;
+    trie->highStart=other->highStart;
+    trie->isCompacted=other->isCompacted;
+
+    return trie;
+}
+
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_clone(const UTrie2 *other, UErrorCode *pErrorCode) {
+    UTrie2 *trie;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(other==NULL || (other->memory==NULL && other->newTrie==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+
+    trie=(UTrie2 *)uprv_malloc(sizeof(UTrie2));
+    if(trie==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return NULL;
+    }
+    uprv_memcpy(trie, other, sizeof(UTrie2));
+
+    if(other->memory!=NULL) {
+        trie->memory=uprv_malloc(other->length);
+        if(trie->memory!=NULL) {
+            trie->isMemoryOwned=TRUE;
+            uprv_memcpy(trie->memory, other->memory, other->length);
+
+            /* make the clone's pointers point to its own memory */
+            trie->index=(uint16_t *)trie->memory+(other->index-(uint16_t *)other->memory);
+            if(other->data16!=NULL) {
+                trie->data16=(uint16_t *)trie->memory+(other->data16-(uint16_t *)other->memory);
+            }
+            if(other->data32!=NULL) {
+                trie->data32=(uint32_t *)trie->memory+(other->data32-(uint32_t *)other->memory);
+            }
+        }
+    } else /* other->newTrie!=NULL */ {
+        trie->newTrie=cloneBuilder(other->newTrie);
+    }
+
+    if(trie->memory==NULL && trie->newTrie==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        uprv_free(trie);
+        trie=NULL;
+    }
+    return trie;
+}
+
+typedef struct NewTrieAndStatus {
+    UTrie2 *trie;
+    UErrorCode errorCode;
+    UBool exclusiveLimit;  /* rather than inclusive range end */
+} NewTrieAndStatus;
+
+static UBool U_CALLCONV
+copyEnumRange(const void *context, UChar32 start, UChar32 end, uint32_t value) {
+    NewTrieAndStatus *nt=(NewTrieAndStatus *)context;
+    if(value!=nt->trie->initialValue) {
+        if(nt->exclusiveLimit) {
+            --end;
+        }
+        if(start==end) {
+            utrie2_set32(nt->trie, start, value, &nt->errorCode);
+        } else {
+            utrie2_setRange32(nt->trie, start, end, value, TRUE, &nt->errorCode);
+        }
+        return U_SUCCESS(nt->errorCode);
+    } else {
+        return TRUE;
+    }
+}
+
+#ifdef UTRIE2_DEBUG
+static long countInitial(const UTrie2 *trie) {
+    uint32_t initialValue=trie->initialValue;
+    int32_t length=trie->dataLength;
+    long count=0;
+    if(trie->data16!=nullptr) {
+        for(int32_t i=0; i<length; ++i) {
+            if(trie->data16[i]==initialValue) { ++count; }
+        }
+    } else {
+        for(int32_t i=0; i<length; ++i) {
+            if(trie->data32[i]==initialValue) { ++count; }
+        }
+    }
+    return count;
+}
+
+static void
+utrie_printLengths(const UTrie *trie) {
+    long indexLength=trie->indexLength;
+    long dataLength=(long)trie->dataLength;
+    long totalLength=(long)sizeof(UTrieHeader)+indexLength*2+dataLength*(trie->data32!=NULL ? 4 : 2);
+    printf("**UTrieLengths** index:%6ld  data:%6ld  serialized:%6ld\n",
+           indexLength, dataLength, totalLength);
+}
+
+static void
+utrie2_printLengths(const UTrie2 *trie, const char *which) {
+    long indexLength=trie->indexLength;
+    long dataLength=(long)trie->dataLength;
+    long totalLength=(long)sizeof(UTrie2Header)+indexLength*2+dataLength*(trie->data32!=NULL ? 4 : 2);
+    printf("**UTrie2Lengths(%s %s)** index:%6ld  data:%6ld  countInitial:%6ld  serialized:%6ld\n",
+           which, trie->name, indexLength, dataLength, countInitial(trie), totalLength);
+}
+#endif
+
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_cloneAsThawed(const UTrie2 *other, UErrorCode *pErrorCode) {
+    NewTrieAndStatus context;
+    UChar lead;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(other==NULL || (other->memory==NULL && other->newTrie==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    if(other->newTrie!=NULL && !other->newTrie->isCompacted) {
+        return utrie2_clone(other, pErrorCode);  /* clone an unfrozen trie */
+    }
+
+    /* Clone the frozen trie by enumerating it and building a new one. */
+    context.trie=utrie2_open(other->initialValue, other->errorValue, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    context.exclusiveLimit=FALSE;
+    context.errorCode=*pErrorCode;
+    utrie2_enum(other, NULL, copyEnumRange, &context);
+    *pErrorCode=context.errorCode;
+    for(lead=0xd800; lead<0xdc00; ++lead) {
+        uint32_t value;
+        if(other->data32==NULL) {
+            value=UTRIE2_GET16_FROM_U16_SINGLE_LEAD(other, lead);
+        } else {
+            value=UTRIE2_GET32_FROM_U16_SINGLE_LEAD(other, lead);
+        }
+        if(value!=other->initialValue) {
+            utrie2_set32ForLeadSurrogateCodeUnit(context.trie, lead, value, pErrorCode);
+        }
+    }
+    if(U_FAILURE(*pErrorCode)) {
+        utrie2_close(context.trie);
+        context.trie=NULL;
+    }
+    return context.trie;
+}
+
+/* Almost the same as utrie2_cloneAsThawed() but copies a UTrie and freezes the clone. */
+U_CAPI UTrie2 * U_EXPORT2
+utrie2_fromUTrie(const UTrie *trie1, uint32_t errorValue, UErrorCode *pErrorCode) {
+    NewTrieAndStatus context;
+    UChar lead;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    if(trie1==NULL) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return NULL;
+    }
+    context.trie=utrie2_open(trie1->initialValue, errorValue, pErrorCode);
+    if(U_FAILURE(*pErrorCode)) {
+        return NULL;
+    }
+    context.exclusiveLimit=TRUE;
+    context.errorCode=*pErrorCode;
+    utrie_enum(trie1, NULL, copyEnumRange, &context);
+    *pErrorCode=context.errorCode;
+    for(lead=0xd800; lead<0xdc00; ++lead) {
+        uint32_t value;
+        if(trie1->data32==NULL) {
+            value=UTRIE_GET16_FROM_LEAD(trie1, lead);
+        } else {
+            value=UTRIE_GET32_FROM_LEAD(trie1, lead);
+        }
+        if(value!=trie1->initialValue) {
+            utrie2_set32ForLeadSurrogateCodeUnit(context.trie, lead, value, pErrorCode);
+        }
+    }
+    if(U_SUCCESS(*pErrorCode)) {
+        utrie2_freeze(context.trie,
+                      trie1->data32!=NULL ? UTRIE2_32_VALUE_BITS : UTRIE2_16_VALUE_BITS,
+                      pErrorCode);
+    }
+#ifdef UTRIE2_DEBUG
+    if(U_SUCCESS(*pErrorCode)) {
+        utrie_printLengths(trie1);
+        utrie2_printLengths(context.trie, "fromUTrie");
+    }
+#endif
+    if(U_FAILURE(*pErrorCode)) {
+        utrie2_close(context.trie);
+        context.trie=NULL;
+    }
+    return context.trie;
+}
+
+static inline UBool
+isInNullBlock(UNewTrie2 *trie, UChar32 c, UBool forLSCP) {
+    int32_t i2, block;
+
+    if(U_IS_LEAD(c) && forLSCP) {
+        i2=(UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2))+
+            (c>>UTRIE2_SHIFT_2);
+    } else {
+        i2=trie->index1[c>>UTRIE2_SHIFT_1]+
+            ((c>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK);
+    }
+    block=trie->index2[i2];
+    return (UBool)(block==trie->dataNullOffset);
+}
+
+static int32_t
+allocIndex2Block(UNewTrie2 *trie) {
+    int32_t newBlock, newTop;
+
+    newBlock=trie->index2Length;
+    newTop=newBlock+UTRIE2_INDEX_2_BLOCK_LENGTH;
+    if(newTop>UPRV_LENGTHOF(trie->index2)) {
+        /*
+         * Should never occur.
+         * Either UTRIE2_MAX_BUILD_TIME_INDEX_LENGTH is incorrect,
+         * or the code writes more values than should be possible.
+         */
+        return -1;
+    }
+    trie->index2Length=newTop;
+    uprv_memcpy(trie->index2+newBlock, trie->index2+trie->index2NullOffset, UTRIE2_INDEX_2_BLOCK_LENGTH*4);
+    return newBlock;
+}
+
+static int32_t
+getIndex2Block(UNewTrie2 *trie, UChar32 c, UBool forLSCP) {
+    int32_t i1, i2;
+
+    if(U_IS_LEAD(c) && forLSCP) {
+        return UTRIE2_LSCP_INDEX_2_OFFSET;
+    }
+
+    i1=c>>UTRIE2_SHIFT_1;
+    i2=trie->index1[i1];
+    if(i2==trie->index2NullOffset) {
+        i2=allocIndex2Block(trie);
+        if(i2<0) {
+            return -1;  /* program error */
+        }
+        trie->index1[i1]=i2;
+    }
+    return i2;
+}
+
+static int32_t
+allocDataBlock(UNewTrie2 *trie, int32_t copyBlock) {
+    int32_t newBlock, newTop;
+
+    if(trie->firstFreeBlock!=0) {
+        /* get the first free block */
+        newBlock=trie->firstFreeBlock;
+        trie->firstFreeBlock=-trie->map[newBlock>>UTRIE2_SHIFT_2];
+    } else {
+        /* get a new block from the high end */
+        newBlock=trie->dataLength;
+        newTop=newBlock+UTRIE2_DATA_BLOCK_LENGTH;
+        if(newTop>trie->dataCapacity) {
+            /* out of memory in the data array */
+            int32_t capacity;
+            uint32_t *data;
+
+            if(trie->dataCapacity<UNEWTRIE2_MEDIUM_DATA_LENGTH) {
+                capacity=UNEWTRIE2_MEDIUM_DATA_LENGTH;
+            } else if(trie->dataCapacity<UNEWTRIE2_MAX_DATA_LENGTH) {
+                capacity=UNEWTRIE2_MAX_DATA_LENGTH;
+            } else {
+                /*
+                 * Should never occur.
+                 * Either UNEWTRIE2_MAX_DATA_LENGTH is incorrect,
+                 * or the code writes more values than should be possible.
+                 */
+                return -1;
+            }
+            data=(uint32_t *)uprv_malloc(capacity*4);
+            if(data==NULL) {
+                return -1;
+            }
+            uprv_memcpy(data, trie->data, (size_t)trie->dataLength*4);
+            uprv_free(trie->data);
+            trie->data=data;
+            trie->dataCapacity=capacity;
+        }
+        trie->dataLength=newTop;
+    }
+    uprv_memcpy(trie->data+newBlock, trie->data+copyBlock, UTRIE2_DATA_BLOCK_LENGTH*4);
+    trie->map[newBlock>>UTRIE2_SHIFT_2]=0;
+    return newBlock;
+}
+
+/* call when the block's reference counter reaches 0 */
+static void
+releaseDataBlock(UNewTrie2 *trie, int32_t block) {
+    /* put this block at the front of the free-block chain */
+    trie->map[block>>UTRIE2_SHIFT_2]=-trie->firstFreeBlock;
+    trie->firstFreeBlock=block;
+}
+
+static inline UBool
+isWritableBlock(UNewTrie2 *trie, int32_t block) {
+    return (UBool)(block!=trie->dataNullOffset && 1==trie->map[block>>UTRIE2_SHIFT_2]);
+}
+
+static inline void
+setIndex2Entry(UNewTrie2 *trie, int32_t i2, int32_t block) {
+    int32_t oldBlock;
+    ++trie->map[block>>UTRIE2_SHIFT_2];  /* increment first, in case block==oldBlock! */
+    oldBlock=trie->index2[i2];
+    if(0 == --trie->map[oldBlock>>UTRIE2_SHIFT_2]) {
+        releaseDataBlock(trie, oldBlock);
+    }
+    trie->index2[i2]=block;
+}
+
+/**
+ * No error checking for illegal arguments.
+ *
+ * @return -1 if no new data block available (out of memory in data array)
+ * @internal
+ */
+static int32_t
+getDataBlock(UNewTrie2 *trie, UChar32 c, UBool forLSCP) {
+    int32_t i2, oldBlock, newBlock;
+
+    i2=getIndex2Block(trie, c, forLSCP);
+    if(i2<0) {
+        return -1;  /* program error */
+    }
+
+    i2+=(c>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK;
+    oldBlock=trie->index2[i2];
+    if(isWritableBlock(trie, oldBlock)) {
+        return oldBlock;
+    }
+
+    /* allocate a new data block */
+    newBlock=allocDataBlock(trie, oldBlock);
+    if(newBlock<0) {
+        /* out of memory in the data array */
+        return -1;
+    }
+    setIndex2Entry(trie, i2, newBlock);
+    return newBlock;
+}
+
+/**
+ * @return TRUE if the value was successfully set
+ */
+static void
+set32(UNewTrie2 *trie,
+      UChar32 c, UBool forLSCP, uint32_t value,
+      UErrorCode *pErrorCode) {
+    int32_t block;
+
+    if(trie==NULL || trie->isCompacted) {
+        *pErrorCode=U_NO_WRITE_PERMISSION;
+        return;
+    }
+#ifdef UCPTRIE_DEBUG
+    umutablecptrie_set(trie->t3, c, value, pErrorCode);
+#endif
+
+    block=getDataBlock(trie, c, forLSCP);
+    if(block<0) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+
+    trie->data[block+(c&UTRIE2_DATA_MASK)]=value;
+}
+
+U_CAPI void U_EXPORT2
+utrie2_set32(UTrie2 *trie, UChar32 c, uint32_t value, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if((uint32_t)c>0x10ffff) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    set32(trie->newTrie, c, TRUE, value, pErrorCode);
+}
+
+U_CAPI void U_EXPORT2
+utrie2_set32ForLeadSurrogateCodeUnit(UTrie2 *trie,
+                                     UChar32 c, uint32_t value,
+                                     UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if(!U_IS_LEAD(c)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    set32(trie->newTrie, c, FALSE, value, pErrorCode);
+}
+
+static void
+writeBlock(uint32_t *block, uint32_t value) {
+    uint32_t *limit=block+UTRIE2_DATA_BLOCK_LENGTH;
+    while(block<limit) {
+        *block++=value;
+    }
+}
+
+/**
+ * initialValue is ignored if overwrite=TRUE
+ * @internal
+ */
+static void
+fillBlock(uint32_t *block, UChar32 start, UChar32 limit,
+          uint32_t value, uint32_t initialValue, UBool overwrite) {
+    uint32_t *pLimit;
+
+    pLimit=block+limit;
+    block+=start;
+    if(overwrite) {
+        while(block<pLimit) {
+            *block++=value;
+        }
+    } else {
+        while(block<pLimit) {
+            if(*block==initialValue) {
+                *block=value;
+            }
+            ++block;
+        }
+    }
+}
+
+U_CAPI void U_EXPORT2
+utrie2_setRange32(UTrie2 *trie,
+                  UChar32 start, UChar32 end,
+                  uint32_t value, UBool overwrite,
+                  UErrorCode *pErrorCode) {
+    /*
+     * repeat value in [start..end]
+     * mark index values for repeat-data blocks by setting bit 31 of the index values
+     * fill around existing values if any, if(overwrite)
+     */
+    UNewTrie2 *newTrie;
+    int32_t block, rest, repeatBlock;
+    UChar32 limit;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if((uint32_t)start>0x10ffff || (uint32_t)end>0x10ffff || start>end) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    newTrie=trie->newTrie;
+    if(newTrie==NULL || newTrie->isCompacted) {
+        *pErrorCode=U_NO_WRITE_PERMISSION;
+        return;
+    }
+#ifdef UCPTRIE_DEBUG
+    umutablecptrie_setRange(newTrie->t3, start, end, value, pErrorCode);
+#endif
+    if(!overwrite && value==newTrie->initialValue) {
+        return; /* nothing to do */
+    }
+
+    limit=end+1;
+    if(start&UTRIE2_DATA_MASK) {
+        UChar32 nextStart;
+
+        /* set partial block at [start..following block boundary[ */
+        block=getDataBlock(newTrie, start, TRUE);
+        if(block<0) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+
+        nextStart=(start+UTRIE2_DATA_MASK)&~UTRIE2_DATA_MASK;
+        if(nextStart<=limit) {
+            fillBlock(newTrie->data+block, start&UTRIE2_DATA_MASK, UTRIE2_DATA_BLOCK_LENGTH,
+                      value, newTrie->initialValue, overwrite);
+            start=nextStart;
+        } else {
+            fillBlock(newTrie->data+block, start&UTRIE2_DATA_MASK, limit&UTRIE2_DATA_MASK,
+                      value, newTrie->initialValue, overwrite);
+            return;
+        }
+    }
+
+    /* number of positions in the last, partial block */
+    rest=limit&UTRIE2_DATA_MASK;
+
+    /* round down limit to a block boundary */
+    limit&=~UTRIE2_DATA_MASK;
+
+    /* iterate over all-value blocks */
+    if(value==newTrie->initialValue) {
+        repeatBlock=newTrie->dataNullOffset;
+    } else {
+        repeatBlock=-1;
+    }
+
+    while(start<limit) {
+        int32_t i2;
+        UBool setRepeatBlock=FALSE;
+
+        if(value==newTrie->initialValue && isInNullBlock(newTrie, start, TRUE)) {
+            start+=UTRIE2_DATA_BLOCK_LENGTH; /* nothing to do */
+            continue;
+        }
+
+        /* get index value */
+        i2=getIndex2Block(newTrie, start, TRUE);
+        if(i2<0) {
+            *pErrorCode=U_INTERNAL_PROGRAM_ERROR;
+            return;
+        }
+        i2+=(start>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK;
+        block=newTrie->index2[i2];
+        if(isWritableBlock(newTrie, block)) {
+            /* already allocated */
+            if(overwrite && block>=UNEWTRIE2_DATA_0800_OFFSET) {
+                /*
+                 * We overwrite all values, and it's not a
+                 * protected (ASCII-linear or 2-byte UTF-8) block:
+                 * replace with the repeatBlock.
+                 */
+                setRepeatBlock=TRUE;
+            } else {
+                /* !overwrite, or protected block: just write the values into this block */
+                fillBlock(newTrie->data+block,
+                          0, UTRIE2_DATA_BLOCK_LENGTH,
+                          value, newTrie->initialValue, overwrite);
+            }
+        } else if(newTrie->data[block]!=value && (overwrite || block==newTrie->dataNullOffset)) {
+            /*
+             * Set the repeatBlock instead of the null block or previous repeat block:
+             *
+             * If !isWritableBlock() then all entries in the block have the same value
+             * because it's the null block or a range block (the repeatBlock from a previous
+             * call to utrie2_setRange32()).
+             * No other blocks are used multiple times before compacting.
+             *
+             * The null block is the only non-writable block with the initialValue because
+             * of the repeatBlock initialization above. (If value==initialValue, then
+             * the repeatBlock will be the null data block.)
+             *
+             * We set our repeatBlock if the desired value differs from the block's value,
+             * and if we overwrite any data or if the data is all initial values
+             * (which is the same as the block being the null block, see above).
+             */
+            setRepeatBlock=TRUE;
+        }
+        if(setRepeatBlock) {
+            if(repeatBlock>=0) {
+                setIndex2Entry(newTrie, i2, repeatBlock);
+            } else {
+                /* create and set and fill the repeatBlock */
+                repeatBlock=getDataBlock(newTrie, start, TRUE);
+                if(repeatBlock<0) {
+                    *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return;
+                }
+                writeBlock(newTrie->data+repeatBlock, value);
+            }
+        }
+
+        start+=UTRIE2_DATA_BLOCK_LENGTH;
+    }
+
+    if(rest>0) {
+        /* set partial block at [last block boundary..limit[ */
+        block=getDataBlock(newTrie, start, TRUE);
+        if(block<0) {
+            *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
+
+        fillBlock(newTrie->data+block, 0, rest, value, newTrie->initialValue, overwrite);
+    }
+
+    return;
+}
+
+/* compaction --------------------------------------------------------------- */
+
+static inline UBool
+equal_int32(const int32_t *s, const int32_t *t, int32_t length) {
+    while(length>0 && *s==*t) {
+        ++s;
+        ++t;
+        --length;
+    }
+    return (UBool)(length==0);
+}
+
+static inline UBool
+equal_uint32(const uint32_t *s, const uint32_t *t, int32_t length) {
+    while(length>0 && *s==*t) {
+        ++s;
+        ++t;
+        --length;
+    }
+    return (UBool)(length==0);
+}
+
+static int32_t
+findSameIndex2Block(const int32_t *idx, int32_t index2Length, int32_t otherBlock) {
+    int32_t block;
+
+    /* ensure that we do not even partially get past index2Length */
+    index2Length-=UTRIE2_INDEX_2_BLOCK_LENGTH;
+
+    for(block=0; block<=index2Length; ++block) {
+        if(equal_int32(idx+block, idx+otherBlock, UTRIE2_INDEX_2_BLOCK_LENGTH)) {
+            return block;
+        }
+    }
+    return -1;
+}
+
+static int32_t
+findSameDataBlock(const uint32_t *data, int32_t dataLength, int32_t otherBlock, int32_t blockLength) {
+    int32_t block;
+
+    /* ensure that we do not even partially get past dataLength */
+    dataLength-=blockLength;
+
+    for(block=0; block<=dataLength; block+=UTRIE2_DATA_GRANULARITY) {
+        if(equal_uint32(data+block, data+otherBlock, blockLength)) {
+            return block;
+        }
+    }
+    return -1;
+}
+
+/*
+ * Find the start of the last range in the trie by enumerating backward.
+ * Indexes for supplementary code points higher than this will be omitted.
+ */
+static UChar32
+findHighStart(UNewTrie2 *trie, uint32_t highValue) {
+    const uint32_t *data32;
+
+    uint32_t value, initialValue;
+    UChar32 c, prev;
+    int32_t i1, i2, j, i2Block, prevI2Block, index2NullOffset, block, prevBlock, nullBlock;
+
+    data32=trie->data;
+    initialValue=trie->initialValue;
+
+    index2NullOffset=trie->index2NullOffset;
+    nullBlock=trie->dataNullOffset;
+
+    /* set variables for previous range */
+    if(highValue==initialValue) {
+        prevI2Block=index2NullOffset;
+        prevBlock=nullBlock;
+    } else {
+        prevI2Block=-1;
+        prevBlock=-1;
+    }
+    prev=0x110000;
+
+    /* enumerate index-2 blocks */
+    i1=UNEWTRIE2_INDEX_1_LENGTH;
+    c=prev;
+    while(c>0) {
+        i2Block=trie->index1[--i1];
+        if(i2Block==prevI2Block) {
+            /* the index-2 block is the same as the previous one, and filled with highValue */
+            c-=UTRIE2_CP_PER_INDEX_1_ENTRY;
+            continue;
+        }
+        prevI2Block=i2Block;
+        if(i2Block==index2NullOffset) {
+            /* this is the null index-2 block */
+            if(highValue!=initialValue) {
+                return c;
+            }
+            c-=UTRIE2_CP_PER_INDEX_1_ENTRY;
+        } else {
+            /* enumerate data blocks for one index-2 block */
+            for(i2=UTRIE2_INDEX_2_BLOCK_LENGTH; i2>0;) {
+                block=trie->index2[i2Block+ --i2];
+                if(block==prevBlock) {
+                    /* the block is the same as the previous one, and filled with highValue */
+                    c-=UTRIE2_DATA_BLOCK_LENGTH;
+                    continue;
+                }
+                prevBlock=block;
+                if(block==nullBlock) {
+                    /* this is the null data block */
+                    if(highValue!=initialValue) {
+                        return c;
+                    }
+                    c-=UTRIE2_DATA_BLOCK_LENGTH;
+                } else {
+                    for(j=UTRIE2_DATA_BLOCK_LENGTH; j>0;) {
+                        value=data32[block+ --j];
+                        if(value!=highValue) {
+                            return c;
+                        }
+                        --c;
+                    }
+                }
+            }
+        }
+    }
+
+    /* deliver last range */
+    return 0;
+}
+
+/*
+ * Compact a build-time trie.
+ *
+ * The compaction
+ * - removes blocks that are identical with earlier ones
+ * - overlaps adjacent blocks as much as possible (if overlap==TRUE)
+ * - moves blocks in steps of the data granularity
+ * - moves and overlaps blocks that overlap with multiple values in the overlap region
+ *
+ * It does not
+ * - try to move and overlap blocks that are not already adjacent
+ */
+static void
+compactData(UNewTrie2 *trie) {
+#ifdef UTRIE2_DEBUG
+    int32_t countSame=0, sumOverlaps=0;
+#endif
+
+    int32_t start, newStart, movedStart;
+    int32_t blockLength, overlap;
+    int32_t i, mapIndex, blockCount;
+
+    /* do not compact linear-ASCII data */
+    newStart=UTRIE2_DATA_START_OFFSET;
+    for(start=0, i=0; start<newStart; start+=UTRIE2_DATA_BLOCK_LENGTH, ++i) {
+        trie->map[i]=start;
+    }
+
+    /*
+     * Start with a block length of 64 for 2-byte UTF-8,
+     * then switch to UTRIE2_DATA_BLOCK_LENGTH.
+     */
+    blockLength=64;
+    blockCount=blockLength>>UTRIE2_SHIFT_2;
+    for(start=newStart; start<trie->dataLength;) {
+        /*
+         * start: index of first entry of current block
+         * newStart: index where the current block is to be moved
+         *           (right after current end of already-compacted data)
+         */
+        if(start==UNEWTRIE2_DATA_0800_OFFSET) {
+            blockLength=UTRIE2_DATA_BLOCK_LENGTH;
+            blockCount=1;
+        }
+
+        /* skip blocks that are not used */
+        if(trie->map[start>>UTRIE2_SHIFT_2]<=0) {
+            /* advance start to the next block */
+            start+=blockLength;
+
+            /* leave newStart with the previous block! */
+            continue;
+        }
+
+        /* search for an identical block */
+        if( (movedStart=findSameDataBlock(trie->data, newStart, start, blockLength))
+             >=0
+        ) {
+#ifdef UTRIE2_DEBUG
+            ++countSame;
+#endif
+            /* found an identical block, set the other block's index value for the current block */
+            for(i=blockCount, mapIndex=start>>UTRIE2_SHIFT_2; i>0; --i) {
+                trie->map[mapIndex++]=movedStart;
+                movedStart+=UTRIE2_DATA_BLOCK_LENGTH;
+            }
+
+            /* advance start to the next block */
+            start+=blockLength;
+
+            /* leave newStart with the previous block! */
+            continue;
+        }
+
+        /* see if the beginning of this block can be overlapped with the end of the previous block */
+        /* look for maximum overlap (modulo granularity) with the previous, adjacent block */
+        for(overlap=blockLength-UTRIE2_DATA_GRANULARITY;
+            overlap>0 && !equal_uint32(trie->data+(newStart-overlap), trie->data+start, overlap);
+            overlap-=UTRIE2_DATA_GRANULARITY) {}
+
+#ifdef UTRIE2_DEBUG
+            sumOverlaps+=overlap;
+#endif
+        if(overlap>0 || newStart<start) {
+            /* some overlap, or just move the whole block */
+            movedStart=newStart-overlap;
+            for(i=blockCount, mapIndex=start>>UTRIE2_SHIFT_2; i>0; --i) {
+                trie->map[mapIndex++]=movedStart;
+                movedStart+=UTRIE2_DATA_BLOCK_LENGTH;
+            }
+
+            /* move the non-overlapping indexes to their new positions */
+            start+=overlap;
+            for(i=blockLength-overlap; i>0; --i) {
+                trie->data[newStart++]=trie->data[start++];
+            }
+        } else /* no overlap && newStart==start */ {
+            for(i=blockCount, mapIndex=start>>UTRIE2_SHIFT_2; i>0; --i) {
+                trie->map[mapIndex++]=start;
+                start+=UTRIE2_DATA_BLOCK_LENGTH;
+            }
+            newStart=start;
+        }
+    }
+
+    /* now adjust the index-2 table */
+    for(i=0; i<trie->index2Length; ++i) {
+        if(i==UNEWTRIE2_INDEX_GAP_OFFSET) {
+            /* Gap indexes are invalid (-1). Skip over the gap. */
+            i+=UNEWTRIE2_INDEX_GAP_LENGTH;
+        }
+        trie->index2[i]=trie->map[trie->index2[i]>>UTRIE2_SHIFT_2];
+    }
+    trie->dataNullOffset=trie->map[trie->dataNullOffset>>UTRIE2_SHIFT_2];
+
+    /* ensure dataLength alignment */
+    while((newStart&(UTRIE2_DATA_GRANULARITY-1))!=0) {
+        trie->data[newStart++]=trie->initialValue;
+    }
+
+#ifdef UTRIE2_DEBUG
+    /* we saved some space */
+    printf("compacting UTrie2: count of 32-bit data words %lu->%lu  countSame=%ld  sumOverlaps=%ld\n",
+            (long)trie->dataLength, (long)newStart, (long)countSame, (long)sumOverlaps);
+#endif
+
+    trie->dataLength=newStart;
+}
+
+static void
+compactIndex2(UNewTrie2 *trie) {
+    int32_t i, start, newStart, movedStart, overlap;
+
+    /* do not compact linear-BMP index-2 blocks */
+    newStart=UTRIE2_INDEX_2_BMP_LENGTH;
+    for(start=0, i=0; start<newStart; start+=UTRIE2_INDEX_2_BLOCK_LENGTH, ++i) {
+        trie->map[i]=start;
+    }
+
+    /* Reduce the index table gap to what will be needed at runtime. */
+    newStart+=UTRIE2_UTF8_2B_INDEX_2_LENGTH+((trie->highStart-0x10000)>>UTRIE2_SHIFT_1);
+
+    for(start=UNEWTRIE2_INDEX_2_NULL_OFFSET; start<trie->index2Length;) {
+        /*
+         * start: index of first entry of current block
+         * newStart: index where the current block is to be moved
+         *           (right after current end of already-compacted data)
+         */
+
+        /* search for an identical block */
+        if( (movedStart=findSameIndex2Block(trie->index2, newStart, start))
+             >=0
+        ) {
+            /* found an identical block, set the other block's index value for the current block */
+            trie->map[start>>UTRIE2_SHIFT_1_2]=movedStart;
+
+            /* advance start to the next block */
+            start+=UTRIE2_INDEX_2_BLOCK_LENGTH;
+
+            /* leave newStart with the previous block! */
+            continue;
+        }
+
+        /* see if the beginning of this block can be overlapped with the end of the previous block */
+        /* look for maximum overlap with the previous, adjacent block */
+        for(overlap=UTRIE2_INDEX_2_BLOCK_LENGTH-1;
+            overlap>0 && !equal_int32(trie->index2+(newStart-overlap), trie->index2+start, overlap);
+            --overlap) {}
+
+        if(overlap>0 || newStart<start) {
+            /* some overlap, or just move the whole block */
+            trie->map[start>>UTRIE2_SHIFT_1_2]=newStart-overlap;
+
+            /* move the non-overlapping indexes to their new positions */
+            start+=overlap;
+            for(i=UTRIE2_INDEX_2_BLOCK_LENGTH-overlap; i>0; --i) {
+                trie->index2[newStart++]=trie->index2[start++];
+            }
+        } else /* no overlap && newStart==start */ {
+            trie->map[start>>UTRIE2_SHIFT_1_2]=start;
+            start+=UTRIE2_INDEX_2_BLOCK_LENGTH;
+            newStart=start;
+        }
+    }
+
+    /* now adjust the index-1 table */
+    for(i=0; i<UNEWTRIE2_INDEX_1_LENGTH; ++i) {
+        trie->index1[i]=trie->map[trie->index1[i]>>UTRIE2_SHIFT_1_2];
+    }
+    trie->index2NullOffset=trie->map[trie->index2NullOffset>>UTRIE2_SHIFT_1_2];
+
+    /*
+     * Ensure data table alignment:
+     * Needs to be granularity-aligned for 16-bit trie
+     * (so that dataMove will be down-shiftable),
+     * and 2-aligned for uint32_t data.
+     */
+    while((newStart&((UTRIE2_DATA_GRANULARITY-1)|1))!=0) {
+        /* Arbitrary value: 0x3fffc not possible for real data. */
+        trie->index2[newStart++]=(int32_t)0xffff<<UTRIE2_INDEX_SHIFT;
+    }
+
+#ifdef UTRIE2_DEBUG
+    /* we saved some space */
+    printf("compacting UTrie2: count of 16-bit index words %lu->%lu\n",
+            (long)trie->index2Length, (long)newStart);
+#endif
+
+    trie->index2Length=newStart;
+}
+
+static void
+compactTrie(UTrie2 *trie, UErrorCode *pErrorCode) {
+    UNewTrie2 *newTrie;
+    UChar32 highStart, suppHighStart;
+    uint32_t highValue;
+
+    newTrie=trie->newTrie;
+
+    /* find highStart and round it up */
+    highValue=utrie2_get32(trie, 0x10ffff);
+    highStart=findHighStart(newTrie, highValue);
+    highStart=(highStart+(UTRIE2_CP_PER_INDEX_1_ENTRY-1))&~(UTRIE2_CP_PER_INDEX_1_ENTRY-1);
+    if(highStart==0x110000) {
+        highValue=trie->errorValue;
+    }
+
+    /*
+     * Set trie->highStart only after utrie2_get32(trie, highStart).
+     * Otherwise utrie2_get32(trie, highStart) would try to read the highValue.
+     */
+    trie->highStart=newTrie->highStart=highStart;
+
+#ifdef UTRIE2_DEBUG
+    printf("UTrie2: highStart U+%06lx  highValue 0x%lx  initialValue 0x%lx\n",
+            (long)highStart, (long)highValue, (long)trie->initialValue);
+#endif
+
+    if(highStart<0x110000) {
+        /* Blank out [highStart..10ffff] to release associated data blocks. */
+        suppHighStart= highStart<=0x10000 ? 0x10000 : highStart;
+        utrie2_setRange32(trie, suppHighStart, 0x10ffff, trie->initialValue, TRUE, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return;
+        }
+    }
+
+    compactData(newTrie);
+    if(highStart>0x10000) {
+        compactIndex2(newTrie);
+#ifdef UTRIE2_DEBUG
+    } else {
+        printf("UTrie2: highStart U+%04lx  count of 16-bit index words %lu->%lu\n",
+                (long)highStart, (long)trie->newTrie->index2Length, (long)UTRIE2_INDEX_1_OFFSET);
+#endif
+    }
+
+    /*
+     * Store the highValue in the data array and round up the dataLength.
+     * Must be done after compactData() because that assumes that dataLength
+     * is a multiple of UTRIE2_DATA_BLOCK_LENGTH.
+     */
+    newTrie->data[newTrie->dataLength++]=highValue;
+    while((newTrie->dataLength&(UTRIE2_DATA_GRANULARITY-1))!=0) {
+        newTrie->data[newTrie->dataLength++]=trie->initialValue;
+    }
+
+    newTrie->isCompacted=TRUE;
+}
+
+/* serialization ------------------------------------------------------------ */
+
+/**
+ * Maximum length of the runtime index array.
+ * Limited by its own 16-bit index values, and by uint16_t UTrie2Header.indexLength.
+ * (The actual maximum length is lower,
+ * (0x110000>>UTRIE2_SHIFT_2)+UTRIE2_UTF8_2B_INDEX_2_LENGTH+UTRIE2_MAX_INDEX_1_LENGTH.)
+ */
+#define UTRIE2_MAX_INDEX_LENGTH 0xffff
+
+/**
+ * Maximum length of the runtime data array.
+ * Limited by 16-bit index values that are left-shifted by UTRIE2_INDEX_SHIFT,
+ * and by uint16_t UTrie2Header.shiftedDataLength.
+ */
+#define UTRIE2_MAX_DATA_LENGTH (0xffff<<UTRIE2_INDEX_SHIFT)
+
+/* Compact and internally serialize the trie. */
+U_CAPI void U_EXPORT2
+utrie2_freeze(UTrie2 *trie, UTrie2ValueBits valueBits, UErrorCode *pErrorCode) {
+    UNewTrie2 *newTrie;
+    UTrie2Header *header;
+    uint32_t *p;
+    uint16_t *dest16;
+    int32_t i, length;
+    int32_t allIndexesLength;
+    int32_t dataMove;  /* >0 if the data is moved to the end of the index array */
+    UChar32 highStart;
+
+    /* argument check */
+    if(U_FAILURE(*pErrorCode)) {
+        return;
+    }
+    if( trie==NULL ||
+        valueBits<0 || UTRIE2_COUNT_VALUE_BITS<=valueBits
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    newTrie=trie->newTrie;
+    if(newTrie==NULL) {
+        /* already frozen */
+        UTrie2ValueBits frozenValueBits=
+            trie->data16!=NULL ? UTRIE2_16_VALUE_BITS : UTRIE2_32_VALUE_BITS;
+        if(valueBits!=frozenValueBits) {
+            *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        }
+        return;
+    }
+
+    /* compact if necessary */
+    if(!newTrie->isCompacted) {
+        compactTrie(trie, pErrorCode);
+        if(U_FAILURE(*pErrorCode)) {
+            return;
+        }
+    }
+    highStart=trie->highStart;
+
+    if(highStart<=0x10000) {
+        allIndexesLength=UTRIE2_INDEX_1_OFFSET;
+    } else {
+        allIndexesLength=newTrie->index2Length;
+    }
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        dataMove=allIndexesLength;
+    } else {
+        dataMove=0;
+    }
+
+    /* are indexLength and dataLength within limits? */
+    if( /* for unshifted indexLength */
+        allIndexesLength>UTRIE2_MAX_INDEX_LENGTH ||
+        /* for unshifted dataNullOffset */
+        (dataMove+newTrie->dataNullOffset)>0xffff ||
+        /* for unshifted 2-byte UTF-8 index-2 values */
+        (dataMove+UNEWTRIE2_DATA_0800_OFFSET)>0xffff ||
+        /* for shiftedDataLength */
+        (dataMove+newTrie->dataLength)>UTRIE2_MAX_DATA_LENGTH
+    ) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return;
+    }
+
+    /* calculate the total serialized length */
+    length=sizeof(UTrie2Header)+allIndexesLength*2;
+    if(valueBits==UTRIE2_16_VALUE_BITS) {
+        length+=newTrie->dataLength*2;
+    } else {
+        length+=newTrie->dataLength*4;
+    }
+
+    trie->memory=uprv_malloc(length);
+    if(trie->memory==NULL) {
+        *pErrorCode=U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    trie->length=length;
+    trie->isMemoryOwned=TRUE;
+
+    trie->indexLength=allIndexesLength;
+    trie->dataLength=newTrie->dataLength;
+    if(highStart<=0x10000) {
+        trie->index2NullOffset=0xffff;
+    } else {
+        trie->index2NullOffset=static_cast<uint16_t>(UTRIE2_INDEX_2_OFFSET+newTrie->index2NullOffset);
+    }
+    trie->dataNullOffset=(uint16_t)(dataMove+newTrie->dataNullOffset);
+    trie->highValueIndex=dataMove+trie->dataLength-UTRIE2_DATA_GRANULARITY;
+
+    /* set the header fields */
+    header=(UTrie2Header *)trie->memory;
+
+    header->signature=UTRIE2_SIG; /* "Tri2" */
+    header->options=(uint16_t)valueBits;
+
+    header->indexLength=(uint16_t)trie->indexLength;
+    header->shiftedDataLength=(uint16_t)(trie->dataLength>>UTRIE2_INDEX_SHIFT);
+    header->index2NullOffset=trie->index2NullOffset;
+    header->dataNullOffset=trie->dataNullOffset;
+    header->shiftedHighStart=(uint16_t)(highStart>>UTRIE2_SHIFT_1);
+
+    /* fill the index and data arrays */
+    dest16=(uint16_t *)(header+1);
+    trie->index=dest16;
+
+    /* write the index-2 array values shifted right by UTRIE2_INDEX_SHIFT, after adding dataMove */
+    p=(uint32_t *)newTrie->index2;
+    for(i=UTRIE2_INDEX_2_BMP_LENGTH; i>0; --i) {
+        *dest16++=(uint16_t)((dataMove + *p++)>>UTRIE2_INDEX_SHIFT);
+    }
+
+    /* write UTF-8 2-byte index-2 values, not right-shifted */
+    for(i=0; i<(0xc2-0xc0); ++i) {                                  /* C0..C1 */
+        *dest16++=(uint16_t)(dataMove+UTRIE2_BAD_UTF8_DATA_OFFSET);
+    }
+    for(; i<(0xe0-0xc0); ++i) {                                     /* C2..DF */
+        *dest16++=(uint16_t)(dataMove+newTrie->index2[i<<(6-UTRIE2_SHIFT_2)]);
+    }
+
+    if(highStart>0x10000) {
+        int32_t index1Length=(highStart-0x10000)>>UTRIE2_SHIFT_1;
+        int32_t index2Offset=UTRIE2_INDEX_2_BMP_LENGTH+UTRIE2_UTF8_2B_INDEX_2_LENGTH+index1Length;
+
+        /* write 16-bit index-1 values for supplementary code points */
+        p=(uint32_t *)newTrie->index1+UTRIE2_OMITTED_BMP_INDEX_1_LENGTH;
+        for(i=index1Length; i>0; --i) {
+            *dest16++=(uint16_t)(UTRIE2_INDEX_2_OFFSET + *p++);
+        }
+
+        /*
+         * write the index-2 array values for supplementary code points,
+         * shifted right by UTRIE2_INDEX_SHIFT, after adding dataMove
+         */
+        p=(uint32_t *)newTrie->index2+index2Offset;
+        for(i=newTrie->index2Length-index2Offset; i>0; --i) {
+            *dest16++=(uint16_t)((dataMove + *p++)>>UTRIE2_INDEX_SHIFT);
+        }
+    }
+
+    /* write the 16/32-bit data array */
+    switch(valueBits) {
+    case UTRIE2_16_VALUE_BITS:
+        /* write 16-bit data values */
+        trie->data16=dest16;
+        trie->data32=NULL;
+        p=newTrie->data;
+        for(i=newTrie->dataLength; i>0; --i) {
+            *dest16++=(uint16_t)*p++;
+        }
+        break;
+    case UTRIE2_32_VALUE_BITS:
+        /* write 32-bit data values */
+        trie->data16=NULL;
+        trie->data32=(uint32_t *)dest16;
+        uprv_memcpy(dest16, newTrie->data, (size_t)newTrie->dataLength*4);
+        break;
+    default:
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+
+#ifdef UTRIE2_DEBUG
+    utrie2_printLengths(trie, "");
+#endif
+
+#ifdef UCPTRIE_DEBUG
+    umutablecptrie_setName(newTrie->t3, trie->name);
+    ucptrie_close(
+        umutablecptrie_buildImmutable(
+            newTrie->t3, UCPTRIE_TYPE_FAST, (UCPTrieValueWidth)valueBits, pErrorCode));
+#endif
+    /* Delete the UNewTrie2. */
+    uprv_free(newTrie->data);
+    uprv_free(newTrie);
+    trie->newTrie=NULL;
+}
diff --git a/thirdparty/icu4c/common/utrie2_impl.h b/thirdparty/icu4c/common/utrie2_impl.h
new file mode 100644
index 0000000000..2a14db3a6b
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie2_impl.h
@@ -0,0 +1,175 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2001-2008, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*   file name:  utrie2_impl.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2008sep26 (split off from utrie2.c)
+*   created by: Markus W. Scherer
+*
+*   Definitions needed for both runtime and builder code for UTrie2,
+*   used by utrie2.c and utrie2_builder.c.
+*/
+
+#ifndef __UTRIE2_IMPL_H__
+#define __UTRIE2_IMPL_H__
+
+#ifdef UCPTRIE_DEBUG
+#include "unicode/umutablecptrie.h"
+#endif
+#include "utrie2.h"
+
+/* Public UTrie2 API implementation ----------------------------------------- */
+
+/*
+ * These definitions are mostly needed by utrie2.cpp,
+ * but also by utrie2_serialize() and utrie2_swap().
+ */
+
+// UTrie2 signature values, in platform endianness and opposite endianness.
+// The UTrie2 signature ASCII byte values spell "Tri2".
+#define UTRIE2_SIG      0x54726932
+#define UTRIE2_OE_SIG   0x32697254
+
+/**
+ * Trie data structure in serialized form:
+ *
+ * UTrie2Header header;
+ * uint16_t index[header.index2Length];
+ * uint16_t data[header.shiftedDataLength<<2];  -- or uint32_t data[...]
+ * @internal
+ */
+typedef struct UTrie2Header {
+    /** "Tri2" in big-endian US-ASCII (0x54726932) */
+    uint32_t signature;
+
+    /**
+     * options bit field:
+     * 15.. 4   reserved (0)
+     *  3.. 0   UTrie2ValueBits valueBits
+     */
+    uint16_t options;
+
+    /** UTRIE2_INDEX_1_OFFSET..UTRIE2_MAX_INDEX_LENGTH */
+    uint16_t indexLength;
+
+    /** (UTRIE2_DATA_START_OFFSET..UTRIE2_MAX_DATA_LENGTH)>>UTRIE2_INDEX_SHIFT */
+    uint16_t shiftedDataLength;
+
+    /** Null index and data blocks, not shifted. */
+    uint16_t index2NullOffset, dataNullOffset;
+
+    /**
+     * First code point of the single-value range ending with U+10ffff,
+     * rounded up and then shifted right by UTRIE2_SHIFT_1.
+     */
+    uint16_t shiftedHighStart;
+} UTrie2Header;
+
+/**
+ * Constants for use with UTrie2Header.options.
+ * @internal
+ */
+enum {
+    /** Mask to get the UTrie2ValueBits valueBits from options. */
+    UTRIE2_OPTIONS_VALUE_BITS_MASK=0xf
+};
+
+/* Building a trie ---------------------------------------------------------- */
+
+/*
+ * These definitions are mostly needed by utrie2_builder.c, but also by
+ * utrie2_get32() and utrie2_enum().
+ */
+
+enum {
+    /**
+     * At build time, leave a gap in the index-2 table,
+     * at least as long as the maximum lengths of the 2-byte UTF-8 index-2 table
+     * and the supplementary index-1 table.
+     * Round up to UTRIE2_INDEX_2_BLOCK_LENGTH for proper compacting.
+     */
+    UNEWTRIE2_INDEX_GAP_OFFSET=UTRIE2_INDEX_2_BMP_LENGTH,
+    UNEWTRIE2_INDEX_GAP_LENGTH=
+        ((UTRIE2_UTF8_2B_INDEX_2_LENGTH+UTRIE2_MAX_INDEX_1_LENGTH)+UTRIE2_INDEX_2_MASK)&
+        ~UTRIE2_INDEX_2_MASK,
+
+    /**
+     * Maximum length of the build-time index-2 array.
+     * Maximum number of Unicode code points (0x110000) shifted right by UTRIE2_SHIFT_2,
+     * plus the part of the index-2 table for lead surrogate code points,
+     * plus the build-time index gap,
+     * plus the null index-2 block.
+     */
+    UNEWTRIE2_MAX_INDEX_2_LENGTH=
+        (0x110000>>UTRIE2_SHIFT_2)+
+        UTRIE2_LSCP_INDEX_2_LENGTH+
+        UNEWTRIE2_INDEX_GAP_LENGTH+
+        UTRIE2_INDEX_2_BLOCK_LENGTH,
+
+    UNEWTRIE2_INDEX_1_LENGTH=0x110000>>UTRIE2_SHIFT_1
+};
+
+/**
+ * Maximum length of the build-time data array.
+ * One entry per 0x110000 code points, plus the illegal-UTF-8 block and the null block,
+ * plus values for the 0x400 surrogate code units.
+ */
+#define UNEWTRIE2_MAX_DATA_LENGTH (0x110000+0x40+0x40+0x400)
+
+/*
+ * Build-time trie structure.
+ *
+ * Just using a boolean flag for "repeat use" could lead to data array overflow
+ * because we would not be able to detect when a data block becomes unused.
+ * It also leads to orphan data blocks that are kept through serialization.
+ *
+ * Need to use reference counting for data blocks,
+ * and allocDataBlock() needs to look for a free block before increasing dataLength.
+ *
+ * This scheme seems like overkill for index-2 blocks since the whole index array is
+ * preallocated anyway (unlike the growable data array).
+ * Just allocating multiple index-2 blocks as needed.
+ */
+struct UNewTrie2 {
+    int32_t index1[UNEWTRIE2_INDEX_1_LENGTH];
+    int32_t index2[UNEWTRIE2_MAX_INDEX_2_LENGTH];
+    uint32_t *data;
+#ifdef UCPTRIE_DEBUG
+    UMutableCPTrie *t3;
+#endif
+
+    uint32_t initialValue, errorValue;
+    int32_t index2Length, dataCapacity, dataLength;
+    int32_t firstFreeBlock;
+    int32_t index2NullOffset, dataNullOffset;
+    UChar32 highStart;
+    UBool isCompacted;
+
+    /**
+     * Multi-purpose per-data-block table.
+     *
+     * Before compacting:
+     *
+     * Per-data-block reference counters/free-block list.
+     *  0: unused
+     * >0: reference counter (number of index-2 entries pointing here)
+     * <0: next free data block in free-block list
+     *
+     * While compacting:
+     *
+     * Map of adjusted indexes, used in compactData() and compactIndex2().
+     * Maps from original indexes to new ones.
+     */
+    int32_t map[UNEWTRIE2_MAX_DATA_LENGTH>>UTRIE2_SHIFT_2];
+};
+
+#endif
diff --git a/thirdparty/icu4c/common/utrie_swap.cpp b/thirdparty/icu4c/common/utrie_swap.cpp
new file mode 100644
index 0000000000..6e8b138394
--- /dev/null
+++ b/thirdparty/icu4c/common/utrie_swap.cpp
@@ -0,0 +1,348 @@
+// © 2018 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+
+// utrie_swap.cpp
+// created: 2018aug08 Markus W. Scherer
+
+#include "unicode/utypes.h"
+#include "cmemory.h"
+#include "ucptrie_impl.h"
+#include "udataswp.h"
+#include "utrie.h"
+#include "utrie2_impl.h"
+
+// These functions for swapping different generations of ICU code point tries are here
+// so that their implementation files need not depend on swapper code,
+// need not depend on each other, and so that other swapper code
+// need not depend on other trie code.
+
+namespace {
+
+constexpr int32_t ASCII_LIMIT = 0x80;
+
+}  // namespace
+
+U_CAPI int32_t U_EXPORT2
+utrie_swap(const UDataSwapper *ds,
+           const void *inData, int32_t length, void *outData,
+           UErrorCode *pErrorCode) {
+    const UTrieHeader *inTrie;
+    UTrieHeader trie;
+    int32_t size;
+    UBool dataIs32;
+
+    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || (length>=0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    if(length>=0 && (uint32_t)length<sizeof(UTrieHeader)) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    inTrie=(const UTrieHeader *)inData;
+    trie.signature=ds->readUInt32(inTrie->signature);
+    trie.options=ds->readUInt32(inTrie->options);
+    trie.indexLength=udata_readInt32(ds, inTrie->indexLength);
+    trie.dataLength=udata_readInt32(ds, inTrie->dataLength);
+
+    if( trie.signature!=0x54726965 ||
+        (trie.options&UTRIE_OPTIONS_SHIFT_MASK)!=UTRIE_SHIFT ||
+        ((trie.options>>UTRIE_OPTIONS_INDEX_SHIFT)&UTRIE_OPTIONS_SHIFT_MASK)!=UTRIE_INDEX_SHIFT ||
+        trie.indexLength<UTRIE_BMP_INDEX_LENGTH ||
+        (trie.indexLength&(UTRIE_SURROGATE_BLOCK_COUNT-1))!=0 ||
+        trie.dataLength<UTRIE_DATA_BLOCK_LENGTH ||
+        (trie.dataLength&(UTRIE_DATA_GRANULARITY-1))!=0 ||
+        ((trie.options&UTRIE_OPTIONS_LATIN1_IS_LINEAR)!=0 && trie.dataLength<(UTRIE_DATA_BLOCK_LENGTH+0x100))
+    ) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR; /* not a UTrie */
+        return 0;
+    }
+
+    dataIs32=(UBool)((trie.options&UTRIE_OPTIONS_DATA_IS_32_BIT)!=0);
+    size=sizeof(UTrieHeader)+trie.indexLength*2+trie.dataLength*(dataIs32?4:2);
+
+    if(length>=0) {
+        UTrieHeader *outTrie;
+
+        if(length<size) {
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        outTrie=(UTrieHeader *)outData;
+
+        /* swap the header */
+        ds->swapArray32(ds, inTrie, sizeof(UTrieHeader), outTrie, pErrorCode);
+
+        /* swap the index and the data */
+        if(dataIs32) {
+            ds->swapArray16(ds, inTrie+1, trie.indexLength*2, outTrie+1, pErrorCode);
+            ds->swapArray32(ds, (const uint16_t *)(inTrie+1)+trie.indexLength, trie.dataLength*4,
+                                     (uint16_t *)(outTrie+1)+trie.indexLength, pErrorCode);
+        } else {
+            ds->swapArray16(ds, inTrie+1, (trie.indexLength+trie.dataLength)*2, outTrie+1, pErrorCode);
+        }
+    }
+
+    return size;
+}
+
+U_CAPI int32_t U_EXPORT2
+utrie2_swap(const UDataSwapper *ds,
+            const void *inData, int32_t length, void *outData,
+            UErrorCode *pErrorCode) {
+    const UTrie2Header *inTrie;
+    UTrie2Header trie;
+    int32_t dataLength, size;
+    UTrie2ValueBits valueBits;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==NULL || inData==NULL || (length>=0 && outData==NULL)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    if(length>=0 && length<(int32_t)sizeof(UTrie2Header)) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    inTrie=(const UTrie2Header *)inData;
+    trie.signature=ds->readUInt32(inTrie->signature);
+    trie.options=ds->readUInt16(inTrie->options);
+    trie.indexLength=ds->readUInt16(inTrie->indexLength);
+    trie.shiftedDataLength=ds->readUInt16(inTrie->shiftedDataLength);
+
+    valueBits=(UTrie2ValueBits)(trie.options&UTRIE2_OPTIONS_VALUE_BITS_MASK);
+    dataLength=(int32_t)trie.shiftedDataLength<<UTRIE2_INDEX_SHIFT;
+
+    if( trie.signature!=UTRIE2_SIG ||
+        valueBits<0 || UTRIE2_COUNT_VALUE_BITS<=valueBits ||
+        trie.indexLength<UTRIE2_INDEX_1_OFFSET ||
+        dataLength<UTRIE2_DATA_START_OFFSET
+    ) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR; /* not a UTrie */
+        return 0;
+    }
+
+    size=sizeof(UTrie2Header)+trie.indexLength*2;
+    switch(valueBits) {
+    case UTRIE2_16_VALUE_BITS:
+        size+=dataLength*2;
+        break;
+    case UTRIE2_32_VALUE_BITS:
+        size+=dataLength*4;
+        break;
+    default:
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    if(length>=0) {
+        UTrie2Header *outTrie;
+
+        if(length<size) {
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        outTrie=(UTrie2Header *)outData;
+
+        /* swap the header */
+        ds->swapArray32(ds, &inTrie->signature, 4, &outTrie->signature, pErrorCode);
+        ds->swapArray16(ds, &inTrie->options, 12, &outTrie->options, pErrorCode);
+
+        /* swap the index and the data */
+        switch(valueBits) {
+        case UTRIE2_16_VALUE_BITS:
+            ds->swapArray16(ds, inTrie+1, (trie.indexLength+dataLength)*2, outTrie+1, pErrorCode);
+            break;
+        case UTRIE2_32_VALUE_BITS:
+            ds->swapArray16(ds, inTrie+1, trie.indexLength*2, outTrie+1, pErrorCode);
+            ds->swapArray32(ds, (const uint16_t *)(inTrie+1)+trie.indexLength, dataLength*4,
+                                     (uint16_t *)(outTrie+1)+trie.indexLength, pErrorCode);
+            break;
+        default:
+            *pErrorCode=U_INVALID_FORMAT_ERROR;
+            return 0;
+        }
+    }
+
+    return size;
+}
+
+U_CAPI int32_t U_EXPORT2
+ucptrie_swap(const UDataSwapper *ds,
+             const void *inData, int32_t length, void *outData,
+             UErrorCode *pErrorCode) {
+    const UCPTrieHeader *inTrie;
+    UCPTrieHeader trie;
+    int32_t dataLength, size;
+    UCPTrieValueWidth valueWidth;
+
+    if(U_FAILURE(*pErrorCode)) {
+        return 0;
+    }
+    if(ds==nullptr || inData==nullptr || (length>=0 && outData==nullptr)) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return 0;
+    }
+
+    /* setup and swapping */
+    if(length>=0 && length<(int32_t)sizeof(UCPTrieHeader)) {
+        *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+        return 0;
+    }
+
+    inTrie=(const UCPTrieHeader *)inData;
+    trie.signature=ds->readUInt32(inTrie->signature);
+    trie.options=ds->readUInt16(inTrie->options);
+    trie.indexLength=ds->readUInt16(inTrie->indexLength);
+    trie.dataLength = ds->readUInt16(inTrie->dataLength);
+
+    UCPTrieType type = (UCPTrieType)((trie.options >> 6) & 3);
+    valueWidth = (UCPTrieValueWidth)(trie.options & UCPTRIE_OPTIONS_VALUE_BITS_MASK);
+    dataLength = ((int32_t)(trie.options & UCPTRIE_OPTIONS_DATA_LENGTH_MASK) << 4) | trie.dataLength;
+
+    int32_t minIndexLength = type == UCPTRIE_TYPE_FAST ?
+        UCPTRIE_BMP_INDEX_LENGTH : UCPTRIE_SMALL_INDEX_LENGTH;
+    if( trie.signature!=UCPTRIE_SIG ||
+        type > UCPTRIE_TYPE_SMALL ||
+        (trie.options & UCPTRIE_OPTIONS_RESERVED_MASK) != 0 ||
+        valueWidth > UCPTRIE_VALUE_BITS_8 ||
+        trie.indexLength < minIndexLength ||
+        dataLength < ASCII_LIMIT
+    ) {
+        *pErrorCode=U_INVALID_FORMAT_ERROR; /* not a UCPTrie */
+        return 0;
+    }
+
+    size=sizeof(UCPTrieHeader)+trie.indexLength*2;
+    switch(valueWidth) {
+    case UCPTRIE_VALUE_BITS_16:
+        size+=dataLength*2;
+        break;
+    case UCPTRIE_VALUE_BITS_32:
+        size+=dataLength*4;
+        break;
+    case UCPTRIE_VALUE_BITS_8:
+        size+=dataLength;
+        break;
+    default:
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+
+    if(length>=0) {
+        UCPTrieHeader *outTrie;
+
+        if(length<size) {
+            *pErrorCode=U_INDEX_OUTOFBOUNDS_ERROR;
+            return 0;
+        }
+
+        outTrie=(UCPTrieHeader *)outData;
+
+        /* swap the header */
+        ds->swapArray32(ds, &inTrie->signature, 4, &outTrie->signature, pErrorCode);
+        ds->swapArray16(ds, &inTrie->options, 12, &outTrie->options, pErrorCode);
+
+        /* swap the index */
+        const uint16_t *inIndex=reinterpret_cast<const uint16_t *>(inTrie+1);
+        uint16_t *outIndex=reinterpret_cast<uint16_t *>(outTrie+1);
+        ds->swapArray16(ds, inIndex, trie.indexLength*2, outIndex, pErrorCode);
+
+        /* swap the data */
+        const uint16_t *inData=inIndex+trie.indexLength;
+        uint16_t *outData=outIndex+trie.indexLength;
+        switch(valueWidth) {
+        case UCPTRIE_VALUE_BITS_16:
+            ds->swapArray16(ds, inData, dataLength*2, outData, pErrorCode);
+            break;
+        case UCPTRIE_VALUE_BITS_32:
+            ds->swapArray32(ds, inData, dataLength*4, outData, pErrorCode);
+            break;
+        case UCPTRIE_VALUE_BITS_8:
+            if(inTrie!=outTrie) {
+                uprv_memmove(outData, inData, dataLength);
+            }
+            break;
+        default:
+            *pErrorCode=U_INVALID_FORMAT_ERROR;
+            return 0;
+        }
+    }
+
+    return size;
+}
+
+namespace {
+
+/**
+ * Gets the trie version from 32-bit-aligned memory containing the serialized form
+ * of a UTrie (version 1), a UTrie2 (version 2), or a UCPTrie (version 3).
+ *
+ * @param data a pointer to 32-bit-aligned memory containing the serialized form of a trie
+ * @param length the number of bytes available at data;
+ *               can be more than necessary (see return value)
+ * @param anyEndianOk If FALSE, only platform-endian serialized forms are recognized.
+ *                    If TRUE, opposite-endian serialized forms are recognized as well.
+ * @return the trie version of the serialized form, or 0 if it is not
+ *         recognized as a serialized trie
+ */
+int32_t
+getVersion(const void *data, int32_t length, UBool anyEndianOk) {
+    uint32_t signature;
+    if(length<16 || data==nullptr || (U_POINTER_MASK_LSB(data, 3)!=0)) {
+        return 0;
+    }
+    signature=*(const uint32_t *)data;
+    if(signature==UCPTRIE_SIG) {
+        return 3;
+    }
+    if(anyEndianOk && signature==UCPTRIE_OE_SIG) {
+        return 3;
+    }
+    if(signature==UTRIE2_SIG) {
+        return 2;
+    }
+    if(anyEndianOk && signature==UTRIE2_OE_SIG) {
+        return 2;
+    }
+    if(signature==UTRIE_SIG) {
+        return 1;
+    }
+    if(anyEndianOk && signature==UTRIE_OE_SIG) {
+        return 1;
+    }
+    return 0;
+}
+
+}  // namespace
+
+U_CAPI int32_t U_EXPORT2
+utrie_swapAnyVersion(const UDataSwapper *ds,
+                     const void *inData, int32_t length, void *outData,
+                     UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) { return 0; }
+    switch(getVersion(inData, length, TRUE)) {
+    case 1:
+        return utrie_swap(ds, inData, length, outData, pErrorCode);
+    case 2:
+        return utrie2_swap(ds, inData, length, outData, pErrorCode);
+    case 3:
+        return ucptrie_swap(ds, inData, length, outData, pErrorCode);
+    default:
+        *pErrorCode=U_INVALID_FORMAT_ERROR;
+        return 0;
+    }
+}
diff --git a/thirdparty/icu4c/common/uts46.cpp b/thirdparty/icu4c/common/uts46.cpp
new file mode 100644
index 0000000000..f25b4e12f1
--- /dev/null
+++ b/thirdparty/icu4c/common/uts46.cpp
@@ -0,0 +1,1494 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2010-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*   file name:  uts46.cpp
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 2010mar09
+*   created by: Markus W. Scherer
+*/
+
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_IDNA
+
+#include "unicode/idna.h"
+#include "unicode/normalizer2.h"
+#include "unicode/uscript.h"
+#include "unicode/ustring.h"
+#include "unicode/utf16.h"
+#include "cmemory.h"
+#include "cstring.h"
+#include "punycode.h"
+#include "ubidi_props.h"
+#include "ustr_imp.h"
+
+// Note about tests for UIDNA_ERROR_DOMAIN_NAME_TOO_LONG:
+//
+// The domain name length limit is 255 octets in an internal DNS representation
+// where the last ("root") label is the empty label
+// represented by length byte 0 alone.
+// In a conventional string, this translates to 253 characters, or 254
+// if there is a trailing dot for the root label.
+
+U_NAMESPACE_BEGIN
+
+// Severe errors which usually result in a U+FFFD replacement character in the result string.
+const uint32_t severeErrors=
+    UIDNA_ERROR_LEADING_COMBINING_MARK|
+    UIDNA_ERROR_DISALLOWED|
+    UIDNA_ERROR_PUNYCODE|
+    UIDNA_ERROR_LABEL_HAS_DOT|
+    UIDNA_ERROR_INVALID_ACE_LABEL;
+
+static inline UBool
+isASCIIString(const UnicodeString &dest) {
+    const UChar *s=dest.getBuffer();
+    const UChar *limit=s+dest.length();
+    while(s<limit) {
+        if(*s++>0x7f) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+static UBool
+isASCIIOkBiDi(const UChar *s, int32_t length);
+
+static UBool
+isASCIIOkBiDi(const char *s, int32_t length);
+
+// IDNA class default implementations -------------------------------------- ***
+
+IDNA::~IDNA() {}
+
+void
+IDNA::labelToASCII_UTF8(StringPiece label, ByteSink &dest,
+                        IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_SUCCESS(errorCode)) {
+        UnicodeString destString;
+        labelToASCII(UnicodeString::fromUTF8(label), destString,
+                     info, errorCode).toUTF8(dest);
+    }
+}
+
+void
+IDNA::labelToUnicodeUTF8(StringPiece label, ByteSink &dest,
+                         IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_SUCCESS(errorCode)) {
+        UnicodeString destString;
+        labelToUnicode(UnicodeString::fromUTF8(label), destString,
+                       info, errorCode).toUTF8(dest);
+    }
+}
+
+void
+IDNA::nameToASCII_UTF8(StringPiece name, ByteSink &dest,
+                       IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_SUCCESS(errorCode)) {
+        UnicodeString destString;
+        nameToASCII(UnicodeString::fromUTF8(name), destString,
+                    info, errorCode).toUTF8(dest);
+    }
+}
+
+void
+IDNA::nameToUnicodeUTF8(StringPiece name, ByteSink &dest,
+                        IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_SUCCESS(errorCode)) {
+        UnicodeString destString;
+        nameToUnicode(UnicodeString::fromUTF8(name), destString,
+                      info, errorCode).toUTF8(dest);
+    }
+}
+
+// UTS46 class declaration ------------------------------------------------- ***
+
+class UTS46 : public IDNA {
+public:
+    UTS46(uint32_t options, UErrorCode &errorCode);
+    virtual ~UTS46();
+
+    virtual UnicodeString &
+    labelToASCII(const UnicodeString &label, UnicodeString &dest,
+                 IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual UnicodeString &
+    labelToUnicode(const UnicodeString &label, UnicodeString &dest,
+                   IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual UnicodeString &
+    nameToASCII(const UnicodeString &name, UnicodeString &dest,
+                IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual UnicodeString &
+    nameToUnicode(const UnicodeString &name, UnicodeString &dest,
+                  IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual void
+    labelToASCII_UTF8(StringPiece label, ByteSink &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual void
+    labelToUnicodeUTF8(StringPiece label, ByteSink &dest,
+                       IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual void
+    nameToASCII_UTF8(StringPiece name, ByteSink &dest,
+                     IDNAInfo &info, UErrorCode &errorCode) const;
+
+    virtual void
+    nameToUnicodeUTF8(StringPiece name, ByteSink &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const;
+
+private:
+    UnicodeString &
+    process(const UnicodeString &src,
+            UBool isLabel, UBool toASCII,
+            UnicodeString &dest,
+            IDNAInfo &info, UErrorCode &errorCode) const;
+
+    void
+    processUTF8(StringPiece src,
+                UBool isLabel, UBool toASCII,
+                ByteSink &dest,
+                IDNAInfo &info, UErrorCode &errorCode) const;
+
+    UnicodeString &
+    processUnicode(const UnicodeString &src,
+                   int32_t labelStart, int32_t mappingStart,
+                   UBool isLabel, UBool toASCII,
+                   UnicodeString &dest,
+                   IDNAInfo &info, UErrorCode &errorCode) const;
+
+    // returns the new dest.length()
+    int32_t
+    mapDevChars(UnicodeString &dest, int32_t labelStart, int32_t mappingStart,
+                UErrorCode &errorCode) const;
+
+    // returns the new label length
+    int32_t
+    processLabel(UnicodeString &dest,
+                 int32_t labelStart, int32_t labelLength,
+                 UBool toASCII,
+                 IDNAInfo &info, UErrorCode &errorCode) const;
+    int32_t
+    markBadACELabel(UnicodeString &dest,
+                    int32_t labelStart, int32_t labelLength,
+                    UBool toASCII, IDNAInfo &info, UErrorCode &errorCode) const;
+
+    void
+    checkLabelBiDi(const UChar *label, int32_t labelLength, IDNAInfo &info) const;
+
+    UBool
+    isLabelOkContextJ(const UChar *label, int32_t labelLength) const;
+
+    void
+    checkLabelContextO(const UChar *label, int32_t labelLength, IDNAInfo &info) const;
+
+    const Normalizer2 &uts46Norm2;  // uts46.nrm
+    uint32_t options;
+};
+
+IDNA *
+IDNA::createUTS46Instance(uint32_t options, UErrorCode &errorCode) {
+    if(U_SUCCESS(errorCode)) {
+        IDNA *idna=new UTS46(options, errorCode);
+        if(idna==NULL) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+        } else if(U_FAILURE(errorCode)) {
+            delete idna;
+            idna=NULL;
+        }
+        return idna;
+    } else {
+        return NULL;
+    }
+}
+
+// UTS46 implementation ---------------------------------------------------- ***
+
+UTS46::UTS46(uint32_t opt, UErrorCode &errorCode)
+        : uts46Norm2(*Normalizer2::getInstance(NULL, "uts46", UNORM2_COMPOSE, errorCode)),
+          options(opt) {}
+
+UTS46::~UTS46() {}
+
+UnicodeString &
+UTS46::labelToASCII(const UnicodeString &label, UnicodeString &dest,
+                    IDNAInfo &info, UErrorCode &errorCode) const {
+    return process(label, TRUE, TRUE, dest, info, errorCode);
+}
+
+UnicodeString &
+UTS46::labelToUnicode(const UnicodeString &label, UnicodeString &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const {
+    return process(label, TRUE, FALSE, dest, info, errorCode);
+}
+
+UnicodeString &
+UTS46::nameToASCII(const UnicodeString &name, UnicodeString &dest,
+                   IDNAInfo &info, UErrorCode &errorCode) const {
+    process(name, FALSE, TRUE, dest, info, errorCode);
+    if( dest.length()>=254 && (info.errors&UIDNA_ERROR_DOMAIN_NAME_TOO_LONG)==0 &&
+        isASCIIString(dest) &&
+        (dest.length()>254 || dest[253]!=0x2e)
+    ) {
+        info.errors|=UIDNA_ERROR_DOMAIN_NAME_TOO_LONG;
+    }
+    return dest;
+}
+
+UnicodeString &
+UTS46::nameToUnicode(const UnicodeString &name, UnicodeString &dest,
+                     IDNAInfo &info, UErrorCode &errorCode) const {
+    return process(name, FALSE, FALSE, dest, info, errorCode);
+}
+
+void
+UTS46::labelToASCII_UTF8(StringPiece label, ByteSink &dest,
+                         IDNAInfo &info, UErrorCode &errorCode) const {
+    processUTF8(label, TRUE, TRUE, dest, info, errorCode);
+}
+
+void
+UTS46::labelToUnicodeUTF8(StringPiece label, ByteSink &dest,
+                          IDNAInfo &info, UErrorCode &errorCode) const {
+    processUTF8(label, TRUE, FALSE, dest, info, errorCode);
+}
+
+void
+UTS46::nameToASCII_UTF8(StringPiece name, ByteSink &dest,
+                        IDNAInfo &info, UErrorCode &errorCode) const {
+    processUTF8(name, FALSE, TRUE, dest, info, errorCode);
+}
+
+void
+UTS46::nameToUnicodeUTF8(StringPiece name, ByteSink &dest,
+                         IDNAInfo &info, UErrorCode &errorCode) const {
+    processUTF8(name, FALSE, FALSE, dest, info, errorCode);
+}
+
+// UTS #46 data for ASCII characters.
+// The normalizer (using uts46.nrm) maps uppercase ASCII letters to lowercase
+// and passes through all other ASCII characters.
+// If UIDNA_USE_STD3_RULES is set, then non-LDH characters are disallowed
+// using this data.
+// The ASCII fastpath also uses this data.
+// Values: -1=disallowed  0==valid  1==mapped (lowercase)
+static const int8_t asciiData[128]={
+    -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
+    -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
+    // 002D..002E; valid  #  HYPHEN-MINUS..FULL STOP
+    -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,  0,  0, -1,
+    // 0030..0039; valid  #  DIGIT ZERO..DIGIT NINE
+     0,  0,  0,  0,  0,  0,  0,  0,  0,  0, -1, -1, -1, -1, -1, -1,
+    // 0041..005A; mapped  #  LATIN CAPITAL LETTER A..LATIN CAPITAL LETTER Z
+    -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+     1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1, -1, -1, -1, -1,
+    // 0061..007A; valid  #  LATIN SMALL LETTER A..LATIN SMALL LETTER Z
+    -1,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
+     0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, -1, -1, -1, -1, -1
+};
+
+UnicodeString &
+UTS46::process(const UnicodeString &src,
+               UBool isLabel, UBool toASCII,
+               UnicodeString &dest,
+               IDNAInfo &info, UErrorCode &errorCode) const {
+    // uts46Norm2.normalize() would do all of this error checking and setup,
+    // but with the ASCII fastpath we do not always call it, and do not
+    // call it first.
+    if(U_FAILURE(errorCode)) {
+        dest.setToBogus();
+        return dest;
+    }
+    const UChar *srcArray=src.getBuffer();
+    if(&dest==&src || srcArray==NULL) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        dest.setToBogus();
+        return dest;
+    }
+    // Arguments are fine, reset output values.
+    dest.remove();
+    info.reset();
+    int32_t srcLength=src.length();
+    if(srcLength==0) {
+        info.errors|=UIDNA_ERROR_EMPTY_LABEL;
+        return dest;
+    }
+    UChar *destArray=dest.getBuffer(srcLength);
+    if(destArray==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return dest;
+    }
+    // ASCII fastpath
+    UBool disallowNonLDHDot=(options&UIDNA_USE_STD3_RULES)!=0;
+    int32_t labelStart=0;
+    int32_t i;
+    for(i=0;; ++i) {
+        if(i==srcLength) {
+            if(toASCII) {
+                if((i-labelStart)>63) {
+                    info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                }
+                // There is a trailing dot if labelStart==i.
+                if(!isLabel && i>=254 && (i>254 || labelStart<i)) {
+                    info.errors|=UIDNA_ERROR_DOMAIN_NAME_TOO_LONG;
+                }
+            }
+            info.errors|=info.labelErrors;
+            dest.releaseBuffer(i);
+            return dest;
+        }
+        UChar c=srcArray[i];
+        if(c>0x7f) {
+            break;
+        }
+        int cData=asciiData[c];
+        if(cData>0) {
+            destArray[i]=c+0x20;  // Lowercase an uppercase ASCII letter.
+        } else if(cData<0 && disallowNonLDHDot) {
+            break;  // Replacing with U+FFFD can be complicated for toASCII.
+        } else {
+            destArray[i]=c;
+            if(c==0x2d) {  // hyphen
+                if(i==(labelStart+3) && srcArray[i-1]==0x2d) {
+                    // "??--..." is Punycode or forbidden.
+                    ++i;  // '-' was copied to dest already
+                    break;
+                }
+                if(i==labelStart) {
+                    // label starts with "-"
+                    info.labelErrors|=UIDNA_ERROR_LEADING_HYPHEN;
+                }
+                if((i+1)==srcLength || srcArray[i+1]==0x2e) {
+                    // label ends with "-"
+                    info.labelErrors|=UIDNA_ERROR_TRAILING_HYPHEN;
+                }
+            } else if(c==0x2e) {  // dot
+                if(isLabel) {
+                    // Replacing with U+FFFD can be complicated for toASCII.
+                    ++i;  // '.' was copied to dest already
+                    break;
+                }
+                if(i==labelStart) {
+                    info.labelErrors|=UIDNA_ERROR_EMPTY_LABEL;
+                }
+                if(toASCII && (i-labelStart)>63) {
+                    info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                }
+                info.errors|=info.labelErrors;
+                info.labelErrors=0;
+                labelStart=i+1;
+            }
+        }
+    }
+    info.errors|=info.labelErrors;
+    dest.releaseBuffer(i);
+    processUnicode(src, labelStart, i, isLabel, toASCII, dest, info, errorCode);
+    if( info.isBiDi && U_SUCCESS(errorCode) && (info.errors&severeErrors)==0 &&
+        (!info.isOkBiDi || (labelStart>0 && !isASCIIOkBiDi(dest.getBuffer(), labelStart)))
+    ) {
+        info.errors|=UIDNA_ERROR_BIDI;
+    }
+    return dest;
+}
+
+void
+UTS46::processUTF8(StringPiece src,
+                   UBool isLabel, UBool toASCII,
+                   ByteSink &dest,
+                   IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return;
+    }
+    const char *srcArray=src.data();
+    int32_t srcLength=src.length();
+    if(srcArray==NULL && srcLength!=0) {
+        errorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return;
+    }
+    // Arguments are fine, reset output values.
+    info.reset();
+    if(srcLength==0) {
+        info.errors|=UIDNA_ERROR_EMPTY_LABEL;
+        dest.Flush();
+        return;
+    }
+    UnicodeString destString;
+    int32_t labelStart=0;
+    if(srcLength<=256) {  // length of stackArray[]
+        // ASCII fastpath
+        char stackArray[256];
+        int32_t destCapacity;
+        char *destArray=dest.GetAppendBuffer(srcLength, srcLength+20,
+                                             stackArray, UPRV_LENGTHOF(stackArray), &destCapacity);
+        UBool disallowNonLDHDot=(options&UIDNA_USE_STD3_RULES)!=0;
+        int32_t i;
+        for(i=0;; ++i) {
+            if(i==srcLength) {
+                if(toASCII) {
+                    if((i-labelStart)>63) {
+                        info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                    }
+                    // There is a trailing dot if labelStart==i.
+                    if(!isLabel && i>=254 && (i>254 || labelStart<i)) {
+                        info.errors|=UIDNA_ERROR_DOMAIN_NAME_TOO_LONG;
+                    }
+                }
+                info.errors|=info.labelErrors;
+                dest.Append(destArray, i);
+                dest.Flush();
+                return;
+            }
+            char c=srcArray[i];
+            if((int8_t)c<0) {  // (uint8_t)c>0x7f
+                break;
+            }
+            int cData=asciiData[(int)c];  // Cast: gcc warns about indexing with a char.
+            if(cData>0) {
+                destArray[i]=c+0x20;  // Lowercase an uppercase ASCII letter.
+            } else if(cData<0 && disallowNonLDHDot) {
+                break;  // Replacing with U+FFFD can be complicated for toASCII.
+            } else {
+                destArray[i]=c;
+                if(c==0x2d) {  // hyphen
+                    if(i==(labelStart+3) && srcArray[i-1]==0x2d) {
+                        // "??--..." is Punycode or forbidden.
+                        break;
+                    }
+                    if(i==labelStart) {
+                        // label starts with "-"
+                        info.labelErrors|=UIDNA_ERROR_LEADING_HYPHEN;
+                    }
+                    if((i+1)==srcLength || srcArray[i+1]==0x2e) {
+                        // label ends with "-"
+                        info.labelErrors|=UIDNA_ERROR_TRAILING_HYPHEN;
+                    }
+                } else if(c==0x2e) {  // dot
+                    if(isLabel) {
+                        break;  // Replacing with U+FFFD can be complicated for toASCII.
+                    }
+                    if(i==labelStart) {
+                        info.labelErrors|=UIDNA_ERROR_EMPTY_LABEL;
+                    }
+                    if(toASCII && (i-labelStart)>63) {
+                        info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                    }
+                    info.errors|=info.labelErrors;
+                    info.labelErrors=0;
+                    labelStart=i+1;
+                }
+            }
+        }
+        info.errors|=info.labelErrors;
+        // Convert the processed ASCII prefix of the current label to UTF-16.
+        int32_t mappingStart=i-labelStart;
+        destString=UnicodeString::fromUTF8(StringPiece(destArray+labelStart, mappingStart));
+        // Output the previous ASCII labels and process the rest of src in UTF-16.
+        dest.Append(destArray, labelStart);
+        processUnicode(UnicodeString::fromUTF8(StringPiece(src, labelStart)), 0, mappingStart,
+                       isLabel, toASCII,
+                       destString, info, errorCode);
+    } else {
+        // src is too long for the ASCII fastpath implementation.
+        processUnicode(UnicodeString::fromUTF8(src), 0, 0,
+                       isLabel, toASCII,
+                       destString, info, errorCode);
+    }
+    destString.toUTF8(dest);  // calls dest.Flush()
+    if(toASCII && !isLabel) {
+        // length==labelStart==254 means that there is a trailing dot (ok) and
+        // destString is empty (do not index at 253-labelStart).
+        int32_t length=labelStart+destString.length();
+        if( length>=254 && isASCIIString(destString) &&
+            (length>254 ||
+             (labelStart<254 && destString[253-labelStart]!=0x2e))
+        ) {
+            info.errors|=UIDNA_ERROR_DOMAIN_NAME_TOO_LONG;
+        }
+    }
+    if( info.isBiDi && U_SUCCESS(errorCode) && (info.errors&severeErrors)==0 &&
+        (!info.isOkBiDi || (labelStart>0 && !isASCIIOkBiDi(srcArray, labelStart)))
+    ) {
+        info.errors|=UIDNA_ERROR_BIDI;
+    }
+}
+
+UnicodeString &
+UTS46::processUnicode(const UnicodeString &src,
+                      int32_t labelStart, int32_t mappingStart,
+                      UBool isLabel, UBool toASCII,
+                      UnicodeString &dest,
+                      IDNAInfo &info, UErrorCode &errorCode) const {
+    if(mappingStart==0) {
+        uts46Norm2.normalize(src, dest, errorCode);
+    } else {
+        uts46Norm2.normalizeSecondAndAppend(dest, src.tempSubString(mappingStart), errorCode);
+    }
+    if(U_FAILURE(errorCode)) {
+        return dest;
+    }
+    UBool doMapDevChars=
+        toASCII ? (options&UIDNA_NONTRANSITIONAL_TO_ASCII)==0 :
+                  (options&UIDNA_NONTRANSITIONAL_TO_UNICODE)==0;
+    const UChar *destArray=dest.getBuffer();
+    int32_t destLength=dest.length();
+    int32_t labelLimit=labelStart;
+    while(labelLimit<destLength) {
+        UChar c=destArray[labelLimit];
+        if(c==0x2e && !isLabel) {
+            int32_t labelLength=labelLimit-labelStart;
+            int32_t newLength=processLabel(dest, labelStart, labelLength,
+                                            toASCII, info, errorCode);
+            info.errors|=info.labelErrors;
+            info.labelErrors=0;
+            if(U_FAILURE(errorCode)) {
+                return dest;
+            }
+            destArray=dest.getBuffer();
+            destLength+=newLength-labelLength;
+            labelLimit=labelStart+=newLength+1;
+            continue;
+        } else if(c<0xdf) {
+            // pass
+        } else if(c<=0x200d && (c==0xdf || c==0x3c2 || c>=0x200c)) {
+            info.isTransDiff=TRUE;
+            if(doMapDevChars) {
+                destLength=mapDevChars(dest, labelStart, labelLimit, errorCode);
+                if(U_FAILURE(errorCode)) {
+                    return dest;
+                }
+                destArray=dest.getBuffer();
+                // All deviation characters have been mapped, no need to check for them again.
+                doMapDevChars=FALSE;
+                // Do not increment labelLimit in case c was removed.
+                continue;
+            }
+        } else if(U16_IS_SURROGATE(c)) {
+            if(U16_IS_SURROGATE_LEAD(c) ?
+                    (labelLimit+1)==destLength || !U16_IS_TRAIL(destArray[labelLimit+1]) :
+                    labelLimit==labelStart || !U16_IS_LEAD(destArray[labelLimit-1])) {
+                // Map an unpaired surrogate to U+FFFD before normalization so that when
+                // that removes characters we do not turn two unpaired ones into a pair.
+                info.labelErrors|=UIDNA_ERROR_DISALLOWED;
+                dest.setCharAt(labelLimit, 0xfffd);
+                destArray=dest.getBuffer();
+            }
+        }
+        ++labelLimit;
+    }
+    // Permit an empty label at the end (0<labelStart==labelLimit==destLength is ok)
+    // but not an empty label elsewhere nor a completely empty domain name.
+    // processLabel() sets UIDNA_ERROR_EMPTY_LABEL when labelLength==0.
+    if(0==labelStart || labelStart<labelLimit) {
+        processLabel(dest, labelStart, labelLimit-labelStart,
+                      toASCII, info, errorCode);
+        info.errors|=info.labelErrors;
+    }
+    return dest;
+}
+
+int32_t
+UTS46::mapDevChars(UnicodeString &dest, int32_t labelStart, int32_t mappingStart,
+                   UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    int32_t length=dest.length();
+    UChar *s=dest.getBuffer(dest[mappingStart]==0xdf ? length+1 : length);
+    if(s==NULL) {
+        errorCode=U_MEMORY_ALLOCATION_ERROR;
+        return length;
+    }
+    int32_t capacity=dest.getCapacity();
+    UBool didMapDevChars=FALSE;
+    int32_t readIndex=mappingStart, writeIndex=mappingStart;
+    do {
+        UChar c=s[readIndex++];
+        switch(c) {
+        case 0xdf:
+            // Map sharp s to ss.
+            didMapDevChars=TRUE;
+            s[writeIndex++]=0x73;  // Replace sharp s with first s.
+            // Insert second s and account for possible buffer reallocation.
+            if(writeIndex==readIndex) {
+                if(length==capacity) {
+                    dest.releaseBuffer(length);
+                    s=dest.getBuffer(length+1);
+                    if(s==NULL) {
+                        errorCode=U_MEMORY_ALLOCATION_ERROR;
+                        return length;
+                    }
+                    capacity=dest.getCapacity();
+                }
+                u_memmove(s+writeIndex+1, s+writeIndex, length-writeIndex);
+                ++readIndex;
+            }
+            s[writeIndex++]=0x73;
+            ++length;
+            break;
+        case 0x3c2:  // Map final sigma to nonfinal sigma.
+            didMapDevChars=TRUE;
+            s[writeIndex++]=0x3c3;
+            break;
+        case 0x200c:  // Ignore/remove ZWNJ.
+        case 0x200d:  // Ignore/remove ZWJ.
+            didMapDevChars=TRUE;
+            --length;
+            break;
+        default:
+            // Only really necessary if writeIndex was different from readIndex.
+            s[writeIndex++]=c;
+            break;
+        }
+    } while(writeIndex<length);
+    dest.releaseBuffer(length);
+    if(didMapDevChars) {
+        // Mapping deviation characters might have resulted in an un-NFC string.
+        // We could use either the NFC or the UTS #46 normalizer.
+        // By using the UTS #46 normalizer again, we avoid having to load a second .nrm data file.
+        UnicodeString normalized;
+        uts46Norm2.normalize(dest.tempSubString(labelStart), normalized, errorCode);
+        if(U_SUCCESS(errorCode)) {
+            dest.replace(labelStart, 0x7fffffff, normalized);
+            if(dest.isBogus()) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+            }
+            return dest.length();
+        }
+    }
+    return length;
+}
+
+// Some non-ASCII characters are equivalent to sequences with
+// non-LDH ASCII characters. To find them:
+// grep disallowed_STD3_valid IdnaMappingTable.txt (or uts46.txt)
+static inline UBool
+isNonASCIIDisallowedSTD3Valid(UChar32 c) {
+    return c==0x2260 || c==0x226E || c==0x226F;
+}
+
+// Replace the label in dest with the label string, if the label was modified.
+// If &label==&dest then the label was modified in-place and labelLength
+// is the new label length, different from label.length().
+// If &label!=&dest then labelLength==label.length().
+// Returns labelLength (= the new label length).
+static int32_t
+replaceLabel(UnicodeString &dest, int32_t destLabelStart, int32_t destLabelLength,
+             const UnicodeString &label, int32_t labelLength, UErrorCode &errorCode) {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    if(&label!=&dest) {
+        dest.replace(destLabelStart, destLabelLength, label);
+        if(dest.isBogus()) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return 0;
+        }
+    }
+    return labelLength;
+}
+
+int32_t
+UTS46::processLabel(UnicodeString &dest,
+                    int32_t labelStart, int32_t labelLength,
+                    UBool toASCII,
+                    IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    UnicodeString fromPunycode;
+    UnicodeString *labelString;
+    const UChar *label=dest.getBuffer()+labelStart;
+    int32_t destLabelStart=labelStart;
+    int32_t destLabelLength=labelLength;
+    UBool wasPunycode;
+    if(labelLength>=4 && label[0]==0x78 && label[1]==0x6e && label[2]==0x2d && label[3]==0x2d) {
+        // Label starts with "xn--", try to un-Punycode it.
+        // In IDNA2008, labels like "xn--" (decodes to an empty string) and
+        // "xn--ASCII-" (decodes to just "ASCII") fail the round-trip validation from
+        // comparing the ToUnicode input with the back-to-ToASCII output.
+        // They are alternate encodings of the respective ASCII labels.
+        // Ignore "xn---" here: It will fail Punycode.decode() which logically comes before
+        // the round-trip verification.
+        if(labelLength==4 || (labelLength>5 && label[labelLength-1]==u'-')) {
+            info.labelErrors|=UIDNA_ERROR_INVALID_ACE_LABEL;
+            return markBadACELabel(dest, labelStart, labelLength, toASCII, info, errorCode);
+        }
+        wasPunycode=TRUE;
+        UChar *unicodeBuffer=fromPunycode.getBuffer(-1);  // capacity==-1: most labels should fit
+        if(unicodeBuffer==NULL) {
+            // Should never occur if we used capacity==-1 which uses the internal buffer.
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return labelLength;
+        }
+        UErrorCode punycodeErrorCode=U_ZERO_ERROR;
+        int32_t unicodeLength=u_strFromPunycode(label+4, labelLength-4,
+                                                unicodeBuffer, fromPunycode.getCapacity(),
+                                                NULL, &punycodeErrorCode);
+        if(punycodeErrorCode==U_BUFFER_OVERFLOW_ERROR) {
+            fromPunycode.releaseBuffer(0);
+            unicodeBuffer=fromPunycode.getBuffer(unicodeLength);
+            if(unicodeBuffer==NULL) {
+                errorCode=U_MEMORY_ALLOCATION_ERROR;
+                return labelLength;
+            }
+            punycodeErrorCode=U_ZERO_ERROR;
+            unicodeLength=u_strFromPunycode(label+4, labelLength-4,
+                                            unicodeBuffer, fromPunycode.getCapacity(),
+                                            NULL, &punycodeErrorCode);
+        }
+        fromPunycode.releaseBuffer(unicodeLength);
+        if(U_FAILURE(punycodeErrorCode)) {
+            info.labelErrors|=UIDNA_ERROR_PUNYCODE;
+            return markBadACELabel(dest, labelStart, labelLength, toASCII, info, errorCode);
+        }
+        // Check for NFC, and for characters that are not
+        // valid or deviation characters according to the normalizer.
+        // If there is something wrong, then the string will change.
+        // Note that the normalizer passes through non-LDH ASCII and deviation characters.
+        // Deviation characters are ok in Punycode even in transitional processing.
+        // In the code further below, if we find non-LDH ASCII and we have UIDNA_USE_STD3_RULES
+        // then we will set UIDNA_ERROR_INVALID_ACE_LABEL there too.
+        UBool isValid=uts46Norm2.isNormalized(fromPunycode, errorCode);
+        if(U_FAILURE(errorCode)) {
+            return labelLength;
+        }
+        if(!isValid) {
+            info.labelErrors|=UIDNA_ERROR_INVALID_ACE_LABEL;
+            return markBadACELabel(dest, labelStart, labelLength, toASCII, info, errorCode);
+        }
+        labelString=&fromPunycode;
+        label=fromPunycode.getBuffer();
+        labelStart=0;
+        labelLength=fromPunycode.length();
+    } else {
+        wasPunycode=FALSE;
+        labelString=&dest;
+    }
+    // Validity check
+    if(labelLength==0) {
+        info.labelErrors|=UIDNA_ERROR_EMPTY_LABEL;
+        return replaceLabel(dest, destLabelStart, destLabelLength,
+                            *labelString, labelLength, errorCode);
+    }
+    // labelLength>0
+    if(labelLength>=4 && label[2]==0x2d && label[3]==0x2d) {
+        // label starts with "??--"
+        info.labelErrors|=UIDNA_ERROR_HYPHEN_3_4;
+    }
+    if(label[0]==0x2d) {
+        // label starts with "-"
+        info.labelErrors|=UIDNA_ERROR_LEADING_HYPHEN;
+    }
+    if(label[labelLength-1]==0x2d) {
+        // label ends with "-"
+        info.labelErrors|=UIDNA_ERROR_TRAILING_HYPHEN;
+    }
+    // If the label was not a Punycode label, then it was the result of
+    // mapping, normalization and label segmentation.
+    // If the label was in Punycode, then we mapped it again above
+    // and checked its validity.
+    // Now we handle the STD3 restriction to LDH characters (if set)
+    // and we look for U+FFFD which indicates disallowed characters
+    // in a non-Punycode label or U+FFFD itself in a Punycode label.
+    // We also check for dots which can come from the input to a single-label function.
+    // Ok to cast away const because we own the UnicodeString.
+    UChar *s=(UChar *)label;
+    const UChar *limit=label+labelLength;
+    UChar oredChars=0;
+    // If we enforce STD3 rules, then ASCII characters other than LDH and dot are disallowed.
+    UBool disallowNonLDHDot=(options&UIDNA_USE_STD3_RULES)!=0;
+    do {
+        UChar c=*s;
+        if(c<=0x7f) {
+            if(c==0x2e) {
+                info.labelErrors|=UIDNA_ERROR_LABEL_HAS_DOT;
+                *s=0xfffd;
+            } else if(disallowNonLDHDot && asciiData[c]<0) {
+                info.labelErrors|=UIDNA_ERROR_DISALLOWED;
+                *s=0xfffd;
+            }
+        } else {
+            oredChars|=c;
+            if(disallowNonLDHDot && isNonASCIIDisallowedSTD3Valid(c)) {
+                info.labelErrors|=UIDNA_ERROR_DISALLOWED;
+                *s=0xfffd;
+            } else if(c==0xfffd) {
+                info.labelErrors|=UIDNA_ERROR_DISALLOWED;
+            }
+        }
+        ++s;
+    } while(s<limit);
+    // Check for a leading combining mark after other validity checks
+    // so that we don't report UIDNA_ERROR_DISALLOWED for the U+FFFD from here.
+    UChar32 c;
+    int32_t cpLength=0;
+    // "Unsafe" is ok because unpaired surrogates were mapped to U+FFFD.
+    U16_NEXT_UNSAFE(label, cpLength, c);
+    if((U_GET_GC_MASK(c)&U_GC_M_MASK)!=0) {
+        info.labelErrors|=UIDNA_ERROR_LEADING_COMBINING_MARK;
+        labelString->replace(labelStart, cpLength, (UChar)0xfffd);
+        label=labelString->getBuffer()+labelStart;
+        labelLength+=1-cpLength;
+        if(labelString==&dest) {
+            destLabelLength=labelLength;
+        }
+    }
+    if((info.labelErrors&severeErrors)==0) {
+        // Do contextual checks only if we do not have U+FFFD from a severe error
+        // because U+FFFD can make these checks fail.
+        if((options&UIDNA_CHECK_BIDI)!=0 && (!info.isBiDi || info.isOkBiDi)) {
+            checkLabelBiDi(label, labelLength, info);
+        }
+        if( (options&UIDNA_CHECK_CONTEXTJ)!=0 && (oredChars&0x200c)==0x200c &&
+            !isLabelOkContextJ(label, labelLength)
+        ) {
+            info.labelErrors|=UIDNA_ERROR_CONTEXTJ;
+        }
+        if((options&UIDNA_CHECK_CONTEXTO)!=0 && oredChars>=0xb7) {
+            checkLabelContextO(label, labelLength, info);
+        }
+        if(toASCII) {
+            if(wasPunycode) {
+                // Leave a Punycode label unchanged if it has no severe errors.
+                if(destLabelLength>63) {
+                    info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                }
+                return destLabelLength;
+            } else if(oredChars>=0x80) {
+                // Contains non-ASCII characters.
+                UnicodeString punycode;
+                UChar *buffer=punycode.getBuffer(63);  // 63==maximum DNS label length
+                if(buffer==NULL) {
+                    errorCode=U_MEMORY_ALLOCATION_ERROR;
+                    return destLabelLength;
+                }
+                buffer[0]=0x78;  // Write "xn--".
+                buffer[1]=0x6e;
+                buffer[2]=0x2d;
+                buffer[3]=0x2d;
+                int32_t punycodeLength=u_strToPunycode(label, labelLength,
+                                                      buffer+4, punycode.getCapacity()-4,
+                                                      NULL, &errorCode);
+                if(errorCode==U_BUFFER_OVERFLOW_ERROR) {
+                    errorCode=U_ZERO_ERROR;
+                    punycode.releaseBuffer(4);
+                    buffer=punycode.getBuffer(4+punycodeLength);
+                    if(buffer==NULL) {
+                        errorCode=U_MEMORY_ALLOCATION_ERROR;
+                        return destLabelLength;
+                    }
+                    punycodeLength=u_strToPunycode(label, labelLength,
+                                                  buffer+4, punycode.getCapacity()-4,
+                                                  NULL, &errorCode);
+                }
+                punycodeLength+=4;
+                punycode.releaseBuffer(punycodeLength);
+                if(U_FAILURE(errorCode)) {
+                    return destLabelLength;
+                }
+                if(punycodeLength>63) {
+                    info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                }
+                return replaceLabel(dest, destLabelStart, destLabelLength,
+                                    punycode, punycodeLength, errorCode);
+            } else {
+                // all-ASCII label
+                if(labelLength>63) {
+                    info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+                }
+            }
+        }
+    } else {
+        // If a Punycode label has severe errors,
+        // then leave it but make sure it does not look valid.
+        if(wasPunycode) {
+            info.labelErrors|=UIDNA_ERROR_INVALID_ACE_LABEL;
+            return markBadACELabel(dest, destLabelStart, destLabelLength, toASCII, info, errorCode);
+        }
+    }
+    return replaceLabel(dest, destLabelStart, destLabelLength,
+                        *labelString, labelLength, errorCode);
+}
+
+// Make sure an ACE label does not look valid.
+// Append U+FFFD if the label has only LDH characters.
+// If UIDNA_USE_STD3_RULES, also replace disallowed ASCII characters with U+FFFD.
+int32_t
+UTS46::markBadACELabel(UnicodeString &dest,
+                       int32_t labelStart, int32_t labelLength,
+                       UBool toASCII, IDNAInfo &info, UErrorCode &errorCode) const {
+    if(U_FAILURE(errorCode)) {
+        return 0;
+    }
+    UBool disallowNonLDHDot=(options&UIDNA_USE_STD3_RULES)!=0;
+    UBool isASCII=TRUE;
+    UBool onlyLDH=TRUE;
+    const UChar *label=dest.getBuffer()+labelStart;
+    const UChar *limit=label+labelLength;
+    // Start after the initial "xn--".
+    // Ok to cast away const because we own the UnicodeString.
+    for(UChar *s=const_cast<UChar *>(label+4); s<limit; ++s) {
+        UChar c=*s;
+        if(c<=0x7f) {
+            if(c==0x2e) {
+                info.labelErrors|=UIDNA_ERROR_LABEL_HAS_DOT;
+                *s=0xfffd;
+                isASCII=onlyLDH=FALSE;
+            } else if(asciiData[c]<0) {
+                onlyLDH=FALSE;
+                if(disallowNonLDHDot) {
+                    *s=0xfffd;
+                    isASCII=FALSE;
+                }
+            }
+        } else {
+            isASCII=onlyLDH=FALSE;
+        }
+    }
+    if(onlyLDH) {
+        dest.insert(labelStart+labelLength, (UChar)0xfffd);
+        if(dest.isBogus()) {
+            errorCode=U_MEMORY_ALLOCATION_ERROR;
+            return 0;
+        }
+        ++labelLength;
+    } else {
+        if(toASCII && isASCII && labelLength>63) {
+            info.labelErrors|=UIDNA_ERROR_LABEL_TOO_LONG;
+        }
+    }
+    return labelLength;
+}
+
+const uint32_t L_MASK=U_MASK(U_LEFT_TO_RIGHT);
+const uint32_t R_AL_MASK=U_MASK(U_RIGHT_TO_LEFT)|U_MASK(U_RIGHT_TO_LEFT_ARABIC);
+const uint32_t L_R_AL_MASK=L_MASK|R_AL_MASK;
+
+const uint32_t R_AL_AN_MASK=R_AL_MASK|U_MASK(U_ARABIC_NUMBER);
+
+const uint32_t EN_AN_MASK=U_MASK(U_EUROPEAN_NUMBER)|U_MASK(U_ARABIC_NUMBER);
+const uint32_t R_AL_EN_AN_MASK=R_AL_MASK|EN_AN_MASK;
+const uint32_t L_EN_MASK=L_MASK|U_MASK(U_EUROPEAN_NUMBER);
+
+const uint32_t ES_CS_ET_ON_BN_NSM_MASK=
+    U_MASK(U_EUROPEAN_NUMBER_SEPARATOR)|
+    U_MASK(U_COMMON_NUMBER_SEPARATOR)|
+    U_MASK(U_EUROPEAN_NUMBER_TERMINATOR)|
+    U_MASK(U_OTHER_NEUTRAL)|
+    U_MASK(U_BOUNDARY_NEUTRAL)|
+    U_MASK(U_DIR_NON_SPACING_MARK);
+const uint32_t L_EN_ES_CS_ET_ON_BN_NSM_MASK=L_EN_MASK|ES_CS_ET_ON_BN_NSM_MASK;
+const uint32_t R_AL_AN_EN_ES_CS_ET_ON_BN_NSM_MASK=R_AL_MASK|EN_AN_MASK|ES_CS_ET_ON_BN_NSM_MASK;
+
+// We scan the whole label and check both for whether it contains RTL characters
+// and whether it passes the BiDi Rule.
+// In a BiDi domain name, all labels must pass the BiDi Rule, but we might find
+// that a domain name is a BiDi domain name (has an RTL label) only after
+// processing several earlier labels.
+void
+UTS46::checkLabelBiDi(const UChar *label, int32_t labelLength, IDNAInfo &info) const {
+    // IDNA2008 BiDi rule
+    // Get the directionality of the first character.
+    UChar32 c;
+    int32_t i=0;
+    U16_NEXT_UNSAFE(label, i, c);
+    uint32_t firstMask=U_MASK(u_charDirection(c));
+    // 1. The first character must be a character with BIDI property L, R
+    // or AL.  If it has the R or AL property, it is an RTL label; if it
+    // has the L property, it is an LTR label.
+    if((firstMask&~L_R_AL_MASK)!=0) {
+        info.isOkBiDi=FALSE;
+    }
+    // Get the directionality of the last non-NSM character.
+    uint32_t lastMask;
+    for(;;) {
+        if(i>=labelLength) {
+            lastMask=firstMask;
+            break;
+        }
+        U16_PREV_UNSAFE(label, labelLength, c);
+        UCharDirection dir=u_charDirection(c);
+        if(dir!=U_DIR_NON_SPACING_MARK) {
+            lastMask=U_MASK(dir);
+            break;
+        }
+    }
+    // 3. In an RTL label, the end of the label must be a character with
+    // BIDI property R, AL, EN or AN, followed by zero or more
+    // characters with BIDI property NSM.
+    // 6. In an LTR label, the end of the label must be a character with
+    // BIDI property L or EN, followed by zero or more characters with
+    // BIDI property NSM.
+    if( (firstMask&L_MASK)!=0 ?
+            (lastMask&~L_EN_MASK)!=0 :
+            (lastMask&~R_AL_EN_AN_MASK)!=0
+    ) {
+        info.isOkBiDi=FALSE;
+    }
+    // Add the directionalities of the intervening characters.
+    uint32_t mask=firstMask|lastMask;
+    while(i<labelLength) {
+        U16_NEXT_UNSAFE(label, i, c);
+        mask|=U_MASK(u_charDirection(c));
+    }
+    if(firstMask&L_MASK) {
+        // 5. In an LTR label, only characters with the BIDI properties L, EN,
+        // ES, CS, ET, ON, BN and NSM are allowed.
+        if((mask&~L_EN_ES_CS_ET_ON_BN_NSM_MASK)!=0) {
+            info.isOkBiDi=FALSE;
+        }
+    } else {
+        // 2. In an RTL label, only characters with the BIDI properties R, AL,
+        // AN, EN, ES, CS, ET, ON, BN and NSM are allowed.
+        if((mask&~R_AL_AN_EN_ES_CS_ET_ON_BN_NSM_MASK)!=0) {
+            info.isOkBiDi=FALSE;
+        }
+        // 4. In an RTL label, if an EN is present, no AN may be present, and
+        // vice versa.
+        if((mask&EN_AN_MASK)==EN_AN_MASK) {
+            info.isOkBiDi=FALSE;
+        }
+    }
+    // An RTL label is a label that contains at least one character of type
+    // R, AL or AN. [...]
+    // A "BIDI domain name" is a domain name that contains at least one RTL
+    // label. [...]
+    // The following rule, consisting of six conditions, applies to labels
+    // in BIDI domain names.
+    if((mask&R_AL_AN_MASK)!=0) {
+        info.isBiDi=TRUE;
+    }
+}
+
+// Special code for the ASCII prefix of a BiDi domain name.
+// The ASCII prefix is all-LTR.
+
+// IDNA2008 BiDi rule, parts relevant to ASCII labels:
+// 1. The first character must be a character with BIDI property L [...]
+// 5. In an LTR label, only characters with the BIDI properties L, EN,
+// ES, CS, ET, ON, BN and NSM are allowed.
+// 6. In an LTR label, the end of the label must be a character with
+// BIDI property L or EN [...]
+
+// UTF-16 version, called for mapped ASCII prefix.
+// Cannot contain uppercase A-Z.
+// s[length-1] must be the trailing dot.
+static UBool
+isASCIIOkBiDi(const UChar *s, int32_t length) {
+    int32_t labelStart=0;
+    for(int32_t i=0; i<length; ++i) {
+        UChar c=s[i];
+        if(c==0x2e) {  // dot
+            if(i>labelStart) {
+                c=s[i-1];
+                if(!(0x61<=c && c<=0x7a) && !(0x30<=c && c<=0x39)) {
+                    // Last character in the label is not an L or EN.
+                    return FALSE;
+                }
+            }
+            labelStart=i+1;
+        } else if(i==labelStart) {
+            if(!(0x61<=c && c<=0x7a)) {
+                // First character in the label is not an L.
+                return FALSE;
+            }
+        } else {
+            if(c<=0x20 && (c>=0x1c || (9<=c && c<=0xd))) {
+                // Intermediate character in the label is a B, S or WS.
+                return FALSE;
+            }
+        }
+    }
+    return TRUE;
+}
+
+// UTF-8 version, called for source ASCII prefix.
+// Can contain uppercase A-Z.
+// s[length-1] must be the trailing dot.
+static UBool
+isASCIIOkBiDi(const char *s, int32_t length) {
+    int32_t labelStart=0;
+    for(int32_t i=0; i<length; ++i) {
+        char c=s[i];
+        if(c==0x2e) {  // dot
+            if(i>labelStart) {
+                c=s[i-1];
+                if(!(0x61<=c && c<=0x7a) && !(0x41<=c && c<=0x5a) && !(0x30<=c && c<=0x39)) {
+                    // Last character in the label is not an L or EN.
+                    return FALSE;
+                }
+            }
+            labelStart=i+1;
+        } else if(i==labelStart) {
+            if(!(0x61<=c && c<=0x7a) && !(0x41<=c && c<=0x5a)) {
+                // First character in the label is not an L.
+                return FALSE;
+            }
+        } else {
+            if(c<=0x20 && (c>=0x1c || (9<=c && c<=0xd))) {
+                // Intermediate character in the label is a B, S or WS.
+                return FALSE;
+            }
+        }
+    }
+    return TRUE;
+}
+
+UBool
+UTS46::isLabelOkContextJ(const UChar *label, int32_t labelLength) const {
+    // [IDNA2008-Tables]
+    // 200C..200D  ; CONTEXTJ    # ZERO WIDTH NON-JOINER..ZERO WIDTH JOINER
+    for(int32_t i=0; i<labelLength; ++i) {
+        if(label[i]==0x200c) {
+            // Appendix A.1. ZERO WIDTH NON-JOINER
+            // Rule Set:
+            //  False;
+            //  If Canonical_Combining_Class(Before(cp)) .eq.  Virama Then True;
+            //  If RegExpMatch((Joining_Type:{L,D})(Joining_Type:T)*\u200C
+            //     (Joining_Type:T)*(Joining_Type:{R,D})) Then True;
+            if(i==0) {
+                return FALSE;
+            }
+            UChar32 c;
+            int32_t j=i;
+            U16_PREV_UNSAFE(label, j, c);
+            if(uts46Norm2.getCombiningClass(c)==9) {
+                continue;
+            }
+            // check precontext (Joining_Type:{L,D})(Joining_Type:T)*
+            for(;;) {
+                UJoiningType type=ubidi_getJoiningType(c);
+                if(type==U_JT_TRANSPARENT) {
+                    if(j==0) {
+                        return FALSE;
+                    }
+                    U16_PREV_UNSAFE(label, j, c);
+                } else if(type==U_JT_LEFT_JOINING || type==U_JT_DUAL_JOINING) {
+                    break;  // precontext fulfilled
+                } else {
+                    return FALSE;
+                }
+            }
+            // check postcontext (Joining_Type:T)*(Joining_Type:{R,D})
+            for(j=i+1;;) {
+                if(j==labelLength) {
+                    return FALSE;
+                }
+                U16_NEXT_UNSAFE(label, j, c);
+                UJoiningType type=ubidi_getJoiningType(c);
+                if(type==U_JT_TRANSPARENT) {
+                    // just skip this character
+                } else if(type==U_JT_RIGHT_JOINING || type==U_JT_DUAL_JOINING) {
+                    break;  // postcontext fulfilled
+                } else {
+                    return FALSE;
+                }
+            }
+        } else if(label[i]==0x200d) {
+            // Appendix A.2. ZERO WIDTH JOINER (U+200D)
+            // Rule Set:
+            //  False;
+            //  If Canonical_Combining_Class(Before(cp)) .eq.  Virama Then True;
+            if(i==0) {
+                return FALSE;
+            }
+            UChar32 c;
+            int32_t j=i;
+            U16_PREV_UNSAFE(label, j, c);
+            if(uts46Norm2.getCombiningClass(c)!=9) {
+                return FALSE;
+            }
+        }
+    }
+    return TRUE;
+}
+
+void
+UTS46::checkLabelContextO(const UChar *label, int32_t labelLength, IDNAInfo &info) const {
+    int32_t labelEnd=labelLength-1;  // inclusive
+    int32_t arabicDigits=0;  // -1 for 066x, +1 for 06Fx
+    for(int32_t i=0; i<=labelEnd; ++i) {
+        UChar32 c=label[i];
+        if(c<0xb7) {
+            // ASCII fastpath
+        } else if(c<=0x6f9) {
+            if(c==0xb7) {
+                // Appendix A.3. MIDDLE DOT (U+00B7)
+                // Rule Set:
+                //  False;
+                //  If Before(cp) .eq.  U+006C And
+                //     After(cp) .eq.  U+006C Then True;
+                if(!(0<i && label[i-1]==0x6c &&
+                     i<labelEnd && label[i+1]==0x6c)) {
+                    info.labelErrors|=UIDNA_ERROR_CONTEXTO_PUNCTUATION;
+                }
+            } else if(c==0x375) {
+                // Appendix A.4. GREEK LOWER NUMERAL SIGN (KERAIA) (U+0375)
+                // Rule Set:
+                //  False;
+                //  If Script(After(cp)) .eq.  Greek Then True;
+                UScriptCode script=USCRIPT_INVALID_CODE;
+                if(i<labelEnd) {
+                    UErrorCode errorCode=U_ZERO_ERROR;
+                    int32_t j=i+1;
+                    U16_NEXT(label, j, labelLength, c);
+                    script=uscript_getScript(c, &errorCode);
+                }
+                if(script!=USCRIPT_GREEK) {
+                    info.labelErrors|=UIDNA_ERROR_CONTEXTO_PUNCTUATION;
+                }
+            } else if(c==0x5f3 || c==0x5f4) {
+                // Appendix A.5. HEBREW PUNCTUATION GERESH (U+05F3)
+                // Rule Set:
+                //  False;
+                //  If Script(Before(cp)) .eq.  Hebrew Then True;
+                //
+                // Appendix A.6. HEBREW PUNCTUATION GERSHAYIM (U+05F4)
+                // Rule Set:
+                //  False;
+                //  If Script(Before(cp)) .eq.  Hebrew Then True;
+                UScriptCode script=USCRIPT_INVALID_CODE;
+                if(0<i) {
+                    UErrorCode errorCode=U_ZERO_ERROR;
+                    int32_t j=i;
+                    U16_PREV(label, 0, j, c);
+                    script=uscript_getScript(c, &errorCode);
+                }
+                if(script!=USCRIPT_HEBREW) {
+                    info.labelErrors|=UIDNA_ERROR_CONTEXTO_PUNCTUATION;
+                }
+            } else if(0x660<=c /* && c<=0x6f9 */) {
+                // Appendix A.8. ARABIC-INDIC DIGITS (0660..0669)
+                // Rule Set:
+                //  True;
+                //  For All Characters:
+                //    If cp .in. 06F0..06F9 Then False;
+                //  End For;
+                //
+                // Appendix A.9. EXTENDED ARABIC-INDIC DIGITS (06F0..06F9)
+                // Rule Set:
+                //  True;
+                //  For All Characters:
+                //    If cp .in. 0660..0669 Then False;
+                //  End For;
+                if(c<=0x669) {
+                    if(arabicDigits>0) {
+                        info.labelErrors|=UIDNA_ERROR_CONTEXTO_DIGITS;
+                    }
+                    arabicDigits=-1;
+                } else if(0x6f0<=c) {
+                    if(arabicDigits<0) {
+                        info.labelErrors|=UIDNA_ERROR_CONTEXTO_DIGITS;
+                    }
+                    arabicDigits=1;
+                }
+            }
+        } else if(c==0x30fb) {
+            // Appendix A.7. KATAKANA MIDDLE DOT (U+30FB)
+            // Rule Set:
+            //  False;
+            //  For All Characters:
+            //    If Script(cp) .in. {Hiragana, Katakana, Han} Then True;
+            //  End For;
+            UErrorCode errorCode=U_ZERO_ERROR;
+            for(int j=0;;) {
+                if(j>labelEnd) {
+                    info.labelErrors|=UIDNA_ERROR_CONTEXTO_PUNCTUATION;
+                    break;
+                }
+                U16_NEXT(label, j, labelLength, c);
+                UScriptCode script=uscript_getScript(c, &errorCode);
+                if(script==USCRIPT_HIRAGANA || script==USCRIPT_KATAKANA || script==USCRIPT_HAN) {
+                    break;
+                }
+            }
+        }
+    }
+}
+
+U_NAMESPACE_END
+
+// C API ------------------------------------------------------------------- ***
+
+U_NAMESPACE_USE
+
+U_CAPI UIDNA * U_EXPORT2
+uidna_openUTS46(uint32_t options, UErrorCode *pErrorCode) {
+    return reinterpret_cast<UIDNA *>(IDNA::createUTS46Instance(options, *pErrorCode));
+}
+
+U_CAPI void U_EXPORT2
+uidna_close(UIDNA *idna) {
+    delete reinterpret_cast<IDNA *>(idna);
+}
+
+static UBool
+checkArgs(const void *label, int32_t length,
+          void *dest, int32_t capacity,
+          UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(U_FAILURE(*pErrorCode)) {
+        return FALSE;
+    }
+    // sizeof(UIDNAInfo)=16 in the first API version.
+    if(pInfo==NULL || pInfo->size<16) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    if( (label==NULL ? length!=0 : length<-1) ||
+        (dest==NULL ? capacity!=0 : capacity<0) ||
+        (dest==label && label!=NULL)
+    ) {
+        *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    // Set all *pInfo bytes to 0 except for the size field itself.
+    uprv_memset(&pInfo->size+1, 0, pInfo->size-sizeof(pInfo->size));
+    return TRUE;
+}
+
+static void
+idnaInfoToStruct(IDNAInfo &info, UIDNAInfo *pInfo) {
+    pInfo->isTransitionalDifferent=info.isTransitionalDifferent();
+    pInfo->errors=info.getErrors();
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_labelToASCII(const UIDNA *idna,
+                   const UChar *label, int32_t length,
+                   UChar *dest, int32_t capacity,
+                   UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(label, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    UnicodeString src((UBool)(length<0), label, length);
+    UnicodeString destString(dest, 0, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->labelToASCII(src, destString, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return destString.extract(dest, capacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_labelToUnicode(const UIDNA *idna,
+                     const UChar *label, int32_t length,
+                     UChar *dest, int32_t capacity,
+                     UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(label, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    UnicodeString src((UBool)(length<0), label, length);
+    UnicodeString destString(dest, 0, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->labelToUnicode(src, destString, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return destString.extract(dest, capacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_nameToASCII(const UIDNA *idna,
+                  const UChar *name, int32_t length,
+                  UChar *dest, int32_t capacity,
+                  UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(name, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    UnicodeString src((UBool)(length<0), name, length);
+    UnicodeString destString(dest, 0, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->nameToASCII(src, destString, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return destString.extract(dest, capacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_nameToUnicode(const UIDNA *idna,
+                    const UChar *name, int32_t length,
+                    UChar *dest, int32_t capacity,
+                    UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(name, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    UnicodeString src((UBool)(length<0), name, length);
+    UnicodeString destString(dest, 0, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->nameToUnicode(src, destString, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return destString.extract(dest, capacity, *pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_labelToASCII_UTF8(const UIDNA *idna,
+                        const char *label, int32_t length,
+                        char *dest, int32_t capacity,
+                        UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(label, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    StringPiece src(label, length<0 ? static_cast<int32_t>(uprv_strlen(label)) : length);
+    CheckedArrayByteSink sink(dest, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->labelToASCII_UTF8(src, sink, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return u_terminateChars(dest, capacity, sink.NumberOfBytesAppended(), pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_labelToUnicodeUTF8(const UIDNA *idna,
+                         const char *label, int32_t length,
+                         char *dest, int32_t capacity,
+                         UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(label, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    StringPiece src(label, length<0 ? static_cast<int32_t>(uprv_strlen(label)) : length);
+    CheckedArrayByteSink sink(dest, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->labelToUnicodeUTF8(src, sink, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return u_terminateChars(dest, capacity, sink.NumberOfBytesAppended(), pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_nameToASCII_UTF8(const UIDNA *idna,
+                       const char *name, int32_t length,
+                       char *dest, int32_t capacity,
+                       UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(name, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    StringPiece src(name, length<0 ? static_cast<int32_t>(uprv_strlen(name)) : length);
+    CheckedArrayByteSink sink(dest, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->nameToASCII_UTF8(src, sink, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return u_terminateChars(dest, capacity, sink.NumberOfBytesAppended(), pErrorCode);
+}
+
+U_CAPI int32_t U_EXPORT2
+uidna_nameToUnicodeUTF8(const UIDNA *idna,
+                        const char *name, int32_t length,
+                        char *dest, int32_t capacity,
+                        UIDNAInfo *pInfo, UErrorCode *pErrorCode) {
+    if(!checkArgs(name, length, dest, capacity, pInfo, pErrorCode)) {
+        return 0;
+    }
+    StringPiece src(name, length<0 ? static_cast<int32_t>(uprv_strlen(name)) : length);
+    CheckedArrayByteSink sink(dest, capacity);
+    IDNAInfo info;
+    reinterpret_cast<const IDNA *>(idna)->nameToUnicodeUTF8(src, sink, info, *pErrorCode);
+    idnaInfoToStruct(info, pInfo);
+    return u_terminateChars(dest, capacity, sink.NumberOfBytesAppended(), pErrorCode);
+}
+
+#endif  // UCONFIG_NO_IDNA
diff --git a/thirdparty/icu4c/common/utypeinfo.h b/thirdparty/icu4c/common/utypeinfo.h
new file mode 100644
index 0000000000..c6663734fc
--- /dev/null
+++ b/thirdparty/icu4c/common/utypeinfo.h
@@ -0,0 +1,32 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 2012-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*/
+
+#ifndef __UTYPEINFO_H__
+#define __UTYPEINFO_H__
+
+// Windows header <typeinfo> does not define 'exception' in 'std' namespace.
+// Therefore, a project using ICU cannot be compiled with _HAS_EXCEPTIONS
+// set to 0 on Windows with Visual Studio. To work around that, we have to
+// include <exception> explicitly and add using statement below.
+// Whenever 'typeid' is used, this header has to be included
+// instead of <typeinfo>.
+// Visual Studio 10 emits warning 4275 with this change. If you compile
+// with exception disabled, you have to suppress warning 4275.
+#if defined(_MSC_VER) && _HAS_EXCEPTIONS == 0
+#include <exception>
+using std::exception;
+#endif
+#if defined(__GLIBCXX__)
+namespace std { class type_info; } // WORKAROUND: http://llvm.org/bugs/show_bug.cgi?id=13364
+#endif
+#include <typeinfo>  // for 'typeid' to work
+
+#endif
diff --git a/thirdparty/icu4c/common/utypes.cpp b/thirdparty/icu4c/common/utypes.cpp
new file mode 100644
index 0000000000..63e05b1249
--- /dev/null
+++ b/thirdparty/icu4c/common/utypes.cpp
@@ -0,0 +1,227 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+*
+*   Copyright (C) 1997-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+******************************************************************************
+*
+*  FILE NAME : utypes.c (previously putil.c)
+*
+*   Date        Name        Description
+*   10/07/2004  grhoten     split from putil.c
+******************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+/* u_errorName() ------------------------------------------------------------ */
+
+static const char * const
+_uErrorInfoName[U_ERROR_WARNING_LIMIT-U_ERROR_WARNING_START]={
+    "U_USING_FALLBACK_WARNING",
+    "U_USING_DEFAULT_WARNING",
+    "U_SAFECLONE_ALLOCATED_WARNING",
+    "U_STATE_OLD_WARNING",
+    "U_STRING_NOT_TERMINATED_WARNING",
+    "U_SORT_KEY_TOO_SHORT_WARNING",
+    "U_AMBIGUOUS_ALIAS_WARNING",
+    "U_DIFFERENT_UCA_VERSION",
+    "U_PLUGIN_CHANGED_LEVEL_WARNING",
+};
+
+static const char * const
+_uTransErrorName[U_PARSE_ERROR_LIMIT - U_PARSE_ERROR_START]={
+    "U_BAD_VARIABLE_DEFINITION",
+    "U_MALFORMED_RULE",
+    "U_MALFORMED_SET",
+    "U_MALFORMED_SYMBOL_REFERENCE",
+    "U_MALFORMED_UNICODE_ESCAPE",
+    "U_MALFORMED_VARIABLE_DEFINITION",
+    "U_MALFORMED_VARIABLE_REFERENCE",
+    "U_MISMATCHED_SEGMENT_DELIMITERS",
+    "U_MISPLACED_ANCHOR_START",
+    "U_MISPLACED_CURSOR_OFFSET",
+    "U_MISPLACED_QUANTIFIER",
+    "U_MISSING_OPERATOR",
+    "U_MISSING_SEGMENT_CLOSE",
+    "U_MULTIPLE_ANTE_CONTEXTS",
+    "U_MULTIPLE_CURSORS",
+    "U_MULTIPLE_POST_CONTEXTS",
+    "U_TRAILING_BACKSLASH",
+    "U_UNDEFINED_SEGMENT_REFERENCE",
+    "U_UNDEFINED_VARIABLE",
+    "U_UNQUOTED_SPECIAL",
+    "U_UNTERMINATED_QUOTE",
+    "U_RULE_MASK_ERROR",
+    "U_MISPLACED_COMPOUND_FILTER",
+    "U_MULTIPLE_COMPOUND_FILTERS",
+    "U_INVALID_RBT_SYNTAX",
+    "U_INVALID_PROPERTY_PATTERN",
+    "U_MALFORMED_PRAGMA",
+    "U_UNCLOSED_SEGMENT",
+    "U_ILLEGAL_CHAR_IN_SEGMENT",
+    "U_VARIABLE_RANGE_EXHAUSTED",
+    "U_VARIABLE_RANGE_OVERLAP",
+    "U_ILLEGAL_CHARACTER",
+    "U_INTERNAL_TRANSLITERATOR_ERROR",
+    "U_INVALID_ID",
+    "U_INVALID_FUNCTION"
+};
+
+static const char * const
+_uErrorName[U_STANDARD_ERROR_LIMIT]={
+    "U_ZERO_ERROR",
+
+    "U_ILLEGAL_ARGUMENT_ERROR",
+    "U_MISSING_RESOURCE_ERROR",
+    "U_INVALID_FORMAT_ERROR",
+    "U_FILE_ACCESS_ERROR",
+    "U_INTERNAL_PROGRAM_ERROR",
+    "U_MESSAGE_PARSE_ERROR",
+    "U_MEMORY_ALLOCATION_ERROR",
+    "U_INDEX_OUTOFBOUNDS_ERROR",
+    "U_PARSE_ERROR",
+    "U_INVALID_CHAR_FOUND",
+    "U_TRUNCATED_CHAR_FOUND",
+    "U_ILLEGAL_CHAR_FOUND",
+    "U_INVALID_TABLE_FORMAT",
+    "U_INVALID_TABLE_FILE",
+    "U_BUFFER_OVERFLOW_ERROR",
+    "U_UNSUPPORTED_ERROR",
+    "U_RESOURCE_TYPE_MISMATCH",
+    "U_ILLEGAL_ESCAPE_SEQUENCE",
+    "U_UNSUPPORTED_ESCAPE_SEQUENCE",
+    "U_NO_SPACE_AVAILABLE",
+    "U_CE_NOT_FOUND_ERROR",
+    "U_PRIMARY_TOO_LONG_ERROR",
+    "U_STATE_TOO_OLD_ERROR",
+    "U_TOO_MANY_ALIASES_ERROR",
+    "U_ENUM_OUT_OF_SYNC_ERROR",
+    "U_INVARIANT_CONVERSION_ERROR",
+    "U_INVALID_STATE_ERROR",
+    "U_COLLATOR_VERSION_MISMATCH",
+    "U_USELESS_COLLATOR_ERROR",
+    "U_NO_WRITE_PERMISSION",
+    "U_INPUT_TOO_LONG_ERROR"
+};
+static const char * const
+_uFmtErrorName[U_FMT_PARSE_ERROR_LIMIT - U_FMT_PARSE_ERROR_START] = {
+    "U_UNEXPECTED_TOKEN",
+    "U_MULTIPLE_DECIMAL_SEPARATORS",
+    "U_MULTIPLE_EXPONENTIAL_SYMBOLS",
+    "U_MALFORMED_EXPONENTIAL_PATTERN",
+    "U_MULTIPLE_PERCENT_SYMBOLS",
+    "U_MULTIPLE_PERMILL_SYMBOLS",
+    "U_MULTIPLE_PAD_SPECIFIERS",
+    "U_PATTERN_SYNTAX_ERROR",
+    "U_ILLEGAL_PAD_POSITION",
+    "U_UNMATCHED_BRACES",
+    "U_UNSUPPORTED_PROPERTY",
+    "U_UNSUPPORTED_ATTRIBUTE",
+    "U_ARGUMENT_TYPE_MISMATCH",
+    "U_DUPLICATE_KEYWORD",
+    "U_UNDEFINED_KEYWORD",
+    "U_DEFAULT_KEYWORD_MISSING",
+    "U_DECIMAL_NUMBER_SYNTAX_ERROR",
+    "U_FORMAT_INEXACT_ERROR",
+    "U_NUMBER_ARG_OUTOFBOUNDS_ERROR",
+    "U_NUMBER_SKELETON_SYNTAX_ERROR",
+};
+
+static const char * const
+_uBrkErrorName[U_BRK_ERROR_LIMIT - U_BRK_ERROR_START] = {
+    "U_BRK_INTERNAL_ERROR",
+    "U_BRK_HEX_DIGITS_EXPECTED",
+    "U_BRK_SEMICOLON_EXPECTED",
+    "U_BRK_RULE_SYNTAX",
+    "U_BRK_UNCLOSED_SET",
+    "U_BRK_ASSIGN_ERROR",
+    "U_BRK_VARIABLE_REDFINITION",
+    "U_BRK_MISMATCHED_PAREN",
+    "U_BRK_NEW_LINE_IN_QUOTED_STRING",
+    "U_BRK_UNDEFINED_VARIABLE",
+    "U_BRK_INIT_ERROR",
+    "U_BRK_RULE_EMPTY_SET",
+    "U_BRK_UNRECOGNIZED_OPTION",
+    "U_BRK_MALFORMED_RULE_TAG"
+};
+
+static const char * const
+_uRegexErrorName[U_REGEX_ERROR_LIMIT - U_REGEX_ERROR_START] = {
+    "U_REGEX_INTERNAL_ERROR",
+    "U_REGEX_RULE_SYNTAX",
+    "U_REGEX_INVALID_STATE",
+    "U_REGEX_BAD_ESCAPE_SEQUENCE",
+    "U_REGEX_PROPERTY_SYNTAX",
+    "U_REGEX_UNIMPLEMENTED",
+    "U_REGEX_MISMATCHED_PAREN",
+    "U_REGEX_NUMBER_TOO_BIG",
+    "U_REGEX_BAD_INTERVAL",
+    "U_REGEX_MAX_LT_MIN",
+    "U_REGEX_INVALID_BACK_REF",
+    "U_REGEX_INVALID_FLAG",
+    "U_REGEX_LOOK_BEHIND_LIMIT",
+    "U_REGEX_SET_CONTAINS_STRING",
+    "U_REGEX_OCTAL_TOO_BIG",
+    "U_REGEX_MISSING_CLOSE_BRACKET",
+    "U_REGEX_INVALID_RANGE",
+    "U_REGEX_STACK_OVERFLOW",
+    "U_REGEX_TIME_OUT",
+    "U_REGEX_STOPPED_BY_CALLER",
+    "U_REGEX_PATTERN_TOO_BIG",
+    "U_REGEX_INVALID_CAPTURE_GROUP_NAME"
+};
+
+static const char * const
+_uIDNAErrorName[U_IDNA_ERROR_LIMIT - U_IDNA_ERROR_START] = {
+     "U_STRINGPREP_PROHIBITED_ERROR",
+     "U_STRINGPREP_UNASSIGNED_ERROR",
+     "U_STRINGPREP_CHECK_BIDI_ERROR",
+     "U_IDNA_STD3_ASCII_RULES_ERROR",
+     "U_IDNA_ACE_PREFIX_ERROR",
+     "U_IDNA_VERIFICATION_ERROR",
+     "U_IDNA_LABEL_TOO_LONG_ERROR",
+     "U_IDNA_ZERO_LENGTH_LABEL_ERROR",
+     "U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR"
+};
+
+static const char * const
+_uPluginErrorName[U_PLUGIN_ERROR_LIMIT - U_PLUGIN_ERROR_START] = {
+     "U_PLUGIN_TOO_HIGH",
+     "U_PLUGIN_DIDNT_SET_LEVEL",
+};
+
+U_CAPI const char * U_EXPORT2
+u_errorName(UErrorCode code) {
+    if(U_ZERO_ERROR <= code && code < U_STANDARD_ERROR_LIMIT) {
+        return _uErrorName[code];
+    } else if(U_ERROR_WARNING_START <= code && code < U_ERROR_WARNING_LIMIT) {
+        return _uErrorInfoName[code - U_ERROR_WARNING_START];
+    } else if(U_PARSE_ERROR_START <= code && code < U_PARSE_ERROR_LIMIT){
+        return _uTransErrorName[code - U_PARSE_ERROR_START];
+    } else if(U_FMT_PARSE_ERROR_START <= code && code < U_FMT_PARSE_ERROR_LIMIT){
+        return _uFmtErrorName[code - U_FMT_PARSE_ERROR_START];
+    } else if (U_BRK_ERROR_START <= code  && code < U_BRK_ERROR_LIMIT){
+        return _uBrkErrorName[code - U_BRK_ERROR_START];
+    } else if (U_REGEX_ERROR_START <= code && code < U_REGEX_ERROR_LIMIT) {
+        return _uRegexErrorName[code - U_REGEX_ERROR_START];
+    } else if(U_IDNA_ERROR_START <= code && code < U_IDNA_ERROR_LIMIT) {
+        return _uIDNAErrorName[code - U_IDNA_ERROR_START];
+    } else if(U_PLUGIN_ERROR_START <= code && code < U_PLUGIN_ERROR_LIMIT) {
+        return _uPluginErrorName[code - U_PLUGIN_ERROR_START];
+    } else {
+        return "[BOGUS UErrorCode]";
+    }
+}
+
+/*
+ * Hey, Emacs, please set the following:
+ *
+ * Local Variables:
+ * indent-tabs-mode: nil
+ * End:
+ *
+ */
diff --git a/thirdparty/icu4c/common/uvector.cpp b/thirdparty/icu4c/common/uvector.cpp
new file mode 100644
index 0000000000..cf19edf646
--- /dev/null
+++ b/thirdparty/icu4c/common/uvector.cpp
@@ -0,0 +1,567 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1999-2013, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   10/22/99    alan        Creation.
+**********************************************************************
+*/
+
+#include "uvector.h"
+#include "cmemory.h"
+#include "uarrsort.h"
+#include "uelement.h"
+
+U_NAMESPACE_BEGIN
+
+#define DEFAULT_CAPACITY 8
+
+/*
+ * Constants for hinting whether a key is an integer
+ * or a pointer.  If a hint bit is zero, then the associated
+ * token is assumed to be an integer. This is needed for iSeries
+ */
+#define HINT_KEY_POINTER   (1)
+#define HINT_KEY_INTEGER   (0)
+ 
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UVector)
+
+UVector::UVector(UErrorCode &status) :
+    count(0),
+    capacity(0),
+    elements(0),
+    deleter(0),
+    comparer(0)
+{
+    _init(DEFAULT_CAPACITY, status);
+}
+
+UVector::UVector(int32_t initialCapacity, UErrorCode &status) :
+    count(0),
+    capacity(0),
+    elements(0),
+    deleter(0),
+    comparer(0)
+{
+    _init(initialCapacity, status);
+}
+
+UVector::UVector(UObjectDeleter *d, UElementsAreEqual *c, UErrorCode &status) :
+    count(0),
+    capacity(0),
+    elements(0),
+    deleter(d),
+    comparer(c)
+{
+    _init(DEFAULT_CAPACITY, status);
+}
+
+UVector::UVector(UObjectDeleter *d, UElementsAreEqual *c, int32_t initialCapacity, UErrorCode &status) :
+    count(0),
+    capacity(0),
+    elements(0),
+    deleter(d),
+    comparer(c)
+{
+    _init(initialCapacity, status);
+}
+
+void UVector::_init(int32_t initialCapacity, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return;
+    }
+    // Fix bogus initialCapacity values; avoid malloc(0) and integer overflow
+    if ((initialCapacity < 1) || (initialCapacity > (int32_t)(INT32_MAX / sizeof(UElement)))) {
+        initialCapacity = DEFAULT_CAPACITY;
+    }
+    elements = (UElement *)uprv_malloc(sizeof(UElement)*initialCapacity);
+    if (elements == 0) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        capacity = initialCapacity;
+    }
+}
+
+UVector::~UVector() {
+    removeAllElements();
+    uprv_free(elements);
+    elements = 0;
+}
+
+/**
+ * Assign this object to another (make this a copy of 'other').
+ * Use the 'assign' function to assign each element.
+ */
+void UVector::assign(const UVector& other, UElementAssigner *assign, UErrorCode &ec) {
+    if (ensureCapacity(other.count, ec)) {
+        setSize(other.count, ec);
+        if (U_SUCCESS(ec)) {
+            for (int32_t i=0; i<other.count; ++i) {
+                if (elements[i].pointer != 0 && deleter != 0) {
+                    (*deleter)(elements[i].pointer);
+                }
+                (*assign)(&elements[i], &other.elements[i]);
+            }
+        }
+    }
+}
+
+// This only does something sensible if this object has a non-null comparer
+UBool UVector::operator==(const UVector& other) {
+    int32_t i;
+    if (count != other.count) return FALSE;
+    if (comparer != NULL) {
+        // Compare using this object's comparer
+        for (i=0; i<count; ++i) {
+            if (!(*comparer)(elements[i], other.elements[i])) {
+                return FALSE;
+            }
+        }
+    }
+    return TRUE;
+}
+
+void UVector::addElement(void* obj, UErrorCode &status) {
+    if (ensureCapacity(count + 1, status)) {
+        elements[count++].pointer = obj;
+    }
+}
+
+void UVector::addElement(int32_t elem, UErrorCode &status) {
+    if (ensureCapacity(count + 1, status)) {
+        elements[count].pointer = NULL;     // Pointers may be bigger than ints.
+        elements[count].integer = elem;
+        count++;
+    }
+}
+
+void UVector::setElementAt(void* obj, int32_t index) {
+    if (0 <= index && index < count) {
+        if (elements[index].pointer != 0 && deleter != 0) {
+            (*deleter)(elements[index].pointer);
+        }
+        elements[index].pointer = obj;
+    }
+    /* else index out of range */
+}
+
+void UVector::setElementAt(int32_t elem, int32_t index) {
+    if (0 <= index && index < count) {
+        if (elements[index].pointer != 0 && deleter != 0) {
+            // TODO:  this should be an error.  mixing up ints and pointers.
+            (*deleter)(elements[index].pointer);
+        }
+        elements[index].pointer = NULL;
+        elements[index].integer = elem;
+    }
+    /* else index out of range */
+}
+
+void UVector::insertElementAt(void* obj, int32_t index, UErrorCode &status) {
+    // must have 0 <= index <= count
+    if (0 <= index && index <= count && ensureCapacity(count + 1, status)) {
+        for (int32_t i=count; i>index; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[index].pointer = obj;
+        ++count;
+    }
+    /* else index out of range */
+}
+
+void UVector::insertElementAt(int32_t elem, int32_t index, UErrorCode &status) {
+    // must have 0 <= index <= count
+    if (0 <= index && index <= count && ensureCapacity(count + 1, status)) {
+        for (int32_t i=count; i>index; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[index].pointer = NULL;
+        elements[index].integer = elem;
+        ++count;
+    }
+    /* else index out of range */
+}
+
+void* UVector::elementAt(int32_t index) const {
+    return (0 <= index && index < count) ? elements[index].pointer : 0;
+}
+
+int32_t UVector::elementAti(int32_t index) const {
+    return (0 <= index && index < count) ? elements[index].integer : 0;
+}
+
+UBool UVector::containsAll(const UVector& other) const {
+    for (int32_t i=0; i<other.size(); ++i) {
+        if (indexOf(other.elements[i]) < 0) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+UBool UVector::containsNone(const UVector& other) const {
+    for (int32_t i=0; i<other.size(); ++i) {
+        if (indexOf(other.elements[i]) >= 0) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+UBool UVector::removeAll(const UVector& other) {
+    UBool changed = FALSE;
+    for (int32_t i=0; i<other.size(); ++i) {
+        int32_t j = indexOf(other.elements[i]);
+        if (j >= 0) {
+            removeElementAt(j);
+            changed = TRUE;
+        }
+    }
+    return changed;
+}
+
+UBool UVector::retainAll(const UVector& other) {
+    UBool changed = FALSE;
+    for (int32_t j=size()-1; j>=0; --j) {
+        int32_t i = other.indexOf(elements[j]);
+        if (i < 0) {
+            removeElementAt(j);
+            changed = TRUE;
+        }
+    }
+    return changed;
+}
+
+void UVector::removeElementAt(int32_t index) {
+    void* e = orphanElementAt(index);
+    if (e != 0 && deleter != 0) {
+        (*deleter)(e);
+    }
+}
+
+UBool UVector::removeElement(void* obj) {
+    int32_t i = indexOf(obj);
+    if (i >= 0) {
+        removeElementAt(i);
+        return TRUE;
+    }
+    return FALSE;
+}
+
+void UVector::removeAllElements(void) {
+    if (deleter != 0) {
+        for (int32_t i=0; i<count; ++i) {
+            if (elements[i].pointer != 0) {
+                (*deleter)(elements[i].pointer);
+            }
+        }
+    }
+    count = 0;
+}
+
+UBool   UVector::equals(const UVector &other) const {
+    int      i;
+
+    if (this->count != other.count) {
+        return FALSE;
+    }
+    if (comparer == 0) {
+        for (i=0; i<count; i++) {
+            if (elements[i].pointer != other.elements[i].pointer) {
+                return FALSE;
+            }
+        }
+    } else {
+        UElement key;
+        for (i=0; i<count; i++) {
+            key.pointer = &other.elements[i];
+            if (!(*comparer)(key, elements[i])) {
+                return FALSE;
+            }
+        }
+    }
+    return TRUE;
+}
+
+
+
+int32_t UVector::indexOf(void* obj, int32_t startIndex) const {
+    UElement key;
+    key.pointer = obj;
+    return indexOf(key, startIndex, HINT_KEY_POINTER);
+}
+
+int32_t UVector::indexOf(int32_t obj, int32_t startIndex) const {
+    UElement key;
+    key.integer = obj;
+    return indexOf(key, startIndex, HINT_KEY_INTEGER);
+}
+
+// This only works if this object has a non-null comparer
+int32_t UVector::indexOf(UElement key, int32_t startIndex, int8_t hint) const {
+    int32_t i;
+    if (comparer != 0) {
+        for (i=startIndex; i<count; ++i) {
+            if ((*comparer)(key, elements[i])) {
+                return i;
+            }
+        }
+    } else {
+        for (i=startIndex; i<count; ++i) {
+            /* Pointers are not always the same size as ints so to perform
+             * a valid comparision we need to know whether we are being
+             * provided an int or a pointer. */
+            if (hint & HINT_KEY_POINTER) {
+                if (key.pointer == elements[i].pointer) {
+                    return i;
+                }
+            } else {
+                if (key.integer == elements[i].integer) {
+                    return i;
+                }
+            }
+        }
+    }
+    return -1;
+}
+
+UBool UVector::ensureCapacity(int32_t minimumCapacity, UErrorCode &status) {
+	if (minimumCapacity < 0) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+	}
+    if (capacity < minimumCapacity) {
+        if (capacity > (INT32_MAX - 1) / 2) {        	// integer overflow check
+        	status = U_ILLEGAL_ARGUMENT_ERROR;
+        	return FALSE;
+        }
+        int32_t newCap = capacity * 2;
+        if (newCap < minimumCapacity) {
+            newCap = minimumCapacity;
+        }
+        if (newCap > (int32_t)(INT32_MAX / sizeof(UElement))) {	// integer overflow check
+        	// We keep the original memory contents on bad minimumCapacity.
+        	status = U_ILLEGAL_ARGUMENT_ERROR;
+        	return FALSE;
+        }
+        UElement* newElems = (UElement *)uprv_realloc(elements, sizeof(UElement)*newCap);
+        if (newElems == NULL) {
+            // We keep the original contents on the memory failure on realloc or bad minimumCapacity.
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return FALSE;
+        }
+        elements = newElems;
+        capacity = newCap;
+    }
+    return TRUE;
+}
+
+/**
+ * Change the size of this vector as follows: If newSize is smaller,
+ * then truncate the array, possibly deleting held elements for i >=
+ * newSize.  If newSize is larger, grow the array, filling in new
+ * slots with NULL.
+ */
+void UVector::setSize(int32_t newSize, UErrorCode &status) {
+    int32_t i;
+    if (newSize < 0) {
+        return;
+    }
+    if (newSize > count) {
+        if (!ensureCapacity(newSize, status)) {
+            return;
+        }
+        UElement empty;
+        empty.pointer = NULL;
+        empty.integer = 0;
+        for (i=count; i<newSize; ++i) {
+            elements[i] = empty;
+        }
+    } else {
+        /* Most efficient to count down */
+        for (i=count-1; i>=newSize; --i) {
+            removeElementAt(i);
+        }
+    }
+    count = newSize;
+}
+
+/**
+ * Fill in the given array with all elements of this vector.
+ */
+void** UVector::toArray(void** result) const {
+    void** a = result;
+    for (int i=0; i<count; ++i) {
+        *a++ = elements[i].pointer;
+    }
+    return result;
+}
+
+UObjectDeleter *UVector::setDeleter(UObjectDeleter *d) {
+    UObjectDeleter *old = deleter;
+    deleter = d;
+    return old;
+}
+
+UElementsAreEqual *UVector::setComparer(UElementsAreEqual *d) {
+    UElementsAreEqual *old = comparer;
+    comparer = d;
+    return old;
+}
+
+/**
+ * Removes the element at the given index from this vector and
+ * transfer ownership of it to the caller.  After this call, the
+ * caller owns the result and must delete it and the vector entry
+ * at 'index' is removed, shifting all subsequent entries back by
+ * one index and shortening the size of the vector by one.  If the
+ * index is out of range or if there is no item at the given index
+ * then 0 is returned and the vector is unchanged.
+ */
+void* UVector::orphanElementAt(int32_t index) {
+    void* e = 0;
+    if (0 <= index && index < count) {
+        e = elements[index].pointer;
+        for (int32_t i=index; i<count-1; ++i) {
+            elements[i] = elements[i+1];
+        }
+        --count;
+    }
+    /* else index out of range */
+    return e;
+}
+
+/**
+ * Insert the given object into this vector at its sorted position
+ * as defined by 'compare'.  The current elements are assumed to
+ * be sorted already.
+ */
+void UVector::sortedInsert(void* obj, UElementComparator *compare, UErrorCode& ec) {
+    UElement e;
+    e.pointer = obj;
+    sortedInsert(e, compare, ec);
+}
+
+/**
+ * Insert the given integer into this vector at its sorted position
+ * as defined by 'compare'.  The current elements are assumed to
+ * be sorted already.
+ */
+void UVector::sortedInsert(int32_t obj, UElementComparator *compare, UErrorCode& ec) {
+    UElement e;
+    e.integer = obj;
+    sortedInsert(e, compare, ec);
+}
+
+// ASSUME elements[] IS CURRENTLY SORTED
+void UVector::sortedInsert(UElement e, UElementComparator *compare, UErrorCode& ec) {
+    // Perform a binary search for the location to insert tok at.  Tok
+    // will be inserted between two elements a and b such that a <=
+    // tok && tok < b, where there is a 'virtual' elements[-1] always
+    // less than tok and a 'virtual' elements[count] always greater
+    // than tok.
+    int32_t min = 0, max = count;
+    while (min != max) {
+        int32_t probe = (min + max) / 2;
+        int8_t c = (*compare)(elements[probe], e);
+        if (c > 0) {
+            max = probe;
+        } else {
+            // assert(c <= 0);
+            min = probe + 1;
+        }
+    }
+    if (ensureCapacity(count + 1, ec)) {
+        for (int32_t i=count; i>min; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[min] = e;
+        ++count;
+    }
+}
+
+/**
+  *  Array sort comparator function.
+  *  Used from UVector::sort()
+  *  Conforms to function signature required for uprv_sortArray().
+  *  This function is essentially just a wrapper, to make a
+  *  UVector style comparator function usable with uprv_sortArray().
+  *
+  *  The context pointer to this function is a pointer back
+  *  (with some extra indirection) to the user supplied comparator.
+  *  
+  */
+static int32_t U_CALLCONV
+sortComparator(const void *context, const void *left, const void *right) {
+    UElementComparator *compare = *static_cast<UElementComparator * const *>(context);
+    UElement e1 = *static_cast<const UElement *>(left);
+    UElement e2 = *static_cast<const UElement *>(right);
+    int32_t result = (*compare)(e1, e2);
+    return result;
+}
+
+
+/**
+  *  Array sort comparison function for use from UVector::sorti()
+  *  Compares int32_t vector elements.
+  */
+static int32_t U_CALLCONV
+sortiComparator(const void * /*context */, const void *left, const void *right) {
+    const UElement *e1 = static_cast<const UElement *>(left);
+    const UElement *e2 = static_cast<const UElement *>(right);
+    int32_t result = e1->integer < e2->integer? -1 :
+                     e1->integer == e2->integer? 0 : 1;
+    return result;
+}
+
+/**
+  * Sort the vector, assuming it constains ints.
+  *     (A more general sort would take a comparison function, but it's
+  *     not clear whether UVector's UElementComparator or
+  *     UComparator from uprv_sortAray would be more appropriate.)
+  */
+void UVector::sorti(UErrorCode &ec) {
+    if (U_SUCCESS(ec)) {
+        uprv_sortArray(elements, count, sizeof(UElement),
+                       sortiComparator, NULL,  FALSE, &ec);
+    }
+}
+
+
+/**
+ *  Sort with a user supplied comparator.
+ *
+ *    The comparator function handling is confusing because the function type
+ *    for UVector  (as defined for sortedInsert()) is different from the signature
+ *    required by uprv_sortArray().  This is handled by passing the
+ *    the UVector sort function pointer via the context pointer to a
+ *    sortArray() comparator function, which can then call back to
+ *    the original user functtion.
+ *
+ *    An additional twist is that it's not safe to pass a pointer-to-function
+ *    as  a (void *) data pointer, so instead we pass a (data) pointer to a
+ *    pointer-to-function variable.
+ */
+void UVector::sort(UElementComparator *compare, UErrorCode &ec) {
+    if (U_SUCCESS(ec)) {
+        uprv_sortArray(elements, count, sizeof(UElement),
+                       sortComparator, &compare, FALSE, &ec);
+    }
+}
+
+
+/**
+ *  Stable sort with a user supplied comparator of type UComparator.
+ */
+void UVector::sortWithUComparator(UComparator *compare, const void *context, UErrorCode &ec) {
+    if (U_SUCCESS(ec)) {
+        uprv_sortArray(elements, count, sizeof(UElement),
+                       compare, context, TRUE, &ec);
+    }
+}
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/uvector.h b/thirdparty/icu4c/common/uvector.h
new file mode 100644
index 0000000000..a2bef923af
--- /dev/null
+++ b/thirdparty/icu4c/common/uvector.h
@@ -0,0 +1,415 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*   Date        Name        Description
+*   10/22/99    alan        Creation.  This is an internal header.
+*                           It should not be exported.
+**********************************************************************
+*/
+
+#ifndef UVECTOR_H
+#define UVECTOR_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "cmemory.h"
+#include "uarrsort.h"
+#include "uelement.h"
+
+U_NAMESPACE_BEGIN
+
+/**
+ * <p>Ultralightweight C++ implementation of a <tt>void*</tt> vector
+ * that is (mostly) compatible with java.util.Vector.
+ *
+ * <p>This is a very simple implementation, written to satisfy an
+ * immediate porting need.  As such, it is not completely fleshed out,
+ * and it aims for simplicity and conformity.  Nonetheless, it serves
+ * its purpose (porting code from java that uses java.util.Vector)
+ * well, and it could be easily made into a more robust vector class.
+ *
+ * <p><b>Design notes</b>
+ *
+ * <p>There is index bounds checking, but little is done about it.  If
+ * indices are out of bounds, either nothing happens, or zero is
+ * returned.  We <em>do</em> avoid indexing off into the weeds.
+ *
+ * <p>There is detection of out of memory, but the handling is very
+ * coarse-grained -- similar to UnicodeString's protocol, but even
+ * coarser.  The class contains <em>one static flag</em> that is set
+ * when any call to <tt>new</tt> returns zero.  This allows the caller
+ * to use several vectors and make just one check at the end to see if
+ * a memory failure occurred.  This is more efficient than making a
+ * check after each call on each vector when doing many operations on
+ * multiple vectors.  The single static flag works best when memory
+ * failures are infrequent, and when recovery options are limited or
+ * nonexistent.
+ *
+ * <p>Since we don't have garbage collection, UVector was given the
+ * option to <em>own</em>its contents.  To employ this, set a deleter
+ * function.  The deleter is called on a void* pointer when that
+ * pointer is released by the vector, either when the vector itself is
+ * destructed, or when a call to setElementAt() overwrites an element,
+ * or when a call to remove() or one of its variants explicitly
+ * removes an element.  If no deleter is set, or the deleter is set to
+ * zero, then it is assumed that the caller will delete elements as
+ * needed.
+ *
+ * <p>In order to implement methods such as contains() and indexOf(),
+ * UVector needs a way to compare objects for equality.  To do so, it
+ * uses a comparison function, or "comparer."  If the comparer is not
+ * set, or is set to zero, then all such methods will act as if the
+ * vector contains no element.  That is, indexOf() will always return
+ * -1, contains() will always return false, etc.
+ *
+ * <p><b>To do</b>
+ *
+ * <p>Improve the handling of index out of bounds errors.
+ *
+ * @author Alan Liu
+ */
+class U_COMMON_API UVector : public UObject {
+    // NOTE: UVector uses the UHashKey (union of void* and int32_t) as
+    // its basic storage type.  It uses UElementsAreEqual as its
+    // comparison function.  It uses UObjectDeleter as its deleter
+    // function.  These are named for hashtables, but used here as-is
+    // rather than duplicating the type.  This allows sharing of
+    // support functions.
+
+private:
+    int32_t count;
+
+    int32_t capacity;
+
+    UElement* elements;
+
+    UObjectDeleter *deleter;
+
+    UElementsAreEqual *comparer;
+
+public:
+    UVector(UErrorCode &status);
+
+    UVector(int32_t initialCapacity, UErrorCode &status);
+
+    UVector(UObjectDeleter *d, UElementsAreEqual *c, UErrorCode &status);
+
+    UVector(UObjectDeleter *d, UElementsAreEqual *c, int32_t initialCapacity, UErrorCode &status);
+
+    virtual ~UVector();
+
+    /**
+     * Assign this object to another (make this a copy of 'other').
+     * Use the 'assign' function to assign each element.
+     */
+    void assign(const UVector& other, UElementAssigner *assign, UErrorCode &ec);
+
+    /**
+     * Compare this vector with another.  They will be considered
+     * equal if they are of the same size and all elements are equal,
+     * as compared using this object's comparer.
+     */
+    UBool operator==(const UVector& other);
+
+    /**
+     * Equivalent to !operator==()
+     */
+    inline UBool operator!=(const UVector& other);
+
+    //------------------------------------------------------------
+    // java.util.Vector API
+    //------------------------------------------------------------
+
+    void addElement(void* obj, UErrorCode &status);
+
+    void addElement(int32_t elem, UErrorCode &status);
+
+    void setElementAt(void* obj, int32_t index);
+
+    void setElementAt(int32_t elem, int32_t index);
+
+    void insertElementAt(void* obj, int32_t index, UErrorCode &status);
+
+    void insertElementAt(int32_t elem, int32_t index, UErrorCode &status);
+    
+    void* elementAt(int32_t index) const;
+
+    int32_t elementAti(int32_t index) const;
+
+    UBool equals(const UVector &other) const;
+
+    inline void* firstElement(void) const;
+
+    inline void* lastElement(void) const;
+
+    inline int32_t lastElementi(void) const;
+
+    int32_t indexOf(void* obj, int32_t startIndex = 0) const;
+
+    int32_t indexOf(int32_t obj, int32_t startIndex = 0) const;
+
+    inline UBool contains(void* obj) const;
+
+    inline UBool contains(int32_t obj) const;
+
+    UBool containsAll(const UVector& other) const;
+
+    UBool removeAll(const UVector& other);
+
+    UBool retainAll(const UVector& other);
+
+    void removeElementAt(int32_t index);
+
+    UBool removeElement(void* obj);
+
+    void removeAllElements();
+
+    inline int32_t size(void) const;
+
+    inline UBool isEmpty(void) const;
+
+    UBool ensureCapacity(int32_t minimumCapacity, UErrorCode &status);
+
+    /**
+     * Change the size of this vector as follows: If newSize is
+     * smaller, then truncate the array, possibly deleting held
+     * elements for i >= newSize.  If newSize is larger, grow the
+     * array, filling in new slots with NULL.
+     */
+    void setSize(int32_t newSize, UErrorCode &status);
+
+    /**
+     * Fill in the given array with all elements of this vector.
+     */
+    void** toArray(void** result) const;
+
+    //------------------------------------------------------------
+    // New API
+    //------------------------------------------------------------
+
+    UObjectDeleter *setDeleter(UObjectDeleter *d);
+
+    UElementsAreEqual *setComparer(UElementsAreEqual *c);
+
+    inline void* operator[](int32_t index) const;
+
+    /**
+     * Removes the element at the given index from this vector and
+     * transfer ownership of it to the caller.  After this call, the
+     * caller owns the result and must delete it and the vector entry
+     * at 'index' is removed, shifting all subsequent entries back by
+     * one index and shortening the size of the vector by one.  If the
+     * index is out of range or if there is no item at the given index
+     * then 0 is returned and the vector is unchanged.
+     */
+    void* orphanElementAt(int32_t index);
+
+    /**
+     * Returns true if this vector contains none of the elements
+     * of the given vector.
+     * @param other vector to be checked for containment
+     * @return true if the test condition is met
+     */
+    UBool containsNone(const UVector& other) const;
+
+    /**
+     * Insert the given object into this vector at its sorted position
+     * as defined by 'compare'.  The current elements are assumed to
+     * be sorted already.
+     */
+    void sortedInsert(void* obj, UElementComparator *compare, UErrorCode& ec);
+
+    /**
+     * Insert the given integer into this vector at its sorted position
+     * as defined by 'compare'.  The current elements are assumed to
+     * be sorted already.
+     */
+    void sortedInsert(int32_t obj, UElementComparator *compare, UErrorCode& ec);
+
+    /**
+     * Sort the contents of the vector, assuming that the contents of the
+     * vector are of type int32_t.
+     */
+    void sorti(UErrorCode &ec);
+
+    /**
+      * Sort the contents of this vector, using a caller-supplied function
+      * to do the comparisons.  (It's confusing that
+      *  UVector's UElementComparator function is different from the
+      *  UComparator function type defined in uarrsort.h)
+      */
+    void sort(UElementComparator *compare, UErrorCode &ec);
+
+    /**
+     * Stable sort the contents of this vector using a caller-supplied function
+     * of type UComparator to do the comparison.  Provides more flexibility
+     * than UVector::sort() because an additional user parameter can be passed to
+     * the comparison function.
+     */
+    void sortWithUComparator(UComparator *compare, const void *context, UErrorCode &ec);
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    void _init(int32_t initialCapacity, UErrorCode &status);
+
+    int32_t indexOf(UElement key, int32_t startIndex = 0, int8_t hint = 0) const;
+
+    void sortedInsert(UElement e, UElementComparator *compare, UErrorCode& ec);
+
+    // Disallow
+    UVector(const UVector&);
+
+    // Disallow
+    UVector& operator=(const UVector&);
+
+};
+
+
+/**
+ * <p>Ultralightweight C++ implementation of a <tt>void*</tt> stack
+ * that is (mostly) compatible with java.util.Stack.  As in java, this
+ * is merely a paper thin layer around UVector.  See the UVector
+ * documentation for further information.
+ *
+ * <p><b>Design notes</b>
+ *
+ * <p>The element at index <tt>n-1</tt> is (of course) the top of the
+ * stack.
+ *
+ * <p>The poorly named <tt>empty()</tt> method doesn't empty the
+ * stack; it determines if the stack is empty.
+ *
+ * @author Alan Liu
+ */
+class U_COMMON_API UStack : public UVector {
+public:
+    UStack(UErrorCode &status);
+
+    UStack(int32_t initialCapacity, UErrorCode &status);
+
+    UStack(UObjectDeleter *d, UElementsAreEqual *c, UErrorCode &status);
+
+    UStack(UObjectDeleter *d, UElementsAreEqual *c, int32_t initialCapacity, UErrorCode &status);
+
+    virtual ~UStack();
+
+    // It's okay not to have a virtual destructor (in UVector)
+    // because UStack has no special cleanup to do.
+
+    inline UBool empty(void) const;
+
+    inline void* peek(void) const;
+
+    inline int32_t peeki(void) const;
+    
+    void* pop(void);
+    
+    int32_t popi(void);
+    
+    inline void* push(void* obj, UErrorCode &status);
+
+    inline int32_t push(int32_t i, UErrorCode &status);
+
+    /*
+    If the object o occurs as an item in this stack,
+    this method returns the 1-based distance from the top of the stack.
+    */
+    int32_t search(void* obj) const;
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    // Disallow
+    UStack(const UStack&);
+
+    // Disallow
+    UStack& operator=(const UStack&);
+};
+
+
+// UVector inlines
+
+inline int32_t UVector::size(void) const {
+    return count;
+}
+
+inline UBool UVector::isEmpty(void) const {
+    return count == 0;
+}
+
+inline UBool UVector::contains(void* obj) const {
+    return indexOf(obj) >= 0;
+}
+
+inline UBool UVector::contains(int32_t obj) const {
+    return indexOf(obj) >= 0;
+}
+
+inline void* UVector::firstElement(void) const {
+    return elementAt(0);
+}
+
+inline void* UVector::lastElement(void) const {
+    return elementAt(count-1);
+}
+
+inline int32_t UVector::lastElementi(void) const {
+    return elementAti(count-1);
+}
+
+inline void* UVector::operator[](int32_t index) const {
+    return elementAt(index);
+}
+
+inline UBool UVector::operator!=(const UVector& other) {
+    return !operator==(other);
+}
+
+// UStack inlines
+
+inline UBool UStack::empty(void) const {
+    return isEmpty();
+}
+
+inline void* UStack::peek(void) const {
+    return lastElement();
+}
+
+inline int32_t UStack::peeki(void) const {
+    return lastElementi();
+}
+
+inline void* UStack::push(void* obj, UErrorCode &status) {
+    addElement(obj, status);
+    return obj;
+}
+
+inline int32_t UStack::push(int32_t i, UErrorCode &status) {
+    addElement(i, status);
+    return i;
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/uvectr32.cpp b/thirdparty/icu4c/common/uvectr32.cpp
new file mode 100644
index 0000000000..d1ae659958
--- /dev/null
+++ b/thirdparty/icu4c/common/uvectr32.cpp
@@ -0,0 +1,335 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1999-2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*   Date        Name        Description
+*   10/22/99    alan        Creation.
+**********************************************************************
+*/
+
+#include "uvectr32.h"
+#include "cmemory.h"
+#include "putilimp.h"
+
+U_NAMESPACE_BEGIN
+
+#define DEFAULT_CAPACITY 8
+
+/*
+ * Constants for hinting whether a key is an integer
+ * or a pointer.  If a hint bit is zero, then the associated
+ * token is assumed to be an integer. This is needed for iSeries
+ */
+ 
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UVector32)
+
+UVector32::UVector32(UErrorCode &status) :
+    count(0),
+    capacity(0),
+    maxCapacity(0),
+    elements(NULL)
+{
+    _init(DEFAULT_CAPACITY, status);
+}
+
+UVector32::UVector32(int32_t initialCapacity, UErrorCode &status) :
+    count(0),
+    capacity(0),
+    maxCapacity(0),
+    elements(0)
+{
+    _init(initialCapacity, status);
+}
+
+
+
+void UVector32::_init(int32_t initialCapacity, UErrorCode &status) {
+    // Fix bogus initialCapacity values; avoid malloc(0)
+    if (initialCapacity < 1) {
+        initialCapacity = DEFAULT_CAPACITY;
+    }
+    if (maxCapacity>0 && maxCapacity<initialCapacity) {
+        initialCapacity = maxCapacity;
+    }
+    if (initialCapacity > (int32_t)(INT32_MAX / sizeof(int32_t))) {
+        initialCapacity = uprv_min(DEFAULT_CAPACITY, maxCapacity);
+    }
+    elements = (int32_t *)uprv_malloc(sizeof(int32_t)*initialCapacity);
+    if (elements == 0) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        capacity = initialCapacity;
+    }
+}
+
+UVector32::~UVector32() {
+    uprv_free(elements);
+    elements = 0;
+}
+
+/**
+ * Assign this object to another (make this a copy of 'other').
+ */
+void UVector32::assign(const UVector32& other, UErrorCode &ec) {
+    if (ensureCapacity(other.count, ec)) {
+        setSize(other.count);
+        for (int32_t i=0; i<other.count; ++i) {
+            elements[i] = other.elements[i];
+        }
+    }
+}
+
+
+UBool UVector32::operator==(const UVector32& other) {
+    int32_t i;
+    if (count != other.count) return FALSE;
+    for (i=0; i<count; ++i) {
+        if (elements[i] != other.elements[i]) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+
+void UVector32::setElementAt(int32_t elem, int32_t index) {
+    if (0 <= index && index < count) {
+        elements[index] = elem;
+    }
+    /* else index out of range */
+}
+
+void UVector32::insertElementAt(int32_t elem, int32_t index, UErrorCode &status) {
+    // must have 0 <= index <= count
+    if (0 <= index && index <= count && ensureCapacity(count + 1, status)) {
+        for (int32_t i=count; i>index; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[index] = elem;
+        ++count;
+    }
+    /* else index out of range */
+}
+
+UBool UVector32::containsAll(const UVector32& other) const {
+    for (int32_t i=0; i<other.size(); ++i) {
+        if (indexOf(other.elements[i]) < 0) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+UBool UVector32::containsNone(const UVector32& other) const {
+    for (int32_t i=0; i<other.size(); ++i) {
+        if (indexOf(other.elements[i]) >= 0) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+UBool UVector32::removeAll(const UVector32& other) {
+    UBool changed = FALSE;
+    for (int32_t i=0; i<other.size(); ++i) {
+        int32_t j = indexOf(other.elements[i]);
+        if (j >= 0) {
+            removeElementAt(j);
+            changed = TRUE;
+        }
+    }
+    return changed;
+}
+
+UBool UVector32::retainAll(const UVector32& other) {
+    UBool changed = FALSE;
+    for (int32_t j=size()-1; j>=0; --j) {
+        int32_t i = other.indexOf(elements[j]);
+        if (i < 0) {
+            removeElementAt(j);
+            changed = TRUE;
+        }
+    }
+    return changed;
+}
+
+void UVector32::removeElementAt(int32_t index) {
+    if (index >= 0) {
+        for (int32_t i=index; i<count-1; ++i) {
+            elements[i] = elements[i+1];
+        }
+        --count;
+    }
+}
+
+void UVector32::removeAllElements(void) {
+    count = 0;
+}
+
+UBool   UVector32::equals(const UVector32 &other) const {
+    int      i;
+
+    if (this->count != other.count) {
+        return FALSE;
+    }
+    for (i=0; i<count; i++) {
+        if (elements[i] != other.elements[i]) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+
+
+
+int32_t UVector32::indexOf(int32_t key, int32_t startIndex) const {
+    int32_t i;
+    for (i=startIndex; i<count; ++i) {
+        if (key == elements[i]) {
+            return i;
+        }
+    }
+    return -1;
+}
+
+
+UBool UVector32::expandCapacity(int32_t minimumCapacity, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return FALSE;
+    }
+    if (minimumCapacity < 0) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    if (capacity >= minimumCapacity) {
+        return TRUE;
+    }
+    if (maxCapacity>0 && minimumCapacity>maxCapacity) {
+        status = U_BUFFER_OVERFLOW_ERROR;
+        return FALSE;
+    }
+    if (capacity > (INT32_MAX - 1) / 2) {  // integer overflow check
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    int32_t newCap = capacity * 2;
+    if (newCap < minimumCapacity) {
+        newCap = minimumCapacity;
+    }
+    if (maxCapacity > 0 && newCap > maxCapacity) {
+        newCap = maxCapacity;
+    }
+    if (newCap > (int32_t)(INT32_MAX / sizeof(int32_t))) {  // integer overflow check
+        // We keep the original memory contents on bad minimumCapacity/maxCapacity.
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    int32_t* newElems = (int32_t *)uprv_realloc(elements, sizeof(int32_t)*newCap);
+    if (newElems == NULL) {
+        // We keep the original contents on the memory failure on realloc.
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    elements = newElems;
+    capacity = newCap;
+    return TRUE;
+}
+
+void UVector32::setMaxCapacity(int32_t limit) {
+    U_ASSERT(limit >= 0);
+    if (limit < 0) {
+        limit = 0;
+    }
+    if (limit > (int32_t)(INT32_MAX / sizeof(int32_t))) {  // integer overflow check for realloc
+        //  Something is very wrong, don't realloc, leave capacity and maxCapacity unchanged
+        return;
+    }
+    maxCapacity = limit;
+    if (capacity <= maxCapacity || maxCapacity == 0) {
+        // Current capacity is within the new limit.
+        return;
+    }
+    
+    // New maximum capacity is smaller than the current size.
+    // Realloc the storage to the new, smaller size.
+    int32_t* newElems = (int32_t *)uprv_realloc(elements, sizeof(int32_t)*maxCapacity);
+    if (newElems == NULL) {
+        // Realloc to smaller failed.
+        //   Just keep what we had.  No need to call it a failure.
+        return;
+    }
+    elements = newElems;
+    capacity = maxCapacity;
+    if (count > capacity) {
+        count = capacity;
+    }
+}
+
+/**
+ * Change the size of this vector as follows: If newSize is smaller,
+ * then truncate the array, possibly deleting held elements for i >=
+ * newSize.  If newSize is larger, grow the array, filling in new
+ * slots with NULL.
+ */
+void UVector32::setSize(int32_t newSize) {
+    int32_t i;
+    if (newSize < 0) {
+        return;
+    }
+    if (newSize > count) {
+        UErrorCode ec = U_ZERO_ERROR;
+        if (!ensureCapacity(newSize, ec)) {
+            return;
+        }
+        for (i=count; i<newSize; ++i) {
+            elements[i] = 0;
+        }
+    } 
+    count = newSize;
+}
+
+
+
+
+/**
+ * Insert the given integer into this vector at its sorted position
+ * as defined by 'compare'.  The current elements are assumed to
+ * be sorted already.
+ */
+void UVector32::sortedInsert(int32_t tok, UErrorCode& ec) {
+    // Perform a binary search for the location to insert tok at.  Tok
+    // will be inserted between two elements a and b such that a <=
+    // tok && tok < b, where there is a 'virtual' elements[-1] always
+    // less than tok and a 'virtual' elements[count] always greater
+    // than tok.
+    int32_t min = 0, max = count;
+    while (min != max) {
+        int32_t probe = (min + max) / 2;
+        //int8_t c = (*compare)(elements[probe], tok);
+        //if (c > 0) {
+        if (elements[probe] > tok) {
+            max = probe;
+        } else {
+            // assert(c <= 0);
+            min = probe + 1;
+        }
+    }
+    if (ensureCapacity(count + 1, ec)) {
+        for (int32_t i=count; i>min; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[min] = tok;
+        ++count;
+    }
+}
+
+
+
+
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/uvectr32.h b/thirdparty/icu4c/common/uvectr32.h
new file mode 100644
index 0000000000..0d81dfb5c1
--- /dev/null
+++ b/thirdparty/icu4c/common/uvectr32.h
@@ -0,0 +1,306 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+//
+//  UVector32 is a class implementing a vector of 32 bit integers.
+//            It is similar to UVector, but holds int32_t values rather than pointers.
+//            Most of the code is unchanged from UVector.
+//
+
+#ifndef UVECTOR32_H
+#define UVECTOR32_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "uhash.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+
+
+/**
+ * <p>Ultralightweight C++ implementation of a <tt>void*</tt> vector
+ * that is (mostly) compatible with java.util.Vector.
+ *
+ * <p>This is a very simple implementation, written to satisfy an
+ * immediate porting need.  As such, it is not completely fleshed out,
+ * and it aims for simplicity and conformity.  Nonetheless, it serves
+ * its purpose (porting code from java that uses java.util.Vector)
+ * well, and it could be easily made into a more robust vector class.
+ *
+ * <p><b>Design notes</b>
+ *
+ * <p>There is index bounds checking, but little is done about it.  If
+ * indices are out of bounds, either nothing happens, or zero is
+ * returned.  We <em>do</em> avoid indexing off into the weeds.
+ *
+ * <p>There is detection of out of memory, but the handling is very
+ * coarse-grained -- similar to UnicodeString's protocol, but even
+ * coarser.  The class contains <em>one static flag</em> that is set
+ * when any call to <tt>new</tt> returns zero.  This allows the caller
+ * to use several vectors and make just one check at the end to see if
+ * a memory failure occurred.  This is more efficient than making a
+ * check after each call on each vector when doing many operations on
+ * multiple vectors.  The single static flag works best when memory
+ * failures are infrequent, and when recovery options are limited or
+ * nonexistent.
+ *
+ * <p><b>To do</b>
+ *
+ * <p>Improve the handling of index out of bounds errors.
+ *
+ * @author Alan Liu
+ */
+class U_COMMON_API UVector32 : public UObject {
+private:
+    int32_t   count;
+
+    int32_t   capacity;
+    
+    int32_t   maxCapacity;   // Limit beyond which capacity is not permitted to grow.
+
+    int32_t*  elements;
+
+public:
+    UVector32(UErrorCode &status);
+
+    UVector32(int32_t initialCapacity, UErrorCode &status);
+
+    virtual ~UVector32();
+
+    /**
+     * Assign this object to another (make this a copy of 'other').
+     * Use the 'assign' function to assign each element.
+     */
+    void assign(const UVector32& other, UErrorCode &ec);
+
+    /**
+     * Compare this vector with another.  They will be considered
+     * equal if they are of the same size and all elements are equal,
+     * as compared using this object's comparer.
+     */
+    UBool operator==(const UVector32& other);
+
+    /**
+     * Equivalent to !operator==()
+     */
+    inline UBool operator!=(const UVector32& other);
+
+    //------------------------------------------------------------
+    // java.util.Vector API
+    //------------------------------------------------------------
+
+    inline void addElement(int32_t elem, UErrorCode &status);
+
+    void setElementAt(int32_t elem, int32_t index);
+
+    void insertElementAt(int32_t elem, int32_t index, UErrorCode &status);
+    
+    inline int32_t elementAti(int32_t index) const;
+
+    UBool equals(const UVector32 &other) const;
+
+    inline int32_t lastElementi(void) const;
+
+    int32_t indexOf(int32_t elem, int32_t startIndex = 0) const;
+
+    inline UBool contains(int32_t elem) const;
+
+    UBool containsAll(const UVector32& other) const;
+
+    UBool removeAll(const UVector32& other);
+
+    UBool retainAll(const UVector32& other);
+
+    void removeElementAt(int32_t index);
+
+    void removeAllElements();
+
+    inline int32_t size(void) const;
+
+    inline UBool isEmpty(void) const;
+
+    // Inline.  Use this one for speedy size check.
+    inline UBool ensureCapacity(int32_t minimumCapacity, UErrorCode &status);
+
+    // Out-of-line, handles actual growth.  Called by ensureCapacity() when necessary.
+    UBool expandCapacity(int32_t minimumCapacity, UErrorCode &status);
+
+    /**
+     * Change the size of this vector as follows: If newSize is
+     * smaller, then truncate the array, possibly deleting held
+     * elements for i >= newSize.  If newSize is larger, grow the
+     * array, filling in new slows with zero.
+     */
+    void setSize(int32_t newSize);
+
+    //------------------------------------------------------------
+    // New API
+    //------------------------------------------------------------
+
+    /**
+     * Returns true if this vector contains none of the elements
+     * of the given vector.
+     * @param other vector to be checked for containment
+     * @return true if the test condition is met
+     */
+    UBool containsNone(const UVector32& other) const;
+
+
+    /**
+     * Insert the given integer into this vector at its sorted position.
+     * The current elements are assumed to be sorted already.
+     */
+    void sortedInsert(int32_t elem, UErrorCode& ec);
+
+    /**
+     * Returns a pointer to the internal array holding the vector.
+     */
+    inline int32_t *getBuffer() const;
+
+    /**
+     * Set the maximum allowed buffer capacity for this vector/stack.
+     * Default with no limit set is unlimited, go until malloc() fails.
+     * A Limit of zero means unlimited capacity.
+     * Units are vector elements (32 bits each), not bytes.
+     */
+    void setMaxCapacity(int32_t limit);
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    void _init(int32_t initialCapacity, UErrorCode &status);
+
+    // Disallow
+    UVector32(const UVector32&);
+
+    // Disallow
+    UVector32& operator=(const UVector32&);
+
+
+    //  API Functions for Stack operations.
+    //  In the original UVector, these were in a separate derived class, UStack.
+    //  Here in UVector32, they are all together.
+public:
+    inline UBool empty(void) const;   // TODO:  redundant, same as empty().  Remove it?
+
+    inline int32_t peeki(void) const;
+    
+    inline int32_t popi(void);
+    
+    inline int32_t push(int32_t i, UErrorCode &status);
+
+    inline int32_t *reserveBlock(int32_t size, UErrorCode &status);
+    inline int32_t *popFrame(int32_t size);
+};
+
+
+// UVector32 inlines
+
+inline UBool UVector32::ensureCapacity(int32_t minimumCapacity, UErrorCode &status) {
+    if ((minimumCapacity >= 0) && (capacity >= minimumCapacity)) {
+        return true;
+    } else {
+        return expandCapacity(minimumCapacity, status);
+    }
+}
+
+inline int32_t UVector32::elementAti(int32_t index) const {
+    return (index >= 0 && count > 0 && count - index > 0) ? elements[index] : 0;
+}
+
+
+inline void UVector32::addElement(int32_t elem, UErrorCode &status) {
+    if (ensureCapacity(count + 1, status)) {
+        elements[count] = elem;
+        count++;
+    }
+}
+
+inline int32_t *UVector32::reserveBlock(int32_t size, UErrorCode &status) {
+    if (ensureCapacity(count+size, status) == false) {
+        return NULL;
+    }
+    int32_t  *rp = elements+count;
+    count += size;
+    return rp;
+}
+
+inline int32_t *UVector32::popFrame(int32_t size) {
+    U_ASSERT(count >= size);
+    count -= size;
+    if (count < 0) {
+        count = 0;
+    }
+    return elements+count-size;
+}
+
+
+
+inline int32_t UVector32::size(void) const {
+    return count;
+}
+
+inline UBool UVector32::isEmpty(void) const {
+    return count == 0;
+}
+
+inline UBool UVector32::contains(int32_t obj) const {
+    return indexOf(obj) >= 0;
+}
+
+inline int32_t UVector32::lastElementi(void) const {
+    return elementAti(count-1);
+}
+
+inline UBool UVector32::operator!=(const UVector32& other) {
+    return !operator==(other);
+}
+
+inline int32_t *UVector32::getBuffer() const {
+    return elements;
+}
+
+
+// UStack inlines
+
+inline UBool UVector32::empty(void) const {
+    return isEmpty();
+}
+
+inline int32_t UVector32::peeki(void) const {
+    return lastElementi();
+}
+
+inline int32_t UVector32::push(int32_t i, UErrorCode &status) {
+    addElement(i, status);
+    return i;
+}
+
+inline int32_t UVector32::popi(void) {
+    int32_t result = 0;
+    if (count > 0) {
+        count--;
+        result = elements[count];
+    }
+    return result;
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/uvectr64.cpp b/thirdparty/icu4c/common/uvectr64.cpp
new file mode 100644
index 0000000000..081565959c
--- /dev/null
+++ b/thirdparty/icu4c/common/uvectr64.cpp
@@ -0,0 +1,214 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+******************************************************************************
+* Copyright (C) 1999-2015, International Business Machines Corporation and
+* others. All Rights Reserved.
+******************************************************************************
+*/
+
+#include "uvectr64.h"
+#include "cmemory.h"
+#include "putilimp.h"
+
+U_NAMESPACE_BEGIN
+
+#define DEFAULT_CAPACITY 8
+
+/*
+ * Constants for hinting whether a key is an integer
+ * or a pointer.  If a hint bit is zero, then the associated
+ * token is assumed to be an integer. This is needed for iSeries
+ */
+ 
+UOBJECT_DEFINE_RTTI_IMPLEMENTATION(UVector64)
+
+UVector64::UVector64(UErrorCode &status) :
+    count(0),
+    capacity(0),
+    maxCapacity(0),
+    elements(NULL)
+{
+    _init(DEFAULT_CAPACITY, status);
+}
+
+UVector64::UVector64(int32_t initialCapacity, UErrorCode &status) :
+    count(0),
+    capacity(0),
+    maxCapacity(0),
+    elements(0)
+{
+    _init(initialCapacity, status);
+}
+
+
+
+void UVector64::_init(int32_t initialCapacity, UErrorCode &status) {
+    // Fix bogus initialCapacity values; avoid malloc(0)
+    if (initialCapacity < 1) {
+        initialCapacity = DEFAULT_CAPACITY;
+    }
+    if (maxCapacity>0 && maxCapacity<initialCapacity) {
+        initialCapacity = maxCapacity;
+    }
+    if (initialCapacity > (int32_t)(INT32_MAX / sizeof(int64_t))) {
+        initialCapacity = uprv_min(DEFAULT_CAPACITY, maxCapacity);
+    }
+    elements = (int64_t *)uprv_malloc(sizeof(int64_t)*initialCapacity);
+    if (elements == 0) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+    } else {
+        capacity = initialCapacity;
+    }
+}
+
+UVector64::~UVector64() {
+    uprv_free(elements);
+    elements = 0;
+}
+
+/**
+ * Assign this object to another (make this a copy of 'other').
+ */
+void UVector64::assign(const UVector64& other, UErrorCode &ec) {
+    if (ensureCapacity(other.count, ec)) {
+        setSize(other.count);
+        for (int32_t i=0; i<other.count; ++i) {
+            elements[i] = other.elements[i];
+        }
+    }
+}
+
+
+UBool UVector64::operator==(const UVector64& other) {
+    int32_t i;
+    if (count != other.count) return FALSE;
+    for (i=0; i<count; ++i) {
+        if (elements[i] != other.elements[i]) {
+            return FALSE;
+        }
+    }
+    return TRUE;
+}
+
+
+void UVector64::setElementAt(int64_t elem, int32_t index) {
+    if (0 <= index && index < count) {
+        elements[index] = elem;
+    }
+    /* else index out of range */
+}
+
+void UVector64::insertElementAt(int64_t elem, int32_t index, UErrorCode &status) {
+    // must have 0 <= index <= count
+    if (0 <= index && index <= count && ensureCapacity(count + 1, status)) {
+        for (int32_t i=count; i>index; --i) {
+            elements[i] = elements[i-1];
+        }
+        elements[index] = elem;
+        ++count;
+    }
+    /* else index out of range */
+}
+
+void UVector64::removeAllElements(void) {
+    count = 0;
+}
+
+UBool UVector64::expandCapacity(int32_t minimumCapacity, UErrorCode &status) {
+    if (U_FAILURE(status)) {
+        return FALSE;
+    }
+    if (minimumCapacity < 0) {
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    if (capacity >= minimumCapacity) {
+        return TRUE;
+    }
+    if (maxCapacity>0 && minimumCapacity>maxCapacity) {
+        status = U_BUFFER_OVERFLOW_ERROR;
+        return FALSE;
+    }
+    if (capacity > (INT32_MAX - 1) / 2) {  // integer overflow check
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    int32_t newCap = capacity * 2;
+    if (newCap < minimumCapacity) {
+        newCap = minimumCapacity;
+    }
+    if (maxCapacity > 0 && newCap > maxCapacity) {
+        newCap = maxCapacity;
+    }
+    if (newCap > (int32_t)(INT32_MAX / sizeof(int64_t))) {  // integer overflow check
+        // We keep the original memory contents on bad minimumCapacity/maxCapacity.
+        status = U_ILLEGAL_ARGUMENT_ERROR;
+        return FALSE;
+    }
+    int64_t* newElems = (int64_t *)uprv_realloc(elements, sizeof(int64_t)*newCap);
+    if (newElems == NULL) {
+        // We keep the original contents on the memory failure on realloc.
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return FALSE;
+    }
+    elements = newElems;
+    capacity = newCap;
+    return TRUE;
+}
+
+void UVector64::setMaxCapacity(int32_t limit) {
+    U_ASSERT(limit >= 0);
+    if (limit < 0) {
+        limit = 0;
+    }
+    if (limit > (int32_t)(INT32_MAX / sizeof(int64_t))) {  // integer overflow check for realloc
+        //  Something is very wrong, don't realloc, leave capacity and maxCapacity unchanged
+        return;
+    }
+    maxCapacity = limit;
+    if (capacity <= maxCapacity || maxCapacity == 0) {
+        // Current capacity is within the new limit.
+        return;
+    }
+    
+    // New maximum capacity is smaller than the current size.
+    // Realloc the storage to the new, smaller size.
+    int64_t* newElems = (int64_t *)uprv_realloc(elements, sizeof(int64_t)*maxCapacity);
+    if (newElems == NULL) {
+        // Realloc to smaller failed.
+        //   Just keep what we had.  No need to call it a failure.
+        return;
+    }
+    elements = newElems;
+    capacity = maxCapacity;
+    if (count > capacity) {
+        count = capacity;
+    }
+}
+
+/**
+ * Change the size of this vector as follows: If newSize is smaller,
+ * then truncate the array, possibly deleting held elements for i >=
+ * newSize.  If newSize is larger, grow the array, filling in new
+ * slots with NULL.
+ */
+void UVector64::setSize(int32_t newSize) {
+    int32_t i;
+    if (newSize < 0) {
+        return;
+    }
+    if (newSize > count) {
+        UErrorCode ec = U_ZERO_ERROR;
+        if (!ensureCapacity(newSize, ec)) {
+            return;
+        }
+        for (i=count; i<newSize; ++i) {
+            elements[i] = 0;
+        }
+    } 
+    count = newSize;
+}
+
+U_NAMESPACE_END
+
diff --git a/thirdparty/icu4c/common/uvectr64.h b/thirdparty/icu4c/common/uvectr64.h
new file mode 100644
index 0000000000..15c9b3f830
--- /dev/null
+++ b/thirdparty/icu4c/common/uvectr64.h
@@ -0,0 +1,279 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+**********************************************************************
+*   Copyright (C) 1999-2014, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+**********************************************************************
+*/
+
+//
+//  UVector64 is a class implementing a vector of 64 bit integers.
+//            It is similar to UVector32, but holds int64_t values rather than int32_t.
+//            Most of the code is unchanged from UVector.
+//
+
+#ifndef UVECTOR64_H
+#define UVECTOR64_H
+
+#include "unicode/utypes.h"
+#include "unicode/uobject.h"
+#include "uhash.h"
+#include "uassert.h"
+
+U_NAMESPACE_BEGIN
+
+
+
+/**
+ * <p>Ultralightweight C++ implementation of an <tt>int64_t</tt> vector
+ * that has a subset of methods from UVector32
+ *
+ * <p>This is a very simple implementation, written to satisfy an
+ * immediate porting need.  As such, it is not completely fleshed out,
+ * and it aims for simplicity and conformity.  Nonetheless, it serves
+ * its purpose (porting code from java that uses java.util.Vector)
+ * well, and it could be easily made into a more robust vector class.
+ *
+ * <p><b>Design notes</b>
+ *
+ * <p>There is index bounds checking, but little is done about it.  If
+ * indices are out of bounds, either nothing happens, or zero is
+ * returned.  We <em>do</em> avoid indexing off into the weeds.
+ *
+ * <p>There is detection of out of memory, but the handling is very
+ * coarse-grained -- similar to UnicodeString's protocol, but even
+ * coarser.  The class contains <em>one static flag</em> that is set
+ * when any call to <tt>new</tt> returns zero.  This allows the caller
+ * to use several vectors and make just one check at the end to see if
+ * a memory failure occurred.  This is more efficient than making a
+ * check after each call on each vector when doing many operations on
+ * multiple vectors.  The single static flag works best when memory
+ * failures are infrequent, and when recovery options are limited or
+ * nonexistent.
+ *
+ * <p><b>To do</b>
+ *
+ * <p>Improve the handling of index out of bounds errors.
+ *
+ */
+class U_COMMON_API UVector64 : public UObject {
+private:
+    int32_t   count;
+
+    int32_t   capacity;
+    
+    int32_t   maxCapacity;   // Limit beyond which capacity is not permitted to grow.
+
+    int64_t*  elements;
+
+public:
+    UVector64(UErrorCode &status);
+
+    UVector64(int32_t initialCapacity, UErrorCode &status);
+
+    virtual ~UVector64();
+
+    /**
+     * Assign this object to another (make this a copy of 'other').
+     * Use the 'assign' function to assign each element.
+     */
+    void assign(const UVector64& other, UErrorCode &ec);
+
+    /**
+     * Compare this vector with another.  They will be considered
+     * equal if they are of the same size and all elements are equal,
+     * as compared using this object's comparer.
+     */
+    UBool operator==(const UVector64& other);
+
+    /**
+     * Equivalent to !operator==()
+     */
+    inline UBool operator!=(const UVector64& other);
+
+    //------------------------------------------------------------
+    // subset of java.util.Vector API
+    //------------------------------------------------------------
+
+    inline void addElement(int64_t elem, UErrorCode &status);
+
+    void setElementAt(int64_t elem, int32_t index);
+
+    void insertElementAt(int64_t elem, int32_t index, UErrorCode &status);
+    
+    inline int64_t elementAti(int32_t index) const;
+
+    //UBool equals(const UVector64 &other) const;
+
+    inline int64_t lastElementi(void) const;
+
+    //int32_t indexOf(int64_t elem, int32_t startIndex = 0) const;
+
+    //UBool contains(int64_t elem) const;
+
+    //UBool containsAll(const UVector64& other) const;
+
+    //UBool removeAll(const UVector64& other);
+
+    //UBool retainAll(const UVector64& other);
+
+    //void removeElementAt(int32_t index);
+
+    void removeAllElements();
+
+    inline int32_t size(void) const;
+
+    inline UBool isEmpty(void) const { return count == 0; }
+
+    // Inline.  Use this one for speedy size check.
+    inline UBool ensureCapacity(int32_t minimumCapacity, UErrorCode &status);
+
+    // Out-of-line, handles actual growth.  Called by ensureCapacity() when necessary.
+    UBool expandCapacity(int32_t minimumCapacity, UErrorCode &status);
+
+    /**
+     * Change the size of this vector as follows: If newSize is
+     * smaller, then truncate the array, possibly deleting held
+     * elements for i >= newSize.  If newSize is larger, grow the
+     * array, filling in new slows with zero.
+     */
+    void setSize(int32_t newSize);
+
+    //------------------------------------------------------------
+    // New API
+    //------------------------------------------------------------
+
+    //UBool containsNone(const UVector64& other) const;
+
+
+    //void sortedInsert(int64_t elem, UErrorCode& ec);
+
+    /**
+     * Returns a pointer to the internal array holding the vector.
+     */
+    inline int64_t *getBuffer() const;
+
+    /**
+     * Set the maximum allowed buffer capacity for this vector/stack.
+     * Default with no limit set is unlimited, go until malloc() fails.
+     * A Limit of zero means unlimited capacity.
+     * Units are vector elements (64 bits each), not bytes.
+     */
+    void setMaxCapacity(int32_t limit);
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for this class.
+     */
+    static UClassID U_EXPORT2 getStaticClassID();
+
+    /**
+     * ICU "poor man's RTTI", returns a UClassID for the actual class.
+     */
+    virtual UClassID getDynamicClassID() const;
+
+private:
+    void _init(int32_t initialCapacity, UErrorCode &status);
+
+    // Disallow
+    UVector64(const UVector64&);
+
+    // Disallow
+    UVector64& operator=(const UVector64&);
+
+
+    //  API Functions for Stack operations.
+    //  In the original UVector, these were in a separate derived class, UStack.
+    //  Here in UVector64, they are all together.
+public:
+    //UBool empty(void) const;   // TODO:  redundant, same as empty().  Remove it?
+
+    //int64_t peeki(void) const;
+    
+    inline int64_t popi(void);
+    
+    inline int64_t push(int64_t i, UErrorCode &status);
+
+    inline int64_t *reserveBlock(int32_t size, UErrorCode &status);
+    inline int64_t *popFrame(int32_t size);
+};
+
+
+// UVector64 inlines
+
+inline UBool UVector64::ensureCapacity(int32_t minimumCapacity, UErrorCode &status) {
+    if ((minimumCapacity >= 0) && (capacity >= minimumCapacity)) {
+        return true;
+    } else {
+        return expandCapacity(minimumCapacity, status);
+    }
+}
+
+inline int64_t UVector64::elementAti(int32_t index) const {
+    return (0 <= index && index < count) ? elements[index] : 0;
+}
+
+
+inline void UVector64::addElement(int64_t elem, UErrorCode &status) {
+    if (ensureCapacity(count + 1, status)) {
+        elements[count] = elem;
+        count++;
+    }
+}
+
+inline int64_t *UVector64::reserveBlock(int32_t size, UErrorCode &status) {
+    if (ensureCapacity(count+size, status) == false) {
+        return NULL;
+    }
+    int64_t  *rp = elements+count;
+    count += size;
+    return rp;
+}
+
+inline int64_t *UVector64::popFrame(int32_t size) {
+    U_ASSERT(count >= size);
+    count -= size;
+    if (count < 0) {
+        count = 0;
+    }
+    return elements+count-size;
+}
+
+
+
+inline int32_t UVector64::size(void) const {
+    return count;
+}
+
+inline int64_t UVector64::lastElementi(void) const {
+    return elementAti(count-1);
+}
+
+inline UBool UVector64::operator!=(const UVector64& other) {
+    return !operator==(other);
+}
+
+inline int64_t *UVector64::getBuffer() const {
+    return elements;
+}
+
+
+// UStack inlines
+
+inline int64_t UVector64::push(int64_t i, UErrorCode &status) {
+    addElement(i, status);
+    return i;
+}
+
+inline int64_t UVector64::popi(void) {
+    int64_t result = 0;
+    if (count > 0) {
+        count--;
+        result = elements[count];
+    }
+    return result;
+}
+
+U_NAMESPACE_END
+
+#endif
diff --git a/thirdparty/icu4c/common/wintz.cpp b/thirdparty/icu4c/common/wintz.cpp
new file mode 100644
index 0000000000..3730232286
--- /dev/null
+++ b/thirdparty/icu4c/common/wintz.cpp
@@ -0,0 +1,147 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 2005-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*
+* File WINTZ.CPP
+*
+********************************************************************************
+*/
+
+#include "unicode/utypes.h"
+
+#if U_PLATFORM_USES_ONLY_WIN32_API
+
+#include "wintz.h"
+#include "charstr.h"
+#include "cmemory.h"
+#include "cstring.h"
+
+#include "unicode/ures.h"
+#include "unicode/unistr.h"
+#include "uresimp.h"
+
+#ifndef WIN32_LEAN_AND_MEAN
+#   define WIN32_LEAN_AND_MEAN
+#endif
+#   define VC_EXTRALEAN
+#   define NOUSER
+#   define NOSERVICE
+#   define NOIME
+#   define NOMCX
+#include <windows.h>
+
+U_NAMESPACE_BEGIN
+
+/**
+* Main Windows time zone detection function.
+* Returns the Windows time zone converted to an ICU time zone as a heap-allocated buffer, or nullptr upon failure.
+*
+* Note: We use the Win32 API GetDynamicTimeZoneInformation (available since Vista+) to get the current time zone info.
+* This API returns a non-localized time zone name, which is mapped to an ICU time zone ID (~ Olsen ID).
+*/
+U_CAPI const char* U_EXPORT2
+uprv_detectWindowsTimeZone()
+{
+    // Obtain the DYNAMIC_TIME_ZONE_INFORMATION info to get the non-localized time zone name.
+    DYNAMIC_TIME_ZONE_INFORMATION dynamicTZI;
+    uprv_memset(&dynamicTZI, 0, sizeof(dynamicTZI));
+    SYSTEMTIME systemTimeAllZero;
+    uprv_memset(&systemTimeAllZero, 0, sizeof(systemTimeAllZero));
+
+    if (GetDynamicTimeZoneInformation(&dynamicTZI) == TIME_ZONE_ID_INVALID) {
+        return nullptr;
+    }
+
+    // If the DST setting has been turned off in the Control Panel, then return "Etc/GMT<offset>".
+    //
+    // Note: This logic is based on how the Control Panel itself determines if DST is 'off' on Windows.
+    // The code is somewhat convoluted; in a sort of pseudo-code it looks like this:
+    // 
+    //   IF (GetDynamicTimeZoneInformation != TIME_ZONE_ID_INVALID) && (DynamicDaylightTimeDisabled != 0) &&
+    //      (StandardDate == DaylightDate) &&
+    //      (
+    //       (TimeZoneKeyName != Empty && StandardDate == 0) ||
+    //       (TimeZoneKeyName == Empty && StandardDate != 0)
+    //      )
+    //   THEN
+    //     DST setting is "Disabled".
+    //
+    if (dynamicTZI.DynamicDaylightTimeDisabled != 0 &&
+        uprv_memcmp(&dynamicTZI.StandardDate, &dynamicTZI.DaylightDate, sizeof(dynamicTZI.StandardDate)) == 0 &&
+        ((dynamicTZI.TimeZoneKeyName[0] != L'\0' && uprv_memcmp(&dynamicTZI.StandardDate, &systemTimeAllZero, sizeof(systemTimeAllZero)) == 0) ||
+         (dynamicTZI.TimeZoneKeyName[0] == L'\0' && uprv_memcmp(&dynamicTZI.StandardDate, &systemTimeAllZero, sizeof(systemTimeAllZero)) != 0)))
+    {
+        LONG utcOffsetMins = dynamicTZI.Bias;
+        if (utcOffsetMins == 0) {
+            return uprv_strdup("Etc/UTC");
+        }
+
+        // No way to support when DST is turned off and the offset in minutes is not a multiple of 60.
+        if (utcOffsetMins % 60 == 0) {
+            char gmtOffsetTz[11] = {}; // "Etc/GMT+dd" is 11-char long with a terminal null.
+            // Note '-' before 'utcOffsetMin'. The timezone ID's sign convention
+            // is that a timezone ahead of UTC is Etc/GMT-<offset> and a timezone
+            // behind UTC is Etc/GMT+<offset>.
+            int ret = snprintf(gmtOffsetTz, UPRV_LENGTHOF(gmtOffsetTz), "Etc/GMT%+d", -utcOffsetMins / 60);
+            if (ret > 0 && ret < UPRV_LENGTHOF(gmtOffsetTz)) {
+                return uprv_strdup(gmtOffsetTz);
+            }
+        }
+    }
+
+    // If DST is NOT disabled, but we have an empty TimeZoneKeyName, then it is unclear
+    // what we should do as this should not happen.
+    if (dynamicTZI.TimeZoneKeyName[0] == 0) {
+        return nullptr;
+    }
+
+    CharString winTZ;
+    UErrorCode status = U_ZERO_ERROR;
+    winTZ.appendInvariantChars(UnicodeString(TRUE, dynamicTZI.TimeZoneKeyName, -1), status);
+
+    // Map Windows Timezone name (non-localized) to ICU timezone ID (~ Olson timezone id).
+    StackUResourceBundle winTZBundle;
+    ures_openDirectFillIn(winTZBundle.getAlias(), nullptr, "windowsZones", &status);
+    ures_getByKey(winTZBundle.getAlias(), "mapTimezones", winTZBundle.getAlias(), &status);
+    ures_getByKey(winTZBundle.getAlias(), winTZ.data(), winTZBundle.getAlias(), &status);
+
+    if (U_FAILURE(status)) {
+        return nullptr;
+    }
+    
+    // Note: Since the ISO 3166 country/region codes are all invariant ASCII chars, we can
+    // directly downcast from wchar_t to do the conversion.
+    // We could call the A version of the GetGeoInfo API, but that would be slightly slower than calling the W API,
+    // as the A version of the API will end up calling MultiByteToWideChar anyways internally.
+    wchar_t regionCodeW[3] = {};
+    char regionCode[3] = {}; // 2 letter ISO 3166 country/region code made entirely of invariant chars.
+    int geoId = GetUserGeoID(GEOCLASS_NATION);
+    int regionCodeLen = GetGeoInfoW(geoId, GEO_ISO2, regionCodeW, UPRV_LENGTHOF(regionCodeW), 0);
+
+    const UChar *icuTZ16 = nullptr;
+    int32_t tzLen;
+
+    if (regionCodeLen != 0) {
+        for (int i = 0; i < UPRV_LENGTHOF(regionCodeW); i++) {
+            regionCode[i] = static_cast<char>(regionCodeW[i]);
+        }
+        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), regionCode, &tzLen, &status);
+    }
+    if (regionCodeLen == 0 || U_FAILURE(status)) {
+        // fallback to default "001" (world)
+        status = U_ZERO_ERROR;
+        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), "001", &tzLen, &status);
+    }
+
+    // Note: cloneData returns nullptr if the status is a failure, so this
+    // will return nullptr if the above look-up fails.
+    CharString icuTZStr;
+    return icuTZStr.appendInvariantChars(icuTZ16, tzLen, status).cloneData(status);
+}
+
+U_NAMESPACE_END
+#endif /* U_PLATFORM_USES_ONLY_WIN32_API  */
diff --git a/thirdparty/icu4c/common/wintz.h b/thirdparty/icu4c/common/wintz.h
new file mode 100644
index 0000000000..ce9c1e9019
--- /dev/null
+++ b/thirdparty/icu4c/common/wintz.h
@@ -0,0 +1,36 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+********************************************************************************
+*   Copyright (C) 2005-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+********************************************************************************
+*
+* File WINTZ.H
+*
+********************************************************************************
+*/
+
+#ifndef __WINTZ
+#define __WINTZ
+
+#include "unicode/utypes.h"
+
+#if U_PLATFORM_USES_ONLY_WIN32_API
+
+/**
+ * \file 
+ * \brief C API: Utilities for dealing w/ Windows time zones.
+ */
+
+U_CDECL_BEGIN
+/* Forward declarations for Windows types... */
+typedef struct _TIME_ZONE_INFORMATION TIME_ZONE_INFORMATION;
+U_CDECL_END
+
+U_CAPI const char* U_EXPORT2
+uprv_detectWindowsTimeZone();
+
+#endif /* U_PLATFORM_USES_ONLY_WIN32_API  */
+
+#endif /* __WINTZ */
diff --git a/thirdparty/icu4c/godot_data.json b/thirdparty/icu4c/godot_data.json
new file mode 100644
index 0000000000..16cfcd2651
--- /dev/null
+++ b/thirdparty/icu4c/godot_data.json
@@ -0,0 +1,9 @@
+{
+    strategy: "additive"
+    featureFilters: {
+        brkitr_rules: include
+        brkitr_dictionaries: include
+        brkitr_tree: include
+        misc: include
+    }
+}
\ No newline at end of file
diff --git a/thirdparty/icu4c/icudt68l.dat b/thirdparty/icu4c/icudt68l.dat
new file mode 100644
index 0000000000..548c1a5a72
--- /dev/null
+++ b/thirdparty/icu4c/icudt68l.dat